AN OVERVIEW OF Na SO AND/OR V O INDUCED HOT CORROSION OF Fe- AND Ni-BASED SUPERALLOYS · ·...

27An overview of Na2SO4 and/or V2O5 induced hot corrosion of Fe- and Ni-based superalloys

© 2007 Advanced Study Center Co. Ltd.

Rev.Adv.Mater.Sci. 16(2007) 27-50

Corresponding author: Harpreet Singh, e-mail: [email protected]

AN OVERVIEW OF Na2SO4 AND/OR V2O5 INDUCED HOTCORROSION OF Fe- AND Ni-BASED SUPERALLOYS

Harpreet Singh1, Devendra Puri2 and Satya Prakash2

1Mechanical Engineering Department, BBSB Engineering College, Fatehgarh Sahib-140407, India2Metallurgical & Materials Engineering Department, Indian Institute of Technology Roorkee, ROORKEE-247 667,

India

Received: May 17, 2005

Abstract. Metals and alloys sometimes experience accelerated oxidation when their surfaces arecovered with a thin film of fused salt in an oxidizing gas atmosphere at elevated temperatures. Thisis known as hot corrosion where a porous non-protective oxide scale is formed at the surfaces andsulphides in the substrate. Hot corrosion has been identified as a serious problem in high tem-perature applications such as in boilers, gas turbines, waste incinerations, diesel engines, coalgasification plants, chemical plants and other energy generation systems. It is basically inducedby the impurities such as Na, V, S, K, and Cl, which are present in the coal or in fuel oil used forcombustion in the abovementioned applications. The use of Ni-, Fe-, and Co- based superalloys inhigh temperature applications such as gas turbines, boilers, etc. is well known, and many moreapplications are still to be explored. Although the superalloys have adequate mechanical strengthfor such high temperature applications, they are prone to degradation by hot corrosion/high tem-perature oxidation during long term exposures. Therefore, the superalloys need to be protected,however the protection system must be practical, reliable, and economically viable. In this paperthe hot corrosion phenomenon has been described in detail with a special reference to Fe- and Ni-based superalloys with the help of a survey of the available literature related to the hot corrosion ofthe superalloys, in the environments constituting mainly of Na2SO4 and/or V2O5.

1. INTRODUCTIONMetals and alloys sometimes experience acceler-ated oxidation when their surfaces are covered witha thin film of fused salt in an oxidizing gas atmo-sphere at elevated temperatures. This is knownas high temperature or ‘hot’ corrosion where a po-rous nonprotective oxide scale is formed at thesurfaces and sulphides in the substrate and themechanism of attack does not involve aqueouselectrolytes, (Rapp and Zhang, [1]).

High temperature corrosion was first recognizedas a serious problem in 1940s in connection withthe degradation of fireside boiler tubes in coal-firedsteam generating plants. Since that time, the prob-lem has been observed in boilers, internal com-bustion engines, gas turbines, fluidized bed com-

bustion and industrial waste incinerators. Turbinemanufacturers and users became aware of the hotcorrosion in the late 1960’s when serious corro-sive attack occurred in the engines of helicoptersand rescue planes in service over and near seawater during the Vietnam conflict, (Rapp, [2,3]).

Hot corrosion is basically the result of attack byfuel and/or ash compounds of Na, V, S, and Cl thatare present in the coal or in fuel oil used for com-bustion in the mentioned applications. In some situ-ations, these impurities may be ingested from theservice environment, for instance in the case ofNaCl contamination of marine atmospheres. Thereis a general agreement that condensed alkali metalsalts including (notably) Na2SO4, are a prerequi-site to hot corrosion, (Beltran and Shores, [4]). Apart

28 Harpreet Singh, Devendra Puri and Satya Prakash

from Na2SO4, V2O5 is a commonly-found high-tem-perature corrosion promotor in these applications.Due to high cost of removing these impurities, theuse of these low grade fuels is usually justified.Sodium vanadyl vanadate (Na2O.V2O4

.5V2O5),which melts at the relatively low temperature of 550°C, is found to be the most common salt depositon boiler superheaters, (Barbooti et al., [5]). Simi-larly, the predominant species in the salt depositsforming on gas turbine surfaces are also expectedto be Na2SO4, V2O5, and Na2V2O6 (Luthra andSpacil, [6]).

Due to the depletion of high-grade fuels and foreconomic reasons, use of residual fuel oil in theseenergy generation systems is well known. Residualfuel oil contains sodium, vanadium, and sulphuras impurities, as well as salt (NaCl) contaminationfrom entrained brine. Sodium and sulphur formNa2SO4 (melting point 884 °C) by reactions in thecombustion system. During combustion of the fuel,vanadium reacts with oxygen to form an oxide V2O5(melting point 670 °C). Thus V2O5 is a liquid at gasturbine operating temperatures. These compounds(known as ash) deposit on the surface of the ma-terials and induce accelerated oxidation (hot cor-rosion) in energy generation systems. When con-sidering coal-gasification processes, hot corrosionis expected to be a problem because the gas envi-ronment generally may contain sulphur and chlo-ride compounds and low oxygen activities, and thefeedstock also contains substantial amount of salts,(Natesan, [7]).

The operating temperatures in the case of gasturbines are usually relatively high and are expectedto increase further with the advances in materialsdevelopment and cooling schemes for the new-generation gas turbine engines. The combinationof such high temperatures with an aircraft environ-ment that contains contaminants such as sodium,sulphur, vanadium, and various halides requiresspecial attention to the phenomena of hot corro-sion, (Eliaz et al., [8]).

Power plants are one of the major industriessuffering from the severe corrosion problems re-sulting in the substantial losses. For instance, thesteam temperature of boilers is limited by corro-sion and creep resistance of boiler components,which affects the thermal efficiency of the boilers.Consequently, the thermal efficiency is reduced andhence the rate of electricity production is reduced(Uusitalo et al., [9]). According to a survey [10] con-ducted over a period of 12 years, encompassing413 investigations, overheating was listed as the

cause of 201 failures or 48.7% of those investi-gated. Fatigue and corrosion fatigue were listedas the next most common causes of failure ac-counting for a total of 89 failures or 21.5%. Corro-sion, stress corrosion and hydrogen embrittlementcaused a total of 68 failures or 16.5%. In anothercase study of a coal fired boiler of a power plant innorth western region of India, Prakash et al., [11],reported that out of 89 failures occurring in oneyear, 50 failures were found to be due to hot corro-sion and erosion by ash.

A superalloy is an alloy developed for elevated-temperature application, usually based on groupVIIIA elements, where relatively severe mechani-cal stress is encountered, and where high surfacestability is frequently required. Iron and nickel basesuperalloys are the commercial alloys commonlyused for the high temperature applications e.g.manufacture of components used in aggressiveenvironments of gas turbines, steam boilers, etc.The superior mechanical performance and goodcorrosion resistance of the superalloys, and espe-cially of the nickel-base superalloys at high tem-perature, make them useful materials for compo-nents that are subjected to high temperature ser-vice environments (such as blades or vanes) inindustrial gas turbines and other energy conver-sion systems. However, the presence of combus-tion gases constitutes an extreme environment andhot corrosion is inevitable when alloys are used athigh temperatures for long periods of time, (Goebelet al., [12]).

2. HOT CORROSIONAccording to [1], metals and alloys sometimes ex-perience accelerated oxidation when their surfacesare covered with a thin film of fused salt in an oxi-dizing gas atmosphere at elevated temperatures.This is known as hot corrosion where a porous non-protective oxide scale is formed at the surfacesand sulphides in the substrate. If concentration ofthe sulphate exceeds the saturation vapor pres-sure at the operating metal temperature for turbineblades, vanes and energy generation components(700-1100 °C), then Na2SO4 will deposit on thesurface of these components. At higher tempera-tures these deposits of Na2SO4 are molten (m.p.884 °C) and can cause accelerated attack of Ni-and Co-base superalloys. Furthermore, the accel-erated corrosion can also be caused by other salts,such as vanadates or sulphate-vanadate mixturesand in the presence of solid or gaseous salts suchas chlorides.


2.1. High temperature (Type I) hotcorrosion (HTHC)

High temperature (Type I) hot corrosion (HTHC) isnominally observed in the temperature range ofabout 825-950 °C when the condensed salt film isclearly liquid. The typical microstructure for HTHCshows the formation of sulphides and a correspond-ing depletion of the reactive component in the al-loy substrate. The external corrosion products fre-quently comprise oxide precipitates dispersed inthe salt film [1]. The dominant salt in HTHC isNa2SO4 due to its high thermodynamic stability. Themacroscopic appearance of HTHC is character-ized in many cases by severe peeling of the metaland by significant color changes (greenish tone,resulting from the formation of NiO) in the area ofthe accelerated attack (Eliaz et al., [8]).

2.2. Low temperature (Type II) hotcorrosion (LTHC)

Low temperature (Type II) hot corrosion (LTHC)occurs well below the melting point of pure Na2SO4.The reaction product morphology for this type ofcorrosion can be characterized by a non-uniformattack in the form of pits, with only little sulphideformation close to the alloy/scale interface and littledepletion of Cr or Al in the alloy substrate [1]. Thisform of corrosion is observed mainly within the tem-perature range 650-800 °C (Nicholls, [13]). Theformation of low melting point eutectics causes typi-cal LTHC pitting, for instance the formation ofNa2SO4-NiSO4 eutectics for nickel-based superal-loys. Wright [14] suggested that a high partial pres-sure of SO3 in the gaseous phase is required inthe LTHC reactions to occur, in contrary to HTHC.

3. HOT CORROSION OFSUPERALLOYS

Hot corrosion degradation process of the superal-loys usually consists of two stages, namely an ini-tiation stage and a propagation stage (Pettit andMeier [15] and Pettit and Giggins [16]). Accordingto Pettit and Meier [15], it is a fact that all corrosionresistant alloys degrade via these two stages andit is the result of using selective oxidation to de-velop oxidation or corrosion resistance. They fur-ther elaborated that the conditions causing hot cor-rosion therefore do nothing more than shorteningthe time for which the superalloys can form pro-tective alumina or chromia scales via selectiveoxidation.

During the initiation stage of hot corrosion, su-peralloys are being degraded at rates similar tothose that would have prevailed in the absence ofthe deposits. Elements in the alloy are oxidized andelectrons are transferred from metallic atoms tothe reducible substances in the deposit. Conse-quently, the reaction product barrier that forms be-neath the deposit on the alloy surface usually ex-hibits primarily those features resulting from thegas–alloy reaction (Pettit and Giggins, [16]).

In some cases of hot corrosion, an increasingamount of sulphide particles become evident in thealloy beneath the protective reaction product bar-rier. In other, small holes become evident in theprotective reaction product barrier where the mol-ten deposits begin to penetrate it. Eventually theprotective barrier formed via selective oxidation isrendered ineffective, and the hot corrosion processenters into the propagation stage. Obviously, in at-tempting to develop resistance to hot corrosion oneshould strive to have the superalloys remain in theinitiation stage as long as possible (Pettit and Meier,[15]).

Numerous factors affect the time at which thehot corrosion process moves from the initiationstage into the propagation stage as shown in Fig.1. These factors also play the dominant role in de-termining the type of reaction product that is formedin the propagation stage. This fact is responsiblefor a variety of hot corrosion processes that havebeen observed when superalloys are exposed todifferent environments (Pettit and Meier, [15]).

The propagation stage of the hot corrosion se-quence is the stage for which the superalloy mustbe removed from service since this stage alwayshas much larger corrosion rates than for the samesuperalloy in the initiation stage (Pettit and Meier,[15] and Pettit and Giggins, [16]). The pertinentmechanism behind this mode of hot corrosion hasbeen summarised in Section 4.

However, Pettit and Giggins [16] opined that thehot corrosion degradation sequence is not alwaysclearly evident, and the time for which protectivereaction products are stable beneath the salt layeris influenced by a number of factors. The alloysmust be depleted of certain elements before non-protective products can be formed, or the compo-sition of the deposits must change to prevent theformation of protective scales. There are caseswhen the initiation stage does not exist at all andthe degradation process directly enters into thepropagation stage.

As per the views of Yoshiba, [17], Ni-based su-peralloys are well known to be susceptible to hot


Fig. 1. Schematic diagram illustrating the conditions that develop during the initiation and the propagationof hot corrosion attack and to identify the factors that determine the time at which the transition from theinitiation to the propagation stage occurs (see [15]).

corrosion associated with combined reaction pro-cesses such as sulphidation-oxidation and occa-sionally together with chlorination. In practice, it isa serious problem for such alloy systems whenused for hot section components of various heatengines such as gas turbines subjected to corro-sion-induced strength degradation.

4. MECHANISMS OF THE HOTCORROSION

Stringer [18] presented an excellent review of thework done in the field of hot corrosion up to about1976, which was aimed at understanding the hotcorrosion processes. Much of the mechanismsproposed during that period were focused on ther-mochemistry. Bornstein and DeCrescente [19,20]and Bornstein et al. [21] proposed a hot corrosionmechanism based on the basic dissolution of theprotective oxide scale by a reaction involving Na2O,the basic minority component of the fused salt.

Goebel and Pettit [22] also interpreted the hot cor-rosion of pure nickel in terms of basic dissolutionand reprecipitation of NiO in the fused salt film.Goebel et al. [12] extended this mechanism to in-clude acidic fluxing and oxide re-precipitation toaccount for the catastrophic oxidation caused bypure Na2SO4 for alloys containing strong acid com-ponents such as vanadium or molybdenum. Thenext important contribution in the fluxing model wasmade by Rapp and Goto [23]), who proposed thatif the gradient in solubility of the protective oxidewith distance into the salt layer becomes negativeat the oxide/salt interface, accelerated attack couldbe sustained. Sodium vanadyl vanadate(Na2O.V2O4

.5V2O5), which melts at a relatively lowtemperature 550 °C is found to be the most com-mon salt deposit on boiler superheaters. Additionof Na2O to liquid V2O5 causes an increase in thebasicity of the melt with a corresponding increasein corrosivity with respect to the acidic oxides.Moreover, Na2O has also been reported to de-


Fig. 2. Schematic representation of the reaction sequence during low-temperature hot corrosion of a Co-30%Cr alloy exposed to O2-SO2-SO3 environment where Na2SO4-CoSO4 liquid and Co3O4 are stable. Athigher concentration of SO3, where Co3O4 is unstable at the gas-salt interface, the outward migratingcobalt will form CoSO4 (s) or Co3O4 and CoSO4(s) (see, e.g. [29]).

crease the viscosity. Therefore the protective ox-ides become porous and non-adherent. The for-mation of binary and tertiary low melting eutecticsincreases the surface attack thereby reducing theuseful life of the component.

Several mechanisms have been proposed todescribe the phenomenon of hot corrosion (Goebeland Pettit [4,22,24], Rapp and Goto, [23], Pettit andGiggins [16], Stringer [18], Otsuka and Rapp, [25],and Zhang et al., [26]). Eliaz et al. [8] briefed intheir review on failures of turbine materials that theinitiation of high temperature hot corrosion (HTHC)is often attributed to failure of the protective oxidelayer, which allows the molten salt to access di-

rectly the substrate metal. This failure may resultfrom erosion, thermal stresses, erosion-corrosion,chemical reactions, etc. The mechanisms pro-posed for the HTHC propagation stage are thesulphidation-oxidation mechanism and the salt flux-ing mechanisms (Stringer, [25]). The salt fluxingmechanism was firstly proposed by Goebel andPetit [22,24]. According to this model, the protec-tion efficiency of the surface oxide layer might belost as a result of fluxing of this layer in the moltensalt. This fluxing may be either due to combinationof oxides with O2- to form anions (basic fluxing) orby decomposition of the oxides into the correspond-ing cations and O2-(acidic fluxing). As compared to


basic fluxing, acidic fluxing causes more severeoxidation. The acidic fluxing takes place when theO2- activity in the molten salt is markedly lowered.In contrast to basic fluxing, the acidic fluxing canbe self-sustaining because the displacement of thesalt from the stoichiometry does not become pro-gressively more difficult as the reaction proceeds(Stringer, [25]). In general, the hot corrosion of su-peralloys with high contents of aluminium and chro-mium is often reported to occur according to thebasic fluxing mechanism, whereas the hot corro-sion of alloys with high contents of tungsten, mo-lybdenum, and vanadium has been reported to fol-low the acidic fluxing mechanism.

Rapp and Goto [23] measured the oxide solu-bilities in molten Na2SO4 as a function of the acid-ity of the salt. They suggested on the basis of thesemeasurements that a negative gradient of the solu-bility of the protective oxide in the salt film at theoxide/salt interface should lead to oxide dissolu-tion at this interface and to precipitation of a non-protective oxide away from the interface, where thesolubility is lower. In this case fluxing arises onlybecause of the local variation of sodium oxide ac-tivity and/or oxygen partial pressure across the saltfilm, without any necessity of sulphide-forming re-actions. This mechanism can explain a self-sus-

taining process of dissolution of the protective ox-ide to maintain an accelerated corrosion attack(Stringer, [25]).

The effect of vanadium on HTHC has also beenstudied by different researchers. Bornstein et al.[21] and Goebel et al. [12] opined that a self-sus-tained acidic dissolution of the protective Cr2O3 orAl2O3 scales could take place when the salt filmcontains vanadium, because V2O5 is a strong acidicoxide. Further, Zhang and Rapp [28] believed thatevery oxide should form an acidic solute with muchhigher solubility in the presence of vanadate, whichshould contribute to the more accelerated attackof oxides by mixed sulphate-vanadate melts thanby a pure sulphate melt.

The phenomenon of LTHC was explained byLuthra [29] with the help of a common model, ac-cording to which LTHC follows two stages. In thefirst stage liquid sodium-cobalt sulphate is formedon the surface, whereas in the second stage propa-gation of the attack takes place, refer Fig. 2. Thepropagation is suggested to occur via migration ofSO3 and cobalt inward and outward, respectively,through the liquid salt.

Some of the hot corrosion studies conductedby various investigators in the Na2SO4 and V2O5environments are summarized in Table 1.

Material Environment Brief Details

Na2SO4 Induced Hot Corrosion

Pure Iron and Fe-5Cr 1 atmosphere Pure iron did not undergo accelerated oxidation, whichAlloy, and Fe-13Cr Alloy of oxygen, 900 °C has been attributed to the thickening of the scale too

rapidly for sulphur to penetrate the oxide and interactdirectly with the metal. Authors further reported an im-mediate acceleration in oxidation rate of Fe-5Cr alloy inthe presence of Na2SO4 deposits this was attributed tosulphide formation mechanism which initially restrictedspinel formation (Trafford and Whittle, [30]).In their study conducted on Fe-13% Cr alloy, Na2SO4coating markedly enhanced the oxidation rate and re-sulted in the formation of thick, compact and stratifiedscales. They postulated that formation of sulphides inthe alloy substrate and mechanical failure of scale wasresponsible for the enhanced oxidation (Trafford andWhittle, [31]).

Table 1. A summary of hot corrosion studies on some Fe- and Ni-base alloys in the Na2SO4 and V2O5environments.


Pure iron, chromium Gas mixture, The enhanced corrosion phenomenon was interpreted and manganese, containing O2 (3.6%), in terms of low melting liquid sulphate formation. Fur- and chromium SO2 (0.25%) and N2 ther the corrosion rate of pure chromium was not ap-

(bal), 600-800 °C preciably affected by the salt deposit (Nanni et al., [32]). Pure iron Simulated combustion The accelerated reaction observed in the presence of

gas containing SO3, Na2SO4 deposits was attributed to the formation of a 600-800 °C liquid salt solution between Na2SO4 and the sulphates

of the corroded metal with the production of duplexscales consisting of a mixture of metal oxides andsulphides (Gesmundo and Viani, [33]).

Iron In oxygen, 750 ° C The accelerated oxidation of iron in the presence ofdeposits of Na2SO4 was attributed to the formation ofabundant sulphide which had a highly defected latticeand allowed rapid diffusion of a liquid phase. The liquidphase was a eutectic melt of Na2SO4 and Na2O. Fur-ther it was concluded that the accelerated oxidation hadmost of the characteristics common with usual low temperature hot corrosion except that it occurred underbasic conditions developed by the removal of sulphurfrom the sulphate deposits instead of the usual acidicconditions established by the SO3 in the atmosphere(Shi, [34]).

Pure iron and Na2SO4 deposit, O2- The accelerated oxidation of pure iron beneath Na2SO4 Fe-Al Alloys rich atmospheres deposit was believed to be induced by the formation of

containing SO2-SO3 eutectic melt of Na2SO4 -iron sulphate. The growth of at 650 –700 °C oxide has been reported to predominate although flux

ing contributed a little to the corrosion. In alloys con-taining Al, fluxing proceeded at a considerable rate inthe early stages due to the different oxygen pressure atthe oxide/melt interface. Reduction of SO3 released SO2,which caused the formation of sulphides in the scale.Sulphides were supposed to contribute considerably tothe acceleration due to their high defect concentration(Shi et al., [35]).

Pure Ni and Al and In oxygen and air, Shi [36] studied the possibility of Na2SO4-Na2O eutectic Fe-Cr, Ni-Cr and 750 °C melt formation on the metals deposited with Na2SO4. In Fe-Al alloys the case of Ni, Co, Al, Cr, and their alloys he could not

detect formation of Na2SO4-Na2O. In the case of iron-based alloy with high Cr or Al content, where Cr2O3 orAl2O3 were formed, again Na2SO4-Na2O eutectic wasnot observed. However, at lower Cr or Al content, thiseutectic melt was suggested to be the possible causefor accelerated rate of corrosion.

Pure iron Na2SO4, Na2SO4+ Iron suffered low temperature hot corrosion in the pres- NaCl, Na2SO4+V2O5, ence of salt deposits at 600 °C. The additions of NaCl combustion gas, and V2O5 to Na2SO4 changed the corrosion kinetics sig

600 °C nificantly and modified the scale structure (Li et al., [37]).Binary iron aluminide In pure oxygen, 827, The faster kinetics observed in the initial stages of oxi-

(Fe3Al) 952, and 1057 °C dation was related to the formation of θ-Al2O3 and slowerkinetics in the later stages of oxidation to the formationof α-Al2O3. The overall rate of hot corrosion was higherthan that of oxidation at all temperatures. The presenceof α-Fe2O3 in addition to alumina was indicated by XRD.


Cross-sectional microscopy revealed that the metalscale interfaces were pitted in hot corrosion conditionsand the pits contained aluminum sulphide. Sulphideswere also detected along grain boundaries in the intermetallic near the scale-metal interface (Das et al., [378]).

Ferritic steels Fused salt mixture The presence of Ti with 12% Cr and 3% Al increased containing Cr, Al, containing Na2SO4 at the proportion of Fe, which diffused to the outer layer of and Ti 927 °C and gas mixture the scale. The excellent behavior of steels at tempera-

CO, CO2, CH4, H2O, H2 tures above 600 °C was explained by the increase in Al and H2S at 600 °C diffusion coefficient with temperature (Coze et al., [39]).

310 Stainless steel Various ratios of The weight gain kinetics in simple oxidation was found Na2SO4/NaCl deposit, to follow a steady state parabolic rate law after 3 hours,

in air at 750 °C while the kinetics with the salt deposits displayed multistage growth rates. The most severe corrosion tookplace with 75% NaCl mixtures. Uniform internal attackwas the morphology of NaCl-induced hot corrosion,while the extent of intergranular attack was more pro-nounced as the content of Na2SO4 in the mixture wasincreased (Tsaur et al., [40]).

Pure nickel and In a high velocity burner Corrosion of Ni in the burner rig produced a relatively Udimet 700 rig, 900 °C compact NiO scale along with some internal grain

boundary corrosion. Corrosion of Udimet 700 was ob-served to occur in two stages. During the first stage,the corrosion proceeded by the reaction of Cr2O3 scalewith the Na2SO4 and evaporation of the Na2CrO4 reac-tion product from the surface of the corroding sample.Cr depletion in the alloy occurred and sulphide particleswere formed in the Cr depletion zone. Extensivesulphidation occurred during the second stage of corrosion and a thick scale was formed (Misra, [41]).

Pre-oxidised nickel In an SO2-O2 gas Chromate anion suppressed the sulphidation of Ni prob atmosphere at 900 °C ably by precipitating solid Cr2O3 from the melt. Vana-

date anions enhanced the onset of the hot corrosionand sulphidation probably via rapid dissolution of theprotective oxide scale at cracks/defects or grain boundaries (Otsuka and Rapp, [26]).

Pure nickel Under SO3 + SO2 + O2 The corrosion loss in mixed atmosphere containing SO3 gas mixture & SO2 and O2 was reported to be larger than those observed in pure atmospheres, 900 °C SO2 and O2 atmospheres. The corrosion loss was found

to correspond to the thickness of the oxide layers. Highcorrosion losses were attributed to the fact that SO3strongly acted as an oxidizing agent for the corrosionprocess (Hara et al., [42]).

Ni-23.1 Nb-4.4AI Dean rig test, 900 °C and Below the outer layer, mainly of NiNb2O6, a high propor and Ni-19.7Nb-6.0 1100°C a high proportion of Al2O3 has been reported to be Cr-2.5AI present. Internal oxidation of’ the metal producing Al2O3

and Cr2O3 was believed to occur before the development of’ external Al2O3 layer. In some isolated regionsthe alloy was found to be much more severely attacked,the pit contained NiNb2O6, NbCrO4, and Ni. At 1100 °C,more uniform corrosion was observed, the outer layerwas mainly NiCr2O4 and beneath it a layer of Cr2O3 containing particles of NbCrO4 was observed. Internal ox-


ides were mainly Al2O3 and there were massive Ni-richsulphides ahead of oxidation front (Johnson et al., [43]).

Ni-30Cr and 600-900 °C The rapid rate of attack was explained on the basis of Co-30Cr sulphation of the transient surface oxides (Ni or Co ox-

ides) and the dissolution of these transition metalsulphates into Na2SO4 to yield a liquid phase (Luthraand Shores, [44]).

B-1900 Na2SO4 Na2SO4 interacted with the alloy to form sodium andsulphur compounds, rapid removal of sulphur fromNa2SO4 by unretarded diffusion of sulphur and precipi-tation of Cr-rich sulphide phases promoted the forma-tion of Na2O.The catastrophic oxidation observed dur-ing sulphidation was due to interactions between Na2Oand the substrate (Bornstein and DeCrescente, [45]).

Ni-base Pure O2, 825-1000 °C Based upon the results, it was concluded that the re- superalloys and duction of the oxide ion content of Na2SO4 was a ne- seven binary cessity, but not sufficient condition for sulphidation inhi- Ni-base alloys bition. The addition of Mo or V to nickel imparted

sulphidation resistance in it because their oxides re-acted with and decreased the oxide ion content ofNa2SO4. The disagreement in the literature regardingthe effect of Molybdenum on hot corrosion was sug-gested to be largely due to differences in testing tech-niques and differences in whether the investigators havebeen more concerned with the initiation or with propa-gation modes of hot corrosion (Bornstein et al., [21]).

Ni- base superalloys 1 atm O2, 650-1000 ° C The alloys underwent catastrophic corrosion. The, ac- (Goebel et al., [12]), O2, celerated oxidation occurred as a result of the forma- 750-950 °C (Misra, [46]) tion of a liquid flux based on Na2SO4 which dissolved

the normally protective oxide scales. Catastrophic orselfsustaining rapid oxidation can occur in alloys whichcontain Mo, W or V because solution of oxides of theseelements with Na2SO4 decrease the oxide ion activityof the molten salts, producing melts which are acidicfluxes for oxide scales (Goebel et al., [12], and Misra,[46]). The evaporation rate of MoO3 from Na2Mo-MoO3melts has been reported to decrease considerably [46](1986) when dissolved in Na2MoO4.

Ni-based industrial Static deposits of Na2SO4 The susceptibility to hot corrosion was found to be cor- superalloys or NaCl or both in still related to the type of scale produced during simple oxi-

air, 850-1000 °C dation. Alloys forming an Al2O3 scale were found to besusceptible to Na2SO4 deposits, independent of theirCr content. The quantity of the Na2SO deposits dictatedthe nature of attack and under certain conditions, therefractory element alloy additions appeared to play anessential role. Alloys containing Cr2O3 or TiO2 in thesimple oxidation scale proved to be sensitive to NaClattack (Bourhis and John, [47]).

Ni-15Cr-Mo (Peters 900 °C (Peters et al., The effect of Mo on the hot corrosion of superalloys et al., [48]), Fe-, Ni- [48]), 600 °C and above has also been reported by Peters et al. [48], Pettit andand Co-based alloys (Pettit and Meier, [15]), Meier [15], and Fryburg et al. [48]). The alloy containing

975 °C in oxygen Mo suffered catastrophic degradation. It has been re-


(Fryburg et al., [50]) ported that MoO2 reacted with Na2SO4 to produce anacid (SO2-rich) salt, leading to acidic fluxing. MoO3 incorporated into the Na2SO4 via the formation of com-pounds such as Na2MoO4, Na2MoO4

.MoO3, andNa2MoO4

.2MoO4. All these phases were liquid and lefta high solubility for Al2O3 and Cr2O3. Peters et at. [48]added that there is a threshold amount of molybdenumbelow which catastrophic attack is not encountered, e.g.for Ni- 15% Cr threshold has been reported to be be-tween 3-4%.

Ni-based In pure O2, 900 °C After an induction period of little corrosion, local basicsuperalloys, B-1900 fluxing attack of the Cr2O3/ Al2O3 scale spread to cover and NASA-TRW the surface and generated catastrophic linear kinetics.

IVA During catastrophic attack of B-1900, the sulphate ionsreacted to release SO2 and formed sulphides in the alloy and salt was converted to Na2MoO4 (Fryburg et al.,[50]).

Fe-, Ni- and Atmospheres containing The corrosion rate was lowest when the chromium con- Co-based alloys O2, N2, and SO2, tent of the alloy was highest. Further Mo and Cu were

729-1076 °C found to increase the corrosion rate. The main corro-sion products formed in air were NiO and Cr2O3. In hotcorrosion tests NiS and Cr5S8 were found (Pehkonen etal., [51]).

Ni and Ni-based Pure Na2SO4 Low alloyed aluminide and Cr-Al coatings showed very Alloy EI 867 with poor resistance to oxidation. After 24 hrs, these had the Aluminide and been almost completely removed. Modification of highly Cr-Al Diffusion alloyed aluminide coatings with Cr resulted in uniform Coatings and relatively slow degradation of the coating. Cr en-

riched zone is supposed to act as a barrier to the oxida-tion of refractory metals such as Mo, W, and V thuspreventing the onset of catastrophic corrosion(Godlewski and Godlewska, [52]).

Ni-16Cr-2Nb, IN738, 900 and 1000 °C Na2SO4 coated coupons of Ni-16Cr-2Nb and Ni-16Cr Ni-16Cr, and developed dense, protective oxide scales and exhib- Superalloy 537 ited good hot corrosion resistance. Alloys 537 and IN

738 experienced a shift from basic fluxing to acidic flux-ing and as the temperature was increased, the rate ofattack increased significantly (Zho et al., [53]).

Inconel 600 and Ar +1% O2, 940 °C Inconel 600 was able to form a protective layer of Incoloy 825 chromia in Ar +1% O2 with or without a thin layer of

Na2SO4 at a temperature of 940 °C, whereas Incoloy825 was not able to form chromia in the same environment. This was attributed to the formation of nickel andchromium molybdates due to 3.22% Mo present in thealloy. When O2 content was only 300 ppm, sodium sul-phate layer strongly enhanced the corrosion of Inconel600 and Incoloy 825 (Santorelli et al., [54]).

Review (Shih, [55]) Low-temperature hot Some superalloys containing molybdenum are sub- corrosion jected to sub-melting point hot corrosion when covered

with a sulphate, which is associated with the formationof a molybdenum-containing melt resulting from the re-action of MoO3 with the sulphate (Shih, [55]).


Superalloys Mach 0.3 burner rig at Accelerated corrosion of the specimens has been re HA-188, S-57, various temperatures, sea ported in the temperature zones above the calculated IN-617, and salt concentrations and dew points of Na2SO4. It was believed that the corro- TD-NiCrAl salt compositions. sion occurred by small amount of salt deposit on those

temperature zones during heat-up following each cooling cycle or by small amounts of salt migrating fromdeposits at cooler zones via a wetting action. Largedeposits of salt appear to inhibit corrosion that is whyless corrosion was noticed at 900 °C when the flamewas doped with 10 ppm sea salt than when it was dopedwith 5 ppm sea salt (Santoro, [56]).

Nimonic 105 900 °C Hot corrosion did not seem to be very detrimental towards the superalloy. It was concluded that SO3 pres-sures below 5.10-3 atm did not affect the electrokineticbehaviour, but pressures greater than 5.10-3 atmo-spheres produced higher corrosion rates which wasattributed to the acid fluxing by the sulphate melt. Addi-tion of NaCl to the molten Na2SO4 resulted in increaseddissolution of Nimonic 105 (Sequeira and Hocking, [57]).

Nimonic 105, 75, With and without Up to 800 °C, the lower oxidation rates for Na2SO4 80A and 90 Cr2(SO4)3, NiSO4 or coated alloys were attributed to a scale morphology

CoSO4 additions, consisting of inner scales of Cr2O3 acting as a protec- 650-1000 °C tive oxide film and external scales of NiO. This mor-

phology was observed to be maintained at high tem-peratures (Malik and Ahmad, [58]).

Alloys B1900 and IN In the presence of The Cr2O3 formers got attacked more aggressively by 100 (Al2O3-formers), Na2SO4 (s) and NaCl NaCl (s) than Na2SO4 (s). Al2O3-formers and to some Inconel 600, 690, (s) separately, and in extent NiO-formers were more resistant to NaCl attack Incoloy 800, IN 738, combination, air, 850 than Cr2O3 formers. While Na2SO4 induced corrosion Nimonic 80A, 100, and 1000 °C preceded by fluxing and sulphidation reactions, the NaCl and 105 induced corrosion was observed to follow a reaction path (Cr2O3-formers) of fluxing, chloridation and oxidation. The Na2SO4–NaCl

induced corrosion involved combination of reactions occurring in the presence Na2SO4 and NaCl separately(Malik et al., [59]).

Single-crystal 704 and 900 °C All the alloys were severely degraded at 900 and 704Ni-based superalloys °C. There was no significant difference between the hot and DS Mar M 200 corrosion of these alloys when tested in air or in oxygen

with SO3, provided a liquid deposit was present in bothcases. It was concluded that all the superalloys understudy require a coating if they were to be exposed toany of the hot-corrosion conditions at temperatures of704 °C or higher (Levy et al., [60]).

Inconel 740 850 – 1000 °C The kinetics curves of the alloy at 950 °C without Na2SO4deposit and at 850 °C with Na2SO4 deposit obeyed theparabolic law, whereas the uniform parabolic growthbehaviour of oxide was not followed at 1000 °C withoutNa2SO4 deposit and at 950 °C with Na2SO4 deposit dueto oxide spallation at 1000 °C and the evaporation ofNa2CrO4 melt at 950 °C, respectively. The oxide scaleswere found to be consisting of Cr2O3, (Ni, Co) Cr2O4,TiO2, and Al2O3 and internal oxide scales at all tem-peratures. The internal suphidation was suggested to


take place due to the existence of Na2SO4 deposit. Thecomplex layered structure of the oxide scales was re-ported to be in favor of the resistance to oxidation. Theaccelerated oxidation of the alloy in the presence ofNa2SO4 deposit was attributed to the dissolution of Cr2O3induced by basic fluxing in molten Na2SO4 (Zhao et al.,[61]).

Co-Cr, Co-Al, and In O2-SO2-SO3, Co-Cr and Co-Cr-Al alloys reacted non-uniformly, usu- Co-Cr-Al Alloys 600 to 750 °C ally in the form of pits and Co-Al alloys suffered broad

frontal attack. Under all conditions, a thin sulphur–richband containing sulphides was observed at the alloy/scale interface and cobalt dissolved near the interfaceand formed Co3O4 /or CoSO4 (S) (Luthra, [62]).

Co-Cr, Co-Al, and In O2-0.15% Accelerated oxidation tests at 750 °C showed that the Co-Cr-Al Alloys (SO2+SO3), 750 °C corrosion resistance of binary Co-Cr and Co-Al alloys

increased with the Cr and Al content of alloys. This pro-tection was offered by the rapid growth of CoCr2O4/Cr2O3and CoAl2O4/Al2O3 oxides in comparison to CoO andCo3O4. At high enough Cr (=40%) and Al (=15%) con-centrations, the growth rates were so fast that liquid didnot even form; consequently the corrosion rates werevery low. Tests on Co-Cr-Al alloys indicated that simul-taneous presence of Cr and Al was deleterious to theresistance against low temperature hot corrosion(Luthra, [63]).

Metal oxides such 627-927 ° (Malik The interaction of the metal oxides with Na2SO4 has as Co3O4, NiO, and Mobin, been studied. As suggested, the high temperature re- Al2O3, Cr2O3, Fe2O3, [64]) and 827-927 °C action products usually contain 3-phase structures and SiO2 (Mobin et al. [65]) in O2 namely, Na2OMO-M2O3 and metal sulphide and or metal

sulphate. The formation of Na2O-M2O was further re-ported to be dependent upon the solid state solubility ofmetal oxide in the molten salt at high temperature andunder limited solubility conditions Na2O-M2O was invari-ably formed, but as soon as conditions relaxed the ox-ide M2O3 precipitated and was observed to form a sepa-rate phase (Malik and Mobin [63], and Mobin et al., [64]).

V2O5 Induced Hot Corrosion

Gas–turbine alloys 750 °C and above Vanadium pentaoxide coatings had a deleterious effectthroughout the useful temperature range of all the al-loys up to 1120 hrs in 70-hr cycles. The effect was re-ported to be more pronounced for iron based alloys attemperatures above 750 °C (Harris et al., [66]).

Review - - - Oxidation of pure Cr in V2O5 occurs with a very rapiddiffusion rate and so only the initial stages of the oxida-tion curve were supposed to be more important. Laterslowing down of the oxidation rate could be attributedmainly to the effect of scale thickening. Loose andspongy appearance of the scale was observed at thebeginning of the process; V2O5 was present in excessand did dissolve the products of oxidation. At the samepoint the liquid was saturated with the oxide which sub-


sequently could get precipitated. The presence of vari-ous phases in a thin layer of scale would impose suchsevere strain on the film that cracking and exfoliationcould be expected. This would permit liquid phase toreach the metal surface again and the conditions to forma spongy scale were seen to prevail. This mechanismwould apply only to iron base alloys, which are suscep-tible to the catastrophic corrosion (Sachs, [67]).

Technical note - - - Fairman [68] reported that the accelerated oxidation due (Fairman,[68]) to V2O5 is most likely to be the consequence of the cata- Review (Pantony lytic action of vanadium pentoxide. It results in splitting and Vasu [69]) the oxygen molecule at the metal surface, so cause

rapid corrosion. Pantony and Vasu [69] concluded fromthe theoretical survey of fireside corrosion of boilers andgasturbines that in residual fuel ash, V2O5 can be themost serious cause of “catastrophic” corrosion.

Pure metals such as Immersion in the melt, A diffusion controlled corrosion process given by equa- iron, cobalt, nickel, in O2 tion below hold good for the initial stages of vanadic molybdenum, corrosion of all metals under study except nickel and titanium, tungsten, chromium.and 99.5% vanadium ∆w = Kt, where ∆w is the weight change per cm2 at time rods, and single t, and K is the velocity constant.crystals of chromium Whereas nickel obeyed a logarithmic rate law. The ve-

locity constant was found to be inversely proportional tothe depth of melt. On the basis of comparison of activa-tion energies of the various rate processes, a singlemechanism underlying the corrosion processes of iron,cobalt, vanadium, titanium, tungsten, and molybdenumwas suggested, which involves an inward diffusion ofoxygen (or other active species) and a sequential outward diffusion of the corrosion products. In case of nickeland chromium, the existence of a coherent corrosionlayer separating the slag from nickel and chromium wasrevealed, which was named as protective barrier forthese metals and was found to be absent for othermetals under study (Pantony and Vasu, [70]).

Iron, nickel, and Crucible and rotating The stainless steels and particularly 440 stainless steel several alloys disk testing (25 wt.% Cr) showed the best corrosion resistance to containing iron, liquid V2O5. The rate of corrosion of Armco iron by liq- nickel, and uid V2O5 has been reported to be controlled initially by chromium the diffusion of oxygen across the metal oxide-liquid

vanadate interface. As the available oxygen ions gotdepleted from the melt, the rate controlling mechanismwas observed to be changed to the sorption of oxygenat the liquid vanadate gas phase interface. Whereas incase of nickel, Inconel 600 and the iron alloys, the ratesof corrosion were found to be affected by the formationand dissolution of the protective metal oxide layers. Theyconcluded that the presence of Fe and Na in the vana-date melts did not alter the rate controlling process ofoxygen diffusion and oxygen sorption, but it increasedthe non stoichiometry and hence increased the oxygendiffusion. The rate of corrosion was reported to be af-


fected by temperature, oxygen partial pressure and rotational speed, volume of liquid vanadate, compositionof the metal and composition of the liquid vanadates.According to them, Inconel showed better corrosion resistance than stainless steels (Kerby and Wilson, [71]).

High temperature - - - Vanadate-induced corrosion of high temperature alloys alloys has been studied by thermogravimetry in laboratory fur

naces (using a limited amount of vanadate deposit), inburner rigs (where small amounts of vanadates arecontinually being deposited on the surface), or by immersion of specimens in crucibles with molten salt. Thereaction products formed during exposure in laboratoryfurnaces and in burner rigs were qualitatively the same,but the kinetics and reaction rates were differing con-siderably. The mechanisms were found to be complex(Beltran and Shores, [4]).

50Cr-50Ni alloy In pure V2O5 and In pure V2O5 at 810 °C, a Cr2O3 scale has been ob sodium vanadates, served on the alloy which subsequently got dissolved 750-950 °C in a rotating slowly into the liquid melt and was thus proposed as a

disk apparatus barrier layer by them. In the case of NV6, this barrierlayer was not observed. The increased basic characterof the melt and its consequently greater fluxing abilitytoward acidic oxides was thought to be a more important part process than the increased ionic conductivityof NV6 over V2O5 at this temperature. At 950 °C thecorrosion in terms of both dissolution rate and corro-sion rate was reportedly greater in V2O5 than in NV6(Dooley and Wilson, [72]).

Superalloys 700 °C The relationship of the high temperature corrosion ofsuperalloys with contaminants has been reported byHancock [73]. It was proposed that to compare con-taminant conditions the contaminant flux rate (CFR)rather than the contaminant level in the fuel or environment should be considered. He further suggested thatat temperatures above 700 °C where vanadates causefluxing of the protective oxide scales, corrosion couldbe determined by the CFR and temperature rather thanby material selection.

Nickel base In air, 700 °C Vanadium, present as vanadium pentoxide, attacked superalloy the alloy severely at the temperature of investigation.

Study of ternary oxide system and spot tests showedthat two low melting eutectics, namely, NiO-V2O5-Cr2O3melting at 550 °C and V2O5-Cr2O3-Fe2O3 melting at480 °C, were formed. The formation of these liquid eutectics and the presence of corrosive V2O5 caused severe damage to the superalloy (Iyer et al., [74]).

Nickel base In air, 650, 700, The hot corrosion kinetics obeyed a parabolic law withsuperalloys Nimonic and 750 °C two regions at 700 and 750 °C, the corrosion rate fall 80A, Hastealloy ing with the formation of a solid Ni3V2O8 layer. The acid- C-276, and base fluxing mechanism and a second mechanism by Superni 600 which nickel vanadate was precipitated from a nickel

vanadium compound located in the short circuit diffusion path has been assumed to be occurring in the growing oxide. The corrosion rate for these alloys got de-creased with the formation of a compact solid vanadatelayer (Swaminathan et al., [75]).


5. STUDIES RELATED TO Na2SO4-V2O5 INDUCED HOT CORROSION

The kinetics of the reactions between Na2SO4 (X)and V2O5 (Y) as studied by Kolta et al. [76] revealedthat the rate of reaction depended both on the tem-perature (600-1300 °C) and the molar ratios of X:Y.They further found that with increase in the reac-tion period (>30 min.), there was a decrease in re-action rate which finally reached to zero order. Thisdecrease in the reaction rate has been attributedto the formation of vanadosulphate complexes suchas (NaV3O8)2

.Na2SO4 and (NaVO3)2.Na2SO4 which

get decomposed at higher temperatures giving themeta- and pyro-vanadates respectively.

Kofstad [77] proposed that during combustion,the vanadium contaminants are oxidized to thehigher valence vanadium oxides (V2O4 and V2O5)and sodium vanadates are formed by the reactionof vanadium oxides and sodium salts, e.g. Na2SO4.Accordingly the composition of the vanadates maybe presented in a simplified form as (Na2O)xV2O5,but the detailed compositions of the individual vana-dates may be more complex as part of the vana-dium may be in the +IV state. The solid compoundscomprise (Na2O)xV2O4(V2O5)12-x (often termed

Fig. 3. Corrosion rate of Type 304 stainless steelin different mixtures of Na2SO4-V2O5 (see [7]).

β-bronze), (Na2O)5(V2O4)x(V2O5)12-x (K bronze),NaVO3 (sodium metavanadate), Na4V2O7 (sodiumpyrovanadate) and Na3VO4 (sodiumorthovanadate). A notable feature of the vanadatesis that they have relatively low melting points, whichextend from 535 °C upwards. Furthermore, metaloxides dissolved in the vanadates may suppressthe melting points and eutectic temperatures evenfurther. He further reported that the slags devel-oped on valves in diesel engines consist predomi-nantly of sodium sulphate and sodium vanadatesand have melting points as low as 400 °C.

Molten vanadates flux oxide ceramics. The solu-bilities of metal oxides may be high and are de-pendent on the Na:V ratio. The solubilities of Cr2O3and Fe2O3 are highest at Na : V ratios close to 5 :12 and amount to about 50 mol.%. For NiO thesolubility is about 60 mol.% for Na : V = 3 : 2 anddecreases to about 55 mol.% for Na : V = 5 : 12.As V2O5 is acidic, it will in general react with morebasic oxides to form the corresponding vanadates.

5.1. Iron and iron-based alloysNatesan [7] reported that the fuels containing only50 ppm vanadium can cause severe attack onstainless steels. Fig. 3 shows the corrosion rate ofType 304 stainless steel in different V2O5-Na2SO4mixtures at temperatures between 790 and 900 °C,as has been presented by him. At temperaturesbelow the melting point of ash, the rate of attack ofstainless steel is low and follows the normal oxida-tion rate. It was concluded that all stainless steelsand other common engineering alloys are severelyattacked by fuel ash that contain vanadium with orwithout sulphates.

Fairman [68] studied some metal specimens inan ash mixture (V2O5 +10% Na2SO4) environmentin air. The corrosion was found to be severe at theash/air or ash/atmosphere interface. The intensityof attack, however, was still higher at the locationswhere the concentration of oxygen and vanadiumpentoxide was greatest, suggesting the transfer ofoxygen atoms or ions by the pentoxide to the metalsurface: 2V2O5→2V2O4+ 2O↓ . He opined that themechanism of accelerated attack could be mostsatisfactorily explained by the catalytic action ofV2O5 operating with an increase in the defect con-centration of the scale.

Valdes et al. [78], in a study of AISI 446 stain-less steel under V2O5 and Na2O.6V2O5 environmentin the temperature range 700-900 °C in air, foundthat the oxide scale was mainly Cr2O3 with somevanadium oxide which acted as a moderate bar-


rier to corrosion. Above 850 °C in V2O5, abreakaway corrosion reaction occurred. No Cr2O3oxide barrier was present but there was a continu-ous oxide that comprised of Cr2O3, Fe2O3, and V2O5at the metal/oxide interface where the crystals grew.It was suggested that the addition of Na2O to V2O5increased the oxide ion (O2-) content of the meltand made it more aggressive to acidic oxides suchas Cr2O3.

Tiwari ([79-81] have reported hot corrosion stud-ies on some industrial superalloys in temperaturerange 700-900 °C in the environments comprisingof pure Na2SO4, Na2SO4-15%V2O5, and Na2SO4-60%V2O5. Very high corrosion rates were observedin the environment having Na2SO4-60%V2O5 com-position. The extremely corrosive nature of thiscomposition was attributed to its low melting pointi.e. 500 °C. Tiwari et al. [80] further revealed that inNa2SO4-60%V2O5 melt, the degradation was dueto the cracking of the protective scale under theinfluence of the fluxing action of the melt for bothFe-base alloy Superfer 800H and Cobase alloySuperco 605. The enhanced degradation was re-ported to be due to the presence of tungsten in theform of Na2WO4-WO3 compound. No oxidation andsulphidation into the substrate was observed in Co-base alloy. Tiwari [81] concluded that the Co-basealloy has inferior corrosion resistance than the Ni-base alloy in Na2SO4-60%V2O5 environment at 900°C.

Almeraya et al. [82] carried out electrochemicalstudies of hot corrosion of steel AISI-SA-213-TP-347H in 80 wt.% V2O5 + 20 wt.% Na2SO4 at 540°C–680 °C and reported the corrosion rate valuesof around 0.58-7.14 mm/year. They observed anincrease in corrosion rate with time. However, thecorrosion potential was observed to be decreasingwith increase in temperature from 540 to 680 °C.Almeraya et al. [83] further conducted similar stud-ies on type 347H stainless steel under the sameenvironment and temperature range, and revealedthe fact that with change in temperature from 540to 680 °C the corrosion potential decreased.

Cuevas-Arteaga et al. [84] used both LPR (Lin-ear Polarisation Resistance) and weight loss tech-niques, in their hot corrosion study on alloy 800 inNa2SO4-20%V2O5, and reported slightly higher cor-rosion rate at 900 °C than at 700 °C. They furtherreported that in both the techniques, corrosion ratesincreased in the beginning of the experiment butdecreased later on until steady values werereached.

Singh [85], during his studies on the hot corro-sion of GrA1, T11, and T22 boiler steels in an envi-

ronment of Na2SO4-60%V2O5 at 900 °C, observedthat the Mo containing T22 steel showed least re-sistance to the hot corrosion attack. This was at-tributed to the formation of a low melting point MoO3phase, which reacted with molten salt resulting information of Na2MoO4. This low melting point ox-ide was suggested to cause the acidic fluxing ofthe protective scale.

The hot corrosion behavior of some Ni- super-alloys viz Superni 75, Superni 600, Superni 601,and Superni 718, and an Fe-based superalloy;Superfer 800H was evaluated by Singh [86] in air,as well as, molten salt environment of Na2SO4-60%V2O5 at 900 °C under cyclic conditions. In gen-eral the superalloys indicated accelerated oxida-tion in Na2SO4-60%V2O5 environment at 900 ºC incomparison to that in air. The Fe-base superalloySuperfer 800H showed a least resistance to thehot corrosion amongst the Superalloys Superni 75,600, 601, and Superfer 800H. The weight gain val-ues for the fifth superalloy Superni 718 could notbe compared with the rest of cases, as data wasavailable for 20 cycles only for this superalloy. Thissuperalloy showed intensive spallation of its scalein the molten salt environment, such that weightgain data could be obtained for 20 cycles only. Onthe basis of cumulative weight gain data for 50cycles, corrosion rate of the superalloys under studycould be arranged in the following order:

Superfer 800H > Superni 601 > Superni 600 >Superni 75

The superior corrosion resistance shown by Ni-base superalloys, Superni 75, and 600 was attrib-uted to the formation of Cr2O3 and nickel vanadatein the oxide scales. Further, it was observed thatall the superalloys under study obeyed parabolicrate law of oxidation up to 50 cycles, except Superni718. In the latter case, the superalloy followed aparabolic rate law upto 20 cycles only. The accel-erated hot corrosion of Superni 718 was ascribedto the presence of higher amount of molybdenum(3.05%) in it. Singh [86] suggested that the molyb-denum might have resulted in the formation ofMoO3, which according to Peters et al. [48] mayreact with Na2SO4 to form a low melting point phaseNa2MoO4 by the following reaction:

Na2SO4 (l)+ MoO3(l) -> Na2MoO4(l) + SO3(g)

This might have caused alloy induced acidicfluxing, subsequently giving rise to acceleratedspalling of the scale. Identical findings have alsobeen reported by Bourhis and John [46], Pettit andMeier [15], Misra [46], and Shih et al. [55]. Fig. 4


Fig. 4. Schematic diagram showing probable hot corrosion mechanism for the uncoated superalloy Superni718 exposed to Na2SO4-60%V2O5 at 900 °C for 50 cycles (see [86]).

shows the plausible mechanism governing the cor-rosion process for superalloy Superni 718 in anenvironment of Na2SO4-60%V2O5 as has been sug-gested by Singh [86].

5.2. Nickel and nickel-based alloysThilakan et al. [87] studied hot corrosion of nickelfree austenitic stainless steel, Cr-Ni stainless steeland Inconel in oxygen atmosphere at different tem-


Fig. 5. Phase Diagram for Na2SO4-V2O5 system(see [91]).

peratures with vanadium-sodium slags of varyingcomposition. They justified the use of saturatedsolution of Na2SO4 as the liquid medium becauseof its low gas solubility, vapor pressure and viscos-ity. Threshold temperature was found to lie between700 and 800 °C. Above the threshold temperaturethe extent of attack initially increased but with in-creasing temperature either it became constant (asobserved for Inconel) or decreased (for other al-loys). Increase in the amount of Na2SO4 in sodiumsulphate-vanadium pentaoxide mixtures first in-creased and then decreased the rate of oxidationat the temperature of 820 and 870 °C. At 950 °C,all Na2SO4 additions to V2O5 decreased its corro-sive effect. The fluidity of slag was observed to beimportant in allowing diffusion of oxygen. Accord-ing to an interesting observation of their study, nickelfree stainless steel was found to have a high resis-tance to attack against V2O5-Na2SO4 mixture, andunder the experimental conditions showed a re-sistance superior to that of even Inconel at tem-peratures above 850 °C especially in 70% V2O5-30% Na2SO4 environment.

Bornstein et al. [21] investigated the relation-ship between vanadium oxide and Na2SO4 accel-erated oxidation for nickel-base superalloy B-1900and seven binary nickelbase alloys. It was estab-lished that the corrosion attack commenced by dis-solution of the normally protective oxide scale intothe fused salt. The dissolution was suggested tooccur by a reaction between oxide ions present inmelt and the oxide scale. They further opined thatthe sulphidation attack can be inhibited by decreas-ing the oxide ion content of the fused salt. In an-other study, Bornstein et al. [88] reported the effectof vanadium and sodium on the accelerated oxi-dation of nickel base alloys. The initial rapid rate ofoxidation between V2O5 and metal substrate wasattributed to the reduction of V2O5 by the substrate.According to them, intermetallics Ni3Al and NiAlwere particularly susceptible to V2O5 corrosion. Itwas further concluded that the sulphidation attackcan be attenuated if the initial oxide ion content ofthe melt is prevented from increasing. Oxides suchas Cr2O3 have been reported to react preferentiallywith oxide ions.

A 50Cr-50Ni commercial alloy was corroded inpure V2O5 and in vanadate melts containing so-dium sulphate and chloride in the temperaturerange 750 to 950 °C in a rotating disc apparatus byDooley et al. [72]. The study reported the forma-tion of chromium rich oxide scale in pure V2O5 whichdissolved only slowly into the liquid melt and thusobserved to act as a barrier layer. In Na2O.6V2O5,

this barrier layer was not observed. Marked inter-nal oxidation of the Cr-rich α-phase was noticed inchloride containing melts throughout the tempera-ture range 750-950 °C.

Thermogravimetric studies which delineate theconditions for simultaneous sulphate and vanadateinduced corrosion at 650 to 800 °C have been car-ried out by Seiersten and Kofstad [89]. They foundthat the corrosion caused by sodium sulphate/so-dium vanadate mixtures have a complex mecha-nism. Samples coated with sodium vanadate wereexposed to O2 + 4% SO2 and the initial reactionwas observed to be the same as that observed inpure oxygen. After an incubation period, the dura-tion was found to decrease with increasing tem-perature and sufficient SO3 got dissolved in themolten vanadate which resulted in formation of amixture of NiSO4 and Na2SO4 near the metal. Whena molten NiSO4-Na2SO4 solution containing smallamounts of vanadate was formed as an intermedi-ate layer, the reaction reportedly proceeded assulphate-induced hot corrosion. The corrosionmechanism was observed to change from initialvanadate-induced to essentially sulphate inducedhot corrosion when the sulphur trioxide pressurewas high enough to form sodium sulphate.

Sidky and Hocking [90] investigated the mecha-nisms of hot corrosion by molten sulphate-vana-


date deposits for Ni-10Cr, Ni-30Cr, Ni-20Cr-3Al, Ni-2lCr-0.3Si, Ni-20Cr-5V alloys, and IN738 superal-loy. The effect of adding Cr to Ni was found to bebeneficial in the Na2SO4 melt, however, on increas-ing the VO3- concentration in the melt, this effectdiminished, becoming harmful in pure NaVO3 dueto the formation of the non protective CrVO4. Ac-cording to them, alloying element Al was found tobe harmful in Na2SO4-NaVO3 melts. Cr depletionwas observed in rich VO3 - melts but internal cor-rosion was more obvious in the SO4

2- rich melts.Corrosion in rich VO3 - melts was aggressive dueto the fluxing action of the salt, which takes placealong internally sulphidised areas. According totheir study, IN738 suffered tremendous internalattack due to its γ precipitates which becamesulphidation prone areas, and were fluxed by theVO3 - melt.

The presence of sulphur and its oxidized com-pounds favor the formation of isolated lobes withradial morphology (Otero et al., [91]) during the hotcorrosion of IN657 at 635 °C in molten salt envi-ronment. These lobes had great permeability whichfacilitated the access of oxygen; therefore the pro-tection character of the scale was reduced. Theyfurther reported that the presence of vanadium andits oxidized products generate compounds withaciculate morphology, which is not very much cov-ering and contribute to reduce the protective char-acter of the scale. The equilibrium diagram for vary-ing composition of Na2SO4 is shown in Fig. 5 andthe mixture of Na2SO4-60%V2O5 is seen to be thelowest eutectic temperature.

The effects of chromate and vanadate anionson the hot corrosion of Ni by a thin fused Na2SO4film in SO2-O2 gas atmosphere at 900 °C wereinvestigated by Otsuka and Rapp [26]. Their re-sults showed that the inhibition of sulphidation mayresult from the precipitation of solid Cr2O3 from themelt and thereby partially sealing/plugging the crackdefects and grain boundaries of the original pro-tective oxide layer. They proposed that vanadateanions enhanced the onset of the hot corrosion andsulphidation probably via rapid dissolution of theprotective oxide scale at cracks/defects or grainboundaries.

Lambert et al. [92] studied the oxidation and hotcorrosion of Ni-17Cr-6Al-0.5Y and Ni-16Cr-5.7Al-0.47Y-5Si alloys at 700 °C in Na2SO4-60%V2O5 toascertain the effect of addition of Si. They observedan outer layer of NiO developed on the surface ofthe Ni-17Cr-6Al-0.5Y alloy whereas a thin Al2O3oxide scale formed on the Si-enriched alloy. Theyfurther opined that the condensed vanadates of

sodium are highly corrosive and can markedly in-crease the rate of oxidation of nickel base superal-loys.

In the same aggressive environment (Na2SO4-60%V2O5), the hot corrosion behaviour of Super-alloy IN-657 at 727 °C has been investigated byOtero et al. [93,94]. They reported that corrosionrate of the alloy in contact with molten salt mix-tures was reduced by approximately one order ofmagnitude over exposure times of 210 hr whenthe amount of molten salt was kept constant. Dur-ing the initial stages of the exposures the corro-sion rate increased with the temperature up to 1273°C and decreased for still higher temperature. Af-ter 100 hr of exposure, it was observed that thetemperature scarcely influenced the corrosion ki-netics.

Swaminathan and Raghvan [95] reported thatcracking and fluxing of the protective scales to-gether with easier crack nucleation and growth atgrain boundaries in the presence of liquid depositsof sodium metavanadate and sodiummetavanadate plus 15 wt.% sodium sulphate at650-750 °C accounted for the enhanced creeprates and reduced rupture life for Superni-600.During a similar study on Superni-C 276 in the sametemperature range, they have reported that eutec-tic NaVO3 + 15 wt.% Na2SO4 mixture was moresevere in degrading the creep properties(Swaminathan and Raghavan, [96]). The corrosiveswere found to reduce the rupture life through en-hanced creep deformation. They further opined thatthe degradation of the alloy in case of NaVO3 meltwas due to the cracking of the protective scalesunder the influence of the applied stress and thefluxing action of the melt. The addition of sodiumsulphate to sodium metavanadate increased thecorrosivity of the deposit by lowering the meltingpoint and by formation of molten Ni-Ni3S2 eutecticwhich initiates a selfsustaining hot corrosion. It wassuggested that the additional presence of molyb-denum compounds Na2MoO4-MoO3 caused en-hanced degradation.

Oxidation and hot corrosion in sulphate, chlo-ride and vanadate environment of a cast nickelbase superalloy have been reported by Deb et al.[97]. Weight gain studies were carried out in air forthe uncoated samples and coated with 100%Na2SO4, 75% Na2SO4 + 25% NaCl, and 60%Na2SO4 + 30% NaVO3 + 10% NaCl. The presenceof sulphur in the form of sulphates has been re-ported to cause internal sulphidation of the alloybeneath the external oxide layer. They observedthe formation of volatile species by chlorides which


Fig. 6. Schematic diagram showing probable hot corrosion mechanism in Na2SO4-60%V2O5 after expo-sure for 50 cycles at 900 °C for alloys (data from [100]) (a) Superni 75 (b) Superni 601.

further led to formation of voids and pits at grainboundaries. These voids and pits were suggestedto provide an easy path for flow of corrodents. Itwas proposed that the presence of vanadate inconjunction with sulphate and chloride providedadditional fluxing action, which destroyed the in-tegrity of the alloy and weakened its mechanicalproperties.

Gurrappa [98] during hot corrosion studies onCM 247 LC Superalloy in Na2SO4 andNa2SO4+NaCl mixtures at 900 °C observed thatthe superalloy CM 247 LC got severely corrodedin just 4 hr, while it was completely consumed in70 hr when tested in 90%Na2SO4+10%NaCl at 900°C. The life of the superalloy, however, further de-creased to just 2 hours in 90%Na2SO4+5%NaCl+5%V2O5 at 900 °C.

Lee and Lin [99] carried out studies on the mol-ten salt-induced oxidation/sulphidation on nickelaluminide intermetallic compound (Ni3Al) in the1%SO2/air gas mixtures. Their observations on hot-corroded specimen tested at different time periods

showed the formation of only NiO at 605 °C, andNiO and NiAl2O4 at 800 and 1000 °C. EDAX analy-sis revealed the presence of AlSx and/or NiSx be-neath the oxide scale at all the temperatures. Basedon the experimental studies, a hot corrosion mecha-nism was proposed by the investigators, accord-ing to which NiO oxide formation consumed theoxygen in the molten salt, which locally reducedthe oxygen and increased the sulphur partial pres-sure in the molten salt. Stability diagram was usedto represent this partial pressure change of sul-phur. They suggested that as the increased partialpressure of the sulphur reaches the equilibriumpartial pressure region of NiSx and/or AlSx, NiSx,and/or AlSx will form at the salt/alloy interfacethrough sulphidation reaction. The consumption ofsulphur will balance out the sulphur, and oxygenpartial pressure increases in the molten salt. Thiswill force NiO to form. This process confirmed thesimultaneous formation of NiO, NiSx, and/or AlSx.They further opined that as the sulphides are ther-modynamically unstable when the oxygen partial


pressure increases, it is possible for sulphide toconvert into oxides NiO, Al2O3, and NiAl2O4 throughnecessary reactions. Two possibilities have beenpointed out for the formation of spinel phase, ei-ther through the reaction of Al and Ni with oxygenin the molten salt or through the evolution ofsulphides.

Gitanjaly [99] has conducted hot corrosion stud-ies on some Ni-, Fe-, and Co-based superalloys inan environment of Na2SO4+60%V2O5 at 900 °C.Whilst the observed corrosion rate was significantin all the superalloys, Ni-base superalloy Superni75 showed the lowest rate of corrosion and theCo-base superalloy Superco-605 the highest. Bet-ter corrosion resistance of Ni-base superalloys wasattributed to the presence of refractory nickel vana-date Ni(VO3)2 which acted as a diffusion barrier forthe oxidizing species. The proposed hot corrosionmechanisms of this study for the superalloysSuperni 75 and Superni 601 have been schemati-cally shown in Fig. 6.

Tzvetkoff and Gencheva [101] presented a re-view on the mechanism of formation of corrosionlayers on nickel and nickel-based alloys in meltscontaining oxyanions. The molten sulphate mix-tures are stated to be particularly aggressive to-wards Ni superalloys. In such environments, theformation of Cr-rich passive films such as spinel-type oxides could be protective to some extent.They further revealed that at metal/oxide interface,sulphides are detected which do not offer plausibleprotection except for a possible positive role ofMoS2 formed on Ni alloys containing significantamounts of Mo. Furthermore, acidic oxides suchas those of V and Mo have been reported to in-duce rapid fluxing of the oxide scale and thereforecatastrophic hot corrosion. Whereas, thechromates were stated to be beneficial for the re-passivation of the surfaces following fluxing of theoxide scale by the molten salt.

6. CONCLUDING REMARKSHot corrosion has been identified as a serious prob-lem in high temperature applications such as inboilers, gas turbines, waste incinerations, dieselengines, coal gasification plants, chemical plants,and other energy generation systems. It is basi-cally induced by the impurities such as Na, V, S,etc. present in the coal or in fuel oil used for com-bustion in the mentioned applications. In some situ-ations, these impurities may be inhaled from theworking environment, for instance NaCl in marineatmospheres. There is a general agreement that

condensed alkali metal salts notably, Na2SO4, area prerequisite to hot corrosion [4]. Due to the highcost of removing these impurities, the use of theselow grade fuels is usually justified.

Although natural gas has also been consideredas an alternative to these traditional fuels for gen-eration of electricity in combined cycle power plants,still there remains strong interest in the develop-ment of clean, high efficiency coal-based systemsfor power generation (Saunders et al., [102]). Thismight be due to anticipated saving of high gradeenergy such as natural gas. Similarly, in coal con-version processes, which are known to produceclean environmentally acceptable fuels, technicalproblems arise from the interaction of sulphur andash in the coal with plant construction materials athigh temperatures and pressures (Natesan, [103]).

For the reason that hot corrosion of the compo-nents in aforesaid high temperature environmentsis inevitable, the phenomenon has maintained itsrelevance from the last more than 60 years. Hotcorrosion often increases the corrosion loss of heatresisting alloys by over hundred times, in compari-son to that incurred by simple oxidation (Shinata etal., [104]).

The use of Ni-, Fe- and Co- based superalloysin the high temperature applications such as gasturbines, boilers, etc. is well known and many moreapplications are still to be explored, where thesealloys may have tremendous potential. Althoughthe superalloys have adequate mechanical strengthfor such high temperature applications, yet they areprone to degradation by high temperature oxida-tion/corrosion during long term exposures. There-fore, the superalloys need to be protected, how-ever the protection system must be practical, reli-able and economically viable.

ACKNOWLEDGEMENTHarpreet Singh et al. thankfully acknowledge theresearch grant under Career Award for YoungTeachers (CAYT) from All India Council for Tech-nical Education (AICTE), New Delhi for carryingout this R&D work on “Role of Post-Coating Treat-ments on the Erosion-Corrosion Performance ofthe Thermal Spray Coatings” vide F.No. 1-51/FD/CA/(17)/2006-07, dated 12/12/2006.

REFERENCES[1] R. A. Rapp and Y. S. Zhang // JOM 46 (1994)

47.[2] R. A. Rapp // Corros. 42 (1986) 568.


[3] R. A. Rapp // Corros. Sci. (2002) 44 209.[4] A.M. Beltran and D.A. Shores, In: The

Superalloys, ed. by C.T. Sims and W.C.Hagel, (Wiley Publ., John Wiley and Sons, N. Y., 1972), Ch. 11: Hot Corrosion.

[5] M. M. Barbooti, S. H. Al-Madfai and H. J.Nassouri // Thermochimica Acta 126 (1988)43.

[6] K. L. Luthra and H.S. Spacil // J. Electrochem.Soc. 129 (1982) 649.

[7] K. Natesan // Corros. 32 (1976) 364.[8] N. Eliaz, G. Shemesh and R.M. Latanision //

Eng. Fail. Anal. 9 (2002) 31.[9] M.A. Uusitalo, P.M.J. Vuoristo and T.A.

Mantyla // Mater. Sci. Eng. A-Struct. 346(2003) 168.

[10] Metals Handbook, Failure analysis andPrevention, Vol.10 (ASM Publication, MetalsPark OH, USA, 1975).

[11] S. Prakash, S. Singh, B. S. Sidhu andA. Madeshia, In: Proc. National Seminar onAdvances in Material and Processing, Nov.,9-10 (2001) (IITR, Roorkee, India,2001) 245.

[12] J. A. Goebel, F. S. Pettit and G. W. Goward// Metall. Trans. 4 (1973) 261.

[13] J.R. Nicholls // JOM 52 (2000) 28.[14] I.G. Wright, In: Metals Handbook, Vol. 13,

9th Ed. (Metals Park, ASM, 1987) 97.[15]. F. S. Pettit and G. H. Meier, In: Superalloys

85, ed. by M. Gell, C. S. Kartovich, R. H.Bricknel, W. B. Kent and J. F. Radovich(Met. Soc. of AIME, Warrendale,Pensylvania, 1985) p. 651.

[16] F. S. Pettit and C. S. Giggins, Hot Corrosion,Ch. 12, In: Superalloys II, ed. by C. T. Sims,N. S. Stollof and W. C. Hagel (Pub. WileyPub., N. Y., 1987).

[17] M. Yoshiba // Corros. Sci. 35 (1993) 1115.[18] J. Stringer // Ann. Rev. Mater. Sci. 7 (1977)

477.[19] N. S. Bornstein and M.A. DeCrescente //

Met. Soc. AIME-Trans. 245 (1969) 1947.[20] N. S. Bornstein and M.A. DeCrescente //

Metall. Trans. 2 (1971) 2875.[21] N.S. Bornstein, M.A. DeCrescente and H.A.

Roth // Metall. Trans. 4 (1973) 1799.[22] J. A. Goebel and F. S. Pettit // Metall. Trans.

1 (1970) 1943.[23] R. A. Rapp and K. S. Goto In: Proc. of

Sympos. Fused Salts, ed. by J. Braunstein,and J. R. Selman (The Electrochem. Soc.,Pennington, N. J., 1981) p.159.

[24] J. A. Goebel and F. S. Pettit // Metall Trans 1(1970) 3421.

[25] J. Stringer // Mater. Sci. Technol. 3 (1986)482.

[26] N. Otsuka and R.A. Rapp // J. Electrochem.Soc. 137 (1990) 53.

[27] J.S. Zhang, Z.Q. Hu, Y. Murata, M. Morinagaand N. Yukawa // Metall. Trans. A 24 (1993)2451.

[28] Y.S. Zhang and R.A. Rapp // Corros. 43(1987) 348.

[29] K. L. Luthra, In: High TemperatureCorrosion, ed. by R.A. Rapp (Houston,NACE, 1983) p. 507.

[30] D. N. H. Trafford and D. P. Whittle // Corros.Sci. 20 (1980) 497.

[31] D. N. H. Trafford and D. P. Whittle // Corros.Sci. 20 (1980) 509.

[32] P. Nanni, V. Buscaglia, C. D. Asmundis andS. K. Roy, In: Proc. of 10th ICMC (Madras,India, 1987) Vol. IV, p. 3413.

[33] F. Gesmundo and F. Viani // Mater. Chem.Phys. 20 (1988) 513.

[34] L. Shi // Oxid. Met. 40 (1993) 197.[35] L. Shi, Y. Zhang and S. Shih // Corros. Sci.

33 (1993) 1427.[36] L. Shi // Corros. Sci. 37 (1995) 1281.[37] L. Li, R. Zhu and F. Gesmundo // J. Mater.

Sci. Technol. 12 (1996) 445.[38] D. Das, R. Balasubramaniam and M. N.

Mungole // J. Mater. Sci. 37 (2002) 1135.[39] L. J. Coze, U. Franzoni, O. Cayla,

A. Devisme and A. Lefort // Mater. Sci. Eng.A-Struct. 120 (1989) 293.

[40] C. C. Tsaur, J. C. Rock, C. J. Wang andY. H. Su // Mater. Chem. Phys. 89 (2005)445.

[41] A.K. Misra // Corros. NACE 43 (1987) 440.[42] M. Hara, T. Hisaichi, K. Itoh and Y. Shinata //

J. Jpn. I. Met. 55 (1991) 1207.[43] D. M. Johnson, D. P. Whittle and J. Stringer

// Oxid. Met. 12 (1978) 273.[44] K. L. Luthra and D.A. Shores //

J. Electrochem. Soc. 127 (1980) 2202.[45] N. S. Bornstein and M.A. DeCrescente //

Corros. 26 (1970) 209.[46] A. K. Misra // J. Electrochem. Soc. 133

(1986) 1029.[47] Y. Bourhis and C. St. John // Oxid. Met. 9

(1975) 507.[48] K. R. Peters, D. P. Whittle and J. Stringer //

Corros. Sci. 16 (1976) 791.


[49] G. C. Fryburg, F. J. Kohl and C. A. Stearns //J. Electrochem. Soc. 131 (1984) 2985.

[50] G. C. Fryburg, F. J. Kohl, C. A. Stearns andW. L. Fielder // J. Electrochem. Soc. 129(1982) 571.

[51] A. Pehkonen, M. H. Tikkanen, S. Ylasaariand O. Forsen, In: Proc. of 10th ICMC,(Madras, India,1987) Vol. IV, p. 3781.

[52] K. Godlewski and E. Godlewska // Mater.Sci. Eng. 88 (1987) 103.

[53] Y. Zho, R. Zhu and M. Guo // Corros. NACE43 (1987) 51.

[54] R. Santorelli, E. Sivieri and R.C. Reggiani //Mater. Sci. Eng. AStruct. 120 (1989) 283.

[55] S. Shih, Y. Zhang and X. Li // Mater. Sci.Eng. A-Struct. 120 (1989) 277.

[56] G. J. Santoro // Oxid. Met. 13 (1979) 405.[57] C.A.C. Sequeira and M.G. Hocking //

Corros. 37 (1981) 392.[58] A. U. Malik and S. Ahmad // Metallkd. (1983)

74 819.[59] A.U. Malik, N. Asrar, S. Ahmad and N. A. B.

Siddiqi // Metallkd. 79 (1988) 285.[60] M. Levy, R. Huie and F. Pettit // Corros.

(1989) 45 661.[61] S. Zhao, X. Xie and G.D. Smith // Surf. Coat.

Technol. 185 (2004) 178.[62] K. L. Luthra // Metall. Trans. A 13A (1982)

1843.[63] K. L. Luthra // J. Electrochem. Soc. 132

(1985) 1293.[64] A. U. Malik and M. Mobin, In: Proc. of 10th

ICMC (Madras, India, 1987) Vol. IV, p. 3345.[65] M. Mobin, A. U. Malik, S. Ahmad, S. K.

Hasan and M. Ajmal // Bull. Mater. Sci. 19(1996) 807.

[66] G.T. Harris, H.C. Child and J.A. Kerr // ISIJInt. 179 (1955) 342.

[67] K. Sachs // Metallurgia Apr (1958) 167.[68] L. Fairman // Corros. Sci. 2 (1962) 293.[69] D.A. Pantony and K. I. Vasu // J. Inorg. Nucl.

Chem. 10 (1968) 423.[70] D.A. Pantony and K.I. Vasu // J. Inorg. Nucl.

Chem. 10 (1968) 755.[71] R. C. Kerby and J. R. Wilson // Trans. ASME

95 Ser. A (1973) 36.[72] R. B. Dooley and J. R. Wilson // Trans.

ASME 97 Ser. A (1975) 422.[73] P. Hancock // Mater. Sci. Technol. 3 (1987)

536.[74] S. R. Iyer, K. J. L. Iyer and V. M.

Radhakrishan, In: Proc. of 10th ICMC(Madras, India, 1987) Vol. IV, p. 3665.

[75] J. Swaminathan, S. Raghavan and S. R. Iyer// T. Indian I. Metals 46 (1993) 175.

[76] G. A. Kolta, L. F. Hewaidy and N. S. Felix //Thermochim. Acta 4 (1972) 151.

[77] P. Kofstad, In: High Temperature Corrosion,Chapter 14 (Elsevier, Applied Science,London & New York, 1988) p. 465.

[78] C.J. Valdes, R.B. Dooley and J.R. Wilson,The Corrosion of A.I.S.I 446 Stainless Steelin Molten Vanadates in the TemperatureRange 700-900 °C (Report DefenceResearch Board Canada, Grant No. 7535-14, 1973).

[79] S. N. Tiwari and S. Prakash, In: Proc. ofSympos. Metals and Materials Research(Indian Institute of Technology, Madras,1996) p. 107.

[80] S. N. Tiwari and S. Prakash, In: Proc. ofSOLCEC (Kalpakkam, India, 1997) PaperC33.

[81] S. N. Tiwari, Investigations on Hot Corrosionof Some Fe-, Ni- and Co-Base Superalloy inNa2SO4-V2O5 Environment under CyclicConditions (Ph. D.Thesis, Met. Mat. Engg.Dept., University of Roorkee, Roorkee, India,1997).

[82] C. F. Almeraya, V. A. Martinez, C. Gaona,M.A. Romero and J.M. Malo // Revista deMetalurgia 34 (1998) 11.

[83] C. F. Almeraya, V. A. Martinez and R. J. G.Gonzalez // Brit. Corros. J. 33 (1998) 288.

[84] C. Cuevas-Arteaga, J. Porcayo-Calderon,G. Izquierdo, A. Martinez-Villafane and J.G.Gonzalez-Rodriguez // Mater. Sci. Technol.17 (2001) 880.

[85] B. Singh, Studies on the Role of Coatings inImproving Resistance to Hot Corrosion andDegradation (Ph.D. Thesis, Met. & Mat. Eng.Dept., Indian Institute of TechnologyRoorkee, Roorkee, 2003).

[86] H. Singh, Hot Corrosion Studies on PlasmaSpray Coatings over Some Ni-and Fe-basedSuperalloys (Ph.D. Thesis, Met. & Mat. Eng.Dept., Indian Institute of TechnologyRoorkee, Roorkee, 2005).

[87] H.R. Thilakan, A.K. Lahiri and T. Banerjee //NML Technical Journal May (1967) 20.

[88] N. S. Bornstein, M. A. Decrescente andH. A. Roth, In: Proc. Conf. on Gas TurbineMater. in the Marine Environment, MMIC-75-27 (Columbus, Ohio, USA, 1975) 115.

[89] M. Seiersten and P. Kofstad // High Temp.Technol. 5 (1987) 115.


[90] P.S. Sidky and M.G. Hocking // Corros. Sci.27 (1987) 499.

[91] E. Otero, M. C. Merino, A. Pardo, M. V.Biezma and G. Buitrago, In: Proc. of 10th

ICMC (Madras, India, 1987) Vol. IV, p. 3583.[92] P. Lambert, B. Champagne and

B. Arseneault // Can. Metall. Quart. 30(1991) 125.

[93] E. Otero, A. Pardo, J. Hernaez and F. J.Perez // Corros. Sci. 30 (1990) 677.

[94] E. Otero, A. Pardo, J. Hernaez and F. J.Perez // Corros. Sci. 34 (1992) 1747.

[95] J. Swaminathan and S. Raghavan // Mater.High Temp. 10 (1992) 242.

[96] J. Swaminathan and S. Raghavan // HighTemp. Mater. Processes 13 (1994) 277.

[97] D. Deb, S. R. Iyer and V. M. Radhakrishnan// Mater. Lett. 29 (1996) 19.

[98] I. Gurrappa // Oxid. Met. 51 (1999) 353.[99] W. H. Lee and R. Y. Lin // Mater. Chem.

Phys. 77 (2002) 86.[100] Gitanjaly, Role of Inhibitors on Hot

Corrosion of Superalloys in Na2SO4-V2O5Environment (Ph.D. Thesis, Met. Mat. Engg.Dept., Indian Institute of TechnologyRoorkee, Roorkee, India, 2003).

[101] T. Tzvetkoff and P. Gencheva // Mater.Chem. Phys. 82 (2003) 897.

[102] S. R. J. Saunders, D. D. Gohil, J. P. Banks,M. U. Sheriff, P. F. Tortorelli, J. H. D. Vanand I. G. Wright // Mater. Sci. Forum 251-254 (1997) 583.

[103] K. Natesan // Corros. 41 (1985) 646.[104] Y. Shinata, F. Takahashi and K. Hashiura //

Mater. Sci. Eng. 87 (1987) 399.

AN OVERVIEW OF Na SO AND/OR V O INDUCED HOT CORROSION OF Fe- AND Ni-BASED SUPERALLOYS · ·...

Documents

Transcript of AN OVERVIEW OF Na SO AND/OR V O INDUCED HOT CORROSION OF Fe- AND Ni-BASED SUPERALLOYS · ·...