Hot corrosion of carbon-alloyed Fe3Al-based iron aluminides

Debashis Das, R. Balasubramaniam *, M.N. Mungole

Department of Materials and Metallurgical Engineering, Indian Institute of Technology, Kanpur 208016, India

Received 22 March 2001; received in revised form 2 October 2001

Abstract

The oxidation and hot corrosion behavior of two Fe3Al-based iron aluminides, Fe�/25Al and Fe�/27.5Al�/3.7C (at.%) have been

studied at 1100, 1225 and 1330 K. Hot corrosion studies were conducted by coating the specimen surfaces with 2.59/0.2 mg cm�2 of

Na2SO4 prior to exposure in pure oxygen. The oxidation kinetics of the carbon-alloyed iron aluminide were generally slower than

that of the binary alloy. Alumina was identified in the scale after oxidation of both the alloys. The rates of hot corrosion were

generally higher than the rates of oxidation for both the alloys. The presence of a-Fe2O3 in addition to alumina was indicated by X-

ray diffraction analysis of the scales present on the surface of the samples after hot corrosion. Fourier transform infrared spectra

from the spalled scales in hot corrosion divulged the presence of a-Al2O3, a-Fe2O3 and sulfate. Cross-sectional microscopy revealed

that the scale�/metal interfaces were pitted under hot corrosion conditions and the pits contained aluminum sulfide. Aluminum

sulfide was also identified along the grain boundaries in the binary aluminide matrix below the scale�/metal interface. The hot

corrosion process has been explained based on sulfide formation and its subsequent oxidation. The lower rate of hot corrosion in the

carbon-alloyed iron aluminide has been related to the blocking effect of carbides, present along the grain boundaries, for the

penetrating sulfur. # 2002 Elsevier Science B.V. All rights reserved.

Keywords: Carbon-alloyed iron aluminide; Oxidation; Hot corrosion; Scale characterization; Sulfide formation

1. Introduction

Iron aluminides, based around the compositions

Fe3Al and FeAl, are candidate high temperature mate-

rials. Iron aluminides generally contain very low (0.01

wt.%) carbon because carbon is known to embrittle

these alloys. Recently, Baligidad et al. have reported

that addition of carbon in the range of 0.6�/2.0% (all

compositions henceforth in atom percent) significantly

increased the room temperature strength of Fe�/28Al

alloys [1]. The increase in room temperature yield

strength was attributed to solid solution strengthening

by the interstitial carbon, as well as precipitation

hardening due to the presence of Fe3AlC0.5 precipitates

[2]. The present study will address the high temperature

oxidation and hot corrosion behavior of a binary and a

carbon-alloyed iron aluminide in oxygen. Aluminum

levels in the Fe3Al-based aluminides are well in excess of

the critical concentration and alumina forms readily

above 500 8C on exposure to oxidizing environments

[3,4]. The oxidation behavior of binary iron aluminides

and the nature of corrosion products have been

reviewed elsewhere [5].

The oxidation behavior of carbon-alloyed iron alu-

minides has not been studied. Earlier studies have

addressed the oxidation of iron alloys containing about

9�/18% Al with carbon addition. The oxidation behavior

of a series of Fe�/Al alloys containing up to 16% Al and

up to 0.4% C have been investigated between 450 and

900 8C [6]. Perforation of the protective alumina oxide

scale resulted in the formation of scattered nodules of

iron oxides, which increased with increasing carbon

content [6]. Kao and Wan [7,8] have studied the

oxidation of two Fe�/Al alloys (Fe�/10.6Al�/2.4C and

Fe�/14.1Al�/2.7C). They did not specifically address the

effect of carbon on oxidation rates.

Few studies have reported the hot corrosion of iron

aluminides using Na2SO4 melts. However, sulphidation

behavior of iron aluminides has been reported. The

presence of molten alkali sulfate salts significantly

increased the corrosion of iron aluminides in SO2-* Corresponding author.

E-mail address: bala@iitk.ac.in (R. Balasubramaniam).

Materials Science and Engineering A338 (2002) 24�/32

www.elsevier.com/locate/msea

0921-5093/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.

PII: S 0 9 2 1 - 5 0 9 3 ( 0 2 ) 0 0 0 7 2 - 2

containing mixed gases [9]. A coating of Na2SO4�/

Li2SO4 on iron aluminides exposed to an oxidizing/

sulphidizing gaseous environment (1% SO2 in air) at 605

and 800 8C resulted in corrosion rates that were at leastten times higher than rates measured in the absence of

the sulfate coating [9]. The degradation by the molten

sulfate decreased with increasing Cr contents (2�/5%)

and increasing Al contents (28�/36%). Stainless steels

(310 and 321) possessed significantly better hot corro-

sion resistance than iron aluminides [9]. In another study

by Gesmundo et al. [10], both Fe3Al (27Al�/2.2Cr�/0.1B)

and FeAl (40Al�/0.05Zr�/0.06B�/0.085C) alloys werecoated with Na2SO4-containing salts and exposed to a

simulated combustion gas at 600 8C. Recently, Rodri-

guez et al. [11] studied the hot corrosion of Fe�/40Al,

Fe�/40Al�/0.1B and Fe�/40Al�/0.1B�/10Al2O3 alloys in

molten NaVO3 at 625 and 700 8C by potentiodynamic

polarization. Both these studies indicated that the sur-

face alumina scale was attacked by the sulfate-contain-

ing salts, resulting in enhanced degradation and inoxidation of the other major element in the alloys, i.e.

Fe.

2. Experimental

The Fe�/25.3Al intermetallic (which will henceforth be

called NC) and the carbon-alloyed iron aluminide of

composition Fe�/27.5Al�/3.7C (which will be called 3C)

were obtained from the Defense Metallurgical Research

Laboratory (DMRL), Hyderabad. The microstructure

of the as-received 3C sample (Fig. 1) revealed bulky

Fe3AlC0.5 precipitates in the matrix and finer carbideprecipitates along grain boundaries. Rectangular speci-

mens were sectioned, mechanically polished to 600-grit

and degreased using acetone and alcohol. Thermogravi-

metric technique was employed for kinetic measure-

ments. The apparatus consisted of a vertical furnace, a

Mettler single pan analytical balance, and gas train. A

vertical furnace of 250 mm length was employed toconduct the oxidation and hot corrosion tests. A mullite

tube (45 mm inner diameter and 460 mm length) acted

as the reaction chamber. The specimen was placed inside

a quartz crucible constructed with three holes at the

bottom of the crucible to allow for passage of gas. The

crucible was 15 mm in diameter and 20 mm in length.

The quartz crucible was suspended from the top of the

furnace using a platinum wire into the reaction zone ofthe reaction chamber. Pure oxygen gas was passed

initially through a bubbler and capillary flow meter,

and then through Ascarite, anhydrous calcium chloride

and Drierite (CaSO4) columns successively before in-

troduction into the reaction chamber. The outlet gas

was passed through a bubbler to ensure that the flow of

gas was maintained through the system.

Oxidation and hot corrosion experiments were carriedout isothermally at temperatures of 1100, 1225 and 1330

K. The gas flow rate was maintained constant at 0.2

cm3 s�1 (STP). The quartz crucible was periodically

removed from the furnace, the weight of the sample

recorded and the crucible again reintroduced into the

furnace. For the hot corrosion experiments, the speci-

mens were initially coated with a thin film of Na2SO4

and then exposed to the environment at the desiredtemperature. The salt deposit was applied to the warm

(�/150 8C) specimen by a brush, to give a uniform coat

of the aqueous solution of Na2SO4 on the surface of the

specimen. A surface coverage of 2.5 mg cm�2 of salt

was used [12]. The kinetics of hot corrosion were

monitored by measuring the weight changes as a

function of time, similar to the oxidation experiments.

There was noticeable scale spallation at the two highertemperatures. The spalled scales were collected and

weighed with the specimen.

The corrosion products were visually observed to

record scale color, adherence and uniformity. X-ray

diffraction (XRD) patterns were obtained from the

surface scales with a Rich�/Seifert 2002D diffractometer

using Cu Ka radiation. A JEOL JSM 840A scanning

electron microscope (SEM) was employed for topologi-cal observation of the surface. A JEOL JXA-8600MX

electron probe micro-analyser (EPMA) was utilized for

qualitative compositional analyses. The cross-sections

were studied after electroless nickel plating the surfaces

and mounting the cross-sections in epoxy resin. The

mounted specimens were polished and etched with

HNO3�/CH3COOH�/H2O�/HF (33:33:33:1) before

observation of the cross-section. The oxidation andhot corrosion products from select experiments were

analysed by Fourier transform infrared (FTIR) spectro-

scopy after pressing them into pellets using spectro-

scopically pure KBr. The FTIR spectra were recorded at

Fig. 1. SEM micrograph of the as-received 3C sample. Bulky

Fe3AlC0.5 precipitates in the matrix and finer carbide precipitates

along grain boundaries can be seen.

D. Das et al. / Materials Science and Engineering A338 (2002) 24�/32 25

room temperature using a Nicolet Magna 750 Series 2,

FTIR system.

3. Results

3.1. Kinetics

In order to analyze the kinetic data, parabolic growth

behavior was assumed and the parabolic rate constantkp was obtained from the slope of the linear regression

fitted line of (DW /A )2 vs. t plot. Fig. 2 shows the nature

of fit of parabolic rate law for the oxidation experi-

ments, while Fig. 3 and Fig. 4 show the same for the hot

corrosion experiments for alloys NC and 3C, respec-

tively. Ideally, parabolic growth would yield straight

lines on such plots, with the slope being equal to the rate

constant kp. It may be noticed from Figs. 1�/3 that theobedience to parabolic rate law was only approximate.

This is due to the discontinuous method of recording

weight gain. The exposure and alloy data have been

summarized in the Fig. 5 by comparing the weight gain

recorded at 234 k s, for all the samples, as a function of

temperature.

3.2. Scale characterization

Visual observation of the scales revealed that the

color of scales on the NC specimens oxidized at 1225and 1330 K was cream white, while the scale after

oxidation at 1100 K was dull white. The scales on the

oxidized 3C specimens were mostly gray in color. The

spalled scales in the oxidation experiments were in the

form of loose fine powders of white color. Severe scale

spalling was noted for the hot corroded samples, mainly

at the two higher temperatures (1330 and 1225 K). The

color of the spalled scale was deep brown. The salt did

not melt at 1100 K and spalled as a loose dry mass. A

white layer of loose scales in all the cases covered the

sample surface (from where the scale had spalled off).

Analysis of the XRD patterns of all the specimens

after oxidation revealed that u-Al2O3 was the major

phase at the lowest temperature of oxidation. At the

intermediate temperature, a-Al2O3 was present in addi-

tion to u-Al2O3, with the former being the major phase.

At the highest temperature, the major constituent of the

scale was a-Al2O3. In the hot corrosion specimens, the

types of Al2O3 phases observed were similar to that

observed after the oxidation experiments. In addition,

peaks corresponding to a-Fe2O3 were identifiable in the

scales of both the alloys after hot corrosion at 1225 and

1330 K. FTIR spectra from the scales of the oxidized

specimens confirmed the presence of alumina, an

example of which is provided in Fig. 6a. The spectra

from the spalled scales after hot corrosion at 1225 and

1330 K indicated the presence of a-Fe2O3 and sulfate in

addition to alumina, for both the alloys. An example of

the same is provided in Fig. 6b. In this figure, the

presence of a-Al2O3 is indicated by the strong peaks at

575�/600 and 450�/432 cm�1 [13], while the presence of

a-Fe2O3 is indicated by the strong peaks at 550�/560 and

474�/467 cm�1 [13]. The presence of sulfate is indicated

by the doublet in the region 1023�/1146 cm�1 [13]. The

scale characterization results are summarized in Table 1.

Scale morphologies were studied by scanning electron

microscopy. In the case of oxidation, the scale on the

NC alloy after oxidation at 1100 K revealed fine faceted

oxides. At higher temperatures, a ridge-like morphology

developed. The development of ridge-like morphology

could be also discerned at 1225 and 1330 K for the 3C

Fig. 2. (DW /A )2 vs. t plots for oxidation of NC and 3C. The lines joining the data points are for visual aid only.

D. Das et al. / Materials Science and Engineering A338 (2002) 24�/3226

alloy. Generally, the oxide covered the surface comple-

tely in the case of the NC alloy compared to the 3C

alloy. In the case of hot corrosion, the NC specimens

were generally covered with uniform scales compared to

the 3C samples. In the NC specimens, nodular features

were observed on the surface at 1100 K, while a ridge-

like morphology could be discerned at 1225 K. At 1330

K, the surface exhibited large nodular features (Fig. 7)

surrounded by clusters of fine needle-like whiskers.

Qualitative EPMA analysis of the whiskers indicated

that it was rich in aluminum and oxygen, thereby

denoting that it was alumina. Therefore, the morphol-

ogy of alumina formed after hot corrosion was different

from that after oxidation at 1330 K. In the 3C samples,

significant scale spallation was observed at higher

temperatures. The surfaces were not completely covered

with the scales and some of the uncovered areas could be

related to the presence of carbides, in both oxidation

and hot corrosion.

4. Discussion

4.1. Oxidation

The difference in the oxidation behavior between un-

doped Fe3Al and Fe3Al�/3.7C was not significant. As

both the alloys contained sufficient Al for formation of

Fig. 3. (DW /A )2 vs. t plots for hot corrosion of NC. The lines joining the data points are for visual aid only.

Fig. 4. (DW /A )2 vs. t plots for hot corrosion of 3C. The lines joining the data points are for visual aid only.

D. Das et al. / Materials Science and Engineering A338 (2002) 24�/32 27

a complete external layer of Al2O3, the external scales on

these intermetallics were alumina (Table 1). This is also

in conformity with published literature on oxidation ofiron aluminides [3�/5]. Therefore, for the conditions

when a complete external scale of alumina forms on the

surface, the effect of carbides on the oxidation behavior

is not significant. In the reported cases of oxidation of

Fe�/Al�/C alloys [6�/8], the aluminum contents were

lower (B/16%) in these alloys and complete alumina

formation was not observed. Localized oxidation of iron

through the protective aluminum oxide scale resulted inthe appearance of corn-like nodules of iron oxide phases

Fe2O3, Fe3O4 and FeAl2O4 [6]. Similarly, oxide nodules

containing mainly FeO, Fe2O3, Fe3O4, FeAl2O4 and

Al2O3 were observed on the surface of oxidized Fe�/

10.6Al�/2.4C and Fe�/14.1Al�/2.7C alloys [7,8]. The

preferential oxidation of Fe in these Fe�/Al�/C alloys

provided higher oxidation rates compared to the Fe�/Al

alloys without carbon.

The cross-section of the 3C specimen oxidized at 1330K was observed (Fig. 8a). The electroless Ni coating can

be seen on top of the oxide layer in this figure. The oxide

layer was analyzed qualitatively in the EPMA and it was

composed of only alumina. There was no preferential

attack along the carbide�/matrix interfaces. The stable

nature of the carbide was also revealed in observations

at several other oxide-substrate locations in the same

sample. The oxidation resistance of the carbide in the

carbon-alloyed iron aluminide could result due to therelatively high Al content of the Fe3AlC0.5 carbide. In

contrast, the iron carbides in the Fe�/10.6Al�/2.4C and

Fe�/14.1Al�/2.7C alloys, especially those near the metal

surface, lost their stability above 800 8C [7,8]. This

resulted in decarburized regions between the oxide scales

and the matrix, and preferential oxidation from the ironcarbide particles exposed to the environment in the Fe�/

Al�/C alloys of lower Al content [7,8].

The parabolic rate constant for the initial stages of

oxidation was higher than in the later stages of oxida-

tion in case of the NC alloy. These two regions could be

distinguished on the parabolic plots of the weight gain

data (Fig. 2). It was earlier pointed by Rommerskirchen

et al. [14] that the kinetics of oxidation of ironaluminides could be analyzed as consisting of two

regions of parabolic behavior, with the two regions

corresponding to the formation of different kinds of

aluminas. For example, the faster initial oxidation

kinetics was related by Rommerskirchen et al. to

transition u-Al2O3 formation while the lower oxidation

kinetics in the later stages was related to a-Al2O3

formation. In order to gain further insights into thepossible relationship of the kinetics with the nature of

alumina formed, the kinetics of the two stages were

compared to a-Al2O3 and u-Al2O3 formation kinetics

[14�/16]. It was found that the initial rate could be

related to u-Al2O3 formation kinetics while the rate from

the later stages of oxidation could be related to a-Al2O3

formation kinetics [5].

4.2. Hot corrosion

The hot corrosion rates were higher than oxidation

rates for both the alloys, except in one case (3C at 1100K). Moreover, the rate of hot corrosion of the NC alloy

was higher than that of the carbon-alloyed intermetallic.

Topological observation of the surface scales after hot

Fig. 5. The variation of weight gain, recorded at 234 k s, as a function of temperature, providing a summary of the alloy and exposure data.

D. Das et al. / Materials Science and Engineering A338 (2002) 24�/3228

corrosion of these alloys indicated that fairly thick scales

covered the surfaces of the NC alloy compared to the 3C

alloy.

Cross-sectional microscopy revealed that the thick-

ness of the surface scale was much higher after hot

corrosion (Fig. 8b). The base metal was degraded in a

characteristic fashion at the scale�/metal interface, where

deep pits could be observed at the interface (Fig. 8b).

The carbide particles were not preferentially attacked.

The corrosion products in the pits at the scale�/metal

interface were analyzed qualitatively by EPMA and it

contained significant amounts of S in addition to Al and

Fe, and a small amount of O. The qualitative EPMA

results obtained from the scale in the same sample did

not reveal any S but only Al, Fe and O. When viewed

along with the XRD results, it can be concluded that the

external scales were composed of a-Fe2O3 and a-Al2O3.

The corrosion products within the pits were sulfides.

The identification of sulfur at the scale�/metal interface

showed that sulfur-bearing compounds are found in the

Fig. 6. FTIR spectra of the scale on NC alloy at 1330 K after (a)

oxidation and (b) hot corrosion.

Table 1

Summary of nature of scales observed on NC and 3C alloys after oxidation (OX) and hot corrosion (HC) experiments

Sample (experiment) Temperature (K) Phases identified by XRD FTIR identification Scale characteristics

Major Minor

NC (OX) 1100 u-Al2O3 �/ �/ Dull white

1225 a-Al2O3 u-Al2O3 �/ Cream white

1330 a-Al2O3 �/ g-FeOOH, a-Al2O3 Cream white

3C (OX) 1100 u-Al2O3 �/ �/ Dull gray

1225 a-Al2O3 u-Al2O3 �/ Gray

1330 a-Al2O3 �/ g-FeOOH, a-Al2O3 Bright gray

NC (HC) 1100 u-Al2O3 �/ �/ Dull brown

1225 a-Al2O3 u-Al2O3, a-Fe2O3 Na2SO4, a-Al2O3, a-Fe2O3 Brown

1330 a-Al2O3 a-Fe2O3 Na2SO4, a-Al2O3, a-Fe2O3 Brown

3C (HC) 1100 u-Al2O3, �/ �/ Dull brown

1225 a-Al2O3 u-Al2O3, a-Fe2O3 Na2SO4, a-Al2O3, a-Fe2O3 Brown

1330 a-Al2O3 a-Fe2O3 Na2SO4, a-Al2O3, a-Fe2O3 Brown

Fig. 7. SEM morphology of the scale on NC alloy after hot corrosion

for 65 h at 1330 K. The surface exhibited large nodular features

surrounded by clusters of fine needle-like whiskers. Qualitative analysis

of the whiskers indicated that they were alumina.

D. Das et al. / Materials Science and Engineering A338 (2002) 24�/32 29

pits below the scale. Interestingly, the inner scale near

the scale�/metal interface in the Fe�/27Al�/2.2Cr�/0.1B

alloys after Na2SO4 induced hot corrosion in a SO2-

containing environment at 600 8C revealed the presence

of Al2O3, Al2S3 and some Fe [10].

The kinetics of hot corrosion were faster in the case of

the NC alloy compared to the 3C alloy. Relatively thick

scales were observed in the NC alloy after hot corrosion

at 1330 K (Fig. 9a). The penetration of a corrosion zone

can be noted. Qualitative EPMA analysis of the corro-

sion product zone revealed that the product consisted

essentially of Fe2O3 and Al2O3. Similar penetrating

corrosion product zones were not consistently observed

on all surfaces. However, the region below the metal�/

scale interface was consistently pitted and these pits

contained corrosion products (Fig. 9b), similar to that

observed for the 3C alloy. Significant presence of

corrosion products was also constantly noticed along

grain boundaries near the scale�/metal interface (Fig.

9b). The corrosion products in the pits along the grain

boundary in Fig. 9b were analyzed qualitatively by

EPMA as aluminum sulfide.

Based on the microstructural and compositional

characterization of the 3C and NC alloys, the process

of hot corrosion can be understood. Hot corrosion of

both the alloys was similar. The protective alumina scale

can be fluxed by both the acidic (SO3) and basic (Na2O)

components of Na2SO4 [17]. The identification of sulfate

in the spalled scales by FTIR spectroscopy probably

indicates the significance of the acid fluxing mechanism

in the initial stages of hot corrosion. Accelerated

corrosion occurs only above the melting point of

Na2SO4 (above 884 8C) [17]. It has been reported that

at lower temperatures the salt deposit may act as a

barrier to oxidation and could reduce the reaction rate

compared to that of metal with no Na2SO4 deposit [18].

This was observed for 3C alloy at 1100 K in the present

Fig. 8. Cross-sectional micrographs of the alloy 3C at 1330 K after 65

h of (a) oxidation and (b) hot corrosion.

Fig. 9. Cross-sectional micrographs of the alloy NC after hot

corrosion at 1330 K for 65 h showing (a) localized attack, and (b)

internal sulfides in the metal below the scale�/metal interface.

D. Das et al. / Materials Science and Engineering A338 (2002) 24�/3230

study. The identification of Fe2O3 in the spalled scales of

the hot corrosion experiments indicated that non-

protective conditions were established when Na2SO4

melted on the surface and resulted in the oxidation ofboth Al and Fe. The results of the present study, that

both Al2O3 was Fe2O3 occurred as corrosion products

during hot corrosion, are in agreement with other hot

corrosion studies of iron aluminides [10,11]. Sustained,

accelerated hot corrosion induced by Na2SO4 appears to

be associated with sulfide formation at or near the

metal�/scale interface. The sulfide phases provide paths

for rapid outward diffusion of the metal. A notablefeature of attack by this mechanism is the formation of

pits at the metal�/scale interface [19]. The observation of

pits at the metal�/scale interface in the present case was

consistent with this mechanism. At the scale�/metal

interface, the low partial pressure of oxygen combined

with the high sulfur activity results in the formation of

aluminum sulfide at the interface. The oxidation of the

sulfide at the metal�/scale boundaries leads to furtherpenetration of the material by the sulfur, which has been

liberated by the oxidation reaction, and in this manner,

the attack of the material appears to be accelerated. It

has been verified by cross-sectional microscopy that the

attack of the material at the scale�/metal interface was

very significant and moreover, the diffusing S had

penetrated deep into the material (Fig. 9b). The

identification of aluminum sulfides in the pits suggestsprobably that sulphidation of Al results in non-protec-

tive scale formation because Al is the metal that is

required to form the protective oxide in iron aluminides.

The sulfur released after oxidation of the sulfides

diffuses inside the material, faster through the grain

boundaries, to cause further degradation.

The hot corrosion rates were lower for the 3C alloy.

The presence of stable carbides on the surface could beone reason. Secondly, the presence of carbide precipi-

tates along the grain boundaries (Fig. 1) would hinder

the diffusion of S deep into the material. Therefore, the

degradation of 3C alloy would proceed at a lower rate

compared to the NC alloy because it was seen that

accelerated attack occurred due to S penetration along

grain boundaries in the NC alloy.

5. Conclusions

The high temperature oxidation and hot corrosion

behavior at 1100, 1225 and 1330 K of a carbon-alloyed

iron aluminide Fe�/27.5Al�/3.7C (3C) was studied and

compared with that of Fe�/25.3Al (NC). The kinetics of

hot corrosion were generally faster than oxidation. The

external scales contained essentially Al2O3 after theoxidation experiments, whereas both Fe2O3 and Al2O3

were identified after hot corrosion experiments at higher

temperatures. Cross-sectional microstructural analysis

revealed pitting just below the scale�/metal interface and

enhanced attack along grain boundaries in the under-

lying metallic matrix. Qualitative compositional analysis

indicated the presence of aluminum sulfides in the pits atthe scale�/metal interface, and in the pits along the grain

boundaries in the metallic matrix below the scale�/metal

interface. The faster hot corrosion kinetics has been

attributed to the formation of sulfides at the scale�/metal

interface. The possible sequence of attack in hot

corrosion has been proposed based on microstructural

and compositional analyses. Fluxing of alumina results

in higher attack rates. The formation of aluminumsulfides results in non-protective scales. The oxidation of

sulfides releases sulfur, which again diffuses inward into

the material to cause further attack. The kinetics of hot

corrosion were lower in the 3C alloy because the

carbides present along grain boundaries in the NC alloy

hinder the diffusion of sulfur into the material.

Acknowledgements

The authors thank Drs D. Banerjee and R. Baligidad

of DMRL, Hyderabad for providing the specimens used

in the present study. The authors also thank Dr A.V.

Ramesh Kumar of the Defense Materials Stores Re-

search and Development Establishment, Kanpur for

performing the FTIR spectroscopic studies.

References

[1] R.G. Baligidad, U. Prakash, A. Radhakrishna, V. Ramakrishna

Rao, P.K. Rao, N.B. Ballal, Scipta Mater. 36 (1997) 667.

[2] R.G. Baligidad, U. Prakash, A. Radhakrishna, V. Ramakrishna

Rao, P.K. Rao, N.B. Ballal, Scipta Mater. 36 (1997) 105.

[3] R. Prescott, M.J. Graham, Oxid. Met. 38 (1992) 73.

[4] P.F. Tortorelli, J.H. DeVan, in: J.H. Schneibel, M.A. Crimp

(Eds.), Processing, Properties and Applications of Iron Alumi-

nides, the Minerals, Metals and Materials Society, Warrendale,

USA, 1994, p. 257.

[5] N. Babu, R. Balasubramaniam, A. Ghosh, Corrosion Sci. 43

(2001) 2239.

[6] W.E. Boggs, J. Electrochem. Soc. (1971) 906.

[7] C.H. Kao, C.M. Wan, J. Mater. Sci. 22 (1987) 3203.

[8] C.H. Kao, C.M. Wan, J. Mater. Sci. 23 (1988) 1943.

[9] W.H. Lee, R.Y. Lin, in: R.R. Judkins, D.N. Braski (Eds.),

Proceedings of the Fourth Annual Conference Fossil Energy

Materials. U.S. Department of Energy, 1990, pp. 475.

[10] F. Gesmundo, Y. Niu, F. Viani, O. Tassa, J. Phys. IV C9 (1993)

375.

[11] M.A.E. Medina, M. Casales, A.M. Villafane, J.P. Calderon, L.

Martinez, J.G.G. Rodriguez, Mat. Sci. Eng. A300 (2001) 183.

[12] K.L. Luthra, D.A. Shores, J. Electrochem. Soc. 127 (1980) 2202.

[13] R.A. Nyquist, R.A. Kagel (Eds.), IR Spectra of Inorganic

Compounds, Academic Press, New York, USA, 1971.

[14] I. Rommerskirchen, B. Eltester, H.J. Grabke, Mater. Corrosion

47 (1996) 646.

[15] G.C. Rybicki, J.L. Smialek, Oxid. Met. 31 (1989) 275.

D. Das et al. / Materials Science and Engineering A338 (2002) 24�/32 31

[16] J.L. Smialek, J. Doychak, D.J. Gaydosh, in: T. Grobstein, J.

Doychak (Eds.), Oxidation of High Temperature Intermetallics,

the Minerals, Metals and Materials Society, Warrendale, USA,

1988, p. 83.

[17] J.A. Goebel, F.S. Pettit, Metall. Trans. 1 (1970) 1943.

[18] K.P. Lillerud, P. Kofstad, Oxid. Met. 21 (1984) 233.

[19] P. Kofstad, High Temperature Corrosion, Elsevier Applied

Science, New York, 1988, p. 465.

D. Das et al. / Materials Science and Engineering A338 (2002) 24�/3232