Aa

BM

a

ARRAA

KOKRH

1

mhuCk[eTtotcatiu

tee

0h

Thermochimica Acta 549 (2012) 57– 62

Contents lists available at SciVerse ScienceDirect

Thermochimica Acta

jo ur n al homepage: www.elsev ier .com/ locate / tca

novel approach to determine oxidation kinetics of Mo–16Cr–xSi(x = 4–6 wt.%)lloy using stepwise isothermal thermo-gravimetry

haskar Paul ∗, S. Koley, A.K. Suriaterials Group, Bhabha Atomic Research Centre, Mumbai 400 085, India

r t i c l e i n f o

rticle history:eceived 4 May 2012eceived in revised form 28 August 2012ccepted 7 September 2012vailable online 24 September 2012

eywords:

a b s t r a c t

A new method was developed for measuring the kinetics of oxidation. In this method, apparent activationenergy and the exponent related to the oxidation phenomena can be evaluated by conducting a singleexperiment. The method was tested for its applicability by measuring the activation energy for oxidationof pure Mo. It was found that the isothermal weight change data of Mo from stepwise isothermal thermo-gravimetry (SITG) could be well analyzed to get kinetics parameters according to the empirical rateequation. ( )

xidation

ineticsefractory metalsot pressing

dY

dt= nk(T)Y(1 − Y)

1 − Y

Y

1/n

Further, the method was extended for evaluating the oxidation kinetics parameters forMo–16Cr–(4–6 wt.%)Si alloys. Effects of concentration of Si on the oxidation behaviors of the alloys werealso studied.

. Introduction

Thermal oxidation problem restrict the widespread use ofolybdenum and its alloys as a structural material, although the

igh temperature mechanical properties are promising. Thereforenderstanding the kinetics of oxidation is of utmost importance.onsiderable efforts have been made in the past to understand theinetics and mechanisms involved in oxidation of Mo and its alloys1–5]. Since oxidation is a thermally activated process, activationnergy and the governing rate laws are important to be known.he activation energies for the oxidation of metals are related toheir band gaps, heats of formation, electronic conductivity of thexides formed [6] and their work functions [7,8]. Band gap, elec-ronic conductivity of the oxides, work function, etc., of the metalsan be altered by suitable alloying additions, which can increase thectivation energy, resulting improvement of the oxidation resis-ance. Thus, determination of the activation energy for oxidations important in correlating with the above physical properties andnderstanding the oxidation behavior of metals and alloys.

Various equations have been suggested to describe the oxida-

ion kinetics and calculate the activation energy [9]. Preferably fourxperiments are required at four different heating rates or differ-nt isothermal temperatures, for calculating activation energy for

∗ Corresponding author. Tel.: +91 22 25595313; fax: +91 22 25505151/19613.E-mail addresses: bpaul@barc.gov.in, er bhaskar2001@yahoo.com (B. Paul).

040-6031/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.tca.2012.09.018

© 2012 Elsevier B.V. All rights reserved.

any thermally activated process by constant rate of heating (CRH)method or isothermal heating [10,11]. In this respect stepwiseisothermal thermo-gravimetry (SITG) would have an advantage,because in the single experiment activation energy could be foundout, using the analogies for evaluating sintering kinetics from step-wise isothermal dilatometry (SID). SID is a relatively new approachas compared to conventional sintering studies and has provenits usefulness in analyzing the sintering mechanism of ceramic[12–15], metal [16] and alloys [17].

There has been a consistent effort to develop the Mo based alloyssuitable for high temperature applications beyond Ni base superalloys, for power and aerospace industries. These alloys shouldhave better high temperature creep and oxidation resistance, andgood room temperature fracture toughness. A significant amountof research has been carried out to enhance the oxidation resis-tance of molybdenum by alloying. In the recent time, Mo–Si–B alloyis the most studied system, in which the relatively more ductileMoss phase reinforced with Mo3Si and Mo5SiB2 phases have shownmarked improvement in the oxidation resistance along with goodroom temperature fracture toughness [18–24].

In the similar approach, Mo–Cr–Si ternary system is also apromising candidate, possessing good properties for high temper-ature applications [25–27]. However, a very little information is

available in the literature on the oxidation properties of Mo–Cr–Sialloys. Si is the most important alloying element for Mo, whichresults in various combinations of fruitful multiphase system. Sicontent less than 25 at.% (10 wt.%) forms a two phase structure of

5 imica

Mpti

okmdail

2

woaorvdofg

amTpgdf1tgiheHwftaffdiatoo(2aneM

3

ws

8 B. Paul et al. / Thermoch

o + Mo3Si. The concentration of Si needed for establishment of arotective layer can be reduced significantly by the addition of ahird element such as Cr. Cr forms a solid solution with Mo andncreases the oxidation resistance of the matrix [25–28].

In the present study, we have attempted to imbibe the method-logies of SID, for using SITG method to determine oxidationinetics of molybdenum and proving the applicability of the novelethod. Further, SITG has been used for evaluating the oxi-

ation kinetics parameters for oxidation resistant Mo–16Cr–xSilloys where x = 4, 5 and 6 wt.%. No such data of oxidation kinet-cs parameters of Mo–16Cr–xSi alloys are available in the openiterature.

. Experimental procedure

Commercially available pure molybdenum strip of known areaas cut from a sheet, and used for oxidation test at 1 atmosphere

xygen pressure in Labsys TG, Setaram, France, make thermobal-nce using alumina crucible. Samples were heated at a heating ratef 20 K/min up to the 523 K then onwards up to 673 K with the sameate of heating and an isothermal hold of 60 min at each 50 K inter-al (step). At higher temperatures (beyond 673 K) a quantitativeescription of oxidation rates of pure Mo is complicated becausef the high vaporization rate of MoO3. Therefore oxidation studyor Mo was limited up to 673 K to avoid any interference of weightain with the vaporization of MoO3.

Elemental powder of Mo, Cr, Si and Ti with a purity of 99.9% andn average size of 5, 8, 6 and 4 �m, respectively, were thoroughlyixed in the desired composition ratios of Mo–16Cr–(4–6)Si (wt.%).

he mixed powder was uniaxial pressed under 200 MPa into discellets. Green densities of the samples were measured using theeometrical dimensions. The compacts were placed into graphiteie and then hot-pressed under vacuum at different temperaturesrom 1300 to 1600 ◦C for 0.5 h to 3 h with an applied pressure of0 MPa. The hot pressed disc pellets were ground from all sideso remove the graphite (C) layer using standard metallographicrinding techniques. Sintered densities were determined throughmmersion method based on the Archimedean principle using alco-ol as the liquid medium. The polished sintered samples weretched by using a mixed solution of 5 ml HNO3, 10 ml HF, 15 ml2SO4 and 50 ml H2O, and the microstructure of the specimensere observed under SEM. X-ray diffraction pattern were recorded

or characterizing the phases evolved after hot pressing. All diffrac-ion patterns were recorded in the 2� range from 35◦ to 50◦, with

slow scanning rate of step size 0.002◦/s. Specimens were cutrom the hot pressed pellets and ground to known surface areaollowed by polishing and ultrasonically cleaning in acetone. Oxi-ation tests were carried out at 1 atmosphere oxygen pressure

n a thermobalance and weight change profiles were recordedt a heating rate of 20 K/min up to the 773 K then onwards upo 1073 K with the same rate of heating and an isothermal holdf 60 min at each 100 K interval. Phases present on the surfacef the oxidized samples were characterized by X-ray diffractionXRD) in an Inel-make unit (model MPD) with Cr-K� radiation at5 mA 35 kV. The morphology and natures of oxides layers werenalyzed by observing the surface and the cross-section in scan-ing electron microscopy (SEM, Camscan MV2300CT/100, UK),quipped with energy dispersive spectrometry (EDS, Oxford, X-ax 80).

. Theory

In general a study of the oxidation behavior of any system beginsith some measurement of the extent of conversion of the original

ystem to oxide as a function of time t under selected condition of

Acta 549 (2012) 57– 62

temperatures and oxygen pressure. Then the data are analyzed toyield a rate law. The conventional equation for oxidation kineticsis generally described as:

xm = kt (1)

where ‘x’ is the oxide film thickness or the mass gain due to oxi-dation, which is proportional to the oxide film thickness, t is timeat isothermal hold, k is the rate constant and ‘m’ is the exponentrelated to the oxidation mechanism. In wider application ‘x’ is aspecific measurement which may be weight change, oxygen con-sumption, thickness of oxide layer, recession of the metallic systemor some combination of these. The conventional kinetic laws statedabove to characterize the oxidation kinetics is based on simple oxi-dation models. Practical oxidation problems usually involve muchmore complicated oxidation mechanisms than considered in thesesimple analyses.

In this study the specific measurement is weight change deter-mined by thermo-gravimetry (TG). The analysis of weight changesdata obtained by TG using the stepwise isothermal approach can becarried out using the similar approach as the shrinkage data duringsintering using SID. Although oxidation and sintering is entirelydifferent phenomena, but both the process is actually thermallyactivated process associated with activation barrier. The major dif-ference between oxidation and sintering is that the alloy oxidationis a surface process while sintering is a volume process which alsoinvolves surface diffusion.

To analyze any type of stepwise isothermal data, “dynamic andrelative” changes with respect to different steps of isothermal hold-ing should be considered [12]. So, for SITG experiments, ‘x’ shouldcorrespond to dynamic relative weight change. An equation todescribe “dynamic” relative specific weight change with time canbe written as in Eq. (2), assuming that the total exposed specificarea (A) of the sample remain constant during the process.

x = (wt − wo)(wf − wt)

= [k(T)(t − to)]n (2)

where

k(T) = ko exp(

− Q

RT

)(3)

where to is the time at which the given isothermal step is started,wo, wt, and wf are the initial weight, weight at particular time ‘t’ andweight of the samples when oxidation is completed respectively.k(T) is the specific rate constant which obeys the Arrhenius law,ko, the frequency factor, Q, the apparent activation energy, R, theuniversal gas constant and ‘n’ is a parameter related to the oxidationmechanism. Eq. (2) can further be modified to include fractionaloxidation function (Y) as:

wt − wo

wf − wt= Y

1 − Y= [k(T)(t − to)]n (4)

where, the fractional oxidation function, Y can be defined as:

Y = wt − wo

wf − wo(5)

The value of fractional oxidation function, Y varies from 0 to 1. Eq.(4) is fairly reasonable for the initial stage of a oxidation, as thechange in a growing process (growth of oxides or weight change)should be proportional to [k(T)t]n. A normalized rate equation canbe written as Eq. (6), by taking derivative and eliminating (t − to).

dY

dt= nk(T)Y(1 − Y)

(1 − Y

Y

)1/n

(6)

Eq. (7), become linear by taking log on both side of Eq. (6):

ln

(dY/dt

Y(1 − Y)

)= 1

nln

[(1 − Y)

Y

]+ ln[nk(T)] (7)

B. Paul et al. / Thermochimica Acta 549 (2012) 57– 62 59

Fig. 1. Plots of ln{(dY/dt)/Y(1 − Y)} vs ln{(1 − Y)/Y} for pure Mo. Variation of tem-perature with time is also shown on the right hand scale. Inset shows the Arrheniusplot of ln k(T) vs 1/T yielding activation energy for oxidation.

Tkmtu

4

4

filiT‘5fisAcTpwoe

(non-isothermal) heating up to 1000 ◦C are given in Fig. 4, whichshows initial stage of oxidation behavior of the alloys in flowing O2.Initial mass gains are mainly due to formation of oxides of Mo, Cr

he plot of ln{(dY/dt)/Y(1 − Y)} vs ln[(1 − Y)/Y] where Y can benown from Eq. (5) as wo, wt, and wf are all experimentally deter-ined using SITG. The plot in the form of a straight lines during

he isothermal steps, indicate the equation is well valid to the casender investigation.

. Result and discussion

.1. Kinetic analysis of Mo based on SITG measurement

Using the above model, the weight change data of Mo, aftertting into Eq. (7) is plotted in Fig. 1 as ln{(dY/dt)/Y(1 − Y)} vs

n[(1 − Y)/Y]. A near straight-line behavior for each isothermal zonendicates the validity of the model for the oxidation data of Mo.he values of slopes ‘1/n’ were 0.94, 1.08, 0.81, 0.91 and intercepts

ln nk(T)’ were −9.5, −9.2, −7.8, −5.7 at isothermal sections of 523,73, 623 and 673 K, respectively, evaluated by least-squares lineartting for each straight-line segments of the curve with regres-ion parameters better than 0.998. The inset of Fig. 1 shows therrhenius plot of ln k(T). The apparent activation energy ‘Q’ was cal-ulated from the slopes of linear fit and found to be 135 ± 6 kJ/mol.he result obtained was in good agreement with the previouslyublished results reported by Gulbransen et al. [3] and Vijh [6],here they reported the value of activation energy for oxidation

f pure Mo to be 151 and 146 kJ/mol, respectively, proving the

ffectiveness and validity of the model.

Fig. 2. BSE-SEM images of the two phase (a) Mo–16Cr–4Si and (b) Mo–16Cr–6Si alloys s

Fig. 3. XRD pattern of hot pressed Mo–16Cr–4Si alloy.

4.2. Oxidation studies and kinetic analysis of Mo–16Cr–xSi alloysbased on SITG measurement

Sintered densities of hot pressed alloys were found to be >96%of theoretical densities. More details on the optimization of hotpressing parameters, density measurements, reactions and phaseformation during sintering and sintering mechanism, etc., are beingreported elsewhere [26], as these are beyond the scope of thepresent study.

Fig. 2a and b shows the SEM images taken in back scatteredmode revealing the microstructures of sintered Mo–16Cr–4Si andMo–16Cr–6Si alloys respectively. The microstructures consisted oftwo different contrast regions as white and black due to differ-ent chemical compositions acquired in those regions. The phase A,appearing white was made up of a solid solution phase basically ofMo and Cr containing small amount of Si. The phase B, appearingblack was made up of (Mo,Cr)3Si intermetallic phase. The volumepercentages of phase A are 28, 26 and 23%, respectively, for thealloy having 4, 5 and 6 wt.% Si. The XRD pattern of the sinteredMo–16Cr–4Si alloy has been indicated in Fig. 3. The well-definedpeaks as exhibited in the XRD plot conformed to molybdenum andMoSi3 phase.

Non-isothermal TGA studies for oxidation have first been car-ried out to determine the onset of temperatures at which weightlosses start taking place, i.e. the temperatures at which volatiza-tion of MoO3 surpass the weight gains due to the formation ofoxides constituents. TGA plots for oxidation during continuous

and Si whereas later mass is due to volatilization of MoO3.

howing difference of contrast for different phase A (Mo rich) and phase B (Si rich).

60 B. Paul et al. / Thermochimica Acta 549 (2012) 57– 62

Fa

hvsfdwlicMntmopo2fiii

Fig. 6. Plot of fractional oxidation function (Y) of Mo–16Cr–4Si alloy with time.Heating schedule is also shown in the plot.

ig. 4. TG plots showing the non-isothermal oxidation behavior of Mo–16Cr–4Si

nd Mo–16Cr–6Si alloy.

Based on the findings, stepwise isothermal heating scheduleas been selected. Fig. 5 shows the plot of ln[(wt − wo)/(wf − wt)]s ln(t − to) using Eq. (2). The linearity at each isothermal stepshows the validity of the model proposed. Fig. 6 shows the plot ofractional oxidation function (Y) with time for Mo–16Cr–4Si alloyuring the isothermal steps. It has been seen that the maximumeight gain was observed during the soaking at 700 ◦C. Weight

oss was observed during heating the sample beyond the heat-ng schedule shown in Fig. 6. So, only weight gain data have beenonsidered in the whole SITG analysis. The weight change data foro–16Cr–4Si alloys, after fitting into Eq. (7) is plotted in Fig. 7. A

ear straight-line behavior has been seen for each isothermal sec-ion which indicates the weight change data is well fitted to present

odel. It is seen from the inset of Fig. 7 that there is increase in ratef fractional weight gain with increase in isothermal holding tem-erature gradually with time up to 973 K. The maximum net ratef weight gain was recorded during an isothermal soak period of5 min at 1073 K. The rate of weight change then decreases with

urther soaking period as seen in the inset of Fig. 7. The gradualncrease in rate of change of weight with increasing temperatures due to enhanced thermally activated diffusion O2. The max-ma in the oxidation rate accounts for all possible mass transport

Fig. 5. Plot of ln[(wt − wo)/(wf − wt)] vs ln(t − to) using Eq. (2).

Fig. 7. Plots of ln{(dY/dt)/Y(1 − Y)} vs ln{(1 − Y)/Y} for Mo–16Cr–4Si alloy. Insetshows the fractional oxidation rate as a function temperature.

mechanisms operating in tandem. Afterward, decrease in the rateof weight change is mainly due to two reasons-firstly, the forma-tion of protective oxide layer which inhibit diffusion of O2 throughoxide layers. The continued chemical reaction of oxide formationthus slows down and so the rate of change of weight. Secondly, dueto simultaneous volatization of MoO3, the net rate of weight change

decreases.

The values of slope ‘1/n’ and intercept ‘ln nk(T)’, evaluated byleast-squares linear fitting for each straight-line segment of thecurve for all three alloys have been presented in Table 1. From

Fig. 8. Arrhenius plot of ln k(T) vs 1/T yielding activation energy for oxidation of thealloy Mo–16Cr–xSi alloys having x = 4, 5, and 6 wt.%.

B. Paul et al. / Thermochimica Acta 549 (2012) 57– 62 61

lloy (b) Mo–16Cr–6Si alloy. Insets show BSE-SEM images of corresponding alloys.

Tdaeci‘ost

tw1rctc

Ms(sflFMtuir

iMaitesc

TK

Fig. 9. SEM image of the oxidized surface for hot pressed (a) Mo–16Cr–4Si a

able 1 it is seen that the parameter ‘1/n’, which relates to the oxi-ation mechanism takes nearly same values up to 973 K for threelloys, which shows the similarity of oxidation mechanism. How-ver, the value of slope is slightly different at 1073 K suggestinghange in mechanism. Here it is worthy to mention that the ‘1/n’s different from the exponent ‘m’ described in Eq. (1). The slope1/n’ cannot predict the oxidation rate laws (parabolic, linear, cubicr other oxidation rate law) as possible by exponent ‘m’. Howeverince the slope ‘1/n’ is related to oxidation mechanism, it wouldake different values for different oxidation mechanisms.

Arrhenius plots between ln k(T) and 1/T for the SITG data for thehree alloys are shown in Fig. 8. The apparent activation energy ‘Q’as calculated from the slopes of straight lines and found to be

34, 168 and 210 kJ/mol for the alloys having 4, 5 and 6 wt.% Si,espectively, in the temperature range of 773–973 K. The probablehange in mechanism above 973 K is shown by dotted lines. Thus,he activation energy for oxidation increases with increase in theoncentration of Si in the Mo–Cr–Si alloys.

SEM images of oxidized surfaces of the Mo–16Cr–4Si ando–16Cr–6Si alloys after stepwise oxidation schedule are pre-

ented in Fig. 9a and b, respectively, while back scattered electronBSE) image of the corresponding alloys before oxidation are alsohown in the respective insets. The oxidized alloys exhibited a uni-orm and continuous oxide surface. XRD pattern of the surface oxideayer of the alloys confirmed the formation of SiO2 and Cr2O3 phase.ig. 10 shows the XRD pattern of the surface oxide layer of theo–16Cr–4Si alloy. The surfaces were crack free but porous and

he morphology of the oxide scales consisted of mainly of partic-lates. The porous nature of the oxide layer is seen in SEM image

n Fig. 9a, while dense oxide scale is seen in SEM image in Fig. 9bevealing better oxidation resistance of the Mo–16Cr–6Si alloy.

Upon exposure to O2 at elevated temperature, at the firstnstance, the surface is covered with competitive oxide formation of

o, Cr and Si. At 633 ◦C, MoO3 begin to grow over the scale surface,nd growth continues up to 725 ◦C [4]. The vapor pressure of MoO3ncreases significantly at 725 ◦C, and MoO3 begins to vaporize from

he oxide–gas interface [4]. The change in oxidation mechanism isxpected due to this transition. The result of the present study alsouggests that there is a change in the operating mechanism, as ahange in the slope is seen in Fig. 8 above 700 ◦C.

able 1inetic parameters of SITG oxidation data of three alloys.

Temp. (K) 4% Si 5% Si 6% Si

1/n ln nk(T) 1/n ln nk(T) 1/n ln nk(T)

773 0.95 −17.90 1.1 −14.35 1.33 −11.96873 1.2 −14.61 1.21 −11.71 1.68 −9.67973 1.1 −11.79 1.15 −8.98 1.16 −7.55

1073 1.45 −8.10 1.32 −6.27 1.71 −5.47

Fig. 10. XRD pattern of oxides formed upon oxidation over Mo–16Cr–4Si alloy sur-face.

Vaporization of MoO3 from the oxide–gas interface leavesbehind the porous oxide layer. Supply of O2 from the atmospherethrough the flawed outer layer continues, due to increase in poros-ity. This causes creation of fresh nucleation sites for formation ofoxides. The Gibbs free energies of the formation are −523.5 and−678.3 kJ/mol O2 for Cr2O3 and SiO2, respectively, at 1273 K. Bothoxides are thermodynamically stable and readily form by exposureof air or oxygen at higher temperatures. With time, Cr2O3 and SiO2precipitates are linked and agglomerated by viscous flow causingclosure of submicron scale porosity to form a complete oxide layerof Cr and Si.

5. Conclusion

In summary, oxidation kinetics of Mo and Mo–16Cr–xSi alloyswere studied by SITG technique. The model was found to fit wellwith the weight change data of both Mo and Mo–16Cr–xSi alloysand hence validated its usefulness for oxidation studies of metaland alloys to evaluate the oxidation kinetics of and gives fairly reli-able values of the activation energies using the thermo-gravimetrydata of a single experiment. Effect of Si on the oxidation resis-tance was evaluated. The activation energy was found to be 134,168 and 210 kJ/mol for the three alloys having Si 4, 5 and 6 wt.%respectively.

Acknowledgements

The authors sincerely thank Dr. R. C. Hubli, Head, MPD and Dr.Alok Awasthi, Head, R&RMDS, BARC for their useful discussions andsupport during the course of this work.

6 imica

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

2 B. Paul et al. / Thermoch

eferences

[1] E.S. Jones, J.F. Mosher, R. Speiser, J.W. Spretnak, Molybdenum oxidation, Cor-ros.: Natl. Assoc. Corros. Eng. 14 (1958) 20–26.

[2] N. Floquet, O. Bertrand, J.J. Heizmann, Structural and morphological studies ofthe growth of MoO3 scales during high-temperature oxidation of molybdenum,Oxid. Met. 37 (1992) 253–280.

[3] E.A. Gulbransen, K.F. Andrew, F.A. Brassart, J. Electrochem. Soc. 110 (1963)952–959.

[4] S. Paswan, R. Mitra, S.K. Roy, Isothermal oxidation behaviour of Mo–Si–B andMo–Si–B–Al alloys in the temperature range of 400–800 ◦C, Mater. Sci. Eng. A424 (2006) 251–265.

[5] T.A. Parthasarathy, M.G. Mendiratta, D.M. Dimiduk, Oxidation mechanisms inMo-reinforced Mo5SiB2 (T2)–Mo3Si alloys, Acta Mater. 50 (2002) 1857–1868.

[6] A.K. Vijh, Relationship between activation energies for the thermal oxidation ofmetals and the semiconductivity of the oxides, J. Mater. Sci. 9 (1974) 985–989.

[7] E.K.R. deal, O.H. Wansbrough, An investigation of the combustion of platinum,Proc. R. Soc. (Lond.) A123 (1929) 202–208.

[8] A.F.H. Ward, N.R. Bharucha, The relationship between the work function of ametal and its energy of activation for oxidation, Recl. Trav. Chim. Pays-Bas 72(1953) 735–741.

[9] D. Monceau, D. Poquillon, Continuous thermogravimetry under cyclic condi-tions, Oxid. Met. 61 (2004) 143–163.

10] J. Wang, R. Raj, Activation energy for the sintering of two-phase alu-mina/zirconia ceramics, J. Am. Ceram. Soc. 74 (1991) 1959–1963.

11] K. Matsui, J. Hojo, Sintering kinetics at constant rates of heating: effect of GeO2

addition on the initial sintering stage of 3 mol% Y2O3-doped zirconia powder,J. Mater. Sci. 43 (2008) 852–859.

12] R. Yan, F. Chu, Q. Ma, X. Liu, G. Meng, Sintering kinetics of samarium dopedceria with addition of cobalt oxide, Mater. Lett. 60 (2006) 3605–3609.

13] G.Y. Meng, O.T. Sorensen, in: Y. Han (Ed.), Advanced Structural Materi-

als, vol. 2, Elsevier Science Publishers B.V., Amsterdam, Netherlands, 1991,pp. 369–374.

14] H.T. Wang, X.Q. Liu, F.L. Chen, G.Y. Meng, Kinetics and mechanism of a sinteringprocess for macroporous alumina ceramics by extrusion, J. Am. Ceram. Soc. 81(1998) 781–784.

[

[

Acta 549 (2012) 57– 62

15] Y.F. Liu, X.Q. Liu, S.W. Tao, G.Y. Meng, O.T. Sorensen, Kinetics of the reactivesintering of kaolinite–aluminium hydroxide extrudate, Ceram. Int. 28 (2002)479–486.

16] B. Paul, D. Jain, A.C. Bidaye, I.G. Sharma, C.G.S. Pillai, Sintering kinetics of sub-micron sized cobalt powder, Thermochim. Acta 488 (2009) 54–59.

17] B. Paul, D. Jain, S.P. Chakraborty, I.G. Sharma, C.G.S. Pillai, A.K. Suri, Sinteringkinetics study of mechanically alloyed nanocrystalline Mo–30 wt.%W, Ther-mochim. Acta 512 (2011) 134–141.

18] M. Berczik, Method for enhancing the oxidation resistance of a molybdenumalloy, and a method of making a molybdenum alloy, US Patent 5,595,616 (1997).

19] K. Yoshimi, S. Nakatani, S. Hanada, S.H. Ko, Y.H. Park, Synthesis and high tem-perature oxidation of Mo–Si–B–O pseudo in situ composites, Sci. Technol. Adv.Mater. 3 (2002) 181–192.

20] M. Berczik, Oxidation resistant molybdenum alloys, US Patent 5,693,156(1997).

21] H. Choe, J.H. Schneibel, R.O. Ritchie, Fracture, Fatigue properties ofMo–Mo3Si–Mo5SiB2 refractory intermetallic alloys at ambient to elevated tem-peratures (25 ◦C to 1300 ◦C), Metall. Mater. Trans. A 34 (2003) 225–232.

22] V. Supatarawanich, D.R. Johnson, C.T. Liu, Effects of microstructure on the oxi-dation behavior of multiphase Mo–Si–B alloys, Mater. Sci. Eng. A 344 (2003)328–339.

23] J.H. Schneibel, R.O. Ritchie, J.J. Kruzic, P.R. Tortorelli, Optimization of Mo–Si–Bintermetallic alloys, Metall. Mater. Trans. A 36 (2005) 525–531.

24] P.R. Subramanian, M.G. Mendiratta, D.M. Dimiduk, High temperature meltingmolybdenum–chromium–silicon alloys, US Patent 5,683,524 (1997).

25] B. Paul, S.P. Chakraborty, J. Kishor, I.G. Sharma, A.K. Suri, Studies on syn-thesis and characterization of Mo based in situ composite by silicothermalco-reduction process, Metall. Mater. Trans. B 42 (2011) 700–710.

26] B. Paul, S. Majumdar, A.K. Suri, Microstructure and mechanical properties of hotpressed Mo–Cr–Si–Ti in situ composite, and oxidation behavior with silicidecoatings, Unpublished work.

27] D.B. Lee, G. Simkovich, Oxidation of molybdenum–tungsten–chromium–siliconalloys, Oxid. Met. 31 (1989) 265–274.

28] S. Ochiai, Improvement of the oxidation-proof property and the scale structureof Mo3Si intermetallic alloy through the addition of chromium and aluminumelements, Intermetallics 14 (2006) 1351–1357.