Letters

Kinetic Study of the Competitive Growth Between h-Al2O3 and a-Al2O3 During the Early Stages of Oxidation of b-(Ni,Pt)Al Bond Coat Systems: Effects of Low Oxygen Partial Pressure and Temperature

Description
Kinetic Study of the Competitive Growth Between h-Al2O3 and a-Al2O3 During the Early Stages of Oxidation of b-(Ni,Pt)Al Bond Coat Systems: Effects of Low Oxygen Partial Pressure and Temperature
Categories
Published
of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Kinetic Study of the Competitive Growth Between  h -Al 2 O 3 and  a -Al 2 O 3  During the Early Stages of Oxidation of  b -(Ni,Pt)AlBond Coat Systems: Effects of Low Oxygen Partial Pressureand Temperature J.M. ALVARADO-OROZCO, R. MORALES-ESTRELLA, M.S. BOLDRICK,G. TRAPAGA-MARTINEZ, B. GLEESON, and J. MUNOZ-SALDANAAn oxidation study of   b -(Ni,Pt)Al commercial bond coat systems was carried by means of TGAanalysis during isothermal treatments at temperatures from 1273 K to 1423 K (1000   C to1150   C). The effect of oxygen partial pressure on their oxidation kinetics was studied andcomplemented by photo-stimulated luminescence spectroscopy and SEM. Pre-oxidation treat-ments performed on as-aluminized samples at 10  5 atm O 2  did not accelerate the  h -Al 2 O 3  fi a -Al 2 O 3  transformation, even after 5 hours of oxidation relative to samples oxidized in 0.21 atmO 2 , with the exception of the sample treated at 1273 K (1000   C), where apparently  a -Al 2 O 3 nucleation started earlier for the sample treated at a low  p O 2 .DOI: 10.1007/s11661-014-2669-3   The Minerals, Metals & Materials Society and ASM International 2014 I. INTRODUCTION H IGH - TEMPERATURE  coatings have been widelyused since the 1950s to protect hot-section componentsin gas turbine engines against oxidation and corrosionphenomena. [1 – 3] In general, such coatings are eitherdiffusion or overlay. [4] Diffusion coatings have beenstudied extensively in recent decades, with those basedon  b -NiAl being the most widely used in the land,marine, and aeronautical applications. [4 – 7] In order toimprove the resistance of   b -based coatings to oxidationand hot-corrosion, the judicious addition of elementssuch as Pt, Pd, Rd, Ir, Cr, Hf, Si, and Y has beeninvestigated. [8 – 11] Pt-modified  b -NiAl systems, the focusof the present study, are diffusion coatings comprising:(1) a Ni-based superalloy as the substrate that providesmechanical strength during operation; (2) an aluminascale forming  b -(Ni,Pt)Al composition that is an alumi-num reservoir to sustain the alumina-scale formation;and (3) a thermally grown oxide (TGO) resulting fromthe  b  oxidation [3,12] and acting as a diffusion barrier toprevent substrate oxidation as well as to increase its hot-corrosion resistance. In the case of thermal barriercoatings (TBC) systems, the  b -based coating server asbond coat to the ceramic top coat (TC) and the intrinsicfailure mechanisms as defined by Evans  et al. , [12] areinvariably associated with detachment along the BC/TGO interface. For instance, thermal expansion mis-match between the TGO and BC can result largecompressive stresses (3 to 5 GPa) in the TGO layer during cooling. [13 – 16] In order to extend the service livesof TBC systems and  b -NiAl coatings in general, theTGO should: (1) be  a -Al 2 O 3  in structure with the largestpossible grain size; (2) have present a uniform columnarmorphology along  b -coatings surface; (3) have goodadhesion between the BC and TC; and (4) be slow growing. [17 – 19] Rybicki and Smialek [20] and separatelyBrumm and Grabke [21] showed that oxidation of theintermetallic  b -NiAl exhibits two polymorphic transi-tions ( c -Al 2 O 3  fi  h -Al 2 O 3  fi  a -Al 2 O 3 ) before reachingthe stable  a -Al 2 O 3  structure. The process to establish astable  a -Al 2 O 3  scale may involve several stages of nucleation and competitive growth depending of theoxidation conditions. For instance, Brumm and Grab-ke [21] also showed that at 1148 K (875   C),  c -Al 2 O 3  isthe controlling phase from about 4 to 10 hours. Asubsequent transition period of nucleation and compet-itive growth of   h -Al 2 O 3  are observed with the continu-ous increase of the parabolic rate constant from about10 to 16 hours, followed for a  h -Al 2 O 3  controllingregime. However, during the early stages of oxidation(below 4 hours for this example), additional transitions(and, therefore, differences in instantaneous growthrate) can be expected due to different factors includingthe heating procedure, surface preparation, and speci-men handling. These factors can affect the interfacial-reaction steps associated with mass at the gas/scale and/or scale/alloy interfaces, growth of transient oxides ( e.g. ,NiAl 2 O 4 ),  etc .Transformation to the stable  a -Al 2 O 3  is accompaniedby a volume decrease of the lattice unit cell. For J.M.ALVARADO-OROZCO,ResearchAssistant,andB.GLEESON,DepartmentChairman,arewiththeMechanicalEngineeringandMaterialsScience Department, University of Pittsburgh, Pittsburgh, PA, 15261.R. MORALES-ESTRELLA, Professor, is with the Instituto de Investiga-ciones Metalurgicas, UMSHN, Edificio ‘‘U’’, C.U., 58060 Morelia,Michoacan, Mexico. M.S. BOLDRICK, Researcher, G. TRAPAGA-MARTINEZ, Principal, and J. MUNOZ-SALDANA, Professor, are withthe Centro de Investigacion y de Estudios Avanzados del IPN, UnidadQuere ´taro, 76230 Quere ´taro, Mexico. Contact e-mail: jmunoz@qro.cinvestav.mxManuscript submitted April 27, 2014. METALLURGICAL AND MATERIALS TRANSACTIONS A  example, the volume change associated with the h -Al 2 O 3  fi  a -Al 2 O 3  transition is a  ~ 8 pct, reductionwhich can lead to defects and thermal stresses in the b /TGO interface that may promote early failure. [22] It is important to emphasize that the depositionprocess of the well-know TBC systems with  b -(Ni,Pt)Albond coats includes stages that can influence the BCbehavior during service, and, therefore, the TBC sys-tem’s lifetime. These processes include electrolytic Ptplating [23] ( i.e. , sulfur impurities), heat treatments andthe aluminizing process, [5,9] surface preparation ( i.e. ,grit-blasting after the aluminizing), and pre-oxidation treatments [17,18,24 – 28] ( i.e. , with the variables being oxy-gen partial pressure, temperature, heating rate, time)prior to TC deposition onto the BC surface. Results of the effect of oxygen partial pressure (  p O 2 ) and temper-ature on the oxidation behavior of as-aluminized b -(Ni,Pt)Al coatings during isothermal (pre-oxidation)treatments are presented in this paper. These results arepart of a long study on early stage TGO formation inTBC systems. [29] Tolpygo and Clarke [18] demonstrated, by way of thermal cycling tests at 1423 K (1150   C) that pre-oxidation treatments in air of   b -(Ni,Pt)Al BC systems toform a scale of   a -Al 2 O 3  TGO prior to TC deposition canimprove time to failure by up to a factor of  three. Ataround the same time, Spitsberg and More [17] reportedthe effect of TGO microstructure on the durability of TBC systems with  b -(Ni,Pt)Al BCs. They found that BCpre-oxidation treatment, when done under specificoxygen partial pressures, can result in greater than afactor of two improvements in TBC lifetime as com-pared with systems having a non-treated BC. It wasconcluded by these authors that pre-oxidation treat-ments at low  p O 2  reduce the steady-state TGO growthrate and, therefore, the growth stress in the TGO.Moreover, it was also inferred that a direct relationexists between the TCB ´ s lifetime and the  a -Al 2 O 3  TGOgrain size. However, the optimum pre-oxidation param-eters ( i.e. ,  p O 2 , time, and temperature) were not reportedin that study.Even though the effects of pre-oxidation been consid-ered, there remains a lack of information concerning theeffects of the temperature,  p O 2 , surface preparation, andtime on the competitive growth between  h -Al 2 O 3  and  a -Al 2 O 3  during the early stages of oxidation ( i.e. , t < 5 hours), which is important for optimizing thesubsequent TGO growth. Thus, the specific aim of thisstudy was to systematically determine the effects of thetemperature and oxygen partial pressure on the oxidegrowth to the better understanding of   h -Al 2 O 3  fi a -Al 2 O 3  phase transition in order to optimize thesubsequent TGO growth of commercial as-aluminized b -(Ni,Pt)Al BC systems during oxidation treatments. II. EXPERIMENTAL PROCEDURES A.  Sample Preparation The samples used in this work were provided by GEAviation (Evendale, OH) as rectangular specimens(1.8  9  1.2  9  0.15 cm), each weighing about 2.9 g. Bondcoats were produced by electroplating a thin layer of Ptonto Rene N5 single-crystal Ni-based superalloy sub-strates. A heat treatment was subsequently performed todiffuse Pt into the superalloy, followed by a vapor phasealuminizing (VPA) process. After a second heat treat-ment the desired  b -(Ni,Pt)Al phase was achieved. Theresulting bond coat was a bilayer structure consisting of a  ~ 50  l m thick  b -(Ni,Pt)Al and a  ~ 20  l m thick inter-diffusion zone (IDZ) as shown in Figure 1. As indicatein Figure 1(c) the coating surface had a root-mean- (a) (b)(c) Fig. 1—As-coated  b -(Ni,Pt)Al BC system morphology: ( a ) surface view, ( b ) cross-sectional, and ( c ) the root mean square roughness (Rq). METALLURGICAL AND MATERIALS TRANSACTIONS A  square surface roughness (Rq) of 2.03  ±  0.29  l m asmeasured using a Mitutoyo Surftest SJ-201P profilom-eter.The nominal compositions of the substrate and BCwere determined using inductively coupled plasmaatomic emission spectroscopy (ICP-AES, Perkin ElmerOptima DV4300) and an electron probe microanalyzer(EPMA, JEOL 8900 WD/ED), respectively. The resultsare summarized in Table I.All samples were ultrasonically cleaned by sequen-tially using xylene, acetone, methanol–water (1:1), anddeionized water for 15 minutes each to remove surfacecontamination prior to oxidation treatments.B.  Oxidation Exposures Thermogravimetric analyses (TGA) were performedusing a Setaram Setsys Evolution 16/18 thermobalance,which had an accuracy of   ± 0.03  l g. The mass changesduring the TGA experiments were recorded at 5 secondsintervals. A given test sample was hung from one end of the beam balance using a 0.4 mm diameter Pt/30 pctRhwire and positioned—at room temperature- in the hot-zone region of the vertical furnace. To minimize anyperturbation resulting from gas flow, buoyancy, dragforces, or sample oxidation before reaching the testtemperature, the heating cycle was programmed asfollows: (1) evacuation of the sample chamber to lessthan 10 Pa; (2) heating to the reaction temperature at amaximum rate of 50 K (50   C)/min under the samevacuum conditions; (3) back filling the chamber withworking gas to atmospheric pressure using a flow rate of 200 mL/min; (4) fixing the flow of the working gas to20 mL/min;andfinally(5)coolingthefurnacechambertoroom temperature at a maximum rate of 50 K (50   C)/min with a gas flow rate of 0.3 mL/min. In order toevaluate the effect of   p O 2  onthe b -(Ni,Pt)Al BC oxidationkinetics, two different conditions were used: dry air with  p O 2  = 0.21 atm and argon with  p O 2  = 10  5 atm in therange from 1173 K to 1473 K (900   C to 1200   C).C.  Characterizations The structural characterization of the TGO wasmonitored after each isothermal exposure (5 hours) byphoto-stimulated luminescence spectroscopy (PSLS)using a micro-Raman mapping spectrometer (RenishawInVia) connected to a Leica microscope equipped with a532 nm line-focus laser. A 5 9 microscope objective wasused to focus the  ~ 4  l m spot-sized laser beam and tocollect the scattered light. The laser power at the samplewas 5 mW, and the acquisition time for each spectrumwas in the 0.5 to 2.0 seconds range. PSLS spectra wereobtained by mapping the sample surface over a ~ 0.09  l m 2 area with a pitch size of 10  l m. The laserpower at the sample was 5 mW, and the acquisition timefor each spectrum was in the 0.5 to 2.0 seconds range.The PSLS technique is based on photon emission fromCr 3+ ions, a typical impurity incorporated in the crystalstructure of Al 2 O 3  formed on alumina coatings. [16,30,31] Once illuminated, the Cr 3+ ions emit fluorescent radi-ation due to radioactive decay of the excited electrons tothe ground state. [32] The identification of the aluminaphases is done on the basis of the well-characterizedfrequencies of the  a -Al 2 O 3  and  h -Al 2 O 3 . [16] Details aboutthe PSLS spectra analysis can be found in Reference 33.The microstructure of the TGO was characterized usingfield emission scanning electron microscopy (JeolJSM7401F FEG-SEM). III. RESULTS The plots in Figure 2(a) compare the net mass-gain( D m )  vs  the square root of time ( t 0.5 ) during isothermaloxidation of as-aluminized  b -(Ni,Pt)Al coatings samplesfor 5 hours at the two different oxygen partial pressuresof 10  5 and 0.21 atm O 2 , and over the temperaturerange 1273 K to 1423 K (1000   C to 1150   C).The TGA plot for the sample exposed at 1373 K(1100   C) and 0.21 atm O 2  is not included becauseatypical noise during the test was observed. For oxida-tions in 10  5 atm O 2 , the rate of mass gain increase of 1273 K to 1373 K (1000   C to 1100   C), but thenabruptly decreased at 1423 K (1150   C); whereas onlyan increase was observed from 1273 K to 1473 K(1000   C to 1150   C) for the samples exposed in0.21 atm O 2 . All the plots exhibit deviations from thestraight line expected for the classic parabolic model todescribe the growth of a diffusion-controlled scale,  i.e. , D m ¼ k p t 0 : 5 ;  ½ 1  where  k p  is the parabolic rate constant. [34,35] Theobserved deviations from a constant  k p  may be associ-ated with dynamic effects such as grain-boundarydiffusion ( i.e. , TGO grain growth), simultaneous reac-tion steps ( i.e. , interfacial-reaction steps associated withmass or defect transfer at the gas/scale and/or scale/ Table I. Chemical Composition of Single-Crystal Superalloy Rene N5 and  b -(Ni,Pt)Al Bond Coat SampleWeight PercentCr Co Mo Re W Al Ti Ta Hf Pt Y NiNominal ReneN5* 7.0 8.0 2.0 3.0 5.0 6.2 — 7.0 0.2 — 0.01 bal.As-received ReneN5** 6.19 8.27 1.39 3.23 5.03 6.37 0.01 6.93 0.15 — — bal.As-coated  b -(Ni,Pt)Al BC*** 0.80 2.66 0.045 — 0.01 24.91 — 0.28 — 31.86 — bal. *Ref. [3].**Measured by ICP-AES.***Measured by EPMA. METALLURGICAL AND MATERIALS TRANSACTIONS A  alloy interfaces [36] ), and/or the polymorphism of Al 2 O 3 and associated transformations within the TGO.Brumm and Grabke [21] investigated the oxidation kinet-ics of NiAl and NiAl-Cr alloys. The oxidation kineticsshowed two phase transformations from  c -Al 2 O 3  fi h -Al 2 O 3  fi  a -Al 2 O 3 . The  c  fi  h  transformations lead toa small increase in the  k p , whereas that the  h  fi  a  leadsto a strong decrease of parabolic rate constant,  k p , overtwo orders of magnitude. Based on the different growthkinetics of the alumina phases, a deviation from theclassic parabolic model is expected if an alumina phasetransformation is taking place during the oxidationexposure.Figure 2(b) shows the differences between the instan-taneous net mass-gain in 0.21 atm O 2  and that measuredin 10  5 atm O 2  at 1273 K, 1323 K, and 1423 K(1000   C, 1050   C, and 1150   C). A marked deviationis shown for the mass-gain difference at 1323 K(1050   C), which suggest a significant change in theoxidation behavior at the two  p O 2  levels.The instantaneous parabolic rate constant,  k p i  , wasestimated using the data presented in Figure 2(a) andusing a local fitting procedure proposed by Monceauand Pieraggi [36] for the general parabolic law: t ¼ A þ B D m þ C  D m 2 ;  ½ 2  where the coefficients  A  and  B  can be related to differentkinetic parameters and  C   is directly related to the inverseof   k p . Figures 3(a) and (b) show the time dependence of  k p i  . As seen in this figure, all samples reached a steadystate after almost 4 hours of exposure, except for the (a) (b) Fig. 2—( a )  D m vs t 0.5 curves showing the early stage (up to 5 h) oxidation behavior of as-aluminized  b -(Ni,Pt)Al BC’s and ( b ) the net mass-gaindifference between the samples oxidized at  p O 2  = 0.21 atm and  p O 2  = 10  5 atm. 0123450.11101000123450.1110100 pO 2 = 2.1x10 -1  atmpO 2 = 1x10 -5  atm       k       i p   x   1   0   -   7    (   m  g    2    /  c  m    4   s    ) t   (h) 1373K1323K1273K1423K       k       i p    x   1   0   -   7    (   m  g    2    /  c  m    4   s    ) t   (h) 1323K1273K1423K (a) (b) Fig. 3—Log  k p i  plotted as a function of time for as-aluminized  b -(Ni,Pt)Al BC system samples during isothermal oxidation at different tempera-tures and oxygen partial pressures: ( a )  p O 2  = 10  5 atm and ( b )  p O 2  = 0.21 atm. METALLURGICAL AND MATERIALS TRANSACTIONS A  sample exposed at 1373 K (1100   C) in 10  5 atm O 2  forwhich  k p i  is still decreasing after 4.5 hours of exposure.Figure 4 summarizes in an Arrhenius plot the para-bolic rate constants estimated after 5 hours ex, with thedata reported by Brumm and Grabke [21] for NiAl alloyoxidation included to provide a frame of reference.These results will be analyzed and discussed in the nextsection; however, at this point it can be sated that asignificant difference is observed in comparison with thereference system, especially at low  p O 2 .In order to determine if the observed deviations arerelated to the polymorphic transformation of the Al 2 O 3 ,a systematic structural characterization by PSLS wascarried out on the TGO surface of all samples. Figure 5compares a semiquantitative estimation of the  a -Al 2 O 3 fraction as a function of temperature for a fixedexposure of 5 hours for all the tested conditions.A sigmoidal behavior is observed for both  p O 2 conditions with a more extensive  h -Al 2 O 3  fi  a -Al 2 O 3 transformation for the samples oxidized in 0.21 atm O 2 from 1273 K to 1373 K (1000   C to 1100   C), reachingthe 100 pct  a -Al 2 O 3  plateau at 1373 K and 1423 K(1100   C to 1150   C) for 0.21 and 10  5 atm O 2 , respec-tively. The results in Figure 5, suggest that at 1173 K to1323 K (900   C to 1050   C) in 10  5 atm O 2  the TGOgrowth kinetics during the first 5 hours of oxidation aremainly controlled by  h -Al 2 O 3  ( c -Al 2 O 3  signal was notobserved during PSLS analysis). IV. DISCUSSION Simultaneous comparisons of the data presented inFigures 2, 3, 4, and 5 are necessary to provide a better understanding of the  h -Al 2 O 3  fi  a -Al 2 O 3  transforma-tion process. Three different exposures temperatures[1273 K, 1323 K, and 1423 K (1000   C, 1050   C, and1150   C)] are selected for detailed discussion in thefollowing.A.  Exposures at 1273 K (1000   C) It can be assumed that  h -Al 2 O 3  was the kineticallycontrolling phase at both oxygen potentials, at 1273 K(1000   C), since  a -Al 2 O 3  fractions at 0.2 and1  9  10  5 atm O 2  were about 13 and 7.5 pct, respec-tively, after 5 hours of oxidation (see Figure 5). Thisassumption is in agreement with the time evolution of the  k p i  shown in Figure 3, where a stable growth rateconstant was established for both atmospheric condi-tions after 2.5 hours of exposure.Since  h -Al 2 O 3  can be modeled as a  p -type oxide [37 – 39] and Al vacancies ( V  Al ¢¢¢ ) predominate in its structureaccording to f ollowing the reaction (using the Kro ¨ger– Vink notation [40] ): O 2  ¼ 2 O X  O þ 43 V  000 Al þ 4 h  ;  ½ 3  where  O O X  represents an oxygen atom in an oxygen siteof Al 2 O 3  and  h Æ represents an electron hole in valenceband. By invoking the law of mass action and the con-straint of electrical neutrality,  h  ½ ¼ 3  V  Al 000   , the alu-minum vacancy concentration is found to be given as V  Al 000   ¼  K  3 81   3 = 16  p 3 = 16O 2 ½ 4  Inasmuch as  V  Al 000    is proportional to  k p , it can beshown that, in accordance with Wagner’s theory, [41] k p  ¼ Z   p 00 O2  p 0 O2 C  1  p 3 = 16O 2 d  ln  p O 2 ;  ½ 5  where  p 00 O 2 and  p 0 O 2 represent the oxygen partial pressureat the gas-scale and metal-scale interfaces, respectively. Fig. 4—Arrhenius plot showing the  k p  values of as-aluminized b -(Ni,Pt)Al BC systems after 5 h of isothermal oxidation at differenttemperatures and oxygen partial pressures: ( a )  p O 2  = 10  5 atm and( b )  p O 2  = 0.21 atm. Bold lines correspond to the  c -,  h -, and  a -Al 2 O 3 lines reported by Brumm and Grabke [21] .Fig. 5—Comparison of the  h -Al 2 O 3  fi  a -Al 2 O 3  transformation as afunction of temperature after 5 h of isothermal oxidation of   b -(Ni,P-t)Al BC’s with ( a )  p O 2  = 10  5 atm and ( b )  p O 2  = 0.21 atm Theseresults were obtained by image analysis of PSLS mappings. A typicalPSLS spectrum for a TGO composed of both  a - and  h -Al 2 O 3  phasesis also shown. METALLURGICAL AND MATERIALS TRANSACTIONS A
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks