Synthesis, Microstructure, and Mechanical Properties of Ti3Sn(1−x)AlxC2 MAX Phase Solid Solutions

Synthesis, Microstructure, and Mechanical Propertiesof Ti3Sn(1�x)AlxC2 MAX Phase Solid Solutions

Sylvain Dubois,* Guo P. Bei, Christophe Tromas, Veronique Gauthier-Brunet, andPascal Gadaud

Institut Prime - UPR 3346, CNRS – Universite de Poitiers – ENSMA, Departement Physiqueet Mecanique des Materiaux, Bat. SP2MI, Bd Marie et Pierre Curie, BP 30179, F-86962Futuroscope – Chasseneuil, France

Ti3Sn(1�x)AlxC2 MAX phase solid solutions are successfully synthesized from different reactant mixtures. Rietveldrefinement allows to carefully characterize their structures and the ocathedra and trigonal prims distortion parameters asa function of the Al content. Mechanical properties of solid solutions are studied from nanoindentation experiments anddynamic resonant method. It is shown that solid solution hardening is not operative in this system. Elastic modulus is found toincrease from Ti3SnC2 to Ti3AlC2, and such a result is discussed in terms of Ti–A bond stiffness.

Introduction

The ternary Mn11AXn carbides and nitrides1–5

compounds, referred to as 211, 312, and 413 MAXfor n 5 1, 2, and 3, respectively, have attracted greatattention recently.6 Here, M denotes an early transition

metal, A is an A-group element (from IIIA to VIA), andX is either C or N. These phases adopt a hexagonalcrystal structure that consists of edge-sharing [M6X]octahedra interleaved with A layers; MAX phases arethus considered as nanolaminated-layered materials. Inthe M2AX or 211’s phases, every third layer is an Alayer1–3; in the M3AX2 or 312’s, every fourth layer,3 andin the M4AX3 or 413’s, every fifth layer.7,8 The metallicnature of the bonding and the nanolayered nature of thestructure give rise to a unique combination of metallicand ceramic properties. For example, these carbides and

Int. J. Appl. Ceram. Technol., 7 [6] 719–729 (2010)DOI:10.1111/j.1744-7402.2010.02554.x

Ceramic Product Development and Commercialization

r 2010 The American Ceramic Society

This work was financially supported by the ‘‘Agence Nationale de la Recherche’’ in the

PLASMAX project (ANR-07-MAPR0015).

*[email protected]

mailto:[email protected]

nitrides exhibit high strength and stiffness at high tem-peratures, good resistance to thermal shock and oxida-tion, high thermal and electrical conductivities as well ashigh damage tolerance; they are also readily machin-able.9 About 60 of such phases are known, most of thembelonging to the 211 family (the so-called H- or Haggphases). Indeed, there are only a few 312 phases knownto date (Ti3SiC2,1 Ti3GeC2,4 Ti3AlC2,10 Ta3AlC2,11,12

Ti3GaC2,13 (V0.5Cr0.5)3AlC2,14 and Ti3SnC215).

Among the 312 phases, Ti3AlC2, Ti2AlC, and theirsolid solution have attractive properties in terms of lowdensity and excellent high-temperature oxidation resis-tance. Indeed, the oxidation kinetics of Ti3AlC2 andTi2AlC have been determined and results show that theydisplay excellent high-temperature oxidation resis-tance.16–18 It is known that the oxidation resistance ofTiAl alloys is not satisfying because no protective Al2O3

scale forms during the high-temperature oxidation pro-cess.19–22 However, it is interesting to notice that, withmuch lower Al content, a continuous and protectiveAl2O3 scales form onto Ti3AlC2 and Ti2AlC surfacesduring high-temperature oxidation in air.16–18 Such anexcellent oxidation resistance has been discussed interms of microstructural investigations of the Ti3AlC2/Al2O3 interface23,24 and in terms of oxide scale adher-ence onto the substrate.25 It has been shown that theadhesive strength of the oxide scale/Ti3AlC2 and Ti2AlCsubstrates is higher than that of Al2O3 scale on othersubstrate.25 Moreover, high-resolution transmissionelectron microscopy (HRTEM) image revealed thatthe Ti3AlC2/Al2O3 interface is free of amorphous phasewhich promote the adhesion of the Al2O3 scale onTi3AlC2 and thus lead to excellent high-temperatureoxidation resistance.23

Moreover, previous work shows that substitutionsat the M, A, and X sites of MAX phases are all feasi-ble.26–29 For example, Wang and Zhou26 have evi-denced solid solution strengthening for MxM

0(1�x)AlC

with M and M05 Ti, V, and Cr. Barsoum et al.27 re-ported solid solution hardening in Ti2AlC0.5N0.5 atroom temperature. Ganguly and colleagues28,29 showedthat hardness values of Ti3Si(1�x)GexC2 lie in betweenthose of Ti3SiC2 and Ti3AlC2, thus concluding thatsolid solution hardening effect is not operative in thissystem. Quite recently, Ti3SnC2 MAX phase has beensynthesized by using Fe as an additive.15 Nevertheless,Fe and FeSn alloys are still present as impurities in the assynthesized products. In this paper, Ti3Sn(1�x)AlxC2

solid solutions are synthesized by using hot isostatic

pressing (HIP) and different reactant mixtures contain-ing Al and Sn. Rietveld refinement allows determiningthe cell parameters and atom positions as a function ofSn content. The consequences on the distortion of the[Ti6C] octahedra and [Ti6Sn1�xAlx] trigonal prism arededuced. Finally, nanoindentation is used to evaluateelastic modulus and hardness of Ti3Sn(1�x)AlxC2 solidsolutions. Nanoindentation being a local solicitation,one can reach the intrinsic mechanical properties of thephases, although MAX phase solid solutions obtainedby HIP are not formed as a single phase.

Experimental Details

Powders of titanium (150–250mm, 99.5% purity),tin (2–20mm, 99% purity), aluminum (45–150mm,99.5% purity), titanium carbide (o45mm, 98% purity),and carbon (graphite, o20mm) were used as startingmaterials. Substoechiometric TiC (i.e., TiC0.66) was alsoused as a reactant; it was synthesized by mechanicalmilling of a mixture of Ti and 0.66C.30,31 DifferentTi:Sn:Al:C, Ti:Sn:Al:TiC, or TiC0.66:Al ratios were cho-sen in order to produce different Ti3Sn(1�x)AlxC2 solidsolutions. It is well known that some A element is lostduring powder metallurgy processing (HIP, hot pressing,or reactive sintering) of the reactant mixture; as a conse-quence, either (Ti1C) substoechiometry or (Sn1Al)over-stoechiometry were used in this study. By usingthese different reactant mixtures, our goal, in this study,was to synthesize a large set of Ti3Sn(1�x)AlxC2 solidsolutions (i.e., with different x values).

To prepare homogeneous reactant mixtures, pow-ders were thoroughly milled for 1 h in a turbula. Thedifferent mixtures were cold-compacted into cylindricalsteel dies using a uniaxial pressure of 800 MPa. Thegreen density, evaluated from weight and geometricmeasurements, ranged from 77% to 81% of the theo-retical density. The green body was then sealed undervacuum in a pyrex container, placed in a hot isostaticpress (HIP) chamber and subjected to the followingtemperature and pressure cycle:� The sample was heated to 14501C under Ar in 1 h.� The HIP was pressurized with Ar to 50 MPa

in 2 h.� Once the processing temperature was reached, the

sample was held at 14501C for 2 h and 50 MPa for1 h before cooling to room temperature and atmo-spheric pressure.

720 International Journal of Applied Ceramic Technology—Dubois, et al. Vol. 7, No. 6, 2010

After HIPing, samples were machined to removethe encapsulating glass container and sliced using adiamond wheel. Samples were thus ground usingsilicon carbide paper and then polished with a diamondsuspension. Finally, in order to produce a very flatsurface and to avoid any work hardening due to con-ventional grinding, chemo mechanical polishing wasperformed using a suspension of alumina particles.

Phase identification was performed by XRD analysesusing a Bruker D501 diffractometer (Karlsruhe, Ger-many) with CuKa radiation. XRD data were refined us-ing the MAUD software32 in order to extract the latticeparameters of the different Ti3Sn(1�x)AlxC2 solid solu-tions and the sample composition. Microstructures wereexamined by scanning electron microscopy (SEM, JEOL5600LV, Tokyo, Japan). Energy-dispersive X-ray spect-rometry (EDXS, Oxford Isis 300, Buckinghamshire,U.K.) was used to determine the Sn over Al and the Tiover (Al1Sn) contents, and it thus allows to extract the Alcontent in the Ti3Sn(1�x)AlxC2 solid solutions.

A Nano Hardness Tester apparatus from Coatingsand Surface Measurements (CSM) Instruments (Peseux,Switzerland) equipped with a Berkovich indenter was usedto perform nanoindentation tests, and the equivalent ind-enter method33,34 was used to analyze the nanoindentationcurves and to extract the hardness and Young’s modulusvalues. The shape of the indenter has been carefully cal-ibrated for true penetration depth as small as 20 nm bothby indenting fused silica samples of known Young’s mod-ulus (72 GPa) and by direct AFM observation of the ind-enter. In order to study hardness variation withpenetration depth, nine loads (1, 2, 3, 6, 10, 15, 30,150, and 300 mN) have been used. Because most sampleswere not single-phase samples, a statistical analysis havebeen performed for each load over 100 nanoindentations.

The Young’s modulus has also been determined bymeans of the dynamic resonant method35 in bendingmode on the beam specimen. Experiments are performedat 1 K/mn under high vacuum (10�4 Pa) from 150 up to1400 K without any harmful contact, the sample beingmaintained horizontally between steel wires located at thevibration nodes. Furthermore, excitation and detectionare insured by an electrostatic device (capacitance createdbetween the sample and a unique electrode36).

The Young’s modulus E of a bulk sample in thebending mode is directly determined by the relation

E ¼ 0:9464rF 2 L4

h2T

h

L; n

� �

where F is the resonance frequency, r the density, nPoisson’s ratio, h and L, the beam thickness and spanlength, and T(h/L, n) a correcting factor close to 1.35,37

The accuracy of the method is better than 1% withregular specimen dimensions.

Results and Discussion

Microstructural Characterization

Ti3Sn(1�x)AlxC2 samples were synthesized fromdifferent reactant mixtures: 3Ti1Sn12C10.1Fe,15

3Ti1xSn1yAl12C, 2.8TiC0.6610.8Sn10.4Al, and1.8TiC1xSn1yAl1Ti. Figure 1 shows SEM obtainedon samples (b), (c), and (d) (see Table I). One canobserve that Ti3Sn(1�x)AlxC2 (gray) phases always appearin between Ti2Sn(1�x)AlxC (light gray) and TiC (darkgray) phases. Such a result strongly suggests that the 211and TiC phases react to form the 312 solid solution.38

SEM results demonstrate that TiC, Ti2Sn(1�x)AlxC andTi3Sn(1�x)AlxC2 are prone to coexist. Such a feature is notsurprising because close structural relationships betweenTi3AlC2, Ti2AlC, and TiC39; Ti3SiC2, Ti5Si3, and TiC40

and Ti2AlN and TiN41 have been demonstrated usingHRTEM and selected-area electron diffraction (SAED).Nevertheless, it has been shown that, depending onthe composition of the reactant mixture, the reactionmechanisms of Ti3AlC2 formation are different.38,42–44 Insuch a context, our experiments, whose goal consists in thesynthesis of a large set of different Ti3Sn(1�x)AlxC2 solidsolutions, do not allow to obtain the reaction mechanismsand the influence of the reactant mixture composition.Lattice parameters, compositions, and Al content (x) ofthe Ti3Sn(1�x)AlxC2 solid solutions obtained are givenin Table I, together with the different initial reactantmixtures. As an example, the Rietveld refined patternobtained on sample c is shown in Fig. 2; the weightedreliability factor is 13.7%. It can be noticed, in Table I,that only samples (c), (f), and (g) contain 495 vol %of Ti3Sn(1�x)AlxC2 solid solutions. Samples (f) and (g),which contain 100 or 80 at.% of Al as A element, appearas a pure sample at the XRD limit. Thus, the formationmechanism of Ti3Sn(1�x)AlxC2, with x 5 0.8, is very likelythe same as the formation mechanism of Ti3AlC2 fromTiC, Al and Ti.38,42 Thus, TiC reacts with Al, Ti, and Snto form either Ti3Sn0.2Al0.8C2 or Ti2Sn0.2Al0.8C.38 Ata high enough temperature, Ti2Sn0.2Al0.8C can reactwith TiC to form Ti3Sn0.2Al0.8C2. Nevertheless, reaction

www.ceramics.org/ACT Synthesis, Microstructure, and Mechanical Properties of Ti3Sn(1�x)AlxC2 MAX 721

syntheses allow having a set of solid solutions where the Alcontent varies in the range 0–1. Figure 3a shows the a andc lattice parameter variations as a function of the Al con-tent in the Ti3Sn(1�x)AlxC2 solid solutions. The resultsindicate that the a-parameter of all Ti3Sn(1�x)AlxC2

decrease linearly as the Al content increases in the struc-ture. The c-parameter does not appear to vary linearlywith Al content; it may be constant for Al content in the

range of 0.6–1. However, as indicated by the change inthe c/a ratio (see Fig. 3b), the lattice contraction ofTi3Sn(1�x)AlxC2 is strongly anisotropic, namely, muchmore reduction along the a axis than that along the c-axis.a-parameter and c/a variations are consistent with the for-mation of an ideal solid solution and follow the Vegard’slaw. Indeed, according to this law, unit cell parametersvary linearly with composition for a continuous substitu-tional solid solution in which atoms that substitutefor each other are randomly distributed. This law is validfor ideal solid solutions when the lattice parameters ofthe pure components differ by o5%,45 which is the casefor Ti3AlC2 and Ti3SnC2 MAX phases. Moreover, as inV2Ga(1�x)AlxC,46 the z-parameter of TiII (ZTi–II) followsVegard’s law (not shown).

Following previous work,47,48 it is of interest toconsider the MAX crystal unit cell as constituted of[M6X] octahedrons and [M6A] trigonal triangularprisms. As shown in Fig. 4, the stacking of two octahe-drons and one trigonal prism allows describing the 312MAX phases. It should be observed that a cubic octa-hedron is the unit block of the binary MX but it losesits fourfold axis in ternary MAX and it results in arelaxation owing to this reduced symmetry. The non-cubic distortion of the octahedron can be estimatedfrom the parameter Od, defined as the ratio of the dis-tances between two faces not in the basal planes (d1) andtwo opposite faces contained in the basal planes (d2)48

and given, in a 312 MAX phase, by

Od ¼d1

d2¼

ffiffiffi3p

2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiZTi

2 ca

� �2þ 112

q

Concerning the trigonal prism, the distortionparameter Pd, defined as the ratio of the M–M distanceand the M–A distance,47 is given, in either 211 or 312MAX phases by

Pd ¼1

2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

14� ZTi

� �2 ca

� �2þ 13

q

For 312 MAX phases, an ideal packing of hardspheres of equal diameter leads to a ratio c=a ¼ 8�ffiffi

23

q� 6:532 and Od 5 Pd 5 1 for an ideal octahedron

(cubic) and trigonal prism. From the position of Tiatoms (ZTi), and c, a lattice parameters given in Table I,one can easily determine the Od and Pd distortionparameters. Figure 5 shows the Od and Pd variationsas a function of the Al content in the Ti3Sn(1�x)AlxC2

Fig. 1. Scanning electron images recorded in backscattered modeon (a) sample b, (b) sample c, (c) sample d.


solid solution. As expected, it can be concluded thatboth polyhedrons are distorted whatever the Al contentin the Ti3Sn(1�x)AlxC2 MAX phase. Ti3SnC2 showsstrong distortion of the octahedron; however, Ti3AlC2

shows small distortion of the octahedron. The trigonalprism distortion does not appear to change a lot withthe Al content of the solid solution. Finally, the octa-hedron and trigonal prism distortion are quite similarfor Al content in the range of 0.4–0.5 (the ratio Od/Pd

departs from unity by 4.10�4). As shown previously,3

anisotropic deformation, that is expansion along thea-axis and contraction along the c-axis, of the [Ti6C]octahedrons occurs in ternary compounds. It is larger inTi3SnC2 than in Ti3SiC2

3 and Ti3AlC2. As shown inTable II, the Ti–C–Ti distances calculated for theTi3Sn(1�x)AlxC2 solid solutions are strongly modifiedcompared with those in TiC. This result indicatesthat the distortion is accommodated by shrinking ofthe Ti–C bonds in the [Ti6C] octahedrons and it differsfrom the one obtained by Gamarnik and Barsoum50

who mentioned that accommodation occurs by Ti–Cbond rotation. Such a shrinking of the Ti–C–Ti dis-tances will affect the Ti–C bond stiffness and may thusmodify the elastic properties of the ternary compounds.

Table I. Initial Reactant Mixture, Lattice Parameters, Composition, and Al Content (x) in the DifferentTi3Sn(1�x)AlxC2 Solid Solutions

Initial reactant mixtureLattice parameters of

Ti3Sn(1�x)AlxC2

Rietveld refinement(vol%)

x value from EDXSexperiments

3Ti1Sn12C10.1 Fe15 a 5 3.1366 A 80.5% of 312 Sample ac 5 18.650 A 16.2% of TiC 0

ZTiII 5 0.1197 3.3% of impurities3Ti10.8Sn10.3Al12C a 5 3.1227 A 57.6% of 312 Sample b

c 5 18.613 A 33.7% of 211 0.25ZTiII 5 0.1297 8.7% of TiC

3Ti10.6Sn10.6Al12C a 5 3.1052 A 97.2% of 312 Sample cc 5 18.604 A 3.8% of 211 0.38

ZTiII 5 0.12972.8TiC0.6610.8Sn10.4Al a 5 3.1114 A 41.8% of 312 Sample d

c 5 18.598 A 43.5% of 211 0.4ZTiII 5 0.1267 14.7% of TiC

1.8TiC10.5Sn10.5Al1Ti a 5 3.1053 A 77.9% of 312 Sample ec 5 18.589 A 2.3% of 211 0.5

ZTiII 5 0.1274 19.8% of TiC1.8TiC10.2Sn1Al1Ti a 5 3.0891 A 98% of 312 Sample f

c 5 18.603 A 2% of impurities 0.8ZTiII 5 0.1315

1.9 TiC1Al1Ti a 5 3.0779 A 100% of 312 Sample gc 5 18.579 A 1

ZTiII 5 0.1326

EDXS, energy-dispersive X-ray spectrometry.

Fig. 2. Rietveld refined pattern obtained on sample c. Blacksquares: experimental data, full line: Rietveld refinement.


The small Ti–Ti spacing, observed in Ti3AlC2, wouldincrease direct bonding between Ti sheets whereashigher values, observed in Ti3SnC2, would favor theformation of C 2p-Ti ds bonds.

Mechanical Properties

Nanoindentation tests were performed on differentTi3Sn(1�x)AlxC2 samples in order to determine elasticmodulus and hardness values as a function of the Sncontent. In the following section, experimental details ofthe different analyses are given in the case of theTi3SnC2 sample which contains (see Table I) Ti3SnC2

(80.5 vol%), TiC (16.2 vol%), and minority phases(3.3 vol%).

The histogram of the elastic modulus, shown inFig. 6 and obtained for a 3mN load, presents threepeaks centered at 370, 245, and 177 GPa. The peak at

370 GPa is ascribed to the TiC phase, and the peak at177 GPa very likely results from the minority phases,and from artefacts induced when the indent is localizedon a grain boundary. The main peak at 245 GPa is un-ambiguously attributed to the Ti3SnC2 MAX phase.This value is moreover in good agreement with resultsobtained by nanoindentation in MAX phase thin

Fig. 3. (a) a (open square) and c (open circle) lattice parametersand (b) c/a ratio versus Al content in Ti3Sn(1�x)AlxC2 solidsolutions.

Fig. 4. Unit cell of the Ti3SnC2 phase. [Ti6C] octahedrons and[Ti6Sn] trigonal prisms are represented. Drawn with VESTA.49

Fig. 5. Octahedron (Od, full triangles) and trigonal prism (Pd,open squares) distortions versus Al content in Ti3Sn(1�x)AlxC2 solidsolutions.


films51 or by ultrasound methods.52 In Fig. 6, Gaussianpeak deconvolution leads to area ratios of 74% forTi3SnC2, 22% for TiC, and 4% for the minority phase.These values are in reasonable agreement with the vol-ume fractions determined from Rietveld refinement.

The histogram of hardness values, shown in Fig. 7and obtained for a 3mN load, presents three peakscentered at 11.2, 15.7, and 22.3 GPa. The highestpeak, centered at 15.7 GPa, is ascribed to the Ti3SnC2

hardness.For each indentation load, the same statistical anal-

ysis has been performed over 100 indents to identify theTi3Sn(1�x)AlxC2 peak on hardness and elastic modulushistograms and to extract a mean value. Correspondinghardness and Young’s modulus values are plotted in Fig.8 as a function of the penetration depth. The strongindentation size effect (ISE) observed on the hardnessversus penetration depth curve is well fitted by the Nixand Gao53 model, except for the 150 and 300 mN loadnanoindentations. This model leads to a Ti3SnC2 bulkhardness of 9.3 GPa. The Young’s modulus remainsconstant (245 GPa) whatever the penetration depth is.

Such a result attests that the ISE is not an artifact due tothe indenter shape calibration. In the case of the 150and 300 mN load indentations, AFM observationsdemonstrate that, for such high loads, several grainsare involved in the deformation process during theindent formation implying that grain boundaries playan important role in macroscopic deformation.54

In such a case, which is close to microindentation testconditions, MAX phase hardness is underestimated asin hardness measurements performed using microin-dentation testing.

Such an analysis has been performed on differentTi3Sn(1�x)AlxC2 solid solutions, which contain 312phases in a sufficient volume fraction. SEM andEDXS were used to check that the indents used to de-termine hardness values and elastic modulus have beenobtained on a single grain of the Ti3Sn(1�x)AlxC2 solidsolutions. Table II shows the hardness and elastic mod-ulus variations as a function of the Al content in theTi3Sn(1�x)AlxC2 solid solutions; it also gives the Ti–Cmean distance and the Ti–A distance of the different

Table II. Hardness (H) and Young’s Modulus of Ti3Sn(1�x)AlxC2 Solid Solutions and Ti–C–Ti and Ti–ADistances as Deduced from the Rietveld Refinement

H (GPa) Young’s modulus (GPa) dTi�C�Ti(A) dTi�A (A)

Ti3SnC2 9.3 245 4.2547 3.0305Ti3Sn0.2Al0.8C2 10.2 251 4.3127 2.8530Ti3AlC2 11.4 260 4.3244 2.8134TiC 22–25 400–460 4.3247 —

Fig. 6. Histogram of elastic modulus determined from 500indents realized with a 3 mN load. Gaussian deconvolution of thehistogram gives the three peaks shown on the figure.

Fig. 7. Histogram, with a bin size of 500 MPa, of hardness valuesdetermined from 500 indents realized with a 3 mN load. Gaussiandeconvolution of the histogram gives the three peaks shown on thefigure.


samples. Firstly, one can observe that hardness valuesdeduced from nanoindentation experiments performedas a function of load (11.4 GPa for Ti3AlC2) are largerthan the currently determined microhardness values55–57

(in the range of 3–7 GPa for Ti3AlC2). Nevertheless,our value is in compatible agreement with the onesdetermined by nanoindentation performed on Ti3AlC2

thin films58 or on Ti3SiC2 bulk material.59 On theother hand, Table II shows that hardness and elasticmodulus slightly increase with Al content. Hardnessvalue (10.2 GPa) of the Ti3Sn0.2Al0.8C2 solutionsolution falls in between the ones of the of Ti3SnC2

(9.3 GPa) and Ti3AlC2 (11.4 GPa). Thus, solid solutionhardening effect is not operative in this system. Thesame kind of results has been obtained by Gangulyet al.28 in Ti3Si(1�x)GexC2 and by Zhou et al.56 inTi3Al1�xSixC2 solid solutions. Based on a quite limitedset of data, it appears that only substitutions on the X siteslead to solid solution hardening27; those on the Msites26,60 and A sites, as shown here and in references28 and 56, do not.

As MAX phases consist of ‘‘soft’’ M–A and ‘‘hard’’M–X bonds, it is reasonable that the strenghening of thesofter bond leads to an increase of the elastic modulus.The effect of intercalation of a weak M–A bond is two-fold: first, enhancement of room temperature ductility,toughness, and damage tolerance, and second degrada-

tion of strength and hardness. Therefore, optimizationsof elastic stiffness and strength of MAX phases have tofocus on the proper strengthening of the M–A bondsby tuning the chemical composition and the valenceelectron concentration.26,47,61–63 It has been shown theo-retically that the increase of p electrons from A atomswhen moving rightward along the periodic table results inan increase of the bulk moduli up to s2p3 column.61 Theshift of the Ti d–A p hybrid to lower energy and its higherfilling leads to stronger bonds. Nevertheless, Al (3s23p)and Sn (5s25p2) do not lie in the same period of theMendeleiev’s periodic table. Hug shows that Ti–Sn(2.975 A) and Ti–Al (2.898 A) bonds have about thesame bond stiffness in Ti2AlC and Ti2SnC MAXphases.47 In Ti3Sn0.2Al0.8C2, Ti–A bond lengths aremore different (see Table II) and hence Ti–A bondstiffness may also differ. As a consequence, the slightdifference observed in the elastic modulus of theTi3Sn(1�x)AlxC2 solid solutions may likely be explainedby the strength of the Ti–A bonds but more work isneeded to confirm such an hypothesis. On the otherhand, the distortion of the [Ti6C] octahedrons and theshrinking of the Ti–C covalent bonds may have someconsequences on the elastic properties of the materials.

Figure 9 shows the temperature variation of theTi3Sn0.2Al0.8C2 Young’s modulus. In the temperaturerange under study, the Young’s modulus decreases

Fig. 8. Ti3SnC2 Young’s modulus (open circle) and Ti3SnC2 hardness (squares) as a function of the indenter penetration depth. Dashed line:fit of the indentation size effect (see text).


slowly and almost linearly with the increasing temper-ature. The loss rate of Young’s modulus, in this tem-perature range, is about 30.8 MPa1/C, which is muchlower than that of Ti3Si(1�x)AlxC2 (49 MPa/1C).64 Theroom temperature Young’s modulus is 268 GPa, whichcorresponds to a 6% difference with the elastic modulusthat we obtain from nanoindentation experiments. Sucha slight difference may result from the small amountof impurities that are present in the Ti3Sn(1�x)AlxC2

solid solution. Moreover, a single-oriented grain is un-der study during nanoindentation experiments and evenif many experiments have been performed to extracta mean value among the different orientations, it ispossible that the statistics is not high enough to recoverthe mean Young’s modulus along all crystallographicorientations. Indeed, after HIP experiments, MAXphase grains are often oriented with basal planes parallelto the sample surface. As shown previously for Ti3AlC2,Ti3SiC2,65 Nb2AlC, and Nb4AlC3,

66 the Ti3Sn0.2Al0.8C2

Young’s modulus does not vary significantly with tem-perature in the temperature range under study. Radovicet al.52 suggested that the critical temperature for theaccelerated decrement of the elastic modulus is closeto the britlle-to-ductile transition temperature (BDTT)for Ti3SiC2 and Ti2AlC. For most MAX phases, such aBDTT is about 10001C.

Summary

Ti3Sn(1�x)AlxC2 MAX phase solid solutions havebeen successfully synthesized from different reactantmixtures. It is shown that 312 solid solutions are likely

formed from a reaction between TiC and the 211 solidsolution. Rietveld refinement allows to carefully char-acterize the structure and the ocathedra and trigonalprism distortion parameters as a function of the Al con-tent. It is shown that, for both Ti3SnC2 and Ti3AlC2,the hexagonal unit cell is strongly distorted. Neverthe-less, [Ti6C] octahedra are mainly distorted in Ti3SnC2

whereas [Ti6Sn(1�x)Alx] trigonal prisms are mainlydistorted in Ti3AlC2. Nanoindentation experimentsallow to determine hardness and elastic modulus ofsome Ti3Sn(1�x)AlxC2 solid solutions. It is demonstratedthat solid solution hardening is not operative in this sys-tem. Elastic modulus sligthly increases with Al content;such an increase may be attributed to the Ti–Al stifferbonds present in Ti3AlC2 than the Ti–Sn bonds inTi3SnC2. More work is needed to confirm such an hypo-thesis. Finally, Young’s modulus of the Ti3Sn0.2Al0.8C2

solid solutions is shown to slightly vary with temperaturein the range 20–4501C.

Acknowledgment

The authors thank A. Baudet for the technical as-sistance and T. Cabioc’h for fruitful discussions.

References

1. W. Jeitschko, H. Nowotny, and F. Benesovsky, ‘‘Kohlen-stoffhaltige ternareVerbindungen (H-Phase),’’ Monatsh. Chem., 94 672–676 (1963).

2. W. Jeitschko, H. Nowotny, and F. Benesovsky, ‘‘Die H-phasen Ti2TlC, Ti2PbC,Nb2InC, Nb2SnC und Ta2GaC,’’ Monatsh. Chem., 95 431–435 (1964).

3. W. Jeitschko and H. Nowotny, ‘‘Die Kristallstructur von Ti3SiC2-Ein NeuerKomplxcarbid-Typ,’’ Monatsh. Chem., 98 329–337 (1967).

4. H. Wolfsgruber, H. Nowotny, and F. Benesovsky, ‘‘Die Kristallstukturvon Ti3GeC,’’ Monatsh. Chem., 98 2403–2405 (1967).

5. V. H. Nowotny, ‘‘Strukturchemie einiger Verbindungen der Ubergangsmetallemit den elementen C, Si, Ge, Sn,’’ Prog. Solid State Chem., 5 27–70 (1971).

6. M. W. Barsoum and T. El-Raghy, ‘‘Synthesis and Characterization of a Re-markable Ceramic: Ti3SiC2,’’ J. Am. Chem. Soc., 79 [7] 1953–1956 (1996).

7. M. W. Barsoum, L. Farber, I. Levin, A. Procopio, T. El-Raghy, andA. Berner, ‘‘HRTEM of Ti4AlN3; or Ti3Al2N2 Revisited,’’ J. Am. Ceram.Soc., 82 [9] 2545–2547 (1999).

8. C. J. Rawn, M. W. Barsoum, T. El-Raghy, A. Procopio, C. M. Hoffmann,and C. Hubbard, ‘‘Structure of Ti4AlN3�x—a layered Mn11AXn Nitride,’’Mater. Res. Bull., 35 1785–1796 (2000).

9. M. W. Barsoum, ‘‘The MN11AXN Phases: A New Class of Solids Thermo-dynamically Stable Nanolaminates,’’ Prog. Solid State Chem., 28 201–281(2000).

10. M. A. Pietzka and J. C. Schuster, ‘‘Summary of Constitutional Dataon the Aluminum–Carbon–Titanium System,’’ J. Phase Equilib., 15 392–400 (1994).

11. J. Etzkorn, M. Ade, and H. Hillebrecht, ‘‘Ta3AlC2 and Ta4AlC3-SingleCrystal Investigation of Two New Ternary Carbides of Tantalum Synthesizedby the Molten Metal Technique,’’ Inorg. Chem., 46 1410–1418 (2007).

12. Z. J. Lin, M. J. Zhuo, Y. C. Zhou, M. S. Li, and J. Y. Wang, ‘‘Microstruc-tures and Theoretical Bulk Modulus of Layered Ternary Tantalum Alumi-num Carbides,’’ J. Am. Ceram. Soc., 89 3765–3769 (2006).

Fig. 9. Ti3Sn0.2Al0.8C2 Young’s modulus, measured by means ofthe dynamic resonant method in the bending mode, as a function oftemperature.


13. H. Hillebrecht and D. Kotzott, ‘‘Ternary Carbides Ti3MC2 (M 5 Al, Si, Ga,Ge)—Synthesis, Crystal Growth, Solid Solutions and Micro-Hardness Mea-surements,’’ Proceedings of the Materials Science and Technology Conference,Pittsburgh, PA, October 2009; 26–29.

14. Y. C. Zhou, F. L. Meng, and J. Zhang, ‘‘New MAX Phase Compounds in theV–Cr–Al–C System,’’ J. Am. Ceram. Soc., 91 1357–1360 (2008).

15. S. Dubois, T. Cabioc’h, P. Chartier, V. Gauthier, and M. Jaouen, ‘‘A NewTernary Nanolaminate Carbide: Ti3SnC2,’’ J. Am. Ceram. Soc., 90 [8] 2642–2644 (2007).

16. X. H. Wang and Y. C. Zhou, ‘‘Oxidation Behaviour of Ti3AlC2 at 1000–14001C in Air,’’ Corros. Sci., 45 891–907 (2002).

17. X. H. Wang and Y. C. Zhou, ‘‘High Temperature Oxidation of Ti2AlC inAir,’’ Oxid. Met., 59 303–320 (2003).

18. M. Sundberg, G. Malmqvist, A. Magnusson, and T. El Raghy, ‘‘AluminaForming High Temperature Silicides and Carbides,’’ Ceram. Int., 30 [7]1899–1905 (2004).

19. S. Becker, A. Rahmel, M. Schorr, and M. Schutze, ‘‘Mechanism of IsothermalOxidation of Intermetallic TiAl and of TiAl Alloys,’’ Oxid. Met., 38 [5–6]427–433 (1992).

20. R. G. Reddy, X. Wen, and M. Divakar, ‘‘Isothermal Oxidation of TiAlAlloy,’’ Metall. Mater. Trans. A, 32 [9] 2357–2361 (2001).

21. M. S. Chu and S. K. Wu, ‘‘The Improvement of High Temperature Oxi-dation of Ti–50Al by Sputtering Al Film and Subsequent InterdiffusionTreatment,’’ Acta Mater., 51 [11] 3109–3120 (2003).

22. M. P. Brady and P. F. Tortorelli, ‘‘Alloy Design of Intermetallics for Pro-tective Scale Formation and for Use as Precursors for Complex Ceramic PhaseSurfaces,’’ Intermetallics, 12 [7–9] 779–789 (2004).

23. Z. J. Lin, M. J. Zhuo, Y. C. Zhou, M. S. Li, and J. Y. Wang, ‘‘InterfacialMicrostructure of Ti3AlC2 and Al2O3 Oxide Scale,’’ Scr. Mater., 54 1815–1820 (2006).

24. Z. J. Lin, M. Li, and Y. C. Zhou, ‘‘TEM Investigations on Layered TernaryCarbides,’’ J. Mater. Sci. Technol., 23 [2] 145–165 (2007).

25. Z. J. Lin, M. Zhuo, Y. C. Zhou, M. Li, and J. Y. Wang, ‘‘Microstructures andAdhesion of the Oxide Scale Formed on Titanium Aluminium Carbide Sub-strates,’’ J. Am. Ceram. Soc., 89 [9] 2964–2966 (2006).

26. J. Y. Wang and Y. C. Zhou, ‘‘Ab Initio Elastic Stiffness of Nano-Laminate(MxM2�x)AlC)(M 5 Ti, V, Nb and Cr) Solid Solution,’’ J. Phys. Condens.Matter, 16 2819–2827 (2004).

27. M. W. Barsoum, M. Ali, and T. El-Raghy, ‘‘Processing and Characterizationof Ti2AlC, Ti2AlN, and Ti2AlC0.5N0.5,’’ Metall. Mater. Trans., 31A [7]1857–1865 (2000).

28. A. Ganguly, T. Zhen, and M. W. Barsoum, ‘‘Synthesis and MechanicalProperties of Ti3GeC2 and Ti3(SixGe1�x)C2 (x 5 0.5, 0.75) Solid Solutions,’’J. Alloys Comp., 376 287–295 (2004).

29. P. Finkel, et al., ‘‘Electronic, Thermal, and Elastic Properties of Ti3Si1xGexC2

Solid Solutions,’’ Phys. Rev. B, 70 085104 (2004).30. Z. G. Liu, J. T. Guo, L. L. Ye, G. S. Li, and Z. Q. Hu, ‘‘Formation Mech-

anism of TiC by Mechanical Alloying,’’ Appl. Phys. Lett., 65 [21] 2666–2668(1994).

31. S. Dubois, E. Heian, N. Karnatak, M. F. Beaufort, and D. Vrel, ‘‘Synthesis ofTiC/Ni Cermets Via Mechanically Activated Self Propagating High Tem-perature Synthesis,’’ Mater. Sci. Forum, 426–432 2033–2038 (2003).

32. L. Lutterotti, S. Matthies, and H.-R. Wenk, ‘‘MAUD: a Friendly Java Pro-gram for Material Analysis Using Diffraction,’’ IUCr: Newsletter of the CPD21, 1999; 14.

33. J. Woirgard and J. C. Dargenton, ‘‘An Alternative Method for PenetrationDepth Determination in Nanoindentation Measurements,’’ J. Mater. Res., 12[9] 2455–2458 (1997).

34. G. M. Pharr and A. Bolshakov, ‘‘Understanding Nanoindentation UnloadingCurves,’’ J. Mater. Res., 17 2660–2671 (2002).

35. ASTM (2001) ‘Standard Test Method for Young’s Modulus, Shear Modulusand Poisson’s Ratio by Impulse Excitation of Vibration,’ Annual Book ofASTM Standards, Vol. 03.01, pp. 1099–1112.

36. P. Mazot, J. de Fouquet, J. Woirgard, and J. P. Pautrot, ‘‘Mesure des const-antes d’elasticite et du frottement interieur des solides par methode deresonance entre 20 et 12001C,’’ Journal de Physique III- May, 751–763(1992).

37. P. Gadaud and S. Pautrot, ‘‘Characterization of the Elasticity and Anelasticityof Bulk Glasses by Dynamical Subresonant and Resonant Techniques,’’J. Non-Cryst. Solids, 316 145–152 (2003).

38. C. Peng, C. A. Wang, Y. Song, and Y. Huang, ‘‘A Novel Simple Method toStably Synthesize Ti3AlC2 Powder with High Purity,’’ Mater. Sc. And Eng. A,428 54–58 (2006).

39. Z. J. Lin, M. J. Zhuo, Y. C. Zhou, M. S. Li, and J. Y. Wang, ‘‘MicrostructuralCharacterization of Layered Ternary Ti2AlC,’’ Acta Mater., 54 1009–1015(2006).

40. Z. J. Lin, M. J. Zhuo, Y. C. Zhou, M. S. Li, and J. Y. Wang, ‘‘MicrostructuralRelationship Between Compounds in the Ti–Si–C System,’’ Scr. Mater., 55445–448 (2006).

41. Z. J. Lin, M. J. Zhuo, M. S. Li, J. Y. Wang, and Y. C. Zhou, ‘‘Synthesis andMicrostructure of Layered-Ternary Ti2AlN Ceramic,’’ Scr. Mater., 56 1115–1118 (2007).

42. A. Zhou, C. A. Wang, and Y. Huang, ‘‘A Possible Reaction Mechanism onSynthesis of Ti3AlC2,’’ Mater. Sci. Eng. A, 352 333–339 (2003).

43. Z. Ge, K. Chen, J. Guo, H Zhou, and J. M. F. Ferreira, ‘‘Combustion Syn-thesis of Ternary Carbide Ti3AlC2 in Ti–Al–C System,’’ J. Eur. Ceram. Soc.,23 567–574 (2003).

44. J. H. Han, S. S. Hwang, D. Lee, and S. W. Park, ‘‘Synthesis and MechanicalProperties of Ti3AlC2 by Hot Pressing TiCx/Al Powder Mixture,’’ J. Eur.Ceram. Soc., 28 979–988 (2008).

45. K. T. Jacob, S. Raj, and L. Rannesh, ‘‘Vegard’s Law: A Fundamental Relationor an Approximation,’’ Int. J. Mater. Res., 9 776–779 (2007).

46. J. Etzkorn, M. Ade, and H. Hillebrecht, ‘‘V2AlC, V4AlC3�x (x � 0.31), andV12Al3C8: Synthesis, Crystal Growth, Structure, and Superstructure,’’ Inorg.Chem., 46 [18] 7646–7653 (2007).

47. G. Hug, ‘‘Electronic Structures of and Composition Gaps among the TernaryCarbides Ti2MC,’’ Phys. Rev. B, 74 184113 (2006).

48. G. Hug, M. Jaouen, and M. W. Barsoum, ‘‘X-Ray Absorption Spectroscopy,EELS, and Full-Potential Augmented Plane Wave Study of the ElectronicStructure of Ti2AlC, Ti2AlN, Nb2AlC, and (Ti0.5Nb0.5)2AlC,’’ Phys. Rev. B,71 024105 (2005).

49. K. Momma and F. Izumi, ‘‘VESTA: A Three-Dimensional VisualizationSystem for Electronic and Structural Analysis,’’ J. Appl. Crystallogr., 41 653–658 (2008).

50. M. Gamarnik and M. W. Barsoum, ‘‘Bond Lengths in the TernaryCompounds Ti3SiC2, Ti3GeC2 and Ti2GeC,’’ J. Mater. Sci., 34 169–174(1999).

51. O. Wilhelmsson, et al., ‘‘Deposition and Characterization of Ternary ThinFilms within the Ti–Al–C System by DC Magnetron Sputtering,’’ J. Cryst.Growth, 291 [1] 290–300 (2006).

52. M. Radovic, et al., ‘‘On the Elastic Properties and Mechanical Damping ofTi3SiC2, Ti3GeC2, Ti3Si0.5Al0.5C2 and Ti2AlC in the 300–1573 K Temper-ature Range,’’ Acta Mater., 54 [10] 2757–2767 (2006).

53. W. D. Nix and H. Gao, ‘‘Indentation Size Effects in Crystalline Materials:A Law for Strain Gradient Plasticity,’’ J. Mech. Phys. Solids, 46 [3] 411–425(1998).

54. C. Tromas, N. Ouabadi, V. Gauthier-Brunet, M. Jaouen, and S. Dubois,‘‘Mechanical Properties of Nanolaminate Ti3SnC2 Carbide Determined byNanohardness Cartography,’’ J. Am. Ceram. Soc., 93 [2] 330–333 (2010).

55. N. V. Tzenov and M. W. Barsoum, ‘‘Synthesis and Characterization ofTi3AlC2,’’ J. Am. Ceram. Soc., 83 [4] 825–832 (2000).

56. Y. C. Zhou, J. X. Chen, and J. Y. Wang, ‘‘Strengthening of Ti3AlC2 byIncorporation of Si to form Ti3Al1�xSixC2 Solid Solutions,’’ Acta Mater.,54 1317–1322 (2006).

57. J. H. Han, S. S. Hwang, D. Lee, and S. W. Park, ‘‘Synthesis and MechanicalProperties of Ti3AlC2 by Hot Pressing TiCx/Al Powder Mixture,’’ J. Eur.Ceram. Soc., 28 979–988 (2008).

58. O. Wilhelmsson, et al., ‘‘Deposition and Characterization of Ternary Thinlms within the Ti–Al–C System by DC Magnetron Sputtering,’’ J. Crys.Growth, 291 290–300 (2006).

59. J. M. Molina-Aldareguia, J. Emmerlich, J.-P. Palmquist, U. Jansson, andL. Hultman, ‘‘Kink Formation Around Indents in Laminated Ti3SiC2 thinFilms Studied in the Nanoscale,’’ Scr. Mater., 49 155–160 (2003).

60. I. Salama, T. El Raghy, and M. W. Barsoum, ‘‘Synthesis and MechanicalProperties of Nb2AlC and (Ti,Nb)2AlC,’’ J. Alloys Comp., 347 [1–2] 271–278(2002).

61. Z. M. Sun, R. Ahuja, and J. M. Schneider, ‘‘Theoretical Investigation of Solubilityin (MxM

02�x)AlC (M and M0 5 Ti, V, Cr),’’ Phys. Rev. B, 68 224112 (2003).

62. R. S. Kumar, S. Rekhi, A. L. Cornelius, and M. W. Barsoum, ‘‘Compress-ibility of Nb2AsC to 41 Gpa,’’ Appl. Phys. Lett., 86 111904 (2005).


63. T. Liao, J. Y. Wang, and Y. C. Zhou, ‘‘Superior Mechanical Properties ofNb2AsC to Those of other Layered Ternary Carbides: A First-PrinciplesStudy,’’ J. Phys. Condens. Matter, 18 L527–L533 (2006).

64. Y. C. Zhou, D. T. Wan, Y. W. Bao, and J. Y. Wang, ‘‘In Situ Processing andHigh-Temperature Properties of Ti3Si(Al)C2/SiC Composites,’’ Int. J. Appl.Ceram. Technol., 3 [1] 47–54 (2006).

65. Y. W. Bao and Y. C. Zhou, ‘‘Evaluating High-Temperature Modulus andElastic Recovery of Ti3SiC2 and Ti3AlC2 Ceramics,’’ Mater. Lett., 57 4018–4022 (2003).

66. J. Wang and Y. C. Zhou, ‘‘Recent Progress in Theoretical Prediction, Prep-aration and Characterization of Layered Ternary Transition-Metal Carbides,’’Annu. Rev. Mater. Res., 39 415–443 (2009).


Synthesis, Microstructure, and Mechanical Properties of Ti3Sn(1−x)AlxC2 MAX Phase Solid Solutions

Documents

Transcript of Synthesis, Microstructure, and Mechanical Properties of Ti3Sn(1−x)AlxC2 MAX Phase Solid Solutions