MAX phase - Alumina composites via exchange reaction in the Mn+1AlCn systems (M=Ti, V, Cr, Nb, or Ta)

Article abstract of DOI:10.1016/j.jssc.2015.10.024

MAX phases have been produced for the first time via an exchange reaction between the M-element oxide and Al leading to an M-Al-C-Al₂O₃ composite in the V-Al-C, Cr-Al-C, Nb-Al-C and Ta-Al-C systems in addition to the previously known

Full text of DOI:10.1016/j.jssc.2015.10.024

Journal of Solid State Chemistry 233 (2016) 150–157
Contents lists available at ScienceDirect
Journal of Solid State Chemistry
journal homepage: www.elsevier.com/locate/jssc
MAX phase – Alumina composites via exchange reaction in the
M
nþ1AlCn systems (M¼Ti, V, Cr, Nb, or Ta)
n
Dylan T. Cuskelly , Erich H. Kisi, Heber O. Sugo
School of Engineering, The University of Newcastle, Callaghan, NSW 2308, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
MAX phases have been produced for the first time via an exchange reaction between the M-element
oxide and Al leading to an M–Al–C–Al2O3 composite in the V–Al–C, Cr–Al–C, Nb–Al–C and Ta–Al–C
systems in addition to the previously known Ti–Al–C system. The reduction reaction was first in-
vestigated by forming the binary M–X carbide and then proven to be generic across all M–Al–C systems
with the production of the M2AlC phase in each case. The work was extended to the other M3AlC2 and
M4AlC3 phases in the respective systems, and was successful in 4 of the 5 cases with moderate yield.
& 2015 Elsevier Inc. All rights reserved.
Received 15 August 2015
Received in revised form
9 October 2015
Accepted 14 October 2015
Keywords:
MAX phases
Exchange reaction
Synthesis
Aluminothermic
1. Introduction
ceramic properties making them attractive engineering materials
[3]. Material properties normally associated with ceramics such as
The MAX phases are an interesting group of ternary carbides
and nitrides first discovered in 1963 [1] and first produced as bulk
single phase materials in 1996 [2]. They have been under intense
study ever since due to their interesting material properties [3–5].
MAX phases are of the form Mnþ1AXn, where M is an early tran-
sition metal, A is a group 3 or 4 element, X is either C or N and n is
an integer either 1, 2 or 3 resulting in what are termed 211, 312 or
413 phases respectively [2]. There are in total nine known M-ele-
ments, twelve A-elements and two X-elements, as well as three
common values of n [6]. The MAX phases have a layered structure
comprised of MX octahedra separated by layers of the A-element;
the integer n indicates the number of octahedra in each layer. Fig. 1
displays the structure of a 312 MAX phase. Not all possible M–A–X
combinations are thermodynamically stable [7]. Out of the 216
possible combinations about 90 MAX phases have been discovered
so far [6,8]. By far the most commonly studied MAX phases are
Ti3SiC2, Ti3AlC2 and Ti2AlC. These were some of the first to be
synthesised by Barsoum et. al [2] as monolithic polycrystals and
have been the subject of the majority of synthesis investigations
and property testing [9]. The systems that form carbide MAX
phases are summarised in Table 1 below [10]. It can be seen that
there is a strong bias towards: the 211 phases; on Ti as the M-
element (row 1); and Al as the A-element (column 1).
high specific stiffness and high temperature oxidation resistance
are coupled with the metallic properties of high electrical and
thermal conductivity as well as fracture tolerance [11,12]. These
properties are present in combination with the ability to be easily
machined and are just some of the interesting engineering prop-
erties of the MAX phases which have been presented in detail
elsewhere [2–5]. However, difficulties in the cost effective synth-
esis of high purity bulk products have severely limited the appli-
cations of MAX phases. Phase stability in these ternary systems is
often quite complex with many factors, both thermodynamic and
kinetic, influencing the conversion of reactants into particular
products [13–15].
In general, ceramic synthesis involves producing large quan-
tities of a phase-pure powdered starting material which is suitable
for downstream processing into components by various shaping
processes (uniaxial pressing, cold isostatic pressing, extrusion
etc.). Shaping is then followed by sintering to produce dense,
strong components. Some processes such as hot pressing or hot
isostatic pressing combine these two stages into a single step. In
other cases, synthesis is combined with the shaping and/or sin-
tering steps (via reactive sintering, reactive hot pressing, reactive
hot isostatic pressing etc.) such that unreacted starting powders
undergo solid-state reactions to form the desired ceramic com-
pound. At the same time, plastic flow and sintering act to produce
the desired final shape. Current MAX phase synthesis methods are
primarily of the latter kind. Methods, such as reactive sintering
(hot iso-static pressing or hot uni-axial pressing) and pulse dis-
charge sintering are effective in the production of high purity, high
MAX phases have a unique combination of metallic and
n
Corresponding author.
E-mail address: Dylan.cuskelly@uon.edu.au (D.T. Cuskelly).
http://dx.doi.org/10.1016/j.jssc.2015.10.024
0022-4596/& 2015 Elsevier Inc. All rights reserved.

D.T. Cuskelly et al. / Journal of Solid State Chemistry 233 (2016) 150–157
151
In an effort to control intermediate phase formation and ulti-
mately prevent unwanted phase retention, non-elemental starting
materials are often utilised. MAX phases have been synthesised
from, AX compounds (e.g. SiC, Al4C3) [14,22,23] or MXx compounds
(TiCx) [24]. The use of AX compounds appears beneficial in con-
trolling the release rate of the A-element, hence reducing sub-
limation losses before reaction.
Recently TiO2 was successfully used as an alternative starting
material in MAX phase formation. The use of the metal oxide as a
starting material is particularly interesting as almost all potential
engineering applications of MAX phases are hampered by the high
material cost. This is due in part to the synthesis difficulties
mentioned above but also to the high cost of metallic starting
materials. Specifically all of the M-elements involved in the for-
mation of MAX phases are difficult to refine and are inherently
expensive, compared with the M-element oxides.
The M-elements used in MAX phase formation are either found
directly as oxides in their ore, or easily refined from the ore into an
oxide (e.g. TiO2 is usually found either as rutile or refined from
Ilmenite). Normally the pure element is then produced from this
oxide via carbothermal or aluminothermic reduction. Carbother-
mal reduction is the industry preferred method of refinement as
carbon is less expensive than aluminium; however some metals
cannot be obtained this way. Aluminothermic reduction is used in
the refinement of both V from the oxide V2O5 and Cr from Cr2O3.
In the case of Cr refinement, carbothermal reduction leads to the
formation of Cr2C3, thus aluminothermic reduction is the preferred
method. Aluminothermic reduction of FeO known as a thermite
reaction is common in railway track welding.
It has been demonstrated in the literature that aluminothermic
reduction can be modified into an exchange reaction which com-
bines M-element reduction and carbide formation in a single-step
self-propagating high temperature synthesis (SHS) reaction giving
TiC and Al2O3 [25] according to:
3TiO2þ4Alþ3C-2Al2O3þ3TiC
(1)
Modification of the stoichiometry by the addition of excess Al
allowed extension to the SHS production of Ti–Al–C MAX
phase-Al2O3 composites mixed according to Eq. (2) [26,27].
3TiO2þ5Alþ2C-2Al2O3þTi3AlC2
(2)
In that work, Chen et al. were successful in demonstrating that
Ti3AlC2 can be synthesised via a single step reduction reaction or
exchange reaction, and the conversion rate was similar to that
often seen in SHS experiments. Products contained both the ex-
pected alumina and Ti3AlC2 along with TiC. Attempts to avoid TiC
formation by modification of reactant ratios was only successful in
limiting the TiC to approximately 5%. Additional modification
introduced additional phases, primarily TiAl3. Nonetheless this
process opens a new route for the production of MAX phase
powders.
The exchange reaction concept has been extended to the pro-
duction of Ti3SiC2–Al2O3 composites via the reduction of TiO2 by Al
in the presence of Si and C [28]. This was an important demon-
stration that showed the process can be applied to systems where
the A-element of the MAX phase and the reducing element are
different i.e. the process can potentially be applied to the phases in
row 2 of Table 1.
Fig. 1. Structure of a 312 MAX phase. Green octahedra have M atom corners and X
atom centres, white atoms separating the octahedra are A-element layers. The two
octahedra in each layer indicate this is a n¼2 or 312 MAX phase. (For interpretation
of the references to color in this figure legend, the reader is referred to the web
version of this article.)
density samples [2,16–19]. These techniques are effective in pro-
ducing laboratory scale cylindrical samples for mechanical prop-
erty testing and have been fundamental in understanding the
formation mechanisms of MAX phases. However, methods which
combine the synthesis and shaping steps are inherently small scale
and unsuited to large scale powder production.
Self-propagating high-temperature synthesis (SHS) is an ap-
pealing method of producing bulk quantities of powders due to its
low input energy and fast reaction times [14,20]. However SHS is
prone to complications due to a lack of control over the reaction
parameters. Problems include extreme temperatures which result
in unequal vapour phase loss of reactants affecting product stoi-
chiometry in systems that are highly sensitive to minor composi-
tional differences. There is also a propensity for the retention of
intermediate phases as the rapidly falling post reaction tempera-
ture does not allow sufficient time for reactions to reach com-
pletion [21].
To date no MAX phases outside of the Ti systems have been
synthesised via a single step exchange reaction, however Al2O3 has
a lower Gibbs free energy of formation per oxygen atom than any
of the M-element oxides [29] as presented in Table 2, and there-
fore should be capable of reducing all of them. The work presented
here expands the concept of the exchange reaction synthesis of
Mnþ1AlXn phases to different M-element oxides using Al as both

152
D.T. Cuskelly et al. / Journal of Solid State Chemistry 233 (2016) 150–157
Table 1
Systematics of MAX phase formation, showing known carbide MAX phases arranged by M and A-element. Higher order phases i.e. 312 and 413 phases are shown in darker
grey and can be seen to exclusively occupy the Al column and the Ti row [10].
Table 2
some of the higher order phases require far more specific condi-
tions. In many cases stable conditions for the higher order phases
have not been found and may not exist. A 211 type MAX phase has
been reported in all five of the systems investigated here while
there are only three 312 and three 413 phases.
In this paper, all of the carbide MAX phase forming systems in
the first column of Table 1 are investigated. Optimisation of the
synthesis procedure and phase purity of a single system was not
the overall goal in this work. Instead the work is focused on de-
termining whether the exchange reactions occurs in these M-
element systems, and if so, which MAX phases form in a single
step pressureless sintering process. Formation of the relevant 211
phase was the success criterion used for MAX phase formation
however the higher order phases were also investigated to test the
generality of the process.
Thermodynamic data for the oxides used in this work [29].
Oxide
ΔH (per O) kJ/mol
ΔS (per O) J
G at 1500 °C (per O) kJ/mol
Al2O3
TiO2
V2O5
Cr2O3
Nb2O5
Ta2O5
ꢀ558.7
ꢀ472
17
ꢀ588.8
ꢀ517.2
ꢀ356.6
ꢀ427.8
ꢀ391.3
ꢀ459.9
25.5
26.2
27
15.3
28.6
ꢀ310.2
ꢀ380
ꢀ368.4
ꢀ409.2
the reducing agent and as the A-element. Consequently, this work
will investigate column 1 of Table 1. The M-elements all have
multiple oxidation states. In general the highest oxidation state is
the most common and was therefore used throughout this work.
The reduction of the oxide does not necessarily mean that the
MAX phase can form under typical conditions of dry-pressed
4-component powders such as these. Formation of MAX phases is
quite complex with specific reaction conditions required for each
phase. The optimum synthesis conditions range in temperature
from 1200 °C to 1650 °C and times from minutes to many hours,
depending on the elements present and the desired n value. Cer-
tain MAX phases also thermally decompose at temperatures which
are lower than the synthesis temperatures of other MAX phases
[30,31]. For example, Ti2GeC decomposes at 1300 °C whereas
Nb2AlC only forms above 1600 °C [32]. The problem of a stable
equilibrium for synthesis is compounded by partial loss of the A-
element due to vapour phase transport [10]. The 211 phases gen-
erally tolerate the widest range of formation conditions whereas
2. Method
For all systems identified in column 1 of Table 1, samples were
initially made using stoichiometric mixing ratios as described by
Eqs. (3)–(12). Samples were first made according to Eqs. (3), (5),
(7), (9) and (11) aimed at producing the binary carbide to de-
termine in the simplest cases, whether the exchange reaction
would occur for each M-element. Following this, compositions
intended to produce 211 MAX phases, Eqs. (4), (6), (8), (10) and
(12), were investigated.
Samples were prepared using TiO2 (þ99.9%, o5
μ
m), Cr2O3
(þ99%, o45 m), Nb2O5 (þ99.9%, o45 m), V2O5 (þ99%,
μ
μ

D.T. Cuskelly et al. / Journal of Solid State Chemistry 233 (2016) 150–157
153
o45
μ
m), Ta2O5 (99.99%, o45
μ
m), Al (þ99%, o45
μ
m) and
The V system was unusual in that the powders reacted in the
ball mill before being placed into the furnace. A lower milling time
of 3 min and lower charge ratio of 3.5:1 was used in subsequent
milling to prevent reaction in the mill and allow the reaction to
occur in the furnace.
graphite (þ99.8%, o100
μ
m) powders mixed according to Eqs. (3)–
(12). Due to a number of factors most notably A-element losses, stoi-
chiometric mixing ratios often lead to poor conversion of the products
to MAX phase. In cases where conversion was so poor that MAX phase
formation was negligible, or vast improvement could be readily
achieved, the initial mixing ratios were varied. In the Ta–Al–C and V–
Al–C systems the mixing ratio was modified to 100% M-element oxide,
120%Al and 95%C compared with the stoichiometric quantities in or-
der to produce significant amounts of MAX phase. As this work was
focused on assessing the feasibility for the reaction to take place rather
than high purity products, further optimisation of starting ratios to
produce phase-pure products was generally not pursued.
The fired pellets were sectioned and XRD patterns (Cu K ) re-
α
corded using a Panalytical ExPert™. XRD data were analysed via
the Rietveld method to determine lattice parameters and atomic
positions whilst the refined scale factors were used for quantita-
tive phase analysis according to the method given by Hill and
Howard [37]. In these refinements, the background, zero offset,
peak width parameters and scale factors of all phases were
refined.
The Cr2AlC system displayed high initial conversion and was
selected as a system to undergo optimisation (i.e. no binary car-
bides or intermetallic compounds present in products). The mixing
ratio, 100% M-element oxide, 115%Al and 100%C of stoichiometric
was found to eliminate unwanted binary carbides from the pro-
ducts, however this was at the expense of Cr5Al8 formation.
For each composition, 10 g of the reactants were ball milled for
15 min (SPEX 8000, steel vial and balls charge ratio 7:1). The powder
was halved and uniaxially pressed at 250 MPa into two pellets 15 mm
diameter and ꢁ10 mm high. Pellets resting in a graphite boat under
flowing argon were heated at 5 °C/min to the appropriate sintering
temperature and held for the associated time in accordance with
Table 3.
3. Results
3.1. Binary carbides
Samples targeting the binary carbides all produced the desired
products thus showing the exchange reaction took place in all
cases. This was indicated by the joint absence of M-element oxides
and presence of Al2O3. The Al2O3 along with the expected carbide
formed the major phases in all samples. Smaller amounts of other
compounds were found in some samples (most notably Cr3C7 and
unreacted graphite). Fig. 2 contains XRD results for samples made
according to Eqs. (3), (5), (7), (9) and (11) highlighting the Al2O3
and carbide phases. Overall the exchange reaction was deemed
3TiO2þ4Alþ3C-3 TiCþ2Al2O3
6TiO2þ11Alþ3C-3Ti2AlCþ4Al2O3
3V2O5þ10Alþ6C-6VCþ5Al2O3
3V2O5þ13Alþ3C-3V2AlCþ5Al2O3
3Cr2O3þ6Alþ3C-2Cr3C2þ3Al2O3
Cr2O3þ3AlþC-Cr2AlCþAl2O3
(3)
(4)
(5)
(6)
(7)
(8)
3Nb2O5þ10Alþ6C-6NbCþ5Al2O3
3Nb2O5þ13Alþ3C-3Nb2AlCþ5Al2O3
3Ta2O5þ10Alþ6C-6TaCþ5Al2O3
3Ta2O5þ13Alþ3C-3Ta 2AlCþ5Al2O3
(9)
(10)
(11)
(12)
Due to the reasons outlined in Section 1, a generic firing se-
quence was not suitable. Instead a heating cycle was determined
for each system from the present state of knowledge in the lit-
erature using other synthesis methods and the cumulative
knowledge gained throughout the work [31,33–36]. As many of
these compounds have not been produced via pressureless re-
active sintering before, and almost none via an exchange reaction,
the sequences used here are not considered to be optimal.
Table 3
Sintering conditions used for different systems.
Systems
Sintering temperature (°C)
Sintering time (h)
Ti, V, Cr
Eqs. (2)–(7)
Nb
Eqs. (8) and (9)
Ta
1400
1600
1550
3
12
0.5
Fig. 2. XRD data of the binary carbide systems. Intensity has been normalised in all
scans. Only selected signature peaks are marked for each phase although all peaks
were identified.
Eqs. (10) and (11)

154
D.T. Cuskelly et al. / Journal of Solid State Chemistry 233 (2016) 150–157
Fig. 4. Rietveld refinement of the Cr2AlC system. Black ‘þ’ represent measured
data, the red line is the calculated fit, and the green line is the difference plot.
Markers represent expected peak positions from the phases Al2O3, Cr2AlC and
Cr5Al8 from top to bottom. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)
3.3. 312 and 413 MAX phase formation
A surprisingly large proportion of the known higher order
phases (312 or 413) are contained within in the M–Al–C systems
studied here (three of the eight 312 phases, and three of the seven
413 phases). Of the six higher order phases (Ti3AlC2, V3AlC2,
V4AlC3, Nb4AlC3, Ta3AlC2, and Ta4AlC3) five were ultimately formed
by exchange reaction, Ta3AlC2 being unsuccessful. The phases
Ti3AlC2, Nb4AlC3 and Ta4AlC3 all formed within samples for-
mulated to produce the M2AlC phase, denoted by the ♦ and ■
markers in Fig. 2. The V3AlC2 and V4AlC3 phases were found in a
sample originally mixed to form VC and Al2O3 as shown in Fig. 4.
The Cr–Al–C system does not contain any known stable higher
order phases. These results show that the exchange reaction can
form 312 and/or 413 MAX phases in all systems where such phases
are known.
Fig. 3. 211 MAX phase XRD data, higher order phases can also be seen in the Ti–Al–
C, Nb–Al–C and Ta–Al–C systems. Each MAX phase can be identified by their
characteristic peak in the 5–15 ° 2θ range. All peaks were identified although only
selected peaks are marked.
successful for these systems under the conditions used.
3.2. 211 MAX phase formation
It was found that all of the known M2AlC MAX phases (i.e.
Ti2AlC, V2AlC, Cr2AlC, Nb2AlC and Ta2AlC) could be successfully
synthesised via a single step exchange reaction using the pres-
sureless reactive sintering method. XRD analysis of samples made
according to Eqs. (4), (6), (8), (10) and (12) are shown in Fig. 3 with
the characteristic 211 MAX phase peaks easily seen in the region
3.4. V–Al–C formation
Contrary to other systems the V2O5, Al and C powders reacted
during the milling process. XRD analysis of the milled reactants
prepared according to Eq. (6) (Fig. 5 lower) proved they had suc-
cessfully undergone an exchange reaction and produced a sig-
nificant amount of the desired V2AlC MAX phase. The milling time
and charge ratio was then lowered for all other samples containing
V2O5 to prevent reaction in the mill.
Even after milling parameters had been adjusted, visual in-
spection of a sample mixed according Eq. (5) targeting VC and
Al2O3, after firing, indicated that the sample had undergone an
SHS reaction in the furnace. XRD analysis of this sample, shown in
Fig. 5 upper, revealed the intended VC was not produced however
it had produced a mixture of 3 MAX phases (V2AlC, V3AlC2 and
V4AlC3) and Al2O3, hence proving that all of these phases can be
formed via an exchange reaction (albeit under SHS conditions).
Other samples prepared under identical conditions but which did
not undergo SHS, produced the intended VC and Al2O3.
12–14° 2θ highlighted by (▼). Along with Al2O3, many of the sys-
tems also produced a mixture of other phases including binary
carbides, intermetallic compounds and/or other MAX phases
which are highlighted on each XRD individually.
Rietveld refinements were carried out on all systems to confirm
the presence of each MAX phase. An example refinement plot for
the optimised Cr2AlCþAl2O3 using the modified mixing ratio 100%
Cr2O3, 115% Al and 100% C is shown in Fig. 4. The modification was
found to have eliminated the large amount of binary carbide Cr3C2
which was present in stoichiometric samples. Unfortunately an
intermetallic compound, Cr5Al8, had formed as a secondary phase
using these modified mixing ratios. Final quantities were de-
termined by Rietveld analysis to be 36 mol% Al2O3 and 55 mol%
Cr2AlC, 9 mol% Cr5Al8, with no other phases found at the detection
limit of the XRD for these systems. It is expected there is an op-
timum mixing ratio between stoichiometric and what was used
here, which would result in a minimum of additional phases.

D.T. Cuskelly et al. / Journal of Solid State Chemistry 233 (2016) 150–157
155
For each of the different M–Al–C systems, each half reaction
can proceed at its own rate on the proviso that the rate of MAX
phase formation cannot exceed the rate of oxygen exchange. The
rates are governed by a complicated interplay of many parameters.
First, the thermodynamic drive is quite different for each system as
may be seen for the oxide decomposition in Table 2. Second, the
reactants are in the form of a mixed powder with local variations
in composition and the density of inter-particle contacts of a
particular type. For example, the oxygen exchange half-reactions
such as Eq. (13) rely upon direct contact between M-oxide parti-
cles and Al droplets.2 This microstructural aspect is strongly in-
fluenced by the particle size, particle shape, green density and the
deformability of the particles. Third, there are likely to be highly
localised temperature variations due to the liberation of the half-
reaction enthalpies at reacting inter-particle contacts. Examination
of Table 2 indicates that this is likely to vary widely between
the different systems studied. A number of features of the res-
ults presented here may be discussed within this qualitative
framework.
The formation of mixed MAX phases occurred in every system
where more than one MAX phase was available. A-element defi-
cient phases are common in MAX phase production and in stan-
dard methods this is normally observed as the MX binary. How-
ever higher order MAX phases are also A-element deficient var-
iants of M2AlC phases, having a closer stoichiometry with respect
to both the M–A ratio and the M–C ratio. A-element losses account
for both the presence of the higher order phases in the 211 for-
mulations of this study and the need to modify starting powder
ratios to achieve enhanced yields in some systems. They are be-
lieved to be the result of vapour transport.
The equilibrium vapour pressure of Al at the temperatures in-
dicated in Table 3 lies in the moderate range 9.7 Pa (M¼Ti, V, Cr)
to 122 Pa (M¼Nb). However, as briefly discussed previously [10],
these are far from closed systems. In a reactive sintering experi-
ment such as these, Al vapour is highly mobile and reactive. It is
continually flushed from the furnace chamber by the flowing ar-
gon atmosphere as well as adsorbing on chamber walls and con-
densing on cooler parts of the furnace system remote from the hot
zone. In a given furnace environment, the degree of A-element
deficiency observed therefore depends upon a combination of the
(overallþlocal) synthesis temperature as well as the hold time.
The temperature, which governs the equilibrium vapour pressure,
is constrained by the stability range and formation kinetics of the
MAX phase and any intermediate phases which may preceed MAX
phase formation. In the Ta and V systems, Al losses were so great
that very little MAX phase was observed in the products prior to
adjustment of the reactant ratios. Although these two systems
suffered from the same problem and both benefited from the same
remedy, the underlying reasons may be different. Preliminary ex-
perimentation in the Ta system with sintering temperatures of
1400 °C (not shown in this work) failed to produce any MAX
phase. When the temperature was raised to 1550 °C synthesis was
successful. Conversely, in the V system, the reactivity is so high
there was a propensity to undergo spontaneous SHS reaction. This
suggests that in the latter system, even under macroscopically
non-SHS conditions, locally high temperatures are caused by the
rapid oxygen exchange reaction and this leads to an excess of Al
loss. In the Ta system, it may be that 1550 °C is at the upper end of
the MAX phase stability range.
Fig. 5. XRD data of the V–Al–C system (top) after SHS in the furnace and (bottom)
MASHS within the ball mill. The V3AlC2 and V4AlC3 phases can be easily identified
in the 5–15° 2θ range. Again only selected peaks are marked.
4. Discussion
Table 4 presents a summary of the M–Al–C MAX phases able to
be produced via an exchange reaction in this work. Overall the
method is applicable to all systems analysed in this study and
indeed almost every composition, indicating the exchange reac-
tion is a very general synthesis technique. The complete absence of
the Ta3AlC2 phase may be due to the sintering conditions being
unsuitable however Ta3AlC2 has only ever been synthesised as
small single crystals and may not form bulk polycrystals [38]. Gi-
ven the success of all other systems, and the formation of both
Ta2AlC and Ta4AlC3, it is unlikely that the exchange reaction
methodology is the reason for the failure to produce Ta3AlC2.
It is qualitatively useful to consider the exchange reactions to
proceed via two half-reactions although the exact reaction me-
chanism is likely to be more complex and would require extensive
rapid in situ diffraction studies to observe.1 The two half-reactions
comprise M-oxide reduction by Al oxidation (or oxygen exchange)
followed by MAX phase formation. For example in the Ti–Al–C
system, Eq. (4) can be written:
6TiO2þ8Al-6Tiþ4Al2O3
(13)
In general, optimisation of a MAX phase system is a compro-
mise between the (often unknown) stability range of the desired
product phase, the kinetics of MAX phase formation and the ki-
netics of A-element losses. It is managed by adjusting both the
6Tiþ3Alþ3C-3Ti2AlC
(14)
1
By analogy with other MAX phase synthesis paths, it may be expected that
intermediate phases will form as transitional steps between these half-reactions.
2
Al melts at 660 °C on the heating ramp.

156
D.T. Cuskelly et al. / Journal of Solid State Chemistry 233 (2016) 150–157
Table 4
Summary table of MAX phase synthesis via exchange reactions. N/A indicates a MAX phase is not known to exist in this system.
ratio of reactants and the reaction conditions. These are related
variables whose separation from ex situ analysis can be difficult as
either can produce the same result – the introduction of additional
phases, primarily binary carbides. The experimental method in-
troduces additional variables. Methods such as hot pressing and
hot isostatic pressing partially supress A-element losses, by rapidly
closing the internal pore structure within the compacted re-
actants. Spark Plasma Sintering significantly reduces the sinte-
ring time and thus would require a different set of optimised
parameters.
In stark contrast to the other systems the V–Al–C sample in-
tended to produce VC, in which all of the available Al should have
been consumed in the reduction of the V2O5, instead underwent
SHS and produced the A-element rich phases V2AlC, V3AlC2 and
V4AlC3. The formation of A-element rich phases in MAX phase
synthesis is very rare. Their formation in the current work is not
fully understood, however it is believed to be related to the SHS
reaction. Subsequent samples made via reactive sintering using
powder from the same batch produced the binary carbide as ex-
pected. A possible explanation is that part of the vanadium oxide
was reduced by the carbon. Carbothermal reduction is possible for
all the M-element oxides investigated however aluminothermal
reduction is thermodynamically preferred at the firing tempera-
tures used. SHS reactions tend to have very short reaction times
and generate high reaction temperatures, as a result the equili-
brium phase may be different at the peak reaction temperature
and kinetically accessible intermediate phases can dominate over
the equilibrium phases. As such it is possible that some of the
vanadium oxide was rapidly reduced by carbon, leaving residual
aluminium to partake in MAX phase formation. It is also possible
that under the extremely short reaction times characteristic of
SHS, the reduction of the vanadium oxide was incomplete and left
partially reduced VxOy, which may have then been subsequently
reduced by C.
produced no V2AlC, whereas the MAX phase and Al2O3 were the
major phases in samples made by MASHS. Unfortunately sub-
sequent firing of samples made by MASHS at 1400 °C caused the
MAX phase to decompose. Although the milling time was reduced
in subsequent samples to prevent reaction in the mill, the powders
were still extremely reactive, making them prone to undergo SHS
when heated. SHS ignition temperature suppression by mechan-
ical activation is a well-documented phenomenon [40].
The exchange reaction synthesis method for MAX phases is
highly appealing from an economic standpoint. Consider the re-
duction of TiO2 by Al as described in Eq. (1). Elemental Ti, is
considerably more expensive than both TiO2 and the Al requ-
ired for reduction. The raw material cost to produce one tonne of
Ti3AlC2 from elemental reactants is approximately $11 500.
(73.8 wt% Ti at $15000/t, 13.8 wt% Al at $2400/t and 12.3 wt%
graphite at $800/t) whereas an exchange reaction to make the
same amount of Ti3AlC2 would cost less than half, only $4350
(60.1 wt% TiO2 at $2100/t, 33.9 wt% Al at $2400/t and 6.0 wt%
graphite at $800/t and accounting for the need to produce 2.05 t of
composite to yield 1 t of MAX phase [41]). If separation of the MAX
phase is desired, then an associated cost would need to be in-
corporated nonetheless the savings in raw materials are sub-
stantial. In addition, the composite itself may find engineering
application [42], removing the need for refinement and greatly
increasing yield.
The concept of MAX phase formation via an exchange reaction
has two main avenues for extension. Firstly there is the application
of Al as a reducing element to other MAX phase systems, primarily
the TiO2–A–C systems. As the Ti3SiC2þAl2O3 composite is already
known to form under these conditions the Ti–A–C MAX phases in
row 2 of Table 1 (A¼Al, Si, S, Ga, Ge, Cd, In, Sn, Ti and Pb) are the
most promising systems. Alternatively the use of a different re-
ducing agent would produce different composites. Choices for a
reducing element are limited but Mg, Ca, B and at high tempera-
tures C may all reduce M-element oxides. These reducing agents
are however are not free from problems. MgO is likely to form
spinels with the A-element (although a MAX phase-spinel com-
posite is possible via this method), Ca is often more expensive than
the M-element, B often results in the formation of very stable MB2
over MAX phases, and C may only be a suitable reducing agent at
temperatures outside that of MAX phase formation. Overall al-
though Al is not the only available reducing agent, finding a
The extreme reactivity of the V–Al–C system was also revealed
when high energy ball milling the samples in preparation for re-
action. MAX phase synthesis via high energy ball milling has been
studied [39], although the milling times and charge ratios used
here are comparatively very low. Mechanically Activated SHS
(MASHS) was more successful in producing V2AlC than pressure-
less reactive sintering with stoichiometric powder ratios in the V–
Al–C system. At these ratios samples synthesised in the furnace

D.T. Cuskelly et al. / Journal of Solid State Chemistry 233 (2016) 150–157
157
different one which still results in MAX phase formation is non-
trivial.
[10] D.T. Cuskelly, et al., Ti3GaC2 and Ti3InC2: first bulk synthesis, DFT stability
calculations and structural systematics, J. Solid State Chem. 230 (2015)
418–425.
An alternative avenue for development is to investigate meth-
ods of separating the MAX phase and the Al2O3. The MAX phases
are resistant to attack from both acids and bases, but Al2O3 is so-
luble in both HCl and NaOH at elevated temperatures. It is possible
that a method analogous to the Bayer process could dissolve the
Al2O3 leaving the MAX phase. Alternatively, physical separation of
crushed powders of composite using mineral separation techni-
ques could allow for refinement. Simple flotation methods have
yielded promising results although still far from producing pure
products.
[11] M.W. Barsoum, et al., Fabrication and electrical and thermal properties of Ti2
InC, Hf2InC and (Ti,Hf)2InC, J. Alloy. Compd. 340 (2002) 173–179.
[12] C. Bruesewilz, et al., Single crystal microcompression tests of the MAX phases
Ti2InC and Ti4AlN3, Scr. Mater. 69 (4) (2013) 303–306.
[13] A. Ganguly, M.W. Barsoum, J. Schuster, The 1300 °C isothermal section in the
Ti–In–C ternary phase diagram, J. Am. Ceram. Soc. 88 (2005) 1290–1296.
[14] D.P. Riley, E.K. Kisi, Self-propagating high-temperature synthesis of Ti3SiC2: 1,
ultra high-speed neutron diffraction study of the reaction mechanism, J. Am.
Ceram. Soc. 85 (10) (2002) 2417–2424.
[15] J.C. Viala, et al., Phase equlibria at 1000 °C in the Al–C–Si–Ti quaternary sys-
tem: an experimental approach, Mater. Sci. Eng. A229 (1997) 95–113.
[16] J.-F. Li, F. Sato, R. Watanabe, Synthesis of Ti3SiC2 polycrystals by hot-isostatic
pressing of the elemental powders, J. Mater. Sci. Lett. 18 (1999) 1595–1597.
[17] Z. Sun, H. Hashimoto, W. Tian, Synthesis of the MAX phases by pulse discharge
sintering, Int. J. Appl. Ceram. Technol. 7 (2010) 704–718.
5. Conclusions
[18] Z.F. Zang, et al., Application of pulse discharge sintering (PDS) technique to
rapid synthesis of Ti3SiC2 from Ti/Si/C powders, J. Eur. Ceram. Soc. 22 (16)
(2002) 2957–2961.
The exchange reaction used to produce Ti3AlC2 via an SHS re-
action mechanism in the past has been investigated here under
pressureless reactive sintering conditions. The concept of using an
exchange reaction to produce MAX phase – Al2O3 composites has
been successfully extended to all the of the M-elements which
produce a M2AlC phase. Many higher order phases (M3AlC2 and
M4AlC3) also exist in these systems and it has been shown that
these phases will also form under the conditions used. It was
observed that the exchange reaction can also take place during
high energy ball milling, successfully producing V2AlC and a fur-
nace ignited SHS exchange reaction in the V–Al–C system can
produce all three of the Vnþ1AlCn phases. Modification of reactant
ratios was shown to be very effective in improving yield of the
target phase.
[19] J. Zhang, et al., Rapid fabrication of Ti3SiC2–SiC nanocomposite using the spark
plasma sintering-reactive synthesis (SPS-RS) method, Scr. Mater. 56 (3) (2007)
241–244.
[20] F. Goesmann, R. Wenzel, R. Schmid-Fetzer, Preparation of Ti3SiC2 by electron-
beam-ignited solid-state reaction, J. Am. Ceram. Soc. 81 (11) (1998)
3025–3028.
[21] E.H. Kisi, D.P. Riley, Diffraction thermometry and differential thermal analysis,
J. Appl. Crystallogr. 35 (2002) 664–668.
[22] M.A. El Saeed, Optimization of the Ti3SiC2 MAX phase synthesis, Int. J. Refract.
Met. Hard Mater. 35 (2012) 127–131.
[23] E. Wu, E.H. Kisi, Synthesis of Ti3AlC2 from Ti/Al4C3/C studied by in situ neutron
diffraction, J. Am. Ceram. Soc. 89 (2) (2005) 710–713.
[24] S.S. Hwang, S.W. Park, T.W. Kim, Synthesis of the Ti3SiC2 by Solid state reaction
below melting temperature of Si, J. Alloy. Compd. 392 (1–2) (2004) 285–290.
[25] H. Hadadzadeh, M. Shafiee Afarani, M. Ekrami, Synthesis of titanium carbide
by the combustion of TiO2–2Mg–C and 3TiO2–4Al–3C systems in a tubular
furnace, Iran. J. Chem. Chem. Eng. 28 (2009) 71–76.
[26] A. Hendaoui, et al., A novel method for synthesis of low-cost Ti–Al–C-based
cermets, Int. J. Self-Propag. High-Temp. Synth. 18 (4) (2010) 267–272.
[27] J. Chen, J. Li, Y. Zhou, In-situ synthesis of Ti3AlC2/TiC–Al2O3 composite from
TiO2–Al–C system, J. Mater. Sci. Technol. 22 (4) (2005) 455–458.
[28] C.L. Yeh, R.F. Li, Y.G. Shen, Formation of Ti3SiC2–Al2O3 in situ composites by
SHS involving thermite reactions, J. Alloy. Compd. 478 (1–2) (2009) 699–704.
[29] G. Aylward, T. Findlay, SI Chemical Data, Wiley, United States, 2008.
[30] W.K. Pang, et al., Comparison of thermal stability in MAX 211 and 312 phases,
J. Phys. 251 (1) (2010) 012025.
Acknowledgements
The first author received financial support from an Australian
Postgraduate Award and Australian Solar Institute
– ARENA
scholarship #2-S015. The assistance of David Phelan and Jennifer
Zobec in the University of Newcastle Electron Microscope and
X-ray unit is also gratefully acknowledged.
[31] W. Zhang, et al., Reactive hot pressing and properties of Nb2AlC, J. Am. Ceram.
Soc. 92 (10) (2009) 2396–2399.
[32] J. Etzkorn, et al., Ti2GaC, Ti4GaC3 and Cr2GaC – synthesis, crystal growth and
structure analysis of the Ga-containing MAX phases Mn¼1GaCn with M¼Ti, Cr
and n¼1, 3, J. Solid State Chem. 182 (5) (2009) 995–1002.
[33] I. Salama, T. El-Raghy, M.W. Barsoum, Synthesis and mechanical properties of
Nb2AlC and (Ti,Nb)2AlC, J. Alloy. Compd. 347 (1–2) (2002) 271–278.
[34] C. Hu, et al., In situ reaction synthesis and mechanical properties of V2AlC, J.
Am. Ceram. Soc. 91 (12) (2008) 4029–4035.
References
[1] W. Jeitschko, H. Nowotny, F. Benesovsky, Die H-Phases Ti2InC, Zr2InC, Hf2InC
und Ti2GeC, Monatsh. Chem. Verwandte Tl. And. Wiss. 94 (6) (1963)
1201–1205.
[2] M.W. Barsoum, T. El-Raghy, Synthesis and characterisation of a remarkable
ceramic: Ti3SiC2, J. Am. Ceram. Soc. 79 (1996) 1953–1956.
[3] M.W. Barsoum, MAX Phases: Properties of Machineable Ternary Carbides and
Nitrides, Wiley, United States, 2013.
[4] M.W. Barsoum, The Mnþ1AXn phases and their properties, in: R. Riedel, I.-
W. Chen (Eds.), Ceramics Science and Technology: Materials and Properties,
John Wiley and Sons, United States, 2010, pp. 299–347.
[35] L.-O. Xiao, et al., Synthesis and thermal stability of Cr2AlC, J. Eur. Ceram. Soc.
31 (8) (2011) 1497–1502.
[36] C. Hu, et al., In situ reaction synthesis and decomposiition of Ta2AlC, Int. J.
Mater. Res. 99 (2008) 8–13.
[37] R.J. Hill, C.J. Howard, Quantitative phase analysis from neutron powder dif-
fraction data using the Rietveld method, J. Appl. Crystallogr. 20 (6) (1987)
467–474.
[38] J. Etzkorn, M. Ade, H. Hillebrecht, Ta3AlC2 and Ta4AlC3 ꢀ single-crystal in-
vestigations of two new ternary carbides of tantalum synthesized by the
molten metal technique, Inorg. Chem. 46 (4) (2007) 1410–1418.
[39] A. Hendaoui, et al., Synthesis of high-purity polycrystalline MAX phases in Ti–
Al–C system through mechanically activated self-propagating high-tempera-
ture synthesis, J. Eur. Ceram. Soc. 30 (4) (2009) 1049–1057.
[5] M.W. Barsoum, M. Radovic, Elastic and mechanical properties of the MAX
phases, Annu. Rev. Mater. Res. 41 (2011) 195–227.
[6] P. Eklund, et al., The Mnþ1AXn phases: materials science and thin-film pro-
cessing, Thin Solid Films 518 (8) (2010) 1851–1878.
[40] D.P. Riley, E.H. Kisi, D. Phelan, SHS of Ti3SiC2: ignition temperature depression
by mechanical activation, J. Eur. Ceram. Soc. 26 (6) (2004) 1051–1058.
[41] InvestmentMine Inc., Mining Markets and Investment, 2015. Available from:
〈http://www.infomine.com/investment/metal-prices/〉.
[7] V.J. Keast, S. Harris, D.K. Smith, Prediction of the stability of the Mnþ1AXn
phases from first principles, Phys. Rev. B 80 (2009) 214113.
[8] M.W. Barsoum, The MNþ1AXN phases: a new class of solids: thermo-
dynamically stable nanolaminates, Prog. Solid State Chem. 28 (2000) 201–281.
[9] I.M. Low, Advances in Science and Technology of the Mnþ1AXn Phases,
Woodhead Publishing, United Kingdom, 2012.
[42] H. Wang, Z.H. Jin, Y. Miyamoto, Effect of Al2O3 on mechanical properties of
Ti3SiC2/Al2O3 composite, Ceram. Int. 28 (8) (2002) 931–934.