Amorphous, monoclinic, and tetragonal porous zirconia through a controlled self-sustained combustion route

Article abstract of DOI:10.1111/j.1551-2916.2010.04334.x

Porous, large surface area, metastable zirconias, are of importance to catalytic, electrochemical, biological, and thermal insulation applications. Combustion synthesis is a very commonly used method for producing such zirconias. However, its rapid nature makes control difficult. A simple modification has been made to traditional solution combustion synthesis to address this problem. It involves the addition of starch to yield a starting mixture with a dough-like consistency. Just 5 wt% starch is seen to significantly alter the combustion characteristics of the dough. In particular, it helps to achieve better control over reaction zone temperature that is significantly lower than the one calculated by the adiabatic approximation typically used in self-propagating high-temperature synthesis. The effect of such control is demonstrated by the ability to tune dough composition to yield zirconias with different phase compositions from the relatively elusive amorphous to monoclinic (>30 nm grain size) and tetragonal pure zirconia (<30 nm grain size). The nature of this amorphous phase has been investigated using infrared spectroscopy. Starch content also helps tailor porosity in the final product. Zirconias with an average pore size of about 50 μm and specific surface area as large as 110 m²/g have been obtained.

Full text of DOI:10.1111/j.1551-2916.2010.04334.x

J. Am. Ceram. Soc., 94 [6] 1747–1755 (2011)
DOI: 10.1111/j.1551-2916.2010.04334.x
r 2011 The American Ceramic Society
ournal
J
Amorphous, Monoclinic, and Tetragonal Porous Zirconia Through a
Controlled Self-Sustained Combustion Route
Venugopal B. Raghavendra,z Swati Naik,y Meera Antony,z Gopalakrishnan Ramalingam,y
Michael Rajamathi,w,z and Srinivasan Raghavanw,y
zMaterials Research Group, Department of Chemistry, St. Joseph’s College, Bangalore 560 027, India
yMaterials Research Centre, Indian Institute of Science, Bangalore 560 012, India
Porous, large surface area, metastable zirconias, are of impor-
tance to catalytic, electrochemical, biological, and thermal in-
sulation applications. Combustion synthesis is a very commonly
used method for producing such zirconias. However, its rapid
nature makes control difficult. A simple modification has been
made to traditional solution combustion synthesis to address this
problem. It involves the addition of starch to yield a starting
mixture with a ‘‘dough-like’’ consistency. Just 5 wt% starch is
seen to significantly alter the combustion characteristics of the
‘‘dough.’’ In particular, it helps to achieve better control over
reaction zone temperature that is significantly lower than the
one calculated by the adiabatic approximation typically used in
self-propagating high-temperature synthesis. The effect of such
control is demonstrated by the ability to tune dough composition
to yield zirconias with different phase compositions from the
relatively elusive ‘‘amorphous’’ to monoclinic (430 nm grain
size) and tetragonal pure zirconia (o30 nm grain size). The
nature of this amorphous phase has been investigated using in-
frared spectroscopy. Starch content also helps tailor porosity in
the final product. Zirconias with an average pore size of about
50 lm and specific surface area as large as 110 m2/g have been
obtained.
phase stability.4–7 This is demonstrated in the research reported
here by the synthesis of pure zirconia in its X-ray amorphous
(hereafter referred to as amorphous), tetragonal, and monoclinic
forms and with varying surface and pore structures. From a
scientific viewpoint, observation of a ‘‘pure amorphous zircon-
ia’’ in particular is interesting as it is a rather elusive phase re-
quiring very small crystallite sizes of the order of 3–4 nm or
less.6,7 From a technological perspective, the combustion syn-
thesis of amorphous zirconia is important as it provides an
alternative route to synthesizing metastable zirconia solid solu-
tions. The production of amorphous and nanocrystalline tetrag-
onal zirconia using a starch-containing gel as the starting
material for combustion synthesis has been reported by Tahma-
sebi and Padyar.8 However, detailed correlation between the
reason for the formation of these phases, their nature and com-
bustion characteristics was not presented. For instance, flame
temperatures were estimated by the adiabatic approximation.
It is shown in this paper that while combustion remains self-
sustained, the addition of starch changes it enough to render the
adiabatic approximation based on completed combustion of
starch invalid. In that sense, it is different from combustion
routes that have been used to prepare ZrO2-based ceramics
either from solution combustion techniques9 or from starting
mixtures containing zirconium metal.10 Also, the exact nature of
the combustion-synthesized amorphous zirconia was not further
probed by Tahmasebi and Padyar.8
I. Introduction
ELF-PROPAGATING high-temperature synthesis (SHS) or com-
bustion synthesis is a well-known method for the prepara-
While the method presented here definitely does not yield or-
dered porous structures, it nevertheless is a simple technique for
producing porous zirconia with the desired phase, in potentially
large quantities. In combination with good chemical stability
and acidic/basic surface adsorption characteristics, such porous
zirconias with a large specific surface area are of importance as
catalyst and catalyst support for a wide variety of reactions.11
The range of temperatures the porous zirconias are exposed to
in these catalytic applications is from 3001 to 8001C and the
surface area involved is 80–100 m2/g. One particular application
of importance is as an anode in solid-oxide fuel cells when com-
bined with Ni to obtain a Ni–YSZ (yttria-stabilized zirconia)
S
tion of oxide ceramic powders in general and porous zirconias in
particular.1 This method is based on the ability of an exothermic
heterogeneous chemical reaction to become self-sustaining and
to propagate as a combustion wave through the mixture of re-
actants. The SHS method has technologically important char-
acteristics like simplicity of the experimental set-up, intrinsic
ability of self-purification, and rapidity of synthesis.2 However,
the rapid self-sustaining nature of the reaction—reaction tem-
peratures, for example, are often estimated by using the adia-
batic approximation—also makes it less amenable to control.
We have recently reported the synthesis of macroporous di-
valent and trivalent metal-oxide foams through a self-sustained
combustion method starting from a ‘‘dough’’ containing a nat-
ural polymer such as starch.3 The addition of starch is a simple
yet effective modification that results in better control over the
time-temperature history of the combustion process. The ad-
vantage of the modification became more obvious when it was
used to synthesize porous zirconia with its grain size-related
12
cermet.
II. Phase Stability in Zirconia-Based Systems
As mentioned before, the effect of starch addition on the com-
bustion process was evaluated using the grain size-related phase
stability of zirconia-based systems. Hence, phase stability and its
importance from the applications mentioned above is briefly
discussed here. Pure zirconia exists in three polymorphic
forms—cubic (c), tetragonal (t), and monoclinic (m). The mono-
clinic phase is the stable form at room temperature, the tetrag-
onal phase above 11701C and the cubic phase above 23701C.13
Among these phases the high-temperature tetragonal phase
is preferred as an active catalyst for organic and petrochemi-
cal reactions. For instance, Stichert et al.14 report that for
M. Cinibulk—contributing editor
Manuscript No. 27826. Received April 10, 2010; approved November 16, 2010.
This work was funded by DST, New Delhi.
wAuthors to whom correspondence should be addressed: e-mails: sraghavan@
mrc.iisc.ernet.in and mikerajamathi@rediffmail.com
1747

1748
Journal of the American Ceramic Society—Venugopal et al.
Vol. 94, No. 6
Fig. 1. Photographs of experimental set-up depicting: (a) electric bunsen with crucible; (b) set-up with thermocouple embedded in the sample;
(c) ‘‘explosive’’ flash at the end of combustion in the case where the reaction mixture contains no starch; (d) black carbon ring in case of mixture
containing 10 mg starch during active combustion, and (e) completely black sample left behind after combustion in case of the 500 mg starch.
acid-catalyzed n-alkane isomerization reactions, sulfated t-ZrO2
is two to five times more active than sulfated m-ZrO2. The high-
temperature phases of ZrO2 can be stabilized at room temper-
ature by incorporating dopants such as MgO, CaO, Y2O3, and
CeO2 into the crystal lattice.13 Alternatively, phase pure t-ZrO2
has also been stabilized at room temperature by preparing nano-
crystalline powders below a critical crystallite size of 20–30 nm,
known as the Garvie limit, above which it reverts back to the
stable monoclinic phase.7,15 Stabilization of the tetragonal phase
at smaller crystallite sizes or more accurately at larger specific
surface areas is due to its smaller specific surface energy. At even
larger specific surface areas, an X-ray amorphous phase can be
thermodynamically stabilized due to possibly an even lower spe-
cific surface enthalpy.6,7 It is pointed out though that it is im-
possible to rule out the presence of small amount of physically
or chemically bound water in X-ray or TEM amorphous zir-
conia. Its observation in general is difficult as zirconia has to be
heated very carefully to remove any water present without caus-
ing crystallization. The tetragonal to amorphous transition
seems to be energetically favored at crystallite sizes o4–5 nm
or equivalent surface area 4230 m2/g.
III. Experimental Details
All the chemicals were used as received without any further pu-
rification. Zirconyl nitrate [ZrO(NO3)2 ꢀ xH2O] of the purity
99.50% was obtained from Acros Organics (Morris Plains,
NJ). Analytical grade urea (99.50% pure) and gravimetric grade
starch was procured from Merck Ltd. (Mumbai, India). In a
typical experiment, hydrated zirconyl nitrate [ZrO(NO3)2 ꢀ
6H2O] and urea were taken in the molar ratio of 1:4.5. Accord-
ing to the principle of propellant chemistry,16 the required ox-
idant to fuel ratio is 1:1.67. However, the ratio of oxidant to fuel
is kept higher thereby maintaining the reaction mixture fuel-
rich. Different amounts of starch {5, 10, 25, 50, 150, 300, 500,
and 600 mg per 100 mg (0.31 mmol of ZrO(NO3)2 ꢀ 6H2O)} was
added to the above mixture and ground until it reached a

June 2011
Amorphous, Monoclinic, and Tetragonal Porous Zirconia
1749
dough-like consistency so that it could be easily manipulated
into pellets of desired shape. Deionized water, typically o1 mL/
g of the reaction mixture, was added, as required, to get the de-
sired consistency. Combustion of undried and predried pellets
showed that this water did not have any significant effect on the
reaction characteristics. To perform the combustion reaction a
porcelain crucible was placed in an electric Bunsen as shown in
Fig. 1(a). It was left to soak for 30 min to remove any possible
volatile contaminants and allow temperatures to stabilize. The
surface temperature of the crucible bottom was measured using
a 0.13 mm diameter, K-type thermocouple to be 4001C. The re-
action pellet was then introduced into this preheated crucible. A
control experiment was also carried out in the absence of starch.
Two types of experiment were performed. In the first type (type-
1), the total mass of the reaction pellet was not maintained con-
stant but allowed to increase with the amount of starch. After
the combustion process, the reaction product was maintained at
the same temperature for 30 min to remove residual carbona-
ceous material. Unless otherwise mentioned in the text, the data
pertain to this type of experiment. In the second type of exper-
iment (type-2), a fixed mass, 175 mg, of the dough was pressed in
a steel die to yield cylindrical pellets 6 mm in diameter and 4 mm
in height. Following combustion, the pellets were removed im-
mediately without subjecting them to the subsequent 30 min of
heat treatment. This set of experiments were designed to factor
out possible effects of total mass and morphology of the reac-
tion mixture and the postcombustion heat treatment as in type-1
experiments. In order to estimate combustion zone tempera-
tures, a pellet was formed around the 0.13 mm K-type thermo-
couple with a 1 s time constant (time to achieve 60% of the value
of an instantaneous temperature change) as shown in Fig. 1(b).
This assembly was then lowered into the crucible.
Type 2
(101)t
(e)
(002)t (200)t
Type 1
(d)
70
75
80
Type 1
(c)
Type 1
(b)
The solids were characterized by powder X-ray diffraction
(PXRD), scanning electron microscopy (SEM), thermogravimet-
ric analysis (TGA), differential thermal analysis (DTA), surface
area measurements, and infrared (IR) spectroscopy. The PXRD
patterns of the materials were recorded on a PANalytical X’Pert
Pro diffractometer (Almelo, the Netherlands) fitted with second-
(111)m
-
˚
ary graphite monochromator. CuKa radiation (l51.5418 A)
Type 2
was used. Data were collected at a typical rate of 11C/min over
the 2y range of 201–801. To identify the tetragonal phase forma-
tion wherever required the data were collected at much slower
scan rate (0.251C/min). Crystallite sizes were calculated using the
Scherrer equation using peak full widths at half maximum cor-
rected for instrumental broadening. Phase fractions were calcu-
lated by using the integrated intensities of the {111} monoclinic
and tetragonal peaks.17 SEM images were obtained using a JEOL
JSM 840A microscope (Jeol, Japan). TGA and DTA were carried
out using TA Instruments (thermal analyzer) system (New Castle,
DE) under continuous air flow with the heating rate of 101/min.
The specific surface area measurements, average pore diameter,
and N2 adsorption–desorption studies were carried out using a
Nova-1000 (Ver. 3.70) instrument (Quantachrome Instruments,
Boynton Beach, FL). The samples were degassed at 1301C before
measurement. IR spectra were recorded on a Thermo-Nicolet
model 400-D FTIR spectrometer (Thermo Scientific, Waltham,
MA) (KBr pellet method 4 cmꢁ1 resolution).
(a)
(111)m
(002)t (200)t
26
28
30
32
34
36
38
40
2θ (degrees)
Fig. 2. PXRD patterns of macroporous ZrO2 obtained when (a) 0 mg
(type-2), (b) 0 mg (type-1), (c) 5 mg (type-1), (d) 10 mg (type-1), and (e)
10 mg (type-2) of starch was used in the reaction mixture. The PXRD
pattern (d) recorded at slower scan rate in the range 2y 5701–801 is
shown as inset. Type-1 and -2 refer to different kinds of synthesis de-
scribed in the text. m, monoclinic; t, tetragonal.
2(d) and (e)) pure tetragonal ZrO2 is obtained. It is difficult to
distinguish between cubic and tetragonal phases of nanocrystal-
line ZrO2 as the additional peaks that differentiate the tetrago-
nal phase from the cubic phase merge due to peak broadening.
However, the (004) and (400) (see inset Fig. 2(d)) reflections
corresponding to the tetragonal phase were observed at lower
scan rates in the 10 mg starch sample thus confirming its iden-
tity. The change in phase composition is accompanied by a re-
duction in X-ray crystallite size (see Table I) from ꢂ 30 nm at
0 mg starch to r10 nm at 10 mg. Further increments in starch
from 10 to 500 mg, (Fig. 3) and (Table I), have very little effect
on phase composition or crystallite size from type-1 experi-
ments. There is a nominal decrease in crystallite size of the te-
tragonal phase from B9–10 to B7–8 nm. However, a significant
IV. Results and Discussions
(1) Effect of Dough Composition on Combustion and Phase
Stability
The XRD patterns of the ZrO2 samples obtained in the control
experiment (without the addition of starch) and when 5 and 10
mg of starch were added are shown in (Fig. 2). Both type-1 and
type-2 experiments have been included. The ZrO2 obtained in
the absence of starch is largely the monoclinic phase (Figs. 2(a)
and (b)) with a small percentage of cubic and/or tetragonal
phase. The percentage of cubic and/or tetragonal phase in-
creases when 5 mg of starch was used (Fig. 2(c)). In contrast,
when the amount of starch used was increased to 10 mg (Figs.

1750
Journal of the American Ceramic Society—Venugopal et al.
Vol. 94, No. 6
Table I. Crystallite Size and Relative Phase Fractions of
Various ZrO2 Samples
(e)
Crystallite size (nm)
Phase fractions (%)
Tetragonal Monoclinic
Type 2
Amount of starch
used (mg)
Tetragonal
Monoclinic
0 (type 1)
0 (type 2)
5 (type 1)
10 (type 1)
10 (type 2)
150 (type 1)
500 (type 1)
500 (type 2)
600 (type 1)
21
—
17
9
10
10
8
31
41
24
—
—
—
—
29
0
59
100
100
100
100
—
71
100
41
0
0
0
0
Type 1
(d)
Amorphous
7
—
100
0
development at large starch contents such as in the 500 mg
sample is the appearance of the amorphous phase in type-2
experiments as seen in Fig. 3 pattern ‘‘e.’’
While the postcombustion heat treatment did not affect the
crystallite size of the tetragonal phase and hence the tetragonal to
monoclinic transition significantly, it does affect the amorphous
to tetragonal transition. As seen from patterns ‘‘d’’ and ‘‘e’’ in
Fig. 3, in the sample containing 500 mg starch, type-1 experi-
ments yield tetragonal zirconia whereas type-2 experiments yield
amorphous zirconia. The 10 mg starch sample in contrast yielded
tetragonal zirconia with a crystallite size of 9–10 nm (see Table I)
irrespective of whether it was subjected to postcombustion heat
treatment or not. The effect of starch addition on the phases
obtained is discussed below by analyzing its effect on the thermal
history of the pellet either during or after combustion.
(c)
(b)
In order to understand the effect of starch addition on the
combustion reaction, the manner in which hydrated zirconyl
nitrate is converted into dry zirconia during combustion needs
to be understood. Hence, pure zirconyl nitrate, pure urea, zir-
conyl nitrate1urea, zirconyl nitrate 110 mg starch, and zirconyl
nitrate1500 mg starch mixtures were subjected to TGA/DTA in
air. As can be seen from Fig. 4(a), the TGA data of pure zirconyl
nitrate indicate two sharp weight losses at about 1001 and 2001C
with corresponding endotherms in the DTA data. These are
most likely due to loss of physically bound water and nitrate,
respectively, leaving behind hydrated zirconia. There is then a
period of gradual weight loss till 4001C which is most likely due
to removal of chemically bound water. The interesting feature
after 4001C is the sharp exothermic peak at 4551C which is not
accompanied with any significant weight loss. The only likely
source of this peak is an exothermic phase transition such as
crystallization. To understand this further and in order to mimic
combustion synthesis, i.e. a fast heating rate but in a more con-
trolled fashion, zirconyl nitrate samples were inserted into a
furnace set to different temperatures and removed after 1–5 min.
As seen from Fig. 5 when inserted at 7001C and removed after
1 min the sample is X-ray amorphous as in Fig. 3, pattern ‘‘e.’’ No
peaks corresponding to zirconyl nitrate (pattern ‘‘a’’ in Fig. 5) is
seen thus implying that it is most likely some form of hydrated
amorphous zirconia. The transition to the tetragonal phase is
seen only after about 3 min. Similar transitions were observed in
all samples heated for 1–5 min from 5001 to 8001C. As can also
be seen from Fig. 4(b), the main exothermic peak in the DTA
data of the mixtures, which is attributed to the combustion pro-
cess, occurs at about 2501C. The X-ray plus TGA/DTA data
unambiguously show that zirconyl nitrate when subjected to
combustion synthesis, first gets dehydrated, then calcined to
amorphous zirconia, which is then converted to crystalline te-
tragonal zirconia presumably due to coarsening. The origin of
the endothermic peaks in the DTA trace of the mixtures can be
easily deduced on comparing with the DTA traces of the pure
components of the mixture. The sharp exothermic peak at 4501C
in the DTA trace of the 500 mg sample is as discussed later, due
to the combustion of residual carbon and not due to the crys-
tallization mentioned above.
(101)t
(a)
(002)t (200)t
25
30
35
40
2θ (degrees)
Fig. 3. PXRD patterns of macroporous ZrO2 obtained using (a) 25 mg,
(b) 50 mg, (c) 300 mg, (d) 500 mg (type-1), and (e) 500 mg (type-2) of
starch in the reaction mixture.
For the purposes of the remaining discussion it will be assumed
that the combustion process can be represented by the reactions
(1) and (2) below
ZrOðNO3Þ2 þ 4:5NH2CONH2 þ 4:25O2
! ZrO2 þ 5:5N2 þ 4:5CO2 þ 9H2O
C6H10O5 þ 6O2 ! 6CO2 þ 5H2O
(1)
(2)
In simple qualitative terms, the reaction, in the absence of
starch and the presence of 10 mg of starch can be described as
‘‘explosive’’ and ‘‘smouldering,’’ respectively. A preliminary
comparison of the manner in which cylindrical pellets 6 mm in
diameter and 4 mm in height, obtained by lightly pressing 175 mg
of the dough in a steel die, underwent combustion, illustrates

June 2011
Amorphous, Monoclinic, and Tetragonal Porous Zirconia
1751
(101)t
(c)
(002)t (200)t
(b)
(a)
26
28
30
32
34
36
38
40
2θ (degrees)
Fig. 5. PXRD patterns of (a) pure ZrO(NO3)2, (b) hydrated amor-
phous ZrO2 (ZrO(NO3)2 calcined at7001C for 1 min), and (c) t-ZrO2
(ZrO(NO3)2 calcined at 7001C for 3 min).
trolled. In contrast, for the 10 mg starch pellet, the reaction was
ignited at the bottom of the pellet as soon as it was placed in the
crucible within 2 s. A combustion front marked by a black ring
due to decomposition of starch could then be seen traveling up the
pellet as shown in Fig. 1(d) (reason for the blurry photograph).
The fact that combustion is marked by a black ring indicates the
heat produced in this case is enough to both decompose starch
and initiate ignition of the carbon so formed. In comparison to
the 3 s in which the entire mixture flares up in the absence of
starch, it took 60 s for the reaction front to go from the bottom to
the top of the 4 mm pellet. In the case of the pellet containing
500 mg starch, the beginning of combustion, blackening at the
bottom of the pellet, was detected 9 s after insertion into the
crucible. It then took the black front 90 s to reach the top of
the pellet. The end product is shown in Fig. 1(e).
Maximum temperatures recorded by the embedded K-type
thermocouple were 4001, 4001, and 3901C for the 0, 10, and 500 mg
samples. The recorded temperature is obviously completely
incorrect for the 0 mg case. The one measured in the 10 mg case
is believed to be an underestimate, while the 3901C temperature
recorded in the 500 mg starch case is probably the most reliable
of the three readings as discussed in the following sections.
If the process can be considered to be adiabatic in the absence
of starch, then a quick calculation using the reaction (1) and
data listed in Table II result in an estimated temperature of
22751C. For doing these calculations, it was assumed that the
Fig. 4. (i) TGA and DTA of ZrO(NO3)2 ꢀ xH2O. The two endotherms
at about 1151 and 2001C are accompanied with weight losses. In contrast
the sharp exotherm at 4551C is not accompanied with any significant
weight loss thereby indicating that it is most likely a phase transition
such as crystallization; (ii) DTA of (a) ZrO(NO3)2 ꢀ xH2O, reaction mix-
tures containing (b) 0 mg, (c) 10 mg and (d) 500 mg starch and (e) urea:
the ignition temperature of the reaction mixtures are approximately
2001C when nitrate decomposition starts and large amount of heat is
evolved.
this difference. In the absence of starch, the time required for
completion of the reaction was about 20 s. The first 17 s involved
large scale frothing due to melting of urea and evolution of steam.
The evolution of reddish brown fumes and slight discoloration of
the frothing mass indicated the beginning of decomposition of
nitrate after about 15 s. The mixture then underwent the com-
bustion reaction given by Eq. (1) between the 17th and 20th sec-
onds in an explosive fashion akin to a flare. This last explosive
stage is shown in Fig. 1(c). Hence, this process is highly uncon-

1752
Journal of the American Ceramic Society—Venugopal et al.
Vol. 94, No. 6
Table II. Data Used for Thermodynamic Calculations
Substance
DHf1 (kJ/mol)
Cp (J ꢀ (K ꢀ mol)ꢁ1 18,19
)
CO2 (g)
H2O (g)
H2O (l)
N2 (g)
NH3 (g)
NH2CONH2 (c)
O2 (g)
(C6H10O5)n (c)
ZrO2 (c)
ꢁ393.5120
ꢁ241.8320
ꢁ285.8320
0.020
44.1419.04ꢃ 10ꢁ3Tꢁ8.58ꢃ 105Tꢁ2
30110.71 ꢃ 10ꢁ3T10.33ꢃ 105Tꢁ2
712.6 ꢃ 10ꢁ3T10.08ꢃ 105Tꢁ2
27.2014.184 ꢃ 10ꢁ3T
ꢁ45.9420
ꢁ333.120
0.020
—
—
29.9614.184 ꢃ 10ꢁ3Tꢁ1.67ꢃ 105Tꢁ2
ꢁ1680.178
ꢁ1100.620
ꢁ1346.4121
ꢁ3086.1221
—
69.6217.53ꢃ 10ꢁ3Tꢁ14.06 ꢃ 105Tꢁ2
ZrO(NO3)2 (c)
ZrO(NO3)2 6H2O
C (s)
—
17.1514.27ꢃ 10ꢁ3Tꢁ8.79ꢃ 105Tꢁ2
13.0011.9 ꢃ 10ꢁ3T11.27ꢃ 105Tꢁ2
H2 (g)
c, crystalline; g, gas; l, liquid; T, absolute temperature.
heat of dehydration of ZrO(NO3)2 ꢀ 6H2O is not involved in the
combustion process. If included it actually makes reaction (1) an
endothermic process. It is assumed that most of this heat is
provided by the electric Bunsen which heats up the pellet to
42501C to initiate the reaction by which time most of the
chemically bound water is removed as seen from Fig. 3. Furnace
experiments showed that it takes 15 min at 10001C for 9 nm
tetragonal powders obtained from combustion synthesis to start
transforming to the monoclinic phase following coarsening to
beyond 30 nm. Given that this process happens in a few seconds
during combustion, the adiabatic estimate of 420001C in the
combustion zone seems reasonable. Thus, in the absence of
starch, temperatures are high enough to cause coarsening of te-
tragonal nuclei beyond the 30 nm limit resulting in the obser-
vation of monoclinic zirconia.
adiabatic temperature as function of starch content from Eqs.
(3a) and (3b) is plotted in Fig. 6. As seen, starch additions
should increase combustion temperature and pure starch would
be expected to burn at 30001C. The experimental results and
actual observations point toward the exact opposite or a much
lower temperature. Indeed if this were to be true, then the ad-
dition of starch should result in a monoclinic product. From
type-2 experiments, the addition of 10 mg of starch yields te-
tragonal zirconia with a 10 nm grain size. Furnace-based exper-
iments to mimic combustion as before indicate that when
zirconyl nitrate is directly inserted into a furnace at 9001C it
gets converted into crystalline tetragonal zirconia in 60 s, the
time required for combustion of the pellet to be completed.
Based on this data and actual experimental observation of the
combustion process, it is estimated that the temperature of the
pellet during combustion on the addition of about 10 mg of
starch is about 9001–10001C, in comparison to 22751C in the
absence of starch. Thus, the addition of just 10 mg (B5 wt%) of
starch causes combustion zone temperatures to drop from
420001 to about o10001C, instead of raising it as would be
expected from adiabatic calculations based on Eqs. (1) and (2).
There are three reasons for the observed temperature drop and
discrepancy between adiabatically calculated and experimentally
In order to estimate combustion zone temperatures in the
presence of starch, if the same adiabatic approximation used
before were to be used again, then the temperature rise would be
given by the equation
0:18ð1 ꢁ f ÞDHnitrateþfDHstarchꢁDHvap:water
2
T
Z
h
i
6
4
¼ 0:18ð1 ꢁ f Þ
CpZrO þ 5:5CpN þ 4:5CpCO dT
2
2
2
298i
2
33
373
T
Z
Z
vap
2
þ15
CpH OðlÞdT þ DHH O
CpH OðgÞdT
2
4
55
H2O
3200
2
298
373
2
3
5
2800
T
373
T
Z
Z
Z
Adiabatically calculated
2
2
2
4
þ f
6
CpCO dT þ 5
CpH OðlÞdT þ DHvap
þ
CpH OðgÞdT
Experimentally estimated
2400
298
298
373
Experimentally measured
2
3
373
T
Z
Z
2000
H2O
þ
CpH OðlÞdT þ DHvap
þ
CpH OðgÞdT
2
2
4
5
1600
1200
800
298
373
(3a)
which simplifies to
399 ꢃ 103ð1 ꢁ f Þ þ 1891 ꢃ 103f
400
¼ T2ð0:018 þ 0:042f Þ þ Tð156 þ 258f Þ ꢁ 51388
ꢁ 94441f
0.0
0.2
0.4
0.6
0.8
1.0
(3b)
Fraction of starch in mixture
Fig. 6. Plot of the combustion temperature as a function of fraction of
starch fraction in the reaction mixture estimated experimentally (using
phase composition data as described in the text) as well as calculated
assuming adiabatic combustion. The experimentally measured temper-
ature using the embedded thermocouple of 3901C for the reaction mix-
ture containing 500 mg of starch is also shown. The comparison shows
that the adiabatic approximation becomes completely invalid when
starch is used.
f is the mole fraction of starch in the reaction mixture and T is
the temperature in K. The left side of the above equation is es-
sentially a measure of the heat that would be released in mix-
tures containing different fractions of starch and the right-hand
side is the temperature raise that could be expected given the
products as per the reactions (1) and (2) given before. The data
used to arrive at Eqs. (3a) and (3b) is listed in Table II. The

June 2011
Amorphous, Monoclinic, and Tetragonal Porous Zirconia
1753
estimated values. Firstly, the fact that the samples containing
starch were brownish to black following combustion, indicates
that not all carbon is oxidized and thus heat produced would be
lower than what is expected from Eq. (2). Secondly, given this
observation, it implies that part of the heat from combustion of
zirconyl nitrate is used for decomposition of starch and raising
the temperature of the decomposition products, namely carbon,
hydrogen, and oxygen. A third and probably most important
reason is that the combustion of pure starch in air, though self-
sustaining, is too slow and cannot be approximated as an adi-
abatic process. This is illustrated by the following calculation for
the 10 mg starch containing sample. Following the reasoning in
the two points above, even if it is assumed that starch only
decomposes and all hydrogen and carbon so produced remain
uncombusted, adiabatic calculations yield a lower bound
temperature of 13001C which is still too high to justify 10 nm
tetragonal zirconia. It, thus, makes adiabatic approximation of
combustion in samples containing even small amounts of starch,
B5 wt%, questionable in the mixtures used herein. Higher
starch contents would involve greater deviations. Thus, slow
rate of combustion (and therefore greater heat losses), energy
consumed in decomposition of starch and incomplete burning of
carbon all contribute to a reduction in temperatures when starch
is added to the nitrate–urea mixture.
Tetragonal
(b)
Amorphous
(a)
O-H Stretching
For greater starch contents up to 500 mg, crystallite size data
from Table I shows that the temperature of the combustion zone
does not change significantly enough to affect coarsening of any
tetragonal nuclei formed. However, it does affect the amorphous
to tetragonal transition, as type-2 experiments with 500 mg
starch yielded amorphous zirconia. Given the TGA/DTA data
showing completion of calcination of nitrate at 4001C, the min-
imum temperature expected is thus 4001C. The fact that the
combusted product remains black, see Fig. 1(e), also indicates
that temperatures were not high enough to ignite carbon in air.
All of these observations are also in fairly good agreement with
the measured temperature of 3901C.
What is nevertheless very interesting is that for the 500 mg
starch sample, the end product from type-2 experiments is amor-
phous zirconia. Observation of amorphous zirconia is difficult
as any hydrolyzed form of zirconia needs to be heated just
enough to remove water without increasing crystallite size to
42–4 nm.7 At larger sizes thermodynamics seems to favor the
tetragonal phase. Thus, pure amorphous zirconia is in general
elusive. While, it cannot be said with complete certainty if
whether the amorphous zirconia observed has any bound wa-
ter or not, it is definitely not the nitrate or tetragonal zirconia as
seen from Fig. 4. The IR spectrum of X-ray amorphous zirconia
produced by combustion synthesis is shown in Fig. 7(a). The
spectrum (a) in Fig. 7 has a broad peak at 3420 cmꢁ1 due to O–
H groups and/or absorbed moisture. However, it does not have
a peak at 1384 cmꢁ1. This observation suggests that there is no
residual nitrate group which is present during the formation of
amorphous ‘‘zirconia’’ by combustion. The formula of the com-
pound could then be written as ZrO2ꢁx(OH)2x ꢀ yH2O. Spectrum
(b) from X-ray nanocrystalline tetragonal zirconia (type-1
experiments) also displays the broad moisture-related peak at
3420 cmꢁ1. The peak at 3420 cmꢁ1 is, however, very difficult to
eliminate in high surface area materials and hence the most that
can be said of the combustion produced amorphous zirconia is
that it is at worst a mildly hydrolyzed form of ZrO2. What the
data nevertheless confirms is that there is very little difference
between the X-ray tetragonal and X-ray amorphous zirconias.
The as-combusted amorphous zirconia is thus ‘‘almost’’ as
‘‘ZrO2’’ as its crystalline tetragonal counterpart.
4000 3500 3000 2500 2000 1500 1000 500
Wavenumber (cm–1
Fig. 7. IR spectra of (a) amorphous ZrO2 and (b) t-ZrO2.
)
spectroscopy in the as combusted mass thereby ruling out this
eventual temperature rise to reaction between any leftover
nitrate and urea. This temperature rise, as can be seen from
Fig. 3(d), is sufficient enough to cause coarsening beyond the
limit required for formation of tetragonal zirconia. TGA studies
on pellets soaked for 30 min following combustion displayed
weight losses of o5% at temperatures 44001C implying that
such heating is enough to get rid of most of the carbon.
The very fact that we have been able to observe an ‘‘amor-
phous’’ form of zirconia by combustion synthesis and the fact
that it has been reproduced over multiple experiments in a fairly
facile fashion is in itself very interesting. Further experiments are
underway to better understand the nature of this amorphous
zirconia. It opens up the possibility of exploiting SHS further for
processing metastable solid solutions through dissolution in
amorphous zirconia followed by controlled annealing. The re-
sults in the previous paragraph, however, clearly show that any
such heat treatment using thermal and/or environmental con-
trol, would have to circumvent the problem associated with re-
sidual carbon.
(2) Effect of Dough Composition on Porosity
The biggest advantage of the process described in this paper is
the ability to produce fine grain zirconia with a large surface
area, listed in Table III and a porous structure. The surface ar-
eas listed in Table III are from samples subjected to the 30 min
posttreatment. The discussion in the previous section showed
that the addition of starch had significant effect on the phase
composition through its influence on reaction zone temperature
and hence grain size, between 0 and 10 mg and relatively little
effect thereafter. In this section we show that in contrast starch
content between 10 and 500 mg has a significant effect on po-
rosity thus enabling the production of nanocrystalline tetragonal
zirconia with qualitatively and quantitatively different surface
areas and pore morphologies.
Postcombustion soaking in the crucible was found to even-
tually ignite the carbon left behind over a period of 2–5 min. The
exothermic peak observed in Fig. 4(b), trace e, is attributed to
the oxidation of carbon. This was verified by the absence of the
peak on performing the experiment in nitrogen instead of air. In
the process a temperature rise of up to 7301C was recorded by
the embedded thermocouple. It must be emphasized that no ni-
trate containing species could be detected by either XRD or IR
The SEM images of ZrO2 samples prepared using 0–500 mg
of starch are shown in Fig. 8 and the morphology-related data
are summarized in Table III. All the samples obtained by the
‘‘dough’’ method used herein have relatively speaking more sur-
face area and pore volume (Table III) than zirconia produced by

1754
Journal of the American Ceramic Society—Venugopal et al.
Vol. 94, No. 6
Table III. Specific Surface Area, Average Pore Diameter, and Total Pore Volume of Various ZrO2 Samples
Amount of starch used (mg)
Average pore diameter (mm)
Specific surface area (m2/g)
Total pore volume (cm3/g)
Morphology (from SEM)
0
42
61
47
47
38
26
110
78
0.04
0.04
0.13
0.09
Flaky
150
300
600
Tubular, porous
Tubes, porous
Flaky, porous
solution combustion synthesis as would be desirable for catalytic
applications. The 150 mg starch sample (Fig. 8(b)) is in partic-
ular interesting wherein pores of the order of 100 mm are sep-
arated by ‘‘walls’’ which in turn contain another level of porosity
in the range of 4–30 mm. The higher magnification images
(Fig. 8(c)) of the same sample show the formation of tubes
of high aspect ratio on the surface of the porous structure with
the average length and the diameter of the tubes being 10–15
and 1 mm, respectively. Such tubular structures were also
observed in the porous structure formed up to the addition of
500 mg of starch. While the morphology is affected by the starch
content given the nature of the combustion process, there is
not much control over tube dimensions. The general trend is
that the length and diameter of the tubular structures formed
increases with increase in the starch content in the reaction mix-
ture up to 500 mg starch. For larger amounts, e.g. 600 mg of
starch, only a flake-like morphology was observed (Fig. 8(d)).
Well-defined anisotropic structures were seen only in the 150 mg
starch sample.
The formation and evolution of tubular structures on the
addition of starch is thought to be due to the method of prop-
agation of the combustion zone, the amount and rate of gases
evolved, and the fraction of zirconia forming constituents in
the mixture. The following qualitative explanation is proposed.
As shown in Fig. 9, it is believed that the combustion from
start to finish can be divided into multiple stages. These are
similar to the ones discussed before in connection with TGA/
DTA data. The length and time scales at which they happen is
not clear yet. Step 1 involves drying and leaves behind a cracked
structure due to shrinkage on removal of water. Step 2 involves
combustion (nitrate decomposition and carbon burn-off)
and results in the conversion of the dried dough to zirconia.
This step is accompanied by the evolution of gases in addition
to water vapor due to drying. This combustion front is preceded
by the drying front and succeeded by the combusted front
as shown in step 3 at any point in time. Its propagation through
the structure leaves behind the porous combusted product as
seen in step 4.
For the formation of anisotropic tubular products, it is es-
sential that the gases produced either due to drying or combus-
tion should be able to leave the system through the existing
porous structure (either due to drying as seen in step 1 or pre-
viously combusted as in step 3) without causing a local pressure
build up. If the rate of combustion is too rapid or if the volume
of gases evolved is too large, then local pressure built up will
cause fragmentation of the combusted product. As mentioned
before in the absence of starch the reaction proceeds almost ex-
plosively. It is believed that the absence of any tubular structure
when no starch is added is thus due to very high rates of reac-
tion, resulting in local pressure build up and fragmentation of
the combusted product. As the starch level increases the reaction
proceeds at a lower rate in a manner that can be best described
as the movement of a smoldering front like that of a cigarette
stick as was also observed by Tahmasebi and Padyar.8 This lat-
ter method of combustion is more conducive to the formation of
anisotropic tubular products seen, as it allows enough time for
gases to escape from the combustion zone. This will result in the
anisotropic structure seen in step 4. However, increasing starch
levels has two effects. It increases the amount of gas evolved per
unit weight of the starting mixture and it also dilutes the con-
centration of zirconia forming elements in the mixture. It is ob-
vious that as the distribution of zirconia forming elements
becomes sparser their ability to form a connected porous prod-
uct as seen in Fig. 8(b) will diminish. Both, these features could
contribute to the absence of tubules at large amounts of starch
in excess of 600 mg.
The nonequiaxed morphology of the product implies that the
zirconias so formed are not suitable for applications that require
Fig. 8. SEM images of macroporous ZrO2 prepared using (a) no starch, (b) and (c) 150 mg, and (d) 500 mg of starch in the reaction mixture.

June 2011
Amorphous, Monoclinic, and Tetragonal Porous Zirconia
1755
References
1K. C. Patil, S. T. Aruna, and S. Ekambaram, ‘‘Combustion Synthesis,’’ Curr.
Opin. Solid State Mater. Sci., 2 [2] 158–65 (1997).
2M. Arimondi, U. Anselmi-Tamburini, A. Gobetti, Z. A. Munir, and G.
Spinolo, ‘‘Chemical Mechanism of the Zirconia Combustion Synthesis Reaction,’’
J. Phys. Chem. B, 101 [41] 8059–68 (1997).
3B. Venugopal, E. Samuel, C. Shivakumara, and M. Rajamathi, ‘‘Macroporous
Metal Oxide Foams Through Self Sustained Combustion Reactions,’’ J. Porous
Mater., 16 [2] 205–8 (2009).
Fig. 9. Stages involved during the combustion process that result in the
formation of anisotropic reaction products. With the passage of time the
uncombusted mixture first undergoes drying, then combustion to yield
the combusted product. UM, uncombusted mixture; DF, drying front;
CF, combusted front; CP, combusted product.
4R. C. Garvie, ‘‘Stabilization of Tetragonal Structure in Zirconia Microcrys-
tals,’’ J. Phys. Chem., 82 [2] 218–24 (1978).
5S. Arun, M. J. Mayo, W. D. Porter, and C. J. Rawn, ‘‘Crystallite and Grain
Size Dependent Phase Transformations in Yttria Doped Zirconia Powders,’’
J. Am. Ceram. Soc., 86 [2] 360–2 (2003).
6I. Molodetsky, A. Navrotsky, M. J. Paskowitz, V. J. Leppert, and S. H. Ris-
bud, ‘‘Energetics of X-Ray Amorphous Zirconia and the Role of Surface Energy
in its Formation,’’ J. Non-Cryst. Solids, 262 [1] 106–13 (2000).
7M. W. Pitcher, S. V. Ushakov, A. Navrotsky, B. F. Woodfield, L. Guangshe,
J. Boerio-Gates, and B. M. Tissue, ‘‘Energy Crossovers in Nanocrystalline
Zirconia,’’ J. Am. Ceram. Soc., 88 [1] 160–7 (2005).
good sinterability. Rather, the intended applications are ones
in which the large specific surface area can be retained, such as
in catalysis. Toward this end, the ability of these zirconias
to weather-specific environments would have to be suitably
assessed. Retention of the tetragonal phase is expected to be a
bigger problem than that of the large surface area. This would
then necessitate the need for addition of suitable stabilizers.
8K. Tahmasebi and M. H. Padyar, ‘‘The Effect of Starch Addition on Solution
Combustion Synthesis of Alumina–Zirconia Nanocomposite Powder Using Urea
as Fuel,’’ Mater. Chem. Phys., 109 [1] 156–63 (2008).
9S. T. Aruna and K. S. Rajam, ‘‘Mixture of Fuels Approach for the Solution
Combustion Synthesis of Alumina–Zirconia Nanocomposite,’’ Mater. Res. Bull.,
39 [2] 157–67 (2004).
V. Summary and Conclusions
10U. Anselmi-Tamburini, M. Arimondi, F. Maglia, G. Spinolo, and Z. A. Munir,
‘‘Ni/Yttria-Stabilized Zirconia Cermets from Combustion Synthesis: Effect of
Process Parameters on Product Microstructure,’’ J. Am. Ceram. Soc., 81 [7] 1765–
72 (1998).
Combustion of a dough, obtained by the addition of starch to
the starting ingredients of solution combustion synthesis, is con-
siderably different from what is usually considered to be SHS. In
particular, while the reaction is self-sustaining, the combustion
zone temperatures are much lower from those obtained by adi-
abatic calculations. In effect, the combustion zone temperature
can be controlled by the starch content of the starting dough.
This effect of such temperature control in turn helps control the
calcination of zirconyl nitrate and coarsening of the zirconia
crystallites formed. It thus enables the controlled synthesis of
pure zirconia in its monoclinic, tetragonal, and an amorphous
form thorough a possible combined effect of crystallite size and
hydrolysis level on phase stability. The X-ray amorphous zir-
conia observed is found to be partly hydrolyzed at worst. By
tuning the amount of starch in the starting dough and hence the
gas volume formed on combustion, nanocrystalline, (crystallite
size between 7 and 10 nm) tetragonal zirconia with a tubular
pore structure and surface area between 38 and 110 m2/g was
synthesized.
11T. Yamaguchi, ‘‘Application of Zirconia as a Catalyst and Catalyst Support,’’
Catal. Today, 20 [2] 199–218 (1994).
12U. Anselmi-Tamburini, G. Chiodelli, M. Arimondi, F. Maglia, G. Spinolo,
and Z. A. Munir, ‘‘Electrical Properties of Ni/YSZ Cermets Obtained Through
Combustion Synthesis,’’ Solid State Ionics, 110 [1–2] 35–43 (1998).
13D. J. Green, R. Hannik, and M. V. Swain, Transformation Toughening of
Ceramics. CRC Press, Boca Raton, FL, USA, 1989.
14W. Stichert, F. Schuth, S. Kuba, and H. Knozinger, ‘‘Monoclinic and
¨
¨
Tetragonal High Surface Area Sulfated Zirconias in Butane Isomerization: CO
Adsorption and Catalytic Results,’’ J. Catal., 198 [2] 277–85 (2001).
15R. C. Garvie, ‘‘The Occurence of Metastable Tetragonal Zirconia as a Crys-
tallite Size Effect,’’ J. Phys. Chem., 69 [4] 1238–43 (1965).
16S. R. Jain, K. C. Adiga, and V. R. P. Vernekar, ‘‘A New Approach to The-
rmochemical Calculations of Condensed Fuel-Oxidizer Mixtures,’’ Combustion
and Flame, 40, 71–9 (1981).
17U. Shulz, ‘‘Phase Transformation in EBPVD Yttria Stabilized Zirconia Ther-
mal Barrier Coatings During Annealing,’’ J. Am. Ceram. Soc., 83 [4] 904–10
(2000).
18Y. S. Touloukian and E. H. Buyco, Specific Heats of Non-Metallic Solids, Vol. 5.
Plenum, New York, 1970, pp. 6–14.
19G. S. Upadhyaya and R. K. Dube, Problems in Metallurgical Thermodynamics
and Kinetics. Pergamon, Oxford, 1997.
20D. R. Lide, CRC Handbook of Chemistry and Physics, Internet Version 2005.
CRC Press, Boca Raton, FL, USA, 2005.
Acknowledgments
21F. D. Rossini, D. D. Wagman, W. H. Evans, S. Levine, and I. Jaffe, Selected
Values of Chemical Thermodynamic Properties. National Bureau of Standards,
B. R. V. thanks CSIR, New Delhi for the award of Senior Research Fellow-
ship. S. N. was supported by DST/SR/S2/CMP-02/2007. The authors thank De-
partment of Materials Engineering, IISc, Bangalore, for recording SEM images,
Department of Chemistry, Bangalore Institute of Technology, Bangalore, for hav-
ing carried out surface area and related measurements, and CMSK Reddy for
carrying out the TGA/DTA. One of the authors S. R., acknowledges discussions
with A.S. Gandhi, Metallurgical and Materials Engineering, IIT, Chennai.
Washington, DC, USA, 1952.
&