Journal of Thermal Analysis and Calorimetry, Vol.

84 (2006) 2, 521–525

THERMAL CHARACTERISATION OF AUTOCLAVED CEMENT MADE

WITH ALUMINA-SILICA RICH INDUSTRIAL WASTE

H. Connan1, D. Klimesch1,2, A. Ray1,2* and P. Thomas1

1
Department of Chemistry, Materials and Forensic Science, University of Technology, Sydney, P.O. Box 123, Broadway
NSW 2007, Australia
2
Centre for Built Infrastructure Research, Faculty of Engineering, University of Technology, Sydney, P.O. Box 123, Broadway
NSW 2007, Australia

The hydration products of hydrothermally cured OPC-quartz blends incorporating clay-brick waste were characterised using simulta-
neous DTA-TG, XRD and SEM. The exotherm above 800°C on the DTA curve, due to the formation of b-wollastonite (b-CS) was re-
lated to the compressive strength. The area and/or height of the exotherm increased with compressive strength and was due primarily to
the presence of calcium silicate hydrate phases including 1.1 nm tobermorite. Supporting evidence is provided by XRD and SEM.

Keywords: autoclaving, clay-brick waste, compressive strength, thermal analysis, tobermorite

Introduction binder. Factors that are considered important in rela-

tion to compressive strength include the amount of
The trend towards the use of renewable resource mate- crystalline to amorphous binding material at a given
rials in the manufacture of building products is driven porosity, the particle size distribution and amount of
by both economic factors and public awareness of en- unreacted starting material and the crystal habit of
vironmental issues. Each year vast quantities of con- tobermorite [7–13]. This article presents preliminary
struction and demolition (C&D) waste are deposited in results from simultaneous DTA-TG, X-ray powder
landfills globally. For example, according to the Envi- diffraction and scanning electron microscopy in
ronmental Protection Authority (EPA) of NSW, Aus- relation to compressive strength development.
tralia, approximately 25% of C&D waste, equating to
approximately 1 million tonnes, ended up in landfills
in 1999 alone [1], and this situation can be avoided by Experimental
making use of emerging technologies. Alumina-silica
rich industrial waste in the form of fired clay-brick is The feasibility of utilizing clay-brick waste for the
generated in abundance on a global scale during pro- manufacture of cement-based building products was
duction, transportation and construction procedures the subject of previous research examining the com-
where large amounts are broken. bined influence of factors on the strength develop-
Previous research has shown that clay-brick, in a ment using a factorial experimental design [14]. The
finely ground state, exhibits high pozzolanic activity factors chosen were: autoclaving time, autoclaving
under ambient conditions [2]. Hydrothermal curing or temperature and cement: quartz replacement level
autoclaving, a process that affords rapid strength de- (additive amount) with the response variable being
velopment by the use of saturated steam under pres- compressive strength. A more detailed account of the
sure, enhances the chemical reactivity of pozzolanic factorial experimental design approach is given else-
materials [3]. Klimesch and Ray demonstrated that where [15]. Table 2 in the results and discussion
the formation of 1.1 nm tobermorite, the principal section provides a summary.
binding agent of calcium silicate based products, is
enhanced by the addition of finely ground clay-brick Materials
to mixtures of ordinary Portland cement (OPC) and
quartz sand [4–6]. • Type A ordinary OPC produced by Australian Ce-
Many studies have been reported on the com- ment, Auburn, NSW, Australia with a Blaine fineness
pressive strength development of autoclaved calcium of 350 m2 kg–1. It contained SiO2 20%, CaO 64.2%,
silicate materials containing tobermorite as the main Al2O3 4.5%, Fe2O3 3.7% and SO3 3.5%.

* Author for correspondence: [email protected]

1388–6150/$20.00 Akadémiai Kiadó, Budapest, Hungary

© 2006 Akadémiai Kiadó, Budapest Springer, Dordrecht, The Netherlands
 CONNAN et al.

• Ground quartz sand (Quartz) supplied by Unimin Pastes were cast into stainless steel moulds and con-
Australia Limited, Sydney, Australia with a mean solidated on a vibratory table, followed by precuring
particle diameter of 22 micron and a Blaine surface for 24 h in a moist cabinet. For each mix three 50 mm3
area of 360 m2 kg–1. It contained SiO2 98.9%, cubes were cast. Demoulded cubes were autoclaved
Al2O3 0.6%, Fe2O3 0.1% and TiO2 0.06%. according to the following autoclave regime: 1.5 h
• Clay-brick waste supplied by a major clay-brick ramping up to the required temperature, followed by 6
manufacturer in Sydney, Australia, was crushed to or 8 h of steam saturation and then ramping down to
less than 2.38 mm followed by dry ball milling to ambient temperature. Autoclaved specimens were
achieve a particle size distribution comparable to then allowed to air dry for 48 h prior to mechanical
that of the quartz used. The major crystalline testing. The bulk density of cubes was determined by
phases of the clay-brick, as determined by X-ray mass and volume measurements. Compressive
powder diffraction (XRD), were: quartz, mullite, strength was measured according to ASTM C109-02
hematite, with minor amounts of cristobalite and [17]. The specimens were loaded under compression
rutile. The major oxides as determined by X-ray until failure using a Tinius Olsen instrument. Average
fluorescence (XRF) are given in Table 1. compressive strength data of all acceptable test speci-
mens per mould are reported to the nearest 0.1 MPa.
Table 1 Major oxide composition of brick fines from XRF
analysis
After mechanical testing samples were oven dried
overnight at around 105°C followed by disc milling.
Major oxides Mass% Duplication of the experimental design as outlined in
SiO2 69.71 Table 2 was performed.
Al2O3 18.64
Fe2O3 7.48 Methods
K2O 1.88 A TA Instruments SDT 2960 simultaneous DTA-TG
MgO 0.95 analyzer was used to measure the mass change and
TiO2 0.84 temperature difference of samples with increasing
Na2O 0.44 temperature. The mass range of powdered samples
CaO 0.21
was between 20–30 mg. Each sample was evenly
packed into an open Pt–Rh crucible with 20 taps.
LOI 0.08
Samples were analysed at a heating rate of 10°C min–1
LOI: loss on ignition at 1050°C for 2 h under flowing air at 20 mL min–1 from 25 to 1000°C.
DTA-TG curves were evaluated using the TA Instru-
Sample preparation ments analysis software with two-point rotations car-
OPC–quartz–clay-brick blends were prepared using a ried out for all curves.
water-to-total solids ratio (w/s) of 0.33 as this yielded X-ray diffraction analysis was carried out using
similar workability for all the mixtures. A cement: a Siemens D5000 X-ray powder diffractometer with
quartz ratio comparable to commercially manufac- symmetrical Bragg–Brentano geometry. An X-ray
tured building products of 0.67 was used. Mechanical tube was used at a power of 40 mA and 40 kV to pro-
mixing of these blends was conducted in accordance duce CuKa radiation (l=0.1540981 nm). Samples
with ASTM C305-99 [16]. Clay-brick was added at were mounted in plastic sample holders. The holder
4.3 and 8.6 mass% as cement:quartz replacement. and sample were placed on the diffractometer and an

Table 2 Results for autoclaved cement blends using TA instruments DTA-TG data analysis software. Compressive strength re-
sults are the average values from duplicate runs
Autoclave Autoclave Additive Compressive Exotherm peak
Sample
time/h temperature/°C amount/% strength/MPa maximum/°C
1 8 170 8.6 77.7 846
2 8 170 4.3 75.0 845
3 6 170 4.3 72.3 847
4 6 170 8.6 68.2 847
5 6 180 4.3 64.7 846
6 8 180 4.3 60.8 844
7 8 180 8.6 59.0 846
8 6 180 8.6 54.4 846

522 J. Therm. Anal. Cal., 84, 2006

 AUTOCLAVED CEMENT MADE WITH ALUMINA-SILICA RICH INDUSTRIAL WASTE

XRD pattern collected for 2q values from 3.000 to

65.000°, at intervals of 0.020°. Step times of 6 s per
interval were employed and the receiving-slit width
was 0.2 mm. The ICDD-JCPDS Powder Diffraction
File CD-ROM was used for qualitative phase identifi-
cation. A fundamental parameters approach was used
for X-ray line profile fitting in order to determine
peak heights and areas [18].
Microstructural analysis of the cement samples
was carried out on a Jeol 6300F field emission scan-
ning electron microscope (FEGSEM) at an accelerat-
ing voltage of 30 kV. Samples were deposited on alu-
minum discs and placed under vacuum prior to carbon
coating. Images were recorded on either a Mitsubishi Fig. 1 Representative DTA curves for hydrothermally treated
video printer or captured digitally using Moran Scien- specimens containing clay-brick waste of increasing
tific X-ray mapping software. Images that were col- strength from bottom to top as Table 2
lected digitally were acquired at 512 by 512 pixel
resolution. and the content of C–S–H(I) in the absence of supple-
mentary addition of Al-containing materials. In contrast
when Al-containing materials were used the peak height
Results and discussion increased and was attributed to Al-tobermorite. It fol-
lows, that the observed increase in exotherm area and
Differential thermal (DTA) curves of autoclaved speci- height can therefore be attributed to the development of
mens are depicted in Fig. 1. The main observed endo- increasing quantities of calcium silicate hydrate phases
therms (endo) and exotherms (exo) are ascribed to: including 1.1 nm tobermorite.
The relationship between the compressive
presence of calcium silicate
ca. 105 to 260°C (endo) strength as a function of bulk density is shown in
hydrates (C–S–Hs) including
not shown Fig. 3. The data show increased compressive strength
1.1 nm tobermorite (C5S6H5)
presence of a member of the with increasing bulk density. Previous work showed
ca. 260 to 350°C (endo) that generally, there is a reverse relationship between
hydrogarnet series, C3AS3–xH2x,
not shown
x=0 to 3 bulk density and apparent porosity. It may be con-
crystalline inversion due to cluded, that increased compressive strength was a
ca. 573°C (endo)
unreacted quartz (Q) consequence of the filling up of available pore vol-
ca. 600 to 700°C (endo) decarbonation of CaCO3 ume with the crystallization of hydration products, in-
crystallization of cluding tobermorite, which correlates well with the
beta-wollastonite (b-CS) from thermal data. We are currently conducting a more de-
ca. 800 to 900°C (exo)
C–S–Hs including (C5S6H5) tailed study of microstructural attributes including the
were present
The variations in the observed exotherm temper-
atures with compressive strength are given in Table 2
along with sample ID and experimental conditions.
The exotherm temperatures of the autoclaved blends
were within a very narrow range, suggesting that the
binding material present was of a similar nature.
It is well known that the exotherm profile due to
b-CS formation is affected by the nature of the pre-
cursor calcium silicate hydrate phases including:
Ca/Si ratio, Al-content and presence of other foreign
ions such a K+ and Na+ [19, 20].
The relationship between compressive strength vs.
exotherm area and height is shown in Fig. 2. Purton [9]
found that for autoclaved cement-bound specimens the
area of the exotherm increased with the amount of Fig. 2 Showing the correlation between compressive strength
binder in the original mix. Alexanderson [8] found a and the area and height of the exothermic peak on DTA
positive correlation between the exotherm peak height curves with duplicates plotted; a. u. – arbitary units

J. Therm. Anal. Cal., 84, 2006 523

 CONNAN et al.

Fig. 5 Tobermorite in sample 1 with compressive strength of

77.7 MPa. Field of view (18.9×18.9 mm); magnifica-
tion ×5000
Fig. 3 Bulk density of autoclaved blends as a function of com-
pressive strength for hydrothermally treated specimens
containing clay-brick waste

nature of the pore structure and will report our find-

ings in future publications.
As the exotherm temperature range for the speci-
mens was within a very narrow range (Table 2) it may
be concluded that the relationship between compres-
sive strength and exotherm area/height is a function
of the amount of binding material present.
Figure 4 shows the ratio of the relative XRD in- Fig. 6 Tobermorite in sample 2 with compressive strength of
tensities of the 2.97/3.08  lines. With one exception, 75.0 MPa. Field of view (18.9×18.9 mm); magnifica-
the data fall into two distinct groups for ratios of tion ×5000
namely 0.35 to 0.40 and 0.45 to 0.50 ranges. Accord-
ing to Kalousek [21], this ratio of intensities can be
used to differentiate between Taylor’s C–S–H(I) (in-
tensity ratio=0) and tobermorite (intensity ra-
tio=0.7–0.8). It may be inferred that either the higher
strength binder contains a more crystalline tober-
morite, or alternatively, the ratio of crystalline to
amorphous binding material has increased. Support-
ing evidence is provided by SEM.
For instance, from SEM observations the high
strength specimen group, depicted in Figs 5 and 6, is
predominantly of a lath-like morphology whereas the Fig. 7 Tobermorite in sample 7 with compressive strength of
59.0 MPa. Field of view (18.9×18.9 mm); magnifica-
tion ×5000

Fig. 8 Tobermorite in sample 8 with compressive strength of

Fig. 4 Ratio of relative 2.97/3.08  X-ray line intensities of the 54.4 MPa. Field of view (18.9×18.9 mm); magnifica-
binder for hydrothermally treated specimens containing tion ×5000
clay-brick waste as a function of compressive strength

524 J. Therm. Anal. Cal., 84, 2006

 AUTOCLAVED CEMENT MADE WITH ALUMINA-SILICA RICH INDUSTRIAL WASTE

low strength specimens (Figs 7 and 8) are more nee- 3 H. F. W. Taylor, Cement Chemistry, 2 nd Ed. Thomas Tel-
dle-like. It has been reported, that the morphology ford, London 1997.
change, plate through lath-like to needle-like, is associ- 4 D. S. Klimesch, A. Ray and J.-P. Guerbois, Thermochim. Acta,
389 (2002) 195.
ated with a decrease in tobermorite crystallinity [12].
5 D. S. Klimesch, M. Gutovic and A. Ray, J. Therm. Anal. Cal.,
It is worth noting that the autoclaving tempera- 75 (2004) 197.
ture (Table 2) appears to play a significant role in the 6 D. S. Klimesch and A. S. Ray, International Conference
development of the particular morphologies ob- on Composites in Construction, Porto, Portugal,
served. We are currently undertaking further investi- Oct. 10–12, 2001, p. 47.
gations into the effect of temperature in relation to 7 H. F. W. Taylor, Autoclaved Calcium Silicate Products,
strength development. T.S.o.C. Industry, Editor, 1967, p. 195.
8 J. Alexanderson, Cement Concrete Res., 9 (1979) 507.
9 M. J. Purton, J. Appl. Chem., 20 (1970) 293.
Conclusions 10 N. Isu, S. Teramura, H. Ishida and T. Mitsuda,
Cement Concrete Res., 25 (1995) 249.
11 P. D. Rademaker and V. Reiman, ZKG International,
From the data presented in this study we conclude: Edition B, 47 (1994) 636.
• DTA-TG is a very useful technique in the study of 12 N. Isu, H. Hibino, S. Teramura, H. Ishida and T. Mitsuda,
hydrothermally cured CaO–SiO2–H2O and Seramikkusu Kenkyu Shisetsu Nenpo (Nagoya Kogyo
CaO–Al2O3–SiO2–H2O systems. Daigaku), 1994, p. 31.
• Compressive strength is related to a combination of 13 M. O’Farrell, S. Wild and B. B. Sabir, Cement Concrete
Composites, 23 (2001) 81.
the quantity of binding material and crystal habit.
14 H. G. Connan, D. S. Klimesch, A. S. Ray and P. S. Thomas,
• In the system studied the exotherm peak above 800°C Advances in Cement Research, 17 (2005) 133.
was a good indicator of the quantity of calcium sili- 15 E. P. B.George, W. G. Hunter and J. S. Hunter, Statistics
cate hydrate phases within OPC–quartz blends incor- for experimenters: an introduction to design, data analysis,
porating clay-brick waste. Compressive strengths and model building, Wiley, New York 1978, p. 653.
were proportional to the exotherm area and height. 16 Standard Practice for Mechanical Mixing of Hydraulic Ce-
• From the relative XRD intensities of the ment Pastes and Mortars of Plastic Consistency, in
2.97/3.08  lines, it may be inferred that either the ASTM C 305-99, ASTM International, Philadelphia 1999.
higher strength binder contains a more crystalline 17 Standard Test Method for Compressive Strength of Hy-
draulic Cement Mortars (Using 2-in. or [50-mm] Cube
tobermorite, or alternatively the ratio of crystalline
Specimens), in ASTM C 109/C 109M-02. 2002, ASTM
to amorphous binding material has increased. Sup- International: Philadelphia.
porting evidence was provided by SEM. 18 R. W. Cheary and A. Coelho, J. Appl. Crystallogr.,
• A lower autoclaving temperature when using 25 (1992) 109.
clay-brick waste was found to be beneficial for the 19 T. Mitsuda, S. Kobayakawa and H. Toraya, 8th International
formation of crystalline lathe-type tobermorite pro- Congress of the Chemistry of Cement, Rio de Janeiro 1986.
ducing higher compressive strengths when com- 20 G. L. Kalousek, J. Am. Ceram. Soc., 40 (1957) 74.
pared to the higher autoclaving temperature. 21 G. L. Kalousek, 5th International Symposium of the Chem-
istry of Cement, Tokyo 1968.

References Received: September 14, 2004

Accepted: February 7, 2006
1 http://www.wrapp.nsw.gov.au/material/cd.shtml.
2 S. Wild, J. Szwabowski, A. Gailius and H. Hansen, European
Commission, 1995–1998, European Commission: Brussels. DOI: 10.1007/s10973-005-6674-z

J. Therm. Anal. Cal., 84, 2006 525