Escolar Documentos
Profissional Documentos
Cultura Documentos
S0261-3069(13)00490-1
http://dx.doi.org/10.1016/j.matdes.2013.05.058
JMAD 5496
To appear in:
Received Date:
Accepted Date:
18 March 2013
17 May 2013
Please cite this article as: Hajjaji, W., Andrejkoviov, S., Zanelli, C., Alshaaer, M., Dondi, M., Labrincha, J.A.,
Rocha, F., Composition and technological properties of geopolymers based on metakaolin and red mud, Materials
and Design (2013), doi: http://dx.doi.org/10.1016/j.matdes.2013.05.058
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1.Geobiotec,
2.Materials
3.Istituto
4.Materials
* Author to whom correspondence should be addressed; phone: +351-234370250; fax: +351234370204; email: w.hajjaji@ua.pt
Abstract
New geopolymer formulations were designed by sodium silicate/NaOH activation of
metakaolin, iron oxide and red mud mixtures. The effects of source materials on the
microstructure and mechanical properties were studied. Each formulation induces
different degree of geopolymerization reaction as reflected by the phase composition
where the amorphous phase is predominant. These vestiges are related to silica provided
by sodium silicate more reactive in the geopolymerization than the silica of metakaolin.
Moreover, the variation in strength between the geopolymers is attributed to the same
factors, with higher porosity and nonreacted phases found in the red mud based
geopolymer matrix. In function of curing time, the mechanical strength increased from
day 1 to 28 for the samples with a low amount of red mud. In these two cases, longer
curing time improves the geopolymerization state resulting in higher compressive load.
The metakaolin and metakaolin/red mud products exhibited comparable water
absorption and density.
1. Introduction
In the Bayer process for alumina production, five tons of bauxite are washed and treated
to produce two tons of aluminum metal [1]. The world annual production of 21 million
ton of aluminum generates 82 million ton of sludge. This extraction residue, highly
alkaline (pH=10) and known as red mud, is produced in huge amounts, evacuated and
disposed in huge landfills [2,3]. Red mud could cause serious environmental problems:
i) contamination of surface and underground water resources with NaOH and metallic
oxide-bearing impurities [4]; ii) direct contact with fauna and flora; iii) evaporation that
could originate highly alkaline rainfalls [5]; iv) visual impact on extensive areas. Some
accidents, the most recent one in Ajka (Hungary, 2010) [6,7], provoked irrecoverable
damages to the environment.
Many recent studies and semi-industrial trials were directed to the incorporation of the
red mud in construction materials; traditional ceramics [8,9], clinker [10,11], mortar and
concrete [10,12], cements [13], light weight aggregates [14] etc. The present work is
aimed at assessing red mud as raw material for new geopolymer-like compounds that
could be used for construction or restoration purposes [15-17].
For several considerations, the geopolymers appears to be a potential alternative to the
classic hydraulic binders. Nevertheless, their chemical composition is very different
from that of cements or lime. Cements owe their mechanical properties to the formation
of the hydrated calcium silicates (C-S-H) while the exothermic reaction of
geopolymerisation generates a structure closer to zeolites or aluminosilicate gels [18].
This alkaline aluminosilicate material, generally amorphous is mainly produced from
metakaolin (source of Si and Al) reacting with hydroxides or alkaline silicate solutions
sialate
siloxo[ Si O Al O Si O]
or
sialate
disiloxo
2. Experimental details
Geopolymers were designed by using metakaolin 1200S (MK) (AGS Mineraux, France)
as source of alumino-silicate and red mud, whose chemical composition is reported in
Table 1. In water medium, alkaline activators NaOH (ACS AR Analytical Reagent
Grade Pellets) and hydrated sodium silicate (Merck, Germany; 8.5 wt.% Na2O, 28.5
wt.% SiO2, 63 wt.% H2O) were used to dissolve aluminosilicate and avoid residual
sodium [30]. The target was the following molar oxide ratios: SiO2/Al2O3 = 1,
Na2O/Al2O3 = 1 [19,31]. The water content for all the samples was kept the same with a
molar ratio of H2O/Na2O = 17. The following compositions were prepared: a
geopolymer with metakaolin (named GMK) and others where metakaolin was
substituted by 1/4, 1/6, 1/8, 1/10 and 1/12 of red mud (named GR4, GR6, GR8, GR10
and GR12, respectively). In addition, 1/12 of Fe2O3 (Sigma-Aldrich, as reference) was
introduced to compare the activation power of red mud (sample GFE).
The mixing of the blends was carried out by Heidolph ST-1 Laboratory stirrer at two
different speeds; 100 rpm for 2 min and 200 rpm for 4 min, to insure their homogeneity
and avoid bubbles and agglomeration into the sample. The pastes were immediately
poured into 20x20x20 mm cubic molds and placed in oven at 50C for 24 h and after
left at room temperature for one day. Curing was carried out by keeping the geopolymer
cubic specimens in distilled water from 1 to 28 days. While standard tests usually
perform curing under controlled relative humidity in environmental chamber [32], we
need extreme conditions (sample immersed in water) to be sure that geopolymers are
fully stable and dissolution does not affect the final properties [33,34].
The X-ray diffraction was conducted on a Rigaku Geigerflex D/max Series instrument
(CuK radiation in the 480 2
X'Pert HighScore Plus. Moreover, to evaluate the phase content, the powders were
characterized by X-ray diffraction (D8 ADVANCE, LynkEye detector-Bruker AXS,
Germany) using CuKa radiation in the 10-80 2 range, scan rate of 0.02 (2 ), and 185
s equivalent per step. The quantitative phase analysis was performed using TOPAS 4.2 BRUKER software following RIR (Reference Intensity Ratio) and Rietveld refinement
techniques. The samples were admixed with 20 wt.% corundum, used as internal
standard. Each X-ray powder diffraction pattern consists approximately of 7000 data
point and 700 reflections; up to 40 independent variables were refined: phase fractions,
zero point, 2530 coefficients of the shifted Chebyschev function to fit the background,
unit cell parameters, profile coefficients (one Gaussian, Gw, and one Lorentzian term,
Lx). The agreement indices, as defined in TOPAS, for the final least-squares cycles of
all refinements are represented by Rp (%), Rwp (%), GOF. For the refined patterns, they
were found in the following ranges: 2.5%<Rp<4.0%, 3.0%<Rwp<5.0% and
1%<GOF<2%. The experimental error is within 5% relative.
The X-Ray Fluorescence (XRF) was conducted using Philips XUNIQUE apparatus.
The WD-XRF spectrometry was performed on glass beads, obtained from a mixture of
66 wt.% lithium tetra-borate and 34 wt.% lithium meta-borate, where 10 wt.% of the
sample was dissolved at high temperature. The bead was then analysed for elemental
composition, according to the calibrated procedure. The experimental error is within 1%
relative.
The XRD patterns of studied metakaolin (MK) and red mud (RM) are reported in Fig. 1.
MK shows a broad reflection centred at 2 =24 attributable to the amorphous
metakaolinite along with residual quartz, illite and anatase (Fig. 2). The SiO2/Al2O3
weight ratio, which should be close to 1.15 in case of metakaolinite, is higher because of
quartz and illite occurrence. The plate-like morphology, characteristic of metakaolinite
particles, is shown in Fig. 2 (particle A: ratio Si/Al
The chemical composition of the red mud (RM), (Table 1) is predominantly iron oxide
and hydroxide (the main crystalline phases being hematite and goethite, Fig. 1) and
aluminium hydroxides (gibbsite and boehmite), while sodium is introduced by
lixiviating agent (caustic soda). The red mud contains also titanium mainly in the form
of rutile, which is a common accessory mineral in bauxite deposits [38] along with
some amorphous compound.
The particle size distribution of MK is coarser (mean size around 5 m) than that of RM
one, whose mean size is below to 1 m) (Fig. 3). The ball-milled red mud presents a
bimodal distribution due to heterogeneity in composition of this sludge.
geopolymerization in both GMK and GR samples. The reaction with the alkaline
solution to form a particulate gel network, characteristic of metakaolin geopolymers
[39], took place at the border of particles (Fig. 6a) then involving the entire surface (Fig.
6b). This amorphous matrix is mainly composed by Si, Al and Na, but Fe seems to be
incorporated as occurs in iron rich GFE and GR4-12 binders [28]. This hypothesis is
supported by the EDS microanalysis (Table 3) and by the difference between the initial
amount of iron oxide and the hematite and goethite residual phases that is as high as
almost 5 to 8% Fe2O3. By increasing the red mud content, the amount of non-dissolved
fillers in the geopolymer binder increased from 1.5 to 3.3%.
In addition, the presence of quartz and metakaolinite is observed as partially reacted
vestiges (Fig. 6) suggests that silica provided by sodium silicate is more active in the
geopolymerization process than the silica introduced by metakaolinite. According to
Pimraksa et al. [33], the Si/Al ratio significantly affects the degree of polymerization.
In the case in hand, the SiO2/Al2O3 ratio was determined by EDS (A1 and A2 in Table
3) to be as low as 1.9 (so well below 3). This EDS quantification matches a threedimensional sodium silico-aluminate structure [40] combining poly-sialate (SiOAl
O) with a Si/Al ratio of 1. A structure like this one, formed by linked SiO4 and AlO4
tetrahedra sharing all the oxygen atoms, requires a lower ratio, i. e. Si/Na the half in the
case of GR (1.7) if compared to GMK (3.3) due to the supplementary content of Na
introduced by the red mud.
The compressive strength (Fig. 7) gave initial maximum values (at 1 day curing) around
8 MPa. The amount of RM had a variable effect on the mechanical properties of
geopolymer. The resistance of sample GR4, for instance, decreased to minimum values
below 5 MPa. Hence, the lower the SiO2/Al2O3 ratio, the weaker is the strength of
geopolymer. The same mechanical performance was observed in the GFE sample (i.e.,
MK/commercial iron oxide ratio 1/12) that exhibits the lowest compressive strength. In
the red mud based equivalent sample GR12 (MK/red mud ratio 1/12), the occurrence of
iron was compensated by the highly alkaline nature of the sludge. By growing the RM
content, also the Na2O/SiO2 ratio increased and a higher strength was measured for the
1/10, 1/8 and 1/6 ratios. This trend is consistent with previous works, as Duxson et al.
[41] and Stevenson and Sagoe-Crentsil [42], which showed that the compressive
resistance is maximal when the of Na2O/Al2O3 molar ratios are between 1 and 3. Over
and below this window, as it occurs for GR4 and GFE, respectively, the polymerized
network is less stable and easily disintegrated [27]. Moreover, the values of mechanical
strength of GR6-10-12 are close that of GMK and in the case of GR10 at 28 days even
better (10.8 MPa). Along with the rising Na2O content, it is the fine granulometry of
RM to probably play an important role by ensuring the cohesion of the geopolymeric
paste. This observation stresses the importance of physical characteristics in addition to
chemical effects.
After curing, the GMK sample showed slight variations in mechanical behavior (Fig. 7).
This is likely stemming from the stable structure and higher polymerization degree. By
extending the curing time, the mechanical strength increased considerably from day 1 to
day 28 for the samples GR10 and GR12 and in lower degree for GR6 (that appears to be
nearly stable). In cases of GR10 and GR12, longer curing time improves the
geopolymerization state resulting in higher compressive strength.
10
On other hand, the geopolymer GR8 showed a opposing behaviour. This could be
explained by dissolution of non-complexed species (like Na or sulphate ions) and
gradual destabilization of the framework in presence of water before 28 days. This
presumed lower degree of geopolymer reaction should be promoted by non-controlled
stoichiometry when RM is added.
Table 4 shows the water absorption, open porosity, bulk density and apparent specific
weight of geopolymers cured for 1 day. The metakaolin-based geopolymers (with or
without RM) basically present similar behavior with high values of water absorption
(44-48 %) and open porosity, while low bulk density (1.0-1.1 g/cm3) and apparent
specific weight are displayed.
A relationship arose between leaching results (Table 5) and compressive strength
values: samples with leachate concentrations close to 100 ppm exhibit the best
mechanical performances. However, both low (GR4 and GR8) and high (GFE) leachate
concentrations result in low strength, this trend might be related also to variation of
sodium content in initial formulations and consequent destabilization of geopolymer
stoichiometry by incorporation of commercial or sludge iron sources. Due to the high
porosity, sodium forming part of the geopolymer network is easily extractable by
Na+
4. Conclusions
11
composition (amorphous phase predominant over residual quartz, illite and anatase) is
slightly affected by the red mud, despite it provides aluminum and iron oxides and oxyhydroxide (hematite, goethite, gibbsite and boehmite). These phases, in fact, are to a
limited extent involved in the geopolymerization process and the alkaline
aluminosilicate amorphous phase is about 90 wt.% in all samples. The structural
features of the amorphous geopolymer, as resumed by the broad hump of XRD patterns,
are not modified by addition of RM.
The physical properties of geopolymers are not significantly affected by RM, as all
samples exhibit high values of water absorption and low apparent density. The
Chemical stability is good: sodium leaching test gave leachate concentrations close to
100 ppm without evidences of deterioration of mechanical performance. Red mud
influences the mechanical strength during curing (especially at the higher amounts of
RM but the lower additions (GR10 and GR12), cured for 28 days showed good
compressive strength.
In general, this study strengthens preliminary observations that red mud constitutes a
suitable material for geopolymer production once applied in reasonable ratios with
respect to metakaolin.
Acknowledgements
The work was supported by FCT-Grant SFRH/BPD/72398/2010 co-financed by
Programa Operacional potencial Humano POPH.
12
References
[1] Liu Yong, Lin Chuxia, Wu Yonggui. Characterization of red mud derived from a
combined Bayer Process and bauxite calcination method. J Hazard Mater 2007;
146:255261.
[2] Rai Suchita, Wasewar KL, Mukhopadhyay J, Kyoo Yoo Chang, Uslu Hasan.
Neutralization and utilization of red mud for its better waste management. Arch
Environ Sci 2012;6:13-33.
[3] Power G, Grfe M, Klauber C. Bauxite residue issues: I. Current management,
disposal and storage practices. Hydrometallurgy 2011;108:33-45.
[4] Ramesh Chandra Sahu, Rajkishore Patel, Bankim Chandra Ray. Neutralization
of red mud using CO2 sequestration cycle. J Hazard Mater 2010;179:2834.
[5] Tamotia SK. Management of Red Mud. In: International Symposium on
Processing of fines - Vol.2. NML Jamshedpur. Jamshedpur, 2000, 430-434.
[6] Burke IT, Mayes WM, Peacock CL, Brown AP, Jarvis AP, Gruiz K. Speciation
of arsenic, chromium and vanadium in red mud samples from the Ajka spill site,
Hungary. Environ Sci Technol 2012;46:3085-3092.
[7] Renforth P, Mayes WM, Jarvis AP, Burke IT, Manning DAC, Gruiz K.
Contaminant mobility and carbon sequestration downstream of the Ajka
(Hungary) red mud spill: The effects of gypsum dosing. Sci Total Environ
2012;421422:253-259.
[8] Amritphale SS, Patel M. Utilisation of red mud, fly ash for manufacturing bricks
with pyrophyllite. Silic Ind 1987;2:31-35.
13
[9] VincenzoMS, Renz C, Stefano M, Giovanni C. Bauxite red mud in the ceramic
industry. Part 2: production of clay based ceramics. J Eur Ceram Soc
2000;20:245252.
[10]
Ribeiro DV, Labrincha JA, Morelli MR. Effect of the addition of red
aggregates using red mud, the by-product from Al2O3 industry. J Korea Solid
Wastes Eng Soc 1998;15:16-23.
[15]
14
[17]
Deventer JSJ. Geopolymer technology: the current state of the art. J Mater Sci
2007;42:2917-2933.
[20]
Wang Jing, Zhang Chi, Xu Jing, Qu Peng, Zhou Ye, Min Han Hui. The
15
geopolymers and composites derived from sodium polysialate. Mater Res Bull
2003;38:319-331.
[31]
16
[33]
leaching of granular waste materials and sludges, parts 14, CEN, Brussel, 1999.
[37]
Zhang Zuhua, Wang Hao, Provis JL, Bullen F, Reid A, Zhu Yingcan.
of
metakaolin
with
sodium
hydroxide.
Thermochim
Acta
2012;539:23-33.
[40]
Duxson P, Provis JL, Lukey GC, Mallicoat SW, Kriven WM, van
17
18
Table captions:
Table 1. Chemical compositions of metakaolin (MK) and red mud (RM).
Table 2. Phase composition of geopolymers (averagestandard deviation).
Table 3. Microchemical analysis of GMK and GR4 samples (see Fig. 6 for spots).
Table 4. Water absorption, open porosity and bulk density of geopolymers.
Table 5. Sodium leaching from samples.
Figure captions:
Fig. 1. XRD patterns of metakaolin (MK) and red mud (RM).
Fig. 2. SEM micrograph (ticks span 8.57 m) and chemical composition (by EDS) at
spots A and B of metakaolin MK.
Fig. 3. Particle size distribution of metakaolin (MK) and red mud (RM).
Fig. 4. Geopolymer cubic specimens (20mm side); from left to right: metakaolin-based
(GMK), with iron oxide (GFE), and mixtures with red mud (GR4, GR6, GR8, GR10
and GR12).
Fig. 5. (a) Plot of Rietveld refinement performed on X-ray powder diffraction data. The
experimental data are indicated by plus signs, the calculated pattern is the continuous
line and the lower curve is the weighted difference between the calculated and observed
patterns. The rows of vertical tick marks shows the allowed reflections for the
crystalline phases present in the sample: 1) Illite, 2) Anatase 3) Corundum (added 20
wt.% as internal Standard 4) Quartz. (b) Comparison of geopolymers based on
metakaolin (GMK) and containing iron-oxide (GFE) or red mud (GR12).
Fig. 6. SEM micrographs of GMK (a) and GR (b) geopolymers.
Fig. 7. Compressive strength of geopolymers at different curing times (1 and 28 days).
Table 1
Oxides (wt.%)
MK
RM
SiO2
TiO2
Al2O3
Fe2O3
MgO
CaO
MnO
Na2O
K 2O
SO3
L.O.I.
54.4
1.55
39.4
1.75
0.14
0.10
0.01
1.03
2.66
5.54
0.23
18.8
51.8
3.27
0.04
6.84
0.08
11.2
1.90
Table 2
% weight
Amorphous phase
GMK
GFE
GR12
GR10
GR8
GR6
GR4
91.5 0.2 91.0 0.5 91.7 0.5 90.9 0.5 89.2 0.6 91.6 0.5 91.4 0.4
Illite
5.6 0.1 4.6 0.2 3.7 0.2 4.2 0.2 6.2 0.3 2.1 0.1 2.5 0.1
Quartz
2.3 0.1 2.4 0.2 2.6 0.2 2.5 0.1 2.3 0.1 3.1 0.2 2.4 0.1
Hematite
absent
1.8 0.1 0.7 0.1 0.9 0.1 1.1 0.1 1.3 0.1 1.5 0.1
Goethite
absent
absent
Anatase
0.6 0.1 0.2 0.1 0.3 0.1 0.2 0.1 0.2 0.1 0.7 0.1 0.4 0.1
1.0 0.1 1.3 0.1 1.0 0.1 1.2 0.1 1.8 0.1
Table 3
Spectrum 1
A1
Spectrum 2
A2
Na
Al
Si
Fe
O
6.2
10.3
21.9
1.6
60
6.1
13.2
19.9
0.4
60.4
8.6
7.9
22.3
0.4
60.8
10.4
11.2
18.2
1
59.2
Na2O
Al2O3
SiO2
Fe2O3
9.1
24.6
64.8
1.5
9.0
32.2
57.3
1.5
13.0
19.7
65.7
1.6
15.6
27.7
53.0
3.6
Si/Al
Si/Na
2.1
3.5
1.5
3.3
2.8
2.6
1.6
1.7
SiO2/Al2O3
SiO2/Na2O
2.6
7.1
1.8
6.4
3.3
5.1
1.9
3.4
Table 4
GMK
GFE
GR12
GR10
GR4
47.2 0.5
44.0 0.5
48.9 0.5
44.3 0.5
44.5 0.5
49.5 0.5
48.2 0.5
50.6 0.5
48.2 0.5
49.3 0.5
1.05 0.03
1.09 0.03
1.03 0.03
1.09 0.03
1.11 0.03
2.08 0.05
2.11 0.05
2.09 0.05
2.09 0.05
2.19 0.05
Table 5
Samples
Leaching of Na (ppm)
GMK
102
GFE
132
GR12
114
GR10
98
GR8
66
GR6
95
GR4
70
Fig. 1
Fig. 2.
Fig. 3
10
Fig. 4
11
(a)
(b)
Fig. 5
12
Fig. 6
13
Fig. 7
14
Highlights
New geopolymer formulations were designed by mixtures of metakaolin and red
mud.
Na leaching values show that the chemical stability of the geopolymers is good.
15