1 s2.0 S0950061821007534 Main
1 s2.0 S0950061821007534 Main
1 s2.0 S0950061821007534 Main
h i g h l i g h t s g r a p h i c a l a b s t r a c t
a r t i c l e i n f o a b s t r a c t
Article history: Red mud (RM) and municipal solid waste incineration bottom ash (MSWIBA) are continually generated in
Received 18 November 2020 large amounts all over the world. In this study, RM and MSWIBA were made into pellets by a disc pel-
Received in revised form 4 March 2021 letizer and then transformed to lightweight aggregate ceramsites by high-temperature sintering. The sin-
Accepted 5 March 2021
tering mechanism and optimal production process were revealed from evaluation of the performance of
Available online 23 March 2021
ceramsites produced under different production processes. The results showed that as the proportion of
RM increased, the required sintering temperature increased at least by 7.34%, while the apparent density,
Keywords:
bulk density, strength, porosity, proportion of macropores and pH of ceramsites reduced up to 9.46%,
Red mud
Municipal solid waste incineration bottom
9.45%, 68.56%, 77.36%, 93.75% and 3.43%, respectively. On the other hand, with the increase in sintering
ash temperature, the apparent density and bulk density of ceramsites made from calcium-rich red mud
Lightweight aggregate (CRM) generally increased while the water absorption and pH generally decreased. The strength, how-
Pelleting ever, first increased up to 27.11 MPa and then decreased to 17.48 MPa. When the ratio of MSWIBA
Sintering and CRM was 1:1 and the sintering temperature was 1070 °C, the ceramsites produced could achieve
the best performance with a bulk density of 1046.73 Kg/m3, an apparent density of 1783.44 Kg/m3, a par-
ticle strength of 27.11 MPa, a 1-hour water absorption rate of 0.8%, and a pH of 8.9. The ceramsites, for
⇑ Corresponding authors at: State Key Laboratory of Geomechanics and Geotechnical Engineering, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences,
Wuhan 430071, China (J.-s. Li and Q. Xue).
E-mail addresses: [email protected] (J.-s. Li), [email protected] (Q. Xue), [email protected] (Q. Sun).
https://doi.org/10.1016/j.conbuildmat.2021.122993
0950-0618/Ó 2021 Elsevier Ltd. All rights reserved.
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
use as lightweight aggregates, can be a promising construction material in particular counting the ben-
efits of waste recycling.
Ó 2021 Elsevier Ltd. All rights reserved.
suppliers: iron-rich RM (IRM) (Fig. 1(a)) was obtained from Zibo process in a muffle furnace was divided into two stages, namely
City, Shandong Province, China, and CRM (Fig. 1(b)) was obtained pre-calcination and calcination. The pellets were heated up at
from Guiyang City, Guizhou Province, China. The variation between 5 °C / min, maintaining the temperature at 400 °C and 1000 °C
the two RMs is likely due to the different chemical compositions of for 30 min and 40 min, respectively, and then immediately
bauxite used as precursors for the production of alumina. MSWIBA removed for cooling at room temperature. After completion of
(Fig. 1(c)) was sourced from Wuhan, Hubei Province, China. the sintering process, the pellets were taken out from the muffle
The chemical compositions of these materials were determined furnace for cooling at room temperature.
by X-ray fluorescence spectrometer (XRF) (ARL PERFORM’X).
Table 1 lists the main chemical compositions of these materials. 2.3. Test methods
Fe2O3 content was the highest in IRM, followed by Al2O3. The con-
tents of CaO, Al2O3, SiO2 and Fe2O3 in CRM were almost the same. The bulk density, apparent density, porosity, 1-hour water
The content of SiO2 in MSWIBA was the highest, followed by CaO. absorption and particle shape coefficient of lightweight aggregates
Since the raw materials did not contain sulfur as confirmed by the were tested in accordance with Chinese standards (GB/T 17431.1-
XRF test results and the calcination temperature exceeded 1010 °C, 2010 and GB/T 17431.2-2010).
the properties of ceramsites were free from influence of sulfide,
sulfate and organic matters. The mineral compositions of the raw 2.3.1. Strength
materials (as shown in Fig. 2) were determined by X-ray diffrac- In order to measure the strength of ceramsites accurately, a
tometer (XRD) (Bruke D8 Advance), scanning from 10° to 80° at a microcomputer controlled strength testing machine (YA91122S)
rate of 5° /min [24]. The main mineral components of IRM were was used [25,26]. 20 samples were selected from each group.
bohmite (AlO(OH)), goethite (Fe + 3O(OH)), hematite (Fe2O3), The calculation of strength (MPa) was according to the following
quartz (SiO2) and rutile (TiO2). The RM is red in color because of equation:
the existence of hematite. The main mineral components of CRM
2:8Pc
were cronstedtite (Fe3FeSiO4(OH)5), katoite (Ca2.93Al1.97Si0.64O2.56 S¼
(OH)9.44), bomite (AlO(OH)), cancrinite (Na8(Si6Al6O24) pX 2
(H0.88(CO3)1.44)(H2O)2), calcite (Ca(CO3)), hematite (Fe2O3) and where Pc is the crushing load (N), and X is the distance between the
brookite (TiO2). The main mineral components of MSWIBA were upper and lower loading plates (mm).
quartz (SiO2) and calcite (Ca(CO3)).
2.3.2. Apparent density and bulk density
About 500 g of the ceramsites were first dried and then soaked
2.2. Preparation of aggregate ceramsites in water for 1 h. After filtering, the surface moisture on the ceram-
sites was wiped off with a wet towel. Finally, the ceramsites were
Fig. 3 outlines the preparation process of ceramsites. The RM put into a 1000 ml measuring cylinder containing 500 ml of water
and MSWIBA were placed in an oven with temperature of 105 °C and the total volume was recorded. Each test was repeated twice
for drying to constant weight. After drying, they were crushed in and the deviation of individual results from the average value
a crusher respectively for at least 1 min and sieved through a was less than 2%. The apparent density was calculated using the
0.25 mm screen. As the original RM particles were fine and the following formula:
crusher could easily break up the weakly agglomerated particles,
m 1000
the RM particles, after crushing were much smaller than the qap ¼
MSWIBA particles. The two types of powder were proportioned Vt 500
according to Table 2 and mixed uniformly. Batch 1 samples were where qap is the apparent density (kg/m3), m is the mass of the
in comparison between IRM and CRM, while batch 2 samples, all dried sample (g), Vt is the total volume of the sample after soaking
made with CRM, were for comparing different calcination temper- for 1 h and water in the cylinder (ml).
atures. The powder mixtures were placed into a disc pelletizer The ceramsites were evenly poured into a 1 L capacity cylinder
(model 1000) for granulation. The fresh pellets prepared by this up to 50 mm above the mouth. The excess ceramsites at the top of
granulation method could meet the requirements of the gradua- the cylinder were scraped away with a ruler, and any surface
tion in standard (GB/T 17461.1–2010). During the granulation pro- depressions were filled with fine particles. The bulk density was
cess, water was sprayed to keep the moisture content at about 25% calculated using the following formula:
and the granulation process was completed within 15 min. The
ðmt mv Þ 1000
green pellets were kept at room temperature for 24 h, and then qbu ¼
dried to constant weight at 105 °C. The subsequent sintering V
Fig. 1. Appearance of raw materials: IRM (a), CRM (b) and MSWIBA (c).
3
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
Table 1
Chemical compositions of IRM, CRM and MSWIBA.
Components (%) CaO Al2O3 SiO2 Fe2O3 Na2O TiO2 K2O MgO SO3 P2O5 Other
IRM 0.84 18.94 7.55 58.51 5.40 6.93 0.06 0.04 – 0.32 1.41
CRM 21.99 20.19 19.24 19.19 7.88 5.40 1.95 1.50 1.42 – 1.24
MSWIBA 13.84 6.49 60.81 3.08 8.40 0.31 1.65 1.54 0.95 1.03 1.90
qbu
t ¼ ð1 Þ 100
qap
2.3.4. pH value
The pH of raw materials, fresh pellets and sintered ceramsites
was measured by a pH meter (ph-3c). Each sample of 30 g in
weight was put into 75 ml deionized water, and stirred for 24 h
before measurement. The average value of triplicate readings was
taken in each test [27].
2.3.3. Void ratio where xa is the 1-hour water absorption rate of the ceramsites (%),
The void ratio can be used to determine the pore characteristic m0 is the mass of the ceramsites soaked for 1 h (g), m1 is the mass of
of ceramsite in the natural accumulation state. It was calculated the ceramsites after drying (g).
4
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
Table 2 and proportions of RMs have great visual differences. The ceram-
Summary of material proportions and sintering temperatures. sites made from IRM were reddish, and the ceramsites made from
Batch 1 Batch 2 CRM were yellowish. The color darkened as the content of RM
Sintering at 1010 °C CRM 50% – MSWIBA 50% increased. The average particle form factor coefficient of the cer-
Sample Material proportion Sample Sintering temperature/ amsite was 1.52, which met the minimum requirements of the Chi-
label label °C nese standard GB/T 17461.1-2010. The main factors that affected
A1 IRM 30% – MSWIBA B1 1030 the particle form factor coefficient included the impact on fresh
70% pellets during the pelletizing process and deformation of contact
A2 IRM 40% – MSWIBA B2 1050 surface due to gravity during the aging period. In this experiment,
60%
crack would not appear on ceramsites made with different formu-
A3 IRM 50% – MSWIBA B3 1070
50% lations and temperature treatment. However, the surface of the
A4 IRM 60% – MSWIBA B4 1090 ceramsites became powdery as the content of RM increased, indi-
40% cating the need for raising the calcination temperature.
A5 IRM 70% – MSWIBA
30%
A6 CRM 30% – MSWIBA
70%
3.2. Ceramsite physical characteristics
A7 CRM 40% – MSWIBA
60%
A8 CRM 50% – MSWIBA 3.2.1. Apparent density, bulk density and void ratio
50% In order to accurately explore the variation regularities of the
A9 CRM 60% – MSWIBA densities which were affected by the porosities appeared in differ-
30%
A10 CRM 70% – MSWIBA
ent ceramsites, ceramsites with voidage varied from 41% to 43%
30% were selected as typical samples. The apparent density and bulk
density of ceramsites are shown in Fig. 5. It can be found that
the apparent density and bulk density of ceramsites, except A1,
2.3.6. Particle form factor were inversely proportional to the content of the RM. This result
The test for particle form factor was conducted according to was contrary to that found by Molineux et al. (2016) [10]. This
GBT17431.2-2010. The grain shape coefficient of a single ceramsite can be attributed to the control in calcination temperature. When
was calculated according to the following formula: the amount of RM increased but the calcination temperature was
not raised, mass transfer in the ceramsite could not evolve com-
0 Dmax pletely, leading to weak migration of particles in ceramsite and
Ke ¼
Dmin decrease in density. However, in the research of Molineux et al.
0 (2016) [10], the sintering temperature was high enough to allow
where K e represents the single particle type coefficient of the cer-
normal mass transfer in ceramsites with the highest RM contents
amsite, Dmax is the largest dimension of the ceramsite (mm), Dmin
selected. Besides, when the calcination temperature was increased,
is the smallest dimension of the ceramsite intermediate interface
the density of ceramsites showed an overall upward trend (Fig. 5
(mm).
(b)).
The average particle form factor of ceramsite was calculated
In addition, it can be seen that the densities of ceramsites made
according to the following formula:
from IRM were higher than that made from CRM. The high bulk
Pn 0 density (1206.09 Kg/m3-1240.74 Kg/m3) of the ceramsites made
i¼1 K e;i
Ke ¼ from IRM actually failed to meet the Chinese standard GB/T
n
17461.1-2010 requirements. In this regard, CRM was more suitable
0
where K e is the average size coefficient of ceramsite, K e;i is the size for producing lightweight aggregates than IRM.
coefficient of a certain ceramsite particle, and n is the number of
ceramsite particles tested.
3.2.2. Aggregate strength
2.3.7. Pore distribution characteristics With the same proportion of the RM, the strengths of most cer-
After splitting a ceramsite particle, the cross-section of the cer- amsites made from IRM (A1-A5) were higher than those of ceram-
amsite was photographed with an electron microscope (RH-2000) sites made from CRM (A6-A10). However, the strength of A1 was
and binarized (Li et al., 2020) [28]. The pore distribution and pore lower than A6 due to better sintering of A1, resulting in the forma-
sizes in the balance section were analyzed and measured. Tripli- tion of more macropore. It can be seen that the type and amount of
cates were selected for observation and measurement. The propor- RM were the main factors affecting the strength of ceramsites. The
tion of pores in the cross section was calculated by the following crushing strength of A1-A5 ceramsites ranged from 8 MPa to
formula: 21 MPa. The strengths of A2 ceramsites was the highest
M (21.02 MPa), and the strength of A5 ceramsite was the lowest
P¼ (8.37 MPa). The crushing strengths of A6-A10 ceramsites ranged
C
from 4 MPa to 22 MPa. The strength of A6 ceramsites was the high-
where P is the porosity (%), M is the total area of pores (lm2), and C est with a value of 21.59 MPa (Fig. 6(a)) and the strength decreased
is the cross-sectional area (lm2). with increase in CRM content. Fig. 6(b) shows the effects of differ-
ent sintering temperatures on the strength of A8 mix. It could be
3. Results found that the strength changed significantly with the tempera-
ture. As the temperature increased, the strength of the ceramsites
3.1. Visual appearance of ceramsites increased first but then decreased beyond 1070 °C due to the
liquid-phase mass transfer and the formed macropores. The result
The ceramsites made in the experiment are shown in Fig. 4. It was in line with that found by Wang et al. (2020) [29]. The highest
can be observed that the ceramsites made from different types strength obtained was 27.11 MPa.
5
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
Fig. 4. Photos of sintered ceramsites made from MSWIBA and RM with different proportions.
Fig. 5. Apparent density and bulk density of ceramsites produced from different mixtures (a) and same constituents but different calcination temperatures (b). The lines 1, 2,
3 and 4 marked in the figure indicate bulk densities of 1200 kg/m3, 1100 kg/m3, 1000 kg/m3 and 900 kg/m3 respectively, representing different grades of lightweight
aggregates.
Fig. 6. Crushing strength of ceramsites with different mixes of materials (a) and with equal amount of SRM and MSWIBA but sintered at different calcination temperatures
(b).
3.2.3. pH value may be due to the existence of the stable alkaline substances in
Fig. 7 shows the pH values of raw materials and ceramsites pro- the RM after mineralization and precipitation [30,31], and part of
duced. The result showed that the type and content of RM as pre- its alkali may react with SiO2 in MSWIBA, triggering the ionization
cursors had little effect on the pH of ceramsites product, and the of chemically bound alkali and increasing the pH. In addition,
pH values varied between 9.4 and 9.9. As the calcination tempera- MSWIBA also contained CaO, which may react with sodium-
ture increased, the pH dropped slightly but were higher than 8. silicon slag (Na2O Al2O3 nSiO2mH2O) in the RM to form calcium
An interesting phenomenon is that the pH of a mixture of RM silicate phase, releasing Na2O into the solution and changing the
and MSWIBA was higher than their respective pH values. This equilibrium of soluble alkaline substances in the RM [2], leading
6
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
Fig. 7. Variation of pH of ceramsites produced from different constituents (a) and same constituents but different firing temperatures (b).
to an increase in pH. This process is similar to the basic principle of These disconnected macropores in the RM-based ceramsites have
alkali recovery from RM. the function of heart preservation and sound insulation.
The MSWIBA is a general solid waste but not a hazardous waste
because of the limited content of heavy metals. Sintering process 3.2.5. Water absorption
can not only neutralize pH value of aggregate, but also immobilize Fig. 10 shows the 1-hour water absorption rate of ceramsites. It
most heavy metals. A large number of researches showed that can be seen that at a constant calcination temperature, the water
heavy metals could be confined within the crystalline structure absorption rate was proportional to the amount of RM added. Actu-
of produced phase during the heat treatment [10,21]. Hwang ally, the water absorption rates of A9 and A10 ceramsites were so
et al. (2012) examined the stabilization behavior of heavy metals high that expulsion of air in bubbles occurred in the first few min-
in the sintered aggregate made by materials rich in heavy metals. utes after submerging them into water. In addition, it can be noted
The results showed that even if the total contents of heavy metals that the water absorption rates of ceramsites made from CRM were
were increased, the leaching concentrations of all elements in sin- higher compared to IRM-based ceramsites. With increase in the
tered aggregate were well under the limit [32]. In addition, the calcination temperature, the water absorption rate of ceramsites
study by Chang et al. (2007) obtained a similar conclusion [33]. decreased, reaching the lowest rate of 0.36% at 1090 °C.
These indicated that the heat treatment is an effective method to It is interesting to note that the cross-sectional porosity and
immobilize most heavy metals in waste materials, thus the aggre- water absorption of ceramsites varied in the opposite trends when
gates prepared in this study are environmentally safe. the sintering temperature or the proportion of RM in the mixes
changed. There was no connection between the change in cross-
sectional porosity and change in water absorption rate.
3.2.4. Pore distribution characteristics Although the cross-sectional porosity of IRM-based ceramsites
The image of a cross-section of the ceramsite captured by a was relatively high, its water absorption rate was relatively low,
microscope is shown in Fig. 8(a). After the binarization process, indicating that the degree of sintering of IRM-based ceramsites
the positions and sizes of the pores were marked with blue lines. was more complete at the same calcination temperature, and more
It can be found that after the same thermal treatment, macropores glassy phases were generated, which hindered the entry of water.
reduced as the proportion of RM increased. However, as the calci-
nation temperature increased, the proportion of macropores 3.3. XRD analysis
increased significantly. Fig. 8(b) shows the interfacial porosity after
the binarization treatment. The interfacial porosities of the two The XRD test results of some ceramsites selected from different
ceramsites made from different RMs decreased as the proportion samples are shown in Fig. 11. It can be observed that there was no
of RM increased, and the porosity of IRM-based ceramsites was phase transformation of hematite in the RM at a sintering temper-
higher than that of CRM-based ceramsites. When the calcination ature of 1010 °C, which was consistent with the finding of Sglavo
temperature increased, the porosity increased due to the decompo- et al. [34]. On the other hand, the peaks corresponding to calcite
sition of some minerals and the migration of particles. in the MSWIBA disappeared as it decomposed into CO2 and CaO
Fig. 9 shows quantitatively the pore size of the cross-section of at 700 – 800 °C [35].
ceramsites produced in this study. The pore size was divided into 6 The XRD patterns of the IRM-based ceramsites are shown in
ranges, namely, 6.5–50 lm, 50–100 lm, 100–150 lm, 150– Fig. 11(a). The main crystal phases were hematite (Fe2O3), quartz
200 lm, 200–250 lm and >250 lm. The result showed that with (SiO2), wollastonite (CaSiO3) and a-Al2O3. The bohmite in IRM
the same thermal treatment, the proportion of pores less than was mainly derived from the chemical digestion process of bauxite,
150 lm in the ceramsite increased with increase in RM content, and OH– was released from this mineral under high temperature
while an opposite trend occurred with macropores larger than environment. The bohmite was then transformed into relatively
200 lm. The two kinds of ceramsites made from different RMs also stable a-Al2O3 [36]. Hematite was dehydroxylated into FeO during
showed an opposite trend. With increase in calcination tempera- the heating process, and was then oxidized into Fe2O3 as the tem-
ture, the boundary line of pore size change was around 150 lm. perature rose. CaO reacted with some SiO2 to form wollastonite
When the ratio of the raw materials was 1:1, the proportion of and the remaining SiO2 could still be detected. The peaks of AlO
macropores in the ceramsites calcinated at 1090 °C was the most. (OH), Fe + 3O(OH) and Ca(CO3) disappeared, replacing by wollas-
7
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
Fig. 8. Cross-sectional images of different ceramsites with a magnification of 40 times (a), and porosity values of ceramsites produced from different proportions of materials
and firing temperatures (b).
tonite and a-Al2O3. With increase in MSWIBA content, the height The hematite content in CRM was relatively small, accounting
of quartz and wollastonite peaks in ceramsites seemed to be for its light yellow color. The XRD patterns of the CRM-based cer-
higher, while the hematite peaks lower. amsites are shown in Fig. 11(b). The main crystal components were
8
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
Fig. 11. XRD patterns of ceramsites produced from different proportions of materials.
9
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
4. Discussion glassy phase produced by higher sintering degree can wrap the
gas produced by calcite decomposition, resulting in the highest
4.1. General sintering mechanism porosity of A1 (Fig. 8 (b)) and the decrease of density (Fig. 5 (a)).
At the same time, these macropores can offset the strength
The intensive energy applied in the crushing of RMs and improvement brought by the densification of ceramsite, which
MSWIBA is transformed to the form of surface energy stored in even lead to the decrease of strength.
the powder. The smaller the broken particles are, the greater the Fig. 14 shows the magnified cross section (350) of different cer-
surface energy stored in them is, which keeps it in a state of ther- amsites under an electron microscope. It can be found that with
modynamically non-equilibrium [37]. Therefore, the fresh pellets higher proportion of RM in the mixes, the glassy phases in the cer-
tended to reduce the surface area by contacting each other from amsites decreased while the powder increased. Besides, compared
mass transfer and migration during calcination, resulting in the with the IRM, the CRM contained more Al2O3 and less flux, which
combination of particles (G. C. Kuczynski’s sintering theory). The generated more ceramsite powder under the same thermal treat-
specific process of particle combination was determined by the ment condition. Therefore, the internal particles of the CRM-based
ceramic ternary phase diagram (Fig. 13). It can be found that all ceramsites were not tightly bonded, resulting in a reduction in the
the mixture of ceramsites contained a large amount of flux ceramsites’s density and strength. Moreover, although CaO in the
(Fe2O3, Na2O, K2O, CaO and MgO). The existence of alkali metal oxi- flux lowered the temperature required for sintering, the Ca2+
des in RM present as katoite, cancrinite, hematite and calcite can exceeded the level required for producing electrical neutrality in
be sintered with Al2O3 and SiO2 at a lower temperature to form the silicate networks [13], resulting in a decrease in the strength of
glassy phase [38,39,40]. The mass transfer occurred at an appropri- the ceramsites.
ate sintering temperature, mainly consisted of viscous flow mass
transfer and dissolution-precipitation mass transfer of the liquid 4.3. Effect of calcination temperature on properties of ceramsites
phase. The whole row of atoms in ceramsite moved along the
stress direction concurrently. The RM particles and some of the fine It can be seen from the XRD patterns that the chemical compo-
particles of MSWIBA were transported into the larger MSWIBA par- sitions of the RM and MSWIBA are very complex. High-strength
ticles under the capillary action crystallized inside. and thermal stable wollastonite produced from some chemical
reaction during the sintering process contributed to a higher
strength of the ceramsites. At the temperature of 1090 °C, the
4.2. Influence of the proportion of RM on the properties of ceramsites
strength of the ceramsites reduced due to the sudden decrease in
wollastonite content. In addition, the CO2 gas generated by the
Si and Al are the main components of the ceramsites skeleton,
decomposition of calcite at high temperature was trapped in the
which provide strength for the ceramsite. The molar ratio of
glassy phases, forming many discrete tiny pores. As diffusion mass
Al2O3/ SiO2 in the products affected the calcination temperature
transfer under sintering temperatures applied in this study was
and range [41]. As the ratio increased, the calcination temperature
almost impossible, the pores could not be filled.
and range also increased. Hence, as the proportion of RM increased,
As the sintering temperature increased, the effect of the liquid
the mass transfer of the internal liquid phase reduced due to the
phase mass transfer became more obvious. More solid particles
decrease of glassy phase in the ceramsite (powder even appeared).
arranged themselves under the transformation to achieve a higher
As a result, it was difficult to hold the gas produced from the
density and to generate crystallization, resulting in a stronger cer-
decomposition of calcite at high temperature, leading to a decrease
amsites (Fig. 6(b)). At the same time, more pores combined together
in porosity (Fig. 8 (b) and Fig. 9). The results also revealed that
forming more macropores. The close packing of particles enhanced
although the porosity of the IRM-based ceramsites was more sig-
the strength of the ceramsites, while the larger pores offset part of
nificant than the CRM-based ceramsites, the density of the IRM-
the additional strength. Therefore, the crushing strength increased
based ceramsites was higher. But in samples A1-A5, the content
first and then decreased (Fig. 6) while the porosity increased with
of RM in A1 is the lowest, thus the sintering degree of A1 is the
the calcination temperature (Fig. 8(a) and Fig. 9). Furthermore, with
highest at the same temperature. However, the density and crush-
the increase in sintering temperature, a more dense glass phase was
ing strength of A1 are lower. This is because a large amount of
formed on the surface and interior part of the ceramsites filling the
capillary channels and blocking the entry of water from outside.
The water absorption rate of the ceramsites dropped to almost zero
after sintering at the high temperature of 1090 °C.
The microstructure presented in Fig. 14 shows that with rise in
calcination temperature, a large number of glassy phase materials
appeared inside the ceramsites and the particles were more closely
packed. During the sintering process, the alkaline materials reacted
to form new mineral components. At the same time, the generated
liquid phases encapsulated the solid particles, reducing the release
of OH– into the water, and lowering the pH value of the ceramsites
(Fig. 7). As the proportion of RM increased, the chemical reaction
became incomplete without a rise in the calcination temperature
and less glassy phases were generated to encapsulate the particles,
resulting in higher pH in the ceramsites. After increasing the sinter-
ing temperature, the positive condition led to the drop of pH value.
4.4. Outlook
Fig. 14. Images of different ceramsites after magnification by 350 times (blue arrows indicate unfired powder particles, red arrows indicate liquid phase formed during
sintering), and unfired powder particles were hardly observed in A1 and B4. (For interpretation of the references to color in this figure legend, the reader is referred to the web
version of this article.)
standard. That can be attributed to relatively higher density of the in a muffle furnace to produce porous ceramsites which could be
IRM due to its high iron content. Besides, the IRM contains more used as lightweight aggregates. Typical properties of lightweight
flux and generates stronger mass transfer at the same thermal sys- aggregates were tested for evaluating the performance of ceram-
tem, leading to a high density of IRM-based ceramsites. Unlike sites made from different mixes of materials and sintering temper-
CRM, no pores are produced during the firing of IRM to help reduc- atures. The optimal production process and the sintering
ing its density. The properties of ceramsites made from equal pro- mechanism were explored. Based on the experimental results,
portions of the two materials are in compliance with the the following conclusions can be drawn:
requirements of GB/T 17461.1-2010. The method adopted by most
scholars for producing RM-based ceramsites is relatively compli- (1) Under the same thermal process, with increase in proportion
cated. The cumbersome process and high sintering temperature of RM (A1-A10 all sintered at 1010 °C) in the raw materials,
undoubtedly increase the production cost, but this method is still the content of glassy phase in ceramsites reduced, the den-
simpler compared with the traditional method for making light- sity and strength of ceramsites also decreased, while the
weight aggregates. Besides, the RM-based ceramsites absorbed lit- water absorption rate increased.
tle water as the pores are disconnected, causing little effect on the (2) With increase in the calcination temperature from 1010 °C
mix design of concrete, and the phenomenon of pump blocking to 1090 °C, the porosity of CRM-based ceramsites increased
rarely occur. Furthermore, no natural minerals are consumed and but the water absorption rate decreased. And the density
waste materials such as shale, clay, etc. are recycled with this basically presented an upward trend. The strength firstly
method, helping to preserve the environment [42]. increased and then decreased, and the pH slightly decreased.
Nevertheless, there are also some drawbacks with the RM- (3) Most of the pores in the ceramsites were formed during the
based ceramsites: decomposition of calcite at high temperature to produce CO2
gas, which was tapped in the glassy phases. At an appropri-
1. The ceramsites may slightly stick to each other due to excessive ate temperature, the pores on the surface of the ceramsites
glassy phases formed on the surface at calcination temperature were mostly sealed due to the limited glassy phase content,
of 1090 °C. making it almost waterproof.
2. The optimal sintering temperature falls in a narrow range due (4) CRM was more suitable than IRM to be a raw material for
to the low aluminum to silicon ratio. Accurate control of furnace producing RM-based lightweight aggregate ceramsites. Tak-
temperature is necessary. ing into account the cost of material preparation, the utiliza-
3. The weight of the ceramsites is still not light enough and some tion rate of RM, and various properties of ceramsites, the
additives may have to be used to further reduce the weight. ratio of CRM:MSWIBA was selected to be 1:1 by weight
and the corresponding sintering temperature should be
5. Conclusion 1070 °C to achieve the best performance.
(5) Hematite contained in RM was not affected by the high tem-
In this study, RM and MSWIBA powders were mixed and made perature sintering. As the proportion of RM decreased in the
into fresh pellets by using a disc pelletizer, and were then sintered raw materials, hematite content in the ceramsites decreased
11
Y. Sun, Jiang-shan Li, Z. Chen et al. Construction and Building Materials 287 (2021) 122993
and the quartz content increased. With rise in the sintering [16] Y. Jiang, B. Xi, X. Li, L. Zhang, Z. Wei, Effect of water-extraction on
characteristics of melting and solidification of fly ash from municipal solid
temperature, more glassy phases were formed in the ceram-
waste incinerator, J. Hazard. Mater. 161 (2–3) (2009) 871–877.
sites and the XRD peaks became less sharp. [17] M.U. Hossain, L. Wang, L. Chen, D.C.W. Tsang, S.T. Ng, C.S. Poon, V.
Mechtcherine, Evaluating the environmental impacts of stabilization and
CRediT authorship contribution statement solidification technologies for managing hazardous wastes through life cycle
assessment: a case study of Hong Kong, Environ. Int. 145 (2020) 106139.
[18] A. Vaitkus, J. Gražulytė, O. Šernas, V. Vorobjovas, R. Kleizienė, An algorithm for
Yuehui Sun: Methodology, Validation, Formal analysis, the use of MSWI bottom ash as a building material in road pavement
Investigation, Writing - original draft, Visualization. Jiang-shan Li: structural layers, Constr. Build. Mater. 212 (2019) 456–466.
[19] Z. Yang, S. Tian, L. Liu, X. Wang, Z. Zhang, Recycling ground MSWI bottom ash
Conceptualization, Resources, Writing - review & editing, in cement composites: long-term environmental impacts, Waste Manage. 78
Supervision, Project administration, Funding acquisition. Zhen Chen: (2018) 841–848.
Writing - review & editing. Qiang Xue: Resources, Supervision, Project [20] M. Behera, S.K. Bhattacharyya, A.K. Minocha, R. Deoliya, S. Maiti, Recycled
aggregate from C&D waste & its use in concrete – a breakthrough towards
administration. Qi Sun: Resources, Funding acquisition. Yifan Zhou: sustainability in construction sector: a review, Constr. Build. Mater. 68 (2014)
Writing - review & editing. Xin Chen: Methodology. Lei Liu: 501–516.
Resources, Supervision. Chi Sun Poon: Writing - review & editing. [21] P. Ren, T.-C. Ling, K.H. Mo, Recent advances in artificial aggregate production, J.
Cleaner Prod. 291 (2021) 125215.
[22] Y. Jiang, T.-C. Ling, Production of artificial aggregates from steel-making slag:
Declaration of Competing Interest influences of accelerated carbonation during granulation and/or post-curing, J.
CO2 Util. 36 (2020) 135–144.
[23] Y. Jiang, T.-C. Ling, M. Shi, Strength enhancement of artificial aggregate
The authors declare that they have no known competing finan-
prepared with waste concrete powder and its impact on concrete properties, J.
cial interests or personal relationships that could have appeared Cleaner Prod. 257 (2020) 120515.
to influence the work reported in this paper. [24] X. Chen, J. Li, Q. Xue, X. Huang, L. Liu, C.S. Poon, Sludge biochar as a green
additive in cement-based composites: mechanical properties and hydration
kinetics, Constr. Build. Mater. 262 (2020) 120723.
Acknowledgement [25] C.R. Cheeseman, G.S. Virdi, Properties and microstructure of lightweight
aggregate produced from sintered sewage sludge ash, Resour. Conserv. Recycl.
This study was financially supported by International Partner- 45 (1) (2005) 18–30.
[26] P. Tang, D. Xuan, H.W. Cheng, C.S. Poon, D.C.W. Tsang, Use of CO2 curing to
ship Program of Chinese Academy of Sciences (Grant No. enhance the properties of cold bonded lightweight aggregates (CBLAs)
115242KYSB20200002), National Science Fund for Distinguished produced with concrete slurry waste (CSW) and fine incineration bottom
Young Scholars (Grant No. 51625903), National Natural Science ash (IBA), J. Hazard. Mater. 381 (2020) 120951.
[27] H. Chen, G. Wang, Y. Xu, Z. Chen, F. Yin, Green process for supercritical water
Foundation of China/Hong Kong Research Grants Council, University oxidation of sewage sludge with red mud as CO2 absorbent, J. Environ. Chem.
Grants Committee Joint Research Scheme (Grant No. 51861165104), Eng. 4 (3) (2016) 3065–3074.
Liaoning province Department of Education fund (Grant No. [28] X. Li, P. Wang, J. Qin, Y. Liu, Y. Qu, J. Liu, R. Cao, Y. Zhang, Mechanical properties
of sintered ceramsite from iron ore tailings affected by two-region structure,
LJ2020JCL001), Special Fund for Basic Research on Science Instru- Constr. Build. Mater. 240 (2020) 117919.
ments of the National Science Foundation of China (Grant No. [29] W. Wang, K. Sun, H. Liu, Effects of different aluminum sources on
51827814) The CAS Pioneer Hundred Talents Program in China morphologies and properties of ceramic floor tiles from red mud, Constr.
Build. Mater. 241 (2020) 118119.
and Youth Innovation Promotion Association CAS (2017376).
[30] B.I. Whittington, B.L. Fletcher, C. Talbot, The effect of reaction conditions on the
composition of desilication product (DSP) formed under simulated Bayer
References conditions, Hydrometallurgy 49 (1–2) (1998) 1–22.
[31] M. Paradis, J. Duchesne, A. Lamontagne, D. Isabel, Long-term neutralisation
[1] L. Wang, L. Chen, B. Guo, D.C.W. Tsang, L. Huang, Y.S. Ok, V. Mechtcherine, Red potential of red mud bauxite with brine amendment for the neutralisation of
mud-enhanced magnesium phosphate cement for remediation of Pb and As acidic mine tailings, Appl. Geochem. 22 (11) (2007) 2326–2333.
contaminated soil, J. Hazard. Mater. 400 (2020) 123317. [32] C.-L. Hwang, L.-A.-T. Bui, K.-L. Lin, C.-T. Lo, Manufacture and performance of
[2] R. Zhang, S. Zheng, S. Ma, Y. Zhang, Recovery of alumina and alkali in Bayer red lightweight aggregate from municipal solid waste incinerator fly ash and
mud by the formation of andradite-grossular hydrogarnet in hydrothermal reservoir sediment for self-consolidating lightweight concrete, Cem. Concr.
process, J. Hazard. Mater. 189 (3) (2011) 827–835. Compos. 34 (10) (2012) 1159–1166.
[3] F. Kehagia, A successful pilot project demonstrating the re-use potential of [33] F.C. Chang, S.L. Lo, M.Y. Lee, C.H. Ko, J.D. Lin, S.C. Huang, C.F. Wang, Leachability
bauxite residue in embankment construction, Resour. Conserv. Recycl. 54 (7) of metals from sludge-based artificial lightweight aggregate, J. Hazard. Mater.
(2010) 417–421. 146 (1–2) (2007) 98–105.
[4] M.A. Khairul, J. Zanganeh, B. Moghtaderi, The composition, recycling and [34] V.M. Sglavo, R. Campostrini, S. Maurina, G. Carturan, M. Monagheddu, G.
utilisation of Bayer red mud, Resour. Conserv. Recycl. 141 (2019) 483–498. Budroni, G. Cocco, Bauxite ‘red mud’in the ceramic industry. Part 1: thermal
[5] H. Sutar, S.C. Mishra, S.K. Sahoo, H.S. Maharana, Progress of red mud behaviour, J. Eur. Ceram. Soc. 20 (3) (2000) 235–244.
utilization: an overview, Chem. Sci. Int. J. (2014) 255–279. [35] K.S.P. Karunadasa, C.H. Manoratne, H.M.T.G.A. Pitawala, R.M.G. Rajapakse,
[6] C. Klauber, M. Gräfe, G. Power, Bauxite residue issues: II. options for residue Thermal decomposition of calcium carbonate (calcite polymorph) as examined
utilization, Hydrometallurgy 108 (1–2) (2011) 11–32. by in-situ high-temperature X-ray powder diffraction, J. Phys. Chem. Solids
[7] E. Mukiza, L. Zhang, X. Liu, N. Zhang, Utilization of red mud in road base and 134 (2019) 21–28.
subgrade materials: a review, Resour. Conserv. Recycl. 141 (2019) 187–199. [36] D. Pei, Y. Li, D. Cang, In situ XRD study on sintering mechanism of SiO2-Al2O3-
[8] S. Wang, H. Jin, Y. Deng, Y. Xiao, Comprehensive utilization status of red mud in CaO-MgO ceramics from red mud, Mater. Lett. 240 (2019) 229–232.
China: a critical review, J. Cleaner Prod. (2020) 125136. [37] Z.Z. Fang, H. Wang, Densification and grain growth during sintering of
[9] H. Blank, Red mud from alumina plants as a possible source of synthetic nanosized particles, Int. Mater. Rev. 53 (6) (2008) 326–352.
aggregate, J. Test. Eval. 4 (5) (1976) 355–358. [38] X. Xu, J. Song, Y. Li, J. Wu, X. Liu, C. Zhang, The microstructure and properties of
[10] C.J. Molineux, D.J. Newport, B. Ayati, C. Wang, S.P. Connop, J.E. Green, Bauxite ceramic tiles from solid wastes of Bayer red muds, Constr. Build. Mater. 212
residue (red mud) as a pulverised fuel ash substitute in the manufacture of (2019) 266–274.
lightweight aggregate, J. Cleaner Prod. 112 (2016) 401–408. [39] T. Liu, J. Zhang, J. Wu, J. Liu, C. Li, T. Ning, Z. Luo, X. Zhou, Q. Yang, A. Lu, The
[11] W. Wang, W. Chen, H. Liu, Recycling of waste red mud for fabrication of utilization of electrical insulators waste and red mud for fabrication of
SiC/mullite composite porous ceramics, Ceram. Int. 45 (8) (2019) 9852–9857. partially vitrified ceramic materials with high porosity and high strength, J.
[12] D.V. Lopes, E. Durana, F.R. Cesconeto, P.V. Almeida, A.V. Kovalevsky, M.J. Quina, Cleaner Prod. 223 (2019) 790–800.
J.R. Frade, Direct processing of cellular ceramics from a single red mud [40] Z.-Y. Chang, K.-X. Jiao, X.-J. Ning, J.-L. Zhang, Novel approach to studying
precursor, Ceram. Int. 46 (10) (2020) 16700–16707. influences of Na2O and K2O additions on viscosity and thermodynamic
[13] J.L. Zou, G.R. Xu, G.B. Li, Ceramsite obtained from water and wastewater sludge properties of bf slags, Metallurgical Mater. Trans. B 50 (3) (2019) 1399–1406.
and its characteristics affected by Fe2O3, CaO, and MgO, J. Hazard. Mater. 165 [41] J. Partyka, M. Sitarz, M. Leśniak, K. Gasek, P. Jeleń, The effect of SiO2/Al2O3 ratio
(1–3) (2009) 995–1001. on the structure and microstructure of the glazes from SiO2–Al2O3–CaO–MgO–
[14] W.D. Kingery, H.K. Bowen, D.R. Uhlmann, Introduction to ceramics, John wiley Na2O–K2O system, Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 134
& sons1976. (2015) 621–630.
[15] L. Chen, Y.S. Wang, L. Wang, Y. Zhang, J. Li, L. Tong, Q. Hu, J.G. Dai, D.C.W. Tsang, [42] Y.M. Wie, K.G. Lee, K.H. Lee, Physicochemical effect of the aeration rate on
Stabilisation/solidification of municipal solid waste incineration fly ash by bloating characterizations of artificial lightweight aggregate, Constr. Build.
phosphate-enhanced calcium aluminate cement, J. Hazard. Mater. 124404 (2020). Mater. 256 (2020) 119444.
12