Construction and Building Materials 296 (2021) 123653

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Red ceramic waste as supplementary cementitious material:

Microstructure and mechanical properties
J. Hoppe Filho a,⇑, C.A.O. Pires a, O.D. Leite b, M.R. Garcez c, M.H.F. Medeiros d
a
Federal University of Western Bahia (UFOB) - Exact Sciences and Technologies Center - Rua Prof. José Seabra de Lemos, 316 – Recanto dos Pássaros, Barreiras, Bahia CEP:
47.808-021, Brazil
b
Federal Technological University of Paraná (UTFPR) - Department of Chemistry, Medianeira, Paran á CEP: 85.844-000, Brazil
c
Federal University of Rio Grande do Sul (UFRGS) - Interdisciplinary Department, Tramandaí, Rio Grande do Sul CEP: 95.590-000, Brazil
d
Federal University of Paraná (UFPR) - Department of Civil Construction – Polytechnical Center, Curitiba, Paraná CEP: 81.530-000, Brazil

h i g h l i g h t s

RCW composed of calcined low-grade kaolinitic clay showed potential to be used as SCM.

The RCW increased the amount of ettringite and stimulated the formation of monocarboaluminate in blended mortars.

The 30% Portland cement replacement by RCW slightly decreased the compressive strength of blended mortars.

The w/b ratio influences the compressive strength and apparent porosity of blended mortars.

The use of RCW can reduce CO2 emissions by reducing the cement demand per unit of compressive strength.

a r t i c l e i n f o a b s t r a c t

Article history: The red ceramic waste (RCW) powder has the potential to be used as supplementary cementitious mate-
Received 7 December 2020 rial (SCM) since ceramic bricks, essentially composed of calcined low-grade kaolinitic clay, presents
Received in revised form 5 April 2021 amorphous phases resulted from the sintering process. This paper investigates the effects of 30%
Accepted 12 May 2021
Portland cement replacement by RCW in the microstructure of cementitious pastes and mortars’ physi-
Available online 29 May 2021
cal–mechanical properties. The use of RCW changed the hydrated pastes’ microstructure by increasing
the amount of ettringite (AFt) and stimulating the formation of monocarboaluminate (AFm). The com-
Keywords:
pressive strength of the blended mortars reduced by 7 to 16% at 182 days. The apparent porosity for mor-
Red ceramic waste powder
Limestone calcined clay reaction
tars with w/b ratio of 0.35 was equivalent to that of reference over the 182 days. The gradual increase in
Pozzolanic activity the w/b ratio of blended mortars tends to maximize the decrease in compressive strength and increase
the apparent porosity. The use of RCW can reduce CO2 emissions by reducing the cement demand per
unit of compressive strength.
Published by Elsevier Ltd.

1. Introduction the Portland cement industry [4–6]. However, the unavailability of

these materials worldwide may limit the use in the near future
Carbon dioxide emissions associated with clinker production [7,8].
causes severe environmental damages. Decarbonation of limestone Among the different types of potential SCM, calcined clay is the
and burn fuel in the production line results in approximately 0.94– mineral addition with greater effectiveness and availability world-
1.00 ton CO2/ton clinker [1,2]. Thus, increasing the cement produc- wide [7,9–11]. Recent publications prove that highly kaolinitic
tion sustainability demands reducing the clinker/cement ratio [3], clays result in higher reactivity [10,12,13]. However, it is possible
which can be achieved using supplementary cementitious materi- to obtain pozzolanic reactivity from composite calcined clay with
als (SCM). The use of SCM as partial clinker replacement is one of 20% or more kaolinite content [10,12–14]. The use of calcined
the most efficient strategies to mitigate emissions associated with low-grade kaolin clay in the cement industry is quite promising
and encouraging [7,15,16], but it is still limited, mainly due to
⇑ Corresponding author. the reactivity behavior that cannot be generalized and the conse-
E-mail addresses: juarez.hoppe@ufob.edu.br (J. Hoppe Filho), claudio.pires@u- quent implications in the mechanical properties [7,9,17]. The
fob.edu.br (C.A.O. Pires), oldair.leite@gmail.com (O.D. Leite), monica.garcez@ufrgs. kaolinite, illite, and montmorillonite contents of composite clays
br (M.R. Garcez), medeiros.ufpr@gmail.com (M.H.F. Medeiros).

https://doi.org/10.1016/j.conbuildmat.2021.123653
0950-0618/Published by Elsevier Ltd.
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

affect the strength and pore characteristics in cementitious sys- RCW was also addressed. The results of this investigation con-
tems [16]. tribute to generating knowledge that leads to an alternative route
The red ceramic industry uses low-grade kaolin clay as a raw for RCW disposal and the use of SCM in the Brazilian Portland
material in conforming and sintering brick production processes, cement industry.
resulting in 0.20 to 0.24 ton CO2/ton brick [18–20]. Ceramic bricks
often break during handling and transportation or are discarded
during quality control, resulting in up to 30% of red ceramic waste 2. Experimental program
(RCW) [15]. The RCW contains quartz, calcined clay, and other
minor phases [14,21]. In the São Paulo State (Southeast Brazil), 2.1. Materials
the clays used as a raw material in the red ceramic industry con-
tain 21 to 36% of kaolinite [14]. The RCW from damaged bricks pro- The RCW was produced from damaged bricks collected in a
duced in this region, with particles smaller than 75 lm, has been ceramic industry located in a small municipality (23,264 inhabi-
proved to fix between 434 and 599 mg Ca(OH)2/g RCW when sub- tants and 5,673.02 km2) of Northeast Brazil, whose sintering tem-
mitted to the modified Chapelle test, which denotes reasonable perature reached 950 °C. The collected damaged bricks were
pozzolanic reactivity [14]. Thus, the fragmentation and milling fragmented with a sledgehammer until reaching approximately
can attribute pozzolanic activity to the waste, adding value to 25 mm. Each 5 kg of fragmented material was then comminuted
the RCW and making it suitable for application as SCM [22,23]. in a Los Angeles abrasion apparatus using 12 iron spheres (Ø
As a result, it reduces environmental impacts caused by clay 47 mm, 445 g each) and cycles of 3.5 h at 33 rpm. After comminu-
extraction and calcining, mitigates environmental liabilities tion, each 400 g of material was sieved for 15 min in an electro-
related to the massive amount of waste generated by the local magnetic stirrer with 150 lm (100 mesh), 75 lm (200 mesh),
ceramic industry, and stimulates reverse logistics in the construc- and 45 lm (325 mesh) sieves. Three Ø 20 mm metal disks were
tion and building sector. added to each sieve to improve particle dispersion. The material
The RCW reactivity is a result of the amorphous and disordered retained in the 150 lm and 75 lm meshes was collected and re-
aluminosilicate structure formed during the sintering process. The comminuted until passing the 45 lm sieve.
calcination temperature allows the micro-structural decomposi- The raw clay material was collected in the brick manufacturing
tion of kaolinite, resulting in metakaolin, primarily responsible plant before the extrusion process, dried for 24 h at 105 °C, and
for the RCW reactivity [24]. The amorphous phase of the RCW sol- ground to produce particles below 150 lm.
ubilizes partially in a high alkaline system, enriching the solution The Portland cement used in the paste and mortar production is
in silica and alumina. These ions react with portlandite to form a Brazilian high-early-strength with sulfate resistance [31].
hydrated cementitious compounds by pozzolanic activity, usually The fine aggregate used in the mortars is natural river sand
C-(A)-S-H (calcium aluminate silicate hydrate), ettringite, and (bulk specific gravity 2.65 g/cm3, loose bulk density 1.57 g/cm3,
AFm phases [25]. Another specific interaction in the presence of fineness modulus 1.95, and maximum size 1.20 mm).
limestone filler is the reaction of alumina from the RCW with car-
bonate, consuming calcium to precipitate hemi and/or monocar-
2.1.1. Mineralogical characterization of RCW and raw clay material
boaluminate (AFm) [16,26]. The RCW reactivity is profoundly
Fig. 1 presents diffraction patterns and mineralogical composi-
affected by the variability in the mineralogical composition and
tion obtained through Rietveld refinement for RCW and raw clay,
physical–chemical properties of the raw material [11,27], which
considering a 5°–75° 2h range, steps of 0.02° for 1 s, Cu Ka Radia-
are related to the material composition and therefore cannot be
tion, 40 kV / 30 mA, and 1° divergence slit. The crystalline phases
generalized [7]. The unfamiliarity regarding mineral composition
were identified by the ICSD (Inorganic Crystal Structure Database).
and sintering temperature requires a comprehensive characteriza-
Lithium fluoride p.a. (10% in mass) was added to the RCW sample
tion to confirm the pozzolanic activity [28], which demands the
as an internal standard. The RCW is composed basically of quartz
availability being assessed locally [25].
and illite phases. The amorphous structure characteristic halo
In the Brazilian cement industry, the use of high amounts of
observed between 17° e 33° 2h represents 61.1% of the sample.
SCM is one of the main strategies to reduce CO2 emissions [29].
The clay material is mainly composed of quartz, illite, kaolinite,
Increasing the volume of SCM in the cement composition, the
and feldspar, with anatase, mica, and magnetite as minority crys-
Brazilian cement industry reduced the clinker/cement ratio from
talline phases. The kaolinite content (21.9%) quantified in the clay
80% to 67% from 1990 to 2014 and aims at reaching 59% by 2030
raw material before sintering at 950 °C was not identified in the
and 52% by 2050 [30]. Depending on the local availability, fly
RCW diffraction pattern.
ash, blast furnace slag, and calcined clay are used, in proportions
Fig. 2 presents the thermogravimetric analysis of the raw clay
up to 50% for pozzolanic additions or up to 75% for slag [31]. In
material (temperature range 30–1000 °C, heating rate 10 °C/min,
Northeast Brazil, calcined clay is the primary SCM used in Portland
nitrogen atmosphere (50 mL/min), sample mass 10 mg in a plat-
cement production, which justifies investigations on the potential
inum crucible). The mass loss between 300 e 600 °C corresponds
use of RCW produced in the region, especially for being a clean and
to the decomposition of kaolinite. The kaolinite content estimated
ready to use waste material that can be used straight from the pro-
by stoichiometry (21.9%) is identical to that quantified by Rietveld
duction line [24] with no need for any calcining.
refinement. The mass loss related to dehydroxylation was 3.05%.
In this context, this paper investigates the use of a low-grade
kaolinite calcined clay waste obtained from broken ceramic bricks
(RCW) as a potential source of SCM. The pozzolanic reactivity of 2.1.2. Physical and chemical characterization of cement and RCW
the RCW has been investigated through chemical, mineralogical, The particle size distribution was performed in isopropyl alco-
and physical characterization. The analysis has been comple- hol (cement) or water (RCW) after ultrasound exposure (60 s) for
mented by a microstructural characterization performed in cemen- particle dispersion. Fig. 3 shows that the RCW is composed pre-
titious paste produced with 30% of RCW. Additionally, the dominantly of particles smaller than those of Portland cement.
evolution of compressive strength and apparent porosity of mor- However, the particle size of both is similar up to 0.8 lm.
tars produced with 30% of RCW in mass and different water/binder Table 1 shows bulk specific gravity [32], BET specific surface
ratios has been evaluated. The eco-efficiency in terms of global area (by hydrogen adsorption), and particle size distribution
warming potential for a Portland cement produced with 30% of parameters.
2
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

Fig. 1. Diffraction pattern and Rietveld refinement of RCW and raw clay material.

Table 2 presents RCW and Portland cement chemical composi-

tion (obtained by X-ray fluorescence spectrometry in samples cal-
cined at 1020 °C for 2 h and fused with anhydrous lithium
tetraborate) and the ability of RCW to fix lime, determined through
the modified Chapelle test [34], equivalent to [35]. Limestone filler
and gypsum contents of cement presented in Table 2 were esti-
mated by thermogravimetric analysis (TG/DTG) using the parame-
ters previously described for TG/DTG of clay raw material. The
presence of mineral addition in the cement composition was inves-
tigated through the insoluble residue (IR) test [36], equivalent to
[37].
The RCW is composed predominantly of a silico-aluminum
matrix, with iron oxide and potassium oxide as secondary con-
stituents. The RCW attends the minimum requirements for use
Fig. 2. Thermogravimetric analysis (TG) of the raw clay material. as pozzolanic material in concrete and mortar [33] regarding the
content of SiO2, Al2O3, and Fe2O3 (>70%), Na2 eq. (<1.50%), maxi-
mum loss on ignition (<10%), and percentage retained in 45 mm
(<20%). The reactivity of the RCW originated from raw clay mate-
rial with low kaolinite content (22%) and particles smaller than
45 lm (684 mg Ca(OH)2/g) complies with the minimum required
for pozzolanic materials (436 mg Ca(OH)2/g [38]) and therefore
can be used as SCM.
The Portland cement SO3 content is higher than the limit of 4.5%
established by the Brazilian standard ABNT NBR 16697 [31]. The
lime (50.17%), silica (20.14%), and alumina (5.82%) contents and
the insoluble residue (7.7%) suggest the presence of pozzolanic
mineral addition in the Portland cement composition [31].

2.2. Methods

2.2.1. Microstructural characterization in cementitious paste

Microstructural analysis was performed in cementitious pastes
produced with cement (reference) and 30% of cement in mass
Fig. 3. Particle size distribution of RCW and Portland cement. replaced by RCW, with a water/binder ratio of 0.45. After produc-
tion, the cementitious pastes were kept for 24 h in 30 mL plastic
recipients and then stored in 50 mL plastic containers with a lid
The higher BET specific surface area of the RCW potentializes
for submerged curing in distillate water. X-ray diffraction (DRX),
the pozzolanic activity. The RCW presents 90% of particles smaller
thermal analysis (TG/DTG), and apparent porosity tests were per-
than 28.53 lm, and 1.87% retained on the 45 lm sieve, which
formed in the paste samples at 7, 28, 56, 91, and 182 days. Before
attends the maximum of 20% required by the Brazilian standard
being submitted to DRX and TG/DTG, small fragments of the pastes
ABNT NBR 12653 [33] for pozzolanic materials used in concrete
were removed from the plastic containers, stored at 80 °C for 24 h
and mortars.
to stop the hydration process, dried by sublimation for 16 h (lyo-
3
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

Table 1
Physical properties of RCW and Portland cement.

Materials Bulk specific BET specific surface Particle size distribution

gravity (g/cm3) area (m2/g)
D10% (lm) D50% (lm) D90% (lm) DMÁX. (lm) Material retained
on # 45 lm (%)
RCW 2.65 12.8 1.18 6.94 28.53 71.00 1.87
Portland cement 3.05 4.3 1.94 19.38 44.08 71.00 8.99

Table 2 duced for each age of test [40]. The mortars’ production started
Chemical composition of RCW and Portland cement. by cement and RCW homogenization, through continuous stirring
Chemical composition (%) RCW Portland for 5 min, followed by mixing (binder and water for 1 min and the
cement paste with sand for 4 min). After mixing, the samples were kept in
CaO 0.22 50.17 the molds for 24 h, demolded, and then submerged in saturate lime
SiO2 72.20 20.14 water. The samples were removed from the curing tanks at the
Al2O3 16.20 5.82 ages of 7, 28, 56, 91, and 182 days and submitted to the axial com-
Fe2O3 6.29 3.10
SO3 — 7.57
pressive strength test in an EMIC DL 30,000 universal testing
MgO 0.82 7.40 machine at 0.25 ± 0.05 MPa/s.
K2O 2.12 1.47 In Table 3, the mix proportion 1.00 cement: 2.06 sand (w/b
Na2O < 0.10 3.20 0.55) corresponds to a common cement/sand proportion used in
TiO2 0.96 0.44
concrete. The mix proportions for w/b ratios of 0.35 and 0.45 con-
MnO < 0.10 0.06
SrO — 0.24 sidered the same water/(cement + sand) ratio to keep similar con-
P2O5 < 0.10 0.22 sistency and avoid superplasticizers.
V2O5 — 0.06 The apparent porosity (P) of the mortars was determined
L.O.I. 1.17 3.29 according to Eq. (1) using fragments with an approximate dimen-
Insoluble Residue (IR) (average ± standard — 7.7 ± 0.6
sion of 8 mm, extracted from the central part of the cylinders sub-
deviation)
Gypsum — 3.11 mitted to the axial compressive strength test.
Limestone filler — 5.18 Statistical analysis of axial compressive strength and apparent
Clinker content (Portland cement – IR – gypsum — 84.00 porosity test results was performed by variance analysis and
– limestone filler)
Tukey’s test with a confidence level of 95%.
Modified Chapelle (mg Ca(OH)2/g) 684.7 ± 12.7 —

RCW: Na2Oeq.: 1.49% (Na2O + 0.658.K2O).

2.2.3. Eco-efficiency
The eco-efficiency of a Portland cement produced with 30% of
RCW was addressed through the Portland cement intensity (PCin-
philizer - condenser temperature 50 °C, 80 lm Hg maximum final
tensity) and carbon dioxide emission (CO2 intensity) indexes proposed
pressure), and finally ground in a mortar and pestle. The X-ray
by Daminelli et al. [41] to quantify the eco-efficiency of cement
diffraction was performed using the same parameters previously
use.
described for clay raw material and RCW and the thermal analysis
The PCintensity index (kg.m3.MPa1), presented in Eq. (2),
with the parameters previously described for TG/DTG of clay raw
expresses the amount of Portland cement consumed in a volumet-
material. The TG/DTG results were corrected to the same non-
ric unit of mortar (kg.m3) necessary to deliver 1.0 MPa of mechan-
volatile basis.
ical strength at the age of ‘‘j” days (7, 28, 56, 91, and 182 days).
The apparent porosity (P) of the cementitious pastes was deter-
mined considering the saturated surface dry mass (MSSD), sub- Portland cement consumption
mersed mass (MSUB), and dry mass (MD), according to Eq. (1). The PC intensity ¼ ð2Þ
compressive strength at }j} days
MSSD was determined in superficially dried paste fragments, and
the MSUB was obtained by hydrostatic weighing. MD was deter- The CO2 intensity index (kg.m3.MPa1), presented in Eq. (3), esti-
mined in paste fragments after being kept at 80 °C for 24 h to mates the global warming potential of a cementitious formulation
stop hydration reactions and dried by sublimation (lyophilization) through the amount of carbon dioxide emitted by the binder (ce-
for 16 h. ment and RCW used to produce a volumetric unit of a cementitious
formulation, in kg.m3) when delivering 1.0 MPa of mechanical
MSSD - MD strength at the age of ‘‘j” days (7, 28, 56, 91, and 182 days).
P ¼ x 100 ð1Þ
MSSD - MSUB
carbon dioxide emission
CO2 intensity ¼ ð3Þ
Statistical analysis of apparent porosity test results was per- compressive strength at }j} days
formed by variance analysis and Tukey’s test with a confidence
Table 4 presents a simplified estimative of carbon dioxide emis-
level of 95%.
sions related to cement and RCW.
The average CO2 emission associated with clinker production in
2.2.2. Evolution of compressive strength and apparent porosity in Brazil is about 0.855 tCO2/t of clinker, considering no allocation for
mortars residual mineral additions [42]. The brick production in Brazil
The evolution of axial compressive strength and apparent releases 0.20 to 0.24 tCO2/t of brick [18,20]. Considering that the
porosity of mortars produced with 30% of Portland cement energy production releases 0.088 tCO2/MWh [43] and four-hour
replaced in mass by RCW (Table 3) has been monitored until the processing (grinding and milling, 750 W) for 5 kg of RCW demands
age of 182 days. Low-grade kaolinite clays displayed the potential 0.60 kWh/kg, the emission corresponding to waste processing is
of replacing cement up to 30% by mass with similar or better prop- about 0.053 tCO2/t of RCW. On the other hand, the transport (trans-
erties [16,25,39]. Four Ø 5  10 cm cylindrical samples were pro- port, freight, lorry 16–32 metric ton EURO 5, cut-off, RoW, Ecoin-
4
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

Table 3
Mix proportion of mortars.

Mixtures Binders (kg) Portland cement Fine aggregate (kg) Water/binder ratio
consumption (kg/m3) (Cement + RCW)
Cement RCW
Reference 0.35 1.00 — 968 0.94 0.35
30% RCW 0.35 0.70 0.30 668
Reference 0.45 1.00 — 744 1.50 0.45
30% RCW 0.45 0.70 0.30 515
Reference 0.55 1.00 — 604 2.06 0.55
30% RCW 0.55 0.70 0.30 419

Table 4
CO2 emission related to cement and RCW.

Mixtures Binders (kg) Binder consumption (kg/m3) Carbon dioxide emission (kg/m3)
Cement RCW Cement RCW Total Clinker1 RCW2 Total
Reference 0.35 1.00 — 968 — 968 695 — 695
30% RCW 0.35 0.70 0.30 668 286 955 486 94 580
Reference 0.45 1.00 — 744 — 744 534 — 534
30% RCW 0.45 0.70 0.30 515 221 736 374 72 446
Reference 0.55 1.00 — 604 — 604 433 — 433
30% RCW 0.55 0.70 0.30 419 180 599 303 59 362
1
Clinker: clinker content in Portland cement was 84%. Carbon dioxide emission from clinker = clinker content in Portland cement  0.855 kg/kg clinker [42]. Ex. Reference
0.35 ? Carbon dioxide emission from clinker = 968  0.84  0.855 = 695 kg/m3.
2
RCW: Carbon dioxide emission from RCW = RCW consumption  0.327 kg/kg RCW. Ex. 30% RCW 0.35 ? Carbon dioxide emission from RCW = 286  0.327 = 94 kg/m3.

vent, CML method) over 200 km would increase the CO2 release by Portland cement paste), occurs due to the increase in the reaction
about 0.034 tCO2/t of RCW. Thus, the CO2 emission related to the kinetics of Portland cement in the presence of fine particles (filler
RCW considering brick production, processing, and transport is effect), and the RCW pozzolanic activity [16].
about 0.287 to 0.327 tCO2/t of RCW. The CO2 emissions for mix- The CBW and portlandite content of the reference paste
tures with 30% of RCW, presented in Table 4, reduce by 12% with increases over time. In the paste produced with RCW, the port-
the brick production emission allocated to the origin (ceramic landite content is similar to that theoretically predicted up to
industry). 7 days. In subsequent ages, the portlandite content reduced grad-
ually as a result of RCW pozzolanic activity. The portlandite con-
sumption at 182 days was 6.4%, which results in a lime content
3. Results and discussion
fixed by the pozzolanic activity of 214 mg Ca(OH)2/g RCW. The
portlandite consumption corresponds to about 31% of the ability
3.1. Microstructural characterization in cementitious paste
to fix lime determined through the Modified Chapelle test
(684 mg Ca(OH)2/g RCW). It means that the lime content fixed
Fig. 4 shows the chemically bound water (CBW) and portlandite
by pozzolanic activity is lower than that determined in conditions
contents of the cementitious pastes obtained through thermal
to maximize the reaction. Thus, it is evident that, after 182 days,
analysis. Masses losses between 35 °C and 385 °C were consid-
the RCW can still develop pozzolanic activity, even with reactions
ered as CBW in the form of C-S-H, AFt, and AFm phases, masses
occurring slowly over time.
losses between 385 °C and 490 °C as portlandite constitution
Fig. 5 presents TG/DTG curves for the paste produced with RCW,
water, and mass loss between 490 °C and 1000 °C as carbon diox-
and the theoretical curves determined considering 70% of Portland
ide bound in calcite [44–46]. The use of RCW as SCM increased the
cement.
CBW content if compared to the Portland content paste (Fig. 4A).
The higher CBW content at 7 days, compared to the propor-
The CBW content was similar to the reference at 7 days, but a grad-
tional Portland cement content, occurs due to the higher content
ual distancing was observed in subsequent ages. The higher CBW
of C-S-H and the formation of AFm phases. The higher C-S-H con-
content in the paste produced with RCW, compared to the propor-
tent is related to the heterogeneous nucleation effect provided by
tional Portland cement content (theoretical corresponds to 70% of

Fig. 4. Chemically bound water as C-S-H and AFt/AFm phases (A). Portlandite content in pastes (B). All results in a non-volatile basis.

5
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

Fig. 6. X-ray diffraction profiles of pastes at different ages. E – Ettringite; Mc –

Monocarboaluminate; Hc – Hemicarboaluminate; B – Brownmillerite (C4AF); I –
Illite.

7 days. The ettringite content remained practically constant, with

no conversion into monosulfoaluminate (AFm), due to the high
SO3 content (7.57%) in the Portland cement composition. The high
SO3 content is related to the coprocessing of residual materials in
the production line. After 28 days, there is evidence of monocar-
boaluminate formation, also identified in a low content at 182 days.
In this case, the monocarboaluminate is related to the calcite and
dissolved alumina reactions. The calcite is part of the limestone fil-
ler composition, and the dissolved alumina is part of the Portland
cement composition (C3A and C4AF). Thus, the limestone filler pre-
sent in the Portland cement composition favored the interaction
Fig. 5. Thermal analysis (TG/DTG) of the paste with RCW at different ages. with calcined clay to form carboaluminate phases [48].
The partial substitution of Portland cement by RCW changed
the formed hydrated compounds’ content. Illite was identified up
the RCW particles [47]. As the hydration of silicate phases results to 91 days with a characteristic peak at 8.82° 2h. The ettringite con-
in C-S-H and CH formation, an increase in both C-S-H and C–H con- tent was proportional to that of reference up to 28 days, similar at
tents might be expected, which was not observed experimentally. 56 and 91 days, and higher at 182 days. The C4AF phase was iden-
Thus, it is assumed that part of the portlandite was consumed for tified over 182 days. Low content of hemicarboaluminate and
the formation of AFm phases, which means that RCW particles monocarboaluminate were identified at 7 days. After 28 days, only
increased the Portland cement degree of hydration at first ages the monocarboaluminate was identified, and the content remained
[16]. There is a similarity between the observed portlandite con- constant up to 182 days. The RCW contributed to intensifying the
tent of the paste with RCW and that theoretically estimated. ettringite content and the monocarboaluminate formation due to
At the ages of 56 and 182 days, the C-S-H content in the pastes the solubilization of alumina from the RCW amorphous matrix.
containing RCW was similar to the theoretically estimated consid- The solubilized alumina reacts in the presence of portlandite, sul-
ering 70% of Portland cement. The higher CBW content is related to fate, and limestone filler to form ettringite and monocarboalumi-
the higher content of AFt and AFm phases. The reduced remaining nate [49].
portlandite content is indicative of the RCW pozzolanic activity. The formation of carboaluminate phases restrings the conver-
Fig. 6 presents the evolution of AFt and AFm phases during the sion of ettringite to monosulphate. Therefore, the formation of C-
paste hydration, identified through X-ray diffractometry. The (A)-S-H and carboaluminate phases, combined with the ettringite
anhydrous phase C4AF and ettringite with characteristic peaks at stabilization, changes the microstructure and increases the cemen-
9.08° and 15.77° 2h were identified in the Portland cement at titious paste performance [9,50].

6
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

water/binder ratio 0.35 are statistically equivalent at 28 and

182 days, according to the Tukey test at a confidence level of 95%.
Fig. 9 presents the evolution of apparent porosity in mortars,
and the red arrows illustrate statistically equivalent results. The
water/binder ratio significantly influenced the apparent porosity
over time. The apparent porosity of 30% RCW mortar produced
with a water/binder ratio of 0.35 is statistically equivalent to that
of reference over the 182 days. The gradual increase in the
water/binder ratio tends to increase the 30% RCW mortar apparent
porosity. The apparent porosity of the mortars produced with
water/binder ratio of 0.45 showed statistical difference in some
ages, according to Tukey’s test at a confidence level of 95%. Increas-
ing the water/binder ratio of the mortar with RCW to 0.55 resulted
in higher apparent porosity.
The RCW potentializes the ettringite content and the formation
of monocarboaluminate, which demands a considerable amount of
Fig. 7. Apparent porosity of pastes with water/binder ratio = 0.45. chemically bound water and fills empty spaces in the paste. The
reduction of water/binder ratio results in a higher decrease in the
voids volume over time due to the Portland cement hydration. In
The apparent porosity evolution of pastes resulting from the such a case, RCW pozzolanic reaction occurs in a denser microstruc-
hydration process is presented in Fig. 7. The initial theoretical ture and causes a higher decrease in the mortars’ apparent porosity
apparent porosity was determined through the relationship (e.g., mortar produced with RCW and water/binder ratio of 0.35).
between the volume of mixing water and the total volume of mate- Fig. 10 presents relative variations of compressive strength and
rials. In the RCW paste, the initial theoretical porosity is lower than apparent porosity of mortars with RCW compared to reference
the reference because the bulk specific gravity of RCW is lower mortars.
than that of Portland cement (Table 2). The variance analysis with The 30% replacement of Portland cement by RCW in mass
Tukey’s test at a confidence level of 95% indicated no significant resulted in compressive strength lower than the reference, regard-
difference for the apparent porosities at different ages. less of the water/binder ratio and test age. Differently, the apparent
porosity varies depending on the water/binder ratio.
The significant variation in the apparent porosity of mortars did
3.2. Evolution of compressive strength and apparent porosity in not influence the compressive strength. The apparent porosity of
mortars mortars produced with RCW was, on average, 22% higher than
the reference for water/binder ratio of 0.55, 13% higher than the
Fig. 8 presents the effects of the partial replacement of Portland reference for water/binder ratio of 0.45, and 5% lower than the ref-
cement by RCW in the compressive strength of mortars, where the erence for water/binder ratio of 0.35. The compressive strength
red arrows show statistically equivalent results. The 30% replace- reached 84 to 96% of the reference after 28 days. The decrease of
ment of Portland cement by RCW decreased the compressive water/binder ratio tends to contribute to the percentual increase
strength over time, regardless of the water/binder ratio used in of compressive strength. Avet et al. [12] estimated the relative
the mixture. Previous studies for 20% to 40% replacement compressive strength of mortars as a function of the raw material
[21,51,52] corroborate these findings. The compressive strength kaolinite content. The authors concluded that, between 7 and
reduction was higher up to 7 days and reduced over time due to 90 days, the relative compressive strength might vary from 85 to
the RCW pozzolanic activity. The lower the water/binder ratio, 95% for a 30% partial replacement (in mass) of Portland cement
the lower the compressive strength difference observed in the by RCW with 22% kaolinite in the raw material. The results
mortars. The compressive strength of mortars with RCW and

Fig. 8. Compressive strength of mortars.

7
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

Fig. 9. Apparent porosity of mortars.

Fig. 10. Relative variation of compressive strength and apparent porosity of mortars with RCW compared to reference mortars.

Fig. 11. Correlation of apparent porosity of paste and mortar with water/binder of 0.45.

presented in this paper corroborates the findings of Avet et al. [12]. Fig. 11 presents the correlation of apparent porosity of pastes
Despite the apparent porosity higher than the reference, the com- and mortars with water/binder of 0.45. The initial theoretical
pressive strength of RCW mortar (water/binder ratio of 0.55) apparent porosity was determined through the relationship
reached 88% of the reference between 28 and 182 days, which between the volume of mixing water and the total volume of mate-
can be considered a satisfactory mechanical performance. rials. The high coefficients of determination (R2) observed in Fig. 11

8
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

Fig. 12. Portland cement intensity of mortars.

Fig. 13. Carbon dioxide intensity of mortars.

confirm the test consistency for evaluating pastes and mortars Fig. 13 presents the carbon dioxide intensity of mortars. The
apparent porosity at different ages. partial replacement of Portland cement by RCW resulted in a
reduction in the carbon dioxide emitted per unit of compressive
3.3. Eco-efficiency strength. Thus, the incorporation of RCW in mortars can reduce
environmental impacts originated by inappropriate disposal,
Fig. 12 shows the Portland cement intensity of mortars. The besides mitigating greenhouse gases emission from the construc-
Portland cement intensity of reference mortars is higher than those tion industry. Data presented in Figs. 12 and 13 show that the
of mortars produced with RCW, regardless of the water/binder modified binder with 30% Portland cement replacement by RCW
ratio. The use of RCW reduces the Portland cement consumption is environmentally more efficient considering the cement con-
and the compressive strength of mortars. On the other hand, the sumption and carbon dioxide emission per unit of compressive.
compressive strength of mortars can be partially re-established
due to the pozzolanic activity of the RCW, which results in a posi-
tive impact regarding Portland cement intensity. Therefore, the use 4. Conclusions
of RCW tends to reduce the Portland cement consumption per unit
of compressive strength. This trend enhances the environmental The potential application of RCW as SCM presented in this
performance of mortars produced with RCW. paper allows the following conclusions:
9
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

- The clay with 22% of kaolinite, after sintering and comminution Resour. Conserv. Recycl. 145 (June 2019) 148–159, https://doi.org/10.1016/j.
(particles lower than 45 lm), results in calcined clay with resconrec.2019.02.032.
[9] K. Scrivener, F. Martinera, S. Bishnoi, S. Maity, Calcined clay limestone cements
enough capacity to fix lime, which allows its classification as (LC3), Cem. Concr. Res. 114 (2018) 49–56, https://doi.org/10.1016/j.
pozzolanic mineral addition. cemconres.2017.08.017.
- The partial replacement of Portland cement by RCW changed [10] F. Avet, K. Scrivener, Investigation of the calcined kaolinite content on the
hydration of Limestone Calcined Clay Cement (LC3), Cem. Concr. Res. 107
the hydrated paste microstructure by intensifying the ettringite (2018) 124–135, https://doi.org/10.1016/j.cemconres.2020.106176.
content and favoring the formation of monocarboaluminate. [11] R. Jaskulski, D. Jóźwiak-Niedźwiedzka, Y. Yakymechko, Calcined clays as
- The CBW content was intensified at the early ages mainly due to supplementary cementitious material, Materials 13 (2020) 1–36, https://doi.
org/10.3390/ma13214734.
the higher degree of hydration of Portland cement and the for- [12] F. Avet, R. Snellings, A.A. Diaz, M.B. Haha, K. Scrivener, Development of a new
mation of AFm phases. In more advanced ages, the CBW content rapid, relevant and reliable (R3) test method to evaluate the pozzolanic
is related to the RCW pozzolanic activity and the higher forma- reactivity of calcined kaolinitic clays, Cem. Concr. Res. 85 (2016) 1–11, https://
doi.org/10.1016/j.cemconres.2016.02.015.
tion of ettringite and monocarboaluminate. [13] A. Tironi, M.A. Trezza, A.N. Scian, E.F. Irassar, Kaolinitic calcined clays: factors
- Portlandite consumption by pozzolanic activity was verified affecting its performance as pozzolans, Constr. Build. Mater. 28 (1) (2012)
after 7 days. The highest consumption, between 7 and 28 days, 276–281, https://doi.org/10.1016/j.conbuildmat.2011.08.064.
[14] E. Garcia, M. Cabral Junior, V.A. Quarcioni, F.F. Chotoli, Avaliação da atividade
coincided with the higher formation of monocarboaluminate. pozolânica dos resíduos de cerâmica vermelha produzidos nos principais polos
- The compressive strength of mortars produced with 30% of ceramistas do Estado de São Paulo, Cerâmica 61 (358) (2015) 251–258, https://
Portland cement replaced by RCW reduced by 7 to 16% at doi.org/10.1590/0366-69132015613581847.
[15] C.M. Carvalho, N.P. Barbosa, U.T. Bezerra, T.B. Simas, Red ceramic industry
182 days. The significant variation in the apparent porosity of
residues: Used to produce Portland cement. Case Studies, Constr. Mater. 13
the mortars did not influence the compressive strength. (2020) e00449, https://doi.org/10.1016/j.cscm.2020.e00449.
- The gradual increase in the water/binder ratio tends to maxi- [16] A. Dixit, H. Du, S.D. Pang, Performance of mortar incorporating calcined marine
mize the decrease in compressive strength and increase the clays with varying kaolinite content, J. Cleaner Prod. 282 (2021), https://doi.
org/10.1016/j.cscm.2020.e00449.
apparent porosity of mortars with 30% of RCW. [17] Msinjili, N. S.; Vogler, N.; Sturm, P.; Neubert, M.; Schöder; H.-J.; Kühne, H.-C.;
- The 30% replacement of Portland cement by RCW can reduce Hünger, K.-J.; Gluth, G. J. G. Calcined brick clays and mixed clays as
CO2 emissions by reducing the cement demand per unit of com- supplementary cementitious materials: Effects on the performance of
blended cement mortars. Constr. Build. Mater., 266, Part A, January 2021.
pressive strength, besides reducing current improper waste https://doi.org/10.1016/j.conbuildmat.2020.120990.
destination and the associated environmental impacts. [18] Caldas, L. R.; Lira, J. S. M. M.; Melo, P. C.; Sposto, R. M. Lyfe cycle carbon
emissions inventory of brick masonry and light steel framing houses in
Brasilia: proposal of design guidelines for low-carbon social housing.
CRediT authorship contribution statement Ambiente Construído, Porto Alegre, Vol. 17, n° 3, p. 71 – 85, Jul./Set. 2017.
http://dx.doi.org/10.1590/s1678-86212017000300163.
[19] C.K. Chau, T.M. Leung, W.Y. Ng, A review on life cycle assessment, life cycle
J. Hoppe Filho: Conceptualization, Methodology, Writing - original energy assessment and cycle carbon emissions assessment on buildings, Appl.
draft. C.A.O. Pires: Methodology, Investigation. O.D. Leite: Methodol- Energy 143 (2015) 395–413.
ogy, Investigation. M.R. Garcez: Data curation, Writing - review & edit- [20] M.R.M. Saade, M.G. Silva, V. Gomes, H.G. Franco, D. Schwamback, B. Lavor,
Material eco-efficiency indicators for Brazilian buildings, Smart Sustainable
ing. M.H.F. Medeiros: Data curation, Writing - review & editing. Built Environ. 3 (1) (2014) 54–71, https://doi.org/10.1108/SASBE-04-2013-
0024.
[21] R.D. Toledo Filho, J.P. Gonçalves, B.B. Americano, E.M.R. Fairbairn, Potential for
Declaration of Competing Interest use of crushed waste calcined-clay brick as a supplementary cementitious
material in Brazil, Cem. Concr. Res. 37 (9) (2007) 1357–1365, https://doi.org/
10.1016/j.cemconres.2007.06.005.
The authors declare that they have no known competing finan-
[22] Araújo, R. A.; Menezes, A. L. R.; Cabral, K. C.; Nóbrega, A. K. C.; Martinelli, A. E.;
cial interests or personal relationships that could have appeared Dantas, K. G. M. Evaluation of the pozzolanic activity of red ceramic waste
to influence the work reported in this paper. using mechanical and physicochemical methods. Cerâmica, Vol. 65, n° 375, p.
461–469, July/Sept. 2019. https://doi.org/10.1590/0366-69132019653752649.
[23] A.E. Lavat, M.A. Trezza, M. Poggi, Characterization of ceramic roof tile wastes as
References pozzolanic admixture, Waste Manage. 29 (5) (2009) 1666–1674, https://doi.
org/10.1016/j.wasman.2008.10.019.
[24] A. Schachow, D. Stringari, L. Senff, S.L. Correia, A.M. Segadães, Influence of fired
[1] Naik, T. R.; Moriconi, G. Environmental-friendly durable concrete made with
clay brick waste additions on the durability of mortars, Cem. Concr. Compos.
recycled materials for sustainable concrete construction. Proceedings of
62 (2015) 82–89, https://doi.org/10.1016/j.cemconcomp.2015.04.019.
International Symposium on Sustainable Development of Cement, Concrete
[25] A. Alujas, R. Fernández, R. Quintana, K.L. Scrivener, F. Martirena, Pozzolanic
and Concrete Structures, p. 485 – 505, Toronto, October 2005.
reactivity of low grade kaolinitic clays: Influence of calcination temperature
[2] Pachego-Torgal, F.; Abdollahnejad, Z.; Miraldo, S.; Kheradmand, M. Alkali-
and impact of calcination products on OPC hydration, Appl. Clay Sci. 108 (May
activated cement-based binders (AACBs) as durable and cost-competitive low-
2015) 94–101, https://doi.org/10.1016/j.clay.2015.01.028.
CO2 binder materials: some shortcomings that need to be addressed. (Chapter
[26] S. Kucharczyk, M. Zajac, J. Deja, The influence of limestone and Al2O3 content
9). In: Handbook of low carbon concrete. 1st edition. Ed. Ali Nazari and Jay G.
in the slag on the performance of the composite cements, Procedia Eng. 108
Sanjayan. p. 195 – 216, 2017. https://doi.org/10.1016/C2015-0-01844-5
(2015) 402–409, https://doi.org/10.1016/j.proeng.2015.06.164.
[3] F. Stafford, F. Raupp-Pereira, J.A. Labrincha, D. Hotza, Life cycle assessment of
[27] A. Schachow, S.L. Correia, C. Effting, Influence of microstructural and
the production of cement: a Brazilian case study, J. Cleaner Prod. 137 (2016)
morphological properties of raw natural clays on the reactivity of clay brick
1293–1299, https://doi.org/10.1016/j.jclepro.2016.07.050.
wastes in a cementitious blend matrix, Cerâmica 66 (2020) 154–163, https://
[4] C. Bataille, Low and zero emissions in the steel and cement industries. Barriers,
doi.org/10.1590/0366-69132020663782852.
technologies and policies, Green Growth and Sustainable Development (GGSD)
[28] J.M. Paris, J.G. Roessler, C.C. Ferraro, H.D. DeFord, T.G. Townsend, A review of
Forum. Theme: Greening heavy and extractive industries: innovation and
waste products utilized as supplements to Portland cement in concrete, J.
fiscal implications, Organization for Economic Co-operation and Development
Cleaner Prod. 121 (May 2016) 1–18, https://doi.org/10.1016/j.
(OECD), Paris, November 2019.
jclepro.2016.02.013.
[5] International Energy Agency (IEA) Technology Roadmap. Low-carbon
[29] Y. Kihara, G. Visedo, Economia de baixo carbono: uma visão para o futuro da
transition in the cement industry. World Business Council for Sustainable
indústria brasileira de cimento, Rev. Concr. Constr. 77 (2015) 86–91.
Development (WBCSD), April 2018.
[30] Sindicato Nacional da Indústria do Cimento (SNIC) Roadmap tecnológico do
[6] UN Environment, K.L. Scrivener, V.M. John, E.M. Gartner, Eco-efficient
cimento: potencial de redução das emissões de carbono da indústria do
cements: Potential economically viable solutions for a low-CO2 cement-
cimento brasileira até 2050. Coordenado por Gonzalo Visedo e Marcelo
based materials industry, Cem. Concr. Res. 114 (2018) 2–26, https://doi.org/
Pecchio. 64 páginas. Rio de Janeiro: SNIC, 2019.
10.1016/j.cemconres.2018.03.015.
[31] ABNT. (2018) ‘‘Portland cement – Requirements.” NBR 16697, Rio de Janeiro,
[7] G. Cardinaud, E. Rozière, O. Martinage, A. Loukili, L. Barnes-Davin, M. Paris, D.
RJ.
Deneele, Calcined clay – Limestone cements: hydration processes with high
[32] ASTM, Standard Test Method for Screening Apparent Specific Gravity and Bulk
and low-grade kaolinite clays, Constr. Build. Mater. 277 (2021) 122271,
Density of Waste, D5057–17, West Conshohocken, PA, 2017.
https://doi.org/10.1016/j.conbuildmat.2021.122271.
[33] ABNT. (2015) ‘‘Pozzolanic materials — Requirements.” NBR 12653, Rio de
[8] S.C. Gomes, J.L. Zhou, W. Li, G. Long, Progress in manufacture and properties of
Janeiro, RJ.
construction materials incorporating water treatment sludge: a review,

10
J. Hoppe Filho, C.A.O. Pires, O.D. Leite et al. Construction and Building Materials 296 (2021) 123653

[34] ABNT. (2010) ‘‘Pozzolanic materials - Determination of calcium hydroxide [45] O. Kizinievič, R. Žurauskienė, V. Kizinievič, R. Žurauskas, Utilization of sludge
fixed - Modified Chapelle’s method.” NBR 15895, Rio de Janeiro, RJ. waste from water treatment for ceramic products, Constr. Build. Mater. 41
[35] AFNOR. (2012) ‘‘Addition for concrete - Metakaolin - Specifications and (2013) 464–473, https://doi.org/10.1016/j.conbuildmat.2012.12.041.
conformity criteria.” NF P18-513, La Plaine Saint-Denis. [46] Y.P. Ling, R.-H. Tham, S.-M. Lim, M. Fahim, C.-H. Ooi, P. Krishnan, A.
[36] ABNT. (2012) ‘‘Portland cement with pozzolanic materials - Chemical analysis Matsumoto, F.-Y. Yeoh, Evaluation and reutilization of water sludge from
- Reference test method.” NM 22, Rio de Janeiro, RJ. fresh water processing plant as a green clay substituent, Appl. Clay Sci. 143
[37] ASTM. (2018) ‘‘Standard Test Methods for Chemical Analysis of Hydraulic (July 2017) 300–306, https://doi.org/10.1016/j.clay.2017.04.007.
Cement.” C114-18, West Conshohocken, PA. [47] C.A.A. Rocha, G.C. Cordeiro, R.D. Toledo Filho, Influence of stone cutting waste
[38] Raverdy, M., Brivot, F., Paillere, A. M., and Dron, R. (1980). ‘‘Appreciation of and ground waste clay brick on the hydration and packing density of cement
pozzolanic reactivity of minor components.” Proc., 7th Int. Congress on the pastes, Ibracon Struct. Mater. J 6 (4) (2013) 661–680.
Chemistry of Cement (ICCC), Vol. 3, Institut Francais des Sciences et [48] B. Lothenbach, G. Le Saout, E. Gallucci, K. Scrivener, Influence of limestone on
Technologies des Transports, de l’Aménagement et des Réseaux (IFSTTAR), the hydration of Portland cements, Cem. Concr. Res. 38 (6) (2008) 848–860,
Champs-sur-Marne, France. https://doi.org/10.1016/j.cemconres.2008.01.002.
[39] H. Du, S.D. Pang, Value-added utilization of marine clay as cement [49] Zaribaf, B. H.; Kurtis, K. E. Admixture compatibility in metakaolin-portland-
replacement for sustainable concrete produxtion, J. Cleaner Prod. 198 (2018) limestone cement blends. Materials and Structures, Vol. 51, Article number:
867–873, https://doi.org/10.1016/j.jclepro.2018.07.068. 33, p. 1 – 13, February 2018. https://doi.org/10.1617/s11527-018-1154-7.
[40] ABNT. (2019) ‘‘Portland cement – Determination of compressive strength of [50] S. Krishnan, S. Bishnoi, Understanding the hydration of dolomite in
cylindrical test specimens.” NBR 7215, Rio de Janeiro, RJ. cementitious systems with reactive aluminosilicates such as calcined clay,
[41] B.L. Damineli, F.M. Kemeid, P.S. Aguiar, V.M. John, Measuring the eco-efficiency Cem. Concr. Res. 108 (2018) 116–128, https://doi.org/10.1016/j.
of cement use, Cem. Concr. Compos. 32 (8) (2010) 555–562, https://doi.org/ cemconres.2018.03.010.
10.1016/j.cemconcomp.2010.07.009. [51] L.A. Pereira-de-Oliveira, J.P. Castro-Gomes, P.M.S. Santos, The potential
[42] V.C.H.C. Oliveira, B.L. Damineli, V. Agopyan, V.M. John, Strategies for the pozzolanic activity of glass and red-clay ceramic waste as cement mortars
minimization of CO2 emissions from concrete, Ambiente Construído 14 (4) components, Constr. Build. Mater. 31 (2012) 197–203, https://doi.org/
(2014) 167–181, https://doi.org/10.1590/S1678-86212014000400012. 10.1016/j.conbuildmat.2011.12.110.
[43] MME. (2019) Balanço Energético Nacional 2019 – Relatório Síntese. http:// [52] M. O’Farrell, B.B. Sabir, S. Wild, Strength and chemical resistance of mortars
www.epe.gov.br containing brick manufacturing clays subjected to different treatments, Cem.
[44] C. Martínez-García, D. Eliche-Quesada, L. Pérez-Villarejo, F.J. Iglesias-Godino, F. Concr. Compos. 28 (9) (2006) 790–799, https://doi.org/10.1016/j.
A. Corpas-Iglesias, Sludge valorization from wastewater treatment plant to its cemconcomp.2006.05.014.
application on the ceramic industry, J. Environ. Manage. 95 (2012) S343–S348,
https://doi.org/10.1016/j.jenvman.2011.06.016.

11