Construction and Building Materials 82 (2015) 31–38

Contents lists available at ScienceDirect

Construction and Building Materials

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

Sugarcane bagasse ash sand (SBAS): Brazilian agroindustrial by-product

for use in mortar
Fernando C.R. Almeida a, Almir Sales a,⇑, Juliana P. Moretti a, Paulo C.D. Mendes b
a
Departament of Civil Engineering, Federal University of São Carlos, Via Washington Luís, km 235, São Carlos, SP 13565-905, Brazil
b
Departament of Chemistry, Federal University of São Carlos, Via Washington Luís, km 235, São Carlos, SP 13565-905, Brazil

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

 The incorporation of SBAS fills mortar

pores with diameters below 150 lm.
mortar pores mortar pores
 The use of SBAS does not affect the without SBAS with SBAS
compressive strength of mortars.
 The carbonation depths of mortars
containing 0% and 30% SBAS are SBAS
equivalent.
 The use of SBAS increases the chloride
penetration resistance of mortars. SBAS incorporation
30%, by mass

↑ Cl- depth ↓ Cl- depth

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

Article history: SBAS (sugarcane bagasse ash sand) is the term used for the residue left over from burning sugarcane
Received 2 October 2014 bagasse. Large amounts of this agro-industrial by-product are generated in Brazilian sugar and ethanol
Received in revised form 19 January 2015 plants, and its disposal is an environmental problem. The application of SBAS as a fine aggregate in mor-
Accepted 18 February 2015
tars can add value to this waste and also reduce the use of natural sand. The growing need to extract nat-
Available online xxxx
ural sand from Brazilian rivers has caused environmental problems. In this article, the effect of SBAS on
mortars was investigated, specifically its compressive strength, porosity, carbonation depth and chloride
Keywords:
penetration. The present study fills the gap in knowledge on the durability of mortars using different
Sugarcane bagasse ash sand
Mortars
levels of SBAS. The substitution of natural sand by SBAS, especially with content of 30%, can lead to main-
Porosity tenance of mechanical properties, micropore clogging and improvement of the durability of mortars, in
Accelerated carbonation comparison with a reference mixture.
Chloride penetration Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction SBAS has high silica (SiO2) content (above 60% by mass), with
variable and coarse particle size distribution [4]. The large amount
Sugarcane production is a major agricultural activity in Brazil. of silica present in SBAS is due to the presence of sand from the
Sugarcane bagasse ash sand (SBAS) is one of the major growing and harvesting processes [3].
by-products from the processing of sugarcane to produce sugar SBAS is the crude residue generated after the burning of sugar-
and ethanol. In Brazil, 4 million tonnes of SBAS are generated per cane bagasse. This residue is collected from boilers at plants and
year [1,2]. SBAS is usually disposed of in crops, despite lacking ade- has low levels of pozzolanic reactivity. SBAS shows a predominant-
quate nutrients for its use as fertilizer [3]. ly crystalline quartz structure, which impairs its pozzolanic activ-
ity [5]. Pozzolanicity in SBAS can be achieved by controlling the
burning of bagasse (via burning temperature and burning/cooling
⇑ Corresponding author. Tel.: +55 16 33518659; fax: +55 16 33518262. time) and/or SBAS grinding conditions. After these treatments,
E-mail address: almir@ufscar.br (A. Sales).

http://dx.doi.org/10.1016/j.conbuildmat.2015.02.039
0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.
32 F.C.R. Almeida et al. / Construction and Building Materials 82 (2015) 31–38

the by-product is called sugarcane bagasse ash (SBAS [4,6], SBA [7] natural sand and mixture (SBAS + natural sand) is shown in
or BA [8,9]). Fig. 2. The composition of the mixture (SBAS + natural sand) was
The reactivity of sugarcane bagasse ash is directly dependent on classified according to the usable area as a fine aggregate, consid-
the conditions when burning the bagasse. Maximum reactivity can ering levels of 30% and 50% by mass [17].
be achieved by burning bagasse at around 500 °C [6]. Sugarcane The SBAS used in this experimental programme had a pre-
bagasse is burned in plants at temperatures between 700 °C and dominantly crystalline structure of SiO2a-quartz, as determined
900 °C, depending on its moisture content [4]. Thus, pozzolanicity by X-ray diffractometry in other studies, which revealed the
in SBAS is not obtained by the uncontrolled burning of bagasse in absence of an amorphous halo in the diffractograms [14]. The val-
plants. ues obtained by chemical analysis (before and after stan-
In addition, sugarcane bagasse ash can become reactive by dardisation) are presented in Table 2. Also, the chemical
ultrafine grinding. Product obtained by grinding to values of D80 composition of the sugarcane bagasse ash (SBAS treated) studied
(80% passing size) below about 60 lm and Blaine specific surface by other authors are shown in Table 2. The SBAS showed low poz-
areas above 300 m2/kg can be classified as pozzolans [4,9,10]. zolanic reactivity, according to the results of the modified Chapelle
Pozzolanicity can also be obtained from sugarcane bagasse ash test. In this test, 48 mg of CaO were consumed per gram of SBAS.
by combining controlled burning with grinding the ash [8,11]. The minimum consumption of CaO for a mineral addition to be
However, grinding and/or controlled burning require energy, and considered pozzolanic is 330 mg of CaO.
treating large volumes of SBAS using these processes is costly. The binder used was a Portland-composite cement with
The low pozzolanic reactivity of SBAS does not prevent its use in blast-furnace slag (CP II E 32). The chemical composition of the
construction materials [5,7,12–15]. An alternative would be to use cement is shown in Table 3.
it as a fine aggregate in mortars. The extraction of natural sand
causes environmental impacts, such as the removal of vegetation 2.2. Production of mortars
cover, siltation of rivers and degradation of waterways [16]. The
use of SBAS as a fine aggregate can reduce the use of natural sand Three series of mortars with different SBAS contents were pro-
and decrease the volume of waste disposed of in the environment. duced, with 0% (reference mortar, RM), 30% (M30) and 50% (M50)
The aim of this paper was to evaluate the use of SBAS (without substitutions of natural sand (by mass). The proportions of materi-
grinding and without controlled thermal treatment) as an alterna- als used in each series of mortar were the same proportions as
tive aggregate substitute for natural sand in the production of mor- determined by other studies [14], considering a mortar content
tar. To do so, the influence of SBAS on the mortar was investigated of 51.3%. The amount of water in each series was adjusted to main-
through measuring the compressive strength and analysing the tain the same levels of mortar consistency. The increase in SBAS
porosity, carbonation depth and chloride penetration. content led to an increased water/cement ratio (w/c) in the mix-
ture. The proportion of materials used in each series of mortar is
2. Experimental methodology shown in Table 4.
The materials used to produce the mortar were mixed in a
2.1. Materials mechanical mixer to obtain a homogeneous mass. Each mortar
was moulded in cylindrical dimensions of 50  100 mm (di-
Sugarcane bagasse ash sand (SBAS) was used as a fine aggregate ameter  height). The samples were kept in a humid chamber for
in partial substitution for natural sand (at levels of 30% and 50%, by 28 days (relative humidity of 95% ± 5%) [18].
mass) to produce mortar. SBAS samples were collected from sugar
and ethanol plants in the state of São Paulo, Brazil. These SBAS 2.3. Physical and mechanical characteristics of the mortars
samples were standardised by sieving (mesh of 4.8 mm) and grind-
ing for three minutes at a mechanic mill (mortar/pestle) [14]. The The compressive strengths of the mortars were tested by apply-
appearance of the SBAS before and after the standardisation pro- ing a load at an average speed of 0.25 MPa/s, in accordance with
cess can be seen in Fig. 1. ABNT NBR 7215:1996 [18].
The characterisation results of the SBAS and natural sand are The physical properties of the mortars were verified by testing
presented in Table 1. The particle size distribution of the SBAS, their water absorption, void ratio and dry bulk density [19]. The
mortar specimens were also characterised via optical microscopy
(OM) in order to analyse the mean pore diameter. OM was carried
out in a HIROX Digital Microscope KH-7700 equipped with a digital
image acquisition system and dark and bright field illumination
techniques. The OM technique was adequate for observing the
effect of varying the porosity of the mortars, as a function of aver-
age pore size and pore distribution.
The results of these tests were submitted to analysis of variance
(ANOVA) and Student’s t-test at a significance level of 5%.

Table 1
Characterisation results of the SBAS and natural sand.

Properties Unit SBAS Natural sand

Specific gravity g/cm3 2.57 2.45

before aer Unit weight

Water absorption
kg/m3
%
1,424.94
0.9
1,531.32
0.5
Maximum dimension mm 1.18 6.3
Fineness modulus – 1.15 2.32
Amount of powdery material finer % 16.2 –
than 75 lm (n° 200) sieve by washing
Fig. 1. SBAS appearance: before and after the standardisation process.
 F.C.R. Almeida et al. / Construction and Building Materials 82 (2015) 31–38 33

The carbonation depth was verified using the colorimetric

method. The surface obtained by the diametrical rupture of the
specimen was treated with a phenolphthalein solution (1% phe-
nolphthalein, 29% distilled water, 70% isopropyl alcohol) [22].
The carbonation depth values were submitted to analysis of vari-
ance (ANOVA) and Student’s t-test, at a significance level of 5%.

2.5. Chloride penetration in mortars

The chloride penetration test was conducted using the col-

orimetry method by sprinkling silver nitrate and fluorescein
solution.
After 28 days of curing in humid chamber, mortar samples were
dried at 50 ± 5 °C until at a constant mass (drying period of 7 days).
Then, the samples were submerged in a saline solution at a concen-
tration of 3.5% NaCl by mass. The samples were completely sub-
merged in this solution up to the age of 30, 60 and 90 days [23].
They were completely submerged in order to ensure that the speci-
Fig. 2. Sieve analysis of the natural sand, SBAS and mixture (natural sand + SBAS).
men’s entire surface was in contact with the saline solution, thus
avoiding different exposure levels in the same sample. After each
of these ages, the samples were dried for 3 days at 105 ± 5 °C.
2.4. Carbonation test in mortars The surface obtained by the diametrical rupture of the specimen
was treated with 0.1 M silver nitrate solution (AgNO3). Then,
Carbonation tests in the mortars were conducted in two stages: fluorescein alcoholic solution (1% fluorescein salt, 29% distilled
pre-conditioning and conditioning in accelerated environment. water, 70% ethanol) was sprinkled onto the specimen’s surface.
The first step was initiated after 28 days of moist curing. The A white precipitate (AgCl) was formed after the sprayed surface
specimens were dried at 50 ± 5 °C for 7 days to obtain mass con- dried. This whitish region, in contrast to the darker area, indicates
stancy. Subsequently, the samples were kept for 5 days in a dry cli- the presence of chlorine ions. The depth of this whitish area was
matic chamber at a temperature of 23 ± 1 °C and relative humidity measured from the outer edge of the specimen, which indicates
of 60% ± 5%. the chloride penetration depth in the mortar matrix. The chloride
In the second phase (conditioning), the specimens were kept in penetration depth values were submitted to analysis of variance
an accelerated carbonation chamber. In this chamber, the CO2 con- (ANOVA) and Student’s t-test at a significance level of 5%.
tent was 15% ± 2% (by volume) and the relative humidity ranged
from 50% to 85%; both were chosen according to other studies. 3. Results and discussion
Some authors recommended carbon dioxide levels below 20%, in
order to avoid microstructural changes developed at higher levels 3.1. Mechanical and physical characterisation
that are not developed under natural conditions [20].
Additionally, the carbonation progress is slower in low-humidity The compressive strength test results of the mortar samples at
conditions (lower than 50%) because there would not be enough 28 days are shown in Fig. 3. The average compressive strength val-
water to dissolve the CO2. Otherwise, in conditions with ues of the three mortar mixtures were statistically equivalent, at a
water-saturated pores, the carbonation would also be retarded significance level of 5% (ANOVA). Thus, the use of SBAS did not
because of the slower CO2 diffusion [21]. The samples were condi- interfere in the mechanical strength of the mortars.
tioned in the chamber until the following ages, during which the For a given consistency, SBAS demanded more water in the
carbonation depth was checked: 7, 14, 28, 56, 84 and 365 days. preparation of mortar without loss of strength. SBAS has larger sur-
The use of mortar specimens was intended to exclude the inter- face area, due to its increased fineness, compared to natural sand.
ference of coarse aggregate in the analysis of concrete carbonation. Thus, the incorporation of SBAS demanded higher amounts of
Coarse aggregate acts as a barrier to the penetration of CO2 into the water in the mixture. However, the mortars with SBAS showed
mortar matrix. Thus, the carbonation depth in concrete is usually equivalent compressive strength values to that of the reference
less than the carbonation depth observed in mortars. sample, even with the expected reduction in compression strength

Table 2
Chemical composition of the SBAS (by mass, %) used on the experimental program and compared with other studies.

Elements SBAS (before standardisation) SBAS (after standardisation) Faria et al. [3] Cordeiro et al. [4] Souza et al. [12] Sales & Lima [14]
SiO2 80.2 80.8 61.6 78.3 85.5 88.2
Fe2O3 5.6 5.8 7.4 3.6 1.3 5.1
K2O 4.0 3.9 6.2 3.5 3.5 1.3
Al2O3 2.6 2.5 5.9 8.6 5.3 2.3
CaO 1.8 1.6 5.0 2.2 2.1 0.6
MgO 1.6 1.5 1.2 1.7 1.1 0.4
P2O5 1.4 1.4 1.0 1.1 0.5 0.4
TiO2 1.4 1.3 1.5 – 0.3 1.0
Na2O 0.2 0.2 – 0.1 – 0.1
MnO 0.2 0.1 – 0.1 – –
SO3 0.1 0.1 0.4 – – <0.1
Loss on ignition 0.80 0.70 9.8 0.42 – 0.35
34 F.C.R. Almeida et al. / Construction and Building Materials 82 (2015) 31–38

Table 3 The explanation for the increased average pore diameter value
Chemical and physical properties of the Portland cement CP II E 32 (source: of mortars with SBAS is in the analysis of Fig. 5. The following dis-
manufacturer).
cussion is based on the analysis of the regions highlighted in yel-
Properties Unit Cement Elements (%) Cement low (light colour) and violet (dark colour).
Specific gravity g/cm3 3.02 MP 1000 °C 4.09 Firstly, the frequency and variation of the mortars’ pore dia-
Granulated slag content % 34.0 SiO2 24.10 meters in the area highlighted in yellow in Fig. 5 will be discussed.
Initial setting time min 197 Al2O3 7.42 The frequency of pores with diameters below 150 lm was higher
Final setting time min 279 Fe2O3 3.08
CaO 52.10
in the reference mortar (without SBAS) because there were fewer
Retained #200 % 3.60 MgO 3.38 small pores in the mortar samples with SBAS. The SBAS provided
Fineness Blaine specific cm2/g 4141 SO3 2.17 a physical effect of filling smaller pores due to its smaller particle
surface size, compared to natural sand. The high content of powder mate-
Na2O 0.28
rial in the SBAS (16.2% by mass, Table 1) favours the filling of
3 days MPa 18.5 K2O 1.05
Compressive 7 days MPa 26.8 Carbonic 1.52 small-diameter pores in mortars. The content of powder material
strength (fcj) anhydride – CO2 was greater than the maximum acceptable amount of 5% for con-
28 days MPa 39.5 Insoluble residue 2.47 ventional fine aggregates [17]. The filling of smaller-diameter
pores may cause an increase in the average pore diameter of a mor-
tar, as shown in Fig. 6 and algebraically in Expressions 1–7. Thus,
with the increased w/c ratio. This fact can be explained by the filler the lack of small-diameter pore values in the average calculation
effect of SBAS. Fine SBAS particles have the physical effect of filling of mortars with SBAS may favour the increase of the final average
mortar micropores, allowing the maintenance of mechanical pore diameter value. This behaviour was similar between the sam-
strength with the increasing w/c ratio. ples with 30% and 50% SBAS.
The results of the dry bulk density, absorption and void ratio
tests on the mortars (RM, M30 and M50) at 28 days are shown in  1 ¼ n:D þ m:d
X ð1Þ
Table 5. nþm
The water absorption and void ratio values showed significant
differences between the samples with and without SBAS. The val-  2 ¼ n:D
X ð2Þ
ues of the M30 and M50 samples were statistically different from n
the RM values. Thus, the use of SBAS may cause an increase in 2
1 < X
Hypothesis : X ð3Þ
the water absorption and void ratio of the mortars. This increase
may be caused by the increased absorption of SBAS, compared to
n:D þ m:d n:D
natural sand. Furthermore, the difference in the porosity of mortars < ð4Þ
due to the increased w/c ratio required for casting may have nþm n
increased the water absorption (around 7%, when comparing
n:D þ m:d < n:D þ m:D ð5Þ
M50 and RM) and void ratio (around 6%, when comparing M50
and RM). Dry bulk density was also influenced by this effect of
m:d < m:D ð6Þ
increased porosity because of the increased w/c ratio. The differ-
ence in the dry bulk density values (around 2%) was considered
d < D ðQ :E:D:Þ ð7Þ
significant, just when comparing the values of the reference mortar
(RM) and M50 sample (50% substitution by SBAS). However, the where:
increases of these porosity indicators can be controlled by reducing  1 = mean pore diameter value of mortar without SBAS.
X
the w/c ratio in each of the three mixtures [11].  2 = mean pore diameter value of mortar with SBAS.
X
The mean pore diameter values of the mortars were obtained D = diameter of larger pores.
using optical microscopy (OM), Table 6. The variance values indi- d = diameter of smaller pores (which can be filled by SBAS).
cate high statistical dispersion of the measures, which indicates n = number of pores with diameter equal to ‘‘D’’.
the variety of pore size diameters in the different mortar samples. m = number of pores with diameter equal to ‘‘d’’.
This variation in the pore diameter measurements of the three Secondly, the frequency and variation of the pore diameter in
mortar samples can be seen in Fig. 4. mortars in the area highlighted in violet in Fig. 5 will be discussed.
The porosity variation was significant when comparing the Many pores with diameters greater than 360 lm were observed in
average values of MR and M50 samples using statistical tests the mortar with 50% SBAS. This increase in the porosity of the M50
applied at a significance level a of 5% (ANOVA and Student’s t).
Thus, the average pore diameter of the reference mortar was statis-
tically equivalent to the average pore diameter of the mortar with
30% SBAS. 50.00 45.94 46.27 45.63
Compressive strength (MPa)

Furthermore, the number of pores decreased as the content of

SBAS in the mortar increased (number of observations, Table 6). 40.00
Nevertheless, the data normality (Fig. 5) showed that the distribu-
30.00
tion of the observations was uniform.
20.00
Table 4
Mix proportion of materials to produce the mortars. 10.00

Group SBAS content (%) Mix proportion (by mass) Consistency (mm) 0.00
RM M30 M50
Cement Sand SBAS Water
Group of mortars
RM 0 1.000 2.010 – 0.420 250 ± 10
M30 30 1.000 1.407 0.603 0.440 251 ± 10
Fig. 3. Compressive strength test results of the mortar samples at the age of
M50 50 1.000 1.005 1.005 0.470 258 ± 10
28 days.
 F.C.R. Almeida et al. / Construction and Building Materials 82 (2015) 31–38 35

Table 5
Physical properties of the mortar samples at the age of 28 days.

Group Absorption (%) Void ratio (%) Dry bulk density (g/cm3)
1 2 1 2
Average value (%) SD (%) CV (%) Average value (%) SD (%) CV (%) Average value (g/cm3) SD1 (g/cm3) CV2 (%)
RM 10.17 0.20 2 20.64 0.30 1 2.03 0.01 1
M30 10.72 0.19 2 21.56 0.30 1 2.01 0.01 1
M50 10.92 0.32 3 21.79 0.52 2 2.00 0.01 1
1
SD – Standard deviation.
2
CV – Coefficient of variation.

Table 6
Average pore diameter values of the mortars by optical microscopy.
L23
Group Average value Standard deviation Variance Number of L21
L22
(lm) (lm) (lm2) observations L26 L25
L11 L24

RM 193.940 114.214 13,044.87 88 L14

L12 L19
M30 226.235 91.008 8,282.43 56
M50 256.424 140.086 19,624.03 42 L13
L20
L2 L9
samples may be related to the increase in the w/c ratio required for L10 L16
casting the specimens [11]. This increased amount of water in the L1
L18
mixture is related to the increased incorporation of SBAS, in order L15 L17
L5 L6
for the mortars to maintain the same consistency. Additional water
was absorbed by the SBAS to maintain the mortar’s consistency
L4
value during casting. This increase in water may have favoured L3 L8
the increase in porosity during the mortars’ curing period. This L29
L7
phenomenon may have contributed to the increase in the average L28
500µm
pore diameter values in the mortars with SBAS, especially when RM
compared to the RM and M50 mixtures.
Besides a possible interference by the w/c ratio, the variation in
L5
the amount of pores in each mortar was confirmed by the incorpo- L6
ration of SBAS (thin material) in smaller pores. The average pore
diameter value increased (comparing RM and M50), but the L11
amount of small pores decreased. SBAS can fill small pores, reduc- L8 L12
ing their amount. Thus, the calculation of the average pore dia- L14
L1
meter value in M50 showed the predominance of larger pores L7
that were not filled with SBAS. This arrangement of SBAS and the
L2
filling of pores can be better understood by the schematic repre- L13
sentation shown in Fig. 6 and by Expressions 1–7. This effect of
increasing the porosity is increased when considering the sig- L9
nificant increment in the w/c ratio of the M50 sample. L4 L10
Thus, the average pore diameter value of the M50 sample was
higher than that observed in the reference sample (RM).
Furthermore, the amount of pores larger than 360 lm increased M30 L3 500µm
as the necessary w/c ratio to involve the specific surface of the
SBAS content that exceeds 30% substitution increased. In other
words, the incorporation of 30% SBAS and the filling of micropores
(smaller than 150 lm) were ideal in compensating for the porosity
increase given by the increased w/c ratio (verified in the pores with
diameters greater than 360 lm), because the average pore dia-
meters of the RM and M30 mortars were statically equivalent. In
summary, the substitution value of 30% SBAS was sufficient to fill L5
L1
pores with smaller diameters (without increasing the amount of
L2 L6
large pores). From this value, the addition of SBAS will require an
increase in the w/c ratio and the consequent formation of pores L7

of larger diameters.
The incorporation of SBAS influenced the filling of smaller pores L8
in the mortar matrix due to their smaller particle size, compared to
natural sand. The higher the SBAS fineness, the better the packing L3
L4
effect of the particles, which can lead to higher compression M50 500µm
strength values, since the w/c ratios are kept constant [6].

Fig. 4. Pore analysis of the mortars by optical microscopy.

36 F.C.R. Almeida et al. / Construction and Building Materials 82 (2015) 31–38

Fig. 5. Normal distribution of pore diameter measures for the mortars obtained by optical microscopy.

Fig. 6. Schematic representation of the calculation to determine the average pore diameter of the mortars without SBAS and with 30% SBAS.

Fig. 8. Average carbonation depth values at different ages for the mortar samples.

Fig. 7. The M50 specimen after colorimetric treatment at 365 days for carbonation
depth analysis.
The carbonation depth magnitudes among the studied mixtures
were similar at early conditioning ages. However, the M50 sample
3.2. Analysis of carbonation depth in mortars at the age of 365 days had significantly greater carbonation depth
than the RM and M30 samples. Carbonation depth can be
Fig. 7 shows the colorimetric treatment applied to the mortar decreased by reducing the w/c ratio and by the packing effect of
specimens. The average carbonation depth values (in mm) in accel- the cementitious matrix provided by the SBAS fineness. The incor-
erated testing are shown in Fig. 8. poration of finer material and the reduction in the w/c lead to a
 F.C.R. Almeida et al. / Construction and Building Materials 82 (2015) 31–38 37

Fig. 9. Comparison of the mean pore diameter and carbonation depth (CD) at the
age of 365 days of the mortar samples. Fig. 11. Average chloride penetration values at different ages for the mortar
samples.

it showed the highest average chloride penetration values at all

of the studied ages. Thus, the incorporation of SBAS increased the
chloride penetration resistance of the mortars.
The increased chloride penetration resistance can be attributed
to the physical and chemical effects of SBAS.
The physical effects are related to transport mechanisms.
Considering the conditions of this study, the entry of chlorides into
mortar essentially occurs by three transport mechanisms: absorp-
tion, permeability and diffusion, occurring solely or in combina-
tion. Diffusion is considered to be the main transport process of
chlorine ions into mortar. In a simplified manner, a description of
this phenomenon is based strictly on Fick’s laws, since the actual
mechanism is much more complex. The presence of other ions in
the pore solution (ions of Na+, K+, Ca2+, OH, etc.) influences the
Fig. 10. The RM specimen after colorimetric treatment at 90 days for chloride
rate of entry of Cl, accelerating or slowing its penetration (via
penetration analysis.
mechanisms of repulsion or attraction among electrical charges)
[25]. Also, the filling of micropores with SBAS particles, especially
denser and stronger mortar matrix; as a result, the penetration of pores with dimensions smaller than 150 lm, may modify the mor-
the carbon dioxide proceeds more slowly [24]. tar microstructure, resulting in a less permeable mortar. Thus, the
Thus, the carbonation front advances similarly in both mortars incorporation of finer materials such as SBAS affects the transport
with 30% SBAS and without SBAS. This statement corroborates the mechanisms.
analysis of the average pore diameters of the mortars (item 3.1): The chemical effect can be attributed to a possible pozzolanic
30% SBAS is sufficient to keep the reference mortar porosity stan- reactivity of SBAS (even if small), leading to the increased pre-
dard (so that both average pore diameters are statistically equiva- cipitation of hydration products. In summary, some authors con-
lent), resulting in similar carbonation depths. cluded that the addition of a pozzolan increases the number of
The carbonation phenomenon is related to strength and poros- nucleation sites for the precipitation of hydration products,
ity. In the present study, the compressive strength of the three reduces Ca(OH)2 and improves the permeability of mortar [9,26].
mortar mixtures did not change significantly. Thus, the higher car- This precipitation effect may have occurred in the M30 and M50
bonation depth advancement was due to the significant porosity samples by reactions between a small portion of the reactive che-
difference between the MR and M50 samples (Fig. 9). mical species present in SBAS (SiO2, Fe2O3 and Al2O3, Table 2) and
Therefore, considering the carbonation progress, mortar with Ca(OH)2 in the Portland cement. Furthermore, precipitation due to
30% SBAS showed similar values to those obtained for the sample the reaction of NaCl with SBAS compounds (including Fe2O3, Al2O3
without SBAS. and Pb2+ [14]) may also have occurred during the immersion of the
mortar samples in the saline solution.
3.3. Analysis of chloride penetration in mortars Therefore, the incorporation of SBAS in mortars had a positive
effect on the chloride penetration resistance, compared with the
Fig. 10 shows the colorimetric treatment applied to the mortar reference mortar.
specimens. The average chloride penetration values in millimetres)
obtained at different immersion times in a 3.5% NaCl solution are 4. Conclusions
plotted in Fig. 11.
The M30 and M50 samples showed significantly different chlo- From the results obtained in the experimental study, the follow-
ride penetration values in the first month of conditioning. ing can be concluded:
However, the samples with SBAS (M30 and M50) showed penetra-
tion values with the same magnitude at older ages. The mortar  The incorporation of SBAS at levels up to 50% does not affect the
without SBAS (RM) was less resistant to chloride attack because compressive strength of mortars with same consistency;
38 F.C.R. Almeida et al. / Construction and Building Materials 82 (2015) 31–38

 The incorporation of SBAS can reduce the amount of mortar [7] Lima AS, Varum H, Sales A, Neto VF. Analysis of the mechanical properties of
compressed earth block masonry using the sugarcane bagasse ash. Constr
pores with smaller diameters (below 150 lm) due to the small-
Build Mater 2012;35:829–37.
er particle size of SBAS. This indicates a possible packing effect [8] Ganesan K, Rajagopal K, Thangavel K. Evaluation of bagasse ash as
of SBAS on the mortar. Furthermore, 30% content is favourable supplementary cementitious material. Cement Concrete Comp
to the mixture because it leads the mortar to have an amount 2007;29(6):515–24.
[9] Rukzon S, Chindaprasirt P. Utilization of bagasse ash in high-strength concrete.
of pores with larger diameters (greater than 360 lm) equivalent Mater Design 2012;34:45–50.
to the reference sample; [10] Somna R, Jaturapitakkul C, Rattanachu P, Chalee W. Effect of ground bagasse
 The carbonation depths are statistically equivalent for the mor- ash on mechanical and durability properties of recycled aggregate concrete.
Mater Design 2012;36:597–603.
tars containing 0% and 30% SBAS, even after 365 days of condi- [11] Paula MO, Tinoco IFF, Rodrigues CS, Saraz JAO. Sugarcane bagasse ash as a
tioning. Substitution contents greater than 30% can lead to partial Portland cement replacement material. Dyna 2010; 77 (163): 47–54.
increased carbonation front; Available: <http://www.scielo.org.co/scielo.php?script=sci_arttext&pid=
S0012-73532010000300005&lng=en&tlng=en>.
 The incorporation of SBAS, both in levels of 30% and 50%, can [12] Souza AE, Teixeira SR, Santos GTA, Costa FB, Longo E. Reuse of sugarcane
significantly increase the chloride penetration resistance of bagasse ash (SCBA) to produce ceramic materials. J Environ Manage
mortars. 2011;92(10):2774–80.
[13] Teixeira SR, Magalhães RS, Arenales A, Souza AE, Romero M, Rincón JM.
Valorization of sugarcane bagasse ash: Producing glass-ceramic materials. J
The substitution of natural sand by SBAS, especially at a content Environ Manage 2014;134:15–9.
of 30%, can lead to the maintenance of mechanical properties, [14] Sales A, Lima SA. Use of Brazilian sugarcane bagasse ash in concrete as sand
replacement. Waste Manage 2010;30(6):1114–22.
micropore clogging and improvement of durability of the mortar,
[15] Modani PO, Vyawahare MR. Utilization of bagasse ash as a partial replacement
in comparison with a reference mixture (without SBAS). of fine aggregate in concrete. Procedia Eng 2013;51:25–9.
[16] Sales A, Souza FR. Concretes and mortars recycled with water treatment sludge
Acknowledgements and construction and demolition rubble. Constr Build Mater
2009;23(6):2362–70.
[17] ABNT – NBR 7211. Agregados para concretos – Especificação (Aggregates for
The authors acknowledge CAPES and MCTI/CNPq for the finan- concrete – Specification). Rio de Janeiro; 2009 [in Portuguese].
cial support; Cosan for supplying the SBAS; Grupo de Materiais [18] ABNT – NBR 7215. Cimento Portland – Determinação da resistência à
compressão (Portland cement – Determination of compressive strength). Rio
Eletroquímicos e Métodos Eletroanalíticos – IQSC/USP for support- de Janeiro; 1996 [in Portuguese].
ing the optical microscopy analysis; and Laboratório [19] ABNT – NBR 9778. Argamassa e concreto endurecido – Determinação da
Interdisciplinar de Eletroquímica & Cerâmica/UFSCar. absorção de água por imersão, índice de vazios e massa específica (Hardened
cement mortar and concrete – Determination of water absorption by
immersion, borehole index and bulk density – Method of test). Rio de
References Janeiro; 1987 [in Portuguese].
[20] Sanjuan MA, Olmo C. Carbonation resistance of one industrial mortar used as a
[1] Conab – National Company of Supply. Acompanhamento da safra brasileira de concrete coating. Build Environ 2001;36(8):949–53.
cana-de-açúcar – Primeiro levantamento da safra 2014/15 (Monitoring of the [21] Pauletti C, Possan E, Dal Molin DCC. Accelerated carbonation: state-of-the art
Brazilian sugarcane crop – First survey of the 2014/15 crop). Brasília: Conab; of the research in Brazil. Ambiente Construído 2007; 7(4): 7–20. Available:
April 2014. Available: <http://www.conab.gov.br/OlalaCMS/uploads/arquivos/ <http://seer.ufrgs.br/index.php/ambienteconstruido/article/view/3750/2103>
14_04_15_15_44_37_boletim_cana_portugues_-_1o_lev_-_14.pdf> [in [in Portuguese].
Portuguese]. [22] RILEM Concrete Permanent Comitee. RILEM draft recommendation –
[2] FIESP/CIESP. Ampliação da oferta de energia através da biomassa (Expanding Measurement of hardened concrete carbonation depth CPC-18. Mater Struct
energy supply from biomass). São Paulo: FIESP/CIESP; 2001. Available: <http:// 1984;17(102):435–40.
www.fiesp.com.br/arquivo-download/?id=4505> [in Portuguese]. [23] AASHTO – T259. Standard method of test for resistance of concrete to chloride
[3] Faria KCP, Gurgel RF, Holanda JNF. Recycling of sugarcane bagasse ash waste in ion penetration. Washington; 1980.
the production of clay bricks. J Environ Manage 2012;101:7–12. [24] Rukzon S, Chindaprasirt P. Strength and carbonation model of rice husk ash
[4] Cordeiro GC, Toledo Filho RD, Tavares LM, Fairbaim EMR. Ultrafine grinding of cement mortar with different fineness. J Mater Civil Eng 2010;22(3):253–9.
sugar cane bagasse ash for application as pozzolanic admixture in concrete. [25] Figueiredo CP, Santos FB, Cascudo O, Carasek H, Cachim P, Velosa A. The role of
Cement Concrete Res 2009;39(2):110–55. metakaolin in the protection of concrete against the deleterious action of
[5] Teixeira SR, Souza AE, Santos GTA, Peña AFV, Miguel AG. Sugarcane bagasse chlorides. Rev IBRACON Estrut Mater 2014; 7(4): 685–708. Available: <http://
ash as a potential quartz replacement in red ceramic. J Am Ceram Soc dx.doi.org/10.1590/S1983-41952014000400008>.
2008;91(6):1883–7. [26] Chindaprasirt P, Rukzon S, Sirivivatnanon V. Resistance to chloride penetration
[6] Cordeiro GC, Toledo Filho RD, Tavares LM, Fairbaim EMR. Pozzolanic activity of blended Portland cement mortar containing palm oil fuel ash, rice husk and
and filler effect of sugar cane bagasse ash in Portland cement and lime mortars. fly ash. Constr Build Mater 2008;22(5):932–8.
Cement Concrete Comp 2008;30(5):410–8.