See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/317310864

A methodology for assessing the chemical and

physical potential of industrially sourced rice
husk ash on strength dev....

Article in Construction and Building Materials · September 2017

DOI: 10.1016/j.conbuildmat.2017.05.187

CITATIONS READS

3 44

2 authors:

Franco Zunino Mauricio Lopez

École Polytechnique Fédérale de Lausanne Pontifical Catholic University of Chile
15 PUBLICATIONS 32 CITATIONS 45 PUBLICATIONS 301 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Limestone Calcined Clay Cement View project

Green Concrete - Cement Replacement View project

All content following this page was uploaded by Franco Zunino on 02 June 2017.

The user has requested enhancement of the downloaded file.

 A methodology for assessing the chemical and physical potential of industrially
sourced rice husk ash on strength development and early-age hydration of
cement paste

Franco Zuninoa,1 and Mauricio Lopeza,b,c

a
Department of Construction Engineering and Management, School of Engineering,
Pontificia Universidad Catolica de Chile, Santiago, Chile
b
Center for Sustainable Urban Development (CEDEUS), Pontificia Universidad Catolica
de Chile, Vicuña Mackenna 4860, Casilla 306, Correo 22, Santiago, Chile
c
Research Center for Nanotechnology and Advanced Materials (CIEN-UC’), Pontificia
Universidad Catolica de Chile, Santiago, Chile

Abstract

Among supplementary cementitious materials (SCMs), rice husk ash (RHA) is a by-
product of the agricultural industry that is recognized by its high reactivity and ability to
enhance performance of cement-based materials. The dependence of RHA reactivity on
burning conditions and grinding procedures makes it sensitive to uncontrolled
combustion/production. This study assesses 5 RHAs from industrial sources to
understand how uncontrolled combustion impacts the reactivity of RHA. A fundamental
model that couple physical (specific surface area) and chemical (amorphous silica
content) characteristics of RHA with hydration and strength gain is proposed,
considering the relative contribution of each effect over time. Isothermal calorimetry was
used to continuously assess strength development at early ages. The results show that
the model is well correlated with strength development at different ages and allows the
computation of an equivalent water-to-cement ratio considering solely the characteristics
of a RHA sample, and prior to producing and testing specimens.

Keywords: calorimetry, hydration, surface area, amorphous material, pore size

distribution

1
Current address: Laboratory of Construction Materials, EPFL-STI-IMX-LMC, Station
12, 1015 Lausanne, Switzerland. franco.zunino@epfl.ch

1
1 Introduction

The effects of SCMs can be analysed as two simultaneous, non-divisible contributions

occurring in a cementitious system:

1. Chemical contribution, referred as the pozzolanic effect. This is the formation of

new hydration products by reaction of portlandite (CH, produced during cement
hydration) and the aluminosilicate phases present in supplementary cementitious
materials [1,2]. The rate of reaction depends on the specific surface available to
react and the amount of CH, aluminosilicate, and water.
2. Physical contribution, referred to as the filler effect, is related to changes in the
particle spacing and surface availability for nucleation of hydration products [2–
4].

Rice husk ash (RHA) is obtained from agricultural waste and has been widely
investigated owing to its potential as an SCM [5–7]. World rice production rose to
approximately 740 billion tonnes in 2014 [8]. The main uses of rice husks are as fuel for
the paddy milling process [9] and electric power plants [10]. RHA is obtained by
combustion of the rice husks, which releases a large proportion of the organic matter
contained, leaving a residue with high silica content [11,12]. The effects of rice husk ash
on concrete strength [5,6,9,12–15] and durability [13,16–20] are well documented in the
literature. The high reactivity observed in RHA is attributed to its high amorphous silica
content and its high specific surface area.

The amorphous silica content in RHA is impacted mainly by the burning temperature and
time of the husks [11,12,21]. While higher burning temperatures are more efficient at
removing organic matter quickly, this may also lead to crystallization of the silica phases
[22]. Della et al. [11] found that burning rice husk at 700ºC for 6 h produced RHA with
95% amorphous silica content. Pre-treatment of rice husk with hydrochloric acid prior to
the burning process has also been shown to be effective to stabilize the pozzolanic
activity but also to reduce the sensitivity of the RHA composition to the burning
conditions [23].

It has been shown in previous studies that in addition to chemical composition, grinding
is a controlling factor to produce high-reactivity RHA [13,24]. The high specific surface
area of RHA results from both its particle size [25] and mesoporous structure [26–28].
Cordeiro et al. [25] found that grinding RHA to particle sizes up to 10 𝜇m decreases the
specific surface area and then increases it. Finer SCMs enhance hydration owing to the
filler effect [3,29]. Furthermore, it has been shown that even crystalline RHA blended
concrete exhibits an increase in compressive strength owing to this physical interaction
[9]. However, Mehta [30] established that grinding RHA to finer particle sizes should be
avoided to preserve its mesoporous structure, which explains its high reactivity.

Whereas the fundamental variables controlling the performance of RHA blended

mixtures have been identified in previous studies, the relationship between them and
their relative contribution to hydration and strength development has not been
established in a way that would allow mixture design based on RHA characterization.
Furthermore, most of the studies present results with ashes obtained in laboratory-

2
controlled conditions, which differ significantly from the feasible industrial applications of
this material. This study focuses on the characterization of five industrially sourced RHA
samples from a chemical and physical perspective and relates these parameters to the
observed strength development of blended mixtures.

2 Materials and methods

Five samples of RHA from different industrial suppliers (four in Brazil and one in China)
were obtained and used in this study to assess the effect of variability in calcination
conditions and control of the manufacturing process in RHA reactivity. No attempts to
modify the RHA samples were made, and they were all used as received. Type I OPC
was used for the entire experimental programme conducted.

The chemical composition was determined by X-ray fluorescence (XRF), and the results
are shown on Table 1 along with phase composition of OPC determined by quantitative
X-ray diffraction (QXRD). In addition, relevant information provided by the suppliers
regarding calcination temperature is included along with the specific gravity (SG) of the
raw materials. No information about calcination time and heating rates was obtained,
which have been found to be relevant parameters [31] to optimize amorphous silica
content by avoiding crystallization along with temperature [11,21,32]. Therefore, this
would be assumed as part of the intrinsic variability of the industrially sourced ashes.

The loss on ignition (LOI) is an indicative of the content of volatiles present in the
sample, particularly carbon. Another factor that significantly affects the residual carbon
content of RHA is calcination time [11,12,32], which was not reported by the suppliers.
Therefore, calcination temperatures should be taken as referential parameters of the
expected LOI because lower LOI values may be obtained by calcination at lower
temperatures for longer periods of time.

Table 1. Chemical composition of OPC and RHA samples, phase composition of OPC
and calcination temperatures of RHA. Chemical composition data is presented as
percentage by wt.
OPC RHA-1 RHA-2 RHA-3 RHA-4 RHA-5
SiO2 21.59 83.22 73.12 78.78 80.01 93.95
Al2O3 3.82 0.00 0.00 0.00 0.00 0.53
Fe2O3 3.05 0.07 0.12 0.29 0.08 0.35
CaO 64.34 0.65 0.52 0.66 0.54 0.91
Na2O 0.54 0.65 0.07 0.12 0.09 0.21
K 2O 0.46 1.14 0.42 0.63 0.52 1.26
MnO 0.06 0.22 0.10 0.17 0.15 0.12
TiO2 0.33 0.00 0.01 0.01 0.00 0.03
MgO 1.84 0.37 0.33 0.40 0.36 0.61
P 2O 5 0.19 0.39 0.43 0.60 0.57 1.09
SO3 2.87 0.00 0.00 0.04 0.00 0.00
LOI 1.60 14.26 24.73 19.51 17.69 2.35
Calcination Tº (ºC) - 450–500 800–1000 650 700 750
C 3S 66.47 - - - - -
C 2S 17.35 - - - - -

3
 C 3A 5.54 - - - - -
C4AF 3.37 - - - - -
Sp. gravity (g/cm3) 3.122 2.348 2.336 2.492 2.861 2.382

2.1 Chemical characterization

2.1.1 X-Ray Diffraction and Rietveld refinement (QXRD)

QXRD analysis was performed to determine the mineral composition of the RHA
samples. Testing was performed between 10 and 75º 2𝜃 at a 0.02º per second sampling
rate. Commercially available software (EVA and TOPAS) was used for phase
identification and Rietveld refinement. Fluorite was selected as the internal standard to
consider the amorphous fraction and correct the Rietveld results, based on the
procedure proposed by Chancey et al. [33]. Because this method allows for an
estimation of the total amount of amorphous materials in the sample, XRF results can be
used in combination with QXRD to estimate the amount of amorphous silica (SiO2-
Amorphous) in the different RHA samples. The calculation is performed by subtracting the
sum of the amounts of crystalline silica polymorphs identified (SiO2-Crystalline) from the total
abundance of SiO2 in each of the samples determined by XRF, according to Eq. (1):

SiO'()*+,-.+/0 = SiO'(234 − SiO'(6,7089::;<= (1)

2.1.2 Fourier transform infrared spectroscopy (FTIR)

The FTIR spectra were obtained using a Shimadzu IRAffinity spectrophotometer, with
the RHA samples prepared by mixing 1 mg of each one with 3000 mg of KBr. Spectral
analysis was conducted at a resolution of 6 cm-1, over the range between 4000 and 400
cm-1. A total of 45 measurements were made for each of the prepared samples.

2.2 Physical characterization

2.2.1 Particle size distribution (PSD)

Particle size distributions (PSD) of the five RHA samples considered and OPC were
measured using a Malvern Mastersizer 2000 laser diffractometer. Iso-propanol
(refractive index 1.378) was used as a dispersant, and the PSD was measured for 20 s
while stirring at 2700 rpm. For PSD calculations, 1.544 and 1.68 were considered
refractive indexes for the RHA samples and OPC, respectively. It should be noted that
PSD analysis neglects the mesoporous structure of RHA. It has been shown that
grinding RHA to finer PSD does not necessarily imply higher reactivity [30], mainly owing
to the destruction of its mesoporous structure. Thus, RHAs with finer PSD do not
necessarily imply higher reactivity of the ashes [25].

2.2.2 BET specific surface area and mesopore volume analysis

4
The specific surface area of the RHA samples was determined by the multipoint
Brunauer-Emmett-Teller (BET) method using a Quantachrome Autosorb-iQ analyser.
Nitrogen was used as the adsorbate, and the outgas time was fixed at 16 h at 25ºC. The
mesopore volume was measured by the density functional theory method (DFT) [34]
using the same equipment. Particle pore volume measurements were collected for pore
diameters between 17 and 78 nm. The outgas conditions were the same as those used
for BET analysis.

The porous nature of the RHA samples was qualitatively observed using a scanning
electron microscope (SEM). Backscatter electron (BSE) images were acquired on gold-
coated specimens, using a 15 kV accelerating voltage.

2.3 RHA strength development and hydration on cement paste

To provide a more exact comparison of results, mixture proportioning was designed on a

volumetric basis, maintaining a constant volume fraction of water and
solidswater/cementitious materials (OPC and RHA) ratio of 0.5 [35,36]. A 100% OPC
mixture with a water-to-cementitious materials (w/cm) ratio by weight of 0.5 was selected
as a reference for volume fraction proportioning. The volumetric adjustment was carried
out by adjusting the volume of powder after the 20% weight replacement of OPC by
RHA to the volume of powders in the reference mixture. This procedure allows
avoidance of the confounding effect of a mass-based replacement between chemical
change in the RHA and initial porosity, given by water volume fraction [35].

Compressive strength and early-age hydration tests were conducted on cement paste
samples with 0% and 20% RHA replacement levels (p) by weight. Cubic specimens with
2 cm sides were used to assess compressive strength. Specimens were removed from
moulds after one day of curing and stored in sealed bags in a chamber at 20 ± 3ºC and
95% relative humidity (RH) until the time of testing. Compressive strength tests were
conducted at 3, 7, 28 and 90 days on 3 cubic specimens at each age, with a constant
1800 N/s load rate in all cases.

Isothermal calorimetry tests were conducted to assess the effect of the different RHAs
considered on early hydration of a cementitious matrix. The measurements were
conducted using a TAM Air isothermal calorimeter at 23ºC. Tests were conducted on
cement paste samples with the same proportioning as the samples used in compressive
strength tests and were kept in the calorimeter up to 7 days.

5
3 Results and discussion

3.1 Chemical characterization

3.1.1 X-ray diffraction and Rietveld refinement (QXRD)

XRD patterns obtained for all RHA samples are shown on Figure 1. It can be seen that
the only crystalline silica polymorph identified was cristoballite and thus corresponds to
the total abundance of crystalline silica (SiO2-Crystalline) in all samples. All patterns exhibit a
pronounced background halo or broad hump between 20 and 30º 2 𝜃 , which is an
indication of the prominent amorphous nature of these materials [37,38].

Cb F Cb Cb
F - Fluorite Cb
(PDF 77-2093) F
Cb - Cristobalite F
(PDF 77-1317)

RHA-05

RHA-04

RHA-03

RHA-02

RHA-01

5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
2q degrees Co-Ka radiation
Figure 1. XRD patterns of the RHA samples studied. Fluorite corresponds to the internal
standard selected.

Cristoballite contents determined using Rietveld refinement and corrected using the
fluorite internal standard, along with calculated amorphous silica contents, are shown in
Table 2. Refinement statistical parameters, goodness of fit (GOF) and Rexp are also
reported for all samples. GOF values close to 1.0 are indicative of good agreement
between computed and experimental XRD patterns.

6
Table 2. Rietveld refinement results, statistical refinement parameters and computed
amorphous silica content of RHA samples.
Cristoballite* Amorphous’’ GOF Rexp SiO2-Amorphous
(% by wt.) (% by wt.) (% by wt.)
RHA-1 25.05 74.95 1.1 11.8 58.17
RHA-2 30.42 69.58 1.1 11.5 42.70
RHA-3 40.66 59.34 1.1 12.2 38.12
RHA-4 6.83 93.17 1.0 11.6 73.18
RHA-5 8.13 91.87 1.0 12.2 85.46
* = Corrected using fluorite I.S.
’’ = Computed as (100 – crystalline forms)

3.1.2 Fourier transform infrared spectroscopy (FTIR)

FTIR spectra obtained for the five RHA samples studied are shown in Figure 2, in which
main peaks in the explored 4000 to 400 cm-1 range are identified.

RHA-05 1630
614

798

620 467
1091

1597
799

RHA-04

621 471
1095
1631 795
RHA-03
477

619
1627 1096 797
RHA-02
474

621
1590
796
RHA-01 1099

474

1096

4000 3500 3000 2500 2000 1500 1000 500

Wavenumber (cm-1)
Figure 2. FTIR spectra of RHA samples studied. Main peaks are identified on each plot,
and broadening of the band between 1200 and 950 cm-1 is observed.

The bands between 467 and 477 cm-1 are assigned to Si–O–Si bending vibration modes
[39]. It is observed that this band is more pronounced in RHA samples with higher silica

7
contents. The bands observed at 614–621 cm-1 are attributed to deformation vibrations
of Si–O–Si bonds [40]. The symmetric stretching band observed between 796 and 799
cm-1 is assigned to the crystalline polymorphs present in the samples, as identified by
XRD (Figure 1). The bands between 1200 and 950 cm-1 are associated with the
asymmetric vibration mode of Si–O–T bonds (T: tetrahedral Si or Al) [41,42]. Reduction
in the wavenumber of this band is associated with reduced crosslinking levels of the
amorphous silica phase [39], induced by increased calcium contents (Table 1).
Furthermore, the broadening of this band is attributed to structural rearrangement of Si–
O–T bonds and increase in crosslinking [43]. The broadening is observed more clearly in
RHA-4 and RHA-5, which also exhibit the highest amounts of amorphous silica
according to QXRD results (Table 2). Bands between 1590 and 1631 cm-1 are identified
as deformation vibration modes of O–H bonds [40], attributable to the water retained in
the RHA mesoporous structure at ambient conditions.

Since FTIR peak intensity are not only affected by the abundance of a particular bond
but also due to differences in particle size and shape [44], the results should be
considered qualitative as no effort was made to eliminate these external contributions.

3.2 Physical characterization

3.2.1 Particle size distribution (PSD)

PSD analysis results for OPC and the five RHA samples are shown on Figure 3. It can
be seen that RHA-4 exhibits finer particles than the other RHAs, but all ashes are
coarser than the OPC employed. It is expected that RHA-4 would exhibit enhanced
activity, owing to a greater specific surface availability for hydration product nucleation
(filler effect) [29] and increased chemical activity due to higher surface available to react.
D10, D50 and D90 percentile diameters of each RHA sample and OPC are summarized in
Table 3.

7 7 100 100
a) 90 b) 90
6 OPC 6
RHA 1 80 80
5 RHA 2 5 70 70
Cumulative (%)
Differential (%)

RHA 3
60 60
4 RHA 4 4
RHA 5 50 50
3 3
40 40

2 2 30 30
20 20
1 1
10 10
0 0 0 0

0.1 1 10 100 1000 0.1 1 10 100 1000

Particle Size (µm) Particle Size (µm)
Figure 3. Particle size distribution (PSD) of OPC and RHA samples a) Differential
curves and b) cumulative PSD. All samples are coarser than OPC, with RHA-4 being the
finest of the studied RHA samples.

8
3.2.2 BET specific surface area and mesopore volume analysis

The specific surface areas (SSA) determined by multipoint BET analysis on the five RHA
samples studied are summarized in Table 3, along with percentiles determined by PSD
analysis. It can be seen that in the case of the RHA-4 sample, a finer PSD correlates
with a higher SSA measured by BET. However, in the case of RHA-5, which has a PSD
similar to that of RHA-1 and RHA-2 samples, the significantly higher SSA observed is
attributed solely to a difference in the mesoporous structure of the particles.
Furthermore, it exhibits a considerably lower LOI (Table 1) than the other samples,
which might indicate an emptier mesoporous structure because most of the organic
compounds were removed on ignition. This is not observed in RHA-4, which has a
higher LOI but a considerably finer PSD, which can explain its increase in SSA. RHA-3
exhibits the lowest SSA among all ashes. By inspection of Table 1 data, it can be
observed that RHA-3 exhibits a relatively high LOI, which suggests a higher residual
carbon content on the calcined ashes.

Table 3. Particle size distribution (PSD) percentiles and specific surface area (SSA)
determined by multipoint BET.
PSD Analysis BET Analysis
2
D10 (𝜇m) D50 (𝜇m) D90 (𝜇m) SSA (m /g)
RHA-1 4.103 20.644 55.022 52.114
RHA-2 3.759 19.123 56.407 31.284
RHA-3 4.126 19.626 58.078 23.582
RHA-4 3.128 14.467 42.530 128.850
RHA-5 4.464 20.953 56.168 114.523
OPC 1.444 11.719 34.728 -

Figure 4a shows the cumulative pore volume over a 1.8–78 nm range of pores analysed,
which covers part of the macroporous (diameter > 50 nm) and mesoporous (2 nm <
diameter < 50 nm) ranges [45]. Figure 4b reflects the accumulation of the SSA with
increasing pore diameter. The significant increase in pore volume over the range
analysed shows that RHA has a prominent mesoporous nature. RHA-4 and RHA-5 show
the highest increases in pore volume, which is consistent with the highest SSA values
measured by multipoint BET (Table 3). The steep increase in pore volume observed in
RHA-5 may explain why, despite its similar PSD to other RHA samples such as RHA-2,
it exhibits a much higher SSA.

Furthermore, Figure 4b shows that this increase in pore volume is reflected in an

increase in SSA, particularly in the range between 1.8 and 20 nm of pore diameter. This
is observed in a higher increase rate in SSA and pore volume in Figure 4a and 4b,
respectively. This aligns with results found by other researchers, in which the high SSA
of RHA is attributed mainly to its mesoporous structure [26–28,45].

9
 0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80

0.11 0.11 140 140

0.1 a) 0.1 b)
Cumulative Pore Volume (cm3 / g)

Cumulative Surface Area (m2 / g)

120 120
0.09 0.09
RHA-1
0.08 RHA-2 0.08 100 100
RHA-3
0.07 RHA-4 0.07
RHA-5 80 80
0.06 0.06
0.05 0.05
60 60
0.04 0.04
0.03 0.03 40 40

0.02 0.02
20 20
0.01 0.01
0 0 0 0

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Particle Pore Diameter (nm) Particle Pore Diameter (nm)
Figure 4. a) Cumulative pore volume of RHA particles, showing the mesoporous
character of the material b) Cumulative surface area of RHA samples. It is observed that
RHA-4 and RHA-5 exhibit a considerably higher pore volume at the mesopore scale that
explains their higher SSA.

The porous nature of the RHA particles is evident in the BSE images at 4000×
magnification (see Figure 5). These micrographs allow a morphological observation of
the porosity present in RHA particles. From inspection and comparison with the provided
scale bars, it can be observed that most of the observed particle porosity lies in the
meso and macro porous range. However, as only a limited amount of images was
collected for each sample, no conclusion on pore volume and size distribution can be
obtained from the micrographs. Thus, they should only be considered as a qualitative
and referential evidence of the mesoporous nature of RHA particles.

a) RHA-1

10
 b) RHA-2

c) RHA-3

d) RHA-4

11
 e) RHA-5
Figure 5. a) to e) SEM micrographs of RHA samples, 4000× magnification, 15 kV, BSE.
Particle pores at the macro- and mesoscales are observed in all samples, some of them
being pointed by arrows.

3.3 Compressive strength and isothermal calorimetry

3.3.1 Compressive strength

The compressive strength at 3 days does not exhibit significant differences between
100% OPC and the RHA mixtures (see Figure 6). However, at 7 days, the RHA-4
mixture shows a strength value that is significantly higher than the rest of the mixtures
(11.7% above 100% OPC average strength), which is consistent with its high SiO2-
Amorphous (Table 2) content and the RHA having the highest SSA (Table 3). At 28 days of
age, RHA-4 and RHA-5 are the mixtures exhibiting the highest strength values, which
are consistent with the combined influence of high SSA and SiO2-Amorphous available to
react. At 90 days, RHA-4 and RHA-5 mixtures also showed strength values that are
statistically higher than those of OPC (10.5% and 7.9% above 100% OPC average
strength, respectively), which explains the relevance of the mesoporous structure of
RHA on the expected reactivity and strength contribution of RHA as an SCM [27], in
combination with high amounts of amorphous silica phases [24]. The lowest mechanical
performance was measured on RHA-3 mixtures, which exhibit the lowest SSA and SiO2-
Amorphous of all ash samples studied.

12
 0 14 28 42 56 70 84 98

70 70

Compressive Strength (MPa)

60 60

50 50

OPC
40 40
RHA-1
RHA-2
RHA-3
30 RHA-4 30
RHA-5

20 20

0 14 28 42 56 70 84 98
Age (days)
Figure 6. Compressive strength results of tests conducted on w/cm 0.5 paste samples
at 3, 7, 28 and 90 days. Samples with higher SSA and amorphous silica contents exhibit
the best results.

3.3.2 Isothermal calorimetry

A strong increase in the heat release (activation) of the hydration reactions is observed
on RHA-1, RHA-4 and RHA-5 mixtures compared with OPC and the other two RHA
mixtures (See Figure 7). This is accompanied by an acceleration of the hydration
reactions on RHA-4 and RHA-5, which is attributed to their higher SSA values, which
can directly contribute to an enhanced physical interaction (filler effect) with cement at
early ages [3]. At 3 days (72 h), RHA-4 and RHA-5 exhibit similar levels of total heat
release (Figure 7b), whereas at 7 days (168 h), RHA-4 reached a higher total heat value
than all other samples. This is in good agreement with compressive strength results, in
which similar strength results at 3 days were observed for RHA-4 and RHA-5, whereas
at 7 days, RHA-4 overcomes all other samples.

RHA-2 and RHA-3 mixtures do not show evidence of significant activation, and a small
degree of retardation on hydration is also observed (Figure 7a). These samples have the
lowest heat released at 3 and 7 days (72 and 168 h) of the five RHA samples considered
in the study (Figure 7b). Furthermore, these mixtures exhibited lower compressive
strength results at 3 and 7 days.

The strong correlation observed between compressive strength and isothermal

calorimetry results is graphically shown in Figure 9. It can be seen that at 3 days (Figure
8a), the total heat normalized by the volume of available water is linearly related with the
compressive strength results for the five RHA samples considered. The water volume
fraction represents the initial porosity of the mixture [35]; thus, hydration products must
fill this porosity to develop strength over time. This analytical approach allows the
differentiation of expected strength based on heat release of systems with different

13
w/cm. The same linear correlation behaviour is observed at 7 days (Figure 8b). This
correlation is based on the fact that strength development is strongly related to the
degree of hydration of the mixture [1,46], which has been found to have a relatively
linear correlation with the total heat released by the mixture [47–49]. Therefore, the total
heat release of a given mixture can be considered a relative estimate of the compressive
strength at a certain age [50].
0 6 12 18 24 30 36 42 48

0.0045 0.0045

0.004 a) 0.004
Normalized Heat Flow (W / g OPC)

OPC
RHA-1
0.0035 RHA-2 0.0035
RHA-3
0.003 0.003
RHA-4
0.0025 RHA-5 0.0025

0.002 0.002

0.0015 0.0015

0.001 0.001

0.0005 0.0005

0 0

0 6 12 18 24 30 36 42 48
Time (hours)
0 24 48 72 96 120 144 168

400 400
b)
350 350
Normalized Heat (J / g OPC)

300 300

250 250

200 200

150 150

100 100

50 50

0 0

0 24 48 72 96 120 144 168

Time (hours)
Figure 7. Isothermal calorimetry results for 100% OPC and 20% OPC-by-RHA
replacement mixtures a) Normalized heat flow plots b) Normalized total heat plots, with
indications of total heat released at 3 and 7 days. High activation of RHA-4 and RHA-5 is
observed, which is consistent with compressive strength results.

14
 475 480 485 490 495 500 505 510 515 585 590 595 600 605 610 615 620 625 630

27 27 40 40
a) 3 days b) 7 days RHA-4

38 38
RHA-5

Compressive Strength (MPa)

Compressive Strength (MPa)

26 26
RHA-4 36 36

RHA-1 34 RHA-1 34
25 25 RHA-5

32 RHA-2 32

24 24
RHA-2 30 30

28 28
23 23
R2=0.9868 26 RHA-3 R2=0.9611 26
RHA-3
22 22 24 24

475 480 485 490 495 500 505 510 515 585 590 595 600 605 610 615 620 625 630
Normalized Heat (J / mL water) Normalized Heat (J / mL water)
Figure 8. Heat release normalized by water volume versus compressive strength at a) 3
days age b) 7 days age. A linear correlation between heat per volume of water and
strength is observed.

Isothermal calorimetry can be used as a reliable tool to estimate mechanical

development over time (Figure 8) as hydration reaction progress measured as heat
release. This offers the advantage of a continuous measurement over a certain time
interval compared with the discrete point measurements that can be achieved with
conventional compressive strength tests.

3.4 Quantification of filler (physical) and pozzolanic (chemical) effects on hydration

Based on the stoichiometric composition of the C–S–H gel produced by RHA pozzolanic
reaction, Jamil et al. [51] established a molar S/CH ratio of 0.54054 on the formed gel.
Furthermore, the hydration reactions of alite (C3S) and belite (C2S) phases can be used
to compute the amount of CH produced per gram of hydrated OPC (CHOPC) [1], based
on phase composition analysis. Using the composition of the employed OPC given in
Table 1, it can be determined that CHOPC equals 0.3609 (grams of CH per gram of OPC).
If impurities in RHA are considered and the actual SiO2-Amorphous content is considered in
the material’s pozzolanic potential, the maximum replacement level (pmax) that is able to
react owing to CH availability constraints can be computed [51] using Eq. (2):

𝑆/𝐶𝐻 ∙ 𝐶𝐻JKL
𝑝BCD = (2)
SiO'()*+,-.+/0 + 𝑆/𝐶𝐻 ∙ 𝐶𝐻JKL

The replacement level (p) selected in this study (20% by weight) is (RHA samples 1-to-
4) below the computed pmax value in 4 cases (25.14%, 31.36%, 33.85% and 21.04% for
RHA 1-to-4 respectively). With regard to RHA-5, in which the computed pmax (18.59%) is
below the dosage level used, it is still expected that complete pozzolanic reaction of the
CH will not occur in the duration of testing (90 days) because of accessibility. CH has

15
been detected in previous studies at this age in mixtures with up to 30% replacement of
OPC-by-RHA [7]. Based on this conceptualization of RHA reaction, the effective silica
content (SiO2-Eff) of RHA may be defined as the amount of amorphous SiO2 that is able
to react, given the replacement level p, the maximum replacement level pmax and the
amorphous silica content of RHA (SiO2-Amorphous) as shown in Eq. (3):

𝑀𝑖𝑛(𝑝, 𝑝BCD ) ∙ SiO'()*+,-.+/0

SiO'(OPP = (3)
1 − 𝑀𝑖𝑛(𝑝, 𝑝BCD )

The relation between SiO2-Eff contents computed using Eq. (3) is shown in Figure 9a and
is indicative of the pozzolanic effect potential of the RHA samples and the heat released
per millilitre of water at 1, 2, 3, and 7 days (24, 48, 72 and 168 h) of hydration. It can be
seen that there is a linear correlation between the effective silica content and the heat
released at a certain time. The latter is also indicative of the compressive strength of the
samples, as shown in Figure 8.

Cyr et al. [29] found that different values of SSA do not have the same effects on
compressive strength. A small amount of particles have a higher impact on strength,
whereas the use of higher amounts has a smaller effect. SSA determined by BET
showed to be the best technique to assess the physical characteristics of RHA. PSD
determined by laser diffractometry neglects the internal porosity of the particles and
therefore underestimates the available surface, losing reliability to accurately predict the
physical effect of this material.

Analogous to the SiO2-Eff concept presented, they defined an effective SSA (SSAEff)
parameter that accounts for replacement level p and the efficiency of the physical
contribution:

𝑝
SSA OPP = SSA ∙ ∙ 𝜉(𝑝) (4)
1−𝑝

where 𝜉(𝑝) represents the efficiency function as defined in [29].

Figure 9b shows the relation between SSAEff, which is indicative of the physical (filler)
effect potential of the RHA samples and the heat released per millilitre of water at 1, 2, 3,
and 7 days (24, 48, 72 and 168 h) of hydration. A logarithmic fit was found between
SSAEff and the heat released at a certain time. This logarithmic correlation has been
observed previously by comparing compressive strength with fineness of different fly
ashes [29] and inert quartz powders [52]. The relationship is attributed to the action of
coupled effects as additional surface is introduced by the RHA: first, the hydration of the
OPC is enhanced as additional nucleation sites and particle spacing are provided [3] and
second, the RHA reaction is accelerated because more surface is exposed to the CH
generated by OPC hydration.

The logarithmic correlation between SSAEff and heat release per millilitre of water (or
compressive strength) may be physically interpreted by considering the increase in SSA
at constant replacement rate as an increment in the amount of RHA particles by unit
volume of material. The RHA particles are able to influence cement hydration only if
cement grains and RHA particles are close enough to interact. When the amount of

16
particles is small (low SSA at constant replacement rate), the probability of the particles
being close to the cement grains is higher. As the particle content increases (increase
SSA at constant replacement rate), some of the RHA particles may be isolated from
cement grains and therefore, the effect of an increment in SSAEff is smaller than at lower
SSA levels [29]. This effect is coupled with the non-linear increase in the pozzolanic
activity of RHA. As SSA of the RHA increase, the rate of reaction is also expected to
increase as additional surface is in contact with CH and water. However, as only RHA
and not OPC is increasing its surface, the rate increase may be limited by CH availability
at some point of the early hydration, consequently reducing the contribution of additional
SSA in RHA.

24h 72h
48h 168h
6 9 12 15 18 21 24 0 5 10 15 20 25 30 35
a) b)
650 650 650 650
RHA-1 RHA-5 RHA-1 RHA-5

Normalized Heat (J / mL water)

Normalized Heat (J / mL water)

RHA-3 RHA-4 RHA-3 RHA-4

600 RHA-2
R2 0.842 600 600 RHA-2 R2 0.917 600

550 550 550 550

RHA-5 RHA-5
RHA-1 RHA-1
500 RHA-3 RHA-4 R2 0.972 500 500 RHA-3 RHA-4 500
RHA-2 RHA-2 R2 0.983
RHA-5 RHA-5
450 RHA-1 450 450 RHA-1 450
RHA-3 RHA-4 RHA-3 RHA-4
R2 0.979
RHA-2 RHA-2 R2 0.965
400 400 400 400
RHA-5 RHA-5
RHA-1 RHA-1
350 RHA-3
350 350 RHA-3
350
RHA-4 R2 0.967 RHA-4
RHA-2 RHA-2 R 0.886
2
300 300 300 300

6 9 12 15 18 21 24 0 5 10 15 20 25 30 35
Effective SiO2-Amorphous (% by wt.) Effective Specific Surface Area (m2 / g)
Figure 9. Correlation between heat released per millilitre of water at 24, 48, 72 and 168
h and a) Amorphous silica content and b) SSA. Linear and logarithmic correlations are
observed between effective amorphous silica and SSA and normalized heat,
respectively.

3.5 Rice husk ash reactivity model

The physical and chemical contributions of RHA on strength can be identified and
coupled based on the effective silica and SSA concepts, as shown in Figure 10. ∆f[;:/8;+<
represents the base strength of the OPC, diluted by the SCM replacement. ∆f\ refers to
the contribution on strength of the SCM particle to OPC hydration by providing additional
nucleation sites for hydration products precipitation. ∆f-] represents the contribution on
strength of pozzolanic reaction of the SCM and CH, which generates additional
hydration products.

17
 (Fig. 13) and g( p) is the efficiency 3.3. Validation of the model for other cement and mineral
admixtures
q. (4), the efficient area of a mineral
0 in three cases: when the replacement Fig. 19 summarizes the decoupling process of the
( p Y 0) or close to 100% (n( p ) / empirical model to separate the physical and chemical
g. 14), or for coarse powders for which effects of mineral admixtures in mortars.
areas S S are very low. This synthetic chart distinguishes:
of the efficiency concept to the exper- Chemical Effect Physical Effect
n in Figs. 15 and 16 for inert (quartz) and – the dilution effect caused by the variation of the water/
powders, respectively. The increase(1)! inf : pozzolanic effect ratio, which (1)can
cement !fdilution
be :evaluated
dilution effectby using Powers’
h as a function of the Strength
efficient surface
pz
law;
Contributions (1)
d by a series of single curves that only – the physical and chemical !f" effects,: nucleation
whicheffectare taken into
on the age of the mortar. The increase of account in very similar ways. For each of them, the
when the efficient surface area is near 0, efficient surface area S eff is calculated from the
case for which the strengths of mortars specific surface area of the admixture, S s, and its
neral admixtures areStrength of
! + ! +!
fRHA = fdilution pz
the same. It can be replacement ratio, fp, byf means of the efficiency factor
Blended Mix
significant increases in strength are (Fig. "
14). The contribution of the physical or
est admixture (Q2 and FAC5) and for pozzolanic effects derives from a simple equation in
of 25%. which two coefficients out of three are identified either
be proposed (Eq. (5)) to express the as a constant (c = 1) or as the specific surface area of
(Df B or Df pz) depending
RHA SiO2-Amorphous : amorphous
on the efficient the cement itself (b). SSAThe : specific surface
third area
coefficient (a) is time
Characterization
silica content
dependent and characterizes the interaction between
the cement and the admixture (Fig. 18). For the case
!c ð5Þ of the physical effect, the curve a(t) looks like a
Min(p, pmaxtypical
) hydration (1)curve (e.g., p compressive strength
Effective SiO2-Eff = SSAEff = SSA · · ξ(p)
over 2-Amorphous
· SiOtime). 1-p
ff Contribution 1-Min(p, pmax)
re empirical parameters; of course this
pozzolanic
chemical
pozzolaniceffect
effect
effect
nly when the specific surface area of 70
70
(1) a pz
s greater than the specific surface area 60
a
Relative Effect Factor

60
∆ f ϕ , pz ( t ) = c
50
50 b
rizes the curves Relative effect for ages
calculated 1+
over time 40
40 S eff
days and Fig. 18 gives the parameters a 30 physical effect
physical effect
30
aϕ
by means of least squares analysis. For
nditions of this study, it was found that b 20
20 specific
• b : 280 surface
nt values: b was close to the specific 10
10 of cement
nt and c was equal to 1. Only parameter 00 • c :1
me of hydration of the mortars. It can be 0
0
30
30 60
60
90
90 120
120 150
150 180
180
Hydration time (days)
at the pozzolanic effect goes beyond the Hydration time (days)
28 days. Figure 10. Chemical and physical effect Fig. 18. Numerical values of parameters of Eq.
coupling conceptualization and their relation (5).

with RHA characterization parameters. (1) Notation, equation and figure taken from [29].

Based on the findings of this particular study, a reactivity index (RI) is proposed as a
means to couple the RHA’s chemical and physical characterization on a unique
parameter, as described in Eq. (5):

log (SSA cOd )

𝑅𝐼 = SiO'(OPP ∙ A + (5)
A

where SiO2-Eff is the effective amorphous silica content (%), SSAEff is the effective
specific surface area determined by BET analysis (m2/g), and A is the hydration time
(days). The increasing relative contribution of the chemical composition of the ash on
compressive strength over time is represented by multiplying SiO2-Eff by hydration time in
accordance with previous findings [29] (see Figure 10a). Similarly, the decreasing
relative contribution of the physical contribution of the ash on compressive strength over
time is represented by dividing the logarithm of SSA by hydration time (see Figure 10b).

Based on the chemical and physical characterization of the five RHA samples studied,
RI was calculated at 1, 2, 3, 7, 28 and 90 days of age to provide performance
comparisons at early and later ages. Considering the good correlation found between

18
volumetric heat release and compressive strength (see Figure 8), the computed RI value
was compared with isothermal calorimetry results for early ages (1-2-3 and 7 days) and
compressive strength results for later ages (28 and 90 days). Correlation plots for each
age of analysis are presented in Figure 11.

1.5 2 2.5 3 3.5 4 0.5 1 1.5 2 2.5 0.5 0.75 1 1.25 1.5 1.75 2

355 355 450 450 515 515

a) 1 day b) 2 days c) 3 days

Normalized Heat (J / mL water)

Normalized Heat (J / mL water)

Normalized Heat (J / mL water)

350 350 445 445 510 510
505 505
345 345 440 440
500 500
340 340 435 435
495 495
335 335 430 430
490 490
330 330 425 425
485 485
325 325 420 420 480 480
R2 = 0.892 R2 = 0.972 R2 = 0.986
320 320 415 415 475 475

1.5 2 2.5 3 3.5 4 0.5 1 1.5 2 2.5 0.5 0.75 1 1.25 1.5 1.75 2
Reactivity Index RI Reactivity Index RI Reactivity Index RI
0.5 0.75 1 1.25 1.5 1.75 2 2 3 4 5 6 8 10 12 14 16 18

625 625 53 53
d) e) 66 f) 66
Normalized Heat (J / mL water)

7 days 52 28 days 52 90 days

Compressive Strength (MPa)

Compressive Strength (MPa)

620 620
51 51 64 64
615 615
50 50
62 62
610 610 49 49
605 605 48 48 60 60
47 47 58 58
600 600
46 46
595 595 56 56
R2 = 0.867 45 R2 = 0.900 45 R2 = 0.880
590 590 44 44 54 54

0.5 0.75 1 1.25 1.5 1.75 2 2 3 4 5 6 8 10 12 14 16 18

Reactivity Index RI Reactivity Index RI Reactivity Index RI
Figure 11. Correlation plots between computed RI and normalized heat released (a–d)
or compressive strength (e–f). Good correlations are observed between computed RI
and normalized heat and strength.

A good agreement between RI and either normalized heat released or compressive

strength can be observed at all ages of analysis; this is an indicator of the RI parameter
potential as a tool to assess RHA effectiveness as SCM considering both chemical and
physical aspects. The proposed approach offers some advantages over conventional
strength activity measures, such as the k-values (or efficiency factors), because it
explicitly considers the chemical and physical properties of RHA. k-values are defined as
the coefficient used to weight the contribution of the SCM on strength in relation to the
contribution of a unit mass of OPC. Therefore, k-values below 1.0 represents a
contribution to strength per mass of material lower than OPC and values higher than 1.0
would imply a higher contribution of the SCM compared to OPC. Furthermore, k-values
are derived uniquely from compressive strength results [24] and thus do not allow
assessment of the effect of additional grinding and/or chemical modifications on the
mechanical performance of a mixture.

The numerical difference between k-values and RI is explained by the fact that k-values
associate compressive strength as a relative measure to OPC performance, whereas RI
was computed based on physical and chemical properties of RHA. RI is correlated
linearly with compressive strength or heat of hydration, as observed in Figure 11. This
suggests that a fundamental relationship between RI and strength may be described,

19
extending the concept to OPC. Whereas RI was defined specifically for a given RHA
based on amorphous silica content and SSA, an equivalent RI for OPC (ERIOPC) may be
computed for the 100% OPC mixture using the linear regressions shown in Figure 11
and compressive strength results of the control mixture. To determine ERIOPC, the
strength (or volumetric heat release) of the 100% control mixture can be used to enter to
one of the plots shown in Figure 11 (depending on the age of interest) and read a
corresponding RI value on the horizontal axis. Using the computed ERIOPC for each age
of analysis, the RI parameter of each RHA sample can be normalized by this measure of
OPC performance as shown in Eq. (6):

𝑅𝐼ghi
ϕ = (6)
𝐸𝑅𝐼JKL

where ϕ is the normalized reactivity factor. Both w/cm and w/c ratios (by mass) are
commonly used parameters to describe strength of cement-based materials. However, it
can be shown that the w/cm ratio by itself fails to account for differences in physical and
chemical interactions between RHA samples. Because the proportioning was made on a
volumetric basis, w/cm is slightly different for each mixture, which clearly does not
explain the variation in compressive strength as shown by the low R2 values in Figure
12a.

3 days 28 days
7 days 90 days

a) 0.505 0.51 0.515 0.52 0.525 0.53 0.535 b) 0.46 0.48 0.5 0.52 0.54 0.56 0.58

70 70 70 70
Compressive Strength (MPa)

Compressive Strength (MPa)

65 65 65 65
60
R2=0.135 60 60 R2=0.854 60
55 55 55 55
50 50 50 50
R2=0.062 R2=0.803
45 45 45 45
40 40 40 40
35 35 35 R2=0.743 35
R2=0.224
30 30 30 30
25 25 25 R2=0.814 25
R2=0.090
20 20 20 20

0.505 0.51 0.515 0.52 0.525 0.53 0.535 0.46 0.48 0.5 0.52 0.54 0.56 0.58
w/cm ratio w/cEQ ratio
Figure 12. Scatterplot between compressive strengths at 3, 7, 28 and 90 days and (a)
w/cm ratio and (b) w/cEQ ratio. The correlation with compressive strength improves
significantly with the equivalent w/cm approach compared with w/cm.

The correlation between compressive strength and w/cEQ is significantly improved as

shown by the R2 values in Figure 12b versus those in Figure 12a. This is attributed to ϕ-
values correctly representing the equivalent effect of each RHA sample compared with
OPC, as seen in Eq. (7) for w/cEQ,

20
 w/cOn = w/(OPCq+: + ϕ ⋅ RHA q+: ) (7)

where OPCVol and RHAVol are the weight fractions of OPC and RHA, respectively,
corrected by volume of solids.

The w/cEQ can be used for RHA blended mixture design based on strength specification
using widely adopted models such as the Bolomey and Abrams equations. This can be
observed as the w/cEQ parameter significantly improves the correlation between Bolomey
exponential model (shown in dashed lines in Figures 12b) compared to the same
regression computed using the w/cm determined by mixture proportioning (Figure 12a).
Therefore, the mathematical operation with ϕ-values is analogue to k-values (ϕ > 1
represents a higher contribution to strength than OPC by unit mass, and ϕ < 1 a lower
contribution). The relationship between ϕ-values and k-values (determined using the
methodology proposed in [24]) can be seen in Figure 13, were a relatively good
agreement between these two parameters was found.

0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6
1.6 1.6
1.5 1.5
R2=0.751
1.4 1.4
Normalized reactivity factor (f)

1.3 1.3
1.2 1.2
1.1 1.1
1 1
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
0.5 0.5
0.4 0.4
0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6
Efficiency factor (k-values)
Figure 13. Scatterplot between ϕ-values and k-values. Good agreement is observed
between the proposed ϕ parameter and the efficiency factors determined from
compressive strength results.

In addition, ϕ-values are determined from the fundamental characteristics of the RHA,
represented by the RI parameter: amorphous silica content and SSA. They provide the
same flexibility and simplicity to predict strength at different levels of OPC-by-RHA
replacement while explicitly considering modifications of the chemical composition and
physical characteristics of the material.

The ϕ-values are not theoretically limited to forecasting on mixtures at a certain level of
replacement because they consider the CH availability on the pmax parameter, which can
be adjusted for any given OPC composition. This is a clear advantage over the
conventional k-values approach, in which forecasts are limited to replacement levels
close to the mixtures used for k-value computation, because CH availability and
differences in physical interactions at different levels of replacement may significantly
affect the accuracy of the model.

21
Because ϕ-values are independent of strength results and are derived from only the
chemical and physical characterization of the RHA sample, they assume that all effective
amorphous silica will react and contribute to strength. Therefore, the model considers
that all required constituents for pozzolanic reaction occurrence are in contact and
available. This is a simplification of the real situation in which CH carbonation,
weathering and pore depercolation may affect the validity of these assumptions. In
addition, the mathematical expression of the RI parameter considers the effect of
hydration time to be multiplicative for the chemical effect and to be divisive for the
physical effect. This is also a simplification from reality, in which this relationship is in
generally nonlinear. However, the model is presented as an initial attempt to capture the
effects of amorphous silica and specific surface on strength in a unique parameter,
derived from a generalized rationale (Figure 10). Further understanding of the nature of
the variation of these effects over time may allow the presented model to be easily
improved.

4 Conclusions

This study conducted an insightful chemical and physical characterization of five RHA
samples obtained from industrial sources. The samples were used as received, and the
feasibility of scaling up the use of this material as SCM to a production regime level is
thus assessed.

Physical characteristics are particularly relevant in RHA because its mesoporous

structure provides additional reaction surface and significantly explains its high reactivity.
Furthermore, the relations among chemical composition, physical characteristics and
expected mechanical performance was explored and compared with the current
modelling approaches found in the literature.

Based on the findings presented, the following conclusions can be stated:

1. A strong correlation between SSA and normalized heat release was observed in
all five RHAs. Based on the use of QXRD analysis and an internal standard, the
amorphous silica content was computed for each RHA. The results showed good
agreement with normalized heat release measurements and also proved to be a
reliable tool to assess the pozzolanic potential of the different RHA samples.

2. The strong correlation between compressive strength and heat release

normalized by water volume showed that isothermal calorimetry might be used
as a reliable tool to predict mechanical performance. Furthermore, calorimetry
measurements provided continuous monitoring of the sample hydration during
the time of testing and thus provided significant advantages compared with
conventional strength testing in terms of the amount of data collected.

3. The RI parameter is proposed based on the relationship between amorphous

silica content and SSA with heat release and also considers the variation of their
relative contribution over time. The results showed that RI predicts with sufficient
precision the heat release and compressive strength of the RHA mixtures at early

22
 and later ages, allowing the effects of additional processing of RHA to be
estimated based on mechanical performance. RI assumes that a certain variation
of the chemical and physical effects over time can be improved based on the
generalized rationale presented and further understanding of these relationships
with time.

4. A normalized reactivity factor (ϕ-values) can be defined by dividing the RI values

of each RHA by an equivalent RI value for OPC. ϕ -values offer significant
advantages because they explicitly consider the chemical and physical
characterization of the material and explicitly account for CH availability. In
addition, it allows an equivalent w/cm ratio to be defined that is suitable for a
strength-based mixture design that accounts for reactivity differences between
RHA samples.

5. This study showed that even considering the variability of raw materials sourced
from industrial suppliers from different countries, RHA can still be considered a
promising cementitious material, based on short- and long-term mechanical
performance results. Whereas process variables such as calcination temperature
and time, grinding and chemical pre-treatment have been found to be relevant
aspects on the expected reactivity of RHA in previous studies, the considered
RHA samples were shown to retain a significant amount of reactivity potential,
even when no effort to control these factors was made. This provides a measure
of the resilience of RHA to the manufacturing process variability inherent in large-
scale production conditions.

5 Acknowledgements

The authors greatly acknowledge the insightful comments of Dale Bentz and Javier
Castro and the support provided by CEDEUS, CONICYT/FONDAP 15110020.

References

[1] P. Mehta, P. Monteiro, Concrete: Microstructure, Properties, and Materials, 3rd

ed., McGraw-Hill Professional, New York, 2005.
[2] B. Lothenbach, K. Scrivener, R.D. Hooton, Supplementary cementitious
materials, Cem. Concr. Res. 41 (2011) 1244–1256.
doi:10.1016/j.cemconres.2010.12.001.
[3] P. Lawrence, M. Cyr, E. Ringot, Mineral admixtures in mortars-- effect of inert
materials on short-term hydration, Cem. Concr. Res. 33 (2003) 1939–1947.
doi:10.1016/S0008-8846(03)00183-2.
[4] D.P. Bentz, C.F. Ferraris, S.Z. Jones, D. Lootens, F. Zunino, Limestone and
Silica Powder Replacements for Cement: Early-Age Performance, Cem. Concr.
Compos. 78 (2017) 43–56. doi:10.1016/j.cemconcomp.2017.01.001.
[5] P. Mehta, Properties of blended cements made from rice husk ash, ACI J. (1977)
440–442.

23
[6] M.H. Zhang, R. Lastra, V.M. Malhotra, Rice-husk ash paste and concrete: Some
aspects of hydration and the microstructure of the interfacial zone between the
aggregate and paste, Cem. Concr. Res. 26 (1996) 963–977.
[7] Q. Yu, K. Sawayama, S. Sugita, M. Shoya, Y. Isojima, The reaction between rice
husk ash and Ca(OH)2 solution and the nature of its product, Cem. Concr. Res.
29 (1999) 37–43. doi:10.1016/S0008-8846(98)00172-0.
[8] Food and Agriculture Organization of the U.N., FAOSTAT, 2014.
http://faostat.fao.org.
[9] G. Rodríguez de Sensale, Strength development of concrete with rice-husk ash,
Cem. Concr. Compos. 28 (2006) 158–160.
doi:10.1016/j.cemconcomp.2005.09.005.
[10] T. Chungsangunsit, S.H. Gheewala, S. Patumsawad, Environmental Assessment
of Electricity Production from Rice Husk : A Case Study in Thailand, Int. Energy J.
6 (2005).
[11] V.P. Della, I. Kühn, D. Hotza, Rice husk ash as an alternate source for active
silica production, Mater. Lett. 57 (2002) 818–821.
http://www.sciencedirect.com/science/article/B6TX9-4625R5D-
1/2/422c3984b452c26411de9fe30479494a.
[12] M.F.M. Zain, M.N. Islam, F. Mahmud, M. Jamil, Production of rice husk ash for
use in concrete as a supplementary cementitious material, Constr. Build. Mater.
25 (2011) 798–805. doi:10.1016/j.conbuildmat.2010.07.003.
[13] A.N. Givi, S.A. Rashid, F.N.A. Aziz, M.A.M. Salleh, Assessment of the effects of
rice husk ash particle size on strength, water permeability and workability of
binary blended concrete, Constr. Build. Mater. 24 (2010) 2145–2150.
doi:10.1016/j.conbuildmat.2010.04.045.
[14] M.. Khan, K. Mohan, H.F.. Taylor, Pastes of tricalcium silicate with rice husk ash,
Cem. Concr. Res. 15 (1985) 89–92.
[15] F. Giannotti, J.B.L. Liborio, P. Helene, F. Giannotti, J.B.L. Liborio, P. Helene,
Improvement of physical and chemical properties of concrete with brazilian silica
rice husk (SRH), Rev. Ing. Construcción. 23 (2008) 18–25. doi:10.4067/S0718-
50732008000100002.
[16] A.L.G. Gastaldini, G.C. Isaia, N.S. Gomes, J.E.K. Sperb, Chloride penetration
and carbonation in concrete with rice husk ash and chemical activators, Cem.
Concr. Compos. 29 (2007) 176–180. doi:10.1016/j.cemconcomp.2006.11.010.
[17] A.L.G. Gastaldini, G.C. Isaia, A.P. Saciloto, F. Missau, T.F. Hoppe, Influence of
curing time on the chloride penetration resistance of concrete containing rice husk
ash: A technical and economical feasibility study, Cem. Concr. Compos. 32
(2010) 783–793. doi:10.1016/j.cemconcomp.2010.08.001.
[18] G.R. De Sensale, Effect of rice-husk ash on durability of cementitious materials,
Cem. Concr. Compos. 32 (2010) 718–725.
doi:10.1016/j.cemconcomp.2010.07.008.
[19] V. Saraswathy, H.-W. Song, Corrosion performance of rice husk ash blended

24
 concrete, Constr. Build. Mater. 21 (2007) 1779–1784.
doi:10.1016/j.conbuildmat.2006.05.037.
[20] F. Zunino, M. Lopez, Decoupling the physical and chemical effects of
supplementary cementitious materials on strength and permeability: a multi-level
approach, Cem. Concr. Compos. 65 (2016) 19–28.
doi:10.1016/j.cemconcomp.2015.10.003.
[21] D.G. Nair, A. Fraaij, A.A.K. Klaassen, A.P.M. Kentgens, A structural investigation
relating to the pozzolanic activity of rice husk ashes, Cem. Concr. Res. 38 (2008)
861–869. doi:10.1016/j.cemconres.2007.10.004.
[22] R. Siddique, Waste Materials and By-Products in Concrete, Springer, Berlin,
2008.
[23] Q. Feng, H. Yamamichi, M. Shoya, S. Sugita, Study on the pozzolanic properties
of rice husk ash by hydrochloric acid pretreatment, Cem. Concr. Res. 34 (2004)
521–526. doi:10.1016/j.cemconres.2003.09.005.
[24] S.K. Antiohos, V.G. Papadakis, S. Tsimas, Rice husk ash (RHA) effectiveness in
cement and concrete as a function of reactive silica and fineness, Cem. Concr.
Res. 61–62 (2014) 20–27. doi:10.1016/j.cemconres.2014.04.001.
[25] G.C. Cordeiro, R.D. Toledo Filho, L.M. Tavares, E.D.M.R. Fairbairn, S. Hempel,
Influence of particle size and specific surface area on the pozzolanic activity of
residual rice husk ash, Cem. Concr. Compos. 33 (2011) 529–534.
doi:10.1016/j.cemconcomp.2011.02.005.
[26] V.T.A. Van, C. Rößler, D.D. Bui, H.M. Ludwig, Mesoporous structure and
pozzolanic reactivity of rice husk ash in cementitious system, Constr. Build. Mater.
43 (2013) 208–216. doi:10.1016/j.conbuildmat.2013.02.004.
[27] V.T.A. Van, C. Rößler, D.D. Bui, H.M. Ludwig, Pozzolanic reactivity of
mesoporous amorphous rice husk ash in portlandite solution, Constr. Build. Mater.
59 (2014) 111–119. doi:10.1016/j.conbuildmat.2014.02.046.
[28] H.T. Le, M. Kraus, K. Siewert, H.-M. Ludwig, Effect of macro-mesoporous rice
husk ash on rheological properties of mortar formulated from self-compacting high
performance concrete, Constr. Build. Mater. 80 (2015) 225–235.
doi:10.1016/j.conbuildmat.2015.01.079.
[29] M. Cyr, P. Lawrence, E. Ringot, Efficiency of mineral admixtures in mortars:
Quantification of the physical and chemical effects of fine admixtures in relation
with compressive strength, Cem. Concr. Res. 36 (2006) 264–277.
doi:10.1016/j.cemconres.2005.07.001.
[30] P. Mehta, Rice husk ash - a unique supplementary cementing material, in: Adv.
Concr. Technol. 2nd Ed. CANMET, 1994: pp. 419–44.
[31] S. Chandrasekhar, P.N. Pramada, J. Majeed, Effect of calcination temperature
and heating rate on the optical properties and reactivity of rice husk ash, J. Mater.
Sci. 41 (2006) 7926–7933. doi:10.1007/s10853-006-0859-0.
[32] R.-S. Bie, X.-F. Song, Q.-Q. Liu, X.-Y. Ji, P. Chen, Studies on effects of burning
conditions and rice husk ash (RHA) blending amount on the mechanical behavior

25
 of cement, Cem. Concr. Compos. 55 (2015) 162–168.
doi:10.1016/j.cemconcomp.2014.09.008.
[33] R.T. Chancey, P. Stutzman, M.C.G. Juenger, D.W. Fowler, Comprehensive
phase characterization of crystalline and amorphous phases of a Class F fly ash,
Cem. Concr. Res. 40 (2010) 146–156. doi:10.1016/j.cemconres.2009.08.029.
[34] J. Landers, G.Y. Gor, A. V. Neimark, Density functional theory methods for
characterization of porous materials, Colloids Surfaces A Physicochem. Eng. Asp.
437 (2013) 3–32. doi:10.1016/j.colsurfa.2013.01.007.
[35] D.P. Bentz, Activation energies of high-volume fly ash ternary blends: Hydration
and setting, Cem. Concr. Compos. 53 (2014) 214–223.
doi:10.1016/j.cemconcomp.2014.06.018.
[36] D.P. Bentz, T. Sato, I. De La Varga, W.J. Weiss, Fine limestone additions to
regulate setting in high volume fly ash mixtures, Cem. Concr. Compos. 34 (2012)
11–17. doi:10.1016/j.cemconcomp.2011.09.004.
[37] R.T. Chancey, P. Stutzman, M.C.G. Juenger, D.W. Fowler, Comprehensive
phase characterization of crystalline and amorphous phases of a Class F fly ash,
Cem. Concr. Res. 40 (2010) 146–156. doi:10.1016/j.cemconres.2009.08.029.
[38] M. Criado, A. Fernandez-Jimenez, a G. de la Torre, M. a G. Aranda, A. Palomo,
An XRD study of the effect of the SiO2/Na2O ratio on the alkali activation of fly
ash, Cem. Concr. Res. 37 (2007) 671–679.
doi:10.1016/j.cemconres.2007.01.013.
[39] I. Ismail, S. a. Bernal, J.L. Provis, R. San Nicolas, S. Hamdan, J.S.J. Van
Deventer, Modification of phase evolution in alkali-activated blast furnace slag by
the incorporation of fly ash, Cem. Concr. Compos. 45 (2014) 125–135.
doi:10.1016/j.cemconcomp.2013.09.006.
[40] I. García Lodeiro, a. Fernández-Jimenez, a. Palomo, D.E. Macphee, Effect on
fresh C-S-H gels of the simultaneous addition of alkali and aluminium, Cem.
Concr. Res. 40 (2010) 27–32. doi:10.1016/j.cemconres.2009.08.004.
[41] J. Serra, P. González, S. Liste, C. Serra, S. Chiussi, B. León, et al., FTIR and
XPS studies of bioactive silica based glasses, J. Non. Cryst. Solids. 332 (2003)
20–27. doi:10.1016/j.jnoncrysol.2003.09.013.
[42] a. M.B. Silva, C.M. Queiroz, S. Agathopoulos, R.N. Correia, M.H. V Fernandes,
J.M. Oliveira, Structure of SiO2-MgO-Na2O glasses by FTIR, Raman and 29Si
MAS NMR, J. Mol. Struct. 986 (2011) 16–21. doi:10.1016/j.molstruc.2010.11.023.
[43] C. Courteille, D. Magni, C. Deschenaux, P. Fayet, A.. Howling, C. Hollenstein,
Gas phase and particle diagnostic of HMDSO plasmas by infrared absorption
spectroscopy, in: 41st Annu. Tech. Conf. - Soc. Vac. Coaters, 1998.
[44] J.L. Rendon, C.J. Serna, IR Spectra of Powder Hematite: Effects of Particle Size
and Shape, Clay Miner. 16 (1981) 375–382. doi:10.1180/claymin.1981.016.4.06.
[45] V.C. Srivastava, I.D. Mall, I.M. Mishra, Characterization of mesoporous rice husk
ash (RHA) and adsorption kinetics of metal ions from aqueous solution onto RHA,
J. Hazard. Mater. 134 (2006) 257–267. doi:10.1016/j.jhazmat.2005.11.052.

26
 [46] A. Neville, Properties of Concrete, 4th Ed., Pearson Education Limited, Essex,
1995.
[47] J. Escalante-Garcia, Nonevaporable water from neat OPC and replacement
materials in composite cements hydrated at different temperatures, Cem. Concr.
Res. 33 (2003) 1883–1888. doi:10.1016/s0008-8846(03)00208-4.
[48] P. Lura, F. Winnefeld, S. Klemm, Simultaneous measurements of heat of
hydration and chemical shrinkage on hardening cement pastes, J. Therm. Anal.
Calorim. 101 (2010) 925–932. doi:10.1007/s10973-009-0586-2.
[49] X. Pang, D.P. Bentz, C. Meyer, G.P. Funkhouser, R. Darbe, A comparison study
of Portland cement hydration kinetics as measured by chemical shrinkage and
isothermal calorimetry, Cem. Concr. Compos. 39 (2013) 23–32.
doi:10.1016/j.cemconcomp.2013.03.007.
[50] D.P. Bentz, T. Barrett, I. De la Varga, W.J. Weiss, Relating Compressive
Strength to Heat Release in Mortars, Adv. Civ. Eng. Mater. 1 (2012) 1–16.
doi:10.1520/ACEM20120002.
[51] M. Jamil, a. B.M. a Kaish, S.N. Raman, M.F.M. Zain, Pozzolanic contribution of
rice husk ash in cementitious system, Constr. Build. Mater. 47 (2013) 588–593.
doi:10.1016/j.conbuildmat.2013.05.088.
[52] P. Lawrence, M. Cyr, E. Ringot, Mineral admixtures in mortars effect of type,
amount and fineness of fine constituents on compressive strength, Cem. Concr.
Res. 35 (2005) 1092–1105. doi:10.1016/j.cemconres.2004.07.004.

27

View publication stats