Construction and 3d Printing Concrete

Construction and Building Materials 412 (2024) 134846
Contents lists available at ScienceDirect
Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat
Optimization of formulation ratios of geopolymer mortar based on

metakaolin and biomass fly ash
Tomás Archer de Carvalho a, *, Florindo Gaspar b, Ana C. Marques c, Artur Mateus b
a
Centre for Rapid and Sustainable Product Development, Polytechnic of Leiria, 2430-028 Marinha Grande, Portugal
b
Centre for Rapid and Sustainable Product Development, School of Technology and Management, Polytechnic of Leiria, 2430-028 Marinha Grande, Portugal
c
CERENA, DEQ, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1049–001 Lisboa, Portugal
A R T I C L E I N F O A B S T R A C T
Keywords: This study investigated the factors affecting the properties of geopolymer (GP) mortars through a two-stage
Circular economy experimental program created through central composite design of experiments. Stage I focused on the effects
Geopolymerization of SiO₂/Al₂O₃, H₂O/Na₂O, and total water-to-total solids (W/S) ratios on mortars formulated using metakaolin
Geopolymer mortar
(MK) as the only geopolymerization precursor, while Stage II evaluated the impact of introducing biomass fly ash
Mix design
(BFA) as a second precursor. Most importantly, the SiO₂/Al₂O₃ ratio was defined based on raw materials’
dissolution tests rather than on their starting chemical composition, setting this study apart from most published
literature. The mortars’ properties of interest, taken as response variables, were the flow test spreading diameter,
and the efflorescence level and compressive strength after 7 days of curing under ambient conditions. The results
of Stage I, regarding GP mortars based on MK only, show that SiO₂/Al₂O₃, H₂O/Na₂O, and W/S all had a sta
tistically significant effect on compressive strength. Regarding the efflorescence level, SiO₂/Al₂O₃ had the most
significant effect, H₂O/Na₂O showed a small contribution, and W/S effect was not statistically significant. In
stage II, the incorporation of BFA led to (MK + BFA)-based mortars with lower spreading diameter and
compressive strength, and higher efflorescence levels, with % BFA being the most important factor affecting
these three response variables. Based on the developed models, appropriate combinations of factor values that
simultaneously maximize the compressive strength while maintaining the spreading diameter and efflorescence
levels within suitable limits were found: GP mortars based only on MK with SiO₂/Al₂O₃ = 4.5, H₂O/Na₂O = 19.0,
and W/S = 0.17 are expected to achieve 53 MPa of compressive strength at 7 days; (MK + BFA)-based GP
mortars with % BFA = 39.0, H₂O/Na₂O = 15.5, SiO₂/Al₂O₃ = 4.5 and W/S = 0.17 are expected to achieve 30
MPa. These results provide valuable insights into the factors affecting the properties of GP mortars and can be
useful for optimizing formulations for different applications.
1. Introduction stone processing [13,14] are among the materials currently being
studied for incorporation in GP formulations. Wood Biomass Fly Ash
Geopolymer (GP) is a synthetic aluminosilicate cement produced by (BFA) is also becoming an important industrial waste stream as a result
reacting a solid aluminosilicate-containing precursor with an aqueous of the shift towards a more sustainable fuel source in power plants, with
alkaline solution [1–5]. GP presents mechanical performances that are the gradual replacement of coal use with residual wood biomass. The
comparable to ordinary Portland cement but is associated with a lower incorporation of BFA as partial replacement for MK in the formulation of
environmental impact as it can recycle various inorganic waste streams GP is scarce and has only recently been reported [15–20]. Additionally,
as its main constituents. In addition to commonly used materials like with the aim of improving the properties of GP for advanced engineering
metakaolin (MK), fly ash (FA), and ground granulated blast furnace slag applications, research on geopolymer composites has gained popularity,
(GGBFS), various industrial waste streams containing reactive silica and including studies related to the incorporation of fibers, nanomaterials,
alumina can be used as geopolymerization precursors. Glass waste [6,7], or phase-change materials [21–26].
ceramics waste [8–10], mine tailings [11,12], and waste from natural According to the results of many authors [27–29], the
* Corresponding author.
E-mail address: [email protected] (T. Archer de Carvalho).
https://doi.org/10.1016/j.conbuildmat.2023.134846
Received 2 June 2023; Received in revised form 11 December 2023; Accepted 29 December 2023
Available online 8 January 2024
0950-0618/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
T. Archer de Carvalho et al. Construction and Building Materials 412 (2024) 134846
geopolymerization reaction mechanism is governed by the following parameters. Accordingly, Yunsheng et al. [47] concluded that
series of steps: dissolution of silicate and aluminate species, as mono Na₂O/Al₂O₃ and H₂O/Na₂O had a very significant effect on the
mers, from solid raw materials; oligomerization of the monomers in compressive strength of harden GP pastes, while comparatively,
solution; polymerization of oligomers and monomers to form alumino SiO₂/Al₂O₃ had negligible influence. On the other hand, the results from
silicate polymers; gelation of aluminosilicate polymer that will produce the experiments and statistical analyses of Lahoti et al. [48] showed that
the setting and hardening of the material. Therefore, the initial alkaline the SiO₂/Al₂O₃ ratio was the most significant parameter in determining
dissolution of precursor particles is recognized as a crucial step, the compressive strength followed by Na₂O/Al₂O₃, while both H₂O/Na₂O
releasing aluminate and silicate species into solution that will form the and W/S were not significant. The same type of approach regarding the
geopolymer matrix through the polymerization process [4,28,30]. Ac effects on the efflorescence level of MK-based GP has not been covered in
cording to Davidovits and Sawyer work [31], in order to obtain geo the literature. Similarly, to the best of our knowledge, simultaneous
polymer materials that present good mechanical properties, the reactant statistical quantification, by means of design of experiments, of the in
mixture should present the following ranges of chemical molar ratios, fluence of formulation parameters on GP based on a mix of MK and BFA
that involve silicates and aluminates: SiO₂/Al₂O₃ from 3.3 to 4.5, has not been carried out.
H₂O/Na₂O between 10 and 25, and Na₂O/Al₂O₃ from 0.8 to 1.6. Finally, it should be stressed that the proposed ranges for SiO₂/Al₂O₃
H₂O/Na₂O ratio, or total water to total sodium oxide in the mortars’ and Na₂O/Al₂O₃ (which, for a cement-like material, should be close to
formulation, affects the solubility of aluminosilicates present in the 4.0 and 1.0, respectively) refer to the desired ratios in the geopolymer
precursors - this ratio is indicative of the pH value used in the prepa matrix rather than to the ratios in the overall composition of the starting
ration of mortars and, typically, the degree of aluminosilicates dissolu mixture. In fact, since the extent of dissolution of silicate and aluminate
tion from the precursors increases with increasing pH [32]. species can vary due to different factors [49,50], it is likely that in most
Furthermore, the structure, stability, and properties of the geopolymer cases, part of the solid materials will not react during the GP prepara
material are impacted by the SiO₂/Al₂O₃ and the Na₂O/Al₂O₃ ratios tion, acting as an inert filler in the final material. Therefore, only a
present in the matrix [30]. For example, a low SiO₂/Al₂O₃ ratio (< 6) fraction of the total silicate and aluminate content will be dissolved and
produces a very rigid 3D network, while a higher SiO₂/Al₂O₃ ratio (> 15) available for participating in the formation of the binder matrix [35,51,
gives the obtained GP a more polymeric character [30]. On the other 52]. Accordingly, when formulating GP, it is crucial to consider the
hand, the occurrence of efflorescence in cured GP, which is considered estimated amounts of dissolved silicate and aluminate species but most
one of the main challenges regarding the use of this material, is linked to published research on GP development does not give enough weight to
a chemical imbalance in the material, involving alkalis and Al sites. In this factor [42,43,47,53–56]. As a result, the chemical ratios and for
fact, according to the amount of available alkalis in the activation so mulations of GP are still, in most cases, based on the precursors’
lution and to the reactivity of the precursors, alkalis will either partic chemical analysis, with only a few exceptions [38,57,58] of experi
ipate in the geopolymerization reaction and become chemically bound mental work relating the amount of dissolved aluminosilicates with the
in Al sites in the 3D-aluminosilicate gel or remain free within the ma properties of the obtained material.
terial’s pores. Free alkalis, usually Na+ and OH− , tend to leach out of the In summary, the importance of chemical and physical ratios in the
pores and onto the surface, where they react with CO₂ from the air, synthesis process of GP is widely acknowledged and the individual ef
resulting in the formation of crystalline sodium carbonates [33–35]. fects on properties such as the workability of the fresh GP mortar and its
According to some authors, in addition to the aesthetic problem, efflo compressive strength and durability after curing have been covered in
rescence may also be related to durability issues, as a high level of ion the literature. However, there is a noticeable research gap regarding the
leaching from the cured material can lead to progressive damage to the statistical quantification of the combined or simultaneous effect of these
geopolymer matrix and thus a deterioration of the mechanical proper key formulation ratios. In particular, no such approach has been re
ties [34–36]. Most proposed strategies for controlling efflorescence in ported for the important case of GP mortars with the incorporation of
GP rely either on promoting a well-balanced geopolymerization reac BFA (an increasingly significant waste stream) and cured in ambient
tion, ensuring that the appropriate and optimized chemical and physical conditions (to ensure the sustainability and future industrial application
ratios are used in the formulation of GP [33,36], or by using curing of the process). Therefore, the main research problem revolves around
conditions other than ambient, namely by applying higher levels of the lack of a comprehensive statistical approach addressing the simul
temperature and relative humidity [18,37]. taneous effect of key formulation ratios in the synthesis of GP in general,
In addition to the chemical ratios, the relation between the liquid and, in particular, in the case of BFA-containing GP cured in ambient
phase and the solid phase is also very important, influencing not only the conditions. Without such an approach, it is difficult to establish the true
behavior of the material, namely in terms of its rheological properties in importance of each ratio and how they interact with each other. This
the fresh state, but also the porosity of the hardened material, and knowledge gap makes it difficult to develop useful correlations between
eventually its mechanical properties [38–41]. Increasing water content formulation ratios and the properties of GP mortars, as well as drawing
in the GP mix generally leads to a decrease in compressive strength due solid conclusions and comparing data from different published studies,
to the formation of voids left by water evaporation in the cured material, therefore hindering advancements in the field and the potential for
while excessively low water content can also result in strength losses due broader industrial applications.
to deficient mixing and hindered geopolymerization process [41]. Based on this, the goal of this study is to statistically quantify and
The effect of the most important chemical and physical ratios on the model the simultaneous effects of SiO₂/Al₂O₃, H₂O/Na₂O, and W/S ratios
properties of MK-based GP has been reported in the literature (see for on the compressive strength, efflorescence level, and spreading diameter
example [40–44]). However, Jan et al. in their review paper [45] present of GP mortars formulated using MK and BFA, by following a central
a wide number of different studies involving the variation of compres composite design of experiments. Moreover, the chemical ratios and
sive strength of GP with some of the relevant mix ratios and conclude formulations are determined based on the estimated amounts of dis
that there is no significant correlation between each ratio or factor solved silicate and aluminates in the mortars, obtained by dissolution
considered individually and the obtained compressive strength. In fact, tests of the two precursors, and the specimens are cured specifically
the use of comprehensive studies based on design of experiments for the under ambient conditions. By systematically analyzing the interaction of
simultaneous quantification and modeling of the effect of chemical ra these formulation ratios and their effect on the response variables, the
tios on MK-based GP is scarce. Ghanbari et al. [46] used the the central current study aims to contribute valuable insights to the understanding
composite design (CCD) method to study these effects on the compres of the GP synthesis process and facilitate more informed decision-
sive strength and bulk density and concluded that H₂O/Na₂O had the making in material formulation for sustainable construction practices.
greatest impact on these properties when compared to the other
2
2. Materials and Methods
2.1. Precursors, aggregates, and alkalination mix
In this work, the geopolymerization process was investigated using

metakaolin (MK) and biomass fly ash (BFA) as the primary
aluminosilicate-containing raw materials. The visual appearance of the
two used geopolymerization precursors is shown in Fig. 1.
MK is a powdered material resulting from the calcination process of a
kaolinitic clay and, for this study, it was supplied by Mota Ceramic So
lutions (Portugal). BFA is a byproduct of the combustion process in
biomass power plants and is obtained as particulate matter from exhaust
gases. BFA resulting from the combustion process of residual forest
biomass and collected using bag filters in a Greenvolt plant located in
Portugal was used in the present work. Fig. 2 displays the particle size
distribution (PSD) curves for each precursor and Table 1 presents the
main PSD parameters, both of which were determined using a Malvern Fig. 2. Frequency PSD curves for MK and BFA.
Masterziser 3000.
D50 represents the particle size below which 50% of the sample
Table 1
volume is found or, in other words, the median particle size with respect
PSD D50 and span parameters for MK and BFA.
to the sample volume. Span is a dimensionless parameter that gives an
indication of the spread of the distribution, calculated as (D90 – D10) / MK BFA
D50 (D10 and D90 are the 10th and 90th percentiles with respect to the D50 (μm) 30 59.3
sample volume, respectively) – the higher the span value, the larger the span 3.96 7.67
spread of the PSD [59]. From the PSD curves and parameters, it can be
concluded that MK is a finer powder and has a narrower distribution of
behavior in the geopolymerization process, and following previous work
particle sizes when compared to BFA. The SEM images in Fig. 3 confirm
[58], the amounts of dissolved aluminate and silicate species from the
the conclusions derived from the PSD analysis. The images show clearly
precursors in an alkaline medium were determined. This will allow the
that the average size of MK particles is smaller than that of BFA.
estimation of the available amounts of silicates and aluminates in the
Moreover, some BFA particles exhibit a characteristic round shape that
fresh mortars, and subsequently formulate them based on these findings,
is typical of fly ash particles.
rather than relying on the overall chemical composition of the mortar.
The chemical composition of the precursors is a key factor that will
By utilizing this approach, it becomes possible to estimate the
impact the geopolymerization process. Therefore, in this study, X-ray
quantities of silicates and aluminates present in the fresh mortars and
Fluorescence was employed to determine the chemical composition of
subsequently formulate them based on these findings, rather than
the precursors. While the composition of MK was given by the supplier, a
relying on the initial overall chemical composition of the mortar.
Malvern PANalytical Zetium XRF equipped with a 4 kW Rhodium x-ray
Suspensions of each powdered precursor in 10 M sodium hydroxide
tube was used for the measurement of BFA (see Table 2).
solution, using a ratio of 200 ml/g, were prepared (see Fig. 5) and
In terms of chemical composition, silica and alumina (and also cal
allowed to react for 120 min at room temperature, with controlled
cium oxide in the case of BFA) were found to be the most abundant
agitation.
oxides.
After decantation and neutralization with hydrochloric acid (Sigma-
Thermogravimetric analysis was carried out using a PerkinElmer
Aldrich, fuming, ≥37%), the determination of silicate and aluminate
STA 6000 instrument, under a nitrogen atmosphere and with a tem
species in the resulting liquid phase was carried out by visible spectro
perature ramp of 10ºC/min. The obtained TG curves (depicted in Fig. 4)
photometry (using a SPECTROstarNano, from BMG LABTECH), based on
show a thermal decomposition of BFA at about 650ºC which is probably
the silicomolybdic acid (SMA) [61] and the Eriochrome Cyanine R (ECR)
due to unburnt organic matter [60]. This is corroborated by the chemical
[62] methods, respectively. The scheme of the alkali dissolution process
analysis by XRF, where carbon and hydrogen are not detected, and the
is shown in Fig. 6.
main oxides sum up to only about 78% in line with the mass loss (or loss
The obtained results are presented in Table 3.
on ignition, LOI) obtained in TG. In the case of MK, the slight mass loss
From Table 3 it can be concluded that the amount of silicate and
observed in the TG curve is probably due to the release of physically
aluminate released by MK is significantly higher than that dissolved
adsorbed water.
from BFA. As referred before, these results were used to estimate the
In order to further characterize the raw materials in terms of their
availability of silicate and aluminate in the fresh mortars and for
Fig. 1. Used GP precursors. MK (left) and BFA (right).
3
Fig. 3. SEM micrographs of MK (left) and BFA (right).
Table 2
Main oxide composition (wt%) of MK and BFA.
MK BFA
SiO₂₂ 58.4% 34.8%

Al₂₂O₃₃ 33.0% 4.4%
CaO 0.1% 27.3%
Na₂₂O 0.1% 0.9%
K2 O 3.1% 4.0%
Fe2O3 3.2% 1.3%
SO3 0.0% 3.1%
MgO 0.2% 2.2%
Sum 98.1% 77.9%
Fig. 5. Suspension of GP precursors during alkali dissolution process with 10 M

sodium hydroxide solution. MK (left) and BFA (right).
2.2. Mortar preparation and testing
GP mortars were prepared by first adding the measured amounts of

precursor mix and alkalination mix to the mortar automatic mixer
(Controls, 65-L0502, depicted in Fig. 8a)) and mixing for 120 s at
140 rpm. After a 60-second pause for scraping the mixer bowl walls, the
required amount of sand was added, and mixing was continued for
another 120 s at 140 rpm. The mixer bowl walls were scraped again
Fig. 4. TG curves of MK and BFA under nitrogen atmosphere.
before continuing to mix for an additional 120 s at 285 rpm.
After mixing, part of the mortars’ volume was used to carry out the
calculating the SiO₂/Al₂O₃ ratio of each GP mortar.
flow table test according to the relevant EN standard [63]. Following
Apart from the precursors, GP mortars were prepared with washed
this method, the flow value was measured by the spreading diameter of a
river sand (supplied by SACT, Lda.) with a median particle diameter
given quantity of fresh mortar (determined by the volume of the stan
(D50) of 370 µm, used as fine aggregates. Additionally, the alkalination
dardized mold used to place it on the table) after it had been subjected to
mix was composed of varying proportions of sodium silicate (SS) solu
15 vertical impacts produced by the flow table mechanism. Fig. 8b)
tion with a silica modulus of 3.3 and 34.9 wt% of dissolved solids (so
shows a picture of the apparatus used.
dium silicate D40, Quimialmel, Lda.), sodium hydroxide (SH) solution
The remaining part of the produced mortars was used to produce cast
used as a main source of Na₂O to avoid the need to dissolve large
specimens for efflorescence assessment and mechanical testing. The
amounts of solid sodium hydroxide (32 wt%, Ângelo Coimbra, S.A.),
fresh mortars were poured into molds measuring 160 mm × 40 mm ×
water and solid SH (Sharlau, pellets, ACS).
40 mm, stroked 30 times to release trapped air, and left to cure for at
4
Fig. 6. Schematic of the alkali dissolution process.
least one day or until the material was hard enough to be demolded.
Table 3
After demolding, the specimens were further cured until a total of seven
Dissolved amounts of SiO₂ and Al₂O₃ from precursors, in terms of mmol/g pre
days of curing was completed. All the specimens were cured indoors,
cursor, determined in 120 min dissolution tests.
under ambient conditions, with relative humidity between 55% and
SiO₂ Al3O2
70% (typically around 60%) and temperature between 18 and 25 ºC
(mmol/g precursor) (mmol/g precursor)
(typically around 22 ºC).
MK 1.2 0.9 Thereafter, the level of efflorescence of the GP mortar specimens was
BFA 0.21 0.11
assessed. Efflorescence is the appearance of a salty deposit at the surface
Fig. 8. a) Mortar mixer, b) flow table apparatus, c) demolded GP mortar specimens curing at room conditions, with different levels of efflorescence, and d) CS test.
5
of the cured material and reveals a chemical imbalance in the formu cations will travel from the bulk of the material to the surface, leaving
lation (see Fig. 7). The level was assessed and recorded according to a void spaces and a weakened structure behind. At the surface, the alka
randomly selected scale ranging from 0 (no signs of efflorescence) to 4 line metal cations will react with CO2 from the air and transform into the
(severe efflorescence). Therefore, each specimen’s efflorescence level whitish carbonated alkaline salt that can be seen in the pictures in Fig. 7.
was assigned a value of 0, 1, 2, 3, or 4. The 5 levels of efflorescence are One commonly used strategy for controlling efflorescence in GP is the
represented by actual test specimens in the following figure. use of high levels of temperature and relative humidity during curing. In
After that, the specimens were tested in a universal mechanical the present study, however, in order to maintain the production process
testing machine (Instron, 4505) to determine their unconfined as simple and sustainable as possible and not hinder its future replica
compressive strength (CS), as shown in Fig. 8d). tion in an industrial environment, all the specimens were cured specif
ically under ambient conditions. Consequently, the issue of efflorescence
formation becomes an important aspect in terms of mortar quality and
2.3. Geopolymerization study methodology was, therefore, studied together with compressive strength and
spreading diameter.
This study of the geopolymerization process was divided into two
stages. The aim of stage I was to determine and understand the effect of 2.3.1. Stage I: effect of the main formulation parameters
the main chemical and physical parameters on the behavior of GP As mentioned above, in stage I, MK was used as a single geo
mortars, both in fresh and in hardened state, using MK as a model raw polymerization precursor, sand was used as fine aggregates at a fixed
material, in a simple single-precursor formulation. Thereafter, in stage Agg/P = 2.0, and the alkalination mix was composed of varying pro
II, based on the ideal ranges for the main chemical and physical pa portions of SS solution, SH solution, water, and SH pellets.
rameters determined before, the goal was to assess the influence of The analysis of the number of degrees of freedom in stage I leads to
introducing BFA as a second precursor in the formulation. the conclusion that the formulation process is defined by the effect of 3
In both stages, for the production of the mortars, the aggregates independent variables or factors. Three of the most important factors in
(sand) were used at a fixed aggregates-to-precursor ratio (Agg/P) of 2.0, the context of the geopolymerization process were, therefore, selected
within the typical range found in published studies focusing on GP for this study: total water-to-total solids ratio (W/S), SiO₂/Al₂O₃, and
mortars cured in room conditions [6–9,13,64–68]. H₂O/Na₂O. Following the discussion in the introduction section, in this
The GP mortar properties of interest, taken as response variables of work the SiO₂/Al₂O₃ ratio is based on the results of alkaline dissolution
the effect of varying the formulation parameters, were selected to be the tests presented in Table 3, rather than on the XRF analysis shown in
spreading diameter in the flow table test, the unconfined compressive Table 2.
strength at 7 days, and the level of efflorescence at 7 days, evaluated as The three used factors and their levels are shown in Table 4.
mentioned before. The spreading diameter in a flow table test represents The simultaneous quantification and modeling of the effect of the 3
a measure of the fluidity and deformability of the fresh mortar, which selected factors on the properties of the produced mortars was carried
will relate to its ability to be easily mixed, placed, and compacted out following a test sequence based on a central composite design of
without the need for excessive force or the addition of excessive amounts experiments. In stage I, there are, therefore, 3 factors with 2 levels each,
of water. A mortar that is too dry will have a low spreading diameter, which will give a total of 23 = 8 combinations or sets of experimental
making it difficult to place and compact. On the other hand, a mortar conditions (also called treatments), that form the corners of a factorial
that is too wet will have a high spreading diameter, making it prone to cube (see Fig. 9). In order to better capture possible non-linearity in the
segregation and resulting in a weaker and less durable final product. response variables, one center point and 6 axial points (points beyond
Compressive strength is a crucial mechanical property of mortars the normal ranges shown in Table 4) were also added, resulting in a
because it measures their ability to withstand external compressive
loads without cracking or collapsing. In order to achieve the desired
compressive strength for a specific application, the mortar mix should be Table 4
optimized by the selection of the most suitable materials and their Stage I factors and their standard levels.
proportions. Finally, high levels of efflorescence are deleterious to the Low Level High Level
quality of the cured mortars because they reveal a chemical imbalance in SiO₂₂/Al₂₂O₃₃ 3.0 5.0
the formulation – alkaline metal cations, Na+ in the present case, are H₂₂O/Na₂₂O 12 25
present in excess in relation to the negatively charged aluminum sites W/S 0.17 0.21
that are formed during the geopolymerization process. The excess
Fig. 7. Examples of tested GP mortar specimens corresponding to each level of efflorescence.
6
Fig. 9. Graphical representation of a) the 15 tested treatments in stage I and b) the 9 tested treatments in stage II.
central composite test program with 15 treatments. In order to obtain GP fixed level in stage II. By applying the same methodology as before, the
mortars with the desired levels of the three selected ratios for each analysis of the process with one new variable (% BFA) and 2 new re
treatment, the proportions of SS solution and SH solution were varied in strictions (fixed SiO₂/Al₂O₃ and W/S) will lead to the conclusion that the
each treatment. Further adjustment to the water content or Na₂O input number of free variables is 2. For this case, the proportion of BFA (wt%)
was achieved by adding water or solid SH to the final activator solution, in the precursor mix and the H₂O/Na₂O ratios were selected as the
respectively. Table 5 summarizes the factor levels and the GP mortar variables to study and their standard high and low levels in stage II are
composition for each of the 15 treatments. shown in Table 6.
The different treatments were labeled as follows: As for the case of stage I, the same type of central composite design of
experiments was carried out in stage II, resulting in a test program with 9
• The Roman numeral (“I” in this case) refers to stage I and is followed treatments. Table 7 summarizes the 9 treatments of stage II, presenting
by their factor levels and GP mortar compositions, using the same labeling
o T1 to T8 – the 8 corners of the factorial cube, corresponding to the system as before.
combinations of factors at their standard low and high levels; For this particular case of design of experiments, where there are 2
o AP1 to AP6 – the 6 axial points in which each of the factors is factors with 2 levels each, the result is a factorial square to which a
tested at a level beyond its standard low and high levels; center point and axial points were added, leading to the central com
o CP – the center point of the factorial cube in which the 3 factors are posite design. In stage II, a total of 19 specimens were produced and
tested at the middle point of their ranges. tested to obtain the experimental data.
The graphical representations of the central composite cube of stage I
In stage I, a total of 34 specimens were produced and tested to obtain
the experimental data.
Table 6
2.3.2. Stage II: effect of introducing BFA Stage II factors and their standard levels.
The second stage focused on the effects of partially replacing MK in Low Level High Level
the precursor mix with varying proportions of BFA, keeping the same % BFA(in precursor mix) 30 70
Agg/P ratio and alkalination mix components as in stage I. Furthermore, H₂₂O/Na₂₂O 16 22
following the results of the first stage, SiO₂/Al₂O₃ and W/S were kept at a
Table 5
Factor levels and composition (wt%) for each treatment of stage I.
Treatment SiO₂/Al₂O₃ H₂O/Na₂O W/S MKd* SS* SH* water* sand
I.T1 5.00 25.00 0.17 25.7% 7.0% 0.4% 14.5% 52.3%

I.T2 3.00 12.00 0.21 24.5% 3.1% 5.5% 17.4% 49.7%
I.T3 3.00 25.00 0.21 25.5% 3.2% 2.1% 17.4% 51.8%
I.T4 5.00 12.00 0.17 24.9% 6.8% 3.3% 14.5% 50.5%
I.T5 5.00 25.00 0.21 24.7% 6.8% 1.0% 17.4% 50.2%
I.T6 3.00 12.00 0.17 25.7% 3.2% 4.4% 14.5% 52.2%
I.T7 5.00 12.00 0.21 23.7% 6.5% 4.4% 17.4% 48.1%
I.T8 3.00 25.00 0.17 26.6% 3.4% 1.5% 14.5% 54.0%
I.CP 4.00 18.50 0.19 25.3% 5.1% 2.3% 16.0% 51.4%
I.AP1 2.37 18.50 0.19 26.0% 2.1% 3.2% 16.0% 52.8%
I.AP2 5.63 18.50 0.19 24.7% 7.9% 1.4% 16.0% 50.1%
I.AP3 4.00 7.89 0.19 23.7% 4.7% 7.5% 16.0% 48.1%
I.AP4 4.00 29.11 0.19 25.8% 5.1% 0.8% 16.0% 52.3%
I.AP5 4.00 18.50 0.157 26.2% 5.2% 1.6% 13.6% 53.3%
I.AP6 4.00 18.50 0.223 24.4% 4.9% 2.9% 18.2% 49.6%
*
In this table, SS represents the solid fraction of the SS solution; SH represents the solid fraction of the SH solution plus the solid SH added as pellets (where
applicable); MKd represents the dry fraction of used MK (excluding 1.5 wt% of residual moisture); water refers to the total water content in the mortar, mainly coming
from the SS and SH solutions, and additional water (where applicable), but also from the residual moisture of solid raw materials.
7
Table 7
Factor levels and composition (wt%) for each treatment of stage II.
Treatment# % BFA H₂O/Na₂O MKd* BFAd* SS* SH* water* sand
(in precursor mix)
II.T1 70% 16.00 7.9% 18.4% 2.3% 3.1% 14.5% 53.7%

II.T2 70% 22.00 8.0% 18.6% 2.3% 2.0% 14.5% 54.5%
II.T3 30% 16.00 18.1% 7.7% 4.5% 2.6% 14.5% 52.6%
II.T4 30% 22.00 18.4% 7.8% 4.5% 1.4% 14.5% 53.3%
II.CP 50% 19.00 13.2% 13.1% 3.4% 2.2% 14.5% 53.6%
II.AP1 17% 19.00 21.4% 4.5% 5.2% 1.7% 14.5% 52.7%
II.AP2 83% 19.00 4.7% 22.0% 1.6% 2.7% 14.5% 54.5%
II.AP3 50% 14.10 13.0% 12.9% 3.4% 3.4% 14.5% 52.8%
II.AP4 50% 23.90 13.3% 13.2% 3.4% 1.5% 14.5% 54.1%
#
In stage II, SiO₂/Al₂O₃ and W/S were fixed as follows: SiO₂/Al₂O₃ = 4.5, W/S = 0.17.
*
See note to Table 5; BFAd represents the dry fraction of used BFA (excluding 2.2 wt% of residual moisture).
and the central composite square of stage II are shown in Fig. 9. process has just started to take place, the amount of liquid in the mix is
The statistical analysis of the results from both stages was carried out expected to be the most determinant factor impacting the fluidity of the
in Minitab 20 Statistical Software. mortar. In fact, according to the geopolymerization reaction mechanism
presented in the introduction section, the initial phase of the process is
3. Results and discussion characterized by the dissolution of silicate and aluminate species from
the solid raw materials, while the polycondensation of monomers and
3.1. Stage I oligomers, which forms the GP matrix and allows the material to set and
harden, will only occur in later stages. Factor B, H₂O/Na₂O, and all the
The flow table test results, reported as the spreading diameter of the squared and interaction terms (A2, B2, C2 A×B, A×C, and B×C) are not
mortars, for the 15 treatments of stage I are presented in Fig. 10. The statistically significant in this case.
higher the spreading diameter of a fresh mortar, the higher its fluidity The model that best fits the data can be mathematically described by
and deformability at the time of testing. The obtained results range from the following equation:
130 mm to about 250 mm, with values in between, which give a good
Spreading diameter (mm) = − 118⋅7 + 12⋅67 × A + 1319⋅8 × C (1)
starting point for developing a model based on the selected factors.
The result of the analysis of each factor’s effect on the spreading This model is also graphically represented by the contour plot in
diameter is shown in the Pareto chart of standardized effects, Fig. 11. For Fig. 12. From the analysis of Eq. 1 and the corresponding contour plot, it
simplicity, factor SiO₂/Al₂O₃ was coded as “A”, H₂O/Na₂O as “B”, and can be concluded that the spreading diameter increases linearly with the
W/S as “C”. increase of W/S and SiO₂/Al₂O₃ ratios. A spreading diameter between
The Pareto chart displays the absolute values of the standardized 140 mm and 180 mm is usually suited for the typical applications of this
effects in descending order, with a reference line to help identify sta kind of mortar and, therefore, for a SiO₂/Al₂O₃ ratio in the range sug
tistically significant factors. The longer the bar, the higher the impor gested by Davidovits and Sawyer [31], W/S should be in the range of
tance of the respective factor in the description of the response variable. about 0.162 to 0.183, as marked in the contour plot with the red square.
The reference line is determined by a chosen significance level (also The same type of analysis was performed for the case of the efflo
called alpha value) and factors that cross this line are statistically sig rescence level of mortars after 7 days of curing in room conditions. The
nificant in the model. In this study the significance level is 0.05 or 5% efflorescence level for each of the treatments is shown in Fig. 13. In this
which implies that the confidence level for the analysis is 95%. In the case, values ranging from 0 to 4 were obtained, with data points in
case of the spreading diameter of the fresh mortars, it can be concluded between.
from Fig. 11 that W/S and SiO₂/Al₂O₃ ratios are the two statistically The Pareto chart for the standardized effects for the case of having
significant factors, with higher importance attributed to W/S. This the efflorescence level as the response variable is shown below. In this
makes sense since, at very early ages, when the geopolymerization
Fig. 10. Spreading Diameter for each treatment of stage I (the black dots represent the individual data points).
8
Fig. 11. Pareto chart of the standardized effects for the spreading diameter of the mortars of stage I (significance level of 5%).
Fig. 12. Contour plot of the spreading diameter of GP mortars as a function of W/S and SiO₂/Al₂O₃.
situation, both SiO₂/Al₂O₃ and H₂O/Na₂O are statistically significant of efflorescence is most relevant are at low H₂O/Na₂O (less than about
terms involved in describing the efflorescence level. Furthermore, all 12) or low SiO₂/Al₂O₃ (less than about 3.75). In the area of low H₂O/
their second-order terms – (SiO₂/Al₂O₃)2, (H₂O/Na₂O)2, and SiO₂/Al₂O₃ Na₂O, the concentration of aluminate species dissolved from the solid
× H₂O/Na₂O – are also statistically significant at a 5% level, making the particles of MK might not be enough to balance the excessively high
underlying model more complex. Taking into consideration the origin of concentration of Na₂O. This will lead to a chemical imbalance in the
efflorescence in the case of these GP mortars – the chemical imbalance formation process of the GP matrix, where the charge-compensating Na+
between sodium cations and negatively charged aluminum sites (as cations are in excess relative to the negatively charged Al sites of the
mentioned before) –, it makes sense that the factors describing the matrix. These excess Na+ cations will be poorly linked to the GP matrix
efflorescence process are the ones including these chemical elements. and will tend to leach out and promote the efflorescence effect.
Fig. 14. On the other hand, one possible mechanism that explains the high
The model that best fits the data regarding the efflorescence level can levels of efflorescence observed in the area of low SiO₂/Al₂O₃ is the
therefore be mathematically described by Eq. 2: formation of more ramified Si-O-Al bonds in the presence of higher
proportions of Al₂O₃. This will lead to a less dense GP matrix structure
Efflorescence Level (a⋅u⋅) = 20⋅42 - 6⋅85 × A - 0⋅471 × B + 0⋅838 × A2 + (see, e.g., [69]) that will allow an easier mobilization of the free Na+
0⋅02188 × B2 - 0⋅0705 × A × B (2)
cations to the surface of the specimens, promoting the development of
Fig. 15 displays the contour plot representing this model, which efflorescence.
visually depicts the impact of the squared and interaction terms (A2, B2, One of the desired properties of any mortar (or generally any con
and AB) on the response surface through its curvature. struction material) is its chemical stability over time and, hence, the
As can be concluded from the contour plot, the areas where the effect efflorescence level should be as low as possible. In the present study, as
9
Fig. 13. Efflorescence level for each treatment of stage I.
Fig. 14. Pareto chart of the standardized effects for the efflorescence level of the mortars of stage I (significance level of 5%).
marked in Fig. 15 with the red square, for achieving a minimum efflo
rescence level within the variable ranges suggested by Davidovits and Comp⋅ Strength (MPa) = − 13⋅7 - 16⋅56 × A + 19⋅065 × B - 356⋅8 × C +
Sawyer [31], a good window would be having SiO₂/Al₂O₃ between 3.8 2⋅552 × A2- 0⋅5874 × B2 (3)
and 4.5, and H₂O/Na₂O between 13.5 and 20.0.
Finally, the results of measured unconfined compressive strength are From a physicochemical point of view, it makes sense that
presented in Fig. 16, ranging from as low as 1,3 MPa (I.T1) to as high as compressive strength is dependent on SiO₂/Al₂O₃, H₂O/Na₂O, and W/S.
around 60 MPa. SiO₂/Al₂O₃, as mentioned before, influences the structure and, therefore,
According to the statistical analysis, summarized in the Pareto chart the mechanical strength of the geopolymer material. In the present case,
shown in Fig. 17, it can be concluded that all three studied factors, SiO₂/ the mathematical analysis of the fitted model leads to the conclusion
Al₂O₃, H₂O/Na₂O, and W/S, are statistically significant in the description that the compressive strength is an increasing function of SiO₂/Al₂O₃ in
of the mortars’ compressive strength variation, with a higher level of the range of about 3.2 to 5.0. These results are not consistent with most
importance attributed to H₂O/Na₂O. This is in line with other reported of the reported studies, where a maximum of compressive strength is
studies on the geopolymerization process of MK [46,47] where achieved for SiO₂/Al₂O₃ of about 4 or Si/Al of about 2 [41]. One possible
H₂O/Na₂O was also found to be among the most important factors cause for this discrepancy could be the fact that in the present study the
affecting compressive strength, followed by SiO₂/Al₂O₃. Furthermore, estimated amounts of dissolved silicates and aluminates were used for
some of the second-order terms should also be considered, namely the the calculation of the SiO₂/Al₂O₃ ratio, while in most other research
squared terms of SiO₂/Al₂O₃ and H₂O/Na₂O, and the interaction between works the ratio was computed based on the initial chemical composition
SiO₂/Al₂O₃ and W/S. of the raw materials. Furthermore, due to concerns regarding the
For the case of compressive strength, a more detailed statistical efflorescence effect and therefore the durability of the produced GP
analysis based on the distribution of residuals leads to the conclusion mortars, the H₂O/Na₂O ratio used in the present study was higher (in the
that the model that best fits the data should exclude the interaction term range of 12 to 25) than is generally the case for other studies, where
SiO₂/Al₂O₃ × W/S. Therefore, the best model is mathematically ratios around 11, and below, are the most common [41,44,48,69]. This
described by Eq. 3. variation in the H₂O/Na₂O ratio can cause a change in the rates of
10
Fig. 15. Contour plot of the efflorescence level of GP mortars as a function of H₂O/Na₂O and SiO₂/Al₂O₃.
Fig. 16. Unconfined compressive strength for each treatment of stage I.
dissolution of silicates and aluminates from the precursors – the first step particle size of the MK used is significantly larger than is generally re
in the mentioned geopolymerization reaction mechanism – eventually ported by other authors (around 30 µm compared to < 5 µm in other
altering the proportion and concentration of silicate and aluminate relevant studies [44,46,70]). A larger mean particle size implies that the
monomers present in solution. The relative proportion of silicate and first step of the geopolymerization reaction mechanism, i.e., the disso
aluminate monomers (or the effective SiO₂/Al₂O₃ ratio), as mentioned in lution of silicate and aluminate monomers from MK (promoted by high
the introduction section, will determine the GP matrix structure, that Na₂O concentration, or low H₂O/Na₂O ratios), will probably become
can vary from a very rigid 3D network at low ratios to a matrix with a limited by diffusion effects, due to the lower specific surface area of the
more polymeric character at high ratios. These effects are possible precursor particles. Therefore, decreasing the H₂O/Na₂O ratio beyond a
causes for the discrepant variation of compressive strength with the certain point will not contribute to a larger extent of dissolution but will
SiO₂/Al₂O₃ ratio obtained in the present experimental study when rather accumulate Na₂O within the formed matrix, leading to efflores
compared to the results from the literature. cence problems and consequently lower compressive strength.
On the other hand, the proportion of Na₂O in the mix can also affect Finally, it is clear that the compressive strength decreases linearly
the quality of the cured mortars, namely in terms of mechanical prop with W/S in the tested range of 0.17 to 0.21. These results are in
erties, due to the process of efflorescence. According to the results ob agreement with the general trend of the data collected by Castillo et al.
tained, compressive strength increases for H₂O/Na₂O in the range of [41] on several geopolymerization studies. Although other authors have
12.0 to about 16.2 and then decreases in the range of 16.2 to 25.0. found no correlation between W/S and compressive strength of geo
Although the reported results found in the literature tend to agree that polymers (refer, e.g., to [48]), it is reasonable to assume that, in addition
compressive strength is a decreasing function of the H₂O/Na₂O ratio in to affecting the rheology of the fresh mortar (as mentioned before), a
MK-based GP [42,46] this was not the case in the present study. One higher proportion of water in the formulation will also affect the me
possible cause for this discrepancy could be the fact that the mean chanical properties of the GP cured mortar through the mechanism of
11
Fig. 17. Pareto chart of the standardized effects for the compressive strength of the mortars of stage I (significance level of 5%).
excess water evaporation during curing, which will create a higher 53 MPa. Since the approach used in the present study – the simultaneous
volume of voids in the final material [41]. optimization of this particular set of parameters and response variables –
For the graphical representation of this function, given its depen has not been found in the literature, a direct comparison between the
dence on three variables, three contour plots will be required, as pre optimized level obtained for the parameters and other authors’ results
sented in Fig. 18. Each contour plot depicts the variation of the mortars’ holds low significance. Nonetheless, it is worth mentioning that the
compressive strength with one combination of two variables, assuming a optimized SiO₂/Al₂O₃ (4.5) is in the upper limit of the optimum range
fixed value for the third variable. Hence, as an example, for the case of found by many other authors (from about 3.5 to about 4.5, refer to
the first contour plot (left) in Fig. 18, the variation of compressive Castillo et al. [41]) when studying the effect of SiO₂/Al₂O₃ on
strength is represented as a function of SiO₂/Al₂O₃ and H₂O/Na₂O, compressive strength alone. On the other hand, regarding H₂O/Na₂O, in
assuming that W/S is fixed. The fixed (or hold) values for SiO₂/Al₂O₃, order to simultaneously achieve high compressive strength and a low
H₂O/Na₂O, and W/S are shown in the legend of the contour plots and are level of efflorescence, the optimized ratio obtained in the present study
4.5, 19.0, and 0.17, respectively. (19.0) is considerably higher than is generally the case for most of the
From the contour plots analysis, it can be concluded that the region published research, that focuses mainly on the compressive strength
of high compressive strength can be achieved with the approximate achieved by the cured GP. In those studies, as mentioned before, ratios
variable ranges: around 11, and below, are the most common [41,44,48,69]. In any case,
a reasonable balance should be pursued between the compressive
• SiO₂/Al₂O₃ above 4.25 strength and the efflorescence level of the GP mortar, and ultimately its
• H₂O/Na₂O between 14.5 and 18.0 durability [37]. Finally, it should be noted that the water-to-solids (W/S)
• W/S between 0.165 and 0.175 ratio is greatly dependent on the properties of the precursors, activator
solution, and aggregates used in the production of the mortar.
Finally, considering the three response variables simultaneously, i.e., Furthermore, in the present study, the workability of the mortar is being
maximizing the compressive strength while maintaining the spreading optimized simultaneously with the compressive strength, and the W/S
diameter and efflorescence levels within suitable limits, an optimized set ratio is a significant factor affecting both response variables. Therefore,
of factor values would be SiO₂/Al₂O₃ = 4.5, H₂O/Na₂O = 19.0, and W/S the significance of a direct comparison with other authors’ results is low.
= 0.17. According to the models, this combination of factors is expected Nevertheless, it is worth noting that the achieved optimum W/S is
to result in a GP mortar with a spreading diameter of about 160 mm, an within the reported range of several studies compiled by Castillo et al.
efflorescence level of around 0, and a compressive strength of about [41] and that its effect on compressive strength is similar to what has
Fig. 18. Contour plots of the compressive strength of GP mortars as a function of H₂O/Na₂O, SiO₂/Al₂O₃, and W/S.
12
been reported. CaO in BFA and increased by high alkalinity (low H₂O/Na₂O) [72–74].
This seemed to be the case of, for instance, treatment II.T1, where the
3.2. Stage II mortar started to harden in a very short period of time, with zero
spreading in the flow table test – the measured diameter of 100 mm is
In stage II, as mentioned before, MK was partially replaced by BFA in equal to the initial sample diameter, meaning that there was no
the precursor mix. % BFA (in the precursor mix) was therefore consid spreading during the test.
ered a new factor in this stage. Focusing on the importance of The efflorescence level results for the 9 treatments of stage II are
compressive strength in the characterization of a mortar, the factor presented in Fig. 22. In this case, the variability of results is not as high
found to have the highest importance in describing this response vari as it would be appropriate for developing an accurate model.
able in stage I – H₂O/Na₂O – was chosen as the second factor for studying From the Pareto chart of standardized effects (Fig. 23), it can be
in stage II. In this case, SiO₂/Al₂O₃ and W/S were kept fixed at 4.5 and concluded that, again, % BFA seems to be the most important factor for
0.17, following the good results obtained in stage I with these factor explaining the variation of efflorescence level. The effect of H₂O/Na₂O is
levels. The statistical analysis of the results of stage II was carried out in present in the square term (H₂O/Na₂O)2. According to the analysis, this
a similar way to that performed in stage I. term is on the limit of statistical significance for a confidence level of
The results regarding the flow table test spreading diameter of the 5%, but considering the chemical process involved in efflorescence, it is
mortars are presented for the 9 treatments of stage II in Fig. 19. reasonable to include it in the model.
The results range from 100 mm (which corresponds to the absence of A more careful statistical analysis considering the distribution of
spreading) to about 275 mm.Fig. 20. residuals leads to the conclusion that the model should additionally
In stage II, factor %BFA was coded as “A” and H₂O/Na₂O as “B”. include the linear term H₂O/Na₂O, making it also more congruent
The Pareto chart shows that all the terms (linear, square, and inter (because it becomes a hierarchical model, where second-order terms are
action terms) are statistically significant affecting the spreading diam only present if the corresponding linear terms are included).
eter of mortars incorporating BFA. The model that best represents the Taking this into consideration, Eq. 5 is the mathematical description
relationship between the spreading diameter and the factors can be of the model that best fits the data of efflorescence level as a function of
mathematically described by the following equation: % BFA and H₂O/Na₂O.
Spreading diameter (mm) = − 79⋅7 - 2⋅033 × A + 31⋅15 × B - 0⋅05009 × A2 - Efflorescence Level (a⋅u⋅) = 22⋅01 - 0⋅1310 × A - 2⋅155 × B + 0⋅002183 ×
1⋅072 × B2+ 0⋅3073 × A × B (4) A2+ 0⋅0562 × B2 (5)
The graphical representation of this model is presented in the con The best model for the efflorescence level in stage II is graphically
tour plot shown in Fig. 21. represented in the contour plot shown in Fig. 24. From the contour plot
From the experimental practice it has become clear that, probably it can be concluded that, in the experimental conditions of stage II, a low
due to the properties of the BFA, its water demand is considerably higher level of efflorescence (< 0.5) can only be achieved if % BFA in the
than that of MK. In fact, according to the literature, one of the main precursor mix is less than 50. At higher levels of BFA incorporation, the
effects of incorporating high LOI FA in the formulation of geopolymer significantly lower reactivity of BFA when compared to MK will affect
mortars is the higher demand for activator solution to maintain a the first steps of the geopolymerization reaction mechanism, namely by
specified level of workability of the fresh mortar, when compared to causing lower concentrations of dissolved aluminosilicates in solution.
precursors with identical particle size distribution [71]. The findings of This, in turn, will lead to the formation of a less dense matrix and to an
the present study appear to support this reported effect since it was excess of Na+ cations relative to the matrix Al sites. These two effects, as
found that the spreading diameter (and thus, the workability) decreases mentioned before, can promote the efflorescence process in GP
as the proportion of BFA included in the precursor mix increases, for the specimens.
same water-to-solids ratio (refer to Eq. 4). This was found despite BFA Finally, the mortars’ compressive strength for each treatment of
having a larger average particle size than MK, as shown in Table 1 and stage II is presented in Fig. 25, ranging from 1.3 MPa to 45.5 MPa, and
Fig. 2. This explains why, for the most part of the factor ranges, the with a good spread of values in between.
spreading diameter decreases with increasing %BFA. As for the influ The Pareto chart of standardized effects (Fig. 17), shows that% BFA
ence of H₂O/Na₂O, this is less clear but might be related to the mecha is clearly the most important factor influencing the variation of the
nism of flash or rapid setting of the mortar promoted by the presence of response variable, compressive strength. Additionally, H₂O/Na₂O and
Fig. 19. Spreading Diameter for each treatment of stage II.
13
Fig. 20. Pareto chart of the standardized effects for the spreading diameter of the mortars of stage II (significance level of 5%).
Fig. 21. Contour plot of the spreading diameter in stage II as a function of % BFA and H₂O/Na₂O.
the higher order terms including % BFA, namely (% BFA)2 and % BFA should lead to less available silicate and aluminate monomers to pro
× H₂O/Na₂O, are also statistically significant factors in the description duce the geopolymer matrix. Therefore, in this situation, a less extensive
of compressive strength and should be included in the best model, that is GP matrix will be produced, and a higher proportion of undissolved and
expressed mathematically in Eq. 6.Fig. 26. unreacted solid material will be present in the cured GP mortar. These
effects will result in lower mechanical performance of the material.
Comp⋅ Strength (MPa) = 137⋅3 - 2⋅363 × A - 3⋅820 × B + 0⋅00556 × A2+ Additionally, according to Diaz et al. [71], the high demand for activator
0⋅0588 × A × B (6)
solution of FA with high LOI could be associated with a loss in me
The contour plot presented in Fig. 27 visually reveals the dominance chanical strength of geopolymers incorporating this raw material,
of % BFA in the variation of the mortars’ compressive strength. It can be although other authors found no correlation between LOI and
concluded that, in the tested range, the GP mortars containing BFA in compressive strength [75]. To achieve a compressive strength higher
the precursor mix show a lower compressive strength than the MK-based than 30 MPa with an incorporation of BFA of around 40%, H₂O/Na₂O as
mortars, for the same set of ratios. Moreover, an incorporation level low as 16.0 to 15.0 must be used.
higher than 50% in the precursor mix yields GP mortars with In a similar way to what was carried out in Stage I, considering the
compressive strength below 20 MPa. The cause behind this effect on three response variables simultaneously, i.e., maximizing the compres
compressive strength is similar to what was presented before for the sive strength, while maintaining the spreading diameter and efflores
efflorescence level. In fact, since BFA seems to contribute considerably cence levels within suitable limits, and also maximizing BFA
lower amounts of aluminosilicates to the mix when compared to MK incorporation, an optimized set of factor values would be % BFA = 39.0
(refer to Table 3), i.e., the dissolution step of the geopolymerization and H₂O/Na₂O = 15.5 (SiO₂/Al₂O₃ = 4.5, W/S = 0.17 were fixed pa
reaction mechanism is less extensive, an increased proportion of BFA rameters). This combination of factors is expected to result in a GP
14
Fig. 22. Efflorescence level for each treatment of stage II.
Fig. 23. Pareto chart of the standardized effects for the efflorescence level of the mortars of stage II (significance level of 5%).
mortar with a spreading diameter of about 175 mm, an efflorescence 4. Conclusions

level of around 0.3, and a compressive strength of about 30 MPa, ac
cording to the models. The number of published studies specifically The effect of main chemical and physical formulation parameters
dedicated to investigating the incorporation of BFA in MK-based GP (SiO₂/Al₂O₃, H₂O/Na₂O, W/S, and % BFA) on the properties of MK-based
mortars is very limited. Moreover, the optimization approach applied and (MK + BFA)-based GP mortars was quantified and modeled. The
herein – simultaneously optimizing the compressive strength, the methodology is based on the statistical analysis of an experimental
spreading diameter, and the efflorescence levels of (MK + BFA)-based program developed through central composite design of experiments
GP – is not found in the literature, and, therefore, any direct compari and was used for the first time for studying (MK + BFA)-based GP
son of the obtained optimized parameters with those used by other au mortars. Moreover, the chemical ratio SiO₂/Al₂O₃ and compositions
thors is not significant. Nevertheless, upon reviewing the literature were established by using the results of dissolution tests conducted on
regarding the development of (MK + BFA)-based GP (and focused the two precursors, which also represents a novel approach in this
mainly on the maximization of the compressive strength), it can be subject.
observed that BFA incorporation usually ranges from 20 to 75% and The main conclusions regarding the development of GP mortar based
SiO₂/Al₂O₃ is typically used between 3.3 and 5.3 [15,16,18,19,76]. It on MK only are as follows:
can be concluded, therefore, that the obtained optimized parameters
herein fall within the typical ranges. Furthermore, Saeli et al. [76], • SiO₂/Al₂O₃, H₂O/Na₂O, and W/S were significant factors affecting
taking into consideration the efflorescence effect of the cured GP, the compressive strength of the GP mortars. The region of higher
established an H₂O/Na₂O ratio range of 12.9 to 18.2. Their results sug compressive strength (> 55 MPa) is achieved with SiO₂/Al₂O₃
gest that an optimized GP formulation requires an H₂O/Na₂O ratio of greater than 4.25, H₂O/Na₂O between 14.5 and 18.0, and W/S be
15.9, similar to the findings of the present study. tween 0.165 and 0.175. The region of minimum efflorescence level
was obtained with SiO₂/Al₂O₃ between 3.8 and 4.5 and H₂O/Na₂O in
the range of 13.5 to 20.0.
15
Fig. 24. Contour plot of the efflorescence level of GP mortars of stage II as a function of H₂O/Na₂O and % BFA.
Fig. 25. Unconfined compressive strength for each treatment of stage II.
• The simultaneous optimization of the three response variables leads These results provide valuable insights into the factors affecting the
to the following set of factor values: SiO₂/Al₂O₃ = 4.5, H₂O/Na₂O properties of (MK + BFA)-based GP mortars and can be useful in
= 19.0, and W/S = 0.17. According to the developed models, this adjusting formulations for different fluidity/deformability and me
combination is expected to result in a GP mortar with a spreading chanical strength requirements, while also addressing the problem of
diameter of about 160 mm, an efflorescence level close to 0, and a efflorescence.
compressive strength of about 53 MPa.
Concerning the study of (MK + BFA)-based GP the following Funding
conclusions can be drawn:
• In general, the incorporation of BFA in the formulation led to lower This research was funded by Fundação para a Ciência e a Tecnologia
spreading diameter and compressive strength, and higher efflores (FCT) through the support of CERENA (Strategic Project FCT-UIDB/
cence levels, when compared with GP mortars based on MK only. 04028/2020), through the support of CDRSP (Strategic Projects UIDB/
This effect was magnified by increasing levels of % BFA, which was 04044/2020 and UIDP/04044/2020; and Associate Laboratory ARISE
the most important factor for explaining the variation in the prop LA/P/0112/2020), and through the individual grant 2020.04531.BD.
erties of (MK + BFA)-based mortars. The research was also funded by the Portugal2020 (Centro2020) project
• The simultaneous maximization of the compressive strength and BFA “3DFiber.concrete” (CENTRO-01–0247-FEDER-072643).
incorporation, while maintaining the spreading diameter and efflo
rescence levels within suitable limits, leads to a formulation with % CRediT authorship contribution statement
BFA = 39.0 and H₂O/Na₂O = 15.5. This GP mortar is expected to
yield little to no efflorescence, a spreading diameter of about Gaspar Florindo: Writing – review & editing, Validation, Method
175 mm, and a compressive strength of about 30 MPa. ology, Conceptualization. Archer de Carvalho Tomás: Writing –
16
Fig. 26. Pareto chart of the standardized effects for the compressive strength of the mortars of stage II (significance level of 5%).
Fig. 27. Contour plot of the compressive strength of GP mortars of stage II as a function of %BFA and H₂O/Na₂O.
original draft, Methodology, Formal analysis, Data curation, Conceptu References

alization. Mateus Artur: Writing – review & editing, Conceptualization.
Marques Ana C.: Writing – review & editing, Conceptualization. [1] R.M. Novais, J. Carvalheiras, M.N. Capela, M.P. Seabra, R.C. Pullar, J.A. Labrincha,
Incorporation of glass fibre fabrics waste into geopolymer matrices: an eco-friendly
solution for off-cuts coming from wind turbine blade production, Constr. Build.
Declaration of Competing Interest Mater. 187 (2018) 876–883, https://doi.org/10.1016/j.conbuildmat.2018.08.004.
[2] C.A. Rees, J.L. Provis, G.C. Lukey, J.S.J. van Deventer, The mechanism of
geopolymer gel formation investigated through seeded nucleation, Colloids Surf. A
The authors declare that they have no known competing financial Physicochem Eng. Asp. 318 (2008) 97–105, https://doi.org/10.1016/j.
interests or personal relationships that could have appeared to influence colsurfa.2007.12.019.
the work reported in this paper. [3] J.L. Provis, Geopolymers and other alkali activated materials: why, how, and what?
Mater. Struct. /Mater. Et. Constr. 47 (2014) 11–25, https://doi.org/10.1617/
s11527-013-0211-5.
Data Availability [4] P. Duxson, A. Fernández-Jiménez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. van
Deventer, Geopolymer technology: the current state of the art, J. Mater. Sci. 42
(2007) 2917–2933, https://doi.org/10.1007/s10853-006-0637-z.
Data will be made available on request. [5] N.B. Singh, B. Middendorf, Geopolymers as an alternative to Portland cement: an
overview (article), Constr. Build. Mater. 237 (2020) 117455, https://doi.org/
Acknowledgments 10.1016/j.conbuildmat.2019.117455.
[6] B. Zhang, P. He, C.S. Poon, Optimizing the use of recycled glass materials in alkali
activated cement (AAC) based mortars (article), J. Clean. Prod. 255 (2020)
The authors would like to express their gratitude to the Portuguese 120228, https://doi.org/10.1016/j.jclepro.2020.120228.
companies that kindly supplied the raw materials used in this study [7] R. Si, Q. Dai, S. Guo, J. Wang, Mechanical property, nanopore structure and drying
shrinkage of metakaolin-based geopolymer with waste glass powder (article),
—Mota Ceramic Solutions and Greenvolt.
17
J. Clean. Prod. 242 (2020) 118502, https://doi.org/10.1016/j. [32] D. Khale, R. Chaudhary, Mechanism of geopolymerization and factors influencing
jclepro.2019.118502. its development: a review, J. Mater. Sci. 42 (2007) 729–746, https://doi.org/
[8] G.F. Huseien, A.R.M. Sam, K.W. Shah, J. Mirza, M.M. Tahir, Evaluation of alkali- 10.1007/s10853-006-0401-4.
activated mortars containing high volume waste ceramic powder and fly ash [33] Y. Wang, X. Liu, W. Zhang, Z. Li, Y. Zhang, Y. Li, Y. Ren, Effects of Si/Al ratio on
replacing GBFS, Constr. Build. Mater. 210 (2019) 78–92, https://doi.org/10.1016/ the efflorescence and properties of fly ash based geopolymer, J. Clean. Prod. 244
j.conbuildmat.2019.03.194. (2020) 118852, https://doi.org/10.1016/J.JCLEPRO.2019.118852.
[9] G.F. Huseien, A.R.M. Sam, K.W. Shah, M.A. Asaad, M.M. Tahir, J. Mirza, Properties [34] E. Ozcelikci, A. Kul, M.F. Gunal, B.F. Ozel, G. Yildirim, A. Ashour, M. Sahmaran,
of ceramic tile waste based alkali-activated mortars incorporating GBFS and fly A comprehensive study on the compressive strength, durability-related parameters
ash, Constr. Build. Mater. 214 (2019) 355–368, https://doi.org/10.1016/j. and microstructure of geopolymer mortars based on mixed construction and
conbuildmat.2019.04.154. demolition waste, J. Clean. Prod. 396 (2023) 136522, https://doi.org/10.1016/j.
[10] G.A. Ramos, F. Pelisser, P.J. Paul Gleize, A.M. Bernardin, M.D. Michel, Effect of jclepro.2023.136522.
porcelain tile polishing residue on geopolymer cement, J. Clean. Prod. 191 (2018) [35] M.A. Longhi, Z. Zhang, E.D. Rodríguez, A.P. Kirchheim, H. Wang, Efflorescence of
297–303, https://doi.org/10.1016/j.jclepro.2018.04.236. alkali-activated cements (geopolymers) and the impacts on material structures: a
[11] H. Paiva, J. Yliniemi, M. Illikainen, F. Rocha, V. Ferreira, Mine tailings critical analysis, Front Mater. 6 (2019), https://doi.org/10.3389/
geopolymers as a waste management solution for a more sustainable habitat FMATS.2019.00089.
(article), Sustainability 11 (2019) 995, https://doi.org/10.3390/su11040995. [36] H.A. Ali, K. Sun, D. Xuan, J.X. Lu, M. Cyr, C.S. Poon, Recycling of high-volume
[12] A. Wang, H. Liu, X. Hao, Y. Wang, X. Liu, Z. Li, Geopolymer synthesis using garnet waste glass powder in alkali-activated materials: an efflorescence mitigation
tailings from molybdenum mines (article), Minerals 9 (2019) 48, https://doi.org/ strategy, J. Build. Eng. 65 (2023) 105756, https://doi.org/10.1016/J.
10.3390/min9010048. JOBE.2022.105756.
[13] W.H. Lee, K.L. Lin, T.H. Chang, Y.C. Ding, T.W. Cheng, Sustainable development [37] A. Allahverdi, E. Najafi Kani, K.M.A. Hossain, M. Lachemi, Methods to control
and performance evaluation of marble-waste-based geopolymer concrete (article), efflorescence in alkali-activated cement-based materials. in: Handbook of Alkali-
Polym. (Basel) 12 (2020) 1924, https://doi.org/10.3390/POLYM12091924. Activated Cements, Mortars and Concretes, Elsevier, 2015, pp. 463–483, https://
[14] B.N. Bayiha, N. Billong, E. Yamb, R.C. Kaze, R. Nzengwa, Effect of limestone doi.org/10.1533/9781782422884.3.463.
dosages on some properties of geopolymer from thermally activated halloysite, [38] S. Samantasinghar, S.P. Singh, Effect of synthesis parameters on compressive
Constr. Build. Mater. 217 (2019) 28–35, https://doi.org/10.1016/j. strength of fly ash-slag blended geopolymer, Constr. Build. Mater. 170 (2018)
conbuildmat.2019.05.058. 225–234, https://doi.org/10.1016/J.CONBUILDMAT.2018.03.026.
[15] L. Silvestro, T.P. Scolaro, A.S. Ruviaro, G.T. dos Santos Lima, P.J.P. Gleize, [39] A.A. Adam, B.R. Ramadhan, S. Maricar, The effects of water to solid ratio, activator
F. Pelisser, Use of biomass wood ash to produce sustainable geopolymeric pastes, to binder ratio, and lime proportion on the compressive strength of ambient-cured
Constr. Build. Mater. 370 (2023) 130641, https://doi.org/10.1016/J. geopolymer concrete, J. Civ. Eng. Forum 5 (2019), https://doi.org/10.22146/
CONBUILDMAT.2023.130641. JCEF.43878.
[16] S. Jurado-Contreras, E. Bonet-Martínez, P.J. Sánchez-Soto, O. Gencel, D. Eliche- [40] F. Bowen, L. Jiesheng, W. Jing, C. Yaohua, Z. Tongtong, T. Xiaoming,
Quesada, Synthesis and characterization of alkali-activated materials containing S. Zhengguang, Investigation on the impact of different activator to solid ratio on
biomass fly ash and metakaolin: effect of the soluble salt content of the residue, properties and micro-structure of metakaolin geopolymer, Case Stud. Constr.
Arch. Civ. Mech. Eng. 22 (2022) 1–21, https://doi.org/10.1007/s43452-022- Mater. 16 (2022) e01127, https://doi.org/10.1016/J.CSCM.2022.E01127.
00444-2. [41] H. Castillo, H. Collado, T. Droguett, S. Sánchez, M. Vesely, P. Garrido, S. Palma,
[17] M. Saeli, D.M. Tobaldi, M.P. Seabra, J.A. Labrincha, Mix design and mechanical Factors affecting the compressive strength of geopolymers: a review, Minerals 11
performance of geopolymeric binders and mortars using biomass fly ash and (2021) 1317, https://doi.org/10.3390/min11121317.
alkaline effluent from paper-pulp industry, J. Clean. Prod. 208 (2019) 1188–1197, [42] T. Meng, S. Ahmed, D. Dai, Y. Yu, Effects of load types and critical molar ratios on
https://doi.org/10.1016/J.JCLEPRO.2018.10.213. strength properties and geopolymerization mechanism, Rev. Adv. Mater. Sci. 60
[18] A. De Rossi, L. Simão, M.J. Ribeiro, D. Hotza, R.F.P.M. Moreira, Study of cure (2021) 216–222, https://doi.org/10.1515/RAMS-2021-0019.
conditions effect on the properties of wood biomass fly ash geopolymers, J. Mater. [43] I. Ozer, S. Soyer-Uzun, Relations between the structural characteristics and
Res. Technol. 9 (2020) 7518–7528, https://doi.org/10.1016/J. compressive strength in metakaolin based geopolymers with different molar Si/Al
JMRT.2020.05.047. ratios, Ceram. Int 41 (2015) 10192–10198, https://doi.org/10.1016/J.
[19] R. Rajamma, J.A. Labrincha, V.M. Ferreira, Alkali activation of biomass fly CERAMINT.2015.04.125.
ash–metakaolin blends, Fuel 98 (2012) 265–271, https://doi.org/10.1016/J. [44] P. Duxson, S.W. Mallicoat, G.C. Lukey, W.M. Kriven, J.S.J. van Deventer, J.S.J. van
FUEL.2012.04.006. Deventer, The effect of alkali and Si/Al ratio on the development of mechanical
[20] S. Candamano, P. De Luca, P. Frontera, F. Crea, Production of geopolymeric properties of metakaolin-based geopolymers, Colloids Surf. A Physicochem Eng.
mortars containing forest biomass ash as partial replacement of metakaolin, Asp. 292 (2007) 8–20, https://doi.org/10.1016/j.colsurfa.2006.05.044.
Environ. 2017, Vol. 4, Page 74 4 (2017) 74, https://doi.org/10.3390/ [45] A. Jan, Z. Pu, K.A. Khan, I. Ahmad, A.J. Shaukat, Z. Hao, I. Khan, A review on the
ENVIRONMENTS4040074. effect of silica to alumina ratio, alkaline solution to binder ratio, calcium oxide +
[21] P. Zhang, X. Zhang, Y. Zhang, Y. Zheng, T. Wang, Gray correlation analysis of ferric oxide, molar concentration of sodium hydroxide and sodium silicate to
factors influencing compressive strength and durability of nano-SiO 2 and PVA sodium hydroxide ratio on the compressive strength of geopolymer concrete,
fiber reinforced geopolymer mortar, Nanotechnol. Rev. 11 (2022) 3195–3206, Silicon 14 (2022) 3147–3162, https://doi.org/10.1007/S12633-021-01130-3.
https://doi.org/10.1515/ntrev-2022-0493. [46] M. Ghanbari, A.M. Hadian, A.A. Nourbakhsh, K.J.D. MacKenzie, Modeling and
[22] R. Shadnia, L. Zhang, P. Li, Experimental study of geopolymer mortar with optimization of compressive strength and bulk density of metakaolin-based
incorporated PCM, Constr. Build. Mater. 84 (2015) 95–102, https://doi.org/ geopolymer using central composite design: a numerical and experimental study,
10.1016/J.CONBUILDMAT.2015.03.066. Ceram. Int 43 (2017) 324–335, https://doi.org/10.1016/J.
[23] K. Korniejenko, W.T. Lin, H. Šimonová, Mechanical properties of short polymer CERAMINT.2016.09.159.
fiber-reinforced geopolymer composites, J. Compos. Sci. 2020, Vol. 4, Page 128 4 [47] Z. Yunsheng, S. Wei, L. Zongjin, Composition design and microstructural
(2020) 128, https://doi.org/10.3390/JCS4030128. characterization of calcined kaolin-based geopolymer cement, Appl. Clay Sci. 47
[24] Z. Abdollahnejad, M. Mastali, M. Mastali, A. Dalvand, Comparative study on the (2010) 271–275, https://doi.org/10.1016/J.CLAY.2009.11.002.
effects of recycled glass–fiber on drying shrinkage rate and mechanical properties [48] M. Lahoti, P. Narang, K.H. Tan, E.H. Yang, Mix design factors and strength
of the self-compacting mortar and fly ash–slag geopolymer mortar, J. Mater. Civ. prediction of metakaolin-based geopolymer, Ceram. Int 43 (2017) 11433–11441,
Eng. 29 (2017) 04017076, https://doi.org/10.1061/(ASCE)MT.1943- https://doi.org/10.1016/J.CERAMINT.2017.06.006.
5533.0001918. [49] K. Komnitsas, D. Zaharaki, Geopolymerisation: a review and prospects for the
[25] L. Qin, Z. Xu, Q. Liu, Z. Bai, C. Wang, Q. Luo, Y. Yuan, Experimental study on minerals industry, Min. Eng. 20 (2007) 1261–1277, https://doi.org/10.1016/j.
mechanical properties of coal gangue base geopolymer recycled aggregate concrete mineng.2007.07.011.
reinforced by steel fiber and nano-Al 2 O 3, Rev. Adv. Mater. Sci. 62 (2023), [50] S. Dadsetan, H. Siad, M. Lachemi, M. Sahmaran, Evaluation of the tridymite
https://doi.org/10.1515/rams-2023-0343. formation as a technique for enhancing geopolymer binders based on glass waste
[26] H.U. Ahmed, A.S. Mohammed, A.A. Mohammed, Engineering properties of (article), J. Clean. Prod. 278 (2021) 123983, https://doi.org/10.1016/j.
geopolymer concrete composites incorporated recycled plastic aggregates modified jclepro.2020.123983.
with nano-silica, J. Build. Eng. 75 (2023) 106942, https://doi.org/10.1016/J. [51] J.G.S. van Jaarsveld, J.S.J. van Deventer, L. Lorenzen, The potential use of
JOBE.2023.106942. geopolymeric materials to immobilise toxic metals: Part I. Theory and applications,
[27] J.L. Provis, J.S.J. van Deventer, Geopolymerisation kinetics. 2. Reaction kinetic Min. Eng. 10 (1997) 659–669, https://doi.org/10.1016/s0892-6875(97)00046-0.
modelling, Chem. Eng. Sci. 62 (2007) 2318–2329, https://doi.org/10.1016/j. [52] R.P. Williams, R.D. Hart, A. Van Riessen, Quantification of the extent of reaction of
ces.2007.01.028. metakaolin-based geopolymers using x-ray diffraction, scanning electron
[28] P.De Silva, K. Sagoe-Crenstil, V. Sirivivatnanon, Kinetics of geopolymerization: role microscopy, and energy-dispersive spectroscopy, J. Am. Ceram. Soc. 94 (2011)
of Al2O3 and SiO2, Cem. Concr. Res 37 (2007) 512–518, https://doi.org/10.1016/ 2663–2670, https://doi.org/10.1111/J.1551-2916.2011.04410.X.
j.cemconres.2007.01.003. [53] J. Kohout, P. Koutník, P. Hájková, E. Kohoutová, A. Soukup, Effect of k/al molar
[29] J.S.J. van Deventer, J.L. Provis, P. Duxson, G.C. Lukey, Reaction mechanisms in the ratio on the thermo-mechanical properties of metakaolinite-based geopolymer
geopolymeric conversion of inorganic waste to useful products, J. Hazard Mater. composites, Polym. (Basel) 13 (2021) 1–20, https://doi.org/10.3390/
139 (2007) 506–513, https://doi.org/10.1016/j.jhazmat.2006.02.044. polym13213754.
[30] J. Davidovits. Geopolymer Chemistry & Applications, 5th ed., Institut [54] A. Dehghani, F. Aslani, N. Ghaebi Panah, Effects of initial SiO2/Al2O3 molar ratio
Géopolymère, Saint-Quentin, 2020. and slag on fly ash-based ambient cured geopolymer properties, Constr. Build.
[31] J. Davidovits, J.L. Sawyer, Early high-strength mineral polymer, U. S. Pat. 4 509 Mater. 293 (2021) 123527, https://doi.org/10.1016/J.
(1985) 985. CONBUILDMAT.2021.123527.
18
[55] J.G.S. Van Jaarsveld, J.S.J. Van Deventer, G.C. Lukey, The characterisation of and hardened-state properties of fly ash-based geopolymer mortars with total
source materials in fly ash-based geopolymers, Mater. Lett. 57 (2003) 1272–1280, replacement of sand, Process Saf. Environ. Prot. 129 (2019) 130–137, https://doi.
https://doi.org/10.1016/S0167-577X(02)00971-0. org/10.1016/j.psep.2019.06.026.
[56] S. Songpiriyakij, T. Kubprasit, C. Jaturapitakkul, P. Chindaprasirt, Compressive [67] N. Cristelo, A. Fernández-Jiménez, C. Vieira, T. Miranda, Á. Palomo, Stabilisation
strength and degree of reaction of biomass- and fly ash-based geopolymer, Constr. of construction and demolition waste with a high fines content using alkali
Build. Mater. 24 (2010) 236–240, https://doi.org/10.1016/J. activated fly ash, Constr. Build. Mater. 170 (2018) 26–39, https://doi.org/
CONBUILDMAT.2009.09.002. 10.1016/j.conbuildmat.2018.03.057.
[57] L.H. Buruberri, D.M. Tobaldi, A. Caetano, M.P. Seabra, J.A. Labrincha, Evaluation [68] X. Gao, Q.L. Yu, H.J.H. Brouwers, Properties of alkali activated slag-fly ash blends
of reactive Si and Al amounts in various geopolymer precursors by a simple with limestone addition, Cem. Concr. Compos 59 (2015) 119–128, https://doi.org/
method, J. Build. Eng. 22 (2019) 48–55, https://doi.org/10.1016/j. 10.1016/j.cemconcomp.2015.01.007.
jobe.2018.11.017. [69] Q. Wan, F. Rao, S. Song, R.E. García, R.M. Estrella, C.L. Patiño, Y. Zhang,
[58] T. Archer de Carvalho, F. Gaspar, A.C. Marques, A. Mateus, Evaluation of the Geopolymerization reaction, microstructure and simulation of metakaolin-based
potential of metakaolin, electric arc furnace slag, and biomass fly ash for geopolymers at extended Si/Al ratios, Cem. Concr. Compos 79 (2017) 45–52,
geopolymer cement compositions, Materials 16 (2023) 2741, https://doi.org/ https://doi.org/10.1016/J.CEMCONCOMP.2017.01.014.
10.3390/ma16072741. [70] S. Yaseri, G. Hajiaghaei, F. Mohammadi, M. Mahdikhani, R. Farokhzad, The role of
[59] W. Li, J. Wang, Y. Zhou, Y. Shao, H. Zhang, J. Zhu, An investigation into the effect synthesis parameters on the workability, setting and strength properties of binary
of particle size distribution on the fluidization behavior of Group C and Group A binder based geopolymer paste, Constr. Build. Mater. 157 (2017) 534–545,
particles, Powder Technol. 398 (2022) 117142, https://doi.org/10.1016/J. https://doi.org/10.1016/J.CONBUILDMAT.2017.09.102.
POWTEC.2022.117142. [71] E.I. Diaz, E.N. Allouche, S. Eklund, Factors affecting the suitability of fly ash as
[60] M. Protić, A. Miltojević, B. Zoraja, M. Raos, I. Krstić, Application of source material for geopolymers, Fuel 89 (2010) 992–996, https://doi.org/
thermogravimetry for determination of carbon content in biomass ash as an 10.1016/J.FUEL.2009.09.012.
indicator of the efficiency of the combustion process, Tech. Gaz. 28 (2021) [72] J. Ghorman Herianto, E. Anastasia, D. Hardjito, Effect of adding acid solution on
1762–1768, https://doi.org/10.17559/TV-20200508110940. setting time and compressive strength of high calcium fly ash based geopolymer,
[61] T. Coradin, D. Eglin, J. Livage, The silicomolybdic acid spectrophotometric method in: AIP Conf Proc, 2017. https://doi.org/10.1063/1.5003525.
and its application to silicate/biopolymer interaction studies, Spectroscopy 18 [73] A. Antoni, S.W. Wijaya, J. Satria, A. Sugiarto, D. Hardjito, The use of borax in
(2004) 567–576, https://doi.org/10.1155/2004/356207. deterring flash setting of high calcium fly ash based geopolymer, Mater. Sci. Forum
[62] APHA, Standard Methods for the Examination of Water and Wastewater, 20th 857 (2016) 416–420, https://doi.org/10.4028/www.scientific.net/MSF.857.416.
Editi, American Public Health Association, American Water Works Association and [74] F. Pacheco-Torgal, J.A. Labrincha, C. Leonelli, A. Palomo, P. Chindaprasirt,
Water Environmental Federation, Washington DC, 1999. Handbook of Alkali-activated Cements, Mortars and Concretes, Elsevier Ltd,
[63] C. 125- Masonry, Methods of test for mortar for masonry - Part 3: Determination of Cambridge, 2015.
consistence of fresh mortar (by flow table), (2006) EN 1015–3:1999/A2:2006. [75] M. Soutsos, A.P. Boyle, R. Vinai, A. Hadjierakleous, S.J. Barnett, Factors
[64] C. Jithendra, S. Elavenil, Effects of silica fume on workability and compressive influencing the compressive strength of fly ash based geopolymers, Constr. Build.
strength properties of aluminosilicate based flowable geopolymer mortar under Mater. 110 (2016) 355–368, https://doi.org/10.1016/J.
ambient curing, Silicon 12 (2020) 1965–1974, https://doi.org/10.1007/s12633- CONBUILDMAT.2015.11.045.
019-00308-0. [76] M. Saeli, R.M. Novais, M.P. Seabra, J.A. Labrincha, Mix design and mechanical
[65] V.C. Prabha, V. Revathi, Geopolymer mortar incorporating high calcium fly ash performance of geopolymer binder for sustainable construction and building
and silica fume, Arch. Civ. Eng. 65 (2019) 3–16, https://doi.org/10.2478/ace- material, : IOP Conf. Ser. Mater. Sci. Eng. (2017), https://doi.org/10.1088/1757-
2019-0001. 899X/264/1/012002.
[66] A. de Rossi, M.J. Ribeiro, J.A. Labrincha, R.M. Novais, D. Hotza, R.F.P.M. Moreira,
Effect of the particle size range of construction and demolition waste on the fresh
19

Construction and 3d Printing Concrete

Uploaded by

Copyright:

Available Formats

Construction and 3d Printing Concrete

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Construction and 3d Printing Concrete

Uploaded by

Copyright:

Available Formats

Construction and Building Materials 412 (2024) 134846

Contents lists available at ScienceDirect

Construction and Building Materials

Optimization of formulation ratios of geopolymer mortar based on

2. Materials and Methods

2.1. Precursors, aggregates, and alkalination mix

In this work, the geopolymerization process was investigated using

Fig. 1. Used GP precursors. MK (left) and BFA (right).

Fig. 3. SEM micrographs of MK (left) and BFA (right).

SiO₂₂ 58.4% 34.8%

Fig. 5. Suspension of GP precursors during alkali dissolution process with 10 M

2.2. Mortar preparation and testing

GP mortars were prepared by first adding the measured amounts of

Fig. 6. Schematic of the alkali dissolution process.

Fig. 7. Examples of tested GP mortar specimens corresponding to each level of efflorescence.

I.T1 5.00 25.00 0.17 25.7% 7.0% 0.4% 14.5% 52.3%

II.T1 70% 16.00 7.9% 18.4% 2.3% 3.1% 14.5% 53.7%

Fig. 13. Efflorescence level for each treatment of stage I.

Fig. 16. Unconfined compressive strength for each treatment of stage I.

Fig. 19. Spreading Diameter for each treatment of stage II.

Fig. 22. Efflorescence level for each treatment of stage II.

mortar with a spreading diameter of about 175 mm, an efflorescence 4. Conclusions

original draft, Methodology, Formal analysis, Data curation, Conceptu­ References

You might also like

original draft, Methodology, Formal analysis, Data curation, Conceptu References