![Figure 4. Pore size distributions obtained from mercury intrusion data for silicate-318 activated fly ash mortars with 28 days of curing 319 320 A remarkable effect of the alkali cation on the pore size distribution is identified in the 321 region of diameters smaller than 1 µm, where an increased pore volume is reported for 322 L300-Na mortars compared with L300-K mortars. In this region two types of porosity can 323 be identified: macropores (50-200 nm) and mesopores (3.6-50nm). Mesopores are present 324 into the aluminosilicate gel network due to short-range ordering which is characteristic of 325 an amorphous material. Macropores are formed during the early stages of 326 geopolymerization, and may transform to mesopores with the progress of the 327 polycondensation of hydrated gels in the binders, as a consequence of the filling of the 328 larger pores with the new reaction products, particularly given the relatively high Ca 329 content of the fly ash used here. On the other hand, pores larger than 200nm in the 330 geopolymer pastes are likely to be associated with the interfacial spaces between partially-331 reacted or unreacted fly ash particles and the geopolymer gel [38, 40]. It is important to 332 note that complete dissolution and reaction of precursors have never been observed in fly 333 ash geopolymer binders, in particular when the solid precursor contains unreactive 334 crystalline phases as is the case for the precursor used in the present study (Figure 1). 335 336 These results indicate that, although L300-Na mortars have higher water content and thus 337 report a higher total porosity than L300-K mortars, a higher extent of polycondensation of 338](/figures/figure-4-pore-size-distributions-obtained-from-mercury-houqa4ex.webp)
![Figure 10. Scanning electron micrographs and corresponding EDS spectra for the binder 537 activated with NaOH/nanosilica-derived activator (L300-Na) 538 539 Dendritic type particles assigned to the ferrite spinels forming the unreacted FA are also 540 identified (Figure 10C), which is coherent with XRD (Figure 5). This is consistent with 541 the observations of Lloyd et al. [60] who observed that iron-rich phases in a fly ash 542](/figures/figure-10-scanning-electron-micrographs-and-corresponding-3iz1gicg.webp)
![Figure 2. Compressive strength of alkali-activated fly ash mortars, as a function of the 246 nature of the activator. Error bars indicate one standard deviation either side of the mean 247 248 As a consequence of the wide range of factors affecting the mechanical strengths of fly ash 249 geopolymer binders, and the physical and chemical differences between fly ash and 250 activator sources used by different research teams, it is difficult to make direct 251 comparisons between the strengths reported by different authors. However, considering 252 the relationship between the molar composition of the precursor and the mechanical 253 strength proposed by Duxson and Provis [34], this ash will be expected to give good 254 strength based on its chemical composition, which is coherent with the results reported 255 here. 256 257 Mortars produced with Na-based activators present higher compressive strength than those 258 with K-based activators. In both cases, slightly higher compressive strengths are obtained 259 when using commercial silicates, especially at early ages. This effect is more notable in 260 samples activated with the L300-K solution, which present compressive strengths up to 261 10MPa lower than the reference specimens (S-K). However, specimens prepared with the 262 modified nanosilica activators present a clear linear trend in the strength development, so 263 that at longer periods of curing (60 days) the compressive strengths of these mortars are 264 comparable to those reported for the reference samples (S-Na and S-K). 265 266 These results are coherent with the mechanisms of activation which are known to take 267 place when different alkali cations are incorporated in the binder. Xu et al. [35] observed 268 that activation with NaOH promotes a higher degree of dissolution of aluminosilicate 269 solids when compared with KOH. This is attributed to the higher capacity of NaOH to 270](/figures/figure-2-compressive-strength-of-alkali-activated-fly-ash-3kn80wbi.webp)
Document downloaded from:
This paper must be cited as:
The final publication is available at
Copyright
http://dx.doi.org/10.1016/j.cemconcomp.2012.08.025
http://hdl.handle.net/10251/37707
Elsevier
Rodriguez Martinez, ED.; Bernal, SA.; Provis, JL.; Paya Bernabeu, JJ.; Monzó Balbuena,
JM.; Borrachero Rosado, MV. (2013). Effect of nanosilica-based activators on the
performance of an alkali-activated fly ash. Cement and Concrete Composites. 35(1):1-11.
doi:10.1016/j.cemconcomp.2012.08.025.
1
Effect of nanosilica-based activators on the performance of an alkali-activated fly ash 1
binder 2
3
Erich D. Rodríguez
1*
, Susan A. Bernal
2†
, John L. Provis
2†*
, Jordi Paya
1
, José M. Monzo
1
, 4
and María Victoria Borrachero
1*
5
6
1
Instituto de Ciencia y Tecnología del Hormigón, Universitat Politècnica de València, 7
46022 Valencia, Spain 8
2
Department of Chemical and Biomolecular Engineering, University of Melbourne, 9
Victoria 3010, Australia 10
†
Current address: Department of Materials Science and Engineering. University of 11
Sheffield, Sheffield, United Kingdom. 12
* To whom correspondence should be addressed. Email: errodnar@disca.upv.es, 13
erichdavidrodriguez@gmail.com (EDR); j.provis@sheffield.ac.uk (JLP) 14
phone: +34-963 877 007 (75648), fax: +34-963 877 569 15
16
Abstract 17
18
This paper assesses the effect of the use of an alternative activator based on 19
nanosilica/MOH (M= K
+
or Na
+
) blended solutions on the performance of alkali-activated 20
fly ash binders. Binders produced with commercial silicate activators display a greater 21
degree of reaction, associated with increased contents of geopolymer gel; however, 22
mortars produced with the alternative nanosilica-based activators exhibited lower water 23
demand and reduced permeability, independent of the alkali cation used. Na-based 24
activators promote higher compressive strength compared with K-based activators, along 25
with a refined pore structure, although K-activated samples exhibit reduced water demand. 26
Zeolite type products are the major crystalline phases formed within these binders. A 27
wider range of zeolites is formed when using commercial silicate solutions compared with 28
the alternative activators. These results suggest that there are variations in the availability 29
of Si in the system, and consequently in the alkalinity, depending on the silicate source in 30
the activator, which is important in determining the nanostructure of the geopolymer gel. 31
32
Keywords: Alkali-activated binders, soluble silicate solutions, nanosilica, X-ray 33
diffraction, scanning electron microscopy 34
Con formato: Inglés (Estados Unidos)
2
1. Introduction 35
36
Interest in the development of alternative building materials such as alkali-activated 37
binders has been promoted by the growth of the building industry, the increased 38
performance requirements placed upon materials, and the higher sustainability criteria 39
applied in construction. Alkali-activated binders represent an attractive alternative for the 40
partial or complete substitution of Portland cement in the production of mortars and 41
concretes, offering comparable performance and cost [1] while reducing greenhouse gas 42
emissions [2]. Specifically, the alkali-activation of low calcium fly ashes (FA) has been 43
extensively assessed over the past decades because these binders exhibit mechanical 44
performance comparable to that reported for Portland cement, and because FA is an 45
industrial by-product available worldwide. FA is produced in high amounts, especially in 46
countries such as India and China, where an increased demand for cement is expected in 47
the coming years. Despite the promising properties of these binders, and their ongoing 48
commercialization, there are technological challenges associated with the variability of the 49
raw materials from different sources, and the low sustainability of the current alkali-50
activators used [3]. 51
52
In the activation of aluminosilicate precursors such as FA, the nature of the activator 53
solution plays a key role in determining structural and mechanical performance. The most 54
relevant characteristics related to the alkali activator are: the type of alkaline salt (usually 55
silicate or hydroxide) [4-6]; the method of addition of the alkaline component (as a 56
solution or in solid-state) [7-9], and the dosage of the alkali component, usually expressed 57
as molar ratios considering the overall composition of the raw material. Additionally, it 58
has been reported [4, 10-11] that the alkali cation supplied by the alkaline solution 59
influences the first stages of binder formation, and consequently the mechanical 60
performance of the final products. The alkali-activators conventionally used are sodium or 61
potassium hydroxides, and/or sodium or potassium silicates [12]. Activation with K-62
containing solutions often leads to increased compressive strength development when 63
compared with Na-containing solutions, where the size and charge density of the alkali 64
cation play an important role in controlling the rate and extent of condensation during the 65
polycondensation or crystallization process [13]. However, these effects are also 66
dependent on the chemical and physical nature of the solid precursor used [5]. 67
68
3
The embodied energy associated with the preparation of an alkali-activated binder is often 69
estimated based on the contributions of the precursor and the alkaline activator. Some 70
studies have analyzed the real energy consumed in fly ash based geopolymer production 71
[14-15], identifying that the major contribution is associated with the type and 72
concentration of the alkaline activator, and is between 0.5 MJ/t and 3.4 MJ/t. This is 73
mainly related to the complex chemical processes required to manufacture these 74
substances. For instance, the production of sodium silicate involves the calcination of 75
sodium carbonate (Na
2
CO
3
) and quartz sand (SiO
2
) at temperatures between 1400-1500ºC, 76
producing large amounts of CO
2
as a secondary product [16-18]. This substantially 77
increases the embodied energy of silicate-activated binders, reducing sustainability. 78
However, sodium silicate (Na
2
O∙rSiO
2
, sometimes referred to as ‘waterglass’) is the 79
activator which generally provides the highest compressive strength development at early 80
ages of curing, and exhibits some technological advantages compared with other activators 81
such as NaOH. 82
83
This then provides motivation for the examination of the current activators used in 84
geopolymerization processes in terms of their sustainability, and the assessment of 85
alternatives that can contribute to reducing the embodied energy of these binders. Some 86
studies assessing alternative activators based on modified silica fume (MSF) have been 87
conducted [19-21]. Likewise, agro-industrial wastes, as well as other silica sources, have 88
been studied as alternative alkali-activators in order to obtain a more environmentally 89
friendly alkali-activated binder with lower cost [22-27]. These results reveal that this 90
alternative activator promotes similar or even better mechanical performance when 91
compared with conventional activators. 92
93
Based on this background, the aim of this paper is to study alkali-activated low calcium fly 94
ash binders, activated by chemically modified nanosilica. The effect of the alkali cation 95
(Na
+
and K
+
) on the structure of the binders is studied by X-ray diffraction (XRD), 96
thermogravimetry and electron scanning microscopy (SEM/EDS). Compressive strength 97
testing and mercury intrusion porosimetry (MIP) are conducted on mortar samples based 98
on the binders produced, in order to generate a better understanding of the effect of the 99
type of activator used, the gel structure formed, and the mechanical strength development 100
of the materials. 101
102
4
2. Experimental program 103
104
2.1. Materials 105
106
The binders studied here were synthesized using a fly ash (FA) from Teruel Power Station 107
in Andorra, Spain, with a specific gravity of 2520 kg/m
3
and a chemical composition as 108
shown in Table 1. The FA was mechanically treated in a high impact mill (Mill2 109
Gabbrielli) to increase its reactivity. The particle size range determined by laser 110
granulometry was 0.2-80 μm, with a mean particle size of 15 μm, and a specific surface of 111
1130 m
2
/kg. 112
113
Table 1. Chemical composition of the fly ash from X-ray fluorescence analysis. LOI is 114
loss on ignition at 950 ˚C 115
116
The X-ray diffraction pattern of the FA (Figure 1) shows that the major crystalline phases 117
present are quartz (SiO
2
; Powder Diffraction File (PDF) card # 00-046-1045), mullite 118
(Al
6
Si
2
O
13
; PDF# 00-015-0776), and Fe-rich phases such as hematite (Fe
2
O
3
; PDF# 00-119
033-0664), iron silicate (Fe
7
SiO
10
; PDF# 00-022-1118), and some ferrite spinels 120
(magnetite - Fe
3
O
4
; PDF#00-019-0629, with and without substituent elements such as Mg 121
and Al on both Fe
2+
and Fe
3+
sites). The presence of these phases is coherent with the high 122
content of iron in the fly ash and has been previously observed in other fly ashes [28-30]. 123
It is important to note that the ferrite spinels in the FA play an important role in the 124
potential hosting of heavy metals, as Fe
3+
sites can be substituted by trivalent
cations such 125
as Cr
3+
[31]. 126
127
Figure 1. Cu-K
α
diffractogram of the fly ash after mechanical treatment 128
129
As alkali-activators, four alkaline solutions derived from hydroxide solutions and soluble 130
silica sources were used. A commercial sodium silicate (SS) from Merck and a potassium 131
silicate (SK) from IQE were used as reference soluble silica sources (Table 2). Two 132
additional soluble silicates based on blends of a nanosilica suspension from H.S. Starck 133
(L300, specific surface 300 m
2
/g; Table 2) were also assessed. Alkali-activators were 134
prepared by the dissolution of analytical sodium hydroxide (99 wt%) or potassium 135