Production of sodium silicate powder from waste glass cullet for alkali

activation of alternative binders

Vinai, R., & Soutsos, M. (2019). Production of sodium silicate powder from waste glass cullet for alkali activation
of alternative binders. Cement and Concrete Research, 116, 45-56.
https://doi.org/10.1016/j.cemconres.2018.11.008

Published in:
Cement and Concrete Research

Document Version:
Peer reviewed version

Queen's University Belfast - Research Portal:

Link to publication record in Queen's University Belfast Research Portal

Publisher rights
Copyright 2019 Elsevier Ltd.
This manuscript is distributed under a Creative Commons Attribution-NonCommercial-NoDerivs License
(https://creativecommons.org/licenses/by-nc-nd/4.0/), which permits distribution and reproduction for non-commercial purposes, provided the
author and source are cited.

General rights
Copyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or other
copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated
with these rights.

Take down policy

The Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made to
ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in the
Research Portal that you believe breaches copyright or violates any law, please contact [email protected].

Open Access
This research has been made openly available by Queen's academics and its Open Research team. We would love to hear how access to
this research benefits you. – Share your feedback with us: http://go.qub.ac.uk/oa-feedback

Download date:08. Apr. 2024

Production of sodium silicate powder from waste glass cullet for alkali

activation of alternative binders.

Raffaele Vinai*1,2 and Marios Soutsos1

1
School of Natural and Built Environment, Queen’s University Belfast, BT9 5AG United Kingdom
2
College of Engineering, Mathematics and Physical Sciences, University of Exeter, Harrison Building, EX4

4QF United Kingdom.

*Corresponding author: [email protected], tel. +44 (0) 1392 72 3627.

Declarations of interest: none

Abstract

A simple process to produce sodium silicate powder from glass cullet has been developed. A mixture of glass

powder, sodium hydroxide powder, and water was heated at temperatures of 150 to 330 °C. The effects of glass

to NaOH ratio, temperature and duration, inclusion of water and fineness of NaOH were investigated. Fly ash and

fly ash/GGBS blends were the precursors for alkali activated binder (AAB) mortars produced with this sodium

silicate. Compressive strengths were similar to or better than those obtained with commercially available sodium

silicate and sodium hydroxide solutions. FT-IR tests suggested that the reactivity of the glass derived sodium

silicate powder was related to the number of non-bridging oxygen atoms in the silicate structure. Cost comparison

between AAB and Portland cement concretes gave similar results for normal strength concretes (35 MPa). AAB

concretes with higher strengths (50 and 70 MPa) can be cheaper than equivalent traditional concrete.

Keywords: alkali activated cement (D), glass (B), fly ash (D), granulated-blast furnace slag (D), waste-derived

activator.

1. Introduction

Alkali activated binders (AAB) rely on the reaction of an aluminosilicate solid powder (called precursor) with an

alkali metal (called activator). There is extensive literature on potential activators. The activator must be able to:

(a) be efficient in triggering and ensuring the full reaction with the precursor in the desired time (too long reaction

times might lead to low early age strength, whilst too short reaction times can result in fast setting problems); (b)

1
be available at low cost and in sufficient quantity; (c) have a low environmental impact; (d) allow simple and cost-

effective use. Among all the available options for activators (hydroxides, silicates, sulfates, carbonates, acidic

solutions), the use of alkali hydroxide (either sodium or potassium) combined with an alkali silicate (again either

sodium or potassium) [1] is favoured as it has been shown to give high compressive strengths when used with fly

ash and slag as precursors. However, the use of alkali hydroxide and silicate solutions has also disadvantages such

as:

1. They are difficult to handle on site as they are highly alkaline liquid solutions, and this poses health and

safety risks.

2. The costs, embodied energy and environmental impact associated with the production of the activators [2-

6], especially those of the sodium silicate which is usually produced by melting sodium carbonate and silicon

dioxide at a temperature of 1200 – 1400 °C. This industrial process emits large quantities of CO2.

Other production methods are available, e.g. dissolution of quartz sand with sodium hydroxide in a reactor, but

the energy requirements are again usually high [2,7-9]. The development of a solid activator, which can be

premixed with a precursor and thus obtaining a “one-part” alkali activated binder where only water is required to

be added to the mix, is highly desired by the industry [1]. A recently published review on one-part alkali-activated

materials is available [10].

Anhydrous sodium silicate in powder form is available on the market but it needs pressure reactor vessels to get

it properly dissolved. Hydrous powders in principle could be dissolved under normal atmospheric pressure

conditions [11], but Rees et al. [12] found that the availability of free Si from solid powder was lower than from

liquid silicates. The development of a powder (solid) activator is therefore a priority in the research agenda [1].

Attempts to find ways of reducing the cost and environmental impact of activators have mainly focused on

hydrothermal processes to produce silicate solutions from by-products such as condensed silica fume [13-14] and

rice husk ash [15-18]. These have the potential of reducing the global warming potential of AAB concretes by

more than 70% compared to Portland cement-based concrete [19]. However, the cost of AAB concretes is not

significantly reduced and this remains higher than Portland cement based concretes.

Waste glass cullet has also been identified as a potential source of silicate, this being 70 – 75% of amorphous

silica. Glass cullet can be used for the production of new glass but colour, composition, and contamination reduce

the amount that can be reused. The percentage of glass rejected by the so-called “closed loop” due to impurities

has been estimated to be between 6% and 15% in Italy [20]. A study carried out in 2003 indicated that, if UK

would meet the recycling target of 60% and under the hypothesis of consuming the recycled glass nationally (i.e.

2
without shipping the excess glass cullet abroad), 400,000 tonnes of glass cullet would still be available for other

uses than production of glass containers [21]. This recycling target of 60% was achieved and even exceeded in

recent years [22], thus, the availability of glass cullet in UK is significant. Even the process of recycling of glass

produces a waste stream which is a very fine powder that is collected by dust extraction units. Due to its chemistry

and its availability, waste glass is therefore a good candidate for the production of AAB. Some recent studies

describe its utilisation as total or partial replacement of usual precursors, suggesting that satisfactory strength can

be achieved by activation of 100% waste glass fibres with NaOH only [23] or by substituting up to 60% of fly ash

in blended fly ash-ground granulated blast furnace slag (GGBS) system [24].

Focussing on the use of waste glass for the production of low cost, environmentally friendly silicate solution,

some patents [25-27] and production methods exist:

 Hydrothermal methods involve heating of the glass in an alkaline solution. Factors influencing the

dissolution rate were (a) temperature, (b) glass composition, (c) particle size and (d) type of alkaline

cation (Na or K) in the solution [28-31]. Attempts to increase the dissolution focussed on increasing the

temperature to between 150 and 250 °C which required pressure reactor vessels as the self-generated

pressure was up to 30 atm. The silica dissolution was about 72% [32]. There are however concerns with

regards to the corrosive nature of the solutions that can damage the pressure reactor vessels.

Temperatures less than 100 °C proved that the process works [33 – 38], but the dissolution rate was low,

e.g., 26% of SiO2 from the 70% in the glass. Although these were used successfully for slag activation

[33, 37-38] their molarity was low compared to commercial solutions and thus the maximum potential

strength of the precursor could not be achieved. A literature analysis of hydrothermal methods for the

extraction of silicate from waste in aqueous alkali solutions was detailed in a previous study [18].

 Thermochemical or fusion methods involve mixing glass and NaOH powders and heating the mixture to

very high temperature, e.g., 500°C [39], 650 °C [40], and 700 to 1300 °C [41]. Conversion of glass to

sodium silicate was high but the solubility of the produced powder at ambient pressure was low thus

requiring reheating at 175 °C for at least one hour and still not fully dissolving. The procedure for

obtaining pure sodium silicate from sand is considered “simple” chemistry and involves mixing sand

with sodium hydroxide in a 2:1 mass ratio and heating to 320 °C for at least 10 minutes, although one

hour is recommended. The melting point of the NaOH is about 318 °C and the following chemical

reaction is produced:

2NaOH + SiO2 → Na2SiO3 + H2O (1)

3
This procedure has the advantage that:

(a) It produces a powder rather than a solution. This makes a one-part alkali activated binder system possible.

(b) It reduces significantly the cost for the production of the activator and thus subsequently reducing the

cost of alkali activated concretes.

(c) It has a lower carbon footprint as it uses low temperature process.

The research aimed at adapting the above procedure so that ground glass cullet could be used with the aim of

lowering even further the required temperature for the process enabling the use of ordinary laboratory ovens. The

work was carried out in four phases:

1. Assessment of the effect of variables such as (a) the SiO2 : Na2O ratio, (b) the inclusion of water, (c) the

temperature and the duration of the process, (d) the fineness of NaOH on the chemical composition of

the sodium silicate powder.

2. Determination of the efficiency of the sodium silicate powder when used as an activator for AAB.

Comparison of compressive strength of mortars produced using (a) glass-derived sodium silicate powder

and (b) commercially available sodium silicate solution.

3. Microstructural investigation of the physico-chemical properties of the produced powder in order to

understand its reactivity potential.

4. A cost analysis for concretes to assess the economic benefits arising from the use of the glass-derived

sodium silicate powder. Comparison of AAB concretes thus produced with those with commercially

available sodium silicate solution and also Portland cement based ones.

2. Materials and experimental methods

Glass cullet was sourced from a local recycler in Northern Ireland and had glass chips of different colours and

sizes as it was from domestic glass containers (bottles). It was milled by means of a Retsch PM400 Ball Mill,

capable of grinding around 440 g of glass (110 g in each of the four 500 ml hardened steel grinding jars, each

loaded with ten hardened steel grinding balls, 15 mm diameter) at a speed of 300 rpm. Grain size distribution was

obtained by means of laser diffraction particle size analyser (Malvern Panalytical Mastersizer 2000). The D50

value (i.e. the value of the particle size at 50% in the cumulative distribution) showed that the milling time

improved the fineness of the material. The D50 reduced from around 17 µm (milling for 5 minutes) to about 12

µm when the material was milled for 10 minutes, see Figure 1. Grinding the material for 10 minutes at 300 rpm

was considered suitable and adopted in this study. The oxide composition of the glass powder was obtained from

X-ray fluorescence (XRF) and it is shown in Table 1. Whilst XRF results can be considered satisfactory reference

4
 values, traces of other elements, such as Cr and Ba, which are frequently found in glass, might have not been

detected. Analysis of trace elements with other techniques such as laser-induced breakdown spectroscopy [42]

would give further insight on the chemistry of the glass, but this was not possible to be done for this work.

Fig. 1. Particle size distribution of used glass powder.

Table 1. Oxide composition (in mass %) of the glass powder, fly ash and GGBS used in the study.

Element SiO2 TiO2 Al2O3 Fe2O3 MnO MgO CaO Na2O K2O P2O5 SO3 LOI
Glass 71.51 0.07 1.74 0.34 0.028 1.34 10.73 13.29 0.64 0.01 0.089 0.27
Fly ash 46.78 1.05 22.52 9.15 0.05 1.33 2.24 0.89 4.09 0.17 0.90 3.57
GGBS 36.5 0.5 10.4 0.7 0.4 8.1 42.4 0.5 0.2 0.00 0.3 -

Commercial grade sodium hydroxide (Atznatron caustic soda from AkzoNobel, Germany) in microprills was used

for the reaction with the ground glass cullet.

An electric furnace (Carbolite RHF 1600, UK) was used for heating samples and preparing powders for X-ray

diffraction tests and for Fourier Tranform Infrared spectroscopy (FTIR) tests. Powders for mortar tests were

produced either with a SNOL 39/1100 electric furnace (UK) at 330 °C process temperature or with conventional

laboratory ovens for lower temperatures.

X-ray diffraction (XRD) tests were carried out with a PANalytical X’Pert PRO diffractometer using pure Cu Kα

1 radiation with wavelength 1.54 Å. The X-ray generator was set to 40 kV and 40 mA, the recorded angular range

was 5 to 65 or 70° (2θ) with a step close to 0.017°. PANalytical X’Pert Highscore software was used for data

interpretation and Rietveld analysis.

5
FT-IR was carried out using a Jasco 4100 series FTIR Spectrometer with Attenuated Total Reflectance attachment

(germanium crystal), recording data in transmittance mode over the range 650 to 4000 cm-1, with a data interval

0.964 cm-1. Jasco software was used for data interpretation.

Mortars were prepared by mixing the required amounts of binder, activator and sand in a 5 l planetary mixer. A

sand:binder ratio of 2.75 was used. Binders consisted either of a blend of fly ash and ground granulated blast

furnace slag (GGBS) or neat fly ash. Fly ash was sourced from a power plant in England (Drax Power Station),

whilst GGBS was supplied by Ecocem (Ireland). Chemical composition of fly ash was obtained with XRF

analysis, whilst chemical composition of GGBS was provided by the producer. Oxide compositions are shown in

Table 1. The powder developed in the study was used for activating the binders for mortars. Control mortar mixes

were produced using commercially available sodium silicate solution (Fisher Scientific with SiO2 content equal

to 25.5% and Na2O content of 12.8%, i.e. SiO2/Na2O = 2) and a sodium hydroxide solution prepared by dissolving

the Atznatron caustic soda in water at 30% mass concentration. Natural sand with grain size distribution from 0

to 4 mm was used. Two 50 mm cubes for each mix were tested at 1, 7 and 28 days for compressive strength.

Cubes were kept wrapped in cling film for the first 24 hours after which they were unwrapped and left to air cure.

Samples produced with fly ash/GGBS blends were cured at room temperature, whilst samples with neat fly ash

were cured in an oven at 70 °C until testing.

3. Results and discussion

3.1. Optimisation of thermochemical process for the production of sodium silicate powder

A preliminary trial was carried out aiming to reproduce the conditions necessary for the reaction described in Eq.

(1) to develop. A dry mix of NaOH and glass powder was prepared with SiO2:Na2O mass ratio equal to 1 and

processed in a furnace at a temperature of 330 °C (i.e. slightly over the NaOH melting temperature) for two hours

(labelled 1-330-2). The various steps in the procedure are shown in Figure 2.

Fig. 2. Procedure steps for the production of the activating powder.

6
XRD analysis on the produced activating powder, see Figure 3, showed that the thermochemical process

transformed the material from a fully amorphous (diffraction pattern with a broad hump and no peaks) to a

crystalline powder with a number of peaks identified as sodium silicates with different SiO2:Na2O ratios, calcium

aluminium oxide and sodium carbonate. The main mineral phase was Na2SiO3 and this indicated that the

procedure was successful.

Fig. 3. XRD patterns of the processed powder activator compared to the raw glass powder.

A number of parameters were therefore subsequently investigated for optimising the sodium silicate production

process (minimising the unreacted glass powder): (1) the SiO2 : Na2O ratio (varied from 4:1 to 1:1), (2) the

presence of water in the initial mix, (3) the process temperature, 150, 250, 330 and 450 °C, (4) the process duration,

1, 2 and 4 hours, and the fineness of NaOH, i.e. prill or ground. The various samples investigated are shown in

Table 2. Labels indicate the SiO2 : Na2O ratio followed by the process temperature, duration and whether the

sample was wet or dry, and whether the NaOH was ground or not.

The SiO2 : Na2O molar ratio was calculated assuming that the SiO2 content in the glass was equal to 70% (see

Table 1) and considering that the Na2O content in NaOH is 77.5%.

7
Processed activator powders were analysed with XRD in order to assess the nature and estimate the quantity of

obtained silicates, as well as the degree of reaction. The presence of a hump is an indicator of the presence of

amorphous material in the powder (and therefore presence of unreacted glass), whilst a nearly flat baseline

suggests almost fully crystalline structures (and therefore full or almost full reaction, although some reaction

products might be amorphous as well).

Table 2. Parameters investigated for the thermochemical process assessment.

Sample SiO2 : Na2O Water included Process Process NaOH fineness

label (mol) (yes/no) temp. (°C) time (h) (prills, ground)
4-330-2 4:1 No 330 2 Prills
2-330-2 2:1 No 330 2 Prills
1-330-2 1:1 No 330 2 Prills
2-330-2w 2:1 Yes 330 2 Prills
1-330-1w 1:1 Yes 330 1 Prills
1-330-2w 1:1 Yes 330 2 Prills
1-150-2w 1:1 Yes 150 2 Prills
1-250-2w 1:1 Yes 250 2 Prills
2-450-2 2:1 No 450 2 Prills
1-330-1 1:1 No 330 1 Prills
2-330-4 2:1 No 330 4 Prills
2-330-2f 2:1 No 330 2 Ground

The estimation of the amorphous content of the powders was carried out according to the method described by

Tiainen et al. [43], which uses the area under the halo in a XRD pattern for semi-quantitative determination of the

amorphous content. The method involves the following steps:

- Calculation of the integrated area under the baseline of the XRD pattern (Ab) in the range 10° to 70° (2θ).

- Calculation of the total integrated area under the XRD pattern (At) in the range 10° to 70° (2θ).

- Calculation of the ratio Ra = Ab/At.

A linear correlation exists between the ratio Ra and the amorphous content of the powder. Tiainen et al. suggested

that a calibration curve should be built using known amount of amorphous contents in the investigated matrix.

However, in their study it was found that significant variation in the matrix composition did not result in major

differences in the calibrating curve [43]. For this investigation, the Ra calculated for raw glass was used as

calibration point for 100% amorphous content, whilst the mix 1-150-2w, being the mix resulting in the flattest

baseline (determined as deviation from the horizontal line), was used as second calibration point, estimated from

[43] at about 30% amorphous content. A sensitivity analysis carried out by varying the assumed amorphous

8
 content of the second calibration point by ± 10% suggested that variations were limited and acceptable under the

assumptions made. Results from this exercise are shown in Table 3.

Table 3. Estimation of the amorphous content of investigated samples.

Sample label Ra (%) Estimated amorphous

content (%)
Raw glass 93.8 100
4-330-2 88.9 90
2-330-2 80.8 70
1-330-2 76.0 60
2-330-2w 75.0 60
1-330-1w 64.2 35
1-330-2w 66.2 40
1-150-2w 61.8 30
1-250-2w 64.9 35
2-450-2 83.4 75
1-330-1 75.2 60
2-330-4 84.6 80
2-330-2f 82.1 75

The identification of mineral phases was carried out using X’Pert software, which was also used for the Rietveld

analysis for the quantification of the mineral phases. Obtained results were then corrected according to the

estimation of the amorphous content, and quantities of mineral compounds were therefore assessed. Due to the

semi-quantitative nature of the adopted approach, results are only precise to +/- 10%. Result are shown in Table

4.

Table 4. Mineral composition from Rietveld analysis on processed powder activator.

Mineral 4-330-2 2-330-2 1-330-2 2-330-2w 1-330-1w 1-330-2w 1-150-2w 1-250-2w 2-450-2 1-330-1 2-330-4 2-330-2f
5% 20% 24% 36% 63% 54% 67% 61% 7% 29% 11% 18%
Na2SiO3

Na4SiO4 1% 0% 3% 0% 1% 0% 0% 0% 3% 0% 1% 2%

0% 1% 4% 0% 0% 0% 0% 0% 11% 0% 1% 0%
Na6Si2O7
2% 8% 9% 3% 0% 6% 0% 0% 2% 10% 6% 4%
Na2CO3

CaAl2O4 1% 1% 1% 1% 1% 0% 3% 2% 0% 1% 1% 1%

0% 0% 0% 0% 0% 0% 0% 2% 2% 0% 0% 0%
Na2CaSiO4
90% 70% 60% 60% 35% 40% 30% 35% 75% 60% 80% 75%
Amorphous

Two main factors were found to be key parameters for improving the efficiency of the reaction (i.e. maximising

the crystalline content of the powder): (a) the amount and availability of Na2O (from NaOH) and (b) the addition

9
of water in the blend before heating. Samples with SiO2 : Na2O ratio equal to 1 had the lowest amorphous content,

indicating high degree of reaction, see Figure 4. Attempts to obtain sodium silicate with a higher silica ratio, by

using high amounts of glass powder, resulted in only partial reaction, e.g. sample 4-330-2 (with SiO2 : Na2O ratio

equal to 4) had a very high amorphous content whilst samples with SiO2 : Na2O ratios of 2 had intermediate

amorphous contents.

XRD patterns indicated that the processed activating powder had higher crystalline content when a small amount

of water was added to the NaOH/glass powder blend to produce a paste before it was heated. The presence of

water in sample 2-330-2w resulted in the highest reflection peaks for Na2SiO3, although certain amorphous

fraction was still observed. Samples 1-330-1w and 1-330-2w, which combined the best SiO2 : Na2O ratio with the

improved mixing procedure, resulted in reacted powder with very high content of sodium silicates, mainly in the

form of Na2SiO3, see Figure 5. It appears that water addition led to the partial dissolution of NaOH which then,

after water evaporation due to the heating, precipitated as a thin layer around the glass powder grains. The micro

(or nano) scale of the deposited NaOH layer, and thus its improved contact with glass powder grains, might have

resulted in an improved reaction.

Fig. 4. XRD patterns of the processed activating powders – effect of SiO2 : Na2O ratio.

10
 Fig. 5. XRD patterns of the processed activating powders – effect of water content.

Higher process temperature resulted in more complex sodium silicate forms, with increased presence of Na-rich

silicates such as Na4SiO4 (i.e. 2Na2OꞏSiO2) and Na6Si2O7 (i.e. 3Na2Oꞏ2SiO2). Sample 2-450-2 (heated at 450 °C)

was mainly Na6Si2O7, but with small quantities of Na2SiO3, Na4SiO4, and sodium calcium silicate (Na2CaSiO4).

Sample 2-330-2 (heated at 330 °C) was mainly Na2SiO3. Phase diagrams for Na2O - SiO2 systems [44,45] suggest

that a temperature above 700 °C is needed, with an excess of SiO2 (about 70%) in the mix in order to obtain

sodium silicates with high SiO2 content, e.g., Na6Si8O19 (3Na2Oꞏ8SiO2).

Surprisingly, when the temperature used was below 330 °C, i.e. at 150 °C and 250 °C, obtained XRD patterns

were very similar to the ones obtained at 330 °C, see Figure 6. A possible explanation for this is that the NaOH

deposited on the glass powder grains after evaporation of water (as previously described) was at a nano-size scale

(as the NaOH was initially dissolved), and the NaOH melting point was consequently suppressed (due to the size

of the deposited particles) [46].

11
 Fig. 6. XRD patterns of the processed activating powders – effect of heating temperature.

No significant difference was observed among XRD patterns of mixes with the same composition and temperature

but different heating duration, i.e. 1, 2 and 4 hours (i.e. comparing 1-330-2 with 1-330-1, 2-330-2 with 2-330-4,

and 1-330-2w with 1-330-1w). Therefore, duration of one hour for heating seems to be sufficient.

Only a marginal increase of Na2SiO3 was observed when the NaOH was ground. There was also some minor

presence of Na4SiO4 in sample 2-330-2f when compared to sample 2-330-2.

The above discussed results indicate that the most important process parameters to achieve the optimum reaction

(i.e. processed activating powder composed of crystalline sodium silicates) were: (a) the availability of enough

NaOH for achieving a theoretical SiO2 : Na2O molar ratio equal to 1, i.e. a glass powder : NaOH mass ratio equal

to 11 : 10, (b) converting the powder into paste by adding water before heating, and (c) heating for 1 hour at a

temperature between 150 °C and 330 °C. These optimum process parameters were therefore used for producing

the activation powder used for the subsequent work.

3.2. Efficiency of the developed powder activator

In order to determine the efficiency of the powder activator, the compressive strengths of AAB mortars with the

developed powder activator were compared to those with commercially available sodium hydroxide and sodium

silicate.

12
The developed sodium silicate powder was used for activating the binder. The amount of powder activator needed

for the mortar mixes was determined based on the required alkali dosage (M+) and alkali modulus (AM). M+ is

the percentage mass ratio Na2O/binder, and AM is the mass ratio Na2O/SiO2. In the literature, silica modulus (SM)

is often used, i.e. the ratio SiO2/Na2O, which corresponds to 1/AM. The required SiO2 mass was calculated based

on the SiO2 mass content in Na2SiO3. AM higher than 1 (i.e. SM < 1) required not only the powder activator but

also extra Na2O, i.e., the Na2O exceeding the quantity available from the dissolution of the sodium metasilicate,

and this was provided by adding NaOH solution (30% mass concentration).

The powder activator was dissolved in the required water volume (i.e. mass of water required for desired water to

solid ratio (w/s), minus the water included in the NaOH solution) just before mixing.

According to the semi-quantitative determination of the chemical composition of powder shown in Table 4, the

activators produced with the optimised process (i.e., samples with SiO2 : Na2O ratio equal to 1 heated for 1 hour

at a temperature between 150 °C and 330 °C) had about 65% in mass of silicates. However, the variability

associated with the determination of the amorphous content as well as the possible presence of reacting silicates

in amorphous form could lead to underestimate the actual amount of silicates. For this reason, a less conservative

value of 80% was also investigated. An “efficiency factor” (EF) defined as the ratio between the available SiO2

and the mass of the powder activator ranging from 65% to 80% was assumed and tried in four series of mortars

with sand-to-binder ratio = 2.75. Mixes were labelled with the following sequence: fly ash content – AM – EF –

powder production temperature. When the powder was added in solid form in the mix, the letter “s” was added

after the EF value.

1. Series I – Mixes 60-1.25-65-330 and 60-1.25-80-330 had 60%/40% fly ash/GGBS binder blend with

activator dosages M+ = 7.5% and AM = 1.25 (i.e. SM = 0.8) and activating powder produced at 330 °C.

The aim of this series was to compare strengths at AM = 1.25 which has previously been shown to result

in the highest possible compressive strength [47]. As such, sodium hydroxide solution in addition to the

activating powder was required for achieving AM =1.25. Efficiency factors of 65% and 80% were

investigated.

2. Series II – Mixes 60-1.0-80-330, 60-1.0-65-330 and 60-1.0-65s-330 had 60%/40% fly ash/GGBS binder

blend with activator dosages M+ = 7.5% and AM = 1.0 (i.e. Na2O/SiO2 = 1, corresponding to SM = 1).

The aim was to use only activating powder without additional sodium hydroxide solution. Mix 60-1.0-

65-330 was repeated by adding the dry powder directly in the mixer with fly ash and GGBS, and then

adding the required water (mix 60-1.0-65s-330).

13
 3. Series III – Mixes 100-1.0-65-330 and 100-1.0-80-330 were with neat fly ash, M+ = 7.5% and AM = 1.0

(i.e. SM = 1). The aim was to investigate the efficiency of the activating powder on neat fly ash mixes.

4. Series IV – Mixes 60-1.0-65-150 and 100-1.0-65-150 were aimed at assessing the efficiency of activation

powder produced at 150 °C instead of 330 °C. Both fly ash/GGBS blend and neat fly ash mixes were

investigated.

Series I: 60%/40% fly ash/GGBS mixes with AM=1.25. Series I aimed at assessing the efficiency factor (EF) of

the activating powder at an AM=1.25 (SM = 0.8) with 60%/40% fly ash/GGBS blends. Table 5 shows the mortar

proportions and Figure 7 shows the obtained compressive strengths at 1, 7 and 28 days. The compressive strengths

of the powder activator at 1-day were much lower than those with the commercial solution. However, the average

compressive strength at 28-days improved considerably and they were comparable with the strength obtained with

the commercially available solution. This indicated that the powder activator was participating in the reaction

(higher strengths than NaOH only activated mortars) but perhaps its dissolution was slow and thus contributed to

strength more at 28-days rather than early ages, i.e. 1- and 7-days. The additional NaOH added to these mixes to

obtain an AM = 1.25 (SM = 0.8) was suspected of interacting with the activator powder. The next series of mixes

therefore investigated an AM = 1.0 (SM = 1.0) which could be achieved with powder activator only.

Table 5. Mix proportions of fly ash/GGBS mortars with AM = 1.25 for Series I.

Control
Parameter 60-1.25-65-330 60-1.25-80-330 NaOH only
(Commercial SS)
M+ (%) 7.5 7.5 7.5 7.5
AM (-) 1.25 1.25 1.25 ∞
EF (%) 75% 90% 100% 100%
w/s ratio (-) 0.387 0.387 0.385 0.386
Fly ash (kg/m3) 316.6 318.5 323.1 330.3
GGBS (kg/m3) 211.1 212.3 215.5 220.2
Sand (kg/m3) 1451.1 1459.8 1480.6 1513.9
Act. powder (kg/m3) 95.3 79.9 - -
SS solution (kg/m3) - - 126.7 -
NaOH solution (kg/m3) 29.5 29.7 103.3 178.4
Added water (kg/m3) 211.1 212.3 86.1 103.5

14
 Fig. 7. Compressive strength results of Series I mortar cubes.

Series II: 60%/40% fly ash/GGBS mixes with AM=1.0. A “one-part” alkali activated binder production (i.e. dry

mix of precursor/activator blend with only add water needed) becomes possible with the powder activator if AM

= 1.0 (SM = 1.0) and not higher is what is required. Table 6 shows the mix proportions of 60%/40% fly ash/GGBS

binder blends investigated and Figure 8 shows the compressive strengths at 1, 7 and 28 days.

The powder activator was dissolved in water for mixes with EF of 70 % and 90%. The mix with EF of 70% was

repeated but with the powder activator added directly to the mixer in solid form.

15
 Table 6. Mix proportions of fly ash/GGBS mortars with AM = 1.0 for Series II.

Control
Parameter 60-1.0-65-330 60-1.0-65s-330 60-1.0-80-330 NaOH only
(Commercial SS)
M+ (%) 7.5 7.5 7.5 7.5 7.5
AM (-) 1.0 1.0 1.0 1.0 ∞
EF (%) 70% 70% 90% 100% 100%
w/s ratio (-) 0.386 0.386 0.386 0.385 0.386
Fly ash (kg/m3) 312.5 312.5 315.8 321.4 330.3
GGBS (kg/m3) 208.3 208.3 210.5 214.3 220.2
Sand (kg/m3) 1432.2 1432.2 1447.4 1473.1 1513.9
Act. powder (kg/m3) 126.0 126.0 99.0 - -
SS solution (kg/m3) - - - 157.6 -
NaOH solution (kg/m3) - - - 86.8 178.4
Added water (kg/m3) 231.2 231.2 233.7 80.4 103.5

Fig. 8. Compressive strength results of Series II mortar cubes.

Despite the different EF of 65 and 80%, and thus different quantities of powder activator, similar compressive

strengths were obtained which exceeded the strength of the control mix which had the commercially available

sodium silicate. Mixes with EF of 80% gave slightly lower strength at 1-day but comparable strengths at 7- and

16
28-days. This result was surprising, since it was expected that increasing the powder content would have led to an

increase in the compressive strength, as the M+ should have increased. However, for room temperature cured

mixes the reaction is driven by the GGBS and the relatively low sensitivity of GGBS to variations in M+ in the

investigated range could explain this result [47]. Further investigations are however needed to confirm this.

Adding the powder activator to the mix in solid form just before mixing did not have any adverse effect on the

compressive strength. This makes possible a one-part alkali activated binder.

Series III: 100% fly ash mixes with AM=1.0. The previous two series investigated blends of fly ash/GGBS whilst

this series aimed at assessing the efficiency factor (EF) with AM=1.0 (SM=1) of a neat fly ash binder which unlike

the previous mixes required curing at 70 °C. Table 7 shows the mortar mix proportions and Figure 9 shows the

compressive strengths at 1, 7 and 28 days.

Table 7. Mix proportions of neat fly ash mortars with AM = 1.0 for Series III (cured at 70 °C).

Control
Parameter 100-10-65-330 100-10-80-330 NaOH only
(Commercial SS)
M+ (%) 7.5 7.5 7.5 7.5
AM (-) 1.0 1.0 1.0 ∞
EF (%) 70% 90% 100% 100%
w/s ratio (-) 0.370 0.370 0.371 0.371
Fly ash (kg/m3) 518.0 523.5 532.2 546.8
GGBS (kg/m3) - - - -
Sand (kg/m3) 1424.5 1439.6 1463.5 1503.7
Act. powder (kg/m3) 125.3 98.5 - -
SS solution (kg/m3) - - 156.5 -
NaOH solution (kg/m3) - - 86.2 177.2
Added water (kg/m3) 220.7 223.0 71.3 94.0

17
 Fig. 9. Compressive strength results of Series III mortar cubes.

The mortar strengths of neat fly ash mixes (cured at 70 °C) using the powder activator assuming EF of 65% and

80% were higher than the control mixes. The efficiency of the powder activator has again been shown to be nearly

equal to the commercially available ones. As previously discussed, the use of a lower EF implies more powder

activator, and therefore potentially increasing the M+ of the mix if the EF assumed is not correct. For example,

the amount of powder activator used for an EF of 65% would lead to an M+ exceeding 9.5% if the actual EF is

80% instead. The 28-day compressive strength would therefore be expected to be higher than the control whose

M+ is only 7.5%. This confirms that an EF of about 80% is reasonable.

Series IV: fly ash/GGBS and fly ash mixes with powder processed at 150 °C. The aim of series IV was to assess

the efficiency of the powder activator produced at 150 °C rather than at 330 °C. Both 60%/40% fly ash/GGBS

blends and neat fly ash mixes were investigated. The efficiency factor (EF) was assumed to be 65%. Perhaps, with

hindsight, an EF equal to 80% should have been more appropriate. This was not possible as series IV was done at

the same time as series III. Table 8 shows the mix proportions and Figure 10 shows the compressive strengths.

18
Results from mixes 60-1.0-65-330 and 100-1.0-65-330 (discussed earlier) as well as those of the control mixes

are also shown for comparison.

The 1- and 7-day strengths of mortars with powder activator produced at 150 °C were similar to the strengths of

mortars with powder activator produced at 330 °C. However, at 28 days, fly ash/GGBS mortars with powder

activator produced at 150 °C had significantly higher strengths than the mortars with powder activator produced

at 330 °C, whilst the opposite was observed for neat fly ash samples. This aspect would require further

investigation.

Table 8. Mix proportions for mortars with activator powder produced at 150 °C.

Parameter 60-1.0-65-150 100-1.0-65-150

M+ (%) 7.5 7.5
AM (-) 1.0 1.0
EF (%) 70% 70%
w/s ratio (-) 0.386 0.370
Fly ash (kg/m3) 312.5 518.0
GGBS (kg/m3) 208.3 -
Sand (kg/m3) 1432.2 1424.5
Act. powder (kg/m3) 126.0 125.3
SS solution (kg/m3) - -
NaOH solution (kg/m3) - -
Added water (kg/m3) 231.2 220.7

19
 Fig. 10. Series IV. Comparison of mortar compressive strengths with power activator produced at 150 °C and
330 °C. (a) 60%/40% fly ash/GGBS mixes; (b) 100% fly ash mixes.

The above four series of mortar mixes have shown that:

- The powder activators produced with the described thermochemical process were efficient in activating

precursors such as fly ash and GGBS which are commonly used in alkali activated binders.

- The actual composition of the powder, the amorphous content and the degree of dissolution in water

make accurate estimates of the availability of silicates very difficult. An efficiency factor (EF) of 80%

has been proposed to take into account these uncertainties.

- An AM = 1.0 (SM = 1) led to satisfactory strengths without the need for adding sodium hydroxide to

increase it to AM = 1.25 (SM = 0.8).

- The powder activator was found to be efficient also when added directly to the mix in solid form together

with the dry precursor. This makes possible a one-part alkali activated binder requiring only water to be

added, i.e. similar to Portland cement.

- The powder activator produced at 150 °C was found to be as efficient as the one produced at 330 °C.

This has considerable savings in terms of the energy requirements and consequently reduces the carbon

20
 footprint of the powder activators.

Compressive strength tests have shown that the efficiency of the powder activator was very good, contrary to

commercially available solid sodium metasilicate (anhydrous or pentahydrate), shown by others [10,12] not to be

good in activating powder precursors. This was confirmed in this research with some tests that used powder

sodium silicate pentahydrate (commercially available) on mortars with neat fly ash and fly ash/GGBS blends. The

powder was either added into the mixer in dry form or pre-dissolved in water, but very low reaction was observed.

The dissolution rate was quite slow and there was no heat developed. Less than one third of the expected

compressive strength was measured even after 56 days of curing in the case of neat fly ash mortar, and less than

half of the expected compressive strength in the case of 60%/40% fly ash/GGBS mix.

Why the solid sodium metasilicate produced from waste glass powder was found to be reactive was therefore

further investigated by a series of infrared spectroscopy tests.

3.3. FT-IR investigation on the produced powder activator

FT-IR spectra of powder activators produced at 150, 250 and 330 °C were investigated, comparing these with the

ones obtained with commercially available sodium silicate pentahydrate and raw glass powder.

The raw glass powder transmission spectrum shows a main peak corresponding to the Si–O–Si stretching band at

999 cm-1, which for amorphous pure silica should occur in the range of 1100 cm-1 [48,49].

The three powder activators showed a recognisable pattern in their spectra, having a broad hump in the region

around 2900 cm-1, which can be attributed to sodium carbonate [50]. No evidence of –OH bonds (which are

usually associated with bands at about 3500 cm-1 and 1650 cm-1 [51]) were observed, which would exclude the

presence of unreacted NaOH, confirming the findings from the XRD analysis. Sodium carbonate presence was

confirmed by the peak at 1423 cm-1 [51, 52]. The main peaks were recorded in the range 1025 to 709 cm-1, see

Figure 11.

The addition of alkali cations in silicates results in a destruction of the silicate polymerised network. One of the

effects of Na inclusion which is observed in IR spectra is the shift towards lower wavenumbers of the main peaks

associated with Si-O-Si bonds stretching and bending [53 – 56]. The inclusion of alkali metal cations forces the

breakdown to Si-O-Si bonds and the formation of Si-O-M (where M is the alkali metal cation) and Si connection

with non-bridging oxygen (NBO) [48,53,57].. In amorphous solid silicate, only Q4 groups are present [30,33],

whilst in silicate solutions a high quantity of Q0 groups is expected [58-59].

21
 Fig. 11. FT-IR spectra of bottle glass and powder activators produced at temperatures of 150, 250 and 330 °C.

An increase of alkali concentration in the silica network has been shown by others [53] to give a higher

concentration of NBO groups, and their presence is known in the literature to be responsible for improving the

dissolution of silica and its availability for the formation of other Si-rich structures [53,58,60].

The presence of peaks that can be attributed to Si-NBO groups in the spectra of processed powder activators could

explain their reactivity, as Si would be readily dissolved and available for geopolymerisation reactions. The

analysis focused therefore on the sequence of peaks in the bands 1025 to 673 cm-1 observed in all processed

powder activators (i.e. irrespective of the process temperature), which might suggest the presence of different

species of silicates, i.e. silicate structures with oxygen bonds having different degrees of connectivity. A

deconvolution of the spectrum of the powder produced at 330 °C was tried in order to better appreciate peaks in

the range 600 – 1200 cm-1 wavenumber. However the deconvolution fit was not suitable for a valid interpretation,

and therefore the following discussion is to be considered qualitative and not quantitative.

Data available from literature [11,49,51-59,61] were used for identifying peaks corresponding to the different

structures. As known, wavenumber decreases with more NBOs present. A strong presence of Q2 groups (at

wavenumber positions 964 to 878 cm-1) was detected, which could suggest that some of the material was

anhydrous sodium metasilicate, which, according to the literature, is composed by a chain of Si-O tetrahedra, each

sharing two oxygens with adjacent Si-O groups, i.e. Q2 [59]. However, there is also a non-negligible presence of

Q0 – Q1 structures, particularly in the band at wavenumber 851 cm-1, which indicates the presence of a certain

number of NBO groups which can be responsible for the good reactivity observed.

22
From qualitative analysis of the IR spectra of the processed powder activators, it can therefore be concluded that

the chemical structure of the obtained powder activators showed a disordered sodium silicate structure with a high

number of reactive NBO groups, presumably concentrated on the surface of the powder grains, which allow the

fast dissolution (and availability for reaction) of Si.

3.4. Cost analysis

Mix proportions for three concretes of 28-day mean compressive strengths of 35, 50 and 70 MPa, described in

[62], were used for estimating the costs per cubic metre when (a) commercially available activators and (b) powder

activators were used. These were compared with equivalent strength Portland cement concretes. Compressive

strength, initial setting time and initial consistency were comparable among alkali activated concretes (AAC) and

Portland cement concretes. Unit costs of the constituents used for estimating the unit cost of a cubic metre of

concrete are shown in Tables 9 to 11.

The powder activator costs were estimated for the thermochemical process described above which involved

heating a mix of NaOH and glass powder with mass proportions 48% / 52% at 150 °C to 330 °C for one to two

hours. One hour duration has been shown to be sufficient. However, the following analysis considers two hours

of heating in order to incorporate possible longer heating time that may be needed for larger quantities likely to

be processed when upscaling the process from laboratory to industrial scale.

The costs associated with the production of the powder activator are:

(a) Cost of raw materials (glass cullet and powder NaOH). Cost of NaOH has been estimated to be 200 £/t

from data available on industrial trading websites. Considering the required mass proportion, cost of

NaOH per tonne of activator powder is about £96. The glass powder has been assumed to be available

for free at source and no transportation costs have been considered. Glass powder is available from bottle

recycling plants and it is currently considered to be a waste. The energy requirements for grinding the

glass in case the material is available in cullet form has been estimated to be 50 kWh/t, as average

(conservative) value obtained from technical specifications (installed power vs capacity) of industrial

grinding equipment. 26 kWh would therefore be required for grinding, which will add an extra 4 £/t when

considering energy cost of £0.15 per kWh.

(b) Cost of energy for the heating process. Large cabinet ovens (700 l capacity) might process 1 tonne of

powder per batch. The power consumption has been estimated to be 6 to 10 kW according to the technical

specifications obtained for industrial ovens with the required capacity. It is therefore estimated that 20

23
 kWh is the energy required for processing 1 tonne of material. A heating cost of 3 £/t is considered

reasonable on the basis of £0.15 per kWh.

Based on the above, the production cost for the powder activator was estimated to be in the range of 100 £/t (out

of which about 95% is due to NaOH cost).

Table 9. Cost analysis for Concrete 1 – target 28-day mean strength of 35 MPa.

AAC with commercial Portland cement AAC with activator

Constituent £/t activators concrete powder
kg £ % kg £ % kg £ %
Portland Cement 70 - - - 355.0 24.85 57% - - -
Fly ash 40 283.2 11.33 18% - - - 283.2 11.33 27%
GGBS 65 70.8 4.60 7% - - - 70.8 4.60 11%
Sodium silicate solution 300 83.3 24.99 39% - - - - - -
Activating powder 100 - - - - - - 66.6 6.66 16%
NaOH 200 20.5 4.10 6% - - - - - -
Water 1 105.9 0.11 0% 220.0 0.22 1% 158.8 0.16 0%
Aggregate 10 1897.5 18.98 30% 1832.0 18.32 42% 1897.5 18.98 45%
Total 2461.2 64.10 100% 2407.0 43.39 100% 2476.9 41.72 100%

Table 10. Cost analysis for Concrete 2 – target 28-day mean strength of 50 MPa.

AAC with commercial Portland cement AAC with activator

Constituent £/t activators concrete powder
kg £ % kg £ % kg £ %
Portland Cement 70 - - - 440.0 30.80 63% - - -
Fly ash 40 258.3 10.33 15% - - - 258.3 10.33 24%
GGBS 65 110.7 7.20 11% - - - 110.7 7.20 17%
Sodium silicate solution 300 86.8 26.04 39% - - - - - -
Activating powder 100 - - - - - - 69.4 6.94 16%
NaOH 200 21.4 4.28 6% - - - - - -
Water 1 102.0 0.10 0% 220.0 0.22 0% 157.0 0.16 0%
Aggregate 10 1897.5 18.98 28% 1750.0 17.50 36% 1897.5 18.98 44%
Total 2476.7 66.92 100% 2410.0 48.52 100% 2492.9 43.60 100%

24
 Table 11. Cost analysis for Concrete 3 – target 28-day mean strength of 70 MPa.

AAC with commercial Portland cement AAC with activator

Constituent £/t activators concrete powder
kg £ % kg £ % kg £ %
Portland Cement 70 0.00 0% 550.0 38.50 69% 0.00 0%
Fly ash 40 99.0 3.96 6% 0.00 0% 99.0 3.96 9%
GGBS 65 231.0 15.02 23% 0.00 0% 231.0 15.02 34%
Sodium silicate solution 300 77.6 23.28 36% 0.00 0% 0.00 0%
Activating powder 100 0.00 0% 0.00 0% 62.1 6.21 14%
NaOH 200 19.1 3.82 6% 0.00 0% 0.00 0%
Water 1 121.2 0.12 0% 220.0 0.22 0% 170.8 0.17 0%
Aggregate 10 1925.0 19.25 29% 1668.0 16.68 30% 1925.0 19.25 43%
Total 2472.9 65.45 100% 2438.0 55.40 100% 2487.9 44.61 100%

As shown in Tables 9 to 11, the alkali activated concrete produced with commercially sourced activators is

considerably more expensive than the equivalent Portland cement concrete. Cost differences vary from 48% for

the normal strength concrete – Concrete 1, decreasing to 18% for high strength concrete – Concrete 3. On the

contrary, alkali activated concrete produced with the powder activator has a slightly lower cost than Portland

cement concrete (4%) for normal strengths, and it is cheaper by 10% and 19% for 50 MPa and 70 MPa 28-day

strength concretes respectively.

The cost of activator is between 41% and 45% of the total cost of alkali activated concretes when commercially

available activators are used but it is only between 14% and 16% when the powder activator is used.

Conclusions

The development of a low cost and with a low environmental carbon footprint powder activator has been

described. This is an alternative to the currently commercially available ones. The cost of currently available

commercials activators is hindering the commercialisation of AAB concretes. Commercially available activators

are not only expensive but also come in liquid form which has health and safety implications in their handling on

site.

Main results from the investigation can be summarised as follows:

- A procedure for the production of a powder activator, composed mainly of reactive sodium silicate has

been developed. This involves blending glass powder and sodium hydroxide in mass proportions of 48%

/ 52%, adding water to produce a paste and then heating the mixture in an oven at 150 °C to 330 °C for

1 to 2 hours.

- The efficiency of the developed powder activator was compared with commercially available activators

25
 by comparing compressive strength of mortars produced with different precursors (a blend of fly

ash/GGBS on the one hand and neat fly ash on the other). The efficiency of the powder activator in

activating fly ash and fly ash/GGBS binders was shown to be around 80%. This is based on the fact that

compressive strengths of mortars produced with the developed powder activator had similar strengths to

those of control mixes produced with commercially available activators.

- Microstructural investigation on the powder activator suggested that it was sodium metasilicate with a

disordered structure showing a significant presence of Q0 and Q1 type silicate. This indicates high

solubility and reactivity.

- The cost of concretes produced with the developed powder activator were comparable, if not cheaper,

than equivalent Portland cement concretes. The developed powder activator is also expected to have a

much lower carbon footprint.

In order to corroborate the results obtained, further work is envisaged in the near future, which will determine

more accurately the solubility of SiO2 under the different conditions employed. The optimisation of the production

step will also investigate suitable methods for the recovery of water used for the treatment, in line with EU strategy

for future industrial processes with no waste.

Microstructural characterisation of pastes produced with the developed powder activator is continuing in order to

determine the reaction products and whether these differ from those obtained when commercially available liquid

activators are used. Porosity and pore size distribution of pastes will also be investigated.

Acknowledgements

The authors would like to thank Innovate UK-EPSRC for providing funding for the project RESCIND “REcovery

and uSe of Cement kIlN Dust as the alkali activator for Geopolymeric (Cementless) Concrete Building Blocks”,

Grant Ref. EP/N508962/1. The authors are also grateful to Prof Kenneth R. Seddon and Dr Natalia Plechkova for

their invaluable advice and guidance in developing the thermochemical process described in this paper. The

assistance of Dr Susan Lawther with the fly ash mixes, Series III, is also gratefully acknowledged.

References

[1] J.L. Provis, Alkali-activated materials. Cement and Concrete Research. Article in press,
http://dx.doi.org/10.1016/j.cemconres.2017.02.009 (2017)

[2] M. Fawer, M. Concannon, W. Rieber, Life cycle inventories for the production of sodium silicates. The
International Journal of Life Cycle Assessment, 4(4) (1999) 207–212.

26
[3] F. Pacheco-Torgal, J.P. Castro-Games, S. Jalali, Alkali activated geopolymeric binder using tungsten mine
waste: preliminary investigation. In Geopolymer, Green Chemistry and Sustainable Development Solutions.
France, (2005) 93–98.

[4] B. Tempest, O. Sanusi, J. Gergely, V. Ogunro, D. Weggel, Compressive strength and embodied energy
optimization of fly ash based geopolymer concrete. In world of coal ash (WOCA) conference, (2009) 1–17.

[5] R. Witherspoon, H. Wang, T. Aravinthan, T. Omar, Energy and emissions analysis of fly ash based
geopolymers. SSEE 2009: Solutions for a Sustainable Planet (2009)

[6] B.C. McLellan, R.P. Williams, J. Lay, A. Van Riessen, G.D. Corder, Costs and carbon emissions for
geopolymer pastes in comparison to ordinary portland cement. Journal of Cleaner Production, 19(9) (2011)
1080–1090.

[7] F.K. Crundwell, On the Mechanism of the Dissolution of Quartz and Silica in Aqueous Solutions. ACS
Omega, 2(3), (2017) 1116-1127.

[8] E. McGuire, J.L. Provis, P. Duxson, R. Crawford, Geopolymer concrete is there an alternative and viable
technology in the concrete sector which reduces carbon emissions. Proceedings concrete, 11 (2011).

[9] G. Habert, C. Ouellet-Plamondon, Recent update on the environmental impact of geopolymers. RILEM
technical Letters, 1, (2016) 17-23.

[10] T. Luukkonen, Z. Abdollahnejad, J. Yliniemi, P. Kinnunen, M. Illikainen, One-part alkali-activated

materials: A review. Cement and Concrete Research, 103 (2017) 21-34.

[11] J.R. Falcone Jr, J.L. Bass, M. Angelella, E.R. Schenk, K.A. Brensinger, The determination of sodium
silicate composition using ATR FT-IR. Industrial & Engineering Chemistry Research, 49(14) (2010) 6287-
6290.

[12] C. Rees, G.C., Lukey, J.S.J. Van Deventer, The microstructural characterization of geopolymers derived
from fly ash and solid silicates. In Geopolymer, Green Chemistry and Sustainable Development Solutions:
Proceedings of the World Congress Geopolymer 2005 (p. 61) (2005). Geopolymer Institute.

[13] V. Živica, Effectiveness of new silica fume alkali activator. Cement and Concrete Composites, 28(1)
(2006) 21-25.

[14] E.D. Rodríguez, S.A. Bernal, J.L. Provis, J. Paya, J.M. Monzo, M.V. Borrachero, Effect of nanosilica-
based activators on the performance of an alkali-activated fly ash binder. Cement and Concrete
Composites, 35(1) (2013) 1-11.

[15] S.A. Bernal, E.D. Rodríguez, R.M. de Gutiérrez, J.L. Provis, S. Delvasto, Activation of metakaolin/slag
blends using alkaline solutions based on chemically modified silica fume and rice husk ash. Waste and Biomass
Valorization, 3(1) (2012) 99-108.

[16] S.A. Bernal, E.D. Rodríguez, R.M. de Gutiérrez, J.L. Provis, Performance at high temperature of alkali-
activated slag pastes produced with silica fume and rice husk ash based activators. Materiales de Construcción,
65(318) (2015).

27
[17] H.K. Tchakouté, C.H. Rüscher, S. Kong, E. Kamseu, C. Leonelli, Geopolymer binders from metakaolin
using sodium waterglass from waste glass and rice husk ash as alternative activators: A comparative study,
Construction and Building Materials, 114 (2016) 276-289.

[18] K.T. Tong, R. Vinai, M.N. Soutsos, Use of Vietnamese rice husk ash for the production of sodium silicate
as the activator for alkali-activated binders. Journal of Cleaner Production, 201 (2018), 272-286.

[19] A. Passuello, E.D. Rodríguez, E. Hirt, M. Longhi, S.A. Bernal, J.L. Provis, A,P, Kirchheim, Evaluation of
the potential improvement in the environmental footprint of geopolymers using waste-derived activators.
Journal of Cleaner Production, 166 (2017) 680-689.

[20] G.A. Blengini, M. Busto, M. Fantoni, D. Fino, Eco-efficient waste glass recycling: Integrated waste
management and green product development through LCA. Waste management, 32(5) (2012) 1000-1008.

[21] J. Flanagan, M. Davies, Glass recycling–life cycle carbon dioxide emissions. A Life Cycle Analysis Report
prepared for British Glass by Enviros Consulting Ltd
http://www.packagingfedn.co.uk/images/reports/Enviros_Report.pdf (2003) (accessed 22 December 2017)

[22] FEVE, Federação dos vidros de embalagem da Europa. Estatísticas. http://feve.org/wp-

content/uploads/2016/10/FEVE-2014-target.jpg (2014) (accessed 22 December 2017)

[23] H. Rashidian-Dezfouli, P.R. Rangaraju, V.S.K. Kothala, Influence of selected parameters on compressive
strength of geopolymer produced from ground glass fiber. Construction and Building Materials, 162 (2018)
393-405.

[24] S. Zhang, A. Keulen, K. Arbi, G. Ye, G, Waste glass as partial mineral precursor in alkali-activated
slag/fly ash system. Cement and Concrete Research, 102 (2017) 29-40.

[25] R. Novotny, A. Hoff, J. Schuertz, J. Process for hydrothermal production of sodium silicate solutions. U.S.
Patent 5,000,933 (1991).

[26] G. Modica, Process of treating end-of-life cathode ray tubes for lead and soluble silicates recovery. U.S.
Patent Application 13/692,136 (2012).

[27] F. Puertas, T.J.D. Torres, C. Varga, Method for the production of alkali cements from industrial and urban
waste glass. WO2012172138A1 (2012).

[28] D.B. Hawkins, Kinetics of glass dissolution and zeolite formation under hydrothermal conditions. Clays
and Clay Minerals, 29(5) (1981) 331-340.

[29] P.M. Dove, N. Han, A.F. Wallace, J.J. De Yoreo, Kinetics of amorphous silica dissolution and the paradox
of the silica polymorphs. Proceedings of the National Academy of Sciences, 105(29) (2008) 9903-9908.

[30] Y. Niibori, M. Kunita, O. Tochiyama, T. Chida,. Dissolution rates of amorphous silica in highly alkaline
solution. Journal of Nuclear Science and Technology, 37(4) (2000) 349-357.

[31] S.S. Kouassi, J. Andji, J.P. Bonnet, S. Rossignol, Dissolution of waste glasses in high alkaline solutions.
Ceramics-Silikáty, 54(3) (2010) 235-240.

28
[32] G. Di Bella, I. Arrigo, P. Catalfamo, F. Corigliano, L. Mavilia, Advances in the extraction of silica from
glass cullet. In Recycl Reuse Waste Mater Proc Int Symp, (2003) 137-142.

[33] F. Puertas, M. Torres-Carrasco, Use of glass waste as an activator in the preparation of alkali-activated
slag. Mechanical strength and paste characterisation. Cement and Concrete Research, 57 (2014) 95-104.

[34] M. Torres-Carrasco, F. Puertas, Waste glass in the geopolymer preparation. Mechanical and
microstructural characterisation. Journal of Cleaner Production, 90 (2015) 397-408.

[35] M. Torres-Carrasco, J.G. Palomo, F. Puertas, Sodium silicate solutions from dissolution of glasswastes.
Statistical analysis. Materiales de Construcción, 64(314) (2014).

[36] F. Puertas, M. Torres-Carrasco, M. Alonso, Reuse of urban and industrial waste glass as novel activator for
alkali-activated slag cement pastes: A case study. In Handbook of alkali-activated cements, mortars and
concretes, Woodhead Publishing, United Kingdom, (2014) p.75.

[37] M. Torres-Carrasco, F. Puertas, Re-use of waste glass as alkaline activator in the preparation of alkali-
activated materials. In 34th Annual Cement and Concrete Science Conference, 14-16 September 2014 -
University of Sheffield, (2014) 207-210

[38] M. Torres-Carrasco, C. Rodríguez-Puertas, M. del Mar Alonso, F. Puertas, Alkali activated slag cements
using waste glass as alternative activators. Rheological behaviour. Boletín de la Sociedad Española de
Cerámica y Vidrio, 54(2) (2015) 45-57.

[39] H. Mori, Extraction of silicon dioxide from waste colored glasses by alkali fusion using sodium hydroxide.
Journal of the Ceramic Society of Japan, 111(1294) (2003) 376-381.

[40] M. Keawthun, S. Krachodnok, A. Chaisena, Conversion of waste glasses into sodium silicate solutions. Int.
J. Chem. Sci., 12(1) (2014) 83-91

[41] B. Hu, S. Zhao, S. Zhang, Removal of lead from cathode ray tube funnel glass by generating the sodium
silicate. Journal of the Air & Waste Management Association, 65(1) (2015) 106-114.

[42] E. Negre, V. Motto-Ros, F. Pelascini, S. Lauper, D. Denis, J. Yu, On the performance of laser-induced
breakdown spectroscopy for quantitative analysis of minor and trace elements in glass. Journal of Analytical
Atomic Spectrometry, 30(2) (2015) 417-425.

[43] M. Tiainen, J. Daavitsainen, R.S. Laitinen, The role of amorphous material in ash on the agglomeration
problems in FB boilers. A powder XRD and SEM-EDS study. Energy & fuels, 16(4), (2002) 871-877.

[44] M.D. Allendorf, K.E. Spear, Thermodynamic analysis of silica refractory corrosion in glass-melting
furnaces. Journal of the electrochemical society, 148(2) (2001) B59-B67.

[45] A.B. Meshalkin, A.B. Kaplun, Phase equilibria in the Na2O–SiO2 system. Russ. J. Inorg. Chem, 48 (2003)
1567-1569.

[46] J. Lee, J. Lee, T. Tanaka, H. Mori, In situ atomic-scale observation of melting point suppression in
nanometer-sized gold particles. Nanotechnology, 20(47) (2009).

29
[47] M. Soutsos, A.P. Boyle, R. Vinai, A. Hadjierakleous, S.J. Barnett, Factors influencing the compressive
strength of fly ash based geopolymers. Construction and Building Materials, 110 (2016) 355-368.

[48] J.R. Ferraro, M.H. Manghnani, Infrared absorption spectra of sodium silicate glasses at high pressures.
Journal of Applied Physics, 43(11) (1972) 4595-4599.

[49] M.E. Simonsen, C. Sønderby, Z. Li, E.G. Søgaard, XPS and FT-IR investigation of silicate polymers.
Journal of materials science, 44(8) (2009) 2079.

[50] F.A. Miller, C.H. Wilkins, Infrared spectra and characteristic frequencies of inorganic ions. Analytical
chemistry, 24(8) (1952) 1253-1294.

[51] V.F. Barbosa, J.K. MacKenzie, C. Thaumaturgo, Synthesis and characterisation of materials based on
inorganic polymers of alumina and silica: sodium polysialate polymers. International Journal of Inorganic
Materials, 2(4) (2000) 309-317.

[52] H. Aguiar, J. Serra, P. González, B. León, Structural study of sol–gel silicate glasses by IR and Raman
spectroscopies. Journal of Non-Crystalline Solids, 355(8) (2009) 475-480.

[53] J. Serra, P. González, S. Liste, C. Serra, S. Chiussi, B. León, M. Pérez-Amor, H.O. Ylänen, M. Hupa,
FTIR and XPS studies of bioactive silica based glasses. Journal of Non-Crystalline Solids, 332(1) (2003) 20-27.

[54] R. Gaggiano, I. De Graeve, J.M.C. Mol, K. Verbeken, L.A.I. Kestens, H. Terryn, An infrared
spectroscopic study of sodium silicate adsorption on porous anodic alumina. Surface and Interface Analysis,
45(7) (2013) 1098-1104.

[55] E.F. Medvedev, A.S. Komarevskaya, IR spectroscopic study of the phase composition for sodium silicate
synthesized in aqueous medium. Glass and Ceramics, 64(1) (2007) 7-11.

[56] D.M. Sanders, W.B. Person, L.L. Hench, Quantitative analysis of glass structure with the use of infrared
reflection spectra. Applied Spectroscopy, 28(3) (1974) 247-255.

[57] D.W. Matson, S.K. Sharma, J.A. Philpotts, The structure of high-silica alkali-silicate glasses. A Raman
spectroscopic investigation. Journal of non-crystalline solids, 58(2-3) (1983) 323-352.

[58] H. Jansson, D. Bernin, K. Ramser, Silicate species of water glass and insights for alkali-activated green
cement. Aip Advances, 5(6) (2015).

[59] I. Halasz, M. Agarwal, R. Li, N. Miller, Vibrational spectra and dissociation of aqueous Na2SiO3
solutions. Catalysis Letters, 117(1-2) (2007) 34-42.

[60] I. Pignatelli, A. Kumar, M. Bauchy, G. Sant, Topological control on silicates’ dissolution kinetics.
Langmuir, 32(18) (2016) 4434-4439.

[61] R.D. Husung, R.H. Doremus, The infrared transmission spectra of four silicate glasses before and after
exposure to water. Journal of Materials Research, 5(10) (1990) 2209-2217.

[62] A. Rafeet, R. Vinai, M. Soutsos, W. Sha, Guidelines for mix proportioning of fly ash/GGBS based alkali
activated concretes. Construction and Building Materials, 147 (2017) 130-142.

30