WO2020037349A1 - Process for preparation of geopolymer foam compositions - Google Patents

Abstract

A process for preparing a geopolymer foam concrete comprising: preparing a geopolymer paste comprising fly ash, blast furnace slag, and an alkaline activator; foaming an aqueous composition comprising surfactant and polysaccharide thickening agent by entraining air, optionally under pressure, into the aqueous composition to provide a foamed aqueous composition; blending the foamed aqueous composition with the geopolymer paste to form a geopolymer foam; and curing the geopolymer foam to form a geopolymer foam concrete.

Description

Process for Preparation of Geopolymer Foam Compositions

Field

[0001 ] The invention relates to a process for preparation of geopolymer foam composition and to a geopolymer foam concrete.

Background of Invention

[0002] Geopolymers are an environmentally friendly alternative to Portland cement. They have desirable properties such as high early strength, excellent fire resistance and high resistance to aggressive chemicals such as acids.

[0003] The chemistry of binder formation of geopolymers is very different from the Portland cement hydration reaction. The setting of a geopolymer composition involves poly-condensation of oligo-(sialate-siloxo) moieties into a poly(sialate-siloxo) cross-linked network, generally initiated by an alkaline solution. In contrast the hardening of a conventional cement composition, such as Portland cement (PC), occurs through simple hydration of calcium silicate into calcium disilicate hydrate and lime Ca(OH)₂. Due to the different hardening mechanism, the known use of Portland cement pastes has little bearing on the use of geopolymer foam concrete.

[0004] Geopolymer foamed concrete may be prepared by a range of methods include generating gas in situ by chemical foaming. Chemical foaming agents include hydrogen peroxide, sodium perborate and aluminium metal which react with the alkaline activating agent such as sodium hydroxide. Air can also be entrained in the geopolymer concrete paste during the mixing process.

[0005] The incorporation of a uniform distribution of fine gas bubbles is important to obtaining dimensional stability during curing and ineffective control or inconsistent foaming has very serious consequence for product strength and reliability.

Summary of Invention

[0006] We have found that the product performance of cured geopolymer foam concrete is improved by a process in which a pre-formed foam having a certain composition is blended with a geopolymer paste. [0007] Accordingly there is provided a process for preparing a geopolymer foam concrete comprising: preparing a geopolymer paste comprising mixing a composition comprising water, fly ash, blast furnace slag, and an alkaline activator;

foaming an aqueous composition comprising surfactant and polysaccharide thickener by entraining air, optionally under pressure, into the aqueous composition; blending the foamed aqueous composition with the geopolymer paste to form a geopolymer foam; and

curing the geopolymer foam to form the geopolymer foam concrete.

[0008] The use of polysaccharide thickener and surfactant in the aqueous composition allows the foam to be formed and blended with the geopolymer paste to provide foam stability and a light weight geopolymer concrete with pore uniformity and size which provides a good balance of insulation and strength. In order to optimise the pore structure in the geopolymer concrete it is preferred that the foamed aqueous composition, prior to blending with the geopolymer paste, has a drainage rate of the aqueous composition from the foamed aqueous composition of no more than 20% of the aqueous composition volume over 20 minutes (preferably no more than 15% of the aqueous composition volume over 20 minutes) from foaming.

[0009] The polysaccharide thickener may be present in an amount which is sufficient to provide a Brookfield viscosity (using spindle No. LV3(63) at 20 rpm and 20°C) in water of at least 200cp (such as 200 cp to 3000 cp), preferably at least 300 cp, more preferably 300 to 3000 cp, still more preferably 400 to 2000 cp. Generally the use of an amount of the polysaccharide thickener which provides a Brookfield viscosity in water of at least 200cp, preferably 300 to 3000 cp, more preferably 400 to 2000 cp (using spindle No. LV3(63) at 20 rpm) results in a foam which can effectively be blended with the paste and provide stabilisation of the foam for a period sufficient for setting and hardening of the paste to form a uniform foamed geopolymer concrete.

[0010] The surfactant may comprise a mixture of anionic surfactant and a non ionic surfactant and in particular such a blend where the HLB of the non-ionic surfactant is at least 12. The presence of the blend provides improved foam stability in both the pre-formed foam and the foam when blended with the geopolymer paste. Brief Description of the Drawings

[0011 ] Examples of the invention are described with reference to the attached drawings.

[0012] In the drawings:

[0013] Figure 1 is a graph showing the influence of polysaccharide gum thickener (XG) on the amount of drainage of liquid from foam prepared by air entrainment, over time

[0014] Figure 2 is a graph showing the influence of polysaccharide gum thickener (XG) on the amount of drainage of liquid from foam prepared by chemical foaming, over time.

[0015] Figure 3 is a three dimensional graph showing the frequency of pore sizes in each of the ranges 0-0.005, 0..5-0.02, 0.02-0.4 and 0.4-1.47 for the foamed concrete compositions formed with foams CS1 , A1 , A2 and A3 in Example 1.

[0016] Figure 4 is a column chart showing the compressive strength of

geopolymer foam concretes prepared with air entrained foams CS1 , A1 , A2 and A3 at 7 days, 14 days and 28 days (columns from left to right respectively, for each foam) referred to in Example 1 .

[0017] Figure 5 is a column chart showing the compressive strength of

comparative geopolymer foam concretes prepared with chemically prepared foams CS2 and B at 7 days, 14 days and 28 days (columns from left to right respectively, for each foam) as discussed in Example 1 .

[0018] Figure 6 is a column chart showing the Ultrasound Pulse Velocity (UPV) test results for each geopolymer foam concrete prepare using foams CS1 , A1 , A2 and A3 formed by mechanical air entrainment and CS2 and B formed by chemical foaming as discussed in Example 1 .

[0019] Figure 7 is a graph of viscosity measurements with different spindle speeds for aqueous solutions containing different concentrations of xanthan gum from 0.18 wt% to 0.65 wt% as discussed in Example 2. [0020] Figure 8 is a column chart showing the compressive strength of

geopolymer foam concrete prepared with no aggregate (control group), with sand aggregate and glass aggregate each shown at 7,14,28 and 56 days in successive columns in each group as discussed in Example 3.

[0021 ] Figure 9 is a column chart showing the strength development of

geopolymer with density of about 1200 Kg/m³ for each of the control, sand aggregate group and glass aggregate group at 7,14, 28 and 56 days as discussed in Example 3.

[0022] Figure 10 is a column chart showing the strength development of geopolymer with density of about 1000 Kg/m³ for each of the control, sand aggregate group and glass aggregate group at 7,14, 28 and 56 days as discussed in Example 3.

[0023] Figure 11 is a column chart showing the strength development of geopolymer with density of about 800 Kg/m³ for each of the control, sand aggregate group and glass aggregate group at 7,14, 28 and 56 days as discussed in Example 3.

[0024] Figure 12 is a column chart showing the strength development of geopolymer with density of about 600 Kg/m³ for each of the control, sand aggregate group and glass aggregate group at 7,14, 28 and 56 days as discussed in Example 3.

[0025] Figure 13 is a graph showing the strength at different densities of the control group (no aggregate), sand aggregate group and glass aggregate group as discussed in Example 3.

Detailed Description

[0026] The terms wt% and %wt are used to refer to the percent by weight of a component based on the weight of the relevant composition.

[0027] Where the terms "comprise", "comprises", "comprised" or "comprising" are used in this specification (including the claims) they are to be interpreted as specifying the presence of the stated features, integers, steps or components, but not precluding the presence of one or more other features, integers, steps or components, or group thereof. [0028] In a preferred embodiment, the geopolymer paste composition used in the process of the present invention includes a binder composition comprising a mineral mixture comprising powder fly ash, blast furnace slag and optionally silica fume.

[0029] Fly ash is by product that is formed from the combustion of coal. For example, electric power plant utility furnaces can burn pulverized coal and produce fly ash. The structure, composition, and other properties of fly ash can depend upon the composition of the coal and the combustion process by which fly ash is formed.

American Society for Testing and Materials (ASTM) C618 standard recognizes differing classes of fly ashes, such as Class C fly ash and Class F fly ash. Class C fly ash can be produced from burning lignite or sub-bituminous coal. Class F fly ash can be produced from burning anthracite or bituminous coal. For one or more

embodiments, the fly ash can be selected from the group consisting of Class F fly ash, Class C fly ash, and combinations thereof. Typically the fly ash contains up to 30 wt% calcium oxide.

[0030] Blast furnace slag (BFS) is a waste product of the glass furnace process. Other blast furnace slag materials are granulated blast furnace slag (GBFS) and ground granulated blast furnace slag (GGBFS), which is granulated blast furnace slag that has been finely pulverized. Ground granulated blast furnace slag varies in terms of grinding fineness and grain size distribution, which depend on origin and treatment method, and grinding fineness influences reactivity here. The Blaine value is used as a parameter for grinding fineness, and typically has an order of magnitude of from 200 to 1000 m² kg ¹, preferably from 300 to 500 m² kg ¹. Finer milling gives higher reactivity.

[0031 ] For the purposes of the present invention, the expression "blast furnace slag" is however intended to comprise materials resulting from all of the levels of treatment, milling, and quality mentioned (i.e. BFS, GBFS and GGBFS). Blast furnace slag generally comprises from 30 to 45% by weight of CaO, about 4 to 17% by weight of MgO, about 30 to 45% by weight of S1O2 and about 5 to 15% by weight of AI2O3, typically about 40% by weight of CaO, about 10% by weight of MgO, about 35% by weight of Si0₂ and about 12% by weight of Al₂0₃. [0032] The preferred binder for use in the process comprises a mixture of fly ash and blast furnace slag. The weight ratio of fly ash to blast furnace slag is generally 1 :5 to 5:1 but significant advantages in the process of preparing the foamed geopolymer are provided in using a weight ratio of fly ash to blast furnace slag of 40:60 to 60:40 and generally it has been found that a weight ratio of about 1 :1 confers good results.

[0033] The binder component of the geopolymer paste may also comprise silica fume. Silica fume is an amorphous silica by-product of the manufacture of ferro-silicon and also silicon metals produced by capturing the finely divided particles from stack gases of electric arc furnaces. Silica fume is a pozzolan and when present forms part of the binder composition of the geopolymer paste. The main constituent is silicon dioxide (Si0₂) and it is usually present in at least about 60% but very good results are achieved in the present invention when the Si0₂ content is at least about 85% by weight of the silica fume.

[0034] Amorphous silica is preferably an X-ray-amorphous silica, i.e. a silica for which the powder diffraction method reveals no crystallinity. Precipitated silica may be obtained on an industrial scale by way of precipitating processes starting from water glass. Precipitated silica from some production processes is also called silica gel.

[0035] Fumed silica may form part of the binder in an amount such as 1 wt% to 10wt% of the binder. It may be produced via reaction of chlorosilanes, for example silicon tetrachloride, in a hydrogen/oxygen flame. Fumed silica is an amorphous Si0₂ powder of particle diameter from 5 to 50 nm with specific surface area of from 50 to 600 m² g ¹. While its use is not excluded fumed silica is generally much more expensive than the other optional binder components and is not required in order to obtain excellent geopolymer concrete properties.

[0036] Microsilica may if desired, be used as an additional binder component in an amount of up to 10 wt%. It is a by-product of silicon production or ferrosilicon production, and likewise consists mostly of amorphous Si0₂ powder. The particles have diameters of the order of magnitude of 0.1 microns. Specific surface area is of the order of magnitude of from 15 to 30 m² g ¹. [0037] In contrast to this, commercially available quartz sand is crystalline and has comparatively large particles and comparatively small specific surface area. It serves as inert filler in the invention.

[0038] The preferred quantity present of the alkaline activator in the invention, based on the geopolymer foam mixture, is from 1 to 55% by weight and in particular from 2 to 50% by weight, based on the solids contents of the alkaline activator.

[0039] The term "alkaline activator" as used herein is intended to include an alkaline bicarbonate activator, an alkaline silicate activator, e.g. sodium silicate and/or potassium silicate and/or an alkaline hydroxide activator, e.g. sodium hydroxide, potassium hydroxide and/or other earth metal hydroxide or alkaline solutions.

The alkaline activators suitable for use in the present invention are those alkaline activators commonly used in the field of geopolymer concrete production. In one preferred embodiment of the present invention, the alkaline activator comprises a mixture of alkali metal hydroxides and of alkali metal silicates. The amount of alkaline activator is typically 3 wt% to 15wt% (such as 3wt% to 10 wt%) of the dry weight of the geopolymer composition.

[0040] In one set of embodiments the alkaline activator is used in an amount of 3 wt% to 15 wt% based on the weight of the total of fly ash and blast furnace slag components

[0041 ] Polysaccharide thickeners are, generally speaking, thickening polymers for aqueous systems bearing sugar units, that is, units derived from a carbohydrate of formula C_n(H₂0)n-i or (CH₂0)_n, which may be optionally modified by substitution and/or by oxidation and/or by dehydration.

[0042] The sugar units that may be included in the composition of the thickening polymers of the invention are preferably derived from the following sugars: glucose, galactose, arabinose, rhamnose, mannose, xylose, fucose, anhydrogalactose, galacturonic acid, glucuronic acid, mannuronic acid, galactose sulfate,

anhydrogalactose sulphate and fructose.

[0043] Polysaccharide thickeners that may especially be mentioned include those of native gums such as: a) tree or shrub exudates, including:

^■ gum arabic (branched polymer of galactose, arabinose, rhamnose and glucuronic acid);

^■ ghatti gum (polymer derived from arabinose, galactose, mannose, xylose and glucuronic acid);

^■ karaya gum (polymer derived from galacturonic acid, galactose, rhamnose and glucuronic acid); and

^■ gum tragacanth (or tragacanth) (polymer of galacturonic acid, galactose, fucose, xylose and arabinose);

b) gums derived from algae, including:

^■ agar (polymer derived from galactose and anhydrogalactose);

^■ alginates (polymers of mannuronic acid and of glucuronic acid); and

^■ carrageenans and furcellerans (polymers of galactose sulfate and of anhydrogalactose sulfate);

c) gums derived from seeds or tubers, including:

^■ guar gum (polymer of mannose and galactose);

^■ locust bean gum (polymer of mannose and galactose);

^■ fenugreek gum (polymer of mannose and galactose);

^■ tamarind gum (polymer of galactose, xylose and glucose);

^■ konjac gum (polymer of glucose and mannose);

d) microbial gums, including:

^■ xanthan gum (polymer of glucose, mannose acetate, mannose/pyruvic acid and glucuronic acid);

^■ gellan gum (polymer of partially acylated glucose, rhamnose and glucuronic acid); and

^■ scleroglucan gum (glucose polymer); e) plant extracts, including:

^■ cellulose (glucose polymer); ^■ starch (glucose polymer) and

^■ insulin.

[0044] These polymers may be physically or chemically modified. A physical treatment that may especially be mentioned is the temperature.

[0045] Chemical treatments that may be mentioned include esterification, etherification, amidation or oxidation reactions. These treatments can lead to polymers that may especially be nonionic, anionic or amphoteric.

[0046] The nonionic guar gums that may be used according to the invention may be modified with CrC₆ (poly)hydroxyalkyl groups.

[0047] Among the C1-C6 (poly)hydroxyalkyl groups that may be mentioned, for example, are hydroxymethyl, hydroxyethyl, hydroxypropyl and hydroxybutyl groups.

[0048] The invention provides a process for preparing a geopolymer foam concrete comprising: preparing a geopolymer paste comprising water, fly ash, blast furnace slag, an alkaline activator and optionally silica fume; foaming an aqueous composition comprising surfactant and thickening agent by entraining air, optionally under pressure, into the aqueous composition; blending the foamed aqueous composition with the geopolymer paste to form a geopolymer foam; and curing the geopolymer foam to form the geopolymer foam concrete.

[0049] The process of the invention allows stable preformed foam to be formed by air entrainment of an aqueous composition comprising thickener and surfactant, preferably containing an anionic and non-ionic surfactant. Stability of the pre-formed foam has a significant bearing on the paste incorporating the foam. The foam components, in accordance with the invention, may readily be combined to provide formation of a stable foam which may be prepared, incorporated into the geopolymer paste composition and the resulting geopolymer foam set without substantial drainage of the foam, to produce a set geopolymer foam concrete of relatively uniform pore size. The pre-formed foam will typically be formulated to provide a drainage rate of aqueous composition from the foamed aqueous composition of no more than 20% of the aqueous composition volume over 20 minutes, preferably no more than 15% over 20 minutes from foaming. The balance of components and amounts used in the composition may be determined based on this drainage test and viscosity test as herein described without undue experimentation.

[0050] The stability of the foam is a result of the use of the thickener, particularly a polysaccharide thickener and the surfactant component which preferably comprises a mixture of ionic and non-ionic surfactant.

[0051 ] We have found that the amount of thickener to be used may be determined by examining the effect of the quantity of thickener in pure water, without other components such as the gaseous phase or surfactants. The preferred amount of amount of polysaccharide thickener, will generally provide a Brookfield viscosity in water of at least 200 cp (such as 200 cp to 3000 cp) preferably 300cp, more preferably 300 to 3000 cp, still more preferably 400 to 2000 cp. The Brookfield viscosity is determined using spindle No. LV3(63) at 20 rpm and at 20°C.

[0052] At lower amounts of polysaccharide thickener the drainage rate of liquid from the foam is increased requiring more rapid preparation of the foam caste composition and curing of the composition. In higher concentration of thickener the foam may become unduly gelatinous reducing the retention of fine bubbles and increasing the force required for mixing of the foam and paste.

[0053] The polysaccharide thickener may, for example comprise at least one selected from the group consisting of agar; alginates, carrageenan, gum Arabic, gum ghatti, gum tragacanth, karaya gum, guar gum, locust bean gum, beta-glucan, chicle gum, dammar gum, glucomannan, mastic gum, psyllium seed husks, spruce gum, tara gum, gellan gum, xanthan gum, pullulan, soybean polysaccharides, pectin, cellulose, carboxymethylcellulose (CMC). The thickener is preferably selected from xanthan gum and galactomannan gums such as fenugreek gum, guar gum, tara gum locust bean gum and cassia gum. More preferably the thickener is guar gum, xanthan gum or mixture thereof and most preferably the thickener is xanthan gum. [0054] The optimal amount of thickener will depend on the specific polysaccharide and its effect on viscosity. The thickener in one embodiment is present in an amount of 0.1 wt % to 3 wt % (more preferably 0.1 wt% to 1 wt%) of the aqueous composition in which the foam is prepared by air entrainment. The preferred thickener for use in these amounts is xanthan gum, guar gum or mixture thereof. In a more specific embodiment the thickener is xanthan gum present in an amount of 0.15 wt % to 0.65 wt % of the aqueous composition.

[0055] The preformed foam composition comprises surfactant preferably including an anionic surfactant and a non-ionic surfactant. We have found that an anionic surfactant is particularly suitable for stabilising the pre-formed foam. However, there is commonly a problem in maintaining stability of the foam on combining the pre-formed foam with the geopolymer paste. While such a process may be carried out using an anionic surfactant without a non-ionic co-surfactant we have found that the resulting stability of the foamed geopolymer concrete derived from mixing of the pre-formed foam and geopolymer paste is significantly enhanced in the presence of a non-ionic surfactant. Without wishing to be bound by theory we believe the performance of the anionic surfactant in stabilising the foam is disrupted to some extent by interaction of the anionic head group of the surfactant with mineral particles present in the geopolymer paste. This interaction may lead to migration of the anionic surfactant from the air-liquid interface to the mineral particles in the paste, resulting in

destabilisation and restructuring of the foam to larger air bubbles with a reduced overall air-liquid interface. This is supported by the finding of larger pore sized in geopolymer foam with an anionic surfactant and without the combination of surfactants.

[0056] The non-ionic surfactant has been found to significantly enhance the stability of the foam on mixing of the pre-formed foam with the paste providing a more consistent result and more uniform pore and pore size distribution.

[0057] The hydrophilic-lipophilic balance (HLB) of the non-ionic surfactant plays a role in the effectiveness of the non-ionic surfactant in stabilising the foam composition and the resulting product. It is preferred that the HLB of the non-ionic surfactant is at least 12, such as from 12 to 16. [0058] Examples of suitable classes of non-ionic surfactant include surfactants of HLB is at least 12, such as from 12 to 16. The surfactants may be selected from alkyl polyglucosides, alcohol ethoxylates including primary and secondary alkyl

ethoxylates, alkylphenol ethoxylates such as nonylphenol ethoxylates and

octylphenol ethoxylates and particularly surfactants of these types in which the HLB is at least 12, such as from 12 to 16. The HLB of a range of surfactants may readily be determined and are reported in texts such as McCutcheon’s Emulsifiers and

Detergents, North American Edition, 2017.

[0059] Among the preferred non-ionic surfactants nonylphenol ethoxylates and octylphenol ethoxylates of HLB at least 12 and preferably 12 to 16 have been found to be useful. Examples of commercially available surfactants of this type include octyl phenol ethoxylates having about 8 to about 12 moles of EO and HLB at least 12 such as TRITON™ X-100 (TRITON is a trademark of the Dow Chemical Company) which has an HLB of about 13.4 and about 9.4 moles of EO.

[0060] The surfactant preferably comprises an anionic surfactant. Suitable anionic surfactants may be selected from the group consisting of C₈ to Ci₈ aliphatic sulfates (particularly C₈ to Ci₈ - alpha-olefin sulfates), C₈ to Ci₈ alkyl ether sulfates, fluorinated alkyl sulfonates such as C₄ to C10 perfluoroalkylsulfonates, C₈to Ci₈alkylsarcosinates. The more preferred anionic surfactants are selected from sodium dodecyl sulfate and sodium lauryl ether sulfate. Another group of useful anionic surfactants are the Gemini surfactants, a family of synthetic surfactants which possess, in sequence, a long hydrocarbon chain an anionic group, a spacer, a second anionic group and a further hydrocarbon tail. Examples of Gemini surfactants are discussed by Menger et al,“Gemini Surfactants: A New Class of Self-Assembling Molecules” J. Am. Chem. Soc. 1993,1 15, 10083-10090.

[0061 ] The surfactant may be present in the liquid composition used in

preparation of the preformed foam in a concentration of from about 0.1 wt% to about 3 wt%, such as about 0.1 wt% to about 2 wt% or about 0.1 wt% to about 1 wt%.

[0062] The composition preferably comprises both an anionic and non-ionic surfactant and the ratio may be balanced to obtain the optimum results depending on the specific choices of respective surfactants. Typically we have found a molar ratio in the range of 1 :5 to 5:1 to be useful with 2:1 to 1 :2 being preferred and in many cases a 1 :1 molar ratio being very effective.

[0063] The process comprises entraining air into an aqueous composition comprising the polysaccharide thickener and surfactant (preferably anionic and non ionic surfactant) to prepare the pre-formed foam. The air may be entrained into the aqueous composition comprising the thickener and surfactant component by intense mixing, by introduction of air under pressure, or a combination of both methods of entraining air.

[0064] In one embodiment the air is entrained in the aqueous composition used to form the foam by using a mixer, such as a mixer under high shear, for example, using a planetary mixer (such as HOBART brand planetary mixer) fitted with a whisk beater.

[0065] In a further embodiment the air is entrained in the aqueous composition by introduction of compressed air. Compressed air may be introduced into the aqueous composition using a compressed air foam generator such as foam generators commercially available under the“DEMA” trademark. In one embodiment the pressure of the introduced air is at least 30 PSI. In a further embodiment both mechanical high shear mixing and compressed air are used to entrain air.

[0066] The density of the aqueous foam resulting from air entrainment is typically in the range of 50 Kg/m³ to 120 Kg/m³

[0067] The bubble size produced in the entraining step is typically in the range of from 0.1 to 3 mm and the volume fraction of air in the foamed aqueous composition may be up to 97% but at the high end of this range the foam is less stable. In general it is preferred that the volume fraction of air is no more than 95 volume % and preferably in the range of 80% to 95% such as 88% to 95% by volume of the aqueous foam composition.

[0068] The foam and geopolymer paste may be combined by blending. The blending process may be conducted so as to not substantially disrupt the foam and provide the required cured geopolymer density. The foam and paste may be combined by addition of the foam to the paste or by addition of the paste to the foam. Generally it is preferred to add the foam to the paste with a blending action to provide a uniform mixture.

[0069] The consistency of the geopolymer paste may be regulated to provide effective mixing of the paste and pre-formed foam. The range of mini-slump height is may for example be from 30 mm to 60 mm such as 35 mm to 55 mm for good consistency of paste before addition of foam.

[0070] The geopolymer paste comprises fly ash, blast furnace slag, activator and water and optionally additional components such as additional binder components such as silica fume, fibre reinforcement, rheology modifiers and aggregates. The paste may be prepared from finely divided mixture of fly ash, blast furnace slag and one or more optional components by addition of an aqueous alkaline activator composition. The alkaline activator preferable comprises sodium silicate and sodium hydroxide. In a preferred embodiment the total of sodium silicate and sodium hydroxide is 3 wt% to 15 wt% based on the total weight of fly ash and blast furnace slag.

[0071 ] The activator, particularly the sodium silicate component of the activator may be used in particulate form and blended with the dry fly ash and blast furnace slag components. In one embodiment an aqueous composition preferably an aqueous solution of sodium hydroxide is mixed with an intimate mixture of dry components comprising the fly ash, blast furnace slag and optionally silica fume binder components, sodium silicate activator and optionally one or more of aggregate and fibre reinforcing components. In another embodiment an aqueous composition comprising sodium hydroxide and sodium silicate is mixed with an intimate mixture of dry components comprising the fly ash, blast furnace slag and optionally silica fume binder components and optionally one or more of aggregate and fibre reinforcing components.

[0072] Generally it is preferred for the pre-formed foam to be added and blended with the geopolymer paste. Continuous blending may take place during the addition process and blending may be continued until a uniform mixture of foam and paste is provided. Combination of the foam and paste and blending to achieve a uniform mixture may take place over a period of up to 20 minutes depending on the setting time of the geopolymer paste. In general the time period for blending the foam and paste is no more than 10 minutes such as from 1 to 10 minutes or from 1 to 5 minutes.

[0073] The amount of air introduced into the geopolymer paste may be sufficient to produce a density in the cured geopolymer of no more than 1600 kg/m³ such as no more than 1200 kg/m³, no more than 1000 kg/m³ or no more than 900 Kg/m³. The density of the final product may be at least 300 Kg/m³ or at least 400 Kg/m³ providing ranges such as, for example, 300 kg/m³ to 1600Kg/m³ or 400 Kg/m³ to 1200 Kg/m³.

[0074] The proportion of foam blended with the paste will depend on the consistency of the paste and composition of the paste including any aggregate or other additives and the foam composition. Typically the foam and geopolymer paste are combined to provide an amount of 5 wt% to 20 wt% of the foam in the geopolymer foam mixture.

[0075] The ratio of aqueous liquid to solids in the geopolymer foam mixture will have a bearing on the workability for the composition and the strength of the cured geopolymer foam. While a wide range of ratios of liquid to solid may be used we have found that a weight ratio of 0.25 to 0.5, preferably from 0.3 to 0.45 is particularly useful.

[0076] The geopolymer composition may comprise further materials such as selected from one or more of additional binder components, fibre reinforcement, rheology modifiers and aggregates.

[0077] Silica fume is a particularly preferred additional component and may be present in the copolymer composition in and amount of, for example up to 15 wt% of the composition. Higher amounts of silica fume are useful but will increase the expense of the composition. The silica fume may be present in amounts of from 2 wt% to 15 wt% such as 2 wt% to 10 wt% or 5 wt% to 10 wt%. of the total of blast furnace slag, fly ash and silica fume.

[0078] Silica fume provides significant advantages in the working of the process of the present invention and in the resulting cured geopolymer foam. We have found that the presence of silica fume in the geopolymer paste combined with the aqueous foam results in maintaining a smaller bubble size and resulting pore size. We believe the presence of the silica fume assists in stabilising the foam formed by air entrainment on admixing with the geopolymer paste.

[0079] The geopolymer foam may, and preferably will, include a fibre reinforcing material. The fibres may be natural or synthetic fibres and may be organic such as natural and synthetic organic polymers or inorganic such as glass and basalt fibres. The preferred fibres are organic polymeric fibres, more preferably selected from the group consisting of polyolefins such as polyethylene and polypropylene, polyesters such as nylon, acrylics and PVA (polyvinyl alcohol).

[0080] The fibres may, for example, be of length such as 5mm to 25mm. The loading of fibres may, for example be 0.1 wt% to 1.5 wt%. The fibres may be monofilament fibers in deniers (diameter of the fiber) such as 7, 15, and 100. The fibres are typically of diameter 20 to 550 microns.

[0081 ] The geopolymer foam composition may, if desired contain a

superplasticiser such as the naphthalene formaldehyde condensates or polyacid polymer such as polyacrylic acid polymers known in the art. We have found however that the presence of super plasticisers may reduce retention of air in the blend of foam and geopolymer paste. The superplasticiser may be present in an amount of 0.1 wt% to 2 wt% however in a preferred embodiment the geopolymer composition is free of superplasticiser.

[0082] The geopolymer composition may comprise an aggregate. While the nature of the aggregate is not limited it is particularly preferred in accordance with the present invention that the aggregate used in the process is selected from particulate silica minerals, particularly comprising at least one of sand and glass and most preferably glass. In ordinary Portland cement concrete the microstructure in the bulk cement is different from that at the aggregate interface resulting in weaknesses for chemical penetration. In contrast the use of silica aggregate, particularly comprising at least one of sand and glass and most preferably glass, very significantly reduces this vulnerability in the presence of these aggregates. Furthermore, we have also found that glass aggregates increase the air content of geopolymer concretes prepared by the process of the invention. Geopolymer paste formed with glass fines provides a geopolymer paste about 100 Kg/m³ less than the corresponding foam paste with sand fines providing better foaming control and higher strength for the glass aggregate cured geopolymer foam. In particular glass aggregates allowed preparation of lighter cured geopolymer foam, preferably 300 Kg/m³ to 1000 Kg/m³, more preferably 400 Kg/m³ to 800 Kg/m³ such as 600Kg/m³to 800 Kg/m³, with a significantly improved strength. The morphology of pores in cured geopolymer foams with glass aggregate was also found to be more consistent and circular with less interconnectivity between pores, particularly at lower densities. The increased population of regular closed pores enhances insulation capacity of lightweight foams with glass aggregate. In one example a thermal conductivity of 0.15 W/mK was achieved in a sample with a density of 600 Kg/m³.

[0083] Accordingly in one aspect the geopolymer foam comprises a particulate glass aggregate. The amount of particulate glass aggregate may be 5 wt% to 65 wt% based on the dry weight of the composition, preferably 10 wt% to 40 wt% based on the dry weight of the composition. The particle size of the particulate glass may, for example, be 100 microns to 2 mm such as 0.1 mm to 2mm.

[0084] The geopolymer foam may be cured at elevated temperature. The geopolymer foam may be heated using radiant or microwave heating and microwave heating may facilitate a high throughput. The curing temperature may be 40°C to 100°C and a temperature range of 60°C to 80°C is particularly preferred.

[0085] The process of the invention may be used in preparing building panels.

The building panel may be formed by casting the geopolymer foam, setting of the geopolymer in the mould. The geopolymer foam may be completely or at least partially cured in the mould. For example the geopolymer foam may be removed from the mould before curing at elevated temperature or remain in the mould during curing at elevated temperature. In one embodiment the geopolymer foam is cast and cured within a frame to form the core of a structural panel.

[0086] The invention will now be described with reference to the following examples. It is to be understood that the examples are provided by way of illustration of the invention and that they are in no way limiting to the scope of the invention. EXAMPLES

[0087] Testing of drainage rate and viscosity effect of thickener

[0088] Drainage Rate of Pre-Formed Foam

[0089] The foamed aqueous composition preferably has a drainage rate of the aqueous composition from the foamed aqueous composition of no more than 20% of the aqueous composition volume over 20 minutes, preferably no more than 15% over 20 minutes, from foaming.

[0090] The drainage rate may be determined by placing a volume of foam, such as 50 ml or 100 ml in a measuring cylinder, such as a 100 ml measuring cylinder. Drainage of liquid from the foam is readily evident and forms a liquid layer at the base of the cylinder allowing the rate of drainage from the foam to be readily determined following foam preparation.

Example 1 - Foaming Process

[0091 ] Materials and methods

[0092] The Granulated blast furnace slag (GBFS) used in this research is obtained from Independent Cement, Australia, and fly ash (FA) with the commercial name of Melbourne Ash is supplied from Cement Australia. The results of X-Ray Fluorescence (XRF) analysis of GBFS and FA are presented in Table 1. Quantitative XRD analysis indicated that 75.5% of Melbourne Ash is amorphous. Sodium hydroxide pellets with a composition of 99.9% wt. NaOH is purchased from Sigma- Aldrich and D-grade sodium silicate solution with a composition of 14.7% wt. Na₂0,

29.4% wt. Si0₂ and 55.9% wt. H₂0 is obtained from PQ Australia.

Table 1 : Oxide Composition of FA and GBFS determined by XRFa. Si0₂

t^(a) : trace amounts detected. [0093] Sodium dodecyl sulphate (SDS) solution with a 0.328% wt. concentration, a critical micelle concentration (CMC) of 8.6x1 O ³ mol/L and a molar mass of 288.372 g/mol is supplied by Chem Supply and used in both mechanical and chemical foaming methods. A 1 :60 surfactant to water weight ratio of solution was used in the mechanical foaming method, and 0.01 % wt. of surfactant was used in the chemical foaming method as the modifier. Hydrogen peroxide solution with a 20%

concentration is supplied by Chem Supply and used as a chemical foaming agent.

[0094] Geopolymer mixes are made by activating a dry mix of 50% wt. fly ash and 50% wt. slag with an alkali solution (solution to solid ratio of 0.38). The alkali solution is a mixture of 33% wt. D-grade sodium silicate solution and 67%wt. NaOH solution (3 molar). The Vicat instrument is used to measure the setting time of the geopolymer paste before foaming according to ASTM C191. The setting time of the geopolymer paste is measured at room temperature (23^°C) from the time that the activating solution is added to the dry mixture. Pre-made foams are manufactured by using a high shear mixer. The SDS solution is foamed with high shear mixing for about 10 minutes until all the solution is converted into foam. In order to quantify the impact of the stabilizer on foam stability, a controlled SDS foam without stabilizer is made and compared with three SDS foams with xanthan gum (XG) polysaccharide thickener concentrations of 0.18%wt., 0.25%wt. and 0.45%wt.The foams are poured in a measuring glass cylinder, and their drainage volume is recorded and compared over time. The same foams are used (immediately after they are made) for foaming geopolymers to compare their influence on the properties of the foamed geopolymers. The foams are firstly combined with geopolymer mixtures manually for 20 seconds and are then mixed for one and half minutes to achieve similar wet densities of about

3

700 kg/m .The apparent density of geopolymer foams are calculated by dividing their weight by their volume in the cubic moulds. The geopolymer control sample foamed with SDS only is named CS1 , and the geopolymer samples foamed with the stabilized foams are named A1 ,A2 and A3 (with increasing concentration of XG from A1 to A3). In order to study the impact of the polysaccharide XG on the chemical foaming method, a geopolymer control sample is first mixed with 0.01 %wt. of SDS solution (to improve the pore size distribution of the chemically foamed samples), and then

3 foamed with hydrogen peroxide solution to reach a wet density of around700 kg/m .

In the sample with stabilizer, 0.45%wt. of XG together with 0.01 %wt. of SDS solution are combined with the geopolymer mixture and foamed with H₂0₂ solution to reach the same density. In order to monitor the impact of XG on the stability of fresh foam, the fresh foamed samples are prepared in glass measuring cylinders and the variations in the volume of the foams over time is measured and compared. The geopolymer control sample in the chemically foamed method with H₂0₂ is named CS₂, and the comparative geopolymer sample foamed with the stabilized foaming method is named B.

Table 2: Geopolymer foam samples.

[0095] For the compressive strength measurement, the wet foams are poured into 50x50x50 mm cubic moulds. The moulds are then sealed and cured in a 60°C oven for 24 hours. After curing, the samples are removed from the oven and kept sealed at room temperature until the day of testing. The Instron 5569A machine is used to measure the compressive strength of the foamed samples. The loading rate of the machine is adjusted with a crosshead displacement rate of 1 mm per minute. The average strength of the three samples is reported. In order to study and compare the quality of the geopolymer matrix, the ultrasonic pulse velocity (UPV) within the cubic samples is measured by a Pundit PL-200 machine (obtained from Proceq). The overall UPV is measured by a 50 mm diameter transducer, and the velocity measurement results are used to compare the compactness of the matrix around the bubbles. The Leica M205FA automated microscopy unit is used to produce

microscopic images of the cross-section of the porous samples. The pore size distribution of the mechanically and chemically foamed samples are then calculated by image analysis using Fiji ImageJ software. Fresh premade foams of controlled SDS foam and the XG modified foam are also studied using Leica M205FA

automated microscopy. A thin layer of the freshly made foams are spread on a microscope slide to capture images of the bubbles and observe the impact of XG on the size of the foam’s lamella (the liquid film border between the bubbles). The thermal conductivity of the geopolymer foams is also measured and compared using the transient method with a needle probe. Small cylindrical samples with a diameter of 50 mm and a height of 1 10 mm are prepared with a hollow core. A heat transfer gel is applied on the probe, and the probe is inserted in the hollow core. The temperature is then recorded for 10 minutes under constant heat dissipation through the probe length with a voltage of 3V. The inverse value of the thermal resistivity is calculated based on the temperature difference to obtain the thermal conductivity.

[0096] Results and Discussion

[0097] The impact of the stabilizer on fresh foams

[0098] XG as a foam stabilizer has been added to the premade foams and the chemical foaming agent. The impact of the stabilizer on the foam’s life is measured and compared with different concentrations of XG in the mechanical foaming technique and a single concentration of XG in the chemical foaming method. Since there is no premade foam in chemical foaming, the impact of the stabilizer on the stability of fresh geopolymer foams is studied and compared.

[0099] Foam stability test in premade foams

[0100] As soon as geopolymer foams are made, the bubbles tend to coalesce.

The faster hardening of the geopolymer paste helps to maintain a larger amount of small sized bubbles in the porous geopolymer matrix and avoid interconnectivity between voids. According to the manual examination, it takes about 80 minutes before the geopolymer foams become hard enough to be handled without collapsing. The result of the Vicat test showed that it takes about 120 minutes for the geopolymer paste to reach its initial setting time and about 135 minutes to reach final setting time.

[0101 ] Another factor that can influence the size of the voids in the matrix is the stability of the premade foams. If the premade foams are strong enough to withstand the mixing and preparation process, the resulting geopolymer foams can hold a larger degree of small sized bubbles to start with. Also, as it takes longer for the stabilized foams to drain out their solution and coalesce i.e., they can easily maintain their size until the hardening of the skeleton governs their permanent size and shape.

[0102] Figure 1 shows the results of the foam stability test on premade foams. Over time, the volume of the solution drainage from the foams is measured and used as an indicator of foam stability. The premade foam without XG stabilizer is used as a control to compare the impact of the stabilizer on the foam’s lifetime and strength. Different concentrations of XG are added to the premade foams to also observe the impact of the stabilizer concentration on foam stability. While all the stabilized foams stay intact within the first 20 minutes after pouring into the cylinder, the control foam without stabilizer drains out 94% of its solution during the first 20 minutes of the test. After 20 minutes, the foam with 0.18% XG starts to drain slowly with an almost constant rate of about 1.1 mL per minute. The foam with 0.25% XG stays intact for 40 minutes and drains only about 20% of its solution within 80 minutes of the test. The foam stability test result shows no drainage in the foam with 0.45% of XG. From the test on premade foam, the impact of XG on foam stability is obvious especially at a 0.45% concentration, which completely stops the drainage during the entire testing period.

[0103] Drainage of the liquid from foams is directly affected by gravity flow which significantly depends on the viscosity of the capillary fluid. By increasing the thickness and viscoelasticity of the liquid films, it is possible to enhance the stability of foams. Polysaccharide thickener such as XG plays an important role in thickening and aggregating the liquid film at the lamella. Therefore, the viscosity of the foams will be increased by holding the liquid column between the bubbles and avoiding gravity deformations. In order to investigate the impact of the stabilizer on the characteristics of premade foam, microscopic images of the control foam and the foam with 0.45% XG were taken. The effect of border densification in stabilized foam was observed. While the size of the lamella is about 30±2 pm in the control foam, the stabilized foam has almost double the thickness (about 65±7 pm). This observation confirms the impact of XG on enhancing the characteristics and improving the lifetime of the foams. [0104] Foam stability test in chemical foaming

[0105] In order to compare the impact of XG on the stability of chemically foamed geopolymers, the freshly made geopolymer foams are poured in measuring cylinders, and the expansion and collapse behaviour of the foams are monitored. A similar foam density is not the objective of this test, and in order to make the behaviour of the foams comparable, a similar amount of foaming agent is added to both the control and stabilized geopolymer foams. In the chemical foaming method, it is common not to capture all the generated gas from the foaming chemical reaction. The workability, viscosity and the setting time of the geopolymer paste and the kinetics of the foaming reaction all influence the effectiveness of the chemical foaming process. Also, chemically foamed geopolymers are not as controllable as the mechanically foamed samples since they tend to collapse after a certain expansion period]. This foam collapse is more common under higher foam reaction rates when the rate of gas generation in the matrix is higher than the rate of hardening and stabilization of the voids. As soon as the bubbles coalesce in the matrix, they escape from the paste and initiate the collapse of the foam. XG will influence the viscosity of the paste and might change the behaviour of the chemically foamed geopolymers.

[0106] Figure 2 shows the volume changes observed in chemically foamed geopolymers with and without XG. In the first 15 minutes, the control sample without XG shows a slightly higher expansion rate compared to the modified sample.

However, this sample starts collapsing after this period and collapses several times until it loses about 30% of its maximum volume at the end of the test. However, the sample with XG shows more steady and reliable expansion behaviour over time. While the rate of expansion and the maximum volume of the geopolymer foam is not a high as the control sample, this sample shows much less collapse behaviour and only about 3% of its maximum volume collapsed by the end of the test. After 70 minutes, no further volume change is observed in the samples. With a similar amount of foaming agent, the sample with XG can successfully expand the geopolymer paste to about 30% vol. more than that of the control sample.

[0107] The characteristics of geopolymer foams

[0108] The obvious changes observed in the behaviour of pre-made foams and the fresh geopolymer foams as a result of adding XG are expected to affect the characteristics and behaviour of geopolymer foams. The size distribution of the pores is the initial characteristic that is expected to change in stabilized geopolymer foams. As a result of using a considerably more stable foam with an improved size

distribution of pores, it is expected that some differences in the compressive strength and thermal properties of the foamed geopolymers will be observed. Another interesting factor to examine is the quality of the binding skeleton in different foams. Even though a similar geopolymer mix design has been used in all samples, some small differences in the foaming process might also impact the quality of the geopolymer binding gel.

[0109] Pore characteristics

[01 10] Microscopic images from the cross section of geopolymer foams are taken to visualize and quantify the differences in pore size distribution of the samples. The control sample CS1 without thickener (XG) has a larger degree of coarse pores compared to the rest of the samples with a different percentage of thickener XG. With an increasing the percentage of thickener XG in the samples from A1 to A3, the size distribution of the pores was found to shift to a finer size range. The destabilization effect in foams, which was more prominent in the control foam with no XG, resulted in a higher likelihood of irregular pore shapes caused by merging and distortion of neighbouring pores . The A1 , A2 and A3 images demonstrated that the addition of polysaccharide thickener such as XG can suppress bubble coalescence and reduce the frequency of irregular shaped pores. This observation shows the impact of polysaccharide thickener XG on foam stability after the geopolymer paste is mechanically mixed with more stable foams. Similar to their behaviour in premade foams, bubbles in the stabilized geopolymer foams have a much lower tendency to drain and coalesce, resulting in the confinement of finer pores within the geopolymer matrix.

[01 11 ] Image analysis was conducted in order to quantify and further highlight the differences in pore size distribution. Figure 3 of the drawings shows the pore size distribution of the foams CS1 , A1 , A2 and A3 formed by mechanical entrainment of air. Pore size distribution analysis showed that while the majority of pores in the stabilized samples A1 , A2 and A3 are in the 0-0.005 mm range, the pore size range in the control sample (CS1 ) was mainly in 0.005-0.2 mm and 0.4-1.47 mm regions. Also, the samples with XG showed a much lower frequency of pores in the 0.4-1.47 mm region. By increasing the concentration of XG from A1 to A3, the pore size distribution becomes narrower. This was evident in the shift in pore size to a smaller diameter. Even a relatively small amount of XG in sample A1 markedly improved the pore size distribution of the control sample. While the initial setting time of the geopolymer paste is measured to be 120 minutes, it was evident that the pore size characteristic of geopolymer foams is mainly defined within just few minutes after they cure. Therefore, if the bubble drainage and collapse can be avoided for a few minutes during the mechanical foaming process and after pouring the samples, it is possible to maintain a considerably large number of fine pores in the matrix. For this purpose, even a small percentage of XG can be very useful.

[01 12] The results of a pore size analysis for the chemically foamed geopolymer samples. By comparing microscopic images of the chemically and mechanically foamed samples, the chemically foamed samples show a higher degree of irregularity in the pore shape. While a small percentage of SDS solution normally helps with the fineness and circularity of pores, it is obvious that there are still noticeable differences in the size and shape of the pores in the mechanical and chemical foaming methods. The amount of modifier (SDS) in these samples might not be high enough to show extensive improvements in the size and shape of the pores. The comparison between the images of the two samples with and without stabilizer in Figure 3 suggests some influences of XG on the pore size of chemically foamed samples. By adding XG, the frequency of large pores in control sample (CS2) has decreased and shifted to an increase in pores of smaller size. The majority of pores are less than 0.03 mm in the stabilized sample while much larger voids are obviously dominant in the control sample. In the stabilized sample, XG helps to increase the cohesiveness of the slurry, thereby forming a thicker geopolymer solution that hampers the growth of gas bubbles. In contrast, the presence of only SDS in the control sample results in the formation of thin pore walls that will easily collapse until the skeleton gains enough strength during the hardening stage of the geopolymer paste. In comparing the pore size distribution between the mechanically and chemically foamed samples, some differences can be observed in the control samples and the samples with a similar percentage of XG. While the density of all the samples is almost similar, the pore size distribution is much finer in the mechanically foamed samples. This suggests that the influence of the SDS modifier in enhancing the pore size of the chemically foamed samples was minor here, and there is some room for improving the characteristics of these foams by changing the type of SDS modifier or increasing its percentage. The chemically foamed samples are expected to generally show lower strength than the mechanically foamed samples due to their larger pore size and their more irregular shape.

[01 13] Strength development

[01 14] While polysaccharide thickener XG showed some remarkable effects on reducing the bubble coalescence, as well as the liquid drainage and collapse time of the foams, its influence on foam stability during the mechanical mixing process and the pouring and handling of the foamed samples is also notable. Due to the weakness of the control foam (without XG), there was some foam collapse during the mechanical mixing of the foam with geopolymer paste. Also, after pouring the samples in the moulds, bubble coarsening was observed over time. In contrast to the control foam, the stabilized foams with XG were notably stronger during the mixing, pouring and handling process. An increase in concentration of XG was accompanied with an increase in the stability of the foams during the mechanical mixing process. This impact is more assessable when a lower amount of foam is required for targeting similar densities in the stabilized samples. By increasing the percentage of XG in foams, we need about 12% wt., 20% wt. and 25% wt. less foam in samples A1 , A2 and A3, respectively, when compared with the amount of foam required in the control sample for achieving similar geopolymer foam densities. Other than the impact of XG on the final pore size distribution, the lower amount of foam needed in the stabilized samples results in a lower total water content in the geopolymer foams. While the increase in the amount of foam is known to decrease the water absorption and sorptivity in foam concretes, higher foam (and water) content will have a negative influence on the mechanical properties of the binding skeleton and the final strength of the porous matrix.

[01 15] Figure 4 shows the compressive strength distribution of the mechanically foamed samples over time. As expected, the stabilized samples with XG show better strength results compared to the control sample at an early age and at 28 days. In particular, sample A3 shows about 40% higher strength at 7 days and about 34% higher strength at 28 days. This significant improvement is attributed to the combined enhancements in pore size distribution and the lower foam content in this sample. While the impact of the small percentage of XG in sample A1 was shown to be significant on enhancing the pore size distribution of porous geopolymers, its impact on the strength development is not as substantial at the highest concentrations (A3). While the small percentage of XG is shown to be adequate for maintaining fine bubbles in the freshly made geopolymer foams (within the critical few minutes after they are manufactured), it is obviously not enough for giving the foams enough stability to fully withstand the mechanical mixing and preparation process. Therefore, compared with A3, a larger amount of foam is still needed for these samples to achieve a similar geopolymer foam density. This is why the improvement in strength development is also not as prominent as in the A3 sample.

[01 16] The impact of XG on the strength development of the chemically foamed samples is also evident. Figure 5 shows the strength development in the chemically foamed samples over time. The overall strength of the samples is observed to be lower than the strength in the mechanical foaming method within the same density range. Also, about 20% strength improvement is observed due to the addition of 0.45% XG to the chemically foamed sample, which is less than 15% of the

improvement observed in the mechanical foaming of a similar sample. Similar to the mechanical foaming method, a higher percentage of foaming agent is required for the control sample (CS2) to target the same density as sample B (about 10%). The addition of more H₂0₂ solution to the mix will similarly impact the total water content of the CS2 sample. Therefore, the narrow size distribution of the pores and the slightly lower water content of the geopolymer binding skeleton both play a role in making a stronger geopolymer foam in sample B.

[01 17] The characteristics of the binding matrix

[01 18] Ultrasound Pulse Velocity (UPV) is a non-destructive technique usually used to study the quality of concrete. The existence of cracks and voids or the formation of honeycomb within the concrete matrix drops the pulse velocity, thereby giving a good indication of the quality of the binding matrix. Recently, this method has been also used for studying the homogeneity of the pore distribution in foam

concretes. Figure 6 shows the result of the UPV test on geopolymer foam samples. It is evident that the increase in XG concentration increases the ultrasonic pulse velocity.

[01 19] There are two factors that may impact the UPV results in these samples, namely the enhanced pore size distribution when the percentage of XG is increased, and the enhanced quality of the binding skeleton due to lower water content required in the stabilized samples. Despite the significant differences between the pore size distributions of the chemically foamed and mechanically foamed samples, there is a small difference in their UPV results. This suggests that the impact of the quality of the binding skeleton on the UPV results should be dominant and the enhanced binding characteristics is the main factor responsible for the change observed between the samples with a different percentage of XG. The similarities between the compressive strength results and the UPV results of samples A1 and B further confirms the close relationship between the quality of the binding skeleton and the UPV results.

[0120] Conclusions

[0121 ] Aggregating the Plateau border and delaying the foam drainage can be the key to control both mechanical and chemical foaming mechanisms during the manufacturing process of foam concrete. Many researchers have utilized different stabilizers for enhancing the properties of foamed concretes. However, it is still not clear that to what extent the stabilizer can improve the properties of the foam concrete. Also, there is no research quantifying and comparing this influence on the same system for both chemical and mechanical techniques. As a thickening agent, polysaccharide thickener (XG) can remarkably influence the viscosity of the foam solution and condense the liquid film around the bubbles.

Example 2 - Test for Effective amounts of thickener

[0122] The amounts of various polysaccharide thickener for providing optimum foam stability (minimum drainage may be determined from the viscosity effect of xanthan gum observed in Example 1. In general the amount of polysaccharide thickener is sufficient to provide a Brookfield viscosity in water of at least 200cp such as at least 300cp, preferably 300 to 3000 cp, more preferably 400 to 2000 cp using spindle No. LV3(63) at 20 rpm at 20°C. The spindle is readily available with standard LV torque Brookfield“Ametek” Viscometers.

[0123] The viscosity of xanthan gum solutions in distilled water was examined at different concentration and the preferred ranges observed in practice are shown in Table 3 below and plotted in Figure 7.

Table 3

[0124] The amounts of polysaccharide thickeners other than XG which may be used can be determined by measuring the amount required to obtain a Brookfield viscosity in water of at least 300cp, preferably 300 to 3000 cp, more preferably 400 to 2000 cp using spindle No. LV3(63) at 20 rpm.

Example 3 - Inclusion of sand and glass aggregate [0125] Materials and Method

[0126] Fly ash (FA) with the commercial name of Melbourne Ash was purchased from Cement Australia. Granulated blast furnace slag (GBFS) used in this study is supplied from Independent Cement, Australia. Anhydrous sodium metasilicate with a composition of 50.5% wt. Na₂0, 46.2% wt. Si0₂ and 3.3% wt. Fl₂0 is supplied from Redox. A solid activator was used for better commercial viability of geopolymers as one-part mixtures. Fine sand is usually used in foam concrete applications. The fine sand used in this research has D50 of 251 pm. Glass fines (with 70% of particles between 0.4-2mm) were obtained from Alex Fraser. In order to remove the organic pollutants, the glass fines were washed as received and dried at 60°C for 24 hours. A Rocklabs ring mill was then used to mill the glass and reduce its particle size. The particle size distribution of the milled glass was measured to be 18 pm using the Malvern Mastersizer 2000 laser-diffraction particle-size analyser. The impact velocity tolerance of glass is known to be much less than that of sand. This characteristic of glass allows it to be grinded to fine particles much easier, which is also very attractive for lightweight concrete applications. The results of X-Ray Fluorescence (XRF) analysis of the source materials are presented in Table 4.

[0127] According to the quantitative XRD analysis, 75.5% of Melbourne Ash is amorphous.

a t tr&se .sssi s etected.

Table 4. Oxide Composition of FA and GBFS determined by XRFa

[0128] Three geopolymer systems were studied, namely a control group, glass group and sand group. The geopolymer system with pure geopolymer paste and no aggregates is called the control group. The control group was synthesized to facilitate the comparison between two mortars. Two mortars (glass group and sand group) were made by mixing 30% wt. of glass or sand as aggregates with 70% wt. of geopolymer dry mix. The mix design of the geopolymer binder consists of 60 % wt. FA, 40% wt. GBFS and 8.5% wt. sodium metasilicate in a dry mix. A consistent water to binder ratio of 0.38 was used for all three systems. Geopolymer dry ingredients were firstly mixed manually for two minutes, and water was gradually added and blended with the dry mix manually for one minute first and then mixed by a Flobart mixer for another five minutes. The resulting paste was used for studying the characteristics of the binding skeleton in geopolymer foams. For mechanical testing, the mixtures were poured into 50x50x50 mm cubic moulds, and sealed and cured at ambient temperature until the day of testing. For testing the dry shrinkage of geopolymers, specimens with dimensions of 40 c40 x160 mm were prepared according to the AS1012.13:2015 standard. After seven days, the specimens were submerged in lime-saturated water, and the samples were dried at 23.3 °CC in a chamber with 60% humidity. The shrinkage of the samples at 7, 14, 21 , 28 and 56 days was determined by measuring the change in length as a percentage of the initial length.

[0129] Fresh geopolymer pastes were also foamed with the mechanical foaming technique in order to study the performance of the porous samples in different densities. Premade foam (with ~100 kg/m³ density) was used to introduce voids in the binder pastes. A commercial surfactant was diluted with water (1 :60 surfactant to water weight ratio) and used as a foaming agent. Foam was then made with the aid of compressed air in a Dema compressed air foam generator. Pre-made foam was added to each group of geopolymers in order to make samples with target dry densities of 1200, 1000, 800 and 600 kg/m³.

[0130] 20% of the required premade foam was initially added to the geopolymer mix to increase the workability of the paste, and the remaining 80% of foam was gently blended in afterwards. The wet foam mixtures were then poured into

50x50x50mm cubic moulds, and sealed and cured at ambient temperature until the day of testing.

[0131 ] The Instron 5569A instrument (with a displacement rate of 1 nmm per minute) was used for measuring the compressive strength of the foamed samples.

For testing the mechanical strength of non-porous geopolymers, an ELE ADRAuto 1500 compression testing machine (with a loading rate of 0.5 kN/sec) was used. The reported compressive strength was the average of the three samples. Microscopic images were taken from the cross-section of the porous samples by a Leica M205FA automated microscope to compare the pore characteristics in three groups of geopolymers. The thermal conductivity of the porous samples was measured by a TCi device developed by C-Therm Technologies Ltd. This device measures the thermal conductivity by using the Modified Transient Plane Source (MTPS) method. Details of the experimental setup and data processing methods have been explained previously by Cha et al.

[0132] Results and discussion

[0133] The properties of the binding skeleton

[0134] One of the critical properties of concretes is their drying shrinkage. When concrete dries over time, the free water in its pores moves to the surface and evaporates. The drying shrinkage of concrete depends on the moisture content, pore network and connectivity of pores. Drying shrinkage has a significant impact on the performance of concrete, whereby the early age shrinkage is a critical parameter affecting its durability. As the water exits from pores over time, hydrostatic tension develops in small capillary pores, which leads to microcracks developing in the area. This is detrimental to the service life of concrete as cracks will eventually allow the penetration of aggressive chemicals.

[0135] Figure 8 shows the drying shrinkage behaviour of the geopolymer pastes over time. It can be observed that the control group without aggregates has the highest degree of shrinkage, which is expected due to the high amount of gel formation per gram of this system. The aggregate portion of the system will not shrink, and therefore affect the final shrinkage percentage of the system. The sand group shows a considerably higher percentage of shrinkage than the glass group.

Compared to the pure paste, mortars are known to have lower shrinkage because of the presence of aggregates. Also, it is known that the increase in the size of aggregates increases the drying shrinkage of the mortars. Therefore, the behaviour observed in Figure 8 is reasonable i.e. decrease of shrinkage from the control group to the sand group, and a further drop in shrinkage from the sand group to the glass group with finer glass particles.

[0136] It is interesting to note that the glass group shows remarkably lower shrinkage at 7 days and 14 days, but the shrinkage suddenly quadruples in 21 days. This is because the glass fine is pozzolanic, and it will eventually take part in the geopolymerization reaction. The amount of silica dissolution from waste glass at ambient temperature is known to be negligible. However, as an aggregate in geopolymer concrete, the surface of the glass aggregates can react with alkaline solution over time and develop good binding with the geopolymer paste in bulk. It is known that increasing the proportion of reactive silica to the mortar increases the ultimate shrinkage. By increasing the extent of glass reaction in the matrix, the reacted glass proportion changes its role from aggregate to paste, and it shows a considerably higher degree of shrinkage in the long term. However, the ultimate shrinkage of the glass group is the lowest among the three systems.

[0137] If the strength of the aggregates is not very different from the strength of the binding matrix, aggregates can increase the strength of the paste. Mortars with glass and sand aggregates are stronger than the newly made geopolymer gel at seven days. However, as the paste ages over time, the geopolymer gel gains higher strength which is comparable to the sand group (slightly higher in 56 days of reaction). The strength of the glass group is lower than the control group and the sand group at 14 and 28 days of reaction, but this group gains noticeable strength in the long-term. After 56 days of reaction, the strength of the glass group is comparable to the strength of the control group. This observation is also suggesting that the surface of glass aggregates will take part in the reaction as the paste ages over time resulting in denser and stronger gel in the glass group.

[0138] The properties of the porous matrix

[0139] After adding foam to the three groups of samples, the difference in strength development between the control group and the aggregate groups (glass and sand) increases. This difference is higher at densities of 1200 (Figure 9) and 1000 kg/m³ (Figure 10). The sand group performs similar to the glass group at a density of 1200

3

kg/m , and as the samples become lighter, the differences become smaller. The

3 control sample is the strongest sample at densities of 1200 and 1000 kg/m , but the

3

glass group is comparable to the control group at 800 and 600 kg/m densities but both become the weakest samples as the density drops further.

[0140] The strength at 56 days of the solid (non-porous) samples is higher than at 28 days in all groups. However, the control and sand groups with densities of 1200

3

and 1000 kg/m , respectively, lose their strength between 28 and 56 days. This is related to the higher extent of shrinkage observed in these systems. The shrinkage in the binding skeleton over time will cause micro-cracks in the sample that can affect their final strength. These tiny cracks are more detrimental in samples with relatively high densities since they can increase the degree of pore interconnectivity, which is

3

seldom observed in samples with densities of 1200 kg/m . However, as the density drops and the number of voids increases, the extent of interconnectivity between the pores also increases, and the impact of micro-cracks on the strength will become less critical. As also observed for samples with 600 and 800 kg/m³ densities, the strength loss at 56 days is not observed in the control and sand groups.

[0141 ] The compressive strength of geopolymer foams for three group of geopolymers: control group, sand group and glass group was determined with densities of 800 kg/m³ (Figure 11) and 600 kg/m³ (Figure 12) densities. [0142] As density drops further, the strength of the samples becomes comparable. While the strength of the non-porous samples is very similar after 56 days,

introduction of voids in the systems brings some differences to the strength

development. The strength of the glass group is very similar to the control group with no aggregates at 800 and 600 kg/m³ densities while the strength of the sand group is noticeably lower at these densities. By comparing the density of the non-porous samples, it is evident that the sand group is about 100 kg/m³ heavier than the control and glass groups. This difference in weight has a negative impact on the strength for lightweight samples with low densities (800 and 600 kg/m³). Since the density of the samples is controlled by air bubbles such that it is similar in all three groups, more air bubbles are required to drop the density of the sand group to a similar range as that of the glass and control groups. The presence of more air in this system will have a negative impact on mechanical properties especially when lighter weights are targeted. Other than the strength of the binding matrix, the possible differences in size, shape and connectivity of the pores have a critical impact on the strength development in porous geopolymers. The structure of the pores will be compared and studied to have a clearer idea on the differences between the three systems.

[0143] The microstructure of the pores in different groups of samples with different densities was examined. For a 1200 kg/m³ density, the glass group and sand group appear more porous than the control group. This is attributed to the number of micro pores in the control group being dominant at a density of 1200 kg/m³, which cannot be observed in the 2 mm scale. The major difference between the strength

development in these systems at a 1200 kg/m³ density is also related to the size of the pores in the samples. Pore size is known to have critical impact on the

compressive strength of concrete. Due to the absence of aggregates in the control group, the paste is more homogenous in this sample. This helps to minimize foam coalescence and collapse, thereby leading to a finer pore size in the sample. It can also be observed that the circularity of the pores is the key difference among the samples with a density of 1000 kg/m³.

[0144] The pores in the sand and glass groups are not as circular as the pores in the control group. Other than the size of the voids, which affects the compressive strength of the samples, their shape and circularity is also known to be vital in governing the strength of porous concrete. While the control group still contains a large number of fine pores, the circularity of the larger pores also helps to increase the strength of the system, which is higher than that of the glass and sand groups.

[0145] As the density of the samples drops further to 800 kg/m³, less differences in the structure of the pores can be observed. The circularity and size of the pores appear more similar at this density. The only noticeable difference among the three groups (800 kg/m³ density) is the connectivity of voids. The black spots on the images represent the surface pores, which are connected to internal pores. This

interconnectivity makes the pores deeper, and the depth, which is not visible at the surface, appears black in the microscopic images. The larger area covered by these black spots is an indication of a higher degree of interconnectivity between pores. In the glass group, there are not many black areas observed on the surface, which is an indication of a lesser degree of interconnectivity between pores in this group. The control group has a relatively higher amount of interconnectivity to the internal pores compared to the glass group. The sand group has a larger proportion of black area in the image, which reflects a higher degree of pore interconnectivity. Pore connectivity is also obvious on the surface pores in this sample.

[0146] As the mechanical performance of the samples showed more similarity at lower densities, the microscopic images of the pores also appear to be more similar in lower densities. At a density of 800 kg/m³, the glass group showed a slightly higher strength compared to the control and sand groups. The difference in pore connectivity observed among the three systems is the main reason affecting mechanical performance at this density. At a density of 600 kg/m³, the interconnectivity of pores is increased in both the control and sand groups. However, in the glass group, the size and extent of the connections are still lower than that of the other two systems. In the sand group, a higher amount of air bubbles resulted in a higher amount of

interconnections between pores. The interconnectivity of pores has left large cavities in this system with a density of 600 kg/m³. These cavities are responsible for the difference between the strength of the sand group and the other two systems.

[0147] The results of the thermal conductivity tests on porous samples was examined. As expected, the thermal conductivity of the non-porous paste was found to increase with the addition of aggregates. However, as the density drops, the sand group appears to be the best thermal insulating foam at densities of 1200, 1000 and 800 kg/m³. However, for a density of 600 kg/m³, all three systems perform very similarly with the glass group showing slightly better insulating capacity. A

combination of several factors impacts the thermal performance in these three systems. The thermal conductivity of the binding material and the thermal conductivity of the porous structure also depend on the amount of moisture retention in the pores, the size and shape of pores, the compactness of the binding matrix and the extent of air voids in the matrix. A higher degree of geopolymer gel in the system means a higher extent of water retention in the pores, especially for higher densities where less interconnection between the pores is developed. Water plays a major role in the geopolymerization process . It initially works as a medium for dissolution and participates in hydration reactions. As the geopolymer gel forms and develops further in its structure, water is released back to the system. Higher thermal conductivity of the control sample for 1200 and 1000 kg/m³ densities is related to the fact that a higher degree of gel exists in this system, which releases water over time as they reorganize and polymerize. This free water will be retained in the pores and will impact the thermal conductivity.

[0148] The lower thermal conductivity in the sand group for 1200 and 1000 kg/m³ densities is also related to the fact that less reacting gel exists in this sample compared to other systems. On the other hand, the participation of glass in the reaction at later ages of the sample and the denser geopolymer matrix in the glass group, which gave it higher strength in 56 days, will make the glass binding matrix more conductive compared to the binding matrix of the sand group. However, the remarkable drop of thermal conductivity in the sand group for similar densities is mainly attributed to the fact that the sand group was about 100 kg/m³ heavier than the other two groups. Therefore, a higher amount of air voids is required in this group to drop the density to a similar range to that of the control and glass groups. With a higher amount of air content, the sand group showed lower strength but better thermal insulation capacity. However, the size and shape of the voids are also known to be critical in determining the insulation capacity in porous structures with smaller and more circular void being more desirable. The large degree of interconnectivity of the voids and the irregularity of the pores deteriorates the insulation capacity in the sand group with lower densities, and the thermal performance of this system becomes more similar to the other two groups. High circularity of pores in the glass group and the lower degree of pore connectivity are reasons for the desirable thermal performance of the glass group with a density of 600 kg/m³.

[0149] Conclusion

[0150] The lighter mass of geopolymer paste with glass aggregates and the pozzolanic behaviour of glass in the long-term provide glass with several advantages for geopolymer foam applications. Glass can be ground easily to very fine particles that can replace fine sand in lightweight geopolymer foams. Over time, the surface of the glass particles reacts with the paste and forms stronger bonds at the interface with the geopolymer binder. These unique characteristics make glass fines a suitable alternative to fine sand in geopolymer foam applications. The degree of shrinkage in the paste decreases by replacing sand with glass, and the strength of the porous matrix shows noticeable improvements at low densities. The higher extent of foaming needed with sand aggregates leads to lower strength development and a higher amount of pore connectivity in the sand group. Therefore, while the thermal conductivity is reduced in the sand group, the larger size and interconnectivity of the voids deteriorates the thermal performance of this group at lower densities. With the

3

lowest thermal conductivity at a density of 600 kg/m (0.15 W/mK), the glass group becomes the most favourable insulating material at this density.

Example 4 - Geopolymer Foam Concrete comprising thickener, surfactant mixture, ground glass aggregate and polymer fibre

[0151 ] The process of Example 3 is repeated using the pre-made foam of about 100 Kg/m³ in which the aqueous solution used in preparation of the pre-formed foam contained XG in an amount of 0.45%wt, a combination of SDS anionic surfactant and TRITON X-100 non-ionic octylphenolethoxylate having an HLB of 13.4 and about 9.5 moles EO. Equal molar amounts of the surfactants were used in the weight ratio of 1 :60 surfactant to aqueous solution.

[0152] The geopolymer paste is prepared as in Example 3. The mortar is made by mixing 30% wt. of glass as aggregates with 70% wt. of geopolymer dry mix. The mix design of the geopolymer binder consists of 60 % wt. FA, 40% wt. GBFS and 8.5% wt. sodium metasilicate in a dry mix as in Example 3 but with the addition of 5 wt% of silica fume based on these three dry mix components. PVA fibre is also with added to the dry mix. The loading of fibres is about 1 wt% of the dry mix. The fibres are PVA RECS 100 (length about 13 mm) and are thoroughly mixed with the dry components of the geopolymer prior to activation.

[0153] A consistent water to binder ratio of 0.38 is used. Geopolymer dry ingredients are firstly mixed manually for two minutes, and water is gradually added and blended with the dry mix manually for one minute first and then mixed by a Hobart mixer for another five minutes.

[0154] The premade foam is added to the geopolymer to provide samples of dry density 1200, 1000, 800 and 600 Kg/m³ and the foamed paste is cured at about 60°C to provide a foamed geopolymer concrete of good strength at 56 days.

Claims

Claims

1. A process for preparing a geopolymer foam concrete comprising:

preparing a geopolymer paste comprising fly ash, blast furnace slag, and an alkaline activator;

foaming an aqueous composition comprising surfactant and polysaccharide thickening agent by entraining air, optionally under pressure, into the aqueous composition to provide a foamed aqueous composition;

blending the foamed aqueous composition with the geopolymer paste to form a geopolymer foam; and

curing the geopolymer foam to form a geopolymer foam concrete.

2. The process of claim 1 , wherein the foamed aqueous composition has a drainage rate of the aqueous composition from the foamed aqueous composition of no more than 20% of the aqueous composition volume over 20 minutes from foaming.

3. The process of claim 1 , wherein the foamed aqueous composition has a drainage rate of the aqueous composition from the foamed aqueous composition of no more than 15% of the aqueous composition volume over 20 minutes from foaming.

4. The process of any one of claims 1 to 3, wherein the polysaccharide thickener is present in an amount which is sufficient to provide a Brookfield viscosity in water of at least 300cp using spindle No. LV3(63) at 20 rpm.

5. The process of any one of claims 1 to 4, wherein the polysaccharide thickener is present in an amount which is sufficient to provide a Brookfield viscosity in water of 400 cp to 2000 cp using spindle No. LV3(63) at 20 rpm.

6. The process of any one of claims 1 to 5, wherein the polysaccharide thickener comprises at least one selected from the group consisting of agar; alginates, carrageenan, gum Arabic, gum ghatti, gum tragacanth, karaya gum, guar gum, locust bean gum, beta-glucan, chicle gum, dammar gum, glucomannan, mastic gum, psyllium seed husks, spruce gum, tara gum, gellan gum, xanthan gum, pullulan, soybean polysaccharides, pectin, carboxymethylcellulose (CMC); preferably selected from galactomannan gums such as fenugreek gum, guar gum, tara gum locust bean gum and cassia gum, more preferably guar gum or xanthan gum and most preferably xanthan gum.

7. The process of any one of claims 1 to 6, wherein the polysaccharide thickener is xanthan gum, guar gum or a mixture thereof.

8. The process of any one of claims 1 to 7, wherein the polysaccharide thickener is present in an amount of 0.1 wt % to 3 wt % (preferably 0.1 wt% to 1 wt%) of the aqueous composition.

9. The process of any one of claims 1 to 8, wherein the polysaccharide thickener is present in an amount of 0.1 wt% to 1 wt% of the aqueous composition.

10. The process of any one claims 1 to 9, wherein the polysaccharide thickener is xanthan gum present in an amount of 0.15 wt % to 0.65 wt % of the aqueous composition.

1 1 . The process of any one of claims 1 to 10, wherein the surfactant comprises anionic surfactant and non-ionic surfactant.

12. The process of claim 1 1 , wherein the HLB of the non-ionic surfactant is at least 12.

13. The process of claim 1 1 or claim 12, whereas the non-ionic surfactant is selected from alkyl polyglucosides, alcohol ethoxylates and alkylphenol ethoxylates such.

14. The process of any one of claims 1 to 13, wherein the surfactant comprises an anionic surfactant selected from the group consisting of C₈ to C-i₈ sulfates, C₈ to C-i₈ alkyl ether sulfates and anionic gemini surfactants.

15. The process of any one of claims 1 to 14, wherein the surfactant comprises an anionic surfactant selected from sodium dodecyl sulfate and sodium lauryl ether sulfate.

16. The process of any one of claims 1 to 15, wherein the surfactant comprises an anionic gemini surfactant.

17. The process of any one of claims 1 to 16, wherein surfactant is present in the liquid composition in a concentration of from 0.1 wt% to 1 wt%.

18. The process of any one of claims 1 to 17, wherein the foamed aqueous composition is formed by entraining air under pressure in the aqueous composition.

19. The process of any one of claim 1 to 18, wherein the foamed aqueous composition has a density of 50 Kg/m³ to 120Kg/m³.

20. The process of any one of claims 1 to 19, wherein the geopolymer paste comprises alkaline activator in an amount of 3 wt% to 15 wt% based on the total weight of fly ash and blast furnace slag.

21. The process of any one of the claims 1 to 20, wherein silica fume is present in the geopolymer paste in an amount of from 5 wt% to 10 wt% of the total of blast furnace slag, fly ash and silica fume.

22. The process of any one of claims 1 to 21 , wherein the weight ratio of liquid to solid in the geopolymer foam is from 0.25 to 0.5.

23. The process of any one of claims 1 to 22, wherein the weight ratio of liquid to solid in the geopolymer foam is preferably from 0.3 to 0.45.

24. The process of any one claims 1 to 23, wherein the geopolymer paste is formed from a dry mixture of finely divided blast furnace slag, fly ash, silica fume and sodium silicate.

25. The process of any one of claims 1 to 24, wherein the weight ratio of fly ash to blast furnace slag in the paste is in the range 60:40 to 40:60.

26. The process of any one of claims 1 to 25, wherein the foamed liquid is combined with the geopolymer paste in a ratio of 5 wt% to 20wt% based on the weight of the combination.

27. The process of any one of claims 1 to 26, wherein the geopolymer foam is cured at a temperature in the range of 60°C to 80°C.

28. The process of any one of claims 1 to 27, wherein the geopolymer foam is cast in a mould and cured within a mould.

29. The process of any one of claims 1 to 28, wherein the foamed geopolymer paste is cast and cured within a frame for a structural panel to form the core of the structural panel.

30. The process of any one of claims 1 to 29, wherein the geopolymer paste further comprises fibres comprising at least one selected from the group consisting of polypropylene, polyethylene, nylon, PVA, glass and basalt.