The Effects of Calcium and Phosphate Compounds on the Mechanical and Microstructural Properties of Fly Ash Geopolymer Mortars

Geopolymers are a kind of inorganic polymer material with a three-dimensional network structure ranging from amorphous to semicrystalline [1]. Initially, geopolymers were strictly defined as alkali-aluminosilicate (AAS) geopolymers produced by the reaction of aluminosilicate precursors (such as clay minerals, metakaolin, fly ash, volcanic ash, slag, and so on) with an alkali activator. Geopolymer, however, can be obtained by two different routes, such as in an alkaline medium (sodium silicate or sodium hydroxide solution) described above and in an acidic medium with phosphoric acid or humic acids [2]. When sodium or potassium silicate is used as a hardener, the geopolymer network is based on poly (sialate), whereas phosphoric acid geopolymers are based on poly (phospho-siloxane). Both systems have been examined separately, with mechanical and microstructural properties as well as reaction processes documented. Geopolymers derived from phosphoric acid solutions were shown to have higher compressive strengths than those derived from sodium silicate solutions [3].

Different calcium sources (slag, limestone, wollastonite, etc.) have been used to investigate their effects on geopolymer composition and mechanical properties [4‐6]. Geopolymer cement can be a mixture of polysialate or sodium, calcium aluminosilicate hydrate (N, C-A-S-H), and calcium silicate hydrate (C-S-H). This system involves the substitution of sodium by calcium during the condensation reaction and the formation of a distinct calcium phase, which is beneficial for strength development [7, 8]. Some researchers, such as Kang et al. [10] analyzed calcium phosphate compounds for producing an inorganic polymer that can be used as bioactive materials for bone tissue scaffolds. Such a system is interesting as calcium could substitute sodium or potassium in the network, and phosphates could replace some silicates during the formation of the geopolymer network, leading to a hybrid system based on Ca, Na-poly(sialate), and poly(phospho-siloxane) networks [11].

However, no literature discusses the influences of phosphate components available in biomass fly ashes on the mechanical and microstructural properties of fly ash-based geopolymers. This work aims to investigate fly ash geopolymer mortars with modified calcium levels and the effects of calcium phosphate compounds on their mechanical and microstructural properties.

The study was performed on geopolymer mortars designed with coal fly ash (RFA) as geopolymerization precursor, standard siliceous natural sand in accordance with EN 196–1 and alkaline activators - sodium hydroxide (NaOH) and calcium additives - quick lime (CaO) or biomass fly ash (BFA). According to Table 1, RFA demonstrates the chemical characteristics needed for Class F fly ash in EN 450–1 and ASTM C618 (the total amount of silica, aluminum, and iron oxides is roughly 83.07%). Due primarily to their contribution to the strength development following alkali-activation, Si and Al elements at high concentrations are essential to the geopolymerization process [12]. Biomass fly ash (BFA) was chosen for its low primary oxide content (less than 1%), but high phosphorus levels (more than 27%).

All geopolymer mortar mixes were prepared and formed according to a procedure modified from EN 196–1 as follows:

A 5M NaOH solution was used to activate the fly ash. Calcium additions (CaO, biomass fly ash - BFA) replaced precursors by 2%, 5%, and 7% of their mass, respectively (Table 2). After 24 h, samples were demoulded and stored in laboratory conditions (20 ± 2 ℃ and 60% relative humidity) until testing. During the conditioning of the samples, no temperature curing was applied. The specimen references are based on general notation RFA-Z#, where “RFA” is the coal fly ash and “Z#” is a symbol of calcium additive – CaO (C), BFA (B) - with the rate of substitution “#”.

After 7, 28, and 56 days, the mortars were tested for flexural and compressive strengths. The test was carried out on a Controls and Instron hydraulic press with a loading rate of 0.33 MPa/min and a sensitivity of 100kN for compressive strength and a loading rate of 0.017 MPa/min and a sensitivity of 5 kN for flexural strength. The result was the average of three flexural strength tests and six compression strength tests.

The Thermal Field Emission Scanning Electron Microscope (FE-SEM Zeiss EVO-40) and Carl-Zeiss EDS analyzer were used to analyze microstructure of mortar samples under high vacuum and 5 kV acceleration.

Samples porosity for chosen samples was measured by micro computed tomography (µCT). µCT test was performed on an XRADIA XCT-400 tomograph at a lamp voltage of 150 kV and a current of 60 A. 900 x-ray images were obtained with exposure time 8 s and resolution of 25 m. Cross-sections of the sample were reconstructed using the Feldkamp algorithm, resulting in over 900 virtual X-Y cross-sections covering a two-dimensional sample area.

At 7, 28, and 56 days, the flexural strength of all CaO mortars was higher than the reference samples (Fig. 1). The increase in strength was further noticed with increasing CaO concentration, allowing for the best flexural results for RFA-N5-C7 (3.8 MPa). RFA mortars containing BFA are at the other end of the spectrum, with no flexural results at 7 days and values below 0.5 MPa at 28 days due to inadequate gepolymerization. The RFA replacement with BFA also had a significant impact on flexural strength growth after 56 days, decreasing from 53% to 73% compared to the RFA-N5-C5.

The Fig. 2 depicts the evolution of compressive strength with increasing calcium content. Compressive strength was less than 2 MPa for the majority of samples after 7 days. The RFA reference sample RFA-N5-R (0.3 MPa) and those with BFA content RFA-N5-B (2,5,7) produce the worst results (0.21 MPa, 0.18 MPa and 0.19, respectively). Only RFA mortars containing 5% and 7% CaO demonstrated greater early compressive strength values (3.0 MPa and 3.7 MPa, respectively). It is consistent with the findings of Buchwald et al., who found that adding Ca(OH)₂ to a geopolymer improved the material's early mechanical performance [13]. The lower RFA reference mortar values were expected because alkali-activated fly ash has a relatively modest rise in mechanical characteristics with time when cured in ambient conditions [14]. RFA mortars with a greater CaO content outperformed reference mortars in terms of strength (CFA-N5-R and RFA-N5-R). The compressive strength growth tendency in CaO mortars is similar to the flexural strength development trend. This conclusion is consistent with recent studies on the increased strength of AAMs with increased calcium concentration [13]. It was also shown that after 56 days, mortars made with 5% and 7% CaO had higher compressive strengths, reaching a peak value close to 10 MPa among all tested fly ash mortars.

In comparison to samples activated with CaO, the mortars activated with high amounts of BFA gave reduced strengths at 28 days, as seen in Fig. 2. At 28 and 56 days, none of the RFA binders with BFA reached the reference value (2.4 MPa and 5.5 MPa, respectively). When compared to RFA-N5-R, compressive strength values at 56 days decreased by 68% when BFA content increased to 7% in RFA mortars.

The μCT test clearly presented differences in pore structure between RFA-N5-R, RFA-N5-C5 and RFA-N5-B5. The RFA-N5-C5 mortar was almost two times less porous than other samples. It had the lowest total porosity (2.74%) with mean pore size 440 µm. s (Fig. 3 b). In RFA-N5-B5 mortar, vast amount of closed spherical pores was observed. This structure can be connected with a production of amorphous calcium phosphate due to high levels of calcium and phosphate in BFA [15]. RFA mortar with BFA content had the highest total porosity (4.29%).

Figure 4 shows SEM images of alkali-activated RFA mortars. The surface morphology of RFA-N5-C5 (Fig. 4c, d) revealed substantially more dense microscale structures than other samples. It can be attributed to faster pozzolanic reaction rates and increased solidification rates of alkali-activated fly ash with calcium addition. According to Lee and Van Deventer [15], the presence of calcium in fly ashes can provide extra nucleation sites for precipitation of dissolved ions and cause fast hardness.

The RFA mortar microstructures revealed reacted amorphous microspheres and some partially reacted fly ash spheres throughout the matrix. During the dissolving of fly ash particles in specific parts of the RFA-N5-C5 sample, some micro shaped pores appeared in the matrix. At higher magnification (Fig. 4 b, f), it is clear that the fly ash spheres in RFA-N5-R and RFA-N5-B5 have not been completely dissolved. That can be justified for the reference sample due to insufficient curing temperature [1]. SEM scans of the sample RFA-N5-B5 revealed that fly ash grains were encased in a thin coating of another chemical substance. EDS analysis revealed that the material comprises phosphorus and is related to amorphous calcium phosphate [16]. Because BFA contains calcium and phosphate compounds, it appears that the retarding effects in binding are linked to the phosphate salts, as stated by Kalina [17].

The following conclusions from the research described above can be drawn: