Influence of metakaolin purities on potassium geopolymer formulation: The existence of several networks

https://doi.org/10.1016/j.jcis.2013.07.024Get rights and content

Highlights

  • Metakaolins reactivity impacts the behavior in the presence of alkaline solution.

  • Highly reactive metakaolin accelerates the consolidation of the material.

  • The metakaolin reactivity can generate the formation of one or several networks.

  • Nanostructure variations after consolidation influence the mechanical properties.

  • A descriptive model of the mechanism of geopolymer formation was proposed.

Abstract

Geopolymer materials are obtained by the alkaline activation of aluminosilicate sources, the best of which is metakaolin. However, every raw material is different, and very few comparative studies have been done on different metakaolin sources. The aim of this work is to develop methods for the prediction of the working properties of geopolymer materials based on the reactivity of the metakaolin employed. Infrared spectroscopy showed direct relationships between the wettability, the Si/Al ratio and the kinetics of conversion of Si–O–Si bonds to Si–O–Al bonds. Moreover, it was demonstrated that the presence of impurities and the reactivity of the metakaolin can generate the formation of one or several networks. Finally, a descriptive model of the mechanism of geopolymer formation was proposed that takes into account the quality of metakaolin used.

Introduction

Geopolymers are synthesized at ambient temperature by the alkaline activation of aluminosilicates derived from natural minerals, calcined clay or industrial byproducts [1]. Usually, this activation is carried out on metakaolin with sodium or potassium silicates. Geopolymers are inorganic binders with good resistance to high temperatures and acid degradation, as well as good mechanical properties. Therefore, they are an attractive alternative to standard cement (Portland), and their use allows for the recycling of large amounts of industrial waste. The mechanical properties of geopolymer materials depend on (i) the alkali cation (Na+ or K+) and (ii) the Si/Al molar ratio [2]. The compressive strengths of potassium-based geopolymers are greater than those of sodium-based materials, each with a Si/Al molar ratio between 1.4 and 1.9 [2]. However, at higher Si/Al ratios (∼2.15) a decrease in mechanical properties is observed due to the presence of unreacted cation after polycondensation reactions [3]. Working properties of these materials also depend on the rate of water add in the mixture [4]. Trapped water in the geopolymer network generates porosity, which results in diminished mechanical properties. The term “geopolymer” is based on the amorphous nature of these materials and the coordination numbers of silicon and aluminum. Their three dimensional structure is composed of SiO4 and MAlO4 tetrahedra, where M is a monovalent cation, typically Na+ or K+. This network is comparable to that of some zeolites but differs in its amorphous character. The polymeric character of these materials increases with the Si/Al ratio, as the aluminum atoms cross link chains of SiO4 tetrahedra. In general, their chemical formulation is in the form {M+ n (SiO2) z, AlO2} n, w H2O, where z is the Si/Al molar ratio, M+ is the monovalent cation and n is the polymerization degree [1]. Altering the Si/Al ratio in geopolymers thus allows the synthesis of materials with different structures. The geopolymerization mechanism is particularly difficult to study on account of the reaction kinetics. However, most authors agree that the mechanism involves dissolution, followed by gel polycondensation [5], [6].

Various characterization techniques, such as infrared spectroscopy [7], thermal analysis (TGA-DTA) [8], rheology [9] or small-angle X-ray scattering (SAXS) [9], can be used to study the formation of the polymer network and test the different mechanisms mentioned above. These techniques allow for the observation of the structural evolution of materials in the polymerization process, including the substitution of Si–O–Si by Si–O–Al bonds (FTIR), dissolution/polycondensation (rheology and TGA-DTA) and oligomer formation (SAXS and TGA-DTA).

The influence of raw material was also studied, focusing on the nature of primary raw materials, the presence of impurities and the role of the metakaolin calcination temperature. Several studies have investigated the use of different clays for the formation of geopolymer materials [10], [11], but it has been shown that metakaolin is ideal [12], [13] because of its high reactivity and purity compared to other materials [14]. A recent study demonstrated the importance of the nature of the kaolin such as kaolinite or halloysite [15] precursor on geopolymer properties as well as the effects of impurities. The results showed that the halloysite-containing kaolin and metakaolin exhibited higher Si and Al dissolution rates than the purer kaolin and metakaolin, which led to better geopolymer properties. The activation temperature of kaolinite (the temperature of dehydroxylation) plays an important role in metakaolin reactivity [16] and also affects geopolymer properties [17], [18]. Indeed, increasing the calcination temperature changes the structure of metakaolin [16], [19] as well as its pozzolanic activity [20] and hydraulicity [21]. However, researchers do not agree on the influence of the calcination temperature on the mechanical properties of metakaolin geopolymers. Elimbi et al. [22] showed that the compressive strength of metakaolin-based geopolymer increases from 11.9 to 36.4 MPa with increasing the calcination temperature from 500 °C to 700 °C, but dropped above 700 °C. In contrast, another study claimed that the mechanical strengths of geopolymers increased with increasing calcination temperature beyond 800 °C [23]. Similarly, Zhang et al. [24] showed an increase in the compressive strength of geopolymers as calcination temperature increased to 900 °C, followed by a sharp decrease in strength at 1000 °C. Nevertheless, several studies have shown that the ideal temperature for the calcination of kaolin to metakaolin is around 750 °C [25], [26], [27], [28]. These differences manifest because of the mineralogical composition of each aluminosilicate source is different. Indeed, each reported material had different purity, cristallinity and secondary phase. These secondary mineral phases may have a low influence on the final properties of the geopolymer [6], [29]. High temperatures may lead to phase transitions [30] or the crystallization of new phases [16]. To eliminate the influence of these impurities, the use of ultrapure kaolin would be ideal. Other works based on mechanochemical pretreatment by high energy grinding showed that it can be possible to obtain setting materials with similar characteristics than materials prepared from dehydroxylated clays [31]. The same author investigated the formation of geopolymers from 2:1 clays and revealed that these materials were not completely amorphous and contained zeolitic compounds [32]. However, the use of ultrapure precursors would result in prohibitively expensive geopolymers; therefore, it is preferable to be able to work with industrial products of varying purity. There are few comparative studies on purity metakaolins in the literature [33]. Those that have been done suggest the difficulty involved in studying the properties of industrial batches, such as surface reactivity of metakaolin, wettability, water demand and structural order [16], [19], [20], [21].

Finally, several studies have been conducted on the temperature behavior of geopolymer materials, demonstrating the crystallization of various phases, such as leucite [33], [34], [35], [36], [37], mullite [34], kalsilite [35] and kaliophilite [36], which produce variations in the composition of geopolymer type materials.

The purpose of this study is the development of a method to predict the working properties of geopolymer type materials from the reactivity of the raw materials (metakaolin). For this, a correlation is initially established between the reactivity of the raw materials and the nanostructural order of the geopolymer type materials. Then, the relation between microstructure and working properties will be determined. The comparison of three metakaolins from different sources, with different purity, particle size and composition, was studied.

Section snippets

Sample preparation

Various samples were synthesized using an alkaline silicate solution (Si/K = 0.7) and permitting the action of the base in solution on metakaolin, as performed in previous works [38]. The alkaline silicate solution was obtained by dissolution of potassium hydroxide pellets (VWR, 85.2% pure) and very fine and highly reactive amorphous silica (denoted S; Aldrich, 99.9% pure) in osmotically purified water. Three kinds of metakaolins were used: (1) M-1000 metakaolin containing 55 wt% SiO2, 40 wt% Al2O3

Characteristics of metakaolin surface reactivity

To study the influence of the aluminosilicate source on the final product and to identify the main differences between raw materials, a preliminary study of the three metakaolins was performed. Metakaolins were characterized by BET analysis, wettability and size distribution (Table 1).

Furthermore, the temperature and the method of calcination are not identical for the 3 metakaolins used which have a direct impact on the powder reactivity. The M-1200 metakaolin is subjected to a 1200 °C flash

Conclusion

In this investigation, the influences of the purity and reactivity of the aluminosilicate precursors (metakaolin) of geopolymer materials were studied. For this purpose, three metakaolins with different BET surfaces, purities, wettabilies and thus different reactivities were used. It was shown that variations in reactivity impact the behavior of metakaolins in the presence of alkaline solution. There are direct relationships between the wettability, the Si/Al ratio and the substitution kinetics

Acknowledgments

The authors thank the IMERYS Ceramic Centre and AGS for the metakaolin samples and Limousin FEDER for financing this study.

References (55)

  • P. Duxson et al.

    Colloids Surf. A

    (2005)
  • E. Prud’homme et al.

    J. Non-Crystal. Solids

    (2011)
  • A. Autef et al.

    J. Non-Crystal. Solids

    (2013)
  • H. Xu et al.

    Int. J. Mineral Process.

    (2000)
  • E. Prud’homme et al.

    Appl. Clay Sci.

    (2011)
  • H. Xu et al.

    Cem. Concr. Res.

    (2002)
  • V.F.F. Barbosa et al.

    Mater. Lett.

    (2003)
  • Y.J. Zhang et al.

    J. Non-Crystal. Solids

    (2012)
  • Z. Zhang et al.

    Cem. Concr. Compos.

    (2012)
  • S. Alonso et al.

    Mater. Lett.

    (2001)
  • R. Cioffi et al.

    Resour. Conserv. Recyc.

    (2003)
  • A.M. Rashad

    Constr. Build. Mater.

    (2013)
  • Ch. Bich et al.

    Appl. Clay Sci.

    (2009)
  • A. Elimbi et al.

    Constr. Build. Mater.

    (2011)
  • M.R. Wang et al.

    Mater. Lett.

    (2010)
  • M. Arikan et al.

    Constr. Build. Mater.

    (2009)
  • F. Zibouche et al.

    Appl. Clay Sci.

    (2009)
  • K.J.D. MacKenzie et al.

    J. Eur. Ceram. Soc.

    (2008)
  • V. Medri et al.

    Appl. Clay Sci.

    (2010)
  • P. Duxson et al.

    J. Non-Crystal. Solids

    (2006)
  • P. He et al.

    J. Eur. Ceram. Soc.

    (2013)
  • A. Autef et al.

    J. Non-Crystal. Solids

    (2012)
  • P. Innocenzi

    J. Non-Crystal. Solids

    (2003)
  • A. Nosrati et al.

    Chem. Eng. Sci.

    (2011)
  • K. Kupwade-Patil et al.

    Comput. Struct.

    (2013)
  • T.M.B. Campos et al.

    J. Eur. Ceram. Soc.

    (2012)
  • P. Duxson et al.

    J. Non-Crystal. Solids

    (2007)
  • Cited by (103)

    View all citing articles on Scopus
    View full text