Synthesis and strength optimization of one-part geopolymer based on red mud

https://doi.org/10.1016/j.conbuildmat.2016.02.099Get rights and content

Highlights

  • One-part geopolymer was synthesized by using Bayer red mud as main raw material.

  • Long-term strength of binder was significantly improved with addition of 20–30 wt% SF.

  • Lower water/solid ratio contributed to increasing the strength.

  • The compressive strength of geopolymer cured for 28 d reached 31.5 MPa.

  • Geopolymerization of dissolved aluminosilicate and silica formed dense matrices.

Abstract

One-part geopolymer was synthesized from alkali–thermal activated Bayer red mud (RM) with addition of silica to optimize its composition. The RM was pretreated through alkali–thermal activation and turned to geopolymer precursor, which could be used by only adding water in blending process. However the long-term strength of the binder with only RM was poor because of the unstable polymerization due to the low SiO2/Al2O3 molar ratio (1.41). Silica fume (SF) was chosen to increase the SiO2/Al2O3 molar ratio of the geopolymer formulation. By adding 25 wt% of SF, the 28 d compressive strength of the geopolymer with a SiO2/Al2O3 molar ratio of 3.45 could reach 31.5 MPa at a water/solid ratio of 0.45. Sodium aluminosilicate in the activated RM dissolved in water and formed an alkaline environment to dissolve SF. The dissolved silica participated in geopolymerization, leading to a satisfactory geopolymer composition. Typical amorphous geopolymer matrices were formed in the binder completely cured.

Introduction

Bayer red mud (RM) or bauxite residue is the residue of bauxite ores after digestion by caustic soda through the Bayer process to produce alumina. It is a high alkaline waste with an average pH of 11.3 ± 1.0 [1] and is classified as a toxic industrial waste [2]. The high alkalinity and superfine particle size make proper disposal of RM difficult. Most of RM is still disposed through storage on land, including lagooning, dry stacking, and dry cake disposal [3]. But land disposal may cause serious environmental pollution, if RM was leaked into the surrounding environment. Ecological disasters caused by RM dam-break have occurred for many times, such as the event in Hungary in 2010 [4].

The research of economical alternatives to utilize red mud have been carried out for more than 50 years. Numerous application possibilities have been researched and developed. The main research areas could be summarized as: metallurgical applications [5], filler or substrate for composite materials [6], catalysts [7], adsorbents [8], construction and building materials [9]. Despite thousands of publications and patents on the subject have been published, large-scale utilization of RM is still absent. Klauber et al. summarized the barriers that need to be overcome for RM utilization as: volume, performance, cost and risk [10]. Research to refine the utilization technology still needs to be conducted. Among the utilization options, construction and building materials pose lower risk for implementation. Manufacture of geopolymers based on RM including controlled low strength materials are one of the suggested research project [10].

Geopolymer poses as a viable alternative for utilizing RM in building materials to avoid the alkali-aggregate reaction since alkali is a necessary component for geopolymer. In recent decades, geopolymer has been attracting worldwide attentions for their low CO2 emissions and high properties. Geopolymers are synthesized by activating solid aluminosilicate sources with alkali metal hydroxide or silicate solutions through a series of dissolution–reorientation–solidification reactions [11]. The binding property of the geopolymer results from the amorphous alkali aluminosilicate gels, which have a general formula as Mn[–(Si–O2)z–Al–O]n·wH2O, wherein M represents one or more alkali metals and z is 1, 2 or 3 [12]. Some geopolymers also contain alkaline earth cations, particularly Ca2+ based on industrial wastes such as granulated blast furnace slag or fly ash, but it’s not sure whether the alkaline earth cations are actually incorporated into the geopolymer structure [13]. The satisfactory geopolymer compositions are suggested to be in the range of M2O/SiO2, 0.2–0.48; SiO2/Al2O3, 3.3–4.5; and H2O/M2O, 10–25 [14].

RM is not a quite ideal material for preparing geopolymer directly due to its poor activity and low SiO2/Al2O3 molar ratio (lower than 2), thus it is usually pretreated and mixed with other materials to prepare geopolymer. Some researches have been done by combining RM with other excellent geopolymer precursors such as metakaolin [15], fly ash [16], [17], [18], and rice husk ash [19] and using sodium hydroxide or sodium silicate solutions as an activator to synthesize geopolymer. In our previous study [20], a type of geopolymer was synthesized from thermal-pretreated RM and granulated blast furnace slag by using sodium silicate as the activator. These geopolymers synthesized by mixing solid aluminosilicate sources with an alkaline activator solution were called as two-part geopolymer for their two-part mix process, which was the conventional design of geopolymer. If the alkali came from the solid phase, and the blending process was just one-part mix (i.e. only need to add water), geopolymer would present the convenience of ordinary Portland cement (OPC). Koloušek et al. proposed the new procedure for synthesizing geopolymer based on direct calcinations of low-quality kaolin with Na/K hydroxides to get one-part geopolymer precursor [21]. Feng et al. synthesized a one-part geopolymer from albite by calcinating it with addition of NaOH and Na2CO3 [22]. One-part geopolymers would present opportunities beyond the traditional two-part geopolymers because one-part geopolymers were ideal for large-scale deployment, as most of the quality control can be dealt with centrally [23].

In the previous work, a one-part geopolymer had been synthesized from Bayer red mud through an alkali–thermal activation process [24]. But the binder collapsed in long-term curing after 7 d since the polymerization of Al–O and Si–O was unstable due to its low SiO2/Al2O3 molar ratio of only 1.41 [24], much lower than the appropriate range of 3.3–4.5. This article presents a research on solving the strength deterioration problem of the one-part geopolymer synthesized from Bayer red mud, by adding another silica-rich material with high activity – silica fume (SF) – to improve the SiO2/Al2O3 molar ratio of the binder, thus to improve the stability of the product.

Section snippets

Raw materials

A local Bayer red mud, provided by an alumina plant of Chalco Co. in Zhengzhou, China, was dried to constant weight at 105 °C and grinded to pass a 0.30 mm mesh sieve. It was a typical residue from the Bayer process to produce alumina using Chinese low-Fe diaspore bauxite ores [25]. The particle size of the RM was in the range of 0.1–70 μm with a median diameter (d50) of 3.5 μm as determined by laser granulometry. A condensed silica fume, provided by China Construction Ready Mixed Concrete Co. Ltd.

The pH transformation of the binders

The pHs of leaching solutions of the one-part geopolymer binders prepared from RM-10N are shown in Fig. 2. The leaching solution of the one-part geopolymer binders pose alkaline, with pHs varying in the range of 11.2–13.2. The alkaline environment in the geopolymer binders results from the dissolution of sodium aluminosilicates in alkali–thermal activated RM, and the pH declines with the increase of SF addition. When the curing age is extended from 3 d to 7 d, the pHs decline, and the reduction

Conclusions

In this work, we have synthesized a novel one-part geopolymer cement by using Bayer red mud as the main raw material through alkali–thermal activation, with silica fume as an additive to improve the SiO2/Al2O3 molar ratio. This approach can greatly contribute to the beneficial reuse of massive Bayer red mud.

The conclusions are as follows:

  • 1.

    SF worked well in improving the long-term strength of the one-part geopolymer binder by improving the stability of the structure. At the addition of SF of 20–30

Acknowledgments

The authors would like to appreciate the financial support of the New Century Excellent Talents Project of Ministry of Education (NCET-10-0392), the Public Welfare Program of Environmental Protection Ministry of China (201509056), and the Project of Innovative and Interdisciplinary Team of Huazhong University of Science and Technology (HUST) (0118261077). The authors would also like to appreciate the Analytical and Testing Center of HUST for the microstructure characterization tests.

References (38)

Cited by (0)

View full text