High calcium fly ash geopolymer stabilized lateritic soil and granulated blast furnace slag blends as a pavement base material

Accepted Manuscript Title: High Calcium Fly Ash Geopolymer Stabilized Lateritic Soil and Granulated Blast Furnace Slag Blends as a Pavement Base Material Authors: Itthikorn Phummiphan, Suksun Horpibulsuk, Runglawan Rachan, Arul Arulrajah, Shui-Long Shen, Prinya Chindaprasirt PII: DOI: Reference:

S0304-3894(17)30576-9 http://dx.doi.org/doi:10.1016/j.jhazmat.2017.07.067 HAZMAT 18759

To appear in:

Journal of Hazardous Materials

Received date: Revised date: Accepted date:

25-1-2017 26-7-2017 29-7-2017

Please cite this article as: Itthikorn Phummiphan, Runglawan Rachan, Arul Arulrajah, Shui-Long Shen, High Calcium Fly Ash Geopolymer Stabilized Lateritic Blast Furnace Slag Blends as a Pavement Base Material, Materialshttp://dx.doi.org/10.1016/j.jhazmat.2017.07.067

Suksun Horpibulsuk, Prinya Chindaprasirt, Soil and Granulated Journal of Hazardous

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Revised manuscript HAZMAT-D-17-00478R2 High Calcium Fly Ash Geopolymer Stabilized Lateritic Soil and Granulated Blast Furnace Slag Blends as a Pavement Base Material Itthikorn Phummiphan, B.Eng., M.Eng., Ph.D. Ph.D. Graduate, Graduate Program in Construction and Infrastructure Management, School of Civil Engineering, Suranaree University of Technology, 111 University Avenue, Muang District Nakhon Ratchasima 30000, THAILAND Suksun Horpibulsuk, B.Eng. (Hons), M.Eng., Ph.D., P.E. Professor and Chair, School of Civil Engineering, and Chair, Center of Excellence in Innovation for Sustainable Infrastructure Development, Suranaree University of Technology, Tel: +66-44-22-4322, Fax: +66-44-22-4607 Email: [email protected] Runglawan Rachan, B.Eng., M.Eng., Ph.D., P.E. Assistant Professor, Department of Civil Engineering, Mahanakorn University of Technology, Nong Chok District, Bangkok 10530, THAILAND E-mail: [email protected] Arul Arulrajah, B.Eng., M.Eng., Ph.D. Professor, Department of Civil and Construction Engineering, Swinburne University of Technology, Melbourne, AUSTRALIA Email: [email protected] Shui-Long Shen, B.Eng., M.Eng., Ph.D. Professor, Department of Civil Engineering, Shanghai Jiao Tong University and State Key Laboratory of Ocean Engineering, 800 Dong Chuan Road, Minhang District, Shanghai 200240, CHINA Tel: (86)21-3420-4301; Fax: (86)21-6419-1030; E-mail: [email protected] Prinya Chindaprasirt, B.Eng., M.Eng., Ph.D. Professor, Sustainable Infrastructure Research and Development Center, Department of Civil Engineering, Khon Kaen University, Khon Kaen, THAILAND Email: [email protected] Date written: 26 July 2017

1

Number of words: 4,799 NOTE: The second author is the corresponding author. Please mail communication to Prof. Suksun Horpibulsuk, School of Civil Engineering, Suranaree University of Technology, 111 University Avenue, Muang District, Nakhon Ratchasima 30000, THAILAND

Research Highlights    

Evaluation of Fly Ash (FA) based geopolymer stabilized lateritic soil/GBFS blend. Role of L, NaOH/Na2SiO3, GBFS content, and curing time were investigated. Microstructural development was examined via XRD and SEM analyses. UCS of FA geopolymer stabilized blends was compared with road authorities’ requirements.

Abstract: Granulated Blast Furnace Slag (GBFS) was used as a replacement material in marginal lateritic soil (LS) while class C Fly Ash (FA) was used as a precursor for the geopolymerization process to develop a low-carbon pavement base material at ambient temperature. Unconfined Compression Strength (UCS) tests were performed to investigate the strength development of geopolymer stabilized LS/GBFS blends. Scanning Electron Microscopy and X-ray Diffraction analysis were undertaken to examine the role of the various influencing factors on UCS development. The influencing factors studied included GBFS content, Na2SiO3:NaOH ratio (NS:NH) and curing time. The 7-day soaked UCS of FA geopolymer stabilized LS/GBFS blends at various NS:NH ratios tested was found to satisfy the specifications of the Thailand national road authorities. The GBFS replacement was found to be insignificant for the improvement of the UCS of FA geopolymer stabilized LS/GBFS blends at low NS:NH ratio of 50:50. Microstructural analysis indicated the coexistence of Calcium Silicate Hydrate (CSH) and Sodium Alumino Silicate Hydrate products in FA geopolymer stabilized LS/GBFS blends. This research enables GBFS, which is traditionally considered as a waste material, to be used as a replacement and partially reactive material in FA geopolymer pavement applications.

2

Keywords: geopolymer; lateritic soil; fly ash; granulated blast furnace slag; pavement base 1. INTRODUCTION Ordinary Portland Cement (OPC) is the most widely used cementing agent in civil engineering infrastructure projects. The production of OPC, however, emits a large amount of greenhouse gases, notably CO2 into the atmosphere. Cement production worldwide discharges up to 4.0 billion tons of CO2 annually (Davidovits 1994; Part et al. 2015; Yusuf et al. 2014). The production of just 1 ton of OPC releases about 1 ton of CO2 (Davidovits 2013). Pollution and global warming coupled with growing public environmental awareness has been increasing rapidly in many developed and developing countries. Environmentally friendly alternative construction materials are increasingly being sought (Part et al. 2015). Geopolymer, a novel green cementing agent manufactured from various industrial waste byproducts, is considered as an alternative materials to OPC. Geopolymer is an inorganic aluminosilicate material synthesized by mixing source materials rich in silica (SiO2) and alumina (Al2O3) such as Fly Ash (FA), metakaolin, Granulated Blast Furnace Slag (GBFS) and Silica Fume (SF) with alkali activators (De Silva et al. 2007). The engineering properties of geopolymers that are sought for civil engineering applications include: high compressive strength (Amnadnua et al. 2013; Bagheri and Nazari 2014; Duxson et al. 2007), rapid controllable setting and hardening (Lee and Van Deventer 2002), fire resistance (Cheng and Chiu 2003; Sakkas et al. 2014; Sarker et al. 2014), acid and salt solution resistance (Palomo et al. 1999), lack of deleterious alkali–aggregate reactions and low shrinkage (Zhang et al. 2013). The commonly used activators for geopolymerization include alkaline metal and alkaline earth metal compounds (Komnitsas and Zaharaki 2007). Generally, the most effective activator providing the best performance for high strength and other advantageous 3

properties is a mixture combining sodium hydroxide (NaOH) and sodium silicate (Na2SiO3) (Rashad 2013). The use of large quantities of sodium silicate is not recommended for the environment, as it imparts a high carbon footprint alkali. The emission factors are 1.514 kg CO2-e/ton (Turner and Collins, 2013) and 0.86 kg CO2-e/ton (McLellan et al., 2011) for sodium silicate and cement production, respectively. Both class F and class C Fly Ash (FA) as precursors have been extensively used for the development of geopolymers (Palomo et al. 2007; Temuujin et al. 2009; Hoy et al., 2015). The class C FA contains high calcium oxide (CaO) and is designated as a self-cementing FA. The Class C Fly ash has been used for making concrete and soil stabilization without cement (Roskos et al. 2011; Misra et al. 2005; Sezer et al. 2006). The reaction between class C FA and the liquid alkaline activator therefore forms Calcium Silicate Hydrate (CSH) and Calcium Alumino Hydrate (CAH), which co-exists with geopolymerization products. Mechanical properties and microstructure of class F FA geopolymer at ambient temperature were found to be improved by including very fine GBFS as an additive (Kumar et al. 2010; Puertas et al. 2000). This is as the FA-GBFS geopolymer system will form aluminummodified calcium silicate hydrate (CASH) gel which coexists with sodium aluminosilicate hydrate (NASH) gel (Bernal et al. 2013; Ismail et al. 2014). The coexistence of CASH and NASH was also evident for other FA and calcium-rich additive based geopolymers (Guo et al. 2010; Somna et al. 2011). Several researchers have used geopolymers in concrete applications. However, only in recent years research has been undertaken on geopolymer improved demolition waste materials and soft soils (Arulrajah et al. 2015; Horpibulsuk et al. 2013; Mohammadinia et al. 2016; Phetchuay et al. 2014; Phummiphan et al. 2016a and b; Sukmak et al. 2013; Suksiripattanapong et al. 2015). Phummiphan et al. (2016a) have first introduced the usage of high calcium FA-based geopolymer to stabilized marginal lateritic soil to develop a green 4

pavement base in Thailand. The early strength of the geopolymer stabilized marginal lateritic soil was found to be enhanced by using a waste Calcium Carbide Residue (CCR) as an additive (Phummiphan et al., 2016b). The CCR was small-sized and was proven to be a binder, which reacted with silica and alumina from soil and FA, to form Calcium Silicate Hydrate (CSH) (Kampala and Horpibulsuk, 2013). Phetchuay et al. (2016) first introduced a FA geopolymer to stabilize soft marine clay in Australia and reported that the CCR not only improved the strength of geopolymer stabilized clay but also provided low carbon-emission when compared to cement stabilization. Similarly, fine sized GBFS as an additive has been demonstrated to enhance the engineering properties of class F FA geopolymer stabilized recycled demolition aggregate (Arulrajah et al. 2016 and Mohammadinia et al. 2016) and FA geopolymer stabilized spent coffee grounds (Kua et al., 2016) for pavement base/subbase and subgrade applications, respectively. The very fine GBFS particles can react rapidly with a liquid alkaline activator and improve the early strength of FA geopolymer stabilized materials. However, crushing of GBFS to very fine particles is a costly and energy-intensive process. A cost-effective and environmental-friendly means is to reuse medium-sized GBFS as a replacement as well as a partially reactive material for geopolymer stabilization. The study on the usage of medium-sized GBFS as a partial replacement of lateritic soil in class C FA geopolymer stabilization for sustainable pavement application is to date limited and is the focus of this research. The usage of by-products (class C FA and GBFS) and liquid alkaline activator without Portland cement to stabilize marginal lateritic soil is novel and significant in terms of engineering, economic and environmental perspectives. Unconfined Compression Strength (UCS) was used to evaluate the strength development of the FA geopolymers stabilized soil/GBFS blends. Scanning Electron Microscopy (SEM) and X-ray Diffraction (XRD) analysis were undertaken to investigate the effect of various 5

influencing factors on the strength development. The factors studied included GBFS content, Na2SiO3:NaOH (NS:NH) ratio and curing time. The outcome of this study will enable GBFS, an industrial by-product, to be used as a replacement and partial reactive material in FA geopolymer pavement applications. 2. MATERIALS AND PROPERTIES 2.1 Soil sample Lateritic Soil (LS) samples were obtained from a quarry in the city of Rayong, Thailand. The Liquid Limit (LL), Plastic Limit (PL), and Plasticity Index (PI) in accordance with ASTM D4318 (ASTM 2010) were 27.72%, 21.65%, and 6.07%, respectively while the specific gravity (Gs) was 2.58. The natural water content of LS was 10%. The grain size distribution was determined by sieve analysis (ASTM 2007a) as shown in Figure 1 and was compared to that specified for base/subbase materials by AASHTO and Department of Highways, Thailand (AASHTO 2012; DOH 1989). The gradation of LS was within the specified limits. The LS was a silty clayey sand (SC-SM) and A-2-4(0) according to the Unified Soil Classification System (USCS) (ASTM 2011) and the AASHTO (ASTM 2009), respectively. The optimum moisture content (OMC) and maximum dry unit weight (γd,max) of LS under modified Proctor energy (ASTM 2012b) were 8.0% and 20.85 kN/m3, respectively. California Bearing Ratio (CBR) value at 95% of γd,max was 14.7%. Los Angeles (LA) abrasion in accordance with ASTM C131 (ASTM 2006) and ASTM C535 (ASTM 2012a) was 52.9%. When comparing the CBR and LA abrasion results to the specification of Department of Highways for subbase and engineering fill materials (DOH 1989) (Table 1), LS did not meet the specified subbase requirement but met the requirements of an engineering fill material. In remote construction sites, located far away from high quality quarry materials, chemical stabilized LS can potentially be used as pavement construction materials. The

6

chemical stabilization of LS will lead to savings in haulage costs and will furthermore minimize negative environmental impacts. Chemical composition of LS obtained from X-ray Fluorescence (XRF) analysis using Horiba XGT-5200 X-ray Analytical Microscope is shown in Table 2. The samples were ground to essentially similar particle sizes and were then compressed by a pressed pellet with 200 kN force for 30 seconds to have a 40 mm diameter specimen prior to XRF analysis. The iron oxides in lateritic soil was 10.9%, which is within the typical value (5.6% and 23.5% with most values less than 12%) (Chanthaburi, Rayong and Chonbuuri) previously reported by Tantiwanit and Changsuwan (1995). The SEM image (Figure 2 (a)) showed that LS was irregular shape. The main mineral components of LS were quartz, muscovite, illite and montmorillonite as shown in Figure 3(a). Peak intensity of crystalline elements composed of calcite, calcium silicate, gehlenite and calcium aluminium oxide in LS occurred between 25o 2 and 43o 2. 2.2 Fly Ash Fly Ash (FA) was collected from the Mae Moh power plant in Thailand. The chemical composition obtained from XRF of FA is shown in Table 2. The main components of the FA (SiO2, Fe2O3, Al2O3) were 70.44% while CaO content was 26.73%. The FA was therefore classified as a high calcium class C. Particle size distribution and microstructure of FA are illustrated in Figure 1 and Figure 2b, respectively. The FA particles were fine-grained and spherical in shape. The XRD pattern of FA showed that FA consisted mainly of glassy phase materials (amorphous humps between 7o 2 and 25o 2) with some crystalline additions of anhydrite, quartz, calcite and maghemite between 20o 2 and 37o 2 (Figure 3b). (c) 2.3 Granulated Blast Furnace Slag

7

Granulated Blast Furnace Slag (GBFS) samples were obtained from Siam Steel Mill Services Co., Ltd., Chonburi Thailand. The GBFS was passed through a 2 mm (No.10) sieve, and air-dried at room temperature. GBFS was non-plastic with a specific gravity of 3.54. The main chemical compositions were 30.38% CaO, 26.78% SiO2, 18.65% Fe2O3, 9.70% MgO and 6.96% Al2O3 as shown in Table 2. The particle size distribution is shown in Figure 1. The particle size was similar to that of LS and much larger than that of FA. The SEM image (Figure 2c) indicated that the particles were generally irregular in shape. The XRD pattern showed that GFBS contained traces of calcite, calcium silicate, gehlenite and calcium aluminium oxide (Figure 3b). The amorphous humps of GBFS were in the region of 8o 2 to 20o 2. Even though the chemical compositions of the GBFS and FA were similar, however the FA was considered more reactive than GBFS due to its higher reactive surface area. 2.4 Liquid alkali activator The liquid alkaline activator (L) used was a mixture of sodium silicate (Na2SiO3, NS) solution and sodium hydroxide (NaOH, NH) solution with a concentration of 5 molars. The suitable concentration for L was generally between 4.5 and 18 molars (Andini et al. 2008; Chindaprasirt et al. 2009; Hanjitsuwan et al. 2014; Rattanasak and Chindaprasirt 2009; Somna et al. 2011). A low NH concentration of 5 molars was considered in this study to avoid health harm of workers and to have cost-effectiveness. Na2SiO3 solution consisted of Na2O, SiO2 and H2O which are 15.50%, 32.75%, and 51.75% by weight, respectively. Distilled water was used throughout the experiments to produce the NH solution. 3. SAMPLE PREPARATION OF FA GEOPOLYMER STABILIZED SOIL In this study, the LS was oven-dried at 40oC for 3 days before mixing with FA and GBFS. The FA content was fixed at 30% of the total mix as suggested by Phummiphan et al. (2016a) and Phummiphan et al. (2016b) while the GBFS contents were varied to replace LS 8

from 10% to 30%. The LS:FA:GBFS ratios were thus 60:30:10, 50:30:20, and 40:30:30. Phummiphan et al. (2016b) reported that the optimal NS:NH ratio providing the highest strength of FA geopolymer stabilized LS was between 100:0 and 50:50. The NS:NH ratios studied were 100:0, 90:10, 80:20, 50:50 and 40:60. Due to the large GBFS particles, even with similar chemical composition to FA, the main geopolymerization reaction in the stabilized LS can be attributed to the presence of FA. The LS, FA and GBFS were firstly mixed together to ensure homogeneity by a soil mixer and then the L was sprayed on the LS, FA and GBFS mixture during mixing for an additional 5 minutes. The mixtures were next compacted in a standard 101.6 mm diameter and 116.4 mm height mold under modified Proctor energy according to ASTM (ASTM 2012b). The compacted samples were demolded and immediately wrapped with plastic sheets and cured at room temperature between 27 to 30oC. It was noted from the laboratory observation that the setting time of class C FS geopolymer stabilized lateritic soil was about 30 minutes. In this study, the sample preparation and compaction were completed within 15 minutes. This stabilization method can be applied in pavement base and subbase applications in practice, where compaction for each section must be finished within the setting time. According to the Department of Highways (DOH) and Department of Rural Roads, Thailand as well as road authorities in Australia and other countries, the UCS is the critical design parameter and as such was used for comparison purposes in this study. However, cyclic tests such as resilient modulus and permanent strain are also important to understand the material performance and are used for a mechanistic design and are recommended for future study for a separate publication. The soaked UCS value of the samples just after mixing was 0. The UCS of soaked geopolymer samples was thus measured after curing periods of 7, 28, and 60 days. The UCS

9

samples were tested using the Universal Testing Machine (UTM) with a compression rate of 1 mm/min in accordance with ASTM (ASTM 2007b), the specification of Department of Highways (DOH 2000) and Department of Rural Roads (DRR 2013), Thailand. The stressstrain and soil modulus were not presented in this paper as they must be determined from sample with a 1 : 2 diameter : height ratio, while the UCS samples were 1 : 1.15 diameter : height ratio, which is specified by various national local road authorities in Thailand and some other countries. The growth of the geopolymerization products was demonstrated using SEM and XRD analysis (Latifi et al., 2016). The small samples were frozen at -195oC by immersion in liquid nitrogen and coated with gold before SEM analysis (Sukmak et al. 2013). The sample surface was scanned with a focused beam of electrons using JEOL JSM-6010LV device. The samples were also ground to fine powder for XRD tests to obtain microstructural information of amorphous and crystalline phases. The XRD scans were performed at 0 – 90o2 by Bruker D8 ADVANCE device. The XRD analysis using Cu X-ray tube was done on powdered samples and patterns were obtained by scanning at 0.1o (2θ) per min and at steps of 0.05o (2θ). The voltage and current of the XRD analysis were set at 45 kV and 40 mA, respectively. 4. RESULTS AND DISCUSSION 4.1 Compaction characteristics of FA geopolymer stabilized LS/GBFS blends Figure 4 illustrates the relationships between γd and liquid alkaline activator, L content of the FA geopolymer stabilized LS/GBFS blends for various LS:FA:GBFS ratios and NS:NH ratios. The γd,max values of FA geopolymer stabilized blends for all GBFS contents tested were higher than the γd,max value of the lateritic soil (without stabilization) possibly because GBFS has a higher specific gravity than lateritic soil. The compaction curves of FA geopolymer stabilized LS/GBFS blends were dependent on the GBFS content 10

and NS:NH ratio. For a particular NS:NH ratio and GBFS content, γd of the FA geopolymer stabilized LS/GBFS blends increased with increasing LA content until γd,max reached an Optimum Liquid Content (OLC). Beyond the OLC, γd decreased with the increased L content. For all the NS:NH ratios, γd,max increased with increasing GBFS content, due to the high specific gravity of GBFS. For 20% and 30% GBFS, γd,max tended to increase with decreasing the NS content while the γd,max for the 10% GBFS sample showed the opposite trend. However, the change in γd,max for all GBFS contents were insignificant when compared to the change in OLC. The OLC significantly decreased with decreasing NS (increasing NH) because NH had lower viscosity than NS (the viscosity of NH and NS was about 1.002 and 400 mPa.s at 20oC, respectively). In other words, NH lubricated the soil particles and hence improved the compactibility. The relationships between OLC and NS contents for various GBFS contents could be represented as linear functions as shown in Figure 5.

4.2 Unconfined compressive strength of FA geopolymer stabilized LS/GBFS blends The relationships between UCS and GBFS content of the FA geopolymer stabilized LS/GBFS blends at different curing times and NS:NH ratios are shown in Figure 6. The 7-day UCS of all stabilized materials at various LS:FA:GBFS ratios compacted at OLC were higher than the UCS specified by the Department of Rural Roads and Department of Highways of Thailand (> 1.724 MPa for light traffic (DRR, 2013) and > 2.413 MPa for heavy traffic (DOH, 2000)) and by the Australia road authority (> 3.5 MPa (VicRoads, 2011)). As shown in Figures 6a – 6d, the 7-day UCS development of the stabilized LS/GBFS blends was dependent upon NS:NH ratio. For NS:NH ratios of 100:0, 90:10 and 80:20, the 7-day UCS values increased with increasing GBFS content until the maximum values at optimal GBFS content were attained and subsequently decreased. The optimal GBFS content decreased with decreasing

11

NS:NH ratios. The highest 7-day UCS was found at 20% GBFS (optimal) and NS:NH = 100:0, which was equal to 10.46 MPa. The same was however not observed for NS:NH = 50:50; i.e., GBFS content did not affect the UCS development. For the long-term curing (28 to 60 days), two strength characteristics were noted for high NS:NH ratios (NS:NH = 100:0, 90:10 and 80:20) and low NS:NH ratio (NS:NH = 50:50). Referring to Figures 6a-6c for the high NS:NH ratios, the LS:FA:GBFS = 60:30:10 was regarded as optimal. The 60-day UCS values at LS:FA:GBFS = 60:30:10 were 18.02 MPa, 20.21 MPa and 18.74 MPa for NS:NH = 100:0, 90:10 and 80:20, respectively. In other words, the highest long-term UCS was found at LS:FA:GBFS = 60:30:10 and NS:NH = 90:10. For the low NS:NH of 50:50 (Figure 6d), GBFS did not affect the UCS development; i.e., GBFS was not required for the NS:NH = 50:50. Even though the 7-day UCS was lowest compared to that at higher NS:NH ratios, the 28-day and 60-day UCS values for LS:FA:GBFS = 70:30:0 (no GBFS) were the highest. NH had a significant effect on long term UCS development because the reaction between NH and FA is time-dependent; hence, the CSH develops at a long curing time of 90 days (Phummiphan et al., 2016b). The highest 7-day UCS was found at NS:NH = 100:0 and LS:FA:GBFS = 50:30:20 while the highest 28-day and 60-day UCS values were found at NS:NH = 90:10 and LS:FA:GBFS = 60:30:10. However, the 7-day UCS at NS:NH = 90:10 and LS:FA:GBFS = 60:30:10 was just slightly lower than that at NS:NH = 100:0 and LS:FA:GBFS = 50:30:20. As such, NS:NH = 90:10 and LS:FA:GBFS = 60:30:10 are recommended in practice. The very high 30% GBFS content (LS:FA:GBFS = 40:30:30) must however be avoided, as it retarded the UCS development especially high NS:NH; i.e. the UCS development was insignificant after 28 days of curing. 4.3 SEM analysis

12

(a)

Investigation of microstructural development of FA geopolymer stabilized LS/GBFS

blends via SEM analysis is advantageous for observing the chemical reaction and the growth of cementation matrix over the curing period. As the 60:30:10 LS:FA:GBFS and 90:10 NS:NH are recommended in practice, the microstructural development with time (7, 28 and 60 days) of the sample with this ingredient was examined and is presented in Figure 7. For early curing time of 7 days, the chemical reaction occured on the FA surface (Figure 7a). The etched holes and cementitious products on FA particles, which indicated the chemical reaction between L and FA, were observed after 28 days of curing (Figure 7b). More etched holes and cementitious products on the shell of FA particles were clearly detected after 60 days of curing (Figure 7c). This indicated the leaching of silica and alumina oxides from FA by alkaline dissolution with time. Hence, the UCS continued to develop even after 60 days of curing. Figure 8 presents SEM images of the FA geopolymer stabilized LS/GBFS blend at high 30% GBFS (LS:FA:GBFS = 40:30:10) for both high and low NS:NH ratios = 90:10 and 50:50 and cured for 7, 14, 28 and 60 days in order to examine the negative impact of very high GBFS contents. For high NS:NH ratio of 90:10, the microstructures changed insignificantly with increasing curing times (Figures 8c, 8e and 8g), which was notably different with those at 60:30:10 LS:FA:GBFS (Figure 7). This indicated that GBFS retarded the geopolymerization process over time. On the contrary, the microstructural change with time was noted for low NS:NH ratios of 50:50 (Figure 8d, 8f and 8h), indicating that high NH caused time-dependent reaction with FA. This also confirmed that GBFS did not affect the UCS development with time even at very high content of 30% as shown in Figure 6d. 4.4 XRD analysis

13

The XRD patterns of FA geopolymer stabilized LS/GBFS blend at 50:30:20 LS:FA:GBFS and at high and low NS:NH ratios of 90:10 and 50:50 are presented in Figure 9 for 7 and 60 days of curing to investigate the role of NS:NH ratios on the UCS development at a particular GBFS content. The main traces consisted of Quartz, Muscovite, Microcline and Albite. The Quartz phases were mainly from the LS. This indicated that LS was inert and insignificant reacted with liquid alkaline activator. The presence of sodium alumino-silicate in the form of Albite and the potassium alumino-silicate in form of Muscovite and Microcline indicated the formation of geopolymer products. For the sample with high NS:NH ratio = 90:10 at early stage (7 days – line A), the main traces were Calcium Silicate Hydrate (CSH); Anorthite; Goethite while at a longer time (60 days) the main traces were Anorthite, Goethite, Calcite with little CSH. This indicated variation of cementitious products present at various curing times. The CSH traces in the 7 days cured sample were considerably higher than those in 60 days cured sample. This confirmed that the significant CSH products occurred in the early stages of the geopolymer process mainly due to the chemical reaction between high calcium FA and NS. On the other hand, for the sample with low NS:NH ratio of 50:50, the main traces were geopolymer products in the form of Anorthite, Goethite and Calcite, with no appearance of CSH in early 7 days (see line C in Figure 9) whereas the 60-day traces were Anorthite, CSH, Goethite and Calcite. It was noted that CSH traces were detected in 60 days of curing but not in 7 days of curing while Anorthite, Goethite and Calcite in 60 days of curing were slightly more than those in 7 days of curing. This implied that the CSH products in geopolymeric system were time-dependent for high calcium FA geopolymer with low NS:NH ratio.

14

Notably, the chemical products for the stabilized samples at each curing time depended on the NS:NH ratio. The high CaO (mainly from FA) in the system, contributed to the development of CSH in the early stage for high NS:NH ratio. Hence, the UCS results of the samples resulted from the coexistence of CSH, calcite and the NASH geopolymer gel. However, the excessive CaO will react with CO2 to form Calcite (Phoo-ngernkham et al. 2015; Ravikumar et al. 2010). A high contents of 30% GBFS was thus found to provide a negative effect on the UCS development. Without GBFS, the 7-day UCS was found at NS:NH = 90:10 while the long-term UCS was found at NS:NH = 50:50. The high short-term UCS was mainly due to the rapid reaction between NS and CaO in FA, hence CSH development while the high long-term UCS was mainly due to the time-dependent reaction between NH and SiO2 and Al2SiO3 in FA, hence NASH development (Phummiphan et al., 2016b). The GBFS replacement increased CaO, SiO2 and Al2SiO3, hence the very high short term and long term UCS were found at NS:NH = 90:10. 5. CONCLUSIONS This paper studied the viability of using two types of wastes (class C FA and GBFS) with a liquid alkaline activator to stabilize marginal lateritic soil to form a sustainable pavement base material. The effect of GBFS, NS:NH ratio and curing time on the strength development and microstructure of FA based geopolymer stabilized LS/GBFS blends was investigated. The 7-day UCS of the FA geopolymer stabilized LS/GBFS blends at various NS:NH ratios and GBFS contents was comparable to the national road authority specifications of Thailand and Australia for cement stabilized pavement material. The following conclusions can be drawn from this study:

15

1. The soaked 7-day UCS of FA geopolymer stabilized LS/GBFS blends at various NS:NH ratios met the strength requirements for both low and high volume roads specified by Department of Highways and Department of Rural Roads, Thailand. 2. GBFS replacement improved the early 7-day UCS of FA geopolymer stabilized LS at the high NS:NH ratios tested (especially at NS:NH ≥ 80:20). The optimal GBFS content providing the highest 7-day UCS tends to decrease with decreasing NS:NH ratio. The GBFS had little effect on early and long-term UCS of FA geopolymer stabilized LS at low NS:NH = 50:50. The highest long term 28-day and 60-day UCS was found at 60:30:10 LS:FA:GBFS and 90:10 NS:NH, which are the recommended optimum ingredients in practice. 3. For high NS:NH ratios (> 80:20), the SEM and XRD results indicated that the cementitious products increased in volume when the curing time increased at optimal GBFS; i.e., NS:NH = 90:10 and GBFS = 10%. 4. The coexistence of CSH and NASH for various NS:NH ratios was explained by the XRD results. The significant CSH was detected at an early stage of curing for high NS:NH while the significant CSH was detected after a long curing time for low NS:NH. The excessive CaO would react with CO2 and form Calcite for both high and low NS:NH ratios; i.e., 30% GBFS provided a negative effect on the UCS development. 5. The outcome of this research will enable GBFS to be used as a replacement material with FA geopolymer to stabilize LS in the development of a low carbon stabilized pavement base. 10% content of GBFS is recommended at high NS:NH ratios (> 80:20). ACKNOWLEDGEMENTS

16

This work was supported by the Thailand Research Fund under the TRF Senior Research Scholar program Grant No. RTA5980005 and Suranaree University of Technology. The first author also acknowledges a financial support from Department of Rural Roads, Thailand for his Ph.D. studies.

REFERENCES AASHTO (2012). ""Standard specification for materials for aggregate and soil-aggregate subbase, base, and surface courses." AASHTO M147-65, American Association of State Highway and Transportation Officials.". Ahmedzade, P., and Sengoz, B. (2009). "Evaluation of steel slag coarse aggregate in hot mix asphalt concrete." J. Hazard. Mater., 165(1-3), 300-305. Amnadnua, K., Tangchirapat, W., and Jaturapitakkul, C. (2013). "Strength, water permeability, and heat evolution of high strength concrete made from the mixture of calcium carbide residue and fly ash." Mater. Des., 51, 894-901. Andini, S., Cioffi, R., Colangelo, F., Grieco, T., Montagnaro, F., and Santoro, L. (2008). "Coal fly ash as raw material for the manufacture of geopolymer-based products." Waste Manag, 28(2), 416-423. Arulrajah, A., Kua, T.-A., Phetchuay, C., Horpibulsuk, S., Maghoolpilehrood, F., and Disfani, M. M. (2015). "Spent coffee grounds-fly ash geopolymer used as an embankment structural fill material." J. Mater. Civ. Eng., 10.1061/(ASCE)MT.1943-5533.0001496, 04015197. Arulrajah, A., Mohammadinia, A., Phummiphan, I., Horpibulsuk, S., and Samingthong, W. (2016). "Stabilization of Recycled Demolition Aggregates by Geopolymers comprising Calcium Carbide Residue, Fly Ash and Slag precursors." Constr. Build. Mater., 114, 864-873. ASTM (2006). "ASTM C131. Standard Test Method for Resistance to Degradation of SmallSize Coarse Aggregate by Abrasion and Impact in the Los Angeles Machine." ASTM, West Conshohocken, PA, USA. ASTM (2007a). "ASTM D422-63. Standard Test Method for Particle-Size Analysis of Soils1." ASTM West Conshohocken, PA, USA. ASTM (2007b). "ASTM D1633. Standard Test Methods for Compressive Strength of Molded Soil-Cement Cylinders." ASTM International, West Conshohocken, PA, USA. ASTM (2009). "ASTM D3282. Standard Practice for Classification of Soils and SoilAggregate Mixtures for Highway Construction Purposes." ASTM, West Conshohocken, PA, USA. ASTM (2010). "ASTM D4318. Standard Test Methods for Liquid Limit, Plastic Limit, and Plasticity Index of Soils." ASTM, West Conshohocken, PA, USA. ASTM (2011). "ASTM D2487. Standard Practice for Classification of Soils for Engineering Purposes (Unified Soil Classification System." ASTM West Conshohocken, PA, USA. ASTM (2012a). "ASTM C535. Standard Test Method for Resistance to Degradation of LargeSize Coarse Aggregate by Abrasion and Impact in the Los Angeles Machine." ASTM, West Conshohocken, PA, USA.

17

ASTM (2012b). "ASTM D1557. Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Modified Effort (56,000 ft-lbf/ft3 (2,700 kN-m/m3)." ASTM International. Bagheri, A., and Nazari, A. (2014). "Compressive strength of high strength class C fly ashbased geopolymers with reactive granulated blast furnace slag aggregates designed by Taguchi method." Mater. Des., 54, 483-490. Bernal, S. A., Provis, J. L., Walkley, B., San Nicolas, R., Gehman, J. D., Brice, D. G., Kilcullen, A. R., Duxson, P., and van Deventer, J. S. J. (2013). "Gel nanostructure in alkali-activated binders based on slag and fly ash, and effects of accelerated carbonation." Cem. Concr. Res., 53, 127-144. Cheng, T. W., and Chiu, J. P. (2003). "Fire-resistant geopolymer produced by granulated blast furnace slag." Miner. Eng., 16(3), 205-210. Chindaprasirt, P., Jaturapitakkul, C., Chalee, W., and Rattanasak, U. (2009). "Comparative study on the characteristics of fly ash and bottom ash geopolymers." Waste Manag, 29(2), 539543. Davidovits, J. (1994). "Global Warming Impact on the Cement and Aggregates Industries." World Resour. Rev., 6(2), 263-278. Davidovits, J. (2013). "Geopolymer Cement." Geopolymer Cement, a review. De Silva, P., Sagoe-Crenstil, K., and Sirivivatnanon, V. (2007). "Kinetics of geopolymerization: Role of Al2O3 and SiO2." Cement and Concrete, 37(4), 512–518. DOH (1989). "DH-S205/1989 Standard of soil aggregate subbase." Department of Highways. DOH (2000). "DH-S204/2000 Standard of soil cement base." Department of Highways. DRR (2013). "DRR244-2013 Standard of soil cement base." Department of Rural Roads. Duxson, P., Provis, J. L., Lukey, G. C., and van Deventer, J. S. J. (2007). "The role of inorganic polymer technology in the development of ‘green concrete’." Cem. Concr. Res., 37(12), 15901597. Guo, X., Shi, H., Chen, L., and Dick, W. A. (2010). "Alkali-activated complex binders from class C fly ash and Ca-containing admixtures." J. Hazard. Mater., 173(1-3), 480-486. Hanjitsuwan, S., Hunpratub, S., Thongbai, P., Maensiri, S., Sata, V., and Chindaprasirt, P. (2014). "Effects of NaOH concentrations on physical and electrical properties of high calcium fly ash geopolymer paste." Cem. Concr. Compos., 45, 9-14. Horpibulsuk, S., Phetchuay, C., Chinkulkijniwat, A., and Cholaphatsorn, A. (2013). "Strength development in silty clay stabilized with calcium carbide residue and fly ash." Soils Found., 53(4), 477-486. Hoy, M., Rachan, R., Horpibulsuk, S., Arulrajah, A. and Mirzababaei, M. (2017). “Effect of wetting-drying cycles on compressive strength and microstructure of recycled asphalt pavement-fly ash geopolymer.” Construction and Building Materials, 144, 624-634 Ismail, I., Bernal, S. A., Provis, J. L., San Nicolas, R., Hamdan, S., and Van Deventer, J. S. J. (2014). "Modification of phase evolution in alkali-activated blast furnace slag by the incorporation of fly ash." Cem. Concr. Compos., 45, 125-135. Kampala, A. and Horpibulsuk, S. (2013), “Engineering properties of calcium carbide residue stabilized silty clay.” Journal of Materials in Civil Engineering, ASCE, 25(5), 632-644. Komnitsas, K., and Zaharaki, D. (2007). "Geopolymerisation: A review and prospects for the minerals industry." Miner. Eng., 20(14), 1261-1277. Kua, T.-A., Arulrajah, A., Horpibulsuk, S., Du, Y.-J., and Shen, S.-L. (2016). "Strength assessment of spent coffee grounds-geopolymer cement utilizing slag and fly ash precursors." Constr. Build. Mater., 115, 565-575. Kumar, S., Kumar, R., and Mehrotra, S. P. (2010). "Influence of granulated blast furnace slag on the reaction, structure and properties of fly ash based geopolymer." Journal of Materials Science, 45(3), 607-615. 18

Latifi, N., Christopher, M., Abd Majid, M.Z. and Horpibulsuk, S. (2016). “Strengthening montmorillonitic and kaolinitic clays using a calcium-based non-traditional additive: A microstructural level study.” Applied Clay Science, doi: 10.1016/j.clay.2016.06.004. Lee, W. K. W., and Van Deventer, J. S. J. (2002). "The effects of inorganic salt contamination on the strength and durability of geopolymers." Colloids Surf., A, 211, 115-126. McLellan, B.C., Williams, R.P., Lay, J., Van Riessen, A., Corder, G.D., 2011. "Costs and carbon emissions for geopolymer pastes in comparison to ordinary Portland cement." J. Clean. Prod. 19(9-10), 1080-1090. Misra, A., Biswas, D., and Upadhyaya, S. (2005). "Physico-mechanical behavior of selfcementing class C fly ash–clay mixtures." Fuel, 84(11), 1410-1422. Mohammadinia, A., Arulrajah, A., Sanjayan, J., Disfani, M. M., Bo, M.W., and Darmawan, S. (2016). "Stabilization of Demolition Materials for Pavement Base/Subbase Applications Using Fly Ash and Slag Geopolymers: Laboratory Investigation." J. Mater. Civ. Eng., 04016033. Palomo, A., Blanco-Varela, M. T., Granizo, M. L., Puertas, F., Vazquez, T., and Grutzeck, M. W. (1999). "Chemical stability of cementitious materials based on metakaolin." Cem. Concr. Res., 27(7), 997-100. Palomo, A., Fernández-Jiménez, A., Kovalchuk, G., Ordoñez, L. M., and Naranjo, M. C. (2007). "Opc-fly ash cementitious systems: study of gel binders produced during alkaline hydration." Journal of Materials Science, 42(9), 2958-2966. Part, W. K., Ramli, M., and Cheah, C. B. (2015). "An overview on the influence of various factors on the properties of geopolymer concrete derived from industrial by-products." Constr. Build. Mater., 77, 370-395. Phetchuay, C., Horpibulsuk, S., Suksiripattanapong, C., Chinkulkijniwat, A., Arulrajah, A., and Disfani, M. M. (2014). "Calcium carbide residue: Alkaline activator for clay–fly ash geopolymer." Constr. Build. Mater., 69, 285-294. Phetchuay, C., Horpibulsuk. S., Arulrajah, A., Suksiripattanapong, C., Udomchai, A. (2016), ”Strength development in soft marine clay stabilized by fly ash and calcium carbide residue based geopolymer.” Applied Clay Science, 127-128, 134-142. Phoo-ngernkham, T., Maegawa, A., Mishima, N., Hatanaka, S., and Chindaprasirt, P. (2015). "Effects of sodium hydroxide and sodium silicate solutions on compressive and shear bond strengths of FA–GBFS geopolymer." Constr. Build. Mater., 91, 1-8. Phummiphan, I., Horpibulsuk, S., Phoo-ngernkham, T., Arulrajah, A., and Shen, S.-L. (2016a). "Marginal Lateritic Soil Stabilized with Calcium Carbide Residue and Fly Ash Geopolymers as a Sustainable Pavement Base Material." J. Mater. Civ. Eng., 04016195. Phummiphan, I., Horpibulsuk, S., Sukmak, P., Chinkulkijniwat, A., Arulrajah, A., and Shen, S.-L. (2016b). "Stabilisation of marginal lateritic soil using high calcium fly ash-based geopolymer." Road Materials and Pavement Design, 1-15. Puertas, F., Martinez-Ramirez, S., Alonso, S., and Vazquez, T. (2000). "Alkali-activated fly ash/slag cement Strength behaviour and hydration products." Cem. Concr. Res., 30, 1625-1632. Rashad, A. M. (2013). "A comprehensive overview about the influence of different additives on the properties of alkali-activated slag – A guide for Civil Engineer." Constr. Build. Mater., 47, 29-55. Rattanasak, U., and Chindaprasirt, P. (2009). "Influence of NaOH solution on the synthesis of fly ash geopolymer." Miner. Eng., 22(12), 1073-1078. Ravikumar, D., Peethamparan, S., and Neithalath, N. (2010). "Structure and strength of NaOH activated concretes containing fly ash or GGBFS as the sole binder." Cem. Concr. Compos., 32(6), 399-410.

19

Roskos, C., Cross, D., Berry, M., and Stephens, J. ( 2011). "Identification and verification of self-cementing fly ash binders for "green" concrete." World of Coal Ash (WOCA) Conference, May 9-12, 2011Denver, CO, USA. Sakkas, K., Panias, D., Nomikos, P. P., and Sofianos, A. I. (2014). "Potassium based geopolymer for passive fire protection of concrete tunnels linings." Tunnelling Underground Space Technol., 43, 148-156. Sarker, P. K., Kelly, S., and Yao, Z. (2014). "Effect of fire exposure on cracking, spalling and residual strength of fly ash geopolymer concrete." Mater. Des., 63, 584-592. Sezer, A., İnan, G., Yılmaz, H. R., and Ramyar, K. (2006). "Utilization of a very high lime fly ash for improvement of Izmir clay." Build. Environ., 41(2), 150-155. Somna, K., Jaturapitakkul, C., Kajitvichyanukul, P., and Chindaprasirt, P. (2011). "NaOHactivated ground fly ash geopolymer cured at ambient temperature." Fuel, 90(6), 2118-2124. Sukmak, P., Horpibulsuk, S., and Shen, S.-L. (2013). "Strength development in clay–fly ash geopolymer." Constr. Build. Mater., 40, 566-574. Suksiripattanapong, C., Horpibulsuk, S., Chanprasert, P., Sukmak, P., and Arulrajah, A. (2015). "Compressive strength development in fly ash geopolymer masonry units manufactured from water treatment sludge." Constr. Build. Mater., 82, 20-30. Tantiwanit, W. and Changsuwan, S. (1995).Geology of Lateritic Soil Sources in Eastern Thailand. Report No. MR 134 Material and Research Division, Department of Highways, Thailand (in Thai). Temuujin, J., van Riessen, A., and Williams, R. (2009). "Influence of calcium compounds on the mechanical properties of fly ash geopolymer pastes." J. Hazard. Mater., 167(1-3), 82-88. Turner, L.K., and Collins, F.G. (2013) “Carbon dioxide equivalent (CO2-e) emissions: a comparison between geopolymer and OPC cement concrete”, Constr Build Mater Vol.43, pp.125–30. VicRoads (2011). Cementitious Treated Crushed Concrete for Pavement Subbase. VicRoads. Yusuf, M. O., Megat Johari, M. A., Ahmad, Z. A., and Maslehuddin, M. (2014). "Strength and microstructure of alkali-activated binary blended binder containing palm oil fuel ash and ground blast-furnace slag." Constr. Build. Mater., 52, 504-510. Zhang, M., Guo, H., El-Korchi, T., Zhang, G., and Tao, M. (2013). "Experimental feasibility study of geopolymer as the next-generation soil stabilizer." Constr. Build. Mater., 47, 14681478.

20

Figure Captions Figure 1. Particle size distribution of LS, FA and GBFS. Figure 2. SEM images of: (a) LS, (b) FA, and (c) GBFS. Figure 3. XRD pattern of (a) LS and (b) GBFS and FA. Figure 4. Compaction curves of FA-GBFS geopolymer stabilized LS. Figure 5. OLC versus NS relationships for various GBFS contents. Figure 6. The UCS of FA-GBFS geopolymer stabilized LS at various GBFS contents, cured at 7 to 60 days for different NS:NH ratios (a) 100:0, (b) 90:10, (c) 80:20 and (d) 50:50. Figure 7. SEM images of stabilized samples with 90:10 NS:NH and 10% GBFS after 7, 28 and 60 days of curing. Figure 8. SEM images of stabilized samples at 30% GBFS NS:NH ratios of 90:10 and 50:50. Figure 9. XRD results of stabilized samples at 20% GBFS and NS:NH ratios of 90:10 and 50:50 for different curing times.

21

 

 

 

  Q= Quartz (SiO2) M= Muscovite (KAl 2(Si3Al) O 10(OH,F)2) I= Illite (KAl 2Si3AlO 10(OH) 2) K= Montmorillonite (Na 0.3Al 4Si6O15(OH) 64H2O)

Intensity (Counts)

Q

LS

M

M K

I

10

Q

M

MI

20

Q

Q Q M Q

M 30

40

Q

Q

50

Q

60

70

2 theta (degree)

Intensity (Counts)

(a) Q= Quartz (SiO2) C= Calcite (CaCO 3) A= Anhydrite (CaSO 4) F= Maghemite (Fe2O3) S= Calcium Silicate (Ca 2SiO 4) G= Gehlenite (Ca 2Al 2SiO 7) L= Calcium Aluminium Oxide (CaO.Al 2O3)

GBFS

C C S G

S G

S CS G S L A Q Q

FA

10

SG

20

C

FA

30

S

L

S

C G C

C

SG

S SG

F A F

40

2 theta (degree) (b)

F

50

F

60

70

 

 

 

 

 

FA FA Etched hole Cementitious product

Reaction product

FA Cementitious product

Etched hole

 

Etched hole

Etched hole

Etched hole

Cementitious product

Etched hole

Cementitious product

Almost completely reaction on FA

Unreacted FA

Cementitious product

Etched hole

Cementitious product

Etched hole

Cementitious product

Etched hole

Cementitious product

 

 

Table Captions Table 1. Engineering properties of LS compared with subbase and engineering fill material specifications (DRR 2013 and DOH 2000). Engineering properties

Soil sample

Subbase

Engineering fill material

Liquid limit (LL) (%)

27.72

< 35

< 40

Plastic index (PI)

6.07

< 11

< 20

California Bearing Ratio

14.70

> 25

> 10

52.90

< 60

< 60

(CBR) at 95% of γd,max (%) Los Angles abrasion (percent of wear) (%)

22

Table 2. Chemical compositions of LS, FA, and GBFS. Chemical composition (%)

LS

FA

GBFS

SiO2

77.81

36.00

26.78

Al2O3

4.42

16.80

6.96

Fe2O3

10.93

17.64

18.65

CaO

1.13

26.73

30.38

MgO

N.D.

N.D.

9.70

SO3

1.36

N.D.

1.74

K2O

2.33

1.83

N.D.

TiO2

1.33

0.48

0.89

MnO2

0.55

0.15

3.56

Br2O

0.38

N.D.

N.D.

Cr2O3

N.D.

N.D.

0.85

ZnO

N.D.

N.D.

0.48

N.D. = not detected

23