Effect of waste glass addition on lightweight aggregates prepared from F-class coal fly ash

Construction and Building Materials 112 (2016) 773–782

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Effect of waste glass addition on lightweight aggregates prepared from F-class coal fly ash Yu-Ling Wei ⇑, Shao-Hsian Cheng, Guan-Wei Ko Department of Environmental Science and Engineering, Tunghai University, 40704, Taiwan

h i g h l i g h t s
F-type coal fly ash can be recycled as lightweight aggregate (LWA) via sintering.
Waste window glass addition enhances coal fly ash sintering due to its rich flux.
When sintering temperature is too high, more open pore volume will be created.

a r t i c l e

i n f o

Article history: Received 1 August 2015 Received in revised form 26 January 2016 Accepted 22 February 2016

Keywords: Coal fly ash Lightweight aggregate Waste glass Sintering Bloating

a b s t r a c t Waste window glass powders are mixed with F-type coal fly ash and fired in an attempt to prepare lightweight aggregates (LWAs) at relatively lower sintering temperatures. The mixtures are formed into cylindrical green pellets with a press operated at 3000-psi pressure. The green pellets are then fired (sintered) in an electric furnace at 1050–1300 °C for 10 min without any pre-heating step, thus simplifying the sintering process and saving energy. Results show that all fired pellets, except that made from pure coal fly ash at 1050 °C, have a particle density less than 2 g cm3 meeting the LWA standard generally accepted by construction industry. The waste glass powders promote flux content in the mixtures, reducing the temperature required to successfully sintering green pellets into LWAs. Pellets lose their weight up to 4.5 weight percent after firing. The weight loss does not go up with increased sintering temperature. Due to enhancing formation of viscous layer to capture gases, addition of glass powder to coal fly ash tends to suppress weight loss when the temperature is above 1175 °C, while such suppression phenomenon in the fired two-component pellets does not occur unless the temperature is greater than 1200 °C. When coal fly ash is sintered above 1175 °C, more big round pores surrounded by glassy layer can be observed on pellets’ microstructure morphologies. However, extremely big pores of distorted shape without distinct solid walls are developed in core fragments at 1300 °C, indicating that the viscosity of viscous layer is too low to capture gases and the fired pellets are just about to melt. After sintering the mixture at 1050 °C and 1250 °C, diopside (MgCaSiO6) and wollastonite (CaSiO3) are formed as new crystalline phases. They are beneficial to sintering reaction because of their relatively low melting points and good sealing property. Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction Lightweight aggregates (LWAs) have extensive practical application. They can be used in environmental engineering as a substrate on which bacteria are grown, in horticulture due to their ability to hold water inside their pore structure and gradually release it, and in civil engineering as construction materials. Tall building can be found almost everywhere in Taiwan nowadays owing to its extremely dense population and limited land space. The thriving activities in construction sector have caused a ⇑ Corresponding author. E-mail address: [email protected] (Y.-L. Wei). http://dx.doi.org/10.1016/j.conbuildmat.2016.02.147 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.

dramatic depletion in the natural resources serving as construction materials. To ease the situation of natural resource shortage and to minimize negative environmental impact from quarrying natural coarse aggregates, manufacturing LWAs from recycled raw materials [1] and various wastes, such as sand sludge, mining and industrial waste, for replacing fractional coarse aggregates in concrete has become an unavoidable trend [2–7]. Having a lower density, LWAs generally weigh less than threefourths of normal-weight aggregates. As a result, the use of LWAs can lessen the demand of reinforced steels and cement, thereby cutting down construction cost. Furthermore, LWAs can reduce heat conduction rate in virtue of their abundance of pores, saving energy cost in both winter and summer for the users of heaters

774

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782

and air conditioners. During a fire occasion, LWAs can also prevent a rapid temperature surge in the rooms next or close to fire origin by retarding heat transference rate. Moreover, level of sound transmission can be lowered through destructive wave interference as sound wave passes through the porous structures of LWAs. Power source in Taiwan mainly includes fossil fuels, renewable energies, and nuclear power. The development of renewable energies is very limited due to Taiwan’s isthmian territory and dense population; it merely represented 3.56% of the total power consumption in 2011. As to extracting power from nuclear source, there has been strong nationwide resistance against it because of the severe damage caused by the nuclear accidents occurred in Japan in 2011 and because of the difficulty in properly handling radioactive wastes. As a result, constructing nuclear power plants takes long drawn-out time in policy debate. In contrast, the resistance to new construction of fossil-fuel plants is much less owing to the rapid advance in pollution control technologies with the capability to considerably abate pollutant emissions. Hence, acquiring power from fossil-fuel power plant seems to be a better alternative for meeting the growing demand of power in Taiwan in near future. Currently, coal-fired power accounts for 49.47% of the total power consumption in Taiwan and the annual output of coal fly ashes is approximately two million tons. Although currently major part of the fly ashes are recycled as an additive in cement and brick manufacturing, still, about one tenth of them remains as un-recycled. The percentage of un-recycled coal fly ash is increasing because brick demand in construction sector is declining. Further, the capacity of landfill for properly treating the fly ash is very limited in Taiwan considering the fact that 23 million people are residing on a 36 thousand square-kilometer island of which approximately 70% is covered with mountains and hills. Worldwide annual coal fly ash production rate is 900 million tons and expected to increase to approximately 2 billion tons in 2020. Some developed countries even dispose of more than 50% of their coal fly ash in open dump sites. Coal fly ash disposal in open dumps is of great environmental and human health concerns such as ground water contamination by toxic metals and polycyclic aromatic hydrocarbons. Therefore recycling of coal fly ash is critical for sustainable society. The present study aims to recycle coal fly ash as LWAs that are increasingly demanded by Taiwan’s construction sector and most of the LWAs are presently imported from other countries. Nevertheless, sintering pure coal fly ash into LWAs requires high temperature, or considerable energy input. The sintering temperature can be lowered if substances rich in flux are blended with coal fly ash. Currently, waste window glass in Taiwan is suffering from poor recycling activity due to a non-rewarded policy practiced by Taiwan government. Their final destiny is to go to incinerator with municipal wastes and then mostly become bottom ashes. Because waste window glass mainly consists of silica and sodium oxide, it can be blended with coal fly ash as raw materials for preparing LWAs. Recycling of coal fly ash and waste window glass as LWAs can not only meet Taiwan government’s zero-waste goal in accordance with the principle of sustainable development, but also reduce reliance on imported LAWs whose price is going up due to worldwide growing environmental concern on quarrying natural resources as LWAs.

coarsely ground into powders with a blade-type shredder, dried at 105 °C, and passed through a 50-mesh sieve for subsequent experiments. Fig. 1 shows the appearance of both coal fly ash (magnification: 400) and waste window glass powders (magnification: 200). The glass powders are basically in triangle-like shape while coal fly ash particles are mostly spherical, and the former is much bigger in size. 2.2. Equipment A constant-temperature oven (RHD-120L, max. temperature 200 °C, Risen, USA) was used to dry the raw materials, coal fly ash and waste window glass powders. The appearance of raw material powders were observed with a microscope (BX 5 system microscope, Olymlus corp., Tokyo, Japan). To analyze chemical compositions of the raw materials, a microwave oven (MWS-2, Berghof Laborprodukte GmbH, Germany) was employed for sample digestion and an inductively coupled plasma atomic emission spectrometer (ICP-AES) (Profile plus, Teledyne Leeman Labs, New Hampshire, USA) was used to quantify the compositions. Thermal property of the raw samples was investigated with a thermogravimetric analyzer/differential scanning calorimetry (TGA/DSC) (Pryis diamond TGA/DSC, Perkin Elmer, Massachusetts, USA). Experiments carried out with a laser diffraction particle analyzer (LS230, Beckman Coulter, Washington, USA) gave particle size distribution of the raw materials. The to-be-sintered cylindrical green pellets were made by pressurizing the raw material powders with a high-pressure press (maximum shaping pressure 8500psi, Pan-Chum Corp., Taiwan). Then, the green pellets were sintered into LWAs in a sintering furnace (maximum temperature 1450 °C, Chung-Lien High Heat Industrial Co. Ltd, Taiwan), and compressive strength of the sintered pellets was measured by the use of a crushing strength machine (HCH-239-20T, Jin-ChingHer Co. Ltd., Taiwan). An X-ray diffractometer (XRD) (D8 Advantac, Bruker AXS, Germany) was employed to identify the crystalline phases present in both the raw materials and sintered pellets. The morphologies of core fragment of the sintered pellets were studied with an environmental scanning electron microscope (ESEM) (FEI Quanta 400F, Oregon, USA). 2.3. Method Fig. 2 is the experimental flowchart depicting the procedure from sample pretreatment, instrumental analysis for raw materials, firing of green pellets, to characterization of fired pellets.

2. Experimental 2.1. Materials The F-class coal fly ash was collected from a coal-fired power plant in Taiwan, oven dried at 105 °C, ground to pass a 50-mesh sieve (less than 297 lm), and stored in PE bottles for subsequent experiments. The ash powders aggregate; thus, for better mixing with waste glass powders, they were subjected to an easy grinding for 15 s. By the way, the coal fly ash is grey in color and low in lime, thus being less desired as cement additive by local cement producers. Waste window glass was

Fig. 1. Appearance of coal fly ash (400) and glass powders (magnification: 200).

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782 2.3.1. Chemical composition analysis The raw material powders were digested in two stages. Firstly, amount of 0.05 g powders were placed in Teflon digestion vessels, mixed with 7 mL concentrated HNO3 plus 3 mL concentrated HF, seated still to allow gas bubbles to disappear, and capped. The mixtures were then subjected to microwave-aid digestion isothermally at 230 °C for 30 min after the temperature has been raised from room temperature to 230 °C in 30 min. During the temperature raise period, the temperature was held isothermally for 5 min each at 140 °C and 160 °C. Secondly, after the samples were cooled to room temperature, the vessels were uncapped and an amount of 20-mL saturated H3BO3 solution was added in, and re-digested by raising the temperature from room temperature to 210 °C in 5 min and then stayed there for 15 min. The digest liquid was then cooled and analyzed with an ICP-AES instrument. Stock solution containing known amount of various metallic ions was used to prepare calibration curves for metallic ion analyses with the ICP-AES. 2.3.2. Sintering process for LWA preparation The mixture to be sintered was prepared by blending coal fly ash and glass powder at a weight ratio of 75:25. This ratio was picked based on the followings. The mixture gains more flux and SiO2 components due to the addition of waste glass powders without losing too much loss on ignition (LOI) of which most will become gases to bloat the fired pellets during sintering. Green pellets made from pure coal fly ash were also sintered for comparison with that made from the mixture (75% coal fly ah + 25% waste glass powders). For an homogeneous mixing, the mixture placed in a polyethylene (PE) bottle was subjected to an end-to-end rotation (30 ± 2 rounds per minute) for 6 h, dried at 105 °C, ground to pass a 50-mesh sieve (less than 297 lm), and then saved in PE bottles for subsequent experiments. To generate LWAs, each of pure coal fly ash and the mixture was separately pressed into cylindrical green pellets at 3000 psi in a stainless-steel mold and then directly introduced into an isothermal electric furnace and heated for 10 min that had reached a pre-set temperature between 1050 °C and 1300 °C. The sintered products were immediately retrieved from the furnace after the 10-min sintering, and were naturally cooled in atmosphere before their physical and chemical properties were investigated. The natural cooling process took 30–35 min to cool the retrieved LWAs down to room temperature. 2.3.3. Measurement of properties of fired pellets The fired pellets were investigated in terms of weight loss, particle density, water sorption rate, morphology of their core fragments, compressive strength, and crystalline phases. 2.3.3.1. Weight loss. After the green pellets were fired at 1050–1300 °C for 10 min, each sample lost some weight. It is defined as below:

weight loss ð%Þ ¼ ½ðW1  W2 Þ=W1   100

2.3.3.2. Water sorption rate. Water sorption rate of fired pellets is defined as the amount of water intake by submerging them in water for 24 h. The pellet was face-dried before its weight was measured. The water sorption rate of a fired pellet is shown below:

coal fly ash 100%

face-dry water sorption rate ð%Þ ¼ ½ðWs  Wd Þ=Wd   100 Wd: weight of fired pellet prior to soaking water. Ws: weight of fired pellet after soaking water for 24 h.

2.3.3.3. LWA particle density. The particle volume was measured based on the Archimedes principle, and the LWA particle density was determined by dividing the weight of a dry fired pellet with its volume. The formulas involved are as below:

qp ¼ Wd =Vs where

qp: particle density of fired pellet. Wd: weight of dry fired pellet. Vs: volume of dry fired pellet. The LWA apparent particle density was measured based on the Archimedes Principle [8]. To determine the size of a LWA pellet which was lighter than water, the pellet was completely pressed below water surface by using a slim metal wire with negligible volume compared with pellet volume.

Vs ¼ Vw  ðWb  Wa Þ=qw where Vw: weight of volumetric flask plus water up to the certified volume mark. Wb: weight of volumetric flask plus dry fired pellet plus water up to the certified volume mark. Wa: weight of volumetric flask plus dry fired sample. qw: water density.

2.3.3.4. LWA compressive strength. Compressive strength of the LWA particles was measured on a crushing strength machine with the to-be-tested particle being seated on the flat seat of the machine. All of the LWA particles in this study have flat bottom because the green pellets are in a shape of cylindrical disk, and during the sintering process they were placed on a flat refractory plate inside the electric furnace. To derive the LWA compressive strength, the force needed to crush the LWA particle was recorded and divided by the area of the LWA bottom.

3. Results and discussion 3.1. Chemical compositions

W1: weight of green pellet prior to firing. W2: weight of fired pellet.

coal fly ash and waste glass (drying, grinding to < 297 μm

775

compositions, particle size distribution, TGA/DTA, XRD

coal fly ash/waste glass 75/25 (by weight)

forming into green pellet with a 3000-psi press weight loss, particle density, sintering at various temperatures for 10

water sorption rate, compressive strength, XRD, and SEM

minutes

morphology for fired pellet Fig. 2. Experimental flowchart.

As indicated in Table 1, main chemical compositions of coal fly ash are SiO2 and Al2O3, and the subsidiary ones are Fe2O3 and CaO with minor components such as Cr2O3, NiO, and others. Percentages of carbon, hydrogen, sulfur, and nitrogen components in the coal fly ash in this study are 1.76%, 0.326%, 0.128%, and 0.0430%, respectively. Because coal fly ash originates from coal combustion, it is mainly composed of various inorganic oxides, with minor organic residual such as carbon [9]. Coal fly ash has great similarity to the ash component of coal fuel in compositions, except that the former contains much more calcium compounds than the latter. During coal combustion, the ash component of coal fuel is dried and cracked and some ash particles are so small in size that it can be carried away by hot flue gas stream from furnace/boiler system into flue gas duct. In the duct, they are captured by air pollutant control devices, such as electric precipitator, scrubber, and bag filters. In contrast, the other ash particles are of big size and falls on furnace grate as bottom ashes [9]. Chemical compositions of ash present in coal fuel are always similar to that present in the minerals where the coal is mined, mainly consisting of silicon oxide and alumina. Coal fly ash always has higher calcium content than its origin due to the application of Ca-based alkaline sorbent to capture sulfur oxides from flue gas stream. Further, during the combustion, very small portion of the combustible component present in coal is incompletely combusted and forms soot particles that are always extremely fine in size and light in weight, thus being readily carried into flue gas duct as part of coal fly ash [9].

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782

40

40

20

20 0

0 100 80

100

Glass (Number percentage) (Cumulative number percentage)

60

40

40

20

20

0

0

Particle size distributions of the raw materials are presented in Fig. 3. The D50 cut diameters of coal fly ash and glass powder are approximately 18 lm and 213 lm, respectively. Coal fly ash is much finer than glass powders. The coal fly ash was sampled from the air pollutant control device of a local power plant. And only fine ashes can be carried away by the flue gas stream [9]. Apparently, as indicated in Fig. 3, the waste window glass was just coarsely ground. During the experimentation period, this study tried to pulverize the glass powder further to a lesser size comparable to the coal fly ashes by prolonging the pulverizing time from 10–15 s to 2 min, but only find little effectiveness in further reducing the size and the pulverizer became over-heated. Considering energy saving and pulverizer protection, the waste window glass was only coarsely ground in present study. For the effect of particle size on sintering reaction, smaller fly ash size generally favors the occurrence of sintering reaction. Under atmospheric pressure, sintering reaction is mainly driven by the change in free energy due to surface area decrease. It’s recognized that atomic diffusion and material transfer during sintering is greatly influenced by particle size [10]. For sintering green pellets consisting of fine particles, the change in free energy becomes much greater, as compared to coarser particles, thereby reducing the sintering temperature required to form glassy layer.

22

8

4.

1-

18

4.

1. 12

23

1. 12 2-

.6

23

2

7 69

69

.7 10

.6

.0

0. -1

<3

52

1

60

.5

3.2. Particle size distribution

80

Number percentage (%)

60

3.

The waste glass used in present study mainly consists of SiO2 and Na2O, and its subsidiary compositions are CaO and MgO with other minors (Table 1). Both silica and flux contents (sum of Na2O, K2O, CaO, MgO, and Fe2O3) in the waste glass are considerably greater than that in coal fly ash. Thus, adding glass powders into coal fly ash enhances the contents of SiO2 and flux in the green pellets. Enhancement of SiO2 promotes the formation of 3dimensional silicate structure when the pellets are fired [10,11]. Further, flux content enhancement in green pellets by glass powder addition is expected to make sintering reaction to occur at lower temperatures, thus saving energy [12]. Nevertheless, the glass powder contains only negligible LOI, 0.58% (i.e., mainly moisture) that is much less than 4.0% as present in coal fly ash, implying that upon sintering less gases can be produced for bloating the two-component pellets.

80

60

2

–: Not detectable. a Analyzed following the protocol of Taiwan CNS 11393. b Analyzed with ICP-AES after microwave digestion (n = 3). c Weight loss after heating from room temperature to 1200 °C at a rate of 10 °C min1 under air flow rate of 100 mL min1.

(Number percentage) (Cumulative number percentage)

80

7-

81.09 ± 1.46 1.02 ± 0.06 12.50 ± 0.28 0.86 ± 0.01 5.29 ± 0.22 2.32 ± 0.09 0.50 ± 0.01 – 0.01 ± 0.00 – – – 0.03 ± 0.01 0.02 ± 0.01 – – – 0.58

100

.0

65.09 ± 0.67 18.80 ± 0.5. 1.17 ± 0.04 0.80 ± 0.03 3.93 ± 0.06 1.32 ± 0.04 6.18 ± 0.03 0.01 ± 0.06 0.02 ± 0.00 0.05 ± 0.00 0.02 ± 0.01 0.02 ± 0.01 0.02 ± 0.00 0.05 ± 0.01 0.09 ± 0.00 0.08 ± 0.00 0.03 ± 0.00 4.0

Coal fly ash

30

SiO2a Al2O3b Na2Ob K2Ob CaOb MgOb Fe2O3b CuOb PbOb MnO2b CoOb NiOb SnO2b SeO2b SrOb BaOb Li2Ob LOIc

100

78

Waste window glass

30

Coal fly ash

Number percentage (%)

Composition

8-

Table 1 Chemical compositions of coal fly ash and waste window glass (wt%).

8-

776

Particle size range (μm) Fig. 3. Particle size distribution of coal fly ash and glass powder.

However, fired pellets might not expand in size unless glassy layer has been developed sooner than generation of gases by pellets. Glassy layer is required to capture generated gases.

3.3. Weight loss of coal fly ash in TGA/DSC slow heating Fig. 4 shows the TGA/DSC results from coal fly ash. The sample was heated to 1200 °C from room temperature at a rising rate of 20 °C min1 in air (flow rate 100 mL min1). Although the heating-up rate in TGA is very slow compared to that experienced by the green pellets during sintering, the TGA/DSC results provide us some basic knowledge about in what temperature region sample weight is lost and whether this occurs as an exothermic or endothermic process. In Fig. 4, the weight loss of coal fly ash around 120 °C is due to loss of sorbed moisture [2,13]. Both dehydration of Mg(OH)2 and

Fig. 4. TGA/DSC results from coal fly ash.

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782

decomposition of MgCO3 occurs at 350 °C. Ca(OH)2 dehydrates at 580 °C while CaCO3 decomposes above 700 °C with the release of CO2 [14]. However, the weight loss above 700 °C can also result from carbon oxidation [14] because the coal fly ash contains some residual carbon with a grey appearance. Further, a previous study suggests that weight loss in the temperature region of 233– 522 °C can also be attributed to the release of structural water, yet this reaction was considered to be negligible in present study because it is exothermic reaction [15]. The endothermic reactions around 1000 °C shown in Fig. 4 can be attributed to lattice formation, implying crystallization of silicates according to a previous study [15]; be noted that the coal fly ash used in present study is rich in silica, 65% in weight. Total weight loss of the coal fly ash heated from room temperature to 1200 °C is approximately 4.0%.

3.4. Weight loss of green pellets due to rapid sintering Weight loss generally reflects the amount of gases emitted to the atmosphere by the pellets during sintering. Amount of these released gases is not necessarily proportional to bloating level of fired pellets because bloating level is mainly influenced by the amount of gases released after viscous layer has been developed. In other words, gases generated prior to formation of viscous layer would likely escape to the atmosphere, leaving tiny pores or tunnels in the fired pellets without effective bloating [12]. Fig. 5 shows the sintering-effected weight loss of the pellets made from pure coal fly ash and the mixtures. Two trends are observed. First, both curves have a peak; in other words, the weight loss does not always go up with increasing temperature. Addition of glass powders into coal fly ash (i.e., the dotted curve) tends to suppress weight loss when the temperature is above 1175 °C, while such suppression in the fired pure coal fly ash pellet does not occur unless the temperature is greater than 1200 °C (i.e., the solid curve). The shift of curve peaks from 1200 °C to 1175 °C by glass addition is due to the fact that glass is rich in flux component (sodium oxide) and silica that can form 3-dimensional silicate structure during sintering [10,11]. In fact, the inversion of curves implies that, above these two temperatures, viscous layer becomes abundant enough to boost the capture of generated gases [8,12,16], thereby reducing the weight loss. It is generally recognized that upon completeness of the firing process, the fired pellets start to cool down and some viscous layer gradually turns into hard external glassy layer which may considerably reduce LWA water sorption rate [8,12,16]. Formation of 3-dimentional silicate structure is considered to be responsible for the desired physical property [11].

Fig. 5. Sintering-induced weight loss of coal fly ash without/with waste glass addition at various temperatures.

777

For the second trend depicted in Fig. 5, in the higher temperature region (=1150 °C) glass powder addition results in less weight loss, as compared to pure coal fly ash. And in the 51150 °C region, the two curves are almost identical; glass addition does not seem to affect weight loss of fired pellets. In fact, the weight loss in this region should be considered as being enhanced by the glass addition, given that the component glass, accounting for 25 wt% of green pellets, contains only negligible LOI, thus making almost no contribution to the experimentally observed weight loss. In other words, the green pellet with glass additive actually contains much less overall LOI component. Thus, if the data is normalized to unit mass of coal ash, the weight loss of the two-component pellets would be higher than that made from pure coal fly ash. The implicit result observed in the region 51150 °C seems to be contradictory to a common recognition, which may not be correct, that with the addition of glass, the weight loss of coal fly ash component of green pellet should be reduced due to an easier formation of viscous layer that can retard the escape of gases during sintering. To explain the contradiction, this study proposes the follows. At the early stage of the 10-min sintering process, i.e., prior to the formation of viscous layer, the tunnels available for gas escape into atmosphere are suggested to be bigger in size for the twocomponent green pellets due to the way the mixed powders were packed. It has been shown that the glass powders are mostly in triangular shape while the coal fly ashes are spherical (Fig. 1). Further, size of the glass powders is approximately ten times coarser than coal fly ash, with their D50 cut diameters being 213 lm and 18 lm (Fig. 3), respectively. For mono-sized spherical powder packing, inter-powder porosity is independent of powder size. For instance, cubic packing gives a porosity of 42% while rhombic packing makes a porosity of 26%. However, other than powder size, powder shape and orientation in the green pellets can also affect the porosity and pore size. Given the triangular shape and the relatively larger size of the waste glass powders, we suggest that the powders in the two-component green pellets were not closely packed. In other words, the interstitial spacing among glass powders was not closely occupied by the coal fly ash powders, leading to the presence of more macro tunnels inside the two-component green pellets. Therefore, upon firing, two-component green pellets are considered to allow more gases to escape. 3.5. Particle density and particle size of fired pellets Particle density is an important physical property to judge whether sintered pellets are classified as LWAs; the borderline is 2 g cm3 [17–19]. As indicated in Fig. 6, for the fired pellets made from pure coal fly ash, their density reduces from 2.07 to 0.67 g cm3 with increasing temperature. This is due to the fact that at higher temperature, pellet surfaces are better sintered, facilitating formation of glassy layer to capture more gases and resulting in more bloating. Fig. 6 shows that the particle density of all fired pellets, except the pure coal ash pellet fired at 1050 °C, is less than 2 g cm3. The pellets sintered at 1300 °C melted with their cylindrical shape collapsed, leaving a melt being stuck to the ceramic plate because the viscosity of glassy layer at 1300 °C is too low to hold gases inside the pellets. Comparing the particle density of LWAs made from pure coal fly ash and twocomponent pellets, Fig. 6 indicates that LWA particle density can be considerably lowered by the addition of glass powders that have a density of 2.47 g cm3, while density of the coal fly ash powders are 2.06 g cm3. Explanation to this result is given as follows. First, the two-component green pellets have more flux content than pure coal fly ash pellets, 21.47% vs. 13.41%, respectively. Flux components are recognized to always make sintering reaction to occur at lower temperature. Second, with glass addition, overall silicon content in the two-component green pellets is also promoted

778

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782

Coal ash Coal fly ash/Glass=75/25

1050

1050

-3

Particle density (g/cm )

2.0

1.5

100μm

1.0

100μm

1100

1100

1150

1150

1175

1175

1200

1200

1250

1250

0.5 1050

1100

1150

1200

1250

1300

o

Temperature ( C) Fig. 6. Particle density of fired pellets made from pure coal ash with/without waste glass addition.

than that in pure coal fly ash. The Si sp3 electronic configuration forms 3-dimentional geometric silicate structure when sintered, which is considered to be apt to form viscous layer, thus more effective in retaining gases inside pellets and leading to a lower particle density [11,12]. The results of average LWA particle volume are presented in Table 2. If expressed in one standard deviation, particle volume scattering of LWA pellets formed at each sintering temperature in the 1050–1250 °C region all falls between 3.2% and 9.1% of their respective average volume. Generally, LWA particles manufactured at higher temperatures have higher volume. Table 2 implies that a rough estimate on the relevant void percentage of the pellets fired at various temperatures can be made if the 1050 °C samples are considered to be non-expansive. The relevant void percentage of the pellets can be as high as 256% and 233% for the coal fly ash LWA (at 1250 °C) and the twocomponent LWA (at 1200 °C), respectively. IN fact, these results will be shown to be consistent with SEM morphologies of core fragments of the LWAs made at various temperatures. 3.6. Microstructure and particle appearance of fired pellets Pore volume is considered to be associated with particle density of fired pellets. Generally, larger pore volume would lead to smaller particle density. Fig. 7 shows the SEM microstructure of core fragments of LWAs made by sintering pure coal fly ash and twocomponent pellets at 1050–1300 °C. For the fired pure coal fly ash pellets, their core SEM morphologies appear to be similar after sintering in 1050–1150 °C region, with their morphologies being abundant in small voids with few large voids. No apparent formation of glassy layer is observed in these core fragment morphologies. The lack of glassy layer results in an inability for the pellets to capture expansion-causing gases. When sintered at temperature =1175 °C, more big round pores surrounded by glassy layer can be observed. When fired at 1300 °C, extremely big pores of distorted shape without distinct

1300

Pure coal fly ash Table 2 Average LWA particle volume prepared at 1050–1250 °C with/without glass power addition.

Coal fly ash/waste glass 75/25

Fig. 7. SEM morphology of LWA core fragment made from pure coal fly ash and two-component pellets at various temperatures.

LWA particle volume (cm3) Temperature (°C)

Coal fly ash LWA

Two-component LWA

1050 1100 1150 1175 1200 1250

2.96 ± 0.18 3.59 ± 0.31 3.90 ± 0.23 4.57 ± 0.38 5.37 ± 0.49 7.57 ± 0.28

3.37 ± 0.14 3.90 ± 0.25 4.11 ± 0.36 5.42 ± 0.37 7.85 ± 0.26 –

glassy walls were developed in core fragments, indicating that during sintering the viscous layer formed was not viscous enough to capture gases and the pellets were almost melted. For the fired two-component pellets, when sintered at 1100 °C, their SEM morphology depicts that glassy layer begins to develop

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782

with a few notable round pores formed. As temperature increases from 1100 °C to 1250 °C, pore size grows quickly. In fact, when fired at 1250 °C, extremely big pores of distorted shape were developed in core fragments, implying that the viscous layer formed during sintering was not viscous enough to capture expansioncausing gases and the fired pellets were nearly melted. As to the expansiveness of LWAs revealed by core fragments’ SEM morphologies, the results correlate with relevant void percentage of the pellets implied in Table 2. Fig. 8 shows the side-viewed photos of the half-cut fired pellets (top row: fired pure coal fly ash; bottom: fired two-component pellets). Size of fired pellets obviously increases with increasing temperature. These results are consistent with the observation of core pore size from SEM morphologies. The two-component pellet fired at 1300 °C was melted down and not shown here. It stuck to the sample holder after cooling. Fig. 9 shows the bird eye’s view picture of the LWA particles made from green pellets of pure coal fly ash (firing at 1200 °C and 1250 °C) and two-component mixture (firing at 1175 °C and 1200 °C). Sintering is known to make objects more compacted if no expansion-causing gases are involved. Fig. 9 shows the effect of gas escape on the external glassy layer; notable pores open to atmosphere are present for the LWAs produced at higher temperature (1200 °C in top and 1250 °C in bottom of Fig. 9). Thus, the water sorption rate of these two LWAs would be expected to be greater than the lower-temperature LWAs (1175 °C in top and 1200 °C in bottom of Fig. 9). Discussion on water sorption rate is given below. 3.7. Water sorption profile versus sintering temperature Water sorption rate represents the pore volume open to atmosphere; it cannot be revealed by examining the SEM morphologies of core fragments. Fig. 10 depicts the temperature-dependent water sorption rate profiles that have minima at 1150–1175 °C and 1100–1150 °C for fired pellets made from coal fly ash and two-component mixture, respectively. For the profiles in the regions less than 1175 °C (solid curve) and 1100 °C (dotted curve), it is understandable that with increasing temperature less open pore volume on LWA’s external surface is available to water sorption due to the cohesion of surface particles although the porosity in core region increases. While in the other temperature regions, water sorption rate increases with increasing temperature, which is due to the increasing open pore volume on the external glassy layer (Fig. 9). The result is an indication of decline of viscosity of the viscous layer when sintered at higher temperature. A viscous layer with lower viscosity allows more gases to escape, leaving more open pores on the shell surface of fired pellets [17].

779

Glass addition shifts the curve minimum point to where the temperature is approximately 75 °C lower. Such result is of no surprise because glass is a good source of flux component. For the fired coal fly ash pellets, with increasing temperature their water sorption rate decreases from 20.1% (at 1050 °C) to the minimum, 5.86% (at 1150–1175 °C), and then picks up to 18.0% (at 1250 °C). Similar trend is observed for the fired two-component pellets, except that its minimum appears at a lower temperature. 3.8. Compressive strength profile versus sintering temperature Combined effect of glass addition and firing temperature on compressive strength performance of fired pellets is presented in Fig. 11. Generally speaking, from 1050 °C to 1200 °C, addition of glass tends to reduce compressive strength of fired pellets. Plus, increase in temperature always reduces the compressive strength of pellets. These results can be understood because both the glass addition and temperature rise can make the fired pellets expand more, reducing particle density. And the reduction in particle density would reduce compressive strength of fired pellets. A dramatic fall in compressive strength is observed for the pellets fired at 1200 °C, which can be explained by recalling the microstructure morphologies (Fig. 7) and side-view pictures (Fig. 8) of fired pellets. At 1200 °C, pellet bloating becomes much more apparent, thus considerably lowering the compressive strength. To derive the LWA compressive strength, the force needed to crush the LWA particle was recorded and divided by the area of the LWA bottom. Because the bottom area that nearly represents the maximum cross section area was used in the denominator to calculate compressive strength for the cone shape-like LWAs, the compressive strength of the cone-like LWAs might be slightly underestimated as compared with standard procedures. 3.9. Change in crystalline phases due to sintering After sintering, crystalline phases present in pellets components always change due to chemical reaction between different components and/or crystal structure rearrangement. Fig. 12 depicts such phase change in the fired pellets of present study. Major crystalline phases present in the coal fly ash raw material (prior to sintering), as shown in top compartment in both columns of Fig. 12, are quartz (SiO2), corundum (Al2O3), hematite (Fe2O3), wustite (FeO), and mullite (Al6Si2O13) [16,20–26]. The waste glass powder used in present study is amorphous. After sintering, the twocomponent pellets fired at 1050–1250 °C, diopside (MgCaSiO4) and wollastonite (CaSiO3) are chemically formed as new phases. However, XRD pattern height of these two phases appears to be

Fig. 8. Side-viewed pictures of half-cut LWAs made from pure coal fly ash and two-component pellets at various temperatures.

780

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782

Fig. 9. Bird eye’s view of LWA surface sintered at higher temperatures.

900

10

5 1050

1100

1150

1200

1250

1300

o

Temperature ( C)

700 600

2

15

Coal fly ash Coal fly ash/Glass=75/25

800

Compressive strength (kgf/cm )

Water absorption (%)

2

Compressive strength (kgf/cm )

Coal fly ash Coal fly ash/Glass=75/25

20

500 400 300

10

200

5

0 1200

1250

1300 o

Temperature ( C)

100 0 1050

Fig. 10. Water sorption of lightweight aggregates made from coal fly ash without/ with waste glass addition (75:25).

15

1100

1150 1175 1200

1250

1300

1350

o

Temperature ( C) Fig. 11. Compressive strength of lightweight aggregates made from coal fly ash without/with waste glass addition (75:25).

in opposite trend with temperature. Diopside pattern height slightly decreases with increasing temperature; while wollastonite phase increases with temperature. It is well known that formation of both diopside and wollastonite phases benefit sintering and

formation of viscous layer [16,17,27]. With a Mohs hardness of 5.5–6.5 and a good sealing property due to Si sp3 electronic

781

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782

15000

M Q

Coal fly ash Q - quartz H - hematite M - mullite Co - corundum K - kyanite Wu - wustite

10000 5000

Wu Co Q Wu Co H M HM Q M Q H

Q M

0 15000

Q

Wu

Q o

Coal fly ash 1050 C K M Q

10000

M

0 15000

Q

Wu Co Q Wu Co H Q HM Q M

K

Q

Wu K Q o

Coal fly ash 1175 C

10000 5000 M Q

0 15000

K M Q K

Co H HM M M

Q

Wu

K Q o

Coal fly ash 1250 C

10000 5000 M Q

0 15000

Co H HM

M Q K

M M

Q

Wu o

Coal fly ash 1300 C

10000 5000 M Q

0 10

20

Co H HM M M

M Q

30

40

Q

Wu

50

60

70

80

Intensity (a.u.)

Intensity (a.u.)

5000

4000 3000 2000 1000 0 4000 3000 2000 1000 0 4000 3000 2000 1000 0 4000 3000 2000 1000 0

Q

M Q

M

Coal fly ash Q - quartz H - hematite Wu M - mullite Co - corundum Co Q Wu H M Co Wu - wustite Q HM Q Wu Q H Q M o

C Q

M QW D

Coal fly ash/Glass=75/25 1050 C C - cristobalite W - wollastonite D - diopside HM M W

Q o

Coal fly ash/Glass=75/25 1175 C

MW C QD QW

M M WW

Q o

Coal fly ash/Glass=75/25 1250 C

M C QW QW D

10

20

o

30

M M WW

Q

40

50

60

70

80

o

2-theta ( )

2-theta ( )

Fig. 12. Crystalline phases of coal fly ash (top compartment), LWAs from fired coal fly ash (left column), and LWAs from fired two-component mixture (right column).

configuration, diopside-based glass ceramics has been applied in nuclear waste immobilization [11,17]. As to wollastonite, with a relatively low melting point, around 1540 °C, it is mostly used by tile sector. Its fluxing property can be used to improve the manufacturing of ceramics by reducing gas escape, object shrinkage, and cracking, as well as by allowing rapid firing [28]. Further, sintering reaction also transforms fractional quartz present in coal fly ash to cristobalite phase. Cristobalite is a high-temperature product of silica de-vitrification, and it is thermodynamically much more stable than silica. The formation of diopside, wollastonite, and cristobalite is considered to benefit properties of the sintered pellets in terms of long-term immobilization of toxic metals. In the coal fly ash pellets sintered at 1050–1300 °C, in contrast to the fired two-component pellets, neither diopside (MgCaSiO6) nor wollastonite (CaSiO3) was detected as new phases. This result may be due to the fact that less calcium and magnesium are contained in coal flay ash that in the two-component mixture and/or because their diffraction pattern intensities are too low to be identified. However, kyanite (Al2SiO5) forms at 1050 and 1250 °C, and disappears at 1300 °C. At temperature =1250 °C, kyanite is suggested to decompose to mullite and silica.

3.10. Potential application of the sintered pellets The formation and shape of green pellets are quite different from what has generally been previously reported by other researchers. In fact the green pellet forming pressure in present study is as high as comparable to what has been used in forming raw brick sheet for refractory brick manufacturers. Although the 3000-psi pressure is energy-consuming, the output rate of raw brick sheet can be as rapid as one piece per second. The size of the raw brick sheet can be as large as 250 mm in length, 150 mm in width, and 6–60 mm in height. The raw brick sheet can be automatically cut into green pellets in cubic shape of various piece sizes for further sintering to make LWAs. The results reported in present study may provide some necessary information for the above-

mentioned type of green pellet formation technique. The LWAs produced by the way reported in present study can be potentially used in environmental engineering as a substrate on which bacteria are grown, in horticulture due to their ability to hold water inside their pore structure and gradually release it, and in civil engineering as construction materials; thus, further study will be focused on the study of rapid and massive production of green pellets and sintered pellets, as well as the effect of using them as partial substitute for normal-weight aggregates on concrete performance. 4. Conclusions The F-type coal fly ash can be successfully recycled as LWAs by one-step rapid sintering in 10 min without the extra 500 °C preheating step previously reported by our group. Waste window glass, which is non-cash-rewarded for Taiwanese recyclers, was pulverized and added to the coal fly ash to lower sintering temperature. LWAs were made by firing at various temperatures the green pellets made from both pure coal fly ash and mixture of coal fly ash and waste glass powders (75:25). Characteristics of the sintered pellets were studied and the following conclusions are made. Glass addition to coal fly ash can effectively reduce sintering temperature by 50–100 °C required for successful preparation of LWAs, due to an enhanced formation of viscous layer that can envelope the expansion-causing gases generated inside the fired pellets. The glass addition can both save energy input and lessen high-grade refractory material requirement for pellet sintering. LWA particle density made from pure coal fly ash and mixture of coal fly ash and waste glass powders (75:25) decreases with increasing sintering temperature, with the pellets starting to melt down at 1250–1300 °C for coal fly ash and 1200–1250 °C for the mixture, respectively. Except the aggregates prepared by firing pure coal ash at 1050 °C, the particle density of all fired pellets are less than 2.0 g cm3, complying with requirement by construction sector.

782

Y.-L. Wei et al. / Construction and Building Materials 112 (2016) 773–782

Sintering at higher temperature tends to enhance LWAs’ pore volume, as supported by the SEM morphologies. However when the temperature was too high, part of the pore volume is connected to the pellet surface and becomes open to atmosphere, causing their water sorption rates to rise. When the two-component mixture was sintered at 1050 °C and 1250 °C, diopside (MgCaSiO6) and wollastonite (CaSiO3) are formed as new phases. They are beneficial to sintering reaction because of their relatively low melting points and good sealing property. Acknowledgment This work was sponsored by Taiwan Ministry of Science and Technology (NSC-99-2221-E-029-008-MY3). References [1] A. Mueller, S.N. Sokolova, V.I. Vereshagin, Characteristics of lightweight aggregate from primary and recycled raw materials, Constr. Build. Mater. 22 (2008) 703–712. [2] B. Gonzalez-Corrochano, J. Alonso-Azcarate, M. Rodas, Characterization of lightweight aggregates manufactured from washing aggregate sludge and fly ash, Resour. Conserv. Recy. 53 (2009) 571–581. [3] B. Gonzalez-Corrochano, J. Alonso-Azcarate, M. Rodas, Production of lightweight aggregates from mining and industrial wastes, J. Environ. Manage. 90 (2009) 2801–2812. [4] B. Gonzalez-Corrochano, J. Alonso-Azcarate, M. Rodas, Effect of prefiring and firing dwell times on the properties of artificial lightweight aggregates, Constr. Build. Mater. 53 (2014) 91–101. [5] Y.C. Liao, C.Y. Huang, Effect of CaO addition on lightweight aggregate produced from water reservoir sediment, Constr. Build. Mater. 25 (2011) 2997–3002. [6] S. Volland, O. Kazmina, V. Vereshchagin, M. Dushkina, Recycling of sand sludge as a resource for lightweight aggregates, Constr. Build. Mater. 52 (2014) 361– 365. [7] S. Volland, J. Brotz, Lightweight aggregates produced from sand sludge and zeolitic rocks, Constr. Build. Mater. 85 (2015) 22–29. [8] C.R. Cheeseman, G.S. Virdi, Properties and microstructure of lightweight aggregate produced from sintered sewage sludge ash, Resour. Conserv. Recy. 45 (2005) 18–30. [9] G. Neupane, R.J. Donahoe, Leachability of elements in alkaline and acidic coal fly ash samples during batch and column leaching tests, Fuel 104 (2013) 758– 770. [10] R.L.W. Popma, Sintering characteristics of nano-ceramic coatings Ph.D. Thesis, University of Groningen, The Netherlands, 2002. April 2002.

[11] J.L. Zou, G.R. Xu, G.B. Li, Ceramsite obtained from water and wastewater sludge and its characteristics affected by Fe2O3, CaO, and MgO, J. Hazard. Mater. 165 (2009) 995–1001. [12] Y.L. Wei, C.Y. Lin, K.W. Ko, H.P. Wang, Preparation of low water-sorption lightweight aggregates from harbor sediment added with waste glass, Mar. Pollut. Bull. 63 (2011) 135–140. [13] S. Bethanis, C.R. Cheeseman, C.J. Sollars, Properties and microstructure of sintered incinerator bottom ash, Ceram. Int. 28 (2002) 881–886. [14] S.C. Huang, F.C. Chang, S.L. Lo, M.Yu. Lee, C.F. Wang, J.D. Lin, Production of lightweight aggregates from mining residues, heavy metal sludge, and incinerator fly ash, J. Hazard. Mater. 144 (2007) 52–58. [15] G. Xu, J. Zou, G. Li, Ceramsite made with water and wastewater sludge and its characteristics affected by SiO2 and Al2O3, Environ. Sci. Technol. 42 (2008) 7417–7423. [16] I. Kourti, C.R. Cheeseman, Properties and microstructure of lightweight aggregate produced from lignite coal fly ash and recycled glass, Resour. Conserv. Recy. 54 (2010) 769–775. [17] C.R. Cheeseman, A. Makinde, S. Bethanis, Properties of lightweight aggregate produced by rapid sintering of incinerator bottom ash, Resour. Conserv. Recy. 43 (2005) 147–162. [18] Taiwan Bureau of Standards, Method of Test for Lightweight Particles in Aggregate, CNS 10990, August 1998. [19] Taiwan Bureau of Standards, Lightweight Aggregate for Structural Concrete, CNS 3691, August 1998. [20] M. Erol, U. Demirler, S. Kucukbayrak, A. Ersoy-Mericboyu, M.L. Ovecoglu, Characterization investigations of glass-ceramics developed from Seyitomer thermal power plant fly ash, J. Eur. Ceram. Soc. 23 (2003) 757–763. [21] M.R. Jones, A. McCarthy, A.P.P.G. Booth, Characteristics of the ultrafine component of fly ash, Fuel 85 (2006) 2250–2259. [22] C. Leiva, L.F. Vilches, J. Vale, J. Olivares, C. Fernandez-Pereira, Effect of carbonaceous matter contents on the fire resistance and mechanical properties of coal fly ash enriched mortars, Fuel 87 (2008) 2977–2982. [23] M. Nascimento, P.S.M. Soares, V.P. Souza, Adsorption of heavy metal cations using coal fly ash modified by hydrothermal method, Fuel 88 (2009) 1714– 1719. [24] L. Qi, Y. Yuan, Characteristics and the behavior in electrostatic precipitators of high-alumina coal fly ash from the Jungar power plant, Inner Mongolia, China, J. Hazard. Mater. 192 (2011) 222–225. [25] X. Wei, H. Baoping, Z. Dong, N. Ange, Physicochemical properties and heavy metals leachability of fly ash from coal-fired power plant, Int. J. Min. Sci. Technol. 22 (2012) 405–409. [26] K. Xu, T. Deng, J. Liu, W. Peng, Study on the phosphate removal from aqueous solution using modified fly ash, Fuel 89 (2010) 3668–3674. [27] A. Karamanov, M. Pelino, Sinter-crystallization in the diopside–albite system: Part II. Kinetics of crystallization and sintering, J. Eur. Ceram. Soc. 26 (13) (2005) 2519–2526. [28] A. Sunny, K.A. Lazer, K.M. Manu, K.P. Surendran, M.T. Sebastian, Effect of glass fluxing on densification and microwave dielectric properties of LiInSiO4 ceramic, J. Alloy Compd. 552 (2013) 83–87.