- Email: [email protected]
Converting hollow fly ash into admixture carrier for concrete
Construction and Building Materials 159 (2018) 431–439
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Converting hollow fly ash into admixture carrier for concrete Peiyuan Chen a, Jialai Wang b,⇑, Fengjuan Liu b, Xin Qian b, Ying Xu a, Jin Li a a b
School of Civil Engineering and Architecture, Anhui University of Science and Technology, Huainan 232001, PR China Department of Civil, Construction, and Environmental Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA
h i g h l i g h t s Perforated cenospheres were produced by chemical etching. Chemical etching removes the amorphous silicon-aluminum matrix on cenospheres. Large inner volume of cenospheres can be used to carry and release admixtures. Perforated cenospheres can survive the mixing of concrete.
a r t i c l e
i n f o
Article history: Received 21 December 2016 Received in revised form 27 October 2017 Accepted 28 October 2017
Keywords: Cenospheres Chemical etching Admixture carrier Internal curing Concrete
a b s t r a c t This study proposes a low-cost method to convert cenospheres into an admixture carrier for concrete manufacturing. Cenospheres are hollow fly ash particles generated in coal burning power plants, having an aluminosilicate shell with high strength and stiffness. The large inner volume of the cenospheres can be used to carry and release admixtures in concrete. However, directly using cenospheres as the carrier is not possible because the inner pore is not accessible to the admixture. To address this problem, chemical etching is employed to produce perforating holes through the shell. Liquid admixtures can be easily loaded into and later released from the produced perforated cenospheres (PCs). A series of characterization tests were carried out to understand the working mechanism of the chemical etching method and its effect on the properties of the PCs. It has been found that chemical etching dissolved a small amount of amorphous materials from the cenosphere shell. As a result, the cenoshpere shell was weakened, as indicated by the reduction of the bulk crushing strength of the PCs. Nevertheless, the bulk crushing strength of all produced PCs are sufficient to guarantee that they can survive the mixing and initial stress in fresh concrete. To experimentally confirm the feasibility of using the PCs as an admixture carrier for concrete, PCs were added into cement mortar as the internal water carrier, which successfully mitigated the autogeneous shrinkage in a low water-to-cement ratio concrete. Scanning electron microscopy analysis of the cement mortar confirms that PCs not only survived the mixing of concrete, but also were dispersed and bonded well to the cement mortar. This study suggests that PCs may provide a versatile tool to integrate various admixtures in concrete. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction Although well designed concrete mixture (i.e. right proportions and sound materials) along with proper construction practices experience little or none deleterious effects, admixtures are essential ingredient of concrete for many reasons. They can greatly improve the performance of concrete when sound materials are too expensive or unavailable, or proportions is not optimal. More importantly, many high performance or multifunctional concretes
⇑ Corresponding author. E-mail address: [email protected] (J. Wang). https://doi.org/10.1016/j.conbuildmat.2017.10.122 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
[1–4] cannot be produced without proper admixtures. For example, most high performance concrete with low water to binder ratio, water reducer is a must. If we want to add new function to concretes, such as thermal energy storage, or self-healing, new admixtures such as phase change materials (PCMs) or selfhealing agents (such as epoxy) will be necessary. In current practice, admixtures are added into concrete at the time of mixing. However, undesired interaction between the admixtures and the hydration of cement may exist which limits the effects of the admixtures, and sometimes, prevents applications of the admixtures in concrete. For example, water reducers, the most commonly used chemical admixtures, can have undesirable interactions with the hydration reaction of cement. These can
432
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
result in many undesirable side effects in concrete such as rapid loss of workability, excessive quickening/retardation of setting, reduced rates of strength gain, and changes in long-term behavior [5]. To minimize the undesirable interaction between the admixture and the concrete, we can isolate the admixture from the concrete through encapsulating it within a proper carrier. Two most commonly used admixture carriers for concrete are microcapsules and porous particles. Microcapsules are usually produced in-situ through physical or chemical methods, such as spray drying, coacervation, and polymerization methods [6]. The admixture can be encapsulated into the microcapsules and isolated from the surrounding concrete. For example, Choi et al. [7] synthesized a microcapsule with degradable shell using soluble acrylic resin-based compound, which can dissolve in concrete with time, releasing the admixtures (accelerator and retarder) to enhance the durability of the concrete. Microencapsulated phase change materials (PCMs) have also been extensively used in concrete to add thermal energy storage capacity to concrete [8]. However, existing microencapsulation method has yet seen successful applications in concretes because of high manufacturing cost of the microcapsules. When porous particles such as light weight aggregates (LWAs) are used as the admixture carrier, the admixture is firstly directly impregnated into the pores of the porous particles through physical absorption, and then released into the concrete as needed. A typical example of this method is the internal curing technique developed for high-performance concrete (HPC), in which saturated LWAs are used as water carriers [9–11]. Since the admixture is only absorbed by the porous inclusions and no protective layer on the surface of the inclusions is provided, this method has little control on the release of the admixture and cannot be used to seal the admixture in concrete. In addition, porous inclusions used in this method either have low strength and stiffness (LWAs) or are soft materials (super absorbent polymers). The presence of these materials in the concrete can reduce the strength and stiffness of the concrete [12]. To overcome the drawbacks in existing admixture carriers, this study develops a low-cost but robust carrier, perforated cenospheres (PCs) for admixtures so that they can be added into concrete to achieve better performance and/or new functions. Cenospheres are hollow fly ash microspheres (Fig. 1(a)) collected from fly ash waste produced during the combustion of coal in thermal power plants. They enjoy many outstanding features, such as low density [13], strong filling ability, chemical inertness and thermal resistance [14]. The size of cenospheres ranges from a few to thousands of micrometers. As shown in Fig. 1, a typical cenosphere consists of a large inner pore and a porous aluminosilicate (Fig. 1 (b)) shell. The large inner pore provides large capacity to carry admixtures. The aluminosilicate shell has high stiffness and
strength, and a thickness in a few micrometers which can provide strong protection for admixtures stored in the inner pore. However, it is not possible to directly use cenopsheres as the admixture carrier because the inner pore is not accessible to admixtures. Therefore, a method is needed to introduce admixtures into the cenospheres. As shown in Fig. 1(b), the shell of the cenosphere has a porous structure formed by gas inclusion and is covered by a glass-crystalline nanosize film. By removing this thin film, perforating holes can be produced on the shell, through which admixtures can be loaded into and released from the inner pore. This can be done by chemical etching, which can dissolve the amorphous nanosize film on the surface of the cenospheres to introduce perforating holes on the shell. The produced perforated cenospheres (PCs) are ideal candidate as the admixture carrier for concrete manufacturing. In this study, the chemical etching process is first introduced in details, through which, PCs were successfully manufactured. A series of characterization tests were then carried out to understand the working mechanism of the chemical etching method and its effect on the properties of the produced PCs. These PCs were also added into cement mortar to demonstrate that they can be used as internal curing water carrier. Internal curing is a technology to mitigate the early age cracking induced by autogeneous shrinkage, which can cause significantly reduced strength and lifetime of concrete structures. This early age cracking due to autogeneous shrinkage is especially serious for HPC because of its low water to cement ratio. These cracking problems cannot be mitigated through conventional full water curing because of HPC’s compact pore structure and very low permeability. To combat this problem, the internal curing method has been developed. In this method, curing water is continuously supplied inside concrete by a water carrier to replenish the empty pore volume that is created by self-desiccation. This will reduce autogenous shrinkage and also improve the curing of concrete at the early age. 2. Materials and methods 2.1. Materials Reagent-grade hydrochloride (HCl) was purchased from ZhongShi Chemicals, China. Ammonium fluoride (NH4F) and sodium bicarbonate (NaHCO3) were purchased from Bodi Chemicals, China. All reagents in this paper were used as received without further purification. Cenospheres were purchased from five different sources in China, and were coded as C1 to C5. XRF-1800 sequential X-ray fluorescence spectroscopy (XRF) was used to determine the chemical compositions of these cenospheres. Particle diameters of cenospheres were analyzed by ASALD7101 laser particle size analyzer (LPSA). 2.2. Manufacturing of PCs As shown in Fig. 1(b), perforating holes can be produced on the shell by removing a thin film of amorphous materials from the cenosphere surface. Admixtures can be loaded into and released from the inner pore through these perforating
Fig. 1. Microstructure of cenospheres under SEM observation: (a) a cenoshpere with impermeable shell; (b) porous shell of the cenosphere.
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439 holes. This can be done by chemical etching, which can dissolve the amorphous nanosize film on the surface of the cenospheres. The detailed manufacture process of PCs using acid etching can be summarized by Fig. 2. The first step of manufacturing PCs is to purify the as-received cenospheres, as shown in Fig. 2, in which broken cenospheres or solid fly ash particles will be removed. To this end, the as-received cenospheres were soaked in water with occasionally stirring for 24 h to remove the broken or solid particles. After that, both the broken cenospheres and the solid fly ash particles settled down to the bottom of the container because they are heavier than water. The cenospheres floating on the surface of the water must be hollow and were then collected and dried in an oven at 105 °C for 24 h to be used as the feedstock in the Step 2. To verify the effectiveness of this method to remove the broken and solid particles, some of these dries particles were soaked in water again with occasionally stirring as before. No more particles were found settled as the bottom of the container, suggesting that the proposed method can effectively remove the broken and solid particles. The percentages of the broken and solid particles were about 4–9% in the as-received cenospheres of C1 to C4. In Step 2, the etching solution was first prepared by mixing 0.6 M HCl and 0.6 M NH4F with a volume of 300 ml in a beaker. Then 40 g of purified cenospheres produced in Step 1 were carefully added into this solution. Due to their low density, all cenospheres were actually floating on the surface of the etching solution and most of them were not in contact with the etching solution. Therefore, a rotation shaker was used to mix the cenospheres with the etching solution to ensure that etching solution can reach cenospheres. After shaking for 8 min at the speed of 130 rpm, the etching process was stopped. At this moment, most cenospheres sank at the bottom of the beaker, as shown in Fig. 2. Three different products can be found in these cenospheres sunk at the bottom of the beaker: cenospheres with small perforated holes (less than 2 lm) (referred to as PCs in Fig. 2), cenospheres with large perforated holes (LPC in Fig. 2), and broken cenospheres (BC in Fig. 2) induced by the stirring during the etching process. The rest cenospheres floating on the surface consist of imperforated ones (IPC in Fig. 2) and the perforated ones with very few or small perforating holes. The cenospheres sunk at the bottom and floating on the surface of the container were collected separately, and then filtered and washed several time by water, and then dried at 105 °C for 24 h, respectively. At the fixed etching parameters (0.6 M HCl-0.6 M NH4F, 130 rpm, 8 min), the percentages of the broken particles in cenospheres C1 to C4 induced by the etching process were about 14–21%. In addition, to minimize the influence of the etching process on the environment, it’s recommended to firstly neutralize the waste etching solution (by Na2CO3) and then reuse it. In step 3, LPCs and BCs in the cenospheres collected at the bottom of the beaker in Step 2 were removed since they cannot effectively hold any admixtures. This can be done in the same way as purifying the as-received cenospheres in Step 1. The dried cenospheres collected at the bottom of the container at the end of Step 2 were soaked in water again with occasionally stirring for 4 h. Since water can flow into LPCs quickly, LPC sank at the bottom together with BCs much faster than PCs. These particles were discarded because they cannot effectively retain the admixture. The rest particles floating on the surface at the end of this step were collected as PCs. Since water entered these PCs very slowly, they can retain the admixtures and then release them slowly, which is desirable for admixture delivery.
433
Similarly, some cenospheres could be already perforated but remained floating at the end of Step 2 because no sufficient solution could enter the PCs due to low permeability of the shell. Therefore, Step 4 was designed to collect these PCs. In this step, cenospheres floating on the surface at the end of Step 2 were dried and then soaked in water again. Unlike in Step 3, vacuum was used to accelerate water to flow into the PCs. At the end of this step, the PCs sunk at the bottom of the container were collected for use as the admixture carrier. By the above steps, about 60% of the as-received cenospheres can be converted into PCs, Fig. 6(c) and (d) shows the typical SEM images of PCs. It should be pointed out the waste solution of this etching process will have some residual HF, which is hazardous. It can be easily removed adding Ca(OH)2 solution. The residual H+ in the waste solution will be neutralized by the OH-, and F- will react with Ca2+ to produce CaF precipitant. Therefore, no HF will exist in the waste solution after this treatment, and the environmental impact of the solution is similar to deicing salt. 2.3. Characterization of the as received cenospheres and the produced PCs The morphologies, mineral compositions, and surface areas of the cenospheres before and after chemical etching were measured to shed light into the working mechanism and to estimate the possible effects of the chemical etching on the produced PCs. To this end, the morphology change of cenospheres induced by the chemical etching was studied by a SEM (JSM-6700F) with an acceleration voltage of 5 kV. Samples prepared for SEM observation were dried and coated with a layer of gold particles. Mineralogical change of the cenospheres induced by the chemical etching was examined by X-ray diffraction (XRD) patterns using a TTR-III h/h rotating anode XRD diffractometer, with Cu Ka radiation at 40 kV and 200 mA at a scan speed of 8°/min between 2h of 10° and 70°. Brunauer, Emmett, and Teller (BET) surface area change of the cenospheres was analyzed for as-received cenospheres and PCs using a Micromeritics Tristar II 3020 M. 2.4. Bulk crushing strength of the cenospheres and PCs Compressive strength is a critical property for PCs to be used as an admixture carrier in concrete. Ideally, the produced PCs should have sufficient strength so that they can not only survive the mechanical mixing of the concrete during manufacturing, but also not significantly reduce the strength of the concrete. Hydraulic pressure method is commonly used to measure the strength of cenospheres [15] because it can provide most reliable result. In this method, cenospheres are loaded into a sealed container with water. Hydraulic pressure can then be applied to these cenospheres through a pump. The compressive strength of the cenospheres is determined based on the broken rates of cenospheres at specific hydraulic pressure. However, this method is not applicable to PCs because the perforating holes on the shell of PCs allow the entrance of water into the inner pore, making it impossible to apply hydraulic pressure on the shell of PCs. As an alternative, the bulk crushing strengths of cenospheres and PCs were measured using a homemade apparatus shown in Fig. 3 [20]. It consists of a guiding cylinder, a cylindrical sample container, and a punching die. The dimension of the sample container is 115 mm in diameter and 100 mm in height. Cenospheres or
Fig. 2. Steps to manufacture perforated cenospheres (PCs).
434
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
qapp ¼
M s qw ; M2 þ Ms M1
ð1Þ
where Ms is the dry weight of the cenospheres, and qw is the density of water. In the similar way, the density of the shell can be determined by adding the broken cenospheres in the container. The bulk density of the cenospheres was measured by filling cenosphere sample within a measuring cylinder. After densifying the container with mechanical vibration, the bulk density was determined by dividing the mass of the cenosphere sample by the volume of the container. For each test, three duplicate samples were tested, and the average value was reported. 2.6. Experimental verification: Carrier of internal curing water
Fig. 3. The apparatus used to measure the bulk crushing strength of cenosphere or PC particles: (a) schematic of the apparatus; (b) experimental set-up. PCs were first carefully loaded into the sample container and then densified by knocking the container wall several times with a rubber hammer. Compressive force was then applied to the sample container by driving the guiding cylinder into the cenosphere or PCs samples through the punching die, as shown in Fig. 3. The applied compressive stress at which the guiding cylinder was driven 20 mm into the sample was chosen as the bulk crushing strength of the sample particles. For each test, the average value of the crushing strengths of three duplicate samples was reported.
2.5. Measurement of densities The volume of the inner pore of the cenosphere determines the carrying capacity of the PC used as the admixture carrier. To estimate this volume, the density of the shell and the apparent density of the cenospheres were measured. Since cenospheres can float on water, an apparatus shown in Fig. 4 was used. Cenospheres were first added into the container with some water. Then a graduated tube was inserted into the container to reach the bottom of the container through a rubber plug. More water was then added into the container through this graduated tube to raise the water level to the bottom of the plug to discharge the air in the container. After that, the rubber plug was further driven into the container to a specified level so that all cenospheres were soaked in water. More water was then added to the tube to bring the water level in the tube to a height H. The corresponding mass of the whole container (including water, cenospheres, the container, and the tube) was measured as M1.Then the same procedure was repeated without adding cenospheres into the container. After water level in the graduated tube reached the same height H, the corresponding mass of the whole container was measured asM2. The apparent density of the cenospheres can be calculated as
The objective of this test is to evaluate whether the produced PCs can survive the mixing of concrete, and release the admixture into concrete properly. To this end, PCs were added into cement mortar with low water to cement ratio as the internal curing water carrier. Compared with most chemical admixtures, water has a lower viscosity. Therefore, if the PCs can retain the water and release it at a sufficient slow rate in the concrete, they should also work for those more viscous admixtures. However, it should be pointed out that if the viscosity of the admixture is too high, it may be unpractical to load it into the PCs. Grade 42.5 Ordinary Poland cement (OPC) as specified in Chinese standard GB175 was used to make cement mortar. Its chemical composition analyzed by XRF is presented in Table 1. River sand with a fineness modulus of 2.6 was chosen as the fine aggregate. PCs manufactured from cenospheres C2 were used as the water carrier. Internal curing water was first loaded into the PCs through vacuum saturation. Superplasticizer was added to achieve proper workability. Four groups of mortar samples, including one control group and three internally cured groups were made, as shown in Table 2. In this table, (w/c)ic represents the mass ratio of the internal curing water to the cement. Mixtures shown in this table were labeled as PC followed by a number standing for the replacement level of the sand by the dry PCs in volume. According to RILEM TC 196 [16], for normal cement paste with w/c less than 0.36, the amount of additional water needed for internal curing can be estimated by Eq. (2) [17] as
w c
ic
¼ 0:18
w c
ð2Þ
Since w/c = 0.34, (w/c)ic was estimated at 0.061 according to Eq. (2). To carry this amount of internal curing water, 3.7% of sand in volume should be replaced by the PC considering that the water absorption of the PC is 180%. This suggests that the internal water in group PC3.7 shown in Table 2 is sufficient to mitigate the autogenous shrinkage of the mortar sample. The measurement of the autogenous shrinkage was carried out from the time of final setting to a specified age, using the corrugated tube method according to ASTM C1698 [18]. Three duplicate specimens were prepared for each group. Fresh cement paste was encapsulated in a corrugated polyethylene tube with a length and diameter of 420 mm and 30 mm, respectively. All specimens were stored and measured at 23.0 ± 1.0 °C all the time. The length measurement was carried out on a dilatometer bench by a dilatometer with the resolution of 1 lm. Measurements were performed at intervals of once per hour in the first day, and once per day thereafter. 50 mm 100 mm cylinder specimens were cast and sealed cured and then fractured without further polishing to create undisturbed microstructure for SEM examination. The distribution of the PCs within the microstructure of the mortar, and the bonding of the PCs to the surrounding cement paste were the focuses of the SEM study.
3. Results and discussions 3.1. Properties of cenospheres
Fig. 4. Schematic of apparatus for measuring the density and apparent density of cenospheres.
The chemical compositions of the cenospheres from five different sources determined by XRF are shown in Table 3. It can be seen that they have the typical chemical compositions of low-calcium fly ash. SiO2 and Al2O3 are two major chemical compounds, accounting for more than 84% of total mass of the cenospheres. The particle size distributions of C1 to C4 are presented in Fig. 5, which shows that particle sizes of these cenospheres are quite different. The average size varying from 221 lm of C1 to 52 lm of C3, representing a large size spectrum of available cenospheres. C5 cenospheres were sieved into six different size groups to evaluate the effect of the size of the cenosphere on the mechanical properties and densities, as described later. Fig. 6(a) and (b) show SEM images of cenospheres particles of C2 and C3, respectively. Both of them have a typical spherical shape as regular fly ash particles. C2 have larger particles than C3, which is in agreement with
435
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439 Table 1 Chemical compositions of cement (XRF). Composition
CaO
SiO2
Al2O3
SO3
MgO
Fe2O3
K2O
TiO2
Na2O
P2O5
Content (wt.%)
49.7
29.6
7.77
4.7
3.6
2.4
0.9
0.4
0.4
0.1
Table 2 Mixture proportion of mortar. Mixture
PC0
PC1.9
PC3.7
PC5.6
Cement (g/kg) Sand (g/kg) Mixing water (g/kg) Additional water (g/kg) PC (g/kg) Superplasticizers(g/kg) Replacement of sand by PC (%) w/c (w/c)ic
448 400 152 0 0 3.3 0 0.34 0
448 391 152 13.5 7.5 3.3 1.9 0.34 0.03
448 383 152 26.8 14.9 3.3 3.7 0.34 0.06
448 371 152 40.3 22.4 3.3 5.6 0.34 0.09
particle size distributions shown in Fig. 5. Compared with C3, more fine powders are attached to C2 cenospheres. 3.2. Characterization of the PCs The SEM images of C2 and C3 cenospheres after chemical etching are presented in Fig. 6(c) and (d). It can be clearly seen that perforating holes have been created on the surface of the cenospheres, confirming that the proposed chemical etching method is really effective to produce PCs. Besides these perforating holes, no other visible difference can be found between the as-received and the etched cenospheres. This confirms that the chemical etching process only dissolves the thin glass-crystalline nanosize film from the cenospheres. Fig. 7(a) shows four typical products from C2 cenospheres after etching: BC, LPC, PC, and IPC as labeled in the figure. LPC have very large perforating holes, which can account for up to 22% of the diameter of the cenosphere, as shown in Fig. 7(b). These large perforating holes will allow admixtures to easily enter the inner pore of the cenosphere. However, this also means that the loaded admixtures can quickly flow out of the cenosphere. Therefore, they cannot be used as admixture carrier and should be removed from the etching products using the method shown in Fig. 2. Compared with LPCs, PCs have much smaller perforating holes. As shown in Fig. 7(d), most holes are smaller than 1 lm. Fig. 7(d) shows a piece of the shell of a broken C2 cenosphere, from which the thickness of the C2 cenosphere can be estimated as around 7.5 lm. With such small holes and much larger wall thickness, considerable capillary tensions can be generated along the perforating holes, making it possible to retain the admixture inside the cenosphere and release it slowly later into the concrete. Therefore, PCs are good candidate as the carrier of admixtures. XRD patterns of C1 to C4 cenospheres before and after chemical etching are presented in Fig. 8. It can be seen that all these patterns exhibit similar features as the regular fly ash. Two major mineral phases can be identified are mullite (Al2.3O4.85Si0.7) and quartz (SiO2). Significant amount of amorphous materials, which very
Fig. 5. Particle size distribution of cenospheres C1–C4.
likely is amorphous SiO2, exists in all tested samples as indicated by the ‘‘humps” between 15° and 35° in all cenospheres. These amorphous materials can react with calcium hydroxide in the concrete to produce more calcium silicate hydrate which can densify the microstructure of the concrete. This pozzolanic reactivity of the cenospheres provides another advantage for the PCs to be used as the admixture carrier since other existing carriers can only serve as inert fillers after releasing admixtures. Compared to the XRD patterns of the as-received cenospheres, phase compositions of the PCs were not significantly affected by the chemical etching process, as shown in Fig. 8. Only some amorphous material and very small amount of mullite were dissolved. Moreover, no new phase was formed on the PCs after etching. This is in agreement with the SEM observation shown in Fig. 6, which suggests that very small amount of shell materials was dissolved during the chemical etching process. This can be further confirmed by the XRF analysis of the chemical compositions of the etched C2. As shown in Table 4, very little change in chemical composition of C2 has been induced by the etching. Compared with the asreceived cenospheres shown in Table 3, the contents of metal oxides such as CaO, MgO, Fe2O3, and Na2O were reduced by etching, suggesting that some of them were dissolved into the etching solution. The reduction of these metal oxides induced by chemical etching leads to relatively increasing of the contents of SiO2 and Al2O3 by 2% and 3%, respectively, in the produced PCs as shown by Tables 3 and 4. Table 5 compares the BET surface area of the as-received cenospheres and the PCs. It can be seen that the BET surface area of all PCs are higher than their as-received counterparts. Clearly, this
Table 3 Chemical composition of cenospheres (wt%). Cenosphere
Sources
SiO2
Al2O3
Fe2O3
K2O
CaO
MgO
TiO2
Na2O
P2O5
C1 C2 C3 C4 C5
Hebei Anhui Henan Hebei Shandong
63.7 57.7 60 64.7 60.7
27.9 26.5 32.8 26.7 30.9
2.3 5.8 1.6 2.7 2.4
2.0 4.0 1.7 1.9 1.7
0.9 1.9 1.1 0.8 1.2
0.7 1.5 0.7 0.7 0.7
1.3 1.3 0.9 1.4 1.0
0.6 0.5 0.5 0.7 0.5
0.2 0.2 0.3 0.1 0.3
436
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
Fig. 6. Typical SEM images of cenospheres before and after chemical etching: (a) as-received C2 cenospheres, (b) as-received C3 cenospheres, (c) perforated C2 cenospheres, and (d) perforated C3 cenospheres.
Fig. 7. Typical etching products of C2 and the pore structures: (a) etched C2, (b) Large-pore cenosphere (LPC), (c) shell of PCs, and (d) Perforated cenosphere (PCs) with two closeups that ‘‘d-A” and ‘‘d-B”, where the diameters of pores were marked locally.
437
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
6000
6000
4000
M: Mullite Q: Quartz
M
3000
M
M
M M
Q
2000
M
Q
M
Q Q M M
1000
MQ
M
M M
M
Un-etched cenospheres Perforated cenospheres M: Mullite Q: Quartz
5000 X-ray Intensity (a.u.)
5000 X-ray Intensity (a.u.)
Un-etched cenospheres Perforated cenospheres
M
M 4000
M 3000
M
2000
M
M
M
Q M Q M
1000
M
Q M M
M
0
0 10
20
30
40 50 2 Theta (degree)
60
10
70
20
30
40 2 Theta (degree)
50
60
70
(c)
(a) 6000
6000
Un-etched cenospheres Un-etched cenospheres
5000
Perforated cenospheres
M
4000
M: Mullite Q: Quartz
Q
M 3000
M
MM
Q 2000
M
M
Q
MQ
1000
M
M
M
Q
M
M
M M
Perforated cenospheres
5000 X-ray Intensity (a.u.)
X-ray Intensity (a.u.)
M
M: Mullite Q: Quartz
M
4000 3000
M
M
2000
M
M
M
M
Q
Q
1000
M
Q M M M
M
Q M M
0
0 10
20
30
40 2 Theta (degree)
50
60
70
10
20
30
40 2 Theta (degree)
50
60
70
(d)
(b) Fig. 8. XRD patterns of un-etched cenospheres and PCs: (a) C1, (b) C2, (c) C3, and (d) C4.
Table 4 Chemical composition of PCs of C2 (XRF). Composition
SiO2
Al2O3
Fe2O3
K2O
CaO
MgO
TiO2
Na2O
P2O5
Content (wt%)
58.92
27.25
5.41
4.18
0.99
1.33
1.33
0.27
0.13
Table 5 BET surface area of as-received cenospheres and PCs (m2/g). Cenospheres
C1
C2
C3
C4
As-received PC
0.3170 0.5705
0.3943 0.7677
0.8999 1.4045
0.7630 0.9671
increase in surface area mainly attributes to the perforated pores on the shell of cenosphere particles. 3.3. Porosity of the cenospheres PCs have large inner volume available for loading admixtures [19]. The porosity of cenospheres indicates the inner volume available for loading admixture. It can be determined by the density and the apparent density of the PC as
£¼
Vv Vs q ¼1 ¼1 a; Vs þ Vv Vs þ Vv q
ð3Þ
where / is the porosity of the cenospheres; Vs and VV are the volumes of the shell and the inner pore, respectively; q is the density s qa is the apparent density of the of the shell and given by; q ¼ W Vs cenospheres and given by qa ¼ V sWþVs v .
Porosity of the C5 cenospheres was measured because they have a large range of size, allowing us to estimate the size dependence of the porosity. To this end, C5 was sieved into six groups based on size: 2000–3000 lm, 1000–2000 lm, 500–1000 lm, 180–380 lm, 96–180 lm, and 75–96 lm. The measured densities and apparent densities and the calculated porosities of these cenospheres samples are presented in Table 6. As anticipated, the apparent densities of all samples are lower than that of water. The porosity of these samples varies from 65% to 77%, confirming that most of the volume of the cenospheres can be used to load admixture. 3.4. Bulk crushing strength reduced by chemical etching Table 7 shows the bulk crushing strength of C1to C4 cenospheres and the corresponding PCs. It can be seen that the bulk crushing strengths of the as-received cenospheres are quite different, varying from 4.3 MPa of C1 to 10.0 MPa of C3, mainly attributed to the differences in the strength of the shell and the size of the particle. Table 7 shows that cenosphere with larger average particle size has lower bulk crushing strength. This size-dependency can be seen more clearly on the C5 cenospheres. As shown in Table 6, the bulk crushing strength of C5 cenospheres decreases quickly with the size of the particles.
438
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
Table 6 Densities and porosity of C5 cenospheres. Particle sizes (lm)
Bulk crushing strength (MPa)
Bulk density (g/cm3)
Apparent density (g/cm3)
Density (g/cm3)
Porosity (%)
2000–3000 1000–2000 500–1000 180–380 96–180 75–96
0.479 ± 0.004 0.658 ± 0.020 1.026 ± 0.011 1.207 ± 0.089 4.175 ± 0.055 11.627 ± 0.100
0.233 ± 0.008 0.264 ± 0.040 0.332 ± 0.001 0.346 ± 0.007 0.482 ± 0.003 0.564 ± 0.004
0.543 ± 0.010 0.667 ± 0.002 0.683 ± 0.008 0.712 ± 0.016 0.774 ± 0.021 0.832 ± 0.014
1.906 ± 0.080 2.854 ± 0.143 1.923 ± 0.091 2.318 ± 0.120 2.266 ± 0.146 2.373 ± 0.184
71.51 76.63 64.48 69.28 65.84 64.93
Table 7 Bulk crushing strength of cenospheres before etching and perforated cenospheres. Cenosphere
Before etching
Perforated cenospheres
Reduction (%)
C1 C2 C3 C4
4.323 ± 0.021 7.564 ± 0.484 10.021 ± 0.045 4.946 ± 0.100
3.232 ± 0.079 5.015 ± 0.057 5.253 ± 0.284 3.357 ± 0.146
25.2 33.7 47.6 32.1
PC0
PC1.9
PC3.7
PC5.6
Autogenous shrinkage (10-6)
300 200 100 0 0 -100
50
100
150 Curing time (hours)
After chemical etching, all these cenospheres exhibit 25%–50% reduction in crushing strength, as shown in Table 7. This reduction is anticipated since chemical etching produces perforating holes on the shell. These holes can introduce strong stress concentration on the shell under compression, leading to lower bulk crushing strength. In addition, this reduction of the bulk crushing strength is also correlated with the size of the cenosphere particle. C3 cenospheres with the minimum mean diameter of 52 lm suffered the highest reduction of 47.6%; while other cenospheres with larger particles size experienced much lower reduction in strength, as shown in Table 7. SEM imaging analysis shows that the pores of all PCs have similar size and structure. Severer stress concentration can be introduced by these pores to the shell of the PC with smaller size, leading to higher strength reduction. Nevertheless, Table 7 shows that the bulk crushing strength of all PCs are still higher than 3.0 MPa, which is sufficient to survive the mixing and pressure in the fresh concrete, as indicated by the SEM images of the microstructure of the cement mortar added with PCs shown in the next session. 3.5. Application verification: Internal curing
-200 -300 -400 Fig. 9. Autogenous shrinkages of four mortar mixtures.
The average autogenous shrinkage of four mixtures with and without PCs as the carrier of internal curing water is presented in Fig. 9. The autogenous shrinkage of all three internally cured mixtures were much less than that of the reference mixture PC0. The reduction in autogenous shrinkage is found in proportion to
Fig. 10. Typical SEM images of PCs incorporated in mortars at the age of 28 days: (a) PCs in mortar; (b) close-up of one intact PC; (c) Residual shell of a broken PC; (d) hydration products within one broken PC.
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
the internal curing water added. For example, the averaged autogenous shrinkages from 48 h to 168 h of the control specimen (PC0) is 215.3 106. With PCs added, these shrinkages was reduced by 50.3%, 108.3%, and 140.3% for PC1.9, PC3.7, and PC5.6, respectively. Significant autogeneous shrinkage still existed in mixture PC1.9, suggesting that internal curing water was not sufficient. When the content of PC was raised to 3.7% to carry more internal curing water in concrete, autogenous shrinkage was essentially eliminated in mixture PC3.7, as shown in Fig. 9. When more PCs were added, volume expansion occurred, as shown by the testing result of mixture PC5.6 shown in Fig. 9. This suggests that the internal curing water brought into the mixture by 5.6% PCs is more than needed. Four typical SEM images obtained from the freshly fractured surface of the internally cured mortar at the age of 28 days are presented in Fig. 10. As shown in Fig. 10(a), PCs were well dispersed within the mortar, as indicated by a few PCs marked in a trapezoid region. Besides a few broken particles, most PCs are intact, as shown in Fig. 10(b), confirming that the PCs can survive the mixing of concrete. The broken PCs identified within the mortarraptured along the fracture surface (Fig. 10(c) and (d)), suggesting that they were produced during the process to create the fracture surface for SEM analysis. Hydration products can be found in some (Fig. 10(c)) but not all broken PCs (Fig. 10(d)). This seems to be determined by the size of the perforating holes on the shell of the PCs. Comparing Fig. 10(c) and (d), it can be seen that the holes on the PC shell shown in Fig. 10(d) are much larger than those in Fig. 10(c). As a result, hydration products are much easier to penetrate inside the PCs shown in Fig. 10(d). In addition, all images shown in Fig. 10 suggest that microstructure of the cement paste surrounding the PCs is very dense, indicating good compatibility between the PCs and the cement mortar. This dense structure can be attributed to two reasons: more complete hydration of cement nearby induced by the internal curing water released from the PCs, and possible pozzolanic reaction between the active silicon-aluminum matrix on the shell of PCs and the Ca(OH)2 produced during the hydration of cement clinker minerals. 4. Conclusions This study demonstrates that PCs can be produced through a simple, low-cost chemical etching process. This process does not significantly affect the chemical and phase compositions, but leads to a 25%–50% reduction in the bulk crushing strength of the cenospheres. Experimental study has shown that PCs can be used as internal curing water carrier. SEM analysis confirms that PC can not only survive the mixing of concrete, but also be dispersed within and bond with cement paste very well. Compared with existing carriers, PCs are particularly fitting for concrete manufacturing for a few reasons: 1) High stiffness and strength: Compared with existing polymer-based microcapsules, cenospheres have much higher stiffness and strength. They can survive the mixing of the concrete and will not adversely affect the strength and modulus of the concrete. 2) Compatibility with the concrete: Cenospheres are essentially the same minerals as bulk fly ash. They are chemically reactive with the hydration products of Portland cement to produce more calcium-silicate-hydrate, which can enhance the strength and durability of the concrete.
439
3) Cost efficiency and ease of manufacturing: Cenospheres are low-cost industrial waste. Chemical etching used to produce the perforating holes on the surface of the cenospheres is easy to implement and requires minimal investment in equipment. Although PCs are only used as the carrier for internal water in this study, it is not difficult to envision other applications, which may open up new opportunities for innovative concrete products. These innovative concrete products can provide viable solutions to enhance the sustainability and resilience of civil infrastructure system. Acknowledgments The authors thank the Instruments’ Center for Physical Science, University of Science and Technology of China for their enthusiastic supports in characterization analysis. This work was supported by the National Natural Science Foundation of China (NO. 51728201 and 51374189), and National Science Foundation – United States under Grant CMMI #100580 and #1563551. References [1] M. Alkaysi, S. El-Tawil, Z. Liu, W. Hansen, Effects of silica powder and cement type on durability of ultra high performance concrete (UHPC), Cem. Concr. Compos. 66 (2016) 47–56. [2] W. Li, Z. Jiang, Z. Yang, H. Yu, Effective mechanical properties of self-healing cement matrices with microcapsules, Mater. Des. 95 (2016) 422–430. [3] A. Kanellopoulos, T.S. Qureshi, A. Al-Tabbaa, Glass encapsulated minerals for self-healing in cement based composites, Constr. Build. Mater. 98 (2015) 780– 791. [4] P. Meshgin, Y. Xi, Effect of phase-change materials on properties of concrete, ACI Mater. J. 109 (2012) 71–80. [5] S. Manu, Evaluation of superplasticizer performance in concrete. Third International Conference on Sustainable Construction Materials and Technology, 2013. http://www.claisse.info/proceedings.htm. [6] B. Boh, S. B, Microencapsulation technology and its applications in building construction materials, RMZ-Mater. Geoenviron. 55 (2008) 329-344. [7] Y.C. Choi, Y.K. Cho, K.J. Shin, S.J. Kwon, Development and application of microcapsule for cement hydration control, KSCE J. Civ. Eng. 20 (2016) 282– 292. [8] T. Khadiran, M.Z. Hussein, Z. Zainal, R. Rusli, Encapsulation techniques for organic phase change materials as thermal energy storage medium: a review, Sol. Energy Mater. Sol. Cells 143 (2015) 78–98. [9] Y. Wei, Y. Xiang, Q. Zhang, Internal curing efficiency of prewetted lwfas on concrete humidity and autogenous shrinkage development, J. Mater. Civ. Eng. 26 (2013) 947–954. [10] D.P. Bentz, P. Lura, J.W. Roberts, Mixture proportioning for internal curing, Concr. Int. 27 (2005) 35–40. [11] J. Castro, L. Keiser, M. Golias, J. Weiss, Absorption and desorption properties of fine lightweight aggregate for application to internally cured concrete mixtures, Cem. Concr. Compos. 33 (2011) 1001–1008. [12] D.P. Bentz, W.J. Weiss, Internal curing: a state-of-the-art review, NISTIR 7765 (2010) 2011. [13] R.G. Puri, A.S. Khanna, Effect of cenospheres on the char formation and fire protective performance of water-based intumescent coatings on structural steel, Prog. Org. Coat. 92 (2016) 8–15. [14] L. Ngu, H. Wu, D. Zhang, Characterization of ash cenospheres in fly ash from australian power stations, Energy Fuels 21 (2007) 3437–3445. [15] X. Fu, J. Zhai, P. Lu, L. Huang, Research on the compressive strength and its influence factors of the float-bead fly ash, Fly Ash Comprehen. Utilization 4 (2002) 27–29 (In Chinese). [16] K. Kovler, O.M. Jensen (Eds.), Internal curing of concrete–State-of-the-art report of RILEM Technical Committee 196-ICC, RILEM Report, 2007. [17] D. Cusson, T. Hoogeveen, Internal curing of high-performance concrete with pre-soaked fine lightweight aggregate for prevention of autogenous shrinkage cracking, Cem. Concr. Res. 38 (2008) 757–765. [18] ASTM C1698,Standard Test Method for Autogenous Strain of Cement Paste and Mortar, ASTM International, West Conshohocken, PA, 2014, http://dx.doi.org/ 10.1520/C1698-09R14. [19] Z.G. Shen, C.L. Li, Fly Ash Cenosphere and Its Applications, National Defense Industry Press, Beijing, 2008 (in Chinese). [20] Y. Li, D. Wu, L. Chang, Y. Shi, D. Wu, Z. Fang, A model for the bulk crushing strength of spherical catalysts, Ind. Eng. Chem. Res 38 (1999) 1911–1916.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Converting hollow fly ash into admixture carrier for concrete Peiyuan Chen a, Jialai Wang b,⇑, Fengjuan Liu b, Xin Qian b, Ying Xu a, Jin Li a a b
School of Civil Engineering and Architecture, Anhui University of Science and Technology, Huainan 232001, PR China Department of Civil, Construction, and Environmental Engineering, The University of Alabama, Tuscaloosa, AL 35487, USA
h i g h l i g h t s Perforated cenospheres were produced by chemical etching. Chemical etching removes the amorphous silicon-aluminum matrix on cenospheres. Large inner volume of cenospheres can be used to carry and release admixtures. Perforated cenospheres can survive the mixing of concrete.
a r t i c l e
i n f o
Article history: Received 21 December 2016 Received in revised form 27 October 2017 Accepted 28 October 2017
Keywords: Cenospheres Chemical etching Admixture carrier Internal curing Concrete
a b s t r a c t This study proposes a low-cost method to convert cenospheres into an admixture carrier for concrete manufacturing. Cenospheres are hollow fly ash particles generated in coal burning power plants, having an aluminosilicate shell with high strength and stiffness. The large inner volume of the cenospheres can be used to carry and release admixtures in concrete. However, directly using cenospheres as the carrier is not possible because the inner pore is not accessible to the admixture. To address this problem, chemical etching is employed to produce perforating holes through the shell. Liquid admixtures can be easily loaded into and later released from the produced perforated cenospheres (PCs). A series of characterization tests were carried out to understand the working mechanism of the chemical etching method and its effect on the properties of the PCs. It has been found that chemical etching dissolved a small amount of amorphous materials from the cenosphere shell. As a result, the cenoshpere shell was weakened, as indicated by the reduction of the bulk crushing strength of the PCs. Nevertheless, the bulk crushing strength of all produced PCs are sufficient to guarantee that they can survive the mixing and initial stress in fresh concrete. To experimentally confirm the feasibility of using the PCs as an admixture carrier for concrete, PCs were added into cement mortar as the internal water carrier, which successfully mitigated the autogeneous shrinkage in a low water-to-cement ratio concrete. Scanning electron microscopy analysis of the cement mortar confirms that PCs not only survived the mixing of concrete, but also were dispersed and bonded well to the cement mortar. This study suggests that PCs may provide a versatile tool to integrate various admixtures in concrete. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction Although well designed concrete mixture (i.e. right proportions and sound materials) along with proper construction practices experience little or none deleterious effects, admixtures are essential ingredient of concrete for many reasons. They can greatly improve the performance of concrete when sound materials are too expensive or unavailable, or proportions is not optimal. More importantly, many high performance or multifunctional concretes
⇑ Corresponding author. E-mail address: [email protected] (J. Wang). https://doi.org/10.1016/j.conbuildmat.2017.10.122 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
[1–4] cannot be produced without proper admixtures. For example, most high performance concrete with low water to binder ratio, water reducer is a must. If we want to add new function to concretes, such as thermal energy storage, or self-healing, new admixtures such as phase change materials (PCMs) or selfhealing agents (such as epoxy) will be necessary. In current practice, admixtures are added into concrete at the time of mixing. However, undesired interaction between the admixtures and the hydration of cement may exist which limits the effects of the admixtures, and sometimes, prevents applications of the admixtures in concrete. For example, water reducers, the most commonly used chemical admixtures, can have undesirable interactions with the hydration reaction of cement. These can
432
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
result in many undesirable side effects in concrete such as rapid loss of workability, excessive quickening/retardation of setting, reduced rates of strength gain, and changes in long-term behavior [5]. To minimize the undesirable interaction between the admixture and the concrete, we can isolate the admixture from the concrete through encapsulating it within a proper carrier. Two most commonly used admixture carriers for concrete are microcapsules and porous particles. Microcapsules are usually produced in-situ through physical or chemical methods, such as spray drying, coacervation, and polymerization methods [6]. The admixture can be encapsulated into the microcapsules and isolated from the surrounding concrete. For example, Choi et al. [7] synthesized a microcapsule with degradable shell using soluble acrylic resin-based compound, which can dissolve in concrete with time, releasing the admixtures (accelerator and retarder) to enhance the durability of the concrete. Microencapsulated phase change materials (PCMs) have also been extensively used in concrete to add thermal energy storage capacity to concrete [8]. However, existing microencapsulation method has yet seen successful applications in concretes because of high manufacturing cost of the microcapsules. When porous particles such as light weight aggregates (LWAs) are used as the admixture carrier, the admixture is firstly directly impregnated into the pores of the porous particles through physical absorption, and then released into the concrete as needed. A typical example of this method is the internal curing technique developed for high-performance concrete (HPC), in which saturated LWAs are used as water carriers [9–11]. Since the admixture is only absorbed by the porous inclusions and no protective layer on the surface of the inclusions is provided, this method has little control on the release of the admixture and cannot be used to seal the admixture in concrete. In addition, porous inclusions used in this method either have low strength and stiffness (LWAs) or are soft materials (super absorbent polymers). The presence of these materials in the concrete can reduce the strength and stiffness of the concrete [12]. To overcome the drawbacks in existing admixture carriers, this study develops a low-cost but robust carrier, perforated cenospheres (PCs) for admixtures so that they can be added into concrete to achieve better performance and/or new functions. Cenospheres are hollow fly ash microspheres (Fig. 1(a)) collected from fly ash waste produced during the combustion of coal in thermal power plants. They enjoy many outstanding features, such as low density [13], strong filling ability, chemical inertness and thermal resistance [14]. The size of cenospheres ranges from a few to thousands of micrometers. As shown in Fig. 1, a typical cenosphere consists of a large inner pore and a porous aluminosilicate (Fig. 1 (b)) shell. The large inner pore provides large capacity to carry admixtures. The aluminosilicate shell has high stiffness and
strength, and a thickness in a few micrometers which can provide strong protection for admixtures stored in the inner pore. However, it is not possible to directly use cenopsheres as the admixture carrier because the inner pore is not accessible to admixtures. Therefore, a method is needed to introduce admixtures into the cenospheres. As shown in Fig. 1(b), the shell of the cenosphere has a porous structure formed by gas inclusion and is covered by a glass-crystalline nanosize film. By removing this thin film, perforating holes can be produced on the shell, through which admixtures can be loaded into and released from the inner pore. This can be done by chemical etching, which can dissolve the amorphous nanosize film on the surface of the cenospheres to introduce perforating holes on the shell. The produced perforated cenospheres (PCs) are ideal candidate as the admixture carrier for concrete manufacturing. In this study, the chemical etching process is first introduced in details, through which, PCs were successfully manufactured. A series of characterization tests were then carried out to understand the working mechanism of the chemical etching method and its effect on the properties of the produced PCs. These PCs were also added into cement mortar to demonstrate that they can be used as internal curing water carrier. Internal curing is a technology to mitigate the early age cracking induced by autogeneous shrinkage, which can cause significantly reduced strength and lifetime of concrete structures. This early age cracking due to autogeneous shrinkage is especially serious for HPC because of its low water to cement ratio. These cracking problems cannot be mitigated through conventional full water curing because of HPC’s compact pore structure and very low permeability. To combat this problem, the internal curing method has been developed. In this method, curing water is continuously supplied inside concrete by a water carrier to replenish the empty pore volume that is created by self-desiccation. This will reduce autogenous shrinkage and also improve the curing of concrete at the early age. 2. Materials and methods 2.1. Materials Reagent-grade hydrochloride (HCl) was purchased from ZhongShi Chemicals, China. Ammonium fluoride (NH4F) and sodium bicarbonate (NaHCO3) were purchased from Bodi Chemicals, China. All reagents in this paper were used as received without further purification. Cenospheres were purchased from five different sources in China, and were coded as C1 to C5. XRF-1800 sequential X-ray fluorescence spectroscopy (XRF) was used to determine the chemical compositions of these cenospheres. Particle diameters of cenospheres were analyzed by ASALD7101 laser particle size analyzer (LPSA). 2.2. Manufacturing of PCs As shown in Fig. 1(b), perforating holes can be produced on the shell by removing a thin film of amorphous materials from the cenosphere surface. Admixtures can be loaded into and released from the inner pore through these perforating
Fig. 1. Microstructure of cenospheres under SEM observation: (a) a cenoshpere with impermeable shell; (b) porous shell of the cenosphere.
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439 holes. This can be done by chemical etching, which can dissolve the amorphous nanosize film on the surface of the cenospheres. The detailed manufacture process of PCs using acid etching can be summarized by Fig. 2. The first step of manufacturing PCs is to purify the as-received cenospheres, as shown in Fig. 2, in which broken cenospheres or solid fly ash particles will be removed. To this end, the as-received cenospheres were soaked in water with occasionally stirring for 24 h to remove the broken or solid particles. After that, both the broken cenospheres and the solid fly ash particles settled down to the bottom of the container because they are heavier than water. The cenospheres floating on the surface of the water must be hollow and were then collected and dried in an oven at 105 °C for 24 h to be used as the feedstock in the Step 2. To verify the effectiveness of this method to remove the broken and solid particles, some of these dries particles were soaked in water again with occasionally stirring as before. No more particles were found settled as the bottom of the container, suggesting that the proposed method can effectively remove the broken and solid particles. The percentages of the broken and solid particles were about 4–9% in the as-received cenospheres of C1 to C4. In Step 2, the etching solution was first prepared by mixing 0.6 M HCl and 0.6 M NH4F with a volume of 300 ml in a beaker. Then 40 g of purified cenospheres produced in Step 1 were carefully added into this solution. Due to their low density, all cenospheres were actually floating on the surface of the etching solution and most of them were not in contact with the etching solution. Therefore, a rotation shaker was used to mix the cenospheres with the etching solution to ensure that etching solution can reach cenospheres. After shaking for 8 min at the speed of 130 rpm, the etching process was stopped. At this moment, most cenospheres sank at the bottom of the beaker, as shown in Fig. 2. Three different products can be found in these cenospheres sunk at the bottom of the beaker: cenospheres with small perforated holes (less than 2 lm) (referred to as PCs in Fig. 2), cenospheres with large perforated holes (LPC in Fig. 2), and broken cenospheres (BC in Fig. 2) induced by the stirring during the etching process. The rest cenospheres floating on the surface consist of imperforated ones (IPC in Fig. 2) and the perforated ones with very few or small perforating holes. The cenospheres sunk at the bottom and floating on the surface of the container were collected separately, and then filtered and washed several time by water, and then dried at 105 °C for 24 h, respectively. At the fixed etching parameters (0.6 M HCl-0.6 M NH4F, 130 rpm, 8 min), the percentages of the broken particles in cenospheres C1 to C4 induced by the etching process were about 14–21%. In addition, to minimize the influence of the etching process on the environment, it’s recommended to firstly neutralize the waste etching solution (by Na2CO3) and then reuse it. In step 3, LPCs and BCs in the cenospheres collected at the bottom of the beaker in Step 2 were removed since they cannot effectively hold any admixtures. This can be done in the same way as purifying the as-received cenospheres in Step 1. The dried cenospheres collected at the bottom of the container at the end of Step 2 were soaked in water again with occasionally stirring for 4 h. Since water can flow into LPCs quickly, LPC sank at the bottom together with BCs much faster than PCs. These particles were discarded because they cannot effectively retain the admixture. The rest particles floating on the surface at the end of this step were collected as PCs. Since water entered these PCs very slowly, they can retain the admixtures and then release them slowly, which is desirable for admixture delivery.
433
Similarly, some cenospheres could be already perforated but remained floating at the end of Step 2 because no sufficient solution could enter the PCs due to low permeability of the shell. Therefore, Step 4 was designed to collect these PCs. In this step, cenospheres floating on the surface at the end of Step 2 were dried and then soaked in water again. Unlike in Step 3, vacuum was used to accelerate water to flow into the PCs. At the end of this step, the PCs sunk at the bottom of the container were collected for use as the admixture carrier. By the above steps, about 60% of the as-received cenospheres can be converted into PCs, Fig. 6(c) and (d) shows the typical SEM images of PCs. It should be pointed out the waste solution of this etching process will have some residual HF, which is hazardous. It can be easily removed adding Ca(OH)2 solution. The residual H+ in the waste solution will be neutralized by the OH-, and F- will react with Ca2+ to produce CaF precipitant. Therefore, no HF will exist in the waste solution after this treatment, and the environmental impact of the solution is similar to deicing salt. 2.3. Characterization of the as received cenospheres and the produced PCs The morphologies, mineral compositions, and surface areas of the cenospheres before and after chemical etching were measured to shed light into the working mechanism and to estimate the possible effects of the chemical etching on the produced PCs. To this end, the morphology change of cenospheres induced by the chemical etching was studied by a SEM (JSM-6700F) with an acceleration voltage of 5 kV. Samples prepared for SEM observation were dried and coated with a layer of gold particles. Mineralogical change of the cenospheres induced by the chemical etching was examined by X-ray diffraction (XRD) patterns using a TTR-III h/h rotating anode XRD diffractometer, with Cu Ka radiation at 40 kV and 200 mA at a scan speed of 8°/min between 2h of 10° and 70°. Brunauer, Emmett, and Teller (BET) surface area change of the cenospheres was analyzed for as-received cenospheres and PCs using a Micromeritics Tristar II 3020 M. 2.4. Bulk crushing strength of the cenospheres and PCs Compressive strength is a critical property for PCs to be used as an admixture carrier in concrete. Ideally, the produced PCs should have sufficient strength so that they can not only survive the mechanical mixing of the concrete during manufacturing, but also not significantly reduce the strength of the concrete. Hydraulic pressure method is commonly used to measure the strength of cenospheres [15] because it can provide most reliable result. In this method, cenospheres are loaded into a sealed container with water. Hydraulic pressure can then be applied to these cenospheres through a pump. The compressive strength of the cenospheres is determined based on the broken rates of cenospheres at specific hydraulic pressure. However, this method is not applicable to PCs because the perforating holes on the shell of PCs allow the entrance of water into the inner pore, making it impossible to apply hydraulic pressure on the shell of PCs. As an alternative, the bulk crushing strengths of cenospheres and PCs were measured using a homemade apparatus shown in Fig. 3 [20]. It consists of a guiding cylinder, a cylindrical sample container, and a punching die. The dimension of the sample container is 115 mm in diameter and 100 mm in height. Cenospheres or
Fig. 2. Steps to manufacture perforated cenospheres (PCs).
434
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
qapp ¼
M s qw ; M2 þ Ms M1
ð1Þ
where Ms is the dry weight of the cenospheres, and qw is the density of water. In the similar way, the density of the shell can be determined by adding the broken cenospheres in the container. The bulk density of the cenospheres was measured by filling cenosphere sample within a measuring cylinder. After densifying the container with mechanical vibration, the bulk density was determined by dividing the mass of the cenosphere sample by the volume of the container. For each test, three duplicate samples were tested, and the average value was reported. 2.6. Experimental verification: Carrier of internal curing water
Fig. 3. The apparatus used to measure the bulk crushing strength of cenosphere or PC particles: (a) schematic of the apparatus; (b) experimental set-up. PCs were first carefully loaded into the sample container and then densified by knocking the container wall several times with a rubber hammer. Compressive force was then applied to the sample container by driving the guiding cylinder into the cenosphere or PCs samples through the punching die, as shown in Fig. 3. The applied compressive stress at which the guiding cylinder was driven 20 mm into the sample was chosen as the bulk crushing strength of the sample particles. For each test, the average value of the crushing strengths of three duplicate samples was reported.
2.5. Measurement of densities The volume of the inner pore of the cenosphere determines the carrying capacity of the PC used as the admixture carrier. To estimate this volume, the density of the shell and the apparent density of the cenospheres were measured. Since cenospheres can float on water, an apparatus shown in Fig. 4 was used. Cenospheres were first added into the container with some water. Then a graduated tube was inserted into the container to reach the bottom of the container through a rubber plug. More water was then added into the container through this graduated tube to raise the water level to the bottom of the plug to discharge the air in the container. After that, the rubber plug was further driven into the container to a specified level so that all cenospheres were soaked in water. More water was then added to the tube to bring the water level in the tube to a height H. The corresponding mass of the whole container (including water, cenospheres, the container, and the tube) was measured as M1.Then the same procedure was repeated without adding cenospheres into the container. After water level in the graduated tube reached the same height H, the corresponding mass of the whole container was measured asM2. The apparent density of the cenospheres can be calculated as
The objective of this test is to evaluate whether the produced PCs can survive the mixing of concrete, and release the admixture into concrete properly. To this end, PCs were added into cement mortar with low water to cement ratio as the internal curing water carrier. Compared with most chemical admixtures, water has a lower viscosity. Therefore, if the PCs can retain the water and release it at a sufficient slow rate in the concrete, they should also work for those more viscous admixtures. However, it should be pointed out that if the viscosity of the admixture is too high, it may be unpractical to load it into the PCs. Grade 42.5 Ordinary Poland cement (OPC) as specified in Chinese standard GB175 was used to make cement mortar. Its chemical composition analyzed by XRF is presented in Table 1. River sand with a fineness modulus of 2.6 was chosen as the fine aggregate. PCs manufactured from cenospheres C2 were used as the water carrier. Internal curing water was first loaded into the PCs through vacuum saturation. Superplasticizer was added to achieve proper workability. Four groups of mortar samples, including one control group and three internally cured groups were made, as shown in Table 2. In this table, (w/c)ic represents the mass ratio of the internal curing water to the cement. Mixtures shown in this table were labeled as PC followed by a number standing for the replacement level of the sand by the dry PCs in volume. According to RILEM TC 196 [16], for normal cement paste with w/c less than 0.36, the amount of additional water needed for internal curing can be estimated by Eq. (2) [17] as
w c
ic
¼ 0:18
w c
ð2Þ
Since w/c = 0.34, (w/c)ic was estimated at 0.061 according to Eq. (2). To carry this amount of internal curing water, 3.7% of sand in volume should be replaced by the PC considering that the water absorption of the PC is 180%. This suggests that the internal water in group PC3.7 shown in Table 2 is sufficient to mitigate the autogenous shrinkage of the mortar sample. The measurement of the autogenous shrinkage was carried out from the time of final setting to a specified age, using the corrugated tube method according to ASTM C1698 [18]. Three duplicate specimens were prepared for each group. Fresh cement paste was encapsulated in a corrugated polyethylene tube with a length and diameter of 420 mm and 30 mm, respectively. All specimens were stored and measured at 23.0 ± 1.0 °C all the time. The length measurement was carried out on a dilatometer bench by a dilatometer with the resolution of 1 lm. Measurements were performed at intervals of once per hour in the first day, and once per day thereafter. 50 mm 100 mm cylinder specimens were cast and sealed cured and then fractured without further polishing to create undisturbed microstructure for SEM examination. The distribution of the PCs within the microstructure of the mortar, and the bonding of the PCs to the surrounding cement paste were the focuses of the SEM study.
3. Results and discussions 3.1. Properties of cenospheres
Fig. 4. Schematic of apparatus for measuring the density and apparent density of cenospheres.
The chemical compositions of the cenospheres from five different sources determined by XRF are shown in Table 3. It can be seen that they have the typical chemical compositions of low-calcium fly ash. SiO2 and Al2O3 are two major chemical compounds, accounting for more than 84% of total mass of the cenospheres. The particle size distributions of C1 to C4 are presented in Fig. 5, which shows that particle sizes of these cenospheres are quite different. The average size varying from 221 lm of C1 to 52 lm of C3, representing a large size spectrum of available cenospheres. C5 cenospheres were sieved into six different size groups to evaluate the effect of the size of the cenosphere on the mechanical properties and densities, as described later. Fig. 6(a) and (b) show SEM images of cenospheres particles of C2 and C3, respectively. Both of them have a typical spherical shape as regular fly ash particles. C2 have larger particles than C3, which is in agreement with
435
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439 Table 1 Chemical compositions of cement (XRF). Composition
CaO
SiO2
Al2O3
SO3
MgO
Fe2O3
K2O
TiO2
Na2O
P2O5
Content (wt.%)
49.7
29.6
7.77
4.7
3.6
2.4
0.9
0.4
0.4
0.1
Table 2 Mixture proportion of mortar. Mixture
PC0
PC1.9
PC3.7
PC5.6
Cement (g/kg) Sand (g/kg) Mixing water (g/kg) Additional water (g/kg) PC (g/kg) Superplasticizers(g/kg) Replacement of sand by PC (%) w/c (w/c)ic
448 400 152 0 0 3.3 0 0.34 0
448 391 152 13.5 7.5 3.3 1.9 0.34 0.03
448 383 152 26.8 14.9 3.3 3.7 0.34 0.06
448 371 152 40.3 22.4 3.3 5.6 0.34 0.09
particle size distributions shown in Fig. 5. Compared with C3, more fine powders are attached to C2 cenospheres. 3.2. Characterization of the PCs The SEM images of C2 and C3 cenospheres after chemical etching are presented in Fig. 6(c) and (d). It can be clearly seen that perforating holes have been created on the surface of the cenospheres, confirming that the proposed chemical etching method is really effective to produce PCs. Besides these perforating holes, no other visible difference can be found between the as-received and the etched cenospheres. This confirms that the chemical etching process only dissolves the thin glass-crystalline nanosize film from the cenospheres. Fig. 7(a) shows four typical products from C2 cenospheres after etching: BC, LPC, PC, and IPC as labeled in the figure. LPC have very large perforating holes, which can account for up to 22% of the diameter of the cenosphere, as shown in Fig. 7(b). These large perforating holes will allow admixtures to easily enter the inner pore of the cenosphere. However, this also means that the loaded admixtures can quickly flow out of the cenosphere. Therefore, they cannot be used as admixture carrier and should be removed from the etching products using the method shown in Fig. 2. Compared with LPCs, PCs have much smaller perforating holes. As shown in Fig. 7(d), most holes are smaller than 1 lm. Fig. 7(d) shows a piece of the shell of a broken C2 cenosphere, from which the thickness of the C2 cenosphere can be estimated as around 7.5 lm. With such small holes and much larger wall thickness, considerable capillary tensions can be generated along the perforating holes, making it possible to retain the admixture inside the cenosphere and release it slowly later into the concrete. Therefore, PCs are good candidate as the carrier of admixtures. XRD patterns of C1 to C4 cenospheres before and after chemical etching are presented in Fig. 8. It can be seen that all these patterns exhibit similar features as the regular fly ash. Two major mineral phases can be identified are mullite (Al2.3O4.85Si0.7) and quartz (SiO2). Significant amount of amorphous materials, which very
Fig. 5. Particle size distribution of cenospheres C1–C4.
likely is amorphous SiO2, exists in all tested samples as indicated by the ‘‘humps” between 15° and 35° in all cenospheres. These amorphous materials can react with calcium hydroxide in the concrete to produce more calcium silicate hydrate which can densify the microstructure of the concrete. This pozzolanic reactivity of the cenospheres provides another advantage for the PCs to be used as the admixture carrier since other existing carriers can only serve as inert fillers after releasing admixtures. Compared to the XRD patterns of the as-received cenospheres, phase compositions of the PCs were not significantly affected by the chemical etching process, as shown in Fig. 8. Only some amorphous material and very small amount of mullite were dissolved. Moreover, no new phase was formed on the PCs after etching. This is in agreement with the SEM observation shown in Fig. 6, which suggests that very small amount of shell materials was dissolved during the chemical etching process. This can be further confirmed by the XRF analysis of the chemical compositions of the etched C2. As shown in Table 4, very little change in chemical composition of C2 has been induced by the etching. Compared with the asreceived cenospheres shown in Table 3, the contents of metal oxides such as CaO, MgO, Fe2O3, and Na2O were reduced by etching, suggesting that some of them were dissolved into the etching solution. The reduction of these metal oxides induced by chemical etching leads to relatively increasing of the contents of SiO2 and Al2O3 by 2% and 3%, respectively, in the produced PCs as shown by Tables 3 and 4. Table 5 compares the BET surface area of the as-received cenospheres and the PCs. It can be seen that the BET surface area of all PCs are higher than their as-received counterparts. Clearly, this
Table 3 Chemical composition of cenospheres (wt%). Cenosphere
Sources
SiO2
Al2O3
Fe2O3
K2O
CaO
MgO
TiO2
Na2O
P2O5
C1 C2 C3 C4 C5
Hebei Anhui Henan Hebei Shandong
63.7 57.7 60 64.7 60.7
27.9 26.5 32.8 26.7 30.9
2.3 5.8 1.6 2.7 2.4
2.0 4.0 1.7 1.9 1.7
0.9 1.9 1.1 0.8 1.2
0.7 1.5 0.7 0.7 0.7
1.3 1.3 0.9 1.4 1.0
0.6 0.5 0.5 0.7 0.5
0.2 0.2 0.3 0.1 0.3
436
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
Fig. 6. Typical SEM images of cenospheres before and after chemical etching: (a) as-received C2 cenospheres, (b) as-received C3 cenospheres, (c) perforated C2 cenospheres, and (d) perforated C3 cenospheres.
Fig. 7. Typical etching products of C2 and the pore structures: (a) etched C2, (b) Large-pore cenosphere (LPC), (c) shell of PCs, and (d) Perforated cenosphere (PCs) with two closeups that ‘‘d-A” and ‘‘d-B”, where the diameters of pores were marked locally.
437
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
6000
6000
4000
M: Mullite Q: Quartz
M
3000
M
M
M M
Q
2000
M
Q
M
Q Q M M
1000
MQ
M
M M
M
Un-etched cenospheres Perforated cenospheres M: Mullite Q: Quartz
5000 X-ray Intensity (a.u.)
5000 X-ray Intensity (a.u.)
Un-etched cenospheres Perforated cenospheres
M
M 4000
M 3000
M
2000
M
M
M
Q M Q M
1000
M
Q M M
M
0
0 10
20
30
40 50 2 Theta (degree)
60
10
70
20
30
40 2 Theta (degree)
50
60
70
(c)
(a) 6000
6000
Un-etched cenospheres Un-etched cenospheres
5000
Perforated cenospheres
M
4000
M: Mullite Q: Quartz
Q
M 3000
M
MM
Q 2000
M
M
Q
MQ
1000
M
M
M
Q
M
M
M M
Perforated cenospheres
5000 X-ray Intensity (a.u.)
X-ray Intensity (a.u.)
M
M: Mullite Q: Quartz
M
4000 3000
M
M
2000
M
M
M
M
Q
Q
1000
M
Q M M M
M
Q M M
0
0 10
20
30
40 2 Theta (degree)
50
60
70
10
20
30
40 2 Theta (degree)
50
60
70
(d)
(b) Fig. 8. XRD patterns of un-etched cenospheres and PCs: (a) C1, (b) C2, (c) C3, and (d) C4.
Table 4 Chemical composition of PCs of C2 (XRF). Composition
SiO2
Al2O3
Fe2O3
K2O
CaO
MgO
TiO2
Na2O
P2O5
Content (wt%)
58.92
27.25
5.41
4.18
0.99
1.33
1.33
0.27
0.13
Table 5 BET surface area of as-received cenospheres and PCs (m2/g). Cenospheres
C1
C2
C3
C4
As-received PC
0.3170 0.5705
0.3943 0.7677
0.8999 1.4045
0.7630 0.9671
increase in surface area mainly attributes to the perforated pores on the shell of cenosphere particles. 3.3. Porosity of the cenospheres PCs have large inner volume available for loading admixtures [19]. The porosity of cenospheres indicates the inner volume available for loading admixture. It can be determined by the density and the apparent density of the PC as
£¼
Vv Vs q ¼1 ¼1 a; Vs þ Vv Vs þ Vv q
ð3Þ
where / is the porosity of the cenospheres; Vs and VV are the volumes of the shell and the inner pore, respectively; q is the density s qa is the apparent density of the of the shell and given by; q ¼ W Vs cenospheres and given by qa ¼ V sWþVs v .
Porosity of the C5 cenospheres was measured because they have a large range of size, allowing us to estimate the size dependence of the porosity. To this end, C5 was sieved into six groups based on size: 2000–3000 lm, 1000–2000 lm, 500–1000 lm, 180–380 lm, 96–180 lm, and 75–96 lm. The measured densities and apparent densities and the calculated porosities of these cenospheres samples are presented in Table 6. As anticipated, the apparent densities of all samples are lower than that of water. The porosity of these samples varies from 65% to 77%, confirming that most of the volume of the cenospheres can be used to load admixture. 3.4. Bulk crushing strength reduced by chemical etching Table 7 shows the bulk crushing strength of C1to C4 cenospheres and the corresponding PCs. It can be seen that the bulk crushing strengths of the as-received cenospheres are quite different, varying from 4.3 MPa of C1 to 10.0 MPa of C3, mainly attributed to the differences in the strength of the shell and the size of the particle. Table 7 shows that cenosphere with larger average particle size has lower bulk crushing strength. This size-dependency can be seen more clearly on the C5 cenospheres. As shown in Table 6, the bulk crushing strength of C5 cenospheres decreases quickly with the size of the particles.
438
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
Table 6 Densities and porosity of C5 cenospheres. Particle sizes (lm)
Bulk crushing strength (MPa)
Bulk density (g/cm3)
Apparent density (g/cm3)
Density (g/cm3)
Porosity (%)
2000–3000 1000–2000 500–1000 180–380 96–180 75–96
0.479 ± 0.004 0.658 ± 0.020 1.026 ± 0.011 1.207 ± 0.089 4.175 ± 0.055 11.627 ± 0.100
0.233 ± 0.008 0.264 ± 0.040 0.332 ± 0.001 0.346 ± 0.007 0.482 ± 0.003 0.564 ± 0.004
0.543 ± 0.010 0.667 ± 0.002 0.683 ± 0.008 0.712 ± 0.016 0.774 ± 0.021 0.832 ± 0.014
1.906 ± 0.080 2.854 ± 0.143 1.923 ± 0.091 2.318 ± 0.120 2.266 ± 0.146 2.373 ± 0.184
71.51 76.63 64.48 69.28 65.84 64.93
Table 7 Bulk crushing strength of cenospheres before etching and perforated cenospheres. Cenosphere
Before etching
Perforated cenospheres
Reduction (%)
C1 C2 C3 C4
4.323 ± 0.021 7.564 ± 0.484 10.021 ± 0.045 4.946 ± 0.100
3.232 ± 0.079 5.015 ± 0.057 5.253 ± 0.284 3.357 ± 0.146
25.2 33.7 47.6 32.1
PC0
PC1.9
PC3.7
PC5.6
Autogenous shrinkage (10-6)
300 200 100 0 0 -100
50
100
150 Curing time (hours)
After chemical etching, all these cenospheres exhibit 25%–50% reduction in crushing strength, as shown in Table 7. This reduction is anticipated since chemical etching produces perforating holes on the shell. These holes can introduce strong stress concentration on the shell under compression, leading to lower bulk crushing strength. In addition, this reduction of the bulk crushing strength is also correlated with the size of the cenosphere particle. C3 cenospheres with the minimum mean diameter of 52 lm suffered the highest reduction of 47.6%; while other cenospheres with larger particles size experienced much lower reduction in strength, as shown in Table 7. SEM imaging analysis shows that the pores of all PCs have similar size and structure. Severer stress concentration can be introduced by these pores to the shell of the PC with smaller size, leading to higher strength reduction. Nevertheless, Table 7 shows that the bulk crushing strength of all PCs are still higher than 3.0 MPa, which is sufficient to survive the mixing and pressure in the fresh concrete, as indicated by the SEM images of the microstructure of the cement mortar added with PCs shown in the next session. 3.5. Application verification: Internal curing
-200 -300 -400 Fig. 9. Autogenous shrinkages of four mortar mixtures.
The average autogenous shrinkage of four mixtures with and without PCs as the carrier of internal curing water is presented in Fig. 9. The autogenous shrinkage of all three internally cured mixtures were much less than that of the reference mixture PC0. The reduction in autogenous shrinkage is found in proportion to
Fig. 10. Typical SEM images of PCs incorporated in mortars at the age of 28 days: (a) PCs in mortar; (b) close-up of one intact PC; (c) Residual shell of a broken PC; (d) hydration products within one broken PC.
P. Chen et al. / Construction and Building Materials 159 (2018) 431–439
the internal curing water added. For example, the averaged autogenous shrinkages from 48 h to 168 h of the control specimen (PC0) is 215.3 106. With PCs added, these shrinkages was reduced by 50.3%, 108.3%, and 140.3% for PC1.9, PC3.7, and PC5.6, respectively. Significant autogeneous shrinkage still existed in mixture PC1.9, suggesting that internal curing water was not sufficient. When the content of PC was raised to 3.7% to carry more internal curing water in concrete, autogenous shrinkage was essentially eliminated in mixture PC3.7, as shown in Fig. 9. When more PCs were added, volume expansion occurred, as shown by the testing result of mixture PC5.6 shown in Fig. 9. This suggests that the internal curing water brought into the mixture by 5.6% PCs is more than needed. Four typical SEM images obtained from the freshly fractured surface of the internally cured mortar at the age of 28 days are presented in Fig. 10. As shown in Fig. 10(a), PCs were well dispersed within the mortar, as indicated by a few PCs marked in a trapezoid region. Besides a few broken particles, most PCs are intact, as shown in Fig. 10(b), confirming that the PCs can survive the mixing of concrete. The broken PCs identified within the mortarraptured along the fracture surface (Fig. 10(c) and (d)), suggesting that they were produced during the process to create the fracture surface for SEM analysis. Hydration products can be found in some (Fig. 10(c)) but not all broken PCs (Fig. 10(d)). This seems to be determined by the size of the perforating holes on the shell of the PCs. Comparing Fig. 10(c) and (d), it can be seen that the holes on the PC shell shown in Fig. 10(d) are much larger than those in Fig. 10(c). As a result, hydration products are much easier to penetrate inside the PCs shown in Fig. 10(d). In addition, all images shown in Fig. 10 suggest that microstructure of the cement paste surrounding the PCs is very dense, indicating good compatibility between the PCs and the cement mortar. This dense structure can be attributed to two reasons: more complete hydration of cement nearby induced by the internal curing water released from the PCs, and possible pozzolanic reaction between the active silicon-aluminum matrix on the shell of PCs and the Ca(OH)2 produced during the hydration of cement clinker minerals. 4. Conclusions This study demonstrates that PCs can be produced through a simple, low-cost chemical etching process. This process does not significantly affect the chemical and phase compositions, but leads to a 25%–50% reduction in the bulk crushing strength of the cenospheres. Experimental study has shown that PCs can be used as internal curing water carrier. SEM analysis confirms that PC can not only survive the mixing of concrete, but also be dispersed within and bond with cement paste very well. Compared with existing carriers, PCs are particularly fitting for concrete manufacturing for a few reasons: 1) High stiffness and strength: Compared with existing polymer-based microcapsules, cenospheres have much higher stiffness and strength. They can survive the mixing of the concrete and will not adversely affect the strength and modulus of the concrete. 2) Compatibility with the concrete: Cenospheres are essentially the same minerals as bulk fly ash. They are chemically reactive with the hydration products of Portland cement to produce more calcium-silicate-hydrate, which can enhance the strength and durability of the concrete.
439
3) Cost efficiency and ease of manufacturing: Cenospheres are low-cost industrial waste. Chemical etching used to produce the perforating holes on the surface of the cenospheres is easy to implement and requires minimal investment in equipment. Although PCs are only used as the carrier for internal water in this study, it is not difficult to envision other applications, which may open up new opportunities for innovative concrete products. These innovative concrete products can provide viable solutions to enhance the sustainability and resilience of civil infrastructure system. Acknowledgments The authors thank the Instruments’ Center for Physical Science, University of Science and Technology of China for their enthusiastic supports in characterization analysis. This work was supported by the National Natural Science Foundation of China (NO. 51728201 and 51374189), and National Science Foundation – United States under Grant CMMI #100580 and #1563551. References [1] M. Alkaysi, S. El-Tawil, Z. Liu, W. Hansen, Effects of silica powder and cement type on durability of ultra high performance concrete (UHPC), Cem. Concr. Compos. 66 (2016) 47–56. [2] W. Li, Z. Jiang, Z. Yang, H. Yu, Effective mechanical properties of self-healing cement matrices with microcapsules, Mater. Des. 95 (2016) 422–430. [3] A. Kanellopoulos, T.S. Qureshi, A. Al-Tabbaa, Glass encapsulated minerals for self-healing in cement based composites, Constr. Build. Mater. 98 (2015) 780– 791. [4] P. Meshgin, Y. Xi, Effect of phase-change materials on properties of concrete, ACI Mater. J. 109 (2012) 71–80. [5] S. Manu, Evaluation of superplasticizer performance in concrete. Third International Conference on Sustainable Construction Materials and Technology, 2013. http://www.claisse.info/proceedings.htm. [6] B. Boh, S. B, Microencapsulation technology and its applications in building construction materials, RMZ-Mater. Geoenviron. 55 (2008) 329-344. [7] Y.C. Choi, Y.K. Cho, K.J. Shin, S.J. Kwon, Development and application of microcapsule for cement hydration control, KSCE J. Civ. Eng. 20 (2016) 282– 292. [8] T. Khadiran, M.Z. Hussein, Z. Zainal, R. Rusli, Encapsulation techniques for organic phase change materials as thermal energy storage medium: a review, Sol. Energy Mater. Sol. Cells 143 (2015) 78–98. [9] Y. Wei, Y. Xiang, Q. Zhang, Internal curing efficiency of prewetted lwfas on concrete humidity and autogenous shrinkage development, J. Mater. Civ. Eng. 26 (2013) 947–954. [10] D.P. Bentz, P. Lura, J.W. Roberts, Mixture proportioning for internal curing, Concr. Int. 27 (2005) 35–40. [11] J. Castro, L. Keiser, M. Golias, J. Weiss, Absorption and desorption properties of fine lightweight aggregate for application to internally cured concrete mixtures, Cem. Concr. Compos. 33 (2011) 1001–1008. [12] D.P. Bentz, W.J. Weiss, Internal curing: a state-of-the-art review, NISTIR 7765 (2010) 2011. [13] R.G. Puri, A.S. Khanna, Effect of cenospheres on the char formation and fire protective performance of water-based intumescent coatings on structural steel, Prog. Org. Coat. 92 (2016) 8–15. [14] L. Ngu, H. Wu, D. Zhang, Characterization of ash cenospheres in fly ash from australian power stations, Energy Fuels 21 (2007) 3437–3445. [15] X. Fu, J. Zhai, P. Lu, L. Huang, Research on the compressive strength and its influence factors of the float-bead fly ash, Fly Ash Comprehen. Utilization 4 (2002) 27–29 (In Chinese). [16] K. Kovler, O.M. Jensen (Eds.), Internal curing of concrete–State-of-the-art report of RILEM Technical Committee 196-ICC, RILEM Report, 2007. [17] D. Cusson, T. Hoogeveen, Internal curing of high-performance concrete with pre-soaked fine lightweight aggregate for prevention of autogenous shrinkage cracking, Cem. Concr. Res. 38 (2008) 757–765. [18] ASTM C1698,Standard Test Method for Autogenous Strain of Cement Paste and Mortar, ASTM International, West Conshohocken, PA, 2014, http://dx.doi.org/ 10.1520/C1698-09R14. [19] Z.G. Shen, C.L. Li, Fly Ash Cenosphere and Its Applications, National Defense Industry Press, Beijing, 2008 (in Chinese). [20] Y. Li, D. Wu, L. Chang, Y. Shi, D. Wu, Z. Fang, A model for the bulk crushing strength of spherical catalysts, Ind. Eng. Chem. Res 38 (1999) 1911–1916.