- Email: [email protected]
Performance of cementitious materials produced by incorporating surface treated multiwall carbon nanotubes and silica fume
Construction and Building Materials 114 (2016) 934–945
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Performance of cementitious materials produced by incorporating surface treated multiwall carbon nanotubes and silica fume Adil Tamimi a, Noha M. Hassan b, Kazi Fattah a,⇑, Amirhooman Talachi a a b
Department of Civil Engineering, American University of Sharjah, United Arab Emirates Industrial Engineering Department, American University of Sharjah, United Arab Emirates
h i g h l i g h t s Use of functionalized CNTs improve compressive and flexural strength of cement mortar. Using CNT-COOH with silica fumes increased compressive strength by 50%. Addition of silica fume enables better dispersion of CNTs in mortar samples. SEM results revealed that the CNTs were able to fill the voids in the mortar.
a r t i c l e
i n f o
Article history: Received 3 February 2016 Received in revised form 28 March 2016 Accepted 30 March 2016 Available online 9 April 2016 Keywords: Carbon nano tube CNT Silica fumes Dispersion Surface functionalization SEM Cement composite Mortar
a b s t r a c t The outstanding mechanical properties of carbon nanotubes (CNTs) highlight them as potential candidates for cementitious material reinforcement. However, their low surface friction and the Van der Waals forces of attraction between them, cause the CNTs to aggregate with each other rather than bind with the cement matrix. A number of methods have been investigated by researchers to reduce the aggregation, improve dispersion and activate the graphite surface to enhance its interfacial interaction. These methods involve surface functionalization and coating, optimal physical blending, use of surfactant and other admixtures. This research investigates the use of silica fumes (an admixture), surface functionalized CNTs and cement paste to overcome those obstacles. CNTs with polar impurities end groups OH and COOH were examined. Mortar samples with non-functionalized CNTs dispersed in water solution, another with non-dispersed, non-functionalized CNTs, and a third batch with no CNTs (as control) was used also studied. Silica fumes volume fraction was varied from 0 to 30% to determine its effect. Compressive and flexural strengths of the different mixes were measured and compared. Qualitative analysis using Scanning Electron Microscope (SEM) and Energy-Dispersive Spectroscopy (EDS) were carried out to study the morphology of each mix. Results reveal a much higher enhancement in strength both compressive and flexural strengths for the functionalized CNTs with 30% silica fumes over the other samples. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction The use of nanomaterials has grown significantly since the first time it was discovered in 1959 [1]. Similar to other industries, the construction industry can benefit from nanotechnology in order to produce a more desirable outcome. One of the important aspects of the construction industry, which can take advantage of nanotechnology, is concrete production. The properties and characteristics of carbon nanotubes (CNT) provide great prospects to improve ⇑ Corresponding author. E-mail addresses: [email protected] (A. Tamimi), [email protected] (N.M. Hassan), [email protected] (K. Fattah), [email protected] (A. Talachi). http://dx.doi.org/10.1016/j.conbuildmat.2016.03.216 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
the properties of concrete when the CNTs are incorporated to the concrete mixture. CNTs can work as nano filler in concrete and has the potential to improve the strength and durability of concrete. The sizes of CNT are much smaller compared with the typical components in cement, and thus can improve mechanical properties at the nano scale [2]. The outstanding mechanical properties of carbon nanotubes (CNTs) highlight them as potential candidates for concrete reinforcement as well. The strength of the CNTs is directly related to the strong C@C bond and the relatively small number of defects present in the tubes. It is said to possess ‘‘a hundred times the strength of steel at one sixth of the weight” [3]. Young modulus is estimated to vary between 1 and 5 TPa, density is around 2000 kg/m3, elongation to failure of 20–30%, which
935
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
correlate to a tensile strength of well above 100 GPa. The CNTs are characterized by thermal stability up to 2800 °C [4]. Previous studies have shown that the compressive strength of concrete can increase with the addition of CNTs [5]. Despite the obvious advantages of adding CNTs in concrete, there are two major problems associated with its use. CNTs have a tendency to agglomerate due to the Van der Waals attraction force between CNT crystalline ropes when it is produced. The triangular network formed by these ropes tends to aggregate and reduce the dispersion of the CNTs in the water that is added to concrete. The second major drawback associated with CNTs is their very low surface friction, making it difficult for them to bind together or with the cement matrix material. The bonding between cement and CNT is often weak and only the hydration products can anchor a few CNTs. However, knowledge in this field is still limited [6]. Different techniques, such as use of silica fume [7,8], sonification [9] and surfactant addition [10] have been tried to disperse CNT in the mortar mix. Each of these methods has its own limitations and drawbacks. Sonication alone can break the CNT tubes resulting in a weaker reinforcement. Sonication and surfactant use can increase the air volume in the mix resulting in a voided structure. Additionally, although hydration product such as cal cium–silicate–hydrates (C–S–H), calcium hydroxide (CH) have same or larger dimensions than CNTs, only few of them are anchored in hydration products. Consequently, they don’t provide effective reinforcement of cement composite. A study by [11] investigated how the length of CNT affects the property of the cement mortar. Sonication and surfactants were used to disperse
the CNTs. Flexural test results show that sonication does not have an effect on surfactant performance since the flexural strength for all the samples were the same. Effective flexural strength can be achieved with high percentage of shorter CNTs, or less percentage of longer CNTs. A study by [7] showed that incorporating a small amount of silica fume to the mix can improve dispersion of CNT and improve its compressive strength. Because of very small particle size of silica fume (in range of 10–500 nm), which is close to the size of CNTs, silica fume can be mixed with agglomerated CNTs and effectively disperse them. Moreover, silica fume particles intermixed with CNTs gets hydrated by Ca2+ ions from the cement due to their high pozzolanic activity. Hydration products of silica fume effectively anchor CNTs, and interfacial interaction between CNTs and the hydration products was thereby enhanced. A study by [12] observed that silica fume and silica functional groups improved the fracture performance of mixtures containing carbon nanotubes and carbon fibers, but further optimization of dosage, size, and interface strength is required to fully utilize carbon nanotubes in cementitious composites [12]. Improvements in flexural strength and fracture toughness are more significant at a later age in mixtures containing silica fume, although higher volumes of silica may lead to an adverse effect by concentrating the dispersed CNT in the silica fume fields [12,13]. Li et al. [14] investigated the mechanical properties of cement composite that contained functionalized CNTs. Results showed that surface treated CNTs exhibited better compressive and flexural strength due to the interfacial interaction with hydration
Fig. 1. Schematic of functionalized CNTs used in the study (a) CNT-OH, (b) CNT-COOH and (c) SEM of MWNTs [17].
Table 1 Mix proportions of CNTs-silica fumes mortars. Set
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 a b c d e f
#
Mix name
CNT type
Silica fumes%
CNT (% wt of cement)
Cement (kg/m3)
Silica (kg/m3)
CNTs (kg/m3)
Sand before water correction (kg/m3)
Water before water correction (kg/m3)
w/ca
s/cb
No CNT 0 CNT-0 CNT-OH-0 CNT-COOH-0 CNT-H2O-0 No CNT-15 CNT-15 CNT-OH-15 CNT-COOH-15 CNT-H2O-15 No CNT-30 CNT-30 CNT-OH-30 CNT-COOH-30 CNT-H2O-30
– CNTc CNT-OHd CNT-COOHe CNT-H2Of – CNT CNT-OH CNT-COOH CNT-H2O – CNT CNT-OH CNT-COOH CNT-H2O
0 0 0 0 0 15% 15% 15% 15% 15% 30% 30% 30% 30% 30%
0.00 0.15 0.15 0.15 0.15 0.00 0.15 0.15 0.15 0.15 0.00 0.15 0.15 0.15 0.15
467.0 466.3 466.3 466.3 466.3 397.0 396.4 396.4 396.4 396.4 327.0 326.5 326.5 326.5 326.5
0 0 0 0 0 70 70 70 70 70 140 140 140 140 140
0.0 0.7 0.7 0.7 0.7 0.0 0.6 0.6 0.6 0.6 0.0 0.5 0.5 0.5 0.5
607 606 606 606 606 516 515 515 515 515 425 424 424 424 424
140 140 140 140 140 140 140 140 140 140 140 140 140 140 140
0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3
1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3
w/c: water to cement ratio. s/c: sand to cement ratio. CNT: Industrial grade MWCNTs. CNT-OH: Industrial grade MWCNTs functionalized with hydroxyl group. CNT-COOH: Industrial grade MWCNTs functionalized with carboxyl group. CNT-H2O: MWCNT dispersed in water.
936
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Fig. 2. Sample preparation equipment.
Fig. 3. (a) Compressive test for cubic sample and (b) flexural test for beam sample.
Fig. 4. (a) Gold etching process and (b) SEM analysis instruments.
product such as C-S-H and surface functionalized CNTs, resulting in a strong connection between CNT and matrix composite, reduction of pores and voids in cement matrix leading to more compact and dense concrete and bridging the connection of CNTs between
cracks and voids. The investigators found that the interfacial interaction with hydration products such as Calcium Silicate Hydrate (C-S-H) and surface functionalized CNT lead to a strong bond between CNT and the matrix composites. Despite the obvious
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
937
Fig. 5. Effect of increasing silica fume in the mix on the compressive strength of mortar samples.
Fig. 6. Ratio of compressive strength of 30% silica fumes mixes compared to No CNT-30 mix (control).
Fig. 7. Flexural strength of samples with 30% silica fume.
and well-documented results on the benefits of nanomaterials in concrete production [2,15,16], the commercial applications for concrete production are being prevented due to the high cost of nanoparticles. If efficient and effective use of nanomaterials in construction can be reached, this might initiate a need for mass production of nanomaterials, thereby decreasing its initial cost, which is currently a limiting factor. This research studied the effect of incorporating silica fumes with functionalized CNTs in the cement mortar to enhance its mechanical properties. The reaction of the polar impurities end groups (OH and COOH) with silica fumes and cement matrix were investigated through SEM and EDS analysis. The effect of this incorporation on the mortar’s mechanical properties was also examined. 2. Materials and methods 2.1. Experimental program 2.1.1. Multiwall carbon nanotubes (MWCNTs) Four industrial grade multiwall carbon nanotubes were used in this study. Regular CNTs, CNTs functionalized with hydroxyl group (CNT-OH) (Fig. 1a), CNTs functionalized with carboxyl groups (CNT-COOH) (Fig. 1b), and CNT that is suspended in water (CNT-H2O) were studied. Aromatic modified polyethylene glycol ether was used as solvent for the dispersion procedure. All CNT grades had 88 + % purity, an outside diameter that varies from 20 to 40 nm, an inner diameter of 5–10 nm and a length varying between 10 and 30 lm as shown in Fig. 1c. The CNT-OH contained 1.55–1.71 wt% –OH. The CNT-COOH contained 1.36– 1.50 wt% –COOH. The water dispersed CNTs had a dispersant content of 1.4–1.6% and are stable for up to 6 months at room temperature. All CNTs were purchased from [17].
Fig. 8. Beam after flexural test showing major failure location.
2.1.2. Mortar mix The mix prepared and tested in this study was composed of normal Portland cement, silica fume, crushed sand and CNTs. Normal Portland cement having a surface area of 355 m2/kg on the Baline test was used. The silica fume was produced in conformance with the ASTM C 1240 specifications with a specific gravity 2.1–2.4, a densified bulk density of 608–720 kg/m3, specific surface of 17 m2/kg and particle size of approximately 0.4 lm. Crushed sands, with particle sizes in the range of 0.15–4.75 mm and specific gravity of 2.65 were used as fine aggregate. The crushed sand had water absorption of 5%, which was tested based on the ASTM test C 128-88.
938
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Fig. 9. Stress Strain diagram of (a) No CNT-30 mix, (b) CNT-COOH-30 mix.
Fig. 10. SEM image of pure CNT without any additive form a clump (a) 1000 magnification, (b) 2000 magnification, (c) 5000 magnification, (d) 10,000 magnification.
2.1.3. Specimen preparation Mortar samples were prepared with three different percentages of silica in the mix and four different types of CNTs for each percentage. Table 1 summarizes the list of 15 sets of mixtures studied. The ratio of CNT/cement in the mix was kept constant to effectively determine the effect of silica on the dispersion of each type of CNT. The amount of CNT in all mixes was kept at 0.15% of the cement weight since CNTs absorb a lot of water and reduce the workability of the mix [7]. Water/cement ratio in the mix was limited to 0.3 to increase the collision of silica molecules with CNT clumps and reduce agglomeration. Water absorption of the aggregate was tested according to ASTM C128-01. The cement-to-sand ratio was kept at 1.3. Three
different percentages (0%, 15% and 30%) of silica fume by weight of cement in the mix were chosen to examine the effect of silica fumes on different types of CNTs. Previous research [7] revealed a nonlinear trend between the compressive strength of CNT reinforced cementitious material and the addition of silica fumes. The choice for the wide range of percentage of silica (0–30%) investigated in the present study was to be able to capture this nonlinearity. A standard Hobart mixer was used for the mixing process (Fig. 2). For the mixing process, all the dry material (cement, sand, silica fume, and CNT) was added to the mixer and mixed for 4 min. After that, water was slowly added to the mix and the mixer was run for another 5 min to make sure all materials were completely mixed with water. Fresh mix was added
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
939
Fig. 11. SEM image of mortar at (a) 5000 magnification of No CNT-0 mix, (b) 5000 magnification of CNT-0 mix, (c) 200 magnification of No CNT-30 mix, (d) 5000 magnification of CNT-15 mix.
to the mold in three layers and each layer was compacted with compacting tools and a shaker table. After one day the mortar was demolded and cured for 14 days. Each set of experiment was repeated three times.
analysis of the mortar mix was performed to determine materials present near the CNTs. The EDS analysis was conducted on 3 samples with 0.15% normal CNT and different ratios (0%, 15% and 30%) of silica fume by weight of cement in the mix.
2.2. Material characterization tests Compressive and flexural tests were conducted on each sample and average of the three samples was taken as final strength for that mix. 2.2.1. Compressive strength The compressive strength of the mortars was tested on 45 cubic samples (15 sets 3 replicates) each of size 50 50 50 mm. The tests were carried out using Instron Universal Testing Machine as shown in Fig. 3a. 2.2.2. Flexural strength Flexural tests were conducted on mortar samples each having sizes 210 50 50 mm. For the flexural test, a four-point loading according to ASTM C651-15 was conducted for the beam with 20 cm span. Fig. 3b illustrates the actual testing conditions for the flexural test. 2.2.3. Morphology and composition characterization The crushed mortar was examined using a scanning electron microscopic to study its surface morphology and microstructure. Microscopic image of sample was made through the interaction of the focused beam of electrons with sample surface atoms revealing information about the microstructure of cement composites. Sample preparation was performed by gold etching for 75 s. The mortar sample was loaded into a high vacuum chamber while a very thin layer of Au, Pd was deposited on the surface (Fig. 4a). A scanning electron microscope (SEM) with energy dispersive spectroscopy (EDS) [18] was used to determine the detailed structure of the nano particles that were used in the study (Fig. 4b). The EDS
3. Results and discussion 3.1. Compressive strength Fig. 5 illustrates the trend of adding silica fumes to the mix. Adding CNT to the mix without the addition of silica slightly increases the compressive strength of all samples. At 0% silica fumes, the CNT-H2O-0 mix has the highest compressive strength as it is initially suspended and dispersed in water. However, the improvement in compressive strength is low at around 11%. Adding 15% silica fume increases the compressive strength of the samples. This increase is more observable for the samples with dispersed CNTs (CNT-H2O-15) and functionalized CNTs with hydroxyl group (CNT-OH-15) whose compressive strength increased by 13.8% and 10.3%, respectively when compared to the sample without CNT (No CNT-15 mix). Sample with CNT-OH-15 and CNT-COOH15 increased to even higher strength values compared to CNT-15 sample. This can be related to the strong bond between hydration products and surface treated CNTs. The interfacial interaction of the hydration product such as C-S-H and surface functionalized CNT (CNT-OH-15 and CNT-COOH-15) is much stronger compared
940
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Fig. 12. SEM image of mixes without any silica fume (a) 5000 magnification of CNT-0 mix, (b) 5000 magnification of CNT-OH-0 mix, (c) 5000 magnification of CNT-COOH0 mix, (d) 5000 magnification of CNT-H2O-0 mix.
to normal CNTs (CNT-15). By increasing the silica fume in the mix to 30%, the compressive strengths for the samples with CNTs (CNT-30) and CNT-H2O (CNT-H2O-30) are reduced, revealing the nonlinear trend that was exhibited by previous researchers using CNTs and silica fumes alone [7]. However, surface functionalized CNTs (CNT-COOH-30 and CNT-OH-30 samples) did not exhibit this decline. A much higher strength was achieved by both samples compared to the NO CNT-30 sample with a 20% and a 16% increase in strength, respectively. Fig. 6 shows the ratio of compressive strength of each of the samples with 30% silica fumes to that of the control (without CNTs – No CNT-30 mix). Samples CNT-30 and dispersed CNT-H2O-30 decreased in strength and only the functionalized samples (CNTOH-30 and CNT-COOH-30) increased in strength. Adding silica fume more than 15% along with CNTs (CNT-30 or CNT-H2O-30) reduced the workability of the mix and increased the voids in the mortar. Additionally, CNT agglomerations were dispersed densely in silica fume fields and re-agglomerated as clumps with smaller sizes, covering cement and hindering its hydration. However, the compressive strength of the sample with surface functionalized CNT (CNT-OH-30 and CNT-COOH-30) increased as much as 20% compared to the control sample (No CNT-30 mix) even with higher percentage of silica fumes (30%). This could be attributed to the development of strong bond between the surface functionalized CNTs and hydration product.
3.2. Flexural strength To ensure consistency in results between compressive and flexural strength, flexural tests were run for 30% silica fumes. Fig. 7 reveals that the samples with CNT-OH-30 and CNT-COOH-30 have the highest flexural strength with strengths of 6.44 MPa and 6.58 MPa, respectively. For all the samples, a crack developed inside the middle area, as shown in Fig. 8. The reason for this is that CNTs with a functionalized surface develop further bridging compared to non-functionalized CNTs leading to better mortar reinforcement. The flexural strength of No CNT-30 mix was equal to 4.37 MPa, the lowest among all the specimens. The flexural strength of the sample with CNT (CNT-30) and CNT-H2O-30 improved by 30%, compared to the No CNT-30 sample. This improvement in flexural strength is much higher for surface treated samples. The flexural strength of CNT-COOH-30 and CNT-OH-30 increased by 50% compared to the control (No CNT30). The interaction between end groups and calcium silicate hydrate (C–S–H) or Ca(OH)2 leads to a strong covalent bond on the interface between the reinforcement and matrix in the composites, bridging micro cracks, and increasing the ultimate displacement. Fig. 9 shows the stress-strain curve of the No CNT-30 mix (control) and CNT-COOH-30 mix illustrating an increase in both flexural strength and ultimate displacement and strain at failure. It can be concluded that the mechanical properties of the
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
941
Fig. 13. SEM image of mixes with 15% silica fume (a) 5000 magnification of CNT-15 mix, (b) 5000 magnification of CNT-OH-15 mix, (c) 5000 magnification of CNT-COOH15 mix, (d) 5000 magnification of CNT-H2O-15 mix .
composites are improved significantly with addition of 30% silica fumes and surface functionalized CNTs. 3.3. Microstructure analysis using Scanning Electron Microscopy (SEM) To reinforce the conclusions deduced from the mechanical tests, an examination of the microstructure was performed. Fig. 10 shows the SEM image of a clump of CNT without any additive. The images were taken with different magnifications (1000, 2000, 5000 and 10,000) and the strings of CNTs clumping together can be clearly visualized. Strong self-attraction between each individual CNT caused by strong van der Waals forces leads to agglomeration of CNTs. To be able to identify CNTs in the mortar sample, it was important to first examine the microstructure without silica fumes and CNTs (No CNT-0) mix (Fig. 11a), and compare it to the mix without silica fumes but with CNTs (CNT-0) mix (Fig. 11b). The main elements of the cement paste are Ca(OH)2, C–S–H gel, pores, cracks, un-hydrated cement particles, fine aggregates [19]. In Fig. 11a, the micro size crystal growth can be seen clearly. The crystals grow in a straight line similar to the branches of a tree. The SEM analysis of the sample in the early age is much simpler compared to an older sample. At this stage, it is easier to observe the CNTs in mortar. At a latter curing stage, the high surface energy of CNTs
cause the hydration product to be attracted to the CNTs, thereby covering and hiding the CNTs from view [20]. By examining Fig. 11b, it is easy to identify the CNTs in the CNT-0 where they agglomerated filling in some of the pores and cracks in the mortar mix. Fig. 11c reveals the microstructure of a mix with 30% silica fumes but without CNTs (No CNT-30) where pores and cracks are visible, most likely due to the reduction of workability in the presence of 30% silica fumes. Results from the SEM analysis in Fig. 13 show that silica fume strongly effects the dispersion of CNT in the mix. Samples with 0% silica in their mix have a clump of CNT formed at a size of 5–20 lm (Fig. 11b), which is not the situation when 15% silica is added to the mixture where it is harder to locate the CNTs as they are more dispersed (Fig. 11d). The clumps of CNTs in samples without silica fume (Fig. 11b) have no negative effect on the compressive strength of the composite. Since the sizes of CNTs are much smaller than the diameter of cement particles, even a clump of CNTs has a smaller size compared to cement particles [21]. As a result, agglomerated CNTs have limited influence on mechanical properties of the mortar. On the contrary, mechanical test results revealed an increase in compressive strength by around 8% when CNTs are added to the sample (CNT-0 mix) and a limited improvement (11%) when using dispersed CNTs (CNT-H2O-0 mix) as illustrated earlier in Fig. 5. This could be attributed to the size of the CNT clump being small enough to fill some of tiny voids that could not be filled otherwise.
942
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Fig. 14. SEM image of mixes with 30% silica fume (a) 5000 magnification of CNT-30 mix, (b) 5000 magnification of CNT-OH-30 mix, (c) 5000 magnification of CNT-COOH30 mix, (d) 5000 magnification of CNT-H2O-30 mix.
Fig. 15. SEM image mix with 30% silica fume (a) 5000 magnification of CNT-COOH-30 mix, (b) 10,000 magnification of CNT-COOH-30 mix.
CNTs provide an excellent field for the crystal growth. Fig. 12 shows the SEM analysis for the samples with 0% silica fume and different types of CNTs. Clumps are apparent with CNT-0 mix, CNT-OH mix and CNT-COOH mix as indicated by Fig. 12a, b and c respectively. In Fig. 12a and b, the start of the crystal growth around the CNT clump can be clearly visualized. Fig. 12c reveals a void in the mortar that is filled with CNT clumps. This justifies the small improvement in strength that was achieved in the three
samples compared to CNT-H2O-0. It is harder to identify the clumps in Fig. 12d where there is a higher dispersion of CNTs in the CNT-H2O-0 mix. This dispersion lead to higher compressive strength compared to other samples as illustrated earlier. Fig. 13 shows the SEM analysis for the samples with 15% silica fume and different type of CNTs. Fig. 13 reveals that the CNTs in all cases are uniformly dispersed in the cement matrix with no agglomeration observed in any of the samples. Each CNT in the
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
943
Fig. 16. EDS analysis for CNT-0 mix at two locations spectrum 2 (far from a CNT) and 3 (near a CNT).
Fig. 17. EDS analysis for CNT-15 mix at two locations spectrum 2 (far from a CNT) and 4 (near a CNT).
Fig. 18. EDS analysis for CNT-30 mix at two locations spectrum 2 (near hydration products) and 3 (near a CNT).
mortar functions as a single fiber. In CNT-15 mix (Fig. 13a), the CNTs are covered by the hydration product between the micro cracks. CNT-OH-15 mix (Fig. 13b) shows silica fumes in the crystal rich area. Since crystals grow in a straight line like branches of a tree, we can differentiate between a crystal and a CNT from the curvature of the fiber. Fig. 13c (CNT-COOH-15 mix) shows the CNTs anchored to the silica fumes forming a strong bond with hydration products. Fig. 13d (CNT-H2O-15 mix) shows the bridging of CNT in a micro crack which enforces the results of the highest compressive strength for that mix.
When the amount of silica fume in the mix rises to 30%, the agglomerated CNTs that are tangled together can be observed (Fig. 14). Meanwhile, the edges of the clumped CNTs were held down in the cement and silica fume hydration products. However, functionalized CNTs (CNT-OH-30 and CNT-COOH-30 samples) show a network-like distribution in the mix that is able to improve the mechanical properties of the mortar mix. This network distribution of CNTs is more visible in the sample in Fig. 14(c) and Fig. 15, enforcing earlier results of a 20% increase in compressive strength and 50% increase in flexural strength for CNT-COOH-30 sample.
944
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Table 2 Cost comparison between No CNT-0 and CNT-COOH-30 mortar mixes. Material
Cement Silica CNT-COOH Water Sand Total cost USD/m3 a
Unit cost (USDa)
0.06/kg 0.27/kg 385/kg 2.09/m3 0.095/kg
Quantity needed
Cost (USD)
No CNT-0
CNT-COOH-30
No CNT-0
CNT-COOH-30
467 kg/m3
326.5 kg/m3 140 kg/m3 0.5 kg/m3 140 kg/m3 424 kg/m3
28.02
19.59 37.80 192.50 2.09 40.28 292.26
140 kg/m3 607 kg/m3
2.09 57.67 87.78
USD – US dollars.
3.4. Energy-Dispersive Spectroscopy (EDS) analysis An EDS analysis was performed to confirm the strength results, locate the CNTs in the samples and determine the composition of the nearby materials. Fig. 16 illustrates the EDS analysis results of the for CNT-0 mix at two locations – spectrum 2 (away from a CNT location) and spectrum 3 (near to a CNT). As we can see in spectrum 2, calcium is the primary element. The amount of the silica in the mix comes from hydration of cement in the mix. Gold and palladium in the analysis are due to gold etching of the samples. Other elements in the analysis might be related to impurity in the components or it might be part of the aggregate. In spectrum 3, carbon is also a main element in addition to calcium, and this indicates a presence of a CNT. Fig. 17 shows the EDS analysis of for CNT-15 mix at two locations – spectrums 2 (far from a CNT) and 4 (near a CNT). By comparing spectrum 2 and spectrum 4, it can be concluded that the percentage of silica is higher near to CNTs. Spectrum 4 shows a higher percentage of C (due to presence of CNTs), small amount of calcium and a high percentage of silica revealing the bonding between the silica and CNTs. On the other hand, spectrum 2 shows a high percentage of calcium and small amounts of silica and no presence of CNTs. Based on this it can be concluded that silica fume works as an intermediate element between CNT and the hydration product. CNTs anchor to the silica fume particles and make a strong bond with the hydration product. Fig. 18 shows the EDS analysis of CNT-30 mix with 30% silica and 0.15% CNT by weight of the cement. The amount of silica is high in all spectrums near CNTs (spectrum 3) and near hydration products (spectrum 2). Silica particles cover the CNTs and CNT re-agglomerate in silica-rich area. 4. Economic feasibility The economic feasibility of using CNTs in concrete was calculated through conducting a preliminary benefit/cost analysis. The benefit is quantized as the reduction in material quantity required for a specific load application. Table 2 illustrates the cost of the individual components in a No CNT-0 mix that exhibited a compressive strength of 59 MPa and a flexural strength of 4.37 MPa and a CNT-COOH-30 mix that exhibited a compressive strength of 70.7 MPa and a flexural strength of 6.58 MPa. Based on the measured compressive strengths of a No CNT-0 and a CNT-COOH-30 (59 MPa and 70.7 MPa, respectively) mix, a 50 50 cm2 column using No CNT-0 mortar is almost equivalent to a 45 45 cm2 column using CNT-COOH-30, that is a saving of nearly 0.05 m2. With an average price of unfinished apartment selling around USD 4100/m2 in UAE, a saving of around USD 205/m2 can be achieved in selling price due to the increase in space provided by using a CNT-COOH-30 mix mortar. Examining benefit/cost
ratio through this simplified example, a savings of USD 205.00 and an increase in cost of USD 204.48 would result through the use of the proposed CNT-COOH-30 mix mortar, resulting in a break even condition. It should be emphasized at this point that the current use of nanoparticles in the construction industry may not be cost effective everywhere due to the high price of nanoparticles. However, commercial applications in the future are promising based on the arguments provided here, as there is no extra burden in terms of construction cost. In fact, with increased use of nanoparticles in all fields of technology, including the construction industry, mass production of nanomaterials will become a norm, and this will assist in decreasing cost of nanoparticles further and lead to lower cost of building materials using CNTs. 5. Conclusions This study investigated the use of silica fumes and surface functionalized CNTs in the production of mortar. Results revealed that adding 0.15% CNTs to the cementitious material improved the compressive strength by 8%. Initially dispersed CNTs in water solution resulted in a compressive strength improvement of 8.5%. It showed that adding 15% silica fume with 0.15% CNTs has increased the compressive strength of all the samples from 7% to 13%. Use of non-dispersed, non-functionalized CNTs resulted in the lowest increase, and initially dispersed CNTs in a water solution the highest increase in the compressive strength. Increasing silica percentage to 30% had a negative impact on the compressive strength of the sample with CNT and CNT-H2O; however, use of functionalized CNTs resulted in an increase of 16–20%. Although silica fume can effectively disperse CNT in the composite, it should be added in conjunction to functionalized CNTs to effectively enhance the mechanical properties of the composite. The flexural strength of the samples increased by 50% and 35% when compared with the control sample (no CNT) when using surface functionalized CNTs (CNT-COOH and CNT-OH) and nonsurface functionalized CNTs, respectively. The increase in flexural strength with CNT-COOH could be related to the presence of carboxylic acid groups on the surfaces of carbon nanotubes that bring about a chemical interaction between the carboxylic acid and the calcium silicate hydrate (C–S–H) or Ca(OH)2. The interaction leads to a strong covalent force on the interface between the CNT reinforcement and matrix in the composites, thereby bridging micro cracks as seen through SEM analysis. SEM results revealed that although all the CNTs were not dispersed properly and formed clumps, they were able to fill the voids in the mortar, and thereby improved the compressive strength. Results validated previous researchers’ work that surface functionalization and proper dispersion lead to an improvement in mechanical properties, but these improvements are limited, especially at low percentage of CNTs.
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Acknowledgement Part of this work was financially supported by the American University of Sharjah Faculty Research Grant FRG14-2-21. References [1] G. Yakovlev, G. Pervushin, I. Maeva, J. Keriene, I. Pudov, A. Shaybadullina, A. Buryanov, A. Korzhenko, S. Senkov, Modification of construction materials with multi-walled carbon nanotubes, Proc. Eng. 57 (2013) 407–413. [2] S. Chuah, Z. Pan, J.G. Sanjayan, C.M. Wang, Nano reinforced cement and concrete composites and new perspective from graphene oxide, Constr. Build. Mater. 73 (2014) 113–124. [3] R. Kelsall, I. Hamley, M. Geoghegan, Nanoscale Science and Technology, Wiley, 2006. [4] A. Cwirzen, K. Habermehl-Cwirzen, A. Nasibulin, E. Kaupinen, P. Mudimela, V. Penttala, SEM/AFM studies of cementitious binder modified by MWCNT and nano-sized FE needs, Mater. Charact. 60 (2009) 735–740. [5] K. Fattah, N.M. Hassan, A. Al-Tamimi, Effect of adding polar impurities on carbon nanotubes and concrete bonding strength, 10th International Conference on Composite Science and Technology, Lisboa, Portugal, 2015. [6] T. Kowald, R. Trettin, Influence of surface-modified carbon nanotubes on ultrahigh performance concrete, International Symposium on Ultra High Performance Concrete, 2004. [7] H. Kim, L. Nam, H. Lee, Enhanced effect of carbon nanotube on mechanical and electrical properties of cement composites by incorporation of silica fume, Compos. Struct. 107 (2014) 60–69. [8] K. Scrivener, R. Kirkpatrick, Innovation in use and research on cementitious material, Cem. Concr. Res. 39 (2008) 128–136. [9] N.M. Hassan, K. Fattah, A. Al-Tamimi, Standardizing protocol for incorporating CNTs in concret, 6th International Conference on Nanotechnology: Fundamentals and Applications (ICNFA’15), Barcelona, Spain, July 15–17, 2015.
945
[10] J. Makar, J. Margeson, J. Luh, Carbon nanotube or cement composites – early results and potential applications, 3rd International Conference on Construction Materials: Performance, Innovations and Structural Implications, Vancouver, 2005. [11] M.S. Konsta-Gdoutos, Z.S. Metaxa, S.P. Shah, Highly dispersed carbon nanotube reinforced cement based materials, Cem. Concr. Res. 40 (7) (2010) 1052–1059. [12] P. Stynoski, P. Mondal, C. Marsh, Effects of silica additives on fracture properties of carbon nanotube and carbon fiber reinforced Portland cement mortar, Cem. Concr. Compos. 55 (2015) 232–240. [13] A. Katz, A. Bentur, Effect of matrix composition on the aging of CFRC, Cem. Concr. Compos. 17 (2) (1995) 87–97. [14] G.Y. Li, P.M. Wang, X. Zhao, Mechanical behavior and microstructure of cement composites incorporating surface-treated multi-walled carbon nanotubes, Carbon 43 (6) (2005) 1239–1245. [15] M. Paiva, B. Zhou, K. Fernando, Y. Lin, J. Kennedy, Y.P. Sun, Mechanical and morphological characterization of polymercarbon nanocomposites from functionalized carbon nanotubes, Carbon 42 (14) (2004) 2849–2854. [16] R. Siddique, A. Mehta, Effect of carbon nanotubes on properties of cement mortars, Constr. Build. Mater. 50 (2014) 116–129. [17] Nanostructured & Amorphous Material Inc., [Online], Available: (accessed 15.9.2015). [18] Tescan VEGA III LMU with Oxford Instruments EDS, [Online], Available: (accessed 15.9.2015). [19] B. Zou, S.J. Chen, A.H. Korayem, F. Collins, C. Wang, W.H. Duan, Effect of ultrasonication energy on engineering properties of carbon nanotube reinforced cement pastes, Carbon 85 (2015) 212–220. [20] S. Xu, J. Liu, Q. Li, Mechanical properties and microstructure of multiwalled carbon nanotube-reinforced cement paste, Constr. Build. Mater. 76 (2015) 16–23. [21] T. Nochaiya, A. Chaipanich, Behavior of multi-walled carbon nanotubes on the porosity and microstructure of cement-based materials, Appl. Surf. Sci. 257 (6) (2011) 1941–1945.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Performance of cementitious materials produced by incorporating surface treated multiwall carbon nanotubes and silica fume Adil Tamimi a, Noha M. Hassan b, Kazi Fattah a,⇑, Amirhooman Talachi a a b
Department of Civil Engineering, American University of Sharjah, United Arab Emirates Industrial Engineering Department, American University of Sharjah, United Arab Emirates
h i g h l i g h t s Use of functionalized CNTs improve compressive and flexural strength of cement mortar. Using CNT-COOH with silica fumes increased compressive strength by 50%. Addition of silica fume enables better dispersion of CNTs in mortar samples. SEM results revealed that the CNTs were able to fill the voids in the mortar.
a r t i c l e
i n f o
Article history: Received 3 February 2016 Received in revised form 28 March 2016 Accepted 30 March 2016 Available online 9 April 2016 Keywords: Carbon nano tube CNT Silica fumes Dispersion Surface functionalization SEM Cement composite Mortar
a b s t r a c t The outstanding mechanical properties of carbon nanotubes (CNTs) highlight them as potential candidates for cementitious material reinforcement. However, their low surface friction and the Van der Waals forces of attraction between them, cause the CNTs to aggregate with each other rather than bind with the cement matrix. A number of methods have been investigated by researchers to reduce the aggregation, improve dispersion and activate the graphite surface to enhance its interfacial interaction. These methods involve surface functionalization and coating, optimal physical blending, use of surfactant and other admixtures. This research investigates the use of silica fumes (an admixture), surface functionalized CNTs and cement paste to overcome those obstacles. CNTs with polar impurities end groups OH and COOH were examined. Mortar samples with non-functionalized CNTs dispersed in water solution, another with non-dispersed, non-functionalized CNTs, and a third batch with no CNTs (as control) was used also studied. Silica fumes volume fraction was varied from 0 to 30% to determine its effect. Compressive and flexural strengths of the different mixes were measured and compared. Qualitative analysis using Scanning Electron Microscope (SEM) and Energy-Dispersive Spectroscopy (EDS) were carried out to study the morphology of each mix. Results reveal a much higher enhancement in strength both compressive and flexural strengths for the functionalized CNTs with 30% silica fumes over the other samples. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction The use of nanomaterials has grown significantly since the first time it was discovered in 1959 [1]. Similar to other industries, the construction industry can benefit from nanotechnology in order to produce a more desirable outcome. One of the important aspects of the construction industry, which can take advantage of nanotechnology, is concrete production. The properties and characteristics of carbon nanotubes (CNT) provide great prospects to improve ⇑ Corresponding author. E-mail addresses: [email protected] (A. Tamimi), [email protected] (N.M. Hassan), [email protected] (K. Fattah), [email protected] (A. Talachi). http://dx.doi.org/10.1016/j.conbuildmat.2016.03.216 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
the properties of concrete when the CNTs are incorporated to the concrete mixture. CNTs can work as nano filler in concrete and has the potential to improve the strength and durability of concrete. The sizes of CNT are much smaller compared with the typical components in cement, and thus can improve mechanical properties at the nano scale [2]. The outstanding mechanical properties of carbon nanotubes (CNTs) highlight them as potential candidates for concrete reinforcement as well. The strength of the CNTs is directly related to the strong C@C bond and the relatively small number of defects present in the tubes. It is said to possess ‘‘a hundred times the strength of steel at one sixth of the weight” [3]. Young modulus is estimated to vary between 1 and 5 TPa, density is around 2000 kg/m3, elongation to failure of 20–30%, which
935
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
correlate to a tensile strength of well above 100 GPa. The CNTs are characterized by thermal stability up to 2800 °C [4]. Previous studies have shown that the compressive strength of concrete can increase with the addition of CNTs [5]. Despite the obvious advantages of adding CNTs in concrete, there are two major problems associated with its use. CNTs have a tendency to agglomerate due to the Van der Waals attraction force between CNT crystalline ropes when it is produced. The triangular network formed by these ropes tends to aggregate and reduce the dispersion of the CNTs in the water that is added to concrete. The second major drawback associated with CNTs is their very low surface friction, making it difficult for them to bind together or with the cement matrix material. The bonding between cement and CNT is often weak and only the hydration products can anchor a few CNTs. However, knowledge in this field is still limited [6]. Different techniques, such as use of silica fume [7,8], sonification [9] and surfactant addition [10] have been tried to disperse CNT in the mortar mix. Each of these methods has its own limitations and drawbacks. Sonication alone can break the CNT tubes resulting in a weaker reinforcement. Sonication and surfactant use can increase the air volume in the mix resulting in a voided structure. Additionally, although hydration product such as cal cium–silicate–hydrates (C–S–H), calcium hydroxide (CH) have same or larger dimensions than CNTs, only few of them are anchored in hydration products. Consequently, they don’t provide effective reinforcement of cement composite. A study by [11] investigated how the length of CNT affects the property of the cement mortar. Sonication and surfactants were used to disperse
the CNTs. Flexural test results show that sonication does not have an effect on surfactant performance since the flexural strength for all the samples were the same. Effective flexural strength can be achieved with high percentage of shorter CNTs, or less percentage of longer CNTs. A study by [7] showed that incorporating a small amount of silica fume to the mix can improve dispersion of CNT and improve its compressive strength. Because of very small particle size of silica fume (in range of 10–500 nm), which is close to the size of CNTs, silica fume can be mixed with agglomerated CNTs and effectively disperse them. Moreover, silica fume particles intermixed with CNTs gets hydrated by Ca2+ ions from the cement due to their high pozzolanic activity. Hydration products of silica fume effectively anchor CNTs, and interfacial interaction between CNTs and the hydration products was thereby enhanced. A study by [12] observed that silica fume and silica functional groups improved the fracture performance of mixtures containing carbon nanotubes and carbon fibers, but further optimization of dosage, size, and interface strength is required to fully utilize carbon nanotubes in cementitious composites [12]. Improvements in flexural strength and fracture toughness are more significant at a later age in mixtures containing silica fume, although higher volumes of silica may lead to an adverse effect by concentrating the dispersed CNT in the silica fume fields [12,13]. Li et al. [14] investigated the mechanical properties of cement composite that contained functionalized CNTs. Results showed that surface treated CNTs exhibited better compressive and flexural strength due to the interfacial interaction with hydration
Fig. 1. Schematic of functionalized CNTs used in the study (a) CNT-OH, (b) CNT-COOH and (c) SEM of MWNTs [17].
Table 1 Mix proportions of CNTs-silica fumes mortars. Set
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 a b c d e f
#
Mix name
CNT type
Silica fumes%
CNT (% wt of cement)
Cement (kg/m3)
Silica (kg/m3)
CNTs (kg/m3)
Sand before water correction (kg/m3)
Water before water correction (kg/m3)
w/ca
s/cb
No CNT 0 CNT-0 CNT-OH-0 CNT-COOH-0 CNT-H2O-0 No CNT-15 CNT-15 CNT-OH-15 CNT-COOH-15 CNT-H2O-15 No CNT-30 CNT-30 CNT-OH-30 CNT-COOH-30 CNT-H2O-30
– CNTc CNT-OHd CNT-COOHe CNT-H2Of – CNT CNT-OH CNT-COOH CNT-H2O – CNT CNT-OH CNT-COOH CNT-H2O
0 0 0 0 0 15% 15% 15% 15% 15% 30% 30% 30% 30% 30%
0.00 0.15 0.15 0.15 0.15 0.00 0.15 0.15 0.15 0.15 0.00 0.15 0.15 0.15 0.15
467.0 466.3 466.3 466.3 466.3 397.0 396.4 396.4 396.4 396.4 327.0 326.5 326.5 326.5 326.5
0 0 0 0 0 70 70 70 70 70 140 140 140 140 140
0.0 0.7 0.7 0.7 0.7 0.0 0.6 0.6 0.6 0.6 0.0 0.5 0.5 0.5 0.5
607 606 606 606 606 516 515 515 515 515 425 424 424 424 424
140 140 140 140 140 140 140 140 140 140 140 140 140 140 140
0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3 0.3
1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3
w/c: water to cement ratio. s/c: sand to cement ratio. CNT: Industrial grade MWCNTs. CNT-OH: Industrial grade MWCNTs functionalized with hydroxyl group. CNT-COOH: Industrial grade MWCNTs functionalized with carboxyl group. CNT-H2O: MWCNT dispersed in water.
936
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Fig. 2. Sample preparation equipment.
Fig. 3. (a) Compressive test for cubic sample and (b) flexural test for beam sample.
Fig. 4. (a) Gold etching process and (b) SEM analysis instruments.
product such as C-S-H and surface functionalized CNTs, resulting in a strong connection between CNT and matrix composite, reduction of pores and voids in cement matrix leading to more compact and dense concrete and bridging the connection of CNTs between
cracks and voids. The investigators found that the interfacial interaction with hydration products such as Calcium Silicate Hydrate (C-S-H) and surface functionalized CNT lead to a strong bond between CNT and the matrix composites. Despite the obvious
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
937
Fig. 5. Effect of increasing silica fume in the mix on the compressive strength of mortar samples.
Fig. 6. Ratio of compressive strength of 30% silica fumes mixes compared to No CNT-30 mix (control).
Fig. 7. Flexural strength of samples with 30% silica fume.
and well-documented results on the benefits of nanomaterials in concrete production [2,15,16], the commercial applications for concrete production are being prevented due to the high cost of nanoparticles. If efficient and effective use of nanomaterials in construction can be reached, this might initiate a need for mass production of nanomaterials, thereby decreasing its initial cost, which is currently a limiting factor. This research studied the effect of incorporating silica fumes with functionalized CNTs in the cement mortar to enhance its mechanical properties. The reaction of the polar impurities end groups (OH and COOH) with silica fumes and cement matrix were investigated through SEM and EDS analysis. The effect of this incorporation on the mortar’s mechanical properties was also examined. 2. Materials and methods 2.1. Experimental program 2.1.1. Multiwall carbon nanotubes (MWCNTs) Four industrial grade multiwall carbon nanotubes were used in this study. Regular CNTs, CNTs functionalized with hydroxyl group (CNT-OH) (Fig. 1a), CNTs functionalized with carboxyl groups (CNT-COOH) (Fig. 1b), and CNT that is suspended in water (CNT-H2O) were studied. Aromatic modified polyethylene glycol ether was used as solvent for the dispersion procedure. All CNT grades had 88 + % purity, an outside diameter that varies from 20 to 40 nm, an inner diameter of 5–10 nm and a length varying between 10 and 30 lm as shown in Fig. 1c. The CNT-OH contained 1.55–1.71 wt% –OH. The CNT-COOH contained 1.36– 1.50 wt% –COOH. The water dispersed CNTs had a dispersant content of 1.4–1.6% and are stable for up to 6 months at room temperature. All CNTs were purchased from [17].
Fig. 8. Beam after flexural test showing major failure location.
2.1.2. Mortar mix The mix prepared and tested in this study was composed of normal Portland cement, silica fume, crushed sand and CNTs. Normal Portland cement having a surface area of 355 m2/kg on the Baline test was used. The silica fume was produced in conformance with the ASTM C 1240 specifications with a specific gravity 2.1–2.4, a densified bulk density of 608–720 kg/m3, specific surface of 17 m2/kg and particle size of approximately 0.4 lm. Crushed sands, with particle sizes in the range of 0.15–4.75 mm and specific gravity of 2.65 were used as fine aggregate. The crushed sand had water absorption of 5%, which was tested based on the ASTM test C 128-88.
938
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Fig. 9. Stress Strain diagram of (a) No CNT-30 mix, (b) CNT-COOH-30 mix.
Fig. 10. SEM image of pure CNT without any additive form a clump (a) 1000 magnification, (b) 2000 magnification, (c) 5000 magnification, (d) 10,000 magnification.
2.1.3. Specimen preparation Mortar samples were prepared with three different percentages of silica in the mix and four different types of CNTs for each percentage. Table 1 summarizes the list of 15 sets of mixtures studied. The ratio of CNT/cement in the mix was kept constant to effectively determine the effect of silica on the dispersion of each type of CNT. The amount of CNT in all mixes was kept at 0.15% of the cement weight since CNTs absorb a lot of water and reduce the workability of the mix [7]. Water/cement ratio in the mix was limited to 0.3 to increase the collision of silica molecules with CNT clumps and reduce agglomeration. Water absorption of the aggregate was tested according to ASTM C128-01. The cement-to-sand ratio was kept at 1.3. Three
different percentages (0%, 15% and 30%) of silica fume by weight of cement in the mix were chosen to examine the effect of silica fumes on different types of CNTs. Previous research [7] revealed a nonlinear trend between the compressive strength of CNT reinforced cementitious material and the addition of silica fumes. The choice for the wide range of percentage of silica (0–30%) investigated in the present study was to be able to capture this nonlinearity. A standard Hobart mixer was used for the mixing process (Fig. 2). For the mixing process, all the dry material (cement, sand, silica fume, and CNT) was added to the mixer and mixed for 4 min. After that, water was slowly added to the mix and the mixer was run for another 5 min to make sure all materials were completely mixed with water. Fresh mix was added
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
939
Fig. 11. SEM image of mortar at (a) 5000 magnification of No CNT-0 mix, (b) 5000 magnification of CNT-0 mix, (c) 200 magnification of No CNT-30 mix, (d) 5000 magnification of CNT-15 mix.
to the mold in three layers and each layer was compacted with compacting tools and a shaker table. After one day the mortar was demolded and cured for 14 days. Each set of experiment was repeated three times.
analysis of the mortar mix was performed to determine materials present near the CNTs. The EDS analysis was conducted on 3 samples with 0.15% normal CNT and different ratios (0%, 15% and 30%) of silica fume by weight of cement in the mix.
2.2. Material characterization tests Compressive and flexural tests were conducted on each sample and average of the three samples was taken as final strength for that mix. 2.2.1. Compressive strength The compressive strength of the mortars was tested on 45 cubic samples (15 sets 3 replicates) each of size 50 50 50 mm. The tests were carried out using Instron Universal Testing Machine as shown in Fig. 3a. 2.2.2. Flexural strength Flexural tests were conducted on mortar samples each having sizes 210 50 50 mm. For the flexural test, a four-point loading according to ASTM C651-15 was conducted for the beam with 20 cm span. Fig. 3b illustrates the actual testing conditions for the flexural test. 2.2.3. Morphology and composition characterization The crushed mortar was examined using a scanning electron microscopic to study its surface morphology and microstructure. Microscopic image of sample was made through the interaction of the focused beam of electrons with sample surface atoms revealing information about the microstructure of cement composites. Sample preparation was performed by gold etching for 75 s. The mortar sample was loaded into a high vacuum chamber while a very thin layer of Au, Pd was deposited on the surface (Fig. 4a). A scanning electron microscope (SEM) with energy dispersive spectroscopy (EDS) [18] was used to determine the detailed structure of the nano particles that were used in the study (Fig. 4b). The EDS
3. Results and discussion 3.1. Compressive strength Fig. 5 illustrates the trend of adding silica fumes to the mix. Adding CNT to the mix without the addition of silica slightly increases the compressive strength of all samples. At 0% silica fumes, the CNT-H2O-0 mix has the highest compressive strength as it is initially suspended and dispersed in water. However, the improvement in compressive strength is low at around 11%. Adding 15% silica fume increases the compressive strength of the samples. This increase is more observable for the samples with dispersed CNTs (CNT-H2O-15) and functionalized CNTs with hydroxyl group (CNT-OH-15) whose compressive strength increased by 13.8% and 10.3%, respectively when compared to the sample without CNT (No CNT-15 mix). Sample with CNT-OH-15 and CNT-COOH15 increased to even higher strength values compared to CNT-15 sample. This can be related to the strong bond between hydration products and surface treated CNTs. The interfacial interaction of the hydration product such as C-S-H and surface functionalized CNT (CNT-OH-15 and CNT-COOH-15) is much stronger compared
940
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Fig. 12. SEM image of mixes without any silica fume (a) 5000 magnification of CNT-0 mix, (b) 5000 magnification of CNT-OH-0 mix, (c) 5000 magnification of CNT-COOH0 mix, (d) 5000 magnification of CNT-H2O-0 mix.
to normal CNTs (CNT-15). By increasing the silica fume in the mix to 30%, the compressive strengths for the samples with CNTs (CNT-30) and CNT-H2O (CNT-H2O-30) are reduced, revealing the nonlinear trend that was exhibited by previous researchers using CNTs and silica fumes alone [7]. However, surface functionalized CNTs (CNT-COOH-30 and CNT-OH-30 samples) did not exhibit this decline. A much higher strength was achieved by both samples compared to the NO CNT-30 sample with a 20% and a 16% increase in strength, respectively. Fig. 6 shows the ratio of compressive strength of each of the samples with 30% silica fumes to that of the control (without CNTs – No CNT-30 mix). Samples CNT-30 and dispersed CNT-H2O-30 decreased in strength and only the functionalized samples (CNTOH-30 and CNT-COOH-30) increased in strength. Adding silica fume more than 15% along with CNTs (CNT-30 or CNT-H2O-30) reduced the workability of the mix and increased the voids in the mortar. Additionally, CNT agglomerations were dispersed densely in silica fume fields and re-agglomerated as clumps with smaller sizes, covering cement and hindering its hydration. However, the compressive strength of the sample with surface functionalized CNT (CNT-OH-30 and CNT-COOH-30) increased as much as 20% compared to the control sample (No CNT-30 mix) even with higher percentage of silica fumes (30%). This could be attributed to the development of strong bond between the surface functionalized CNTs and hydration product.
3.2. Flexural strength To ensure consistency in results between compressive and flexural strength, flexural tests were run for 30% silica fumes. Fig. 7 reveals that the samples with CNT-OH-30 and CNT-COOH-30 have the highest flexural strength with strengths of 6.44 MPa and 6.58 MPa, respectively. For all the samples, a crack developed inside the middle area, as shown in Fig. 8. The reason for this is that CNTs with a functionalized surface develop further bridging compared to non-functionalized CNTs leading to better mortar reinforcement. The flexural strength of No CNT-30 mix was equal to 4.37 MPa, the lowest among all the specimens. The flexural strength of the sample with CNT (CNT-30) and CNT-H2O-30 improved by 30%, compared to the No CNT-30 sample. This improvement in flexural strength is much higher for surface treated samples. The flexural strength of CNT-COOH-30 and CNT-OH-30 increased by 50% compared to the control (No CNT30). The interaction between end groups and calcium silicate hydrate (C–S–H) or Ca(OH)2 leads to a strong covalent bond on the interface between the reinforcement and matrix in the composites, bridging micro cracks, and increasing the ultimate displacement. Fig. 9 shows the stress-strain curve of the No CNT-30 mix (control) and CNT-COOH-30 mix illustrating an increase in both flexural strength and ultimate displacement and strain at failure. It can be concluded that the mechanical properties of the
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
941
Fig. 13. SEM image of mixes with 15% silica fume (a) 5000 magnification of CNT-15 mix, (b) 5000 magnification of CNT-OH-15 mix, (c) 5000 magnification of CNT-COOH15 mix, (d) 5000 magnification of CNT-H2O-15 mix .
composites are improved significantly with addition of 30% silica fumes and surface functionalized CNTs. 3.3. Microstructure analysis using Scanning Electron Microscopy (SEM) To reinforce the conclusions deduced from the mechanical tests, an examination of the microstructure was performed. Fig. 10 shows the SEM image of a clump of CNT without any additive. The images were taken with different magnifications (1000, 2000, 5000 and 10,000) and the strings of CNTs clumping together can be clearly visualized. Strong self-attraction between each individual CNT caused by strong van der Waals forces leads to agglomeration of CNTs. To be able to identify CNTs in the mortar sample, it was important to first examine the microstructure without silica fumes and CNTs (No CNT-0) mix (Fig. 11a), and compare it to the mix without silica fumes but with CNTs (CNT-0) mix (Fig. 11b). The main elements of the cement paste are Ca(OH)2, C–S–H gel, pores, cracks, un-hydrated cement particles, fine aggregates [19]. In Fig. 11a, the micro size crystal growth can be seen clearly. The crystals grow in a straight line similar to the branches of a tree. The SEM analysis of the sample in the early age is much simpler compared to an older sample. At this stage, it is easier to observe the CNTs in mortar. At a latter curing stage, the high surface energy of CNTs
cause the hydration product to be attracted to the CNTs, thereby covering and hiding the CNTs from view [20]. By examining Fig. 11b, it is easy to identify the CNTs in the CNT-0 where they agglomerated filling in some of the pores and cracks in the mortar mix. Fig. 11c reveals the microstructure of a mix with 30% silica fumes but without CNTs (No CNT-30) where pores and cracks are visible, most likely due to the reduction of workability in the presence of 30% silica fumes. Results from the SEM analysis in Fig. 13 show that silica fume strongly effects the dispersion of CNT in the mix. Samples with 0% silica in their mix have a clump of CNT formed at a size of 5–20 lm (Fig. 11b), which is not the situation when 15% silica is added to the mixture where it is harder to locate the CNTs as they are more dispersed (Fig. 11d). The clumps of CNTs in samples without silica fume (Fig. 11b) have no negative effect on the compressive strength of the composite. Since the sizes of CNTs are much smaller than the diameter of cement particles, even a clump of CNTs has a smaller size compared to cement particles [21]. As a result, agglomerated CNTs have limited influence on mechanical properties of the mortar. On the contrary, mechanical test results revealed an increase in compressive strength by around 8% when CNTs are added to the sample (CNT-0 mix) and a limited improvement (11%) when using dispersed CNTs (CNT-H2O-0 mix) as illustrated earlier in Fig. 5. This could be attributed to the size of the CNT clump being small enough to fill some of tiny voids that could not be filled otherwise.
942
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Fig. 14. SEM image of mixes with 30% silica fume (a) 5000 magnification of CNT-30 mix, (b) 5000 magnification of CNT-OH-30 mix, (c) 5000 magnification of CNT-COOH30 mix, (d) 5000 magnification of CNT-H2O-30 mix.
Fig. 15. SEM image mix with 30% silica fume (a) 5000 magnification of CNT-COOH-30 mix, (b) 10,000 magnification of CNT-COOH-30 mix.
CNTs provide an excellent field for the crystal growth. Fig. 12 shows the SEM analysis for the samples with 0% silica fume and different types of CNTs. Clumps are apparent with CNT-0 mix, CNT-OH mix and CNT-COOH mix as indicated by Fig. 12a, b and c respectively. In Fig. 12a and b, the start of the crystal growth around the CNT clump can be clearly visualized. Fig. 12c reveals a void in the mortar that is filled with CNT clumps. This justifies the small improvement in strength that was achieved in the three
samples compared to CNT-H2O-0. It is harder to identify the clumps in Fig. 12d where there is a higher dispersion of CNTs in the CNT-H2O-0 mix. This dispersion lead to higher compressive strength compared to other samples as illustrated earlier. Fig. 13 shows the SEM analysis for the samples with 15% silica fume and different type of CNTs. Fig. 13 reveals that the CNTs in all cases are uniformly dispersed in the cement matrix with no agglomeration observed in any of the samples. Each CNT in the
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
943
Fig. 16. EDS analysis for CNT-0 mix at two locations spectrum 2 (far from a CNT) and 3 (near a CNT).
Fig. 17. EDS analysis for CNT-15 mix at two locations spectrum 2 (far from a CNT) and 4 (near a CNT).
Fig. 18. EDS analysis for CNT-30 mix at two locations spectrum 2 (near hydration products) and 3 (near a CNT).
mortar functions as a single fiber. In CNT-15 mix (Fig. 13a), the CNTs are covered by the hydration product between the micro cracks. CNT-OH-15 mix (Fig. 13b) shows silica fumes in the crystal rich area. Since crystals grow in a straight line like branches of a tree, we can differentiate between a crystal and a CNT from the curvature of the fiber. Fig. 13c (CNT-COOH-15 mix) shows the CNTs anchored to the silica fumes forming a strong bond with hydration products. Fig. 13d (CNT-H2O-15 mix) shows the bridging of CNT in a micro crack which enforces the results of the highest compressive strength for that mix.
When the amount of silica fume in the mix rises to 30%, the agglomerated CNTs that are tangled together can be observed (Fig. 14). Meanwhile, the edges of the clumped CNTs were held down in the cement and silica fume hydration products. However, functionalized CNTs (CNT-OH-30 and CNT-COOH-30 samples) show a network-like distribution in the mix that is able to improve the mechanical properties of the mortar mix. This network distribution of CNTs is more visible in the sample in Fig. 14(c) and Fig. 15, enforcing earlier results of a 20% increase in compressive strength and 50% increase in flexural strength for CNT-COOH-30 sample.
944
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Table 2 Cost comparison between No CNT-0 and CNT-COOH-30 mortar mixes. Material
Cement Silica CNT-COOH Water Sand Total cost USD/m3 a
Unit cost (USDa)
0.06/kg 0.27/kg 385/kg 2.09/m3 0.095/kg
Quantity needed
Cost (USD)
No CNT-0
CNT-COOH-30
No CNT-0
CNT-COOH-30
467 kg/m3
326.5 kg/m3 140 kg/m3 0.5 kg/m3 140 kg/m3 424 kg/m3
28.02
19.59 37.80 192.50 2.09 40.28 292.26
140 kg/m3 607 kg/m3
2.09 57.67 87.78
USD – US dollars.
3.4. Energy-Dispersive Spectroscopy (EDS) analysis An EDS analysis was performed to confirm the strength results, locate the CNTs in the samples and determine the composition of the nearby materials. Fig. 16 illustrates the EDS analysis results of the for CNT-0 mix at two locations – spectrum 2 (away from a CNT location) and spectrum 3 (near to a CNT). As we can see in spectrum 2, calcium is the primary element. The amount of the silica in the mix comes from hydration of cement in the mix. Gold and palladium in the analysis are due to gold etching of the samples. Other elements in the analysis might be related to impurity in the components or it might be part of the aggregate. In spectrum 3, carbon is also a main element in addition to calcium, and this indicates a presence of a CNT. Fig. 17 shows the EDS analysis of for CNT-15 mix at two locations – spectrums 2 (far from a CNT) and 4 (near a CNT). By comparing spectrum 2 and spectrum 4, it can be concluded that the percentage of silica is higher near to CNTs. Spectrum 4 shows a higher percentage of C (due to presence of CNTs), small amount of calcium and a high percentage of silica revealing the bonding between the silica and CNTs. On the other hand, spectrum 2 shows a high percentage of calcium and small amounts of silica and no presence of CNTs. Based on this it can be concluded that silica fume works as an intermediate element between CNT and the hydration product. CNTs anchor to the silica fume particles and make a strong bond with the hydration product. Fig. 18 shows the EDS analysis of CNT-30 mix with 30% silica and 0.15% CNT by weight of the cement. The amount of silica is high in all spectrums near CNTs (spectrum 3) and near hydration products (spectrum 2). Silica particles cover the CNTs and CNT re-agglomerate in silica-rich area. 4. Economic feasibility The economic feasibility of using CNTs in concrete was calculated through conducting a preliminary benefit/cost analysis. The benefit is quantized as the reduction in material quantity required for a specific load application. Table 2 illustrates the cost of the individual components in a No CNT-0 mix that exhibited a compressive strength of 59 MPa and a flexural strength of 4.37 MPa and a CNT-COOH-30 mix that exhibited a compressive strength of 70.7 MPa and a flexural strength of 6.58 MPa. Based on the measured compressive strengths of a No CNT-0 and a CNT-COOH-30 (59 MPa and 70.7 MPa, respectively) mix, a 50 50 cm2 column using No CNT-0 mortar is almost equivalent to a 45 45 cm2 column using CNT-COOH-30, that is a saving of nearly 0.05 m2. With an average price of unfinished apartment selling around USD 4100/m2 in UAE, a saving of around USD 205/m2 can be achieved in selling price due to the increase in space provided by using a CNT-COOH-30 mix mortar. Examining benefit/cost
ratio through this simplified example, a savings of USD 205.00 and an increase in cost of USD 204.48 would result through the use of the proposed CNT-COOH-30 mix mortar, resulting in a break even condition. It should be emphasized at this point that the current use of nanoparticles in the construction industry may not be cost effective everywhere due to the high price of nanoparticles. However, commercial applications in the future are promising based on the arguments provided here, as there is no extra burden in terms of construction cost. In fact, with increased use of nanoparticles in all fields of technology, including the construction industry, mass production of nanomaterials will become a norm, and this will assist in decreasing cost of nanoparticles further and lead to lower cost of building materials using CNTs. 5. Conclusions This study investigated the use of silica fumes and surface functionalized CNTs in the production of mortar. Results revealed that adding 0.15% CNTs to the cementitious material improved the compressive strength by 8%. Initially dispersed CNTs in water solution resulted in a compressive strength improvement of 8.5%. It showed that adding 15% silica fume with 0.15% CNTs has increased the compressive strength of all the samples from 7% to 13%. Use of non-dispersed, non-functionalized CNTs resulted in the lowest increase, and initially dispersed CNTs in a water solution the highest increase in the compressive strength. Increasing silica percentage to 30% had a negative impact on the compressive strength of the sample with CNT and CNT-H2O; however, use of functionalized CNTs resulted in an increase of 16–20%. Although silica fume can effectively disperse CNT in the composite, it should be added in conjunction to functionalized CNTs to effectively enhance the mechanical properties of the composite. The flexural strength of the samples increased by 50% and 35% when compared with the control sample (no CNT) when using surface functionalized CNTs (CNT-COOH and CNT-OH) and nonsurface functionalized CNTs, respectively. The increase in flexural strength with CNT-COOH could be related to the presence of carboxylic acid groups on the surfaces of carbon nanotubes that bring about a chemical interaction between the carboxylic acid and the calcium silicate hydrate (C–S–H) or Ca(OH)2. The interaction leads to a strong covalent force on the interface between the CNT reinforcement and matrix in the composites, thereby bridging micro cracks as seen through SEM analysis. SEM results revealed that although all the CNTs were not dispersed properly and formed clumps, they were able to fill the voids in the mortar, and thereby improved the compressive strength. Results validated previous researchers’ work that surface functionalization and proper dispersion lead to an improvement in mechanical properties, but these improvements are limited, especially at low percentage of CNTs.
A. Tamimi et al. / Construction and Building Materials 114 (2016) 934–945
Acknowledgement Part of this work was financially supported by the American University of Sharjah Faculty Research Grant FRG14-2-21. References [1] G. Yakovlev, G. Pervushin, I. Maeva, J. Keriene, I. Pudov, A. Shaybadullina, A. Buryanov, A. Korzhenko, S. Senkov, Modification of construction materials with multi-walled carbon nanotubes, Proc. Eng. 57 (2013) 407–413. [2] S. Chuah, Z. Pan, J.G. Sanjayan, C.M. Wang, Nano reinforced cement and concrete composites and new perspective from graphene oxide, Constr. Build. Mater. 73 (2014) 113–124. [3] R. Kelsall, I. Hamley, M. Geoghegan, Nanoscale Science and Technology, Wiley, 2006. [4] A. Cwirzen, K. Habermehl-Cwirzen, A. Nasibulin, E. Kaupinen, P. Mudimela, V. Penttala, SEM/AFM studies of cementitious binder modified by MWCNT and nano-sized FE needs, Mater. Charact. 60 (2009) 735–740. [5] K. Fattah, N.M. Hassan, A. Al-Tamimi, Effect of adding polar impurities on carbon nanotubes and concrete bonding strength, 10th International Conference on Composite Science and Technology, Lisboa, Portugal, 2015. [6] T. Kowald, R. Trettin, Influence of surface-modified carbon nanotubes on ultrahigh performance concrete, International Symposium on Ultra High Performance Concrete, 2004. [7] H. Kim, L. Nam, H. Lee, Enhanced effect of carbon nanotube on mechanical and electrical properties of cement composites by incorporation of silica fume, Compos. Struct. 107 (2014) 60–69. [8] K. Scrivener, R. Kirkpatrick, Innovation in use and research on cementitious material, Cem. Concr. Res. 39 (2008) 128–136. [9] N.M. Hassan, K. Fattah, A. Al-Tamimi, Standardizing protocol for incorporating CNTs in concret, 6th International Conference on Nanotechnology: Fundamentals and Applications (ICNFA’15), Barcelona, Spain, July 15–17, 2015.
945
[10] J. Makar, J. Margeson, J. Luh, Carbon nanotube or cement composites – early results and potential applications, 3rd International Conference on Construction Materials: Performance, Innovations and Structural Implications, Vancouver, 2005. [11] M.S. Konsta-Gdoutos, Z.S. Metaxa, S.P. Shah, Highly dispersed carbon nanotube reinforced cement based materials, Cem. Concr. Res. 40 (7) (2010) 1052–1059. [12] P. Stynoski, P. Mondal, C. Marsh, Effects of silica additives on fracture properties of carbon nanotube and carbon fiber reinforced Portland cement mortar, Cem. Concr. Compos. 55 (2015) 232–240. [13] A. Katz, A. Bentur, Effect of matrix composition on the aging of CFRC, Cem. Concr. Compos. 17 (2) (1995) 87–97. [14] G.Y. Li, P.M. Wang, X. Zhao, Mechanical behavior and microstructure of cement composites incorporating surface-treated multi-walled carbon nanotubes, Carbon 43 (6) (2005) 1239–1245. [15] M. Paiva, B. Zhou, K. Fernando, Y. Lin, J. Kennedy, Y.P. Sun, Mechanical and morphological characterization of polymercarbon nanocomposites from functionalized carbon nanotubes, Carbon 42 (14) (2004) 2849–2854. [16] R. Siddique, A. Mehta, Effect of carbon nanotubes on properties of cement mortars, Constr. Build. Mater. 50 (2014) 116–129. [17] Nanostructured & Amorphous Material Inc., [Online], Available: