Construction and Building Materials 98 (2015) 112–118

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

WS2 nanotube – Reinforced cement: Dispersion matters

http://dx.doi.org/10.1016/j.conbuildmat.2015.08.0850950-0618/� 2015 Elsevier Ltd. All rights reserved.

Abbreviations: BSE, Back-Scattered Electron; CNT, Carbon NanoTube; CNC,Cement Nano-Composites; EDS, Energy-Dispersive X-ray Spectroscopy; NM,Nano-Material; NT, nanotube; PC, Plain Cement paste; SEM, Scanning ElectronMicroscope; TEM, Transmission Electron Microscope; WS2NT, Tungsten di-SulfideNanoTube.⇑ Corresponding authors at: Chemical Engineering, Ben-Gurion University of the

Negev, 84105 Beer-Sheva, Israel.E-mail addresses: roeynadiv@gmail.com (R. Nadiv), oregev@bgu.ac.il (O. Regev).

Roey Nadiv a,⇑, Michael Shtein a,b, Alva Peled c, Oren Regev a,b,⇑aDepartment of Chemical Engineering, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israelb Ilse Katz Institute for Nanoscale Science and Technology, Ben-Gurion University of the Negev, Beer-Sheva 84105, IsraelcDepartment of Structural Engineering, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel

h i g h l i g h t s

� Tungsten di-Sulfide NanoTubes(WS2NTs) enhance the cementflexural strength by 74%.

� We developed a method to disperseWS2NT individually in cement paste.

� Optimal enhancement occurs atextremely low WS2NT concentration(0.15 wt%).

� WS2NTs inhibit crack propagation bybridging, and fail via pulloutmechanism.

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 30 March 2015Received in revised form 3 August 2015Accepted 9 August 2015

Keywords:CementDispersionFractographyMechanical propertiesNano-compositesNanotubes

a b s t r a c t

Nanotubes are considered as promising nano-reinforcement in cement-based materials. The main chal-lenge towards achieving a significant enhancement in cement properties is an effective dispersion of theagglomerated nanotubes. In this paper, we demonstrate a novel dispersion method of Tungsten di-SulfideNanoTubes (WS2NTs) that results in substantial flexural and compressive strength enhancements at opti-mal nanotube concentration as low as 0.15 wt%. The reinforcement by WS2NTs remains significant after avariety of curing processes, suggesting a genuine nanoscale reinforcing effect. Finally, by employing acomprehensive fractography we found that the WS2NTs inhibit crack propagation by bridging with apullout failure mechanism.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Cement, one of the most widely used composite materials, ischaracterized by high compressive strength on the one hand andby low tensile, flexural and fracture toughness properties on theother [1,2]. The latter properties are expected to improve by load-ing appropriate Nano-Materials (NMs) into the cement pastematrix [3].

A wide range of NMs were used in Cement Nano-Composites(CNC), including nano-silica [4–6], nano-titanium dioxide [4,7,8],

R. Nadiv et al. / Construction and Building Materials 98 (2015) 112–118 113

carbon nanofiber [9,10] and Carbon NanoTube (CNT) [10–18].Tungsten di-Sulfide NanoTube (WS2NT) [19,20], which possessesattractive mechanical and geometrical properties (Table 1)[21,22], may be a promising alternative for cement nano-reinforcement.

NMs tend to agglomerate due to strong interfacial van derWaals interactions [23–25], or to entangele due to their wavedstructure (e.g. CNT) [26,27]. Therefore, their effective surface areais reduced, leading to a decrease in stress transfer between thematrix and the NMs. Furthermore, NM agglomerate act as stressconcentrators and may initiate crack propagation [28]. As such, itis essential to develop a dispersion method yielding individualNMs. WS2NTs also tend to form agglomerates, however, theirstraight shape does not allow entanglements (in contrast to CNT)and, consequently, simplifies the dispersion process.

NMs may serve as nucleation sites in the cement hydrationreaction and accelerate its kinetics [29–31]. Such accelerationinsinuates a genuine, although short-term, mechanical propertiesenhancement in the early stages of the hydration. However, aftera longer hydration time (>10 d), this increase vanishes almost com-pletely [32]. As such, in order to observe signs of long-term nano-reinforcement, the mechanical properties should be evaluatedafter a longer period. Alternatively, hydration processes may beaccelerated by heating [1] and the outcomes of this process predictthe long-term mechanical properties of the CNC.

Recently, we have shown that WS2NTs can potentially serve asnano-reinforcement in cement systems [33]. We now aim atexpanding the study on the benefits of incorporating this nano-reinforcement. In this paper, we demonstrate a novel dispersionmethod, facilitating the integration of only individually dispersedWS2NTs. The mechanical properties of the produced CNCs are mea-sured and compared to the Plain Cement paste (PC) in a variety ofcuring ages, hence evaluating both WS2NT’s short and long termreinforcing effects. Concerning our methodology, we distinguishbetween structurally similar cement ettringites and the WS2NTsby employing both quantitative and qualitative techniques toallow an accurate evaluation of both the toughening and failuremechanisms of WS2NT-based CNCs.

2. Experimental section

2.1. Materials

Portland cement CEM I 52.5 R (Nesher Israel cement enterprises Ltd.), protein-based dispersant b-lactoglobulin (P90%, Sigma–Aldrich) and WS2NT (batch no.TWPO-MB023, received as a gift from NanoMaterials Ltd.), were used as received.

2.2. Specimen preparation

2.2.1. WS2NT dispersion [34]WS2NTs are mixed with deionized water (6.0 mg/ml) containing b-lactoglobulin

(2.0 mg/ml). The solution is bath-sonicated (Elma, model S10; 30 W, 37 kHz,Singen) for 30 min (540 J). The vial (20 ml) is placed at the center of the sonicatorand kept at 0 �C during the whole sonication process. To allow the precipitationof large agglomerates, a phase separation by decantation is conducted an hour afterthe sonication process. The WS2NT concentration in the supernatant is then calcu-lated using a combination of thermo-gravimetric and spectroscopic techniques[35]. The supernatant (exfoliated WS2NT) is freeze-dried (Lobanco Freezone 2.5)

Table 1Properties of WS2NT.

Property Value

Young’s modulus (TPa) 0.15–0.17 [21,22]Tensile strength (GPa) 19.6 [22]Diameter (nm)a 30–100Length (lm)a 1–4 lm

a Measured by electron microscopy (WS2NT counts = 180).

in a 40 ml plastic flask for 72 h. The product of which is a sticky–fluffy powder ofconcentrated WS2NTs wrapped in b-lactoglobulin dispersant. This product is aready to use additive in any cement preparation technology.

2.2.2. CNC preparationThe lyophilized WS2NTs are mixed in water and bath-sonicated for 2 min. The

cement is then gradually added and manually mixed into the solution using a spat-ula (water/cement ratio of 0.4). To reach a uniform dispersion, the mixture ismechanically mixed for 4 min (800 rpm, R50D overhead stirrer, CAT). Finally, thisCNC mixture is cast in silicone molds (specimen’s dimensions are 8 � 8 � 60 mm3

and 12 � 12 � 12 mm3 for flexural strength and compressive strength, respec-tively). The molds are placed inside a vibration machine (Lab Line Orbital Shaker)for 4 min to remove large air bubbles. The CNC samples are removed from themolds (24 h after casting) and cured in water vessel (33.7 mg/l calcium) in a main-tenance room (23 �C, 60% humidity) for either 14 or 28 days. An accelerated agingprocess is performed by immersing CNC samples (after 14 d of normal curing atroom temperature) in a hot bath (50 �C) for an additional 21 d, to evaluate theCNC durability and the WS2NT’s effect on the composite’s properties over time[36–38]. The effect of the ionic strength of the curing medium on the CNC perfor-mances has been studied by replacing the water curing medium by a lime saturatedone. CNC samples without dispersant (Section 2.2.1) were also prepared for control.

2.3. Characterization

2.3.1. Dispersion characterizationThe WS2NT dispersion quality is examined by the use of a Transmission

Electron Microscope (TEM) (FEI Tecnai 12 G2 TWIN TEM operated at 120 kV). TheTEM samples are prepared by placing a droplet of dispersion on carbon-coated cop-per grids (Ted Pella, lacey carbon, 300 mesh), followed by drying at 80 �C for 2 hbefore TEM examination.

2.3.2. Measurements of mechanical propertiesThe flexural strength of the WS2NT-based CNC and PC is determined by per-

forming a three-point bending test, using prism-shaped specimens. These measure-ments are performed by a LRX, LLOYD instrument (capacity of 5 kN) at a constantextension rate of 0.5 mm/min (>5 specimens for each wt%). The compressivestrength is determined by compression measurements of cube specimens, per-formed by an Instron 5982 instrument (capacity of 100 kN), with a constant exten-sion rate of 2 mm/min (>5 specimens for each wt%).

2.3.3. Fractographic characterizationTo understand the role of WS2NT in both toughening and failure mechanisms of

WS2NT-based CNC, it is essential to distinguish between WS2NT and other cementcomponents with similar morphologies.

A fractographic study is carried out on the specimens’ fractured surfaces using ahigh-resolution field-emission gun-SEM (JEOL, JSM-7400F) equipped with Energy-Dispersive X-ray Spectroscopy (EDS) instrument (Noran Vantage) and a Back-Scattered Electron (BSE) detector (AutraDet, AUTORATA YAG) operated at 20 keV.The samples are Pt-coated (few Ångstroms) with Sputter Coater (Emitech K575X).The WS2NTs are qualitatively identified (1) using BSE imaging mode, whichincreases mass contrast, since heavy elements (e.g., tungsten Z = 74) backscatterelectrons stronger than lighter elements in the matrix (e.g., calcium Z = 20). Thus,in BSE imaging mode, WS2NT appears brighter than ettringite in comparison withsecondary electron imaging mode (Fig. 1).

(2) WS2NT and ettringite have different elemental compositions. As such, whena suspected WS2NT is found (Fig. 2a), tungsten identification by EDS also confirmsits presence (white coloration in Fig. 2b).

The EDS elemental spectrum provides a sulfide-to-tungsten atom ratio of �2, inagreement with the WS2 composition (Fig. 2c) as opposed to spectrum of CNC with-out WS2NT, which does not show evidence of tungsten atoms (Fig. 2d).

3. Results and discussion

3.1. Dispersion method and characterization

As stated above, the major challenge in achieving a significantproperties enhancement of CNC by means of nanotubes (NTs)nano-reinforcement is their efficient dispersion. Therefore, a noveldispersion method, based on sonication, decantation andlyophilization is implemented (Section 2.2), yielding only individ-ually dispersed WS2NTs in the cement matrix.

The WS2NTs’ structural integrity and dispersion quality weremonitored by electron microscopy throughout the dispersion pro-cedure to verify that only individual and defect-free NTs are inte-grated into the CNC. The as-received WS2NTs (Fig. 3a) are in anagglomerated state (�0.1 mm in diameter), which calls for bothsonication and decantation steps. The individually dispersed

Fig. 1. SEM micrographs of WS2NT-based CNC demonstrating the imaging mode effect on WS2NT detection: (a) Secondary electron mode; (b) BSE mode. WS2NT location isindicated by arrows.

Fig. 2. Analysis of WS2NT-based CNC fractured surfaces: (a) SEM micrograph of three WS2NTs (white arrows) in the cement matrix; (b) EDS elemental mapping: W, S, Sielements in green, purple and blue, respectively. Right panel: elements’ superposition. The white color indicates WS2NTs; (c) EDS spectrum of the marked area in (b)indicating a 1:2 sulfide-to-tungsten ratio; (d) EDS spectrum of CNC without WS2NT. (For interpretation of the references to color in this figure legend, the reader is referred tothe web version of this article.)

114 R. Nadiv et al. / Construction and Building Materials 98 (2015) 112–118

WS2NTs (Fig. 3b, left), are defect-free (Fig. 3b, right). Thelyophilization yields WS2NTs coated by the dispersant(b-lactoglobulin, dotted arrows in Fig. 3c). Finally, the SEM imageof the WS2NT-loaded CNC fractured specimen indicates that indi-vidual WS2NTs are well-dispersed in the CNC (Fig. 3d).

3.2. Mechanical properties

3.2.1. WS2NT concentration effectFirst, we aim at exploring the effect of WS2NT concentration on

the mechanical performance of the CNC. Therefore, both flexural

and compressive strengths were measured as a function ofWS2NT wt% (Fig. 4a) (14 d of curing).

The concentration of the WS2NT has a major effect on the prop-erties of the resulting CNC. At low nano-reinforcement loading, theflexural strength increases with WS2NT concentration (full dia-monds, Fig. 4a) up to an optimal NT concentration of 0.15 wt%, inwhich an enhancement of 74% (�6 MPa increase) over the PC isachieved (14 d of curing). Above this optimal concentration, theflexural strength drastically decreases by �30% with respect toits maximal value and remains unchanged with further increasein NT concentration. Similar trend was also observed for the

Fig. 3. Micrographs of WS2NTs along the dispersion method: (a) SEM image of the as-received agglomerated WS2NTs; (b) TEM: left – individually dispersed WS2NT; right:defect-free structure (marked area in the left image); (c) SEM image of lyophilized WS2NT coated with b-lactoglobulin (dotted arrows); (d) SEM image of 0.15 wt% WS2NT-based CNC fractured specimen, indicating well-dispersed WS2NTs (full arrows) as well as ettringites (dashed arrows-see Section 2.3.3).

Fig. 4. (a) Flexural (full diamonds) and compressive (full squares) strengths of CNC following 14 d curing as a function of WS2NT wt%, indicating an optimal concentration at0.15 wt%. The lines are only visual guides. Hollow markers indicate control samples prepared without dispersant. The starred samples are imaged in Figs. 3d(*) and 4b(**); (b)SEM micrograph of 0.5 wt% WS2NT-based CNC fractured surfaces indicating agglomerated WS2NT near pores (full arrow).

R. Nadiv et al. / Construction and Building Materials 98 (2015) 112–118 115

CNCs under compression (full square, Fig. 4a). We found that, atthe optimal NT concentration (0.15 wt%), the compressive strengthincreases by 29% with respect to the PC, whereas, at a higher NTconcentration (0.5 wt%), a decrease of 19% in compressive strengthwas observed (Fig. 4a).

Interestingly, at the optimal NT concentration (0.15 wt%), bothcompressive and flexural strengths reach their maximal values.In other words, a single nano-reinforcement agent, WS2NT, pro-vides dual enhancements, namely, compressive and flexuralstrengths, at low concentration.

A possible interpretation to this phenomenon relates to the NTarrangement within the matrix. Below the optimal NT concentra-tion, only a mild mechanical properties enhancement is achieved

(Fig. 4a), since a large volume in the matrix is not occupied byany NTs. With increasing NT concentration, the volume of NT-free matrix diminishes, while at the optimal NT concentration(Fig. 3d), the CNC is loaded by well-dispersed NTs, which effec-tively reinforce the matrix and lead to a maximal NT reinforce-ment. Above the optimal NT concentration, the embeddedindividual NTs re-agglomerate and interconnect, hence resistingdeformation [39]; this, in turn, decreases the mixture’s workability,as manifested by the large pores in the resulting CNC (Fig. 4b,Table 3). Moreover, the NT agglomerate may be considered aspoints of stress concentration rather than reinforcing units andhence lead to lower mechanical performance. A similar trend ofan abrupt increase in strength at the optimal NT concentration,

116 R. Nadiv et al. / Construction and Building Materials 98 (2015) 112–118

was also observed in NT-based polymer nano-composites [34], andis considered as a general nanocomposite phenomenon rather thanmatrix or property related.

Finally, the critical role of the dispersant (Section 2.2.2) is stud-ied by control samples in which the dispersant is excluded (Fig. 4a,hollow markers). These samples demonstrate milder enhance-ments compared to dispersant-assisted preparation of CNC.

Fig. 6. The effect of curing medium on the flexural (diamond) and compressive(square) strengths of PC (solid fill) and 0.15 wt% WS2NT-based CNC (line fill) at 14 dof curing in water or lime saturated water media. The error bars of flexural strengthmeasurements assimilate into the markers due to small statistical errors.

Table 2b-Lactoglobulin’s effect on the flow and mechanical properties of NT-free cementpaste after 14 d of curing in water.

Property PC b-Lactoglobulin-based cementpaste

Flow (%) 106 126Flexural strength (MPa) 8.1 ± 0.5 9.0 ± 0.7Compressive strength

(MPa)39.8 ± 2.9 40.6 ± 0.9

3.2.2. Curing effectFollowing the determination of the optimal NT concentration

(0.15 wt%) after 14 d curing in water (Fig. 4a), we studied the effectof prolonging the curing time to 28 d (Fig. 5). Interestingly, while itseems that the relative flexural strength enhancement diminishes(43% after 28 d curing, as opposed to 74% after 14 d curing,Fig. 5), the absolute strength contribution of the WS2NT to thematrix is not altered (�6 MPa in both cases), insinuating a genuinereinforcing effect, regardless of the progress in the hydrationprocess.

The long-term reinforcement efficiency and durability ofWS2NT-based CNCs is also explored by studying the effect of accel-erated aging, i.e., 14 d curing at room temperature followed by 21 dof accelerated curing at 50 �C [37] (Fig. 5, acc.). After acceleratedaging, the WS2NT-based CNC enhances the flexural strength by35% (�4.6 MPa increase) with respect to the accelerated cured PC.This clearly suggests that WS2NTs offer genuine long-term nanos-cale reinforcement of PC. The nano-reinforcing effect is indepen-dent of various experimental parameters including curingduration (Fig. 5) or curing medium (lime-saturated water, Fig. 6).

Finally, control experiments, evaluating the effect of the disper-sant (b-lactoglobulin) alone (without NT) were made, indicatingthat the dispersant acts as a superplasticizer in the cement matrix,increasing the flow by 20% (ASTM C1437, Table 2). The dispersanthas only little effect on the mechanical properties of the CNC(Table 2), verifying that indeed the observed reinforcing effect isonly the outcome of the incorporation of WS2NTs.

Our flexural strength enhancement (Fig. 5) is among the highestreported so far [11–13,15,16], and most probably stems from theefficiency of the dispersion method, yielding highly dispersedindividual WS2NTs at 0.15 wt% (Fig. 3d).

Fig. 5. Flexural strength of 0.15 wt% WS2NT-based CNC (line fill) and PC (solid fill)following 14 d, 28 d and 14 d + 21 d accelerated (acc.) curing; (percent ofenhancement (red) over PC under the same ageing is indicated for each CNC).(For interpretation of the references to color in this figure legend, the reader isreferred to the web version of this article.)

3.3. Failure and toughening mechanisms

A NTmay fail in two main scenarios, according to the traditionalfiber-matrix failure mechanisms [40]: either via fiber pullout fromthe matrix or through fiber fracture, in weak or strong NT-matrixinterfacial adhesion, respectively. The reported fractographic stud-ies of NT-based CNC have focused mainly on the bridging mecha-nism [14,15,32], in which the tough NTs inhibit crackpropagation by holding the matrix together. Nevertheless, in theevent of complete failure, the NTs will either be pulled out orfractured.

The porous structure of the cement paste matrix insinuates lowinterfacial area and adhesion; hence, a NT pullout mechanism isexpected.

The identified WS2NTs (Section 2.3.3 and Fig. 7a) indicate thatthey indeed reinforce the matrix via a bridging mechanism, hold-ing cracks in the cement at both ends. Eventually, when the spec-imen completely fails, the WS2NTs exhibit a pullout mechanism,manifested by a substantial part of the NT protruding from thespecimen surface (Fig. 7b).

One can also learn about the failure mechanism by statisticallymeasuring the length of the protruding NT from the matrix [34,41](Fig. 7b). A very short protrusion indicates a fracture mechanism,inwhich theNT remains in thehostingmatrix,while a longerprotru-sion indicates a pullout failure mechanism. A statistical analysis ofthe initial length of the NTs and their protrusion length after thespecimens’ failure is given in Fig. 8 (based on 545 NT counts), indi-cating a clear pulloutmechanism,with an averageprotruding lengththat is approximately equal to half of the initial WS2NT length

Fig. 7. BSE–SEM images of 0.15 wt% WS2NT-based CNC fractured surfaces, indicating (a) NT bridging mechanism and (b) NT pullout mechanism, the protrusion length isindicated; the EDS spectra of the marked areas are in line with Fig. 2c.

Fig. 8. Statistical analysis of WS2NT protruding length: relative frequency of WS2NTinitial and protrusion lengths. The protrusion length is about half of the originallength indicating a pullout failure mechanism. N = number of NTs counted.

R. Nadiv et al. / Construction and Building Materials 98 (2015) 112–118 117

(2.36 ± 0.62 lm) after 14 days of curing (1.24 ± 0.67 lm). The pull-outmechanism is independent of the hydration period, with similaraveragepullout lengths following28 d (1.19 ± 0.73 lm)andacceler-ated curing processes (0.94 ± 0.68 lm) (Fig. 8).

Given these results and based on prior knowledge regarding theNTs failure mechanism in polymer composites [34,42], it is clearthat, in order to further increase the reinforcing effect, it is vitalto increase the NT-cement interfacial adhesion. We plan to explorethe interfacial adhesion between WS2NT and the cement pastematrix by preforming pullout measurements using atomic forcemicroscopy [43–45].

The role of WS2NT as a filler [11,46] is explored by measuringthe porosity by Archimedes’ method (ASTM C948) (Table 3). Thecomplex nature of the cement matrix results in only minordecrease (1.7%) in the apparent porosity up to the optimalWS2NT concentration (0.15 wt%), i.e., the void-filling mechanism

Table 3CNC apparent porosity as a function of WS2NT wt%. In Bold, the optimal WS2NT concentr

WS2NT (wt%) 0 0.1 0

Apparent porosity (%) 31.8 ± 0.6 31.3 ± 0.6 3

[11,46] is not dominant in the cement reinforcement by WS2NT.The porosity increases above 0.15 wt% agrees with the aforemen-tioned pores formation (Fig. 4b).

The low NT concentration required to achieve such long-termflexural strength enhancement demonstrates the potential ofWS2NT to serve as nano-reinforcement in cement-based systems.Its attractiveness is further increased since it reinforces the matrix,rather than inhibiting catastrophic failure (e.g., fibers) [47–49].Consequently, WS2NT may be combined with fiber reinforcement,producing even stronger composites.

4. Conclusions

In this paper, we employed WS2NTs to reinforce cement sys-tems. Using a novel dispersion method, we were able to integrateonly individually dispersed WS2NTs, achieving substantialmechanical properties enhancements (over PC) following a varietyof hydration processes, suggesting genuine nanoscale reinforce-ment by WS2NTs. The loading of WS2NTs also enhances the com-pressive strength of the CNC. These properties enhancementswere obtained by adding only a very small concentration ofWS2NT (0.15 wt%), making the WS2NTs an attractive nano-reinforcement for cement systems.

By employing a combined spectroscopic-scattering fractogra-phy, we were able to observe the inhibition of crack propagationby WS2NT bridging and failure via pullout mechanism.

References

[1] H.F. Taylor, Cement Chemistry, Thomas Telford, 1997.[2] G.C. Bye, Portland Cement: Composition, Production and Properties, Thomas

Telford, 1999.[3] F. Sanchez, K. Sobolev, Nanotechnology in concrete – a review, Constr. Build.

Mater. 24 (11) (2010) 2060–2071.[4] H. Li, M.-h. Zhang, J.-p. Ou, Abrasion resistance of concrete containing nano-

particles for pavement, Wear 260 (11–12) (2006) 1262–1266.[5] J.-Y. Shih, T.-P. Chang, T.-C. Hsiao, Effect of nanosilica on characterization of

Portland cement composite, Mater. Sci. Eng., A 424 (1–2) (2006) 266–274.[6] K. Sobolev, I. Flores, L.M. Torres-Martinez, P.L. Valdez, E. Zarazua, E.L. Cuellar,

Engineering of SiO2 nanoparticles for optimal performance in nano cement-based materials, in: Z. Bittnar, P.M. Bartos, J. Nemecek, V. Šmilauer, J. Zeman(Eds.), Nanotechnology in Construction 3, Springer, Berlin Heidelberg, 2009,pp. 139–148.

ation.

.15 0.2 0.3 0.5

0.1 ± 0.3 30.6 ± 0.8 31.5 ± 0.5 34.0 ± 0.4

118 R. Nadiv et al. / Construction and Building Materials 98 (2015) 112–118

[7] C. Lan, H. Li, Y. Ju, Ductility of High Strength Concrete Containing Nano-particles, (2009) p. 74932F-F-6.

[8] A.R. Jayapalan, B.Y. Lee, S.M. Fredrich, K.E. Kurtis, Influence of additions ofanatase TiO2 nanoparticles on early-age properties of cement-based materials,Transport. Res. Record: J. Transport. Res. Board 2141 (1) (2010) 41–46.

[9] G. Di, S. Mariel, Y.L. Mo, Electrical resistance of carbon–nanofiber concrete,Smart Mater. Struct. 18 (9) (2009) 095039.

[10] R. Abu Al-Rub, B. Tyson, A. Yazdanbakhsh, Z. Grasley, Mechanical properties ofnanocomposite cement incorporating surface-treated and untreated carbonnanotubes and carbon nanofibers, J. Nanomech. Micromech. 2 (1) (2011) 1–6.

[11] G.Y. Li, P.M. Wang, X. Zhao, Mechanical behavior and microstructure of cementcomposites incorporating surface-treated multi-walled carbon nanotubes,Carbon 43 (6) (2005) 1239–1245.

[12] J. Luo, Z. Duan, H. Li, The influence of surfactants on the processing of multi-walled carbon nanotubes in reinforced cement matrix composites, Phys. StatusSolidi A 206 (12) (2009) 2783–2790.

[13] S. Musso, J.-M. Tulliani, G. Ferro, A. Tagliaferro, Influence of carbon nanotubesstructure on the mechanical behavior of cement composites, Compos. Sci.Technol. 69 (11–12) (2009) 1985–1990.

[14] S. Kumar, P. Kolay, S. Malla, S. Mishra, Effect of multiwalled carbon nanotubeson mechanical strength of cement paste, J. Mater. Civ. Eng. 24 (1) (2012) 84–91.

[15] A. Cwirzen, K. Habermehl-Cwirzen, V. Penttala, Surface decoration of carbonnanotubes and mechanical properties of cement/carbon nanotube composites,Adv. Cem. Res. 20 (2) (2008) 65–73.

[16] M.S. Konsta-Gdoutos, Z.S. Metaxa, S.P. Shah, Multi-scale mechanical andfracture characteristics and early-age strain capacity of high performancecarbon nanotube/cement nanocomposites, Cem. Concr. Compos. 32 (2) (2010)110–115.

[17] J. Li, P.C. Ma, W.S. Chow, C.K. To, B.Z. Tang, J.K. Kim, Correlations betweenpercolation threshold, dispersion state, and aspect ratio of carbon nanotubes,Adv. Funct. Mater. 17 (16) (2007) 3207–3215.

[18] A. Sobolkina, V. Mechtcherine, V. Khavrus, D. Maier, M. Mende, M. Ritschel,et al., Dispersion of carbon nanotubes and its influence on the mechanicalproperties of the cement matrix, Cem. Concr. Compos. 34 (10) (2012) 1104–1113.

[19] R. Tenne, L. Margulis, M. Genut, G. Hodes, Polyhedral and cylindrical structuresof tungsten disulphide, Nature 360 (6403) (1992) 444–446.

[20] R. Tenne, G. Seifert, Recent progress in the study of inorganic nanotubes andfullerene-like structures, Annu. Rev. Mater. Res. 39 (2009) 387–413.

[21] I. Kaplan-Ashiri, S.R. Cohen, K. Gartsman, V. Ivanovskaya, T. Heine, G. Seifert,et al., On the mechanical behavior of WS2 nanotubes under axial tension andcompression, Proc. Natl. Acad. Sci. U.S.A. 103 (3) (2006) 523–528.

[22] D.-M. Tang, X. Wei, M.-S. Wang, N. Kawamoto, Y. Bando, C. Zhi, et al., Revealingthe anomalous tensile properties of WS2 nanotubes by in situ transmissionelectron microscopy, Nano Lett. 13 (3) (2013) 1034–1040.

[23] J. Jiang, G. Oberdörster, P. Biswas, Characterization of size, surface charge, andagglomeration state of nanoparticle dispersions for toxicological studies, J.Nanopart. Res. 11 (1) (2009) 77–89.

[24] L.X. Benedict, N.G. Chopra, M.L. Cohen, A. Zettl, S.G. Louie, V.H. Crespi,Microscopic determination of the interlayer binding energy in graphite, Chem.Phys. Lett. 286 (5–6) (1998) 490–496.

[25] L.A. Girifalco, M. Hodak, R.S. Lee, Carbon nanotubes, buckyballs, ropes, and auniversal graphitic potential, Phys. Rev. B 62 (19) (2000) 13104–13110.

[26] D. Weidt, Ł. Figiel, Effect of CNT waviness and van der Waals interaction on thenonlinear compressive behaviour of epoxy/CNT nanocomposites, Compos. Sci.Technol. 115 (2015) 52–59.

[27] U.A. Joshi, S.C. Sharma, S.P. Harsha, Effect of waviness on the mechanicalproperties of carbon nanotube based composites, Physica E 43 (8) (2011)1453–1460.

[28] L. Vaisman, H.D. Wagner, G. Marom, The role of surfactants in dispersion ofcarbon nanotubes, Adv. Colloid Interface Sci. 128–130 (2006) 37–46.

[29] J.M. Makar, G.W. Chan, Growth of cement hydration products on single-walledcarbon nanotubes, J. Am. Ceram. Soc. 92 (6) (2009) 1303–1310.

[30] J. Björnström, A. Martinelli, A. Matic, L. Börjesson, I. Panas, Accelerating effectsof colloidal nano-silica for beneficial calcium–silicate–hydrate formation incement, Chem. Phys. Lett. 392 (1–3) (2004) 242–248.

[31] A.R. Jayapalan, B.Y. Lee, K.E. Kurtis, Effect of nano-sized titanium dioxide onearly age hydration of Portland cement, in: Z. Bittnar, P.M. Bartos, J. Nemecek,V. Šmilauer, J. Zeman (Eds.), Nanotechnology in Construction 3, BerlinHeidelberg, Springer, 2009, pp. 267–273.

[32] J.M. Makar, M.J., J. Luh, Carbon nanotube/cement composite – early results andpotential applications, 3rd International Conference on ConstructionMaterials: Performance, Innovation and Structural Implications, Vancouver,BC, (2005) pp. 1–10.

[33] R. Nadiv, M. Shtein, A. Peled, O. Regev, Cement reinforcement by nanotubes, in:K. Sobolev, S.P. Shah (Eds.), Nanotechnology in Construction, SpringerInternational Publishing, 2015, pp. 231–237.

[34] M. Shtein, R. Nadiv, N. Lachman, H. Daniel Wagner, O. Regev, Fracture behaviorof nanotube–polymer composites: insights on surface roughness and failuremechanism, Compos. Sci. Technol. 87 (2013) 157–163.

[35] M. Shtein, I. Pri-bar, O. Regev, A simple solution for the determination ofpristine carbon nanotube concentration, Analyst 138 (5) (2013) 1490–1496.

[36] K.L. Litherland, D.R. Oakley, B.A. Proctor, The use of accelerated ageingprocedures to predict the long term strength of GRC composites, Cem. Concr.Res. 11 (3) (1981) 455–466.

[37] Z. Cohen, A. Peled, Controlled telescopic reinforcement system of fabric–cement composites — durability concerns, Cem. Concr. Res. 40 (10) (2010)1495–1506.

[38] Z. Cohen, A. Peled, Effect of nanofillers and production methods to control theinterfacial characteristics of glass bundles in textile fabric cement-basedcomposites, Compos. A 43 (6) (2012) 962–972.

[39] S.S. Rahatekar, K.K.K. Koziol, S.A. Butler, J.A. Elliott, M.S.P. Shaffer, M.R.Mackley, et al., Optical microstructure and viscosity enhancement for an epoxyresin matrix containing multiwall carbon nanotubes, J. Rheol. (1978–present)50 (5) (2006) 599–610.

[40] D. Hull, T. Clyne, An Introduction to Composite Materials, CambridgeUniversity Press, 1996.

[41] L.-c. Tang, H. Zhang, X.-p. Wu, Z. Zhang, A novel failure analysis of multi-walled carbon nanotubes in epoxy matrix, Polymer 52 (9) (2011) 2070–2074.

[42] N. Lachman, Wagner H. Daniel, Correlation between interfacial molecularstructure and mechanics in CNT/epoxy nano-composites, Compos. A 41 (9)(2010) 1093–1098.

[43] A.H. Barber, S.R. Cohen, H.D. Wagner, Measurement of carbon nanotube–polymer interfacial strength, Appl. Phys. Lett. 82 (23) (2003) 4140–4142.

[44] A.H. Barber, S.R. Cohen, S. Kenig, H.D. Wagner, Interfacial fracture energymeasurements for multi-walled carbon nanotubes pulled from a polymermatrix, Compos. Sci. Technol. 64 (15) (2004) 2283–2289.

[45] A.H. Barber, S.R. Cohen, A. Eitan, L.S. Schadler, H.D. Wagner, Fracturetransitions at a carbon-nanotube/polymer interface, Adv. Mater. 18 (1)(2006) 83–87.

[46] A. Chaipanich, T. Nochaiya, W. Wongkeo, P. Torkittikul, Compressive strengthand microstructure of carbon nanotubes–fly ash cement composites, Mater.Sci. Eng. A 527 (4–5) (2010) 1063–1067.

[47] Y. Akkaya, A. Peled, S.P. Shah, Parameters related to fiber length and processingin cementitious composites, Mater. Struct. 33 (8) (2000) 515–524.

[48] A. Peled, B. Mobasher, Tensile behavior of fabric cement-based composites:pultruded and cast, J. Mater. Civ. Eng. 19 (4) (2007) 340–348.

[49] A. Peled, A. Bentur, Fabric structure and its reinforcing efficiency in textilereinforced cement composites, Compos. A 34 (2) (2003) 107–118.