Effect of nanosilica and montmorillonite nanoclay particles on cement hydration and microstructure

Abstract

The need to produce sustainable cements has driven research towards nanotechnology. The main cement hydration product, calcium silicate hydrate, is nanosized; hence, the addition of nanoparticles to blended Portland cement formulations can remarkably modify mechanical strength, porosity and durability. The present paper discusses the material aspects of two different nanoparticles, nanosilica and montmorillonite nanoclay, the complications that arise from their addition to cement pastes and ways to mitigate these limitations. It is deduced that nanosilica solids in blended cement pastes should be limited to 0.5, whereas nanoclay solids to almost 1 mass- binder. Competitive reactivity of nanoparticles with other constituents is expected, and the possible pozzolanic activity is critically addressed. Notwithstanding progress made, there are significant potentials related to inorganic nanoclays.

Keywords

Nanosilica Montmorillonite nanoclay Pozzolanic reaction Optimum dosage Cement hydration Microstructure Blended cements

Introduction

Cement production has raised environmental concerns with regards to the amount of CO₂ emitted during Portland cement (PC) clinkering. The total contribution of the cement industry to the global manmade CO₂ emission reaches 8, not taking into account the energy consumed for the rest of the industrial processes.^1–3 It has, therefore, become a priority to develop cements in which PC clinker is limited to the maximum, primarily by substituting it with materials, such as fillers, for instance limestone (LS), and/or pozzolanic byproducts, for example fly ash (FA) or microsilica (μS).

The need for sustainability in construction is calling for more durable and more ecofriendly concretes and consequently cements. Knowing the application for which a concrete is intended (exposure conditions, use, availability of raw materials and others), the cement paste, which is a highly engineered material, can be designed to meet specific requirements in terms of properties and performance. It can, therefore, become ‘greener’ not only through the process of replacing PC by supplementary cementitious materials but also by creating a novel blend of constituents that improves the durability and the life span of structures. In this way, cement science can provide the bridge between structural design and protection of natural resources especially taking into consideration the worldwide concern with respect to the deterioration of concrete structures and the need to preserve them for longer periods. This requirement has sparked research towards understanding the internal structure of cement and investigating the modifications applicable in the submicrometre to nanolevel. In fact, the addition of nanomaterials in cement formulations opens a new horizon to the research and industry, to the extent that many experts consider nanotechnology as bringing about the next industrial revolution.⁴

The field of science focusing on the study, experimentation and observation of phenomena and modification of materials at atomic and molecular level is known as ‘nanoscience’, whereas the design, characterisation, manufacturing and application of materials limited to the nanosize is termed ‘nanotechnology’.⁵ As for the definition of nanomaterial or nanoparticles, the European Commission in 2011 specified that ‘“Nanomaterial” means a natural, incidental or manufactured material containing particles, in an unbound state or as an aggregate or as an agglomerate and where, for 50 or more of the particles in the number size distribution, one or more external dimensions is in the size range 1 nm–100 nm. In specific cases and where warranted by concerns for the environment, health, safety or competitiveness the number size distribution threshold of 50 may be replaced by a threshold between 1 and 50’ (Ref. 6, p. 40).

Nanoparticles by definition lie in the area below 100 nm and should not be confused with other cementitious compounds, such as silica fume (100–1100 nm)^7,8 (also known as μS), metakaolin (800–5000 nm)⁷ or FA (>1100 nm),^9,10 which belong to different scale ranges, therefore exhibit inherently different properties and have been extensively studied in the past.

With the evolution of nanotechnology, researchers have gained more insight into the parameters affecting the products of cement hydration, the structure of which is nanosized. Therefore, one of the most stimulating questions to be answered is, what effect would the addition of nanoparticles have on the hydration of cement? Are some nanoparticles pozzolanic materials in their own right and therefore produce additional calcium silicate hydrate (C–S–H) responsible for the mechanical strength and the durability of the paste, or do they primarily act as nucleation centres or as catalysts for cement hydration? This subsequently raises more questions, such as, how does the hydration chemistry change, how are the mechanical properties modified and what would, then, be the optimal proportions in such novel blended cement formulations?

The present paper clarifies the state of the art on these matters, the contradictions and omissions related to the addition of nanoparticles of silica or montmorillonite (MMT) nanoclay (nC) in cement formulations. A short reference to the chemistry and products of cement hydration, as well as the microstructure of cement, is regarded necessary to facilitate the understanding of the effect of the specific nanoparticles on the properties of cement.

Portland cement hydration

When the two major compounds of cement, that is to say the silicate phases (C₃S and C₂S), react with water, two very significant products are formed; crystalline calcium hydroxide and an almost amorphous C–S–H, which comprises over 60 of the hydration products in total. The hyphens imply that the stoichiometry is not constant. The specific surface area of C–S–H ranges between 7 × 10⁷ and 10⁸ m² g^− 1 of dry paste measured by the small angle scattering method,¹¹ and for this, it is usually described as gel. However, there are many reasons for calling C–S–H a gel, and surface area is not the main one. Under X-ray diffraction (XRD), C–S–H is amorphous, which in fact could constitute another criterion, as discussed later. Furthermore, the setting and hardening of cement is primarily determined by the physical properties of the C–S–H. In particular, the setting time and the early strength development of cement are attributed to the hydration of alite, whereas the strength development up to 1 year is related to the hydration of belite.¹² It should be noted that the hydration of these two major compounds, the silicate phases, is often considered individually, for simplicity. Effectively, strength, shrinkage and durability of the hardened cement paste can be attributed to the C–S–H.^12–14

The chemical formulas expressing the hydration of alite and belite can only be approximated respectively as¹⁵ 1 2 where x and y are not necessarily integers, C = CaO, Si = SiO₂, H = H₂O, CH = Ca(OH)₂ and x = C/S.

Apart from C–S–H, the second main hydration product is calcium hydroxide [Ca(OH)₂]. Owing to the presence of impurities during its production, the mineralogical term, portlandite, is often adopted. The monitoring of the portlandite level is significant, as it can yield information on the reactivity of pozzolanas.¹⁶

The hydration of belite produces a C–S–H, different to the one produced by the hydration of alite in terms of the C/S ratio and the chemically bound water.¹⁷

The molar ratio of calcium to silicon (C/S) is one of the most determinant parameters for the structure of C–S–H. This ratio ranges from 0.7 to 2.1. The higher values may be obtained under extreme hydration or curing conditions and the lower values in presence of supplementary cementitious materials.¹⁸ Calcium silicate hydrate has been historically classified according to the C/S value. In fact, Taylor¹⁹ suggested an average value of 1.5, dividing C–S–H in two categories: C–S–H (I), when C/S < 1.5, and C–S–H (II), when C/S>1.5. Nonat and Lecoq²⁰ further divided the first category into two: C–S–H (α) for C/S < 1.0 and C–S–H (β) for 1 < C/S < 1.5. The C/S is affected by the clinker components and their particle size distribution, degree of hydration of PC (and curing conditions), age of the paste¹³ and addition of supplementary cementitious materials.¹⁸ A reduction in C/S causes an increase in the mean length of the silicate chains and the interlayer distance, affecting C–S–H's structural morphology.

The products of cement hydration are summarised in Table 1.

Table 1

Main cement hydration products¹⁴

Abbreviation	Chemical formula of compound	Name or mineral phase
CH	Ca(OH)₂ or CaO·H₂O	Calcium hydroxide/portlandite
C–S–H	2(CaO)·SiO₂·0·9–1·25(H₂O), and/or 0·8–1·5(CaO)·SiO₂·1·0–2·5(H₂O)	Calcium silicate hydrate
C–A–H	Not clearly defined	Calcium aluminate hydrate
AFt	C₃AŠ₃H_30–32	Aluminate ferrite trisulphate/ettringite
AFm	C₂AŠH₁₂	Aluminate ferrite monosulphate/monosulphate
C₃AH₆	3CaO·Al₂O₃·6H₂O	Hydrogarnet

Taking into consideration the perspective that the way to enhance the macroscopic mechanical properties of cement is by understanding and modifying the structure of the C–S–H gel at the nanolevel,²¹ the addition of nanoparticles can fundamentally modify the nanostructure of the hydration products. Therefore, it would be interesting to the reader to go one level down and acquire some insight into the nanostructure of C–S–H, which will help to lay the path for the understanding of the mechanisms affecting it when nanoparticles are added in the different cement formulations. Analytical information can be traced in Papatzani et al.,²² Fonseca et al.²³ or Selvam et al.²¹

As discussed in these publications, C–S–H is nanosized, and within the hydrated cement paste, apart from the nanostructured solids, there are also nanopores filled with water and nano air voids. Hence, nanotechnology can offer a twofold advantage by providing nanosized constituent materials and by allowing the study of the chemical and structural modifications induced by the addition of nanomaterials at the nanometric scale range.

Effect of pozzolanic materials on cement hydration

The effect of the addition of pozzolanic particles in cement pastes has been studied through a series of tests, such as heat of hydration, compressive strength, shrinkage, frost resistance and microstructural characterisation and has been found to accelerate the hydration of cement, through the creation of nucleation sites assisting the depletion of Ca and Si in the pore solution.²⁴

During the process of hydration, pozzolanas consume one of the hydration products, portlandite, producing supplementary C–S–H or C–A–H or calcium alumina silicate hydrates.^25,26 The consumption of CH and alkalis causes the lowering of the pH of the paste. There is a maximum proportion of μS that is allowed to be added in the mix, based on the quantity of Ca(OH)₂ available. In broader terms, the hydration of C₃A is delayed, whereas the hydration of C₃S is accelerated. The chemical formulas representing the reactions taking place with the addition of μS in the paste are as follows²⁴ 3

The secondary C–S–H (2) produced by the reaction of additional μS with the primary C–S–H (1) exhibits a lower C/S. According to Vogt, the higher the μS content, the slower the compressive strength gain, since at high concentrations, there will not be enough Ca(OH)₂ to react with μS. Adding to this, higher shrinkage and autogeneous shrinkage values are also expected for higher proportions of μS. On the positive side, packing density is improved, enhancing frost resistance. It has also been observed that, with the addition of μS, the capillary porosity is reduced because the additional C–S–H produced by the pozzolanic reaction is much denser.²⁷

The effect of the addition of FA depends on the amorphous phase of FA; in general, the reaction is not as fast as the one with μS. The particle size distribution and the exact oxide composition of FA has a significant influence on the hydration procedure. Grains >10 μm can act as fillers, while those < 10 μm can accelerate the hydration of C₃S. Adding to this, Al₂O₃ and SiO₂ present in FA also participate in the pozzolanic reaction.

Studies have been presented with the addition of a combination of FA and μS particles to PC: addition of the FA with the aim to address consistence issues when μS is present, and addition of μS with the aim to counterbalance the low early strength gain and total porosity induced by FA,²⁸ as compared by Supit and Shaikh.²⁹ Their results indicate that the introduction of even finer particles, i.e. nanoparticles, could be even more promising.

Cement paste microstructure

Having touched upon the topic of hydration products, a short reference to the microstructure of the hardened cement paste is necessary. The hydrated cement paste consists of solids, voids and water. In Table 1, the solid hydration products are listed. Residual unhydrated cement should also be regarded in the solids of the paste. The strength of the paste is highly affected by the amount of pores present and their distribution. As a matter of fact, cylinder compressive strength f_c and porosity p are inversely related³⁰ 5 where k is the strength of voidless mortar (∼234 MPa).

It is acknowledged that a number of different definitions exist with respect to the pore sizes and characteristics. The water filled pores in the fresh paste are called capillary pores, and the water is termed capillary water, whereas the nanosized pores in the interlayer space within the gel itself are called gel pores. The latter have a nominal diameter of 0.5–3 nm, while the former lie in a range of 1–100 nm and have an irregular shape. In capillaries of diameter up to 50 nm, removal of water results in shrinkage because new attractive forces can form between the C–S–H surfaces. In capillaries of diameter from 50 to 100 nm, water is in the form of ‘free water’, it is not attracted by forces on the surface of solids and its removal does not yield volume changes. Capillary pores are created from the excess of water in the paste; they are interconnected and largely responsible for the permeability of the hardened paste and its susceptibility to freezing and thawing. As hydration progresses (affected by curing time, temperature or relative humidity), the volume of the capillary pores is reduced. Gel pores occupy ∼28 of the total volume of gel. Unlike capillary pores, the gel pores’ volume is independent of the water/cement ratio or the degree of hydration. As the total volume of gel increases, the volume of the relevant pores increases, too. Therefore, it is postulated that the addition of nanoparticles could have a nanoporosity filling effect. In general, the micropores ( < 50 nm) are significant for drying shrinkage and creep, while the macropores (>50 nm) affect strength and permeability.¹⁷ Last but not least, there is entrained air (purposely trapped air in the mix, usually with the help of admixtures) in the form of spherical voids of 70–500 μm in size, which mainly increase freeze/thaw resistance and irregular voids of entrapped air (air trapped due to, for example, inadequate compaction), varying in size (often larger than 1 mm), usually exhibiting a distorted shape and accounting for 1–2 in most concrete mixes.^13,14,30

The third significant component of the hardened cement paste is water. Apart from the capillary water, which has already been mentioned, there is also the adsorbed water. Most of the hydration products are colloidal; hence, according to Jennings,³¹ the surface area of C–S–H expands during hydration and adsorbs free water, which for this, it is termed adsorbed water. The layers of adsorbed water are held by hydrogen bonding and can be removed if the hydrated cement paste is dried to below 30 relative humidity, causing shrinkage. Furthermore, some other water molecules are trapped within the interlayers of C–S–H, comprising the interlayer water, a monomolecular layer strongly held by hydrogen bonding. If the interlayer water is lost (only under severe drying), collapse of the C–S–H layers at 11 relative humidity occurs. Last, the chemically combined water, which is water in molecular combination with the hydration products, cannot be removed by drying, but can be decomposed on heating.

Effect of nanoparticles on cement hydration and microstructure

The main reason for nanomodifications of materials is that as its size decreases to the nanosize, its specific surface area is multiplied. The nanoparticles added to cement are more chemically reactive, since a greater surface area is available for reactions.^32–34 Furthermore, the nanoparticles strengthen the hydrating cement nanostructure by minimising the nanosized, gel pores (0.5–5 nm wide) within the C–S–Hs.^35–41 They are, consequently, expected to react rapidly and enhance the precipitation of hydrates, resulting in a denser matrix.^36,42–47

These observations have been rendered possible with the advancement of nanotechnology and related experimental methods, namely, scanning electron microscopy (SEM), transmission electron microscopy, Fourier transform infrared spectroscopy, thermogravimetric analysis (TGA), XRD, atomic force microscopy (AFM) and AFM nanoidentation, nuclear magnetic resonance (NMR), small angle neutron and X-ray scattering.²² In light of this, nanotechnologically enhanced cement composites can be engineered to deliver innovative cement blends targeted at specific strength and durability characteristics.⁴⁸

The introduction of nanoparticles in cement matrixes affects hydration, strength, durability and microstructure in a number of ways. An overview of the various roles that the nanoparticles may develop within the cement paste is briefly presented below and analytically discussed in the following sections: (i)

nanofillers, reducing the nanoporosity within the hydrated paste and extending the concept of particle packing^49–51

(ii)

nanoreinforcements, increasing the tensile and flexural strength and consistence of the paste especially the sheet/platelet like nanoparticles^40,47,52

(iii)

catalysts due to their high surface area^34,53,54

(iv)

nucleation sites for the reactions due to their high specific surface area and the unsaturated bonds on the surface of the particles^52,55

(v)

highly reactive pozzolanas, consuming Ca(OH)₂ to produce additional C–S–H^{47,51,56–58}

(vi)

counter balance against high amounts of supplementary cementitious materials (such as reduction in the early strength gain with high amounts of FA or need for superplasticisers with large quantities of μS).⁴⁹

Some of the nanoparticles more extensively researched in the cement industry are nanotitanium dioxide,⁵⁹ nano-Fe₂O₃,⁴⁵ nanoalumina (nano-Al₂O₃),⁶⁰ nanosilica (nS),^51,54,57,61 nCs,^41,47,62,63 nano-LS (nano-CaCO₃)^64,65 and others. In the present review, the effect of the addition of two types of nanoparticles, namely, nS and MMT nC, was investigated. Both nanoparticles attract the attention due to the innumerable abilities they offer, if used appropriately, for the production of low carbon footprint cement with enhanced properties leading to pragmatic sustainability.

Adding to the above, the cost of nanosized silica containing particles is significantly lower than the cost of μS. In fact, nS is produced from the extraction of silica sand avoiding the silica fume industrial processes.³⁵ Nanomontmorillonite is produced from MMT, which is a clay material found in abundance in nature. It is interesting to note that nS characteristics and reactivity are affected by the production method used, and the same is valid for nC, which, on top of that, being produced by generic clays also have a significant variability according to the place they are collected from. For this, the various production methods are also discussed in the present review.

A number of inconsistencies, contradicting results and omissions have been observed in several studies calling for a holistic understanding of the effect of nS or nS on the C–S–H formation, while simultaneously discussing about the optimal dosage, the limits of the nanoparticle's reactivity in blended cement pastes, potential competition with supplementary cementitious materials and use of superplasticisers in the pastes in the fresh and hardened state. All these will be discussed extensively in the following subsections.

Addition of nS particles to cement pastes

On production of nS particles

According to Fares and Khan,⁶⁶ the foreseen global need for silica (natural, biogenic or synthesised) for concrete will reach 2.7 million tonnes with a total cost estimation approaching 3.9 billion British pounds by 2014. Nanosilica is essentially nanosized silica, produced by reducing the particle size of silica powder to the nanometric level through a number of different processes. Silica has the chemical formula of silicon dioxide (SiO₂). Every atom of Si is connected with four atoms of oxygen, creating tetrahedra, the basic building block. Hence, silicate chains are produced by oxygen sharing silicate tetrahedra. In crystalline silicas, there is some long range order with chains formed by siloxane bonds ( ≡ Si–O–Si ≡ ). Amorphous SiO₂ has no long range order. In both cases, the Si atom is unsaturated on its outer side, allowing for reactions with water leading to the creation of silanol groups ( ≡ Si–OH), governing adsorption and surface reactivity of silica.⁶⁷

Different synthesis processes have been suggested, believed to have an effect, in turn, on the reactivity of nS.^68,69 The nS particles delivered are either hydrophilic or hydrophobic, with respect to the condition of the OH– group on the surface of the nS particle. There are two major nS production methods used, encompassing thermal treatment or not: pyrogenic or non-pyrogenic, accordingly.^66,70 Amorphous nS particles are produced by polymerisation of monomers and can be classified in six major categories, depending on the method of production. These are the following: (i)

colloidal or sol with uniformly shaped hydrophilic particles of diameter ranging from 1 to 1000 nm

(ii)

Stöber silica with spherical hydrophilic particles of diameter ranging from 10 to 1000 nm

(iii)

silica gel with hydrophilic particles of diameter of 0.5–5 nm

(iv)

precipitated silica with hydrophilic particles of diameter of 5–6 nm

(v)

mesoporous silica with hydrophobic particles of diameter of 50–1000 nm

(vi)

pyrogenic/fumed silica with hydrophobic particles of diameter of 2–50 nm.⁶⁶

Overall, the use of dry powders in concrete can only be achieved with the use of superplasticisers as dispersion media. For practical reasons, therefore, colloidal silica is more often preferred by cement scientists. More information on the synthesis of amorphous nS, as well as the synthesis of crystalline silica and their relative properties, can be traced in Napierska et al.⁷⁰ Production costs in terms of energy requirements can be reduced through nS production by the dissolution of olivine (Mg, Fe)₂SiO₄, a much promising technique, since the Si–O–Si bonds it contains can be easily broken; however, it is still at experimental level.⁷¹

Optimum dosage of nS particles

Different optimum proportions have been suggested by various researchers, and contradictions have been observed. Some researchers managed to maximise compressive strength by keeping nS substitution to a minimum, for example < 1;^{40,47,56,72,73} between 1 and 3.5nS,^58,74,75 0.5, 1, 1.5 or 2;⁶⁸ or up to 10 wt-nS of cement.⁴⁶ The justification provided was on the grounds that, at higher dosages, the nS particles conglomerate and dispersion is inhibited. Others recommended even higher percentages of replacement for best results. For example, Bi et al.³⁵ tested concrete specimens containing 3, 5, 10 and 15nS and compared them against similar specimens that contained 5 of μS in addition, in a study discussed in detail later. In another study, Mondal et al.,⁷⁶ in a nanoindentation and ²⁹Si magic angle spinning NMR spectra analyses, compared pastes containing 6 and 18 mass- colloidal nS and found that the higher the percentage of nS, the more the high density C–S–H (38 for the sample containing 6nS and 50 for the sample of 18nS) is formed against low density C–S–H. According to this finding, the higher the nS replacement, the more durable the product is (since more high stiffness C–S–H, which is more resistant to calcium leaching, is produced). However, recent studies⁴⁷ suggest that the optimum dosage should be expressed with respect to the constituent materials. The author investigated the effect of nS in a non-pozzolanic reference pastes, with high LS content, 60PC and 40LS, and found that, for long term strength maintenance, the amount of nS should not exceed 0.5nS solids by mass of binder. Pastes that contained high proportions of FA (e.g. 43PC, 20LS and 36FA) performed best at 0.5nS solids content by mass of binder, a finding confirmed with the use of two different commercial colloidal nS solutions.⁴⁷

One of the research objectives is, hence, to engineer and suggest a matrix of colloidal nS dosages with respect to the intended application. Furthermore, it would be interesting to investigate the different C–S–H density areas of composite cement pastes with NMR studies and the polymerisation of silicate chains with AFM.

Pozzolanic activity of nS particles

In a comparative study presented by Lim et al.,³² the pozzolanic reactivity of nS in mixes containing nS dispersion at 5 replacement by mass of cement was confirmed through TGA, XRD and compressive strength tests. Importantly, the fact that the addition of nS increases the thermal stability of the paste was affirmed. Two series of samples were prepared (with and without nS) and were cured in an environmental chamber at 25°C and 100 relative humidity for 28 days. They, then, underwent different circles of heat treatment, ranging from 105 to 500°C. To ensure the validity of results, six specimens were used for the heat treatment and three specimens for control. Thermogravimetric analysis showed a 5 mass loss related to the Ca(OH)₂ consumption for the sample containing nS. The observed reduction in Ca(OH)₂ for the specific samples supports the opinion of the enhanced pozzolanic reaction that nS is offering, a conclusion also confirmed by the XRD analyses. It is interesting to note that the XRD showed that the percentage of nS was insufficient for the full transformation of Ca(OH)₂ into C–S–H. The compressive strength of the samples with nS was always higher by 10–20 with maximum performance achieved for the sample with the highest heat treatment. Still, in the present research, the optimum content was not investigated; TGA were carried out only up to 500°C, and tests were presented only for day 28, not allowing further conclusions to be drawn. Senff et al.⁵⁸ studied the combined effect of nS and μS in PC pastes (CEM I-52.5R) admixed with a polycarboxylic acid based superplastictiser. Thermogravimetric analyses at days 7, 28 and 90 showed pozzolanic activity of the nS dispersion used at a dosage of 3.5 or 2 mass- cement (nS Levasil 300/30, containing 30nS by mass of solids).

Recent studies, discussed in greater detail in the next subsection⁷⁷ on PC-FA-nS mortar systems at 1, 2, 7, 28, 56 and 90 days, suggested a reduction of Ca(OH)₂ over time to the extent that the pozzolanic reaction of FA at later days is delayed and strength gain is stabilised. Particularly in water cured mortars at 70°C in which high contents of colloidal nS (5 or 7.5) was added to a 60PC and 40FA system, depletion of Ca(OH)₂ was observed.

Thermogravimetric and XRD analyses of non-pozzolanic and pozzolanic reference cement pastes, enhanced with 0.1, 0.2, 0.5, 1.0 or 1.5nS solids at days 1, 7, 28, 56, 90 or 170 showed undoubted pozzolanic activity of the nS particles, pronounced on the early days, comprising of Ca(OH)₂ consumption and production of additional C–S–H.^47,57 In the high FA and LS content systems (43PC, 36FA and 20LS by mass of binder) a delayed pozzolanic reaction of the FA was also postulated, attributed to the high pozzolanic performance of nS during early ages^47,51 and due to the ion penetration barriers created by the different C/S ratios of C–S–H produced by the nS addition.^47,54,57,77 An AFM nanoindentation study could provide more information on the characteristics of the different areas of C–S–H produced, and a series of field emission SEM investigations could assist in the identification of the ion penetration barrier ‘coats’ of the A particles in PC–FA–LS–nS systems.

Effect of nS particle size also related to pozzolanic activity

Al-Otaibi³⁶ compared the effect of two different sizes of nS: a 15 nm (fine) nS and an 80 nm (coarse) nS to cement mixes. Replacement (1, 2, 3, 4 and 5) of PC was investigated. X-ray diffraction, scanning electron microscopy (SEM), TGA and compressive strength tests were carried out at 7 and 28 days. Some results for the 3 and 5 replacement were presented. Overall, the addition of nS was found to increase hydration products and accelerate the reaction. The microstructure was once again improved. Conclusions would have been more robust if supported by later ages and all percentages of nS.

A similar comparison, although this time using colloidal nS, was carried out by Kawashima et al.⁷⁸ who blended nS in FA mortars. Nanosilica is believed to counteract the negative effects that the addition of FA has on the early age properties of pastes. The importance of incorporating nS in the mix is then enhanced since, in parallel, it allows for greater substitutions of cement by greener replacements. Their intention was to combine the high consistence, provided by FA, with high early strength gain provided by nS. The slump flow was determined for mortars containing 0, 2.25 or 5nS together with 20, 40 or 60FA. Fluidity increased with increasing FA and decreased with more nS. At the same time, compressive strength of later ages was also slowed down. Microscopic analyses showed that this effect can be attributed to the consumption of most of the Ca(OH)₂ by nS at early ages, leaving small amounts of Ca(OH)₂ for the hydration of FA. However, some FA particles exhibited a thick coating of hydration products comprised of two layers with a low calcium/silicon (C/S) ratio, lower than that of the adjacent C–S–H, implying that this double coating is acting as an ‘ion penetration barrier’ not allowing the hydration of these FA particles. They also concluded that the higher the FA and nS contents are, the more insufficient the amount of Ca(OH)₂ is likely to be. In the present study, arrest of hydration was carried out with the use of acetone, which is not the most recommended alcohol for this procedure.⁵⁴ Oltulu and S¸ahin⁷⁹ studied the effect of powder nS on PC mortars containing 15 mass-FA and concluded that increase in compressive strength can be expected if nS addition is limited to 1.25 mass- when FA is present. A 2.25nS addition led to reductions in compressive strength due to agglomeration of the nanoparticles as suggested by SEM analysis.

An interesting conclusion with respect to particle packing was drawn by Senff et al. in the study mentioned earlier.⁵⁸ Analyses (SEM) of low water/solids ratio mortars indicated that nS additions enhance early age strength, an effect less pronounced at later ages (Fig. 1) due to agglomeration of the nS particles, responsible for a poorer arrangement and packing of particles. It should be stressed, though, that a 3.5nS solids addition is not the right proportioning for mortars containing only PC. In the same research, the 2nS and 10.2μS combination provided better results. However, it should be noted that, since the nS content was greatly reduced, one cannot tell with certainty that the particle packing was improved solely by the presence of μS and not by the more limited agglomeration of the nS particles, as well. A series of density studies and particle size distribution tests at different nS additions, starting as low as 0.1, could provide more information about this effect, especially if compared with mathematical simulations of the microstructure hydration, the pore space characterisation and permeability.⁸⁰

Improvement in compressive strength of mortars with addition of nS and/or microsilica⁵⁸

Effect of nS on blended cements

Colloidal nS produced by sol–gel and stabilised with sodium was added to cement and calcium hydroxide to study the effects on cement hydration and C–S–H gel properties in early and later ages by Hou et al.⁸¹ Binary mixes containing up to 5nS and up to 5μS was also developed for comparison at a water/binder mass ratio (w/b) ratio of 0.4. Specimens were demoulded 1 day after casting and were stored in a limewater curing tank for 3, 7, 28 and 84 days. For the microstructural characterisation, the samples were crushed, the hydration was arrested with acetone and then they were oven dried at 105°C for 4 h. Once again, it was confirmed that the pozzolanic reaction of nS was nearly finished after 7 days, but for μS, a month was needed. In agreement with Ye et al.,⁷⁴ they attributed this difference to the chemical structure of the two materials; nS having many unsaturated Si–O bonds [due to high surface area, there are more terminations of the silica network at the surface, which can, in turn, extend pozzolanic reactivity (and subsequently the amount of C–S–H formed)82], whereas μS has saturated ones, slowing down the reactions. The compressive strength tests on mortars having additionally 40FA as cement replacement revealed that nS has a stronger effect at the early ages. Indicatively, for the 5nS mix the 7 day strength was enhanced by 45. However, the 5nS and μS mix exhibited a 10 reduction for the same age. They also concluded that nS accelerated cement hydration due to the nucleation sites created. The severe pozzolanic reaction that is taking place when nS is added is leading to the depletion of Ca(OH)₂, an effect responsible for the high early age reactivity of nS and catching up of μS mixes with the nS ones at later ages. As also pointed out in an earlier study of Hou's team,⁷⁸ backscattered SEM image analysis confirmed this finding, while revealing a coating of hydrates surrounding unhydrated cement particles. This coating of very low permeability is not allowing alite's hydration. For this reason, only 89.4 of cement with 5nS added participated in the 8 month hydration. However, as discussed earlier, it has been found⁵⁷ that, in ternary cement formulations (43PC, 20LS and 37FA), depletion of Ca(OH)₂ was not an issue and that there was enough Ca(OH)₂ should the FA be able to react with, although the antagonism between FA and nS has been clearly discussed.⁵⁴

Last, AFM nanoindentation tests showed that the addition of nS can cause an expected reduction of LD C–S–H in favour of the HD C–S–H, as also proven by Mondal et al.⁷⁶

Effect of conditioning, grinding and mechanochemical activation (MCA)

Soleymani⁶⁸ investigated the effect that curing in saturated limewater or plain water or a combination of both has on pastes where OPC (CEMI) is replaced by 0.5, 1, 1.5 and 2nS powder by mass of cement. Water/binder (cement+nS) ratio was held constant at 0.4. Compressive strength tests were carried out on mortars at days 7, 28 and 90. The optimal mix cured in water contained 1nS, where the 2nS mix was less strong than the 0.5nS mortar. Overall, the highest strengths were achieved by the 2nS mortar cured for the first 28 days in limewater and, from then on, in plain water.

At this stage, special attention should be given to the preparation method followed for the addition of nS to the mix. Various methods have been studied in an effort to accelerate hydration and avoid clustering of the nanoparticles. Stirring the solutions in an ultrasonic sonicator seems to be one of the most effective mitigation measures.⁸³ Very recently, Elkady et al.⁸⁴ presented a study on the effect of different deagglomeration techniques (sonication, homogenisation and stirring) on the compressive strength and workability of nS concrete. The specimens were water cured and tested at days 7 and 28. Five minutes sonication proved to be the most effective method, enhancing the compressive strength by 23 and the concrete workability, using only 1nS (as opposed to 0, 1.5 and 2nS), substituting cement. Sonication was considered to have increased the surface area and the nucleation sites. The sonicated nS exhibited a higher pozzolanic activity.

Last, apart from superfine grinding of the materials or ultrasonic stirring or high speed mixing at different stages, Lin et al.⁴³ presented a study of mortars prepared with nanomilling and MCA. They created a slurry believed to be ‘mechanochemically activated’ containing colloidal nS at 10–20 cement replacement and superplasticiser. This slurry was then tested in standard mortars as replacement of PC, cured at room temperature, at ages 1, 3, 7 and 28 days, again lacking the study of the effect at later ages. Adding to this, two specimens were reported to be tested for each mix. The results showed that the MCA did not reduce the binding characteristics of cement and, in effect, nanocement can be used as slurry for the design of concrete mixes. The early age compressive strength of cement can be improved by the proposed method as well as the microstructure of the cement thanks to the higher densification of the nanoparticles.

Limitation on addition of nS particles: Consistence

Another divergence of findings is related to consistence. According to several studies, consistence is enhanced by nS.^56,85,86 More recent studies proved that, with increasing amounts of nS, additional water and/or superplasticisers are needed.^4,35,87 Senff et al.,⁸⁸ after testing samples containing 0–7nS (in suspension having 30 wt-solids), μS of 0–20 and w/b ratio of 0.35–0.59 found that the maximum unrestrained shrinkage increased 80 for the 7 day old nS mortars and 54 for the 28 day old ones, compared against the μS mortars. Moreover, the maximum proportion of nS yielded the maximum water absorption and porosity. It is necessary, therefore, to clarify the effect the addition of various percentages of nS has on the consistence of cement pastes.

Mitigating limitations: Use of superplasticisers and other additives

Bi et al.³⁵ produced a reference concrete mix, including μS, and the admixture Viscocrete 10 ex by Sika and compared its compressive strength, permeability and microstructure against four mixes containing 3, 5, 10 and 15nS by mass of cement and four mixes containing 3, 5, 10 and 15nS together with 5μS by mass of cement. Testing was carried out only at ages 1, 3, 7 and 28 days; therefore, the effect nS has on the mixes at later ages, which is of high importance and interest, was not reported. The 10nS with 5μS mix showed the highest compressive strength and the lowest permeability and chloride ion penetration. The highest amount of nS (with or without μS) showed a decrease in the compressive strength and a dramatic increase in permeability and chloride ion penetration, a fact attributed by the authors to the agglomeration of the particles, implying that excessive nS does not allow μS to react as intended. It is of interest to note, though, that the lower dosage of nS (3nS and 5μS) exhibited very high values on permeability and chloride ion penetration, and that for higher amounts of nS, the w/b ratio and dosage of viscocrete was increased, proving once again the point that workability would have been reduced with higher dosages of nS if superplasticisers had not been employed.

Oltulu and Sahin³³ compared the compressive strength and capillary permeability of cement mortars combined with one, two or three nanoparticles of nS, nano-Al₂O₃ and nano-Fe₂O₃ in powder form and μS. It is worth noting that a water reducing agent (commercially available as Viscocrete PC-15) was added in the mix at 0.75 mass- binder, as well as a defoamer at 1 mass- binder. After mixing, compacting and demoulding a day later, the specimens (three for each mix and age) were cured in limewater at 20 ± 2°C. In the single combinations, only nS was added at 0.5, 1.25 and 2.5 mass- binder. In the binary combinations, apart from nS, μS was added at 5 of the cement, keeping the w/b ratio constant and equal to 0.4. Compressive strength was tested at days 3, 7, 28, 56 and 180. In all mixes, higher values were achieved with age, the later ones being marginal. Best results were achieved for the 1.25nS mortar. In specific, both compressive strength and capillary permeability were greatly enhanced, in both single and binary combinations compared to the specimen control (mortar of CEMI+5μS by weight of cement). In contrast, strength reduction between 43 and 60 and increase in the capillary permeability of the mortar by 54 compared to the control specimen was observed for the 2.5nS mortar. The general increase in compressive strength and reduction in ‘capillary permeability’ (as a function of the amount of water absorbed by immersed in water specimens, the area of the specimen in contact with water and the time elapsed) was attributed to the effective filling of the micro- and nanopores of the paste by the nanoparticles, supported by SEM images. There is a threshold, though, that, if exceeded, the permeability of the mortar instead of being further decreased, as intended and expected due to the filler effect, is reduced. This was attributed to agglomeration of the nanoparticles that took place at higher concentrations (2.5nS) due to their very high specific surface area, which in turn caused uneven dispersion of the nanoparticles in the mortar. Hence, the two researchers brought attention to the fact that this threshold must be determined in all efforts of improving the pore structure of pastes.

It is a fact, though, that superplasticisers have been used in older studies, too. For instance, Ye et al.⁷⁴ kept a constant w/b ratio of 0.22 and a superplasticiser content of 2.5 mass- cement to compare mixes with 3nS or 3μS content. They concluded that nS, due to its high specific surface area, increases the amount of adsorbed water and the wettable surface area, producing a denser paste. They also concluded that, due to the great number of unsaturated ≡ Si–O– and ≡ Si– bonds, nS accelerates hydration, as was also expected by Bye,¹⁷ as opposed to the hydration reactions taking place in mixes containing only μS, which has many saturated bonds siloxane ( ≡ Si–O–Si ≡ ) in the surface.

Last, the compressive strength results they presented included ages 1, 3, 28 and 60 days. Ye et al.⁷⁴ also concluded that compressive strength increases with increasing percentage of nS addition, as opposed to μS, which had an adverse effect for 1 and 3 day old samples.

Li et al.⁴⁶ tested the mechanical properties and produced SEM images of cement mortars containing 3, 5 and 10nS by weight of cement mixed with cement, sand defoamer and UNF water reducing agent. In Li et al.,⁴⁵ the mixing procedure for the prementioned mortars is described as follows: the defoamer and the water reducing agent were dissolved in water, then nS was added and the mix was stirred for 2 min at high rotation. Next, cement was added and mixed for 1 min at low speed in a mortar mixer. Sand was added last and mixed again at low speed for 1.5 min. At a constant w/b ratio of 0.5 (the binder being considered as the sum of cement and nanoparticles), the three different mixes showed an enhancement in both 7 and 28 days compressive and flexural strength, with values increasing with the nS content. The conglomeration ability and the formation of nucleus were once again observed with the use of SEM. Furthermore, it was confirmed that the nanoparticles may prevent the expansion of the Ca(OH)₂ crystals, and although they are indeed capable of filling the nanopores, in cases where they are not well dispersed (such as at high nanoparticle contents), they create weak zones, an opinion also shared by Sobolev et al.⁴⁰ At the same time, they claimed that, at lower nS contents, even if nS is not well dispersed, the weak zones will be prevented, as the nS will be consumed for the production of C–S–H. They argued that the addition of nS had an overall enhancing effect, providing mortars with higher compressive strength for the pastes containing higher nS percentage, whereas flexural strength was found to be inversely proportional to the nS content. It should be stressed that these two researches were carried out for the ages of 7 and 28 days, hence still lacking the later age effect.

Discussion and further research

To sum up, it is still uncertain what is the optimum dosage of nS in cement, with researchers suggesting dosages ranging from 0.5⁶⁸ to 1^74,85 and even as high as 10 mass- cement.⁴⁶ Furthermore, a maximum limit on nS addition in blended cements cannot be determined, but a matrix of permitted nS additions with respect to the constituents of the cement formulation should be developed. Adding to this, most researchers are reporting pozzolanic reactivity of the nS particles, whereas others are supporting the seeding effects of nS. Therefore, the way the nS particles affect cement pastes, both at early and later ages, should be further investigated, particularly in relation to the constituent materials since evidence of antagonistic performances have been presented. Future research could investigate the effect of different percentages of colloidal nS in blended and composite PC pastes in an effort to define the following: (i)

the limits of nS addition

(ii)

the competing constituents in the formulations and the effects of their interaction

(iii)

the roles nS particles play within the composite hydrating pastes

(iv)

optimal engineered formulations towards more sustainable PC blends.

Nanotechnology, with AFM, NMR or TEM investigations, could facilitate the mapping of the various areas of C–S–H, while advanced molecular modelling could reduce experimentation time and expand current knowledge.

Addition of MMT nC particles to cement pastes

Supply of natural clays

After an extensive number of studies in the field of clay–polymer nanocomposites, clay–cement nanocomposites have gained momentum as a cutting edge alternative to the engineering of sustainable modern construction materials. An nC is essentially a clay whose layers have been separated and are individually available for reactions. The popularity of clays as additions to cement is due to their vast surface area, which lays in the order of 10–800 m² g^− 1,^89,90 since it has been proven that the extent of pozzolanic reactivity (and subsequently the amount of C–S–H formed) is proportional to the surface area available for reactions.⁸² Unlike most nanoparticles, clays are naturally occurring minerals; therefore, their use is the most economical among most nanoparticles.⁹¹ Clays are classified in two categories with respect to the geographical location they are extracted/excavated/mined from: residual and transported clays. The forgmer are the result of rock weathering and are found where they were created. They are usually formed by either disintegration of volcanic ash or by the hydrothermal modification or solution of volcanic rocks.⁹² The latter category encompases all clays naturally removed from the site they were created and deposited elsewhere.⁹³ The main constituents of clays are silica and alumina.

Clays can also be classified with respect to their structure into four main groups: smectites (comprising of MMTs and others), kaolinites, illites and chlorites. The latter is classified by some scholars as a separate phyllosilicate group.⁹² Smectites are groups of clay minerals, with a general chemical formula: (Ca, Na, H)(Al, Mg, Fe, Zn)₂(Si, Al)₄O₁₀ (OH)₂.nH₂O.⁹³ Montmorillonite is a specific mineral, which took its name after the place where it was discovered, Montmorillon, in the Vienne prefecture of France, in 1847. Bentonite, also known as Wyoming bentonite, was given its name by Knight in 1898, after the cretaceous Benton shale near Rock River, Wyoming. The Wyoming bentonite is an impure clay soil, containing a number of minerals the majority of which (∼80) is MMT.⁹⁰ Through purification processes with the use of sodium cations, MMT can also be derived from bentonites, justifying its second name, ‘sodium bentonite’, as discussed later.

Nanostructure of MMT clays

In order to understand how the nC is produced, the structure of the natural MMT clay will be discussed first. Montmorillonite clay is a two-dimensional hydrophilic nanoparticle. It is composed of crystalline layers or platelets stuck together. Each layer is composed of a ‘sheet’ of octahedra of Al₂O₃, which is bonded at the top and bottom by a ‘sheet’ of tetrahedra of SiO₂. The silicon–oxygen tetrahedra share three corners with neighbouring SiO₂ tetrahedra to form hexagonal networks, whereas the fourth corner is bonded with the octahedra of AlO6.9³ The total thickness of this 2:1 ratio, structure, also known as 2:1 layer silicate (Fig. 2), is estimated to range from ∼0.95 to 1 nm.^90,94–96 The lateral dimensions varying from a few nanometres to a few hundreds of micrometres.⁹⁵

Structure of montmorillonite⁹⁵

Production of MMT nC particles for use in cement and examples of other applications

The agglomeration of the nanosized layers in MMT is due to the surface attraction between them.⁹⁷ A certain number of layers can be held together by electrostatic force, by hydrogen bonding, by interlayer cations or by van der Waals forces, inhibiting the solubility or miscibility.⁹⁴ However, these forces are moderately weak forces in MMT.^93,94 Clays can only be considered to comprise of individual nanoparticles, charged for reactions, if and only if the bonds bridging the layers are weakened to the extent of separation of the layers. This happens usually in two steps: purification of the MMT through washing with distilled water and shearing/vigorous stirring leading to the dilamination of the particles⁹⁸ (heating being another or additional method for delamination which is mostlty used for the expansion of metakaolin) and introduction of either a modifier, within the galleries, having a chemical effect by causing an exchange of cations, hence a change in the charge of the molecules inducing electrostatic repulsion of the platelets and hence a furter separation of the layers or/and of a surfactant, introducing solely surface tension and repulsion of the platelets. The addition of either of the two should be followed by vigorous striring, reflux or other processes.^90,98

Depending on the medium in which these processes take place, the nCs can be further classified and used in a number of different applications. That is to say, nCs can be prepared in ceramic or metallic matrixes, in polymeric matrixes [producing the polymer/clay nanocomposites (PCNs)] or in water.⁹⁰ The production of the nCs is a highly complex process, affected by a number of parameters, depending on the application that the nC is designed for. A large number of papers have been published on nCs prepared for various applications such as biopolymer–clay nanocomposites for drug delivery.⁹⁹ or nCs for the biodegradation of polyolefins.¹⁰⁰ Other examples indicatively referred to include the study by Hetzer et al.¹⁰¹ in which the effect of the processing conditions and the nC loading on the properties of the resulting material was identified. In Meng et al.,¹⁰² the effect of the preparation method of PCNs was discussed; in Utracki,⁹⁰ the effect of the agents, mixing equipment and methods on the thermal stability, mechanical behaviour, rheology and crystallisation of PCNs was compared. Furthermore, in Aktas and Altan,¹⁰³ the effect of nC loading on the thermal stability and interlaminar shear and flexural strength of glass fibre reinforced waterborne epoxy laminates was analysed. Bandyopadhyay et al.¹⁰⁴ characterised clay and silica nanocomposites dispersed in rubber mediums in terms of mechanical strength, morphology and thermal stability.

It should be stressed that, although there is a very large number of papers on nCs for various usages, the ones related to cement science are very much limited. In the present review, the MMTs are analytically discussed, due to their properties discussed above and in the paragraphs that follow. It is acknowledged that there are a few papers on other nCs such as nanokaolinites;^105,106 however, this paper is limited to the discussion of the MMTs in cement formulations. In light of this, for an MMT nC to be suitable for use in a cement formulation, platelet separation should take place in water, as hydrating cement is regarded an aqueus environment. Furthermore, metallic, ceramic or polymeric matrixes would add enormous complexity in assesing the effects of the nanoparticles on the hydration, morphology and mechanical characteristics of nC enhanced cement blends.

Montmorillonite particles are naturaly agglomerated inorganic, hydrophilic nanoparticles, and they are considered incompatible to be added directly in cement formulations because the clay will swell in presence of water.⁵³ For this reason, significant research has been dedicated to the modification of the clay in an organophilic material, compatible with cement, which is the case if the modifier is organic. The clay is then said to be organomodified [e.g. organomodified MMT (OMMT)], becoming organophilic, therefore hydrophobic by cation exchange. The cation exchange mechanism can be explained as follows: Si and Al are bonded with oxygen molecules. Molecules of OH⁻ are also found at the edges of the octahedral sheet. Since each of the layers is charged, including their edges, a major feature of the MMT is the fact that ion exchange can take place within their structure affecting the properties of the clay. Specifically, mostly Na, Mg, Al or Ca cations can be exchanged for organic cations between the interlayer to produce the OMMT. When only Na⁺ cations are used for the modification, the clay is known as sodium MMT or, else, bentonite, with chemical formula: Na0.3₃{(Al1.6₇Mg0.3₃)[O(OH)]₂(SiO₂)₄}.¹⁰⁷ A number of studies on the preparation of OMMT are mentioned in De Paiva et al.,⁹¹ as one of the main usages of this technology of modifying clay into a highly absorptive, organophilic material with enormous specific surface area has been extensively implemented in sea water pollutant removal applications or remediation of contaminated soils.^89,108 Theoretically, most MMTs have a total total content of SiO₂ and Al₂O₃ equal to 92, and the SiO₂/Al₂O₃ ratio is equal to 2.6. These values for the sodium MMT, however, are 83.6 and 4.79 respectively, indicating greater amounts of SiO₂. Therefore, if the SiO₂ and Al₂O₃ platelets are dispersed, becoming individually available for reactions, they could enhance in cement hydration, towards the production of additional C–S–H and C–A–H. This fact is rendering nCs a very interesting material for cement scientists.

The degree of expansion in the OMMT is reflected by the number of cations exchange between the layers, of the crystal which is termed cation exchange capacity (CEC) of clays. Montmorillonite has a CEC in the order of 80–120 meq/100 g.⁹⁶

Apart from their chemical composition, MMTs are also preferred due to their high CEC, their high surface area and the relatively easy ‘expansion’ in water. However, the higher the consentration of the nC particles in these aqueous solutions, the greater the possibility of flocculation.

Kuo et al.¹⁰⁹ have stressed the need for the organomodification of the clays by cation exchange in order to prepare the clay for mixing with water and consecutively with cement. Notwithstanding the significant research involved in the expansion and separation of the platelets in the presence of organomodifiers in aqueous solutions, there is one issue pertaining; since the nC has become hydrophobic, it has become incompatible with water, causing extensive flocculation of particles. This difficulty can be overcome with the use of surfactant technology. The surfactants, are amphiphilic compounds, that is to say, they contain both; a water insoluble component and a water soluble component. They can surround the platelets, increase the surface tension and allow the dispersion of the nC platelets in aqueous solutions. However, if used in organomodified clays, the product becames even more chemically complex. This technology is of particular importance as it can potentially keep the platelets dispersed without the need for organomodification, possibly allowing the use of inorganic nC. At the same time, it can be utilised in OMMT dispersions. It should be noted that some modifiers may act as surfactants, as well, but the surfactants cannot act as modifiers as chemically they are not designed to facilitate cation exchange.

Intercalated or exfoliated MMT nC particles

The organomodification and subsequent expansion of the platelets can yield three different states.¹¹⁰ The first one, which is the flocculation/phase separation, results into a microcomposite (Fig. 3a). In this case, layers of clay are mixed with a polymer, but the two phases are weakly interacting with each other. As a result, the modifier or surfactant cannot enter in the clay galleries, and the composite exhibits poor mechanical properties. The next two modification techniques produce nanosized particles of clay: the intercalation and the exfoliation (or delamination).

a phase separation (microsized); b intercalated clay (nanosized); c exfoliated clay (nanosized)

Through the process of intercalation, the silicate platelets shown in Fig. 3b can be forced apart by the selected agent, modifier/surfactant, but still remain at regular, orderly distances. When the agent is very reactive, the platelets are completely separated and randomly dispersed; they are exfoliated as shown in Fig. 3c, transforming a small mass of MMT in multiples of platelets, therefore creating a material with very large surface area. Once it is exfoliated, the interlaminar distance increases, reaching 5–10 or more nanometres,⁹⁵ and the environment is transformed to hydrophobic (organophilic),^62,89,96 facilitating the nC dispersion in water (since water will not be absorbed causing subsequent agglomeration of particles). When added to cement, the hydrophobic nC is expected to inhibit water from entering the interlayer space, lowering the water demand. Simultaneously, OMMT nanoparticles are expected to contribute to the permeability reduction by surrounding the capillary pores and impeding the diffusion of chemicals or pore solution. Adding to this, Xi⁸⁹ noted that the basal spacing of the exfoliated nC is also significantly expanded. It is generally acknowledged that composites with exfoliated clays perform better in terms of mechanical properties than system with intercalated clays.⁹⁵ This is because the specific surface area when the platelets are fully separated (exfoliation) increases drastically and by far more than when the platelets are just pushed apart (intercalation). This condition renders the nC highly reactive.

Furthermore, there are some other arrangements that may occur (Fig. 4b–d), apart from exfoliated dispersed (or in other words highly separated and suspended in water) (Fig. 4a). Aggregation of the nC platelets is represented by Fig. 4b, the distance being < 2 nm. This face to face arrangement reduces the strength of the nC gel as smaller surface area is available for reactions. Moreover, face to face stacking appears to be the preferred inherent arrangement of the nanolayers rendering their dispersion in monolayers challenging.¹¹¹ Additionally, it could induce stress paths at the nanolevel. The other two arrangements are typical of nC flocculation forming a continuous gel-like structure in the nC dispersions. Other arrangements have also been suggested.¹⁰⁷

a dispersed; b face to face; c edge to face; d edge to edge

According to De Paiva et al.,⁹¹ the lead commercial suppliers of organoclays are Laviosa Chemica Mineraria (Dellite®), Nano-cor (Nanomer), Southern Clay Products (Cloisite and Garamite), Süd Chemie (Nanofil) and Elementis Specialites (Bentone nCs).²⁰ Laviosa Chimica Mineraria, Southern Clay Products and Elementis Specialities produce organoclays based only on quaternary alkylammonium salts, also known as quats, modified MMT.

De Paiva et al.⁹¹ also stated that the ‘quats’ are the most frequently used cationic surfactants since they can reduce the density of the dispersed particles due to the large amount of organic material. These organic compounds comprise of four functional groups covalently linked to a central nitrogen atom. Of the four functional groups, there is one or more long chain alkyl group, and the remaining are either benzyl of methyl groups. Different compounds induce different basal spacing.

The modifications totally alter the structure and morphology of the clays. In order to offer a visual understanding of the nanostructure of the purified inorganic clay and the organomodified nC, TEM imaging was carried out by the author on the two samples in powder form: the initial inorganic clay (Fig. 5) and the OMMT product (Figs. 6 and 7). For the TEM imaging, 10 mg of each clay/nC powder was diluted in 100 mL of distilled water, and small drops of the diluted solutions were dripped on copper mesh grids coated with a thin carbon film. Grids were dried at 25°C before the insertion in the instrument. Samples were examined at a voltage of 120 kV, and micrographs were acquired with GATAN Jeol view camera. The agglomeration areas of the natural, inorganic clay (purified MMT) can be clearly seen in Fig. 5, whereas the platelet separation can be observed in Figs. 6 and 7.

Image (TEM) of powder purified MMT (Dellite® HPS) at × 75 000 magnification

Image (TEM) of powder OMMT at × 100 000 magnification

Image (TEM) of powder OMMT at × 60 000 magnification

Effect of modifiers/surfactants and optimum content of nC particles

A number of different modifiers/surfactants have been utilised in the published studies. For example, Birgisson and Dham⁵³ used organic ammonium chloride (also known as Cloisite Na⁺ or sodium MMT) and concluded that the amount of modifier/surfactant must be carefully chosen with the use of zeta potential, by which the electrostatic repulsion between near and similarly charged particles in a dispersion can be determined. If the zeta potential is high, the dispersion is considered stable, i.e. resists conglomeration. This condition can lead to the increase in the compressive strength of cement pastes. However, if the surfactant is excessive, the zeta potential will indicate and electrically destabilised suspension whose particles tend to agglomerate and create weak zones. After studying different samples, containing 0, 2, 5, 9 and 13OMMT by mass of cement, they reached the conclusion that 5 exhibited the highest compressive strength. Kuo et al.⁶² used dimethyl, dihydrogenated allow, quaternary ammonium chloride as organic modifier and assessed the effect that 0, 0.5 and 2OMMT had on the strength and coefficient of permeability of cement mortars for different w/b ratios: 0.4, 0.48 and 0.55. They concluded that, in dosages larger than 1 mass-, the OMMT particles tend to conglomerate. They suggested that the optimal amounts are < 1, depending on the w/b ratio. Furthermore, they clarified that the OMMT particles surround the capillary pores and impede the diffusion of pore solution, reducing the permeability. In a later study, Kuo et al.¹⁰⁹ used the same organic modifier as before. They compared the effect of different dosages of OMMT, 0, 0.25, 0.5, 0.75, 1 and 2 wt- cement replaced, and different w/b ratios of 0.485 and 0.6 have on the microstructure and mechanical properties of cement mortars. Mercury intrusion porosimetry tests proved that the pore size distribution is greatly affected by the addition of OMMT with a significant reduction of the pores between 0.1 and 0.2 μm. It should be stressed though that they also recorded an increase in the total volume of the larger pores (>0.2 μm), which could be attributed to the agglomeration of the OMMT particles while they were dispersed in water. Adding to that, they found that the compressive strength and the elastic moduli initially increase when OMMT was added; as the proportions of OMMT increase further, they drop. For this, the ideal dosage was found to be ∼0.5 for the lower w/b and ∼0.75 for the higher one. According to their results, compressive strength can be more than 33 of equivalent plain cement mortars, and elastic moduli can be increased over 65, which, in combination with the total porosity and intermediate pore reduction, justify the use of OMMT for the enhancement of the properties of cements. Chang et al.⁴¹ added OMMT to cement at 0, 0.2, 0.4, 0.6 and 0.8 wt- cement at affixed w/b ratio of 0.55. The highest compressive strength was achieved by the 0.6nC mix in a twofold manner: OMMT acted as filler, reducing the porosity of the microstructure and as a catalyst pozzolana forming supplementary hydration products. The lowest permeability coefficient was derived for the 0.4 mix. The microstructural analyses revealed a denser structure with more stable bonds.

Tregger et al.¹¹² tested concrete specimens containing 0.5, 1 and 1.5 of highly purified commercial nC by mass of cement for use in slipform self-consolidating concretes for paving. Although the early ages strength properties were presented for a concrete pavement application, the optimal dosage was found to be 1. The reference paste contained 70 of cement and 30FA by mass of binder, whereas the nC enhanced ones contained an increasing amount with the nC content amount of admixture.

Hosseini et al.¹¹³ considered that the maximum dosage of nC or nS should be 1. Therefore, they added 0 and 1 Cloisite 15A and aminosilane at 0 or 0.75 mass- binder to produce PC type I cement pastes and mortars, at a water/cement ratio of 0.4. They conducted a comparative study of the effect of 1nC or 1nS or 1 nano-Al₂O₃ or 1 nano-CaCO₃ and the combined effect of either of the four nanoparticles enhanced by 0.75 aminosilane. With respect to the nC addition, they observed an increase in the viscosity of the paste and a slower compressive strength gain of mortars at days 7, 28 and 90 than flexural strength gain of mortars at days 7 and 28.

In Papatzani⁴⁷ and Papatzani et al.,⁶³ the effect of two aqueous dispersions of OMMT and one aqueous dispersion of inorganic nC in Portland LS cement pastes was investigated in terms of compressive or flexural strength tests carried out at days 1, 7, 28, 56, 90 and 170. The nC dispersions were added at 0, 0.5, 1, 2, 4 and 5.5nC solids by mass of binder to a reference paste comprising of 60PC and 40LS, at a constant w/b ratio of 0.3. Compressive strength results until day 170 showed that marginal improvements were attained for < 1nC addition. This was attributed to the high nC loading of the dispersions and the difficulty in the miscibility of the nC enhanced Portland LS cement pastes. Interestingly, flexural strength results until day 90 of nC enhanced fibre cement pastes containing superplasticiser, based on the same reference paste (60PC and 40LS) presented significant improvements, at 1nC addition. This indicated that the presence of superplasticisers in highly viscous nC dispersions could allow the nC to be more homogeneously mixed in the cement matrix. This condition was considered necessary for strength improvement.

Pozzolanic activity of MMT nC particles

Aly et al.³⁷ created waste glass cement mortars of different dosages with 2.5OMMT (Cloisite 30B) modified with quaternary ammonium salt. The flexural and compressive strength were enhanced. The TGA they carried out proved that the nC acted as a catalyst pozzolana. In the mixture with the nC, most of Ca(OH)₂ was consumed for the production of supplementary C–S–H. As a result, a very strong increase in the C–S–H formation (strong increase in the first peak dehydration) at the expense of Ca(OH)₂ (dramatic decrease in the second peak loss of carbon dioxide from the inorganic compound) was observed. The improvement of the microstructure due to the packing effect of OMMT was also confirmed by SEM analysis.

In Hosseini et al.,¹¹³ no chemical analyses have been shown in an effort to distinguish possible pozzolanic reactivity of the nC.

However, the potentials for pozzolanic activity of nCs in cement pastes and in fibre cement pastes were extensively investigated in Papatzani⁴⁷ and Papatzani et al.⁶³ In the present research, a thorough investigation with the help of TGA, XRD, SEM and W-ray energy dispersive spectroscopy of the sensitivity of the OMMT to the surfactant was presented. Two different OMMT nC dispersions were produced by the addition of the same OMMT dispersed in water with the aid of two different dispersing agents: a non-ionic fatty alcohol or an anionic alkyl aryl sulfonate. The effect of the addition of either of these two types of OMMT dispersions, in comparison with one inorganic nC dispersion and one OMMT in powder form, in various Portland LS cement and Portland LS fibre cement formulations was examined by XRD and TGA. It was concluded that the inorganic nC dispersion developed the most pronounced pozzolanic activity and that the surfactant of the OMMTs had a significant effect on the pozzolanic performance of the nC dispersion. It should be noted that the agglomeration observed, causing a reduction in the compressive strength especially with time, was also attributed to the high nC loading (15 mass-nC solids) of the dispersions.

Microstructural enhancement with MMT nC particles

In 2008, He and Shi⁵² prepared mortars with type I PC containing an OMMT or an inorganic nC only at 1 mass- cement to study the chloride permeability and microstructure. Pozzolanic activity was hypothesised but not chemically tested. Still, denser microstructure was observed, and ordered arrays were captured, suggesting the nanoreinforcement role of the nCs.

The changes in the morphology of nC (two different dispersed OMMTs, and one inorganic dispersed nC) enhanced Portland LS cement pastes (60PC and 40LS) were examined with the help of field emission SEM in Papatzani.⁴⁷ It was concluded that, with the advancement of age, pastes seemed to become denser and the formation of C–S–H and ettringite more noticeable with the addition of the inorganic nC dispersion. In these pastes, the pozzolanic activity of the inorganic nC was further justified by the obvious reduction in Ca(OH)₂ crystals with age with respect to the reference paste. On the contrary, the addition of the OMMT produced less coherent pastes. Nanoclay sheets forming ordered arrays were distinguished in the pastes in Papatzani,⁴⁷ just like the ones shown in mortars in Hosseini et al.¹¹³ or He and Shi⁵².

Discussion and further research

It is evident from the present literature review that the optimal dosage of nC for possible increase in compressive strength is dependent upon a number of factors: (i)

the nC type (it is acknowledged that only OMMT was considered)

(ii)

the nC loading of the dispersions

(iii)

the surfactants used to disperse the nCs in water

(iv)

the constituents of the reference cement formulation.

Given the reactions with various chemicals for the development of nC particles, careful characterisation of the resulting nCs should precede their use in cement pastes. The nature of the specific nanomaterial does not allow for overall conclusions to be drawn. However, it is reasonable to assume that the inorganic dispersions can act as both nanofillers and pozzolanas, whereas the performance of the OMMT dispersions depends on the effectiveness of the surfactant used to retain the platelets dispersed in the aqueous solutions. The effect on the flexural strength development is another area of research, since the different arrangements that the platelets may take could act as nanoreinforcement of the paste, and further investigation is required. Furthermore, the long term effect of nC on strength is worth being studied since some researchers claim reductions in strength over time. Up to what proportions on nC solids by mass of binder allow for strength increase? Although much has been discussed about OMMT, research seems to have been mainly focused only on one direction, that of inserting organic modifiers in MMT to transform it to hydrophobic in fear of its swelling capabilities if allowed to maintain its natural characteristics in a cement paste. However, the fact that the organomodified intercalated or exfoliated nC particles should be dispersed in an aqueous solution and remain dispersed until insertion in the cement paste constitutes an additional difficulty. One research provided this comparison with inorganic nC, in which only surfactants and no modifiers were present; however, further research with the addition of superplasticisers is expected to extend the application of nCs in cement pastes.

Conclusions

The present review was selective and by no means exhaustive. The main conclusion can be summarised as follows.

The hydration product of PC is nanosized and can be manipulated at the nanolevel by adding nanoparticles. Properties such as chemical reactivity, mechanical characteristics and durability of the nanomodified cement product are remarkably altered compared to cement pastes containing materials at the microlevel. Therefore, nanoparticles can significantly enhance the properties of low carbon cements, or produce new sustainable cements.

Moreover, it was identified that the effect the addition of nS or nC has on the microstructure, on long term strength and hydration of cement pastes, is still unclear, particularly in composite cement blends; so is the optimal dosage. Additionally, the combined effect of the addition of nS and μS on the properties of blended cement pastes is still unresolved.

Nanoclays are extremely complicated constituents, due to the chemistry and physics involved in their production. It is still believed that the organomodified nCs can offer advantageous solutions; however, new research suggests that the future belongs to the inorganic nCs. For this, the effect of inorganic nC on composite cement paste hydration and strength development must be further investigated.

Some areas of future research could include the following: (i)

the determination of C–S–H chain polymerisation with the addition of the nanoparticles by solid state NMR

(ii)

the detection of the changes in the C–S–H density with the help of an AFM mounted with a nanoindentation probe

(iii)

the definition of the effect of nanoparticles on changes in the early age properties with the help of vicat and calorimetry time tests

(iv)

the assessment of changes in the pore structure with Ar or N₂ porosimetry.

Acknowledgement

The author acknowledges the European Commission (FIBCEM project, grant no. 262954) for funding of the present research.

References

Olivier

J. G. J.

, Janssens-Maenhout

and Peters

J. A. H. W.

: ‘Trends in global CO₂ emissions’, 2012 Report, Joint Research Centre, Ispra, Italy, and PBL Netherlands Environmental Assessment Agency, The Hague, The Netherlands; 2012..

Meyer

: ‘The greening of the concrete industry’, Cem. Concr. Compos., 2009, 31, (8), 601–605. doi: 10.1016/j.cemconcomp.2008.12.010

Mindess

and Young

J. F.

: ‘Concrete’; 1981, London/Englewood Cliffs/London, Prentice-Hall.

Sobolev

, Flores

, Hermosillo

and Torres-Martínez

L. M.

: ‘Nanomaterials and nanotechnology for high-performance cement composites’. Proc. ACI Session on ‘Nanotechnology of concrete: recent developments and future perspectives. Denver, MA, USA, November 2006, National Science Foundation, 93–120.

Arivalagan

, Ravichandran

, Rangasamy

and Karthikeyan

: ‘Nanomaterials and its potential applications’, Int. J. ChemTech Res., 2011, 3, (2), 534–538.

European Commission . Commission recommendation of 18 October 2011 on the definition of nanomaterial. 2011/696/EU2011, European Commission, Brussels, Belgium, 40..

Barluenga

, Palomar

and Puentes

: ‘Early age and hardened performance of cement pastes combining mineral additions’, Mater. Struct., 2013, 46, 921–941. doi: 10.1617/s11527-012-9944-9

Bubshait

A. A.

, Tahir

B. M.

and Jannadi

M. O.

: ‘Use of microsilica in concrete construction’, Build. Res. Inf., 1996, 24, (1), 41–49. doi: 10.1080/09613219608727497

Dhir

R. K.

, McCarthy

M. J.

, Paine

K. A.

, 2002, ‘Use of fly ash to BS EN 450 in structural concrete’, ‘Technology Digest 1’, ‘Technology Digest TD12002’, 20; 2002, Camberley, The Concrete Society; 2002..

10.

Dhir

R. K.

, El-Mohr

M. A. K.

and Dyer

T. D.

: ‘Developing chloride resisting concrete using PFA’, Cem. Concr. Res., 1997, 27, (11), 1633–1639. doi: 10.1016/S0008-8846(97)00146-4

11.

Jennings

H. M.

: ‘Refinements to colloid model of C–S–H in cement: CM-II’, Cem. Concr. Res., 2008, 38, (3), 275–289. doi: 10.1016/j.cemconres.2007.10.006

12.

Ridi

, Fratini

and Baglioni

: ‘Cement: a two thousand year old nano-colloid’, J. Colloid Interface Sci., 2011, 357, (2), 255–264. doi: 10.1016/j.jcis.2011.02.026

13.

Alizadeh

R. A.

: ‘Nanostructure and Engineering properties of basic and modified calcium-silicate-hydrate systems’, 250; 2009, Ottawa, Ont, Department of Civil Engineering, University of Ottawa.

14.

Neville

A. M.

: ‘Properties of concrete’, 4th edn, 844; 1995, Harlow, Longman Group.

15.

Brouwers

H. J. H.

: ‘Chemical reactions in hydrated ordinary Portland cement based on the work by Powers and Brownyard’, in ‘Proc. 15th Int. Conf. on ‘Building materials’’, (ed. Fisher

H. B.

), ; 2003, Weimar, F.A., Finger-Institut fur Baustoffkunde.

16.

Dyer

T. D.

, Halliday

J. E.

and Dhir

R. K.

: ‘An investigation of the hydration chemistry of ternary blends containing cement kiln dust’, J. Mater. Sci., 1999, 34, (20), 4975–4983. doi: 10.1023/A:1004715806829

17.

Bye

: ‘Portland cement’; 2011, London, ICE Publishing.

18.

Lothenbach

, Scrivener

and Hooton

R. D.

: ‘Supplementary cementitious materials’, Cem. Concr. Res., 2011, 41, (12), 1244–1256. doi: 10.1016/j.cemconres.2010.12.001

19.

Taylor

H. F. W.

: ‘Hydrated calcium silicates. Part I: compound formation at ordinary temperatures’, J. Chem. Soc., 1950, 3682–3690. doi: 10.1039/jr9500003682

20.

Nonat

and Lecoq

: ‘The structure, stoichiometry and properties of C–S–H prepared by C₃S hydration under controlled condition’, in ‘Nuclear magnetic resonance spectroscopy of cement-based materials’, (ed. Colombet

., 197–207; 1998, Berlin/Heidelberg, Springer.

21.

Selvam

R. P.

, Subramani

V. J.

, Murray

and Hall

: ‘Potential application of nanotechnology on cement based materials’, Final refport no. MBTC DOT 2095/3004, National Technical Information Service, Alexandria, VA, USA; 2009..

22.

Papatzani

, Paine

and Calabria-Holley

: ‘A comprehensive review of the models on the nanostructure of calcium silicate hydrates’, Constr. Build. Mater., 2015, 74, 219–234. doi: 10.1016/j.conbuildmat.2014.10.029

23.

Fonseca

P. C.

, Jennings

H. M.

and Andrade

J. E.

: ‘A nanoscale numerical model of calcium silicate hydrate’, Mech. Mater., 2011, 43, (8), 408–419. doi: 10.1016/j.mechmat.2011.05.004

24.

Vogt

: ‘Ultrafine particles in concrete – influence of ultrafine particles on concrete properties and application to concrete mix design’, Stockholm, School of Architecture and the Built Environment, Division of Concrete Structures, Royal Institute of Technology 177; 2010..

25.

Mendoza

and Tobón

J. I.

: ‘An alternative thermal method for identification of pozzolanic activity in Ca(OH)₂/pozzolan pastes’, J. Therm. Anal. Calorim., 2013, 114, 589–596. doi: 10.1007/s10973-013-2973-y

26.

McCarthy

M. J.

and Dhir

R. K.

: ‘Development of high volume fly ash cements for use in concrete construction’, Fuel, 2005, 84, (11), 1423–1432. doi: 10.1016/j.fuel.2004.08.029

27.

Lindgreen

, Geiker

, Krøyer

, Springer

and Skibsted

: ‘Microstructure engineering of Portland cement pastes and mortars through addition of ultrafine layer silicates’, Cem. Concr. Compos., 2008, 30, (8), 686–699. doi: 10.1016/j.cemconcomp.2008.05.003

28.

Barbhuiya

S. A.

, Gbagbo

, Russell

and Basheer

: ‘Properties of fly ash concrete modified with hydrated lime and silica fume’, Constr. Build. Mater., 2009, 23, (10), 3233–3239. doi: 10.1016/j.conbuildmat.2009.06.001

29.

Supit

and Shaikh

: ‘Durability properties of high volume fly ash concrete containing nano-silica’, Mater. Struct., to be published.

30.

Kurtis

: ‘Structure of the hydrated cement paste’. School of Civil Engineering, Georgia Institute of Technology, Atlanta, GA, USA, 2007. http://people.ce.gatech.edu/kk92/hcp.pdf (accessed 16 February 2013)..

31.

Jennings

H. M.

: ‘A model for the microstructure of calcium silicate hydrate in cement paste’, Cem. Concr. Res., 2000, 30, (1), 101–116. doi: 10.1016/S0008-8846(99)00209-4

32.

Lim

, Mondal

and Cohn

: ‘Effects of nanosilica on thermal degradation of cement paste’, Proc. 4th Int. Symp. on ‘Nanotechnology in construction’, Agios Nikolaos, Crete, Greece, May 2012, AEROCOINS..

33.

Oltulu

and Sahin

: ‘Single and combined effects of nano-SiO₂, nano-Al₂O₃ and nano-Fe₂O₃ powders on compressive strength and capillary permeability of cement mortar containing silica fume’, Mater. Sci. Eng. A, 2011, 528, 7012–7019. doi: 10.1016/j.msea.2011.05.054

34.

Björnström

, Martinelli

, Matic

, Borjesson

and Panas

: ‘Accelerating effects of colloidal nano-silica for beneficial calcium–silicate–hydrate formation in cement’, Chem. Phys. Lett., 2004, 392, (1–3), 242–248. doi: 10.1016/j.cplett.2004.05.071

35.

, Pane

, Hariandja

and Imran

: ‘The use of nanosilica for improving of concrete compressive strength and durability’, Appl. Mech. Mater., 2012, 204–208, 4059–4062. doi: 10.4028/www.scientific.net/AMM.204-208.4059

36.

Al-Otaibi

S. F.

: ‘Influence of nano-SiO₂ on the properties of cement pastes’, Proc. 4th Int. Symp. on ‘Nanotechnology in construction’, Agios Nikolaos, Crete, Greece, May 2012, AEROCOINS.

37.

Aly

, Hashmi

M. S. J.

, Olabi

A. G.

, Messeiry

and Hussain

A. I.

: ‘Effect of nano clay particles on mechanical, thermal and physical behaviours of waste–glass cement mortars’, Mater. Sci. Eng. A, 2011, A528, (27), 7991–7998. doi: 10.1016/j.msea.2011.07.058

38.

Vera-Agullo

, Chozas-Ligero

, Portillo-Rico

, García-Casas

M. J.

, Gutiérrez-Martínez

, Mieres-Royo

J. M.

and Grávalos-Moreno

: ‘Mortar and concrete reinforced with nanomaterials’, in ‘Nanotechnology in construction 3’, (ed. Bittnar

., 383–388; 2009, Berlin, Springer-Verlag.

39.

Morsy

, Alsayed

and Aqel

: ‘Effect of nano-clay on mechanical properties and microstructure of ordinary Portland cement mortar’, Int. J. Civil Environ. Eng., 2009, 10, (01), 23–27.

40.

Sobolev

, Flores

, Torres-Martinez

L. M.

, Valdez

P. L.

, Zarazua

and Cuellar

E. L.

: ‘Engineering of SiO₂ nanoparticles for optimal performance in nano cement-based materials’, Proc. 4th Int. Symp. on ‘Nanotechnology in construction. Agios Nikolaos, Crete, Greece; 2009..

41.

Chang

T. P.

, Shih

J. -Y.

, Yang

K. -M.

and Hsiao

T. -C.

: ‘Material properties of Portland cement paste with nano-montmorillonite’, J. Mater. Sci., 2007, 42, (17), 7478–7487. doi: 10.1007/s10853-006-1462-0

42.

Sobolev

and Sanchez

: ‘The application of nanoparticles to improve the performance of concrete’, Proc. 4th Int. Symp. on ‘Nanotechnology in construction’, Agios Nikolaos, Crete, Greece, May 2012, AEROCOINS.

43.

Lin

, Vivian

I. F.

and Sobolev

: ‘Nano-engineered cements with improved early strength’, Proc. 4th Int. Symp. on ‘Nanotechnology in construction’, Agios Nikolaos, Crete, Greece, May 2012, AEROCOINS..

44.

Balaguru

and Chong

: ‘Nanotechnology and concrete: research opportunities’, Proc. Session on ‘Nanotechnology of concrete: recent developments and future perspectives’, Denver, CO, USA, November 2006, ACI..

45.

, Xiao

and Ou

: ‘A study on mechanical and pressure-sensitive properties of cement mortar with nanophase materials’, Cem. Concr. Res., 2004, 34, (3), 435–438. doi: 10.1016/j.cemconres.2003.08.025

46.

, Xiao

, Yuan

and Ou

: ‘Microstructure of cement mortar with nano-particles’, Composites B, 2004, 35B, (2), 185–189. doi: 10.1016/S1359-8368(03)00052-0

47.

Papatzani

: ‘Nanotechnologically modified cements: effects on hydration, microstructure and physical properties’, PhD thesis, University of Bath, Bath, UK, 2014, 329.

48.

Raki

, Beaudoin

and Alizadeh

: ‘Nanotechnology applications for sustainable cement-based products’, in ‘Nanotechnology in construction 3’, (ed. Bittnar

., 119–124; 2009, Berlin, Springer-Verlag.

49.

Scrivener

: ‘Nanotechnology and cementitious materials’, in ‘Nanotechnology in construction 3’, (ed. Bittnar

., 37–42; 2009, Berlin, Springer-Verlag.

50.

B.-W.

, Kim

C. H.

, Tae

G. H.

and Park

J. B.

: ‘Characteristics of cement mortar with nano-SiO₂ particles’, Constr. Build. Mater., 2007, 21, (6), 1351–1355. doi: 10.1016/j.conbuildmat.2005.12.020

51.

Papatzani

, Paine

and Calabria-Holley

: ‘Strength and microstructure of colloidal nanosilica enhanced cement pastes’, Cem. Appl., 2014, 4, 80–85.

52.

and Shi

: ‘Chloride permeability and microstructure of Portland cement mortars incorporating nanomaterials’, Transp. Res. Rec., 2008, 2070, (1), 13–21. doi: 10.3141/2070-03

53.

Birgisson

and Dham

: ‘Optimization of clay addition for the enhancement of pozzolanic reaction in nano-modified cement paste’, in ‘Nanotechnology in civil infrastructure: a paradigm shift’, (ed. Gopalakrishnan

., 225–236; 2011, Berlin/Heidelberg, Springer.

54.

Calabria-Holley

, Paine

and Papatzani

: ‘Effects of nanosilica on the calcium silicate hydrates in Portland cement–fly ash systems’, Adv. Cem. Res., 2014, 26, (1), 1–14. doi: 10.1680/adcr.11.00052

55.

Birgisson

, Mukhopadhyay

A. K.

, Geary

, Khan

and Sobolev

: ‘Nanotechnology in concrete materials – a synopsis’, Transportation Research Circular E-C1702012, Transportation Research Board, Washington, DC, USA; 2012..

56.

Sobolev

and Gutiérrez

M. F.

: ‘How nanotechnology can change the concrete world.Part 1 of a two-part series’, Am. Ceram. Soc. Bull., 2005, 84, (10), 14.

57.

Papatzani

, Paine

and Calabria-Holley

: ‘The effect of the addition of nanoparticles of silica on the strength and microstructure of blended Portland cement pastes’, Proc. 2014 Int. Concrete Sustainability Conf., Boston, MA, USA, May 2014, NRMCA..

58.

Senff

, Hotza

, Repette

W. L.

, Ferreira

V. M.

and Labrincha

J. A.

: ‘Effect of nanosilica and microsilica on microstructure and hardened properties of cement pastes and mortars’, Adv. Appl. Ceram., 2010, 109, (2), 104–110. doi: 10.1179/174367509X12502621261659

59.

Lee

B. Y.

and Kurtis

K. E.

: ‘Influence of TiO₂ nanoparticles on early C₃S hydration’, J. Am. Ceram. Soc., 2010, 93, (10), 3399–3405. doi: 10.1111/j.1551-2916.2010.03868.x

60.

, Wang

, He

, Lu

and Wang

: ‘Investigations on the preparation and mechanical properties of the nano-alumina reinforced cement composite’, Mater. Lett., 2006, 60, (3), 356–359. doi: 10.1016/j.matlet.2005.08.061

61.

Rashad

A. M.

: ‘A comprehensive overview about the effect of nano-SiO₂ on some properties of traditional cementitious materials and alkali-activated fly ash’, Constr. Build. Mater., 2014, 52, 437–464. doi: 10.1016/j.conbuildmat.2013.10.101

62.

Kuo

W. Y.

, Huang

J. S.

and Lin

C. H.

: ‘Effects of organo-modified montmorillonite on strengths and permeability of cement mortars’, Cem. Concr. Res., 2006, 36, (5), 886–895. doi: 10.1016/j.cemconres.2005.11.013

63.

Papatzani

, Paine

and Calabria-Holley

: ‘Dispersed and modified montmorillonite clay nanoparticles for blended Portland cement pastes: effects on microstructure and strength’, Proc. 5th Int. Symp. on ‘Nanotechnology in construction, Springer. Available at: http://doi.org/10.1007/978-3-319-17088-6_16.

64.

Camiletti

, Soliman

A. M.

and Nehdi

M. L.

: ‘Effects of nano- and micro-limestone addition on early-age properties of ultra-high-performance concrete’, Mater. Struct., 2013, 46, (6), 881–898. doi: 10.1617/s11527-012-9940-0

65.

Liu

, Chen

, Liu

and Wang

: ‘Effect of nano-CaCO₃ on properties of cement paste’, Energy Procedia B, 2012, 16B, 991–996. doi: 10.1016/j.egypro.2012.01.158

66.

Fares

and Khan

M. I.

: ‘Nanosilica and its future prospects in concrete’, Adv. Mater. Res., 2013, 658, 50–55. doi: 10.4028/www.scientific.net/AMR.658.50

67.

Oertel

: ‘The influence of amorphous, sub-micrometer silica particles in cement pastes and mortars with very low water-to-cement ratios (ultra-high performance concrete)’; 2013, Würzburg, Der Fakultät für Biologie, Chemie und Geowissenschaften, Universität Bayreuth.

68.

Soleymani

: ‘Optimum content of SiO₂ nanoparticles in concrete specimens’, J. Am. Sci., 2012, 8, (4), 432–437.

69.

Singh

L. P.

, Bhattacharyya

S. K.

, Singh

and Ahalawat

: ‘Granulometric synthesis and characterisation of dispersed nanosilica powder and its application in cementitious system’, Adv. Appl. Ceram., 2012, 111, (4), 220–227. doi: 10.1179/1743676112Y.0000000002

70.

Napierska

, Thomassen

L. C. J.

, Lison

, Martens

J. A.

and Hoet

P. H.

: ‘The nanosilica hazard: another variable entity’, Part. Fibre Toxicol., 2010, 7, (1), 39. doi: 10.1186/1743-8977-7-39

71.

Lazaro

, Brouwers

H. J. H.

, Quercia

and Geus

J. W.

: ‘The properties of amorphous nano-silica synthesized by the dissolution of olivine’, Chem. Eng. J., 2012, 211–212, 112–121. doi: 10.1016/j.cej.2012.09.042

72.

Sobolev

and Gutiérrez

M. F.

: ‘How nanotechnology can change the concrete world. Part 2 of a two-part series’, Am. Ceram. Soc. Bull., 2005, 84, (11), 16–19.

73.

Stefanidou

and Papayianni

: ‘Influence of nano-SiO₂ on the Portland cement pastes’, Composites B, 2012, 43B, (6), 2706–2710. doi: 10.1016/j.compositesb.2011.12.015

74.

, Zhang

, Kong

and Chen

: ‘Influence of nano-SiO₂ addition on properties of hardened cement paste as compared with silica fume’, Constr. Build. Mater., 2007, 21, (3), 539–545. doi: 10.1016/j.conbuildmat.2005.09.001

75.

, Zhang

, Sheng

and Chen

: ‘A comparative study on the pozzolanic activity between nano-SiO₂ and silica fume’, J. Wuhan Univ. Technol., 2006, 21, (3), 153–157. doi: 10.1007/BF02840907

76.

Mondal

, Shah

S. P.

, Marks

L. D.

and Gaitero

J. J.

: ‘Comparative study of the effects of microsilica and nanosilica in concrete’, Transport. Res. Rec., 2010, 2141, 6–9. doi: 10.3141/2141-02

77.

Kawashima

, Hou

, Corr

D. J.

and Shah

S. P.

: ‘Modification of cement-based materials with nanoparticles’, Cem. Concr. Compos., 2013, 36, 8–15. doi: 10.1016/j.cemconcomp.2012.06.012

78.

Kawashima

, Hou

, Corr

D. J.

, and Shah

S. P.

: ‘Modifications of cement-based materials with nanoparticles’, Cem. Concr. Compos., 2012, 36, 8–15. doi: 10.1016/j.cemconcomp.2012.06.012

79.

Oltulu

and S¸ahin

: ‘Pore structure analysis of hardened cement mortars containing silica fume and different nano-powders’, Constr. Build. Mater., 2014, 53, 658–664. doi: 10.1016/j.conbuildmat.2013.11.105

80.

Pignat

, Navi

and Scrivener

: ‘Simulation of cement paste microstructure hydration, pore space characterization and permeability determination’, Mater. Struct., 2005, 38, (4), 459–466. doi: 10.1007/BF02482142

81.

Hou

: ‘Modification effects of colloidal nanoSiO₂ on cement hydration and its gel property’, Composites B, 2013, 45B, (1), 440–448. doi: 10.1016/j.compositesb.2012.05.056

82.

B. W.

, Kim

C. -H.

and Lim

J. -H.

: ‘Characteristics of cement mortar with nano-SiO₂ particles’, Constr. Build. Mater., 2007, 21, (6), 1351–1355. doi: 10.1016/j.conbuildmat.2005.12.020

83.

Bhattarai

: ‘Mechanical properties of nanoclay based nanocomposites: a review’; 2006, Pittsburg, KS, Pittsburg State University.

84.

Elkady

, Serag

M. I.

and Elfeky

M. S.

: ‘Effect of Nano silica de-agglomeration, and methods of adding super-plasticizer on the compressive strength, and workability of nano silica concrete’, Civil Environ. Res., 2013, 3, (2), 21–35.

85.

Sobolev

, Flores

, Torres Martinez

L. M.

, Valdez

P. L.

, Zarazua

and Cuellar

E. L.

: ‘Engineering of SiO₂ nanoparticles for optimal performance in nano cement-based materials’, in ‘Nanotechnology in construction 3’, (ed. Bittnar

., 139–148; 2009, Berlin/Heidelberg, Springer.

86.

: ‘Properties of high-volume fly ash concrete incorporating nano-SiO₂’, Cem. Concr. Res., 2004, 34, (6), 1043–1049. doi: 10.1016/j.cemconres.2003.11.013

87.

Sanchez

and Sobolev

: ‘Nanotechnology in concrete – a review’, Constr. Build. Mater., 2010, 24, (11), 2060–2071. doi: 10.1016/j.conbuildmat.2010.03.014

88.

Senff

, Hotza

, Repette

W. L.

, Ferreira

V. M.

and Labrincha

J. A.

: ‘Mortars with nano-SiO₂ and micro-SiO₂ investigated by experimental design’, Constr. Build. Mater., 2010, 24, (8), 1432–1437. doi: 10.1016/j.conbuildmat.2010.01.012

89.

: ‘Synthesis, characterisation and application of organoclays’; 2006, Brisbane, Qld, School of Physical and Chemical Sciences–Inorganic Material Research Group, Queensland University of Technology.

90.

Utracki

L. A.

: ‘Clay-containing polymeric nanocomposites’ Vol. 1, ; 2004, Shawbury, Rapra Technology Limited.

91.

De Paiva

L. B.

, Morales

A. R.

and Valenzuela Díaz

F. R.

: ‘Organoclays: properties, preparation and applications’, Appl. Clay Sci., 2008, 42, (1), 8–24. doi: 10.1016/j.clay.2008.02.006

92.

Briell

: ‘Nanoclays – counting on consistency’, Proc. Conf. Nanocomposite 2000, Gonzales, TX, USA, 2000.

93.

Uddin

: ‘Clays, nanoclays, and montmorillonite minerals’, Metall. Mater. Trans. A, 2008, 39A, (12), 2804–2814. doi: 10.1007/s11661-008-9603-5

94.

Dalir

, Farahani

R. D.

, Nhim

, Samson

, Lévesque

and Therriault

: ‘Preparation of highly exfoliated polyester–clay nanocomposites: process–property correlations’, Langmuir, 2012, 28, 791–803. doi: 10.1021/la203331h

95.

Sapalidis

A. A.

, Katsaros

F. K.

and Kanellopoulos

N. K.

: ‘PVA/montmorillonite nanocomposites: development and properties’, in ‘Nanocomposites and polymers with analytical methods’, (ed. Cuppoletti

); 2011, IntechOpen (e-book, available at: http://doi.org/10.5772/18217 [accessed10 January 2015].

96.

Jahromi

S. G.

, Andalibizade

and Vossough

: ‘Engineering properties of nanoclay modified asphalt concrete mixtures’, Arab. J. Sci. Eng. B, 2010, 35, (1B), 90.

97.

Batra

, Gotam

, Dadarwal

, Nainwani

and Sharma

: ‘Nano-clay as polymer porosity reducer: a review’, J. Pharm. Sci. Technol., 2011, 3, (10), 709–716.

98.

Vazquez

, López

, Kortaberria

, Martín

and Mondragon

: ‘Modification of montmorillonite with cationic surfactants. Thermal and chemical analysis including CEC determination’, Appl. Clay Sci., 2008, 41, (1–2), 24–36. doi: 10.1016/j.clay.2007.10.001

99.

Viseras

, Aguzzi

, Cerezo

and Bedmar

M. C.

: ‘Biopolymer–clay nanocomposites for controlled drug delivery’, Mater. Sci. Technol., 2008, 24, (9), 1020–1026. doi: 10.1179/174328408X341708

100.

Annamalai

and Martin

: ‘Can clay nanoparticles accelerate environmental biodegradation of polyolefins?’, Mater. Sci. Technol., 2014, 30, (5), 593–602. doi: 10.1179/1743284713Y.0000000498

101.

Hetzer

, Zhou

H. X.

, Poloso

and De Kee

: ‘Influence of compatibiliser blends on mechanical and thermal properties of polymer–clay nanocomposites’, Mater. Sci. Technol., 2011, 27, (1), 53–59. doi: 10.1179/174328409X418964

102.

Meng

, Wang

and Tang

: ‘Controlling dispersed state and exfoliation process of clay in polymer matrix’, Mater. Sci. Technol., 2006, 22, (7), 780–786. doi: 10.1179/174328406X101238

103.

Aktas

and Altan

: ‘Effect of nanoclay content on properties of glass–waterborne epoxy laminates at low clay loading’, Mater. Sci. Technol., 2010, 26, (5), 626–629. doi: 10.1179/174328409X428864

104.

Bandyopadhyay

, Maiti

and Bhowmick

: ‘Synthesis, characterisation and properties of clay and silica based rubber nanocomposites’, Mater. Sci. Technol., 2006, 22, (7), 818–828. doi: 10.1179/174328406X101265

105.

Farzadnia

, Ali

A. A. A.

, Demirbogaa

and Anwar

M. P.

: ‘Effect of halloysite nanoclay on mechanical properties, thermal behavior and microstructure of cement mortars’, Cem. Concr. Res., 2013, 48, 97–104. doi: 10.1016/j.cemconres.2013.03.005

106.

Zhang

, Fan

and Li

: ‘Pore structure and freezing resistance of nanoclay modified cement based materials’, Mater. Res. Innov., 2014, 18, (S2), S2-358–S2-362. doi: 10.1179/1432891714Z.000000000434

107.

Luckham

P. F.

and Rossi

: ‘The colloidal and rheological properties of bentonite suspensions’, Adv. Colloid Interface Sci., 1999, 82, (1–3), 43–92. doi: 10.1016/S0001-8686(99)00005-6

108.

, Martens

, He

and Frost

R. L.

: ‘Thermogravimetric analysis of organoclays intercalated with the surfactant octadecyltrimethylammonium bromide’, J. Therm. Anal. Calorim., 2005, 81, (1), 91–97. doi: 10.1007/s10973-005-0750-2

109.

Kuo

W.-Y.

, Huang

J. -S.

and Yu

B. -Y.

: ‘Evaluation of strengthening through stress relaxation testing of organo-modified montmorillonite reinforced cement mortars’, Constr. Build. Mater., 2011, 25, (6), 2771–2776. doi: 10.1016/j.conbuildmat.2011.01.001

110.

Sinha Ray

and Okamoto

: ‘Polymer/layered silicate nanocomposites: a review from preparation to processing’, Prog. Polym. Sci., 2003, 28, (11), 1539–1641. doi: 10.1016/j.progpolymsci.2003.08.002

111.

LeBaron

P. C.

, Wang

and Pinnavaia

T. J.

: ‘Polymer-layered silicate nanocomposites: an overview’, Appl. Clay Sci., 1999, 15, (1), 11–29. doi: 10.1016/S0169-1317(99)00017-4

112.

Tregger

N. A.

, Pakula

M. E.

and Shah

S. P.

: ‘Influence of clays on the rheology of cement pastes’, Cem. Concr. Res., 2010, 40, (3), 384–391. doi: 10.1016/j.cemconres.2009.11.001

113.

Hosseini

, Hosseinpourpia

, Pajum

, Khodavirdi

M. M.

, Izadi

and Vaezi

: ‘Effect of nano-particles and aminosilane interaction on the performances of cement-based composites: an experimental study’, Constr. Build. Mater., 2014, 66, 113–124. doi: 10.1016/j.conbuildmat.2014.05.047