Acta Polytechnica doi:10.14311/APP.2020.27.0101 Acta Polytechnica 27(0):101–106, 2020 © Czech Technical University in Prague, 2020 available online at http://ojs.cvut.cz/ojs/index.php/app MICROMECHANICAL CHARACTERIZATION OF CEMENT PASTE MODIFIED BY NANOCLAYS Vojtěch Zacharda∗, Jiří Němeček Czech Technical University in Prague, Faculty of Civil Engineering, Department of Mechanics, Thákurova 7, 166 29 Prague 6, Czech Republic ∗ corresponding author: vojtech.zacharda@fsv.cvut.cz Abstract. This work deals with the comparison of micromechanical properties and microstructure of cement pastes with additives used in concrete to reduce the lateral pressure on the formwork. This work is a steppingstone for broader investigation of the lateral pressures on different cement mixtures. The work focused on two additives used for the purpose, namely calcined clay (metakaolin) and a type of nanoclay sepiolite. A scanning electron microscope was used to describe their microstructure. Micromechanical properties of both cement composites were investigated by nanoindentation. Large statistical grids of indents were performed on three sample types: plain cement paste and two mixtures containing the enhancing additive of metakaolin and nanoclay. From the evaluated results in the form of property histograms, the modulus of elasticity, hardness and creep parameter were derived. It was found that in the cement paste with metakaolin the amount of C-S-H gel increased compared to the control mixture. Increased portlandite and the amount of unhydrous clinker was found in the cement paste with nanoclay. Nanomechanical response of individual phases was derived from overall property histograms by statistical deconvolution. The results were confirmed by electron microscopy. The micromechanical research was supplemented with the measurement of the compressive strength on cubes at the macroscopic level. Keywords: Cement, hydration, metakaolin, nanoclay, nanoindentation. 1. Introduction In the construction industry, concretes in their fresh state are cast into formworks. Ordinary concretes are vibrated to support their flow and to overcome their high viscosity. The casting process of tall struc- tures is done in steps to allow mixture compaction and stepwise vibration. Lateral pressures introduced by the ordinary concretes on formwork diminish rela- tively soon after casting due to the very high mixture viscosity and prolonged casting times. In contrary, self-compacting concretes (SCC) are characterized by low viscosity with no need of additional vibration to fill the formwork [1–4]. SCC are usually poured into formwork in one step filling the whole height of the structure (e.g. a column or a wall). The long-lasting mixture flowability is usually maintained by additives (plasticizers) and hydration retarders and even after casting the mixture behaves like a fluid that introduces high lateral pressures on the formwork. However, the development of high lateral pressures in the SCC mixtures can be controlled by special ad- ditives that control viscosity and workability of the mixtures in the fresh state and simultaneously con- trol development of thixotropic behavior and early strength development of the mixture. Such additives can be in the form of hydration modifiers or some nanoparticles [1–6]. So far, the effect of the additives on the microstructure and micromechanical proper- ties of the resulting hardened concrete was not well studied. Thus, this work deals with the comparison of micromechanical properties [7–10] and microstruc- ture of cement pastes modified with additives that are used in SCCs to reduce the lateral pressure on the formwork [1–6]. There are several ways how to reduce the lateral pressure in the mixture. One of the ways is to change thixotropy of the concrete by non-reactive additives. It was found that the thixotropy can be increased and the pressure can be reduced by nanoclay min- erals or derived materials [1–6]. The other way of reducing pressures is to modify hydration kinetics and hydration products by additives such as metakaolin. Metakaolin reacts with dissolved clinker minerals in cement paste and forms calcium-silica hydrates (C-S- H) [6]. 2. Tested materials Cement pastes modified by two additives, metakaolin and nanoclay, were tested in hardened state in this work. Firstly, metakaolin which is a calcined clay with the chemical composition: 53.1% SiO2, 41.7% Al2O3, 1.1% Fe2O3 and other minor oxides, was used. It is prepared as a granulated powder with a similar grain distribution as cement. Metakaolin particles have angular shape with mean size which can range from 1 to 10 µm (Figure 1). Metakaolin reacts with cement during hydration. It reduces the amount of portlandite in the hydrated cement and increases the main hydration product that is calcium-silica-hydrate gel (C-S-H) [6]. The nanoclay sepiolite was used for 101 http://dx.doi.org/10.14311/APP.2020.27.0101 http://ojs.cvut.cz/ojs/index.php/app Vojtěch Zacharda, Jiří Němeček Acta Polytechnica Sample Cement [g] Metakaolin [g] Sepiolite [g] Water [g] C 450 − − 202.5 CM 445.5 4.5 − 202.5 CS 445.5 − 4.5 202.5 Table 1. Composition of cement pastes. Figure 1. SEM image of metakaolin. the second part of the samples. Its chemical formula is Mg4Si6O15(OH)2·6H2O. Sepiolite particles have tiny elongated shape (Figure 2). They have typical length of 1 to 2 µm and width of several tens of nanometers. Sepiolite does not dissolve in water and it does not react with cement, but it works on reorganization of the crystallites in plastic stage of the mixture. It also binds more water upon mixing which lowers worka- bility. The resulting microstructure contains larger amount of unhydrous clinker. The micromechanical properties of the three ce- ment paste mixtures with the composition defined in Table 1 were tested. All mixtures are based on Port- land cement CEM I 42,5R and the additives described earlier. Equal water to binder ratio 0,45 was used for all samples. The amount of additives was chosen as 1% of weight of cement. Samples were cast into cylindrical plastic molds with height of 65 mm and diameter of 30 mm. They were demolded after three days from casting and stored in water for additional 25 days. After hardening, the samples were cut into discs and placed in an acetone bath to prevent fur- ther hydration. The surface of samples were grinded and polished by a metallographic procedure and pre- pared for testing by nanoindentation. Along, six cubic specimens (40x40x40 mm) were made for compressive strength testing. Figure 2. SEM image of sepiolite. 3. Methods 3.1. Nanoindentation The main objective of the paper was to compare nanomechanical properties of the cement paste modi- fied by metakaolin or sepiolite with pure cement paste. Therefore, series of 400 indents were done on each sam- ple. Each consisted of a matrix of 20x20 indents. The separation of indents was 10 µm. A load-controlled test with the load function lasting for 26 seconds to the maximum force of 2000 µN was prescribed for each indent. The function had a trapezoidal shape (Figure 3). The first part of the function was a linear loading with speed of 40 mN/min lasting for 3 sec- onds. The second part was a constant loading for 20 seconds and the last part of the function was a linear unloading with speed of 40 mN/min lasting for 3 seconds. A typical response (load-depth) for various phases in the material is shown in Figure 5. Material properties were evaluated for each indent by the Oliver and Pharr method [11]. Some defective indentations caused by local porosity or roughness were eliminated from considerations. The Oliver and Pharr [11] method was used for obtaining reduced modulus, Er and hardness, H for single indents and property histograms were constructed for each matrix. The evaluation was complemented by creep indenta- tion parameter CIT . The parameters are defined as 102 vol. 27/2020 Cement paste modified by nanoclays Figure 3. Load function of indents. Figure 4. Topography of the surface. follows. Er = S √ π 2β √ Ac (1) where S is the unloading tangential stiffness, β is a tip shape correction coefficient and Ac is the contact area, H = Pmax Ac (2) where Pmax is the maximum indentation force, CIT(P,t1,t2) = h2 − h1 h1 × 100 (3) which is defined as a relative change between indenta- tion depths h1 encountered at the time t1 and h2 at the time t2, respectively. The CIT is dependent on the contact force P and time of holding period. Note, that the reduced modulus can easily be used to calculate Young’s modulus of isotropic materi- als [11] knowing Poisson’s ratio of the material (ν=0.2 was assumed for all phases in this work). Figure 5. Indentation curves. Figure 6. BSE images (phase composition). 3.2. Microstructural analysis The basic microstructural analysis was performed by scanning electron microscope (SEM). Back scattered electrons (BSE) images together with energy disper- sive x-ray spectroscopy (EDS) analysis are very effec- tive way for exploring the phase of cement paste at the level of micrometers. Topography of the surface (Figure 4), BSE images (phase composition) (Figure 6) and EDS maps of cement pastes were acquired. The distinction between the microstructural phases could be done from these images but it was not attempted in this work that concentrated on statistical evaluation of micromechanical results as shown later. 3.3. Macroscopic level The paper was supplemented by the compressive strength measurements performed in a standard electro-mechanical press on 40x40x40 mm cubes (Fig- ure 7). 103 Vojtěch Zacharda, Jiří Němeček Acta Polytechnica Figure 7. Test of cement paste cube for compressive strength. 4. Results and discussion 4.1. Modulus of elasticity Figure 8 shows histograms of modulus of elasticity for mixtures of cement pastes. Values were calcu- lated from merged results of nanoindentation. The histograms show a similar trend for all mixtures. The main peak in CM sample appears between 11 to 51 GPa, between 16 to 66 GPa on CS mixture and 13 to 61 GPa on C sample, respectively, with the mean of 25 GPa for CS and C and 22 GPa for CM (Figure 8). The results show that the wrapping curve of elastic modulus for CM has shifted to the left compared to the control mixture. This confirms the assumption that the amount of portlandite in the hydrated cement is reduced and the calcium-silica-hydrate gel (C-S-H) content increases. The results of CS mixture show similar values compared to the control mixture C. 4.2. Hardness Figure 9 shows histograms of hardness of cement pastes. Values were calculated from merged results of nanoindentation. Similarly, to results of E, the histograms show similar trend for all mixtures. The main peak of hardness measured on CM and C is around 1.0 GPa. In CS mixture the hardness has the main peak around 0.9 GPa, thus similar. It can be concluded that the results are qualitatively as well as quantitatively equivalent. 4.3. Creep indentation parameter CIT Figure 10 shows histograms of CIT of cement pastes. Values were calculated from merged results of nanoin- dentation. Again, similar trend for all mixtures was encountered. The highest mean CIT value is exhibited by the mixture with metakaolin suggesting the role of increased C-S-H content on increased creep. How- ever, the differences between the samples are not large and the mixtures can be considered as equivalently creeping. 4.4. Compressive strength The results of compressive strength were measured on the six samples cubes for each mixture. Average value was calculated and was supplemented by standard deviation. The average compressive strength of control mixture (C) was 65,9 ± 5,7 MPa. For mixture with sepiolite (CS) the average compressive strength was 63,6 ± 1,1 MPa. The average compressive strength of mixture with metakaolin (CM) was 63,1 ± 3,4 MPa. It can be concluded that the average compressive strength of mixtures with additives were at about 95% of the control mixture. Thus, the additives of metakaolin and sepiolite do not have considerable effect on macromechanical properties. 4.5. EDS analysis On the samples EDS analysis map were made (Fig- ure 11). The results are shown in Table 2. Comparison of results of mixture from EDS analysis shows small change in volume concentration of chemical phases except carbon that is largely present in CS samples. This is a consequence of organic pollution of the raw sepiolite used for sample preparation. Element C CM CS W. C. [%] W. C. [%] W. C.[%] O 72.63 68.74 59.17 Ca 15.89 20.21 17.78 Si 5.98 7.07 6.24 Al 3.28 2.11 1.82 Mg 1.4 0.89 0.53 S 0.82 0.97 0.48 C - - 12.95 Table 2. Weight concentrations of elements. 5. Conclusions The presented paper provides nanomechanical and microstructural results of cement pastes with special additives in the form of nanoclays and calcined clay minerals used for the control of lateral pressures in fresh mixtures. Nanoindentation was used to obtain microscale hardness, elastic and creep properties. Val- ues of Young’s moduli, hardness and creep indentation parameter, CIT , were calculated from large statisti- cal sets of nanoindentation. The results of elastic moduli indicate a statistically significant shift towards lower values on CM samples. This is a consequence of the hydration process where the amount of port- landite in the hydrated cement is reduced and the main hydration product, the calcium-silica-hydrate gel increases. The sepiolite in CS samples does not 104 vol. 27/2020 Cement paste modified by nanoclays Figure 8. Histogram of modulus of elasticity with trend line for cement pastes CS, CM and C. Figure 9. Histogram of hardness with trend line for cement pastes CS, CM and C. Figure 10. Histogram of CIT with trend line for cement pastes CS, CM and C. 105 Vojtěch Zacharda, Jiří Němeček Acta Polytechnica Figure 11. Maps of chemical elements from EDS analysis: a) C, b) CM and c) CS. appear to statistically influence elastic modulus on microscale. The hardness and creep were also found statistically equivalent on all samples. The average compressive strength of modified sam- ples was found to be 95% of the control mixture mean- ing very little macroscopic influence of additives was observed. EDS analysis shows small change in concen- trations of chemical elements except the presence of carbon in CS samples which is a consequence of the raw material organic pollution. The work described in this paper confirmed me- chanical compatibility of the additives with nil or very little influence on micro and macroscopic mechani- cal properties of the hardened samples. The work is a steppingstone for future lateral pressure investiga- tions. Acknowledgements This work was performed under the support of Tech- nological Agency of the Czech Republic (Trend FW- 01010521) and Czech Technical University in Prague (project SGS20/107/OHK1/2T/11). References [1] J. H. Kim, N. Noemi, S. P. Shah. Effect of powder materials on the rheology and formwork pressure of self-consolidating concrete. Cement and Concrete Composites 34(6):746 – 753, 2012. doi:10.1016/j.cemconcomp.2012.02.016. [2] N. Roussel. A thixotropy model for fresh fluid concretes: Theory, validation and applications. Cement and Concrete Research 36(10):1797 – 1806, 2006. doi:10.1016/j.cemconres.2006.05.025. [3] S. Kawashima, J. H. Kim, D. J. Corr, S. P. Shah. Study of the mechanisms underlying the fresh-state response of cementitious materials modified with nanoclays. Construction and Building Materials 36:749 – 757, 2012. doi:10.1016/j.conbuildmat.2012.06.057. [4] J. H. Kim, M. Beacraft, S. P. Shah. Effect of mineral admixtures on formwork pressure of self-consolidating concrete. Cement and Concrete Composites 32(9):665 – 671, 2010. doi:10.1016/j.cemconcomp.2010.07.018. [5] J. P. Melo, A. S. Aguilar, F. H. Olivares. Rheological properties of aerated cement pastes with fly ash, metakaolin and sepiolite additions. Construction and Building Materials 65:566 – 573, 2014. doi:10.1016/j.conbuildmat.2014.05.034. [6] R. Siddique, J. Klaus. Influence of metakaolin on the properties of mortar and concrete: A review. Applied Clay Science 43(3):392 – 400, 2009. doi:10.1016/j.clay.2008.11.007. [7] V. Nezerka, J. Nemecek, Z. Slizkova, P. Tesarek. Investigation of crushed brick-matrix interface in lime- based ancient mortar by microscopy and nanoindentation. Cement and Concrete Composites 55:122 – 128, 2015. doi:10.1016/j.cemconcomp.2014.07.023. [8] J. Nemecek, V. Kralik, V. Smilauer, et al. Tensile strength of hydrated cement paste phases assessed by micro-bending tests and nanoindentation. Cement and Concrete Composites 73:164 – 173, 2016. doi:10.1016/j.cemconcomp.2016.07.010. [9] S. Diamond, J. Huang. The itz in concrete - a different view based on image analysis and sem observations. Cement and Concrete Composites 23(2):179 – 188, 2001. Special Theme Issue on Image Analysis, doi:10.1016/S0958-9465(00)00065-2. [10] J. Nemecek, V. Kralik, J. Vondrejc. Micromechanical analysis of heterogeneous structural materials. Cement and Concrete Composites 36:85 – 92, 2013. Special issue: Nanotechnology in Construction, doi:10.1016/j.cemconcomp.2012.06.015. [11] W. C. Oliver, G. M. Pharr. An improved technique for determining hardness and elastic modulus using load and displacement sensing indentation experiments. Journal of materials research 7(6):1564–1583, 1992. doi:10.1557/JMR.1992.1564. 106 https://doi.org/10.1016/j.cemconcomp.2012.02.016 https://doi.org/10.1016/j.cemconres.2006.05.025 https://doi.org/10.1016/j.conbuildmat.2012.06.057 https://doi.org/10.1016/j.cemconcomp.2010.07.018 https://doi.org/10.1016/j.conbuildmat.2014.05.034 https://doi.org/10.1016/j.clay.2008.11.007 https://doi.org/10.1016/j.cemconcomp.2014.07.023 https://doi.org/10.1016/j.cemconcomp.2016.07.010 https://doi.org/10.1016/S0958-9465(00)00065-2 https://doi.org/10.1016/j.cemconcomp.2012.06.015 https://doi.org/10.1557/JMR.1992.1564 Acta Polytechnica 27(0):101–106, 2020 1 Introduction 2 Tested materials 3 Methods 3.1 Nanoindentation 3.2 Microstructural analysis 3.3 Macroscopic level 4 Results and discussion 4.1 Modulus of elasticity 4.2 Hardness 4.3 Creep indentation parameter CIT 4.4 Compressive strength 4.5 EDS analysis 5 Conclusions Acknowledgements References