From microstructure to macrostructure: an integrated model of structure formation in polymer-modified concrete

Available online at www.rilem.net Materials and Structures 38 (July 2005) 601-607 From microstructure to macrostructure: an integrated model of structure formation in polymer-modified concrete A. Beeldens 1, D. Van Gemert 2, H. Schorn 3, Y. Ohama 4 and L. Czarnecki 5 (1) Belgian Road Research Centre, Belgium (2) Department of Civil Engineering, Katholieke Universiteit Leuven, Belgium (3) Department of Building Materials, University of Technology, Dresden, Germany (4) Department of Architecture, College of Engineering, Nihon University, Koriyama, Japan. (5) Institute of Technology, Faculty of Civil Engineering, Warsaw University of Technology, Warsaw, Poland Received: 31 January 2004; accepted: 17 November 2004 ABSTRACT A model is proposed for the formation of the microstructure in polymer-modified cementitious materials. Cement hydration and polymer film formation were studied, with specific emphasis on the synergetic effect between cement particles and polymer particles. Alterations at the microstructure level result in macroscopic changes in the properties of the modified material. In this paper, the influence of the polymer addition on the appearance of the cement hydrates and the presence of the polymer film through the cement hydrates are presented in relation to the minimum film forming temperature. Owing to the presence of the cement particles and to cement hydration, film formation can take place at lower temperatures, so that a polymer dispersion with a slightly higher MFT (minimum film forming temperature) can be used. This is important for the physical and mechanical properties of the polymer-modified materials. The findings have been included in an integrated model based on the threestep model of Ohama, in which the polymer film formation and the cement hydration processes are integrated in relation to each other. A timedependent evaluation of both processes was incorporated. The research presented in this paper was part of a PhD research at the Civil Engineering Department, University of Leuven, Belgium [1]. 1359-5997 2004 RILEM. All rights reserved. RÉSUMÉ L auteur présente un modèle de la formation de la microstructure dans les matériaux hydrauliques modifiés aux polymères. L étude porte non seulement sur l hydratation du ciment, ainsi que la formation du film polymérique, mais aussi spécifiquement sur les effets synergétiques entre les particules de ciment et les particules de polymères. Les altérations au niveau de la microstructure entraînent des changements macroscopiques des propriétés du matériau modifié. Cette communication présente l influence de l ajout de polymères sur l aspect des hydrates de ciment et la présence du film polymérique parmi les hydrates de ciment, en relation avec la température minimale de formation du film. En raison de la présence des particules de ciment et de l hydratation du ciment, le film peut se former à des températures plus basses, si bien qu on peut utiliser une dispersion polymérique ayant une température minimale de formation du film légèrement supérieure. Cela est important pour les propriétés physiques et mécaniques du matériau modifié aux polymères. Les résultats sont synthétisés dans un modèle intégré, basé sur le modèle à trois niveaux d Ohama, dans lequel sont intégrés les processus de formation du film et d hydratation du ciment, l un par rapport à l autre. Une évaluation des deux processus en fonction du temps est incorporée. La recherche présentée dans cette communication faisait partie d une thèse de doctorat au département d ingénierie civile de l Université catholique de Louvain en Belgique [1]. 1. INTRODUCTION Polymer modification of cementitious materials is frequently used nowadays in restoration and repair works and in specific situations where high demands are made on adhesion, durability and weatherability. The properties of the composite material derive from those of its constituents, but there is also a synergetic effect. The super-structurality, i.e., Editorial Note The Belgian Road Research Centre is a RILEM Associate Member. Dr. Ir. Anne Beeldens participates in RILEM TCs 192-ECM Environmentconscious construction materials and systems and 194-TDP Application of Titanium dioxide photocatalysis to construction materials. Prof. Dr. Dionys Van Gemert, Prof. Dr.-Ing. Harald Schorn, Prof. Dr. Yoshihiko Ohama and Prof. Lech Czarnecki are RILEM Senior Members. They all participate in RILEM TC 184-IFE Industrial floors for withstanding harsh environmental attacks, including repair and maintenance. Prof. Van Gemert is the Secretary of both RILEM TCs 192-ECM and 194-TDP. Prof. Ohama is the Chairman of RILEM TC 194-TDP and he also participates in RILEM TC 192-ECM. 1359-5997 2004 RILEM. All rights reserved. doi:10.1617/14215

602 A. Beeldens et al. / Materials and Structures 38 (2005) 601-607 the geometric distribution of the different phases in the volume of the material, allows sound reasoning about variation of the different phases in relation to mechanical and physical properties. However, it does not consider the interaction between the different phases and, therefore, makes it difficult to incorporate e.g. increased adhesion strength between filler and matrix as is the case in polymer-modified concrete. The formation of the microstructure of polymer-modified cement mortar and concrete is described by various authors [2-5]. The process of polymer film formation coincident with cement hydration is described in various qualitative models. The basic model was proposed by Ohama [3]. Several adjustments and refinements of this model are presented. Based on Ohama s model combined with the adjustments and additional experimental work, an integrated model is presented in which the reciprocal influences between the polymer and the cement particles are incorporated and implemented on a time scale. An experimental program was set up to investigate the mutual influence between the polymer particles and the cement hydrates. Altered conditions for film formation and minimum film forming temperature (MFT) were studied, i.e., higher relative humidity, dilution in an alkaline pore solution, and a decreased drying rate. Another important item was to position the polymer film or polymer particles throughout the cement hydrates. Film formation is influenced by the presence of the cementitious products and, conversely, the cement particles and the cement hydration process are altered by the presence of the polymer particles and film. Changes are observed in the fresh mixture, during hydration and in the hardened material. 2. POLYMER FILM FORMATION IN THE PRESENCE OF CEMENTITIOUS MATERIAL Three important alterations of the curing conditions for the polymer dispersion can be observed when comparing the film formation process of the pure polymer emulsion and the film formation process in the presence of cement particles and cement hydrates. A high relative humidity is present owing to curing in moisture or water, especially during the first days. This results in an altered drying rate. A large dilution of the dispersion in a solution, which turns into an alkaline solution, occurs immediately after the different compounds have been mixed. The influence of relative humidity (RH) on the film formation process is closely related to the influence of the drying rate on film formation. The lower the drying rate, the lower the temperature is at which a continuous film is formed. This is related to the thermodynamic energy of the polymer particles and is comparable to the influence of the drying rate on the crystallization process of semi-crystalline materials. In the case of a modified cementitious material, loss of water occurs as a result of hydration of the cement particles and of evaporation. This depends not only on RH and on the temperature of the surrounding atmosphere, but also on the porosity of, and the pore size distribution in, the material itself, as the loss of water becomes diffusion-controlled. 2.1 Influence of relative humidity on film formation and drying rate To investigate the influence of RH on the film forming capacity of the polymer dispersion, three different curing conditions were considered: curing at a RH higher than 98%, a RH of 86% and a RH of 60%. An atmosphere with 86% RH was obtained in a sealed desiccator partly filled with a saturated KCl solution. The temperature was 20 ± 3 C. Fig. 1 presents the results for a SBR polymer dispersion. The pure polymer dispersion was tested as well as a diluted solution containing 50% of polymer dispersion, i.e., 25% of solids, 25% of emulsion water and 50% of tap water. Samples were prepared by pouring a thin layer of the polymer dispersion on a glass plate. The amount of solids poured on the glass plate was kept constant to obtain an equal thickness for all samples. The results indicated a rapid linear decrease in weight, due to the evaporation of the water until only the solid part of the dispersion was left. The results for the various dispersions reflected a constant drying rate until the 50 and 25% of relative weight, respectively, was reached, corresponding to the solid weight of the polymer dispersion added. Very little change in drying rate due to the dilution of the polymer dispersion was observed when comparing the drying rates of the 100-% polymer dispersion and the diluted 50-% solution. The increased relative humidity, on the other hand, strongly reduced the drying rate of the polymer dispersion. This resulted in a film forming capacity at a lower temperature than the minimum film forming temperature, as the polymer particles were able to come closer to each other during the slow evaporation of the water. 2.2 Influence of alkalinity The high alkalinity which appears in the pore water during cement hydration could cause interaction with the surfactants and result in delayed or even completely stopped film formation. To test the influence of the alkalinity of the pore solution, a simulation was made with a pure dispersion. The SBR dispersion was mixed with a NaOH solution having a ph value of 13. The solution contained 50% of polymer dispersion and had a ph of 12.7. The solution was poured on a glass plate and stored at 60 and 86% RH, respectively, and at Relative weight (%) 100 90 80 70 60 50 40 30 20 10 SBR 100%/R.H.60% SBR 100%/R.H.86% SBR 50%-water 50%/R.H.60% SBR 50%-water 50%/R.H.86% 0 0 5 10 15 20 25 30 35 40 45 50 Time (hour) Fig. 1 - Weight variation of polymer dispersion after curing at 60 and 86% RH.

A. Beeldens et al. / Materials and Structures 38 (2005) 601-607 603 20 C. The weight variations are presented in Fig. 2. The results indicate a small reduction in drying rate for the dispersions diluted with the NaOH solution as well as for the dispersion diluted with the H 2 SO 4 solution. However, the effect was very limited. In all cases a continuous film was formed. Again, RH had a much greater influence than the ph value of the solution. Care has to be taken with experimental dispersions, however, since the SBR dispersion was developed for use in combination with cement and had, therefore, a good resistance to the alkaline solution. 2.3 Appearance of the polymer film in the binder matrix The variations observed in the film forming process in the pore solution of cementitious materials can be ascribed to the influence of the altered conditions on the drying rate of the polymer dispersion. In addition, owing to cement hydration, additional forces may be exerted on the water of the polymer particles. Extra water withdrawal results in more closely packed polymer particles, and coalescence can take place at even lower temperatures (6). In this way, a clear continuous polymer film was formed with SAE modification, even at temperatures of 20 C, which was lower than the MFT of 32 C. Measurements of temperature inside the specimens showed only a limited increase due to the release of heat in the hydration process. Figs. 3 to 6 show the etched surfaces of SAE-modified mortar samples, at various polymer/cement ratios. A polymer film is visible in all cases. The film formation is not prohibited by the curing temperature of 20 C. Owing to the presence of the cement and the cement hydration process, extra forces acted on the polymer particles and a film was formed. The influence of the polymer/cement ratio, i.e., the weight of solids of the polymer emulsion/weight of cement, is also clearly visible. At a p/c ratio of 5%, the polymer particles and the polymer film are preferably situated at the transition zone between the aggregate and the polymer-cement binder. Since the connections between the aggregates are very thin, the polymer particles acts more as an admixture for the cementitious material and has no major influence on mechanical properties or durability. In the case of a p/c ratio of 10%, a more continuous film is observed, Fig. 4. The polymer film is situated at the surface of the aggregates as well as in the bulk binder matrix. At higher p/c ratios, the polymer film becomes denser, and the bridges Relative weight (%) 100 90 80 70 60 50 40 30 20 10 SBR 50% -water 50% /R.H.60% SBR 50%-water 50%/R.H.86% SBR 50%-NaOH 50%/R.H.60% SBR 50%-NaOH 50%/R.H.86% SBR 50%-HSO 2 450%/R.H.60% SBR 50% -HSO 50% /R.H.86% 2 4 0 0 5 10 15 20 25 30 35 40 Time (hour) Fig. 2 - Weight variation of the SBR dispersion after dilution with a NaOH solution and a H 2 SO 4 solution, respectively. Fig. 4-10-% SAE modification, etched. Fig. 3-5-% SAE modification, etched. Fig. 5-15-% SAE modification, etched.

604 A. Beeldens et al. / Materials and Structures 38 (2005) 601-607 Fig. 6-20-% SAE modification, etched. between the aggregates are wider. The continuity of the polymer film throughout the binder matrix is striking. It is as if a second matrix is formed throughout the material, which is intermingled with the cement hydrates. The porosity of this second matrix becomes smaller with increasing p/c-ratio, which indicates that the polymer particles and the polymer film are intimately mixed with the cement hydrates and cement particles. 3. CEMENT HYDRATION IN POLYMER- MODIFIED MATERIALS Cement hydration in polymer-modified materials is influenced by the presence of the polymer particles and polymer film in the fresh state, during hydration and in the hardened state. The properties of the fresh mixture are strongly influenced by the surfactants present at the surface of the polymer particles. The cement particles are better dispersed in the mixture and a more homogeneous material is formed. The hydration of the cement is reflected in the strength development of the material. The influence of polymer modification is twofold. The cement hydration process is retarded by the polymer and the surfactants. This is visible especially in the compressive strength of mortar beams. After 28 days, the compressive strength of the polymer-modified mortar was lower than that of the reference mortar. After 90 days, however, the strength of the modified samples was equal to and even slightly higher than that of the unmodified samples. This is due to water retention by the surfactants of the polymer dispersion and to the partial or full encapsulation of the cement particles by the polymer dispersion. On the other hand, owing to the film formation or to the interaction between the cement hydrates and the polymer particles, the tensile strength of the binder matrix increases, as well as the adhesion strength between the aggregate and the binder. This is seen especially in the flexural strength of the mortar beams. Flexural strength increases by 25% in the case of a modified sample after 7 days, if a dry curing period (at 60% RH) is included. The results are presented in [1] and [7]. From the results of the mechanical tests for compressive and flexural strength, conclusions may be drawn for the film formation mechanism and especially for the time at which film formation takes place. Considering the compressive and flexural strengths after 7 days of dry curing, together with the reduction or increase, respectively, of the strength in relation to the strength of the unmodified mortar, an increase in flexural strength is observed, although the hydration of the cement is retarded when the polymer/cement ratio is increased (lower compressive strength). This points to the existence, at an early stage of curing, of a polymer film, or at least to an interaction between polymer particles and cement particles. However, the incapacity of the modified porous mortar specimens to overcome the large drying shrinkage stresses after 7 days or 28 days of moist curing indicates that the continuous polymer film is not yet formed in the case of water-saturated conditions. No influence of the polymer modification on flexural strength is noticed in the case of standard cured and watercured samples as long as no dry curing period is applied. It may, therefore, be concluded that at high relative humidity, the influence of polymer modification on short-term flexural strength is limited. As soon as a dry curing period is introduced, a polymer film starts to build up through the binder phase and an increase in flexural strength is measured with increasing p/c ratio. The influence of the retardation of the cement hydration process on flexural strength is compensated by the presence of the polymer film. When long-term behavior is considered, a maximum of flexural strength is achieved with a p/c ratio of about 15%. 4. INTEGRATED MODEL OF STRUCTURE FORMATION The mutual influences between the cement hydrates and the polymer particles and film have been incorporated in an integrated model of structure formation. The model is based on the three-step model as proposed by Ohama [3], but stresses the positioning of the mechanisms on the time scale and the interaction between the different components. The findings are supported by images taken with an environmental scanning electron microscope [8]. The formation of the polymer film plays an important role in the development of modified cementitious mortar and concrete. It can take as soon as two polymer droplets have sufficient energy to overcome the repulsion forces originating from the surfactants. In other words, if the temperature is high enough to cause sufficient Brownian motion, or if additional forces are acting on the liquid layer around the polymer droplets such as capillary forces or water withdrawal by further cement hydration, two droplets can come close to each other and coalesce into each other, and a polymer film is formed. This process can take place simultaneously with the cement hydration, especially under dry curing conditions. This allows partial or full encapsulation of the cement hydrates, which retards the hydration process. The different steps of the conclusive model are presented in Figs. 7 to 9. Immediately after mixing, the cement particles and polymer particles are dispersed in the water. The first hydration of the

A. Beeldens et al. / Materials and Structures 38 (2005) 601-607 605 Step 1: Immediately after mixing particles Polymer particles Water Step 2: Partial deposit of polymer particles, cement hydration, film formation particles and cement gel Dispersed polymer particles, coalesced polymer film Water, saturated with Ca(OH) 2 Fig. 7a - Step 1, immediately after mixing. Fig. 8a - Step 2, after mixing, the polymer particles interact with the cement particles and the aggregates. A continuous film may be formed if a dry curing period is introduced. Fig. 7b - Step 1, cement particles (small spheres) in a surrounding polymer particle, small ettringite needles are formed. cement takes place, which results in an alkaline pore solution. This is indicated as step 1, Fig. 7. The second step is presented in Figs. 8a and 8b. A portion of the polymer particles is deposited on the surface of the cement particle and the aggregate. The polymer/cement ratio determines the amount of polymers present in the pore solution and at the aggregate s surface. Part of the polymer particles may coalesce into a continuous film. This preferably takes place at the surface of the cement hydrates, where extra forces are exerted on the polymer particles owing to the extraction of water for cement hydration. The polymer film can partly or completely envelop a cement grain, which results in a retardation or even a complete stop of the hydration process. The following step, Fig. 9, consists of cement hydration, polymer flocculation and possibly polymer coalescence into a film. The processes which take place depend on the curing conditions. If no dry curing period, i.e., curing at a lower RH, is included, the overall film formation is retarded and the influence on the properties of the fresh mixture is limited at this stage. If a dry curing period is included, polymer film formation takes place during this step, which affects the cement hydration process as well as the strength development at early ages. In the bulk liquid phase, hydrate precipitations are present, which form a combined inorganic and organic product. The fractions of the different types of product formed depend on the polymer/cement ratio used. The polymer fractions included in these hydration products do not contribute to the strength Fig. 8b - Step 2, the polymer particles flocculate together, at restricted places; coalescence is visible by the elongation of polymer particles. As most particles are round, no coalescence has taken place at this stage. development of the specimen (5). It is found that only the polymers and the polymer film present at the aggregatematrix interface and in the pores of the material contribute to the strength development of the material. The final step, Fig. 10, includes further hydration and final film formation. Through the cement hydrates, a continuous polymer film forms as water is further removed from the pore solution. The part of the polymer particles that is still present in the dispersion is restricted to the capillary pores and at the interface of the aggregates and the bulk polymer-cement phase. It is this part which contributes the most to the elastic and final strength properties. The continuity of the polymer phase through the binder matrix is more marked in the case of a higher polymer/cement ratio. If the MFT of the polymer dispersion is higher than the curing temperature, the polymer particles may not coalesce into a continuous film, but remain as closely packed polymer particles. The use of the integrated Beeldens-Ohama-Van Gemert model, a refinement of Ohama s model, can be illustrated with the different curing conditions. From the results it is concluded that the best conditions for strength development are a wet curing period followed by a dry curing period. The longer the moist and water curing period, the higher the final flexural strength if shrinkage is prevented and if a curing

606 A. Beeldens et al. / Materials and Structures 38 (2005) 601-607 Step 3: Mixture of cement gel and unhydrated cement particles, enveloped with a close-packed layer of polymer particles and with polymer film. The cement hydrates grow partly through the polymer film particles and cement gel Polymer particles, closely packed coalesced into a film Combined inorganic and organic product precipitated in the bulk phase Water, pore solution Fig. 9a - Step 3, cement hydration proceeds, polymer film formation starts on specific spots. Step 4: Hardened structure, cement hydrates enveloped with polymer film particles Polymer film Combined inorganic and organic product precipitated in the bulk phase Entrained air Fig. 10a - Final step, cement hydration continuous, the polymer particles coalesce into a continuous film. Fig. 9b - Step 3, polymer particles have coalesced into a continuous film. The polymer particles shape is no longer spherical, but deformed. period at lower RH is introduced. This means that first cement hydration takes place, with only limited film formation. The polymer particles remain in the pore solution and a larger amount of polymer particles will be incorporated into the continuous film which is formed in the final stage. If the drying period is introduced earlier in the process, film formation will start sooner, i.e., before and simultaneously with the cement hydration, resulting in more encapsulation of the cement hydrates an incorporation of the polymer phase in the hydration product precipitated from the pore solution. 5. CONCLUSIONS The influence of polymer modification on the cement hydration process was investigated, as well as the influence of the presence of the cement particles on the polymer film formation process. The results were incorporated in a descriptive model for the structure formation of polymermodified materials, based on Ohama s model and including the modifications and developments proposed by various researchers. The polymer film formation process is influenced by the presence of the cement hydrates in different ways. First, there is the large dilution of the dispersion by the mixing water, which quickly changes into an alkaline solution. Results of tests on a diluted polymer emulsion showed little Fig. 10b - Step 4, cement particles are hydrated (angular shapes), polymer film (plastic film) is formed in the pores of the cement hydrates. or no influence of alkalinity on the film formation process for the investigated types of polymer dispersion. The relative humidity of the surrounding atmosphere had a strong influence on film formation and especially on the drying rate. The higher the relative humidity of the surrounding atmosphere, the lower the drying rate became. This greatly affects the film forming temperature of the dispersion. The lower the drying rate, the lower the amount of energy needed for the polymer particles to coalesce into a continuous film. Therefore, the MFT is reduced with a reduced drying rate. Tests indicated that even under laboratory conditions, i.e., 20 C, the SAE dispersion with a MFT of 32 C was able to form a continuous film as long as the drying rate was low enough. This is also revealed in the microstructure of the SAEmodified specimens. Through the cement matrix, a continuous polymer film is visible. At a p/c ratio lower than 5%, the continuity is only present through small tiny bridges on a limited number of spots. At higher p/c ratios, the film is denser. The film connects the different aggregates and is homogeneously present over the material. However, in the case of small p/c ratios, a preferred position close to the aggregate is to be noticed.

A. Beeldens et al. / Materials and Structures 38 (2005) 601-607 607 Cement hydration is retarded by polymer modification. As soon as a dry curing period is introduced, polymer film formation and cement hydration coincide. This results in partial or complete encapsulation of the cement particles, which reduces the hydration rate. Encapsulation also occurs as a result of the extra water withdrawal from the polymer particles situated at the surface of the cement grains. The cement hydration process is also influenced by the fact that water is retained longer, owing to the presence of the surfactants at the surface of the polymer particles. This results in a better dispersion of the polymer particles and the cement hydrates, but also retards the hydration of the cement. The influence increases with increasing p/c ratio. The previous findings have been incorporated in a model of structure formation. This model features two major changes with respect to the original model by Ohama. First, a relation to the time scale of the different processes is made. If a dry curing period is included, cement hydration and polymer film formation coincide and cement particles can be encapsulated. Secondly, the formation of an interstitial phase, consisting of inorganic and organic precipitates in the bulk phase, is pointed out. This is important with a view to deriving maximum benefit from polymer modification, since the polymers present in this phase are contributing less to the final properties of the material. The optimum conditions come forward from these findings, i.e., a long period of water or moist curing (up to 28 days), during which the cement hydrates develop, followed by a period of curing at lower relative humidity, during which the polymer film formation is promoted. REFERENCES [1] Beeldens, A., Influence of polymer modification on the behaviour of concrete under severe conditions, PhD dissertation, Faculty of Engineering, Katholieke Universiteit Leuven (2002) 248. [2] Bijen, J.M. and Su, Z., Polymer cement concrete: a contribution to modelling of the microstructure, Technical Committee TC-113, Symposium on Properties and Test Methods for Concrete-Polymer Composites, Oostende (Belgium) 19-27. [3] Ohama, Y., Handbook of Polymer-Modified Concrete and Mortars, Properties and Process Technology (Noyes Publications, 1995) 236. [4] Puterman, M. and Malorny, W., Some doubts and ideas on the microstructure formation of PCC, IX th International Congress on Polymers in Concrete ICPIC'98, Bologna, 1998, 165-178. [5] Su, Z., Larbi, J.A. and Bijen, J.M., The interface between polymer-modified cement paste and aggregate, Cement and Concrete Research 21(6) (1991) 983-990. [6] Tabor, L.J., Dispersed polymers, Revision of Concrete Society Technical Report No.9 - POLYMER CONCRETE, chapter 7, Contribution to the Fifth International Congress on Polymers in Concrete at Brighton, UK, 1987, 83-94. [7] Beeldens, A., Van Gemert, D., Ohama, Y. and Czarnecky, L., Integrated model of structure formation in polymer modified concrete, Proceedings of the 11 th International Congress on the Chemistry of Cement, 11-16 May 2003, Durban, CD-Rom. [8] Czarnecki, L. and Schorn, H., Nanomonitoring of polymercement concrete microstructure, 2 nd Conference Concrete Days, Szczyrk, Poland, 2002, 409-426.