Paper
Pore size distribution of hydrated cement pastes modified with polymers

https://doi.org/10.1016/S0008-8846(01)00549-XGet rights and content

Abstract

Mercury intrusion porosimetry (MIP) tests were performed on pure and polymer-modified cement pastes in order to evaluate the influence of hydroxyethyl cellulose (HEC polymer) and vinyl acetate/ethylene copolymer (EVA) on the pore size distribution. Cement pastes containing 0%, 10%, and 20% of EVA and 0%, 0.5%, and 1.0% of HEC of the weight of cement were prepared and the water/cement ratio was kept constant (0.4). The effects of two curing methods (dry cure: 27 days at 23°C and 75% relative humidity (RH); mixed cure: 7 days at 23°C under sealed conditions and 20 days at 23°C and 75% RH) were also evaluated. Analysis of variances (ANOVA) showed that the most important factors affecting pore size distribution are the curing method and EVA content.

Introduction

Polymers are employed as mortar and concrete modifiers due to the improvement of some properties such as fracture toughness, impermeability, durability, and bond strength to various substrates [1]. Water retention capacity and rheology are also improved. Basically, four polymeric compounds are used for these purposes: latexes or polymeric dispersions, redispersible powders, water-soluble polymers, and liquid polymers [2]. Some examples of polymer-modified cement-based materials currently used on construction sites are adhesive mortars for ceramic tile installation (dry-set mortars), structural repair mortars, overlays for bridge and parking decks, and impermeable mortars.

Research dealing with polymer addition in mortars for ceramic tile installation started in the 1960s in Europe and the USA. Methylcellulose was the most studied water-soluble polymer. The effects of other cellulosic water-soluble polymers on mortar properties have also been investigated, and they have been used for adhesive mortar production [3].

Nowadays in Brazil, most adhesive mortars are modified with water-soluble polymers and water-redispersible powders. Hydroxyethyl cellulose (HEC) and vinyl acetate/ethylene copolymer (EVA) are the most employed. HEC is a water-soluble polymer added in powder form to dry materials (cement and aggregates). After dissolution in water, it forms a gel and has major effect on fresh mortar properties, such as water retention capacity, rheology, and adhesiveness. EVA is a water-redispersible powder, and is also added to anhydrous cement and aggregates before mixing with water. EVA markedly influences hardened mortars behavior [4].

One of the most important characteristics of adhesive mortars is the pore structure since it affects the hardened mortars' behavior. Some mortars properties can be estimated from pore size distribution and total porosity. According to Ohama et al. [5], the total pore volume of the polymer-modified mortars, determined by mercury porosimetry, tends to decrease with an increase in polymer/cement ratio and decrease in water/cement ratio. However, the isolated effect of polymer/cement ratio was not studied, since in these experiments the water/cement ratio is changed to produce mortars with the same consistency.

The purpose of this study is to evaluate the effects of EVA and HEC on the pore structure of adhesive mortars by mercury intrusion porosimetry (MIP). Portland cement pastes were modified with different contents of both polymers. The water/cement ratio was kept constant in order to evaluate the actual effect of polymers. The threshold diameter, total pore volume, and pore size concentration over four distinct diameter ranges were evaluated and the paste behavior was then estimated.

Section snippets

Materials, experimental design, and methods

Materials employed for preparation of the pastes were ordinary Portland cement with up to 5% of limestone filler (type CPI-S 32 according to Brazilian standard NBR 5732/91), deionized water, water-redispersible EVA powder and water-soluble HEC powder. Table 1, Table 2 present the materials characteristics.

The following procedure was employed for paste mixing: (i) dry mixing of polymers and cement, in a low-speed mechanical mixer for 120 s, (ii) dry mixture flowing over deionized water, (iii)

MIP experimental procedure

At 28 days old, the specimens were cut with a diamond saw and four 1.0±0.2 mm thick slices were taken from different regions. Ethyl alcohol was used to cool the saw. The slices were then immersed on ethyl alcohol and submitted to 20 min in ultrasonic cleaning equipment, manually fragmented, and stored in alcohol for at least 6 days for hydration interruption and solvent replacement drying.

Before the test, the pieces were removed from the alcohol and vacuum-oven-dried at 40°C for 25 h. This

Results and discussion

Fig. 1 shows the characteristic curves of intruded mercury volume in function of pore diameter (incremental volume) for mixed-cured pastes. It is possible to observe that pore size distribution curves of pure (reference) pastes and HEC- and EVA-modified pastes typically exhibit at least two peaks. The first one lies approximately at the 3.9-nm diameter, while the second and sharpest one corresponds to a diameter around 40–75 nm. For dry-cured pastes (results not shown), the profile is quite the

Conclusions

From the MIP tests performed with EVA and/or HEC-modified cement pastes it was possible to conclude that:

1. Pore size distribution curves of pure and polymer-modified pastes show at least two peaks corresponding to micropores and capillaries that lie around 3.9 and 40–75 nm, respectively. High contents of HEC change this profile, and the curve shows two more peaks. This is probably due to the lower degree of cement hydration, to the transformation of water into a gel and, for larger pores, to

Acknowledgements

The authors acknowledge the financial support of CNPq and FINEP and the tests realization by Building Materials Microstructure Laboratory, Department of Civil Construction Engineering, Escola Politécnica da Universidade de São Paulo (BMML-USP).

References (18)

There are more references available in the full text version of this article.

Cited by (0)

View full text