© 2003 The Royal Microscopical Society Journal of Microscopy, Vol. 212, Pt 2 November 2003, pp. 186–196 Received 18 December 2002; accepted 7 July 2003 Blackwell Publishing Ltd. Quantitative microstructure analysis of polymer-modified mortars A. JENNI, M. HERWEGH, R. ZURBRIGGEN*, T. ABERLE* & L. HOLZER† Institute of Geological Sciences, University of Berne, Baltzersrtasse 1, CH-3012 Berne, Switzerland *Elotex AG, Sempach Station, Switzerland †EMPA, Dübendorf, Switzerland Key words. Ceramic tile adhesive, element mapping, fluorescence, polymer- modified cement mortar, quantitative microstructure analysis, staining. Summary Digital light, fluorescence and electron microscopy in combi- nation with wavelength-dispersive spectroscopy were used to visualize individual polymers, air voids, cement phases and filler minerals in a polymer-modified cementitious tile adhesive. In order to investigate the evolution and processes involved in formation of the mortar microstructure, quantifications of the phase distribution in the mortar were performed includ- ing phase-specific imaging and digital image analysis. The required sample preparation techniques and imaging related topics are discussed. As a form of case study, the different tech- niques were applied to obtain a quantitative characterization of a specific mortar mixture. The results indicate that the mortar fractionates during different stages ranging from the early fresh mortar until the final hardened mortar stage. This induces process-dependent enrichments of the phases at specific locations in the mortar. The approach presented provides important information for a comprehensive understanding of the functionality of polymer-modified mortars. Received 18 December 2002; accepted 7 July 2003 Introduction Polymer-modified mortars exist in a broad variety of applica- tions, e.g. tile adhesives. They are commercially available as premixed dry compounds, so-called dry mortars, which can basically be grouped into binders, fillers and additives (see Table 1). Typical additives are cellulose ether (CE) and redis- persible powder (RP). CE acts as thickener and air entraining agent, providing appropriate fresh mortar properties. RP is a polymeric powder typically gained by spray-drying of a polyvi- nyl alcohol (PVA)-containing latex emulsion. The RP is used to improve the fresh mortar properties and increase the flexi- bility and strength of the hardened mortar. Ordinary Portland cement (OPC) is the most typical mineral binder, which is added at 20 – 40 wt%. Sixty to 80 wt% of the mortar is com- posed of filler materials, typically siliceous and/or carbonate sand and finer grained fractions (< 100 µm). At the construction site, dry mortar is mixed with a prede- fined amount of water to gain a ready-to-use fresh mortar, giving a creamy and homogeneous mass with a considerable component of entrained air (15–30 vol.%). This so-called fresh mortar is applied by a trowel onto the substrate (wall or floor) and then covered by the tiles. Curing of the mortar over the following days and weeks is strongly dependent on the availability and migration of the added water, and involves three main processes (Fig. 1): (1) cement hydration, (2) drying and (3) polymer film formation. 1 Water reacts with anhydrous clinker phases and forms various cement hydrates. With the growth of these hydrates the mineral grains (clinker and fillers) are cemented together (Taylor, 1997). 2 Pore water that is not consumed by the hydration of the cement evaporates and leaves capillary pores (Taylor, 1997). 3 Polymer film formation occurs, dependent primarily on loss of water and curing time (Routh & Russel, 1999). The finally cured mortar consists of mineral grains (cement clinker and fillers) that are bound by the interstitial cement hydrates and polymer films. The bulk porosity of a cured mortar is between 30 and 50 vol.% (Fig. 1). Despite the low polymer level, the application performance is strongly dependent on the functionality of the polymers and their distribution in the mortar. The dynamic evolution of the mortar results in distinct sizes, shapes, spatial distributions and degree of con- nectivity of the different phases, i.e. a typical microstructure, which defines the mortar’s final physical properties. Therefore, the characterization of this microstructure is the essential key for a profound knowledge of both evolution and final properties. Correspondence to: A. Jenni. Tel.: +41 31 631 87 58; fax: +41 31 631 48 43; e-mail: andreas.jenni@geo.unibe.ch