Clinker phases

A knowledge of the relevant high-temperature phase equilibria is necessary for understanding the factors that govern acceptable bulk compositions for Portland cement clinker, the conditions under which the latter can be manufactured, and the phase composition and microstructure of the resulting material. This chapter deals with these equilibria and with the phases to which they relate, with the exception of the major clinker phases, which were described in Chapter I. Some anhydrous phases primarily of interest in relation to other types of cement are also considered here. Principles underlying the preparation of anhydrous silicate, aluminate and other high-temperature phases are outlined. 

Portland cement clinkers contain small amounts of alkalis and sulphates derived from the raw materials and fuel. Both alkalis and SO3 can be present in the major clinker phases, but tend to combine preferentially with each other to form alkali or potassium calcium sulphates, and it is necessary to consider these components together. In addition, silicate and aluminate phases containing sulphate can form either as intermediates or in undesirable deposits in eement making, and a calcium aluminate sulphate is a major constituent of some expansive and other speeial cements. 

Table 2.3 lists some phases containing MgO that are in varying degrees relevant to cement chemistry. It is not a complete list of phases with essential MgO in the CaO-MgO-AljOj-SiOj system. As seen in Chapter 1, some MgO is also taken up by all four of the major clinker phases, typical contents being 0.5-2.0% for alite, 0.5% for belite, 1.4% for the aluminate phase, and 3.0% for the ferrite phase. Magnesium oxide (periclase), like calcium oxide, has the sodium chloride structure it is cubic, with a = 0.4213 nm, space group Fm3m, Z = 4, = 3581 kgm (S5) and refrac-... 

Swayze assumed that, for the bulk compositions that he studied, all the MgO was in the liquid phase at equilibrium, apart from the small proportions present as periclase. As noted in the preceding section, all four of the major clinker phases take up significant proportions of MgO. The contents of the latter component in the liquid phase were therefore probably substantially lower than 5%. 

Fig. 4.3 Backscattered electron images of polished sections of (A) a Portland cement clinker and (B) grains of a Portland cement in a fresh paste. In both sections, alite is the predominant clinker phase. In (A), the relatively large, darker areas are of belite, and the interstitial material consists of dendritic ferrite (light) in a matrix of aluminate (dark) cracks and pores (black) are also visible. In (B), the belite forms well-defined regions, which are rounded, striated and darker than the alite the interstitial material, present, for example, in a vertical band left of centre within the larger grain, consists mainly of ferrite (light) and aluminate (dark). Scrivener and Pratt (S28).

The potential uses of XRD powder diffraction in the study of clinker or anhydrous cement include the qualitative and quantitative (QXDA) determination of phase composition, and the determination of polymorphic modification, state of crystallinity and other features of individual phases. In principle, information on compositions of phases is obtainable through cell parameters, but, due to the lack of adequate reference data, XRD is generally less satisfactory for the clinker phases than X-ray microanalysis. Table 4.2 gives the pattern of a typical Portland cement, with indications of the assignments of peaks to phases. 

The Appendix gives calculated patterns based on the results of X-ray structure determinations for the clinker phases. For each pattern, a Reference Intensity Ratio (H23) is included. This is the integrated intensity of the strongest individual reflection (which may be a component of an overlap) relative to that of the strongest peak of corundum in a 1 1 mixture by... 

Unless otherwise stated, this chapter relates to ordinary Portland cements hydrated in pastes at 15-25°C and w/c ratios of 0.45-0.65. XRD powder studies on such pastes have been reported by many investigators (e.g. C38,M67). The rates of disappearance of the phases present in the unreacted cement are considered more fully in Section 7.2.1. Gypsum and other calcium sulphate phases are no longer detectable after, at most, 24 h, and tbe clinker phases are consumed at differing rates, alite and aluminate phase reacting more quickly than belite and ferrite. The ratio of belite to alite thus increases steadily, and after about 90 days at most, little or no alite or aluminate phase is normally detectable. 

The brightest areas are of unreacted clinker phases. Individual cement grains are usually polymineralic and, as in a clinker (Section 4.3.1), the different phases within them can be distinguished by their differing grey levels or by X-ray microanalysis. 

CH can be observed as areas darker than the unreacted clinker phases but brighter than the other hydration products. As in calcium silicate pastes, these appear to have grown in regions initially occupied by water. Although the areas appear discrete on two-dimensional sections, they are not necessarily so in the three-dimensional material. They can engulf small cement grains. 

The experimental considerations applying to calcium silicate pastes (Sections 5.1 and 5.2) are equally relevant to cement pastes. Of the methods so far used in attempts to determine the degrees of reaction of the individual clinker phases as a function of time, QXDA (C39,D12,T34,P28) has proved much the most satisfactory. Procedures are essentially as for the analysis of a clinker or unreacted cement (Section 4.3.2), but it is necessary to take account of overlaps with peaks from the hydration products, and especially, with the C-S-H band at 0.27-0.31 nm. The water content of the sample must be known, so that the results can be referred to the weight of anhydrous material. If a sample of the unhydrated cement is available, and its quantitative phase composition has been determined, it may be used as the reference standard for the individual clinker phases in the paste. 

Fig. 7.3 QXDA results for the fractions of the clinker phases reacted in Portland cement pastes. Filled circles Copeland and Kantro (C39), w/c = 0.65. Diamonds Bezjak et al. (B98), sample C2. Vertical lines Osbaeck and Jons (08), range for 7 samples. Open circles Dalziel and Gutteridge (DI2). Open squares Patel et al. (P28), samples cured at 100% RH. Filled squares Tang and Gartner (T34), clinker interground with gypsum.

Water retained after D-drying, known as non-evaporable water, has often been wrongly identified with chemically bound water. It excludes much of the interlayer water in C-S-H, AFm and hydrotalcite-type phases and much of the water contained in the crystal structure of AFt phases. It is often used as a measure of the fraction of the cement that has reacted, but can only be approximate in this respect, because the clinker phases react at different rates and yield products containing different amounts of non-evaporable water. Fully hydrated cement pastes typically contain about 23% of non-evaporable water, referred to the ignited weight. Copeland et al. (C38) determined the non-evaporable water contents of a series of mature cement pastes and carried out regression analyses on the cement composition. For pastes of w/c ratio 0.8 and aged 6.5 years, they obtained the approximate expression ... 

The and Na" are present in the cement partly as sulphates and partly in the major clinker phases (Section 3.5.6). When the phases containing them react, the accompanying anions enter products of low solubility and equivalent quantities of OH are produced. The K, Na and OH" ions are partitioned between the pore solution and the hydration products. Early estimates of the fractions remaining in solution were too high because non-evaporable water was used as a measure of bound water and the quantity of pore solution thereby overestimated. 

Factors affecting the rate at which AljOj is supplied include its content in the clinker, its distribution among the clinker phases and especially the amount present in the aluminate, the specific surface area of the ground clinker, the reactivities of the aluminate and ferrite, and the microstructure of the clinker particles, or, more specifically, the areas of surface composed of aluminate and ferrite phases and the manner in which these phases are... 

Sulphate ion is also supplied by the clinker, especially as alkali sulphates or calcium langbeinite, and Ca " ions are supplied by the clinker phases, including free lime. The alkali sulphates provide a highly available source of S04, but the alkali cations, or more probably the OH" ions that they produce, have additional effects (Section 7.6.3). 

Major influences on the kinetic curve of a cement include the phase composition of the clinker, the particle size distribution of the cement and the RH and temperature regimes during curing. Other influences include the w/c ratio, the content and distribution of admixtures, including gypsum, the reactivities of individual clinker phases and probably others, such as the microstructures of the clinker and of the cement particles. 

By definition, the kinetic curve of a cement is the weighted sum of the curves for its constituent phases as they occur in that cement. The reactivities of individual clinker phases were considered in Section 4.5 and some effects of particle size distribution, which is a particularly important variable, in Section 4.1.4. Although many data relating particle size distribution directly to strength exist, much less is known about its relation to degrees of reaction. Parrott and Killoh (P30) presented data indicating that the rate of hydration, as represented by that of heat evolution, was proportional to the specific surface area during the period of hydration in which the rate was controlled by diffusion. 

The rates of reaction of the clinker phases are greatly influenced by the RH of the atmosphere in which curing occurs. For a typical Portland cement paste of w/c ratio 0.59 cured at 20°C and 100% RH, Patel el al. (P28) found the fractions of the alite, belite, aluminate and ferrite phases hydrated after 90 days to be respectively 0.94, 0.85, 1.00 and 0.51. If the RH was lowered to 80%, the corresponding values were 0.77, 0.19, 0.83 and 0.32. The hydration rate of the belite thus appears to be especially sensitive to RH. On the basis of earlier data from the literature, Parrott and Killoh (P30) concluded that the effect of RH on the hydration rate (da/d/) of each of the phases could be represented by a factor (RH — 0.55)/0.45. ... 

Clinker phases C-S-H Ca(OH)2 AFm AFt Slag HP FeHP Mg HP CaCOj Slag glass Slag res. Other Pores... 

Rates of consumption of clinker phases and pfa, and contents of calcium hydroxide... 

Uchikawa (UI7) reviewed the hydration chemistry of pfa and other composite cements. Pfa cements differ from pure Portland cements notably in (i) the hydration rates of the clinker phases, (ii) CH contents, which are lowered both by the dilution of the clinker by pfa and by the pozzolanic reaction, (iii) the compositions of the clinker hydration products and (iv) formation of hydration products from the pfa. The two last aspects cannot be wholly separated. 

Quantitative data on the rate of consumption of pfa are few and somewhat variable. Those based on differences between the CH contents of pure Portland and pfa cements are suspect, because the calculation involves the effects of pfa substitution both on the rate of consumption of the clinker phases and on the compositions of the products, which are not fully understood. Unreacted pfa has been directly determined by dissolution of the other phases with HCl (C43) or with salicylic acid in methanol followed by HCl (T44), chemical separation of the residual pfa followed by QXDA determination of its content of crystalline phases (D12) and a trimethylsily-lation method (U19). A method based on EDTA extraction was found unsatisfactory (L46). 

---