Tag Archives: Reactivity

Influence of Particle Size on the Early Hydration of Slag Particle Activated by Ca(OH)2 Solution

Authors: Zhijun Tan, Geert De Schutter, Guang Ye, Yun Gao, Lieven Machiels


This paper investigates the influence of slag particle size on its hydration speed at particle level in the early age. Slags were separated with sieves into groups of different size fractions, considering a wide range of sizes. The chemical compositions of each group were analyzed by X-ray Fluorescence (XRF). Activated by 15% Ca(OH)2 (by mass) at water/powder ratio 1:1, the hydration heat evolution was recorded by isothermal calorimetry up to 84 hours and converted to hydration degree. Based on the hydration degrees and particle size distributions, the rate of increase of hydrating layer thickness of each single slag particle (k value) was calculated. Results reveal that k values of coarse particles are higher than that of fine particles. Coarse particles contain higher content of CaO but relatively lower content of MgO, Al2O3 and SiO2, resulting in higher reactivity index of (CaO+Al2O3+MgO)/SiO2.

Keywords: Blast-furnace slag, Particle size distribution, Reactivity, Hydration degree

Full paper link.

Does particle size of slag influence its reactivity?

It is mentioned previously that the reactivity of slag is influenced by its fineness of grinding (specific surface area) that is determined by the sizes of slag particles. The finer the slag particles, the larger the surface area, which means higher hydration degree of slag.

However, if disregarding the fineness, in other words, take two slag particles with significantly different sizes (diameters of “spherical” particles) as the study target, do the reacting front layers of the two particles grow at same rate? The growing rate of reacting layer can be a good tool to calculate the hydration degree of slag at particle level.

Simulated results conducted by Wei Chen showed the thickness of the hydrated layer is not significantly influenced by the particles size of slags, which is in good agreement with experiment result in which Na(OH)2 solution was used as activator [Ref. Three-dimensional computer modeling of slag cement hydration].


Another research carried out in 1986 reported similar result: “the thickness of hydrated layer of particle did not depend on the particle size, that is, the hydration reactivity had nothing to do with particle size and the rate of hydration was actually in proportion to the total surface area.” Detailed information is quoted as follows [Ref. Hydration of blast furnace slag particle, from 8th International Congress on the Chemistry of Cement Vol 4 (1986) Pages: 98-103],

It is most practical to evaluate the hydraulic reactivity of water granulated blast furnace slag in terras of the strength of hardened body. In order to make the extensive use of slag, however, it is important to investigate the hydration properties of individual particles. Hydration properties mainly investigated here were the change in the amount of hydration and the thickness of hydrated layer with time, and the effect of alkaline activator.

We examined the hydration of ground commercial slag with various particle sizes. Experiment was carried out as follows: samples with five different particle sizes (3.27, 4.05, 5.86, 8.66 and 13.36 μm by volume mean diameter) were prepared. They were hydrated in suspension with the water-slag ratio of 10 with NaOH, Ca(OH)2 or ordinary Portland cement as alkaline activator.

The amount of hydration of the slag under such an alkali-condition was measured by determining the soluble part of the hydration product using salicylic acid-acetone-methanol solution, and also by loss on ignition.

The following results were obtained. The thickness of hydrated layer of particle did not depend on the particle size, that is, the hydration reactivity had nothing to do with particle size and the rate of hydration was actually in proportion to the total surface area. It was suggested, on the other hand, that each alkaline activator show the characteristic effect on the hydration of each sample. Some of these results were supported by further experiments.

According to these research results, it is concluded that the reactivity of slag has nothing to do with particle size at particle level, while the surface area (fineness) determines hydration degree of slag.

It is should be noted that the two research results mentioned here only cover a narrow particle size distribution (0-15 micrometers). To apply this conclusion to wide size range, further research work is needed.

Factors affecting the reactivity of slag

The exact composition of slag varies over a range. In general, factors that determine the suitability of slag for usage in composite cement mainly include the fineness of grinding, glass content and the chemical composition.

  • Fineness

Like most of other cement materials, the reactivity of slag is influenced by its surface area. Increased surface area leads to better strength development and more water requirement; however, the fineness of slag is limited from practical aspects, such as economic and performance considerations, setting time and shrinkage. The following table shows typical fineness data of market slag in some countries.

Table: the surface area of slag in some countries (m2/Kg)

UK USA Canada India
Blaine surface area 375-425 450-550 450 350-450
  • Glass content

During the quenching process, the liquid slag forms glassy and crystalline contents. Practical glass content of slag depends on the cooling rate, in general, rapid rate results in high glass content. The main difference between glass content and crystal content of slag is that the former part has a latent hydraulic property that makes the glass content of slag a very important factor affecting the engineering performance of slag cement.

Though some researchers did obtain a roughly linear relationship between glass content and strength, there is no well-defined relationship between the glass content and strength of slag cement.

As for the relationship between hydraulicity and glass content, increasing glass content of slag improves its hydraulicity; however, research data that slag samples with 30-65% glass contents are still suitable has not shown exact correlation between them. Due to this uncertainty, most international standards classify slag reactivity by testing its direct strength rather than requiring minimum glass content. But from a practical standpoint, the glass content of slag should exceeds 90% to guarantee satisfactory properties.

  • Chemical composition

As stated above, the chemical composition of slag is mainly the four components, namely, MgO, Al2O3, SiO2, and CaO. From a metallurgical standpoint, slag can be sorted as either basic or acidic, and the more basic of slag, the greater its hydraulic activity in the presence of alkaline activators, Lea also reported that the hydraulic values of slag increase with the increasing CaO/SiO2 ratio up to a limiting value (not precisely stated). Further, in European Standard EN 197-1:1992 and British Standards, the ratio of the mass MgO plus CaO to SiO2 must exceed 1.0, by which the high alkalinity is guaranteed and otherwise the slag would be hydraulically inactive.

With a constant CaO/SiO2 ratio, the strength of hydrated slag increases with the Al2O3 content, and a large amount of Al2O3 can compensate the deficiency of CaO. Further research, by a regression analysis of compressive strength on composition using a wide range of west European slags, showed that increase in Al2O3 content above 13% tended to increase the early strengths but to decrease late strengths. Moreover, the content of Al2O3 also influences the sulfate resistance of slag concrete.

The influence of MgO as a replacement of CaO seems depending on both the basicity and the MgO content of slag. Variations in the MgO content up to 8-10% may have little effect on strength development, but high content have an adverse effect. It also reported that MgO in amount up to 11% was quantitatively equivalent to CaO. Frearon and Higgins reported that to get a satisfactory sulfate resistance the content of MgO should be about 13%.

Many researchers attempted to quantify the reactivity of slag considering the four major components together. Among these results, ratio (CaO+MgO+Al2O3)/SiO2 is the simplest and most widely used one. It was observed that the hydraulic activity of slags increases with the increasing contents of CaO, MgO and Al2O3 but decreases with the increasing content of SiO2. Furthermore, minimum values for this ratio, such as 1.0 (Germany) and 1.4 (Japan) have already been adopted in some countries’ standard specifications.

Apart from the four major components, there are also some minor components that may have important effect on the properties of slag, such as MnO is always negative, P2O5 and alkalis are more complicated.