Improved hydration resistance of magnesia by EDTA and ammonium phosphate as additives

2.1 Materials

Suspensions of light burned magnesia (D₅₀=6.67 μm and 99.0 wt.% MgO, China) with 10 wt.% of solid content were used for the experiments in water or ethanol. Compounds with 0.5 wt.% relative to MgO were evaluated as additive candidates, such as EDTA and AP. Ethylene diamine tetraacetic acid (EDTA), ammonium phosphate (AP), were supplied by Tianjin Chemical Reagent Co., Ltd, Tianjin, China. MgO and ethanol were purchased from Sinopharm Chemical Reagent Co., Ltd., Shanghai, China. Distilled water was used as solvent for all experiments.

2.2 Characterization

A laser particle size analyzer (Microtrac S3500; Microtrac Inc., USA) was used to measure particle size distribution before and after MgO particle hydration, and the morphologies of unhydrated MgO particle were characterized by using a SEM (JSM-6360 LV; JEOL Ltd., Japan). In RSV test, hydrated and unhydrated MgO particles were sedimented under the action of gravity. Our experiments were carried out in a graduated cylinder (10 ml) to determine the relationship between settling velocity and particle size at settling time; in this method, the best antihydration additive, with which the RSV was the least, was determined. The ζ potential and pH value were measured by using a ζ potential tester (ZetaPlus2002; Brookhaven Instruments Corporation), and the hydration ratio (HR) was calculated as follows:

where m₁ is the weight of MgO after hydration and m₂ is the weight of MgO before hydration.

3.1 Size changes in plain MgO suspensions

Figure 1 shows that the MgO particle has a size of only several micrometers. When the MgO particle surface is exposed to water, its cubic structure is changed to brucite’s hexagonal one or to the rhombohedra when magnesite is formed. The volumetric expansion produces a striped micro-cracked pattern, as a result of the hexagonal structure morphology; the density of the particles was reduced while, at the same time, its volume increased. The particle size change of MgO in ethanol and in water is schematically shown in Figure 2. The hydration of MgO is carried out by the contact of this oxide with liquid water (as in the mixing and curing steps of castables), water vapor (during the drying process), or moisture (during the storage period). The hydration products formed are described in the literature as follows:

Figure 1:

Morphology of MgO powder.

Figure 2:

Particle distribution of MgO powders in water and in ethanol.

Magnesium hydroxide or brucite is crystalline, and it is formed when there is a great availability of water [Eq. (2)], as in castable structures. Brucite formation promotes a significant increase in the pH, allowing CO₂ to be dissolved in water and the generation of carbonic acid (H₂CO₃). This acid reacts with magnesium oxide (MgO), resulting in magnesium carbonate or magnesite, as shown in Eq. (3). Magnesite can also be formed during the storage of magnesia sinters and preshaped parts, and presents a mold aspect. Both hydration routes present a similar effect: a remarkable 2.5-fold volumetric expansion. Nevertheless, as magnesite is preferably formed on the surface of the exposed MgO, brucite is normally associated with the volumetric variation and its effects on the refractory structures and properties.

3.2 Hydration in plain MgO and additive suspensions

Figure 3 shows the ionic conductivity and pH for the plain MgO suspensions. The low value of ionic conductivity at the beginning of the experiment and then the gradual increase suggest that magnesia surface protonation [Eq. (4)] occurs as soon as it comes in contact with water, At the same time, OH^- ions are released to the aqueous medium, which explains the relatively high initial values of pH.

Figure 3:

Changes of pH and electrical conductivity of untreated MgO with time.

In the presence of EDTA (AP), there are two dissociation reactions that occur simultaneously in the suspensions: the reaction between water and EDTA (AP). In the case of EDTA (AP), these reactions are more complex because they depend on their structure and on how many dissociable groups there are in their molecule.

EDTA presents various negative sites in the same molecule and because of this, it could be adsorbed onto the positively charged magnesia surface, while AP can react with magnesium ions. In the first contact with the water-EDTA (AP) system, the MgO particles’ surfaces become protonated [Eq. (5)] by Mg²⁺ ions from both water and the EDTA (AP) dissociation. Thereafter, as the negative ions approach the positively charged MgO surface, two different negative species are available: (i) the OH^- ions due to the water dissociation and (ii) the negative ions released from the EDTA (AP). As seen previously, the OH^- ions adsorbed on the MgOH⁺ surface are desorbed after some period of time, releasing magnesium cations into the solution [Eq. (3)]. However, because these Mg²⁺ ions can be trapped by the chelant, the EDTA (AP) negative ions could react with them [1]:

Thus, depending on the stability constant of the complex formed and the energy reduction caused by the ionic adsorption on the MgO surface, the complexation trend can be inferred by the stability constant (Ks) of the complex compound, as presented in Table 1 for the chelants studied here. The higher the Ks value, the more stable is the complex formed. Figure 4 shows different additives (EDTA or AP) for MgO suspensions, the ζ potential values of them increase with time. However, the higher initial zeta potential values of EDTA additive indicate lower degree of surface adsorption. So, AP additives present the higher adsorption on the magnesia surface, resulting in a lower ζ potential values. Therefore, the complexation reaction is favored and magnesia dissolution takes place. Conversely, citric acid and tartaric acid additives present the highest adsorption on the magnesia surface, as can be predicted from their low Ks values [22].

Figure 4:

Change of electrical conductivity of MgO slurry treated by AP or EDTA with time.

3.3 The effect of additives on RSV

According to the theoretical calculation of density of MgO and magnesium hydrate, the expansion that follows the hydration of the MgO can reach up to 2.5 times its initial volume. Nevertheless, the results of the sedimentation volume of MgO slurry at different conditions showed that the expansion observed is frequently smaller (less than two times), which is because a part of MgO does not hydrate; both EDTA and AP can partly inhibit MgO hydration, and the effect of AP is better than that of EDTA. Thus, the RSV observed in formulations can be evaluated by measuring the hydration difference of the samples before and after being exposed to a certain hydration condition (Figure 5). The RSV test consists in measuring the sedimentation volume of a cuvette sample under a certain hydration condition (over 24 h in a humid saturated environment at 25°C, and the content of MgO is 10 wt.%).

Figure 5:

RSV of MgO slurry at different conditions.

3.4 The effect of additives on hydration ratio

Table 2 shows the hydration degrees of different MgO with time tests results carried out in samples cured at 25°C for different periods. It can be pointed out that the hydration degree of the different MgO in each hydration stage changed with the increase in the hydration time. At 30 min, the hydration ratio of untreated MgO has reached 10.45%, while it was only 7.83% and 6.72% in MgO treated with EDTA and AP, respectively. This is an indication that the additives EDTA and AP can evidently decrease the hydration ratio in the formulation in their early stages of hydration, all processes are similar. The results show that the hydration-resistant effect of AP in MgO is, once again, better than that of EDTA.