5

Corrosion of Specific

Crystalline Materials

The most beautiful thing we can experience is the mysterious. It is the source of all true art and science.

ALBERT EINSTEIN

5.1 ATTACK BY LIQUIDS

5.1.1 Attack by Glasses

In the indirect corrosion of oxides by glasses, the crystalline phase that forms at the interface is dependent upon the glass composition and the temperature. Various interface phases that form in some silicate melts are listed in Table 5.1. Whether the system is under forced convection or not will also play an

165

Copyright © 2004 by Marcel Dekker, Inc.

TABLE 5.1 Interfacial Reaction Products Caused by Molten Liquid Attack

important role in the formation of a crystalline interface phase. An excellent study of the effects of forced convection is that by Sandhage and Yurek [5.1], who, in their studies of the indirect dissolution of chrome-alumina crystalline solution materials in CaO–MgO–Al2O3–SiO2 melts at 1550°C, reported that the reaction layer thickness of the spinel that formed decreased with increasing rotational rpm, but did not change with time at constant rpm. The reaction layer was an order-of-magnitude thinner (30 vs. 300 µm) at 1200 rpm when compared to the case with no forced convection.

The investigator must be careful in his interpretation of the crystalline phases present after an experiment has been

Copyright © 2004 by Marcel Dekker, Inc.

completed, so that he does not confuse phases that precipitate during cooling with those that were present during the experiment. The reader, if interested in a particular system, should examine the original articles of those listed in Table 5.1 to determine the exact experimental conditions. The following sections describe some of the more important systems that have been investigated but no attempt has been made for an exhaustive survey.

Alumina-Containing Materials

The corrosion of multicomponent materials proceeds through the path of least resistance. Thus, those components with the lowest resistance are corroded first. This is really a form of selective corrosion and may proceed through either the direct or indirect corrosion process. The corrosion of a fusion cast alumina-zirconia-silica (AZS) refractory will be used as an example of a case when selective direct corrosion is operative. This particular material is manufactured by fusing the oxides, casting into a mold, and then allowing crystallization to occur under controlled conditions. The final microstructure is composed of primary zirconia, alumina, alumina with included zirconia, and a glassy phase that surrounds all the other phases (Fig. 5.1). The glassy phase (about 15% by volume) is necessary for this material to provide a cushion for the polymorphic transformation of zirconia* during cooling and subsequent use. This material is widely used as a basin-wall material in soda- lime-silica glass furnaces. The corrosion proceeds by the diffusion of sodium ions from the bulk glass into the glassy phase of the refractory. As sodium ions are added to this glass, its viscosity is lowered and it becomes corrosive toward the refractory. The corrosion next proceeds by solution of the

* Tetragonal zirconia transforms to monoclinic zirconia athermally upon cooling at about 1000°C accompanied by an expansion of about 5 vol.%.

Copyright © 2004 by Marcel Dekker, Inc.

FIGURE 5.1 Microstructure of an alumina-zirconia-silica refractory. Reflected oblique illumination (magnification 200×). Brightest areas are ZrO2, next darker areas are A12O3, next darker areas are silicate glass (diagonal area across middle) and the few darkest spots are pores.

alumina and finally by partial solution of the zirconia. Under stagnant conditions, an interface of zirconia embedded in a high-viscosity, alumina-rich glass is formed (Fig. 5.2). If the diffusion of sodium ions into the glassy phase is sufficient, the glassy phase may contain sufficient sodium so that upon cooling, nepheline (Na2Al2Si2O8) crystals precipitate, or if the temperature is proper, the nepheline may form in service. The presence of nepheline has been reported by several investigators [5.6,5.12,5.13]. In actual service conditions, however, the convective flow of the bulk glass erodes this interface, allowing

Copyright © 2004 by Marcel Dekker, Inc.

FIGURE 5.2 ZrO2 interface on AZS refractory corroded by a soda- lime-silica glass at 1450°C for 7 days. Reflected light illumination. (Courtesy of Corning, Inc.)

continuous corrosion to take place until the refractory is consumed. This type of corrosion can take place in any multicomponent material where the corroding liquid diffuses into a material that contains several phases of varying corrosion resistance.

Hilger et al. [5.7] reported the corrosion of an AZS refractory by a potassium-lead-silicate glass at 1200°C to be very similar to that discussed above. In this case, the potassium ions diffused into the glassy phase of the refractory, dissolved the alumina of the refractory, and formed a glassy phase with a composition very similar to leucite (K2Al2Si4O12). Actual crystals of leucite were found upon examination of used blocks. It was interesting that very little lead diffused into the refractory.

In these refractory materials containing ZrO2, one should note that the ZrO2 is very insoluble in soda-lime-silicate and

Copyright © 2004 by Marcel Dekker, Inc.

potassium-lead-silicate glasses. Thus the corrosion of AZS refractories in these glasses is very similar to that which occurs in alumino-silicate (e.g., mullite) refractories. The major difference being the skeletal interface layer of undissolved ZrO2 that forms on the AZS materials. The presence of lead in the corroding glass acts predominantly to lower the viscosity, with increasing lead contents producing more severe corrosion [5.11].

Lakatos and Simmingskold [5.14] studied the effects of various glass constituents upon the corrosion of two pot clays, one with 21% alumina and one with 37% alumina. Their silicate glasses contained K2O, Na2O, CaO, and PbO in varying amounts. They found that PbO had no significant effect upon corrosion, that Na2O was 2–3 times more corrosive than K2O, and that CaO followed a cubic function. As their tests were conducted at 1400°C, it should be obvious that the glass viscosities varied considerably. They concluded that 95–96% of the total variance in corrosion was a result of viscosity differences, and that the specific chemical effects existed only to a small extent.

Lakatos and Simmingskold [5.15] later found in isoviscosity tests that the corrosion of alumina depended upon the lime and magnesia content of the glass, whereas the corrosion of silica depended upon the alkali content.

During the testing of refractories for resistance toward coalash slags, Bonar et al. [5.3] determined that AZS type refractories exhibited complete dissolution at the slag line, alumina exhibited significant corrosion, and a chrome-spinel refractory exhibited negligible attack at 1500°C and 10-3 Pa oxygen pressure for 532 hr. These results were consistent with the determined acid/ base ratios of the slags and what one would predict knowing the acid or base character of the refractories.

Fig. 5.3 shows the results of a mullite refractory that was removed from the regenerator division wall of a soda-lime- silicate container glass furnace. The sample was in service for 1 year at a temperature of approximately 1480°C. The attacking glass was from batch particulate carryover and condensation of volatiles. A small amount of convective flow down the vertical

Copyright © 2004 by Marcel Dekker, Inc.

FIGURE 5.3 Corrosion of mullite refractory: (a) XRD pattern, (b) reflected light optical micrograph (magnification 100×, lighter areas are corundum and darker areas are nepheline), and (c) EDS spot maps for several elements.

Copyright © 2004 by Marcel Dekker, Inc.

FIGURE 5.3 Continued.

face of the wall was present due only to gravity. The alteration of the refractory due to corrosion occurred to a depth of about 25 mm. As can be seen from Fig. 5.3, the mullite has completely converted to predominantly corundum and nepheline. Fig. 5.3a is the X-ray diffraction (XRD) pattern supporting the presence of only corundum and nepheline. The optical micrograph shown in Fig. 5.3b indicated the presence of an additional phase. Upon examination of elemental maps via scanning electron microscopy (SEM)/energy dispersive spectroscopy (EDS) shown in Fig. 5.3c, it was determined that the nepheline contained a reasonable amount of dissolved calcium and that the crystalline nepheline was embedded in a matrix of vitreous potassium- titanium-silicate. The potassium diffused into the refractory from glass batch impurities and the titanium was present in the original refractory in minor amounts.

Copyright © 2004 by Marcel Dekker, Inc.