As petroleum resources diminish, the future will bring strong incentives for the use of low quality fuel in engines employing advanced materials. In its present usage, the term advanced materials refers principally to the ceramics silicon carbide (SiC), silicon nitride (Si3N4), and stabilized zirconia (ZrOz); to high-chromium content, silicon-augmented, or specially processed metals; and to ceramic-fiber reinforced metals or ceramics. Other ceramics such as mullite, Ca2Si04, CaTi03, ZrSi04, and ZrTi04 have been suggested for engine usage, but these are as yet little developed.

Of the contaminants occurring in low quality fuel, experience to date has indicated that sodium, sulfur, vanadium, phosphorus, and lead are the most likely to cause degradation of the advanced materials projected for future engines.

This review is organized therefore to first look briefly at how these contaminants, especially vanadium, react in the molten salt corrosion of metals, and then to examine how the same contaminants react in the molten salt degradation of ceramics.

ACID-BASE REACTIONS OF OXIDES During combustion^ fuel contaminants are converted to oxides such as NazO, S02-S03, V205, etc. which have strong acid-base characteristics. Fuel contaminant corrosion of metals and ceramics involves reaction between these corrosive oxides and protective oxide scales on the metal or ceramic surface (e.g., SiC and Si3N4 rely on a Si02 surface layer for high temperature stability in air). When molten salt deposits are present, the acid-base reactions are often best treated by the Lux-Flood (1) acid-base theory where the various compounds are related by equilibrium reactions such as

and the activities of the acidic and basic components in the melt are fixed by the dissociation constant (e.g., 10"16-7 for Na2S04 at 1200°K, cf. ref. 6) of the reaction.

In solid state oxide reactions, the Lewis theory, which defines acid-base behavior in terms of the ability of the species to donate (base) or accept (acid) electrons, is the more useful. This allows, for instance, ready understanding of reactions of the type

An excellent insight into the influence of Lewis acid-base character in determining oxide reactions is provided by the works of Duffy (2,3) on silicate glass and other oxide systems.

VANADIC HOT CORROSION OF METALS Since metal engines (with metallic coatings) will be in use for many years to come, the first "new" low quality fuel problem, as the vanadium level in the world's oil supplies rises, is likely to be molten vanadate corrosion of metals. Before discussing vanadic hot corrosion, however, it is beneficial to review some aspects of molten sulfate hot corrosion.

Molten sulfate corrosion has been studied extensively, and acid-base oxide reactions, although certainly not the only critical reaction in sulfate hot corrosion, have been shown to be important. Initiation of corrosive attack at 900°C' (i.e., in the so-called "high temperature" hot corrosion regime) appears often to involve "basic" fluxing of the protective surface oxide, e.g.,

2NaA102("soluble salt") [3]

where the NazO activity in the molten sulfate deposit has been raised by metal sulfide formation (4). Similarly, in 700°C "low temperature" hot corrosion, acidic fluxing of cobalt oxide by S03 in the turbine gas® is the apparent cause of attack

Acid-base oxide reactions with mplten sulfate have been examined by Rapp (6) through the measurement of oxide solubilities as a function of Na2G activity in fused Na2S04 (Fig. 1). Fig. 1 shows clearly the influence of

Fig. 1. Oxide solubilities in fused Na2SC>4 at 1200°K and 1 atm 02. (Taken from ref. 6; used with permission.)

Na20 activity on oxide solubility. It also serves1 to rank' thé acid-base character of ; the individual oxide's:'' Cobalt oxide, e.gv reacts as an acid with Na2Ô (forming sodium cobaltate) down to Na20 activity of about 10'9-3, but as à base (forming; cobalt sulfate) below that activity. Conversely, A1203 and Cr203 are sufficiently acidic that they react as acids (forming sodium chrômate and sodium aluminate) down to NazO activity'of TO"15-5 (corresponding to an S03 activity of 10'1 •2), and as bases only below this activity. The data for Si02 in Fig. I illustrate a third impdrtaiit factor —that all oxides may not show ttiè same acid-base reaction behavior. Although other work!; indicates • that sodium silicates would be produced at higher NazO activities, only physical solubility ,of Si02 (with no evidence of reaction with either Na20 or S03) was found over the. activity range in Fig. 1. No explanation has been given, but, as noted by Duffy (3), "network forming" oxides such as SiOz (also B203 and P205) may possess lattice bonding energies that can override weak acid-base driving forces.

Compared to molten sulfate corrosion which involves only Na2S04, vanadic hot corrosion appears potentially more complex because five compounds (Fig. 2), ranging from high V2Os-activity Na2V12031 to high

0 0

Post a comment