Ageing of chromium(III)-bearing slag and its relation to the atmospheric oxidation of solid chromium(III)-oxide in the presence of calcium oxide

Abstract

Slag arising in ferrochromium and stainless steel production is known to contain residual levels of trivalent chromium. As the chromium is normally bound in the slag matrix in various silicate or spinel phases, and hence not easily mobilised, utilisation or controlled disposal of such slag is generally considered unproblematic. Experimental test work with a number of slag materials indicates, however, that very gradual oxidation of trivalent to hexavalent chromium does occur when the slag is exposed to atmospheric oxygen, rendering a quantifiable but small portion of chromium in this much more mobile and toxic form. Mechanisms and rates of the oxidation reaction were investigated in a number of long-term studies using both original slag materials and artificial mixes of chromium and calcium oxides. Powders of these materials, some of them rolled into balls, were left to age under different conditions for periods of up to 12 months. In the slag samples, which contained between 1 and 3 wt.% chromium, 1000–10 000 μg Cr(VI) were found per gram of chromium within 6–9 months of exposure to an ambient atmosphere. The rate of the oxidation reaction decreased exponentially, and the reaction could generally be said to have ceased within 12 months. In mixtures of calcium and chromium oxides the oxidation reaction is presumed to occur at the boundaries between chromium oxide and calcium oxide phases through diffusion of oxygen along the grain boundaries and of Cr³⁺ across the boundaries, resulting in the formation of calcium chromate. In the slags, where calcium and chromium oxide can form a solid solution, the oxidation is likely to occur at the exposed surface of grains containing this solution.

Introduction

Chromium occurs in the natural environment primarily in two oxidation states, trivalent (Cr(III)) and hexavalent (Cr(VI)). Chemical compounds containing hexavalent chromium are generally considered far more toxic than those containing the trivalent atom. Toxic effects associated with Cr(VI) range from mucosal ulceration to lung cancer (Yassi and Nieboer, 1988; Sheehan et al., 1991). Numerous documented case studies of chromium poisoning have focussed on the impacts of hexavalent chromium released into groundwater, and on the consequences of occupational exposure to chromium in this oxidation state (Mancuso and Hueper, 1951; Franchini et al., 1983; Calder, 1988; Burke et al., 1991; Sheehan et al., 1991). Hence, regulatory bodies tend to prescribe strict codes in terms of permissible Cr(VI) levels in any form of industrial effluent, and safe occupational practice.

Concerns regarding trivalent chromium are less stringent, both because it has a much lower human toxicity rating and because it is less environmentally mobile––due to its low solubility and its tendency to become absorbed or complexed by organic molecules (Schroeder and Lee, 1975; Rai et al., 1987; Bartlett and James, 1988). Therefore, utilisation and disposal practises for wastes containing trivalent chromium (which by implication would be as solids) are considered to be of acceptable environmental risk, as it is presupposed that this oxidation state will be maintained indefinitely.

One prime exponent of such waste is slag arising from the stainless steel industry. A study by Kilau and Shah (1984) has presented evidence that for certain slag compositions atmospheric oxidation of trivalent to hexavalent chromium can nevertheless occur at ambient conditions. This presents some concern since a relatively benign waste material could potentially be rendered hazardous through this process.

The oxidation of trivalent chromium under environmental conditions has been studied at great length, primarily in aqueous environments. Numerous studies have focussed on a quantitative description of the oxidation reaction at the surface of various types of MnO₂ in an aqueous environment (Eary and Rai, 1988; Johnson and Xyla, 1991; Fendorf and Zasoski, 1992), driven by the belief that manganese oxides are the only naturally occurring substances capable of oxidising trivalent chromium to a significant extent. Direct oxidation of Cr(III) in aqueous solutions at natural pH levels of 4–9 by dissolved oxygen is known to occur, but with reaction half-lives in the order of years (Van der Weijden and Reith, 1982; Saleh et al., 1989). The rate of oxidation tends to increase with solution alkalinity. Schroeder and Lee (1975) found appreciable oxidation in aerated Cr(III) solutions at pH 10.5. Rapid oxidation has been observed in highly alkaline conditions at pH 12 and higher (Wood and Black, 1916; Petersen, 1998), although such conditions are rarely found in the natural environment.

Little is reported in the literature, however, about direct Cr(III) oxidation in the solid state at ambient conditions. Barnhart (1997) refers to the reaction, but dismisses it as very slow under environmental conditions. Petersen (1998) reports continuous leaching of Cr(VI) in small but appreciable concentrations from a column lysimeter filled with stainless steel slag over a 2 year period. The material was initially found to be benign and yielded very little Cr(VI) over the first few months, but concentrations increased thereafter. In the same study, appreciable formation of Cr(VI) was also reported to occur in balls prepared from powders of reagent grade Cr₂O₃ mixed with CaO.

Kilau and Shah (1984) propose the following reaction equation for solid state oxidation of Cr(III):

$$\text{Cr}{}_2\text{O}{}_3\text{+2}\text{CaO}\text{+1.5}\text{O}{}_2\text{→2}\text{CaCrO}{}_4$$

They postulate that certain stainless steel slags have a composition (in terms of Ca and Si concentrations), which favours the formation of a calcium chromite (CaO · Cr₂O₃) phase, which is essentially a solid solution of the two solid reactants in Eq. (1). The reaction is well known to occur at high temperatures (e.g. from the ceramics industry), but Hattori et al. (1978) have shown that the reaction is thermodynamically feasible also at ambient temperatures.

A description of the reaction mechanism and quantification of the rates, however, are, to the authors’ knowledge, still outstanding. This paper aims to present some further evidence with regard to the reaction and the factors that contribute toward it. Experimental work is focussed on stainless steel slags on the one hand, and mixtures of pure Cr₂O₃ and CaO powders on the other. The second study was done in order to investigate the reaction on the key reactants identified in reaction (1) in isolation of the remaining mineral phases present in the slag. By direct comparison between the two sets of experiment we draw some conclusions with respect to reaction mechanisms. Further, we provide an empirical description of rate and extent of the oxidation reaction, and finally discuss the significance of our findings for slag handling and disposal practices.