Benchmarking the flotation performance of ores

Abstract

A porphyry copper ore containing chalcopyrite as the principal copper bearing mineral, and pyrite as the only other sulphide mineral, was treated in batch flotation tests under well defined physical conditions. The size-by-size flotation response was benchmarked against established calibration curves to infer an operational contact angle of the sulphide minerals as a function of particle size. The inferred operational contact angle values of the sulphide minerals were validated by independent measurements of contact angle on the concentrates and, in the case of chalcopyrite, by an indirect approach using Time of Flight Secondary Ion Mass Spectrometry (ToF-SIMS).

Recovery, flotation rate, and inferred operational contact angle increased with collector addition across all size fractions, with the intermediate and coarse size fractions benefitting the most from increased collector addition. The directly measured and inferred operational contact angles were in reasonable agreement, with an R² value of 0.7 across all size fractions. There was good agreement between the advancing contact angle values determined using ToF-SIMS and those calculated from direct contact angle measurement on the 53–75 μm size fraction for the case of chalcopyrite. A method for benchmarking flotation response has been developed, which may lead to better flotation process diagnostics and modelling.

Introduction

The flotation behaviour of model chalcopyrite particles has been characterised as a function of particle size fraction and advancing contact angle range under well defined physical conditions (Muganda et al., 2011a). The model chalcopyrite particles were liberated and of high purity such that the flotation response could be attributed to the effects of particle size and advancing contact angle. In the case of a natural ore, the principal scope of this current study, the degree of liberation has been identified as an important parameter determining floatability (Runge et al., 2003). Sutherland (1989) observed that copper bearing value minerals within all size fractions and in all liberation classes including those fully liberated, displayed two-component flotation behaviour. In the study by Sutherland (1989) it was not possible to link one floatability component to the degree of liberation or composition of the composite particle. Sutherland (1989) assumed that the hydrophobicity of the chalcopyrite surface was independent of the particle size and degree of liberation. Rather, differences in floatability were attributed to the amount of chalcopyrite in a particle, and not to differences in chalcopyrite hydrophobicity, implying that all exposed chalcopyrite surfaces had the same contact angle, irrespective of the particle size. In other studies, hydrophobicity was shown to depend on particle size and surface roughness (Drelich and Miller, 1992). In these studies, a pseudo-line tension was introduced to describe more realistically the effect of surface roughness on the line tension, which controls the critical contact angle of flotation, and therefore is a measure for particle hydrophobicity.

Previous investigations on the relationship between the particle size, contact angle and flotation behaviour of chalcopyrite (Muganda et al., 2011a) demonstrated that the advancing contact angle varies with particle size due to differences in surface species even for narrow particle size fractions subjected to the same treatment. This finding necessitates the determination of contact angle and flotation behaviour on a size-by-size basis, at least. For coarse particle size fractions, incomplete liberation may reduce the particle contact angle, while at fine size fractions (less than 10 μm) oxidation and metal ion hydrolysis may play key roles. Heterogeneity in surface oxidation of sulphides results in the variation of hydrophobicity, even within the same particle size fraction. Furthermore, the development of the contact angle with increasing collector addition is also apparently particle size dependent (Trahar, 1981). A distribution of particle contact angles within individual size fractions of a feed sample was demonstrated by direct contact angle measurements on flotation concentrates and tailings (Muganda et al., 2011b). In this current paper, it is assumed that multi-component flotation behaviour within a size fraction, and within any liberation class, may be explained in terms of the distribution of particle advancing contact angle. Thus floatability, and more specifically, the flotation rate and floatability components, is a function of particle size and contact angle, for a given hydrodynamic condition.

While the contact angle is evidently one of the most important parameters determining particle floatability, its measurement on heterogeneous particle mixtures in which the bulk is hydrophilic gangue is very problematic. As a result, researchers have resorted to characterising flotation response of natural ores in terms of particle size, liberation class, and collector surface coverage (Sutherland, 1989, Runge et al., 2003, Vianna, 2004). The surface coverage of collector molecules largely determines the hydrophobicity of the particle. The particle contact angle controls the rate of bubble–particle attachment (Holuszko et al., 2008), and hence the flotation rate constant. It may be possible to infer the contact angle of particles of a natural ore by benchmarking the flotation rate constant as demonstrated in our previous paper which focussed on model, single mineral chalcopyrite and pyrite particles (Muganda et al., 2011b).

The flotation rate constant has been considered to be the best descriptor of floatability (Imaizumi and Inoue, 1963). The challenge now, is to be able to predict flotation behaviour through a description of the rate constants for a given particle size fraction and contact angle range, but under strictly controlled and specific hydrodynamic conditions (Prestidge and Ralston, 1996). An attempt is made here to characterise the flotation response of chalcopyrite under well-defined physical conditions, and to use the floatability of single minerals to infer the effective operational contact angles of particles in an ore. The characterisation of the flotation response of chalcopyrite, as a single mineral, has led to the development of calibration curves of the undistributed rate constant, k^*, and collection efficiency, E_coll, against particle size fraction for different contact angle ranges (Muganda et al., 2011a). These curves have been used to benchmark the flotation responses of chalcopyrite with a different type of collector (Muganda et al., 2011b). It was shown that the flotation response of chalcopyrite is the same when the advancing contact angle, measured by the Washburn method on particle ensembles of a given size fraction, is the same, within experimental error. It was also shown that the calibration curves were valid at a higher pulp density (30% solids) of a mixture of chalcopyrite and non-interacting gangue mineral, paving the way for benchmarking the flotation response of a natural ore against the established calibration.

This paper tackles the determination of the operational advancing contact angle of both chalcopyrite and combined sulphide minerals (i.e., chalcopyrite and pyrite combined) in an ore feed using their flotation behaviour. A porphyry copper ore is floated at different collector additions and the flotation behaviour of chalcopyrite in each size fraction is benchmarked against the calibration curves. An attempt is made to verify the inferred contact angle values by directly measuring the advancing contact angles on size fractions of the concentrates and tailings from the flotation tests, and back calculating the contact angle of the feed. The approach is somewhat difficult because the bulk of the tailing is composed of hydrophilic gangue and the Washburn method, as demonstrated in this paper, is insensitive to low concentrations of hydrophobic particles in a bulk hydrophilic matrix. In the same vein, direct contact angle measurements on the feed size fractions were not successful because the hydrophobic component was also too low. Furthermore, there is the potential issue of surface oxidation which may take place during sample preparation of sulphide minerals for the contact angle measurement. However, the techniques developed for sample preparation seemed to preserve the surface species (Muganda et al., 2011b). Indeed, these difficulties were precisely the primary motivation for developing the method of inferring the advancing contact angle by benchmarking flotation response against a well-defined standard. The low concentration of hydrophobic copper bearing mineral in the ore precluded the direct measurement of contact angle using the Washburn technique, as discussed in this paper.