Degree Name

Doctor of Philosophy


School of Mechanical, Materials and Mechatronic Engineering - Faculty of Engineering


A significant body of work exists around coke dissolution into liquid iron, however there are key aspects of this important reaction that are not well described. This study was focused on gaining the answers to the questions “How does the coke mineral matter behave during coke dissolution?” and “Can the effects of sulphur and oxide layer formation on the dissolution rate be separated and quantified?”. Issues that must be addressed if the understanding of the kinetics of this reaction is to be advanced and coke's use in the blast furnace further optimised. To this end, a detailed investigation was conducted examining the influence of coke mineral matter on coke (carbon) dissolution into liquid iron. This project was focused on the mineral matter layer that forms at the coke/iron interface and how the presence of this layer affects the kinetics of carbon dissolution from the coke into the liquid iron. A range of experimental techniques were used to identify and characterise the mineral layer as it formed at the coke/iron interface and to assess the layers influence on the carbon dissolution kinetics. Carbon dissolution experiments, utilising a carburiser cover technique, were carried out where carbon and sulphur transfer to an iron-carbon melt was measured over time at temperatures of 1450°C, 1500°C and 1550°C. This technique allowed fundamental data on the dissolution rate constant to be calculated, and the effects of temperature, melt sulphur and different carbonaceous materials to be explored. Development of the mineral layer at the coke/iron interface as a function of both time and temperature was studied utilising a quenched carbon dissolution technique that was developed during the project. This technique had the additional benefit of allowing the composition of the melt to be determined for a larger range of elements than the dissolution experiments. The quenched carbon dissolution experiments complemented the carbon dissolution experiments and allowed direct comparisons between the dissolution behaviour measured in the dissolution experiments and the composition and morphology of the mineral layer observed in the quenched samples. The dissolution studies were further complemented by sessile drop measurements of the wetting behaviour of iron on the mineral phases that were identified in the mineral layer present at the coke/iron interface, thermodynamic modelling utilising the MTDATA software package and a conceptual model of the interfacial mineral layer. A mineral layer was observed to have formed at the coke/iron interface during coke dissolution into liquid iron at experimental temperatures ranging from 1450°C to 1550°C. The mineral layer was solid at these temperatures and persistent at the interface. The amount of mineral matter present in the mineral layer was observed to be increasing with increased reaction time. The composition and structure of the mineral layer changed with both experimental time and temperature. The composition of the mineral layer was principally composed of oxides of aluminium and calcium, present as various calcium aluminates and calcium sulphides. Initially the mineral layer was a loose agglomeration of particles of which a majority were alumina particles. As the dissolution reaction proceeded, the loose agglomeration of particles was replaced by an open acicular layer that was predominantly the calcium aluminate CaO.6Al2O3 (CA6). As the dissolution reaction continued further, the calcium aluminates became increasingly richer in calcium oxide, with the predominate phase present in the mineral layer progressing through the calcium aluminates phases CA6 to CaO.2Al2O3 (CA2) and onto CaO.Al2O3 (CA). The apparent calcium enrichment of the mineral layer was observed to occur more rapidly as the experimental temperature increased. Accompanying the compositional changes in the mineral layer, the morphology of the mineral layer was also observed to change. The mineral layer was formed as an initial loose agglomeration of alumina particles, changing to an open acicular structure consisting of CA6/CA2 before being converted to a dense layer as the dissolution reaction proceeded and the composition of the mineral layer changed to CA and calcium sulphide (CaS) appeared at the interface. It was found that the formation of the calcium sulphide layer was preceded by the formation of the calcium aluminate layer. Only after the calcium aluminate layer had experienced progressive calcium enrichment and the CA and CA2 phases had formed did the CaS phase appear at the iron interface. Thermodynamic analysis of the experimental results confirmed that the formation of the calcium enriched calcium aluminates, CA2 and CA, were a necessary requirement to stabilise the calcium sulphide layer for the coke composition studied. The kinetics of carbon dissolution from the coke to the liquid iron were observed to be dependent on the structure of the interfacial mineral layer. This dependence was manifest as two stage behaviour in the first order mass transfer plots. The initial stage, characterised by rapid carbon dissolution, was observed while the mineral layer was developing or had the open acicular structure of the CA6/CA2 layer. The second stage was characterised by a significant reduction in the rate of carbon dissolution. The onset of the second stage was coincident with the change in the composition of the mineral layer from CA6/CA2 to CA2/CA and the associated densification of the mineral layer. In stating that the nature of the mineral layer affects the dissolution kinetics, a change in the reaction control mechanism is implied. This represents a change in the coke dissolution kinetics from simple mass transfer control to a mixed control regime where both mass transfer and the mineral (product) layer are active. In an attempt to overcome the problems associated with the heterogeneity of coke a coke analogue was developed. In the coke analogue the composition and dispersion of the carbonaceous and mineral matter (oxides) are controlled and the porosity is fixed. When single phase calcium aluminates were introduced into the coke analogues, calcium enrichment of the resulting calcium aluminate mineral layer was observed. The observed carbon dissolution kinetics were dependant on the structure of the interfacial calcium aluminate layer. Consistent with the coke dissolution studies, the calcium aluminate layer formed at the coke analogue iron interface during carbon dissolution was at least in part rate controlling the carbon dissolution reaction for the coke analogues studied. Utilising the sessile drop experimental technique the wettability with liquid ironcarbon-sulphur alloys of the predominate phases that were observed in the mineral layer were measured. It was observed that the contact angle decreased as the proportion of lime (CaO) in the calcium aluminate increased. Further it was observed that while the presence of sulphur in the melt increased the contact angle for the alumina and CA6 substrates, on the CA2 and CA substrates the contact angle was decreased. The improvement in the wetting of the CA2 and CA substrates with sulphur was attributed to the formation of CaS at the substrate/droplet interface. This study has produced new fundamental data on the growth and development of the mineral layer and the wettability of the predominate calcium aluminates observed in the mineral layer. These detailed studies have illuminated the changing nature of the layer in terms of both composition and morphology and found that the kinetics of carbon dissolution from the coke to the liquid iron were dependant on the structure of the interfacial mineral layer.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.