investigations of phase composition of three-component

METAL 2003 20.-22.5.2003, Hradec nad Moravicí ___________________________________________________________________________________

1

Jan Mróz, Przemysław Francik, Ryszard Budzik Politechnika Częstochowska Poland

Investigations of phase composition of three-component slags of type CaO-FeO-SiO2

Abstract

The reduction of iron oxides from liquid slag is the fundamental stage of many newly developed, future-oriented technologies of obtaining liquid iron (pig iron), which are referred to as reduction iron smelting processes, whose main distinguishing feature is the use of non-coking coals. The main components of such slags are: CaO, SiO2 and FeO. The main aim of investigations of FeO reduction from the system CaO-FeO-SiO2 is to evaluate the kinetic parameters of the process. One should suppose that these parameters are strongly affected by the phase composition of slags. In this paper the results of investigations of phase composition of slag of type CaO-FeO-SiO2 are presented. The slags were of basicities CaO/SiO2 equal to 1.27 and 1.80 and were heated up to 1420oC. The microscopic observations were made on the scanning microscope equipped with the elements microanalyser. The results revealed great differences between particular microphases as far as the chemical composition are concerned. Moreover, one may notice characteristic areas of specific arrangements of some phases. These results strongly supports the theses that the knowledge of structure of slags is indispensable to explain some phenomena of FeO reduction from liquid slags.

I. INTRODUCTION

The present technology of the production of liquid iron in blast furnaces continues to be a

technology prevailing in the world’s manufacture; however, the awareness of the very limited

resources of coking coals demands searching for new production technologies that would be

based on the immense resources of non-coking coals. The most advanced new technological

processes of liquid iron production are Corex [1] and Romelt [2], as well as the Hismelt, Dios,

CCF and AISI-DOE processes. Despite the visible advancement of some technological

processes, laboratory studies are still the basic element of the future progress and development

of those processes. As commonly known, the processes of reduction of iron oxides from liquid

slags are characterized by a specific cumulation of partial processes, such as the processes of

Fe2O3 and Fe3O4 dissociation, melting down, the dissolution of FeO in the bath, the conversion

of hard coal in the metal bath, i.e. the decomposition of volatile constituents, the carbonization

of coal and its dissolution and gasification in the slag-metal bath and reduction reactions

between oxides and reducing agents in the slag-metal bath [3-13]. Moreover, during studies

conducted on reduction, new phenomena appear, which represent new research problems in

the sphere of iron oxide reduction from liquid slag phases. The present study reports the

results of the investigation of FeO reduction from liquid slags of increased basicity. As the

investigation reveals, the process of reduction from these slags is a process which is more

complex than that from slags of the “classical” basicity of 1.0-1.2, or from acid slags. New

phenomena appear, which indicate a more complex mechanism of reduction.


2

II. THE RATE OF FeO REDUCTION FROM LIQUID SLAGS

Investigations of FeO reduction from liquid slags was carried out using synthetic ternary

slags of CaO-FeO-SiO2 type [14]. The basicity of those slags, as defined by the CaO/SiO2

ratio, was 1.27; 1.6; 1.8; and 2.0, respectively. All slags contained 60% FeO. The reducer was

graphite in the shape of a cylinder of a diameter of 19.5 mm and a length of 45 mm. Reduction

was conducted at a temperature of 1420°C. The rate of reduction was determined based on the

readings of an automatic carbon monoxide and carbon dioxide analyzer indicating the contents

of those gases in the carrier gas, i.e. argon. A computer was connected to the measuring

system, which recorded data indicated by the analyzer at 10s intervals.

A summary of testing results is shown in Fig. 1.

Fig. 1. Summary figure of reduction rates vs time for slags of given basicities

In the case of the reduction of slags with basicities of 1.8 and 2.0, two reduction stages are

distinguished: in the first, initial stage, for a period of several tens of seconds, the reduction

proceeds at a very high rate reaching, respectively, 95⋅10-6 mole FeO⋅cm-2⋅s-1 and 195⋅10-6

mole FeO⋅cm-2⋅s-1, then it passes to the second stage of running, stable in terms of the rate,

with an average rate of approx. 40 ⋅10-6 mole FeO⋅cm-2⋅s-1. It is characteristic that the rate of

reduction from the slags of an increased basicity of 1.6 and 1.8 stabilizes in all the three cases

at a level of approx. 40 ⋅10-6 mole FeO⋅cm-2⋅s-1. It is visible, at the same time, that the rate of

reduction from the 1.2-basicity slag is approx. 4 times lower than the rate of reduction from

slags of a higher basicity.

The phenomenon of the several-times higher rate of reduction in the case of increased-basicity

slags is not easy to explain because of the fact that the viscosity of slags with increased


3

basicity – as calculated from the model proposed by Ji et al [15] – is higher than that of the

reference slag of a basicity of 1.2, thus the transfer conditions are more difficult in slags with

increased basicity.

Reference [14] provides the qualitative description of a likely mechanism of FeO reduction

from increased-basicity slags. It assumes that this mechanism is associated with the problem

of uniformity and homogeneity of liquid slag phases. The reported investigation results

suggest indirectly that, in terms of phase composition, slag is not homogeneous in its liquid

state, too, and this inhomogeneity grows with increasing CaO content. It can be hoped that

further investigations, also those concerning the structure of slag in the liquid state or in a state

very close to liquid, will help explain the phenomenon of the high rate of FeO reduction from

slags of increased basicity.

III. MINERALOGICAL EXAMINATION AND PHASE COMPOSITION EXAMINATION OF MELTED CaO-FeO-SiO2 TERNARY SLAGS

Slags of highly diverse basicities of 1.27 and 1.80, containing 60% FeO (slags after

reduction carried out at a temperature of 1420°C with the use of a rotating graphite disc) were

selected for mineralogical examination. The degree of reduction did not exceed 2%. The slags

were cooled at a rate of 15K/min. The tests were carried out at the Institute of Iron Metallurgy

in Gliwice. They included the determination of the morphology of specimen surfaces and the

identification of the phase composition of the characteristic, diverse micro-areas of specimens

tested. The tests were carried out on an XL-30 electron scanning microscope equipped with

an Edax elementary composition analyzer by Philips.

Examination results for the 1.27-basicity slag. The microscopic examination of this slag

showed a high diversification in its structure both in terms of phase composition and the forms

of occurrence. This is visible in Fig. 2a (magn. 200x) and in Figs. 2b and 2c representing

magnifications of selected areas from Figure 2a. The characteristic features is the occurrence

of highly dispersed “light” phases shaped in the form of spikes and in the form of small balls

(Figs. 2b,c). It is interesting to note that these phases form specific areas of geometric forms

with clearly outlined interfaces. In Fig. 2a, a light phase is also visible in the form of large

aggregates (X1), which is reduced metallic iron. For the purposes of microanalysis, another

magnification (2000x) of the examined areas was made, which is shown in Figure 2d. The

elementary composition of the dispersed light phases and the phases forming a “background”

was given in Table 1. It can be seen from the data shown in the table that the light phases in

the form of spikes and small balls contain (in addition to oxygen) chiefly iron and calcium.

Silicon, also occurring in this phase, is in much less concentration compared with calcium

(with concentration being lower by 3-4 times). This would suggest that the “light” phases

correspond mainly to calcium ferrites.

META

L 2003 20.-22.5.2003, Hradec nad M

oravicí ___________________________________________________________________________________

4

Surface morphology of slag sample of type CaO – FeO – SiO2 of basicity 1.80, 60% FeO, temp. 1420 OC a). magn. 100 x, b). magn. 2000 x

Fig. 3 a, b. Surface morphology of slag sample of type CaO – FeO – SiO2 of basicity 1.27, 60% FeO, temp. 1420 OC c). magn. 500 x, d). magn. 2000 x

Fig. 2 c, d. Surface morphology of slag sample of type CaO – FeO – SiO2 of basicity 1.27, 60% FeO, temp. 1420 OC a). magn.. 200 x, b). magn. 400 x

Fig. 2 a, b.


5

The light grey phase (X3) contains iron in a much smaller quantity (13.32%) than the light

phase, with silicon occurring here in a larger amount (10.93%). Also calcium occurs in this

phase in the amount of 13.42%, which allows one to suppose that this phase corresponds to

iron-calcium olivines (CaxF1-x)s⋅SiO4.

Table 1. Percentage contribution of particular elements in melted slag of type CaO-FeO-SiO2 (60% FeO), basicity 1.27, temp. 1420oC.

number of phase (description of phase)

element X1

X2 (dark-grey)

X3 (lihgt-grey)

X4 (light spikes) light balls

O - 42.60 51.09 49.84 47.50

Si - 8.68 10.93 2.92 2.82

Ca - 25.84 13.42 9.72 7.41

Fe 100 1.46 13.32 28.40 34.48

The dark grey phase (X2) contains minimum quantities of iron (1.46%), with the main

constituents of this phases (beside oxygen) being calcium (25.84%) and silicon (8,68%). The

mutual quantitative proportions of calcium and silicon allow one to suppose that this can be a

bicalcium silicate, 2CaO⋅SiO2 (larnite), phase. As this compound has a very high melting point

(2150°C), it can be supposed that its occurrence is associated only with the alloy solidification

phase, while during melting, in the presence of other phases of the CaO-FeO-SiO2 system, it

transforms into other low-melting phases and compounds. This mechanism is likely to apply

also to other phases, therefore it can be supposed that in the liquid state – just as in the solid

state – there is a significant diversification in the concentration of individual constituents in

the micro-volumes of the slag alloy. It should also be stressed that the presence of specific

constituents in an amount close to the stoichiometric amount for a particular compound does

not necessarily mean the occurrence of that compound in a fully crystallized form, which is

associated with the cooling rate. This appears to be confirmed by the existence of calcium

ferrites in small amounts and in poorly crystallized forms, which was shown by diffraction

analysis [16].

Examination results for the 1.80-basicity slag. The results of the examination of surface

morphology of this slag are shown in Figure 3. In Figure 3a (magn. 100x), numerous light

elongated “dendrite” phases are visible on a non-uniform light grey background. In Fig. 3b

there are shown (in a magnitude of 2000x) another phase in the form of a white “mosaic”

against a dark grey background. The quantitative description of these phases, which takes into

account main elements forming these phases, is shown in Table 2.

As can be seen from the tabular data, iron oxide FeO, is concentrated chiefly in two phases,

i.e. in the “dendritic” phase with an iron content of 52.18%, and in the “mosaic” phase


6

constituting mainly a mineralogical background for the slag, with an iron content of 30.53 Fe.

It is characteristic that no silicon at all occurs in the “dendritic” phase, which means that this

phase does not contain silica. In the mosaic phase, both silicon and calcium occur, and the

CaO/SiO2 ratio is relatively high, amounting to 3.43.

Table 2. Percentage contribution of particular elements in melted slag of type CaO-FeO-SiO2 (60% FeO), basicity 1.80, temp. 1420oC.

description of phase

element light phase, big conglomerates

(fig. 3b)

light-grey phase (fig.3b)

black phase

(fig. 3b)

light, mosaic phase, (fig. 3b)

O - 45.41 41.73 48.32

Si - - 7.97 3.60

Ca - 2.41 26.42 12.34

Fe 100 52.18 1.48 30.53

IV. SUMMARY

The investigation of the reduction of FeO from higher basicity slags of CaO-FeO-SiO2 type, that is slags of basicities of 1.6, 1.8 and 2.0, has shown that the behaviour of the reduction is characterized by occurrence of the very high rate of reduction process, compared to typical basicities of slags. Considering the fact that the higher basicity slags investigated are characterized by a higher dynamic viscosity than that of the reference slag, the phenomenon of the unexpectedly high rate of reduction from the higher basicity slags can be explained by presuming the occurrence of a reduction mechanism other than the diffusion control or activation control mechanisms. It has now been assumed that an important role in FeO reduction from higher basicity slags may be played by the Marangoni phenomenon. This implies at the same time, that liquid slags constitute a set of microphases of a very diverse chemical composition and diverse physico-chemical properties. Further experimental and theoretical studies on the structure of liquid slags would explain the phenomenon of the high rates of reduction from higher basicity slags.

V. LITERATURE 1. J. Flickenschild: Metal. Bull. Monthly Suppl., 1996, vol. 308, August, pp. 10-13 2. V. Romenets: The Romelt Process, ISS Ironmaking 2000, Myrtle Beach SC, October, 1994. 3. G. Heinrich, H. Pook, H. D. Waldhecker, K. Yamamori: 1984, nr. 10, s. 45. 4. H. Nakajima, K. Okane, S. Furujo, S. Okamura, M. Sueyasu, T. Tanoue, S. Anezaki, T. Matsuo:

Ironmaking a. Steelmaking, 1983, nr 3, s. 130. 5. H. Pook, D. Waldhecker, Y. Yamata, M. Sato: Seminar Steel Committee (Economic Commision

for Europe, United Nations). Izmir, Turkey, 5-9 May, 1986. 6. J. Hartwig, D. Neushütz: Ironmaking a. Steelmaking, 1983, nr 3, s. 124. 7. V. Romenets: The Romelt Process, ISS Ironmaking 2000, Myrtle Beach SC, October, 1994. 8. F. Oeters, S. Orsten: Steel Research, 1989, nr 3-4, s. 145. 9. K. Ragland, C. A. Weiss: Energy, (Oxford), 1979, vol. 4, nr.2, s. 341. 10. J. Mróz: Steel Research, 1998, nr. 12, s. 465. 11. J. Mróz: Metall. Trans. B, 2001, vol. 32B, nr. 5, pp. 263-67 12. R. Pattipati, C. Wen: Chem. Eng., Montreal, Canada, Oct. 4, 1981, Canad. Soc. Chem., 1982, s. 42 13. R.J. Fruehan, K. Ito, B. Ozturk: Steel Research, 1989, vol. 60, nr 3-4, p .129 14. J. Mróz: Mills Symposium: „Metals, Slags, and Glasses: High Temperature Phenomena”, London,

22-23 August 2002. 15. F.-Z. Ji, D. Sichen, S. Seetharaman: Metall. a. Mat. Trans. B, 1997, vol. 28B, Oct., s. 827. 16. J. Mróz: Investigations of reduction of iron oxides from liquid phase using rotating disc

methodology, Report of grant (no. 7 TOB 031 13) of State Committee for Scientific Research, 1999, (not published).

investigations of phase composition of three-component

Documents