Activation of the carbothermic reduction of manganese ore

https://doi.org/10.1016/S0301-7516(02)00045-5Get rights and content

Abstract

The effect of extended milling on the carbothermic reduction of a manganese ore has been examined using a combination of thermal analysis and X-ray diffraction (XRD). Thermodynamic modelling indicated that reduction of MnO2 to MnO was possible at 25 °C, although no reaction was found to occur during milling of the ore with graphite for up to 10 h. For a physical mixture, cryptomelane, KMn8O16, reduced at 500 °C and braunite, Mn7SiO12, at 700 °C after 10 h milling these temperatures were reduced by 200 °C. The initial product was Mn3O4, although in the 10-h-milled powder, the reduction of braunite may have been directly to MnO. Reduction at 600 °C only formed Mn3O4 in the unmilled powder but the major product in the 10-h-milled powder was MnO. The increased extent of reaction after premilling may allow current processing plants to expand their throughput without increasing the size of reduction kiln.

Introduction

One of the major components in alkaline batteries is γ-manganese dioxide; this phase is also present in lithium batteries as the intercalation host. The major commercial route to γ-MnO2 is by electro-oxidation of purified manganese sulphate solution on titanium anodes. The crude solution is commercially prepared from natural manganese ores via thermal reduction to soluble MnO using carbon as reductant. The wide variation in the chemical composition of manganese ores makes comparison between published reduction reports difficult Akdogan and Eric, 1994, Akdogan and Eric, 1995, Bakardjieva et al., 2000, Berg and Olsen, 2000, Gillot et al., 2001, Ostrovski and Webb, 1995, Paixao et al., 1995, Stobbe et al., 1999, Terayama et al., 1996, Zaki et al., 1997, Zaki et al., 1998, Zhang and Schleich, 1994. In general, carbon reduction has been examined at >1100 °C Akdogan and Eric, 1994, Akdogan and Eric, 1995, Ostrovski and Webb, 1995, Terayama et al., 1996, Zhang and Schleich, 1994 with resultant formation of carbide phases. Lower temperatures have been examined, but these have used hydrogen (Zaki et al., 1997), carbon monoxide Berg and Olsen, 2000, Zaki et al., 1998 or methane (Stobbe et al., 1999) as reductant. It is certain that the differing minerals present in the ores used will all have different reduction characteristics and there is likely to be some synergetic effects between phases at different stages of reduction (Ostrovski and Webb, 1995), making each ore unique in its characteristics.

A number of other nonthermal reduction routes are under investigation Das et al., 1998, Kholmogorov et al., 2000, Nayak et al., 1999, Petrie, 1995, Trifoni et al., 2001, Veglio et al., 1997, Veglio and Toro, 1994a, Veglio and Toro, 1994b; most have yet to be applied beyond the bench scale, but reduction using sulphur dioxide is being tested at a demonstration plant (Anon, 2001).

The leach solution is neutralised with lime and both gypsum (CaSO4·2H2O) and jarosite [MFe3(SO4)2(OH)6, M=Na+, K+, H+ and/or NH4+] precipitate, the latter removing the dissolved iron and a range of other impurities. The resultant solution is high purity manganese(II) sulphate at circumneutral pH and electro-oxidation at >90 °C is used to convert this into γ-MnO2.

As the world grows, the number of batteries in use is increasing and will continue to do so for many years to come. The present consumption of MnO2 in batteries is around 220,000 tons per annum and this is expected to grow by around 5% per annum (Anon, 2000). Remarkably, little manganese is recycled from dry cell batteries, with the major commercial interest being in recycling lithium and rechargeable batteries. Consequently, a number of companies are presently expanding by building new plants or increasing the capacity of current operations to meet anticipated demand.

Methods of increasing capacity within an existing plant include replacement of equipment with a larger model, duplication of existing plant and increasing throughput in existing equipment. The first route may require the plant to be closed whilst replacement is made if building space is limited, the second route essentially builds a new plant alongside the first increasing the number of streams within the plant and increasing the engineering complexity. In many plants, there is no method of increasing the throughput of existing equipment as optimisation ensures that each operation is at the maximum capacity to maximise production.

It has been shown previously Welham, 1996, Welham, 1997a, Welham, 1998a, Welham, 2000b that the rate of carbothermic reduction reactions can be greatly increased by premilling the mineral and carbon together when compared with powders milled separately and then mixed. This rate increase leads to a larger amount of material reduced per unit time and therefore increases the throughput of the thermal reduction stage. Thus, were the same to be observed for the carbothermic reduction of MnO2, existing plants may not have to replace their current kiln but simply add a suitable mill prior to the kiln feed storage bin. The typical particle size after milling is <50 μm which would also be expected to lead to a greater rate in the subsequent dissolution in sulphuric acid. Indeed, milling of minerals has also led to rates of dissolution increasing faster than the surface area increase, even with no chemical reaction Welham, 1997b, Welham, 2000a, Welham, 2001, Welham and Llewellyn, 1998.

This paper examines the effect of premilling on the carbothermic reduction of manganese ore.

Section snippets

Thermodynamic appreciation

The Ellingham diagram for the Mn–O–C system, calculated for a total pressure of 1 atm, is shown in Fig. 1. This diagram indicates that MnO2 should spontaneously reduce upon heating to >510 °C and the Mn2O3 formed would also reduce by heating to >915 °C, leaving Mn3O4. The loss of oxygen during heating of MnO2 has been observed previously at 500 °C and for Mn2O3 at 900 °C (Stobbe et al., 1999). The reduction of MnO by carbon at atmospheric pressure is only feasible at >1410 °C, which explains

Experimental

A sample of commercial manganese ore was obtained from a mine site in Western Australia and comprised lumps which were typically >30 mm. One lump was smashed using a hammer to <5 mm and then ground in a ball mill for 2 min; the resultant powder was screened at 212 μm and the oversize reground. This cycle was repeated three times and any remaining oversize was discarded.

Three powders were prepared in accordance with the anticipated reaction:2MnO2+C=2MnO+CO2but using a 20 mol% excess of graphite

Results and discussion

XRF analysis of the powdered concentrate showed the presence of 48.83 wt.% Mn, 1.03% Si, 1.17% K, 1.11% Fe and 1.38% alkaline-earth metals. XRD of the raw concentrate is shown in Fig. 3, the major phases present are braunite, Mn7SiO12, and cryptomelane, KMn8O16. If it is assumed that Si and K are only present as braunite and cryptomelane, respectively, then from the analysis, the concentrate is 28.4% cryptomelane, 28.5% braunite, 4.3% CaO+MgO. The remaining 38.8 wt.% was composed of a phase

Conclusions

Premilling of a manganese ore, comprised of cryptomelane, braunite and an unidentified manganese phase with graphite led to enhanced reduction at decreased temperatures. The longer premilled powder showed complete reduction to MnO within 30 min at 600 °C, the unmilled powder showed Mn3O4 as the major phase after 30 min at 800 °C. Cryptomelane was found to reduce at a lower temperature than braunite, although both phases were reduced to Mn3O4. The reduction of Mn3O4 was enhanced more by the

References (38)

Cited by (0)

View full text