pct concentrations, simulating the environment created by either evaporation or dilution of seawater. Communications Raman spectroscopy was used to identify corrosion prod- ucts, by comparison of spectra obtained, with standard spec- tra for known iron oxides [6] after removal from test Reoxidation of Hot Briquetted Iron in environment. Surface area analysis was done in a Flow Salt Water Sorb II 2300 BET surface area analyzer. Briquettes were accurately weighed before immersion in salt solution, and JUANITA McKAY, ROBYN ARCHER, the difference in mass was determined by measuring the VEENA SAHAJWALLA, DAVID YOUNG, and mass subsequent to being washed and dried after salt solution TOM HONEYANDS immersion. Samples were kept in air after being removed from solutions and observed again after a period of about 4 weeks. Direct reduced iron (DRI) materials are sensitive to oxida- All briquettes immersed in solution were seen to corrode, tion, [1] resulting in metallization and economic loss and pos- but to different extents depending on the time of immersion, ing a potential safety hazard. Hot briquetting, reported to as shown in Table I. Briquettes immersed for only 1 to 2 be the best countermeasure to this reactivity, reduces the weeks showed little corrosion deposits. By 8 weeks, the porosity and surface to volume ratio of DRI. briquette surface was almost entirely covered by corrosion Thermodynamics predict that iron will corrode forming product that emerged at some points up to approximately 5 iron oxide products. The usual air oxidation of iron is slow mm above the original briquette surface. On the top surface at ambient temperatures and becomes prominent only at of the briquettes, the primary product was orange in color, temperatures above 500 °C; however, in the presence of which if removed had both green and black material below water, iron corrodes electrochemically. [2] The presence of it. The lower surface of the briquettes tended to be darker chloride ions in a corrosive environment accelerates rusting, in color. Briquettes exhibited significant surface cracking, as the chloride ion improves conduction within the electro- suggesting oxidation is not limited to the surface of the lytic cell. [3] briquettes, but occurs also within the interior. The limited work done on hot briquetted iron (HBI) reoxi- After a period of about 4 weeks of dry storage in air, HBI dation has shown that the porosity of the briquettes has a that had been in distilled water showed little change in very important influence on both the reaction rate and amount of corrosion product from when it had been removed extent. [4] The nature of the briquettes’ surface also effects from the solution. However, HBI, which had been immersed the corrosion: briquettes passivated in air have a much slower for 4 weeks or longer in any salt solution, showed significant corrosion rate in air at room temperatures than unpassiv- further degradation, in the form of lamination within the ated HBI. [5] samples, and red-brown coloring of the interior of the sam- Very little work has been reported in literature on the ples. Degradation behavior may indicate that once HBI has corrosion behavior and products of HBI. The focus of the been contaminated by a saline solution, salt precipitates are present investigation is to simulate conditions encountered sufficiently hygroscopic to render it more susceptible to during shipping or storage, to understand the reoxidation of HBI under various conditions, specifically the presence of atmospheric oxidation than briquettes exposed to moisture salt water, distilled water, and a dry environment. alone. The DRI was reduced from Mt Newman ore in a labora- Corrosion products were identified as lepidocrocite (- tory scale fluidized bed. Chemical analysis indicated it had FeOOH), iron trihydroxide (Fe(OH) 3 ), goethite (-FeOOH), a metallization of 96.6 pct and contained 0.09 wt pct carbon and maghemite (-Fe 2 O 3 ). The orange corrosion product on and 0.004 wt pct sulfur, with an O/Fe ratio of 0.03655. The the external surface of the briquettes was a mixture of various size distribution used was 0.063 1 mm. Laboratory HBI was iron oxides and hydroxides, particularly lepidocrocite. The produced from this DRI in a briquetter designed in the School green corrosion product was identified as goethite. Investiga- of Materials Science, UNSW. Briquettes were nominally 40 tion of a cross section of a briquette immersed in saline mm in diameter and 9 mm in height, having a weight of 50 solution for 4 days indicated that the center core of the g and a density of 5 g/cm 3 . briquette had traces of iron trihydroxide (Fe(OH) 3 ) present, Samples were placed in solutions of different salt concen- whereas the outer portions had magnetite (Fe 3 O 4 ) and hema- tration for up to 8 weeks, at 40 °C. Solutions simulated the tite (-Fe 2 O 3 ). After 14 days submersion, magnetite (Fe 3 O 4 ) chlorine concentration of seawater (26.7 g/L NaCl) and  10 was found throughout the center. These results can be compared to earlier reports of corro- sion products found under various conditions. Music et al. [7] said that -FeOOH, Fe 3 O 4 , and amorphous oxides were JUANITA McKAY, Postgraduate Student, VEENA SAHAJWALLA, Senior Lecturer, and DAVID YOUNG, Professor and Head, are with the found in distilled water at 20 °C, and the fraction of - School of Materials Science and Engineering, University of New South FeOOH increased as the chloride concentration increased. Wales, Sydney 2052, Australia. ROBYN ARCHER, formerly Undergradu- In seawater, they found -FeOOH and -FeOOH and FeOCl. ate Student, School of Material Science and Engineering, University of In an earlier study, [8] they reported the transformation of - New South Wales, is Metallurgical Engineer, Tyco Water, Sydney, Australia, 2161. TOM HONEYANDS, Senior Research Engineer, is with the BHP FeOOH to -FeOOH in solutions containing chloride ions Centre for Metallurgy and Resource Processing, Wallsend, NSW, Austra- at 90 °C. lia 2287. To establish the kinetics, the change in mass was moni- This article is based on a presentation made in the “Geoffrey Belton tored with respect to time. Corrosion reactions were slow at Memorial Symposium,” held in January 2000, in Sydney, Australia, under the joint sponsorship of ISS and TMS. low temperatures, reflecting the chemical reaction activation METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 31B, OCTOBER 2000—1133