A. Sample preparation

The mill scale has previously been characterised (Wang et al., Reference Wang, Pinson, Chew, Monaghan, Pownceby, Webster, Rogers and Zhang2016b). In summary, its bulk composition is 71.0 wt% total Fe, 0.2 wt% CaO, 0.1 wt% Al₂O₃, 0.9 wt% SiO₂, 0.2 wt% MgO, and 0.6 wt% Mn. Its relative crystalline mineral phase composition determined by XRD is 53 wt% wüstite, 35 wt% magnetite, 10 wt% hematite, and 1 wt% goethite (minor green rust, Fe₂O₃.H₂O, also observed). The mill scale was added in progressively higher amounts to fine grained (<20 µm) synthetic Fe₂O₃ (Acros Organics, 99.999% purity), calcite, CaCO₃ (Thermo Fisher, 99.95%), quartz, SiO₂ (Sigma Aldrich, 99.995%), and gibbsite, Al(OH)₃ (Alcan OP25 Super White, 99.9%), which were mixed under acetone in a mortar and pestle with an intermediate drying and remixing stage to ensure homogenisation. Three sinter mixtures were prepared, with mill scale contents of 2.6–10.6 and to 21.2 wt%. Each mixture had bulk composition – 77.36 wt% Fe₂O₃, 14.08 wt% CaO, 3.56 wt% SiO₂, and 5 wt% Al₂O₃ – matching that of the bulk composition investigated in detail previously by Webster et al. (Reference Webster, Pownceby, Madsen and Kimpton2012, Reference Webster, Pownceby, Madsen and Kimpton2013a, Reference Webster, Pownceby, Madsen, Studer, Manuel and Kimpton2014, Reference Webster, Churchill, Tufaile, Pownceby, Manuel and Kimpton2016), and within the SFCA compositional stability domain established by Patrick and Pownceby (Reference Patrick and Pownceby2001) (i.e. each mixture was designed to form SFCA during heating).

B. In situ XRD data collection

Details of the in situ XRD experimentation have been described in detail previously (Webster et al., Reference Webster, Pownceby, Madsen and Kimpton2013a, Reference Webster, Pownceby and Madsen2013b, Reference Webster, Churchill, Tufaile, Pownceby, Manuel and Kimpton2016, Reference Webster, O'Dea, Ellis and Pownceby2017). The mill scale sample, and the three sinter mixture samples were heated over the range 20–1350 °C, at a heating rate of 20 °C min⁻¹ from 20 to 600 °C as the decomposition temperature of CaCO₃ was approached, and with individual datasets collected for 1 min continuously during heating. The rate was then reduced to 10 °C min⁻¹ for the range 600–1350 °C, which corresponded to the period of Ca-rich ferrite phase formation, reaction and decomposition for the sinter mixture samples. Samples were heated under a flow of a 0.5 vol% O₂ in N₂ gas mixture (to give a nominal oxygen partial pressure of pO₂ = 5 × 10⁻³ atm). The oxygen partial pressure has been used in all of our in situ XRD work, and the selection was based on the work of Hsieh and Whiteman (Reference Hsieh and Whiteman1989) who determined that this pO₂ maximised the formation of Ca-rich ferrites whilst still producing mineral assemblages similar to those found in industrial sinters.

C. In situ XRD data analysis

Rietveld refinement-based QPA was performed using TOPAS (Bruker, 2014). The crystal structure data of Blake et al. (Reference Blake, Hessevick, Zoltai and Finger1966), Hamilton (Reference Hamilton1958), Fjellvåg et al. (Reference Fjellvåg, Grønvold, Stølen and Hauback1996), Hazemann et al. (Reference Hazemann, Berar and Manceau1991), and Lager et al. (Reference Lager, Jorgensen and Rotella1982) were used for Fe₂O₃, Fe₃O₄, FeO, FeOOH and SiO₂, respectively. Corrections to account for sample displacement and peak intensity variation in asymmetric diffraction geometry were incorporated into the TOPAS refinement. The use of the Hill and Howard QPA algorithm (Hill and Howard, Reference Hill and Howard1987) embodied in TOPAS returns relative, rather than absolute, concentrations for crystalline phases in a system if amorphous material, including melt phases, are present. The previous in situ work performed in this context has demonstrated amorphous Al₂O₃ is present in these systems after the decomposition of Al(OH)₃. Therefore, absolute phase concentrations as a function of temperature were determined using the “external standard” approach given by Webster et al. (Reference Webster, Pownceby and Madsen2013b)

(1) $$W_i = \displaystyle{{\mu _mS_i{(ZMV)}_i} \over K}$$

Here W _i is the weight fraction of phase i, S _i is the Rietveld scale factor, ZM is the unit-cell mass, V is the unit-cell volume, and μ _m is the mass absorption coefficient of the entire mixture. K is an experiment constant used to put W _i on an absolute basis. It was calculated using the “known” concentrations of phases in the starting sample (the assumption is made that the sample is 100% crystalline), and the Rietveld-refined S values for phases in the sample in the first dataset collected at 20 °C according to

(2) $$K = \displaystyle{{\mu _m\sum\nolimits_{i = 1}^n {S_i{(ZMV)}_i}} \over {\sum\nolimits_{i = 1}^n {W_i}}} $$

D. Ex situ XRD data collection and analysis

Ex situ XRD data were collected on the mill scale sample, mixed in a 50:50 wt ratio with a very high purity and highly crystalline corundum (α-Al₂O₃; Baikalox Alumina Polishing Powder), over the range 5°–140°2θ using a PANalytical MPD instrument fitted with a cobalt long-fine-focus X-ray tube operated at 40 kV and 40 mA. The incident beam path was defined using 0.04 radian Soller slits, a 20 mm mask, a 0.5° fixed divergence slit, and a 1° anti-scatter slit. The diffracted beam incorporated a second set of Soller slits, a graphite monochromator to eliminate unwanted wavelengths and a 4.6 mm anti-scatter slit. An X'Celerator detector was used in scanning line (1D) mode with an active length of 2.122°2θ. Quantitative phase analysis was performed using the internal standard method (Madsen and Scarlett, Reference Madsen, Scarlett, Dinnebier and Billinge2008), in order to calculate the amorphous content of the mill scale.

A. Thermal decomposition of mill scale

Figure 1 shows a plot of accumulated in situ XRD data, viewed down the intensity axis and with temperature plotted vs. 2θ, for the experiment performed just using the mill scale sample. Figure 2 shows the results of the Rietveld-based QPA on these in situ XRD data, showing absolute phase concentrations as a function of temperature. Based on known phase equilibria between hematite/magnetite and magnetite/wüstite at a pO₂ of 5 × 10⁻³ atm, magnetite should be the stable Fe-oxide phase between about 1180 and 1750 °C (Huebner, Reference Huebner and Ulmer1971). Wüstite begins transforming first to magnetite at temperatures as low as ~300 °C. At these low temperatures, even though the pO₂ is at 5 × 10⁻³ atm, the Fe-oxide stability field is for hematite. As the temperature is further increased, the hematite is transformed to magnetite with a transition beginning above 1150 °C and being completed by 1300 °C. Graininess of this Fe₃O₄ above 1250 °C results in poor particle statistics, which has an influence on the QPA in this region.

Figure 1. (Colour online) In situ XRD data collected for the mill scale sample, viewed down the intensity axis, over the range 20–1350 °C at pO₂ = 5 × 10⁻³ atm.

Figure 2. (Colour online) Results of Rietveld-based quantitative phase analysis, showing absolute phase concentrations as a function of temperature, for the mill scale sample.

What is striking about the QPA plot is that the calculated absolute Fe₂O₃ concentration at 850 °C, which is after all of the crystalline FeO and Fe₃O₄ has oxidised, is 133 wt% and greater than was expected (106 wt%) from the sample gaining oxygen during heating. This means that there is unaccounted-for amorphous material in the mill scale sample at room temperature, which crystallises during heating. The results of the internal standard-based determination outlined in the section IID confirm this. Since the mill scale sample exhibited negligible loss on ignition (Wang et al., Reference Wang, Pinson, Chew, Monaghan, Pownceby, Webster, Rogers and Zhang2016b) this amorphous material is considered likely to be amorphous/nanocrystalline Fe-oxide material, rather than hydrated mineral(s). This work demonstrates the power of in situ XRD, combined with the external standard approach to QPA, for gaining insights into the quantity and nature of amorphous material in mineral mixtures.

Another noteworthy feature of the data shown in Figure 1 is the shift of the FeO reflections to lower 2θ (i.e. unit cell expansion) as this phase converts to Fe₃O₄. This is considered to indicate that the wüstite at room temperature is non-stoichiometric (i.e. Fe_1−x O), since the lattice parameter of Fe_1−x O is smaller than that of FeO (Minervini and Grimes, Reference Minervini and Grimes1999). The wüstite appears to become more stoichiometric as it converts to Fe₃O₄ during heating. A small shift to higher 2θ (i.e. unit cell contraction) as the FeO converts to Fe₂O₃ is attributed to the smaller size of Fe³⁺ relative to Fe²⁺ (Shannon, Reference Shannon1976).

B. Effects of mill scale addition on SFCA-I and SFCA formation

Figure 3 shows the plot of accumulated in situ XRD data for the 21.2 wt% mill scale mixture. Reflections for Fe₂O₃, Fe₃O₄, Fe_1−x O, CaCO₃, α-SiO₂, and Al(OH)₃ were observed at room temperature. The first event during heating (note that only the events distinct from those discussed in the section IIIA are discussed here) was the decomposition of Al(OH)₃ to amorphous Al-oxide, which was complete by ~300 °C. Then, CaCO₃ decomposed to CaO, which was complete by ~680 °C. The first Ca-rich ferrite to form was alumina-substituted dicalcium ferrite [designated hereafter as C₂(F_1−x A _x ) at ~800 °C, followed by CF (i.e. monocalcium ferrite, where C = CaO and F = Fe₂O₃] and CFA (a phase with average composition 71.7 mass% Fe₂O₃, 12.9 mass% CaO, 0.3 mass% SiO₂, and 15.1 mass% Al₂O₃) (Webster et al., Reference Webster, Pownceby, Madsen and Kimpton2012) at ~960 °C. As the temperature increased further, SFCA-I and SFCA formed at ~1110 and 1160 °C, respectively. Melting, which was complete by ~1260 °C, produced a phase assemblage of Fe₃O₄ in a Fe₂O₃–(Fe₃O₄)–CaO–SiO₂–Al₂O₃-rich melt. Each of these temperatures is consistent with those reported by Webster et al. (Reference Webster, Pownceby, Madsen and Kimpton2012) for the mixture SM4/5, which had the same bulk composition as the 21.2 wt% mill scale mixture analysed here, but did not contain mill scale. This suggests that the addition of mill scale has minimal effect of the formation of SFCA-I and SFCA.

Figure 3. (Colour online) In situ XRD data collected for the 21.2 wt% mill scale mixture.

Figure 4 shows stack plots of the in situ XRD collected for the 2.6, 10.6, and 21.2 wt% mixtures in the range 1100–1310 °C. Indeed, the effect of increasing mill scale concentration of the formation of the SFCA-I and SFCA phases does appear to be minimal. This is attributed, firstly, to the low impurity (i.e. Mg, Mn) content in the mill scale. Secondly, any Fe₂O₃ which forms through oxidation of the FeO and Fe₃O₄ in the mill scale only begins to convert to Fe₃O₄ above ~1150 °C (see Figures 1 and 2) and so no appreciable conversion of Fe₂O₃ to Fe₃O₄, which is unreactive in the context of SFCA-I and SFCA phase formation (Webster et al., Reference Webster, Pownceby, Madsen and Kimpton2013a), can occur before formation of SFCA-I and SFCA and the Fe₂O₃ is available for reaction to form the SFCA phases. If the pO₂ the experiments were conducted under was more reducing, then the effect of mill scale addition would likely become significant, in accordance with the results of Wang et al. (Reference Wang, Pinson, Chew, Monaghan, Pownceby, Webster, Rogers and Zhang2016b).

Figure 4. Stack plots of the in situ XRD collected for the (a) 2.6, (b) 10.6, and (c) 21.2 wt% mixtures in the range 1100–1310 °C. Datasets are offset in the vertical axes for clarity. The temperature values provided are those at the start – which is when the temperatures were automatically recorded – of the datasets. Reflections for key phases are marked in (a).