H 2 adaptation in ironmaking: a review with focus on suspension smelting and its reaction mechanisms

Abstract

Suspension smelting technology is emerging as a promising alternative in ironmaking to conventional hydrogen-based direct reduction (H-DRI) for low-carbon ironmaking. Unlike the slower solid-state reduction reactions in DRI processes, suspension smelting uses rapid in-flight reduction of fine iron ore particles. Within seconds, the ore fines undergo several process steps like heating up, phase transformation, reduction and partially melting, due to the high temperatures. This review explores hydrogen adaptation in suspension smelting technologies such as the flash ironmaking technology (FIT) and the HIsarna^® process. The key reaction mechanisms, thermal decomposition and gas reduction with H₂/CO, are examined through thermodynamic and kinetic analyses. A comparison of reduction kinetics between classical DRI processes and suspension smelting is provided, highlighting the significantly faster reaction rates of suspension smelting. Finally, the influence of the water–gas shift reaction and its potential negative impact on the reduction behaviour is analysed.

Keywords

suspension smelting flash ironmaking hydrogen iron ore kinetics water–gas shift reaction

Introduction

The steelmaking industry accounts for 7–9% of the global anthropogenic emitted CO₂ emissions,¹ with an annual contribution estimated at around 2.6 Gt (2019).² The conventional iron- and steelmaking route is commonly based on either the blast furnace (BF) or the direct reduction of iron ores as the initial ironmaking step.³ In 2019, the amount of pig iron produced by the blast furnace was 1208.7 Mt, while the amount produced by direct reduced iron (DRI) was 107.6 Mt, indicating the much stronger industrial presence of the blast furnace.⁴ In BF ironmaking, iron ore is transferred to molten metallic iron, which can be directly cast to pig iron or mixed and treated with recycled steel in a basic oxygen furnace (BOF), whereas solid iron is produced by direct reduction, which is treated together with recycled steel in the electric arc furnace (EAF) to create molten steel.³ Around 75% of iron used in the BOF steelmaking route is based on the BF hot metal, and around 25% is scrap. In contrast, the EAF can run on up to 100% scrap, but also utilise large shares of DRI.^5,6 The World Steel Association summarises CO₂-emissions and energy consumption for the main steelmaking routes.⁷ Data from 2021⁷ (Table 1) shows that most of the global steel production (71%) is based on the BF-BOF route. However, this route is responsible for high energy consumption and the most CO₂ emissions. Even though DRI-based steelmaking shows (slightly) higher energy intensity compared to the BF-BOF route, its CO₂ emissions are lower. Lastly, the recycling route using scrap as an input in the EAF shows the best environmental performance, both in terms of energy consumption and CO₂ emissions. While recycling is clearly the most sustainable option, recycling alone cannot fulfil the ever-growing demand for steel. Hence, primary production will remain significant in the future.⁸

Table 1.

CO₂ emissions and energy consumption of steelmaking routes in 2021.⁷

	CO₂ emission intensity[t_CO2/t_{crude steel]}	Energy intensity[GJ/t_{crude steel}]	Share in global production (%)
Global average	1.91	21.31	100
BF-BOF	2.32	24.43	71
Scrap-EAF	0.67	10.04	22
DRI-EAF	1.65	25.29	7

The conventional steelmaking route heavily depends on the use of carbon in the form of coal or coke as a reducing agent.⁹ The EU aims to become carbon neutral by 2050 and to cut 55% of its emissions by 2030,¹⁰ which forces the steel industry, as a large emitter, to move towards more sustainable production. Multiple strategies have been investigated to help reduce carbon emissions. Examples include the use of carbon capture and storage/use (CCS/U), the use of biomass in the reduction process, plasma-based iron ore reduction or electrolytic approaches.¹¹ Yet, the most popular approach to actively decrease the emitted CO₂ from the production process is using green hydrogen to substitute carbon as a reducing agent for iron ore.¹¹ The adaptation of H₂ in ironmaking processes could lead to various advantages compared to carbon-based reduction. First, pure H₂-based reduction will only emit H₂O. Secondly, H₂ shows faster kinetics compared to carbon-based reduction,⁹ resulting in a higher productivity. However, problems can also arise from utilising H₂. In contrast to CO reduction, the reduction of iron oxides with hydrogen is endothermic,⁶ requiring additional (external) energy supply to keep the high temperatures needed during the process. In addition, carbon is an essential component for the transition of iron to steel, which cannot be completely omitted.¹²

To decrease the CO₂ emissions arising from this intense use of carbon, hydrogen is increasingly investigated as a substitute. Ironmaking processes utilising hydrogen as a reducing gas can be classified based on the type and shape of input material, the reducing gas composition or furnace technology, where, for example, shaft furnace processes like MIDREX operate with pellets and fluidised bed applications like CIRCORED with fine ore as an input material.¹³ Most H₂ adaptation nowadays focuses on the DRI processes, where the shaft furnace is currently at the furthest state considering the technology readiness.¹¹ However, processes like HIsarna^®, developed by Tata Steel, and the Flash Ironmaking process, developed by the University of Utah, are two new alternative ironmaking approaches based on the principle of suspension smelting.^14,15 A significant difference of the suspension smelting process compared to the DRI processes is the much higher temperatures, which lead to a faster reduction, within seconds.^15,16 Since such high temperatures are not achieved during DRI processes, the resulting reaction mechanisms differ and include thermal decomposition next to gas reduction.¹⁵ Moreover, these processes are more flexible in their input material, leading to a more sustainable ironmaking by eliminating the material pre-treatment steps.¹¹ This review explores the differences between the two methods, covering H₂ adaptation in both DRI and suspension smelting processes.¹⁴ Next, the two reaction mechanisms in suspension smelting, thermal decomposition and gas reduction, are discussed in terms of thermodynamics and kinetics and compared to the more established DRI technology.

Hydrogen in DRI processes

DRI in shaft furnace

Direct reduction refers to the one-step conversion of iron ore to solid metallic iron using gas-based or coal-based reduction technologies.⁶ In direct reduction ironmaking, gas-based reduction is responsible for over 90% of the annual DRI output.^17,18 DRI is primarily used as feedstock for the EAFs or as a clean alternative to scrap iron in the BOF.¹⁹ The most established DRI technology is the MIDREX process, operating in the shaft furnace. Around 60% of global DRI production can be assigned to it. It uses pellets or lump ore as input material and reformed natural gas (primarily hydrogen and carbon monoxide) as reducing agents. The raw material is charged into the furnace from the top, while the reducing gas is blown in at the middle. The exhaust gas leaves the shaft furnace at the top and is cleaned, cooled and reintroduced into the process. To create the reducing gases, the exhaust gas (containing H₂O and CO₂) is pressurised, mixed with natural gas, heated and passed through a reformer to produce CO and H₂. After reduction, the DRI is discharged at the bottom of the shaft furnace.¹⁹ The reducing gas composition in the MIDREX process contains 24–36% CO and 40–60% H₂, with N₂ as the remaining part.¹⁷ To decrease emissions even further, natural gas can be mixed with pre-heated hydrogen before the reducing gas mixture is injected into the shaft furnace. This can decrease the use of natural gas in a MIDREX plant by up to one-third.²⁰ To achieve a complete transition from natural gas to hydrogen, ArcelorMittal is investigating a direct reduction demonstration plant based on the MIDREX process in Hamburg, Germany, in the framework of the H₂ project.²¹ The plant in Hamburg currently operates with natural gas. The objective is to utilise 100% hydrogen as reductant with a production of 100,000 t_DRI/a by the end of 2025.^21,22 Another commercial shaft furnace-based DRI process is HYL, operating with natural gas or coal-based (Energiron HYL).²³ An advantage of HYL is the possibility to recycle its reducing gases and recharge some H₂ and CO back into the DR plant.²⁴ However, the production share of HYL is currently lower compared to MIDREX and accounts for approximately 15% of DRI output.²⁵

Figure 1 shows a general illustration of a possible decarbonised steelmaking route by hydrogen-based direct reduction (H-DRI) compared to natural gas in a classic MIDREX furnace, with the resulting DRI converted to steel together with scrap in an EAF. In this illustration, emissions are further reduced with the hydrogen supply to DRI generated based on electrolysis with renewable electricity. The hydrogen and iron ore are fed to the DRI process, resulting in either DRI or being made into hot briquetted iron (HBI) as output. The resulting product can then be fed into the EAF (with the option of adding more steel scrap), resulting in crude steel as a final product.

Figure 1.

Decarbonization of steelmaking based on H-DRI and EAF route, adapted from¹⁹ with permission from Kobelco and from²⁶.

DRI in a fluidised bed reactor

In the framework of gas–solid reduction processes based on hydrogen as the reducing gas, fluidised bed (FB) technology is an alternative to shaft furnaces. The reaction takes place in fluidised bed reactor chambers, where fine iron ores and reducing gases are continuously injected and brought into contact with each other.¹¹ The use of fine ores as input material leads to further reduction in the environmental footprint since no prior agglomeration of the material is necessary, eliminating the CO₂-emissions arising in these pre-processing step.²⁷ In general, fluidised bed technology can achieve higher metallisation (∼95%) compared to the shaft furnace (∼90%) due to less sticking problems of the fine ores compared to the pellets in the shaft furnace.¹¹ However, the application of fluidised bed technologies also brings disadvantages, leading to less market dominance than the shaft furnace-based processes. First, fine ores have a wide distribution of particle sizes, meaning that not all particles can be fluidised equally during the process. A second drawback is the trade-off between low and high temperatures during the process. While a high temperature leads to kinetic advantages and counteracts the endothermic reaction behaviour of hydrogen-based reduction, a lower temperature would prevent sticking of the fine ores during the process.⁵ Industrial processes using fluidised bed technology are FINMET and the Metso Outotec CIRCORED. In contrast to the FINMET process, the CIRCORED process uses pure hydrogen as a reducing gas.²⁷ The CIRCORED process is based on a two-stage reduction process. First, the ores are treated in a circulating fluidised-bed (CFB) reactor. The CFB reactor can be defined as a pre-reduction stage, where 70% metallisation is reached. Since the reactions taking place in the CFB reactor are endothermic, energy is supplied by pre-heating the fine ores and through the combustion of natural gas and air. Processing of fine ores at high temperatures can lead to the sticking of the material. To prevent this, the CFB reactor only operates at 630 ˚C. Next, the material from the CFB is supplied to the second stage, the fluidised bed reactor. Here, final reduction takes place to achieve metallisation rates of 92–93% at temperatures of 850–900 °C. The off-gas from the second-stage fluidised-bed reactor is reintroduced into the CFB reactor as a secondary process gas. Briquetting requires a material temperature of approximately 700 ˚C. Therefore, the reduced iron is supplied to a flash heater before briquetting.^13,14,28 If pure hydrogen is used, the final product of the CIRCORED process shows no carbon contents.¹³

Smelting reduction ironmaking

Even though the dominant trend in Europe is moving in the direction of establishing H-DRI processes for ironmaking,^2,29 smelting reduction is another alternative that can reduce the carbon footprint of iron production. Smelting reduction currently accounts for 7.5–7.8 Mt iron per year,¹⁷ a much lower number compared to the 1354 Mt (2021) produced by the blast furnace and 114 Mt (2021) produced by DRI technology.¹⁴ The COREX and FINEX processes (illustrated in Figure 2) are the only commercially available smelting reduction installations. In general, smelting reduction combines an iron ore pre-reduction step, followed by smelting in a reactor without coke usage. The product is liquid pig iron, which is comparable to the hot metal exiting the blast furnace.⁶ The initial pre-reduction stage follows the DRI principle and can be performed in a shaft furnace (Figure 2(a)) or fluidised bed (Figure 2(b)), at comparable low temperatures (800–900 °C). However, it is now followed by a melting stage (Figure 2(c)) operating at a much higher temperature of 1000–1800 °C, depending on the zones of the melting reactor.³⁰ Another difference from the DRI processes is the ratio of CO and H₂ in the reducing gases. In smelting reduction, the hydrogen content is usually much lower.¹⁷ The hydrogen content in COREX is 20–25%¹⁴ and therefore relatively lower than the 40–60% in MIDREX.¹⁷

Figure 2.

Simplified presentation of a COREX and FINEX process, adapted from¹⁴ with permission from Elsevier.

The COREX process is commercially the most common form of smelting reduction.⁶ A commercial example of COREX implementation is JSW (Jindal South West) Steel in India, implementing the world's biggest COREX-based DRI plant (2014).³¹ COREX works in two stages. In the first step, the reduction from iron ore to sponge iron takes place. This pre-reduction works according to the shaft furnace principle with CO and H₂ as reducing gases. The reducing gases are obtained by the gasification of coal with a resulting composition containing 65–70% CO, 20–25% H₂ and 2–4% CO₂. The reduced iron ore is then melted during the second step in a melter-gasifier. The resulting heat, from the coal combustion, is used to melt the reduced ore, which drops from the shaft furnace to the melter-gasifier. The hot gas then leaves the melter-gasifier and is mixed with a cooling gas to reach a temperature of 850 ˚C. After purification of the gas, it is fed back to the shaft furnace as a reducing gas.⁶

The FINEX process can be considered as a variant of the COREX process, where the first commercial plant of POSCO (Pohang Iron and Steel Company) in Pohang started operation in 2007. The main difference is that the FINEX process uses fine ore, while COREX uses lump ore and pellets. Furthermore, the reduction process is based on the fluidised bed principle instead of the shaft furnace.⁶ Generally, FINEX operates as a four-unit process with a multi-stage fluidised bed reactor as pre-reduction steps, hot-compacting, coal-briquetting and a melter-gasifier needed for the final reduction.³² Compared to the shaft furnace approach of the COREX process, the fluidised bed approach leads to a better contact between the reducing gases and the material. The more uniform gas flow and ore distribution lead to more even reduction degrees of the material compared to the pellets in the shaft furnace.³² FINEX allows to operate with low-cost fine ores with a higher content of iron oxide, thereby reducing costs and emissions for the input material and generating gases which can be reused in power generation and metallurgical applications.³³ To further reduce CO₂-emissions, POSCO plans to continue the transition from coal to H₂ as reducing agent in its ironmaking processes by 2050.³³ The newly planned HyREX approach, based on FINEX technology, will help POSCO achieve carbon neutrality. In this new approach, DRI will be produced in fluidised bed reactors with 100% H₂ (compared to the current 25% H₂) before being processed in an EAF to molten iron.^34,35

Another commercially available smelting reduction technology is HIsmelt, developed by Rio Tinto. HIsmelt is a single-reactor smelting reduction process, currently operating in China (Molong plant).^14,36 Before being injected into the smelting bath, fine ores can be pre-heated and partially reduced, for example, by means of a two-stage circulating fluidised bed reactor.^37,38 In the HIsmelt reactor, reduction and melting of pre-reduced fine iron ores occur simultaneously.^39,40 The fine ores, fluxes, and coal are injected directly into a molten metal bath (∼1673 K), which is covered by a layer of molten slag. The injection is performed via water-cooled lances, while a hot oxygen-enriched air blast post-combusts the off-gases generated during iron oxide reduction.³⁶ The submerged injection of the solid reagents further induces stirring, causing slag droplets to erupt into the upper regions of the reactor, promoting heat transfer from the post-combustion zone into the liquid slag/metal zone.^36,39

At the Molong plant (in operation since 2016), rotary kilns are used for the pre-treatment of the ores. The first kiln dries the hematite ores, while the second heats them up to around 1073 K using pulverised coal combustion.³⁶ An advantage of HIsmelt over the blast furnace is its ability to process ores with high phosphorus content without significantly increasing the phosphorus content in the liquid iron, due to the more oxidising conditions in the reactor.³⁹

From a sustainability perspective, HIsmelt theoretically allows partial substitution of coal with H₂.^37,41 However, no industrial implementation exists to date.³⁷ Other potential improvements include the use of sustainable biomass (converted to biochar) to replace coal, enrichment of H₂ during the ore pre-reduction, or CO₂-capture from off-gases.⁴¹

Suspension smelting ironmaking

Two new approaches currently under development utilise a method known as suspension smelting ironmaking. The reduction part of the process takes place in suspension, using flash or cyclone processes, leading to the reduction of iron ore fines at high temperatures and short particle residence times (seconds).⁴² One such process under development is the flash ironmaking technology (FIT), which is investigated at the University of Utah.^43–48 In the FIT, iron ore concentrates are reduced in a flash reactor using reducing gases such as hydrogen, natural gas, bio/coal gas or their mixtures. In the case of natural gas and pure hydrogen, the injected mixtures are H₂ + CO + H₂O + CO₂ and H₂ + H₂O, respectively. This technology could be scaled up to industrial applications with the prerequisite that the concentrate is smaller than 100 µm.¹⁶ In terms of energy efficiency and emission savings, the advantages of this technology arise from the input material, since no agglomeration or coking steps are necessary as pre-treatment. Another advantage is the very short residence times (only seconds) needed for reduction at the operating temperatures of 1473–1873 K.¹⁶ Total energy savings using FIT are expected to be around 38% compared to the traditional blast furnace route.⁴⁸ An additional advantage is the combination of the two steps previously discussed in smelting reduction.⁴⁴ The process could be established in an overall continuous direct steelmaking route, leading to products in either molten- or solid state for EAF steelmaking.¹⁶

The development of FIT started in 2005, primarily at the University of Utah with support from the US Department of Energy.⁴⁸ The article by Sohn, Elzohiery and Fan⁴⁹ provides a comprehensive overview of the research progress achieved within FIT-related projects. Early studies focused on understanding reaction kinetics under FIT conditions and on formulating reduction rate equations in the temperature range of 1423–1873 K.^50–54 These investigations demonstrated that high metallisation degrees of magnetite fine ores could be achieved within seconds. Building on the kinetic analysis, studies were subsequently conducted utilising a laboratory flash reactor operating at temperatures between 1423 and 1473 K.^55–57 Applying H₂–H₂O mixtures, reduction degrees exceeding 90% were obtained. In contrast, experiments employing post-combustion of methane with pure oxygen generated reducing gas mixtures of H₂ + CO + H₂O + CO₂, but resulted in lower overall reduction degrees (approx. 80%).

In a subsequent step, upscaling of the process to pilot scale was investigated. The pilot flash reactor, with a reactor volume of 1.01 m³, was operated at temperatures between 1473 and 1823 K and material feeding rates ranging from 1 to 7 kg/h. The pilot-scale experiments demonstrated good reproducibility, and complete reduction was achieved at temperatures between 1573 and 1673 K using H₂ with an excessive driving force of 1.51, a fine ore feeding rate of 5 kg/h, and a nominal particle residence time of 14.6 s. Overall, these results indicated the technical feasibility of up-scaling FIT towards industrial iron production.^49,58 To further support the scale-up efforts, a CFD (Computational Fluid Dynamics) model of the laboratory flash reactor was extended and refined based on the experimental findings obtained from the pilot-scale reactor.⁵⁹ Despite FIT being regarded as a promising technology for next-generation, low-emission ironmaking, challenges remain with respect to large-scale implementation. To the author's knowledge, no additional upscaling efforts beyond the pilot-scale investigations reported in the literature have been made public.

Another process utilising the principle of suspension smelting is the HIsarna^® process. HIsarna was developed in the framework of the ULCOS (ultra-low-CO₂-steelmaking) initiative.¹⁴ The initiative (2004) aimed to reduce 50% of CO₂ emissions per ton of steel in a production based on iron ores as input material.⁶⁰ As part of the initiative, a pilot plant was built at Tata Steel in IJmuiden (Netherlands), with construction completed in 2010. The hot metal capacity of the pilot plant is approx. 65 ktpa (∼8 t/h).⁶¹ HIsarna combines the two process steps of a cyclone converter furnace (CCF) and a smelting reduction vessel (SRV) into a single furnace (Figure 3), operating in a counter-current principle, where the injected iron ores are in contact with the hot processing gases.⁶² Overall, HIsarna can be categorised as smelting reduction technology, due to final melting in the SRV, which is based on HIsmelt technology.¹⁴ However, the mechanism occurring in the CCF (top part) of the furnace can be compared to suspension smelting. HIsarna operates by injecting the iron ores at the top of the CCF section of the furnace, together with oxygen and potentially also fluxes.⁶³ The upcoming off-gases from the SRV containing pre-heated H₂ and CO gases result in the pre-reduction and even melting of the injected ores, since temperatures in the CCF can reach up to 1773 K.⁶³ Due to the high temperatures, two reaction mechanisms can influence the reduction of the injected ores: thermal decomposition and gas reduction.¹⁵ The gas reduction ability of the SRV off-gas is described by the post-combustion ratio (PCR) that gives the percentage of the total CO and H₂ that was converted to CO₂ and H₂O, as in equation (4.1)⁶²:

PCR = \frac{% C O_{2} + % H_{2} O}{% CO + % C O_{2} + % H_{2} + % H_{2} O} \times 100

(4.1)

Figure 3.

Schematic of HIsarna process with (a) the CCF top and (b) the SRV bottom section shown separately, adapted from⁶³, CC BY 4.0 license.

When the off-gases from the SRV enter the CCF, the gas composition is characterised by a PCR of around 60% and usually contains more CO (∼32%) than H₂ (3.4%).⁶³ While reacting with the injected ores, the off-gas is further oxidised, reaching PCR values of around 98% at the end of the CCF, consisting of mainly CO₂ and H₂O and only marginal amounts of reducing gases.⁶³ In the CCF section, reduced ore exhibits pre-reduction degrees of 20–25%, which indicates FeO as the main product. Following the pre-reduction, the molten droplets directly fall into the SRV section, where the final reduction to liquid iron is achieved by the injection of coal into the molten bath.⁶⁴

Hot commissioning and the first pilot campaign (campaign A) were carried out in 2011. Between 2011 and 2014, a total of four pilot campaigns (A–D) were conducted, each consisting of multiple production runs over several months. The main objective was the verification of the technical feasibility and stable operation of HIsarna. These campaigns further revealed high flexibility in raw materials quality (including steel scrap, dust and sludge utilisation) and supplied the first hot metal to a BOF plant.⁶¹ Campaigns B and C specifically aimed to improve the production rate compared to campaign A, to increase the run length, and optimise key parameters such as the PCR, oxygen distribution in the reactor and coal injection. The increase of PCRs from approx. 80% to 90–95% was achieved and led to higher production rates (up to 6 t_{hot metal}/h) and reduced specific coal consumption. During these campaigns, the influence of ore particle size and dust losses was quantified and stable hot metal chemistries with very low Si (<0.02%), Mn (<0.02%), and P (0.015–0.03%) contents were demonstrated.⁶⁵ After the conclusion of campaign D, the main objective was to demonstrate the industrial feasibility of the process, and therefore demonstrate strong CO₂ reduction potential at long runs with stable operation. To achieve this, the pilot plant underwent major upgrades between 2014 and 2017. These changes were, for example, the off-gas handling system, addition of raw materials handling and preparation area, and a scrap handling system.⁶¹

In terms of environmental footprint, HIsarna provides the same advantages as FIT due to the direct production of liquid hot metal from iron ores without pre-treatment of the coke (coking) or ore agglomeration.³⁹ In combination with CCS technology, HIsarna can save up to 80% of CO₂-emissions compared to the traditional blast furnace ironmaking.¹⁴ In order to further improve HIsarna from an environmental perspective, it is possible to partly substitute the injected coal with biomass, natural gas or hydrogen.⁶⁶ Earlier studies indicated the potential to replace around 40% of fossil coal with the more sustainable charcoal. Even though complete removal of the solid carbon source is not possible for HIsarna to guarantee carburisation of the hot metal, injection of gaseous reducing agents (CH₄, H₂) could further improve the process.⁶⁶ Another advantage of HIsarna is the flexibility in the input material, since it is also possible to use low-grade ores and secondary materials from the steel production.^62,66 The first pilot plant trials utilising scrap and biomass-derived charcoal were conducted in 2017–2018, as part of the H2020 LoCO₂Fe project, demonstrating low CO₂ potential (55% lower than HIsarna base case). Stable operation with up to 37% hot metal produced from scrap and up to 42% renewable carbon input was achieved.⁶¹

More recently (2022), accumulated pilot data have been used to develop fundamental process models for HIsarna, describing the chemical, physical and thermodynamic phenomena of the process. This model is used to support further up-scaling of the process by combining theory with plant data. The model divides HIsarna into two parts, consisting of the ‘main process’ and the ‘secondary process’. The two parts consist of seven zones. The main process includes the hot metal (zone 1), slag (zone 2) and evolved gas (zone 3), while the secondary process includes metal splash, char and volatiles above the slag layer (zone 4), metal droplets in the slag layer (zone 5), the surplus carbon in the metal (zone 6), as well as the process gases from both the smelting and pre-reduction section (zone 7).⁶⁷ In 2025, following the successful pilot-scale operations and trials in IJmuiden, Tata Steel has approved the initiation of engineering and regulatory activities for a 1 million t/a HIsarna demonstration plant in Jamshedpur (India), signalling its strategic intent to advance the low-carbon technology towards industrial development.⁶⁸

Another process worth mentioning in the framework of flash ironmaking technologies, utilising H₂, is ZESTY (Zero Emissions Steel Technology) (Figure 4), which is currently under development and tested in pilot plant scale by Calix in Australia.^69,70 Due to its lower reaction temperatures of 950–1100 °C compared to HIsarna and the FIT, the ZESTY process creates a DRI powder product, even though it follows more the principle of the flash ironmaking technology.^70,71 Therefore, ZESTY doesn’t operate in suspension, but in the temperature range of classic DRI processes. Pilot testing has been done with a large range of particle size distributions (<500 µm) of several Australian iron ore samples, where most were mixtures of hematite and goethite, but also ores with magnetite as the iron-bearing phase.^69–71 The iron ores are fed from the top using a screw feeder and are reduced over the length of the furnace (18 m) by H₂, flowing in counter-current contact with the ores. The product is water-cooled to stop the following reactions.⁷⁰ With approximate residence times of 60 s, the reduction is longer compared to the FIT and pre-reduction in HIsarna, but significantly shorter than the reduction in a MIDREX shaft furnace, which can take several hours.⁷⁰ In pilot testing, ZESTY was able to achieve high metallisation degrees of 98%. The tested magnetite ores achieved lower metallisation compared to the hematite/goethite mixtures, explained by lower surface area and porosity.⁶⁹

Figure 4.

Schematic of the ZESTY process, adapted from^70,71, CC BY 4.0 license and public report.

Comparing HIsarna and the FIT to other smelting reduction processes like COREX and FINEX, the main difference is the pre-reduction of the ores. While in COREX and FINEX, pre-reduction is based on the DRI principle, HIsarna and FIT rely on suspension smelting technology, in which the ores are pre-reduced at significantly higher temperatures and shorter residence time, making use of the accelerated kinetics of gas reduction and thermal decomposition at high temperatures.¹⁵ The following sections discuss in detail the two reaction mechanisms, namely, thermal decomposition and gas reduction, responsible for reduction in suspension smelting and compare them to the reduction mechanisms in DRI processes.

Reaction mechanisms in suspension smelting

Due to the much higher temperatures in suspension smelting ironmaking, thermal decomposition of the iron ore plays a crucial role in addition to gas reduction. In considering the thermal decomposition, it is important to account for the presence of multiple iron-bearing phases in the ore. Some interesting ores to consider from an industrial perspective are magnetite (Fe₃O₄), hematite (Fe₂O₃) and goethite (FeO*OH), while many others are mined as well.¹⁴ The main reactions of interest are the dehydration of the goethite phase, as well as oxygen removal from the hematite phase, which are discussed separately.

Thermal decomposition of goethite

At relatively low temperatures, goethite transforms into hematite under the release of water–vapour, following (reaction 5.1)⁷²:

2 α FeOOH \to α F e_{2} O_{3} + H_{2} O

(reaction \;5.1 )

The temperature range of goethite thermal decomposition has been studied by Walter et al.⁷³ for several goethite ores with varying particle sizes and specific surface areas. They determined that the initiation temperature for dehydration varies between 465 and 523 K and found a single or two-stage decomposition process, depending on ore properties (specific surface area and particle geometry). The peak temperature varied between 545 and 551 K for the first stage and between 578 and 595 K for the second stage. Ammasi⁷⁴ investigated the decomposition for several heating rates and observed similar initiation temperatures, with complete conversion occurring between 593 and 623 K (higher temperature corresponding to higher heating rates).

Several experimental studies^75–78 have been conducted under non-isothermal conditions to determine the kinetic parameters and mechanisms of the goethite dehydration, with the results of selected studies presented in Table 2. Przepiera and Przepiera⁷⁷ described goethite decomposition based on (reaction 5.2), in which the removal of water first creates an amorphous hematite phase before forming crystalline hematite at increasing temperature.

α - FeOOH \to {Fe}_{2} O_{3} (amorphous phase) \to α - {Fe}_{2} O_{3} (hematite)

(reaction \;5.2 )

Table 2.

Kinetic findings of goethite thermal decomposition.

Reference	Temperature (K)	Atmosphere	Heating rate (K/min)	Method	Mechanisms	Activation energy (kJ/mol)
López et al.⁷⁵	RT-1723	Ar	5–30	Iso-conversional	unclear	105.44–124.97
Leerhoff et al.⁷⁶	RT-1773	Ar	12,5	Coats-Redfern	Chemical reaction Diffusion	46.55–134.47 7.79–44.76
Przepiera et al.⁷⁷	RT-800	Air	3	Coats-Redfern	Nucleation (Avrami-Erofeev)	Dehydroxylation: 22.17 Hematite transformation: 27.32
Grygar et al.⁷⁸	453–533	Air	80	GRO kinetic law	Chemical reaction Diffusion	123–132
Beuria et al.⁷⁹	673–1473	Air	-	Model fitting	Random nucleation	58.74–65.74

They determined that the overall decomposition reaction is controlled by nucleation. Grygar et al.⁷⁸ and Leerhoff et al.⁷⁶ also reported a two-stage decomposition of goethite but proposed separate mechanisms. Both identified chemical reaction control in the first stage and diffusion in the second. Other studies, such as Beuria et al.⁷⁹ suggest that the overall reaction is governed by nucleation in isothermal conditions. While the studies which report nucleation as a dominant mechanism^77,79 indicate lower activation energies, the values of Beuria et al. are comparable to Leerhoff et al. for ores with lower goethite content during the reaction-controlled stage. In contrast, Grygar et al. reported significantly higher activation energies across both stages. Leerhoff et al. further found that an increasing goethite content in the ore raises the activation energy for the chemical reaction controlled, but reduces it for the diffusion-controlled stage. These results indicate that the amount of goethite in the ore does not have an influence on the decomposition mechanism, but on the activation energy needed to initiate the stage-dependent reaction.

Strezov et al.⁸⁰ studied the influence of goethite transformation to hematite on the structural changes within the ore particle. Their proposed mechanism, based on initial findings from Ariyama et al.⁸¹ is presented in Figure 5. They observed diameter increase upon heating, with a volume expansion of 110% at 673 K and 112.5% at 1173 K (measured with a custom-designed spring-loaded linear velocity displacement transducer). This was followed by cracking of the particle, which they assigned to the formation of water–vapour.

Figure 5.

Mechanism of goethite decomposition through heating proposed by and adapted from Strezov et al.⁸⁰ with permission from Elsevier and Ariyama et al.⁸¹ with permission from the author.

They found that most strain occurs at the particle surface, reaching up to 20% of its depth. They further found the influence of dehydration on the particle size of the goethite particle to be dependent on the initial particle size. For a particle size between 1.5 and 10 µm, the particle size decreased and for a particle size of 40–100 µm the particle size increased. This indicates that the expansion and not the fragmentation to finer fractions was the dominant effect. A change in material properties after heating was also observed by Li et al.,⁸² who studied the behaviour of fluxes and goethite ore during sintering. During the heating stage, crack formation on the goethite ore was observed once more, which led to a more porous structure. This change in the particle structure enhanced the diffusion of the fluxes during the process. Murakami et al.⁸³ found a link between the water content in the ore and the affected surface area during heating. A higher amount of water led to a higher specific surface area at a heating temperature of 673 K. Upon further heating the goethite to 1073 K, the specific surface area still increased compared to the raw material but decreased compared to the values measured after heating to 673 K, which they explained by the formation of larger pores at higher temperatures as also proposed in the study of Hirokawa et al.⁸⁴

Thermal decomposition of hematite

Iron can be present in different oxide states, namely hematite (Fe₂O₃), magnetite (Fe₃O₄) and wustite (FeO), depending on the applied temperature and oxygen potential. Besides their composition, these iron oxides differ in their thermal stability. Magnetite exhibits the highest melting point (1870 K), followed by hematite (1838 K) and wustite (1644 K).⁸⁵ From a thermodynamic viewpoint, the thermal decomposition and gas reduction of the hematite phase describe the same fundamental principle: the removal of oxygen from the iron oxide. The main reactions of hematite thermal decomposition are shown in (reaction 5.3) and (reaction 5.4)⁸⁶:

6 F e_{2} O_{3} (s) = 4 F e_{3} O_{4} (s) + O_{2} (g)

(reaction \;5.3 )

2 F e_{3} O_{4} (s) = 6 FeO (l) + O_{2} (g)

(reaction \;5.4 )

Figure 6 shows calculations for the thermal decomposition of 100 g pure hematite in an argon atmosphere, calculated with the equilib module of FactSage 7.0.⁸⁷

Figure 6.

Thermodynamic equilibrium of thermal decomposition of Fe₂O₃ in Ar atmosphere, calculated with the equilib module of FactSage 7.0⁸⁷ (Database FactPS⁸⁸).

Based on these thermodynamic calculations in Figure 6, the first significant thermal decomposition of hematite to magnetite starts at 1200 K, even though marginal amounts can be detected before. At 1367 K, all hematite is converted to magnetite. At temperatures present during suspension smelting, only minor amounts of wustite can be created based on thermal decomposition. The thermodynamic calculations indicate that, for example, at 1800 K, 0.03 g of wustite is released into the gas phase, indicating that (reaction 5.3) is the main reaction of thermal decomposition.

To determine the decomposition temperature of hematite in the ores in practice, TGA (thermogravimetric analysis) measurements have been widely used in the literature. Chen et al.⁸⁹ reported that the hematite within a low-grade ore decomposes via a two stage process; the first stage occurs between 1423 and 1593 K, while the second stage takes place at temperatures above 1593 K. Similarly, Qu et al.⁸⁵ determined the thermal decomposition temperature of hematite in higher grade ores (94.94 wt.% Fe₂O₃) to be between 1473 and 1573 K. Leerhoff et al.⁷⁶ studied the decomposition behaviour of varying low-grade iron ores and also identified a two stage process. Depending on the mineralogy and heating rate, the first stage occurred between 1312 and 1419 K, and the second stage at temperatures above 1676 K. However, the influence of mineralogy on the decomposition temperature was only marginal and mainly influenced by the applied heating rate. Xing et al.⁸⁶ examined the effect of particle size on the decomposition temperature of hematite ores and observed no significant influence. The decomposition range they reported was between 1480 and 1532 K, depending on the heating rate and particle size. Generally, it is clear that the initiation temperatures of the thermal decomposition determined with the TGA are significantly higher compared to the theoretical thermodynamic calculations shown in Figure 6, indicating kinetic barriers to thermal decomposition.

Several studies^76,86,89,90 were carried out to determine the reaction mechanisms of hematite thermal decomposition. The main findings from selected studies are presented in Table 3. These kinetic mechanisms are usually determined under non-isothermal conditions using several heating rates, with kinetic parameters commonly determined via model-free iso-conversional methods, like the Ozawa Flynn and Wall method⁹¹ or the model-fitting Coats-Redfern method.⁹² Overall, there seems to be a consensus that the hematite thermal decomposition is controlled by chemical reaction. Only Chen et al.⁸⁹ defined different reaction mechanisms, consisting of inner diffusion of iron at the first stage, followed by slag diffusion. The reported activation energy varies significantly across studies. Salmani et al.⁹⁰ reported comparably lower values for the main decomposition stage, while Chen et al.⁸⁹ observed higher values, particularly in the first stage. Leerhoff et al.⁷⁶ found a wide range of activation energies, influenced by ore mineralogy, and Xing et al.⁸⁶ reported higher values than those observed in other studies. These trends highlight a strong dependence of the activation energy on the applied method, but also on material properties.

Table 3.

Kinetic findings of hematite thermal decomposition.

Reference	Temperature (K)	Atmosphere	Heating rate (K/min)	Method	Mechanisms	Activation energy (kJ/mol)
Chen et al.⁸⁹	RT-1773	N₂	2, 3, 5, 10	Iso-conversional	1423–1593K: Inner diffusion of iron in solid phase >1593K: Diffusion of iron in slag phase	1423–1593K: 636 ± 75 >1593K: 327 ± 41
Leerhoff et al.⁷⁶	RT-1773	Ar	1, 2, 5	Coats-Redfern	Chemical Reaction	424–832
Xing et al.⁸⁶	RT-1773	N₂	10, 15, 20	Coats-Redfern	Chemical Reaction	1256
Salmani et al.⁹⁰	RT-1723	Air, N₂	5, 7.5, 10	Iso-conversional	Chemical Reaction	Air: 382 N₂: 324

Gas reduction of iron oxides

During the pre-reduction stage of smelting reduction and DRI processes, iron ores are reduced via gas–solid reactions with H₂ and/or CO. Below 843 K, only hematite and magnetite are stable, and reduction proceeds in two steps via the disproportionation reactions (5.5, 5.6, 5.9, 5.10), producing magnetite and metallic iron. Above 843 K, wustite becomes more thermodynamically stable, and the reduction shifts to a three-step process involving wustite. This temperature marks the threshold for a change in the dominant reduction pathway.^6,93 The chemical reactions of iron oxides with H₂ and CO are presented in Table 4 (reaction 5.5)–(reaction 5.12) with their reaction enthalpy given at room temperature (25 ˚C)^94,95:

Table 4.

Step-wise reduction reactions of iron oxide with H₂ or CO.^6,93–95

	H₂ reduction
	$3 F e_{2} O_{3} + H_{2} = 2 F e_{3} O_{4} + H_{2} O$	$Δ H = - 11.72 kJ / mol$	(reaction 5.5)
T < 843 K	${Fe}_{3} O_{4} + 4 H_{2} = 3 Fe + 4 H_{2} O$	$Δ H = 148 .96 kJ / mol$	(reaction 5.6)
T > 843 K	${Fe}_{3} O_{4} + H_{2} = 3 FeO + H_{2} O$	$Δ H = 77 .41 kJ / mol$	(reaction 5.7)
	$FeO + H_{2} = Fe + H_{2} O$	$Δ H = 23 .85 kJ / mol$	(reaction 5.8)

	CO reduction
	$3 F e_{2} O_{3} + CO = 2 F e_{3} O_{4} + {CO}_{2}$	$Δ H = - 52. 87 kJ / mol$	(reaction 5.9)
T < 843 K	${Fe}_{3} O_{4} + 4 CO = 3 Fe + 4 C O_{2}$	$Δ H = - 15.66 kJ / mol$	(reaction 5.10)
T > 843 K	${Fe}_{3} O_{4} + CO = 3 FeO + {CO}_{2}$	$Δ H = 36 .25 kJ / mol$	(reaction 5.11)
	$FeO + CO = Fe + {CO}_{2}$	$Δ H = - 17.31 kJ / mol$	(reaction 5.12)

Examining H₂-based reductions (reaction 5.5)–(reaction 5.8), only the reduction from hematite to magnetite is exothermic. The following steps of the reduction are endothermic. This indicates that additional energy has to be supplied to the system after the pre-reduction of hematite to magnetite. For CO-based reduction, the reduction of magnetite to wustite is endothermic. All other reduction steps of the iron oxides are exothermic and, therefore, release energy to the system. Hence, CO reduction has a better influence on the overall energy balance of the system.

Figure 7(a) represents the classic Bauer-Glaessner diagram, where the influence of the reducing gas partial pressure is displayed for the ratio of CO to the total gas volume of CO + CO₂ and similar for the case of H₂ and H₂O. The main takeaway from the diagram is that CO is a better reducing agent at lower temperatures, while H₂ is preferred at higher temperatures, from a thermodynamic perspective. This indicates that under the DRI conditions, H₂ is better suited as the reducing agent. Figure 7(b) shows a broader temperature range, which is more relevant for the suspension smelting processes. The gas composition displays the post-combustion ratio, present in the cyclone converter of the HIsarna furnace (equation 4.1) and, therefore, considers the presence of both reducing gases H₂ and CO and the resulting gases H₂O and CO₂. A pre-reduction degree of 33.33% would indicate the reduction from hematite to wustite.⁸⁵ Therefore, the figure highlights that wustite is the thermodynamic limit of pre-reduction in the HIsarna furnace, where a decrease in the PCR would increase the pre-reduction degree (PRD). During the flash ironmaking, which operates at higher amounts of reducing gases, the reduction to metallic iron in solid and liquid form is theoretically possible.

Figure 7.

Bauer–Glaessner Diagram for (a) the general reduction of hematite with CO and H₂, adapted from¹⁴ with permission from Elsevier and (b) the Bauer-Glaessner diagram with special focus on the conditions during the cyclone pre-reduction in HIsarna ironmaking, adapted from³⁹ with permission from Springer Nature.

Reaction kinetics of hematite reduction with H₂ and/or CO

Spreitzer et al.⁹⁶ and Heidari et al.⁹⁷ summarised multiple parameters that influence the reduction kinetics of iron oxides and therefore their reduction rate. These parameters can be categorised into process parameters (such as temperature, pressure and gas composition) and material properties (such as particle size, morphology and porosity). Comparing differences between suspension smelting and DRI ironmaking processes, most interesting parameters to consider are the influence of the processing temperatures, material properties like particle size and mineralogy and of course the reducing gas composition. From a kinetic point of view, another important processing parameter is the reduction time. Thermodynamics indicate that reduction of iron ores is already possible at lower temperatures, but the significantly higher temperatures in suspension smelting processes have the potential of decreasing the reduction time.

A general approach to describe the kinetics of the gas–solid reduction of pellets or particles is the (un-)reacted shrinking core model.^98,99 In the shrinking core model, the first reactions take place at the outer surface of the particle. The reaction zone then progresses inward towards the core, while the reducing agents CO and H₂ diffuse through the developing product layer and towards the unreacted core. The product layer is the already converted material phase surrounding the unreacted solid iron ore core. The developing reduction products (CO₂ and H₂O) diffuse back through the product layer and into the gas film.⁹⁹ The shrinking core model considers interfacial chemical reactions, internal diffusion or a mixed mechanism of both as possible reaction mechanisms.^99–102 Generally, chemical reaction-controlled mechanisms are much more temperature dependent than diffusion-controlled mechanisms. Therefore, chemical reaction is often the rate-limiting step at lower temperatures, whereas diffusion becomes more likely to be rate-controlling at higher temperatures.⁹⁶ One disadvantage of the shrinking core model is that it assumes a reduction from the outside to the inside of the particle, disregarding the possibility of (internal) nucleation and growth being the underlying mechanism. Based on the selected model, the reaction rate constant can be determined using the Arrhenius equation (equation 5.1).¹⁰³

k = A * e^{- \frac{E_{a}}{R T}}

(equation \;5.1 )

where A is the collision factor (1/s), E_a is the activation energy (J/mol), R is the ideal gas constant (J/(mol*K) and T is the temperature (K). The influence of temperature on the reduction kinetics is generally described by its exponential dependency on the rate constant indicating an increase of the reaction rate with increasing temperature,⁹⁷ which is highly beneficial for the suspension smelting processes. A lot of research has been done on the reduction kinetics of iron ore pellets and fines with H₂ and CO, and a selection of studies focusing on the reduction under isothermal conditions are summarised in Table 5. In terms of kinetic aspects of the gas reduction, it is important to evaluate the influence of H₂ and CO on the reaction rate, the underlying reaction mechanism and the arising activation energy.

Table 5.

Overview of isothermal kinetic studies on H₂–CO reduction in ironmaking.

Reference	Input material	Reduction step	Temperature (K)	Atmosphere	Mechanism	Activation energy (kJ/mol)	Reduction degree (%)	Set-up
DRI conditions
Zhang et al.¹⁰⁴	Pellets	Fe₂O₃ → Fe	1043–1443	H₂	1043 K: interfacial reaction + diffusion>1143 K: interfacial reaction	31.3 ± 1.3	∼ 97	TGA
Sohn et al.¹⁰⁵	Pellets	Fe₂O₃ → Fe	973–1173	H₂	Initial: topohemical reactionLater: pore diffusion	15–29	∼ 100	TGA
Piotrowski et al.¹⁰⁶	Fines	Fe₂O₃ → FeO	973–1173	90%N₂ + 10%CO90%N₂ + 10%H₂90%N₂ + 5.7%CO + 4.3%H₂	Initial: topochemical reaction at the surfaceLater: diffusion (after a thin layer of lower iron oxides is formed)	122.5228.0893.72	-	TGA
Zuo et al.¹⁰⁷	Pellets	Fe₂O₃ → Fe	1073–1273	H₂-CO	Initial: chemical reactionFinal: internal gas diffusion	-	50.01–98	Reducing furnace
Moon et al.¹⁰⁸	Compacted fines	Fe₂O₃ → Fe	1073–1223	H₂-CO	Initial: chemical reaction at the interfaceLater: particle diffusion through the product layer	19.8–42.1	∼ 90	Cahn balance
Lyu et al.¹⁰⁹	Fines	Fe₂O₃ → Fe	1073–1373	30%H₂ + 70%N₂	Initial: gaseous diffusion interfacial chemical reactionLater: interfacial chemical reaction and solid diffusion	27–93	∼90–100	TGA
Hammam et al.¹¹⁰	Compacted fines	Fe₂O₃ → Fe	1073–1373	H₂	Initial: gaseous diffusion and interfacial chemical reactionLater: interfacial chemical reaction	R = 5–10%: 28.34R = 85–95%: 41.15	100	TGA
Kuila et al.¹¹¹	Fines	Fe₃O₄ → Fe	973–1273	H₂	Pore diffusion	Fe₃O₄ → Fe: 42FeO → Fe: 55	70–100	TGA
Suspension smelting conditions
Qu et al.¹¹²	Fines	Fe₂O₃ → FeO	1550–1750	42.2%CO + 57.8%CO₂	-	-	3.3–26.3	Drop tube
Qu et al.¹⁰²	Fines	Fe₂O₃ → FeO	1550–1650	42.2%CO + 57.8%CO₂	Chemical reaction	270	3.3–26.3	Drop tube
Qu et al.¹¹³	Fines	Fe₂O₃ → FeO	1700–1750	42.2%CO + 57.8%CO₂	Diffusion	156	3.3–26.3	Drop tube
Qu et al.¹¹³	Fines	Fe₂O₃ → Fe	1650	CO-CO₂-H₂-N₂	-	-	23.1–97	Drop tube
Chen et al.¹¹⁴	Fines	Fe₂O₃ → Fe	1423–1623	H₂	Nucleation and growth	214	>90	Drop tube
Chen et al.¹¹⁵	Fines	Fe₂O₃ → Fe	1423–1623	CO	Nucleation and growth	231	>90	Drop tube
Elzohiery et al.⁵²	Fines	Fe3O₄ → Fe	1423–1623	H₂	Nucleation and growth	193	>90	Drop tube

Literature reports that the presence of hydrogen gas in the reducing atmosphere offers an advantage over carbon monoxide, due to increased reduction rates.^107,108,116 This is generally attributed to the faster diffusion of hydrogen due to its smaller molecular size and lower viscosity.¹⁰⁷ Regarding the reaction mechanism of H₂ or CO reduction, literature indicates a change in the reaction mechanism throughout the reduction process from hematite to metallic iron. Sohn et al.¹⁰⁵ investigated hematite pellet reduction in pure H₂ at 973–1173 K and found a transition from a topochemical reaction in the early stages to pore diffusion in later stages. Similar observations were reported by other studies,^106–108 explained by the formation of an initial iron or wustite layer on the surface, limiting the gas diffusion through the product layer. However, some authors propose slightly different mechanisms. Kuila et al.¹¹¹ studied the reduction of magnetite fines with H₂ at temperatures between 973 and 1273 K and found that both the reduction from Fe₃O₄ to FeO and the subsequent reduction from FeO to Fe is controlled by pore diffusion. Lyu et al.¹⁰⁹ used a mixture of 30% H₂ and 70% N₂ for the reduction of hematite fines and proposed that the first stage of the reduction is already a mixed mechanism of gaseous diffusion and chemical reaction at the interface, before it shifts to the mixed mechanism of interfacial chemical reaction and solid diffusion.

The influence of H₂ and CO on the activation energy was investigated by Piotrowski et al.¹⁰⁶ The reduction of hematite to wustite was investigated for temperatures between 973 and 1173 K in reducing gas mixtures of H₂, CO and N₂. The calculated activation energies varied between 25 and 125 kJ/mol, depending on the gas composition. For the reduction with pure hydrogen, significantly lower activation energies (ΔE_a = 28.08 kJ/mol) are determined in comparison to CO-based gas mixtures (ΔE_a = 122.52 kJ/mol). For the mixture of the two gases, an activation energy of ΔE_a = 93.72 kJ/mol was determined. Conversely, Moon et al.¹⁰⁸ observed an increasing activation energy with an increasing amount of H₂ in the atmosphere.

The influence of particle size on the reduction rate of hematite fines in H₂ atmosphere has been investigated by Hou et al.¹¹⁷ (pure Fe₂O₃), Wagner et al.¹¹⁸ (pure Fe₂O₃) and Pang et al.¹¹⁹ (62.7 wt.-% T.Fe, Fe₂O₃ + Fe₃O₄). All observed a general increase in reaction rate with decreasing particle size. Hou et al.¹¹⁷ found 0.045 mm as the lower limit of the influence because the internal diffusion resistance doesn’t have an influence on the overall reduction rates at these dimensions. Pang et al.¹¹⁹ further state a co-dependency of the temperature and particle size. The influence of temperature on the reduction is stronger with decreasing particle sizes.

Even though a significant reduction with H₂ is already achieved at lower DRI processing temperatures, the advantage of higher temperatures is the acceleration of the reaction and, therefore, decrease in the needed reduction time for the pre-reduction under industrial process conditions relevant to HIsarna and Flash Ironmaking Technology. In contrast to DRI ironmaking, thermal decomposition of the hematite ore becomes a crucial part of the process. Considering the reduction of iron ore fines at elevated temperatures, significant research was conducted at Delft University of Technology^{95,102,112,113,120–122} in the framework of investigating fine ore reduction in the pre-reduction section of HIsarna and at the University of Utah^16,43,44,52 for Flash Ironmaking. Both approaches focus on a fast reduction at high temperatures. A common experimental apparatus used in these studies is the drop tube furnace, where the ore concentrates are transported through a defined reaction zone filled with the reducing gases present in laminar flow. The rough principle of a laminar flow reactor as it is used in the studies at Delft University of Technology and the University of Utah is portrayed in Figure 8. The particles are generally introduced into the hot reaction zone via a pneumatic powder feeder, with the powders being carried with an injected gas stream into the furnace, where they are gradually reduced via the laminar reducing gas flow. Next, the reduced particles are quenched to prevent reoxidation and collected at the bottom using a sample collector.^52,95,114

Figure 8.

Setup for laminar flow drop tube furnace experiments, adapted from¹²² (CC.BY.4.0) and adapted from ⁵² with permission from John Wiley and Sons.

Qu et al.^102,112,113 studied the impact of gas reduction mechanisms on the overall reduction of fine hematite ores at short particle residence times (210–2020 ms). They investigated the reaction kinetics for a gas mixture containing 42.2% CO and 57.8% CO_2, and found that the reduction was chemical reaction-controlled in the temperature range of 1550–1650 K, with an activation energy of 270 kJ/mol. With an increase in temperature (1700–1750 K), the reaction mechanisms shifted to diffusion and the activation energy decreased to 156 kJ/mol. They also found a significant influence of the temperature, with only 3.3% of reduction degree at 1550 K and a particle residence time of 210 ms and an increase to 26.3% at 1750 K and a residence time of 970 ms. Based on their observations, they were able to explain the influence of thermal decomposition and gas reduction on the pre-reduction degree in the drop tube furnace.¹¹² The results are summarised in Figure 9. Until 1700 K, thermal decomposition is the dominant reduction mechanism, with a particularly low contribution of gas reduction at lower temperatures. Gas reduction becomes the significant reduction mechanism at 1750 K. However, the influence of thermal decomposition is only significant in the first 210 ms of the reaction. Since thermal decomposition is mostly completed in this fast timeframe, gas reduction becomes more influential at increasing particle residence times.

Figure 9.

Influence of gas reduction (42.2% CO + 57.8% CO₂) and thermal decomposition (100% CO₂) on the pre-reduction degree at a particle residence time of 210 ms and a particle size of 45–53 μm, adapted from¹¹² with permission from ISIJ.

Additionally, Qu et al.¹¹³ studied the impact of hydrogen enrichment on the reduction degree in the drop tube furnace by testing 5 different atmospheres at 1650 K and a relatively long particle residence time of 2020 ms. They showed that by changing the atmosphere from 25% CO and 75% N₂ to 25% H₂ and 75% N₂, the reduction degree increased from 26.4% to 77.1%. Moreover, increasing the hydrogen content in the atmosphere to 75%, the reduction degree increased further to 97%, even at the short particle residence times.

The influence of the reducing gas atmosphere in suspension smelting processes was further studied by Chen et al.^114,115 in the framework of developing the flash ironmaking technology. They studied the reduction of hematite concentrates (d = 21 μm) in a temperature range of 1423 to 1623 K in H₂ and CO separately. Both H₂ and CO reduction followed nucleation and growth as reduction mechanisms and achieved reduction degrees above 90%. However, they observed faster reduction with H₂, achieving a similar amount of reduction in 3 s, as it took 5 s for the CO reduction. Building on the reduction of hematite concentrates, Elzohiery et al.⁵² tested the H₂ reduction of magnetite concentrates (d = 20–25, 32–38, 45–53 μm). Again, they found that the reaction completed within seconds, whereas the duration of the reduction was influenced by the temperature and particle size. All three studies determined nucleation and growth as the governing reaction mechanism, and reported the lowest activation energy for the reduction of the magnetite concentrates (193 kJ/mol), followed by the reduction of hematite concentrates with H₂ (214 kJ/mol) and CO (231 kJ/mol), respectively.

Qu et al.¹²³ investigated the impact of temperature and reduction degree on particle morphology during flash reduction (Figure 10). In the temperature range of 1450–1600 K, with corresponding reduction degrees of approximately 9.6–47.5%, the particles maintained their irregular shape. Under these temperatures, the reaction occurs as gas–solid reduction. However, at temperatures exceeding 1600 K and a reduction degree of 52.2%, the reduction mechanism transitioned to a gas–liquid reduction, leading to the formation of spherical particles, with melting mostly driven by the developing wustite phase. Initially, iron products disperse across the wustite particle before changing into a spherical form as the temperature increases.

Figure 10.

Influence of temperature on hematite particle during gas reduction in a drop tube furnace, adapted from¹²³ with permission from Elsevier.

The particle size plays a crucial role in the reduction process in suspension smelting. Here again, Qu et al.¹¹² investigated the influence of particle size on the reduction behaviour of iron ores at high temperatures (1650 K). They studied four particle size fractions (38–45, 45–53, 53–75 and 75–90 µm) in gas atmospheres consisting of CO–CO₂ and CO–CO₂–H₂–N₂ and varying contents of reducing gases. For all the tested gas compositions, the reduction degree increases approximately linearly with decreasing particle size. Chen et al.^120,121 found an influence of the particle size on the reaction rate constant for the reduction of hematite, described by a nucleation and growth mechanism of magnetite grains, within the hematite particle. The reaction rate constant increased with an increase in particle size from 67 to 85 μm. However, after a threshold of 85 μm, the reaction rate constant decreased with a further increase in particle size. An increase in particle size, however, can yield a negative influence on the heating of the particle, which has a significant influence on the thermal decomposition at the short particle residence times during suspension smelting and therefore leads to a negative effect on the reduction degree.¹²⁰

When comparing the reaction kinetics of suspension smelting studies to those of DRI-based ironmaking, a key difference is the substantially higher activation energy observed in suspension smelting. This increased energy requirement is attributed to the high degrees of reduction achieved within very short reduction times. The higher activation energy is therefore higher in suspension smelting processes. However, its significant advantage is the rapid reduction, which enables the formation of metallic iron under the appropriate atmospheric and temperature conditions.

Water–gas shift reaction

The reduction of iron ores with H₂ and CO produces H₂O and CO₂ as combustion products, both of which can participate in the water–gas shift reaction (WGSR) (reaction 5.13), wherein steam (water vapour) reacts with carbon monoxide to form carbon dioxide and hydrogen gas.¹²⁴ This reaction plays an important role in influencing the local gas composition during reduction, potentially enhancing the availability of H₂ and thereby affecting the overall reduction kinetics.

H_{2} O (g) + CO (g) ⟷ C O_{2} (g) + H_{2} (g), Δ H = - 40 .6 kJ / mol

(reaction \;5.13)

The reaction (5.13) is exothermic and reversible.¹²⁴ An important factor for the evaluation of the WGSR is its equilibrium constant, K_P, which is defined by equation (5.2):

K_{P} = \frac{P_{{CO}_{2}} * P_{H_{2}}}{P_{CO} * P_{H_{2} O}}

(5.2)

with P_i being the partial pressure of the different gaseous species. While pressure does not affect the equilibrium due to constant mol numbers,¹²⁴ it can increase the reaction rate and CO conversion before the equilibrium is reached. If large amounts of steam are present (H₂O/CO > 1), the equilibrium shifts towards product formation, increasing CO conversion.¹²⁵ Temperature strongly affects the WGSR via the equilibrium constant K_P, given in (equation 5.3) (best applicable: 588 to 753 K).^124,125

K_{P} = e^{(\frac{4577.8}{T} - 4.33)}

(5.3)

Figure 11 compares K_P values calculated via the Reactions Module in FactSage 7.0¹²⁶ and (equation 5.3) across 298–1800 K. Both show decreasing K_P with rising temperature, indicating that the forward reaction is favoured at lower temperatures. A brief plateau appears in the FactSage calculations at low temperatures but is negligible for high-temperature ironmaking. At ∼1058 K (equation 5.3) and 1101 K (FactSage), K_P becomes negative, favouring CO and H₂O formation. This shift is crucial in suspension smelting as it may lead to more endothermic conditions, requiring an external energy supply.^6,127

Figure 11.

Temperature dependence of WGSR after FactSage7.0¹²⁶ calculations and (equation 5.3).

Another issue with the WGSR at high temperatures is the development of water–vapour. Water vapour decreases the thermodynamic driving force for the reduction of the iron oxides. In addition, the water vapour can be absorbed at the reaction interfaces and thus block the reactive surfaces and inhibit reduction.⁹⁶ Abdelrahim et al.¹²⁸ studied iron ore reduction using CO–CO₂–H₂–H₂O–N₂ gas mixtures. They found that when the H₂O content exceeds 2% at 1073–1173 K, the reduction efficiency decreases. This was explained by H₂O blocking the active sites on the wustite. Lorente et al.¹²⁹ observed that up to 5% steam has little effect on Fe₂O₃ to Fe₃O₄ reduction but significantly limits the Fe₃O₄ to Fe step. Similarly, Kim et al.¹³⁰ studied magnetite reduction in H₂–H₂O mixtures (p_H2O = 0.1 atm) and observed that H₂O negatively affects the reduction rate, particularly at temperatures between 773 and 973 K. However, the influence diminished at higher temperatures.

The influence of the WGSR reaction becomes increasingly interesting due to the intense use of H₂–CO mixtures and high temperatures in flash ironmaking processes. If H₂O and CO₂ are present in the system, the occurrence of the WGSR is possible. From Figure 11, the equilibrium constant is smaller than one around 1100 K, meaning that at this temperature, the H₂/CO ratio decreases as CO formation is favoured. Gas-based DRI processes operate at these high temperatures. The MIDREX process, for example, operates with a gas feeding temperature around 1173–1223 K.⁵ Particularly in the temperature range relevant to suspension smelting ironmaking processes, where the equilibrium strongly favours the formation of H₂O and CO, it is essential to verify the actual composition of reducing gases and assess if the WGSR could reduce the available H₂ content. This can become an issue for processes like the cyclone converter of the HIsarna furnace, where large amounts of CO₂ are present in the atmosphere, together with H₂. Figure 12 shows the influence of the WGSR on a potential gas atmosphere for PCR ∼ 60% for two cases: one based on plant data from literature⁶³ and another from a theoretical calculation assuming a potential increase in H₂ content in the CCF, both calculated with the Equilib module of FactSage 7.0⁸⁷ and the FactPS database.⁸⁸

Figure 12.

Influence of WGSR reaction of PCR ∼ 60% atmospheres in HIsarna CCF based on plant data⁶³ and a potential H₂-enriched scenario.

In both cases, the equilibrium shifts towards CO and H₂O with increasing temperatures, leading to lower amounts of H₂ in the reducing gas atmosphere. At high temperatures, the total concentration of reducing and oxidising gases saturates, indicating that although the overall amount of reducing gases remains constant, the proportion of H₂ decreases. This shift could lead to a significant decline in the reduction potential of the gas mixture, which is especially crucial in the smelting reduction process where the reactions occur in a very short time.

Conclusion

The transition of the steel industry towards a more sustainable production is becoming an urgent necessity to reduce emissions and comply with the EU climate goals. Currently, the blast furnace is still the dominant technology for ironmaking, but this will change soon. New approaches with either less limitations in terms of input material or more sustainable reducing agents are being developed. In this framework, the adaptation of hydrogen in DRI processes, especially shaft furnace-based, seems to be most promising. In the current state of the art, hydrogen is already utilised in DRI ironmaking, where the transition to high hydrogen concentrations and consequently, a significant reduction in CO₂-emissions appears to be most feasible. Compared to the more established DRI ironmaking processes, this article focuses on suspension smelting technologies as a potential alternative, which are currently still under development, including the FIT and the pre-reduction stage of the HIsarna process. The FIT and HIsarna both bring an advantage by eliminating material pre-treatment steps, since fine ores can be directly used without agglomeration. The objective of this review is to understand the reaction mechanisms involved in suspension smelting ironmaking processes, examine how they differ from the more conventional DRI routes and evaluate the role of hydrogen within these processes.

According to the literature, the two primary reaction mechanisms in suspension smelting are thermal decomposition and gas–solid reduction, with a transition to gas-melt reduction at elevated temperatures. Regarding thermal decomposition, an important parameter is the nature of the iron ore. In the case of goethite being present in the ore, its conversion to hematite can influence the material properties due to volume expansion and cracking. This can potentially accelerate the gas-reduction, for example due to the change in particle size and potential reactive surfaces. The thermal decomposition of the hematite phase is mostly described by the reduction to magnetite and has a stronger influence on the reduction during the first 210 ms of suspension smelting. Most literature agrees that the thermal decomposition of hematite to magnetite is chemical reaction-controlled. The kinetic onset of thermal decomposition was determined to lie between 1312 and 1480 K, making it an insignificant mechanism in DRI ironmaking.

From a thermodynamic perspective, reduction with H₂ has the disadvantage of being largely endothermic, in contrast to the typically exothermic reduction with CO. This results in a less favourable energy balance and the need to apply additional energy input to sustain the process. In terms of thermodynamic preference, the reduction with CO is favoured at lower, while H₂ becomes advantageous at higher temperatures. However, more critical than thermodynamics is the impact of H₂ on the reduction kinetics. Overall, H₂ enables higher reduction rates compared to CO, which is explained by its better diffusivity. While various parameters influence the reaction mechanism, many studies report a mixed-controlled mechanism. The process is first controlled by chemical reaction at the pellet or particle interface and subsequently transitioned to diffusion control as a product layer of wustite or iron is formed.

The most commonly used equipment to study the kinetics of suspension smelting processes is a laminar flow drop tube furnace, which allows for studying the in-flight flash reduction of the particles under isothermal conditions without the influence of turbulence. In suspension smelting, the particles are reduced within seconds. In this rapid process, the kinetic advantage of H₂ becomes particularly relevant for achieving high reduction degrees in a very short residence time. A significant difference from DRI is the higher activation energy observed for suspension smelting, attributed to the fast reduction kinetics. In processes such as HIsarna, achieving high reduction degrees in the pre-reduction cyclone could further improve the environmental performance by reducing the amount of carbon needed for the final reduction in the smelting section. One important consideration during reduction with CO or H₂ is the influence of the WGSR in the presence of CO₂ or H₂O in the gas phase. At elevated temperatures, the equilibrium of the WGSR shifts towards greater production of CO and H₂O, which can lower the effective H₂ content in the atmosphere and diminish its kinetic advantage. Furthermore, the presence of excessive water–vapour in the system may further negatively affect the reduction performance.

In conclusion, the suspension smelting reduction process remains a relatively new technology compared to the more established DRI processes. Further research is essential to understand how material properties and gas atmosphere influence the reduction kinetics. Nevertheless, hydrogen-enriched atmospheres must be applied with caution, particularly due to the possible limitations imposed by the water–gas shift reaction at high temperatures and the endothermic nature of the hydrogen reduction.

Footnotes

ORCID iD

Philipp Leerhoff

Ethical considerations

There are no human participants in this article, and informed consent is not required.

Consent to participate

Not applicable.

Authorship contribution statement

Philipp Leerhoff: writing – review and editing, writing – original draft, visualisation, validation, software, methodology, investigation, formal analysis, data curation, conceptualization. Christiaan Zeilstra: writing – review and editing, supervision, resources, project administration, funding acquisition. Koen Meijer: writing – review and editing, supervision. Jan van der Stel: writing – review and editing, supervision. Shoshan T. Abrahami: writing – review and editing, supervision. Neslihan Dogan: writing – review and editing, supervision. Yongxiang Yang: writing – review and editing, supervision, resources, project administration, funding acquisition.

Funding

This research was carried out under project number T20009 in the framework of the Research Program of the Materials Innovation Institute (M2i) () supported by the Dutch government.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability statement

Data will be made available on request.

References

Worldsteel Association. Climate change and the production of iron and steel. Brussels, 2021.

IEA. Iron and steel technology roadmap, https://www.iea.org/reports/iron-and-steel-technology-roadmap (2020, accessed 4 January 2023).

Vehec

. Technology roadmap research program for the steel industry. American Iron and Steel Institute, 2010. DOI: 10.2172/1011081.

Worldsteel association. 2020 World steel in figures, 2020.

Battle

Srivastava

Kopfle

, et al. The direct reduction of iron. In: Treatise on process metallurgy. Elsevier Ltd., 2014, pp.89–176.

Yang

Raipala

Holappa

. Ironmaking. In: Treatise on process Metallurgy. Elsevier Ltd., 2014, pp.2–88.

Worldsteel association. Sustainable performance of steel industry 2003–2021. Brussels, December 2022.

World Steel Association. 2024 World Steel in Figures. Brussels, https://worldsteel.org/data/world-steel-in-figures-2024/ (2024, accessed 12 March 2025).

Patisson

Mirgaux

. Hydrogen ironmaking: how it works. Metals (Basel) 2020; 10: 1–15.

10.

Hydrogen Europe. Hydrogen - Enabling a Zero-Emission Society. Brussels, 2021.

11.

Roland Berger GMBH. The Future of Steelmaking - How the European Steel Industry Can Achieve Carbon Neutrality. Munich, 2020.

12.

Vogl

Åhman

Nilsson

. Assessment of hydrogen direct reduction for fossil-free steelmaking. J Clean Prod 2018; 203: 736–745.

13.

Markotic

Dolic

Trujic

. State of the direct reduction and reduction smelting processes. J Min Metall Sect B 2002; 38: 123–141.

14.

Seetharaman

Guthrie

McLean

, et al. Treatise on Process Metallurgy Volume 3: Industrial Processes. 2nd ed. Elsevier Ltd., 2024.

15.

Chen

Zeilstra

van der Stel

, et al. Review and data evaluation for high-temperature reduction of iron oxide particles in suspension. Ironmaking Steelmaking 2020; 47: 741–747.

16.

Sohn

. Energy consumption and CO2 emissions in ironmaking and development of a novel flash technology. Metals (Basel) 2020; 10: 54.

17.

Liu

Zuo

Wang

, et al. The production and application of hydrogen in steel industry. Int J Hydrogen Energy 2021; 46: 10548–10569.

18.

Ramakgala

Danha

. A review of ironmaking by direct reduction processes: quality requirements and sustainability. Procedia Manuf 2019; 35: 242–245.

19.

Atsushi

Uemura

Takashi

. MIDREX Process. Kobelco Technology Review 2010; 29: 50–57.

20.

Ripke

Kopfle

. MIDREX H2: Ultimate low CO2 Ironmaking and its place in the new Hydrogen Economy, https://www.midrex.com/tech-article/midrex-h2-ultimate-low-co2-ironmaking-and-its-place-in-the-new-hydrogen-economy/ (2017, accessed 17 March 2022).

21.

ArcelorMittal. Hamburg H2, https://corporate.arcelormittal.com/climate-action/decarbonisation-technologies/hamburg-h2-working-towards-the-production-of-zero-carbon-emissions-steel-with-hydrogen# (accessed 17 March 2022).

22.

ArcelorMittal. Hydrogen-based steelmaking to begin in Hamburg, https://corporate.arcelormittal.com/media/case-studies/hydrogen-based-steelmaking-to-begin-in-hamburg (accessed 10 January 2022).

23.

Cavaliere

. Clean ironmaking and steelmaking processes. Cham: Springer International Publishing, 2019. DOI: https://doi.org/10.1007/978-3-030-21209-4.

24.

Duarte

Tavano

Sales

, et al. Achieving Carbon Free Emissions via the ENERGIRON DR Process, https://www.energiron.com/media/papers-publications?page=1 (June 2010, accessed 24 July 2025).

25.

Miškovičová

Legemza

Demeter

, et al. An overview analysis of current research Status in iron oxides reduction by hydrogen. Metals (Basel) 2024; 14: 89.

26.

Fuel Cells and Hydrogen 2 Joint Undertaking. Hydrogen Roadmap Europe: A sustainable pathway for the European Energy Transition. Belgium, 2019.

27.

Spreitzer

Schenk

. Iron ore reduction by hydrogen using a laboratory scale fluidized bed reactor: kinetic investigation—experimental setup and method for determination. Metall Mater Trans B 2019; 50: 2471–2484.

28.

Nuber

Eichenberger

Rollinger

. Circored fine ore direct reduction. MILLENIUM Steel 2006: 37–40.

29.

IEA. The Future of Hydrogen - Seizing today’s opportunities, 2019.

30.

Anameric

Kawatra

. DIRECT IRON SMELTING REDUCTION PROCESSES. Miner Process Extr Metall Rev 2008; 30: 1–51.

31.

Singh

LR.

JSW Steel—Journey with Speed and Innovation. Indian Metallurgy. Singapore: Springer Nature, 2024, pp.99–103.

32.

S-H

Choi

M-E

Kim

D-H

, et al. FINEX® as an environmentally sustainable ironmaking process. Ironmak Steelmak 2019; 46: 625–631.

33.

Zhang

Jiao

Zhang

, et al. A review on low carbon emissions projects of steel industry in the world. J Clean Prod 2021; 306: 127259.

34.

Posco newsroom. POSCO starts to design the HyREX demonstration plant, https://newsroom.posco.com/en/posco-starts-to-design-the-hyrex-demonstration-plant/ (2022, accessed 18 January 2023 ) .

35.

Park

Lee

Y-J

Kim

J-H

, et al. Simulation of the hydrogen reduction process for the production of direct reduced iron. In: Kim

Song

, et al. (eds) Nano-, bio-, info-tech sensors, and wearable systems 2022. Long Beach, California: SPIE, 2022.

36.

Goodman

. Operation of the first HIsmelt plant in China. In: AISTech2019 proceedings of the iron and steel technology conference. Pittsburgh: AIST, 2019, pp.593–600.

37.

Shahabuddin

Brooks

Rhamdhani

. Decarbonisation and hydrogen integration of steel industries: recent development, challenges and technoeconomic analysis. J Clean Prod 2023; 395: 136391.

38.

Orth

Anastasijevic

Eichberger

. Low CO2 emission technologies for iron and steelmaking as well as titania slag production. Miner Eng 2007; 20: 854–861.

39.

Meijer

Zeilstra

Teerhuis

, et al. Developments in alternative ironmaking. Trans Indian Inst Met 2013; 66: 475–481.

40.

Cao

Meng

Yan

, et al. Analysis on energy efficiency and optimization of hismelt process. In: Wang

Chen

Guillen

, et al. (eds) Energy Technology 2019. Cham: Springer International Publishing, 2019. DOI: https://doi.org/10.1007/978-3-030-06209-5.

41.

Goodman

. The production of green steel via HIsmelt. In: Iron Ore 2021. AusImm, 2021.

42.

Cavaliere

. Hydrogen Assisted Direct Reduction of Iron Oxides. Cham: Springer, 2022. DOI: https://doi.org/10.1007/978-3-030-98056-6.

43.

Sohn

. From sulfide flash smelting to a novel flash ironmaking technology. In: Mackey

Grimsey

Jones

Brooks

(eds) Celebrating the Megascale. Cham: Springer, 2014, pp.69–76.

44.

Sohn

Mohassab

. Development of a novel flash ironmaking technology with greatly reduced energy consumption and CO2 emissions. Journal of Sustainable Metallurgy 2016; 2: 216–227.

45.

Pinegar

Moats

Sohn

. Flowsheet development, process simulation and economic feasibility analysis for novel suspension ironmaking technology based on natural gas: part 2 – flowsheet and simulation for ironmaking combined with steam methane reforming. Ironmak Steelmak 2013; 40: 32–43.

46.

Pinegar

Moats

Sohn

. Flowsheet development, process simulation and economic feasibility analysis for novel suspension ironmaking technology based on natural gas: part 1 – flowsheet and simulation for ironmaking with reformerless natural gas. Ironmak Steelmak 2012; 39: 398–408.

47.

Pinegar

Moats

Sohn

. Process simulation and economic feasibility analysis for a hydrogen-based novel suspension ironmaking technology. Steel Res Int 2011; 82: 951–963.

48.

Sohn

. Suspension Hydrogen Reduction of Iron Ore Concentrate in AISI/DOE Technology Roadmap Program for the Steel Industry. Utah: American Iron and Steel Institute, 2008.

49.

Sohn

Elzohiery

Fan

D-Q

. Development of the flash ironmaking technology (FIT) for green ironmaking with low energy consumption. Journal of Energy and Power Technology 2021; 03: 1–30.

50.

Fan

D-Q

Elzohiery

Mohassab

, et al. The kinetics of carbon monoxide reduction of magnetite concentrate particles through CFD modelling. Ironmak Steelmak 2021; 48: 769–778.

51.

Elzohiery

Fan

Mohassab

, et al. Kinetics of hydrogen reduction of magnetite concentrate particles at 1623–1873K relevant to flash ironmaking. Ironmak Steelmak 2021; 48: 485–492.

52.

Elzohiery

Sohn

Mohassab

. Kinetics of Hydrogen Reduction of Magnetite Concentrate Particles in Solid State Relevant to Flash Ironmaking. Steel Res Int 2017; 88. DOI: 10.1002/srin.201600133

53.

Fan

Elzohiery

Mohassab

, et al. Rate-enhancement effect of CO in magnetite concentrate particle reduction by H2 +CO mixtures. Ironmak Steelmak 2021; 48: 1064–1075.

54.

Fan

Mohassab

Elzohiery

, et al. Analysis of the hydrogen reduction rate of magnetite concentrate particles in a drop tube reactor through CFD modeling. Metall Mater Trans B 2016; 47: 1669–1680.

55.

Elzohiery

Fan

Mohassab

, et al. Experimental investigation and computational fluid dynamics simulation of the magnetite concentrate reduction using methane-oxygen flame in a laboratory flash reactor. Metall Mater Trans B 2020; 51: 1003–1015.

56.

Fan

D-Q

Sohn

Mohassab

, et al. Computational fluid dynamics simulation of the hydrogen reduction of magnetite concentrate in a laboratory flash reactor. Metall Mater Trans B 2016; 47: 3489–3500.

57.

Mohassab

Elzohiery

Sohn

. Flash reduction of magnetite concentrate based on partial oxidation of natural gas relevant to a novel ironmaking process. Utah: AISTech, 2016.

58.

Elzohiery

. Flash reduction of magnetite concentrate related to a novel flash ironmaking process. Utah: The University of Utah, 2018.

59.

Abdelghany

Fan

D-Q

Elzohiery

, et al. Experimental investigation and computational fluid dynamics simulation of a novel flash ironmaking process based on partial combustion of natural gas in a reactor. Steel Res Int; 90. 2019. DOI: 10.1002/srin.201900126.

60.

Meijer

Denys

Lasar

, et al. ULCOS: ultra-low CO2 steelmaking. Ironmak Steelmak 2009; 36: 249–251.

61.

Van Boggelen

Meijer

Zeilstra

, et al. HIsarna-demonstrating low CO2 ironmaking at pilot scale In: SteelVIA Conference. Dubai, 2018.

62.

Meijer

Guenther

Dry

. HIsarna pilot plant project. Dusseldorf: METEC InSteelCon, 2011.

63.

Yan

Htet

Hage

, et al. HIsarna process simulation model: using FactSage with macro facility. Metall Mater Trans B 2023; 54: 868–879.

64.

Peeters

Meijer

Van Der Stel

. Development of HIsarna Ironmaking Technology for Low CO2 Steelmaking. Dusseldorf: VDEh-CSM Seminar, 2014.

65.

European Commission: Directorate-General for Research and Innovation. Hisarna experimental campaigns B and C (HISARNA B and C) : final report, https://data.europa.eu/doi/10.2777/107258 (2015, accessed 13 January 2026).

66.

Boggelen

Hage

Meijer

, et al. HIsarna: benefits of a new developed smelting-reduction option for ironmaking. In: AISTech 2021 proceedings of the iron and steel technology conference. Nashville: AIST, 2021, pp.1388–1399.

67.

Gavel

Barnes

Zeilstra

, et al. Descriptive modelling of the HIsarna process. In: 4th International Conference on Science and Technology of Ironmaking and Steelmaking, Mumbai, 2022.

68.

Tata Steel. Tata Steel Board affirms the long-term strategy for growth in India, https://www.tatasteel.com/newsroom/press-releases/india/2025/tata-steel-board-affirms-the-long-term-strategy-for-growth-in-india/ (2025, accessed 13 January 2026).

69.

Ormston

Xia

Dufty

, et al. Breakthrough pilot testing results for the hydrogen direct flash reduced iron process. In: AISTech 2025 — proceedings of the iron & steel technology conference. Nashville: Association for Iron & Steel Technology, 2025, pp.2251–2265.

70.

Dufty

Calix Zesty tech – zero emissions iron and steel – ks5 pilot plant study report, https://arena.gov.au/knowledge-bank/calix-zesty-tech-zero-emissions-iron-and-steel-ks5-pilot-plant-study-report/ (2023, accessed 25 July 2025).

71.

Mokhtarani

Jafarian

Sabah

, et al. New insights into hydrogen reduction of hematite in an indirectly heated flash reactor from measurements and first-order modeling. Metall Mater Trans B 2025; 56: 1928–1941.

72.

Diamandescu

Mihàilà-Tàràbàşanu

Feder

. On the solid phase transformation goethite → hematite. Mater Lett 1993; 17: 309–311.

73.

Walter

Buxbaum

Laqua

. The mechanism of the thermal transformation from goethite to hematite. J Therm Anal Calorim 2001; 63: 733–748.

74.

Ammasi

. Effect of heating rate on decomposition temperature of goethite ore. Trans Indian Inst Met 2020; 73: 93–98.

75.

López

Ramirez

Pons

, et al. Kinetic study of the thermal decomposition of low-grade nickeliferous laterite ores. J Therm Anal Calorim 2008; 94: 517–522.

76.

Leerhoff

Brouwer

Armaki

, et al. Characterisation of varying iron ores and their thermal decomposition kinetics under HIsarna ironmaking conditions. Metals (Basel) 2024; 14: 1271.

77.

Przepiera

. Kinetics of thermal transformations of precipitated magnetite and goethite. J Therm Anal Calorim 2001; 65: 497–503.

78.

Grygar

Ruan

IHD

Gilkes

. Re-examination of the kinetics of the thermal dehydroxylation of goethite. J Therm Anal Calorim 1999; 55: 301–309.

79.

Beuria

Biswal

Mishra

, et al. Study on kinetics of thermal decomposition of low LOI goethetic hematite iron ore. Int J Min Sci Technol 2017; 27: 1031–1036.

80.

Strezov

Evans

Zymla

, et al. Structural deterioration of iron ore particles during thermal processing. Int J Miner Process 2011; 100: 27–32.

81.

Ariyama

Isozaki

Matsubara

, et al. Fluidization and degradation characteristics of iron ore fines in prereduction fluidized bed. ISIJ Int 1993; 33: 1220–1227.

82.

Pinson

Zulli

, et al. Interaction between mineral phases in a goethitic iron ore and fluxing materials during sintering. Metall Mater Trans B 2021; 52: 996–1011.

83.

Murakami

Nishimura

Kasai

. Lowering reduction temperature of iron ore and carbon composite by using ores with high combined water content. ISIJ Int 2009; 49: 1686–1693.

84.

Hirokawa

Naito

Yamaguchi

. Effect of atmosphere on the goethite decomposition and pore structure of product particles. J Colloid Interface Sci 1986; 112: 268–273.

85.

Yang

Zou

, et al. Thermal decomposition behaviour of fine iron ore particles. ISIJ Int 2014; 54: 2196–2205.

86.

Xing

Wang

, et al. Kinetic study on thermal decomposition behavior of hematite ore fines at high temperature. Metall Mater Trans B 2020; 51: 395–406.

87.

CRCT -ThermFact Inc. GTT-Technologies. FactSage, Equilib Module.

88.

CRCT -ThermFact Inc. GTT-Technologies. FactPS Database.

89.

Chen

Zeilstra

Van Der Stel

, et al. Thermal decomposition reaction kinetics of hematite ore. ISIJ Int 2020; 60: 65–72.

90.

Salmani

Alamdari

Firoozi

. Isoconversional analysis of thermal dissociation kinetics of hematite in air and inert atmospheres. J Therm Anal Calorim 2017; 128: 1385–1390.

91.

Ozawa

. Kinetics of non-isothermal crystallization. Polymer (Guildf) 1971; 12: 150–158.

92.

Coats

Redfern

. Kinetic parameters from thermogravimetric data. Nature 1964; 201: 68–69.

93.

Jozwiak

Kaczmarek

Maniecki

, et al. Reduction behavior of iron oxides in hydrogen and carbon monoxide atmospheres. Appl Catal A Gen 2007; 326: 17–27.

94.

Sun

. A study of kinetics and mechanisms of iron ore reduction in ore/coal composites . PhD Thesis, McMaster University, 1997.

95.

. Experimental Study of the Melting and Reduction Behaviour of Ore Used in the HIsarna Process. Proefschrift, Technische Universiteit Delft, 2013.

96.

Spreitzer

Schenk

. Reduction of iron oxides with hydrogen—A review. Steel Res Int 2019; 90: 1900108.

97.

Heidari

Niknahad

Iljana

, et al. A review on the kinetics of iron ore reduction by hydrogen. Materials (Basel) 2021; 14: 1–19.

98.

Meshram

Govro

OMalley

, et al. Modeling isothermal reduction of iron ore pellet using finite element analysis method: experiments & validation. Metals (Basel) 2022; 12: 2026.

99.

Levenspiel

. Fluid-Particle Reactions: Kinetics. Chemical Reaction Engineering. New York: John Wiley & Sons, 1999, pp.566–588.

100.

Han

Wang

Lan

, et al. Kinetic analysis of iron oxide reduction in top gas recycling oxygen blast furnace. Ironmak Steelmak 2012; 39: 313–317.

101.

Xing

Zou

, et al. Gas–solid reduction behavior of in-flight fine hematite ore particles by hydrogen. Steel Res Int 2019; 90: 1–10.

102.

Yang

Zou

, et al. Reduction kinetics of fine hematite ore particles with a high temperature drop tube furnace. ISIJ Int 2015; 55: 952–960.

103.

Šimon

. Isoconversional methods. J Therm Anal Calorim 2004; 76: 123–132.

104.

Zhang

Monaghan

Longbottom

, et al. Reduction kinetics of oxidized New Zealand ironsand pellets in H2 at temperatures up to 1443K. Metall Mater Trans B 2020; 51: 492–504.

105.

Sohn

Jung

. Effect of metal additions to the reduction of iron oxide composite pellets with hydrogen at moderate temperatures. Steel Res Int 2011; 82: 1345–1354.

106.

Piotrowski

Mondal

Lorethova

, et al. Effect of gas composition on the kinetics of iron oxide reduction in a hydrogen production process. Int J Hydrogen Energy 2005; 30: 1543–1554.

107.

Zuo

Wang

Dong

, et al. Reduction kinetics of iron oxide pellets with H2 and CO mixtures. Int J Miner Metall Mater 2015; 22: 688–696.

108.

Moon

I-J

Rhee

C-H

Min

D-J

. Reduction of hematite compacts by H2-CO gas mixtures. Steel Res 1998; 69: 302–306.

109.

Lyu

Wang

Yang

, et al. Kinetic analysis of isothermal and non-isothermal reduction of iron ore fines in hydrogen atmosphere. Metals (Basel) 2022; 12: 1754.

110.

Hammam

Nie

, et al. Isothermal and non-isothermal reduction behaviors of iron ore compacts in pure hydrogen atmosphere and kinetic analysis. Min Metall Explor 2021; 38: 81–93.

111.

Kuila

Chatterjee

Ghosh

. Kinetics of hydrogen reduction of magnetite ore fines. Int J Hydrogen Energy 2016; 41: 9256–9266.

112.

Yang

Zou

, et al. Melting and reduction behaviour of individual fine hematite ore particles. ISIJ Int 2015; 55: 149–157.

113.

Yang

Zou

, et al. Kinetic study on gas molten particle reduction of iron ore fines at high temperature. Ironmaking Steelmaking 2015; 42: 763–773.

114.

Chen

Mohassab

Jiang

, et al. Hydrogen reduction kinetics of hematite concentrate particles relevant to a novel flash ironmaking process. Metall Mater Trans B 2015; 46: 1133–1145.

115.

Chen

Mohassab

Zhang

, et al. Kinetics of the reduction of hematite concentrate particles by carbon monoxide relevant to a novel flash ironmaking process. Metall Mater Trans B 2015; 46: 1716–1728.

116.

Huang

Peng

, et al. Action rules of H2 and CO in gas-based direct reduction of iron ore pellets. J Cent South Univ 2012; 19: 2291–2296.

117.

Hou

Zhang

, et al. Study on kinetics of iron oxide reduction by hydrogen. Chin J Chem Eng 2012; 20: 10–17.

118.

Wagner

Devisme

Patisson

, et al. A laboratory study of the reduction of iron oxides by hydrogen. In: Sohn International Symposium, San Diego, United States, pp.111–120.

119.

Pang

Guo

Zhao

, et al. Influence of size of hematite powder on its reduction kinetics by H2 at low temperature. J Iron Steel Res Int 2009; 16: 7–11.

120.

Chen

Zeilstra

van der Stel

, et al. Reduction kinetics of fine hematite ore particles in suspension. Metall Mater Trans B 2021; 52: 3393–3402.

121.

Chen

Zeilstra

van der Stel

, et al. Reduction mechanism of fine hematite ore particles in suspension. Metall Mater Trans B 2021; 52: 2239–2252.

122.

Chen

Zeilstra

, et al. Prediction of density and volume variation of hematite ore particles during in-flight melting and reduction. J Iron Steel Res Int 2019; 26: 1285–1294.

123.

Xing

Wang

, et al. Kinetic characterization of flash reduction process of hematite ore fines under hydrogen atmosphere. Int J Hydrogen Energy 2020; 45: 31481–31493.

124.

Newsome

. The water-gas shift reaction. Catal Rev 1980; 21: 275–318.

125.

Reddy

Smirniotis

. Introduction about WGS reaction. In: Water gas shift reaction. Elsevier, 2015, pp.1–20.

126.

CRCT - ThermFact Inc. GTT - Technologies. FactSage, reactions module.

127.

Idriss

Scott

Subramani

. Introduction to hydrogen and its properties. In: Compendium of hydrogen energy. Elsevier, 2015, pp.3–19.

128.

Abdelrahim

Iljana

Omran

, et al. Influence of H2–H2O content on the reduction of acid iron ore pellets in a CO–CO2–N2 reducing atmosphere. ISIJ Int 2020; 60: 2206–2217.

129.

Lorente

Herguido

Peña

. Steam-iron process: influence of steam on the kinetics of iron oxide reduction. Int J Hydrogen Energy 2011; 36: 13425–13434.

130.

Kim

W-H

Lee

Kim

S-M

, et al. The retardation kinetics of magnetite reduction using H2 and H2–H2O mixtures. Int J Hydrogen Energy 2013; 38: 4194–4200.

H 2 adaptation in ironmaking: a review with focus on suspension smelting and its reaction mechanisms

Abstract

Keywords

Introduction

Hydrogen in DRI processes

DRI in shaft furnace

DRI in a fluidised bed reactor

Smelting reduction ironmaking

Suspension smelting ironmaking

Reaction mechanisms in suspension smelting

Thermal decomposition of goethite

Thermal decomposition of hematite

Gas reduction of iron oxides

Reaction kinetics of hematite reduction with H2 and/or CO

Water–gas shift reaction

Conclusion

Footnotes

ORCID iD

Ethical considerations

Consent to participate

Authorship contribution statement

Funding

Declaration of conflicting interests

Data availability statement

References

Reaction kinetics of hematite reduction with H₂ and/or CO