Iron Oxide Direct Reduction and Iron Nitride Formation Using Ammonia: Review and Thermodynamic Analysis

Abstract

The steel industry is one of the main contributors to global greenhouse gas emissions, responsible for about 7 to 9% of the world’s total output. The steel sector is under pressure to move toward net-zero emissions by reducing its consumption of coke as the main method of reducing iron-rich feed materials to iron. Due to its well-developed synthesis process, high supply chain, straightforward handling technologies, and highly developed long-standing infrastructure, ammonia has the potential to become a replacement for coke as a future iron ore reductant. This work reviews previous research on ammonia direct reduction of iron oxides and the possible formation of iron nitrides. A thermodynamic assessment using FactSage 8.2 thermochemical software was carried out examining the behavior of ammonia gas as the reductant upon heating, detailed evaluations of the stable phases present under different reaction conditions and using different feed materials, and the formation and stability of iron nitride phases. The results suggest that the reduction of hematite with ammonia occurs in two steps below 570 °C and three steps above 570 °C. The ratio of Fe₂O₃/NH₃ was predicted to affect the reduction reactions by promoting a greater reduction degree and simultaneously lowering the initial temperature needed for reduction, while the excess gas concentration can suppress FeO formation. A predominance area diagram was developed showing the main areas of stable phases as a function of the partial pressure of NH₃ and temperature. The formation of iron nitrides during the process was predicted and these were not expected to cause issues for the formation of iron due to their instability under the conditions studied. This analysis can be used to inform further experimental studies regarding ammonia reduction of iron oxide.

Graphical Abstract

Introduction

Steel is the largest manufactured metal reaching a global production capacity of 1,878.5 Mt of crude steel in 2022 (see Fig. 1a) [1]. The amount of steel produced has increased by up to 130% in the past two decades and is anticipated to further grow. According to recent IEA data, around 70% of the total steel production worldwide was produced via the well-established BF-BOF technology [2]. The blast furnace has been used worldwide for more than 700 years and remains the most efficient continuous process of ironmaking as its counter-current flow system ensures good transference of mass and heat due during operation [3]. It also has outstanding process stability, an extensive operating lifetime (12 to 20 + years), and a high iron production rate up to 13,000 tons of hot metal per day [3]. However, minimizing carbon dioxide emissions and the environmental impact of each facility is now becoming a significant challenge in BF ironmaking. The integrated BF-BOF steel plant releases 2.32 tons CO₂ per ton of crude steel cast on the basis of scope 1, 2, and 3 emissions [4]. It is also reported that 24.43 GJ of energy is required to produce 1 ton of cast crude steel in 2021 [4]. Numerous studies have been conducted examining the potential for coke and coal replacement with materials such as biomass and hydrogen. However, both these new process options face challenges, such as acquiring adequate supplies of sustainably sourced biomass and also issues with disrupting the heat distribution in the BF through partial use of hydrogen [5,6,7,8,9,10,11,12]. Furthermore, the blast furnace is a mature technology and the potential to improve its energy and emission minimization is approaching its limits. Therefore, developing a sustainable and innovative technology with low emissions to contribute to global warming mitigation is an important priority in metallurgy.

Fig. 1

a World crude steel production from 1950 to 2022 (data taken from World Steel Association Annual Report 2023), b World ammonia production from 2013 to 2022 (data taken from International Fertilizer Association and The Royal Society)

Various pathways have been launched globally to develop new low-emission iron and steelmaking technologies e.g., COURSE50 in Japan, ULCOS in Europe, hydrogen flash ironmaking technology in US, and HYBRIT in Sweden [13]. The goal is to create low-CO₂ breakthrough technologies that combine method changes or new process units such as top gas recycling for the BF, direct reduction with an electric arc furnace (HyDR-EAF), electrolysis, and carbon capture and storage (CCS). Hydrogen metallurgy has been regarded as the most promising way to mitigate CO₂ emissions from ironmaking and it has been shown that the HyDRI-EAF process can significantly reduce CO₂ levels to less than 0.3 ton CO₂/ton crude steel and potentially eliminate 84% of total CO₂ emissions compared to the conventional BF-BOF route [14]. However, significant barriers regarding the use and transport of hydrogen exist that prevent expansion of hydrogen applications. These include a high evaporation rate, a low volumetric density, and low ignition energy, coupled with techno-economic challenges in transportation, distribution, and long-term storage of hydrogen [15,16,17].

Ammonia (NH₃), whose previous main use has been as a fertilizer for agriculture, has recently attracted attention as a potential option for supplying hydrogen as an alternative low-CO₂ reductant. Ammonia is an attractive hydrogen carrier compared to liquid H₂ and LOHC (e.g., methylcyclohexane, 12H-N-ethylcarbazole, 18H-dibenzyltoluene) suitable for long-distance transport, as hydrogen constitutes 17.6 wt.% of ammonia [18,19,20]. Among all hydrogen carriers, ammonia is the easiest to be liquefied by pressurizing at – 33 °C at 1 atm (10 bar at room temperature) compared with liquid H₂, which must be liquefied at − 252.9 °C, at the same pressure, resulting in significant energy usage [16, 19]. Liquid ammonia also has a 1.5 to 1.7 times higher volumetric hydrogen density than liquid H₂, which is 120 kg-H₂/m³ in liquid ammonia compared to 70 kg-H₂/m³ in liquid H₂ [15]. As well, the density of gaseous ammonia is lower than air (0.769 kg/m³ compared to 1.225 kg/m³) ensuring it dissipates rapidly in air at atmospheric pressure leading to lower risk of explosion in the case of leakages [19]. Moreover, at current production rates of hydrogen, ammonia may be considered as an excellent alternative hydrogen carrier because of its availability in larger quantities, reaching a worldwide production rate of up to 180 million tons NH₃ (see Fig. 1b) [21] making it the world’s second most-produced chemical. Ammonia’s overall global infrastructure maturity for manufacture (through the commercialized Haber–Bosch process), ease of transportation, and well-established distribution technologies provide ammonia with a significant advantage over hydrogen [22]. Furthermore, the viability of converting the existing LNG facilities to distribute ammonia also offers wider opportunities [2].

Since its first commercial plant commenced operation in 1913, the Haber–Bosch process has been used as the primary ammonia synthesis technique. The ammonia synthesis currently relies on fossil fuel for hydrogen generation via coal gasification and the steam methane reforming (SMR) process using natural gas (> 70%) was responsible for 1.3% (450 Mt) of global CO₂ emissions in 2020 [23]. To reduce emissions, modern ammonia plants are transitioning to using green hydrogen produced through water electrolysis and to a modified Haber–Bosch process that is powered by renewable energy sources to attain decarbonization and net-zero emissions. These modifications consist of the use of process units such as alkaline water electrolysis (AWE), solid oxide water electrolysis (SOE), and polymer electrolyte membrane water electrolysis (PEM WE) that reached TRL of 9, 5, and 8, respectively [24]. Due to its high TRL and low investment cost, the AWE has been industrialized and will potentially become the method for a sustainable green NH₃ production system in the future [24].

Ammonia is considered as a potential chemical option to replace fossil resources for iron ore reductant [25]. Direct reduction of iron oxide with ammonia offers a novel process for sustainable ironmaking at low temperatures and low carbon emissions. This process can potentially pave the way to having a more environmentally friendly ironmaking process by using renewable energy resources in all streams (see Fig. 2). The use of ammonia as a reductant also solves many of the logistic and energy-related problems that would be associated with transporting and storing hydrogen. However, there are only few studies that have been published on metal oxide reduction using ammonia and the thermodynamics of the process have not yet been adequately explored. This work reviews previous studies on ammonia direct reduction (ADR) of iron oxides, followed by a detailed thermodynamic study of iron ore reduction using ammonia that covers aspects such as the behavior of ammonia decomposition, an evaluation of the phases stabilized under different reaction conditions, and the formation and stability of iron nitrides as one of the potential phases when producing iron under ammonia atmosphere.

Fig. 2

Schematic flow diagram outlining the process of ammonia direct reduced iron prepared using green hydrogen from water electrolysis for ammonia synthesis and as a renewable energy source

Thermodynamic Calculations

Due to its ability to examine the thermodynamic behavior of multicomponent and multiphase system under variety of conditions (temperature, pressure, volume, chemical compositions, chemical activity, etc.), computational thermochemistry is widely utilized for the rapid assessment of equilibrium state and potential stable phase of pyrometallurgical processes [26]. In this study, FactSage 8.2 thermochemical software and databases were used to perform phase equilibrium computations based on Gibbs Free Energy. The following databases were used in the calculations in the current study: FactPS (pure substances), FToxid (pure oxide compounds and solutions), and FSstel (steel alloy compounds and solutions).

The thermodynamic evaluation was carried out through equilibrium state analysis of possible chemical reactions together with possible stable phases analysis under different processing conditions, such as molar ratio, pressure, and iron oxide type. The evaluation was calculated using two modules, i.e., “Reaction” and “Equilib.” The Reaction module computes change in thermochemical properties under several constraints for chemical reactions involved in the reduction process, including enthalpy (H), Gibbs free energy (G), Volume (V), entropy (S), specific heat (C_p), and activity (A). The calculation using Reaction module included the ammonia decomposition reaction, the reduction reaction for each of the iron oxides (i.e., Fe₂O₃, Fe₃O₄, and FeO) with hydrogen and ammonia at standard state, and the reduction reaction of various iron types with ammonia at standard state. The Equilib module employs Gibbs free minimization to compute the conditions for multiphase and multicomponent equilibria under a wide range of possible conditions [27,28,29]. Calculations using the Equilib module included the ammonia decomposition reaction at various pressure (0.01, 0.1, 0.5, 1, 2, and 3 atm), the reduction of hematite with Fe₂O₃/NH₃ mole ratio of 1/2, 1/4, 1/8, and 1/20 at standard state, the reduction of hematite with ammonia at different pressures (0.001, 0.01, 0.1 1, 2, 3, 5, and 10 atm), and the reduction of various other iron oxide types with ammonia at standard state.

Previous Studies on Direct Reduction of Iron Oxides with Ammonia

Considering the properties and world trading capacity, ammonia is considered as a potential reductant for iron reduction to support green steelmaking [17]. However, there are only four laboratory-based studies related to ammonia ironmaking that have been published to date (see Table 1) [25, 30,31,32]. The basic reaction of the process will depend on the reduction temperature and can be categorized as either a direct or an indirect reduction mechanism [30, 31]. In direct reduction, iron oxide will be directly reduced by ammonia to metallic iron, while for indirect reduction, the reduction will occur only after reaching the decomposition temperature of ammonia which will generate hydrogen and subsequently reduce the iron oxide (i.e., similar to existing hydrogen ironmaking principles). The key generic reactions are presented in Eqs. 1, 2, and 3.

Table 1 Previous studies on ammonia ironmaking

Direct reduction reaction:

$$ {\text{Fe}}_{2} {\text{O}}_{{3({\text{s}})}} + \, 2{\text{NH}}_{{3({\text{g}})}} \to \, 2{\text{Fe}}_{{({\text{s}})}} + {\text{ N}}_{{2({\text{g}})}} + \, 3{\text{H}}_{{2}} {\text{O}}_{{({\text{g}})}} $$

(1)

Indirect reduction reaction:

$$ {\text{NH}}_{{\text{3(g)}}} \to \, \raise.5ex\hbox{$\scriptstyle 1$}\kern-.1em/ \kern-.15em\lower.25ex\hbox{$\scriptstyle 2$} {\text{ N}}_{{2({\text{g)}}}} + \, 3/2{\text{ H}}_{{2({\text{g}})}} $$

(2)

$$ {\text{Fe}}_{2} {\text{O}}_{{\text{3(s)}}} + \, 3{\text{H}}_{{2{\text{(g)}}}} \to \, 2{\text{Fe}}_{{\text{(s)}}} + \, 3{\text{H}}_{2} {\text{O}}_{{\left( {\text{g}} \right) }} $$

(3)

Hosokai et al. (2011) observed that direct reduction of hematite with ammonia started at 430 °C as indicated by an increase of steam and nitrogen in the off-gas and the decrease of ammonia flux as recorded using QMS (Quadrupole Mass Spectrometry). This reaction produced a mixture of metallic iron and iron nitride in the form of Fe₄N. At 450 °C, the reduction ratio increased with an increasing amount of magnetite formation and hydrogen flux. This observation suggested that iron served as a catalyst for ammonia decomposition. The reaction shifted to an indirect reaction mechanism at 530 °C, as demonstrated by a decrease in the amount of the steam flux and a decrease in the amount of hydrogen measured by QMS. The final product obtained at 600 to 700 °C was metallic iron as the iron nitride is easily decomposed at temperatures higher than 600 °C. They also reported that the reaction rate was increased as the temperature was increased. However, this study did not examine the thermodynamics and reaction mechanisms of the reduction process, including the interaction between the flowing ammonia gas and the single solid iron ore particle.

Yasuda et al. (2015) conducted laboratory experiments to test the reduction and nitriding behavior of hematite (Fe₂O₃) powder at 5, 10, 20, and 30% NH₃. They reported that hematite was directly reduced to magnetite by ammonia at temperatures between 500 and 590 °C. The magnetite formed by the hematite reaction with ammonia and then was reduced to iron by ammonia-decomposed hydrogen at 600 °C (Eq. 4). According to in situ XRD results, they found that hematite was reduced to magnetite at 500 °C and then nitrided to form ε-Fe_3-xN at 600 °C (Eq. 5). The γ’-Fe₄N phase was formed at 700 °C due to decomposition of the ε-Fe_3-xN. Iron nitride was subsequently decomposed at a higher temperature through Eq. 6. An increase in the ammonia concentration from 5 to 20% resulted in an increase of the reduction rate, while the nitriding potential was proportional to the ammonia concentration [31]. However, there was no clear understanding on the kinetics of this process including the gas diffusion behavior in a single solid iron ore particle and the rate controlling factors.

$$ \raise.5ex\hbox{$\scriptstyle 1$}\kern-.1em/ \kern-.15em\lower.25ex\hbox{$\scriptstyle 2$} {\text{Fe}}_{2} {\text{O}}_{{3({\text{s}})}} + \, 1/9{\text{NH}}_{{3({\text{g)}}}} \to \, 1/3{\text{Fe}}_{3} {\text{O}}_{{\text{4(s)}}} + \, 1/6{\text{H}}_{2} {\text{O}}_{{\text{(g)}}} + \, 1/18{\text{N}}_{{\text{2(g)}}} $$

(4)

$$ {\text{yFe}}_{{\text{(s)}}} + {\text{ NH}}_{{\text{3(g)}}} \to {\text{ Fe}}_{{\text{y}}} {\text{N}}_{{\text{(s)}}} + \, 3/2{\text{ H}}_{{2\left( {\text{g}} \right) }} $$

(5)

$$ {\text{Fe}}_{{\text{y}}} {\text{N}}_{{({\text{s)}}}} \to {\text{ yFe}}_{{({\text{s)}}}} + \, 1/2{\text{N}}_{{2({\text{g}})}} $$

(6)

Iwamoto et al. (2022) conducted ammonia ironmaking experiment with different types of natural goethite (FeO.OH)-rich ores under various temperatures and cooling techniques. Because goethite is an oxy-hydroxide iron phase, they also examined the effect of water by testing high-combined water ores (high-CW or ore A) and low-combined water ores (low-CW or ore B). It was reported that high-CW ores showed higher reducibility with a maximum reduction degree of 96% achieved after 20 min of reduction, compared to low-CW ores that reached 65% of reduction degree at 700 °C after the same holding time. It was suggested that due to the higher surface area of pores induced by the goethite dehydration/dehydroxylation process, more Fe₂O₃ was in contact with NH₃, which considerably boosted the reaction kinetics for direct reduction in the initial stage. The fact that only N₂ was produced without any H₂ generation in the high-CW ore reduction experiments confirmed that direct reduction occurred at the beginning of reduction. On the contrary, H₂ and N₂ showed comparable increases in flow rate, suggesting that no initial direct reduction with NH₃ took place with low-CW ore reduction. They also assumed that hematite has a low catalytic activity for ammonia decomposition. Thus, the reduction of hematite to magnetite, wűstite, and metallic iron may increase the catalytic activity. During cooling process, Fe was nitrided in the presence of NH₃ gas and the iron nitride decomposed into Fe in the absence of NH₃. However, the knowledge of impurities behavior of iron ore on the ammonia direct reduction process is also inadequate.

Ma et al. (2023) conducted reduction experiments on commercial iron ore pellets with ammonia of 99.999% purity and with a gas flow rate of 10 l/h in thermogravimetric (TG) configuration. Results were compared to hydrogen reduction (HyDR) using 99.999% hydrogen at the same experimental conditions (i.e., at 700 °C isothermally for 2 h). Ammonia direct reduction (ADR) reaction achieved a reduction degree of 98%, and was kinetically similar to HyDR according to the TG analysis [25]. However, the sample mass was increased at the final process, and as a result, the reduction degree declined by 7.5% due to partial nitriding of the reduced Fe to form passive Fe₄N on the surface of the DRI during the cooling process. The QMS results indicated that there was no evidence of NO_x formation in the off-gas throughout the process. The formation of iron nitride at the surface was also confirmed through a phase map, constructed by electron backscatter diffraction (EBSD) and compositional analysis from a 3D elemental distribution analysis across the interface.

Ma et al. (2023) suggested that the formation of iron nitride at the surface of DRI could contribute to improve the corrosion resistance of DRI and prevent the porous DRI from re-oxidation. The presence of iron nitride will also be beneficial to shipping and handling of DRI without any necessity for subsequent hot briquette compaction [25]. The iron nitride formation was proved to be dissolved and removed during the melting process experimentally. They proposed three stages of the process during heating according to mass spectrometry signals: (1) no significant changes below 350 °C; (2) ammonia decreased significantly from 350 to 650 °C due to ammonia decomposition; and (3) reduction of iron ore with hydrogen derived from ammonia decomposition above 650 °C. However, none of these studies indicated how the iron nitride forms and its competition with iron formation was also not fully identified. The morphological progression of pellet during heating was also not studied.

Formation of Iron Nitride

Since 1905, researchers have been examining the equilibrium between iron, iron nitride, ammonia, and hydrogen at high temperatures. In 1930, Lehler constructed the equilibrium diagram for the iron and nitrogen system, consisting of iron-NH₃-iron nitride-H₂ and the iron–nitrogen systems, by magnetometric and x-ray methods [33]. Through this diagram (Fig. 3), the equilibrium phase of iron nitriding at a certain temperature in the range of 400–750 °C and nitrogen composition can be predicted. The ɣ’ iron nitride (ɣ’-Fe₄N_1−x) is the phase containing FCC iron lattice with nitrogen in octahedral interstitial sites and ε iron nitride (ε-Fe₃N_1+x or ε-Fe₂N_1−x) contains HCP iron lattice with nitrogen in octahedral interstitial sites.

Fig. 3

Temperature and nitriding potential relationship (Lehrer Diagram) corresponding to ammonia content in NH₃–H₂ gas mixture at 1 atm [36]

Iron nitrides exist as following compounds: Fe₁₂N, Fe₈N, Fe₆N, Fe₄N, Fe₇N₂, Fe₆N₂, Fe₆N₂, Fe₅N₂, Fe₂N, Fe₃N₂, and FeN, but Fe₄N and Fe₂N are the most common forms [34]. Fe₄N is a ferromagnetic compound, having face-centered cubic crystal structure, while Fe₂N is paramagnetic with hexagonal close packed structure. Fe₂N can be produced by applying gaseous ammonia at 450 °C over iron powder. The Fe₂N can be converted to Fe₄N by heating it up in vacuum pressure at 440–550 °C [34].

Thermodynamics studies have been carried out regarding the formation of iron nitrides. Satoh (1932) calculated the equilibrium constants of both iron nitrides, i.e., Fe₄N and Fe₂N based on the reactions:

$$ 8{\text{Fe }} + \, 2{\text{NH}}_{3} = \, 2{\text{Fe}}_{4} {\text{N }} + \, 3{\text{H}}_{{2}} $$

(7)

$$ 2{\text{Fe}}_{4} {\text{N }} + \, 2{\text{NH}}_{3} = \, 4{\text{Fe}}_{2} {\text{N }} + \, 3{\text{H}}_{2} $$

(8)

The equilibrium equation of reaction Eqs. 7 and 8 was expressed as log Kp₁ and log Kp₂, respectively, according to the equations [34]:

$$ \log K_{P1} = - \frac{2305}{T} + 3.2676\log T + 0.0021174T - 3.72 \times 10^{ - 7} T^{2} - 6.8 $$

(9)

$$ \log K_{P2} = - \frac{3548}{T} + 3.2676\log T + 0.0021174T - 3.72 \times 10^{ - 7} T^{2} - 6.8 $$

(10)

They also compared the calculation with experimentally observed equilibrium constants at 461 °C and 500 °C. The calculations showed comparable results where the experiment equilibrium constant of reaction is 0.752 and 1.107 at each temperature and calculated values were 0.876 and 1.073, respectively. Brunauer et al. (1930) carried out an experiment by using a series of NH₃-H₂ gas mixture that was passed over a hydrogen-reduced iron from ferric oxalate (C₆Fe₂O₁₂) [33]. They also attempted to use iron obtained from reduced ferric oxide; however, it was significantly decomposed ammonia at 444 °C. They found that at 1 atm, Fe was converted into Fe₄N using no more than 37% NH₃ at 400 °C and 14.5% at 575 °C.

Thermochemical surface engineering techniques through nitriding have previously been used to increase the corrosion, fatigue, and wear resistance as well as magnetic properties of steel which has a similar magnetic properties to permanent rare-earth magnets [35]. Mittemeijer and Somers (1997) suggested that a considerable amount of iron nitride formation does not occur at 1 atm through this following equation [36]:

$$ 4{\text{Fe}} + 1/2\left( {1 - x} \right){\text{N}}_{2} = {\text{Fe}}_{4} {\text{N}}_{1 - x} $$

(11)

Instead, it can be formed by reacting Fe with NH₃–N₂ gas mixtures at 1 atm and temperature of 227 to 727 °C by the reaction [36]:

$$ 4{\text{Fe}} + \left( {1 - x} \right)\,{\text{NH}}_{{3}} = {\text{Fe}}_{{4}} {\text{N}}_{{{1 - }x}} + 3/2\,\left( {1 - x} \right)\,{\text{H}}_{2} $$

(12)

The relationship between temperature and nitriding potential (develop by Lehrer and known as Lehrer Diagram) was plotted together with corresponding ammonia content in ammonia-hydrogen mixtures at 1 atm as can be seen in Fig. 3. The nitriding potential (r_N) is expressed by reaction Eq. 13 below:

$$ r_{N} = \frac{{p_{{NH_{3} }} }}{{p_{{H_{2} }}^{3/2} }}. $$

(13)

According to this figure, it can be concluded that Fe₄N (ɣ′) can be formed by reacting Fe using a 10 vol% of NH₃-H₂ mixture at 600 °C, 30 vol% at 450 °C, and as low as 350 °C when using 40 vol% of NH₃. In addition, Fe₂N (ε) was produced using 10 vol-% of NH₃ at 650 °C and as low as 450° under 70 vol% of ammonia. All of the solid iron nitride phases are interstitial solid solutions, based on a Fe and N sublattice made up of all of the octahedral Fe sublattice interstices. The nitrogen atoms are dispersed randomly over the sites of N sublattice in both α (with BCC Fe lattice) and ɣ (with FCC Fe lattice), while they exhibit long-range ordering on N sublattice in the ɣ′ and ε phase, with FCC and HCP Fe sublattice, respectively [36].

Voorthuysen et al. (2002) extended the Lehrer and Fe–N phase diagram further to lower temperatures using chemical nitriding technique down to 240 °C. They found the triple point α/ɣ'/ε at 240 °C ± 10 °C and r_N of − 2.04 from calculation of the new boundaries and the ɣ' phase was no longer stable below this temperature, remaining as the stable phase of α and ε phases. They also derived some calculations regarding the relationship of nitrogen content as a function of temperature and nitriding potential for the α, ɣ', and ε phases as presented in Eqs. 14, 15, and 16, respectively.

$$ \alpha - {\text{FeN}}_{{\text{c}}} \left( {300\,^\circ {\text{C to }}590\,^\circ {\text{C}}} \right): $$

$$ c = 3\exp [\ln \left( {r_{N} } \right) - \frac{9096}{T} + 11.56] $$

(14)

$$ \gamma ^{\prime} - Fe_{4} N_{(1 + z)} \left( {400\,^\circ {\text{C to }}600\,^\circ {\text{C}}} \right): $$

$$ z = - 4\exp [ - \ln \left( {r_{N} } \right) - \frac{1210}{T} - 10.906] $$

(15)

$$ \varepsilon - {\text{Fe}}_{x} {\text{N }} = \, \varepsilon - {\text{FeN}}_{y} \left( {400{\text{ to }}550\,^\circ {\text{C}}} \right): $$

$$ \ln \left( {r_{N} } \right) = 82.87 y^{2} - 30.76 y + \frac{3590}{T} - 8.76 $$

(16)

A complete Fe–N phase diagram was calculated using FactSage 8.2 as shown in Fig. 4. Iron exists in solid solution with nitrogen for each phase transformation. The highest amount of nitrogen that nitrogen ferrite (bcc, α-Fe) can dissolve is 1.6 × 10^–2 wt% nitrogen, while in comparison nitrogen austenite (fcc, ɣ-Fe), stable above 589 °C, can contain up to 10.3 wt% nitrogen [37]. Fe₄N is the only stable phase at around 0.19 to 0.2 of concentration (mol N/ (mol Fe + mol N)) from 427 to 692 °C.

Fig. 4

Fe–N phase diagram (calculated using FactSage 8.2)

A thermodynamic assessment was conducted using FactSage 8.2 to see the possibility of the iron oxide nitriding reaction occurring through an ammonia decomposition scheme according to the reactions:

$$ {\text{Fe}}_{2} {\text{O}}_{{\text{3(s)}}} + \, 3{\text{ H}}_{{2{\text{(g)}}}} + \, 0.25{\text{ N}}_{{\text{2(g)}}} = \, 0.5{\text{ Fe}}_{4} {\text{N}}_{{({\text{s}})}} + \, 3{\text{ H}}_{{2}} {\text{O}}_{{\text{(g)}}} $$

(17)

$$ {\text{Fe}}_{3} {\text{O}}_{{\text{4(s)}}} + \, 4{\text{ H}}_{{\text{2(g)}}} + \, 0.375{\text{ N}}_{{\text{2(g)}}} = \, 0.75{\text{ Fe}}_{{4}} {\text{N}}_{{\text{(s)}}} + \, 4{\text{ H}}_{{2}} {\text{O}}_{{\left( {\text{g}} \right){ }}} $$

(18)

$$ {\text{FeO}}_{{\text{(s)}}} + {\text{ H}}_{{\text{2(g)}}} + \, 0.125{\text{ N}}_{{\text{2(g)}}} = \, 0.25{\text{ Fe}}_{{4}} {\text{N}}_{{\text{(s)}}} + {\text{ H}}_{{2}} {\text{O}}_{{\text{(g)}}} $$

(19)

The change in the Gibbs free energy of each reaction was calculated. The formation of iron nitride through reaction Eq. 17 begins at 810 °C. However, hematite may no longer present at this temperature as the reduction reaction of hematite with hydrogen to form metallic iron is more spontaneous, which starts from 550 °C. It was calculated that the reactions Eq. 18 and Eq. 19 were unlikely to occur because the ΔG for each reaction was positive at all temperatures. The Gibbs free energy of iron nitride formation from magnetite ores through the ammonia decomposition mechanism was also computed at various gas pressure conditions as shown in Fig. 5. Iron nitrides were predicted to not form through iron oxide reduction with nitrogen generated from ammonia decomposition. Pressure shifts from 1 atm to 0.1 and 10 atm had no effect on the reaction thermodynamic feasibility.

Fig. 5

The Gibbs free energy of reaction with temperature of nitride formation from magnetite ores under different pressures (0.1, 1, and 10 atm) showing positive values at all conditions

Another calculation was carried out to investigate the possibility of iron nitride formation through iron nitriding process under an ammonia atmosphere. From some related studies, the common iron nitrides that can be formed are the phases Fe₂N, Fe₃N, and Fe₄N [38,39,40]. The reaction can be expressed by the following equations:

$$ 8{\text{Fe }} + \, 2{\text{NH}}_{3} = \, 2{\text{Fe}}_{4} {\text{N }} + \, 3{\text{H}}_{2} ; \Delta H_{298} = \, 69.6{\text{ kJ}}/{\text{mol}};\,{\Delta }G < 0{\text{ at }}T = 370\,^\circ {\text{C}} $$

(20)

$$ 8{\text{Fe }} + \, 4{\text{NH}}_{{3}} = \, 4{\text{Fe}}_{2} {\text{N }} + \, 6{\text{H}}_{2} ; \Delta H_{298} = \, 168.5{\text{ kJ}}/{\text{mol}};\,\Delta G < 0{\text{ at }}T = 470\,^\circ {\text{C}} $$

(21)

The formation of iron nitride in the form of Fe₄N is likely to exist compared to the formation of Fe₂N compound. Reactions Eq. 20 and Eq. 21 take place at temperatures of 370 °C and 470 °C, respectively. It was found that Fe₄N can be formed when Fe and NH₃ co-exist. However, this condition is practically not available during reduction of iron oxide using ammonia as the Fe formation from reduction of iron ore occurs simultaneously with the gas decomposition. Therefore, decomposition of NH₃ must be suppressed and Fe formation enhanced at the same time, which can be achieved only at extreme condition, with a pressure of 10⁷ atm and temperatures ranging from 300 to 600 °C. There may be some kinetics reasons why the formation of nitride is favored over the metal formation. However, this is not yet established in this study. Iron nitride is thermodynamically an unstable compound at room temperature, and it will decompose as the ΔG for decomposition reaction (Eq. 22) is always negative. The phase change of iron nitride decomposition per 100 g of Fe₄N during heating is shown in Fig. 6.

$$ 2{\text{Fe}}_{4} {\text{N }} \to \, 8{\text{Fe }} + {\text{ N}}_{2} ;\,\Delta H_{298} = \, 4.67{\text{ kJ}}/{\text{mol}};\,\Delta G_{298} = \, - \,1.56{\text{ kJ}} $$

(22)

Fig. 6

Equilibrium calculation of iron nitride decomposition at 1 atm (calculated using data from FactSage 8.2) showing that Fe₄N is unstable at all temperatures forming N₂ and Fe

Thermodynamic Calculation Results

Ammonia Decomposition

Ammonia decomposition is the reverse reaction of ammonia synthesis. The reaction is a typical endothermic process with an enthalpy change of 45.9 kJ/mol. This reaction commences at 185 °C as the ΔG is negative. The decomposition reaction of ammonia can be presented by the following reaction:

$$ {\text{NH}}_{{\text{3(g)}}} \to \, \raise.5ex\hbox{$\scriptstyle 1$}\kern-.1em/ \kern-.15em\lower.25ex\hbox{$\scriptstyle 2$} {\text{ N}}_{{\text{2(g)}}} + \, 3/2{\text{ H}}_{{\text{2(g)}}} ; \,{\Delta }H_{298} = \, 45.9{\text{ kJ/mol}} $$

(23)

The FactSage Equilib module was used to evaluate the stable gas phases under different temperature conditions during the ammonia decomposition process. This calculation was run between 100 and 500 °C at 20 °C intervals under 0.01, 0.1, 0.5, 1, 2, and 3 atm and the results are plotted in Fig. 7. The equilibrium conversion of NH₃ can reach 99% at 400 °C and ambient pressure. From this figure, it can also be concluded that the decomposition of ammonia is thermodynamically favorable at low pressures as 99% of ammonia conversion can be achieved at 180 °C in 0.01 atm and 260 °C in 0.1 atm, whereas it can be achieved at temperatures 460 °C and 480 °C in 2 and 3 atm, respectively.

Fig. 7

Equilibrium conversion of ammonia decomposition at different pressures (calculated using data from FactSage 8.2)

The gas profile during ammonia decomposition at 1 atm is shown in Fig. 8. This graph was plotted using gas composition data from the FactSage Equilib module calculation during heating using a temperature range of 0–500 °C, following Eq. 23. Results showed that ammonia gas gradually decomposed during heating, resulting in increasing hydrogen and nitrogen gas. One mol of NH₃ generated 1.5 mol hydrogen and 0.5 mol nitrogen. The decomposition proceeds linearly from 0 to 300 °C then becomes flattened from 460 °C, predicting that decomposition reactions have been completed.

Fig. 8

Gas profile during decomposition of ammonia at 1 atm (calculated using FactSage 8.2)

Phase Stability Diagrams and Reaction Mechanisms

Evolution in process methods used in the iron and steel industry is required within the upcoming decades due to increasing environmental challenges. Hydrogen ironmaking is currently the favored option for carbon replacement to reduce greenhouse gases (GHG) emissions. The reduction of iron ores by hydrogen produces water vapor as a by-product instead of carbon dioxide when a carbon-based reductant is utilized. The reaction sequences of iron oxide reduction with hydrogen depend on the reduction temperature [41,42,43]. Hematite (Fe₂O₃) is not directly reduced to metallic iron and if the reduction temperature is less than 570 °C, reduction to Fe occurs by a two-step mechanism, sequentially from hematite (Fe₂O₃) to magnetite (Fe₃O₄) and then with continued reduction from magnetite to Fe without any wűstite (FeO) formation as an intermediate oxide. Otherwise, if the reduction temperature is higher than 570 °C, the reduction proceeds by a three-step mechanism from hematite to magnetite, magnetite to wűstite, and then wűstite to Fe according to the following reactions [43]:

$$ {\text{If }}T \, < \, 570\,^\circ {\text{C}}: $$

$$ 3{\text{Fe}}_{{2}} {\text{O}}_{{\text{3(s)}}} + {\text{ H}}_{{\text{2(g)}}} = \, 2{\text{Fe}}_{{3}} {\text{O}}_{{\text{4(s)}}} + {\text{ H}}_{{2}} {\text{O}}_{{\text{(g)}}} {,}\,\,K_{{\text{A}}} = \frac{{{\text{PH}}_{2} {\text{O}}}}{{{\text{PH}}_{2} }} $$

(24)

$$ 0.5{\text{Fe}}_{{3}} {\text{O}}_{{\text{4(s)}}} + {\text{ H}}_{{\text{2(g)}}} = \, 0.75{\text{Fe}}_{{\text{(s)}}} + {\text{ H}}_{{2}} {\text{O}}_{{\text{(g)}}} ,\,K_{{\text{B}}} = \frac{{{\text{PH}}_{2} {\text{O}}}}{{{\text{PH}}_{2} }} $$

(25)

$$ {\text{If }}T \, > \, 570\,^\circ {\text{C}}: $$

$$ 3{\text{Fe}}_{{2}} {\text{O}}_{{\text{3(s)}}} + {\text{ H}}_{{\text{2(g)}}} = \, 2{\text{Fe}}_{{3}} {\text{O}}_{{\text{4(s)}}} + {\text{ H}}_{{2}} {\text{O}}_{{\text{(g)}}} ,\,K_{{\text{A}}} = \frac{{{\text{PH}}_{2} {\text{O}}}}{{{\text{PH}}_{2} }} $$

(26)

$$ 1.202{\text{Fe}}_{{3}} {\text{O}}_{{\text{4(s)}}} + {\text{ H}}_{{\text{2(g)}}} = \, 3.808{\text{Fe}}_{{{0}{\text{.947(s)}}}} + {\text{ H}}_{{2}} {\text{O}}_{{\text{(g)}}} ,\,K_{{{\text{B1}}}} = \frac{{{\text{PH}}_{2} {\text{O}}}}{{{\text{PH}}_{2} }} $$

(27)

$$ {\text{Fe}}_{0.947} {\text{O}}_{{\text{(s)}}} + {\text{ H}}_{{\text{2(g)}}} = \, 0.947{\text{Fe}}_{{\text{(s)}}} + {\text{ H}}_{{2}} {\text{O}}_{{\text{(g)}}} ,\,K_{{{\text{B2}}}} = \frac{{{\text{PH}}_{{2}} {\text{O}}}}{{{\text{PH}}_{2} }}, $$

(28)

where K is the equilibrium constant, PH₂O is the partial pressure of steam, and PH₂ is the partial pressure of hydrogen as reductant. By using this equation together with Gibbs free energy of each reaction at certain temperatures, the stability diagram can be constructed as shown by Fig. 9. This figure shows the stability areas of iron-iron oxide phases in Fe–O–H₂ system at different temperatures and ratios of PH₂/PH₂O. The condition of high PH₂ will be more favorable for iron oxides reduction. The reduction of iron oxides to iron using hydrogen as reductant is preferable at the highest possible temperatures because the stability area of iron increases with rising temperature. Therefore, energy must be added to the system to ensure a constant reduction temperature when reducing with hydrogen because of the endothermic reaction.

Fig. 9

Stability diagram of Fe–O–H₂ system (calculated using data from FactSage 8.2)

Hosokai et al. (2011) performed thermodynamic calculation using HSC Chemistry v.5.11 (Outotec) to examine if there was any advantage of hematite iron ore reduction using ammonia. Their calculations were reproduced here by using FactSage’s Reaction module and the results are presented in Fig. 10. Direct ammonia reduction has a thermodynamic advantage because the reduction of hematite with ammonia promotes a greater equilibrium conversion to metallic iron compared to hydrogen reduction at temperature above 200 °C, as the change of Gibbs free energy of ammonia reduction is thermodynamically more favorable than for hydrogen at the same temperature [30]. As explained previously, ammonia commences decomposing at 185 °C under ambient pressure and achieves 99% of equilibrium conversion at 400 °C at 1 atm. The hydrogen generated from ammonia decomposition can reduce the iron oxide in this circumstance. The direct formation of metallic iron from Fe₂O₃ in H₂ reducing conditions occurs at temperature above 550 °C which is beyond the decomposition temperature of ammonia. On the other hand, direct reduction of hematite with NH₃ was predicted to start at temperatures above 300 °C, but is thermodynamically no longer possible above 400 °C at 1 atm as the ammonia is fully decomposed at this condition. Heat must be supplied to conduct this process due to the positive enthalpy reaction.

Fig. 10

The Gibbs free energy of hematite reduction with hydrogen and ammonia, as well as ammonia decomposition at standard state (calculated using data from FactSage 8.2)

A predominance area diagram for the Fe–O–NH₃ system was constructed based on equilibrium data from FactSage as shown in Fig. 11. This diagram depicts the stability areas of Fe, FeO, Fe₃O₄, and Fe₃O₄ phases depending on the temperature and partial pressure of reducing NH₃ gas. It can be seen from Fig. 11 that Fe₄N was predicted to only form when the partial pressure of NH₃ is very high. The predominance diagram was constructed using the equations listed below:

$$ \frac{9}{2}{\text{Fe}}_{2} {\text{O}}_{{3\left( {\text{s}} \right)}} + {\text{NH}}_{{3\left( {\text{g}} \right)}} = 3{\text{Fe}}_{3} {\text{O}}_{{4\left( {\text{s}} \right)}} + \frac{1}{2}{\text{N}}_{{2\left( {\text{g}} \right)}} + \frac{3}{2}{\text{H}}_{2} {\text{O}}_{{\left( {\text{g}} \right)}} ;{\mkern 1mu} \Delta H_{{298}} = - {\mkern 1mu} 19{\mkern 1mu} {\text{kJ}}/{\text{mol}} $$

(29)

$$\frac{3}{8}{\text{Fe}}_{3} {\text{O}}_{{4\left( {\text{s}} \right)}} + {\text{NH}}_{{3\left( {\text{g}} \right)}} = \frac{9}{8}{\text{Fe}}_{{\left( {\text{s}} \right)}} + \frac{1}{2}{\text{N}}_{{2\left( {\text{g}} \right)}} + \frac{{12}}{8}{\text{H}}_{2} {\text{O}}_{{\left( {\text{g}} \right)}} ;{\mkern 1mu} \Delta H_{{298}} = {\mkern 1mu} 97{\mkern 1mu} {\text{kJ}}/{\text{mol}} $$

(30)

$$ \frac{3}{2}{\text{Fe}}_{3} {\text{O}}_{{4\left( {\text{s}} \right)}} + {\text{NH}}_{{3\left( {\text{g}} \right)}} = \frac{9}{2}{\text{FeO}}_{{\left( {\text{s}} \right)}} + \frac{1}{2}{\text{N}}_{{2\left( {\text{g}} \right)}} + \frac{3}{2}{\text{H}}_{2} {\text{O}}_{{\left( {\text{g}} \right)}} ;{\mkern 1mu} \Delta H_{{298}} = 37.1{\mkern 1mu} {\text{kJ}}/{\text{mol}} $$

(31)

$$ \frac{3}{2}{\text{FeO}}_{{\left( {\text{s}} \right)}} + {\text{NH}}_{{3\left( {\text{g}} \right)}} = \frac{3}{2}{\text{Fe}}_{{\left( {\text{s}} \right)}} + \frac{1}{2}{\text{N}}_{{2\left( {\text{g}} \right)}} + \frac{3}{2}{\text{H}}_{2} {\text{O}}_{{\left( {\text{g}} \right)}} {\mkern 1mu} ;{\mkern 1mu} \Delta H_{{298}} = 15.9{\mkern 1mu} {\text{kJ}}/{\text{mol}} $$

(32)

$$ 4{\text{Fe}}_{{\text{(s)}}} + {\text{ NH}}_{{\text{3(g)}}} = {\text{ Fe}}_{{4}} {\text{N}}_{{\text{(s)}}} + \, 1.5{\text{H}}_{2} ; \, \Delta H_{298} = 15.9{\text{ kJ}}/{\text{mol}} $$

(33)

Fig. 11

Predominance area diagram of Fe–O–NH₃ system (calculated using data from FactSage 8.2)

The change in Gibbs energy becomes

$$ \Delta G = \Delta G^{{\text{o}}} + {\text{RT}}\,\ln \,\frac{{\left( {P_{{N_{2} }} } \right)^{m} \left( {P_{{H_{2} O}} } \right)}}{{P_{{NH_{3} }} }}^{n} $$

(34)

where P_N2, P_H2O, and P_NH3 are relative partial pressures of the gases in the product and reactant mixture. However, the competing reaction with H₂ is neglected in this calculation for simplification. At chemical equilibrium, ΔG = 0, therefore,

$$ \Delta G^{{\text{o}}} = - RT\,\ln \,\frac{{\left( {P_{{{\text{N}}_{{2}} }} } \right)^{m} \left( {P_{{{\text{H}}_{{2}} {\text{O}}}} } \right)^{n} }}{{P_{{{\text{NH}}_{{3}} }} }} = - {\text{RT}}\,\ln K $$

(35)

$$ \log P_{{{\text{NH}}_{{3}} }} = \frac{{\Delta G^{{\text{o}}} }}{{2.303 {\text{RT}}}} - \log \frac{1}{{(P_{{{\text{N}}_{{2}} }} )^{m} (P_{{{\text{H}}_{{2}} {\text{O}}}} )^{n} }} $$

(36)

Effect of Mole Ratio on Product Phase Composition

Calculations were carried out to study the effect of using different mole ratios of Fe₂O₃/NH₃ on product phase composition at equilibrium using the Equilib module. The calculations were run using 1/2, 1/4, 1/8, and 1/20 mol ratios of Fe₂O₃/NH₃ at 1 atm. The results of the solid and gas phase composition change at equilibrium during reduction with different mole ratios at different temperatures are shown in Fig. 12. At 1/2 mol ratio, hematite is completely reduced to Fe₃O₄ through direct reduction with ammonia up to 400 °C. The total amount of Fe atoms in Fe₃O₄ (0.67 mol) is equal to the former amount of Fe₂O₃ (1 mol). Reduction with H₂ is assumed to take place at temperatures higher than 400 °C as H₂ declined from that point. It is assumed that reduction of some Fe₃O₄ to Fe occurs at 400 °C by H₂ generated as reductant. Fe₃O₄ is stable up to 570° and then reduced with H₂ generated to FeO. Iron metallic (α-Fe or Fe bcc) begins to appear at temperatures higher than 670 °C produced from further reduction of FeO with H₂. Furthermore, at temperatures higher than 912 °C, metallic iron will be stable as γ-Fe or Fe-FCC.

Fig. 12

Phase composition of iron reduction with ammonia using Fe₂O₃/NH₃ ratios of 1/2, 1/4, 1/8, and 1/20 [left: gas phase composition; right: solid phase composition] (calculated using FactSage 8.2)

A thermodynamic calculation to study the effect of NH₃ gas composition on the metal recovery is shown in Fig. 13. The simulation was carried out using ammonia concentrations ranging from 5 to 100% at 400, 600, and 800 °C, respectively. It shows that at the lowest gas composition (5% of ammonia), the metal recovery was 14.8% at 400 °C, while it reached 44.2% and 71.3% at 600° and 800 °C, respectively. Complete metallization at 400, 600, and 800 °C was attained at 25, 8.5, and 6.5% of ammonia gas concentration. At 400 °C, the metallization was significantly enhanced at gas concentrations higher than 10%, ranging from 41.9% using 10% of NH₃ to 69. 2% and 96.5% under 15% and 20% of NH₃.

Fig. 13

Effect of ammonia gas composition to Fe metal recovery at 400, 600, and 800 °C and 1 atm (calculated using data from FactSage 8.2)

At 1/4 mol ratio, the reduction shows similar reaction behavior with the 1/2 mol ratio despite the higher Fe and lower FeO composition of the final product. Fe-metallization achieved 0.14 mol at a mole ratio of 1/2, while 0.54 mol of Fe was attained at 1/4 of mol ratio and a temperature of 500 °C. A significant amount Fe was generated at 1000 °C (1.55 mol of Fe) at a 1/4 of mol ratio compared to 1/2 mol ratio (0.23 mol of Fe) at the same temperature. This result provides a possible explanation as to why the portion of FeO decreases as the NH₃ portion increases. At 1/8 mol ratio, only 0.07 mol of FeO was generated at 600 °C, compared to 1/2 and 1/4 at which the reduction produced 1.74 and 1.3 of FeO, respectively. The gas composition shows that the reduction of hematite using 1/2, 1/4, and 1/8 mol ratios of Fe₂O₃/NH₃ takes place under the same mechanism. The increase of H₂ mole indicates that the reduction occurs via ammonia reduction up to 300 to 400 °C of temperature depending on the mole ratio, and subsequent reaction is assumed to occur via hydrogen reduction generated from ammonia decomposition. Reduction of 1 mol hematite at a high NH₃ concentration (20 mol) showed different results. Fe₃O₄ is stable up to 400 °C, while Fe (bcc) is formed at temperatures higher than 220 °C and becomes the dominant phase from 300 °C. Hematite is directly reduced by ammonia to Fe₃O₄ and Fe up to 300 °C and subsequent reduction is dominated by reduction with hydrogen. The FeO phase is not produced at these conditions. To summarize, higher amounts of ammonia promote greater reduction of Fe₂O₃, which simultaneously lowers the temperature required for reduction. Additionally, higher concentration of ammonia suppresses the conversion of Fe₃O₄ to FeO which could be beneficial as the conversion of FeO to Fe is generally considered a rate-limiting step in iron oxide reduction [44].

Effect of Pressure on Product Phase Composition

Investigations were carried out using the Equilib module to examine the impact of pressure on product phase composition resulting from iron oxide reduction with ammonia. The assessment was computed by varying the pressure condition from 10^–3, 10^–2, 10^–1, 1, 2, 3, 5, and 10 atm using 1 mol Fe₂O₃ and 2 mol NH₃ as the reactant species. The results are shown in Figs. 14 and 15. These figures show that pressure has no direct effect on solid phase composition as the quantity of each phase is similar at all pressure conditions. It has been explained previously in “Ammonia Decomposition” Section that ammonia is thermodynamically favored to decompose at low pressure. At extremely low pressure, 10^–3 atm, NH₃ is stable as the decomposed species, hydrogen, and nitrogen, as the amount of NH₃ is nearly zero at all temperatures. The reduction of Fe₂O₃ is predicted to commence at 400 °C through hydrogen reduction mechanism, indicated by the increasing of H₂O simultaneously with the decreasing of H₂. These figures also show that pressure has no direct effect on solid phase composition as the quantity of each phase is similar at all pressure conditions.

Fig. 14

Gas phase composition of Iron reduction with ammonia under 10^–3, 10^–2, 10^–1, 1, 2, 3, 5, and 10 atm of pressure (calculated using data from FactSage 8.2)

Fig. 15

Solid phase composition of Iron reduction with ammonia under 10^–3, 10^–2, 10^–1, 1, 2, 3, 5, and 10 atm of pressure (calculated using data from FactSage 8.2)

Effect of Iron Ore Type on Product Phase Composition

Calculations have been conducted to investigate the possibility of using the ammonia as reductant for different types of iron ores, such as magnetite and goethite. The reduction reaction for each iron ore type is as follows:

$$ {\text{Magnetite}}:\,{\text{Fe}}_{{3}} {\text{O}}_{{{4}\left( {\text{s}} \right)}} + 8/3\,{\text{NH}}_{{{3}\left( {\text{g}} \right)}} = 3{\text{Fe}}_{{\left( {\text{s}} \right)}} + 4/3\,{\text{N}}_{{{2}\left( {\text{g}} \right)}} + 4{\text{H}}_{{2}} {\text{O}}_{{\left( {\text{g}} \right)}} $$

(37)

$$ {\text{Goethite}}: {\text{Fe}}_{{2}} {\text{O}}_{{3}} \left( {{\text{H}}_{{2}} {\text{O}}} \right)_{{\left( {\text{s}} \right)}} + \, 2{\text{NH}}_{{\text{3(g)}}} = \, 2{\text{Fe}}_{{\left( {\text{s}} \right)}} + {\text{ N}}_{2} + \, 4{\text{H}}_{{2}} {\text{O}}_{{\text{(g)}}} $$

(38)

The changes in Gibbs free energy of each reaction at various temperatures and at standard state are shown in Fig. 16. Reactions Eq. 37 and Eq. 38 start at 380 and 240 °C, respectively, according to ∆G data obtained from calculations using Reaction module. It can be concluded that magnetite and goethite show similar solid phase products, i.e., spinel, FeO, and Fe. Magnetite is predicted to reduce to Fe-BCC directly with ammonia until 400° and to be continuously reduced with hydrogen produced from ammonia decomposition. The reduction of Fe₃O₄ is completed at 500 °C as magnetite phase is no longer present at temperatures higher than 500 °C. The reduction is shifted to FeO being the most stable and dominant phase at a temperature of 570 °C (2.33 mol of FeO at 1200 °C), while Fe-FCC is only formed at a minimum temperature of 1000 °C but gradually increases with further temperature rise (0.39 mol of Fe-FCC at 1200 °C). The utilization of goethite as the primary source of iron oxide for ammonia reduction shows higher Fe-metallization as the Fe-FCC phase can reach up to 1 mol at 1200°.

Fig. 16

Phase composition of magnetite and goethite reduction with ammonia; left: solid phase composition; right: gas composition (calculated using data from FactSage 8.2)

Conclusion

Direct reduction of iron oxide with ammonia offers a novel process for sustainable ironmaking at low temperatures and low carbon emissions. Previous studies indicated the reduction of iron oxide with ammonia is potentially possible at temperatures as low as 600 °C. The formation of iron nitrides occurs during the process; however, these are not expected be an issue as they decompose at temperatures higher than 700°C. Iron nitride formation occurrence may be beneficial in the case of DRI handling due to its oxidation resistance.

Thermodynamic assessment of ammonia reduction of iron oxides was performed using FactSage 8.2 thermochemical software and databases and results compared with previous studies. Ammonia decomposition reaction is thermodynamically desirable at low pressures as 99% of ammonia conversion can be performed at 180 °C in 0.01 atm and at 400 °C (both under ambient pressure conditions). Similar to hydrogen ironmaking, reduction of hematite with ammonia was found to occur in two steps at temperatures lower than 570 °C and three steps when temperatures are over 570 °C. The reduction sequence with ammonia is Fe₂O₃ → Fe₃O₄ → Fe below 570 °C and Fe₂O₃ → Fe₃O₄ → FeO → Fe above 570 °C. The last reaction suggested to occur with generated hydrogen from ammonia decomposition as ammonia has fully decomposed at this temperature. The ratio of Fe₂O₃/NH₃ is predicted to affect the reduction reaction of hematite with ammonia. Complete metallization was attained at 400, 600, and 800 °C at 25, 8.5, and 6.5% of ammonia gas concentration. The greater amount of ammonia promotes the greater reduction degree of Fe₂O₃, which simultaneously lowers the starting temperature needed for reduction. An excess ammonia concentration hinders the conversion of Fe₃O₄ to FeO, which is anticipated to be advantageous due to prevent the FeO to Fe conversion which is the rate-limiting step. Pressure has no direct impact on the thermodynamics composition of solid phases in reduction of iron oxide with ammonia. Thermodynamically, Fe₄N is unlikely to form directly from hematite reduction. It can be formed when Fe and NH₃ co-exist. However, this condition is not available as the Fe formation occurs simultaneously with decomposition of NH₃ according to previous assessment. Therefore, decomposition of NH₃ must be suppressed and enhanced Fe formation at the same time, which only occurs under extreme conditions at pressures of 10⁷ atm and temperatures ranging from 300 to 600 °C.

The use of ammonia to reduce iron oxides during ironmaking may open a new opportunity to decarbonize steel industries by using renewable energy and green ammonia, especially without any CO₂ and NO_x GHG emissions produced throughout the process. At the same time, it can solve the supply chain and energy issue in transporting and storing hydrogen as iron metallization of both hydrogen and ammonia-based reduction comparably showed similar degrees. This work has only provided a detailed thermodynamic assessment of the feasibility of the process and needs to be followed up by additional research, such as kinetics studies, reaction mechanisms, and optimization of reduction variables, including operating parameters and feedstock compositions. Utilization of different types of reactors with certain heat integration will need to be investigated in order to develop reactor configurations and suitable scale-up. Stoichiometric calculations show that DRI is reduced by NH₃ at a minimum consumption rate of 450 kg NH₃/ton of DRI for 95% of metallisation degree. Therefore, experiment testing and techno-economic analysis of this process should also be considered.