Hydrogenation of Inorganic Metal Carbonates: A Review on Its Potential for Carbon Dioxide Utilization and Emission Reduction.

Carbonaceous minerals represent a valuable and abundant resource. Their exploitation is based on decarboxylation at elevated temperature and under oxidizing conditions, which inevitably release carbon dioxide into the atmosphere. Hydrogenation of inorganic metal carbonates opens up a new pathway for processing several metal carbonates. Preliminary experimental studies revealed significant advantages over conventional isolation technologies. Under a reducing hydrogen atmosphere, the temperature of decarboxylation is significantly lower. Carbon dioxide is not directly released into the atmosphere, but may be reduced to carbon monoxide, methane, and higher hydrocarbons, which adds value to the overall process. Apart from metal oxides in different oxidation states, metals in their elemental form may also be obtained if transition-metal carbonates are processed under a hydrogen atmosphere. This review summarizes the most important findings and fields of the application of metal carbonate hydrogenation to elucidate the need for a detailed investigation into optimized process conditions for large-scale applications.

Introduction

Carbon dioxide (CO2) is an abundant chemical species. On earth, it is present in all three aggregate states: in gaseous form in the atmosphere (approximately 2.5×1012 tons), in the dissolved state in the hydrosphere (approximately 1016 tons), and in the solid state fixed in carbonate rocks (approximately 1014 tons).1 Hence, more (by a factor of 104) terrestrial carbon is fixed in carbonate rocks in the earth's crust than is actually present in the gaseous atmosphere. Carbonate rocks have been known as a valuable and abundant resource for a long time. The history of inorganic metal carbonate chemistry dates back to the early ages of solid‐state chemistry. The thermal decomposition, meaning decarboxylation reactions of metal carbonates in an oxidizing atmosphere, also known as calcination, roasting, or burning, depending on the metal species involved, represent processes that were developed first on experience, later on research, and finally on technological optimization in industrial production plants. Cato, for instance, mentioned the burning of limestone (calcite, CaCO3) in kilns in 184 B.C. In the 1700s, Joseph Black gave the first systematic technical explanation of the calcination of limestone, also including the evolution of gaseous carbon dioxide.2 Lime (CaO) is one of the cheapest and most widely used alkalizing chemicals and the main constituent of cement production. The production of magnesium oxide (magnesia, MgO) through calcination of magnesite (MgCO3) for refractory manufacture was established over 100 years ago in Austria.3 The roasting of iron carbonate (siderite, FeCO3) is used for iron and steel production in areas with vast mineral iron carbonate reserves, such as Austria4 and China.5, 6 The cement industry, iron production, and manufacture of magnesia sinter make use of large quantities of the respective metal carbonates. Further industrial applications of metal carbonates of, for instance, nickel, cobalt, manganese, and zinc, or more precisely of their respective oxides or the metals in their elemental form, include the production of catalysts (catalytic material and porous support), pigments and glass, the ceramics industries, and the electronics industries.3, 7, 8

During the decarboxylation of metal carbonates to yield the respective metal oxide (MeO), inevitably one mole of CO2 and/or carbon monoxide (CO) evolve per mole of metal oxide formed. The metal oxide, in turn, gives access to retransformation into the corresponding carbonate through CO2 uptake. This characteristic provides the basis for carbon capture and storage (CCS) technologies, referred to as mineral carbonation, in which CO2 is fixed to and stored as carbonate minerals, mainly in the form of calcium and magnesium carbonate.9

Whereas the decarboxylation products of main‐group elements (e.g., alkaline‐earth‐metal carbonates) are the corresponding metal oxides and CO2 [Eq. (1)], the decomposition of transition‐metal carbonates follows a more complex reaction pathway [Eq. (2)] because redox processes may take place. The redox behavior of the transitions metals allows the reduction of CO2 to CO by means of thermodynamic fundamentals. The solid products can be metal oxides and mixtures of metal oxides that adopt different oxidation states.10

In metal carbonate decarboxylation, the reaction conditions, especially the nature of the gas atmosphere, play a crucial role in the course of the reaction. If carried out in a reducing atmosphere with hydrogen (H2), a fascinating reaction network is observed. Equation (3) shows the reaction for the hydrogenation of alkaline‐earth‐metal carbonates, whereas Equation (4) gives a potential reaction pathway for the hydrogenation of transition‐metal carbonates. Apart from metal oxides and mixtures of metal oxides in different oxidation states, metals in their elemental form can be formed as solid products from transition‐metal carbonates. Gaseous products may include methane (CH4) in addition to or in place of CO2 and CO.10 Some citations even report the formation of higher hydrocarbons (C1–C).11, 12, 13 In this context, it is noteworthy that an admixture of iron oxides enables the direct conversion of calcium carbonate into C1–C3 hydrocarbons.13

Transition metals are well established catalysts in Fischer–Tropsch (FT) synthesis,14, 15, 16 and the water‐gas shift reaction.17, 18, 19, 20 FT synthesis, namely, the catalytic polymerization and hydrogenation of CO, gives access to the synthesis of hydrocarbons (alkanes, alkenes, and oxygenated hydrocarbons) from syngas, CO and H2.21 The reaction for the synthesis of alkanes is shown in Equation (5) as an example. The reverse water‐gas shift reaction yields CO from CO2 through reduction with hydrogen [Eq. (6)]. Consequently, the presence of transition metals in metal carbonate hydrogenation opens up a pathway for the synthesis of hydrocarbons, instead of simply releasing CO2 into the gas phase.

In addition to different gaseous products that are released from metal carbonates under reducing conditions, changing the gaseous atmosphere from inert to reducing also has an effect on the morphology of the solid products.22 In most cases, the reducing agent in reductive metal carbonate decarboxylation is hydrogen. Until now, the production of hydrogen on an industrial scale has mainly (>95 %) been based on fossil fuels, for instance, through steam reforming of methane or gasification of coal and hydrocarbons, which, apart from hydrogen, generates CO2.23, 24 Apart from conventional production routes, hydrogen can also be produced renewably and sustainably from various sources (e.g., water electrolysis or water splitting through photocatalysis,25, 26 solar thermal production,27 photosynthesis by algae,28 reforming of biomass29). Hydrogen storage is challenging, however, and requires a completely new distribution system.23

Whereas the decomposition of various metal carbonates in a vacuum or under an oxidizing or inert atmosphere has been well investigated and reported (e.g., for FeCO3 in vacuum,30 oxygen,31, 32, 33 and nitrogen34, 35, 36), information on metal carbonate hydrogenation is scarce. Reduction of carbonates in aqueous solution is also well described in the literature and is not covered herein.37, 38, 39, 40, 41, 42 Metal carbonate hydrogenation is not only challenging from a reaction mechanism point of view, but also features some characteristics that render it promising in terms of carbon capture and utilization (CCU), hydrogen storage, and novel production technologies in metallurgy. The purpose herein is to provide a comprehensive literature review on metal carbonate hydrogenation and to discuss potential applications. After describing their natural occurrence (Section 2), thermodynamics of metal carbonate hydrogenation are specified (Section 3). Experimental studies that have been conducted so far are listed in Section 4. After summing up the main benefits of metal carbonate hydrogenation (Section 5), feasible process options are illustrated and evaluated (Section 6). Finally, Section 7 identifies future needs for detailed research.

Natural Occurrence

Inorganic carbonates, characterized by planar [CO3]2− complexes with metal ions, form one of the most important mineral groups. Taxonomy is based on cations and/or anions outside of the [CO3]2− complexes. Water‐free mineral carbonates may feature calcite‐ (trigonal, scalenohedric), dolomite‐ (trigonal, rhombohedral), and aragonite‐type (orthorhombic) structures. Carbonates with additional anions may contain OH− anions or water.

Main‐group elements

Alkaline metal carbonates

Mineral lithium carbonate (Li2CO3, zabuyelite) occurs in the lithosphere as an ore companion, mainly imbedded in halite in rock salt and in saline lakes. Sodium carbonate (Na2CO3, natrite, also known as soda) appears in saline lakes and minerals such as trona Na3(CO3)(HCO3)⋅2 H2O and natron Na2CO3⋅10 H2O. It undergoes a rapid change superficially in air to thermonatrite (Na2CO⋅H2O). Potassium carbonate (K2CO3) is called potash because it was historically produced by treating wood ash with water in a pot. There are no K2CO3‐containing ores known that would be worth mining. It is generally produced from electrolytically produced KOH and CO2. Rubidium (Rb2CO3) and cesium (Cs2CO3) carbonate are naturally found together in low concentrations accompanying sodium‐ and potassium‐containing ores. Similar to the other alkaline metals, they also collect in saline lakes. Francium isotopes (Fr2CO3) are radioactive and only found in traces in the lithosphere.43, 44

Alkaline‐earth‐metal carbonates

Beryllium carbonate (BeCO3) barely exists in minerals because it is easily decomposed into beryllium oxide and CO2, and therefore, only stable in a CO2 atmosphere. It exists naturally in the mineral niveolinite NaBe(CO3)2(OH)⋅2 H2O. Magnesium carbonate (MgCO3) is found in the lithosphere as magnesite MgCO3 and dolomite CaMg(CO3)2, such as in the Southern Alps—the Dolomites. Mining of magnesite is one of the main resources for magnesium carbonate. Calcium carbonate (CaCO3) mainly forms calcite CaCO3 and dolomite CaMg(CO3)2. Mining of these minerals is the main source for calcium carbonate. Three different naturally occurring crystal structures of calcium carbonate exist: calcite, aragonite, and vaterite. In calcite, the central Ca2+ ion is coordinated by six oxygen atoms, whereas in aragonite nine oxygen atoms coordinate the central Ca2+ ion. Strontium carbonate (SrCO3) is found in the naturally occurring mineral strontianite. Barium carbonate (BaCO3) forms in the lithosphere as witherite.43, 44

Further main‐group metals

Lead carbonate (PbCO3) is found as cerussite. Caustic aluminum carbonate is found in the mineral dawnsonite NaAl(CO3)(OH)2.43, 44

Transition metals

Manganese is present in the lithosphere nearly as frequently as carbon or phosphorus. Manganese carbonate (MnCO3) is found in the mineral rhodochrosite and as an ore companion of iron. Iron carbonate (FeCO3) is found in the lithosphere as siderite and ankerite (Ca(Mg,Fe)[CO3]2), for example, at the Erzberg in Styria, Austria, and in China. Cobalt occurs in diverse forms in the lithosphere, often as an ore companion. Nevertheless, no major cobalt carbonate (CoCO3)‐containing ore is known. Cobalt carbonate can be produced by precipitation of water‐soluble cobalt(II)salts with alkaline‐earth carbonates. Nickel exists in diverse forms in the lithosphere, but no nickel carbonate ores are known. The industrially most important nickel carbonate is caustic nickel carbonate, 2 NiCO3 ⋅3 Ni(OH)2 ⋅4 H2O, produced by precipitation of aqueous nickel sulfate with sodium carbonate. Caustic nickel carbonate can be dehydrated to give anhydrous nickel carbonate or the hexahydrate NiCO3 ⋅6 H2O. Copper carbonate (CuCO3) occurs naturally as azurite 2 CuCO3 ⋅Cu(OH)2 (blue) and malachite CuCO3⋅Cu(OH)2 (green). Silver carbonate (Ag2(CO3)3) is not found in minerals, but precipitates from water by using soluble silver species and alkaline‐metal carbonates, such as soda. Zinc carbonate (ZnCO3) exists as smithsonite. Cadmium carbonate (CdCO3) is mostly found as an ore companion of smithsonite.43, 44

Thermodynamics of Metal Carbonate Hydrogenation

Thermodynamic analysis of alkaline, alkaline‐earth, and transition‐metal carbonates between 400 and 1200 K at ambient pressure shows increasing standard free energies of reaction, ΔG R 0, for methane formation with increasing temperature. Due to strongly positive ΔG R 0 values (>60 kJ mol−1), methane formation is not possible through hydrogenation of alkaline‐metal carbonates (Figure 1 a). Among alkaline‐earth‐metal carbonates, only hydrogenation of MgCO3 features a negative ΔG R 0, favoring CH4 formation (Figure 1 b). Hydrogenation of the transition‐metal carbonates MnCO3, FeCO3, CoCO3, NiCO3, CuCO3, and ZnCO3 to their respective bivalent oxides, CH4, and H2O is favorable (Figure 1 c).

Figure 1

Standard free energies of reaction, ΔG R 0, for methane formation through hydrogenation of metal carbonates between 400 and 1200 K at ambient pressure calculated with HSC Chemistry 8 software; a) alkaline‐metal carbonates: Me2CO3+4 H2↔Me2O+CH4+2 H2O, b) alkaline‐earth‐metal carbonates, and c) transition‐metal carbonates: MeCO3+4 H2↔MeO+CH4+2 H2O.

Figure 2 compares conventional beneficiation of FeCO3 under oxidizing conditions with hydrogenation of FeCO3. Both routes are thermodynamically favorable in the temperature range examined (400–1200 K), although reduction of hematite (Fe2O3) features significantly higher standard free energies of reaction.

Figure 2

Standard free energies of reaction, ΔG R 0, for decarboxylation of FeCO3 relative to hydrogenation of Fe2O3 between 400 and 1200 K at ambient pressure calculated with HSC Chemistry 8 software.

Hydrogenation of the transition‐metal carbonates and MgCO3 to yield CH4 exhibit pronounced exothermic behavior (ΔH R 0<−50 kJ mol−1).

Experimental Studies

Experimental studies on the hydrogenation of metal carbonates are scarce. Although first reports date back to the late 1960s, to the best of our knowledge, there are only 20 scientific publications from six research groups. Interestingly, hardly any cross reference exists between the studies.

Giardini and Salotti from the University of Georgia, USA, were the first to report on reactions between mineral calcite, dolomite, and siderite with pressurized hydrogen and the concomitant formation of hydrocarbons.11, 45, 46 The primary purpose of their study was to address geological issues, such as the formation of hydrocarbons from inorganic sources occurring in the earth's crust. A patent was filed on this topic.47 The experimental apparatus consisted of a 25 cm3 externally heated “cold seal”‐type vessel, in which the carbonate charge was enclosed in a platinum foil, and left unscaled and suspended in the heated part of the vessel. Before usage, the minerals were handpicked for impurities, crushed, and heated in 30 % hydrogen. The natural mineral deposit and the composition of the minerals were not stated.

From 1987, Reller and co‐workers from the University of Zürich, Switzerland, investigated the thermal reactivity of pure alkaline‐earth‐metal carbonates, 3d transition‐metal carbonates, and metal‐doped alkaline‐earth‐metal carbonates in pure and dilute hydrogen by thermogravimetric (TG) experiments.10, 48, 49

Two groups from Japan—one from the Tokyo Institute of Technology;50 the other from Kobe University51—investigated metal carbonate hydrogenation in the 1990s. Tsuneto et al. used NiCO3 and CoCO3 without additional catalysts and a series of metal carbonates doped with catalytically active metals, and hydrogenated them in a fixed‐bed flow reactor at atmospheric pressure.50 Formation of gaseous products was reported in μmol h−1 at a total gas flow of 9.6 cm3 min−1 and feed samples of 2 g after 0.5–1 h. To estimate the reaction rate, we calculated the hourly conversion of the respective carbonate to CH4 (CtM in % h−1), as depicted from Equation (7). Data listed in the table coincided only partially with the data given in the main text of their paper.50 In case of discrepancies, data listed tabularly were used herein.

Yoshida et al. focused on CH4 formation from metal‐catalyzed CaCO3 hydrogenation, in comparison to hydrogenation of pure CaCO3.51 Experiments were carried out by means of the temperature‐programmed hydrogenation technique, and for isothermal kinetic runs two apparatuses were applied: a Cahn electrobalance and a closed circulation apparatus.

In 2009 and 2013, Jagadeesan et al. from the Jawaharlal Nehru Centre for Advanced Scientific Research, India, investigated CaCO3 and mixed‐metal/CaCO3 hydrogenation promoted with catalytically active metallic nanoparticles in a continuous‐flow, packed‐bed, stainless‐steel reactor and managed to directly convert the inorganic carbonates into C1–C3 hydrocarbons.13, 52

The effect of reaction temperature, pressure, and gas atmosphere on the reaction kinetics of mineral FeCO3, MgCO3, and CaMg(CO3)2 hydrogenation has recently become subject of detailed investigation by Baldauf‐Sommerbauer et al. from Graz University of Technology, Austria.12, 53, 54 Experiments were carried out in a thermobalance and a tubular reactor setup.

A detailed list of the abovementioned studies, starting with the most recent ones, is given in Table 1. The investigations can be divided into two groups: 1) hydrogenation of single‐metal carbonates without additional catalysts, applied either as mineral ore or in synthetic form; 2) mixed‐metal/metal carbonates, mainly synthetic, in which one of the metals, mostly transition metals, acts as an internal catalyst. In general, synthetic carbonates and mixed‐metal/metal carbonates were produced by precipitation and coprecipitation from aqueous solution of NaHCO3.

Table 1

Papers on metal carbonate hydrogenation dating back to the 1960s until the present (2018), ranked in chronological order.

Material component	Additional	Feed gas	Experimental	Sample		Flow rate	T	P	Ref.	Year
	catalyst	comp.	apparatus	size	mass [g]	[cm3 min−1]	[K]	[MPa]
mineral MgCO3/CaMg(CO3)2 [a]	no	H2 (90 vol %)N2 (10 vol %)	tubular reactor (T316 SS, d=2.5 cm, l=80 cm)	5–8 mm	121	500	768–808	amb0.3, 0.8 (o.p)	54	2017
mineral MgCO3 [b]	no	H2 (90 vol %)N2 (10 vol %)	tubular reactor (T316 SS, d=2.5 cm, l=80 cm)	5–8 mm	115	500	748, 763, 778	amb0.3, 0.81.2 (o.p.)	99	2016
mineral FeCO3 [c]	no	H2 (70 vol %)N2 (30 vol %)	thermobalance (HR=1.8,3, 5, 10 K min−1)	100–200 μm	0.02	100	<1023	amb	12	2016
mineral FeCO3 [c]	no	H2 (90 vol %)N2 (10 vol %)	tubular reactor (T316 SS, d=2.5 cm, l=80 cm)	0.5–1 mm	60	867	623, 648	amb	96	2016
CaCO3	Fe	H2	continuous‐flow, packed‐bed reactor (SS)	n.a.	0.04	3	573–873	amb	13	2013
MeCa(CO3)2 (Me=Co, Ni, Fe;Me/Ca=1:1); Me1Me2Ca(CO3)3 (Me1Me2=CoNi, NiFe, FeCo;Me1/Me2/Ca=1:1:2)	no/yes, NPs of Co/CaO/CoO,Ni/CaO, CoNi/CaO/CoOFe/CaO/Fe3O4, FeCo/CaO/CoO,NiFe/CaO	H2	tubular reactor (SS, d=0.6 cm, l=30 cm)	n.a.	0.05	3.5–8	823	amb	52	2009
CaCO3	no/yes, Pd, Ir (5 wt %)	H2 (<0.027 MPa)	temp.‐programmedhydrogenation (HR=3 K min−1),Cahn electrobalance(3.62 cm3), closed circul.apparatus (275 cm3)	n.a.	0.015, 0.1, 570	60	573–698	amb	51	1999
NiCO3 ⋅Ni(OH)2 ⋅4 H2O,CoCO3, Li2CO3, Na2CO3, K2CO3,MgCO3⋅Mg(OH)2 ⋅5 H2O,BaCO3, CaCO3	no/yes; Ni, Fe, Co, Pd, Pt, Cu(2 wt %)	H2 (50 vol %)He (50 vol %)	fixed bed flow reactor	n.a.	2	9.6	473–673	amb	50	1992
MnCO3, FeCO3, CoCO3, NiCO3,ZnCO3, caustic 3d transition‐metal carbonates (Cu, Ni, Zn,Cu/Ni, Cu/Co, Cu/Zn, Ni/Zn),MeMg(CO3)2 (Me=Fe, Co, Ni,Cu, Zn), Me1Me2Mg(CO3)2 (Me1Me2=CuZn, Ni/Zn, Cu/Ni)	no, influence of type of energy	H2 (5 % H2 in Ar)	TG/MS (HR=10 K min−1)		≈15×10−3	30	<900	amb	22	1992
Mg5(OH)2(CO3)4 ⋅4 H2O	no/yes, Ni	H2	TG/DTA, TG/MS	n.a.	8.2×10−3	30	<900	amb	56	1991
MgCO3, CaCO3, SrCO3, BaCO3,FeCO3	no/yes; Ni, Ru, Rh for CaCO3	H2 (0.1 MPa)	TG/MS (HR=10 K min−1)	n.a.	≈15×10−3	30	<1200	amb	10, 55	19911989
CaCO3	Fe, Ni, Co, Cu, Ru, Rh, Pd, Ag	H2	TG/MS	n.a	10–12×10−3	30	<950	amb	49, 55	19901989
mineral MgCO3,[d] mineral CaMg(CO3)2,[e] mineral CaCO3 [f]	Co, Ni, Cu (10 %) as(Mg,Me)(CO3)2, (Ca,Me)(CO3)2	H2 (0.1 MPa)	TG/MS (HR=10 K min−1)	n.a.	8.35–10.99×10−3	n.a.	<1000	amb	48, 55	19871989
mineral CaCO3,mineral CaMg(CO3)2,mineral FeCO3	metallic Ni, Pt, Cu, Ti, Mg, Fe;mixtures of Pd, Pt, Rh on Al2O3 and dried silica gel; activatedAl2O3, Fe2O3, Fe3O4, Cr2O3, CrO3,Kieselguhr mixtures; hydrousand anhydrous oxides,pyrolytic carbon	H2 (1.4–55 MPa)	25 cm3 cold seal, SS vessel	40–60 mesh	n.a.	DA	618–1143	amb	11, 46, 47	196819691971
mineral CaCO3	no	H2 (0.7–80 MPa)	25 cm3 cold seal, SS vessel	n.a.	n.a.	DA	693–1243	amb	45	1968

[a] Breitenau, Austria. [b] Eskişehir, Turkey. [c] Erzberg, Austria. [d] Ural, Russia. [e] Binn Valley, Switzerland. [f] Gonzen mine, Switzerland. amb=ambient pressure, o.p.=overpressure, n.a.=not available, TG=thermogravimetric experiments, MS=mass spectrometry, DTA=differential thermal analysis, SS=stainless steel, HR=heating rate, NPs=nanoparticles, DA=discontinuous operation.

Single‐metal carbonates

Hydrogenation of single‐metal carbonates includes the alkaline‐earth‐metal carbonates MgCO3, CaCO3, SrCO3, BaCO3, and CaMg(CO3)2 and the 3d transition‐metal carbonates MnCO3, FeCO3, CoCO3, NiCO3, and ZnCO3. Decomposition of the respective carbonates in a reducing hydrogen atmosphere occurs at lower temperatures than that of decomposition under inert or oxidizing conditions.

Alkaline‐earth metals

MgCO3, CaCO3, BaCO3, and SrCO3

Reller et al.48 and Padeste55 performed TG measurements with finely ground MgCO3 and CaCO3 mineral, and synthetic SrCO3 and BaCO3 at atmospheric pressure. Equal amounts of CO2 and CO, together with H2O, were formed as gaseous products from MgCO3. In the case of CaCO3, the main volatile compound was CO (CO/CO2≈10:1) from which a change in the degradation mechanism was concluded. The concomitant formation of CO, in addition to CO2 and H2O, was dedicated to the reduction of CO2 through the reverse water‐gas shift reaction [Eq. (6)]. The amount of CO increases with increasing atomic mass of the alkaline‐earth‐metal cation due to higher decarboxylation temperatures (T <800 K, T ≈900 K) and the endothermic nature of the reverse water‐gas shift reaction, which shifts the equilibrium composition towards the product CO at higher temperatures. The reaction temperature for decarboxylation in hydrogen was lowered by at least 150 K compared with the analogous reaction in a nitrogen atmosphere. In a hydrogen atmosphere, all four alkaline‐earth‐metal carbonates fully degraded below temperatures of 1200 K into their respective oxides MgO, CaO, SrO, and BaO. They formed as solid conglomerates of microcrystalline domains with diameters of 10–20 nm and showed a pronounced reactivity towards the respective hydroxides and towards recarbonation.48 If synthetic CaCO3 crystals were degraded in nitrogen and hydrogen atmospheres, distinct destruction in nitrogen occurred, which suggested that diffusion of H2 into the carbonate, where it directly reacted with fixed CO2, and reverse diffusion of formed CO and H2O proceeded faster than that of CO2 diffusion.10, 48

Kinetic studies with CaCO3 were performed by Yoshida et al. in a closed circulation apparatus at a fixed H2 pressure of 0.13×105 Pa and 748 K.51 The reaction was of half‐order with respect to H2 with an activation energy of 236 kJ mol−1. Initial reaction rates did not differ for varying CaCO3 amounts, which was explained by decomposition of CaCO3 and hydrogenation of released CO2 to CO. The lowest temperature at which hydrogenation occurred was 700 K. At temperatures of 700–773 K, only CO formed. Above 773 K, CO and CO2 formed. The formation of the two gaseous products differed over the course of time. Whereas CO pressure steadily increased with time, CO2 pressure abruptly reached a maximum value of 22.7 Pa and did not change subsequently.

Baldauf‐Sommerbauer et al. examined the effect of reaction temperature (748–778 K) and pressure (ambient to 1.2 MPa overpressure) on the reductive calcination of magnesite (5–8 mm) in a fixed‐bed tubular reactor with 70 % hydrogen and nitrogen.53 MgCO3 conversion rose with increasing reaction temperature (e.g., from 45 % at 748 K to 96 % at 778 K). Under isothermal conditions, the conversion decreased with increasing pressure (e.g., at 763 K from 76 % at ambient pressure to 67 % at 1.2 MPa overpressure). In contrast to the findings of Reller et al.,48 CH4 was found in addition to CO2, CO, and H2O as gaseous products [Eq. (8)].

Low temperature and elevated pressure facilitated CH4 formation. Moderate to high temperature and low pressure facilitated CO formation. The CH4 yield was 38.6 % after 20 % MgCO3 conversion at 748 K and 1.2 MPa overpressure. With respect to the reaction mechanism, decreasing CH4 formation with increasing magnesite conversion indicated a dependency on the amount of MgCO3. Baldauf‐Sommerbauer et al. related the increase of CO concentration with rising MgCO3 conversion to the amount of MgO. To scrutinize this interpretation, they examined reductively calcined MgO for its catalytic properties for CO2 conversion with H2. At ambient pressure and 0.3 MPa overpressure, only CO formed. At 0.8 MPa overpressure, traces of CH4 occurred in addition to the major product CO. These findings revealed significant reverse water‐gas shift activity of reductively calcined MgO, but no CH4 formation. CH4 formation during reductive calcination of magnesite seemingly proceeds through a different mechanism.

Giardini and Salotti reported the formation of CH4 and ethane (C2H6) through heating (693–1243 K) of calcite and dolomite under a pressurized hydrogen atmosphere (0.7–80 MPa H2); however, it was not comprehensible whether the pure minerals were used or if the reaction was catalytically accelerated.11, 45, 46, 47 In their main publication,11 they stated that metallic Ni, Pt, Cu, Ti, Mg, and Fe; commercial mixtures of 0.5 % Pd, Pt, Rh on alumina and dried silica gel; activated alumina; hematite, magnetite; chromic oxide; chromium trioxide; and Kieselguhr mixtures were added, but no precise information was given on the type of catalyst for individual experimental data. The catalyst admixture was neither mentioned in the first publication46 nor in the patent application.47 Apparently, none of the catalytically active materials had a discernible effect on the rate of reaction. The reaction kinetics were thus expected to fit the uncatalyzed hydrogenation of calcite. Due to high initial hydrogen concentrations relative to calcite, the reaction kinetics simplified to pseudo‐first‐order kinetics, with an activation energy of 75 kJ mol−1 at 14 MPa.45 Hydrogenation of calcite started at 773 K. It was primarily dependent on temperature and secondarily dependent on pressure and time and proceeded in a crystallographically anisotropic manner. As solid products, CaO, Ca(OH)2, graphite (C), and a black residue, possibly solid hydrocarbons, formed. At higher temperatures, CaO was the principal solid product. CH4 and H2O were ubiquitous gaseous products, whereas C2H6 and CO appeared under certain specific conditions. CO2 was never detected in remarkable amounts (>0.01 %). CO2, if formed, immediately converted into CH4, CO, and H2O. The maximum CH4 content (on a dry basis) was 3.2 % (and 96.75 % H2) at 988 K, 61 MPa, and 16 h. The reaction was dependent on the type of surface because powdery CaCO3 reacted faster than a single rhomb of equal weight.

Mg5(OH)2(CO3)4⋅4 H2O

Hydrogenation of hydromagnesite [Mg5(OH)2(CO3)4⋅4 H2O] to MgO occurred in two steps and finished at temperatures lower than 750 K. The first step corresponded to loss of water. In the second step, CO2 released from Mg5(OH)2(CO3)4 and yielded MgO as the final solid product.56

CaMg(CO3)2

The effect of reaction temperature (793 K–1108 K) and initial hydrogen pressure (14–34 MPa) on the hydrogenation of dolomite (40–60 mesh) was investigated by Giardini and Salotti.46 As for calcite, there was no clear information on the catalyst admixture. We assume that the main findings apply to the hydrogenation of dolomite without additional catalysts. Solid products included CaCO3, Ca(OH)2, CaO, noncrystalline Mg(OH)2, graphite, and a “soot‐like” material; gaseous products CH4, C2H6, CO, and CO2. H2O was the oxygenated product in all experiments. A two‐step reaction mechanism was proposed in which noncrystalline Mg(OH)2 or MgO formed [Eq. (9)]. At 34 MPa, the reaction started at 793 K.

Reller et al. found that the decarboxylation temperature of dolomite and the ratio of the gaseous products CO and CO2 lay between the corresponding values of the two pure carbonates CaCO3 and MgCO3.48

Baldauf‐Sommerbauer et al. suggested the hydrogenation of mixed magnesite/dolomite (1:1 mol/mol) for the synthesis of CO.54 TG measurements showed two decomposition steps. Compared with the reaction in nitrogen, a hydrogen atmosphere leads to a decrease of the decomposition temperature of 60 K for the first step and 100 K for the second step. In the first step, the decomposition of concomitant MgCO3, FeCO3, and MnCO3 to the respective bivalent oxides and CO2 was observed. Then CaMg(CO3)2 decomposed into CaO, MgO, and CO2. Reductive calcination experiments in a tubular reactor indicated a sequential mechanism of calcination followed by hydrogenation of CO2. CH4 was only formed in traces, even at elevated pressure. It is assumed that CH4 formation was kinetically hindered. A CO yield of 61–73 % was achieved for partial reductive calcination of the magnesite content below 813 K. An increase of pressure did not affect the formation of CO, but caused a slight retardation of the reaction.

Transition‐metal carbonates

The minerals MnCO3 (rhodochrosite) and FeCO3 (siderite), and synthesized CoCO3, NiCO3, and ZnCO3 were investigated. Synthesis requires hydrothermal conditions at high pressure. According to the size of their cations, the 3d transition‐metal carbonates crystallize in a trigonal calcite‐type structure.22, 57

As postulated by Emmenegger22 from hydrogenation experiments in pure and dilute hydrogen (5 % H2 in Ar), the partial hydrogen pressure decisively influences the hydrogenation reaction. Different solid products—transition‐metal oxides and elemental transition metals arise—depending on the selected atmosphere, and morphological features may be controlled. In hydrogen, decarboxylation temperatures drop, in comparison to the respective reaction under an inert atmosphere. Dilute hydrogen results in a lower temperature drop. The degradation temperature drops with decreasing radii of the transition‐metal cations: Mn2+ (0.8 Å)>Fe2+ (0.76 Å)>Co2+ (0.74 Å)>Ni2+ (0.72 Å)>Mg2+ (0.65 Å). The ratio of gaseous products CO2, CO, CH4, higher hydrocarbons, and H2O varies, depending on the transition‐metal species. According to Reller et al., CO2 is released during transition‐metal carbonate hydrogenation, and further catalytically hydrogenated.10 In the course of decarboxylation, the catalysts Me or Me/MeO form in situ. Fe, Co, and Ni act as efficient hydrogenation catalysts.

MnCO3

In pure and dilute (5 % H2 in Ar) hydrogen, the mineral MnCO3 decarboxylated to MnO without changing its oxidation number at 643 K. As gaseous products, traces of unconverted CO2, CO, and H2O formed. Consequently, manganese efficiently catalyzed the reverse water‐gas shift reaction. In dilute hydrogen, this reaction was retarded. After a period of induction, released CO2 was further converted into CO and H2O. Compared with pure hydrogen, the chemical equilibrium barely lay on the side of the product.22

FeCO3

Salotti and Giardini first studied the hydrogenation of siderite (40–60 mesh) at reaction temperatures of 618 to 878 K and initial partial hydrogen pressures of 1.4 to 34 MPa (H2).46 Because there was no precise information on the catalyst admixtures reported, we assumed that the main findings applied to siderite only. At temperatures above 728 K, the solid products consisted of elemental Fe and wüstite (FeO). At lower temperatures (673 K and 14 MPa H2), magnetite (Fe3O4) formed. From minute and rare flecks, Giardini and Salotti assumed that also graphite was formed.11 Wüstite was the primary alteration product [Eq. (10)]. Further reduction of wüstite to elemental Fe [Eq. (11)] or oxidation of wüstite to magnetite [Eq. (12)] depended on the reaction temperature and the ultimate H2O/H2 ratio. A low reaction temperature and dry hydrogen need to be provided to effect reduction.

CH4 and H2O were ubiquitous gaseous products, whereas CO, CO2, and the higher hydrocarbons ethane (C2H6), propane (C3H6), and butane (C4H10) were present over a limited temperature and pressure range. An inverse relationship existed between the temperature and the length of the hydrocarbon chain, meaning that a lower initial reaction temperature results in a more complex hydrocarbon species. At 673 K (14 MPa H2), the gaseous product on a dry basis contained 4.45 mol % CH4, 0.28 mol % ethane, 0.01 mol % propane, and 0.03 mol % butane. At 728 K (1.4 MPa H2), the concentration of CH4 slightly decreased (4.34 mol %), whereas the concentrations of higher hydrocarbons increased (0.42 mol % ethane, 0.23 mol % propane, 0.05 mol % butane). At 878 K and 14 MPa H2, only CH4 was detected. CO2 only formed at 798 K (34 MPa H2). Giardini and Salotti explained this by the thermal stability of the hydrocarbons.11 The reaction temperature for siderite hydrogenation is low enough to ensure thermal stability of ethane, propane, and butane. They concluded that higher hydrocarbons formed directly through a reaction on the mineral surface, rather than in a subsequent reaction between released gases and hydrogen.

This conclusion contradicts the findings of Reller et al., who explained the formation of CO and CH4 by in situ formation of catalytically active transition‐metal species.10 Emmenegger compared the hydrogenation of siderite in pure and dilute hydrogen (5 % H2 in Ar).22 In pure hydrogen, the formation of mainly elemental Fe together with FeO as solid products was observed. At elevated temperature above 823 K, decomposition was slow and not yet finished at 973 K. The gaseous products CH4, H2O, and CO were formed. Higher hydrocarbons were not found. CO2 was also released. The reaction pathways in Equations (13), (14) were postulated for siderite hydrogenation. Because Fe was not the main product, according to Equation (14) reduction occurred only partially. The reverse water‐gas shift reaction for the reduction of CO2 to CO was reported.

In dilute hydrogen, CH4 did not form. The major amount of CO2 released from siderite [Eq. (13)] was not converted, but formed the main product of the product gas. Reduction of intermediate FeO was incomplete and only minor amounts of Fe were formed. In addition to the distinct difference in product composition between conversion in pure and dilute hydrogen atmospheres, solid products also differed in morphology. In pure hydrogen, the solid product contained a high amount of crystalline parts, whereas the product in dilute hydrogen featured a higher amount of fine particles. The conversion temperature significantly reduced in pure hydrogen (603 K), with respect to an inert helium atmosphere (653 K). In dilute hydrogen, decarboxylation started at 623 K.

Baldauf‐Sommerbauer et al. suggested the hydrogenation of siderite as a means for sustainable iron production.12 They performed kinetic computations to study the reaction kinetics of iron formation, and considered the concomitant decomposition of the accessory matrix carbonates of calcium, magnesium, and manganese The original mineral consisted of three main carbonate components of siderite FeCO3 with partial Mg and Mn substitution, ankerite (CaFeMgMn)CO3, and dolomite CaMg(CO3)2. Potassium, aluminum, and silicon existed in the form of muscovite KAl2(AlSi3O10)(OH)2, whereas a major part of the silicon was quartz (SiO2). A concentrated siderite specimen (size fraction 100–200 μm) was used for kinetic analysis in a thermobalance. During conversion under a hydrogen atmosphere, the FeCO3 content of mineral siderite was converted into elemental Fe [(79±2) wt %; Eq. (15)]. From calcium, magnesium, and manganese carbonates, the respective oxides formed [Eq. (16)].

The model‐free kinetic analysis, according to the Ozawa–Flynn–Wall,58, 59 Kissinger–Akahira–Sunose,60 and Friedman61 approaches, gave a parallel kinetic model. With multivariate nonlinear regression, the kinetic parameters (E a=151.8 kJ mol−1, log 10(A)=8.751 s−1) were determined. A two‐dimensional Avrami–Erofeev model A2 was applicable for the conversion of FeCO3 into Fe. Therefore, a temperature‐controlled nucleation and diffusional growth mechanism was suggested. For the concomitant formation of CaO, MgO, and MnO, multiparameter autocatalysis models were used without applying multistep kinetics. At 723 K, more than 95 % conversion within less than 60 min reaction time was observed. Increasing temperature leads to a faster reaction, but a lower yield of elemental Fe.

CoCO3 and NiCO3

Hydrogenation of synthetic CoCO3 and NiCO3 was investigated by Emmenegger22 and Tsuneto et al.50 Tsuneto et al. hydrogenated commercial transition‐metal powders of CoCO3 and NiCO3⋅Ni(OH)2⋅4 H2O in a fixed‐bed flow reactor at atmospheric pressure and 473 K.50 In both cases, CH4 formed after a period of induction, in which only CO2 evolved through thermal decomposition. In the case of CoCO3, a maximum methane formation rate of 97 μmol h−1 was achieved after 55 h. With NiCO3, the methane formation rate reached a maximum value (72 μmol h−1) after 7 h and hydrogen (12 mmol h−1) was consumed completely. At 523 K, CH4 and CO2 formed promptly. The formation of metallic Co and Ni, together with the period of induction for methane formation, suggests that Ni and Co act as hydrogenation catalysts for CO2.

Emmenegger performed TG experiments with synthetic CoCO3 and NiCO3 in helium and pure and dilute hydrogen (5 % H2 in Ar).22 The temperature at which carbonate degradation started dropped from 603 K in helium to 543 K in pure hydrogen for CoCO3. For NiCO3, a similar temperature dependency of conversion, with 623 K in helium, 548 K in dilute hydrogen, and 513 K in pure hydrogen, was observed. Under a hydrogen atmosphere (pure and dilute), elemental Co and Ni formed as solid products. Similar behavior between CoCO3 and NiCO3 was also visible, in terms of the gaseous product stream. The main gaseous products CH4 and H2O formed. The byproducts CO2 and barely any CO were detected. The formation of gaseous products proceeded simultaneously. CO2 fully converted after a certain period of induction, which indicated that cobalt and nickel catalyzed the hydrogenation of CO2 into CH4. Conversion differed with respect to the rate of reaction in dilute hydrogen: at the beginning, the conversion of NiCO3 was slower than that of CoCO3. In both cases, the catalytic activity for CO2 hydrogenation dropped drastically and the main gaseous products were CO2 and H2O [Eqs. (13) and (14) for CoCO3 and NiCO3] and only minor amounts of CO. These reactions seemed to occur simultaneously because CO2 and H2O formed concurrently. From CoCO3 conversion, barely any CH4 formed, but still significant amounts of CH4 arose from NiCO3. Consequently, Ni was confirmed to be a more efficient catalyst in a dilute hydrogen atmosphere than Co.

ZnCO3

Contrary to other 3d transition‐metal carbonates, Emmenegger found that decomposition temperatures of ZnCO3 increased from 583 K in pure hydrogen to 628 K in dilute hydrogen (5 % H2 in Ar).22 Due to a lower hydrogen concentration, the equilibrium composition of the reverse water‐gas shift reaction preferably consists of H2 and CO2. Zinc did not show any catalytic activity for the reduction of evolved CO2. The composition of the gaseous product mixture (CO2, CO, and H2O) resembled that expected for the water‐gas equilibrium.

Caustic 3d transition‐metal carbonates

Emmenegger investigated the hydrogenation of caustic 3d transition‐metal carbonates of the type Me(OH)(CO3)⋅z H2O (Me=Cu, Ni, Zn), combinations thereof (Cu/Ni, Cu/Co, Cu/Zn, Ni/Zn), and mixed caustic transition‐metal/Mg carbonates (MeMg(CO3)2 with Me=Fe, Co, Ni, Cu, Zn; Me1Me2Mg(CO3)2 with Me1Me2=CuZn, Ni/Zn, Cu/Ni).22 Decomposition temperatures were lower than those of the respective neutral carbonates. Loss of crystal water and condensation of OH− with caustic carbonates was observed. Higher water concentration shifts the equilibrium composition of potential CO2 hydrogenation to the reactant side. Consequently, all caustic metal carbonates show low catalytic hydrogenation activity compared with that of the respective neutral carbonates. Mixed transition‐metal/Mg carbonate systems show pronounced catalytic activity, which highlights the effect of the noncatalytic support material MgO. MgO forms fine particles that give access to high dispersion of the transition metals or alloys.22

Main‐group metal carbonates combined with transition metals

An admixture of transition metals to main‐group metal carbonates opens up a new pathway in metal carbonate hydrogenation: because many transition metals catalyze hydrogenation reactions, CO2 evolved from the carbonate is converted into CO, CH4, or higher hydrocarbons CH and CHO. The catalytically active transition‐metal species is formed in situ during the decomposition reaction. A wide range of transition metals (Fe, Ni, Co, Cu from the 3d group; Ru, Rh, Pd, Ir, Ag from the 4d group) was used for doping of various main‐group metal carbonates, such as Li2CO3, Na2CO3, K2CO3, CaCO3, MgCO3, SrCO3, 4 MgCO3⋅Mg(OH)2⋅5 H2O, and BaCO3. Mixed alkaline‐earth‐metal/transition‐metal carbonates of the type MeCa(CO3)2 and Me1Me2(CO3)3 were also investigated.

In general, two promising effects take place. First, the decarboxylation temperature of the alkaline and alkaline‐earth‐metal carbonates drops. Second, different gaseous compounds evolve during decomposition due to the catalytic activity of the transition‐metal species. The product composition depends on the transition metal in the carbonate.

Alkaline‐metal carbonates

Tsuneto et al. hydrogenated nickel‐doped (2 wt %) Li2CO3, Na2CO3, and K2CO3 at 400 K and atmospheric pressure.50 Nickel powder was mechanically mixed with the carbonate sample. CO2 was not detected after 1 h from Li2CO3 and Na2CO3 decomposition, according to thermodynamics. Traces of CO2 were detected from K2CO3 (<2 μmol h−1 at a total gas flow rate of 9.6 cm3 min−1). With all three carbonates, minor amounts of CH4 were formed (Li2CO3: 1.3 μmol h−1, CtM=0.005 %, Na2CO3: 2.5 μmol h−1, CtM=0.013 %, K2CO3: 1.8 μmol h−1, CtM=0.012 %).

Hydrogenation of calcite and dolomite minerals at elevated temperatures (calcite: 693–1143 K, dolomite: 793–1108 K) and elevated pressure (calcite: 1.4–55 MPa H2, dolomite: 14–34 MPa H2) was investigated by Giardini and Salotti.11 In their report, it is ambiguous to which experimental runs metallic Ni, Pt, Cu, Ti, Mg, and Fe, or commercial mixtures of 0.5 % Pd, Pt, and Rh on alumina and dried silica gel, activated alumina, hematite, magnetite, chromic oxide, chromium trioxide, Kieselguhr mixtures, hydrous and anhydrous oxides, and pyrolytic carbon were added as potential catalysts. None of the catalytically active materials had a discernible effect on the rate of reaction. The metals Pt, Fe, and Ni catalytically promoted pyrolytic dissociation of formed CH4 [Eq. (17)]. For calcite, they reported the formation of CaO [Eq. (18)]; Ca(OH)2 [Eq. (19)]; graphite; and, at temperatures above 973 K, carbon soot‐like material, which was seemingly amorphous carbon formed through thermal dissociation of CH4 [Eq. (17)]. Gaseous products were CH4, C2H6, CO, CO2, and H2O. Below the dehydration temperature, Ca(OH)2 was the stable solid product.

They assumed that CH4, and its higher homologues, if thermally stable under the reaction conditions, formed directly through methanation of calcite, rather than through reactions between H2, CO2, and CO. CO2 and CO were only detected at low pressure (1.4 MPa) and high reaction temperature (973 K).

Reller et al. investigated the effects of Co, Ni, and Cu doping (10 %) on the hydrogenation of MgCO3 and CaCO3.48 Mixed alkaline‐earth‐metal/transition‐metal carbonates, Mg−Me and Ca−Me carbonates, were prepared by coprecipitation with sodium carbonate from the respective nitrate solutions. In a second study, the effect of coprecipitated Ni, Ru, and Rh was reported.10 The decarboxylation temperatures dropped in the range of 200 K to, at most, 400 K in the case of Ni, compared with the decarboxylation of the pure carbonates in a nonreducing atmosphere. Mixtures of CO and CO2 were released if Cu was used as a coprecipitate. In the case of Co, predominantly CH4 formed together with minor amounts of CO. The formation of CO2 was negligible. With Ni on MgCO3 and CaCO3, more than 90 % of the gaseous product was CH4. The formation of CHO species from CaCO3 in H2 was not detected at atmospheric pressure (p=0.1 MPa). The solid products consisted of a mixture of microcrystalline alkaline‐earth‐metal oxides and elemental transition metals. Reller et al. confirmed that the formed solid products acted as effective catalysts for the partial reduction of CO2 to CO or direct conversion of CO2 into CH4. For Ni−Ca carbonate systems, the activity of the system was explained by the high dispersion of the catalytically active transition metal in the CaCO3 matrix.48 Consequently, the combination of transition‐metal carbonates with alkaline‐earth‐metal carbonates improves the catalytic activity of the in situ formed transition‐metal species if an appropriate dispersion or active surface area is generated during decarboxylation; an effect that cannot be accomplished with pure transition‐metal carbonates only.

Padeste et al. published a detailed study comparing the influence of the 3d transition metals Fe, Ni, Co, and Cu and the 4d transition metals Ru, Rh, Pd, and Ag on the thermal decomposition of CaCO3 in hydrogen.49 The mixed‐metal carbonates were produced by coprecipitation. Whereas 3d metal carbonates, except for FeCO3, normally precipitate as caustic carbonates (hydroxocarbonates) from aqueous solutions, ions of the 4d metals Rh, Ru, and Pd precipitate as oxides or oxide hydrates and Ag predominantly forms the simple carbonate Ag2CO3. Samples from NiCO3/CaCO3 and CoCO3/CaCO3 showed that two‐phase systems formed, even at transition‐metal carbonate concentrations of 5 %, rather than replacing large amounts of Ca2+ by transition metals with similar ionic radii. TG measurements showed that thermal decomposition in hydrogen proceeded in two steps: First, decomposition and reduction of the transition‐metal carbonate below temperatures of 600–700 K with H2O evolution. Second, CaCO3 decomposition at temperatures above 600–700 K. H2O evolution might result from loss of coprecipitated water, decomposition of hydroxides to oxides [Eq. (20)], reduction of CO2 [Eqs. (21), (22)], and reduction of the metal oxide formed [Eq. (23)].

The reaction in Equation (23) was observed for all transition‐metal carbonates. CO formation [Eq. (21)] did not occur in this temperature range and CH4 formation [Eq. (22)] was only observed for Ni and Co. During the CaCO3 decomposition step, CO2 [Eq. (24)], CO [Eq. (25)], and CH4 [Eq. (26)] formed as volatile carbon products.

Most CO2 reduces to CO and CH4. A correlation between the decomposition temperature and the distribution of the volatile products based on the thermodynamics of the different reactions was stated: carbonates that evolve predominantly CH4 decompose at the lowest temperature, whereas carbonates that release CO as the main gaseous compound have the highest decomposition temperature, which resembles the behavior of pure CaCO3. Whereas Fe, Cu, and Ag showed little influence (T decomposition=730–880 K, CH4>1 %, CO>90 %) on the thermal decomposition of CaCO3 in hydrogen, Co and Pd had a medium effect (T decomposition=680–850 K, 20 %<CH4<70 %), and Ni, Ru, and Rh had a pronounced effect (T decomposition=620–780 K, CH4>95 %).

Hydrogenation of 4 MgCO3⋅Mg(OH)2⋅5 H2O (at 573 K), CaCO3 (at 473, 573 and 673 K), and BaCO3 (at 673 K) in the presence of Ni powder in a fixed‐bed reactor was compared by Tsuneto et al.50 MgCO3 was readily hydrogenated to form a considerable amount of CH4 (1600 μmol h−1) with an hourly conversion to CH4 of 38.9 % and minor amounts of CO2 (290 μmol h−1). Contrary to MgCO3, no CO2 was released from CaCO3 and BaCO3 under the conditions investigated. CH4 formed in minor amounts from BaCO3 (1.8 μmol h−1, CtM: 0.018 %) and higher amounts from CaCO3 (95 μmol h−1, CtM: 0.475 %) at 673 K. The effect of reaction temperature was also investigated for Ni‐doped CaCO3. Although CH4 formation was not suppressed at temperatures as low as 473 K, the rate of CH4 formation decreased with decreasing temperature (473 K: 3.7 μmol h−1, CtM: 0.019 %; 573 K: 8.5 μmol h−1, CtM: 0.053 %; 673 K: 95 μmol h−1, CtM: 0.475 %). To connote, a remarkable shift of 400–500 K towards lower hydrogenation temperatures occurred in comparison to undoped CaCO3 (1172 K). Similar to the work of Padeste et al.,49 Tsuneto et al.50 investigated the effect of the transition metals Fe, Co, Ni, Pd, Pt, and Cu on CaCO3 hydrogenation at 673 K. The order of the catalytic activity was Ni>Co>Pt>Fe>Cu>Pd. This sequence confirms the results of Padeste et al.,49 apart from Pd, which was attributed medium influence in the former study. Long‐term hydrogenation was investigated for Ni‐doped CaCO3 at 673 K. The CH4 formation rate was constant for 7 days. Within 15 days, a total of 95 % of CaCO3 was converted. As solid products, CaO and Ca(OH)2 formed according to Equations (26) and (27).

Because the rate of CH4 formation was greater than that of CO2 evolution, Tsuneto et al. concluded that direct CaCO3 hydrogenation occurred on the CaCO3 surface by hydrogen spillover from a Ni surface, rather than thermal decomposition of CaCO3 followed by hydrogenation of released CO2.50

Yoshida et al. studied methane formation through the hydrogenation of CaCO3 catalyzed by Pd and Ir (5 wt %) over a temperature range of 573–698 K.51 Temperature‐programmed hydrogenation data revealed high‐temperature tails and lower temperatures for the start of decomposition. These findings were explained by an increase of available reactive hydrogen at low temperature, probably due to adsorbed H atoms. An admixture of transition metals that are able to dissociate H2 molecules lower the starting temperature of metal oxide reduction.62 Due to the low equilibrium decomposition pressure of CaCO3 at 573 K (0.0015 Pa), it was assumed that, with transition‐metal catalysts, hydrogenation occurred through direct interaction of CaCO3 and H atoms to form CH4. It is possible that reaction intermediates formed in the CaCO3 surface layer, but this assumption has not been validated. Activation energies derived from experiments in an electrobalance were 105 and 111 kJ mol−1 for the Ir‐ and Pd‐catalyzed reactions, respectively. The reaction rate increased steadily with increasing H2 pressure and remained constant at sufficiently high pressures. Yoshida et al. dedicated this tendency to a transition in reaction kinetics from slightly higher than first order at low pressures to zero order at high pressures. From reaction kinetics, it was suggested that the rate of dissociative adsorption of H2 on the metal surface was fast relative to the overall reaction rate.51

Jagadeesan et al. extensively studied hydrocarbon formation from various mixed inorganic carbonates and made remarkable conclusions about hydrocarbon selectivity for C1–C3 chain lengths.13, 52 In the first study, Jagadeesan et al. examined methane formation from mixed alkaline‐earth‐metal/transition‐metal carbonates at 823 K.52 The operating pressure was not stated and the assumption could thus be made that the work was carried out at atmospheric pressure. In a second paper by Jagadeesan et al., however, it was stated that the earlier study was performed at 0.3–0.5 MPa.13 The mixed carbonates were prepared by precipitation form aqueous solutions of NaHCO3 and had a composition of MeCa(CO3)2, in which Me was Co, Ni, or Fe in a Me/Ca ratio of 1:1, and Me1Me2Ca(CO3)3, in which Me1Me2 was CoNi, NiFe, or FeCo in a ratio Me1/Me2/Ca of 1:1:2. Solid decomposition products consisted of nanoparticles of metal dispersed on metal oxide. Hydrogenation of CoCa(CO3)2 gave CO2 and CH4 as major gaseous products. As the reaction proceeded, transition‐metal nanoparticles formed on the carbonate surface, which catalyzed the subsequent conversion of CO2 to CH4. An increasing amount of H2 facilitated the formation of transition‐metal nanoparticles through the reduction of metal ions. The carbonates completely decomposed. For CoCa(CO3)2 at 823 K, the optimal H2 flow rate for maximum conversion and CH4 selectivity was 8 cm3 min−1 for 5 h. These conditions applied to all mixed carbonates. Again, the type of transition metal had a leading role on product gas composition. With Co, complete carbonate conversion occurred with a selectivity to CH4 of 80 % (20 % CO2). All other transition metals and transition‐metal combinations yielded 100 % CH4 selectivity at reduced conversion. CO, H2O, and coke were not found. Carbonate conversion dropped in the order of NiCa (81 %)>CoNi (77 %)>FeCo (76 %)>NiFe (16 %)>Fe (4 %). Fe exhibited poor conversion. However, in the presence of Fe (Fe, NiFe, FeCo), traces of higher hydrocarbons up to C3 formed. The introduction of Pt or K into FeCa(CO3)2 did not increase the formation of higher hydrocarbons. Because reduced transition‐metal particles appeared to be essential for high CH4 selectivity, catalyst nanoparticles were prepared separately by heating freshly prepared transition‐metal carbonates in H2. They contained nanoparticles of transition metals, bivalent metal oxide, and CaO and were highly efficient in catalyzing the conversion of CO2 to CH4. With CoCa(CO3)2, if mixed in a 50:50 weight ratio with metal–metal oxide nanoparticles, the amount of CH4 formed during the first 2 h was four times higher. Studies on the effect of transition‐metal–metal oxide nanoparticles on hydrogenation of mixed carbonates indicated a change in reaction kinetics. In the presence of Co/CaO/CoO catalysts, H2 efficiency improved to yield 100 % CH4 selectivity for CoCa(CO3)2, in contrast to 80 % in the absence of the catalyst. The effect of Co on the methanation of carbonates was higher than that of Ni and Fe, and combinations thereof. The catalyst nanoparticles were also capable of decomposing the natural minerals calcite and dolomite. Complete conversion of MgCO3 and CaCO3 with 100 % CH4 selectivity was achieved with Co/CaO/CoO. The type of catalyst significantly influenced the ability to convert CaCO3 into CH4. Ni and Co, for instance, worked well individually, but were less active upon combination. Fe was not very active itself, but effectively catalyzed CaCO3 decomposition in combination with Ni and even more effectively with Co. CaCO3 conversion decreased in the order of Co (100 %)>Ni (80 %)>Fe (18 %) for single transition metals and FeCo (89 %)>NiFe (40 %)>CoNi (34 %) for combinations thereof.

In a second study, Jagadeesan et al. directly focused on the formation of C1–C3 hydrocarbons from CaCO3 through iron oxides.13 The starting carbonate, denominated FeCaCO, consisted of CaCO3 and Fe oxides with Fe/Ca molar ratios, x, of 0–5. The carbonate contained Fe in the form of Ca1−FeCO3 in the calcite structure (for x<2). Excess Fe was present as Fe2O3 and increased with increasing x. The effect of reaction temperature (573–873 K) was investigated at ambient pressure and a H2 flow rate of 3 cm3 min−1 in a continuous‐flow packed‐bed reactor. The reaction time was 2 h, after which time no further decomposition of the carbonate occurred. In any case, carbonate conversion was not complete. At 673 K, the carbonate conversion and yield of carbohydrates (23 % at x=5) were highest. Further gaseous products were CO2 and CO. Iron metal (α‐Fe), iron oxides (Fe3O4, γ‐Fe2O3, α‐FeOOH, CaFe2O4), and carbide (d‐Fe3C, χ‐Fe5C2) particles formed as solid residues supported on Ca‐rich oxides. The level of conversion and yield of hydrocarbons, suggesting higher hydrocarbon selectivity, increased with increasing molar ratios of Fe/Ca. At x=5, the yield of the gaseous products CH4, C2H4, C2H6, C3H8, CO, and CO2 was 5, 7, 4, 4, 6, and 21 %, respectively. In the absence of Fe, mainly CO2 formed with traces of CO. Consequently, Fe not only improved carbonate decomposition, but also the subsequent hydrogenation of released CO2 to higher hydrocarbons. For all molar ratios, the relative C2H4 yield was highest among all hydrocarbons. The yield of C3H8 increased with increasing amounts of Fe, which suggested that higher concentrations of Fe favored C−C coupling, rather than dehydrogenation. The total hydrocarbon yield from Fe‐catalyzed CaCO3 hydrogenation was comparable to that in the FT synthesis. A reaction mechanism based on carbonate decomposition to CO2, which underwent further reduction to CO and hydrocarbons, was stated. According to Jagadeesan et al., the formation of CO was crucial because CO and H2 were adsorbed on the catalyst surface and gave rise to hydrocarbon formation. They speculated that particle size played an important role in the selectivity of the reaction. From FT synthesis, it is known that smaller particles show lower H2 chemisorption potential, which favors the formation of olefins instead of C−C coupling.63

Main Benefits

The main benefits of metal carbonate hydrogenation may be summed up as outlined in the following sections.

CO2 emission reduction

CO2 is the primary greenhouse gas emitted through human activities. In 2015, the global CO2 concentration in the atmosphere reached an average value of (399.4±0.1) ppm.64 At present, the industrial sector is responsible for approximately one‐third of the total anthropogenic CO2 equivalent (CO2 e) emissions.65 Decarboxylation of metal carbonates under reducing conditions may contribute to a substantial decrease of CO2 emissions, especially in high‐emission industrial sectors, such as the iron and steel industry, in which carbonaceous ores are used.12 As opposed to conventional decarboxylation processes, CO2 is not released into the atmosphere, but reduced to CO, CH4, and higher hydrocarbons. The formation of reduced carbon species in the gas atmosphere adds value to the overall process compared with state of the art decarboxylation under oxidizing conditions.

Renewable production of chemicals and fuels

At present, the three major carbon feedstocks are still petroleum, coal, and biomass. Because most of earth's carbon (>99.9 %) exists as carbonates, carbonaceous minerals may provide a potential carbon source for hydrocarbon synthesis in the future. The conversion of carbonaceous inorganic rocks (e.g., calcite, magnesite, dolomite) to organic compounds may help to fulfil future energy requirements and provide a renewable and nearly inexhaustible resource for the production of chemicals.66

Because methane naturally occurs in biogas, power generation is still its main purpose. The conversion of biogas to electricity is standard technology. Apart from the production of synthesis gas (syngas) and its utilization, further uses of methane include catalytic and noncatalytic oxidative coupling (OCM) to C2+ hydrocarbons, direct oxidation to methanol or formaldehyde, oxidative methylation of hydrocarbons by methane, and oxidative carboxylation of methane by CO to acetic acid.67 The generation of H2 from CH4 is industrially accomplished through steam methane reforming, but the conversion can also be achieved by pyrolysis.68, 69, 70

From syngas, which is the feed material for a FT process, a synthetic crude oil (syncrude) is obtained. The syncrude consists of a multiphase mixture of hydrocarbons, oxygenates, and water. Refining of the syncrude yields products that are traditionally produced from conventional crude oil, such as transportation fuels and chemicals.21, 71

Hydrogen storage

Hydrogenation of metal carbonates can also be seen as a means of chemical hydrogen storage in the form of CH4 72 or fuels produced from syngas.73, 74 This is similar to the power‐to‐methane (PtM) concept that converts electrical into chemical energy by using captured CO2 and H2 from water electrolysis.75, 76 The main advantage of these alternative hydrogen‐storage technologies is the availability of storage and distribution systems prepared for natural gas and liquid hydrocarbon fuels. In regions where a natural gas infrastructure exists, both concepts provide a promising option to absorb and exploit surplus renewable energy.

Catalyst preparation

Tailor‐made solid products, regarding composition and morphology, may be achieved through adjusting the process conditions, especially the gas atmosphere. Morphology plays a crucial role if the solid products represent catalytically active materials, both catalytically active material and support material.49 A hydrogen atmosphere opens up a new pathway for transition‐metal carbonate transformation into finely dispersed, active catalysts for in situ or ex situ use.10 Although initial morphological studies seem promising, the catalytic activity of reductively calcined material has not yet been tested. To the best of our knowledge, only reductively calcined MgO was studied for its catalytic properties for CO2 conversion with H2, revealing reverse water‐gas shift activity.53 Long‐term stability was not considered.

Chemical solar energy storage

Experiments with energetic light, for instance Vis or UV radiation, indicate that irradiation affects the course of metal carbonate decarboxylation.10 The mechanistic course could depend on the wavelength of the radiation source, giving access to new pathways for the application of solar energy in solar furnaces. In principle, solar energy can be stored and transformed by means of cyclic inorganic processes. Primary electron spectroscopy for chemical analysis (ESCA) experiments revealed the difference between thermal hydrogenation and hydrogenation induced by irradiation.22 The results confirmed an effect of the type of energy (UV/Vis radiation) on the mechanism and kinetics of decomposition, but no precise conclusion was feasible and no further reports were made.77

Potential Fields of Application

Two fields of application were identified and analyzed regarding potential, limitations, and domains requiring further research. Both elucidate the need for further studies in industrial‐sized reactors, allowing appropriate process conditions to be stated for industrial application.

CCU based on mineral carbonation

Mineral carbonation is a novel, widely investigated concept for CO2 capture and storage based on weathering of limestone in nature.9, 78, 79, 80, 81, 82, 83 In mineral carbonation, high concentrations of captured CO2 from an industrial or power‐sector source react with metal oxide [MeO, mainly CaO or MgO; Eq. (28)] or metal hydroxide bearing materials to form the corresponding insoluble carbonate.

Due to the lower energy state of inorganic carbonates relative to CO2, the reaction is exothermic. Therefore, in theory, the process does not require any energy input, but produces heat. Unfortunately, extensive preparation of the solid reactants (including mining, transportation, grinding, and activation, if necessary); the use, recycling, and loss of additives and catalysts; and disposal of carbonates and byproducts render the overall process energy intensive and require external high‐grade energy sources.84 Due to its thermodynamics, carbonate formation is favored at low temperatures. High temperatures favor the reverse reaction, namely, decarbonation, which is generally referred to as calcination.

Appropriate carbonaceous feedstock sources include abundant silicate rocks that involve laborious mining and alkaline industrial residues that are readily available, but only on a small scale (e.g., slag from steel production or fly ash).85 Pure calcium and magnesium oxides and hydroxides provide the ideal source material because they are more readily carbonated than that of the corresponding silicates. However, due to their high reactivity, they are scarce in nature.78, 84

There are several single (direct) or multistep (indirect) dry or wet process routes. In aqueous environment, carbonation is faster, but, due to higher dilution and lower reaction temperatures, the heat of reaction is difficult to retrieve. The dry process is a simpler approach that brings gaseous CO2 into contact with particulate metal oxide bearing materials. Easy recovery of the heat of reaction is beneficial for this method, but the bottleneck is the slow rate of reaction at suitable temperature levels. It is only feasible at elevated pressures for refined, rare materials, such as the oxides and hydroxides of calcium and magnesium.9, 84

The generated carbonates (CaCO3, MgCO3), if not disposed of, are used for mine reclamation or in construction.

It is expected that alkaline‐earth‐metal carbonates will give access to reversible thermal decarbonation/recarbonation cycles if decarbonation is carried out under a reducing hydrogen atmosphere.55 In H2, CO2 is not released into the flue gas, but further reduced to CO.86 Reductively calcined MgO, CaO, SrO, and BaO were found to be constituted of solid conglomerates of microcrystalline domains featuring pronounced reactivity towards recarbonation; a fact that renders them promising as potential CO2‐trapping systems.48 Repeated carbonation/recarbonation cycles omit excessive measures for reactant preparation, makeup of additives and catalysts, and product disposal. Consequently, the net energy input required is potentially lower. Once prepared, refined, small particles of metal oxides can be repeatedly used; this poses a pronounced advantage for the reaction of dry CO2 gas with solid oxides not only as far as labor input is concerned, but also for the rate of reaction. It is well known that small particle sizes facilitate high reaction rates.78 In conventional mineral carbonation of CO2, most of the energy required is needed for grinding of the feedstock to particles of 100 μm. During carbonation, the formation of silica and carbonate layers on the mineral surface hinders the reaction and limits conversion. Carbonate layers are barely prevented, but silica layers do not form because pure metal oxides are applied. Conventional CO2 mineralization is criticized for its tremendous environmental impact associated with large‐scale mining directly leading to land clearing and product disposal; an issue that does not need any consideration if the metal oxides are repeatedly used.

Through transition‐metal doping of the solid reactant, a composite system may be generated, in which CO2 is trapped on metal oxides (carbonation/recarbonation step) and subsequently transformed into higher organic species through hydrogenation of the metal carbonate (decarbonation step; Figure 3).55

Figure 3

Concept of a closed CO2 circuit based on decarboxylation (decarbonation) of metal carbonates MeCO3 through hydrogenation followed by recarbonation (adapted from Ref. 55).

The technology for mineral carbonization is still immature. The concept of a closed CO2 circuit based on decarbonation through hydrogenation followed by recarbonation is a promising concept. Nevertheless, it is merely based on primary laboratory‐scale experiments and further research is required for a feasible economic analysis.

For a new 600 MWe coal‐fueled power plant with an annual CO2 emission of 4 Mt year−1, a total energy requirement of 580 kWh for CO2 capture and carbonation has been estimated.9 Because most of the energy is needed for grinding of the feedstock material (280 kW h) and this step is omitted for repeated decarbonation/recarbonation cycles, power plant efficiencies may increase from 23.6 to 33 % and CO2 avoidance rates from 72.5 to 82 % for the concept of a closed CO2 circuit.9 In the literature, reported cost estimations of various carbonation routes differ significantly. At present, direct aqueous technologies seem to be the most realistic ones with costs ranging from € 60 to 100 t−1 CO2 fixed.87 Additional CO2 emissions associated with the energy required for the carbonation process will boost the costs to € 80 to 130 t−1 CO2 fixed. Further taking into account the costs for capturing CO2 from a power plant yields total costs of € 150 t−1 CO2 avoided for a full CCS system with mineral carbonation.9

At this point, composite systems derived from transition‐metal doping that may not only trap CO2, but also transform it into higher organic species are not considered. Further energy, and consequently, cost savings are expected due to increasing reaction rates of carbonation based on the use of refined metal oxide reactants with small particle sizes. Metal doping also allows for lower reaction temperatures for the carbonation step and decreasing activation energies. Furthermore, dry carbonation will allow for easy accessibility of the heat of reaction of carbonation. Currently, cost estimation is not feasible because the type of catalyst, hydrocarbon species generated, and hydrocarbon selectivity still need to be established. Initial studies highlight its potential and elucidate the need for further research into process conditions and the long‐term stability of the system, which are crucial factors for its economic viability.

Direct reduction of mineral iron carbonate

Austria and China have major siderite reserves for iron and steel production. Siderite beneficiation is challenging because of the low iron content of the ore compared with magnetite and hematite ores. The industrial practice is to blend siderite with other high‐grade ores in the sinter plant. During the sintering process, siderite is converted into hematite through roasting in air. The sinter product is fed to the blast furnace (BF), in which it is preferably reduced with coke via CO, producing at least 1.5 mol CO2 per mole of iron due to the stoichiometry of reaction. Consequently, at least 2.5 mol CO2 are emitted during the production of 1 mol iron from iron carbonate.

Direct hydrogen reduction of the mineral iron carbonate represents a novel process concept for sustainable pig iron production. It is a high‐potential approach for significant energy savings and CO2 emission reduction, especially if coupled with catalytic CO2 hydrogenation (e.g., methanation) to further convert inevitably released CO2 (Figure 4).12 Due to the debate about a sustainable energy supply, research into methanation has increased tremendously in recent years and is readily available.88, 89, 90, 91, 92, 93, 94, 95

Figure 4

Process concept for direct iron production from the mineral siderite followed by off‐gas valorization by catalytic CO2 hydrogenation, for example, methanation.

Proof of concept

TG studies revealed that mineral iron carbonate was directly reduced to elemental iron under a hydrogen atmosphere. Iron carbonate reduction was represented by a distinct mass loss below 723 K, which was followed by a small relative mass loss spanning over a broad temperature range (723–923 K) allocated to the concomitant decomposition of manganese, magnesium, and calcium carbonate to the respective oxides.12

In the ideal case of complete carbonate conversion, elemental iron is formed together with CO2, CO, CH4, and potentially even higher hydrocarbons (CH). Baldauf‐Sommerbauer et al. investigated the effect of temperature (Figure 5) and pressure on the composition of the product gas that consisted of CO2, CO, and CH4.96 Elevated pressure and low temperature increased the yield of CH4. CO formation was preferred at low pressure and higher temperatures.

Figure 5

Dry product gas composition for the direct hydrogen reduction of the mineral siderite at 623 and 648 K; 60 g siderite, size fraction: 0.5–1 mm, H2/N2=9:1, 867 cm3 min−1, ambient pressure.

CO2 emission savings of at least 60 % are possible because, at most, 1 mol CO2 is released per mole of elemental iron. If hematite is reduced with hydrogen, 1.5 mol hydrogen is required per mole of elemental iron. Consequently, up to 33 % less reducing agent is needed if direct siderite reduction is applied, due to circumventing the hematite route. Direct siderite reduction can be run at relatively low temperatures (673–773 K) compared with other metallurgical iron carbonate beneficiation processes, such as the Midrex© process for direct iron oxide reduction with natural gas (1053–1073 K)97 or the classical BF process (1773 K).98

Case studies

A comparison of four different case studies with the state‐of‐the art BF process highlights the potential of the concept of direct hydrogen reduction of the mineral iron carbonate.

As feed material, the mineral siderite with a characteristic composition of a concentrated siderite sample from the Styrian Erzberg in Austria (Table 2) was chosen. Iron was assumed to occur in the form of FeCO3 (79 wt %) and CaFe(CO3)2 (5 wt %). Minor components were MgCO3, MnCO3, SiO2, and Al2O3.

Table 2

Mean composition of the mineral siderite from the Styrian Erzberg in Austria.

Component	Mass fraction
FeCO3	0.79
CaFe(CO3)2	0.05
MgCO3	0.07
MnCO3	0.05
SiO2	0.03
Al2O3	0.01

The classical BF process serves as benchmark. It consists of a sintering step under oxidizing conditions [Eqs. (29), (30), (31), (32), (33), 1373 K) in which FeCO3 is transformed into Fe2O3 for reduction in the BF [Eq. (34), 1773 K].

During direct hydrogen reduction to elemental iron, several reactions are conceivable for FeCO3 [Eqs. (35), (36), (37)] and CaFe(CO3)2 [Eqs. (38), (39)], depending on the product gas. The reaction temperature was set to 773 K.

To calculate CO2 emissions, direct emissions of CO2 and CO are summed due to CO oxidation to CO2 [Eq. (40)]. Oxidation of CO contributes to CO2 emission, but reduces the total energy demand. A reduction in energy demand because of CH4 was also considered. Hydrogen supply was assumed to be accomplished by water electrolysis (4.8 kWh Nm−3).24, 25

Four case studies, Red1, Red2, Red3, and Red4, represent the extreme cases (Red1 and Red2) and mixed cases (Red3 and Red4). In Red1, full carbon conversion to CO was assumed [Eqs. (35) and (38)]. CH4 formation was postulated for Red2 [Eqs. (36) and (39)]. For Red3, 50 % CO and 50 % CH4 were assumed. Red4 reproduces the experimental product composition depicted in Figure 5 (49 % CO2, 27 % CH4, 24 % CO).

Total CO2 emissions and the total energy demand for all four cases are compared with the benchmark BF process in Table 3. The results quantify the capability of CO2 emission reduction. The classical BF process releases 2212 kg CO2 t−1 pig iron. 100 % CO formation (Red1) saves 64 % of the CO2 emitted in the benchmark process. No CO2 is released if full conversion to CH4 is hypothesized (Red2). This scenario exhibits the highest energy demand of 5267 kW h (111 % compared to the BF process) and is not aspired to from an economic point of view. Scenarios with CO and CH4 (Red3 and Red4) show excellent CO2 emission reduction (82 and 74 %) and decreased energy demand (11 and 28 %, although a hydrogen supply from water electrolysis was chosen), which underlines the high potential of the proposed concept.

Table 3

Total CO2 emission and total energy demand for the case studies Red1 (100 % CO formation), Red2 (100 % CH4 formation), Red3 (50 % CO and 50 % CH4), and Red4 (49 % CO2, 27 % CH4, 24 % CO) compared with the benchmark BF process.

Case study	CO2 emission	Energy demand
	[kg CO2 t−1 Fe]	[GJ t−1 Fe]	[kWh]
BF	2212	17.1	4755
Red1	788.5	11.4	3156
Red2	0	19.0	5267
Red3	394.3	15.2	4211
Red4	569.9	12.2	3401

Direct iron carbonate reduction is a high‐potential candidate to open up a new route for environmentally benign pig iron production. The findings are based on TG experiments with siderite12 and tests in a tubular reactor setup;96 thus direct conclusions for application in large‐scale reactors and optimized process conditions (e.g., particle size, temperature) cannot be drawn yet. Iron separation from the unconverted siderite matrix and gangue through magnetic separation was suggested in the literature, but still needs verification. Nevertheless, the presented case studies highlight the potential of reductive calcination of siderite and the need for ongoing research in this field.

Summary and Outlook

Various aspects render metal carbonate hydrogenation a powerful means for direct and indirect CO2 emission reduction, CO2 utilization, and metal carbonate exploitation.

Under a hydrogen atmosphere, the decarboxylation temperature is significantly lower than that of the respective reaction under inert conditions. Doping with transition metals further lowers the temperature level. The combination of decarboxylation and CO2 reduction with the renewable energy carrier hydrogen transforms the conventional endothermic process into an overall exothermic process, which allows for significant energy savings.

In reductive metal carbonate decarboxylation, CO2 is not (or only partially) released, but reduced to CO, CH4, and higher hydrocarbons. The composition of the gaseous reaction product strongly depends on the gas atmosphere (pure or dilute hydrogen); the presence of transition‐metal species acting as in situ catalysts; and the reaction temperature, pressure, and residence time. Apart from metal oxides in various oxidation states, elemental metals are obtained as solid reaction products from transition‐metal carbonates. Tailormade products, in terms of composition and morphology, would give access to novel production routes for catalysts.

Until now, preliminary studies focusing on feasibility and chemism have mainly been made with metal carbonates in small‐scale apparatus, lacking transferability to industrial scale. Additionally, disagreement exists concerning the reaction mechanisms. Whereas some researchers propose the direct reaction of hydrogen with fixed CO2 in the carbonate, others assume that hydrogen reacts with released CO2. Degradation studies in hydrogen and nitrogen revealed differences in morphology that indicate the direct reaction of H2 with CO2; however, several aspects require closer examination, especially when it comes to optimized process conditions for industrial applications. Once clarified, metal carbonate hydrogenation could provide a quantum leap in high‐emission industrial sectors, such as the iron and steel industry, if a renewable hydrogen supply is accomplished. Further potential fields of applications include the renewable production of chemicals and catalyst preparation.

Conflict of interest

The authors declare no conflict of interest.

Biographical Information

Susanne Lux studied technical chemistry at the Graz University of Technology in Austria. In 2009, she obtained her Ph.D. in chemical engineering under the supervision of Prof. Matthäus Siebenhofer. In 2012, she accepted a position as Assistant Professor at the Institute of Chemical Engineeering and Environmental Technology at Graz University of Technology. Her research group focuses on process intensification, with the main focus on reactive separation technologies and heterogeneous catalytic systems to address environmental challenges.

Georg Baldauf‐Sommerbauer studied chemistry (B.Sc.) and chemical engineering (M.Sc./Dipl.‐Ing.) at the University of Graz and Graz University of Technology in Austria. In 2017, he successfully defended his Ph.D. thesis (supervised by M. Siebenhofer and S. Lux) on the reductive calcination of mineral carbonates. His research interest focuses on fundamental kinetic studies, as well as on applied studies towards a possible industrial implementation of renewable technologies.

Matthäus Siebenhofer studied chemical engineering and received his doctoral degree in 1983 from the Graz University of Technology in Austria. Currently, he is Professor for Chemical Reaction Engineering and Head of the Institute of Chemical Engineering and Environmental Technology at Graz University of Technology.