Ali Reza Kamali
*
Energy and Environmental Materials Research Centre (E2MC), School of Metallurgy, Northeastern University, Shenyang, 110819, People's Republic of China. E-mail: ali@smm.neu.edu.cn; a.r.kamali@cantab.net
First published on 1st October 2020
Green and low cost production of strategic materials such as steel and graphene at large scale is a critical step towards sustainable industrial developments. Hydrogen is a green fuel for the future, and a key element for the clean production of steel. However, the sustainable and economic production of hydrogen is a barrier towards its large scale utilisation in iron and steelmaking, and other possible applications. As a key challenge, the water electrolysis, which is commonly used for the carbon-free production of hydrogen, is uneconomic and involves various problems including the corrosion of equipment, the use of expensive catalysts and high over-potentials, limiting its viability. Moreover, the hydrogen transportation from the electrolyser to the utilisation unit is problematic in terms of cost and safety. From a thermodynamic point of view, the potential and efficiency of the water splitting process can greatly be improved at high temperatures. Therefore, a practical approach to resolve the above-mentioned shortcomings can be based on the electro-generation of hydrogen in high temperature molten salts, and the utilisation of the generated hydrogen in situ to produce metals, alloys or other commercially valuable materials. Clean production of alloy powders is particularly interesting due to the rising of advanced manufacturing methods like additive manufacturing. The hydrogen produced in molten salts can also be used for the large scale preparation of high value advanced carbon nanostructures such as single and multi-layer high quality graphene and nanodiamonds. The combination of these findings can lead to the fabrication of hybrid structures with interesting energy and environmental applications. Surprisingly, the production of a large variety of materials such as Fe, Mo, W, Ni and Co-based alloys should be achievable by the electrolytic hydrogen produced in molten salts at a potential of around 1 V, which can easily be powered by advanced photovoltaic cells. This review discusses the recent advancements on these topics.
Hydrogen is also considered as a green reducing agent for the sustainable production of a large variety of metals such as molybdenum,12,13 nickel,14 germanium,15 tungsten,16 and cobalt17 as well as alloys such Fe–Mo.18 On important possible application of hydrogen is the clean reduction of iron oxides either from oxidised alloys,19 or iron ores20–22 at 600–800 °C. This application is particularly attractive due to the fact that the steel industry is at the heart of global development with a world production of over 1.8 billion tons, producing more than 3.2 billion tons of CO2 every year.23
Moreover, the natural gas resources mostly contain impurities such as H2S and CO2, and should be purified before transportation due to the toxicity and corrosion-enhancing characteristics of these compounds. Therefore, the natural gas is difficult to be used in the future sustainable developments.35–37 A concept to tackle the CO2 emissions is based on the CO2 capture.38,39 However, its high associated costs, based on the current technologies (US$60–80 per ton of CO2), are far more expensive to make this concept commercially viable at large scales.40
Fe2O3 + 3H2 → 2Fe + 3H2O | (1) |
However, practically, around four times the stoichiometric gas flow is required to support the reduction process.41,42 It should be mentioned that hydrogen produced by different methods costs between $1.5 to $6.2 per kilogram.43 The lower costs associate with steam methane reforming (SMR) methods that also produce CO2 emissions, whilst higher costs (over $2.2 per kilogram) relates to electrolysis approaches, with an efficiency of around 30%.43,44 Therefore, the hydrogen required for the clean production of 1 kg Fe would cost around $0.5 which is not economic, considering that the price of producing iron in blast furnace is considerably less expensive.
It should be noticed that, theoretically, for the decomposition of water, a minimum voltage of 1.23 V is required to achieve the cathodic hydrogen and anodic oxygen evolution reactions. Practically, however, the threshold voltage for the water electrolysis is considerably higher at 1.8–2.0 V either in alkaline or acidic electrolytes, due to the presence of ohmic losses, and the activation over-potential caused by electrode kinetics and the sluggish mass transfer at the electrode–electrolyte interfaces.45–48 Therefore, a large amount of energy is required to split water (∼4 kW h m−3) which is greater than the energy content of produced hydrogen (∼3.5 kW h m−3),49 making the whole process non-economical on an industrial scale. Other technical problems include the corrosion of the electrodes, containers and compartments;50 and expensive catalysts required such as Pt- and Ru-based compounds.51 Due to these limitations, the production of hydrogen by water electrolysis accounts for only 4% of the world's hydrogen production.
On the other hand, while the production of hydrogen powered by photovoltaic (PV) panels represents an environmentally preferable way for the green production of hydrogen, this approach is not economic using current technologies. The efficiency (E) of such a system comprising the PV panel and the water electrolyser can be expressed by eqn (2):49
E = (M × C)/(V × I) | (2) |
ΔH° = ΔG° + TΔS° | (3) |
Fig. 1 The theoretical electrolysis voltage and the energy required for the electro-decomposition of water into hydrogen and oxygen; H2O → H2 + 1/2O2. |
At present, solid oxide steam electrolysis (SOE) operating at 700–1000 °C is the only way to produce hydrogen at high temperatures,52,53 but these systems require expensive materials such as yttria doped zirconia, gadolinia-doped ceria, and lanthanum strontium cobalt ferrite.54 Moreover, the SOE electrolysis cells are prone to degrade primarily due to the anode delamination at the interface with the solid electrolyte due to the elevated pressure caused by the electrochemically formed oxygen.55,56 Moreover, the transportation of hydrogen from the production unit to the utilisation unit is highly problematic, expensive and dangerous.57,58
2LiClmelt + H2Oatmosphere → 2HCl[LiCl] + Li2O[LiCl] | (4) |
Although the standard equilibrium constant of this reaction is low (K = 1.05 × 10−10 at 700 °C), yet the reaction can proceed at a finite rate by the dissolution of the reaction products in the melt. It should be mentioned that the formation of Li2O in the melt is highly beneficial for the fast production of Li-containing ceramic materials such as LiNbO3,64 Li2Fe3O5,67 and Li2TiO3 (ref. 68) as well as Li2CO3.69,70
Under this situation, the cations present in the molten salt (H+ or Li+) may be discharged on the cathode under the influence of a cathodic potential applied. In LiCl, possible electrochemical reactions at 660 °C can be described as below:
2LiCl → 2Li + Cl2(g), ΔG° = 666.3 kJ per mole, E° = −3.4 V | (5) |
2Li2O → 4Li + O2(g), ΔG° = 948.7 kJ per mole, E° = −2.5 V | (6) |
2HCl → H2(g) + Cl2(g), ΔG° = 200.8 kJ per mole, E° = −1.0 V | (7) |
2H2O → 2H2(g) + O2(g), ΔG° = 303.9 kJ per mole, E° = −0.79 V | (8) |
As can be observed, despite the presence of Li+ in the melt, these cations might not be discharged on the cathode, if the cell potential is restricted smaller than 2.5 V, since the corresponding reactions involving Li+ require larger potentials to occur. In other words, the cathodic formation of hydrogen is favorable. It can also be depicted that the evolution of oxygen at the anode is more favorable than that of chlorine. Therefore, at lower cell potentials, hydrogen can solely be discharged on the cathode:
2H+ + 2e → H2 on the cathode | (9.1) |
O2− → 1/2O2 + 2e on the anode | (9.2) |
H2O[LiCl] → H2 + 1/2O2, ΔG° = 153 kJ, E° = −0.79 V, T = 650 °C | (9) |
The formation of hydrogen in molten salts at temperatures around 600–900 °C can provide unique applications as briefly explained in next sections.
(10) |
Fig. 2 High quality graphene can be produced in molten salts containing hydrogen cations. (a) Protons are neutralised at the surface of cathodically polarised graphite to form hydrogen atoms adsorbed on the graphite surface at 800 °C.The adsorbed hydrogen atoms diffuse into the graphite structure, and then combine to form hydrogen molecules, as identified in (b). The hydrogen gas formed within the graphite structure can exfoliate the material into high quality graphene sheets, as indicated by a typical SEM micrograph in (c).59,60,71 |
The rapid diffusion of into the graphite lattice is highly possible, as confirmed by the temperature dependency of the diffusion coefficient of atomic hydrogen into the graphite structure (DH°, cm2 s−1):76
DH° = 2.0 × 10−3exp(−6.09 × 10−20/kBT), T = 300–1700 °C | (11) |
(12) |
Diffused hydrogen atoms can then combine to form hydrogen molecules within the graphite structure:
(13) |
Such hydrogen molecules (with a size of 0.25 nm) created within the hexagonal graphite lattice (with an interlayer spacing of 0.33 nm) have sufficient kinetic energy to exfoliate the graphite material in a green and economic way.60 The consumables of the process are electrical energy and water, and no harmful by product is produced, hence the process is environmentally safe. The cost and the specific energy consumption for the preparation of graphene in molten LiCl can be estimated to be low at about US $10–20 and 25 kW h per kilogram of the graphene product, respectively. The molten salt graphene has presented superior performance in various applications including composite anode materials for high performance lithium ion batteries,78–80 electrode materials for supercapacitors,62 reusable adsorbents,81,82 high performance ceramic composites83 and precursors for the facile fabrication of nanodiamonds.84–87 The application of molten salt produced graphene nanosheets as an efficient reusable adsorbent is based on its high surface area and density of edge sites, as well as its high structural stability, providing the nanosheets with the potential to become recovered after the adsorption of organic contaminants from aqueous solutions by a simple heat treatment process is air.81
The crystallinity, and therefore, the electrical conductivity of molten salt produced nanostructured carbon materials have been reported to be high, due to the improving effects brought about by the ionic melt, including the reactive dissolution of impurities from carbonaceous materials into molten salts, the reduction of the d-spacing fluctuation between graphitic basal planes, and high diffusion rate of species in the molten salt environment, enhancing the crystallisation of carbon products.87–92
3Fe2O3 + 2Li2O + H° + H+ + e → 2Li2Fe3O5 + H2O | (14) |
Fig. 3 Hydrogen can be produced in molten LiCl under a humid argon atmosphere, and the hydrogen produced can reduce Fe2O3 to Fe.93 |
Interestingly, it was found that the electrolytic Li2Fe3O5 formed by the reaction (14) has an excellent performance as anode material in lithium ion batteries.94 At prolonged electrolysis time, the removal of oxygen from the cathode, and therefore, a complete metallisation of the cathode could be achieved:
Li2Fe3O5 + 4H2 → 3Fe + Li2O + 4H2O, ΔG° = −1.1 × 10−16 kJ mol−1, T = 660 °C | (15) |
Whilst the reaction (15) is energetically close to the equilibrium, the presence of a small cathodic potential as well as the dissolution of the reaction products (Li2O and H2O) into the melt provide further driving forces for the reaction to proceed. At 660 °C, a cell voltage of 0.79 V should be theoretically sufficient for the decomposition of water dissolved in molten salt to form cathodic hydrogen. This result93 demonstrated that Fe2O3 precursor could be reduced to Fe under a low applied voltage of 1.0 V. This achievement can revolutionise the iron production technology, as can be depicted from Table 1.
Interestingly, individual iron oxide particles simply immersed in molten LiCl at 660 °C can also be reduced to metallic iron particles by the hydrogen gas generated in situ in the melt by the electrochemical decomposition of water at a low voltage of only 0.97 V,61 as illustrated in Fig. 4, under the influence of hydrogen gas generated on the graphite crucible:
Fe2O3 + 3H2(electro-generated) → 2Fe + 3H2O[LiCl], ΔG° = −12.5 kJ, T = 650 °C | (16) |
Fig. 4 Schematic representation of the molten salt production of hydrogen and its utilisation for direct reduction of Fe2O3 powders suspended in the melt. Water can repeatedly be decomposed and formed during the process to reduce iron oxides.61 |
In another work, it was demonstrated that water can repeatedly be decomposed and regenerated in molten LiCl at 680 °C under a low cell voltage of only 1 V, leading to the clean production and the in situ consumption of hydrogen to reduce cobalt oxide into highly faceted cobalt microcrystals with a low energy consumption of 1150 kW h per ton of reduced cobalt, providing considerable advantages over the alternative approaches available for cobalt production.99
It was explained that the hydrogen gas generated on the cathode (reaction (9.1)) can either escape from the cathode (reaction (17)) or alternatively be consumed for the reduction of cobalt oxide (reaction (18)).
H2 (melt) → H2 (atmosphere) | (17) |
4H2 + Co3O4 (pellet) → 3Co + 4[H2O]LiCl, ΔG° = −236.3 kJ at 670 °C | (18) |
Fig. 5 exhibits the voltage contributions between the graphite anode and the metal oxide cathode, and a Mo pseudo-reference electrode immersed in the melt during the process. As can be observed, although the constant cell voltage of 1 V is applied throughout the process, the cathodic and anodic potentials vary with the electrolysis time. The I–t and V–t curves exhibit various distinct stages, based on which the cathodic generation of hydrogen and its utilisation can be explained. The electrolysis initiates with a cathodic and anodic potential of −0.35 V and 0.65 V, respectively, contributing to the high capacitive current observed (this section is not clear in Fig. 5). Then, the values of cathodic and anodic voltages change to around −0.25 V and 0.75 V, respectively, corresponding to the stage I in Fig. 5. During this stage, the cathodic reaction is mainly characterised by the high yield generation of hydrogen, reaction (9.1), on the fresh Ni wire surfaces wrapping the cobalt oxide pellet. After around 30 min, a continuous increase in the cathode potential (and the counterpart reduction of the anodic voltage) is observed, related to the stage II of the process. This stage is associated with the gradual passivation of the Ni wire surfaces, increasing the cathodic impedance, and consequently, the increase of the cathode voltage and the resulting current drop. During the stage I and II, the hydrogen released on the cathode have the opportunity to reduce the surface of the cobalt oxide to form metallic cobalt, proving more active surfaces for the electrochemical production of hydrogen, and hence, leading to a gradual recovery of current and the initiation of stage III. In this stage, the combination of reactions (9) and (18) will maintain the concentration of hydrogen cations adjacent to the cathode, and therefore, an efficient reduction of the oxide phase occurs. Then, at the end of the process, the cathodic voltage sharply increases to the values around −0.5 V due to the passivation of the metallic cobalt cathode, and the shortage of hydrogen cations available adjacent to the cathode. The passivation is mainly due to the formation of hydrogen gas on the cathode with no subsequent metal reduction. Since the experiment was conducted at a low cell voltage of 1 V, the increase in the impedance of the cathode results in a sharp current decay, as observed in the I–t curve. The mechanism explained above, is responsible for the electrochemical production of hydrogen and subsequent chemical reduction of cobalt oxides by the newly generated hydrogen through different stages.
Fig. 5 The changes of cell current and cathodic/anodic voltages measured versus electrolysis time during the electrolysis of LiCl at 1 V using a cobalt oxide pellet wrapped with Ni wire served as the cathode, and a graphite crucible served as the anode under humid Ar flow. The cathodic and anodic potential differences were measured between the cathode/anode and a Mo pseudo-reference electrode immersed in the melt.99 |
The low cost and green fabrication of metallic powders using this innovative approach should be interesting, particularly considering the rise of additive manufacturing that utilises metallic powders for the precise, fast and economical manufacturing of components for various industries including aerospace, automotive, energy, and biomedical applications.100–102
It is worth mentioning that in the traditional hydrogen reduction processes, metal oxides are reduced by a hydrogen gas stream, from which only a small fraction of hydrogen molecules can contribute to the reduction process41,42 and the rest flows away from the reduction zone. In contrast, in the molten salt process, hydrogen is gradually formed in the vicinity of metal oxides, and therefore, a substantially less amounts of hydrogen is required, contributing to the high efficiency of the process.
I should be mentioned that in the water assisted molten salt reduction of metal oxides, hydrogen cations in the molten salt can be reduced on metal oxide cathodes to form metal and water. The latter is instantly dissolved in the molten salt to re-form hydrogen cations. Therefore, theoretically, a small amount of water in molten salts should be sufficient to reduce a high amount of metal oxides immersed in the melt, at a low cell potential which just high enough to decompose water at high temperatures (<1 V). This mechanism, which is highlighted in Fig. 6, can be applied for a variety of metal oxides.
Table 2 presents a selected number of metal oxides that can thermodynamically be reduced in molten salts under a low cell potential of around 1 V, under the influence of the generated hydrogen. The changes of Gibbs free energy of these reactions are negative, and therefore, can proceed instantly upon the generation of cathodic hydrogen. For the case of oxides such as WO3, although the ΔG° has a positive value at 700 °C, negative values of ΔG° can still be achieved at T > 825 °C. Although this temperature can easily be obtained using LiCl (melting point ≈ 605 °C, boiling point ≈ 1380), other molten salts with lower values of vapour pressure, such as CaCl2 (melting point ≈ 770 °C, boiling point ≈ 1930 °C) can be more appropriate at higher temperatures. It should be mentioned that, unlike salts such as SnCl2,105 molten salts like CaCl2 (ref. 106 and 107) and ZnCl2 (ref. 108) have an obvious capability of being hydrolysed at high temperatures to produce HCl, indicating their potential for being considered appropriate candidates to be included in the molten salt mediums for the production of hydrogen. Further studies should be conducted to elucidate the performance of alternative salt systems.
Hydrogen reduction reaction | ΔG° (kJ) |
---|---|
Fe2O3 + 3H2 → 2Fe + 3H2O | −16.7 |
MoO3 + 3H2 → Mo + 3H2O | −83.0 |
PbO + H2 → Pb + H2O | −72.3 |
CuO + H2 → Cu + H2O | −125.7 |
SnO2 + 2H2 → Sn + 2H2O | −9.1 |
CoO + H2 → Co + H2O | −28.8 |
NiO + H2 → Ni + H2O | −43.3 |
WO3 + 3H2 → W + 3H2O | 9.7 |
It is worth mentioning that the thermal energy required to prepare molten salts can be supplied from waste thermal energy recovered from industrial processes. The wide range of possibilities, shown in Table 2, provides a new horizon in the green production of metals and alloys using water by applying a potential in the range of 1–2.5 V. This includes the green production of stainless steels, Ni-based super alloys, and novel high entropy alloys in powder form. Green production of metal and alloy powders should support the novel manufacturing techniques such as additive manufacturing. The other areas include the reprocessing of nuclear fuels, the recovery of waste metals and alloys, as well as the green and CO2-free extraction of metals and alloys from natural ores. Moreover, the intermediate products produced during the molten salt processes, such as Li2Fe3O5 (reaction (14)),94 can be interesting for energy applications. Molten salt preparation of metal hydrides should also be highly interesting for applications such as advanced catalysis103 and energy storage.104
Fig. 7 Schematic representation of the solar powdered hydrogen production and utilisation in molten salts. |
The characteristics mentioned above are very interesting especially where the large-scale green production of metallic materials is concerned. However, the molten salt approach, like many other high temperature processes, has its own challenges, such as the corrosion of the reactors at high temperatures under the influence of molten salts.111–114 Moreover, the level of expertise required to conduct molten salt electrochemical experiments may be higher than those in low temperature operations creating a barrier for researchers to effectively utilise this approach. These limitations can be significantly reduced/eliminated in large-scale molten salt operations, such as the case of Hall–Héroult process for aluminium smelting.
Despite its interesting nature, however, there is only limited experimental information available in the literature concerning the solubility of hydrogen containing compounds such as water,115–120 HCl121,122 and HF123,124 in molten salts. This is mainly due to the complications involved in such experiments at high temperatures, and also the fact that in traditional molten salt processes, particularly, water has commonly been considered as an undesired impurity that should be eliminated as much as possible from the salt.125–128
From the available literature, it is known that the solubility of various gases in high temperature molten salts can be classified into two groups; i.e. physical and chemical. The former is represented by the dissolution of nonpolar noble gases in molten metal halides, in which the dissolution can be explained by the creation of a cavity with the same size as the gas molecule in the melt. In this case, the solubility can be correlated by equating the free energy of the gas-melt solution to the free energy of formation of holes considering the surface tension of the melt as the solvent.129,130 Such solutes might still exhibit polarisation effects, for instance by the polarisation of the salt ions around the solute particles due to the symmetry breaking created by the presence of such particles.117
The solubility of polar water in molten LiNO3 and LiNO3–KNO3 can be as high as 3.5 × 10−4 mole of water per mole of the salt. However, the water dissolution process involved was found to be reversible, and based on electrostatic ion–dipole interactions without the occurrence of a hydrolysis reaction.111
In contrast to the physical type, the chemical dissolution of water in molten salts is considered technologically important and rather undesirable in various fields such as the electrolytic production of aluminium in cryolite-based melts. It this case, the dissolution process proceeds through the hydrolysis reaction between water and fluoride salts, leading to the formation of hydrolysis products including OH− and H+ in the bulk of molten salt. These ionic species can act as charge carriers, influencing the current efficiency of the process. Furthermore, the excessive formation of corrosive HF is of concern.123,131 The chemical solubility of water in pure NaF and CaF2 was estimated to be as high as 100 and 36 ppm at around 1000 and 1400 °C, respectively.123
In chloride molten salts, the eutectic LiCl–KCl is technologically important for applications such as the processing of used nuclear fuels,132,133 and therefore, its interaction with water has received some research interest.
Accordingly, it is known that the water uptake of this molten salt system can reach around 0.001 mole of H2O per mole of LiCl134 at around 400 °C. Also, the solubility of HCl in this eutectic salt has been studied, which is also relevant to our case since protons might be formed by introducing HCl (as the solute) into the molten salt (as the solvent), without the involvement of O2−. The protons can be supplied by bubbling of HCl in the melt for several hours.
An interesting discussion can be found in the work of Laitinen and Plambeck135 who saturated the purified eutectic LiCl–KCl with HCl, and studied the cell voltage required to form H2 and Cl2 on Pt electrodes:2
HCl → H2 + Cl2, Ecell = +1.0161 ± 0.0050 V |
In another work, Van Norman and Tivers122 employed chronopotentiometry to investigate the same system using a glassy carbon working electrode, and determined the apparent diffusion coefficient of protons to be around 2 × 10−4 cm2 s−1, which is an order of magnitude higher than those of most other species in molten salts.
Minh and Welch136 studied the cathodic reduction of hydrogen chloride dissolved in the melt by a combination of chronopotentiometry and linear sweep voltammetry using platinum served as the working electrode and also the pseudo-reference electrode, and graphite as the counter electrode. After the saturation of the melt with HCl, a well-defined cathodic reduction wave could be observed in the corresponding voltammogram at −0.55 V vs. the Pt pseudo-reference electrode. The peak disappeared when HCl was purged from the eutectic melt with argon. They further found that the reduction reaction proceeds by a reversible one-electron transfer process, and that the diffusion coefficients obtained are considerably high (1.8–2.7 × 10−4 cm2 s−1). The high diffusion coefficients observed could be attributed to the small ionic radius of non-solvated protons in comparison with other ions, providing protons with high mobility within the free volume of the melt rather than its ionic lattice. Moreover, the activation energy for inter-diffusion of protons was estimated to be low at 3.7 kJ mol−1, which is an order of magnitude lower than those of alkali metal ions.136
Future investigations should be designed to more specifically evaluate key issues influencing the molten salt generation of hydrogen such as the selection of molten salt and electrodes materials, as well as the processing conditions including the temperature and cell voltage. Electrochemical measurements should also provide useful information about the mechanisms and kinetics of hydrogen evolution in more details. Finally, the generated hydrogen should provide an opportunity to investigate the green production of a range of metals and alloys.137 This research is ongoing in my laboratory.
This journal is © The Royal Society of Chemistry 2020 |