Elsevier

Journal of Power Sources

Volume 203, 1 April 2012, Pages 4-16
Journal of Power Sources

Review
Recent advances in high temperature electrolysis using solid oxide fuel cells: A review

https://doi.org/10.1016/j.jpowsour.2011.12.019Get rights and content

Abstract

New and more efficient energy conversion systems are required in the near future, due in part to the increase in oil prices and demand and also due to global warming. Fuel cells and hybrid systems present a promising future but in order to meet the demand, high amounts of hydrogen will be required. Until now, probably the cleanest method of producing hydrogen has been water electrolysis. In this field, solid oxide electrolysis cells (SOEC) have attracted a great interest in the last few years, as they offer significant power and higher efficiencies compared to conventional low temperature electrolysers. Their applications, performances and material issues will be reviewed.

Highlights

High temperature electrolysis is the cheapest way to produce clean hydrogen. ► The materials used for these devices are reviewed. ► Their performance and degradation issues are also reviewed. ► Modeling of systems is analyzed. ► Chemical flexibility of these devices is finally discussed.

Introduction

Renewable energy resources have attracted great interest in recent years. A fundamental problem associated with renewable energy sources such as solar energy, wind power, hydropower or geothermal power is that they have to match supply with demand, and therefore energy storage is essential. Battery storage has been proposed as an alternative for some applications, although several problems such as high cost for large storage requirements, or loss of charge overtime are also associated. Energy storage in the form of hydrogen will also be essential and has been widely discussed for many years with an increasing drive toward the hydrogen economy. Hydrogen is probably the preferred energy carrier for a future zero-carbon economy but several research efforts are required in order to supply inexpensive and plentiful amounts of fuel. Although hydrogen is the most abundant element in nature, it is usually found as a compound combined with other elements, and thus, the production of hydrogen always requires energy. Current hydrogen production methods need the use of fossil fuels, such as steam reforming, partial oxidation of heavy hydrocarbons and gasification of coal. Other processes currently under development include reforming and pyrolysis using biomass and other carbon waste, direct methanol reforming, as well as fermentation of biomass and biological production. Moreover, there are also other hydrogen production methods that are generally categorized as electrochemical processes, including photoelectrochemical methods, thermochemical water splitting, and water electrolysis. Of these, only water electrolysis is currently commercially available. In addition, of all the methods to produce hydrogen, water electrolysis is probably the cleanest when combined with a renewable energy source to produce the electricity. Additional information regarding hydrogen production can be found in the following excellent reviews [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19]. This review will focus on the production of hydrogen by high temperature electrolysis. Materials, performance and degradation issues of those devices will be reviewed in the manuscript.

Water electrolysis to produce hydrogen and oxygen gases is a well-known established process. Basically, the principle of a water electrolyser is to convert water and DC electricity into gaseous hydrogen and oxygen, that is to say the reverse of a hydrogen fuel cell. This process was firstly demonstrated by Nicholson and Carlisle in 1800. In the 1820s Faraday clarified the principles and in 1934 he introduced the word “electrolysis”. Electrolysis was not used commercially to produce hydrogen from water until 1902 by the Oerlikon Engineering Company. During the same period, Nernst developed the high-temperature electrolyte ZrO2 with 15% Y2O3, this being the basis for solid oxide electrolysis (SOEC) and solid oxide fuel cells (SOFC). In 1951, the first commercially available high pressure electrolyser (30 bar) was presented by Lurgi. Nowadays, low temperature electrolysis technology is available with at least 13 manufactures (3 using alkaline electrolysers and 10 using polymer membranes). On the other hand, SOEC technology is still under development. This technology attracted great interest in the 1980s because of the studies curried out by Donitz and Erdle [20], where they reported the first SOEC results within the HotElly project from Dornier System GmbH using electrolyte supported tubular SOEC. In this program, single cells have been operated during long-term periods with current densities of −0.3 A cm−2 and 100% Faraday efficiency at a voltage as low as 1.07 V. In addition, Westinghouse Electric Corporation Research and Development Centre contributed to the development of SOEC. They reported Area Specific Resistance (ASR) values of about 0.6 Ω cm2 per cell in a seven-cell stack at 1003 °C [21]. Research in high temperature electrolysis has increased significantly in recent years, as will be described in the present review.

The electrochemical reactions that take part in an SOEC are the inverse reactions to those that take part in an SOFC. Cell polarization is the opposite and anode and cathode interchange their roles. In an SOEC, water acts as a reactant and is supplied to the cathode side of the cell (anode electrode in SOFC mode). Oxygen ions are transported to the anode through the electrolyte, and hydrogen is produced in the cathode side, as shown in Fig. 1.

The overall reaction of the water electrolysis is:H2O  H2 + 1/2O2

The reactions in the cathode and anode sides are:H2O + 2e  H2 (g) + O2− (cathode)O2−  1/2O2 (g) + 2e (anode)

There are mainly two types of electrolysers, depending on their operation temperature: low temperature electrolysers (LTE) and high temperature electrolysers (HTE). LTE are also divided into alkaline and proton-exchange membrane, and these devices are proven technologies that can achieve energy efficiencies of about 75% [22].

The major problem associated with LTEs is the high electric energy consumption which can degrade the competitiveness of the process. Although LTE is a mature technology, HTE presents a greater potential as the electrolysis of water is increasingly endothermic with increasing temperature. The required electrical power is reduced at higher temperatures as the unavoidable joule heat of an electrolysis cell is used in the H2O splitting process. Another advantage of the high temperature is the reduction of electrode overpotentials which cause power losses in the electrolysis cell.

The minimum electric energy supply required for the electrolysis process is equal to the change in the Gibbs free energy (ΔG):ΔG=ΔHTΔSwhere ΔH is the enthalpy change, T the temperature and ΔS the entropy change. The electrical energy demand, ΔG, decreases with increasing temperature; for example, the ratio of ΔG to ΔH is about 93% at 100 °C and about 70% at 1000 °C.

The thermodynamics of water electrolysis are given in Fig. 2. In this figure we can observe how ΔG decreases and heat energy demand (T ΔS) increases with increasing temperature at a steam pressure of 0.1 MPa. Even though total energy demand is increasing, the decrease in electrical energy demand is more noticeable, as over two thirds of the cost of electrolytic hydrogen arises from the use of electricity. Operating at higher temperature can therefore decrease the cost of the hydrogen produced, especially if the increase in heat energy demand can be fulfilled by an external heat source, such as nuclear power, renewable energy, or waste heat from high-temperature industrial processes.

As previously described, since the thermal energy required for the electrolysis reaction can be obtained from Joule heat produced within the cell as a consequence of the passage of electrical current through the cell, the electrical energy demand is reduced and therefore the H2 production price also decreases. For these devices, the thermoneutral potential is defined as the potential at which the generated Joule heat in the cell and the heat consumption for the electrolysis reaction are equal:Vf=ΔHfnFwhere ΔHf is the total energy demand for the electrolysis reaction, n is the number of electrons involved in the reaction and F is the Faraday constant. At the typical temperature of SOEC operation (900–950 °C), this voltage is around 1.29 V. At this level, the cell can be theoretically operated at thermal equilibrium with an electrical conversion efficiency of 100%. If we operate below the thermoneutral voltage (endothermic mode), the electric energy is lower than the enthalpy of reaction and heat must be supplied to the cell to maintain the temperature. In this mode of operation, electrical efficiencies above 100% could be achieved. On the contrary, if we operate above the thermoneutral voltage (exothermic mode), electrical efficiencies below 100% are obtained. However, operating in the exothermic mode (at moderate overpotentials) can present some advantages, for instance in wind farms during high-wind conditions and no electric demand. Although the electrical efficiency will be lower, high current densities can be obtained and therefore the hydrogen production rate will be higher.

According to Hauch et al. [23], in the case of H2O being fed into the system as liquid water, we should also take into account the heat demand for water evaporation at 100 °C which results in an increase in the operation voltage given by,Vvap=ΔHvapnFwhere ΔHvap is the molar energy demand for steam raising, n is the number of electrons involved in the reaction and F is the Faraday constant. As the water vaporization enthalpy (ΔHvap) is 40.65 kJ mol−1, the voltage Vvap corresponds to 0.21 V. Bearing in mind that all the energy necessary to heat up the incoming gases is obtained from the outcoming gases using a perfect heat exchanger, the thermoneutral potential is then defined as the sum of both Eqs. (5), (6), and at a temperature of 950 °C this value is about 1.5 V. In order to calculate an accurate thermoneutral point of the real stack, more complex calculations including all heat losses will be required.

The typical materials used in SOEC are basically similar to those used for SOFC. Detailed information of SOFC materials can be found in the following references [24], [25], [26], [27]. The most common electrolyte material is a dense ionic conductor consisting of ZrO2 doped with 8 mol% of Y2O3 (YSZ) [23]. This material presents high ionic conductivity as well as thermal and chemical stability at the operation temperatures (800–1000 °C). Other materials are also considered, such as Scandia stabilized zirconia (ScSZ) [28], [29], ceria-based electrolytes (fluorite structure) [30], [31] or the lanthanum gallate (LSGM, perovskite structure) materials [32], [33], as will be discussed in further sections. For the fuel electrode (cathode in electrolysis mode), the most commonly used material is a porous cermet composed of YSZ and metallic nickel [23]. Other alternative materials also used for the fuel electrode include samaria doped ceria (SDC) with nickel dispersed nanoparticles [34], titanate/ceria composites [35], or the perovskite material lanthanum strontium chromium manganite (LSCM) [36]. Finally, for the oxygen electrode the most common material used to date is the lanthanum strontium manganite (LSM)/YSZ composite [23]. Different electrode materials have also been proposed, including La0.8Sr0.2FeO3 (LSF), and La0.8Sr0.2CoO3 (LSCo) [37]; lanthanum strontium cobalt ferrite (LSCF) and lanthanum strontium copper ferrite (LSCuF) [35]; nickelate based materials such as Nd2NiO4+δ [38] or the Sr2Fe1.5Mo0.5O6−δ (SFM) perovskite [39]. Although the materials typically employed for SOEC until now have been basically the same as those used for SOFC, we should take into account that operation conditions in electrolysis mode have also changed drastically. As a consequence, several issues are emerging such as example those associated with the high steam concentrations at the fuel electrode, the high oxygen partial pressures at the electrolyte/oxygen electrode interface, or the presence of electronic conduction in zirconia based electrolytes. All these issues will be discussed in more detail in further sections.

Instead of using an oxygen conductor electrolyte, another possibility for SOEC is the use of a proton conductor. In this case, the reactions that take place in both electrodes are:H2O  2H+ + 1/2O2 + 2e (anode)2H+ + 2e  H2 (cathode)

The main advantage of using proton conductors over an oxide ion conductor is that using these systems allows the production of pure and dry hydrogen gas at the cathode, whereas when using oxide conductors, the non-utilized steam is mixed with the hydrogen produced and the use of gas separators is required. This also means that proton conducting SOECs can be coupled directly with a high temperature reactor steam cycle. However, proton conductors have been proven to be able to conduct oxide ions as well as protons at higher temperatures [40]. Mixed conduction can also be beneficial for these devices, as will be discussed in the following sections.

Section snippets

Status of high temperature electrolysis using SOFC

As previously described in the introduction section, the first significant results were reported by Donitz et al. in the 1980s [20], [22], [41], [42]. The HotElly project (High Operating Temperature Electrolysis) system was led by Dornier GmbH, and consisted of research into electrolysis single cells as well as pilot plant tests. They used the typical materials: YSZ (8–12 mol% Y2O3) electrolyte (thickness of 300 μm), LSM (250 μm) as the oxygen electrode, Ni–YSZ cermet (100 μm) for the fuel

Materials degradation issues in solid oxide electrolysis cells

As previously mentioned, long term degradation is the main issue for the viability of this technology as a practical hydrogen production system. Several long-term degradation studies have been performed to date and all of them have concluded that further improvements are required prior to commercialization.

For example, aging studies of metal supported cells at DLR [67] showed a degradation rate of 3.2% per 1000 h at 800 °C, −0.3 A cm−2 and using 43% RH of steam at the fuel electrode. Their AC

Modeling of solid oxide electrolysis cells and systems

Mathematical models are of great importance for the design of technological devices, especially if they are still under development, as in the case of SOECs. Prediction of the performance under different conditions is essential. A large number of works have been done in the last 5 years. In the present section, a short summary of these activities will be given. Additional information can be found in the following references [103], [104], [105]. Of great interest are the works of Udagawa et al.

Other applications using solid oxide electrolysis cells

Apart from hydrogen production, in very recent years SOFC cells have been proposed for a wide range of different applications [124]. An interesting approach was made by Martinez-Frias et al. [125]. They proposed a novel and highly efficient solid oxide natural gas-assisted steam electrolyser (NGASE), where natural gas reacts with the oxygen produced in the electrolysis, reducing the chemical potential across the electrolyser, thus minimizing electricity consumption. In this system, the oxygen

Summary

High temperature electrolysis using SOFC cells were presented as a promising alternative to the existing water electrolysis methods for hydrogen production. In addition, due to the chemical flexibility of those devices, it has been demonstrated that they could be used for the electrolysis of CO2 to CO, and also for the co-electrolysis of H2O/CO2 to H2/CO (syngas). In the present manuscript, current state in terms of electrolyte materials, fuel and oxygen electrodes, and material degradation

Acknowledgements

I would like to thank grant no. MAT2009-14324-C02-01 financed by the Spanish Government, and also the JAEprogram (CSIC) for financial support.

References (140)

  • H.F. Abbas et al.

    Int. J. Hydrogen Energy

    (2010)
  • M. Balat

    Int. J. Hydrogen Energy

    (2008)
  • M. Balat

    Int. J. Hydrogen Energy

    (2009)
  • R. Elder et al.

    Prog. Nucl. Energy

    (2009)
  • P.C. Hallenbeck et al.

    Trends Biotechnol.

    (2009)
  • J.D. Holladay et al.

    Catal. Today

    (2009)
  • Y. Kalinci et al.

    Int. J. Hydrogen Energy

    (2009)
  • R. Kothari et al.

    Renew. Sust. Energy Rev.

    (2008)
  • M. Momirlan et al.

    Renew. Sust. Energy Rev.

    (1999)
  • M. Ni et al.

    Int. J. Hydrogen Energy

    (2007)
  • M. Ni et al.

    Renew. Sust. Energy Rev.

    (2007)
  • J.W. Phair et al.

    Sci. Technol. Adv. Mater.

    (2006)
  • I. Valdez-Vazquez et al.

    Renew. Sust. Energy Rev.

    (2009)
  • A. Yilanci et al.

    Progr. Energy Combust. Sci.

    (2009)
  • K. Zeng et al.

    Progr. Energy Combust. Sci.

    (2010)
  • W. Donitz et al.

    Int. J. Hydrogen Energy

    (1985)
  • A.O. Isenberg

    Solid State Ionics

    (1981)
  • M.A. Laguna-Bercero et al.

    J. Power Source

    (2009)
  • K. Eguchi et al.

    Solid State Ionics

    (1996)
  • B. Zhu et al.

    Electrochem. Commun.

    (2006)
  • T. Ishihara et al.

    Solid State Ionics

    (2011)
  • F. Chauveau et al.

    J. Power Sources

    (2010)
  • Q.A. Liu et al.

    Int. J. Hydrogen Energy

    (2010)
  • W. Suksamai et al.

    Solid State Ionics

    (2007)
  • W. Donitz et al.

    Int. J. Hydrogen Energy

    (1988)
  • E. Erdle et al.

    Int. J. Hydrogen Energy

    (1992)
  • K.H. Quandt et al.

    Int. J. Hydrogen Energy

    (1986)
  • N.J. Maskalick

    Int. J. Hydrogen Energy

    (1986)
  • A. Brisse et al.

    Int. J. Hydrogen Energy

    (2008)
  • P. Kim-Lohsoontorn et al.

    Curr. Appl. Phys.

    (2011)
  • T. Schober

    Solid State Ionics

    (2001)
  • H. Iwahara et al.

    Solid State Ionics

    (1986)
  • W. Coors

    J. Power Sources

    (2003)
  • H. Iwahara et al.

    Solid State Ionics

    (1981)
  • H. Matsumoto et al.

    Solid State Ionics

    (2002)
  • T. Kobayashi et al.

    Solid State Ionics

    (2002)
  • P.A. Stuart et al.

    Solid State Ionics

    (2008)
  • T. Sakai et al.

    Int. J. Hydrogen Energy

    (2009)
  • F. He et al.

    J. Power Sources

    (2010)
  • R. Hino et al.

    Nucl. Eng. Des.

    (2004)
  • S. Hashimoto et al.

    Int. J. Hydrogen Energy

    (2009)
  • M. Liang et al.

    J. Power Sources

    (2009)
  • M. Backhaus-Ricoult et al.

    Solid State Ionics

    (2008)
  • C. Yang et al.

    Int. J. Hydrogen Energy

    (2010)
  • R.J. Gorte et al.

    J. Catal.

    (2003)
  • J. Kong et al.

    J. Power Sources

    (2009)
  • M.A. Laguna-Bercero et al.

    Solid State Ionics

    (2011)
  • Y. Bo et al.

    Int. J. Hydrogen Energy

    (2008)
  • P. Kim-Lohsoontorn et al.

    Int. J. Hydrogen Energy

    (2010)
  • F. Chauveau et al.

    Int. J. Hydrogen Energy

    (2011)
  • Cited by (882)

    View all citing articles on Scopus
    View full text