Using Solid Oxide Electrolysis Cells to Scale Green Hydrogen Production
Abstract
Un-electrifiable sectors such as shipping, steelmaking, and the chemical industry pose a great threat to de-carbonizing the global economy. Water electrolysis is currently the cheapest way of producing green hydrogen, and the fall of renewable electricity prices will continue to make it so. However, the major setback of water electrolysis is its low electrical efficiency. Water electrolysis using solid oxide electrolysis cells (SOECs) can reach near 100% electrical efficiency with increased hydrogen production thanks to favourable thermodynamics and kinetics. This paper covers SOEC technology; how it works, the challenges it faces, recent advancements/ideas in the technology as well as thoughts on hydrogen as a key for a sustainable energy future.
Contents
Introduction
1. Basic principles and functioning of water electrolysis
2. Introduction to SOECs
3. The chemical advantages of SOECs
3.1. Thermodynamics
3.2. Kinetics
4. Basic principles and functioning of solid oxide electrolysis
4.1. Overpotential
4.2. Potentiostatic operation at thermoneutral voltage
4.3. Current density
5. Choosing materials
5.1. Chemical stability and thermal expansion coefficients
5.2. Electrode requirements
5.3. Electrolyte requirements
5.4. The triple phase boundary
6. Degradation
6.1. Delamination
6.2. Grain surfaces and contact area
6.3. Formation of insulating phases
7. Recent advancements/ideas
7.1. Hybrid (proton and oxygen) electrolyte
7.2. Heterogeneous design
7.3. Infiltration of nanoparticles
Concluding remarks
Introduction
Cargo transportation, steelmaking, and parts of the chemical industry contribute upwards of 15% of global greenhouse gas emissions. All of these sectors are un-electrifiable, meaning that batteries or electricity are not viable solutions to de-carbonizing them.
Batteries are way too heavy and don’t pack enough of a punch for cargo; a 900-mile range cargo truck would need so many batteries it could hardly carry any cargo. Not to mention recharging it would take hours. Steelmaking requires iron which is extracted from its oxides using coal as the reducing agent (releasing CO₂), and some chemical processes require extremely high temperatures that are impractical to reach with electricity.
Hydrogen, however, is a plausible alternative; it has a high gravimetric energy density for shipping, can be used as the reducing agent when extracting pure iron from iron oxides, and can be used to generate extremely high temperatures for chemical processes. The big promise of hydrogen has always been that when combusted in a reaction, it releases water as a byproduct. The challenge in using hydrogen has been finding a clean source — hydrogen gas in pure form is extremely rare. Right now, 95% of hydrogen produced comes from steam methane reforming, a process using methane that releases carbon monoxide as a byproduct. To de-carbonize un-electrifiable sectors with hydrogen, it must be produced from renewable sources (AKA green hydrogen). Currently, water electrolysis using renewable electricity is the most viable method of producing green hydrogen and the fall of renewable electricity prices has made this even more relevant today:
However, mediocre efficiencies (~80%) with conventional alkaline electrolyzers and the high cost of noble metals used in polymer exchange membrane electrolyzers remain challenges for scaling commercial water electrolysis. Solid oxide electrolysis cells (SOECs) offer a solution to both of these problems as they can reach up to 100% electrical efficiency while being made of abundant materials such as yttrium, zirconium, and nickel. In this work, hydrogen production using SOECs is reviewed.
1. Basic principles and functioning of water electrolysis
Conventional electrolysis plants use alkaline water electrolyzers. These electrolyzers have a cathode, anode, electrolyte (either NaOH or KOH solution), and a DC current.
For example, assume a NaOH solution (NaOH dissolved in water) — a much more conductive fluid than water. NaOH can be thought of as an escort for the electricity provided to the water.
In this example, two electrodes — a cathode and an anode — are submerged into the electrolyte solution of Na+ and OH- ions. The anode is connected to the positive terminal of the battery and the cathode is connected to the negative terminal.
Because of the charges of the electrodes, the ions in the solution will be attracted to their opposite charges. Na+ will be attracted to the cathode (-) and OH- attracted to the anode (+). When OH- touches the anode which wants to receive electrons, OH- loses an electron and turns into oxygen gas and water by this reaction:
At the cathode which is negatively charged and wants to give electrons, something different happens. Instead of the Na+ ion taking an electron which would require a half-cell potential of -2.71V, H₂O takes on the electrons which requires -0.83V by this equation:
A reminder of terms and facts:
- 1V = 1J/1C. In other words, one volt is equal to a certain amount of energy (one joule) per a fixed amount of charges (one coulomb which is approximately 6.242 quintillion electrons).
- A positive cell potential indicates that the reaction occurs spontaneously while a negative cell potential means that the reaction occurs spontaneously in the opposite direction. Therefore, a reaction with a negative cell potential means that you must supply energy for it to react (e.g., separating water).
- A basic principle of science is that nature tends towards the reaction requiring the least amount of energy. That is why the water reaction is favoured over the sodium (Na).
Notice that when the two half-reactions are summed, a balanced water-splitting reaction is the result:
Since the top reaction has a standard oxidation potential of -0.4V and the bottom has a standard reduction potential of -0.83V, by adding these together, the overall standard cell potential is -1.23V. This means that separating water into hydrogen and oxygen gases requires a theoretical minimum voltage of 1.23V. In practice — because of resistance and other factors — the reaction requires a higher voltage.
The goal of advanced electrolyzers is to get as close as possible to the theoretical voltage. More terms to clarify:
- 1 amp (A) = 6.242*10¹⁸ electrons/second
- Power (in watts W) = Current (in amps A) * Voltage (in volts V)
- An electricity bill is paid based on how much electrical energy is consumed.
Electrical energy (in Watt-hour Wh) = Power (W) * Time (h)
Since it takes two electrons to separate one molecule of water, that means that only the amount of current supplied determines the amount of hydrogen produced, given you provide at least 1.23V. In other words — given the same current — an electrolyzer running at a voltage of 1.8V will produce the same amount of hydrogen as an electrolyzer with a voltage of 1.23V, but it will be more expensive with the former.
Efficiency = (minimum required electrical energy to produce X amount of hydrogen) ÷ (consumed electrical energy used to produce X amount of hydrogen)
Higher efficiency → less electrical energy consumed → lower cost.
2. Introduction to SOECs
Among the main water electrolyzers, there are alkaline, polymer exchange membrane (PEM), and solid oxide electrolysis cells (SOECs).
Alkaline is the current industry standard because it is old, reliable technology that uses affordable materials, but its peak efficiency is around 80%. PEM is a slightly newer technology that can reach higher efficiencies but is not very scalable because some of its parts require noble metals (platinum, iridium) which are too expensive to commercialize.
On the contrary, SOECs are made of abundant and affordable materials such as yttrium, zirconium, and nickel while also having important thermodynamic and kinetic advantages. Instead of water, they separate steam and operate at temperatures in the range of 700–1000˚C.
3. The chemical advantages of SOECs
3.1. Thermodynamics
Because SOECs operate at extremely high temperatures, they are much more electrically efficient than other types of electrolyzers. Recall that:
Efficiency = (minimum required electrical energy to produce X amount of hydrogen) ÷ (consumed electrical energy used to produce X amount of hydrogen)
The total energy required for water splitting can be represented by ∆H, where W is the electrical energy and Q is the heat supplied.
As the heat energy supplied increases, the electrical energy demand decreases. In other words, SOECs require a lower voltage than other electrolyzers to split water because of the thermal energy supplied, making them more efficient.
Given a free source of thermal energy (from nuclear reactors, solar thermal, or waste heat), SOECs can reach a 100% electrical efficiency.
3.2. Kinetics
SOECs also have a kinetic advantage being that chemical reactions occur faster in hotter environments. This increases the average hydrogen production rate which is key for scalable hydrogen production.
4. Basic principles and functioning of solid oxide electrolysis
Solid oxide electrolysis requires steam, a cathode, anode, solid electrolyte, and DC current. Two half-reactions occur: one at the cathode and one at the anode. Steam enters at the cathode which is negatively charged, and accepts electrons which split water into hydrogen gas and Oˆ(2-) ions. The oxygen ions travel through the solid electrolyte to the anode and bond with other oxygen ions to form O₂, releasing 4 electrons per molecule of O₂.
4.1. Overpotential
As mentioned before, to make electrolysis most efficient, the voltage must be as low as theoretically possible. Any extra voltage used to split water is called overpotential. The overpotential of the electrolysis cell must be minimized for the highest efficiency. The most important contributors to overpotential are:
- Electrolyte degradation
- Electrode degradation
- Electrode materials/structure
- Electrolyte materials/structure
To minimize overpotential, impedance (or resistance) must be minimized with the right materials and in every part of the cell.
4.2. Potentiostatic operation at thermoneutral voltage
The thermoneutral voltage is the voltage at which the electrical input into the cell and the total energy demand for the electrolysis reaction are equal. This includes the heat generated from the internal resistance of the cell, balancing the heat generated by the cell and the heat absorbed by electrolysis so that no cooling or heating is required. Potentiostatic operation at thermoneutral voltage is the act of maintaining the thermoneutral voltage (~1.29V for steam electrolysis) throughout the entirety of the experiment.
As well as 100% electrical efficiency, an advantage of operating at the thermoneutral voltage is that the heat distributes itself properly in the stack (multiple cells) which minimizes thermal fatigue degradation.
Although these advantages are great, other factors like degradation and current density must be kept in mind for commercial use.
4.3. Current density
Current density is defined as “the amount of electric current flowing per unit cross-sectional area of a material.” Imagine having a cylindrical wire and current flowing through it. If you were to cut that wire in half and look inside, you would be looking at its cross-sectional area.
Let’s say the cross-sectional area of your wire was 2cm² and the current flowing through it was 6A. Then the current density would be 6A/2cm² = 3A/cm². Current density, rather than simply current, is used as a measurement to compare the ability of different SOECs to let current flow.
While operating at the lowest possible voltage is a good intention, it is not always possible given the constraints of commercialization. Running at the lowest possible voltage would mean high efficiency, but if the current density were very low, the hydrogen produced would be minimal and thus non-profitable. Recall that the amount of hydrogen produced depends on the current, and so a minor amount of current flow = a minor amount of hydrogen produced.
When operating SOECs, it is desirable to both have low overpotential and high current density so as to have the highest electrical efficiency and fastest hydrogen production, respectively.
5. Choosing materials
5.1. Chemical stability and thermal expansion coefficients
The greatest challenge with SOECs is finding the perfect materials that will decrease all the different impedances to as little as possible. This goes for the fuel electrode (where hydrogen is produced), the air electrode (where oxygen is produced), and the electrolyte. Not only should the parts be chemically stable (i.e., not be oxidized or reduced such as to create unwanted insulating phases), the thermal expansion coefficients of the electrodes and electrolyte must be near identical to avoid breaking the cell.
5.2. Electrode requirements
The structural requirements for a good electrode are:
- Stable chemical structure → Does not oxidize or reduce easily.
- A porous microstructure that allows for gases to permeate through the electrode into the electrolyte
- Chemical compatibility → The electrode should not react with the electrolyte at the risk of creating insulating phases and it should not break down at high temperatures.
- Thermal expansion co-efficient match with the electrolyte → The electrode and electrolyte, when heated, should expand identically together. SOECs are made by sintering (heating a powdered material into a solid mass without liquefaction), so they are one entity. If one part of the entity expands disproportionately to another part, there is a high risk of breaking or inefficiency from unused area.
An electrode must also have the following properties:
- Ionic conductivity → lets ions pass through
- Electronic conductivity → lets current pass through
- Catalytic activity → ability to allow the water-splitting reaction to occur
5.3. Electrolyte requirements
The same challenge persists with finding excellent electrolytes that meet all the conditions while being cheap and having identical thermal expansion coefficients as the electrodes. Some criteria for a good electrolyte include:
- Chemical stability (i.e., no formation of cracks, insulating phases, or other undesirable changes)
- High ionic conductivity: the job of the electrolyte is to conduct ions.
- Low electronic conductivity: electrolytes should not conduct electrons because the electrons should be used at the triple phase boundary (see below) to create hydrogen gas.
Electrolytes can be of two types: oxygen-conducting or proton (hydrogen ion)-conducting. With oxygen-conducting electrolytes, the steam enters at the cathode, the hydrogen gas produced stays at the fuel electrode and the oxygen ions move through the electrolyte to form oxygen at the air electrode.
Proton-conducting electrolytes are less researched but have gained more interest because of their ability to be highly efficient at lower temperatures (500–800˚C instead of 700–1000˚C). Another upside of proton-conducting electrolytes is there is no need to separate steam from hydrogen because the fuel electrode is no longer where the steam enters. Instead, steam enters at the air electrode (anode) and splits into O₂ and H+. The O₂ stays at the air electrode while the hydrogen ions travel through the electrolyte and bond together at the fuel electrode, fully separated from the steam.
Currently, the most common electrolyte is oxygen-conducting made of yttria-stabilized zirconia (YSZ) because of its relatively high ionic conductivity at temperatures above 800˚C. Scandia stabilized zirconia also demonstrates excellent ionic conductivity but is too expensive.
5.4. The triple phase boundary
The triple phase boundary is the surface where the interaction of electrons, ions, electrode particles, and electrolyte particles form the hydrogen gas. Naturally, it is desirable to maximize the area of this surface by making the particles as dense as possible.
6. Degradation
Degradation is the limiting factor of an electrolyzer. Depending on the degradation process, cell impedance will be increased (causing overpotential) while sometimes also decreasing the current density. For investment costs to be decreased, the operation of SOECs at high current densities (-1 A/cm²) with low overpotential for approximately five years is desirable.
To achieve these performance metrics, the following degradation processes must be avoided:
6.1. Delamination
Delamination usually occurs with electrodes (such as strontium-doped lanthanum manganite (LSM)) which can accept excess oxygen when in oxidation environments. The excess oxygen accepted creates cation vacancies which can cause the formation of insulating cation oxides and result in delamination due to the microstructural damage near the electrode/electrolyte interface.
The loss of contact between the electrode and electrolyte is — by definition — an increase in the contact resistance of the cell which negatively affects ionic conductivity and results in overpotential.
6.2. Grain surfaces and contact area
In the electrolyte are grains which often have different structures. On the grain surfaces, there are different geometrically shaped voids of sizes 5–50nm. When these voids become larger and combine, fractures in the electrolyte can result; these new voids inside the electrolyte decrease the contact area in between electrolyte grains.
Decrease in contact area → increased contact resistance → overpotential.
6.3. Formation of insulating phases
This is a degradation process that can happen to both electrolyte and electrodes. Insulating phases, i.e. substances that are less conductive than the ones being used at first, are usually formed by oxidation of cations to form an oxide. For example, an oxygen electrode La₀.₅₈Sr₀.₄Co₀.₂Fe₀.₈O₃ forming insulating phase Co₃O₄ after hours of operation.
These oxides add ohmic resistance which decreases both electronic and ionic conductivity of the electrode and the ionic conductivity in the case of the electrolyte. The only ways to avoid this are by choosing more stable species or by better sintering, heterogeneous electrode-electrolyte, or infiltration which will be discussed later.
7. Recent advancements
Recent SOEC research has demonstrated designs to increase current density with low overpotential while minimizing degradation. A hybrid electrolyte, heterogeneous design, and nanoparticle infiltration were among my favourite.
7.1. Hybrid electrolyte
It was demonstrated that a hybrid proton and oxygen-conducting electrolyte could have better performance than conventional either-or electrolytes. They used a BaZr₀.₁Ce₀.₇Y₀.₁Yb₀.₁O₃-δ (BZCYYb) electrolyte, NdBa₀.₅Sr₀.₅Co₁.₅Fe₀.₅O₅+δ (NBSCF)-BZCYYb composite air electrode and a Ni-BZCYYb composite fuel electrode.
This simple yet novel idea achieved very high performance compared to previous electrolytes. 2.41 A/cm² with 1.3V was achieved at 700˚C, which is much higher than what has been reported for either only oxygen (0.40~0.52 A/cm²) or proton (0.08~0.21 A/cm²) conducting cells. Furthermore, no degradation for more than 60 hours was observed.
7.2. Heterogeneous design
One novel idea to circumvent the drawbacks of proton-conducting materials while leveraging the advantages is a heterogeneous design SOEC. Y-doped BaZrO₃ (BZY) electrolyte or BaZr₀.₈-xCexY₀.₂O₃-δ (BZCY) are well-known proton-conducting materials for SOECs. However, BZY has high grain boundary resistance and BZCY has not-so-great chemical stability in the long term.
Researchers always use the same proton-conducting material for the electrolyte and electrode, but Lei et al. used BZY as the electrolyte for its high density to protect the BZCY17-Ni fuel electrode from stability issues. Also, the sintering of these materials together allowed the grain growth of the BZY electrolyte, which helped reduce the grain boundary resistance of BZY.
Note: It has already been shown that the boundary of grains is much more ionically resistant than the grain itself.
BZCY17-NiO on BZY has higher sinterability and therefore larger grains, less resistance. Also, the higher sinterability makes BZY electrolyte denser and therefore more ionically conductive (since the average cross-section available for conduction increases with density).
Also, from mixing BZY and BZCY17 powders and sintering them at 1723K, Lei et al. noticed the lattice spacing of the mixture was in between the lattice spacing of BZY and BZCY17. This confirmed the interaction between the two substrates, speculating that Ce element diffused from BZCY17-NiO substrate to the BZY electrolyte layer to form BaZr₀.₈-xCexY₀.₂O₃ solid solution. From previous studies, adding Ce to BaZr₀.₈-xCexY₀.₂O₃ makes it more proton-conductive. Yet another advantage of mixing BZY and BZCY17.
7.3. Infiltration of nanoparticles
Ovtar et al. also showed an excellent way to slow down the degradation rate of Ni/YSZ electrodes without developing a completely different electrode material: infiltrating Ce₀.₈Gd₀.₂O₂-δ (CGO) nanoparticles into the electrode to achieve stable cell performance.
A long-term test was conducted on a bare Ni/YSZ electrode and a CGO infiltrated Ni/YSZ electrode at 800 °C and -1.25 A/cm². The degradation rate of the bare electrode was 699 mV/kh and only 66 mV/kh for the infiltrated one.
The great thing about catalyst infiltration is that it doesn’t have to obey the thermal expansion coefficients of the used electrode and electrolyte materials. Between the bare and infiltrated electrodes, the largest difference in the total resistance was the triple phase boundary resistance. The infiltrated electrode had much smaller TPB resistance than the bare electrode. Two reasons for this were speculated: i) CGO is a better catalyst material and ii) with the CGO infiltration, due to the presence of nanoparticles, the surface area of the TPB was increased.
Concluding remarks
As far as electrolyzers go, SOECs are the most promising technology to date thanks to their ability to use waste heat to reach 100% electrical energy efficiency. Despite their promise, more research needs to be done on minimizing degradation and proven designs have to be scaled for SOECs to become the industry electrolyzer of choice.
Even if this were to happen overnight, the “hydrogen economy” still has many barriers to overcome such as the lack of infrastructure, investment, and the increased acceptance around nuclear fission and biofuels to compete with.
It is still unclear which of the three main technologies — hydrogen, biofuels, or nuclear — will become the leading solution for un-electrifiable sectors. It is probable that SOECs will play a big part in lowering the cost and determining the winner.
As of right now, there are at least five companies commercializing this technology. Only time will tell whether the hype around hydrogen was foolish or foresighted.
Resources
Papers:
Reviews:
Other:
