GB2305169A - Solid oxide fuel cells - Google Patents

Abstract

A material for use in solid oxide fuel cells comprising a nickel-chromium oxide spinel. The material is electronically conducting and may also comprise free nickel oxide. In another embodiment of the present invention there is provided a wire for use in fuel cells wherein the wire may be surface coated, preferably by being buried within electrode material. Yet another embodiment of the invention provides a fuel cell stack wherein individual cells are connected together electrically as anode-anode and cathode-cathode pairs using porous interconnectors.

Description

"Solid Oxide Fuel Cells" The present invention relates to solid oxide fuel cells and fuel stacks.

Fuel cell stacks and the housing thereof are employed for the production of electrical power using a variety of fuel gases and oxidants. There are three main stack designs as follows: 1. Tubular (as manufactured by Westinghouse) Advantages are ease of sealing.

Disadvantages are relatively low power density, necessity of support tube, high cost.

2. Monolithic (as manufactured by Argonne National Laboratory) Advantages are high power density, 'one-piece' manufacture, and moderate cost.

Disadvantages are sealing difficulties, manufacturing without component fracture, restrictions in manufacturing flexibility.

3. Planar (as manufactured by Ceramatec, Siemens etc) Advantages are high power density, easier control of manufacturing stages, and moderate cost.

Disadvantages are stack configuration complexity and sealing.

Within a solid oxide fuel cell (SOFC) the cathode is usually made from doped lanthanum manganite, the anode from a cermet of nickel in stabilised zirconia, and the interconnect is made from doped lanthanum chromite. In planar SOFC the electrolyte is commonly yttriastabilised zirconia and is in plain sheet form.

A SOFC as described has the disadvantages that 1. The cathode material (lanthanum manganite) is easily poisoned by chromium based vapours causing loss of electronic conduction. These vapours arise from chromium based interconnects of the commonly employed corrosion resistant alloys.

2. The anode material (nickel cermet) during its life in the fuel cell can see repeated cycles of oxidation (to nickel oxide) and reduction back to nickel, especially during warm-up and cool-down conditions. The change from nickel to nickel oxide (and the reverse) is accompanied by a volume change, causing anode-electrolyte delamination, or cracking of the electrolyte itself leading to cell failure.

3. Interconnect sheet material (where used in series arranged stacks of cells) is based mainly on doped lanthanum chromite. this material is extremely expensive, requires excessively high firing temperature for is densification and possesses rather poor electronic conductance. The requirement for lanthanum chromite hampers many conventional approaches to stack construction due to its high cost and difficult (extremely high temperature) processing and poor operational performance.

4. The electrolyte (stabilised zirconia) has to be made as thin as possible (about 20 microns) to work at the reduced temperatures preferred by interconnect alloys (ie 800"C as opposed to 1000"C). This leads to loss of strength, increased fragility and considerable difficulties in handling.

The present invention aims to provide improved solid oxide fuel cells and fuel cell stacks.

According to one aspect of the present invention there is provided a material suitable for use in fuel cells, the material comprising a nickel-chromium oxide spinel and being electronically conducting.

The material may further comprise free nickel oxide.

The material may further comprise oxidation resistance metallic particles. Suitable particles include silver, platinum, Nichrome (Trade Mark) and other oxygen resistant high temperature resistant alloys.

In one embodiment the invention provides the use of the material in a porous form as an anode or cathode in fuel cells.

In an alternative embodiment the material may be used in a dense form as an interconnect in fuel cells.

The present invention also provides an electrolyte support structure, the structure comprising a thin web, comprising electrolyte material.

In a preferred embodiment the structure comprises a honeycomb wall structure of electrolyte material which may be supported on a cell such as fuel cell stack cell.

Preferably the structure is at least 15 microns thick.

Most preferably the structure is 25-30 microns thick and is stable to operation at 800 C.

The invention further provides supported thin electrolyte films.

Suitably the electrolyte support structure may comprise stabilised zirconia or ceria-gadolinia.

In yet another aspect of the present invention there is also provided wire for use in a solid oxide fuel cell wherein the wire is surface coated.

The invention further provides solid oxide fuel cells and solid oxide fuel cell stacks containing wires wherein at least some of the wires are surface coated.

Preferably the wire is surface coated by being buried in electrode material.

In particular embodiment the wires contain chromium and they are coated with doped nickel chromite.

Alternatively, the wires may be coated with high temperature oxidant resistant cobalt base brazing alloys.

The alloy coated wires may be used in a fuel cell or fuel cell stack comprising lanthanum strontium manganite (LSM) cathodes wherein the coating reduces migration of chromium from wire into LSM.

In one embodiment of the invention there is provided a monopolar fuel cell stack comprising solid oxide fuel cell wherein porous alumina is loosely sandwiched between surfaces of cells.

In yet another aspect of the present invention there is provided a fuel cell stack wherein individual cells are connected together electrically as anode-anode pairs and cathode-cathode pairs, using porous interconnectors.

Preferably each fuel cell comprises a planar solid electrolyte; on one side of the electrolyte a planar anode; on the other side of the electrolyte a planar cathode; and means for providing fluid fuels to the faces of the anode and the cathode remote from the electrolyte.

Preferably the electrolyte comprises stabilised zirconia.

Preferably the anode comprises a porous nickel cermet.

Preferably the cathode comprises a porous lanthanum manganite.

Further, according to the invention there is also provided a cell stack wherein the cells are arranged such that the anode of one cell faces the anode of the next; the cathode of one cell faces the cathode of the next, and the interconnectors are mechanically sandwiched between electrode pairs.

Preferably the inter-connectors are actually joined to the electrodes.

Preferably the inter-connectors are actually joined to the electrodes by brazing to give improved electrical contact.

Preferably spacer bars separate the cells.

Preferably the spacer bars are made of a ceramic material.

Preferably the thickness of the spacer bars matches the thickness of the interconnectors. They may be cemented down to form a gas tight seal.

Examples The invention is illustrated in non-limiting manner by reference to the following Examples and the figures wherein: Figure la shows series connected arrangement; Figure lb shows parallel connected arrangement; Figures 2a & 2b show honeycomb arrangement of the electrolyte; Figure 3 shows the basics of a monopular fuel cell stack; Figure 4 illustrates a typical fuel cell stack; Figure 5 illustrates a plan view of a manifolding arrangement; Figure 6 shows an arrangement whereby the exhaust fuel gas is fully used up; Figure 7 illustrates seven cells in a monoplanar arrangement; Figure 8 is a plan view of the stack as shown in Figure 7; and Figure 9 illustrates the construction of a fuel cell stack.

Example 1 relates to figures 1-4.

The operating principles of planar solid oxide fuel cell stacks is indicated in Fig la for a series connected, and Fig lb for a parallel connected arrangement. In both cases, gas is presented to the anode (An), air to the cathode (Ca), and the concomitant reactions via the electrolyte (El) give an open circuit voltage of about 1 volt per cell. In the series arrangement, this results in an open circuit voltage of 3 volts, but requires an extra interconnect layer (lc) to join the cathode of one cell with the anode of the next, and keep the air and gas streams separate. In the parallel arrangement, no interconnect is used or needed, and the cells are arranged such that anode faces anode, and cathode faces cathode. This simplifies gas and fuel fee arrangements, but, for the three cells shown, results in only one volt open circuit.Irrespective of the arrangement, however, a similar wattage will be delivered. The connections between the cell in both arrangements are shown as metallic wires, although sculpted metal plates (Siemens design) or ceramic materials (lanthanum chromite) are traditionally used. Only three cells are shown in each arrangement, but clearly there is no limit to the number of cells which may be connected together.

The invention described herein relates to solid oxide fuel cells (SOFC), novel materials for use therein, the design and layout of individual, or sets of, components for use in SOFC.

Due to the necessity of the cathode to tolerate chromium vapour in stacks employing chromium based alloys or ceramics, it was necessary to find a material which would not be poisoned in a similar manner to that exhibited by the conventionally used lanthanum manganite. By incorporating metallic oxidationresistant particles in amounts necessary to exceed the percolation density, into the spinel of the invention, a highly electronic conducting cermet results. It is then possible to join wires to this cermet, either by embedding their ends in the material, or by brazing them in place, such that low electrical resistance connections can be made between cells and interconnects (series stacking), and between individual cells themselves (parallel stacking). Making the cermet porous is the final step in producing the cathode.

This cermet has application also as an anode material, the spinel part itself, unlike the traditional zirconia matrix of the zirconia-nickel cermet, being electronically conducting, thus promoting further electrical conductance in the final spinel cermet.

Present drawbacks of the conventional nickel-zirconia cermet are its requirement to undergo redox reactions during start-up or close-down conditions. These cycle the nickel to nickel oxide and back, with each change accompanied by a volume change, which eventually results in delamination from the electrolyte, or fracture of the electrolyte. Replacing the system with a cermet spinel, where the metallic network particles will remain largely chemically unchanging in both the oxidation or reduction conditions, will avoid the corresponding dimensional changes, and deleterious effects on the electrolyte. Again, the cermet needs to be manufactured to be porous, to allow percolation of the gases up to the electrolyte interface.

As the spinel cermet can tolerate both oxidising and reducing conditions, it has potential for use as an interconnect. Unlike both cathode and anode, the interconnect requires to be fully dense to prevent contact of the air and gas feeds (see Fig la - series stacking arrangement).

Nickel-chromium oxide may be formed by mixing nickel oxide and chromium oxide in equal molar ratios: NiO + Cr2O3 NiCr2O4 The nickel-chromium oxide may then be pressed (approximately 201b/in2) to form a disc and fired in the air at approximately 1200"C.

To produce a porous form of the nickel chromium oxide material, large particle starting materials may be used or plastic particles may be incorporated which melt and leave pores when the material is fired.

To produce a dense form of the nickel chromium oxide material, finer powder starting material may be used, greater pressure may be used and/or the material may be fired at a higher temperature or for a longer time than is necessary for the porous form.

The use of metallic materials in the stack design is favoured if the operating temperature can be lowered.

This, unfortunately, for any given electrolyte, results in slower ionic transfer, but the situation may be mitigated by reducing the path length through which the ions have to pass. This then requires the electrolyte to be as thin as possible, with around 25 microns being a suitable thickness for acceptable operation at 8000C.

The electrolyte has then little mechanical strength, but the situation may be improved by 'encasing' it in a honeycomb-wall structure of the same material. This is illustrated in plan view in Figure 2a, with a typical cross section being shown in Figure 2b. The cell walls will then support the thin connecting web. The 'honeycomb', after firing to shape, may be 'filled' with electrode material on each side, with one or more connecting wires being 'attached' to each honeycomb cell unit.

According to the invention the cathode can be made as a cermet, and being based on an electronically conducting nickel-chromium oxide (spinel), cannot be "poisoned" by chromium bearing materials. Further, the introduction of oxidation resistant alloy particles into this spinel improves its electrical conductivity far beyond that of the conventional lanthanum manganite, and also allows connection of chromium bearing interconnect wires directly into the cermet, or by brazing them onto the cermet. Chromium bearing alloys, such as the sculpted material used by Siemens, can be used in close contact with the cathode without cathode poisoning effects.

A similar cermet to that described above can be used for the anode. As the metallic particles in the cermet are oxidation resistant, there is no expansion change resulting from redox cycles. In addition, the spinel matrix is electronically conducting giving improved performance over the non-conducting zirconia matrix of the conventional anode cermet.

The cermet spinel material used for anode and cathode, can also be used, in dense form, as an interconnect, giving a cheap, highly conducting, and easily processed alternative to the conventionally used lanthanum chromite interconnect.

The integral honeycomb-wall/web structure design allows the production of supported thin electrolyte films which give handleability and processability for fuel cell operation at reduced temperatures. Being modular, multiple units can be assembled with common gas feeds, as indicated in the British Patent Application No 9502970.8. The stack design is sufficiently flexible to allow the use of other planar cell concepts, such as supported electrolyte designs for low temperature operation.

Example 2 Example 2 relates to Figures 3, 4, 5 and 6. Figure 3 shows the basics of a monopolar fuel cell stack, containing three active cells. All the cell components are solid, and comprise a central oxygen-ion-carrying electrolyte (typically stabilised zirconia), with on one of its sides an anode (generally a porous nickel cermet), and on its other side a cathode (typically a porous lanthanum manganite). Oxygen gas (or air) is presented to the cathode, becomes ionised by taking up electrons, and diffuses through the electrolyte lattice to the anode. Here it reacts with hydrogen gas in contact with the anode, to form water, and give up its electrons, thus creating a voltage difference across the cell.

Figure 3 shows a method of connecting these individual cells together electrically, as anode-anode pairs, and cathode-cathode pairs, using porous interconnectors.

Wires are joined to these interconnectors as shown, and supply approximately 1 volt (open circuit) at high current. The arrangement may be imagined as a series of batteries all connected in parallel.

Figure 4 illustrates a typical stack hardware, based on the schematic of Figure 3.

The anode/electrolyte/cathode assembly comprising the cell, is now, for simplicity, shown as one block. As indicated above, these cells are arranged such that anode of one cell faces the anode of the next; likewise cathode faces cathode. The interconnect is mechanically sandwiched between these electrode pairs, and where possible actually joined to the electrode by, for example, brazing to give improved electrical contact.

As it is essential to keep the air an combustive gas streams apart, the arrangement allows the gas to be presented to the front face, say, and leave at the opposite face, having passed through the porous interconnector. The air, on the other hand is presented to the face lying at 90 degrees to the face used for the gas, and likewise emerges at its oppose face. Spacer bars, conveniently made of ceramic material, lie along the sides of the cell plate assemblies, to separate them. Their thickness is chosen to match the thickness of the interconnect, and they are cemented down to form a gas-tight seal.

Figure 5 illustrates a plan view of a manifolding arrangement, again probably of ceramic, and joined to the fuel cell 'block' to form four gas-tight channels.

Block and channels would further be cemented to an endslab which would act to close off the channels at one end, gas and air being fed in through the remaining open ends. In practice, weight would be applied mechanically to compress the stack, to ensure good electrical contact of all the layered element. Wires from each interconnector would be joined to a larger conductor wire which would pass up the respective gas or air channels, and pass out through seals to supply the electrical power. A convenient size of stack would be a cube of side 50mm, and would have approximately 25 cells, and deliver about 200 watts at 0.7 volts. These building blocks would then be connected electrically in series, to give higher voltage outputs, for example 18 connected in series would give approximately twelve volts and delivering 3.6 kilowatts.

Figure 6 shows an arrangement whereby the exhaust fuel gas is fully used up by passing it through two or three 'scavenger' cells. As these cells are supplied with already depleted gas, their output will be less than the 'main' cells. Even in a worst case, where one of these cells is supplied with a totally exhausted gas, there will be no 'shorting' effect presented to the parallel arrangement, as the typical working cell voltage (0.7 volts or less) will be too low to drive a reverse reaction of the electrolysis of the water (needlessly wasting power), to produce hydrogen. This allows a degree of flexibility in the system to allow for changes in fuel gas flow patters, and the effect on those from varying power demands.

Example 3 relates to figure 7, 8 and 9 wherein: Figure 7 illustrates seven cells in a monoplaner arrangement.

Figure 8 is a plan view of the stack as shown in Figure 3.

Figure 9 illustrates the construction of a fuel cell stack.

This example relates to a compact, modular, planar SOFC stack, comprising a minimum of seven, 50mm square 'PEN' units, no lanthanum chromite or other interconnect, and developing approximately 50 watts at 5 volts (based on a current density of 0.5 amps/sq.cm.).

This design has been carefully engineered to give a practical, easy to build, fuel cell stack, without the expensive, difficult to process, and poorly performing traditional lanthanum chromite interconnect. In fact, as no interconnect at all is required, each PEN is self-contained, and truly modular, and it s performance can be optimised independent of other stack elements.

Because this design uses a buried-wire-in-electrode technique, in principle akin to that used for heating elements encapsulated in ceramic, and because these wires generally contain chromium, use may be made of a porous cathode made of doped nickel chromite, a more compatible material. Alternatively, surface coating of these wires would allow their use with LSM cathodes, the coating reducing the migration of chromium from the wire into the LSM, with its resulting 'poisoning' effect.

Figure 7 indicates, schematically, the arrangement proposed, and shows seven PEN cells, arranged in a novel manner, with the anode of one call facing the anode of the next cell, and, likewise, cathode facing cathode. This allows a common gas feed to two similar electrode surfaces, and considerably aids gas sealing and manifolding arrangements. Porous alumina felt is loosely sandwiched between each surface, and acts as a gas diffuser, and electrical separator. For ease of visualisation, the wire terminations or lead-outs from the anode and cathode of any given PEN, are shown lying in the same orientation. As shown in Figure 8, they are, in fact, offset by ninety degrees. Figure 1 also shows the connection strategy, in this case a cells-inseries arrangement, giving about 5 volts (7 x 0.7 volts), with the cells under load. These connections are made externally, and can be re-arranged at will for other desired voltages. At 0.5 amps/sq.cm. of cell surface, 10 amps per 'PEN' is expected, yielding a 50 watt output from a stack less than 3 cms high. this represents an extremely high power density (0.7 watts/cc or 0.7 MW/m3) Figure 8 is a plan view of the stack, with the end cap removed, and shows 'PEN' with anode face upwards, supported on the Ythrium, Strontium, Zirconia electrolyte. Although there are many different arrangements possible for current collection from the electrodes, shown is a continuous wire embedded in the electrodes, and co-fired with the structure.From work on metallic foams at least for the nickel-chromium composition available, it was shown that non-uniform composition, high surface area, and localised thin connection webs, led to unacceptable lifetimes in oxidising atmospheres at 900"C. Commercial wire compositions, on the other hand, which have been designed for use as furnace windings, are available in thin sections, with temperature oxidation resistance to at least 1200 . Considerable scope exists within this programme to optimise the pitch and wire diameter of the flat winding design, in conjunction with electrode thickness and electrical conductivity. It should also be possible to heat the stack to working temperature by powering the windings. This would need to be carried out in a controlled manner to avoid thermal gradients causing structural failure.

Figure 9 shows the probably construction. Again, for clarity, the other walls forming the plenum chambers have been removed to show the 'PEN' support housing only, and the gas entry ports, out of which the current collector leads are fed, and thence either up and through the top plate, or individually through the plenum other walls. The hidden faces of the stack have an identical arrangement, and the design is contrived as a cross-flow. Glass bonding, ceramic to bonding, and ceramic metal seals are used in this construction.

While the invention relates particularly to solid oxide fuel cells (referred to as SOFC in this document), the new materials should have application to other types of fuel cells (polymeric, molten, carbonate, phosphoric acid etc), and also to gas sensors, especially high temperature sensors.

Claims (29)

Claims

1. A material for use in fuel cells comprising a nickel-chromium oxide spinel.

2. A material as claimed in Claim 1 further comprising free nickel oxide.

3. A material as claimed in Claim 1 or 2 further comprising oxidation resistant metallic particles.

4. A material as claimed in claim 3 wherein the oxidation resistant metal is selected from a group including silver, platinum or Nichrome (Trademark).

5. A material as claimed in any preceding claim wherein the material is porous.

6. An interconnect material for use in fuel cells comprising a material as claimed in any of Claims 1 to 5.

7. A material for use as a support structure for use in fuel cells comprising of a thin web of electrolyte material.

8. A material as claimed in claim 7 wherein the electrolyte material is of a honeycomb structure.

9. A material as claimed in Claim 7 or 8 having a thickness of at least 15 microns.

10. A material as claimed in any of Claims 7 to 9 having a thickness of from 25 to 30 microns.

11. A material as claimed in any of Claims 7 to 10 comprising stabilised zirconia or ceria-gadolinia.

12. A wire for use in solid oxide fuel cells and cell stacks wherein the wire is surface coated with an electrode material.

13. A wire as claimed in Claim 12 wherein the wire contains chromium.

14. A wire as claimed in Claim 12 or 13 wherein the coating consists of doped nickel chromite.

15. A wire as claimed in Claim 12 or 13 wherein the coating consists of high temperature oxidant resistant cobalt base brazing alloys.

16. A fuel cell stack comprising individual fuel cells wherein the individual fuel cells are connected together as anode-anode and cathode-cathode pairs.

17. A fuel cell stack as claimed in claim 16 wherein the connectors are porous.

18. A fuel cell stack as claimed in claim 16 or 17 comprising a planar solid electrolyte wherein on one side of the electrolyte there is a planar anode and on the other side of the electrode is a planar cathode.

19. A fuel cell stack as claimed in any of Claims 16 to 18 further comprising means to provide fluid fuels to the faces of the anode and the cathode remote from the electrolyte.

20. A fuel cell stack as claimed in any of Claims 16 to 19 wherein the electrolyte comprises stabilised zirconia.

21. A fuel cell stack as claimed in any of Claims 16 to 20 wherein the anode comprises a porous nickel cermet.

22. A fuel cell stack as claimed in any of Claims 16 to 21 wherein the cathode comprises a porous lanthanum manganite.

23. A fuel cell stack as claimed in Claim 16 wherein the individual cells are arranged such that the anode of one cell faces the anode of the next cell; the cathode of one cell faces the cathode of the next cell; the interconnectors are mechanically sandwiched between electrode pairs.

24. A fuel cell stack as claimed in Claim 16 or 23 wherein the interconnectors are joined to the electrodes.

25. A fuel cell stack as claimed in Claim 24 wherein the interconnectors are joined to the electrodes by brazing to give improved electrical contact.

26. A fuel cell stack as claimed in Claim 25 wherein spacer bars separate the cells.

27. A fuel cell stack as claimed in Claim 26 wherein the spacer bars are made of a ceramic material.

28. A fuel cell stack as claimed in Claim 26 or 27 wherein the thickness of the spacer bars matches the thickness of the interconnectors.

29. A fuel cell stack as claimed in any of Claims 26 to 28 wherein the spacer bars may be cemented down to form a gas-tight seal.