Background of the invention

The fuel cell technology is over 170 years old. A fuel cell is an electrochemical device for converting chemical energy into electrical energy without an intermediate thermal energy stage. In a fuel cell, a fuel generally in the form of a gas is oxidised with an oxygen-containing gas to produce electricity directly. A typical cell has two electrodes immersed into and separated by an electrolyte. The fuel gas such as hydrogen is fed to the anode and the oxygen-containing gas is supplied to the cathode. At the anode the fuel is oxidised releasing electrons which are picked up by the anode, carried through an external circuit to the cathode where they reduce oxygen and, with hydrogen ions from the electrolyte, for water. Fuel cells can be classified into three general types:

- The high temperature fuel cell operates at temperatures higher than 600 °C using solid oxide or molten carbonate electrolytes. Besides the advantages of using relatively inexpensive electrode materials such as iron or stainless steel and not requiring anode catalysts, they suffer the disadvantages of slow start up time, inefficient use of the fuel and corrosion problems due to the high operating temperature;

The low temperature fuel cell operates at temperatures below 100°C. They use either aqueous acidic proton conducting membranes presenting high cost, short lifetimes and requiring expensive precious metal catalysts or a fairly fluid highly conductive electrolyte such as aqueous potassium hydroxide. They may offer the advantage of using inexpensive electrode catalysts, but they suffer the disadvantages of porous electrodes being easily flooded by the aqueous electrolyte, resulting in a sharp decrease of the power output of the cell, and the electrolyte being easily poisoned by impurities such as carbon monoxide and carbon dioxide; the need to efficiently remove carbon dioxide from air has practically limited this technology to aerospace applications, where pure oxygen can be supplied and where cost is not a prime consideration.

- The intermediate temperature fuel cell operates at temperatures from 100°C to 300 °C. It offers improved performance and durability over low temperature fuel cells. It often uses acid aqueous or polymeric electrolytes, but it suffers the disadvantage of requiring relatively expensive active materials for the catalyst or the electrodes semi permeable separation membrane. High pressures have been considered in order to limit the evaporation of volatile components. Phosphoric acid fuel cells have represented the first commercial successes for this source of electrical energy. These intermediate temperature fuel cells proved to be durable and reliable, though expensive and delicate to manage. A significant disadvantage resides in their relatively high temperature of operation, which prevents rapid start up and their poor tolerance to cooling and interruptions of operation. The dilatation of the electrolyte, when it solidifies tends indeed to damage tightly mounted stacks. Phosphoric acid fuel cells also require topping up acid to compensate for losses by evaporation of the anhydride near 200 °C at atmospheric pressure. The high corrosion capacity of the electrolyte also means that exclusively precious metals could be considered as catalysts. Large scale units have been installed in the United States, but the cost of the catalyst has remained a major deterrent.

Numerous fuel cell principles based on proton exchange membranes have appeared over the last few years. Recently, aqueous environment low temperature acidic fuel cells using Nation type membranes have prevailed, probably due to their flexibility and relative ease of use, as they could be started up rapidly, at room temperature. But they present serious deficiencies, such as the poor durability of the membrane. Some reports indicate lifetimes up to two thousand hours, which is unacceptable for most durable power applications. The presence of water saturated environments inside the cell favours production of carbon monoxide through the oxidation of the carbon from bipolar plates. Carbon monoxide is known to poison the platinum catalyst at relatively low operating temperatures. The strongly acid character of Nation also restricts practically the catalyst system to precious metals or precious metals alloys. Most Nation type cells present on the market are used for demonstration purposes. Their commercialisation is being hampered by poor durability, and by the high cost of the membrane and of the catalyst.

The fact remains that low and intermediate temperatures acidic fuel cells operating up to 200 °C retain most of the attention for their evident advantages in terms of flexibility, safety and limited use of advanced materials. Many phosphoric acid based gel electrolytes have been proposed as fuel cell electrolytes for the intermediate temperature devices. Each unit of a phosphoric acid fuel cell has an electrode assembly comprising an anode, a cathode and an electrolyte layer impregnated with phosphoric acid and interposed between the anode and the cathode. Usually the phosphoric acid fuel cell has a predetermined number of unit cells which are stacked together and electrically connected in series to form a stack body.

U.S. Patent No. 3,375,138 teaches that ortho-phosphoric acid gelled with carbon black permits short circuits to develop in the cell, and phosphoric acid gelled with silica were converted to high resistance solids by heating above 150°C, but these disadvantages are overcome by an electrolyte gel comprising borophosphoric acid and ortho-phosphoric acid in a mole ratio between 1 :6 and 1 :1 which maintains a constant fluidity throughout an operating temperature range from 100°C to 250 °C.

U.S. Patent No. 3,490,953 discloses high strength ion exchange membranes useful in fuel cells, formed by pre-sintering a water-insoluble hydrous metal oxide (e.g. wherein the metal is zirconium, scandium, titanium or molybdenum) or water-insoluble acid salt (e.g. zirconium phosphate) and an inorganic acid (e.g. phosphoric acid) and mixing said pre-sintered material with an inorganic acid and a material such as an alumina-silicate, colloidal silica, silica gel and the like. European Patent No. 1 ,372,205 discloses a method of operating a phosphoric acid fuel cell so as to be able to increase the temperature of the phosphoric acid fuel cell up to an operating temperature while reliably preventing the cell characteristics of the phosphoric acid fuel cell from being lowered, without increasing the size of a facility for operating the phosphoric acid fuel cell.

European Patent No. 1 ,975,947 discloses pasty composite electrolytes useful in a fuel cell and being obtained by combining at least two kinds of oxoacid groups (e.g. phosphoric or tungstic acid and sulphuric acid) and at least one alkaline-earth element, for instance a mixture of calcium sulphate and a phosphoric acid aqueous solution, wherein the content of one kind of oxoacid groups to the total amount of oxoacid groups in the composite is at most 70 mole%.

United States Patent Application Publication No. 2005/186480 discloses a gel electrolyte useful for a fuel cell, comprising an acid (e.g. orthophosphoric acid or a condensed phosphoric acid), a linear polymer (e.g. polybenzimidazole) and a crosslinkable preferably fluorine based polymer.

Chandra et al in Solid state Ionics (2002) 154-155, 609-619 discloses a method of obtaining xerogel/aerogel electrolytes by a sol-gel process starting from an inorganic metal salt precursor like sodium metasilicate.

SUMMARY OF THE INVENTION

It has now been found novel fuel cell systems which can significantly minimize the inconveniences of known acidic fuel cells. The fuel cell systems of this invention involve an electrolyte made up of specific aqueous gel compositions of partially neutralised phosphoric acid and solubilized silica which is useful for filling an inert open porosity separator. Open porosity, as used herein, refers to the fraction of the total volume in which fluid flow is effectively taking place and includes Catenary and dead-end (as these pores cannot be flushed, but they can cause fluid movement by release of pressure like gas expansion) pores and excludes closed pores (or non-connected cavities).

The specific aqueous gel compositions of the present invention are suitable to produce efficient proton conducting electrolytes which can operate at relatively low temperatures, are compatible with a wide range of catalysts including inexpensive transition metal catalysts and represent a very moderate materials cost. Their mineral character and the stability of their components offer fair guarantees of long-term durability for the electrolyte and the fuel cell itself. Their comparatively lower operating temperatures substantially eliminates disadvantages such as, but not limited to, evaporative losses of phosphoric anhydride.

DETAILED DESCRIPTION OF THE INVENTION

In a first aspect this invention relates to a gel composition consisting of : - from 30 to 89 % by weight phosphoric acid ;

- from 5 to 60% by weight silica ;

- from 1 to 7% by weight of a strong base derived from alkaline hydroxide, alkaline earth hydroxide or equivalent amounts of corresponding anhydrous oxide; and - from 5 to 30% by weight water.

In one embodiment of the present invention; the proportion by weight of phosphoric acid in the gel composition may be from 70 to 85%.

Each of the individual components of the gel composition of this invention will be described in more details hereinafter.

Phosphoric acid, also known as orthophosphoric acid, is an inorganic acid having the chemical formula H ₃ PO ₄ . Orthophosphoric acid molecules can combine with themselves to form a variety of compounds which are also referred to as orthophosphoric acid, a white solid or colorless viscous liquid depending upon the temperature (melting point : 42 °C in anhydrous form , 29 °C as a hemihydrate). Pure 70 - 85% phosphoric acid aqueous solutions are clear, colorless, odorless, non-volatile, rather viscous, syrupy liquids, but still pourable. Phosphoric acid is very commonly used as an aqueous solution of 85% phosphoric acid or H ₃ P0 ₄ . Because it is a concentrated acid, such a solution can be corrosive, although non-toxic when diluted. Because of the high percentage of phosphoric acid, at least some of the orthophosphoric acid may be condensed into polyphosphoric acids in a temperature-dependent equilibrium, but, for the sake of labeling and simplicity, the 85% represents H ₃ P0 ₄ as if it were all orthophosphoric acid.

One preferred form of silica for performing the present invention is colloidal silica, i.e. a suspension of fine amorphous, non-porous, and typically spherical silica particles in a liquid phase, preferably an aqueous phase. Usually colloidal silica is suspended in an aqueous phase that is stabilized electrostatically, and exhibits particle densities in the range of 2.1 to 2.3 g/cm ³ . Particle sizes typically range from 1 to 50 nm. The maximum concentration obtainable depends upon particle size. For example, 50 nm particles can be concentrated to greater than 50 weight% solids while 10 nm particles can only be concentrated to approximately 30 weight% solids before the suspension becomes too unstable.

One out of many useful preparation procedures for the aqueous gel composition of the invention involves a step of combining, e.g. mixing, the desired amount (e.g. 80% by weight) of a 1 15% equivalent polyphosphoric acid liquid sample and a desired amount (e.g. 20% by weight) of an aqueous colloidal silica such as, but not limited to, Ludox AS-40 a 40 weight% aqueous colloidal silica (220 m ² /g) commercially available from W.R. Grace & Co. (East Chicago, Indiana) or a similar 30 weight% aqueous colloidal silica. Heat is generated during this mixing step, while the solution clarifies and becomes fluid after a few minutes. The strong alkaline or alkaline earth hydroxide is then added in a second process step. The solution is then brought for a sufficient time to a temperature of about 120°C, for instance about five minutes, whereby the hydroxide is solubilized. Higher concentrations of colloidal silica can be added to the system at this stage. The final composition is then brought for about 2 to 20 hours at a higher temperature, for instance about 200 °C, till it reaches the desired concentration by weight of water, i.e. from 5 to 30% by weight, preferably from 8 to 20% by weight, more preferably from 10 to 15% by weight. A pasty non mobile homogenous transparent gel is then formed. Without wishing to be bound by theory, it is assumed that water acts, among other functions, as a plasticizer. The major ingredients of the electrolyte are relatively weak polar acids, strongly linked via a so-called hydrogen bonding, which would solidify in the absence of water. Water also confers a high proton conductivity to the resulting electrolyte gel composition. Higher concentrations of water in the gel composition typically help generation of more power in a related fuel cell by diminishing the proton impedance of the system. However, the preservation of a high water concentration in the gel requires a comparatively high water partial pressure in incoming gasses. This may limit the partial pressure of the involved fuel. Higher concentrations of water will also fluidify the gel and may cause mobility of the gel inside the cell and ultimately gas percolation across the separator. The type of fuel cell catalyst, as well as the manageable gas throughput, shall command the best compromise with respect to water concentration. The presence of silica, for instance colloidal silica, provides the gel composition of the present invention with higher water retention properties at temperatures above its natural boiling point, hence lower partial pressure at equilibrium in the overhead space. It also confers the gel composition with an inherent viscosity which prevents uncontrolled diffusion of phosphoric acid across the cell or cell stacks. It also provides an appropriate physical barrier to the gases after impregnation of the porous separator. Suitable operating temperatures for the fuel cell system of this invention range from 101 °C to about 150°C, preferably from about 120°C to about 140°C, at normal ambient pressure. A large variety of relatively inexpensive construction materials can be used for the porous separator at such temperatures. Risks of evaporative losses of phosphoric anhydrides are inexistent at such temperatures and consequently no addition of phosphoric acid is required during operation, by contrast with most conventional phosphoric acid fuel cells operating in nearly anhydrous conditions close to 200 °C. Higher or lower (with respect to ambient) total vapour pressures can also be operated in the context of this invention, but the operating temperature should most preferably range above the water boiling point in order to limit risks of condensation of water inside the porous gas diffusion section of the electrodes. Condensation of water inside the electrodes damages the performance of the cell, as is often the case with Nation type proton exchange membrane cells where operating conditions are often close to dew point, at temperature below water boiling point for a series of operational reasons, such as the membrane inherent stability.

The addition of a low concentration, i.e. from 1 to 7% by weight, preferably from 1 to 5% by weight, more preferably from 1 .5 to 4% by weight, of a strong alkaline or alkaline earth hydroxide base, to the electrolyte gel composition of this invention, is meant, among other reasons, to moderate the corrosivity of phosphoric acid aqueous solutions at operating temperatures of the fuel cell system.

The presence of such concentrations of a strong alkaline or alkaline earth hydroxide base has been found to surprisingly improve the power generation of the fuel cell, in particular within the operating conditions described herein. Without willing to be bound by theory, one can assume that this behaviour is related to a lower proton impedance of the phosphoric acid solution based electrolyte. The strong proton interaction with relatively weak but highly polar acids produces a three dimensional network that hinders mobility and slows down exchanges. The presence of a strong alkaline or alkaline earth base producing discrete ions may help breaking to some extent the three dimensional network created by proton bonding and facilitate proton mobility. Beyond a certain concentration of the strong alkaline or alkaline earth base, free proton concentration becomes low enough to hinder proton conductivity, which explains the optimum concentrations of alkaline or alkaline earth base selected for this invention. Suitable strong bases are bases which completely dissociate in water into the cation and OH ^" (hydroxide anion). The following hydroxides or their equivalent anhydrous oxides of most Group I and Group II metals usually are considered to be strong bases suitable for the purpose of the present invention:

LiOH - lithium hydroxide

NaOH - sodium hydroxide

KOH - potassium hydroxide

RbOH - rubidium hydroxide

CsOH - cesium hydroxide

Ca(OH)2 - calcium hydroxide

Sr(OH)2 - strontium hydroxide

Ba(OH)2 - barium hydroxide

It should be understood that one or more such strong alkaline or alkaline earth bases can be used in the electrolyte gel composition of the invention. The optimal amount of such one or more strong alkaline or alkaline earth bases, although always included within a range from 1 to 7% by weight, may depend upon a range of parameters such as the selected type of alkaline or alkaline earth base(s), the respective amounts of phosphoric acid and colloidal silica, the specific grade of colloidal silica (in particular the particle size, and the kind of hydroxide stabilizer commonly incorporated by the silica manufacturer) and the like. The skilled person is able to determine such optimal amount through a series of simple experimental trials. Water concentration in the gel is also an important parameter in the definition of an electrolyte gel composition of the invention. A particularly preferred feature of this invention resides with supplying the fuel and oxygen gases with an adequate partial pressure of water in order to maintain a substantially constant concentration of water in the electrolyte. The partial pressure of water in incoming gases can be, for instance, adjusted by setting the temperature of gas humidifiers. Other techniques such as, but not limited to, controlled injections of water into incoming gases are also possible. Condensation inside the gas ducts should be avoided by keeping their wall temperature above the dew point. The water concentration of the electrolyte composition of the present invention can be monitored, for instance, via its proton or electrical conductivity. The composition of the evolving gases can also provide suitable feedback for the adjustment of the gas humidifying system.

The preferred concentration of water in the electrolyte composition can be readily determined according to a number of factors including power and durability requirements. Elevated temperatures and water concentrations enhance power. They are also known to shorten the lifetime of carbon electrodes and of catalysts, via the production of carbon monoxide (and dioxide) through the oxidation of the carbon components of the cell, such as the bipolar plates. The utilisation of hydrocarbon fuels or petrochemical origin hydrogen contaminated by carbon monoxide from its manufacturing process, hence ahead of the cell, can favour higher operating temperatures, as the poisoning of platinum catalysts will be more limited.

Lower operating temperatures shall normally improve durability of the fuel cell electrolyte, but at the same time shall limit power generation. In circumstances where reaching high power generation shall be attempted by increasing the water content in the electrolyte, this will require higher water partial pressure in incoming gases and consequently limit the partial pressure of fuel gases. Enhanced gas flow and catalyst activity shall be important under such circumstances. The open porosity inert separation membrane of the fuel cell system of this invention can be made up of acid resistant mineral and/or organic materials. One may, for instance, apply any polymeric woven or non-woven fabric withstanding the above-mentioned cell operating conditions, in particular temperature and pressure. Examples of such suitable polymeric fabrics include, but are not limited to, polyphenylene sulfide and polytetrafluoro-ethylene and mixtures thereof. A suitable porous separator shall preferably have filtration particle retention characteristics above one micron. The thickness of the porous separator fabric can typically range from about 20 microns up to about 5 mm, for instance from about 50 microns to about 2 mm, or from 0,1 to about 1 mm, depending upon the size of the cell and of pressure differences to be operated. Similar pressures shall preferably be operated on either side of the membrane.

The preparation of the electrodes separator involves impregnating the abobe-mentioned porous separator with an electrolyte gel composition according to any one of the embodiments of the invention, preferably without any volume of gel in excess of the apparent volume of the porous separator. The catalyst bearing electrodes can be applied in direct contact with the separator. Physical contact with the gel shall be established, but the porous electrodes shall preferably not be soaked with the electrolyte gel composition. The resulting assembly can then be inserted between conventional fuel cell bipolar plates for gas supply and with conventional electrical connections. Adequate tightening force shall preferably be applied in order to avoid limiting the porosity of the gas diffusion zone of the electrodes, whilst maintaining the overall fluids tightness of the device. In one embodiment of this invention, one or both electrodes may be made from precious metals such as, but not limited to, platinum, gold or rhodium, or alternatively from non-precious metals such as but not limited to, cobalt, nickel and alloys thereof.

The following examples are given for the purpose of illustrating various ways of performing the present invention. They are merely describing specific embodiments within the framework of the general description hereinabove, together with the resulting advantages, but are not intended to limit the scope of the invention, which is defined only by the appended claims

EXAMPLES

Fuel cell assembly materials and methods The following experiments involved A Pragma Industries 25 cm ² cell equipped with heating cartridges and a temperature control station. A Zahner Ennium PP201 potentiostat was operated for monitoring the cell performance.

The platinum electrodes were supplied by Quintech (Goppingen, Germany). References BC-H25-05F (0,5 mg/cm ² platinum) and BC-H25-10S (1 mg/cm ² platinum) were applied respectively on the anode and the cathode.

The cobalt electrodes were made from a GP4 cobalt on carbon black catalyst powder supplied by GP Materials (Aix-les-Bains, France) as follows. The catalytic ink was prepared by suspending 5,5 % by weight of the catalyst powder in ethanol. The suspension was then supplemented with 6,0 % by weight of an electronic grade polytetrafluoroethylene emulsion in order to facilitate filtration and adhesion to its substrate. This suspension was then filtered under vacuum on a Buchner device equipped with a 400 mg carbon cloth supplied by Quintech (Goppingen, Germany) under the reference CC- 060T, on top of a filter paper. After drying at ambient temperature, the catalyst loaded graphite cloth makes up the electrode. 21 mg/cm ² GP4 are used as replacement for platinum, which equates to about 0.84 mg/cm ² actual cobalt based catalyst complex

The separator was made from a 550 g/m ² grade non woven polyphenylene sulfide fabric having a 1 .7 mm thickness supplied by T.T.L France SA (Sausheim, France) . The separator was impregnated with a phosphoric acid based gel composition as defined herein, and the electrodes were assembled and applied manually on the latter.

Example 1 A Platinum/Platinum catalyst system was operated at 120°C. The electrolyte contained (by weight) 79% phosphoric acid, 2% calcium hydroxide, 9% colloidal silica (Ludox AS-40 commercially available from W.R. Grace & Co) and 10% water. The resulting potential (V) versus current diagram (mA) represented in figure 1 shows a power density comparable to conventional phosphoric acid fuel cells.

Example 2

This cell was operated under the same conditions as in example 1 , except that a cobalt catalyst was used instead of platinum at the cathode. The potential versus current diagram represented in figure 2 shows that cobalt can advantageously replace platinum at the cathode without significant losses in performance. This cell was maintained in operation for several hours. This experiments thus demonstrates that significant cost savings can be achieved through the present invention on both the catalyst system (cobalt is less expensive than platinum) and the proton conductive separator.

Example 3 (comparative)

This cell was operated in same conditions as in example 2, except that no calcium hydroxide was added to the electrolyte gel composition. The potential versus current diagram represented in figure 3 shows that a significantly lower power is being generated. The slope of the curve is significantly higher than in example 2, also indicating a higher impedance for the circuit.

Example 4

The following data confirm the critical role of a strong alkaline hydroxide for the power generation of an electrolyte based on a phosphoric acid aqueous solution operating at relatively low temperatures. They confirm that improved power generation is related to the enhancement of proton conductivity of the electrolyte, through the measurement of the proton conductivity of phosphoric acid aqueous solutions at different levels of neutralisation. Measurements were performed according to the technique described by Molenberg et al in Microwave and optical technology letters (201 1 ) 53(9).

The four following samples were prepared:

109iw01 : track etched polyimide membrane impregnated with a composition comprising 96% by weight phosphoric acid and 4% by weight NaOH.

109iw02: glass fiber non woven fabric impregnated with a composition comprising 96% by weight phosphoric acid and 4% by weight NaOH.

109iw03: glass fiber non woven fabric impregnated with a composition comprising 90% by weight phosphoric acid and 10% by weight NaOH. 109iw04: glass fiber non woven fabric impregnated with 98% a composition comprising 98% by weight phosphoric acid and 2% by weight NaOH.

A polytetrafluorethylene plate was used as a blank reference. The non woven glass fiber fabrics were filter type tissues presenting a particle retention of about 100 nm. The pore size of the track etched film was 60 nm diameter. The diagram shown in figure 4 provides data on the proton conductivity observed at 140°C under 30% relative humidity. The sample with 4% NaOH displays by far the highest proton conductivity, as compared to lower or higher concentrations of NaOH. Comparable results were obtained at room temperature. This experiment also illustrates that the non woven material displays a significantly higher conductivity than the track etched polyimide film.