Fuel cells with a proton-conducting membrane that use hydrazine as a fuel were mentioned in Section 4.12.2. But long before 2003, when the first reports on these cells began to appear, hydrazine–oxygen fuel cells with alkaline electro- lytes had been the subject of research and engineering work. Hydrazine N2H4is a strong reducing agent that can be used as a rocket fuel. It can be synthesized from ammonia and can exist in the form of stable aqueous solutions in concentrations of up to 100% hydrazine hydrate (N2H4H2O or N2H5OH).
6.2 ALKALINE HYDRAZINE FUEL CELLS 117
Hydrazine is extremely toxic. This point, and its high price (about $3/kg), make it difficult to use widely, so it has been used only in special cases.
In a hydrazine–oxygen (or air) fuel cell with alkaline electrolyte, the following reactions take place:
Anode: N2H4ỵ4OH!N2ỵ4eỵ4H2O E0ẳ 1:22 V ð6:5ị Cathode: O2ỵ4eỵ2H2O!4OH E0ẳ0:40 V ð6:6ị Overall: O2ỵN2H4!N2ỵ2H2O E0ẳ1:62 V ð6:7ị
DG0ẳ623:5 kJ=molẳ1:62 eV DH0 ẳ534:2 kJ=molẳ1:39 eV First news about hydrazine–oxygen (or hydrazine–air) fuel cells with alka- line electrolytes appeared in the early 1960s. Gillibrand and Lomax (1962) reported building a rather primitive hydrazine–oxygen fuel cell of the jar type without a diaphragm using four pairs of nickel electrodes (anodes and cathodes). This cell had the defect of admitting direct contact of the oxygen electrode with the hydrazine solution, which made it much less efficient. The electrolyte was 7 M KOH and contained 1.6% hydrazine. The cell produced 50 mA/cm2at a cell voltage of 0.65–0.70 V.
During the decade from 1962 to 1972, a large volume of work on hydrazine–
air fuel cells was done in the Shell research center in the United Kingdom (Andrew et al., 1972). Highly active hydrazine electrodes were developed on supports of expanded nickel mesh receiving additional surface activation. A thin layer of nickel–aluminum alloy was formed on the surface of this screen, then leached with alkali, producing a highly disperse nickel catalyst of the skeleton type. This catalyst was activated additionally by depositing ruthenium on it. Silver catalyst was used on the oxygen (or air) electrode; it was deposited on a support of porous poly(vinyl chloride) (PVC) and contained rhodium. In cells having a thickness of about 4 mm, a flowing alkaline electrolyte containing 1 to 2% hydrazine was used (the alkali concentration was not reported). There was no special diaphragm in the cell. Apparently, the porous PVC support for the oxygen electrode provided sufficient protection against a direct contact between the catalyst of the oxygen electrode and the hydrazine in the electrolyte. It is also possible that the silver catalyst was sufficiently inert toward hydrazine that its activity toward oxygen reduction was unaffected by any hydrazine that was present. A battery assembled from 10 cells of this type working at a temperature of 601C gave a current density of 160 mA/cm2and a voltage per cell of 0.6 V. A battery consisting of 120 cells weighing 57 kg was built which had a useful electrical power output (after subtracting power consumed for its internal needs) of 10 kW.
118 ALKALINE FUEL CELLS
In 1963, work on hydrazine–oxygen fuel cells was also under way at Allis- Chalmers. A 3-kW battery with a voltage of 36 V was developed and intended for golf carts. In the battery, an electrolyte of 5.5 M KOH containing 3%
hydrazine was circulated. As a catalyst for the hydrazine electrode, Jasinski (1963) first suggested nickel boride (Ni2B), which proved to be more active than palladium-activated nickel used in the original experiments. Silver was the catalyst for the oxygen electrode. The working temperature of the battery was 701C. At a current density of 100 mA/cm2, the average voltage of the individual cells was 0.63 V (see Vielstich, 1965, Sec. 4.3.2.2).
Karl Kordesch equipped a motorcycle with a similar battery that he developed at UCC (Figure 6.1) and ran it for 300 miles, giving an electro- chemical periodical the occasion to publish a hilarious German–English poem.*
Many research efforts went into elucidating the behavior of different catalysts in the hydrazine oxidation reaction. Like the oxygen electrode, the open-circuit potential of a hydrazine electrode deviates from the thermody- namic value (1.22 V) and in certain cases comes close to that of the hydrogen electrode (0.828 V). This is due to the fact that thermodynamically hydrazine is unstable and may decompose catalytically to nitrogen and hydrogen:
N2H4!N2ỵ2H2 ð6:8ị
a reaction that will, of course, detract from the faradaic efficiency of hydrazine utilization in a fuel cell. The rate of catalytic decomposition depends on many factors: the nature of the catalyst, solution composition, temperature, and so on. When the metal catalyst for hydrazine oxidation is a good catalyst for cathodic hydrogen evolution as well, hydrazine decomposition is greatly accelerated due to an electrochemical mechanism coupling the two reactions:
anodic hydrazine oxidation by reaction (6.5) and cathodic hydrogen evolution by the reaction
Hỵỵ2e!H2 ð6:9ị This coupling of the two reactions (a kind of local-cell action) is possible because the equilibrium potential of the hydrazine electrode is markedly more negative (by almost 0.4 V) than that of the hydrogen electrode. In an operating hydrazine–air fuel cell, the potential of the hydrazine electrode moves in the positive direction when an anodic current flows, and the rate of cathodic hydrogen evolution [reaction (6.9)] decreases accordingly. In an ideal case, catalytic decomposition should cease completely in an operating fuel cell, so
* Der Kordesch mit sei Motorrad/der racet de highway krumm und grad/Downhill goes it mighty schnell,/aber uphill push like hell.
6.2 ALKALINE HYDRAZINE FUEL CELLS 119
nitrogen and water will then be the only reaction products according to reaction (6.7).
This implies that catalyst research should focus on finding materials and conditions where anodic hydrazine oxidation is accelerated as much as possible, while cathodic hydrogen evolution would be hindered as much as possible.
Tamura and Kahara (1976) studied in detail the composition of the products evolved in hydrazine–air fuel cells working at a temperature of 251C when different catalysts were used for the hydrazine electrode. Platinum, palladium, cobalt, and silver supported by a nickel substrate were studied. The highest faradaic efficiency for hydrazine utilization was found for silver, the lowest for platinum. The efficiency was found to increase with increasing anodic current density (for the reasons stated previously). The efficiency decreased with increasing hydrazine concentration in the solution. Obviously, the nonelectro- chemical mechanism of catalytic hydrazine decomposition then becomes more FIGURE 6.1 Karl Kordesch rides his hydrazine fuel cell motorcycle, 1967. (From Smithonian Institution, neg. EMP059006, from the Science Service Historical Images Collection, courtesy of Union Carbide Corp.)
120 ALKALINE FUEL CELLS
important. At temperatures above 601C, traces of ammonia are detected in the gases evolved. This indicates that a slow catalytic decomposition of hydrazine might occur according to
3N2H4!N2ỵ4NH3 ð6:10ị