Initially, before sealing, the cell was filled with deuterium gas at approximately 1 atm pressure. Calorimetry The electrochemical cell described above was contained within a mass flow calorimeter, Figure 2, the design and operation of which have been described previously. 2,3 Briefly, the calorimeter consisted of an approximately adiabatic enclosure - comprised largely of a silvered, evacuated Dewar which contained the electrochemical cell and through which the calorimetric fluid (water) was pumped. The calorimeter was situated in a constant temperature bath, maintained at 30 ± 0.003°C, which also served as the source of the calorimetric fluid. The mass flow rate of the calorimetric fluid was determined gravimetrically, using an auto-siphon device mounted on an electronic balance, after passing through the calorimeter. internal heater, described above, permitted operation at constant total input power, so as to maintain approximately constant the mean electrochemical cell temperature. The power output from the calorimeter was determined essentially by the mass flow rate, the change in the temperature of the calorimetric fluid on its transit through the calorimeter, and a power loss term, discussed further below. Experimental control and data acquisition were achieved with a Macintosh microcomputer. Data analysis The The difference between the output power and the power input to the calorimeter (both electrochemical and heater) may be referred to as an "excess power", Pxs. For the calorimetric system employed here, this quantity is given by where Cp is the heat capacity of water, dm / St the mass flow rate, Tout the outlet temperature of the calorimetric fluid, Tin the corresponding inlet temperature, Pel the input electrochemical power and Ph the input heater power. The power loss term k' is retained in order to account for the fact that the adiabatic calorimeter boundary is inevitably imperfect, and some conductive heat loss is expected. The methods employed both for the determination of k' at the outset of an experiment, and for the confirmation of its 5 The measurement uncertainty in the excess power, treated as an example of a single-sample measurement, was calculated as described previously3 and is quoted (approximately) at the 95% confidence level ( 2 σ). RESULTS Electrochemical and calorimetric data for the experiment described here during the time period 300 780 h are presented in Figures 3 to 5. Prior to 300 h, either statistically significant quantities of excess power were not produced, or complete calorimetric data were not obtained (due to a bath malfunction). For the calorimeter employed in this experiment, k' was 0.46 ± 0.05 W K-1. Figure 3 shows the variation of input electrochemical and heater powers, and the resulting total input power. Figure 4 describes the measured cell voltage and the electrochemical current during the time period 300 780 h. Note that a cell current of, for example, 5 A is equivalent to a current density of 0.44 A cm-2. The calculated excess power with its associated measurement uncertainty and the average cathode loading are shown in Figure 5. Figures 6 and 7 depict the variation of excess power with electrochemical current and average cathode loading, respectively. DISCUSSION During the time period of interest, excess power up to approximately 1.2 W was produced. Although significant with respect to the measurement uncertainty, the excess power in this particular experiment was relatively small, in particular when compared to the total input power. The excess energy produced during the time period of interest was 1.2 ± 0.3 MJ or approximately 4.3 MJ cm-3 of palladium cathode. this period, the total input electrochemical and heater energies were 36.3 ± 0.07 and 12.6 ± 0.03 MJ, respectively. During In common with previous experiments,2,3 the excess power production observed here appears to conform to a certain phenomenology, discussed below. In addition, |