In the Fall of 2006, Drs. John Dash and Wu-Shou Zhang came to our lab to assist in a replication effort of their CMNS experiment. During the previous year, they published a paper entitled Seebeck Envelope Calorimetry With A Pd/D2O+H2SO4 Electrolytic Cell in which they report excess power of nearly 10% in several experiments (see Table 1). This performance was reported to be reproducible as well.
Dash and Zhang brought their cell (the ‘DZ cell’) to EarthTech to be tested in our high-accuracy calorimeter (MOAC). We tested the DZ cell in the same configuration they use in their laboratory as well as in various modified configurations. We did observe some apparent excess heat in our testing but not as large as previously reported by Dash and Zhang. However, after much investigation, the evidence points only to mundane causes for the excess heat signals observed in our lab.
The DZ cell is contained in a relatively tall glass vessel. The electrodes are located near the bottom immersed in approximately 100 mL of electrolyte. The upper half of the vessel contains approximately 100 gm of Pt-coated alumina recombiner pellets. The cell lid is not gas tight. The electrolyte is composed of 100 gm D2O and 19.12 gm H2SO4 – a volumetric ratio of 8.5 to 1. The electrodes are thin sheets of similar surface area placed about 3 cm apart in the cell. The palladium cathode was slightly curved with the concave surface towards the platinum anode. Dash and Zhang assert that this is essential for excess heat and that the curvature is in fact created during electrolysis.
According to Dash and Zhang, several conditions are conducive to the production of excess heat. The run should be started with the cell at room temperature. Full current (approximately 4 A) should be applied at the beginning of the run and maintained throughout. Electrolyte temperature should closely approach the boiling point during the run. Zhang further recommended that the cell should be enclosed in a close-fitting cardboard box.
Additionally, Zhang claimed it to be advantageous to reverse the polarity of the electrodes for a period of time, usually an hour. When the normal polarity is restored, some cells will perform better.
During their visit to our lab, all electrolyte mixes were prepared by Zhang and the cells always contained electrodes provided by Dash and Zhang.
I – EarthTech Cell and Unsealed DZ Cell
After reviewing the details of the DZ cell, we expressed concern over the fact that the cardboard box would barely fit inside MOAC’s calorimetry chamber. Primarily we were worried about reduced air circulation around the box because air is the primary heat exchange medium inside MOAC’s chamber. After some discussion, Dash and Zhang encouraged us to conduct the first run (DZ1) using only their electrodes and electrolyte in our standard cell design: a 200 mL beaker with an o-ring sealed Teflon male gland top with 7 o-ring sealed feed-through holes. A machined groove at the bottom of the gland holds a single row of platinum coated alumina recombiner pellets (about 1.3 g of pellets). In order to achieve the desired electrolyte temperature, the cell was insulated. This arrangement produced an electrolyte temperature of approximately 85°C with 10 W of input power. The insulation covered the entirety of the electrolyte leaving the portion of the beaker containing the recombiner pellets exposed.
Using our standard methods, the cell was run at a constant power of 13 W instead of constant current. Figure 1 shows the power balance for the run. The blue trace is the electrical input power and the red trace is the heat output power. No excess heat was observed. The polarity of the electrodes was reversed for an hour during the run and the disturbance from this can clearly be seen in the input power on the graph at about 1630 and 1730 on 08/29. At about 0300 on 08/30, the recombiners became insufficiently active and were unable to react all of the D2 and O2 gas being produced. MOAC is designed to vent gases from the cell when the pressure exceeds a set point, typically 800 torr. Since “fuel gas” was being vented from the cell, the measured heat power output decreased. The power, and therefore the current, was reduced in an attempt to reduce the gas production rate to a level that the ailing recombiners could handle. After several unsuccessful attempts, the run was abandoned.
Figure 1: Power Signal from DZ1
Because of the failure to observe excess heat in this run, we realized that there could be something important about the DZ setup that was missing from our first attempt. We therefore decided that the next run (DZ2) should test the DZ cell with little or no deviation from the procedures and practices recommended by Dash and Zhang, including use of the cardboard box. At this time MOAC did not have constant current capability so we manually adjusted the input power at the beginning to maintain an approximately constant current.
As shown in Figure 2 below, DZ2 showed what appeared to be a small excess heat signal. However, the output power eventually dropped below the input power and remained there until the end of the run.
Figure 2: Power Signal from DZ2
The negative power balance at the end of the run suggested either that the unsealed cell was losing significant amounts of unreacted D2 and O2 gas or that MOAC was not responding accurately to the unusually sized and shaped heat source in the calorimetry chamber. However, the apparent excess heat at the beginning of the run did not have a ready explanation.
This behavior convinced us to perform a calibration run with the DZ cell and cardboard box. The heavy water electrolyte was replaced with light water electrolyte and two platinum electrodes were used. Unfortunately the light water electrolyte boiled much more readily than the heavy water and the recombiners were unable to function when the head space contained such large amounts of water vapor. After many failed attempts to obtain a proper calibration, open cell testing was abandoned and we returned to a modified version of the ET standard cell.
II – Modified EarthTech Cell
Our first test was unsatisfactory because the recombiners failed and allowed gas to escape. Our second test was unsatisfactory because the cell was open and probably allowed gas to escape. Dash and Zhang suggested that we modify our standard cell to more closely mimic theirs. We did so by adding a second row of recombiner pellets, more than doubling their number (the standard ET cell holds just over 1 gm of pellets while the DZ cell contains 100 gm of them). The modified cell top now contained 3 gm of recombiner pellets.
Considerable effort and time was needed to configure the ET cell to operate within the conditions set forth. Eventually, we succeeded in constructing a cell that operated well enough to complete a proper calibration in MOAC. We performed four tests (under the common run name DZ5) on this cell in order to explore the techniques typically employed by Dash and Zhang. None of them showed excess heat.