Calorimetric Measurement of Pd/Ag Alloy Permeated by a Steady Hydrogen Flux
The phenomena known collectively as “cold fusion” include numerous reports of caloric anomalies in metal-hydride systems. As part of an ongoing effort to observe these anomalies, we have recently performed a calorimetric measurement on a membrane of 75Pd/25Ag alloy with a steady flux of hydrogen passing through it. No excess heat was detected in this measurement.
The 75Pd/25Ag membrane is part of a hydrogen purifier manufactured by REB Research. This photo shows the purifier along with its wrap-around heater. The Pd/Ag membrane is inside the cylindrical stainless steel housing. The tube with the pressure gauge brings hydrogen gas to the purifier, the tube with the valve permits the air in the purifier to be flushed out, and the other tube conveys the purified hydrogen out.
For this experiment, we attached two K thermocouple probes to the body of the purifier, one near the top end of the cylinder and the other in the center. They are strapped onto the cylinder with copper wire.
The heater serves to raise the temperature of the Pd/Ag membrane so the diffusion rate of hydrogen through it will be satisfactory. A typical operating temperature is 350-400° C. The physical configuration of the Pd membrane is such that inlet pressures up to 200 psi can be safely tolerated.
After wrapping the heater onto the purifier body, the assembly is wrapped with several layers of wrinkled aluminum foil. The resulting blanket of air pockets serves as high-temperature insulation.
In this photo, the fully wrapped purifier is being inserted into a coil of copper tubing. This coil serves as a heat exchanger for the calorimetric measurement. This measurement is accomplished by flowing water through the heat exchanger at a constant rate while the purifier is operating. Sensitive thermistor probes monitor both the inlet and outlet temperatures of the water. Ideally all the heat released by the purifier is carried away by the flowing water and thus will be reflected in the measured temperature difference.
The photo below shows the encased purifier all connected up to the water system. The temperature probes are located in the Tee fittings that connect to the ends of the copper coil. A metering pump manufactured by FMI, Inc circulates the water at approximately 5 ml/sec.A temperature regulated bath just below the pump serves to keep the inlet temperature constant during operation of the calorimeter.
This photo shows the entire system ready to run. The encased purifier has been wrapped with several inches of cotton insulation that prevent significant heat loss to the surroundings. The hydrogen supply cylinder is visible in the foreground on the left. In the background you can see the inverted graduated cylinder employed to measure the hydrogen gas flow rate through the purifier (the vacuum chamber visible below is not involved in this experiment).
The heater is supplied with AC power from a motor-driven Variac. Delivered power is computed by a TEK 720 digital oscilloscope with one channel monitoring voltage across the heater and the other channel monitoring current via a precision 1 ohm current-viewing resistor. A PC-compatible computer with a Computer Boards DAS-802 data acquisition card monitors the temperature sensors, computes the results, logs data to disk, and displays a live plot of selected parameters during the experiment. The computer communicates with the TEK 720 via RS-232 to learn the input power. The computer uses two digital I/O lines to control the Variac motor to keep the input power constant.
We started each run with no hydrogen pressure on the purifier. We allowed time for the calorimeter system to come to equilibrium and produce a heat output result that agreed satisfactorily with the measured electrical input power. Then, without changing anything else, we applied a selected pressure of hydrogen to the purifier. While the pressure was on we periodically measured the flow rate with a stopwatch and an inverted graduated cylinder (initially completely filled with water). We studied the heat output signal carefully before and after the hydrogen flow looking for any sign of change. After a significant period of constant hydrogen flow, we removed the hydrogen pressure so the baseline heat output could be reconfirmed.
The plot above shows the results of a typical run. The input power, Pin, is plotted in purple on a scale from 76 to 86 watts (1 watt/division). The measured heat output power, Pout, is plotted on the same scale in blue. Note that Pin stays fairly constant throughout the run at the setpoint of 80 watts. Pout varies somewhat but typically runs 78-79 watts. The duration of this run was nearly 6 hours. The horizontal scale is 1 hour/division.
The purifier case temperatures, Tcaseupper and Tcaselower, are displayed in red and black on a scale from 280° C to 320° C (10° C/division). Room temperature, Troom, is displayed in gray on a scale from 15° C to 25° C (1° C/division). Finally, the inlet water temperature, Tin, is displayed in cyan on a scale from 25° C to 35° C (1° C/division).
Periodically throughout the run, the TEK 720 scope was recalibrated to compensate for accuracy drift. These events are marked by sharp excursions in the Tin trace which correspond precisely in time to larger excursions in the Pout trace. These excursions should therefore essentially be ignored.
At the time marked by the first arrow, a hydrogen pressure of 150 psig was applied to the purifier. This resulted in a hydrogen flow of about 7 cm3/sec (measured at room temperature and pressure). This pressure (and flow) was maintained until the time marked by the second arrow. Then the hydrogen cylinder valve was closed and the pressure was allowed to decline naturally as the remaining hydrogen in the lines passed through the purifier. 30 minutes after closing the valve, the hydrogen pressure was essentially gone.
During the period in which hydrogen gas was flowing, we made an effort to monitor the temperature rise of the hydrogen gas during its passage through the calorimetric enclosure. We applied K thermocouples to the exterior surfaces of the inlet and outlet gas tubes approximately where these tubes crossed the boundaries of the calorimetric enclosure. A number of factors make it difficult to determine precisely where these boundaries are located so this data is useful only for rough approximations. We observed a 20° C delta-T with these sensors.
We performed a total of seven runs in this study. We varied the purifier case temperature from 210° to 400° and the hydrogen gas pressure from 20 psig to 150 psig. The data presented above is representative of all of these runs.
It is evident from the Pout trace that no large thermal effects were produced by the hydrogen gas. When the gas flow was started there is a small dip in Pout but it apparently recovers in about 1/2 hour. The overall appearance of the Pout trace also indicates a gradual positive drift in the system.
With a little imagination one can estimate that, during the period of hydrogen flow, the Pout trace was depressed perhaps a few tenths of a watt. If we take the observed hydrogen flow rate of 7 cm3/sec and the crudely measured gas delta-T of 20° C we can compute that the flowing hydrogen should have cooled the purifier by about 0.2 watts. This value is not inconsistent with the appearance of the Pout trace.
We can conservatively estimate our detection limit for anomalous heat production in this experiment at 1 watt. The measured hydrogen flow rate corresponds to about 3.6 *1020 hydrogen atoms/sec. Thus our 1 watt heat detection limit can also be expressed as a 0.017 eV/atom detection limit. That is, if each H atom passing through the Pd/Ag alloy somehow generated 0.017 eV of anomalous energy, the result would have been a 1 watt increase in the Pout trace which would have been easily recognizable.
Thus we can conclude that, if there is any anomalous heat generation due to the passage of H atoms through a Pd/Ag alloy under these conditions, it is substantially less that 0.017 eV/atom of hydrogen.
The experiment should be repeated with deuterium instead of hydrogen (protium). It also seems prudent to try high pressures combined with lower temperatures in order to make the Pd/Ag alloy load significantly with hydrogen.