240) Helium is produced in spent CF cathodes

Ludwik Kowalski (15/7/05)
Department of Mathematical Sciences
Montclair State University, Upper Montclair, NJ, 07043

Searching for something else over the Internet I discovered an interesting paper published in 2002 by Robert Bass. That paper, a file in the pdf format, is automatically downloaded, as a pdf file, when one clicks on the first link that appears after the following URL is specified in the browser.

http://www.google.com/search?hl=en&q=donquixote%40radix.net&btnG=Google+Search

The title of the paper is “Five frozen needles CF protocol;” it appeared in Journal of New Energy, 2002. 6(2): p. 30. The protocol is designed for a layman who, presented with data, would be convinced, with 95% certainty, than helium is produced in a cathode during the electrolysis (in a Fleischmann type cell). In the last paragraph Bass wrote: “I showed my one-page proof to an internationally reputable expert in Statistics and he agreed that it was correct and said that if the experiment turns out as I predict then he will give me a Certificate that the Confidence Level is that asserted and be willing to testify in Court as an Expert Witness that according to accepted principles of Statistics the preceding protocol is definitive as stated.”

The emphasis in the paper is on statistics. But experimental details for the proposed experiments are worth paying attention to. The difficulties, as we know, are often in details. I see no evidence that the experiment has been actually performed by the author. Here is what would be required:

1) Having a strip-shaped cathode of cold-fusion-suitable Palladium. As emphasized by Bass, recognizing suitable cathode material is not a trivial task.

2) Loading the strip with deuterium up to a level of about one atom of ²H per one atom of Pd. This task is also not trivial. Quoting E. Storms Bass lists four methods of measuring ²He/Pd ratios.

3) Keeping the foil at the liquid air temperature till five main experiments can be performed (to minimize losses of ²H due to diffusion).

4) The foil should be cut into five narrow strips, called needles. These needles should then be used in five nearly identical experiments (k=0, 1, 2, 3 and 4). The only difference between the experiments is duration of electrolysis, t_k=k*T. If T=2 hrs, for example, then the durations of consecutive experiments, would be 0, 2, 4, 6 and 8 hours. The difference of potential between the electrodes should be 17.7 volts (I do not know why) and the current should be kept constant. The k=0 experiment would be a control.

5) All five cathodes should be sent to a laboratory specializing in measurements of the amounts of helium, A_k. Using the results -- A₀, A₁, A₂, A₃, and A₄ -- one would calculate statistical parameters defined by Bass. Unfortunately, I am lost in the conclusive part of the paper. And I am puzzled by the fact that random errors, associated with individual A_k, are not being used in the analysis. Plotting A_k versus t_k, and showing the bars of A_k errors, is likely to be more convincing. I saw that kind of a plot in several cold fusion reports; the most recent was a paper describing a real experiment performed by Russ George.

The issue that Russ had to address was presence of helium in air. The challenge was to show that helium from air, or from something else, was not confused with helium from nuclear reactions in the cathode. Those who remember the 1989 DOE report know that absence of nuclear ashes (such as He), in the amounts commensurate with the amount of excess heat, was one of the decisive arguments against the claim of Pons and Fleischmann. The topic was discussed by pannelists of the second DOE report, several months ago. This time some of them were convinced that generation of He in cold fusion experiments is real.

Appended on 7/18/05:
The protocol described by Bass should not be viewed as applicable only to generation of helium. But the title of this unitis is “Production of helium.” It is therefore appropriate to quote what Michael McKubbre wrote on that subject meny years ago. His paper on the subject is in the compilation of articles publishe in Infinite Energy. That compilation, a file named samplearticle.pdf, can be downloaded from <www.infinite-energy.com>. The method used by McKubre was not electrolytic; generation of ⁴He was reported to occur in foils that were kept in a vessel filled with duterium (at high temperature and pressure). Here is the quotation:

“. . . we submit a sample of gas from each of the cells for analysis to the mass spectrometer, a high-resolving, low-mass mass spectrometer. We’re capable of separating the two masses of species, deuterium ²D and ⁴He. . . . On the monitor you see displayed, in fact, the mass spectrum from one of these samples. This is a relatively high level of helium-4. The peak here is the helium-4 peak, the deuterium peak would normally appear here; it’s completely absent. This particular example shows 10.5 ppm helium. We compare the samples each day that we perform the analysis, we compare the samples of gas from the various active cells and blanks with a sample of room air, which we have measured many, many times and know to be 5.22 ppm. And we have some standards, which we typically use—that is, gas samples of helium in deuterium and argon which we submit to the mass spectrometer for the purpose of calibration.

The mass spectrometer simply sweeps a mass from low mass to high mass, in this case from 3.96 mass units to 4.06 mass units, which encompasses the range in which helium is to be found. In fact, this peak is helium, and deuterium D2 is to be found which will be found somewhere in this region. We use a liquid nitrogen cooled carbon trap in order to remove D2 so that we’re able to see quite low levels of helium. We’re accurate to probably 0.1 ppm helium and we can clearly resolve the presence of ²D and ⁴He. This spectrum is, in fact, the sum of a number of spectra that the mass spectrometer simply sweeps for the period of time that we pre-program, and this is the cumulative signal representing the integral of all helium which was present in the sample when we submitted it for analysis. To acquire this spectrum takes us about five minutes.

It’s clearly not possible to produce helium from a chemical process. If we observe helium in our experiments it’s either because it leaked in from the atmosphere—we can rule that out by the blanks that we do and the fact that the helium signal that we have seen is larger than the helium in the ambient. It’s possible that the helium pre-existed in the sample and was simply released to the gas phase with long term exposure. We can rule that out largely because we’ve analyzed the catalyst that we’re using and found that it contains no measurable levels of helium. The only possibility that remains, and remains to be checked, is that the helium is produced by a nuclear process. If the helium is produced by a nuclear process, then necessarily there will be an associated release of heat. . . .”

Recognizing the major problem in the entire cold fusion field McKubre wrote: “One of the difficulties in the cold fusion field is the apparent apparent lack of replicability of experiments: many people performing the same experiment get apparently different results; different experiments performed in the same laboratory give apparently different results. So it’s obvious that if you do the same thing you must always get the same result. What this is telling us is that there are some important parameters of our experiments that are not under our control. Some of them I know and understand, and still [we] can’t control some of these parameters we don't know about yet. We just don’t know what the process is that we are studying, so we don’t know what parameters we need to control in order to yield a consistent result. An experiment which always gives the same result --can be performed in several different laboratories to yield the same result--would be very valuable to us, in part in helping to convince the remaining skeptical scientists in the world that there is a phenomenon to observe. But, in fact, in order to use the scientific method to observe scientific results, we have to be able to reproduce the results of our own experiments so that we can see what the effects of small changes are on these experiments.

One experiment which has been reported to produce consistent and reproducible results is that of Professors Arata and Zhang, both of them are very, very experienced and very well recognized scientists in Japan. They performed a very careful experiment, reproduced it apparently a number of times in their own laboratory—producing both anomalous excess heat in fairly significant levels and helium-4 and, perhaps more interestingly, helium-3. The helium-3 to helium-4 ratio that they observed in their experiments is different from that in the air that we're breathing. Sufficiently different to indicate that there is clearly an anomalous nuclear reaction occurring. The difficulty with Arata and Zhang’s experiment is that it’s only been performed by them and only in their laboratory. What we’re attempting to do here is to produce their same results with their apparatus and with their help. This is a collaborative effort between Arata and Zhang and the SRI group, to produce in our laboratory the same results as they have obtained repeatedly over the years, which would indicate that we have some degree of mastery over the experiment.

The experiment that we have running here, in fact, is relatively young; it hasn’t been operating for very long. One of the difficulties with Arata’s experiment is that it requires many, many months to produce a result, and quite literally we’re not very experienced with Arata’s methods, so we’ve had some difficulty getting his experiment set up and operational. Certainly, it’s caused me to have an increased level of respect for Arata and Zhang’s technical competence. They are very, very good scientists. Within a month or two, we hope to have reproduced their experiment faithfully and reproduced their result.”

It should be clear that factors mentioned by McKubre are not at all addressed in the statistical protocol described by Bass. His protocol was not designed to deal with systematic errors; it was designed for experiments in which all uncertainties are due random fluctuations about mean values.