(Paper submitted to Rev. Modern Physics and rejected because it discussed cold fusion (See letter at end). Eventually was published in Infinite Energy ,4, #21 (1998) 16.)

Cold Fusion Revisited

Edmund Storms,
2140 Paseo Ponderosa, Santa Fe,
NM 87501-6319


A collection of recent studies of what is incorrectly called "Cold Fusion" shows the present status of the phenomenon and provides some explanations for the claims. An understanding of many claims continues to improve while new discoveries still challenge the most creative theories. A case is made to support a new field of study involving chemically assisted nuclear reactions.


Nucleons and electrons are separated by vast differences in energy and in the time needed for its release. As a result, chemistry and nuclear physics occupy two different worlds. Observations of Profs. Stanley Pons, Martin Fleischmann, and Mr. M. Hawkins(University of Utah)[1] and independently by Prof. Steven Jones et al.[2](BYU), suggest that a bridge exists between these two worlds. Thus was born what is conventionally called "Cold Fusion", an unfortunate name. Because no accepted theory properly supports the claims and because much experimental data were poorly presented and hard to duplicate, most scientists continue to reject the notion. Adding to the difficulty is the conflict between the picture of nature these claims would imply and the dearly held conventional understanding.

In spite of good reasons for skepticism, a few scientists succeeded in duplicating the observations. Many of these studies are on going at the present time, especially outside of the U.S. This world-wide effort is underway in at least eight countries involving hundreds of scientists and hundreds of millions of dollars per year. Very persuasive results supporting the initial claims, as well as many new ones, have created a need for a fresh look.

Because the phenomenon combines solid-state chemistry and nuclear physics, the backgrounds and expectations of potential readers are vastly different. Consequently, a level of detail required to be persuasive to everyone is not possible in a brief review. So, only a sampling of experimental observations and theoretical models are presented to give a general background. I hope the reader will take the trouble to read other sources of information before reaching a conclusion. Unfortunately, much of the supporting evidence is still published only in conference proceedings because many scientific journals have been reluctant to publish papers about the subject. Consequently, peer reviewed information, the core of conventional scientific evaluation, is largely missing. The reader will, to some extent, have to evaluate some of the claims using their own good judgment. Six books describe the history of the field, some with objectivity[3-6] and some containing serious distortions[7,8]. Several general, peer reviewed scientific reviews are available[9-12]. These sources can be consulted to understand the background, both pro and con. A complete bibliography can be obtained from Dieter Britz [13] or Hal Fox [14]. Over 2000 publications now address issues related to the field.

The field has expanded beyond the original claims and can now be called "Chemically Assisted Nuclear Reactions" (CANR) instead of "Cold Fusion". Production of anomalous energy far in excess of any conventional chemical source, production of many elements not present in the original environment, and production of radiation that can only result from nuclear reactions are all characteristic of the field. Each is found to occur in a variety of materials using numerous techniques when deuterium as well as normal hydrogen are present. The nuclear products include helium, tritium, and various isotopes resulting from several types of transmutation. Radiation in the form of neutrons, gamma rays, X-rays and charged particles have been detected, all at very low levels compared to the rate at which energy is produced. A change in radioactive decay rate of tritium, produced by the physical environment, also has been proposed.[15] The extraordinary nature of this growing list adds to the challenge of presenting a review that encompasses the entire field without sounding so incredible to cause alienation of the reader.

To understand CANR, we need to address the issue of why the claims have been so hard to believe. Disbelief is based on the experimental results being inadequate and in conflict with conventional understanding of nuclear interaction. In addition, many failures to produce excess energy, especially those at several high-profile institutions[16,17,18], gave the impression that the phenomenon could not be reproduced at will.[19] At one time this was true. As the variables have been better understood, success in producing energy as well as nuclear products has improved. Certain methods and results are now highly reproducible in the hands of skilled experimenters. Many negative results are known to be caused by failure to understand and control important variables. For example, the effect will not occur in an electrolytic cell using a palladium cathode when too much light water (H2O) is present in the heavy-water (D2O). Many early studies used open cells which readily absorbed H2O from the atmosphere. Added to these experimental problems, is the highly variable nature of palladium such that only a small number of samples will have the necessary characteristics. Of course, random error can still play a role, but a growing understanding indicates error is not the main cause of positive results. Several of the better done studies will be described in detail.

Scientists quite rightly have asked to be shown nuclear products along with energy production before a nuclear source is accepted. This request has now been honored with the detection of helium (4He) by several laboratories, as described in a later section. In addition, a growing body of data now support production of several nonradioactive nuclear products not associated with a fusion reaction, so-called transmutation products. Detection of such products is claimed in several vastly different chemical environments including, surprisingly, living systems. Of course, some of these claims have simple explanations, but not all.

When conventional methods are used to produce fusion at high temperatures in a gas plasma, in the Tokamak for example, neutrons and tritium are produced in equal amounts and at high levels. Such expectations fueled early efforts to detect these products during CANR. Lack of success added to widespread skepticism but also demonstrated that whatever occurs is not conventional fusion. CANR apparently does not generate these products in the expected ratio or quantities. Indeed, anomalous heat is seldom associated with tritium production and very few neutrons are detected. These contrasting behaviors are summarized in Table 1.

Tritium is one isotope whose presence is difficult to mistake. If its formation by CANR can be demonstrated, a powerful support for the general idea would exist. For this approach to be successful, the tritium must be shown not to result from another process. The detected tritium could have entered the cell from the environment or be a contaminate in palladium, two of the more common reasons given for rejecting the nuclear explanation. However, tritium has been produced in such large quantities and in such isolated environments to make such prosaic explanations very difficult to support. Except in certain areas of government laboratories, tritium is very rare indeed. Numerous studies have shown that commercial palladium does not contain significant tritium.[20] Therefore, it is no longer appropriate to dismiss tritium claims as being caused by contaminated palladium. These arguments will be developed in more detail in a later section (Section 2.2.1).




Fraction of Product with Resulting Energy shown for Hot and Cold Fusion


t(1.0 MeV + p(3.05 MeV)

3He(0.83 MeV) + n(2.45 MeV)

4He + g(24 MeV)

d+t = 4He + n(14 MeV)


Although neutron emission is well below the expected rate, it is not completely absent. Many studies have detected small emissions with energies corresponding to the conventional fusion reaction as well as neutrons produced by other sources. Because the emission rates are close to background and are erratic, much doubt still remains as to their source.

Helium-4 production in a hot plasma by fusion is rare and always accompanied by 24 MeV gamma emission. Absence of this radiation is a major reason why many people reject this claim when applied to CANR. They fail to consider at least five other reactions which result in helium without accompanying gamma radiation, as listed in Table 2. Each of these reactions has a theoretical basis and, in a few cases, some experimental support. Therefore, helium cannot be rejected as a nuclear product just because gamma radiation is absent.

Well established theory predicts that the coulomb barrier can not be overcome using the small energies available within a chemical structure. Clearly, if fusion or any other nuclear reaction is to occur at all, a mechanism must exist that does not require high energy, that does not produce the usual reaction paths, and that is able to absorb the resulting nuclear energy without significant radiation. This is a lot to ask of any process. To make matters worse, energy production from light hydrogen and the newly discovered transmutation reactions introduce a whole new set of problems that tax any explanation. We can either accept the conventional viewpoint and throw out all experimental observations involving CANR, or we can search for new theories that consider the novel characteristics of the chemical environment while keeping most claims. In other words, the results obtained at high energy are correct for the environment in which they were obtained, and the CANR results are also correct for their environment. Most of the new explanations take this approach. No lack of possibilities exist. A few will be discussed later in the paper. The challenge is to demonstrate which one, if any, is correct. Although it is satisfying to believe one explanation fits all, the increasing variety of nuclear reactions and chemical environments make this approach increasingly difficult to justify.



Reactions Producing Helium with No Gamma Emission

1. Multiple fusion to produce helium

d + d + d = 4He + d

d + d + d + d = 2 4He

2. Reaction with boron impurity to produce helium

d + 10B= 4He + 8Be

8Be = 2 4He

3. Reaction with lithium impurity to produce helium

d + 6Li = 8Be = 2 4He

4. Fusion with a metal leading to alpha decay

d + nM = a + (n-2)M

5. Collapse of deuterium atom to produce beta decay and helium

d + e- = 2n

2n + d = 4H = 4He + e-



The CANR phenomena has expanded from simple electrolytic loading of palladium and titanium with deuterium to include the methods listed in Table 3. Each method has produced unusual results when both light (H) and heavy (D) hydrogen were used. Unfortunately, only a few of the methods have been replicated by independent investigators. Most of methods will be discussed starting with electrolytic production of heat. Recent work not covered by previous reviews will be emphasized. Radiation and nuclear products will be discussed in separate sections.



Methods Used to Produce the CANR Effect


Gas Reaction

Electric Discharge

Phonon conduction through semiconductors

Cavitation involving bubble formation

Mechanical changes

Sudden decomposition of hydride

Biological systems


II.1. Electrolysis

Electrolysis is the most widely used method to produce the claimed CANR effects. When a current is passed through a water-based electrolyte, hydrogen (or deuterium) is formed at the cathode (negative electrode) and oxygen is formed at the anode (positive electrode). The metal used as the cathode reacts with the hydrogen to form a hydride. Usually palladium is used as the cathode with D2O, and various forms of nickel are used with H2O. Hydrogen with palladium and deuterium with nickel have not been found to produce unusual results. Other metals have given some claimed success but without the necessary replication. For reasons that are not yet fully understood, certain batches of metal are much more likely to show the CANR effects than are others. Because these additional variables exist, negative results can not be used to off-set positive results such as would be appropriate if the positive results resulted only from random error, as some skeptics have suggested.

II. 1.1 Heavy-water

In the case of palladium electrolyzed in D2O, the hydride formed is b-PdD1-x, a face-centered-cubic structure containing vacancies (x) in the deuterium sublattice. Because the effective deuterium activity and concentration are both very high at the cathode surface, normally unstable hydrides can also form in a region which is microns deep. In addition, the surface region is rich in deposited impurities which influence the deuterium activity and phase stability of various compounds. Consequently, the active chemical environment is highly variable, accounting partly for the difficulty in reproducing the claims. For these reasons, an understanding can not be based only on the chemical environment known to exist in b-PdD, as has been done extensively in the past.

Over 50 studies reporting repeated examples of excess energy production have been done, most of which have been published at least in conference proceedings. Studies published before March 1996 are reviewed by Storms [11]. To understand and eventually accept the claims for heat production, it is necessary to understand the various possible errors in calorimetry and how they have been eliminated. For example, open cells were extensively used during the early studies in which an uncertain recombination of the D2 and O2 gases could take place, as pointed out by Jones et al.[22]. This process would cause a variable amount of energy to be captured by the calorimeter, resulting in what might be interpreted as excess energy. Most studies now use closed cells or measure the amount of recombination. Another early issue was the effect of temperature gradients within a nonstirred calorimeter.[23] To answer this criticism, Pons and Fleischmann[24], Klein et al.[25] , Takahashi et al.[26] and Guruswamy et al.[27], among others, demonstrated that mixing produced by electrolytic generated gas bubbles is sufficient to eliminate serious gradients in their cells. Not given much attention is the effect of the stagnate water layer at the cell wall on the thermal conductivity of the wall. The magnitude of this layer is very sensitive to the flow patterns within the cell. Consequently, the accuracy of simple isoperibolic calorimetry can be compromised by changes in stirring rate and by changes in the amount of bubbles generated. In addition, calibrations using internal heaters, where bubbles are absent, can not be applied to the electrolytic conditions where bubbles are present. It is, therefore, difficult to evaluate claims based on calorimeters stirred only by bubbles or when the stability of the mechanical stirring rate is uncertain. In contrast, some workers use calorimeter designs that are completely immune to the effect of temperature gradients such as flow calorimetry, double-wall isoperibolic calorimetry [28,29], and devices based on the Seebeck effect [30,31].

In their original studies, Profs. Pons and Fleischmann used a calorimeter requiring a rather complex mathematical analysis, thereby opening the door to easy rejection. Several independent analyses were made to evaluate their claims, the most detailed being a paper by Wilson et al.[32] This and other criticisms are answered by Pons and Fleischmann in several subsequent publications [24] and by several other workers[33,34,35]. In addition, an independent study by Lonchampt [36], using an identical calorimeter, duplicated the Pons-Fleischmann claims for anomalous energy. Although the heat measurements by Pons and Fleischmann can be analyzed in different ways to give slightly different results, their conclusion for excess energy production has been supported.

One extensive study done at SRI (Stanford Research International) stands out for its completeness and its use of state-of-the-art equipment and methods.[37] EPRI (Electric Power Research Institute) invested over $6M to give us this information, and the New Energy Development Organization (NEDO) (Japan) has continued to support the work for the last 3 years after EPRI withdrew from the field. Figure 1 shows the cross-section of one of several flow-type calorimeters used for most of their studies. The cell is closed and contains a recombiner that returns to the calorimeter all energy invested in decomposing the water. Over 98% of the energy applied to the cell is recovered by the flowing water. Composition of the palladium cathode is determined from its change in resistance. An internal heater keeps the internal temperature constant when different electrolytic currents are applied and provides a means to check instrument stability. The device is used as an absolute calorimeter based on the flow rate and temperature change of the cooling fluid. This design has shown very stable behavior and has a sensitivity to energy change of ±10 mW although an uncertainty of ±50 mW is used when data are reported. Typical data are shown in Fig. 2. This sample produced no excess heat until the current was increased at 400 h, thereby causing an increase in composition. Figure 3 shows how excess power relates to the average composition. Other workers have reported seeing this behavior as well. Applied current also has an effect as can be seen in Fig. 4. Studies done at other laboratories, including the work of Pons and Fleischmann, are consistent with there being a critical current below which no excess heat is detected. This behavior occurs even when a variety of sample forms and calorimeter types are used. Rejection of these studies must explain why the relationship between bulk composition and applied current is so consistent. Conclusions obtained from the work at SRI are summarized in Table 4. These conclusions are completely consistent with all published studies showing excess energy and each has been confirmed by several investigators. Unfortunately, all attempts to initiate the CANR effect using palladium suffer from a large fraction of failures. Adding to the confusion are the many negative studies which failed to consider even the basic requirements for success.



Conclusions From Studies Done at SRI

1. The D/Pd ratio must exceed a critical value.

2. Current must be maintained for a critical time.

3. The current density must be above a critical value.

4. Inert palladium can sometimes be activated by adding certain impurities to the electrolyte.

5. The effect occurs in only a small fraction of samples but more often in certain batches than in others.


II. 1.2 Light-water

Surprisingly, deuterium and hydrogen both produce excess energy and nuclear products. However, as expected, the nuclear products frequently are different.

Energy production using light water was first demonstrated by Mills et al.[38] Their reasons for making such an attempt were not based on an expected nuclear reaction, as described later. Energy production has been duplicated by other studies[11], but with evidence for a transmutation reaction being the source. Recently, Clean Energy Technology, Inc. (CETI) [40], using nickel coated plastic beads electrolyzed in a flowing electrolyte containing Li2SO4 in H2O, have demonstrated production of significant energy. However, most efforts to duplicate this work have been unsuccessful for various reasons. Apparently, the nature of the metal coating on the bead is very important and difficult to duplicate even when the patent is consulted.

Light hydrogen has also been reacted as a gas with nickel to produce heat[41] and nuclear products and it has been used as an energetic ion to bombard nickel as described below. Preliminary claims for production of 3He[42] suggest that d-p fusion may occur when deuterium and hydrogen are both present.

II. 2.1 Nuclear Products

II. 2.2.1 Tritium

The first nuclear product supporting the claims of Pons and Fleischmann was the claimed detection of tritium[43] in an electrolytic cell. This isotope is seldom produced when excess energy is detected and seems to be associated with the presence of certain impurities in the cell, copper being the more likely one, and the presence of dendrites on the cathode surface. When detected in anomalous amounts, the tritium exists in the electrolyte rather than as DT gas.[44] In contrast, tritium previously dissolved in palladium as contamination is always released during electrolysis as DT gas. This contrasting behavior demonstrates that observed tritium is not caused by contamination and that it was not formed within the bulk material. Only tritium formed on the surface can dissolve in the electrolyte. Unfortunately, tritium production has proven to be more difficult to duplicate than excess energy.

Two of approximately a dozen studies stand out in demonstrating the presence of anomalous tritium when heavy-water is used with palladium. Will et al.[45], at the now disbanded National Cold Fusion Institute, used cells completely isolated from the environment which contained either D2SO4 or H2SO4 electrolytes and a recombiner catalyst. Tritium atoms in the range of 7x1010 to 2.1x1011 were found in four heavy-water cells while fewer than 4x109 atoms (the detection limit) were present in light-water cells using palladium from the same batch. Other batches were below the detection limit in both cell types. The T/D ratio in the palladium was found to be significantly higher than in the electrolyte, indicating the tritium originated in the palladium. Careful analysis of many virgin palladium samples showed no indication of tritium contamination.[20] They conclude that the probability of the effect being caused by contamination was 1 in 2380.

Workers at Texas A & M[46] produced tritium in one cell using a LiOD electrolyte which was in series with a similar, nonactive cell. The production rate was sensitive to the applied potential (current) as shown in Fig. 5. Production could be stopped by changes in current or by agitating the cell. The reaction could then be restarted. This behavior is totally inconsistent with tritium originating either as contamination or from the external environment. A total of 1015 tritium atoms were produced during the study compared to fewer than 5.1x109 atoms detected in unused palladium, the background of the detection method.

Claytor et al.(Los Alamos National Laboratory (LANL)[47] continue to explore the production of tritium. A pulsed-gas discharge is used between a small palladium wire and an inert electrode in low pressure D2 gas. The voltage is too low to initiate a normal fusion reaction known to occur at high voltages. All wires are previously outgassed to remove any tritium contamination. No tritium has been found in any commercial palladium even though scores of samples from many sources have been examined. This experience is consistent with several other detailed studies as noted previously. Tritium pickup from the atmosphere, a very unlikely source even at LANL, is prevented by using a vacuum-tight, stainless steel apparatus. Deuterium gas is pumped through a loop containing the plasma cell, in which the discharge occurs, and through a Femtotech Tritium Gauge which provides a continuous measurement of tritium. Provisions are included to remove some gas, convert it to water, and measure its tritium content using standard scintillation techniques. Consequently, two independent methods are used to insure the claimed product is indeed tritium. Figure 6 compares the behavior of several samples exposed to D2 as well as the behavior when H2 or platinum are used in place of the active materials. Active samples begin producing tritium immediately after discharge is started. Changes in current or other operating conditions cause changes in the observed tritium production rate. Similar conditions produce no effect when D2 is replaced by H2 or when Pd is replaced by Pt, thus providing a null check. Like all experiences in the field, the ability to produce the effect is very sensitive to the nature of the palladium sample. For this reason, the results have been inconsistent. Nevertheless, certain batches of palladium or its alloys have a sufficiently high success rate to allow many variables to be explored. Of the many examples of tritium production, this work stands out as being the most difficult to refute. Although this work has not been published in a reviewed journal, it has been extensively reviewed at LANL and is available on the Internet.[47]

II. 2.2.2 Helium

Helium was recognized early in the field's history as being a likely nuclear product, but its difficult detection had to await the careful work of Miles et al.[48] Confirming studies using the electrolytic technique have been done at the University of Texas[31], University of Roma La Sapienza[49] and at INFN in Italy[50], and in Japan[51] using gas loading. In each case, care was taken to eliminate helium that might enter the cell from surrounding air. However, not all attempts to detect helium have been successful partly because excess energy is difficult to produce and partly because measuring small amounts of helium in deuterium is difficult. When helium is reported, the levels are lower than expected based on a fusion reaction being the sole source of measured energy. Unfortunately, the reported values are not the sum total of all helium, i.e. that contained both in the gas and dissolved in the palladium metal. Only one of the two sources is typically reported. Obviously, all helium needs to be considered when the results are compared to energy production. Alpha emission has also been detected[52] and recent claims for 3He have been published[42]. The latter claim is being actively investigated by other workers. No gamma radiation of any significance has been detected when 4He is produced. Again, this only means that the helium producing reaction is not caused by "hot" fusion.

Miles et al.[48] culminated an extensive investigation sponsored by the U.S. Navy with the results shown in Fig. 7. A similar but independent study was reported by Bush and Lagowski [31] which shows excellent agreement. Gas evolving from energy-producing cells was collected in metal flasks and analyzed by a sensitive mass spectrometer. Samples producing no detectable heat produced no detectable helium, except when a leak was evident. Alloys of Ce-Pd produced detectable heat without helium being found in the evolving gas. Although the errors are large, a consistent pattern is apparent. The production rate is about a factor of 2 lower than expected if the helium were produced by a fusion reaction. This difference might be caused by helium being retained by the palladium. Presence of any helium at all in the surrounding gas implies that a significant number of the helium producing reactions occur within a few microns of the surface. Helium formed deeper within the structure is known to be completely retained by the metal. Helium production by the reaction sequence d + 10B = 4He + 8Be followed by 8Be = 2 4He would seem to be ruled out by the consistent behavior of the boron-containing sample.

II.2.2.3 Neutron Emission

Over 300 studies have attempted to find neutrons, most with no success at all. A few careful and lucky studies have demonstrated neutron emission and determined their energy. Takahashi [53] was one of the first and the most successful. Figure 8 shows the gross emission rate while a cell was making excess energy. An energy spectrum was obtained using a NE-213 detector with pulse-height analysis. The upper figure shows both signal and background while the lower figure gives the difference between these two values. Energies near 2.45 MeV, 4.5 MeV, and 7 MeV are evident. Other studies have reported similar energy values and emission rates. Although this work was able to correlate excess energy production with neutron emission, many studies have failed to detect neutrons when excess energy was being produced. Part of this failure may be caused by inadequate sensitivity and part may result from an absence of neutrons. It is possible that neutron emission and heat result from different reactions so their occasional synchronicity may be coincidence. Many questions still remain unanswered.

II. 2.2.4 Transmutation Products

Recent studies have revealed a variety of elements that seem to result from fragmentation of a heavy nucleus or its fusion with deuterium or hydrogen. Sometimes fusion seems to proceed fission and sometimes elements other than palladium and nickel are involved. Many claims can be cited but, for the purpose of this review, only two will be examined. The Journal of New Energy [54], Fall 1996, and Infinite Energy, Vol. 3, #13 & 14, (1997)[55] contain many papers claiming to have produced anomalous transmutation. A number of organizations [56] are proposing to use this effect to quickly convert radioactive isotopes to stable ones. These claims are still uncertain and are being actively pursued using a variety of methods and chemical environments.

Miley (Univ of Illinois)[57] and Patterson (CETI) have collaborated to examined thin coatings of mainly nickel on plastic beads after they were electrolyzed in a flowing electrolyte of Li2SO4 and H2O. Great care was taken to analyze the material before hand and to remove possible contaminates from the electrolyte. Significant quantities of Fe, Ag, Cu, Mg, and Cr were detected using neutron activation analysis (NAA), energy dispersive X-ray (EDX), Auger electron spectrometry (AES) and secondary ion mass spectrometry (SIMS). Many other elements were also found but at lower concentrations. When the concentrations are plotted as a function of atomic number, four regions of enhanced concentration are produced with peaks at 15, 30, 50 and 80 au. Many of the minor elements probably result from the expected localization of impurities. However, the major elements are at such high concentrations making this explanation difficult to support. The main anomalous elements are noted in the Periodic Table shown as Fig. 9. Many of the detected elements show an abnormal isotopic ratio and have a higher concentration within the Ni layer in contrast to being found on the surface as would be expected if they plated out of the electrolyte. While most questions about the analytical methods have been answered, the nature of the nuclear process is still very much in doubt. The main problems involve how elements much heavier than Ni are produced, how the neutron/proton ratio between the proposed reactants and products can be balanced, and why the measured energy production is so small compared to the amount of nuclear transformation. Of course, the basic question remains as to how such reactions can occur in the first place.

Mizuno et al. [58] (Hokkaido Univ. Japan) subjected palladium to electrolysis at high pressure and high temperature. The electrodes were analyzed using EDX, AES and SIMS. Although this study is not as complete as the one described above, many of the same elements were found with abnormal concentrations and isotopic ratios. Surprisingly, significant excess Xenon was detected within the palladium metal using SIMS. Changes in the 104Pd and 110Pd isotopic ratio were also seen as a function of depth with the largest deviations from natural abundance at the surface. Abnormal isotopic ratios resulting from formation of metal hydride molecules, which distort SIMS measurements, or because of isotopic separation caused by electromigration may occur but are difficult to justify in all cases.

II. 3.1 Gas Reaction

Arata and Zhang (Osaka Univ., Japan) explored energy production when finely divided palladium was exposed to high-pressure D2 gas. In their case, the gas was produced by electrolyzing a palladium tube containing powdered palladium (palladium-black) in a mixture of LiOH and D2O. Heat was seen after a delay of many days and after the internal pressure had risen to as high as 800 atm. Power levels between 10 and 20 watts were measured using flow calorimetry. Samples maintained this power for months and excess power could be re-established after the samples had been stored for over a year. Helium-4 and now 3He have been extracted from energy-producing samples by heating them to 1200° C. The surrounding gas has not yet been analyzed. Because the work has been done with great care, interest in the results is growing. Several attempts in Japan to duplicate the results have failed because, according to Prof. Arata, the proper protocols were not followed. Additional efforts are underway in the U.S.

II. 3.2 Electric Discharge

Dufour and co-workers[59](Shell/CNAM, France) initiate anomalous effects by producing a silent AC discharge through an insulator surrounding a cell containing hydrogen (deuterium) gas and a metal electrode. This method involves a so-called "Ozoniser". The voltages are not sufficient to cause conventional nuclear reactions. Unusual effects have been seen including excess power production up to 10 W for H2 and 14 W for D2, significant loss of hydrogen isotopes from the cell, emission of ionizing radiation for days after the discharge is stopped, and increased concentration of lithium. The results depend on the nature of the electrode and the gas.

A group at "Luch"[60] (Scientific Industrial Association, Russian Federation) pioneered and continues to study a technique involving high voltage-pulsed discharge in low-pressure deuterium or hydrogen. Excess energy, various radiations, and transmutation products have been reported. The amounts and types of anomalous products depend on the amount of applied current, the metal used as the cathode, and whether H2 or D2 is used. The combination of D2 with one special palladium sample produced 21 W of excess power when 95 mA was applied with a nearly linear relationship between heat and applied current. Other palladium cathodes were not as productive. Various stable elements were found in the cathode surface after excess energy was seen including Na, Mg, Ti, Fe, Ni, Cu, Rb, Zr, Nb, Rh and Ag, some with abnormal isotopic ratios. Silver was also found to produce excess power of 9 W at 51 mA when bombarded with deuterium ions. Normal hydrogen produced excess power when nickel or niobium cathodes were used but the number of transmutation products was reduced. Minor amounts of neutron and gamma emission have been detected, sometimes continuing after the current was stopped. Efforts to duplicate the results at the Naval Research Laboratory (NRL, US) were only marginally successful, possibly because the method was not completely duplicated and the same palladium was not used.

Prelas et al.[61] (Univ Missouri, US) used a microwave heated deuterium plasma to bombard palladium with ions having bulk temperatures between 0.5 and 10 eV. They detected a significant increase in neutron and gamma emission only when the sample was being bombarded. In one case, a broad gamma peak at 8.11 MeV was seen. The results were sensitive to the type of palladium used and its temperature. Higher energies near 11 kV were used by workers in China[62]. X-ray emission near 27 keV was routinely observed when various metals were bombarded with deuterium or hydrogen ions.

II. 3.3 Proton conduction through semiconductors

When certain semiconductors are heated in deuterium and a voltage is applied across their thickness, a small current can be made to flow, caused by dissolved hydrogen ions moving within the structure. Biberian [63](Faculté des Sci. de Luminy, France) applied this technique to AlLaO3 while Mizuno et al.[64](Hokkaido Univ., Japan) used Sr(Ce,Nb,Y)O3. Both studies observed significant excess energy production. Oriani [30] (Univ. of Minnesota, US) succeeded in duplicating the results of Mizuno using a high-temperature Seebeck-type calorimeter. Although the amount of excess energy is small, it represents a large increase over the amount being applied. Oriani measured some excess even after the applied current was turned off. Another duplication using SrCeO3 and BaCeO3 ceramics was done in Russia [65] but the description lacks much needed detail. In this case, occasional excess energy was measured as well as neutron emission during thermal cycling. The behavior depended on the chemical purity and structure of the ceramic, an experience shared by the other studies.

II. 3.4 Rapid decomposition of a hydride

Yamaguchi and Nishioka[66] (NTT, Japan) first showed charged particle emission when palladium containing deuterium is rapidly heated in vacuum. Alpha particles with an energy of 4.5-6 MeV and protons with an energy of 3 MeV were detected. The samples were palladium coated on one side with MnOx and the other with gold. Iwamura et al.[67](Mitsubishi, Japan) used palladium coated with gold or aluminum and detected neutron (5s) and tritium production in a few samples. However, extensive efforts to reproduce this work in Japan were unsuccessful. Recently, Lipson et al.[68] (Russian Academy of Sciences, Russia) detected neutron and gamma emission when palladium, coated on one side with a complex oxycarbide (based on an unpublished method of application) and the other with gold, was heated in air or oxygen. They report seeing 100-500 neutrons/sec-cm2 and gamma ray peaks at 2.22 MeV, 3.8 MeV, and 6.3 MeV as well as excess energy over that expected from the D2-O2 reaction. Attempts to duplicate this work are presently underway. Unfortunately, the nature of the palladium, once again, plays a major role in producing the effects. In addition, the presence of other chemical environments, applied as layers, creates additional regions for abnormal nuclear reactions to occur, thereby creating additional variables.

II .3.5 Cavitation involving bubble formation

Stringham and George [69](E-Quest, US) have been perfecting a method based on generating bubbles in D2O using an intense acoustic field. When these bubbles collapse against a metal surface, they inject deuterium and oxygen ions into the metal as a high-temperature plasma. The deuterium diffuses away from the surface while the oxygen remains trapped and forms a colored oxide. Anomalous behavior is immediate. Silver and palladium are especially good producers of anomalous energy, helium, and various transmutation products. The fact that many other metals produce no unusual results indicates absence of reactions associated with sonoluminessence within the bubbles. Unfortunately, details of the process and the results are not available to the public although a general description can be found in Infinite Energy magazine.[70] This is one of the few methods having high reproducibility and producing significant amounts of energy and nuclear products. People interested in the method can obtain more information by contacting the inventors.

A related approach has been developed by Griggs[71]. In this case, bubbles are generated by a perforated rotor which is rotated within normal water by a powerful motor. Steam is produced and the bubbles collapse against aluminum and steel. Several independent tests of the method have all found more energy produced by the device than used to rotate the rotor. Nevertheless, the company sells the units only as a way to obtain efficient, maintenance-free energy conversion.

II.3.6 Biological system

Beginning in 1954, Kervran[72] was the first to make a systematic study of nuclear reactions in biological systems using experiences by farmers and biologists. A modern study claiming transmutation was made in Japan using various cultures [73] and conventional analytical techniques. A recent study done in Russia[74] has added powerful support to the accumulating evidence. In this case, fusion between 55Mn and deuterium to give 57Fe in various yeast and bacteria cultures was demonstrated using the Mössbauer effect. The relative velocity between a 57Co gamma emitter and the solution was changed until the gamma energy matched that required for absorption by any 57Fe nucleus present. The result for yeast and bacteria cultures can be seen in Fig. 10. The 57Fe is produced only when Mn and D are both present in the culture and at a rate of (1.9±0.5) x 10-8 57Fe per sec per 55Mn. A double hump in the spectra occurred for the bacteria culture whether the 57Fe was made in or added to the culture. Apparently, 57Fe made by a nuclear reaction and 57Fe added to the culture both occupy the same chemical environment, but an environment that is different from the one in the yeast culture. This study is particularly persuasive because it uses a conventional method which is only sensitive to the presence of 57Fe, an isotope easily excluded from the environment. Therefore, strong evidence exists for at least one nuclear reaction. What other reactions are possible?


Well over 100 explanations of CANR have been attempted, most of which have little relationship to reality or to being useful. However, several models do seem to offer important, partial insights into possible processes. Many of these models have continued to be changed and improved over the years. Unfortunately, most theories address the nuclear process while ignoring the unique environment in which such reactions must occur. In addition, most models still conflict with some observations within the field or rest on assumptions not supported by observations made outside the field.

Because the nuclear events occur in a variable number of random sites within the bulk material, a quantitative relationship between theory and observation is not yet possible even though some attempts have been made. No theory has explained

why only these rare sites are active or has predicted their chemical characteristics except in general terms. Confusion still swarms around the nature of the nuclear reactions. Is the detected helium produced by d-d fusion or does it result from alpha decay of destabilized heavy nuclei? Can mass be added to heavy elements by fusion between multiple hydrogen (deuterium) nuclei or are heavier nuclei involved? Why do some nuclear processes produce detectable energy while others produce none? How many theories will it take to finally understand all observations?

A few examples are summarized below to give a partial insight into the approaches being explored. It is still very risky, and likely to raise the emotional temperature in some quarters to suggest which of these ideas might be right or wrong. Therefore, neither the choice of examples nor their sequence represent a judgement of value. Courageous evaluations of several theories have been provided by Chechin[75] and Preparata[76].

Chubb and Chubb[77] (OIC, US)

This model uses ion band state theory involving stationary, three dimensional Bloch states to explain the fusion process. Thus, a few deuterium nuclei are thought to act as a wave when the deuterium concentration has achieved a critical value and when other conditions are present. These waves have a period equal to the distance between the deuterium lattice positions, thereby reducing coulombic repulsion. Occasionally, fusion occurs in small steps during D+-D+ wave overlap thereby producing a helium nucleus having gradually increasing stability. The resulting nuclear energy is coupled to the lattice by a coherent process involving transfer of energy in small packets to the Fermi levels from which it is dissipated throughout the lattice. The theory predicts that a critical crystal size is required and that 4He is the only nuclear product without gamma or any other radiation being emitted. However, the proposed particle-wave transition has yet to be demonstrated to occur in a crystal lattice.

Preparata[78] (INFN, Italy)

Using an approach called QED (Quantum Electro Dynamics), the model proposes that various coherent plasma fields exists in a crystal structure, i.e. the fields act like electron lasers which are completely contained within the structure. These fields are proposed to combine and provide sufficient screening of deuterium nuclei to allow fusion and other reactions to take place. The model requires the fusion process to occur at tetrahedral sites within b-PdD when the deuterium concentration has reached very high values. Released energy is absorbed by the coherent fields which then emit X-rays. However, no evidence exists for tetrahedral occupancy by deuterium in b-PdD.

Bazhutov [79](Erzion Center, Russia)

Small quantities of massive, stable hadrons left over from the Big-Bang are thought to be present in all matter. Under certain conditions, these particles can be released from their bound state and used to catalyze nuclear reactions. McKibben[80] (LANL, US) has taken a similar approach by proposing the presence of fractionally charged particles. These particles can stabilize composite nuclei which act chemically like normal matter and, when destabilized, can produce energy by catalyzing various nuclear reactions. The existence of these particles has yet to be established.

Li [81](Tsinghua Univ., China)

A very narrow energy level is proposed to exist in the nucleus which is able to resonate with certain energy levels in the surrounding atomic lattice. The process is proposed to promote barrier penetration and nuclear interaction. The presence of this narrow level or its relationship to nuclear stability has not been established.

Hagelstein [82](MIT, US)

He has abandoned the virtual neutron model and replaced it by one involving energy transfer from a phonon laser or strong resonance vibrations operating within the lattice. Vibrations of individual atoms caused by their temperature combine to produce pockets of higher energy (temperature). Energy is accumulated in the phonon bands by nonlinear frequency shifting involving fluctuations in the phonon spectrum. This enhanced energy is transferred to a few atoms by the usual vibrational processes causing adjacent nuclei to approach each other with sufficient energy to allow various nuclear interactions including fusion and transmutation. The "Lattice Quake" model of Arata[83] (Univ. Osaka) uses similar features to explain a fusion reaction. Kucherov[84] (ENECO, US) has employed phonons to propose a gradual accumulation of energy directly in the energy levels of various metal nuclei. When sufficiently high, this energy is released by a emission or fission. The quantized nature of the nuclear constituents and their very high energy levels would appear to make such an incremental transfer of energy impossible.

Mills[85] (Blacklight Power, US)

All isotopes of hydrogen are proposed to have fractional energy levels below those described by conventional quantum theory. Electrons can be caused to access these levels if a suitable repository for the released energy is available. Consequently, under certain conditions, energy can be released by the formation of collapsed hydrogen atoms called Hydrinos. These smaller hydrogen atoms leave the system and return to their initial size and energy elsewhere in the world. Therefore, this is an energy transfer process which does not involve nuclear reactions. However, tritium can sometimes result when complete atom collapse produces a neutron which reacts with a nearby deuterium nucleus. Variations on this approach have been proposed by Dufour[86] (Shell, France) and Yi-Fang and Zheng-Rong (Yannan Univ., China)

Kozima[87] (Shizuoka Univ., Japan)

Thermal neutrons are proposed to be trapped in crystals where they can, under the proper circumstances, interact with nuclei. This is called the TNCF model (Trapped Neutron Catalyzed Fusion). The neutrons are thought to be stabilized by forming neutron Cooper pairs and by acting as Bloch waves which prevent normal neutron decay and prevent interaction with nearby nuclei until a large perturbation is suffered by the crystal. This perturbation is proposed to be caused by certain surface impurities, some of which subsequently react with the released neutrons. The approach is handicapped by the need to make several arbitrary assumptions to make the model consistent with observation. These assumptions include the need to justify why neutrons are not emitted after they become available to interact with the surrounding nuclei and why significant tritium is not formed when lithium is present.

Miley et al.[88](Univ. Illinois, US)

The region between two metals having large differences in Fermi electron levels is proposed to provide an environment in which the coulomb barrier can be reduced, the so-called swimming electron layer (SEL). Conduction electrons which concentrate at a metal surface, producing a characteristic work function, are thought to form a plasma of sufficient magnitude to partially shield deuterium nuclei located within an interface region. The reduced average distance between nuclei caused by this screening increases the normal fusion rate. Recent work by this group[89] applies the concept to transmutation reactions. In this context, nuclear interaction is proposed to occur at much larger distances than normally experienced for d-d fusion. The SEL is proposed to reduce the distance by an amount sufficient to cause enhanced nuclear interaction. However, anomalous nuclear reactions are seen when conditions required for the swimming electron layer to form are absent.

Kim [90](Purdue Univ.,US)

This is not a general model but a more exact calculation of the magnitude for the coulomb barrier using conventional methods. The analysis concludes that barrier penetration for fusion is easier than first thought, but not as easy as is required to explain many CANR claims.

Many models propose a resonance processes but they differ in what is resonating and how the resonance structure interacts with the nucleus. Preparata sees a wave-like electron structure neutralizing the coulomb barrier while Miley proposes enhanced concentration alone can do the trick. The Chubbs visualize direct interaction of deuterium waves; Kucherov and Li would have vibrational energy of the atom-electron structure add energy directly to the nucleus thereby causing destabilization; and Hagelstein and Arata have this vibrational energy cause direct nuclear interaction between adjacent nuclei. Each model has the electron structure carry away the resulting nuclear energy which is dissipated throughout the lattice as heat.

The models proposed by Mills and Dufour, involving a change in energy and size of the hydrogen atom, are not part of CANR unless this change results in a nuclear reaction. Although most examples of heat production are not consistent with this model, occasionally results are seen which are consistent. Perhaps several mechanisms are possible depending on the imposed conditions.


Separating fact from fiction is the main problem in evaluating the value of these extraordinary claims. The basis for accepting evidence for such unconventional nuclear reactions is very much in the mind of the beholder. Is the reader familiar with the technique? Are the investigators known to the reader and can they be trusted? Is the work peer reviewed by an acceptable journal? Unfortunately, most people working in this field are not known to the general physics community and most of the work is not peer reviewed by acceptable journals. I have attempted in this review to present enough detail about some of the better work so that a reader will at least appreciate why this field is receiving increased attention. Although definite conclusions are still not possible, some general trends are apparent. Unfortunately, many studies can not be judged because so little detail has been properly described and many of the techniques have not been used by several independent investigators. On the other hand, some very good studies are available which use conventional, well accepted techniques. In addition, similar results have been obtained using a wide range of methods and techniques in laboratories throughout the world, only a small fraction of which are described in this review. While this common experience is not proof, it encourages continued open-minded interest in the field.

World-wide experience using many techniques show that the mix of nuclear products and the ability to make excess energy are both very sensitive to the nature of the chemical environment in which the reactions are thought to occur. Because few experiments use identical chemical environments, the results seem to have no clear pattern and are difficult to reproduce. Although error and incompetence add to the problem, these are clearly not the main variables. Accepting this insight is fundamental to accepting the claims.

Neutron production, the favorite of many physicists, does not occur at a significant level and may not be associated with heat production. Therefore, the unusual energy of the emissions can not be used with confidence to understand heat production. On the other hand, some neutron emission is detected and must be explained. Intense bursts are sometimes observed and are usually associated with changes in the chemical or physical environment. This behavior has prompted interest in the effect of crack formation, so-called fractofusion. Such an environment would be expected to produce "hot" fusion caused by the high voltages produced therein. However, the required amount of tritium is not observed when neutrons are detected.

Tritium can be produced, but only with difficulty. Although it is seldom found in cells making excess energy, its production may be accompanied by a few emitted neutrons. Dendrites on the cathode surface seem to be present when it is formed in an electrolytic cell. Intense electric fields or high voltage discharge encourage its formation even though the voltages are lower than conventional theory would require.

Helium or a few transmutation products are present after excess energy is detected. The mix between these products depend on the raw materials present. Deuterium produces helium while light hydrogen produces various transmutation products involving at least the alkali metals. Helium-3 may be produced when both H and D are present in sufficient concentrations. Some of the helium is produced with sufficient energy and sufficiently close to the surface to be identified as alpha particles.

More complex reactions giving a spectrum of products seem to be possible. Various fragments of nickel or palladium have been detected, some with abnormal isotopic ratios. Occasionally, short-lived radioactive isotopes are detected. More difficult to justify are the heavy elements which can only result from multiple addition of hydrogen nuclei or other light elements to palladium or nickel. These new claims are still being debated and offer a significant challenge to any theory. Particularly troublesome are the lack of resolution in the neutron/proton ratio between the apparent reactants and products, and the lack of corresponding energy production.

X-rays are occasionally detected using various methods. However, the emissions are sporadic, they do not act like bremsstrahlung nor can they always be identified as characteristic X-rays. An intriguing observation is reported by Rout et al. [91](Bhabha Atomic Research Center, India). Every palladium sample, after being loaded with either deuterium or hydrogen, produced an emission which is influenced by electric and magnetic fields, able to pass through certain thin absorbers, and able to produce fogging of X-ray film. Presence of oxygen enhanced the effect. A very complete study rules out conventional types of radiation or chemical products. The source of this radiation as well as X-rays detected in other studies is still uncertain. However, these observations continue to be consistent with early studies and continue to support the initial conclusion that the nuclear energy is somehow dissipated throughout the atomic lattice rather than at the site of its production.

Gamma emission is occasionally seen sometimes with a measurable half-life after the experiment is stopped. However, no study has observed gamma emission having a consistent intensity or energy. Once again, the behavior apparently depends on the nature of the chemical environment.

Everyone can agree that a nuclear reaction involving coulomb barrier penetration does not occur within ordinary matter at any significant rate. The problem remains as to whether it can occur at all in some rare and unusual environments. Experience shows that a few materials have the capacity to contain these environments, i.e. to produce anomalous effects. This selectivity not only demonstrates absence of random error but also suggests the existence of a universal nuclear-electron interaction, albeit negligeable in most cases. Once a nuclear-active material is formed, a wide spectrum of nuclear reactions appear to be possible within the same sample. How is this possible? How is the choice made as to which reaction will occur? A logical possibility involves inhomogeneities within the material in which one reaction takes place while other isolated regions cause a different reaction. This is consistent with the common observation of anomalous elements occupying isolated regions, generally near the surface. If this conclusion is correct, we are faced with the possibility that, once a nuclear-active material is produced, minor variations can have a large effect on the type of nuclear reaction produced. Understanding the nature of these isolated regions is fundamental to making this phenomenon occur at high levels and with predictability. Such knowledge is also basic to any successful theory.


I can sympathize with anyone struggling with the reality of and meaning behind these disparate observations. Because skepticism is so wide spread, an intense effort by skilled and well funded scientists required to answer many of the questions has not been applied to the problem. The need for proof to be "compelling", as many scientists require, adds to the difficulty. Nevertheless, I suggest sufficient information is now available to strongly support claims for certain nuclear reactions taking place under conditions not sanctioned by conventional theory.

The human race is desperate to find energy sources which are pollution-free and renewable. In addition, we have the serious problem of disposing of radioactivity from fission reactors and the nuclear arms race. The CANR phenomenon may offer a solution to both problems. I suggest that lack of certainty about its reality or a lack of knowledge about many of the details should be viewed as temporary distractions and not used as justifications for ignoring the potential benefits. Even if many of the claims have trivial explanations, the evidence overwhelmingly indicates the existence of a novel phenomenon having unexplored benefits. Why not test the possibilities no matter how remote they seem?


After receiving a blanket rejection, I sent the following letter to the editor of Rev. Modern Physics.

Dharam V. Ahluwalia
MS-H846, P-25, LANL
Los Alamos, NM 87545

Dear Dr.Ahluwalia,

I would like to comment on the review of my manuscript you e-mailed to me.

I have worked in science for over 40 years, publishing during that time over 70 reviewed publications and several books. I know what a review is supposed to do and how to be critical. The problem is that when the subject concerns cold fusion, the rules change. What the reviewer wants is either a demonstration that the claims are wrong or a proof so compelling that no room for doubt can exist. Such a proof is only accepted if the specialized language and theories previously accepted by the reader are used. Unfortunately, the field is not yet at that level of understanding. I had hoped that a few people in the physics profession, like yourself, would be interested in knowing where knowledge and claims in the field stand at the present time.

An example of this need for using the theoretical approach previously accepted by the reviewer is demonstrated by the reviewer's comment about my lack of knowledge of quantum mechanics. He has no basis for this judgement, being only able to know my description of the approach other people have taken. Based on his judgement of me, I conclude that he would only be satisfied if I rejected these approaches as he would - because they do not fit with his understanding of quantum mechanics. He thereby demonstrates that he is unwilling to consider any approach which does not fit his reconceptions. I call this attitude closed minded and not befitting a good scientist.

It is sad to me, that in spite of your interest in publishing information about the field, you could not find a few people in physics who would at least open a dialogue about the form for such a review. A blanket rejection is totally unexpected and will only add to the bitter reactions to conventional physics when the phenomena are eventually demonstrated and applied.


Edmund Storms