Anomalous effects in hydrogen-charged palladium — A review

Abstract

There are more than 10 groups world wide that have reported the measurement of excess heat in 1/3 of their experiments in open and/or closed electrochemical cells with a Pd solid metal cathode and deuterium containing electrolyte, or D2 gas loading of Pd powders (see Table 1 of the main text). Most of these groups have occasionally experienced significant events lasting for time periods of hours to days with 50–200% excess heat measured as the ratio between electrical input energy and heat output energy. Moreover, these experimenters have improved their methods over time and it is to be noted that the reported excess heat effect has not diminished in frequency or magnitude. This paper cites selected data generated over the past 15 years to briefly summarize what has been reported about the production of excess heat in Pd cathodes charged with deuterium. A set of new materials experiments is suggested that, if performed, may help to reveal the underlying mechanism(s) responsible for the reported excess heat.

Full text

Anomalous effects in hydrogen-charged palladium —A review
G.K. Hubler
U.S. Naval Research Laboratory, Washington, DC 20375, United States
Available online 13 March 2007
Abstract
There are more than 10 groups world wide that have reported the measurement of excess heat in 1/3 of their experiments in open and/or closed
electrochemical cells with a Pd solid metal cathode and deuterium containing electrolyte, or D
2
gas loading of Pd powders (see Table 1 of the main
text). Most of these groups have occasionally experienced significant events lasting for time periods of hours to days with 50–200% excess heat
measured as the ratio between electrical input energy and heat output energy. Moreover, these experimenters have improved their methods over
time and it is to be noted that the reported excess heat effect has not diminished in frequency or magnitude. This paper cites selected data generated
over the past 15 years to briefly summarize what has been reported about the production of excess heat in Pd cathodes charged with deuterium. A
set of new materials experiments is suggested that, if performed, may help to reveal the underlying mechanism(s) responsible for the reported
excess heat.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Electrochemistry; Palladium; Hydrogen
Contents
1. Introduction ............................................................. 8568
2. Lack of excess heat in 1989–1990 ................................................. 8569
3. Summary of past work ....................................................... 8569
4. Recent examples of excess heat ................................................... 8569
5. Material properties of Pd at high hydrogen concentration ...................................... 8571
6. Summary and conclusions ...................................................... 8572
Acknowledgements ............................................................ 8572
References ................................................................ 8572
1. Introduction
Fig. 1 is a schematic diagram of a modified, planar geometry
Fleishmann and Pons cell presented here to review its main
features [1]. A Pd cathode plate positioned between two parallel
Pt-plate anodes are immersed in electrolyte (0.1 M LiOD in
D
2
O). Voltage applied between the electrodes causes hydrogen
to enter the cathode. Hydrogen that evolves from the cathode
and oxygen that evolves from the anodes are recombined by a
catalyst residing above the liquid in closed cells, and allowed
escape in open cells. A thermocouple measures the temperature
of the electrolyte. The H/Pd ratio of the cathode is measured in
situ by means of a four-point probe resistivity ratio R/R
0
, where
R
0
is the initial resistivity value, and the R/R
0
versus H/Pd is
compared to literature values. This in-situ monitoring of the
hydrogen concentration in Pd was not in use in the first year
after the Fleishmann and Pons announcement [2]. Details of
calorimetry will not be discussed in this short review.
Motivated by the report of excess heat by Fleischmann and
Pons [1], a number of research groups from around the world
have been conducting experiments on the Pd–D materials sys-
tem more or less continuously since 1989. Considerable pro-
gress has been made on several fronts that include improved
reproducibility of high loading (i.e., D/Pd N0.90) of deuterium
Surface & Coatings Technology 201 (2007) 8568 –8573
www.elsevier.com/locate/surfcoat
E-mail address: hubler@ccs.nrl.navy.mil.
0257-8972/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.surfcoat.2006.03.062

in Pd cathodes, improved reproducibility of measured excess
heat in highly deuterium-loaded Pd cathodes, and improved in-
situ measurement of the status of the Pd cathodes [2]. The
materials science of hydrogen loading is now well enough
understood that it may offer a reason why the many groups that
tried to reproduce the heat effect in 1989–1990 were
unsuccessful. Less encouraging are the facts that there is still
no viable physical mechanism to explain the heat effect, and
triggering the heat effect is still not empirically understood.
2. Lack of excess heat in 1989–1990
Many research groups attempted to reproduce the excess
heat production and the neutron and gamma ray signals reported
by Fleishmann and Pons in the year following their spectacular
announcement. The nuclear data that Fleishmann and Pons
presented proved to be in error and have never been reproduced.
However, there remains the possibility that their excess heat
production results may have been correct. One reason that most
researchers were unsuccessful in achieving heat production may
have been at least in part due to the lack of understanding of
how to achieve H/Pd ratio of N0.90. Procedures required to
achieve this loading [3] that were not generally known or
followed include:
–Pd cracks under loading (∼4% vol. change), especially for
pure, annealed Pd, and the internal surface area of cracks are
recombination sites for hydrogen and deloading in the cracks
competes with electrolytic charging and the net loading is
b0.9.
–Pd must be strengthened and toughened by alloying or
mechanical treatment to avoid surface cracks.
–Pd must have an optimum grain size to load. H entry is
primarily along grain boundaries.
–Pd must be loaded gently (gradually or cycled in current) or
it will crack even after the above cathode preparation steps
(loading time diffusion time).
–Achievement of high loading with D is many times more
difficult than for H.
–Catalytic surfaces are easily contaminated and often will not
maintain water electrolysis efficiency over the time that is
required for high H loading.
Unable to achieve high loading, and, therefore, excess heat,
most researchers declared that heat production in Fleishmann
and Pons cells is not a real effect and ceased working on the
experiments.
3. Summary of past work
Table 1 is a partial list of groups that are or have been active
in this area [4–13]. Previous work by many researchers has
determined some necessary but not sufficient conditions to
observe excess of heat [14]. These are:
High D Loading (xN0.90; PdD
x
)
High electrical current (∼250 mA/cm
2
)
Dynamic trigger that imposes a D flux in, out or along the
cathode (Δtemperature (ΔT), Δcurrent flow (ΔI), laser)
Abrupt changes in any of these parameters can stimulate the
production of excess heat with greater frequency than a system
maintained at steady state. It has been empirically found that a
He–Ne laser (∼10 mW CW) impinging on the Pd surface can
stimulate the excess heat effect as well [15]. Two emerging
trends are, that the incubation time for heat production decreases
as the cathode volume decreases, and, that increased surface-to-
volume ratio increases the specific energy/mass value of excess
heat production.
4. Recent examples of excess heat
Two examples cells that are reported to produce excess heat
are presented in the next 3 figures. In Figs. 2, 3 and 4, the cell
was in a planer geometry (as in Fig. 1), closed system with
100 μm thick, ∼4-cm
2
-area Pd foil, and run in current control
mode. Fig. 2 shows time versus input power on the right
ordinate and versus integrated energy on the left ordinate, for a
Fig. 1. Schematic representation of modified Fleishmann–Pons electrolysis cell
used at ENEA, Frascati, Italy, and Energetics Tech., Ltd.
Table 1
Selected list of research groups that have measured excess heat in the palladium
deuteride material system
Research group Institution Cathode type Reference
Arata Osaka University Powder [4] (1994)
Fleischmann University of Utah Rod [1] (1990)
Lautzenhiser Amoco Research Laboratory Ingot [5] (1990)
Lesin Energetics Tech., Ltd., Israel Foil [6] (2004)
McKubre Stanford Research International Rod [7] (1993)
Mengoli CNR IPELP, Italy Foil/rod [8] (1998)
Oriani University of Minnesota Rod [9] (1990)
Swartz Jet Thermal Products, USA Wire [10] (2006)
Violante ENEA, Rome, Italy Foil [11] (2004)
Will University of Utah Rod [12] (1990)
LabA Government lab, USA Rod [13] (2001)
LabB Company lab, USA Powder [13] (1998)
8569G.K. Hubler / Surface & Coatings Technology 201 (2007) 8568–8573

control cell in which the electrolyte is H
2
O in 0.1 M LiOH [11].
Input power is determined by measurement of current and
voltage and output power is measured by calorimetry. Note the
time lag of response of the input and output power that is
characteristic of the time constant of the calorimeter. The total
integrated input energy is a well-behaved straight line, and the
integrated output energy shows a curve that is below the input
energy and is characteristic of small energy losses due to the
97% efficiency of the calorimeter.
Fig. 3 shows time versus excess power (output power–input
power) on the right ordinate versus integrated energy on the left
ordinate, for a cell identical to that in Fig. 2 and in which the
electrolyte is D
2
O and 0.1 M LiOD [11]. At time 25,000 s, the
cell begins to display excess power that behaves erratically until
the cell is turned off. The integrated input energy monotonically
increases and the integrated output energy initially displays a
transient often seen as the cell heats up due to the initial turn on
and is dependent on the initial cell temperature and input power
history. The integrated output energy rises faster at the time the
excess power is observed, and rises above the input energy by
the end of the run. Taking into account the 3% energy loss of the
calorimeter, this cell displayed a total of 8% excess power in the
form of heat for the duration of this experiment.
Fig. 4 shows input and output power versus time for a cell in
which the electrolyte is D
2
O and 0.1 M LiOD. The input power
was cycled between high and low values. At 220,000 s, the cell
begins to display excess power that continues until the experi-
ment is shut down. This cell produced 50% excess power, or
2.2 net W, averaged over a period of 12 days [6].
The primary criticisms of experiments that measure heat
production during electrolysis of Pd in deuterium containing
electrolytes are:
1. Energy is stored by some as yet unknown but straightforward
mechanism during long incubation times and then released
(battery).
2. Excess heat due to recombination of oxygen and hydrogen in
cell (battery).
Fig. 2. The integrated input and output energy (left-hand scale) and the instantaneous input and output power (right-hand scale) versus time for a reference experiment
with hydrogen at ENEA using a modified Fleishmann–Pons electrolysis cell (H
2
O + 0.1 M LiOH). Calorimeter power error is ± 10 mW at 100–3000 mW [11].
Fig. 3. The integrated input and output energy (left-hand scale) and the instantaneous excess power (right-hand scale) versus time at ENEA using a modified
Fleishmann–Pons electrolysis cell (D
2
O + 0.1 M LiOD). Calorimeter excess power error is ± 10%, or about 6 mW at the peak in excess power at 50,000 s [11].
8570 G.K. Hubler / Surface & Coatings Technology 201 (2007) 8568–8573

3. Calorimeter is not calibrated correctly (experimental error).
4. Energy inventory not measured correctly (experimental
error).
Criticisms 1 and 2 are stored chemical energy explanations.
Number 1 appears to be questionable since incubation times in
some experiments that use small volume cathodes are as short as
a few minutes and experiments that integrate the total energy do
not detect endothermic processes that would signal energy
storage prior to excess heat release. Number 2 appears to be
questionable since excess heat has been measured in many
closed-cell experiments where the energy from recombination
of hydrogen and oxygen is continuously recovered. Individual
experiments can always be criticized as having experimental
errors and one might conclude that all researchers in Table 1
make similar subtle mistakes in order to refute their reports of
excess heat. One might also conclude that there is ample
evidence that excess heat is produced under certain conditions,
and, that the results in Table 1 are consistent over time and
collectively suggest that a closer look at this materials system is
warranted.
5. Material properties of Pd at high hydrogen concentration
While there was a flurry of activity in the 1980s concerning
superconductivity in Pd–H alloys, most of the research at that
time involved hydrogen concentrations far less than Pd/H ratio
of 0.90. The remainder of this paper discusses experiments that
if done, could shed light on possible mechanism(s) that produce
excess heat.
First, it is desirable to perform detailed tracking of the
morphological and impurity changes to the cathodes and anodes
ex situ before and after charging with hydrogen using a variety of
methods such as SEM, XRD, XRF and ICP-OES, to name a few.
Second, in-situ experiments are necessary since electro-
chemical loading is one of the few methods to achieve high
loading. Once the cell voltage is turned off, the hydrogen
evolves from the cathode very rapidly, so ex-situ experiments
on this material are impractical. In-situ experiments require
propagation of signals into the liquid to the Pd foils and return.
This restricts the signals that can be used to X-rays, gamma
rays, neutrons and light waves in the transmission band of
water, and to transducers at or near the foil.
Third, a reproducible materials system for such experiments
is an absolute necessity. Fortunately, such a reproducible Pd
cathode material has emerged from the work of group headed by
V. Violante. His group has performed metallurgical studies of
the effects of mechanical treatment and annealing on the ability
to electrolytically load Pd foils with hydrogen [16,17]. Follo-
wing Violante's procedures, it is now possible to load Pd foils
up to H/Pd N0.90, with 100% reproducibility.
Armed with a reproducible PdH alloy materials system, what
experiments can or should be performed? The following sug-
gested experiments are by no means an exhaustive list, but do
represent a cross-section of experiments that investigate dif-
ferent aspects of this material system.
1. Tensile stress —the first-order materials parameter of
pressure and its related quantity tensile stress has never been
investigated. Use of in-situ tensile apparatus on the cathode
would assess the primary effect of stress on the loading, voltage,
current, and temperature characteristics of the basic loading
experiment and would characterize the stress behavior of this
material system in any of the suggested experiments below.
2. High-energy X-ray scattering —this experiment monitors
the Pd lattice as the hydrogen concentration increases. It will
measure lattice expansion and any phase changes that might
occur around Pd:H ratios of 1:1. This experiment has been
performed only up H/Pd ratio of 0.76 [18,19].
3. Neutron scattering —this technique characterizes the
deuterium sub-lattice positions and provides information on Pd
and D phonons with inelastic scattering. This experiment cannot
be performed with hydrogen due to the 9× shorter neutron
scattering length in H
2
O compared to D
2
O.
4. Radioactive isotope spectroscopy —it has been suggested
that a nuclear process is responsible for unusual effects in PdD
[1]. This experiment turns this supposition on its head by
purposely injecting isotopic material into experiments. One
introduces a radioactive isotope or isomer into the Pd by thermal
diffusion, and observes the effects of the PdD environment, if
any, on gamma and X-ray radiation emitted from the decay of
the excited nuclei. For example, one might observe a small
energy shift, or change in the lifetime of the isomer at high
loadings that would signal a chemical effect on the nuclei. Such
influences have been observed [20], and this would be a survey
experiment to determine if there are unusual excitations in this
materials system that affects the nucleus directly. Candidate
isomers are 270-day half-life
57
Co electron capture decaying to
57
Fe, and 2.7-day half-life
198
Au Beta decaying to
198
Hg. Both
have relatively low energy gamma lines (14–412 keV), and Au
also produces a 70 keV X-ray that probes the electron K-shell of
Hg. Other candidates are 367-day half-life
106
Ru Beta decaying
to
106
Pd emitting ∼600 keV gammas, and 4-day half-life
100
Pd
electron capture decaying to
100
Rh emitting high-energy gam-
mas. All four isotopes are soluble in Pd. An alpha emitter might
be monitored by Pd K X-ray excitation.
5. Mössbauer spectroscopy [21] —The isomer
57
Co is
commonly used to explore the hyperfine fields [22] acting at the
57
Fe nucleus in solids. One can survey effects of H environment
Fig. 4. The instantaneous input power (bottom curve) and output power (top
curve) versus time at Energetics Tech., Ltd., using an ENEA Pd Foil in a
modified Fleishmann–Pons electrolysis cell (D
2
O+0.1 M LiOD). This
experiment produced ∼2.2 W (+50%) of average power for ∼12 days
(300 h), with a ∼3 day incubation time [6].
8571G.K. Hubler / Surface & Coatings Technology 201 (2007) 8568–8573

on magnetic and/or electric quadrupole hyperfine fields caused
by distortion of the electron cloud in ns time resolution. In
particular, the isomer shift, δ, indicates the degree of s-electron
distortion that might be caused by the PdH lattice and exci-
tations therein. It also can provide the magnitude of electric and/
or magnetic field at the Fe nucleus. An external magnetic field is
required for hyperfine magnetic studies.
6. Perturbed angular correlations (PAC) [21] —Internal
hyperfine fields can be measured using gamma–gamma coin-
cidence techniques on gamma emissions from radioactive iso-
topes diffused into Pd. A candidate is 367-day half-life
106
Ru
that Beta decays to excited
106
Pd that emits gamma rays in a 624
and 512 keV cascade in time coincidence. The electric qua-
drupole moment of the excited state couples to the hyperfine
electric field and processes. Measuring the time dependence of
the anisotropic angular distribution of the emitted gamma rays
captures this precession. One can obtain the lattice location of
the Pd and the electric field acting at the Pd nucleus. This would
assess possible disturbance of the s-electron orbitals around the
nucleus that might be caused by the present of hydrogen in the
lattice with time resolution of nanoseconds. Another candidate
is 2.8-day half-life
111
In that decays by electron capture to
excited
111
Cd that emits 419 and 247 keV gamma rays in time
coincidence. Many others are possible.
7. Nuclear acoustic resonance (NAR) [23] —It has been
suggested that acoustic excitation of PdD can trigger heat
producing events. The NAR technique is usually used in con-
junction with nuclear magnetic resonance (NMR). NMR is not
well suited to metals and conducting liquids. However, NAR or
just acoustic resonance can be used in conjunction with all of
the experiments listed above, to assess the influence of natural
acoustic resonance in the PdD determined by internal friction
mechanisms and/or the geometry of the cathode. Natural reso-
nances (up to hundreds of kHz) in the cathode excited using
an in-situ or ex-situ acoustic transducer, can exchange energy
with phonons and defects in the cathode that may influence the
measurements of experiments 1–6 above.
6. Summary and conclusions
In this paper reports of anomalous heat in the materials system
of highly hydrogen-loaded Pd were selected from the literature
and highlighted. It was suggested that evidence for anomalous
heat effects is now strong enough to warrant fundamental
investigations of this system. Based upon the availability of new
reproducible Pd–H foil materials with H/Pd ratio N0.90, a case
was made that these foils could be a reliable platform for the
exploration of the Pd–H system at high hydrogen fractions with a
variety of sophisticated in-situ materials science techniques. A
selected list of possible experiments was presented that if exe-
cuted, may help to reveal the underlying mechanism(s) respon-
sible for the excess heat data. The experiments would provide
fundamental materials data on the primary phases and lattice
position of the Pd and H, phonon modes of the H sub-lattice,
stress-modified H-diffusion, influence of the H-rich chemical
environment on nuclear decay, electron cloud distortion around
the nucleus, electric and magnetic hyperfine fields at impurity
nuclei and Pd, time dependence of these fields with nanosecond
resolution, and the effect of acoustic waves on nuclear alignment
in external and hyperfine fields. Individuals acting in isolation
could not conduct these experiments. They require sophisticated
experimental infrastructure, interested participants acting as a
team, and sustained financial support.
Acknowledgements
I would like to thank D. Knies and A. Ehrlich for in-depth
criticisms, P. Hagelstein, J. Aviles, K. Grabowski and D. Nagel
for useful discussions, M. Melich and M. McKubre for assis-
tance with references and history, J. Baglin for helpful insight
and V. Violante and G. Dearnaley for encouragement and
inspiration.
References
[1] M. Fleischman, S. Pons, M.W. Anderson, L.-J. Li, M. Hawkins,
J. Electroanal. Chem. 287 (1990) 293.
[2] M.C.H. McKubre, R.C. Rocha-Filho, S. Smedley, F.L. Tanzella, J. Chao,
B. Chexal, T.O. Passell, J. Santucci, Proceedings of The First Annual
Conference on Cold Fusion, National Cold Fusion Institute, University of
Utah Research Park, Salt Lake City, Utah, 1990.
[3] R.A. Oriani, Fourth International Conference on Cold Fusion, Electric
Power Research Institute, Lahaina, Maui, 1993, 312 Hillview Ave., Palo
Alto, CA 94304, available at (www.lenr-canr.org/LibFrame1.html).
[4] Y. Arata, Y.C. Zhang, Proc. Jpn. Acad 70B (1994) 106.
[5] T.V. Lautzenhiser, D.W. Phelps, Report by Amoco Production Company
Research Department, 1990 available at (www.lenr-canr.org/LibFrame1.html).
[6] I. Dardik, T. Zilov, H. Branover, A. El-Boher, E. Greenspan, B.
Khachaturov, V. Krakov, S. Lesin, M. Tsirlin, in: J.-P. Biberian (Ed.),
Proc. ICCF11 Marseille (FR), World Scientific, 2006, p. 84.
[7] M.C.H. McKubre, et al., Second Annual Conference on Cold Fusion, “The
Science of Cold Fusion”, Societa Italiana di Fisica, Como, Italy, 1991,
Bologna, Italy; and in M.C.H, McKubre et al., EPRI report no. TR-104195,
1993.
[8] G. Mengoli, M. Bernardini, C. Manduchi, G. Zannoni, J. Electroanal.
Chem. 444 (1998) 155.
[9] R.A. Oriani, J.C. Nelson, S.-K. Lee, J.K. Broadhurst, Fus. Technol. 18
(1990) 652.
[10] M. Swartz, G. Verner, in: P.L. Hagelstein, S.R. Chubb (Eds.), Condensed
Matter Nuclear Science, Proc. of ICCF-10, World Scientific Publishing,
NJ, ISBN: 981-256-564-7, 2006, p. 29, (the author brings to the reader's
attention the data in Figs. 1–5 of this paper and does not endorse models
and explanations of the effect therein).
[11] V. Violante, M. Apicella, E. Castagna, L. Capobianco, L.D. Aulerio,
G. Mazzitelli, M. McKubre, F. Sarto, C. Sibilia, A. Rosada, E.
Santoro, F. Tanzella, in: A. Takahashi, K. Ota, Y. Iwamura (Eds.),
Proc. ICCF12 Yokohama (JP), World Scientific, 2006, p. 117.
[12] F.G. Will, K. Cedzynska, M.-C. Yang, J.R. Peterson, H.E. Bergeson, S.C.
Barrows, W.J. West, B.C. Linton, Proc. 2nd International Conference on
Cold Fusion, Como, Italy, Societa Italiana di Fisica, Bologna, Italy, 1990,
available at (www.lenr-canr.org/LibFrame1.html).
[13] Unpublished work. These laboratories are no longer working due to lack of
institutional and funding support in the case of the government laboratory,
and due to lack of funding in the case of the company. Their inclusion in
this list is subjective and is based on my personal viewing of their data and
operation of their experiments.
[14] M.C.H. McKubre, et al., 5th International Conference on Cold Fusion,
IMRA Europe, Sophia Antipolis Cedex, France, Monte-Carlo, Monaco,
1995.
[15] D. Letts, D. Cravens, in: P.L. Hagelstein, S.R. Chubb (Eds.), Condensed
Matter Nuclear Science, Proc. of ICCF-10, World Scientific Publishing,
NJ, ISBN: 981-256-564-7, 2006, p. 159.
8572 G.K. Hubler / Surface & Coatings Technology 201 (2007) 8568–8573

[16] A. De Ninno, V. Violante, A. LaBarbera, Phys. Rev. B56 (1997) 2417;
A. Adrover, M. Giona, L. Capobianco, P. Tripodi, V. Violante, J. Alloys &
Comp. 358 (2003) 268;
A. Adrover, M. Giona, L. Capobianco, P. Tripodi, V. Violante, J. Alloys &
Comp. 358 (2003) 157;
A. Adrover, M. Giona, L. Capobianco, P. Tripodi, V. Violante, J. Hydrogen
Energ. 28 (2003) 1279.
[17] A. De Ninno, V. Violante, A. LaBarbera, J. Alloys Compd. 253–254
(1997) 181.
[18] E.F. Skelton, P.L. Hagans, S.B. Qadri, D.D. Dominguez, A.C. Ehrlich, J.Z.
Hu, Phys. Rev., B 58 (1998) 14775.
[19] R. Felici, L. Bertalot, A. DeNinno, A. LaBarbera, V. Violante, Rev. Sci.
Instrum. 66 (1995) 3344.
[20] C.-A. Huh, Earth Planet. Sci. Lett. 171 (1999) 325.
[21] For an informative discussion of the Mössbauer Effect (ME) and Perturbed
Angular Correlations (PAC), see http://defects.physics.wsu.edu/.
[22] For a rigorous discussion of hyperfine interactions and experimental data,
see Hyperfine Interactions, eds. A.J. Freeman and R.B. Frankel,
(Academic, New York, 1967).
[23] D.I. Bolef, R.K. Sundfors, Nuclear Acoustic Resonance, Academic,
New York, 1993.
8573G.K. Hubler / Surface & Coatings Technology 201 (2007) 8568–8573