- Email: [email protected]
Recent results on mixed conductors containing hydrogen or deuterium
Solid State lonics 40/41 (1990) 519-524 North-Holland
RECENT RESULTS ON MIXED CONDUCTORS CONTAINING HYDROGEN OR DEUTERIUM Andreas BELZNER, Ulrike BISCHLER, Steven C R O U C H - B A K E R l, Turgut M. G U R , George LUCIER, Martha SCHREIBER and Robert A. H U G G I N S Department of Materials Science and Engineering, Stanford University. StanJbrd, CA 94305, USA
A number of experiments have been undertaken to determine whether there is a significant difference in thermal effects when deuterium, rather than hydrogen, is inserted into palladium by the use of electrochemicaltechniques involvingextremely negative potentials. They involved the measurement of heat inputs and outputs using careful calorimetric techniques, and employing experimental conditions for the two systems as comparable as possible in order to reduce the possibility of a major thermal influence due to chemical or metallurgical effects. These experiments have shown not only a different behavior in the case of the hydrogen-palladium system and the deuterium-palladium system, but also the generation of true excess power in the deuterium case. The magnitude of the effects observed, of the order of one to two Watts of excess power per gram of palladium, are comparable to those which were reported earlier by Fleischmann and Pons.
I. Introduction Martin F l e i s c h m a n n a n d Stanley Pons a n n o u n c e d on March 23, 1989 that they had performed electrochemical experiments involving the insertion of deuterium into palladium that produced excess heat generation, neutrons, g a m m a rays, and tritium [ 1 ], a n d claimed that these must be due to some previously unrecognized form of atomic fusion. If this is found to be correct, it might represent a very important breakthrough in the search for a source of energy that is widely available, relatively inexpensive, and not accompanied by severe radiation hazards or e n v i r o n m e n t a l pollution. The experiments reported in this paper were designed to provide a direct comparison of the thermal behavior during electrolysis of the d e u t e r i u m - p a l ladium system with that of the chemically and metallurgically similar h y d r o g e n - p a l l a d i u m system under comparable experimental conditions, by the careful use of c o n v e n t i o n a l calorimetric methods. Except for n e u t r o n and g a m m a ray monitors used Present address: Department of Materials Science and Engineering, University of California, Los Angeles, Los Angeles, CA 90024, USA. 0167-2738/90/$ 03.50 © Elsevier Science Publishers B.V. ( North-Holland )
for safety purposes, no radiation detection or reaction product m e a s u r e m e n t s were undertaken in this work.
2. Experimental aspects
Experiments were conducted using isoperibolic calorimetry. This is a two-compartment steady state power balance method in which heat is generated within one compartment, and is conducted through an intermediate thermally-conducting wall into the other c o m p a r t m e n t , which is m a i n t a i n e d at a fixed lower temperature. U n d e r steady slate conditions, a temperature distribution is established in which the temperature difference across the thermally-conducting wall between the two c o m p a r t m e n t s transports heat at a rate that just balances the power generated within the first compartment. Thus, m e a s u r e m e n t of the difference in the temperature of the two c o m p a r t m e n t s (T~-T2) provides information about the thermal power generated in the first c o m p a r t m e n t , P, he~m, that is passed out as heat flux through the thermally-conducting wall into the sec-
520
A. Belzner et al. / M i x e d conductors with hydrogen or deuterium
ond c o m p a r t m e n t . This can be expressed as T h e r m a l power generated = Ptherm = Kdr ( T~ -- 72 ) , ~1) where Kc~ris the c a l o r i m e t e r calibration constant. As will be shown later, the thermal power generated P~he~m is not equal to the applied electrical power Papp~ in the case o f a reaction involving a change in the enthalpy o f the system. Experiments were c o n d u c t e d in which the first c o m p a r t m e n t consisted o f an electrochemical cell that was operated at a sufficient voltage that electrolysis o f the D 2 0 (or H 2 0 ) took place. The electrochemical cells were submerged inside a large, well stirred cold water bath, which acted as the second c o m p a r t m e n t . The i n t e r m e d i a t e thermally-conducting wall thus was the glass container surrounding the electrochemical cell. Two closely c o m p a r a b l e electrochemical cells were employed, one containing a 0.1 M solution o f LiOH in light water ( H 2 0 ) , and the other a 0.1 M solution o f LiOD in D 2 0 as electrolytes. Each also contained a resistance heater for calibration purposes, and a protected t h e r m o c o u p l e for t e m p e r a t u r e measurements. There was also a thermoccruple to measure the t e m p e r a t u r e o f the surrounding water bath. Electrolysis voltages in the range 3 to 15 V were applied to the cells in both increasing and decreasing directions, and the currents varied between 10 and 1000 mA. Temperatures, as well as the cell currents were m o n i t o r e d for steady state conditions. The t e m p e r a t u r e differences between the cells and the water bath ranged up to about 10°C. The p a l l a d i u m was o b t a i n e d by arc melting pieces cut from a p a l l a d i u m crucible that c o n t a i n e d an appreciable a m o u n t o f hydrogen, and other possible impurities. In o r d e r to get rid o f such c o n t a m i n a n t s , the samples were re-melted at least ten times, using an arc-melting a p p a r a t u s with a water-cooled tungsten electrode and a water cooled copper hearth, which acted as the other electrode. The e n v i r o n m e n t was argon, and the p r o c e d u r e consisted o f first melting some t i t a n i u m sponge in o r d e r to getter species from the argon, such as oxygen, j u s t prior to melting the palladium. After melting the palladium, the argon a t m o s p h e r e was replaced, and the p r o c e d u r e repeated again. It was quite obvious from the change in the color o f the arc during repeated melting that
the i m p u r i t y content o f the p a l l a d i u m was gradually being changed. This process was repeated until remelting caused a clear, white arc color. The resulting palladium, in the form o f a distorted sphere, was then mechanically converted into coinshaped disks 2 - 4 m m thick and roughly 1 cm in diameter, with weights in the range of 1 to 2 g each. Fine gold wire was e m p l o y e d as the current collector. The anodes were m a d e from a p p r o x i m a t e l y 2 meters o f fine p l a t i n u m wire, and were coiled just inside the cell periphery in order to be as far away as possible from the p a l l a d i u m samples, which were attached by gold wire to the calibration heater in the center o f the cell. H y d r o g e n acts as a poison in a system o f this type, and therefore precautions were taken to prevent contact between the p a l l a d i u m and ( m o i s t ) air or water, both during assembly and operation o f the cells. The cell c o m p o n e n t s were stored, as well as assembled and loaded into the cells, inside a dry nitrogen-filled glove box. Gases were allowed to exhaust from the cell through a simple one-way bubbler system containing silicone oil. In o r d e r to be sure that there was sufficient stirring, both cells were m o u n t e d in a gyrotary water bath shaker. This p r o v i d e d additional mechanical motion, and therefore stirring, both within the cell and in the external water bath. Calibration m e a s u r e m e n t s were m a d e over the whole range o f electrolysis power level by a d d i n g to the electrolysis power three different values o f additional heater power for each data point. By measuring the change in ( T I - ~ ) as a function o f the additional electrical power supplied to the heater, one can d e t e r m i n e the calibration constant over a specified range o f total input power in the presence o f the convective stirring. Thus, every d a t a point that was o b t a i n e d was a c c o m p a n i e d by a direct calibration of the calorimeter. The characteristic relaxation time o f the calorimeters was o f the order o f 15 to 20 rain. If we introduce power to the cell electrically by passing c u r r e n t ]appl at voltage Eappl , the applied electrical power is all converted to thermal power Papvl = Pth . . . . = l~ppl Z R .
(2)
However, in an electrochemical cell passing current so that electrolysis is taking place, we can divide the externally applied voltage across the cell into sev-
A. Belzner et al. / Mixed conductors with hydrogen or deuterium
eral parts, a chemical term, and a purely resistive term Eappl = E° + Iappl ( Zc + Za + Ze ) ,
(3)
where E ° is the equilibrium t h e r m o d y n a m i c voltage due to any difference in the chemical potentials at the two electrodes, and Z¢, Za and Z¢ are the impedances at the cathode/electrolyte interface, the anode/electrolyte interface, and in the bulk o f the electrolyte, respectively. The products IapplZ¢ and Iapp~Za are often called overvoltages in the electrochemical literature. In experiments that involve electrolysis the values of E ° can be related to the standard Gibbs free energy change per mole AG O by E°= AG°/2F,
(4)
where F is the Faraday constant. Furthermore, the Gibbs free energies can be divided into their enthalpic and entropic components: AGO= AH ° - TAS ° ,
(5)
where AG °, AH ° and AS ° are all positive. We can thus write Papp, = lapp, ( A H ° - T A S ° ) / 2 F + IEppl(Zc + Za + Ze ) . (6) The enthalpy change can be converted to an associated voltage, the thermoneutral voltage E,n, which is the voltage applied across the electrochemical cell that neither generates nor consumes heat as a result of the chemical reaction taking place. A l l ° = 2FE,~ .
(7)
Thus we can rewrite eq. (6) as
Pappl= IapplE,, -- I~ppl( T A S ° ) / 2 F +
I 2ppl( Z¢ + Za +
Z~).
(8) From elementary thermodynamics, one can relate the enthalpy change of a system A H to the difference between electrical work done on it and the heat evolved. Thus in the presence o f electrolysis, the thermal power generated will be less than the electrical power introduced. This can be expressed as Ptherm = Pappl --
IapplEt..
(9 )
521
The values of AH ° for such reactions are essentially independent of temperature, and are reported to be + 2 8 6 k J / m o l for the electrolysis o f H20, and + 297 k J / m o l for the electrolysis of D20. Thus the values o f Era are 1.48 V and 1.54 V, respectively. We see that the positive values of AH ° mean that a cell in which such reactions are occurring should be generating less heat than an identical cell containing only resistive components by an amount corresponding to the value o f Iapp~Et, as a result of entropic cooling. If, in addition, there were some other phenomenon present in the cell that produced thermal power, Pint, it would add an additional positive term, giving TAS°)/2F +12pp,(Zc+ Z , + Zc) + P, nt
Ptherm = - - l a p p l (
(10) and therefore an increase in (T~-T2) over that for the case of only endothermic electrolysis and exothermic resistive behavior for any level of Pappt. In addition to the resistive components in eq. (10), other exothermic phenomena can also be present related to chemical phenomenon taking place within the cell, and contribute positively to Pro,. One of these is the heat of solution when hydrogen or deuterium is inserted into the palladium, and the enthalpy change related to the ct-[3 phase transformation. The magnitudes o f these effects are relatively small, 9.55 kcal/mole for the ct-13 transformation in the H - P d case, and 8.55 kcal/mole in the D - P d case [2] and only contribute to the observed behavior when these particular processes are under way. As the insertion into the interior is relatively slow in the time span of a particular measurement these produce only relatively small effects in comparison to the large thermal effects of the electrolysis phenomenon itself. In addition, the rate o f the insertion reaction decreases with time (approximately proportional to t-~/2) and so this effect becomes less and less important the longer the cell has been in operation. Any recombination o f the hydrogen (or deuterium) and oxygen gases that are formed by the electrolysis reaction, either within the electrolyte or in the adjacent gas phase, would contribute exothermally, but no exothermic effects related to the reversal of the primary electrolysis cell reaction can be larger than the endothermic effects o f the reaction
522
A. Belzner et al. / M i x e d conductors with hydrogen or deuterium
itself. Therefore, one cannot get more heat from rec o m b i n a t i o n than that which was a b s o r b e d by the electrolysis reaction. M e a s u r e m e n t s in heavy water cells, were m a d e i m m e d i a t e l y upon the i n t r o d u c t i o n o f fresh pallad i u m cathodes into the cell, so that there would be no influence related to the presence of appreciable a m o u n t s o f d e u t e r i u m in the p a l l a d i u m . It was found that the m e a s u r e d e n d o t h e r m i c effects related to electrolysis were within a few percent o f calculated values. These results showed that the a m o u n t o f rec o m b i n a t i o n , or any other exothermic p h e n o m e n o n present, must be very small in c o m p a r i s o n with the m a j o r processes that are occurring.
14
12
10
8
.s
3. Experimental results The relation between the thermal power out through the c a l o r i m e t e r system and the applied electrical power was d e t e r m i n e d for cells containing light water. The o u t p u t thermal power during electrolysis was found to be lower than that when the electrical power was put into the calibration heater due to the entropic cooling. M e a s u r e m e n t s taken at different times showed that there is essentially no time dependence to this behavior. On the other hand, in e x p e r i m e n t s with heavy water cells, the thermal b e h a v i o r u n d e r electrolysis conditions was found to vary with time. This is illustrated in fig. 1, which shows the t e m p e r a t u r e difference between the two c o m p a r t m e n t s o f the calorimeter system with the heavy water cell as a function o f a p p l i e d electrical power for three situations. The d a t a on the m i d d l e line are those o b t a i n e d when the electrical power was a p p l i e d only to the calibration heater. The d a t a on the lower line are those which were o b t a i n e d initially when a fresh piece o f pallad i u m not containing d e u t e r i u m was inserted into the cell and the electrolysis reaction was m a d e to occur. One sees that the e n d o t h e r m i c b e h a v i o r related to electrolysis was initially present in this case, as with the light water cell m e n t i o n e d earlier. However, over a p e r i o d o f t i m e the t e m p e r a t u r e in the heavy water electrolysis cell rose relative to its initial value, in contrast to the light water case. The d a t a on the u p p e r line were o b t a i n e d after electrolysis had been under way for 66 h. This indicates that
2
0
~ 0
2
4
6
8
10
12
14
Pappl (W) Fig. 1. Variation of the relation between the thermal power output P~h~,~ and the applied electrical power Papp~of a D - P d cell with time. The data on the middle line were from the heater calibrations, those on the lower line were obtained shortly after the introduction of a fresh sample of palladium, and those on the top line were measured 66 h later.
there was some process taking place, generating heat, and giving Pl.t a positive value. In this case, at any applied electrical power level, the thermal power out had increased b e y o n d that observed when electrical power was a p p l i e d to the resistive calibration heater. Thus under these conditions Pint was greater than IapptEtn, and the electrochemical cell was producing excess power Pexc above a n d beyond that which was being input electrically, Papp~, where Pexc = P~herm -- Pappl -- Pin, -
IapplEm -
(11)
Thus the cell in which an appreciable a m o u n t o f d e u t e r i u m had diffused into the p a l l a d i u m had an a d d i t i o n a l exothermic p h e n o m e n o n present produc-
A. Belzner et al. / M i x e d conductors with hydrogen or deuterium 1.2'
523
dence to the behavior of the light water cell, in which hydrogen is inserted into the palladium. In the heavy water cells, with deuterium insertion into the palladium, there was always a time-dependent generation of additional power Pint above that which is due to the exothermic Joule heating and the endothermic electrolysis process. Over time, the values of thermal power out of the cell come to exceed the electrical power input, so that the value of Pexc becomes positive.
1.1
Q.
£ k
1.0
[]
n
4. Discussion rn
[] 0.9
0.8
•
,
20
•
i
40
.
,
.
,
60 T i m e
80
•
,
100
•
120
(Hr)
Fig. 2. Longer term time dependence of the ratio of thermal power output Pthermto electrical power input P.pp~ for two different D Pd cells.
ing Pint that was not observed in the case of the light water cell where hydrogen was inserted into the palladium. Measurements have been taken of the ratio of power out, Pthe~m, to the power supplied electrically, Papp~, at various values of time after the insertion of a fresh palladium electrode. Ptherm/Pappl = ( Pappl "~-Pexc)/Pappl
•
(
12 )
Two sets of such data are shown in fig. 2. At the beginning, one sees the electrolysis cooling effect, but this is gradually overcome with time by an additional phenomenon that produces Pint, so that after about 60 h we see that Pexcbecomes greater than zero, and the thermal power evolved Pth~m is greater than the electrical power applied PapplFour sets of light water/heavy water comparisons were run. In every case, there was no time depen-
A direct comparison has been made, using careful calorimetric techniques, of the behavior of the light water ( H20 )-palladium system and the heavy water (D20)-palladium system under conditions in which electrolysis was taking place at high rates. It was found that these two systems behaved quite differently. In the light water case, the electrochemical cell temperature was lower than that expected from Joule heating alone, because of the endothermic electrolysis reaction taking place. This behavior was found to be independent of time. On the other hand, electrolysis of the heavy water system showed different behavior. Initially, such electrochemical cells behaved in a manner directly analogous to the light water cells, showing the combination of Joule heating and electrolysis cooling. However, with time, the cell thermal power output gradually increased, and after a number of days reached values greater than the input electrical power. These experiments thus demonstrate not only a different behavior in the case of the hydrogen-palladium system and the deuterium-palladium system, but also the generation of true excess power in the deuterium case. They are consistent with the claim of Fleischmann and Pons [ 1 ] of the observation of excess power generation in the deuteriumpalladium system under electrolytic conditions. The magnitude of the effects observed here, of the order of one to two Watts of excess power per gram of palladium, is comparable to those which they reported. Our calorimetric measurements give no information about the mechanism causing the internal generation of this excess power. However, the observations which have come from several different
524
A. Belzner et al. / M i x e d conductors with hydrogen or deuterium
laboratories o f the generation o f t r i t i u m in such cells are especially interesting, a n d deserve further consideration. If, indeed, tritium is an i m p o r t a n t product o f the reaction occurring in the electrochemical systems that are p r o d u c i n g Pint, it is an indication that some type o f nuclear process must be taking place.
experiments is gratefully acknowledged. In addition, this work was partially s u p p o r t e d by the U.S. Department o f Energy under Subcontracts LBL 4556010 and B N L 274318-S.
Acknowledgement
[l]M. Fleischmann and S. Pons, J. Electroanal. Chem. 261 (1989)301. [2] A. Maeland and T.B. Hanagan, J. Phys. Chem. 68 (1964) 1419.
The assistance o f Dr. P r o d y o t Roy o f the General Electric Co. in acquiring the p a l l a d i u m used in these
References
RECENT RESULTS ON MIXED CONDUCTORS CONTAINING HYDROGEN OR DEUTERIUM Andreas BELZNER, Ulrike BISCHLER, Steven C R O U C H - B A K E R l, Turgut M. G U R , George LUCIER, Martha SCHREIBER and Robert A. H U G G I N S Department of Materials Science and Engineering, Stanford University. StanJbrd, CA 94305, USA
A number of experiments have been undertaken to determine whether there is a significant difference in thermal effects when deuterium, rather than hydrogen, is inserted into palladium by the use of electrochemicaltechniques involvingextremely negative potentials. They involved the measurement of heat inputs and outputs using careful calorimetric techniques, and employing experimental conditions for the two systems as comparable as possible in order to reduce the possibility of a major thermal influence due to chemical or metallurgical effects. These experiments have shown not only a different behavior in the case of the hydrogen-palladium system and the deuterium-palladium system, but also the generation of true excess power in the deuterium case. The magnitude of the effects observed, of the order of one to two Watts of excess power per gram of palladium, are comparable to those which were reported earlier by Fleischmann and Pons.
I. Introduction Martin F l e i s c h m a n n a n d Stanley Pons a n n o u n c e d on March 23, 1989 that they had performed electrochemical experiments involving the insertion of deuterium into palladium that produced excess heat generation, neutrons, g a m m a rays, and tritium [ 1 ], a n d claimed that these must be due to some previously unrecognized form of atomic fusion. If this is found to be correct, it might represent a very important breakthrough in the search for a source of energy that is widely available, relatively inexpensive, and not accompanied by severe radiation hazards or e n v i r o n m e n t a l pollution. The experiments reported in this paper were designed to provide a direct comparison of the thermal behavior during electrolysis of the d e u t e r i u m - p a l ladium system with that of the chemically and metallurgically similar h y d r o g e n - p a l l a d i u m system under comparable experimental conditions, by the careful use of c o n v e n t i o n a l calorimetric methods. Except for n e u t r o n and g a m m a ray monitors used Present address: Department of Materials Science and Engineering, University of California, Los Angeles, Los Angeles, CA 90024, USA. 0167-2738/90/$ 03.50 © Elsevier Science Publishers B.V. ( North-Holland )
for safety purposes, no radiation detection or reaction product m e a s u r e m e n t s were undertaken in this work.
2. Experimental aspects
Experiments were conducted using isoperibolic calorimetry. This is a two-compartment steady state power balance method in which heat is generated within one compartment, and is conducted through an intermediate thermally-conducting wall into the other c o m p a r t m e n t , which is m a i n t a i n e d at a fixed lower temperature. U n d e r steady slate conditions, a temperature distribution is established in which the temperature difference across the thermally-conducting wall between the two c o m p a r t m e n t s transports heat at a rate that just balances the power generated within the first compartment. Thus, m e a s u r e m e n t of the difference in the temperature of the two c o m p a r t m e n t s (T~-T2) provides information about the thermal power generated in the first c o m p a r t m e n t , P, he~m, that is passed out as heat flux through the thermally-conducting wall into the sec-
520
A. Belzner et al. / M i x e d conductors with hydrogen or deuterium
ond c o m p a r t m e n t . This can be expressed as T h e r m a l power generated = Ptherm = Kdr ( T~ -- 72 ) , ~1) where Kc~ris the c a l o r i m e t e r calibration constant. As will be shown later, the thermal power generated P~he~m is not equal to the applied electrical power Papp~ in the case o f a reaction involving a change in the enthalpy o f the system. Experiments were c o n d u c t e d in which the first c o m p a r t m e n t consisted o f an electrochemical cell that was operated at a sufficient voltage that electrolysis o f the D 2 0 (or H 2 0 ) took place. The electrochemical cells were submerged inside a large, well stirred cold water bath, which acted as the second c o m p a r t m e n t . The i n t e r m e d i a t e thermally-conducting wall thus was the glass container surrounding the electrochemical cell. Two closely c o m p a r a b l e electrochemical cells were employed, one containing a 0.1 M solution o f LiOH in light water ( H 2 0 ) , and the other a 0.1 M solution o f LiOD in D 2 0 as electrolytes. Each also contained a resistance heater for calibration purposes, and a protected t h e r m o c o u p l e for t e m p e r a t u r e measurements. There was also a thermoccruple to measure the t e m p e r a t u r e o f the surrounding water bath. Electrolysis voltages in the range 3 to 15 V were applied to the cells in both increasing and decreasing directions, and the currents varied between 10 and 1000 mA. Temperatures, as well as the cell currents were m o n i t o r e d for steady state conditions. The t e m p e r a t u r e differences between the cells and the water bath ranged up to about 10°C. The p a l l a d i u m was o b t a i n e d by arc melting pieces cut from a p a l l a d i u m crucible that c o n t a i n e d an appreciable a m o u n t o f hydrogen, and other possible impurities. In o r d e r to get rid o f such c o n t a m i n a n t s , the samples were re-melted at least ten times, using an arc-melting a p p a r a t u s with a water-cooled tungsten electrode and a water cooled copper hearth, which acted as the other electrode. The e n v i r o n m e n t was argon, and the p r o c e d u r e consisted o f first melting some t i t a n i u m sponge in o r d e r to getter species from the argon, such as oxygen, j u s t prior to melting the palladium. After melting the palladium, the argon a t m o s p h e r e was replaced, and the p r o c e d u r e repeated again. It was quite obvious from the change in the color o f the arc during repeated melting that
the i m p u r i t y content o f the p a l l a d i u m was gradually being changed. This process was repeated until remelting caused a clear, white arc color. The resulting palladium, in the form o f a distorted sphere, was then mechanically converted into coinshaped disks 2 - 4 m m thick and roughly 1 cm in diameter, with weights in the range of 1 to 2 g each. Fine gold wire was e m p l o y e d as the current collector. The anodes were m a d e from a p p r o x i m a t e l y 2 meters o f fine p l a t i n u m wire, and were coiled just inside the cell periphery in order to be as far away as possible from the p a l l a d i u m samples, which were attached by gold wire to the calibration heater in the center o f the cell. H y d r o g e n acts as a poison in a system o f this type, and therefore precautions were taken to prevent contact between the p a l l a d i u m and ( m o i s t ) air or water, both during assembly and operation o f the cells. The cell c o m p o n e n t s were stored, as well as assembled and loaded into the cells, inside a dry nitrogen-filled glove box. Gases were allowed to exhaust from the cell through a simple one-way bubbler system containing silicone oil. In o r d e r to be sure that there was sufficient stirring, both cells were m o u n t e d in a gyrotary water bath shaker. This p r o v i d e d additional mechanical motion, and therefore stirring, both within the cell and in the external water bath. Calibration m e a s u r e m e n t s were m a d e over the whole range o f electrolysis power level by a d d i n g to the electrolysis power three different values o f additional heater power for each data point. By measuring the change in ( T I - ~ ) as a function o f the additional electrical power supplied to the heater, one can d e t e r m i n e the calibration constant over a specified range o f total input power in the presence o f the convective stirring. Thus, every d a t a point that was o b t a i n e d was a c c o m p a n i e d by a direct calibration of the calorimeter. The characteristic relaxation time o f the calorimeters was o f the order o f 15 to 20 rain. If we introduce power to the cell electrically by passing c u r r e n t ]appl at voltage Eappl , the applied electrical power is all converted to thermal power Papvl = Pth . . . . = l~ppl Z R .
(2)
However, in an electrochemical cell passing current so that electrolysis is taking place, we can divide the externally applied voltage across the cell into sev-
A. Belzner et al. / Mixed conductors with hydrogen or deuterium
eral parts, a chemical term, and a purely resistive term Eappl = E° + Iappl ( Zc + Za + Ze ) ,
(3)
where E ° is the equilibrium t h e r m o d y n a m i c voltage due to any difference in the chemical potentials at the two electrodes, and Z¢, Za and Z¢ are the impedances at the cathode/electrolyte interface, the anode/electrolyte interface, and in the bulk o f the electrolyte, respectively. The products IapplZ¢ and Iapp~Za are often called overvoltages in the electrochemical literature. In experiments that involve electrolysis the values of E ° can be related to the standard Gibbs free energy change per mole AG O by E°= AG°/2F,
(4)
where F is the Faraday constant. Furthermore, the Gibbs free energies can be divided into their enthalpic and entropic components: AGO= AH ° - TAS ° ,
(5)
where AG °, AH ° and AS ° are all positive. We can thus write Papp, = lapp, ( A H ° - T A S ° ) / 2 F + IEppl(Zc + Za + Ze ) . (6) The enthalpy change can be converted to an associated voltage, the thermoneutral voltage E,n, which is the voltage applied across the electrochemical cell that neither generates nor consumes heat as a result of the chemical reaction taking place. A l l ° = 2FE,~ .
(7)
Thus we can rewrite eq. (6) as
Pappl= IapplE,, -- I~ppl( T A S ° ) / 2 F +
I 2ppl( Z¢ + Za +
Z~).
(8) From elementary thermodynamics, one can relate the enthalpy change of a system A H to the difference between electrical work done on it and the heat evolved. Thus in the presence o f electrolysis, the thermal power generated will be less than the electrical power introduced. This can be expressed as Ptherm = Pappl --
IapplEt..
(9 )
521
The values of AH ° for such reactions are essentially independent of temperature, and are reported to be + 2 8 6 k J / m o l for the electrolysis o f H20, and + 297 k J / m o l for the electrolysis of D20. Thus the values o f Era are 1.48 V and 1.54 V, respectively. We see that the positive values of AH ° mean that a cell in which such reactions are occurring should be generating less heat than an identical cell containing only resistive components by an amount corresponding to the value o f Iapp~Et, as a result of entropic cooling. If, in addition, there were some other phenomenon present in the cell that produced thermal power, Pint, it would add an additional positive term, giving TAS°)/2F +12pp,(Zc+ Z , + Zc) + P, nt
Ptherm = - - l a p p l (
(10) and therefore an increase in (T~-T2) over that for the case of only endothermic electrolysis and exothermic resistive behavior for any level of Pappt. In addition to the resistive components in eq. (10), other exothermic phenomena can also be present related to chemical phenomenon taking place within the cell, and contribute positively to Pro,. One of these is the heat of solution when hydrogen or deuterium is inserted into the palladium, and the enthalpy change related to the ct-[3 phase transformation. The magnitudes o f these effects are relatively small, 9.55 kcal/mole for the ct-13 transformation in the H - P d case, and 8.55 kcal/mole in the D - P d case [2] and only contribute to the observed behavior when these particular processes are under way. As the insertion into the interior is relatively slow in the time span of a particular measurement these produce only relatively small effects in comparison to the large thermal effects of the electrolysis phenomenon itself. In addition, the rate o f the insertion reaction decreases with time (approximately proportional to t-~/2) and so this effect becomes less and less important the longer the cell has been in operation. Any recombination o f the hydrogen (or deuterium) and oxygen gases that are formed by the electrolysis reaction, either within the electrolyte or in the adjacent gas phase, would contribute exothermally, but no exothermic effects related to the reversal of the primary electrolysis cell reaction can be larger than the endothermic effects o f the reaction
522
A. Belzner et al. / M i x e d conductors with hydrogen or deuterium
itself. Therefore, one cannot get more heat from rec o m b i n a t i o n than that which was a b s o r b e d by the electrolysis reaction. M e a s u r e m e n t s in heavy water cells, were m a d e i m m e d i a t e l y upon the i n t r o d u c t i o n o f fresh pallad i u m cathodes into the cell, so that there would be no influence related to the presence of appreciable a m o u n t s o f d e u t e r i u m in the p a l l a d i u m . It was found that the m e a s u r e d e n d o t h e r m i c effects related to electrolysis were within a few percent o f calculated values. These results showed that the a m o u n t o f rec o m b i n a t i o n , or any other exothermic p h e n o m e n o n present, must be very small in c o m p a r i s o n with the m a j o r processes that are occurring.
14
12
10
8
.s
3. Experimental results The relation between the thermal power out through the c a l o r i m e t e r system and the applied electrical power was d e t e r m i n e d for cells containing light water. The o u t p u t thermal power during electrolysis was found to be lower than that when the electrical power was put into the calibration heater due to the entropic cooling. M e a s u r e m e n t s taken at different times showed that there is essentially no time dependence to this behavior. On the other hand, in e x p e r i m e n t s with heavy water cells, the thermal b e h a v i o r u n d e r electrolysis conditions was found to vary with time. This is illustrated in fig. 1, which shows the t e m p e r a t u r e difference between the two c o m p a r t m e n t s o f the calorimeter system with the heavy water cell as a function o f a p p l i e d electrical power for three situations. The d a t a on the m i d d l e line are those o b t a i n e d when the electrical power was a p p l i e d only to the calibration heater. The d a t a on the lower line are those which were o b t a i n e d initially when a fresh piece o f pallad i u m not containing d e u t e r i u m was inserted into the cell and the electrolysis reaction was m a d e to occur. One sees that the e n d o t h e r m i c b e h a v i o r related to electrolysis was initially present in this case, as with the light water cell m e n t i o n e d earlier. However, over a p e r i o d o f t i m e the t e m p e r a t u r e in the heavy water electrolysis cell rose relative to its initial value, in contrast to the light water case. The d a t a on the u p p e r line were o b t a i n e d after electrolysis had been under way for 66 h. This indicates that
2
0
~ 0
2
4
6
8
10
12
14
Pappl (W) Fig. 1. Variation of the relation between the thermal power output P~h~,~ and the applied electrical power Papp~of a D - P d cell with time. The data on the middle line were from the heater calibrations, those on the lower line were obtained shortly after the introduction of a fresh sample of palladium, and those on the top line were measured 66 h later.
there was some process taking place, generating heat, and giving Pl.t a positive value. In this case, at any applied electrical power level, the thermal power out had increased b e y o n d that observed when electrical power was a p p l i e d to the resistive calibration heater. Thus under these conditions Pint was greater than IapptEtn, and the electrochemical cell was producing excess power Pexc above a n d beyond that which was being input electrically, Papp~, where Pexc = P~herm -- Pappl -- Pin, -
IapplEm -
(11)
Thus the cell in which an appreciable a m o u n t o f d e u t e r i u m had diffused into the p a l l a d i u m had an a d d i t i o n a l exothermic p h e n o m e n o n present produc-
A. Belzner et al. / M i x e d conductors with hydrogen or deuterium 1.2'
523
dence to the behavior of the light water cell, in which hydrogen is inserted into the palladium. In the heavy water cells, with deuterium insertion into the palladium, there was always a time-dependent generation of additional power Pint above that which is due to the exothermic Joule heating and the endothermic electrolysis process. Over time, the values of thermal power out of the cell come to exceed the electrical power input, so that the value of Pexc becomes positive.
1.1
Q.
£ k
1.0
[]
n
4. Discussion rn
[] 0.9
0.8
•
,
20
•
i
40
.
,
.
,
60 T i m e
80
•
,
100
•
120
(Hr)
Fig. 2. Longer term time dependence of the ratio of thermal power output Pthermto electrical power input P.pp~ for two different D Pd cells.
ing Pint that was not observed in the case of the light water cell where hydrogen was inserted into the palladium. Measurements have been taken of the ratio of power out, Pthe~m, to the power supplied electrically, Papp~, at various values of time after the insertion of a fresh palladium electrode. Ptherm/Pappl = ( Pappl "~-Pexc)/Pappl
•
(
12 )
Two sets of such data are shown in fig. 2. At the beginning, one sees the electrolysis cooling effect, but this is gradually overcome with time by an additional phenomenon that produces Pint, so that after about 60 h we see that Pexcbecomes greater than zero, and the thermal power evolved Pth~m is greater than the electrical power applied PapplFour sets of light water/heavy water comparisons were run. In every case, there was no time depen-
A direct comparison has been made, using careful calorimetric techniques, of the behavior of the light water ( H20 )-palladium system and the heavy water (D20)-palladium system under conditions in which electrolysis was taking place at high rates. It was found that these two systems behaved quite differently. In the light water case, the electrochemical cell temperature was lower than that expected from Joule heating alone, because of the endothermic electrolysis reaction taking place. This behavior was found to be independent of time. On the other hand, electrolysis of the heavy water system showed different behavior. Initially, such electrochemical cells behaved in a manner directly analogous to the light water cells, showing the combination of Joule heating and electrolysis cooling. However, with time, the cell thermal power output gradually increased, and after a number of days reached values greater than the input electrical power. These experiments thus demonstrate not only a different behavior in the case of the hydrogen-palladium system and the deuterium-palladium system, but also the generation of true excess power in the deuterium case. They are consistent with the claim of Fleischmann and Pons [ 1 ] of the observation of excess power generation in the deuteriumpalladium system under electrolytic conditions. The magnitude of the effects observed here, of the order of one to two Watts of excess power per gram of palladium, is comparable to those which they reported. Our calorimetric measurements give no information about the mechanism causing the internal generation of this excess power. However, the observations which have come from several different
524
A. Belzner et al. / M i x e d conductors with hydrogen or deuterium
laboratories o f the generation o f t r i t i u m in such cells are especially interesting, a n d deserve further consideration. If, indeed, tritium is an i m p o r t a n t product o f the reaction occurring in the electrochemical systems that are p r o d u c i n g Pint, it is an indication that some type o f nuclear process must be taking place.
experiments is gratefully acknowledged. In addition, this work was partially s u p p o r t e d by the U.S. Department o f Energy under Subcontracts LBL 4556010 and B N L 274318-S.
Acknowledgement
[l]M. Fleischmann and S. Pons, J. Electroanal. Chem. 261 (1989)301. [2] A. Maeland and T.B. Hanagan, J. Phys. Chem. 68 (1964) 1419.
The assistance o f Dr. P r o d y o t Roy o f the General Electric Co. in acquiring the p a l l a d i u m used in these
References