APPLIED PHYSICS LETTERS 89, 051918 共2006兲
H2 cold plasma on Pd/ H system at low hydrogen pressure A. Baldi,a兲 F. Di Pascasio, and D. Gozzib兲 RIMLab, Dipartimento di Chimica, Università degli Studi di Roma La Sapienza, P.le A. Moro 5, 00185 Rome, Italy
共Received 3 March 2006; accepted 20 June 2006; published online 4 August 2006兲 The effect of hydrogen cold plasmas, in negative corona regime, was investigated on the palladium-hydrogen system in the ␣ +  region at room temperature and H2 pressures below 0.037 bar. Three independent experimental quantities have been simultaneously measured: hydrogen concentration, electrical resistance, and elongation of the samples. An increase in the above quantities was always observed when plasma was lighted. The ability of an external electric field 共the corona discharge兲 to perturb the thermodynamic state of the Pd/ H system was experimentally demonstrated. © 2006 American Institute of Physics. 关DOI: 10.1063/1.2267181兴 A hydrogen economy has been proposed as a secure and clean solution to the dramatic energy consumption increase expected in the next decade.1 Hydrogen storage for transportation is one of the major challenges in building a hydrogen economy, which requires the realization of high capacity storage systems.2 This letter deals with the experimental procedures designed to increase the hydrogen storage capacity of metals or alloys. These procedures perturb the thermodynamic equilibrium between hydrogen gas and hydrogen dissolved in the metal. In this work, the hydrogen absorbing metal is one of the electrodes exposed to the hydrogen cold plasma. Palladium was chosen among other metals since the Pd/ H system is one of the most investigated metal/hydrogen systems and it has been frequently considered as a model system to understand the behavior of hydrogen in metals. In a cold plasma the ion concentration is low 共⬃1011 cm−3兲 and the temperature of ions and neutral species is approximately the environment temperature, while electrons have a much higher temperature 共⬃1 eV兲. Due to the low concentration of electrons, the heat transfer is controlled by the dominant neutral species and the surfaces in the plasma are at the temperature of the neutral species. The experiments are performed in the corona region of the current versus voltage discharge curve of H2. The corona discharge is a self-sustained electric discharge in a gas, in which the primary ionization processes are confined to a region surrounding the electrode where E is maximum.3–7 The ability of the corona to perturb the thermodynamic state of bulk hydrogen in the  phase of the Pd/ H was recently shown.8 The results indicate that the ultimate effect of the cold plasma is to raise the hydrogen concentration in the system. The presented work investigates the effect of negative corona on the Pd/ H system in the ␣ +  region at room temperature. The recorded data are consistent with the results previously found8 and the amplitude of the observed effect is greater. Details on the experimental apparatus and experimental procedure are found elsewhere.8 Samples are 99.99% pure Pd wires 0.5 mm in diameter and 15 cm long. Negative corona is performed, in cylindric coaxial configuration, lowering the electric potential of the sample wire 共SW兲 with a兲
Present address: Condensed Matter Physics, Department of Physics and Astronomy, Faculty of Sciences, Vrije Universiteit De Boelelaan 1081, 1081 HV Amsterdam, The Netherlands. b兲 Electronic mail:
[email protected]; URL: http:// www.chem.uniroma1.it/⬃gozzi/
respect to an aluminum counterelectrode 共CE兲 of 10 mm in inner diameter and 5 cm in length. The CE covers one-third of the SW length. Plasma current is set to ⬃110 A and the voltage difference between SW and CE is ⬃0.5 kV, depending on the pressure values. The pressure range explored is between 0.012 and 0.037 bar at 299 K. A preliminary approach to characterize the plasma was carried out by a Langmuir probe9,10 orthogonally positioned with respect to the SW. The i共V兲 characteristic was measured between −100 and +100 V at 0.08 bar under a discharge current of 200 A by a platinum wire probe 0.1 mm in diameter and 1.3 mm long. Under positive polarization, the probe current quite soon reaches an intensity comparable to that of the discharge current, so that the perturbation of the plasma prevents the possibility of extracting any information from the positive tract of the i共V兲 curve. For small negative polarization the probe current shows an exponential trend, while for large negative polarization no saturation is observed due to the cylindrical shape of the probe.11 An approximate method12 was used to estimate the electronic temperature Te: Te = −
V p e ⬃ 10 eV, k ln共i − i0兲
共1兲
where V p is the probe voltage, i0 ⬃ i共1 − 冑1 − 2⌬兲, and ⌬ is determined from a linear fit of the ln共i兲 vs V p curve. From this value, the plasma density ne was calculated:13 ne ⬃
冉 冊
i共V p兲 2me · eS kTe
1/2
⬃ 2 ⫻ 1011 cm−3 ,
共2兲
with S being the probe surface area. In the following the atomic hydrogen concentration in the SW will be expressed as atomic ratio x = nH / nPd—the ratio between the number of H atoms and Pd atoms in the SW; x is determined volumetrically at constant temperature T through the H2 pressure change ⌬p in the cell due to the gas absorption by the SW. Using the law of perfect gas, its value is obtained as follows: x=
nH 2M SWvcell · 兩⌬p兩 = const · 兩⌬p兩, = nPd mSWRT
共3兲
where M SW, mSW, vcell, and R are, respectively, the atomic weight and mass of the SW, the free volume of the cell, and the ideal gas constant. The electrical resistances of the whole SW and its two halves—of which only one is affected by
0003-6951/2006/89共5兲/051918/3/$23.00 89, 051918-1 © 2006 American Institute of Physics Downloaded 29 Aug 2006 to 130.37.34.203. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
051918-2
Baldi, Di Pascasio, and Gozzi
Appl. Phys. Lett. 89, 051918 共2006兲
FIG. 2. Comparison between R / R0 data measured in this work 共circles and squares兲 for the ␣ +  region and data collected 共Ref. 8兲 共triangles兲 for the  phase at 299 K.
the same batch. Each one was oxidized in air at 600 ° C for 10 min before being exposed to hydrogen in order to speed up the absorption kinetics.16 A tensile load ⑀ = 7.8 MPa was applied to each SW. The procedure adopted for assembling the SW in the apparatus and to clean the system has already been reported.8 The samples were exposed to a given pressure of H2 at 299 K until a steady state was reached 共⬃20 h兲. Thereafter, negative corona was lighted until a different steady state was again reached 共⬃30 h兲. In all the cases, the plasma discharge produces a decrease in the pressure due to the absorption of H2 and a time-related increase in all three observed variables. Scanning electron microscopy analysis was performed ex situ on each SW at the end of the experiment. Dendritic morphologies similar to those previously found8 were not observed in the present work. They were observed only on SWs exposed to plasma for hundreds of hours. In Fig. 1 the steady-state values of x, R / R0, and ⌬l / l0 collected prior to the plasma lighting 共E = 0兲 and after 共E ⫽ 0兲 are compared; it is evident that, at any given pressure, the values of all the recorded variables are always higher under plasma. Therefore, the effect of the plasma is to raise the hydrogen concentration in the SW beyond the value expected in the same experimental conditions but without plasma. Regarding the R / R0 values, it must FIG. 1. Steady-state values of x 共a兲, R / R0 共b兲, and ⌬l / l0 共c兲 at 299 K. 共쎲兲 Data measured without plasma; 共䊐兲 data measured under plasma.
theplasma—are obtained by four-probe measurements. A couple of 0.35 mm diameter Pt wires are crimped to each SW end and another Pt wire is crimped at its middle. Three of these contacts are sliding contacts which move as the SW length changes. The electrical resistance values will be expressed as relative resistance R / R0—the ratio between the actual resistance value of SW and the value when SW is free from H, namely, at x = 0. The SW elongation ⌬l, produced by absorption of hydrogen, is measured by an inductive displacement transducer 共Penny and Giles, model 1354, UK兲. The head of the transducer follows the motion of one end of SW, while the other end of SW is fixed. The elongation will be expressed in terms of relative elongation ⌬l / l0, where the SW length l0 at x = 0 is measured prior to its assembly in the system. It must be underlined that the three observed variables are strictly physically correlated14,15 but each one is measured independently from the other ones. The data were collected on six different SWs cut from
FIG. 3. Changes in 共R / R0兲in, relative resistance of the portion of SW in plasma, and 共R / R0兲out, relative resistance of the portion of SW out of plasma, measured after the plasma lighting in a typical run. The gas temperature change measured outside the CE is shown in the insert. Downloaded 29 Aug 2006 to 130.37.34.203. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp
051918-3
Appl. Phys. Lett. 89, 051918 共2006兲
Baldi, Di Pascasio, and Gozzi
be recalled that an increase in the electrical resistance indicates an increase of the hydrogen concentration, since a strict linear correlation between the two variables is known to exist at constant temperature in the x range explored here.17–19 The comparison between R / R0 values collected in the ␣ +  region in this work and R / R0 values found in the  phase8 is shown in Fig. 2; it is clear that the effect of the plasma in increasing the H concentration in SW is much more intense in the region where both phases coexist. For instance, in the  phase at ln共PH2 / Pbar兲 = −3.8, the change
冏
⌬共R/R0兲 R/R0
冏
= 
共R/R0兲E⫽0 − 共R/R0兲E=0 ⬃ 0.008 共R/R0兲E=0
共4兲
was found between the conditions E ⫽ 0 and E = 0, see Fig. 2. In the ␣ +  region at the same pressure value—panel 共b兲 of Fig. 1—a wider change was found
冏
⌬共R/R0兲 ⌬R/R0
冏
= ␣+
共R/R0兲E⫽0 − 共R/R0兲E=0 ⬃ 0.43. 共R/R0兲E=0
共5兲
The change in x observed in the ␣ +  region at the same pressure value—panel 共a兲 of Fig. 1—is
冏 冏 ⌬x x
= ␣+
共x兲E⫽0 − 共x兲E=0 ⬃ 3.3. 共x兲E=0
共6兲
The corresponding value for ⌬x / x in the  phase is calculated through the relationship R / R0 vs x 关see Eq. 共4兲 in Ref. 8兴 and it is equal to
冏 冏 ⌬x x

⬃ 0.02.
共7兲
Then, the mean increase of x due to the plasma action in negative corona regime can be estimated, comparing Eq. 共6兲 with Eq. 共7兲, to be approximately 165 times greater when moving from the  phase to ␣ +  region. A simultaneous measurement of the relative resistance 共R / R0兲in of the SW half portion affected by plasma and the relative resistance 共R / R0兲out of the unaffected portion was carried out during the experiments. The typical time profiles of 共R / R0兲in and 共R / R0兲out observed after the plasma lighting are shown in Fig. 3. The changes were found only in the SW portion affected by the discharge. This suggests that the variations produced locally by the plasma must be larger than those reported in Fig. 1, since R / R0 and 共R / R0兲in changes are measured, respectively, on the whole SW and on one half of it whereas the plasma action is confined to only one-third of the SW length. The difference in the R / R0 values registered prior to the plasma lighting on the two SW portions, which is greater than the instrumental error, may be attributed to some additional resistance produced by the sliding contacts used to measure the SW resistances. It is worth noticing that the main observed change due to the plasma cannot be attributed to the thermal effect of the discharge. In fact, once the plasma is lighted, the final steady state is reached in ⬃30 h, while the thermal equilibrium is reached much faster: the gas temperature outside the CE, measured by a Pt-100 thermoresistance, undergoes within ⬃10 min a mean increment of ⬃0.03 ° C and then remains constant—see insert of Fig. 3. In conclusion, the ability of the hydrogen plasma to increase the loading of hydrogen in the ␣ +  region of the Pd/ H system without changing any other experimental condition has been experimentally demonstrated. The magnitude
of the observed changes caused by plasma depends on the initial position in the absorption isotherm curve of the Pd/ H system. In the ␣ +  region, the changes are wider with respect to the  phase. Thus, at low H2 pressure and under the action of the plasma, it is possible to maintain higher values of x inside Pd than expected at 299 K. The reason why the condition E ⫽ 0 affects the H loading in Pd has been extensively discussed in a previous work8 where it has been shown that there is a different thermodynamic state of H with respect to the condition E = 0, at the same values of T and x. It was suggested that, at E ⫽ 0, the thermodynamic activity of H in Pd is higher than the activity at E = 0. This means a higher chemical potential of H , H 共E ⫽ 0—see Fig. 13 in Ref. 8兲. A tentative explanation of the observed changes can be summarized in two key points: 共i兲 H in Pd is partially positively charged,20,21 H␦+, and 共ii兲 an electric field E* exists within SW, dependent on E, that produces gradients of electrochemical potential of H␦+ and electrons e, ⵜH␦+ and ⵜe, respectively. Due to this, the mass transfer of H␦+ inside the SW is controlled by diffusion and electromigration, i.e., by the gradient of the chemical potential ⵜH␦+ and gradient of electric potential , E* = −ⵜ. At E ⫽ 0, both these terms contribute to the driving force for the SW loading. The corona polarity determines also the sign of the electrical component in ⵜH␦+ but the resulting effect on the driving force on H␦+ is determined from the magnitude of the respective chemical and electrical terms that affect both ⵜH␦+ and ⵜe. This can explain the different results previously found in the  phase when passing from negative to positive corona. In the present work, it was found that the effect of negative corona is much more intense in the ␣ +  region than in the  phase. Therefore, it is expected that under positive corona and in the ␣ +  region, the effect of the plasma is much more intense. Basic Energy Sciences Advisory Committee Report 共http://www. sc.doe.gov/bes/BESAC/Basic_Research_Needs_To_Assure_A_Secure_ Energy_Future_FEB2003.pdf 2 L. Schlapbach and A. Zuttel, Nature 共London兲 414, 353 共2001兲. 3 P. A. Durbin and L. Turyn, J. Phys. D 20, 1490 共1987兲. 4 L. B. Loeb, Phys. Rev. 76, 255 共1949兲. 5 Y. Takahashi, M. Yoshida, Y. Anma, S. Kobayashi, and M. Endo, J. Phys. D 15, 639 共1982兲. 6 J. Chen and J. H. Davidson, Plasma Chem. Plasma Process. 23, 83 共2003兲. 7 U. Fantz, Max-Planck-Institut fYr Plasmaphysik Report No. IPP 10/21, 2002 共unpublished兲. 8 F. Di Pascasio, D. Gozzi, B. Panella, and C. Trionfetti, J. Appl. Phys. 97, 043304 共2005兲. 9 P. C. Stangeby, Plasma Diagnostics: Surface Analysis and Interactions 共Academic, New York, 1989兲. 10 V. I. Demidov, S. V. Ratynskaia, and K. Rypdal, Rev. Sci. Instrum. 73, 3409 共2002兲. 11 F. F. Chen, J. D. Evans, and D. Arnush, Phys. Plasmas 9, 1449 共2002兲. 12 T. S. Pak, J. Sci. Instrum. 41, 180 共1964兲. 13 F. F. Chen, Phys. Plasmas 8, 3029 共2001兲. 14 G. Alefeld and J. Volkl, Hydrogen in Metals 共Springer, Berlin, 1978兲, Vol. 1, p. 2. 15 F. A. Lewis, The Palladium-Hydrogen System 共Academic, New York, 1967兲. 16 H. Uchikawa, T. Okazaki, and K. Sato, Jpn. J. Appl. Phys., Part 1 32, 5095 共1993兲. 17 B. Baranowski, F. A. Lewis, W. D. McFall, S. Filipek, and T. C. Witherspoon, Proc. R. Soc. London, Ser. A 386, 309 共1983兲. 18 R. J. Smith and D. A. Otterson, J. Phys. Chem. Solids 31, 187 共1970兲. 19 T. B. Flanagan and F. A. Lewis, Z. Phys. Chem., Neue Folge 27, 104 共1961兲. 20 A. Coehn, Z. Elektrochem. Angew. Phys. Chem. 35, 676 共1929兲. 21 A. Herold and J. C. Rat, Bull. Soc. Chim. Fr. 1, 80 共1972兲. 1
Downloaded 29 Aug 2006 to 130.37.34.203. Redistribution subject to AIP license or copyright, see http://apl.aip.org/apl/copyright.jsp