clear from the table that an implant can constitute a continuous source of nickel resulting in far higher plasma levels that those due to the heaviest atmospheric pollution.

Although there is no direct evidence as yet that these high levels of plasma nickel are necessarily toxic, one should certainly exercise caution with regard to the potential hazards associated with its use in implants. Indeed our results reinforce the opinion voice by Hueper [35] according to which "the evidence on hand indicates that metal implants which contain nickel and which remain over long periods in human tissues might create delayed potential cancer hazards to their recipients."

FIGURE 7. Computer-plotted curves for the predicted concentrations of 63 Ni(II) in plasma or serum of rats and rabbits, assuming a continuous i.v. infusion of 63 NiCl, at a rate of 1 μg Ni/kgh. We can add finally that the steady state under a continuous input is reached very quickly, as illustrated in figure 7 which pictures the transient curve from the time at which the continuous infusion is started. As can be seen from the figures, it takes about 7-8 hours to achieve a plasma concentration half the maximum value.

This research is supported by the National Science Foundation and the Institute of Materials Science, University of Connecticut.

7. References

[1] Recommended Practice for Experimental, Testing for Biological Compatibility of Metals for Surgical Implants, ASTM Committee F-4 on Surgical Implants, ASTM, . Philadelphia, Pa.

[2] Laing, P. G., Ferguson, A. B., Jr., and Hodge, E. S., J. Biomed. Mat. Res. 1, 135 (1967).

[3] Jones, D. A., and Greene, N. D., Corrosion 22, 198 (1966).

[4] Colangelo, V. J., Greene, N. D., Kettelkamp, D. B., Axelander, H., and Campbell, C. J., J. Biomed. Mater. Res. 1, 405 (1967).

[5] Skold, R. V., and Larson, T. E., Corrosion 13, 139t (1957). [6] Stern, M., Corrosion 14, 440t (1958).

[7] Stern, M., and Weisert, E. D., Proc. ASTM 59, 1280 (1959). [8] Revie, R. W., and Greene, N. D., J. Biomed. Mater. Res. 3,

465 (1969).

[9] Jones, D. A., and Greene, N. D., Corrosion 25, 367 (1969). [10] Revie, R. W., and Greene, N. D., Corr. Sci. 9, 755, 763 (1969).

[11] Picard, R. J., and Greene, N. D., Electrochemical Reactions at Paper Shielded Metal Surfaces to be published. [12] Greene, N. D., Moebus, G. A., and Baldwin, M. H., The Mini Potentiostat: A Versatile Power Source for Electrochemical Studies, accepted publication in Corrosion. [13] Keller, H. O., and Sorkin, E., Separatum Experientia 24, 641 (1968). [14] Mergenhagen, S. E., Snyderman, R., Gewurz, H., and Shin, H. S., Significance of complement to the mechanism of action of endotoxin, Current Topics in Microbiology and Immunology 50, 37 (1969).

[15] Tempel, T. R., Synderman, R., Jordan, H., and Mergenhagen, S. E., J. Periodont. 41, 71 (1970).

[16] Ward, P. A., Arth. & Rheum. 13, 181 (1970).

[17] Ward, P. A., J. Exp. Med. 124, 209 (1966).

[18] Ward, P. A., Lepow, I. H., and Newman, L. J., Am. J. Path.

52, 725 (1968).

[19] Boyden, S., J. Exp. Med. 115, 453 (1962).

[20] Ward, P. A., Cochrane, C. G., and Muller-Eberhard, H. J., J. Exp. Med. 122, 327 (1965).

[21] Ward, P. A., J. Exp. Med. 128, 1201 (1968).

[22] Shin, H. S., Snyderman, R., Friedman, E., Mellors, A., and Mayer, M. M., Science 162, 361 (1968).

[23] Taubman, S. B., Goldschmidt, P. R., and Lepow, I. H., Fed. Proc. 29,434 (1970).

[24] Hegyeli, R. J., Limitations of Current Techniques for the Evaluation of Biohazards and Biocompatibility of New Candidate Materials, Medical Applications of Plastics, H. P. Gregor, Ed. (Wiley, New York, N.Y., 1971) pp. 1-14. [25] Ferguson, A. B., Akahoshi, Y., Laing, P. G. Hodge, E. S., J. Bone and Joint Surg. 44-A, 317 (1962).

[26]

Ferguson, A. B., Akahoshi, Y., Laing, P. G., Hodge E. S.,
J. Bone and Joint Surg. 44-A, 323 (1962).

[27] Sunderman, F. W., Jr., Food and Cosmetics Toxicology 9, 105 (1971).

[28] Ariens, E. J., Molecular Pharmacology (Academic Press, New York, N.Y., 1964).

[29] Riggs, D. S., The Mathematical Approach to Physiological Problems, (Williams and Wilkins, Baltimore, 1963). [30] Berman, M., and Weiss, M. F., Users Manual for SAAM, Publication No. 1703, U.S. Public Health Service. [31] Aubert, J. P., Bronner, F., Richelle, L. J., J. Clin. Invest. 42, 885 (1963).

[32] Bronner, F., Aubert, J. P., Richelle, L. J., Saville, P. D., Nicholas, J. A., Cobb, J. R., J. Clin. Invest. 42, 1095 (1963). [33] Richelle, L. J., Bronner, F., in Proc. First Europ. Symp. Calcif, Tissues, H. J. J. Blackwood, Ed. (Pergamon Press. London, 1964).

[34] Moore, C. V., Dubach, R., in Mineral, Metabolism, and Advanced Treatise, Vol. 3, C. Comar, F. Bronner, eds. (Academic Press, New York, N.Y. 1963).

[35] Heuper, W. E., Carcinogenic hazards from arsenic and metal containing drugs, Potential Carcinogenic Hazards from Drugs, R. Truhart, Ed., (Springer-Verlag, Berlin. 1967), pp. 79-104.

[36] Sunderman, F. W., Jr., The current status of nickel carcinogenesis, Am. Clin. Lab. Sci. (in press).

[37] Van Soestberger, M., and Sunderman, F. W., Jr. 63 Nicomplexes in serum and urine after injection of NiCl2 (in press).

[38] McNeely, M. D., Nechay, M. W., and Sunderman, F. W., Jr., Clinical Chemistry 18, 992 (1972).

(Issued May 1975).

and

1. Introduction

1

Implantable hybrid cells consisting of catalytic cathodes where dissolved oxygen is reduced and sacrificial anodes are galvanically oxidized are capable of powering pacemakers [1-10], carotidsinus, bladder and electrophrenic stimulators [11]. The implantable hybrid cell has a long shelf life, costs no more than conventional batteries and has a potential for lasting 10 years or longer. The power generating capabilities of the hybrid cell depend upon the cathode electrode potential and current drain. Under zero current drain the electrodes in the body develop an open circuit voltage. When current is drawn through the load, the electrode potential decreases from its open circuit value due to the overpotential developed at the electrode. Higher power outputs are produced by electrodes which can maintain higher potentials for given current drains. In-vivo and in-vitro studies have demonstrated that the power of the cell is decreased due to the overpotential at the catalytic cathode. There is sufficient oxygen available in the subcutaneous tissue according to Roy et al. [12], to depolarize for current loads in excess of 200 μA which will produce sufficient power to drive pace

'Figures in brackets indicate the literature reference at end of this paper.

makers. At Drexel, implanted hybrid cells powering commercial pacemakers pacing dogs use porous cathodes which have operated with current loads of 610 μA or 25 μA/cm [12, 13].

In-vivo and in-vitro studies on platinum and palladium black cathodes consisting of hydrophobic binders of teflon and hydrophilic binders of polyvinyl chloride and silicone rubber have demonstrated considerable overpotential with a resulting power loss [8, 13].

In order to increase the number of available sites for reducing the oxygen and to obtain a more biologically compatible interface [14], platinum and palladium porous cathodes were fabricated using the techniques of powder metallurgy [15, 16]. The polarization or overpotential at the various types of cathodes has been studied using the galvanostatic technique. In-vivo studies have been made implanting the electrodes under load subcutaneously in the lateral abdomen and thorax of dogs. Material and biological compatibility characterization studies have been made on the various types of cathodes. The surface areas of the palladium and platinum powders and compacts have been determined using the Brunauer-Emmett-Teller (BET) gas adsorption technique and the scanning electron microscope. The compatibility of the electrodes have been determined by histopathologic evaluation of the

tissues surrounding the porous metal implants [15, 16].

Besides the problem of decreasing the electrode polarization (i.e., overpotential) of the power source, there is also the problem of a significant waste of stimulation energy at the cardiac pacemaker electrode surface due to polarization [17, 18, 19, 20]. A decrease in the electrode surface increases the current density, lowering the threshold voltage for stimulation, but increasing the polarization voltage [20]. There is therefore an optimum size electrode and pulse shape [17, 42] from the standpoint of both stimulation threshold and pacemaker energy.

Although a variety of electrodes have been investigated for delivering an electrical stimulus to the heart [17, 18, 20, 21, 22, 23], noble metals such as platinum and platinum alloys are commonly used and recommended because of their inertness to body tissue [17]. Very high thresholds due to infection have been reported [23]. The current threshold is determined by the minimum current density at the electrode required for stimulation of the surrounding myocardium. Fluctuations which occur in the current threshold immediately after implantation are attributed to the process of tissue formation. When the myocardial tissue reaction is complete the threshold stabilizes. Infection around the electrode with tissue degeneration results in a further increase in threshold current [23]. The above investigations [17, 19, 21, 22, 23] have been limited to electrode materials currently used in cardiac. pacemaker stimulating electrodes. The polarization that occurs at the smooth platinum electrodes used as pacemaker stimulating electrodes is much greater than in the case of platinum black electrodes.

The porous stimulating electrode on the other hand can be made small with the electrode polarization being reduced without effecting the current density just outside the electrode surface area. The rise in threshold is attributed not to an increase in the electrical impedance of the scar tissue surrounding the electrodes, but rather due to an increase in the size of the electrode surface due to the conductive scar tissue with a resulting decrease in current density [20]. Porous ceramics with 45 to 100 μm pore size have shown thin fibrous encapsulating layers of 4-6 cells in thickness with a good blood supply or no observable encapsulation as contrasted to thick encapsulation around impervious discs [24, 25]. Thus, porous stimulating electrodes should have lower electrode polarization and a lower threshold of stimulation due to tissue ingrowth. Tissue ingrowth will also help to fix the stimulating electrode in place. One of the principal complications remaining with the transvenous technique has been reported to be electrode displacement in the right ventricular inflow tract [26].

Electrode polarization is also a problem in impedance measurements on biological materials, and corrections for this polarization must be made.

if excessive measurement errors are to be avoided [27]. Corrections are not only large but uncertain at sufficiently low alternating current frequencies where polarization becomes large [28, 29]. At present, electrodes in our laboratory and elsewhere [27] are coated with platinum black by electroplating from a Kohlrausch solution. This normally reduces electrode polarization to a relatively low level for frequencies above 100 kHz. However, for measurements on physiological fluids at lower frequencies, corrections for polarization must be made [18, 19, 28, 29, 30, 31] or different electrode configurations used. For example, the four-electrode null technique [28] has been applied to impedance measurements over the range of 10 Hz to 1 kHz. Here voltage across the sample is applied by one pair of electrodes and another pair of noncurrentcarrying electrodes is used to measure the voltage across the sample. The electrodes have a platinum black coating to further decrease polarization. Platinum black electrodes deteriorate with use in biological fluids which is believed due to macromolecules effecting the platinum black surface [27]. If macromolecules are blocking the sites of the platinum black interface and decreasing the ease of charge transfer, this would account for the increased polarization when the electrodes are used in protein solutions. Again, the development of porous electrodes could be applied to improve biological impedance measurements.

DeRosa and Beard [30, 40] at Drexel have considered polarization in terms of rate determining steps. The effects of electrode d.c. potential, pH and oxygen coverage on the mechanism of cathodic reduction of oxygen and their influence on the cathode polarization has been studied [30]. Anodic polarization under the variation of the above parameters has also been studied. The anodic as well as the cathodic reactions are important in a.c. polarization in electrical impedance measurements and at bipolar pacemaker electrodes.

2. Experimental Procedures

2.1. Modified Transient Galvanostatic Method for Studying Polarization [33]

When using the galvanostatic technique to study the variation of electrode potential as a function of current density, the cell current is changed suddenly from zero to a given value, and the resulting variation of potential with time is recorded. The electrode under test and a counter electrode are placed in a Ringer's solution which has an ionic concentration approximating that of subcutaneous tissue fluid. The potential of test electrodes with reference to a silver-silver chloride electrode is measured as shown in figure 1. The cell current is changed from zero to a given current density by a current pulse of approximately 1 ms. width and 500 μA amplitude. The electrode potential is measured with respect

to the reference electrode. Since the current through the reference electrode is very small, there is negligible overpotential and voltage drop in the solution between it and the electrode under test.

When a stimulus such as a step current is applied to an electrode interface, most of the initial current is used to charge up the double layer while later on the constant current density is used to promote electronation or a charge transfer reaction [33]. By recording the variation of the potential with time between a nonpolarizable reference electrode and the test electrode, the kinetic parameters of the electrode reaction can be obtained [33, 34].

2.2. The Brunauer, Emmett, and Teller (BET) Gas Adsorption Method

The Brunauer, Emmett, and Teller (BET) Gas Adsorption Method has been used to measure the surface area of porous compacts and powder [16, 35-38]. The utilization of the technique requires one to obtain an experimental gas adsorption isotherm from which the volume corresponding to monomolecular gas coverage is estimated. Once this volume is known along with the average area of the molecules of the adsorbate, the surface area of the specimen can be calculated.

The experimental setup which was used is similar to Kreiger [38] and is basically a combined gas buret and mercury manometer. To obtain more accurate pressure measurements, a thermocouple was also used. Before testing began, the specimens were outgassed at 0.5 μmHg at 200 °C for approximately 16 hours. Helium was used for dead space determination and nitrogen for the adsorbate during the actual measurement portion of the tests. All tests were made with the sample container immersed in a liquid nitrogen bath. The resulting

The product of Vm, the cross-sectional area of the absorbate and Avogadro's number corrected to standard temperature and pressure gives the surface area, SBET.

2.3. Scanning Electronmicrographs

A scanning electron-microscope 2,3 was used to study the loose powders and compacts. The preparation of biological materials for the scanning electron microscope was as follows: Upon removal from the animal, the sample was washed in saline to remove mucous and fixed in glutaraldehyde. The sample was then using a succession of ethyl alcohol-water baths, i.e., 58, 90, 95, and 100 percent alcohol. The alcohol was dehydrated then replaced with amyl acetate. The samples were then freeze-dried in conventional manner, coated with gold, and examined in the scanning electron microscope.

2.4. Impedance Measurements

Electrode polarization, which greatly influences electrical impedance measurements, can be studied both from time and frequency domain measurements [19, 21, 33, 40]. Plastic cells of poly(tetrafluoroethylene) and poly(methyl methacrylate), consisting of thin-walled measuring chambers surrounded by circulating water for temperature control, were used for studying a variety of porous electrodes. For time domain measurements constant current pulse source is connected to the electrodes and the amplitude of the current pulse is increased in steps. The potential across the electrodes and current are measured at each level. The

a

impedance measurements were made using a Wayne Kerr B221 Admittance Bridge which measures the parallel conductance and capacitance. The amplitude of the applied sinusoidal voltage is only 25 millivolts RMS which insures the electrical impedances are measured in the linear range, i.e., not in the range where impedance is a function of the amplitude of the applied signal. Measurements were made at 25 °C over the frequency range from 200 Hz to 20 kHz. While cell temperature was controlled and the pH and oxygen concentration was monitored with an oxygen meter.

3. Materials

3.1. Fabrication of Porous Electrodes Following powder metallurgy techniques, the loose powder is placed in a hardened metal die and

is compressed under high pressure. Under pressure, the particles become cold welded together and the magnitude of the compacting pressure greatly influences the density and strength of the resulting compact. The required compacting pressure depends upon the ductility of the base metal from which the powder is formed, e.g., aluminum compacts at 10,000-30,000 psi (0.07-0.2 GPa), ferrous powders at 30,000-60,000 psi (0.2–0.4 GPa), and tungsten at 80,000-150,000 psi (0.5-1.0 GPa). Since no published information was available regarding compacting pressure versus density or strength, a number of platinum and palladium powder electrodes were prepared at different compacting pressures. These pressures were 10,000, 30,000 and 50,000 psi, which resulted in compacts with gradually increasing density. The green strength of the compacts was sufficient to withstand sterilization in a steam autoclave and they were therefore left unsintered.

The stimulating electrodes have been fabricated under a variety of techniques. Some electrodes were fabricated from platinum- and palladium-black powders with an isostatic compaction technique, then sintered. In the fabrication of large pore sizes, poly(methyl methacrylate) (PMMA) powder was blended with the palladium black and compacted under a hot press technique [41].

4. Results

In-vivo results for the porous electrodes used as power source electrodes demonstrated that there was a six-fold increase in power density for these electrodes [15]. Figure 3 shows the comparison between a porous palladium electrode and others with plastic binders. Porous Pd electrodes implanted in dogs over a period of 146 days under loads of 70 μA/cm2 had no adverse tissue reactions [43, 44].

Figure 4 illustrates that the surface area of platinum and palladium is of the order of 104 times greater than foil. The large surface area, with a large increase in sites, accounts for the marked decrease in overpotential.

Figure 5 demonstrates the decrease in polarization of the porous stimulating electrodes which are compared to the standard platinum-iridium tips. The potential difference versus time across the platinum-iridium electrode has, as shown by its initial jump, a considerable potential across its electrodes and, as shown, in the curve of figure 5. a large overpotential across the interface. The amplitude of the 1 ms. current pulse was held to 200 μA, which gave a current density of 200 μA/cm2.

* Type JSM-2, Japan Electron Optics Laboratory Company, Ltd.

3 Certain commercial materials may be identified in this publication in order to adequately specify the experimental procedure. In no case does such identification imply recommendation or endorsement by the National Bureau of Standards, nor does it imply that the material are the best available for the purpose.

The metal powders were obtained from Matthey Bishop, Inc., Malvern, Pa.