Medical Plastics and Biomaterials Magazine
MPB Article Index

Originally published May 1998

METALLIC BIOMATERIALS

Novel Electrochemical Techniques for Analysis of Metallic Biomaterials Surfaces

Jeremy L. Gilbert

Electrochemical phenomena play an important role in the performance of metallic biomaterials. When a metal is introduced to the body, a wide variety of processes and interactions with the biological environment can take place. Such processes include metal ion release; oxide formation, either as a passive oxide film or as a particulate oxide; and corresponding reduction reactions that typically involve oxygen reduction but may also include biochemical species such as proteins. These phenomena vary microscopically over the alloy surface and depend on the local environment, local alloy composition and structure, and local mechanical factors.

Metallic biomaterial products—like these femoral caps made from a CoCrMo alloy—must function under conditions of considerable stress and motion. Photo: Carpenter Technology Corp.

Metallic biomaterials are normally considered to be highly corrosion resistant, owing to the presence of an extremely thin passive oxide film that spontaneously forms on their surfaces. These films serve as a barrier to corrosion processes in alloy systems that would otherwise experience very high corrosion rates. That is, in the absence of passive films, the driving force for corrosion for typical implant alloys (e.g., titanium-based, cobalt chromium (CoCr)–based, and stainless-steel alloys) is very high, and corrosion rates would also be high. Figure 1 reproduces an atomic-force-microscope (AFM) image of a polished and etched surface of a titanium-6% aluminum-4% vanadium (Ti-6Al-4V) alloy showing the oxide film that spontaneously forms on its surface. Note its domed-shape appearance and the difference between the alpha-phase oxide and the beta-phase oxide (raised portion in the center of image). The properties of these passive oxide films depend to a large extent on their structure and chemistry, which are themselves dependent on the substrate's prior thermal, mechanical, and electrochemical history.

Figure 1. Atomic-force-microscope image of oxide film on Ti-6Al-4V. Note the dome-shaped morphology and the difference between the alpha and beta oxide films. Beta grains are shown at center top, with alpha grains surrounding the beta. The image is 4 µm on each side, with a z range of 23 nm.

The in vivo environment is complicated by several factors that are not common to other applications. For instance, most metallic biomaterials are designed to function under conditions of considerable stress and motion, typically cyclic stress (i.e., fatigue) and wear- or fretting-type motion. When stresses and motion are sufficiently large, localized abrasion of the surface oxide film can occur, changing the local electrochemical environment and dramatically increasing the rate of corrosion. In studies previously published by the author, it was found that fretting of oxides will increase corrosion currents by two to three orders of magnitude and decrease the open circuit potential of the alloy by up to –500 mV.^1–3 Such changes can have a profound impact on the nature of the oxide film and on the chemistry of the local solution adjacent to the alloy.

Furthermore, the role and effect of viable cells, proteins, enzymes, and catalysts present in the body are poorly understood. Most traditional electrochemical techniques used to evaluate metallic biomaterial performance—for example, polarization tests or electrochemical impedance tests—do not allow for testing of this complex, microscopically varying interaction between material and environment, nor do they investigate the effect of mechanical factors on corrosion. Standard polarization tests do not provide any detailed microscopic information on the distribution of corrosion over a surface or other information regarding the nature of the oxide film. Nor is it easy to gather detailed information regarding the electrical properties of the oxide films that form on these surfaces using available impedance spectroscopy methods. These latter techniques are typically performed at a fixed potential and do not obtain information through the range of potentials that can be experienced by the alloy in vivo.

This article will present three techniques that have been developed and used by the author to provide more detailed and quantitative information concerning the electrochemical behavior of metallic biomaterials. The three techniques are scanning electrochemical microscopy (SECM), high-speed electrochemical scratch testing, and step-polarization impedance spectroscopy. Each provides unique information relating to local electrochemical and mechanical phenomena, and to the measurement of electrical properties of surface oxides.

Scanning Electrochemical Microscopy

SECM is a technique that has evolved from other scanning probe technologies.^4–6 The basic instrumentation features a three-dimensional translation stage and an electrochemical cell that allows for two independent working electrodes, the sample, and the probe (see Figure 2). Movement actuators attached to the probe can precisely place it (with accuracy to 1 µm or less) within a 2 x 2-mm area and within approximately 1 to 10 µm from the surface of the sample.

Figure 2. Schematic of the scanning electrochemical microscope. The probe is moved by a series of dc motors and piezoelectric translation actuators with positioning. The probe and sample are independently biased to cause a variety of different electrochemical reactions to take place.

Both the sample and the probe can be polarized to whatever potentials are desired in order to react with the solution environment. For instance, if the sample is anodically polarized to corrode (i.e., release ions into solution) and the probe is cathodically polarized to reduce any corrosion product (i.e., lower its valence) coming from the sample, then, as the probe is scanned over the surface of the sample, any variation in corrosion rate should be detected by the probe. If this information is plotted with respect to x-y position, then a three-dimensional image—representing probe current versus position—will result. Thus, in this scenario, the distribution of corrosion on a surface can be imaged while it is actually taking place.

Figure 3. SECM image of ion release from CoCr through a pinhole in the TiN/A1N superlattice coating. The peaks in the figures represent regions of high current and high ionic release. (a) SECM image with sample at +1000 mV and probe at –700 mV. (b) SECM image of identical region with sample at +500 mV and probe at –700 mV. Note clear sign of pinhole in (a) (dome in center) but not in (b).

An example of such an image is depicted in Figure 3a, which shows an SECM image of a cobalt-chromium alloy (ASTM F 1537⁷) that has had a 5-µm intermetallic coating applied (TiN/AlN superlattice coating). The sample has been polarized to above its breakdown potential, 1000 mV (versus AgCl), and the probe has been polarized to –700 mV to reduce any ionic species present in solution. The sample was subjected to these high potentials for more than 16 hours, and several regions of the surface have developed pinholes in the coating. The image shows one such pinhole, approximately 200 µm in diameter, as it is corroding. In Figure 3b, the same region is imaged; however, the sample potential has been reduced to 500 mV (AgCl), which causes a corresponding drop in probe currents since less ionic dissolution is taking place. An SEM image of this same surface is shown in Figure 4. It can be seen that significant amounts of corrosion have taken place through the pinhole.

Figure 4. Scanning-electron-microscopy image of superlattice-coated CoCr after SECM imaging, showing the pinholes that have formed due to transpassive polarization of the substrate.

A wide variety of imaging methods can be used in SECM (see Figure 5). These include monitoring of ion release, monitoring of reduction-oxidation of solution-dissolved species, and competitive reduction of a species in solution. Furthermore, because of the ability to independently control both the sample and probe potentials, one can obtain complete information concerning which combinations of sample and probe potentials will result in probe current changes. These so-called "current-potential response surfaces" are useful not only in providing qualitative information for imaging purposes, but also in quantifying (with the appropriate calibrations) the type and amount of species being detected so that data regarding spatial distribution of concentrations (i.e., diffusion fields) can be obtained.^8,9

Figure 5. Schematic of potential imaging modes for the SECM. From left to right: ion-release mode, solution reduction-oxidation mode (positive feedback mode), competitive reduction mode.

An example of one such current-potential response surface is shown in Figure 6. This is an image of the probe current (10-µm PtIr tip) resulting from sweep voltammetry (at 50 mV/sec) while the probe is positioned approximately 10 to 20 µm above a polarized cobalt-chromium-molybdenum (CoCrMo) sample immersed in pH-7 saline. During sweeping of the probe voltage, the sample is held at a fixed potential; the sample potential is then increased, and the probe-voltage sweep is repeated. From these curves, it can be seen that imaging can take place when the sample is either negatively biased below about –300 mV or polarized above 500 to 600 mV due to changes in the probe current response with potential.

Figure 6. Current-potential response surface for CoCrMo in physiological saline solution. Arrow indicates the potentials for sample and probe where imaging was performed in Figure 3a.

SECM has many potential applications to biomaterials research. These include the study of corrosion heterogeneity in different microstructures, the monitoring of electroactive species distribution over a surface, and even the potential to monitor biologically based electrochemical reactions near implant surfaces or near viable cell membranes. For example, it has recently been reported that using high-sweep-rate cyclic voltammetry to monitor species such as epinephrine or other biological species released from cells in culture can be performed with the microelectodes used in SECM.¹⁰ With the scanning capabilities of SECM, one may be able to spatially and temporally resolve electrochemical events at or near viable cells while they are situated on a material surface. The future of this technique is quite promising, and may provide new insights into corrosion behavior or other electrochemical processes relevant to biomaterials research.

Electrochemical Scratch Tests

As stated earlier, most alloys used in medicine and dentistry rely on a surface oxide film (a so-called passive film) to provide protection against corrosion. In the absence of mechanical factors, these oxide films are in fact highly corrosion resistant. However, when stress, fretting, or wear are present, the possibility of mechanically assisted corrosion significantly increases, and the rate of corrosion can increase as well.^11,12

The author has investigated mechanically assisted corrosion processes with a number of techniques. The one that shows the most promise as a tool to evaluate the combined role of mechanical and electrochemical factors is the electrochemical scratch test.^13,14 In this test, a passivating alloy is first immersed in solution and polarized to a fixed potential (see Figure 7). Next, a diamond stylus attached to a calibrated load cell is brought into contact with the sample surface at a known contact stress. High-speed piezoelectric actuators are then used to apply a scratch of fixed length (typically 20 to 70 µm, and ranging in width from 1 to 10 µm, depending on applied stress) in approximately 1 to 2 milliseconds. The actuators and computer control and acquisition systems are a part of the scanning electrochemical microscope described earlier.

Figure 7. Schematic of scratch test. A controlled load is applied to a diamond stylus while it is moved a fixed length over the surface in 1 to 2 milliseconds. Then, the film reforms anodically and ions are released from the scratch region. = oxide thickness; = oxide density; z = cation valence; F = 96,500 C/mole; A = scratch area; M_w = molecular weight of oxide; = time constant for repassivation; = overpotential for ionic dissolution; b_a = Tafel slope.

Scratching of the surface results in removal of the surface oxide as well as localized deformation of the surface alloy. Immediately after scratching, the alloy undergoes three processes: ionic dissolution through the scratched region, reestablishment of the electrical double layer, and repassivation of the oxide film layer (i.e., the film reforms). These are highly transient events, lasting only about 3 to 10 milliseconds. The current transient and other relevant parameters are monitored with computer data-acquisition techniques; a typical example for a Ti-6Al-4V alloy in phosphate-buffered saline is shown in Figure 8. The characteristics of the current transient are used to describe the behavior of the surface during a scratch test, with peak current and time constant the primary features used.

Figure 8. Current transient response of a Ti-6Al-4V surface to a controlled scratch. The area scratched is approximately 300 µm², which yields a current density of 10 A/cm².

To date, this technique has been employed to assess the effect of sample potential, applied scratch load, solution chemistry, and addition of proteins on the electrochemical response to scratches for CoCrMo and titanium alloys.^13,14 From the effect of potential, one can determine the potential below which an oxide film does not form when the surface is scratched: between –600 and –800 mV versus silver–silver chloride (AgCl) for titanium, and between –400 and –500 mV (AgCl) for CoCr. Above these potentials, an oxide film will form, whose thickness can be directly approximated using a Faraday's law–type calculation. That is, the charge associated with the current transient is equated with the volume of film that reforms. If the area scratched (A_o) is known, then the thickness () can be determined as shown below:

where is the density of the oxide film, z is the charge per cation, F is Faraday's constant, and M_w is the molecular weight of the oxide. One result from this testing is that the film charge is approximately linear—with potential above the onset potential for oxide film formation—which implies that film thickness varies approximately linearly with potential. This is known as the anodization rate, and from these tests it has been measured to be approximately 1.8 nm/V for CoCr (with an onset potential of about –500 mV) and approximately 2 nm/V for Ti-6Al-4V (with an onset potential of about –700 mV).^13,14

It was also noted that the contact stresses required to induce oxide fracture were on the order of the microhardness for CoCr alloys (about 5 to 7 GPa), whereas titanium oxides could be abraded without measurable surface deformation. That is, small-scale surface deformation was needed to induce a transient response for CoCr, but no plastic deformation of the titanium alloy was required for oxide fracture.¹² Testing revealed current densities in the scratched region to be on the order of 1 to 10 A/cm²—approximately 1 million times higher than for the unscratched surface). pH was shown to affect the transient response for both alloys, as did the addition of serum proteins or albumin.^13,14

A significant advantage of the scratch test system is that it is capable of directly imaging alloy surface topography before and after a scratch and so can directly measure the depth of a scratch resulting from a set of conditions (e.g., applied load, scratch length, solution, potential, etc.). The technique is thus ideal for evaluating new surface treatments or alloy variations and for investigating the role of varying solution chemistries on the behavior of alloy surfaces exposed to both mechanical wear and electrochemical corrosion.

Step-Polarization Impedance Spectroscopy

In the last of the three techniques for analyzing metallic biomaterial surfaces, the electrical nature of the alloy-solution interface is of primary importance. Most classical electrochemical investigations involve either polarization testing (anodic polarization, cyclic polarization, etc.), in which the current-potential relationship for an alloy is monitored, or testing of interface impedance using electrochemical impedance spectroscopy methods. In the former process, typically low scan rates (voltage over time) are employed and the results are plotted as potential versus log of the current density. In the latter, the impedance characteristics are typically determined only for one fixed potential (usually the at-rest open circuit potential). The data obtained from these methods are very useful in assessing some properties of alloys, but do not provide a complete picture of the surface electrical properties. Step-polarization impedance spectroscopy is a technique whereby both polarization and impedance behavior can be simultaneously determined over a wide range of potentials.¹⁵

In this technique, a metallic sample is immersed in solution and electrically connected to a potentiostat. A computer with A/D capabilities is then used to control the potential of the sample (through the potentiostat) and to collect the signals (current and voltage) coming from the potentiostat. Then, starting with the sample at some fixed initial potential, a step (typically 25 to 50 mV) is applied and the current transient response is instantaneously collected at an initially high rate (approximately 20,000 counts/sec). After a period of anywhere from 10 to 100 seconds, another step in potential is applied and the current transient is collected again.

The current transient data reflect the response of the alloy-oxide-solution interface, and can be used to obtain both the polarization and the impedance behavior simultaneously. The polarization response is simply the potential versus the log of the current density, taken at any time after the step in potential is applied. Thus, by monitoring how the current changes over time after the step in potential, investigators can determine a whole family of polarization plots. The impedance characteristics, on the other hand, are derived from the current transient after it is subjected to a piece-wise numerical Laplace transform technique, which converts the time-domain data to the frequency domain.¹⁵ This results in a quantity called the admittance (A^*[]), which is a complex number containing both a real (A') and an imaginary (A'') part:

The admittance of an electrical system is simply the reciprocal of the system's impedance (i.e., it is like a conductance). This means that, from the admittance, one can determine the electrical nature of the interface. Some of the parameters that can be obtained from this type of analysis are the high-frequency resistance (Rs), the low-frequency resistance (Rp), and the interfacial capacitance (C).

These kinds of data provide the ability to assess aspects of the interfaces such as resistance to charge transport (ionic and electronic), resistance to oxide film growth and dissolution, density of defects present in the film, film thickness, and so on. Because the details of these analyses are beyond the scope of the present article, the reader is referred to other studies for further information.¹⁵

The following example of the step-polarization impedance technique involved a high-purity platinum electrode subjected to 50-mV steps every 10 seconds while immersed in a phosphate-buffered saline solution. Figure 9 presents four graphs that resulted from this methodology; in all plots, the closed symbols represent the upward steps in potential and the open symbols the downward steps.

Figure 9. Results from a step-polarization impedance spectroscopy test of high-purity platinum. (a) Current density versus potential plots for both 50-microsecond currents (top) and 10-second currents. (b) Capacitance versus potential. (c) High-frequency resistance (Rs). (d) Low-frequency resistance (Rp). Closed symbols represent upward steps and open symbols represent downward steps.

Figure 9a shows the current versus potential relationship for two time points after each potential step—50 microseconds after and 10 seconds after. Note the typical appearance for platinum, with a hydrogen desorption peak at –800 mV in the upward scan and an oxygen adsorption peak at around –300 to 0 mV in the upward scan (closed symbols) for both time points. Similarly, on the downward scan, the oxygen desorption peak starts at about 0 mV in the 10-second curve and at about –300 mV in the 50-microsecond curve.

The capacitance of the interface shows two transitions with potential in the upward and downward scans, as seen in Figure 9b. Changes in C indicate a change in the surface oxide thickness, with increasing capacitance corresponding to decreasing thickness, and vice versa. A noticeable decrease in C occurs in the upward scan at approximately –150 mV, which is about where oxygen adsorption is known to occur. The capacitance increases in the downward scan at approximately –350 mV, once again about where oxygen desorption occurs. Figure 9c shows the high-frequency resistance Rs, which demonstrates only small changes with potential, but these too vary in accordance with ongoing behavior at the interface (adsorption/ desorption events). Figure 9d illustrates the polarization resistance of the interface (i.e., the resistance to charge transfer across the interface). At potentials more negative than –600 mV, the polarization resistance decreases significantly (to about 100 W), whereas above –600 mV it increases to between approximately 8 and 50 kW. The decrease in Rp occurs at the potential at which water is being reduced.

Figures 10a and 10b show the complex admittance of the interface, revealing how it varies as a function of potential and frequency. The changes seen in Figure 9 can also be observed in Figure 10, as can the nature of the current transient behavior. It should be remembered that the admittance is the reciprocal of the impedance, and can be thought of as a frequency-dependent conductance.

Figure 10. Complex admittance for platinum in phosphate-buffered saline. (a) Real admittance versus potential and frequency. (b) Imaginary admittance versus potential and frequency.

The admittance data thus provide detailed information regarding the electrical nature of the interface. With the passivating alloys used in medicine and dentistry, these plots can help to assess the behavior of oxide films as well. For example, the capacitance measures are reflective of the presence of the oxide film and of the density of lattice defects in the oxide.¹⁶ Similar examples of CoCr and titanium alloy behavior have been reported elsewhere.¹⁵

Conclusion

The surface analysis of biomaterials can take many forms. One critical feature of the techniques presented in this article is that they assess the surface of metallic biomaterials in situ—while samples are immersed in a physiological saline. This is important, since it is within the biological milieu that these materials must work and survive for prolonged periods. Because no single technique will provide all of the structural, chemical, and property information needed to understand metallic biomaterial surface performance, complementary methods need to be employed. The three techniques discussed are simply examples of in situ analytical methods that can be used to gather detailed quantitative information that cannot be obtained using other methods. These techniques also provide critical information as to how metallic biomaterial surfaces respond to the real environment of the human body.

Acknowledgments

Portions of the work described in this article were supported in part by the National Institutes of Health and by Osteonics Corp. (Allendale, NJ). The author would also like to acknowledge the assistance provided by Jay Goldberg, Samuel Smith, Christine Buckley, and others.

References

1. Goldberg JR, Buckley CA, Jacobs JJ, et al., "Corrosion Testing of Modular Implants," Modularity of Orthopedic Implants, ASTM STP 1301, Parr JE, Mayor MB, and Marlow DE (eds), Philadelphia, American Society for Testing and Materials (ASTM), pp 157–176, 1997.

2. Gilbert JL, and Buckley CA, "Mechanical-Electrochemical Interactions during In-Vitro Fretting Corrosion Tests of Modular Taper Connections," Total Hip Revision Surgery, Galante JO, Rosenberg AG, and Callaghan JJ (eds), New York, Raven Press, pp 41–50, 1994.

3. Gilbert JL, Buckley CA, Jacobs JJ, et al., "In-Vitro Mechanical-Electrochemical Testing of the Fretting Corrosion Process in Modular Femoral Tapers," Trans Orthop Res Soc, 20:240, 1995.

4. Gilbert JL, Smith SM, and Lautenschlager EP, "Scanning Electrochemical Microscopy of Metallic Biomaterials: Reaction Rate and Ion Release Imaging Modes," J Biomed Mater Res, 27(11):1357–1366, 1993.

5. Bard AJ, Fan FRF, and Pierce DT, "Scanning Electrochemical Microscopy," Science, 254:68–74, 1991.

6. Pierce DT, and Bard AJ, "Scanning Electrochemical Microscopy. 23. Reaction Localization of Artificially Patterned and Tissue-Bound Enzymes," Anal Chem, 65(24):3598–3604, 1993.

7. Standard Specification for Wrought Cobalt-28 Chromium-6 Molybdenum Alloy for Surgical Implants, ASTM Specification F 1537, Philadelphia, ASTM, 1996.

8. Smith SM, and Gilbert JL, "Scanning Electrochemical Microscopy of Co-Cr-Mo Microstructures: Imaging and Current-Potential Response," Compatibility of Biomedical Implants, in Proceedings of the Electrochemical Society, Kovacs P, and Istephanous N (eds), vol 94-15:229–240, 1994.

9. Mueller WD, and Gilbert JL, "TEA-CO₂ Laser Effects on Ti Surface," Compatibility of Biomedical Implants, in Proceedings of the Electrochemical Society, Kovacs P, and Istephanous N (eds), vol 94-15:155–166, 1994.

10. Wightman RM, Hochstelter S, Michael D, et al., "Using Microelectrodes as Secretion Probes in Biological Systems," J Electrochem Soc, Interface, Fall, pp 22–26, 1996.

11. Gilbert JL, Buckley CA, and Jacobs JJ, "In-Vivo Corrosion of Modular Hip Prosthesis Components in Mixed and Similar Metal Combinations: The Effect of Crevice, Stress, Motion and Alloy Coupling," J Biomed Mater Res, 27(12):1533–1544, 1993.

12. Gilbert JL, and Jacobs JJ, "The Mechanical and Electrochemical Processes Associated with Taper Fretting Crevice Corrosion," Modularity of Orthopedic Implants, ASTM STP 1301, Parr JE, Mayor MB, and Marlow DE (eds), Philadelphia, ASTM, pp 45–59, 1997.

13. Gilbert JL, Buckley CA, and Lautenschlager EP, "Titanium Oxide Fracture and Repassivation: The Effect of Potential, pH and Aeration," Medical Applications of Titanium and Its Alloys—The Materials and Biological Issues, ASTM Special Technical Publication 1272, Philadelphia, ASTM, pp 199– 215, 1996.

14. Goldberg JR, Lautenschlager EP, and Gilbert JL, "Electrochemical Response of CoCrMo to High Speed Fracture of Its Metal Oxide Using an Electrochemical Scratch Test Method," J Biomed Mater Res, 37:421–431, 1997.

15. Gilbert JL, "Step-Polarization Impedance Spectroscopy of Medical Alloys in Physiologic Solutions," J Biomed Mater Res, 40:233–243, 1998.

16. Pan J, Leygraf C, Thierry D, et al., "Corrosion Resistance for Biomaterial Applications of TiO₂ Films Deposited on Titanium and Stainless Steel by Ion Beam Assisted Sputtering," J Biomed Mater Res, 35:309–318, 1997.

Jeremy L. Gilbert, PhD, is associate professor in the Division of Biological Materials at Northwestern University, where he has been a faculty member since 1988. His areas of expertise include electrochemistry and corrosion of implantable materials, mechanically assisted corrosion processes, and the development of highly oriented amorphous polymer composites (self-reinforced composites). He also serves as a biomaterials consultant to several medical device manufacturing companies.