MANUFACTURING STRATEGIES Implanted devices and prostheses have revolutionized many branches of medicine. Products such as vascular stents, prosthetic hips, and dental implants use a variety of materials to provide the required combination of structural strength, biocompatibility, osseointegration, and, in the case of joint replacements, low-friction load-bearing surfaces.

Wayne Mozer

September 1, 2006

13 Min Read
Surface Metrology for Stent and Implant Manufacturing

The physical parameters of surface topography and roughness, component dimensions, radius of curvature, and coating thickness all affect the medical benefit of these devices. Device manufacturers need surface metrology tools that can simultaneously map all of these parameters for a range of material surfaces and over a wide range of roughness values for both flat and highly shaped surfaces.

White-light interferometry is an optical profiling technique that is ideally suited to this task. This technique combines the requisite sensitivity and accuracy for research work with the speed necessary to support both process-monitoring and quality control (QC) functions. In addition, the technology uses advanced software algorithms to simplify data acquisition and analysis.

Stents

Stents were developed to treat various vascular conditions and blockages and to provide an alternative to highly invasive, life-threatening surgeries, particularly in the treatment of coronary artery disease and blocked carotid arteries. In fact, the current worldwide coronary stent market is estimated at $6 million.1 A typical stent is a mesh-like tube used to support the vessel wall after minimally invasive treatments such as balloon angioplasty. The stent material depends on the intended use; stainless steel and cobalt-nickel are preferred for cardiovascular applications because of their mechanical strength. But these metals aren't appropriate for carotid devices because there is no rib cage protection and the vulnerable artery may be pinched accidentally. Instead, carotid stents often use nitinol, a shape-memory alloy that automatically recovers its shape after deformation. In most cases, the metal stent is produced by a three-step process: laser cutting, followed by chemical (acidic) etching, and then electropolishing.

One problem with stenting is restenosis—when the body coats the stent with scar tissue and reblocks the treated vessel. Consequently, many stents are now coated after the bare metal has undergone its final electropolishing. The coating consists of alternate layers of a barrier material, such as parylene, and a biodegradable polymeric material that elutes an antirestenosis drug as the material dissolves. The parylene is applied using vapor deposition and the polymer is often applied by an ultrasonic spray coating or dipping. Coating thickness is typically in the range of 1–16 µm.

Each fabrication step has a target process window that must be monitored. Critical parameters for stents include width, feature shape, and radius of curvature of the device. In addition, surface roughness must be minimized to avoid physical damage to the treated vessel and to ensure adhesion of the coatings. Potential defects include broken struts, scratches, pits, improperly carved junctions, surface contamination, and incorrect thickness and lack of uniformity of the various coating layers.

Dental Implants

With endosseus dental implants, the optimum outcome depends on new tissue adhering to the implant. Early implants were simply based on a threaded titanium post. In the past decade, implant manufacturers have produced a range of variants to improve long-term outcomes as well as to support one-step procedures where the implant and a crown are inserted in a single visit.

Newer implants are available with various surface characteristics, ranging from relatively smooth machined surfaces to rougher surfaces created by coatings that have been blasted by various substances, acid treatments, or combinations of the treatments. Bioglasses, calcium phosphates such as fluorapatite and tricalcium phosphate, and inert ceramics such as alumina are all commonly used as coating materials. The most popular material is plasma-sprayed hydroxylapatite. All of these coatings have been shown to enhance bone apposition better than uncoated metal implants.

Similarly, surface roughness plays a key role in the biocompatibility and long-term survival of these titanium implants. But the roughness also affects the long-term integrity of any coating, so the optimum surface roughness for an implant is based on multiple factors. As a result, both researchers and manufacturers need to perform quantitative, reliable surface metrology. This includes mapping the roughness of coated and uncoated surfaces, as well as measuring coating thickness.

Hip Implants

It is estimated that in the United States alone, the market for hip implants will surpass $2 billion by 2007, a number that will continue to grow as baby boomers enter retirement.2 Unfortunately, these prostheses have a finite lifetime. The most significant problem affecting long-term replacement success is the minute but progressive shedding of material from the implant. The resulting wear particles cause bone resorption and produce granuloma tissue. They may cause pain and ultimately loosen the prosthesis. Also, the parts can wear through, destroying the portion of the prosthesis supporting the bearing surface. That these hip implants degrade over time is not surprising, considering that forces of four to seven times one's body weight are transmitted millions of times every year through each hip joint. Biomaterials advances have allowed device manufacturers to experiment with new bearing materials, and several different options are now available for hip replacement surgery. But what constitutes the best option is problematic because the results of laboratory tests do not always correlate with real-world results. A metrology tool that examines coated and uncoated surfaces and provides a means for ongoing quantitative analysis can help resolve this uncertainty.

mddi0609p74a_thumb.jpg

Figure 1. (click to enlarge) Figure 1. When light bounces off a thin transparent film, interference can cause light and dark bands to appear in the reflected image. These bands are called fringes, and their location is a function of film thickness.

White-Light Interferometry

Optical profiling is a metrology method that is quickly gaining popularity for these types of devices and implants. It relies on white-light interferometry technology, which utilizes the wave nature of light to create an extremely precise ruler. This is accomplished using the same interference phenomenon that produces vivid iridescent colors when sunlight is reflected off a thin film of gasoline floating in a water puddle.

When light waves are reflected off a thin film, there are two reflections: one from the top and one from the bottom of the film (see Figure 1). The two reflected waves recombine in a process known as interference. Specifically, if the two recombining waves have the same phase (i.e., the wave crests coincide), they combine to make a single, more-intense wave. If they are out of phase (i.e., the crest of one wave coincides with the trough of the second), they cancel each other out. The relative phase of the two reflected waves depends on the path difference, which is set by the film thickness and the angle at which the wave hits the surface. As a result of this interference, a series of dark and light bands is visible. These bands, called fringes, map the thickness of the film at nanometer resolution. The film thickness (z dimension) can be mapped with high lateral (x–y dimension) resolution when viewed through a microscope.

mddi0609p76a_thumb.jpg

Figure 2. (click to enlarge) Figure 2. In an optical profiler, a digital camera records fringes that result from reflections off a test surface and a reference surface. The system converts these fringes into high-resolution topographic information.

The white-light interferometer is a type of microscope in which the two waves are split with a partially reflecting mirror called a beam splitter. As shown in Figure 2, light is reflected from two different surfaces rather than two sides of a film. (There is a test surface and a reference surface.) But the same pattern of light and dark fringes will still be seen, provided the path difference is very small (micrometers or less). If the reference surface is extremely flat (<0.5 nm rms), the fringes can be used to quantitatively map the topography of the test surface. The topography can be measured in two ways: first, the interference effect drops off dramatically as the path difference increases because fringes are only formed when the difference is close to zero. Therefore, the surface can be mapped by measuring the microscope focus position that gives the highest-contrast fringes. Second, the exact position of these light and dark fringes provides even-higher-resolution depth information.

Practical Instrumentation

Figure 3 shows a white-light optical profiler designed for benchtop work. The illumination source is a long-life halogen lamp. Reflections from the test surface and reference surface are recombined in the microscope objective and recorded using a low-noise charge-coupled device digital camera. The measurement speed depends on the camera size and how the instrument is used: the desired range of the vertical (z) scale and the z-axis resolution. A typical data set takes less than 10 seconds for full z-axis resolution over the entire x–y field of view. Microscope objectives are available with fields of view ranging from 0.05 to 8.45 mm in diameter. The objective and the number of pixels determine the lateral (x–y) resolution of the instrument, which can be as fine as 500 nm. The z-axis (surface height) resolution is the same for all objectives and is better than 0.1 nm. The technique also provides absolute accuracy (3 nm) and repeatability (30 nm).

Once the test surface is placed on the microscope stage, the system rapidly steps the microscope objective vertically to record the fringes as different parts of the surface pass through focus. A series of algorithms is then applied to the data to map the precise height of each point on the surface with subnanometer resolution. When surfaces have a transparent coating, the instrument's computer automatically separates the reflections from the coating and the underlying surface, allowing accurate mapping of both surfaces and the coating thickness.

mddi0609p76b.jpg

Figure 3. A typical optical profiler is a rugged self-contained instrument (the Veeco Wyko NT1100 is shown here).

In addition, system software includes pattern-recognition routines that calculate feature widths and relative positions. It can also be customized to identify deviations from the ideal shape. The software screens for defects such as scratches and pits at user-specified lateral and vertical thresholds. Automatic part rejection and cause-logging improve process control. But even for simple pass-fail tests, full surface maps are recallable for additional review. An optical profiler is capable of absolute accuracy and every instrument is fully calibrated by the manufacturer. In the critical z-axis, there are two steps to this calibration: first, the linearity of the z-axis response is verified using a series of glass slides with known height-step features. Then, a final single-point calibration is performed to recheck the absolute accuracy of the z scales, again using a glass-height standard. These step-height standards are commercial products that are NIST (National Institute of Standards and Technology) traceable.

The x-y scale is a function of the digital camera and the microscope objective. The scale is fully calibrated at the factory because a microscope objective is rarely the precise quoted magnification; for example, a 10× objective rarely has a magnification of 10.000. The x-y scale is calibrated using NIST-traceable striped optical targets.

mddi0609p78a_thumb.jpg

Figure 4. (click to enlarge) Figure 4. A 3-D surface profile of a defective stent. Pits are found automatically and their dimensions, depths, and volumes are logged into a database.

mddi0609p78b_thumb.jpg

Figure 5. (click to enlarge) Figure 5. Measurement of part of a coated stent showing top (a) and bottom (b) surfaces, as well as the coating thickness (c). Pits in the coating can be seen in the lower half of the image.

Advantages of Optical Profilometry

Measurement Speed. Optical profilometry offers a number of critical advantages over other topographic metrology techniques. Measurement speed is often cited as the most important of these advantages, particularly for production applications requiring 100% sampling. This speed is a direct result of digital imaging technology, which allows an image of the entire depth profile to be recorded and analyzed in just a few seconds. All digital cameras exhibit noise in their readout. Since optical profiling depends on intensity measurements of fringes (stripe contours), noise can affect the measurement accuracy. It typically translates into data noise of a few nanometers or less. But many optical profilers can be operated in a phase shifting mode—where camera noise translates into accuracy errors of less than 0.1 nm. Furthermore, because each image contains hundreds of thousands of pixels, the speed does not come at the expense of the sampling area or spatial resolution. When this speed is combined with optional automated scanning routines, even a complex stent can be characterized quickly. By contrast, traditional quantitative techniques sample a single point or line transect, making them too slow for complete surface characterization.

Multiple Parameters. Another important advantage of optical profilometry is the ability to simultaneously measure multiple parameters. The instrument software can use a single data set to determine all topographic parameters of interest. These include local and average roughness, radius of curvature, dimensions of all structures and components, identification and location of defects, and coating thickness and uniformity. No other metrology technology can match this level of broad utility.

Dynamic Range. Optical profilometry also stands out because of its dynamic range. In the z-axis, the technique delivers resolution of less than 1 nm, yet it can measure features and step heights as large as 8 mm.

Noncontact. Another advantage of optical profilometry is that it is a completely noncontact method. This enables softer materials and coatings to be measured without the risk of altering the surface. The instrument can also look through an open structure. For example, the technique can be used to measure the internal surface of a stent by observing it through the mesh on the opposite side.

Ease of Use. Ease of use is an important practical advantage of optical profilometry for medical device manufacturers. The combination of rugged, turnkey instrumentation with powerful, customizable software means that these instruments can deliver state-of-the-art data for a variety of applications. These range from demanding research programs to in-line operation for automated pass-fail QC tests to process monitoring and feedback.

mddi0609p78c_thumb.jpg

Figure 6. (click to enlarge) Figure 6. Measurement of a variety of implant surfaces: ceramic hip implant head (a), polymer hip implant cup (b), steel knee implant (load-bearing surface) (c), and titanium dental implant (d).

Sample Results

These instruments can log and graph data in a variety of ways to support both research and production applications. Figure 4 shows the 3-D surface profile data for a defective (uncoated) stent. This data set has been presented with the commonly used approach of showing false color to represent the height and depth dimension. The pits have been automatically located by the instrument and their dimensions, depths and volumes are logged into a database. Average surface parameters are located in the upper left of this screen capture. When needed, arbitrary 2-D transects can also be plotted from the full 3-D data set.

Figure 5 presents typical data for a coated stent. Even though this particular surface has a curved (cylindrical) profile, the optical profilometer can still graph the top and bottom of the coating and measure and plot film thickness. Global averages for the film-thickness data set are calculated and shown.

Figure 6 illustrates the broad applicability of optical profiling for medical implants, including false-color topographic maps for a selection of implant surfaces: a hip implant (both head and cup surfaces), a knee implant (load-bearing surface), and a dental implant.

Conclusion

Accurate surface metrology is critical to the continued development and production of more-sophisticated medical implants and prostheses. White-light interferometry offers a unique approach because it can deliver high-resolution data for many different topographic parameters on a wide range of devices. Furthermore, the technology's rugged, turnkey packaging makes it useful for production-line personnel.

Wayne Mozer is a product specialist in white-light interferometry at Veeco Instruments Inc. (Woodbury, NY). He can be contacted at [email protected].

References

1. Boston Scientific Corp., Quarterly Report, May 2006.

2. Datamonitor, Commercial Perspectives: US Hip and Knee Replacement—Market Surges in New Millennium, Market Report, January 2004.

Copyright ©2006 Medical Device & Diagnostic Industry

Sign up for the QMED & MD+DI Daily newsletter.

You May Also Like