Metallurgical Analysis of Unimplanted Stem

Optical Analysis

Although the precise thermo-mechanical history of the stem examined for this research is not known, several features of its bulk microstructure provide important clues as to how it was processed. Figure 5** provided in the section Mechanical Metallurgy and Passivation of Titanium shows a typical fine and equiaxed a + b grain morphology. A photomicrograph of the sectioned, mounted and polished stem is given in Figure 22. Due to the similarities between these two micrographs, it is reasonable to conclude that this alloy may also have been heat treated at about 950°C for one hour, air cooled, and annealed 2 hours at 700°C. However, instead of a 700°C anneal, it is likely that the diffusion bonding of the mesh to the stem elicited similar effects as the anneal. A micrograph of the narrow section of the stem (Figure 23) also exhibits distinct 'banding' in the vertical direction, indicative of a forging operation. The outlined region of the upper right hand corner of this figure is detailed in Figure 24.

Figure 23

Figure 24

The large bright inclusions visible in both micrographs were located several microns below (and followed the contours of) the stem surface and are raised above the surrounding structure. Although too small to measure with Knoop hardness, they did exhibit much less deformation than the surrounding matrix when a hardness indentation was made in their vicinity. Analysis with EDS indicates that these inclusions are primarily composed of titanium. Only a small indication of aluminum was detected, thus they are probably not an intermetallic phase. It is possible that, due to their proximity to the surface, these inclusions are interstitial hydrogen hardened (and stabilized) retained b titanium resulting from hydrogen absorption by the stem. Another possibility is that they consist of titanium carbide. If this is indeed the case, carbide presence in this quantity indicates that the starting alloy may not have been 'Extra Low Interstitial' (ELI) Grade as specified by the ASTM for orthopedic applications. These specifications are given in Table 10.

An optical micrograph of the mesh-stem interface are given in Figure 25. Figure 25 shows the microstructural variation across the diffusion bonded interface from commercially pure (a) titanium at the mesh to acicular a with small amounts of b, to a final a + b equiaxed structure. Figure 25 also shows evidence of the needle-shaped hydride phase (TiH2) within one of the fibers. This micrograph also shows that the mesh is not completely bonded to the stem; pores are readily visible at the interface.

SEM Analysis of Unimplanted Stem

The sectioned and mounted stem was examined in its Bakelite mount in order to augment the findings of optical microscopy. Particular attention was paid to the mesh-stem interface (Figure 26), regions of hydride evidence, and stem surface. Figure 27 and Figure 28 show that although not prevalent, hydrides are present in many locations of the fiber mesh. As discussed above in the section Effect of Hydrogen on Microstructure, we would expect hydrides to form preferentially in this region for three reasons: 1) the mesh consists of a phase titanium which exhibits very low hydride solubility at room temperature, 2) the mesh fibers are at the outermost surface of the implant and more accessible to hydrogen absorption, and 3) this region was exposed to a high temperature treatment and slowly cooled to room temperature, allowing coarse hydride platelets to form (as opposed to a fine dispersion of hydride commonly observed in quenched a phase). An oblique view of the fluted (longitudinally grooved) stem surface is given in Figure 29. Two large particles (confirmed by EDS to be Al and Si-rich) can be discerned above the surface layer. A higher magnification, normal view of the surface (Figure 30) shows the extent of glass grit particle embeddement. These micrographs correlate well with those presented by Ricci et al.

Figure 27

Figure 28

Figure 29

Figure 30

Microhardness

Microhardness results show that the bulk stem material is over two times harder than the mesh material. Approximate values are: Bulk KHN(300g) = 55, Fiber KHN(300g) = 22. These values are comparable to those seen in brass and aluminum alloys, thus it is evident that this alloy has not been hardened to a considerable extent. The mesh is expected to be 'dead' soft since it primarily consists of annealed, commercially pure metal. The stem is twice as hard owing to its interstitial and substitutional hardening additions of hydrogen, oxygen, carbon, aluminum, nitrogen, and vanadium. However, it is far from its hardest condition of KHN = 275.

Particle Characterization

Control Particles

Nickel Control Particles

Owing to their very ductile origin, these particles have been smashed during the ball milling process into very flat, rounded edged flakes, as shown in Figure 31. Surface markings are not visible under electron microscopy. Particles have agglomerated and there is a wide distribution of sizes visible. No evidence of brittle behavior or chemical attack is evident.

Titanium Control (Wear Generated) Particles

Some of these sliding wear generated particles were easily visible with the naked eye. When viewed under the electron microscope (Figures 32 and 33), many of these were helical, indicative of extensive and ductile wear. In fact, they resembled cold butter when it is scraped with a knife. These shear banded particles bear a striking resemblance to those observed by Xiaoxia et al. in their study of the wear debris of Ti6Al4V in an acidic medium, Figure 8**. This morphology was determined to exhibit a relatively low hydrogen content by these researchers. On the other hand, some of these microscopic particles were thin and jagged, and only small amounts of cracking was evident (left edge of Figure 33). Most of the debris consists of metallic particles ranging in size from about 500 µm to less than 1 µm in diameter.

Figure 33

Retrieved Titanium Debris Particles

Scanning Electron Microscopy Results

Features Common to All Particles Examined:

SEM micrographs of the six extracted debris specimens some common features. The range of particle sizes for these six specimens varied from about 0.5µm (or below the electron microscope's resolution) to over 175µm with diameters of 5-10 µm being the most abundant. The larger particles always exhibited much smaller particles attached to their surfaces. Peaks for (in order of decreasing intensity) P, Ca, Cl, Al, Si, Na, and S were detected when non-metallic debris was analyzed with EDS. Similar results were also found by other researchers (Hanlon, et al.) This paper by Hanlon, et al. indicates that when some particles were subjected to EDS, either Na, Cl, Ca, and P only or Ti in combination with these elements were detected. He concludes that these elements are present because not all of the mineral component of bone tissue had been dissolved and often remained strongly attached to titanium surfaces. The research conducted on the residual grit particles found at the surface of titanium implants by Ricci et al. also asserts that Si and some Al rich particles are likely to be alumina or silica grit particles. Another interesting trend for all specimens is that no Co, Cr, or Mo was detected, even though these implants had been in direct contact with a CoCrMo alloy head during their service.

Variable Features of Examined Particles:

Figures 34(a) and 34(b) show the variability in the quantity of metallic debris present in three typical in vivo specimens. Two specimens exhibited extensive quantities of metallic debris. Electron micrographs of these specimens are given in Figures 35, 36, and 37. Thin, rough particles with surface cracking, very similar to those shown to contain large quantities of hydrogen by Xiaoxia et al. (Figure 9** above) are common (Figures 35 and 36), as are 'intermediate' particles in the other of these two 'high-burden' specimens that exhibit signs of both surface grooving indicative of sliding wear and thin or jagged particles (Figure 37).

Figure 34(a)	Figure 34(b)
Photographs of Three Typical Debris Specimens in (a) Settled State and (b) Agitated State Showing Different Starting Debris Concentrations

Figure 36

Figure 37

The third specimen exhibited primarily polymeric/bone debris macroscopically and very large, smooth, relatively infrequent metallic debris particles (shown in Figure 38). The remaining three specimens exhibited intermediate amounts of metallic debris, and exhibited large, flat, jagged or sometimes ductile particles. These particles are depicted in Figures 39, 40, and 41. One specimen exhibited iron-based particles that are believed to be artifacts from the tools and mechanical procedures used by the surgeon to remove the implant.

Figure 39

Figure 40

Figure 41

Transmission Electron Microscopy

The electron diffraction pattern shown in Figure 42 was obtained from a particle from the same specimen as shown in Figure 41. The appearance of the electron diffraction pattern (Figure 42) coupled with a STEM photo (Figure 43) of the area penetrated by the electron beam indicate very small grain size. Table 11 shows that the particle is likely to be primarily b phase titanium. Thus, since EDS indicates that the particle was composed of titanium, aluminum, and possibly vanadium, (Figure 44) we can conclude that particle was indeed a fairly homogenous chunk of the parent alloy of the stem or, owing to its shape and microstructure, from the stem-mesh interface. It is an important result that the particle is not an oxide, intermetallic, or hydride, all of which can exhibit similar energy dispersive analysis spectra (the tool commonly used in previous debris particle analyses). Up to this point, it appears that TEM techniques have not been used by investigators when they examined implant wear debris. The use of STEM and electron diffraction now confirm what investigators have been assuming, i.e., observed wear debris does not primarily consist of oxides, aluminides, or hydrides of titanium.

Figure 43

Figure 44

Next Chapter (Chapter 10) Previous Chapters (Chapters 6-8) Table of Contents

This page is maintained by Prof. Stacy Gleixner.  
Questions or problems please send email to gleixner@email.sjsu.edu or call (408)924-4051. 
The page was last updated 09/27/00 .

Some files on this site are in .pdf format. To read these you will need the Adobe Acrobat Reader.