Characterization of Ti Used for Implants

Jeffrey J. Weimer

Titanium is a widely used implant material. We are characterizing the surface properties of Ti using XPS and AFM. This work involves collaboration with faculty and students in the Department of Biomaterials at UAB.

"Surface Analysis of Titanium Based Implant Materials"
H. Placko, W. Perry, S. Mishra, L. C. Lucas, and J. J. Weimer

The following sections show the layout of the work so far. Sections are being added as the data are evaluated.

AFM Section

Experimental Methods

One set (grit blasted, electropolished, and polished) of each type of sample (commercially pure or alloy) was scanned with an atomic force microscope (AFM) (Nanoscope III, Digital Instruments) in tapping mode using standard etched Si tips with a nominal radius of 5 - 10 nm. Two regions were imaged on each sample. For each region, scans were done in a sequence of sizes ranging from 200 x 200 microns to 0.1 x 0.1 microns. The scan rate was about 0.1 Hz at the largest scan size and about 10 Hz or less for the smallest scan (faster scans were possible on smoother surfaces). All images were taken with a resolution of 512 x 512 pixels. The data resolution normal to the image plane was adjusted to be the best possible under the given scan conditions, accounting for sample tilt and the height of the surface features. The raw images were all fit with a quadratic plane in the x and y directions to remove image bow. Where appropriate, the image was also flattened with a zeroth order line fit to remove scan line anomolies. Roughness values were calculated for each corrected image over the entire image frame. For display purposes, the images were also passed through a low pass filter to remove spurious noise.

Results

Fig. AFMA
Images of commercially pure Ti (Ti CP) and Ti alloy (Ti 64) surfaces at 10 x 10 microns using an AFM. The z scale is 2 microns in all cases. 		Fig. AFMB
Images of commercially pure Ti (Ti CP) and Ti alloy (Ti 64) surfaces at 1 x 1 microns using an AFM. The z scale is 200 nm in all cases.
Type Ti CP Ti 64 Ti CP Ti 64
gb AFM of Ti CP gb sample at 10 microns. AFM of Ti 64 gb sample at 10 microns. AFM of Ti CP gb sample at 1 microns. AFM of Ti 64 gb sample at 1 microns.
ep AFM of Ti CP ep sample at 10 microns. AFM of Ti 64 ep sample at 10 microns. AFM of Ti CP ep sample at 1 microns. AFM of Ti 64 ep sample at 1 microns.
pl AFM of Ti CP pl sample at 10 microns. AFM of Ti 64 pl sample at 10 microns. AFM of Ti CP pl sample at 1 microns. AFM of Ti 64 pl sample at 1 microns.

Representative images obtained from the surfaces with the AFM are shown in Figs. AFMA and AFMB. The images in Fig. AFMA are at 10 microns x 10 microns in scan length and those in Fig. AFMB are for 1 micron x 1 micron scans.
Roughness values for Ti implants. Fig. AFMC
Root mean square (RMS) roughness values for commercially pure Ti (Ti CP) and Ti alloy (Ti 64) samples subjected to various surface treatments. The values are plotted versus the length of one side of the scan region.

The root mean square (RMS) roughness values determined from the scans with the AFM are shown in Fig. AFMC. The values are plotted versus the length of one side of the (square) image frame. This length is proportional to the pixel resolution in the plane of the image according to the formula: image resolution = scan length / 512. The points are the average values for the two regions scanned, and the error bars are determined by the differences between the maximum and minimum values measured. Error bars that do not show are smaller than the size of the point. This representation of the data is not meant to imply that two points are a statistically reliable sample of the roughness value for any substrate. The solid curves in Fig. AFMC show the overall trends in the data for the three sets of samples (grit blasted, electropolished, and polished) where both Ti types (commercially pure and alloy) are considered together. Mean roughness values were lower in all cases than the values in Fig. AFMC, and they followed the same trends. The mean roughness values are not shown for clarity. The roughness values were generally lowered by the off-line flattening process, in some cases by as much 30%. In most cases, the two values measured for each region differed by more than the change caused by flattening, and the error bars and scatter at each point in Fig. AFMC encompassed such variations.

jjweimer@matsci.uah.edu
6.Aug.96