5.5 Surface characterization of laser-treated samples
A close inspection of the grainy surface after the laser treatment and washing verifies the existence of bound ceramic particles. No pure metal surface can develop such a modification under laser irradiation (compare Fig. 1). Individual particles side by side are visible with the optical microscope and the SEM. Figure 5 illustrates the possibility to use the fixation method for covering large surface areas by moving the sample through the laser beam. Here parallel lines of fixed ceramic particles were produced. The left side of the image shows the GB14 surface after the laser treatment with a fluence slightly above 0.4 J/cm2. Along the centre of the laser beam path the maximum intensity causes some ablation in the ceramic powder, visible as dark lines in this microscopic image. The sample under the microscope was illuminated from the left side, emphasizing the partial roughness of the original surface. Nevertheless, the right side of Fig. 5 shows the fixed ceramic powder after the excess GB14 was washed away. Fixed grains are bright in front of the darker metal. The bound particles seem to be fixed with a very constant density along the lines. The regular density variation perpendicular to the lines is due to an incomplete overlap of the parallel lines of the laser irradiation. A reduction of this distance results in an even higher homogeneity.
Another example clearly shows how simple the fixation method works. Figure 6 depicts a sample which was only covered in the upper part with dip-coated Ca10. The lower part is pure metal. All images show the sample after it was moved through the laser beam (vertical movement) and after washing. From the left to the right only the magnification of the images is enhanced. The fluence near the ablation threshold for the ceramic first caused a strong ablation in the Ti6Al4V (Fig. 6a,b,d,e) which has a lower ablation threshold. Where the laser beam touched the edge of the ceramic layer, the metal ablation was locally enhanced (Fig. 6b,c) serving as a mark here. Further following the trace of the laser beam upwards, a streak of fixed Ca10 grains is visible.
The fixation works best at the laser fluence, as close as possible, but slightly below the ablation threshold of the metal. This is strong evidence for the assumption that a local melting of the metal binds the particles.
Figure 7 contrasts the Ti6Al4V without and with fixed Ca10. It is the same sample as in Fig. 6. The two pictures from the optical microscope (Fig. 7a,b) and the two SEM images (Fig. 7c,d) have the same scale, respectively. The remaining roughness of the grinded Ti6Al4V plates (Fig. 7a,c) does not disturb the Ca10 fixation. The grains with a diameter in the range of 1 - 2 µm are big enough to homogeneously cover the small irregularities of the metal (Fig. 7b,d). The lower part of the SEM image (Fig. 7d) with the bound ceramic includes the previous coating edge, proving the direct and very abrupt transition between metal ablation and ceramic fixation.
The four pictures with different magnification in Fig. 8 present another sample of fixed Ca10, partly with a higher resolution compared to Fig. 7. It is clearly visible that the Ca10 particles are sharp-edged to make clear that their shape and structure was not changed by the laser treatment. This becomes even more plausible when taking into account that during the irradiation these fixed grains were still shielded by other particles above. The top layer of up to 40 µm thickness is not ablated, not melted, and possibly remains unaltered. Thus the bottom layer is most likely unharmed, too. The radiation reaching the grains at the metal surface travelled through a layer as thick as at least 20 times the diameter of the average particle. A slightly different perspective of the same sample as in Fig. 8 is shown in Fig. 9, only tilted by 40 degrees from the normal. The perspective view underlines the remaining roughness of the surface, a property regarded as supportive for cell growth as desired for bone implants.
The necessary proof that the chemical composition of the fixed particles is unaltered can be seen in the measurements of energy dispersive X-ray spectroscopy (Symietz, 2011). Figure 10 is an example of an energy dispersive X-ray spectroscopy (EDX) measurement, integrated over the complete GB14 surface shown in the picture inset. Signals from the metal are also detected and are due to the existence of only a “monolayer” of ceramic particles bound to the Ti6Al4V. But the ratio between the counts from the ceramic elements supports qualitatively an unchanged GB14 composition.
So far one can assume that the proper material is fixed unharmed. Another most important question needs to be answered concerning the unknown bonding strength. Ultrasonic cleaning shows that the particles are fixed quite well. An additional test is the scratching of the fixed grain layer. A diamond tip scratching under defined forces over the surface of a sample of Ca10 fixed to Ti6Al4V was performed. Figure 11 presents pictures from the optical microscope of three traces of the diamond tip. The mechanical forces were one, five and ten Newton. One can clearly recognize the fixed grains as little dark spots. They do not disappear inside the visible trench the diamond tip has cut at the two lower forces. At F = 10 N though, there seems to be a very shiny and smooth track dug deep into the metal. One is led to believe from this picture that the Ca10 grains are removed from the metal. EDX measurements prove that this is wrong. There is a clear EDX signal of all the elements Ca10 consists of, not just next to the scratch but everywhere inside the seemingly polished trace. In the optical microscope just the topography is seen, the chemistry is invisible to the eye. The fact that Ca10 still exists in all the scratch lines shows that the particles are overrun by the diamond tip and pressed inside the metal. The ceramic grains must remain intact, because if they had been destroyed and spread along the direction the tip is pulled, a linearly smeared out and stretched signal from the debris would appear. Instead the EDX count distribution is the same inside and outside the scratch lines. If and how much the particles are overrun and partly covered by the metal cannot be seen directly. A thin covering of metal over the ceramic would go undetected by EDX because of the depth sensitivity. The most important message from this test is that such a strong mechanical friction does not remove the fixed particles from the metal.
Another experiment led to more proof of the high binding strength. It is intended to look at the contact line between the fixed grains and the metal more closely. The substances were not suitable for a polishing procedure. Too much material was smeared over the surface, disturbing the delicate transition zone. Instead, the freeze fracturing of a sample resulted in interesting SEM images. Figure 12 presents a sample, actually the very same one as depicted in Figs. 8 and 9, but here it was additionally covered electrolytically, first with a very thin gold layer, followed by a thick layer of nickel. This protective cover layer was meant to stabilize the thin Ca10 layer for breaking the sample plate. The sample was made as thin as possible by producing a saw cut from the side of the titanium alloy. Immediately after cooling in liquid nitrogen, the sample was broken at the predetermined breaking point. The Ti6Al4V looked distorted at the edges, caused by the mechanical force during the break. This made a detailed study of the surface impossible. Instead the Ca10 grains are still bound to the Ti6Al4V under some big portions of the nickel cover that had come off and remained high above the ceramic particles. The striking similarity to Fig. 9 shows that the direct cover of gold and nickel, grown in contact to the Ca10 particles and unintentionally being bent upwards, did not pull the Ca10 away from the titanium alloy. The fixation to Ti6Al4V was stronger than the binding to the new nickel cover.
All these results show that the calcium alkali phosphate powder can be fixed to Ti6Al4V, and that the fixation is very strong. It remains to be proven that the titanium alloy itself does not lose its mechanical stability. This will be documented below in section 5.6.