WO2003080140A1

WO2003080140A1 - Use of bioactive glass for air abrasive blasting of implant surfaces

Info

Publication number: WO2003080140A1
Application number: PCT/GB2003/001223
Authority: WO
Inventors: Richard James Cook; Larry Leroy Hench; Timothy Frederick Watson; Ian Dennis Thompson; Paul Derek Robinson
Original assignee: Ic Innovations Limited; King's College London
Priority date: 2002-03-22
Filing date: 2003-03-21
Publication date: 2003-10-02
Also published as: AU2003214436A1; GB0206838D0

Abstract

The present invention relates to a method for treating a metal surface of a medical implant which comprises contacting the metal surface with a bioactive glass using an air abrasion system.

Description

USE OF BIOACTIVE GLASS FOR AIR ABRASIVE BLASTING OF IMPLANT SURFACES.

The present invention relates to the use of bioactive glass as an air abrasive agent for treating metal surfaces, in particular metal surfaces which form part of a surgical or dental implant, and to implants so treated. Bioactive glasses such as Bioglass^® are well known in restorative surgery and dentistry for their proven efficacy in bone regeneration. They represent a class of synthetic materials that react in the presence of body fluids to enhance the body's ability to regenerate tissue and heal itself. Bioactive glasses can be processed into many forms but are most commonly produced as granules or a fine powder. The excellent biological behaviour, superior corrosion resistance under physiological conditions and acceptable modulus matching of titanium justify its status as the material of choice for load bearing medical implants.

Due to the reactive nature of titanium and its alloys, a native passivating oxide layer about 2-5 nm thick is formed spontaneously when exposed to air, water or an aqueous biological environment. The gaseous oxidation series in Titanium proceeds according to the following scheme: Ti + O → Ti(O) → Ti₄O → Ti₃O → Ti₂O → TiO → Ti₂O₃ → Ti₃O₅ → TiO₂. The end product in this series, TiO₂, exists in three structural forms: anatase, brookite and rutile. The TiO₂ (rutile) form is the most stable thermodynamically and is considered to be the usual protective oxide covering titanium and its alloys. Figure 1 illustrates the non stoichiometry of the titanium surface oxide.

Dental implants provide an abutment for the support of a prosthetic tooth. Endosseous implants, usually threaded, cylindrical commercially pure titanium (c.p Ti), are those which are inserted into mandibular or maxillary bone.

Clinical evidence demonstrates that topographic modification of commercially pure titanium implants leads to improved bone behaviour at endosseous implants (Cochran DL, "A Comparison of Endosseous Dental Implant Surfaces" Journal of Peridontology 70 1523 - 1539 1999). In a study Bowers et al found that significantly higher levels of cellular attachment were found using rough, sandblasted surfaces with irregular morphologies (Bowers KT, Keller JC, Randolph BA, Wick DG, Michaels CM, "Optimisation of Surface Micromorphology for Enhanced Osteoblast Responses In Vitro" International Journal of Oral Maxilliofacial Implants 7 302-310 1992). This supports the use of roughened surfaces at bony contact areas. From a clinical point of view, rough titanium surfaces offer significantly better anchorage in bone with shorter healing periods and the option to utilise shorter implants with a good long term prognosis (Schenk RK, Buser D "Osseointegration: A reality" Peridontology 2000 17 23-35 1998). Accordingly, current implant designs employ microtopographically roughened surfaces with macroscopic grooves, threads, or porous surfaces to provide sufficient bone ingrowth for mechanical stabilisation and the prevention of detrimental micro-motion. Grit blasting aims to enlarge the contact area between the implant and the bone in order to optimise interaction with the bioenvironment. The implant surface is bombarded with particles of aluminium oxide or titanium oxide, and by abrasion, a rough surface is produced with irregular pits and depressions. Roughness depends inter alia on particle size, time of blasting, pressure and distance from the source of particles to the implant surface.

It is current practice to achieve a high surface roughness by grit blasting with Al O₃.

However, Al₂O₃ particles may become embedded in the metal which results in detrimental micro-motion around the implant. Also, release of aluminium ions can poison the local area leading to a fibrous capsule formation and ultimately failure. Accordingly, implants are usually acid etched to remove the Al₂O once the desired roughness has been acquired.

Grit blasting with TiO₂ is one way to change surface roughness without influencing the surface chemistry of titanium implants. However TiO₂ blasted surfaces are not as rough as those blasted with Al O₃. Consequently TiO₂ basted surfaces have a smaller contact area between the implant and the bone.

We have now found that using bioactive glass to roughen the metal surface of medical implants increases the surface area of the metal thus optimising the interaction between the bone and the implant without the disadvantages of the prior art. Moreover, benefits are provided in terms of the stimulating tissue growth properties of bioactive glass.

Accordingly the present invention provides a method of treating a metal surface which comprises contacting the metal surface with a bioactive glass using an air abrasion system.

In a further aspect, the present invention provides the use of a bioactive glass as an air abrasive agent for treating metal surfaces. Preferably the metal surface forms part of a surgical or medical implant such as hip stems, knee joints, dental implants, fracture fixation devices, elbow joints and shoulder joints.

Most preferably the metal surface forms part of a dental implant, in particular an endosseous dental implant. Preferably the metal surface comprises titanium or an alloy thereof. Preferably the metal surface comprises is essentially pure titanium or Ti6A14V.

Preferably bioactive glass impregnates the metal surface of the implant during abrasion.

Thus the present invention is based on the observation that when applied through an appropriate air abrasion system bioactive glass can roughen the metal surface of a medical implant to a greater extent than TiO and to the same extent as alumina without the associated deleterious effects of alumina. Thus the present mvention obviates the need for acid etching of blasted surfaces. Moreover, bioactive glass may be impregnated in the titanium surface thereby stimulating growth of the bone.

Brief Description of the Figures: Figure 1 is a schematic diagram illustrating the range of oxygen containing titanium oxide.

Figure 2 illustrates SEM image of untreated, non-imersed c.p Ti surface.

Figure 3 illustrates SEM image of untreated, imersed c.p Ti surface.

Figure 4 illustrates SEM image of Al₂O₃ blasted, non-imersed c.p Ti surface.

Figure 5 illustrates SEM image of Al₂O₃ blasted, imersed c.p Ti surface. Figure 6 illustrates SEM image of Bio glass^® blasted, non-imersed c.p Ti surface.

Figure 7 illustrates SEM image of Bioglass^® blasted, imersed c.p Ti surface.

Figure 8a illustrates an oblique plot of the results of a white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer) analysis of an untreated c.p Ti surface. Measurement area 294 mm x 220 mm. Figure 8b illustrates an oblique plot of the results of a white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer) analysis of an untreated c.p Ti surface, immersed in SBF for 48h. Measurement area 294 mm x 220 mm. Figure 9a illustrates an oblique plot of the results of a white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer) analysis of a Al O₃ blasted c.p Ti surface. Measurement area 294 mm x 220 mm.

Figure 9b illustrates an oblique plot of the results of a white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer) analysis of a Al₂O₃ blasted c.p Ti surface after immersion in SBF for 48 hours. Measurement area 294 mm x 220 mm.

Figure 10a illustrates an oblique plot of the results of a white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer) analysis of a Bioglass^® blasted c.p Ti surface. Measurement are 294 mm x 220 mm. Figure 10b illustrates an oblique plot of the results of a white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer) analysis of a Bioglass^® blasted Ti Surface after immersion in SBF for 48 hours. Measurement area 294 mm x 220 mm.

Figure 11 shows the macro-topography of a machined c.p Ti implant screw thread.

Figure 12 shows where on the commercial implant the white light interferometer (Zygo) measurements were taken.

Figure 13a illustrates an oblique plot of the results of a white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer) analysis of a machined titanium screw implant. Measurement area 294 mm x 220 mm.

Figure 13b illustrates an oblique plot of the same area as in figure 13a but corrected for the curvature.

Figure 14a illustrates an oblique plot of the results of a white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer) analysis of a TiO₂ blasted titanium screw implant. Measurement area 294 mm x 220 mm.

Figure 14b illustrates an oblique plot of the same area as in figure 14a but corrected for the curvature.

Figure 15 depicts a low magnification FJB SIMS image of a Bioglass^® blasted c.p Ti surface after immersion in SBF for 48 hours.

Figure 16 depicts the image of figure 15 at a higher magnification. Figure 17 shows a FIB SIMS image of an ion beam milled cross section of a Bioglass^® blasted c.p Ti surface after immersion in SBF for 48 hours.

Figure 18a illustrates the region over which elemental mapping was carried out on a Bioglass^® blasted c.p Ti surface. Figure 18b illustrates the FIB SIMS elemental mapping for titanium carried out over the region of figure 18a.

Figure 18c illustrates the FIB SJJVIS elemental map for calcium carried out over the region of figure 18 a.

Figure 18d illustrates the FJJB SIMS elemental mapping for sodium carried out over the region of figure 18 a.

Figure 19a shows a FIB SIMS image of an ion beam milled cross section of a Al₂O₃ blasted c.p Ti surface.

Figure 19b shows a FIB SJJVIS elemental map of titanium over area of figure 19a.

Figure 19c shows a FIB SIMS elemental map for aluminium over area of figure 19a. The term "bioactive glass" as used herein refers to a glass or ceramic material comprising Si-oxide or Si-hydroxide which is capable of developing a surface calcium phosphate/hydroxy-carbonate apatite layer in the presence of an aqueous medium, or at the interface of body tissues and the glass, so producing a biologically useful response.

Bioactive glasses suitable for use with the present invention include the silicon based bioactive glasses derived from the Sol-Gel process (Hench LL., West JK., 1990, The Sol-gel Process, Chem. Reviews, 90, 33-72) or the Melt process (Hench LL., Wilson J., 1993 Introduction to Bioceramics. Publisher : World Scientific).

Although it may be possible for a bioactive glass lacking a source of calcium or phosphorus to generate an apatite layer in vivo by utilising endogenous sources of these ions, typically a bioactive glass will comprise a source of at least one of calcium or phosphorous in addition to a source of Si-oxide or Si-hydroxide. Typically the bioactive glass will comprise a source of calcium. Optionally the bioactive glass may contain further hardening and/or softening agents. Such softening agents may be selected from: sodium, potassium, calcium, magnesium, boron, titanium, aluminum, nitrogen, phosphorous and fluoride. Additions of sodium, potassium, calcium and phosphorus are most commonly used, to reduce the melting temperature of the glass and to disrupt the Si networks within it. Optionally, hardening agents such as TiO may be included in the glass composition. Its presence would allow crystallization to occur within its structure, so producing a glass - ceramic material, whose hardness will be greater than that of the glass alone. Thus, composition ranges for bioactive glasses which may be used with the present invention are as follow:

SiO₂ or Si(OH)₂ : 1-100%

CaO : 0-60%

P₂O₅ : 0-60% Na₂O : 0-45%

K₂O : 0-45%

MgO : 0-40%

Plus additions of Na, K, Ca, Mg, B, Ti, Al, P, N and F as necessary.

Preferably, a bioactive glass will contain between 30 and 100 % Si-oxide or Si- hydroxide, more preferably between 40 and 85 %.

In a further preferred embodiment the bioactive glass will contain between 5 and 60 % Ca, more preferably between 30 and 55 %.

With respect to a source of phosphorus, the bioactive glass will contain between 5 and 40 % P, more preferably between 10 and 30 %. Thus, in one embodiment the bioactive glass will comprise SiO₂, CaO and P₂O₅.

Preferably the bioactive glass includes from 44 to 86 weight % SiO₂, from 4 to 46 weight % CaO and from 3 to 15 weight % P₂O₅. Preferably the bioactive glass is prepared by the sol gel route and comprises from 55 to 86 weight % SiO , from 4 to 33 weight % CaO and from 3 to 15 weight % P₂O5. Preferably such a bioactive glass has the composition 58 weight % SiO₂, 33 weight % CaO and 9 weight % P₂O₅.

In an alternative embodiment the bioactive glass composition may be prepared by the Melt method such as that described in US 5,981,412. Such a glass may have a composition of from 40 to 51 weight % SiO₂, 23 to 25 weight % CaO, 23 to 25 weight % Na₂O and 0 to 6 weight % P₂O₅. Preferably such a bioactive glass has the composition (by weight); SiO₂ - 45%

NaO₂ - 24.5%

CaO - 24.5%

P₂O₅ - 6%. Such a bioactive glass is available commercially as Bioglass^® 45S5.

The manufacturing and processing methods used in the silicon based bioactive glass family are ideally suited to the production of tailored particles for use with the present invention.

Typically, alumina has a Vickers Hardness of 2000 to 2300. Where as commercially available bioactive glasses have Vickers Hardness in the range of 400-500 Bioglass^® 45S5 has Vickers Hardness of 458+/-9.4. As mentioned above, hardening (and softening) components may be added to modulate the hardness of the bioactive glass. Thus, either by selecting from known bioactive glasses or by varying the amounts of hardening agents the skilled man will be able to prepare bioactive glass air abrasive agents suitable for use with the present invention.

Further, by controlling the processing conditions in the densification phase of the sol gel process (Hench LL., West JK., 1990, The Sol-gel Process, Chem. Reviews, 90, 33-72.

Hench LL., West JK., 1996, Biological applications of Bioactive glasses, Life Chemistry

Reports, 13, 187-241) sol-gel variants of bioactive glass can be processed to differing densities and ultimate strengths and hardnesses.

Thus, by increasing the quantity of network modifier (non - silica species species, eg Na, K, Ca, Mn, Br, Al, N, P, FI etc) the hardness of the finished glass decreases. These modifiers may be added to the melt derived glasses while in their molten states, or to sol-gel materials at the mixing phase of production. Hardness may also be decreased by increasing the porosity within the glass, achieved by variations in the drying and stabilisation and densification phases of the sol-gel process. As described above, the hardness of glasses can be increased by allowing crystal formation within them, so the use of TiO₂ can act as a hardening agent, as the glass becomes a glass ceramic. Also modifications to the sol-gel processing phases allowing a more dense glass product will result in a harder product. The strength required of the bioactive glass will depend on the metal to be treated.

Moreover, for some metals, such as titanium and alloys thereof, which, as mentioned previously, have a layered structure of different forms, bioactive glasses of different hardnesses may be required in treating the same sample. Typically a sutiable bioactive glass will have a Vickers Hardness of at least 50, preferably at least 300.

A further consideration when preparing a bioactive glass for use in the present invention is the shape of the bioactive glass particles. The shape of bioactive glass particles may be controlled by selecting the appropriate particulation process from, for example, grinding, crushing or air-collision milling during their manufacture. Thus, crushing produces sharper angulated particles, whereas, air collision milling will produce more rounded particles. Grinding (e.g. ball milling) however, will produce particles of a more intermediate shape. These processes being suitable for glasses produced by both the sol-gel and melt routes.

Particles most suitable for use in the present invention will have a diameter in the range of 1 μm to 1mm, more preferably in the range of 5μm to 500μm, most preferably in the range lOμm to lOOμm.

The present invention may be used with conventional air abrasion systems well known to those skilled in the art. Examples of suitable air abrasion systems include the Velopex® Alycat marketed by Medivance Instruments Ltd., which permits switching the source of the abrasive agent during operation. The use of other gases as a propellant (eg CO₂ or N₂) is included in the definition of "air abrasion" and the use of water or other fluids to act as dust supression agents (regardless of potential contribution to the overall cutting effect) are also included, however delivered - either included in the gas stream or entrained around it (e.g. The Aquacut air abrasive machine - Medivance Instruments Ltd, Harlesden, London). It is to be understood that the present invention covers all combinations of suitable and preferred groups described hereinabove.

The present invention will now be illustrated, but is not intended to be limited, by means of the following examples.

Examples A rod of commercially pure titanium was turned on a lathe down to a diameter of 8 mm from which 6 discs of 1.5 mm thickness were cut. Two of the samples were left untreated, two were grit blasted with Al₂O₃ powder with a particle size of 53 μm and two were blasted with 45S5 Bioglass^® powder, composition (in weight per cent) 45% SiO₂, 24.5% Na₂O, 24.4% CaO, 6% P₂O₅, with a particle size of 20-90 μm. A Velopex Alycat Aluminium Oxide Cutting System driven on dry atmospheric air was used. The powder was delivered at 5 MPa pressure at a constant feed rate. The samples were blasted for a time of one minute and a consistent grey coverage was achieved.

One of each type of disc was immersed in 50 ml of simulated body fluid (S.B.F) for 48 hours, and maintained at a constant temperature of 37.5°C in a New Brunswick Scientific C24 Incubator Shaker. The remaining, non-immersed discs were retained as controls. The ionic composition in mM of S.B.F is given in Table 3. The solution has a pH of 7 to represent conditions in the body. At the end of the 48 hour period, samples were removed from the solution and dried in air at 60°C for 20 minutes.

Table 3. Ionic composition, in mM of SBF Surface Morphology

The surface morphology of the samples was evaluated qualitatively by scanning electron microscopy (SEM) using a Jeol JSM T200 Scanning Microscope. Standardised conditions of 25keV accelerating voltage and a working distance of 20mm were used.

A white light interferometer (Zygo Newview 100 3D imaging Surface Analyzer), was used to investigate the surface topography of the samples quantitatively. This non contact technique with high vertical resolution (1 nm) allows 3D image data to be produced and is suitable for use in the topographical characterisation of implants. The principle of operation involves the generation of white light fringes and automatic image analysis converts information from these fringes into measurements of height. Three surface parameters were measured: Ra, peak to valley height and the rms value corresponding to Ra, where:

Ra is the arithmetic mean of the departures of the roughness profile from the mean line and is measured in micrometres; rms is the root mean square parameter corresponding to Ra, in micrometres; and

PV is the maximum peak to valley height in the measured area, also measured in micrometres. For each of the samples, 3 values of Ra, peak to valley height and of rms were obtained with a measured area of 294 mm x 220 mm in each case.

Chemical Surface Analysis

Semi quantitative energy dispersive spectra (ED AX) were obtained using a Jeol JSM T200 Scanning Microscope.

Secondary Ion Mass Spectroscopy was carried out with a Perkin Elmer Atomika 6500 Ion Microprobe. A Xenon beam was used with a discharge current of lOOnA. The scan width was 800 mm and the sweep time was 0.5 seconds. Both positive and negative mass to charge ratio spectra were obtained for all samples. Further study was carried out on two of the samples using the Focused Ion Beam

Secondary Ion Mass Spectroscopy (FIB-SIMS) technique. The instrument used was the FEI Focused ion beam workstation with SIMS. SIMS imaging is a direct method of chemically mapping the distribution of materials with high sensitivity, extreme surface specificity and full range of elemental and molecular information available from the SIMS technique (Vickerman JC, Brown A, Reed NM (editors) SIMS Principles and Applications Oxford Science Publications, 1989). Elemental maps of the surface of the two samples and of a cross section, produced by ion beam milling into the surface, were obtained.

Zygo analysis and SIMS were also carried out on commercial implants, supplied by Astra Tech. Two types of threaded c.p Ti implant, with a machined surface and a roughened surface respectively, were studied. The roughened implants had been previously grit blasted with c.p Ti powder. Samples were received in sterile individual glass vials containing ethylene oxide. The machined samples had diameter 4.0 mm and length 19mm and the roughened samples 4.0 mm diameter and 13 mm length. The unthreaded ends of the implants were selected for study for ease of access. Results

Surface Morphology

On removal from the S.B.F it was observed that the control sample and the sample blasted with Al₂O₃ appeared unchanged but that the sample blasted with Bioglass^® appeared lighter in comparison with the equivalent sample not immersed. SEM analysis of the samples revealed a stark difference in the appearance of samples not blasted and those samples blasted with either Al₂O₃ or Bioglass^®, as can be seen by comparing Figures 2 and 3 with Figures 4, 5, 6 and 7. The control samples have a streaked appearance, consistent with their preparation as they were cut on a saw and no further polishing was carried out. The slightly raised feature in Figure 3 is thought to be a surface irregularity rather than any deposit on the surface. The samples blasted with both Al O₃ and Bioglass^® show a rough irregular topography with numerous randomly oriented sharper features and protrusions from the surface. On visual inspection, it appears that samples blasted with Al₂O₃ and samples blasted with Bioglass^® have an equivalent qualitative surface roughness.

The results of the Zygo analysis for the experimental samples are summarised in Table 4. Ra is the value most commonly referred to and most frequently quoted in the literature. The results for peak to valley height and for rms follow a similar trend to that for

Ra. It must be noted that maximum peak to valley height is an extreme value and as such may have large scatter even on the same sample. As can be seen in the table, the Ra value of the c.p Ti has been increased from a value of 0.27 μm in the case of the untreated control to 1.30 μm in the case of Al₂O₃ and 1.26 μm in the case of the Bioglass^®. Thus it is concluded that blasting with Bioglass^® particles appears to roughen the c.p Ti surface to an equivalent quantitative degree as the established method of blasting with alumina particles. Wennerberg et al (Journal of Biomedical Materials Research 30 251-260 1996) propose that an Ra value of 1.1 μm to 1.4 μm is particularly desirable, based on experimental studies. This invention presents surfaces blasted with Bioglass^® particles that lie within this range.

Table 4. Zygo results for experimental samples The oblique plots for the measured area for each of the three surface conditions prior to and after immersion in SBF are shown in figures 8 to 10. The data shown are from one of three replicate measurements, from which means were taken.

Again, significant differences in surface morphology can be seen. The untreated samples are relatively flat and both types of blasted sample exhibit sharply defined edges. It should be noted that the y axis scales differ slightly in these figures.

The effect of soaking in SBF for 48 hours is not clear and seems to depend on the surface treatment. In the case of the untreated samples, the Ra has increased slightly.

However in the case of the alumina blasted samples the Ra has decreased and for Bioglass^® Ra remains virtually unchanged after soaking, as can also be seen graphically by comparing

Figures 10a and 10b.

Figure 11 shows the macro-topography of the machined c.p Ti implant screw thread. The commercial implants were also analysed using the Zygo instrument. Measurements were made in several places: the flat end of the tip, the unthreaded head of the screw, the thread peak and the thread trough as indicated in Figure 12.

The surface topography parameters measured in these regions for the machined and TiOBlast™ implants are given in Table 5. Again, by comparing Ra values it can be seen the TiO₂ blasted implant has a rougher surface than the machined surface in all the regions studied. There is a small amount of variation of roughness depending on the region within the two samples, with the peak being the roughest region in the TiOBlast™ sample (Ra = 0.86 μm) and the tip being the roughest in the machined sample (Ra = 0.39 μm).

Table 5. Zygo results for Astra Tech Dental Implants Examples of the oblique plots for the implants are shown in Figures 13 and 14. They show visually the numerical differences in surface roughness in the two types of implant. Due to the geometry of the implant, it was necessary to correct for the curvature of the threads to obtain meaningful values for the peaks and the troughs. This was done be choosing 'remove cylinder' in the Zygo software. The effect of this can be seen by comparing Figure 13(a) with 13(b), and 14(a) and 14(b), in the case of the peak.

Thus an Ra value of around 0.8 μm was measured for the TiO blasted surface, depending slightly on the position of the measurement on the sample. In comparing the Zygo results of the commercial implants with the experimental samples, it is evident that the TiO₂ blasted surface is not as rough as the Al₂O₃ blasted or the Bioglass^® blasted surface (Ra =

1.31 μm and 1.26 μm respectively).

Chemical Analysis

EDAX analysis revealed the elemental composition of the various surfaces. The untreated control sample was shown to comprise solely of titanium, with Ti and Tiβ peaks at 4.5 and 4.9 keV respectively. The non-treated, immersed sample showed a small amount of chlorine and trace amounts of silicon in addition to the titanium peaks. It is thought that this chlorine could have come from the SBF solution, or could be contamination.

The samples blasted with Al₂O₃ both showed, as one would expect, peaks corresponding to titanium and aluminium, at 1.5 keV. In addition, for the sample immersed in SBF, chlorine and silicon were also detected. Thus the samples blasted with Al₂O₃ showed that alumina particles remained on the surface, effectively contaminating the surface since once the particles become embedded, they are difficult to remove (Papadopolous T, Tsetsekou A, Eliades G "Effect of Aluminium Oxide Sandblasting on Cast Commercially Pure Titanium Surfaces" Eur. J. Prosthodont, Rest. Dent, vol 7 no 1 15-21 1999). In the Bioglass^® blasted, non-immersed sample, peaks corresponding to titanium, silicon, and calcium were observed. After immersion, the silicon peak appeared smaller and the presence of chlorine and phosphorous was detected alongside a marginal increase in the amount of calcium.

SIMS The main findings from the S S analysis were that, as one would expect, Ti, TiO and TiO₂ were found on all samples. The untreated, immersed sample had a similar spectrum to the untreated, non-immersed sample with the addition of Mg⁺, Ca⁺ and CaO, as well as an apparent increased count of Na and K⁺, although this is not directly quantifiable. The appearance of these species is thought to be a result of soaking in SBF. It was found that Na⁺ and K were over-represented on all samples, probably due to the higher ionisation 5 probabilities of these elements. The negative species spectra were very similar for both samples, with no notable differences. There was a high count for chlorine, as for all the samples studied. This may be explained by the high ionisation probability for chlorine, and possible contamination.

In the alumina blasted samples, aluminium and oxides of aluminium were well 10 represented. In addition, CaO⁺, Mg⁺ and H₃O were observed on the immersed samples. Thus, although the alumina blasted sample proved to be problematic in terms of sample charging this study does indicate aluminium is still present on the surface after immersion, implying that alumina particles have indeed become embedded in the c.p titanium surface. Figure 19a shows an ion beam milled cross section and figures 19b and 19c show the L5 elemental mapping for titanium and aluminium respectively for this section. No significant amounts of calcium were found

The following species were found on the Astra Tech machined implant: Ti, TiO and TiO₂, as can been seen in Appendix 2(b). Difficulty was encountered for the TiO₂ implant due to sample charging.

Ϊ0 On the Bioglass^® blasted samples, Ca⁺, SiO₂, CaO⁺, and Na O were observed, suggesting that some quantity of Bioglass^® had remained on the surface after blasting. However Figure 15 shows a SJJVIS image of the surface of the Bioglass^® blasted sample after immersion in SBF. Noticeable features observed were voids with dimensions comparable to Bioglass particle size (-20 μm). The same surface at closer magnification can be seen in

'5 Figure 16, where the roughness becomes more apparent. A possible explanation for the voids is particle break up on blasting and subsequent falling out, maybe after immersion. It is interesting to note that the immersed sample had slightly less silicon than the non-immersed sample in the ED AX analysis. Following immersion in SBF, there was an apparent increase in Ca⁺, although the amount of P^" was largely unchanged. H₃O was also observed on the

0 immersed sample, suggesting ion exchange had taken place.

Figure 17 shows a cross section produced by ion beam milling where the bright oxide layer is contrasted with the duller metallic bulk. Elemental mapping was carried out for titanium, calcium, sodium and silicon on the section depicted in Figure 18a. The images for titanium, calcium and sodium can be seen in Figures 18(b) to (d). Bright white regions on the images represent a high count for the particular element. The results indicate that the cross section is metallic titanium all the way up to the surface, where a very thin layer of TiO₂ is observed. In Figure 18c the calcium was observed on raised parts of the surface and may be present either as a result of the Bioglass^® blasting, or as a deposit from the SBF, as a precursor to HCA. There was no evidence of a silicon layer following blasting with Bioglass^®, elemental mapping revealed an insignificant amount of silicon on the surface.

After 48 hours immersion in SBF, one may have expected little evidence of hydroxyapatite (HA) formation on the untreated control samples, even less on the alumina blasted samples, as aluminium is detrimental to nucleation of HA, and a hydroxyapatite layer up to 1 μm thick on the surface of titanium blasted with Bioglass. In fact, ED AX analysis of c.p Ti after grit blasting with Bioglass^® indicated that only marginally more Ca and P were present on the surface after immersion in SBF for 48 hours suggesting that the precursors for HCA may had only just begun to mineralise.

Conclusion

Grit blasting with Bioglass^® has been found to roughen the surface of commercially pure titanium to the same extent, both qualitatively and quantitatively, as the established method of Al₂O₃ blasting and to a greater extent than TiO₂. However, despite the observance of Ca⁺, SiO₂, CaO⁺, and Na₂O on the Bioglass^® blasted samples before immersion in SBF, Bioglass^® particles do not appear to have become embedded in the surface of titanium, whereas alumina particles were found to have become embedded in the samples blasted with Al₂O₃. ED AX analysis of c.p Ti after grit blasting with Bioglass^® indicated that only marginally more Ca and P were present on the surface after immersion in SBF for 48 hours suggesting that the precursors for HCA may had only just begun to mineralise.

Claims

1. A method of treating a metal surface which comprises contacting the metal surface with a bioactive glass using an air abrasion system.

2. The method of claim 1 wherein the metal surface forms part of a medical implant

3. The method of claim 2 wherein the medical implant is a dental implant, preferably an endosseous dental implant.

4. The method of any of claims 1 to 3 wherein the metal comprises titanium or an alloy thereof.

5. The method of claim 4 wherein the metal is essentially commercially pure titanium.

6. The method of claim 4 wherein the metal is essentially Ti6A14V.

7. The method of any preceding claim wherein the bioactive glass becomes impregnated in the metal surface of the implant.

8. The method of any preceding claim wherein the treatment causes roughening of the metal surface.

9. The method of any preceding claim wherein the bioactive glass comprises a source of SiO or Si(OH)₂, and a source of CaO₂ and/or P₂O₅.

10. The method of claim 9 wherein the bioactive glass further comprises at least one hardening agent and/or at least one softening agent.

11. The method of claim 10 wherein the softening agent is selected from Na, K, Ca, Mg, B, Al, P, N, F and the hardening agent is TiO₂.

12. The method of any of claims 9 to 11 wherein the bioactive glass comprises 1 to 100% SiO₂ or Si(OH)₂, 0 to 60% CaO, 0 to 60% P₂O₅, 0 to 45% Na₂O, 0 to 45% K₂O and 0 to 40%MgO.

13. The method of any of claims 9 to 12 wherein the bioactive glass is obtainable by the Sol-Gel method.

14. The method of any of claims 9 to 12 wherein the bioactive glass is obtainable by the Melt method.

15. The method of claim 13 wherein the bioactive glass comprises 44 to 86 weight % SiO₂, 4 to 46 weight % CaO and 3 to 15 weight % P₂O₅.

16. The method of claim 13 wherein the bioactive glass comprises 58 weight % SiO₂, 33 weight % CaO and 9 weight % P₂O₅.

17. The method of claim 14 wherein the bioactive glass comprises 40 to 51 weight % SiO₂, 23 to 25 weight % CaO, 23 to 25 weight % Na₂O and 0 to 6 weight % P₂O₅.

18. The method of claim 14 wherein the bioactive glass comprises (by weight) :

SiO₂ - 45%

NaO₂ - 24.5%

CaO - 24.5%

P₂O₅ - 6%.

19. The method of any preceding claim wherein the bioactive glass has a Vickers Hardness of at least 50.

20. The method of claim 19 wherein the bioactive glass has a Vickers Hardness of at least 300.

21. The method of any preceding claim wherein the bioactive glass particles are substantially non-spherical .

22. The method of any preceding claim wherein the bioactive glass particles are substantially spherical.

23. The method of any preceding claim wherein the bioactive glass particles have a diameter of from lOμm to 500μm.

24. Use of a bioactive glass as an air abrasive agent for treating metal surfaces.

25. Use according to claim 24 wherein the metal surface forms part of a medical implant

26. Use according to claim 25 wherein the medical implant is a dental implant, preferably an endosseous dental implant..

27. Use according to any of claims 24 to 26 wherein the metal comprises titanium or an alloy thereof.

28. Use according to claim 27 wherein the metal is essentially commercially pure titanium.

29. Use according to claim 27 wherein the metal is essentially Ti6A14V.

30. Use according to any of claims 24 to 29 wherein the bioactive glass is impregnated in the metal surface of the implant.

31. Use accordmg to any of claims 24 to 30 wherein the treatment causes roughening of the metal surface.

32. Use according to any of claims 24 to 31 wherein the bioactive glass comprises a source of SiO or Si(OH)₂, and a source of CaO and or P₂O₅.

33. Use according to claim 32 wherein the bioactive glass further comprises at least one hardening agent and/or at least one softening agent.

34. Use according to claim 33 wherein the softening agent is selected from Na, K, Ca, Mg, B, Al, P, N, F and the hardening agent is TiO₂.

35. Use according to any of claims 32 to 34 wherein the bioactive glass comprises 1 to 100% SiO₂ or Si(OH)₂, 0 to 60% CaO, 0 to 60% P₂O₅, 0 to 45% Na₂O, 0 to 45% K₂O and 0 to 40%MgO.

36. Use according to any of claims 32 to 35 wherein the bioactive glass is obtainable by the Sol-Gel method.

37. Use according to any of claims 32 to 35 wherein the bioactive glass is obtainable by the Melt method.

38. Use according to claim 36 wherein the bioactive glass comprises 44 to 86 weight % SiO₂, 4 to 46 weight % CaO and 3 to 15 weight % P₂O₅.

39. Use according to claim 36 wherein the bioactive glass comprises 58 weight % SiO , 33 weight % CaO and 9 weight % P₂O₅.

40. Use according to claim 37 wherein the bioactive glass comprises 40 to 51 weight % SiO₂, 23 to 25 weight % CaO, 23 to 25 weight % Na₂O and 0 to 6 weight % P₂O₅.

41. Use according to claim 37 wherein the bioactive glass comprises (by weight): SiO - 45%

NaO₂ - 24.5% CaO - 24.5% P₂O₅ - 6%.

42. Use according to any of claims 24 to 41 wherein the bioactive glass has a Vickers Hardness of at least 50.

43. Use according to claim 42 wherein the bioactive glass has a Vickers Hardness of at least 300.

44. Use according to any of claims 24 to 41 wherein the bioactive glass particles are substantially non-spherical.

45. Use according to any of claims 24 to 41 wherein the bioactive glass particles are substantially spherical.

46. Use according to any of claims 24 to 41 wherein the bioactive glass particles have a diameter of from 1 Oμm to 500μm.

47. A medical implant obtainable by a method of any of claims 1 to 23.