Dental implant materials. 11. Preparative procedures and surface spectroscopic studies D.C. Smith,* R. M. Pilliar, J. B. Metson,’ and N. S. McIntyret University of Toronto, Faculty of Dentistry and ‘Surface Science Western, University of Western Ontario, Canada The tissue response to an implant may involve both physical and chemical factors. There is little reliable information on the effects of these parameters and the associated ionic release on the cell-material interaction because the majority of studies have not fully characterized the implant material. In this work surface spectroscopy using ISS, ESCA, and SIMS was carried out on Ti6A14V Co-Cr-Mo,
A1203, and hydroxyapatite dental implant materials that had been subjected to six commonly used preparative procedures. The results showed that each procedure generated an individualistic composition for the outermost surface of each material. These differences could be significant in cellular and tissue response. Improved understanding of these factors requires defined and reproducible surfaces.
IN T RODUCTION
The tissue response to a dental implant may involve physical factors such as size, shape, surface topography, and relative interfacial movement, as well as chemical factors associated with the composition and structure.’” There is still relatively little reliable information on the influence of these variables because the majority of in vitro and in vivo studies on implant materials and systems has not characterized in a precise manner the bulk and surface characteristics of the materials being used, nor the effects of the particular preparative cleaning and sterilization systems on such characteristic^."^ Further, until very recently, little or no attempt has been made in most studies to evaluate ionic release into the tissues and its relationship to the observed tissue r e a ~ t i o n . ~ These - ~ issues have been discussed in Part I of this series8 and elsewhere .6,7,y,1fl Characterization of implant materials to evaluate tissue reaction necessarily includes analysis of surface composition and contamination.’-13 It is now well understood that the surface composition of most implant materials is substantially different from the bulk, and a variety of techniques have been used to characterize such surfaces.’’ Several such techniques are needed to fully characterize implant surfaces. *To whom correspondence should be addressed at Centre for Biomaterials, University of Toronto, 170 College Street, Toronto, Ontario, Canada M5S 1Al. Journal of Biomedical Materials Research, Vol. 25, 1069-1084 (1991) CCC 0021-9304/91/091069-16$4.00 0 1991 John Wiley & Sons, Inc.
SMITH ET AL.
1070
Such a combined approach is only just beginning to be used in biomateria1s.l" L a ~ s m a a , Kasemo '~ and c o - w ~ r k e r s , ' ~Sundgren,16 ~'~~'~ Ducheyne and cow o r k e r ~ have ' ~ applied ESCA and Auger analysis to the surface characterization of titanium and stainless-steel (dental) implants and found, in common with the present that topographical and compositional changes occur in the surface of such implants as a result of both in vitro treatment and in vivo implantation. We also observedz2in earlier work extensive changes in the surface composition of nickel-chromium alloys after implantation for at least 3 months in animal models. In the present work, four presently commonly used dental implant materials have been subjected to customarily used preparative techniques, and the effects on their surface characteristics investigated. In a previous paper: we described the results of SEM/EDX and surface wetting studies. The present paper describes the results of ISS, SIMS, and ESCA studies on these materials. ISS gives information on the outer 1-2 atomic layers and is especially sensitive to high mass elements such as elements above Ti. The signal amplitude is related to atomic concentrations in the surface and is not usually strongly affected by changes in matrix. In SIMS, all elements are detectable, including H, as well as clusters related to molecular fragments that may indicate the nature of the chemical bonding at the surface. However, the relative instrumental sensitivities to the elements can vary over a range of about five orders of magnitude and can be strongly affected by the matrix. These processes lead to sputter erosion of the surface such that successive spectral scans can give depth profiling information. ESCA provides information from depths as great as 5 nm on the specific chemical binding of atoms for all elements except H and He. In some cases, semiquantitative data (210%) can be obtained. Detailed reviews of these techniques and their biological applications1"" can be found elsewhere. MATERIALS A N D METHODS
The following materials were investigated: (a) Ti6A14V ELI conforming to ASTM F-136, (b) Co-Cr-Mo conforming to ASTM F-75 (Canox Ltd.), (c) singlecrystal A1203(Kyocera, Kyoto) (AO), (d) dense hydroxyapatite (HA) (Durapatite; Sterling Winthrop Laboratories). Chemical analyses of these materials have been given previously.* Discs 3.5 mm in diameter and 1.5 mm thick were fabricated by cutting from rod stock with a low-speed diamond saw. The alloys and HA specimens were finished by wet grinding to 600 grit on Sic paper. The alumina discs were polished by a proprietary process. Similar porous-surfaced alloy discs were prepared by sintering spherical particles of 50-300 pm diameter onto the solid substrate as described elsewhere? For the Co-Cr-Mo alloy powders (Canox Ltd.), a 1300°C, 3-h sinter in uucuo was employed, whereas for the Ti6A14V powder (Dynament Inc.) 1250°C for 1 h in vacuo was used.' The solid discs were subjected to the same heat treatment (simulated sintering process). Fully porous alumina discs were made by a proprietary process (Kyocera Inc.).
DENTAL IMPLANT MATERIALS. I1
1071
After fabrication, the specimens were washed in distilled water and airdried. This was the "as-received condition. Groups of three specimens for each material were subjected to the following cleaning regimens: (1) Retained in the "as-received" condition (2) Washing in an ultrasonic cleaner as follows: (a) Sonicated for 60 min in a 2% solution of Decon detergent (B.D.H Chemicals Canada Ltd.). (b) Then sonicated in deionized water for 2 min, three changes. (c) Air-dried in a closed container. (3) As in (2), followed by steam sterilization by autoclaving at 121°C for 30 min. (4) As in (2), followed by radiation sterilization. (5) As in (2), followed by treatment with 40% vol HNO, at RT (ASTM F-86) for 60 min in the sonicator followed by similar rinsing with deionized water for 3 min, three changes. (6) As in (2), followed by exposure to an argon plasma for 30 min at 0.3 mm Hg (Harrick Model PDC-3XG Plamsa Cleaner). The materials were coated with Au for SEM and carbon for energy dispersive analysis (EDX) in a Polaron sputter coater. Surface analyses were carried out on duplicate samples using ISS, SIMS, and ESCA. ISS and SIMS measurements were made using a 3M-Kratos Model 555 BX combined ISS/SIMS spectrometer with the assistance of Dr. G.R. Sparrow (Advanced R and D Inc.). A surface area of about 6 mm2 was examined for SIMS and about 2 mm2 for ISS. The positive ion beams used were 3Heor 40Arwith a beam energy of 2 keV or, occasionally 4 key At 2 keV the beam current was typically 3-500 mA/cm2 over the irradiated area. ESCA Spectra were obtained using a Surface Science Labs SSX-100 spectrometer mm Hg, using monochromatized A1 K, radiation at a pressure of lO-'-lO-' a small flood gun compensation was applied until satisfactory charge compensation was obtained. Additional spectra were obtained using a similar instrument with the assistance of Dr. B. Ratner (National ESCA Centre for Biological Materials, University of Washington). Most data were obtained using a 1-mm-diameter spot but some spectra were also obtained using the 150-pm-diameter spot capability. Stoichiometry could be obtained from the relative peak areas using the form factors derived by S~hofield?~ modified to take account of the mean free path of the escaping electron. A constant transmission function was assumed for the electron analyzer. Under the present conditions the data can be considered significant only to 210%.
RESULTS
Ti6A14V alloy Sequential ISS broad-scan spectra are shown in Figures 1 and 2 for this alloy surface treated by two different processes: (4) radiation sterilization
SMITH ET AL.
1072
Z
9
L
z z 0
SPUTTERING TIME (SECS) Figure 1. 3He+ ISS sequential rapid scan spectra showing depth profile changes in TiAlV surface ion concentrations as sputtering time is increased. Sputtering rate approximately 0.5 nm/60 s. The surface was treated by process 4 (radiation sterilization).
(Fig. 1) and (5) nitric acid passivation (Fig. 2). The depth profiling indicated by the successive spectra reveals the presence of Pb in the outer 1 nm or so of the surface, more prominently for process 4. The sputter etch rate was approximately 1 nm/150 s as determined from standard calibrated samples. Continued depth profiles showed that the nitric-acid treatment (5) and an argon plasma treatment (6) produced the thickest oxide layers, which, as shown by the expanded ISS spectra, typically contained sodium, aluminum, titanium, and lead oxides. The nitric-acid treatment resulted in nitrogen within the oxide layer. Vanadium was only detected in the oxide layer when SIMS and ESCA were used. As noted in a previous paper: nitric-acid treatment induced pronounced etching. Small-area ESCA analyses of the nitric-acid-treated surface suggested that different compositions existed on the face of the grain and on the intersection of grain boundary with the surface (see Figs. 3 and 4) with phosphorus and sulphur more abundant in the boundary regions and nitrogen and molybdenum depleted (Table I). Titanium oxides predominated in both phases, and there was relatively little variation in the AI/Ti and V/Ti ratios with other treatments used (see Table 11). In general, ESCA exhibited greater elemental sensitivity than other techniques used (see Fig. 5 and Table I). In comparison to processes 1-4, nitric-acid passivation introduced additional elements including Fe, Pb, Mo, P, and S.
DENTAL IMPLANT MATERIALS. I1
1073
z
0 la
rY I-
z
W
0
6
0
z
0
SPUTTERING TIME (SECS)
Figure 2. 'He' ISS sequential spectra (as Fig. 1)for a TiAlV surface treated with process 5 (nitric-acid passivation).
ESCA studies showed that plasma cleaning produced a surface lower in carbonaceous contamination (see Fig. 6), but there was contamination from silicon and nitrogen. The silicon contamination is believed to result from the quartz plasma vessel.
Figure 3. BSE image of TiAlV alloy surface treated with H N 0 3 (process 5) showing prior P-grain boundaries and a-plate colonies (original magnification ~ 1 2 8 ) .
SMITH ET AL.
1074
Figure. 4. Detail of Figure 3, showing differentiated composition and segregation at grain boundary (original magnification ~ 1 0 0 0 ) .
C 0-Cr- Mo The ISS spectra indicated that a thin but definite layer of organic material was present on all the surfaces corresponding to procedures 1 to 5, with minimal amounts on those treated by plasma (F) (see Figs. 7 and 8). These depth profiles also demonstrated that the oxide layer on this alloy was relatively thin compared to those on the Ti alloys. An Al/Si oxide layer was detected on top of the CoKr layer. Throughout the oxide layer, there was evidence of Na and K as well as I'b. Mo was not observed until a depth of about 5 nm, i.e., at TABLE I ESCA Determinations of Elemental Compositions for Ti6A14V Alloy at Grain Boundary and Grain Interior
Element
Photoelectron Line Used in Analysis
Composition (atom %) Grain Interior
Grain Boundary
70.0 2.0 16.0 4.0 0.8 0.3 1.0 -
67.0 1.3 17.0 2.4 1.2 0.1 3.1 1.4 1.7 1.2 2.2 0.8 99.4
~~
0
Is
V Ti N
2P3 2P3 1s 2P 3d 2s 2P 2P 4f 2s 2P 2P
Ca Mo P
s (so;) s (s=) Pb A1 Si Fe Total
-
0.8 2.8 1.1 98.8
DENTAL IMPLANT MATERIALS. I1
1075
TABLE I1 ESCA Determinations of Atomic Ratios in Ti6A14V Alloy Surfaces after Various Preparative Procedures Atomic Ratios ~
Preparation procedures As received
Detergent-wash Nitric acid Plasma
(1) (2) (5) (6)
O/Ti
Al/Ti
V/ Ti
2.7 3.5 4.2 3.3
0.05 0.1 0.2 0.1
0.05 0.05 0.1 0.1
the oxide/metal interface. Process 2 showed high surface concentrations of Si and C1. As before, procedures 4 and 5 appeared to increase the oxide layer thickness; however, no Pb was evident after process 5. The ESCA data showed similar features for the material at a greater depth from which information is obtained by this technique. Differences between ul
c
0
I
1000
BINDING ENERGY (eV)
0
Figure 5. ESCA spectrum of TiAlV surface treated with process 5. I-
ul
z
7
0
3
I
0 0
I
1000
BINDING ENERGY ( e V )
Figure 6. ESCA spectrum of TiAlV surface subjected to plasma cleaning (process 6); cf. Figure 5.
0
SMITH ET AL .
1076
z
0
z
0
I
Figure 7. ISS 3He+ sequential spectra as Figure 1 for Co-Cr--Mo alloy treated with detergent (process 2).
z
0 I-
Q
lY I-
z
W
0 Z 0
0
z
0
SPUTTERING TIME (SECS)
Figure 8. ISS 3He' sequential spectra as Figure 1 for plasma-cleaned CoCr-Mo alloy (process 6).
DENTAL IMPLANT MATERIALS. I1
1077
the 2 and 5 processes over the lower part of the binding energy spectrum (
A1203, and hydroxyapatite dental implant materials that had been subjected to six commonly used preparative procedures. The results showed that each procedure generated an individualistic composition for the outermost surface of each material. These differences could be significant in cellular and tissue response. Improved understanding of these factors requires defined and reproducible surfaces.
IN T RODUCTION
The tissue response to a dental implant may involve physical factors such as size, shape, surface topography, and relative interfacial movement, as well as chemical factors associated with the composition and structure.’” There is still relatively little reliable information on the influence of these variables because the majority of in vitro and in vivo studies on implant materials and systems has not characterized in a precise manner the bulk and surface characteristics of the materials being used, nor the effects of the particular preparative cleaning and sterilization systems on such characteristic^."^ Further, until very recently, little or no attempt has been made in most studies to evaluate ionic release into the tissues and its relationship to the observed tissue r e a ~ t i o n . ~ These - ~ issues have been discussed in Part I of this series8 and elsewhere .6,7,y,1fl Characterization of implant materials to evaluate tissue reaction necessarily includes analysis of surface composition and contamination.’-13 It is now well understood that the surface composition of most implant materials is substantially different from the bulk, and a variety of techniques have been used to characterize such surfaces.’’ Several such techniques are needed to fully characterize implant surfaces. *To whom correspondence should be addressed at Centre for Biomaterials, University of Toronto, 170 College Street, Toronto, Ontario, Canada M5S 1Al. Journal of Biomedical Materials Research, Vol. 25, 1069-1084 (1991) CCC 0021-9304/91/091069-16$4.00 0 1991 John Wiley & Sons, Inc.
SMITH ET AL.
1070
Such a combined approach is only just beginning to be used in biomateria1s.l" L a ~ s m a a , Kasemo '~ and c o - w ~ r k e r s , ' ~Sundgren,16 ~'~~'~ Ducheyne and cow o r k e r ~ have ' ~ applied ESCA and Auger analysis to the surface characterization of titanium and stainless-steel (dental) implants and found, in common with the present that topographical and compositional changes occur in the surface of such implants as a result of both in vitro treatment and in vivo implantation. We also observedz2in earlier work extensive changes in the surface composition of nickel-chromium alloys after implantation for at least 3 months in animal models. In the present work, four presently commonly used dental implant materials have been subjected to customarily used preparative techniques, and the effects on their surface characteristics investigated. In a previous paper: we described the results of SEM/EDX and surface wetting studies. The present paper describes the results of ISS, SIMS, and ESCA studies on these materials. ISS gives information on the outer 1-2 atomic layers and is especially sensitive to high mass elements such as elements above Ti. The signal amplitude is related to atomic concentrations in the surface and is not usually strongly affected by changes in matrix. In SIMS, all elements are detectable, including H, as well as clusters related to molecular fragments that may indicate the nature of the chemical bonding at the surface. However, the relative instrumental sensitivities to the elements can vary over a range of about five orders of magnitude and can be strongly affected by the matrix. These processes lead to sputter erosion of the surface such that successive spectral scans can give depth profiling information. ESCA provides information from depths as great as 5 nm on the specific chemical binding of atoms for all elements except H and He. In some cases, semiquantitative data (210%) can be obtained. Detailed reviews of these techniques and their biological applications1"" can be found elsewhere. MATERIALS A N D METHODS
The following materials were investigated: (a) Ti6A14V ELI conforming to ASTM F-136, (b) Co-Cr-Mo conforming to ASTM F-75 (Canox Ltd.), (c) singlecrystal A1203(Kyocera, Kyoto) (AO), (d) dense hydroxyapatite (HA) (Durapatite; Sterling Winthrop Laboratories). Chemical analyses of these materials have been given previously.* Discs 3.5 mm in diameter and 1.5 mm thick were fabricated by cutting from rod stock with a low-speed diamond saw. The alloys and HA specimens were finished by wet grinding to 600 grit on Sic paper. The alumina discs were polished by a proprietary process. Similar porous-surfaced alloy discs were prepared by sintering spherical particles of 50-300 pm diameter onto the solid substrate as described elsewhere? For the Co-Cr-Mo alloy powders (Canox Ltd.), a 1300°C, 3-h sinter in uucuo was employed, whereas for the Ti6A14V powder (Dynament Inc.) 1250°C for 1 h in vacuo was used.' The solid discs were subjected to the same heat treatment (simulated sintering process). Fully porous alumina discs were made by a proprietary process (Kyocera Inc.).
DENTAL IMPLANT MATERIALS. I1
1071
After fabrication, the specimens were washed in distilled water and airdried. This was the "as-received condition. Groups of three specimens for each material were subjected to the following cleaning regimens: (1) Retained in the "as-received" condition (2) Washing in an ultrasonic cleaner as follows: (a) Sonicated for 60 min in a 2% solution of Decon detergent (B.D.H Chemicals Canada Ltd.). (b) Then sonicated in deionized water for 2 min, three changes. (c) Air-dried in a closed container. (3) As in (2), followed by steam sterilization by autoclaving at 121°C for 30 min. (4) As in (2), followed by radiation sterilization. (5) As in (2), followed by treatment with 40% vol HNO, at RT (ASTM F-86) for 60 min in the sonicator followed by similar rinsing with deionized water for 3 min, three changes. (6) As in (2), followed by exposure to an argon plasma for 30 min at 0.3 mm Hg (Harrick Model PDC-3XG Plamsa Cleaner). The materials were coated with Au for SEM and carbon for energy dispersive analysis (EDX) in a Polaron sputter coater. Surface analyses were carried out on duplicate samples using ISS, SIMS, and ESCA. ISS and SIMS measurements were made using a 3M-Kratos Model 555 BX combined ISS/SIMS spectrometer with the assistance of Dr. G.R. Sparrow (Advanced R and D Inc.). A surface area of about 6 mm2 was examined for SIMS and about 2 mm2 for ISS. The positive ion beams used were 3Heor 40Arwith a beam energy of 2 keV or, occasionally 4 key At 2 keV the beam current was typically 3-500 mA/cm2 over the irradiated area. ESCA Spectra were obtained using a Surface Science Labs SSX-100 spectrometer mm Hg, using monochromatized A1 K, radiation at a pressure of lO-'-lO-' a small flood gun compensation was applied until satisfactory charge compensation was obtained. Additional spectra were obtained using a similar instrument with the assistance of Dr. B. Ratner (National ESCA Centre for Biological Materials, University of Washington). Most data were obtained using a 1-mm-diameter spot but some spectra were also obtained using the 150-pm-diameter spot capability. Stoichiometry could be obtained from the relative peak areas using the form factors derived by S~hofield?~ modified to take account of the mean free path of the escaping electron. A constant transmission function was assumed for the electron analyzer. Under the present conditions the data can be considered significant only to 210%.
RESULTS
Ti6A14V alloy Sequential ISS broad-scan spectra are shown in Figures 1 and 2 for this alloy surface treated by two different processes: (4) radiation sterilization
SMITH ET AL.
1072
Z
9
L
z z 0
SPUTTERING TIME (SECS) Figure 1. 3He+ ISS sequential rapid scan spectra showing depth profile changes in TiAlV surface ion concentrations as sputtering time is increased. Sputtering rate approximately 0.5 nm/60 s. The surface was treated by process 4 (radiation sterilization).
(Fig. 1) and (5) nitric acid passivation (Fig. 2). The depth profiling indicated by the successive spectra reveals the presence of Pb in the outer 1 nm or so of the surface, more prominently for process 4. The sputter etch rate was approximately 1 nm/150 s as determined from standard calibrated samples. Continued depth profiles showed that the nitric-acid treatment (5) and an argon plasma treatment (6) produced the thickest oxide layers, which, as shown by the expanded ISS spectra, typically contained sodium, aluminum, titanium, and lead oxides. The nitric-acid treatment resulted in nitrogen within the oxide layer. Vanadium was only detected in the oxide layer when SIMS and ESCA were used. As noted in a previous paper: nitric-acid treatment induced pronounced etching. Small-area ESCA analyses of the nitric-acid-treated surface suggested that different compositions existed on the face of the grain and on the intersection of grain boundary with the surface (see Figs. 3 and 4) with phosphorus and sulphur more abundant in the boundary regions and nitrogen and molybdenum depleted (Table I). Titanium oxides predominated in both phases, and there was relatively little variation in the AI/Ti and V/Ti ratios with other treatments used (see Table 11). In general, ESCA exhibited greater elemental sensitivity than other techniques used (see Fig. 5 and Table I). In comparison to processes 1-4, nitric-acid passivation introduced additional elements including Fe, Pb, Mo, P, and S.
DENTAL IMPLANT MATERIALS. I1
1073
z
0 la
rY I-
z
W
0
6
0
z
0
SPUTTERING TIME (SECS)
Figure 2. 'He' ISS sequential spectra (as Fig. 1)for a TiAlV surface treated with process 5 (nitric-acid passivation).
ESCA studies showed that plasma cleaning produced a surface lower in carbonaceous contamination (see Fig. 6), but there was contamination from silicon and nitrogen. The silicon contamination is believed to result from the quartz plasma vessel.
Figure 3. BSE image of TiAlV alloy surface treated with H N 0 3 (process 5) showing prior P-grain boundaries and a-plate colonies (original magnification ~ 1 2 8 ) .
SMITH ET AL.
1074
Figure. 4. Detail of Figure 3, showing differentiated composition and segregation at grain boundary (original magnification ~ 1 0 0 0 ) .
C 0-Cr- Mo The ISS spectra indicated that a thin but definite layer of organic material was present on all the surfaces corresponding to procedures 1 to 5, with minimal amounts on those treated by plasma (F) (see Figs. 7 and 8). These depth profiles also demonstrated that the oxide layer on this alloy was relatively thin compared to those on the Ti alloys. An Al/Si oxide layer was detected on top of the CoKr layer. Throughout the oxide layer, there was evidence of Na and K as well as I'b. Mo was not observed until a depth of about 5 nm, i.e., at TABLE I ESCA Determinations of Elemental Compositions for Ti6A14V Alloy at Grain Boundary and Grain Interior
Element
Photoelectron Line Used in Analysis
Composition (atom %) Grain Interior
Grain Boundary
70.0 2.0 16.0 4.0 0.8 0.3 1.0 -
67.0 1.3 17.0 2.4 1.2 0.1 3.1 1.4 1.7 1.2 2.2 0.8 99.4
~~
0
Is
V Ti N
2P3 2P3 1s 2P 3d 2s 2P 2P 4f 2s 2P 2P
Ca Mo P
s (so;) s (s=) Pb A1 Si Fe Total
-
0.8 2.8 1.1 98.8
DENTAL IMPLANT MATERIALS. I1
1075
TABLE I1 ESCA Determinations of Atomic Ratios in Ti6A14V Alloy Surfaces after Various Preparative Procedures Atomic Ratios ~
Preparation procedures As received
Detergent-wash Nitric acid Plasma
(1) (2) (5) (6)
O/Ti
Al/Ti
V/ Ti
2.7 3.5 4.2 3.3
0.05 0.1 0.2 0.1
0.05 0.05 0.1 0.1
the oxide/metal interface. Process 2 showed high surface concentrations of Si and C1. As before, procedures 4 and 5 appeared to increase the oxide layer thickness; however, no Pb was evident after process 5. The ESCA data showed similar features for the material at a greater depth from which information is obtained by this technique. Differences between ul
c
0
I
1000
BINDING ENERGY (eV)
0
Figure 5. ESCA spectrum of TiAlV surface treated with process 5. I-
ul
z
7
0
3
I
0 0
I
1000
BINDING ENERGY ( e V )
Figure 6. ESCA spectrum of TiAlV surface subjected to plasma cleaning (process 6); cf. Figure 5.
0
SMITH ET AL .
1076
z
0
z
0
I
Figure 7. ISS 3He+ sequential spectra as Figure 1 for Co-Cr--Mo alloy treated with detergent (process 2).
z
0 I-
Q
lY I-
z
W
0 Z 0
0
z
0
SPUTTERING TIME (SECS)
Figure 8. ISS 3He' sequential spectra as Figure 1 for plasma-cleaned CoCr-Mo alloy (process 6).
DENTAL IMPLANT MATERIALS. I1
1077
the 2 and 5 processes over the lower part of the binding energy spectrum (
