Okefjord and Hansson Et Al

JOMI on CD-ROM, 1993 Jan (32-40 ): Surface Analysis of Four Dental Implant System… Copyrights © 1997 Quinte…
Surface Analysis of Four Dental Implant Systems

Ingemar Olefjord, Prof/Stig Hansson, Lic Eng
Dental implants obtained from four suppliers were analyzed by electron

spectroscopy for chemical analysis and scanning electron microscopy. Three of
the implants were delivered in a sterilized condition, while the fourth implant
was delivered in a plasma-sprayed condition. The covering oxide layer consisted
mainly of TiO2. Divalent and trivalent states of titanium were also detected,
showing that TiO and Ti2O3 layers occurred. The thickness of the oxide formed
on the plasma-sprayed implant was 3.4 nm. The oxide thicknesses of the
sterilized implants were 4.6 ± 0.4 nm. The surfaces of all samples were covered
with organic contaminants. A strong fluorine signal was obtained from one
sample, indicating that the supplier etches the implants in hydrofluoric acid.
Calcium and zinc were found on the surfaces of all samples from one supplier,
while calcium and silicon were found on the surfaces of the implants from
another supplier. It is suggested that inorganic contaminants should be avoided
because these species can possibly provoke the dissolution of titanium. (INT J
O RAL MAXILLOFAC IMPLANTS 1993;8:32-40.)
Key words: cleanliness, dental implant, oxide thickness, surface analysis, titanium
Clinical studies1 have shown that dental implants produced from commercially pure
titanium can support fixed prostheses in the edentulous jaw for long periods of time.
Titanium is one of the metals best tolerated by the body in spite of the fact that it is
one of the most reactive metals in the periodic table. The "poisoning" effect of Ti is
low because the metal is passivated by the immediate formation of a surface oxide
during manufacturing. The thickness of the oxide increases during sterilization in
water vapor. It is this oxide layer that the biomolecules meet when the implant is
placed into bone.
Titanium is known for its high corrosion resistance, for example, in
chloride-containing environments, because of the high stability of TiO2. Titanium, as
all passivable metals and alloys, is dissolved (from a thermodynamic point of view)
in its passive state. However, the dissolution rate is very low. This property is
probably the primary factor in the successful use of Ti as an implant material.
During production and handling of implants, their surfaces are contaminated.
The contaminants are organic and inorganic in nature. Inorganic contaminants can be
harmful to the biocompability of an implant because they can catalyze the oxygen
reaction (O2 + 2H2O + 4e- → 4OH-) and thereby enhance the dissolution of Ti (Ti →
Ti4+ + 4e-). Further, adsorption of ions on titanium oxide may increase the defect
concentration of the oxide and thereby increase the dissolution rate of the metal.
Good long-term prognosis requires that the implant be in direct contact with
living bone tissue. Brånemark et al2 coined the term "osseointegration," defined as
"A direct structural and functional connection between ordered living bone and the
surface of a load-carrying implant." However, the term "direct . . . connection" is
based upon observations using light microscopy, the resolution of which is of the
order of 1-µm magnitude. Later, Hansson et al3 analyzed Ti implants that had been
removed from patients after implantation times ranging from 6 months to 7 years.
The implants adhered firmly to the bone. Decalcified samples were studied by
electron microscope. Bundles of collagen were not found closer to the implant
surface than 1,000 to 3,000 nm, whereafter they were replaced by collagenous
filaments arranged perpendicularly to the implant surface. The collagenous filaments
were always separated from the implant surface by a layer of ground substance
approximately 20 nm thick. Cell processes adjacent to Ti implants also seem to be
separated from the implant surface by a ground substance layer,3-7 the thickness of
which is about 20 nm. The thickness of the ground substance layer seems to depend
on the implant material.3-7 Among the metals Au, Zr, stainless steel, and Ti, the
thinnest ground substance layer was obtained on Ti. The property of the metal
determines the tissue response and thereby controls the thickness of the ground
substance layer.3 The main factors involved are the properties of the surface oxide
layer and the composition of the contamination layer.
The surface energy of clean metal oxides is high. This may be advantageous
because it results in the adsorption of molecules and thereby in an improved cellular
attachment.8 A contamination layer on the implant surface would, even if very thin,
in most cases decrease the surface energy. It has been speculated that the adhesion
between an implant and bone is created by different chemical bonds, some of which
are of very short range.9 Examples of such potential bonds are the hydrogen bonds
between TiO2 and the hydroxyl groups of the glycoproteins in the ground substance
layer. Such bonds would be severely disturbed by even a very thin contamination
layer. The Basic Science Committee of the International Congress on Tissue
Integration in Oral and Maxillo-Facial Reconstruction10 stated that "A monolayer of
contaminants from the environment or bulk material can invalidate the utility of the
implant."
A recent improvement in implant technology is that the implant is delivered
cleaned and sterilized by the manufacturer. To obtain the necessary quality
assurance, an implant supplier should perform the cleaning and sterilization process.
In the present study, the cleanliness of four different dental implants, three of which
were sterile-packed, were investigated. In addition, the oxide composition and
thickness of the surface oxide after sterilization were determined.
Materials and Methods
Implants obtained from four suppliers were analyzed by electron spectroscopy for
chemical analysis (ESCA) and scanning electron microscopy (SEM): Implant A—

Astra dental implant system (Astra Tech AB, Mölndal, Sweden); Implant B—
Brånemark System (Nobelpharma AB Göteborg, Sweden); Implant C— Screw-Vent
(Dentsply/Implant Div, Encino, CA); and Implant D—IMZ (Friedrichsfeld,
Mannheim, Germany). The implants were shaped as screws or cylinders; the length
and the diameter were approximately 11 and 3.5 mm, respectively. The diameter of
the Nobelpharma implant was 3.75 mm. Three of the four manufacturers (Astra
Tech, Nobelpharma, and Dentsply) delivered sterilized implants. Two samples of
each of the implant types A, B, and C, and one sample of the implant type D were
analyzed. When two implants of the same type were analyzed, the samples were
received from different production batches.
The implant capsules were broken and the samples were immediately mounted
in the ESCA instrument (HP 5950 A). The analysis started with a survey scan to
record signals from all elements present in the surface region. Chemical and
quantitative information related to the elements was obtained from narrow scans
over representative peaks. To obtain in-depth information, the analysis was repeated
after removal of about 1nm of the top layer by ion etching.
After the ESCA analyses, SEMs (Jeol 733) were taken to obtain information
about the topography of the surfaces.
Quantification
The spectrum shown in Fig 1 is selected to illustrate the information obtained from
the ESCA analysis. An ion-etched Ti sample (ie, the air-formed oxide was removed)
was oxidized in oxygen at 26 kPa (200 torr) for 15 minutes. The oxidation was
performed in the reaction chamber of the ESCA instrument. The elements present on
the surface were O, Ti, and C. The recorded signals were deconvoluted into
components representing chemical states and bondings. The dashed lines represent
the backgrounds, and the dotted lines are the sum of the deconvoluted peaks and the
background. It appears that there is agreement between the recorded signal and the
sum of the components.
The Ti spectrum is complex; the metallic state and the divalency, trivalency and
four-valency oxide states of Ti occur. (In fact, these four states occur twice because
the Ti 2p level is split into Ti 2p1/2 and Ti 2p3/2 states, which are separated by 6.1
eV.) The binding energies of the Ti 2p3/2 signals representing Tim (metallic state),
Ti 2+, Ti3+, and Ti4+ are 454.1, 455.6, 457.5, and 459.4 eV, respectively. (The
calibration of the spectrometer was Au4f7/2 84.0 eV and Cu 2p3/2 932.7 eV.)
The occurrence of the peak representing the metallic state shows that the oxide is
very thin. The dominating peak is Ti4+ recorded from TiO2. The other two oxide
peaks represent Ti in TiO and Ti2O3. The oxides form surface layers on each other.
From a thermodynamic point of view, TiO is the most stable of the oxides.
Therefore, TiO is situated closest to the metal phase; followed by Ti2O3 and TiO2.
The carbon peak consists of two peaks at 285.6 and 287.0 eV, representing C in
hydrocarbon and C bound to OH. The carbon signal is enlarged five times, thus the
amount of hydrocarbon present on the surface is low. The source of the carbon
compounds is the atmosphere in the vessel in which the sample was oxidized.
The strong oxygen peak at 530.7 eV represents the oxygen in TiO2. The second
oxygen peak at 531.8 eV is mainly emitted from TiO and Ti2O3. About one third of
the latter oxygen peak represents oxygen bound to OH groups in hydrocarbons
present in the contamination layer.
The composition and thickness of surface layers (oxide and contamination
layers) are estimated from the measured intensities. In the model described below, it
is assumed that the layers are plane-parallel on the surface. The intensities (I) of the
peaks representing the oxide and the metallic states of Ti are expressed by the
formula11,12:
(1) Iox = αDoxYλox sinθ {1 – exp [ –aox/(λox sinθ)]} exp [ –a c/(λcsinθ)]
(2) Im = αDmYλm sinθ exp [ –aox/(λox sinθ)] exp [ –ac/(λcsinθ)]
where aox and ac are the thicknesses of the oxide and the contamination layers
respectively, α is a variable instrument constant, D is the atomic density, Y is an
experimentally determined photoelectron yield factor, λ is the attenuation length,
and θ is the photoelectron take-off angle. Assuming that λox = λm, the thickness of
the oxide layer is obtained from the formula:
(3) a ox = λox sinθ ln(1 + DmIox/(DoxIm)).
The following parameter values have been used: Dox(Ti) = 0.053 mole/cm3;
Dm(Ti) = 0.094 mole/cm3; λ = 1.7 nm; θ = 34°.
The composition of the surface layer is calculated from the measured intensities
and equation (1) above. The concentration, Ci, of the element, i, is:
(4) Ci = Di/ΣDn.
In this study the variation of the λ values is small thus the exponential functions
can be neglected when determining the composition.
Using formula (3) and the intensities recorded in Fig 1, the oxide formed on the
surface of Ti at room temperature is found to be 1.6 nm thick. The inner layers of
TiO and Ti2O3 are about one monolayer each.
Results
Figure 2 shows representative survey spectra recorded from all four implant
materials involved in this study. The analyzed energy range was between the binding
energies 0 and 1,200 eV. All detectable signals occurred within this range. It
appeared that the elements O, Ti, and C gave the strongest signals. The peaks
represent the elements in the surface oxide and in the contamination layer. The
spectra also show that the following elements were present in the surface zone:
implant A, N and Cl (only a weak signal from the 2p level—above the noise level—
is detected; there should also have been a signal from the 2s level); implant B, Zn,
Ca, N, and Si; implant C, Fl, Ca, Cl, N, and Si; implant D, N and Ca. Zinc signals, as
shown in Fig 2(B), were obtained from three different samples from two different
batches. The spectra from sample C, Fig 2(C), indicate that this implant had been
etched in hydrofluoric acid. The occurrence of a strong F- signal shows that the
rinsing procedure used had not removed fluoride from the surface.
Narrow scans recorded from samples A and B are shown in Figs 3 and 4,
respectively. The spectra of the "as-received" state and after removal of about 1nm
of the surface layer are shown. The deconvolution has been done in the same way as
in Fig 1. To make the spectra more easily readable, the intensity of the background
and the summation of peak intensities arc not shown for all signals.
The dominating oxide state of Ti was the four-valency state. However, it appears
from Fig 3 that even the metallic and the divalency and trivalency states were
detected. Applying formula (3), the oxide thickness was found to be about 5 nm.
After ion etching, the intensity of the lower valency oxide states increased. To some
extent this is an artifact caused by reduction of the oxide by the ion beam. The
oxygen signal is split into three components. The peak denoted O2- represents
oxygen in TiO2, and the peak on the left side of the main peak represents the lower
valency oxides, TiO and Ti2O3. The third peak, located at 533.1 eV, may represent
oxygen in a carboxyl group and water.
The carbon signals have been deconvoluted into four peaks. The position of the
carbon peak depends on the bonding characteristic of the atom. In Figs 3 and 4,
possible groups are denoted. The alcohol notation also represents ether, ester, and
hydroperoxide. The double bond is an aldehyde or ketone, and the triple represents
carboxylic acid, ester, or carbonate. The intensity of the carbon peak from the triple
bond corresponds to the intensity of the oxygen at 533.1 eV. It is therefore suggested
that the peak represents a carboxyl group. The oxygen could also represent oxygen
in carbonate, but in that case its intensity should have been three times larger to
correlate with the carbon signal at 289.5 eV.
In the survey scan (Fig 2, A) a weak signal appeared at the position of Cl 2p. The
narrow scan over this peak (Fig 3) shows that its peak height intensity is about the
same as the noise level. Thus the content of Cl- is low.
Nitrogen was detected from all samples. The peaks are located at 401 eV, which
is the position of N in the groups – CN, – NH2, – OCONH – and – CONH2. The
origin of these compounds is not known. Calcium and K are present on sample B.
The source of these elements may be the rinsing water used in the process.
Using the formulae previously described, the composition and thickness of the
surface layers were calculated. The results are shown in Table 1. The model used
implies that all elements were uniformly distributed through the surface layer.
However, the elements were present in the metal phase (Tim), the oxide film, and the
contamination layer. The compositions given in Table 1 are therefore apparent
compositions. The measured "apparent" content of Ti in its metallic state is low
because of the attenuation of electrons in the oxide and contamination layers. The
oxide was of the same thickness for all samples, except for sample D, hence the
"apparent" concentration is a good measure of the amount of the contaminants on the
surface. The data in rows 1 and 4 are obtained from the spectra in Figs 3 and 4,
respectively.
From Table 1 it can be seen that the thickness of the oxide formed on samples A,
B, and C is 4.6 ± 0.4 nm. Sample D was not sterilized and therefore the oxide
thickness is less: 3.4 nm. The average ratio between the contents of O2- and Ti4+ is
1.58 ± 0.18. This is lower than the expected value for TiO2, which is 2.0. The result
can to some extent be explained by the fact that the O2- electrons were more
attenuated than the Ti4+ electrons in the contamination layer.
The contents of hydrocarbon (C – H) varied markedly between the samples.
There are even large differences between the samples from the same suppliers
(samples A and C). It should be noted that the content of the carboxyl group is
almost the same on all sterilized samples, the average value and the standard
deviation being 4.0 ± 0.7. Sample D, which was not sterilized, is covered with much
less of the carboxyl compound. However, only one sample was available so the
finding is only an indication.
Figures 5 to 8 show micrographs of the four implants. The surface morphology
of samples A (Figs 5a and 5b) and B (Figs 6a and 6b) is about the same. The
characteristic turning pattern appears in the pictures. It is suggested that the light
particles on the surfaces of both implants are metal particles that were formed during
turning and are loosely bonded to the surface. The surface of sample B shows
pronounced fracturing of metal chips.
The surface of sample C (Figs 7a and 7b) shows a noticeable pitting attack. The
ESCA analysis shows strong signals from F. It is therefore likely that the implant
had been etched in hydrofluoric acid. (Auger analysis indicated that the white small
particles consist of a fluorine compound left on the surface.) From the photographs,
it appears that the diameter of the pits varied in the range of 0.1 to 10 µm.
Implant D had been prepared by plasma spraying. The morphology shown in
Figs 8a and 8b is characteristic for a surface layer formed from the melt of a metal
with high solidification temperature. The surface becomes rough with deep craters
and trenches.
Discussion
The natural oxide formed on Ti consists of layers of TiO, Ti2O3, and TiO2. The
divalency oxide, TiO, is (from a thermodynamic point of view) the most stable
oxide, estimated per mole oxygen, and thus occurs at the oxide-metal interface. The
outer oxide layer is TiO2. Oxidation for a short time at room temperature provides a
total oxide thickness of approximately 1.6 nm. The thicknesses of each of the inner
oxide layers are one monolayer, about 0.35 nm.
The analyses show that the surface oxide grows during sterilization. The average
oxide thickness was 4.6 ± 0.4 nm. The plasma-sprayed sample had not been
sterilized. The oxide for this implant, 3.4 nm thick, was formed during solidification
and cooling of the surface layer. It was possible to detect the low valency states of
titanium from the sterilized samples. Their intensities were very low because the
overlayer of TiO2 had increased in thickness when compared to the room
temperature-exposed sample. In addition, the attenuation of the signals from the
underlying compounds was increased.
Carbon-containing compounds were adsorbed on the oxide surface of all
samples. The dominating signals recorded represent carbon bound to hydrogen in
aliphatic and/or aromatic compounds. These groups varied considerably between the
samples. Sample A1 shows about three times more "apparent" content of
hydrocarbon than the sample with the lowest content. However, the hydrocarbon is
present in the outermost layer and thereby attenuates the signals from the underlying
layers. Thus, the outermost compound becomes overrepresented when the average
composition is calculated without taking into consideration the real distribution of
the elements.
The functional carbon groups are bound to the surface. The change of the surface
energy will depend on the nature of these molecules. The concentrations of C – OH
and C = O did not vary much between the samples. The average percentage value of
the apparent concentrations of C – OH (before ion etching) is 7.3 ± 1.2. The
corresponding values for C = O and C ≡ O are 3.4 ± 0.9 and 3.6 ± 1.3, respectively.
The samples D1 and B2 showed significantly higher contents of C – OH and C = O,
respectively, than the others. In the same way, sample A1 showed a significantly
lower concentration of C = O than the other samples. However, in this case the low
value is partly caused by the fact that the surface was covered by more of the C – H
compound. Thus it can be concluded that the surface of the implants was saturated
with functional carbon groups. The influence of the adsorbed carbon molecules on
the surface energy depends on whether these molecules are dissolved in the body
fluids during the healing process.
Calcium was detected on all samples except A1, A2, and C2. The two B samples
showed that about 1% of the total analyzed volume consists of Zn. A survey scan
was recorded from a third B sample. This also showed that Zn was present on the
surface. If it is assumed that the Zn atoms are situated on the oxide surface, the Zn
concentration in this plane is much higher than the percentage level shown in Table
1. Silicon was observed on samples B2, C1, C2, and D1. A high concentration of F-
was found on sample C2. This sample had probably been etched in hydrofluoric acid
and the ensuing rinsing procedure had not been sufficient.
The influence of contaminants on the dissolution rate of Ti in body fluids has not
been studied. It is suggested that, in principle, organic compounds lower the
dissolution rate of titanium because they block the sites for the oxygen cathodic
reaction. Foreign ions on the titanium oxide surface may catalyze the oxygen
reaction and thereby promote the dissolution of titanium. In engineering, ions of Pt,
Fe, Cr, and Mo, for example, are added to reducing acids (H3PO4, H2SO4) to enhance
the electrochemical reactions so that passivation can occur. It has been reported13
that the corrosion products formed on the surface of Ti during its exposure to NaNO3
and NaCl containing aqueous solution catalyze the reduction process of NO3- and
thereby increase the dissolution rate of Ti. Zinc in both solid and liquid form may
cause embrittlement of Ti.14 It is not known whether the Ca2+, Zn2+, and Si4+ found
on the surface in this study have the effect of promoting dissolution of Ti and thus
"poison" the tissue or cause stress-corrosion cracking of the implant. However, until
this fundamental question has been clarified, a process for producing implants
should be developed so that inorganic contaminants are avoided.
Conclusions
1. Exposure of pure Ti to oxygen at room temperature (200 torr for 15 minutes)
provided an overall 1.6-nm-thick oxide layer.
2. The thickness of the Ti oxide formed during sterilization in water vapor was 4.6
nm.
3. The implant surfaces were contaminated with organic and inorganic compounds.
4. The variation of the concentrations of the functional groups C – OH, C = O, and
C ≡ O was relatively small between the implants from the different suppliers.
5. Zinc and Ca were found on the surfaces of the implants from two different
suppliers, respectively.
6. Surface treatment, sterilization, and handling of the implants should be
performed so that contamination of inorganic compounds is avoided.
1. Adell R, Lekholm U, Rockler B, Brånemark P-I. A 15-year study of

osseointegrated implants in treatment of the edentulous jaw. Int J Oral Surg
1981;10:387-416.
2. Brånemark P-I, Hansson BO, Adell R, Breine U, Lindström J, Hallén O, et al.
Osseointegrated implants in the treatment of the edentulous jaw. Experience from
a 10-year period. Scand J Plast Reconstr Surg 1977;11(suppl 16).
3. Hansson H-A, Albrektsson T, Brånemark P-I. Structural aspects of the interface
between tissue and titanium implants. J Prosthet Dent 1983;50:108-113.
4. Albrektsson T, Hansson H-A, Ivarsson B. Interface analysis of titanium and
zirconium bone implants. Biomaterials 1985;6:97-101.
5. Albrektsson T, Hansson H-A. An ultrastructural description of the interface
between bone and sputtered titanium or stainless steel surfaces. Biomaterials
1986;7:201-205.
6. Johansson C, Lausmaa J, Ask M, Hansson H-A, Albrektsson T. Ultrastructural
differences of the interface zone between bone and Ti-6Al-4V or commercially
pure titanium. J Biomed Eng 1989;11:3-8.
7. Johansson C, Hansson H-A, Albrektsson T. A qualitative, interfacial study between
bone and tantalum, niobium or commercially pure titanium. Biomaterials
1990;11:277-280.
8. Baier RE, Meyer AE. Surface energetics and biological adhesion. In: Mittal KL
(ed). Physicochemical Aspects of Polymer Surfaces, vol 2. New York: Plenum,
1983:895.
9. Kasemo B. Biocompatibility of titanium implants: Surface science aspects. J
Prosthet Dent 1983;49:832.
10. Bjursten L-M, et al. Basic Science Committee. Tissue Integration in Oral and
Maxillo-Facial Reconstruction [Proceedings of an International Congress, May
1985, Brussels.] Amsterdam: Excerpta Medica, 1986:511.
11. Olefjord I, Elfström BO. The composition of the surface during passivation of
stainless steel. Corrosion NACE 1982;38:46.
12. Olefjord I. XPS — qualitative and quantitative aspects. In: Baer DR, Clayton CR,
Davis GD (eds). The Application of Surface Analysis Methods to
Environmental/Material Interaction S. [proceedings, vol 7]. Pennington, NJ: The
Electrochemical Society, 1991;22:91-97.
13. Shekun IF, Dikusar AI, Molin AN, Darydov AD. A possible mechanism for the
reduced effective valency of Ti during anodic dissolution in chloride-nitride
solutions. Elektrokhimiya 1990;26:754-757.
14. Schutz RW, Thomas DE. Corrosion of Ti and Ti-alloys. In: Davis JR (ed). Metals
Handbook, ed 9, vol 13. Ohio: ASM-International, 1987:669.
Fig. 1
ESCA spectrum recorded from Ti oxidized at room temperature.
Fig.
Fig.
3 ESCA signals recorded from sample A before and after ion etching.
Fig.
!700 and !7,000. Figs
Figs
Figs
Figs

Okefjord and Hansson Et Al

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Okefjord and Hansson Et Al

Загружено:

Авторское право:

Доступные форматы

JOMI on CD-ROM, 1993 Jan (32-40 ): Surface Analysis of Four Dental Implant System… Copyrights © 1997 Quinte…

Surface Analysis of Four Dental Implant Systems

Dental implants obtained from four suppliers were analyzed by electron

chemical analysis (ESCA) and scanning electron microscopy (SEM): Implant A—

1. Adell R, Lekholm U, Rockler B, Brånemark P-I. A 15-year study of

!700 and !7,000. Figs

Вам также может понравиться