Bioreactivity of titanium implant alloys

Bioreactivity of titanium implant alloys
Susan J. KerberMaterial Interface, Incorporated, Sussex, Wisconsin 53089-2244 ͑Received 30 September 1994; accepted 3 June 1995͒ A study was conducted regarding the adsorption of peptides on commercially pure ͑cp͒ Tiand Ti-6Al-4V. The peptides used were arginine-glycine-aspartic acid-alanine ͑RGDA͒,arginine-glycine-aspartic acid-serine ͑RGDS͒, and arginine-phenylalanine-aspartic acid-serine ͑RFDS͒. The tripeptide RGD is known to be important for biologically specific adhesion reactions.
This research was conducted to investigate the reason for a tendency toward thrombus formationwith Ti-6Al-4V that is not observed with cp Ti. After argon plasma cleaning, coupons of the titaniumalloys were inserted into solutions with variable concentrations ͑0.0625–2 mg/ml͒ of an individualpeptide group under constant temperature and time conditions. The samples were rinsed, dried, andanalyzed with x-ray photoelectron spectroscopy ͑XPS͒. Adsorption isotherms were obtained byplotting the relative amount of peptide adhesion as a function of solution concentration. It waspostulated through the XPS and adsorption isotherm data that the major adhesion mechanism for thepeptides to the titanium alloys was hydrogen bonding. CP titanium and Ti-6Al-4V are hypothesizedto react differently as implants because Ti-6Al-4V has a more electropositive surface, which allowsfewer hydrogen bonds to form. Hydrophilic reactions were proposed to be of secondary importanceduring bioadhesion, influencing the structure of the second layer adsorbed. There was no correlationfound between the net charge of the peptide groups and their adhesion to the alloys.
I. INTRODUCTION
adhesion. Although it will not have specific binding to im- When a material is implanted into living tissue, proteins plant surfaces in the same fashion as it does to other biomol- immediately adsorb onto the surface of the foreign object.
ecules, this peptide series is well characterized and is known Any subsequent reaction between the material and the host is to be important at implant sites. Adsorption isotherms of a function of these adsorbed proteins and surrounding tissue.
RGD-related peptides on titanium alloys were obtained.
An understanding of the interaction between materials used The mechanism of bioadhesion to titanium and its alloys for implants and proteins may help to increase the success of may be due to hydrogen bonding, hydrophilic interactions, implants. The most common titanium alloy is Ti-6Al-4V and/or charge transfer. To study these parameters, the substi- ͑titanium–6 wt % aluminum–4 wt % vanadium͒, and at this tuted RGD-based peptides had different amino acid side time it is the only titanium alloy used in load-bearing im- chains and different total charges. The four amino acids of plants such as hips and knees because it offers the best com- arginine-glycine-aspartic acid-serine ͑RGDS͒ have charges bination of strength, ductility, and freedom from environ- of ϩ, 0, ϩ, Ϫ for a net charge of ϩ1. In RFDS, the second mental effects such as stress corrosion cracking. Heart position amino acid glycine ͑with a –H side group͒ is re- valves, however, are usually made of unalloyed titanium placed by phenylalanine, an aromatic-containing amino acid.
͑also called cp titanium, for ‘‘commercially pure’’͒ because The charges on the RFDS amino acids are ϩ, Ϫ, ϩ, Ϫ. In of a reported tendency toward thrombus formation1 with Ti- the tetrapeptide RGDA, the final serine ͑with an –OH side 6Al-4V. A study by Johansson et al.2 determined a more group͒ is replaced by alanine ͑with a –CH3 group͒. The natural-like tissue reaction occurred with cp titanium than charges on the RGDA amino acids are ϩ,0,ϩ,ϩ. These three with Ti-6Al-4V. It is theorized that the clotting in the region relatively differently charged peptide groups were adsorbed of the Ti-6Al-4V structures is associated with the aluminum onto the two titanium alloy surfaces. Additionally, because of in the alloy. The differences in which cp titanium and Ti- the difference in the side chains, RFDS is more hydrophobic 6Al-4V interact with known, well characterized peptide sys- than RGDA and RGDS, and RGDS and RFDS should form more hydrogen bonds than RGDA. The relative amount of Both fibronectin and fibrinogen are proteins that exist in peptide adsorbed will be determined with x-ray photoelec- high concentration in animals. Fibronectin is a large protein tron spectroscopy ͑XPS͒. A review of the use of XPS to that can mediate adhesion and spreading of cells on an ex- study protein adhesion was completed by Paynter and tracellular matrix. Fibrinogen is essential to blood clotting.
To perform these adhesive functions, both proteins are re-quired to interact with other proteins, primarily receptors. A II. EXPERIMENT
chain of amino acids with a specific electronic configuration Samples of commercially pure titanium ͑cp Ti, grade 2͒ acts as a receptor area to accomplish this interaction. Both were obtained from Titanium Industries ͑Fairfield, NJ͒ and of fibronectin and fibrinogen have the same specific sequence Ti-6Al-4V from Intermedics Orthopedics, Inc. ͑Austin, TX͒.
of amino acids ͑arginine-glycine-aspartic acid, abbreviated The surfaces were mechanically polished to a bright finish by RGD͒ that act as this specific cell receptor and mediate cell Efco Finishing Corporation ͑Butler, WI͒, utilizing No. SS-35 J. Vac. Sci. Technol. A 13(5), Sep/Oct 1995
0734-2101/95/13(5)/2619/5/$6.00
1995 American Vacuum Society
Susan J. Kerber: Bioreactivity of Ti implant alloys
stainless steel compound ͑alumina in an organic binder͒ fromKocour Co. ͑Chicago, IL͒. The samples were cleaned with asolvent method consisting of 5 minutes each in ultrasonicisopropanol and acetone. The test coupons were subse-quently argon plasma etched. Cleaning of implants by thismethod has been shown to yield clean surfaces that have abeneficial reaction with neighboring tissue.4–7 The sputteringsystem used for plasma etching was a Materials ResearchCorporation model 822 Sputtersphere with an argon radiofrequency ͑rf͒ plasma operated at 13.56 MHz. The 8 in. di-ameter pallet voltage was 1350 V. The etching process wasdone for 15 minutes. The samples were allowed to cool inargon before exposure to the atmosphere. The cleanliness ofthe samples was verified with XPS. The analysis system usedwas a VG ESCALAB system operating with an Al K␣ an-ode. Survey spectra were collected with a pass energy of 50eV; high resolution spectra with a pass energy of 20 eV.
FIG. 1. XPS high resolution O 1s spectra from ͑a͒ RGDS powder, ͑b͒Ti-6Al-4V exposed to distilled water, and ͑c͒ 2 mg/ml RGDS adsorbed from The cp titanium and Ti-6Al-4V samples were cut into 15 aqueous solution on Ti-6Al-4V. The centroids of deconvoluted peaks are mmϫ7 mm coupons. One coupon of each alloy was inserted noted with vertical dashes. Peaks are assigned as No. 1 TiO ͑ vertically ͑back, unpolished sides touching͒ into new, 532.0͒, No. 4 TiO2 nH2O and C–OH cleaned 9 mm test tubes. Solutions of RGDA, RGDS, and ͑532.5͒, No. 5 speculated to be oxygen* in Ti–O*–H–peptide, and No. 6speculated to be oxygen* in C–O*–H–O–Ti.
RFDS obtained from Bachem, Inc. ͑Torrance, CA͒ weremade with double distilled, deionized water ͑resistance Ͼ10M⍀) by successive dilution. The nominal peptide concentra- surface. Nitrogen on both the RGDS powder and on the tions were 2, 1, 0.5, 0.25, 0.125, 0.0625, and 0.00 mg/ml͑ peptide-exposed titanium samples occurred at 400.9 eV.
distilled water control͒. These nominal weights were cor- The C 1s peak of the RGDS powder ͑not shown͒ readily rected for the percentage of the peptide in the vial as given demonstrated the expected carbonyl and COH groups as well by Bachem: RGDA 76.3%, RGDS 90.9%, and RFDS 93.7%.
All of these weights are Ϯ3.0%. The titanium alloys were xHy and CmNn backbone. On the peptide-exposed titanium samples, the primary carbon peak coincided with exposed to the solutions for 26 hours at 25.0Ϯ0.5 °C using a that of adventitious carbon, except for minor high energy temperature controlled water bath. Each concentration ex- broadening. At higher peptide concentrations, there was in- periment for each peptide was replicated fifteen times. After creasing carbon with an increase in solution concentration, the exposure was completed, the solution was decanted and but the carbon background present on the control samples the samples were rinsed twice with double distilled, deion- was too large to allow the detection of peptide carbon at ized water and allowed to air dry. The samples were analyzed with XPS. The XPS spectra were quantified by the As with carbon, oxygen was not a good quantitative indi- ESCALAB system using the sensitivity factors of Wagner cator of peptide at the surface because it is also present not et al.8 The surface concentration of carbon, nitrogen, oxygen, only in the peptide, but also to a large degree on the surface and titanium was tabulated. The nitrogen signal from XPS is of distilled-water exposed titanium. The oxygen spectrum used by Ratner3 as a quantitative measure of the amount of from the RGDS powder is shown in Figure 1͑a͒. The primary protein present on a surface. Adsorption isotherms were ob- peak occurs at 531.2 eV, coinciding with CvO; high energy tained by plotting the sensitivity factor-corrected nitrogen͑ tailing is also found, due apparently to C–OH oxygen at 400.9 eV͒/titanium ͑458.5 eV͒ ratio as a function of solution 532.5 eV. High resolution oxygen spectra of the as-cleaned concentration. The ratios were additionally corrected to ac- titanium and distilled water-exposed titanium were very count for the different theoretical amounts of nitrogen in the similar and are illustrated in Figure 1͑b͒; the results are con- peptide. The amount of peptide on the surface cannot be precisely quantified due to uncertainties in sensitivity factors 530.7 eV and rutile titanium dioxide occurs at 530.6 eV.9 and attenuation lengths in these systems. Throughout this There was high energy broadening ͓Figure 1͑b͔͒. High reso- study, the data are used on a comparison basis. Additionally, lution XPS of the oxygen peaks for the samples exposed to a powder sample of RGDS peptide was analyzed by embed- peptide solutions yielded additional information about the ding the material in indium foil and immediately inserting adsorption processes. A series of peaks ͓Figure 1͑c͔͒ ranging from 531.5 to 533.5 eV increased ͑relative to the main oxy-gen peak at 530.7 eV͒ with increasing peptide solution con-centration. The spectra depicted in Figure 1 are typical of III. RESULTS
both alloys and all three peptides analyzed.
XPS analysis of the cleaned samples indicated the two As expected, no aluminum was detected on the surface of alloys have a relatively similar surface composition. A small cp titanium; only a small Al 2p peak at 75.2 eV was visible amount of nitrogen detected on the surface of the control on the Ti-6Al-4V. This peak is due to the native oxide from samples occurred at 397.4 eV. This correlated to nitride at the aluminum.8 No vanadium was detected. The low aluminum J. Vac. Sci. Technol. A, Vol. 13, No. 5, Sep/Oct 1995
Susan J. Kerber: Bioreactivity of Ti implant alloys
titanium and Ti-6Al-4V. There was more RGDA adsorbed oncp titanium than on Ti-6Al-4V.
IV. DISCUSSION
A. Interpretation of the O 1s binding energy
The complexity of these spectra precludes any absolute judgements, but the many repetitive detailed features that aredetected and their obvious relationship with the results ofprevious studies strongly support the following scenario. ͑1͒The O 1s spectrum of the RGDS powder ͓Figure 1͑c͔͒ con-firms the substantial presence of carbon–oxygen groups withthe major peak at about 531.2 eV, typical of the presence of FIG. 2. Adsorption isotherms of RFDS, RGDS, and RGDA on cp titanium.
CvO units with obvious tailing upfield, indicative of C–OHunits. ͑2͒ The O 1s spectra for the water-exposed titanium ͓Figure 1͑b͔͒ is typical of the oxide portion of TiO surface concentration and lack of detectable vanadium is high energy tailing in the spectrum reflects the terminal hy- similar to the data of Maeusli.10 Low levels of aluminum were only sporadically detected on the survey scans of the peptide-exposed Ti-6Al-4V. Because of this fact, the surface 2 is well known to exhibit several n values that should result in a manifold of peaks, the most aluminum was not included in the peptide:alloy ratio. The prominent of which should occur about 1.3 eV upfield from qualitative ramifications of this omission are discussed sub- that for the oxides, i.e., at about 532 V. ͑3͒ For the peptide- exposed samples, there is a series of peaks suggested by the Typical titanium spectra were obtained from both alloy numerical designations in Figure 1͑c͒. First, there is evidence surfaces, with 2p peaks occurring at 458.5 eV ͑due to Ti4ϩ in peak 1 for the retention of substantial amounts of largely titanium oxide͒ and at 453.8 eV ͑metallic titanium͒. There were no substantial changes in the titanium metal or titanium 2 . It should be apparent from the compari- son with Figure 1͑b͒ that most of the balance of the upfield O oxide peak energies as a function of peptide concentration.
1s structure found in Figure 1͑c͒ is due to the presence of Because the TiO2 signal was so strong, peaks from titanium and interaction with the peptide. The broadened low binding hydroxides were obscured. Ti͑OH͒4 has been reported to oc- energy structure suggests an additional peak near point 2 that cur as a leading edge of the Ti 2 p3/2 peak at 457.9 eV. One is indicative of the CvO part of the metal adsorbed carboxyl may presume that titanium hydroxides are present because of the oxygen upfield tailings that are consistent with a hydrated significant peak structures in the vicinity of positions 5 and 6 surface ͓Figure 1͑b͔͒. The titanium peak ratio of metal to ͑about 533 eV͒ are indicative of adsorptive bonding through oxide was also essentially constant. It can be argued that the peptide was adhering in discrete islands with a thickness ofgreater than 50 Å. The size of the peptides are on the order of the analysis depth for XPS—30–50 Å.
Adsorption isotherms are illustrated in Figures 2 and 3; carboxyl unit, thus confirming that the peptide is interacting error bars represent one standard deviation of 15 analyses.
with the titania with its oxygen down. On the other hand, With the exception of one data point ͑0.91 mg/ml RGDS on peaks 2– 4 are indicative of unattached units of the peptide cp titanium͒ out of 42 points, a surface concentration pro- plus very important reflections of the retention of titanium gression of RFDSϾRGDSϾRGDA was found on both cp metal hydroxide peak structures that still exist on the outersurface of the metal oxide. Some of the latter reflect undis-turbed surfaces of TiO • 2 n H2O, but some of these metal hy- droxide peaks are additionally shifted to new positions ͑peak5͒, suggesting possible involvement with the adsorbed pep-tide. These shifts may reflect the creation of a hydrogenbond. Surface reactions with organic acids can occur by elec-tron donation:13 FIG. 3. Adsorption isotherms of RFDS, RGDS, and RGDA on titanium-6aluminum-4 vanadium.
JVST A - Vacuum, Surfaces, and Films
Susan J. Kerber: Bioreactivity of Ti implant alloys
with the transfer of electrons and hydrogen ions in the direc- In RGDS, the final amino acid of RGDA ͑alanine, –CH3 tions shown. R represents the balance of the peptides. Using side chain͒ is replaced with serine ͑–CHϪ–OH side chain͒.
the above diagram, it is speculated that peak 5 could be due Serine hydrogen bonds through the removal of its hydrogen to the oxygen* in Ti–O*–H-peptide and peak 6 could be due atom. Theoretically, it should be easier for RGDS than to C–O*–H–O–Ti. Neither of these oxygen states was RGDA to form a hydrogen bond to the surfaces because present in the initial systems, consistent with the XPS data there is a higher likelihood that the molecule is in the correct obtained. It should be noted that several characteristics of the orientation for hydrogen bonding to occur in at least one of materials involved support this supposition. For example, the the three molecular sites. Therefore, it is expected that more surface titanium involved is completely oxidized before ad- RGDS would be on the surface of the materials if hydrogen sorption of the peptide units, thus no direct positions exist on bonding was the dominant force; this was indeed found.
the metal itself to accommodate the OϪ unit of the peptide.
More RGDS than RGDA was detected on the surface of both Also, the latter ͑OϪ) is not compatible with direct attachment to the oxygens of any terminal TiO2 . Therefore, it seems RGDS can be altered by changing the second position consistent to reason that the attachment of at least some of glycine amino acid ͑–H side chain͒ to phenylalanine the peptide to the titania surface is through surface hydroxyls ͑–CH2–benzyl group side chain͒ to become RFDS. The phe- in the form of a hydrogen bond. As one finds in the case of nylalanine side chain is very hydrophobic.14 RFDS could be DNA and related systems, there is also a driving force for expected to bond to the surface of hydrated titanium oxide in organic units such as the one illustrated above to form a the same fashion as RGDS; the end group acid, aspartic acid, and serine are all present in the same configuration as in The terminal ends of the tetrapeptide groups each have an RGDS and all near the terminal side. As a result, the hydro- organic acid of the form shown above for reaction with the phobic phenylalanine is probably pointed outward, away hydrated titanium oxide surface. In addition, each of them from the surface. Water tends to order at this type of hydro- has an aspartic acid ͑abbreviated D͒ group that can allow phobic interface. Because this ordering is entropically unde- bonding in this fashion and RFDS and RGDS also contain sirable, there is a driving force to minimize the water/ serine ͑S͒, with an OH-containing side chain. Thus, addi- phenylalanine interface in solution. This could cause the tional bonding seems to occur between the OH and OϪ form- adsorption of a second layer, with the phenylalanine side ing six-member ring structures, very similar to the manner chains grouping together, in much the same fashion as hy- that paired carboxylic acids form two hydrogen bonds.
drophobic forces are responsible for protein folding. Inagreement with this theory, the concentration of RFDS was B. Relative peptide surface concentration
generally higher than RGDS on each alloy.
This type of double layer adsorption behavior has been Trends in the data are visible. There is a pattern in the noted in the literature for protein structures. For example, an surface concentration of RFDSϾRGDSϾRGDA on both al- extensive study of protein adsorption isotherms on polymeric loys; this is more apparent above concentrations of 0.4 mg/ biomaterials was published by Young et al.;15 the isotherms ml. The only difference in the surface composition of the two showed the possibility of two layers of adsorbed protein.
alloys studied is a small amount of aluminum on the surface Studies by Arnebrant et al.16 of hydrophilic substrates of Ti-6Al-4V. The native surface of cp titanium is negatively showed that the plateau values of the adsorption isotherms charged while the native surface of Ti-6Al-4V is relatively correspond to a bilayer. In that case, the protein adsorbed more positively charged because of the aluminum.1 The net into a bilayer, with the bottom layer unfolded and attached charge of RFDS is 0, of RGDS it is ϩ1, and of RGDA it is by strong polar bonds to the surface. On top of that layer, 3. If net peptide charge was the strongest motivation for additional protein molecules are attached by hydrophobic in- the adsorption to occur, the peptide with the highest surface teraction and/or ionic forces. The upper layer results in large concentration on cp titanium should be RGDA because it has electrical charges. Similarly, Johnston et al. reported17 that, the highest net positive charge. This was not the case; in- as the bulk fibrinogen concentration was increased, the pro- deed, RGDA had the lowest concentration on cp titanium.
tein fractional coverage detected on polytetrafluoroethylene This lack of effect is due to the high dielectric constant of reached a constant value but the adsorption isotherm contin- water, i.e., only when unlike charges are very close and there ued to increase, indicating multilayer growth in patches on are no water molecules in between would this type of elec- the surface. This was consistent with the incomplete cover- age indicated by their XPS results. Incomplete substrate cov- RGDA, with an end group acid and aspartic acid avail- erage was also detected in our titanium study.
able, is hypothesized to form a hydrogen bonded structure onthe surface. Because of the electropositive nature of the hy- C. Proposed mechanism for the shape of the RFDS
drogen atom in covalent bonds, fewer hydrogen bonds would isotherm
form on the relatively electropositive surface of Ti-6Al-4Vthan on the more electronegative surface of cp titanium. This The isotherms for RFDS are illustrated in Figures 2 and 3.
seems to be the case. Inspecting Figures 2 and 3, there ap- The plots for cp titanium and Ti-6Al-4V are different in pears to be slightly more RGDA on cp titanium than on form, with substantially smaller standard deviations in the Ti-6Al-4V. This difference would be accentuated if a correc- data than the other peptides. Although not fully illustrated in tion could have been done for the amount of aluminum at the the plots, there were no overlaps in the surface concentration error bars above 0.25 mg/ml. The plot for cp titanium is a J. Vac. Sci. Technol. A, Vol. 13, No. 5, Sep/Oct 1995
Susan J. Kerber: Bioreactivity of Ti implant alloys
classic Brunauer–Emmett–Teller ͑BET͒ isotherm with a Differences in the isotherms were noted and possible expla- monolayer concentration corresponding to a solution concen- nations discussed. In correlation with the high resolution tration of 0.1 mg/ml. The BET theory implies that multilayer oxygen XPS data, it has been postulated that biomolecules adsorption is present. Within the concentration range used, may adsorb differently on cp titanium and Ti-6Al-4V due to the amount of adsorbed peptide does not plateau. In contrast, the difference in the abilities of the alloys to form hydrogen RFDS on Ti-6Al-4V shows a distinct bilayer adsorption, bonds and the ultimate effect that this has on hydrophobic with each layer behaving in accordance with Langmuir interactions. There was no correlation found between the net charge of the peptide groups and their adhesion to the alloys.
A possible mechanism for RFDS adsorption can be pro- posed. At low solution concentrations, the major adhesionmechanism is the formation of a hydrogen bond through theterminal acid group, the aspartic acid, and the serine end ACKNOWLEDGMENTS
group. The nitrogen/titanium ratio is the same for both alloysand the distribution of RFDS is random on the surface The author wishes to express her thanks to David Licht- throughout the solution concentration range of 0–0.25 mg/ man and David Amrani for their support in planning these ml. As solution concentrations continue to increase, the experiments and to Tery Barr for his help in interpretation of RFDS continues to be adsorbed onto the initial surface of cp titanium; additionally, hydrophobic interactions with the ex-posed phenylalanine can begin and a second layer is ad-sorbed in spots. It is conceivable that adjacent, bonded RFDSmolecules could rearrange themselves on the surface due to hydrophobic forces of adjacent phenylalanine groups. Be- H.A. Luckey, in Titanium for Energy and Industrial Applications, editedby D. Eylon ͑American Institute of Mining, Metallurgical and Petroleum cause the ability to form hydrogen bonds is high, the RFDS Engineers, New York, 1981͒, pp. 293–312.
density is relatively high on cp titanium and this rearrange- 2C. Johansson, J. Lausmaa, M. Ask, H.A. Hansson, and T. Albrektsson, J.
Biomed. Eng. 11, 3 ͑1989͒.
3R.W. Paynter and B.D. Ratner, in Surface and Interfacial Aspects of Bio- In contrast, the mechanism for Ti-6Al-4V could be differ- medical Polymers, Protein Adsorption Vol. 2, edited by J.D. Andrade ent because it is more electropositive than cp titanium. At ͑Plenum, New York, 1985͒, pp. 189–216.
low solution concentrations, the major adhesion mechanism 4M. Meenaghan, J.R. Natiella, J.L. Moresi, H.E. Flynn, J.E. Wirth, and is once again bonding through the terminal acid group, the R.E. Baier, J. Biomed. Mater. Res, 13, 631 ͑1979͒.
5J.H. Doundoulakis, J. Prosthet. Dent. 58, 471 ͑1987͒.
aspartic acid, and the serine. As solution concentrations in- 6G.L. Grobe, PHI Interface 11, 6 ͑1988͒.
crease, the sites for the hydrogen bonding on the alloy begin 7B. Liedberg, I. Lundstro¨m, C.R. Wu, and W.R. Salaneck, J. Colloid In- to be used up, sooner than on cp titanium. The surface con- terface Sci. 108, 123 ͑1985͒.
centration of RFDS increases and becomes saturated on Ti- C.D. Wagner, W.M. Riggs, L.E. Davis, J.F. Moulder, and G.E. Mullen-berg, Handbook of X-Ray Photoelectron Spectroscopy ͑Perkin-Elmer Cor- 6Al-4V at a lower solution concentration than for cp tita- nium. The RFDS groups on Ti-6Al-4V may be too far apart 9R.N.S. Sodhi, A. Weninger, J.E. Davies, and K. Sreenivas, J. Vac. Sci.
to minimize the hydrophobic surface energy by rearrange- Technol. A 9, 1329, ͑1991͒.
ment and as a result they could then adsorb additional hy- P.A. Maeusli, P.R. Bloch, V. Geret, and S.G. Steinemann, in Biologicaland Biomechanical Performance of Biomaterials, edited by P. Christel drophobic RFDS groups from solution. This is done rela- ͑Elsevier Science, Amsterdam, 1986͒, p. 57.
tively quickly until saturation of the initial RFDS layer is 11N.S. McIntyre, in Practical Surface Analysis by Auger and X-Ray Photo- complete. After this second layer of adsorption has occurred, electron Spectroscopy, edited by D. Briggs and M.P. Seah ͑Wiley, New there is no hydrophobic outer layer to allow additional ad- 12T.L. Barr, Modern ESCA ͑Chemical Rubber, Boca Raton, FL, 1994͒, p.
sorption, and equilibrium has been attained. Consistent with the data observed, this proposed mechanism could allow for 13K.L. Mittal, Pure Appl. Chem. 52, 1295 ͑1980͒.
a gradual increase of RFDS on cp titanium and bilayer ad- R.J. Fletterick, Molecular Structure: Macromolecules in Three Dimen-sions ͑Blackwell Scientific, Palo Alto, CA, 1985͒, pp. 42– 44.
15B.R. Young, W.G. Pitt, and S.L. Cooper, J. Colloid Interface Sci. 124, 28
V. CONCLUSION
16T. Arnebrant, B. Ivarsson, K. Larsson, I. Lundstro¨m, and T. Nylander, Prog. Colloid Polym. Sci. 70, 62 ͑1985͒.
XPS nitrogen/titanium ratios were used to obtain adsorp- 17A.B. Johnston, B.D. Ratner, and R.A. Horbett, The Third World Bioma- tion isotherms of RGD-based peptides on titanium alloys.
terials Congress, April 1988, Japan ͑unpublished͒.
JVST A - Vacuum, Surfaces, and Films

Source: http://www.uran.donetsk.ua/~masters/2009/fizmet/leoha/library/lib6.pdf