Silanization with APTES for Controlling the Interactions Between Stainless Steel and Biocomponents: Reality vs Expectation

These contribution books collect reviews and original articles from eminent experts working in the interdisciplinary arena of biomaterial development and use. From their direct and recent experience, the readers can achieve a wide vision on the new and ongoing potentialities of different synthetic and engineered biomaterials. Contributions were selected not based on a direct market or clinical interest, but based on results coming from very fundamental studies. This too will allow to gain a more general view of what and how the various biomaterials can do and work for, along with the methodologies necessary to design, develop and characterize them, without the restrictions necessarily imposed by industrial or profit concerns. The chapters have been arranged to give readers an organized view of this research area. In particular, this book contains 25 chapters related to recent researches on new and known materials, with a particular attention to their physical, mechanical and chemical characterization, along with biocompatibility and hystopathological studies. Readers will be guided inside the range of disciplines and design methodologies used to develope biomaterials possessing the physical and biological properties needed for specific medical and clinical applications.


Introduction
The surface of biomaterials is frequently chemically modified with the aim to modify the physicochemical properties (hydrophobicity, electrical charge, solvation) which control the interactions with biomolecules and consequently with cell surfaces, or to retain biochemical entities which are specifically recognized by the cells (Williams, 2010). Regarding inorganic materials, widespread procedures involve self-assembly of alcane thiols on gold, silver, copper or platinum (Wink et al., 1997). However, these substrates have limited interest in biomedical applications. Other procedures consist in grafting organosilanes on silica and other metal oxides (Weetall, 1993). The use of silane coupling agents has been reported in various biomaterials researches, such as surface modification of titanium (Nanci et al., 1998), natural fiber/polymer composites (Xie et al., 2010) or dental ceramics (Matinlinna et al., 2004;Matinlinna & Vallittu, 2007). The silanization reaction at interfaces is complex and there is still considerable debate on the retention mechanisms and on the organization of the interface (Gooding & Ciampi, 2011;Haensch et al., 2010, Suzuki & Ishida, 1996. Depending on the nature of reactive moieties bound to Si in the silane (typically Cl or alkoxy group) and their number, and on the reaction conditions (particularly the presence of water), the relative importance of covalent binding to the surface, oligomerization, polymerization along the surface plane, threedimensional polymerization may possibly vary. The efficiency of the surface modification is often demonstrated by its influence on biochemical or biological activity. However the nature of the interface produced is difficult to characterize, which limits the guidelines available to improve the procedures. Moreover organic contaminants are always present on high energy solids. They are mainly of hydrocarbon nature and are readily adsorbed from surrounding air or in surface analysis spectrometers (Caillou et al., 2008;Landoulsi et al., 2008a). The possible influence of contaminants on the silanization process and product is usually not considered. In the case of silicon wafer silanized with 3-[methoxy(polyethyleneoxy)]propyl trimethoxysilane and trichlorosilane in organic solvents under a controlled atmosphere, the surface obtained was described as a 1 to 2 nm thick grafted silane layer covered by a thin layer of adventitious contaminants, suggesting that contamination was posterior to the silanization reaction. On the other hand, the silane layer was not stable in phosphate buffered saline at 37°C (Dekeyser et al., 2008). Aminopropylalkoxysilanes are attractive for surface modification (Plueddemann, 1991), as their bifunctional nature is expected to offer the possibility of covalently attaching a biomolecule, either directly or through a linker. 3-Aminopropyl(triethoxysilane) (APTES) is one of the most frequently used organosilane agents for the preparation of amineterminated films (Asenath Smith & Chen, 2008;Howarter & Youngblood, 2006;Kim et al., 2009a;Lapin & Chabal, 2009;Pasternack et al., 2008). Table 1 presents a list of references in which APTES was used to hopefully graft biomolecules on different substrates. The survey is exhaustive for stainless steeel substrates relevant for the field of biomaterials and illustrative for other substrates. Additional references are: El-Ghannam et al., 2004;Kim et al., 2010;Sasou et al., 2003;Sarath Babu et al., 2004;Quan et al., 2004 ;Subramanian et al., 1999 ;Jin et al., 2003 ;Cho & Ivanisevic, 2004 ;Katsikogianni & Missirlis, 2010 ;Sordel et al., 2007 ;Toworfe et al., 2006 ;Balasundaram et al., 2006 ;Doh & Irvine, 2006 ;Palestino et al., 2008 ;Son et al., 2011 ;Koh et al., 2006 ;Mosse et al., 2009 ;Weng et al., 2008 ;Charbonneau et al., 2011 ;Iucci et al., 2007 ;Chuang et al., 2006 ;Schuessele et al., 2009 ;Ma et al., 2007 ;Toworfe et al., 2009 ;Zile et al., 2011 ;Sargeant et al., 2008 ;Lapin & Chabal, 2009. Table 1 indicates the substrate and linker used, the main conditions of the APTES treatment and the evaluation of the surface treatment regarding biomolecule activity with the blank used for comparison. The table also presents the main data obtained by surface characterization. In some systems, no covalent grafting was aimed. In other systems, although it was aimed, there is no direct evidence for the formation of covalent bonds between the biomolecules and the substrate surface. On the other hand, the evaluation of the bio-efficacy was never based on comparisons involving a complete set of blanks: treatment with the biomolecule without silanization, without linker, without silanization and linker. In a study of surface modification with the aim to enhance mineralization, it has been demonstrated that APTES-coated glass retains a homopolymer with monoester phosphate groups, poly[(2-methacryloyloxy)ethyl phosphate], by proton transfer and electrostatic interaction, while the retention of a neutral homopolymer, poly[2-(acatoacetoxy)ethyl methacrylate], was attributed to covalent linkage by reductive amination between the keto groups of the polymer and the surface amine functions (Jasienak et al., 2009). The retention of the diblock copolymer seemed to occur via segments allowing covalent bonds to be formed. In Table 1, several systems show an improved behavior which may only be attributed to non covalent bonding between the biomolecule and the silanized substrate. In contradiction with frequent implicit considerations, the occurrence or improvement of bioactivity as a result of surface treatments does not demonstrate that the chemical schemes which motivated the treatments worked in reality. This question is crucial as many organic reactions that work well in solution are difficult to apply at solid surfaces (Kohli et al., 1998).
Stainless steels (SS) are extensively used in biomaterials researches and other applications involving contact with biologic compounds, owing to their adaptable mechanical properties, their manufacturability and their outstanding corrosion resistance. For instance, SS may be used in the manufacture of vascular stents, guide wires, or other orthopedic implants (Hanawa, 2002, Ratner, 2004. In these conditions, SS are subjected to the adsorption of biomolecules (proteins, polysaccharides, lipids) and biological materials (cellular debris). The surface modification of SS may thus be important to orient the host response as desired. The present work is dedicated to the surface composition of 316L SS surfaces at different stages of the procedure used to graft a protein via the use of APTES and of a bifunctional agent expected to link the NH 2 -terminated silane with NH 2 groups of the protein. Glucose oxidase was chosen as a model protein for reasons of convenience owing to previous works related to microbiologically influenced corrosion (Dupont et al., 1998, Landoulsi et al., 2009, Landoulsi et al., 2008b, Landoulsi et al., 2008c. A particular attention is given to (i) the real state of the interface (composition, depth distribution of constituents) at different stages and (ii) the mode of protein retention. Therefore, X-ray photoelectron spectroscopy is used in a way (angle resolved measurements, reasoned peak decomposition, validation by quantitative relationships between spectral data) to provide a speciation in terms of classes of compounds (silane, protein, contaminants), using guidelines established in previous works Rouxhet & Genet, 2011). Water contact angle measurements are used to address the issue of the presence of contaminants and the perspectives of avoiding it.

Surface treatment procedure
The silanization of nat samples was performed in the gas phase with APTES. To this end, nat samples were placed, together with a small vial containing 100 µL APTES, in a 7 dm 3 closed recipient under vacuum for 30 min at room temperature. The samples were cured for 1 h at 100°C under vacuum, rinsed and incubated for 6 h in milliQ water to eliminate the excess of non attached silanes. These are called "sil". Both nat and sil samples were subjected to treatments with the coupling agent (BS) and/or the enzyme (Gox) as detailed below: i. "+BS" and "sil+BS" obtained after BS treatment (10 mM in milliQ water) for 30 min on nat and sil samples, respectively, ii. "+Gox" and "sil+Gox" obtained after Gox treatment (0.1 mg.mL -1 in phosphate buffer pH~6.8) for 2 h, on nat and sil samples, respectively, iii. "+BS+Gox" and "sil+BS+Gox" obtained after both BS and Gox treatment (according to the procedure described above) on nat and sil samples, respectively. After BS or Gox treatment, the samples were rinsed three times with milliQ water and dried under nitrogen gas flow. In order to clarify the issue of surface contamination, nat samples were further cleaned by UV-ozone treatment (UVO Cleaner, Jelight Co, Irvine, Ca, USA) during 20 minutes. They were then placed in a Petri dish, left in the laboratory environment, and submitted to water contact angle measurements as a function of time.

X-ray photoelectron spectroscopy
XPS analyses were performed using a Kratos Axis Ultra spectrometer (Kratos Analytical, UK), equipped with a monochromatized aluminum X-ray source (powered at 10 mA and 15 kV) and an eight channeltrons detector. No charge stabilization device was used on these conductive samples. Analyses were performed in the hybrid lens mode with the slot aperture; the resulting analyzed area was 700 µm 300 µm. A pass energy of 40 eV was used for narrow scans. In these conditions, the full width at half maximum (FWHM) of the Ag 3d 5/2 peak of clean silver reference sample was about 0.9 eV. The samples were fixed on the support using a double-sided adhesive conducting tape. The pressure in the analysis chamber was around 10 -6 Pa. The photoelectron collection angle θ between the normal to the sample surface and the electrostatic lens axis was 0° or 60°. The following sequence of spectra was recorded: survey spectrum, C 1s, O 1s, N 1s, P 2p, Cr 2p, Fe 2p, Ni 2p, Mo 3d, Na 1s, S 2p, Si 2p and C 1s again to check for charge stability as a function of time, and absence of sample degradation. The binding energy scale was set by fixing the C 1s component due to carbon only bound to carbon and hydrogen at 284.8 eV. The data treatment was performed with the Casa XPS software (Casa Software Ltd., UK). The peaks were decomposed using a linear baseline, and a component shape defined by the product of a Gauss and Lorentz function, in the 70:30 ratio, respectively. Molar concentration ratios were calculated using peak areas normalized according to the acquisition parameters and the relative sensitivity factors and transmission functions provided by the manufacturer.

Atomic force microscopy (AFM)
The surface topography was examined using a commercial AFM (NanoScope III MultiMode AFM, Veeco Metrology LLC, Santa Barbara, CA) equipped with a 125 µm 125 µm 5 µm scanner (J-scanner). A quartz fluid cell was used without the O-ring. Topographic images were recorded in contact mode using oxide-sharpened microfabricated Si 3 N 4 cantilevers (Microlevers, Veeco Metrology LLC, Santa Barbara, CA) with a spring constant of 0.01 N.m -1 (manufacturer specified), with a minimal applied force (<500 pN) and at a scan rate of 5-6 Hz. The curvature radius of silicon nitride tips was about 20 nm. Images were obtained at room temperature (21-24°C) in milliQ water. All images shown in this paper were flattened data using a third order polynomial. The surface roughness (R rms ) was computed over an area of 1 µm 1 µm using the Veeco software.

Water contact angle measurements
Water contact angles were measured at room temperature using the sessile drop method and image analysis of the drop profile. The instrument, using a CCD camera and an image analysis processor, was purchased from Electronisch Ontwerpbureau De Boer (The Netherlands). The water (milliQ) droplet volume was 0.3 L, and the contact angle was measured 5 s after the drop was deposited on the sample. For each sample, the reported value is the average of the results obtained on 5 droplets.

Results
AFM images obtained in water on SS samples after different treatments are presented in Figure 1. The nat sample showed the presence of nanoparticles with different sizes (nat, Figure 1, R rms = 3.2 nm), in agreement with previous results. The formation of nanoparticles, presumably made of ferric hydroxide, resulted from oxidation occurring during the 48 h of immersion subsequent to polishing (Landoulsi et al., 2008a). The surface of silanized SS exhibited particles with a bigger size (sil, Figure 1) and the roughness decreased slightly (sil, Figure 1, R rms = 2.4 nm). The treatment with Gox, with or without previous treatment with BS, led to the formation of particles with a more uniform size and a higher density, in comparison with nat sample, and no appreciable change of surface roughness (R rms = 1.7 nm for sil+Gox and 2.5 nm for sil+BS+Gox, Figure 1). XPS is a suitable method to obtain information regarding the different constituents at the surface (substrate, silane, other organic compounds). The elemental composition of the samples is given in Table 2. Representative C 1s and O 1s peaks recorded on SS surface prior to and after silanization, and after further treatments are given in Figure 2. After BS and/or Gox treatment of silanized SS, a relative increase of the component around 531.2 eV in the O 1s peak was observed (Figure 2). In the C 1s peak, an increase of the components at 286.3 and 288.7 eV was also clear, while the main component remained at 284.8 eV. The same www.intechopen.com tendency regarding the C 1s and O 1s peaks was also observed after BS and/or Gox treatment of nat sample, but was less pronounced (data not shown). It may be attempted to Fig. 1. AFM height images (1 1 µm 2 , contact mode, in water; z scale 10 nm) of native (nat), silanized stainless steel (sil), the same after adsorption of glucose oxidase (sil+Gox) and after treatment with glucose oxidase subsequent to treatment with the coupling agent (sil+BS+Gox). Cross sections were taken at the place indicated by the dashed lines. www.intechopen.com extract chemical information by careful decomposition of the peaks. This requires to impose reasonable constraints (number of components, full width at half maximum FWHM) in order to insure reliable comparisons, and to check the chemical relevance of the results by examining correlations between spectral data of different natures Rouxhet & Genet, 2011). In previous studies (Landoulsi et al., 2008a;Landoulsi et al., 2008b), we have demonstrated the usefulness of this approach, even when the evolution of the C 1s and O 1s peak shape is weak, in order to obtain information on the amount and the nature of organic and inorganic constituents on SS surfaces.  Fig. 2. O 1s and C 1s peaks of native (nat), of silanized stainless steel (sil), of the same after treatment with the coupling agent (sil+BS), after adsorption of glucose oxidase (sil+Gox) and after treatment with glucose oxidase subsequent to treatment with the coupling agent (sil+BS+Gox). Figure 3 presents typical O 1s, N 1s and C 1s XPS peaks recorded on native SS (nat), silanized (sil) and the same after Gox treatment (sil+Gox). For the decomposition of these peaks, reasonable constraints were applied, based on our experience with the XPS analysis of biosurfaces , Rouxhet & Genet, 2011. The C ls peak was decomposed in four components, the FWHM of which were imposed to be equal: (i) a component at 284. The O 1s peak was decomposed in three components (Landoulsi et al., 2008a Fig. 3. Decomposition of O 1s, N 1s and C 1s peaks recorded on: native stainless steel (nat), silanized stainless steel (sil) and the same after adsorption of glucose oxidase (sil+Gox). The N 1s peak is overlapped with a Mo 3p 3/2 contribution.
The N 1s peak showed a main component at 400.0 eV attributed to amide or amine (N-C). An additional component appeared clearly near 401.6 eV in silanized SS samples (sil and sil+Gox, Figure 3), indicating the presence of protonated amines. The contributions at lower binding energies in the N 1s spectral window are due to Mo 3p 3/2 components (Olefjord & Wegrelius, 1996); the reliability of their quantification was checked by comparison with the Mo concentration deduced from the Mo 3p 3/2 and Mo 3d peaks (Landoulsi et al., 2008a). The N 1s contribution of sil shows a shape which is in agreement with spectra reported in the literature for APTES-modified surfaces (Suzuki et al., 2006;Xiao et al., 1997). It was not found justified to decompose it in three components attributed to amine, amide and protonated amine, respectively, as done in (Suzuki et al., 2006). The surface concentrations (mole fraction) associated with the components of C 1s, O 1s and N 1s peaks are given in Table 2.  Table 2. Surface concentration (mole fraction (%) computed over the sum of all elements except hydrogen) of elements determined by XPS (θ = 0°) on stainless steel samples.
The Si 2p peak was decomposed in two components, at 99.3 and 101.8 eV, attributed to non oxidized silicon in SS (Si 0 ) and to silicon of silane (Si org ), respectively. It was not decomposed in Si 2p 3/2 and Si 2p 1/2 contributions because these are very close in energy. The decomposition procedure for Fe 2p 3/2 , Cr 2p, Ni 2p 3/2 , and Mo 3d peaks was described before (Landoulsi et al., 2008a). The Fe concentration may be underestimated due to the procedure used to treat the complex baseline of the Fe 2p peak. The distinction between contributions of oxidized (M ox ) and nonoxidized (M met ) metal elements was easily made. The concentrations obtained are also given in Table 2 (1) where the name of an element in italic designates its concentration and the number in subscript designates the binding energy of the peak component. Errors would occur in case of a high concentration of polysaccharides (C ox /O = 6/5) or carboxyl (C ox /O = 1/2) Landoulsi et al., 2008a). The sum of the concentrations of the elements present in organic compounds is then given by: For sake of uniformity, all spectral data involved in correlations below are ratioed to ∑org ( Table 2). The concentration of the main elements or functions due to organic compounds, obtained at photoelectron collection angle θ = 0°, is plotted in Figure 4 as a function of the same quantity obtained at θ = 60°. A 1:1 relationship is obtained for all elements or functions and all samples, indicating no significant effect of the photoelectron collection angle θ on the relative contribution of the constituents of the organic adlayer. Table 2 shows that the apparent concentrations of metal elements varied only slightly according to the surface treatment. The main change concerned the decrease of the Fe ox concentration for sil+BS, sil+Gox and sil+BS+Gox samples. However, no change in the shape of the Fe 2p 3/2 peak was observed (data not shown). For these samples, a significant decrease of the molar concentration of O 529.7 was also noticed ( Table 2). It appears that the oxide layer of SS passive film, after incubation in the aqueous medium for 48 h (nat sample), was mainly constituted with a mixture of Fe and Cr oxides/hydroxides and small amounts of partially oxidized Ni and Mo. This is in agreement with a previous study (Landoulsi et al., 2008a), however in the latter, the stoichiometry of the passive film was not computed. The O 1s component at 529.7 eV is due to metal oxides. By considering that Mo ox is in the form of MoO 3 (Landoulsi et al., 2008a), the difference between the O 529.7 concentration and three times the Mo ox concentration should be due to Fe and Cr oxides. Figure 5 presents the relation between this difference and the sum of Fe ox and Cr ox concentrations. All data show reasonable linear regressions. The shift of the dots along the line when the photoelectron collection angle changes from 0° to 60° is due to the presence of the organic constituents on top of stainless steel. The average ratio between the y and x scales is 0.80 and 0.96 at θ = 0 and 60°, respectively; the slope of the regression lines is 1.03 (s.d. 0.23) and 1.42 (s.d. 0.12), respectively. Thus, the ratio oxide/metal ions in chromium and iron oxyhydroxides is of the order of 1 to 1.5. The evaluation of the quantity of hydroxide associated to Fe and Cr is complex due to the multiple chemical functions overlapping in the O 531.2 component. Ni ox is in the form of Ni(OH) 2 (component at ~855.6 eV (Briggs & Seah, 1990, Zhou et al., 2006, spectra not shown). The amount of oxygen associated to silane depends on the products of APTES reaction. The different possibilities, corresponding to the relative importance of grafting with respect to polymerization, are shown in Figure 6 and characterized by a, defined as the sum, over oxygen atoms which are not bound to a metal element, of the inverse of the number of bonds oxygen forms with silicon. Accordingly the concentration of inorganic hydroxide in the passive film may be given as follows:

Passive film composition
where a can take the values of 0, 0.5, 1, 1.5 or 2 ( Figure 6). Figure 7 shows the plot of the concentration of hydroxide which should be associated with Fe and Cr, considering different values of a. Taking silicon into account (a≠0) brings the silanized substrates better in line with the non-silanized substrates and the correlation improves as a increases, indicating that silane is polymerized and not just grafted. Depending on the value of a, the y : x ratio varies from 1.1 (a = 0) to 0.8 (a = 2). These observations indicate that the stoichiometry of the Cr and Fe oxyhydroxide at the surface is close to (Fe,Cr)OOH. Many studies have reported a stratification in the passive film and the presence of Fe and Cr oxyhydroxides in the outermost layer (Le Bozec et al., 2001). In our case, no significant effect of the photoelectron collection angle appears in the OH inorg /O 529.7 ratio (Figure 8), whatever the value selected for a to evaluate OH inorg . However it must be kept in mind that the surface roughness revealed by Figure 1, which is of the order of the inelastic mean free path of photoelectrons in oxides, may mask the effect of a stratification.  Table 2 reveals an increase of N tot concentration as a result of surface treatments, and an increase of Si org concentration for samples prepared with APTES treatment. However, the concentration of carbon is high and remains almost unchanged, suggesting that organic contaminants, mainly hydrocarbon-like compounds, are always dominating in the organic adlayer. If nitrogen was exclusively due to amide functions (N-C=O) as in the peptide link of proteins, and if the C 1s component at 287.8 eV was exclusivley due to amide, a 1:1 correlation would be found between the concentrations of C 287.8 and N tot . This is indeed observed (Figure 9a) for the set of samples involving the silanized substrate. As nitrogen is partly in the form of silane, relevant alternatives for the abscissa scale may be the concentration of N 400 or the difference between the concentrations of N tot and Si org . If polysaccharides were present with protein, the C 287.8 concentration should be corrected by subtracting the contribution of acetal and thus replaced by [C 287.8 − (C 286.3 − N 400 )/5] (Ahimou et al., 2007, Landoulsi et al., 2008a or [C 287.8 − (C 286.3 − N tot + Si org )/5]. A comparison between different plots in Figure 9 shows that the dots representative of samples prepared with nonsilanized substrate remain clustered. The shift of the cluster along the ordinate scale according to the plot indicates that C 287.8 concentration is higher than what can be attributed to amide. On the other hand, the samples prepared with silanized substrate preserve a unit slope whatever the plot, with much higher values of the coordinates for samples exposed to the enzyme, with or without the linker. This reveals an excellent agreement between the increases of concentrations of nitrogen and of carbon attributed to peptidic links (N−C=O), which result from the Gox treatment. It also validates the C 1s peak decomposition and component attribution. The meaning of the surface composition appears more clearly if it is summarized in terms of concentration of model molecular compounds. This approach was already used for microbial surfaces (Dufrêne & Rouxhet, 1996;Tesson et al., 2009), for food products  and for stainless steel aged in different conditions (Landoulsi et al., 2008b), considering proteins (Pr), polysaccharides (PS), and hydrocarbon-like compounds (HC) which represent mainly lipids in biological systems. The novelty here is to take silane (Sl) into account in addition to the three classes of biochemical compounds; the chemical composition and density considered for these model compounds are listed in Table 3. Accordingly, the proportion of carbon atoms due to each model compound X (C X /C tot ) can be computed by solving the following system of equations:

Chemical speciation of the organic adlayer
(N tot -Si org )/C tot = 0.273 × (C Pr /C tot ) Si org / C tot = 0.333 × (C Sl / C tot ) 1 = (C Pr / C tot ) + (C PS / C tot ) + (C Sl / C tot ) + (C HC / C tot ) The experimental concentration ratios (C X /C tot ) can be converted into weight percentages of model compounds , Rouxhet & Genet, 2011, using the carbon concentration specific to each compound (Table 3).  Table 3. Chemical composition of organic compounds (molar concentration ratio of elements and concentration of carbon) and their densities.
Results, presented in Table 4, show the often dominating presence of hydrocarbon-like compounds and polysaccharides, for all samples. Both compounds are due to adventitious contamination which may originate from adsorption from air, as always observed for high surface energy solids (Caillou et al., 2008;Mantel et al., 1995), but also from aqueous media (Landoulsi et al., 2008a). They may be taken as a global way to reflect the amount of compounds which contain hydrocarbon chains and oxygen, such as esters, but excluding proteins and silane. The concentration of silane deduced for non-silanized samples is non negligible but highly variable, and may also be attributed to contamination. A drastic increase of silane concentration is observed for all silanized samples (Table 4).  Table 4. Chemical composition (weight %) of the organic adlayer present on stainless steel samples, as deduced from XPS data and expressed in terms of classes of molecular compounds*. Thickness of the organic adlayer deduced from measurements at photoelectron collection angle θ = 0° and 60°.
The Gox treatment leads to a marked increase of the protein concentration on nat sample and a much stronger increase on sil sample. This may be attributed to physical adsorption www.intechopen.com and to the fact that protonated amine of silane favors adsorption by electrostatic attraction (Jasienak et al., 2009). The pH of the Gox solution (6.8) is indeed higher than the isoelectric point of glucose oxidase (4.9). In the above computation, the concentration of the BS coupling agent could not be evaluated. The reaction of BS with NH 2 transforms an amine function into amide. If only one end of BS reacts with silane, the N 400 concentration should be doubled, which is consistent with the increase found in Table 2. However, converting it into protein-equivalent gives a number with no physical meaning. When the second end of the coupling agent reacts with the protein, no additional nitrogen is incorporated, the evaluation of the protein concentration is correct but the suberate (CH 2 ) 6 chain is counted in the HC concentration. This has no important impact owing to the high concentration of hydrocarbon-like compounds. Despite the limitations regarding the accuracy of the data in Table 4, it is clear that prior silanization increases markedly the concentration of glucose oxidase (sil+Gox and sil+BS+Gox compared to nat+Gox); however the treatment with the coupling agent does not increase the amount of immobilized enzyme (sil+BS+Gox compared to sil+Gox).

State of stainless steel surface
It appears in Table 4 that the SS surface prior to and after silanization or enzyme immobilization is bearing a high amount of organic contaminants. It may be argued that this is due to improper cleaning protocols, inappropriate sample manipulation or contamination in the XPS spectrometer. Actually, a clean stainless steel surface is getting quickly contaminated in contact with the surrounding atmosphere, as revealed by water contact angles, which can be measured quickly in the same environment. A nat sample showed a water contact angle of 44° which increased to about 60° within a delay of a few hours (Figure 10). When a nat sample was further treated with UV-ozone to oxidize organic compounds, the water contact angle was lowered down to 12°. However it increased rapidly in contact with the surrounding atmosphere ( Figure 10) to reach values above 40° in a few hours. Similar results were obtained with 304L stainless steel. Wet cleaning essentially standardizes the surface contamination; further cleaning leaves a material with a high surface energy, which adsorbs quickly significant amounts of contaminants (Caillou et al., 2008). It must be emphasized that outgassing a material of high surface energy is not suitable to prevent adventitious contamination of the surface. Materials cleaned with UV-ozone treatment and showing a water contact angle below 5° reached an appreciable contact angle (20° for silica, 40° for stainless steel and gold) after a stay of 5 minutes in the vacuum of the spectrometer chamber. The high rate of contamination may be due to evacuation itself, owing to the increased proportion of organic compounds in the residual gas and to their increased rate of transfer (Caillou et al., 2008).

Thickness of the organic adlayer
No variation of the relative concentrations of organic compounds was revealed by angleresolved XPS analysis (Figure 4) but an effect of stratification may be masked by the surface roughness. The evaluation of the adlayer thickness may clarify whether the increase of the silane or protein concentration resulting from the respective treatments reflects an addition of or a substitution by new compounds in the adlayer.
If the surface is considered as atomically smooth and the organic layer is continuous with a constant thickness, the apparent concentration ratio i C and i Cr are the relative sensitivity factors of C and Cr, respectively, provided by the spectrometer manufacturer. The photoionization cross sections σ are 1 for C 1s and 11.7 for Cr 2p (Scofield, 1976). The superscripts Org and Ox designate the organic adlayer and the passive oxide layer, respectively. The concentration of Cr in the oxide layer ( Org Cr C ; between 0.015 and 0.020 mol.cm -3 , depending on the sample), was determined on the basis of the above discussion indicating that the oxide layer is constituted with FeOOH, CrOOH, Ni(OH) 2 and MoO 3 (section 4.1). Note that this is in agreement with a concentration of inorganic oxygen close to (O tot − O org ), owing to the low concentration of silane. The concentration of carbon in the organic adlayer was determined on the basis of the surface composition modeled as detailed above, using the densities given in Table 3. The density of silane, was taken as the average between the densities of 3-aminopropyl(trimethylsilane) (0.8 g.cm -3 ) and 3-aminopropyl(trimethoxysilane) (1.0 g.cm -3 ). The electron inelastic mean free paths (IMFP) were calculated using the Quases program (http://www.quases.com ) and the TPP2M formula (Tanuma et al., 1997), considering the matrix composition deduced above. For The thickness of the organic adlayer deduced for photoelectron collection angles θ = 0° and 60° is given in Table 4. The difference between the values computed for the two www.intechopen.com photoelectron collection angles is due to the approximation of a smooth surface while the analyzed surface is rough at the scale of the inelastic mean free paths. The real organic adlayer thickness should be between the values given in Table 4. Comparison between nat and sil samples suggests that silane just adds up to the contaminants. A 3.0 nm thick adlayer containing 25 wt.% silane corresponds to 4.5 molecules.nm -2 . This value is consistent with a monolayer of silane, however the retained silane is mixed with a much larger amount of contaminants. The protein treatment of silanized substrates (compare sil+Gox and sil+BS+Gox to sil) led to a significant decrease of the amount of hydrocarbon-like compounds, while the adlayer thickness did not change appreciably. This suggests that the protein adsorption caused the displacement of part of contamination present on the silanized stainless steel surface in the form of hydrocarbon-like compounds.

Conclusion
The decomposition of the C 1s and O 1s peak provided a distinct evaluation of oxygen present in inorganic oxide, inorganic hydroxide and organic compounds. This led to a stoichiometry close to (Fe, Cr)OOH for iron and chromium species in the passive layer. The elemental composition of the organic adlayer was converted into concentration of model organic compounds: protein, silane and contaminants. For the latter, extreme poles of hydrocarbons and polysaccharides were taken as models.
Silanization increased markedly the retention of glucose oxidase, whether or not using a linker expected to couple the enzyme with the silane. Direct retention of the enzyme by the silane may be attributed to electrostatic attractions with the protonated amine groups (Jasienak et al., 2009). It is thus demonstrated, with the same biomolecule, that its retention by using an APTES-treated substrate and a linker does not infer covalent grafting through the linker. The occurrence of covalent binding might possibly be evaluated by examining the retention upon aging in electrolytes, keeping in mind that the silane layer itself may alter in these conditions (Dekeyser et al., 2008). The thickness of the organic adlayer was of the order of 3 nm. In all cases, the concentration of contaminants exceeded the concentration of silane and protein. Angleresolved measurements did not reveal any stratification in the organic adlayer, but this was not conclusive since the effect of a stratification may be masked owing to the roughness created by nanoparticles of inorganic oxyhydroxide present at the stainless steel surface. The presence of organic contaminants is unavoidable when high energy materials are exposed to air for a few hours or to vacuum for a very short time. However the influence of contaminants on the desired surface reactions is not known and is difficult to establish. The best to reduce contamination is to thoroughly clean the substrate by oxidation, to minimize the time of contact of substrates with air at any stage, and to strictly avoid outgassing. Water contact angle measurement is the best way to assess the cleanliness of the native substrate but the information will not be unambiguous after silanization and further treatments.

Acknowledgments
The authors thank Simon Degand for his help in statistics. They acknowledge the support of the "Conseil Régional de Picardie" (France) and the National Fondation for Scientific Research (F.N.R.S. -Belgium). www.intechopen.com