Langmuir 2009, 25, 203-209
203
Aminosilane Multilayer Formed on a Single-Crystalline Diamond Surface with Controlled Nanoscopic Hardness and Bioactivity by a Wet Process Yosuke Amemiya,† Akiko Hatakeyama,‡ and Nobuo Shimamoto*,†,‡ Structural Biology Center, National Institute of Genetics, Mishima, Shizuoka 411-8540, Japan and Department of Genetics, School of Life Science, The Graduate UniVersity for AdVanced Studies, Mishima, Shizuoka 411-8540, Japan ReceiVed May 21, 2008. ReVised Manuscript ReceiVed August 13, 2008 Diamond could be an excellent support for nanodevices utilizing biomolecules if it is covered with a polymer layer immobilizing a variety of biomolecules. We report a wet method to form a 3-aminopropyltriethoxysilane (APTES) multilayer with a controlled hardness, roughness, and capacity for immobilizing protein. The method is feasible in typical biochemical laboratories where biomolecules are prepared. Atomic force microscopy (AFM) revealed that the surface geometries and nanoscopic hardness of the multilayers on an oxygen-terminated single-crystalline diamond surface depended on the dielectric constant of the solvent; the smaller the constant, the harder the layer. The hard multilayers had holes and APTES aggregates on the surfaces, while less hard ones had homogeneous surfaces with rare holes and little aggregates. The secondary deposition of APTES in a solvent with a large dielectric constant on a hard multilayer removed the holes, and further treatment of the multilayer in acidic ethanol solution diminished the aggregates. Such a surface can immobilize streptavidin with enough specificity against nonspecific adsorption using a combination of polyethylene glycol reagents. The results of a scratching test and nanoindentation test with AFM provided consistent results, suggesting some universality of the scratching test independent of the tip structure of the cantilever. The mechanism of formation of multilayers on the diamond surface and their binding to it is discussed.
Introduction Diamond is a promising material for nanoelectric devices or nanomechanical parts due to its stiffness, potential conductivity, and chemical stability. Because of these properties, diamond has been considered for biological applications,1 and thus, immobilization of functional biomolecules on its surface has attracted increasing attention. For utilization of the distinct properties, biomolecules must be immobilized on a diamond surface with enough stability or mechanical strength. However, the mechanical strength of the immobilization in nanoscale is greatly unknown, and there are only a few methods to test the strength. Because direct immobilization of biomolecules, especially proteins, on inorganic surfaces tends to inactivate them due to adsorption, an interface layer between a diamond surface and the functional biomolecules is generally required. Since the merit of chemical stability of diamond is also its demerit of difficulty in chemical modification, it is a challenge of general importance to develop a method for forming such an interface layer with a relevant mechanical strength in nanoscale by chemical modification. To form a hard layer on diamond the layer is preferred to be covalently bound to a diamond surface, either a hydrogenterminated (H-diamond) surface or an oxygen-terminated (Odiamond) one. They have opposite properties and thus require different processes for preparation and modification. The H-diamond surface is hydrophobic, while the O-diamond surface is hydrophilic. The H-diamond surface is prepared by exposing hydrogen plasma at elevated temperatures2 and can be photo* To whom correspondence should be addressed. Phone: +81-55-9816843. Fax: +81-55-981-6844. E-mail:
[email protected]. † National Institute of Genetics. ‡ The Graduate University for Advanced Studies. (1) Nebel, C. E.; Shin, D.; Rezek, B.; Tokuda, N.; Uetsuka, H.; Watanabe, H. J. R. Soc. Interface 2007, 1–23. (2) Kawarada, H. Surf. Sci. Rep. 1996, 26, 205–259.
chemically3-5 or electrochemically6 modified. The O-diamond surface is prepared by various processes involving acid treatments,2,7,8 anodic polarization,9 oxygen plasma treatment,10 treatment with the thermally activated oxygen,11 or irradiation of ultraviolet (UV) light in the presence of ozone.12 Formation of aminosilane layers on the O-diamond surface has been reported.12-14 However, the nature of the binding between the layers and diamond as well as the mechanical properties are mostly unknown. Only two layers on the H-diamond surface have been tested by scratching with a cantilever of atomic force microscope (AFM).15,16 Two other properties of the layer are essential to practical application: high adaptability to a wide variety of biomolecules (3) Strother, T.; Knickerbocker, T.; Russell, J. N.; Butler, J. E.; Smith, L. M.; Hamers, R. J. Langmuir 2002, 18, 968–971. (4) Takahashi, K.; Tanga, M.; Takai, O.; Okamura, H. Diamond Relat. Mater. 2003, 12, 572–576. (5) Zhang, G. J.; Song, K. S.; Nakamura, Y.; Ueno, T.; Funatsu, T.; Ohdomari, I.; Kawarada, H. Langmuir 2006, 22, 3728–3734. (6) Wang, J.; Firestone, M. A.; Auciello, O.; Carlisle, J. A. Langmuir 2004, 20, 11450–11456. (7) Ri, S. G.; Nebel, C. E.; Takeuchi, D.; Rezek, B.; Tokuda, N.; Yamasaki, S.; Okushi, H. Diamond Relat. Mater. 2006, 15, 692–697. (8) Ushizawa, K.; Sato, Y.; Mitsumori, T.; Machinami, T.; Ueda, T.; Ando, T. Chem. Phys. Lett. 2002, 351, 105–108. (9) Notsu, H.; Yagi, I.; Tatsuma, T.; Tryk, D. A.; Fujishima, A. Electrochem. Solid-State Lett. 1999, 2, 522–524. (10) Yagi, I.; Notsu, H.; Kondo, T.; Tryk, D. A.; Fujishima, A. J. Electroanal. Chem. 1999, 473, 173–178. (11) Pehrsson, P. E.; Mercer, T. W. Surf. Sci. 2000, 460, 49–66. (12) Zhang, G. J.; Umezawa, H.; Hata, H.; Zako, T.; Funatsu, T.; Ohdomari, I.; Kawarada, H. Jpn. J. Appl. Phys 2005, 44, L295-L298. (13) Hernando, J.; Pourrostami, T.; Garrido, L. A.; Williams, O. A.; Gruen, D. M.; Kromka, A.; Steinmu¨ller, D.; Stutzmann, M. Diamond Relat. Mater. 2007, 16, 138–143. (14) Notsu, H.; Fukazawa, T.; Tatsuma, T.; Tryk, D. A.; Fujishima, A. Electrochem. Solid-State Lett. 2001, 4, H1-H3. (15) Uetsuka, H.; Shin, D.; Tokuda, N.; Saeki, K.; Nebel, C. E. Langmuir 2007, 23, 3466–3472. (16) Yang, N.; Uetsuka, H.; Watanabe, H.; Nakamura, T.; Nebel, C. E. Chem. Mater. 2007, 19, 2852–2859.
10.1021/la801556x CCC: $40.75 2009 American Chemical Society Published on Web 11/26/2008
204 Langmuir, Vol. 25, No. 1, 2009
to be immobilized and high enough density of immobilization to overcome nonspecific adsorption. The former is satisfied by introducing amino residues on the interface layer,5,13,15,17-20 and the latter must be tested in each case. Furthermore, there is another property as important for progress of this technological field: feasibility of the developed method. Because the techniques in a new field are developed by trial and error and because biological systems are extremely diverse, an application to biological systems must be established by bountiful trials which are performed in various places involving standard biological laboratories. Unfortunately, the methods for the H-diamond surface as well as some for the O-diamond surface require commercially unavailable reagents or sophisticated instruments which are not readily available in a standard biological laboratory where target biomolecules are prepared, such as a hightemperature furnace, plasma generator, sensitive electrochemical instrument, and powerful UV illuminator used in an inert gas atmosphere. Therefore, we developed a wet method for forming an interface layer of a silane on diamond without using any sophisticated instruments. When 3-aminopropyltriethoxysilane (APTES), a common aminosilane, was used, the formed layers showed various nanoscopic hardness and capacity for immobilizing streptavidin depending on the solvent used for deposition process. We formed APTES layers immobilizing streptavidin with high enough density to overcome nonspecific binding. We further controlled the hardness of APTES layers so that the layer can be manipulated with AFM but mechanically stable enough to stay on the diamond surface, providing an alternative way to manipulate the hard surface of diamond. The origin of the hardness of silane layers on a diamond surface is discussed.
Experimental Section Deposition of Aminosilane on a Diamond Surface. All experiments in this section were carried out at room temperature unless otherwise noted. The surface of single-crystalline synthetic (100) Ib diamond substrate with 3 mm × 3 mm × 0.5 mm size (Sumitomo Electric Industries) was oxidized by stepwise exposure to HCl-HNO3 (3:1, v/v) for 30 min, H2SO4-H2O2 (3:1, v/v) at 160 °C for 15 min, and H2SO4-HNO3 (3:1, v/v) at 240 °C for 60 min followed by being boiled in water for 10 min and rinsed with water. APTES was deposited on the surface in a single step or two steps. In the single-step deposition, APTES was deposited for 1 h in its 2% solution of the mixture of 95% ethanol and 5% water. The pH of the ethanol mixture was adjusted to give a reading of 4.7 with acetic acid. The diamond substrate was then rinsed with the ethanol mixture and baked at 120 °C for 30 min. The substrate was washed twice for 5 min in a glass beaker containing 20 mL of the ethanol mixture, which was placed in a water-filled VC-1 ultrasonication bath (AsOne), followed by being air dried. Alternatively, APTES was deposited in its 2% solution for 10 h in an organic solvent of either acetone, n-octanol, chloroform, toluene, or n-heptane. The samples were washed in the same organic solvent with the same ultrasonication and then baked at 120 °C for 30 min. In place of the deposition in a single step, APTES deposition was carried out in two steps; first, in 2% APTES solution of toluene for 10 h and then in 2% APTES solution of the ethanol mixture or in that of acetone for (17) Coffinier, Y.; Szunerits, S.; Jama, C.; Desmet, R.; Melnyk, O.; Marcus, B.; Gengembre, L.; Payen, E.; Delabouglise, D.; Boukherroub, R. Langmuir 2007, 23, 4494–4497. (18) Ha¨rtl, A.; Schmich, E.; Garrido, J. A.; Hernando, J.; Catharino, S. C.; Walter, S.; Feulner, P.; Kromka, A.; Steinmu¨ller, D.; Stutzmann, M. Nat. Mater. 2004, 3, 736–42. (19) Rubio-Retama, J.; Hernando, J.; Lo´pez-Ruiz, B.; Ha¨rtl, A.; Steinmu¨ller, D.; Stutzmann, M.; Lo´pez-Cabarcos, E.; Garrido, J. A. Langmuir 2006, 23, 4494– 4497. (20) Yang, W.; Auciello, O.; Butler, J. E.; Cai, W.; Carlisle, J. A.; Gerbi, J. E.; Gruen, D. M.; Knickerbocker, T.; Lasseter, T. L.; Russell, J. N., Jr.; Smith, L. M.; Hamers, R. J. Nat. Mater. 2002, 1, 253–257.
Amemiya et al. 14 h. The samples were washed with the same ultrasonication in the solvent used for the second deposition and then baked under the same conditions above. In some cases, the substrate treated with APTES in toluene and then acetone was immersed in the ethanol mixture for 1 h just after the deposition process. The substrate was washed with ultrasonication in the ethanol mixture and baked under the same conditions as given above. Quantification of Amino Residues on an APTES-Coated Surface. A drop of 20 mM sulfosuccinimidyl 6-(3′-[2-pyridyldithio]propionamido) hexanoate (Sulfo-LC-SPDP) dissolved in PBS (pH 7.6) was introduced between the APTES-coated substrate and a glass slide and incubated for 2 h under the saturated water vapor pressure for prevention of drying. The modified substrate was washed manually with water and air dried. Five microliters of 20 mM dithiothreitol was spread on the modified surface and incubated for 15 min under the saturated water vapor pressure to release 2-pyridylthiol from amino-reacted Sulfo-LC-SPDP. The amount of released 2-pyridylthiol was measured at 343 nm with a spectrophotometer, model ND-100 (NanoDrop Technologies). The density of the amino residues on APTES-coated surfaces was calculated by subtracting the background value, which was obtained for an uncoated O-diamond surface. Atomic Force Microscopy. AFM, model SPI3700 (Seiko Instruments) was used for both topographic imaging and surface scratching to examine the hardness of the APTES layers.21 We calibrated a silicon AFM cantilever, SI-DF3 (Seiko Instruments), with the Cleveland method22 and selected tips with a spring constant of 1.2-1.4 N/m. The thickness of the APTES layers was determined as the difference between the observed height of the layer surface and the surface exposed by the scratch. As a control, ω-unsaturated 10-trifluoroacetic amide-dec-1-ene (TFAAD) was attached photochemically on the diamond surface,16 which was a kind gift from Dr. Nebel and Dr. Uetsuka, and the hardness of its layer was examined by scrabbling the surface. The nanoindentation test was performed by pressing the APTES layer with no horizontal movements in the contact mode and then imaging the surface in the noncontact mode. Measurement of Specificity of Immobilizing Streptavidin. For introduction of biotin residues on a part of the surface, a drop of 0.5 µL of PBS containing 10 mM N-hydroxyl succinimide-polyethyleneglycol-biotin (NHS-PEG-biotin) was placed for 3 h on a surface of diamond substrate coated with APTES. Two kinds of NHS-PEGbiotin with Mw of 5000 (Nektar) and 589 (Pierce) were used. The modified substrates were manually rinsed with water and then incubated for 12 h in PBS containing 10 mM NHS-PEG. Three kinds of NHS-PEG with a molecular weight of 5000 (Nektar) and 333 (Pierce) and their one-to-one mixture were used. The modified substrate was rinsed with water again, and a thin layer of the solution of 200 µg/mL Alexa Fluor 488-labeled streptavidin (Invitrogen) was introduced between the modified surface and a glass slide. The modified substrate was incubated for 1 h under the saturated water vapor and rinsed with PBS containing 0.5% Tween-20 and then with water. Fluorescence images of the labeled streptavidin immobilized on diamond surfaces were obtained with an IX-70 fluorescence microscope (Olympus) equipped with a PhotonMAX EMCCD camera (Princeton Instruments), 460-480 nm excitation filter, and 495-540 nm emission filter. The fluorescent intensities of the images were determined using Adobe Photoshop 7.0 (Adobe Systems). The specific binding of streptavidin to a biotin spot was indexed by the following equation
specificity ) (Iin - Iout) ⁄ (Iout - Ib)
(2)
where Iin, Iout, and Ib are the average fluorescent intensities inside the biotin spot, outside the spot, and from the background of the uncoated O-diamond surface, respectively. Quantification of Streptavidin Fixed on a Surface. Immobilization of the long-chain biotin, blocking with the short NHS(21) Xiao, X. D.; Liu, G. Y.; Charych, D. H.; Salmeron, M. Langmuir 1995, 11, 1600–1604. (22) Cleveland, J. P.; Manne, S.; Bocek, D.; Hansma, P. K. ReV. Sci. Instrum. 1993, 64, 403–405.
Aminosilane Multilayer
Figure 1. AFM images in noncontact mode after the scratching test at 100 nN. (A) APTES layer was formed by the single-step method using the ethanol mixture and subjected to the test. (B) The same image of noncoated diamond surface after the test. (C and D) Plotting of height along the lines shown in A and B, respectively.
PEG, and immobilization of Alexa Fluor 488-labeled streptavidin were carried out on the overall surface of an APTES-coated substrate as mentioned in the previous section. The immobilized streptavidin was dissociated from the surface by its heat denaturation at 98 °C for 2 h in a phosphate-sodium buffer containing 1% Triton X-100 and 1% 2-mercaptoethanol. The fluorescence intensity of the solution was measured by an F-4500 fluorescence spectrophotometer (Hitachi) with excitation at 480 nm and emission at 520 nm with a series of standard solutions of the labeled streptavidin. The distance between flanking streptavidin molecules was calculated under the assumption of the hexagonal packing of streptavidin molecules on a flat surface.
Results and Discussion Scratching Test of APTES Layers on a Diamond Surface. APTES was deposited on the acid-washed O-diamond surface in its 2% solution of the ethanol mixture as described in the Experimental Section, and the surface was examined by the scratching test. This test was composed of a contact by an AFM cantilever at a constant pressing force to scrabble the surface and subsequent imaging of the surface in a noncontact mode to detect a scratch.21 The imaging was carried out with a force small enough to preserve the topography of the surface subjected to the scrabbling, and preservation of the topology was confirmed by triplicate observation. Figure 1A shows the topography of the diamond surface modified with APTES after a scratching test at 100 nN. The observed excavation had the same size as had been scrabbled (squares of 400 nm × 400 nm). The bottom of the excavation showed a similar roughness to the uncoated O-diamond surface (Figure 1B) showing the significant removal of APTES by a cantilever. The thickness of the layer was measured to be 2.5 nm as the depth of the relevant cavity (Figure 1C). Since the thickness of the APTES monolayer is 7 Å,23,24 the layer formed here was multilayer. The index of the hardness of the surface was supposed to be the smallest force under which a scratch was observed.25 This index was used practically in the following analysis, although the true index of the hardness would be the energy cost in dissociating the layer from the substrate. The (23) Flink, S.; Van Veggel, F. C. J. M.; Reinhoudt, D. N. J. Phys. Org. Chem. 2001, 14, 407–415. (24) Kurth, D. G.; Bein, T. Langmuir 1993, 9, 2965–2973. (25) Rezek, B.; Shin, D.; Nakamura, T.; Nebel, C. E. J. Am. Chem. Soc. 2006, 128, 3884–3885.
Langmuir, Vol. 25, No. 1, 2009 205
reproducibility of the index was satisfactory with the same sample and type of the cantilever, although this result did not deny the possibility that the index depends on the structure of the tip of a cantilever. The APTES multilayers were formed by single-step deposition in six different solvents, and the scratching test was carried out with a set of constant pressuring forces from 1 to 500 nN (Figure 2A). Notably, the index of hardness depended on the dielectric constant of the solvent. The smaller the dielectric constant of the solvent, the harder the layer (Figure 2B-G). The layer formed in toluene or heptane could not be scratched at 500 nN (Figure 2F and 2G), but a larger force was not applied because of the irreversible damage of the cantilever. In contrast, the APTES multilayer formed in ethanol mixture was scratched at 1 nN, the weakest force applied. In all cases, the observed index was independent of the lot of the cantilever, the position on the surface to be tested, and the lot of diamond substrate with a resolution of 50 nN. To examine the effect of the difference in the cantilever used in the previous study,16 we tested the same material used in the study, ω-unsaturated 10-trifluoroacetic amide-dec-1-ene (TFAAD). The TFAAD layer photochemically formed on a diamond surface was scratched under a force more than 10 nN, which had been 25 nN in the provider’s AFM system.16 Since the TFAAD layer is bound to a surface by a covalent bond, the agreement of these forces may suggest the existence of covalent bonds between APTES and the O-diamond surface. The relationship of the index with the binding energy of the layer will be analyzed in a subsequent section. The tip structure of the cantilever is shown to have an effect only for a force smaller than 10 nN (see Supporting Information). Aggregates and Defects. On the APTES layers formed in solvents with the two smallest dielectric constants, toluene and heptane, there were many aggregates of APTES and holes which were not observed much on the layer formed in solvents with large dielectric constants (Figure 2). The presence of holes denies the possibility that the absence of scratches on the layer formed in toluene and heptane is not due to the absence of APTES. The existence of APTES on the surfaces was further evidenced by the altered surface topography from the uncoated surface (Figure 1B) as well as by the significant existence of amino residue as mentioned below. We quantified the amino residues accessible to Sulfo-LCSPDP, which has an amino-reactive NHS group, in solution. The observed densities are of similar order of the densities as previously reported for the layer formed on silicon.26 As shown in Table 1, the APTES layers formed in chloroform, toluene, and heptane involved the largest amount of the accessible amino residues, while the layers formed in the ethanol mixture and acetone had much smaller amounts, showing a negative dependence on dielectric constants. The observed large densities for nonpolar solvents are partly, but not completely, explained by the aggregates. The sizes of the holes ranged from 100 nm to 1 µm, and their distribution was inhomogeneous. The holes were not reduced after an extended incubation up to 24 h, indicating establishment of a quasi-equilibrium in their formation. To eliminate the holes we invented a two-step deposition process. After primary deposition of APTES in toluene, secondary deposition was performed in the ethanol mixture or acetone, which greatly removed the defects (Figure 3A and 3B). The layer formed initially in toluene was scratched at 100 nN, if further deposition was (26) Nakagawa, T.; Tanaka, T.; Niwa, D.; Osaka, T.; Takeyama, H.; Matsunaga, T. J. Biotechnol. 2005, 116, 105–111.
206 Langmuir, Vol. 25, No. 1, 2009
Amemiya et al.
Figure 2. Scratching test of APTES layers formed in different solvents. (A) Schematic illustration of scrabbled areas (dashed squares) with the pressing forces (1-500 nN). (B-F) AFM images of APTES layers formed in various solvents with different dielectric constants: (B) ethanol mixture, (C) acetone, (D) octanol, (E) chloroform, (F) toluene, and (G) heptane. Thicknesses of the layers, measured as the largest depth of scratches and holes, in B-F were respectively 3, 2, 5, 2, 30, and 130 nm. Table 1. Density of Amino Residues of APTES Layer Formed in Various Solvents on a Diamond Surface solventa ethanol mixture acetone octanol chloroform toluene heptane a
minimum force dielectric surface amine constant density (×1015/[cm]2) to scratch (nN) >24 20.7 10.3 4.9 2.2 1.9
1.1 ( 0.5 (1 h) 0.9 ( 0.4 0.5 ( 0.2 5.4 ( 1.8 15.0 ( 0.9 11.0 ( 1.2 13.6 ( 1.7
N × (energy of the essential bond) εED N or < (1) Sc (energy of the essential bond) where ε is the strain of the silicon cantilever at the force, E is Young’s modulus of silicon, and D is the vertical displacement of the cantilever. The strain ε is calculated to be 0.003-0.005 at the force of several hundred nN irrespective of the detailed structure of the tip (Supporting Information). Because a C-O bond is weaker than a Si-O one, the essential bond is a C-O bond. Since its free energy of formation is 6 × 10-19 J,31 E is 130-170 GPa, and D is on the order of 1 nm, the density of the essential bond is smaller than 0.6-1.4/(nm)2 according to eq 1. Therefore, the essential covalent bond of the layer that is indented or scratched at several hundred nN distributes on the order of one bond in a square of 1 nm. As evidenced by the roughness on the order of nanometers (Figure 1D), the polished surface used in this study is not composed of a single (100) face but involved the (111) faces as the minor face of the surface. The (100) faces in the O-diamond surface contain oxygen atoms in the form of carbonyl or ether, while the (111) faces can attach them in the form of hydroxyl.14,32 Since APTES is known to react with a hydroxyl residue by releasing ethanol,33 a covalent bond between APTES and the O-diamond surface should be limited to the (111) face. Therefore, (31) Atkins, P.; de Paula, J. Atkin′s physiscal chemistry, 7th ed.; Oxford University Press: New York, 2002. (32) Notsu, H.; Yagi, I.; Tatsuma, T.; Tryk, D. A.; Fujishima, A. J. Electroanal. Chem. 2000, 492, 31–37. (33) Vandenberg, E. T.; Bertilsson, L.; Liedberg, B.; Uvdal, K.; Erlandsson, R.; Elwing, H.; Lundstro¨m, I. J. Colloid Interface Sci. 1991, 147, 103–118.
Amemiya et al.
the estimated low density of the covalent bond is consistent with the results of these previous studies. Speculated Mechanism Yielding the Hardness of APTES Layer on Diamond. The deposition mechanism of silane layers including APTES layers on hydroxylated surfaces has been extensively studied, and the consensus mechanism of the deposition involves polymerization mediated by hydrolysis of silanes. The terminal silanol group of hydrolyzed silanes reacts with hydroxyl groups and thus attaches to hydroxylated surfaces and polymerize with each other to form silane layers on the surface. The amounts of water both in reaction solvents and on substrate surfaces are therefore critical in formation of silane layers. An excessive amount of water in the solvent tends to make aggregates of polymerized silane suspended in the solvent, while a deficiency of water results in poor coverage of the surface.33-44 In this study, we deposited APTES on the hydrophilic surface of O-diamond in a single step or two steps. In a single step the dielectric constant of the solvent used for deposition varied the hardness of the APETS multilayer. In a solvent with a large dielectric constant water is homogeneously dissolved rather than condensed on the diamond surface. In such a condition, the watercatalyzed polymerization of APTES generates aggregates with homogeneous size in solution but not at the hydroxyl residues on the (111) diamond surface (Figure 5A). The low reactivity of the hydroxyl residue on the (111) diamond surface may be due to its lack of the coordinated water and its difficult accessibility to unreacted APTES molecules in aggregates. Therefore, the APTES layer is formed on the O-diamond surface by physical adsorption of the aggregates as well as further polymerization, making a smooth surface but few covalent bonds to the diamond surface (Figure 5B). In contrast, in solvents with small dielectric constants, water exists as clusters both in the solvent and on the hydrophilic surface, allowing formation of aggregates of heterogeneous sizes in solution and on the surface to form a rough APTES layer with holes (Figure 5C and 5D). Some of the hydroxyl residues on the (111) diamond surface will be involved in polymerization of APTES because of the water coordinated to them, and thus, the APTES layer is covalently bound to some (111) surfaces (Figure 5D). The holes will be removed by exposing the surface to the homogeneous smaller aggregates formed in a solvent with a large dielectric constant in a two-step deposition method. (Figure 5E). This model can explain the reason why a rougher and harder surface was observed in a solvent with a small dielectric constant. However, since the surface reactions are not a simple extrapolation of the reactions in a solution, this model cannot be considered more than a speculation.
Conclusion A method has been developed for formation of APTES multilayer on a single-crystalline O-diamond surface with a (34) Angst, D. L.; Simmons, G. W. Langmuir 1991, 7, 2236–2242. (35) Arkles, B.; Steinmetz, J. R.; Zazyczny, J.; Mehta, P. Factors contributing to the stability of alkoxysilanes in aqueous solution; VSP B. V.: Utrecht, 1992; pp 91-104. (36) Fadeev, A. Y.; McCarthy, T. J. Langmuir 2000, 16, 7268–7274. (37) Howarter, J. A.; Youngblood, J. P. Langmuir 2006, 22, 11142–7. (38) McGovern, M. E.; Kallury, K. M. R.; Thompson, M. Langmuir 1994, 10, 3607–3614. (39) Silberzan, P.; Leger, L.; Ausserre, D.; Benattar, J. J. Langmuir 1991, 7, 1647–1651. (40) Simon, A.; Cohen-Bouhacina, T.; Porte, M. C.; Aime, J. P.; Baquey, C. J. Colloid Interface Sci. 2002, 251, 278–83. (41) Tripp, C. P.; Hair, M. L. Langmuir 1992, 8, 1120–1126. (42) Vallant, T.; Brunner, H.; Mayer, U.; Hoffmann, H.; Leitner, T.; Resch, R.; Friedbacher, G. J. Phys. Chem. B 1998, 102, 7190–7197. (43) Wang, M.; Liechti, K. M.; Wang, Q.; White, J. M. Langmuir 2005, 21, 1848–1857. (44) Wang, Y.; Lieberman, M. Langmuir 2003, 19, 1159–1167.
Aminosilane Multilayer
Langmuir, Vol. 25, No. 1, 2009 209
Figure 5. Speculated mechanism for formation of the APTES multilayer on the O-diamond surface. A hydroxyl residue is illustrated as -OH in A-C and its oxygen covalently bound to silicon of APTES is as Si-O in D and E. APTES aggregates are colored in yellow and water clusters in blue. Formation of APTES layer in a solvent with a large dielectric constant is shown in A and B and that in a solvent with a small dielectric constant in C-E. See details in the text.
controlled hardness, roughness, and capacity for immobilizing streptavidin. The hardness and density of holes of the layer depended on the dielectric constant of the solvent used in the deposition process. The solvents with large dielectric constants formed soft layers with high coverage and smoothness, while solvents with small dielectric constants formed hard layers with holes. Formation of a hard APTES layer with few holes was attained by sequential deposition steps first in a solvent with a small dielectric constant and then in a solvent with a large constant. Further treatment of the multilayer with the acidic ethanol solution improved the capacity for immobilizing streptavidin. Formation of a hard and bioactive multilayer was achieved with commercially available reagents and familiar equipment in a wet laboratory. The APTES multilayer with a controlled hardness can be fabricated in nanoscale by AFM. This method for functionalization
of the diamond surface is a useful technique for the research and development of biological applications using diamond. Acknowledgment. We thank Dr. Nebel Christoph and Dr. Hiroshi Uetsuka of the National Institute of Advanced Industrial Science and Technology for a TFAAD layer formed on diamond, experimental advice, discussions, and critical reading of the manuscript. We also thank Dr. Takatoshi Yamada, Dr. Nianjun Yang, Dr. Shinichi Shikata, and Dr. Naoji Fujimori of the same institute for useful advice and help. This research was supported by the New Energy and Industrial Technology Development Organization (NEDO). Supporting Information Available: Analysis of the distortion of a cantilever under a given force. This material is available free of charge via the Internet at http://pubs.acs.org. LA801556X
