Langmuir 2004, 20, 7499-7506
7499
Preparation and Characterization of n-Alkanoic Acid Self-Assembled Monolayers Adsorbed on 316L Stainless Steel Galit Shustak,†,‡ Abraham J. Domb,‡ and Daniel Mandler*,† Department of Inorganic and Analytical Chemistry, The Hebrew University of Jerusalem, Jerusalem 91904, Israel, and Department of Medicinal Chemistry and Natural Products, School of PharmacysFaculty of Medicine, The Hebrew University of Jerusalem, Jerusalem 91120, Israel Received December 29, 2003. In Final Form: June 15, 2004
The electrochemical formation and characterization of decanoic, myristic, palmitic, and stearic acid self-assembled monolayers on a native oxide surface of 316L stainless steel have been studied. This work describes a new approach to surface modification of stainless steel in which the self-assembly of n-alkanoic acids is facilitated by applying a potential to the stainless steel in an organic electrolyte solution. While decanoic acid forms a disorganized monolayer as a result of sweeping the potential in an acetonitrile solution containing 0.1 mM of the respective acid, longer acids, that is, myristic and palmitic acids, form highly ordered closed-packed monolayers. This electrochemical approach results in highly reproducible monolayers that are deposited within a shorter time than the traditional assembly process. The monolayers were characterized by cyclic voltammetry, double-layer capacity (ac voltammetry), contact angle measurements, X-ray photoelectron spectroscopy, and external reflection-absorption Fourier transform infrared spectroscopy. The utilization and implications of this modification technique are discussed.
Introduction The surface properties of metals, such as adhesion, wetting, and lubrication, can be tailored by attaching various organic modifiers that can be as thin as a monolayer. Among metals, stainless steel has been extensively used for manufacturing orthopedic implants and other implantable medical devices, owing to its corrosion resistance and superior mechanical properties.1 The biocompatibility of stainless steel implants can be significantly improved by modifying their surfaces with organic molecules, polymers, or inorganic coatings.2-7 This clearly requires good adhesion, that is, strong chemical interactions, between the monolayer and the metal substrate. The monolayer can then be the substrate for additional organic specific coatings, which can increase the implant biocompatibility, prevent its corrosion, or function as controlled-release drug carriers. Deposition of self-assembled monolayers (SAMs) offers one of the highest quality routes for preparing chemically and structurally well-defined surfaces on solid supports and particularly on metal surfaces.8-11 Despite the * Corresponding author. Tel: +972-2-658-5831. Fax: +972-2658-5319. E-mail:
[email protected]. † Department of Inorganic and Analytical Chemistry. ‡ Department of Medicinal Chemistry and Natural Products. (1) Metals as Biomaterials; Helsen, J. A., Breme, H. J., Eds.; Wiley: New York, 1998. (2) Harm, U.; Burgler, R.; Furbeth, W.; Mangold, K.-M.; Juttner, K.; Macromol. Symp. 2002, 187, 65. (3) Reinartz, C.; Furbeth, W.; Stratmann, M. Fresenius’ J. Anal. Chem. 1995, 353, 657. (4) Stratmann, M.; Volmeruebing, M. Appl. Surf. Sci. 1992, 55, 19. (5) Ruan, C. M.; Bayer, T.; Meth, S.; Sukenik, C. N. Thin Solid Films 2002, 419, 95. (6) Marsh, J.; Scantlebury, J. D.; Lyon, S. B. J. Appl. Polym. Sci. 1996, 59, 897. (7) Shih, C.-C.; Shih, C.-M.; Su, Y.-Y.; Su, L. H. J.; Chang, M.-S.; Lin, S.-J. Corros. Sci. 2004, 46, 427. (8) Aramaki, K.; Shimura, T. Corros. Sci. 2003, 45, 2639. (9) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir-Blodgett to Self-Assembly; Academic Press: New York, 1991.
increasing interest and efforts devoted to self-assembly technology in recent years, only a few reports dealing with SAMs on stainless steel have been published.2-5,8,12-15 Stratmann4 and Sukenik5 reported on the formation and characterization of n-alkanethiol SAMs on electrochemically reduced stainless steel surfaces. They applied a negative potential in order to remove the native oxide layer of stainless steel and then added the thiols to the electrolyte solution to form a monolayer. Carboxylic acid monolayers have been assembled on iron16 and steel17 using the Langmuir-Blodgett (LB) technique as a means of corrosion inhibition and increasing lubrication of the surface. The purpose of this study is to describe the preparation and characterization of carboxylic acid SAMs deposited on 316L stainless steel via an electrochemical method. There are two principal possible interactions between an n-alkanoic acid monolayer and a metal surface. Acidbase interaction will result in the chemisorption of the acid to form a metal carboxylate salt.18-19 Samant and co-workers19 reported that docosanoic acid (C22) spontaneously adsorbs on Ag(111). They suggested that the nature of the carboxylic acid and silver interaction is predominantly ionic, as opposed to the covalent interactions in the previously characterized thiol films.20 Tao and co(10) Schriber, F. Prog. Surf. Sci. 2000, 65, 151. (11) Netzer, L.; Sagiv, J. J. Am. Chem. Soc. 1983, 105, 674. (12) Tsuji, N.; Nozawa, K.; Aramaki, K. Corros. Sci. 2000, 42, 1523. (13) Aramaki, K. Corros. Sci. 1999, 41, 1715. (14) Nozawa, K.; Aramaki, K. Corros. Sci. 1999, 41, 57. (15) Nozawa, K.; Nishihara, H.; Aramaki, K. Corros. Sci. 1997, 39, 1625. (16) Xing, W.; Shan, Y.; Guo, D.; Lu, T.; Xi, S. Corrosion 1995, 51, 45. (17) Domiguez, D. D.; Mowery, R. L.; Turner, N. H. Tribol. Trans. 1994, 37, 59. (18) Lin, S-.Y.; Tsai, T. K.; Lin, C. M.; Chen, C. H. Langmuir 2002, 18, 5473. (19) Samant, M. G.; Brown, C. A.; Gordon, J. G. Langmuir 1993, 9, 1082.
10.1021/la036470z CCC: $27.50 © 2004 American Chemical Society Published on Web 07/29/2004
7500
Langmuir, Vol. 20, No. 18, 2004
workers21 studied the stability of carboxylic acid SAMs under conditions of chemical stress. They showed that a sudden exposure of n-alkanoic acid SAMs on silver to H2S resulted in sulfidation of the silver surface, destabilization of the strong -CO2-|Ag+ interaction, protonation of the carboxylate headgroups, and rapid reorganization of the film. On the other hand, in most studies involving longchain n-alkanoic acids on reactive metals having a native oxide overlayer, such as aluminum,22-24 silver,24-27 and copper,24-25,27 the acids are probably chemisorbed via proton transfer to a surface oxygen atom to form ionic bonds. The classic study of Allara and Nuzzo22,23 concluded that carboxylic acids with chain lengths of C11 and longer form close-packed monolayers with structural features akin to those of Langmuir-Blodgett type films on aluminum covered with an oxide layer. Tao27 compared n-alkanoic acid SAMs on copper, aluminum, and silver oxide surfaces and showed that their structures are dictated by the substrate as well as their chain length. The carboxylic headgroups were more highly ordered on silver than on copper and aluminum, leading to less densely packed films on the latter. Our results described below support the general concept of forming a stable and well-ordered carboxylic acid monolayer via mainly ionic interaction between the carboxylic headgroup and the 316L stainless steel oxide surface. The self-assembly technique of amphiphilic molecules that was utilized in all the studies described above involved modification under open-circuit potential for typically 3 h to several days. This study describes a different approach to surface modification of stainless steel, which clearly shows that the self-assembly of n-alkanoic acids can be facilitated and better controlled by applying a potential in the course of the modification process. The approach is based on the electrochemical activation of 316L stainless steel in an organic aprotic solvent containing low amounts of water in the presence of the carboxylic acids. The formation of highly reproducible, stable, well-ordered and closely packed monolayers has been studied and characterized by electrochemistry, external reflection-absorption Fourier transform infrared (RA-FTIR) spectroscopy, X-ray photoelectron spectroscopy, and water contact angle measurements. Experimental Section Materials. Decanoic acid (DA, 99+%), myristic acid (MA, 99.5%), stearic acid (SA, 99+%), tetrabutylammonium tetrafluoroborate (TBATFB, 99%), hexaamineruthenium(III) trichloride (Ru(NH3)6Cl3, 98%), NaNO3 (99.9%), and NaCl (99.5%) were purchased from Aldrich. Acetonitrile (ACN, >99.8%) was from J.T. Baker, while palmitic acid (PA, 99%) was obtained from BDH Technologies (U.K.). 316L Stainless steel plates and rods were purchased from Mashaf Co. (Jerusalem, Israel). The stainless steel plates (9 × 40 mm) were used for infrared spectroscopy and contact angle measurements, while the rods were applied for electrochemical measurements. The stainless steel rod (3 mm diameter) was embedded in a Teflon sheath exposing only a disk, which served as the electrode surface. (20) Samant, M. G.; Brown, C. A.; Gordon, J. G. Langmuir 1991, 7, 437. (21) Tao, Y. T.; Lin, W. J. Phys. Chem. B 1997, 101, 9732. (22) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 45. (23) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 52. (24) Tao, Y. T.; Hietpas, G. D.; Allara, D. L. J. Am. Chem. Soc. 1996, 118, 6724. (25) Tao, Y. T.; Lee, M. T.; Chang, S. C. J. Am. Chem. Soc. 1993, 115, 9547. (26) Allara, D. L.; Bright, B. T.; Porter, M. D.; Schlotter, N. E. Chem. Phys. Lett. 1986, 132, 93. (27) Tao, Y. T. J. Am. Chem. Soc. 1993, 115, 4350.
Shustak et al. Instrumentation. Electrochemical measurements were conducted with an AUTOLAB PGSTAT10 potentiostat (EcoChemie, Utrecht, The Netherlands) and a BAS-100B/W electrochemical analyzer (Bioanalytical Systems, Lafayette, IN), using a singlecompartment three-electrode glass cell. The reference electrode was either a saturated Hg|Hg2SO4|K2SO4(sat) electrode when aqueous solutions were employed or a Ag|AgBr wire that was used in organic media (see below for a preparation procedure). The latter, which has a potential of 0.448 V versus ferroceneferrocenium (Fc/Fc+),28 was found to have a much more stable potential in the organic media than the commonly used Ag|AgCl wire. A 6 mm diameter graphite rod was used as an auxiliary electrode. External Fourier transform infrared (FTIR) spectra were recorded using an Equinox 55 (Bruker) spectrometer, at a resolution of 2 cm-1, equipped with a nitrogen-cooled MCT detector. The spectra were acquired with a grazing angle accessory having an incident angle of 80° to the normal and a p-polarized beam. Normally, 1500 scans of the sample were collected versus a reference which was a bare and freshly polished stainless steel surface. X-ray photoelectron spectra were recorded using an Axis Ultra spectrometer (Kratos), and Mg KR radiation of 1486.71 eV. Data were collected and analyzed by a vision processing program. Contact angles were measured with a Rame´-Hart model 100 contact angle goniometer. Advancing and receding contact angles were determined by adding and withdrawing a fixed amount of deionized water to and from the drop, respectively. This measurement was repeated three times for each sample, and the average values are reported. All aqueous solutions were prepared from deionized water (Barnstead Easypure UV system). Procedures. An Ag|AgBr reference electrode (which was used for the electrochemical modification in ACN) was prepared by sweeping the potential of a polished silver wire electrode in 50% HBr solution from -0.3 to 1 V versus Ag|AgCl at a scan rate of 10 mV s-1. The potential was held at 1 V for 2 min before the electrode was pulled out from the solution under potential (1 V), washed with water, and dried in air. The potential of the resulting Ag|AgBr wire was frequently checked versus ferrocene/ferrocenium.28 The stainless steel disk electrodes were polished first with 240, 600, and 2000 grit emery paper (Buehler), followed by fine polishing by alumina paste (1 and 0.05 µm) on a microcloth polishing pad. The plates were received polished and treated only with a 2000 grit emery paper. Then the electrodes were washed and sonicated (∼15 min) in ACN and dried with a stream of nitrogen at room temperature prior to the modification. The clean electrodes were immersed into a deaerated modification solution containing 0.1 mM carboxylic acid and 0.1 M TBATFB in ACN at room temperature. A potential sweep between -0.8 and 1.2 V versus Ag|AgBr was typically applied (10 cycles unless otherwise mentioned). The modified surfaces were rinsed with pure ACN and dried with a gentle stream of nitrogen. Double-layer capacity, Cdl, was measured in a 10 mL deaerated aqueous solution of 0.1 M NaNO3. After the electrode was allowed to equilibrate for 10 min, an ac voltage of 7 mV peak-to-peak and 320 Hz was superimposed on the dc potential (0-0.3 V versus Hg|Hg2SO4|K2SO4(sat)) and the real and imaginary parts of the ac current were detected with an EcoChemie potentiostat equipped with a frequency analyzer (FRA). All measurements were performed at room temperature (21 °C ( 2 °C).
Results and Discussion The formation of SAMs on reactive metals, such as stainless steel, is not trivial and needs to take into account the interactions between the amphiphiles and the native oxide layer. The latter affects significantly the adhesive properties of the surface. We have been motivated to form SAMs on stainless steel as a means of increasing the surface adhesion for the further attachment of organic biocompatible substances. An appealing approach for attaching these substances is through electropolymer(28) Gritzner, G.; Kuta, G. J. Pure Appl. Chem. 1984, 56, 461.
n-Alkanoic Acid SAMs Adsorbed on Stainless Steel
Langmuir, Vol. 20, No. 18, 2004 7501 Table 1. Advancing and Receding Water Contact Angles Measured on 316L Stainless Steel Surface Modified with n-Alkanoic Acids
monolayer composition bare 316L SS bare 316L SS after 10 cycles in 0.1 M TBATFB/ACN decanoic acid (10 cycles) myristic acid (10 cycles) palmitic acid (1 cycle) palmitic acid (5 cycles) palmitic acid (10 cycles)
Figure 1. First and second cyclic voltammograms of 316L stainless steel in a solution of 0.1 mM decanoic acid and 0.1 M TBATFB in acetonitrile; scan rate, 100 mV s-1.
ization. Therefore, a critical prerequisite of the oxide layer, in our case, has been to control its thickness in order to maintain electron transfer across it. Growing a controlled oxide layer can be accomplished by different methods, such as chemical and thermal treatment or electrochemical oxidation.29-33 The formation of SAMs under potential control, where a gold surface was oxidized prior to the self-assembly process, has been described previously.34-36 Nevertheless, to the best of our knowledge the potentialassisted SAM formation has never been described in conjunction with assembling monolayers on stainless steel. In the next section, we describe the formation of a wellordered SAM of long-chain carboxylic acids on 316L stainless steel by sweeping the potential in the presence of the acids. Figure 1 shows the cyclic voltammetry (CV) of a stainless steel electrode in ACN in the presence of 0.1 mM decanoic acid. The first two cycles are shown, where an irreversible anodic wave is clearly observed only in the course of the first scan. A similar electrochemical behavior is seen in the absence of the acid, suggesting that the anodic wave is due to the oxidation of the surface. Subsequent cycles overlapped with the second scan. Furthermore, the oxidation wave was not detected when dry ACN was used, implying that water participates in the electrochemical process. Indeed, the magnitude of the oxidation wave increased upon adding water (up to 20% v/v) to the electrochemical cell. The effect of water content in the modification solution on the deposition nature of alkanoic acid on the 316L stainless steel surface will be further discussed; vide infra. Interestingly, we found that the water contact angle of an electrode electrochemically treated in the presence of decanoic acid was significantly higher than in the case of cycling the stainless steel surface in the absence of an acid (Table 1). At the same time, the contact angle of an electrode, which was not cycled, however, immersed in the modification solution did not change noticeably. The contact angle of a decanoic film was 87° as a result of cycling the electrode 10 scans between -0.8 to 1.2 V versus (29) Jervis, T. R.; Hirvonen, J.-P. Wear 1991, 150, 259. (30) Ohmi, T.; Morita, M.; Teramoto, A.; Makihara, K.; Tseng, K. S. Appl. Phys. Lett. 1992, 60, 2126. (31) Nieminen, M.; Niinisto, L.; Rauhala, E. J. Mater. Chem. 1996, 6, 27. (32) Niinisto, L.; Ritala, M.; Leskela, M. Mater. Sci. Eng. 1996, 41, 23. (33) Brunner, H.; Vallant, T.; Mayer, U.; Hoffmann, H. Langmuir 1996, 12, 4614. (34) Ron, H.; Rubinstein, I. J. Am. Chem. Soc. 1998, 120, 13444. (35) Ron, H.; Rubinstein, I. Langmuir 1994, 10, 4566. (36) Turyan, I.; Mandler, D. J. Am. Chem. Soc. 1998, 120, 10733.
advancing contact angle/ deg ( 1°
receding contact angle/ deg ( 1°
hysteresis
65 59
47 45
18 19
87 97 98 102 109
70 82 80 90 100
17 15 18 12 9
Ag|AgBr. Increasing the chain length is known to yield more highly ordered SAMs22-27 almost regardless of the substrate. Table 1 summarizes the advancing and receding water contact angles of three films made of decanoic (C10), myristic (C14), and palmitic (C16) acids formed under the same conditions. A clear trend of increase in the advancing contact angles can be seen, indicating the enhancement of film packing. The advancing contact angle values for myristic and palmitic acid films are typical for a densely packed array of n-alkyl chains and agree with those previously reported for comparable structures.18,22,25,27 The contact angle was also affected by the number of potential scans. The data in Table 1 show that the contact angle increases up to an upper limit of 109° upon cycling the stainless steel surface in the presence of palmitic acid. We did not observe a further increase of the contact angle when scanning the potential beyond 10 cycles, suggesting that the layer reached its final organization. A control experiment in which a stainless steel electrode was cycled in the absence of an acid gave (Table 1) a contact angle of 59°, which is relatively high and is presumably attributed to some adsorption of the electrolyte (as detected by X-ray photoelectron spectroscopy (XPS)) on the oxide layer. Evidently, the oxide layer that is formed in the absence of an acid is likely to be thicker than that formed in the course of assembling the organic layer. This issue will be addressed further by XPS and when studying the double-layer capacity and the rate of electron transfer in these systems. The only contact angle data of SAMs on stainless steel of which we are aware are those by Sukenik and co-workers,5 however, using alkanethiols and alkylamines. They obtained contact angles of 104° and 105° for hexadecaneamine and hexadecanethiol, respectively. The hysteresis of the films is also an indicator for their packing; that is, as the ordering of the film increases the hysteresis drops. This trend is also seen in our results (Table 1). The reproducibility of the electrochemical modification procedure, as evidenced by the precision of the contact angle, is very high. The contact angles reported here are the average of numerous measurements where the standard deviation of the measurements was very small, alluding to the high reproducibility of the film assembly. The formation of a thin film on the stainless steel surface is also expressed in the cyclic voltammetry of the modified electrodes in the presence of a redox couple. Figure 2 shows the effect of the number of potential cycles on the CV of Ru(NH3)63+. The modification was carried out in ACN as previously described using 0.1 mM palmitic acid. Then the electrode was transferred to an aqueous solution containing the redox probe. It is evident that the magnitude of the reduction and oxidation waves of Ru(NH3)63+ is reduced as the number of cycles increases. After 10 cycles, a complete blocking of the Ru(NH3)63+ is attained. A control experiment (Figure 2, bare) in which a stainless
7502
Langmuir, Vol. 20, No. 18, 2004
Shustak et al.
Figure 2. Cyclic voltammetry of 1 mM Ru(NH3)63+ in 0.1 M NaCl recorded with (a) a bare 316L stainless steel electrode and after electrochemical modification in 0.1 mM palmitic acid and 0.1 M TBATFB/ACN solution with (b) 1 cycle, (c) 5 cycles, and (d) 10 cycles, applying a scan rate of 100 mV s-1.
Figure 4. Cyclic voltammetry of 1 mM Ru(NH3)63+ in 0.1 M NaCl (scan rate, 100 mV s-1) recorded with a 316L stainless steel electrode: (a) freshly polished, (b) electrochemically cycled for 10 scans in 0.1 M TBATFB/ACN solution, and (c) polished and left under ambient conditions for 1 day.
Figure 3. Cyclic voltammetry of 1 mM Ru(NH3)63+ in 0.1 M NaCl (scan rate, 100 mV s-1) recorded with (a) a bare 316L stainless steel electrode and after electrochemical modification in 0.1 M TBATFB/ACN and 0.1 mM (b) decanoic acid, (c) myristic acid, and (d) palmitic acid. Table 2. Peak Currents and Potentials of the Cathodic Wave of Ru(NH3)63+ Shown in Figure 3 stainless steel electrode
peak current/µA
peak potential/V vs Hg/Hg2SO4
bare electrode decanoic acid myristic acid palmitic acid
140.0 93.9 47.0 no peak
0.670 0.733 0.845 no peak
steel electrode was swept for 10 cycles in an acid-free ACN solution affected only the oxidation wave, yet, not affecting the electrochemical reversibility, as will be discussed below. This clearly indicates that blocking is due to the deposition of an acid film, which is induced by potential and not by oxide growth. Further experiments in which the potential window, the scan rate, and the nature of the electrolyte were examined led us to determine the optimum conditions for assembling the films of alkanoic acids. A potential sweep (100 mV s-1) of 10 cycles between -0.8 to 1.2 V versus Ag/AgBr rather than a potential step gave the best and highly reproducible results. Figure 3 illustrates the blocking properties of electrodes modified with decanoic, myristic, and palmitic acids using Ru(NH3)63+ in an aqueous solution. The CV of a bare electrode, which was cycled in the absence of an acid, is also shown. For a decanoic acid film, the reduction peak potential was only 63 mV more negative and its current was reduced by only 33%, as compared with a bare stainless steel electrode (Table 2). The small influence of
Figure 5. (A) Cyclic voltammetry of 1 mM Ru(NH3)63+ in 0.1 M NaCl recorded with a decanoic acid modified 316L stainless steel electrode (after cycling for 10 cycles in 0.1 mM decanoic acid in 0.1 M TBATFB/ACN). Scan rate: (a) 10, (b) 100, (c) 200, (d) 500, (e) 1000, and (f) 2000 mV s-1. Inset: dependence of the peak current as a function of square root of the scan rate. (B) Cathodic peak potential as a function of the logarithm of the scan rate.
the decanoic acid film on the redox is not surprising, as a closed-packed self-assembled monolayer is expected for chain lengths of C11 and longer.22-23 For a myristic acid modified electrode, the peak potential was shifted by 175 mV to more negative potentials and the current was reduced by 66%, as compared with an unmodified stainless steel electrode, whereas the electrode modified with palmitic acid showed almost complete blocking. The kinetics of electron transfer using Ru(NH3)63+ was further studied by a bare stainless steel electrode and an electrode modified by decanoic acid (Figures 4 and 5).
n-Alkanoic Acid SAMs Adsorbed on Stainless Steel
Langmuir, Vol. 20, No. 18, 2004 7503
Figure 4 shows the effect of polishing and electrochemical cycling on the voltammetric behavior of Ru(NH3)63+ using a stainless steel electrode. It can be seen that the freshly polished electrode exhibits a quasireversible behavior with a potential peak difference of 160 mV and a reductionoxidation current ratio close to unity. On the other hand, the CV of a freshly polished and electrochemically cycled electrode shows a chemically irreversible behavior where the oxidation wave is almost absent. A similar CV was obtained with a polished electrode, which was left under ambient conditions for 1 day (Figure 4), suggesting that the formation of an oxide layer has a pronounced effect on the oxidation of Ru(NH3)62+. Figure 5 shows the effect of the scan rate on the CV of Ru(NH3)63+ using a decanoic acid modified electrode. The measurements were carried out using IR compensation. A linear dependence between the cathodic peak current and the square root of the scan rate is obtained (Figure 5A, inset) indicating that the reduction of the redox species is diffusion controlled. Furthermore, the kinetic parameters, that is, transfer coefficient (R) and standard heterogeneous rate constant (k0), were determined by plotting the peak potential as a function of the logarithm of the scan rate, v (eq 1).37
[
( )
D01/2 RFv RT Ec,p ) E ′ 0.780 + ln + ln 0 RF RT k 0
]
1/2
( )
(1)
Ec,p is the cathodic peak potential, and E0′, R, T, and F have their usual meanings. Figure 5B shows that indeed a linear dependence is obtained, allowing one to extract R (0.3) and kmonolayer0 (1.6 × 10-3 cm s-1) from the slope and intercept, respectively. The same experiment and data treatment were performed with an electrode that was subjected to electrochemical cycling (2 and 10 cycles) in the absence of an acid. The transfer coefficient was ca. 0.5 for both cases, and kbare0 was ca. 4.0 × 10-3 and 2.3 × 10-3 cm s-1 for 2 and 10 cycles, respectively. The decrease of the rate of electron transfer upon increasing the number of potential scans is obviously due to the thickening of the oxide layer. Following Amatore’s approach,40 θ, which is the fractional coverage (the fraction of the surface that is covered with the monolayer and through which the rate of electron transfer is negligible), can be derived by comparing the heterogeneous rate constant of bare and modified electrodes (eq 2).
Figure 6. Double-layer capacity as a function of potential measured in 0.1 M NaNO3 solution for a bare 316L stainless steel electrode that was freshly polished (left-pointing triangle) and after cycling (10 cycles) in 0.1 M TBATFB/ACN (9) and 0.1 mM decanoic acid (b), 0.1 mM myristic acid (2), 0.1 mM palmitic acid (1), and 0.1 mM stearic acid ([).
on the electrode. This treatment could not be applied to longer acids due to complete blocking of electron transfer. Verifying that a monolayer rather than a multilayer was formed was accomplished by studying the doublelayer capacity of the different films formed by four acids. Figure 6 shows the dependence of the capacity on the potential of the modified and bare electrodes in 0.1 M NaNO3. The capacity was measured by alternating current voltammetry in an aqueous solution between 0 and 0.3 V versus Hg|Hg2SO4|K2SO4(sat) and was found to be potential independent. The capacity of decanoic, myristic, palmitic, and stearic acid films was 7.3, 3.57, 1.78, and 1.36 µF cm-2, respectively. These values are significantly lower than the capacity measured for a stainless steel electrode before (17.5 µF cm-2) and after (13.0 µF cm-2) it was cycled in the absence of an acid. The fact that the capacity is not dependent on potential indicates that it can be described by the Helmholtz model37,38 which assumes that the double layer behaves as a capacitor plate. The capacity is then given by eq 3:
Cdl )
0A d
(3)
Apparently, the fractional coverage depends to a large extent on the heterogeneous rate constant of a bare electrode. Introducing the values for the heterogeneous rate constants into eq 2 yields a fractional coverage of 0.6 and 0.3 for a bare electrode cycled for 2 and 10 scans, respectively. It is conceivable that the oxide layer that is formed in the presence of an acid, as a result of cycling the stainless steel surface 10 cycles, is significantly thinner than that formed in the absence of an acid. Nevertheless, θ is still lower than expected even when the rate constant of a bare electrode that was cycled only 2 scans was used. This can probably be attributed to the fact that some tunneling occurs across the decanoic acid film, an effect that is not included in Amatore’s approach. In other words, electron transfer takes place not only in uncoated areas
where is the dielectric constant, 0 is the permittivity of free space, A is the area of the working electrode, and d is the film thickness. This means that plotting the reciprocal capacity versus the film thickness should give a straight line. Figure 7 shows that a linear relationship is obtained between the reciprocal capacity and the length of the acid chains (assuming 1.3 Å per methylene, see below). Linearity is increased when the decanoic acid is omitted, which is understandable based on our electrochemical observations. Assuming a length of either 1.1 or 1.3 Å per methylene in an all-trans chain configuration, which corresponds to a tilt angle of 30° and 0° from the normal, respectively, gives a slope that varies between 1.0 × 107 and 8.7 × 106 cm µF-1. Since the slope equals 1/0, the dielectric constant of the layer can be estimated. A value of 1.13 and 1.30 is obtained for the dielectric constant of the layer for 30° and 0° tilt angles, respectively. These values are somewhat lower than the typical dielectric constants for pure aliphatic hydrocarbons and polyethylene, which are 2.0 and 2.3, respectively.38 This
(37) Bard, A. J.; Faulkner, L. R.Electrochemical Methods: Fundamentals and Applications, 2nd ed.; Wiley: New York, 2001; pp 230250.
(38) Finklea, H. O. In Electroanalytical Chemistry: A Series of Advances; Bard, A. J., Ed.; Marcel Dekker: New York, 1996; Vol 19, pp 110-318.
kmonolayer0 ) kbare0(1 - θ)
(2)
7504
Langmuir, Vol. 20, No. 18, 2004
Shustak et al.
Figure 7. Dependence of the reciprocal double-layer capacity (from Figure 6) as a function of the length of the carboxylic acid used for modifying a 316L stainless steel surface.
can be explained by recalling that our interface is a series of two capacitors representing the oxide layer and the organic film. Nevertheless, the fact that a linear dependence is obtained (Figure 7) suggests that the total capacity of the interface is governed primarily by the organic film. Investigating the thickness of the monolayers by ellipsometry was very intricate due to a continuous oxidation process of the stainless steel surface. The bare stainless steel surface continuously oxidized while exposed to air, preventing the establishment of an adequate model of the oxide film that could be used for the organically modified stainless steel. This behavior was reported by Tao25,27 and Nuzzo39 when measuring the thickness of carboxylic acid monolayers on copper and aluminum oxides. The solid-film interface was further examined using XPS. Figure 8 shows the high-resolution X-ray photoelectron spectra of a freshly polished and electrochemically cycled stainless steel surface in the absence and presence of palmitic acid. Care must be taken in drawing quantitative conclusions from these spectra due to the relatively high roughness of the surface. Nevertheless, it is possible to qualitatively compare the XPS measurements. Only the peaks of the most relevant elements, namely, Fe2p3/2, Fe2p1/2, and O1s, are shown. Iron oxide surfaces are characterized by two Fe2p3/2 peaks at 709.7 and 711.2 eV assigned to Fe2+ and Fe3+, respectively. The Fe2p1/2 peaks are separated by 14.3 eV and appear at a higher binding energy. The small peak at 707 eV is assigned to Fe0. Therefore, it is evident that the iron surface is composed of mostly Fe2O3. Comparing the Fe2p3/2 spectra of the freshly polished surface and that of an electrode, which was cycled in ACN without palmitic acid, shows only negligible difference (Figure 8A, curves 2 and 3). On the other hand, the intensity of the iron peaks of a surface modified with palmitic acid decreased because of increasing the mean path of the generated photoelectrons due to the organic layer. Two peaks are evident at 530.1 eV (ν1) and 531.8 eV (ν2) that are associated with the O1s electron binding energy (Figure 8B). However, while the peak at ca. 530.1 eV is attributed to lattice oxygen, O2-, in the metal oxide, that at 531.8 eV is assigned to the hydroxyl anion, OH-. The latter can originate from two sources, the carboxylic acid and surface Fe-OH groups. We find that the ratio ν2/ν1 increased as the electrodes were cycled in the absence (39) Laibinis, E.; Whitesides, G. M.; Allara, D. A.; Tao, Y. T.; Parikh, A. N.; Nuzzo, R. G. J. Am. Chem. Soc. 1991, 113, 7152. (40) Amatore, C.; Saveant, J.-M.; Tessier, D. J. Electroanal. Chem. 1983, 147, 39.
Figure 8. High-resolution Fe2p3/2 and Fe2p1/2 (A) and O1s (B) X-ray photoelectron spectra of a freshly polished (3) and electrochemically cycled stainless steel surface in the absence (2) and presence (3) of palmitic acid.
(ν2/ν1 ) 1.3) or presence (ν2/ν1 ) 1.4) of palmitic acid (ν2/ν1 ) 1.1 for a freshly polished surface). This is understandable since cycling the electrode should increase the density of OH surface groups because of either oxidation of the surface or attachment of the carboxylic acid. The carboxylic acid monolayer attenuates both oxygen peaks. The fact that the ratio ν2/ν1 does not increase significantly as a result of cycling the stainless steel surface suggests that the electrochemical treatment does not thicken the oxide layer. To verify that the n-alkanoic acid films formed on 316L stainless steel are stable and structurally comparable to n-alkanoic acid monolayers previously formed at other metal oxide surfaces,22-25 the films were further characterized by external RA-FTIR. The C-H stretching region of the infrared spectrum of decanoic, myristic, and palmitic acid monolayers at 316L stainless steel substrates is shown in Figure 9. The labeled peaks22 represent the symmetric and asymmetric stretching modes of the methylene [(νs, CH2) and (νa, CH2)] and methyl [(νs, CH3) and (νa, CH3)] groups. Both the absolute intensities and peak locations indicate that the surface coverage and structure within the hydrocarbon chains of the myristic and palmitic acid films are comparable to those previously reported22-27 for carboxylic acid monolayers on metal oxide surfaces. More specifically, the data presented for myristic (C14) and palmitic (C16) acid layers fit reasonably well with the monolayer assembly model in which the CH3 group has its C-CH3 rotation axis tipped closer to parallel to the stainless steel surface, where the alkyl chain axes are tilted off the surface normal. The orientation of the alkyl chain can be estimated from the ratio between the intensities of the methylene and methyl stretching vibrations (Aνs,CH2/Aνs,CH3 and Aνa,CH2/Aνa,CH3) as has been shown
n-Alkanoic Acid SAMs Adsorbed on Stainless Steel
Langmuir, Vol. 20, No. 18, 2004 7505
Figure 9. Reflection-absorption FTIR spectra of the C-H stretching region of decanoic (solid line), myristic (dashed line), and palmitic (dashed-dotted line) acid modified 316L stainless steel surfaces.
Figure 10. Reflection-absorption FTIR spectra of the carbonyl stretching region of palmitic (dashed-dotted line), myristic (dashed line), and decanoic (solid line) acid modified 316L stainless steel surfaces.
by Ulman et al.41 In essence, as the chain becomes close to the normal, the intensity of the methylene group decreases since the dipole moment component that is parallel to the surface increases. The intensity ratios (Aνs,CH2/Aνs,CH3 and Aνa,CH2/Aνa,CH3) found for palmitic and myristic acids were 1.3 and 2.8, and 1.7 and 3.0, respectively. Such ratios correspond to a tilt angle that is slightly larger than that reported for alkanoic acid SAMs on Cu and Al and fairly close to that on Ag.27 Yet, the fact that the asymmetric and symmetric methylene vibration modes appear at 2917 and 2849 cm-1, respectively, is evident of the highly ordered SAMs, which is similar to data found for carboxylic acid monolayers on metal oxides.27 Comparing the spectra of myristic and palmitic acid films reveals that their alignment is identical but the absorbance intensity of myristic acid is half that of palmitic acid. The relative absorbance intensity can be attributed to three distinct contributions. The dominant contribution is due to the difference in the tilt angle of the myristic and palmitic monolayers. As the tilt angle increases, the component of the dipole moment perpendicular to the surface decreases, causing a decrease of the vibration intensity. Obviously, as the number of carbons in the hydrocarbon chain increases, the intensity will increase accordingly. This by itself cannot account for the difference between myristic (C14) and palmitic (C16) acid (Figure 9). Finally, the difference in surface coverage will also affect the relative absorbance intensity. Since we were unable to determine the surface coverage of myristic and palmitic acids (by electrochemical means), we can only speculate that the difference observed in the absorbance intensities between the two acids is a result of all these three contributions. On the other hand, the spectrum of the shorter chain, decanoic acid (C10), on oxidized stainless steel shows conformational disordering. This result is in accordance with the above-described interfacial properties of this layer and can be related to earlier studies on chemisorption of carboxylic acid SAMs on other metal oxide surfaces. Highly oriented and closely packed monolayers were formed by amphiphiles bearing hydrocarbon chains longer than C11.22-25 To test the stability of the films, samples of decanoic, myristic, and palmitic acid SAMs modified on 316L stainless steel were measured immediately after modi-
fication and after 1, 3, 5, and 7 days of exposure to ambient (not shown). There was not a noticeable change in the spectrum, indicating that the films are stable under these conditions. Figure 10 shows the low-frequency region IR spectra of palmitic, myristic, and decanoic acid monolayers. Three vibration modes associated with the carboxylic group are noticeable: a symmetric and asymmetric -CO2- stretching at 1452 and 1595 cm-1, respectively, and a CdO stretching in a -CO2H moiety at 1728 cm-1. The presence of these peaks in conjunction with the lack of a CdO stretching mode at 1703 cm-1, which is characteristic of a hydrogenbonded carboxylic moiety, indicates that the carboxylic headgroup undergoes partial dissociation to form surface carboxylate species. The asymmetric -CO2- stretching mode is known to be dependent on the local environment and the nature of the ionic interaction with the substrate.22,24 Indeed, this signal is broadened and contains several overlapping peaks, suggesting that the headgroup interacts via a number of different modes with the surface. The presence of a symmetric -CO2- mode suggests that the carboxylate headgroup is attached to the surface in a bidentate configuration. On the other hand, the existence of an asymmetric mode alludes to a monobinding configuration via a single oxygen atom.21,27 Thus, the relative intensities of these two bands provide a qualitative measure of the surface configuration of the carboxylate headgroup. Yet, a small fraction of the headgroups do not undergo proton dissociation and are entrapped at the oxide|monolayer interface as carboxylic species as is evidenced by the peak at 1728 cm-1. The latter is presumably due to the fast deposition induced by applying an external potential. Comparing the spectra of the acids (Figure 10) reveals that the intensity ratio between the symmetric and asymmetric carboxylate stretching modes increases as the hydrocarbon chain lengthens. This is in accordance with the spectra shown in Figure 9 suggesting that the longer the acid, the more organized it is. In conclusion, we interpret that the majority of the SAM molecules are adsorbed on 316L stainless steel as carboxylate species, which are attached to the surface via ionic interactions. Increasing the chain length induces organization and increases binding to the surface through a bidentate mode. Nevertheless, we believe that the fact that the modification time is relatively short results in a mixture of binding configurations. Since the interaction between the carboxylic acid and the surface is via an oxide layer, it is conceivable that the
(41) Yillman, N.; Ulman, A.; Schildkraut, J. S.; Penner, T. L. J. Am. Chem. Soc. 1988, 110, 6136.
7506
Langmuir, Vol. 20, No. 18, 2004
Shustak et al.
of water increases. Increasing the water content above 20% did not yield a detectable monolayer of myristic acid. Conclusions
Figure 11. Reflection-absorption FTIR spectra of the C-H stretching region of myristic acid modified 316L stainless steel surfaces as a function of water content (v/v) in the modification solution. The ACN solutions contained 0.1 M TBATFB and (1) 0%, (2) 5%, (3) 10%, and (4) 20% water. Frequencies of the bands are summarized in the inset.
latter will depend on the percentage of water in the ACN modification solution. Moreover, water can compete with the alkanoic acids on the surface binding sites. Figure 11 shows the IR spectra of a myristic acid SAM as a function of water level (v/v) in the modification solution. The effect is evident and is expressed in both the relative intensities of the C-H vibrations and their wavenumbers (inset of Figure 11). Since the conditions of assembly and inspection of the layers were identical, the difference in the intensities of the bands suggests that the surface coverage and the organization of the monolayer decrease as the percentage
Self-assembled monolayers of n-alkanoic acids have been assembled on 316L stainless steel via electrochemical deposition. We find, based on a number of surface techniques, that long-chain acids form close-packed and highly ordered SAMs, while short chains result in disordered arrays. The IR spectra of long acids are consistent with monolayers of mainly carboxylate species in which the chains are conformationally ordered and aligned nearly perpendicular to the surface. The assembling procedure has been optimized and resulted in highly reproducible and stable films. Assembling the monolayers under potential control provides a number of advantages over the conventional method of self-assembly carried out under open-circuit potential. The major benefit lies in better control of the monolayer-substrate interface, which plays a major role in assembling monolayers, in particular on reactive metals such as stainless steel. Controlling the thickness of the oxide layer, which is the substrate on which the monolayer is deposited, is crucial to maintaining electron transfer across the monolayer. In addition, application of a potential decreases significantly the time of organization and seems to decrease the kinetic barrier of the assembly process. Acknowledgment. This research is partially supported by Elutex, Kiryat Shmona and by the Israel Science Foundation (200/02). LA036470Z
