5776
Langmuir 2004, 20, 5776-5781
Self-Assembled Organic Monolayers Terminated in Perfluoroalkyl Pentafluoro-λ6-sulfanyl (-SF5) Chemistry on Gold R. Winter, P. G. Nixon, and G. L. Gard Department of Chemistry, Portland State University, Portland, Oregon 97201
D. J. Graham and D. G. Castner NESAC/Bio, Departments of Bioengineering and Chemical Engineering, Box 351750, University of Washington, Seattle, Washington 98195
N. R. Holcomb and D. W. Grainger* Department of Chemistry, Colorado State University, Fort Collins, Colorado 80523-1872 Received January 27, 2004. In Final Form: April 21, 2004 Recently synthesized (Winter, R.; Nixon, P. G.; Gard, G. L.; Radford, D. H.; Holcomb, N. R.; Grainger, D. W. J. Fluorine Chem. 2001, 107, 23-30) SF5-terminated perfluoroalkyl thiols (SF5(CF2)nCH2CH2SH, where n ) 2, 4, and 6) and a symmetric SF5-terminated dialkyl disulfide ([SF5-CHdCH-(CH2)8-S-]2) were assembled as thin films chemisorbed onto gold surfaces. The adsorbed monolayer films of these SF5-containing molecules on polycrystalline gold were compared using ellipsometry, contact angle, X-ray photoelectron spectroscopy (XPS), time-of-flight secondary ion mass spectrometry (ToF-SIMS), and infrared spectroscopy (FTIR) surface analytical methods. The resulting SF5-dialkyl disulfide monolayer film shows moderate angle dependence in depth-dependent XPS analysis, suggesting a preferentially oriented film. The SF5-terminated perfluoroalkyl thiols exhibit angular-dependent XPS compositional variance depending on perfluoroalkyl chain length, consistent with improved film assembly (increasingly hydrophobic, fewer defects, and more vertical chain orientation increasing film thickness) with increasing chain length. TofSIMS measurements indicate that both full parent ions for these film-forming molecules and the unique SF5 terminal group are readily detectable from the thin films without substantial contamination from other adsorbates.
Introduction Perfluorinated alkyl silanes, alkyl thiols, and disulfides are recognized, interesting alternative components to hydrocarbon-based constituents for organic self-assembled monolayers (SAMs).2-13 Fluorine chemistry in perfluoroalkylated thin film components provides unambiguous chemical markers for surface analytical methods (e.g., X-ray photoelectron spectroscopy (XPS), near-edge X-ray * To whom correspondence should be addressed. E-mail:
[email protected]. (1) Winter, R.; Nixon, P. G.; Gard, G. L.; Radford, D. H.; Holcomb, N. R.; Grainger, D. W. J. Fluorine Chem. 2001, 107, 23-30. (2) Genzer, J.; Sivaniah, E.; Kramer, E. J.; Wang, J.; Xiang, M.; Char, K.; Ober, C. K.; Bubeck, R. A.; Fischer, D. A.; Graupe, M.; Colorado, R., Jr.; Shmakova, O. E.; Lee, T. R. Macromolecules 2000, 33, 6068. (3) Scho¨nherr, H.; Vancso, G. J. Langmuir 1997, 13, 3769. (4) Scho¨nherr, H.; Ringsdorf, H.; Jaschke, M.; Butt, H.-J.; Bamberg, F.; Allinson, H.; Evans, S. D. Langmuir 1996, 12, 3898. (5) Scho¨nherr, H.; Ringsdorf, H. Langmuir 1996, 12, 3891. (6) Smith, D. L.; Wysocki, V. H.; Colorado, R., Jr.; Shmakova, O. E.; Graupe, M.; Lee, T. R. Langmuir 2002, 18, 3895. (7) Tsao, M.-W.; Hoffman, C. L.; Rabolt, J. F.; Johnson, H. E.; Castner, O. G.; Erdelen, C.; Ringsdorf, H. Langmuir 1997, 13, 4317. (8) Biebuyck, H. A.; Whitesides, G. M. Langmuir 1993, 9, 1766. (9) Liu, G.; Fenter, P.; Chidsey, C. E. D.; Ogletree, D. F.; Eisenberger, P.; Salmeron, M. J. J. Chem. Phys. 1994, 101, 4301. (10) Kim, H. I.; Koini, T.; Lee, T. R.; Perry, S. S. Langmuir 1997, 13, 7192. (11) Miura, Y. F.; Takenaga, M.; Koini, T.; Graupe, M.; Garg, N.; Graham, R. L.; Lee, T. R. Langmuir 1998, 14, 5821. (12) Ishida, T.; Yamamoto, S.; Mizutani, W.; Motomatsu, M.; Tokumoto, H.; Hokari, H.; Azehara, H.; Fujihira, M. Langmuir 1997, 13, 3261. (13) Chidsey, C. E. D.; Lociano, D. N. Langmuir 1990, 6, 682.
absorption fine structure (NEXAFS) spectroscopy, Fourier transform infrared (FTIR) spectroscopy), distinct chemical and physical properties compared to purely hydrocarbon analogue materials, and extremely apolar, nonwetting surfaces interesting for various technologies. Work to date has shown that perfluoroalkyl thiols and disulfides form high-quality SAMs on noble and coinage metal supports with distinctly different epitaxial commensurate structures c(7 × 7)14 versus p(2 × 2) known for hydrocarbon analogues.13 Nonetheless, these fluorinated adlayers form stable, highly organized surfaces exhibiting highly apolar, hydrophobic water contact angles consistent with long-known surfaces terminated in oriented -CF3 terminal groups.15,16 Specific interfacial properties, including nonwetting, lubrication, unique thin film structures, and inert SAM chemistry, have been a substantial interest for perfluoroalkyl surface species and continue to be investigated.17 A new alternative, the perfluoroalkyl pentafluoro-λ6-sulfanyl terminal group (-(CF2)n-SF5),1,18-20 has the potential to provide new low-energy fluorinated (14) Scho¨nherr, H.; Vancso, G. J. In Fluorinated Surfaces, Coatings and Films; Castner, D. G., Grainger, D. W., Eds.; ACS Symposium Series 787; American Chemical Society: Washington, DC, 2001; pp 55-30. (15) Zisman, W. A.; Schulman, F. J. Am. Chem. Soc. 1952, 74, 2133. (16) Grainger, D. W. In Fluorinated Surfaces, Coatings and Films; Castner, D. G., Grainger, D. W., Eds.; ACS Symposium Series 787; American Chemical Society: Washington, DC, 2001; pp 1-14. (17) Fluorinated Surfaces, Coatings and Films; Castner, D. G., Grainger, D. W., Eds.; ACS Symposium Series 787; American Chemical Society: Washington, DC, 2001. MRS Bull. 1996, 21, 16-53 (special issue on Polymer Surfaces and Interfaces; Koberstein, J. T., Ed.).
10.1021/la040011w CCC: $27.50 © 2004 American Chemical Society Published on Web 06/11/2004
SAMs Using Pentafluoro-λ6-sulfanyl Chemistry
surfaces with interfacial properties analogous to those of previous perfluoroalkyl systems,2-17 in addition to interesting new surface analytical and physical chemistry of more fundamental relevance. Substantially less is reported about materials or surfaces containing SF5 chemistry. Acrylate monomers bearing perfluoralkyl-SF5 side chain chemistry were shown to enrich surfaces of photopolymerized films comprising mostly aliphatic acrylic comonomers to very high nonstoichiometric levels, exhibiting apolar, nonwetting surface properties comparable to those of known fluorinated polymer surfaces (e.g., fluorinated ethylene propylene (FEP), poly(tetrafluoroethylene) (PTFE)).17-19 Silicon oxide supported alkylsilane thin films terminated in (-CF2)n-SF5 groups (n ) 2 or 4) exhibit high aqueous contact angles comparable to those of perfluoroalkyl silanes.20 This initial evidence shows that the -SF5-terminated perfluoralkyl group has some properties analogous to those of known fluorinated systems. This chemistry might serve as an interesting alternative to more generic perfluorocarbons in thin film applications and warrants further research on its utility in applications. One important question regards thin film organization on surfaces critical to perfluorinated interfacial properties16,21 and lateral packing ability due to bulky pentafluoro sulfanyl bonding geometry in the -SF5 terminal group that may compromise film organization.18-20 In this contribution, new asymmetric bis(sulfur-derivatized) long-chain hydrocarbon derivatives containing the perfluoroalkyl -SF5 or alkyl-SF5 terminal groups at one end and thiol or disulfide anchoring groups at the other end, respectively,1 were fabricated as SAM-analogue thin films by chemisorption onto polycrystalline gold supports. Surface analysis of these films using ellipsometry, angle-resolved XPS, time-of-flight secondary ion mass spectrometry (ToF-SIMS), FTIR spectroscopy, and aqueous contact angles distinguishes differences in film quality between these two SF5-terminated perfluoroalkyl chain chemistries in thin films. Experimental Section Materials. All water was polished Millipore grade (18 MΩ cm resistivity). Methylene chloride (Burdick and Jackson, highpurity grade) and hexane (J.T. Baker, high-purity grade) were used as received. Film-forming compounds SF5(CF2)nCH2-CH2SH (n ) 2, 4, and 6) and [SF5-CHdCH-(CH2)8-S-]2 shown in Figure 1 were prepared as previously described.1 These SAM components were characterized by NMR, elemental analysis, and gas chromatography-mass spectrometry (GC-MS) as previously described1 prior to SAM formation and further study. Gold Substrate Preparation. Silicon wafers (Wacker Siltronic, Portland, OR) were cleaned by 24-h immersion in a Piranha solution (1:1 mix of concentrated sulfuric acid and 30% H2O2) etch (caution: this solution contains reactive peroxides!), followed by copious rinsing in pure water, then anhydrous ethanol, and dry nitrogen gas. Silicon wafer pieces were cut by scoring with a diamond scribe. Cleaned wafers were vapordeposited with approximately 100 Å of chromium first as an adhesive underlayer and then with 2000 Å of gold (99.998%) under a (1.6-2) × 10-6 Torr vacuum in a metal evaporator. Goldcoated wafers were cut into 2.2 × 3.7 cm pieces for FTIR measurement and 0.9 × 2.0 cm pieces for ellipsometry, contact angle, ToF-SIMs, and XPS measurements using a diamond scribe on their opposite side. (18) Nixon, P. G.; Gard, G. L.; Hu, Y. H.; Castner, D. G.; Holcomb, N. R.; Grainger, D. W. Chem. Mater. 1999, 11, 3044. (19) Winter, R.; Nixon, P. G.; Terjeson, R. J.; Mohtasham, J.; Holcomb, N. R.; Grainger, D. W.; Graham, D.; Castner, D. G.; Gard, G. L. J. Fluorine Chem. 2002, 115, 107-113. (20) Nixon, P. G.; Winter, R.; Castner, D. G.; Holcomb, N. R.; Grainger, D. W.; Gard, G. L. Chem. Mater. 2000, 12, 3108. (21) Genzer, J.; Efimoto, K. Science 2000, 290, 2130.
Langmuir, Vol. 20, No. 14, 2004 5777
Figure 1. Structures for the SF5 (pentafluoro sulfanyl)containing molecules used for thin film formation on gold (for synthesis, see ref 1). Circles evident on the left SF5-terminated alkyl thiol structure indicate relative cross-sectional areas for (solid line) the SF5 pentafluoro sulfanyl terminal unit (32.6 Å2, from space-filling molecular model),20 (dashed-dotted line) the CF2 perfluoroalkyl methylene unit (range ∼ 26.4-33.2 Å2, depending on consideration of isolated units versus helical chains),13,14,30,31 and (dotted line) the hydrocarbon methylene unit (∼18.5-20.5 Å2 depending on consideration of isolated units versus close-packed all-trans chains).22,29 For comparison, published values for the terminal perfluoromethyl group crosssectional areas range from 29.0 Å2 (space-filling molecular model)20 to 31-32 Å2 (monolayer experimental measurements).28,32 Nearest neighbor close-packed intermolecular spacing for perpendicularly oriented perfluoroalkyl chains is 5.5 Å,33 similar to that found on perfluoroalkyl SAM systems on gold.3,14 Monolayer Formation. Gold substrates were placed in 1.03.0 mM (based on the thiol or dithio group content) pure solutions of the sulfur-containing molecules SF5(CF2)nCH-CH2-SH (n ) 2, 4, and 6) and [SF5-CHdCH-(CH2)8-S-]2 in hexanes at room temperature for 24 h. Adsorption solutions were first deoxygenated by bubbling with nitrogen for 5 min. Gas bubbles absorbed to gold substrate surfaces after immersion were removed by gentle agitation. Monolayer samples were then removed and immersed in fresh solvent for about 1 min, rinsed with solvent, chloroform, ethanol, and Millipore water in sequence, and blown dry under nitrogen prior to analysis. Surface Analysis of SF5-Perfluoroalkyl Thiol and -Dialkyl Disulfide Adlayers on Gold. Ellipsometry. Film thicknesses were measured using a Gaertner L117 ellipsometer with a wavelength of 6328 Å (He-Ne laser) at an incident angle of 70°. Each underivatized, freshly deposited gold surface on a silicon wafer was used as its own internal reference, and 3-4 measurements were made on each sample. SAM film thicknesses were calculated numerically assuming a refractive index of 1.3822 for the adsorbed SF5-perfluoroalkyl and 1.4522 for the hydrocarbonrich SF5-dialkyl disulfide films on gold, respectively. X-ray Photoelectron Spectroscopy. XPS experiments were performed on a Surface Science SSX-100 spectrometer (Mountain View, CA) equipped with a monochromatic Al KR source, a hemispherical analyzer, and a multichannel detector. Typically, spectra were collected with the axis of the analyzer lens at 55° with reference to the sample surface normal, and the operating pressure was approximately 3 × 10-9 Torr. High-resolution spectra were obtained at a pass energy of 50 eV using a 1000 µm spot size. Both survey spectra and data for quantitative analysis were collected at a pass energy of 150 eV and a spot size of 1000 µm. The binding energy (BE) scales for all spectra were referenced to the C1s C-H peak at 285.0 eV. Peak fitting of the highresolution spectra was done using Gaussian peak shapes with commercial software supplied by Surface Science Instruments. For calculation of XPS elemental composition, the analyzer transmission function was assumed not to vary with photoelectron kinetic energy (KE),23 the photoelectron escape depth was assumed to vary as KE0.7,23 and Scofield’s photoionization cross sections were used.24 (22) Ulman, A. An Introduction to Ultrathin Organic Films from Langmuir-Blodgett to Self-Assembly; Academic Press: New York, 1991. (23) Application note from Surface Science Instruments, Mountain View, CA, 1987.
5778
Langmuir, Vol. 20, No. 14, 2004
Angle-dependent XPS data were collected at nominal photoelectron takeoff angles of 0°, 55°, and 80°. The takeoff angle was defined as the angle between the surface normal and the axis of the analyzer lens system. Using mean free paths calculated from the equations given by Seah and Dench,25 the sampling depth (3 times the mean free path) for C1s photoelectrons should decrease from 90 to 15 Å as takeoff angle increases from 0° to 80°.26 From analysis of replicates, the typical XPS uncertainties were observed to be less than ×b11.0 atomic % for gold, carbon, and fluorine, and less than ×b10.5 atomic % for sulfur.27 Static Contact Angle Analysis. Sessile drop contact angle analysis (Rame´-Hart 100 apparatus) used purified (Millipore 18 MΩ cm resistivity) water drops (2 µL) on three separate spots on each film surface in a controlled environment (100% relative humidity). Measurements were taken on both sides of water drops at ambient temperature 30-40 s after drops were applied to surfaces. Contact angle data report the average of three drops at different surface locations. Reflection Fourier Transform Infrared Spectroscopy. Grazing angle external reflection spectra were recorded with a variable angle specular reflectance accessory (Spectra Tech) at an incident angle of 83° with respect to the surface normal, using a p-polarized infrared beam (wire-grid polarizer, Perkin-Elmer). The spectrometer and sample chamber were purged with nitrogen or dry, CO2-free air. Clean, freshly deposited gold substrates were used as background, and 1024 scans of both the sample and reference were collected to obtain good signal-to-noise ratios. Interchange of monolayer sample and bare gold reference was required to avoid long-term instabilities and minimize baseline artifacts. Reference subtraction and flattening were achieved using Spectracalc software (Galactic Industries), but no curve smoothing or other spectral alterations were applied. Time-of-Flight Secondary Ion Mass Spectrometry. The ToFSIMS data were acquired using a model 7200 Physical Electronics instrument (PHI, Eden Prairie, MN). The 8 keV Cs+ ion source was operated at a current of 1.5 pA and a pulse width of 0.9 ns. Data were acquired over a mass range from m/z ) 0-2000 for both positive and negative secondary ions. The ion beam was moved to a new spot on the sample for each spectrum. The total ion dose used to acquire each spectrum was less than 2 × 1012 ions/cm2. The area of analysis for each spectrum was 0.01 mm2. Secondary ions were extracted into a two-stage reflectron timeof-flight mass analyzer with a potential of 3 kV. A secondary ion-focusing lens between the analyzer entrance and drift region was held at 1 kV, promoting high angular acceptance and good transmission of ions. The band-pass of the analyzer is 100 eV, and an independent adjustable grid voltage (deceleration) allows energy focusing to be performed. The ions were postaccelerated to 10 kV and converted to charge pulses by a stacked pair of chevron-type multichannel plates (MCP). The signals were detected using a 256 stop time-to-digital converter (TDC) with 156 ps time resolution. The mass scales for the negative ion spectra were calibrated using the CH-, C2H-, and AuS- peaks. The positive ion spectra were calibrated using the CH3+, C2H5+, and AuSCH2+ peaks. The mass resolution (m/∆m) for both positive and negative secondary ions was between 6000 and 10 000.
Results and Discussion SF5-containing perfluoroalkyl thiol and dialkyl disulfide films (see structures in Figure 1) adsorbed on polycrystalline gold were characterized using ellipsometry, aqueous wetting (contact angles), glancing angle FTIR spectroscopy, angle-resolved XPS, and ToF-SIMS. Ellipsometry provided film thicknesses of 8 Å for the shortest (n ) 2) SF5-thiol, 11 Å for the intermediate (n ) 4) length SF5thiol, 13 Å for the longest (n ) 6) SF5-thiol, and 12 Å for the SF5-dialkyl disulfide on gold. These values assume an (24) Scofield, J. H. J. Electron Spectrosc. Relat. Phenom. 1976, 8, 129. (25) Seah, M. P.; Dench, W. A. Surf. Interface Anal. 1979, 1, 2. (26) Tyler, B. J.; Castner, D. G.; Ratner, B. D. Surf. Interface Anal. 1989, 14, 443. (27) Castner, D. G.; Hinds, K.; Grainger, D. W. Langmuir 1996, 12, 1235.
Winter et al.
average refractive index of n ) 1.38 for the chemisorbed perfluoroalkyl thiol monolayer and 1.45 for the dialkyl disulfide monolayer.22 These film thickness values are smaller than expected and consistent with values for tilted or disordered monolayers of the thiols and disulfide on gold.2-13 Aqueous static contact angles for these same films are 90°, 92°, 112°, and 81° for the short (n ) 2), intermediate (n ) 4), and long (n ) 6) SF5-perfluoroalkyl thiols and SF5-dialkyl disulfide, respectively, again generally lower than expected for highly ordered perfluoroalkyl SAMs2-13 except for the longest thiol, consistent with the disorder proposed from ellipsometry, but also indicating that these SF5-thiol and disulfide film surfaces are approximately as hydrophobic as other analogous apolar alkyl thiol monolayers.13 Reduced water contact angles over that expected for highly perfluorinated thiols2-13 indicate reduced chain order or poor lateral organization in the shorter (n ) 4, 6) chain monolayers with likely defects contributing to exposure of more polar chemistry (i.e., gold substrate or contaminant adsorbates). This is perhaps anticipated, based on the spatial mismatch of the bulky SF5 umbrella-like terminal group and short mixed perfluoroalkyl/alkyl underlying chain above the thiol and disulfide anchor groups. We have previously published space-filling models for analogous SF5-perfluoroalkyl silane molecules (generated using PC Spartan Pro, non-energy-minimized).20 Both side-on and top-down perspectives for these molecules illustrate differences in molecular dimensions for the SF5 terminal group over more conventional perfluoroalkyl perfluoromethyl terminal analogues as well as perfluoroalkyl versus alkyl chains, demonstrating larger molecular area requirements for the SF5-perfluoroalkyl chains over hydrocarbon and perfluorocarbon SAM constituents in monolayers.20 The top-down cross-sectional view shows that the SF5 terminal space-filling model group area is ∼10% larger than the CF3 terminal group and published cross-sectional area requirements for close-packed perfluoroalkyl chains.14,28 The cis-olefinic hydrocarbon group adjacent to -SF5 in the SF5-terminated dialkyl disulfide forces substantially larger area requirements than typical saturated all-trans hydrocarbon chains in general,29 and the SF5-perfluoroalkyl thiol20 and aligned perfluorocarbon chains in particular.30,31 Hence, steric problems in monolayer packing might be anticipated for these pure SF5-terminated systems. Attempts to fill proposed SF5-monolayer defects with mixed coadsorbed perfluoroalkyl thiol/SF5-perfluoroalkyl thiol monolayers were unsuccessful (XPS analysis presented below). FTIR p-polarized surface reflection spectra from these SAMs (data not shown) are consistent with bulk FTIR spectra recently reported for similar SF5-containing molecules1,19 and specifically show surface-derived vibrational bands unique to S-F stretching modes in the films (835-876 cm-1), C-F stretching modes (1100 and 1225 cm-1), and other CF2 group vibrational modes (1050-1250 cm-1). The hydrocarbon-derived C-H vibrational modes in the 2915-2980 cm-1 region are evident but notably weaker due to either surface orientational selection effects (p-polarization) or low relative abundance in the films. Generally, these reflection spectra were not of sufficient (28) Takahara, A.; Koijo, K.; Ge, S. R.; Kajiyama, T. J. Vac. Sci. Technol., A 1996, 14, 1747. (29) Lagaly, G.; Stuke, E.; Weiss, A. Prog. Colloid Polym. Sci. 1976, 60, 102. (30) Bunn, C. W.; Howells, E. R. Nature 1954, 174, 549. (31) Rabolt, J. F.; Fanconi, B. Polymer 1977, 18, 1258. (32) Bernett, M. K.; Zisman, W. A. J. Phys. Chem. 1963, 67, 1534. (33) Polymer Handbook, 3rd ed.; Brandup, J., Immergut, E. H., Eds.; Wiley: New York, 1989; Part V, p 37.
SAMs Using Pentafluoro-λ6-sulfanyl Chemistry
Langmuir, Vol. 20, No. 14, 2004 5779 Table 2. Angle-Dependent XPS Data (Depth Dependence) for Monolayer Films of SF5(CF2)6(CH2)2SH on Gold depth dependence
Figure 2. XPS high-resolution C1s spectra for SF5-(CF2)6(CH2)2-SH monolayers on gold substrates at a takeoff angle of 80° (sampling depth ∼ 15 Å).
Figure 3. XPS high-resolution S2p spectrum for SF5-(CF2)6(CH2)2-SH monolayers on gold substrates. Takeoff angle ) 0°, sampling depth ∼ 90 Å. Table 1. XPS Data Compiled for SF5-Terminated Perfluoroalkyl Thiol and Dialkyl Disulfide Monolayers on Gold (Fixed Sampling Depth of ∼45 Å) monolayer film chemistry
F
XPS elemental atomic percent C S O Au
SF5-(CF2)2-CH2-CH2-SH all elements (film) 22.6 29.8 4.2 without Aua 35.9 47.3 6.7 theory w/o Au 60.0 26.7 13.3
6.4 10.2
37.0
SF5-(CF2)4-CH2-CH2-SH all elements in film 31.9 27.6 3.4 without Aua 47.3 41.0 5.0 theory w/o Au 61.9 28.6 9.5
4.6 6.8
32.5
SF5-(CF2)6-CH2-CH2-SH all elements in film 51.1 20.9 4.4 without Aua 65.3 26.7 5.7 theory w/o Au 63.0 29.6 7.4
1.8 2.4
21.8
[SF5-CHdCH-(CH2)8-S-]2 all elements in film 18.4 37.8 4.2 without Aua 29.4 60.3 6.8 theory w/o Au 29.4 58.8 11.8
2.2 3.5
37.3
a Au compositional weighting removed to eliminate substrate contribution for theoretical comparison.
quality to permit assessment of orientational anisotropy, indicative of film disorganization. Only qualitative comparative FTIR analysis was performed. XPS data for the SF5 monolayers are shown for select energies in Figures 2 and 3 and summarized in Tables 1 and 2. The SF5-perfluoroalkyl thiol high-resolution spectra (Figures 2 and 3) show features very similar to those from surfaces of SF5-perfluoroalkylacrylate films18 and SF5-terminated perfluoroalkylsilane monolayers.20 Specifically, the high-resolution carbon C1s spectrum for all SF5-thiol SAMs exhibits components characteristic of both perfluoroalkyl and hydrocarbon chemistry in the thiol molecules and hydrocarbon chains in the dialkyl disulfide
XPS elemental atomic percent -S-Au SF5 C O Au (-S-Au/SF5)
angle (deg)
depth (Å)
0 55 80
90 45 15
SF5-(CF2) 6-(CH)2-SH 44.8 1.3 1.7 18.3 1.6 32.3 51.1 1.9 2.5 20.9 1.8 21.8 59.8 1.2 4.0 22.5 1.7 10.8
80
15
67.0
F
Without Aua 1.3 4.5 25.2 1.9
Theory without Au 63.0 3.7 3.7 29.6
0.77 0.76 0.30 0.29 1.00
a
Au compositional weighting removed to eliminate substrate contribution for theoretical comparison.
molecules (typified by Figure 2). Additionally, the sulfur S2p spectrum (Figure 3) is interesting, showing the characteristic overlapping spin doublets for sulfur26 in two forms from each of these molecules: (1) peaks shifted to extremely high binding energy due to the fluorine electron withdrawing influence at high stoichiometry in -SF518 and (2) peaks at low binding energy due to formation of a sulfur-metal thiolate species.27 For the fixed 55° XPS takeoff angle (sampling depth ∼ 45 Å, Table 1), all pure monolayers exhibit elemental XPS signatures characteristic of the expected constituent chemistry. In addition to near-stoichiometric amounts of F, C, and S, each sample has detectable amounts of oxygen in the monolayer. Because the S2p spectrum (Figure 3) shows no evidence for oxidized sulfur species,27 this oxygen is likely from adventitious contaminants. When monolayer elemental composition is normalized to eliminate substrate (Au) contributions, differences between the XPS analyses of these various adlayer films are evident. The shorter (n ) 2, 4) SF5-thiols and SF5-dialkyl disulfide monolayers demonstrate actual elemental compositions containing less F and more C than predicted from calculated theoretical compositions of these surfaces (Table 1). This suggests the presence of adventitious contamination that was not removed from the surface during assembly, resulting in disruption of surface order. This is consistent with the somewhat reduced thickness and contact angles also seen for these surfaces. Again, this might be expected from films comprising short underlying alkyl chains precluded from high lateral packing density by a large -SF5 terminal umbrella-like structure. The longest (n ) 6) SF5-perfluoroalkyl thiol exhibits improved monolayer quality. Table 1 XPS data at ∼45 Å sampling depth for this film elemental composition are close to expected theoretical bulk thiol and disulfide composition. This supports a more uniform, organized film structure consistent with observed ellipsometry and film wetting data. Additional insight is gained from depth-dependent XPS data for the longest (n ) 6) SF5-perfluoroalkyl thiol (Figure 2 and Table 2). Figure 2 shows the high-resolution C1s spectrum at a takeoff angle of 80° (sampling depth of 15 Å). The concentration of perfluorocarbon (70%) and hydrocarbon (30%) species in this spectrum is comparable to the expected theoretical composition (75% perfluorocarbon and 25% hydrocarbon). This is consistent with the elemental composition results in Table 1 at a takeoff angle of 80°. Table 2 data show some additional trends in film composition for the longer (n ) 6) thiol that support improved film organization. As sampling depth moves from deep within the film (takeoff angle of 0°) to very shallow 15 Å depths (takeoff angle of 80 Å), fluorine, sulfur, and carbon compositions all increase substantially, consistent
5780
Langmuir, Vol. 20, No. 14, 2004
Winter et al.
Figure 4. ToF-SIMs negative ion spectrum of the longest SF5-thiol on gold. Only the major peaks are labeled on the spectrum to maintain clarity. The SF5 terminal group and its subfragments are clearly seen in the spectrum.
with more upright orientation of SF5-perfluoroalkyl thiol and SF5-disulfides on this support. Gold substrate signals decrease proportionally. Normalization of 15-Å-depth film composition to F, S, O, and C elements alone (Table 2) shows nearly stoichiometric agreement of actual with theoretical film composition. In particular, the decrease in the S-Au to SF5 ratio with increasing takeoff angle (decreasing sampling depth) is consistent with the presence of the gold thiolate species25 (clearly shown at 162 eV in Figure 3) at the gold-monolayer interface and the SF5 species at the vacuum-monolayer interface. XPS data for surface enrichment support successful fabrication of nonrandom, preferentially oriented (but not highly ordered) thiol and disulfide monolayer films from this longer (n ) 6) SF5-thiol and SF5-disulfide, even with the anticipated structural problems from the large SF5 terminal cap on these layers. Stronger surface enrichment to nonstoichiometric levels might be expected in the XPS angle-resolved data for fluorine and sulfur if these films were highly organized laterally over large length scales, indicating that the observed XPS enrichment for fluorine (∼6% increase over glancing angle theoretical composition) and for the SF5 chemistry (∼22% increase over theoretical composition, Table 2) derive from films with structural anisotropy but not high alignment over large lateral (or area) length scales. In mixed monolayers formed from simultaneously adsorbed 1 mM solutions of the longer (n ) 6) SF5-thiol and 8-carbon CF3-perfluoroalkyl thiol molecules,12 the XPS data (not shown) almost uniformly exhibited nearly complete exclusion of the SF5-containing thiol component from these monolayers at all mixing ratios studied. XPS C1s spectra for all mixing ratios studied resembled those previously published for pure perfluoroalkyl thiol monolayers.12 Only at 70% SF5-component mixing bias was a trace of the distinctive SF5 S2p XPS peak observed. This indicates preferential adlayer formation from the CF3perfluoroalkyl thiol, even at very low (1%) feed ratios in solution, that outcompetes SF5-thiol for gold adlayer sites.
Figure 5. ToF-SIMs positive ion spectrum of the longest SF5-thiol on gold. Only the major peaks are labeled.
ToF-SIMS analysis of both the SF5-thiol and SF5disulfide samples provides further evidence of the presence of the long-chain SF5-functionalized molecules (Figures 4 and 5). Ion yields were compared with bulk mass spectrometry (GC-MS) information published for the SF5perfluoroalkyl long-chain systems1 and ToF-SIMS results from SF5-perfluoroacrylate polymer films.19 The low-mass region of the negative ion ToF-SIMS spectra of the longest SF5-thiol on Au (see Figure 4) shows a series of strong peaks related to the SF5 functional group, including F(m/z ) 18.99), F2- (m/z ) 37.99), SF2- (m/z ) 69.97), SF3(m/z ) 88.97), SF4- (m/z ) 107.96), and SF5- (m/z ) 126.95). In the high-mass region (not shown), representative thiol peaks were seen (at low intensity) for the complete parent disulfidemoleculeatm/z)1170.86(Au[SF5(CF2)6CH2CH2S]2-) and for an oxidized parent SF5-thiol at m/z ) 534.93 (SF5(CF2)6CH2CH2SO3-). Other peaks in the negative ion spectra included Cl- (m/z ) 34.97), CN- (m/z ) 26.003), CNO- (m/z ) 41.99), and various recombination fragments of Au, S, F, and Cl in the form of AuxS, AuxFy, and AuxClF.
SAMs Using Pentafluoro-λ6-sulfanyl Chemistry
Since gold has high affinity for adventitious chlorine and chloride and also since precursor synthons for the SF5containing disulfide molecules described here derive from chlorohexane adducts,1 these trace halo-fragment ions are not unexpected but must be present at low levels (
