Langmuir 2009, 25, 1427-1433
1427
A Study on the Formation and Thermal Stability of 11-MUA SAMs on Au(111)/Mica and on Polycrystalline Gold Foils Johanna Stettner,*,† Paul Frank,† Thomas Griesser,‡ Gregor Trimmel,‡ Robert Schennach,† Eduard Gilli,† and Adolf Winkler† Institute of Solid State Physics, Graz UniVersity of Technology, Petersgasse 16, A- 8010 Graz, Austria, and Institute for Chemistry and Technology of Materials, Graz UniVersity of Technology, Stremayrgasse 16, A-8010 Graz, Austria ReceiVed August 5, 2008. ReVised Manuscript ReceiVed October 20, 2008 In this article we present a comprehensive study of 11-mercaptoundecanoic acid self-assembled monolayer (SAM) formation on gold surfaces. The SAMs were prepared in ethanolic solution, utilizing two different substrates: Au(111)/ mica and polycrystalline gold foils. Several experimental methods (X-ray photoelectron spectroscopy, Fourier transform infrared spectroscopy, and atomic force microscopy) reveal a well-defined SAM. The main focus of this work, however, was to test the stability of these SAMs by thermal desorption spectroscopy. The spectra show different desorption peaks indicating different adsorption states and/or decomposition products on the surface. The assumed monolayer peak, which can be attributed to desorption of the intact molecule, is detected at 550 K. Further desorption peaks can be found, which result, e.g., from cracking of the S-C bond on the surface, depending on the substrate quality and on the residence time under ambient conditions.
Introduction There are lots of potential applications for self-assembled monolayers (SAMs), such as for corrosion inhibition, biosensors, lithography, organic electronics, and so forth, as already described comprehensively in several excellent review articles.1-6 The most frequently used organic molecules for SAM formation are alkane thiols, adsorbed on gold surfaces. The reason is that methylterminated, long alkane thiols (n > 6) form well-ordered, closely packed monolayers on Au(111).7 One quite often investigated system in view of applications is mercaptoundecanoic acid (11MUA, (HS-(CH2)10-COOH)) on gold surfaces, which not only forms well-ordered SAMs, but also provides the opportunity to modify the acid end group by other functionalized groups.8 Even though this system has already been studied in some detail, there are still a lot of discrepancies and unsolved questions. There is a discussion about the ideal preparation procedure as well as about the alignment of the molecules on the surface.9-12 Because of the importance of 11-MUA as a precursor for further modifications,8,13-17 a good understanding of the SAM structure and thermal stability is indispensable. Various experimental techniques have been used to characterize this system, such as X-ray photoelectron spectroscopy * Corresponding author. † Institute of Solid State Physics. ‡ Institute for Chemistry and Technology of Materials.
(1) Ulman, A. Chem. ReV. 1996, 96, 1533. (2) Schreiber, F. J. Phys.: Condens. Matter 2004, 16, R881. (3) Li, X. M.; Huskens, J.; Reinhoudt, D. N. J. Mater. Chem. 2004, 14, 2954. (4) Evans, S. D.; Williams, L. M. In Functional Organic and Polymeric Materials; Richardson, T. H., Ed.; Wiley: Chichester/New York, 2000; p 149. (5) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem. ReV. 2005, 105, 1103. (6) Poirier, G. E. Chem. ReV. 1997, 97, 1117. (7) Nuzzo, R. G.; Dubois, L. H.; Allara, D. L. J. Am. Chem. Soc. 1984, 112, 558. (8) Frey, B. L.; Corn, R. M. Anal. Chem. 1996, 68, 3187. (9) Ito, E.; Konno, K.; Noh, J.; Kanai, K.; Ouchi, Y.; Seki, K.; Hara, M. Appl. Surf. Sci. 2005, 244, 584. (10) Mendoza, S. M.; Arfaoui, I.; Zanarini, S.; Paolucci, F.; Rudolf, P. Langmuir 2006, 23, 582. (11) Wang, H.; Chen, S.; Li, L.; Jiang, S. Langmuir 2005, 21, 2633. (12) Arnold, R.; Azzam, W.; Terfort, A.; Wo¨ll, C. Langmuir 2002, 18, 3980.
(XPS),9,10,13,14,16,17 scanning tunneling microscopy (STM)9,10,18 and Fourier transform infrared spectroscopy (FTIR).17,19-21 FTIR, even though in principle being one of the most appropriate methods to study SAMs, has the disadvantage of requiring an ideal reference sample. Polarization-modulated infrared reflection-absorption spectroscopy (PM-IRRAS), discarding the need of a reference, was performed by Duevel and Corn22 for 11MUA on gold. The resulting spectra suggested a quite wide range of disorder. Also STM has shown that it is difficult to grow good SAMs of 11-MUA on Au(111).9,10 With respect to the thermal stability of 11-MUA on gold and to the influence of the substrate structure, however, there is still little information available. In this contribution we focus on these issues. By applying thermal desorption spectroscopy (TDS), in addition to XPS, FTIR, and atomic force microscopy (AFM), we have studied 11-MUA SAMs for three different scenarios: freshly prepared SAMs on Au(111)/mica, 1 month aged SAMs on Au(111)/mica, and SAMs on recrystallized gold foils, which consist of grains with stepped surfaces.
Experiment SAMs were prepared on gold surfaces ex situ in a 1-3 mM ethanolic solution of 11-MUA. No influence of the concentration in the denoted range on the SAM formation could be noticed. The immersion time was typically 48 h. After removal out of solution, (13) Cecchet, F.; Rudolf, P.; Rapino, S.; Margotti, M.; Paolucci, F.; Baggerman, J.; Brouwer, A. M.; Kay, E. R.; Wong, J. K. Y.; Leigh, D. A. J. Phys. Chem. B 2004, 108, 15192. (14) Czanderna, A. W.; King, D. E.; Spaulding, D. J. Vac. Sci. Technol. A 1991, 9(5), 2607. (15) Leopold, M. C.; Bowden, E. F. Langmuir 2002, 18, 2239. (16) Cavalleri, O.; Natale, C.; Stroppolo, M. E.; Relini, A.; Consulich, E.; Thea, S.; Novi, M.; Gliozzi, A. Phys. Chem. Chem. Phys. 2000, 2, 4630. (17) Jiang, P.; Liu, Z. F.; Cai, S. M. Langmuir 2002, 18, 4495. (18) Gorman, C. B.; He, Y.; Carroll, R. L. Langmuir 2001, 17, 5324. (19) Cutler, E. C.; Lundin, E.; Garabato, B. D.; Choi, D.; Shon, Y. S. Mater. Res. Bull. 2007, 42, 1178. (20) Nakano, K.; Yoshitake, T.; Yamashita, Y.; Bowden, E. F. Langmuir 2007, 23, 6270. (21) Shi, W.; Sahoo, Y.; Swihart, M. T. Colloids and Surfaces A: Physicochem. Eng. Aspects 2004, 246, 109. (22) Duevel, R. V.; Corn, R. M. Anal. Chem. 1992, 64, 337.
10.1021/la802534q CCC: $40.75 2009 American Chemical Society Published on Web 01/02/2009
1428 Langmuir, Vol. 25, No. 3, 2009 the samples were carefully rinsed with ethanol and dried with CO2 spray. To extend the field of our investigation, two different substrates were investigated. Au(111)/Mica. Gold grows epitaxially on a mica surface with a strong preference for the (111) orientation. The Au(111)/mica samples with a nominal Au-film thickness of 300 nm were purchased from Georg Albert PVD.23 With low-energy electron diffraction (LEED), the preferential (111) orientation could be verified. XPS showed a very low carbon contamination on these surfaces. Nevertheless, the samples were further cleaned under ultrahigh vacuum (UHV) conditions by Ar+-sputtering and subsequent annealing at 850 K, before they were put into the solution. The commonly performed cleaning in Piranha solution was not possible for these substrates, as the gold layer flaked off from the mica surface. Recrystallized Gold Foils. Gold foils (100 µm thick) were annealed and sputtered at 900 K for 24 h in UHV. As demonstrated in a previous paper,24 this procedure yields a clean, polycrystalline sample with grains up to several 100 µm in diameter. Since there exist different orientations of the grains, the future goal is to investigate the surface structure dependence of SAM formation simultaneously on a single sample. Recrystallization was verified by LEED, and the cleanliness of the surface was checked by XPS. Various sharp diffraction spots could be observed in a wide range of energies, indicating many differently oriented domains. The such prepared gold substrates were dismounted from the UHV chamber and immediately put into an ethanolic solution of 11-MUA. To characterize the 11-MUA SAM on the gold substrates, various surface-sensitive techniques have been used. Most experiments were performed in a UHV chamber, with a base pressure of 10-10 mbar. However, some of the experiments were performed without baking the UHV chamber, in which case the pressure was about 1 × 10-8 mbar. The apparatus was equipped with an X-ray photoelectron spectrometer (Leybold Heraeus, EA 10/100) with a twin X-ray source (Al and Mg cathodes), low energy electron diffraction optics with a multichannel plate (Omicron), and a quadrupole mass spectrometer (QMS; Balzers QMA 400) with a mass range from 1 to 500 amu. Furthermore, an Ar+-sputter gun was available to clean the surface. For TDS, the samples were mounted to a sample holder, which could be cooled by liquid nitrogen. In order to heat the sample, it was fixed to a steel plate, which could be heated by tantalum wires spot welded on the backside of the plate. Typically, the heating rate was 2 K/s. Further experimental details on the sample mounting and heating can be found elsewhere.25 AFM was performed ex situ using a Nanosurf Easyscan 2 scanning probe microscope in tapping mode. FTIR was performed with a Bruker IFS 66 v/s spectrometer using p-polarized light incident at 82° with respect to the surface normal.
Results and Discussion Characterization of the 11-MUA SAM. Evidence of the formation of a 11-MUA SAM was provided by means of XPS, FTIR, AFM, and TDS. Figure 1 shows an overview XPS spectrum of a well-prepared 11-MUA SAM on Au(111)/mica. The C1s and O1s peaks can clearly be seen in addition to the gold peaks, whereas the sulfur S2p peak is barely visible. In Figure 2a,b, detailed spectra of the S2p and C1s peaks, as obtained by using a higher data acquisition time and a narrower energy range, are shown, respectively. For the least-squares fit, Gauss distributions were used with a full width at half-maximum (fwhm) of 2.1 eV and a spin-orbit splitting separation of 1.2 eV.9 The observation of the S2p peak maximum at 162.7 eV is characteristic for sulfur bonded to a gold substrate; the strong attenuation indicates that the sulfur is indeed at the bottom of the SAM.10 Considering the C1s signal, the observed binding energy of 284.8 eV is (23) Georg Albert PVD Beschichtungen, www.georg-albert-pvd.de. (24) Mu¨lleger, S.; Mitsche, S.; Po¨lt, P.; Ha¨nel, K.; Birkner, A.; Wo¨ll, C.; Winkler, A. Thin Solid Films 2005, 484, 408. (25) Frank, P.; Hlawacek, G.; Lengyel, O.; Satka, A.; Teichert, C.; Resel, R.; Winkler, A. Surf. Sci. 2007, 601, 2152.
Stettner et al.
Figure 1. Overview XPS spectrum for an 11-MUA SAM on Au(111)/ mica. The sulfur signal is too small to be seen in this spectrum.
Figure 2. Detailed XPS spectra for an 11-MUA SAM on Au(111)/mica. The S2p peak at 162.7 eV indicates sulfur bonded to the substrate (a). The C1s peak at 284.8 eV stems from aliphatic carbon, and the broad shoulder at 289 eV is an indication of carboxylic carbon (b).
characteristic for the alkane chain. In addition, using the Gaussian fit procedure, one can denote a second peak at around 289 eV, which is attributed to the carboxylic carbon.10,13,17 A third peak might exist around 286 eV, which can be attributed to emission from the C atoms next to the sulfur atoms and the C atoms next to the carboxylic C atoms, respectively.10 Thus, the XPS results are compatible with the features that one would expect for a well-defined SAM. The verification of a SAM with AFM is a delicate endeavor, as it is difficult to demonstrate the presence of a well-ordered monolayer on a smooth surface. We use AFM therefore mainly to observe large-scale defects and multilayer agglomeration. Figure 3 shows a series of AFM images of 11-MUA on Au(111)/ mica. Figure 3a was taken immediately after removing the sample out of solution. One can see a large island (labeled “A”) with a height of about 9 nm, according to the cross section. The following images in Figure 3b,c were taken after 3 and 6 h under ambient conditions, respectively. The disappearance of the island is attributed to the multilayer desorption of 11-MUA. We have made sure that the island features do not stem from the ethanolic solvent, by applying AFM under comparable conditions to samples that had just been immersed in pure ethanol. It is shown
11-MUA SAM Formation on Gold Surfaces
Figure 3. Series of AFM images of 11-MUA on Au(111)/mica immediately after removing the sample out of solution (a), 3 h under ambient conditions (b), and 6 h under ambient conditions (c). The island labeled “A” stems from a multilayer that disappears under ambient conditions.
by vapor deposition experiments that the multilayer of 11-MUA desorbs under vacuum conditions around room temperature.26 The SAM, located underneath the multilayer, can not be distinguished from the clean substrate by AFM in this range of resolution. The relatively high desorption probability of the multilayer at room temperature can be seen as an advantage in comparison to other SAM/substrate systems, where long and careful ethanol rinsing is necessary to get rid of the multilayer.27 Further information of the SAM quality has been obtained by FTIR. Figure 4 shows the FTIR spectrum of 11-MUA on Au(111)/ mica. In Table 1, the respective band positions are specified. Asymmetric and symmetric stretch vibrations of the CH2 groups of the alkane chain yield the absorption bands at 2919 cm-1 and 2850 cm-1, respectively (Figure 4a); the absorption band of the deformation vibration of the CH2 group is observed at 1468 cm-1. The most important feature is the CdO stretching band of the acid end group, which occurs at 1710 cm-1 (Figure 4b). A similar value is described in several papers7,22 and is usually attributed to an acid group that is participating in the hydrogen bonding processes. The absorption band at 1435 cm-1 is attributed (26) Frank, P.; Stettner, J.; Nuβbacher, F.; Winkler, A.; Springer Proceedings in Physics; Rubahn, H.-G., Sitter, H., Horowitz, G., Al-Shamery, K., Eds.; 2009, Vol. 129. (27) Ka¨fer, D.; Witte, G.; Cyganik, P.; Terfort, A.; Wo¨ll, C. J. Am. Chem. Soc. 2006, 128, 1723. (28) Porter, M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc. 1987, 109, 3559.
Langmuir, Vol. 25, No. 3, 2009 1429
Figure 4. IRRAS spectra of an 11-MUA SAM on Au(111)/mica. The position of the CH2 stretch vibrations at 2919 cm-1 and 2850 cm-1, respectively, indicate a well-defined SAM (a). The CdO stretch vibration is observed at 1710 cm-1. The peak at 1435 cm-1 is due to the C-O-H deformation vibration; the feature at 1468 cm-1 is attributed to CH2 deformations (b). Table 1. Spectral Mode Assignments for 11-MUA SAMs on Au(111)/Mica freq, cm-1
mode assignment
2919 2850 1710 1468 1435
CH2 stretch, asymm. CH2 stretch, symm. CdO stretch CH2 deformation C-O-H deformation
to the C-O-H deformation vibration.12 As reported in ref 28, the location of the CH2 stretch vibrations gives information about lateral interactions between the n-alkyl chains. A so-called “liquidlike state” is characterized by peak positions at 2928 cm-1 and 2856 cm-1 for the νa(CH2) and νs(CH2) modes, respectively, while the peak positions of 2920 cm-1 (νa(CH2)) and 2850 cm-1 (νs(CH2)) correspond to the “solid-like state”. Since for these measurements the resolution of the FTIR spectrometer was set to 4 cm-1, our observed peak positions at 2919 cm-1 and 2850 cm-1 clearly correspond better with the “solid-like state”. The main features of the IR spectra are in good agreement with the experimental and theoretical results described in the paper by Nuzzo et al.7 Our main effort, however, was the characterization of the exsitu-prepared SAMs by TDS under vacuum conditions. TDS was performed after taking the sample out of solution, rinsing it with ethanol and mounting it in the UHV chamber. The measurements were performed on the following day after installation, so as to achieve good vacuum conditions without baking the system (p ≈ 1 × 10-8 mbar). When performing TDS measurements, one important issue is to distinguish between the
1430 Langmuir, Vol. 25, No. 3, 2009
Figure 5. TDS of a freshly prepared 11-MUA SAM on Au(111)/mica. The peak at 550 K is attributed to the monolayer (a). Some molecules decompose on the surface and desorb at higher temperatures, e.g., H2S (m ) 34) (b).
cracking of the molecules in the QMS upon ionization and the decomposition of the molecules on the surface. In order to figure out the cracking pattern of 11-MUA, it was, in a separate experiment, directly evaporated from a Knudsen cell into the QMS. Additional confirmation of this cracking pattern was obtained from multilayer desorption.26 In addition to the intact molecules at 218 amu, a quite complex cracking pattern exists, showing bunches of CxHy (with intense lines at m ) 27, 28, 39, 41 and 55), the COOH signal at m ) 45, and several discrete lines at 149 amu (e.g., C11H17), 167 amu (e.g., C11H19O), 182 amu (e.g., C11H18O2), and 199 amu (e.g., C11H19SO). For our TDS measurements, we multiplexed the QMS to various masses, from which only the most important ones are presented here. The desorption spectrum of a well-prepared 11-MUA SAM on Au(111)/mica is shown in Figure 5a,b. The most important feature is that only one main desorption peak of the intact molecules at 550 K exists, denoted as the R peak in the following. This peak shows a cracking pattern as observed for multilayer desorption, characterized in the figure by masses 218, 199, 45, and 39. Therefore we attribute this peak to the well-defined monolayer. There is still a discussion about a possible associative desorption of alkane thiols in the form of disulfides, as suggested by Lavrich et al.,29 Kondoh et al.,30 and Hara et al.31 We indeed could find a rather small feature of mass 434 at this temperature, which corresponds to twice the mass of the molecule. On the other hand, the observed cracking pattern is quite similar to the (29) Lavrich, D. J.; Wetterer, S. M.; Bernasek, S. L.; Scoles, G. J. Phys. Chem. B 1998, 102, 3456. (30) Kondoh, H.; Kodama, C.; Sumida, H.; Nozoye, H. J. Chem. Phys. 1999, 111(3), 1175. (31) Hara, M.; Tamada, K.; Hahn, C.; Nishida, N.; Knoll, W. Supramol. Sci. 1996, 3, 103.
Stettner et al.
Figure 6. TDS of an 11-MUA SAM on Au(111)/mica. The sample was under ambient conditions for 1 month. One can still see the monolayer peak at 550 K. A second peak with a different cracking pattern appears at higher temperature.
recorded gas phase spectrum of 11-MUA. We therefore suppose that associative desorption takes place, but it is a minor effect. However, a rather large amount of molecules seems to decompose on the surface and desorb at higher temperature. In particular, the C-S bond breaks, which leads to the desorption of H2S molecules (m ) 34) at around 650 K. In the same temperature range, CxHy fragments (mass 39) can also be observed (see Figure 5b; note the different y-scale). Interestingly, in this range, only little desorption of mass 45 appears, which would be characteristic of the acid group. This means that, in this case, not only is the S-C bond broken, but also the acid group is already separated from the molecule. It is difficult to determine the relative amount of the desorbing species quantitatively for the intact molecules and the fragment molecules, since the cracking pattern of the fragments is not known. In a first approximation, an estimate can be made by comparing the TDS areas of mass 39 (C3H3). This mass is an intense cracking product of the intact molecule and can be assumed to be a cracking product of the fragments also, which have lost either the HS anchor group or the acid headgroup. The evaluation yields that the ratio between the intact molecules and the decomposed molecules is about 50%:50%. Long-Term Stability of the 11-MUA SAM. TDS is an appropriate method to study long-term effects that occur on the 11-MUA SAM on Au(111)/mica under ambient conditions. Even small changes of the SAM, which cannot be detected by, e.g., XPS, can easily be seen in TDS. A TD spectrum of an 11-MUA SAM, which was kept 1 month under ambient light (day light and Neon light), is shown in Figure 6a,b. The R monolayer desorption peak at 550 K is still present, even though it is much smaller. Interestingly, one additional desorption peak around 700 K, denoted γ, can be observed. Moreover, a small desorption
11-MUA SAM Formation on Gold Surfaces
Langmuir, Vol. 25, No. 3, 2009 1431
Figure 7. The evolution of mass 434 (disulfide) during thermal desorption of a freshly prepared 11-MUA SAM on Au(111)/mica (corresponding to Figure 5), and from a sample after 1 month under ambient conditions (corresponding to Figure 6).
feature at a low temperature of approximately 360 K can be seen, which will be described in more detail later. The main feature, however, is the large γ peak at higher temperature, which suggests much more strongly bonded molecules. At first view it seems unlikely that the molecules get stronger bonded to the substrate while staying under ambient conditions. An important observation is the different cracking pattern of the γ peak, showing a higher amount of mass 185 (the sulfur-free mass), which indicates a significant change of the bonding of the molecule to the substrate (Figure 6a). A possible explanation for these more strongly bonded molecules could be a higher tilt angle, as, for example, reported by Horn et al.,32 or even flat lying molecules. Such molecules are not only chemisorbed by the sulfur but also physisorbed by the carbons of the chain. It is known from the literature that the physisorption enthalpy can become even larger than the chemisorption enthalpy for lying molecules with long chain lengths.29 These lying molecules will dissociatively desorb at high temperature, leading to desorption peaks with a different cracking pattern. The appearance of the significant amount of mass 197 at 700 K can only be explained by an MUA molecule that has either lost one oxygen atom and five hydrogen atoms or 21 hydrogen atoms on the surface. As mass 197 is not observed in the cracking pattern of the pure MUA molecule, we can exclude fragmentation in the mass spectrometer. The fact that in the temperature range above 600 K hydrogen is also released is a strong indication of the former reaction, because at this high temperature hydrogen cannot exist in the form of purely adsorbed hydrogen.33 Thus, we have to assume that additional bonding via unsaturated bonds of the aliphatic chain and the acid group can also contribute to the binding energy of the γ-peak. A rough estimate from the TDS peak areas of the mass 39 signal yields a ratio between the intact desorbing molecules at 550 K and the desorption of the lying fragments of 12%:88%. An interesting additional feature in the TDS of this sample is the appearance of mass 434 at about 360 K, as compared to the freshly prepared sample (Figure 7). This is clear evidence for the oxidation of the mercaptans to disulfides as a consequence of the long exposure to air. The small peak of mass 199 in the same temperature range (Figure 6a) can be attributed to cracking of the disulfide. TDS of 11-MUA on Gold Foil. To investigate a possible influence of the substrate crystallinity on the SAM formation, TDS of 11-MUA on recrystallized gold foils was also performed. A TD spectrum of a freshly prepared 11-MUA SAM on (32) Horn, A. B.; Russell, D. A.; Shorthouse, L. J.; Simpson, T. R. E. J. Chem. Soc., Faraday Trans. 1996, 92(23), 4759. (33) Stobinski, L.; Dus, R. Surf. Sci. 1993, 298, 101.
Figure 8. TDS of a freshly prepared 11-MUA SAM on polycrystalline gold foil.
recrystallized gold foil is shown in Figure 8. A quite different desorption behavior can be seen when compared with Figure 5. In this case, already on the freshly prepared SAM a strong indication of lying molecules can be observed, as indicated by the γ peak at 700 K. In addition, the R peak, which is attributed to the standing molecules, is much smaller. Moreover, a third peak appears at around 580 K, denoted as the β peak (Figure 8b). The cracking pattern for all three peaks is different, which tells us that different species desorb already from the surface. In particular, the β peak is characterized by a strong mass 34 (H2S) and mass 185 (the sulfur - free molecule) signal. Thus, the β peak is attributed to dissociative desorption upon cracking of the sulfur-carbon bond. Two similar desorption peaks were observed for related molecules of decanethiol by Lavrich et al.29 and anthracene-2-thiol on Au(111) by Ka¨fer et al.27 In the latter paper, the authors also attributed the second peak to the sulfurfree molecule. The interesting point is that we could not find the β peak in the TD spectrum for 11-MUA on Au(111)/mica. This means that the adsorption and desorption behavior of 11-MUA must be strongly dependent on the structure of the substrate. The recrystallized gold foil shows many crystallites with different surface orientation and surface atom packing density.24 In particular, many stepped surface planes are exposed on this recrystallized foil. One can assume that the sulfur is more strongly bonded on such step sites, and therefore the S-C bond can be broken more easily. In the literature, this β peak is sometimes also correlated with a dilute phase.27 One can assume that such a phase can more easily be formed on less close packed crystallographic surface planes. The γ peak is, in the case of the recrystallized foil, already strongly pronounced on the freshly prepared 11-MUA SAM (Figure 8). This is in good agreement with the assumed higher reactivity of individual surface planes. This could lead to more
1432 Langmuir, Vol. 25, No. 3, 2009
Stettner et al.
Figure 9. The desorption of hydrogen for 11-MUA on polycrystalline gold foil. The γ peak is characterized by desorption of mass 197 as well as desorption of mass 2. Purely adsorbed hydrogen cannot exist at this high temperature.
Figure 11. XPS signal of the C1s peak for 11-MUA on the gold foil, before heating (a), showing the peak of the carboxylic carbon around 289 eV, after heating at 650 K (b), where no peak of the carboxylic carbon can be detected, and after heating at 900 K (c). One observes a shift of the aliphatic carbon binding energy toward lower values.
Figure 10. XPS signal of the S2p peak for 11-MUA on the gold foil, before heating (a), after heating at 650 K (b), and after heating at 900 K (c). One observes a shift of the signal toward lower binding energies.
flat-lying molecules since, in this case, more energy is released. A rough estimate of the amount of molecules in the individual states from TDS yields R:β:γ ) 20%:30%:50%. The additional desorption of hydrogen, which has already been mentioned in the previous section, is shown in Figure 9. Thus, there is strong evidence that, at around 700 K, hydrogen is stripped off of the aliphatic backbone, leading to a backbone consisting of a linear sp carbon chain (polyyine). Linear carbon chains with sp hybridization represent one of the most simple but quite intriguing one-dimensional systems that have attracted considerable interest for many years.34 For further characterization of this high-temperature peak, XPS measurements were performed at different temperatures. Figure 10 displays the data for the S2p XPS signal, and Figure 11 shows the data for the C1s signal, before heating (curve a), after heating to 650 K (curve b), and after heating to 900 K (curve c). Several important features can be noticed: First, the (34) Martin, R. E.; Diederich, F. Angew. Chem,. Int. Ed. 1999, 38, 1350.
sulfur signal increased after heating the SAM to 650 K, although a significant amount of molecules had desorbed as intact molecules or in the form of H2S. This is a strong indication that now at least some part of the molecules are flat lying on the substrate. Furthermore, the binding energy of the S2p photoelectrons decreases, indicating a weakening of the S-C bond strength and a strengthening of the S-Au bond. In general, the binding energy of electrons decreases when the atom is bonded to a substrate. Similarly, the C1s spectrum also shifts toward lower binding energies and the shape of the carbon peak changes. The feature arising from the carboxylic carbon as observed before heating to 650 K disappears after heating. We believe that this is due to flat-lying molecules, where the chemical configuration of the carboxylic carbon has changed so strongly that no or only little related emission can take place. In addition, the C1s peak becomes more narrow, indicating less different chemical environments for the individual carbon atoms in the adsorbed species. This feature thus is in good agreement with our assumed configuration of flat-lying molecules, strongly bonded to the substrate by the carbons of the chain. Again, the appearance of mass 197, which stems from MUA molecules that have lost many H atoms already on the surface, corroborates the proposed model. After heating to 900 K, still a considerable amount of carbon and some sulfur remains on the surface, which are now even more strongly bonded to the surface, as deduced from the further decrease of the XPS binding energies of the S2p and C1s peaks, respectively.
Conclusion The formation and thermal stability of 11-MUA SAMs on different gold samples (Au(111)/mica, recrystallized gold foil) and under different aging conditions were investigated by means of XPS, AFM, FTIR, and TDS. XPS indicated sulfur bonded to the substrate and the existence of the carboxylic acid. AFM showed the existence of multilayer islands of 11-MUA, which disappeared under ambient conditions. More detailed information was obtained by FTIR measurements, indicating all the expected absorption bands of a 11-MUA SAM.
11-MUA SAM Formation on Gold Surfaces
The main focus of this work, however, was to investigate the thermal stability of the SAMs, the influence of the substrate structure, and the influence of aging on the 11-MUA SAM. A general feature for all investigated scenarios is the stability of the SAM up to 500 K. At this temperature, intact molecules of the SAM start to desorb with a peak maximum at 550 K. Some of the molecules decompose, leading to desorption of fragments at a higher temperature (between 600 and 800 K). As a rough estimate, about 50% of the molecules decompose for the freshly prepared SAM on Au(111)/mica. A 1 month aged SAM on Au(111)/mica yields a very different TDS, showing that only around 12% of the SAM molecules desorb intact at 550 K. A further desorption peak of various masses can be observed at 700 K. An interesting feature is the desorption of mass 197 at this temperature, a mass that is not a cracking product of 11MUA in the mass spectrometer. A similar desorption feature is observed for the freshly prepared SAMs on the recrystallized gold foils. In this case, about 20% of the molecules desorb intact at 550 K. Immediately afterward, at 580 K, about 30% of the molecules desorb by breaking the S-C bond, which leads to
Langmuir, Vol. 25, No. 3, 2009 1433
simultaneous detection of the sulfur-free molecule as well as of SH2. Obviously, this S-C bond breaking takes place more easily on the recrystallized gold foil, which mainly consists of high index planes and stepped surfaces. About 50% of the molecules again desorb as other fragments at approximately 700 K with a significant amount of mass 197. There are indications by XPS that the latter desorption feature might stem from lying molecules. The mass 197 can be best explained by desorption of molecule fragments which have lost either one oxygen and 5 hydrogen atoms or all the hydrogen atoms of the aliphatic backbone. The latter would yield molecules containing a linear sp carbon chain (polyyine) as the backbone. The many different features in the desorption spectra as a function of the substrate structure and the sample aging might explain the disagreements existing in the literature about 11MUA SAM formation on gold. Acknowledgment. We thank the Austrian Science Fund (FWF), P19197 and S9702-N08, for the financial support. LA802534Q
