Stability of a Surface Adlayer at Elevated Temperature: Coronene and Heptanoic Acid on Au(111).

Transcription

1 "This document is the Accepted Manuscript version of a Published Work that appeared in final form in J. Phys. Chem. C, copyright American Chemical Society after peer review and technical editing by the publisher. To access the final edited and published work see Journal of Physical Chemistry C, 2008; 112(6); Stability of a Surface Adlayer at Elevated Temperature: Coronene and Heptanoic Acid on Au(111). William A. English + and K.W. Hipps * Department of Chemistry and Materials Science Program Washington State Universiy Pullman, WA Hipps@wsu.edu ABSTRACT The temperature dependence of the stability of the adlayer formed at the interface between Au(111) and a coronene-heptanoic acid solution is reported. At coronene concentrations above 1.5x10-4 M and at 21 C, a dense coronene adlayer is observed. At higher temperatures (up to about 60 C) or in a more dilute solution, a coadsorbed coronene-heptanoic acid structure is observed. This coadsorbed structure is very hardy, surviving exposure to the pure solvent for periods of many minutes and temperatures as high as 105 C. Loss of the ordered monolayer with high temperature solvent exposure is also accompanied by the formation of at least two additional surface structures involving both coronene and heptanoic acid. This work demonstrates that there is much to be learned by exploring the solution-solid interface at elevated temperatures with STM.

2 INTRODUCTION Supramolecular chemistry is chemistry that uses molecules rather than atoms as building blocks. Weak intermolecular forces, not covalent bonds, are used to assemble by design large structures from tailored molecules. 1,2 Since the pioneering work of Lehn, Cram, and Pedersen, 3 there has been a steadily increasing interest in the development and application of supramolecular chemistry. In its beginning, focus was on molecular recognition, which is the selective binding of a guest by a host using non-covalent interactions. 4 As the field grew, the rational design of molecular crystals using supramolecular interactions became an area of interest. 5-7 Most recently, the desire to understand the solution-solid interface 8-12 and the design of nanostructures by bottom-up methods, has driven the field of supramolecular chemistry from the three dimensional (3D) realm to the two dimensional (2D). The discovery and application of the scanning tunneling microscope has made this evolution to 2D supramolecular studies possible. The design strategies discovered for 3D supramolecular chemistry can be applied to the adsorbed state. Moreover, there is a particular relevance of the surface state to supramolecular synthesis with physisorbed molecules, because many of the weaker interactions used in generating synthons are probably distorted or destroyed in fluid solution and in chemisorption. The weak lateral forces exerted by the surface upon physisorbed molecules, and the image charges that occur in metal substrates, allow these weaker intermolecular forces to play a significant role in the formation of long-range order in the adsorbed phase. Scanning tunneling microscopy (STM) is the only technique that can provide detailed sub nanometer structural analysis at the solid solution interface in real time. Through its application, a view is provided of the elegant architectures that occur in what one might think was the relatively disordered interface between solid and solution. 8-12,14,17,18 STM has also been key to

3 characterizing 2D supramolecular structures resulting from vapor deposition on solid surfaces. 15,16,19,20 In the present study we will use STM to probe the interfacial layer that results when Au(111) comes into contact with a solution of coronene in heptanoic acid at various temperatures. While there have been many studies of the solution-solid interface, and more recently detailed inquiries about the role of the solvent, 18,21 they have been either at room temperature or below. The supposition has been that the molecular forces that drive the formation of these adlayers are weak and would not persist at elevated temperatures. Coronene itself is known to physisorb from the vapor phase on graphite in a close packed hexagonal structure with the coronene lying flat on the surface and having a lattice spacing of about 1.1 nm. 22 STM studies of coronene adsorbed on Au and Ag reported a coronene spacing of about 1.2 nm Gyarfas et al 18 used STM to study the solution-solid interface formed between Au(111) and dilute solutions of coronene in hexanoic, heptanoic, and octanoic acid. In all three cases adsorbed coronene is observed and lays flat on the metal surface. For the heptanoic and hexanoic cases, dipole-image dipole interactions and H bonding stabilize a hexagonal surface structure in which 12 acid molecules surround each coronene and produce a coronene spacing of 1.45 nm. In the case of octanoic acid as solvent, the incorporation of the solvent into the monolayer is not so favored. The coronene spacing can range from close packed (1.2 nm) with no solvent presumed present in the monolayer, to 1.50 nm with up to 12 solvent molecules possible. The close packed regions have hexagonal symmetry, as do those with the largest (1.5 nm) spacing. Heptanoic acid solutions give the clearest STM images and are associated with the most stable two-component monolayer, Figure 1. In the present work, we explore the temperature and concentration dependence of the goldcoronene-heptanoic acid system. We find that it is significantly more complex than might have

4 been expected. Moreover, this work suggests that temperature dependent STM studies (above room temperature) of the solution-solid interface in general are a worthwhile and productive activity. EXPERIMENTAL Heptanoic acid (C 7 O 2 H 14 ) 99%, was used as supplied by Sigma-Aldrich. Coronene was also provided by Aldrich and was labeled as sublimed and 99% purity. Epitaxial Au(111) films with well defined terraces and single atomic steps were prepared on mica by previously described methods These films were 0.1 to 0.2 m thick and had a mean single grain diameter of about 0.3 m. The gold films were hydrogen flame annealed just prior to use. A new piece of gold was used for each days experiments and for each temperature run. The STM head used was produced by Digital Instruments (now Veeco Metrology). A Digital Instruments Nanoscope E controller was used to acquire the reported data. STM image analysis was performed with the SPIP 30 commercial software package. Constant current images are reported and any filtering is indicated in the appropriate figure caption. All images were acquired with solution or solvent in contact with the gold surface and the tip penetrating the solution layer. Both etched and cut Pt 0.8 Ir 0.2 tips were used. In-plane spacing measured by STM was calibrated using the graphite lattice. In situ Heating: The heating stage for in situ heating was adapted from Sattan and Goh s design 31 using a Thermoptics DN505 subminiature proportionally controlled heater in the form of an integrated circuit. The temperature was controlled by a potentiometer and measured with a K type thermocouple as indicated in Figure 2. A Au(111) surface supported on mica was

5 exposed to a few microliters of coronene in heptanoic acid (5x10-5 M) and placed on the heater. Sample temperature was allowed to reach equilibrium (several minutes) and then STM images were obtained. Timed Submersion in Heated Solvent: The second method for heat treating the 1 cm 2 monolayer samples was to make a sample with 3 L of 1.5x10-4 M coronene in heptanoic acid, verify a complete monolayer by STM, rinse excess coronene off with about 10 L of pure heptanoic acid at room temperature, and then immerse the sample in 30 ml of continuously stirred and heated pure heptanoic acid for a given length of time. This sample was then removed from the solvent, excess solution was removed from the substrate by edge contact to a Kimwipe, and then re-imaged at room temperature. From these images, the percentage of the surface covered with an ordered film was determined. The sample was re-immersed in the heated heptanoic acid, cooled, and scanned again to monitor the progress of loss of the coronene on the surface. This process was repeated until individual coronene molecules were not observable on the surface. RESULTS AND DISCUSSION If the coronene in heptanoic acid solution is less than about 1.5x10-4 M, a complex coadsorption monolayer such as reported by Gyarfas 18 and displayed in Figure 1 is observed. At or above this concentration, we have found that a dense monolayer of molecular spacing near 1.2 nm is produced at room temperature. This spacing is consistent with coronene densely packed on Au(111) with the exclusion of heptanoic acid. 18,22-25 Figure 3A shows such a dense monolayer. If this surface is then washed with pure heptanoic acid (still at room temperature), the surface converts to the low density structure seen at lower concentrations heptanoic acid

6 molecules surrounding each coronene with a coronene spacing of about 1.45 nm, as shown in Figure 3B. If this same sample is then heated to 40 C, and measured at 40 C, one sees the same low density structure as seen at room temperature. This is shown in Figure 3C. Attempts to observe the solution-surface equilibrium at higher temperatures met with mixed success. Shown in Figure 4 are two images acquired at 55 C. Although there is considerable thermal drift, one can clearly observe the coronene adlayer and determine through sequential scanning that the spacing is the larger (1.45 nm) value associated with the low temperature, low concentration surface structure. However, as time progresses and the solvent evaporates, multilayer and crystallization structures begin to obscure the image (Figure 4B). Another problem associated with high temperature in situ STM measurements is the condensation of solvent onto the piezo and preamp. Even relatively low conductivity molecules when bridging the tip wire and the high voltage going to the piezos can produce large leakage currents. Because of these combined effects, we were not successful in acquiring quality images above 55 C, but it was apparent that adsorbed coronene existed well above 55 C. In order to track the effects of temperature significantly above 55 C, we attempted to wash (for varying times) the monolayer off the surface at various temperatures, and then cooled the solution-solid interface back to 21 C for STM analysis. Unlike the experiments described previously, it is important to note that the STM images are acquired from cooled samples. Details of the procedure are given in the experimental section. Considering that the volume of heated and stirred solvent is more than 10,000 times greater than the initial volume of solution in contact with the surface, it is surprising that any coronene is left on the surface after even a minute treatment above 60 C. As shown in Figure 5, some structured coronene persists above 100 C for more than 90 seconds! Before proceeding, we should note that the typical scan size

7 was 70 nm on a side, or 50 coronene molecules across. This means a typical image has about 2500 coronene molecules in it. This is far fewer molecules than one would normally use to determine thermodynamic values. As there is no way to re-image the same spot after taking the sample out, each time increment represents a different spot on the surface. In order to get some significance in the individual time measurements, multiple spots on the surface were imaged within the same time increment. Variations of up to 20% difference in coverage between two different locations on the same sample were not uncommon. Another aspect of this experimental data is that it is not completely clear that coronene molecules are lost from the surface, only that ordered regions become disordered. Thus, we will focus on tracking the extent of disorder which may include either loss or simple rapid diffusion. Figure 6 is representative of how this was done. The degree of molecular coverage of each sample at a particular time in the experimental sequence was determined by hand in Paint Shop Pro photo editing software. Regions of order/disorder were identified and the software gave a pixel count of each. Figure 6A and B show one of the many images before and after this editing process. Using the resulting fraction of ordered surface, a graph of fractional ordered monolayer versus time was made at each temperature studied (54, 67, 69, 76, 80,89,92, 97, and 102 C). Three typical graphs are shown in Figure 7. Initially there is a variable amount of time where no molecular depletion takes place. Thus, the particular 80 C sample in Figure 7 resisted thermal treatment longer than the particular sample measured at 89 C. This delay is not temperature specific, but rather sample specific. It is theorized that this is due to variations in the starting quality of the monolayer. Although great care is taken to ensure that the starting substrate is systematically cleaned by flame annealing, each Au(111) sample will have slightly different grain sizes, different degrees of reconstruction

8 due to hydrogen flame annealing, and different degrees of residual contamination. The monolayer will also have varying degrees of defects in the surface. Disorder is conjectured to start at individual defect sites in the monolayer and the grain boundaries of the islands and to continue until no molecular order can be seen. The smooth curves in Figure 7 are intended as a guide for the eye and have no special significance. Attempts to fit these curves to a delayed onset followed by a first order decay were not successful. We note that once the disordering process starts, it always occurs faster as the temperature increases. In the course of collecting the hundreds of STM images taken after various exposures, we began to notice the presence of additional ordered structures never previously seen. These new phases are never seen in the freshly deposited monolayers. The lower parts of Figure 8, for example, show the surface of a sample that had been heated for 5 minutes at 67 C. In the large image (lower right) one clearly sees islands having distinctly different structures. The "normal" hexagonal structured layer with coronene laying flat on the surface is adjacent to islands having a more rectangular structure. The image at the lower left of Figure 8 is an expanded view of one of these islands. In this image, the structural repeat distances are 1.55 nm and 0.82 nm. In other images, such as Figure 9, a repeat structure of 2.1 nm by 0.75 nm is observed. What are these new structures? A coronene molecule is approximately 0.95 nm in diameter and 0.3 nm thick. An unfolded heptanoic acid is approximately 0.95 nm long and 0.35 nm wide. The 0.75 to 0.82 nm spacing excludes the possibility that the coronene is laying flat on the surface. If no coronene were present then we might expect to see the heptanoic acid lie flat on the surface with the acid groups head to head as seen by Fang et al, 32 with the oxygens forming an apparent trough in the constant current image due to their electron withdrawing nature. Instead of a narrow trough between long

9 rows of high density, the image is much more complex. What seems most probable to us is that the coronene molecules are in alternating rows with heptanoic acid in between. A possible structure for the 2.1 nm x 0.75 nm spacing is given in Figure 9. This structure has the coronenes approximately perpendicular to the surface with a heptanoic acid dimmer pair spanning the distance between the rows, laying flat on the surface, and separating the coronenes. The extent to which the coronene molecules are upright is unclear and they may be significantly canted down towards the surface. In order to explain the 1.55 nm x 0.82 nm structure seen in Figure 8, we assumed that the alkyl chains were partially overlapping, as seen in the insert to that figure. Both of these proposed models have the advantage of predicting the correct height contrast. The high spots on the rows are the coronene molecules and the dark spots between the rows are where we predict the electron withdrawing carboxylate groups to be located. CONCLUSIONS The application of scanning tunneling microscopy to the solid solution interface at elevated temperatures was explored. With the particular solvent used, heptanoic acid, 55 C was the maximum temperature at which good quality images of the adlayer could be obtained. This maximum temperature is heavily influenced by the vapor pressure versus temperature curve of the solvent. Temperatures well above this limit (up to 105 C) were probed by a heat and sample technique. The coronene and heptanoic acid two-component adlayer was found to tenaciously adhere to the Au(111) surface in heated pure solvent. Lifetimes of the ordered layer in heated pure solvent were about one minute at 100 C and six minutes at 70 C. The mechanism of disruption of the monolayer appears to include a slow step where defects are formed. Depending on the quality of the initial sample, it can take several minutes in heated solvent before any change in the adlayer is observed.

10 An unexpected outcome of this study was the observation of several different surface structures. In our preliminary room temperature study at low concentration of coronene, only the 1:6 coronene:heptanoic acid (1 coronene surrounded by 12 acid molecules) structure was observed in heptanoic acid. In this study we find that at least three other structures are present, two of which are composed of both coronene and heptanoic acid. Our preliminary analysis of the structure suggests that these have a 1:2 coronene:heptanoic acid composition. We expect that an even richer zoo of structures will be disclosed by heating the substrate in a coronene containing solution, rather than the pure solvent used here. This work will hopefully encourage others to begin to explore the surface-solution interface at elevated temperatures using STM. Clearly there are systems where such studies can yield new insights and guide in the formation of new interface structures. Acknowledgements This material is based upon work supported by the National Science Foundation under Grant CHE We gratefully acknowledge their support. This work, in part, was submitted towards completion of the doctoral degree.

11 REFERENCES 1 J. M. Lehn, Supramolecular Chemistry: Concept and Perspectives, VCH, Weinheim, Lehn, Jean-Marie PNAS 2002, 99, Noble Prize in Chemistry. Nobel Lectures: Chemistry , Editor-in-Charge Tore Frängsmyr, Editor Bo G. Malmström, World Scientific Publishing Co., Singapore, Gale, P. A. Phil. Trans. R. Soc. Lond. A 2000, 358, Reddy, D. S.; Craig, D. C.; Desiraju, G. M. J. Am. Chem. Soc. 1996, 118, Gillard, R.; Stoddard, J.; White, A.; Williams, B.; Williams, D.; J. Org. Chem. 1996, 61, Desiraju, G. R.; Angew. Chem. Int. Ed. Engl, 1995, 34, De Feyter, S.; De Schryver, F. C.; Chem. Soc. Rev. 2003, 32, Yablon, D. G.; Giancarlo, L. C.; Flynn, G. W.; J. Phys. Chem. B 2000, 104, Yablon, D.; Guo, J.; Knapp, D.; Fang, H.; Flynn, G.W. J. Phys. Chem. B 2001, 105, Cai, Y.; Bernasek, S. L. J. Am. Chem. Soc. 2003, 125, Yoshimoto, S.; Suto, K.; Itaya, K.; Kobayashi, N. Chemical Communications, 2003, Whitesides, G. M.; Boncheva, M. PNAS 2002, 99, Griessl, S. J. H.; Lackinger, M.; Jamitzky, F.; Markert, T.; Hietschold, M.; Heckl, W. M.. J. Phys. Chem. B, 2004, 108, Hipps, K. W.; Scudiero, L.; Barlow, D. E.; Cooke, M. P. J. Amer. Chem. Soc. 2002, 124, Scudiero, L.; Hipps, K. W.; Barlow, D. E. J. Phys. Chem. B 2003, Lei, S.; Wang, C.; Wan, L.; Bai, C. J. Phys. Chem. B 2004, 108, Gyarfas, B.; Wiggins, B.; Zosel, M.;. Hipps, K. W. Langmuir 2005, 21, 919.

12 19 Griessl, S.; Lackinger, M.; Edelwirth, M.; Hietschold, M.; Heckl, W. M. Single Mol. 2002, 3, Barth, J.V.; Weckesser, J.; Lin, N.; Dmitriev, A.; Kern, K. Appl. Phys. A 2003, 76, Kampschulte, L.; Maie, A.; Kishore, R.; Schmittel, M.; Heckl, W.; Lackinger, M. J. Phys. Chem 2006, 110, a) Lackinger, M.; Griessl, S.; Heckl, W. M.; Hietschold, M. Anal Bioanal Chem., 2002, 374, 685; b) Walzer, K.; Sternberg, M.; Hietschold, M. Surf. Sci., 1998, 415, Uemura, S.; Sakata, M.; Taniguchi, I.; Hirayama, C.; Kunitake, M. Thin Solid Films, 2002, 409, Yoshimoto, S.; Narita, R.; Itaya, K. Chemistry Letters 2002, Lackinger, M.; Griessl, S.; Heckl, W. M.; Hietschold, M. J. Phys. Chem. B 2002, 106, Lu, X.; Hipps, K. W.; J Phys. Chem. B 1997, 101, Lu, X.; Hipps, K. W.; Wang, X.; Mazur, U. J. Amer. Chem. Soc. 1996, 118, Barlow, D.; Hipps, K. W. J. Phys. Chem B 2000, 104, Scudiero, L.; Barlow, D.; Hipps, K. W. J. Phys. Chem B 2000, 104, SPIP software, Version 4.1. Published by Image Metrology A/S, Hørsholm, Denmark. 31 Sattin, Bernie D.; Goh, M. Cynthia. Rev. Sci. Instrum. 2004, 75, Fang, Hongbin; Giancarlo, Leanna C.; Flynn, George W. J. Phy. Chem. B 1998, 102, 7421.

13 FIGURE CAPTIONS Figure 1: High resolution constant current STM image of the self-assembled structure having 12 heptanoic acid molecules surround each coronene on Au(111) at 19 C. Inserts are a CPK model depicting the packing structure and a correlation average from the STM image. Figure 2: Schematic of sample heater and heating setup. Block diagram of a commercial Thermoptics DN505 subminiature proportionally controlled heater is shown on left. Figure 3: Constant current STM images (50 pa and -700 mv) of the Au(111)/solution interface. A) Dense coronene structure (1.2 nm spacing) seen at 21 C with 1.5x10-4 M coronene in heptanoic acid solution. B) Low coronene density structure due to coadsorbed heptanoic acid seen at 21 C after washing sample A in pure heptanoic acid. C) The low density structure (1.45 nm coronene spacing) persists at 40 C. Figure 4: Constant current STM image of the Au(111) interface with a 5x10-5 M solution of coronene in heptanoic acid at 55 C. Image A was taken shortly after the heater stabilized at 55 C, while B was acquired about 3 minutes later. Settings were -900 mv and 200 pa. Figure 5: Constant current STM image (-700mV, 50 pa) of a monolayer of coronene and heptanoic acid formed on Au(111) and acquired at 21 C after immersion in pure heptanoic acid at 102 C. A) 30 second immersion, B) 90 second immersion.

14 Figure 6: Sample of coronene and heptanoic acid that has been submerged in pure 92 C heptanoic acid for 60 seconds. Left image is as taken, right image has had the regions of disorder shaded. Figure 7. Figure 7. Fraction of ordered monolayer remaining after the indicated time in pure solvent at the indicated temperature. Note that, as discussed in the text, the initial number of defects in the films varied from one experiment to the next. Figure 8: Possible surface structure with approximately 1.55 nm x 0.82 nm spacing. Sample was heated in heptanoic acid for 5 minutes at 67 C. The scan size on the right is 75nm x 75nm. Left image is a zoomed image of the box in right image. Figure 9. Constant current STM image (21 C) of initially monolayer sample after heating in heptanoic acid for 4 minutes at 67 C. Cartoon of a possible structure of the reconstructed areas of the monolayer with 2.1 nm x 0.75 nm spacing is shown by CPK models inserted in the figure. Note that the coronene molecules are probably tilted rather than completely vertical.

15 Figure 1: High resolution constant current STM image of the self-assembled structure having 12 heptanoic acid molecules surround each coronene on Au(111) at 19 C. Inserts are a CPK model depicting the packing structure and a correlation average from the STM image.

16 Figure 2: Schematic of sample heater and heating setup. Block diagram of a commercial Thermoptics DN505 subminiature proportionally controlled heater is shown on left.

17 Figure 3: Constant current STM images (50 pa and -700 mv) of the Au(111)/solution interface. A) Dense coronene structure (1.2 nm spacing) seen at 21 C with 1.5x10-4 M coronene in heptanoic acid solution. B) Low coronene density structure (1.45 nm spacing) due to coadsorbed heptanoic acid seen at 21 C after washing sample A in pure heptanoic acid. C) The low density structure (1.45 nm coronene spacing) persists at 40 C.

18 Figure 4: Constant current STM image of the Au(111) interface with a 5x10-5 M solution of coronene in heptanoic acid at 55 C. Image A was taken shortly after the heater stabilized at 55 C, while B was acquired about 3 minutes later. Settings were -900mV and 200 pa.

19 Figure 5: Constant current STM image (-700mV, 50 pa) of a monolayer of coronene and heptanoic acid formed on Au(111) and acquired at 21 C after immersion in pure heptanoic acid at 102 C. A) 30 second immersion, B) 90 second immersion.

20 Figure 6: Sample of coronene and heptanoic acid that has been submerged in pure 92 C heptanoic acid for 60 seconds. Left image is as taken, right image has had the regions of disorder shaded.

21 1 Fraction Covered C 92 C 80 C Time (seconds) Figure 7. Fraction of ordered monolayer remaining after the indicated time in pure solvent at the indicated temperature. Note that, as discussed in the text, the initial number of defects in the films varied from one experiment to the next.

22 Figure 8: Possible surface structure with approximately 1.55 nm x 0.82 nm spacing. Sample was heated in heptanoic for 5 minutes at 67 C. The scan size on the right is 75nm x 75nm. Left image is a zoomed image of the box in right image.

23 Figure 9. Constant current STM image (21 C) of initially monolayer sample after heating in heptanoic for 4 minutes at 67 C. Cartoon of a possible structure of the reconstructed areas of the monolayer with 2.1 nm x 0.75 nm spacing is shown by CPK models inserted in the figure. Note that the coronene molecules are probably tilted rather than completely vertical.