Langmuir 2004, 20, 2803-2806
2803
On the Existence of Ordered Organic Adlayers at the Cu(111)/Electrolyte Interface Daniel Friebel,* Thomas Mangen, Britta Obliers, Christian Schlaup, Peter Broekmann, and Klaus Wandelt Institut fu¨ r Physikalische und Theoretische Chemie, Wegelerstrasse 12, D-53115 Bonn, Germany Received November 12, 2003. In Final Form: December 12, 2003 We have reinvestigated the behavior of a Cu(111) electrode in pure and cinchonidine containing aqueous 0.1 M HClO4 solution by cyclic voltammetry (CV) and in situ electrochemical scanning tunneling microscopy (STM). In contrast to previous publications by Wan et al. (Langmuir 2000, 19, 1958-1962 and references cited therein) on Cu(111) in pure 0.1 M HClO4 which claimed an adsorbate-free Cu(111) surface in the entire potential range, we have found a highly ordered hexagonal adsorbate structure with a (4 × 4) unit cell, which is stable in the potential range from hydrogen evolution at -350 to -150 mV (RHE). The adsorbate-free (1 × 1) Cu(111) surface is only visible in a fairly small potential range from -150 to +50 mV. A disordered surface structure is formed at more positive potentials which is interpreted by adsorption of an oxygen-containing species. Furthermore, the formation of a highly ordered cinchonidine adlayer on Cu(111) in 0.1 M HClO4 as reported by Wan et al. (J. Am. Chem. Soc. 2002, 124, 14300-14301) could not be reproduced here. In fact, the similarity of all structures reported by Wan et al. for a great variety of different organic adlayers on Cu(111) in HClO4 solution including cinchonidine with the (4 × 4) superstructure found here already in pure HClO4 solution (i.e., without organic solute) casts serious doubts on the validity of those previous results by Wan et al. in general.
Introduction Organic thin films are of special interest due to a wide variety of possible applications including organic electronic devices, catalysis, and corrosion inhibition. A review of approaches which have been successful to prepare wellordered organic films on metal surfaces under electrochemical conditions has been given by Dretschkow and Wandlowski.1 However, the controlled formation of organic films at the metal/electrolyte interface requires, as a basis, a detailed knowledge of the influence of the supporting electrolyte itself on the metal surface. Adsorption of a vast number of organic compounds2-14 ranging from simple arenes2 to crown ethers3 and, more * To whom correspondence may be addressed. Tel: +49 228 732572. Fax: +49 228 732551. E-mail:
[email protected]. (1) Dretschkow, Th.; Wandlowski, Th. In Solid-Liquid Interfaces; Wandelt, K., Thurgate, S., Eds.; Topics of Applied Physics 85; SpringerVerlag: Berlin Heidelberg, 2003; pp 259-321. (2) Wan, L.-J.; Itaya, K. Langmuir 1997, 13, 7173-7179. (3) Wang, D.; Xu, Q.-M.; Wan, L.-J.; Wang, C.; Bai, C.-L. Surf. Sci. 2001, 489, L568-L572. (4) Han, M.-J.; Xu, Q.-M.; Wan, L.-J.; Bai, C.-L. J. Phys. Chem. B 2002, 106, 11272-11276. (5) Wei, G.-X.; Pan, G.-B.; Wan, L.-J.; Zhao, J.-C.; Bai, C.-L. Surf. Sci. 2002, 520, L625-L632. (6) Wang, D.; Xu, Q.-M.; Wan, L.-J.; Wang, C.; Bai, C.-L. Langmuir 2002, 18, 5133-5138. (7) Wang, D.; Wan, L.-J.; Xu, Q.-M.; Wang, C.; Bai, C.-L. Surf. Sci. 2001, 478, L320-L326. (8) Xu, Q.-M.; Zhang, B.; Wan, L.-J.; Wang, C.; Bai, C.-L.; Zhu, D.-B. Surf. Sci. 2002, 517, 52-58. (9) Xu, Q.-M.; Wang, D.; Wan, L.-J.; Wang, C.; Bai, C.-L.; Feng, G.Q.; Wang, M.-X. Angew. Chem., Int. Ed. 2002, 41, 3408-3411. Angew. Chem. 2002, 114, 3558-3561. (10) Xu, Q.-M.; Wang, D.; Wan, L.-J.; Bai, C.-L.; Wang, Y. J. Am. Chem. Soc. 2002, 124, 14300-14301. (11) Yin, X.-L.; Wan, L.-J.; Yang, Z.-Y.; Yu, J.-Y.; Bai, C.-L. Surf. Sci. 2003, 531, 226-230. (12) Wang, D.; Xu, Q.-M.; Wan, L.-J.; Bai, C.-L.; Jin, G. Langmuir 2003, 19, 1958-1962. (13) Han, M.-J.; Zeng, Q.-D.; Wan, L.-J.; Bai, C.-L. Chem. Lett. 2003, 32, 702-703. (14) Wang, D.; Lei, S.-B.; Wan, L.-J.; Bai, C.-L. J. Phys. Chem. B 2003, 107, 8474-8478.
recently, even chiral compounds9,10,12-14 on Cu(111) in 0.1 M HClO4 has been studied by scanning tunneling microscopy (STM) by Wan et al.2-14 According to their results, all these compounds form highly ordered adlayers on the Cu(111) surface with lattice constants around 1 nm at negative potentials with respect to the reversible hydrogen electrode (RHE). In the absence of the organic compounds, the presence of the (1 × 1)-structure of the adsorbate-free Cu(111) surface is reported for the entire potential range. In contrast to these results, our own STM measurements on Cu(111) in pure 0.1 M HClO4 reveal a highly ordered (4 × 4) superstructure in the potential range between -350 and -150 mV vs RHE. Furthermore, cyclic voltammograms (CVs) of Cu(111) in 0.1 M HClO4 shown by Wan et al.2 differ significantly from respective CVs in the literature15,16 and from our own experimental data. Given all these differences from the previous publications,2-14 we felt a need to reinvestigate the adsorption of organic molecules under these conditions. As an example, we show and discuss here STM results for the system Cu(111)/0.1 M HClO4 + 0.01 mM cinchonidine. Experimental Section Our measurements were carried out in a home-built scanning tunneling microscope17 under an inert atmosphere of purified argon. STM and CV measurements are performed in the same electrochemical cell with a large cell volume of 2.3 cm3. The tunneling tips were electrochemically etched from a 0.25 mm diameter tungsten wire in a thin lamella of 2 M KOH and insulated with nail polish. The HClO4 and cinchonidine solutions were prepared from 70% HClO4 (suprapur, Merck), cinchonidine (Merck) and ultrapure water (Millipore). Even suprapure HClO4 contains trace amounts of chloride which might influence experimental data. However, we can exclude every known (15) Brisard, G. M.; Zenati, E.; Gasteiger, H. A.; Markovic´, N. M.; Ross, P. N. Langmuir 1995, 11, 2221-2230. (16) Chu, Y.-S.; Robinson, I. K. (advisor) Ph.D. dissertation, University of Illinois, 1997. (17) Wilms, M.; Kruft, M.; Bermes, G.; Wandelt, K. Rev. Sci. Instrum. 1999, 70, 3641-3650.
10.1021/la036130d CCC: $27.50 © 2004 American Chemical Society Published on Web 02/28/2004
2804
Langmuir, Vol. 20, No. 7, 2004
Friebel et al.
Figure 1. Cyclic voltammogram of Cu(111) in 0.1 M HClO4 dE/dt ) 10 mV/s: H, hydrogen evolution; O/O′, oxygencontaining species ad-/desorption; D, anodic copper dissolution; X, unknown. influence15,18 by Cl- contamination on our CV as well as on our STM data. A commercial Cu(111) single crystal (MaTeck GmbH, Germany) was prepared prior to each experiment by electropolishing in 50% H3PO4 at a voltage of +2 V. All electrochemical potentials E are given with respect to the RHE.
Results and Discussion A typical CV of a Cu(111) electrode in 0.1 M HClO4 (Figure 1) at a scan rate of 10 mV/s shows close to the potential region of anodic copper dissolution (D) at +350 mV a pair of peaks (O/O′) at +250 and +150 mV indicating the formation of a passivating oxygen-containing adlayer. Kinetic hindrance of the adsorption/desorption of this adlayer leads to a considerable peak to peak separation of 100 mV. Moreover, no sharp copper redeposition peak can be found during the negative potential sweep. Only after sweeping the potential to higher potentials than +400 mV and a very high copper dissolution current, redeposition can be discerned by a broad bump. This can be explained by the kinetic hindrance of the dissolution as well as the redeposition of copper due to the oxygencontaining adlayer. At negative potentials close to hydrogen evolution (H), another cathodic current peak (X) appears at -280 mV. The origin of this peak which does not correspond to any anodic current peak is still unclear. In agreement with our findings, Brisard et al.15 and Chu16 published very similar CV curves for Cu(111)/HClO4 with a peak system related to the adsorption of an oxygencontaining species. In strong contrast to these results, the CV of Cu(111) in 0.1 M HClO4 shown by Wan et al.2 exhibits only a featureless double layer region between hydrogen evolution and anodic copper dissolution. We were able to reproduce such a CV only by restricting the potential scan range between -300 and +300 mV, while scanning over the entire potential range leads to the CV shown in Figure 1. From our cyclic voltammogram, we conclude that Cu(111) exposed to 0.1 M HClO4 cannot be considered as an inert metal/electrolyte interface as claimed in the works of Wan et al.,2-14 but instead, a complex potential dependent phase behavior seems to occur. On the basis (18) Inukai, J.; Osawa, Y.; Itaya, K. J. Phys. Chem. B 1998, 102, 10034.
Figure 2. STM images of the (4 × 4) structure in pure 0.1 M HClO4 at different tunneling parameters, 6.2 nm × 6.2 nm: (a) E ) -150 mV, UB ) 50 mV, IT ) 20 nA; (b) E ) -325 mV, UB ) 5 mV, IT ) 50 nA; (c) E ) -250 mV, UB ) 100 mV, IT ) 10 nA; (d) E ) -250 mV, UB ) 40 mV, IT ) 40 nA.
of our results and those of Brisard et al.15 and Chu16 we therefore conclude that the CV measurement shown by Wan et al. is not completely representative for the system Cu(111)/HClO4. The complex potential dependent phase behavior of Cu(111) in 0.1 M HClO4 is clearly supported by our STM measurements. STM images of Cu(111) in pure 0.1 M HClO4 solution show in the potential range between massive hydrogen evolution at -350 and -150 mV a long-range hexagonal structure covering the entire Cu(111) surface which can be described by a (4 × 4) unit cell which is not rotated with respect to the unit cell of the bare Cu(111) substrate. The lattice parameters of the superstructure could be clearly determined in several experiments using the (1 × 1)Cu(111) structure for internal calibration. The (4 × 4) structure shows as a characteristic feature a strong dependence of its imaging properties on the tunneling current IT and the bias voltage UB. By careful variation of the tunneling parameters, many different substructural features of this (4 × 4) structure in pure HClO4 solution can be imaged (Figure 2), all of which originate only from contributions of different electronic states of the observed superstructure to the tunneling process. The same (4 × 4) structure has also been observed on Cu(111) in less acidic and neutral electrolytes as well as in the presence of many different anions and cations, namely, dilute H2SO4,19 HBr, HI, and K2SO4.20 Therefore, we propose this superstructure to originate from a solvent species such as H3O+, OH-, or OH species. A pure water adlayer (“two-dimensional ice”) can be easily ruled out on structural considerations and in view of the weak interaction of water molecules with metal surfaces.21 Only at potentials between -150 and +50 mV, the adsorbate-free (1 × 1) Cu(111) structure can be imaged in HClO4 solution with atomic resolution (Figure 3a). (19) Broekmann, P.; Wilms, M.; Wandelt, K. Surf. Rev. Lett. 1999, 6, 907-916. (20) Friebel, D.; Broekmann, P.; Wandelt K. In preparation, 2004. (21) Henderson, M. A. Surf. Sci. Rep. 2002, 46, 1-308.
Ordered Organic Adlayers
Langmuir, Vol. 20, No. 7, 2004 2805
Figure 3. (a) STM image of the (1 × 1) Cu(111) structure under pure 0.1 M HClO4, 4.3 nm × 4.3 nm, E ) -160 mV, UB ) 4 mV, IT ) 50 nA. (b) Surface topography at more positive potentials, 40.6 nm × 40.6 nm, E ) +50 mV, UB ) 100 mV, IT ) 1 nA.
At higher potentials, the surface topography becomes more complex with the formation of furrows (Figure 3b), and atomic resolution can no longer be achieved. Instead, a disordered structure forms which originates from the adsorption of an oxygen species (most likely hydroxide). This result is in good agreement with an in situ X-ray diffraction study of Cu(111) in 0.1 M HClO4 by Chu.16 In this work, oxygen or hydroxide was reported to form a laterally disordered adlayer by adsorption on face-centered cubic and hexagonal close packed 3-fold hollow sites of the Cu(111) surface at potentials close to anodic copper dissolution. A more detailed discussion of the phase behavior of Cu(111) in dilute HClO4 at positive potentials will be given in a forthcoming publication.22 At the onset of anodic copper dissolution at +250 mV, STM imaging becomes impossible because the tungsten tip deteriorates. After the characterization of Cu(111) in pure 0.1 M HClO4, the electrolyte was exchanged for a solution of 0.1 M HClO4 + 0.01 mM cinchonidine at an electrode potential of -200 mV and re-examined with STM. For the entire potential range from -350 to +250 mV, exactly the same surface phases were found as in the absence of cinchonidine. To prevent that eventually, adsorbed cinchonidine molecules could have been swept away by the tip, we also recorded STM images with very low tunneling currents between 0.1 and 0.5 nA and moderate to high bias voltages from 50 to 300 mV. However, under none of these conditions did we find any evidence pointing to the formation of an ordered cinchonidine adlayer. The only adlayer structure at potentials lower than -150 mV could be clearly identified as the (4 × 4) structure, which was already observed in pure 0.1 M HClO4 (Figure 4). Wo¨ll et al.23 have extensively studied the adsorption of arenes on Cu(111) under ultrahigh vacuum conditions by thermal desorption spectrometry (TDS) and low-energy electron diffraction (LEED). Their work shows that arenes are only weakly adsorbed on Cu(111) and tend to form disordered two-dimensional gas phases. A preferential adsorption occurs at defect sites and on upper step edges. The latter is explained by the stronger attraction between the enhanced density of empty electronic states on the upper step edges (Smoluchowski effect) and the nucleophilic arenes. While no ordered films are observed on Cu(111), well-ordered films are formed on its vicinal Cu(221) and Cu(443) surfaces.23 In contrast, Wan et al. have reported highly ordered films of benzene, naphtha(22) Obliers, B.; Friebel, D.; Mangen, T.; Broekmann, P.; Wandelt, K. In preparation, 2004. (23) Lukas, S.; Vollmer, S.; Witte, G.; Wo¨ll, Ch. J. Chem. Phys. 2001, 114, 10123-10130.
Figure 4. STM images of the (4 × 4) structure in the presence of cinchonidine at different tunneling parameters, 6.2 nm × 6.2 nm: (a) E ) -300 mV, UB ) 80 mV, IT ) 20 nA; (b) E ) -300 mV, UB ) 60 mV, IT ) 20 nA; (c) E ) -300 mV, UB ) 40 mV, IT ) 50 nA; (d) E ) -350 mV, UB ) 9 mV, IT ) 50 nA.
lene, and anthracene on Cu(111) in electrochemical environment,2 and these arenes have been postulated to adsorb preferentially at potentials close to hydrogen evolution, while desorption is claimed to occur at potentials higher than -150 mV (benzene, naphthalene) or 0 mV (anthracene). However, by contrast, following the interpretation of Wo¨ll et al.,23 due to the nucleophilic character of these arenes, adsorption, if any, should be expected at much higher potentials and cathodic desorption should occur at these low potentials where ordered arene adlayers where claimed to be stable by Wan et al. Several other reports by Wan et al.3-14 on organic adlayers have been based on their questionable assumption that arenes adsorb at potentials close to hydrogen evolution and on their incorrect assumption that the adsorbate-free (1 × 1) Cu(111) structure is present in the entire potential range. The incorrect arene adsorption model was extended and generalized for a series of organic compounds including prominent examples for chirality,9,10,12-14 natural products such as amino acids12,14 and alkaloids,10 crown ethers,3 and magnetic molecules,8 all of which have been stated to form identical adlayer structures with (4 × 4) unit cells or extremely similar structures regardless of strongly diverging physical and chemical properties, van der Waals radii, symmetry, and even ionic charge8,11 of the organic species studied. There is no obvious reason that all these different molecules show this uniform behavior, and for most of the organic species studied by Wan et al., the assumed adsorbateadsorbate distance of 4aCu ) 1.02 nm is much too high in comparison to their van der Waals radii. Moreover, most of these molecules have a reduced symmetry with respect to the hexagonal Cu(111) surface and should therefore form different rotational or mirror domains, but no such domains have ever been reported by Wan et al. Furthermore, Wan et al. reported to have imaged these organic adlayers by STM at fairly high tunneling currents of 10-20 nA.2-14 From our own successful STM measure-
2806
Langmuir, Vol. 20, No. 7, 2004
ments on weakly adsorbed organic adlayers,24 however, we suspect that these adsorbates will be swept away by the tip if such high tunneling currents are applied. Summarizing, we have shown the complex behavior of Cu(111) in pure 0.1 M HClO4. At negative potentials, a (4 × 4) structure is observed which is a general feature of Cu(111) in aqueous solutions. At positive potentials close to anodic copper dissolution, a disordered adlayer is formed by the interaction of the Cu surface with an oxygencontaining species. Only in a small potential range in the double layer region, the bare (1 × 1) Cu(111) surface can be observed. From the comparison of our own STM data with the results of Wan et al., we conclude that none of the adlayer structures reported in refs 2-14 is due to the (24) Safarowsky, C.; Broekmann, P.; Wandelt, K. In preparation, 2004.
Friebel et al.
presence of adsorbed organic molecules because the structures described by Wan et al. are identical to the “inorganic” (4 × 4) structure which, according to our measurements, exists already in pure 0.1 M HClO4 solution. All the STM images which were reported to show internal structures of the organic compounds2-14 can be reproduced with this “inorganic” (4 × 4) structure, i.e., in the absence of these compounds by just varying the tunneling conditions, i.e., the tunneling current and the bias voltage. Moreover, our measurements show that even in the presence of cinchonidine in the HClO4 solution, the “inorganic” (4 × 4) structure is not displaced. Therefore we express our serious doubts that any of the STM images shown by Wan et al. in refs 2-14 correspond to the claimed organic adlayers. LA036130D
