Understanding the Contrast Mechanism in Scanning Tunneling

Langmuir 2008, 24, 1897-1901

1897

Understanding the Contrast Mechanism in Scanning Tunneling Microscopy (STM) Images of an Intermixed Tetraphenylporphyrin Layer on Ag(111) Karmen Comanici, Florian Buchner, Ken Flechtner, Thomas Lukasczyk, J. Michael Gottfried,* Hans-Peter Steinru¨ck, and Hubertus Marbach* Lehrstuhl fu¨r Physikalische Chemie II and Interdisciplinary Center for Molecular Materials (ICMM), UniVersita¨t Erlangen-Nu¨rnberg, Egerlandstraβe 3, D-91058 Erlangen, Germany ReceiVed October 19, 2007. In Final Form: NoVember 22, 2007 W This paper contains enhanced objects available on the Internet at http://pubs.acs.org/journals/langd5. The appearance of tetraphenylporphyrins in scanning tunneling micrographs depends strongly on the applied bias voltage. Here, we report the observation and identification of certain features in scanning tunneling microscopy (STM) images of intermixed layers of tetraphenylporphyrin (2HTPP) and cobalt-tetraphenylporphyrin (CoTPP) on Ag(111). A significant fraction of an ordered monolayer of commercially available CoTPP appears as “pits” at negative bias voltages around -1 V. The obvious possibility that these pits are missing molecules within the ordered layer could be ruled out by imaging the molecules at reduced bias voltages, at which the contrast of the pits fades, and at positive bias voltages around +1 V, at which the image contrast is inverted. With the investigation of the electronic structure, in particular the density of states (DOS) close to the Fermi level, of CoTPP and 2HTPP layers by means of ultraviolet photoelectron spectroscopy (UPS) and scanning tunneling spectroscopy (STS), the contrast mechanism was clarified. The correlation of the bias dependent contrast with the UPS data enabled us to interpret the “pits” as individual 2HTPP molecules. Additional evidence could be provided by imaging layers of different mixtures of 2HTPP and CoTPP and by high-resolution STM imaging of the features in CoTPP.

Introduction Studying porphyrins on surfaces is of general interest for manifold reasons. From a general standpoint, porphyrins show a large variety of interesting functional properties, which are mostly based on the redox activity of the central metal ion. Prominent examples are metalloporphyrin derivatives in hemoglobin (iron porphyrin) and chlorophyll (magnesium porphyrin), which are responsible for the functionality (oxygen transport in hemoglobin and photosynthesis in chlorophyll, respectively) of these important biological molecules. Therefore, porphyrin derivatives adsorbed on surfaces can be regarded as model systems to study their functionality in biological molecules in detail. Numerous studies demonstrate that monolayers or submonolayers of porphyrins tend to arrange themselves in well-defined long range ordered lateral structures.1-19 The combination of both * To whom correspondence should be addressed. E-mail: [email protected] (J.M.G.); hubertus.marbach@ chemie.uni-erlangen.de (H.M.). (1) Jung, T. A.; Schlittler, R. R.; Gimzewski, J. K. Nature 1997, 386, 696698. (2) Yoshimoto, S.; Tada, A.; Suto, K.; Narita, R.; Itaya, K. Langmuir 2003, 19, 672-677. (3) Barlow, D. E.; Scudiero, L.; Hipps, K. W. Langmuir 2004, 20, 44134421. (4) Shubina, T. E.; Marbach, H.; Flechtner, K.; Kretschmann, A.; Jux, N.; Buchner, F.; Steinru¨ck, H.-P.; Clark, T.; Gottfried, J. M. J. Am. Chem. Soc. 2007, 129, 9476-9483. (5) Scudiero, L.; Barlow, D. E.; Mazur, U.; Hipps, K. W. J. Am. Chem. Soc. 2001, 123, 4073-4080. (6) Scudiero, L.; Barlow, D. E.; Hipps, K. W. J. Phys. Chem. B 2000, 104, 11899-11905. (7) Yanagi, H.; Mukai, H.; Ikuta, K.; Shibutani, T.; Kamikado, T.; Yokoyama, S.; Mashiko, S. Nano Lett. 2002, 2, 601-604. (8) Buchner, F.; Comanici, K.; Jux, N.; Steinru¨ck, H.-P.; Marbach, H. J. Phys. Chem. C 2007, 111, 13531-13538. (9) Terui, T.; Sekiguchi, T.; Wakayama, Y.; Kamikado, T.; Mashiko, S. Thin Solid Films 2004, 464-465, 384-387.

functionality and structure makes them ideal candidates for molecular devices. Scanning tunneling microscopy (STM) proved to be a powerful technique to characterize this long range order20,21 and even the internal conformation of different porphyrin derivatives.1,8,22 It is mostly conducted on metal surfaces either under ultrahigh vacuum (UHV) conditions or in solution.2,19 The investigation of multicomponent porphyrin layers (e.g., NiTPP and CoTPP6) and mixed films of phthalocyanines (Pc) and porphyrins (e.g., CoPc and NiTPP,17 CoPc and CoTPP3) by means of STM is a topic of special interest. Furthermore, an important step toward (10) Terui, T.; Yokoyama, S.; Suzuki, H.; Mashiko, S.; Sakurai, M.; Moriwaki, T. Thin Solid Films 2006, 499, 157-160. (11) Grill, L.; Stass, I.; Rieder, K. H.; Moresco, F. Surf. Sci. 2006, 600, L143L147. (12) Terui, T.; Kamikado, T.; Okuno, Y.; Suzuki, H.; Mashiko, S. Curr. Appl. Phys. (AMN-1, First International Conference on Advanced Materials and Nanotechnology) 2004, 4, 148-151. (13) Auwa¨rter, W.; Weber-Bargioni, A.; Brink, S.; Riemann, A.; Schiffrin, A.; Ruben, M.; Barth, J. V. ChemPhysChem. 2007, 8, 250-254. (14) Buchner, F.; Schwald, V.; Comanici, K.; Steinru¨ck, H.-P.; Marbach, H. ChemPhysChem. 2007, 8, 241-243. (15) Bonifazi, D.; Spillmann, H.; Kiebele, A.; de Wild, M.; Seiler, P.; Cheng, F. Y.; Gu¨ntherodt, H. J.; Jung, T.; Diederich, F. Angew. Chem., Int. Ed. 2004, 43, 4759-4763. (16) Arima, V.; Fabiano, E.; Blyth, R. I. R.; Delia Sala, F.; Matino, F.; Thompson, J.; Cingolani, R.; Rinaldi, R. J. Am. Chem. Soc. 2004, 126, 1695116958. (17) Scudiero, L.; Hipps, K. W.; Barlow, D. E. J. Phys. Chem. B 2003, 107, 2903-2909. (18) Scudiero, L.; Barlow, D. E.; Hipps, K. W. J. Phys. Chem. B 2002, 106, 996-1003. (19) Yoshimoto, S.; Tsutsumi, E.; Suto, K.; Honda, Y.; Itaya, K. Chem. Phys. (Molecular Charge Transfer in Condensed Media - from Physics and Chemistry to Biology and Nanoengineering in honour of Alexander M. Kuznetsov on his 65th birthday) 2005, 319, 147-158. (20) Gimzewski, J. K.; Joachim, C. Science 1999, 283, 1683-1688. (21) De Feyter, S.; De Schryver, F. C. Chem. Soc. ReV. 2003, 32, 139-150. (22) Moresco, F.; Meyer, G.; Rieder, K.-H.; Ping, J.; Tang, H.; Joachim, C. Surf. Sci. 2002, 499, 94-102.

10.1021/la703263e CCC: $40.75 © 2008 American Chemical Society Published on Web 01/25/2008

1898 Langmuir, Vol. 24, No. 5, 2008

the engineering of porphyrins was achieved with the in situ metalation of 2H-porphyrins by vapor-deposition of different metals (Fe,13,14 Co,4 and Zn4,23,24), which also included the investigation of mixed layers of 2H-porphyrins and metalloporphyrins by STM.4,13,14 In these studies, discrimination and identification of the corresponding molecules was possible on the basis of the appearance of submolecular features.3,4,6,13,14,17 Generally, at negative sample bias, 2H-porphyrins appear as features with a central cavity in high-resolution STM, whereas metalloporphyrins exhibit a central protrusion caused by the central metal ion. In binary mixtures of molecules with different central metal ions (different M(II)TPPs, M(II)Pc and M(II)TPP), CoPc17 or CoTPP6 was found to appear particularly bright at negative bias voltages, due to an enhanced tunneling contribution through the half filled dz2 orbital of the Co ion. Evidence for a bias dependent appearance of single Fe-tetrapyridylporphyrin (FeTPyP) molecules has also been found in a study by Auwa¨rter et al.,13 where a contrast inversion of FeTPyP (protrusion at negative bias, depression at positive bias voltage) within a layer of 2HTPyP was reported and correlated with corresponding scanning tunneling spectroscopy (STS) data. A general discussion of the appearance of organic molecules in STM can be found in refs 25 and 26. In this work, we systematically address the bias dependent contrast mechanism in a monolayer of cobalttetraphenylporphyrin (CoTPP) on Ag(111) in great detail by correlating the local probes STM and STS with laterally averaging ultraviolet photoelectron spectroscopy (UPS) measurements. Experimental Section The STM experiments were carried out in a two-chamber UHV system at a background pressure in the low 10-10 mbar regime. Preparation of the CoTPP layers was performed in UHV via sublimation onto the substrate held at room temperature. The evaporator is a home-built Knudsen cell. The evaporation time was ∼20 min for one monolayer and ∼45 min for multilayers, with the evaporator at 330 °C. Prior to sublimation, the evaporant was outgassed at slightly higher temperatures. The substrate was a Ag(111) single-crystal surface. It was cleaned by repeated cycles of Ar+ ion sputtering (500 eV) followed by annealing up to 850 K. Cut Pt/Ir tips were used as STM probes. The experiments were carried out with a commercial VT-STM instrument (RHK technologies). All STM images were recorded in the constant current mode at room temperature. Moderate low pass filtering was applied for noise reduction. The STM data were processed with the software WSxM (a description of the program can be found in ref 27). The scanning tunneling spectroscopy (STS) data were acquired in the continuous imaging tunneling spectroscopy mode (CITS).28,29 In the CITS mode, the STM tip follows the constant current contour, acquiring an STS spectrum at each point. Thus, in contrast to the spatially averaging UPS method, the local electronic structure can be probed. The UV photoelectron spectra in Figure 2 were recorded in a separate UHV system with a commercial photoelectron spectrometer (Scienta ESCA-200). This apparatus is equipped with a differentially pumped gas discharge lamp, which was used as a source of He-I radiation (21.22 eV), and other techniques for surface preparation and characterization. The UP spectra were recorded in normal emission and with a sample bias of -10 V. CoTPP and tetraphenylporphyrin (23) Kretschmann, A.; Walz, M. M.; Flechtner, K.; Steinru¨ck, H. P.; Gottfried, J. M. Chem. Commun. 2007, 568-570. (24) Flechtner, K.; Kretschmann, A.; Bradshaw, L. R.; Walz, M. M.; Steinru¨ck, H. P.; Gottfried, J. M. J. Phys. Chem. C 2007, 111, 5821-5824. (25) Chiang, S. Chem. ReV. 1997, 97, 1083-1096. (26) Sautet, P. Chem. ReV. 1997, 97, 1097-1116. (27) Horcas, I.; Fernandez, R.; Gomez-Rodriguez, J. M.; Colchero, J.; GomezHerrero, J.; Baro, A. M. ReV. Sci. Instrum. 2007, 78, 013705. (28) Stroscio, J. A., Kaiser, W. J., Eds. Scanning Tunneling Microscopy; Academic Press: New York, 1993. (29) Ruben, M.; Lehn, J. M.; Mu¨ller, P. Chem. Soc. ReV. 2006, 35, 10561067.

Comanici et al.

Figure 1. Constant current STM images of a monolayer of CoTPP on Ag (111). All STM micrographs were acquired with a current of Iset ) 0.3 nA and the indicated bias voltages at room temperature. In (a-c), the appearance of depressions, denoted as “pits”, is evident for bias voltages of -1.2 and -0.6 V. The images in (c) and (d) were acquired at the identical surface region but at different bias voltages; in both images, the same specific molecule is identified with the black circle landmark. At -0.1 V, the contrast of the pits has obviously faded. (2HTPP) with a specified purity of 98% were purchased from Porphyrin Systems.

Results and Discussion In this work, we address the bias dependent appearance of a monolayer of CoTPP on Ag(111) in STM. Our investigations were triggered by the observation that a significant portion of an ordered monolayer, prepared by evaporation of commercially available CoTPP molecules, appears as depressions or “pits” at negative bias voltages around -1 V, as can be seen, for example, in Figure 1a-c. In these STM images, the majority of molecules appear as protrusions arranged in a square order with a lattice constant of ∼1.4 nm. This type of appearance and arrangement is well-known for different metallotetraphenylporphyrins (MTPP) on various substrates.3,5,14 Within this arrangement, a significant number of depressions are clearly visible, which is especially apparent in the 3D plot in Figure 1b, as indicated by the white arrows. A statistical analysis over large scan areas resulted in an amount of 7-10% of “pit” features. Similar observations have occasionally been reported in the literature and were interpreted as missing molecules2 or impurity molecules.5 Interestingly, the STM images collected with a bias voltage of -0.1 V do not show these “pits”. This is most evident when comparing the images in Figure 1c and d, which show the same region with bias voltages of -1.2 and -0.1 V, respectively. Obviously, at -0.1 V, the pits are not visible anymore. Instead, bright features at the former positions of the pits indicate the presence of molecules. Therefore, the identification of the pits as missing molecules can be immediately ruled out. Since the observed pits are apparently not missing molecules, additional investigations of the electronic structure, in particular of the density of states of Co- and 2H-porphyrin layers, were conducted by means of ultraviolet photoelectron spectroscopy

Contrast in STM Images of a CoHTPP/2HTPP Layer

Langmuir, Vol. 24, No. 5, 2008 1899

Figure 2. Correlation of UP spectra of homogeneous monolayers of 2HTPP (green line) and CoTPP (blue line) with a bias series (-1.2 V to -0.1 V in 0.1 V steps) of constant current STM images acquired at a current of 0.3 nA. The corresponding bias voltages are marked with the red arrows, referring to the energy scale of the UP spectrum (relative to EF ≡ 0).

(UPS). Of particular relevance are the UPS signals near the Fermi level, since this region is accessible with the STM instrument. Figure 2 shows a comparison between the STM and UPS measurements. The UPS spectra for CoTPP and 2HTPP monolayers are shown in the energy range of 0 eV () EF) to -1.3 eV, for example, near the Fermi level. The spectrum of 2HTPP is almost featureless in the investigated energy region. For CoTPP, the spectrum is dominated by a peak centered at -0.6 eV below EF. As has been discussed in an earlier publication, this signal can neither be attributed to the isolated CoTPP molecule (because the signal is absent for multilayer coverages) nor is it a feature characteristic of the Ag valence band. Most likely, this signal indicates the formation of a new valence state, which results from the electronic interaction of the Co ion with the Ag surface. This initially unoccupied state is located below the Fermi level and can therefore be filled up with electrons from the Fermi sea. This interpretation, which implies a partial reduction of the Co ion by substrate electrons, is in line with the strong influence of the surface on the positions of the Co core level signals. It is also interesting to note that the signal at -0.6 eV is suppressed when an additional axial ligand such as NO is attached to the Co ion.30,31 In Figure 2, it is evident that the pits in the STM images start to disappear at bias voltages around -0.6 eV. This behavior can be understood by considering the UP spectra. Since, for a given bias voltage, all electronic states between the corresponding binding energy and the Fermi level contribute to the tunneling current, the STM signal and thus the contrast can be obtained by integration of all DOS in the UP spectra in this energy region, that is, on the right-hand side of the arrows in Figure 2. Consequently, with increasing negative bias voltage, the contribution of the peak at 0.6 eV due to the Co ion-substrate interactions increases, enhancing the contrast of Co-porphyrins and 2H-porphyrins in the STM images. On the other hand, at low energies, for example, at -0.1 eV, where this peak does not contribute to the integral, the contrast vanishes.

W A QuickTime movie demonstrating the correlation between the UPS spectra of 2HTPP and CoTPP layers and the bias dependent contrast of both species in an intermixed layer is available online. Additional evidence for the fact that the pits are indeed 2HTPP molecules is obtained from the high-resolution STM images shown in Figure 3a-c. In these images, the internal structure of the pits is resolved as four dim protrusions, which are interpreted as the corresponding phenyl groups of 2HTPP. Figure 3c shows a “pit” molecule with an overlaid and scaled model of a 2HTPP molecule: the left and the right phenyl groups fit perfectly to the corresponding protrusion, whereas the upper and lower protrusions deviate from the position of the remaining phenyl groups, possibly due to tip artifacts and/or a different molecular conformation of the adsorbed TPP molecule compared to the model. However, the most important feature with respect to the topic of this paper is the central cavity in the STM image in Figure 3b at the center of the molecule, indicating the absence

(30) Lukasczyk, T.; Flechtner, K.; Merte, L. R.; Jux, N.; Maier, F.; Gottfried, J. M.; Steinru¨ck, H. P. J. Phys. Chem. C 2007, 111, 3090-3098.

(31) Flechtner, K.; Kretschmann, A.; Steinru¨ck, H.-P.; Gottfried, J. M. J. Am. Chem. Soc. 2007, 129, 12110-12111.

Figure 3. Constant current STM images of a monolayer of CoTPP (a-c) and a 50:50 mixture of 2HTPP and CoTPP (d) on Ag(111). The tunneling parameters are Iset ) 0.4 nA, Ubias ) -1.1 V (a-d) and Iset ) 0.3 nA, Ubias ) -1.0 V (d).

1900 Langmuir, Vol. 24, No. 5, 2008

Comanici et al.

Figure 4. Constant current STM images (both with Iset ) 0.27 nA) of a monolayer of CoTPP on Ag(111) at negative bias voltage -1 V (a) and the same surface area at positive bias voltage +1V (b). The apparent contrast inversion is visualized in (c), which represents an overlay of the processed STM images in (a) and (b) as described in the text. The profiles of the apparent height along the colored lines in (a) and (b) are shown in (d). In graph (e), local ST spectra reflecting the LDOS of CoTPP (blue line) and 2HTPP (red line) are plotted. The ST spectra were extracted from CITS data of a mixed layer of CoTPP and 2HTPP.

of the central metal ion. This observation is also in line with recent studies on the in situ metalation of 2HTPP with Fe and Co,4,13,14,32 where a similar appearance of 2H-porphyrins was observed. To further confirm the assignment of the pits to 2Hporphyrins, a mixture of roughly 50% CoTPP and 50% 2HTPP was prepared and thermally evaporated in the same way as the “pure” CoTPP. Figure 3d shows an STM image of a monolayer of this porphyrin mixture. It is immediately evident that the number of pits increased dramatically as compared to the “pure” CoTPP layer studied in Figures 1 and 2. Statistical analysis reveals an increase of the number of pits from 7-10% in the “pure” CoTPP layer to 40-55% in the mixture layer, which serves as strong additional evidence that the pits are indeed due to 2HTPP. It is interesting to note that a concentration of ∼10% 2HTPP was determined by NMR in the purchased CoTPP powder. This explains the higher concentration in the prepared CoTPP monolayers as compared to the value expected from the specified purity of >98%. So far, the presented data only refer to negative bias voltages, that is, contributions from the highest occupied molecular orbital (HOMO) of the porphyrins to the tunneling contrast. Figure 4a and b shows two STM images of the same surface area, which were acquired almost simultaneously with bias voltages of -1 and +1 V, respectively (this was achieved by switching the bias voltage when switching the scan direction). The STM image obtained with -1 V (Figure 4a) exhibits the appearance described above, with CoTPP as protrusions and 2HTPP as depressions. The same surface area imaged at +1 V (Figure 4b) displays an inverted contrast. This becomes particularly evident by comparing the apparent height along the (32) Gottfried, J. M.; Flechtner, K.; Kretschmann, A.; Lukasczyk, T.; Steinru¨ck, H.-P. J. Am. Chem. Soc. 2006, 128, 5644-5645.

same line marked in Figure 4a and b, as plotted in Figure 4d. Obviously, the two depressions (pits) at -1 V appear as protrusions at +1 V. The inversion of the contrast can be further visualized in Figure 4c, which represents an overlay of the two STM images in Figure 4a and b, after some data processing. For Figure 4a, a threshold operation was applied, such that all points exceeding an apparent height of 0.15 nm (see the scale in Figure 4d) are marked red. The resulting image was superimposed onto the contrast-enhanced and green color-coded image shown in Figure 4b. From Figure 4c, it is evident that almost all depressions visible at -1 V coincide with the protrusions found at +1 V, which demonstrates that the contrast of isolated 2HTPP molecules within an ordered array of CoTPP strongly depends on the polarity of the bias voltage. This is because in one case it is the HOMO (negative bias) and in the other it is the lowest unoccupied molecular orbital (LUMO) (positive bias) of the molecules, which is imaged in STM. A similar effect was recently also observed by Auwa¨rter et al.13 for 2H-porphyrins mixed with Fe-porphyrins. There are just a few lattice sites (indicated with blue arrows) in Figure 4a-c, which do not follow the described behavior for CoTPP or for 2HTPP. These features can be interpreted as damaged or chemically altered TPP molecules. To understand the origin of the contrast inversions in the STM images, the electronic structure of the LUMO has to be investigated, which can be done by scanning tunneling spectroscopy using positive bias voltages. In Figure 4e, the corresponding ST spectra of CoTPP (blue) and 2HTPP (red) ranging from -1.5 V to +1.2 V are plotted. In agreement with the UPS data in Figure 2, a dominant peak at -0.6 V is found for CoTPP, while the 2HTPP spectra do not exhibit significant intensity maxima in the negative bias voltage region. At positive bias

Contrast in STM Images of a CoHTPP/2HTPP Layer

Langmuir, Vol. 24, No. 5, 2008 1901

voltages, the 2HTPP spectrum has a dominant peak at +0.6 V, whereas the spectrum for CoTPP exhibits a peak slightly below +1 V, which contributes only with half of its area. Consequently, for an STM image acquired at +1 V, the integral of the DOS for 2HTPP leads to a higher value (shaded in blue) compared to the one of CoTPP (shaded in red). At this point, it is noteworthy to mention that the STS data shown in Figure 4e are the differential conductivities without normalization. The corresponding normalized differential conductivity (dI/dU)/(I/U) leads to similar peak heights for the discussed LUMO DOS of 2HTPP and CoTPP. The drawback of the normalization is an unrealistic shift of all observed peaks (e.g., ∼0.1 V for the dominating CoTPP HOMO peak), which is a known problem in STS of organic molecules as discussed in the work of Wagner et al.33 At this point it should also be noted that, even though we found no indications for that, a bias induced change of the intramolecular conformation of the TPPs would not affect our findings or conclusions.

2HTPP layers in STM is related to the corresponding electronic structure. These findings, in addition with high-resolution STM images, prove that the features observed as depressions (pits) within CoTPP layers at bias voltages around -1 V are 2HTPP molecules. The origin of the contrast, that is, the appearance of CoTPP as protrusions and 2HTPP as depressions at negative bias voltages around -1 V, can be traced back to an increased DOS around -0.6 V below the Fermi level in the case of CoTPP; this contribution to the DOS is due to the interaction of the central Co ion with the Ag(111) substrate. At positive bias voltages around +1 V, the 2HTPP molecules appear brighter than the CoTPP molecules; that is, the contrast is inverted, due to a higher density of unoccupied states for the former. If one assumes distinguishable DOS fingerprints for different metalloporphyrins at energies accessible to STM (∼ -1.5 V to +1.5 V), the described contrast mechanism can be exploited to discriminate and identify individual molecules even at medium STM resolution.

Conclusions

Acknowledgment. This work has been supported by the DFG through SFB 583. The authors thank Dr. Norbert Jux for performing the NMR measurements and the group of Prof. Dr. Paul Mu¨ller for their support concerning the CITS measurements.

In summary, we have developed a consistent picture on how the bias voltage dependent appearance of mixed CoTPP and (33) Wagner, C.; Franke, R.; Fritz, T. Phys. ReV. B 2007, 75.

LA703263E