Molecular heterostructures are formed from meso-tetraphenyl porphyrins-Zn(ii) (ZnTPP) and Cu(ii)-phthalocyanines (CuPc) on the rutile TiO2(011) surface. We demonstrate that ZnTPP molecules form a quasi-ordered wetting layer with flat-lying molecules, which provides the support for growth of islands comprised of upright CuPc molecules. The incorporation of the ZnTPP layer and the growth of heterostructures increase the stability of the system and allow for room temperature scanning tunneling microscopy (STM) measurements, which is contrasted with unstable STM probing of only CuPc species on TiO2. We demonstrate that within the CuPc layer the molecules arrange in two phases and we identify molecular dimers as basic building blocks of the dominant structural phase.

In recent years, we observe an increasing interest in the development of electronic devices, in which organic molecules are connected with inorganic substrates. To this day, numerous devices, such as dye sensitized solar cells (DSSC),1 organic light-emitting diodes,2 or organic field effect transistors,3 have been commercialized and brought into market. Apart from applications in electronic devices, organic molecules are frequently utilized in other areas of technology and science, such as gas sensing or catalysis.4,5 Performance of optoelectronic devices based on light harvesting molecules is strongly dependent on the properties of the organic-inorganic and organic-organic interfaces.6–11 It is of great importance, therefore, to analyze the details of molecule binding, the role of different anchoring groups, and the application of heteromolecular assemblies in a context of resulting geometrical and electronic properties.12,13 Among different inorganic substrates, titanium dioxide is regarded as the model transition metal oxide system,14 which has a variety of applications ranging from catalysis, corrosion protective layers, optoelectronics, self-cleaning materials, chemical sensing, up to medical applications to name just a few. One of the most rapidly growing areas of interest is related to functionalized surfaces with modified chemical and (opto)electronic properties by adsorption of molecules,15–19 often focused on DSSC with organic molecules applied as dyes.1 In microscopic studies, the research was often focused on metal equipped porphyrins (Porph) and phthalocyanines (Pc), such as Cu(ii)- phthalocyanines (CuPc)/TiO2(011),15,20 CoPc/TiO2(110),21 FePc/TiO2(110),22,23 CoPc/TiO2(110)-(1 × 2) and TiO2[210],24 metal free Pc/TiO2(110),25 ZnEtioPorph/TiO2(110),26 2HTPPorph/TiO2(110),27 and CuTCPPorph/TiO2(110).28 Up to day a lot of studies have been devoted to increase understanding of the molecule-substrate interaction and resulting device efficiency.10 From the atomistic point of view, the best characterization of the molecule geometry and the interface properties could be achieved when measurements are performed at low temperature, when the system is stabilized and the thermal effects are minimized. However, in view of potential applications, the formation of stable molecular overlayers, which could be utilized at room temperature, becomes crucial. A few years ago, it was shown that CuPc molecules could be grown on the (011) face of rutile titania in an ordered manner and that by applying annealing of the sample after molecule evaporation the orientation of the molecules in the second molecular layer could be significantly changed from flat-lying into upright oriented,15,20 which, in principle, influences the overlap of the molecular orbitals, and hence, may strongly modify the molecule-substrate electronic coupling.29–32 However, it was shown that the layered structure could only be probed by scanning tunneling microscopy (STM) at cryogenic temperature, due to the high instability of the system against measurements at room temperature.15,20

In this paper, we show that by switching from homomolecular into heteromolecular structures, by substituting the wetting layer of CuPc molecules on TiO2(011)-(2 × 1) by meso-tetraphenyl porphyrins-Zn(ii) (ZnTPP) species, the stability of the CuPc overlayers is greatly enhanced and high resolution room temperature STM imaging could be performed. We demonstrate that the consecutive layers exhibit an increased level of ordering and that the limiting factor for the size of CuPc domains is imposed by the extent of TiO2 substrate terraces. The results indicate that a key factor in the formation of ordered overlayers is the quality of the wetting layer. Furthermore, we show that the ordered CuPc structures containing up-right oriented molecules on a ZnTPP wetting layer could be formed at lower annealing temperatures (150 °C) compared to the previously observed reorientation for homomolecular CuPc structures (200 °C).15,20 Together with the absence of the metastable phases comprising mostly flat-lying molecules observed for CuPc molecules and formed at moderate annealing temperatures (100 °C–150 °C),15,20 the observation may indicate that the interaction of the islands with the substrate is altered due to an exchange of the CuPc wetting layer by the ZnTPP layer. Finally, the high resolution imaging enables the revision of a previously proposed tentative structural model.20 

In the present study, the molecules were investigated on the TiO2(011)-(2 × 1) surface,33–36 which is the third most stable face of rutile titania. The choice of that particular face was driven by the previous experience with rutile TiO2 substrates functionalized with organic molecules. It was shown that, in general, the molecules tend to exhibit higher mobility and form more ordered structures on the (011) surface compared to the (110) one.17 It was argued that the higher mobility may originate from the fact that on the (011) surface the 5-fold coordinated Ti atoms are less accessible to the molecules compared to the most stable and the most studied rutile (110) surface.37 Moreover, the (011) face does not exhibit a high density of surface oxygen vacancies, typical for the (110) surface, and furthermore, the hydroxyl groups are less strongly bound to the surface. Recently, it was even demonstrated that on the rutile TiO2(110) surface the hydroxyl groups play an important role in molecule migration38 and, furthermore, our previous studies showed that the movement of carboxyl-equipped porphyrin molecules on rutile TiO2(011) is correlated with the displacement of surface hydroxyls.37 Recently, it was also demonstrated that surface hydroxyls play a crucial role in the on-surface synthesis of polymers.39,40 It is important to note that in hybrid optoelectronic devices, molecules are often deposited on nanostructured substrates; therefore, it is instructive to analyze the behavior of organic species on differently oriented surfaces.

All experiments were performed in an ultra-high vacuum (UHV) system equipped with an Omicron Variable Temperature scanning tunneling microscope (VT-STM). The base pressure in the system was kept at 1 × 10−10 mbar. TiO2 samples purchased from MaTeck were placed on a sample holder with Si wafer underneath acting as a resistive heater and subsequently prepared by cycles of Ar+ ion sputtering at 1 keV energy followed by annealing at 1040 K. The quality of the prepared surface was monitored by STM measurements. The molecules were evaporated from the Knudsen effusion cell (Kentax GmbH) after careful outgassing. The CuPc and ZnTPP molecules were provided by Sigma-Aldrich and TriPorTech, respectively. Electrochemically etched tungsten tips were applied as STM probes. Data processing and analysis were performed with the utilization of SPIP and WSxM software.41 

The schematic image of the TiO2(011)-(2 × 1) surface and typical empty state STM images are shown in Figure 1. In STM images, the dominant feature is the characteristic zig-zag pattern (marked with white circles in Figure 1(d)), which corresponds to the topmost oxygen zig-zag rows; however, the pattern does not directly mark the positions of oxygen atoms.36 The bright dots clearly seen within the reconstructed rows are attributed to surface hydroxyl groups.

FIG. 1.

TiO2(011)-(2 × 1) surface. (a) Top and (b) perspective views. White, yellow, and green circles indicate oxygen atoms corresponding to the double zig-zag row, remaining oxygen atoms are highlighted in blue, red, and black and small dots indicate titanium atoms; (c) and (d) empty state STM images, bright dots visible over surface reconstruction rows are hydroxyl groups, white circles in (d) mark the zig-zag pattern of the STM image. Scanning parameters: I = 10 pA, U = + 1.7 V (b) and +2.0 V (d).

FIG. 1.

TiO2(011)-(2 × 1) surface. (a) Top and (b) perspective views. White, yellow, and green circles indicate oxygen atoms corresponding to the double zig-zag row, remaining oxygen atoms are highlighted in blue, red, and black and small dots indicate titanium atoms; (c) and (d) empty state STM images, bright dots visible over surface reconstruction rows are hydroxyl groups, white circles in (d) mark the zig-zag pattern of the STM image. Scanning parameters: I = 10 pA, U = + 1.7 V (b) and +2.0 V (d).

Close modal

Following our previous experiments, we have decided to form a wetting layer from ZnTPP molecules (Figure 2(a)), which were found to create an ordered layer stable at room temperature.37 This was motivated by the fact that CuPc molecules are instable during STM measurements at room temperature and could only be probed at cryogenic conditions.15 The chemical structure of a CuPc molecule is shown in Figure 2(b). Furthermore, the microscopic investigation on this system is relevant, due to the reported effect of such a mixture on the photovolatic parameters of dye-sensitized solar cells.42,43 Therefore, we have decided to evaporate CuPc molecules on a wetting layer formed by ZnTPP species leading to the formation of stable overlayers.

FIG. 2.

Chemical structure of the molecules used in the study: (a) meso-tetraphenyl porphyrin-Zn(ii) (ZnTPP), (b) Cu(ii)-phthalocyanine (CuPc); color coding, dark gray — carbon, light gray — hydrogen, blue — nitrogen, violet — zinc, yellow — copper.

FIG. 2.

Chemical structure of the molecules used in the study: (a) meso-tetraphenyl porphyrin-Zn(ii) (ZnTPP), (b) Cu(ii)-phthalocyanine (CuPc); color coding, dark gray — carbon, light gray — hydrogen, blue — nitrogen, violet — zinc, yellow — copper.

Close modal

In the first step of multilayer formation, ZnTPP molecules were evaporated on the TiO2 surface to form an ordered wetting layer. As recently demonstrated by Olszowski et al.,37 the molecules adopt a flat-lying geometry and are stabilized by intermolecular interactions. The layer exhibits very similar ordering as the CuPc layer reported by Godlewski et al.20 However, contrary to CuPc molecules, the ZnTPP layer could be probed with STM at room temperature exhibiting sufficient stability. A typical empty state image of the ZnTPP layer is shown in Figure 3(a). Because the TiO2(011)-(2 × 1) surface reconstruction zig-zag rows separated by approximately 0.92 nm are too densely packed to accommodate neighboring molecules over every surface row, the molecules are located over every second surface row as confirmed by cross sections shown in Figure 3(b). This leads to the formation of mixed straight lines with chessboard-like structure.

FIG. 3.

Quasi-ordered wetting layer formed by ZnTPP molecules on TiO2(011). (a) Empty state STM image acquired with I = 5 pA and U = + 2.0 V, (b) cross sections along the [100] direction over the chessboard-like pattern (blue) and straight lines (green) showing adsorption of molecules over every second surface reconstructed row.

FIG. 3.

Quasi-ordered wetting layer formed by ZnTPP molecules on TiO2(011). (a) Empty state STM image acquired with I = 5 pA and U = + 2.0 V, (b) cross sections along the [100] direction over the chessboard-like pattern (blue) and straight lines (green) showing adsorption of molecules over every second surface reconstructed row.

Close modal

In the following step, CuPc molecules were evaporated onto the layer of ZnTPP molecules. The deposition of a submonolayer coverage at room temperature results in an unordered growth with formation of small clusters as shown in Figure 4(a). The analysis of the cluster apparent heights indicates that some molecules are lying flat on the ZnTPP wetting layer, whereas others adopt a more vertical position as indicated by cross sections in Figure 4(c). In general, however, the growth clearly proceeds in a disordered manner. It is worth to note that the stability of the system against STM imaging was much lower than for the pure ZnTPP wetting layer. It was often observed that some mobile molecules on top of the first layer were influencing the measurements and leading to streaky patterns. The situation was changed notably when the heteromolecular system was annealed at 150 °C (Figure 4(b)). It can clearly be seen that the CuPc molecules form large ordered islands on top of the ZnTPP wetting layer. The islands exhibit a rectangular shape and are mostly elongated along the [11-1] or [-11-1] surface directions. Their internal structure is similar to the one observed previously for CuPc molecules on the bare TiO2(011)-(2 × 1) surface.20 The analysis of the island apparent heights, which reach 1.0-1.2 nm, indicates that the CuPc molecules within the islands are standing upright. Flat lying species are significantly lower, by approximately 40%, reaching 0.6-0.7 nm in height.20 

FIG. 4.

Empty state STM images of CuPc structures formed on top of the ZnTPP wetting layer. (a) Unordered clusters obtained after deposition of 0.15 ML of CuPc molecules at room temperature, (b) ordered islands comprised of upright CuPc molecules formed after additional deposition of 0.1 ML on the sample shown in (a) and subsequent annealing at 150 °C, (c) cross sections from STM topographies shown in panels (a) and (b) indicating flat or upright oriented molecules, (d) enlarged STM image showing the part marked with yellow dashed rectangle in panel (c), blue rectangles indicate the row structures, yellow rectangle marks the ZnTPP molecules forming a straight pattern within the wetting layer. Scanning parameters: I = 10 pA, U = + 2.0 V (a), +2.4 V (b), and +2.6 V (d).

FIG. 4.

Empty state STM images of CuPc structures formed on top of the ZnTPP wetting layer. (a) Unordered clusters obtained after deposition of 0.15 ML of CuPc molecules at room temperature, (b) ordered islands comprised of upright CuPc molecules formed after additional deposition of 0.1 ML on the sample shown in (a) and subsequent annealing at 150 °C, (c) cross sections from STM topographies shown in panels (a) and (b) indicating flat or upright oriented molecules, (d) enlarged STM image showing the part marked with yellow dashed rectangle in panel (c), blue rectangles indicate the row structures, yellow rectangle marks the ZnTPP molecules forming a straight pattern within the wetting layer. Scanning parameters: I = 10 pA, U = + 2.0 V (a), +2.4 V (b), and +2.6 V (d).

Close modal

The findings indicate that upon thermal annealing the molecules gain enough energy to reorient and create compact islands, in which the total energy is minimized by mutual interactions. Islands exhibit a row structure as indicated by dashed rectangles in Figure 4(d), with slightly meandering rows running in general parallel to longer island edges. We assume that the shape of the islands is driven by mutual CuPc - CuPc intermolecular interactions. This means that the structure is dominated by the assembly of CuPc molecules within rows forming the islands. In the most often studied adsorption of organic molecules on metals, phthalocyanines equipped with different central metal atoms are reported to form usually ordered layers of flat-lying molecules and the assembly is governed by the balance between the substrate influence and intermolecular interactions.44–48 The situation is, however, different for semiconducting and insulating substrates. Upright orientation of phthalocyanine molecules was already reported for systems, in which the interaction between the molecules dominates over the binding to the underlying substrate, like, for example, on Ag(111)/C6012,13 or passivated silicon.49 Therefore, it is plausible to suppose that the interaction of the CuPc molecules with the ZnTPP layer is relatively weak and the islands mimic the CuPc crystal structure similar to the behavior on other substrates, which interact weakly with molecules, e.g., KCl.50,51 Recently, it was also demonstrated that the molecule — substrate and molecule — molecule interaction might be significantly influenced by the central metal atom. Wagner et al. have shown that ZnPc, CuPc, and CoPc molecules form distinctly different structures on B passivated Si(111) due to the differences in the coupling between the substrate and the central metal atom.52 

Further room temperature deposition of the CuPc molecules on the substrate with already formed well-ordered islands results in the formation of small clusters located between the islands. This is shown in Figures 5(a) and 5(b). STM images show that the newly deposited molecules form either short straight assemblies marked with a white dashed rectangle, or agglomerates which are indicated by purple dashed rectangles. Within the straight assemblies, the molecules are flat-lying, as deduced from the cross sections shown in Figure 5(c), and they are following the direction of the substrate reconstruction rows. The agglomerates are formed from upright oriented molecules but exhibit slightly less degree of ordering compared to the islands. They are also often found attached to the edges of the molecular islands. The upright position of molecules within the agglomerates could be concluded from the cross sections shown in Figure 5(c). The apparent height of previously fabricated islands and newly created agglomerates is identical reaching approximately 1.0-1.2 nm. Post-deposition annealing of the system at 150 °C leads to an enlargement of the ordered islands and disappearance of the agglomerates and straight assemblies.

FIG. 5.

0.4 ML (0.2 ML annealed at 150 °C + 0.2 ML deposited at room temperature) of CuPc molecules on a ZnTPP wetting layer. (a) and (b) empty state STM images, dark blue rectangles mark unordered agglomerates of upright CuPc molecules, white rectangle indicates straight assemblies of flat-lying CuPc species, (c) cross sections of STM images shown in panels (a) and (b), STM images acquired with I = 5 pA and U = + 2.1 V.

FIG. 5.

0.4 ML (0.2 ML annealed at 150 °C + 0.2 ML deposited at room temperature) of CuPc molecules on a ZnTPP wetting layer. (a) and (b) empty state STM images, dark blue rectangles mark unordered agglomerates of upright CuPc molecules, white rectangle indicates straight assemblies of flat-lying CuPc species, (c) cross sections of STM images shown in panels (a) and (b), STM images acquired with I = 5 pA and U = + 2.1 V.

Close modal

Further deposition of CuPc molecules followed by annealing leads again to the formation of large ordered islands, which in fact cover the whole surface. The limiting factor for their planar growth is the size of the surface terraces. Figures 6(a) and 6(b) show the surface almost entirely covered by CuPc islands formed from upright molecules. It is worth to note that in the areas that are not covered by islands, on the ZnTPP wetting layer, the STM imaging is much less stable and streaky patterns were frequently observed. These areas are imaged as dark depressions in Figures 6(a) and 6(b). On top of the upright oriented CuPc molecules, some additional small bright features are observed, as marked by dashed dark blue circles in Figure 6(b). We attribute them to single CuPc molecules located usually at some CuPc layer defects, i.e., domain boundaries and molecule vacancies. These molecules are not strongly attached to the support and often exhibit fuzzy appearance or frequently move over the surface during imaging. Figure 6(c) shows the magnified image of the upright layer of CuPc molecules, which provides higher resolution and allows us to draw some conclusions on the structure of the layer. As demonstrated by dashed blue rectangles, the layer is composed of molecular rows running approximately along [11-1] or [-11-1] surface directions. From Figure 6(c), it is clear that the CuPc rows are formed by smaller building blocks marked by full blue rectangles. They are most often oriented along the [01-1] surface direction, i.e., parallel to surface rows.

FIG. 6.

High coverage layer of CuPc on ZnTPP wetting layer. (a)–(c) show an almost complete layer of upright CuPc molecules, lower areas in panels (a) and (b) are the ZnTPP wetting layer not covered by CuPc, green square in (a) marks the area shown in (b), dark circle in (b) indicates single molecule immobilized on top of the layer by defects, dashed blue rectangles in (b) and (c) show the direction of molecular rows, blue rectangles in (c) mark the basic building blocks of the molecular rows; (d)–(f) show the sample after deposition of additional 0.15 ML of CuPc molecules without annealing, green square in (e) indicates the area shown magnified in (f), red rectangles in (f) mark basic building blocks of the second layer CuPc island, the contrast in (f) was adjusted to the CuPc island, the inset in (f) shows the structure of individual building blocks in the first and second CuPc layers; scanning parameters: I = 2 pA ((a) and (b)), 3 pA ((e) and (f)), 5 pA (c) and 20 pA (d), U = + 2.2 V ((a), (b), and (d)–(f)), +2.6 V (c).

FIG. 6.

High coverage layer of CuPc on ZnTPP wetting layer. (a)–(c) show an almost complete layer of upright CuPc molecules, lower areas in panels (a) and (b) are the ZnTPP wetting layer not covered by CuPc, green square in (a) marks the area shown in (b), dark circle in (b) indicates single molecule immobilized on top of the layer by defects, dashed blue rectangles in (b) and (c) show the direction of molecular rows, blue rectangles in (c) mark the basic building blocks of the molecular rows; (d)–(f) show the sample after deposition of additional 0.15 ML of CuPc molecules without annealing, green square in (e) indicates the area shown magnified in (f), red rectangles in (f) mark basic building blocks of the second layer CuPc island, the contrast in (f) was adjusted to the CuPc island, the inset in (f) shows the structure of individual building blocks in the first and second CuPc layers; scanning parameters: I = 2 pA ((a) and (b)), 3 pA ((e) and (f)), 5 pA (c) and 20 pA (d), U = + 2.2 V ((a), (b), and (d)–(f)), +2.6 V (c).

Close modal

Figures 6(d) and 6(e) present the contrasting behavior of CuPc deposited at room temperature onto the ZnTPP wetting layer and on top of the first CuPc ad-layer. Deposition of the molecules at room temperature results in spontaneous — without post-deposition annealing — formation of second layer islands containing upright molecules. The images show the presence of islands elongated mostly in the [11-1] and [-11-1] surface directions. CuPc molecules which are deposited on the ZnTPP wetting layer do not form islands but again unordered agglomerates following the behavior of the 0.15 ML of CuPc, as was shown in Figure 4(a). The inspection of the STM images indicates that some molecules are also decorating the step edges of the first layer islands. The structure of the second layer CuPc islands is similar to the one of the first layer underneath. The magnified image of the second layer island with the contrast adjusted to its surface is shown in Figure 6(f). The detailed analysis indicates that the building blocks of the rows, which are marked by red rectangles in Figure 6(f), are oriented perpendicular in neighboring layers. This is schematically shown in the inset of Figure 6(f). That points to the fact that the interaction of the molecules in neighboring layers governs the orientation of the building blocks of the islands. However, the direction of the island growth and their elongation is in both layers identical.

As shown in Figs. 7(a)7(c), annealing of the structures shown in Figures 6(d) and 6(e) allows for the complete filling of the first CuPc layer of upright molecules and does not leave uncovered areas. High resolution images show that the molecules form rows aligned mostly along the [11-1] or [-11-1] directions of the surface, which resembles the structure reported for CuPc molecules observed at low temperature on TiO2.20 These rows are built up by molecular building blocks which are indicated by dashed blue rectangles in Figure 7(c). In the previous studies,20 it was argued that the molecular building block corresponds to single molecules. However, our present high resolution study does not support such interpretation anymore. In the zoomed image shown in Figure 7(f), the white arrow indicates a single molecule vacancy within the layer and the violet arrows mark single molecules, which forms only half of the typical building blocks. This suggests that the single building block contains typically two CuPc molecules. On the basis of the STM measurements, it is not possible to unquestionably define the orientation of the molecules within the building blocks. In Figure 7(c), blue filled rectangles indicate one of the possible orientations. The analysis suggests that the molecules form dimers due to the balance between the mutual interaction within the layers and the influence of the flat-lying wetting layer. However, it is clear from Figure 7(c) that within the blocks the relative orientation and separation of the molecules vary slightly leading to the non-identical appearance of these blocks as indicated by a profile shown in Figure 7(e) and the zoomed image in Figure 7(f). Further, that leads also to meandering shape of the columns, which are clearly not completely straight. The tentative model of the structure is shown in Figure 7(g) with dashed blue ovals indicating the STM recorded building blocks.

FIG. 7.

Fully covered CuPc molecule layer on top of the ZnTPP wetting layer. (a)–(c) show empty state STM images, white contours in (b) show the parts of the sample with the second CuPc phase, dashed blue rectangles indicate the direction of molecular rows, blue square marks the area shown in (c), yellow arrow indicates the single molecule vacancy, blue and green rectangles indicate individual molecules within the first and the second phases, respectively, dashed rectangles mark basic building blocks of the first phase, white line shows the distance covered by ten CuPc molecules, and yellow line indicates the width of 5 molecular rows of the second phase; (d) and (e) show profiles along blue and green lines in (c) showing molecule spacing in the second and the first phases, respectively; (f) shows the zoom of the area marked with white, dashed square in (c) with white arrow indicating the molecule vacancy, violet arrows marking single molecules (a dashed one points a molecule that is mobile during imaging) and green arrows indicating molecular dimers, all arrows show the molecules positioned along the cross section profile displayed in (d); (g) shows the tentative model of the first phase with blue ellipses indicating the STM recorded basic building block; (h) presents the structural model of the second phase, dashed green rectangles indicate single molecules. Scanning parameters: I = 5 pA ((a) and (b)), 3 pA (c), voltage = + 2.0 V ((a) and (c)), +2.2 V (b).

FIG. 7.

Fully covered CuPc molecule layer on top of the ZnTPP wetting layer. (a)–(c) show empty state STM images, white contours in (b) show the parts of the sample with the second CuPc phase, dashed blue rectangles indicate the direction of molecular rows, blue square marks the area shown in (c), yellow arrow indicates the single molecule vacancy, blue and green rectangles indicate individual molecules within the first and the second phases, respectively, dashed rectangles mark basic building blocks of the first phase, white line shows the distance covered by ten CuPc molecules, and yellow line indicates the width of 5 molecular rows of the second phase; (d) and (e) show profiles along blue and green lines in (c) showing molecule spacing in the second and the first phases, respectively; (f) shows the zoom of the area marked with white, dashed square in (c) with white arrow indicating the molecule vacancy, violet arrows marking single molecules (a dashed one points a molecule that is mobile during imaging) and green arrows indicating molecular dimers, all arrows show the molecules positioned along the cross section profile displayed in (d); (g) shows the tentative model of the first phase with blue ellipses indicating the STM recorded basic building block; (h) presents the structural model of the second phase, dashed green rectangles indicate single molecules. Scanning parameters: I = 5 pA ((a) and (b)), 3 pA (c), voltage = + 2.0 V ((a) and (c)), +2.2 V (b).

Close modal

Additionally, at the saturation coverage, we observe also the appearance of a second phase not reported previously, in which the molecular rows are running along the [01-1] direction, i.e., parallel to the TiO2 surface rows. These structures are less frequently observed and are marked by white contours in Figure 7(b). The high resolution image of this phase is shown in the middle part of the magnification in Figure 7(c). STM images show that each protrusion corresponds to one CuPc molecule, further indicated by the single molecule vacancy, pointed by a yellow arrow. Periodicity of molecular rows indicates that the molecules are accommodated over every surface rows, i.e., are separated by approximately 0.92 nm. This dimension is smaller than the size of the molecule indicating that CuPc adopt a slightly rotated geometry as shown by green blue rectangles in Figure 7(c). The spacing of the molecules along the [01-1] direction reaches approximately 0.6 nm (see Figure 7(d)), which is more than the distance between molecules within a crystal formed on ZnO,53 but the difference most likely originates from the influence of the underlying ZnTPP layer. The tentative model of the structure is shown in Figure 7(h). It is worth to note that although the structures are stable enough to be probed by STM at room temperature, the individual molecules tend to exhibit some mobility, especially when located close to defects, i.e., molecule vacancies, where we can often observe hoping of molecules and motion of vacancies.

When a complete layer of upright CuPc molecules is formed, it provides the support for further growth of ordered structures. As demonstrated in Figure 8(a) after annealing at 150 °C, the molecules are forming second layer islands, which exhibit a higher degree of ordering compared to the first layer. It is clearly seen that the density of defects and vacancies is highly reduced, whereas the structure resembles the ordering of the first layer with again the presence of both the above described phases. Interestingly, the STM image of the second layer islands shows the presence of an additional pattern, which may originate from the role of the second CuPc layer in tunneling from the tip to the substrate through a multilayer structure. This pattern is marked with black rectangles in Figures 8(c) and 8(d).

FIG. 8.

Empty state STM images of molecular islands obtained from 1.25 ML of CuPc annealed at 150 °C; dashed square in (a)–(c) indicates the area shown in (b)–(d), respectively, dashed contour in (c) marks the part of the 2nd layer island forming the second phase, dashed black rectangles in (c) and (d) indicate the additional pattern obtained on top of the island. Scanning parameters: I = 4 pA, U = + 2.2 V.

FIG. 8.

Empty state STM images of molecular islands obtained from 1.25 ML of CuPc annealed at 150 °C; dashed square in (a)–(c) indicates the area shown in (b)–(d), respectively, dashed contour in (c) marks the part of the 2nd layer island forming the second phase, dashed black rectangles in (c) and (d) indicate the additional pattern obtained on top of the island. Scanning parameters: I = 4 pA, U = + 2.2 V.

Close modal

We have shown that application of ZnTPP molecules to form a wetting layer on the TiO2(011)-(2 × 1) surface provides good support for further growth of CuPc overlayers. Room temperature STM measurements demonstrated that annealing of the system after CuPc molecule deposition leads to the formation of ordered islands built from upright CuPc molecules. The limiting factor for the size of molecular islands is the dimension of the TiO2 terraces. The detailed analysis of the high resolution images indicates that within the islands CuPc molecules form dimers, which are arranged in slightly meandering rows, which mostly follow the [11-1] or [-11-1] directions of the substrate. The experiments indicate that the stability of the molecular layers could be influenced and increased by substitution of homomolecular structures by heteromolecular overlayers, which might be advantageous in the construction of the prototypical DSSC devices.

This work was supported by a grant from Switzerland through the Swiss Contribution to the enlarged European Union (Joint Polish-Swiss Research Program) No. PSPB-085/2010 “Molecular assemblies on semiconductors and insulating surfaces.” http://www.molSurf.eu. Furthermore, the Swiss National Science Foundation (SNF) and the Swiss Nanoscience Institute (SNI) are acknowledged for financial support.

1.
B.
O’Regan
and
M.
Grätzel
,
Nature
353
,
737
740
(
1991
).
2.
R.
Capelli
,
S.
Toffanin
,
G.
Generali
,
H.
Usta
,
A.
Facchetti
, and
M.
Muccini
,
Nat. Mater.
9
,
496
503
(
2010
).
3.
H.
Sirringhaus
,
Adv. Mater.
26
,
1319
1335
(
2014
).
4.
X.
Chen
and
S. S.
Mao
,
Chem. Rev.
107
,
2891
2959
(
2007
).
5.
R.
Rella
,
A.
Rizzo
,
A.
Licciulli
,
P.
Siciliano
,
L.
Troisi
, and
L.
Valli
,
Mater. Sci. Eng. C
22
,
439
(
2002
).
6.
H.
Ishii
,
K.
Sugiyama
,
E.
Ito
, and
K.
Seki
,
Adv. Mater.
11
,
605
625
(
1999
).
7.
A.
Kahn
,
N.
Koch
, and
W. Y.
Gao
,
J. Polym. Sci. B
41
,
2529
(
2003
).
9.
L.
Zhang
and
J. M.
Cole
,
ACS Appl. Mater. Interfaces
7
,
3427
3455
(
2015
).
10.
R.
Ambre
,
K. B.
Chen
,
C. F.
Yao
,
L.
Luo
,
E. W. G.
Diau
, and
C. H.
Hung
,
J. Phys. Chem. C
116
,
11907
11916
(
2012
).
11.
A. S.
Hart
,
C. B.
KC
,
H. B.
Gobeze
,
L. R.
Sequeira
, and
F.
D’Souza
,
ACS Appl. Mater. Interfaces
5
,
5314
5323
(
2013
).
12.
W.
Chen
,
S.
Chen
,
H.
Huang
,
D. C.
Qi
,
X. Y.
Gao
, and
A. T. S.
Wee
,
Appl. Phys. Lett.
92
,
063308
(
2008
).
13.
H.
Huang
,
W.
Chen
,
S.
Chen
,
D. C.
Qi
,
X. Y.
Gao
, and
A. T. S.
Wee
,
Appl. Phys. Lett.
94
,
163304
(
2009
).
14.
15.
S.
Godlewski
,
A.
Tekiel
,
J. S.
Prauzner-Bechcicki
,
J.
Budzioch
,
A.
Gourdon
, and
M.
Szymonski
,
J. Chem. Phys.
134
,
224701
(
2011
).
16.
S.
Godlewski
,
A.
Tekiel
,
W.
Piskorz
,
F.
Zasada
,
J. S.
Prauzner-Bechcicki
,
Z.
Sojka
, and
M.
Szymonski
,
ACS Nano
6
,
8536
8545
(
2012
).
17.
S.
Godlewski
and
M.
Szymonski
,
Int. J. Mol. Sci.
14
,
2946
2966
(
2013
).
18.
J. S.
Prauzner-Bechcicki
,
S.
Godlewski
,
A.
Tekiel
,
P.
Cyganik
,
J.
Budzioch
, and
M.
Szymonski
,
J. Phys. Chem. C
113
,
9309
(
2009
).
19.
F.
Zasada
,
W.
Piskorz
,
S.
Godlewski
,
J. S.
Prauzner-Bechcicki
,
A.
Tekiel
,
J.
Budzioch
,
P.
Cyganik
,
M.
Szymonski
, and
Z.
Sojka
,
J. Phys. Chem. C
115
,
4134
(
2011
).
20.
S.
Godlewski
,
A.
Tekiel
,
J. S.
Prauzner-Bechcicki
,
J.
Budzioch
, and
M.
Szymonski
,
ChemPhysChem
11
,
1863
(
2010
).
21.
N.
Ishida
and
D.
Fujita
,
J. Phys. Chem. C
116
,
20300
20305
(
2012
).
22.
P.
Palmgren
,
S.
Yu
,
F.
Hennies
,
K.
Nilson
,
B.
Akemark
, and
M.
Goethelid
,
J. Chem. Phys.
129
,
074707
(
2008
).
23.
P.
Palmgren
,
K.
Nilson
,
S.
Yu
,
F.
Hennies
,
T.
Angot
,
C. I.
Nlebedim
,
J.-M.
Layet
,
G.
Le Lay
, and
M.
Goethelid
,
J. Phys. Chem. C
112
,
5972
(
2008
).
24.
Y.
Wang
,
Y.
Ye
, and
K.
Wu
,
J. Phys. Chem. B
110
,
17960
(
2006
).
25.
P.
Palmgren
,
B. R.
Priya
,
N. P. P.
Niraj
, and
M.
Goethelid
,
Sol. Energy Mater. Sol. Cells
90
,
3602
(
2006
).
26.
M.
Lackinger
,
M. S.
Janson
, and
W.
Ho
,
J. Chem. Phys.
137
,
234707
(
2012
).
27.
C.
Wang
,
Q.
Fan
,
S.
Hu
,
H.
Ju
,
X.
Feng
,
Y.
Han
,
H.
Pan
,
J.
Zhu
, and
J. M.
Gottfried
,
Chem. Commun.
50
,
8291
8294
(
2014
).
28.
R.
Jöhr
,
A.
Hinaut
,
R.
Pawlak
,
A.
Sadeghi
,
S.
Saha
,
S.
Goedecker
,
B.
Such
,
M.
Szymonski
,
E.
Meyer
, and
T.
Glatzel
,
J. Chem. Phys.
143
,
094202
(
2015
).
29.
C.
Videlot
,
A.
El Kassimi
, and
D.
Fichou
,
Sol. Energy Mater. Sol. Cells
63
,
69
82
(
2000
).
30.
Ch. D.
Dimitrakopoulos
and
P. R. L.
Malenfant
,
Adv. Mater.
14
,
99
(
2002
).
31.
M.
Szybowicz
,
T.
Runka
,
M.
Drozdowski
,
W.
Bala
,
A.
Grodzicki
,
P.
Piszczek
, and
A.
Bratkowski
,
J. Mol. Struct.
704
,
107
(
2004
).
32.
K. O.
Sylvester-Hvid
,
J. Phys. Chem. B
110
,
261
(
2006
).
33.
X.
Torrelles
,
G.
Cabailh
,
R.
Lindsay
,
O.
Bikondoa
,
J.
Roy
,
J.
Zegenhagen
,
G.
Teobaldi
,
W. A.
Hofer
, and
G.
Thornton
,
Phys. Rev. Lett.
101
,
185501
(
2008
).
34.
X.-Q.
Gong
,
N.
Khorshidi
,
A.
Stierle
,
V.
Vonk
,
C.
Ellinger
,
H.
Dosch
,
H.
Cheng
,
A.
Selloni
,
Y.
He
,
O.
Dulub
, and
U.
Diebold
,
Surf. Sci.
603
,
138
(
2009
).
35.
S. E.
Chamberlin
,
C. J.
Hirschmugl
,
H. C.
Poon
, and
D. U.
Saldin
,
Surf. Sci.
603
,
3367
(
2009
).
36.
T.
Woolcot
,
G.
Teobaldi
,
C. L.
Pang
,
N. S.
Beglitis
,
A. J.
Fisher
,
W. A.
Hofer
, and
G.
Thornton
,
Phys. Rev. Lett.
109
,
156105
(
2012
).
37.
P.
Olszowski
,
Ł.
Zając
,
S.
Godlewski
,
B.
Such
,
R.
Jöhr
,
T.
Glatzel
,
E.
Meyer
, and
M.
Szymonski
,
J. Phys. Chem. C
119
,
21561
21566
(
2015
).
38.
S.-C.
Li
,
L.-N.
Chu
,
X.-Q.
Gong
, and
U.
Diebold
,
Science
328
,
882
884
(
2010
).
39.
M.
Kolmer
,
A. A. A.
Zebari
,
J. S.
Prauzner-Bechcicki
,
W.
Piskorz
,
F.
Zasada
,
S.
Godlewski
,
B.
Such
,
Z.
Sojka
, and
M.
Szymonski
,
Angew. Chem., Int. Ed.
52
,
10300
10303
(
2013
).
40.
M.
Kolmer
,
R.
Zuzak
,
A. A. A.
Zebari
,
S.
Godlewski
,
J. S.
Prauzner-Bechcicki
,
W.
Piskorz
,
F.
Zasada
,
Z.
Sojka
,
D.
Bleger
,
S.
Hecht
, and
M.
Szymonski
,
Chem. Commun.
51
,
11276
11279
(
2015
).
41.
I.
Horcas
,
R.
Fernandez
,
J. M.
Gomez-Rodriguez
,
J.
Colchero
,
J.
Gomez-Herrero
, and
A. M.
Baro
,
Rev. Sci. Instrum.
78
,
013705
(
2007
).
42.
D.
Wróbel
,
A.
Boguta
, and
A.
Wójcik
,
J. Phys. IV
117
,
101
113
(
2004
).
43.
H.
Deng
,
Y.
Zhou
,
H.
Mao
, and
Z.
Lu
,
Synth. Met.
92
,
269
274
(
1998
).
44.
K. W.
Hipps
,
D. E.
Barlow
, and
U.
Mazur
,
J. Phys. Chem. B
104
,
2444
(
2000
).
45.
H.
Karacuban
,
M.
Lange
,
J.
Schaffert
,
O.
Weingart
,
Th.
Wagner
, and
R.
Moller
,
Surf. Sci.
603
,
L39
(
2009
).
46.
M.
Takada
and
H.
Tada
,
Chem. Phys. Lett.
392
,
265
(
2004
).
47.
J. Y.
Grand
,
T.
Kunstmann
,
D.
Hoffmann
,
A.
Haas
,
M.
Dietsche
,
J.
Seifritz
, and
R.
Moller
,
Surf. Sci.
366
,
403
(
1996
).
48.
K. W.
Hipps
,
X.
Lu
,
X. D.
Wang
, and
U.
Mazur
,
J. Phys. Chem.
100
,
11207
(
1996
).
49.
J.
Gardener
,
J. H. G.
Owen
,
K.
Miki
, and
S.
Heutz
,
Surf. Sci.
602
,
843
(
2008
).
50.
A.
Hoshino
,
T.
Takenaka
,
M.
Izukashi
,
H.
Yoshida
, and
H.
Miyaji
,
Jpn. J. Appl. Phys., Part 1
43
,
4344
(
2004
).
51.
A.
Hoshino
,
Y.
Takenaka
, and
H.
Miyaji
,
Acta Crystallogr., Sect. B: Struct. Sci.
59
,
393
(
2003
).
52.
S. R.
Wagner
,
B.
Huang
,
C.
Park
,
J.
Feng
,
M.
Yoon
, and
P.
Zhang
,
Phys. Rev. Lett.
115
,
096101
(
2015
).
53.
A. C.
Cruickshank
,
C. J.
Dotzler
,
S.
Din
,
S.
Heutz
,
M. F.
Toney
, and
M. P.
Ryan
,
J. Am. Chem. Soc.
134
,
14302
14305
(
2012
).