Mn-Coordinated Stillbenedicarboxylic Ligand Supramolecule

Oct 19, 2007 - The fabrication of Mn-coordinated metal−organic networks on the reconstructed Au(111) surface was investigated by scanning tunneling ...
4 downloads 0 Views 806KB Size
16946

J. Phys. Chem. C 2007, 111, 16946-16950

Mn-Coordinated Stillbenedicarboxylic Ligand Supramolecule Regulated by the Herringbone Reconstruction of Au(111) Yan-Feng Zhang,† Na Zhu,† and T. Komeda*,†,‡ Institute of Multidisciplinary Research for AdVanced Materials, Tohoku UniVersity, Japan, and CREST, JST, Japan ReceiVed: June 25, 2007; In Final Form: July 29, 2007

The fabrication of Mn-coordinated metal-organic networks on the reconstructed Au(111) surface was investigated by scanning tunneling microscopy(STM), with stillbenedicarboxylic acid (SDA) as the linker molecules. Two network phases have been observed, sorted by the geometry of node Mn atoms with a parallel or a perpendicular geometry with regard to the adjacent units. A commensurate ordering of the networks with the substrate herringbone reconstructions was identified. The periodic alterations of the Mn unit geometry may provide us with clear evidence of the modulation effects from the substrate reconstructions. A ladder structure was obtained for a SDA oversaturated surface, which revealed another coordination interaction between the Mn atoms and the SDA molecules.

Introduction The fabrication of materials with a reduced dimensionality and a designed structure is a very intriguing project from both scientific and technological point of views. In the traditionally bottom-up approach, superstructures are assembled by relatively weak interactions such as hydrogen bonding, van der Waals, or electrostatic forces with a limited thermal stability in these systems.1-3 Metal coordination interactions are stronger, more directional, and more selective than the hydrogen bond, and proved to be a widely adapted method for fabricating the molecule building blocks. Meanwhile, the metal-organic nanosystems attract our wide attention because of their potential applications in device design that related to sensing, switching, and information storage.4-6 Novel physical and chemical properties involved with magnetic, electronic, and catalysis are also very promising topics in these nanosystems.7-10 In recent years, metal-organic coordination networks (MOCNs), formed by coordination bonds between metallic center atoms and organic linker molecules, have been realized on metal surfaces under ultrahigh vacuum conditions.11-18 On an inhomogeneous 2D surface, the adsorption of organic molecules and magnetic atoms may be influenced by the different adsorption sites of the substrate. The investigation of the substrate effects on the fabrication of 2D networks should be a very interesting orientation because some new structures with unique properties will appear with regard to the traditional assembly principle in 3D coordination fabrication. The identification of the bonding configuration with the scanning tunneling microscopy (STM) can push this investigation to a new level of atomic precision. In this work, we report the STM investigation of stillbenedicarboxylic acid molecule (SDA) linked metal-organic networks fabricated on a reconstructed Au(111) surface. Here, a * Corresponding author. E-mail address: [email protected]. Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, Katahira 2-1-1, Aoba-ku, Sendai 980-8577, Japan. Tel: +81 22 2175368. Fax: +81 22 2175404. † Tohoku University. ‡ CREST.

unique magnetic material Mn is selected as the node atom, and this is different from some published investigations with Fe, Co, Cu, or other materials. The morphology of the Mn-based coordination networks is investigated systematically; thus, the phase transitions, the long-range modulation from the herringbone structure of the Au(111) substrate, and the various network shape corresponding to the different bonding nature of Mn atoms is illustrated. Experimental Section The experiment was carried out with a home-built ultrahigh vacuum low-temperature STM, which was equipped with a mini-chamber suitable for sample preparation and molecule evaporation. The Au(111) substrate was prepared by a standard method with Ar+ sputtering and then an annealing process. The commercially available SDA molecule (Alfa Aesar, g99%) was degassed up to 530 K in a Ta boat for several hours and evaporated with a flux rate of about 1/3 monolayer (ML) per minute. The growth rate can be checked by an in situ thickness monitor. The Au(111) substrate was held at room temperature during the SDA molecule evaporation. Then, Mn atoms were evaporated with another Ta boat in the same preparation chamber, with the growth rate controlled below 0.05 ML/min. In our experiments, the Mn-coordinated SDA networks were fabricated by depositing Mn atoms on the SDA precursor layer, followed by annealing the sample to about 420 K. Another way for network fabrication is to deposit Mn first and then SDA molecules, and further sample annealing should be performed at about 420 K. All STM measurements were conducted subsequently at room temperature. Results and Discussion We first investigated the assembly behavior of the SDA molecule on the Au(111) surface. The inserted pictures in Figure 1a show the high-resolution STM image and the molecule structure of SDA with an axial length of about 1.55 nm. The high-resolution image reveals a dumbbell shape of SDA molecules, the two bright spots should correspond to the benzene

10.1021/jp074925d CCC: $37.00 © 2007 American Chemical Society Published on Web 10/19/2007

Stillbenedicarboxylic Ligand Supramolecule

J. Phys. Chem. C, Vol. 111, No. 45, 2007 16947

Figure 1. (a) Large-scale STM morphology showing SDA assembly on Au(111), with a high-resolution image and a molecule structure inserted. (b) Patterning of Mn clusters at the elbow sites of the chevron reconstructions of Au(111). The underlying broad and dark stripes correspond to the domains with surface atoms in fcc and hcp array, respectively.

Figure 2. (a) Mixed surface characterized with a hydrogen-bonded SDA phase and Mn-coordinated SDA compounds with irregular shape structures. (b) Formation of Mn-coordinated networks with a complete complexation of Mn islands after thermal annealing. A long-range modulation from the substrate herringbone can be obtained on the networks. The included area in the dashed circle shows the defects in network surface.

rings, and the connecting node is likely the ethylene group. Large-scale morphology of SDA molecules is illustrated by the STM image in Figure 1a, where stripped domains with straight boundary or single (double) molecule rows can be observed to be the typical feature of the surface. The driving force for such molecule assembly is expected to be the formation of head-totail and lateral hydrogen bonds as reported in a similar system.19,20 Our result is a little bit different from a similar assembly of SDA on Cu(100). The SDA molecules were found to exist as deprotonated dicarboxylate species on the Cu(100) surface, and the intermolecular coupling can be rationalized as the following: carboxylate groups point to the C-C single bond to form hydrogen bonds with a phenyl hydrogen and an ethenylene hydrogen.21,22 Thus, the SDA molecules accumulate into 2D supramolecular networks on a relatively reactive Cu(100) surface. For our experiment with Au(111) as the substrate, the carboxylate groups remain unaffected and the H bonds are decisive for the 2D assembly of SDA molecules into stripped islands. As we know, the (22 × x3) reconstruction of Au(111) is induced by a local contraction of the surface layer along the [11h0] direction with the contraction direction periodically rotated by 120°; thus, mesoscopic reconstruction domains with chevron

(or herringbone) patterns can be observed from STM observations. Because the dislocations at the elbows of the herringbone structure can provide preferential nucleation sites, which is typical of patterning of transitional metal clusters involved with Fe, Co, Mn, and so forth.23 In our experiments, the fabrication of such clusters is realized by a deposition of 0.1 ML Mn on Au(111), and corresponding STM observation is illustrated in Figure 1b. In the fabrication of Mn-coordinated networks, a SDA precursor layer was first prepared with coverage less than 0.5 ML, and enough bare Au surface should be left for the inhabitant of networks. After Mn deposition on the SDA-decorated Au(111) surface, our STM observation reveals that the surface is greatly changed. The picture in Figure 2a exhibits the complicated morphology where the bright spot surrounded by some irregularly arranged SDA molecules should correspond to the residual Mn cluster. Triangular or rectangular structures near the Mn cluster may imply a limited interaction of the Mn atoms with the carboxylic groups of the adjacent SDA molecules. The Mn atoms at room temperature are not mobile enough to form saturated coordination bonds with SDA molecules; therefore, no long-range ordering of the networks can be obtained under

16948 J. Phys. Chem. C, Vol. 111, No. 45, 2007

Zhang et al.

Figure 3. (a) Detailed sturcture of the networks coexisted with R and β phase, with rectangular (the axis of the Mn atoms parallel to the adjacent units) and square shapes (the axis of the Mn units perpendicular to the adjacent ones). (b and c) Schemetic molecule models of R and β phase. The unit cells of the Mn-Mn arrays are also indicated, respectively.

the current conditions. On the upper part of Figure 2a, the SDA striped phase bears indiscernible influence from the adjacent Mn islands. To improve the mobility of the Mn atoms and the SDA molecules, thus to increase the reactivity of the coordination interaction, we then anneal the sample to about 420 K for 2 min, which transforms the complicated surface (Figure 2a) into the one as indicated in Figure 2b. Mn-coordinated SDA networks with a macroscopic uniformity are clearly observed but with minor defects, which is reflected with the missing of SDA molecules or the twisting of the near-rectangular shape of the networks (shown in the dashed circles). Interestingly, we become conscious of the substrate herringbone structure on the nanogrid surface, and this is expected to imply modulation effects on the specific configuration of the networks. The details of this effect will be discussed in the latter part of the paper. A series of STM images are displayed in Figure 3. We can obtain two phases, and they can be categorized into two types of network geometries: a rectangular or a near-square geometry. Further high-resolution STM observations indicate that the difference for the two network shapes is derived mainly from the distinct arrangement of the two Mn atoms at the network node. The axis of the node Mn unit presents two sets of configurations: they are parallel or perpendicular to the adjacent Mn units. We nominate these two cases to be R and β phase, and corresponding domains are indicated in the STM image of Figure 3a, respectively. In this case, we can deduce that the stoichiometry ratio of Mn/SDA for the network formation equals ∼1:1. It is seen clearly that the two phases appear in different network domains with different orientations, which are rotated by ∼30° from each other. A further investigation of the different phases is obtained from our STM images and the related schematic molecule models. The unit cells of the various structures can be deduced from the point view of the Mn-Mn units, and they are plotted in corresponding images with a and b delegating the two vectors. For the R phase shown in Figure 3b, two parallel SDA molecules contained in a rectangular network form monodentate bonds with two Mn atoms at both ends, while the other two parallel SDA molecules bind to a single Mn atom with a bidentate bond for two carboxylic groups on both ends. Comparatively, a schematic illustration of the β phase is shown in Figure 3c, where each single SDA molecule forms monodentate bonds with two Mn atoms at one end and a bidentate bond with a single Mn atom on the other end. The unit cell of the Mn-Mn array for the two phases should correspond to a rectangular and a square shape, with a ) 21.4 Å and b ) 17.3 Å for the R phase and a ) b ) 19.3 Å for β phase. If we introduce the Mn-O

distance to be 2.0 Å, which falls in the range of the typical Mn-O metal organic coordination bond length (1.96-2.2 Å),24 then the Mn-Mn spacing can be deduced from our STM results and proven to be 4.0-4.2 Å. These parameters are comparable with some recent results about the Fe-TPA, Fe-BDA, and FeTDA network systems.18 Although we propose a model of dimerized Mn atoms, it should be noted that it is a tentative model and further investigations by theoretical calculations or by more-specific STM observations are necessary. Usually, we cannot acquire a large domain of a single phase with a size more than 20 × 20 nm2. A mixing of different phases with different orientations, or only irregular arrays of the networks with quasi-rectangular configurations dominate the morphology of the Mn-coordinated SDA networks. It is worth noting that the different configurations of the node Mn atoms, with parallel or a perpendicular geometries in connecting the SDA molecules, are isometric states if we judge from the point view of surface adsorption. In a recent publication, SDA-based metal-organic networks have been realized on Cu(100), where the coordination of Fe atoms with the deprotonated carboxylato groups build up infinite tetragonal network structures. The reported Fe-coordinated SDA networks are a little bit distorted and with a small domain size.25 The difference between this result and our study are derived from the different substrate and the node metal atoms. For the Au(111) substrate used in our experiment, it has weak chemical reactivity and low surface corrugation, which provide an inert surface. The carboxylato groups of the SDA molecules cannot be deprotonated before their coordination interaction with the Mn atoms. The intrinsic reactivity of the Mn complexation reactions should be expressed on the ideal Au(111) surface. All of these are expected to be the reasons that we can acquire larger single domains of Mncoordinated SDA networks on the Au(111) reconstructed surface. In some related work from other groups, it has been suggested that the driving force for the network formation derives mainly from the coordination interaction between the carboxylic groups of organic molecules and the node metal atoms.17 Meanwhile, the specific adsorption sites of the metal atoms and the adsorption geometry of the aromatic rings from the organic molecules, which are determined by the reconstruction of the substrate surface, are the other two factors that will influence the network configuration.18 In our experiment, the adopted Au(111) substrate is comprised of uniaxial domains of alternating face-centered cubic (fcc) and hexagonal close-packed (hcp) stacking of the surface atoms. The discommensuration regions or the surface dislocations related with an inhomogeneous

Stillbenedicarboxylic Ligand Supramolecule

J. Phys. Chem. C, Vol. 111, No. 45, 2007 16949

Figure 4. (a) Periodic array of the Mn-based SDA networks modulated by the Au(111) herringbone structure, with the dashed lines separating the hcp and fcc regions. (b) High-resolution STM image of a with the schemetic model superimposed (balls: Mn atoms).

Figure 5. (a) Ladder-shape networks appearing on the SDA oversaturated surface, (b) Specific configuration for the ladder structure with two Mn atoms coordinating with three SDA molecules.

substrate are expected to imply strong influence on the nucleation of Mn atoms and thus the network fabrication. Relatively high-resolution STM images of the networks are captured in Figure 4 to illustrate the detailed arrangement of the center Mn atoms. The dashed lines in Figure 4a indicate the boundaries of fcc and hcp regions of Au(111), and the red balls demonstrate the position and the geometry of the Mn units. In this image, the networks present no clear character of a single phase (R or β phase) as described above, while irregular or quasi-rectangular shape networks build up the morphology of the surface. Interestingly, we can notice a commensurate ordering of the networks with the substrate herringbone, that is, the axis or the geometry of the Mn units in the network node present a periodic array, according to the different reconstruction regions of the hcp or fcc parts. A clear image is shown in Figure 4b, if we use (s) to indicate the parallel and (|) to delegate the perpendicular arrangement of the Mn unit axis. In the fcc region, we found a | | s s array of the Mn unit axis and a | s geometry in the hcp region. This periodic alternation of the Mn unit axis, and thus the particular configuration of the networks, can be repeated by several periods along with the herringbone structure. To know about the periodic alternations of the Mn unit axis, the identification of the possible locations of the energy minimum on Au(111) may be the first step because they can directly determine the nucleation sites of Mn atoms. A related theoretical calculation has been performed about a single Co atom adsorption on Au(111), where five types of adsorption sites can be observed: (1) two ternary sites exist on fcc or hcp

stacking in the commensurate regions; (2) two pseudoternary sites appear on each side of the discommensuration line; (3) a most stable adsorption site derives from a bridge position in the highest part of the line defect.26 Although these are only theoretical calculations for a single atom adsorption, it might correspond well with our experimental results because five typical Mn dimers can be observed on top of an hcp and fcc alternation. The reconstructed surface of Au(111) is very complicated, here we cannot give a visual theoretical model to illustrate the Mn unit axis alternations on the hcp and the fcc regions. The specific adsorption sites for the two Mn atoms in the network node should correspond to a ternary site and a nextneighboring bridge site, with the Mn-Mn distance estimated to be 4.0 Å. Anyhow, the perpendicular or parallel geometry of the Mn unit is expected to be influenced by the various adsorption energies for a single atom or Mn dimer adsorption, thus depending strongly on the Au relaxation effects and the segregation behavior involved with the first and sublayer of Au(111). In our experiments, the long-range alternations of the network configurations provide us with a clear manifestation that the substrate reconstructions can also play an important role in deciding the specific structures of metal organic coordination networks. Another type of network nominated as a ladder-type structure appears when a shortage of the deposited Mn atoms or a surplus of the SDA molecule occurs for the network fabrication. A representative feature of the surface is illustrated in Figure 5a and a superimposed molecule model is shown in Figure 5b.

16950 J. Phys. Chem. C, Vol. 111, No. 45, 2007 One array of networks connected by two SDA striped domains illustrates this ladder structure, and a superimposed schematic graph shows a stoichiometry ratio for SDA/Mn ) 3:2. Because this ladder structure depends strongly on the ratio of SDA versus Mn under specific surrounding conditions, it is generally not easy to acquire this structure in a macroscopic uniformity. Anyhow, this new network indicates a versatile coordination interaction between the Mn atoms and the SDA molecules. In conclusion, we have succeeded in fabricating novel Mncoordinated SDA networks on the reconstructed Au(111) surface. The formation of the different network phases according to the specific arrangement of the Mn units is identified. For the large-scale configuration of the networks, they are strongly influenced by the substrate herringbone structure, where periodic alternations of the network configurations can be observed above the fcc and hcp regions of the Au(111) surface. For a SDA oversaturated surface, a novel ladder-type structure corresponding to a stoichiometry ratio of SDA/Mn ) 3:2 is observed, thus revealing a different binding nature of the Mn atoms. This fabrication of the Mn-coordinated SDA networks is expected to be an ideal system for the investigation of other important phenomena, such as host-guest interactions, molecule recognition, and catalysis.14,27-29 Acknowledgment. This research was partially supported by the Ministry of Education, Science, Sports and Culture, Grantin-Aid for Scientific Research on Priority Areas, 448, 2005. References and Notes (1) Yokoyama, T.; Yokoyama, S.; Kamikado, T.; Okuno, Y.; Mashiko, S. Nature 2001, 413, 619-621. (2) Bo¨hringer, M.; Morgenstern, K.; Schneider, W.; Berndt, R.; Mauri, F.; De Vita, A.; Car, Roberto. Phys. ReV. Lett. 1999, 83, 324-327. (3) Barth, J. V.; Weckesser, J.; Cai, Cheng Zhi; Gu¨nter, P.; Bu¨rgi, L.; Jeandupeux, O.; Kern, K. Angew. Chem., Int. Ed. 2000, 39, 1230-1234. (4) (a) Holliday, B. J.; Mirkin, C. A. Angew. Chem. 2001, 113, 20762078. (b) Angew. Chem., Int. Ed. 2001, 40, 2022-2043. (5) Swiegers, G. F.; Malefetse, T. J. Chem. ReV. 2000, 100, 34833537. (6) Yaghi, O. M.; O’Keeffe, M.; Ockwig, N. W.; Chae, H. K.; Eddaoudi, M.; Kim, J. Nature 2003, 423, 705-714. (7) (a) Coperet, C.; Chabans, M.; Saint-Arroman, R. P.; Basset, J. -M. Angew. Chem. 2003, 115, 164-191. (b) Angew. Chem., Int. Ed. 2003, 42, 156-181.

Zhang et al. (8) Nozaki, C.; Lugmair, C. G.; Bell, A. T.; Tilley, T. D. J. Am. Chem. Soc. 2002, 124, 13194-13203. (9) Srikanth, H.; Hajndl, R.; Moulton, B.; Zaworotko, M. J. J. Appl. Phys. 2003, 93. 7089-7091. (10) Lehn, J.-M. Supramol. Chem.-Concepts and PerspectiVes; VCH: Weinheim, 1995; Chapter 9, p 200. (11) (a) Barth, J. V.; Costantini, G.; Kern, K. Nature 2005, 437, 671679. (b) Barth, J. V.; Weckesser, J.; Lin, N.; Dmitriev, A.; Kern, K. Appl. Phys. A 2003, 76, 645-652. (12) Dmitriev, A.; Spillmann, H.; Lin, N.; Barth, J. V.; Kern, K. Angew. Chem., Int. Ed. 2003, 421, 2670-2673. (13) Spillmann, H.; Dmitriev, A.; Lin, N.; Messina, P.; Barth, J. V.; Kern, K. J. Am. Chem. Soc. 2003, 125, 10725-10728. (14) Stepanow, S.; Lingenfelder, M.; Dmitriev, A.; Spillmann, H.; Delvigne, E.; Lin, N.; Deng, X.; Cai, C.; Barth, J. V., Kern, K. Nat. Mater. 2004, 3, 229-233. (15) Dmitriev, A.; Spillmann, H.; Lingenfelder, M.; Lin, N.; Barth, J. V.; Kern, K. Langmuir 2004, 41, 4799-4801. (16) Classen, T.; Fratesi, G.; Costantini, G.; Fabris, S.; Stadler, F. L.; Kim, C.; Gironcoli, S. de.; Baroni, S.; Kern, K. Angew. Chem., Int. Ed. 2005, 44, 6142-6145. (17) Clair, S.; Pons, S.; Fabris, S.; Baroni, S.; Brune, H.; Kern, K.; Barth, J. V. J. Phys. Chem. B 2006, 110, 5627-5632. (18) Stepanow, S.; Lin, Nian.; Barth, J. V.; Kern, K. J. Phys. Chem. B 2006, 110, 23472-23477. (19) Barth, J. V.; Weckesser, J.; Trimarchi, G.; Vladimirova, M.; De Vita, A.; Cai, Cheng Zhi.; Brune, H.; Gunter, P.; Kern, K. J. Am. Chem. Soc. 2002, 124, 7991-8000. (20) Zhu, N.; Osada, T.; Komeda, T. Surf. Sci. 2007, 601, 1789-1794. (21) Lin, N.; Dmitriev, A.; Weckesser, J.; Barth, J. V.; Kern, K. Angew. Chem., Int. Ed. 2002, 41, 4779. (22) Stepanow, S.; Lin, N.; Vidal, F.; Landa, A.; Ruben, M.; Barth, J. V.; Kern, K. Nano Lett. 2005, 5, 901-904. (23) (a) Chambliss, D. D.; Wilson, R. J.; Chiang, S. Phys. ReV. Lett. 1991, 66, 1721-1724. (b) Voigtla¨nder, B.; Meyer, G.; Amer, N. M. Phys. ReV. B 1991, 44, 10354-10357. (c) Fonin, M.; Dedkov Y. S.; Rudiger, U.; Guntherodt, G. Surf. Sci. 2003, 529, 275-280. (24) King, R. B. Ed. Encyclopedia of Inorganic Chemistry; Wiley: Chichester, 1994. (25) Lin, N.; Stepanow, S.; Vidal, F.; Kern, K.; Alam, S. M.; Stromsdorfer, S.; Dremov, V, Muller, P.; Landa, A.; Ruben, M. Dalton Trans. 2006, 2794-2800. (26) Goyhenex, C.; Bulou, H. Phys. Rev. B 2001, 63, 235404. (27) Davis, M. E. Nature 2002, 417, 813-821. (28) Sotein, A. AdV. Mater. 2003, 15, 763-775. (29) Stepanow, S.; Lin, N.; Barth, J. V.; Kern, K. Chem. Commun. 2006, 2153-2155.