COMMUNICATIONS First Steps Towards Ordered Monolayers of Ligand-Stabilized Gold Clusters** Stefanie Peschel and G i i n t e r Schmid* Dedicated to Professor Peter Paetzold on the occasion of his 60th birthday The achievement of an ordered arrangement of chemically produced quantum dots is very important in the light of future applications in nanotechnology. Most of the work carried out in this field concerns the use of semiconductor quantum dots, since it is hoped that the transition of a semiconductor from the bulk state to the nanometer region, will give rise to novel physical properties, particularly in the field of electrooptics.['] Apart from the difficulty of synthesizing semiconductor clusters with largely uniform size and defined structure, the main problem is to achieve a two- or three-dimensional arrangement of the particles, without them touching. Numerous attempts have been made to use zeolites or porous materials as host lattices.[21Polymers have also been used to prevent a coalescence of the semiconductor particles. To date there has been no recognizable breakthrough in the practical effectiveness of these types of systems. With the synthesis of ligand-stabilized transition metal clusters a different kind of approach to quantum dots has become possible, which has advantages over conventional systems: 1. uniform particle size, 2. the stabilization by largely inert ligand shells, 3. the equidistant arrangement as a result of defined ligand shells, 4. the avoidance of coalescence processes, 5. the possibility of varying the separation between the clusters by chemical modification of the ligand shell. Figure 1 illustrates these possibilities which are decisive for the utilization of quantum properties. identical clusters ' d' ' separated by the ligand shells.'31 Investigations of clusters on the surface of pellets using scanning tunnel microscopy (STM) and scanning tunneling spectroscopy (STS) produce images which show that the ligand shells of the individual clusters are in contact, and in the current/voltage diagram there is not only the expected Coulomb barrierr4] but also, for the first time, energy fine line splitting of the electronic states in individual clusters at room t e r n p e r a t ~ r e .Therefore, ~~] ligand-stabilized metal clusters seem to fulfill the requirements for their application in the nanometer range. In contrast to the readily synthesized, but complicated workings of three-dimensional cluster arrangements in crystals and pellets, ordered monolayers on suitable substrates should facilitate electronic investigations. This applies to an even greater extent to one-dimensional cluster arrangements (quantum wires), which we are currently trying to produce. We were able to synthesize largely ordered monolayers of ligand-stabilized Au,, clusters for the first time. In order to accomplish this we used the established principle of self-assembly with the help of poIy-electrolytes.[~' However, we could not use the [Au,,(PPh,),,CI,J cluster itself but had to synthesize a derivative in which the PPh, was replaced by a PPh,(m-C,H,SO,H) ligand. To this end we used the previously reported sodium salt [ A U , , ~ P P ~ , ( ~ - C , H , S O , N ~ ) , , ~ and CI~]~~~ converted it into the free acid by ion exchange. When a freshly prepared thin layer of poly(ethy1enimine) (PEI) on a mica surface is immersed into a solution of the cluster there is a strongly attractive interaction between the sulfonic groups and the imino groups. The coating is carried out by immersion of a mica platelet for 24 hours in a molar aqueous solution of PEI. After intensive washing with distilled water and drying in vacuo, the coated platelet was then immersed in a 3 x molar aqueous solution of the Au cluster. In order to remove any additional cluster molecules sticking to the surface, the platelet was washed with a total of 0.5 L of distilled water and dried in vacuo. Because of the ready solubility of the cluster, this procedure ensures that only the strongly adhesive monolayer produced by the acidbase interaction remains. Figure 2 is a simplified illustration of the principle of the cluster/PET interaction. different ligand shells Fig. 1. The intermolecular distances and therefore the Coulomb barriers between the clusters can be modified by varying the ligand shell. oso 0 H NH oso 0 H NH oso 0 H NH oso 0 o$o H H NH NH ) Poly(ethy1enimine) Recently, we were able to answer unequivocally the question A / / / / / / / / / / / / / / / / Mica of whether besides semiconductor clusters also metal clusters could be applied as quantum dots. Gold clusters of composition Fig. 2. Schematic illustration of the interaction between cluster molecules and poly(ethy1enimine). A mica platelet (ca. 1 cm') coated with PEL was immersed for [Au,,(PPh3),,C1,] are composed of a cluster core about I .4 nm 24 h in a 3 x 10-' M aqueous cluster solution. As a consequence the rulfonic groups in diameter and a ligand shell 0.35 nm thick, which results in an of the cluster ligands interact with the imino groups of the polymer. approximately spherical particle with a total diameter of about 2.1 nm. In a close packed three-dimensional arrangement these clusters show conductivity, which is characterized by singleFigure 3 (top) shows a three-dimensional atomic force microelectron tunneling (SET) processes between the cluster cores scopic (AFM) image''] of a section of the cluster monolayer. [*] Prof. Dr. G. Schmid. DipLChem. S. Peschel Institut fur Anorganische Chemie der UniversitPt Universitdtsstrasse 5-7. D-45117 Essen (Germany) Telefitx: Int. code + (201) 183-2402 [**I This work was supported by the Deutsche Forschungsgemeinschaft. We would likc to thank the Fiirderverein der Stadt Essen [or their generous donations. We are grateful for the valuable discussions with Prof H. Ringsdorf. 1442 ?Z: VCH Vcrlq~.~gesell.schufl mhH,D-69451 Wcinheim.1995 Sections of comparable quality can be obtained from all areas of the approximately 1 cm2 surface. Shorter reaction times between the PEI layer and the cluster solution, for example six or twelve hours, gave a relatively closely packed occupation, however, the level shown in Figure 3 top occurs only after 24 hours. Numerous cross-sections show that less than 5 '/o of the surface is disordered or uncovered. 0570-0X33~9S/13/3-144~ $ 10.00+ .ZS,'O Angen. Cliem. In!. Ed. Engl. 1995. 34, N o . 13/14 COMMUNICATIONS  U. Simon, G. Schon. G . Schmid. .Inpew. Chm. 1993. 105, 264; Angew. Chcm. I n t . Ed. tii,ql. 1993, 32. 250; U. Simon, G. Schmid, G. Schon. M r r / w . Rr.s. Sol, S,~?np. Proc. 1992,272. 167: G. Schmid in Clusters and Colloids - From Theory to Applications (Ed : G. Schmid), VCH, Weinheim, 1994.  R. Houbertz, T. Feigeuspan, F. ,Mielke, U. Memmert. U. Hartrnann. U. Simon, G . SchBn. G . Schmid, Europhj..s. Le/t. 1994, 28, 641.  U. Hartinann. unpublished  G. Mao, Y Tsao, M . Tirrell, H. T. Davis, V. Hessel, H. Ringsdorf. Langmuir 1993, 9, 3461 ; L. F Chi, R. R. Johnston, H. Ringsdorf, N. Kimizuka. T. Kunirake, Thin Solid Films 1992,210,'2//. 11 1 ; G . Decher, Nachr. Chem. Techn. Lub. 1993, 41. 793; G. Decher. J. D. Schinitt. T h n So/ir/ Film.\ 1992, 210!2/1. 831; Hong, .I G. Decher, J. D. Hong. Ber. Buii.wn,qes. Ph,~,s.Chem. 1991, 95, 1430. 171 G. Schmid, N. Klein. L. Korste. U . Kreibig. D. Schonauer, Polyhedron 1988, 7, 605.  ARlS 3300, Burleigh Instruments 8.54nm 7.90nm t , 2.4nm Self-Assembly of a Ferromagnetically Coupled Manganese(I1) Tetramer"" - 7.27nm 0.OOnm 8.09nm I 1 16.17nm 24.26nm 32.34nm Fig. 3. Top: Thicc-dimensional representation of a largely ordered monolayer of [Au,,(PPh,(mC6H4S03H),2i(.IJ on poly(ethy1enimine). Bottom: Cross-section of a single cluster row. The separation between thc clusters i s 2.4 nm, which is in agreement with the calculated cluster srze. The equidistance of the clusters within a row and the resulting organization can be seen at the bottom of Figure 3 which shows a typical cross-section. The average distances of 2.4 i 0.2 nm are in very good agreement with those expected for S0,H-substituted clusters. These easily synthesized cluster monolayers can now be used to study electronic transitions in two-dimensional arrangements of quantum dots (cluster cores) separated by the ligand shells and spaced 1.4 nm apart, a distance which cannot be obtained either by the usual semiconductor technology or by the semiconductor clusters described above. Furthermore, this principle can be used for the formation of sandwich structures because the cluster monolayer can, in turn, be coated with PEL Since it should be possible to synthesize the corresponding cluster monolayers with clusters from other metals, additional possibilities are available for the generation of multilayer systems with different metals. Received: November 30. 1994 Revised version: February 2, 1995 [Z7513IE] German version: Angex. Chrm. 1995, 107. 1568 1569 Angelo J. Amoroso, John C. Jeffery, Peter L. Jones, Jon A. McCleverty,* Peter Thornton, and Michael D. Ward* We recently described the preparation of the podand ligand tris[3-(2'-pyridyl)pyrazol-l-yl]hydroborate,L . [ I 1 This ligand contains three N,N-bidentate chelating arms linked at the apical boron atom and is the first example of a podand ligand based on the tris(pyrazo1-1 -yl)hydroborate core. The cavity size wdS found to be appropriate for lanthanide(n1) ions (ionic radius x 1 A), and some 1 : 1 nine-coordinate complexes such as [Eu(L)F(MeOH),]PF, were prepared in which the lanthanide ion was located in the hexadentate ligand cavity."] We thought it would be of interest to see how the ligand behaved with firstrow transition-metal dications (ionic radius = 0.7-0.8 A), which are both smaller than lanthanides and prefer lower coordination numbers. Molecular modeling studies indicated that 1 :1 complexes with sensible metal-ligand bond lengths would involve considerable strain on the ligand; conversely if the ligand were allowed to adopt a strain-free conformation then the metal-pyridyl bond lengths would be unrealistically long (>3A). ~ Keywords: clusters . gold compounds . monolayers . nanostructures [ I ] H Weller. .lngrw. Chcvn. 1993. 105. 43: A n g m . Chem. In!. Ed. EngI. 1993, 32. 41, Adis .&fuwr. 1993. 5, SX; J. N. Fendler. Chrm. Rev. 1987. 87, 877; A. Hengk i n . TO/J C w r . Chmi. 1988, 143, 115; M. C. Steigerwald. L. E. Brus, Ann. Rev. Mu/er. Sci. 1989, 19.471,O. V. Salata. P. J. Dobson, P. J. Hull, J. L. Hutchison. A h . ,MU/CI 1994.6, 772; J. E. B. Katari, V. L. Colvin, A. P. Alivisatos, J. f h p Clicwi 1994. 98. 4109.  Y. Wang, N Herron, J f h ) , s . Chem. 1988, 92, 4988; ihrd. 1987, 91, 257. G . A. Osiin. S iizkar, A h M r r t ( ~ 1992, . 4. 11 1 ; J. Kuczyuski. J. K. Thomas. Chrm. Phi..s. Lcf!. 1985. 89.2720. ['I Prof. J. A. McCleverty, Dr. M. D . Ward. Dr. A. J. Amoroso, Dr. J. C. Jeffery, P. L. Jones School of Chemiatry, University of Bristol Cantock's Close, Bristol BSX ITS (UK) Telefax: Int. code (117) 929-0509 e-mail: wardmd(u siva.bristol.ac.uk + I**] Dr. P. Thornton Department of Chemistry, Queen Mary and Westfield College, London (UK) This work was partly funded by the EPSRC and Unilever Research. We thank Dr. J. P. Maher for recording the EPR spectrum, Dr J. Crosby and Miss Kate Byrom for recording the electrospray mass spectra. and Dr. S. G. Carling and Prof. P Day of the Royal Institution for assistance with recording the magnetic data.