Nanorobotic Manipulation of Microspheres for On-Chip Diamond Architectures** By Florencio García-Santamaría, Hideki T. Miyazaki, Alfonso Urquía, Marta Ibisate, Manuel Belmonte, Norio Shinya, Francisco Meseguer,* and Cefe López* Controlled manipulation of particles can be used for the construction of macroporous structures. This would find an echo in different fields depending on the size of the particles involved. For instance, if the periodicity of the lattice is com- parable with optical wavelengths, the crystals may show re- markable photonic effects. [1,2] Alternatively, for sizes around several tens to hundreds of micrometers, materials useful for tissue engineering might be found. [3] Indeed, porosity, pore size, and opening diameter of connecting windows are funda- mental parameters for the control of cell or tissue induction within scaffolds [4] that can be made of biocompatible polymer blocks. [5] A major target in materials science is the production of three- dimensional (3D) microporous structures that are periodic at an optical scale and have appropriate symmetry. Conventional li- thography can be used to produce complex small-feature struc- tures, not only in two dimensions. [6,7] Holographic lithography is quickly developing and has a great potential. [8] Colloidal tech- niques are attractive because they can produce very stable and robust structures with periodicity in the micrometer and submi- crometer region. However, fabricated colloidal crystals have insufficient control of symmetry and produce mostly close- packed lattices. [9] Epitaxial growth of crystalline structures on patterned surfaces, gravitationally [10] or entropically driven, [11] and microfluidics [12] with layer-by-layer growth [13] present novel opportunities to build non-close-packed structures. The trapping of micrometric spheres by radiation pres- sure [14,15] or manipulation of nanometer-sized particles with an atomic force microscope [16] have attracted attention in recent decades. The utilization of a nanorobot [17] attached to a scanning electron microscope (SEM) is particularly suitable for building 3D structures from microscopic objects of size above 100 nm. The probe can be controlled with high accu- racy (a few nanometers) and used to pick and position the particles. This happens because electrostatic and van der Waals forces dominate the dynamics of micro-objects. [18] However, assembling particles is subjected to a limitation. Regardless of lattice parameter or feature size, there is a con- dition that must always be fulfilled: the location where the particle is to be placed has to be mechanically stable. This is usually a problem when highly porous structures are used. Our work uses a sacrificial scaffold that can be eliminated when the structure has been assembled. In this communication we describe the use of a nanorobot to assemble diamond-lattice crystals from silica microspheres, as an example of how our method can be applied. Such a ma- terial has potential as a biocompatible scaffold (for larger lat- tice parameters). Also the photonic bandgap community is very interested in this sort of structure because it can easily be made to sustain a full photonic bandgap. [19] The diamond lat- tice is a non-close-packed arrangement and, as a consequence, colloidal methods have so far failed to produce it. Inserting latex spheres that act as a temporary supporting scaffold has allowed us to build diamond-type lattices of silica micro- spheres by placing single microspheres in designated locations with the use of a nanorobot attached to an SEM. This method potentially enables fabrication of structures including line and point defects for optical circuitry. [20] Finite systems (as op- posed to infinite crystals) are of great interest for theoretical approaches to the build up of photonic bands and as techno- logical components of future photonic devices. [21,22] In particu- lar, Ozin has invented [23] and adapted [24] techniques that en- able the formation of planar opal-based microphotonic crystal chips with controlled shape, size, and orientation. The idea of stacking spheres in a diamond lattice has been proposed recently. [25] It is conceptually based on two simple facts: first, a body centered cubic (BCC) lattice is formed by two interpenetrating diamond lattices, [26] and second, the BCC lattice is easier to grow since, in it, sphere sites are stable. Therefore, the proposed route comprises the assembly of a BCC lattice of mixed silica and latex spheres of equal di- ameter (mixed body centered cubic, mBCC hereafter) and subsequent removal of the latex set. Silica spheres would sit in the positions of one of the interpenetrating diamond lattices while latex ones would lie on the other sub-lattice. The result- ing structure, after latex removal, is a diamond lattice of silica spheres. To test the stability of the structure after latex removal, ten- tative samples were prepared using silica and latex submicro- meter spheres. The beads were placed on a silicon substrate in a face centered cubic (FCC) disposition in two different orien- tations (Fig. 1a and b). The next step was to remove the latex spheres without disturbing the silica ones. Latex calcination was discarded for this purpose because it produced liquid la- tex, which dragged the silica beads by surface tension and the structure collapsed. Therefore, a more gentle method was used: oxygen plasma etching [27] at 65 mW cm ±2 and 30 Pa. The 1144 Ó WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim, 2002 0935-9648/02/1608-1144 $ 17.50+.50/0 Adv.Mater. 2002, 14, No. 16, August 16 COMMUNICATIONS ± [*] Dr. C. López, Prof. F. Meseguer,F. García-Santamaría, M. Ibisate Instituto de Ciencia de Materiales de Madrid (CSIC) Cantoblanco, E-28049 Madrid (Spain) and Unidad Asociada UPV-CSIC Camino de Vera s/n, E-46022 Valencia (Spain) E-mail: cefe@icmm.csic.es, fmese@fis.upv.es Dr. H. T. Miyazaki, Prof. N. Shinya National Institute for Materials Science 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047 (Japan) Dr. A. Urquía, Dr. M. Belmonte Agere Systems Espaæa Polígono Industrial de Tres Cantos s/n, Tres Cantos E-28760 Madrid (Spain) [**] This work was partially supported by the Spanish CICyT project MAT2000-1670-C04 and the European Commission Project IST-1999- 19009 PHOBOS. We thank J. A. Peinador, J. P. Gonzalez, G. Sanchez Plaza, M. Holgado, L. Muæoz, and J. de la Hoz for their help with the fabrication of a patterned substrate on a silicon wafer. We are grateful to T. Sato and H. Morishita for support.