Prospects of lithium liquid metal ion sources in nanometer-scale focused ion beam technology

Abstract

Focused ion bean (FIB) systems are well established tools in micro-scale material processing and surface analysis. State-of-the-art FIB systems can produce a minimum beam diameter of some tens of nanometer. Following a very recent study on the applicability of FIB for nanotechnology, such a system has to satisfy some basic requirements: (1) the use of light ions like Li, as these ions can transfer enough energy to a specimen to include a chemical reaction that result in etching or deposition on a scale of about 1 nm; (2) the need to minimize the chromatic aberration contribution to the probe size of the final lens above the specimen to less than 1 nm; (3) for observation of the fabricated structures a scanning electron beam has to be used as ions would destroy the nanostructures during observation. The production rate of the nanostructures depends on the probe current; the probe size and current are directly related to the brightness of the ion and the setup of the ion optical system. Therefore a liquid metal ion (LMIS) would be a favorite candidate, because of its high brightness and small virtual size; nevertheless, a major drawback is its energy spread of about 4.5 eV. Very recently, a new type of metal ion has been developed which may overcome the disadvantage of a rather high energy spread. The Li surface diffusion metal ion (SDMIS) is capable to emit a Li/sup +/ ion beam at a current of 10/sup -12/ A with an extremely narrow energy distribution, i.e. a FWHM of 0.25 eV. In this ion Li atoms diffuse in the solid state from the shank of a [111] oriented tungsten field emitter towards its apex, where the Li/sup +/ ions are field desorbed. Even just recently, stable field electron emission from a liquid Li cone formed on the apex of a [111] oriented W tip by field induced surface diffusion has been reported. In both cases, the initial Li supply was by means of vapor deposition from a Li perpendicular to the beam axis. Combining these forthcoming concepts and considering the demands mentioned above, first of all it appears obvious to amalgamate a Li field desorption ion for fabrication with a Li field emission electron for observation into one single hybrid field effect source which can in principle deliver both an ion or electron beam separately only by changing the polarity. Furthermore, in order to extend the operation period of the and to minimize evaporation losses, in any case it would be advantageous to improve the Li supply to the field emitter. This can be realized by a concept similar to the impregnated-electrode type LMIS which has been proposed in particular for miniaturized and microstructured liquid metal ion sources (MILMIS). Eventually, even the idea has to be considered to micro-miniaturize a single Spindt-type microvolcano field ion to an ultimate limit that depends above all on the surface tension and viscosity of liquid Li. Such a nanosource probably could deliver much higher ion currents, but at the expense of an extended energy spread.