Geometry optimization future developments

Abstract This contribution focuses upon the application of evolutionary algorithms to the non-deterministic polynomial hard problem of global cluster geometry optimization. The first years of method development in this area are sketched briefly followed by a characterization of the current state of the art by an overview of recent application work. Strengths and weaknesses of this approach are highlighted by comparison with alternative methods. Last but not least, current method development trends and desirable future development directions are summarized. 

The remainder of this review is outlined as follows. The historical method development of EA use for global cluster geometry optimization is briefly recalled in Sect. 2. An overview of typical application work in recent years is provided in Sect. 3. In Sect. 4, we take a side glance at other methods to tackle the same and related problems, and briefly discuss advantages and disadvantages of some the prominent alternatives to EAs. Finally, in Sect. 5 recent method development work is summarized, and we try to give some (personal, biased) opinions on which open questions such developments should address in the future. 

More recently, as computing power has continued to increase along with cost efficiency, and with the development of methods for geometry optimization in ab initio calculations, the latter have become much more pervasive. With large basis sets, including electron correlation, one can now study reasonable size molecules by rather high level ab initio calculations. This trend will only increase in the future. 

We hope that this brief review has given the reader a general feeling of the development and application of CE in the separation of nucleic acids. With the advent of capillary array electrophoresis and microchip electrophoresis, as well as remarkable improvements in separation matrices, CE has become a standardized and cost-effective technique in the separation of nucleic acids. Novel thermo-responsive polymer solutions combine the merits of different monomers, and offer the possibility to fine-tune the desirable properties of polymer molecular architecture and chemical composition. Artificial entropic trapping systems obviate the use of viscous polymer solutions, and even offer fast, unattended, miniaturized, and multiplexed platforms. Optimizing the geometry of these electrophoretic systems to both increase the separation and reduce the diffusion (band broadening) is the main topic for future research. 

One future prospect for microfluidic control will be the development of tailor-made vesicles with more complex geometries, triple, quadruple, multiple emulsions and artificial cells (artificiells), i.e. liposomes containing hposomes with each a different functionality, closely mimicking a biological cell structure, all designed for optimal performance for the specific function demanded from the vesicle. Double emulsions had already been made in a microfluidic device in a very controlled way [60, 61] (Figures 19.7 and 19.8). The number of secondary droplets inside the primary droplet... 

---