Table of contents

Existing capabilities in chemistry and scanning probe microscopy make it feasible to undertake the development of instruments providing positional control of sequences of chemical reactions. Once developed, such instruments will enable the construction of nanoscale systems with precise molecular control. Improvements to previously proposed positional synthesis mechanisms are suggested, as are means for attaching protein molecules to surfaces with substantial rigidity. The latter capability would facilitate the development of positional synthesis systems by providing a means for binding and positioning reagent molecules in an atomic force microscope mechanism.

An adhesive interaction between a silicon nitride AFM tip and glass substrate in water is described. This adhesion is in the range 5-40 nN, of which a large component is likely to be due to hydrogen bonding between the silanol groups on both surfaces. The interaction can be modulated by a variety of buffers commonly used in biochemical and biological research, including sodium phosphate, tris(hydroxymethyl)aminomethane, glycine, and N-2-hydroxyethyl-piperazine N'-2-ethanesulfonic acid. Using these buffers it appears that there are effects of ion concentration, ion type and pH on the measured adhesion. Of the conditions examined, phosphate was most effective at reducing adhesion and could be used at concentrations as low as 10 mM at neutral pH. The results demonstrate that the chemical interactions between tip and sample can be modulated, and provide a basis for designing conditions for imaging and manipulating biological molecules and structures.

The bacterial flagellar motor is a molecular machine that converts chemical energy stored in a transmembrane proton gradient into the mechanical work of rotation. Knowledge of the structure, genetics and dynamics of this motor has increased steadily. Recent progress is reviewed, with an emphasis on the identification of parts most likely to be involved in torque generation and on the assignment of specific mechanistic roles to these parts. An explicit hypothesis for torque generation is described, and some tests of this hypothesis are proposed.

The major research objectives in molecular nanotechnology are the design, modeling, and fabrication of molecular machines and molecular devices. While the ultimate objective must clearly be economical fabrication, present capabilities preclude the manufacture of any but the most rudimentary molecular structures. The design and modeling of molecular machines is, however, quite feasible with present technology. More to the point, such modeling is a cheap and easy way to explore the truly wide range of molecular machines that are possible, allowing the rapid evaluation and elimination of obvious dead ends and the retention and more intensive analysis of more promising designs. While it can be debated exactly how long it will take to develop a broadly based molecular manufacturing capability, it is clear that the right computational support will substantially reduce the development time. With appropriate molecular CAD software, molecular modeling software (including available computational chemistry packages, e.g., molecular mechanics, semi-empirical and ab initio programs) and related tools, one can plan the development of molecular manufacturing systems on a computer.

While computational chemistry can be used to predict properties of existing and hypothetical structures, molecular engineering remains an expensive, time consuming and largely experimental challenge. With the aim of working towards reducing the high cost of molecular engineering, a computational counterpart to experimentation and synthesis is presented which utilizes mixed-integer nonlinear programming. The optimization formulation uses a superstructure representing all candidate molecules to determine group contributions to functionality. Furthermore, partial experimental or synthetic knowledge can be incorporated directly in the problem formulation. Three example problems are solved: (i) design of sulphide molecules with optimal selectivity, (ii) design of sulphide molecular switches and (iii) design of optimal binding amido groups. The key contribution of this article is a framework for automated molecular design which integrates independent sources of mathematical and experimental knowledge.

An approach to molecular nanotechnology is presented that utilizes branched DNA molecules to form stick figures. These branched molecules are directed to associate in a particular fashion by utilizing the addressability of DNA sticky ends. The sequences of molecules are designed using an algorithm that minimizes sequence symmetry. This system has been used to construct macrocycles, a DNA quadrilateral, a molecule with the connectivity of a cube and a single-stranded DNA knot. Possible uses for this system include the construction of macromolecular zeolites for structure determination by diffraction, the assembly of new catalysts, the solubilization and delivery of drugs, and the assembly of molecular electronic devices.

A preliminary study has been carried out to evaluate the possibility of patterning biological molecules on surfaces for scanning tunnelling microscopy and atomic force microscopy. The approach uses the concept of treating surfaces that are suitable for semiconductor lithography with bridging molecules that can be patterned to micrometre and nanometre scales. Indium tin oxide and flame-treated gold have been selected as the immobilization surfaces, and initial results indicate good anchorage of proteins for scanning tunnelling microscope imaging using the bridging procedure developed.