We propose a systematic method of synthesizing robust macromolecules of any desired shape with precise and total structural control. It involves assembly of 3D objects from a few fundamental and carefully designed ``molecular building blocks'' in a "treelike leaf-to-root" fashion, followed by a later "global rigidizing" reaction.

We propose three different sets of abstract primitive operations, (which the building block molecules must support) and show that each suffices for essentially universal synthetic power. We present a mathematical theory of "assembly" including apolynomial time algorithm to find an "optimal" (e.g. with maximum possible synthetic yield) "tree sliced iso-oriented" assembly of any "lattice animal." Finally, we present designs of actual molecules and chemistry to show that all of our required primitive operations should indeed be achievable compatibly. Analysis suggests that the major limit on all this will be imperfect chemical specificity.

This idea is still in its early stages and will need further investigation and development, especially by synthetic organic chemists, to create building blocks with the right properties.

This method may provide a way to achieve something similar to K.Eric Drexler's notions of "nanosystems" (popularized in several books authored or coauthored by Drexler). Although I think much of the "nanosystems" area is closer to science fiction than to science, the ideas in the present paper, especially if it is possible to "raise" them to "higher levels" of the "size hierarchy," may provide a way to realize some science fiction visions with some resemblance to Drexler's. In particular, perhaps they may eventually lead to ultrafast and/or ultrasmall computers. In the nearer term, these techniques should lead to micromold or microstencil techniques for creating extremely small structures, and the creation of small quantities of: photonic bandgap materials, highly controllable "molecular sieve" materials, and zero density solids.

Keywords:Nanosystems, dendrimers, molecular sieve, low density solids, micromolds, microstencils, photonic bandgap, molecular tinkertoys, automated synthesis.