Foresight Institute Stage 2: Implementation

Archival Page—See updated Roadmap Project

The proposals described on this page were developed in late 2003 and announced in Foresight Update 53. Since then, Foresight has focused its technical work on the International Technology Roadmap for Productive Nanosystems. If either the Roadmap project or the project proposal below interests you, please contact

Speeding the development of molecular nanotechnology

By Ralph C. Merkle
Distinguished Professor
Georgia Tech College of Computing
and Robert A. Freitas, Jr.
Research Scientist, Zyvex Corp.

Proposed Budget: $5 million over five years


Molecular nanotechnology (MNT) is a projected manufacturing technology that should be able to inexpensively arrange atoms in most of the ways permitted by physical law. Besides computers several orders of magnitude more powerful than any that exist today, MNT will also permit inexpensive construction of light, strong, smart materials that will revolutionize aerospace and other applications, and molecular medical devices that will revolutionize medicine. Inexpensive solar cells and batteries will eliminate our dependence on coal, oil and nuclear fuels. Critical performance parameters for almost all manufactured products will see major and often revolutionary improvements that will fundamentally alter our relationship with the material world.

The next step

The most important next step in achieving MNT is to provide clear and accurate descriptions of the fundamental chemical reactions that will be involved. The claim is that molecular tools with chemically specific tip structures can be used, sequentially, to modify a work piece and build a wide range of molecular structures. This process, called "mechanosynthesis," is readily accessible to theoretical and computational investigation. Indeed, several papers have already been published describing specific tools and specific chemical reactions (a bibliography is available). Perhaps the best studied reaction is site-selective hydrogen abstraction using a positionally controlled radical (see, for example,

The proposal

The proposal is basically simple: validate the feasibility of mechanosynthesis by analyzing in specific detail some relevant molecular tools and positional synthesis operations, using computational chemistry to model and verify the feasibility of the unit operations.

MNT proposes to arrange elements drawn from the entire periodic table into useful structures. Here, to make the analysis tractable, we propose to analyze only the synthesis of stiff hydrocarbons—structural components made from carbon with surfaces terminated by hydrogen. Other elements will be introduced only sparingly, and only when necessary to support basic catalytic functions (see, for example, A proposed "metabolism" for a hydrocarbon assembler).

Stiff hydrocarbons include graphite, diamond, fullerenes, nanotubes, and a host of other useful structures that, by themselves, would be quite sufficient to revolutionize manufacturing.


The purpose of this research is to influence further research, particularly experimental research. That is, a theoretical study, regardless of how penetrating or insightful, will not let us build molecularly precise products. However, it can influence the research climate and redirect research funding and targets. Today, there is no well-funded project with the explicit aim of developing MNT. A significant reason for this absence is the lack of a clear, well-defined target whose feasibility has been well established by appropriate research.

Because this research is aimed at influencing further research and funding, a fundamental objective must be the dissemination of the results in a form that can be both readily understood and which has a high degree of credibility. The credibility will be generated by replication of the computational results using multiple computational chemistry tools, along with peer review and publication in generally accepted journals on computational chemistry. The dissemination is best done by an organization with a track record in communicating the concepts of MNT to a wider audience. The organization with the best and longest track record in this regard is the Foresight Institute. While the precise percentage of the overall effort that should be devoted to disseminating the results can be debated, it should be noted that many successful organizations devote as much as half of their gross revenues to sales and marketing. Allocating 20% of project resources explicitly to education would seem reasonable.


The deliverables from this work will be:

  1. The specifications for a set of molecular tools which, taken collectively, will be able to synthesize a wide range of stiff hydrocarbons.
  2. The description of a set of molecular machine components made from stiff hydrocarbons that can be synthesized by the use of the aforesaid tools (including details of the synthetic sequence).
  3. Papers submitted to the relevant computational chemistry journals (and possibly to some journals with a broader reach) describing the specifics of the molecular tools and chemical reactions involved. Each of these papers will form a small piece of the larger whole, and will describe and analyze some specific reaction on some specific surface in such detail that the feasibility of that reaction will be generally accepted.
  4. Graphical images (suitable for television or other media coverage as well as for talks to both technical and more general audiences) showing the various reactions, along with the sequences required to make some selected molecular machine components.

At the present time, the best analyzed reaction is the hydrogen abstraction tool (see This tool has been analyzed by multiple computational chemistry programs, all producing very similar results. We will need to provide a similar depth of analysis for at least six to ten tools to develop a minimal set of tools able to synthesize stiff hydrocarbons, and will need to explore perhaps five to ten times as many tools (i.e., 30 to 100) because many tool proposals will simply fail to work as desired, requiring the development of additional tool proposals.

The minimal effort required to achieve credibility might involve a team of 4-5 people working for several (perhaps five) years. Depending on the exact overhead being assumed (institutional support/overhead, computational requirements, etc) the burn rate is likely to be ~ $1,000,000 per year. Note that the caliber of researchers involved must necessarily be high to produce the kind of persuasive results that are the goal of this project. It is worth noting that more people working on the problem will be able to speed development in an almost linear fashion: there will be many different tools, and different sub-teams can analyze different tools entirely independently of each other. In other words, more money implies more results more rapidly.

The objective is simple: once there is an easily understood video of the relevant synthetic steps, backed up by in depth technical analysis and published articles in appropriate journals, there should be sufficient acceptance of the basic feasibility of mechanosynthesis to allow the raising of additional funds from a wider set of funders to support the next steps in the development process.


The single greatest obstacle to the development of MNT is a clear and credible description of the molecular tools and the site-specific chemical reactions involved in mechanosynthesis. This clear and credible description can be achieved in a few years, using existing off-the-shelf computational chemistry software and a small development team.

In the absence of such a clear and credible description, the development of MNT might be delayed by decades. This is a very high payoff project.

Technical Bibliography for Research on Positional Mechanosynthesis


Foresight Institute Stage 2: Implementation