1.5.1. Overview of Part I Chapter 2 summarizes classical scaling laws for mechanical, electrical, and thermal systems, describing the magnitudes they imply for various physical parameters, and describing where and how these laws break down (requiring the substitution of molecular and quantum mechanical models). These relationships are provided for perspective and as an aid in making preliminary estimates of physical quantities. They are seldom used in later chapters, where most calculations proceed directly from physical principles, rather than from a scaling analysis. Chapter 3 provides an overview of molecular and intersurface potential energy functions, describing in some detail the molecular mechanics models that later chapters apply to the description of molecular machines. The concepts developed in this chapter are fundamental to much of what follows, since the potential energy function of a molecular system provides the basis for an essentially complete description of its mechanical properties. Its chief conclusion is that the limitations and inaccuracies of molecular mechanics models, although a serious handicap in solutionphase chemistry, are compatible with the use of these models in designing certain classes of molecular machinery. Chapter 4 provides an overview of various models of molecular dynamics and describes the basis for the choice of models made in later chapters. Its chief conclusion is that classical mechanics and classical statistical mechanics, with occasional quantum mechanical corrections, provide an adequate basis for analyzing most molecular mechanical systems operating at ordinary temperatures. Chapter 5 examines several classes of mechanical structures and derives relationships that describe the positional uncertainties resulting from the combined effects of quantum mechanics and thermal excitation; later chapters use these results to calculate error rates. This is the most heavily mathematical chapter in the book, but it chiefly concludes that classical statistical mechanics adequately describes the positional uncertainties of nanometerscale mechanical systems at ordinary temperatures. Only Eq. (5.4) is directly applied in later chapters. Chapter 6 examines various processes that cause transitions among ° potential wells in a nanomechanical system, including transitions that cause errors and structural damage. It describes standard theoretical models used to calculate chemical reaction rates based on potential energy functions; these are applied later in analyzing molecular manufacturing processes. Regarding damaging transitions, the chapter concludes that systems at room temperature can be designed to limit rates of damage caused by thermal, mechanical, and photochemical effects to low enough levels that the overall rate of damage is chiefly determined by the background level of ionizing radiation. Damage caused by radiation imposes major constraints on the design of large systems. Chapter 7 examines various processes that degrade mechanical energy into heat, causing frictional losses; among these are acoustic radiation, phonon scattering, shearreflection drag, phonon viscosity, thermoelastic damping, nonisothermal compression of mobile components, and transitions among potential wells that vary with time. It develops a set of analytical models applied in later chapters to estimate the power dissipated by various nanomechanical devices. Chapter 8 compares and contrasts the established capabilities of solution chemistry to those expected from mechanochemical systems, considering speed, efficiency, versatility, and reliability; it also examines an illustrative set of mechanochemical processes in detail. It concludes that a large set of mechanochemical reactions can be made extremely reliable, with error rates <10^{–15}. (Although not strictly necessary, this degree of reliability substantially simplifies molecular manufacturing.) It also concludes that mechanochemical processes can be used to construct a wide range of diamondoid structures; this motivates the consideration of diamondoid components in Part II.
Chapter 9 discusses nanoscale structural components and relates molecular mechanics models to descriptions based on bulk material properties. It develops the concept of a bounded continuum model that omits atomic detail while treating surfaces in a way that takes account of interatomic potentials. It concludes that diamondoid structures are desirable nanoscale components, that bounded continuum models can provide useful preliminary descriptions of component properties, and that a wide range of shapes can be constructed on a nanometer scale, despite the discrete nature of atoms and bonds. Chapter 10 uses molecular mechanics models and analytical models developed in Part I to describe the mechanical properties and performance characteristics of active devices such as gears, bearings, and drive mechanisms. It concludes that structures on a severalnanometer scale can serve as mechanical devices of most classes familiar on the macroscopic scale, and that these nanomechanical devices can in many instances move with negligible static friction. Models from Chapter 7 are applied to describe energy dissipation from dynamic friction. The conditions for lowfriction motion derived in this chapter are assumed as background in Chapters 11 to 14. Chapter 11 describes various subsystems of intermediate size and complexity. These include devices capable of measuring and distinguishing molecular features, stiff drive mechanisms, fluid handling systems (including walls, seals, and vacuum pumps), cooling systems, and electromechanical devices such as motors and actuators. These subsystems and their capabilities are applied or assumed as background in Chapters 12 to 14. Chapter 12 describes nanomechanical computer systems. It starts with mechanical digital logic systems, including logic gates, signal transmission channels, registers, and their integration into finitestate machines (with an analysis of thermally induced error rates). It then discusses carry chains, random access memory, tapebased storage, power supplies, clock distribution, information input and output to macroscopic systems, and overall system performance (volume, speed, and power dissipation). Its chief conclusion is that a 1000 MIPS computer can occupy less than one cubic micron and consume less than 0.1 microwatt of power. Chapter 13 describes systems for converting impure feedstock solutions into diamondoid molecular objects, using molecular concentration and purification systems, followed by specialpurpose mechanochemical systems (molecular mills) capable of performing repetitive operations efficiently, and by generalpurpose mechanochemical systems (i.e., molecular manipulators) capable of performing a complex series of operations under programmable control. A key conclusion is that, after an initial purification and ordering process, molecular assembly can be sufficiently reliable that cycles of inspection and rework to correct failures can be avoided. The conclusions of this chapter are directly exploited in the next. Chapter 14 describes molecular manufacturing systems that use purification systems, mills, and manipulators to transform an impure feedstock solution into any one of a large set of macroscopic (kilogramscale) products within a few thousand seconds. It focuses on systems integration and overall performance, and discusses a range of issues including design complexity and product cost.

Copyright © 1998 by John Wiley & Sons, Inc.