in the form of ATP. Biomotors often involve “walking” carrier proteins along “rails” that are themselves biopolymers, such as tubulin or actin (Figure 11). These systems are receiving increased attention in the context of studying molecular-level transport with modern probes such as optical tweezers (magnetic beads with fluorescent labeling). There are also several theoretical groups modeling biomotors in the framework of the statistical mechanics of irreversible processes.
Section 2, “Status,” provides many examples of current research in biomolecular materials and its applications, from which the immediate direction of the field can be discerned. In this section the panel suggests a few possibilities for the longer term and describes ways in which biomolecular materials can have an impact on national needs. For example, it appears likely that it will be possible to produce self-repairing composites of biologically active molecular arrays with passive structural materials, using the same chemistry that nature uses to produce operational devices and structures. Much of the content of this section is speculative. The developments described are meant to represent a vision of what might occur in the field over the next 10 to 20 years.
Liquid Crystalline Polymers35
Monomers similar in structure to the molecules that form liquid crystals can be linked together in a large number of ways, including linear and comb-like arrays, to form polymer liquid crystals. Liquid-crystal-forming molecules are beginning to be incorporated into polymer hosts to form anisotropic gels. These, in turn, can be cross-linked or used to form composites with structures on larger length scales. Polymer liquid crystals can combine the optical properties of liquid crystals with the ease of processing and good mechanical properties of polymers.
Surfactant-based lyotropics will play a basic role in self-assembling and self-repairing systems. For example, lipid-water systems are known to display cubic phases that are bicontinuous; i.e., the entire aqueous region of the sample is divided by the lipid bilayer into two disconnected regions that are
simultaneously continuous. The bilayer has been shown to be an infinite periodic minimal surface. It has cubic symmetry and long-range three-dimensional periodicity, even though the lipid molecules within the bilayer possess no positional order. Although the structures of biological cell membranes most often resemble those of the liquid-crystalline lamellar phase exhibited by phospholipids, there are several examples of organelles with structures very similar to bicontinuous cubic phases. Cell membranes resembling periodic minimal surfaces have been observed in cytoplasmic organelles such as mitochondria and chloroplasts. It has been suggested that in certain invertebrates the endoplasmic reticulum (a system of interconnecting membranes inside the cell) may exhibit gyroid (spiral) cubic structures.
Electron micrographs show that nature has elevated these structures to a high level of sophistication. The lamellar-like body of chloroplasts has revealed a structure in which the pore sizes of the two intertwining aqueous channels are different, suggesting a difference in osmotic pressure between them. It is believed that these two channels act as reaction chambers for the synthesis of chlorophyll. Even though the formation of these saddle surfaces in lipid-water systems is determined by the balance of forces between the polar head groups and the nonpolar tails, in biological membranes the periodic curvature may be due to the presence of some membrane-spanning proteins.
The bicontinuous cubic phases provide thermodynamically stable structures on the nanoscale whose characteristic size can be precisely controlled. They are not rigid, however. If these structures could be stabilized, they would provide continuous, triply-periodic pore space with very uniform nano-sized pores. Such structures could find many useful technological applications in such areas as controlled release and ultrafiltration. “Smart” release vehicles can be envisioned that would allow first-order drug release in response to stimuli. These phases can also be used as templates for synthesizing nanoporous materials and nanocomposites. Another important property of the cubic phase that can be harnessed is the large bilayer surface area that it provides (103 to 104 m2/g). Immobilization of proteins can be envisaged, either by covalent attachment to the head group or by simple incorporation into the bilayer, which could lead to the development of biosensors. It is believed that the functionality of the integral proteins will be optimal since the fluid bilayer provides an environment closest to the natural condition in vivo.
Fabrication of Devices by Self-Assembly37
The potential for fabrication of electronic devices by self-assembly has often been cited as a long-term goal of research in organic thin films. It is attractive to consider devices in which the components are individual molecules or molecular complexes self-assembled on substrates from solution or by deposition from interfaces. There are three steps that must be accomplished in order to achieve this goal. Functioning molecular units must be designed and synthesized, they must be organized on a surface into defect-free structures, and they must be interconnected to form functioning networks. While progress needs to be made in each of these areas, the last one is the most difficult and needs to be addressed in the long-term. It is clearly possible to make connections by photolithography, but this cannot be accomplished at the molecular scale. Methods based on scanning microscopy, especially with chemically active tips, may provide a solution, but it is difficult to envision how such a process can be carried out on a practical scale. A biomimicking process of self-assembly, in which connections are made by enzyme-like molecules that either form bonds or activate functional groups so that they can be photochemically linked, is more attractive.
Polymers—Synthesis and Processing.
The biosynthesis of polymers is discussed in Section 2 of this report. The generality of this approach remains uncertain. Although it seems clear that it is possible to use the technique to make virtually any copolymer of the 20 naturally occurring amino acids, extension to other classes of monomers is not simple. To date, two successful approaches to this problem have been reported. The first involves chemical acylation of transfer RNAs and in vitro protein synthesis. This is an elegant approach and has been shown38 to succeed not only with non-natural amino acids but also with lactic acid, a hydroxyacid related to the natural amino acid alanine. This result suggests that templated ribosomal catalysis might be extended to polymerization reactions other than polyamidation and that other classes of polymers might be prepared with the exquisite architectural control that is the hallmark of protein biosynthesis. On the other hand, the scale of these reactions is limited by the cost and inefficiency of cell-free biosynthesis, and the modest efficiency of incorporation of artificial monomers limits the in vitro method's utility in the preparation of periodic chains that require high levels of substitution. A second approach to the preparation of protein-like polymers of non-natural amino acids relies on the observation in the 1950s that bacteria can utilize as substrates a surprising number of amino acid analogues, including several with functionally interesting side chains.39 The scope of this second approach remains to be defined, but it appears likely that in vivo protein biosynthesis will prove to be a more versatile route to new materials than was previously anticipated.
Although the morphologies so far discovered in polymer systems are already quite diverse, it is nevertheless expected that advances in synthetic capabilities for greater control and complexity will lead to macromolecules with new structural geometries. Control of the morphology of polymers, and thereby control of their physical properties, has been well demonstrated in the anionically synthesized A/B block copolymers; biosynthesis will provide greater opportunities to tailor-make such materials.
Some possible future directions are to use block copolymer architecture to produce liquid-crystalline domains of prescribed size and shape embedded in a matrix of high-temperature thermoplastic. Such an approach could generate materials of novel mechanico-optical properties.
There is much additional scope for the production of advantageous materials by the coupling of dissimilar molecules. Not only are flexible-stiff combinations of interest (including the combination just mentioned), but so also is tailoring of the polymer backbone (persistence length and chemical bonding), which controls subsequent physical structure. Structural control could be of special interest for materials with rheological applications, because chain geometry can influence the nature of entanglements and the motion of the chain.
Control of molecular diffusivity is a critical factor in the organization of molecules at different length scales, but as yet it is almost unused. As an example, consider a blend of an A homopolymer and a B/C diblock copolymer. At high temperature or with sufficient solvent the system is in a single homogeneous phase. Depending on the interactions of A with B, A with C, and B with C, and on the relative selectivity of the solvent, macrophase separation of A liquid from B/C liquid or microphase separation of swollen B from swollen C may occur first, followed at some later stage by aggregation of A within the existing structure. Depending on the various chemical interactions, and especially on the relative mobilities of the components, a vast range of morphologies is possible. Arresting the evolving structure can be done thermally (via quench below the glass transition temperature Tg) or through a chemical trigger.
Another opportunity in this area is the challenge of inserting polymer genes into plants and then using biomass conversion to make protein-and polyester-based polymers. This would be an entirely new method of polymer production that might be more environmentally benign than present techniques.
Materials with highly specialized functions are likely candidates for the first applications of polymer biosynthesis. The major advantage of polymers is their easy and versatile processing into useful shapes such as fibers and films. Single-step processing in which the overall shaping of a part is achieved simultaneously with its detailed structural arrangements (possibly on several length scales) will be an important factor in acceptance into the marketplace. Likely early candidates for thin film applications are those in which surface, optical, electrical, or transport characteristics are critical. Because of the small quantities of materials required, the first specialty applications of biosynthesized materials will probably be biocompatible coatings.
Perhaps the earliest example of a biosensor is the use of canaries to detect lethal fumes in mines. Even today animals are the method of choice to search for the highly prized truffle, and dogs are used to track missing persons and to search for earthquake victims.
One of the goals of biomolecular materials research is to couple the sensitivity and selectivity of biosensing with the robustness and mass-production attributes of silicon and the reliability of electronics.40 To this end, optical-fiber-based and microelectronics-based biosensors have been fabricated to detect a large number of chemicals, including glucose, nerve gas, and ethanol.41 Many of these devices take advantage of highly selective antigen-antibody recognition events, others employ receptors as the sensing element, and yet others use catalytic selectivity of enzymes such as horseradish peroxidase to produce a detectable byproduct.
One of the major as yet underexplored opportunities in sensor research is the coupling of biological sensing units, whether they be receptors, antibodies, or enzymes, with microelectromechanical machines (MEMs). MEM devices have been fabricated with free-standing components that can be made to oscillate at a frequency that changes with the binding of a very small number of molecules.
Biomolecular-based sensors will have a wide range of applications, including the detection of low levels of toxic or harmful chemicals, the detection of biological warfare agents, and diagnostic applications in health care, agriculture, and food quality and safety. For example, one can envision nanoscale reactors and sensors that are safe and reliable parts of artificial organs, depending on the seamless integration of biomaterials with other high-performance materials.
Active transport by biomotors (discussed in Section 2, “Status”) suggests the possibility of molecular-level bioengineering and construction on a unit-by-unit basis. This capability would require the use of in situ biorecognition sensors. In other words, integrated systems might be fabricated that would use biomotors to build up supermolecular assemblies much as children construct tinker toy models.
Other examples of molecular machines are discussed elsewhere in this report, such as bacteriorhodopsin in the subsection titled “Membrane-associated Proteins” and RNA polymerases in the subsection titled “Polymers—Synthesis and Processing.” The idea of combining such machines presents a number of interesting opportunities. For example, one might couple a biomotor with an energy