1
The Discovery, Development, and Applications of Fullerenes
Kristen Leskow and Laura Weaver
March 2007
Figure 1. C60 buckyball (Haymet).
History and Discovery:
Until the discovery of fullerenes in 1985, the only known forms of pure carbon were graphite and diamond. Graphite consists of sheets of carbon atoms bonded hexagonally, and diamond is a tetrahedral crystal lattice structure, making diamond very hard (“Buckyballs”). In 1985, British chemist Harry Kroto studied combinations of carbon atoms that appeared to be primarily molecules with exactly sixty carbons. Kroto collaborated with Richard Smalley and Robert Curl to recreate in the laboratory the conditions near the red giant stars where Kroto had detected the carbon molecules. The scientists could not quickly determine the structure of the C60 molecules, however. They hypothesized that the molecules must be spherical in nature, due to its high stability (“Buckyballs”).
Since these carbon-60 molecules could be produced by laser vaporization of graphite, they believed that individual hexagonal six-carbon rings were blasted apart from the graphite structure, and these rings collide with each other and form an equilibrium structure. The most common equilibrium structure combined sixty carbons in a unique arrangement, although seventy-carbon and other structures were also found. Kroto and his colleagues searched for a geometric arrangement of the sixty carbons based upon combinations of six-membered rings, and reasoned that the molecule must be spherical in order for all of the sp2 valences to be satisfied (Kroto).
Eventually, they found that the geometric shape must be a series of interconnected hexagons and pentagons, such as in the soccer ball (football), with the hexagons in white leather and the pentagons in black leather. Thinking that the shape might be a spherical version of the hexagonal sheets of graphite, they tried to make spheres of hexagons. Finally, they determined that the combination of twelve pentagons and twenty hexagons formed a perfect ball with sixty vertices. The pentagonal rings sit at the vertices of an icosahedron.
The architect R. Buckminster Fuller had designed a geodesic dome for the 1967 Montreal World Exhibition with the same combination of hexagons and pentagons. Kroto, Smalley, and their colleagues named the C60 molecule buckminsterfullerene in his honor (“Buckyballs”). When other spherical carbon molecules were found, the name was shortened to fullerene to refer to the family of molecules.
Curl, Kroto, and Smalley received the Nobel Prize in 1996 for this discovery.
Independently, near the same time, Tony Haymet of the University of California at Berkeley published a paper predicting the existence of a compound of this kind (“Buckyballs”). Based upon the molecule corannulene, a similar nonplanar assembly of carbons in pentagonal and hexagonal rings, he estimated that such a molecule existed, which he named “footballene,” with sixty carbons in twelve pentagons and twenty hexagons (Haymet).
Figure 2. Corannulene molecule (Haymet).
Chemistry:
All sp2 valences are satisfied with the buckyball structure. The hexagonal rings consist of three single bonds and three double bonds, and the pentagonal rings have only single bonds.
Figure 3. C60 buckyball bond structure. The red bonds indicate single bonds and the yellow bonds indicate double bonds (
The stability of the molecule is largely determined by the location of the double bonds. A double bond in a pentagonal ring further shortens the bond length, straining the ring. The structure shown in Fig. 3 is the only structure for C60 without double bonds in pentagonal rings. This leads to poor delocalization of electrons, since the limitations on the location of double bonds limit the movement of electrons, causing the molecule to be less stable than originally predicted and thus more reactive. Thus C60 is not an aromatic molecule, but in fact more like a large closed-cage electron-deficient alkene which can react in similar ways to chain alkenes (Taylor).
Up to six electrons can be added to C60 and C70 fullerenes, creating aromaticity in one of the pentagonal rings in each paracyclene unit, comprised of two pentagonal rings separated by a double bond. Thus the anions of C60 and C70 are more stable than the neutral molecules, which can be used as oxidizing agents in this way. Anions of C60 can be reacted with methyl iodide to add methyl groups. Adding larger groups to fullerenes is affected by the bulkiness of the group because of the steric constraints of the molecule. Additions are possible in the form of cycloadditions, bridging (such as forming epoxides, carbon bridges, and metallic bridges), and group additions (such as halogenation, hydrogenation, anionic reaction with electrophiles, addition of neutral and charged nucleophiles, and radical addition). Up to twenty-four small molecules can be added to C60 without any two being adjacent, and this number decreases for more bulky additions. Substitutions and polymerization are also possible (Taylor).
Figure 4. Summary of possible reactions that C60 and C70 can undergo (Taylor).
Other fullerenes:
Although the C60 buckyball is the most famous fullerene, many other molecules of the family of interlocking hexagons and pentagons have been discovered. The most important groups of fullerenes include exohedral fullerenes (with molecules attached to the external cage), endohedral fullerenes (encaging some other molecule), and nanopeapods (fullerene molecules enclosed in carbon nanotubes).
Conventional Synthesis:
Kroto, Smalley, and Curl were only able to produce very small quantities of buckyballs. In 1990, both German and American scientists independently developed a procedure for making larger quantities of buckyballs. They sent a large current between two graphite electrodes, resulting in a carbon plasma arc which cooled into soot, in which the fullerenes could be isolated. The fullerenes in the soot were comprised of about 75% C60, 23% C70, and the rest larger molecules (“Buckyballs”).
The method published by Krätschmer et al for mass production of C60 uses pure carbon soot produced by evaporating graphite electrodes in helium at about 100 torr. The soot is gently scraped from the surfaces of the evaporation chamber and dispersed in benzene. The dissolved material gives rise to C60 and is a wine-red to brown liquid, varying with concentration. The liquid is separated from the soot and dried with gentle heat, producing a residue of dark brown to black crystalline material. The material can also be dissolved in other nonpolar solvents, such as carbon disulfide and carbon tetrachloride. Krätschmer’s paper also mentions concentrating the material by heating the soot to 400ºC in a vacuum or inert atmosphere, subliming the C60 out of the soot. The paper recommends purifying by washing the initial soot with ether before the concentration procedure is applied. These techniques produce a thin film or powder, which can be handled without special precautions and which are somewhat stable in air, with some deterioration after several weeks. The material can be sublimed repeatedly without decomposition, and at least 100mg of purified C60 material can be produced by one person per day. The solid produced after the benzene has been evaporated forms a variety of crystals of packed fullerenes, all with hexagonal symmetry. However, the diffraction patterns of larger crystals show no clear pattern, indicating that the crystals do not have long-range periodicity in all directions due to the disorder within the crystal (Krätschmer).
Subsequent articles provide further details about the Krätschmer-Huffman contact-arc technique. Carbon electrodes are arced in an atmosphere of helium, typically with an alternating current. This causes the deposition of soot, from which can be extracted the fullerenes. The use of a flow-through reactor assembly is helpful for removing the soot from the arcing zone; the soot generated by arcing is flushed by the helium carrier gas (Weston). The graphite rod electrode is kept in gentle contact with a graphite disk by an adjustable spring. The graphite rod and disk are connected to an external alternating-current power source at 60 Hz. The graphite is vaporized by driving current between the electrodes between 100 and 200 A, with an rms voltage of 10-20 V. The spring tension is varied to optimize the contact between the graphite rod and disk in order to ensure that the bulk of the power is dissipated in the arc as opposed to Ohmic heating of the rod (Haufler).
Weston and Murthy discuss the optimization of the contact arc method for fullerene synthesis. Since UV radiation results from the current applied for arcing, removal of the soot helps to prevent UV decomposition of the fullerenes. Although UV radiation is proportional to reactor current, lower currents give lower vaporization rates and require longer arcing times, so the reactor current must be optimized in order to minimize the UV effect. The helium pressure should be optimized and other gases such as hydrogen and water vapor should be eliminated. Weston and Murthy report an increased yield with application of a black coating to the inner reactor surface and with an increased residence time of the carbon vapor. They recommend using cyclohexane as the solvent for chromatographic separation of C60 and C70 products for reasons of time, volume, and toxicity of solvent; C70 is more difficult to elute out than C60 and requires more solvent. Weston and Murthy report fullerene yields of 15 wt%, and yields of up to 40% with direct laser vaporization in the single-walled nanotube experiments of Guo et al. (Weston). Despite these optimization methods for contact arc fullerene synthesis, the process variables are relatively rigid and can be changed only so much in order to maximize the product yield. However, the procedure is relatively simple and can be carried out in many laboratories with minimal requirements.
Selective-Size Synthesis:
The first fullerene discovered was C60, the buckyball, and it is known to be the most stable of all neutral carbon fullerenes. Since that discovery, research has progressed toward investigating various sizes of fullerenes to learn more about the compound and explore more possible applications. Large fullerenes could possibly fit needs that fall in the gap between classic fullerenes and carbon nanotubes because of their small band gaps. Small fullerenes may prove to be important for astrophysical implications (Umeda). The need for varying cage sizes arises with endohedral fullerenes. This family of fullerenes capture other atoms or molecules within the carbon cage. An important property that determines the ability of the cage to enclose another molecule is the relative sizes of the cage and molecules. For example, the larger the molecule encaged, the larger the cage must be. This has led to an entire area of research and synthesis. There is a great challenge in the control of the size of the carbon fullerene that is synthesized. In the conventional arc-discharge method, the final product is a function of the final product’s stability. C60 and C70 are the most stable fullerenes and therefore will be synthesized before other fullerenes, which may be the preferred product.
The first successful synthesis of C60 by size-selective flash vacuum pyrolysis (FVP) at 1100°C was accomplished in 2002 by Scott and co-workers. They explain the entire twelve-step synthesis to form the cholorhydrocarbon, C60H27Cl3, precursor, and the ensuing pyrolysis to form the carbon 60 fullerene (Scott). This process only provided a product with a maximum yield of 1%, but showed promise that C60 could be selectively formed (Umeda).
Figure 5. First known precursor to produce C60 by FVP.
Where X=Cl. (Umeda)
This result motivated other researchers to look for precursors to size-selectively synthesize C60. The reason these precursors are so hard to determine and use is the lack of knowledge of how the fullerene formation mechanism proceeds. Many will agree that cyclic polyynes are one of the most probable precursors to fullerenes. Two different research groups, Tobe and Rubin, spent substantial effort attempting to synthesize carbon-60 molecules through three-dimensional cyclic polyynes. Both groups are trying to create the closed fullerene molecule by the FVP process. The following synthesis is the same for both groups where they differ is in the way by which they create the precursor (Umeda).
Both groups report proudly that one should be able to achieve C60 by their method of synthesis. However, neither ever produced any feasible amount of carbon 60 product. They both have formed C60+ or C60- and report this as a breakthrough (Rubin, Tobe). This may be some proof of the ability to have control over the size of the fullerene by the precursors, but the ion formation is still very different than the classical fullerene cage. Since charge transfer can be a very important characteristic of an endo- or exohedral fullerene, a charge on the carbon-60 cage cannot be ignored. More work must be done on this synthesis to prove the synthesis actually proceeds as suggested and to synthesize the desired product, not an ion corresponding to the molecule, due to the importance of charges in this area of science.
Exohedral Fullerenes Application:
Soon after buckyballs were first discovered, researchers began thinking about how the cage could be changed or adjusted. A popular use of a fullerene cage is the addition of molecules to the external cage. A recent application of this group of fullerenes has been in hydrogen storage.
Zhao et al report that “the U.S. Department of Energy has determined that a hydrogen storage density of 9 wt% will be required for fuel-cell powered vehicles to able to replace petroleum-fueled vehicles on a large scale.” They suggest that by using organometallic molecules based upon C60 to store hydrogen, nearly 9 wt% can be retrieved reversibly at room temperature and near ambient pressure. Complexes of transition metals with hydrogen on pentadiene rings can store up to six dihydrogen species, but the complexes may polymerize when the hydrogen is removed, rendering the process irreversible. Arranging the complexes on buckyballs, such as C60[ScH2]12 and C48B12[ScH]12, leads to stable species which can reversibly absorb additional hydrogen. Doping the buckyballs with boron has been experimentally observed. This reduces the fullerene weight and enhances the complex’s stability by increasing the binding energy. The boron-doped buckyball also allows the binding of an additional H2 molecule per Sc, increasing the amount of retrievable H2 to 8.77 wt% (Zhao).
Figure 6. Optimized Molecular Structures of C60[ScH2(H2)4]12 (a) and C48B12[ScH(H2)5]12 (b) (Zhao).
Endohedral Fullerene Synthesis:
Endohedral fullerenes are a class of chemicals that are of great interest, especially in the biochemical world. Endohedral fullerenes consist of a carbon cage with another molecule enclosed within the cage. For possible applications, some imagine the ability to capture a drug in this very stable carbon cage in order to control the delivery time of the drug to the body. Another possibility is to place a radioactive “tracer” in the fullerene and then safely inject it into the blood stream. Also, endohedral fullerenes are the most likely type of fullerene to be used studied for use in nanoscale electronic devices because of their ability to have a specific band gap (0.2-1.0 eV) as a function of chemical structure (Shinohara).
The possibility of what fullerenes can hold inside their cage is always growing. Conventional endohedral fullerenes were monometallofullerenes like La@C82. The formation of these first endohedral fullerenes was achieved by a high-temperature laser-vaporization process in 1991. Knowledge of these molecules was limited to the fact that the metal was truly enclosed in the cage and that there exists a charge transfer from the caged metal to the carbon cage. For example, La@C82 is better described by La3+@C823-. There was great difficulty in applying these products to any application or even obtaining accurate analysis of them, because the synthesis produced such a small amount of product (less than 1% yield). Research of endohedral fullerenes was then extended to encaging nonmetals and noble gases. Nonmetals and noble gases were confined primarily in C60 fullerenes, and they do not carry a charge. This family of endohedral fullerenes was produced by placing the already-made C60 under high pressure in the presence of the gas to be encaged (Dunsch 2006).
Multimetallofullerenes were the next group to be investigated. These endohedral fullerenes were the first to break the well-accepted isolated pentagon rule (abbreviated as IPR). The violation of this rule means that the carbon cage can have pentagons fused together, creating a fullerene that isn’t perfectly spherical, as can be seen in figure 7. Only endohedral fullerenes are known to be able to violate this stability rule. These endohedral fullerenes were originally synthesized by the Krätschmer-Huffman arc method. This conventional process produced only a maximum of 2% yield of metallofullerenes in the soot (Dunsch 2004).
Figure 7. Calculated Sc2@C66 structure: top view at left, side view at right. Violation of IPR. (Dunsch 2)
The production of each new endohedral fullerene suggests new possibilities for atoms to be enclosed and new applications. However, all of the preceding endohedral fullerenes and their respective formation processes have one major disadvantage. They produce extremely low yields (less than 2%) of the endohedral fullerene as final product. No application would be possible if only such a small amount of material could be produced.
The first noticeable improvement in yield of endohedral fullerenes came in 1999 when a small amount of nitrogen gas (from air) was introduced into the Krätschmer-Huffman generator during the vaporization of the metal oxide containing graphite rods. This new process was named trimetallic nitride template (TNT) and noticeably improved endohedral fullerene yields. This synthesis first created Sc3N@C80. The empty cage of C80 has never been isolated and the trimetallic nitride is not stable as a single molecule. Therefore the endohedral fullerene must stabilize itself. After the success of the first trimetallic nitride fullerene various others were investigated. The general formula for these molecules is M3N@C80 where M=Sc, Y, Tb, Ho, and Er (Dunsch 2006).