As with many important scientific discoveries, buckyball was discovered by accident. In 1985, the American chemist R. E. Smalley at Rice University, the British chemist H. F. Kroto at Sussex, and graduate students working under their direction were studying the nature of interstellar matter. They wanted to know what forms of carbon-containing materials can be found between the stars. The overall strategy of the research was to compare spectroscopic readings from unidentified matter in interstellar space with those obtained from well-characterized materials in the laboratory. If a match is found, then one can infer the nature of the interstellar matter. This strategy reveals a fundamental principle of the rules describing matter: they apply equally throughout the universe. The intrinsic properties (such as color, mass, etc.) exhibited by diamond or graphite or buckyball in space are identical to those on earth.
When buckyball was discovered, Kroto and his coworkers already knew that long chains of carbons were present in space. This knowledge was based on the readings obtained from a radio telescope. Every molecule exhibits a characteristic reading on this telescope that is like a fingerprint. This fingerprint can be compared to the fingerprint given by known molecules on earth, hence, the molecules in interstellar space are characterized. The nature of spectroscopic analyses used to characterize interstellar matter is a topic that is too involved for further discussion at this point. Instead, we turn our attention to the methods used to discover buckyball.
Smalley's apparatus (Figure 2.1) was designed to generate long chain carbon molecules so that their spectroscopic fingerprints could be measured. In the Smalley apparatus a laser is aimed at a rotating graphite disk in a helium-filled vacuum chamber. This apparatus uses the ability of lasers to deliver short, high energy bursts of energy in the form of light. The rapid, intense heating of the graphite surface by the laser enables many of the C-C bonds in the graphite to rupture. As a result, carbon atoms and small clusters of carbon atoms sputter from the graphite surface. Thus, the energy of the light produced in the laser is used to break the bonds between atoms in graphite, a process that involves the conversion of light energy to chemical energy. The high energy C atoms and small clusters of carbon atoms cool and collide in the He atmosphere yielding new bonding arrangements of C atoms. These new materials can be characterized by different instruments. The two instruments that played a key role in the discovery of buckyball are mass spectrometers and nuclear magnetic resonance spectrometers.
Figure 2.1 Smalley's apparatus for generating and detecting buckyball.
A mass spectrometer measures the mass of molecules and atoms. First a source of atoms or molecules is volatilized, or converted into a gas. In the Smalley apparatus, carbon atoms and clusters of carbon atoms are volatilized by the rapid heating of the rotating graphite block by the laser. The flow of He gas carries the volatilized matter into the mass spectrometer. At the entrance to the mass spectrometer, the atoms and clusters of carbon are ionized by bombardment with high energy electrons. Consider what might happen to a carbon atom that enters the mass spectrometer. Prior to ionization the carbon atom will have 6 protons in the nucleus and 6 electrons that occupy most of the volume of the atom. Thus the carbon atom is neutral (no charge). Upon collision with a high energy electron the carbon atom expels one of its electrons, leaving a carbon atom with 6 protons and just 5 electrons. The resulting ion has a net +1 charge. Ions with a net positive charge are called cations.
Mass spectrometers separate ions according to their mass-to-charge ratio. Because most of the ions have a charge of +1, we can consider the mass spectrometer to separate ions with different masses. A schematic of a mass spectrometer is shown in Figure 2.2. A simple way to understand how a mass spectrometer works is by analogy to a car on racecar track. The ion is analogous to the cars. First the ions are accelerated to high speed by electrical plates; the positively charged cations are attracted to the negative charge of the plates. Heavy ions will move more slowly than light ions just as heavy cars will move slower than light cars when they are pushed by the same amount of force. One might expect that the ions would simply crash into the plates. However, the plates have many large holes in them so that many of the ions are accelerated and then pass through the holes. Next, the ions enter a region in which a magnetic field is present. The trajectory of ions moving through a magnetic field is bent. This is analogous to a set of racecars entering a banked curve. As the racecars travel around a curve at high speed, the cars that are too slow (heavy) will slide down the bank toward the inside wall, and the cars that are too fast(light) will hit the outside wall. The racecars with just the right mass will continue along the track without crashing into either wall. Similarly, ions moving at high speeds through the mass spectrometer will travel along different trajectories from which their masses can be determined. The detector of the mass spectrometer measures how many ions travel through the curved trajectory without colliding into the walls of the instrument.
Figure 2.2 Schematic of a mass spectrometer
Consider what the mass spectrometer would measure if 1000 atoms of C in a graphite sample were subjected to such a high energy pulse of laser light that all of the bonds were broken. The matter moving toward the spectrometer entrance would consist of 1000 separated C atoms. Although each atom would be a carbon atom, not all atoms would have the same mass. On average 11 of the atoms would have mass 13 amu (due the 1.1% abundance of the 13C isotope) and 989 would have mass 12 amu (due to the 98.9% abundance of 12C. These atoms would be ionized at the entrance of the mass spectrometer to yield carbon cations, written as C+. If we set the strength of the magnetic field first so that masses of 12 amu were measured (or kept on track) and then switched the magnetic field such that masses of 13 amu were detected, the resulting mass spectrum would exhibit two peaks (Figure 2.3). The peak corresponding to 12 amu would be larger than the peak at 13 amu by approximately 
times. In fact the peak at 13 amu would be so small that it would be easy to miss.
Figure 2.3 Mass spectrum of carbon atoms
Smalley, Kroto, and coworkers found surprising results when the intensity of the laser was decreased. At high laser energies they observed large peaks at many different masses, suggesting that high energy atoms and clusters of carbon were volatilized from the graphite surface (Figure 2.4). However, decreasing the laser power led to the predominance of a single peak in the mass spectrum with a mass of 720 amu. Because the starting material, graphite, has only carbon atoms the matter with mass of 720 amu must consist of carbon, only. Chemical reaction cannot change the nature of the individual atom; only their arrangements are changed. Because most carbon atoms are 12C isotopes, this matter must correspond to 
carbon atoms that are bonded together. Apparently, the formation of molecules with exactly 60 carbon atoms must be very favorable.
Figure 2.4 Smalley mass spectra at high(a) and low(b) laser power
(b) 
On the basis of the information provided by mass spectrometry, Kroto, Smalley, and coworkers were faced with the problem of rationalizing the unusual stability of 60 carbon atoms bound together to make a molecule of buckyball. The fundamental issue concerned how to construct a molecular structure that satisfied normal bonding (four bonds per carbon atom) and comprised exactly 60 atoms of C. With a great leap of both insight and faith, as well as considerable influence from the geodesic dome structures designed by Buckminster Fuller, these collaborators proposed that C60 adopts an arrangement of carbon atoms that is similar to the stitching on a soccer ball (Figure 2.5).
Figure 2.5. Soccer Ball and Buckyball (ball-and-bond)
Why a soccer ball? If one traces the stitching on a soccer ball, one finds exactly 60 vertices, or points where three lines of stitching intersect. A soccer ball has both 6-membered (hexagonal) and 5-membered (pentagonal) patches. These are sewn together to make a round ball. Taken in by the beautiful symmetry of such structures, Kroto and Smalley conjectured that if 5-membered and 6-membered rings of carbon were placed together in the same pattern then a round molecule containing 60 carbon atoms would result. This structure can be reconciled with the tendency of carbon to form four bonds if one assumes that each C engaged in one double bond and two single bonds with the neighboring carbon atoms. Overall, this results in a structure that has patterns of alternating single and double bonds as one traces the C-C bonding framework. Each carbon lies at the vertex of fused 5- and 6-membered rings.
But why should C60 be so stable? The alternation of single and double bonds in a molecule has been found to correlate with unexpected stablility in molecules closely related to C60. By stability we mean that the forces, or bonds, holding the carbon atoms close together are stronger than one might have expected from comparison with simpler, related materials. Consider the example of C6H6. Various experimental measurements indicate that benzene (C6H6) contains six carbon atoms arranged at the vertices of a hexagon. Attached to each carbon atom by a single bond is a hydrogen atom (Figure 2.6). One can satisfy the bonding rules for carbon and hydrogen by (1) joining each hydrogen to a carbon atom with a single bond and (2) arranging bonds between the carbon atoms in an alternating single-double-single pattern as shown in the figure. Two arrangements of C-C double bonds in benzene are possible; the two structures differ only in the positioning of the double and single bonds.
Figure 2.6 Benzene and resonance structures
Which of the two structures for benzene is correct? Neither! Although each of the two structures implies alternation of C-C bond distances, experimentally it is found that all of the C-C distances are equal. Furthermore, benzene is found to be more stable than is expected for a molecules that has C-C single bonds and C=C double bonds. The American chemist, Linus Pauling, suggested a way to resolve this dilemma: benzene is best considered to be both structures simultaneously. In the language originated by Pauling, modern chemists state that the best description of benzene is a superposition of the two resonance structures shown above. By resonance we mean that benzene is at all times a mixture of the two limiting structures. This is subtly, but importantly, different from saying that benzene is one structure half of the time and the other structure the rest of the time. The best description of benzene is that is always half of each structure. Molecules that can engage in this type of bonding resonance are found to have more stablity than molecules that do not. Thus, we say that the effect of resonance is to (1) even out the alternation of bond lengths and (2) add stability to the molecule. A general postulate of Pauling's resonance theory is that the greater the number of resonance forms, the greater the stability of the material.
The special stablilization of C-C bonds associated with resonance in benzene is manifested in the C-C bond lengths. An average C-C single bond is 1.54 Å in length, whereas an average C-C double bond length is 1.34 Å. Based on the two resonance strucutures for benzene one might approximate each bond to be equivalent to 3/2 of a bond (the average of 1 and 2). This would lead to a predicted bond length of about 1.44 Å. The experimental C-C bond length is observed to be 1.39 Å, illustrating that the forces of attraction between carbon atoms in benzene are unexpectedly strong.
We can extend the concept of resonance to understand the unusual stability of C60. Just as different patterns of alternating single and double bonds can be drawn for benzene, different resonance structures can be drawn for C60. Thus, by arranging the carbon atoms in hexagonal and pentagonal rings, we find that a nearly spherical molecule with substantial resonance stablilization can be formed from 60 carbon atoms. These features are consistent with the properties observed for buckyball. For example, by analogy to benzene we would expect the average C-C bond distances in buckyball that is less than 1.44 Å the average C-C bond lengths of buckyball are found to be ??.
No. Although the soccer-ball shape of C60 is consistent with the observation of a peak at 720 amu and the unusual stability associated with resonance structures, the mass spectrum alone does not reveal the spatial arrangements of atoms in buckyball. The mass spectrum reveals the mass only, not the bonding arrangement. The critical piece of evidence that effectively "proved" the structure of buckyball awaited the isolation of sufficiently large amounts of material to perform a different type of spectroscopic measurement called nuclear magnetic resonance spectroscopy.
Magnetic Resonance Imaging (MRI) (MRI) is an increasingly common source of physiological information. For example, MRI can be used to probe the brain for irregularities. The information obtained from MRI is complementary to X-Ray photographs and CAT scans. Whereas X-Rays produce images of body regions with contrasting density (such as the dense bones vs. low density soft tissue), MRI yields images of regions with contrasting water content (such as different areas of the brain). A MRI image of a human head is shown in Figure 2.7. It may come as a surprise to you that MRI is actually a form of a more general spectroscopic technique with the ominous sounding name of Nuclear Magnetic Resonance (NMR) spectroscopy. NMR is a very powerful spectroscopic method that reveals direct information about the environment of atoms in different materials.
Figure 2.7 MRI of a human head2.4 Does the Kroto-Smalley experiment prove the structure of
buckminsterfullerene?
2.5 How does nuclear magnetic resonance spectroscopy work?
Movies of a rotating brain obtained with MRI
Just as the name implies, NMR (as well as MRI) involves observation of nuclei in the presence of a magnetic field. Atoms that have odd numbers of protons (2H, 14N, etc.) or odd numbered sums of protons and neutrons (1H, 15N, 31P, 13C, 19F, 17O, etc) have nuclei that behave like tiny, spinning bar magnets. These nuclei can be detected in nuclear magnetic resonance spectrometers; such nuclei are called NMR active nuclei. It is common to refer to these spinning bar magnets as nuclear spins or, more simply, just spins. If these tiny magnets, or spins, are placed in a strong magnetic field such as that in an NMR or MRI spectrometer, they align so that they are either parallel (in the same direction of the large magnetic field) or antiparallel (opposite in direction to the large magnetic field) to the field of the spectrometer's large magnet. The parallel alignment is more stable than the antiparallel alignment, so that slightly more spins are aligned with the external magnetic field than with it.
You may wonder why all of the spins are not in the more stable, parallel arrangement. They would be at temperatures approaching absolute zero. However, at room temperature the spins have thermal energy. Thermal energy is the amount of kinetic energy characteristic of that temperature. At room temperature the thermal energy available exceeds the difference in energy between the parallel and antiparallel spin orientations in a magnetic field. We say that the spins are constantly excited into the less stable spin orientation by thermal energy. On average, just a slight excess of spins have the more stable orientation than the less stable orientation.
The energy difference between the parallel and antiparallel orientations depends on the size of the applied magnetic field (Figure 2.8). Larger magnetic fields yield greater energy differences. A spin in the lower energy orientation can be induced to change its orientation by the action of electromagnetic radiation. This process is called a spin flip. Spin flips require that the energy of the electromagnetic radiation, or light, match exactly the energy difference between the parallel and antiparallel spin orientations. Nuclear magnetic resonance spectrometers measure the energy of the radiation required to induce spin flips. Soon we will see that such measurements yield detailed information about the structure of molecules with NMR active nuclei. But first, let's explore what is meant by electromagnetic radiation.
Figure 2.8 Behavior of nuclear spins in the presence and absence of magnetic fields.
We normally think of light in the context of what we can see: blue light, green light, red light, etc. However you probably have heard of other forms of light without realizing that they are closely related to the light that we see. For example, concerns about ultraviolet radiation are frequently associated with discussions of the "ozone hole", we use microwave ovens (which emit microwave radiation) to warm foods, and x-rays are used for medicinal purposes. Each of these is an example of electromagnetic radiation. All electromagnetic radiation consists of oscillating electric and magnetic fields that move at the speed of light (186,000 miles per second in vacuum). The oscillatory behavior of the electric and magnetic fields of light can be described as waves. What distinguishes one form of light from another is the wavelength and frequency of the electromagnetic radiation (Figure 2.9). For example, radio waves have long wavelengths (1-100 meters) and low frequencies (106 or one million cycles/second). X-rays have short wavelengths (1/10,000,000,000th of a meter) and very high frequencies (1018 cycles/second).
Figure 2.9 Electromagnetic Spectrum and Wave Illustration
The energy of electromagnetic radiation varies with the frequency. High energy radiation such as x-rays, gamma-rays, and ultraviolet light have high frequencies and short wavelengths. These forms of light are particularly hazardous because of their high energy and penetrating power. In particular, exposure to x-rays and gamma-rays must be monitored carefully. Lower energy forms of radiation such as microwaves and radiowaves pose less of a hazard than high energy radiation. Radio stations, television stations, and NMR instruments all use long wavelength, low energy radiation.
When a sample that contains NMR active nuclei is placed in the magnetic field of an NMR instrument, it is possible to induce a "spin-flip" by the absorption of electromagnetic radiation. That is, by absorbing light energy the nuclear spins may be stimulated to change from the lower energy, parallel orientation of the spin to the higher energy, antiparallel orientation. When this spin-flip occurs the nuclei are said to be in resonance, hence the name nuclear magnetic resonance. Resonance is not brought about by just any form of electromagnetic radiation, the energy of radiation precisely must match the difference in energy between the parallel and antiparallel orientations.
To get a better feel for the resonance phenomenon consider an analogy with musical instruments. Imagine that you removed all but one string from a guitar and tuned that string to concert A. When plucked this string will vibrate at the frequency characteristic of concert A. An NMR sample that is analogous to the one-string guitar would be a sample that had just one NMR active nucleus. When a trumpet is pointed at the one-string guitar and a scale is played, the guitar string will exhibit little vibration until concert A is played on the trumpet. At that point, the natural frequency of the guitar string and the frequency emitted by the trumpet are in resonance. The frequencies match and some of the energy of the trumpet's sound waves will be absorbed by the guitar string. As a result the guitar will play the note although the string was not plucked. Note that this type of resonance, which is very different from the kind of resonance that we have used to describe bonding in certain molecules, occurs only when the frequencies of the sound waves and the guitar string match. Similarly, when the radio-frequency transmitter of an NMR instrument is pointed at the NMR sample containing one NMR active nucleus, no energy is absorbed until the frequency of the radio waves match the natural frequency of the nuclear spin. That frequency corresponds to the radiation energy that equals the energy difference between the different spin alignments. At resonance, the radio-frequency light is absorbed and the nuclear spin flips its orientation. In considering this analogy it is important that you realize that sound and light are different, although they share wave-like properties. Sound waves are due to the movement of air or other matter, and are not forms of electromagnetic radiation. As a result sound does not move at the speed of light; instead, sound waves propagate at a far slower rate.
The power of NMR spectroscopy arises from the ability of the measurements to distinguish not only the natural frequencies of different nuclei (e.g. C and H have very different resonance frequencies) but also similar nuclei in different chemical environments. The bonding environment around a nucleus influences its resonance frequency. Consider two molecules that we have seen before, benzene and methane. NMR spectrometers could be used to measure the resonance frequencies of either the 1H or the 13C nuclei in these molecules since both elements have NMR-active nuclei. Let's look at the 13C NMR spectra first. The methane molecule has just one carbon atom so that there can be just one bonding environment for the carbon nuclei. Therefore the 13C NMR spectrum of methane has just one peak (Figure 3.0). Now consider benzene. Unlike methane in which the carbon atom has four single bonds to hydrogen atoms, each benzene carbon atom has two single bonds (one to carbon and one to hydrogen) and one double bond (to another carbon atom). Clearly the bonding environment of a carbon atom in benzene is different from that of a carbon atom in the methane molecule. As a result the NMR resonance frequency for benzene is different from that of methane as shown in Figure 3.0. As a general rule, atoms in benzene rings have resonance frequencies that are lower than those of methane (assuming of course that the magnetic field strength of the NMR instrument is the same).
Figure 3.0 NMR Spectra of methane, benzene, and hexane.
Note that the 13C NMR spectrum of benzene has just one peak! Although there are six carbon atoms in a benzene ring, the bonding environment of each C is identical. We can understand this in two different ways. First, trace the bonding topology (or connectivity) of benzene and you find that each carbon is identical. Each carbon has the equivalent of one and one-half bonds to adjacent carbon atoms and a single bond to hydrogen. A second approach utilizes symmetry considerations. The flat, hexagonal structure of benzene is highly symmetric. This is particularly clear if we use the simple hexagonal representation of benzene in Figure 2.6. This kind of representation is very common for carbon-containing molecules. None of the atoms are drawn explicitly, rather the vertices of the polygon represent carbon atoms and the hydrogen atoms are implied. Visually one can see that there is no distinction between the vertices of a hexagon. The hexagon can be rotated or flipped such that different vertices are interchanged without any apparent change in the hexagon. As a result of this symmetry, there can be no distinction between carbons. Hence, just one NMR signal is observed.
Now we consider a more complicated molecule, hexane. Hexane has a chain of six carbons, but the NMR spectrum exhibits just three peaks. This is because the carbons are in three different environments as shown by the a, b, and c notations in Figure 3.0. The carbons at the ends of the molecule (type a) have three bonds to hydrogens and one to another carbon. This bonding environment is clearly different from the four internal carbons which each have two bonds to hydrogens and two bonds to carbons. The distinction between carbon types b and c is more subtle. Look closely at the C-C bonds. Carbon type b is joined on one side to a carbon with three bonds to H and on the other side to a carbon with only two attached hydrogens. In contrast, carbon type c is joined on either side to a carbon that has two bonds to hydrogens. Although the difference is subtle, it is sufficient to create different NMR resonance frequencies for type b and c carbons. Thus, hexane has three carbon environments and three 13C NMR peaks.
More info: symmetry and benzene
Finally returning to buckyball, note that the structure proposed by Kroto and Smalley (Figure 2.5) is highly symmetric. Each carbon atom is located at the intersection of two hexagons and one pentagon. The 60 vertices of the buckyball are identical. Therefore, all the carbon atoms in buckyball are in the same environment. One expects such a structure to exhibit just one 13C NMR resonance despite the fact that there are 60 carbon atoms! By 1990, methods of obtaining the milligram quantities of buckyball required for NMR measurements had been developed. Consistent with the structure proposed by Kroto and Smalley, just one peak was observed in the spectrum of C60! This observation severely limited the number of acceptable structures: there is no simpler way of arranging 60 carbon atoms in agreement with the bonding rules of carbon such that the bonding environment of each C atom is identical. For most chemists, the structure of buckyball was considered proven by the observation of a single resonance in the 13C NMR spectrum.
After scientists discovered how to make buckyball in large enough quantity for characterization and study, a whole new field of chemistry was opened to creative exploration. As many scientists began working with buckyball in their own research labs, a flood of information hit the scientific journals. So many papers on buckyball were submitted that, in 1993, a journal entitled "Fullerene Science and Technology" was born.
Two of the most interesting areas of current research with buckminsterfullerene deal with building onto the outside and catching smaller molecules in the inside of buckyball. In 1991, addition of an osmium-containing sidearm to buckyball to make a derivative of C60 allowed for absolute confirmation of the structure of buckyball using a process called x-ray diffraction. (A full explanation of x-ray diffraction is presented in the unit of graphite and carbon fibers.) The osmium derivative and another deriviative of C60 are both shown in Figure 2.10. Derivatives are analogs of an original compound which can be said to be derived from the original molecule. Analogs of buckyball that have extra atoms attached to the original 60 atom carbon framework are derivatives of buckyball.
Figure 2.10 Fullerene Pyrrolidine and Osmium Tetroxide derivatives of C60.
Trapping gases such as helium and neon in the interior of a buckyball appears to be quite easy. The common method of making buckyball involves the creation of an electronic arc between two graphite rods that are placed in a helium atmosphere. An arc welder supplies the power. A helium atmosphere is used because helium is an unreactive gas that is capable of transfering heat rapidly. Therefore the tremendous energy at the site of the arc is rapidly dissipated as the fragments of carbon cool and form buckyball. Because helium is abundant in the buckyball generator, helium is trapped inside some of the molecules of C60. Indeed, molecules of buckyball prepared in helium have probably always generated some of the material with helium gas trapped in the interior cavity of buckyball. Only one out of every 880,000 molecules of buckyball are believed to have helium trapped inside them, though, and these special complexes remained undetected until very sophisticated mass spectroscopy techniques were applied to the problem in 1993. The interior of C60 is large enough to accommodate an atom of any element in the periodic table, and many metals including lanthanide, uranium, cesium, and scandium have been trapped by fullerenes ranging in size from C28 to C82. The hope of scientists working in this area is that fullerenes with metals trapped in their interiors will be superconductors.
Although practical applications for buckyball have not yet been realized (Figure ??--newspaper headline "Buckyballs apps slow to develop"), promising groundwork has been laid in two areas. In one area, materials called buckytubes have been made in much the same way as buckyball is made. Buckytubes typically consist of 2 to 50 concentric tubes, and each tube is an array of carbon atoms linked together in a curved sheet. The ends of each set of concentric tubes fit together like Russian nesting dolls and consist of hemispherical arrays of carbon atoms similar to the spherical arrays of atoms found in buckyball. Shaped like needles, each buckytube is about one micron long (one micron is about 1/70 of a human hair). If buckytubes can be lengthened into fibers, they will likely be far stronger than ordinary reinforcing fibers.
In another area of work with buckyball, a derivative of C60 was shown to inhibit HIV-1 and HIV-2, the human immunodeficiency viruses that cause AIDS. Researchers at the Univeristy of California, San Francisco, noticed that buckyball fit perfectly into the active site of HIV proteases. (The active site is where reactions occur.) A water-soluble derivative of C60 was made by Fred Wudl and co-workers at the University of California, Santa Barbara, and this compound was indeed shown to disarm the HIV virus and block HIV protease from cutting proteins. (Figure 2.11 shows the computer-generated model of C60 docked in the protease active site.) The infected cells themselves, however, were not damaged. At Emory University, inhibition of HIV's ability to infect cells was also shown using a water-soluble derivative of buckyball. Unfortunately, the potency of the buckyball analog is low when compared to AZT and other HIV enzyme-inhibiting drugs. To be useful, the buckyball compound must be at least 1000 times more potent than it is now.
Figure 2.11 Model of C60 docked in the binding site of HIV-1 protease
2.11 What are the possible applications of Buckyball?
Currently, no actual applications of buckyball have been realized. Groundwork using buckyball in reinforcing materials and in drug design have both produced positive results, and it is likely that more uses for buckyball will evolve as more and more scientists enter this new and challenging field.
This site is under construction.