Fullerene Research in the Cross Group
Yale University, Department of Chemistry
In collaboration with Prof. Martin Saunders we make fullerenes with atoms and small molecules trapped inside and study their properties. These are van der Waals molecules in that there is no chemical bond between the trapped atom or molecule and the carbon cage. Yet, they are very stable, since the atom cannot escape unless several bonds are broken. Numbers refer to publications listed below. So far we have put He, Ne, Ar, Kr, Xe and tritium and nitrogen atoms, as well as He2, Ne2, CO, and N2 inside a variety of fullerenes. The tritium atoms are inserted by generating them at high energies in a nuclear reaction. The noble gases are put in by heating the fullerenes in the presence of the gas at high temperatures and pressures or by shooting them in as ions or metastable atoms. The noble gas compounds can be detected by mass spectroscopy either as intact molecules or by decomposing them and detecting the noble gas. In the case of 3He we can see the 3He NMR signal. When a molecule is put inside a fullerene, the vibration-rotation spectroscopy can have some unusual features. 
The chief method used to make the noble-gas fullerene compounds is to heat fullerene in the presence of the gas at 650oC and 3000 atm. We make an ampoule from a tube of OFHC copper by crimping one end, filling it, and crimping off the top. The ampoule is placed in a high-pressure bomb which is then filled with water, closed and heated to 650o. The pressure rises to 3000 atm, and the copper ampoule is squashed flat, compressing the gas within. After about 8 hrs., the bomb is cooled, and the ampoule is opened. The fullerene is extracted in CS2. About 85% of the fullerene is soluble, and about 0.1% of the molecules contain a noble gas atom. In the cases of He and Ne, we find small amounts of C70 containing two helium atoms or two neon atoms . We have found even smaller amounts of C60 containing two helium atoms.
Tritium atoms with high kinetic energy are generated using a nuclear reaction. We prepare either a lithium salt of C60 or use 3He@C60 or a mixture of C60 and 3He gas. 6Li and 3He each absorb thermal neutrons in a reactor to give tritium. The tritium then loses energy by ionizing the fullerene until it eventually stops. Some of the time it stops inside a fullerene molecule which then remains stable. We can isolate tritium labeled C60. In the case of the T@C60 generated from lithium, we obtain trace amounts of 3He@C60 formed by the radioactive decay of the tritium. If the tritium were on the outside, the 3He would be on the outside and would be lost.
We have constructed a beam machine to put atoms inside fullerenes. In the center is a cylindrical target, rotating slowly. On one side is an oven which produces a continuous beam of fullerene. Thus we have a freshly deposited surface of fullerene on the target. On the other side is a source of noble gas ions or metastable neutrals which hit the surface. The ions and mtastables are made in an electric discharge. Ions are extracted by an electric field and bent by 90o. The amount of incorporation for He+ is small at 30eV and rises to a maximum near 100eV and then decreases. Above 100eV the fullerene is partially destroyed. For Ne+, the yield is smaller, and the threshold is about 100eV. In the metastable mode, the ions are bent away from the target, while the metastables hit it. We find incorporation of both He* and Ne*. The method also works for nitrogen atoms. N @C60 consists of a free nitrogen atom with three unpaired electrons unbound to the carbon cage. It gives a clean atomic-like ESR signal. We have used the beam method to put He into dodecahedrane, C20H20, a hydrocarbon cage.
3He labeled fullerenes and their derivatives can be studied by NMR spectroscopy. The pi electrons around the fullerene molecules cause large diamagnetic shielding and an upfield shift of the 3He line relative to disolved 3He gas. C60 has an upfield shift of 6.4 ppm and C70 28 ppm. Higher fullerenes fall between these limits. Adding groups to the outside changes the pi electron structure and the chemical shift of the 3He. The most common adduct is across one of the 6,6 double bond joining two hexagons. Single addition usually causes an upfield shift of about 3 ppm from C60, the exact amount depends on the group being added. Addition across a 5,6 single bond joining a pentagon and a hexagon gives a much smaller shift. Multiple additions give a more complicated picture. Each fullerene molecule and each fullerene adduct gives a different, unique NMR line. Adding six electrons to C60 and to C70 gives another closed-shell species. The anions have very different chemical shifts from the neutrals. We also see the NMR signal for 129Xe in Xe@C60. The NMR of 129Xe@C60 is different from that of 3He@C60.
We have constructed a mass spectrometer to analyse the noble gas inside the fullerenes. We find that very pure C60 is extraordinarily stable. At 630oC the half life for decomposition is greater than one month, but even trace quantities of solvent or air absorbed in it will catalyze its decomposition. At 900oC the half life is 10 hours. In both cases the gas is largely or completely released by the decomposition of the C60. Using a different mass spectrometer, we can directly see the peaks for the various compounds.
Using multiple passes through an HPLC column gave us nearly pure Kr@C60. We could see the small shift in the 13C NMR due to the presence of the Kr atom. There were small shifts in the IR, visible, and UV lines and a 10% decrease in the life time of the lowest triplet state as well. A similar separation has been achieved with Xe@C60.
Using classical statistical mechanics, we can claculate the equilibrium constant for the incorporation of a noble gas atom into C60. We start with a potential function V(R) for the gas atom as a function of the distance from the center. V(R) is obtained by using one of several literature potentials between the gas atom and each carbon atom. We can then calculate the equilibrium constant using classical statistical mechanicsq* is the the usual partition function for X@C60 not including the gas motion, qint is the internal partition function for C60, and pX. The values for He and Ne seem to be relatively independent of the potential model used, but the values for the higher noble gases are much more sensitive to the choice of potential. For He and Ne the equilibrium values are much higher than the amounts that we can get from our experiments, so that the experiments are far from reaching equilibrium.
9,10 dimethyl anthracene (DMA) reacts reversibly with C60. By measuring the 3He NMR peak heights as a function of DMA concentration, we can get the equilibrium constants for the addition of successive DMA molecules to C60 and C70. By doing this as a function of temperature, we can get the enthalpy changes as well. Using mixtures of 3He@C60 and 129Xe@C60, we measured the ratio of the equilibrium constants for the two species. 3He is favored at low T and 129Xe at high T. Thus, putting Xe inside C60 changes both ΔH and ΔS for the DMA addition. We found that the various isomers of C84 have very different equilibrium constants for the addition of DMA, and this can be used as a basis for separation of the isomers. We used H2@C60 to measure the rate using an NMR T-jump mmethod. .
By adding suitable groups to the outside of fullerenes, it is possible to open a hole in the cage. We have measured both equilibrium constants and kinetics for noble gases entering and leaving chemically opened fullerenes.  We have put ammonia  and methane  into one of the opened fullerenes.The analysis required some unusual NMR spectroscopy.
Research support for this research from the US National Science Foundation is gratefully appreciated.
From here you can go to:
Contact me at firstname.lastname@example.org
Phone: (203) 432-5203, Fax: (203) 432-6144
Mail: Yale Chemistry Dept., PO Box 208107, New Haven, CT 06520-8107, USA
Last Updated: 7/26/12.