Anal. Chem. 1997, 69, 2312-2316
Multielemental Characterization of Several Brands of Fullerenes and Fullerene Precursors by Instrumental Neutron Activation Analysis Henrik Rausch† and Tibor Braun*,‡
KFKI Atomic Energy Research Institute of the Hungarian Academy of Sciences, P.O. Box 49, H-1525 Budapest 114, Hungary, and Institute of Inorganic and Analytical Chemistry, Lorand Eo¨ tvo¨ s University, P.O. Box 123, H-1443 Budapest, Hungary
The characterization of the purity of fullerenes is based on the concept of defining the percentage amount of the main component (e.g., C60, C70, etc.) versus the minor fractions or microfractions of the homologue species and/ or other polycyclic organic compounds. It has also to be considered that fullerene products may be contaminated in some degree by element impurities as well, both from the production processes and from the various precursor materials themselves. We report here for the first time detailed data on trace element impurities in various fullerene precursors as well as in the main C60 and C70 fullerene products of different producers, measured by instrumental neutron activation analysis. The concentration distributions of about 35 trace elements have been studied, including the determination of the relevant limits of detection. It could be established that all fullerene materials investigated contain a broad scale of trace elements distributed within a fairly wide concentration range, from ppb (ng/g) up to ppm (µg/g) levels. For some impurity elements, extremely high concentration levels were found, indicating that elemental impurities have to be considered when studying the electrical, magnetic, and other features of the fullerenes. During our studies of the effect of nuclear radiation on the physical and chemical properties of the fullerenes using reactor neutrons, we have noticed an unexpected side effect of the neutron irradiation. Namely, it was supposed that neutron irradiation will not induce remarkable levels of radioactivity in the fullerenes, being pure carbon atoms, since carbon is known as a nonactivable nuclide. To our surprise, we observed that, after the neutron irradiations, all our fullerene samples showed a considerable rate of radioactivity which, in fact, was emitted by radioactive isotopes of trace element impurities produced by nuclear reactions in the fullerene samples. As a consequence, it was decided to undertake a detailed instrumental neutron activation analysis (INAA) study of the trace element impurities in C60 and C70 fullerenes as well as in graphite and soot carbon precursors of different fullerene producers (e.g., Hoechst A.G., Frankfurt, Germany; F&J Co., Moscow, Russia; MER Corp., Tucson, AZ; and Southern Chemical Group LLC, USA). † ‡
KFKI Atomic Energy Research Institute. Lorand Eo ¨tvo ¨s University.
2312 Analytical Chemistry, Vol. 69, No. 13, July 1, 1997
It is well known that trace and ultratrace element impurities significantly affect the electrical, magnetic, optical, etc. properties of modern conductors, semiconductors, and superconductors. Thus, the chemical identity, concentration level, spatial distribution, and lattice position of the trace element impurities are important parameters when studying the physical properties of the fullerenes. As a detailed literature retrieval study revealed, the quantity of trace element impurities in fullerenes has, so far, been completely neglected. Our first results on the determination of trace element impurities in fullerenes by INAA were recently presented.1,2 In this paper, we extend the analytical investigations to additional fullerene samples and give a summary report of the measured concentration and/or limit of detection (LOD) values for about 35 elemental impurities. EXPERIMENTAL SECTION General Concept of the INAA Multielement Analysis. Commonly, in order to analyze multicomponent samples by INAA covering many detectable elemental impurities, measurements had to be optimized within a dual neutron irradiation concept. First, a short neutron irradiation of 300 s was used to determine elements producing short-lived radionuclides (T1/2 < 4-6 h). On a parallel sample, a subsequent long neutron irradiation of 24 h served for the determination of impurity elements producing middle- or long-lived isotopes (T1/2 J 6 h). The induced γ radiation of the isotopes was counted by γ spectroscopy after subsequent optimized decay periods. Sample Preparation. For the INAA measurements, sample amounts of about 20-50 mg were prepared both in high-purity polyethylene capsules for short-period neutron irradiations and in high-purity SUPRASYL-AN type quartz ampules for long-period neutron activations, respectively. Both types of sample holders were hermetically sealed. Neutron Irradiation. Samples were irradiated in the recently renewed WWR-M type 10 MW nuclear research reactor operated by the Atomic Energy Research Institute, Budapest, Hungary. Short neutron irradiations were performed by means of a computer-controlled pneumatic fast rabbit system running within a distance of 82 m between the NAA laboratory and the reactor core. At the irradiation position, a thermal neutron flux of 6.0 × 1013 n‚cm-2‚s-1 and a thermal/epithermal flux ratio (φs/φe) of 39.6 were available. During the neutron irradiations, the polyethylene (1) Braun, T.; Rausch, H. Anal. Chem. 1995,65, 1512. (2) Braun, T. J. Radioanal. Nucl. Chem. Art. 1996, 203, 377. S0003-2700(96)01092-X CCC: $14.00
© 1997 American Chemical Society
Table 1. Elemental Impurity Concentrations (Ci, in ng/g) in Several Fullerene Precursors, Measured by INAA MERGRAa
MERSOOTa
HOESOOTa
element
Ci
SD (%)
Ci
SD (%)
Ci
SD (%)
Ag Al Ar As Au Ba Br Ca Ce Cl Co Cr Cu Fe Ga I In Ir K La Mg Mn Mo Na Pt Sb Sc Sm Ta Th Ti U V W Zn
