Anal. Chem. 1996, 68, 841-844
Accurate Measurement of Ruthenium Isotopes by Negative Thermal Ionization Mass Spectrometry Min Huang,* Yongzhong Liu, and Akimasa Masuda
Department of Chemistry, The University of Electro-Communications, Chofugaoka 1-5-1, Chofu-shi, Tokyo 182, Japan
The method for measurement of ruthenium isotopic composition as RuO3- by negative thermal ionization mass spectrometry (NTI-MS) is shown to be sensitive and accurate. Precise measurement of the 18O/16O ratio, which is important for oxygen correction in NTI-MS, has also been made. Both Re and Pt filaments were tested, and the latter was proved to be more efficient for negative ion production. The mechanism of ion production with the addition of HI as a reducing reagent and Ba(NO3)2 as an ionizing enhancer was also studied. Sensitivity was found to be about 100 times higher than that of the positive mode. Factors related to negative ion formation are discussed, and parameters are optimized. The ionization efficiency has been improved to 0.7%. Ten nanograms of Ru yielded a total ion current of 3 × 10-12 A for 1 h. The precisions of all Ru isotope ratios with a 100 ng sample size were better than 0.009%. Thermal ionization mass spectrometry (TIMS) has been widely used in many research fields, due to its high accuracy and precision for isotopic measurement. However, some elements, like ruthenium and osmium, are very difficult to ionize by positive thermal ionization mass spectrometry (PTI-MS), the conventional technique, because of their high ionization potentials. The isotopic composition of Ru is important because, according to the theory of nucleosynthesis, its seven stable isotopes are produced by several different types of nuclear process: p-process only (96Ru and 98Ru), s-process only (100Ru), r-process only (104Ru), and mixed r- and s-process (99Ru, 101Ru, and 102Ru). Since radiogenic isotope abundance anomalies in 98Ru and 99Ru may result from the decays of 98Tc and 99Tc with half-lives of 4.2 × 106 and 0.21 × 106 y, respectively, accurate measurement of Ru isotope ratios may provide information about the production of technetium isotopes. The high ionization potential (7.364 eV) of Ru and the high volatility of its oxides greatly limit the ionization efficiency of positive Ru ions. Furthermore, the intensity of the signal decreases quickly. Devillers et al.1 and Poths et al.2 tried to improve the sensitivities by adding silica gel and boric acid. But they also reported interferences at masses of 100 and 104, which are probably from 40Ca28SiO2 and 88Sr16O, respectively. Negative thermal ionization mass spectrometry (NTI-MS) is becoming a powerful technique for the isotopic measurement of nonmetals and metals with very high ionization potentials. Heu(1) Devillers, C.; Lecomte, T.; Lucas, M.; Hagemann, R. Adv. Mass Spectrom. 1978, 7A, 553-564. (2) Poths, H.; Schmitt-Strecker, S.; Begemann, F. Geochim. Cosmochim. Acta 1987, 51, 1143-1149. 0003-2700/96/0368-0841$12.00/0
© 1996 American Chemical Society
mann’s group3-7 has contributed significantly to this development since the 1980s and has demonstrated NTI-MS to be a highly sensitive and accurate technique. Rokop et al.8 reported measurement of trace level technetium by NTI-MS, with lanthanum oxide as an ion enhancer in conjunction with Ca(NO3)2. Creaser et al.9 successfully determined nanogram quantities of osmium and rhenium and predicted correctly that the technique would get wide application in 187Re-187Os chronometry. Theoretically, ruthenium should produce intense beams of negatively charged oxide ion under proper conditions. However, no isotopic measurements by NTI-MS have been published. In this work, a method for Ru isotope measurement by NTI-MS is demonstrated. EXPERIMENTAL SECTION Instrumentation. The mass spectrometer used in this study was a VG Sector 54, which is equipped with a nine Faraday cup, variable multicollector system. Oxygen was leaked into the source chamber through a sensitive gas pressure regulator (Edwards). The polarity change operation from positive to negative mode required a few seconds. The Faraday amplifiers were bipolar, requiring no settling time. Reagents. A 1000 µg mL-1 solution was prepared by dissolving ruthenium chloride (Aldrich Chemical Co., Inc.) in 5% HCl medium and was used for isotopic measurement of terrestrial Ru. A 20 mg mL-1 Ba solution was prepared by dissolving Ba(NO3)2 (Kanto Chemical Co. Inc., Cica-Reagent) and was used as an ionization enhancer. Hydroiodic acid (assay 56%, Cica-Reagent) was used for reduction of Ru compounds and prevention of their loss due to volatilization before being ionized. High-purity HCl was prepared by three distillations. All aqueous solutions were prepared from water with 18 MΩ resistance by Millipore (MilliXQ) ion exchange. Filament Preparation and Sample Loading. A single platinum filament system was found to produce an ion beam intense enough for measurement. Platinum foil (0.03 mm × 0.7 mm) was obtained from Tanaka Rare-metal Co., Japan. After the (3) Heumann, K. G.; Schindlmeier, W.; Zeininger, H.; Schmidt, M. Fresenius Z. Anal. Chem. 1985, 320, 457-462. (4) Walczyk, T.; Heumann, K. G. Int. J. Mass Spectrom. Ion Processes 1993, 123, 139-147. (5) Walczyk, T.; Hebeda, E. H.; Heumann, K. G. Fresenius Z. Anal. Chem. 1991, 341, 537-541. (6) Volkening, J.; Walczyk, T.; Heumann, K. G. Int. J. Mass Spectrom. Ion Processes 1991, 105, 147-159. (7) Walczyk, T.; Hebeda, E. H.; Heumann, K. G. Int. J. Mass Spectrom. Ion Processes 1994, 130, 237-246. (8) Rokop, D. J.; Schroeder, N. C.; Wolfsberg, K. Anal. Chem. 1990, 62, 12711274. (9) Creaser, R. A.; Papanastassiou, D. A.; Wasserburg, G. J. Geochim. Cosmochim. Acta 1991, 55, 397-401.
Analytical Chemistry, Vol. 68, No. 5, March 1, 1996 841
filament was degassed at 3 A current in a high-vacuum chamber to remove impurities, the sample was loaded on the degassed filament. The Ru solution was dried at 0.7 A and covered with 0.1 µg of HI. The filament was put into the high-vaccum chamber again and heated with 2.5 A current for 30 min to reduce Ru compounds to Ru metal. The sample was then covered with 20 µg of Ba taken from the 20 mg mL-1 Ba(NO3)2 solution mentioned above and dried in air at 0.7 A. A homogeneous coating of Ba was found to be one of the most important parameters affecting precision and signal intensity. Mass Spectrometry. The sample was heated continuously to 2.5 A in 2 h. During this time, all focusing parameters were adjusted to maximize the intensity of signal. Usually, a set intensity (2 × 10-12-4 × 10-12 A) could be reached when the heating current was >2.7 A. We checked for the ions MO3- of Zr, Rh, and Pd, but no signals from them were found. It would be worthwhile to note that it is difficult to form negative oxide ions MO3- under the conditions used. In addition, peaks from Mo were hardly detected, and the corresponding corrections were made if necessary. RESULTS AND DISCUSSION Filaments. Theoretically, the lower the work function of the filament, the higher the ionization efficiency of the analyte in NTI-MS. The ion yields, β, for positive (eq 1) and negative (eq 2) ions are controlled by the following parameters: I is the first ionization potential of the analyte; W is the work function of electrons for the filament material; and EA is the electron affinity of the analyte.
β+ )
β- )
[
)]
g0 N+ I-W ) 1+ exp 0 + g kT N +N +
[
(
-1
)]
g0 N W - EA ) 1+ exp 0 g kT N +N -
(
(1) -1
(2)
In this work, both Re and Pt filaments were tested for NTI-MS. Only Pt produced RuO3- intensities strong enough for precise measurement of Ru isotopes, although Re should be better because of its lower work function. This result was repeatedly confirmed by our experiments. The work functions of Re and Pt are 4.98 and 5.13 eV, respectively. Such a small difference in work function between the two elements can be ignored when ionization enhancer (
