An enclosed inductively coupled plasma (ICP) was designed to overcome the limitations of the conventional ICP torch for the analysis of toxic and reactive gases. In particular, the extreme toxicity of arsine prevents the safe application of a standard ICP torch and gas exhaust for the direct determination of impurities in arsine. The enclosed ICP provides containment of toxic gases in a quartz discharge container. The total volume of toxic gas consumed is minimized as well. Parameter characterization of a sealed ICP system was investigated. The choice and role of the additive gas, effect of flow rate, discharge container size and geometry, rf power, signal reproducibility, operating parameters, and procedures were determined. Modifiers were investigated to prevent deposition of arsenic and metallic impurities to the cooler container walls. The first reported direct qualitative analysis of semiconductor-grade arsine is described. Chlorine was found to be the most effective additive gas for arsenic vaporization for both flowing and static ICP operation. Chlorine addition to the argon stream extended the arsine introduction to concentrations of up to 10% into the discharge. An rf generator (40.68 MHz) power of 1.0 kW and 30% chlorine content for 7.11% arsine in a 65-mm diameter spherical container were applied to identify eight impurities qualitatively: C, Fe, Ge, Mg, Mo, Ni, Sn, and V. A vapor phase introduction system was developed to calibrate the SICP. Theoretical detection limits for tin in arsine and chlorine were calculated as 2.00 and 0.218 ppb, respectively. A hypothesis was formulated to describe the stability of the chlorine-containing arsine plasma. Proof of this hypothesis will require techniques to probe the presence and distribution of ion and atom species within the sealed discharge. The absolute noise power spectra of atomic emission signals from the SICP for flowing and static operation demonstrated that noise below 5 Hz was lower than observed in conventional ICP discharges. White noise levels were lower for the SICP than a conventional ICP. The implications of this result is the improvement in signal-to-noise ratio signal averaging techniques. This can provide very low analyte detection limits measured in the sealed inductively coupled plasma.
| Identifer | oai:union.ndltd.org:UMASS/oai:scholarworks.umass.edu:dissertations-8419 |
| Date | 01 January 1992 |
| Creators | Jacksier, Tracey |
| Publisher | ScholarWorks@UMass Amherst |
| Source Sets | University of Massachusetts, Amherst |
| Language | English |
| Detected Language | English |
| Type | text |
| Source | Doctoral Dissertations Available from Proquest |
Page generated in 0.0109 seconds