Changes in sensitivity with gap width were recorded for various elemental +1 and +2 ions and molecular ions utilizing electrical detection peak switching techniques. The peak switching was accomplished magnetically rather than electrostatically through use of a Hall probe magnetic field monitor, thus allowing coverage of the entire mass range at a constant accelerating voltage. For metallic matrices, gap effects are on the order of 5-10% for most elements. In graphite matrices, however, changes in sensitivity with gap of up to a factor of three are not uncommon.A correlation between gap effects and ionization potential is also drawn. The ratio of Ionization potentials between the matrix and a particular measured species appears to exert a definite influence on the behavior of sensitivity changes with gap width.Investigators have previously recognized that the width of the spark gap could have an effect on trace elemental analysis by spark source mass spectrometry (SSMS). One of the earliest studies of these effects was made by Franzen (1), who showed how gap variations could affect elemental sensitivities in an iron matrix. He attributed these effects to changes in ion energy distributions upon varying the spark gap width. Woolston and Honig (2) demonstrated that the energy distribution of ions in the beam is also affected by gap width. Effects of spark gap on sensitivity have also been suggested by other workers as tangential observations to various studies. Bingham and Elliot (3) cited gap effects in high accuracy peak switching analysis. Colby and Morrison (4) have shown gap effects for vanadium and chromium. Konishi and Nakamura ( 5) have described spark gap effects on ion ratios.We have routinely monitored spark gap width (6) and recently described an automatic spark gap control unit (7). The present study was undertaken to obtain a clearer understanding as to the direction and magnitude of spark gap effects in a number of different matrices.
EXPERIMENTALApparatus. The basic AEI MS702 mass spectrometer used for this study has been previously described (8) as has the standard electrical detection equipment (3), Modifications or additions in our laboratory to the standard system have been detailed previously ( 6). An automatic spark gap control unit (7) was used to monitor all selected gap widths.Additional modifications used in this study include: (a) The electrical detection monitor head and amplifier are used for both log ratio scanning and peak switching. A wiring modification al-
