Ionization yield

V is the volume of the source (it is constant), P is the partial pressure in analyte M within the source (P depends on the quantity of analyte introduced), a is a value that depends on the physico-chemical properties of M (IP in particular), and i is the ionization current (current of electrons produced by the heated filament). The ionization yield is variable from one molecule to another. It is considered not to exceed 1/1000 in the best of cases, which means that one forms, at best, 1 ion for 1000 introduced molecules  

Forming so few ions is not really a problem in terms of detection limits because the detectors in mass spectrometers (Chapter 2) are very sensitive. The problem concerns what will become of the non-ionized species. If one neglects gravity, there are essentially two forces present in the source of the spectrometer (1) the force exerted by the electric field that extracts formed ions and (2) the suction force exerted by the pumping system. The first force exerts itself only on the ions and the second concerns only neutral species in a first approximation, because the force exerted on the ions by pumping is negligible compared to that exerted by the electric field. 

If the extraction of ions is quasi-instantaneous since it is undergone in approximately 10 seconds, the extraction of neutrals is much slower. Therefore, the non-ionized molecules diffuse in the source and will be adsorbed (insofar as they have low volatility) on the panels of the source and on the electrostatic lenses of the extraction system. These neutrals contribute vastly to the pollution of the system and make it compulsory to periodically dismantle the source to cleanse it. When possible, one should be careful not to inject more analytes than necessary as to avoid polluting the source prematurely. 

---

Operationally, a procedure may be based on measuring the yield of a reaction traceable to ionization, usually giving a lower limit to the ionization yield. Thus, in the radiation chemistry of hydrocarbon liquids, the product of an electron scavenging reaction (for example, C2H3- radical from the scavenger C2H5Br)... 

For highly polar media, the yield of the solvated electron can serve as a lower limit to the ionization yield. This method needs short-time measurement and may work for liquid water and ammonia. Farhataziz et al. (1974) determined the G value—that is, the 100-eV yield—of solvated electrons in liquid NH3 to be about 3.1 at -50 ns. This corresponds to a W value of 32 eV, compared with the gas-phase value of 26.5 eV. The difference may be attributed to neutralization during the intervening time. In liquid water, it has been found that G(eh) increases at short times and has a limiting value of 4.8 (Jonah et al., 1976 Sumiyoshi et al, 1985). This corresponds to W,. = 20.8 eV compared with Wgas = 30 eV (Combecher, 1980). Considering that the yield of eh can only be a lower limit of the ionization yield, suggestions have... 

Scavenging experiments in hydrocarbon liquids (Rzad et al, 1970 Kimura and Fueki, 1970) tend to give low observed ionization yield, although the primary yield may be greater. The situation is similar for free-ion yield measurement under a relatively large external field. Both processes require large extrapolations to obtain the W value. 

Ionization yield is readily measured this offers relatively simple dosimetry, as the W value for ionization is nearly independent of the quality and energy of the incident particle (see Sect. 4.8). 

Following Platzman (1967), Magee and Mozumder (1973) estimate the total ionization yield in water vapor as 3.48. The yield of superexcited states that do not autoionize in the gas phase is 0.92. Assuming that all of these did autoion-ize in the liquid, we would get 4.4 as the total ionization yield. This figure is within the experimental limits of eh yield at 100 ps, but it is less than the total experimental ionization yield by about 1. The assumption of lower ionization potential in the liquid does not remove this difficulty, as the total yield of excited states in the gas phase below the ionization limit is only 0.54. 

Discovery of the hydrated electron and pulse-radiolytic measurement of specific rates (giving generally different values for different reactions) necessitated consideration of multiradical diffusion models, for which the pioneering efforts were made by Kuppermann (1967) and by Schwarz (1969). In Kuppermann s model, there are seven reactive species. The four primary radicals are eh, H, H30+, and OH. Two secondary species, OH- and H202, are products of primary reactions while these themselves undergo various secondary reactions. The seventh species, the O atom was included for material balance as suggested by Allen (1964). However, since its initial yield is taken to be only 4% of the ionization yield, its involvement is not evident in the calculation. 

The first subnanosecond experiments on the eh yield were performed at Toronto (Hunt et al., 1973 Wolff et al., 1973). These were followed by the subnanosecond work of Jonah et al. (1976) and the subpicosecond works of Migus et al. (1987) and of Lu et al. (1989). Summarizing, we may note the following (1) the initial (-100 ps) yield of the hydrated electron is 4.6 0.2, which, together with the yield of 0.8 for dry neutralization, gives the total ionization yield in liquid water as 5.4 (2) there is -17% decay of the eh yield at 3 ns, of which about half occurs at 700 ps and (3) there is a relatively fast decay of the yield between 1 and 10 ns. Of these, items (1) and (3) are consistent with the Schwarz form of the diffusion model, but item (2) is not. In the time scale of 0.1-10 ns, the experimental yield is consistently greater than the calculated value. The subpicosecond experiments corroborated this finding and determined the evolution of the absorption spectrum of the trapped electron as well. 

In conclusion we may state that there is evidence for multiple ion-pair recombination in spurs yet a theoretical analysis of free-ion yield and scavenging at low-LET based on the geminate ion-pair picture is meaningful in view of the similarity of the recombination process in the geminate and multiple ion-pair cases. However, if this analogy holds, the geminate ionization yield has to be somewhat less than the true ionization yield. 

Although free-ion yield has been measured in a number of polar liquids (see Allen, 1976, and Tabata et al, 1991, for tables), and in some as a function of temperature, neither the free-ion yield nor the total ionization yield is understood... 

Finally, when comparing precision, number of samples, analyte concentration, and ionization yield should be considered. A small amount of data can lead to unrepresentative values. Furthermore, when the concentration is close to the LOQ, the variability increases and analytes with higher ionization efficiencies yield present lower relative standard deviations (RSD) values because of H/N differences. 

---