798 Nature Vol. 280 30 August 1979 In situ identification of various ionic species in Jupiter's magnetosphere James D. Sullivan & F. Bagenal Center for Space Research, Department of Physics, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 Positive ions with various mass per charge values up to -160 have been identified within 20 jovian radii of Jupiter from the analysis of data from the plasma science experi- ment on Voyager 1. ONE of the major surprises from the Voyager 1 encounter with Jupiter in March 1979 was the discovery 1 of active volcanoes on Io. This result emphasises the importance of Io as the principal source of the bulk plasma in the jovian magnetosphere 2 • 3 • The plasma science experiment 4 on Voyager 1 has made the first detailed in situ measurements of the plasma (10-5,950 V) near Jupiter 2 . Continuing analysis of these measurements for the ionic composition of this plasma has revealed the existence of further atomic and molecular ions as minor constituents of the plasma. We have now identified ions with mass per charge (A/Z*) values of 1, 8, 10-2/3 , 16, 23, 32, 64, -104 and -160 within 20 jovian radii, RJ, of Jupiter in the dayside magnetos- phere. The measurements used in the preliminary analysis reported here were taken in the high resolution mode (M mode) which has -3.6% resolution in energy per charge between 10 and 5,950 V. The energy per charge scan effectively becomes a mass per charge scan whenever all ionic species have a common component of velocity into any sensor; this common component of velocity as well as the spacecraft potential are parameters of the present analysis. (The instrument contains four independent sensors 4 A, B, C and D, looking in different directions 5 .) The measurements are usually displayed as relative distribution function versus energy per charge; the relative distribution function is the measured current divided by the energy per charge width of the particular measurement step. Since the step widths increase with increasing energy per charge, a constant background will result in a smaller value of the distribution function at higher values of energy per charge. Significant heavy ion abundances are found beyond the inner magnetosphere where ions with A/ Z* = 8, 16, 32 and 64 were reported by Bridge et a/. 2 • For example, at 19.6 R" the energy I charge spectrum (Fig. 1) shows several distinct peaks. 10°k-L---L-__ 3, 6,000 Energy/charge (V) Fig. 1 D-sensor energy per charge spectrum obtained on day 63 at 15.50 UT when the spacecraft was at 19.6 R, and a magnetic latitude of - 8.5°. The heavy arrow labelled co-rotation marks the expected location of a proton peak moving with the geometrically expected co-rotation velocity. The li ght arrows mark the locations of various A/ z• values assuming the lowest peak is a proton pea k. 00 28-083fi / 79 / 35079R - Ol $01 .00 This spectrum is typical of the spectra for -15 min about this time. Each spectrum is analysed with a simultaneous fit to a sum of convected isotropic maxwellian distribution functions with a common component of velocity; the sum is over peaks at A/ Z* = 1, 8, 10-2/3 , 16, 23 and 32. The analysis of this spectrum gives a common component of velocity of = 180 kms- 1 and a spacecraft potential of less than ±5 V. The common component of velocity is not identical to the value of =240 km s- 1 expected geometrically from co-rotation; this difference and related measurements are discussed by McNutt et a/. 5 • The mass/charge values for the individual peaks obtained from the fit are within half a unit of the indicated numbers. Thus, we can make the following tentative assignments: A/ Z* = 1 is H+ ; 8 is 0 2 + or S 4 +; 10-2/3 is S 3 ... or even B ... ; 16 is o+ or S 2 +; 23 is Na·; and 32 iss + (or 0 2 ... ?).The peak observed at =3 .9 kV is most probably sodium and not magnesium or neon. This conclusion is based on the temperatures of the ions derived from the maxwellian fits; within ±20% all ionic species have the same temperature. Under these circumstances, it is possible to produce a peak at A/ Z* = 23 by an appropriate combination of Mg ... and Ne+. However, the temperature of the 'sodium' peak produced in this way is too high in comparison with tempera- tures of the other ions. There is a clear absence of a He+ or He 2 + peak; a rough upper limit can be put at n (He+, He 2 +) < ts n (H ... ). Any signal from A/ Z* = 64 would come in at =10.8 kV well above our energy per charge scan range. The mass ratio of heavy ions to protons can be computed independently of the actual Energy /charge ( Y) Fig. 2 C sensor energy per charge spectrum obtained on day 64 at 10.00 UT when the spacecraft was at 5.4 R, and a magnetic latitude of +4.4". The light arrows mark the locations of various A/ z• values assuming strict co- rotation. The notch at =950 V is caused by interference from another experiment on the spacecraft. mass and charge of the individual peaks as it depends on (A/ Z*) 312 • The heavy ion to proton mass ratio is > 95; consequently the mass density is dominated by heavy ions. It is instructive to compare these results with the inner magnetos- phere results of ref. 2. First, the A/ Z* = 23 peak would have been masked by the prominent A/ Z* = 32 peak so that its absence at 5.3 R 1 is not significant. Second, the relative abun- dances of 32: 16: 10-2/ 3 differ markedly; the higher ionisation states are enhanced at the earlier time, greater distance, repor- ted here. Direct measurement of values of A/ Z* > 32 were restricted to regions where the densities are high and the common component of velocity is small. Such is the case in the Io plasma <f) Macmillan Jnurn als Ltd 1979