John D. Hey

Criteria for local thermal equilibrium in non-hydrogenic plasmas

Abstract Rate equations for collisional processes are formulated and used to obtain the minimum electron density necessary to ensure LTE of a particular atomic species, with particular attention to the existence of low-lying metastable levels. The final expression is compared with an earlier criterion of Griem. The results of various experiments on plasmas containing one or more ionization stages of helium, nitrogen, oxygen, and argon are discussed. The importance of making reliable estimates of collision strengths, for excitation involving metastable levels, is emphasized.

Thermal equilibrium criteria in a nitrogen plasma

Abstract A method for obtaining the lower electron density limit for LTE in a nitrogen plasma is described, whereby the experimentally determined ratio of the collisional-radiative ionization and recombination coefficients ( S /α) is compared with the corresponding LTE value (Saha ratio). It is argued that if the electron density is increased from values of about 10 16 cm -3 , S /α should tend to the Saha ratio as LTE is approached. For NII and NIII spectral lines, this is found to happen at an electron density of a few times 10 16 cm -3 when the electron temperature is about 3 eV, in good agreement with the LTE criterion of Griem.

Calculated stark widths of oxygen ion lines

Calculations have been performed on the electron impact broadening of isolated lines from singly-ionized and doubly-ionized oxygen emitted from a plasma of electron density 1017 cm-3 and temperature about 2 eV. These have been compared with results of measurements performed by Platisa, Popovic, and Konjevic on a plasma produced by a low pressure pulsed arc. Good overall agreement has been obtained for both ionization stages, which we interpret as strong support for a recently derived expression for the effective Gaunt factor in line broadening calculations. This in turn indicates the important role that the curvature of the perturber trajectory plays in the broadening process, and that by proper allowance for this effect, classical path calculations of the isolated ion line widths can be extended to spectra of the multiply-charged ions. Some ambiguity still remains, however, as to the proper method of extrapolation of the effective Gaunt factors below threshold energies in the classical path calculation of the elastic contribution to the broadening. The present comparison appears to indicate that for the higher ionization stages, extrapolation of ḡ as a constant equal to its threshold value, is satisfactory.

The role of the oscillator strength in the determination of plasma densities

Abstract The problem is considered of deducing the electron density of a plasma from the widths of isolated, optically thin lines radiated by the ionic species. From the close connection between the line width calculated in the impact approximation and the atomic oscillator strengths for the collision-induced transitions which contribute to the width, it follows that accurate knowledge of these oscillator strengths is important, particularly in those cases where a few transitions are predominant. This, in turn, implies that careful consideration should be given to the coupling scheme which is used to describe the configuration of the atomic electrons. An important example of a situation where strong deviations occur from the commonly used LS coupling is the case of a single outer electron highly escited above the energies of the inner electrons. Although a form of pair coupling should, in general, be used in such a situation, special cases arise for which LS coupling is still valid. Criteria are derived for distinguishing these cases, an example of which is the 4041A˚line of singly-ionised nitrogen. The calculation of the Stark width of this line is discussed in detail and the result compared with a recent measurement. An example of a line whose width cannot be calculated in pure LS coupling is the 4552A˚line of singly-ionised nitrogen. This case is also evaluated and the result compared with experiment.

Estimates of Stark broadening of some Si(II) lines

Abstract Recent measurements of spectral line widths for the first five multiplets in the visible of singly-ionized silicon, are compared with the corresponding values predicted by the semi-empirical method of Griem, as well as two other Stark broadening theories. Although there is considerable disagreement between different measurements, as well as with and between the theories involved, agreement to better than 30% on average is obtained between these calculations and the computations of Sahal-Brechot. Some of the experimental and theoretical questions involved are discussed and suggestions are made for future work on the subject.

On the stark broadening of isolated lines of F(II) and Cl(III) by plasmas

Abstract Recent measurements of Stark widths of isolated lines emitted by singly-ionized fluorine atoms and doubly-ionized chlorine atoms are compared with the corresponding values calculated by the semiempirical impact approximation of Griem. A discussion is given of some of the difficulties which arise particularly when this method is applied to lines from multiply charged ions. On the basis of these comparisons, some new values are proposed for the threshold Gaunt factors of the ions under consideration, and these are compared with values deduced earlier for a number of other ions. Some systematic trends are observed.

Stark broadening of isolated lines emitted by singly-ionized tin

Abstract Calculations have been performed on the electron impact broadening of isolated lines from singly-ionized tin from a cool plasma. These have been compared with results of measurements recently performed by Miller, Roig, and Bengtson on a plasma produced by a conventional shock tube, and satisfactory overall agreement has been obtained. Our method is similar in procedure to that of the semi-empirical approximation of Griem, but we take into account the details of the radiator structure, including configuration interaction for two of the terms. We propose a new expression from which the relevant effective Gaunt factors for singly-charged ions can be calculated for incident energies above threshold. In some cases, a significant variation of line width within multiplets is noted. Some difficulties have been experienced with transitions to states involving equivalent electrons.

Estimates of Stark broadening of some Ar(III) and Ar(IV) lines

Abstract Recent measurements of Stark widths for doubly-and triply-ionized argon atoms are compared with the corresponding values predicted by the semi-empirical method of Griem. The results tend to confirm the basic validity of this approach and lend support for such attempts to refine and extend the method to the multiply-charged ions. On the basis of these comparisons, some new values are proposed for the threshold Gaunt factors of the ions under consideration.

ION DYNAMICAL CORRECTIONS TO THE HOLTSMARK THEORY OF SPECTRAL LINE BROADENING

Summary A method is described whereby earlier work by Kogan and Griem on ion dynamical corrections to the Holtsmark theory of spectral line shapes may be generalized to include the effects of Debye shielding and ion-ion correlations. Expressions for the dynamical corrections to the Holtsmark profile of a single Stark component are obtained to second order in the inverse Debye length. The correction terms are evaluated in the case of a high-density plasma (Ne ∼ 1017 cm−3), and the results used to discuss the central structure of the hydrogen Balmer β line profile. On the basis of our model, it is concluded that allowance for ion motion, with or without shielding, cannot account for outstanding discrepancies between the calculated and measured central structure of this line, under these conditions.

Oscillator strengths for six multiplets of neutral oxygen

Abstract Calculated and measured absorption oscillator strengths from the literature are compared for multiplets 1, 3, 3.01, 4, 5, and 6 of neutral oxygen. Considerable discrepancies are found in some cases between the values published by various authors. This comparison suggests the need for new high-precision measurements on this atomic species.