The level 2s2p2(4P) of OIV is a metastable state and the line intensity Ii (2p3(4S) − 2s2p2(4P)) is mainly generated by the excitation from the metastable state 2s2p2(4P) to 2p3(4S). Then the line intensity ratio is Ii to the resonance line intensity Ir (2s2p2(2D) − 2s22p(2P)) has a dependence on electron density until the metastable level is saturated. The ratio Ii/Ir obtained from the measurements of JIPPT-IIU Tokamak plasmas has Been analyzed. It is found that the recombination from a metastable state of OV 2s2p(3P) to the metastable state of OIV 2s2p2(4P) is appreciable in high temperature plasmas where the abundance of OV is larger than that of OIV. The recombination rate coefficient between metastable states is determined from-the line intensity ratios; I(2p2(3P) − 2s2p(3P))/I(2s2p(1P) − 2s2(1S)) of OV, (2s2p2(2D) − 2s22p(2P)) and Ii/Ir of OIV. This recombination process consists of dielectronic recombination and radiative recombination.

We present new calculations for ls−2p transitions in C−, N−, and O-like Ti, Cr, and Ni for application to tokamak X-ray spectra. This extends our earlier work for the corresponding iron ions (Phillips et al. 1983, Ap.J. 265, 1120) in which a density dependence of line intensities was pointed out. These lines are principally formed by dielectronic recombination; the density effect arises due to the density-dependent populations of the fine structure levels in the ground configuration of the recombining ion. In the present work the calculations of the Ti, Cr, and Ni lines have been made for the wavelengths and the dielectronic recombination rates using a Hartree-Fock technique with statistical exchange. The density dependence of the collisionally excited transitions was also included. Density-dependent populations of the fine structure levels of the ground configuration were obtained from Feldman et al. (1980 J. Appl. Phys. 51, 190) and Bhatia et al. (1980, J. Appl. Phys. 51, 1464). In calculating the ionization fractions we have assumed coronal equilibrium; new ionization rate coefficients were calculated using the formalism given by Burgess and Chidichimo (1983). The spectra were synthesized with Voigt line profiles and typical tokamak radius-dependent temperature and density profiles were assumed. The N-like lines, in-particular show a strong dependence over tha range: <1011 −3.1013 cm−3 (Ti XVI) <1011 −3.1014 cm−3 (Cr XVIII), 1012 −1015 cm−3 (Ni XXII). If a density profile across the torus is available from Thomson-scattered laser radiation, observations of these lines may serve to show where the N-like ions are formed in the tokamak plasma.

Observations of cool stars with the IUE satellite have shown that although main sequence stars have uv spectra similar to that of the sun, the spectra of the giants and supergiants are dominated by lines of neutral atoms and singly charged ions. The atmospheres of the low gravity stars are greatly extended and there is a large opacity in the stronger lines. This leads to multiple scattering and the appearance of many transitions through line ‘leakage’.

A review is given of how high resolution IUE spectra have allowed line identifications to be established and some current problems are discussed.

The spectroscopic diagnostic techniques for the low gravity, low density atmospheres are also briefly outlined.

With the 500 and 1000 l/mm transmission gratings aboard the European x-ray Observatory SATellite (EXOSAT) we have measured medium-resolution (Δλ 5 A at 100 A) spectra of some ten objects of various categories such as isolated white dwarfs, cool stars with convective mantles, cataclysmic variables (e.g. AM Her) and a high-luminosity X-ray source (Sco X-1).

The Instrument configuration was mostly such that one low-energy telescope was used as a photometer, while the other telescope was used as a spectrometer with the 500 l/mm grating inserted.

The white dwarf spectra were measured between about 60 and 300 A. They show a continuum with no clear evidence of aborption and emission lines except for the He II absorption edge at 227 A in the spectrum of Feige 24. For the cooler (28 000 K) white dwarf Sirius B the emission is peaked between about 100 and 160 A and limited to about 200 A. which can be expected from atmospheric model spectra of DA white dwarfs. The soft X-ray emission of the hotter (> 60 000 K) DA white dwarfs (HZ43. Feige 24) is also interpreted in terms of photospheric emission. In the HZ43 spectrum the absorption edge is apparently absent which sets a stringent upper limit to the abundance ratio He/H of about 10−5. On the other hand the spectrum of Feige 24 shows a dominant absorption edge, implying He/H > 10−3. Moreover, here the shape of the continuum may be indicative of a stratification of element abundances in the outer atmosphere.

Optical spectra of nebulae and novae are dominated by recombination lines of H I, Be I and He II and by forbidden lines, such as those of [0 III]. Their IR spectra show the fine-structure lines and continua due to dust. Most of the lines observed in the UV (say from 3000 to 1200 Å) are either due to resonance transitions (e.g. C IV 1549) or are of intercombination type (e.g. C III] 1908). Some elements do not have many lines in the optical - nitrogen gives only [N II] and there are no forbidden lines of ionized carbon - and for such elements much valuable information is provided by UV lines (e.g. N III], N IV], N V, C II], C III] and C IV). Carbon/oxygen abundance ratios deduced from combined UV and optical observations agree well with those deduced from IR continuum features (Seaton, 1983).

We analyze 6–300Å observations of late-type stars obtained with EXOSAT. We discuss the problem of deriving source temperature and other physical parameters from broad-band EXOSAT observations.

For the past year a Joint Working Group of NASA and ESA scientists and engineers has been defining the scientific objectives and instrument parameters for a proposed satellite to obtain far and extreme ultraviolet spectra of stars, interstellar gas, solar system objects, and galaxies. The project, now called Columbus, incorporates the scientific goals of the previously proposed NASA Far Ultraviolet Spectrograph Explorer (FUSE) and ESA Magellan missions.

The prime spectral range of Columbus, 900–1200 Å, cannot be observed by IUE or Space Telescope. In this spectral range Copernicus was able to observe bright stars (mv ≤ 6) with high resolution and the Hopkins Ultraviolet Telescope (HUT) will observe faint sources at low resolution, but Columbus will be the first instrument capable of high spectral resolution observations of faint sources (mv ≈ 17). High resolution spectra in the 900–1200 Å region will permit studies of the Lyman lines of atomic H and D, the molecules H2 and HD, resonance lines of C III and O VI, and other species listed in Table 1. Columbus also is being designed to observe the 1200–2000 Å spectral region at high resolution, permitting measurements of many stages of ionization for the same atom (i.e. N I, II, III, V; C II, III, IV; and S II, III, IV, VI). The broad coverage of ionization states is essential for the analysis of interstellar and stellar plasmas where the ionization balance can be quite complex.

When a cool plasma is shock-heated to X-ray temperatures, the ionization structure does not attain its final, equilibrium value immmediately, but proceeds toward it through electron-ion collisions with a timescale τ ≡ net of order 1012 cm−3 sec. For supernova remnants (SNRs), where 0.1 ≤ ne ≤ 10 cm−3 typically, the time required to achieve collisional ionization equilibrium (CIE) can be greater than the age of the remnant. Even if the SNR is quite old, that part of the remnant which is emitting most of the X-rays may have been shocked relatively recently, so that the assumption of CIE may be inappropriate (see below).

The question of ionization equilibrium is of great astrophysical importance in the study of SNRs because it affects the deduced values of their masses and elemental abundances (e.g. Shull 1982). Mass determinations are affected because underionized plasma generally has a much higher emissivity in soft X-rays than equilibrium plasma. Unless this is accounted for, the deduced value of the density and therefore of the mass, will be considerably overestimated.

The presence of objective transmission gratings on the two recent X-ray telescope experiments, the Einstein and EXOSAT Observatories, has opened up a new avenue of research in X-ray astronomy by enabling us, for the first time, to obtain moderate-to-high resolution spectra of cosmic sources in the soft X-ray band (λ ~ 5–200 Å). Both experiments incorporated gold bar transmission gratings with line densities of 500 1/mm and 1000 1/mm which could be inserted into the X-ray optical path at the exit from a grazing incidence mirror. At short wavelengths, the resolution was determined principally by the spatial resolution of the detector-telescope combination. For Einstein, this was Δλ ~ 0.4 Å for the 1000 1/mm grating and ~ 0.8 Å for the 500 1/mm grating. For EXOSAT, the resolution was somewhat worse: Δλ ~ 1.5 Å for the 1000 1/mm grating, and ~ 3 Å for the 500 1/mm grating. (More complete descriptions of these instruments can be found in Seward et al. 1982 and de Korte et al. 1981.)

Spectroscopic analysis is a powerful technique for the diagnosis of temperatures and compositions of astrophysical plasmas. The EUV (100–1000Å) and soft x-ray (10–100Å) bands contain hundreds of potentially useful diagnostic lines. Unfortunately, traditional types of grating spectrometer become inefficient or unwieldy when adapted to stellar spectroscopy onboard a spacecraft. At grazing incidence, the required length of a high-resolution plane-grating spectrometer can easily exceed the length of the telescope feeding it. For these reasons, we have systematically explored ways to introduce a reflection grating into the converging beam formed by a given objective optical system ahead of its first focus. A spectrometer of this type results in an optical train no longer than the telescope’s existing prime-focus beam.

The mixed selection of strong emission lines present in the spectra of WR stars suggests that we are observing plasma with an electron temperature of the order of 105 K somewhere in the atmospheres of these rare stars. In the spectra of some WR stars emission lines of H are detected; this suggests that plasma with an electron temperature of the order of 104 K may be present also. Since the observations made in the last 30 years show that the masses, luminosities, effective temperatures, and general distribution in space of WR stars are similar to those of stars with spectral types in the range from about B2 to O9, a prime question is why are the spectra of WR stars so different from those of the B stars with which they are associated.

During the last few years, observations of solar phenomena have been carried out by rocket flights, manned satellites like Skylab, unmanned satellites like Orbiting Solar Observatories and more recently Solar Maximum Mission. The wavelengths, line intensities and line profiles in UV and X-ray regions of the solar spectra have been measured. The spectroscopic data obtained are of high spatial and spectral resolution. The goal is to understand the physical properties of the emitting plasma and determine the electron temperatures, densities and volume of the emitting plasma from UV and X-ray spectra.

Continuum emission due to bremsstrahlung and direct radiative recombination is studied in two types of fusion-plasmas with non-Maxwellian electron distributions. The calculated emissivity coefficients Jω(hω) are shown to depart from the Maxwellian ones. The main consequence of the modification of the spectrum is to invalidate the usual electron temperature diagnostic which consits in equating dlnjω(hω)/d(hω) to l/kTe.

Case I corresponds to the critical density region of a laser-produced plasma ( Langdon 1980, Mora and Yahi 1982 ). The distribution function is cm exp (-v/vo,m)m with m=5 for all values of the incident electron velocity v. Case II corresponds to very steep temperature gradients for which the spacial scale of the transport region is larger than the gradient scale itself (Albritton 1983, Luciani et al. 1984, Campbell 1984). On the higher temperature side of the heat front, the tail of the distribution is depleted with respect to the Maxwellian one. At a particular distance from it, this tail can be roughly described by the above analytical function but with m=3.

The process of dielectronic recombination has been the subject of intense theoretical activity mainly because of the pioneering work by Burgess (1964), who demonstrated that this process can be the dominant recombination mechanism for multiply-charged atomic ions in low-density high-temperature astrophysical and laboratory plasmas. Recent attempts to measure dielectronic recombination cross sections and rate coefficients have renewed interest in the development of a rigorous quantum-mechanical theory of the resonant electron-ion recombination process. A precise theory is clearly required for the interpretation of the dielectronic satellite lines, which have been found to be of great value in the spectroscopic determination of temperatures, densities, and departures from ionization equilibrium.

A comparison among nine computer codes shows surprisingly large differences where it had been believed that the computational physics was well understood. The codes simulate a plasma that is in steady state but not in local thermodynamic equilibrium (NLTE codes). In this study each code treats an “easy” problem, which is an argon plasma, optically thin and with no external photon flux; densities are varied from near-coronal to an intermediate 1021 electrons/cc and above. The temperatures are high enough that most ions have two or fewer bound electrons, which for this plasma means temperatures above about 300 eV.

The present study asks only if the codes compute similar charge-state populations (a surprising “no”), and, if not, why not. It does not claim accuracy for any code either by comparison to experiment or by appeal to a concensus.

It is well known that radiation from the levels of the ls2ℓ2ℓ' and ls2ℓ3ℓ' configurations of Li-like ions form satellites of spectral lines from He-like ions that are of high interest for plasma diagnostic purposes. Recently Sampson et al (1984) have made calculations of the collision strengths for innershell excitation to these upper levels from the ls22s and ls22p levels for a large number of Li-like ions using a Coulomb-Born-exchange method. Here we consider the alternative mechanism for populating the ls2ℓ3ℓ' levels by collisional excitation from the ls23s, ls23p and ls23d levels. Although the poDulations of the ls23ℓ levels are lower than those of the ls22ℓ levels, this is at least partially compensated for by the fact that the collision strengths for the ls23ℓ − ls2ℓ'3ℓ" transitions are considerably larger on the average than those for the ls22ℓ − ls2ℓ'3ℓ" transitions. The method of calculation is the same as that used by Sampson et al (1984), where all possible configuration mixing, parentage mixing and intermediate coupling type mixing among the states in a complex (the states having the same n values, parity and J values) was included. Also the present application is sufficiently simple that after summation over J values for the initial level the results for the collision strength Ω can be expressed in a simple manner in terms of the Z scaled hydrogenic collision strengths Z2ΩH and analogous to Eqs (8)-(15) of Sampson et al (1984). The appropriate formulae and the numerical results for both excitation to upper fine structure levels and to upper energy terms for 19 Li-like ions have been submitted for publication to Atomic Data and Nuclear Data Tables. Here we give only the pertinent formula and sample numerical results for transitions between energy terms in Li-like iron. This formula is

where due to configuration mixing neither ℓ need equal ℓ" nor ℓa need equal ℓ'. The recommended values for σ and σe are zero for transitions involving ls−2s and are −0.5 and 0.3, respectively, for transitions involving ls−2p. Numerical values for the coefficients A and Ae for Li-like iron are given in Tables I-III. Values for Z2ΩH and for impact electron energies up to 15 times threshold are given in Golden et al (1981) and Clark et al (1982) and are repeated in Table IV for convenience. Finally we note that the line strengths S for transitions involving ls−2p are obtained by multiplying the corresponding value for A(ls−2p) by the scaled hydrogenic line strength 1.665 and dividing by (Z-σ)2.

The formulas for mean wavelengths and widths of Unresolved Transition Arrays (UTA)[1,2], have been extended to include spectator electrons and jj transitions. These were used to interpret several sorts of satellite spectra, e.g., 3d10 nℓq− −3d94fnq transitions in atoms Tm thru W from laser produced plasmas, and 3dm4s − 3dm−1 4p4s in Mo and Pd spectra. The Unresolved Character of the UTA will be discussed thanks to an evaluation of the number of lines in the array. This will be applied to 4d-4f transitions in Ionized Rare Earths.

The breakthrough in 1978 by Brooks et al. and by Breton et al. provided the first plasma rate measurements on dielectronic recombination (DR), and their basic technique has been followed up on in a few other investigations. An important new plasma technique was introduced in 1982. Another breakthrough in 1982–1983 led to colliding beams cross section measurements for DR and to more direct comparisons with theory. Experiment/theory disagreements and agreements have led to follow-up experimental and theoretical efforts which are pointing to better understanding of DR and which emphasize issues that should be paid attention to by plasma modelers. An overview will be given with strong emphasis on the beams experiments and the implications.

For over ten years an expanding collaborative programme of experimental work on the absorption spectra of neutral atomic and simpler molecular gases and vapours, and on similar species “matrix-isolated” in rare-gase solids, has been in progress. The experimental work has centered on use of background continua from the 0.5 and 2.6 GeV synchrotrons in Bonn, - high resolution work being based mainly on the smaller machine. The instrumentation includes - most significantly for the work reported here - a 3-metre Eagle-style vacuum spectrograph equipped with “holographic” gratings of 5,000 – 6,000 1/mm spacing, and usable over the range 200 – 2,000Å.

It has been shown that is is possible to determine absolute wavelength values of X-ray radiation emitted as a result of electron transitions in ions or atoms, without precise Spectrometers and without reference “lines”. Knowledge of the lattice constants of the diffracting crystal is sufficient for the determination of the wavelengths. Here the first experiment describing the feasibility of the method is shown.

Usually spectra are obtained in form of “lines”. In the X-ray case, however, it is possible to obtain spectra in form of points, ordered along lines. The lines represent the white radiation [Fig. 1].