1: %\documentstyle[preprint,eqsecnum,aps,psfig]{revtex}
2: \documentstyle[eqsecnum,aps,psfig]{revtex}
3: \def\btt#1{{\tt$\backslash$#1}}
4: \def\etal{{\it et\thinspace al.}\ }
5: \def\eion{{(e~+~ion)}\ }
6: \def\efe17{{\rm (e~+~Fe~XVII)}}
7: \def\fe17{{\rm Fe~\sc xvii}}
8: \def\eie{{electron impact excitation}}
9: \def\cc{{close coupling}}
10: \def\bprm{{Breit-Pauli R-matrix}\ }
11: \def\apjl{{Astrophys. J. (Lett.)}}
12: \def\lr{{$\longrightarrow$}\ }
13: \def\ra{{$\rightarrow$}\ }
14: \def\ll{{$\lambda$}\ }
15: \newcommand{\abi}{{\it ab~initio}\ }
16: \newcommand{\lettt}{{\it this Letter}\ }
17: \newcommand{\rC}{{\rm 3C}}
18: \newcommand{\rD}{{\rm 3D}}
19: \newcommand{\rE}{{\rm 3E}}
20: \newcommand{\rF}{{\rm 3F}}
21: \newcommand{\ro}{{\rm o}}
22: \begin{document}
23: \draft
24: \preprint{HEP/123-qed}
25: \title{Influence of Resonances on Spectral Formation of X-Ray
26: Lines In \fe17}
27: \author{Guo Xin Chen and Anil K. Pradhan}
28: \address{
29: Department of Astronomy, The Ohio State University, Columbus, Ohio
30: 43210\\
31: }
32: \date{\today}
33: \maketitle
34: \begin{abstract}
35: New theoretical results from large-scale relativistic close coupling
36: calculations reveal the precise effect of resonances in collisional
37: excitation of {\sc x}-ray lines of Ne-like \fe17. Employing the Breit-Pauli
38: R-matrix method and a 89-level eigenfunction expansion including up to
39: $n = 4$ levels shows significant resonance enhancement of
40: the collision strengths of forbidden and intercombination transitions.
41: The present results
42: differ from all previous calculations, heretofore without detailed
43: resonance structures,
44: and should help resolve longstanding discrepancies. In particular,
45: the present line ratios of three benchmark diagnostic lines 3C, 3D, and 3E
46: at 15.014, 15.265, and 15.456 $\AA$ respectively,
47: are in excellent agreement with two independent measurements on
48: Electron-Beam-Ion-Traps [Laming \etal, Astrophys.~J {\bf 545}, L161~(2000)
49: and Brown \etal, Astrophys.~J {\bf 502}, 1015~(1998)]. The strong
50: energy dependence due to resonances in these and other cross sections
51: is demonstrated for the first time. It is of general importance and
52: strongly manifests itself in {\sc x}-ray plasma diagnostics.
53:
54: \end{abstract}
55: \pacs{PACS number(s): 34.80.Kw}%, 32.80.Dz, 32.80.Fb}
56:
57: %\narrowtext
58: Prominent \fe17 {\sc x}-ray lines have long been observed from laboratory and
59: astrophysical sources [e.g. \cite{rs,smi,hut}],
60: most recently in the {\it Chandra} and the {\it
61: XMM-Newton} spectra of active galactic nuclei,
62: {\sc x}-ray binaries, supernovae, and flaring and non-flaring active stellar
63: coronae \cite{can,sab}.
64: In spite of a number of theoretical, experimental, and observational
65: studies over decades there remain outstanding
66: discrepancies related to atomic processes and astrophysical effects
67: responsible for \fe17 spectral formation in high temperature sources [1-15].
68: However, the interpretation of experimental and observational ratios of line
69: intensities usually relies on collisional-radiative (C-R) models using
70: theoretical cross sections that neglect the fundamental role of
71: resonant excitation, which preferentially affects the forbidden and
72: intercombination transitions as opposed to dipole allowed ones.
73: Of particular interest are three prominent {\sc x}-ray transitions to the
74: ground level 1s$^2$2s$^2$2p$^6$~$^1$S$_0$ from excited levels:
75: 3C ($\lambda$~15.014$\AA$)
76: 1s$^2$2s$^2$2p$^5$[1/2]3d$_{3/2}$~$^1$P$^\ro_1$
77: (level 27),
78: 3D $\lambda$~15.265$\AA$
79: 1s$^2$2s$^2$2p$^5$[3/2]3d$_{5/2}$~$^3$D$^\ro_1$ (level
80: 23), and
81: 3E $\lambda$~15.456$\AA$:
82: 1s$^2$2s$^2$2p$^5$[3/2]3d$_{5/2}$~$^3$P$^\ro_1$ (level
83: 17), (level numbers denote energy order relative to the ground level 1).
84: While the 3C is dipole allowed, the
85: 3D and 3E are spin-forbidden intercombination transitions.
86: All but one of the previous calculations used the distorted wave
87: (DW) approximation that neglects channel coupling, and hence resonances \cite{zha,gol,chem,bha}.
88: A point of confusion has been that the only previous coupled
89: channel (CC) calculation \cite{moh}
90: also yielded the same results as the DW calculations \cite{bha}, because the
91: CC R-matrix calculations were at energies {\it above all target levels
92: in the eigenfunction expansion}; and therefore resonances were not
93: included. Since both the DW and the CC results agreed, it has hitherto
94: been assumed that resonance effects are not important, and that the cross
95: sections are relatively constant with energy (e.g. \cite{bro,bro1}).
96: On the other hand,
97: the experiments on Electron-Beam-Ion-Traps (EBIT) at Lawrence Livermore
98: National Laboratory (e.g. \cite{bro}), and at the National Institute of
99: Standards and Technology (NIST, \cite{lam}),
100: do not measure absolute excitation cross sections
101: directly. Rather, they measure relative intensity ratios R1=3C/3D
102: and R2=3E/3C at a few selected energies. Therefore the presence
103: of resonances, and more generally the energy variations, are not readily
104: discernible in experimental data (discussed later). The measured
105: values differ from theoretical ones by up to 50\% for R1, and a factor of
106: 2 for R2 \cite{lam,bro,bro1,bro2,can}. In order to accout for these discrepancies several
107: mechanisms based on atomic and astrophysical effects have been
108: put forward, such as polarization, resonance scattering, and
109: dielectronic satellite blending, which may be of varying importance
110: under appropriate conditions \cite{lam,bro,sab,wal}.
111:
112: However, as evident from the results in this {\it Letter},
113: there are extensive resonance
114: structures in cross sections for all transitions in \fe17 due to many
115: Rydberg series of resonances converging on to a number of $n=3$ and $n=4$
116: levels. Many infinite and interacting series of resonances arise due to
117: coupling between open and closed scattering channels
118: which, in principle, must be included in order to obtain the cross
119: sections precisely. But for a complex ion such as \fe17 the number
120: of channels is very large and the CC calculations become enormously
121: difficult, especially since relativistic fine structure
122: must also be considered in addition to other atomic effects. Using the
123: Breit-Pauli R-matrix (BPRM) method \cite{ip,ben95} we construct a large
124: eigenfunction
125: expansion including 89 levels corresponding to 49 LS terms up to the
126: $n=3$ and the $n=4$ complexes of \fe17. Full details of the
127: calculations will be presented elsewhere, but we briefly outline these below.
128: The coupled-channel wavefunction expansion for the
129: \efe17 system may be expressed as
130: %\begin{equation}
131: $\Psi(E; {\rm e + \fe17}) = \sum_{i} \chi_{i}(\fe17)\theta_{\rm e}(\ell_i) +
132: \sum_{j} c_{j} \Phi_{j}({\rm Fe}~{\sc xvi})$,
133: %\end{equation}
134: where the $\Psi$ denote the continuum (E $>$ 0)
135: states of given each total angular momemtum and parity J$\pi$,
136: expanded in terms of the core ion eigenfunctions
137: $\chi_i(S_iL_iJ_i)$; the $\theta_{\rm e}(\ell_i)$ refer to the free-electron
138: partial wave, and the $\Phi_j$ are short-range correlation functions that
139: also serve to compensate for orthogonality constraints. The 89 levels
140: belong to the configurations $
141: 2s^22p^6 ,\ 2s^22p^5 (3s,3p,3d),\ 2s^22p^5 (4s,4p,4d,4f),
142: \ 2s^12p^6 (3s,3p,3d),\ 2s^12p^6 (4s,4p,4d,4f)$. We consider total
143: symmetries 2J$ \leq 51$ of both parities explicitly in the BPRM
144: calculations. The size of the calculations (possibly the largest
145: \eion scattering calculations to date) may be gauged by the fact
146: that the dimension of the Hamiltonian matrices ranges up to 10286, for J
147: = 3.5 with 395 free channels and 486
148: bound channels ($\Phi_j$) (the largest number of free channels per symmetry
149: is 401); 25 continuum basis functions are used to
150: represent the $\Psi(\rm e~+~\fe17)$ in the
151: inner R-matrix region. Relativistic distorted wave (RDW) and
152: Coulomb-Born-Bethe approximations are employed to `top-up' the partial
153: wave summations. Particular attention is paid to the resolution of
154: resonances, with cross sections computed at about 20,000 energies.
155:
156: Figs.~1(a-c) display the dense resonance structures in
157: excitation collision strengths for the 3C, 3D, and 3E transitions
158: respectively. The results are compared with the RDW values (filled
159: circles, \cite{zha}) that
160: essentially represent previous DW calculations (such as from
161: the HULLAC code \cite{gol,chem}), and the one {\it non-resonant} value from
162: the previous CC R-matrix calculation (filled square, \cite{moh}). As seen
163: from the figures, the other
164: data correspond roughly to the `background' collision strength compared
165: to the present detailed results. The energy behavior of the three
166: transitions may be ascertained from the eigenfunction expansions of the
167: upper levels: 3C: $0.7857|27\rangle+0.1753|23\rangle+0.0305|17\rangle$,
168: 3D: $0.7479|23\rangle+0.2010|27\rangle+0.0491|17\rangle$, and 3E:
169: $0.9150|17\rangle+0.0767|23\rangle+0.0030|27\rangle$, where the eigenkets
170: correspond to the energy level indices. As the mixing coefficients
171: indicate the 3D has a significant component of $|^1$P$^{\rm o}\rangle$, but
172: the 3E considerably less so. Therefore the 3D would depart from LS coupling
173: to intermediate coupling to jj-coupling schemes along the neon
174: isoelectronic sequence with Z, {\it and with $\Omega(\rD)$ increasing with
175: energy} like $\Omega(\rC)$.
176: On the other hand the 3E remains largely a (spin) forbidden
177: transition and $\Omega(\rE)$ decreases with energy. The Einstein A-coefficients
178: of the 3C, 3D, and 3E are $2.47 \times 10^{13}, 6.01 \times 10^{12}$, and
179: $9.42 \times 10^{10}$ sec$^{-1}$ respectively. Other interesting atomic
180: physics aspects of these and other transitions in neon-like ions will be
181: discussed in later publications, with particular reference to laser
182: transitions.
183:
184: Spectral formation in laboratory and astrophysical plasmas needs
185: to be distinguished from each other. While we describe the latter using a
186: C-R model for \fe17, generally using
187: collision strengths averaged over a Maxwellian characterizing a
188: temperature, the former are measured using a mono-energetic beam with a
189: certain veocity spread assumed to be a gaussian. We compute the
190: collision strengths averaged three ways: (i) Maxwellian, (ii) gaussian,
191: and (iii) numerical. Present results for (ii) and (iii) are shown in
192: Figs.~1(a-c). In order to compare
193: with EBIT experiments the gaussian FWHM is
194: taken to be 30 eV. We note that the energy variations, dependent
195: on resolution and density of resonances in a given energy region, are
196: reflected in the gaussian averages (ii); the numerical averages are
197: slowly varying.
198: The theoretical line ratios using gaussian and numerical
199: averages $\langle\sigma\cdot v\rangle$ (GA and NA respectively; $\sigma$
200: denotes the cross section), are computed taking account of the
201: radiative branching ratios for 3C, 3D, and 3E which are: 1.0, 1.0, and 0.89
202: respectively. Therefore we have
203: %\begin{eqnarray}
204: $R1=\rC/\rD=\langle\sigma_{\rC}\cdot v\rangle/\langle\sigma_{\rD}\cdot
205: v\rangle$ and
206: %\hskip 1in
207: %\end{eqnarray}
208: %\begin{eqnarray}
209: $R2=\rE/\rC= 0.89 \langle\sigma_{\rE}\cdot v\rangle/\langle\sigma_{\rC}\cdot
210: v\rangle$.
211: %\end{eqnarray}
212: The results are compared with several EBIT measurements in Table I.
213: These direct
214: theoretical results agree with measured values to within experimental errors.
215: Although our calculations extend up to the $n = 4$ levels, the
216: highest threshold (level 89) is at $\sim$84.5 Ryd or 1.15 keV;
217: no resonances above
218: this energy are therefore included. While the line ratios at 0.85 keV and
219: 0.9 keV are computed directly from the collision strengths as in Figs.~1(a-c),
220: some extrapolation of the averaged values is made to compare with
221: the experimental value at 1.15 keV since we expect similar resonance
222: enhancement above the $n = 4$ due to still higher thresholds.
223: The GA results differ rather more
224: than the NA results but still agree with experiment, in contrast to
225: all other theoretical values. However,
226: there is significant variation with energy in the GA collision strengths
227: as they `oscillate' irregularly with energy depending on the density of
228: resonances prevalent in a given range. This
229: appears to be reflected in similar `oscillatory' structure seen in the
230: experimental values reported in \cite{bro2} throughout the energy range
231: 0.1 - 4 keV.
232: %\newpage
233: \begin{table}
234: \caption{Comparison of present line ratios for R1=3C/3D and R2=3E/3C with EBIT
235: measurements}
236: \begin{center}
237: \begin{tabular} {ccccc}
238: &&$E_i$=0.85 keV&0.9 keV& 1.15 keV\\
239: \hline
240: &EBIT&2.77$\pm$0.19$^a$&2.94$\pm$0.18$^b$&(3.15$\pm$0.17,2.93$\pm$0.16)$^a$\\
241: R1=3C/3D&Theory$^c$:NA&2.80&3.16&3.05$^\dag$\\
242: &Theory$^c$:GA&2.95&3.27&3.10$^\dag$\\
243: &Other Theory&\multicolumn{3}{c}{3.78$^d$;4.28$^e$;3.99$^f$}\\
244: \hline
245: &EBIT&&0.10$\pm$0.01$^b$&\\%0.09$\pm$0.01$^b$\\
246: R2=3E/3C&Theory$^c$:NA&0.11&0.085&0.07$^\dag$\\
247: &Theory$^c$:GA&0.11&0.083&0.07$^\dag$\\
248: &Other Theory&\multicolumn{3}{c}{0.04$^d$;0.05$^e$;0.05$^f$}\\
249: \end{tabular}
250: \end{center}
251: $^a$ EBIT experiments at LLNL \cite{bro};
252: $^b$ EBIT experiments at NIST \cite{lam};
253: $^c$ present theory with NA and GA;
254: $^d$ \cite{zha};
255: $^e$ \cite{bha};
256: $^f$ \cite{moh};
257: $^\dag$present values with extrapolation of resonance enhancement from
258: \abi collision strengths from E $\leq$ 1.02 keV (see text).
259: \end{table}
260: We also find that it is the $n = 4$ resonances, rather than the $n = 3$,
261: that dominate the enhancement of electron impact cross sections in
262: \fe17. The $n = 4$
263: resonances begin with resonant configurations $2s^22p^53\ell4\ell'$
264: that manifest themselves from $\sim$47 Ryd, considerably {\it
265: below} the excitation thresholds of the 2p \lr 3d lines 3C, 3D, and 3E
266: at $\sim$60 Ryd. Therefore there are relatively fewer $n = 3$
267: resonances and the $n = 4$ resonances greatly influence the
268: near-threshold behavior of these cross sections. In the region
269: 75 - 84.5 Ryd the resonances are
270: due to thresholds corresponding to two-electron-excitation
271: configurations $2s2p^64\ell$ with weakly coupled channels and
272: resonances.
273: This region is expected to be dominated by $n > 4$ resonances.
274: It might also be mentioned that radiation
275: damping of autoionizing resonances in cross sections
276: of \fe17 is known to be negligible \cite{pz97,petal,znp}.
277:
278: We note a few specific cases that exemplify the
279: findings in this {\it Letter}. An application of the present rates would
280: be to estimate more preciesly the
281: degree of resonance scattering of the 15.014 $\AA$ line in the solar
282: and stellar coronae as discussed in \cite{lam,bro1}. Based on our
283: C-R model with the present collisional data and
284: extensive new radiative calculations for \fe17 transition
285: probabilities using the BPRM method (as in \cite{np})
286: and from the SUPERSTRUCTURE code \cite{eis}, we obtain the 3C/3D ratio
287: to vary between 2.63 - 3.20, which agrees with the values from EBIT
288: and from flaring solar corona, but is still higher than the
289: observed value \cite{wal} of 1.87$\pm$0.21 from non-flaring active
290: region. This further suggests that
291: resonance scattering of 3C, or possibly some other mechanism, may be operative
292: under astrophysical conditions \cite{lam}. Another example is the
293: discussion of the diagnostic
294: utility of the 3C/3D line ratio by Brown \etal \cite{bro1} who (a) assume the
295: cross sections to be relatively constant, and (b) mutiply the RDW cross
296: section of Zhang and Sampson by 1.25 to agree with the measured value.
297: From the present work neither (a) nor (b) are necessary. Furthermore,
298: Brown \etal obtain the temperature dependence of 3C/3D including an
299: inner-shell satellite line blended with the 3D, but using constant
300: cross sections. A revised analysis with the present
301: cross sections should yield a different temperature diagnostics.
302:
303: More generally, the excitation of most of the
304: $n = 3$ levels of \fe17 is similarly affected. Using our
305: 89-level C-R model, with newly computed collsional and radiative
306: data employing the BPRM method, we investigated the prominent transitions
307: corresponding to the `coronal' {\sc x}-ray lines $2s^22p^53s$ \lr $2s^22p^6$ at
308: \ll\ll $16.780, 17.055$ and $17.100~\AA$ (e.g. \cite{mlf}). Fig.~2(a) shows the
309: detailed collision strength for the forbidden J = 1 \lr 0 \ll 16.780 $\AA$
310: (3F) transition, compared with previous DW values (filled circles
311: \cite{zha} and square \cite{bha}).
312: The forbidden transitions are expected to be most enhanced by resonances,
313: as in Fig.~2(a), since the background cross sections are much smaller than
314: for allowed transitions.
315: The temperature dependence of the forbidden (3F) to the allowed (3C)
316: line ratio R3 = 3F/3C = I(16.780)/I(15.014)
317: is demonstrated in Fig.~2(b), and compared
318: with that calculated using DW cross sections (filled
319: squares \cite{bha}); at higher temperatures the resonance
320: contribution decreases progressively with energy due to the Maxwellian
321: factor. The electron density dependence is small;
322: solid-line and dot-line correspond to 10$^{13}$ and 10$^9$ cm$^{-3}$
323: respectively. The 4 open
324: circles with error bars are observed and experimental values.
325: At all temperatures T $< 10^7$ K the present line ratio departs considerably
326: from those using DW data without resonances, to more than a factor of 3
327: at about 10$^6$ K---a fact of considerable importance in
328: photoionized {\sc x}-ray plasmas that have
329: temperatures of maximum abundance much lower than that in coronal
330: equilibrium T$_m \sim 4 \times 10^6$ K for \fe17, as marked.
331: \begin{table}
332: \caption{Rate coefficients at
333: $T_{\rm e}$=200 eV for resonant excitation to the $2s^22p^53s$ levels}
334: \begin{center}
335: \begin{tabular} {ccccc}
336: final&\multicolumn{4}{c}{Rate Coefficient ($\times10^{-13}$cm$^3$sec$^{-1}$)}\\
337: state&Present&Smith \etal~\cite{smi}&Goldstein \etal~\cite{gol}&Chen\&Reed~\cite{chem}\\
338: \hline
339: $2p^53s~^3$P$^{\rm o}_1$&22.7&48.0&15.12&14.2\\
340: All 2p$^5$3s states&70.4&147.0&46.68&42.7
341: \end{tabular}
342: \end{center}
343: \end{table}
344: Resonant excitation may be considered indirectly by approximate methods
345: such as quantum
346: defect theory \cite{znp} or isolated resonance approximations (IRA)
347: \cite{smi,gol,chem}.
348: Table II compares the present rate coefficients for the individual 3F line,
349: and for all transitons to the $2s^22p^53s$ levels, with IRA calculations
350: using HULLAC \cite{gol}, multi-configuration Dirac-Fock \cite{chem}, and
351: semi-relativistic Hartree-Fock \cite{smi} methods.
352: The present data are
353: $\sim$50\% higher and differ from \cite{gol,chem} which
354: (i) neglect interference between resonances, and (ii)
355: take limited account of the several infinite Rydberg series of
356: resonances resulting in a considerable underestimate of resoance enhancement.
357: The imprecision of the IRA is also indicated by more than a factor of 3
358: discrepancy between \cite{chem} and \cite{smi}, both
359: using IRA but with different decay channels. Present results
360: are about a factor of 2 lower than \cite{smi}.
361: Although resonances are generally less important for highly charged ions,
362: the resulting enhancement in the BP R-matrix cross sections for
363: forbidden and intercombination lines, compared to the allowed line,
364: yields the ratios R1 and R2 closer to experiments compared to other works.
365: We also note that ionization and recombination processes are not
366: included in the present C-R model; dielectronic recombination from
367: Fe~XVIII to Fe~XVII may also contribute to line emissions.
368:
369: The main conclusions of this {\it Letter} are:
370: (I) resonances are present in \fe17 collision cross sections at all
371: energies, (II)
372: the energy dependence of different types of forbidden, intercombination,
373: and allowed spectral transitions is demonstrated, (III) the effect on
374: diagnostic line ratios is significant, and may be up to several factors
375: compared to previous data neglecting resonance effects, (IV)
376: the present work is expected to be generally applicable to the
377: temeprature and density analysis
378: of many astrophysical and laboratory {\sc x}-ray sources (especially
379: photoionized plasmas and forbidden lines) not only for \fe17
380: but also other neon-sequence ions of importance in {\sc x}-ray
381: lasers \cite{ros}, (V) this work is part of a large-scale
382: program to compute excitation, photoionization, \eion recombination
383: \cite{petal,znp}, and
384: transition probabilities of \fe17 using the accurate BPRM method; for
385: example, a set of A-values for $\sim$30~000 transitions
386: have been computed.
387:
388:
389: We would like to thank Dr.~Werner Eissner for assistance with
390: the Iron Project BPRM codes.
391: This work was partially supported by the National Science Foundation
392: and the NASA Astrophysical Theory Program. The computational work was
393: carried out utilising well over 1000 CPU hours on the Cray YMP, T94, and
394: SV1 at the Ohio Supercomputer Center in Columbus Ohio.
395:
396: \def\amp{{Adv. At. Molec. Phys.}\ }
397: \def\apj{{ Astrophys. J.}\ }
398: \def\apjs{{Astrophys. J. Suppl.}\ }
399: \def\apjl{{Astrophys. J. (Lett.)}\ }
400: \def\aj{{Astron. J.}\ }
401: \def\aa{{Astron. Astrophys.}\ }
402: \def\aas{{Astron. Astrophys. Suppl.}\ }
403: \def\adndt{{At. Data Nucl. Data Tables}\ }
404: \def\cpc{{Comput. Phys. Commun.}\ }
405: \def\jqsrt{{J. Quant. Spectrosc. Radiat. Transf.}\ }
406: \def\jpb{{J. Phys. B}\ }
407: \def\pasp{{Pub. Astron. Soc. Pacific}\ }
408: \def\mn{{Mon. Not. R. Astron. Soc.}\ }
409: \def\pra{{Phys. Rev. A}\ }
410: \def\ps{{Phys. Scr.}\ }
411: \def\prl{{Phys. Rev. Lett.}\ }
412: \def\zpds{{Z. Phys. D Suppl.}\ }
413: \def\adndt{At. Data Nucl. Data Tables}
414:
415: \begin{references}
416: \bibitem{lam} J.M. Laming \etal, \apjl, {\bf 545}, L161~(2000)
417: \bibitem{bro}G.V. Brown, P. Beiersdorfer, D.A. Liedahl, and K. Widmann,
418: \apj, {\bf 502}, 1015~(1998)
419: \bibitem{rs} J.C. Raymond and B.W. Smith, \apj, {\bf 306}, 762~(1986); N.S. Brickhouse,
420: J.C. Raymond, and B.W. Smith, \apjs, {\bf 97}, 551~(1995)
421: \bibitem{smi} B.W. Smith, J.C. Raymond, J.B. Mann, and R.D. Cowan, \apj, {\bf 298}, 898~(1985)
422: \bibitem{hut}R.F. Hutcheon, F.P. Paye and K.D. Evans, \mn, {\bf 175}, 489~(1976)
423: \bibitem{can}C.R. Canizares \etal, \apjl, {\bf 539}, L41~(2000)
424: \bibitem{sab} J.L.R. Saba, J.T. Schmelz, A.K. Bhatia, and K.T. Strong, \apj, {\bf 510},
425: 1064~(1999);
426: \bibitem{zha}H.L. Zhang and D.H. Sampson, \adndt, {\bf 43}, 1~(1989)
427: \bibitem{gol}W.H. Goldstein, A. Osterheld, J.Oreg, and A. Bar-Shalom, \apjl, {\bf 344}, L37~(1989)
428: \bibitem{chem}M.H. Chen and K.J. Reed, \pra, {\bf 40}, 2292~(1989)
429: \bibitem{bha}A.K. Bhatia and G.A. Doschek, \adndt, {\bf 52}, 1~(1992);
430: M.Cornille, J. Dubau, F. Faucher, F. Bely-Dubau, and C. Blancard, \aas, {\bf 105}, 77~(1994);
431: O. Bely and F. Bely, Sol. Phys., {\bf 2}, 285~(1967)
432: \bibitem{moh}M. Mohan, R. Sharma, and W. Eissner, \apjs, {\bf 108}, 389~(1997)
433: \bibitem{bro1} G.V. Brown, P. Beiersdorfer, H. Chen, and K.J. Reed, \apjl, {\bf 557}, L75~(2001)
434: \bibitem{bro2} G.V. Brown, P. Beiersdorfer, and K. Widmann, \pra, {\bf 63}, 032719~(2001)
435: \bibitem{wal}K. Waljeski, \etal, \apj, {\bf 429}, 909~(1994)
436: \bibitem{ip} D.G. Hummer, K.A. Berrington, W. Eissner, A.K. Pradhan,
437: H.E. Saraph and J.A. Tully, Astron. Astrophys. {\bf 279}, 298~(1993).
438: \bibitem{ben95} K.A. Berrington, W. Eissner, and P. H. Norrington, \cpc
439: {\bf 92}, 290~(1995).
440: \bibitem{pz97}A.K. Pradhan and H.L. Zhang, \jpb {\bf 30}, L571~(1997).
441: \bibitem{petal} A.K. Pradhan, S.N. Nahar, and H.L. Zhang, \apjl {\bf 549}, L265~(2001).
442: \bibitem{znp} H.L. Zhang, S.N. Nahar and A.K. Pradhan, \pra {\bf 64}, 032719-1~(2001).
443: \bibitem{np} S.N. Nahar and A.K. Pradhan, Physica Scripta {\bf 61}, 675~(1999).
444: \bibitem{eis} W.B. Eissner, M. Jones and H. Nussbaumer, \cpc, {\bf 8}, 270~(1974)
445: %W.B. Eissner, \cpc, {\bf 114}, 295~(1998)
446: \bibitem{mlf} C.W. Mauche, D.A. Liedahl, and K.B. Fournier, astro-ph/0106518~(2001).
447: %\bibitem{che} G.X. Chen and A.K. Pradhan, \jpb, {\bf 32}, 1809~(1999)
448: \bibitem{ros}M.D. Rosen, P. Hagelstein, D.L. Matthews, E.M. Campbell, A.U. Hazi,
449: B.L. Whitten, B. MacGowan, R.E. Turner, R.W. Lee, G. Charatis,
450: G.E. Busch, C.L. Shepard, and P.D. Rochett, \prl {\bf 54}, 106~(1985)
451:
452: %\bibitem{zha1} H.L. Zhang, S.N. Nahar and A.K. Pradhan, \pra {\bf 64}, 032719-1~(2001);
453: %A.K. Pradhan, S.N. Nahar, and H.L. Zhang, \apjl {\bf 549}, L265~(2001);
454: %A.K. Pradhan and H.L. Zhang \jpb {\bf 30}, L571~(1997).
455: \end{references}
456:
457: \newpage
458: % Fig.1
459: \begin{figure}
460: \centering
461: \psfig{figure=fig1.eps,height=17.0cm,width=18.0cm}
462: \caption{\bprm collision strength $\Omega$ for 3C, 3D and 3E lines
463: with detailed resonance structures as a function of incident electron energy;
464: filled dots are RDW results. The gaussian (FWHM = 30 eV) and
465: numerical averages are also shown (see text).}
466: \end{figure}
467:
468: \begin{figure}
469: \centering
470: \psfig{figure=fig2.eps,height=17.0cm,width=18.0cm}
471: \caption{(a): \bprm collision strength $\Omega$ for the 3F line;
472: the filled circles and square are non-resonant DW calculations;
473: (b): line ratio 3F/3C vs.~T from a 89-level C-R model.
474: The electron densities for
475: solid-line and dot-line curves are 10$^{13}$ and 10$^9$ cm$^{-3}$ respectively.
476: The 4 open circles with error bars are observed and experimental values: from the solar corona
477: at T$_m \sim$~4MK \protect\cite{hut},
478: from the corona of solar-type star Capella at $\sim$~5MK \protect\cite{can},
479: and from the EBIT experiment at 0.9~keV (log T = 7) \protect\cite{lam}.
480: The filled squares are values using DW results \protect\cite{bha}.}
481: \end{figure}
482:
483: %*** E n d o f p a g e 6 o f g a l l e y - m o d e o u t p u t
484: %***
485:
486:
487: \end{document}
488:
489:
490:
491:
492: