1: \documentclass[11pt,a4paper]{article}
2: \usepackage{cite,indentfirst}
3: \usepackage{amssymb,amsmath}
4: \usepackage[dvips]{epsfig}
5: \usepackage{epsf}
6: \usepackage{graphics}
7: %\usepackage[active]{srcltx}% for Linux Etex
8:
9: \topmargin -0.4cm
10: \headsep=0.0cm
11: \headheight=0.0cm
12: \textheight 24.6cm
13: \oddsidemargin -0.3cm
14: \evensidemargin -0.3cm
15: \textwidth 15.9cm
16:
17: \def\mprp{\mbox{\tiny $\bot$}}
18: \def\mprl{\mbox{\tiny $\|$}}
19: \def\th{\mbox{th}}
20:
21: \def\beq{\begin{eqnarray}}
22: \def\eeq{\end{eqnarray}}
23: \def\ee{\varepsilon}
24: \def\lm{\lambda}
25: \newcommand{\prp}[1]{#1_{\mbox{\tiny $\bot$}}}
26: \newcommand{\prl}[1]{#1_{\mbox{\tiny $\|$}}}
27: \def\HH{H\!\!\left(\frac{4 m^2}{\prl{q}^2}\right)}
28: \def\HHi{H\!\!\left(\frac{4 m^2}{\prl{q'^2}}\right)}
29: \def\HHii{H\!\!\left(\frac{4 m^2}{\prl{q''^2}}\right)}
30: \def\ggg{\gamma \rightarrow \gamma \gamma}
31: \def\ff{\Lambda}
32: \def\tff{\widetilde \Lambda}
33: \def\1{1 \to 1 \, 2}
34: \def\2{1 \to 2 \, 2}
35: \def\S{{\cal S}}
36: \def\J{{\cal J}}
37: \def\A{{\cal A}}
38: \def\P{{\cal P}}
39: \def\M{{\cal M}}
40:
41: \title{\vspace*{-20mm}
42: \begin{flushright}
43: {\normalsize Yaroslavl State University\\
44: Preprint YARU-HE-06/04\\
45: hep-ph/0609192} \\[10mm]
46: \end{flushright}
47: {\bf The Compton effect in strongly magnetized plasma}}
48:
49: \author{M. V. Chistyakov, D. A. Rumyantsev \\[3mm]
50: {\small\it Division of Theoretical Physics,} \\
51: {\small\it Yaroslavl State (P.G.~Demidov) University,} \\
52: {\small\it Sovietskaya 14, 150000 Yaroslavl, Russian Federation}\\
53: {\small\tt E-mail: mch@uniyar.ac.ru, rda@uniyar.ac.ru}
54: }
55:
56: \date{}
57:
58: \begin{document}
59:
60: \maketitle
61:
62: \thispagestyle{empty}
63:
64: \begin{abstract}
65:
66: \baselineskip=22pt The process of Compton scattering $\gamma e^{\pm}
67: \to \gamma e^{\pm}$ in strongly magnetized hot electron-positron plasma is considered.
68: The analytical expressions for the partial cross sections in rarefied plasma are obtained.
69: The numerical estimations of the absorption rates for various scattering channels
70: are presented with taking into account of the photon dispersion and wave function renormalization
71: in strong magnetic field and plasma. The comparison of the scattering absorption rate with photon splitting probability
72: shows the existence of the temperature range where these values are comparable with each other.
73: The astrophysical applications of the obtaining results are discussed.
74: \end{abstract}
75:
76: \section{Introduction}
77:
78: \unitlength 1mm
79: Recent observations~\cite{Kouveliotou:1998ze,Kouveliotou:1998fd,Gavriil:2002mc,Ibrahim:2002zw} give ground to believe that some astrophysical objects (SGR and AXP) are magnetars,
80: a distinct class of isolated neutron stars with magnetic field strength of
81: $B \sim 10^{14}-10^{16}$ G~\cite{Duncan:1992,Duncan:1995,Duncan:1996}, i.e. $B \gg B_e$, where
82: $B_e = m^2/e \simeq 4.41\times 10^{13}$~G~\footnote{We use natural units
83: $c = \hbar = k = 1$, $m$ is the electron mass, $e > 0$ is the elementary
84: charge.} is the critical magnetic field.
85: The spectra analysis of these objects is also providing evidence for the
86: presence of electron-positron plasma in magnetar environment. It is
87: well-known that strong magnetic field and/or plasma could influence
88: essentially on different quantum processes~\cite{Duncan:2000pj,Lai:2001,KM_Book,Harding:2006qn}. Therefore the effect of magnetized plasma on microscopic physics should be incorporated in the magnetosphere models of strongly magnetized neutron stars.
89:
90:
91: The various studies indicate that electromagnetic processes such as
92: Compton scattering and photon splitting $\gamma \to \gamma \gamma$ (merging $\gamma \gamma \to \gamma$)
93: could play a crucial role in these models. For example, the process of photon splitting in strong magnetic field could
94: suppress the production of electron-positron pair required for a
95: detectable radio emission from magnetar~\cite{BH:1998, BH:2000}. In turn, Compton
96: scattering may play an important role in the models of atmosphere emission
97: from strongly magnetized neutron stars~\cite{Miller:1995, Bulik:1997}.
98: The both processes are very important in the models the radiation emission from
99: SGR burst~\cite{Duncan:1995, HBG:1996, HBG:1997}.
100:
101: Previously, Compton scattering was studied under various conditions. The
102: case of the magnetic field without plasma was investigated
103: in~\cite{Herold:1979, Melrose:1983, Harding:1986, Harding:2000}. It was found that Compton scattering
104: becomes strongly anisotropic and essentially depends on photon
105: polarization. In~\cite{Bulik:1997} it was shown that the
106: dispersion properties of photon in strongly magnetized cold plasma could
107: significantly influence on the process under consideration.
108:
109: In the present work the process of photon scattering, $\gamma e^{\pm} \to \gamma
110: e^{\pm}$, is investigated in the presence of strong magnetic field and
111: charge-symmetric electron positron plasma,
112: when the magnetic field strength $B$ is the maximal physical
113: parameter, namely $\sqrt{eB} \gg T,\, \omega, \, E$. Here $T$ is
114: the plasma temperature, $\omega$ and $E$ is the initial photon and electron
115: energies. In this case almost all electrons and positrons in plasma are on the
116: ground Landau level.
117: %Moreover, we consider the region the electron and photon momenta far
118: We also consider the none resonance case, when electron and photon momenta
119: satisfy the condition $eB \gg (pk)$.
120: Such conditions could be realized e.g. in the Thompson \& Duncan model of SGR
121: burst~\cite{Duncan:1995} in which magnetically trapped high temperature ($\sim 1$
122: MeV) plasma fireball could be created.
123:
124:
125: The main purposes of our work are:
126: \begin{itemize}
127: \item
128: to investigate the influence of the strongly magnetized hot plasma on the
129: process of Compton scattering taking account of photon dispersion and wave
130: function renormalization;
131: \item
132: to compare the Compton scattering with the process of photon splitting;
133: \item
134: to estimate the possible effect of the process under consideration on
135: spectral formation of SGR.
136: \end{itemize}
137:
138: The plan of the paper is following. In the section 2 we consider the
139: propagation of the electromagnetic radiation in magnetized medium and give
140: the expression of Compton scattering amplitudes for different polarization
141: configurations of the initial and final photons. The analytic and numerical calculations of the
142: photon absorption rate and cross section of the process are
143: presented in Section 3. The discussion of the possible astrophysical
144: application of Compton scattering are given in Section 4.
145:
146: \section{Photon propagation in strongly magnetized hot plasma}
147:
148: \label{Sec2}
149:
150: The propagation of the electromagnetic radiation in any active medium is
151: convenient to describe in terms of normal modes (eigenmodes). In turn, the
152: polarization and dispersion properties of normal modes are connected with eigenvectors and
153: eigenvalues of polarization operator correspondingly. In the case of
154: strongly magnetized charge-symmetric plasma the eigenvalues of the
155: polarization operator can be obtained from the results of ~\cite{Shabad}:
156:
157: %
158: \begin{eqnarray}
159: \P^{(1)}(q) &\simeq& - \frac{\alpha}{3 \pi}\,
160: q_{\mbox{\tiny $\bot$}}^2 -
161: q^2\, \Lambda(B), \label{P1}
162: \\[2mm]
163: {\cal P}^{(2)}(q) &\simeq& -\frac{2 eB \alpha}{\pi}\left[
164: H\left(\frac{q^2_{\mprl}}{4m^2} \right) + {\cal J}(\prl{q}) \right]
165: - q^2 \, \Lambda(B), \label{P2}
166: \\[2mm]
167: \P^{(3)}(q) &\simeq& - \, q^2 \, \Lambda(B),
168: \label{P3}
169: \end{eqnarray}
170:
171: where
172: %
173: $$
174: \Lambda(B) = \frac{\alpha}{3 \pi}\,\left[1.792 - \ln
175: (B/B_e)\right].
176: $$
177: %
178: \beq
179: \nonumber
180: {\cal J}(\prl{q}) = 4 \prl{q}^2 m^2 \int \frac{dp_z}{E} \,
181: \frac{f_E}
182: {(\prl{q}^2)^2 - 4 \prl{(pq)}^2}\,, \qquad E = \sqrt{p_z^2 + m^2},
183: \eeq
184: %
185: $f_E \, = \, [\exp{(E/T)} \, + \, 1]^{-1}$
186: is the electron distribution function,
187: %
188: \beq
189: \nonumber
190: H(z)=\frac{1}{\sqrt{z(1 - z)}} \arctan \sqrt{\frac{z}{1 - z}} - 1,
191: \ z \le 1.
192: \eeq
193:
194:
195: The four-vectors with indices $\bot$ and $\parallel$
196: belong to the Euclidean \{1, 2\}-subspace and the Minkowski
197: \{0, 3\}-subspace correspondingly in the frame were the magnetic
198: field is directed along third axis; $\varphi_{\alpha \beta} = F_{\alpha
199: \beta} /B$ and
200: ${\tilde \varphi}_{\alpha \beta} = \frac{1}{2} \varepsilon_{\alpha \beta
201: \mu \nu} \varphi_{\mu \nu}$ are the dimensionless field tensor and dual
202: field tensor correspondingly.
203: The tensors
204: $\Lambda_{\alpha \beta} = (\varphi \varphi)_{\alpha \beta}$,\,
205: $\widetilde \Lambda_{\alpha \beta} =
206: (\tilde \varphi \tilde \varphi)_{\alpha \beta}$, with equation
207: $\widetilde \Lambda_{\alpha \beta} - \Lambda_{\alpha \beta} =
208: g_{\alpha \beta} = diag(1, -1, -1, -1)$ are introduced.
209:
210:
211: The dispersion properties of the normal modes are defined from the
212: dispersion equations
213: %
214: \beq
215: q^2 - \P^{(\lm)}(q) = 0 \qquad (\lm = 1, 2, 3).
216: \label{disper}
217: \eeq
218: %
219: Their analysis shows that the modes
220: with $\lambda = 1, 2$ and with polarization vectors
221: %
222: \beq
223: \ee_\alpha^{(1)}(q) = \frac{(q \varphi)_\alpha}{\sqrt{q_{\mprp}^2}},
224: \qquad
225: \ee_\alpha^{(2)}(q) = \frac{(q \tilde \varphi)_\alpha}
226: {\sqrt{q_{\mprl}^2}}.
227: \label{epsilon}
228: \eeq
229: %
230: are only physical in the case under consideration, just as in the
231: pure magnetic field~\footnote{ Symbols 1 and 2 correspond to
232: the $\|$ and $\perp$ polarizations in pure magnetic
233: field~\cite{Adler:1971} and $E$- and $O$- modes in magnetized
234: plasma~\cite{Duncan:1995}.}. However, it should be emphasized that the
235: coincidence of the polarization vectors in the plasma is
236: approximate, to within $O(1/\beta)$.
237:
238: Notice, that in plasma only the eigenvalue $\P^{(2)}(q)$ modifies in
239: comparison with pure magnetic field case. It means that the dispersion
240: law of the mode 1 is the same one as in the magnetized vacuum, where its
241: deviation from the vacuum law, $q^2 = 0$, is negligibly small.
242: From the other hand, the dispersion properties of the mode 2 essentially differs from the
243: magnetized vacuum ones. In the Figure~\ref{fig:dis1} the photon dispersion in both
244: strong magnetic field and magnetized plasma at various temperatures are depicted.
245: %
246: \begin{figure*}
247: \centerline{\includegraphics{fig31.eps}}
248: \vspace*{-2mm}
249: \caption{Photon dispersion laws in strong magnetic field $B/B_e = 200$
250: and neutral plasma vs. temperature: $T = 1$ MeV
251: (is upper curve), $T = 0.5$ MeV (is middle curve), $T = 0.25$ MeV
252: (is lower curve).
253: Photon dispersion without plasma is depicted by dashed line.
254: Dotted line corresponds to the vacuum dispersion law, $q^2 = 0$. The angle
255: between the photon momentum and the magnetic
256: field direction is $\pi/2$.
257: }
258: \label{fig:dis1}
259: \end{figure*}
260: %
261: One can see that in the presence of the magnetized plasma there exist the kinematical region, where $q^2 >0$ contrary
262: to the case of pure magnetic field. This fact could lead to the alteration of the different photon processes kinematic.
263: For example, the photon splitting channel $\gamma _2 \to \gamma _1 \gamma_1$ forbidden in the magnetic field without plasma becomes allowed~\cite{RCh05}.
264: Another important distinction is an essentially
265: different dependence of the dispersion law in variables $q_{\mprl}^2$
266: and $q_{\mprp}^2$ on the angle between the photon momentum
267: and the magnetic field direction (see Fig.~\ref{fig:dis2}).
268:
269: \begin{figure*}
270: \centerline{\includegraphics{fig3.eps}}
271: \vspace*{-2mm}
272: \caption{
273: Photon dispersion in a strong magnetic field ($B/B_e = 200$) and a charge-symmetric plasma
274: ($T = 1$ MeV) at
275: various angles between the photon momentum and the magnetic
276: field direction: $\theta = \pi/2$ (upper solid curve), $\pi/6$ (middle
277: solid curve), and $\pi/12$ (lower solid curve). The dashed curve
278: represents the dispersion in the absence of a plasma.
279: }
280: \label{fig:dis2}
281: \end{figure*}
282: %
283:
284: It follows from Eq.~(\ref{P2}) that the eigenvalue of the
285: polarization operator ${\cal P}^{(2)}$ becomes large near the electron-
286: positron pair production threshold, $q_{\mprl}^2=4 m^2$. This suggests that
287: the renormalization of the wave function for a photon
288: of this polarization should be taken into account:
289: %
290: \beq
291: \ee_{\alpha}^{(2)}(q) \to \ee_{\alpha}^{(2)}(q) \sqrt{Z_2}, \quad
292: Z^{-1}_2 = 1 - \frac{\partial {\cal P}^{(2)}(q)}{\partial \omega^2}.
293: \label{eps}
294: \eeq
295:
296:
297: To calculate the amplitude of process in strong magnetic field one
298: should use the Dirac equation solutions at the ground Landau level.
299: For the electron propagator it is relevant to use its asymptotic
300: form~\cite{KM_Book}. Then substituting the polarization
301: vectors~(\ref{epsilon}) one can obtain partial amplitudes for
302: different polarization configuration of the initial and final
303: photons. It is possible to present them in the covariant form:
304: %
305: \begin{eqnarray}
306: {\cal M}_{1 \to 1} = -\frac{8 \pi \alpha m}{eB}\,
307: \frac{(q \varphi q')(q \tilde \varphi q')}
308: {\sqrt{q^2_{\mprp} q'^2_{\mprp} (-Q^2_{\mprl})}},
309: \label{eq:M11}
310: \end{eqnarray}
311: %
312: \begin{eqnarray}
313: {\cal M}_{1 \to 2} = -\frac{8 \pi \alpha m}{eB}\,
314: \frac{(q \Lambda q')(q' \tilde \Lambda Q)}
315: {\sqrt{q^2_{\mprp} q'^2_{\mprl} (-Q^2_{\mprl})}},
316: \label{eq:M12}
317: \end{eqnarray}
318: %
319: \begin{eqnarray}
320: {\cal M}_{2 \to 1} = \frac{8 \pi \alpha m}{eB}\,
321: \frac{(q \Lambda q')(q \tilde \Lambda Q)}
322: {\sqrt{q^2_{\mprl} q'^2_{\mprp} (-Q^2_{\mprl})}},
323: \label{eq:M21}
324: \end{eqnarray}
325: %
326: \begin{eqnarray}
327: {\cal M}_{2 \to 2} = 16 i \pi \alpha m\, \frac{\sqrt{q^2_{\mprl}
328: q'^2_{\mprl}}\,\sqrt{(-Q^2_{\mprl})}\, \varkappa} {(q \tilde \Lambda
329: q')^2 - \varkappa^2 (q \tilde \varphi q')^2}, \label{eq:M22}
330: \end{eqnarray}
331: %
332:
333: \noindent where $\varkappa = \sqrt{1 - 4m^2/Q^2_{\mprl}}$ and
334: $Q^2_{\mprl} = (q - q')^2_{\mprl} <0$. We can see from last equations that all
335: amplitudes except ${\cal M}_{2 \to 2}$ are suppressed by magnetic field
336: strength. Therefore one could expect that mode 2 has the largest scattering
337: absorption rate.
338:
339:
340:
341: \section{Photon absorption rate and cross section in the strongly magnetized medium}
342: \label{Sec3}
343:
344: To analyse the efficiency of the process under consideration and to compare it
345: with other competitive reactions we calculate the photon absorption rate which
346: can be defined in the following way:
347: %
348: \begin{eqnarray}
349: &&W_{\lambda e^{\pm} \to \lambda^{'} e^{\pm}} = \frac{eB}{16 (2\pi)^4 \omega_{\lambda}}
350: \int \mid {\cal M_{\lambda \to \lambda^{'} }}\mid^2
351: Z_{\lambda}Z_{\lambda^{'}} \times
352: \label{eq:Wscatt}\\
353: &&\times f_{E}\, (1-f_{E'}) \, (1 + f_{\omega'})
354: \delta (\omega_{\lambda}({\bf k}) + E - \omega_{\lambda^{'}}({\bf k'}) - E')
355: \frac{dp_z\,d^3 k^{'}}{ E E' \omega_{\lambda^{'}}},
356: \nonumber
357: \end{eqnarray}
358: %
359: where $f_{\omega'} \, = \, [\exp{(\omega'/T)} \, - \, 1]^{-1}$ is the
360: photon distribution function. In the case of the rarefied plasma ($T \ll m$)
361: these absorption rates can be expressed in term of partial cross sections~\footnote{Hereafter the initial photon propagation across magnetic field is considered}
362: $W_{\lambda \to \lambda^{'}} = W_{\lambda e^{-} \to \lambda^{'} e^{-}} +
363: W_{\lambda e^{+} \to \lambda^{'} e^{+}} =
364: n_e \sigma_{\lambda \to \lambda^{'}}$:
365: %
366: \begin{eqnarray}
367: \sigma_{1 \to 1} = \frac{3}{8}\, \sigma_T \left ( \frac{B_e}{B} \right )^2 \frac{\omega^2}{m^2} \,
368: \left [\frac{\omega + 2m}{2(\omega + m)} + \frac{m}{\omega}
369: \ln \left (1 + \frac{\omega}{m} \right)\right ],
370: \label{eq:sigma11}
371: \end{eqnarray}
372: %
373: %
374: \begin{eqnarray}
375: \sigma_{2 \to 1} = \frac{3}{8}\, \sigma_T \left ( \frac{B_e}{B}\right )^2 \frac{q^2_{\mprp}}{m^2} \, Z_2 \,
376: \left [\frac{\omega + 2m}{2(\omega + m)} - \frac{m}{\omega}
377: \ln \left (1 + \frac{\omega}{m} \right)\right ],
378: \label{eq:sigma21}
379: \end{eqnarray}
380: %
381: %
382: \begin{eqnarray}
383: \sigma_{1 \to 2} = \frac{3}{4}\, \sigma_T \left ( \frac{B_e}{B} \right)^2
384: \frac{(\omega + 2m)^2}{\omega (\omega + m)} \!\!\!\! \int \limits_0^{\,\,\,\,\,\,\omega^2\!/4m^2} \!\!\!\!\!\!\! dz
385: \left(1 + \frac{3}{2} \xi \,\frac{H(z)}{z} \right)
386: \sqrt{\frac{\omega^2 - 4 m^2 z}{(\omega + 2m)^2 - 4 m^2 z}},
387: \label{eq:sigma12}
388: \end{eqnarray}
389: %
390: %
391: \begin{eqnarray}
392: \!\!\!\!\!\!\!\!\!\!\!\!\sigma_{2 \to 2}& = & 6\, \sigma_T \, \frac{m^4 (\omega + m)}
393: {\omega^3 (\omega + 2m)^2}\, Z_2 \,
394: \left [\frac{\omega(\omega + 2m)}{(\omega + m)(2m - \omega)} -
395: \ln \left (1 + \frac{\omega}{m} \right) + \right.
396: \nonumber
397: \\
398: &+&\left.\frac{2 \omega (\omega - m) (2m + \omega)}
399: {(\omega + m)(2m - \omega)\sqrt{4m^2 - \omega^2}}\,
400: \arctan{\frac{\omega}{\sqrt{4m^2 - \omega^2}}} \right ],
401: \label{eq:sigma22}
402: \end{eqnarray}
403: %
404: \noindent where $\sigma_T$ is the Thompson cross section, $\xi = \frac{\alpha}{3 \pi} \frac{B}{B_e}$ is the parameter characterizing magnetic field influence, $q^2_{\mprp}= \omega^2 - \P^{(2)}(\omega^2)$ and the number of electron (positron) density in a
405: strongly magnetized, charge-symmetric rarefied plasma can be estimated as
406: %
407: \begin{eqnarray}
408: n_e \simeq eB \sqrt{\frac{mT}{2\pi^3}}\,e^{-m/T}.
409: \label{eq:ne}
410: \end{eqnarray}
411: %
412: In the low energy limit ($\omega \ll m$) the formulas (\ref{eq:sigma11}-\ref{eq:sigma22}) can be presented in the following form:
413: %
414: \begin{eqnarray}
415: \sigma_{1 \to 1} &=& \frac{3}{4}\, \sigma_T \left ( \frac{B_e}{B} \right )^2 \frac{\omega^2}{m^2},
416: \quad
417: \sigma_{1 \to 2} = \frac{1}{4}\, \sigma_T \left ( \frac{B_e}{B} \right )^2 \frac{\omega^2}{m^2}\, (1 + \xi),
418: \label{eq:sigma1sT}
419: \\
420: \sigma_{2 \to 1} &=& \frac{1}{16}\, \sigma_T \left ( \frac{B_e}{B} \right )^2 \frac{\omega^2 (\omega^2-\omega_{pl}^2)}{m^4} \theta(\omega - \omega_{pl}) ,
421: \quad
422: \sigma_{2 \to 2} = \frac{\sigma_T}{1+\xi},
423: \label{eq:sigma2sT}
424: \end{eqnarray}
425: %
426: where $\omega_{pl}$ is plasma frequency defined by equation $\omega_{pl}^2 - \P^{(2)}(\omega_{pl}, \vec k \to 0 )$ = 0.
427: One can see that the presence of magnetized plasma slightly influences on the process cross-sections in this limit. Moreover, the corrections connected with photon dispersion and wave function renormalization are significant only for $\xi \sim 1$, i.e. when magnetic field is rather strong $B \sim 10^3 B_e$. In the case $\xi \ll 1$ which is relevant for the models of magnetar magnetosphere emission the formulas (\ref{eq:sigma1sT}, \ref{eq:sigma2sT}) coincide with the well-known result~\cite{Herold:1979}. It is interesting also to consider the case of $\omega \sim 2 m$. In this region the magnetized vacuum influence is large and the cross-sections (\ref{eq:sigma11}-\ref{eq:sigma22}) have the form:
428: %
429: \begin{eqnarray}
430: \sigma_{1 \to 1} &=& \sigma_T \left ( \frac{B_e}{B} \right )^2 \, \Big [1 + \frac{3}{4} \ln 3 \Big ],
431: \quad
432: \sigma_{1 \to 2} = 4\, \sigma_T \left ( \frac{B_e}{B} \right )^2 \,\Big [1 - \frac{3}{4} \ln 3 \Big ]\, (1 + 2 \xi),
433: \label{eq:sigma12m}
434: \\
435: \sigma_{2 \to 1} &=& 2 \, \sigma_T \left ( \frac{B_e}{B} \right )^2 \, \Big [1 - \frac{3}{4} \ln 3 \Big ]\, \left (1 - \frac{\omega^2}{4 m^2} \right ),
436: \quad
437: \sigma_{2 \to 2} = \frac{\sigma_T}{2 \xi}.
438: \label{eq:sigma22m}
439: \end{eqnarray}
440: %
441: To investigate the Compton scattering under hot plasma conditions ($T \sim m$) we have made the numerical calculations of photon absorption rates for the various channels. The results are represented in Figures \ref{fig:W22} and \ref{fig:W11}.
442:
443: %
444: %
445: \begin{figure*}
446: \centerline{\includegraphics{c2n.eps}}
447: \vspace*{-2mm}
448: \caption{The dependence of the photon absorption rate for channels
449: $\gamma_2 e^{\pm} \to \gamma_2 e^{\pm}$ (solid line) and
450: $\gamma_2 e^{\pm} \to \gamma_1 e^{\pm}$ (dashed line) on energy of initial photon in
451: strong magnetic field $B/B_e \, = \, 200$ at $T \,=\,1 MeV$ -- {\it a}, $T \,=\,250 keV$ -- {\it b} and
452: $T \,=\,50 keV$ -- {\it c}. The dotted line corresponds to the probability of photon splitting,
453: $\gamma_2 \to \gamma_1 \gamma_1$, at $T \, = \, 1 MeV$~\cite{RCh05}.
454: Here $W_0 \, = \, (\alpha/\pi)^3\, m \simeq 3.25\cdot 10^2 cm^{-1}$.}
455: \label{fig:W22}
456: \end{figure*}
457: % %
458: \begin{figure*}
459: \centerline{\includegraphics{c1n.eps}}
460: \vspace*{-2mm}
461: \caption{The dependence of the photon absorption rate for channels
462: $\gamma_1 e^{\pm} \to \gamma_1 e^{\pm}$ (solid line) and
463: $\gamma_1 e^{\pm} \to \gamma_2 e^{\pm}$ (dashed line) on energy of initial photon in
464: strong magnetic field $B/B_e \, = \, 200$ at $T \,=\,1 MeV$ -- {\it a}, $T \,=\,250 keV$ -- {\it b} and
465: $T \,=\,50 keV$ -- {\it c}.
466: The dotted and chain lines correspond to the probability of photon splitting,
467: $\gamma_1 \to \gamma_1 \gamma_2$ and $\gamma_1 \to \gamma_2 \gamma_2$,
468: respectively, at $T \, = \, 50 keV $~\cite{RCh05}.
469: Here $W_0 \, = \, (\alpha/\pi)^3\, m \simeq 3.25\cdot 10^2 cm^{-1}$.}
470: \label{fig:W11}
471: \end{figure*}
472:
473: \section{Discussion}
474: \label{Sec4}
475:
476: In the models of soft gamma repeaters spectrum formation the dependence of photon absorption rates on energy and temperature plays an important role. It could influence on the shape of emergent spectrum and defines the temperature profile in the emission region during bursts in SGRs~\cite{Duncan:1995, Lyubarsky:2002}.
477: In the both figures \ref{fig:W22} and \ref{fig:W11} one can see that photon absorption rates corresponding to Compton scattering are the fast increasing functions of temperature. At the same time, the channels with initial photon of mode 1 and mode 2 have different character of the absorption coefficient energy dependence. As shown in the Fig.\ref{fig:W22} the absorption rates for the reactions
478: $\gamma_2 e^{\pm} \to \gamma_2 e^{\pm}$ and $\gamma_2 e^{\pm} \to \gamma_1 e^{\pm}$ have thresholds. They are caused by mode 2 dispersion relation with plasma frequency and indicated the fact that the electromagnetic wave corresponding to mode 2 can't propagate with energy below $\omega_{pl}$. In the region above plasma frequency $W_{\gamma_2 e^{\pm} \to \gamma_2 e^{\pm}}$ is slightly depends on energy while $W_{\gamma_2 e^{\pm} \to \gamma_1 e^{\pm}}$ tends to zero in the vicinity of $\omega_{pl}$ and pair creation threshold, $\omega = 2m$. We would like to note that in the vicinity of pair creation threshold taking account of wave function renormalization and photon dispersion becomes very important and defines the processes rates dependence on energy, temperature and magnetic field. In the region $\omega \ge 2 m$ the photon of mode 2 is unstable due to the process $\gamma _2 \to e^+ e^-$ with rate much lager than Compton scattering. The energy dependence of $W_{\gamma_1 e^{\pm} \to \gamma_1 e^{\pm}}$ and
479: $W_{\gamma_1 e^{\pm} \to \gamma_2 e^{\pm}}$ is depicted in Fig.\ref{fig:W11}. On can see the fast increasing of absorption coefficients at low energies and rather slow dependence at $\omega \gtrsim 2 m$.
480:
481: The previous investigation of radiation transfer problem in strongly magnetized plasma have shown that along with Compton scattering process the photon splitting $\gamma \to \gamma \gamma$ could play a significant role as mechanism of photon production~\cite{Duncan:1995}. To compare these processes we depict the probabilities of the photon splitting channels on the same plots (see dotted and chain lines in Fig.\ref{fig:W22} and Fig.\ref{fig:W11}). On can see that at temperature $T \sim m$ and in kinematical region $\omega \le 2 m $ the main photon splitting process is $\gamma_2 \to \gamma_1 \gamma_1$ forbidden in pure magnetic field~\cite{RCh05}. It is seen also that the rate of the process is much less than Compton scattering ones. Nevertheless, it could be an effective photon production mechanism at temperatures under consideration. The probabilities of the channels $\gamma_1 \to \gamma_1 \gamma_2$ and $\gamma_1 \to \gamma_2 \gamma_2$ increase with temperature falling and become comparable and even larger than Compton scattering rates $W_{\gamma_1 e^{\pm} \to \gamma_1 e^{\pm}}$ and $W_{\gamma_1 e^{\pm} \to \gamma_2 e^{\pm}}$. As shown in Fig.~\ref{fig:W11} the process of photon splitting strongly dominates over Compton scattering at $T=50 keV$.
482: It was claimed previously that the effect of the cold strongly magnetized plasma on photon splitting is not pronounced and the vacuum approximation can be used in the most calculations~\cite{Bulik:1997,Elmfors:1998}. We can see now that in the presence of hot plasma the process of photon splitting could not be only intensive source of photon production but also an effective absorption mechanism.
483:
484: Let us illustrate this fact in the framework of the magnetar model of SGR burst. It is known that the radiation transfer in the magnetically trapped plasma may be described as s diffusion of the 1-mode photons whereas 2-mode photons are locked~\cite{Duncan:1995,Lyubarsky:2002}. The last circumstance is connected especially with weak dependence of the 2-mode absorption coefficient on photon energy (see Fig.~\ref{fig:W22}). In the assumption that the properties of the radiation field do not change much over absorption length $W^{-1}$ the radiation transfer may be conveniently described in terms of the Rosseland mean free path of 1-mode photons~\cite{Duncan:1995,Lyubarsky:2002}
485: %
486: \begin{equation}
487: \lambda_R = \frac{\int W^{-1} \omega^4 \exp(\omega/T) [\exp(\omega/T)-1]^{-2} d \omega}{\int \omega^4 \exp(\omega/T) [\exp(\omega/T)-1]^{-2} d \omega},
488: \label{eq:lR}
489: \end{equation}
490: %
491: where $W$ is the total absorption coefficient. Then the temperature distribution in the magnetically trapped plasma region could be defined from the diffusion equation:
492: %
493: \begin{equation}
494: F(z)= \frac{16 \lambda_R \, \sigma}{3}\,T^3 \, \frac{\partial T}{\partial z},
495: \label{eq:Difuss}
496: \end{equation}
497: %
498: where $\sigma$ is the Stefan-Boltzmann constant, $F(z)$ is the radiation flux.
499: If only Compton scattering process is considered then
500: %
501: \begin{equation}
502: W = W_{\gamma_1 e^{\pm} \to \gamma_1 e^{\pm}} + W_{\gamma_1 e^{\pm} \to \gamma_2 e^{\pm}}.
503: \label{eq:WCompt}
504: \end{equation}
505: %
506: To take into account the effect of photon splitting one should add term
507: $W_{\gamma_1 \to \gamma_1 \gamma_2} + W_{\gamma_1 \to \gamma_2 \gamma_2}$ to the equation (\ref{eq:WCompt}). In the figure~\ref{fig:RC_CSp} the ratio of the Rosseland mean free path for only Compton scattering to Rosseland mean free path with photon splitting is depicted. One can see that at low temperatures the additional absorption process of photon splitting leads to the significant decreasing of the Rosseland mean free path. The solution of the diffusion equation (\ref{eq:Difuss}) is out of the scope of this article. However, we would like to note that the more detail analysis of the radiation transfer needs the consistent solution of Boltzmann equation for the photon occupation number and radiative transfer equation in the wide range of temperatures
508: ($ 10 keV \lesssim T \lesssim 1 MeV$).
509: %
510: \begin{figure*}
511: \centerline{\includegraphics{rc_csp.eps}}
512: \vspace*{-2mm}
513: \caption{The ratio of the Rosseland mean free paths without and with
514: taking into account of photon splitting process as a function of the
515: inverse temperature at $B = 200 B_e$}
516: \label{fig:RC_CSp}
517: \end{figure*}
518: %
519:
520: In conclusion, we have investigated the influence of the strongly
521: magnetized hot plasma on the Compton scattering process with taking into
522: account of photon dispersion and large radiative corrections. It was found
523: that absorption rate of the dominant scattering channel of the 2-mode
524: photon weakly depends on energy in wide temperature range whereas the
525: 1-mode absorption coefficient is fast increasing function of photon
526: energy. The comparison of the photon splitting effect and Compton
527: scattering shows that the influences of these reaction on the 1-mode
528: radiation transfer are competitive in rarefied plasma $(T \ll m)$.
529:
530:
531: \medskip
532:
533: \noindent{\bf Acknowledgements}
534:
535: We express our deep gratitude to the organizers of the
536: Seminar ``Quarks-2006'' for warm hospitality.
537: This work supported in part by the Council on Grants by the President
538: of Russian Federation for the Support of Young Russian Scientists and
539: Leading Scientific Schools of Russian Federation under the Grant
540: No. NSh-6376.2006.2 and No. MK-1097.2005.2 (MC), and by the Russian Foundation for Basic Research
541: under the Grant No. 04-02-16253. Work of MC is supported
542: by the DAAD (Deutscher Akademischer Austauchdienst).
543:
544: %%%%%%%%%%%%%%%%%%%
545:
546:
547: \begin{thebibliography}{99}
548: %
549: %%%%%%%%%Observation%%%%%%%%%%%%%%
550: %\cite{Kouveliotou:1998ze}
551: \bibitem{Kouveliotou:1998ze}
552: C.~Kouveliotou {\it et al.},
553: % ``An X-ray pulsar with a superstrong magnetic field in the soft gamma-ray
554: %repeater SGR 1806-20,''
555: Nature {\bf 393} (1998) 235.
556: %%CITATION = NATUA,393,235;%%
557:
558: %\cite{Kouveliotou:1998fd}
559: \bibitem{Kouveliotou:1998fd}
560: C.~Kouveliotou {\it et al.},
561: % ``Discovery of a magnetar associated with the Soft Gamma Repeater SGR
562: %1900+14,''
563: Astrophys.\ J.\ {\bf 510} (1999) L115
564: % [arXiv:astro-ph/9809140].
565: %%CITATION = ASTRO-PH 9809140;%%
566:
567: %\cite{Gavriil:2002mc}
568: \bibitem{Gavriil:2002mc}
569: F.~P.~Gavriil, V.~M.~Kaspi, P.~M.~Woods,
570: %``Magnetar-like X-ray Bursts from an Anomalous X-ray Pulsar,''
571: Nature {\bf 419} (2002) 142.
572: %%CITATION = ASTRO-PH 0209202;%%
573:
574: %\cite{Ibrahim:2002zw}
575: \bibitem{Ibrahim:2002zw}
576: A.~I.~Ibrahim, S.~Safi-Harb, J.~H.~Swank, W.~Parke and S.~Zane,
577: % ``Discovery of Cyclotron Resonance Features in the Soft Gamma Repeater SGR
578: %1806-20,''
579: Astrophys.\ J.\ {\bf 574} (2002) L51
580: % [arXiv:astro-ph/0210513].
581: %%CITATION = ASTRO-PH 0210513;%%
582: % \bibitem{Israel2005}
583: % Israel~G.L., Belloni~T., Stella~L. et al.
584: % // Preprint astro-ph/0505255.
585: %
586: %%%%%%%%%%%%Model%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
587: \bibitem{Duncan:1992}
588: R.C.~Duncan, C.~Thompson,
589: Astrophys.~J. {\bf 392} No.~1. (1992) L9
590: %
591: \bibitem{Duncan:1995}
592: C.~Thompson, R.C.~Duncan,
593: Mon.~Not.~R.~Astron.~Soc. {\bf 275} (1995) 255
594: %
595: \bibitem{Duncan:1996}
596: R.C.~Duncan, C.~Thompson,
597: Astrophys.~J. {\bf 473} (1996) 322
598: %%%%%%%%%%%%%%Matter in strong magnetic field %%%%%%%%%%%%%
599: \bibitem{Duncan:2000pj}
600: R.~C.~Duncan,
601: Fifth Huntsville Gamma-Ray Burst Symposium, Eds: R. Marc Kippen, Robert
602: S. Mallozzi, Gerald J. Fishman. AIP Vol. 526 (American Institute of Physics, Melville, New York)
603: p.830 [astro-ph/0002442].
604: %
605: \bibitem{Lai:2001}
606: D.~Lai,
607: Rev. Mod. Phys. {\bf 73} (2001) 629
608: %
609: \bibitem{KM_Book}
610: A.~V.~Kuznetsov, N.~V.~Mikheev,
611: {\it Electroweak Processes in External Electromagnetic Fields},
612: Springer-Verlag, New York (2003), p.1.
613: %
614: \bibitem{Harding:2006qn}
615: A.~K.~Harding, D.~Lai,
616: %``Physics of Strongly Magnetized Neutron Stars,''
617: Rept.\ Prog.\ Phys.\ {\bf 69} (2006) 2631
618: %
619: \bibitem{BH:1998}
620: A.K.~Harding, M.G.~Baring,
621: Astrophys.~J. {\bf 507} (1998) L55
622: %
623: \bibitem{BH:2000}
624: M.~G.~Baring and A.~K.~Harding,
625: Astrophys.~J. {\bf 547} (2001) 929
626: %
627: \bibitem{Miller:1995}
628: M.C.~Miller,
629: Astrophys.~J.~Lett. {\bf 448} (1995) L29
630: %
631: \bibitem{Bulik:1997}
632: T.~Bulik, M.C.~Miller,
633: Mon.~Not.~R.~Astron.~Soc. {\bf 288} (1997) 596
634: %
635: \bibitem{HBG:1996}
636: A.~K.~Harding, M.~G.~Baring, P.~L.~Gonthier,
637: Astro. Astrophys. Suppl. Ser. {\bf 120} (1996) 111
638: %
639: \bibitem{HBG:1997}
640: A.~K.~Harding, M.~G.~Baring, P.~L.~Gonthier,
641: Astrophys.~J. {\bf 476} (1997) 246
642: %
643: \bibitem{Herold:1979}
644: H.~Herold,
645: Phys. Rev. {\bf D19} (1979) 2868
646: %
647: \bibitem{Melrose:1983}
648: D.B. Melrose, A.J. Parle,
649: Aust.~J. Phys. {\bf 36} (1983) 799
650: %
651: \bibitem{Harding:1986}
652: J.K.~Daugherty, A.K.~Harding,
653: Astrophys.~J. {\bf 309} (1986) 362
654: %
655: \bibitem{Harding:2000}
656: P.L. Gonthier, A.K. Harding, M.G. Baring et al.,
657: Astrophys.~J. {\bf 540} (2000) 907
658: %
659: \bibitem{Shabad}
660: A.E. Shabad,
661: Tr. Fiz. Inst. Akad. Nauk SSSR {\bf 192} (1988) 5
662: %
663: \bibitem{Adler:1971}
664: S.~L.~Adler,
665: Ann. Phys. (N.Y.) {\bf 67} (1971) 599
666: %
667: \bibitem{RCh05}
668: D.~A.~Rumyantsev, M.~V.~Chistyakov, JETP {\bf 101} (2005) 635
669: %
670: \bibitem{Lyubarsky:2002}
671: Y.~E.~Lyubarsky,
672: Mon. Not. R. Astron. Soc. {\bf 332} (2002) 199
673: %
674: \bibitem{Elmfors:1998}
675: P~Elmfors., B~Skagerstam.,
676: Phys.~Lett. {\bf B427} (1998) 197
677: \end{thebibliography}
678:
679: %
680: \end{document}
681: