1: \documentclass{aastex}
2: \usepackage{emulateapj5,graphicx}
3: %\documentclass[12pt,preprint]{aastex}
4: %\usepackage{graphicx}
5: \begin{document}
6:
7: \newcommand{\bm}[1]{\mbox{\boldmath $#1$}}
8: %%%%%%%%%%%%%%%%%
9: % (1)TITLE PAGE %
10: %%%%%%%%%%%%%%%%%
11: \title{
12: Secondary Star Formation in a Population III Object
13: }
14: \author{Hajime Susa\altaffilmark{1}
15: \vskip 0.2cm
16: \affil{Department of Physics, Rikkyo University, Nishi-Ikebukuro,
17: Toshimaku, Japan}
18: \vskip 0.3cm
19: Masayuki Umemura\altaffilmark{2}
20: \vskip 0.2cm
21: \affil{Center for Computational Sciences, University of
22: Tsukuba, Japan}
23: \altaffiltext{1}{susa@rikkyo.ac.jp}
24: \altaffiltext{2}{umemura@rccp.tsukuba.ac.jp}
25: }
26:
27: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
28: % (2)Abstract & Subject Headings %
29: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
30: \begin{abstract}
31: We explore the possibility of subsequent star formation
32: after a first star forms in a Pop III object,
33: by focusing on the radiation hydrodynamic (RHD) feedback
34: brought by ionizing photons as well as H$_2$ dissociating photons.
35: For the purpose, we perform three-dimensional RHD
36: simulations, where the radiative transfer of ionizing photons
37: and H$_2$ dissociating photons from a first star
38: is self-consistently coupled with hydrodynamics based on
39: a smoothed particle hydrodynamics method.
40: As a result, it is shown that density peaks above a threshold density
41: can keep collapsing owing to the shielding of H$_2$ dissociating radiation
42: by an H$_2$ shell formed ahead of a D-type ionization front.
43: But, below the threshold density, an M-type ionization front
44: accompanied by a shock propagates, and density peaks are radiation
45: hydrodynamically evaporated by the shock.
46: The threshold density is dependent on the distance from a source star,
47: which is $\approx 10^2 {\rm cm^{-3}}$ for the source distance of 30pc.
48: Taking into consideration that the extent of a Pop III object is $\approx 100$pc
49: and density peaks within it have the density of $10^{2-4}$cm$^{-3}$,
50: it is concluded that the secondary star formation is allowed
51: in the broad regions in a Pop III object.
52:
53: \end{abstract}
54: \keywords{theory:early universe --- galaxies: formation --- radiative transfer
55: --- molecular processes --- hydrodynamics}
56:
57: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
58: %(3)TEXT & Acknowledgements %
59: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
60:
61: \section{Introduction}
62: \label{intro}
63: In the last decade, the formation of Population III (hereafter Pop III) stars
64: has been explored extensively
65: (Bromm, Coppi \& Larson 1999, 2002;
66: Nakamura \& Umemura 1999, 2001; Abel, Bryan, Norman 2000, 2002;
67: Yoshida 2006).
68: The Pop III stars can significantly
69: affect the reionization history in the universe
70: \citep{Cen03,Cia03,Wyi04,Somer03,Sokasian04,Murakami05},
71: and the metal enrichment of intergalactic medium
72: \citep{NU01,Scan02,Ricotti04}.
73: In previous analyses
74: \citep{Tegmark97,Fuller00,Yoshida03},
75: it is shown that Pop III objects
76: have the halo mass of $\approx 10^6 M_\odot$ and
77: the extent of $\approx 100$pc. In Pop III objects, density peaks
78: collapse owing to H$_2$ cooling, forming cloud cores with
79: the density of $10^{2-4}$cm$^{-3}$ \citep{Bromm02}.
80: A highest peak in the halo collapses earlier to form a first Pop III star.
81: Hence, subsequently collapsing cores are affected by
82: the radiative feedback by the first star.
83: However, the range of the feedback and the possibility of
84: secondary star formation in a Pop III object are still under debate.
85:
86: If a first star is distant by more than 1pc,
87: dense cores are readily self-shielded from the ultraviolet (UV) radiation
88: \citep{TU98, Kitayama01,SU04a,SU04b,Kitayama04,Dijkstra04,Alvarez06}.
89: Thus, the photoevaporation
90: by UV heating is unlikely to work devastatingly.
91: However, the photodissociating radiation (11.18-13.6eV)
92: of hydrogen molecules (H$_2$) in Lyman-Werner band (LW band)
93: can preclude the core from collapsing,
94: since H$_2$ are the dominant coolant to enable the core collapse
95: \citep{Haiman97,ON99,Haiman00,GB01,Machacek01}.
96: Hence, this may lead to momentous negative feedback.
97:
98: On the other hand, ionizing radiation
99: for hydrogen ($\geq$13.6eV) drives an ionization
100: front (I-front), which propagates in a collapsing core.
101: The enhanced fraction of electrons promotes H$_2$ formation
102: \citep{SK87, KS92, SUNY98, OhH02}. In particular,
103: the mild ionization ahead of the I-front can generate an H$_2$ shell,
104: which potentially shields H$_2$ dissociating photons \citep{Ricotti01}.
105: This mechanism is likely to work positively to form Pop III stars.
106: %Also, H$_2$ molecules form efficiently in relic HII regions
107: %if an ionizing source fades out
108: %\citep{Ricotti02, Oshea05, Nagakura05}.
109: In practice, the propagation of I-front is complex.
110: When UV irradiates a dense core, the I-front changes from R-type
111: on the surface to D-type inside the core. The transition occurs
112: via an intermediate type (M-type), which is accompanied with
113: the generation of shock \citep{Kahn54}. The shock can affect
114: significantly the collapse of the core. This is a totally radiation
115: hydrodynamic (RHD) process, which has not been hitherto explored
116: in detail as feedback by first stars.
117:
118: In this Letter, we scrutinize the radiation hydrodynamic (RHD) feedback
119: by a first star through the propagation of I-front in a dense core.
120: For the purpose, we solve the three-dimensional RHD,
121: where the radiative transfer of ionizing photons
122: as well as H$_2$ dissociating photons is self-consistently coupled
123: with hydrodynamics.
124: In \S 2, the key physical processes associated with I-front
125: propagation are overviewed. The setup of simulation is
126: described in \S 3, and numerical results are presented in \S 4.
127: \S 5 is devoted to the conclusions.
128: %The fully self-consistent radiation hydrodynamics (RHD) of
129: %Pop III star formation is still in incunabula.
130: %In this Letter, we present a 3D RHD simulation, where
131: %the radiation transfer is accurately solved
132: %with non-equilibrium chemistry regarding the photoionization
133: %and H$_2$ formation. The goal is to elucidate whether
134: %multiple Pop III stars can form in a single first object
135: %as a result of negative and positive radiative feedback processes.
136:
137:
138: %Moreover, according to the one dimensional radiation hydrodynamic
139: %simulations on the formation of HII region in interstellar media
140: %e.g. \citep{Hosokawa05}, ionization front could be D-type, and the
141: %gas swept by the HII region form a dense shell which can fragment
142: %gravitationally. In these cases, radiative feedback promotes the
143: %structure formation, thus it is a positive one.
144:
145:
146: \section{Basic Physics}
147:
148: If a collapsing core is irradiated by an ionizing source located
149: at the distance $D$, the propagation speed of I-front
150: in the core is given by
151: \begin{equation}
152: v_{\rm IF}=
153: 21~ {\rm km~s^{-1}}
154: \left(\frac{\dot{N}_{\rm ion}}{10^{50}{\rm s^{-1}}}\right)
155: \left(\frac{D}{20{\rm pc}}\right)^{-2}
156: \left(\frac{n_{\rm core}}{10^3{\rm cm^{-3}}}\right)^{-1},
157: \end{equation}
158: where $\dot{N}_{\rm ion}$ is the number of ionizing photons per
159: unit time and $n_{\rm core}$ is the number density in the cloud core.
160: The sound speed in the regions cooling by H$_2$
161: is $a_1 \approx 1{\rm km~s^{-1}}$,
162: while that in the ionized regions is $a_2 \approx 10{\rm km~s^{-1}}$.
163: Hence, $v_{\rm IF} > 2a_2$ for low density cores or
164: a nearby ionizing source, and therefore I-front
165: becomes R-type.
166: Supposing the cloud core size $r_{\rm core}$ is on the order of
167: $a_1\sqrt{\pi/G\rho_{\rm core}}$,
168: the propagation time of R-type front across the core satisfies
169: \begin{eqnarray}
170: t_{\rm IF}\equiv \frac{r_{\rm core}}{v_{\rm IF}} <
171: \frac{a_1}{2a_2}
172: \sqrt{\frac{\pi}{G\rho_{\rm core}}}
173: < \sqrt{\frac{\pi}{G\rho_{\rm core}}}.
174: \end{eqnarray}
175: This means that an R-type front sweeps the core before the core
176: collapses in a free-fall time. Thus, the core is likely to undergo
177: photoevaporation.
178: On the other hand, if the density of cloud core
179: is high enough and the source distance is large, then
180: $v_{\rm IF} < a_1^2/2a_2$ and a D-type I-front emerges.
181: The propagation time of D-type front across the core is
182: \begin{eqnarray}
183: t_{\rm IF} >
184: \frac{2a_2}{a_1}\sqrt{\frac{\pi}
185: {G\rho_{\rm core}}} > \sqrt{\frac{\pi}{G\rho_{\rm core}}}.
186: \end{eqnarray}
187: Thus, the core can collapse
188: before the I-front sweeps the core.
189:
190: The above arguments are based on the assumption that
191: the ionizing photon flux does not change during the propagation
192: of I-front.
193: However, the core could be self-shielded from the ionizing
194: radiation if
195: $\dot{N}(\pi r_{\rm core}^2/4\pi D^2)<4\pi r_{\rm core}^3
196: n_{\rm core}^2 \alpha_{\rm B}/3$, where $\alpha_{\rm B}$
197: is the recombination coefficient to all excited levels of hydrogen.
198: The critical density for self-shielding is given by
199: \begin{eqnarray}
200: &\;& n_{\rm shield} \simeq \left(\frac{3\dot{N}_{\rm ion}}
201: {16\pi D^2 a_1 \alpha_{\rm B}} \sqrt{\frac{G m_{\rm p}
202: }{\pi}}\right)^{2/3}\nonumber\\
203: &=& 5.1~ {\rm cm^{-3}}
204: \left(\frac{\dot{N}_{\rm ion}}
205: {10^{50}{\rm s^{-1}}}\right)^{2/3}\left(\frac{D}
206: {20{\rm pc}}\right)^{-4/3}
207: \left(\frac{a_1}{1{\rm km~ s^{-1}}}\right)^{-2/3}. \label{eq:shield}
208: \end{eqnarray}
209: If $n_{\rm core}> n_{\rm shield}$, the ionizing photon flux
210: diminishes significantly during the I-front propagation.
211: Hence, even if the I-front is R-type on the surface of
212: cloud core, the front can change to M-type accompanied with shock,
213: and eventually to D-type inside the core (Kahn 1954).
214:
215: In comparison to ionizing photons, H$_2$ dissociating (LW band)
216: radiation is less shielded \citep{DB96}.
217: The self-shielding of LW band flux ($F_{\rm LW}$) is expressed by
218: \begin{equation}
219: F_{\rm LW} = F_{\rm LW,0} f_{\rm sh}
220: \left( N_{\rm H_2,14 } \right) \label{LW}
221: \end{equation}
222: where $ F_{\rm LW,0}$ is the incident flux,
223: $ N_{\rm H_2,14}= N_{\rm H_2}/10^{14} {\rm cm^{-2}}$, and
224: \begin{equation}
225: f_{\rm sh}(x) = \left\{
226: \begin{array}{cc}
227: 1,~~~~~~~~~~~~~~x \le 1 &\\
228: x^{-3/4},~~~~~~~~~x > 1 &
229: \end{array}
230: \right.
231: \end{equation}
232: However, dissociating radiation could be shielded effectively
233: if an H$_2$ shell forms ahead of I-front.
234: This process is tightly coupled with the propagation of I-front.
235:
236: \section{Setup of Simulation}
237: \label{simulation}
238: We perform radiation hydrodynamic simulations
239: using a novel radiation transfer solver based
240: on smoothed particle hydrodynamics (SPH)
241: \citep{Susa06}.
242: We suppose a primordial gas cloud collapsing in a run-away fashion.
243: The chemical compositions are initially assumed to be
244: the cosmological residual value \citep{GP98}.
245: The mass of cloud is $M_{\rm b} = 8.3\times 10^4 M_\odot$
246: in baryonic mass.
247: When the central density of cloud exceeds a certain value,
248: $n_{\rm on}$, we ignite a 120$M_\odot$ Pop III star
249: ($\dot{N}_{\rm ion}=1.3\times 10^{50}{\rm s^{-1}}$),
250: which is located at $D=20$pc for fiducial models.
251: We also explore the dependence on the distance of a first star
252: by changing the relative location between a source star and
253: a collapsing cloud.
254: The luminosity and the effective temperature
255: of the source star are taken from \citet{Baraffe01}.
256: The fiducial models are \\
257: \hspace*{2mm}
258: Model A --- $n_{\rm on} = 3\times 10^3 {\rm cm^{-3}}$ with H$_2$ dissociating photons but without ionizing photons \\
259: \hspace*{2mm}
260: Model B --- $n_{\rm on} = 3\times 10^3 {\rm cm^{-3}}$
261: with ionizing and H$_2$ dissociating photons \\
262: \hspace*{2mm}
263: Model C --- same as Model B but for $n_{\rm on} = 3\times 10^2 {\rm cm^{-3}}$ \\
264: The physical simulation time after the ignition of the source star is
265: $4$Myr except $1.54$Myr for model B,
266: since the central part collapses below the resolution limit.
267: In Model A, it is assumed
268: that only LW band photons escape from the neighbor of a source star,
269: whereas ionizing photons do not because of large opacity.
270: This model also can be
271: regarded as the reference to Model B.
272:
273: \section{Results}
274: \subsection{Failed Collapse (Model A)}
275:
276: In Fig. \ref{timeevol}, the time evolution of
277: density profiles along the axis of symmetry is shown for three models.
278: In this figure, the red curve represents the distribution at 1Myr,
279: at which epoch the spatial distributions of physical quantities are shown
280: in Fig. \ref{3models}.
281: For Model A, as shown in the left panel of Fig. \ref{timeevol},
282: the collapse of central regions stops virtually
283: and forms a quasi-hydrostatic core between $\sim$1Myr and $\sim$4Myr.
284: The failure of collapse is caused by the photodissociation of H$_2$ by
285: LW band photons.
286: In upper panels of Fig. \ref{3models}, the spatial distribution of H$_2$
287: fraction on a slice along the symmetry axis is shown
288: by colored dots at the positions of SPH particles.
289: In lower panels of Fig. \ref{3models}, physical quantities along
290: the symmetry axis are shown at this epoch.
291: In Model A (left panel), the temperature is several $10^2$K in the envelope
292: and $\sim 10^3$K in the central regions.
293: The H$_2$ column density
294: is lower than $10^{14}{\rm cm^{-2}}$ in the envelope and
295: therefore H$_2$ is highly photodissociated by LW band radiation.
296: In the central regions, H$_2$ column density
297: exceeds $10^{14}{\rm cm^{-2}}$ and therefore
298: H$_2$ fraction is raised up by
299: the self-shielding of LW band photons according to (\ref{LW}).
300: However, this level of shielding is not enough to allow
301: the cloud to keep collapsing by H$_2$ cooling.
302: Eventually, the collapse is halted by the thermal pressure.
303: The numerical results of Model A show that gravitational
304: instability is hindered by permeated H$_2$ dissociating photons.
305:
306: \subsection{H$_2$ Shielded Collapse (Model B)}
307: In Model B, ionizing photons are included for the same model
308: parameter as Model A. In this model, I-front propagates into the
309: cloud. Although the I-front is R-type far from the cloud center,
310: it changes to M-type and a shock is generated as shown by
311: a peak at 0.5Myr in Fig. \ref{timeevol} (middle panel).
312: Around 1Myr, the I-front changes to D-type, and therefore
313: the core collapse proceeds faster than the propagation of the I-front,
314: as argued in \S 2.
315: The density distribution is highly changed by photoionization,
316: as shown by the angelfish-shaped distribution in Fig. \ref{3models}.
317: In particular, it is worth noting that an H$_2$ shell forms ahead of
318: the I-front due to the enhanced ionization fraction.
319: %as originally argued by \citet{Ricotti01} in different context.
320: As a result, the H$_2$ column density is steeply raised up.
321: This H$_2$ shell effectively shields LW band radiation from a source star,
322: and the resultant H$_2$ fraction ($y_{\rm H_2}$) becomes larger by an order
323: of magnitude than that of Model A.
324: Eventually, due to H$_2$ cooling,
325: the cloud core can continue to collapse to the level
326: of $n_{\rm H} > 10^7 {\rm cm^{-3}}$, as shown in Fig. \ref{timeevol} (middle).
327: Since the only difference between Models A and B is the presence of
328: ionizing radiation, we can conclude that photoionization can restrain the
329: negative feedback effect by H$_2$
330: photodissociation.
331:
332: \subsection{Shock-driven Evaporation (Model C)}
333: Model C is the lower density version of Model B, where $n_{\rm on}$
334: is ten times smaller than that of Model B.
335: In this case, quite similarly to Model B, the I-front becomes to M-type
336: at 0.5Myr (Fig. \ref{timeevol}, right panel).
337: As shown in Fig. \ref{3models},
338: there forms an H$_2$ shell that shields LW band radiation
339: from a source star. But, before the core collapses in the free-fall time,
340: the shock accompanied with M-type I-front sweeps up the central
341: regions at $\approx$ 1Myr.
342: Eventually, the shock blows out the collapsing core.
343: Hence, it is concluded that if an M-type I-front passes through
344: the cloud core, the whole cloud is evaporated
345: radiation hydrodynamically by the shock.
346:
347: \subsection{Dependence on Position of a Source Star }
348: As shown above, the criterion of radiation hydrodynamic
349: evaporation of a collapsing core is whether an M-type I-front
350: sweeps up the core.
351: By changing the position of a source star, we can obtain the
352: threshold density for the evaporation depending on the source distance.
353: As a result of numerical simulations, it is shown that
354: the threshold density is $\approx 10^3 {\rm cm^{-3}}$
355: for $D=20$pc, $\approx 10^2 {\rm cm^{-3}}$ for $D=30$pc,
356: and $\approx 10 {\rm cm^{-3}}$ for $D=50$pc.
357: Thus, cloud cores with $n_{\rm core} \gtrsim 10^2 {\rm cm^{-3}}$
358: can collapse at $D \gtrsim 30$pc.
359: Further details of the analysis on the parameter dependence will be
360: described in the forth-coming full paper.
361:
362: \section{Conclusions}
363: We have performed three-dimensional radiation hydrodynamic
364: simulations to scrutinize the feedback by a first star in a Pop III
365: object. As a result, it has been found that a collapsing core is
366: evaporated by a shock if an M-type I-front sweeps the core.
367: In order for a collapsing core to evade the radiation hydrodynamic
368: evaporation, the core density should exceed a threshold density
369: that depends on the distance from a source star.
370: Above the threshold density, the I-front is quickly changed to
371: the D-type, and an H$_2$ shell forms ahead of the I-front,
372: which effectively shields H$_2$ dissociating radiation
373: from a source star. Eventually, the core can keep collapsing
374: owing to H$_2$ cooling. The present numerical study has shown
375: that if a source star is distant by more than 30pc, then
376: a collapsing core denser than $\approx 10^2 {\rm cm^{-3}}$
377: is absolved from evaporation.
378: Taking into account that the density of cloud cores in
379: a Pop III object is expected to be $10^{2-4}$cm$^{-3}$,
380: stars can form in the regions of $\gtrsim 30$ pc
381: even after a first star forms.
382: Since the extent of a Pop III object is $\approx 100$pc,
383: it is concluded that
384: the subsequent star formation is allowed in the broad regions
385: in a Pop III object.
386:
387: \bigskip
388: We are grateful to N. Shibazaki for continuous encouragement,
389: to T. Nakamoto and K. Ohsuga for intense discussion,
390: and to all the collaborators in
391: {\it Cosmological Radiative Transfer Codes Comparison Project}
392: (astro-ph/0603199) for fruitful discussions during
393: the workshop in CITA and Lorentz Center in Leiden.
394: The analysis has been made with the {\it FIRST} simulator
395: at Center for Computational Sciences in University of Tsukuba and
396: with computational facilities in Rikkyo University.
397: This work was supported in part by Grants-in-Aid, Specially
398: Promoted Research 16002003 and Young Scientists (B) 17740110
399: from MEXT in Japan.
400:
401: %%%%%%%%%%%%%%%%%%
402: % (4) Appendices %
403: %%%%%%%%%%%%%%%%%%
404: %No Appendix
405: %%%%%%%%%%%%%%%%%
406: % (5)References %
407: %%%%%%%%%%%%%%%%%
408: \begin{thebibliography}{}
409: \bibitem[Abel, Bryan, Norman(2000)]{Abel00}
410: Abel, T., Bryan, G. L., \& Norman, M. L. 2000, \apj, 540, 39
411: \bibitem[Abel, Bryan, Norman(2002)]{Abel02}
412: Abel, T., Bryan, G. L., \& Norman, M. L. 2002, Science, 295, 93
413: \bibitem[Alvarez et al.(2006)]{Alvarez06}
414: Alvarez, M.~A., Bromm, V., \& Shapiro, P.~R.\ 2006, \apj, 639, 621
415: \bibitem[Baraffe, Heger \& Woosely (2001)]{Baraffe01}
416: Baraffe, I., Heger, A. \& Woosely, S.E. \ 2001, \apj, 550, 890
417: \bibitem[Bromm, Coppi \& Larson(1999)]{Bromm99}
418: Bromm, V., Coppi, P. S., \& Larson, R. B. 1999, \apj, 527, L5
419: \bibitem[Bromm, Coppi \& Larson(2002)]{Bromm02}
420: Bromm, V., Coppi, P. S., \& Larson, R. B. 2002, \apj, 564, 23
421: \bibitem[Cen(2003)]{Cen03} Cen, R. 2003, \apj, 591, L5
422: \bibitem[Ciardi Ferrara, \& White(2003)]{Cia03}
423: Ciardi, B., Ferrara, A., \& White, S. D. M. 2003, \mnras, 344, L7
424: \bibitem[Dijkstra et al.(2004)]{Dijkstra04}
425: Dijkstra, M., Haiman, Z., Rees, M.~J., \& Weinberg, D.~H.\ 2004, \apj, 601, 666
426: \bibitem[Draine \& Bertoldi (1996)]{DB96}
427: Draine, B. T., \& Bertoldi, F. 1996, \apj, 468, 269
428: \bibitem[Fuller \& Couchman(2000)]{Fuller00}
429: Fuller, T.~M., \& Couchman, H.~M.~P.\ 2000, \apj, 544, 6
430: \bibitem[Galli \& Palla (1998)]{GP98}
431: Galli D. \& Palla F. 1998, \aap, 335, 403
432: \bibitem[Glover \& Brand(2001)]{GB01}
433: Glover, S.~C.~O., \& Brand, P.~W.~J.~L.\ 2001, \mnras, 321, 385
434: \bibitem[Haiman et al.(1997)]{Haiman97}
435: Haiman, Z., Rees, M.~J., \& Loeb, A.\ 1997, \apj, 476, 458
436: \bibitem[Haiman et al.(2000)]{Haiman00}
437: Haiman, Z., Abel, T., \& Rees, M.~J.\ 2000, \apj, 534, 11
438: \bibitem[Kahn(1954)]{Kahn54}
439: Kahn, F.~D.\ 1954, \bain, 12, 187
440: \bibitem[Kang \& Shapiro (1992)]{KS92}
441: Kang, H., \& Shapiro, P., \apj, 386, 432
442: \bibitem[Kitayama et al.(2001)]{Kitayama01}
443: Kitayama, T., Susa, H., Umemura, M., \& Ikeuchi, S.\ 2001, \mnras, 326, 1353
444: \bibitem[Kitayama et al. (2004)]{Kitayama04}
445: Kitayama,T., Yoshida, N., Susa, H. \& Umemura, M., 2004, \apj, 613, 631
446: \bibitem[Machacek et al.(2001)]{Machacek01}
447: Machacek, M.~E., Bryan, G.~L., \& Abel, T.\ 2001, \apj, 548, 509
448: %\bibitem[Maselli, Ferrara \& Ciardi (2003)]{MFC03}
449: %Maselli, A., Ferrara, A. \& Ciardi, B. 2003, \mnras, 345, 379
450: %\bibitem[Mellema et al. (2005)]{Mellema05}
451: %Mellema, G., Iliev, I. T., Alvarez, M. A.\& Shapiro, P.R., 2005, astro-ph/0508416
452: %\bibitem[Monaghan (1992)]{Monaghan92}
453: %Monaghan, J.J. 1992, ARA\&A, 30, 543
454: \bibitem[Murakami et al.(2005)]{Murakami05}
455: Murakami, T., Yonetoku, D., Umemura, M., Matsubayashi, T.,
456: \& Yamazaki, R.\ 2005, \apjl, 625, L13
457: %\bibitem[Nagakura \& Omukai (2005)]{Nagakura05}
458: %Nagakura, T. \& Omukai, K. astro-ph/0505599 (2005)
459: %\bibitem[Nakamoto, Umemura, \& Susa (2001)]{NUS01}
460: %Nakamoto, T., Umemura, M., \& Susa, H.\ 2001, \mnras, 321, 593
461: \bibitem[Nakamura \& Umemura(1999)]{NU99}
462: Nakamura F. \& Umemura M. 1999, \apj, 515, 239
463: \bibitem[Nakamura \& Umemura(2001)]{NU01}
464: Nakamura, F., \& Umemura, M. 2001, \apj, 548, 19
465: \bibitem[Oh \& Haiman (2002)]{OhH02}
466: Oh, P \& Haiman, Z. \ 2002, \apj, 569, 558
467: \bibitem[Omukai \& Nishi (1999)]{ON99}
468: Omukai, K. \& Nishi, R. \ 1999, \apj, 518, 64
469: %\bibitem[O'Shea et al. (2005)]{Oshea05}
470: %O'Shea, B.~W., Abel, T., Whalen, D., Norman, M.~L. \apj, 628, 5L
471: %\bibitem[Razoumov et al. (2002)]{Raz02}
472: %Razoumov, A.~O., Norman, M.~L., Abel, T., \& Scott, D.\ 2002, \apj, 572, 695
473: \bibitem[Ricotti, Gnedin, \& Shull(2001)]{Ricotti01}
474: Ricotti, M. Gnedin, N.~Y., Shull, M. \ 2001, \apj, 560, 580
475: %\bibitem[Ricotti, Gnedin, \& Shull(2001)]{Ricotti02}
476: %Ricotti, M. Gnedin, N.~Y., Shull, M. \ 2002, \apj, 575, 49
477: \bibitem[Ricotti \& Ostriker(2004)]{Ricotti04}
478: Ricotti, M.~\& Ostriker, J.~P.\ 2004, \mnras, 350, 539
479: %\bibitem[Rijkhorst et al.(2005)]{saru}
480: %Rijkhorst, E., Plewa, T., Dubey, A. \& Mellema, G. \ 2005, astro-ph/0505213
481: \bibitem[Scannapieco, Ferrara, \& Madau(2002)]{Scan02}
482: Scannapieco, E., Ferrara, A., \& Madau, P.\ 2002, \apj, 574, 590
483: \bibitem[Shapiro \& Kang (1987)]{SK87}
484: Shapiro, P.R., \& Kang, H., 1987, \apj, 318, 32
485: \bibitem[Sokasian et al.(2004)]{Sokasian04}
486: Sokasian, A., Yoshida, N., Abel, T., Hernquist, L., \& Springel, V.\ 2004, \mnras, 350, 47
487: \bibitem[Somerville \& Livio(2003)]{Somer03}
488: Somerville, R.~S.~\& Livio, M.\ 2003, \apj, 593, 611
489: %\bibitem[Spitzer(1978)]{Spitzer78}
490: %Spitzer, L. Jr. 1978, in Physical Processes in the Interstellar Medium
491: %(John Wiley \& Sons, Inc. 1978)
492: %\bibitem[Susa \& Kitayama (2000)]{SuKi00}
493: %Susa, H. \& Kitayama, T. 2000, \mnras, 317, 175
494: \bibitem[Susa (2006)]{Susa06}
495: Susa, H. 2006, PASJ, in press (astro-ph/0601642)
496: \bibitem[Susa et al.(1998)]{SUNY98}
497: Susa, H., Uehara, H., Nishi, R., \& Yamada, M. 1998,
498: Prog. Theor. Phys., 100, 63
499: %\bibitem[Susa \& Umemura(2000)]{SU00}
500: %Susa, H. \& Umemura, M. 2000, \apj, 537, 578
501: \bibitem[Susa \& Umemura(2004a)]{SU04a}
502: Susa, H. \& Umemura, M. 2004a, \apj, 600, 1
503: \bibitem[Susa \& Umemura(2004b)]{SU04b}
504: Susa, H. \& Umemura, M. 2004b, \apj, 610, 5L
505: \bibitem[Tajiri \& Umemura(1998)]{TU98}
506: Tajiri, Y., \& Umemura, M.\ 1998, \apj, 502, 59
507: \bibitem[Tegmark et al.(1997)]{Tegmark97} Tegmark, M., Silk, J.,
508: Rees, M.~J., Blanchard, A., Abel, T., \& Palla, F.\ 1997, \apj, 474, 1
509: %\bibitem[Thacker et al. (2000)]{Thac00}
510: %Thacker, J., Tittley, R., Pearce, R., Couchman, P. \& Thomas, A. 2000, \mnras 319, 619
511: \bibitem[Wyithe \& Loeb(2004)]{Wyi04}
512: Wyithe, J.~S.~B.,~\& Loeb, A.\ 2004, \nat, 427, 815
513: \bibitem[Yoshida et al. (2003)]{Yoshida03}
514: Yoshida, N., Abel, T., Hernquist, L. \& Sugiyama, N., 2003, \apj, 592, 645
515: \bibitem[Yoshida(2006)]{Yoshida06}
516: Yoshida, N.\ 2006, New Astronomy Review, 50, 19
517: \end{thebibliography}
518:
519: %\end{document}
520:
521:
522: %%%%%%%%%%%%%%%%%%%%%%%
523: % (6) Figure Captions %
524: %%%%%%%%%%%%%%%%%%%%%%%
525:
526:
527: %\clearpage
528: \setcounter{figure}{0}
529: \begin{figure}[ht]
530: \begin{center}
531: \includegraphics[angle=0,width=14cm]{f1.eps}
532: \caption{
533: The time evolution of density distributions along the symmetry axis
534: for three models.
535: Four curves correspond
536: to $t=0$, 0.5Myr, 1Myr, and 4Myr, for models A and C,
537: whereas the final plot for model B corresponds to $t=1.5{\rm Myr}$.
538: The red curves denote the profiles at $1{\rm Myr}$, at which
539: epoch the detailed structure is shown in Fig.\ref{3models}. }
540: \label{timeevol}
541: \end{center}
542: \end{figure}
543:
544: \begin{figure}[ht]
545: \begin{center}
546: %\plotone{fig1.eps}
547: \includegraphics[angle=0,width=14cm]{f2.eps}
548: \caption{Spatial distributions of physical quantities for three models
549: at 1 Myr. Colored dots on upper three panels
550: represent the distributions of SPH particles on
551: a slice which includes the symmetry axis (dashed line).
552: The colors of particles
553: denote the H$_2$ fraction, whose legend is shown on the upper left.
554: The yellow star represents the position of a source Pop III star.
555: Lower three panels show various physical quantities along the
556: symmetry axis
557: as $T$[K], $n_{\rm H} [{\rm cm^{-3}}]$, $y_{\rm HI}$, $y_{\rm H_2}$,
558: and $N_{\rm H_2,14} ($H$_2$ column density from the source star in units of
559: $10^{14}{\rm cm^{-2}}$).
560: Horizontal axis shows the distance from the source star.
561: Arrows show the correspondence of coordinates
562: between the SPH distributions and the horizontal axis. }\label{3models}
563: \end{center}
564: \end{figure}
565:
566: \end{document}
567: