1: \documentclass[12pt,preprint]{aastex}
2:
3: \def\lya{Ly$\alpha$}
4: \def\year{yr}
5: \def\kms{km s$^{-1}$}
6: \def\msun{M$_{\odot}$}
7: \def\ergcm2s{\ifmmode {\rm\,erg\,cm^{-2}\,s^{-1}}\else
8: ${\rm\,ergs\,cm^{-2}\,s^{-1}}$\fi}
9: \newcommand{\fluxunit}{\ergcm2s}
10:
11: \begin{document}
12:
13: \title{X-ray nondetection of the Ly$\alpha$ Emitters at $z$ $\sim$ 4.5}
14:
15: \author{J. X. Wang\altaffilmark{1}, J. E. Rhoads\altaffilmark{2}, S. Malhotra\altaffilmark{2}, S. Dawson\altaffilmark{3}, D. Stern\altaffilmark{4}, A. Dey\altaffilmark{5}, T. M. Heckman\altaffilmark{1}, C. A. Norman\altaffilmark{1,2}, H. Spinrad\altaffilmark{3}}
16: \begin{abstract}
17:
18: The \lya\ emitters found at $z \sim 4.5$ by the Large Area Lyman
19: Alpha (LALA) survey have high equivalent widths in the \lya\ line,
20: which can be produced by either narrow-lined active galactic nuclei (AGNs)
21: or by stellar populations with a very high proportion of young,
22: massive stars. To investigate the AGN scenario,
23: we obtained two deep $Chandra$ exposures to study the
24: X-ray nature of the \lya\ emitters.
25: The 172 ks deep $Chandra$ image on the LALA Bo\"{o}tes field
26: was presented in a previous paper (Malhotra et al. 2003),
27: and in this paper we present a new $Chandra$ deep exposure (174 ks)
28: on the LALA Cetus field, which doubled our sample of X-ray
29: imaged \lya\ sources, and imaged the brightest source among our
30: \lya\ emitters.
31: None of the 101 \lya\ sources covered by two $Chandra$ exposures
32: was detected individually in X-ray, with a $3 \sigma$ limiting
33: X-ray flux of $F_{\rm 0.5-10.0 keV} < 3.3 \times 10^{-16}$ \fluxunit\
34: for on-axis targets.
35: The sources remain undetectable in the stacked image, implying
36: a 3$\sigma$ limit to the average luminosity of $L_{\rm 2-8keV} < 2.8 \times 10^{42}$ ergs s$^{-1}$.
37: The resulting X-ray to \lya\ ratio is $>$ 21 times lower than the
38: ratios for known high redshift type-II quasars.
39: Together with optical spectra obtained at Keck, we conclude that
40: no evidence of AGN activity was found among our \lya\ emitters
41: at $z \sim 4.5$.
42: \end{abstract}
43:
44: \keywords{galaxies: active --- galaxies: high-redshift --- galaxies: starburst --- X-rays: galaxies}
45:
46: \altaffiltext{1}{Department of Physics and Astronomy, Johns Hopkins University, 3400 N. Charles Street, Baltimore, MD 21218; jxw@pha.jhu.edu, heckman@pha.jhu.edu, norman@stsci.edu.}
47: \altaffiltext{2}{Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218; san@stsci.edu, rhoads@stsci.edu.}
48: \altaffiltext{3}{Department of Astronomy, University of California at Berkeley, Mail Code 3411, Berkeley, CA 94720; sdawson@astron.berkeley.edu, spinrad@astro.berkeley.edu.}
49: \altaffiltext{4}{Jet Propulsion Laboratory, California Institute of Technology, Mail Stop 169-506, Pasadena, CA 91109; stern@zwolfkinder.jpl.nasa.gov.}
50: \altaffiltext{5}{National Optical Astronomy Observatory, 950 North Cherry Avenue, Tucson, AZ 85719; dey@noao.edu}
51: \section {Introduction}
52: More than three decades ago Partridge \& Peebles (1967) predicted
53: that galaxies undergoing their first throes of star formation
54: should be strong emitters in the \lya\ line.
55: Recently, narrow-band surveys have been successful in finding \lya\
56: emitters at ever increasingly high redshifts.
57: These include our Large Area
58: Lyman Alpha (LALA) survey (e.g., Rhoads et al. 2000) and other recent
59: searches over smaller volumes (Cowie \& Hu 1998; Hu et al 1998, 2002, 2004;
60: Kudritzki et al 2000; Fynbo, Moller, \& Thomsen 2001; Pentericci et
61: al. 2000; Stiavelli et al. 2001; Ouchi et al. 2003; Fujita et
62: al. 2003, Shimasaku et al. 2003, Kodaira et al. 2003).
63: The equivalent widths (EW) of
64: \lya\ emitters selected using narrow-band surveys tend to be large
65: (Malhotra \& Rhoads 2002, hereafter MR02; Kudritzki et al. 2000).
66: With the LALA survey we have found about 400 \lya\ candidates at
67: $z \sim 4.5$, 101 of which are covered by our deep $Chandra$ images
68: (Malhotra et al. 2003, hereafter M03; and this letter).
69: We have obtained optical spectra for a subset of the
70: \lya\ candidates, resulting
71: in 47 spectroscopical confirmations to date and a spectroscopical
72: success rate of $>$ 72\% (Dawson et al. 2004, in prep).
73: The median equivalent width of the \lya\ line
74: is greater than 240\AA\ in the LALA $z \sim 4.5$ sample (MR02), and
75: comparable at z $\sim$ 5.7 (Rhoads \& Malhotra 2001, Rhoads et al. 2003).
76: Hu et al. (2004) have found a somewhat smaller fraction (25\% with
77: EW $>$ 240\AA) at z $\sim$ 5.7 based on a sample that is spectroscopically
78: confirmed but much smaller than the MR02 sample.
79: Normal stellar populations can produce \lya\ emission with equivalent width
80: 240\AA\ or less (Charlot \& Fall 1993), unless they have
81: a top-heavy initial mass function (IMF),
82: zero (or very low) metallicity, and/or extreme youth (age $< 10^7$
83: years).
84: By comparison, stellar populations older than 10$^7$ years and with a Salpeter IMF = 2.35 can only produce \lya\ emission with equivalent width of $\sim 100$ \AA\ (see Fig.2 of MR02).
85: The high equivalent widths could also be explained if narrow-lined
86: (type II) active
87: galactic nuclei (AGNs) were present in our \lya\ emitter sample.
88: The broad-lined (type I) AGNs are ruled out because we see no evidence
89: of broad emission lines from either narrow-band imaging or spectroscopy.
90:
91: A large population of high redshift \lya\ emitting AGNs would have
92: interesting implications both for the pace of black hole formation
93: and for cosmic background radiation from the
94: gamma ray to the far infrared. Of particular interest is the
95: possibility of a large population of type II quasars, i.e.,
96: high luminosity AGNs ($L_{\rm X} >$ 10$^{44}$ ergs s$^{-1}$)
97: whose broad line regions and soft X-rays are greatly attenuated by
98: large column densities of gas and dust.
99: Deep X-ray surveys have uncovered the first X-ray selected type~II quasars
100: (Norman et al 2002, Stern et al. 2002, Dawson et al. 2003, Della Ceca et al. 2003), and
101: more candidates (i.e., Fiore et al. 2003, Crawford et al. 2000, Mainieri et al.
102: 2002).
103: Inspired by these examples, and the fact that the \lya\ line luminosities
104: of the LALA z $\sim$ 4.5 candidates are comparable to these of the high
105: redshift type~II quasars, we set out to image the LALA survey fields with
106: $Chandra$ Advanced CCD Imaging Spectrometer (ACIS) to explore the possibility
107: that some of the \lya\ emitters might be type II quasars.
108:
109: Two deep $Chandra$ ACIS images were obtained to check for the presence
110: of type II quasars among our \lya\ emitters.
111: The results of the 172 ks $Chandra$ deep image on the LALA Bo\"{o}tes field
112: was provided by M03. In this paper,
113: we present a new 174 ks $Chandra$ deep image on LALA Cetus field,
114: which doubled our sample to 101 X-ray imaged \lya\ emitters.
115: The new $Chandra$ image was also designed to cover the ``boomer"
116: of our \lya\ emitters: LALA J020438.95-051116.2,
117: which is the brightest confirmed \lya\ emitter selected by
118: by the LALA survey, with line flux of $9.3 \times 10^{-17} \fluxunit$ and a
119: rest frame equivalent width of $650$~\AA.
120:
121: \section{Optical and $Chandra$ imaging}
122: The LALA survey comprises two primary fields, located in Bo\"{o}tes
123: (J142557+3532) and in Cetus (J020520--0455).
124: Each field is $36\arcmin \times 36\arcmin$ in size,
125: corresponding to a single field of the 8192$\times$8192 pixel Mosaic CCD
126: cameras at the National Optical Astronomy Observatory's 4 meter telescopes.
127: For each field, 5 partially overlapping narrow band filters covering
128: $4.37 < z < 4.57$ for \lya\ were utilized to obtain deep narrowband
129: images. These filters have full
130: width at half maximum (FWHM) $\approx 80$\AA\ and are spaced at
131: $\approx 40$\AA\ intervals (central wavelengths $\lambda_c \approx
132: 6559$, $6611$, $6650$, $6692$, and $6730$\AA).
133: For the Bo\"{o}tes field, 3 more narrowband images covering
134: $z \sim 5.7$ (Rhoads et al. 2003) and $z \sim 6.5$ (Rhoads et al. 2004)
135: are also available.
136: Imaging data reduction followed the methods described in Rhoads et al
137: (2000), and \lya\
138: candidates were selected using criteria described by Rhoads \& Malhotra (2001).
139: This resulted in a total of $\sim$ 400 good candidates at $z \sim 4.5$,
140: which will be presented in a future paper (Rhoads et al. 2004,
141: in preparation).
142:
143: Two deep $Chandra$ ACIS images were obtained for the LALA Bo\"{o}tes
144: and Cetus fields.
145: The $Chandra$ observations were designed to maximize the number of
146: large equivalent width sources within the ACIS-I field of view.
147: The 172 ks (net exposure) $Chandra$ exposure of the LALA Bo\"{o}tes field
148: was taken in 2002, and details of the X-ray data analyses can be found in
149: M03 and Wang et al. (2004).
150: The new 174 ks (net exposure) $Chandra$ ACIS exposure on the LALA Cetus field
151: was obtained in very faint (VFAINT) mode on UT 2003 June 13-15.
152: Data reduction was done with the package CIAO 2.2.1 (see
153: http://asc.harvard.edu/ciao), following the procedures described in M03.
154: The average offset between the X-ray and optical images was obtained by
155: comparing X-ray source positions with optical counterparts
156: (whenever found). The derived 0.3$\arcsec$ offset has been
157: applied for all subsequent analyses.
158: No obvious rotation and plate-scale effects were discovered.
159: We ran WAVDETECT (Freeman et al. 2002) on the soft (0.5 -- 2.0 keV),
160: hard (2.0 -- 7.0 keV), and total band (0.5 -- 7.0 keV) X-ray images.
161: The detailed X-ray data reduction and the detected X-ray sources will be
162: presented in a future paper (Wang et al. 2004, in preparation).
163:
164:
165: \section{X-ray imaging Results}
166: \subsection{Non-detection of individual sources}
167: A total of 101 \lya\ sources were imaged by the two $Chandra$
168: exposures, 49 in the Bo\"{o}tes field, and 52 in the Cetus field.
169: None were detected.
170: To perform X-ray photometry analyses, for each \lya\ source
171: we defined a circular source region
172: centered at the source position on the X-ray images, with radius R$_s$ set
173: to the 95\% encircled-energy radius of {\it Chandra\/} ACIS PSF at the
174: position.
175: The source regions defined above were used to extract source
176: photons for the \lya\ sources, and the backgrounds were extracted
177: from an annulus with $1.2 R_s < R < 2.4 R_s$ after masking out nearby
178: X-ray sources. The net count of each source was derived after accounting
179: for differences of effective exposure time between source regions
180: and background regions (mainly due to CCD edge effects and bad columns).
181: We found that none of the \lya\ sources has an X-ray net count above
182: 2 $\sigma$ level in either 0.5 -- 2.0 keV, 2.0 -- 7.0 keV or 0.5 -- 7.0 keV band. In Fig. \ref{hist}
183: we present the histogram distributions of the signal-to-noise ratios (S/N)
184: of the X-ray net counts from \lya\ sources.
185:
186: We conclude that none of the \lya\ sources are detected by the
187: X-ray observations, allowing us to place upper limits on their
188: X-ray fluxes. For the \lya\ source nearest
189: to the axis of X-ray observation (with an off-axis angle of
190: 1.7$\arcmin$), there are no photons
191: within R$_s$ in the 0.5 -- 7.0 keV band.
192: The 3$\sigma$ level upper limit of X-ray count is
193: 6.61 (Gehrels 1986).
194: Assuming a powerlaw spectrum with photon index of 2.0, this corresponds
195: to a 3$\sigma$ upper limit of X-ray flux $F_{\rm 0.5 - 10.0 keV} <
196: 3.3 \times 10^{-16}$ \fluxunit.
197: Note that to make our fluxes in this paper directly comparable to those in
198: other literatures,
199: we convert the 0.5 -- 7.0 keV band X-ray counts to $F_{\rm 0.5 - 10.0 keV}$.
200: Other sources have higher upper
201: limits due to the lower effective areas and larger PSF sizes.
202: The ``boomer", LALA J020438.95-051116.2, is located at off-axis angle
203: of 6.11\arcmin\ in the $Chandra$ image.
204: We found no evidence of X-ray emission from the ``boomer": X-ray
205: photometry yields a net count of 1.0 in the 0.5 -- 7.0 keV band,
206: with 3$\sigma$ upper limit of 12.2.
207: The corresponding 3$\sigma$ upper limit to the X-ray flux is
208: $F_{\rm 0.5 - 10.0 keV} = 6.6 \times 10^{-16}$ \fluxunit.
209:
210: \subsection{Stacking analysis}
211: The symmetric histograms in Fig. \ref{hist} suggest that X-ray emission
212: will not be detectable in summed images of either the total sample of
213: \lya\ sources, or of the sub-sample of 64 \lya\ sources with rest frame
214: EWs $>$ 240\AA.
215: We performed a stacking analysis to calculate
216: the upper limit of the average X-ray flux for the \lya\ sources.
217: We note that the X-ray fluxes of \lya\ sources located at
218: larger off-axis angles ($>$ 8\arcmin) are less constrained,
219: because of their larger PSF sizes and lower effective exposures.
220: Thus we only stacked the $Chandra$ images for the 70 \lya\ sources
221: (69\% of the whole sample) with off-axis angles $<$ 8\arcmin,
222: providing an effective exposure time of $\sim$ 10.6 Ms.
223: The stacked 0.5 -- 7.0 keV image is presented in Fig. \ref{stack}.
224: The second panel of Fig.~2 presents the
225: stacked image for the subset of these sources with larger \lya\ equivalent
226: widths (EW $>$ 240 \AA\ in the rest frame). No ~cumulative~ X-ray emission
227: is detected in either stacked image.
228: For each of the 70 sources, X-ray photons within
229: 80\% encircled-energy radius of {\it Chandra\/} ACIS PSF were
230: extracted and summed up. After subtracting the local background
231: and applying aperture corrections, we obtained a net count of
232: -12 in the 0.5 -- 2.0 keV band, and -20 in the 0.5 -- 7.0 keV band,
233: with 3$\sigma$ upper limits of 29 and 52 respectively.
234: The corresponding 3$\sigma$ upper limit on the average X-ray flux for
235: the $z \sim$ 5 \lya\ source population is
236: $F_{\rm 0.5 - 2.0 keV} < 1.3 \times 10^{-17}$ \fluxunit,
237: which is 67\% of the limit in M03\footnote{
238: Using the 95\% encircled energy radius for the ACIS
239: PSF provides a somewhat degraded upper limit to the average X-ray flux.
240: The method for calculating the average X-ray flux upper limit described in
241: $\S$3.2 of M03 is somewhat inaccurate. However, the limit presented there
242: is very close to the correct value determined using the method provided here.
243: },
244: and $F_{\rm 0.5 - 10.0 keV} < 4.2 \times 10^{-17}$ \fluxunit.
245: Note that including sources with off-axis angles $>$ 8\arcmin\ will
246: increase the summed effective exposure time to $\sim$ 14 Ms;
247: however, it does not provide an improved upper limit on the average
248: X-ray flux.
249:
250: \section {Discussion and Conclusions}
251:
252: We imaged 101 \lya\ emitters at z $\sim$ 4.5 with two
253: 170 ks deep $Chandra$ ACIS exposures. None were clearly detected
254: at X-ray energies.
255: In Fig. 3, we present the 1$\sigma$ upper limits
256: of the X-ray fluxes against their \lya\ line fluxes, which in most cases are
257: the only well measured properties of the \lya\ sources.
258: For comparison, three high redshift type II quasars,
259: CDF-S202 ($z$ = 3.7, Norman et al. 2002), CXO52 ($z$ = 3.288, Stern et al. 2002) and HDFX28 ($z$ = 2.011, Dawson et al. 2003) are
260: also plotted in the figure.
261: We found that after scaling the X-ray fluxes with \lya\ line fluxes, all the
262: \lya\ emitters at $z \sim 4.5$ are fainter in X-ray than HDFX28,
263: $>$ 96\% of them are fainter than CDF-S202,
264: and $>$ 50\% are fainter than CXO52.
265: This indicates that $<$ 4\% of our \lya\ emitters at $z \sim 4.5$ can be
266: type II quasars like CDF-S202, and $<$ 50\% of them can be like CXO52.
267: However, that average X-ray to \lya\ ratio (1$\sigma$ upper limit) is
268: $>$ 21 times fainter than that of CXO52, which means actually $<$ 4.8\% of
269: the \lya\ emitters could resemble CXO52.
270:
271:
272: Could some of our \lya\ emitters still be Seyfert 2 galaxies (i.e., lower
273: luminosity type II AGNs)?
274: Kriss (1984, 1985) presented samples of low redshift ($z<1$) Seyferts
275: and quasars with good measurements of \lya\ line and X-ray luminosities.
276: The AGNs in Kriss's samples have X-ray luminosities
277: spreading from L$_X$ $\sim$ 10$^{42}$ erg s$^{-1}$ to 10$^{47}$ erg s$^{-1}$,
278: but no evidence of a correlation between their X-ray to \lya\ ratios and
279: X-ray luminosities was found.
280: This indicates that the X-ray to \lya\ ratios of low luminosity AGNs (i.e.,
281: Seyferts) are similar to these of luminous ones (i.e., quasars), thus the average X-ray to \lya\ ratio (upper limit) of our
282: \lya\ emitters is not only significantly fainter than luminous quasars,
283: but also lower luminosity AGNs.
284: The average X-ray to \lya\ ratio of Kriss's samples is also plotted in Fig. 3,
285: which is pretty close to these of the three type II quasars, and 8.5
286: times higher than the average ratio (upper limit) of our \lya\ emitters at
287: $z \sim 4.5$.
288:
289: No evidence of AGN activity among our \lya\ emitters was found in
290: optical spectra either.
291: At $z \sim 4.5$, we have 18 confirmed
292: \lya\ lines with Keck/LRIS, and 29 with Keck/DEIMOS (Dawson et al. 2004).
293: The \lya\ lines are found to be
294: narrow ($<$ 500 \kms) in all the spectroscopically confirmed \lya\ emitters
295: (e.g., Rhoads et al. 2000, Rhoads et al. 2003, Dawson et al. 2004).
296: This is in fact narrower than the typical
297: physical line widths of even type II quasars (see Rhoads et al 2003).
298: None of the spectra shows high excitation emission lines either (e.g., NV, CIV, HeII and CIII]),
299: even in the stacked spectrum (Dawson et al. 2004).
300: In Fig.4, we presented the Keck LRIS spectrum of our ``boomer",
301: LALA J020438.95-051116.2, which has been confirmed to be normal
302: star forming galaxy at z=4.457. We can clearly see a narrow
303: (with a full width half maximum of $<$ 400 \kms) and asymmetric
304: \lya\ line, but no other lines indicating AGN activity.
305:
306:
307: We conclude that no evidence of AGN activity among our \lya\ emitters
308: at z $\sim$ 4.5 was found, from either deep $Chandra$ images, or
309: optical spectra.
310: The $3\sigma$ upper limit of 0.5 -- 2.0 keV flux on the average LALA sources
311: corresponds to an X-ray luminosity of $2.8 \times 10^{42}$ ergs s$^{-1}$ at
312: $z \sim 4.5$ (for either 0.5 -- 2.0 keV rest frame or 2.0 -- 8.0 keV rest
313: frame bandpass, H0=65 km s$^{-1}$ Mpc$^{-1}$, $\Omega_m$=1/3,
314: $\Omega_\lambda$=2/3).
315: The corresponding $3\sigma$ upper limits of L$_{\rm 2-8keV}$ for on-axis
316: individual \lya\ emitters and the ``boomer" are 3.3 $\times 10^{43}$ and 6.6 $\times 10^{43}$
317: ergs s$^{-1}$ respectively.
318: Our analysis indicates that $<$ 4.8\% of our \lya\ emitters could
319: be possible AGNs based on their average X-ray to \lya\ ratio.
320: Steidel et al. (2002) found that 3\% of their Lyman break galaxies
321: at $z \sim 3$ show clear evidence for AGN activity.
322: It's interesting to note that the upper limit of AGN fraction of our
323: \lya\ emitters is well consistent with that of Lyman break galaxies.
324:
325: \acknowledgements
326: This work was supported by the CXC grant GO2-3152x and GO3-4148X, and
327: has benefited from images provided by the NOAO Deep Wide-Field
328: Survey (NDWFS; Jannuzi \& Dey 1999, Brown et al. 2003),
329: which is supported by the National
330: Optical Astronomy Observatory (NOAO). NOAO is operated by AURA,
331: Inc., under a cooperative agreement with the National Science
332: Foundation. We would like to thank B. T. Jannuzi for his help on
333: narrow band imaging observations, and L. Kewley for
334: checking the \lya\ flux of CDF-S 202 in the observed frame.
335: The work of SD was supported by IGPP--LLNL University
336: Collaborative Research Program grant \#03--AP--015.
337: The work of DS was carried out at Jet Propulsion Laboratory, California Institute of
338: Technology, under a contract with NASA.
339: This work includes observations
340: made at the W.M. Keck Observatory, which is operated as a scientific
341: partnership among the California Institute of Technology, the University
342: of California and the National Aeronautics and Space Administration.
343:
344: \begin{references}
345: \reference{} Brown, M, Dey, A, Jannuzi, B. T., Lauer, T. R., Tiede, G. P., \& Mikles, V. J. 2003, ApJ, 597, 225
346: \reference{} Cowie, L. L., \& Hu, E. M. 1998, AJ, 115, 1319
347: \reference{} Charlot, S. \& Fall, S. M. 1993, ApJ, 415, 580
348: \reference{} Crawford, et al. 2002, MNRAS, 333, 809
349: \reference{} Dawson, S., McCrady, N., Stern, D., Spinrad, H., Graham, J., \& Dey, A. 2003, AJ, 125, 1236
350: \reference{} Della Ceca, R. et al. 2003, A\&A, 406, 555
351: \reference{} Fiore, F. et al. 2003, A\&A, 409, 79
352: \reference{} Freeman, P.E., Kashyap, V., Rosner, R., \& Lamb, D. Q., 2002,
353: ApJS, 138, 185
354: \reference{} Fujita, S. S., et al 2003, AJ, 125, 13
355: \reference{} Fynbo, J. U., Moller, P., \& Thomsen, B. 2001, A\&A, 374, 443
356: \reference{} Gehrels, N. 1986, \apj, 303, 336
357: \reference{} Hu, E. M., Cowie, Lennox L., \& McMahon, Richard G. 1998, ApJ, 502, L99
358: \reference{} Hu, E. M., et al. 2002, ApJ, 568, L75
359: \reference{} Hu, E. M., Cowie, L. L., Capak, P., McMahon, R. G., Hayashino, T., Komiyama, Y. 2004, AJ, 127, 563
360: \reference{} Jannuzi, B. T., \& Dey, A. 1999, in ASP Conf. Ser. 191, Photometric
361: redshifts and High-Redshift Galaxies, ed. R. J. Weymann, L. J. Storrie-Lombardi,
362: M. Sawicki, \& R. J. Brunner (San Francisco: ASP), 111
363: \reference{} Kodaira, K. et al. 2003, PASJ, 55, L17
364: \reference{} Kriss, G. A. 1984, \apj, 277, 495
365: \reference{} Kriss, G. A. 1985, \aj, 90, 1
366: \reference{} Kudritzki, R.-P. et al. 2000, ApJ, 536, 19
367: \reference{} Malhotra, S. \& Rhoads, J.E. 2002, \apj, 565, L71
368: \reference{} Malhotra, S., Wang, J, Rhoads, J. E., Heckman, T. M., \& Norman, C. A., 2003, ApJ, 585, L25
369: \reference{} Mainieri, V. et al. 2002, A\&A, 393, 245
370: \reference{} Norman, C. et al. 2002, \apj, 571, 218
371: \reference{} Ouchi, M. et al 2003, ApJ, 582, 60
372: \reference{} Partridge, R. B. \& Peebles, P. J. E. 1967, ApJ, 148, 377
373: \reference{} Pentericci, L. et al. 2000, A\&A, 361, 25
374: \reference{} Rhoads, J.~E., Malhotra, S., Dey, A., Stern, D.,
375: Spinrad, H., \& Jannuzi, B.~T.\ 2000, ApJ, 545, L85
376: \reference{} Rhoads, J. E. \& Malhotra, S. 2001, ApJ, 563, L5
377: \reference{} Rhoads, J. E. et al. 2003, AJ, 125, 1006
378: \reference{} Rhoads, J. E. et al. 2004, ApJ in press, astro-ph0403161
379: \reference{} Shimasaku, K. et al. 2003, ApJ, 586, L111
380: \reference{} Stern, D. et al. 2002, ApJ, 568, 71
381: \reference{} Steidel, C. et al 2002, ApJ, 576, 653
382: \reference{} Stiavelli, M., Scarlata, C., Panagia, N., Treu, T.,
383: Bertin, G., \& Bertola, F. 2001, ApJ, 561, L37
384: \reference{} Wang, J. et al. 2004, AJ, 127, 213
385: \end{references}
386:
387: \clearpage
388: \begin{figure}
389: \plotone{f1.eps}
390: \caption{
391: Histogram distributions of the X-ray signal-to-noise ratios (S/N)
392: for \lya\ sources in the 0.5 -- 2.0 keV, 2.0 -- 7.0 keV and 0.5 -- 7.0 keV
393: band. The total
394: sample of 101 \lya\ sources is shown as the solid line. The sub-sample of
395: 64 \lya\ sources with rest frame EWs $>$ 240 \AA\ is overplotted as the
396: dotted line. None of the \lya\ sources has an X-ray net count above
397: 2 $\sigma$ level in any band.
398: }
399: \label{hist}
400: \end{figure}
401:
402: \newpage
403: \begin{figure}
404: \plotone{f2.eps}
405: \caption{
406: Stacked 0.5 -- 7.0 keV $Chandra$ images of \lya\ sources.
407: Left: all \lya\ emitters with off-axis angle $<$ 8\arcmin;
408: Right: \lya\ emitters with off-axis angle $<$ 8\arcmin\ and rest frame
409: equivalent widths EW $>$ 240\AA.
410: The effective exposure time of the stacked images
411: is 11.2 Ms (left panel). The images are $\sim$ 20\arcsec $\times$ 20\arcsec\
412: in size, and the circles are centered on the stacking positions with a radius of
413: 2.5\arcsec.
414: }
415: \label{stack}
416: \end{figure}
417:
418: \begin{figure}
419: \plotone{f3.eps}
420: \caption{
421: \lya\ fluxes vs 1$\sigma$ upper limits of the X-ray fluxes
422: for our \lya\ sources at z $\sim$ 4.5, compared with three known high redshift
423: type II quasars. The ``boomer" of our \lya\ sources is plotted by
424: a big triangle. The average \lya\ flux is plotted against
425: the 1$\sigma$ upper limit of the average X-ray flux.
426: Lines indicate constant X-ray to \lya\ flux ratios.
427: The average X-ray to \lya\ ratio for the lower luminosity, low redshift
428: ($z<1$) Seyferts and quasars samples (Kriss 1984, 1985) is also plotted.
429: }
430: \end{figure}
431:
432: \begin{figure}
433: \plotone{f4.eps}
434: \caption{
435: Keck LRIS Spectrum of the ``boomer" of our \lya\ emitters at z $\sim$ 4.5.
436: The source is confirmed to have a spectroscopic redshift of 4.457.
437: The photon counting errors are plotted in the bottom, and the insert shows
438: the profile of the narrow and asymmetric \lya\ line.
439: }
440: \end{figure}
441: \end{document}
442:
443:
444: