0707.1663/ms.tex
1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass[preprint]{aastex}
3: 
4: \newcommand{\lya}{Ly$\alpha$}
5: \newcommand{\nv}{N\,{\sc v}}
6: \newcommand{\siii}{Si\,{\sc ii}}
7: \newcommand{\oi}{O\,{\sc i}}
8: \newcommand{\cii}{C\,{\sc ii}}
9: \newcommand{\siiv}{Si\,{\sc iv}}
10: \newcommand{\civ}{C\,{\sc iv}}
11: \newcommand{\heii}{He\,{\sc ii}}
12: \newcommand{\oiii}{O\,{\sc iii]}}
13: \newcommand{\aliii}{Al\,{\sc iii}}
14: \newcommand{\siiii}{Si\,{\sc iii]}}
15: \newcommand{\ciii}{C\,{\sc iii]}}
16: \newcommand{\mgii}{Mg\,{\sc ii}}
17: \newcommand{\feii}{Fe\,{\sc ii}}
18: \newcommand{\hb}{H$\beta$}
19: 
20: \newcommand{\civd}{C\,{\sc iv} $\lambda\lambda$1548,1550}
21: \newcommand{\feiiaa}{Fe\,{\sc ii} $\lambda$1608}
22: \newcommand{\aliia}{Al\,{\sc ii} $\lambda$1670}
23: \newcommand{\aliiid}{Al\,{\sc iii} $\lambda\lambda$1854,1862}
24: \newcommand{\feiiab}{Fe\,{\sc ii} $\lambda$2600}
25: \newcommand{\mgiid}{Mg\,{\sc ii} $\lambda\lambda$2796,2803}
26: \newcommand{\mgia}{Mg\,{\sc i} $\lambda$2852}
27: 
28: \newcommand{\mgit}{Mg\,{\sc i}}
29: \newcommand{\aliit}{Al\,{\sc ii}}
30: 
31: \begin{document}
32: 
33: \title{GEMINI NEAR-INFRARED SPECTROSCOPY OF LUMINOUS {\boldmath $z\sim6$} 
34:   QUASARS: CHEMICAL ABUNDANCES, BLACK HOLE MASSES, AND \mgii\ 
35:   ABSORPTION\footnote{
36:   Based on observations obtained at the Gemini Observatory (acquired through 
37:   the Gemini Science Archive), which is operated by the Association of 
38:   Universities for Research in Astronomy, Inc., under a cooperative agreement 
39:   with the NSF on behalf of the Gemini partnership: the National Science 
40:   Foundation (United States), the Particle Physics and Astronomy Research 
41:   Council (United Kingdom), the National Research Council (Canada), CONICYT 
42:   (Chile), the Australian Research Council (Australia), CNPq (Brazil) and 
43:   CONICET (Argentina).}}
44: 
45: \author{Linhua Jiang\altaffilmark{1}, Xiaohui Fan\altaffilmark{1},
46:   Marianne Vestergaard\altaffilmark{1}, Jaron D. Kurk\altaffilmark{2},
47:   Fabian Walter\altaffilmark{2}, Brandon C. Kelly\altaffilmark{1},
48:   and Michael A. Strauss\altaffilmark{3}}
49: \altaffiltext{1}{Steward Observatory, University of Arizona,
50:   933 North Cherry Avenue, Tucson, AZ 85721}
51: \altaffiltext{2}{Max-Planck-Institut f\"{u}r Astronomie, K\"{o}nigstuhl 17,
52:   D-69117 Heidelberg, Germany}
53: \altaffiltext{3}{Department of Astrophysical Sciences, Princeton University, 
54:   Princeton, NJ 08544}
55: \begin{abstract}
56: 
57: We present Gemini near-infrared spectroscopic observations of six luminous
58: quasars at $z=5.8\sim6.3$. Five of them were observed using 
59: Gemini-South/GNIRS, which provides a simultaneous wavelength coverage of 
60: 0.9--2.5 $\mu$m in cross dispersion mode. The other source was observed in $K$
61: band with Gemini-North/NIRI. 
62: We calculate line strengths for all detected emission lines and use their 
63: ratios to estimate gas metallicity in the broad-line regions of the quasars. 
64: The metallicity is found to be supersolar with a typical value of $\sim4$ 
65: Z$_{\sun}$, and a comparison with low-redshift observations shows no strong 
66: evolution in metallicity up to $z\sim6$. The \feii/\mgii\ 
67: ratio of the quasars is $4.9\pm1.4$, consistent with low-redshift 
68: measurements. We estimate central BH masses of $10^9$ to $10^{10}$ M$_{\sun}$ 
69: and Eddington luminosity ratios of order unity. 
70: We identify two \mgiid\ absorbers with rest equivalent width 
71: $W_0^{\lambda2796}>1$ \AA\ at $2.2<z<3$ and three \mgii\ absorbers with 
72: $W_0^{\lambda2796}>1.5$ \AA\ at $z>3$ in the spectra, with the two most 
73: distant absorbers at $z=4.8668$ and 4.8823, respectively. The redshift number 
74: densities ($dN/dz$) of \mgii\ absorbers with $W_0^{\lambda2796}>1.5$ \AA\ are 
75: consistent with no cosmic evolution up to $z>4$.
76: 
77: \end{abstract}
78: 
79: \keywords
80: {galaxies: high-redshift --- galaxies: active --- quasars: emission lines 
81: --- quasars: absorption lines ---  quasars: general}
82: 
83: \section{INTRODUCTION}
84: 
85: In recent years we have witnessed the discoveries of quasars at $z\sim6$; 
86: at this epoch the universe was less than one Gyr old 
87: \citep[e.g.][]{fan03,fan06,coo06,mcg06}. These high-redshift quasars are among 
88: the most luminous objects known and provide direct probes of the early 
89: universe when the first generations of galaxies and quasars formed. Quasar 
90: activity and the formation processes of galaxies and supermassive black holes 
91: (BHs) are closely correlated \citep[e.g.][]{kau00,wyi03,hop05,cro06}, so these 
92: $z\sim6$ objects are essential for studying the accretion history of BHs, 
93: galaxy formation, and chemical evolution at very early epochs. High-redshift 
94: luminous quasars harbor BHs with masses of $10^9$--$10^{10}$ M$_{\sun}$ and 
95: emit near the Eddington limit \citep[e.g.][]{bar03,ves04,jia06,kur07}, 
96: revealing the rapid growth of central BHs. Their emission line ratios show 
97: solar or supersolar metallicity, as is found in low-redshift quasars 
98: \citep[e.g.][]{bar03,die03a,fre03,mai03}, indicating that there was vigorous 
99: star formation and element enrichment in the first Gigayear.
100: 
101: Understanding quasars requires observations from X-ray to radio wavelengths;
102: each spectral region is due to different physical mechanisms and probes 
103: different regions of the active nucleus. In particular, the rest-frame UV 
104: spectrum, from the \lya\ emission line to the \feii\ bump at 2000--3000 \AA, 
105: contains strong diagnostic emission lines and provides key information on the 
106: physical conditions and emission mechanisms of the broad-line region (BLR). 
107: The bulk motions of the BLR can be used to determine the mass of the central 
108: BH, while chemical abundances in the BLR are important in understanding
109: the history of star formation in the host galaxy. Photoionization models show 
110: that various emission-line ratios provide reliable estimates of metallicity in 
111: the BLRs of quasars \citep[e.g.][]{ham02,nag06}. Strong permitted line ratios, 
112: such as \nv\ $\lambda$1240/\civ\ $\lambda$1549 (hereafter \nv/\civ),
113: \nv\ $\lambda$1240/\heii\ $\lambda$1640 (hereafter \nv/\heii), and 
114: \feii/\mgii\ $\lambda$2800 (hereafter \feii/\mgii), are relatively easy to 
115: measure even in $z\sim6$ quasars \citep[e.g.][]{ham93,ham99}.
116: Fe abundance is of particular interest \citep[e.g.][]{ham93,ham99}: most of 
117: the Fe in the solar neighborhood is generated by Type Ia supernovae 
118: (SNe Ia), whose precursors are believed to be long-lived intermediate-mass 
119: stars in close binaries, so appreciable Fe enrichment can only happen on a 
120: timescale of one Gyr after the initial starburst
121: \citep[e.g.][]{gre83}. On the contrary, the production of $\alpha$ elements
122: such as Mg and O is dominated by SNe of types II, Ib and Ic, which 
123: explode very soon after the initial burst. Therefore, the Fe/$\alpha$ ratio is 
124: expected to be a strong function of age in young systems, and thus it puts a 
125: useful constraint on the age and chemical enrichment of the gas in the quasar
126: environment.
127: 
128: At low redshift, elemental abundances measured from strong emission and 
129: intrinsic absorption lines have revealed that quasar environments have solar 
130: or supersolar metallicity. At $z\sim6$, measurement of these spectral lines 
131: is difficult, as the rest-frame UV waveband is 
132: redshifted to the NIR range. NIR spectroscopy has been obtained for a few 
133: $z\sim6$ quasars \citep[e.g.][]{goo01,pen02,bar03,fre03,mai03,ste03,iwa04}. 
134: These observations only provide NIR spectra with low signal-to-noise ratios 
135: (SNRs) or partial wavelength coverage. Even basic properties such as the 
136: average continuum slope are not well measured. Recently, \citet{kur07} 
137: obtained NIR spectra with high SNRs for a sample of five $z\sim6$ quasars with 
138: ISAAC on VLT. 
139: They used the ISAAC data to study the \feii/\mgii\ ratios, BH masses, and the 
140: Str\"omgren spheres around quasars. In this paper we present Gemini NIR 
141: spectroscopy of six $z\sim6$ quasars.
142: All these quasars were discovered by the Sloan Digital Sky Survey 
143: \citep[SDSS;][]{yor00} and were selected from Fan et al. 2000, 2001, and 2004 
144: (hereafter quasar discovery papers). They have extensive follow-up 
145: observations in X-ray \citep[e.g.][]{she06}, $Spitzer$ bands 
146: \citep[e.g.][]{jia06}, millimeter/submillimeter and radio wavelengths 
147: \citep[e.g.][]{wan07}, providing a fundamental dataset to study luminous 
148: quasars at $z\sim6$ in the context of the growth of early BHs and their 
149: relation to galaxy formation. The Gemini data have high SNRs and excellent 
150: NIR wavelength coverage. They are used to measure the metal abundances in
151: high-redshift environments and estimate central BH masses using empirical mass 
152: scaling relations \citep[e.g.][]{mcl04,ves06}.
153: 
154: We also use these high-redshift spectra to study the evolution of strong 
155: intergalactic \mgii\ absorption at $z<6$. Quasar absorption lines are an 
156: unbiased tracer of intervening gaseous material along the lines of sight to 
157: quasars. 
158: Strong \mgii\ absorption systems are usually directly associated with 
159: galaxy halos or disks \citep[e.g.][]{cha98,chu00,ste02}. The statistical
160: properties of the \mgiid\ doublet absorbers at low redshift ($z\le2.2$) have 
161: been well studied (e.g. Steidel \& Sargent 1992; Nestor et al. 2005; Prochter 
162: et al. 2006; see a summary of \mgii\ absorption surveys by Churchill
163: et al. 2005). At higher redshift, the \mgii\ doublet is shifted to the NIR 
164: wavelength range, where detectors are much less efficient than are detectors 
165: at the optical region. The study of high-redshift \mgii\ absorbers is 
166: therefore
167: limited by the small number of quasars with good NIR spectra available. Our 
168: Gemini/GNIRS spectra with high SNRs have excellent NIR wavelength coverage,
169: so they are efficient to probe the \mgii\ absorption systems at $z<6$.
170: 
171: The structure of the paper is as follows. In $\S$ 2 we describe our Gemini 
172: observations and data reduction of the six quasars. In $\S$ 3 we conduct
173: a detailed spectral analysis, including the measurement of emission-line 
174: strengths. Chemical abundances are calculated and central BH masses are 
175: estimated in $\S$ 4. \mgii\ absorption systems in the quasar spectra are 
176: identified and analyzed in $\S$ 5. We give a brief summary in $\S$ 6. 
177: In this paper we use $\lambda$ ($\lambda_0$) to denote 
178: observed-frame (rest-frame) wavelength. We use a $\Lambda$-dominated flat 
179: cosmology with H$_0$ = 70 km s$^{-1}$ Mpc$^{-1}$, $\Omega_{m}$ = 0.3 and
180: $\Omega_{\Lambda}$ = 0.7 \citep{spe07}.
181: 
182: \section{OBSERVATIONS AND DATA REDUCTION}
183: 
184: We obtained NIR spectra for the six luminous $z\sim6$ quasars using 
185: Gemini-South/GNIRS and Gemini-North/NIRI. Five of them were chosen to be 
186: easily reached by Gemini-South, and were observed using GNIRS in February and 
187: March 2006. The observations on GNIRS were carried out in cross-dispersion 
188: mode, giving a simultaneous wavelength coverage of 0.9--2.5 $\mu$m with 
189: excellent transmission. At $z\sim6$, strong diagnostic emission lines such as 
190: \civ\ and \mgii\ are redshifted to this wavelength range, so GNIRS provides us 
191: an efficient way to collect NIR information for high-redshift quasars. We used 
192: the short camera on GNIRS with a pixel scale of $0.15\arcsec$/pixel. The slit 
193: length in cross-dispersion mode is $6.1\arcsec$ and we used a slit width of 
194: $0.675\arcsec$ to optimize the combination of resolution and light gain. The 
195: resolving power is about 800 and the dispersion varies from $\sim$3 \AA\ per
196: pixel at 0.9 $\mu$m to $\sim$6 \AA\ per pixel at 2.5 $\mu$m. This is 
197: sufficient to separate the \mgiid\ doublet absorption lines. We used 
198: the standard ABBA nodding sequence between exposures. The exposure time at 
199: each nod position was five minutes and the distance between the two positions 
200: was $3\arcsec$.
201: Before or after the exposure of each quasar, a nearby A or F spectroscopic 
202: standard star was observed for flux calibration and to remove telluric
203: atmosphere absorption. The log of observations is given in Table 1, where
204: redshifts, $z_{AB}$ (AB magnitudes) and $J$ (Vega-based magnitudes) are taken 
205: from the quasar discovery papers.
206: 
207: The GNIRS data were reduced using the Gemini package within IRAF\footnote{IRAF 
208: is distributed by the National Optical Astronomy Observatories, which are 
209: operated by the Association of Universities for Research in Astronomy, Inc., 
210: under cooperative agreement with the National Science Foundation.}.
211: Briefly, for each quasar the NIR spectroscopic data were sky-subtracted for 
212: each pair of images taken at the two nod positions. Then the images were 
213: flat-fielded. In cross-dispersion mode, GNIRS uses orders 3--8 to cover 
214: 0.9--2.5 $\mu$m. A single flat has non-uniform illumination between orders,
215: so three sets of flats with different exposure times were taken each night.
216: The final flat is extracted from the three flats so that each order has good 
217: illumination but without saturation. After flat-fielding, distortion 
218: correction and wavelength calibration were applied, the frames were 
219: combined, and one-dimensional spectra were extracted. Then the spectra
220: were flux calibrated and corrected for telluric atmosphere absorption using
221: the spectra of the spectroscopic standard stars. Finally the spectra in 
222: different orders were scaled and connected to a single spectrum. 
223: Adjacent orders cover a short common wavelength range, which enables the 
224: individual orders to be scaled to form a continuous, smoothly connected
225: spectrum. This arbitrarily scaled spectrum was then calibrated to the $J$ 
226: broad-band magnitude listed in Table 1 and thereby placed on an absolute flux 
227: scale with an uncertainty of 5$\sim$10\%. 
228: The typical SNR at 1.2 $\mu$m reaches $\sim8$ per pixel.
229: 
230: Figure 1 shows the spectra of the five quasars in the observed frame. The 
231: spectra at $\lambda\leq1.0$ $\mu$m are taken from the quasar discovery papers. 
232: The spectra have been smoothed by three pixels. The bottom panel shows the sky 
233: transparency at an airmass of 1.0 and water vapor of 1.6 mm. There are strong
234: telluric absorption bands around 1.4 and 1.9 $\mu$m, so we do not show
235: the spectra in these ranges. In Figure 1 we also show a spectrum of a standard 
236: star HIP 54027 calibrated using the same procedure as we do for our science 
237: objects. For comparison, a spectrum (in red) of a black body with the 
238: effective temperature of HIP 54027 is given on top of the spectrum of 
239: HIP 54027. The good agreement between the two spectra indicate that spectra
240: in different dispersion orders are properly scaled and connected.
241: 
242: We obtained a $K$-band spectrum for SDSS J1623+3112\footnote{The naming 
243: convention for SDSS sources is SDSS JHHMMSS.SS$\pm$DDMMSS.S, and the 
244: positions are expressed in J2000.0 coordinates as given in Table 1. We use 
245: SDSS JHHMM$\pm$DDMM for brevity.} using Gemini-North/NIRI in August and 
246: September 2004 and August 2005. The observations were made with a f/6 
247: camera in spectroscopic mode. The long-slit has a length of $110\arcsec$
248: and we chose to use a slit width of $0.75\arcsec$, which provides a resolving
249: power of $\sim500$. The observing strategy and data reduction are similar to
250: those used for the GNIRS data. The standard ABBA mode was used with a 
251: nodding distance of $20\arcsec$ and an exposure time of five minutes at each 
252: nod position. A nearby A or F spectroscopic standard star was observed before
253: or after the target observation. The NIRI data were also reduced using the 
254: Gemini package within IRAF.
255: After sky-subtraction, flat-fielding and wavelength calibration, the 
256: frames were combined. Then one-dimensional spectra were extracted from the 
257: combined images, and were flux calibrated using the spectra of standard stars.
258: The observing log is given in Table 1 and the spectrum is shown in Figure 2.
259: Since we do not have $K$-band photometry for this object, we cannot apply an
260: absolute flux calibration. In Figure 2, we have scaled the spectrum to 
261: its $J$ magnitude assuming a power-law continuum slope of $\alpha_{\nu}=-0.5$
262: ($f_{\nu}\sim\nu^{\alpha_{\nu}}$).
263: 
264: \section{SPECTRAL FIT AND EMISSION-LINE MEASUREMENTS}
265: 
266: To analyze broad emission lines such as \civ\ and \mgii\ we fit and subtract 
267: the UV \feii\ emission that contaminates most of the UV-optical spectral 
268: region; thousands of \feii\ lines are emitted at UV-optical energies that 
269: blend with other lines in the spectrum. We use the UV \feii\ template 
270: presented by \citet{ves01} and adopt the fitting procedure described in detail 
271: in $\S$ 4.2 by these authors. Briefly, the fitting is done as an iteration 
272: over a power-law fit to the continuum emission and a model fit to the \feii\ 
273: emission using a scaled and broadened version of the \feii\ template. The
274: \feii\ template is broadened by convolving the template with Gaussians with
275: a range of sigmas. In the first iteration a power-law is fitted to regions 
276: with very little or no contribution from line emission (e.g. 1275--1295 \AA; 
277: 1325--1335 \AA; 1425--1480 \AA; 2180--2220 \AA; 3000--3040 \AA). Upon 
278: subtraction of this primary continuum fit multiple broadened copies of the 
279: \feii\ template are scaled to regions which predominantly contain Fe
280: emission. The broadened, scaled \feii\ template that provides the best overall 
281: match to the observed \feii\ emission in the observed spectrum is then 
282: selected and subtracted. Some adjustments of the \feii\ fitting regions are 
283: sometimes necessary to optimize the fit, as judged by a minimization of the
284: residual flux upon subtraction of the \feii\ model. After subtraction of the 
285: best matching \feii\ model from the original data (i.e., with the continuum 
286: not subtracted) another power-law fit to the continuum emission is performed; 
287: the lack of (most of the) contaminating \feii\ enables the fitting to be done 
288: over a broader wavelength range, often resulting in a better and more 
289: realistic 
290: continuum fit. After subtraction of this continuum model, the \feii\ fitting 
291: is repeated. The iteration of the continuum and \feii\ fitting is repeated 
292: until these individual fits do not change and the subtraction of the \feii\ 
293: model leaves a smooth, featureless spectrum (barring the strong, broad non-Fe 
294: emission lines) with a realistic power-law slope. Convergence is often 
295: obtained in the second or third iteration. For SDSS J1623+3112, the wavelength 
296: range of the spectrum is too short to allow a reasonable fit to both the 
297: \feii\ emission and a power-law continuum, so we assume a slope of 
298: $\alpha_{\nu}=-0.5$ during the model fitting. This gives a lower limit of the 
299: \feii\ emission. Balmer continuum could be non-negligible at 
300: $\lambda_0\geq2000$ \AA\ \citep[e.g.][]{die02,die03b}, however, 
301: the wavelength coverage of our spectra is not long enough to provide a 
302: realistic fit to it, so we do not consider it in our fits. 
303: The power-law continuum fits to each of the quasars in our sample are listed 
304: in column 3 of Table 2. We note that the accuracy of the derived line width 
305: depends on the quality of the spectra. For widths broader than 5000--6000 km/s 
306: our ability to accurately determine the intrinsic \feii\ line width decreases.
307: 
308: We use \mgii\ emission lines to determine the redshifts of the quasars.
309: Accurate redshifts are important for various studies such as molecular line
310: searches. The redshifts given in the quasar discovery papers were mostly 
311: determined from the \lya\ emission lines, where strong \lya\ absorption 
312: systems usually cause large uncertainties in redshift measurements. While most 
313: of high-ionization, broad emission lines such as \civ\ can be significantly
314: blueshifted with respect to quasar systematic redshifts, the \mgii\ emission 
315: line has a small blueshift and thus provides a reliable redshift measurement
316: \citep[e.g.][]{mci99,van01,ric02}. The \mgii\ emission line at $z\sim6$ is 
317: close to a strong telluric absorption band at 1.9 $\mu$m, as seen in Figure 1.
318: To avoid the effect of the telluric absorption, we measure the line center of 
319: each \mgii\ emission line using a Gaussian profile to fit the top $\sim50$\% 
320: of the line. The line center is also estimated by calculating the 
321: $3\times$median--$2\times$mean mode of the top $\sim50$\% of the line. The two 
322: methods give similar results because the top $\sim50$\% of the line is almost 
323: symmetric. The measured redshifts are given in Column 2 of Table 2. They
324: are in good agreement with the values derived in \citet{kur07}, but differ by
325: 0.02--0.03 from the values in the discovery papers. We did not detect the 
326: \mgii\ emission line for SDSS J0836+0054 and SDSS J1044$-$0125, which were 
327: observed with high humidity. In Table 2, the redshift of SDSS 
328: J0836+0054, measured also using the \mgii\ emission line, is taken from 
329: \citet{kur07}, and the redshift of SDSS J1044$-$0125 is measured from the 
330: \ciii\ $\lambda$1909 emission line using the same method.
331: 
332: We fit individual emission lines after the subtraction of the \feii\ emission
333: and power-law continuum. We use one Gaussian profile to fit most isolated 
334: lines and use multiple Gaussian profiles to fit blended lines. We allow both 
335: the line center and the line width to vary unless stated otherwise. Broad 
336: emission lines in quasars usually show asymmetric velocity profiles, so in the 
337: cases in which a single Gaussian profile does not provide a reasonable fit, we 
338: use double Gaussian profiles. There is no particular physical meaning to the
339: two Gaussian components. For each quasar, the following emission lines 
340: (if detected) are fitted.
341: (1) \lya\ + \nv\ + \siii\ $\lambda1262$ (hereafter \siii). We use four 
342: Gaussian profiles to fit the three lines, with the first two profiles
343: representing 
344: the broad and narrow components of \lya\ and the last two profiles
345: representing \nv\ and \siii, respectively. During the fit, the line centers 
346: are fixed at the values given by \citet{van01}. Since the blue side of the 
347: \lya\ line is strongly absorbed, we only fit the red side of the \lya\ 
348: line. (2) \oi\ $\lambda1304$, \cii\ $\lambda1335$, and \siiv\ $\lambda1397$
349: (hereafter \oi, \cii, and \siiv). We use one Gaussian profile to fit each of 
350: these lines. (3) \civ. The strong \civ\ line is fitted using one or two 
351: Gaussian components. (4) \heii\ + \oiii\ $\lambda1663$ (hereafter \oiii). 
352: These two lines are very weak, but \heii\ is important for determining 
353: chemical abundances. They were tentatively detected in three quasars and 
354: fitted simultaneously using two Gaussians. (5) \aliii\ $\lambda1857$ + \ciii\ 
355: $\lambda1909$ (hereafter \aliii\ and \ciii). The weak \aliii\ line is often 
356: detected on the blue wing of the strong \ciii\ line. We use two Gaussian 
357: profiles to fit these two lines simultaneously if \aliii\ is detected, 
358: otherwise the \ciii\ line is fitted using a single Gaussian profile. We ignore 
359: the weak \siiii\ $\lambda1892$ line on top of \ciii.
360: (6) \mgii. The \mgii\ emission line can be well fitted by double Gaussians
361: \citep[e.g.][]{bar03}, with a weak component located on the blue side of
362: a much stronger component. If the blue side of \mgii\ is affected by telluric 
363: absorptions, we fit the line using a single Gaussian. For the emission lines 
364: that are severely affected by broad-absorption lines (BALs), we fix their 
365: central wavelengths and fit their red half sides using a single Gaussian 
366: profile.
367: 
368: Figures 1 and 2 show the fitting results for the six quasars. The blue solid 
369: lines are the best fits to the emission lines and the blue dashed lines are
370: the best model fits to the power-law continuum and \feii\ emission. The red 
371: lines are the sums of all the components. For comparison, we show the 
372: low-redshift composite spectrum of \citet{van01} in gray. Table 3 shows the 
373: rest-frame EW and FWHM of emission lines derived from the model fitting 
374: results. For the lines which were fitted using double Gaussian profiles, their 
375: EW and FWHM are calculated from the combination of the two components. The 
376: errors given in the table are the formal uncertainties obtained from our 
377: fitting process, so the real errors could be larger in the cases in which the 
378: lines are heavily blended, or we fit a profile of the wrong shape.
379: The \heii\ emission is tentatively detected in
380: three quasars at a significance level of $\sim2\sigma$. We note from Table 2
381: that three quasars have continuum slopes consistent with the slope 
382: $\alpha_{\nu}=-0.44$ of the composite spectrum of low-redshift SDSS quasars 
383: \citep{van01}, while SDSS J1030+0524 and SDSS J1306+0356 show significantly
384: bluer continua with $\alpha_{\nu}\sim0.5$. \citet{pen03} obtained an average 
385: slope of $\alpha_{\nu}=-0.57$ with a scatter of 0.33 from 45 
386: intermediate-redshift SDSS quasars at $3.6<z<5.0$. This indicates that the 
387: continuum slopes in the two unusual quasars deviate from the average slopes
388: at a $\sim3\sigma$ significance level. Table 2 also shows the rest-frame 
389: optical continuum slopes derived from the $Spitzer$ IRAC broad-band photometry 
390: of the same objects \citep{jia06}. The optical continua are generally redder 
391: than the UV continua measured from the Gemini spectra, as also reported by 
392: \citet{van01}.
393: 
394: \section{CHEMICAL ABUNDANCES AND CENTRAL BH MASSES}
395: 
396: \subsection{Chemical Abundances}
397: 
398: Emission-line ratios can be used to measure gas metallicity in the quasar BLR
399: and track the chemical evolution with redshift. When weak emission lines are 
400: difficult to detect in high-redshift quasars, strong broad emission lines such 
401: as \nv, \civ, and \mgii\ become a powerful tool to probe metallicity at 
402: high redshift \citep[e.g.][]{ham92,ham93}. Different elements form on 
403: different timescales. Elements such as C, O, and Mg are 
404: formed in the explosions of massive stars, and their enrichment is rapid; 
405: while the second generation element N is produced from previously 
406: generated C and O, and its enrichment is relatively slow. Helium is 
407: a primordial element and its abundance changes little with cosmic time. 
408: Chemical abundances measured from emission lines and intrinsic absorption 
409: lines show solar or supersolar metallicity in quasar environments at low
410: redshift \citep[e.g.][]{ham07}, and can be as high as 15 times solar 
411: metallicity \citep{bal03}. Furthermore, studies have shown that there is 
412: little chemical evolution up to the highest known redshift 
413: \citep[e.g.][]{bar03}.
414: 
415: We calculate emission-line fluxes from the model fitting results described
416: in $\S$ 3. The fluxes are normalized to the \civ\ fluxes and shown in 
417: Table 4 and Figure 3
418: (filled circles with errors). \citet{nag06} used a sample of over 5000 quasars
419: from the SDSS Data Release Two to study quasar BLRs. They made quasar composite 
420: spectra in the ranges of $2.0\le z \le4.5$ and $-29.5\le M_B \le-24.5$, and 
421: measured emission-line ratios and metallicities in the composite spectra for 
422: each redshift and luminosity bin. They found that most of the line ratios 
423: (with respect to the \civ\ line) that they investigated do not show strong 
424: evolution with redshift. For the purpose of comparison, we show in Figure 3 
425: the line ratios of quasars (filled squares) in the luminosity range of
426: $-27.5<M_B<-26.5$ from \citet{nag06}. Although there are correlations between 
427: quasar luminosities and most line ratios \citep{nag06}, most of our quasars 
428: have luminosities in a small range of $-27.5<M_B<-26.5$. Figure 3 shows 
429: that most of the flux ratios do not exhibit strong evolution up to $z\sim6$. 
430: There is a trend for higher \cii/\civ\ and \oi/\civ\ ratios at higher 
431: redshift. \citet{nag06} already noted that the \oi/\civ\ flux ratio is 
432: marginally correlated with redshift. This trend could also be caused by 
433: the intrinsic dispersion of the flux ratios.
434: 
435: Photoionization models have shown that a series of emission-line ratios can
436: be used to estimate gas metallicity in the BLRs of quasars. \citet{ham02} 
437: studied the relative N abundance as a metallicity indicator based on 
438: the locally optimally emitting cloud (LOC) model \citep{bal95}, and calculated 
439: theoretical emission-line ratios as a function of metallicity Z/Z$_\sun$. We 
440: estimate the metallicity from the \nv/\civ\ and \nv/\heii\ ratios using Figure 
441: 5 of \citet{ham02}. There is a 30\% difference between the solar abundances
442: used by \citet{ham02} and the latest solar abundances \citep{bal03}, so we
443: scale the abundances of \citet{ham02} to match the latest values using the
444: method by \citet{bal03}. The results are shown as the open circles 
445: in Figure 4.
446: \citet{nag06} also used the LOC model, but included more UV emission lines in 
447: their photoionization model. We use Figure 29 of \citet{nag06} to measure the 
448: metallicity from various line ratios and show the measurements as the filled 
449: circles in Figure 4. The two models give similar metallicities from the 
450: \nv/\civ\ and \nv/\heii\ ratios. The metallicities in these high-redshift 
451: quasars are supersolar, with a typical value of $\sim4$ Z$_{\sun}$.
452: \citet{nag06} computed chemical abundances as a function of redshift for the 
453: SDSS composite spectra using their photoionization model. The filled squares 
454: in Figure 4 show the abundances averaged in the luminosity range of 
455: $-28.5<M_B<-25.5$. The filled triangles represent the chemical abundances of 
456: eleven $3.9<z<5.0$ quasars from \citet{die03a}. Figure 4 shows that the
457: metallicity is consistent with no strong evolution within the errors up to 
458: $z\sim6$.
459: 
460: The \feii/\mgii\ ratio is important in understanding chemical evolution at 
461: high redshift. We use the \feii\ emission line complex at 
462: $\lambda_0=2000\sim3000$ \AA\ as the \feii\ flux indicator 
463: \citep[e.g.][]{die02,bar03}. The flux of the \feii\ complex is integrated from 
464: 2200 to 3090 \AA\ over the best-fitting \feii\ template. The derived 
465: \feii/\mgii\ ratios (filled circles) with a typical value of $4.9\pm1.4$ are 
466: shown in Table 4 and Figure 5. Using a similar method \citet{kur07} 
467: obtained a lower ratio of $2.7\pm0.8$ from their quasar sample.
468: The discrepancy is partly due to the fact that they corrected for Balmer 
469: continuum in their spectral fits. For 
470: comparison, we also show the \feii/\mgii\ ratios of $z\sim6$ quasars from 
471: previous measurements \citep{bar03,mai03,fre03,iwa04} as open circles in 
472: Figure 5. \citet{iwa02} measured the
473: \feii/\mgii\ ratios from the spectra of quasars at $0<z<5.3$, and computed
474: the median values of the ratios for a range of redshifts. These median
475: ratios are shown as the filled squares in Figure 5. To study the evolution
476: of \feii/\mgii, \citet{die03b} made quasar composite spectra from $z=0$ to
477: 5. The \feii/\mgii\ ratios measured from the composite spectra are shown
478: as the filled triangles in Figure 5. We note that the \feii\ flux in some
479: studies was integrated over the wavelength range of $2150<\lambda_0<3300$ \AA\
480: \citep[e.g.][]{fre03,iwa04}, slightly different from the range that we used. 
481: We also note that some of these studies used different \feii\ templates to 
482: measure \feii\ fluxes. These issues usually affect the \feii/\mgii\ 
483: measurements by less than a factor of two and cause a relatively large scatter 
484: in the measurements at $z\sim6$. Figure 5 shows that our results are 
485: consistent with those derived from both low-redshift samples and other 
486: $z\sim6$ samples within errors. Because the details of how the \feii\ bump is 
487: formed are not well understood \citep[e.g.][]{bal04}, so we do not derive 
488: actual abundances from the \feii/\mgii\ ratios.
489: 
490: We have shown that the metallicity in the BLRs of high-redshift quasars is
491: supersolar, and the lack of strong evolution in metallicity continues to 
492: $z\sim6$. The high 
493: metallicity at $z\sim6$ indicates that vigorous star formation and element 
494: enrichment have occurred in quasar host galaxies in the first Gyr of 
495: cosmic time. Millimeter and submillimeter observations revealed that
496: luminous $z\sim6$ quasars are extremely far-infrared (FIR) luminous 
497: ($\sim10^{13}$ L$_{\sun}$) and have a large amount of cool dust 
498: ($10^8$--$10^9$
499: M$_{\sun}$) \citep{car01,ber03a,pri03,rob04}. If the FIR emission
500: is mainly powered by star formation in host galaxies, the star formation rates
501: are estimated to be $\sim1000$ M$_{\sun}$ yr$^{-1}$. Strong starbursts can be 
502: induced by merging gas-rich galaxies \citep[e.g.][]{hop06,li07}.
503: CO observations have already revealed the presence of 
504: $\sim2\times 10^{10}$ M$_{\sun}$ of molecular gas in the highest redshift 
505: quasar SDSS J1148+5251 \citep{wal03,ber03b,wal04}. This amount of gas can 
506: sustain the star formation rate of $\sim3000$ M$_{\sun}$ yr$^{-1}$, inferred
507: from the FIR luminosity, for $10^7$--$10^8$ years. If the typical duty cycle 
508: time of luminous quasars is $10^7$--$10^8$ years 
509: \citep[e.g.][]{kau00,wyi03,she07}, it would have enough time to form a large 
510: amount of first generation elements from massive stars.
511: 
512: We have also shown that the \feii/\mgii\ ratios of the $z\sim6$ quasars
513: are consistent with low-redshift measurements. Most of the Fe in the solar
514: neighborhood is produced in intermediate-mass stars, and their enrichment is 
515: delayed by $\sim1$ Gyr compared to Mg.
516: Mg$^+$ and Fe$^+$ have similar ionization potentials and thus the \feii/\mgii\
517: ratio reflects the Fe/Mg abundance \citep{die03b}. Therefore, if Fe was 
518: generated in the same way as it is in our neighbors, the \feii/\mgii\ ratio is 
519: expected to be a strong function of cosmic time at high redshift.
520: However, we do not see any evolution in this ratio even at $z\sim6$. 
521: It has been shown that the timescale of Fe enrichment from SNe Ia strongly
522: depends on star formation histories of host galaxies, and can be much shorter
523: in high-redshift quasar environments \citep[e.g.][]{fri98,mat01}. 
524: \citet{mat01} found that the timescale for the maximum SN Ia rate (Fe
525: enrichment) can be as short as $\sim0.3$ Gyr in an elliptical galaxy with a 
526: high star formation rate but a short star formation history. On the other 
527: hand, \citet{ven04} pointed out that stars with a present-day initial
528: mass function are sufficient to produce the observed \feii/\mgii\ ratios in
529: $z\sim6$ quasars, and SNe Ia are not necessarily the main contributor.
530: Fe could also be generated in Population III stars, which are suggested to 
531: be first generation stars with masses between 100 and 1000 M$_{\sun}$. These 
532: very massive stars produce a large amount of 
533: Fe within a few Myr due to their short lifetimes \citep{heg02}. 
534: 
535: \subsection{Central BH Masses}
536: 
537: The central BH masses of high-redshift quasars are important in understanding
538: the growth of BHs and quasar accretion rates in the early universe. For 
539: high-redshift quasars where direct mass measurements are difficult, BH masses 
540: can be estimated using mass scaling relations based on broad emission line 
541: widths and continuum luminosities. Strong emission lines, including \hb, \civ, 
542: and \mgii, have been used to determine BH masses 
543: \citep[e.g.][]{wan99,die04,mcl04,ves06}. Based on empirical scaling relations, 
544: a few luminous $z\sim6$ quasars have been found to harbor central BHs with 
545: masses of $10^9\sim10^{10}$ M$_{\sun}$ \citep[e.g.][]{bar03,ves04}.
546: 
547: We estimate the central BH masses for the six quasars using the following BH
548: mass scaling relations,
549: \begin{equation}
550:   \rm M_{BH}(\mbox{\civ}) = 4.57 \left( \frac{FWHM(\mbox{\civ})}{km\ s^{-1}}
551:   \right)^2 \left(\frac{\lambda L_{\lambda}(1350\mbox{\AA})}
552:   {10^{44}\ ergs\ s^{-1}}\right)^{0.53} M_{\sun}
553: \end{equation}
554: and
555: \begin{equation}
556:   \rm M_{BH}(\mbox{\mgii}) = 3.2 \left( \frac{FWHM(\mbox{\mgii})}{km\ s^{-1}}
557:   \right)^2 \left(\frac{\lambda L_{\lambda}(3000\mbox{\AA})}
558:   {10^{44}\ ergs\ s^{-1}}\right)^{0.62} M_{\sun}
559:   \label{eq:mgii}
560: \end{equation}
561: derived by \citet{ves06} and \citet{mcl04}, respectively.
562: $\rm L_{\lambda}(1350\mbox{\AA})$ and $\rm L_{\lambda}(3000\mbox{\AA})$ are
563: luminosities at rest-frame 1350 and 3000 \AA. The results are listed in 
564: Columns 4 and 5 of Table 5. They are in agreement with the masses derived by
565: \citet{kur07}. For SDSS J1030+0524 and SDSS J1306+0356, the \civ--based BH 
566: masses are greater than the \mgii--based BH masses by a factor of $\sim$3, but 
567: still within the intrinsic scatter (a factor of 3--4) in these scaling 
568: relations. The unusual blue continua in the two quasars could cause this 
569: discrepancy. The \mgii--based BH masses are also estimated using the latest
570: scaling relation based on FWHM(\mgii) and $\rm L_{\lambda}(1350\mbox{\AA})$
571: by Vestergaard et al. (in preparation). 
572: The results are shown in Column 6, and 
573: they are consistent with the \civ--based BH masses.  Similar to other luminous 
574: quasars at $z\sim6$, our quasars have central BH masses of 
575: $10^9$--$10^{10}$ M$_{\sun}$ \citep[e.g.][]{bar03,ves04,jia06}. 
576: 
577: We calculate the Eddington luminosity ratios of the quasars using their
578: bolometric luminosities and BH masses. We use the \mgii--based BH mass 
579: for SDSS J1623+3112 and the \civ--based BH masses for the others. The 
580: bolometric luminosities were calculated from multiband observations of the
581: quasars by \citet{jia06}. The results
582: are given in Column 3 of Table 5. These luminous $z\sim6$ quasars have 
583: Eddington ratios of order unity, comparable to quasars with similar 
584: luminosities at lower redshift \citep[e.g.][]{mcl04,ves04,kol06}.
585: 
586: It is remarkable that billion-solar-mass BHs can form less than one Gyr after 
587: the Big Bang. With the reasonable assumption that BH accretion is at the 
588: Eddington rate, the BH mass at time $t$ is $M_t=M_0e^{t/\epsilon\tau}$,
589: where $\epsilon$ is the radiative efficiency, $\tau=4.5\times10^8$ years, and 
590: $M_0$ is the initial BH mass or the seed BH mass. Seed BHs can be produced 
591: from the collapse of Population III stars or gas clouds, and their masses are
592: roughly $10^2$--$10^4$ M$_{\sun}$ \citep[e.g.][]{mad01,vol06,lod07}. Consider
593: the case in which a massive BH at $z=6$ formed from a seed BH with $M_0=10^3$ 
594: M$_{\sun}$ at $z=20$. The $e$-folding time $\epsilon\tau$ for the BH growth is 
595: roughly $4.5\times10^7$ years if $\epsilon\sim0.1$. The BH grows from $z=20$ 
596: to 6 by 15 $e$-foldings, or a factor of $3.3\times10^6$, which results in a 
597: massive BH with $M_t=3.3\times10^9$ M$_{\sun}$ at $z=6$, comparable to the 
598: observed BH masses in our sample. If a quasar is shining at half of the 
599: Eddington limit, its BH grows from $z=20$ to 6 by only 7.5 $e$-foldings, or 
600: a factor of $\sim2000$, making it very difficult to form billion-solar-mass 
601: BHs in this scenario. In addition, if Eddington-limited accretion is via 
602: standard thin disks, BHs are likely to be spun up and the radiative efficiency 
603: and 
604: Eddington timescale will increase \citep{vol06,ree06}. In this case it would
605: take much longer to form massive BHs. So super-Eddington accretion or lower 
606: radiative efficiency is probably required to form BHs with 
607: $M_t=10^9$--$10^{10}$ M$_{\sun}$ by $z=6$.
608: 
609: \subsection{Notes on Individual Objects}
610: 
611: {\bf SDSS J0836+0054 ($z=5.810$).}
612: SDSS J0836+0054 was discovered by \citet{fan01}. Its redshift estimated from 
613: the \lya\ emission line is 5.82,
614: comparable to the redshift $5.810\pm0.003$ determined from the \mgii\ emission
615: line \citep{kur07}. Due to high humidity, we did not detect the \mgii\ 
616: emission line. \citet{ste03} found a very red continuum slope of
617: $\alpha_{\nu}=-1.55$ in their NIR spectrum of this quasar. However, using the
618: GNIRS spectra with a longer wavelength coverage and a higher spectral quality
619: we obtained a slope of $\alpha_{\nu}=-0.62$, which is close to
620: $\alpha_{\nu}=-0.44$, the slope of the SDSS composite spectrum \citep{van01}. 
621: SDSS J0836+0054 is the most luminous
622: quasar known at $z>5.7$. The central BH mass estimated from the \civ\ emission
623: line is $9.5\times10^9$ M$_{\sun}$. It was also detected by the
624: Faint Images of the Radio Sky at Twenty-cm \citep[FIRST;][]{bec95}.
625: The flux at 1.4 GHz measured by FIRST is 1.11$\pm$0.15 mJy and measured by
626: \citet{pet03} is 1.75$\pm$0.04 mJy.
627: 
628: {\bf SDSS J1030+0524 ($z=6.309$).}
629: SDSS J1030+0524 is the second highest redshift quasar known to date. The
630: redshift estimated from \lya\ is 6.28 \citep{fan01}. We obtained a more
631: accurate and slightly higher redshift $6.309\pm0.009$ from \mgii, consistent
632: with the value 6.311 of \citet{iwa04}. The emission-line ratios and
633: metallicity in this object have been studied by a few groups. In the discovery
634: paper \citet{fan01} estimated lower limits for the \nv/\civ\ and \nv/\heii\
635: ratios: \nv/\civ\ $\ge0.4$ and \nv/\heii\ $\ge3.0$. Using VLT NIR observations
636: \citet{pen02} found \nv/\civ\ $=0.35$ and \nv/\heii\ $>4.3$. From our
637: measurements we obtained \nv/\civ\ $=0.21\pm0.02$ and \nv/\heii\
638: $=2.85\pm1.70$, comparable to the previous studies. The \feii/\mgii\ ratios
639: of this object measured in previous papers are quite different. The values
640: reported by \citet{mai03}, \citet{fre03}, and \citet{iwa04} are $8.65\pm2.47$,
641: $2.1\pm1.1$, and $0.99_{-0.99}^{+1.86}$, respectively. We obtained
642: \feii/\mgii\ $=5.46\pm0.90$, consistent with the result of \citet{mai03}. Note 
643: that the \feii\ flux in \citet{fre03} and \citet{iwa04} was integrated over 
644: the wavelength range $2150<\lambda_0<3300$ \AA, slightly different from the 
645: range $2200<\lambda_0<3090$ \AA\ that \citet{mai03} and we used. In addition,
646: our NIR spectra have much higher SNRs than those of previous studies.
647: 
648: {\bf SDSS J1044$-$0125 ($z=5.778$).}
649: SDSS J1044$-$0125 was the first quasar discovered at $z>5.7$ \citep{fan00}.
650: The redshift estimated from \lya\ is 5.80 \citep{fan00} and that estimated 
651: from \civ\ is 5.74 \citep{goo01}. However, these measurements could be 
652: severely affected by absorption since this is a BAL quasar 
653: \citep{goo01,djo01}. Due to high humidity we 
654: did not detect its \mgii\ emission line. The redshift measured from \ciii\ 
655: (Figure 1 shows that this line is not affected by absorption) is 
656: $5.778\pm0.005$, consistent with the value 5.78 obtained by \citet{fre03}. 
657: \citet{fre03} also reported a \feii/\mgii\ ratio of $5.0\pm2.1$ for this 
658: quasar. SDSS J1044$-$0125 has the highest BH mass of the objects in our 
659: sample. Submillimeter observations \citep{pri03} and $Spitzer$
660: observations \citep{jia06} revealed a large amount of cool and hot dust
661: in this object. SDSS J1044$-$0125 is the only BAL quasar in our sample.
662: The fraction of BAL quasars in the sample is 16.7\%, similar to the 
663: low-redshift fraction \citep{tru06}.
664: 
665: {\bf SDSS J1306+0356 ($z=6.016$).}
666: The discovery paper for SDSS J1306+0356 indicated a redshift $z=5.99$ 
667: based on \lya\ \citep{fan01}.
668: We obtained a more accurate and slightly higher redshift $6.016\pm0.005$.
669: \citet{pen02} reported a \nv/\civ\ ratio of 0.67 for this quasar. The ratio
670: we measured is $0.55\pm0.05$. We also obtained a \feii/\mgii\ ratio of 
671: $5.77\pm0.72$, slightly smaller than the value $9.03\pm2.26$ derived by 
672: \citet{mai03} based on low SNR spectra. We detected four strong \mgiid\ 
673: absorption systems in the spectra of this quasar, including the two highest 
674: redshift \mgii\ absorbers known (see $\S$ 5).
675: 
676: {\bf SDSS J1411+1217 ($z=5.927$).}
677: SDSS J1411+1217 has the narrowest emission lines in our sample. Its \lya\
678: and \nv\ emission lines are well separated. Therefore the redshift $z=5.93$
679: measured from \nv\ and \oi\ in the discovery paper \citep{fan04} is accurate,
680: and is consistent with the redshift $5.927\pm0.004$ measured from \mgii\
681: in the Gemini spectrum. SDSS J1411+1217 has the lowest BH mass in the sample.
682: It also has a very weak $Spitzer$ 24$\mu$m ($\lambda_0 \sim 3.5$ $\mu$m) flux,
683: and its SED does not show any hot dust emission \citep{jia06}. Another quasar
684: SDSS J0005$-$0006 (not in this sample) shows similar IR properties and does 
685: not have hot dust emission either. This may reflect dust evolution at high 
686: redshift, as such dustless quasars are unknown at lower redshift.
687: 
688: {\bf SDSS J1623+3112 ($z=6.247$).}
689: The redshift of SDSS J1623+3112 measured from \lya\ in the discovery paper
690: is 6.22. We obtained a higher redshift of $6.247\pm0.005$ from \mgii.
691: This makes it the third highest redshift quasar known to date.
692: 
693: \section{\mgii\ ABSORPTION SYSTEMS AT $z<6$}
694: 
695: We search for \mgiid\ intervening systems in the spectra of five $z\sim6$ 
696: quasars. SDSS J1623+3112 is excluded in this analysis because it has a short 
697: wavelength coverage. The SNRs of the spectra for different quasars are 
698: different, as are the dispersions from one order to another. The SNRs at 
699: $\lambda\le 1.2$ $\mu$m are usually higher than those at longer wavelengths, 
700: especially in the range of $1.4<\lambda<1.8$ $\mu$m (roughly $H$ band). 
701: However, a \mgii\ $\lambda2796$ absorption line with rest equivalent width
702: $W_0^{\lambda2796}$ greater than 1.0 \AA\ at $z<3$ or a \mgii\ $\lambda2796$ 
703: absorption line with $W_0^{\lambda2796}$ greater than 1.5 \AA\ at $z>3$ can be 
704: detected at a $\sim7\sigma$ significance level in the lowest SNR parts of the 
705: spectra shown in Figure 1. 
706: Moreover, previous studies have used similar limits \citep[e.g.][]{nes05}. Our 
707: sample is small, and does not provide good statistics by itself, so the 
708: comparison with other samples is important. Thus to achieve unbiased 
709: statistics, we search for \mgii\ absorbers with $W_0^{\lambda2796}>1.0$ \AA\ 
710: at $z<3$ and \mgii\ absorbers with $W_0^{\lambda2796}>1.5$ \AA\ at $z>3$.
711: 
712: We search for pairs of absorption lines with separation matching the \mgiid\
713: doublet. Then we search for absorption lines of other species associated 
714: with the identified \mgii\ absorbers by matching various strong quasar 
715: absorption lines known in the literature to the spectra. Five \mgii\ 
716: absorption systems are identified, with one in the spectrum of SDSS J0836+0054 
717: and the other four in the spectra of SDSS J1306+0356.
718: We normalize the absorption lines using the local continua measured
719: by fitting low-order spline curves to the spectra. Figure 6 shows the 
720: normalized spectra of the five absorption systems, with the identified lines 
721: marked. These lines are usually strong and commonly detected 
722: in quasar absorption systems, such as \civd\ (marked as \civ-D in the figure),
723: \feiiaa, \aliia, \aliiid\ (marked as \aliii-D), \feiiab, \mgiid\
724: (marked as \mgii-D), and \mgia. The four absorption systems found in the
725: spectra of SDSS J1306+0356 are denoted as (a), (b), (c), and (d).
726: 
727: We measure the redshifts ($z_{abs}$) and rest-frame equivalent widths ($W_0$) 
728: of the absorption lines by fitting a Gaussian to each line. Unresolved or 
729: very weak absorption lines are not fitted. The results are given in Table 6. 
730: The \mgii\ doublet ratios are close to one, indicating that these strong 
731: \mgii\ absorbers are saturated. The \mgii\ absorber identified in the spectra 
732: of SDSS J0836+0054 is at $z=3.742$. A pair of moderate strong \aliiid\ lines 
733: in the same absorber are also detected. \civd, which is usually associated 
734: with \mgii, is beyond the spectral coverage. We do not detect any strong 
735: \feii\ absorption in this system. There are four \mgii\ absorption systems 
736: identified in the spectra of SDSS J1306+0356. Systems (a) and (b) are at 
737: relatively low redshifts. System (a) at $z=2.202$ has a strong \mgia\ 
738: absorption line. System (b) at $z=2.532$ has a strong \feiiab\ absorption 
739: line. \civd\ in both systems is beyond the spectral coverage. We discovered
740: two strong \mgiid\ absorbers at $z=4.8668$ and 
741: 4.8823. They are the most distant \mgii\ absorbers known to date. The \mgii\ 
742: $\lambda2803$ line in system (c) and the \mgii\ $\lambda2796$ line in system 
743: (d) almost exactly overlap with each other, so we do not calculate $W_0$ for 
744: the two lines. Note that this is not the so-called line-locking effect, which 
745: mainly occurs in BAL quasars or other associated absorption systems with 
746: $z_{abs}\sim z_{em}$ \citep[e.g.][]{fol87,vil01,fec04}, since the two 
747: absorbers are far from SDSS J1306+0356. The proper distance between the two 
748: absorbers is about 1.4 Mpc, so they are not likely associated with the same 
749: galaxy, but could still belong to the same large-scale structure. Both 
750: absorption systems have strong \aliia\ and \feiiab\ absorption and weak 
751: \feiiaa\ and \aliiid\ absorptions. The \civd\ absorption is 
752: strong in system (c) but very weak in system (d). It has been found that 
753: \civd\ is almost always detected in strong \mgii\ absorption systems at low 
754: redshift \citep[e.g.][]{ste92}. The weakness of \civ\ in system (d) indicates 
755: that it is a rare, \civ-deficient system \citep{chu00}. \citet{els96} 
756: discovered a strong \mgii\ absorber at $z=4.38$ in the spectrum of a $z=4.68$
757: quasar. They found that the \civ\ absorption in this system is also very weak. 
758: Thus among the three strong \mgii\ absorbers at $z>4$, two have very weak 
759: associated \civ, probably indicating that $z>4$ absorption systems are in a
760: lower ionization state than their low-redshift counterparts \citep{els96}.
761: 
762: We calculate the comoving line-of-sight number densities ($dN/dz$) of the 
763: \mgii\ absorbers. The sample of the five \mgii\ absorbers is divided into two 
764: subsamples, one with $W_0^{\lambda2796}>1$ \AA\ at $z<3$ and the other one 
765: with $W_0^{\lambda2796}>1.5$ \AA\ at $z>3$. The redshift path covered by a 
766: spectrum is,
767: \begin{equation}
768: \Delta Z(W_0^{\lambda2796}) = \int_{\Delta z} g(W_0^{\lambda2796},z) dz,
769: \end{equation}
770: where $g(W_0^{\lambda2796},z)=1$ if a \mgii\ absorber at redshift $z$ 
771: and of the given EW would be detected in the spectrum in the redshift range 
772: $\Delta z$ and 
773: $g(W_0^{\lambda2796},z)=0$ otherwise. The total redshift paths covered by
774: the two subsamples are 3.4 and 10.4, respectively. Then the number density 
775: along the redshift path for each subsample is estimated by,
776: \begin{equation}
777: dN/dz = \frac{N_a} {\sum\ \Delta Z(W_0^{\lambda2796})_i},
778: \end{equation}
779: where $N_a$ is the number of \mgii\ absorbers and the sum is over all spectra 
780: in the subsample. We show our results in Figure 7. The filled
781: diamond represents the density for $W_0^{\lambda2796}>1.0$ \AA\ at $z<3$ and 
782: the filled circle represents the density for $W_0^{\lambda2796}>1.5$ \AA\ 
783: at $z>3$. The two data points are positioned at the median redshifts $z=2.6$ 
784: and 4.3, respectively. The statistical uncertainties are Poisson.
785: 
786: The number density of \mgii\ absorbers at low redshift has been well studied. 
787: \citet{ste92} collected the spectra of 103 quasars and found that the density 
788: of \mgii\ absorbers with $W_0^{\lambda2796}>0.3$ \AA\ does not shows cosmic
789: evolution, but the density of strongest \mgii\ absorbers significantly 
790: increases with redshift. \citet{nes05} reached similar results by studying 
791: \mgii\ absorptions in the spectra of thousands of quasars from the SDSS Early 
792: Data Release \citep{sto02}. They divided the sample into subsamples with 
793: $W_0^{\lambda2796}>$ 0.3, 0.6, 1.0, 1.5, 2.0, 2.5, and 3.0 \AA, respectively, 
794: and concluded that $dN/dz$ of the $W_0^{\lambda2796}>$ 0.3, 0.6, 1.0, and 
795: 1.5 \AA\ subsamples are consistent with no evolution within errors, while the 
796: densities of the other three subsamples show strong evolution with redshift. 
797: For comparison, we show in Figure 7 the $W_0^{\lambda2796}>$ 1.0 and 1.5 \AA\ 
798: samples (open diamonds and circles) of \citet{nes05}. \citet{pro06} studied 
799: strong \mgii\ absorption systems using an even larger quasar sample from the 
800: SDSS Data Release Three \citep{aba04}. They found that there is no evolution
801: in the densities of $W_0^{\lambda2796}>$ 1.0 \AA\ absorbers at $0.8<z<2.0$. 
802: The crosses in Figure 7 show the densities 
803: for their $W_0^{\lambda2796}>1.0$ \AA\ sample 
804: (they did not use a sample of $W_0^{\lambda2796}>1.5$ \AA). In Figure 7 we 
805: also show the non-evolution curves (dashed lines) for the cosmology of 
806: $(\Omega_m, \Omega_{\Lambda}, h)=(0.3, 0.7, 0.7)$ with zero curvature,
807: \begin{equation}
808: dN/dz = N_0 (1+z)^2 [\Omega_m(1+z)^3 + \Omega_{\Lambda}]^{-1/2},
809: \end{equation}
810: where $N_0$ is a normalization constant. The curves have been scaled to 
811: minimize the $\chi^2$ to the data of \citet{nes05}. The density of \mgii\ 
812: absorbers with $W_0^{\lambda2796}>1.5$ \AA\ that we have measured is 
813: consistent with no cosmic evolution up to $z>4$. However, due to the large 
814: uncertainty, we cannot rule out the density evolution of \mgii\ absorbers.
815: 
816: \section{SUMMARY}
817: 
818: We have obtained high-quality NIR spectra of six luminous $z\sim6$ SDSS 
819: quasars from our Gemini observations. Five of them were observed using 
820: Gemini-South/GNIRS in cross dispersion mode, which provides a simultaneous 
821: coverage of 0.9--2.5 $\mu$m. The sixth was observed in $K$ band with 
822: Gemini-North/NIRI. These spectra have higher SNRs and better wavelength 
823: coverage than those in previous studies. We use the NIR spectra 
824: combined with optical data to study chemical abundances and BH masses in 
825: the $z\sim6$ quasars and strong intergalactic \mgii\ absorption at $2.2<z<6$. 
826: 
827: The spectra are fitted using a combination of a power-law continuum, an \feii\ 
828: emission template, and a series of emission lines. We calculate the fluxes for 
829: detected emission lines based on the Gaussian model fitting results. We find 
830: that the 
831: line flux ratios (normalized to the \civ\ fluxes) that we have investigated 
832: do not show strong evolution with redshift. We calculate gas metallicity
833: from emission-line ratios using the results of photoionization models given by 
834: previous studies and find that the metallicity in the BLRs of these 
835: high-redshift quasars is supersolar with a typical value of $\sim4$ 
836: Z$_{\sun}$. The comparison with low-redshift observations shows no show strong
837: evolution in metallicity up to $z\sim6$. 
838: The \feii/\mgii\ ratio is also measured for each quasar. We find a typical 
839: \feii/\mgii\ ratio of $4.9\pm1.4$, which is consistent with 
840: low-redshift samples. All these measurement indicate the existence of vigorous 
841: star formation and element enrichment in host galaxies in the first Gyr after 
842: the Big Bang. We have estimated central BH masses from the \civ\ and \mgii\ 
843: emission lines using empirical mass scaling relations. As found in other 
844: luminous $z\sim6$ quasars, these quasars have BH masses of $10^9$--$10^{10}$ 
845: M$_{\sun}$ and Eddington luminosity ratios of order unity. 
846: 
847: We have searched for strong \mgiid\ intervening systems in the spectra of five
848: quasars. Two \mgii\ absorbers with rest equivalent width 
849: $W_0^{\lambda2796}>1$ \AA\ at $2.2<z<3$
850: and three absorbers with $W_0^{\lambda2796}>1.5$ \AA\ at $z>3$ are identified
851: in the spectra of two quasars. The two most distant absorbers are at 
852: $z=4.8668$ and 4.8823, respectively. These are the highest redshift \mgii\ 
853: absorbers known to date. We calculate the comoving line-of-sight
854: number densities for the five identified \mgii\ absorbers. By comparing with
855: low-redshift studies we find that the densities ($dN/dz$) of \mgii\ absorbers
856: with $W_0^{\lambda2796}>1.5$ \AA\ are consistent with no cosmic evolution up 
857: to $z>4$. We note that our sample is small. A larger sample is needed to 
858: provide a good constraint on the density evolution of \mgii\ absorbers.
859: 
860: \acknowledgments
861: 
862: We thank the Gemini staff for their expert help in preparing and carrying out 
863: the observations. We thank T. Nagao and D. B. Nestor for providing useful 
864: data. We acknowledge support from NSF grant AST-0307384, a Sloan Research 
865: Fellowship and a Packard Fellowship for Science and Engineering 
866: (LJ, XF, MV, BCK). We acknowledge support from DFG grant SFB 439 (JDK).
867: 
868: Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan
869: Foundation, the Participating Institutions, the National Science Foundation,
870: the U.S. Department of Energy, the National Aeronautics and Space
871: Administration, the Japanese Monbukagakusho, the Max Planck Society, and the
872: Higher Education Funding Council for England. The SDSS is managed by the 
873: Astrophysical Research Consortium for the Participating Institutions.
874: The SDSS Web Site is http://www.sdss.org/.
875: 
876: \begin{thebibliography}{}
877: \bibitem[Abazajian et al.(2004)]{aba04} Abazajian, K., et al.\ 2004, 
878:   \aj, 128, 502 
879: \bibitem[Baldwin et al.(1995)]{bal95} Baldwin, J., Ferland, G., Korista, K., 
880:   \& Verner, D.\ 1995, \apjl, 455, L119
881: \bibitem[Baldwin et al.(2003)]{bal03} Baldwin, J.~A., Hamann, F., Korista, 
882:   K.~T., Ferland, G.~J., Dietrich, M., \& Warner, C.\ 2003, \apj, 583, 649
883: \bibitem[Baldwin et al.(2004)]{bal04} Baldwin, J.~A., Ferland, G.~J., 
884:   Korista, K.~T., Hamann, F., \& LaCluyz{\'e}, A.\ 2004, \apj, 615, 610 
885: \bibitem[Barth et al.(2003)]{bar03} Barth, A.~J., Martini, P., Nelson, C.~H.,
886:   \& Ho, L.~C. 2003, \apj, 594, L95
887: \bibitem[Becker et al.(1995)]{bec95} Becker, R.~H., White, R.~L.,
888:     \& Helfand, D.~J. 1995, \apj, 450, 559
889: \bibitem[Bechtold et al.(2003)]{bec03} Bechtold, J., et al. 2003, \apj,
890:   588, 119
891: \bibitem[Bertoldi et al.(2003a)]{ber03a} Bertoldi, F., et al. 2003, \aap, 406,
892:   L55
893: \bibitem[Bertoldi et al.(2003b)]{ber03b} Bertoldi, F., et al. 2003, \aap, 409,
894:   L47
895: \bibitem[Carilli et al.(2001)]{car01} Carilli, C. L., Bertoldi, F., Omont, A.,
896:   Cox, P., McMahon, R. G., \& Isaak, K. G. 2001, \aj, 122, 1679
897: \bibitem[Charlton \& Churchill(1998)]{cha98} Charlton, J.~C., \& 
898:   Churchill, C.~W.\ 1998, \apj, 499, 181
899: \bibitem[Churchill et al.(2005)]{chu05} Churchill, C. W., Kacprzak, G. G.,
900:   \& Steidel, C. C. 2005, Proceedings IAU Colloquium No. 199, Probing
901:   Galaxies through Quasar Absorption Lines, P. R. Williams, C. Shu, and
902:   B. M\'enard, eds., astro-ph/0504392
903: \bibitem[Churchill et al.(2000)]{chu00} Churchill, C.~W., Mellon, R.~R., 
904:   Charlton, J.~C., Jannuzi, B.~T., Kirhakos, S., Steidel, C.~C., \& 
905:   Schneider, D.~P.\ 2000, \apj, 543, 577
906: \bibitem[Cool et al.(2006)]{coo06} Cool, R.~J., et al.\ 2006, \aj, 132, 823
907: \bibitem[Croton et al.(2006)]{cro06} Croton, D. J., et al. 2006, \mnras,
908:     365, 11
909: \bibitem[Dietrich et al.(2002)]{die02} Dietrich, M., Appenzeller, I., 
910:   Vestergaard, M., \& Wagner, S. J. 2002, \apj, 564, 581
911: \bibitem[Dietrich et al.(2003a)]{die03a} Dietrich, M., Appenzeller, I., 
912:   Hamann, F., Heidt, J., J{\"a}ger, K., Vestergaard, M., 
913:   \& Wagner, S.~J.\ 2003, \aap, 398, 891
914: \bibitem[Dietrich et al.(2003b)]{die03b} Dietrich, M., Hamann, F., Appenzeller,
915:   I., \& Vestergaard, M. 2003, \apj, 596, 817
916: \bibitem[Dietrich \& Hamann(2004)]{die04} Dietrich, M., \& Hamann, F. 2004,
917:   \apj, 611, 761
918: \bibitem[Djorgovski et al.(2001)]{djo01} Djorgovski, S. G., Castro, S.,
919:   Stern, D., \& Mahabal, A. A. 2001, \apj, 560, L5
920: \bibitem[Elston et al.(1996)]{els96} Elston, R., Bechtold, 
921:   J., Hill, G.~J., \& Ge, J.\ 1996, \apj, 456, L13 
922: \bibitem[Elvis et al.(1994)]{elv94} Elvis, M., et al. 1994, \apjs, 95, 1
923: \bibitem[Fan et al.(2000)]{fan00} Fan, X., et al. 2000, \aj, 120, 1167
924: \bibitem[Fan et al.(2001)]{fan01} Fan, X., et al. 2001, \aj, 122, 2833
925: \bibitem[Fan et al.(2003)]{fan03} Fan, X., et al. 2003, \aj, 125, 1649
926: \bibitem[Fan et al.(2004)]{fan04} Fan, X., et al. 2004, \aj, 128, 515
927: \bibitem[Fan et al.(2006)]{fan06} Fan, X., et al. 2006, \aj, 131, 1203
928: \bibitem[Fechner et al.(2004)]{fec04} Fechner, C., Baade, R., 
929:   \& Reimers, D.\ 2004, \aap, 418, 857 
930: \bibitem[Foltz et al.(1987)]{fol87} Foltz, C.~B., Weymann, R.~J., Morris, 
931:   S.~L., \& Turnshek, D.~A.\ 1987, \apj, 317, 450 
932: \bibitem[Freudling et al.(2003)]{fre03} Freudling, W., Corbin,
933:   M. R., \& Korista, K. T. 2003, \apj, 587, L67
934: \bibitem[Friaca \& Terlevich(1998)]{fri98} Friaca, A.~C.~S., 
935:   \& Terlevich, R.~J.\ 1998, \mnras, 298, 399 
936: \bibitem[Goodrich et al.(2001)]{goo01} Goodrich, R. W., et al. 2001, \apj,
937:   561, L23
938: \bibitem[Greggio \& Renzini(1983)]{gre83} Greggio, L., \& Renzini, A. 1983,
939:   \aap, 118, 217
940: \bibitem[Hamann \& Ferland(1992)]{ham92} Hamann, F., \& Ferland, G. 1992,
941:   \apj, 391, L53
942: \bibitem[Hamann \& Ferland(1993)]{ham93} Hamann, F., \& Ferland, G. 1993,
943:   \apj, 418, 11
944: \bibitem[Hamann \& Ferland(1999)]{ham99} Hamann, F., \& Ferland, G. 1999,
945:   \araa, 37, 487
946: \bibitem[Hamann et al.(2002)]{ham02} Hamann, F., Korista, K.~T., Ferland, 
947:   G.~J., Warner, C., \& Baldwin, J.\ 2002, \apj, 564, 592 
948: \bibitem[Hamann et al.(2007)]{ham07} Hamann, F., Warner, C., Dietrich, M.,
949:   \& Ferland, G. 2007, in ASP Conference Proceedings, the Central Engine of 
950:   Active Galactic Nuclei, ed. L. C. Ho and J.-M. Wang (San Francisco: ASP), 
951:   in press (astro-ph/0701503)
952: \bibitem[Heger \& Woosley(2002)]{heg02} Heger, A., \& 
953:   Woosley, S.~E.\ 2002, \apj, 567, 532 
954: \bibitem[Hopkins et al.(2005)]{hop05} Hopkins, P. F., Hernquist, L., 
955:   Cox, T. J., Di Matteo, T., Martini, P., Robertson, B., \& Springel, V.
956:   2005, \apj, 630, 705
957: \bibitem[Hopkins et al.(2006)]{hop06} Hopkins, P.~F., Hernquist, L., Cox, 
958:   T.~J., Di Matteo, T., Robertson, B., \& Springel, V.\ 2006, \apjs, 163, 1 
959: \bibitem[Iwamuro et al.(2002)]{iwa02} Iwamuro, F., Motohara, K., Maihara, 
960:   T., Kimura, M., Yoshii, Y., \& Doi, M.\ 2002, \apj, 565, 63 
961: \bibitem[Iwamuro et al.(2004)]{iwa04} Iwamuro, F., Kimura, M., Eto, S.,
962:   Maihara, T., Motohara, K., Yoshii, Y., \& Doi, M. 2004, \apj, 614, 69
963: \bibitem[Jiang et al.(2006)]{jia06} Jiang, L., et al. 2006, \aj, 132, 2127
964: \bibitem[Kauffmann \& Haehnelt(2000)]{kau00} Kauffmann, G., \& Haehnelt, M.
965:     2000, \mnras, 311, 576
966: \bibitem[Kollmeier et al.(2006)]{kol06} Kollmeier, J.~A., et al.\ 2006,
967:   \apj, 648, 128
968: \bibitem[Kurk et al.(2007)]{kur07} Kurk, J.~D. et al.\ 2007, \apj, submitted
969: \bibitem[Li et al.(2007)]{li07} Li, Y., et al. 2007, \apj, in press
970:   (astro-ph/0608190)
971: \bibitem[Lodato \& Natarajan(2007)]{lod07} Lodato, G., \&   Natarajan, P.\ 
972:   2007, \mnras, 377, L64
973: \bibitem[Madau \& Rees(2001)]{mad01} Madau, P., \& Rees, M.~J.\ 2001, \apjl, 
974:   551, L27
975: \bibitem[Maiolino et al.(2003)]{mai03} Maiolino, R., Juarez, Y., Mujica, R.,
976:   Nagar, N. M., \& Oliva, E. 2003, \apj, 596, L155
977: \bibitem[Matteucci \& Recchi(2001)]{mat01} Matteucci, F., \& 
978:   Recchi, S.\ 2001, \apj, 558, 351 
979: \bibitem[McGreer et al.(2006)]{mcg06} McGreer, I.~D., Becker, R.~H.,
980:   Helfand, D.~J., \& White, R.~L.\ 2006, \apj, 652, 157
981: \bibitem[McIntosh et al.(1999)]{mci99} McIntosh, D. H., Rix, H. W., Rieke,
982:   M. J., \& Foltz, C. B. 1999, \apj, 517, L73
983: \bibitem[McLure \& Dunlop(2004)]{mcl04} McLure, R.~J., \&
984:   Dunlop, J.~S.\ 2004, \mnras, 352, 1390
985: \bibitem[Nagao et al.(2006)]{nag06} Nagao, T., Marconi, A., 
986:   \& Maiolino, R.\ 2006, \aap, 447, 157
987: \bibitem[Nestor et al.(2005)]{nes05} Nestor, D.~B., Turnshek, 
988:   D.~A., \& Rao, S.~M.\ 2005, \apj, 628, 637 
989: \bibitem[Pentericci et al.(2002)]{pen02} Pentericci, L., et al. 2002, \aj,
990:   123, 2151
991: \bibitem[Pentericci et al.(2003)]{pen03} Pentericci, L., et al. 2003, \aap,
992:   410, 75
993: \bibitem[Petric et al.(2003)]{pet03} Petric, A. O., et al. 2003, \aj, 126, 15
994: \bibitem[Priddey et al.(2003)]{pri03} Priddey, R. S., Isaak, K. G., McMahon,
995:   R. G., Robson, E. I., \& Pearson, C. P. 2003, \mnras, 344, L74
996: \bibitem[Prochter et al.(2006)]{pro06} Prochter, G.~E., Prochaska, J.~X., 
997:   \& Burles, S.~M.\ 2006, \apj, 639, 766 
998: \bibitem[Rees \& Volonteri(2006)]{ree06} Rees, M.~J., \& Volonteri, M.\ 2006, 
999:   Proceedings IAU Symposium No. 238, Black Holes: from Stars to Galaxies --
1000:   across the Range of Masses, V. Karas and G. Matt, eds., astro-ph/0701512
1001: \bibitem[Richards et al.(2002)]{ric02} Richards, G. T., et al. 2002,
1002:   \aj, 124, 1
1003: \bibitem[Robson et al.(2004)]{rob04} Robson, I., Priddey, R. S., Isaak, K. G.,
1004:   \& McMahon, R. G. 2004, \mnras, 351, L29
1005: \bibitem[Shemmer et al.(2006)]{she06} Shemmer, O., et al. 2006, \apj,
1006:   644, 86
1007: \bibitem[Shen et al.(2007)]{she07} Shen, Y., et al.\ 2007, \aj, 133, 2222 
1008: \bibitem[Spergel et al.(2007)]{spe07} Spergel, D.~N., et al.\ 2007, \apjs, 
1009:   170, 377 
1010: \bibitem[Steidel et al.(2002)]{ste02} Steidel, C.~C., Kollmeier, J.~A.,
1011:   Shapley, A.~E., Churchill, C.~W., Dickinson, M., \& Pettini, M.\ 2002,
1012:   \apj, 570, 526
1013: \bibitem[Steidel \& Sargent(1992)]{ste92} Steidel, C.~C., \& 
1014:   Sargent, W.~L.~W.\ 1992, \apjs, 80, 1
1015: \bibitem[Stern et al.(2003)]{ste03} Stern, D., et al. 2003, \apj,
1016:   596, L39
1017: \bibitem[Stoughton et al.(2002)]{sto02} Stoughton, C., et al.\ 2002, 
1018:   \aj, 123, 485 
1019: \bibitem[Trump et a.(2006)]{tru06} Trump, J.~R., et al. 2006, \apjs, 165, 1
1020: \bibitem[Vanden Berk et al.(2001)]{van01} Vanden Berk, D. E., et al. 2001,
1021:     \aj, 122, 549
1022: \bibitem[Venkatesan et al.(2004)]{ven04} Venkatesan, A., Schneider, R., \& 
1023:   Ferrara, A.\ 2004, \mnras, 349, L43 
1024: \bibitem[Vestergaard \& Wilkes(2001)]{ves01} Vestergaard, M. \& Wilkes, B. J.
1025:   2001, \apjs, 134, 1
1026: \bibitem[Vestergaard(2004)]{ves04} Vestergaard, M. 2004, \apj, 601, 676
1027: \bibitem[Vestergaard \& Peterson(2006)]{ves06} Vestergaard, M., \&
1028:   Peterson, B.~M.\ 2006, \apj, 641, 689
1029: \bibitem[Vilkoviskij \& Irwin(2001)]{vil01} Vilkoviskij, E.~Y., \& 
1030:   Irwin, M.~J.\ 2001, \mnras, 321, 4 
1031: \bibitem[Volonteri \& Rees(2006)]{vol06} Volonteri, M., \& Rees, M.~J.\ 2006, 
1032:   \apj, 650, 669 
1033: \bibitem[Walter et al.(2003)]{wal03} Walter, F., et al. 2003, Nature, 424, 406
1034: \bibitem[Walter et al.(2004)]{wal04} Walter, F., et al. 2004, \apj, 615, L17
1035: \bibitem[Wandel et al.(1999)]{wan99} Wandel, A., Peterson, B.~M., \& 
1036:   Malkan, M.~A.\ 1999, \apj, 526, 579 
1037: \bibitem[Wang et al.(2007)]{wan07} Wang, R., et al. \aj, in press
1038:   (astro-ph/0704.2053)
1039: \bibitem[Wyithe \& Loeb(2003)]{wyi03} Wyithe, J. S. B., \& Loeb, A. 2003,
1040:     \apj, 595, 614
1041: \bibitem[York et al.(2000)]{yor00} York, D. G., et al. 2000, \aj, 120, 1579
1042: \end{thebibliography}
1043: 
1044: \clearpage
1045: \begin{figure}
1046: \epsscale{0.8}
1047: \plotone{f1.eps}
1048: \caption{Optical and NIR spectra of five $z\sim6$ quasars in our sample. 
1049: The spectra at $\lambda>1.0$ $\mu$m were obtained with Gemini-South/GNIRS and
1050: the spectra at $\lambda<1.0$ $\mu$m are taken from the quasar discovery 
1051: papers. The spectra have been smoothed by three pixels.
1052: The blue solid lines are the best fits to emission lines and the blue 
1053: dashed lines are the best fits to the power-law continuum and \feii\ emission.
1054: The red lines are the sums of all the components. For comparison, we show
1055: the low-redshift composite spectrum of \citet{van01} in gray. 
1056: A spectrum of a standard star HIP 54027 is also shown in the figure with 
1057: a black body (with the effective temperature of HIP 54027) spectrum (in red) 
1058: on top of it. The bottom panel 
1059: shows the sky transparency at an airmass of 1.0 and water vapor of 1.6 mm.}
1060: \end{figure}
1061: 
1062: \clearpage
1063: \begin{figure}
1064: \epsscale{0.6}
1065: \plotone{f2.eps}
1066: \caption{Same as Fig. 1 for SDSS J1623+3112, which was observed in $K$ band 
1067: using Gemini-North/NIRI.}
1068: \end{figure}
1069: 
1070: \clearpage
1071: \begin{figure}
1072: \epsscale{1.0}
1073: \plotone{f3.eps}
1074: \caption{Emission line fluxes (filled circles with errors) of the $z\sim6$
1075: quasars compared to low-redshift measurements. The fluxes have been normalized 
1076: to the \civ\ fluxes. The filled squares represent the flux ratios measured in 
1077: the composite spectra of quasars in the luminosity range $-27.5<M_B<-26.5$ 
1078: from \citet{nag06}.
1079: Most of the flux ratios do not exhibit strong evolution up to $z\sim6$.}
1080: \end{figure}
1081: 
1082: \clearpage
1083: \begin{figure}
1084: \epsscale{0.6}
1085: \plotone{f4.eps}
1086: \caption{Chemical abundances of quasars derived from line 
1087: ratios as a function of redshift. The open circles represent the metallicities
1088: derived from the \nv/\civ\ and \nv/\heii\ ratios using Figure 5 of 
1089: \citet{ham02}. The filled circles represent the metallicities derived from the 
1090: line ratios using Figure 29 of \citet{nag06}. The filled squares show the 
1091: averaged metallicities of quasars for the luminosity range $-28.5<M_B<-25.5$ 
1092: from \citet{nag06}. The filled triangles show the metallicities of eleven 
1093: $3.9<z<5.0$ quasars from \citet{die03a}. The metallicities estimated from each 
1094: line ratio show no strong evolution up to $z>6$ within the errors.}
1095: \end{figure}
1096: 
1097: \clearpage
1098: \begin{figure}
1099: \epsscale{0.6}
1100: \plotone{f5.eps}
1101: \caption{The \feii/\mgii\ ratio as a function of redshift. The filled 
1102: circles are our results. The open circles show the \feii/\mgii\ ratios of
1103: $z\sim6$ quasars from previous measurements \citep{bar03,mai03,fre03,iwa04}. 
1104: The filled squares and triangles represent the low-redshift results from 
1105: \citet{iwa02} and \citet{die03b}. The lack of strong evolution of the 
1106: \feii/\mgii\ abundance continues to $z\sim6$.}
1107: \end{figure}
1108: 
1109: \clearpage
1110: \begin{figure}
1111: \epsscale{0.8}
1112: \plotone{f6.eps}
1113: \caption{Normalized spectra of the five absorption systems in the spectra
1114: of SDSS J0836+0054 and SDSS J1306+0356. The positions of the identified lines
1115: are marked as vertical lines, and the names of the lines are given nearby.
1116: Doublets are expressed as `--D'. The four absorption systems in the spectra
1117: SDSS J1306+0356 are denoted as (a), (b), (c), and (d).}
1118: \end{figure}
1119: 
1120: \clearpage 
1121: \begin{figure}
1122: \epsscale{0.6}
1123: \plotone{f7.eps} 
1124: \caption{Comoving number densities of \mgii\ absorbers. The filled
1125: diamond represents the density for $W_0^{\lambda2796}>1$ \AA\ at $z=2.6$ and
1126: the filled circle represents the density for $W_0^{\lambda2796}>1.5$ \AA\
1127: at $z=4.3$. The open diamonds and circles show the number densities for 
1128: $W_0^{\lambda2796}>1.0$ and 1.5 \AA\ from Figure 9 of \citet{nes05}. The 
1129: crosses show the densities of the $W_0^{\lambda2796}>1.0$ \AA\ absorbers from 
1130: \citet{pro06}. The dashed lines represent non-evolution 
1131: curves for the cosmology of $(\Omega_m, \Omega_{\Lambda}, h)=(0.3, 0.7, 0.7)$. 
1132: The curves have been scaled to minimize the $\chi^2$ to the data of 
1133: \citet{nes05}. The densities are consistent with no cosmic evolution.}
1134: \end{figure} 
1135: 
1136: \clearpage
1137: \begin{deluxetable}{ccccllc}
1138: \rotate
1139: \tablecaption{Log of Observations}
1140: \tablewidth{0pt}
1141: \tablehead{\colhead{Quasar (SDSS)} & \colhead{Redshift} & 
1142:   \colhead{$z_{AB}$ (mag)} & \colhead{$J$ (mag) } & \colhead{Date} & 
1143:   \colhead{Instrument} & \colhead{$t_{exp}$ (min)} }
1144: \startdata
1145: J083643.85+005453.3    & 5.82 & 18.74$\pm$0.05 & 17.89$\pm$0.05 
1146:   & 2006 Feb 19     & GNIRS &  70  \\
1147: J103027.10+052455.0    & 6.28 & 20.05$\pm$0.10 & 18.87$\pm$0.10 
1148:   & 2006 Mar 20, 26 & GNIRS & 140  \\
1149: J104433.04$-$012502.2  & 5.80 & 19.23$\pm$0.07 & 18.31$\pm$0.10
1150:   & 2006 Mar 28     & GNIRS & 120  \\
1151: J130608.26+035626.3    & 5.99 & 19.47$\pm$0.05 & 18.77$\pm$0.10 
1152:   & 2006 Mar 26, 27 & GNIRS & 140  \\
1153: J141111.29+121737.4    & 5.93 & 19.65$\pm$0.08 & 18.95$\pm$0.05
1154:   & 2006 Mar 28     & GNIRS & 120  \\
1155: J162331.81+311200.5    & 6.22 & 20.09$\pm$0.10 & 19.15$\pm$0.10
1156:   & 2004 Aug 16, Sep 14 & NIRI (K) & 180 \\
1157:   &      &       &       & 2005 Aug 22, 24  &         &     \\
1158: \enddata
1159: \tablecomments{Redshifts, $z_{AB}$ (AB magnitudes) and $J$ (Vega-based 
1160:   magnitudes) are from the quasar discovery papers \citet{fan00}, 
1161:   \citet{fan01}, and \citet{fan04}.}
1162: \end{deluxetable}
1163: 
1164: \clearpage
1165: \begin{deluxetable}{lcrc}
1166: \tablecaption{Redshifts and Continuum Slopes}
1167: \tablewidth{0pt}
1168: \tablehead{\colhead{Quasar (SDSS)} & \colhead{Redshift} 
1169:   & \colhead{$\rm Slope_{uv}$($\alpha_{\nu}$)\tablenotemark{a}} 
1170:   & \colhead{$\rm Slope_{opt}$($\alpha_{\nu}$)\tablenotemark{b}}}
1171: \startdata
1172: J0836+0054 & $5.810\pm0.003$\tablenotemark{c}& $-0.62_{-0.06}^{+0.06}$ & $-0.52$\\
1173: J1030+0524 & $6.309\pm0.009$                 & $0.46_{-0.25}^{+0.18}$  & $-0.31$\\
1174: J1044$-$0125&$5.778\pm0.005$\tablenotemark{d}& $-0.27_{-0.10}^{+0.09}$ & $-0.30$\\
1175: J1306+0356 & $6.016\pm0.005$                 & $0.50_{-0.14}^{+0.12}$  & $-0.12$\\
1176: J1411+1217 & $5.927\pm0.004$                 & $-0.21_{-0.19}^{+0.16}$ & $-0.51$\\
1177: J1623+3112 & $6.247\pm0.005$                 & $\ldots$                & $-0.32$\\
1178: \enddata
1179: \tablenotetext{a}{UV continuum slopes measured from the Gemini spectra.}
1180: \tablenotetext{b}{Optical continuum slopes taken from \citet{jia06}.}
1181: \tablenotetext{c}{Taken from \citet{kur07}.}
1182: \tablenotetext{d}{Measured from the \ciii\ $\lambda$1909 emission line.}
1183: \end{deluxetable}
1184: 
1185: \clearpage
1186: \begin{deluxetable}{llllllll}
1187: \tabletypesize{\small} 
1188: \rotate
1189: \tablecaption{Emission Line Properties}
1190: \tablewidth{0pt}
1191: \tablehead{\colhead{} & \colhead{} & 
1192:  \colhead{J0836+0054} & \colhead{J1030+0524} & \colhead{J1044$-$0125} & 
1193:  \colhead{J1306+0356} & \colhead{J1411+1217} & \colhead{J1623+3112}}
1194: \startdata
1195: \nv    & EW   & 15.9$\pm$0.7  & 4.9$\pm$0.1   & 3.9$\pm$0.2
1196:               & 9.3$\pm$0.3   & 28.9$\pm$0.4  & $\ldots$ \\
1197:        & FWHM & 31.5$\pm$1.0  & 8.7$\pm$0.2   & 10.4$\pm$0.3
1198:               & 15.0$\pm$0.2  & 22.3$\pm$0.3  & $\ldots$ \\
1199: \siii  & EW   & 3.6$\pm$0.3   & 3.0$\pm$0.2   & 2.3$\pm$0.1
1200:               & 1.7$\pm$0.1   & 0.9$\pm$0.1   & $\ldots$ \\
1201:        & FWHM & 15.0$\pm$0.6  & 15.8$\pm$0.7  & 12.7$\pm$0.3
1202:               & 12.8$\pm$0.4  & 3.7$\pm$0.3   & $\ldots$ \\
1203: \oi    & EW   & 7.4$\pm$0.1   & 6.7$\pm$0.3   & 2.6$\pm$0.2
1204:               & $\ldots$      & 6.7$\pm$0.2   & $\ldots$ \\
1205:        & FWHM & 24.6$\pm$0.5  & 18.8$\pm$0.7  & 24.3$\pm$1.3
1206:               & $\ldots$      & 8.1$\pm$0.2   & $\ldots$ \\
1207: \cii   & EW   & 5.8$\pm$0.5   & 4.0$\pm$0.4   & $\ldots$
1208:               & $\ldots$      & $\ldots$      & $\ldots$ \\
1209:        & FWHM & 41.9$\pm$3.1  & 17.0$\pm$1.5  & $\ldots$
1210:               & $\ldots$      & $\ldots$      & $\ldots$ \\
1211: \siiv  & EW   & 9.3$\pm$4.2   & 8.5$\pm$1.7   & 8.9$\pm$0.7
1212:               & 9.5$\pm$0.8   & 8.1$\pm$6.4   & $\ldots$ \\
1213:        & FWHM & 25.3$\pm$5.9  & 18.4$\pm$2.9  & 35.7$\pm$1.8
1214:               & 29.0$\pm$2.0  & 21.5$\pm$12.2 & $\ldots$ \\
1215: \civ   & EW   & 15.0$\pm$1.6  & 40.0$\pm$3.1  & 24.7$\pm$2.1
1216:               & 29.6$\pm$2.4  & 76.2$\pm$5.8  & $\ldots$ \\
1217:        & FWHM & 37.1$\pm$3.0  & 27.3$\pm$2.9  & 42.2$\pm$3.1
1218:               & 24.5$\pm$2.3  & 17.5$\pm$1.6  & $\ldots$ \\
1219: \heii  & EW   & $\ldots$      & 3.4$\pm$2.0   & $\ldots$
1220:               & 2.6$\pm$1.8   & 5.1$\pm$2.5   & $\ldots$ \\
1221:        & FWHM & $\ldots$      & 21.7$\pm$8.8  & $\ldots$
1222:               & 15.5$\pm$7.5  & 39.0$\pm$26.8 & $\ldots$ \\
1223: \oiii  & EW   & $\ldots$      & 2.3$\pm$1.3   & $\ldots$
1224:               & 2.7$\pm$2.0   & 2.8$\pm$2.5   & $\ldots$ \\
1225:        & FWHM & $\ldots$      & 22.9$\pm$11.0 & $\ldots$
1226:               & 26.9$\pm$19.4 & 42.2$\pm$34.0 & $\ldots$ \\
1227: \aliii & EW   & 1.8$\pm$0.9   & $\ldots$      & 5.0$\pm$1.6
1228:               & 7.0$\pm$3.1   & 5.1$\pm$2.3   & $\ldots$ \\
1229:        & FWHM & 16.3$\pm$6.8  & $\ldots$      & 31.0$\pm$6.7
1230:               & 34.5$\pm$14.7 & 25.6$\pm$9.2  & $\ldots$ \\
1231: \ciii  & EW   & 18.9$\pm$2.3  & 19.3$\pm$2.8  & 24.0$\pm$1.8
1232:               & 19.1$\pm$3.9  & 26.1$\pm$3.2  & $\ldots$ \\
1233:        & FWHM & 53.2$\pm$4.9  & 31.6$\pm$3.3  & 50.3$\pm$3.2
1234:               & 37.6$\pm$7.4  & 31.3$\pm$2.6  & $\ldots$ \\
1235: \mgii  & EW   & $\ldots$      & 46.1$\pm$3.7  & $\ldots$
1236:               & 24.9$\pm$2.1  & 37.5$\pm$5.5  & 27.4$\pm$3.9 \\
1237:        & FWHM & $\ldots$      & 33.8$\pm$2.2  & $\ldots$
1238:               & 33.2$\pm$2.2  & 23.7$\pm$2.7  & 34.4$\pm$2.4 \\
1239: \enddata
1240: \tablecomments{Rest-frame FWHM and EW are in units of \AA.}
1241: \end{deluxetable}
1242: 
1243: \clearpage
1244: \begin{deluxetable}{lcccccc}
1245: \tabletypesize{\small}
1246: \rotate
1247: \tablecaption{Flux Ratios}
1248: \tablewidth{0pt}
1249: \tablehead{\colhead{} &
1250:  \colhead{J0836+0054} & \colhead{J1030+0524} & \colhead{J1044$-$0125} &
1251:  \colhead{J1306+0356} & \colhead{J1411+1217} & \colhead{J1623+3112}}
1252: \startdata
1253: \nv/\civ    &  1.44$\pm$0.17  &  0.21$\pm$0.02  &  0.23$\pm$0.03  
1254:             &  0.55$\pm$0.05  &  0.56$\pm$0.04  &  $\ldots$ \\
1255: \siii/\civ  &  0.32$\pm$0.04  &  0.12$\pm$0.02  &  0.13$\pm$0.02 
1256:             &  0.10$\pm$0.01  &  0.02$\pm$0.01  &  $\ldots$ \\
1257: \oi/\civ    &  0.62$\pm$0.07  &  0.26$\pm$0.02  &  0.14$\pm$0.02 
1258:             &  $\ldots$       &  0.12$\pm$0.02  &  $\ldots$ \\
1259: \cii/\civ   &  0.47$\pm$0.06  &  0.14$\pm$0.02  &  $\ldots$
1260:             &  $\ldots$       &  $\ldots$       &  $\ldots$ \\
1261: \siiv/\civ  &  0.72$\pm$0.33  &  0.27$\pm$0.06  &  0.43$\pm$0.05
1262:             &  0.41$\pm$0.05  &  0.13$\pm$0.10  &  $\ldots$ \\
1263: \heii/\civ  &  $\ldots$       &  0.07$\pm$0.04  &  $\ldots$
1264:             &  0.08$\pm$0.05  &  0.06$\pm$0.05  &  $\ldots$ \\
1265: \oiii/\civ  &  $\ldots$       &  0.05$\pm$0.03  &  $\ldots$
1266:             &  0.08$\pm$0.06  &  0.03$\pm$0.02  &  $\ldots$ \\
1267: \aliii/\civ &  0.10$\pm$0.05  &  $\ldots$       &  0.15$\pm$0.05
1268:             &  0.15$\pm$0.07  &  0.05$\pm$0.02  &  $\ldots$ \\
1269: \ciii/\civ  &  0.95$\pm$0.15  &  0.29$\pm$0.05  &  0.68$\pm$0.08
1270:             &  0.38$\pm$0.08  &  0.24$\pm$0.03  &  $\ldots$ \\
1271: \feii/\mgii &  $\ldots$       &  5.45$\pm$0.58  &  $\ldots$ 
1272:             &  5.77$\pm$0.72  &  5.27$\pm$0.72  &  3.21$\pm$0.62 \\
1273: \enddata
1274: \end{deluxetable}
1275: 
1276: \clearpage
1277: \begin{deluxetable}{cccccc}
1278: \tablecaption{Central BH Masses ($10^9$ M$_{\sun}$)}
1279: \tablewidth{0pt}
1280: \tablehead{\colhead{Quasar (SDSS)} &\colhead{$\rm L_{Bol}$\tablenotemark{a}} &
1281:   \colhead{$\rm L_{Bol}/L_{Edd}$\tablenotemark{b}} & 
1282:   \colhead{$\rm M_{BH}$(\civ)} & \colhead{$\rm M_{BH}$(\mgii)} & 
1283:   \colhead{$\rm M_{BH}^{'}$(\mgii)\tablenotemark{c}} }
1284: \startdata
1285: J0836+0054   & 47.72 & 0.44 & 9.3$\pm$1.6  & $\ldots$    & $\ldots$    \\
1286: J1030+0524   & 47.37 & 0.50 & 3.6$\pm$0.9  & 1.0$\pm$0.2 & 2.1$\pm$0.4 \\
1287: J1044$-$0125 & 47.63 & 0.31 & 10.5$\pm$1.6 & $\ldots$    & $\ldots$    \\
1288: J1306+0356   & 47.40 & 0.61 & 3.2$\pm$0.6  & 1.1$\pm$0.1 & 2.2$\pm$0.3 \\
1289: J1411+1217   & 47.20 & 0.94 & 1.3$\pm$0.3  & 0.6$\pm$0.1 & 0.9$\pm$0.2 \\
1290: J1623+3112   & 47.33 & 1.11 & $\ldots$     & 1.5$\pm$0.3 & $\ldots$    \\
1291: \enddata
1292: \tablenotetext{a}{Bolometric luminosity in log[erg s$^{-1}$] from 
1293:   \citet{jia06}.}
1294: \tablenotetext{b}{$\rm L_{Edd}$ is derived from $\rm M_{BH}$(\civ) except for 
1295:   SDSS J1623+3112, whose $\rm L_{Edd}$ is derived from $\rm M_{BH}$(\mgii).}
1296: \tablenotetext{c}{$\rm M_{BH}^{'}$ is estimated from the new relation by 
1297:   Vestergaard et al. (in preparation).}
1298: \end{deluxetable}
1299: 
1300: \clearpage
1301: \begin{deluxetable}{lrrcc}
1302: \tablecaption{Absorption Lines}
1303: \tablewidth{0pt}
1304: \tablehead{\colhead{Absorption system} & \colhead{$\lambda_{abs}$ (\AA)} 
1305:   & \colhead{ID} & \colhead{$z_{abs}$} & \colhead{$W_0$ (\AA)}}
1306: \startdata
1307: SDSS J0836+0054    & 13261.81 &  \mgii(2796) &  3.7426 &   2.59\\
1308:                    & 13295.43 &  \mgii(2803) &  3.7424 &   2.08\\
1309:                    & & & & \\
1310: SDSS J1306+0356 (a)&  8952.03 &  \mgii(2796) &  2.2013 &   1.86\\
1311:                    &  8977.30 &  \mgii(2803) &  2.2022 &   1.99\\
1312:                    &  9137.13 &  \mgit(2852) &  2.2028 &   1.29\\
1313:                    & & & & \\
1314: SDSS J1306+0356 (b)&  9877.64 &  \mgii(2796) &  2.5323 &   3.24\\
1315:                    &  9903.27 &  \mgii(2803) &  2.5324 &   3.15\\
1316:                    &  9185.23 &  \feii(2600) &  2.5325 &   1.94\\
1317:                    & & & & \\
1318: SDSS J1306+0356 (c)& 16405.55 &  \mgii(2796) &  4.8668 &   2.92\\
1319:                    & 16447.67 &  \mgii(2803) & $\ldots$&  $\ldots$\\
1320:                    &  9795.84 &  \aliit(1670)&  4.8632 &   1.11\\
1321:                    & 15252.21 &  \feii(2600) &  4.8658 &   1.49\\
1322:                    & & & & \\
1323: SDSS J1306+0356 (d)& 16448.85 &  \mgii(2796) & $\ldots$&  $\ldots$\\
1324:                    & 16491.08 &  \mgii(2803) &  4.8823 &   2.11\\
1325:                    &  9822.93 &  \aliit(1670)&  4.8795 &   0.76\\
1326:                    & 15293.94 &  \feii(2600) &  4.8819 &   1.97\\
1327: \enddata
1328: \end{deluxetable}
1329: 
1330: \end{document}
1331: