STAR FORMATION AND GAS KINEMATICS OF QUASAR HOST GALAXIES AT z ∼ 6: NEW INSIGHTS FROM ALMA

Quasars at redshift 6 and higher provide a unique sample to study the formation of the first supermassive black holes (SMBHs) and their host galaxies at the epoch of cosmic reionization. There are currently more than 60 quasars known at z ∼ 6, selected from large optical and near-infrared surveys (e.g., the Sloan Digital Sky Survey [SDSS], Fan et al. 2006; Jiang et al. 2008, 2009; the Canada–France High Redshift Quasar Survey, Willott et al. 2007, 2009, 2010; the UKIDSS Large Area Survey [ULAS], Mortlock et al. 2009, 2011; and the Panoramic Survey Telescope and Rapid Response System [Pan-STARRS], Morganson et al. 2012) with optical z-band magnitudes from 18.8 to 24.4 and inferred SMBH masses from 10⁸ to a few 10⁹ M_☉ (Kurk et al. 2007; Jiang et al. 2007; Willott et al. 2010; De Rosa et al. 2011). The formation of 10⁹ M_☉ SMBHs at the highest redshift suggests fast black hole accretion and significant SMBH-galaxy evolution within 1 Gyr after the big bang (Li et al. 2007; Narayanan et al. 2008; Di Matteo et al. 2012). The co-evolution of the first SMBHs and their host galaxies are studied with observations of dust continuum, molecular CO, and [C ii] line emission at submillimeter and millimeter ((sub)mm) wavelengths (e.g., Priddey & McMahon 2001; Walter et al. 2003; Maiolino et al. 2005). Dust heated by the UV photons from young, massive stars presents strong thermal FIR continuum emission. Due to the negative K-correction (Blain & Longair 1993), observations of dust continuum emission at (sub)mm wavelengths provide an efficient way to search for star forming activity at high redshifts (Omont et al. 1996, 2003; Priddey et al. 2003; Bertoldi et al. 2003a; Wang et al. 2007). Molecular gas traced by CO transitions provides the requisite fuel for star formation (Narayanan et al. 2008; Carilli & Walter 2013), and the [C ii] 158 μm fine structure line emission is a principal interstellar coolant, which probes the star forming-powered photodissociation regions (PDRs) and interstellar medium in these earliest quasar host galaxies (Walter et al. 2009; De Looze et al. 2011; Gallerani et al. 2012; Wagg et al. 2012; Venemans et al. 2012; Carilli & Walter 2013).

Dust and high-order molecular CO line emission in the host galaxies of the z ∼ 6 quasars have been searched for using the Max Planck Millimeter Bolometer Array (MAMBO) on the IRAM-30 m telescope (Bertoldi et al. 2003a; Petric et al. 2003; Wang et al. 2007, 2008, 2011a; Omont et al. 2013), the Submillimeter Common User Bolometer Array (SCUBA) on the James Clerk Maxwell Telescope (Priddey et al. 2003; Robson et al. 2004; Priddey et al. 2008), and the IRAM Plateau de Bure interferometer (PdBI; Bertoldi et al. 2003b; Walter et al. 2003; Carilli et al. 2007; Wang et al. 2010, 2011a). The MAMBO survey of SDSS z ∼ 6 quasars at 250 GHz (typical 1σ errors of 0.6 mJy) found that about 30% of them show strong continuum emission from dust at a temperature of 40–60 K, with FIR luminosities of a few 10¹² to 10¹³ L_☉ and inferred dust masses of a few 10⁸ M_☉ (Priddey & McMahon 2001; Bertoldi et al. 2003a; Petric et al. 2003; Wang et al. 2011a). Most of the FIR-luminous z ∼ 6 quasars also show bright CO line emission, implying that there is ∼10¹⁰ M_☉ of molecular gas in the quasar host galaxies (Walter et al. 2003; Carilli et al. 2007; Wang et al. 2010, 2011b). The strong FIR continuum and molecular CO line emission suggest that there is active star formation in the host galaxies of these FIR luminous z ∼ 6 quasars. The star formation rates inferred from the FIR luminosities range from about a few hundred to one thousand M_☉ yr⁻¹ (Bertoldi et al. 2003a; Wang et al. 2008, 2011a).

The CO detections from the FIR-luminous z ∼ 6 quasars also provide constraints on the spatial distribution of molecular gas and the dynamical masses of the quasar host galaxies (Walter et al. 2004; Wang et al. 2011b). Deep CO line imaging of one of the most FIR-luminous quasars, SDSS J114816.64+525150.3 at z = 6.42 (hereafter J1148+5251), using the Very Large Array and the PdBI (Walter et al. 2004; Riechers et al. 2009) at subarcsecond resolution reveals a CO source size of about 3.6 kpc × 1.4 kpc (FWHM) in the central region of the quasar host galaxies. Recent JVLA observations of the CO (2–1) line emission from another FIR-luminous quasar, SDSS J092721.82+200123.7 at z = 5.77, at 2'' resolution has constrained the source size to be about 10 kpc (Wang et al. 2011b). The line widths of the CO-detected z ∼ 6 quasars range widely from 160 to 860 km s⁻¹. The host galaxy dynamical masses (within the CO emitting region) estimated from the CO line widths and the sizes of the bright detections indicate a median SMBH–bulge mass ratio about one order of magnitude higher than the present-day value (Walter et al. 2004; Wang et al. 2010).

Strong [C ii] 158 μm line emission was also detected from the host galaxies of the FIR and CO-luminous quasars at z ∼ 6 (Maiolino et al. 2005, 2009; Willott et al. 2013). The PdBI image of [C ii] line of J1148+5251 shows two components. The narrow-line component with a line width of 345 km s⁻¹ and a spatial extension of FWHM ∼ 15 (8 kpc) traces the distributed star formation in the quasar host galaxy (Maiolino et al. 2012) and the compact core of this component within the central ∼03 region suggests a very high peak star formation rate surface density of ∼1000 M_☉ yr⁻¹ kpc⁻² (Walter et al. 2009). The broad-line component shows a line width of FWHM ∼ 2000 km s⁻¹, distributed over ∼16 kpc, suggesting gas outflow driven by the central luminous quasar (Maiolino et al. 2012; Valiante et al. 2012). The detection of dust continuum, CO, and [C ii] line emission in the quasar host galaxies at z ∼ 6 suggests an early phase of SMBH-galaxy evolution, in which the SMBHs are accreting at their Eddington limit and the quasar stellar bulges are still accumulating their mass via massive star formation. The interferometer observations of J1148+5251 have demonstrated that the [C ii] line emission is an ideal tracer of star-forming activity and gas dynamics in the nuclear region of the starburst quasar host galaxies at the highest redshift.

The Atacama Large Millimeter/submillimeter Array (ALMA) in the early science phase provides a wide frequency coverage for observations of fine structure line emission at high redshift. It also provides the required sensitivity and allows detections of line emission with peak flux density of a few mJy in only one hour of observing time. Additionally, the spatial resolution of ∼07 in Band 6 and 7 in the extended configuration (with a maximum baseline of ∼400 m) can measure or tightly constrain the distribution of the gas component from high redshift quasar host galaxies. Thus, the Cycle 0 phase of ALMA opens a unique opportunity to expand the observation of [C ii] line emission to a large sample of quasar-starburst systems at the highest redshift. In this paper, we present ALMA observations of the [C ii] line emission from five FIR luminous quasars at z ∼ 6. We also report IRAM observations of the 250 GHz dust continuum and CO (6–5) line emission from the quasar SDSS J231038.88+185519.7 at z = 6.00 (X. Fan et al. 2013, in preparation). We describe the observations in Section 2, present the results in Section 3, discuss the star forming activity and gas dynamics based on the [C ii] line detections in Section 4, and summarize the results in Section 5. A ΛCDM cosmology with H₀ = 71 km s⁻¹ Mpc⁻¹, Ω_M = 0.27 and Ω_Λ = 0.73 is adopted throughout this paper (Spergel et al. 2007).

2.1. IRAM Observations of SDSS J231038.88+185519.7

The quasar SDSS J231038.88+185519.7 (hereafter J2310+1855) is a broad absorption line quasar discovered in the SDSS (X. Fan et al. 2013, in preparation). The redshift measured with the quasar rest-frame UV line emission is z = 6.00 ± 0.03. It is one of the optically brightest sources among the known z ∼ 6 quasars with rest frame 1450 Å magnitude of m₁₄₅₀ = 19.3. We observed the 250 GHz dust continuum from this object with the MAMBO-II 117-element array on the IRAM 30 m telescope (Kreysa et al. 1998) in the winter of 2010–2011. We adopted the standard on–off photometry mode with a chopping rate of 2 Hz by 32'' in azimuth. The total observing time was 100 minutes with 50 minutes on source. The data were reduced with the MOPSIC pipeline (Zylka et al. 1998). We reached a 1σ sensitivity of 0.63 mJy and the source is detected at 8.29 mJy (∼13σ), making it the brightest 250 GHz detection among all the known z ∼ 6 quasars.

We then searched for the CO (6–5) line emission from this object using the PdBI. The observation was carried out in 2011 in D configuration with a synthesized beam size (FWHM) of 54 × 39. We spent 3 hr on source using the new wide band correlator WideX, with a bandwidth of 3.6 GHz in dual polarization. The phase was checked with the quasar 3C 454.3 and the flux calibration uncertainty is 5%. The data were reduced with the IRAM GILDAS software package (Guilloteau & Lucas 2000). We detected the line at 11σ with a typical rms of 0.5 mJy beam⁻¹ per 100 km s⁻¹ channel (Table 1 and Figure 1). We also detected the 99 GHz dust continuum at 8σ with a continuum 1σ sensitivity of 0.05 mJy beam⁻¹.

Zoom In Zoom Out Reset image size

Table 1. Summary of Previous Observations

Source	m1450	S250	zCO	FWHMCO	SΔvCO(6–5)	LCO(1–0)
		(mJy)		(km s−1)	(Jy km s−1)	(105 L☉)
(1)	(2)	(3)	(4)	(5)	(6)	(7)
SDSS J231038.88+185519.7	19.30	8.29 ± 0.63	6.0025 ± 0.0007	456 ± 64	1.52 ± 0.13	32.1
ULAS J131911.29+095051.4	19.65	4.20 ± 0.65	6.1321 ± 0.0012	537 ± 123	0.43 ± 0.09	9.4
SDSS J205406.49−000514.8	20.60	2.38 ± 0.53	6.0379 ± 0.0022	360 ± 110	0.34 ± 0.07	7.3
SDSS J012958.51−003539.7	22.28	2.37 ± 0.49	5.7794 ± 0.0008	283 ± 87	0.37 ± 0.07	7.4
SDSS J104433.04−012502.2	19.21	1.82 ± 0.43	5.7824 ± 0.0007	160 ± 60	0.21 ± 0.04	4.2

Notes. Column 1: name; Column 2: magnitudes at rest-frame 1450 Å (X. Fan et al. 2013, in preparation; Mortlock et al. 2009; Jiang et al. 2008, 2009; Fan et al. 2000); Column 3: continuum flux density at 250 GHz; Columns 4 and 5: redshift and FWHM line width of the CO (6–5) line; Column 6: CO (6–5) line flux of the five objects; Column 7: CO (1–0) line luminosity calculated from the CO (6–5) line emission assuming a CO excitation ladder similar to J1148+5251 (Riechers et al. 2009). The CO (6–5) line measurements for J2310+1855 are from this work, and from Wang et al. (2010, 2011a) for the other four sources.

Download table as:  ASCIITypeset image

2.2. ALMA Observations

To study the [C ii] 158 μm line emission and the underlying continuum with ALMA, We selected J2310+1855 and another four z ∼ 6 quasars that were detected in bright millimeter dust continuum and CO (6–5) line emission (Petric et al. 2003; Priddey et al. 2003, 2008; Wang et al. 2008, 2010, 2011a; X. Fan et al. 2013, in preparation) and have declination of δ ⩽ +20°. We list the sources and measurements from previous MAMBO observations in Table 1. The ALMA observations were carried out in Cycle 0 in 2012 with 14–21 12 m diameter antennas in the extended array with baselines from 36 m to 400 m. The typical synthesized beam size (FWHM) is 07, corresponding to 4 kpc at z ∼ 6. We observed the [C ii] 158 μm (1900.5369 GHz) line emission in Band 6 or 7, with one 2 GHz window centered at the line frequency and the other three 2 GHz windows observing the dust continuum. The correlator channel width is 15.625 MHz, or 16–18 km s⁻¹ at the observed line frequency. The phase stability was checked every ∼8 minutes by observing nearby phase calibrators (e.g., bright quasars). We spent 50–90 minutes on source for each target and reached 1σ spectral sensitivities of 0.4–0.7 mJy beam⁻¹ per 62.5 MHz re-binned channel after continuum subtraction. The continuum sensitivities are 0.04–0.09 mJy beam⁻¹ (Table 2).

Table 2. ALMA Observations

Source	ton	$z_{[C\,\scriptsize{II}]}$	SΔv	$\rm FWHM_{[{\rm C\,\scriptsize{II}}]}$	$\rm rms_{[{\rm C\,\scriptsize{II}}]}$	νcon	Scon	rmscon
	(min)		(Jy km s−1)	(km s−1)	(mJy beam−1)	(GHz)	(mJy)	(mJy beam−1)
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
J2310+1855	50	6.0031 ± 0.0002	8.83 ± 0.44	393 ± 21	0.5	263	8.91 ± 0.08	0.06
J1319+0950	80	6.1330 ± 0.0007	4.34 ± 0.60	515 ± 81	0.7	258	5.23 ± 0.10	0.08
J2054−0005	40	6.0391 ± 0.0001	3.37 ± 0.12	243 ± 10	0.4	262	2.98 ± 0.05	0.04
J0129−0035	60	5.7787 ± 0.0001	1.99 ± 0.12	194 ± 12	0.4	287	2.57 ± 0.06	0.05
J1044−0125	87	5.7847 ± 0.0007	1.70 ± 0.30	420 ± 80	0.6	287	3.12 ± 0.09	0.09

Notes. Column 1: source; Column 2: on-source time of the ALMA [C ii] observations; Columns 3–5: redshift, flux, and FWHM line width of the [C ii] line fitted to a single Gaussian line profile; Column 6: ALMA 1σ line sensitivity bin to a channel width of 62.5 MHz (∼70 km s⁻¹); Columns 7 and 8: frequency and flux density of the continuum measured with the line-free windows; Column 9: ALMA 1σ rms continuum sensitivity integrated over a total bandwidth of 5.8–6 GHz. The flux errors quoted here are from our fitting process described in Section 3. The calibration uncertainties are not included here.

Download table as:  ASCIITypeset image

We reduced the data using the CASA package.¹⁶ The flux scale was calibrated in four of the sources by observing Titan, Neptune, and Uranus, and the calibration uncertainties are better than 15%. For the source J1044−0125, no primary amplitude calibrator was observed. The flux scale was determined with 3C 273 based on the ALMA and SMA observations in 2012 March and April at 225 GHz and 343 GHz. Considering the uncertainties in the flux densities and spectral index of the calibrator, we estimate a total flux calibration error of 20% for the observation of J1044−0125. We derive the line frequencies in the kinematic local standard of rest frame to compare the results to previous PdBI CO (6–5) line observations.

We detected strong [C ii] line and dust continuum emission in the host galaxies of all the five z ∼ 6 quasars. The line emission of the five objects and the continuum emission from four of them (the exception is J1044−0125) are marginally resolved. The line intensity-weighted velocity maps of J0129−0035, J1319+0950, and J2310+1855 show clear velocity gradients (Figure 2). We fit the sizes of the line or continuum sources using the IMFIT task in CASA, which performs synthesized beam deconvolution and two-dimensional (2D) Gaussian fitting to the images. The resulting deconvolved full width at half maximum (FWHM) major axis sizes are 03–06 (1.7–3.5 kpc) for the line and 02–04 (1.2–2.3 kpc) for the dust continuum. We list the fitted parameters for each object below. However, we emphasize that, due to the limited spatial resolution, these measurements should be considered as tight constraints on the spatial extent of the sources, rather than accurate source morphology measurements. Interferometer imaging at higher resolution is required to finally determine the source morphology. To measure the line center, FWHM line width, and flux, we subtract the underlying dust continuum in the U − V plane, calculate the line spectra by integrating the intensity over the line-emitting region in each channel, and fit the spectra with a Gaussian profile. The [C ii] results are summarized in Table 2, and the line and continuum maps are presented in Figures 2 and 3.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

SDSS J2310+1855. This object is the brightest detection among all the z ∼ 6 quasars from our MAMBO observations. The 250 GHz continuum flux density is 8.29 ± 0.63 mJy. Our PdBI observation of the CO(6–5) line emission shows a line flux of 1.52 ± 0.13 Jy km s⁻¹, which is also by far the strongest CO line flux. The redshift and FWHM measured with the CO (6–5) line are z_CO = 6.0025 ± 0.0007 and FWHM_CO = 456 ± 64 km s⁻¹. The dust continuum at 99 GHz has also been detected in the line-free channels with a flux density of 0.40 ± 0.05 mJy (Figure 1).

We detect the [C ii] line emission with a flux which is again more than twice stronger than the four other sources. The derived FWHM source size, after deconvolving the 072 × 051 synthesized beam, is (056 ± 003) × (039 ± 004) with a position angle (P.A.) of 142° ± 10° east of north. The line intensity-weighted velocity map derived with the line-emitting channels indicates a velocity gradient from southwest to northeast (see the right panels of Figure 2). The 263 GHz continuum flux density imaged with the line-free windows and fitted to a 2D Gaussian is 8.91 ± 0.08 mJy, which is in very good agreement with the MAMBO value. This constrains the (deconvolved) continuum source size to be (025 ± 002) × (020 ± 002) with P.A. = 162° ± 18°.

J1319+0950. This quasar was discovered in the UKIRT Infrared Deep Sky Survey (UKIDSS) and has m₁₄₅₀ = 19.65 (Mortlock et al. 2009). The [C ii] line emission is imaged by ALMA with a synthesized beam size of 069 × 049. The 2D Gaussian fit suggests line emission with a deconvolved source size of (057 ± 007) × (032 ± 015) oriented at P.A. = 28° ± 18°. The line velocity map shows a clear gradient from the southwest to the northeast. We plot the [C ii] line channel map of this object in Figure 3. The map shows a clear position shift of the [C ii] line peak (by ∼04) from the +222 km s⁻¹ channel to the −200 km s⁻¹ channel. The 258 GHz continuum flux density is 5.23 ± 0.10 mJy and the deconvolved continuum source size is (039 ± 002) × (034 ± 003) with P.A. = 121° ± 148°.

J2054−0005. The quasar was selected from SDSS stripe 82 with m₁₄₅₀ = 20.60, i.e., about one magnitude fainter than the objects discovered from the SDSS main survey (Jiang et al. 2008). We observed the [C ii] line emission at 064 × 058 resolution. The deconvolved FWHM source size is (035 ± 004) × (032 ± 005) and P.A. = 91° ± 173°. The 262 GHz continuum flux density measured in the line-free windows is 2.98 ± 0.05 mJy, and the deconvolved continuum source size is (027 ± 003) × (026 ± 003) with P.A. = 168° ± 68°.

J0129−0035. This quasar was discovered in SDSS stripe 82 with m₁₄₅₀ = 22.16 (Jiang et al. 2009). It is the faintest optical source among the MAMBO-detected z ∼ 6 quasars, but previous millimeter observations found bright 250 GHz dust continuum and CO (6–5) line emission from this object (Wang et al. 2011a). The [C ii] line was observed at 057 × 049 resolution. The deconvolved source size is estimated to be (041 ± 006) × (023 ± 012) with P.A. = 12° ± 22°. A velocity gradient is found from the southeast to the northwest. A 2D Gaussian fit to the continuum image gives a 287 GHz flux density of 2.57 ± 0.06 mJy, with a deconvolved continuum source size of (023 ± 004) × (014 ± 007) at P.A. = 115° ± 154°.

J1044−0125. This is another broad absorption line quasar at z ∼ 6 discovered from the SDSS main survey with m₁₄₅₀ = 19.2 (Fan et al. 2000). A deep near-IR spectrum of the quasar C iv line emission of this object yields a SMBH mass of 10.5 × 10⁹ M_☉ (Jiang et al. 2007). The beam size of the [C ii] line observation is 066 × 045 and the deconvolved source size is estimated to be (061 ± 011) × (033 ± 022) oriented at P.A. = 65° ± 48°. No gradient is seen in the line intensity-weighted velocity map of this object. The 287 GHz continuum flux density is 3.12 ± 0.09 mJy and the source is unresolved in the continuum.

We calculate the [C ii] line luminosity of each source as $L_{[{\rm C\,\scriptsize{II}}]}/L_{\odot }=1.04\times 10^{-3}S\Delta v\nu _{0}(1+z)^{-1}{D_{L}}^{2}$ (Solomon & Vanden Bout 2005), where ν₀ = 1900.5369 GHz is the rest frame [C ii] line frequency, SΔv is the integrated line flux in Jy km s⁻¹, and D_L is the luminosity distance in Mpc. The derived line luminosities are in the range of 1.6–8.7 × 10⁹ L_☉. We also calculate the CO (6–5) line luminosity for J2310+1855, which is (5.4 ± 0.5) × 10⁸ L_☉ or (5.1 ± 0.4) × 10¹⁰ K km s⁻¹ pc².

We then estimate their FIR luminosities (L_FIR) by fitting a modified blackbody (i.e., S_ν ∼ ν^{3 + β}/(exp(hν/kT_dust) − 1); Priddey & McMahon 2001; De Breuck et al. 2003; Kovács et al. 2006) to the ALMA continuum flux densities, MAMBO (Petric et al. 2003; Wang et al. 2007, 2008, 2011a), and available PdBI and SCUBA data (Priddey et al. 2003, 2008; Wang et al. 2010, 2011a). For four of the five objects (except J2054−0005), we adopt a dust temperature of T_dust = 47 K and emissivity index of β = 1.6, which are the mean values found in the sample of high-z FIR luminous quasars (Beelen et al. 2006). This yields the FIR luminosities (L_FIR) in the range 42.5–122.5 μm of 5–17 × 10¹² L_☉.¹⁷ The estimates of L_FIR could be larger by up to 20% if a higher dust temperature of 50 K is assumed, or lower by ⩽8% if the best fit of z > 4 quasars from Priddey & McMahon (2001; i.e., T_dust  =  41 K and β  =  1.95) is adopted. J2054−0005 was recently detected by Herschel/SPIRE at 250 μm and 350 μm, which samples the continuum at wavelengths close to or shorter than the peak of the starburst-powered thermal dust emission (Leipski et al. 2013). As the 250 μm flux density might be significantly contaminated by the emission from the active galactic nucleus (AGN) dust torus (Leipski et al. 2013), we fit L_FIR for this object with the Herschel 350 μm, the MAMBO 250 GHz, and ALMA 262 GHz measurements, and leave T_dust as a free parameter. The fit suggests T_dust = 52  ±  6 K and L_FIR = (8.0  ±  3.3)  ×  10¹² L_☉. We have listed the derived [C ii] and FIR luminosities in Table 3.

Table 3. Luminosities and Dynamical Masses

Source	$L_{[{\rm C\,\scriptsize{II}}]}$	LFIR	Lbol	MBH	Mdynsin 2i	Mdyn	MBH/Mdyn
	(109 L☉)	(1012 L☉)	(1013 L☉)	(109 M☉)	(1010 M☉)	(1010 M☉)
(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
J2310+1855	8.7 ± 1.4	17.0 ± 1.8	9.3	2.8	4.9 ± 0.6	9.6	0.030
J1319+0950	4.4 ± 0.9	10.7 ± 1.3	7.0	2.1	8.5 ± 2.9	12.5	0.017
J2054−0005	3.3 ± 0.5	8.0 ± 3.3	2.8	0.86	1.2 ± 0.2	7.2	0.012
J0129−0035	1.8 ± 0.3	4.6 ± 0.6	0.57	0.17	0.9 ± 0.2	1.3	0.013
J1044−0125	1.6 ± 0.4	5.5 ± 0.7	11.6	10.5	...	...	...

Notes. Column 1: source; Column 2: [C ii] line luminosity; Column 3: FIR luminosity in the wavelength range from 42.5 μm to 122.5 μm (see Section 3 for details). We considered both measurement errors listed in Table 2 and the 15%–20% calibration uncertainties in the calculation of $L_{[{\rm C\,\scriptsize{II}}]}$ and L_FIR. Column 4: quasar bolometric luminosities. The bolometric luminosity for J1044−0125 is taken from Jiang et al. (2006), and we estimate the bolometric luminosities for the other four objects from their rest-frame 1450 Å magnitude, assuming an isotropic bolometric correction of L_bol = 4.2L₁₄₅₀ (Runnoe et al. 2012a, 2012b); Column 5: SMBH masses. The SMBH mass of J1044−0125 is calculated from the quasar C iv line emission (Jiang et al. 2007), and we estimate the SMBH masses for the other four sources from the bolometric luminosities assuming Eddington accretion; Column 6: dynamical mass without inclination angle correction, estimated from the [C ii] line width and source size (see Section 4.2); Column 7: inclination angle-corrected dynamical mass within the [C ii]-emitting region. We estimate the disk inclination angle from the [C ii] minor and major axis ratio as i = 46°, 56°, 24°, and 56° for J2310+1855, J1319+0950, J2054−0005, and J0129−0035, respectively; Column 8: SMBH–bulge mass ratio.

Download table as:  ASCIITypeset image

4.1. Distributions of Gas, Dust, and Star Formation in the Quasar Host Galaxies

The new detections of strong [C ii] line emission toward the five FIR-luminous quasars at z ∼ 6 provide further evidence for active star formation in the quasar host galaxies. The [C ii] line emission is marginally resolved in all these objects. The ALMA observations at 07 resolution yield estimates of the intrinsic [C ii] source sizes (FWHM of the major axis) of 03–06 (1.7–3.5 kpc). Clear velocity gradients have been found in the line intensity-weighted velocity maps of J2310+1855, J1319+0950, and J0129−0035, which suggests that the gas could be in rotation, and gravitationally bound, e.g., a non-face-on rotating disk. The dust continuum flux densities measured in the vicinity of the [C ii] line frequencies for all five objects are consistent with previous MAMBO measurements at 250 GHz. The continuum sources of four of them (except J1044−0125) are marginally resolved, indicating deconvolved FWHM major axis sizes of 02–04, or 1.2–2.3 kpc. These results constrain the spatial extent of star forming activity to be 2.6–5.3 kpc in diameter¹⁸ in the nuclear region. We will observe these sources with ALMA in Cycle 1 at 02 resolution to fully resolve the line and dust continuum sources, and measure what fraction of the dust continuum emission is from the central compact AGN. These observations will finally measure the star formation rates (SFR) and SFR surface densities in these earliest quasar-starburst systems. The higher resolution imaging will also address if the gas components in the nuclear starburst region are uncoalesced and show multiple-peak morphology in line emission (Walter et al. 2004), which was suggested by the galaxy merger models of quasar-galaxy formation (e.g., Narayanan et al. 2008)

In Figure 4, we plot the [C ii]-to-FIR luminosity ratios of the [C ii]-detected z ⩾ 5.8 quasars, including the five new detections in this work, J1148+5251 (Maiolino et al. 2005), CFHQS J0210−0456 at z = 6.43 (Willott et al. 2013), and ULAS J1120+0640 at z = 7.08 (Venemans et al. 2012). We compare them to samples of [C ii]-detected local normal star forming galaxies, ULIRGs, submillimeter galaxies, and FIR-luminous quasars at high redshift (Malhotra et al. 2001; Luhman et al. 2003; Stacey et al. 2010; Maiolino et al. 2009; Ivison et al. 2010; De Breuck et al. 2011; Swinbank et al. 2012; Wagg et al. 2012; Pety et al. 2004; Gallerani et al. 2012; Carilli et al. 2013; Valtchanov et al. 2011; Riechers et al. 2013; Marsden et al. 2005). The five z ∼ 6 quasars presented in this work, as well as J1148+5251, show luminosity ratios of 2.9–5.1 × 10⁻⁴, which are comparable to the typical values found in local ULIRGs and 1 ⩽ z ⩽ 5 [C ii]-detected quasars, and a few to 10 times lower than that of the disk star forming galaxies and submillimeter galaxies. We also notice that the other two z > 6 quasars with moderate FIR luminosities (10¹¹–10¹² L_☉) show higher $L_{[{\rm C\,\scriptsize{II}}]}/L_{{\rm FIR}}$ than most FIR luminous objects (Venemans et al. 2012; Willott et al. 2013). We estimate the CO (1–0) line luminosities for the five objects from the PdBI CO (6–5) detections (last column of Table 1), assuming a CO excitation ladder similar to J1148+5251 (Riechers et al. 2009). The calculated [C ii]-to-CO (1–0) line luminosity ratios are about 2400–4700, which is slightly higher than the values found in local ULIRGs (Luhman et al. 2003), and the highest value is close to the median luminosity ratios found in starburst galaxies (i.e., $L_{[{\rm C\,\scriptsize{II}}]}/L_{{\rm CO}}\sim 4400$; Stacey et al. 1991, 2010; Swinbank et al. 2012).

Zoom In Zoom Out Reset image size

The [C ii], FIR, and CO luminosity ratios of the five [C ii]-detected z ∼ 6 quasars can be reproduced by PDR models (Kaufman et al. 1999; Stacey et al. 2010; Luhman et al. 2003) with gas density on orders of 10⁴–10⁵ cm⁻³ and FUV (6 eV <hν < 13.6 eV) radiation field G₀ of a few 10³–10⁴ (in units of the Habing Field, 1.6 × 10⁻³ erg cm⁻² s⁻¹; Kaufman et al. 1999). However, such strong radiation fields will also produce strong [O i] 63 μm line emission with intensities comparable or higher than the [C ii] line (Kaufman et al. 1999; Luhman et al. 2003; Carilli et al. 2013). Thus future observations of the [O i] and other fine structure lines from these objects with the full configuration of ALMA will be a crucial test of the physical conditions inferred from these PDR models. It is also possible that the central AGN has substantial contribution to the dust heating and FIR emission, which results in lower [C ii]-to-FIR luminosity ratios in the nuclear region (Luhman et al. 2003; Sargsyan et al. 2012).

We compare the [C ii] and CO (6–5) line profiles of the five objects in Figure 5. For four of the five sources, the redshifts measured with [C ii] and CO (6–5) are consistent within the 1σ errors; there are no large velocity offsets between the gas components traced by [C ii] and CO (6–5) lines. The [C ii] FWHM line widths of J2310+1855, J2054−0005, and J0129−0035 are about 60 to 115 km s⁻¹ smaller than the CO (6–5) measurements. But these differences are within the 1σ–2σ error bars, as the CO (6–5) line width uncertainties for these objects are between 60 and 120 km s⁻¹ (Table 1). The other object, J1044−0125, shows a larger [C ii] redshift with $\Delta z=z_{[{\rm C\,\scriptsize{II}}]}-z_{{\rm CO}}=0.0023\pm 0.0010$ (i.e., a velocity difference of 100 ± 44 km s⁻¹), and a much broader [C ii] line width (Figure 5). This may indicate different kinematical properties between the two gas components in this object. However, it is also possible that a large fraction of the CO line emission is undetected and the CO line width is underestimated due to the low signal-to-noise ratio of the line spectrum. Thus, deep imaging of the CO line emission with better measurements of the CO line profile and spatial distribution is needed to address whether the [C ii]-emitting gas is more centrally concentrated in J2310+1855, J2054−0005, and J0129−0035, and to understand the origin of the narrower CO line in J1044−0125.

Zoom In Zoom Out Reset image size

4.2. Dynamical Masses Traced by the [C ii] Line Emission

4.3. Search for [C ii] Line Emitters Associated with the Quasar Environments

We detected [C ii] fine structure line and dust continuum emission from the host galaxies of five quasars at z ∼ 6, using ALMA in the Cycle 0 phase at ∼07 resolution. Complementary IRAM observations of the CO (6–5) line and 250 GHz dust continuum emission from the z = 6.00 quasar J2310+1855 are also presented. Our ALMA observations measure the FWHM major axis sizes of the [C ii] emission from the five objects to be 03–06 (1.7–3.5 kpc) and the sizes of the dust continuum source for four of them to be 02–04 (1.2–2.3 kpc). The detections of [C ii] line and dust continuum emission indicate active star formation in the central few kpc region of the quasar host galaxies. The derived [C ii]-to-FIR luminosity ratios are of the order 10⁻⁴, which are comparable to the typical values found in local ULIRGs and other FIR-luminous quasars at high redshift. The intensity-weighted velocity maps of J2310+1855, J1319+0950, and J0129−0035 show velocity gradients. Such velocity gradients are consistent with rotation, although they are not uniquely interpreted as such. We estimate the dynamical masses within the [C ii]-emitting region for four of the five objects assuming that the gas is distributed in a rotating disk. The derived ratios between the SMBH masses and the dynamical masses are one order of magnitude higher than that of local normal galaxies.

The detections of [C ii] 158 μm line emission from quasar host galaxies at z ∼ 6 have demonstrated the power of ALMA in observing signatures of star formation at the earliest cosmic epoch. With the full configuration of ALMA, we should be able to resolve the line and dust continuum emission on kpc or sub-kpc scales, which will measure the surface densities of the gas components and star forming activity in the nuclear region, and better address the gas kinetics and dynamical masses of the spheroidal quasar host galaxies. The full frequency coverage of ALMA will also allow a search of other ionized/atomic interstellar cooling lines from these FIR-luminous z ∼ 6 quasars to measure the physical conditions (e.g., density, radiation field, temperature, etc.) of the interstellar medium in these earliest quasar-starburst systems.

We thank M. Lacy at the National Radio Astronomy Observatory for help with the observation and data analysis. This work is based on observations carried out with ALMA (NRAO), the Max Planck Millimeter Bolometer Array (MAMBO) on the IRAM 30 m telescope, and the Plateau de Bure Interferometer. The National Radio Astronomy Observatory (NRAO) is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc. This paper makes use of the following ALMA data: ADS/JAO.ALMA 2011.0.00206.S. ALMA is a partnership of ESO (representing its member states), NSF (USA) and NINS (Japan), together with NRC (Canada) and NSC and ASIAA (Taiwan), in cooperation with the Republic of Chile. The Joint ALMA Observatory is operated by ESO, AUI/NRAO and NAOJ. IRAM is supported by INSU/CNRS (France), MPG (Germany) and IGN (Spain). Frank Bertoldi and Fabian Walter acknowledge support through the DFG priority program 1573 and the SFB 956. Desika Narayanan acknowledges support from the NSF via grant AST-1009452. Kirsten Knudsen acknowledges support from the Swedish Research Council. X. Fan acknowledges support from NSF grant AST 08-06861 and 11-07682 and a David and Lucile Packard Fellowship.

Facilities: ALMA - Atacama Large Millimeter Array, IRAM:30m (MAMBO) - Institute de Radioastronomie Millimetrique 30 meter telescope, IRAM: Interferometer Europe -