[CII] 158 Micron Emission from z~4 HI Absorption-Selected Galaxies

TL;DR

This study uses ALMA to detect [CII] 158 micron emission from galaxies at z~4 associated with high-metallicity DLAs, revealing properties similar to luminous Lyman-break galaxies and showing that DLAs trace gas in these galaxies.

Contribution

First detection of [CII] emission from high-metallicity DLA galaxies at z~4, linking absorption features to galaxy emission properties.

Findings

[CII] emission detected in 5 of 6 targeted galaxies.

Galaxies have star-formation rates of 7-110 solar masses per year.

Absorption and emission line velocities are aligned, indicating the DLA gas is in the galaxy itself.

Abstract

We report on a search for the [CII] 158 micron emission line from galaxies associated with four high-metallicity damped Ly-alpha absorbers (DLAs) at z ~ 4 using the Atacama Large Millimeter/sub-millimeter Array (ALMA). We detect [CII] 158 micron emission from galaxies at the DLA redshift in three fields, with one field showing two [CII] emitters. Combined with previous results, we now have detected [CII] 158 micron emission from five of six galaxies associated with targeted high-metallicity DLAs at z ~ 4. The galaxies have relatively large impact parameters, ~16 - 45 kpc, [CII] 158 micron line luminosities of (0.36 - 30) x 10^8 Lsun, and rest-frame far-infrared properties similar to those of luminous Lyman-break galaxies, with star-formation rates of ~7 - 110 Msun yr-1. Comparing the absorption and emission line profiles yields a remarkable agreement between the line centroids, indicating that the DLA traces gas at velocities similar to that of the [CII] 158 micron emission. This disfavors a scenario where the DLA arises from gas in a companion galaxy. These observations highlight ALMA's unique ability to uncover a high redshift galaxy population that has largely eluded detection for decades.

Tables

Table 1. Table 1: Properties of the DLAs and Associated ALMA [C  ii ] 158  μ 𝜇 \mu m Emitters

	QSO J0834$+$2140	QSO J1101$+$0531	QSO J1253$+$1046	QSO PSS1443$+$27
DLA Properties
Right Ascension (J2000)	08:34:29.44	11:01:34.36	12:53:53.35	14:43:31.17
Declination (J2000)	$+$21:40:24.7	$+$05:31:33.8	$+$10:46:03.1	$+$27:24:36.7
Redshift	4.3900	4.3446	4.6001	4.2241
$\log (N$(H i)/cm${}_{}{}^{-2})$	$21.00\pm 0.20$	$21.30\pm 0.10$	$20.30\pm 0.15$	$21.00\pm 0.10$
$\left[M/H\right]$	$-1.30\pm 0.20$	$-1.07\pm 0.12$	$-1.36\pm 0.16$	$-0.95\pm 0.20$
$\Delta V_{90}$ (km s-1)	$290\pm 10$	$60\pm 10$	$70\pm 10$	$284\pm 10$
ALMA Observations
Continuum Resolution (${}_{}{}^{\prime \prime }\times _{}^{\prime \prime }$)a	($0.70\times 0.56$)	($0.99\times 0.91$)	($0.95\times 0.83$)	($0.67\times 0.48$)
Continuum RMS ($\mu$Jy beam-1)	16.4	24.5	22.7	26.2
Cube Resolution (${}_{}{}^{\prime \prime }\times _{}^{\prime \prime }$)a	($0.69\times 0.56$)	($0.97\times 0.89$)	($0.96\times 0.82$)	($0.67\times 0.47$)
Channel width (km s-1)	53.1	52.7	55.2	51.5
RMS per channel (mJy beam-1)	0.17	0.27	0.19	0.29
[C ii] 158 $\mu$m Emitter Properties
Right Ascension (J2000)	08:34:29.71	11:01:34.34	—	14:43:31.29	14:43:31.41
Declination (J2000)	$+$21:40:23.3	$+$05:31:37.8	—	$+$27:24:38.3	$+$27:24:38.8
Redshift	4.3896	4.3433	—	4.2227	4.2276
Impact Parameter (′′)	4.0	4.0	—	2.3	4.0
$S_{\mathrm{cont}}$ ($\mu$Jy)	$70\pm 22$	b	b,c	$161\pm 44$	b
$\int S_{\left[C\mathrm{II}\right]}𝑑V$ (Jy km s-1)	$0.173\pm 0.020$	$0.062\pm 0.012$	b,c	$0.846\pm 0.063$	$0.274\pm 0.043$
$[C$ ii]-$\Delta V_{90}$ (km s-1)d	$270\pm 60$	$370\pm 60$	—	$510\pm 60$	$620\pm 60$
$L_{\left[C\mathrm{II}\right]}$ (${10}^8$ L⊙)	$1.02\pm 0.12$	$0.36\pm 0.07$	b,c	$4.7\pm 0.4$	$1.5\pm 0.2$
$L_{\mathrm{TIR}}$ (${10}^{10}$ L⊙)e	$6.6\pm 2.1$	b	b,c	$14\pm 4$	b
SFR (M⊙ yr-1)	$7\pm 2$	b	b,c	$15\pm 4$	b
a The FWHM of the synthesized beam of the image.
b All upper limits are 3$\sigma$ and assuming the source is unresolved in the observations.
c Limit for a source within 10′′ of the DLA sight line, and assuming a FWHM line width of 100 km s-1.
d Velocity interval that contains 90 % of the velocity-integrated [C ii] 158 $\mu$m flux density.
e Systematic uncertainties are 0.5 dex.

Full text

quasars: absorption lines — galaxies: high-redshift — galaxies: ISM — submillimeter: galaxies —galaxies: kinematics and dynamics

††journal: ApJL

1 Introduction

Neutral atomic gas plays a central role in the formation and evolution of galaxies. Although simulations based on a $\Lambda$-cold dark matter ($\Lambda$-CDM) cosmology suggest that gas falling onto dark matter halos from the cosmic web is predominantly ionized, as the accreted gas migrates inward, the increase in density and pressure causes most of the gas to become neutral (e.g., Kereš et al., 2005; Dekel et al., 2009). The neutral gas can then cool further to form molecular gas, which eventually forms stars. Neutral atomic gas thus acts as a reservoir of fuel for future star formation; studying its properties, and its evolution with redshift, allows us to understand how galaxies assemble their baryonic mass and convert this into stars.

Unfortunately, it is not possible today to directly detect H i emission from galaxies at $z\gtrsim 0.5$ in reasonable integration times (Kanekar et al., 2016; Fernández et al., 2016). The Ly$\alpha$ absorption signature provides the only opportunity to detect this gas phase at high redshifts, during the epoch of galaxy assembly. The strongest Ly$\alpha$ absorbers, the so-called damped Ly$\alpha$ absorbers (DLAs; Wolfe et al., 2005), trace the bulk of the neutral atomic gas in the universe at all redshifts ($\approx 80\%$; e.g., O’Meara et al., 2007; Noterdaeme et al., 2012). Studying DLAs therefore allows us to directly probe the evolution of H i. However, to study the interplay between these H i reservoirs and galaxies, we need to characterize the galaxies with which the DLAs are associated.

Detecting the galaxies associated with high-$z$ DLAs is challenging (e.g., Kulkarni et al., 2006; Fumagalli et al., 2015). This has previously been assumed to be due to the inherent faintness of the DLA galaxies compared to the bright background quasar. Today, there are $\approx 20$ DLAs at $z\sim 2-3$ with detections of their associated galaxies at optical/near-infrared wavelengths (e.g., Krogager et al., 2017). Complementary to this approach, Neeleman et al. (2016) showed that the galaxies associated with Ly$\alpha$ absorbers can be detected at millimeter wavelengths, in their CO emission with the Atacama Millimeter/sub-millimeter Array (ALMA). Subsequent ALMA studies have yielded a high success rate in CO detections from galaxies associated with DLAs out to $z\approx 2.5$ (Møller et al., 2018; Kanekar et al., 2018; Neeleman et al., 2018; Fynbo et al., 2018; Klitsch et al., 2019), and possibly beyond (D’Odorico et al., 2018).

For galaxies at even higher redshifts, the fine structure line of singly-ionized carbon, [C ii] 158 $\mu$m, shifts into the ALMA observing bands. In Neeleman et al. (2017), we showed that the [C ii] 158 $\mu$m line can be used to identify galaxies associated with DLAs at $z\sim 4$. In this Letter, we report on an ALMA search for [C ii] 158 $\mu$m emission in four additional DLAs at $z\gtrsim 4$. We assume a standard, flat, $\Lambda$-CDM cosmology with $\Omega_{\lambda}=0.7$ and $H_{0}=70~{}\text{km}~{}\text{s}^{-1}~{}\text{Mpc}^{-1}$.

2 Sample Selection, Observations and Reduction

We used the ALMA Band-7 receivers to search for [C ii] 158 $\mu$m emission in the fields of four DLAs at $z\sim 4$ between UT 2016 March 16 and UT 2017 January 29 (ALMA proposal ID’s: 2015.1.01564.S and 2016.1.00569.S; PI: Neeleman). The four targets were selected from a parent sample of DLAs for which high-resolution ($R>10\,000$) optical spectra are available, allowing an accurate determination of the gas metallicity (Rafelski et al., 2012, 2014). All selected DLAs have relatively high metallicities, [M/H] $\geq-1.36$, which is higher than the median DLA metallicity at $z\sim 4$ ($\text{[M/H]}~{}\approx-1.8$; Rafelski et al., 2014). The high metallicity is an indicator that these DLAs are associated with more massive galaxies (Møller et al., 2013; Neeleman et al., 2013; Christensen et al., 2014), and therefore relatively high star formation rates (SFRs). The correlation between [C ii] 158 $\mu$m line luminosity and SFR in the local Universe (e.g., De Looze et al., 2014) further indicates that such high-metallicity DLAs are the best candidates for detections of [C ii] 158 $\mu$m emission.

The ALMA observations used four 1.875 GHz bands, each sub-divided into 128 channels. For each target, one of the bands was centered on the expected redshifted [C ii] 158 $\mu$m line frequency, while the remaining three bands were used to obtain a continuum image of the field. The on-source times were $1.2-2.8$ hours. All of the data were calibrated using the ALMA pipeline, in the Common Astronomy Software Applications (CASA; McMullin et al., 2007) package. Additional data editing was performed in CASA after the initial calibration. Two of the targets, QSO J1101$+$0531 and QSO J1253$+$1046, were sufficiently bright to perform self-calibration, which was done in the Astronomical Image Processing System (AIPS; Greisen, 2003) package. The final spectral cubes and continuum images were obtained using natural weighting, to maximize sensitivity, using the task tclean in CASA. Details of the sources and observations are provided in Table 1, and the integrated [C ii] 158 $\mu$m flux density and continuum images of the four fields are shown in Fig. 1.

3 Results

3.1 DLA J0834$+$2140

The $z=4.3900$ DLA towards QSO J0834$+$2140 has an H i column density of $\log(N_{\rm H{\scriptscriptstyle I}}/{\rm cm^{-2}})$ = 21.0 $\pm$ 0.2 and a metallicity of $-1.30\pm 0.20$. The multi-component velocity profile spans a large velocity interval, with a velocity spread $\Delta V_{90}\approx 290$ km s*-1* (e.g., Prochaska & Wolfe, 1997). This is significantly larger than the median $\Delta V_{90}$ in DLAs, $\approx 72$ km s*-1* (Neeleman et al., 2013), and indicates that the DLA is tracing either neutral gas belonging to a massive galaxy or a complex of lower-mass galaxies (e.g., Bird et al., 2015). Many metal line species are seen in the high resolution absorption spectrum of the DLA, including C+ (Rafelski et al., 2012), evidence that the gas has been enriched.

Our ALMA data yield a clear ($8.7\sigma$) detection of an emission line at an impact parameter of 4$\farcs$0 southeast of the quasar (Fig. 1), which we identify as the [C ii] 158 $\mu$m emission line at a redshift of $z=4.3896$. At this redshift, an impact parameter of 4.0*′′* corresponds to 27 kpc. The total velocity-integrated line flux density is $0.173\pm 0.020$ Jy km s*-1*, implying a [C ii] 158 $\mu$m line luminosity, $L_{\rm[C{\scriptscriptstyle II}]}$, of $(1.02\pm 0.12)\times 10^{8}$ L⊙ (e.g., Solomon et al., 1992). Within the positional uncertainty of the ALMA observations, a $3.7\sigma$ continuum feature is also seen at the position of the [C ii] 158 $\mu$m emission with a flux density of $70\pm 22$ $\mu$Jy. Besides the quasar, this is the most significant emission of the continuum image. Therefore, we tentatively take this as the continuum flux measurement of the [C ii] 158 $\mu$m emitter. The total infrared luminosity, $L_{\rm TIR}$, as determined from a modified black-body fit to the continuum measurement is $(6.6\pm 2.1)\times 10^{10}$ L⊙. We note that the quoted uncertainties on $L_{\rm TIR}$ are observational, and there is an additional systematic uncertainty of $\approx 0.5$ dex when estimating $L_{\rm TIR}$ from a measurement at a single wavelength (Neeleman et al., 2017). Using the conversion rate between SFR and 160 $\mu$m continuum emission (Calzetti et al., 2010), we obtain an SFR of $(7\pm 2)$ M⊙ yr*-1* for this galaxy.

3.2 DLA J1101$+$0531

The $z=4.3446$ DLA towards QSO J1101$+$0531 has an H i column density of $\log(N_{\rm H{\scriptscriptstyle I}}/{\rm cm^{-2}})$ = 21.3 $\pm$ 0.1 and a metallicity of $-1.07\pm 0.12$. The low-ionization metal lines show a narrow velocity spread, $\Delta V_{90}=60$ km s*-1*. This is lower than expected from the correlation between metallicity and $\Delta V_{90}$ (Neeleman et al., 2013), and may suggest that the DLA traces metal-rich gas around a low-mass galaxy or that the sightline intersects relatively unperturbed gas around a more massive system.

The ALMA observations show a $5.2\sigma$ emission feature 4$\farcs$0 north of the quasar (Fig. 1). The redshift of this [C ii] 158 $\mu$m emission line is 4.3433, which is within 75 km s*-1* of the redshift of the DLA, and the velocity-integrated line flux density is $(0.12\pm 0.04)$ Jy km s*-1* yielding $L_{\rm[C{\scriptscriptstyle II}]}$ = $(7.0\pm 2.3)\times 10^{7}$ L⊙. No continuum is detected at the position of the [C ii] 158 $\mu$m emission, putting an upper limit on the $L_{\rm TIR}$ and SFR of $<7.0\times 10^{10}$ L⊙ and $<7$ M⊙ yr*-1*, respectively.

3.3 DLA J1253$+$1046

The $z=4.6001$ DLA towards QSO J1253$+$1046 is both our highest-redshift and our lowest-metallicity target ([M/H] = $-1.36\pm 0.16$). Note that the DLA metallicity is still above the average DLA metallicity at $z\sim 4$. Its H i column density and $\Delta V_{90}$ are both low, $\log(N_{\rm H{\scriptscriptstyle I}}/{\rm cm^{-2}})$ = 20.3 $\pm$ 0.15 and $\Delta V_{90}\approx 70$ km s*-1*  perhaps indicating that the DLA is probing a less-massive galactic halo compared to our other targets (Neeleman et al., 2013).

We find no evidence of significant ($\gtrsim 5\sigma$) line emission in our ALMA images, near the DLA redshift. However, the 339 GHz continuum image shows a bright source 1.9*′′* southeast of the quasar. No emission lines are detected from this source in any of our four ALMA bands, consistent with a scenario in which the source is at the quasar redshift (which was not covered in the [C ii] 158 $\mu$m line). Assuming that the source is indeed at the quasar redshift ($z=4.908$), the measured continuum flux density of $(9.9\pm 0.6)$ mJy yields $L_{\rm TIR}$=$1\times 10^{13}$ L*⊙* and a SFR of $\approx 1000$ M⊙ yr*-1*, typical of high-$z$ sub-millimeter galaxies (SMGs) (Carilli & Walter, 2013). If the SMG is confirmed to be at the quasar redshift, the system would be similar to the well-studied quasar-SMG pair, QSO BR1202$-$0725 at $z\approx 4.7$ (e.g., Omont et al., 1996; Carilli et al., 2013), except that the separation between SMG and QSO is smaller here, $\approx 12$ kpc. We stress that this galaxy is not associated with the DLA, as the expected [C ii] 158 $\mu$m line luminosity of the SMG is significantly above the detection limit of our ALMA observations.

3.4 DLA PSS1443$+$2724

The $z=4.2241$ DLA towards PSS1443$+$2724 has $\log(N_{\rm H{\scriptscriptstyle I}}/{\rm cm^{-2}})$ = 21.3 $\pm$ 0.1 and a metallicity of [M/H]$=-0.95\pm 0.20$ (Prochaska et al., 2001). Strong low-ion absorption is seen over a wide range of velocities, with $\Delta V_{90}=284$ km s*-1*. In simulations, such large velocity spreads are typically seen in lines of sight passing through multiple halos (Bird et al., 2015). No optical counterpart was identified in deep ground-based imaging, with a $3\sigma$ $R$-band limit of $M_{\rm AB}=26.9$ (Prochaska et al., 2002).

Our ALMA data reveal strong ($>10\sigma$) line emission at an impact parameter of 2$\farcs$3 northeast of the quasar. We identify this as redshifted [C ii] 158 $\mu$m line emission from a $z=4.2256$ galaxy, only $-84$ km s*-1* offset from the DLA redshift. The velocity-integrated line flux density of $(0.846\pm 0.063)$ Jy km s*-1* yields a [C ii] 158 $\mu$m line luminosity of $(4.7\pm 0.4)\times 10^{8}$ L⊙. Additionally, at the position of the [C ii] 158 $\mu$m emitter, a $3.8\sigma$ excess is observed in the 356 GHz continuum. As in J0834$+$2140, besides the QSO, this is the most significant continuum emission. We therefore assume this emission arises from the [C ii] 158 $\mu$m emitter. The total flux density of $(161\pm 44)$ $\mu$m yields $L_{\rm TIR}$=$(14\pm 4)\times 10^{10}$ L⊙, and an SFR = $(15\pm 4)$ M⊙ yr*-1*. This is the only [C ii] 158 $\mu$m emitter of our present sample that is resolved, with a size of ($0\farcs 58\pm 0\farcs 17\times 0\farcs 4\pm 0\farcs 3$), corresponding to an emission region of only a few kpc.

In addition to the above [C ii] 158 $\mu$m line emission, the ALMA cube shows a second line feature, close to the DLA redshift. This $6.4\sigma$ feature is detected at an impact parameter of 4$\farcs$0 northeast of the quasar, which is 1$\farcs$5 distance from the other line emission. Assuming that this corresponds to redshifted [C ii] 158 $\mu$m emission, the emission velocity is offset by $+200$ km s*-1* from the DLA redshift. This is well within the velocity spread of the absorption (see Fig. 2). The velocity-integrated line flux density is $(0.274\pm 0.043)$ Jy km s*-1*, yielding $L_{\rm[C{\scriptscriptstyle II}]}$=$(1.5\pm 0.2)\times 10^{8}$ L⊙. No continuum emission is detected at the position of the [C ii] 158 $\mu$m emission, yielding a $3\sigma$ SFR limit of 7 M⊙ yr*-1*.

4 Discussion and Summary

4.1 [C ii] 158 $\mu$m Emission from Absorption-Selected Galaxies

The primary aim of the ALMA observations was to use [C ii] 158 $\mu$m emission to identify the galaxies associated with DLAs at $z\sim 4$. Together with the results of Neeleman et al. (2017), the success rate of this program has been remarkable, with 5 detections out of 6 targets. Fig. 3 shows the total far-infrared (FIR) luminosity plotted against the [C ii] 158 $\mu$m line luminosity for the galaxies of our sample, along with a set of low-redshift and high-redshift galaxy populations. The figure indicates that the FIR properties of galaxies selected through absorption are similar to those of luminous Lyman-break galaxies (LBGs) and color-selected galaxies (CSGs), supporting earlier conclusions that at least the most metal-enriched DLAs are associated with galaxies similar to luminous LBGs (Møller et al., 2002; Neeleman et al., 2017). Selecting galaxies through absorption, therefore provides a unique, complementary approach to study ‘normal’ galaxies at high redshift.

4.2 Sizes and Impact Parameters of [C ii] 158 $\mu$m Emitters

One of the most interesting results of our ALMA studies is the relatively large impact parameter ($\approx 16-45$ kpc) between the absorbers and the [C ii] 158 $\mu$m emitters (including the two systems of Neeleman et al., 2017). This indicates that the neutral gas probed by high-redshift, high-metallicity DLAs is not directly responsible for star formation in the associated galaxies, but probes a more extended H i envelope around these galaxies. While H i in nearby galaxies is also typically far more extended than the stars (e.g., Briggs et al., 1980), the high H i column densities ($\gtrsim 10^{21}$ cm*-2*) of our DLAs suggests that H i reservoirs around high-$z$ galaxies are more clumpy, since such high H i column densities are rarely seen at these large radial distances in nearby galaxies (e.g., Zwaan et al., 2005). In addition, the high gas metallicity ($\sim 1/10^{\rm th}$ solar) indicates that metals are efficiently mixed with the gas and transported to these distances.

Further, the large impact parameters imply that the associated galaxies are outside the point spread function (PSF) of the quasar for most modern optical/near-infrared instruments. Previous searches for high-$z$ DLA hosts have often attributed non-detections to the possibility that the galaxy might lie below the quasar PSF (e.g., Kulkarni et al., 2006). None of our $z\approx 4$ galaxies would have been below the quasar PSF for typical imaging studies. This is in agreement with the non-detection of star formation at the location of the DLA for a study where the quasar PSF was not an issue (Fumagalli et al., 2015). Our results thus indicate that at least for high-metallicity, high-redshift DLAs, the quasar PSF may not play a significant role in obscuring galaxies associated with DLAs. The non-detection of these systems in typical optical imaging may instead be explained by significant dust obscuration of the galaxy. ALMA CO and [C ii] 158 $\mu$m studies of DLA host galaxies thus provide a unique, complementary view to traditional optical imaging. We note, in passing, that some of these galaxies would not have been detected by the triple long-slit experiment (Fynbo et al., 2010; Krogager et al., 2017), due to their large angular separation (even in the absence of significant dust obscuration).

4.3 Kinematics of Emission and Absorption Lines

Fig. 2 shows the [C ii] 158 $\mu$m emission line profile compared with a low-ionization metal absorption line profile from the DLA. There is a striking agreement between the DLA redshift and the redshift of the [C ii] 158 $\mu$m emission in the figure (and in the corresponding figure of Neeleman et al., 2017). In all of the 5 detected [C ii] 158 $\mu$m emitters, the velocity difference between the emission and absorption line centroids is $<100$ km s*-1*. Only the second, fainter and more distant, [C ii] 158 $\mu$m emitter towards PSS 1443$+$2724, has a velocity offset of $+200$ km s*-1* from the centroid of the absorption line. This striking agreement disfavors a scenario whereby the DLA is probing gas solely associated with another, fainter —and presumably closer— galaxy, as one would then expect to see, on average, a larger velocity offset between the DLA absorption and the emission from the unrelated galaxy detected here (see also Neeleman et al., 2018; Fynbo et al., 2018). To be specific, if we assume the bright [C ii] 158 $\mu$m emitter resides in a moderate halo mass of $10^{11.5}$ M⊙, then the expected virial velocity is $\approx 350$ km s*-1*, which —accounting for projection effects— results in typical velocity offsets of $\approx 200$ km s*-1*.

The distribution of $\Delta V_{90}$ values in DLAs has long been known to be skewed towards large values (Prochaska & Wolfe, 1997), implying either large massive galaxies, or that the sightline crosses multiple smaller density structures/peaks (e.g., Prochaska & Wolfe, 1997; Bird et al., 2015). Fig. 7 of Bird et al. (2015) shows that the fraction of DLAs arising from multiple density peaks in simulations rises sharply for $\Delta V_{90}\gtrsim 200$ km s*-1*. Interestingly, two of the three systems in our sample with $\Delta V_{90}>200$ km s*-1*, show evidence for multiple [C ii] 158 $\mu$m emitters: DLA J1201$+$2117 appears to have two distinct [C ii] 158 $\mu$m components in the process of merging (Neeleman et al., 2017), whereas DLA PSS1443$+$2724 shows two clearly distinct (both spatially and spectrally) [C ii] 158 $\mu$m emitters. Our ALMA observations thus corroborate the hypothesis that DLAs with large $\Delta V_{90}$ values, $\gtrsim 200$ km s*-1*, are likely to arise from sightlines that intersect multiple density peaks (see Fig. 4).

4.4 Concluding Remarks

ALMA searches for [C ii] 158 $\mu$m emission from galaxies associated with high-metallicity DLAs are proving to be an efficient way to identify and study this hitherto-elusive high-$z$ galaxy population, allowing us to detect [C ii] 158 $\mu$m emission in galaxies with SFRs as low as 7 M⊙ yr*-1*. Our ALMA observations (including the two systems of Neeleman et al., 2017) have identified the galaxies associated with five DLAs at $z\approx 3.8-4.4$ via their redshifted [C ii] 158 $\mu$m emission. The inferred [C ii] 158 $\mu$m line and FIR continuum luminosities are consistent with the DLA galaxies being similar to luminous LBGs at these redshifts. While this correspondence may not be surprising in hindsight, it may pose a serious challenge to the current paradigm of galaxy formation which predicts that the majority of galaxies at $z\sim 4$ have low SFRs and masses. Our approved Cycle 6 program will complete our survey of $z\sim 4$ DLAs by including lower metallicity systems, and, thereby, establish the properties of the entire host population.

The large impact parameter ($\approx 16-45$ kpc) and high H i column density along the DLA sightline suggest that a large fraction of the H i resides in clumpy regions in the halo of high-$z$ galaxies, away from the bulk of the star formation. The excellent agreement between the absorption and emission redshifts for all five galaxies detected in [C ii] 158 $\mu$m emission disfavors a scenario where the absorbing gas is solely in a companion galaxy as one would then expect velocity offsets between the centroids of the emission and absorption. The relatively high gas metallicity ($\sim 1/10^{\rm th}$ solar) along the DLA sightline indicates that metals must be effectively mixed with the gas and can escape from the main star-forming regions out to large distances. In summary, our ALMA observations suggest that the H i distributions surrounding high-redshift galaxies are markedly different from the H i distributions seen around galaxies in the local Universe.

Finally, we note that all quasars are detected in continuum emission, which is consistent with expectations based on the results of Decarli et al. (2018). In addition, we have detected a bright continuum source close to one of our target quasars, QSO J1253$+$1046. If the source is at the quasar redshift ($z\approx 4.908$), it would have a total infrared luminosity of $\approx 10^{13}$ L⊙ and an SFR of $\approx 1000$ M⊙ yr*-1*, typical of high-$z$ sub-mm galaxies. This would be a second case of a quasar-SMG pair at these redshifts, after BR 1202$-$0725 at $z\approx 4.7$, but at an even smaller transverse separation, of only $\approx 12$ kpc.

We would like to thank the anonymous referee for helpful comments that have improved this manuscript. This paper makes use of the following ALMA data: ADS/JAO.ALMA #2015.0.01564.S and #2016.0.00569.S. ALMA is a partnership of ESO (representing its member states), NSF (USA) and NINS (Japan), together with NRC (Canada), MOST and ASIAA (Taiwan), and KASI (Republic of Korea), in cooperation with the Republic of Chile. The Joint ALMA Observatory is operated by ESO, AUI/NRAO and NAOJ. M.N. acknowledges support from ERC Advanced grant 740246 (Cosmic_Gas). N.K. acknowledges support from the Department of Science and Technology via a Swarnajayanti Fellowship (DST/SJF/PSA-01/2012-13).