1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass{emulateapj}
3:
4: \newcommand{\myemail}{riechers@mpia.de}
5: %\usepackage{amssymb}
6:
7: %% -------Author macros----------------------
8: %% DO NOT USE IN MATH MODE!
9: \def\ts {\thinspace}
10: \def\kms {\ts km\ts s$^{-1}$}
11: \def\etal {{\rm et\ts al.}}
12: \def\msol {M$_{\odot}$}
13: \def\lsol {L$_{\odot}$}
14: \def\ci {C\ts {\scriptsize I}}
15: \def\hi {H\ts {\scriptsize I}}
16: \def\aco {$^{12}${\rm CO}($J$=1$\to$0)}
17: \def\bco {$^{12}${\rm CO}($J$=2$\to$1)}
18: \def\cco {$^{12}${\rm CO}($J$=3$\to$2)}
19: \def\dco {$^{12}${\rm CO}($J$=4$\to$3)}
20: \def\eco {$^{12}${\rm CO}($J$=5$\to$4)}
21: \def\fco {$^{12}${\rm CO}($J$=6$\to$5)}
22: \def\gco {$^{12}${\rm CO}($J$=7$\to$6)}
23: \def\hco {$^{12}${\rm CO}($J$=8$\to$7)}
24: \def\ico {$^{12}${\rm CO}($J$=9$\to$8)}
25: \def\jco {$^{12}${\rm CO}($J$=10$\to$9)}
26: \def\kco {$^{12}${\rm CO}($J$=11$\to$10)}
27: \def\cho {{\rm H$_2$O}(3$_{13}$$\to$2$_{20}$)}
28: \def\ahcn {{\rm HCN}($J$=1$\to$0)}
29: \def\bhcn {{\rm HCN}($J$=2$\to$1)}
30: \def\chcn {{\rm HCN}($J$=3$\to$2)}
31: %%-------------------------------------------
32:
33: \slugcomment{draft version \today, accepted for publication in the Astrophysical Journal}
34:
35: \shorttitle{A search for H$_2$O in MG\,0751+2716}
36: \shortauthors{Riechers et al.}
37:
38: \begin{document}
39:
40: \title{A search for H$_2$O in the strongly lensed QSO MG\,0751+2716 at
41: $z$ = 3.2}
42:
43: \author{Dominik A. Riechers\altaffilmark{1}, Axel Weiss\altaffilmark{2},
44: Fabian Walter\altaffilmark{1}, Christopher L. Carilli\altaffilmark{3},
45: and Kirsten K. Knudsen\altaffilmark{1}}
46:
47:
48: \altaffiltext{1}{Max-Planck-Institut f\"ur Astronomie, K\"onigstuhl
49: 17, Heidelberg, D-69117, Germany}
50:
51: \altaffiltext{2}{Max-Planck-Institut f\"ur Radioastronomie, Auf dem
52: H\"ugel 69, Bonn, D-53121, Germany}
53:
54: \altaffiltext{3}{National Radio Astronomy Observatory, PO Box O,
55: Socorro, NM 87801, USA}
56:
57: \email{riechers@mpia.de}
58:
59:
60: \begin{abstract}
61: We present a search for 183\,GHz \cho\ emission in the
62: infrared--luminous quasar MG\,0751+2716 with the NRAO Very Large
63: Array (VLA). At $z = 3.200 \pm 0.001$, this water emission feature
64: is redshifted to 43.6\,GHz. As opposed to the faint rotational
65: transitions of HCN (the standard high--density tracer at high--$z$),
66: \cho\ is observed with high maser amplification factors in Galactic
67: star--forming regions. It therefore holds the potential to trace
68: high--density star--forming regions in the distant universe. If
69: indeed all star--forming regions in massively star--forming galaxies
70: at $z > 3$ have similar physical properties as e.g.\ the Orion or
71: W49N molecular cloud cores, the flux ratio between the
72: maser--amplified \cho\ and the thermally excited \aco\ transitions
73: may be as high as factor of 20 (but has to be corrected by their
74: relative filling factor). MG\,0751+2716 is a strong \dco\ emitter,
75: and therefore one of the most suitable targets to search for \cho\
76: at cosmological redshifts. Our search resulted in an upper limit in
77: line luminosity of $L'_{\rm H_2O} < 0.6 \times
78: 10^9\,$K\,\kms\,pc$^2$. Assuming a brightness temperature of $T_{\rm
79: b}{\rm (H_2O)} \simeq 500$\,K for the maser emission and CO
80: properties from the literature, this translates to a \cho/\dco\ area
81: filling factor of less than 1\%. However, this limit is not valid if
82: the \cho\ maser emission is quenched, i.e.\ if the line is only
83: thermally excited. We conclude that, if our results were to hold for
84: other high--$z$ sources, H$_2$O does not appear to be a more
85: luminous alternative to HCN to detect high--density gas in
86: star--forming environments at high redshift.
87: \end{abstract}
88:
89: \keywords{galaxies: active, starburst, formation, high redshift --- cosmology: observations
90: --- masers}
91:
92: \section{Introduction}
93:
94: Over the past years, massive amounts of dust and gas have been
95: detected in distant quasars, allowing
96: to study the properties of molecular gas in the early epoch of galaxy
97: formation.
98: Molecular gas in high--redshift galaxies is commonly traced by CO
99: emission and was found in $>$30 galaxies at $z>2$ to date, out to the
100: highest redshift quasar,
101: SDSS\,J1148+5251 at $z = 6.42$ (Walter \etal\ \citeyear{wal03},
102: \citeyear{wal04}; Bertoldi \etal\ \citeyear{ber03}).
103: The observed molecular gas masses in excess of 10$^{10}$\,\msol\
104: provide the requisite material for star formation (SF, e.g.\ Solomon
105: \& Vanden Bout \citeyear{sv05}).
106:
107: The presence of abundant molecular gas, the fuel for SF, has led to
108: the hypothesis that the tremendous far--infrared (FIR) luminosities
109: ($>10^{12}$\,\lsol) of these high--redshift objects are not only
110: powered by active galactic nuclei (AGN) but also by major starbursts
111: (SBs), suggesting that this population represents the formation of
112: large spheroidal galaxies beyond redshift 2 (Blain \etal\
113: \citeyear{bla02}).
114: However, the relative contribution to the FIR luminosity from dust
115: heated by SF and AGN activity in high--$z$ sources remains subject to
116: discussion (e.g.\ Andreani \etal\ \citeyear{and03}).
117:
118: While the lower--$J$ transitions of CO are a good indicator for the
119: total molecular gas content of a system (e.g.\ Carilli \etal\
120: \citeyear{car99}, \citeyear{car02}; Riechers \etal\ \citeyear{rie05}),
121: they can be excited at relatively low densities; the critical density
122: lies at only $n_{\rm H_2} \simeq 10^3\,$cm$^{-3}$. Hence, it is a
123: relatively poor tracer of the denser gas directly involved in massive
124: star formation. The standard tracer of the dense molecular gas phase
125: is HCN, and the critical density to excite its lower--$J$ transitions
126: is $n_{\rm H_2} \simeq 10^5$ cm$^{-3}$ (due to a higher dipole moment
127: in comparison to CO). In this context, recent studies of the dense
128: molecular gas phase in local ($z<0.3$) luminous and ultra--luminous
129: infrared galaxies (LIRGs/ULIRGs) have shown a correlation between the
130: HCN luminosity and the star--formation rate (SFR) as traced by the FIR
131: luminosity. This correlation is much tighter than that between the CO
132: and FIR luminosities (Solomon \etal\ \citeyear{sol92a}; Gao \& Solomon
133: \citeyear{gao04a}, \citeyear{gao04b}).
134: Thus, observations of HCN towards high--$z$ sources hold the potential
135: to pin down the contribution of SF to the total FIR luminosity.
136: Unfortunately, emission lines from HCN in LIRGs/ULIRGs are typically
137: by a factor of 4 to 10 fainter than those of CO (even a factor of
138: 25--40 in ordinary spiral galaxies, Gao \& Solomon \citeyear{gao04b}),
139: which hinders systematic HCN surveys at cosmological distances using
140: current telescopes. Until today, deep observations of HCN at $z>2$
141: resulted in four detections and four upper limits (Solomon \etal\
142: \citeyear{sol03}; Vanden Bout \etal\ \citeyear{vdb04}; Carilli \etal\
143: \citeyear{car04}; Wagg \etal\ \citeyear{wag05}).
144:
145: Due to the faintness of emission connected with thermally excited
146: rotational transitions of HCN, it is important to investigate whether
147: another physical process can be found which produces emission lines
148: with significantly higher luminosities.
149: The 183\,GHz $3_{13} \to 2_{20}$ emission line of para--H$_2$O
150: holds the potential to be such a tracer: in warm, dense SF regions
151: ($n_{\rm H_2} > 10^5$\,cm$^{-3}$), this line is collisionally pumped
152: at relatively low kinetic temperatures ($T_{\rm kin} = 50 - 100\,$K,
153: e.g. Cernicharo \etal\ \citeyear{cer94}). In contrast to compact
154: H$_2$O maser sources at 22\,GHz, the 183\,GHz maser emission is
155: spatially extended in high--mass (Orion, Cernicharo \etal\
156: \citeyear{cer94}; W49N, Gonz{\'a}lez-Alfonso \etal\ \citeyear{gon95};
157: Sgr\,B2, Cernicharo \etal\ \citeyear{cer06}) and even low--mass
158: (HH7--11, Cernicharo \etal\ \citeyear{cer96}) star--forming regions.
159: Due to maser amplification, the observed 183\,GHz \cho\ line
160: brightness temperatures in these regions exceed those observed in
161: \aco\ by up to a factor of 20 (e.g.\ for Orion, Cernicharo \etal\
162: \citeyear{cer94}; Schulz \etal\ \citeyear{sch95}).
163:
164: However, the \cho\ line is strongly absorbed by the terrestrial
165: atmosphere, rendering detection in nearby galaxies very difficult. The
166: first successful extragalactic detection of \cho\ has been reported in
167: NGC\,3079 (Humphreys \etal\ \citeyear{hum05}). Previous searches in
168: low--$z$ (U)LIRGs have only resulted in upper limits (Combes \etal\
169: \citeyear{com97}). Recently, the \cho\ transition has also been
170: detected towards Arp\,220 (Cernicharo, Pardo \& Weiss 2006, in prep.,
171: see Table 1).
172: At high redshift, only
173: the $2_{11} \to 2_{02}$ transition of para--H$_2$O at 752.033\,GHz
174: (rest frame) has tentatively been detected in the $z \simeq 2.3$ QSO
175: IRAS\,F10214+4724 (Encrenaz \etal\ \citeyear{enc93}; Casoli \etal\
176: \citeyear{cas94}).
177:
178: \indent In this paper, we report on a search for maser emission from
179: the 183\,GHz water line in the strongly lensed, radio--loud $z = 3.2$
180: QSO MG\,0751+2716 using the VLA\footnote{The Very Large Array is a
181: facility of the National Radio Astronomy Observatory, operated by
182: Associated Universities, Inc., under a cooperative agreement with
183: the National Science Foundation.}.
184: Due to its strong magnification, MG\,0751+2716 is the brightest CO
185: source at high redshift that can be observed in the \cho\ transition
186: with the VLA. Its $L'_{\rm CO}$ and $L_{\rm FIR}$ (see Table 1) are
187: comparable to local starburst galaxies like Arp\,220. This suggests
188: that this source is undergoing massive star--formation, which
189: contributes significantly to its far--IR luminosity (e.g.\ Fig.\ 8 in
190: Solomon \& Vanden Bout \citeyear{sv05}).
191: We use a standard concordance cosmology throughout, with $H_0 =
192: 71\,$\kms\,Mpc$^{-1}$, $\Omega_{\rm m} =0.27$, and $\Omega_{\Lambda} =
193: 0.73$ (Spergel \etal\ \citeyear{spe03}).
194:
195: \section{Observations}
196:
197: We observed the \cho\ transition ($\nu_{\rm rest} = 183.3101\,$GHz)
198: towards MG\,0751+2716 using the VLA in D configuration on 2004 June 19
199: and 27. At the target $z$ of 3.200, this transition is redshifted to
200: 43.6453\,GHz (6.87\,mm). The total on--sky integration time amounts to
201: 16\,hr. Observations were performed in fast--switching mode using the
202: nearby source 0745+241 for secondary amplitude and phase calibration.
203: Observations were carried out under good weather conditions with 26
204: antennas. The phase stability in all runs was good (typically
205: $<$25$^\circ$ rms for the longest baselines). For primary flux
206: calibration, 3C286 was observed during each run.
207:
208: Two 50\,MHz (corresponding to 344\kms\ at 43.6\,GHz) intermediate
209: frequencies (IFs) were observed simultaneously in the so--called
210: 'quasi--continuum' mode.
211: One IF was centered on the line frequency at 43.6351\,GHz (closest
212: possible tuning frequency, offset by $\sim$70\kms\ from the \dco\ line
213: center of Barvainis \etal\ \citeyear{bar02}), and the second IF was
214: centered at two frequencies symmetrically offset by 150\,MHz
215: (1030\kms, i.e.\ at 43.4851\,GHz and 43.7851\,GHz) from the line
216: frequency to monitor the 7\,mm continuum of MG\,0751+2716
217: simultaneously. Despite the small offset from the line center, the
218: frequency/velocity coverage is well--matched to a line with a FWHM
219: similar to the \dco\ line reported by Barvainis \etal\
220: (\citeyear{bar02}).
221:
222: For data reduction and analysis, the
223: ${\mathcal{AIPS}}$\footnote{www.aoc.nrao.edu/aips/} package was used.
224: The two continuum channels were concatenated in the uv/visibility
225: plane. The data were mapped using the CLEAN algorithm and 'natural'
226: weighting; this results in a synthesized beam of
227: 1.8\,$''$$\times$1.7\,$''$ ($\sim$13\,kpc at $z = 3.2$) at a major
228: axis position angle of 78$^\circ$. The final rms in both the line
229: channel and the combined continuum channel is 60\,$\mu$Jy beam$^{-1}$.
230: \section{Results}
231:
232: In the top panel of Fig.\ 1, the final map of the 'ON' channel
233: (line+continuum emission) is shown, while the middle panel shows the
234: map of the two combined continuum ('OFF') channels, representing the
235: continuum--only flux. The source is clearly detected in both maps. To
236: derive the total flux density of the source, a 2--dimensional Gaussian
237: was fitted to the detected source structure. Our best fit for
238: MG\,0751+2716 gives a Gaussian diameter of 2.3\,$''$$\times$2.3\,$''$
239: (source convolved with the synthesized beam); i.e.\ the emission
240: appears only marginally resolved in our observations, as our beam is
241: too large to recover the lens image substructure seen at higher
242: resolution (Carilli \etal\ \citeyear{car04}). The integrated flux
243: density in the 'ON' channel is 10.3 $\pm$ 0.2\,mJy (peak: 6.12 $\pm$
244: 0.06\,mJy beam$^{-1}$), and the continuum flux in the combined 'OFF'
245: channels is 10.3 $\pm$ 0.2\,mJy (peak: 5.88 $\pm$ 0.06\,mJy
246: beam$^{-1}$). From the VLA 42.2\,GHz measurement of 13.2\,mJy (Carilli
247: \etal\ \citeyear{car04}) and using a spectral index of $\beta = -1.2$
248: (Leh{\'a}r \etal\ \citeyear{leh97}), an extrapolated flux density of
249: 12.7\,mJy can be calculated, which is higher than derived from our
250: observations. Therefore, continuum variability cannot be excluded.
251:
252: In the bottom panel of Fig.\ 1, the continuum-subtracted 'line' map is
253: shown. This map was generated by subtracting a CLEAN component model
254: of the continuum emission (combined 'OFF' channel) from the visibility
255: data of the 'ON' channel, which was then imaged applying the CLEAN
256: algorithm with the same parameters used to create the continuum model.
257: No clear evidence for \cho\ line emission is found. Thus, we set a
258: 3$\sigma$ upper limit of $S_{\rm H_2O} < 180\,\mu$Jy on the peak flux
259: of \cho\ emission. For MG\,0751+2716, we thus derive an upper limit
260: for the \cho\ line luminosity of $L'_{\rm H_2O} < 0.6 \times
261: 10^9\,$K\,\kms\,pc$^2$ (3$\sigma$ limit, corrected for gravitational
262: magnification, $\mu_{\rm L} = 17$, Barvainis \etal\ \citeyear{bar02},
263: cf.\ Table \ref{tab-1}).
264:
265: \section{Discussion}
266:
267: \subsection{Intrinsic line brightness temperatures} \label{temp}
268:
269: Given that the 183\,GHz water maser line most likely arises from
270: extended regions associated with star formation (Cernicharo \etal\
271: \citeyear{cer94}), we can use our upper limit together with the \dco\
272: line luminosity (Barvainis \etal\ \citeyear{bar02}) to estimate how
273: much of the molecular gas in MG\,0751+2716 may be associated with
274: massive star forming regions. Radiative transfer models based on the
275: \cho\ emission line show that it is inverted in warm and dense
276: environments ($T_{\rm kin}>40$\,K, $n_{\rm H_2} > 5 \times
277: 10^{4}$\,cm$^{-3}$) which are typically found in star--forming
278: regions, and that the resulting line brightness temperature critically
279: depends on the underlying physical conditions. It is known to range
280: from thermalized emission up to line temperatures of $\sim$10000\,K
281: (Cernicharo \etal\ \citeyear{cer94}; Combes \etal\ \citeyear{com97}).
282: As an example, observations of the 183\,GHz water maser towards Orion
283: IRc2 show that emission arises from narrow (few \kms) line features
284: with up to $\simeq 2000$\,K as well as a broad ($\simeq 200$\,\kms)
285: emission plateau with $\simeq 500$\,K peak line brightness temperature
286: (Cernicharo \etal\ \citeyear{cer94}, \citeyear{cer99}). As an order of
287: magnitude estimate, we assume in the following that the \cho\ line in
288: MG\,0751+2716 arises from a similar environment as Orion with an
289: intrinsic brightness temperature of $T_{\rm b}{\rm (H_2O)} \simeq
290: 500$\,K. Only the \dco\ line is observed so far in MG\,0751+2716,
291: therefore the brightness temperature of the CO emission is not
292: well--constrained. However, the dust continuum temperature of $T_{\rm
293: dust}\simeq 40$\,K in MG\,0751+2716 (Barvainis \& Ivison
294: \citeyear{bi02}) is consistent with those observed in most other
295: high--$z$ QSOs (e.g.\ Beelen et al.\ \citeyear{bee06}). In addition,
296: the CO brightness temperature in well--studied high--$z$ objects such
297: as IRAS\,F10214+4724 agrees within a factor of $\sim 2$ with the dust
298: temperature (e.g. Downes \etal\ \citeyear{dow95}). By analogy, we here
299: assume $T_{\rm b}{\rm (CO)}\simeq 40$\,K.
300:
301: \subsection{Area filling factor} \label{filf}
302:
303: The combined ratios of the brightness temperatures and line
304: luminosities of H$_2$O and CO ($T_{\rm b}{\rm (H_2O)} \simeq 500$\,K,
305: $T_{\rm b}{\rm (CO)}\simeq 40$\,K, $L'_{\rm H_2O} < 0.6 \times
306: 10^9\,$K\,\kms\,pc$^2$, $L'_{\rm CO} = 10.0 \times
307: 10^9\,$K\,\kms\,pc$^2$) can be used to estimate the relative area
308: filling factor (FF) of both molecules. By this means, our observed
309: upper limit on the 183\,GHz H$_2$O line luminosity together with the
310: \dco\ line luminosity translates into an upper limit on the H$_2$O/CO
311: area filling factor of $<1\%$.
312:
313: To better understand the relevance of this filling factor limit
314: derived from our observations, we now give an independent estimate for
315: the filling factor based on the FIR luminosity and geometrical
316: arguments. For this purpose, we assume in the following that all
317: 183\,GHz H$_2$O emission in MG\,0751+2716 is created in hot,
318: star--forming cores like the one found in the central region of Orion.
319: Assuming $L_{\rm FIR} = 1.2 \times 10^{5}$\,\lsol\ for the central
320: 1\,$\square '$ of the Orion Nebula (Werner \etal\ \citeyear{wer76},
321: corresponding to an area of 0.013\,pc$^2$ at a distance of 450\,pc),
322: we find that 10$^7$ of these cores are needed to account for the
323: far--IR luminosity in MG\,0751+2716 (Table 1). Assuming a size of
324: 0.013\,pc$^2$ each, 10$^7$ of such cores would fill a disk with an
325: equivalent radius of 200\,pc. Taking the observed \dco\ line FWHM and
326: luminosity into account, and assuming $T_{\rm b}{\rm (CO)}\simeq
327: 40$\,K, for MG\,0751+2716, we can calculate that the {\em dust and CO}
328: are distributed over a region with an equivalent radius of 450\,pc.
329: Comparing the derived disk sizes, we thus find an expected H$_2$O/CO
330: area filling factor of $\sim$20\%. This estimate is much higher than
331: what is actually observed.
332:
333: \subsection{Comparison with other (U)LIRGs and NGC\,3079} \label{comp}
334:
335: Our results for MG\,0751+2716 are in line with observations of the
336: radio--quiet QSOs Mkn\,1014 (PG\,0157+001, $z=0.16$) and VII\,Zw\,244
337: (PG\,0838+770, $z=0.13$, see Table 1), the only two other
338: infrared--luminous galaxies for which upper limits on the 183\,GHz
339: H$_2$O line have been reported (Combes \etal\ \citeyear{com97}) to
340: date. Using the same assumptions on the intrinsic line brightness
341: temperatures as for MG\,0751+2716, the non--detections in these
342: galaxies imply filling factors of the water line relative to \aco\ of
343: $<5\%$ and $<2\%$ for Mkn\,1014 and VII\,Zw\,244 respectively. As CO
344: emission appears to be close to thermalized up to the \dco\ transition
345: in high--$z$ QSOs (Weiss et al.\ \citeyear{wei05}), these limits are
346: directly comparable to MG\,0751+2716.
347:
348: In Arp\,220 ($z=0.018$), the \cho\ line was detected with a line
349: luminosity of $L'_{\rm H_2O} = 2.9 \times 10^8\,$K\,\kms\,pc$^2$
350: (Cernicharo, Pardo \& Weiss 2006, in prep.). With the CO luminosity
351: given in Solomon \etal\ (\citeyear{sol92a}), this translates to a
352: H$_2$O/CO filling factor of 0.4\%. Arp\,220 has a far--IR luminosity
353: similar to that of MG\,0751+2716, which might suggest a similar
354: H$_2$O/CO filling factor.
355: This suggests that the \cho\ line might be detectable in MG\,0751+2716
356: if the sensitivity was increased only by a factor of few.
357:
358: For the recent first extragalactic \cho\ detection in the LINER
359: NGC\,3079 (Humphreys \etal\ \citeyear{hum05}), the H$_2$O/CO
360: luminosity ratio (using same brightness temperatures as before)
361: translates into a H$_2$O/CO area filling factor of only $0.05\%$.
362: However, it is difficult to assess the relevance of the NGC\,3079
363: results for MG\,0751+2716, as the former is not a ULIRG. For
364: lower--luminosity galaxies, it has been found that the HCN/CO
365: luminosity ratio is significantly lower (2.5--4\% rather than
366: 10--25\%, Gao \& Solomon \citeyear{gao04b}); by analogy, also lower
367: H$_2$O/CO luminosity ratios may be expected in galaxies with FIR
368: luminosities that are significantly lower than in a ULIRG.
369:
370:
371: \subsection{Possible interpretations of the non--detection} \label{conc}
372:
373: Our observations suggest that the H$_2$O/CO area filling factor in
374: MG\,0751+2716 is significantly lower than expected from simple
375: estimates based on the distribution and temperature of the dust and
376: CO.
377:
378: However, other reasons for the weakness of the water emission should
379: also be considered: as discussed in Combes \etal\ (\citeyear{com97}),
380: the physical conditions to efficiently pump the 183\,GHz line to the
381: assumed 500\,K may not be present. There are strong differences in
382: the conditions for H$_2$O line emission among Galactic star--forming
383: regions. E.g., the \cho\ maser emission in Sgr\,B2(N/M) is much
384: weaker than that in Orion\,IRc2. The extent of the maser emitting
385: regions in both sources is of order 1\,pc, but the observed brightness
386: temperature in Sgr\,B2 is by 1--2 orders of magnitude lower, reaching
387: only few tens of Kelvin
388: (Cernicharo \etal\ \citeyear{cer94}, \citeyear{cer06}).
389: This is much lower than the 500\,K we assumed for a typical H$_2$O
390: emitting region. The 183\,GHz H$_2$O lines are thus only by a factor
391: of 3 amplified relative to CO in this source. For high densities
392: and/or column densities, which are known to be present in the compact
393: molecular distributions of luminous IR galaxies, the maser emission of
394: the 183\,GHz line may even be quenched. This would lead to thermalized
395: emission without maser amplification (Cernicharo \etal\
396: \citeyear{cer94}; Combes \etal\ \citeyear{com97}). For thermalized
397: emission in an optically thick environment, the \cho\ line is not
398: expected to be stronger than those of other high density gas tracers
399: such as HCN. In Arp\,220, the HCN luminosity is $L'_{\rm HCN} = 9.4
400: \times 10^8\,$K\,\kms\,pc$^2$, i.e.\ by a factor of 3 brighter than
401: \cho\ (Cernicharo, Pardo \& Weiss 2006, in prep.; Graci\'a-Carpio
402: \etal\ \citeyear{gra06}). The H$_2$O and HCN luminosities in Arp\,220
403: would thus even be consistent with $T_{\rm b}{\rm (H_2O)} = T_{\rm
404: b}{\rm (HCN)}$ and an area filling factor that is smaller for H$_2$O
405: than for HCN. This suggests that only a small fraction of the dense
406: gas is giving rise to H$_2$O emission.
407: Carilli \etal\ (\citeyear{car04}) set a 3$\sigma$ limit of $L'_{\rm
408: HCN} < 1.0 \times 10^9\,$K\,\kms\,pc$^2$ on the \bhcn\ line
409: luminosity in MG\,0751+2716, which is twice as high as our limit for
410: $L'_{\rm H_2O}$.
411:
412: It also remains a possibility that the reason for our non--detection
413: is due to limitations in our observing mode. Due to the small
414: fraction of gas giving rise to the \cho\ emission compared to CO, its
415: linewidth could be much smaller than that of the \dco\ transition.
416: This has recently been found in NGC\,3079 (Humphreys \etal\
417: \citeyear{hum05}), where the difference in linewidth is more than a
418: factor of 10. As our observations had to be done using 50\,MHz
419: (344\,\kms) channels, the emission from a narrow line would be diluted
420: over the full velocity range covered by that channel, rendering
421: detection unlikely even at the achieved sensitivity. However, in
422: Arp\,220, which is likely more similar to MG\,0751+2716, the \cho\
423: line has a width of 350\,\kms. Such a linewidth would match our
424: observing mode very well.
425: Finally, differential lensing may have an impact on the measured
426: luminosity ratios ($\mu_{\rm L} = 17$), in particular if the CO and
427: H$_2$O emission do not emerge from the same regions.
428:
429: Given the aforementioned results, we find that water maser activity
430: does not outshine emission from rotational transitions of CO or HCN in
431: the $z=3.2$ QSO MG\,0751+2716. If our results can be generalized, we
432: conclude that H$_2$O (despite the fact that it potentially traces hot,
433: star--forming cores) does not appear to be a more luminous alternative
434: to HCN to detect high--density gas in star--forming environments at
435: high redshift.
436:
437: \acknowledgments
438: The National Radio Astronomy Observatory is operated by Associated
439: Universities Inc., under cooperative agreement with the National
440: Science Foundation. D.~ R.\ acknowledges support from the Deutsche
441: Forschungsgemeinschaft (DFG) Priority Programme 1177. C.~C.\
442: acknowledges support from the Max-Planck-Gesellschaft and the
443: Alexander von Humboldt-Stiftung through the
444: Max-Planck-Forschungspreis. The authors would like to thank the
445: referee for useful comments which helped to improve the manuscript.
446:
447: %\clearpage
448:
449: \begin{thebibliography}{}
450:
451: \bibitem[Alloin et al. (1992)]{all92} Alloin, D., Barvainis, R.,
452: Gordon, M.~A., \& Antonucci, R.~R.~J.\ 1992, A\&A, 265, 429
453: \bibitem[Andreani et al.(2003)]{and03} Andreani, P., Christiani, S.,
454: Grazian, A., LaFranca, F., \& Goldschmidt, P.\ 2003, AJ. 125, 444
455: \bibitem[Barvainis \& Ivison(2002)]{bi02} Barvainis, R., \& Ivison,
456: R.\ 2002, ApJ, 571, 712
457: \bibitem[Barvainis et al.(2002)]{bar02} Barvainis, R., Alloin, D., \&
458: Bremer, M.\ 2002, A\&A, 385, 399
459: \bibitem[Beelen et al.(2006)]{bee06} Beelen, A., Cox, P., Benford,
460: D.~J., \etal\ 2006, ApJ, in press (astro-ph/0603121)
461: \bibitem[Bertoldi et al.(2003)]{ber03} Bertoldi, F., Cox, P., Neri,
462: R., \etal\ 2003, A\&A, 409, L47
463: \bibitem[Blain et al.(2002)]{bla02} Blain, A., Smail, I., Ivison, R.,
464: Kneib, J.-P., \& Frayer, D.\ 2002, Phys.\ Rep., 369, 111
465: \bibitem[Carilli et al.(1999)]{car99} Carilli, C.~L., Menten, K.~M.,
466: \& Yun, M.~S.\ 1999, ApJ, 521, L25
467: \bibitem[Carilli et al.(2002)]{car02} Carilli, C.~L.,Kohno, K.,
468: Kawabe, R., Ohta, K., \etal\ 2002, AJ, 123, 1838
469: \bibitem[Carilli et al.(2005)]{car04} Carilli, C.~L., Solomon, P.~M.,
470: Vanden Bout, P.~A., \etal\ 2005, ApJ, 618, 586
471: \bibitem[Casoli et al.(1994)]{cas94} Casoli, F., Gerin, M., Encrenaz,
472: P.~J., \& Combes, F.\ 1994, A\&A, 287, 716
473: \bibitem[Cernicharo et al.(1994)]{cer94} Cernicharo, J.,
474: Gonz{\'a}lez-Alfonso, E., Alcolea, J., \etal\ 1994, ApJ, 432, L59
475: \bibitem[Cernicharo et al.(1996)]{cer96} Cernicharo, J., Bachiller,
476: R., Gonz{\'a}lez-Alfonso, E., \etal\ 1996, A\&A, 305, L5
477: \bibitem[Cernicharo et al.(1999)]{cer99} Cernicharo, J., Pardo, J.~R.,
478: Gonz{\'a}lez-Alfonso, E., \etal\ 1999, ApJ, 520, L131
479: \bibitem[Cernicharo et al.(2006)]{cer06} Cernicharo, J., Goicoechea,
480: J.~R., Pardo, J.~R., \& Asensio-Ramos, A.\ 2006, ApJ, in press
481: \bibitem[Combes et al.(1997)]{com97} Combes, F., Nguyen-Q-Rieu, \&
482: Dinh-V-Trung 1997, A\&A, 323, 357
483: \bibitem[Downes et al.(1995)]{dow95} Downes, D., Solomon, P.~M., \&
484: Radford, S.~J.~E.\ 1995, ApJ, 453, L65
485: \bibitem[Encrenaz et al.(1993)]{enc93} Encrenaz, P.~J., Combes, F.,
486: Casoli, F., \etal\ 1993, A\&A, 273, L19
487: \bibitem[Gao \& Solomon(2004a)]{gao04a} Gao, Y., \& Solomon, P.~M.\
488: 2004a, ApJS, 152, 63
489: \bibitem[Gao \& Solomon(2004b)]{gao04b} Gao, Y., \& Solomon, P.~M.\
490: 2004b, ApJ, 606, 271
491: \bibitem[Gonz{\'a}lez-Alfonso et al.(1995)]{gon95}
492: Gonz{\'a}lez-Alfonso, E., Cernicharo, J., Bachiller, R., \& Fuente,
493: A.\ 1995, A\&A, 293, L9
494: \bibitem[Graci\'a--Carpio et al.(2006)]{gra06} Graci\'a-Carpio, J.,
495: Garc\'ia-Burillo, S., Planesas, P., \& Colina, L.\ 2006, ApJ, 640,
496: L135
497: \bibitem[Humphreys et al.(2005)]{hum05} Humphreys, E.~M.~L.,
498: Greenhill, L.~J., Reid, M.~J., \etal\ 2005, ApJ, 634, L133
499: \bibitem[Kennicutt(1998)]{ken98} Kennicutt, R.\ 1998, ARA\&A, 36, 189
500: \bibitem[Koda et al.(2002)]{kod02} Koda, J., Sofue, Y., Kohno, K.,
501: \etal\ 2002, ApJ, 573, 105
502: \bibitem[Leh{\'a}r et al.(1997)]{leh97} Leh{\'a}r, J., Burke, B.~F.,
503: Conner, S.~R., \etal\ 1997, AJ, 114, 68
504: \bibitem[Riechers et al.(2006)]{rie05} Riechers, D.~A., Walter, F.,
505: Carilli, C.~L., \etal\ 2006, ApJ, submitted
506: \bibitem[Schulz et al.(1995)]{sch95} Schulz, A., Henkel, C., Beckmann,
507: U., \etal\ 1995, A\&A, 295, 183
508: \bibitem[Solomon et al.(1992a)]{sol92a} Solomon, P., Downes, D., \&
509: Radford, S.\ 1992a, ApJ, 387, L55
510: \bibitem[Solomon et al.(1992b)]{sol92b} Solomon, P.~M., Radford,
511: S.~J.~E., \& Downes, D.\ 1992b, Nature, 356, 318
512: \bibitem[Solomon et al.(1997)]{sol97} Solomon, P.~M., Downes, D.,
513: Radford, S., \& Barrett, J.~W., 1997, ApJ, 478, 144
514: \bibitem[Solomon et al.(2003)]{sol03} Solomon, P., Vanden Bout, P.,
515: Carilli, C., \& Guelin, M.\ 2003, Nature 426, 636
516: \bibitem[Solomon \& Vanden Bout(2005)]{sv05} Solomon, P.~M., \& Vanden
517: Bout, P.~A.\ 2005, ARA\&A, 43, 677
518: \bibitem[Spergel et al.(2003)]{spe03} Spergel, D.~N., Verde, L.,
519: Peiris, H.~V., \etal\ 2003, ApJS, 148, 175
520: \bibitem[Vanden Bout et al.(2004)]{vdb04} Vanden Bout, P., Solomon,
521: P., \& Maddalena, R.\ 2004, ApJ, 614, L97
522: \bibitem[Wagg et al.(2005)]{wag05} Wagg, J., Wilner, D.~J., Neri, R.,
523: Downes, D., \& Wiklind, T.\ 2005, ApJ, 634, L13
524: \bibitem[Walter et al.(2003)]{wal03} Walter, F., Bertoldi, F.,
525: Carilli, C.~L., \etal\ 2003, Nature, 424, 406
526: \bibitem[Walter et al.(2004)]{wal04} Walter, F., Carilli, C.,
527: Bertoldi, F., \etal\ 2004, ApJ, 615, L17
528: \bibitem[Weiss et al.(2005)]{wei05} Wei\ss, A., Downes, D., Walter,
529: F., \& Henkel, C.\ 2005, A\&A, 440, L45
530: \bibitem[Werner et al.(1976)]{wer76} Werner, M.~W., Gatley, I.,
531: Harper, D.~A., \etal\ 1976, ApJ, 204, 420
532:
533: \end{thebibliography}
534:
535: %\clearpage
536:
537: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
538: %%%% Tab.1: Previous detections of high-z water
539: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
540: \begin{deluxetable}{ l c c c c c c c c c }
541: %\rotate
542: \tabletypesize{\scriptsize}
543: \tablecaption{Extragalactic para--\cho\ emission: fluxes, luminosities, and area filling factors.\label{tab-1}}
544: \tablehead{
545: & $z$ & $D_{\rm L}$& $\nu_{\rm obs}$ & $S_{\rm H_2O}$ & $L'_{\rm CO}$\tablenotemark{a} & $L'_{\rm H_2O}$\tablenotemark{a} & $L_{\rm FIR}$ & FF & Ref. \\
546: & & [Mpc] & [GHz] & [mJy] & [10$^9$\,K\,\kms\,pc$^2$] & [10$^9$\,K\,\kms\,pc$^2$] & [10$^{12}\,$\lsol] & & }
547: \startdata
548: MG\,0751+2716\tablenotemark{b} & 3.200 & 27940 & 43.6453 & $<$0.18 & 10.0 & $<0.6$ & 1.2 & $<$1\% & 1,2,3 \\
549: Mkn\,1014 & 0.1631 & 774 & 157.6048 & $<$26 & 7.8 & $<2.9$ & 2.2 & $<$5\% & 4,5 \\
550: VII\,Zw\,244 & 0.1324 & 616 & 161.8775 & $<$26 & 3.9 & $<0.7$ & 0.14 & $<$2\% & 4,5 \\
551: \tableline
552: NGC\,3079 & 0.003723 & 16 & 182.6302 & 550 & 1.1 & 0.0041 & 0.021 & 0.05\% & 6,7 \\
553: Arp\,220 & 0.018126 & 78 & 180.0680 & 170 & 5.9 & 0.29 & 1.3 & 0.4\% & 8,9,10 \\
554: \vspace{-3mm}
555: \enddata
556: \tablenotetext{a}{Derived as described by Solomon \etal\ (\citeyear{sol92b}):
557: $L'_{\rm X}[{\rm K \ts km\ts s^{-1} pc^2}] = 3.25 \times 10^7 \times I \times \nu_{\rm obs}^{-2} \times D_{\rm L}^2 \times (1+z)^{-3}$,
558: where X is the molecule, $I$ is the velocity--integrated line flux in Jy \kms, $D_{\rm L}$ is the luminosity
559: distance in Mpc, and $\nu_{\rm obs}$ is the observed frequency in GHz.}
560: \tablenotetext{b}{This QSO is lensed by a factor of $\mu_{\rm L} = 17$ (Barvainis \etal\ \citeyear{bar02}). All given luminosities are corrected for lensing. }
561: \tablerefs{${}$[1] This work, [2] Barvainis \etal\ (\citeyear{bar02}), [3] Carilli \etal\ (\citeyear{car04}),
562: [4] Combes \etal\ (\citeyear{com97}), [5] Alloin \etal\ (\citeyear{all92}),
563: [6] Koda \etal\ (\citeyear{kod02}), [7] Humphreys \etal\ (\citeyear{hum05}),
564: [8] Cernicharo, Pardo \& Weiss (2006), in prep., [9] Solomon \etal\ (\citeyear{sol92a}), [10] Solomon \etal\ (\citeyear{sol97}).}
565: \tablecomments{For MG\,0751+2716, the \dco\ FWHM linewidth
566: (Barvainis \etal\ \citeyear{bar02}, corrected for the VLA bandpass) of 350\,\kms\ is utilized to derive $L'_{\rm H_2O}$, whereas the
567: \aco\ FWHM linewidths of 210\,\kms\ and 80\,\kms\ (Combes \etal\ \citeyear{com97}) are assumed for Mkn\,1014 and VII\,Zw\,244.}
568: \end{deluxetable}
569:
570: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
571:
572: %\clearpage
573:
574: \begin{figure}
575: \epsscale{.5}
576: \plotone{f1.eps}
577: \caption{VLA observations of MG\,0751+2716 at a resolution of 1.8\,$''$$\times$1.7\,$''$ at a major axis position angle of
578: 78$^\circ$. The circle in the bottom left corner of each map represents the FWHM of the restoring CLEAN beam.
579: \textbf{Top}: Map of the central 50\,MHz (344\kms) 'ON' channel at 43.6\,GHz (183.3\,GHz rest--frame),
580: the frequency for which the H$_2$O line is expected. \textbf{Middle}:
581: Combined map of the two 50\,MHz 'OFF' (continuum only) channels symmetrically offset by 150\,MHz (1030\kms) each from
582: the H$_2$O line frequency. The source is marginally resolved in both maps.
583: \textbf{Bottom}: Continuum--subtracted ('ON--OFF') map. No clear evidence for \cho\ line emission is found
584: within the uncertainties of the observations.
585: All three maps are shown with (-4, -3, 3, 4, 8, 16, 32, 48, 64, 80, 96)$\times 60\,\mu$Jy beam$^{-1}$ contours.
586: \label{f1}}
587: \end{figure}
588: \end{document}
589: