1: \documentclass[preprint2]{aastex}
2: %\documentclass[12pt,preprint]{aastex}
3: \shorttitle{NGC 1333 IRAS 4A Disks}
4: \shortauthors{Choi et al.}
5: \received{}
6:
7: % For preprint only
8: \usepackage{times}
9: \frenchspacing
10:
11: % Custom macros
12: \newcommand{\Jypb}{{Jy beam$^{-1}$}}
13: \newcommand{\kms}{\mbox{km s$^{-1}$}}
14: \newcommand{\Msun}{{$M_\odot$}}
15: \newcommand{\NHt}{{\rm NH$_3$}}
16: \newcommand{\Vlsr}{{$V_{\rm LSR}$}}
17: \usepackage{epsfig}
18: \newcommand{\ploteps}[2]{\centerline{\ifdim#2=0mm\epsfig{file=#1}%
19: \else\epsfig{file=#1,width=#2}\fi}}
20:
21:
22: \slugcomment{To appear in the Astrophysical Journal}
23:
24: \begin{document}
25: \fontsize{10}{10.6}\selectfont
26: \title{Ammonia Imaging of the Disks
27: in the NGC 1333 IRAS 4A Protobinary System}
28: \author{\sc Minho Choi\altaffilmark{1},
29: Ken'ichi Tatematsu\altaffilmark{2},
30: Geumsook Park\altaffilmark{1,3},
31: and Miju Kang\altaffilmark{1,3}}
32: \altaffiltext{1}{International Center for Astrophysics,
33: Korea Astronomy and Space Science Institute,
34: Hwaam 61-1, Yuseong, Daejeon 305-348, South Korea;
35: minho@kasi.re.kr.}
36: \altaffiltext{2}{National Astronomical Observatory of Japan,
37: 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan.}
38: \altaffiltext{3}{Department of Astronomy and Space Science,
39: Chungnam National University, Yuseong, Daejeon 305-764,
40: South Korea.}
41: \setcounter{footnote}{3}
42:
43: \begin{abstract}
44: \fontsize{10}{10.6}\selectfont
45: The NGC 1333 IRAS 4A protobinary was observed
46: in the ammonia (2, 2) and (3, 3) lines and in the 1.3 cm continuum
47: with a high resolution (about 1.0$''$).
48: The ammonia maps show two compact sources, one for each protostar,
49: and they are probably protostellar accretion disks.
50: The disk associated with IRAS 4A2 is seen nearly edge-on
51: and shows an indication of rotation.
52: The A2 disk is brighter in the ammonia lines
53: but dimmer in the dust continuum than its sibling disk,
54: with the ammonia-to-dust flux ratios
55: different by about an order of magnitude.
56: This difference suggests
57: that the twin disks have surprisingly dissimilar characters,
58: one gas-rich and the other dusty.
59: The A2 disk may be unusually active or hot,
60: as indicated by its association with water vapor masers.
61: The existence of two very dissimilar disks in a binary system suggests
62: that the formation process of multiple systems
63: has a controlling agent lacking in the isolated star formation process
64: and that stars belonging to a multiple system
65: do not necessarily evolve in phase with each other.
66: \end{abstract}
67:
68: \keywords{accretion disks --- binaries: general
69: --- ISM: individual (NGC 1333 IRAS 4A) --- stars: formation}
70:
71: \section{INTRODUCTION}
72:
73: The NGC 1333 region in Perseus is a site of active star formation
74: producing a young cluster of Sun-like low-mass stars
75: (Aspin et al. 1994; Bally et al. 1996; Rodr{\'\i}guez et al. 1999).
76: IRAS 4A is one of the brightest submillimeter sources
77: among the deeply embedded protostars in NGC 1333 (Sandell \& Knee 2001).
78: The cold spectrum of dust emission suggests that IRAS 4A is very young
79: (Sandell et al. 1991),
80: and IRAS 4A drives molecular outflows with an interesting morphology
81: (Blake et al. 1995; Girart et al. 1999; Choi 2001).
82: Radio interferometric observations revealed
83: that there are two sources in IRAS 4A,
84: probably gravitationally bound to each other
85: (Lay et al. 1995; Looney et al. 2000).
86:
87: In all the continuum images published so far,
88: the primary, A1, is more luminous than the secondary, A2,
89: by a factor of $\sim$3
90: (Looney et al. 2000; Reipurth et al. 2002; Girart et al. 2006),
91: which suggests that A1 may be more massive than A2.
92: However, recent observations in the SiO and the H$_2$ lines
93: puzzlingly showed
94: that the outflow driven by A2 (northeast-southwestern outflow)
95: is much more powerful and larger
96: than the one driven by A1 (southern outflow) (Choi 2005; Choi et al. 2006).
97: This inverse relation between the (dust) mass and the outflow activity
98: challenges conventional models of star formation
99: since both of them are supposed to be positively correlated
100: with the mass accretion rate (Cabrit \& Andr{\'e} 1991; Bachiller 1996).
101:
102: To investigate the structure of the NGC 1333 IRAS 4A system,
103: we observed the region in several molecular tracers.
104: In this Letter, we present our high angular resolution observations
105: in the \NHt\ lines.
106: In \S~2 we describe our \NHt\ observations.
107: In \S~3 we report the results of the \NHt\ imaging
108: and discuss the physical process in the IRAS 4A system.
109:
110: \section{OBSERVATIONS}
111:
112: The NGC 1333 IRAS 4 region was observed using the Very Large Array (VLA)
113: of the National Radio Astronomy Observatory\footnote{
114: The NRAO is a facility of the National Science Foundation
115: operated under cooperative agreement by Associated Universities, Inc.}
116: in the \NHt\ (2, 2) and (3, 3) lines
117: (23722.6336 and 23870.1296 GHz, respectively)
118: and in the $\lambda$ = 1.3 cm continuum.
119: Twenty-five antennas were used
120: in the C-array configuration on 2004 March 5.
121: The continuum was observed for 20 minutes at the beginning
122: and for 10 minutes at the end of the observing track,
123: and the \NHt\ lines were observed
124: for 5 hours in the midsection of the track.
125: For each of the \NHt\ lines,
126: the spectral windows were set to have 64 channels
127: with a channel width of 0.049 MHz,
128: giving a velocity resolution of 0.62 \kms.
129: For the 1.3 cm continuum,
130: the observations were made in the standard $K$-band continuum mode
131: (22.5 GHz or $\lambda$ = 1.33 cm).
132:
133: The phase tracking center was ($\alpha$, $\delta$)
134: = (03$^{\mathrm h}$29$^{\mathrm m}$10.41$^{\mathrm s}$,
135: 31\arcdeg13$'$32.2$''$) in J2000.0.
136: The nearby quasar 0336+323 (PKS 0333+321) was observed
137: to determine the phase and to obtain the bandpass response.
138: The flux was calibrated by observing the quasar 0713+438 (QSO B0710+439)
139: and by setting its flux density to 0.49 Jy,
140: which is the average of the flux density
141: measured within a day of our observations (VLA Calibrator Flux Density
142: Database\footnote{See http://aips2.nrao.edu/vla/calflux.html.}).
143: Comparison of the amplitude gave a flux density of 1.97 Jy for 0336+323,
144: and the flux uncertainty is $\sim$10\%.
145: To avoid the degradation of sensitivity owing to pointing errors,
146: pointing was referenced
147: by observing the calibrators at the $X$-band (3.6 cm).
148: This referenced pointing was performed
149: about once an hour and just before observing the flux calibrator.
150:
151: Maps were made using a CLEAN algorithm.
152: The \NHt\ data produced synthesized beams
153: of FWHM = 0.97$''$ $\times$ 0.94$''$
154: and P.A. = 29\arcdeg\ for the (2, 2) line
155: and 0.98$''$ $\times$ 0.95$''$ and 47\arcdeg\ for the (3, 3) line,
156: when the imaging was done with a natural weighting.
157: The 1.3 cm continuum data produced a synthesized beam
158: of 1.02$''$ $\times$ 0.91$''$ and 71\arcdeg,
159: with a robust weighting.
160:
161: \section{RESULTS AND DISCUSSION}
162:
163: Figure 1 shows the \NHt\ (3, 3) line image
164: with an angular resolution similar to the SiO outflow image of Choi (2005).
165: While the \NHt\ map shows clumpy structures around the outflows,
166: the most striking feature is the compact structure
167: associated with the central objects,
168: which were not seen in the SiO map.
169: (The outflows will be discussed separately in a future paper.)
170: Figure~2 compares the \NHt\ maps with the 1.3 cm continuum map,
171: Table 1 lists the continuum parameters,
172: and Figure 3 shows the \NHt\ spectra.
173: The deconvolved FWHM sizes of the \NHt\ (3, 3) sources are
174: 350 $\times$ 260 AU$^2$ (1.1$''$ $\times$ 0.8$''$) for A1
175: and 260 $\times$ 60 AU$^2$ (0.8$''$ $\times$ 0.2$''$) for A2,
176: assuming a distance of 320 pc (de Zeeuw et al. 1999).
177: Their compact nature suggests that the \NHt\ lines
178: are tracing accretion disks.
179: This interpretation is especially strong in the case of A2
180: because the emission structure is elongated
181: in the direction perpendicular to the main bipolar outflow.
182: The position angle difference
183: between the minor axis of the A2 disk and the outflow axis
184: is $\sim$20\arcdeg.
185: In addition, the blue/redshifted emission peaks of A2 are
186: displaced in a way suggestive of a rotating disk (Fig. 2$c$).
187:
188: \input{fig1.tex}
189: \input{fig2.tex}
190: \input{tab1.tex}
191:
192: \subsection{Flux Anti-correlation}
193:
194: Comparisons of the \NHt\ and the continuum maps
195: reveal a remarkable anti-correlation:
196: A1 is brighter than A2 in the continuum map,
197: but it is in the other way around in the \NHt\ maps.
198: For a quantitative analysis,
199: we may define an \NHt-to-dust flux ratio
200: $R = F({\rm NH}_3) / F({\rm 2.7\ mm})$,
201: which is the ratio of the total flux densities
202: of the \NHt\ (3, 3) line to the 2.7 mm continuum.
203: The flux densities from the \NHt\ data shown in Figure 2$b$
204: and the 2.7 mm data of Looney et al. (2000)
205: give $R$(A1) = (7.1 $\pm$ 1.1) $\times$ 10$^{-3}$
206: and $R$(A2) = (4.8 $\pm$ 0.7) $\times$ 10$^{-2}$.
207: (The uncertainties were estimated
208: by assuming a 10\%\ uncertainty in flux scales.)
209: That is, the \NHt-to-dust flux ratio of A2 is larger
210: than that of A1 by a factor of 6.8 $\pm$ 1.4.
211: (If we use the 1.3 cm flux densities,
212: $R$(A2)/$R$(A1) is about 12.)
213: Before making interpretations of the flux ratio as column density ratio,
214: several factors affecting the flux densities should be considered.
215:
216: First, can the line optical depth affect the flux ratio?
217: Since the satellite hyperfine components
218: were not covered in our observations,
219: the optical depth cannot be deduced directly.
220: The line profiles (Fig. 3), however, provide useful information.
221: The spectra toward A2 have larger line widths and higher intensities
222: than those of A1,
223: which suggests that A2 may have relatively larger optical depths.
224: In addition, the (2, 2) spectrum toward A2 shows a self-absorption dip,
225: suggesting that the line may be optically thick.
226: Therefore, if the line optical depth effects are considered,
227: the degree of the anti-correlation would be severer.
228:
229: Note that, however,
230: there are alternative explanations for the line profiles.
231: The line width difference can be explained
232: if source A2 is a nearly edge-on disk
233: and A1 is either a relatively more face-on disk or a static core.
234: Also, the central dip of the spectra can be caused
235: by a missing flux problem owing to large-scale structures.
236: This issue can be addressed by future observations
237: either with a higher angular resolution
238: or with a spectral coverage wide enough
239: to include the satellite hyperfine components.
240:
241: Second, can the high \NHt\ flux of A2 be caused by a peculiar excitation?
242: The ratio between the (3, 3) and the (2, 2) lines
243: is not very useful for a quantitative analysis
244: because they belong to different species,
245: ortho-\NHt\ and para-\NHt, respectively (Ho \& Townes 1983).
246: Even so, the (3, 3) to (2, 2) line ratio
247: is similar in both sources (Fig. 3),
248: suggesting that they have similar excitation conditions.
249: In fact, the critical density of the \NHt\ inversion transitions
250: are low ($\sim$2 $\times$ 10$^3$ cm$^{-3}$; Ho \& Townes 1983),
251: and the \NHt\ molecules are expected to be thermalized.
252: Therefore, it is unlikely
253: that the difference in the \NHt-to-dust flux ratio
254: is caused by a peculiar excitation condition of \NHt\
255: in one of the sources.
256:
257: Finally, can the dust properties affect the flux ratio?
258: The spectral index in principle
259: provides some information on the dust opacity index.
260: The spectral index of A1 is slightly lower than that of A2 (Table 1),
261: but these values cannot be used directly
262: because the contributions from free-free emission
263: to the 1.3 cm fluxes are not known.
264: Flux measurements at submillimeter wavelengths
265: are desirable for estimating the dust opacity index.
266: Girart et al. (2006) presented a 345 GHz map
267: that resolves the continuum peaks,
268: but their beam size was not quite small enough
269: for measuring the total flux densities of each source separately.
270: Therefore, this issue cannot be resolved with the currently available data,
271: and we presume that the dust opacity index of the two sources are similar.
272: Another variable that can affect the flux ratio is the dust optical depth.
273: The dust is most likely optically thin at centimeter wavelengths,
274: but it could be optically thick at submillimeter.
275: While the measurement of the optical depth is not easy,
276: if this effect is significant,
277: it would affect the stronger source, A1, more severely.
278: Therefore, a correction for the dust optical depth, if necessary,
279: would also make the degree of the anti-correlation severer.
280:
281: \input{fig3.tex}
282:
283: In summary, the results of our observations suggest
284: that the IRAS 4A system contains
285: two sources of contrasting conditions.
286: The \NHt-to-dust flux ratio of A2 is $\sim$7 times larger than that of A1,
287: and the \NHt-to-dust column density ratios
288: may be different by a similar factor,
289: or by a larger factor
290: if the optical depth of the line or the dust emission is considered.
291: The difference between the two sources is huge,
292: considering that they are accreting matter
293: from a common protostellar envelope,
294: and must be caused by the physical and chemical processes
295: happening in each component.
296:
297: \subsection{Possible Explanations}
298:
299: There are two possible explanations
300: for the flux/column-density anti-correlation of the IRAS 4A system.
301: The main difference between them is the evolutionary status of A1.
302: While A2 is almost certainly a protostar,
303: the nature of A1 is less certain.
304:
305: (1)
306: A1 and A2 are roughly coeval,
307: the \NHt\ lines trace two accretion disks,
308: and the anti-correlation is caused
309: by a peculiar physical/chemical condition in one of them.
310: In this scenario,
311: A2 is a protostar driving the northeast-southwestern outflow,
312: and A1 is another protostar driving the southern outflow.
313: (We will elaborate on this model in \S~3.3.)
314:
315: (2)
316: A1 and A2 are in quite different stages of evolution,
317: the \NHt\ maps show
318: a (spherical) dense core (A1) and an accretion disk (A2),
319: and the anti-correlation is
320: an indication of the difference in their nature.
321: In this scenario, A2 is an actively accreting protostar,
322: and A1 is a pre-protostellar object without an outflow activity.
323: The strong millimeter continuum of A1 suggests a high concentration of dust.
324: Nevertheless, the compact structure detected by interferometers
325: suggests that A1 may not be a usual pre-protostellar object, either.
326: Then A1 could be a transitionary object,
327: either a pre-protostellar object on the verge of collapse
328: or a protostar immediately after the onset of collapse.
329:
330: We prefer the first explanation for several reasons.
331: First, A1 is bright in the centimeter continuum (Reipurth et al. 2002).
332: This free-free emission is a clear sign of outflow activity.
333: Second, comparison of the mass estimates over a range of size scale
334: indicates that A1 has a steep density gradient.
335: [For example, the mass of the IRAS 4A envelope
336: within a diameter of $\sim$9$''$ is 3.2~\Msun,
337: when scaled to the distance of 320 pc (Sandell \& Knee 2001).
338: The mass of A1 within a $\sim$2.5$''$ box is 1.9 \Msun,
339: when scaled to 320 pc and scaled by the flux ratio between A1 and A2
340: (Looney et al. 2000).
341: Then the density of A1 (average within a diameter of 820 AU) is higher
342: than the density of the envelope (average within 2900 AU)
343: by a factor of 26 $\pm$ 4.]
344: Third, in the second scenario,
345: the southern outflow might be driven by an unknown protostellar object,
346: which is unlikely because NGC 1333 IRAS 4 is
347: one of the most extensively observed regions
348: of star formation near the Sun.
349: Therefore, we suppose that both A1 and A2 are protostars.
350: In the following section,
351: we will discuss the implications of our observations
352: based on the first explanation.
353:
354: \subsection{Peculiarity of the IRAS 4A2 Disk}
355:
356: The difference in the column density ratio suggests
357: that one of the IRAS 4A disks is peculiar.
358: There are several lines of evidence
359: indicating that the A2 disk is unusually active.
360: First, most of the water maser spots in this region
361: are intimately associated with A2 (Fig. 2),
362: and their velocities are close to the systemic velocity of the cloud core
363: (Furuya et al. 2003; Park \& Choi 2007),
364: suggesting the existence of shocked gas in the A2 disk.
365: Second, the outflow driven by A2 (northeast-southwestern bipolar outflow)
366: is stronger than the one driven by A1 (southern outflow)
367: (Choi 2005; Choi et al. 2006),
368: also suggesting that the outflow engine (the accretion disk)
369: is more active in A2 than in A1.
370: The northeast-southwestern outflow of IRAS A2
371: is one of the best collimated molecular outflows (Blake et al. 1995).
372: Third, A2 seems to have an unusually large $R$ value.
373: Comparison with other protostars may tell
374: which disk is the peculiar one.
375: Examples of protostellar disks detected in the \NHt\ (3, 3) line is rare,
376: but fortunately IRAS 4BI was detected.
377: IRAS 4BI is a single protostar located within the field of view
378: of our observations (Fig. 1).
379: Measurements of the flux densities
380: give $R$(BI) = (4 $\pm$ 2) $\times$ 10$^{-3}$,
381: which is similar to $R$(A1)
382: and suggests that A2 may be the abnormal one.
383: Therefore, the A2 disk may be unusually gas-rich or dust-poor.
384: Such a condition may be possible if the disk is very active or hot
385: so that the dust grains infalling from the protostellar envelope
386: may be destroyed and converted to gaseous molecules,
387: probably via evaporation of molecules in the grain mantle.
388:
389: The anti-correlation between the gas and the dust flux densities
390: of the IRAS 4A disks
391: has important implications on the star formation process.
392: The standard models of accretion
393: are based on the cases of single central star,
394: and the mass accretion rate is
395: mainly related with the density structure of the protostellar envelope
396: (Shu et al. 1987).
397: In binary systems, however, as the IRAS 4A system shows,
398: even though the two components share a common envelope,
399: each of them can evolve in a distinctive way.
400: That is, there is an important controlling agent
401: that may be lacking in the case of isolated star formation.
402: A possibly crucial factor can be
403: the distribution and (mis-)alignment of angular momentum
404: (Bodenheimer 1995).
405: If the mass outflow rate is a good indicator of the mass accretion rate,
406: A2 may be growing much faster than A1,
407: by accreting matter through an active disk.
408:
409: One interesting issue to be addressed in the future is the \NHt\ abundance.
410: Does the high column-density ratio of A2 mean
411: an overall enhancement of gaseous molecules relative to dust?
412: Or does it mean a selective enhancement
413: of \NHt\ (and related species) only?
414: Estimating the degree of enhancement is difficult
415: because comparison with CO lines cannot be interpreted easily
416: owing to the complicated chemistry of nitrogen-bearing molecules
417: (Charnley 1997)
418: and confusion with outflows.
419: Currently there is no reliable estimates of \NHt\ abundance
420: in protostellar disks,
421: and it is needed to make high-resolution images of the two disks
422: in a variety of molecular lines.
423:
424: Since planets do exist in multiple-star systems (Raghavan et al. 2006),
425: our results have interesting implications on the planet formation.
426: If the A2 disk is indeed gas-rich/dust-poor,
427: and if such a condition can persist
428: until the planet-forming phase of the disk evolution,
429: planetary systems produced in such disks
430: may look very different from our solar system.
431: We may speculate that, for example, such a system may strongly favor
432: the growth of gas-giant planets.
433: Thus, multiple-protostar systems can produce
434: diverse types of planetary systems.
435:
436: \acknowledgements
437:
438: We thank J. Cho, K.-T. Kim, and Y. Lee
439: for helpful discussions and encouragement.
440: %This work was partially supported by the LRG program of KASI.
441: G. P. was partially supported by the Brain Korea 21 project
442: of the Korean Government.
443:
444: \begin{references}
445: \reference{} Aspin, C., Sandell, G., \& Russell, A. P. G. 1994,
446: A\&AS, 106, 165
447: \reference{} Bachiller, R. 1996, ARA\&A, 34, 111
448: \reference{} Bally, J., Devine, D., \& Reipurth, B. 1996, ApJ, 473, L49
449: \reference{} Blake, G. A., Sandell, G., van Dishoeck, E. F.,
450: Groesbeck, T. D., Mundy, L. G., \& Aspin, C. 1995,
451: \reference{} Bodenheimer, P. 1995, ARA\&A, 33, 199
452: \reference{} Cabrit, S., \& Andr{\'e}, P. 1991, ApJ, 379, L25
453: \reference{} Charnley, S. B. 1997, MNRAS, 291, 455
454: \reference{} Choi, M. 2001, ApJ, 553, 219
455: \reference{} Choi, M. 2005, ApJ, 630, 976
456: \reference{} Choi, M., Hodapp, K. W., Hayashi, M., Motohara, K., Pak, S.,
457: \& Pyo, T.-S. 2006, ApJ, 646, 1050
458: \reference{} de Zeeuw, P. T., Hoogerwerf, R., de Bruijne, J. H. J.,
459: Brown, A. G. A., \& Blaauw, A. 1999, AJ, 117, 354
460: \reference{} Furuya, R. S., Kitamura, Y., Wootten, A., Claussen, M. J.,
461: \& Kawabe, R. 2003, ApJS, 144, 71
462: \reference{} Girart, J. M., Crutcher, R. M., \& Rao, R. 1999, ApJ, 525, L109
463: \reference{} Girart, J. M., Rao, R., \& Marrone, D. P. 2006,
464: Science, 313, 812
465: \reference{} Ho, P. T. P., \& Townes, C. H. 1983, ARA\&A, 21, 239
466: \reference{} Lay, O. P., Carlstrom, J. E., \& Hills, R. E. 1995,
467: ApJ, 452, L73
468: \reference{} Looney, L. W., Mundy, L. G., \& Welch, W. J. 2000, ApJ, 529, 477
469: \reference{} Park, G., \& Choi, M. 2007, ApJ, 664, L99
470: \reference{} Raghavan, D., Henry, T. J., Mason, B. D., Subasavage, J. P.,
471: Jao, W.-C., Beaulieu, T. D., \& Hambly, N. C. 2006,
472: ApJ, 646, 523
473: \reference{} Reipurth, B., Rodr{\'{\i}}guez, L. F., Anglada, G.,
474: \& Bally, J. 2002, AJ, 124, 1045
475: \reference{} Rodr\'{\i}guez, L. F., Anglada, G., \& Curiel, S. 1999,
476: ApJS, 125, 427
477: \reference{} Sandell, G., Aspin, C., Duncan, W. D., Russell, A. P. G.,
478: \& Robson, E. I. 1991, ApJ, 376, L17
479: \reference{} Sandell, G., \& Knee, L. B. G. 2001, ApJ, 546, L49
480: \reference{} Shu, F. H., Adams, F. C., \& Lizano, S. 1987, ARA\&A, 25, 23
481: \end{references}
482:
483: %\TimeStamp
484:
485: \end{document}
486:
487: