Observations of Cyanopolyynes toward Four High-Mass Star-Forming Regions Containing Hot Cores

TL;DR

This study surveys cyanopolyynes in four high-mass star-forming regions, detecting HC5N and HC7N, revealing warm gas presence, and suggesting chemical differentiation among the regions.

Contribution

First detection of high-excitation HC5N lines in warm regions around massive young stars, indicating its existence in warm gas within 0.07-0.1 pc radii.

Findings

HC5N detected in all sources; HC7N in three.

HC5N rotational temperatures are 13-20 K.

Evidence of chemical differentiation among regions.

Abstract

We carried out line survey observations at the 26-30 GHz band toward the four high-mass star-forming regions containing hot cores, G10.30-0.15, G12.89+0.49, G16.86-2.16, and G28.28-0.36, with the Robert C. Byrd Green Bank Telescope. We have detected HC5N from all of the sources, and HC7N from the three sources, except for G10.30-0.15. We further conducted observations of HC5N at the 42-46 GHz and 82-103 GHz bands toward the three sources, G12.89+0.49, G16.86-2.16, and G28.28-0.36, with the Nobeyama 45 m radio telescope. The rotational lines of HC5N with the high excitation energies (Eu/k=63-100 K), which are hardly excited in the cold dark clouds, have been detected from the three sources. The rotational temperatures of HC5N are found to be 13-20 K in the three sources. The detection of the lines with the high excitation energies and the derived rotational temperatures indicate that HC5N exists in the warm gas within 0.07-0.1 pc radii around massive young stellar objects. The column densities of HC5N in the three sources are derived to be (2.0+-2.8)*10^(13) cm-2 . We compare the ratios between N(HC5N) the column density of HC5N and W(CH3OH) the integrated intensity of the thermal CH3OH emission line among the three high-mass star-forming regions. We found a possibility of the chemical differentiation in the three high-mass star-forming regions; G28.28-0.36 shows the largest N (HC5N)/W (CH3OH) ratio of > 8.0*10^(14) in units of (K km s-1 )-1 cm-2 , while G12.89+0.49 and G16.86-2.16 show the smaller values (~2*10^(13) ).

Tables

Table 1. Table 1: Observing parameters

Frequency	Telescope	Beam size	${\eta }_{\mathrm{mb}}$	$T_{\mathrm{sys}}$	$\Delta \nu$	$T_{\mathrm{rms}}$
(GHz)		(′′)	(%)	(K)	(kHz)	(mK)
26$-$30	GBT	27	77	70$-$100 (April)	66	10$-$17 ($T_{\mathrm{mb}}$)
				90$-$120 (July)
42$-$46	Nobeyama	37	71	120$-$150	122.07	6$-$14 ($T_A^{\ast }$)
82$-$103	Nobeyama	18aaThe values are at the 86 GHz.	54aaThe values are at the 86 GHz.	120$-$200	244.14	3$-$6 ($T_A^{\ast }$)

Table 2. Table 2: Properties of our target sources

Source	R.A.aaPurcell et al. (2006)	Decl.aaPurcell et al. (2006)	$D$	$V_{\mathrm{LSR}}$aaPurcell et al. (2006)	Other AssociationbbThe symbols of ”Y” and ”N” represent detection and non-detection, respectively. “UCHII” indicates an ultracompact H II region lies within the Mopra beam ($\sim 38$′′). The 6.7 GHz methanol masers are associated with all of the four sources.
	(J2000)	(J2000)	(kpc)	(km s-1)	UCHIIaaPurcell et al. (2006)	outflow
G10.30-0.15	18h08m55$\stackrel{s}{.}$5	-20°05′58$\stackrel{\prime \prime }{.}$0	2.1ccGreen et al. (2014)	13.0	Y	NffCyganowski et al. (2008)
G12.89+0.49	18h11m51$\stackrel{s}{.}$4	-17°31′30$\stackrel{\prime \prime }{.}$0	2.50ddReid et al. (2014)	33.3	N	YeeLi et al. (2016)
G16.86-2.16	18h29m24$\stackrel{s}{.}$4	-15°16′04$\stackrel{\prime \prime }{.}$0	1.67eeLi et al. (2016)	17.8	N	YeeLi et al. (2016)
G28.28-0.36	18h44m13$\stackrel{s}{.}$3	-04°18′03$\stackrel{\prime \prime }{.}$0	3.0ccGreen et al. (2014)	48.9	Y	YffCyganowski et al. (2008)

Table 3. Table 3: Spectral line parameters observed with the GBT

			G10.30-0.15				G12.89+0.49				G16.86-2.16				G28.28-0.36
Species	FrequencyaaTaken from the Cologne Database for Molecular Spectroscopy (CDMS) (M$\ddot{u}$ller et al., 2005).	$E_u/k$	$T_{\mathrm{mb}}$	FWHM	$\int T_{\mathrm{mb}}𝑑v$		$T_{\mathrm{mb}}$	FWHM	$\int T_{\mathrm{mb}}𝑑v$		$T_{\mathrm{mb}}$	FWHM	$\int T_{\mathrm{mb}}𝑑v$		$T_{\mathrm{mb}}$	FWHM	$\int T_{\mathrm{mb}}𝑑v$
(Transition)	(GHz)	(K)	(K)	(km s-1)	(K km s-1)		(K)	(km s-1)	(K km s-1)		(K)	(km s-1)	(K km s-1)		(K)	(km s-1)	(K km s-1)
HC3N
($J=3-2$)	27.29429	2.6	1.14 (5)	7.9 (4)	9.6 (6)		1.60 (11)	4.5 (4)	7.7 (8)		1.94 (5)	4.42 (14)	9.1 (4)		2.24 (11)	2.38 (14)	5.7 (4)
HC5N
($J=10-9$)	26.62654	7.0	0.107 (14)	2.9 (4)	0.33 (7)		0.209 (13)	3.5 (3)	0.78 (8)		0.37 (2)	2.6 (2)	1.03 (11)		0.37 (3)	2.08 (17)	0.81 (9)
($J=11-10$)	29.28915	8.4	0.158 (11)	4.7 (4)	0.79 (8)		0.283 (16)	3.6 (2)	1.10 (9)		0.42 (3)	3.1 (3)	1.41 (16)		0.53 (2)	1.91 (9)	1.08 (7)
HC7N
($J=24-23$)	27.07181	16.2		…	…		0.076 (8)	4.5 (5)	0.36 (6)			…	…		0.086 (10)	2.4 (3)	0.22 (4)
($J=25-24$)	28.19981	17.6		…	…		0.051 (8)	3.1 (6)	0.16 (4)			…	…		0.083 (12)	2.0 (3)	0.17 (4)
($J=26-25$)	29.32777	19.0		…	…			…	…		0.104 (12)	1.17 (15)	0.13 (2)		0.098 (12)	1.8 (2)	0.19 (3)
CH3OH
($J_K=4_0-3_1$ $E$)	28.31607	36.3	0.186 (5)	7.5 (4)	1.48 (10)		0.259 (14)	4.5 (3)	1.25 (10)		0.180 (14)	5.5 (5)	1.06 (13)			…	…
SO2
($J_{Ka,Kc}=4_{0,4}-3_{1,3}$)	29.32133	9.2	-	-	-		0.16 (4)	5.1 (1.3)	0.9 (3)		0.27 (3)	4.8 (7)	1.4 (2)		0.084 (12)	2.6 (4)	0.23 (5)
H2CO
($J_{Ka,Kc}=3_{1,2}-3_{1,3}$)	28.97480	33.4		…	…		0.284 (14)	4.5 (3)	1.37 (10)		0.285 (19)	3.5 (3)	1.05 (10)			…	…

Table 4. Table 4: Spectral line parameters of HC 5 N observed with the Nobeyama 45 m telescope

			G12.89+0.49					G16.86-2.16					G28.28-0.36
Transition	FrequencyaaTaken from the Cologne Database for Molecular Spectroscopy (CDMS) (M$\ddot{u}$ller et al., 2005).	$E_u/k$	$T_{\mathrm{mb}}$	FWHM	$\int T_{\mathrm{mb}}𝑑v$	$V_{\mathrm{LSR}}$bbThe errors were 0.85 km s-1, which corresponds to the velocity resolution (Section 2.2).		$T_{\mathrm{mb}}$	FWHM	$\int T_{\mathrm{mb}}𝑑v$	$V_{\mathrm{LSR}}$bbThe errors were 0.85 km s-1, which corresponds to the velocity resolution (Section 2.2).		$T_{\mathrm{mb}}$	FWHM	$\int T_{\mathrm{mb}}𝑑v$	$V_{\mathrm{LSR}}$bbThe errors were 0.85 km s-1, which corresponds to the velocity resolution (Section 2.2).	${\eta }_B$
($J^{\prime }-J"$)	(GHz)	(K)	(K)	(km s-1)	(K km s-1)	(km s-1)		(K)	(km s-1)	(K km s-1)	(km s-1)		(K)	(km s-1)	(K km s-1)	(km s-1)	(%)
16$-$15	42.60215	17.4	0.186 (7)	3.36 (14)	0.67 (4)	32.7		0.296 (17)	2.69 (18)	0.85 (7)	19.0		0.251 (14)	1.97 (12)	0.53 (4)	49.6	71
17$-$16	45.26472	19.6	0.148 (8)	3.5 (2)	0.56 (5)	32.4		0.253 (17)	2.8 (2)	0.75 (8)	19.0		0.239 (13)	1.88 (11)	0.48 (4)	49.1	71
31$-$30	82.53904	63.4	0.096 (7)	3.8 (3)	0.39 (4)	33.6		0.063 (6)	4.0 (5)	0.27 (4)	17.8		0.068 (6)	2.2 (2)	0.16 (2)	49.5	56
32$-$31	85.20134	67.5	0.092 (8)	3.4 (3)	0.34 (4)	33.7		0.084 (8)	2.8 (3)	0.25 (3)	17.4		0.053 (4)	2.6 (2)	0.144 (17)	49.3	54
34$-$33	90.52589	76.0	0.109 (7)	4.4 (3)	0.50 (5)	33.3		0.133 (7)	3.1 (2)	0.44 (4)	17.1		0.036 (5)	2.5 (4)	0.10 (2)	49.5	52
36$-$35	95.85034	85.1	0.063 (5)	4.0 (3)	0.27 (3)	34.1		0.042 (6)	4.4 (7)	0.20 (4)	17.5		0.031 (4)	2.8 (5)	0.09 (2)	49.1	49
38$-$37	101.17468	94.7	0.034 (4)	6.5 (9)	0.23 (4)	32.3		0.106 (9)	2.9 (3)	0.33 (4)	17.7		0.018 (5)	2.1 (6)	0.039 (16)	49.0	46
39$-$38	103.83682	99.7	0.035 (5)	4.1 (6)	0.15 (3)	33.0		0.124 (9)	2.7 (2)	0.35 (4)	17.5		0.039 (6)	1.6 (3)	0.069 (17)	49.7	45

Table 5. Table 5: The rotational temperatures and column densities of HC 5 N in the four sources

Source	filling factor = 1		With beam-size correction
	$T_{\mathrm{rot}}$ (K)	$N$ ($\times {10}^{12}$ cm-2)	$T_{\mathrm{rot}}$ (K)	$N$ ($\times {10}^{12}$ cm-2)
G10.30-0.15	$21\pm 11$aaThe value is the average rotational temperature of the three sources except for G10.30-0.15. The error originates from those of the rotational temperatures in the three sources.	${8.5}_{-1.7}^{+2.0}$	$16\pm 4$aaThe value is the average rotational temperature of the three sources except for G10.30-0.15. The error originates from those of the rotational temperatures in the three sources.	${7.5}_{-0.6}^{+0.8}$
G12.89+0.49	${23}_{-5}^{+7}$	${9.4}_{-2.0}^{+1.8}$	$18\pm 2$	${23.9}_{-1.7}^{+1.5}$
G16.86-2.16	${23}_{-4}^{+7}$	$11\pm 2$	$17\pm 2$	${27.8}_{-2.0}^{+1.6}$
G28.28-0.36	${17}_{-3}^{+4}$	${7.5}_{-2.1}^{+1.9}$	${13.8}_{-1.1}^{+1.5}$	${20.5}_{-0.5}^{+2.0}$

Table 6. Table 6: The values of 850 µm µm \micron continuum flux, dust temperature, N 𝑁 N (H 2 ), and X 𝑋 X (HC 5 N) in the four hot cores

	850 $\mathrm{\mu m}$ flux	$T$aaTaken from online material of Purcell et al. (2009).	$N$(H2)bbThe errors were derived from the uncertainties of the dust temperatures of 20 K.	$X$(HC5N)
Source	(Jy beam-1)	(K)	($\times {10}^{22}$ cm-2)	($\times {10}^{-10}$)
G10.30-0.15	4.1	71	${1.2}_{-0.3}^{+0.7}$	${6.3}_{-2.6}^{+3.2}$
G12.89+0.49	5.8	54	${2.4}_{-0.8}^{+2.5}$	${9.9}_{-5.4}^{+6.1}$
G16.86-2.16	5.8	70	${1.7}_{-0.5}^{+1.0}$	$16\pm 7$
G28.28-0.36	1.3	58 ccThe average value derived from 65 good fits (Purcell et al., 2009).	${0.49}_{-0.16}^{+0.42}$	${42}_{-20}^{+26}$

Equations

$$\ln \frac{3k\int T_{\mathrm{mb}}dv}{8{\pi }^3\nu S{\mu }^2}=\ln \frac{N}{Q(T_{\mathrm{rot}})}-\frac{E_{\mathrm{u}}}{k{T}_{\mathrm{rot}}},$$

$$N({\rm{H}}_{2})=2.02\times 10^{20}{\rm{cm}}^{-2}\Bigl{(}e^{1.439(\lambda/{\rm{mm}})^{-1}(T/10{\rm{K}})^{-1}}-1\Bigr{)}\Bigl{(}\frac{\kappa_{\nu}}{0.01\>{\rm{cm}}^{2}\>{\rm{g}}^{-1}}\Bigr{)}^{-1}\Bigl{(}\frac{S_{\nu}^{\rm{beam}}}{{\rm{mJy}}\>{\rm{beam}}^{-1}}\Bigr{)}\Bigl{(}\frac{\theta_{\rm{HPBW}}}{10\>{\rm{arcsec}}}\Bigr{)}^{-2}\Bigl{(}\frac{\lambda}{{\rm{mm}}}\Bigr{)}^{3}.$$

Full text

astrochemistry — ISM: molecules — stars: formation

††journal: ApJ††facilities: Robert C. Byrd Green Bank Telescope, Nobeyama 45 m radio telescope††software: GBTIDL, Java Newstar

1 Introduction

As the star formation process progresses, the chemical composition of molecular gas changes. The star formation starts with the cold dark molecular clouds ($T\sim 10$ K, $n\sim 10^{4}$ cm*-3*). In the dark clouds, carbon-chain molecules such as CCS and HC5N are abundant (Suzuki et al., 1992; Hirota et al., 2009). After formation of protostars, the cloud temperature rises. Such physical changes significantly affect the chemical composition, leading to hot core chemistry in the high-mass star-forming regions and hot corino chemistry in the low-mass star-forming regions. In hot core and hot corino regions ($T\geq 100$ K, $n\sim 10^{6}$ cm*-3*), while carbon-chain molecules except for HC3N are deficient, saturated complex organic molecules (COMs) such as CH3OH and CH3CN are rich. These COMs are formed by the combination of the grain-surface reactions and the following gas-phase reactions (e.g. Garrod & Herbst, 2006; Herbst & van Dishoeck, 2009).

Conversely, two low-mass star-forming cores with a variety of carbon-chain molecules have been discovered; L1527 in Taurus (Sakai et al., 2008) and IRAS 15398-3359 in Lupus (Sakai et al., 2009a). The rotational temperatures of carbon-chain molecules in L1527 were derived to be $\sim 12-15$ K (Sakai et al., 2009b), which are higher than those in the cold dark clouds. Aikawa et al. (2008) and Hassel et al. (2008) showed that CH4, which is evaporated from grain mantles in the warm regions ($T\sim 20$ K) around the protostar, could form carbon chains in the gas phase. In fact, carbon-chain molecules were found to be enhanced in the warm regions where the temperature becomes higher than 20$-$30 K (Sakai et al., 2010). Such a chemical mechanism was named warm carbon chain chemistry (WCCC, Sakai et al., 2008).

While there have been many studies about carbon-chain molecules in the low-mass star-forming regions, our knowledge of carbon-chain molecules in the high-mass star-forming regions is still poor. As with the low-mass star-forming cores, understanding chemical mechanisms of carbon-chain molecules in the high-mass star-forming cores is one of the keys to revealing the initial conditions of massive star formation. Cyanoacetylene (HC3N), the shortest member of cyanopolyynes, is one of the hot core tracers (a dense gas tracer; $n_{\rm{H}_{2}}\sim 10^{5}-10^{6}$ cm*-3*, Bergin et al., 1996), whereas longer cyanopolyynes are rarely detected. Chapman et al. (2009) demonstrated their chemical model calculation, and suggested that cyanopolyynes (HC2n+1N; $n=1-4$) could be efficiently formed by the neutral-neutral reactions between CN and C2nH2. The latter is formed from C2H2 evaporated from grain mantles in hot core regions. Taniguchi et al. (2016b) proposed that the main formation pathway of HC3N in the high-mass star-forming region G28.28-0.36 is the neutral-neutral reaction between C2H2 and CN, which is the same reaction suggested by Chapman et al. (2009), based on their observations of the 13C isotopic fractionation with the Nobeyama 45 m radio telescope. Motivated by the chemical model calculation of Chapman et al. (2009), Green et al. (2014) conducted survey observations of HC5N using the $J=12-11$ rotational line toward 79 hot cores with the Tidbinbilla 34 m radio telescope, and detected HC5N toward 35 sources. They could not derive the excitation temperatures due to the detection of the single rotational line which can be excited even in the cold dark clouds. Thus, the detected emission lines may come from the cold envelopes around the protostars. We still cannot confirm that long cyanopolyynes are formed and exist in hot cores.

In this paper, we carried out observations at the 26$-$30 GHz band toward four high-mass star-forming regions containing hot cores, G10.30-0.15, G12.89+0.49, G16.86-2.16, and G28.28-0.36, with the Robert C. Byrd Green Bank Telescope (GBT)111The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.. Two rotational transitions of HC5N ($J=10-9$ and 11$-$10) and three transitions of HC7N ($J=24-23$, 25$-$24, and 26$-$25) are in the frequency band. HC7N has been detected in several star-forming regions (Sakai et al., 2008; Cordiner et al., 2012; Friesen et al., 2013; Feng et al., 2015), which is considered as an indicator of cyanopolyyne-rich sources. We also observed the rotational lines of HC5N at the 42$-$46 GHz and 82$-$103 GHz bands toward the three sources, except for G10.30-0.15, with the Nobeyama 45 m radio telescope222The Nobeyama 45 m radio telescope is operated by the Nobeyama Radio Observatory, a branch of the National Astronomical Observatory of Japan.. Several lines of HC5N with the high excitation energies ($E_{\rm{u}}/k\sim 63-100$ K) are in the 82$-$103 GHz band. Such lines are hardly excited in the cold environments, and the detection of these lines means that HC5N exists in the warm regions, as discussed later. From the results with the two radio telescopes, we derive the rotational temperatures and the column densities of HC5N in the three sources using the rotational diagram. We confirm that HC5N exists in the warm gas in the high-mass star-forming regions from the rotational temperatures. The thermal emission line of CH3OH, which is one of the hot core tracers and representative COMs, is in the frequency band of the GBT observations. The origin of CH3OH is considered to be different from that of carbon-chain species. We then investigate the relationship between HC5N and CH3OH in the three high-mass star-forming regions in Section 4.2.1. We found a possibility of the chemical differentiation among the three sources.

2 Observations

We describe the observational details of each frequency band in this section. Table 1 summarizes the observing parameters of each frequency band.

The observed positions, distances, systemic velocities, and related objects (ultracompact H II regions and outflows) of our four targets are summarized in Table 2. We selected these targets from the list of sources detected in HC5N (Green et al., 2014) by applying the following three criteria:

the source declination is above -21°, 2. 2.

the distance ($D$) is within 3 kpc, and 3. 3.

CH3CN was detected (Purcell et al., 2006).

The first criterion enables us to observe with the GBT and the Nobeyama 45 m telescope. The second criterion is required for preventing from heavy beam dilution. The last one ensures that target sources contain hot cores. CH3CN is one of the hot core tracers, and it is always detected in the hot gas ($>100$ K, Bisschop et al., 2007). Eight sources in the list meet the three criteria. Since our purpose is to derive the accurate rotational temperatures, we chose three sources among the eight selected sources which show the highest peak intensities of HC5N (G12.89+0.49, G16.86-2.16, and G28.28-0.36). In order to investigate different conditions, we selected one source which shows the low peak intensity of HC5N (G10.30-0.15) with the high peak intensities of CH3CN.

\floattable

\floattable

2.1 Observations with the Robert C. Byrd Green Bank Telescope (GBT)

We carried out observations with the Robert C. Byrd Green Bank Telescope (GBT) of the National Radio Astronomy Observatory in 2016 April and July. We used the Ka-band receiver to cover the 26.1$-$29.7 GHz range, except for 28.7$-$28.9 GHz. The beam size (HPBW), the aperture efficiency ($\eta_{\rm{A}}$), and the main beam efficiency ($\eta_{\rm{mb}}$) of the Ka-band receiver were approximately 27*′′, 67%, and 77%, respectively. The typical system temperatures were from 70 to 100 K in April, and from 90 to 120 K in July, depending on the weather conditions and elevations. We used the VEGAS spectrometer whose bandwidth and frequency resolution were 1080 MHz and 66 kHz, respectively. The frequency resolution of 66 kHz corresponds to the velocity resolution of 0.73 km s-1* at 27 GHz. The center frequencies of four spectral windows were set at 26.525, 27.425, 28.325, and 29.275 GHz, respectively.

We employed the position-switching mode. The integration time was 60 seconds per on-source and off-source positions. The observed positions of the four target sources are summarized in Table 2. The off-source positions were set to be +15*′* away in the declination.

The pointing sources were J1850-0001 at ($\alpha_{2000}$, $\delta_{2000}$) = (18h50m31$\fs$17, -00°01′55$\farcs$1) for G28.28-0.36, and J1833-2103 at ($\alpha_{2000}$, $\delta_{2000}$) = (18h33m39$\fs$93, -21°03′40$\farcs$769) for the others. We checked the pointing accuracy every 1.5-2 hours, and the pointing error was less than 8*′′*, depending on the weather conditions. We observed the bright radio continuum source 3C 286 at ($\alpha_{2000}$, $\delta_{2000}$) = (13h31m08$\fs$288, +30°30′32$\farcs$96) in the beginning of every observing session in order to conduct the absolute flux calibration. The intensity calibration error was less than 5%.

2.2 Observations with the Nobeyama 45 m radio telescope

The observations were carried out between 2017 January and March. We observed toward the three sources G12.89+0.49, G16.86-2.16, and G28.28-0.36. The on-source and off-source positions were set at the same ones with the GBT observations. The position-switching mode was employed, and the integration time was 20 seconds per scan.

We used the Z45 receiver (Nakamura et al., 2015) for the 42$-$46 GHz band observations. The system temperatures of the Z45 receiver were 120$-$150 K, depending on the weather conditions and elevations. The main beam efficiency ($\eta_{\rm{mb}}$) and the beam size (HPBW) at 43GHz were 71% and 37*′′, respectively. The TZ receiver (Nakajima et al., 2013) was used for the 82$-$103 GHz band observations. The system temperatures were from 120 to 200 K. The main beam efficiency and the beam size at 86 GHz were 54% and 18′′*, respectively.

We used the SAM45 FX-type digital correlator (Kamazaki et al., 2012) in frequency settings whose bandwidths and resolutions are 500 MHz and 122.07 kHz for observations of the Z45 receiver, and 1000 MHz and 244.14 kHz for those of the TZ receiver, respectively. The frequency resolutions correspond to the velocity resolution of $\sim 0.85$ km s*-1*.

We checked the telescope pointing every 1.5 hr by observing SiO maser line ($J=1-0$) from OH39.7+1.5 at ($\alpha_{2000}$, $\delta_{2000}$) = (18h56m03$\fs$88, +06°38′49$\farcs$8). We used the Z45 receiver for the 42$-$46 GHz band observations, and the H40 receiver for the 82$-$103 GHz band observations. The pointing error was less than 3*′′*.

3 Results and Analyses

3.1 Results

3.1.1 Observational results with the GBT

We conducted data reduction333We set the velocity units as “lsrk” in the observational scripts, but we wrote the “$V_{\rm{LSR}}$” values in the source catalog. We then shifted the frequency so that the emission peaks of HC3N correspond to its rest frequency, when we made the final spectra. using the GBTIDL444http://gbtidl.nrao.edu. The overviews of obtained spectra toward the four sources are shown in Figures 1 and 2. The rms noise levels are 14, 10, 17, and 11 mK in $T_{\rm{mb}}$ for G10.30-0.15, G12.89+0.49, G16.86-2.16, and G28.28-0.36, respectively. We fitted the spectra with a Gaussian profile and obtained the spectral line parameters, as summarized in Table 3. The lower detection limit is set to be $4\sigma$.

Two and three rotational lines of HC5N ($J=10-9$ and $11-10$) and HC7N ($J=24-23$, $25-24$, and $26-25$) are in the observed frequency range, respectively. All of the observed rotational lines of HC5N were detected from all of the four sources. We also detected HC7N from the three sources, except for G10.30-0.15. Detection of HC7N has been reported in several star-forming regions (Sakai et al., 2008; Cordiner et al., 2012; Friesen et al., 2013; Feng et al., 2015). The detection of HC7N from the three high-mass star-forming regions suggests that our target cores contain the long cyanopolyynes abundantly.

In addition, several other lines were detected. The hydrogen recombination lines are prominent only in G10.30-0.15, which seems to be caused by a few ultracompact H II (UCH II) regions associated with G10.30-0.15 (Thompson et al., 2006). Although a UCH II region is associated with G28.28-0.36 (Walsh et al., 2003), we did not detect any hydrogen recombination lines in G28.28-0.36. The rotational lines of HC3N were detected toward all of the four sources. HC3N is considered as one of the hot core tracers (a dense gas tracer), and it is usually detected in hot core regions, as mentioned in Section 1. The thermal CH3OH emission lines ($J_{K}=4_{0}-3_{1}$ $E$) with the low-excitation energy ($E_{\rm{u}}/k=36.3$ K) were detected except for G28.28-0.36, and its lines with high-excitation energies ($E_{\rm{u}}/k\sim 120-340$ K) were detected only toward G12.89+0.49. Metastable inversion transition lines of NH3 with very high-excitation energies, i.e., ($J$, $K$) = (8, 8) at 26.51898 GHz ($E_{\rm{u}}/k\sim 686$ K) and ($J$, $K$) = (9, 9) at 27.47794 GHz ($E_{\rm{u}}/k\sim 852$ K), were detected toward G12.89+0.49 and G16.86-2.16. The results indicate that extremely hot gas is contained in these regions (Goddi et al., 2011). The H2CO ($J_{Ka,Kc}=3_{1,2}-3_{1,3}$) emission lines were detected toward G12.89+0.49 and G16.86-2.16.

3.1.2 Observational results with the Nobeyama 45 m radio telescope

We conducted data reduction using the Java Newstar, which is an astronomical data analyzing system of Nobeyama data555http://www.nro.nao.ac.jp/ jnewstar/html/. Figures 3 and 4 show the spectra of HC5N at the 42$-$46 GHz and 82$-$103 GHz bands, respectively. The rms noise levels are from 6 to 14 mK in $T_{\rm{A}}^{*}$ at the 42$-$46 GHz band and from 3 to 6 mK at the 82$-$103 GHz band. Table 4 summarizes the line parameters obtained from the Gaussian fitting. The main beam efficiencies which we used in the analysis are summarized in Table 4.

The signal-to-noise ratios were between 4 and 23. The two rotational lines of HC5N at the 42$-$46 GHz band ($J=16-15$ and 17$-$16) were detected from all of the three sources. Moreover, we detected six rotational lines of HC5N with the high excitation energies ($E_{\rm{u}}/k\sim 63-100$ K) from all of the sources. The $V_{\rm{LSR}}$ values of the detected lines are summarized in Table 4. The $V_{\rm{LSR}}$ values of the HC5N lines are consistent with the systemic velocities of each source (Table 2).

\floattable

\floattable

3.2 Analyses

We derived the rotational temperatures and column densities of HC5N in the three sources (G12.89+0.49, G16.86-2.16, and G28.28-0.36) from the rotational diagram analysis, using the following formula (Goldsmith & Langer, 1999);

[TABLE]

where $k$ is the Boltzmann constant, $S$ is the line strength, $\mu$ is the permanent electric dipole moment, $N$ is the column density, and $Q(T_{\rm{rot}})$ is the partition function. We used 4.330 D (Alexander et al., 1976) for $\mu$ of HC5N. Figure 5 shows the fitting results of the rotational diagram for the three sources. The errors include the Gaussian fitting errors, the uncertainties from the main beam efficiency of 10%, the chopper-wheel method of 10%, and from the other factors such as calibration and filling factor of 30%. Since we do not know the spatial distributions of HC5N, we assume the filling factor of unity. In the fitting, the values of the $J=16-15$ and 17$-$16 transitions are systematically lower than the fitting lines. This suggests a smaller filling factor than unity, because the beam size at the 42$-$46 GHz band with the Nobeyama 45 m telescope is significantly larger than those at the 26$-$30 GHz band with the GBT and at the 82$-$103 GHz band with the Nobeyama 45 m telescope. The ranges of the derived rotational temperatures and column densities in the three sources are 17$-$23 K and (7.5$-$11)$\times 10^{12}$ cm*-2*, as summarized in Table 5. In the case of G10.30-0.15, we detected only two lines with the similar energies using the GBT, and we cannot conduct the rotational diagram analysis. We derived the column density of HC5N in G10.30-0.15 to be $8.5^{+2.0}_{-1.7}\times 10^{12}$ cm*-2*, using the average value of the rotational temperatures in the three sources ($21\pm 11$ K). The error of the rotational temperature in G10.30-0.15 was estimated from the relative errors of the rotational temperatures in the three sources, and the error of the column density corresponds to the change in the rotational temperature.

Assuming a small beam filling factor, we multiplied the integrated intensities of the GBT data by ($\frac{27^{\prime\prime}}{18^{\prime\prime}}$)2 and the 45GHz band data by ($\frac{37^{\prime\prime}}{18^{\prime\prime}}$)2 for the correction of the different beam sizes. Figure 6 shows the rotational diagram with the beam-size correction. All of the data, including the 42$-$46 GHz band, are better fitted. From the above considerations of beam sizes and a small beam filling factor, the emission region sizes are limited to below 18*′′*, which correspond to 0.07$-$0.1 pc radii at the target source distances (1.67$-$3.0 kpc, Table 2).

Table 5 summarizes the derived values of the rotational temperatures and column densities. The rotational temperatures are derived to be 13.8$-$18 K, and the corresponding column densities are (2.05$-$2.78) $\times 10^{13}$ cm*-2*. The derived rotational temperatures with and without the beam-size correction are consistent within their $1\sigma$ errors. The rotational temperature and the column density in G10.30-0.15 were derived by the same manner as mentioned in the previous paragraph ($T_{\rm{rot}}=16\pm 4$ K, $N=7.5^{+0.8}_{-0.6}\times 10^{12}$ cm*-2*). We use the results with the beam-size correction in the following sections, because the fitting results are better.

\floattable

4 Discussion

4.1 Comparisons of the rotational temperature of HC5N

We compare the results in the high-mass star-forming regions with that in the cyanopolyyne peak in Taurus Molecular Cloud-1 (TMC-1 CP, $d=140$ pc), which is a well studied dark cloud and carbon-chain-species-rich source. The derived excitation temperature of HC5N in TMC-1 CP is $6.5\pm 0.2$ K (Taniguchi et al., 2016a). The rotational temperatures of HC5N in our three target high-mass star-forming regions except for G10.30-0.15, $\sim 13-20$ K, are significantly higher than the excitation temperature in the dark cloud TMC-1 CP.

The rotational temperature of HC5N in the low-mass star-forming region L1527 ($d=140$ pc) is known to be $14.7\pm 5.3$ K ($3\sigma$, Sakai et al., 2009b). The rotational temperatures of HC5N in our target high-mass star-forming regions are comparable with or slightly higher than that in L1527. L1527 is one of the WCCC sources, in which carbon-chain species are formed in the warm gas around the protostar with the temperature of $\sim 20-30$ K (Sakai & Yamamoto, 2013). Therefore, HC5N seems to exist in the warm gas in our target high-mass star-forming regions.

The detection of the emission lines with the high excitation energies also allows us to confirm the warm components of HC5N by itself. The detected emission lines of HC5N with the high-excitation energies ($E_{\rm{u}}/k\sim 90-100$ K) should not come from the cold environments with the gas kinetic temperature of 10 K. We assume that the column density of HC5N is comparable with that in TMC-1 CP ($N$(HC5N) = ($6.2\pm 0.3$) $\times 10^{13}$ cm*-2*, Taniguchi et al., 2016a) and the rotational temperature is a typical gas kinetic temperature in cold dark cloud (10 K). The peak intensities of the $J=38-37$ and 39$-$38 lines are calculated to be 0.8 mK in $T_{\rm{A}}^{*}$, even though the column density of the particular carbon-chain-rich source is considered. Thus, the detection of the emission lines of HC5N with the energies of $\sim 90-100$ K implies that HC5N exists in the warm gas. There is a possibility that the emission lines with the energies of $<20$ K come from not only the warm gas but also the cold gas. In that case, the derived rotational temperatures will be the lower limits due to the mixing of the warm gas and the cold gas in the single-dish beams. To summarize, we can conclude that HC5N exists in the warm gas within 0.07$-$0.1 pc radii around the massive young stellar objects.

4.2 Possibility of chemical differentiation in the high-mass star-forming cores

4.2.1 $N$(HC5N)/$W$(CH3OH) ratios in the four high-mass star-forming regions

We investigate the relationship between HC5N and CH3OH in the high-mass star-forming cores. CH3OH is one of the representative hot core tracers and COMs, and its origin is considered to be different from carbon-chain molecules. Figure 7 shows the comparison of the ratios between $N$(HC5N) the column density of HC5N and $W$(CH3OH) the integrated intensity of the thermal CH3OH emission line summarized in Table 3. We cannot derive the excitation temperature and column density of CH3OH due to the detection of only single line with low-excitation energy ($E_{\rm{u}}<50$ K). The other detected lines with high-excitation energies in G12.89+0.49 could possibly be masers (M$\ddot{\rm u}$ller et al., 2004). Therefore, we cannot use these lines to derive the excitation temperatures.

Since the column density of HC5N in G10.30-0.15 was derived using its average rotational temperature of the other three high-mass star-forming regions, we do not discuss the result of G10.30-0.15 in detail due to its large uncertainty. We can divide three high-mass star-forming regions into two groups; G28.28-0.36 shows the largest $N$(HC5N)/$W$(CH3OH) ratio of $>8.0\times 10^{14}$ in units of (K km s*-1*)-1 cm*-2*, while those of G12.89+0.49 and G16.86-2.16 are only ($1.9\pm 0.3$) $\times 10^{13}$ and ($2.6\pm 0.5$) $\times 10^{13}$, respectively. The ratio in G28.28-0.36 is significantly larger than those in G12.89+0.49 and G16.86-2.16 by one order of magnitude, even though there is a possibility that the chemical and physical structures in the beams cause the observed large difference. The $N$(HC5N) values are almost comparable among the three sources. Therefore, the differences in the $N$(HC5N)/$W$(CH3OH) ratio come from the $W$(CH3OH) values.

The integrated intensity is proportional to both the column density and the rotational temperature. When we assume the same $N$(HC5N)/$N$(CH3OH) ratios in all of the three sources, the differences in the $N$(HC5N)/$W$(CH3OH) ratios depend on only the rotational temperature of CH3OH. If the rotational temperature of CH3OH in G28.28-0.36 is the typical value in hot cores (100$-$250 K, Bisschop et al., 2007), the large difference in the integrated intensities of CH3OH by an order of magnitude (e.g., for the case of G28.28-0.36 and G16.86-2.16) cannot be explained given any rotational temperature assumed in G16.86-2.16. Alternatively, if the rotational temperatures of CH3OH in G12.89+0.49 and G16.86-2.16 are 100$-$250 K, G28.28-0.36 will have an unlikely high rotational temperature of CH3OH, 1200$-$2600 K. Thus, the differences in the $N$(HC5N)/$W$(CH3OH) ratio among the three sources cannot be explained only by changes in the rotational temperature of CH3OH, but the column density of CH3OH in G28.28-0.36 should be lower than those in G12.89+0.49 and G16.86-2.16. In summary, the observed variation in the $N$(HC5N)/$W$(CH3OH) ratio reflects the chemical differentiation in the single-dish beams.

4.2.2 Fractional abundances of HC5N in the four high-mass star-forming regions

We also derived the fractional abundances of HC5N defined as $X$(HC5N) = $N$(HC5N)/$N$(H2) toward the four sources. We calculated the column densities of H2 from the SCUBA, installed on the James Clerk Maxwell Telescope, 850 $\micron$ continuum data using the following formula (Kauffmann, 2005);

[TABLE]

The continuum data were obtained from the Canadian Astronomy Data Center, JCMT Science Archive666http://www.cadc-ccda.hia-iha.nrc-cnrc.gc.ca/en/jcmt/. We used the continuum fluxes ($S_{\nu}^{\rm{beam}}$) toward the observed positions and the beam size (18*′′) as summarized in Table 6777Proposal IDs are m05ai02 for G10.30-0.15, m03bc05 for G12.89+0.49, and m00au01 for G16.86-2.16 and G28.28-0.36, respectively.. We used 0.0182 cm2* g*-1* for $\kappa_{\nu}$ (Kauffmann, 2005), and 18*′′* for $\theta_{\rm{HPBW}}$. $T$ is the dust temperature obtained by SED fitting (Purcell et al., 2009), as summarized in Table 6. In G28.28-0.36, the dust temperature was not derived, and we used the average value of 58 K, determined from 65 good fits (Purcell et al., 2009). Table 6 summarizes the column densities of molecular hydrogen $N$(H2) and the $X$(HC5N) values. The errors of $N$(H2) are derived assuming that uncertainty of the dust temperatures is 20 K.

Figure 8 shows the comparison of the $X$(HC5N) values among the four sources. The $X$(HC5N) values show the same tendency as the $N$(HC5N)/$W$(CH3OH) ratios. G28.28-0.36 shows the larges $X$(HC5N) value ($4.2^{+2.6}_{-2.0}\times 10^{-9}$) in the four sources, and G12.89+0.49 and G16.86-2.16 show the similar values, which is smaller than that in G28.28-0.36. As mentioned before, the column density of HC5N in G10.30-0.15 was derived using the average rotational temperature of the three sources, and so we do not discuss in detail.

Both Figures 7 and 8 imply a possible signature of chemical differentiation among the high-mass star-forming cores, although we selected only four biased samples associated with HC5N emission and the 6.7 GHz methanol masers. We cannot exclude a possibility of the chemical and physical structures in the beam, and one possible explanation for such chemical differentiation is the different beam filling factor of hot core regions (Hatchell et al., 1998). Observations to reveal the spatial distributions of carbon-chain molecules and COMs with the interferometry are essential for revealing the origin of the chemical differentiation.

The fractional abundance of HC5N in L1527 was derived to be $\sim 2\times 10^{-10}$ ($N$(HC5N) = ($6.8\pm 1.4$)$\times 10^{12}$ cm*-2* and $N$(H2) = $2.8\times 10^{22}$ cm*-2* were derived by Sakai et al. (2009b) and Jørgensen et al. (2002), respectively). Even though there are uncertainties in column densities of H2 and HC5N due to assumed parameters (e.g. temperature, dust opacity, source size), the derive fractional abundances of HC5N in the three high-mass star-forming regions (except for G10.30-0.15) are comparable to or slightly higher than that in L1527 by a factor of 5$-$20. Such high HC5N abundances in our target sources suggest that the warm gas around these massive young stellar objects is rich in HC5N. Moreover, the high HC5N abundances would suggest efficient formation mechanisms of carbon chains in these high-mass star-forming regions (Aikawa et al. (2008) and Hassel et al. (2008) for WCCC and Chapman et al. (2009) for hot core).

\floattable

5 Conclusions

We carried out line survey observations at the 26$-$30 GHz band toward the four high-mass star-forming regions containing hot cores, G10.30-0.15, G12.89+0.49, G16.86-2.16, and G28.28-0.36, with the Robert C. Byrd Green Bank Telescope. We have detected HC5N from all of the sources, and HC7N from the three sources except for G10.30-0.15. The detection of HC7N means that these high-mass star-forming regions contain plenty of long cyanopolyynes compared to typical star-forming regions.

We also conducted observations of HC5N at the 42$-$46 GHz and 82$-$103 GHz bands toward the three high-mass star-forming regions, G12.89+0.49, G16.86-2.16, and G28.28-0.36, with the Nobeyama 45 m radio telescope. We have detected the rotational lines of HC5N with the high excitation energies ($E_{\rm{u}}/k\sim 63-100$ K), which are hardly excited in the cold dark clouds. We conducted the rotational diagram analysis, combining the GBT data with the Nobeyama data. The rotational temperatures of HC5N in the three high-mass star-forming regions are derived to be $\sim 13-20$ K. The detection of the lines with the high excitation energies and the derived rotational temperatures indicate that HC5N exists in the warm gas within 0.07$-$0.1 pc radii around the massive young stellar objects. The derived column densities of HC5N in the three sources are ($\sim 2.0-2.8$) $\times 10^{13}$ cm*-2*.

We compare the ratios between the column density of HC5N, $N$(HC5N), and the integrated intensity of the thermal CH3OH emission line, $W$(CH3OH) among the three high-mass star-forming regions. We found a possibility of the chemical differentiation in the three high-mass star-forming regions; G28.28-0.36 shows the largest $N$(HC5N)/$W$(CH3OH) ratio of $>8.0\times 10^{14}$ in units of (K km s*-1*)-1 cm*-2*, while G12.89+0.49 and G16.86-2.16 show the smaller values ($\sim 2\times 10^{13}$). Such chemical differentiation may originate from the complex chemical and physical structures in the beam.

We deeply appreciate the anonymous referee for constructive suggestions, which are very useful to develop our discussion. We would like to express our great thanks to the staff members of the National Radio Astronomy Observatory. The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc. K. T. deeply appreciates Dr. Nichol Cunningham (NRAO) for her kind help to prepare and carry out observations. We would like to express our thanks to the staff members of the Nobeyama Radio Observatory. The Nobeyama Radio Observatory is a branch of the National Astronomical Observatory of Japan, National Institutes of Natural Sciences.