A 3 mm Spectral Line Survey toward the Barred Spiral Galaxy NGC 3627

Observations toward SA and BE in NG 3627 were performed with the IRAM 30 m telescope at Pico Veleta in 2014 July and December. The observed positions are shown in Figure 1 and Table 1. The beam sizes are 30''–20'' and 19''–17'' at the 3 mm and 2 mm bands, respectively, which correspond to linear scales of 1.2–0.9 kpc and 0.8–0.7 kpc at a distance of 11.1 Mpc, respectively. The observed frequency range is from 85.0 to 116.0 GHz and from 140.0 to 148.0 GHz for SA, while it is from 85.0 to 100.5 GHz, from 108.4 to 116.0 GHz, and from 140.0 to 148.0 GHz for BE. The frequency range from 100.5 to 108.4 GHz does not involve strong spectral lines according to previous spectral line surveys toward the spiral arm of M51 (Watanabe et al. 2014). Therefore, we put a lower priority on this frequency range, and did not observe it for BE because of the limited observation time. Two EMIR (Eight MIxer Receivers) bands, E090 and E150, were used simultaneously with the dual polarization mode. The EMIR is the sideband separating receiver, which outputs the USB and LSB signals separately. The image rejection ratio is confirmed to be better than 13 dB and 10 dB for E090 and E150, respectively, according to the status report of the 30 m telescope.¹⁰ The system noise temperature ranged from 75 to 310 K for E090 and from 100 to 210 K for E150. Detailed frequency settings and system noise temperatures are summarized in Table 3. Backends were eight FTS (Fourier Transform Spectrometers) autocorrelators whose bandwidth and channel width are 4050 MHz and 195 kHz, respectively. The telescope pointing was checked every hour by observing the continuum source 1055+018 near the target position, and was found to be better than ±5''. The wobbler switching mode was employed with a beam throw of ±120'' and switching frequency of 0.5 Hz. The wobbler throw is toward the azimuth direction, hence the absolute off position depends on the hour angle. Nevertheless, the off position was always out of the CO disk of NGC 3627. The intensity scale was calibrated to the antenna temperature (${T}_{{\rm{A}}}^{* }$) scale using cold and hot loads. ${T}_{{\rm{A}}}^{* }$ was converted to the main beam temperature T_mb by multiplying F_eff/B_eff. Here, F_eff is the forward efficiency, which is 95% and 93% for the 3 mm and 2 mm bands, respectively, and B_eff is the main beam efficiency, which is 81% and 74% for the 3 mm and 2 mm bands, respectively. Spectral baselines were subtracted for each correlator band (4050 MHz) and for each scan by fitting a 7th–9th order polynomial to the line-free part before averaging all the scans. Then, the final spectrum for each correlator band was obtained by integrating all the scans. After the integration, the spectrum was smoothed to the frequency resolution of 5 MHz and 10 MHz for the 3 mm band and the 2 mm band, respectively (velocity resolution of ~15 km s⁻¹ and ~20 km s⁻¹, respectively).

Observations toward the nuclear region of NGC 3627 (NGC 3627 NR: Figure 1 and Table 1) were performed with the Nobeyama 45 m telescope in the 3 mm band in 2015 January and March. The beam size is 20''–15'', which corresponds to a linear scale of 1.2–0.9 kpc at the distance of 11.1 Mpc. The observed frequency range is from 85.0 to 115.5 GHz. The TZ1H/V receiver (Nakajima et al. 2008) was used as the front end with the dual polarization mode. It is the sideband separating receiver, which outputs the USB and LSB signals separately for the two linear polarizations. The image rejection ratio is assured to be better than 10 dB. The system temperature ranged from 110 to 250 K. Detailed frequency settings and system noise temperatures are summarized in Table 3. The backends were 16 SAM45 autocorrelators (Kamazaki et al. 2012) whose bandwidth and channel width each are ~1600 MHz and 0.5 MHz, respectively. The telescope pointing was checked every hour by observing the SiO maser source W-Leo, and was confirmed to be better than ±6''. The position switching mode was employed with an off position separated from the nuclear region by 10' along the azimuthal direction. The intensity scale was calibrated to the antenna temperature (${T}_{{\rm{A}}}^{* }$) scale using the chopper-wheel method. ${T}_{{\rm{A}}}^{* }$ was converted to the main beam temperature T_mb by dividing B_eff. Here, B_eff is the main beam efficiency, which is 41%, 37%, and 30% at 86 GHZ, 110 GHz, and 115 GHz, respectively. Spectral baselines were subtracted for each correlator band (~1600 MHz) and for each scan by fitting a 6th–9th order polynomial to the line-free part before averaging the scans. Then, the final spectra for each correlator band were obtained by integrating all the scans. After the integration, the spectrum was smoothed to the frequency resolution of 20 MHz corresponding to the velocity resolution of ~60 km s⁻¹ in the 3 mm band.

Figure 2 shows the composite spectra (Figures 2(a)–(c)) and those with a magnified vertical scale (Figures 2(d)–(f)) for NGC 3627 SA, NGC3627 BE, and NGC 3627 NR in the 3 mm band. Typical rms noise levels of the 3 mm spectra are 0.4–1.3 mK, 0.5–1.6 mK, and 0.8–3.2 mK for SA, BE, and NR, respectively. Figure 3 shows the expanded version of the 3 mm spectra. In addition to the 3 mm band, the 2 mm band (140–148 GHz) was observed for SA and BE, as shown in Figure 4. Typical rms noise levels of the 2 mm spectra are 0.4 mK and 0.5–0.6 mK for SA and BE, respectively. The frequency range, the frequency resolution, and the rms noise level for the overall spectra are summarized in Table 4. Line-of-sight velocities (V_LSR) of 655 km s⁻¹, 870 km s⁻¹, and 715 km s⁻¹ are used for SA, BE, and NR, respectively, in the spectra (Figures 2–4). These velocities are derived from the ¹³CO(J = 1–0) mapping observation toward NGC 3627 with the Nobeyama 45 m telescope (Watanabe et al. 2011).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Tables 5–7 summarize the parameters of emission lines detected in SA, BE, and NR, respectively. We identify molecular emission lines with the aid of the spectral line databases, the Cologne Database for Molecular Spectroscopy (CDMS) maintained by University of Cologne (Müller et al. 2001, 2005) and the Submillimeter, Millimeter and Microwave Spectral Line Catalog provided by Jet Propulsion Laboratory (Pickett et al. 1998). A criterion for the line detection is that the peak intensity and the integrated intensity exceed 3σ evaluated from the rms noise at the expected frequency of the line. If the peak intensity or the integrated intensity is less than 3σ, we regard the emission line as a marginal detection. In addition to the detected molecular lines, the upper limits to the peak intensity and the integrated intensity for important undetected lines are also given in Tables 5–7. V_LSR and the FWHMs of each emission line are derived by Gaussian fitting, except for CCH and CN. V_LSR and FWHM are not derived for the two molecules, since the spectral lines are blended by several hyperfine structure components. Although some lines show an asymmetric profile in BE, we simply applied Gaussian fitting to the spectra. As a result, V_LSR is sometimes different from the velocity of intensity peak, and the uncertainties of V_LSR and FWHM are slightly larger than those in SA. The derived V_LSR values are almost consistent with the V_LSR values obtained by the observations of ¹³CO(J = 1–0) (Watanabe et al. 2011) in the three positions (Table 1). Moreover, we also confirm that the V_LSR values are consistent with those obtained by the other observations such as CO(J = 1–0) (Sorai et al. 2019) and CO(J = 2–1) (Leroy et al. 2009).

Table 5.  Line Parameters in NGC 3627 SA

	Transition	Freq.	Eu	Sμ2	Tmb Peaka,b	$\int {T}_{\mathrm{mb}}{dv}$ a,c	VLSR,d	FWHMd
		(GHz)	(K)		(K)	(K km s−1)	(km s−1)	(km s−1)
c-C3H2	${2}_{12}\mbox{--}{1}_{01}$	85.338894	4.1e	16.1f	<0.001	<0.04	⋯	⋯
CH3CCH	50–40	85.457300	12.3	6.15	<0.001	<0.04	⋯	⋯
H42α		85.688390			<0.001	<0.04	⋯	⋯
H13CN	1–0	86.340163	4.1	8.91g	<0.001	<0.03	⋯	⋯
H13CO+	1–0	86.754288	4.2	15.2	<0.001	<0.03	⋯	⋯
SiO	2–1	86.846985	6.3	19.2	<0.001	<0.03	⋯	⋯
HN13C	1–0	87.090825	4.2	9.30g	<0.001	<0.03	⋯	⋯
CCHh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=2\mbox{--}1$	87.316898	4.2	0.988	0.003 (1)	0.23 (7)	i	⋯
CCHh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=1\mbox{--}0$	87.328585	4.2	0.492			⋯	⋯
CCHh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=1\mbox{--}1$	87.401989	4.2	0.492	0.001 (1)	0.10 (6)	i	⋯
CCHh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=0\mbox{--}1$	87.407165	4.2	0.198			⋯	⋯
HNCO	${4}_{04}\mbox{--}{3}_{03}$	87.925237	10.5	9.99g	<0.001	<0.04	⋯	⋯
HCN	1–0	88.631602	4.3	8.91g	0.007 (2)	0.39 (7)	663 (2)	58 (4)
HCO+	1–0	89.188525	4.3	15.2	0.008 (1)	0.46 (6)	658 (1)	59 (3)
HNC	1–0	90.663568	4.4	9.30g	0.003 (1)	0.13 (4)	657 (2)	41 (5)
HC3N	10–9	90.979023	24.0	139.3g	<0.001	<0.04	⋯	⋯
CH3CN	50–40	91.987088	13.2	153.8	<0.001	<0.03	⋯	⋯
H41α		92.034430			<0.001	<0.03	⋯	⋯
13CS	2–1	92.494308	6.7	15.3	<0.001	<0.04	⋯	⋯
N2H+	1–0	93.173770	4.5	104.1	<0.002	<0.04	⋯	⋯
C34S	2–1	96.412949	6.9	7.67	<0.001	<0.03	⋯	⋯
CH3OHh	2−1–1−1, E	96.739362	4.6j	1.21	<0.001	<0.04	⋯	⋯
CH3OHh	${2}_{0}\mbox{--}{1}_{0},{{\rm{A}}}^{+}$	96.741375	7.0	1.62			⋯	⋯
CH3OHh	${2}_{0}\mbox{--}{1}_{0},E$	96.744550	12.2j	1.62			⋯	⋯
OCS	8–7	97.301209	21.0	4.09	<0.001	<0.08	⋯	⋯
CS	2–1	97.980953	7.1	7.67	0.003 (1)	0.14 (5)	661 (3)	47 (7)
H40α		99.022950			<0.001	<0.03	⋯	⋯
SO	JN = 32–21	99.299870	9.2	6.91	<0.001	<0.03	⋯	⋯
HC3N	11–10	100.076392	28.8	153.2g	<0.001	<0.03	⋯	⋯
HC3N	12–11	109.173634	34.1	167.1g	<0.002	<0.04	⋯	⋯
OCS	9–8	109.463063	26.3	4.60	<0.002	<0.04	⋯	⋯
C18O	1–0	109.782173	5.3	0.0122	0.007 (2)	0.42 (7)	657 (1)	49 (3)
HNCO	${5}_{05}\mbox{--}{4}_{04}$	109.905749	15.8	12.5g	<0.005	<0.04	⋯	⋯
13CO	1–0	110.201354	5.3	0.0122	0.047 (2)	2.50 (7)	661.3 (3)	50.2 (7)
C17O	1–0	112.359284	5.4	0.0122	<0.002	<0.04	⋯	⋯
CNh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=1/2\mbox{--}3/2$	113.144157	5.4	1.25	0.003 (2)	0.15 (8)	i	⋯
CNh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=3/2\mbox{--}1/2$	113.170492	5.4	1.22			⋯	⋯
CNh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=3/2\mbox{--}3/2$	113.191279	5.4	1.58			⋯	⋯
CNh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=5/2\mbox{--}3/2$	113.490970	5.4	1.58	0.005 (2)	0.30 (8)	i	⋯
CNh	N = 1–0, J = 3/2–1/2, F = 3/2–1/2	113.488120	5.4	4.21			⋯	⋯
CNh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=1/2\mbox{--}1/2$	113.499644	5.4	1.25			⋯	⋯
CO	1–0	115.271202	5.5	0.0121	0.498 (4)	29.7 (2)	662.9 (3)	52.7 (7)
H2CO	${2}_{12}\mbox{--}{1}_{11}$	140.839502	6.7k	8.16f	0.0014 (9)	0.05 (3)	l	⋯
CH3OHh	${3}_{0}\mbox{--}{2}_{0},{\rm{E}}$	145.093707	19.2j	2.42	0.001 (1)	0.06 (5)	l	⋯
CH3OHh	${3}_{-1}\mbox{--}{2}_{-1},{\rm{E}}$	145.097370	11.6j	2.16			⋯	⋯
CH3OHh	${3}_{0}\mbox{--}{2}_{0},{{\rm{A}}}^{+}$	145.103152	13.9	2.43			⋯	⋯
H2CO	${2}_{02}\mbox{--}{1}_{01}$	145.602949	10.5	10.9	<0.001	<0.04	⋯	⋯
CS	3–2	146.969029	14.1	11.5	0.001 (1)	0.06 (6)	657 (7)	40 (20)

Notes.

^aThe numbers in parentheses represent 3σ errors. ^bUpper limits to the peak temperature are 3σ. ^cThe upper limit to the integrated intensity is calculated as: $\int {T}_{\mathrm{mb}}{dv}\lt 3\sigma \times \sqrt{{\rm{\Delta }}V\times {\rm{\Delta }}{v}_{\mathrm{res}}}$, where ΔV is the assumed line width (50 km s⁻¹) and ${\rm{\Delta }}{v}_{\mathrm{res}}$ is the velocity resolution per channel. ^dThe numbers in parentheses represent 1σ errors in units of the last significant digits. ^eThe upper state energy is calculated from the lowest ortho state (${1}_{01}$). ^fThe spin weight of 3 is not included. ^gThe nuclear spin multiplicity of 3 for the N nucleus is not included. ^hThe line is blended with other lines. ⁱGaussian fitting is not successful due to blending with other lines. ^jThe upper state energy is calculated from the lowest E state (${1}_{-1}$, E). ^kThe upper state energy is calculated from the lowest ortho state (${1}_{11}$). ^lGaussian fitting is not successful due to the poor S/N.

Download table as:  ASCIITypeset image

Table 6.  Line Parameters in NGC 3627 BE

	Transition	Freq.	Eu	Sμ2	Tmb Peaka,b	$\int {T}_{\mathrm{mb}}{dv}$ a,c	VLSR,d	FWHMd
		(GHz)	(K)		(K)	(K km s−1)	(km s−1)	(km s−1)
c-C3H2	${2}_{12}\mbox{--}{1}_{01}$	85.338894	4.1e	16.1f	<0.002	<0.09	⋯	⋯
CH3CCH	50–40	85.457300	12.3	6.15	<0.003	<0.1	⋯	⋯
H42α		85.688390			<0.002	<0.09	⋯	⋯
H13CN	1–0	86.340163	4.1	8.91g	<0.002	<0.09	⋯	⋯
H13CO+	1–0	86.754288	4.2	15.2	<0.002	<0.1	⋯	⋯
SiO	2–1	86.846985	6.3	19.2	<0.003	<0.1	⋯	⋯
HN13C	1–0	87.090825	4.2	9.30g	<0.003	<0.1	⋯	⋯
CCHh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=2\mbox{--}1$	87.316898	4.2	0.988	0.006 (2)	0.6 (1)	i	⋯
CCHh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=1\mbox{--}0$	87.328585	4.2	0.492			⋯	⋯
CCHh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=1\mbox{--}1$	87.401989	4.2	0.492	0.003 (2)	0.3 (1)	i	⋯
CCHh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=0\mbox{--}1$	87.407165	4.2	0.198			⋯	⋯
HNCO	${4}_{04}\mbox{--}{3}_{03}$	87.925237	10.5	9.99g	<0.003	<0.1	⋯	⋯
HCN	1–0	88.631602	4.3	8.91g	0.017 (3)	1.6 (2)	876 (4)	102 (8)
HCO+	1–0	89.188525	4.3	15.2	0.017 (3)	1.5 (2)	872 (3)	101 (8)
HNC	1–0	90.663568	4.4	9.30g	0.008 (3)	0.6 (1)	872 (6)	90 (10)
HC3N	10–9	90.979023	24.0	139.3g	<0.002	<0.09	⋯	⋯
CH3CN	50–40	91.987088	13.2	153.8	<0.003	<0.1	⋯	⋯
H41α		92.034430			<0.003	<0.1	⋯	⋯
13CS	2–1	92.494308	6.7	15.3	<0.003	<0.1	⋯	⋯
N2H+	1–0	93.173770	4.5	104.1	0.002 (1)	0.21 (8)	j	⋯
C34S	2–1	96.412949	6.9	7.67	0.002 (2)	0.13 (9)	j	⋯
CH3OHh	2−1–1−1, E	96.739362	4.6k	1.21	0.003 (2)	0.28 (8)	863 (5)	110 (10)
CH3OHh	${2}_{0}\mbox{--}{1}_{0},{{\rm{A}}}^{+}$	96.741375	7.0	1.62			⋯	⋯
CH3OHh	${2}_{0}\mbox{--}{1}_{0},E$	96.744550	12.2k	1.62			⋯	⋯
OCS	8–7	97.301209	21.0	4.09	<0.001	<0.05	⋯	⋯
CS	2–1	97.980953	7.1	7.67	0.006 (2)	0.52 (9)	868 (5)	110 (10)
H40α		99.022950			<0.002	<0.06	⋯	⋯
SO	${J}_{N}={3}_{2}\mbox{--}{2}_{1}$	99.299870	9.2	6.91	0.002 (2)	0.13 (8)		⋯
HC3N	11–10	100.076392	28.8	153.2g	<0.001	<0.05	⋯	⋯
HC3N	12–11	109.173634	34.1	167.1g	<0.002	<0.06	⋯	⋯
OCS	9–8	109.463063	26.3	4.60	<0.002	<0.06	⋯	⋯
C18O	1–0	109.782173	5.3	0.0122	0.011 (2)	0.9 (1)	883 (3)	82 (7)
HNCO	${5}_{05}-{4}_{04}$	109.905749	15.8	12.5g	<0.008	<0.07	⋯	⋯
13CO	1–0	110.201354	5.3	0.0122	0.057 (2)	4.9 (1)	870 (3)	103 (7)
C17O	1–0	112.359284	5.4	0.0122	0.003 (2)	0.1 (1)	892 (5)	40 (10)
CNh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=1/2\mbox{--}3/2$	113.144157	5.4	1.25	0.006 (3)	0.7 (2)	i	⋯
CNh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=3/2\mbox{--}1/2$	113.170492	5.4	1.22			⋯	⋯
CNh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=3/2\mbox{--}3/2$	113.191279	5.4	1.58			⋯	⋯
CNh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=5/2\mbox{--}3/2$	113.490970	5.4	1.58	0.012 (3)	1.3 (1)	i	⋯
CNh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=3/2\mbox{--}1/2$	113.488120	5.4	4.21			⋯	⋯
CNh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=1/2\mbox{--}1/2$	113.499644	5.4	1.25			⋯	⋯
CO	1–0	115.271202	5.5	0.0121	0.497 (4)	52.9 (2)	862 (3)	115 (6)
H2CO	${2}_{12}\mbox{--}{1}_{11}$	140.839502	6.7l	8.16f	0.003 (1)	0.31 (8)	868 (5)	100 (10)
CH3OHh	${3}_{0}\mbox{--}{2}_{0},{\rm{E}}$	145.093707	19.2k	2.42	0.003 (1)	0.15 (8)	877 (3)	34 (9)
CH3OHh	${3}_{-1}\mbox{--}{2}_{-1},{\rm{E}}$	145.097370	11.6k	2.16			⋯	⋯
CH3OHh	${3}_{0}\mbox{--}{2}_{0},{{\rm{A}}}^{+}$	145.103152	13.9	2.43			⋯	⋯
H2CO	${2}_{02}\mbox{--}{1}_{01}$	145.602949	10.5	10.9	0.002 (1)	0.17 (8)	874 (7)	80 (20)
CS	3–2	146.969029	14.1	11.5	0.005 (1)	0.42 (6)	878 (4)	90 (10)

Notes.

^aThe numbers in parentheses represent 3σ errors. ^bUpper limits to the peak temperature are 3σ. ^cThe upper limit to the integrated intensity is calculated as $\int {T}_{\mathrm{mb}}{dv}\lt 3\sigma \times \sqrt{{\rm{\Delta }}V\times {\rm{\Delta }}{v}_{\mathrm{res}}}$, where ΔV is the assumed line width (100 km s⁻¹) and Δv_res is the velocity resolution per channel. ^dThe numbers in parentheses represent 1σ errors in units of the last significant digits. ^eThe upper state energy is calculated from the lowest ortho state (${1}_{01}$). ^fThe spin weight of 3 is not included. ^gThe nuclear spin multiplicity of 3 for the N nucleus is not included. ^hThe line is blended with other lines. ⁱGaussian fitting is not successful due to blending with other lines. ^jGaussian fitting is not successful due to the poor S/N. ^kThe upper state energy is calculated from the lowest E state (1₋₁, E). ^lThe upper state energy is calculated from the lowest ortho state (${1}_{11}$).

Download table as:  ASCIITypeset image

Table 7.  Line Parameters in NGC 3627 NR

	Transition	Freq.	${E}_{{\rm{u}}}$	${\rm{S}}{\mu }^{2}$	${T}_{\mathrm{mb}}$ Peaka,b	$\int {T}_{\mathrm{mb}}{dv}$ a,c	${V}_{\mathrm{LSR}}$ d	FWHMd
		(GHz)	(K)		(K)	(K km s−1)	(km s−1)	(km s−1)
c-C3H2	${2}_{12}\mbox{--}{1}_{01}$	85.338894	4.1e	16.1f	<0.006	<0.7	⋯	⋯
CH3CCH	50–40	85.457300	12.3	6.15	<0.006	<0.7	⋯	⋯
H42α		85.688390			<0.004	<0.5	⋯	⋯
H13CN	1–0	86.340163	4.1	8.91g	<0.006	<0.6	⋯	⋯
H13CO+	1–0	86.754288	4.2	15.2	<0.004	<0.5	⋯	⋯
SiO	2–1	86.846985	6.3	19.2	<0.003	<0.4	⋯	⋯
HN13C	1–0	87.090825	4.2	9.30g	<0.003	<0.3	⋯	⋯
CCHh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=2\mbox{--}1$	87.316898	4.2	0.988	0.004 (3)	1.2 (6)	i	⋯
CCHh	$N=1\mbox{--}0,J=3/2\mbox{--}1/2,F=1\mbox{--}0$	87.328585	4.2	0.492			⋯	⋯
CCHh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=1\mbox{--}1$	87.401989	4.2	0.492			⋯	⋯
CCHh	$N=1\mbox{--}0,J=1/2\mbox{--}1/2,F=0\mbox{--}1$	87.407165	4.2	0.198			⋯	⋯
HNCO	${4}_{04}\mbox{--}{3}_{03}$	87.925237	10.5	9.99g	<0.003	<0.3	⋯	⋯
HCN	1–0	88.631602	4.3	8.91g	0.021 (4)	6.0 (8)	710 (6)	240 (20)
HCO+	1–0	89.188525	4.3	15.2	0.017 (3)	4.6 (6)	703 (9)	250 (20)
HNC	1–0	90.663568	4.4	9.30g	0.007 (3)	1.8 (5)	740 (20)	260 (40)
HC3N	10–9	90.979023	24.0	139.3g	<0.004	<0.5	⋯	⋯
CH3CN	50–40	91.987088	13.2	153.8	<0.003	<0.4	⋯	⋯
H41α		92.034430			<0.003	<0.4	⋯	⋯
13CS	2–1	92.494308	6.7	15.3	<0.003	<0.4	⋯	⋯
N2H+	1–0	93.173770	4.5	104.1	<0.003	<0.3	⋯	⋯
C34S	2–1	96.412949	6.9	7.67	<0.004	<0.5	⋯	⋯
CH3OHh	${2}_{-1}\mbox{--}{1}_{-1},E$	96.739362	4.6j	1.21	0.006 (4)	1.3 (6)	740 (20)	240 (60)
CH3OHh	${2}_{0}\mbox{--}{1}_{0},{{\rm{A}}}^{+}$	96.741375	7.0	1.62			⋯	⋯
CH3OHh	${2}_{0}\mbox{--}{1}_{0},E$	96.744550	12.2j	1.62			⋯	⋯
OCS	8–7	97.301209	21.0	4.09	<0.002	<0.3	⋯	⋯
CS	2–1	97.980953	7.1	7.67	0.012 (1)	2.9 (2)	740 (6)	210 (10)
H40α		99.022950			<0.004	<0.4	⋯	⋯
SO	JN = 32–21	99.299870	9.2	6.91	<0.004	<0.5	⋯	⋯
HC3N	11–10	100.076392	28.8	153.2g	0.003 (1)	0.4 (1)	790 (8)	100 (20)
HC3N	12–11	109.173634	34.1	167.1g	<0.005	<0.5	⋯	⋯
OCS	9–8	109.463063	26.3	4.60	<0.004	<0.5	⋯	⋯
C18O	1–0	109.782173	5.3	0.0122	0.008 (5)	1.7 (8)	720 (30)	340 (90)
HNCO	${5}_{05}-{4}_{04}$	109.905749	15.8	12.5g	<0.005	<0.5	⋯	⋯
13CO	1–0	110.201354	5.3	0.0122	0.042 (5)	10.6 (9)	728 (4)	220 (10)
C17O	1–0	112.359284	5.4	0.0122	<0.006	<0.7	⋯	⋯
CNh	N = 1–0, J = 1/2–1/2, F = 1/2–3/2	113.144157	5.4	1.25	0.013 (8)	3 (1)	i	⋯
CNh	N = 1–0, J = 1/2–1/2, F = 3/2–1/2	113.170492	5.4	1.22			⋯	⋯
CNh	N = 1–0, J = 1/2–1/2, F = 3/2–3/2	113.191279	5.4	1.58			⋯	⋯
CNh	N = 1–0, J = 3/2–1/2, F = 5/2–3/2	113.490970	5.4	1.58	0.018 (8)	5 (1)	i	⋯
CNh	N = 1–0, J = 3/2–1/2, F = 3/2–1/2	113.488120	5.4	4.21			⋯	⋯
CNh	N = 1–0, J = 3/2–1/2, F = 1/2–1/2	113.499644	5.4	1.25			⋯	⋯
CO	1–0	115.271202	5.5	0.0121	0.626 (7)	144 (1)	737 (3)	219 (7)

Notes.

^aThe numbers in parentheses represent 3σ errors. ^bUpper limits to the peak temperature are 3σ. ^cThe upper limit to the integrated intensity is calculated as $\int {T}_{\mathrm{mb}}{dv}\lt 3\sigma \times \sqrt{{\rm{\Delta }}V\times {\rm{\Delta }}{v}_{\mathrm{res}}}$, where ΔV is the assumed line width (220 km s⁻¹) and ${\rm{\Delta }}{v}_{\mathrm{res}}$ is the velocity resolution per channel. ^dThe numbers in parentheses represent 1σ errors in units of the last significant digits. ^eThe upper state energy is calculated from the lowest ortho state (${1}_{01}$). ^fThe spin weight of 3 is not included. ^gThe nuclear spin multiplicity of 3 for the N nucleus is not included. ^hThe line is blended with other lines. ⁱGaussian fitting is not successful due to blending with other lines. ^jThe upper state energy is calculated from the lowest E state (1₋₁, E).

Download table as:  ASCIITypeset image

Figures 5–7 display profiles of the lines listed in Tables 5–7. In SA, 14 emission lines of 8 molecular species (CCH, CN, CO, HCN, HNC, HCO⁺, H₂CO, and CS) and two isotopologues (¹³CO and C¹⁸O) are detected in the 3 and 2 mm bands. In addition to these molecules, the emission line of CH₃OH is seen at 145.1 GHz with a signal-to-noise ratio (S/N) of 3 in the peak intensity and the integrated intensity. However, we do not count CH₃OH in the detected molecule because another emission line expected at 96.74 GHz is not detected. In BE, 20 emission lines of 11 molecular species (CCH, CN, CO, HCN, HNC, HCO⁺, H₂CO, CH₃OH, N₂H⁺, CS, and SO) and 4 isotopologues (¹³CO, C¹⁸O, C¹⁸O, and C³⁴S) are detected in the 3 mm and 2 mm bands. In NR, 12 emission lines of 8 molecular species (CCH, CN, HCN, HNC, CO, HCO⁺, CH₃OH, and CS) and two isotopologues (¹³CO and C¹⁸O) are detected in the 3 mm band. Although one line of HC₃N (J = 11–10; 100.076392 GHz) is detected with the S/N of 3 and 4 in the peak intensity and the integrated intensity, respectively, no other transition lines, such as J = 10–9 and J = 12–11, are detected in NR. Therefore, we treat HC₃N as a marginal detection.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The undetected lines in Tables 5–7 include the lines of the isotopologues whose normal species lines are strongly detected, such as H¹³CN, HN¹³C, H¹³CO⁺, ¹³CS, and C³⁴S. The tables also include the lines detected with relatively strong intensity in the nuclear region of other nearby galaxies such as c-C₃H₂, CH₃CCH, CH₃CN, HNCO, and OCS (e.g., Martín et al. 2009; Lindberg et al. 2011; Nakajima et al. 2011, 2018; Aladro et al. 2013, 2015). The hydrogen recombination lines of H42α, H41α, and H40α are also included as the undetected lines in Tables 5–7.

Column densities of molecules are derived by assuming optically thin emission and local thermodynamic equilibrium conditions. Since only one transition line is detected for each molecule except for BE, a rotation temperature cannot be derived from the observations. We therefore adopt the three rotation temperatures, 10, 15, and 20 K, because a kinetic temperature of molecular gas is evaluated to be 10–20 K in the disk region of NGC 3627 (Law et al. 2018). A temperature of cold dust in NGC 3627 is also found to fall in this range (Galametz et al. 2012).

Before deriving the column densities, the integrated intensities are divided by the beam filling factor ${\theta }_{\mathrm{source}}^{2}/({\theta }_{\mathrm{source}}^{2}+{\theta }_{\mathrm{beam}}^{2})$, where θ_source and θ_beam are the source size and the FWHM of the telescope beam, respectively, to compensate for frequency dependence of the beam size. Here, the source size (θ_source) is assumed to be 10'', because distributions of HCO${}^{+}(J=1\mbox{--}0)$ are resolved to be ~10'' with ALMA in BE and NR (Murphy et al. 2015). We assume the same source size in SA for simplicity. The θ_beam value is 2460/f arcsec and 1710/f arcsec for the IRAM 30 m telescope and the Nobeyama 45 m telescope, respectively, where f is the observation frequency in GHz.

The column densities are evaluated by the following formula, assuming the optically thin condition (e.g.,: Goldsmith & Langer 1999; Yamamoto 2017):

$$\begin{eqnarray}\begin{array}{rcl}N & = & \displaystyle \frac{3{W}_{\nu }{k}_{{\rm{B}}}U({T}_{\mathrm{rot}})}{8{\pi }^{3}S{\mu }_{0}^{2}\nu }\exp \left(\displaystyle \frac{{E}_{{\rm{u}}}}{{k}_{{\rm{B}}}{T}_{\mathrm{rot}}}\right)\\ & & \times \,{\left\{1-\displaystyle \frac{\exp (h\nu /{k}_{{\rm{B}}}{T}_{\mathrm{rot}})-1}{\exp (h\nu /{k}_{{\rm{B}}}{T}_{\mathrm{bg}})-1}\right\}}^{-1},\end{array}\end{eqnarray} \tag{ 1 }$$

where W_ν is integrated intensity, k_B is the Boltzmann constant, U(T) is the partition function, T_rot is the assumed rotation temperature, S is the line strength, μ₀ is the dipole moment, ν is the frequency, E_u is the upper state energy, h is the Planck constant, and T_bg is the cosmic microwave background temperature of 2.7 K. The derived column densities are summarized in Tables 8–10 for SA, BE, and NR, respectively. The partition function U(T) is numerically calculated by the following formula:

$$\begin{eqnarray}&&U(T)=\displaystyle \sum _{i}{g}_{i}\exp \left(-\displaystyle \frac{{E}_{i}}{{k}_{{\rm{B}}}T}\right),\end{eqnarray} \tag{ 2 }$$

where g_i and E_i are the statistical weight and the energy for the ith state. The energy levels up to 300 K or more listed in CDMS are involved in the calculation. We confirm consistency between our values and those in CDMS at temperatures of 9.375 and 18.75 K.

In addition to the column densities, the fractional abundances relative to the molecular hydrogen (H₂) are evaluated. For this purpose, the column densities of H₂ are derived from the column densities of C¹⁸O by assuming the [C¹⁸O]/[H₂] ratio of 1.7 × 10⁻⁷ (Frerking et al. 1982; Goldsmith et al. 1997). Here, we assume that the [C¹⁸O]/[H₂] ratio in NGC 3627 is the same as that in our Galaxy, because the elemental abundance in NGC 3627 is also similar to that in our Galaxy (Contini 2017). The derived fractional abundances are also summarized in Tables 8–10 for SA, BE, and NR, respectively. Figure 8 shows histograms of the fractional abundances derived at T_rot = 10 K for the three positions.

Zoom In Zoom Out Reset image size

Variations of the fractional abundances by changing the assumed rotation temperature from 10 to 20 K are smaller than those of the column densities for most molecules. This is because the temperature effects are mostly compensated by taking the ratio of the two column densities. In fact, variations of the fractional abundances are less than 10% except for CH₃OH, c-C₃H₂, CH₃CCH, HC₃N, and OCS, while those of the column densities are 10%–30%. Although variation of the fractional abundances is slightly higher for CH₃OH, c-C₃H₂, and CH₃CCH, it is evaluated to be ~20% or less. On the other hand, the fractional abundances of HC₃N and OCS show larger variation by a factor of two, because their column densities are derived using the lines with relatively high upper state energies of ~20 K. As shown in Equation (1), the column densities are proportional to $\exp ({E}_{{\rm{u}}}/{k}_{{\rm{B}}}{T}_{\mathrm{rot}})$, hence the column densities are sensitive to the assumed rotation temperature for the lines with ${E}_{{\rm{u}}}/{k}_{{\rm{B}}}\gt {T}_{\mathrm{tot}}$. However, this effect does not cause a serious impact on the following discussion, because only upper limits are derived for HC₃N and OCS. In the following sections, we will use the fractional abundances relative to H₂ or C¹⁸O to compare chemical compositions in different sources.

The SFR (Figure 1(c)) is calculated using the Hα and Spitzer 24 μm data (Kennicutt et al. 2003). Before the calculation, the Hα image is convolved by a Gaussian kernel to be the angular resolution of 5'' corresponding to the angular resolution of the 24 μm data. The calibration of SFR is done by the following equation (Calzetti et al. 2007):

$$\begin{eqnarray}\begin{array}{rcl}\mathrm{SFR} & = & 5.3\times {10}^{-42}[L{({\rm{H}}\alpha )}_{\mathrm{obs}}+0.031\\ & & \times L(24\ \mu {\rm{m}})]({M}_{\odot }\,{\mathrm{yr}}^{-1}),\end{array}\end{eqnarray} \tag{ 3 }$$

where L(Hα) and L(24 μm) are luminosities of the Hα and 24 μm emission, respectively, in erg s⁻¹. The surface density of SFR (Σ_SFR) is evaluated within the observation beam of 22'', 22'', and 16'' for SA, BE, and NR, respectively. The surface density of molecular gas mass Σ_gas is calculated from the column density of H₂ derived from the C¹⁸O data assuming the rotation temperature of 10 K. The contribution of He is involved in Σ_gas by multiplying a factor of 1.36. Here, the source size of 10'' is not taken into account for both Σ_SFR and Σ_gas. The SFE is calculated as Σ_SFR/Σ_gas. The Σ_SFR, Σ_gas, and SFE values are summarized in Table 2.

We calculate integrated intensity ratios of the CO isotopologues in the three positions. The ¹²CO(J = 1–0)/¹³CO(J = 1–0) intensity ratios are 11.9 ± 0.3, 10.8 ± 0.2, and 14 ± 1 in SA, BE, and NR, respectively. These values are consistent with the intensity ratios in NGC 3627 previously observed with the Nobeyama 45 m telescope (Watanabe et al. 2011) and the IRAM 30 m telescope (Cormier et al. 2018). Law et al. (2018) reported the ¹²CO(J = 2–1)/¹³CO(J = 2–1) ratios to be ~4 and 2.5 ± 0.2 in BE and NR, respectively, from the SMA observation at a resolution of 233 × 185. Although the ratios are for the J = 2–1 transition, these values are much lower than our values. As discussed in Law et al. (2018), ¹²CO can trace much more diffuse molecular gas than that of ¹³CO and be detectable everywhere. As the result, the ratios obtained with a larger beam size tend to be larger.

However, this trend does not always work for any case. For instance, Tan et al. (2011) observed the ¹²CO(J = 1–0)/¹³CO(J = 1–0) ratio with the PMO 14 m telescope toward various positions in NGC 3627. Although their ratio is consistent with our value in NR, their value of 5.3 ± 1.3 at the position covering BE is significantly lower than our value of 10.8 ± 0.2 at BE in spite of a larger observation beam (60''). This implies the ¹²CO/¹³CO ratio is sensitive to the physical properties and the distributions of the molecular cloud involved in the beam.

The ¹³CO(J = 1–0)/C¹⁸O(J = 1–0) intensity ratios observed in the present study are 6 ± 1, 5.4 ± 0.6, and 6 ± 3 in SA, BE, and NR, respectively. The ratio for NR is similar to the mean intensity ratio of 6.0 ± 0.9 for the nuclear regions of nearby galaxies (Jiménez-Donaire et al. 2017). On the other hand, the ratios in SA and BE are found to be lower than the mean intensity ratio of 9.0 ± 1.1 in the disk region of NGC 3627 (Jiménez-Donaire et al. 2017).

The lower ratios in SA and BE would originate from the optical depth effect. If the optical depth of ¹³CO is higher in SA and BE than in the other disk regions, the ratio would be lower in SA and BE. In our observation, the optical depths of the ¹³CO line are evaluated to be 0.018 for both SA and BE with the temperature of 15 K and the assumed source size of 10''. However, the optical depth of the ¹³CO line is likely higher than the above value, because the distribution of molecular gas is expected to be clumpy. In addition to the high angular resolution observation of ¹²CO in NGC 3627 with the radio interferometers (Gallagher et al. 2018a, 2018b; Law et al. 2018; Sun et al. 2018), observations of the CO isotopologues have recently been conducted with the ALMA at a scale of 1'' (Gallagher et al. 2018a). The result will be useful for examining the possibility of a clumpy structure in the future.

Another possibility is different ¹⁸O abundances for the different positions. Jiménez-Donaire et al. (2017) reported that the intensity ratios are lower in the vicinity of active star-forming regions than quiescent regions among the nearby galaxies. They claimed that the trend can be explained, for instance, by enhancement of the ¹⁸O abundance by nucleosynthesis in the massive stars. Since the SFR is relatively higher in SA and BE than the other disk regions in NGC 3627 (Table 2 and Watanabe et al. 2011), the active star formation might result in lower ¹³CO(J = 1–0)/C¹⁸O(J = 1–0) intensity ratios in SA and BE in principle. However, it is not obvious whether such an effect could have an observable effect on the ratios. Note that, for this reason, we employ the standard [C¹⁸O]/[H₂] ratio to derive the H₂ column density, as mentioned in Section 3.3.

Figure 9(a) shows a correlation plot of the fractional abundances between SA and BE, which are derived under the assumption of a rotation temperature of 10 K. Although the SFE is higher in BE than in SA by a factor of 5 (Table 2), the chemical compositions are similar. In fact, the Pearson correlation coefficient is found to be 0.999. Here, we do not take into account the uncertainties in the calculation. In addition, we exclude the upper limit values from the correlation-coefficient calculations. The correlation coefficient changes by 0.001, even if the correlation coefficients are calculated from the fractional abundances assuming different rotation temperatures. Because of the high SFE in BE, star formation feedback such as UV radiation from high-mass stars and shocks by outflows from protostars is expected to be prominent in BE, which consequently affects the chemical composition in BE in comparison with those in SA. However, no significant effects of such star formation feedback can be seen in the chemical composition observed at a scale of ~1 kpc. The feedback effects would be diluted by the large observation beam, while the chemical composition of extended molecular gas makes a dominant contribution to the spectrum. Similar results have been reported in observations of nearby galaxies and Galactic molecular clouds (e.g., Watanabe et al. 2014, 2017; Nishimura et al. 2016a, 2017).

Zoom In Zoom Out Reset image size

On the other hand, most of the fractional abundances are found to be systematically higher in BE than SA by factors of 2–3, except for CCH and ¹³CO. The difference cannot be explained by the assumed beam filling factors. Even if we assume a smaller source size of 1'' instead of 10'', the fractional abundances become higher by only a factor of 1.1. Therefore, the higher fractional abundances indicate that there is more dense molecular gas in BE than SA, because these molecular transition lines likely trace denser molecular gas than the gas traced by CO isotopologues due to their higher critical densities (e.g., Nishimura et al. 2017).

However, the relation of the higher fraction of dense gas to higher SFE in BE is not straightforward, because the relation itself is still controversial. Muraoka et al. (2016) reported that a volume density of H₂ (${n}_{{{\rm{H}}}_{2}}$) is correlated with SFE in their observation of the nearby galaxy NGC 2903 with the Nobeyama 45 m telescope. Since high ${n}_{{{\rm{H}}}_{2}}$ is expected in the region with the high dense gas fraction, the high dense gas fraction is a possible origin of the high SFE. On the other hand, Law et al. (2018) found no correlation between the ${n}_{{{\rm{H}}}_{2}}$ and SFE in NGC 3627 at a scale of kpc. In addition, Usero et al. (2015) also reported that no correlation can be seen between SFE and the dense gas fraction evaluated from CO(1–0) and HCN(1–0) observed with the IRAM 30 m telescope in nearby galaxies. Therefore, we need more investigations of the relation in NGC 3627.

While the critical density of CCH is higher than those of CO isotopologues, the fractional abundance in SA is similar to that in BE. This could be a result of the wider spatial distribution of CCH compared to those of the other molecules except for CO and its isotopologues. In fact, the extended CCH distributions are found in the Galactic molecular clouds (e.g., Nishimura et al. 2017; Pety et al. 2017). Nishimura et al. (2017) conducted a wide-field mapping observation of the Galactic molecular cloud W3(OH), and found that the CCH/HCO⁺ ratio is higher in the outer regions of molecular clouds than in the central parts. According to Nishimura et al. (2017), the abundance of CCH is high in the periphery of a molecular cloud where the UV photon can penetrate, because CCH is efficiently produced from C⁺ ionized by the interstellar UV field. Another reason could be the excitation conditions necessary for producing the CCH emission. Because the dipole moment of CCH (0.77 Debye) is smaller than those of other molecules such as HCN (2.99 Debye) and CS (1.96 Debye), the critical density of CCH is lower than these molecules (e.g., Nishimura et al. 2017). Therefore, CCH can trace more diffuse gas than the other molecules due to chemical and physical reasons.

Here, we compare the chemical composition in NGC 3627 BE with those in a spiral arm in M51 and a giant molecular cloud W51 in our Galaxy. Since these sources are molecular clouds located in a disk region of galaxies, their chemical compositions are free from the effects of nuclear activities such as AGNs and starbursts. The chemical composition in SA is similar to that in BE, hence we focus only on BE in this section.

M51 is a nearby face-on spiral galaxy (d = 8.4 Mpc Feldmeier et al. 1997; Vinkó et al. 2012). The elemental abundance in M51 (N/O:~0.25, S/O:0.03: Bresolin et al. 2004) is similar to that in NGC 3627 (N/O:~ 0.23, S/O:~ 0.025: Contini 2017). Figure 9(b) is a correlation plot of the molecular abundances between BE and M51 P1 (Watanabe et al. 2014). The rotation temperature assumed for BE is 10 K, and that for M51 P1 is 5 K. The plot shows a good correlation with a correlation coefficient of 0.989. The fractional abundances in BE are thus similar to those in M51 P1 for most of the species. Because the beam size of the observation (~25'') corresponds to ~1 kpc at the distance of M51, the spatial scale of the M51 observation is similar to that of our observation toward NGC 3627. It is reported that the SFE is lower in M51 P1 (5.9 × 10⁻¹⁰ yr⁻¹) than in BE (5.7 × 10⁻⁹ yr⁻¹) by a factor of 10 (Watanabe et al. 2014). In spite of the different SFE, the chemical compositions in BE and M51 P1 are similar to each other. Therefore, the chemical compositions of molecular gas at a scale of ~1 kpc are thought to be unaffected by the star formation activities.

Figure 9(c) shows a correlation plot between BE and the Galactic molecular cloud W51 (Watanabe et al. 2017). The rotation temperature assumed for BE is 10 K. W51 is the molecular-cloud complex with the vigorous massive star-forming regions. The fractional abundances in W51 were obtained from the spectrum stacked over the 39 pc × 47 pc area centered at the W51A molecular cloud mapped with the Mopra 22 m telescope. Therefore, the fractional abundances in W51 represent those at the scale of a molecular cloud. The elemental abundance in the Galaxy is also similar to that in NGC 3627 (Contini 2017). A strong positive correlation (0.989) seen in Figure 9(c) indicates that the chemical composition in BE is also similar to the molecular-cloud-scale (~50 pc) chemical composition in the spiral arm of our Galaxy.

Figures 9(d) and (e) are correlation plots of the fractional abundances between SA and NR, and between BE and NR, where the rotation temperature of 10 K is assumed. Although the fractional abundances are higher in NR than SA and BE by factors of ~5 and ~2, respectively, the correlation coefficients are as high as 0.991–0.993 for both plots. The strong positive correlation suggests that the chemical compositions are similar to each other between BE and NR, although NR possesses a LINER/Seyfert 2 nucleus (Ho et al. 1997; Filho et al. 2000). The X-ray luminosity in the 2–10 keV band is reported to be 3.2 × 10³⁹ erg s⁻¹, which is lower than that of prototypical Seyfert nuclei in NGC 1068 by three orders of magnitude (Brightman & Nandra 2011). Therefore, the effect of nuclear activities would be limited to very small region (<100 pc), and would not significantly affect the chemical composition in the major part of the molecular gas in the beam.

In spite of the overall similarity, we can see small differences between SA and NR and between BE and NR. For example, the fractional abundances are higher in NR than SA and BE, except for CCH and N₂H⁺. The trend of higher fractional abundances in NR is not artificially caused by the assumption of the low rotation temperature of 10 K, although higher rotation temperatures were reported in the nuclear region of other galaxies at a scale of ~1 kpc (e.g., Aladro et al. 2011, 2013; Nakajima et al. 2018). As seen in Table 10, the fractional abundances derived by assuming the temperature of 20 K are similar to or slightly higher than those derived by assuming the temperature of 10 K. The similar trend can be reproduced by assuming higher temperature (e.g., 100 K), except for HC₃N. Therefore, even if the rotation temperature in NR is higher than 10 K due to the nuclear activity, we will see the same trend of higher fractional abundances in NR. The higher fractional abundances could be a result of a higher fraction of denser molecular gas in NR than in the other positions. On the other hand, the abundances of CCH are similar among the three regions (Figures 9(d) and (e)). As discussed in Section 4.1, CCH traces widely extended molecular gas as CO and its isotopologues rather than dense gas traced by other molecules. Hence, the trend of CCH in Figures 9(d) and (e) seems reasonable.

The N₂H⁺ abundance in NR is found to be similar to or lower than that in BE (Figure 9(e)). This result indicates inefficient production of N₂H⁺ and/or efficient destruction working in NR. There are two major destruction processes for N₂H⁺; the reaction with CO and the dissociative recombination with an electron. However, the destruction by CO would not be the cause of the different N₂H⁺ abundances, because the fractional abundances of various molecules, for which H₂ column density is evaluated from C¹⁸O, are higher in NR than in SA and BE. If the CO fractional abundance were enhanced in NR, the fractional abundances of the other molecules would be calculated to be lower in NR than in SA and BE. Moreover, the dissociative recombination would not be the cause, either. Because the SFE is derived to be higher in BE than in NR, the UV radiation field is expected to be higher in BE than NR. The electron abundance is consequently expected to be higher in BE than NR, which contradicts the lower N₂H⁺ abundance in NR. However, the electron abundance could be enhanced by strong X-ray radiation in NR due to nuclear activities, which may contribute to destruction of N₂H⁺ in NR.

Pety et al. (2017) reported that N₂H⁺ traces the densest regions of the Orion B cloud, while HCN and HCO⁺ trace the less dense molecular gas with an H₂ density of ~1500 cm⁻³. They discussed that N₂H⁺ can only survive in cold and dense gas where CO is frozen on the dust grain. Kauffmann et al. (2017) also reported that N₂H⁺(1–0) traces the dense gas at a characteristic H₂ density of ~4000 cm⁻³, while HCN(1–0) can trace rather diffuse gas at a H₂ density of ~870 cm⁻³, on the basis of the observations of the Orion molecular cloud. If this picture is applied to the present case, a fraction of "cold dense gas" is lower in NR than in BE. In any case, the lower abundance of N₂H⁺ in NR seems to suggest some important physical conditions. Interferometric studies revealing the distributions of N₂H⁺ and other molecules are awaited for further discussions.

Here, we compare the chemical composition of NR with that of the Seyfert 2 type AGN NGC 1068, because NGC 3627 possesses LINER or low-luminous Seyfert 2 AGNs (Ho et al. 1997; Filho et al. 2000). The AGNs are thought to be powered by accretion of material to the super-massive black hole at the center of galaxy and to emit strong X-rays. As a result, the chemical compositions of molecular gas around the AGNs are expected to be affected by it (X-ray dominated region) (e.g., Lepp & Dalgarno 1996; Maloney et al. 1996; Meijerink & Spaans 2005; Meijerink et al. 2007). The AGN activity is thought to be higher in NGC 1068 than in NGC 3627, because the intrinsic luminosity of X-rays in the 2–10 keV band is reported to be higher in NGC 1068 (1.6 × 10⁴² erg s⁻¹) than in NGC 3627 (3.2 × 10³⁹ erg s⁻¹) by three orders of magnitude (Brightman & Nandra 2011).

Figure 9(f) shows a correlation plot of fractional abundances relative to C¹⁸O between NR and the nuclear region of NGC 1068 (Aladro et al. 2013, 2015). Here, the rotation temperature assumed for NR is 10 K. Aladro et al. (2015) derived column densities of molecules in NGC 1068 using the spectral line survey data observed with the IRAM 30 m telescope by assuming the rotation temperature of 10 K. The fractional abundances of NGC 1068 are therefore the ones averaged over ~1.5 kpc. The correlation coefficient of the plot is as high as 0.84. Although it is slightly lower than those among the disk regions in nearby galaxies and our Galaxy (W51), it indicates that the chemical compositions are similar to each other despite different X-ray luminosities. The similarity implies that the effect of X-rays is diluted by an overwhelming contribution of ambient molecular clouds due to the large observation beam (~1 kpc).

The fractional abundances are generally found to be higher in NGC 1068 than NGC 3627 NR by a factor of two except for CH₃OH, HC₃N, and CS, although the chemical compositions are almost similar to each other as a whole. The higher abundances indicate a higher fraction of dense molecular gas in NGC 1068 than NGC 3627 NR. On the other hand, the abundances of CH₃OH, CS, and HC₃N are higher in NGC 3627 NR than NGC 1068 by a factor of two, although HC₃N is a tentative detection. It is reported that CH₃OH and CS are enhanced by liberation from ice mantle of dust grains due to shocks in Galactic sources such as, for instance, outflow shocks (e.g., Bachiller & Pérez Gutiérrez 1997). In external galaxies, the enhancement of CH₃OH is reported in the bar region, where shocks due to cloud–cloud collisions are expected (Meier & Turner 2005). It has also been seen in shocked regions caused by interactions of molecular gas in merging galaxies (Saito et al. 2017; Ueda et al. 2017). However, no sign of shocks such as outflows from the nuclear region and active star formation is reported for NGC 3627 NR even by high angular resolution observations with interferometers (Casasola et al. 2011; Murphy et al. 2015; Law et al. 2018). Therefore, the shock could not be the case for the enhancement, although there may exist accretion shocks in the nuclear region of NGC 3627 in principle. Another possibility is that these molecules could more efficiently be destroyed in NGC 1068 by strong radiation from the AGN. However, CS, CH₃OH, and HC₃N are found to be associated with the circumnuclear disk in NGC 1068 by an ALMA observation (Takano et al. 2014). Thus, destruction by the radiation would also not explain the differences. Overabundant CH₃OH and CS in NGC 3627 should be investigated in future studies.

The chemical compositions of molecular clouds at a scale from 1 kpc to 10 pc have been studied for NGC 3627 and M51 (Watanabe et al. 2014, 2016), as well as the Galactic molecular clouds W51 (Watanabe et al. 2017) and W3 (Nishimura et al. 2017). All these sources have similar elemental abundances. From these studies, we find that the chemical composition of molecular gas is similar among different galaxies and among different regions with different star formation/nuclear activities, when we observe them in the 3 mm band at a scale larger than a molecular cloud (>10 pc). Figures 10(a)–(e) show the spectra observed in NGC 3627 SA, BE, NR, M51 P1, and W51 in the 3 mm bands. Indeed, the relative intensities of molecular lines look similar among the five sources.

Zoom In Zoom Out Reset image size

In contrast, the spectrum of the hot core (Orion KL: Figure 9(f)) observed at a scale of 0.04 pc is much different from the other spectra. The difference reflects a different chemical composition in the hot core from those in molecular clouds as well as a different excitation condition: namely hotter and denser gas in the hot core than in molecular clouds. At a scale of a molecular-cloud core (~0.01 pc), the chemical composition is largely controlled by its evolutionary state (e.g., Suzuki et al. 1992) and ambient UV environments (e.g., Tielens & Hollenbach 1985). Therefore, we can make use of the chemical compositions as tracers of their evolutionary stages and physical environments in the Galactic objects. However, such chemical features of molecular-cloud cores are smeared out when we observe them at scales larger than the molecular clouds. The sound-crossing timescale of the molecular clouds (>10⁷ yr), which is estimated by assuming a typical size of 10 pc (e.g.,: Roman-Duval et al. 2010) and a sound speed of 0.3 km s⁻¹, is longer than the timescale of chemical equilibrium (~10^5–6 yr). Here, the sound speed c_s is calculated by the following formula:

$$\begin{eqnarray}&&{c}_{{\rm{s}}}=\sqrt{\gamma \displaystyle \frac{{k}_{{\rm{B}}}}{\mu {m}_{{\rm{H}}}}T},\end{eqnarray} \tag{ 4 }$$

where γ is the specific heat ratio of gas (5/3), μ is the molecular mass, and m_H is the hydrogen mass. We evaluated the sound speed with μ = 2 and T = 10 K. Therefore, the chemical composition in the molecular clouds is thought to be close to that in the chemical equilibrium state, which is almost independent of the evolutionary history of molecular clouds.

Harada et al. (2019) recently reported that the chemical compositions of the molecular clouds can be reproduced by their time-dependent chemical model with a relatively short time of ~10⁵ yr. Because this timescale is much shorter than the lifetime of the molecular cloud and the timescale for chemical equilibrium, they suggested that a turbulence in the molecular clouds regulates the chemical clock. Namely, the chemical composition is refreshed by the interstellar UV radiation when the turbulence brings the molecular gas from the inside to the surface of the molecular cloud where the visual extinction is low.

In spite of the similarity of the chemical composition at a scale of molecular clouds, a different chemical composition has been found in metal-poor galaxies (Nishimura et al. 2016a, 2016b), shocked regions of merging galaxies (e.g., Saito et al. 2017; Ueda et al. 2017), and nuclear regions of galaxies observed at a scale of a few hundred parsecs (e.g., Costagliola et al. 2015; Meier et al. 2015; Ando et al. 2017). In these objects, the chemical compositions are thought to reflect low elemental abundances, gas dynamics at molecular-cloud scales, and extreme environments due to feedback from nuclear activity and vigorous star formation. To highlight chemical characteristics in these objects, the result of this study can be used as a "standard" template of chemical composition without such influences.