Physical conditions in star-forming regions around S235

Abstract

Gas density and temperature in star-forming regions around Sh2-235 are derived from ammonia line observations. This information is used to evaluate formation scenarios and to determine evolutionary stages of the young embedded clusters S235 East 1, S235 East 2 and S235 Central. We also estimate the gas mass in the embedded clusters and its ratio to the stellar mass. S235 East 1 appears to be less evolved than S235 East 2 and S235 Central. In S235 East 1 the molecular gas mass exceeds that in the other clusters. Also, this cluster is more embedded in the parent gas cloud than the other two. Comparison with a theoretical model shows that the formation of these three clusters could have been stimulated by the expansion of the Sh2-235 H ii region (hereafter S235) via a collect-and-collapse process, provided the density in the surrounding gas exceeds 3 × 10³ cm⁻³, or via collapse of pre-existing clumps. The expansion of S235 cannot be responsible for star formation in the southern S235 A-B region. However, formation of the massive stars in this region might have been triggered by a large-scale supernova shock. Thus, triggered star formation in the studied region may come in three varieties, namely collect-and-collapse and collapse of pre-existing clumps, both initiated by expansion of the local H ii regions, and triggered by an external large-scale shock. We argue that the S235 A H ii region expands into a highly non-uniform medium with increasing density. It is too young to trigger star formation in its vicinity by a collect-and-collapse process. There is an age spread inside the S235 A-B region. Massive stars in the S235 A-B region are considerably younger than lower mass stars in the same area. This follows from the estimates of their ages and the ages of associated H ii regions.

stars: formation, H ii regions, ISM: molecules, open clusters and associations: general

1 INTRODUCTION

Among the various modes of star formation, triggered star formation attracts a special interest due to an increasing availability of multi-wavelength observations of expanding shells in the Milky Way (Churchwell et al. 2006; Simpson et al. 2012). The idea of self-stimulating star formation is quite old (Elmegreen & Lada 1977). However, there are still only a few examples of star-forming regions, where triggering can be inferred more or less reliably (Deharveng, Zavagno & Caplan 2005; Dewangan et al. 2012). The reason is that expanding shells often have irregular shapes, so it is hard to deduce an evolutionary sequence based on morphology alone.

In our previous paper (Kirsanova et al. 2008, hereafter Paper I) we presented arguments in favour of triggered star formation around an extended H ii region, Sh2-235 (hereafter S235), using both morphological and kinematical data. S235 is located in the Perseus Spiral Arm towards the anticentre of the Galaxy at |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}8$| and δ₂₀₀₀ = +35°48^′27|${^{\prime\prime}_{.}}$|0. The region surrounding S235 is a site of active star formation. Young star clusters in its vicinity were recently investigated, e.g. by Allen et al. (2005), Klein et al. (2005), Kumar, Keto & Clerkin (2006), Paper I, Camargo, Bonatto & Bica (2011) and Dewangan & Anandarao (2011). In Paper I we argued that the formation of three clusters, located close to the edge of S235, could have been triggered by an expansion of the H ii region via a ‘collect-and-collapse’ (C&C) process due to gravitational instability in a dense material accumulated between the ionization front and the shock wave. In the present paper we refer to these clusters as S235 East 1, S235 East 2 and S235 Central, following the notation introduced in Paper I. Their cross-identification can be found in Camargo et al. (2011).

The inhomogeneously distributed molecular gas around the S235 H ii region (Heyer, Carpenter & Ladd 1996) complicates an analysis of the triggering scenario. In Paper I arguments in favour of triggering were found on the basis of (1) relative positions of the three embedded star clusters, dense molecular clumps and the H ii region itself, and (2) kinematics of molecular gas emitting ¹³CO (1–0) and CS (2–1) lines. It was shown that the young star clusters S235 East 1, S235 East 2 and S235 Central are embedded in dense molecular gas which forms a shell-like structure at the south-eastern side of S235. Each cluster is embedded in a separate dense clump, which is clearly visible on a CS (2–1) map. The morphology and location of the sources and the distribution of the molecular gas are shown in Fig. 1. Note that S235 is not completely surrounded by the dense shell. There is some gas emitting in ¹²CO (1–0) and ¹³CO (1–0) lines to the north and west from S235 (Heyer et al. 1996). However, the ¹³CO (1–0) brightness significantly decreases in these directions, so we had not observed that area during observing sessions described in Paper I. Also, the column density of molecular gas located in front of the H ii region decreases to the south of S235 as this gas becomes transparent to the optical emission. In Paper I we proposed that expansion of S235 triggered the formation of a dense shell and subsequent formation of three young embedded stellar clusters.

Figure 1.

Star-forming regions around S235. Reference position is |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8 and δ₂₀₀₀ = +35°48^′27|${^{\prime\prime}_{.}}$|0 in both panels. Distance to S235 is about 1.8 kpc (Evans & Blair 1981), so that 1 arcmin corresponds to ∼0.5 pc. Left-hand panel: map of ¹³CO(1–0) emission from Paper I. A cross marks location of the exciting star. Levels of grey-scale correspond to 5, 15, 25, 35, 45 and 55 K  km s⁻¹. Contour levels show an optical image of the nebulae taken from the Digitized Sky Survey (POSS 2 red filter). The levels correspond to 30, 50, 70 and 90 per cent of the peak value. We removed point sources from the contour image to show the nebulae clearer. H ii regions S235 and S235 A as well as the reflecting nebulae S235 B are marked by arrows. Dashed circles indicate regions that were observed with the Pushchino 22-m telescope (RT-22). Thick black solid lines outline regions observed with the Effelsberg 100-m telescope. Sizes of dashed and small solid circles correspond to the FWHP beam sizes of the RT-22 and the Effelsberg telescopes, respectively. Right panel: map of CS(2–1) emission (grey) and stellar density at 2 μm (contours) taken from Paper I. The grey-scale levels correspond to 1, 3, 5, 7, 9 and 11 K km s⁻¹. Levels of stellar density are given from 20 to 60 stars arcmin⁻² with a step of four stars arcmin⁻². An asterisk marks location of IRAS source 05382+3547.

Three kinematic components were distinguished in Paper I that may represent basic stages of interaction of young stellar objects with the surrounding medium. The first component that we refer to as primordial arguably comprises quiescent clumpy gas of the parent giant molecular cloud G174+2.5 that has not yet been disturbed by the expansion of S235. The second kinematic component represents gas that was entrained in motion and compressed by the expansion of the H ii region. This is the gas out of which the young star clusters have been formed. This gas still partially surrounds them. The third kinematic component supposedly corresponds to gas blown away by stellar winds from S235 East 2 and S235 Central (see Fig. 1, right panel). It was proposed in Paper I that these two clusters are, probably, more evolved than S235 East 1.

The star-forming region S235 A-B encompasses both S235 A and S235 B nebulae and is located almost 10 arcmin to the south of S235 Central. The S235 A nebula is an expanding H ii region (Israel & Felli 1978; Felli et al. 1997). Boley et al. (2009) have shown that S235 B is a reflection nebula. A water maser was found towards the region between S235 A and S235 B (Lo, Burke & Haschick 1975) suggesting that massive star formation takes place there. Stellar mass estimates by Dewangan & Anandarao (2011) indicate that this water maser is associated with a young stellar object of high mass. The S235 A-B region is not located in the compressed shell around the S235 H ii region. In Paper I we suggested that star formation in S235 A-B is not related to the expansion of S235.

This paper presents new observational data related to the star-forming regions S235 East 1, S235 East 2, S235 Central in the immediate vicinity of S235, and to regions around the S235 A and S235 B nebulae. The aim of this work is to determine physical conditions in these star-forming regions. Ammonia emission was observed to estimate gas temperature and number density as this molecule is known to be both a thermometer and a densitometer for dense molecular gas (Ho & Townes 1983; Walmsley & Ungerechts 1983). The derived physical structure turns out to be consistent with our star formation scenario in this region and the inferred evolutionary stages of the young embedded star clusters.

2 OBSERVATIONS AND DATA REDUCTION

Ammonia emission was mapped towards the S235 East 1, S235 East 2 and S235 Central star clusters as well as towards S235 A-B and primordial molecular gas at other locations.

2.1 Pushchino (RT-22) observations

We first searched for (J, K) = (1,1) and (2,2) emission (ν = 23.694496 and 23.722631 GHz) towards three selected positions with the Pushchino 22-m telescope (RT-22, Astro Space Centre of Lebedev Physical Institute, Pushchino, Russian Academy of Sciences) in 2006 April. Rest frequencies of observed ammonia transitions were taken from the catalogue of Lovas (1992). We used an ON–ON observing technique based on beam switching (Berulis et al. 1983) with a beam throw of 10 arcmin at 23 GHz. The main beam efficiency was 0.38 and the calibration uncertainty is estimated to be 10–15 per cent. We used the 50 MHz band of the Pushchino autocorrelator with 2048 channels. The spectral resolution was about 24 kHz. The full width at half power (FWHP) beam size of the 22-m telescope is about 2.6 arcmin which is larger than the size of the individual sources. Therefore these observations are suitable to estimate average parameters in the region. The ammonia lines were detected towards the three selected positions, shown in Fig. 1 with dashed circles. The typical system temperature was about 150 K on an antenna temperature (⁠|$T_{\rm A}^*$|⁠) scale. Pointing was checked twice a day. The values of introduced corrections ranged from 0 arcsec to 20 arcsec.

2.2 Effelsberg observations

Emission from the (J, K) = (1,1), (2,2), (3,3) and (4,4) ammonia transitions was mapped with the 100-m Effelsberg telescope in 2007 April and October. We observed areas in star-forming regions around S235 that are bright in the CS (2–1) emission and host young embedded star clusters. Long integrations were also performed towards locations of primordial gas. The regions observed in ammonia are shown in Fig. 1 (left panel).

The FWHP of the Effelsberg telescope beam is about 40 arcsec at 23 GHz. We used an 8192 channel correlator (AK90) split into eight separate bands with a bandwidth of 10 MHz and 1024 channels. Thus we were able to obtain data on all four ammonia lines in two orthogonal linear polarizations simultaneously. The chosen backend configuration provided a channel spacing of 0.12 km s⁻¹ (9.77 kHz). Observations were done in a position switching mode. The off-position was chosen to be located outside the giant molecular cloud G174+2.5 (Heyer et al. 1996) and has offset coordinates (Δα, Δδ)= (−15 arcmin, 0 arcmin) with respect to our reference position of |$\alpha _{2000}=05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8 and δ₂₀₀₀ = +35°48^′27|${^{\prime\prime}_{.}}$|0. We checked the telescope pointing every hour. Corrections for azimuth and elevation were typically <5 arcsec, but sometimes ∼10 arcsec. The focus was checked after sunrise and sunset. Corrections for focusing were of the order of 1 mm. Data calibration was performed by observing the calibrators NGC 7027, 3C123 and 3C286. A standard 22 GHz gain curve (Gallimore et al. 2001) was used to account for a change of the telescope gain with elevation. Absolute and relative calibration uncertainties were estimated to be 15 per cent and 5 per cent, respectively. We spent about 12 min of on-time towards the majority of the positions. Positions of particular interest were measured twice. The map spacing was chosen to be 40 arcsec. The system temperature was 200–300 K on a T_mb scale. The average noise level was 0.118 K on a T_mb scale.

Observations of the star-forming region S235 A-B were carried out in 2007 October. Frontend and backend configurations were the same as for the previous observations, but the map spacing was reduced to 20 arcsec implying full sampling. The position of reference for the S235 A-B spectra is |$\alpha _{2000} = 05^{\rm h}40^{\rm m}53{^{\mathrm{s}}_{.}}$|4 and δ₂₀₀₀ = +35°41^′48|${^{\prime\prime}_{.}}$|0, the location of the HCO⁺(1–0) emission peak (Felli et al. 2004). It is situated to the south of the S235. The mapped area is also shown in Fig. 1.

Data obtained with the RT-22 and the Effelsberg telescope were reduced with the gildas software.¹ We used gildas routine ‘nh3(1,1)’ to estimate the optical depth τ_1,1 of the ammonia (J, K) = (1,1) line. Spectra of other ammonia transitions were fitted with the ‘gauss’ routine to determine their intensity and width.

3 RESULTS

The standard interpretation of ammonia observations, described by Stutzki & Winnewisser (1985b), was utilized to determine ammonia column density |$N_{\rm NH_3}$| and rotational temperature T_rot. We calculated the gas number density (⁠|$n_{\rm H_2}$|⁠, essentially, number density of molecular hydrogen), using equation (2) from Ho & Townes (1983), towards positions with determined values of τ_1,1 and NH₃ excitation temperature T_ex. Rate coefficients for rotational excitation from Danby et al. (1988) were used to determine the kinetic temperature T_kin at positions with known T_rot. If data quality at a certain position was insufficient to determine τ_1,1, i.e. the uncertainty of the value was higher than the value itself, we did not use corresponding τ_1,1 and T_ex estimates for further analysis.

Ammonia spectra, measured with the Pushchino 22-m telescope, are shown in Fig. 2. Satellites of the NH₃(1,1) line are seen towards all observed positions, while the hyperfine structure of the NH₃(2,2) line is not pronounced. The quality of our spectra allows estimating T_kin towards all young clusters under study. The values are presented in Table 1. The gas number density could be determined towards S235 Central because we could determine T_ex towards this position. Numbers in parentheses in all tables show the values of uncertainties. Table 1 indicates that T_kin in the direction of S235 East 1 is considerably lower than T_kin towards S235 East 2 and S235 Central. As it was mentioned already, results of observations with RT-22 represent physical parameters averaged over a large area.

Figure 2.

Spectra of NH₃(1,1) and NH₃(2,2) emission towards selected positions, obtained with the RT-22 telescope. The two lines were observed simultaneously within a single band. Solid lines show spectra fitted with the gildas ‘nh3(1,1)’ and ‘gauss’ routines for NH₃(1,1) and NH₃(2,2) lines, respectively. All spectra are Hanning-smoothed. For the reference position, see Fig. 1.

Table 1.

Observational results and physical parameters derived from RT-22 telescope data. Reference position is |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8 and δ₂₀₀₀ = +35°48^′27|${^{\prime\prime}_{.}}$|0.

Offset	(J,K)	Tmb	Width	Vlsr	τ	Tkin	n
(arcsec, arcsec)		(K)	(km s−1)	(km s−1)		(K)	(103 cm−3)
S235 East 1
(−40, 20)	(1,1)	0.08 (0.02)	3.04 (0.29)	−20.00 (0.16)	–	23.38 (8.44)	–
	(2,2)	0.04 (0.02)	3.47 (0.72)	−20.31 (0.38)	–
S235 East 2
(−80,240)	(1,1)	0.08 (0.03)	2.43 (0.47)	−19.53 (0.20)	–	32.25 (20.16)	–
	(2,2)	0.05 (0.03)	3.29 (0.98)	−20.15 (0.38)	–
S235 Central
(−350,40)	(1,1)	0.15 (0.03)	2.24 (0.26)	−20.40 (0.10)	0.98 (0.64)	30.24 (11.58)	0.49
	(2,2)	0.09 (0.03)	1.88 (0.42)	−20.62 (0.21)	–

With the Effelsberg telescope, we detected NH₃(1,1) and NH₃(2,2) lines towards almost all selected positions, while the (3,3) transition was detected in 10 of 43 positions. The (4,4) line remained undetected. Representative NH₃(1,1), NH₃(2,2) and NH₃(3,3) lines are shown in Fig. 3. The hyperfine structure of NH₃(1,1) emission provides a possibility to determine the optical depth of the line. The NH₃(1,1) emission lines are optically thin in most of the observed positions. Observational results obtained with the Effelsberg telescope are presented in Table 2 . The signal-to-noise ratio allowed estimating T_kin towards 37 positions. The gas number density was estimated in 15 positions. Several positions were identified where intensity ratios of hyperfine components differ from their LTE values. This information can be used to put some additional constraints on the cloud structure (e.g. Section 3.2).

Figure 3.

Spectra of NH₃(1,1) (left), NH₃(2,2) (middle) and NH₃(3,3) (right) emission towards three star-forming clumps (three rows of spectra from the top) and primordial gas (bottom) of the S235 molecular complex. The spectra towards the offsets (−40 arcsec, 40 arcsec), (−120 arcsec, 240 arcsec) and (−320 arcsec, 40 arcsec) correspond to the S235 East 1, S235 East 2 and S235 Central clumps, respectively. The spectra are Hanning-smoothed.

Table 2.

Parameters of (J, K) = (1,1), (2,2) and (3,3) ammonia lines measured with the Effelsberg 100-m telescope.

Offset	(J, K)	Tmb	Width	Vlsr	τ
(arcsec, arcsec)		(K)	(km s−1)	(km s−1)
S235 East 1, ref. pos. |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8, δ2000 = +35°48′27|${^{\prime\prime}_{.}}$|0
0, 160	(1,1)	0.55 (0.17)	2.14 (0.28)	−21.39 (0.10)	–
	(2,2)	< 0.20	–	–	–
	(3,3)	< 0.20	–	–	–
0, 120	(1,1)	0.77 (0.18)	2.34 (0.25)	−20.90 (0.11)	–
	(2,2)	<0.18	–	–	–
	(3,3)	<0.20	–	–	–
0, 80	(1,1)	0.77 (0.20)	1.95 (0.21)	−19.95 (0.09)	–
	(2,2)	0.17 (0.16)	3.78 (1.11)	−18.97 (0.63)	–
	(3,3)	<0.20	–	–	–
0, 40	(1,1)	0.50 (0.19)	1.52 (0.22)	−19.37 (0.11)	–
	(2,2)	<0.18	–	–	–
	(3,3)	<0.20	–	–	–
−40, 160	(1,1)	0.52 (0.18)	1.06 (0.43)	−20.70 (0.09)	–
	(2,2)	<0.19	–	–	–
	(3,3)	<0.20	–	–	–
−40, 120	(1,1)	1.16 (0.17)	2.30 (0.14)	−20.87 (0.07)	0.16 (0.09)
	(2,2)	0.41 (0.08)	2.42 (0.21)	−20.78 (0.09)	–
	(3,3)	0.21 (0.08)	3.59 (0.38)	−20.17 (0.17)	–
−40, 80	(1,1)	2.38 (0.16)	1.95 (0.08)	−20.20 (0.03)	0.41 (0.19)
	(2,2)	0.37 (0.08)	2.22 (0.68)	−20.10 (0.20)	–
	(3,3)	< 0.24	–	–	–
−40, 40	(1,1)	1.87 (0.06)	1.97 (0.03)	−19.62 (0.01)	0.29 (0.08)
	(2,2)	0.67 (0.06)	2.21 (0.10)	−19.56 (0.04)	0.10 (0.07)
	(3,3)	0.31 (0.09)	2.43 (0.30)	−19.57 (0.13)	–
−80, 120	(1,1)	0.23 (0.21)	2.69 (1.08)	−14.31 (0.51)	–
	(2,2)	<0.33	–	–	–
	(3,3)	<0.20	–	–	–
−80, 80	(1,1)	0.53 (0.18)	1.54 (0.31)	−19.78 (0.11)	–
	(2,2)	<0.21	–	–	–
	(3,3)	<0.20	–	–	–
−80, 40	(1,1)	1.10 (0.11)	1.73 (0.07)	-19.21 (0.03)	0.29 (0.21)
	(2,2)	0.45 (0.08)	1.81 (0.19)	-19.14 (0.07)	–
	(3,3)	< 0.24	–	–	–
−80, 0	(1,1)	1.27 (0.18)	1.27 (0.09)	-19.06 (0.04)	–
	(2,2)	0.49 (0.08)	1.11 (0.23)	-19.20 (0.10)	–
	(3,3)	<0.24	–	–	–
S235 East 2, ref. pos. |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8, δ2000 = +35°48′27|${^{\prime\prime}_{.}}$|0
−40, 240	(1,1)	0.67 (0.19)	1.68 (0.19)	−21.14 (0.09)	–
	(2,2)	<0.17	–	–	–
	(3,3)	<0.20	–	–	–
−40, 200	(1,1)	0.61 (0.06)	1.91 (0.10)	-21.24 (0.04)	–
	(2,2)	0.19 (0.05)	1.58 (0.36)	-21.35 (0.12)	–
	(3,3)	0.07 (0.03)	2.22 (0.49)	-20.84 (0.16)	–
−80, 200	(1,1)	0.55 (0.11)	1.78 (0.16)	-20.96 (0.06)	0.60 (0.44)
	(2,2)	0.23 (0.08)	1.89 (0.37)	-20.66 (0.14)	–
	(3,3)	<0.10	–	–	–
−80, 240	(1,1)	1.33 (0.34)	1.27 (0.12)	-21.28 (0.12)	–
	(2,2)	0.37 (0.14)	1.31 (0.31)	-21.32 (0.10)	–
	(3,3)	<0.30	–	–	–
−80, 280	(1,1)	0.97 (0.10)	1.48 (0.09)	-21.37 (0.03)	–
	(2,2)	0.30 (0.06)	1.18 (0.21)	-21.10 (0.08)	–
	(3,3)	<0.18	–	–	–
−120, 200	(1,1)	0.90 (0.16)	1.34 (0.18)	-21.13 (0.07)	–
	(2,2)	0.48 (0.07)	0.84 (0.18)	-20.96 (0.05)	–
	(3,3)	<0.15	–	–	–
−120, 240	(1,1)	1.58 (0.30)	1.19 (0.12)	-21.10 (0.04)	–
	(2,2)	0.49 (0.06)	1.44 (0.20)	-21.19 (0.08)	–
	(3,3)	0.14 (0.07)	1.61 (0.98)	-21.39 (0.38)	–
−120, 280	(1,1)	1.13 (0.16)	1.47 (0.15)	-21.00 (0.05)	–
	(2,2)	<0.16	–	–	–
	(3,3)	<0.16	–	–	–
−160, 200	(1,1)	0.18 (0.07)	1.42 (0.40)	-20.99 (0.16)	–
	(2,2)	0.17 (0.06)	1.56 (0.25)	-20.90 (0.10)	–
	(3,3)	0.16 (0.04)	0.44 (0.25)	-20.90 (0.06)	–
S235 Central, ref. pos. |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8, δ2000 = +35°48′27|${^{\prime\prime}_{.}}$|0
−280, 80	(1,1)	0.75 (0.13)	1.44 (0.12)	−19.97 (0.05)	–
	(2,2)	0.33 (0.09)	1.47 (0.34)	−19.90 (0.11)	–
	(3,3)	0.18 (0.11)	0.98 (0.42)	−19.92 (0.20)	–
−280, 40	(1,1)	0.35 (0.10)	2.84 (0.12)	−20.26 (0.11)	0.98 (0.67)
	(2,2)	0.23 (0.11)	1.24 (0.43)	−20.81 (0.19)	–
	(3,3)	< 0.20	–	–	–
−320, 80	(1,1)	1.12 (0.17)	1.46 (0.10)	−20.08 (0.04)	1.05 (0.35)
	(2,2)	0.50 (0.05)	1.54 (0.17)	−20.01 (0.07)	−
	(3,3)	< 0.15	–	–	−
−320, 40	(1,1)	1.45 (0.08)	1.13 (0.03)	−20.18 (0.01)	0.90 (0.12)
	(2,2)	0.45 (0.07)	1.24 (0.09)	−20.08 (0.04)	−
	(3,3)	0.11 (0.06)	1.56 (0.48)	−20.00 (0.19)	−
−320, 0	(1,1)	0.35 (0.17)	2.03 (0.50)	−18.95 (0.24)	−
		0.38 (0.17)	1.40 (0.35)	−21.01 (0.17)	−
	(2,2)	0.16 (0.15)	1.41 (0.66)	−20.27 (0.35)	−
	(3,3)	< 0.16	–	–	−
−360, 80	(1,1)	0.53 (0.17)	2.41 (0.28)	−20.23 (0.12)	−
	(2,2)	<0.18	–	–	−
	(3,3)	<0.18	–	–	−
−360, 40	(1,1)	0.53 (0.11)	2.24 (0.19)	−20.42 (0.08)	−
	(2,2)	0.25 (0.09)	2.22 (0.28)	−20.26 (0.11)	−
	(3,3)	< 0.10	–	–	−
−360, 0	(1,1)	0.56 (0.16)	0.82 (0.19)	−18.65 (0.07)	−
		0.28 (0.16)	2.41 (0.58)	−20.78 (0.25)	−
	(2,2)	<0.14	–	–	−
	(3,3)	<0.17	–	–	−
−400, 80	(1,1)	0.42 (0.17)	3.06 (0.38)	−21.52 (0.17)	−
	(2,2)	<0.17	–	–	−
	(3,3)	<0.20	–	–	−
−400, 40	(1,1)	0.93 (0.13)	1.66 (0.17)	−21.56 (0.06)	−
	(2,2)	<0.18	–	–	−
	(3,3)	<0.20	–	–	−
−400, 0	(1,1)	0.79 (0.20)	2.55 (0.23)	−20.80 (0.09)	0.61 (0.45)
	(2,2)	0.20 (0.18)	1.44 (0.51)	−21.17 (0.31)	−
	(3,3)	<0.20	–	–	−
−440, 80	(1,1)	0.37 (0.16)	1.54 (0.42)	−21.73 (0.15)	−
	(2,2)	<0.11	–	–	−
	(3,3)	<0.11	–	–	−
−440, 40	(1,1)	0.61 (0.16)	1.87 (0.21)	−21.70 (0.09)	−
	(2,2)	<0.09	–	–	−
	(3,3)	<0.15	–	–	−
−440, 0	(1,1)	0.54 (0.16)	1.46 (0.20)	−21.00 (0.07)	1.07 (0.69)
	(2,2)	0.09 (0.06)	1.68 (0.80)	−20.70 (0.33)	−
	(3,3)	<0.18	–	–	−
−480, 80	(1,1)	0.98 (0.08)	1.07 (0.06)	−21.84 (0.02)	0.52 (0.17)
	(2,2)	0.50 (0.08)	1.34 (0.12)	−21.68 (0.05)	−
	(3,3)	0.12 (0.10)	2.05 (1.16)	−21.72 (0.42)	−
Primordial gas, ref. pos. |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8, δ2000 = +35°48′27|${^{\prime\prime}_{.}}$|0
−600, −150	(1,1)	0.46 (0.05)	1.10 (0.09)	−17.43 (0.03)	−
	(2,2)	0.11 (0.05)	0.64 (0.16)	−17.29 (0.08)	−
	(3,3)	0.15 (0.03)	0.51 (0.22)	−17.29 (0.08)	−
−600, 0	(1,1)	0.32 (0.09)	1.69 (0.20)	−16.82 (0.09)	−
	(2,2)	0.17 (0.09)	1.36 (0.34)	−17.11 (0.14)	−
	(3,3)	< 0.10	–	–	−
−200, 400	(1,1)	0.21 (0.08)	0.67 (0.27)	−20.72 (0.08)	−
	(2,2)	<0.06	–	–	−
	(3,3)	<0.08	–	–	−
S235 A-B, ref. pos. |$\alpha _{2000} = 05^{\rm h}40^{\rm m}53{^{\mathrm{s}}_{.}}$|4, δ2000 = +35°41′48|${^{\prime\prime}_{.}}$|0
0,40	(1,1)	0.47 (0.22)	2.25 (0.81)	−16.42 (0.28)	–
	(2,2)	<0.18	–	–	–
	(3,3)	<0.18	–	–	–
0,20	(1,1)	1.12 (0.22)	2.12 (0.21)	−16.61 (0.10)	0.67 (0.54)
	(2,2)	0.56 (0.17)	1.95 (0.37)	−16.29 (0.15)	−
	(3,3)	0.79 (0.40)	1.84 (0.74)	−16.60 (0.27)	−
−40,0	(1,1)	0.05 (0.03)	2.36 (0.81)	−23.38 (0.35)	−
		0.08 (0.03)	2.42 (0.55)	−16.58 (0.23)	−
	(2,2)	<0.13	–	–	–
	(3,3)	<0.30	–	–	–
0,0	(1,1)	2.07 (0.13)	2.17 (0.07)	−16.83 (0.03)	0.33 (0.17)
	(2,2)	1.50 (0.21)	2.35 (0.15)	−16.90 (0.07)	−
	(3,3)	0.73 (0.09)	2.70 (0.27)	−16.83 (0.11)	−
40,0	(1,1)	0.70 (0.21)	1.96 (0.38)	−15.94 (0.15)	−
	(2,2)	0.23 (0.15)	2.85 (0.80)	−16.28 (0.38)	−
	(3,3)	<0.30	–	–	–
20,−20	(1,1)	2.71 (0.24)	2.14 (0.10)	−16.70 (0.05)	0.276(0.223)
	(2,2)	0.84 (0.20)	2.01 (0.26)	−16.72 (0.12)	−
	(3,3)	0.72 (0.37)	3.06 (0.70)	−16.29 (0.31)	−
0,−20	(1,1)	3.25 (0.20)	1.98 (0.08)	−16.84 (0.03)	−
	(2,2)	0.90 (0.13)	2.26 (0.29)	−16.87 (0.10)	−
	(3,3)	0.78 (0.25)	2.35 (0.51)	−17.16 (0.24)	−
−20,−20	(1,1)	0.55 (0.21)	2.29 (0.60)	−16.05 (0.31)	−
	(2,2)	0.40 (0.20)	4.29 (1.04)	−15.99 (0.40)	−
	(3,3)	0.39 (0.38)	2.10 (1.21)	−17.17 (0.62)	−
0,−40	(1,1)	3.35 (0.24)	1.83 (0.08)	−16.56 (0.03)	−
	(2,2)	0.66 (0.17)	2.96 (0.43)	−16.03 (0.17)	−
	(3,3)	0.44 (0.32)	5.88 (1.25)	−15.15 (0.63)	−
0,−60	(1,1)	2.84 (0.23)	2.29 (0.10)	−16.07 (0.04)	−
	(2,2)	0.61 (0.18)	1.68 (0.32)	−16.10 (0.14)	−
	(3,3)	0.38 (0.26)	2.02 (0.79)	−16.04 (0.47)	−
−20,−60	(1,1)	1.70 (0.20)	2.20 (0.17)	−15.91 (0.08)	−
	(2,2)	0.77 (0.30)	1.84 (0.38)	−16.36 (0.18)	−
	(3,3)	0.50 (0.27)	2.62 (0.99)	−15.18 (0.45)	−
20,−60	(1,1)	2.40 (0.25)	1.96 (0.13)	−16.24 (0.05)	−
	(2,2)	0.72 (0.33)	2.64 (0.89)	−16.33 (0.28)	−
	(3,3)	0.73 (0.33)	2.28 (0.69)	−16.44 (0.24)	−
0,−80	(1,1)	2.61 (0.19)	2.16 (0.08)	−16.01 (0.03)	0.48 (0.18)
	(2,2)	0.48 (0.13)	3.03 (0.34)	−15.63 (0.16)	−
	(3,3)	0.45 (0.24)	2.48 (0.86)	−15.94 (0.30)	−
0,−100	(1,1)	1.74 (0.21)	2.49 (0.17)	−15.95 (0.07)	−
	(2,2)	0.65 (0.27)	3.80 (0.68)	−15.58 (0.28)	−
	(3,3)	<0.20	–	–	–
0,−120	(1,1)	0.47 (0.13)	3.74 (0.58)	−16.01 (0.27)	−
	(2,2)	0.20 (0.17)	2.00 (0.80)	−17.24 (0.39)	−
	(3,3)	<0.18	–	–	–

Maps of NH₃(1,1) and NH₃(2,2) emission from S235 East 1, S235 East 2 and S235 Central are shown in Fig. 4. The overall structure of the star-forming regions is similar to that outlined by the CS (2–1) map in Fig. 1. Values of physical parameters in these star-forming regions are shown in Fig. 5. Estimates of T_kin derived from RT-22 and Effelsberg telescope data are in agreement within the limits of uncertainties. A spectral map of NH₃(1,1) towards the S235 A-B star-forming region is shown in Fig. 6. Values of T_kin and |$n_{\rm H_2}$| are also indicated. A complete list of the physical parameters in all these regions is presented in Table 3.

Figure 4.

Maps of NH₃(1,1) (grey) and NH₃(2,2) (contours) emission towards the embedded clusters in the S235 complex. Levels of grey-scale correspond to 1.0, 3.0, 5.0, 7.0 and 9.0 K km s⁻¹, contours correspond to 0.25, 0.50, 0.75, 1.00, 1.25 and 1.50 K km s⁻¹. For the reference position, see Fig 1.

Figure 5.

Physical conditions in the embedded clusters around S235. Left panel: T_kin (grey) and NH₃(1,1) emission map (contours). The contours correspond to 1.0, 3.0, 5.0, 7.0 and 9.0 K km s⁻¹. Levels of grey-scale are given from 10 K to 25 K with a step of 2.5 K. Right panel: stellar density map taken from Paper I (contours) and |$N_{\rm NH_3}$| distribution (grey). Contours are the same as in Fig. 1 (right panel). Levels of grey-scale are given for 10¹³ to 2 × 10¹⁴ cm⁻² with a step of 5 × 10¹³ cm⁻². For the reference position, see Fig. 1.

Figure 6.

Spectra of NH₃(1,1) emission towards the S235 A-B region. Offsets are given with respect to |$\alpha _{2000} = 05^{\rm h}40^{\rm m}53{^{\mathrm{s}}_{.}}$|4 and δ₂₀₀₀ = +35°41^′48|${^{\prime\prime}_{.}}$|0. Values of T_kin (in K) and gas number density (in units of 10³ cm⁻3) are given in the upper right and upper left hand sides of the NH₃(1,1) main component, respectively. Asterisks denote values with uncertainties exceeding 40 per cent.

Table 3.

Physical parameters of molecular clumps related to star formation.

Offset	N(1,1)	N(2,2)	N(3,3)	Tex	Trot	Tkin	n(H2)
(arcsec, arcsec)	(1013 cm−2)	(1013 cm−2)	(1013 cm−2)	(K)	(K)	(K)	(103 cm−3)
S235 East 1, ref. pos. |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8, δ2000 = +35°48′27|${^{\prime\prime}_{.}}$|0
0,160	2.49 (0.52)	–	–	–	–	–	–
0,120	4.94 (0.63)	–	–	–	–	–	–
0,80	4.17 (0.58)	0.76 (0.23)	–	–	18.69 (3.90)	21.20 (4.42)	–
0,40	4.17 (0.43)	–	–	–	–	–	–
−40,160	1.12 (0.41)	–	–	–	–	–	–
−40,120	8.05 (0.51)	1.40 (0.24)	0.80 (0.63)	10.54 (4.23)	18.37 (1.93)	20.73 (2.18)	18
−40, 80	19.75 (10.76)	1.39 (0.93)	–	8.36 (2.21)	16.82 (3.25)	18.57 (3.58)	19
−40, 40	15.40 (5.30)	2.43 (0.81)	1.05 (0.38)	9.24 (1.63)	24.96 (4.34)	31.53 (5.48)	19
−80,120	1.42 (0.86)	–	–	–	–	–	–
−80,80	1.25 (0.45)	–	–	–	–	–	–
−80,40	5.85 (0.27)	0.89 (0.17)	–	7.02 (2.74)	17.36 (1.76)	19.31 (1.95)	13
−80, 0	4.92 (0.35)	0.72 (0.22)	–	–	17.13 (2.72)	18.99 (3.02)	−
S235 East 2, ref. pos. |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8, δ2000 = +35°48′27|${^{\prime\prime}_{.}}$|0
−40, 240	2.79 (0.47)	–	–	–	–	–	–
−40, 200	3.42 (0.17)	0.43 (0.11)	–	–	16.06 (1.91)	17.56 (2.09)	−
−80, 280	4.36 (0.22)	0.48 (0.09)	–	–	15.39 (1.34)	16.69 (1.45)	−
−80, 240	5.17 (0.65)	0.66 (0.20)	–	–	16.17 (2.83)	17.69 (3.10)	−
−80, 200	6.46 (4.92)	1.81 (1.45)	–	3.91 (0.69)	23.31 (9.64)	28.56 (11.81)	3
−120, 200	3.45 (0.34)	0.44 (0.15)	–	–	16.15 (2.67)	17.68 (2.92)	−
−120, 240	5.64 (0.37)	0.97 (0.15)	0.23 (0.20)	–	18.42 (1.30)	20.81 (1.47)	−
−120, 280	5.08 (0.40)	–	–	–	–	–	–
−160, 200	0.94 (0.12)	0.35 (0.10)	0.11 (0.04)	–	27.44 (2.65)	36.45 (3.52)	−
S235 Central, ref. pos. |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8, δ2000 = +35°48′27|${^{\prime\prime}_{.}}$|0
−280, 80	3.48 (0.28)	0.54 (0.19)	0.17 (0.12)	–	19.56 (2.40)	22.48 (2.75)	−
−280, 40	12.37 (9.29)	2.43 (2.37)	−	3.30 (0.35)	19.40 (6.80)	22.25 (7.79)	1
−320, 80	20.80 (7.20)	2.85 (1.43)	−	4.35 (0.40)	22.94 (4.26)	27.92 (5.19)	4
−320, 40	16.10 (2.30)	1.50 (0.31)	0.38 (0.23)	5.14 (0.24)	18.97 (1.19)	21.61 (1.34)	6
−320, 0	2.56 (0.93)	0.13 (0.20)	–	–	13.49 (10.67)	14.32 (11.34)	–
−360, 80	2.44 (0.60)	–	–	–	–	–	−
−360, 40	3.59 (0.36)	0.62 (0.22)	–	–	18.33 (3.79)	23.65 (4.35)	−
−360, 0	3.55 (0.89)	–	–	–	–	–	−
−400, 80	2.56 (0.75)	–	–	–	–	–	−
−400, 40	4.12 (0.37)	–	–	–	–	–	−
−400, 0	11.96 (9.38)	0.76 (0.95)	–	4.42 (1.01)	12.72 (3.16)	13.41 (3.33)	5
−440, 80	1.89 (0.41)	–	–	–	–	–	−
−440, 40	2.95 (0.45)	–	–	–	–	–	−
−440, 0	8.87 (5.84)	8.29 (7.23)	–	3.52 (0.38)	14.40 (3.37)	15.44 (3.62)	2
−480, 80	8.71 (3.20)	1.84 (0.70)	0.52 (0.49)	5.14 (0.66)	30.27 (7.46)	43.04 (10.62)	5
Primordial gas, ref. pos. |$\alpha _{2000} = 05^{\rm h}41^{\rm m}33{^{\mathrm{s}}_{.}}$|8, δ2000 = +35°48′27|${^{\prime\prime}_{.}}$|0
−600, −150	1.52 (0.09)	0.45 (0.15)	0.07 (0.03)	–	22.66 (2.18)	27.44 (2.65)	–
−600, 0	1.66 (0.23)	0.38 (0.14)	–	–	20.75 (5.48)	24.33 (6.42)	–
−200, 400	0.49 (0.11)	–	–	–	–	–	–
S235 A-B, ref. pos. |$\alpha _{2000} = 05^{\rm h}40^{\rm m}53{^{\mathrm{s}}_{.}}$|4, δ2000 = +35°41′48|${^{\prime\prime}_{.}}$|0
0,40	1.55 (0.84)	–	–	–	–	–	−
0,20	19.26 (16.64)	3.26 (2.58)	4.06 (4.28)	4.92 (1.31)	26.06 (13.28)	33.63(17.14)	5
−40,0	0.51 (0.11)	–	–	–	–	–	−
0,0	21.07 (12.76)	5.83 (3.53)	2.71 (1.62)	10.18 (3.22)	37.68 (17.27)	66.09 (30.40)	16
40,0	3.50 (0.49)	0.72 (0.42)	–	–	19.80 (7.80)	22.86 (9.00)	−
20,−20	22.31 (23.34)	2.35 (2.24)	2.70 (2.64)	13.72 (7.80)	19.66 (7.56)	22.64 (8.71)	51
0,−20	18.60 (0.70)	2.22 (0.41)	1.79 (0.84)	–	21.86 (1.49)	26.12 (1.78)	−
−20,−20	3.87 (0.80)	1.91 (1.07)	0.80 (0.74)	–	33.26 (12.70)	51.41 (19.63)	−
0,−40	13.78 (0.67)	2.17 (0.60)	2.50 (2.05)	–	18.01 (1.92)	20.22 (2.15)	−
0,−60	19.55 (0.83)	1.12 (0.34)	0.75 (0.55)	–	14.73 (1.30)	15.85 (1.40)	−
−20,−60	10.59 (0.83)	1.56 (0.56)	1.28 (0.79)	–	22.50 (2.90)	27.17 (3.51)	−
20,−60	14.05 (0.81)	2.09 (0.95)	1.63 (0.70)	–	25.45 (2.78)	32.36 (3.54)	−
0,−80	26.43 (11.61)	2.30 (1.11)	1.58 (1.01)	9.49 (2.05)	18.38(3.08)	20.76(3.48)	23
0,−100	11.41 (0.83)	2.73 (1.23)	–	–	21.38 (6.14)	25.35 (7.28)	−
0,−120	1.39 (0.77)	0.45 (0.25)	–	–	25.27 (7.10)	32.10 (8.40)	−

3.1 S235 East 1

S235 East 1 shows the brightest ammonia emission among S235 East 1, S235 East 2 and S235 Central. The peak of the NH₃(1,1) emission is located close to the position with offset (Δα, Δδ) = (−40 arcsec,80 arcsec), while the peak of the NH₃(2,2) emission is shifted to the south (Fig. 4) by 40 arcsec. Ammonia lines in this clump are optically thin. The observed linewidth is ∼2 km s⁻¹, which is larger than the thermal linewidth for ammonia emission (0.2–0.7 km s⁻¹ for T_kin from 20 to 200 K). So we can infer the presence of non-thermal motions in this clump. Such ammonia linewidths are typical for massive star-forming regions (Jijina, Myers & Adams 1999; Wienen et al. 2012). Radial velocities of the NH₃(1,1) line centres are in the range from −19 to −21 km s⁻¹, being redder than V_lsr in the S235 East 2 and S235 Central clumps.

3.2 S235 East 2

Unlike the other two clumps in the dense shell around S235, this region represents only a modest enhancement of ammonia emission. Spectra of ammonia towards the emission peak at (Δα, Δδ) = (−120 arcsec, 240 arcsec) are shown in Fig. 3. The spatial distribution of the stellar density is shifted relative to the ammonia NH₃(1,1) emission and partially coincides with the patch of warm gas at (−160 arcsec, 200 arcsec; see Fig. 5). The ammonia lines are optically thin, so lower values of |$N_{\rm NH_3}$| and |$n_{\rm H_2}$| in this clump in comparison with the other clumps are expected because of lower brightness of ammonia emission. In contrast to S235 East 1, there are no ammonia lines centred at −19 km s⁻¹. The emission is shifted bluewards here. Linewidths exceed thermal values, as in the S235 East 1 clump.

There are regions in this clump, where the NH₃(1,1) line shows signatures of a non-LTE excitation. One of such regions is found towards the (Δα, Δδ) = (−80 arcsec, 280 arcsec) position. To describe the anomaly, we follow Longmore et al. (2007) and define α_hyp as a ratio of intensities of an outer left (F₁ = 0 → 1) to an outer right (F₁ = 1 → 0) satellite

\begin{equation} \alpha _{\rm hyp} = \frac{T_{\rm mb}(F_1 = 0\rightarrow 1)}{T_{\rm mb}(F_1 = 1\rightarrow 0)}. \end{equation}

(1)

Similarly, for the inner satellites

\begin{equation} \beta _{\rm hyp} = \frac{T_{\rm mb}(F_1 = 2\rightarrow 1)}{T_{\rm mb}(F_1 = 1\rightarrow 2)} \end{equation}

(2)

and for the sum of right and left satellites

\begin{equation} \gamma _{\rm hyp} = \frac{T_{\rm mb}(F_1 = 0\rightarrow 1)+T_{\rm mb}(F_1 = 2\rightarrow 1)}{T_{\rm mb}(F_1 = 1\rightarrow 0)+T_{\rm mb}(F_1 = 1\rightarrow 2)}. \end{equation}

(3)

In our case, β_hyp < 1 and γ_hyp < 1, namely, β_hyp = 0.52 ± 0.22 and γ_hyp = 0.66 ± 0.20. Radiative transfer modelling of ammonia line formation in a medium with systematic velocity gradients was performed by Park (2001). He showed that the case of γ_hyp < 1 is realized in a cloud with inward motion. Therefore, it is possible that the NH₃(1,1) hyperfine intensity anomaly at the offset position (Δα, Δδ) = (−80 arcsec, 280 arcsec) is created by the line-of-sight velocity gradient in the region emitting in NH₃ lines. This case can correspond either to remaining inward motion or deceleration of outflowing gas with distance from the cluster which is typical for expanding flows.

3.3 S235 Central

Ammonia emission in this clump has several peaks. The spatial distribution of NH₃(1,1) is different from that of NH₃(2,2) as shown in Fig. 4. In some positions the optical depth of ammonia is close to unity in contrast to lower values encountered in S235 East 1 and S235 East 2 (see Table 2). Spectra towards the strongest peak of NH₃(1,1) emission at (Δα, Δδ) = (−320 arcsec, 40 arcsec) are shown in Fig. 3. Radial velocities obtained from Gaussian fits range from −19 km s⁻¹ to −22 km s⁻¹. The scatter in radial velocities between different parts of S235 Central is higher than that in S235 East 1 and S235 East 2. Widths of ammonia lines are also super-thermal here.

A hyperfine intensity anomaly is possible around the position with the offset (Δα, Δδ)= (−320 arcsec, 40 arcsec) as can be seen in Fig. 3. The significance of the deviation is about 1.5 sigma. Here we have α_hyp = 1.33 ± 0.21 and β_hyp = 0.84 ± 0.11, that is, the case of α_hyp > 1 and β_hyp < 1, which was explained by Matsakis et al. (1977) and Stutzki & Winnewisser (1985a). According to these studies, several small (in comparison to the beam size of the Effelsberg telescope at 23 GHz, i.e. 40 arcsec) dense clumps contribute to the observed NH₃(1,1) line. Individual clumps should have linewidths of about 0.3–0.6 km s⁻¹ which leads to selective trapping in the hyperfine components of the far-infrared (J, K) = (2, 1) → (1,1) transition via Doppler shift.

3.4 Quiescent primordial gas

We have obtained ammonia spectra towards three positions where primordial gas is supposed to be located. It can be seen from Fig. 4 and Table 2 that the brightness of NH₃(1,1) and NH₃(2,2) lines is significantly lower there in comparison with ammonia emission from the dense clumps. The spectra towards the (−600 arcsec, −150 arcsec) position are shown in Fig. 3. Linewidths exceed thermal values, so non-thermal motions exist in the primordial gas, too. Radial velocities of line centres exceed −18 km s⁻¹, that is, lines from the dense clumps are shifted blueward relative to those from primordial gas. The primordial gas is quite warm, with T_kin of about 27 K. |$N_{\rm NH_3}$| is about 2× 10¹³ cm⁻² there.

3.5 S235 A-B

The brightest peak of ammonia emission in the entire S235 area is found towards the star-forming region S235 A-B. Spectra of NH₃(1,1) emission are shown in Fig. 6. The H ii region S235 A itself is located to the north from (0 arcsec, 0 arcsec) position at Fig. 6 (reference position is |$\alpha _{2000} = 05^{\rm h}40^{\rm m}53{^{\mathrm{s}}_{.}}$|4 and δ₂₀₀₀ = +35°41^′48|${^{\prime\prime}_{.}}$|0), with offset (Δα, Δδ) = (0 arcsec, 34 arcsec) (see also Fig. 1). The vicinity of the S235 A and S235 B nebulae is only partially covered by our data, mainly consisting of spectra along the north–south direction. It allows investigating the gradient of physical parameters along the major axis of the region. Ammonia emission in S235 A-B is optically thin (Fig. 2). Linewidths exceed thermal values. Radial velocities are redshifted relative to those in S235 East 1, S235 East 2, S235 Central and are similar to velocities of the primordial gas.

4 PHYSICAL STRUCTURE OF THE STAR-FORMING REGIONS

4.1 Summary of densities and temperatures

The ammonia emission in the entire S235 area is very non-uniform. It consists of several bright clumps that generally coincide with peaks in stellar density and emission of other molecular tracers.

In those locations within the northern clumps (S235 East 1, S235 East 2 and S235 Central), where physical parameters can be determined, ammonia column density varies from ∼10¹³ cm⁻² to 2 × 10¹⁴ cm⁻². The value of |$N_{{\rm NH}_3}$| towards the primordial gas locations is about 10¹³ cm⁻² (solid circles in Fig. 1). Temperatures over the same area vary from 13 to 43 K. Temperature distribution apparently does not follow any simple pattern, with ‘warm’ and ‘cold’ spots being generally intermixed with each other. In S235 Central and S235 East 2 clumps high temperature spots are located at their respective inner boundaries, in relative isolation. While this arrangement may be indicative of gas heating by UV emission from S235, heating by internal sources is also possible. For example, the warm spot in S235 East 2 at (−160 arcsec, 200 arcsec) is situated only ∼1 arcmin off a bright millimetre continuum source (Klein et al. 2005), while the warm spot in S235 East 1 at (−40 arcsec, 40 arcsec) may have appeared due to illumination from IRAS 05382+3547. Also, abrupt T_kin changes observed in the studied clumps may be an indication of unresolved embedded heating sources. Were the H ii region entirely responsible for gas heating, we would expect warm regions to be smoother and more extended.

The brightest NH₃(1,1) emission in the three clumps, embedded in the molecular shell around S235, comes from gas in S235 East 1. This gas is in average denser and colder than gas in the other clumps of the shell. An ammonia column density |$N_{\rm NH_3}$|∼2 × 10¹⁴ cm⁻² is found towards the NH₃(1,1) peak in S235 East 1. This high column density corresponds to low T_kin (about 19 K) and high gas volume density (about 2 × 10⁴ cm⁻³ towards the centre of the clump).

There is no notable emission and column density peak in S235 East 2. The |$N_{\rm NH_3}$| distribution is smooth with values of about a few times 10¹³ cm⁻² at nearly all the observed locations. Unlike S235 East 1, the highest column density in S235 East 2 corresponds to relatively high temperature, but in other locations within the clump T_kin values are low. The gas volume density is only determined at the location of the highest column density and is nearly an order of magnitude lower than in S235 East 1.

S235 Central contains at least two dense subclumps (Fig. 5) with column densities of about 10¹⁴ cm⁻² (Central W) and 2 × 10¹⁴ cm⁻² (Central E). The greatest column density in S235 Central exceeds that in S235 East 1 despite emission here being fainter as effects of the optical depth could be important in S235 Central [τ(1,1) ∼ 1 at the emission peaks]. Even higher values of |$N_{\rm NH_3}$| can be expected in the densest part of S235 Central because of effects of self-absorption.

The volume density in the dense subclumps is 4–6 × 10³ cm⁻³ and an order of magnitude lower (∼5 × 10² cm⁻³) in the region between them (see Table 1). We note again that the actual density can be higher because of the optical depth effects. As in S235 East 2, the location with the highest |$N_{\rm NH_3}$| in S235 Central is quite warm, with T_kin being above 25 K. The lowest temperatures in the shell are found in the southern part of S235 Central, where the peak of CO emission is observed.

The highest values of T_kin and |$n_{\rm H_2}$| among all the star-forming regions around S235 are found in S235 A-B (see Table 3 and Fig. 6). The value of T_kin decreases from more than 50 K at the water maser position (0, 0) to 16 K at the position 1 arcmin to the south, and then rises again to 30 K at the position 2 arcmin to the south. The value of T_rot = 37 K towards the water maser is consistent with that of the CH₃CN(5–4) transition (Felli et al. 2004). The maximum gas density is found in about 20 arcsec to the east from the centre of the S235 B nebula (see Fig. 1 for location of the S235 B in S235 A-B star-forming region, and Fig. 6 for the spectra). Here |$n_{\rm H_2}$| is larger by a factor of 2 than the density in the densest part of S235 East 1 (5.1 × 10⁴ cm⁻³ and 1.9 × 10⁴ cm⁻³, respectively). Gradients of T_kin, |$n_{\rm H_2}$| and |$N_{\rm NH_3}$| in the north–south direction, that we found, reflect the large-scale structure of the star-forming region to the south of S235 B. Gas temperature rises as we move southward from S235 B to the region with bright HCO⁺ and ¹³CO emission (Boley et al., in preparation). Young star cluster and H ii region S235 C are located there (beyond our maps).

4.2 Radial velocities of the molecular gas

In Paper I, discrimination between the quiescent undisturbed gas and gas which has been compressed by the shock from the expanding H ii region was made on the basis of gas kinematics reflected in CS (2–1) and ¹³CO (1–0) emission. Specifically, the ‘red’ V_lsr range (from −15 km s⁻¹ to −18 km s⁻¹) represents quiescent undisturbed primordial gas. The range from −18 km s⁻¹ to −21 km s⁻¹ (hereafter the ‘central’ range) corresponds to gas compressed by the shock from expanding S235 H ii region. The ‘blue’ range from −21 km s⁻¹ to −25 km s⁻¹ presumably corresponds to gas expulsion from the embedded young star clusters driven by combined effect of the cluster stars.

Ammonia emission supports this picture. Radial velocities of NH₃ lines in the dense molecular clumps S235 East 1, S235 East 2 and S235 Central mostly belong to the ‘central’ range, while V_lsr values from the primordial gas are about −17 km s⁻¹, that is, belong to the ‘red’ velocity range. These velocities are consistent with V_lsr of ¹³CO (1–0) and CS (2–1) emission lines which have higher optical depths. Radial velocities of ammonia lines towards the star-forming region S235 A-B are similar to velocities of the primordial gas and belong to the ‘red’ range. This is consistent with our suggestion from Paper I that star formation in S235 A- B is independent of the expansion of S235. In summary, the overall kinematic structure of the S235 environment, suggested in Paper I, is confirmed by optically thin ammonia emission. This is illustrated in Fig. 7 where we show ¹³CO and ammonia velocities determined from Gaussian fits as a function of offset in RA (note that in both panels we only show data for positions where ammonia emission is present). The two subclumps in S235 Central apparently have distinct average velocities, namely, about −20 km s⁻¹ for S235 Central E and about −21 km s⁻¹ for S235 Central W.

Figure 7.

Line centre velocities for CO (left panel) and NH₃ (right panel) emission as a function of offset in right ascension (at all declinations). Different symbols correspond to various gas clumps as indicated in the legend. A symbol size corresponds to the column density. The colour version of this figure is available in the electronic edition of the journal.

Apart from the three velocity components, yet another kinematic feature found in Paper I is a velocity shift between ¹³CO (1–0) and CS (2–1) lines in S235 East 1 (Fig. 8a), reaching about 2 km s⁻¹ in line wings. As both CS and NH₃ trace dense gas, we may expect a similar velocity difference between ¹³CO (1–0) and ammonia lines, and it indeed exists. While both these molecules trace dense gas, the spatial distributions of carbon-bearing molecules and nitrogen-bearing molecules differ (Pirogov et al. 2007) due to the difference in the chemical evolution of these molecules. An example of the shift of NH₃(1,1) and ¹³CO (1–0) lines towards the (−40 arcsec, 40 arcsec) offset position is presented in Figs 8(a) and (b) along with the ¹³CO (1–0) versus CS (2–1) shift. Both ¹³CO (1–0) –CS (2–1) velocity shift and ¹³CO (1–0) –NH₃(1,1) velocity shift presumably indicate the relative motion of a dense clump in S235 East 1 and its outer envelope. NH₃(1,1) and CS (2–1) line profiles are also slightly shifted relative to each other, but in this case the shift is seen mostly in the red wings of the lines (Fig. 8c). To show the shifts more clearly, in panels of Figs 8(d)–(f) we display a channel-by-channel subtraction of ¹³CO (1–0) –CS (2–1) and ammonia spectra, scaled as indicated in the caption. The shifts ¹³CO (1–0) versus CS (2–1) and ¹³CO (1–0) versus NH₃ are smaller than the widths of the ¹³CO (1–0), NH₃and CS (2–1) lines themselves and are comparable to the thermal width for molecular gas at the temperature of 20–30 K (0.3–0.4 km s⁻¹).

Figure 8.

Scaled CS (2–1), ¹³CO (1–0)and NH₃(1,1) spectra towards the (−40 arcsec, 40 arcsec) position (top row) and the difference between the corresponding spectra (bottom row). (a) ¹³CO (1–0)· 0.09 and CS (2–1), (b) ¹³CO (1–0)· 0.06 and NH₃(1,1), (c) CS (2–1)· 0.7 and NH₃(1,1), (d) ¹³CO (1–0)· 0.09 −CS (2–1), (e) ¹³CO (1–0)· 0.06 −NH₃(1,1), (f) CS (2–1)· 0.7 −NH₃(1,1). For the reference position, see Fig. 1.

While the ¹³CO (1–0)–NH₃ line shift is most clearly seen towards the (−40 arcsec, 40 arcsec) position, it is not a unique feature of this location. ¹³CO (1–0) and ammonia lines are somewhat displaced relative to each other at other locations as well. To quantify this shift, we subtracted ammonia line centre velocities (main component) from ¹³CO (1–0) line centre velocities as determined from Gaussian fits. A shift like this has been analysed previously in several star-forming regions. In particular, Walsh, Myers & Burton (2004) looked for motions of dense cores [in the N₂H⁺(1–0) line] with respect to their more rarefied envelopes [using ¹³CO (1–0) and C¹⁸O(1–0)] towards a sample of low-mass star-forming cores. Very small velocity differences (less or about 0.1 km s⁻¹) were found, so Walsh et al. (2004) concluded that isolated low-mass cores generally do not move ballistically relative to their surrounding envelopes. Similar analysis for regions of massive star formation was performed by Wienen et al. (2012). They discussed relative motions in several high-mass cold clumps based on observed ¹³CO (1–0) and NH₃ lines and mentioned some trends that can be compared to our results.

First, Wienen et al. (2012) found that ¹³CO linewidths are as a rule broader than widths of NH₃ (1,1) lines. This holds in our study as well, for both CO and CS lines, even though at a smaller scale, as clumps that we have observed are less dense and more compact. Ammonia (1,1) linewidths range from 1 to 3 km s⁻¹, while ¹³CO and CS linewidths are in average 20–30 per cent larger and extend to 3.5 km s⁻¹ (Fig. 9, top row). Among positions with high column density, the only one with a wider ammonia line and narrower CO and CS lines is the position (−280 arcsec, 40 arcsec) in S235 Central E. Apart from this, all the gas clumps seemingly follow the same pattern in ‘width versus width’ plots.

Figure 9.

Top row: widths of the ¹³CO lines and CS lines compared to those of ammonia (1,1) lines. Solid line corresponds to equality of both widths. Bottom row: velocity shift between a low-density tracer (CO) and higher density tracers (CS and NH₃) as functions of the ammonia (1,1) linewidth. Different symbols correspond to various gas clumps as indicated in the legend. A symbol size corresponds to the column density. The colour version of this figure is available in the electronic edition of the journal.

This is not true in the case of line velocity differences. In S235 East 1 lines of a low density tracer [¹³CO (1–0)] are shifted relative to higher density tracers [CS (2–1) and NH₃], as we noted in Paper I. The maximum shift between ¹³CO (1–0) and CS (2–1) line profiles is about 0.5 km s⁻¹ which is a bit higher than the thermal velocity and significantly lower than the linewidths (Fig. 9, bottom left, filled circles). With respect to ¹³CO (1–0)-NH₃, some regularity is seen in S235 East 1: larger ammonia linewidths correspond to higher velocity differences between ¹³CO (1–0) and ammonia (Fig. 9, bottom right, filled circles).

In S235 East 2 and S235 Central E, there is almost no difference between ¹³CO and CS line velocities, and the difference between ¹³CO and NH₃ line velocities is much smaller than in S235 East 1 (but also positive). In S235 Central W the difference between CO and ammonia velocities is close to zero, and CO−CS difference is notably negative.

The analysis performed by Wienen et al. (2012) has shown that supersonic velocity differences are typical of high-mass clumps, unlike low-mass star-forming regions, where velocity differences between low- and high-density tracers are mostly subsonic. Our data show a mixed picture. While the CO−CS difference is at best mildly supersonic (in S235 East 1 and S235 Central W), CO−NH₃ differences exceed the sound speed more significantly, reaching almost 1 km s⁻¹ in S235 East 1. The important difference between our results and Wienen et al. (2012) results is that CO−NH₃ differences that we measure are positive at almost all locations where ammonia emission is observed. A possible explanation is that Wienen et al. (2012) studied large-scale regions. In this case, the velocity differences can be related to chaotic motions of individual dense clumps within a low-density medium. We study individual clumps themselves, so our results are, probably, better explained in the ‘core-to-envelope’ paradigm. More specifically, as we relate gas clumps to star clusters, they may represent gas expelled from the clusters, with different lines tracing denser and less dense parts of an expanding envelope.

In Paper I we argued that S235 East 1 is the youngest cluster in the northern part of the S235 region. The ammonia data are in qualitative agreement with this suggestion. The ammonia lines being blueshifted with respect to CO indicates the presence of a significant gas velocity gradient along the line of sight. An outflow in the region is not well developed and contains considerable amounts of weakly processed gas. This gas is mostly in molecular form and displays considerable variability in the outflow kinematics with the distance from the cluster. Molecular gas is observed in projection on to the cluster, so the difference in velocities of molecular lines which trace regions with different kinematics is well pronounced. It arises due to both chemo-dynamical processing of the gas in a propagating shock and ionization front (see e.g. Kirsanova, Wiebe & Sobolev 2009) and also due to line excitation details. Ammonia emission arises in the dense gas that has been swept up from the cluster by the combined action of young stars and now moves towards the observer. CO emission forms in less dense regions and attains significant optical depth much closer to the observer. The more red-shifted CO velocity component may trace the deceleration of the expanding outflow or leftovers of the parent gas clump which are still moving towards the S235 East 1 cluster.

The absence of a similar velocity shift towards S235 East 2 and S235 Central E is also consistent with our suggestion that clusters embedded in these clumps are actively losing leftover molecular gas that shows up observationally as ¹³CO (1–0) and CS (2–1) emission in the ‘blue’ range. At this later evolutionary stage any initial kinematical structure, including velocity shifts between various molecular tracers, has probably been destroyed. It must be noted that while both these clumps possess CO and CS emission in the blue range and are somewhat displaced relative to corresponding clusters, otherwise they are kinematically different. The S235 East 2 clump is very quiescent, with small linewidths that do not differ much from one direction to another. On the other hand, S235 Central E is more disturbed showing both wide and narrow lines, sometimes in adjacent directions.

Kinematic properties of S235 Central W are harder to interpret as it differs from the other clumps in a couple of aspects. First, this is the coldest clump among those covered with our observations. Also, in the other clumps CS and ammonia line positions are generally consistent with each other, in the sense that CO−CS and CO−NH₃ velocity differences are either both close to zero (in S235 East 2 and S235 Central E) or both positive (in S235 East 1). In S235 Central W velocities of CO and NH₃ are similar, but CS emission is shifted significantly bluewards relative to CO emission. This difference does not fit into our simple picture and craves for more data.

5 STAR FORMATION TRIGGERING AROUND S235

5.1 Sizes and triggering time-scales

In Paper I we proposed that both the C&C process and compression of pre-existing clumps could be responsible for the star formation in this region. Here we re-consider this suggestion using S235 parameters derived from our ammonia observations. Specifically, we compare our results with predictions of a theoretical model by Whitworth et al. (1994) to investigate the history of star formation around S235 and to determine main characteristics of a C&C process: time, when fragmentation of a dense shell starts (t_frag), the radius of the shell (R_frag) at this moment and mass (M_frag) of the fragments. We also calculate the radius of an initial Strömgren sphere, R_s(0), (Dyson & Williams 1997). Then we relate these values to parameters of S235 and S235 A in order to estimate ages of the H ii regions. Given the clumpiness of matter, surrounding both regions, the model of Whitworth et al. (1994) should be applied with care, and our conclusions can only be considered as approximate.

To use the relations from Whitworth et al. (1994), one needs to specify some parameters. To calculate the speed of sound in the undisturbed medium, we adopt 25 K as a typical temperature for S235 and 34 K as a typical temperature for S235 A. The gas number density in the undisturbed medium is supposedly equal to the density of primordial gas for S235 and to the density of surrounding gas for S235 A. Several values of n₀ are considered for S235. The lowest one, 500 cm⁻³, is an average large-scale value measured with the RT-22 (see Table 1). The observationally motivated upper limit is taken to be 3 × 10³ cm⁻³ which is a reasonable value for the region of S235 according to Evans & Blair (1981) (their ‘−17 km s⁻¹’ cloud). We also consider a higher value of 7 × 10³ cm⁻³ to find the density at which the estimated H ii region age becomes comparable to t_frag. The flux of hydrogen-ionizing photons (S_ion) was taken from Sternberg, Hoffmann & Pauldrach (2003) in accordance with the spectral types of ionizing stars which are O9.5 in S235 (Georgelin, Georgelin & Roux 1975) and B0.5 in S235 A (Thompson, Thronson & Campbell 1983).

Distance estimates D to S235 vary from 1.6 to 2.0 kpc. We adopted an intermediate value 1.8 kpc (Evans & Blair 1981) for our estimates. The angular size φ of S235 is 380 × 450 arcsec² in accordance with radio continuum measurements (Israel & Felli 1978). Thus, the linear radius of S235 is constrained by values from 1.5 pc (D = 1.6 kpc, φ = 380 arcsec) to 2.2 pc (D = 2.0 kpc, φ = 450 arcsec). We adopt 2 pc as a representative value for S235. The angular size of S235 A is 20 arcsec (Israel & Felli 1978), so its linear radius is between 0.08 and 0.10 pc. All the adopted parameters for S235 and S235 A are summarized in Table 4. Results of a comparison with calculations based on Whitworth et al. (1994) are given in Table 5 . In the last row we present an H ii region age, estimated using equation (4) from Whitworth et al. (1994). Adopted sizes are listed in Table 6.

Table 4.

Parameters of S235 and S235 A used for calculation of characteristics of the triggering process.

Parameter	S235	S235 A
Sion (s−1)	1048.26	1048.02
Tkin(K)	25	34
n0 (cm−3)	5 × 102–3 × 103	5 × 103
Radius (pc)	2.0|$^{2.2}_{1.5}$|	0.09|$^{0.10}_{0.08}$|

Table 5.

Estimated characteristics related to a C&C process for the parameters derived for S235 and S235 A.

Value	S235				S235 A
n0	5 × 102	103	3 × 103	7 × 103	5 × 103
Rs(0)(pc)	0.7	0.4	0.2	0.1	0.1
Rfrag(pc)	9	6	3	2	3
tfrag(Myr)	3	2	2	1	1
Mfrag|$\text{(M}_{\odot }$|⁠)	200	150	90	60	140
Age (Myr)	0.3	0.4	0.6	1	0.004

Table 6.

Sizes and masses of dense gas in the young clumps.

Cluster	Angular	Physical	Mgas	|$\frac{M_{\rm star}}{M_{\rm gas}}$|
	size (arcsec)	size (pc)	(⁠|$\text{M}_{\odot }$|⁠)
S235 East 1	50×70	0.4×0.6	250	–
S235 East 2	60×110	0.5×1	100	0.4–0.6
S235 Central E	50×60	0.4×0.5	40	0.9–1.5
S235 Central W	50×20	0.4×0.2	12	0.9–1.5

The fragmentation radius R_frag coincides with the actual linear radius of the S235 H ii region within a factor of 2 only for n₀ of 3 × 10³ cm⁻³ or greater. So, it is possible that triggering of star formation in S235 East 1, S235 East 2 and S235 Central by a C&C process took place in the recent past if the density in the surrounding medium is close to the upper limit of the range that is indicated by our observations (typically about 3–5 × 10³ cm⁻³ towards non-peak positions). In this case, star-forming clumps should have masses of about 50–100 |$\text{M}_{\odot }$|⁠.

It is interesting that the projected distance from the exciting star of S235, BD +35°1201 (Georgelin et al. 1975), to S235 A and S235 B is close to R_frag for n₀ ∼ 3 × 10³ cm⁻³. However, as already mentioned, the star-forming regions S235 A-B are located outside the common dense envelope that contains S235 East 1, S235 East 2 and S235 Central. The outer boundary of the S235 H ii region is about two times closer to S235 East 1 and S235 East 2 than to S235 A (see Fig. 1). A bright bubble of the S235 photon dominated region on Spitzer-IRAC images does not have a common border with S235 A. Therefore, the formation of S235 A and S235 B exciting stars could not have been triggered by the expansion of S235. This conclusion is confirmed by Dewangan & Anandarao (2011). They found that young stellar objects in the star-forming region S235 A-B are about 10–20 times older than YSOs in S235 East 1 and S235 East 2. Ages of embedded clusters around S235 and in S235 A-B were also estimated by Camargo et al. (2011) to be 3 ± 2 Myr. However, the low accuracy of these results precludes them from being used to elucidate the star formation history in the area.

A central star in S235 A has a spectral type close to that of the star in S235. Its ionizing power is only about two times smaller. Therefore, the much smaller size of S235 A suggests that it is much younger. Note, however, that S235 A expands in a medium with increasing density (see Table 3 and Fig. 1). Specifically, in S235 A |$n_{\rm H_2}$| is about 5 × 10³ cm⁻³ towards the position with offset (Δα, Δδ)= (0 arcsec, 20 arcsec) (see Fig. 6) and ∼16 × 10³ cm⁻³ towards the position with offset (Δα, Δδ)= (0 arcsec,0 arcsec). Remember that the H ii region S235 A itself is located at (0 arcsec, 34 arcsec). Given this, it is hard to estimate an actual age of S235 A. If we take the lower value of |$n_{\rm H_2}$| ∼5 × 10³ cm⁻³ and assume a uniform density distribution, we can estimate the minimum possible age of S235 A, using formulae from Whitworth et al. (1994). The adopted physical size of the S235 A H ii region is close to R_s(0) and significantly smaller than R_frag (see Table 5). If we again use equation (4) from Whitworth et al. (1994), we obtain an age of only 4000 years for S235 A. The upper limit of |$n_{\rm H_2}$|⁠, quoted above, increases this value by a factor of 2. Anyway, the estimated age of S235 A is significantly smaller than t_frag. So, even though S235 A expands into a highly non-uniform medium with increasing density, and our age estimates are very uncertain, it is definitely too young to trigger star formation in its vicinity by a C&C process. The dense shell around S235 A seems to be just at the beginning of formation. Currently, S235 A can trigger star formation only if it encounters some pre-existing dense clumps.

Triggered star formation may actually arise in three varieties in the studied region. Apart from the C&C mechanism and compression of pre-existing clumps by the shock front around an H ii region, triggering by external shocks is also possible. It is important to note that the star formation in the area is not coeval. Our age estimates for the S235 and S235A H ii regions differ quite significantly. An age spread is seen even within the S235 A-B cluster itself. There are three currently known massive stars in the S235 A-B area: the exciting stars for S235 A, S235 B (Israel & Felli 1978; Boley et al. 2009) and MIR-AB from Dewangan & Anandarao (2011) (‘s5’ source; see their table C8). All of them are considerably younger than lower mass stars in the same area. Ages of the lower mass stars in the cluster are about 10⁵–10⁶ years (Dewangan & Anandarao 2011), while an H₂O maser indicates a very recent high-mass star formation episode (Felli et al. 1997). This is supported by its association with the young strong HCO⁺ outflow (Felli et al. 2004). The age of the S235 A H ii region is ∼4000 years only. Ages of MIR-AB and the exciting star in S235 B are ∼3000 and ∼40000 years, respectively (Dewangan & Anandarao 2011). The age spread indicates that there were multiple star formation episodes in the vicinity of S235 A-B.

The high-velocity gas feature FVW 172.8+1.5 probably representing an old supernova remnant (Kang, Koo & Salter 2012) could be responsible for some of the episodes. It was found by Kang et al. (2012) that there is H i emission with positive (from +25 to +20 km s⁻¹) and negative velocities (from −28 to −25 km s⁻¹). These negative velocities are consistent with the velocities of the molecular gas in the vicinity of S235. It means that the possible supernova is located beyond the giant molecular cloud G174+2.5 and star-forming regions S235 and S235 A-B (S235 C also) belonging to it. We see optical emission from the H ii region S235 slightly obscured by interstellar gas and dust. So, S235 is situated at the near edge of the giant molecular cloud. The formation of S235 cannot be triggered by that supernova. The optical extinction towards S235 A is about 10 mag (Krassner, Pipher & Sharpless 1979). So, this region is more deeply embedded in the molecular gas. Streamlining of the S235 A-B star-forming region by the shock from the supernova as was modelled for example by Kimura & Tosa (1993) could be responsible for formation of those young stars in S235 A-B whose age is less than about 0.3 Myr [age of the proposed SN remnant from Kang et al. (2012)].

5.2 Masses of gas and stars in the young embedded clusters

Our data allow us to estimate gas masses of the dense clumps (see Table 6). Here we assumed that the clump borders correspond to positions where |$N_{\rm NH_3}$| is equal to a half of the maximum value given in Table 3 for a particular clump. The gas mass is derived from n(H₂) values given in Table 3. To calculate M_star/M_gas, stellar mass estimates for main-sequence and pre-main-sequence stars in S235 East 2 and S235 Central from Camargo et al. (2011) are used.

We noted above that S235 East 1 is the densest clump in the vicinity of S235. Apparently, it is also the most massive clump in the S235 area. M_gas in S235 East 1 is higher than M_frag for any considered value of |$n_{\rm H_2}$| (Table 5). No stellar mass estimate is given for S235 East 1 in Camargo et al. (2011) but by analogy with other clusters we may assume that it is comparable to the gas mass. In other words, the total mass of S235 East 1 (gas+stars) significantly exceeds the estimated mass of an average dense shell fragment formed due to the C&C mechanism. Thus, the formation of S235 East 1 looks more like a result of a contraction of a pre-existing molecular clump.

Total masses (gas+stars) of S235 East 2 and S235 Central are lower. We have seen previously that our age estimate for S235 is comparable to t_frag only if the density of the surrounding gas is relatively high. In this case, clumps having appeared due to fragmentation of the shell should have masses of the order of 50–100 |$\text{M}_{\odot }$|⁠. Masses of S235 East 2 and S235 Central E fall into this range, while the mass of S235 Central W is somewhat lower. It must be kept in mind that our mass estimates are quite uncertain, in particular, because the stellar component and the gas component in S235 East 2 and S235 Central are displaced relative to each other. But still, order-of-magnitude mass estimates – tens of solar masses for S235 East 2 and S235 Central E (versus hundreds of solar masses for S235 East 1) – indicate that the C&C mechanism may have been responsible for the formation of S235 East 2 and S235 Central E. As we mentioned above, the status of S235 Central W is uncertain.

Mass is not the only parameter that distinguishes S235 East 1 from the other two clusters. The gas clump associated with this cluster is the only one of the considered clumps where lines of various molecular tracers are noticeably shifted relative to each other. Also, it lacks the blue CO and CS emission component that is present in S235 East 2 and S235 Central clumps. All these differences hint at the different evolutionary track for S235 East 1 in comparison to S235 East 2 and S235 Central (or, at least, to S235 Central E).

6 CONCLUSIONS

We describe ammonia observations towards star-forming regions around S235 and present maps of NH₃(1,1) and NH₃(2,2) emission. The dense gas content in young embedded clusters S235 East 1, S235 East 2, S235 Central, as well as in the star-forming region S235 A-B is investigated. We determine temperature, ammonia column density and gas number density towards the clusters. Gas kinematic structure and physical parameters suggest that at least two (S235 East 2 and S235 Central) of the three clusters nearest to S235 have been formed via triggering by a C&C process if the density of the surrounding gas exceeds 3 × 10³ cm⁻³. The embedded cluster S235 East 1 could have been formed as a result of an interaction of the shock front from S235 with a pre-existing dense clump. The new observations are consistent with our previous conclusion that S235 East 1 is at an earlier evolutionary stage than S235 East 2 and S235 Central. Parameters of the dense gas towards S235 A imply that this H ii region expands into a medium with increasing density. The S235 A H ii region is too young to trigger star formation around itself by a C&C process. However, it can trigger formation of stars in pre-existing clumps. External large-scale shocks could have been responsible for shaping the S235 A-B star-forming region.

CO, CS and NH₃ lines are shifted relative to each other in S235 East 1. A shift between CO and NH₃ lines confirms our conclusion from Paper I that there is a relative motion between dense and diffuse gas in this object. We note also that star formation in S235 A-B is not coeval. Massive stars in the region are considerably younger than the lower mass stars in the region.

This work was supported by the RFBR (grants 13-02-00642, 11-02-01332) and by the President of the RF grant NSh-3602.2012.2. MSK and DSW are grateful to the OFN-17 program of the Russian Academy of Sciences. MSK and AMS thank the staff of the Effelsberg Observatory for their hospitality and assistance during observations. We thank Aurora Sicilia-Aguilar for help in observations of S235 A. The Second Palomar Observatory Sky Survey (POSS-II) was made by the California Institute of Technology with funds from the National Science Foundation, the National Aeronautics and Space Administration, the National Geographic Society, the Sloan Foundation, the Samuel Oschin Foundation and the Eastman Kodak Corporation. The Oschin Schmidt Telescope is operated by the California Institute of Technology and Palomar Observatory. This work is based [in part] on observations made with the Spitzer Space Telescope, which is operated by the Jet Propulsion Laboratory, California Institute of Technology under a contract with NASA. Support for this work was provided by NASA through an award issued by JPL/Caltech.

1

http://www.iram.fr/IRAMFR/GILDAS