SURVEY OBSERVATIONS OF A POSSIBLE GLYCINE PRECURSOR, METHANIMINE (CH₂NH)

We conducted survey observations of CH₂NH with the 45 m radio telescope of Nobeyama Radio Observatory (NRO), the National Astronomical Observatory of Japan, in 2013 April, May, and December, and the Sub-Millimeter Radio Telescope (SMT), Arizona Radio Observatory, in 2013 October. The observed lines and their parameters are summarized in Table 2. Various species are observed simultaneously in our observations. The observation and analysis of other species will be presented in a separate paper, with comparisons to the chemical modeling in order to discuss the chemistry of glycine formation.

At NRO, we used the TZ receiver (Nakajima et al. 2013) in the dual polarization mode. The 4₀₄–3₁₃ transition at 105 GHz was observed. The SAM45 spectrometer was used for the backend with a frequency resolution of 122 kHz, corresponding to a velocity resolution of 0.35 km s⁻¹ at 105 GHz. The system temperature ranged from 150 to 300 K. The main beam efficiency (η_mb) was 0.31, and the beam size (FWHM) was 16''. Observations were performed in the position switching mode. The pointing accuracy was checked by observing nearby SiO masers, and the pointing error was typically 5'', but 10'' under windy conditions.

With the SMT telescope, the 1₁₁–0₀₀, 4₁₄–3₁₃, and 4₂₃–3₂₂ lines at 225, 245, and 255 GHz regions, respectively, were observed. Their parameters are summarized in Table 2. We used the 1 mm ALMA SBS SIS receiver in the dual polarization mode. The system temperature was between 250 and 400 K. The main beam efficiency was 0.74. The beam sizes (FWHM) were 35'' and 40'' depending on the observing frequencies. The position switching mode was employed. The pointing accuracy was checked using Venus, and its accuracy was kept within 10''. We used Forbes filters as the spectrometer with a frequency resolution of 250 kHz, which is equivalent to the velocity resolution of 0.29 km s⁻¹. Since the Forbes filter has 512 channels, the observed frequency range covered 128 MHz at 225 GHz. For the transitions in 245 and 255 GHz, we split the Forbes filter into two 256 channel sets, and two lines were observed simultaneously.

Theule et al. (2011) suggested that CH₂NH is formed via the hydrogenation process of HCN (HCN $\to$ CH₂NH). To select our target sources, we noted that CH₃OH is well known to be formed via hydrogenation process from CO on the grain surface (CO $\to$ H₂CO $\to$ CH₃OH). If the formation process to CH₂NH is similar to that of CH₃OH, CH₂NH would potentially be rich toward CH₃OH-rich sources. Thus, our sources were selected from "The Revised Version of Class I Methanol Maser Catalog"⁷ (Val'tts et al. 2012), and from CH₃OH maser sources (Minier & Booth 2002). We also took into account the results of survey observations of COMs toward hot core sources by Ikeda et al. (2001). Hot cores are high-mass star-forming regions traced by high temperature and dense molecular gas, which are sometimes associated with H ii regions. Further, we also selected two famous low-mass star-forming regions where complex molecules have been observed. These 14 sources for our survey are summarized in Table 3.

In this table, W3(H₂O), Orion KL, NGC 6334F, G10.47+0.03, G19.61–0.23, G31.41+0.3, W51 e1/e2, and DR21(OH) are molecule-rich sources (Ikeda et al. 2001). G75.78+0.34 and Cep A were selected from the CH₃OH maser catalogs. The above sources are high-mass star-forming regions.

In this subsection we will describe the methodologies used to derive the fractional abundances of CH₂NH. In the following discussion, we will show that reliable CH₂NH abundances can be obtained when we employ the same source size as that of the continuum emission. If continuum data are not available, we assume a source size of 10'' in the derivation. The estimated CH₂NH abundances under compact source sizes, which will be described later in detail, and a source sizes of 10'' are summarized in Tables 5 and 6, respectively.

Table 5.  CH₂NH Abundances for Compact Sources

Source	Size (Components)	N[H2] (1024cm−2)	Tex (K)	N[CH2NH] (1016 cm−2)	X[CH2NH] (10−8)
Orion KL	2'' (3)	1.4(a)	88(±23)	4.6(±0.7)	3.3(±0.5)
G10.47+0.03	14	6.7(b)	84(±57)	21(±7.3)	3.1(±1.1)
G31.41+0.3	17	3.5(b)	28(±17)	3.1(±3.0)	0.88(±0.85)
W51 e1/e2	12 (2)	4.0(b)	20(±2)	1.1(±0.63)	0.28(±0.16)
NGC 6334F	37	1.4(b)	[40]	0.3(±0.2)	0.24(±0.14)
G19.61–0.23	25	6.7(c)	[40]	0.94(±0.65)	0.14(±0.10)

Note. CH₂NH column densities "N," or its upper limits and fractional abundances "X" are shown. The distributions of CH₂NH were assumed to be equal to the dust continuum. Since it is known that Orion KL has three CH₂NH peaks and W51 e1/e2 has two continuum components, beam coupling factors were modified for these sources. Excitation temperatures are fixed at 40 K (shown with square brackets if only one transition was available) and the uncertainty associated with this assumption is assumed to be 50% of its central value, considering the wide range of excitation temperatures derived with the rotation-diagram method.

References. (a) Hirota et al. 2015; (b) Hernández-Hernández et al. 2014; (c) Wu et al. 2009.

Download table as:  ASCIITypeset image

For sources where more than two transitions are available, we calculated the column densities of CH₂NH using the rotation-diagram method described in Turner (1991), and the following equation was employed:

$$\begin{eqnarray}&&\mathrm{log}\displaystyle \frac{3{kW}}{8{\pi }^{3}\nu S{\mu }^{2}{g}_{{\rm{I}}}{g}_{{\rm{K}}}}=\mathrm{log}\displaystyle \frac{N}{{Q}_{\mathrm{rot}}}-\displaystyle \frac{{E}_{{\rm{u}}}}{k}\displaystyle \frac{\mathrm{log}e}{{T}_{\mathrm{rot}}},\end{eqnarray} \tag{ 1 }$$

where W is the integrated intensity, S is the intrinsic line strength, μ is the permanent electric dipole moment, g_I and g_K are the nuclear spin degeneracy and the K-level degeneracy, respectively, N is the column density, Q_rot is the rotational partition function, E_u is the upper level energy, and T_rot is the rotation temperature. g_I and g_K are unity for CH₂NH because CH₂NH is an asymmetric top with no identical H atoms. The column density and the excitation temperature can be derived by utilizing least-squares fitting. The column density is derived from the interception of a diagram, and its slope will give us the excitation temperature.

In the excitation analysis above, the beam coupling factors are key parameters. We investigated the distribution of the CH₂NH 1₁₁–0₀₀ transition at 225.55455 GHz toward Orion KL using the calibrated Science Verification data of ALMA cycle 0⁹ to estimate CH₂NH source sizes. The data were imaged with the Common Astronomy Software Applications (CASA) package. These data have the spectral resolution of 488 kHz and the synthesized beam size of ∼2''. We selected line-free channels for continuum subtraction in the (u, v) domain using the CASA task uvcontsub. The channel maps of the CH₂NH 1₁₁–0₀₀ distribution were made using the CASA task CLEAN. We employed the equal weighting of uv, and set the threshold value to stop the CLEAN task at 72 mJy, considering the rms level. After that, we created the integrated intensity map from 225.55348 to 225.55641 GHz, corresponding to the V_LSR range from 6.5 to 10.4 km s⁻¹.

Our integrated intensity map of Orion KL is shown in Figure 4(a). The region corresponding to the beam of the NRO 45 m telescope in our survey is shown by the black circle. We found that there are three strong CH₂NH peaks in this region, which are unresolved with the synthesized beam of these data (∼2''). In Figure 4(b), CH₂NH distributions are compared with the 0.8 mm dust continuum emission in Figure 4 of Hirota et al. (2015). We found that the peak positions of CH₂NH are in both the Orion Hot Core and Compact Ridge. There is also a weak peak of CH₂NH in the northwest clump. Since the distribution of the CH₂NH and dust continuum emission look similar to each other, dust continuum emissions could be good indicators to estimate the source sizes of CH₂NH. Thus, if dust continuum data were available, we employed the source size of the dust continuum emission for CH₂NH to calculate the beam coupling factors. In this case, column densities of molecular hydrogen were estimated from the dust continuum emission to calculate the fractional abundances of CH₂NH (Table 5).

Zoom In Zoom Out Reset image size

For G10.47+0.03 and G31.41+0.3, we created rotation diagrams to derive the column densities using the three transitions that we observed. For Orion KL and W51 e1/e2, we created rotation diagrams by including other CH₂NH transitions reported by Dickens et al. (1997). These results are shown in Figure 5. For these four CH₂NH sources, we derived excitation temperatures ranging from 20 to 90 K. Since only one transition was observed toward the other sources, we assumed that the excitation temperature is similar to that of HCOOH (40 K) reported by Ikeda et al. (2001) because the dipole moment of HCOOH is close to that of CH₂NH. It would be important to know the error associated with the fractional abundances when the excitation temperature is fixed to be 40 K. Considering that the wide range of excitation temperatures was derived from the rotation-diagram method, we assumed that the excitation temperatures can be changed within ±20 K, which can change the column densities by 40%. We note that the optical depths calculated with our method were less than 0.3, even for strong CH₂NH transitions. The CH₂NH column densities and the fractional abundances derived assuming compact sources are summarized in Table 5.

Zoom In Zoom Out Reset image size

For sources for which dust continuum emission data are not available (G34.3+0.2, DR21(OH), W3(H₂O), G75.78+0.34, Cep A, NGC 2264 MMS3, NGC 1333 IRAS4B, and IRAS 16293-2422) we assumed, based on extended CO data, that CH₂NH is spatially extended (a source size of 10'') in calculating the fractional abundances. In this case, both CH₂NH and molecular hydrogen may be underestimated due to beam dilution. It is important to know how fractional abundances tend to be affected if we employ a source size of 10''. For this purpose, we calculated CH₂NH fractional abundances toward Orion KL, G10.47+0.03, G31.41+0.3, NGC 6334F, and W51 e1/e2 assuming source sizes of 10'' (top half of Table 6), and the CH₂NH fractional abundance ratio of the compact source to the 10'' source were investigated. As a result, we found that the differences of the fractional abundances were within a factor of ∼3 (Table 6). Considering that the dependence on the source sizes was small, we employed a compact source size when dust continuum data were available (Table 5), and we used the source size of 10'' to derive the CH₂NH fractional abundances for the other sources (the bottom half of Table 6).

In Tables 5 and 6, we summarize the results of our analysis. Fractional abundances of CH₂NH are compared in Figure 6. Our CH₂NH fractional abundance is highest in Orion KL. The column density of CH₂NH toward Orion KL is 6.0 (±1.8) × 10¹⁴ cm⁻² in Dickens et al. (1997) , and our value, assuming a beam filling factor of unity, is 1.2 (±0.4) × 10¹⁵ cm⁻², which agrees within a factor of 2. In our calculation with a compact source size, the column density and the corresponding fractional abundance are, respectively, 5.5 (±0.8) × 10¹⁵ cm⁻² and 3.3 (±0.5) × 10⁻⁸.

Zoom In Zoom Out Reset image size

G10.47+0.03 and G31.41+0.3 are CH₂NH-rich sources next to Orion KL. In G10.47+0.03, the fractional abundance is 3.1 (±1.1) × 10⁻⁸. G31.41+0.3 has the fractional abundance of 8.8 (±8.5) × 10⁻⁹. The large error of the fractional abundance in G31.41+0.3 is due to scattering of the plots in its rotation diagram (Figure 5). Transitions at much higher energy levels will enable us to determine the column densities more accurately.

The fractional abundances of other sources are, 2.4 (±1.6) × 10⁻⁹, 2.8 (±1.6) × 10⁻⁹, 2.4 (±1.7) × 10⁻⁹, 1.4 (±1.1) × 10⁻¹⁰, and 1.4 (±1.0) × 10⁻⁹, in G34.3+0.2, W51 e1/e2, NGC 6334F, DR21 (OH), and G19.61–0.23, respectively. If an extended source size was assumed, our column density toward W51 e1/e2, 6.6 (±3.6) × 10¹³ cm⁻², was close to that reported by Dickens et al. (1997), 8 (±0.4) ×10¹³ cm⁻². From these results, we found that CH₂NH abundances range from ∼10⁻⁹ to ∼10⁻⁸.

CH₂NH was not detected with sufficient S/N ratios toward the other sources (W3(H₂O), G75.78+0.34, Cep A, NGC 2264 MMS3, NGC 1333 IRAS4B, IRAS 16293-2422). We derived upper limits for the column densities of CH₂NH for these sources corresponding to three sigma rms noise levels and excitation temperatures of 40 K. We summarize these results in Table 6.

As presented in Figure 6, the observed CH₂NH abundances are different by a factor of about 20. What characteristics cause such a large difference between CH₂NH abundances? In this section, we will discuss three hypotheses to explain the different CH₂NH abundances in terms of the distance, kinetic temperatures, and evolutionary phase of sources.

First, we will discuss the possibility that the differences in CH₂NH abundances simply originate from the different distances to the sources. In fact, the beam averaged CH₂NH column densities would be high for closer sources since they are less affected by beam dilution. However, the differences depending on the source sizes were small when we compared their CH₂NH fractional abundances. In Figure 7, we plotted the CH₂NH fractional abundances against the distance to the sources. It is difficult to see a correlation between CH₂NH fractional abundances and distances in Figure 7. The difference in CH₂NH fractional abundances should due to source properties rather than just an effect of beam dilution.

Zoom In Zoom Out Reset image size

The second possibility is that the different temperatures of CH₂NH sources might contribute to CH₂NH abundances. Theule et al. (2011) demonstrated that CH₂NH can be formed via the hydrogenation process from HCN on the dust surface (Figure 1). If CH₂NH is produced on the dust surface, the gas-phase CH₂NH abundance would be lower until the source temperature became high enough to sublimate species from the dust surface. Hernández-Hernández et al. (2014) reported rotation temperatures of CH₃CN, which is often used as a tracer of the kinetic temperature, toward 17 sources including G10.47+0.03, NGC 6334F,W51 e2, and W3(H₂O). The rotation temperatures were 499 K for G10.47+0.03, 408 K for NGC 6334F, 314 K for W51 e2, and 182 K for W3(H₂O). It is notable that W3(H₂O), where CH₂NH was not detected, has a lower temperature than the CH₂NH sources. Thus, this scenario could be convincing when the desorption temperature of CH₂NH is high (>180 K). One easy way to roughly estimate the desorption temperature is to use the Clausius–Clapeyron equation (2) and the ideal gas law (3),

$$\begin{eqnarray}&&{\mathrm{log}}_{10}\;P=4.559(1-(T/{T}_{{\rm{B}}}))\end{eqnarray} \tag{ 2 }$$

$$\begin{eqnarray}&&P=8.314{nT}\end{eqnarray} \tag{ 3 }$$

where T_B is the boiling point under the standard state, T is the sublimation temperature under the interstellar gas pressure P, and n is the number density of the species. We have to be cautious as the desorption temperature derived using this method is an approximate value, and it has uncertainty for two reasons. First, the Clausius–Clapeyron equation is applicable for the phase transition between a pure solid or liquid and a gas. However, the actual interstellar molecules would be far from pure material, and physisorbed on H₂O ice. Despite the differences in physical condition, the Clausius–Clapeyron equation is successful in reproducing the commonly accepted sublimation temperature of H₂O (from 90 to 100 K) and CO (from 15 to 18 K) in the ISM, assuming a density from 1 × 10⁵ to 1 × 10⁷ cm⁻³. Second, since the boiling point for CH₂NH has not been measured, we had to use the theoretically predicted value of 215(±15) K from a chemical database¹⁰ .

If we use the CH₂NH fractional abundance of 3.1 × 10⁻⁸ (the value observed in G10.47+0.03) and the molecular hydrogen number density of 10⁷ cm⁻³, the CH₂NH number density n is 0.31 cm⁻³. In our calculation, we assume initial T values of from 20 to 200 K (=T₀) and P is obtained from Equation (2), then, the next desorption temperature, T₁ is obtained with Equation (3). In these iterations between Equations (2) and (3) we are able to determine the final T value when the difference of successive temperatures becomes within 1 K. As a result, we found that the estimated desorption temperature of 40 K is quite low compared to the gas temperatures of star-forming regions.

We also calculated the desorption temperature in terms of the equilibrium of microscopic accretion and desorption processes, which may better approximate the interstellar situation. The accretion rate R_acc(i) (cm⁻³ s⁻¹) and the desorption timescale t_evap (s) for the species i are described in Hasegawa et al. (1992) as shown below:

$$\begin{eqnarray}&&{R}_{\mathrm{acc}}(i)={\sigma }_{d}\lt v(i)\gt {n}_{\mathrm{gas}}(i){n}_{d}\end{eqnarray} \tag{ 4 }$$

$$\begin{eqnarray}&&{t}_{\mathrm{evap}}={\nu }_{0}^{-1}\mathrm{exp}({E}_{D}/{{kT}}_{\mathrm{dust}}),\end{eqnarray} \tag{ 5 }$$

$$\begin{eqnarray}&&\mathrm{where}\;{\nu }_{0}=(2{n}_{s}{E}_{D}/{\pi }^{2}m).\end{eqnarray} \tag{ 6 }$$

σ_d, $\langle v(i)\rangle$, n_gas(i), n_dust(i), n_d, E_D, n_s, and m, respectively, represent the cross section of the dust, the averaged velocity of the gas species under a certain temperature, the gas and the dust surface phase number density of the species, the dust number density, the desorption energy, the surface density of the site (∼1.5 × 10¹⁵cm⁻²), and the mass of the molecule (4.83 × 10⁻²³ g for CH₂NH). Using the desorption timescale t_evap, the desorption rate R_des(i) (cm⁻³ s⁻¹) is given by ${t}_{\mathrm{evap}}^{-1}$n_dust(i). The desorption temperature is defined as the temperature that makes R_des(i) and R_acc(i) equal. Assuming a grain radius of 1 × 10⁻⁵ cm⁻¹, the physical cross section of 3.14 × 10⁻¹⁰ cm⁻² is used for σ_d. In Ruaud et al. (2015), E_D for CH₂NH is assumed to be the same as fpr CH₃OH (5530 K). Although the masses of CH₃OH and CH₂NH are similar, this assumption may overestimate the desorption temperature for CH₂NH, since the OH group in CH₃OH can strongly interact with H₂O molecules in the interstellar ice. Thus, if we use the desorption temperature of CH₃OH, we should be able to estimate the upper limit to the desorption temperature for CH₂NH.

We used our simulation results from model A, described in the Section 4.2, to estimate the number density of CH₂NH (n_gas and n_dust) just before the warm-up of the core. Under the molecular hydrogen number density of 1 × 10⁷ cm⁻³, n_gas (CH₂NH) and n_dust (CH₂NH) are roughly 1 × 10⁻⁷ cm⁻³ and 1 × 10⁻⁴ cm⁻³. With this gas density, the dust number density n_d can be estimated to be 1 × 10⁵ cm⁻³, assuming a gas-to-dust ratio of 0.01. Then, the upper limit to the desorption temperature can be calculated to be 160 K, which is comparable to the dust temperature in W3(H₂O). Since CH₂NH should be sufficiently desorbed even in W3(H₂O), we can conclude that temperature differences between hot core sources could not explain our observational results.

The third hypothesis is that the different CH₂NH abundances are related to the evolutionary phases of H ii regions inside a hot core. In the ISM, it is well known that star formation starts in a cold pre-stellar core. The gas temperature gradually rises when a protostar is formed inside the core, and finally frozen species on the dust are evaporated into the gas phase when the dust temperature becomes high enough. (As we have already mentioned, some of our sources are known to have high gas temperatures.) Following on from that, CH₂NH will gradually be photo-dissociated due to the strong UV field in the vicinity of an evolved H ii region inside the core. In this case, the CH₂NH-rich sources can be explained as less evolved H ii regions than the other sources, as is shown below.

The hydrogen recombination lines, which originate from H ii regions, are good tools to study the evolutionary state of an H ii region. In Figure 8 we compare the strength of a recombination line, H54β, between five CH₂NH sources. The observed line properties are summarized in Table 7. In G10.47+0.03 and G31.41+0.3, the observed H54β line is very weak or below the 3 sigma level. For Orion KL, it is well known that recombination lines originate from the foreground H ii region; Plambeck et al. (2013) reported that recombination lines were not detected inside the hot core of Orion KL. Our results suggest that the top three CH₂NH-rich sources possess less evolved H ii regions, while the H54β line is prominent toward W51 e1/e2 and NGC 6334F. These sources are considered to have more evolved H ii regions, supporting our scenario. We also note that past observational results also agree with this scenario. Cesaroni et al. (2010) claimed that G10.47+0.03 contains two deeply embedded hyper-compact H ii regions in the hot core while no detectable H ii region is present in G31.41+0.3. Although it is difficult to make direct comparisons with our observations, Jiménez-Serra et al. (2011) detected H40α, H34α, and H31α toward Cep A. Cep A may have an evolved H ii region inside it and the dissociation process might result in the non-detection of CH₂NH. Considering the above, the difference in evolutionary phase is the most plausible reason to explain the different CH₂NH abundances.

Zoom In Zoom Out Reset image size

It is difficult to discuss the actual formation processes of CH₂NH from our observations alone. In this section, we present the results of a gas-grain chemical model and discuss the chemistry of CH₂NH.

4.2.1. The Nautilus Chemical Model

4.2.2. The Physical Model

4.2.3. Gas and Ice Phase Formation of CH₂NH

As a result of our model, CH₂NH is formed both in the gas phase and on the dust grain surface. We summarize the main chemical reactions that form CH₂NH on the surfaces in Table 10. The dust surface reaction rates are proportional to both the probability of the reaction κ and the diffusion rates of reactant species (Hasegawa et al. 1992). κ is proportional to exp(-E${}_{A}^{1/2}$), where E_A represents the value of the activation barrier. The diffusion rates depend on the exponential of the dust temperature. For the activation barriers of the hydrogenation processes of HCN and CH₂NH, theoretically predicted values by Woon (2002) were employed.

Table 10.  Dust Surface Reactions Related to CH₂NH

Reaction	EA (K)
N + CH3 $\longrightarrow$ CH2NH	EA = 0 K
NH + CH2 $\longrightarrow$ CH2NH	EA = 0 K
NH2 + CH $\longrightarrow$ CH2NH	EA = 0 K
HCN + H $\longrightarrow$ H2CN	EA = 3647 K
HCN + H $\longrightarrow$ HCNH	EA = 6440 K
H2CN + H $\longrightarrow$ CH2NH	EA = 0 K
HCNH + H $\longrightarrow$ CH2NH	EA = 0 K
CH2NH + H $\longrightarrow$ CH3NH	EA = 2134 K
CH2NH + H $\longrightarrow$ CH2NH2	EA = 3170 K
CH3NH + H $\longrightarrow$ CH3NH2	EA = 0 K
CH2NH2 + H $\longrightarrow$ CH3NH2	EA = 0 K

Note. The dust surface reactions related to CH₂NH are shown. E_A represents the value of the activation barrier. Since radical species are so reactive, radical–radical reactions would have no activation barriers. The activation barriers for HCN and CH₂NH were cited from the theoretical study by Woon (2002).

Download table as:  ASCIITypeset image

The main reactions in the gas phase that form CH₂NH are summarized in Table 11. CH₂NH is formed both by dissociative recombination processes and neutral–neutral reactions. The rate coefficients of gas-phase reactions are given by the Arrhenius equation, using the coefficients α, β, and γ:

$$\begin{eqnarray}&&k(T)=\alpha {(T/300)}^{\beta }{e}^{-\gamma /T},\end{eqnarray} \tag{ 7 }$$

where T represents the gas temperature. As shown in Table 11, the dissociative recombination processes have larger reaction rates.

Table 11.  Gas-phase Reactions Related to CH₂NH

Reaction	α	β	γ
NH + CH3 $\longrightarrow$ CH2NH + H	1.3 × 10−10	1.7 × 10−1	0
CH + NH3 $\longrightarrow$ CH2NH + H	1.52 × 10−10	−5 × 10−2	0
CH2NH2+ + e− $\longrightarrow$ CH2NH + H	1.5 × 10−7	−5 × 10−1	0
CH3NH${}_{3}^{+}$ + e− $\longrightarrow$ CH2NH + H2 + H	1.5 × 10−7	−5 × 10−1	0
CH3NH${}_{2}^{+}$ + e− $\longrightarrow$ CH2NH + H2	1.5 × 10−7	−5 × 10−1	0

Note. α, β, and γ represent the coefficients for the Arrhenius equation.

Download table as:  ASCIITypeset image

4.2.4. Modeling Results and Discussion

The simulated gas-phase CH₂NH abundances (relative to the total proton density; hereafter represented by the symbol X) for the four models A, B, C, and D are presented in Figure 9 using different colors. The origin in the X axis corresponds to the time of start the collapse of the gas. We found that the general behavior of X[CH₂NH] along with the cloud evolution does not depend on the model. X[CH₂NH] is almost the same in each model in the collapsing phase (<1.4 × 10⁶ years). X[CH₂NH] decreases just before the warm-up phase (∼1.4 × 10⁶ years), since gas-phase species are accreted on the dust surface due to the high gas density and low temperature environment. After the warm-up phase, X[CH₂NH] increases dramatically. It is important to examine the detailed CH₂NH formation process in this phase to understand the dominant formation path. Following this, X[CH₂NH] decreases via gas-phase reactions with ions and/or radicals.

Zoom In Zoom Out Reset image size

Which chemical processes are the most dominant formation processes for CH₂NH? In Figure 10, we compare X[CH₂NH] and X[CH₃NH₂] in both the gas phase and the dust grain surface using model A. We can easily recognize that the dust surface abundance before the warm-up phase (∼1.4 × 10⁶ years) and the gas-phase abundance of CH₃NH₂ after warm-up are almost the same. CH₃NH₂ would mainly be formed on the dust surface before the birth of a star and evaporate to the gas phase. On the other hand, the dust surface X[CH₂NH] is very low (<10⁻¹²) compared to the gas-phase X[CH₂NH], since dust surface CH₂NH is efficiently hydrogenated to CH₃NH₂ during the collapsing phase. We also confirmed the same trend in models B, C, and D. Thus, gas-phase reactions would be the dominant formation process to explain gas-phase X[CH₂NH] in the protostellar stage.

Zoom In Zoom Out Reset image size

To discuss the most dominant gas-phase reactions in Table 11, we compared their formation rates k_rate(T, t) (cm⁻³ s⁻¹). Formation rates are represented by k_rate(T, t) = $k(T)$ × n₁(t) × n₂(t), where $k(T)$ is the temperature-dependent rate constant (cm³ s⁻¹) calculated with the Arrhenius equation, and n₁(t) and n₂(t) are the number densities (cm⁻³) of reactant species in each time step. In Figure 11(A)–(D), the time-dependent behaviors of k_rate(T, t) are shown for models A, B, C, and D, after the collapsing phase (t > 1.4 × 10⁶ years). In these figures, the formation rates of NH + CH₃, CH + NH₃, CH₂NH₂⁺ + e⁻, CH₃NH₃⁺ + e⁻, and CH₃NH₂⁺ + e⁻ are compared. As we can see through the comparison of models A and C, formation rates tend to be lower under a low density environment, due to the low number densities of reactant species. The difference between model A and model D originates from the different rate coefficients determined by the gas temperatures. However, we confirmed that NH + CH₃ reactions would be the most dominant formation process in all the models.

Zoom In Zoom Out Reset image size

Finally, we will mention the origin of these radicals. The dust surface abundances of radicals are very low during the collapsing phase, since unstable species are quickly hydrogenated. In Figure 12(A), we show the X[CH₃] in the gas phase and on the dust surface with red and blue lines, and the X[NH] in the gas phase and on the dust surface with green and black lines. The dust surface X[CH₃] and X[NH] are both nearly 10⁻¹² before the warm-up phase (∼1.4 × 10⁶ years). Since X[CH₃] and X[NH] in the gas phase are much higher than 10⁻¹² after the warm-up phase, gas-phase reactions are essential to explain the abundant X[CH₃] and X[NH]. In the data set of KIDA, radicals are formed via the destruction of complex species. We show that the gas-phase abundances of some species are related to the formation of CH₃ and NH using model A in Figures 12(B) and (C). The CH₃ radical originates from the destruction of larger saturated molecules, like CH₄, CH₃OH, and CH₃NH₂, by atoms and/or secondary UV radiation (Figure 12(B)). The NH radical is formed via the destruction of NH₃ (Figure 12(C)).

Zoom In Zoom Out Reset image size

We summarize the CH₂NH formation process here. During the cold collapsing phase, saturated complex molecules like NH₃, CH₄, CH₃OH, and CH₃NH₂ are formed on the dust surface. We note that the dust surface CH₂NH abundance is quite low in the collapsing phase. During the warm-up phase, these molecules evaporate to a gas, and NH and CH₃ radicals are produced through the gas-phase reactions using these evaporated molecules. Finally, newly formed radicals form CH₂NH.