betaFeedback

WorksPhysics and Astronomy

Rotational Excitation of HC5_5NH+^+ by He: Collision Dynamics and Astrophysical Applications

Abstract

The protonated cyanopolyyne HC5_5NH+^+ is an interstellar molecular ion of astrochemical interest, and understanding its rotational excitation is essential for interpreting its emission in molecular clouds. In this work, we construct a two-dimensional potential energy surface (PES) for the HC5_5NH+^+–He system using the CCSD(T)-F12B/cc-pVTZ-F12 method. Rotationally inelastic cross sections are calculated using the quantum close-coupling (CC) method, and the corresponding collisional rate coefficients are obtained by thermal averaging over a Maxwell–Boltzmann distribution. These rate coefficients are incorporated into a non-local thermodynamic equilibrium (non-LTE) radiative transfer model to investigate the excitation of HC5_5NH+^+ in the cold dark cloud TMC-1. A global χ2χ^2 minimization analysis of seven observed transitions yields best-fitting values of n(H2)=1.63×105n(\mathrm{H}_2)=1.63\times10^{5} cm−3^{-3} and N(HC5NH+)=5.09×1011N(\mathrm{HC}_5\mathrm{NH}^+)=5.09\times10^{11} cm−2^{-2}, with a minimum reduced χ2χ^2 of 0.87. The modelled line intensities are in good agreement with the observations. This study provides the first-principles collisional data needed to model the rotational excitation of HC5_5NH+^+ and offers new constraints on its physical and spectroscopic properties in TMC-1.

Introduction

Protonation gives molecules large permanent dipole moments and small rotational constants, making the resulting ions detectable radio tracers in cold dark clouds and providing indirect evidence for their neutral parents (Agúndez & Wakelam 2013). Among the carbon-chain species found in these environments, cyanopolyynes and their protonated counterparts, HC2n+1_{2n+1}N and HC2n+1_{2n+1}NH+^+ (n=0,1,2,…n = 0, 1, 2, \ldots), form an important homologous series. While neutral cyanopolyynes have been extensively studied (McGuire 2018), their protonated counterparts have been far less extensively detected. To date, the members reported in the protonated series include HCNH+^+ and the longer-chain species with n=1n = 1–3 (Ziurys & Turner 1986; Kawaguchi et al. 1994; Marcelino et al. 2020; Cabezas et al. 2022), with their column densities generally decreasing as the carbon-chain length increases.

HC5_5NH+^+ is the focus of this work. Its proposed interstellar detection is based on seven lines with nearly harmonic frequency spacing found in the Yebes 40 m survey of TMC-1 (Marcelino et al. 2020), with HC5_5NH+^+ as their most likely carrier. This assignment is supported by theoretical rotational constants and by the similarity between the abundance ratios of HC5_5N/HC5_5NH+^+ and HC3_3N/HC3_3NH+^+. However, in the absence of laboratory rotational spectroscopy, its detection remains tentative. Laboratory infrared spectroscopy has provided direct frequency references for the identification of HC3_3NH+^+ (Lee & Amano 1987; Kawaguchi et al. 1990), whereas the assignments of longer-chain protonated species have relied largely on rotational constants and dipole moments predicted by high-level coupled-cluster calculations (Botschwina et al. 1997; Heyl et al. 1997). We therefore adopt this assignment as a working hypothesis for the collisional and radiative transfer analyses.

The observationally inferred column densities of these ions depend on the rotational level populations, and collisional excitation in cold dark clouds is generally insufficient to maintain local thermodynamic equilibrium (LTE). Studies of HCNH+^+ collisions with He and H2_2, together with hyperfine-resolved data for its isotopologues, have demonstrated how dedicated potential energy surfaces (PESs) and state-to-state rate coefficients affect spectral predictions (Bop & Lique 2023; Bop et al. 2024, 2026). However, because accurate rate coefficients for long-chain molecules are extremely demanding to compute, quantum scattering results have so far been reported only for HC3_3NH+^+–He [3], and none are available for HC5_5NH+^+ or longer chains. Moreover, the effects of increasing chain length on the potential anisotropy and transition propensities cannot be directly extrapolated from shorter-chain systems.

In this work, we construct a two-dimensional ab initio PES for HC5_5NH+^+–He and perform CC calculations to obtain rotationally inelastic cross sections and low-temperature rate coefficients. The resulting collisional data are then used in non-LTE radiative transfer calculations and fitted to the observed line intensities towards TMC-1, in order to determine its excitation properties and constrain its column density. The remainder of this paper is organised as follows. Section 2 describes the construction and Legendre expansion of the HC5_5NH+^+–He PES. Section 3 presents the calculated cross sections and rate coefficients. Section 4 discusses the astrophysical applications and the non-LTE radiative transfer results, and § 5 summarises our conclusions.

Theoretical methods and computational details

Potential energy surface

Within the rigid-rotor approximation, the molecular geometry of HC₅NH⁺ is fixed at the equilibrium structure of its ground state, and only its overall rotation is considered. The lowest vibrational frequency of the ion is approximately 99 cm⁻¹ (Marcelino et al. 2020). At a typical interstellar temperature of 20 K, the Boltzmann factor for the first vibrationally excited state, exp⁡(−99/kBT)\exp(-99/k_{\mathrm{B}}T), is only about 0.1%. Vibrational excitation can therefore be neglected over the temperature range considered in this work (Lique et al. 2006; Faure et al. 2020). This approximation has been widely adopted for similar weakly interacting systems, including SiC₄–He (Chefai et al. 2024), HNSi/HSiN–He (Tian et al. 2026a), and CaO/CaS–He (Tian et al. 2026b).

HC₅NH⁺ is a linear conjugated polyyne ion. In the centre-of-mass frame, the position of He relative to the linear HC₅NH⁺ molecule in its electronic ground state is described by the Jacobi coordinates (R,θ)(R,\theta), where RR is the distance between He and the centre of mass of the ion, and θ=0∘\theta=0^\circ corresponds to the arrangement HC₅NH⁺—He, as illustrated in Figure 1. The bond lengths are calculated at the CCSD(T)-F12B/cc-pVTZ-F12 level of theory and are as follows: rH1C1=1.0688r_{\mathrm{H}_{1}\mathrm{C}_{1}}=1.0688 Å, rC1C2=1.2148r_{\mathrm{C}_{1}\mathrm{C}_{2}}=1.2148 Å, rC2C3=1.3517r_{\mathrm{C}_{2}\mathrm{C}_{3}}=1.3517 Å, rC3C4=1.2247r_{\mathrm{C}_{3}\mathrm{C}_{4}}=1.2247 Å, rC4C5=1.3413r_{\mathrm{C}_{4}\mathrm{C}_{5}}=1.3413 Å, rC5N=1.1527r_{\mathrm{C}_{5}\mathrm{N}}=1.1527 Å, and rNH2=1.0064r_{\mathrm{N}\mathrm{H}_{2}}=1.0064 Å.

Geometry of the HC₅NH⁺–He collision system and the Jacobi coordinates

Figure 1. Geometry of the HC₅NH⁺–He collision system and the Jacobi coordinates.

The interaction energy is calculated using the explicitly correlated coupled-cluster method CCSD(T)-F12B and the cc-pVTZ-F12 basis set (Knizia et al. 2009; Werner et al. 2012). The radial grid covers R=2R=2–30 Å, spanning the short-, intermediate-, and long-range regions. The angle θ\theta is sampled uniformly from 00 to 180∘180^\circ in steps of 5∘5^\circ, yielding a total of 2553 geometries. The basis-set superposition error (BSSE) is corrected using the Boys–Bernardi counterpoise scheme:

V(R,θ)=EHC5NH+–He(R,θ)−EHC5NH+(R,θ)−EHe(R,θ)(1)V(R,\theta)=E_{\mathrm{HC}_{5}\mathrm{NH}^{+}\text{--He}}(R,\theta)-E_{\mathrm{HC}_{5}\mathrm{NH}^{+}}(R,\theta)-E_{\mathrm{He}}(R,\theta) \tag*{(1)}

where EHC5NH+–HeE_{\mathrm{HC}_{5}\mathrm{NH}^{+}\text{--He}}, EHC5NH+E_{\mathrm{HC}_{5}\mathrm{NH}^{+}}, and EHeE_{\mathrm{He}} are the energies of the complex, the HC₅NH⁺ monomer, and the He atom, respectively, all calculated within the same set.

The resulting PES is shown in Figure 2. It exhibits pronounced axial asymmetry and a lateral double-minimum structure. Its global minimum lies on the axis at the protonated end (θ=0.2∘\theta=0.2^\circ), reaching a depth of 194.55 cm⁻¹ at R=6.0R=6.0 Å. By contrast, the well at the N end (θ=180∘\theta=180^\circ) is much shallower. Laterally, the interaction exhibits two asymmetric minima, separated by a shallow saddle point near θ=90∘\theta=90^\circ with a depth of about 58 cm⁻¹; consequently, the 60 cm⁻¹ contour in the lateral region splits into two closed loops. The PES exhibits four distinct low-energy regions associated with the H₁ (+0.45680), C₂ (+1.30156), C₄ (+1.63499), and H₂ (+0.29715) sites of HC₅NH⁺, corresponding to the four main low-energy regions in the contour map. Their spatial distribution reflects the pronounced anisotropy and end-group asymmetry of the interaction potential, with the deepest well located near the protonated H₂ end.

Contour plot of the HC₅NH⁺–He interaction potential $V(R,\theta)$

Figure 2. Contour plot of the HC₅NH⁺–He interaction potential V(R,θ)V(R,\theta) (in cm⁻¹). Blue contours indicate positive values, while red contours indicate negative values.

Legendre expansion of the PES

In the short-range region, the raw interaction energy VaV_{\mathrm{a}} exhibits an unphysically steep increase due to orbital overlap and the numerical behaviour of the Boys–Bernardi correction. To address this issue, before performing the Legendre expansion, we apply the smoothing and scaling method of González-Sánchez et al. (2024) to the raw ab initio energies. Denoting the raw and smoothed energies by VaV_{\mathrm{a}} and VsV_{\mathrm{s}}, respectively, the procedure is defined as

Vs={Va,Va≤VT1,VT1+(VT2−VT1)S(Va−VT1VT2−VT1),VT1<Va≤VT2,VT1+2π(VT2−VT1),Va>VT2,(2)V_{\mathrm{s}}= \begin{cases} V_{\mathrm{a}}, & V_{\mathrm{a}}\le V_{\mathrm{T1}},\\ V_{\mathrm{T1}}+(V_{\mathrm{T2}}-V_{\mathrm{T1}})S\left(\dfrac{V_{\mathrm{a}}-V_{\mathrm{T1}}}{V_{\mathrm{T2}}-V_{\mathrm{T1}}}\right), & V_{\mathrm{T1}}<V_{\mathrm{a}}\le V_{\mathrm{T2}},\\ V_{\mathrm{T1}}+\dfrac{2}{\pi}(V_{\mathrm{T2}}-V_{\mathrm{T1}}), & V_{\mathrm{a}}>V_{\mathrm{T2}}, \end{cases} \tag*{(2)}

where the threshold parameters are VT1=2000V_{\mathrm{T1}}=2000 cm⁻¹ and VT2=10000V_{\mathrm{T2}}=10000 cm⁻¹, and the switching function is S(x)=2πsin⁡[π2sin⁡(π2x)]S(x)=\dfrac{2}{\pi}\sin\left[\dfrac{\pi}{2}\sin\left(\dfrac{\pi}{2}x\right)\right]. This scheme ensures that VsV_{\mathrm{s}} and its first derivative are continuous at the joining points, while limiting the high-energy repulsive wall to a finite value, thereby avoiding numerical instabilities in the scattering calculations.

For use in the scattering calculations, the two-dimensional PES is expanded in Legendre polynomials:

V(R,θ)=∑λ=0λmax⁡Vλ(R)Pλ(cos⁡θ),(3)V(R,\theta)=\sum_{\lambda=0}^{\lambda_{\max}} V_{\lambda}(R)P_{\lambda}(\cos\theta), \tag*{(3)}

where PλP_{\lambda} is the Legendre polynomial of order λ\lambda and Vλ(R)V_{\lambda}(R) are the radial coefficients. To reduce the computational cost while maintaining accuracy, the expansion is truncated at λmax⁡=56\lambda_{\max}=56, for which the mean deviation of the reconstructed PES from the ab initio values is less than 0.5%.

Figure 3(a) shows the RR dependence of the first five radial coefficients. Among the anisotropic terms with λ>0\lambda>0, V2V_{2} has a markedly larger amplitude than the others, followed by V4V_{4}, indicating the dominance of the even-order terms. The odd terms V1V_{1} and V3V_{3} have appreciable amplitudes only at short range and decay rapidly with increasing RR. The dominance of the even-order terms reflects the strong long-range induction and dispersion attraction experienced by He as it approaches the conjugated carbon chain from the side, which also explains the propensity for Δj=±2\Delta j=\pm2 transitions during collisions.

Radial dependence of the first five Legendre radial coefficients and long-range interaction potential

Figure 3. (a) Radial dependence of the first five Legendre radial coefficients Vλ(R)V_\lambda(R) (λ=0\lambda= 0–4, in cm−1^{-1}), as functions of RR for the HC5_5NH+^+ system. (b) Radial dependence of the long-range interaction potential.

The long-range fit provides a further test of the asymptotic behaviour of the PES. Figure 3(b) shows that, in the long-range region R≥16R\geq16 Å, the ab initio points agree well with both V0V_{0} and the −C/R4-C/R^{4} fit, consistent with an R−4R^{-4} asymptote dominated by the charge-induced-dipole interaction (Buckingham 1978). For R<16R<16 Å, the ab initio points lie systematically above the −C/R4-C/R^{4} asymptote, indicating that contributions from repulsion and higher-order terms such as R−6R^{-6} are no longer negligible. This also explains why the analytic potential adopts the asymptotic form only in the long-range region, while the inner region is described by a Legendre expansion. The two branches are joined such that their first derivatives are continuous, avoiding unphysical kinks or barriers in the low-energy scattering calculations. The deviation between the ab initio points and the fit decreases rapidly with increasing RR and tends to zero, supporting the use of this analytic PES in subsequent low-energy scattering calculations.

Scattering cross sections and rate coefficients

Using the PES described above, we employ the MOLSCAT code (Hutson & Le Sueur 2019) to calculate the inelastic scattering of an atom with HC5_5NH+^{+} within the CC framework (Arthurs & Dalgarno 1960), obtaining state-to-state cross sections for collision energies up to 300 cm−1^{-1}. For HC5_5NH+^{+}, we adopt the rotational constant B0=0.043224B_{0}=0.043224 cm−1^{-1} and the centrifugal distortion constant D0=9.11×10−10D_{0}=9.11\times10^{-10} cm−1^{-1} (Marcelino et al. 2020).

The partial-wave sum for the integral cross section from jj to j′j' is given by:

σj→j′(E)=πkj2(2j+1)∑J=0∞∑l=∣J−j∣J+j∑l′=∣J−j′∣J+j′(2J+1)∣δjj′δll′−Sjj′,ll′J(E)∣2,(4)\sigma_{j\to j'}(E)=\frac{\pi}{k_{j}^{2}(2j+1)}\sum_{J=0}^{\infty}\sum_{l=|J-j|}^{J+j}\sum_{l'=|J-j'|}^{J+j'}(2J+1)\left|\delta_{jj'}\delta_{ll'}-S_{jj',ll'}^{J}(E)\right|^{2}, \tag*{(4)}

where EE is the collision energy, JJ is the total angular momentum, ll and l′l' are the orbital angular momenta of the initial and final channels, respectively, and δ\delta is the Kronecker delta. The channel wavenumber is defined as kj=[2μ(E−Ej)]1/2/ℏk_{j}=[2\mu(E-E_{j})]^{1/2}/\hbar, where μ=3.8023948\mu=3.8023948 amu (unified atomic mass units) is the reduced mass of the HC5_5NH+^{+}–He system. The rotational energy levels EjE_{j} of the molecule are given by:

Ej=B0j(j+1)−D0j2(j+1)2(5)E_{j}=B_{0}j(j+1)-D_{0}j^{2}(j+1)^{2} \tag*{(5)}

To ensure the convergence of transition cross sections involving rotational levels up to j≤20j\leq20, the rotational basis truncation parameter jmax⁡j_{\max} was systematically optimized for different energy ranges. Accordingly, an energy-dependent rotational basis was adopted: jmax⁡=29j_{\max}=29 for E≤100E\leq100 cm−1^{-1} and jmax⁡=35j_{\max}=35 for 100<E≤300100<E\leq300 cm−1^{-1}. For the transitions considered, the numerical uncertainty of the computed cross sections using these jmax⁡j_{\max} values is less than 5%, validating that the adopted basis provides sufficiently converged cross sections for the target levels. The corresponding integration step size control ensures 50, 20, and 10 steps per half-wavelength for E=0E=0–20, 20–100, and 100–300 cm−1^{-1}, respectively. To adequately sample the threshold and resonance regions, the collision-energy grid spacings are set to 0.05, 0.1, 1, and 5 cm−1^{-1} for the energy intervals of E≤30E\leq30, 30<E≤8030<E\leq80, 80<E≤15080<E\leq150, and 150<E≤300150<E\leq300 cm−1^{-1}, respectively. At each energy point, the total angular momentum JtotJ_{\mathrm{tot}}, controlled by JTOTU, is iteratively increased until convergence is reached (with an upper limit set to 99999). The summation process terminates automatically when the maximum changes in both diagonal and off-diagonal cross sections fall below their respective thresholds (DTOL = 0.1 Å2^{2}, OTOL = 0.005 Å2^{2}). To fully resolve the resonance features, the radial coupled equations are divided into two propagation stages: a log-derivative propagator is used in the short-range region (Manolopoulos 1986), and a variable-step Airy propagator is used in the long-range region (Alexander & Manolopoulos 1987). All calculations are solved using the MOLSCAT program (Hutson & Le Sueur 2019).

Figure 4 shows the de-excitation dynamics for high-jj initial states, and Figure 4(a) presents the de-excitation cross sections for j=13j=13–19. Below 10 cm−1^{-1}, shape resonances associated with tunnelling through the centrifugal barrier and Feshbach resonances arising from quasi-bound states in the potential wells produce dense oscillations [3]; at very low energies, the cross section decreases rapidly with increasing energy, approximately following the long-range capture E−1/2E^{-1/2} scaling (Gioumousis & Stevenson 1958). Above 30 cm−1^{-1}, the cross sections depart from this scaling and decline slowly, indicating that long-range capture is no longer the limiting factor and that the dynamics become governed by the short-range anisotropic potential and exchange repulsion. At the low-energy end, the Δj=−1\Delta j=-1 cross section is the largest, whereas the Δj=−3\Delta j=-3 cross section is the smallest over most of the energy range; as the energy increases, the decrease in the Δj=−2\Delta j=-2 cross section becomes markedly less pronounced and its cross section becomes the largest, with the clearest advantage at high energy. This energy-dependent selectivity is consistent with the anisotropy of the PES being dominated by the V2V_{2} term.

De-excitation integral cross sections and rate coefficients

Figure 4. (a) De-excitation integral cross sections for initial states j=13j = 13–19 as functions of collision energy (on a log–log scale). (b) The corresponding de-excitation rate coefficients as functions of temperature.

Maxwell–Boltzmann averaging of these state-to-state cross sections over a 1 K grid from 3 to 50 K yields the state-to-state (de-)excitation rate coefficients for j≤19j\leq19.

kj→j′(T)=(8πμkB3T3)1/2∫0∞Ekσj→j′(Ek)e−Ek/kBT dEk,(6)k_{j\to j'}(T)=\left(\frac{8}{\pi\mu k_{\mathrm{B}}^{3}T^{3}}\right)^{1/2}\int_{0}^{\infty}E_{k}\sigma_{j\to j'}(E_{k})e^{-E_{k}/k_{\mathrm{B}}T}\,dE_{k}, \tag*{(6)}

where kj→j′k_{j\to j'} is the de-excitation rate coefficient from the upper level jj to the lower level j′j', and kBk_{\mathrm{B}} is the Boltzmann constant. The corresponding excitation rate coefficients, kj′→jk_{j'\to j}, are obtained from the principle of detailed balance (En Abdelhadi et al. 2012):

kj′→j(T)=2j+12j′+1exp⁡(−Ej−Ej′kBT)kj→j′(T).(7)k_{j'\to j}(T)=\frac{2j+1}{2j'+1}\exp\left(-\frac{E_{j}-E_{j'}}{k_{\mathrm{B}}T}\right)k_{j\to j'}(T). \tag*{(7)}

Figure 4(b) shows the temperature dependence of the de-excitation rate coefficients: the Δj=−2\Delta j=-2 channel increases monotonically and substantially with temperature, becoming dominant at higher temperatures. This behaviour is directly related to its cross section, which decreases most slowly at high collision energies and forms a plateau above 30 cm−1^{-1} (Figure 4a): as the temperature rises, high-energy collisions gain weight and the rate continues to grow. The Δj=−1\Delta j = -1 cross section decreases steadily with energy, and its thermally averaged rate coefficient becomes nearly constant above approximately 20 K, consistent with ion–induced-dipole Langevin capture (Gioumousis & Stevenson 1958); the Δj=−3\Delta j = -3 channel has the smallest rate and the weakest temperature dependence. Differences among initial states within the same Δj\Delta j are small; only for Δj=−2\Delta j = -2 does the rate coefficient increase slightly with increasing initial jj, following the same trend as the high-energy cross sections in Figure 4(a). Thermal averaging smooths out the dense resonances at low collision energies, resulting in smooth temperature dependences of the rate coefficients. The weak dependence on the initial rotational level also indicates that this system is less sensitive to the rotational quantum number than neutral molecular systems.

Astrophysical applications

In low-density clouds such as TMC-1, collisional excitation is generally insufficient to fully thermalise the level populations (Bop & Lique 2025), so the column density derived by Marcelino et al. (2020) under the LTE assumption may be subject to systematic bias. Using the rate coefficients derived above, we can quantitatively assess this effect with non-LTE modelling.

Non-LTE radiative transfer simulation is computed with RADEX (van der Tak et al. 2007). The code adopts the escape-probability approximation and iteratively solves the statistical equilibrium equations for the level populations, assuming a uniform-sphere geometry. Because no direct HC5_5NH+^+–H2_2 collisional data are currently available, the H2_2 rate coefficients are obtained by scaling those for HC5_5NH+^+–He. Since He is not a direct surrogate for H2_2, the scaling must account for differences in both reduced mass and polarisability:

f2=kH2kHe=αst(H2)/μHC5NH+–H2αst(He)/μHC5NH+–He(8)f_2 = \frac{k_{\mathrm{H}_2}}{k_{\mathrm{He}}} = \sqrt{\frac{\alpha^{\mathrm{st}}(\mathrm{H}_2)/\mu_{\mathrm{HC}_5\mathrm{NH}^+\text{--}\mathrm{H}_2}}{\alpha^{\mathrm{st}}(\mathrm{He})/\mu_{\mathrm{HC}_5\mathrm{NH}^+\text{--}\mathrm{He}}}} \tag*{(8)}

where αst(He)=1.384 a03\alpha^{\mathrm{st}}(\mathrm{He}) = 1.384\ a_0^3 and αst(H2)=5.314 a03\alpha^{\mathrm{st}}(\mathrm{H}_2) = 5.314\ a_0^3 are the polarisabilities of He and H2_2, respectively; the calculated scaling factor is f2=2.7f_2 = 2.7 (González-Sánchez et al. 2024). Because calculations involving He are computationally less demanding, He is often used as a proxy when H2_2 collision data are unavailable (Bop & Lique 2025). The linewidth is set to the mean of the seven transitions listed in Table 1 of Marcelino et al. (2020), approximately 0.65 km s−1^{-1}.

Figure 5(a) and Figure 5(b) present the non-LTE radiative transfer results for the representative transitions J=14→13J = 14 \to13 (top) and J=18→17J = 18 \to17 (bottom), respectively. Each transition is shown in four panels corresponding to Tkin=5T_{\mathrm{kin}} = 5, 10, 15, and 20 K. Solid lines show the brightness temperature TBT_{\mathrm{B}}, long-dashed black lines the excitation temperature TexT_{\mathrm{ex}}, and short-dashed purple lines the optical depth τ\tau multiplied by 100; the three colours correspond to the three column densities. Because τ<0.12\tau< 0.12 in all models, the lines are entirely optically thin and radiative trapping is absent, and TBT_{\mathrm{B}} is approximately proportional to the column density at low densities.

Plots of $T_{\rm ex}$ and $T_{\rm B}$ for two rotational transitions

Figure 5. TexT_{\rm ex} and TBT_{\rm B} for the J=14→13J = 14 \rightarrow13 (top) and J=18→17J = 18 \rightarrow17 (bottom) transitions as functions of the H2_2 volume density at three HC5_5NH+^+ column densities.

At the low-density (n(H2)≲103 cm−3n(\mathrm{H}_2) \lesssim10^3\ \mathrm{cm}^{-3}), both transitions are subthermally excited (Tex<TkinT_{\mathrm{ex}} < T_{\mathrm{kin}}). As n(H2)n(\mathrm{H}_2) increases, collisions become increasingly dominant, and TexT_{\mathrm{ex}} approaches TkinT_{\mathrm{kin}}. TBT_{\mathrm{B}} levels off for n(H2)≳105 cm−3n(\mathrm{H}_2) \gtrsim10^5\ \mathrm{cm}^{-3}. The variation of τ\tau with density merely reflects differences in the populations of the upper and lower levels. Under these conditions, the plateau level of TBT_{\mathrm{B}} is determined mainly by TexT_{\mathrm{ex}}, which has approached TkinT_{\mathrm{kin}} at these densities.

A comparison of the two transitions reveals a clear rotational-level effect: because J=18→17J = 18 \to17 involves a higher upper-level energy and a larger energy gap, a higher density is required for it to thermalise. At the same n(H2)n(\mathrm{H}_2), as long as the transitions are not fully thermalised, TexT_{\mathrm{ex}} of J=14→13J = 14 \to13 always exceeds that of J=18→17J = 18 \to17; the difference is most pronounced at intermediate densities and then narrows as both lines approach TkinT_{\mathrm{kin}}. For the Tkin=20T_{\mathrm{kin}} = 20 K model, at n(H2)=104 cm−3n(\mathrm{H}_2) = 10^4\ \mathrm{cm}^{-3}, TexT_{\mathrm{ex}} of J=14→13J = 14 \to13 is already about 16.6 K, whereas that of J=18→17J = 18 \to17 is only about 8.4 K, roughly half as high, showing that the latter thermalises significantly later. As the density rises further (e.g. to 3×104 cm−33 \times10^4\ \mathrm{cm}^{-3}), the J=18→17J = 18 \to17 transition still lags behind and approaches the kinetic temperature only at about 105 cm−310^5\ \mathrm{cm}^{-3}, indicating that the system is progressing towards a near-LTE state.

The density at which these plateaus occur can be understood in terms of critical density (ncrn_{\mathrm{cr}}). For a transition u→lu \to l, the critical density is the collider density at which the rates of radiative and collisional de-excitation are equal (van der Tak et al. 2007):

ncr(u→l,T)=Au→l∑l<uku→l(T)(9)n_{\mathrm{cr}}(u \to l,T) = \frac{A_{u\to l}}{\sum_{l<u} k_{u\to l}(T)} \tag*{(9)}

where uu and ll denote the initial and final levels, Au→lA_{u\to l} is the Einstein A coefficient for spontaneous emission from uu to ll. For the transitions considered here, we calculated the Einstein A coefficients for all Δj=−1\Delta j = -1 channels within j≤19j \leq19; the data are available as described in the Data availability section. Here, ku→lk_{u\to l} is the rate coefficient for collisional de-excitation from the initial level uu to each lower level ll, the summation in the denominator runs over all possible de-excitation channels.

J→J′J \to J'5 K10 K15 K20 K
13 → 126.086×1036.086 \times10^36.272×1036.272 \times10^36.266×1036.266 \times10^36.188×1036.188 \times10^3
14 → 137.430×1037.430 \times10^37.652×1037.652 \times10^37.645×1037.645 \times10^37.551×1037.551 \times10^3
15 → 148.929×1038.929 \times10^39.181×1039.181 \times10^39.171×1039.171 \times10^39.059×1039.059 \times10^3
16 → 151.064×1041.064 \times10^41.094×1041.094 \times10^41.094×1041.094 \times10^41.080×1041.080 \times10^4
17 → 161.246×1041.246 \times10^41.285×1041.285 \times10^41.286×1041.286 \times10^41.272×1041.272 \times10^4
18 → 171.460×1041.460 \times10^41.506×1041.506 \times10^41.507×1041.507 \times10^41.491×1041.491 \times10^4
19 → 181.686×1041.686 \times10^41.741×1041.741 \times10^41.742×1041.742 \times10^41.721×1041.721 \times10^4

Table 1. Critical densities ncrn_{\mathrm{cr}} of the seven observed transitions at 5–20 K.

Table 1 lists the values for the seven observed transitions at 5–20 K. As the rotational quantum number increases, Au→lA_{u\to l} increases rapidly with the cube of the transition frequency, and ncrn_{\mathrm{cr}} grows from 6.1×103 cm−36.1 \times10^3\ \mathrm{cm}^{-3} for 13→1213 \to12 to 1.7×104 cm−31.7 \times10^4\ \mathrm{cm}^{-3} for 19→1819 \to18; for a given transition, the ncrn_{\mathrm{cr}} varies by less than 4% over 5–20 K, indicating a much weaker temperature dependence than that found in neutral molecular systems (e.g. HC3_3N–He; Wernli et al. 2007). The typical density of TMC-1 (about 104 cm−310^4\ \mathrm{cm}^{-3}) lies within this range: the low-jj transitions are nearly thermalised but not fully, with TexT_{\mathrm{ex}} values slightly below TkinT_{\mathrm{kin}}, and their line intensities mainly constrain the column density; the high-jj transitions remain subthermally excited and are sensitive to density, providing the density information. This analysis indicates the density sensitivity of each line without detailed fitting, and guides the density range of the radiative transfer grid and the joint fit. Furthermore, it demonstrates that non-LTE treatment is essential under the physical conditions of TMC-1 and that the density sensitivity is confined to the subthermal regime, where the line ratios respond to collisions.

To constrain the observational parameters, we perform a joint fit to the seven lines observed by Marcelino et al. (2020) using the χ2\chi^2 statistic of Lamption et al. (1976):

χ2=∑i[Icalc,i−Iobs,iσI,i]2(10)\chi^2 = \sum_i \left[\frac{I_{\mathrm{calc},i} - I_{\mathrm{obs},i}}{\sigma_{I,i}}\right]^2 \tag*{(10)}

where Icalc,iI_{\mathrm{calc},i} and Iobs,iI_{\mathrm{obs},i} are the calculated and observed integrated line intensities, respectively, and σI,i\sigma_{I,i} is the observational uncertainty.

Figure 6 shows the two-dimensional χ2\chi^2 minimisation grid, where the colour scale represents Δχ2=χ2−χmin2\Delta\chi^2 = \chi^2 - \chi^2_{\mathrm{min}} and the silver star marks the best-fitting solution; the contours, from the innermost to the outermost, correspond to Δχ2=2.30\Delta\chi^2 = 2.30, 6.18, and 11.83, i.e. the 1σ1\sigma, 2σ2\sigma, and 3σ3\sigma confidence regions for a two-parameter fit (Lamption et al. 1976). The best fit is located at n(H2)=1.63×105 cm−3n(\mathrm{H}_2) = 1.63 \times10^5\ \mathrm{cm}^{-3} and N(HC5NH+)=5.09×1011 cm−2N(\mathrm{HC}_5\mathrm{NH}^+) = 5.09 \times10^{11}\ \mathrm{cm}^{-2}. The Δχ2\Delta\chi^2 contours form a horizontal band: the 1σ1\sigma region constrains the column density to approximately (4.5(4.5–5.5)×1011 cm−25.5) \times10^{11}\ \mathrm{cm}^{-2} and extends towards higher densities. Because the seven lines have different ncrn_{\mathrm{cr}} and respond differently to density in the subthermal regime, the joint fit sets a lower limit on n(H2)n(\mathrm{H}_2). At the best-fit density of 1.63×105 cm−31.63 \times10^5\ \mathrm{cm}^{-3}, which exceeds all ncrn_{\mathrm{cr}}, the TexT_{\mathrm{ex}} of the lines approaches TkinT_{\mathrm{kin}}, and the contours flatten along the density direction; the resulting density degeneracy is therefore a natural consequence of the loss of diagnostic information after thermalisation, rather than a deficiency of the fitting procedure itself.

Distribution of $\Delta\chi^2$ in the $n({\rm H}_2)$–$N({\rm HC}_5{\rm NH}^+)$ plane

Figure 6. Distribution of Δχ2=χ2−χmin⁡2\Delta\chi^2 = \chi^2 - \chi^2_{\min} in the n(H2)n({\rm H}_2)–N(HC5NH+)N({\rm HC}_5{\rm NH}^+) plane at Tkin=7.8T_{\rm kin} = 7.8 K. The best-fitting solution is marked by the silver star. The white solid, black dashed, and red short-dashed contours (from inside out) denote the 1σ1\sigma, 2σ2\sigma, and 3σ3\sigma confidence level (Δχ2=2.30\Delta\chi^2 = 2.30, 6.18, and 11.83, respectively).

To further evaluate the goodness-of-fit and compare it with previous studies, Figure 7 displays the one-dimensional χ2\chi^2 curve as a function of column density at the best-fit density of n(H2)=1.63×105 cm−3n(\mathrm{H}_2) = 1.63 \times10^5\ \mathrm{cm}^{-3}. The curve exhibits a single minimum at N=5.09×1011 cm−2N = 5.09 \times10^{11}\ \mathrm{cm}^{-2}, yielding a reduced χmin2=0.87\chi^2_{\mathrm{min}} = 0.87. A reduced χ2\chi^2 value ≤1\leq1 is not uncommon in astrochemical analyses, typically indicating either overestimated observational uncertainties or an excellent fit to the data, as seen in other non-LTE studies (e.g. Agúndez et al. (2023)). Compared to the value of N=(7.5±2.2)×1011N = (7.5 \pm2.2) \times10^{11} cm−2^{-2} derived from the LTE rotational diagram by Marcelino et al. (2020), our result is lower by a factor of ∼1.5\sim1.5, yet the two values are consistent within 1σ1\sigma. The direction and magnitude of this offset align exactly with the expected systematic difference from the LTE assumption: LTE tends to overestimate the populations of high-jj levels, and consequently the column density (Agúndez et al. (2023)). The agreement in magnitude indicates that the current rate coefficients introduce no significant systematic bias in the radiative transfer modelling, thereby further validating the non-LTE treatment and the spectral line assignment.

Variation of $\chi^2$ with the column density $N({\rm HC}_5{\rm NH}^+)$

Figure 7. Variation of χ2\chi^2 with the column density N(HC5NH+)N({\rm HC}_5{\rm NH}^+) at the best-fitting value of n(H2)n({\rm H}_2). The solid blue curve shows the fit, the dashed red line marks the best-fit column density, and the grey band shows the column density range derived from the LTE rotational diagram of Marcelino et al. (2020).

Conclusions

We have investigated the rotationally inelastic collisions of HC5_5NH+^+ with He and examined their implications for the excitation of HC5_5NH+^+ in interstellar clouds. The calculated collision dynamics exhibit a clear propensity for ∣Δj∣=2|\Delta j| = 2 transitions, which is closely related to the dominant contribution of the V2V_2 component to the PES anisotropy. This highlights the important role of the interaction potential in determining the rotational transition propensities of the system.

The non-LTE analysis towards TMC-1 indicates that the observed HC5_5NH+^+ lines are subthermally excited. While the derived column density agrees with the LTE estimate within 1σ1\sigma, the volume density is constrained only from below, suggesting that HC5_5NH+^+ is therefore a robust column-density tracer, while its density diagnostic power is limited by thermalisation. Future studies could extend this work to other homologous ions (e.g. HC7NH+\mathrm{HC}_{7}\mathrm{NH}^{+}) and isotopologues (e.g. HC15NH+\mathrm{HC}^{15}\mathrm{NH}^{+}) to investigate whether the “even-Δj\Delta j dominance” is a general collisional characteristic of long-chain protonated cyanopolyynes. Furthermore, the highly accurate state-to-state collisional rate coefficients presented here will provide a robust dataset to support radio observations and non-LTE analyses of HC2n+1NH+\mathrm{HC}_{2n+1}\mathrm{NH}^{+} and related species in molecular clouds.

Data Availability

The collisional rate coefficients and the associated molecular data required for the RADEX calculations will be deposited in the Zenodo repository and submitted to the LAMDA database upon acceptance for publication.

Acknowledgements

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

This paper has been typeset from a \TeX/\LaTeX file prepared by the author.

References

  1. [3]Al Mogren M. M., Ben Abdallah D., Al Harbi S. D. A., Al Salhi M. S., Hochlaf M., 2021, Mon. Not. R. Astron. Soc., 503, 2902

Paper details

vlnc:W_98snmw7b
Version
v2 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 2

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents