https://orcid.org/0000-0002-1254-7738
ABSTRACT
The recent interstellar detection of individual polycyclic aromatic hydrocarbons (PAHs) in the Taurus Molecular Cloud (TMC-1) brings with it interest in related species that could be present in this astronomical environment. The interstellar PAHs detected in TMC-1 consist of a few pure PAHs while the majority that have been detected are their cyano-derivative counterparts due to their larger dipole moment components. Bowl-shaped PAHs, such as sumanene (C|$_{21}$|H|$_{12}$|), represent another important target for radio astronomy as they are very polar species, in spite of their high symmetry, increasing their chances of detection. Here, we present the laboratory rotational spectroscopic study of the PAH sumanene, characterized in the gas-phase using a chirped-pulse Fourier-transform microwave spectrometer operating between 2 and 8 GHz. Accurate spectroscopic parameters are derived from the spectral analysis and compared to those obtained for corannulene. These parameters have been employed to achieve reliable frequency predictions for their astronomical search in TMC-1. We do not detect either sumanene or corannulene in our QUIJOTE line survey of TMC-1 but upper limits to their abundance in this source are derived.
1 INTRODUCTION
Polycyclic aromatic hydrocarbons (PAHs) are considered the most widespread compounds in the Universe based on the intensity and omnipresence of the unidentified infrared (UIR) bands, of which PAHs are the likely carriers (Tielens 2008). The presence and abundance of PAHs in space is strongly supported by IR observations, which suggest that |$\sim$|10–25 percent of all carbon in the interstellar medium (ISM) may be incorporated into PAHs (Allamandola, Tielens & Barker 1985; Léger & Puget 1985; Tielens 2008). However, no individual PAHs have been identified in the ISM via the UIR features, despite the IR observations with unprecedented spatial resolution provided by the JWST (Chown et al. 2024). To date, a total of nine particular PAHs have been discovered in the ISM, all of them in the Taurus Molecular Cloud (TMC-1), through radio-astronomical observations. Radio-astronomy allows for the distinguishment and identification of molecules with no null electric dipole moment, since each one has a distinct rotational spectrum with narrow emission lines. In this manner, sensitive line surveys of TMC-1 have allowed the unequivocal identifications of the PAHs indene (Burkhardt et al. 2021; Cernicharo et al. 2021a), 2-cyanoindene (Sita et al. 2022), 1- and 2-cyanonaphthalene (McGuire et al. 2021), 1- and 5-cyanoacenaphthylene (Cernicharo et al. 2024), and 1-, 2-, and 4-cyanopyrene (Wenzel et al. 2024a, b). Nevertheless, even with the detections of individual PAHs there are still unknown aspects about the formation and evolution of these interstellar molecules.
Two opposite pathways have been proposed for PAH production: the ‘top-down’ and the ‘bottom-up’ scenarios. The first one suggests that very large multiringed PAHs, like fullerenes, are formed in the hot and dense outflows of carbon-rich evolved stars (Tielens 2008; Martínez et et. 2019). Subsequent processing of these species in diffuse clouds produces smaller species in high abundances. In the ‘bottom-up’ scenario, PAHs are built up from small carbon chains in the cold and shielded environments of dark clouds. Recent observational evidences indicate that a ‘bottom-up’ approach is compatible with the identifications and abundances of the PAHs detected in TMC-1. It seems that the PAH indene and the cyano derivatives of naphthalene can be formed with just a few reactions starting from neutral and cationic molecules also detected in this cloud, like c-C|$_5$|H|$_6$| (Cernicharo et al. 2021a), CH|$_2$|CHCCH (Cernicharo et al. 2021b), H|$_2$|CCCH (Agúndez et al. 2021, 2022), c-C|$_5$|H|$_5$|CCH|$_2$| (Cernicharo et al. 2022), CH|$_2$|CCHC|$_4$|H (Fuentetaja et al. 2022), and l-C|$_3$|H|$_3$||$^+$| (Silva et al. 2023). However, it is unlikely that the PAHs observed in TMC-1 arise from a reservoir of PAHs existing since the early stages of the cloud because these relatively small PAHs would not have survived the diffuse cloud stage. Nevertheless, the detailed chemical routes for the growth of PAHs are beginning to emerge (Kaiser & Hansen 2021), and they are still missing in the chemical networks. Consequently, it is highly important to detect more PAHs in space in order to provide additional clues as to their chemistry.
From the nine PAHs discovered in TMC-1, only one of them is an unsubstituted (‘pure’) PAH, indene, while the other eight species are CN-functionalized PAHs. This is because PAHs are often highly symmetric or, when they are not symmetric, they are weakly polar at best. This fact makes them notoriously difficult – if not impossible – to detect by their rotational fingerprints. In contrast, replacing even a single hydrogen on a pure PAH with a polar functional group, such as the -CN unit, yields a surrogate with a largely increased dipole moment with respect to the PAH counterpart, making it much easier to detect. There are, however, some highly symmetric pure PAHs that present non-negligible electric dipole moments, the bucky-bowls. These molecules are PAHs that possess bowl-shaped round |$\pi$| systems (Amaya & Hirao 2011) arising from strain induced pyramidalization of the carbon atoms in the central five or six-membered rings. These bowl-shaped geometries confer on these molecules large dipole moments, making them potential candidates to be detected in the ISM (Lovas et al. 2005).
The smallest bucky-bowls are named corannulene (C|$_{20}$|H|$_{10}$|) and sumanene (C|$_{21}$|H|$_{12}$|), see Fig. 1. Both are fragments of the fullerene C|$_{60}$| that retain the C|$_{5v}$| and C|$_{3v}$| symmetries, respectively. Corannulene has been extensively investigated since its first synthesis in 1966 (Barth & Lawton 1966, 1971), including a gas-phase electron diffraction study (Hedberg et al. 2000), infrared, raman, and UV/Vis spectroscopy studies (Rouillé et al. 2008) and rotational spectroscopy investigations (Lovas et al. 2005; Pérez et al. 2017). In contrast, the synthesis of sumanene (Sakurai, Daiko & Hirao 2003) was achieved later than that for corannulene, and the studies on this molecule mostly focus on its electronic properties (Amaya et al. 2009) and promising applications as functional materials for endeavours such as hydrogen storage (Della & Suresh 2018). Sumanene spectroscopy, on the contrary, has barely been explored. To the best of our knowledge, only two spectroscopic studies on sumanene have been reported. Kunishige et al. (2013) recorded the fluorescence excitation spectrum of jet-cooled gas-phase sumanene and very recently Weber et al. (2022) investigated the infrared, electronic fluorescence excitation, and dispersed fluorescence spectra of sumanene isolated in |$para$|-H|$_2$|.
A: Sumanene, C|$_{21}$|H|$_{12}$|. The top figure shows sumanene from a top view while the bottom figure shows it from a side-view. B: Corannulene, C|$_{20}$|H|$_{10}$|. Top and side-view of corannulene. Both structures were calculated at B3LYP/cc-pVTZ level of theory with empirical dispersion correction.
In this work, we report the first microwave spectroscopy investigation of sumanene in order to provide precise data to allow for its astronomical detection. We have observed the rotational spectrum of this large PAH in the gas-phase using high-resolution chirped-pulse Fourier transform microwave spectroscopy in the 2–8 GHz frequency region. Our data analysis allowed us to determine the molecular parameters, which have been used to obtain reliable predictions. Based on these predictions we have searched for sumanene in the starless molecular cloud TMC-1, which is known to harbor all the individual PAHs detected so far in the ISM.
2 QUANTUM CHEMICAL CALCULATIONS
Quantum chemical calculations were performed using the gaussian16 programme package (Frisch et al. 2016) to optimize the molecular structure of sumanene and obtain the values of its molecular constants and electric dipole moment components. Geometry optimization calculations were carried out using the B3LYP (Becke 1993) hybrid density functional with the polarised valence triple-|$\zeta$| basis set (cc-pVTZ; Woon & Dunning 1993). For the optimized geometries, harmonic frequencies were computed at the same level of theory to estimate the values of the quartic centrifugal distortion constants. The results from these calculations are shown in Table 1.
Spectroscopic parameters of sumanene.
| Parameter | Experimental | Theoretical |
|---|---|---|
| B /MHz | 461.671909(92) a | 463.75 |
| C /MHz | [246.74] b | 246.74 |
| |$D_J$| /kHz | [0.0038] | 0.0038 |
| |$D_{JK}$| /kHz | [−0.0057] | −0.0057 |
| |$|\mu |$| /D | – | 2.3 |
| |$N_{\mathrm{ lines}}$| | 5 | – |
| |$\mathrm{ rms}$| /kHz | 2.5 | – |
| Parameter | Experimental | Theoretical |
|---|---|---|
| B /MHz | 461.671909(92) a | 463.75 |
| C /MHz | [246.74] b | 246.74 |
| |$D_J$| /kHz | [0.0038] | 0.0038 |
| |$D_{JK}$| /kHz | [−0.0057] | −0.0057 |
| |$|\mu |$| /D | – | 2.3 |
| |$N_{\mathrm{ lines}}$| | 5 | – |
| |$\mathrm{ rms}$| /kHz | 2.5 | – |
Numbers in parentheses are 1σ uncertainties in units of the last digits.
Values in brackets have been kept fixed to those obtained by quantum chemical calculations.
Spectroscopic parameters of sumanene.
| Parameter | Experimental | Theoretical |
|---|---|---|
| B /MHz | 461.671909(92) a | 463.75 |
| C /MHz | [246.74] b | 246.74 |
| |$D_J$| /kHz | [0.0038] | 0.0038 |
| |$D_{JK}$| /kHz | [−0.0057] | −0.0057 |
| |$|\mu |$| /D | – | 2.3 |
| |$N_{\mathrm{ lines}}$| | 5 | – |
| |$\mathrm{ rms}$| /kHz | 2.5 | – |
| Parameter | Experimental | Theoretical |
|---|---|---|
| B /MHz | 461.671909(92) a | 463.75 |
| C /MHz | [246.74] b | 246.74 |
| |$D_J$| /kHz | [0.0038] | 0.0038 |
| |$D_{JK}$| /kHz | [−0.0057] | −0.0057 |
| |$|\mu |$| /D | – | 2.3 |
| |$N_{\mathrm{ lines}}$| | 5 | – |
| |$\mathrm{ rms}$| /kHz | 2.5 | – |
Numbers in parentheses are 1σ uncertainties in units of the last digits.
Values in brackets have been kept fixed to those obtained by quantum chemical calculations.
3 EXPERIMENTAL
The broad-band rotational spectrum of sumanene was observed between 2–8 GHz using a chirped-pulse Fourier transform microwave (CP-FTMW) spectrometer located at the University of Valladolid that has followed previous designs (Brown et al. 2008; Pérez et al. 2013). The instrument was purchased from Brightspec Inc., and it has been upgraded to include a 300 W traveling wave tube amplifier from Applied Systems Engineering which operates over the 2–8 GHz regime (Cabezas et al. 2024).
The sample of sumanene was purchased from Tokyo Chemical Industry (TCI) Europe with a purity of |$>$|99.0 percent. Sumanene is a solid at room temperature with a melting point of |$\sim$|115|$^{\circ }$| C. In order to vapourize the sample, we employed a modified pulsed Parker valve, series 9. The valve was modified in a way such that solid samples can be placed next to the orifice where the injection into the vacuum takes place and where the sample can be heated. Several trials were made to get sufficient vapour pressure of sumanene and most of the measurements were made with a reservoir temperature of 150|$^{\circ }$| C. Neon, at three bars, were passed over the sample, seeding the inert gas with sumanene before being supersonically expanded into the chamber, at a repetition rate of 7 Hz. We used the fast frame option on the Tektronix oscilloscope with eight frames to obtain an effective repetition rate of 56 Hz. A total of 800 000 free induction decays were collected and fast Fourier transformed to yield the spectrum in Fig. 2. The spectral resolution is better than 100 kHz and given the signal-to-noise ratio observed, the frequency measurements have an estimated accuracy of 2.5 kHz.
Broad-band rotational spectrum of sumanene from 2 to 8 GHz. The black trace is the experimental spectrum (800 000 averages) and the red one is the simulated (at 2 K) spectrum based on the fitted molecular constants sumanene.
4 ROTATIONAL SPECTRA ANALYSIS
The broadband rotational spectrum of sumanene is shown in Fig. 2. Sumanene belongs to the C|$_{3v}$| point group and consequently is an oblate top with a characteristic spectrum. The predicted rotational constants B and C, and the large dipole moment of 2.3 D (Table 1) guided its identification in the spectrum. Five nearly equally spaced rotational transitions with |$\Delta$|J = +1 were observed, with J being the total quantum number of the angular momentum. The frequencies of the observed transitions for sumanene are listed in Table 2. As an oblate symmetric top, sumanene could present a K structure, with K being the quantum number representing the projection of the angular momentum along the principal axis of rotation. However, as occurred with corannulene (Lovas et al. 2005; Pérez et al. 2017), no splitting was observed. The analysis of the rotational frequencies from Table 2, using a symmetric top Hamiltonian (Pickett 1991), rendered the experimental value of the rotational constant B collected in Table 1. The number of observed transitions included in the fit did not allow for the derivation of the experimental values of the rotational constant C and the quartic centrifugal distortion constant D|$_J$|, which were kept fixed to those obtained from quantum chemical calculations. The experimental rotational constant B shows good agreement (within 0.4 per cent) with the experimental value, indicating that the calculated geometry accurately represents the structure of sumanene in the gas phase. In contrast to that observed in corannulene (Pérez et al. 2017), the weak intensity of the sumanene spectrum hindered the detection of |$^{13}$|C isotopologues in natural abundance. This can be attributed to the structural differences between the two bowl-shaped molecules. Sumanene has three cyclopentane rings, which are believed to make it more rigid with a higher bowl-to-bowl inversion energy barrier, a higher bowl depth (1.11 Å compared to 0.87 Å for corannulene) and a higher strain energy (Priyakumar & Sastry 2001). Although the melting point of sumenene is lower than that for corannulene, the mentioned structural differences between the two molecules may be the reason for the different vapour pressures displayed by the two chemical species. This will explain the difficulties to get better signal-to-noise ratio spectrum for sumanene. The small value of the calculated quartic centrifugal constant D|$_J$| indicate that sumanene is a very rigid molecule. Due to the fact that the B rotational constant for sumanene is determined very accurately, the frequency predictions derived from our data can serve as a good basis for the astronomical search of this species in the Q-band (30–50 GHz) region of astronomical objects such as TMC-1.
Laboratory observed transition frequencies and residuals for sumanene.
| |$J^{\prime }$| | |$J^{\prime \prime }$| | |$\nu _{\mathrm{ obs}}$| (MHz) | Obs-Calc (MHz) |
|---|---|---|---|
| 4 | 3 | 3693.370 | –0.004 |
| 5 | 4 | 4616.719 | 0.001 |
| 6 | 5 | 5540.062 | 0.002 |
| 7 | 6 | 6463.404 | 0.002 |
| 8 | 7 | 7386.741 | –0.002 |
| |$J^{\prime }$| | |$J^{\prime \prime }$| | |$\nu _{\mathrm{ obs}}$| (MHz) | Obs-Calc (MHz) |
|---|---|---|---|
| 4 | 3 | 3693.370 | –0.004 |
| 5 | 4 | 4616.719 | 0.001 |
| 6 | 5 | 5540.062 | 0.002 |
| 7 | 6 | 6463.404 | 0.002 |
| 8 | 7 | 7386.741 | –0.002 |
Laboratory observed transition frequencies and residuals for sumanene.
| |$J^{\prime }$| | |$J^{\prime \prime }$| | |$\nu _{\mathrm{ obs}}$| (MHz) | Obs-Calc (MHz) |
|---|---|---|---|
| 4 | 3 | 3693.370 | –0.004 |
| 5 | 4 | 4616.719 | 0.001 |
| 6 | 5 | 5540.062 | 0.002 |
| 7 | 6 | 6463.404 | 0.002 |
| 8 | 7 | 7386.741 | –0.002 |
| |$J^{\prime }$| | |$J^{\prime \prime }$| | |$\nu _{\mathrm{ obs}}$| (MHz) | Obs-Calc (MHz) |
|---|---|---|---|
| 4 | 3 | 3693.370 | –0.004 |
| 5 | 4 | 4616.719 | 0.001 |
| 6 | 5 | 5540.062 | 0.002 |
| 7 | 6 | 6463.404 | 0.002 |
| 8 | 7 | 7386.741 | –0.002 |
5 ASTRONOMICAL SEARCH
The astronomical observations presented in this work are from the ongoing Yebes 40m Q-band line survey of TMC-1, QUIJOTE 1 line survey (Cernicharo et al. 2021c). A detailed description of the line survey and the data-analysis procedure are provided in Cernicharo et al. (2022). Briefly, QUIJOTE consists of a line survey in the Q band (31.0–50.3 GHz) at the position of the cyanopolyyne peak of TMC-1 (|$\alpha _{J2000}=4^{\rm h} 41^{\rm m} 41.9^{\rm s}$| and |$\delta _{J2000}=+25^\circ 41^{\prime } 27.0^{\prime \prime }$|). This survey was carried out using a receiver built within the Nanocosmos project2 consisting of two cooled high-electron-mobility-transistor amplifiers covering the Q band with horizontal and vertical polarization. Fast Fourier transform spectrometers with |$8\times 2.5$| GHz and a spectral resolution of 38.15 kHz provide the whole coverage of the Q band in both polarizations. Receiver temperatures are 16 K at 32 GHz and 30 K at 50 GHz. The experimental setup is described in detail by Tercero et al. (2021).
All observations were performed using frequency-switching observing mode with a frequency throw of 10 and 8 MHz. The total observing time on the source for the data taken with frequency throws of 10 MHz and 8 MHz is 772.6 and 736.6 h, respectively. Hence, the total observing time on source is 1509.2 h. The QUIJOTE sensitivity varies between 0.08 mK at 32 GHz and 0.2 mK at 49.5 GHz. The main beam efficiency can be given across the Q band as |$B_{\rm eff}$| = 0.797 exp[|$-(\nu$|(GHz)/71.1)|$^2$|]. The forward telescope efficiency is 0.97. The telescope beam size at half power intensity is 54.4 arcsec at 32.4 GHz and 36.4 arcsec at 48.4 GHz. The absolute calibration uncertainty is 10 |${{\ \rm per\ cent}}$|. The data were analysed with the gildas package.3
The frequency predictions for sumanene were implemented in the madex code (Cernicharo 2012) to compute synthetic spectra assuming local thermodynamic equilibrium. We used the dipole moment component from Table 1 and assumed a rotational temperature of 9 K and a |$\nu _{\rm LSR}$| = 5.83 km s|$^{-1}$| (Cernicharo et al. 2020). Even though a total of 21 lines for sumanene are covered by our QUIJOTE line survey, we focused on the lower frequency transitions with J values from 34 to 42, predicted in the frequency range from 31 to 39 GHz. Fig. 3 shows the spectrum of TMC-1 at the frequencies of the mentioned transitions for sumanene, together with the synthetic spectra calculated with madex. There are some signals observed at the predicted frequencies for three transitions of sumanene with J = 35-34, 36-35, and 39-38. The line coincident with the J = 35-34 correspond to a rotational transition of 1-cyanonaphthalene, while the other two are unidentified lines. However, no signals are clearly seen in our TMC-1 data for the other predicted lines of sumanene. The sensitivity of the QUIJOTE line survey varies between 0.08 and 0.2 mK in the 31–50 GHz domain. Adopting the observed 3|$\sigma$| limits to the intensity of strongest components of sumanene, we derive a 3|$\sigma$| upper limit to the column density of this species in TMC-1 of 1.0|$\times 10^{10}$| cm|$^{-2}$|.
Predicted rotational transitions of sumanene in TMC-1 in the 31–39 GHz range. The abscissa corresponds to the rest frequency, assuming a local standard of rest velocity of 5.83 km s|$^{-1}$|. The ordinate is antenna temperature in millikelvins. Curves shown in red are the computed synthetic spectra for N = 1.0|$\times 10^{11}$| cm|$^{-2}$|.
Despite the fact that some astronomical searches of corannulene have been reported (Thaddeus 2006; Pilleri et al. 2009; Klemperer 2011), we have searched for this molecule also in our QUIJOTE data. Like with sumanene, we focused in the lower frequency transitions of corannulene predicted between 32 and 42 GHz. Using the predicted frequencies derived from the spectroscopic data and the dipole moment value from Lovas et al. (2005) we searched for this molecule, but no lines attributable for corannulene were seen in our survey. Adopting the observed 3|$\sigma$| limits to the intensity of strongest components of corannulene, we derived a 3|$\sigma$| upper limit to the column density of this species in TMC-1 of 1.0|$\times 10^{10}$| cm|$^{-2}$|, similar to that of sumanene.
Although there are no obvious rotational features of either of these two PAHs in our TMC-1 data, it is not ruled out that molecular species of this size may be detected in the near future, given the increases in sensitivity achieved in the current radio astronomical observations. It must be noted that searching for these molecules and other species with similar molecular size using statistical techniques requires a detailed analysis of each line participating in the stacking of data. As shown in Fig. 3, there are some fortuitous feature coincidences at intensities below 0.5 mK that could bias the final signal when a stacking procedure is employed. Future data from QUIJOTE project will allow us to search for these particularly large PAHs using a line-by-line detection technique.
6 CONCLUSIONS
In this work, we have investigated the pure rotational spectra of the bucky-bowl sumanene in the 2 to 8 GHz frequency range using a high-resolution broadband microwave spectrometer. Through the analysis of the rotational spectrum, accurate rotational parameters have been determined for sumanene, enabling us to make frequency predictions within the Q-band range. We have searched for sumanene and corannulene in our QUIJOTE line survey. Although they have not been detected towards TMC-1, upper limits have been established for both species.
ACKNOWLEDGEMENTS
CC and JC acknowledge thank Ministerio de Ciencia e Innovación through the PID2019-107115GB-C21 and PID2019-106110GB-I00. CC and JC thank the Consejo Superior de Investigaciones Científicas for funding through project PIE 202250I097. ALS acknowledges grant RYC2022-037922-I financed by MCIU/AEI/10.13039/501100011033 and by the FSE+. CP thanks Spanish Ministerio de Universidades for the BG20/00160 Beatriz Galindo Senior Researcher at the University of Valladolid, and the ERC for the CoG HydroChiral (grant agreement no. 101124939). CP and FSH acknowledge funding from the Spanish Ministerio de Ciencia e Innovación and the European Regional Development Fund (MICINN-ERDF, grant no. PID2021-125015NBI00). JRM and IP thank the Spanish Ministerio de Ciencia e Innovación for funding support through the project PID2020-117925GA-I00. JRM and FSH thank the University of Valladolid for a predoctoral contract. The authors thank the Junta de Castilla y León (INFRARED-FEDER IR2020-1-UVa02) for research funds.
DATA AVAILABILITY
The data underlying this article will be shared on reasonable request to the corresponding authors.
Footnotes
Q-band Ultrasensitive Inspection Journey to the Obscure TMC-1 Environment.
