EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 662 (June 2022) A&A, 662 (2022) A111 Full HTML
Press Release
Open Access
Issue
A&A
Volume 662, June 2022
Article Number A111
Number of page(s) 17
Section Atomic, molecular, and nuclear data
DOI https://doi.org/10.1051/0004-6361/202243571
Published online 28 June 2022

© O. Zingsheim et al. 2022

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe-to-Open model. Subscribe to A&A to support open access publication.

1 Introduction

To date, more than 260 molecules have been detected in space1. Many rotational molecular spectra are collected in databases such as the Cologne Database for Molecular Spectroscopy (CDMS; Müller et al. 2001, 2005; Endres et al. 2016), allowing astronomers to search for their spectroscopic signatures in space. Several complex organic molecules (COMs) have been found thus far, for instance, in the warm gas around protostars in star forming regions, such as formamide (Rubin et al. 1971), ethyl cyanide (Johnson et al. 1977), and ethanediol (Hollis et al. 2002). One prominent example is the Protostellar Interferometric Line Survey (PILS; Jørgensen et al. 2016) with the Atacama Large Millimeter/submillimeter Array (ALMA). Among the COMs, alcohols have also been found in space. So far, methanol (CH3OH; Ball et al. 1970) and ethanol (C2H5OH; Zuckerman et al. 1975; Pearson et al. 1997) have been detected. Ethanol occurs as two conformers, so-called anti-ethanol (Zuckerman et al. 1975) and gauche-ethanol (Pearson et al. 1997), and both have been referenced based on their first detection. The next heavier alcohol, propanol (C3H7OH), is thus another candidate for a detection in space, particularly in environments with large amounts of ethanol. The possible detection of propanol may be further supported by the detection of its isomer, ethyl methyl ether (or methoxyethane, CH3CH2OCH3) in space (Tercero et al. 2015, 2018; Manigand et al. 2020).

Propanol itself occurs in two isomers, as a primary alcohol n-propanol (normal-propanol, propan-1-ol, normal propyl alcohol; CH3CH2CH2OH), where the hydroxyl group (-OH) is attached to the outer carbon atom, and as a secondary alcohol i-propanol (iso-propanol, propan-2-ol, or iso-propyl alcohol; CH3CH(OH)CH3), where the hydroxyl group is attached to the central carbon atom. n-Propanol, the subject of this study, occurs in five different conformers: Ga, Gg, Gg', Aa, and Ag, see Abdurakhmanov et al. (1968, 1970, 1976); Dreizler & Scappini (1981); Lotta et al. (1984); Maeda et al. (2006); Kisiel et al. (2010) and references therein. The capital letters G and A refer to the rotation of the heavy nuclei plane C-C-C compared to C-C-O. An “anti” (A) configuration2 describes a rotation by 180° and "gauche" (G) by about 60°. The small letters refer to the rotation of the dihedral angle of the hydroxy group (-OH).

The G family of conformers (Ga, Gg, and Gg') of n-propanol has been rather extensively studied in rotational spectroscopy, as in, for instance, Abdurakhmanov et al. (1970); Maeda et al. (2006); Kisiel et al. (2010) and references therein. A single state analysis of Ga was able to fit 2865 lines up to 375 GHz, but systematic deviations occurred for some transitions with higher J quantum numbers (Maeda et al. 2006). Based on this observation, a combined analysis of the three G conformers was carried out by taking into account Coriolis interaction and Fermi resonances (Kisiel et al. 2010). Thanks to the quantum mechanical models of rotational spectra with experimental accuracy (Kisiel et al. 2010), conformers of the G family can be found in space. Furthermore, this distinctive analysis gives confidence that the G and A conformers can be treated separately as within the G family, the conformers seem to be unaffected by conformers of the A family (Aa and Ag).

The ab initio calculations have shown some scatter for the relative energies of n-propanol conformers, as, for instance, in Lotta et al. (1984); Kahn & Bruice (2005) and references therein (a detailed overview is given in Maeda et al. 2006, but they agree in the fact that all conformers, in particular from both G and A families, are rather close in energy)3. Considering only Boltz-mann statistics, the Aa and Ag conformers are expected to have about the same chance of being detected in the warm interstellar medium as the G family conformers.

The most advanced rotational analysis of the Aa n-propanol was derived by using a Fourier transform microwave (FTMW) spectrometer (8-18 GHz) and a Stark modulated microwave spectrometer (up to 40 GHz) (Dreizler & Scappini 1981). Four conformers of n-propanol occur as mirror configurations (twofold degeneracy due to "left-" and "right-handed" versions, C1 symmetry), except for the Aa conformer, which has a plane of symmetry and one-fold degeneracy (Cs symmetry), as shown in Fig. 1. The Ag conformer occurs in mirror images (see Fig. 1), however, the barrier for hydrogen tunneling between the two equivalent Ag forms is low enough to distinguish symmetric Ag+ and antisymmetric Ag− tunneling states (Abdurakhmanov et al. 1976; Kisiel et al. 2010). The tunneling transition state has Cs symmetry and tunneling-rotation interactions need to be taken into account, as has been done, for instance, for gauche-ethanol (Pearson et al. 2008) and gauche-propanal more recently (Pickett & Scroggin 1974; Zingsheim et al. 2017, 2022). So far, only eleven doublet features up to 30 GHz were tentatively assigned to Ag+ and Ag− tunneling states with Jmax = 4 (Abdurakhmanov et al. 1976).

The main aim of the present study is to facilitate the astronomical detection of Aa n-propanol and Ag n-propanol. For this purpose, the rotational spectra are measured in selected frequency ranges from 18 to 505 GHz. The employed experimental setups are presented in Sect. 2. The experimental data are analyzed in Sect. 3 with a focus on secure assignments of the perturbed systems. Next, the spectroscopic results and the astronomical search of n-propanol are discussed in Sect. 4. Finally, our main conclusions are given in Sect. 5.

thumbnail Fig. 1

Aa and Ag conformers of n-propanol. Hydrogen, carbon, and oxygen atoms are depicted with white, black, and red spheres, respectively. The atoms C-C-C-O-H span a (symmetry) plane for the Aa conformer (Cs symmetry) and it is shown as face-on and side view for clarification. Then, Ag n-propanol occurs in two mirror-image forms leading to the Ag+ and Ag− tunneling states. The G conformers are not depicted here, but the oxygen is out of the heavy nuclei plane for all of them.

2 Experimental techniques

We recorded rotational spectra of n-propanol (i) in the 17.5-26.5 GHz region using chirped-pulse Fourier transform microwave (CP-FTMW) jet spectroscopy and (ii) in the regions of 33-67 GHz, 70-129 GHz, and 160-505 GHz using conventional 2f demodulation absorption techniques. Additional (iii) double-modulation double-resonance (DM-DR) measurements support assignments of weaker transitions to the Aa conformer and, in particular, secure the identification of various qR series4 of the Ag conformer. The (i) CP-FTMW, (ii) conventional (sub-)millimeter wave (MMW) absorption spectrometers, and (iii) the DM-DR setup are introduced in detail in (i) Hermanns et al. (2019), (ii) Ordu et al. (2019); Martin-Drumel et al. (2015), and (iii) Zingsheim et al. (2021), respectively. Thus, their main characteristics are only briefly summarized in the following.

In terms of (i) the 17.5-26.5 GHz range was covered in steps of 1 GHz bandwidth using a CP-FTMW spectrometer. In contrast to the above-described setup (Hermanns et al. 2019), mixing up the CP and mixing down the molecular signal is no longer required as the CP is generated directly in the target frequency region using an arbitrary waveform generator (Keysight M8195A 65 GSa s−1). In the present setup, each 1 GHz broad CP spanned 2 μs and was amplified with a 4 W amplifier. These amplified pulses were then broadcast via a horn antenna to excite the molecular ensemble. After a pulse was stopped, the resulting free induction decay (FID) of the molecules was collected with another horn antenna, amplified using a low noise amplifier, and recorded for 10 μs on a 100 GS s−1 oscilloscope. Each measurement consists of 65 000-100000 acquisitions, which were performed with a 10 Hz repetition rate. The averaged FIDs were then Fourier transformed using a Hanning window, which resulted in a frequency resolution of about 290kHz, considering the full width at half maximum (FWHM). The sample was prepared by mixing 1 mL of liquid n-propanol (Sigma-Aldrich, 97.7% purity) with 7 bar neon in a 10 L container. The mixture was supersonically expanded with a relative stagnation pressure of 1 bar and an opening duration of 500 μs.

For (ii) either an Agilent E8257D or a Rohde & Schwarz SMF100A synthesizer generated electromagnetic waves. Electromagnetic waves were directly coupled to a horn antenna to record single spectra in the 33-67 GHz region. We used in-house developed electronics for operating an amplifier tripler chain in full saturation mode to fully cover the 70-129 GHz region and used commercially available cascaded doublers and triplers from Virginia Diodes Inc. (VDI) to fully cover the 160-505 GHz region. Frequency modulation (FM) increases the signal-to-noise ratio (S/N) and the 2f demodulation creates absorption features that are close to the second derivative of a Voigt profile. We used a modulation frequency of f ≈ 47 kHz and its amplitude is on the order of the FWHM of the absorption profiles. The absorption path was 14 m in a single pass configuration for measurements up to 130 GHz (Ordu et al. 2019) and 10 m in double pass configuration otherwise (Martin-Drumel et al. 2015). Here, n-propanol was vaporized into the cell to pressures in the range of 10-40 μbar.

For (iii) we used both synthesizers, namely, Agilent as the so-called probe and Rohde & Schwarz as the pump, simultaneously for DM-DR measurements; a modulated pump radiation was applied in addition to the frequency modulated probe radiation. The probe signal was usually frequency tripled, and in-house developed electronics used for operation in full saturation mode ensured a stable output power of about 1 mW (0dBm). The pump source, an active frequency multiplier (AFM6 70-110 + 14 from RPG Radiometer Physics), delivered output powers of up to 60mW (17.8 dBm). The final pump and probe frequencies were both located in the W-band. We performed some additional measurements with the probe frequency being around 200 GHz using commercial components by VDI. It is only the probe transitions that share an energy level with the pump transition that are affected by the additional source. An additional 1 f' demodulation of the pumped signal is experimentally realizing a difference spectrum of the on- and off-resonant measurements, or, in other words, a subtraction of a conventional and a DR measurement. In this way, it is only the lines of interest remaining and the assignment of even weak or blended features is facilitated (Zingsheim et al. 2021).

3 Results: Spectroscopic signatures

The measurement results for Aa and Ag n-propanol are presented in this section. The central focus of this chapter is the analyses and the resulting quantum mechanical models of rotational spectra, as well as the verification of assignments, since both conformers show perturbed patterns. Our assignment procedure benefited, in terms of speed and unambiguity, from the LLWP software, which is based on Loomis-Wood plots5 (Bonah et al. 2022). Adjacent transitions of certain series are plotted around their predicted frequency in Loomis-Wood plots to increase the confidence of correct assignments or to visualize systematic deviations between observed and predicted transition frequencies.

3.1 Aa n-propanol

The Aa conformer is an asymmetric rotor quite close to the prolate limit with κ = -0.978 (κ = [2B - A - C]/[A - C]). The dipole moments were determined with Stark effect measurements to be μα = 0.21(7) D and μb = 1.54(2) D (Abdurakhmanov et al. 1970). In total, 73 lines (118 transitions) were assigned and fit up to 40 GHz in a previous study (Dreizler & Scappini 1981). Therein, all lines from Abdurakhmanov et al. (1968) were remeasured; α- and b-type transitions were assigned and fit and in particular the assignments of weak α-type transitions were secured with the help of MW-MW DR measurements (Dreizler & Scappini 1981). A splitting due to the internal rotation of the methyl group (-CH3) could sometimes be resolved (14 A + 14 E transitions) and together with 45 blended doublet features, the transition frequencies were reproduced with a standard deviation of 195 kHz (Dreizler & Scappini 1981). Thereby, the potential barrier of the methyl group internal rotation was determined to be V3 = 955(21)cm−1 (or 2730(60)cal mol−1) and the angle of the internal rotation axis i to the principal moment of inertia axis α to be <(i, a) = (29 ± 1)° (Dreizler & Scappini 1981).

We used the available parameters and transition frequencies (Dreizler & Scappini 1981) to set up an initial spectroscopic model of an asymmetric rotor, including three-fold internal rotation in ERHAM (Groner 1992, 1997, 2012). Thanks to the initial prediction and with the application of an iterative assignment and fit strategy, we could assign strong and also weak b-type transitions in a straightforward fashion (see Fig. 2). Transitions from Dreizler & Scappini (1981) were remeasured if possible6. Furthermore, we secured assignments of weak α-type transitions in the W-band region with the DM-DR technique. Additionally, transitions of the rR series with J up to 13) have been unambiguously assigned above 200 GHz with DM-DR measurements (see Fig. 3). The assignment of internal rotation components, A and E, is unambiguous as we observed E* transitions, which are forbidden in rigid rotors (Herschbach & Swalen 1958). However, the rR series transitions appeared to be noticeably shifted from their prediction and not all of them are incorporated in the final fit. Including these transitions is possible if others are excluded instead. The final choice of fit transitions is somewhat ambiguous: the quantum number coverage is visualized in Fig. C.1. Larger systematic deviations, Δν = vObs- vCalc, occur from our final predictions for various series (cf. Fig. 2). These deviations start at similar J values for series involving identical Ka values, see Table 1. However, the assignment of these transitions is unambiguous as demonstrated in Fig. 2 and the assignment of internal rotation components is also secured (cf. Fig. 3). Finally, we could assign 1255 transitions with frequencies up to 505 GHz and Jmax = 70, thereof, 928 transitions are well reproduced by the prediction. The determined spectroscopic parameters can be found in Table 2.

thumbnail Fig. 2

Loomis-Wood plot of Aa n-propanol - screenshot of the LLWP software (Bonah et al. 2022). Shown here is a doublet series (A and E components of the methyl group internal rotation) of the J3,J-3 ← J2,J-2 b-type transitions from J = 6 (bottom panel) to J = 36 (top panel). Predictions of the A components are set as center frequencies and are depicted by green sticks. The E components are depicted by red ones. This rQ series with Ka + Kc = J and ΔΚa, ΔKc = +1,-1 is well reproduced by our spectroscopic model described in Table 2 up to J = 22, however, transitions up to J = 41 (Δν ≈ 90 MHz) are straightforwardly assignable.
thumbnail Fig. 3

DM-DR measurements of Aa n-propanol of the transition around 232.55 GHz. The lower panels show two DM-DR measurements where the A and E components of the transition are used as pump frequencies at 113 811.6420 MHz and 113 810.5391 MHz, respectively. A conventional measurement is shown in the upper panel. E* transitions, which are forbidden in rigid rotors (Herschbach & Swalen 1958), secure the assignment of internal rotation components. We note that slightly off-resonant pumping leads to small, asymmetric features of A and E components in the DM-DR spectrum of the other one, as pumped A and E components are rather close in frequency, cf. Zingsheim et al. (2021).

3.2 Ag n-propanol

The three members of the G family were extensively described in the literature plus the Aa conformer to some extent. The situation is different for the fifth conformer of n-propanol: Ag. There is no spectroscopic model of rotational spectra with experimental accuracy available, and only 22 qR series transitions (1, 3, 5, and 2 transitions for J = 1,2, 3, and 4, respectively, for each of the two tunneling states), were tentatively assigned in the MW region (Abdurakhmanov et al. 1976). Therein, doublet features were observed and assigned to the tunneling states Ag+ and Ag−. The average B + C was determined to be about 7.3 GHz (Abdurakhmanov et al. 1976). A repeating signature with this spacing was also observed in the MMW region (see Fig. 2 of Kisiel et al. 2010) and can be attributed to Ka structures with increasing J quantum numbers of the Ag conformer. However, to our knowledge no spectroscopic model with experimental accuracy nor assignments are publicly available for that latter work. Namely, Ag n-propanol is a perturbed system that cannot be analyzed easily. The focus of this work is on providing initial assignments of qR series transitions, as the dipole moments are calculated to be μa = 1.15 D, μb = 0.39 D, and μc = 1.15 D7, but the energy difference of Ag+ and Ag− tunneling states is unknown so far. Then, Ag n-propanol is expected to be close to the prolate limit with κ = -0.98 (Kisiel et al. 2010).

We performed CP-FTMW spectroscopy in particular to check the assignments of a-type transitions from Abdurakhmanov et al. (1976). We measured not only the aforementioned transitions of Ag n-propanol, but also of gauche-propanal as its resulting tunneling spectroscopic pattern is already well-understood (Zingsheim et al. 2022). Similar spec-troscopic signatures are expected for both molecules and by comparing both spectra; literature assignments of Ag n-propanol could not be confirmed, but the first tentative assignments of typical doublet transitions, originating in Ag+ and Ag− tunneling states, could be made (cf. Fig. B.1).

DM-DR measurements secured the linkages of various transitions that occur about every 7.3 GHz in the W-band region. Afterwards, we assigned qR series transitions in the frequency region from 36 to 505 GHz with the help of the LLWP software (Bonah et al. 2022). Some of the qR series of the Ag conformer are visualized exemplarily in a Fortrat diagram (Fig. 4). Assignments of series transitions, , are done as they start in rows of the Fortrat diagram with J = Ka + 1 (). Two patterns are found, with each consisting of a doublet series for each qR4, qR5, qR6, qR7, qR8, and qR9 series (4 ≤ Ka ≤ 9) proving the existence of 0+ and 0− tunneling states (see Fig. 4). The nomenclature 0+ and 0− is used to highlight that the vibrational ground state ν = 0 is observed. We assigned the tunneling state to transitions with 4 ≤ Ka ≤ 9 in the way that we assumed 0+ transitions to be always lower in frequency than ones of 0−, which is usually the case if identical transitions are compared (transitions with same J's, Ka's, and Kc's).

A further confirmation of unambiguous assignments is the observed asymmetry splitting for qR4, qR5, qR6, and qR7 series. The two asymmetry components, transitions with identical J and Ka but either Kc = J - Ka or Kc = J - Ka + 1, are frequently blended for low J and higher Ka values (prolate pairing), but this degeneracy is lifted for higher J (cf. colored lines in Fig. which are split into two lines at higher J). The asymmetry splitting is the frequency difference of the two α-type transitions with and . The predicted asymmetry splittings, from the ab initio calculations of Kisiel et al. (2010), show the right order of magnitude when compared to the observed ones of both tunneling states; more importantly, they show similar systematic deviations for all Ka numbers (see Fig. 5).

At this point, we tried to reproduce the determined frequencies of assigned transitions with 4 ≤ Ka ≤ 9 by applying a common asymmetric rotor Hamiltonian, which is comparable to what has been done for the Aa conformer (cf. Table 2); This time the SPFIT program was used (Pickett 1991). Many series seem to deviate from asymmetric rotor spectra. We expect identical series (same Ka and same asymmetry side) of the two tunneling states to appear as nearly parallel lines in Fig. 4, whose spacing should decrease for increasing Ka quantum numbers. This is not the case, for instance, the two series with Ka = 8 (purple lines) have quite different slopes and the tunneling splitting of series with Ka = 6 (green lines) is larger than that of the series with Ka = 5 (red lines). We tried to incorporate even order Coriolis-type parameters (Fbc, Fac, and Fab) to account for possible tunneling-rotation interaction between the two tunneling states, Ag+ and Ag-, as has been done successfully several times for similar molecular systems in the past; for instance, in Pickett (1972); Christen & Müller (2003); Pearson et al. (1997); Zingsheim et al. (2022). However, for the case of Ag ra-propanol, still no experimental accuracy for our predictions has been reached.

We continued assigning J, Ka, and Kc quantum numbers to transitions with Ka < 3. However, transitions from one of the two series, Ag+ and Ag−, are not strictly lower in frequency than the other, in contrast to series with 4 ≤ Ka ≤ 9. Crossing series are observed in Loomis-Wood plots (Figs. B.2–B.5). Therefore, the tunneling state assignments remain speculative for transitions with Ka < 3. The two tunneling series (qR series with identical Ka and Kc quantum numbers) with Ka < 3 are even less parallel than observed ones with 4 ≤ Ka ≤ 9. We expect states with Ka < 3 to be strongly influenced by tunneling-rotation interaction between Ag+ and Ag- tunneling states, cf. Fig. C.2. Finally, we could not securely assign qR3 transitions as too many candidate series are close in frequency (Fig. B.6).

In summary, we could prove the existence of the two tunneling states Ag+ and Ag− and assign lines with Jmax = 69 and Ka,max = 9. A quantum number overview of assigned transitions is visualized in Fig. C.1. Deriving a spectroscopic quantum mechanical model with experimental accuracy is beyond the scope of this work because of the multiple perturbations whose origins are unknown at present.

Table 1

Quantum number overview of the final analysis of Aa n-propanol.

thumbnail Fig. 4

Fortrat diagram of n-propanol. Top panel: a broadband measurement in the W-band region is shown, where respective regions with signatures of the Ag conformer, leading to the Fortrat diagram, are highlighted. In some regions, the spectrometer is not as sensitive, e.g. around J=11, compared to others. Bottom panel: The colored lines in the Fortrat diagram mark linkages of series transitions, . DM-DR spectroscopy secured assignments in the W-band region (from J = 10 to J = 16). The existence of two tunneling states is proven as for each Ka partner series are found; the tunneling states are marked by 0+ and 0−. More series can be found, but are not assigned yet, e.g. QR3 ones. We note the larger tunneling splitting of QR6 series (green lines), e.g., in comparison to QR5 ones, and the different slopes for QRS series (thick purple lines).
Table 2

Spectroscopic parameters of Aa n-propanol.

thumbnail Fig. 5

Asymmetry splitting of the qR4, qR5, qR6, and qR7 series transitions, (a) Exemplary Loomis-Wood plot for qR6 transitions o7 0+. Two transitions, either with Kc = J - Ka + 1 or with Kc = J - Ka, are observed and marked by green qtars in each row. The average frequency of both transitions is the center frequency (Frequeney Offset is 0 MHz). Sceeenshot of the LLWP software (Bonah et al. 2022). The frequency difference of both components is the asymmetry splitting for Ka = 6 of 0+ for a given J. (b) Οbserved and calculated asymmetry splittings for qR4 (blue), qR5 (red), qR6 (green) and qR6 (orange). Similar observed and calculated asymmetry splittings confirm the correct assignment of Ka quantum numbers. We note that small asymmetry splittings are not resolvable, both transitions are partly blended, and missing experimental values for J = 17-22 eriginate in thx experimental gap from 130-170 GHz.

4 Discussion

In this work, we present an extended spectroscopic quantum mechanical model based on additional assignments in the MMW region for Aa n-propanol and secured assignments of both tunneling states, Ag+ and Ag−, for Ag n-propanol. In the following, we discuss the spectroscopic results of the Aaand Ag conformers (Sects. 4.1 and 4.2). Subsequently, we offer some ideas to overcome certain limitations, specifically, thanks to a global analysis of both conformers (Sect. 4.3). Next, spectroscopic results of n-propanol are judged in regard of astronomical searches (Sect. 4.4). The rotational data are used in a companion article for its astronomical detection (Sect 4.5). Moreover, we showed that DM-DR measurements can straightforwardly be extended to frequencies higher than the W-band, see Fig. 3, as was demonstrated in the original DM-DR manuscript (Zingsheim et al. 2021).

4.1 Aa n-propanol

We extended spectroscopic assignments of Aa n-propanol up to 505 GHz with Jmax = 70 and Ka,max = 6. Finally, 928 transitions are well reproduced by the spectroscopic parameters in Table 2. Larger systematic deviations Δν from our predictions occur for certain series (cf. also Fig. 2); deviations start from transitions with similar J when series with identical Κa values are compared. The deviations occur at rather low Κa values and the origins of these perturbations are unknown so far. Transitions involving Κa = 0 and 1 seem to be unperturbed as b-type transitions involving Κa = 1 ↔ 0 are well reproduced in our final spectroscopic model. On the other hand, b-type transitions with Κa = 2 ↔ 1, Κa = 3 ↔ 2, and Κa = 4 ↔ 3 are only well reproduced up to around J = 30, J = 20, and J = 10, respectively (see Table 1). This case is similar to that of a-ethanol, as discussed for the 13C isotopomers (Bouchez et al. 2012).

If different states are interacting, consecutive transitions of one series are frequently perturbed. In particular, systematic deviations occur around the strongest perturbed energy levels. So-called avoided crossing patterns may be observed; in this case, transitions belonging to the same series with somewhat smaller and larger quantum numbers than heavily perturbed ones can often be assigned with confidence and, more importantly, reproduced to experimental accuracy without treating the interaction. In this way, perturbed energy levels can be located and transitions from these levels can be unweighted in the analysis if the origin of the interaction is unknown or indescribable. For instance, at the start of the analysis of gauche-propanal, certain υ = 0 series had to be unweighted (Zingsheim et al. 2017), but they could be properly described by including Fermi resonances and Coriolis interaction in a global analysis with υ24 = 1 (Zingsheim et al. 2022). In the case of Aa n-propanol, no transitions with higher J values than the ones that deviate from the prediction could be fit to experimental accuracy; the strongest perturbed energy levels and therefore the center of the perturbation, or the strongest perturbed levels and their exact energies, are unknown to this point. Therefore, we expect that our final spec-troscopic model is effective since already perturbed transitions may be included in the line list.

4.2 Ag n-propanol

We thoroughly examined assignments of Ag n-propanol in the MW and MMW regions and proved the existence of the two tunneling states Ag+ and Ag−. First, DM-DR measurements linked various qR series transitions and assignment of these series was undertaken with the support of the LLWP software. We could not confirm doublet patterns from the literature (cf. Fig. B.1; Abdurakhmanov et al. 1976), but we did find others instead (see Fig. 4). The ab initio calculations (Kisiel et al. 2010) allowed us to pinpoint patterns and, in particular, the verification of assignments by studying the asymmetry splitting, cf. Fig. 5. In this way, assignments up to Jmax = 69 and Κa,max = 9 with frequencies up to 505 GHz have been achieved. A derivation of a spectro-scopic quantum mechanical model to fit the experimental data accurately is beyond the scope of this work.

4.3 Global analysis: Aa, Ag+, and Ag− conformers

In the future, Aa n-propanol should be described in a more sophisticated spectroscopic quantum mechanical model and an initial description of Ag n-propanol with experimental accuracy should be derived. This may be possible only through a global analysis of Aa, Ag+, and Ag− conformers. The rotational energy levels of the G family seem to be isolated from A conformers ones (Kisiel et al. 2010), therefore, an additional incorporation of G conformers in a global analysis is expected to be unnecessary. A global analysis of the A family is expected to be successful only if Coriolis interaction and Fermi resonance between the Aa conformer and Ag+ or Ag− tunneling states and, very importantly, also tunneling-rotation interactions between the Ag+ and Ag- tunneling states are considered8. The first interaction terms which should be investigated are the ones used to describe the system anti, gauche+, and gauche− ethanol (Pearson et al. 2008), as the group theoretical considerations and probably also the energy splittings are similar9. A quantum mechanical treatment taking interactions into account would also allow to determine relative energies as has been done for the G family (Kisiel et al. 2010). Furthermore, the potential of the -OH group rotation may be studied in detail by taking also higher vibrationally excited states into account.

4.4 Astronomical search of n-propanol

The partition function used for creating the predictions of all five conformers, given in the supplementary material, is QGx = 513984.0861 (log Q = 5.7109) at 300 K10. This value is derived from the extensive parameter set from Kisiel et al. (2010) and takes the three G conformers and their relative energies into account. Thereby, the degeneracy factor was set to 4 (2 for the internal rotation of the -CH3 group plus 2 when taking the mirror images into account). Conformational correction factors taking into account the existence of the A conform-ers as well as vibrational correction factors, can be found in Table A.1.

Predictions of Ga, Gg, and Gg' are based on the parameter set of Kisiel et al. (2010). We note that the relative energy of Ga was set to EGa = 0 cm−1 (consequently, EGg = 47.8 cm−1 and EGg' = 50.8 cm−1), instead of using Gg as reference point, as done in the original data. This adaption is inevitable to derive meaningful intensities and lower state energies in the predictions. Predictions in Pickett's SPCAT format (*.cat files; Pickett 1991) can be found in the supplementary material and will be added to the CDMS as well.

The predictions of Aa n-propanol are limited to transitions with J up to 70, 30, 20, and 10 for Κa = 1, 2, 3, and 4, respectively. All assigned transitions with Κa ≤ 5 can be used in astronomical observations, but measured transition frequencies should replace predicted ones. Transitions with Κα > 5 should not be used in astronomical observations due to the limitations of our spectroscopic quantum mechanical model. The intensities and lower state energies in our predictions were corrected by an assumed relative energy of EAa = 30 cm−1. Overall, the number of unambiguously assigned transitions up to 505 GHz and the quantum number coverage for low Ka values is sufficient to search for rotational transitions of the Aa conformer in space.

An astronomical detection of Ag n-propanol is facilitated thanks to many unambiguously assigned transitions. For creating meaningful predictions, despite the fact that a quantum mechanical model with experimental accuracy is missing, we used, in a first step, ab initio calculations that are based on Kisiel et al. (2010) to derive theoretical predictions. Even though predicted frequencies do not match experimental accuracy at all, meaningful physical parameters, such as the intensities, lower-state energy, and degeneracies of transitions are obtained by assuming a relative energy of EAg = 45.5 cm−1. Finally, we replaced the predicted frequencies by the measured ones and neglected all unassigned transitions. In doing so, an astronomical detection of Ag n-propanol is possible, but deriving a proper spectro-scopic quantum mechanical model should be done in particular to derive relative energies and to enlarge the number of usable transitions.

4.5 Astronomical detection of n-propanol

Some members of our team searched for distinctive rotational signatures of all five conformers of n-propanol in a companion astronomical study that uses the survey called Re-exploring Molecular Complexity with ALMA (ReMoCa; details given in Belloche et al. 2019) toward Sagittarius (Sgr) B2(N) (Belloche et al. 2022). This astronomical study reports the identification of n-propanol toward the secondary hot molecular core, Sgr B2(N2), thanks to the spectroscopic predictions from this work. The detection relies on eight and five lines of the Gg' and Ag conformers, respectively. The catalogs produced in this work were instrumental in securing this first astronomical detection of n-propanol in a hot core.

5 Conclusion

We recorded spectra of n-propanol in the region of 18-505 GHz, which allowed us to analyze the rotational signatures of its Aa and Ag conformers, in addition to the already extensively studied Ga, Gg, and Gg' conformers (Kisiel et al. 2010). Therefore, rotational signatures of all five conformers of n-propanol are available at present and the main conclusions of this work are as follows:

  1. An extended quantum mechanical model of rotational spectra of Aa n-propanol has been made available.

  2. The existence of the two tunneling states, Ag+ and Ag−, is proven for the first time by unambiguously assigning transitions of Ag n-propanol.

  3. A global analysis of Aa and Ag conformers is recommended if the spectroscopic description of the Aa conformer needs to be extended or if a quantum mechanical model with experimental accuracy of the Ag conformer is the target.

  4. Predictions of the rotational spectra of all five conformers of n-propanol (*.cat files in the supplementary material) enable their astronomical search.

In addition, the spectroscopic analysis presented in this work allowed for the astronomical detection of n-propanol toward a hot core in a companion article (Belloche et al. 2022).

Acknowledgements

We would like to thank Dr. Matthias Justen and Dr. Leonid Surin for their help in translating Russian manuscripts and providing literature data from the VINITI database. B.H. acknowledges financial support by the Deutsche Forschungsgemeinschaft (DFG; project ID WE 5874/1-1). This work has been supported via Collaborative Research Centre 956, sub-projects B3 and B4, and the "Cologne Center for Terahertz Spectroscopy", both funded by the DFG (project IDs 184018867 and SCHL 341/15-1, respectively).

Appendix A Partition function Q of n-propanol

We calculated the partition function of the three G conformers QGx with SPCAT to be 513984.0861 at 300 K, which takes into account the Ga, Gg, and Gg' conformers and a degeneracy factor of 4 (a factor of 2 for the internal rotation of the -CH3 group and another factor of 2 for mirror images of the respective conform-ers; see next paragraph for more details).11 This value is based on the quantum mechanical description from Kisiel et al. (2010). The spectroscopic quantum mechanical models of Aa and Ag conformers are not as accurate as the ones of the G family and the relative energies of EAa and EAg to EGa are somewhat uncertain, thus, we used the partition function derived from the G family QGx for creating predictions (*.cat files) of all five con-formers. For the predictions of Aa and Ag conformers, the degeneracy factor is 2 and 4, respectively. Both conformers have an internal rotor (A and E transition, whereby both components are blended for the Ag conformer), but the Aa conformer does not occur in a mirror configuration as the other four conformers do. In this way, we derived correct relative intensities of all five conformers.

For predicting correct absolute intensities, which are fundamental for deriving meaningful column densities of molecules in space, the existence of both A conformers and all vibrational states need to be taken into account in order to derive QTot. Therefore, we present conformational and vibrational correction factors in Table A.1. The conformational correction factors are derived as follows: (A.1)

where the partition function is calculated with degeneracy factor 1, for instance, to be QGa (1) = 49318.3481 at 300 K. The first factor f = 2 takes the internal rotation into account and the second factor, namely, f" = 1 or 2, the conformer multiplicity of Aa and Ag, respectively. The Boltzmann factor of both conformers is derived by assuming EAa = 30.0 cm−1 and EAg = 45.5 cm−1. The vibrational correction factors were calculated using fundamental vibrational frequencies (with ν ≤ 1200 cm−1) of the Aa conformer (Fukushima & Zwolinski 1968), that is, ν30 = 131 cm−1, ν29 = 240 cm−1, ν28 = 286 cm−1, ν27 = 3 32 cm−1, ν26 = 463 cm−1, ν25 = 758 cm−1, ν24 = 8 58 cm−1, ν23 = 898 cm−1, ν22 = 971 cm−1, ν21 = 1013 cm−1, ν20 = 1047 cm−1, and ν19 = 1066 cm−1. Applying the correction factors from Table A.1 allows to derive accurate column densities of n-propanol and derive relative abundances to other molecular species in astronomical sources.

Table A.1

Partition function QGx of all three G conformers (with spin weight 4), QGa(1) of Ga (with spin weight 1), together with conformational correction factors FConf, vibrational correction factors FVib, and their product FTot for various temperatures of interest for an astronomical detection.

Appendix B Rotational signatures of Ag n-propanol

First, a part of the MW spectrum of n-propanol and propanal, depicting the a-type transitions, recorded with the CP-FTMW spectrometer, are shown in Fig. B.1. Rotational signatures are shown in the frequency regions 21.5-22.5 GHz and 25.0-26.0 GHz, respectively. Both measurements were performed under similar conditions. The rotational spectrum of gauche-propanal is well-understood (Zingsheim et al. 2022), in particular doublet structures due to two tunnelings states are observed; Two Ka = 1 transitions which are close in frequency and have similar intensities. We note that tunneling-rotation interaction mixes the wave functions of involved rotational energy levels, in this case intensity borrowing may play a role and nominally forbidden transitions can be observed (cf. transitions marked with υ = 0+ ↔ 0−). The interpretation of the rotational spectrum of Ag n-propanol is more complicated. Assigned doublet transitions for Ag n-propanol from Abdurakhmanov et al. (1976) are marked in red, but cannot be verified in our CP-FTMW jet experiment. However, two transitions with similar intensities close in frequency are found twice. Comparison of the rotational spectra of both molecules leads to tentative assignments of Ka = 1 transitions of Ag n-propanol. The appearance of two transitions close in frequency were first hints of the existence of two tunneling states (Ag+ and Ag−), similar to what is observed for gauche-propanal. The different tunneling splittings of transitions with Ka = 1 for Ag n-propanol suggest that involved rotational energy levels are perturbed due to so far untreated interactions. We note that the transition marked with an asterisk (*) is another possible candidate for a transition with Ka = 1, if qR series are tracked, cf. Fig. B.4. Unambiguously assigning these transitions can be done if a spectroscopic quantum mechanical model with experimental accuracy is derived.

Second, typical spectroscopic signatures of series transitions allow for unambiguous assignments of the Ag conformer, despite a missing quantum mechanical description. A broader Fortrat diagram than shown in Fig. 4 is presented in Fig. B.2. Then, Loomis-Wood plots are shown for series with 0 ≤ Ka ≤ 3 to further document these assignments (see Figs. B.3-B.6). Series with 4 ≤ Ka ≤ 9, which have been assigned in this work as well, are already well represented by Fig. 4 in the main article. The Loomis-Wood plots are all screenshots of the LLWP software used for assigning transitions within this work (Bonah et al. 2022).

thumbnail Fig. B.1

Microwave spectra of n-propanol and propanal. Assignments of α-type transitions with are shown for both molecules (Ka's are given in the figure). Top panel: Rotational spectrum of gauche-propanal with assignments from Zingsheim et al. (2022) marked in blue. Bottom panel: Assigned doublet transitions for Ag ra-propanol from Abdurakhmanov et al. (1976) are marked in red, but cannot be verified in our CP-FTMW jet experiment. Our tentative assignments of Ka = 1 transitions of Ag n-propanol are marked in blue. We note that the transition marked with an asterisk (*) is another possible candidate for a transition with Ka = 1, more information can be found in the text.
thumbnail Fig. B.2

Fortrat diagram of n-propanol. The colored lines in the Fortrat diagram mark linkages of series transitions of Ag n-propanol, - DM-DR spectroscopy secured assignments in the W-band region (from J = 10 to J = 16). Tentative assignments of Κa = 1 transitions derived from the CP-FTMW measurement (Fig. B.1) are marked by the blue arrows. Considering found qR1 series, another candidate transition is marked by the blue asterisk. A quantum mechanical model with experimental accuracy can solve the ambiguity of these tentative assignments. The gray area depicts the region of the Fortrat diagram in Fig. 4.
thumbnail Fig. B.3

Loomis-Wood plot around the assigned qR0 series transitions in the W-band region of the Ag conformer, which are shown by red sticks.
thumbnail Fig. B.4

Loomis-Wood plot around the assigned qR1 series transitions in the W-band region of the Ag conformer, which are shown by red sticks. Theblue asterisk (*) marks the same transition as is marked in Fig. B.1.
thumbnail Fig. B.5

Loomis-Wood plot around the assigned qR2 series transitions in the W-band region of the Ag conformer, which are shown by red sticks. The series higher in frequency with Ka + Kc = J + 1 is probably strongly interacting with a so far unknown state around J = 12.
thumbnail Fig. B.6

Loomis-Wood plot around assumed qR3 series transitions in the W-band region. No assignments with Ka = 3 are secured yet, but the existence of such can be seen by the many unassigned transitions in the shown region. Four series close in frequency are expected; Two asymmetry sides for both tunneling states.

Appendix C Overview of the rotational energy levels of Aa and Ag n-propanol

We present a quantum number overview of our analysis for Aa and Ag conformers of n-propanol here as we could not reproduce all assigned transitions to experimental uncertainty. However, all assigned transitions can be used for astronomical searches of the rotational spectra of both conformers, therefore, quantum number overviews are given for clarification (see Fig. C.1). Furthermore, an estimated reduced energy diagram of Ag n-propanol is shown, which illustrates the location of tunneling-rotation interactions (see Fig. C.2). 0f course, interactions between rotational energy levels of Ag n-propanol and those of Aa n-propanol should also be considered if a global analysis is performed, but these interactions are not illustrated here.

thumbnail Fig. C.1

Quantum number overview of the analysis of (a) Aa n-propanol and (b) Ag n-propanol. All rotational energy levels of assigned transitions are visualized with the J quantum numbers on the x-axis, Ka quantum numbers on the y-axis, and Kc quantum numbers are either depicted by squares or circles for Kc = J - Ka or Kc = J - Ka+ 1, respectively. Filled markers depict quantum numbers of assigned transitions whose frequencies are fit to experimental accuracy using our final spectroscopic model in Table 2. The empty circles mark assigned transitions which are not included in the final analysis, but can be used for astronomical searches. For Aa n-propanol, we note that we do not differentiate between A and Ε components of each rotational energy level as in overwhelming majority of cases both components are assigned per transition. For Ag n-propanol, the rotational energy levels of the 0+ and 0− tunneling states are given in blue and red, respectively. Triangles are used for transitions whose tunneling state assignments have only been guessed so far.
thumbnail Fig. C.2

Reduced energy diagram of Ag n-propanol. The reduced energy is Ered = ΕT + Erot - J • (J + 1) • (B + C)/2. Thereby, the energies of the rotational levels Erot are calculated based on ab initio calculations of Kisiel et al. (2010). We used the same parameters for both tunneling states as ab initio calculations are presented for a single state only. We note that actual parameters of the two tunneling states will differ. We set the energy difference of the two tunneling states to 3 cm−1, i.e. and , on the assumption that the splitting is similar to what has been observed for gauche-ethmol (Pearson et al. 1997). (a) The reduced energy diagram for rotational energy levels with Ka < 9 shows that series with Ka > 5 may not cross each other. We note that it may happen if different rotational parameters for the two tunneling states are taken into account, (b) The reduced energy diagram for rotational energy levels with Ka < 4 is a zoom of the y-axis of (a). We note that many series cross each other and even though the exact energy difference between the two tunneling states Ag+ and Ag- is unknown so far, many tunneling-rotation interactions occur in this region and determining an accurate energy difference is essential to properly take them into account.

References

  1. Abdurakhmanov, A., Rahimova, R., & Imanov, L. 1968, Izv. Akad. Nauk. Azerb. SSR, 3, 15 [Google Scholar]
  2. Abdurakhmanov, A., Rahimova, R., & Imanov, L. 1970, Phys. Lett. A, 32, 123 [NASA ADS] [CrossRef] [Google Scholar]
  3. Abdurakhmanov, A., Ragimova, R., Veliyulin, E., & Imanov, L. 1976, VINITI (a Russian database), 3883-76, 6 [Google Scholar]
  4. Ball, J. A., Gottlieb, C. A., Lilley, A. E., & Radford, H. E. 1970, Astrophys. J. Lett., 162, L203 [NASA ADS] [CrossRef] [Google Scholar]
  5. Belloche, A., Garrod, R. T., Müller, H. S. P., et al. 2019, A&A, 628, A10 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  6. Belloche, A., Garrod, R. T., Zingsheim, O., Müller, H. S. P., & Menten, K. M. 2022, A&A, 662, A110 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  7. Bonah, L., Zingsheim, O., Müller, H. S., Guillemin, J.-C., & Schlemmer, S. 2022, J. Mol. Spectrosc., submitted [Google Scholar]
  8. Bouchez, A., Walters, A., Müller, H. S., et al. 2012, J. Quant. Spectr. Rad. Transf., 113, 1148 [NASA ADS] [CrossRef] [Google Scholar]
  9. Christen, D., & Müller, H. S. P. 2003, Phys. Chem. Chem. Phys., 5, 3600 [NASA ADS] [CrossRef] [Google Scholar]
  10. Dreizler, H., & Scappini, F. 1981, Z. Naturforsch. A, 36, 1187 [NASA ADS] [CrossRef] [Google Scholar]
  11. Endres, C. P., Schlemmer, S., Schilke, P., Stutzki, J., & Müller, H. S. 2016, J. Mol. Spectr., 327, 95 [Google Scholar]
  12. Fukushima, K., & Zwolinski, B. J. 1968, J. Mol. Spectr., 26, 368 [NASA ADS] [CrossRef] [Google Scholar]
  13. Groner, P. 1992, J. Mol. Spectr., 156, 164 [NASA ADS] [CrossRef] [Google Scholar]
  14. Groner, P. 1997, J. Chem. Phys., 107, 4483 [Google Scholar]
  15. Groner, P. 2012, J. Mol. Spectr., 278, 52 [NASA ADS] [CrossRef] [Google Scholar]
  16. Hermanns, M., Wehres, N., Lewen, F., Müller, H., & Schlemmer, S. 2019, J. Mol. Spectr., 358, 25 [NASA ADS] [CrossRef] [Google Scholar]
  17. Herschbach, D. R., & Swalen, J. D. 1958, J. Chem. Phys., 29, 761 [NASA ADS] [CrossRef] [Google Scholar]
  18. Hollis, J. M., Lovas, F. J., Jewell, P. R., & Coudert, L. H. 2002, ApJ., 571, L59 [NASA ADS] [CrossRef] [Google Scholar]
  19. Johnson, D. R., Lovas, F. J., Gottlieb, C. A., et al. 1977, ApJL, 218, 370 [NASA ADS] [CrossRef] [Google Scholar]
  20. Jørgensen, J. K., van der Wiel, M. H. D., Coutens, A., et al. 2016, A&A, 595, A117 [Google Scholar]
  21. Kahn, K., & Bruice, T. C. 2005, ChemPhysChem, 6, 487 [CrossRef] [Google Scholar]
  22. Kisiel, Z., Dorosh, O., Maeda, A., et al. 2010, Phys. Chem. Chem. Phys., 12, 8329 [NASA ADS] [CrossRef] [Google Scholar]
  23. Lotta, T., Murto, J., Räsänen, M., & Aspiala, A. 1984, Chem. Phys., 86, 105 [NASA ADS] [CrossRef] [Google Scholar]
  24. Maeda, A., Lucia, F. C. D., Herbst, E., et al. 2006, ApJS, 162, 428 [NASA ADS] [CrossRef] [Google Scholar]
  25. Manigand, S., Jørgensen, J. K., Calcutt, H., et al. 2020, A&A, 635, A48 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  26. Martin-Drumel, M.-A., van Wijngaarden, J., Zingsheim, O., et al. 2015, J. Mol. Spectr., 307, 33 [NASA ADS] [CrossRef] [Google Scholar]
  27. Müller, H. S., Thorwirth, S., Roth, D., & Winnewisser, G. 2001, A&A, 370, L49 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  28. Müller, H. S., Schlöder, F., Stutzki, J., & Winnewisser, G. 2005, J. Mol. Struct., 742, 215 [Google Scholar]
  29. Ordu, M. H., Zingsheim, O., Belloche, A., et al. 2019, Astron. Astrophys., 629, A72 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  30. Pearson, J. C., Sastry, K. V. L. N., Herbst, E., & Lucia, F. C. D. 1997, ApJ, 480, 420 [NASA ADS] [CrossRef] [Google Scholar]
  31. Pearson, J. C., Brauer, C. S., & Drouin, B. J. 2008, J. Mol. Spectr., 251, 394 [Google Scholar]
  32. Pickett, H. M. 1972, J. Chem. Phys., 56, 1715 [NASA ADS] [CrossRef] [Google Scholar]
  33. Pickett, H. M. 1991, J. Mol. Spectr., 148, 371 [Google Scholar]
  34. Pickett, H. M., & Scroggin, D. G. 1974, J. Chem. Phys., 61, 3954 [NASA ADS] [CrossRef] [Google Scholar]
  35. Rubin, R. H., Swenson, G. W. J., Benson, R. C., Tigelaar, H. L., & Flygare, W. H. 1971, ApJL, 169, L39 [NASA ADS] [CrossRef] [Google Scholar]
  36. Tercero, B., Cernicharo, J., López, A., et al. 2015, A&A, 582, L1 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  37. Tercero, B., Cuadrado, S., López, A., et al. 2018, A&A, 620, L6 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  38. Zingsheim, O., Müller, H. S. P., Lewen, F., Jørgensen, J. K., & Schlemmer, S. 2017, J. Mol. Spectr., 342, 125 [NASA ADS] [CrossRef] [Google Scholar]
  39. Zingsheim, O., Bonah, L., Lewen, F., et al. 2021, J. Mol. Spectr., 381, 111519 [NASA ADS] [CrossRef] [Google Scholar]
  40. Zingsheim, O., Müller, H. S., Bonah, L., et al. 2022, J. Mol. Spectr., 384, 111565 [NASA ADS] [CrossRef] [Google Scholar]
  41. Zuckerman, B., Turner, B., Johnson, D., et al. 1975, ApJ, 196, L99 [NASA ADS] [CrossRef] [Google Scholar]
2

The term “trans” (T) was often used instead of “anti” historically, although it should only be used to refer to planar molecules.

3

For instance, the Ga conformer was predicted to be the energetically lowest energy conformer with the others higher in energy by E(Aa) = 24 cm−1 (35 K), E(Gg) = 35 cm−1 (50 K), E(Ag) = 42 cm−1 (60 K), and E(Gg') = 49 cm−1 (70 K) (Kahn & Bruice 2005). Experimentally derived values for the G family are E(Ga) = 0 cm−1, E(Gg) = 47.8 cm−1 (69 K), and E(Gg') = 50.8 cm−1 (73 K) (Kisiel et al. 2010).

4

Transitions with identical selection rules are called a series, here abbreviated as , with ΔJ represented by capital letters P, Q, and R for ΔJ = -1, ±0 or +1, respectively, and ΔKa = -1, ±0, or +1 by superscripts p, q, or r, respectively. For instance, a-type transitions (ΔKa = 0) with ΔJ = +1 are called qR series transitions. A subscript may be used additionally to highlight the respective Ka quantum number ( series).

6

Two Q-branch α-type lines (or four transitions) were not included in the analysis of Dreizler & Scappini (1981), but were secured by them with MW-MW DR spectroscopy; These lines are successfully incorporated to the fit in this study. However, six other transitions have been unweighted in this study due to conspicuously large deviations.

7

Allowed a- and b-type transitions occur within a tunneling state: Ag+ ↔ Ag+ and Ag− ↔ Ag−, whereas c-type transitions occur between the tunneling states: Ag+↔ Ag−.

8

Future analyses may also benefit by taking into account preliminary results of Kisiel et al. (2010), presented at the International Symposium on Molecular Spectroscopy (ISMS) 2006. We were made aware of this after submission; assignments are not publicly available.

9

Caution is advised concerning the number of interaction parameters; Using too many leads to high correlations and convergence problems. Using too few may lead to a more stable fit, but may not account for all perturbations, or, if it does, at the expense of unphysical parameters.

10

Our predictions (*.cat files) will be available together with further information; cf. Appendix A, in the CDMS soon.

11

Reducing the total degeneracy factor of 4 is not possible if only one column density for all five conformers of n-propanol are to be derived, e.g., in local thermal equilibrium conditions in space, as the Aa conformer does not exist in a mirror image form and A and E internal rotation components are resolved.

All Tables

Table 1

Quantum number overview of the final analysis of Aa n-propanol.

In the text
Table 2

Spectroscopic parameters of Aa n-propanol.

In the text
Table A.1

Partition function QGx of all three G conformers (with spin weight 4), QGa(1) of Ga (with spin weight 1), together with conformational correction factors FConf, vibrational correction factors FVib, and their product FTot for various temperatures of interest for an astronomical detection.

In the text

All Figures

thumbnail Fig. 1

Aa and Ag conformers of n-propanol. Hydrogen, carbon, and oxygen atoms are depicted with white, black, and red spheres, respectively. The atoms C-C-C-O-H span a (symmetry) plane for the Aa conformer (Cs symmetry) and it is shown as face-on and side view for clarification. Then, Ag n-propanol occurs in two mirror-image forms leading to the Ag+ and Ag− tunneling states. The G conformers are not depicted here, but the oxygen is out of the heavy nuclei plane for all of them.
In the text
thumbnail Fig. 2

Loomis-Wood plot of Aa n-propanol - screenshot of the LLWP software (Bonah et al. 2022). Shown here is a doublet series (A and E components of the methyl group internal rotation) of the J3,J-3 ← J2,J-2 b-type transitions from J = 6 (bottom panel) to J = 36 (top panel). Predictions of the A components are set as center frequencies and are depicted by green sticks. The E components are depicted by red ones. This rQ series with Ka + Kc = J and ΔΚa, ΔKc = +1,-1 is well reproduced by our spectroscopic model described in Table 2 up to J = 22, however, transitions up to J = 41 (Δν ≈ 90 MHz) are straightforwardly assignable.
In the text
thumbnail Fig. 3

DM-DR measurements of Aa n-propanol of the transition around 232.55 GHz. The lower panels show two DM-DR measurements where the A and E components of the transition are used as pump frequencies at 113 811.6420 MHz and 113 810.5391 MHz, respectively. A conventional measurement is shown in the upper panel. E* transitions, which are forbidden in rigid rotors (Herschbach & Swalen 1958), secure the assignment of internal rotation components. We note that slightly off-resonant pumping leads to small, asymmetric features of A and E components in the DM-DR spectrum of the other one, as pumped A and E components are rather close in frequency, cf. Zingsheim et al. (2021).
In the text
thumbnail Fig. 4

Fortrat diagram of n-propanol. Top panel: a broadband measurement in the W-band region is shown, where respective regions with signatures of the Ag conformer, leading to the Fortrat diagram, are highlighted. In some regions, the spectrometer is not as sensitive, e.g. around J=11, compared to others. Bottom panel: The colored lines in the Fortrat diagram mark linkages of series transitions, . DM-DR spectroscopy secured assignments in the W-band region (from J = 10 to J = 16). The existence of two tunneling states is proven as for each Ka partner series are found; the tunneling states are marked by 0+ and 0−. More series can be found, but are not assigned yet, e.g. QR3 ones. We note the larger tunneling splitting of QR6 series (green lines), e.g., in comparison to QR5 ones, and the different slopes for QRS series (thick purple lines).
In the text
thumbnail Fig. 5

Asymmetry splitting of the qR4, qR5, qR6, and qR7 series transitions, (a) Exemplary Loomis-Wood plot for qR6 transitions o7 0+. Two transitions, either with Kc = J - Ka + 1 or with Kc = J - Ka, are observed and marked by green qtars in each row. The average frequency of both transitions is the center frequency (Frequeney Offset is 0 MHz). Sceeenshot of the LLWP software (Bonah et al. 2022). The frequency difference of both components is the asymmetry splitting for Ka = 6 of 0+ for a given J. (b) Οbserved and calculated asymmetry splittings for qR4 (blue), qR5 (red), qR6 (green) and qR6 (orange). Similar observed and calculated asymmetry splittings confirm the correct assignment of Ka quantum numbers. We note that small asymmetry splittings are not resolvable, both transitions are partly blended, and missing experimental values for J = 17-22 eriginate in thx experimental gap from 130-170 GHz.
In the text
thumbnail Fig. B.1

Microwave spectra of n-propanol and propanal. Assignments of α-type transitions with are shown for both molecules (Ka's are given in the figure). Top panel: Rotational spectrum of gauche-propanal with assignments from Zingsheim et al. (2022) marked in blue. Bottom panel: Assigned doublet transitions for Ag ra-propanol from Abdurakhmanov et al. (1976) are marked in red, but cannot be verified in our CP-FTMW jet experiment. Our tentative assignments of Ka = 1 transitions of Ag n-propanol are marked in blue. We note that the transition marked with an asterisk (*) is another possible candidate for a transition with Ka = 1, more information can be found in the text.
In the text
thumbnail Fig. B.2

Fortrat diagram of n-propanol. The colored lines in the Fortrat diagram mark linkages of series transitions of Ag n-propanol, - DM-DR spectroscopy secured assignments in the W-band region (from J = 10 to J = 16). Tentative assignments of Κa = 1 transitions derived from the CP-FTMW measurement (Fig. B.1) are marked by the blue arrows. Considering found qR1 series, another candidate transition is marked by the blue asterisk. A quantum mechanical model with experimental accuracy can solve the ambiguity of these tentative assignments. The gray area depicts the region of the Fortrat diagram in Fig. 4.
In the text
thumbnail Fig. B.3

Loomis-Wood plot around the assigned qR0 series transitions in the W-band region of the Ag conformer, which are shown by red sticks.
In the text
thumbnail Fig. B.4

Loomis-Wood plot around the assigned qR1 series transitions in the W-band region of the Ag conformer, which are shown by red sticks. Theblue asterisk (*) marks the same transition as is marked in Fig. B.1.
In the text
thumbnail Fig. B.5

Loomis-Wood plot around the assigned qR2 series transitions in the W-band region of the Ag conformer, which are shown by red sticks. The series higher in frequency with Ka + Kc = J + 1 is probably strongly interacting with a so far unknown state around J = 12.
In the text
thumbnail Fig. B.6

Loomis-Wood plot around assumed qR3 series transitions in the W-band region. No assignments with Ka = 3 are secured yet, but the existence of such can be seen by the many unassigned transitions in the shown region. Four series close in frequency are expected; Two asymmetry sides for both tunneling states.
In the text
thumbnail Fig. C.1

Quantum number overview of the analysis of (a) Aa n-propanol and (b) Ag n-propanol. All rotational energy levels of assigned transitions are visualized with the J quantum numbers on the x-axis, Ka quantum numbers on the y-axis, and Kc quantum numbers are either depicted by squares or circles for Kc = J - Ka or Kc = J - Ka+ 1, respectively. Filled markers depict quantum numbers of assigned transitions whose frequencies are fit to experimental accuracy using our final spectroscopic model in Table 2. The empty circles mark assigned transitions which are not included in the final analysis, but can be used for astronomical searches. For Aa n-propanol, we note that we do not differentiate between A and Ε components of each rotational energy level as in overwhelming majority of cases both components are assigned per transition. For Ag n-propanol, the rotational energy levels of the 0+ and 0− tunneling states are given in blue and red, respectively. Triangles are used for transitions whose tunneling state assignments have only been guessed so far.
In the text
thumbnail Fig. C.2

Reduced energy diagram of Ag n-propanol. The reduced energy is Ered = ΕT + Erot - J • (J + 1) • (B + C)/2. Thereby, the energies of the rotational levels Erot are calculated based on ab initio calculations of Kisiel et al. (2010). We used the same parameters for both tunneling states as ab initio calculations are presented for a single state only. We note that actual parameters of the two tunneling states will differ. We set the energy difference of the two tunneling states to 3 cm−1, i.e. and , on the assumption that the splitting is similar to what has been observed for gauche-ethmol (Pearson et al. 1997). (a) The reduced energy diagram for rotational energy levels with Ka < 9 shows that series with Ka > 5 may not cross each other. We note that it may happen if different rotational parameters for the two tunneling states are taken into account, (b) The reduced energy diagram for rotational energy levels with Ka < 4 is a zoom of the y-axis of (a). We note that many series cross each other and even though the exact energy difference between the two tunneling states Ag+ and Ag- is unknown so far, many tunneling-rotation interactions occur in this region and determining an accurate energy difference is essential to properly take them into account.
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.