Skip to main content
UKPMC Funders Author Manuscripts logoLink to UKPMC Funders Author Manuscripts
. Author manuscript; available in PMC: 2019 Mar 15.
Published in final edited form as: Chemphyschem. 2018 Sep 20;19(23):3182–3185. doi: 10.1002/cphc.201800744

Direct Evidence of the Benzylium and Tropylium Cations as the two long-lived Isomers of C7H7+

Pavol Jusko a, Aude Simon b, Shreyak Banhatti c, Sandra Brünken d, Christine Joblin a,*
PMCID: PMC6420061  EMSID: EMS80180  PMID: 30238585

Abstract

Disentangling the isomeric structure of C7H7+ is a longstanding experimental issue. We report here the full mid-infrared vibrational spectrum of C7H7+ tagged with Ne obtained with infrared-predissociation spectroscopy at 10 K. Saturation depletion measurements were used to assign the contribution of benzylium and tropylium isomers and demonstrate that no other isomer is involved. Recorded spectral features compare well with density functional theory calculations. This opens perspectives for a better understanding and control of the formation paths leading to either tropylium or benzylium ions.

Keywords: benzylium, tropylium, structure elucidation, IR spectroscopy, cryogenic ion trap


Benzylium (Bz+) and tropylium (Tr+) ions are key isomers of C7H7+ commonly produced from energized toluene (C7H8).[1,2,3] Whereas Bz+consists of a benzene ring substituted with a methylene group, the Tr+ isomer is a fully aromatic ion made of a 7-membered CH ring. The possibility that similar structures can be involved in the fragmentation of methyl-substitued polycyclic aromatic hydrocarbon ions has been recently discussed.[4,5] Still, one of the major issues in these studies is the limited understanding of the formation paths of these species, that would allow us to predict and therefore control the production of one or the other isomer. On the one hand, it was shown that the choice of the precursor is important, e. g., the use of halogen-substitued toluene is expected to optimise the production of Bz+.[6] But on the other hand, the Bz+/Tr+ ratio depends strongly on the various experimental conditions.[7,8] Another point, in photoionization experiments of toluene, is that it is not possible to produce solely Tr+ at appearance threshold,[9] despite calculations show Tr+ being energetically more favorable than Bz+ (by ~ 38 kJ · mol−1; cf. Tab S3 in the SI). The latter results were interpreted as due both to the presence of a barrier in the dissociation path from toluene towards Tr+, and to the role of autoionizing states in promoting a nonstatistical formation of Bz+ over Tr+.[9] In contrast to Tr+, theoretical studies show that paths towards Bz+ lack high barriers, and no barrier is found in the path from benzyl chloride ionization towards Bz+ formation.[10]

It is known since earlier studies on C7H7+ that a convenient way to identify Bz+ is through ion-molecule reaction with toluene, leading to efficient formation of C8H9+.[11,1] The non-reactive part of the ion population is then attributed to Tr+. In parallel, there have been many attempts to characterize the structure of C7H7+ directly, i. e., via spectroscopy. Electronic absorption features have been studied in both Ne matrix[12] and in gas-phase by photodissociation.[13,14] For Bz+, the vibronic structure was resolved for the S1←S0 transition but much broader features were observed for the higher excited states. Absorption features of Tr+ are expected in the UV range,[12] but the data obtained in Ne matrices could not be confirmed in the gas-phase.[13] On the opposite, infrared and Raman spectra of Tr+could be obtained in solutions or in solid phase using salts.[15,16,17] In gas-phase, one can think of infrared multiple photon dissociation (IRMPD) vibrational spectroscopy as the technique of choice when characterizing the structure of ions in mass spectrometry experiments.[18] Information on the Bz+/Tr+ dichotomy[19] could be obtained from the IRMPD spectra of derivative ions with lower dissociation thresholds compared to C7H7+.[20,21,19,22] The spectroscopy of polycyclic hydrocarbon cations containing Bz+ and Tr+ structure was also investigated.[8,5] However, the IRMPD spectra of pure Bz+ and Tr+ are still lacking, despite the efforts.[20] This could be due to the exceptional stability of these ions, or possible isomerisation issues, as found in the case of C17H11+.[5]

An alternative approach to IRMPD not owning the disadvantages of the multi-photon process, is infrared pre-dissociation (IR-PD) spectroscopy of a weakly bound complex of the ion with a rare gas atom. Thanks to this technique, we were able to obtain the first complete mid-IR-PD spectra of Ne tagged C7H7+ recorded in gas-phase and at low temperature. We used the cryogenic 22 pole ion trap,[23] which is coupled to the free electron laser FELIX, as described in the experimental section below. We introduced different precursors and varied the ionization conditions (cf. Tab. S1 in the SI). Figure 1 gathers the recorded experimental spectra, which appear to agree well with the calculated spectra of Bz+ and Tr+ using density functional theory (DFT) at the B3LYP/6-31G(d,p) level of theory as implemented in Gaussian09 suite of programs.[24] In addition, the contribution of Bz+ and Tr+ in each mixture could be determined directly during the spectroscopic experiment by using saturation depletion measurements. To achieve the latter, the laser is tuned in resonance with a vibrational band of interest. By applying a large number of photon pulses, the optically active isomer-Ne complex can be dissociated completely leaving only the optically inactive isomer-Ne complex in the trap. Recording the ion-Ne (m = 111 u) number as a function of time reveals the relative abundance of active to inactive isomers. Depletion tests for selected bands of both isomers and under varying ion source conditions are shown in Fig. 2. They show that one condition leads to 90% Bz+ population and another to 60-70% of Tr+ with 40-30% of Bz+. These measurements provide the first complete demonstration that Bz+ and Tr+ are the only long-lived isomers of C7H7+. In a preparatory study, we also recorded the chemical reactivity with toluene and found a good agreement with the above measurements, although only Bz+ can be directly monitored (cf. Fig. S2 in the SI).

Figure 1.

Figure 1

Mid-IR spectrum of C7H7+. Top panel – IR pre-dissociation spectrum of Bz+ tagged with Ne. Middle panel – stick spectrum corresponding to the scaled DFT harmonic spectra for Bz+ (blue) and Tr+ (red). Convoluted spectra with σ = 0.5% BW are provided for comparison with the experiment. Bottom panel – IR pre-dissociation spectrum of Tr+ tagged with Ne. Bands marked by (a), (b), (c), (d) have been confirmed to belong to Tr+ and those marked (a), (u), (v) to Bz+ by saturation depletion measurements. Intensities for experimental spectra are in arbitrary units. All recorded bands are listed in Tab. 1 for Bz+ and Tab. 2 for Tr+.

Figure 2.

Figure 2

Depletion of C7H7+ · Ne in the ion trap for the different experimental conditions used in Fig.1. Color code: Blue – wavelength at which the Bz+ isomer is active. Red – wavelength at which the Tr+ isomer is active. Magenta – both isomers active. Black – corresponds to non-resonant dissociation of the Ne tag (~ 8 – 10 mJ/pulse). Numbers in the right column correspond to saturation depletion in percent. Letters refer to the bands shown in Fig. 1.

Assignment of the experimental bands of Bz+ and Tr+ is made in Tables 1 and 2. Additional levels of theory were carried out for benchmarking (cf. Section 2 in the SI). For each level, the calculated band positions are linearly scaled with a scaling factor S, in order to match with the strongest Bz+ band. From calculations on Ne tagged ions, we found that the top (T) and molecular plane (P) isomers are quasi degenerate (cf. Tab. S3 in the SI) and that the Ne tag is not expected to induce band shifts (cf. Figs. S12 and S13, Table S4 in the SI), nor splitting of the bands due to the degeneracy lifting in the Tr+ case, that would be observable in our experiments (cf. Fig. S12 in the SI, degenerate bands always split by less than 1 cm−1). Overall, we found a good agreement between the measured and calculated band positions. However, in the case of Tr+, one major difference exists. The harmonic calculations provide only three infrared active bands, which can be rationalized by the high symmetry of Tr+ which pertains to the D7h point group symmetry (cf. Tab. S2 in the SI for descent in symmetry from D7h to C2v, in which calculations are performed). In the experimental spectrum, we found four bands, two bands including the additional one falling at low frequency. We note that two close-by bands in this range are also observed in some experiments in condensed phase.[15,16,17]

Table 1.

Experimental mid-IR band positions recorded for Bz+ and calculated values using DFT (B3LYP/6-31G(d,p)) obtained in the present work (see Figure 1). Only the modes which are IR active are reported. Their symmetry in the point group of the molecular ion (C2v) are reported.

Mode Exp. Calc.
ν ν Inten. Sym.
ν5 524 521 4.3 A1
ν6 596 594 1.8 B2
ν8 632 628 42.8 B1
ν9 790 780 26.3 B1
ν10 820 803 2.6 A1
ν12 962 0.9 B2
ν13 974 976 5.4 A1
ν16 992 1000 2.4 A1
ν17 1036 1026 2.3 B1
ν18 1070 1077 23.2 B1
ν19 1114 1109 4.0 B2
ν20 1184 1174 8.4 B2
ν21 1198 1184 18.5 A1
1232
1302
ν22 1332 1321 5.6 B2
ν23 1356 1355 105.5 A1
ν24 1400 1393 16.1 B2
ν25 1448 1446 96.4 B2
ν26 1467 0.2 A1
ν27 1541 0.6 B2
ν28 1554 1564 1.8 A1
ν29* 1630 1630 250.9 A1

Note: All frequencies in cm−1, intensities in km · mol−1.

*

mode used to determine the scaling factor 0.974. Cf. Fig. S15. in the SI for the visualisation of modes.

Table 2.

Experimental mid-IR band positions recorded for Tr+ and calculated values using DFT (B3LYP/6-31G(d,p)) obtained in the present work (see Figure 1). Only the modes which are IR active are reported. Although they were computed in the C2v symmetry point group, their symmetry in the point group of the molecular ion (D7h) are listed.

Mode Exp. Ref. Calc.
ν ν ν (Inten.) Sym.
νc 630 633,[15] 634,[16] 655[17]
ν4 652 658,[15] 660,[16] 685[17] 646(85.8) A”2
ν8 994 992,[15] 992[16] 989(3.7) E’1
ν14 1486 1477,[15] 1477[16] 1480(42.9) E’1

Note: All frequencies in cm−1, intensities in km · mol−1. Scaling factor 0.974. νc is a combination band of the two lowest lying fundamental bands of Tr+. Cf. Fig. S16 – S18. in the SI for the visualisation of modes.

Saturation depletion measurements established that bands (b) – (d) in the experimental spectrum (Fig. 1), belong solely to Tr+, and that both Bz+ and Tr+ are active at the 630 cm−1 band (a). Bands (c) and (d) can be assigned to the doubly degenerate in plane CH bending mode and in plane CC stretching mode of Tr+, respectively. The third Tr+ band, the CH out-of-plane bending mode (not degenerate) should lay around 650 cm−1 (close to band (b), cf. Fig. S3). The performed saturation depletion measurements invalidate the hypothesis that both bands (a) and (b) are caused by distinct (T) and (P) Tr+ · Ne isomers. We also excluded contribution from a triplet spin state, whose presence can lead to shifts in low energy modes for PAHs.[25] In the case of Tr+, it was found distorted and 3.3 eV above the ground state at the B3LYP/6-31G(d,p) level of theory (cf. Fig. S14 in the SI for its computed IR spectrum). The excited state can not survive many collisions with cold buffer gas and its spectrum does not contain features in the wavelength range of interest. The most likely scenario to account for bands (a) and (b), is that a combination band comes into close resonance with the fundamental CH out-of-plane bending mode, and borrows a significant part of its intensity. Our calculations show that the combination band involves the two lowest vibrational modes of Tr+ and that the combination band is red shifted relative to the fundamental band (Tab. 2 and Tab. S5 in the SI). In our experiment, we can assign band (b) as the fundamental mode (i. e. the strongest band) since the initial dissociation rate of this band is higher and its band width at saturation is larger as compared to band (a) (Fig. 2). Finally, our spectra do not include the CH stretch range. However, we learned while writing this manuscript that this range has recently been studied using a jet experiment and a table-top laser.[26]

We recorded the mid-IR-PD spectra of C7H7+ tagged with Ne at 10 K. The comparison of the experimental spectra with the harmonic calculated spectra show, that neither significant band shifts, nor band splitting are induced by the presence of the Ne tag, that can in particular induce a symmetry decrease in the case of Tr+. The Tr+ spectrum exhibits an interesting strong resonance at low frequency between a combination band and the fundamental CH out-of-plane bending mode. Anharmonic calculations would help in providing a detailed quantitative study of this resonance.[27,28]

We used saturation depletion measurements to confirm the existence of only two long-lived isomers of C7H7+, and the obtained spectra support their initial structural assumption as Bz+ and Tr+.[1,9] The depletion technique in an ion trap has the advantage of being able to track the abundance of a well-defined isomer, contrary to the earlier reactivity method, which only has access to Bz+. This technique opens new perspectives for our understanding of the isomerisation paths of C7H7+.

Experimental Section

The ions are produced by electron bombardment in a storage type ion source[29] from benzyl chloride and toluene precursors. Ions of mass 91 u (C7H7+) are mass selected in a quadrupole mass filter prior to being injected into the cryogenic 22 pole rf ion trap (nominal temperature 8 − 9 K), where they are cooled by a 100 ms long 3:1 He:Ne gas pulse. The high number gas density achieved during this pulse (n ~ 1014 cm−3) promotes ternary attachment of Ne to the bare ion producing Bz+ · Ne and Tr+ · Ne complexes. After the cooling and tagging sequence, the number density in the trap decreases quickly to a level where no further attachment takes place. The ion–Ne complex is then dissociated by infrared radiation provided by the FELIX FEL-2 free electron laser,[30] operated in the 500 − 1650 cm−1 range. The laser delivers up to 30 mJ in a single macropulse into the 22 pole trap with a repetition rate of 10 Hz and spectral bandwidth better than σ = 0.5%. After an irradiation time of typically 2.6 s the trap content is emptied through a mass filter onto a counting detector.

Supplementary Material

Supporting information

Acknowledgements

The research leading to this result is supported by the European Research Council under the European Union’s Seventh Framework Programme ERC-2013-SyG, Grant Agreement n. 610256 NANOCOSMOS. We acknowledge support from the project CALIPSOplus under the Grant Agreement 730872 from the EU Framework Programme for Research and Innovation HORIZON 2020. We greatly appreciate the experimental support provided by the FELIX team. We gratefully acknowledge the Nederlandse Organisatie voor Wetenschappelijk Onderzoek (NWO) for the support of the FELIX Laboratory. We thank the Cologne Laboratory Astrophysics group for providing the FELion ion trap instrument and the Cologne Center for Terahertz Spectroscopy (core facility, DFG grant SCHL 341/15-1) for supporting its operation. S.Ba. is supported by the H2020-MSCA-ITN-2016 Program (EUROPAH project, G. A. 722346). A. S. thanks the computing facility CALMIP for generous allocation of computer resources.

References

  • [1].Dunbar RC. Journal of the American Chemical Society. 1975;97:1382–1384. doi: 10.1021/ja00839a016. [DOI] [Google Scholar]
  • [2].Kuck D. Mass Spectrometry Reviews. 1990;9:583–630. doi: 10.1002/mas.1280090602. [DOI] [Google Scholar]
  • [3].Lifshitz C. Accounts of Chemical Research. 1994;27:138–144. doi: 10.1021/ar00041a004. [DOI] [Google Scholar]
  • [4].Rapacioli M, Simon A, Marshall CCM, Cuny J, Kokkin D, Spiegelman F, Joblin C. The Journal of Physical Chemistry A. 2015;119:12845–12854. doi: 10.1021/acs.jpca.5b09494. [DOI] [PubMed] [Google Scholar]
  • [5].Jusko P, Simon A, Wenzel G, Brünken S, Schlemmer S, Joblin C. Chemical Physics Letters. 2018;698:206–210. doi: 10.1016/j.cplett.2018.03.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Jackson JAA, Lias SG, Ausloos P. Journal of the American Chemical Society. 1977;99:7515–7521. doi: 10.1021/ja00465a020. [DOI] [Google Scholar]
  • [7].Fridgen TD, Troe J, Viggiano AA, Midey AJ, Williams S, McMahon TB. The Journal of Physical Chemistry A. 2004;108:5600–5609. doi: 10.1021/jp031328s. [DOI] [Google Scholar]
  • [8].Morsa D, Gabelica V, Rosu F, Oomens J, De Pauw E. The Journal of Physical Chemistry Letters. 2014;5:3787–3791. doi: 10.1021/jz501903b. [DOI] [PubMed] [Google Scholar]
  • [9].Lifshitz C, Gotkis Y, Laskin J, Ioffe A, Shaik S. The Journal of Physical Chemistry. 1993;97:12291–12295. doi: 10.1021/j100149a032. [DOI] [Google Scholar]
  • [10].Choe JC. The Journal of Physical Chemistry A. 2008;112:6190–6197. doi: 10.1021/jp802641c. [DOI] [PubMed] [Google Scholar]
  • [11].Shen J, Dunbar RC, Olah GA. Journal of the American Chemical Society. 1974;96:6227–6229. doi: 10.1021/ja00826a070. [DOI] [Google Scholar]
  • [12].Nagy A, Fulara J, Garkusha I, Maier JP. Angewandte Chemie International Edition. 2011;50:3022–3025. doi: 10.1002/anie.201008036. [DOI] [PubMed] [Google Scholar]
  • [13].Dryza V, Chalyavi N, Sanelli JA, Bieske EJ. The Journal of Chemical Physics. 2012;137:204304. doi: 10.1063/1.4767402. [DOI] [PubMed] [Google Scholar]
  • [14].Féraud G, Dedonder-Lardeux C, Soorkia S, Jouvet C. The Journal of Chemical Physics. 2014;140:024302. doi: 10.1063/1.4858409. [DOI] [PubMed] [Google Scholar]
  • [15].Fateley WG, Lippincott ER. The Journal of Chemical Physics. 1957;26:1471–1481. doi: 10.1063/1.1743566. [DOI] [Google Scholar]
  • [16].Sourisseau C. Spectrochimica Acta Part A: Molecular Spectroscopy. 1978;34:881–887. doi: 10.1016/0584-8539(78)80007-5. [DOI] [Google Scholar]
  • [17].Howard J, Graham D. Spectrochimica Acta Part A: Molecular Spectroscopy. 1985;41:815–824. doi: 10.1016/0584-8539(85)80027-1. [DOI] [PubMed] [Google Scholar]
  • [18].Oomens J, Sartakov BG, Meijer G, von Helden G. International Journal of Mass Spectrometry. 2006;254:1–19. doi: 10.1016/j.ijms.2006.05.009. [DOI] [Google Scholar]
  • [19].Chiavarino B, Crestoni ME, Dopfer O, Maître P, Fornarini S. Angewandte Chemie International Edition. 2012;51:4947–4949. doi: 10.1002/anie.201200558. [DOI] [PubMed] [Google Scholar]
  • [20].Chiavarino B, Crestoni ME, Fornarini S, Dopfer O, Lemaire J, Maître P. The Journal of Physical Chemistry A. 2006;110:9352–9360. doi: 10.1021/jp0628380. [DOI] [PubMed] [Google Scholar]
  • [21].Schröder D, Schwarz H, Milko P, Roithová J. The Journal of Physical Chemistry A. 2006;110:8346–8353. doi: 10.1021/jp056962f. [DOI] [PubMed] [Google Scholar]
  • [22].Zins E-L, Pepe C, Schröder D. Faraday Discuss. 2010;145:157–169. doi: 10.1039/B907236E. [DOI] [Google Scholar]
  • [23].Asvany O, Bielau F, Moratschke D, Krause J, Schlemmer S. Review of Scientific Instruments. 2010;81:076102. doi: 10.1063/1.3460265. [DOI] [PubMed] [Google Scholar]
  • [24].Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G, Barone V, Mennucci B, Petersson GA, Nakatsuji H, et al. Gaussian 09, Revision D.01. Gaussian Inc.; Wallingford CT: 2016. [Google Scholar]
  • [25].Falvo C, Calvo F, Parneix P. The Journal of Chemical Physics. 2012;137:064303. doi: 10.1063/1.4739468. [DOI] [PubMed] [Google Scholar]
  • [26].Wagner JP, McDonald DC, Duncan MA. The Journal of Physical Chemistry Letters. 2018;9:4591–4595. doi: 10.1021/acs.jpclett.8b02121. [DOI] [PubMed] [Google Scholar]
  • [27].Mackie CJ, Candian A, Huang X, Maltseva E, Petrignani A, Oomens J, Buma WJ, Lee TJ, Tielens AGGM. The Journal of Chemical Physics. 2015;143:224314. doi: 10.1063/1.4936779. [DOI] [PubMed] [Google Scholar]
  • [28].Mulas G, Falvo C, Cassam-Chenaï P, Joblin C. J Chem Phys. 2018 doi: 10.1063/1.5050087. http://arxiv.org/abs/1809.05669. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [29].Gerlich D. In: Adv Chem Phys: State-Selected and State-to-State Ion-Molecule Reaction Dynamics. Ng C-Y, Baer M, editors. LXXXII. Wiley; New York: 1992. pp. 1–176. [DOI] [Google Scholar]
  • [30].Oepts D, van der Meer AFG, van Amersfoort PW. Infrared Physics and Technology. 1995;36:297–308. doi: 10.1016/1350-4495(94)00074-U. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting information

RESOURCES