New Fragmentation Pathways in K–THF Collisions As Studied by

Article pubs.acs.org/JPCA

New Fragmentation Pathways in K−THF Collisions As Studied by Electron-Transfer Experiments: Negative Ion Formation D. Almeida,† F. Ferreira da Silva,† S. Eden,‡ G. García,§,⊥ and P. Limaõ -Vieira*,†,‡ †

Laboratório de Colisões Atómicas e Moleculares, CEFITEC, Departamento de Fı ́sica, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal ‡ Department of Physical Sciences, The Open University, Walton Hall, Milton Keynes MK7 6AA, United Kingdom § Instituto de Fı ́sica Fundamental, Consejo Superior de Investigaciones Cientı ́ficas, Serrano 113-bis, 28006 Madrid, Spain ⊥ Centre for Medical Radiation Physics, University of Wollongong, NSW 2522, Australia ABSTRACT: Time-of-flight (TOF) negative ion mass spectra have been obtained in collisions of 20−100 eV neutral potassium atoms with tetrahydrofuran (C4H8O), an analogue for the sugar unit in DNA/RNA. Major enhancements in O− and C2H3O− production were observed compared with earlier dissociative electron attachment (DEA) experiments. In further contrast with DEA, no evidence was observed for dehydrogenated parent anions, and three new fragment anions were detected: CH−, C2−, and C2H−. These contrasting results for potassium impact and DEA highlight significant differences in the reaction pathways initiated by the two electron delivery processes.

1. INTRODUCTION

was shown that the DEA profiles consist of Feshbach resonances where the parent neutral is in a Rydberg (excited) state. The electronic state spectroscopy of THF has been explored by photoelectron and VUV photoabsorption experiments together with quantum chemical calculations. These studies determined that Cs and C2 conformers (see Figure 1) coexist (55

Further to its applications as a solvent in chemical industries, tetrahydrofuran (THF) has attracted considerable interest as a model for the sugar unit in DNA/RNA. Electron-induced processes in THF and other molecules analogizing parts of key biological macromolecules have received particular attention due to evidence that subionization energy electrons can cause efficient damage in DNA through dissociative electron attachment (DEA) to its molecular subunits.1,2 As low-energy electrons are the most abundant products of ionizing radiation in the biological environment, detailed understanding of electron attachment and transfer processes is essential to characterize and model radiation damage mechanisms at the nanoscale level.3 This can have significant impact on radiotherapy developments, notably in the emerging field of nanodosimetry.4 Previous theoretical5−7 and experimental8−17 gas-phase studies have provided detailed descriptions of the molecular chemistry of gas-phase THF under low-energy (20 eV potassium impact energy), these fragments were initially not observed in DEA studies, although recent unpublished studies show the formation of these fragments.34 It therefore stands to reason that the presence of the potassium cation post collision strongly influences the reaction pathways. While the dominant DEA channel produces C2H3O−, the potassium cation has the effect of enabling other dissociation channels to successfully compete with C2H3O− production. Not only is the potassium cation responsible for repressing electron autodetachment, as was already proposed in other molecular targets,29,30 it also apparently affects the way that the fragmentation channels progress. An intriguing possibility is initial C−O cleavage followed by further bond breaking between the alkyl carbons of the no-longer furanosic intermediate ester anion. This proposed pathway can be considered competitive with the C3H5− and C2H3O− formation channels, which involve cleavage of different C−C bonds. In this case, the localization of the excess electron (either at the end of the alkyl group or at the ester) can be decisive. The aforementioned m/z 13, 14, 24, and 25 fragments are obtained through cleavages at different points of the alkyl chain, with the extra electron remaining in the alkyl side. As such, we propose that the initial stages of fragmentation are common to all of these fragments and hence we have plotted the summed BR for these channels in Figure 3. Another discrepancy when comparing electron transfer with DEA relates to OH− production, for which, although reported in both DEA and electron transfer, the relative yields for each technique are quite different. The formation of this anion requires quite a plausible tautomeric transition of a hydrogen atom bonded to an adjacent carbon followed by hydroxyl anion abstraction from the chain ester. H− formation is also observed in the present results, albeit with a very low yield at lower collision energies. By contrast, a significant H− yield was obtained in one of the DEA studies.12 This difference is most likely attributable to competition with other fragments such as C2H3O−, which can be seen to possess DEA resonances at similar energies.12

energy) does not show any signal reminiscent of a 6.2 eV resonance being accessed. Owing to the low perturbation of the positive ion core by the Rydberg electron, it is argued that the energy difference between the cationic state and the anionic state depends dominantly on the type of Rydberg state, that is, ∼4.5 eV for ns and ∼1.8 eV for np states.14 We propose an extension of this rationale to THF, a cyclic ether. As the general shape of the first ionization band6,25 and the DEA band for 41 m/z fragments with a maximum at 7.65 eV12 are similar, we can tentatively attribute this resonance to an initial occupation of an np Rydberg state associated with the first ionization energy. The DEA resonance with a maximum at 6.7 eV12 can be tentatively attributed to an initial excitation into a ns Rydberg state associated with the second ionization limit. Giuliani et al.25 reported that Rydberg series converging to the second ionization limit are exclusive of the C2 isomer of THF. Hence, it is plausible that perturbation of the ns and np Rydberg states by the potassium cation may be responsible for the contrasting C3H5−/C2H3O− ratios in DEA and electron-transfer experiments. The most recent study of DEA to THF proposes that the first step following electron capture is ring-opening via C−O bond breakage,12 with the excess electron primarily localized in the vicinity of the oxygen atom (see Figure 1). The high internal energy of the resulting metastable intermediate anion will lead to fragmentation or autodetachment.12 Fragmentation can occur in different parts of the alkyl chain. C3H5− and C2H3O− production can be traced to cleavage of the C2−C3 bond together with additional hydrogen abstractions. As the threshold energies for the production of these fragment anions via DEA and electron transfer appear to be consistent, we expect the reaction pathway described above to apply to both electron delivery mechanisms. Finally, an analysis of the BRs (Figure 3) shows a major change in the relative yields of C3H5− and C2H3O−as a function of collision energy. The ratio of C3H5− production/total anion production decreases from a 30 to 70 eV collision energy but slightly increases from 70 to 100 eV. Conversely, the C2H3O− BR remains constant from 30 to 70 eV but falls significantly from 70 to 100 eV. 3.2. O−. The fragment at 16 m/z (O−) is the most abundant anion. An analysis of the 20 eV collision energy spectrum (Figure 2) shows a very weak O− signal, indicating that the threshold for its formation is only slightly below 7.2 eV. DEA studies do not report O− formation at this energy; therefore, it is not possible to ascertain which of the anionic states is playing a role in yielding this fragment in the present experiments. However, it is interesting to look once again at the proposed intermediate anion state in Figure 4 and Figure 3 of ref 12. Owing to the localization of the extra electron, it is quite plausible to suggest that this intermediate state may also have an antibonding character along the C−O bond, eventually leading to bond breakage and O− formation. The absence of O− in DEA experiments is most likely due to O− formation competing with autodetachment or alternative fragmentation channels. In the former case, the presence of the potassium cation may suppress autodetachment long enough for fragmentation to successfully compete with re-ejection of the extra electron. In the latter case, the potassium cation may act as a perturbing third-body that changes the probability for the extra electron to undergo intramolecular transfer, thereby allowing different fragmentation channels. Finally, it is interesting to compare THF with D-ribose in the context of electron-transfer collisions. Both of these molecules

4. CONCLUSIONS The present work provides novel data on low-energy neutral atom collisions with THF, a possible model for the DNA/RNA sugar unit. These data strongly highlight the major differences in the fragmentation dynamics between capture of a free electron (DEA) and electron donation (as in electron-transfer experi694

dx.doi.org/10.1021/jp407997w | J. Phys. Chem. A 2014, 118, 690−696

The Journal of Physical Chemistry A

Article

Notes

ments). The difference between these two mechanisms is attributed to the presence of the potassium cation post collision, which can affect the decaying process of the temporary negative ion (TNI). Within this context, it is clear that the study of different electron delivery mechanisms can greatly increase our understanding of electron-driven processes in DNA/RNA damage. Indeed, it is worth noting that the DEA and electrontransfer fragmentation patterns for THF are the most dissimilar for all of the molecular targets studied so far. More specifically, some DEA studies report the formation of the parent anion and its dehydrogenated analogue,10 which are absent in the present study. This point highlights a tendency to observe an enhancement of ring breaking in electron transfer, as opposed to DEA, where the dominant fragmentation normally does not proceed through loss of ring integrity. The present data indicate that collision-frame threshold energies for the production of specific fragment anions by electron transfer are broadly consistent with the onsets of corresponding DEA resonances. This suggests that the initially accessed states in DEA and electron transfer are the same. A recent DEA study assigns such profiles to Feshbach resonances in which the neutral parent is in a Rydberg excited state.12,14 However, it is clear that these states can decay into different fragments, presumably due to the presence of the potassium cation. Indeed, it is proposed that the fragmentation of this molecule is sequential whereupon a chain ester anion is initially formed by C−O bond cleavage. This is followed by a subsequent break along the different parts of the alkyl chain. Depending on the site of bond cleavage, different fragments are formed. For example, a bond break of the remaining C−O bond will most likely lead to O− formation, whereas cleavages of the various C− C bonds will entail the formation of all other fragments (with the exception of OH−and H−). We then tentatively suggest that fragmentation will most likely consist of an intramolecular electron transfer from the initial intermediate state into different highly antibonding valence states along the C−C bonds. In other words, bond cleavage in the C1−C2 bond will most likely result in formation of CH−, CH2−, and C3H5−, whereas in the C2−C3 bond, it leads to formation of C2− and C2H− fragments. As for C2H3O− and O−, a similar mechanism also seems plausible but with the electron remaining on the ester side of the chain. It is striking that the fragmentation pattern in electron transfer is significantly richer than that reported in DEA experiments. Furthermore, the strongest of the fragment anions is O−, which is in complete contrast with DEA, where it was only recently reported and with a small relative yield. In conclusion, owing to the similar ring structure, THF appears to be a good sugar unit surrogate in the context of electron-transfer processes, mainly as far as the ring breaking dynamics are concerned. This is further supported by comparing the DEA resonance profiles of THF in the gas phase6,12 with the profiles for single- and double-strand breaks of DNA.1 Finally, the fragmentation patterns of D-ribose26 have shown the importance of the hydroxyl groups as the main anionic species formed in electron-transfer experiments. As such, while a possible surrogate for the sugar unit, the use of THF will not be able to provide information on the relevance of the hydroxyl groups in the actual DNA/RNA sugar unit.

■

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS D.A. and F.F.S. acknowledge the Portuguese Foundation for Science and Technology (FCT-MEC) for postgraduate SFRH/ BPD/99261/2013 and postdoctoral grants SFRH/BPD/68979/ 2010, respectively. We also acknowledge partial funding from the Portuguese research Grants PEst-OE/FIS/UI0068/2011 and PTDC/FIS-ATO/1832/2012 through FCT-MEC and from the Spanish Ministerio de Economı ́a y Competitividad (Project No. FIS 2012-31230). S.E. acknowledges the financial support of the British EPSRC (CAF fellowship). P.L.V. acknowledges his visiting Professor position at The Open University, U.K. Some of this work forms part of the EU/ESF COST Actions Nano-IBCTMP1002.

■

REFERENCES

(1) Boudaiffa, B.; Cloutier, P.; Hunting, D.; Huels, M. A.; Sanche, L. Science 2000, 287, 1658. (2) Antic, D.; Parenteau, L.; Lepage, M.; Sanche, L. J. Phys. Chem. B 1999, 103, 6611. (3) Baccarelli, I.; Bald, I.; Gianturco, F. A.; Illenberger, E.; Kopyra, J. Phys. Rep. 2011, 508, 1. (4) Sanz, A. G.; Fuss, M. C.; Munoz, A.; Blanco, F.; Limão-Vieira, P.; Brunger, M. J.; Buckman, S. J.; Garcia, G. Int. J. Radiat. Biol. 2012, 88, 71. (5) Winstead, C.; McKoy, V. J. Chem. Phys. 2006, 125, 244302. (6) Winstead, C.; McKoy, V. J. Chem. Phys. 2006, 125, 074302. (7) Tonzani, S.; Greene, C. H. J. Chem. Phys. 2006, 125, 094504. (8) Zubek, M.; Dampc, M.; Linert, I.; Neumann, T. J. Chem. Phys. 2011, 135, 134317. (9) Zecca, A.; Perazzolli, C.; Brunger, M. J. J. Phys. B: At. Mol. Opt. Phys. 2005, 38, 2079. (10) Sulzer, P.; Ptasinska, S.; Zappa, F.; Mielewska, B.; Milosavljevic, A. R.; Scheier, P.; Märk, T. D.; Bald, I.; Gohlke, S.; Huels, M. A.; Illenberger, E. J. Chem. Phys. 2006, 125, 44304. (11) Możejko, P.; Ptasińska-Denga, E.; Domaracka, A.; Szmytkowski, C. Phys. Rev. A 2006, 74, 012708. (12) Ibănescu, B. C.; May, O.; Allan, M. Phys. Chem. Chem. Phys. 2008, 10, 1507. (13) Ibănescu, B. C.; Allan, M. Phys. Chem. Chem. Phys. 2009, 11, 7640. (14) Ibănescu, B. C.; Allan, M. Phys. Chem. Chem. Phys. 2008, 10, 5232. (15) Dampc, M.; Szymańska, E.; Mielewska, B.; Zubek, M. J. Phys. B: At. Mol. Opt. Phys. 2011, 44, 055206. (16) Colyer, C. J.; Vizcaino, V.; Sullivan, J. P.; Brunger, M. J.; Buckman, S. J. New J. Phys. 2007, 9, 41. (17) Aflatooni, K.; Scheer, A. M.; Burrow, P. D. J. Chem. Phys. 2006, 125, 054301. (18) Lepage, M.; Letarte, S.; Michaud, M.; Motte-Tollet, F.; HubinFranskin, M.-J.; Roy, D.; Sanche, L. J. Chem. Phys. 1998, 109, 5980. (19) Jäggle, C.; Swiderek, P.; Breton, S.-P.; Michaud, M.; Sanche, L. J. Phys. Chem. B 2006, 110, 12512. (20) Breton, S.-P.; Michaud, M.; Jäggle, C.; Swiderek, P.; Sanche, L. J. Chem. Phys. 2004, 121, 11240. (21) Antic, D.; Parenteau, L.; Sanche, L. J. Phys. Chem. B 2000, 104, 4711. (22) Young, R. M.; Yandell, M. A.; Niemeyer, M.; Neumark, D. M. J. Chem. Phys. 2010, 133, 154312. (23) Park, Y. S.; Cho, H.; Parenteau, L.; Bass, A. D.; Sanche, L. J. Chem. Phys. 2006, 125, 074714. (24) Fuss, M.; Muñoz, A.; Oller, J. C.; Blanco, F.; Almeida, D.; LimãoVieira, P.; Do, T. P. D.; Brunger, M. J.; García, G. Phys. Rev. A 2009, 80, 052709. (25) Giuliani, A.; Limão-Vieira, P.; Duflot, D.; Milosavljevic, A. R.; Marinkovic, B. P.; Hoffmann, S. V.; Mason, N.; Delwiche, J.; HubinFranskin, M.-J. Eur. Phys. J. D 2008, 51, 97.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: (+351) 21 294 78 59. Fax: (+351) 21 294 85 49. 695

dx.doi.org/10.1021/jp407997w | J. Phys. Chem. A 2014, 118, 690−696

The Journal of Physical Chemistry A

Article

(26) Almeida, D.; Ferreira da Silva, F.; García, G.; Limão-Vieira, P. J. Chem. Phys. 2013, 139, 114304. (27) Almeida, D.; Ferreira da Silva, F.; García, G.; Limão-Vieira, P. Phys. Rev. Lett. 2013, 110, 023201. (28) Ferreira da Silva, F.; Almeida, D.; Antunes, R.; Martins, G.; Nunes, Y.; Eden, S.; Garcia, G.; Limão-Vieira, P. Phys. Chem. Chem. Phys. 2011, 13, 21621. (29) Almeida, D.; Antunes, R.; Martins, G.; Eden, S.; Ferreira da Silva, F.; Nunes, Y.; Garcia, G.; Limão-Vieira, P. Phys. Chem. Chem. Phys. 2011, 13, 15657. (30) Antunes, R.; Almeida, D.; Martins, G.; Mason, N. J.; Garcia, G.; Maneira, M. J. P.; Nunes, Y.; Limão-Vieira, P. Phys. Chem. Chem. Phys. 2010, 12, 12513. (31) Guler, L. P.; Yu, Y.; Kentta, H. J. Phys. Chem. A 2002, 106, 6754. (32) Ptasińska, S.; Denifl, S.; Scheier, P.; Märk, T. D. J. Chem. Phys. 2004, 120, 8505. (33) Kleyn, A. W.; Moutinho, A. M. C. J. Phys. B: At. Mol. Opt. Phys. 2001, 4075, R1. (34) Janecková, R.; May, O.; Milosavljevic, A. R.; Fedor, J. Private Communication. (35) Allan, M. J. Phys. B At. Mol. Opt. Phys. 2007, 40, 3531.

696

dx.doi.org/10.1021/jp407997w | J. Phys. Chem. A 2014, 118, 690−696