Dissociative electron attachment to gas-phase N-methylformamide

ARTICLE IN PRESS

G Model MASPEC-15061; No. of Pages 9

International Journal of Mass Spectrometry xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Mass Spectrometry journal homepage: www.elsevier.com/locate/ijms

Dissociative electron attachment to gas-phase N-methylformamide a,b,∗ ´ M. Michele Dawley a , Sylwia Ptasinska a b

Radiation Laboratory, University of Notre Dame, Notre Dame, IN 46556, USA Department of Physics, University of Notre Dame, Notre Dame, IN 46556, USA

a r t i c l e

i n f o

Article history: Received 14 November 2013 Received in revised form 3 December 2013 Accepted 4 December 2013 Available online xxx Keywords: Dissociative electron attachment Anions N-methylformamide Peptide Low energy electrons

a b s t r a c t Gas-phase dissociative electron attachment (DEA) measurements to N-methylformamide (NMF, HC(O)NHCH3 ), a simple molecule containing the peptide linkage, were performed. The anion yields as a function of electron energy below 15 eV were obtained using a crossed molecular beam/Quadrupole Mass Spectrometer (QMS) setup. DEA to NMF produces CN− , OCN− , NH2 − , O− , HCO− , and NH− , similar to formamide. Isotopic studies with N-methyl-d3 -formamide (d3 -NMF, HC(O)NHCD3 ) have confirmed and clarified assignments. Newly observed dissociation channels from NMF result in the formation of CH3 NHC(O)− /CH3 NC(O)H− , CH3 NH− , CO− , and CH3 − . The dominant fragmentation pathways involve the formation of CN− and OCN− near 1.5 eV, which are produced from the electron capture by the ␲* orbital of the C O combined with the scission of the N CH3 bond and simultaneous formation of a C N ␲ bond. The methyl neutral by-product assists in lowering the thermodynamic threshold for CN− and OCN− production as compared with the DEA to formamide. The [Parent-H]− signal at m/z 61 from d3 -NMF confirms site-selectivity for formation of this ion and excludes H loss from the methyl group of NMF. Peptide bond (O C N H) dissociation is observed resulting in both CH3 NH− and HCO− , but these channels require higher energies (>5 eV) to occur. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Dissociative electron attachment (DEA) is an electron-induced fragmentation process occurring at low energies (<15 eV) that is of interest to the fields of radiation chemistry [1], physics [2], radiation therapy [3], astrochemistry [4], and materials and surface science [5]. Specifically, high energy impact of ionizing radiation can produce secondary low-energy electrons that can induce chemical reactions or degradation of molecules via the DEA process. In DEA, a transient negative ion (TNI) forms, (AB)−# , due to the temporary trapping of an impinging electron into a LUMO or a higher empty molecular orbital state of a molecule [6]. This is a resonant scattering event; thus, the TNI is often called an anion resonance because the impinging electron energy must coincide with the TNI state, and the Frank–Condon transition controls the width of the resonance. A TNI can be classified as either a single-particle resonance at low energies (<5 eV) in which the impinging electron fills one of the unoccupied molecular orbitals (MOs) or as a core-excited resonance (5–15 eV) in which two electrons fill empty MOs. When a core-excited resonance state is above the electronically excited parent molecule, it is called an open channel resonance, but if it

∗ Corresponding author at: Radiation Laboratory, University of Notre Dame, Notre Dame, IN 46556, USA. Tel.: +1 574 631 4506. ´ E-mail addresses: [email protected], [email protected] (S. Ptasinska).

is below the excited state, then it is referred to as a closed channel or Feshbach resonance [2,7,8]. Likewise, when a single-particle resonance is above the ground state, it is referred to as a shape resonance, and when it is below the ground state, it is referred to as a nuclear-excited Feshbach resonance [8]. The TNI may then decay into a neutral fragment or molecular radical (A• ) and a negatively charged (B− ) product via the following general reaction: e− + AB → (AB)−# → B− + A• The resonance thermodynamic threshold (or onset), H(B− ), for anion formation can be given by the following formula: H(B− ) = D(A-B) − EA(B) where D(A-B) is the bond dissociation energy of AB that can be determined from the enthalpies of formation, f H◦ , of the reactants and products, and EA(B) is the electron affinity of neutral B [2]. Simple amides have gained interest in the prebiotic [9–15], radiation research [7,16], and astrophysical communities [17–19] because of their simple structure that includes the peptide bond linkage (O C N H). Formamide (HC(O)NH2 ), the simplest amide, has been the subject of many studies because of its ability to be formed from simpler molecules, including CO + NH3 [18] or H2 O + HCN [20] under both heat [21] and radiation environments [11], its ability to form larger biomolecules, including several nucleobases [22–24], and its detection in astrophysical environments [19,25]. Formamide also has an ability to intercalate into clay

1387-3806/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijms.2013.12.005

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005

G Model MASPEC-15061; No. of Pages 9 2

ARTICLE IN PRESS M.M. Dawley, S. Ptasi´ nska / International Journal of Mass Spectrometry xxx (2014) xxx–xxx

Fig. 1. Photographs of the internal (left) and external (right) view of the experimental vacuum chamber setup used in this study. The inlet gas was dosed directly into the aperture of the QMS, where the molecules (NMF and d3 -NMF) were internally ionized, and the anions were then analyzed and detected. The setup is also equipped with a low-energy electron gun that was not used for these measurements.

minerals to act as catalysts or to create new materials [9,26,27]. Finally, a study of formamide’s DEA mechanisms has recently been motivated by an interest in its ability to assist surface functionalization using electron-induced reactions [28]. N-methylformamide (HC(O)NHCH3 , hereafter denoted as NMF), another simple amide that is a derivative of formamide, has also recently gained attention because it contains the peptide bond and can serve as a model peptide. Additionally, NMF carries a methyl group, which leads to an interest in the differences in its interaction with electrons compared to formamide. Unlike formamide, NMF has a trans and a cis isomer due to the methyl group, both of which have been studied to understand the influence of the substituent on the peptide linkage [29,30]. Two recent studies by Desfranc¸ois et al. have concluded that formamide’s electron affinity decreases with the addition of the methyl group due to a molecular size effect [31,32]. Such a decrease in the electron affinity could result in changes in the formation of stable anions induced by DEA. Recently, Maljkovic´ et al. determined elastic electron scattering cross sections from NMF to understand the influence of the methyl group in the place of an H atom, and they suggested that the elastic electron scattering processes are not significantly affected by this substitution [33]. However, a recent study by Puschnigg et al. suggested that methylation causes site selective loss of H atoms from a dipeptide (dialanine), possibly altering the thresholds for anion formation [34]. Thus, the product ion yields, resonant energies, and energetic thresholds for the DEA to NMF need to be studied and compared with formamide to elucidate the effects of the methyl group. The radiation damage research community is interested in interactions of biomolecules, such as NMF, with electrons because low-energy electrons have been shown to damage DNA [2,35]. Secondary electrons of low energies (<20 eV) are primarily formed within cells during radiation cancer therapies; thus, interactions with biomolecules have become vital to understand the short- and long-term effects of such therapies. A pioneering Science article published by Boudaiffa et al. [3] in 2000 reported that secondary electron irradiation of DNA led to significant strand breakage. Since then, multiple reports on low-energy electron interactions with biomolecules have been published, including studies on amino acid–nucleotide pairs [36], amino acids [37,38], and nucleobases and nucleosides [39–41]. Additionally, electron attachment studies on model peptide molecules suggested possible denaturation

of the peptide structure [42] and dissociation of the peptide backbone [43]. Thus, an investigation of the electron interactions with NMF can contribute to radiation research of simple peptide linkages. Finally, NMF is likely one of the larger prebiotic molecules in the interstellar medium, as it has recently been identified in VUV photo-processed binary ices containing methylamine (CH3 NH2 ) and CO, both of which are interstellar molecules [44]. Therefore, interactions of model proteins, such as NMF, with low-energy electrons are also of interest to the exobiology community due to molecular interactions with radiation, including electrons in space [45]. In this investigation, we report for the first time the behavior of NMF upon electron impact below 15 eV. Anion yields as a function of electron energy, proposed fragmentation pathways, and thermodynamic thresholds for anion formation are presented. A comparison with DEA studies to formamide [28,46] also assists in understanding the influence of methylation. Our results show that precise control of the impinging electron energy does affect the resulting fragmentation pathways induced by the DEA to NMF, especially regarding peptide bond cleavage. In addition, this study broadly contributes to understanding ionizing radiation effects on amino acid, peptide, and protein structure. 2. Experimental setup The experimental data presented here was obtained using a new and recently optimized experimental high vacuum chamber (base pressure ∼1 × 10−8 mbar) that was designed and built at the Notre Dame Radiation Laboratory. Fig. 1 presents photographs of the internal (left) and external (right) view of this apparatus. Using this new system, crossed molecular beam/low-energy electron experiments can detect anions, cations, radical, and neutral species using mass spectrometric detection. The chamber is equipped with a quadrupole mass spectrometer (QMS) from Hiden Analytical, Inc. that is in the 3F Series and is commercially called an Ion Desorption Probe (IDP). The QMS is capable of gas-phase studies as well as electron, photon, and laser stimulated desorption studies, and it can be operated in both positive and negative modes to detect cations and anions, respectively. The system was used for gas-phase detection of anions in this work. Ionization can either occur via secondary ionization using an attached external Kimball Physics

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005

G Model MASPEC-15061; No. of Pages 9

ARTICLE IN PRESS M.M. Dawley, S. Ptasi´ nska / International Journal of Mass Spectrometry xxx (2014) xxx–xxx

3

Table 1 Anionic species, masses, and maximum peak positions for observed fragments formed upon DEA to N-methylformamide. The anions are given in order of largest to smallest mass. The peak position for CD3 − from d3 -NMF is also presented for comparison.

Fig. 2. Chemical structures of (a) N-methylformamide and the isotope (b) N-methyld3 -formamide investigated in this work.

electron gun (1–2000 eV) or via internal ionization using the oxidecoated iridium filament in the QMS. All results shown here were obtained with ionization using the internal QMS filament (negative Residual Gas Analyzer mode), and thus no secondary ionization occurs in the processes included in this study. Anions formed upon electron bombardment in the ion source were then detected by an ion counting detector in the QMS. Anion mass spectra were obtained at several electron energies, and then energy scans were performed from 0 to 15 eV for each observed anion (dwell time – 1 s, energy step – 0.1 eV) using MASsoft version 7 Professional software provided by Hiden. For these experiments, the following IDP-QMS settings were used: multiplier +1800 V, extractor +5 V, lens 0 V, filter energy was varied from +0.8 to 8 V, focus +10 V, and suppressor +200 V. The Hiden IDP-QMS can achieve unit resolution (1 m/z) over the entire mass range of 0–300 m/z. Using the oxide-coated iridium filament, an average electron energy resolution of ∼500 meV at an electron current of 2.0 ␮A was determined from the SF6 − signal at 0 eV. The vacuum chamber is also internally equipped with four 500 W halogen lamps for bake-out at ∼373 K prior to the experiments to reduce signals due to background gases (e.g., H2 O, CO2 , CO, etc.). In addition, a pair of Helmholtz coils affixed to the outside of the chamber created a uniform magnetic field and eliminated Earth’s magnetic field to allow electron attachment to the molecule with low energy electrons. The Helmholtz coils were spaced ∼35 cm apart from each other and were affixed around the chamber so that the magnetic field lines at the center flowed perpendicular to the entrance of the QMS. There were 50 loops per coil with each coil having a radius of ∼25 cm. Thus, the magnetic flux density, B, is calculated to be ∼7.0 G at the midpoint between the coils when set at a current of ∼4.0 A (∼22.5 V), which was the typical current used during these experiments. NMF (molecular weight: 59 g/mol, Fig. 2a) of a stated purity of 99% was purchased from Sigma–Aldrich, USA, and N-methyld3 -formamide (hereafter denoted as d3 -NMF, molecular weight: 61 g/mol, Fig. 2b) was purchased from C/D/N Isotopes, Inc., USA. The certificate of analysis for the d3 -NMF stated 99.9 atom% D purity using NMR analysis and 99.8% chemical purity using GC analysis. Isotopic d3 -NMF anion yields were measured as a function of electron energy to confirm the anion NMF assignments. Both molecules are liquids at room temperature, and NMF has a vapor pressure at ∼313 K of ∼0.95 mbar [47]. However, heating of the samples before dosing was not performed to prevent molecular decomposition in the dosing line. Although the exact thermal decomposition temperature of NMF is not available to our knowledge, the simpler formamide molecule is known to decompose to NH3 and CO with prolonged heating at 458 K, and HCN is also a decomposition product at ∼493 K [48]. Thus, the thermal decomposition of the NMF sample in the ion source of the QMS can most likely be ruled out because the temperature that was monitored with a thermocouple on the QMS stayed below ∼313 K even when the filament was operating for several hours. The samples were both freeze-pump-thawed several times to remove contaminants prior to introduction into the vacuum chamber via the dosing line. The dosing line was connected to an internal capillary in the vacuum chamber that directed the dosed gas to the aperture of the QMS.

Mass (Da)

Anion formula

Resonance positions (eV) ± 0.25 eV

58 43 42 30 29 28 27 26 18

CH3 NHC(O)− /CH3 NC(O)H− O13 CN− OCN− CH3 NH− HCO− CO− 13 CN− CN− CD3 − from d3 -NMF NH2 − O− NH− /CH3 −

2.9, 6.6 (weak shoulder), 8.6 1.7 1.7 6.0, 7.1 6.8, 10.6 7.4, 10.3 1.5 1.4 6.4 6.6, 7.9 10.5 8.4

16 15

The pressure during vapor dosing was 1 × 10−5 –8 × 10−6 mbar and was kept constant during each measurement. 3. Results and discussion Table 1 provides the anionic species produced by electron attachment to NMF below 15 eV with their maximum peak positions. Figs. 3–7 display the anion yields as a function of electron energy from 0 to 15 eV. The isotope, d3 -NMF, anion yields were also measured and are presented to clarify some assignments. At least five different anion resonant peaks were observed, and the specific positions depend upon the individual fragment with maxima ranging from 1.4 to 1.7, 2.9 to 3.4, 6.0 to 6.8, 7.1 to 8.6, and 10.3 to 10.6 eV. Following previous work on the DEA to formamide [28], this is the first study on the DEA to NMF. Much of our discussion will center around comparison with the formamide DEA study, as well as with studies of low-energy electron interactions with similar molecules such as acetamide (CH3 C(O)NH2 ) [42,49]. 3.1. CN− and OCN− The most dominant anions observed in the mass spectra are m/z 26 and m/z 42 representing CN− and OCN− , respectively, as shown in Fig. 3. In the energy scans, both CN− and OCN− display a sharp peak at 1.4–1.7 eV that is the only peak observed for these anions. CN− has the highest yield among all anions observed and is formed through an intense resonance at a maximum of ∼1.4 eV. This low energy resonance is possibly a single-particle shape resonance in which the impinging electron fills a previously half-filled or empty molecular orbital of the ground state [2,49]. However, this would need to be confirmed with molecular orbital calculations of the anionic states of NMF. For acetamide and formamide, ab initio [49] and DFT calculations [50], respectively, have described an intense low energy resonance near 2 eV that occurs via electron capture into the LUMO of the C O bond that has ␲* character. For model peptides, the process has been further described as involving subsequent scission of the N R bond (for NMF, R CH3 ) with formation of a C N ␲ bond [51]. This low energy resonance has been observed for CN− in many studies such as in the DEA to nitromethane (CH3 NO2 ) [52] and formamide [28]. Several amide derivatives have generated the resonance near 2 eV, and a previous study on acetamide and glycolamide (NH2 C(O)CH2 OH) suggested that its formation must involve the formation of new bonds (e.g., to form the H2 O molecule) that supply the energy necessary for the reaction [49]. For NMF, although multiple bonds dissociate to form CN− , the high electron affinity of CN (3.86 eV [53]) helps explain the low energy resonance.

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005

G Model MASPEC-15061; No. of Pages 9 4

ARTICLE IN PRESS M.M. Dawley, S. Ptasi´ nska / International Journal of Mass Spectrometry xxx (2014) xxx–xxx

reaction (1), which agrees with previous formamide [28], acetamide [49], and glycolamide [49] experiments. The unimolecular decomposition of NMF leading to OCN− can be explained similarly. Experimentally, we observe a low energy resonance maximum at 1.7 eV, which is close to that obtained for formamide [28] and could be interpreted as a shape resonance [49]. The estimated thermodynamic threshold for the reaction: e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → OCN− + H2 + • CH3 (or H• + CH4 )

(3)

H2 + • CH3

gives a value of 1.48 eV (for the by-products) or 1.46 eV (for the H• + CH4 by-products). For these determined thresholds, the electron affinity used for OCN was 3.61 eV [49,55], which is also relatively high and lowers the threshold significantly. In addition, the f H◦ of OCN (1.60 eV [49,56]), H2 (0 eV [53]), CH3 (1.51 eV [53]), CH4 (−0.77 eV [53]) and H (2.26 eV [53]) assist to lower the thresholds. Both of these estimates (∼1.5 eV) are close to the experimental resonance of 1.7 eV, which suggests that OCN− formation is endothermic via the (3) pathways and involves C H, N H, and N CH3 bond breaking. Similar to CN− formation, the • CH3 or CH4 by-products specifically contribute to the lower threshold for this reaction from the 2.22 eV predicted value for formamide to 1.46–1.48 eV for OCN− production induced by electron attachment to NMF. Of note, the low energy resonance of OCN− was not observed in a recent study on acetamide [49], but a resonance was observed near 1.4 eV for glycolamide [49]. Higher core-excited resonances were observed for acetamide [49], glycolamide [49], and formamide [28] in contrast to our results, which indicates that recombination of abstracted H atoms does not readily occur during formation of OCN− . For the following reactions: e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → OCN− + 2H• + • CH3 (4) e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → OCN− + 2H2 + • CH

(5)

e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → OCN− + 4H• + • CH (6)

Fig. 3. Anion signals as a function of electron energy below 15 eV for the dominant fragments, CN− and OCN− , and their naturally abundant carbon-13 isotopes, 13 CN− and O13 CN− , induced by the electron attachment to N-methylformamide.

The neutral stable by-products, H2 O and • CH3 , both likely form to give the following pathway to CN− : e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → CN− + H2 O + • CH3 (1) Using the electron affinity of CN and the f H◦ of CN (4.54 eV [53]), H2 O (−2.51 eV [53]), CH3 (1.51 eV [53]), and HC(O)NHCH3 (−1.98 eV [54]), we estimate the thermodynamic threshold to form CN− via pathway (1) is 1.66 eV, which agrees well within our experimental error. Of note, the • CH3 by-product assists in lowering this threshold from the 2.33 eV predicted value for formamide [28]. An alternative reaction forming CN− could be: e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → CN− + • OH + CH4

(2)

However, this channel is shown to be energetically less favorable (threshold at 2.29 eV) based again on the electron affinity of CN, the enthalpies given above, and the f H◦ of OH (0.40 eV [55]) and CH4 (−0.77 eV [53]). A similar reaction was considered for acetamide, but the threshold was higher (2.74 eV [49]) for formation of CN− with the by-products • OH and CH4 . Thus, we conclude that the most favorable CN− formation occurs via excision of the CN group of NMF and formation of the by-products, H2 O and • CH3 as shown in

the estimated thermodynamic thresholds are at 6.0 eV, 6.15 eV, and 15.19 eV, respectively, and are not observed in Fig. 2 as clear peaks for OCN− . A gradual increase in signal is recorded beginning at approximately 3 eV for both CN− and OCN− in our experiments. In the case of formamide, such a gradual increase in signal was also observed for CN− ; however, distinct peaks were observed for OCN− near 7 and 11 eV [28]. Also, fulminate (CNO− ) formation was considered but ruled out in the case of DEA to formamide [28]. The formation of this anion would require intramolecular rearrangement and is also precluded for NMF because the threshold is estimated at 3.23 eV using the f H◦ of CNO (3.35 eV) [52], the calculated electronic affinity of CNO (3.61 eV) [57], and considering H2 and • CH3 as the by-products. The 3.23 eV threshold is above our experimental onset of formation of m/z 42; thus for NMF, our assignment of OCN− is supported. Because of the high yields of CN− and OCN− , the carbon-13 isotopes, 13 CN− and O13 CN− , are also detectable (m/z 27 and m/z 43, respectively, Fig. 3) and have an abundance of approximately 1.5% of their carbon-12 counterparts with peak maxima that also follow those of CN− and OCN− . 3.2. CH3 NHC(O)− /CH3 NC(O)H− Fig. 4 presents the [Parent-H]− yields induced by the electron attachment to NMF and d3 -NMF. Two peaks are observed: one with a maximum near 3 eV and a broader peak with a maximum near 9 eV. The signal from ∼13 eV might be due to ion pair formation because this signal was not present in background experiments when NMF was not dosed. In addition, a shoulder near 7 eV is also slightly visible for m/z 58, but this shoulder was not reproducible in

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005

G Model MASPEC-15061; No. of Pages 9

ARTICLE IN PRESS M.M. Dawley, S. Ptasi´ nska / International Journal of Mass Spectrometry xxx (2014) xxx–xxx

Fig. 4. Anion signals as a function of electron energy below 15 eV for m/z 58 (top) and m/z 61 (bottom) induced by electron attachment to N-methylformamide and N-methyl-d3 -formamide, respectively. The resonances at m/z 61 confirm that siteselective cleavage of the C H (forming CH3 NHC(O)− ) and N H bonds (forming CH3 NC(O)H− ) can occur, while H loss from the methyl group is less likely.

every experiment and was not clearly visible in the isotopic experiments of m/z 61. The main broad peak near 9 eV is most likely due to a core-excited resonance in which two electrons fill an empty orbital to form a temporary anion [2]. In many studies, the [ParentH]− has been attributed to reactions involving site selective single bond cleavage with dehydrogenation [39,40,58]. The anion signals for m/z 58 could be due to the loss of H• from the methyl group, the amide group, or the aldehyde group of the NMF molecule. The anion signals at m/z 61 from d3 -NMF disprove the H loss from the methyl group of the NMF molecule, along with the lack of other resonances from DEA to d3 -NMF at the following masses: m/z 60 [Parent-1D or -2H]− , m/z 58 [Parent-2D]− , and m/z 56 [Parent-3D]− . Additionally, the lack of signal at m/z 59 [Parent-3H or -1D-1H]− precludes significant isotopic exchange, while the lack of signal at m/z 57 [Parent-2D-1H]− also indicates that H• loss from multiple locations (methyl+either amino or aldehyde site) does not occur. Thus, dehydrogenation of NMF is site-selective and occurs either at the aldehyde or amide sites, as described by the following reactions: e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → CH3 NHC(O)− + H•

(7)

e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → CH3 NC(O)H− + H•

(8)

The thermodynamic thresholds for H• loss from each location can be estimated based on DFT (B3LYP functional) calculations [59] of the bond dissociation energies for the homolytic scission to give H atoms at either the aldehyde (C H, 3.94 eV), amide (N H, 4.59 eV), or methyl (CH2 H, 3.91 eV) groups and based on the calculated electron affinities of the neutral radical products (0.65 eV, 2.57 eV, and 0.49 eV, respectively). From these bond dissociation energies and electron affinities, we estimate the H(B− ) to be 3.29 eV for the C H cleavage at the aldehyde site, 2.07 eV for N H cleavage at the amide site, and 3.42 eV for CH2 H cleavage at the methyl site. Both the thresholds for H• loss from the aldehyde and amide sites agree well with the experimental anion thresholds seen for both m/z 58 and m/z 61 (Fig. 4), whereas the onset for dehydrogenation from the methyl site is slightly higher. This is consistent with the incident electron energy control observed for methylated

5

thymine and uracil in which the dehydrogenation site could be precisely controlled and showed a preference for H• loss due to N H dissociation relative to C H dissociation [60]. Although dehydrogenation could only occur from the aldehyde site or the amide site in the case of formamide, interestingly, the [Parent-H]− resonances from DEA to NMF are similar to those observed for the [Parent-H]− from DEA to formamide [28] (m/z 44). We observe two resonances at similar maxima (2.9 eV vs. 2.5 eV and 8.6 eV vs. 7.0 eV for NMF vs. formamide, respectively). The peaks are broader in our case, possibly due to our lower energy resolution, but they overlap the formamide resonances significantly. Hamann et al. used isotopically labeled formamides to conclude that the two resonances observed are due to H• loss from two different sites. Their 2.5 eV resonance is due to H• loss from the amide site (N H), while the 7.0 eV resonance is due to cleavage at the aldehyde site (C H) [28], and we tentatively follow these assignments. Our 2.9 eV resonance could be due to cleavage of the N H bond in NMF, while the 8.6 eV resonance could be due to cleavage of the C H bond at the aldehyde site of NMF. Additionally, the broad resonance with a maximum near 8.6 eV could represent the capture of an impinging electron by the NMF molecule to form a coreexcited or two-electron one-hole state [2,42]. The high electron affinity calculated for N H bond cleavage (2.57 eV, [59]) explains the low threshold for formation of CH3 NC(O)H− , especially because the N H bond energy in NMF is high (4.59 eV [59]). Of note, the [Parent-H]− was not included but may have been observed in recent acetamide DEA studies [42,49], and the [Parent-H]− was observed but at a lower energy (∼2 eV) for glycolamide [49]. 3.3. CH3 NH− Fig. 5 presents the anion yield for m/z 30 (CH3 NH− ) induced by the DEA to NMF. This anion was not necessarily expected especially because of the high yields of both CN− and OCN− , indicating a lower probability of cleavage of the peptide linkage. However, two closely overlapping peaks were recorded that have maxima at 6 and 7.1 eV. The isotopic anion yields for m/z 33 (CD3 NH− ) confirms the

Fig. 5. Anion signals as a function of electron energy below 15 eV for CH3 NH− (top) and CD3 NH− (bottom) induced by electron attachment to N-methylformamide and N-methyl-d3 -formamide, respectively, confirming the assignment. The dosing pressure for the top experiment at m/z 30 (CH3 NH− ) was 1 × 10−5 mbar, while the dosing pressure for the bottom experiment at m/z 33 (CD3 NH− ) was 8.7 × 10−6 mbar.

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005

ARTICLE IN PRESS

G Model MASPEC-15061; No. of Pages 9

nska / International Journal of Mass Spectrometry xxx (2014) xxx–xxx M.M. Dawley, S. Ptasi´

6

assignment, while also displaying the two resonant peaks at the same energies. A broad weak peak near 3 eV that was not reproducible is observed in Fig. 5 for CD3 NH− . However, no such peak for m/z 30 (CH3 NH− ) is observed. For the two main peaks, a dose dependence is observed. At a dosing pressure of 1 × 10−5 mbar, the 6 eV peak had a higher relative ion yield than the 7.1 eV peak for both CH3 NH− (Fig. 5, top) and CD3 NH− (not shown). But, at a slightly lower dose (8.7 × 10−6 mbar), the ion yield of the two main peaks is below 100 c/s for m/z 33 (Fig. 5, bottom) and results in equal relative intensities for the two peaks. Based on the bond energy for C N dissociation in formamide (4.37 eV [28]) (this value is not known for NMF) and the adiabatic electron affinity of the CH3 NH radical (0.432 eV [61]), the thermodynamic threshold, H(CH3 NH− ), is estimated to occur at 3.94 eV. This is in good agreement with the experimentally determined onset of the resonance observed for CH3 NH− in Fig. 5, which is approximately at 4.0–4.5 eV. The broad double-peak resonance is likely to occur via the following reaction: e− + HCONHCH3 → (HCONHCH3 )−# → CH3 NH− + HCO• (or H• + CO)

(9)

in which either formyl radicals (HCO• ) or H• and CO neutral radicals are the by-products to the anion formation. This C N dissociation, although not dominant for electron attachment to NMF, has also been observed during photo-dissociation of NMF using 248 nm UV [30] and 193 nm excimer laser irradiation [62]. Ruzi and Anderson [62] reported that CH3 NH + HCO is one of three primary competing dissociation channels and that peptide bond scission (O C N H) can occur by exciting the n → ␲* transition at ∼5.5 eV that is localized on the peptide bond. Additionally, previous calculations for formamide have described a Feshbach resonant state near 6.10 eV [7] that could be the resonance observed at ∼6 eV for the formation of CH3 NH− , as shown in Fig. 5. 3.4. O− , NH2 − , HCO− , and CO− Although the formation of CH3 NH− from NMF cannot be compared with formation of this ion following the DEA to

formamide because formamide lacks the methyl group, peptide bond (O C N H) dissociation was induced by the DEA to formamide to form NH2 − and HCO− [28]. For NMF, we expect and observe HCO− formation (Fig. 6) but do not necessarily expect NH2 − unless NH recombines with an available H atom. However, as shown from the anion yields in Fig. 6 for m/z 16, we do certainly observe NH2 − and O− at 6.6–7.9 eV and 10.5 eV, respectively. The two anions can be distinguished from each other based on previous high resolution and isotopically labeled DEA experiments with formamide at m/z 16 [28]. The similarity of the signal shapes and positions allows us to follow the assignments for NMF. In the formamide DEA study [28], a shoulder at ∼5.6 eV and dominant signal at 6.8 eV, both representing NH2 − , are similar to the two peaks we observe at 6.6 and 7.9 eV. Shifting of these peaks may be the result of the necessary recombination with H atoms to form NH2 − from NMF, which did not occur in the case of formamide. Two Feshbach resonant states (6.64 and 6.86 eV) and one core-excited shape resonance (6.41 eV), which are similar to the 6.6 eV resonance observed in the present study, were recently calculated for formamide [7]. We suggest the following dissociation channels leading to NH2 − : e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → NH− + HCO• + • CH3 → NH2 − + CO + • CH3

(10)

e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → NH− + HCO• + • CH3 → NH2 − + HCO• + • CH2

(11)

As shown, the formation of NH2 − is likely a two-step process. The formyl radical (HCO• ) could dissociate and provide a neutral H atom that can then recombine with NH− to yield NH2 − , as shown in reaction (10). The alternative is dissociation of the methyl radical, as in reaction (11), leaving neutral HCO• intact. Using the known f H◦ of HC(O)NHCH3 (−1.98 eV [63]), NH2 (1.92 eV [53]), NH (3.64 eV [53]), CO (−1.15 eV [53]), HCO (0.45 eV [53]), CH2 (4.05 eV [53]), and CH3 (1.51 eV [53]) and the electron affinity of neutral NH2 (0.77 eV [55]), the thermodynamic thresholds, H(NH2 − ), leading to NH2 − formation via these two-step processes can be estimated at 3.49 eV for reaction (10) and 5.99 eV for reaction (11). The value

Fig. 6. Anion signals as a function of electron energy below 15 eV for m/z 16 for O− and NH2 − (left), m/z 28 for CO− (top right), and m/z 29 for HCO− (middle and bottom right) induced by electron attachment to N-methylformamide and N-methyl-d3 -formamide. The resonances observed for the anions from d3 -NMF follow those from NMF as expected. The peak marked with a “*” is due to a contributed signal from m/z 27 (13 CN− ) near 1–2 eV. The dosing pressure for m/z 29 for d3 -NMF was 8.6 × 10−6 mbar, while all others were near 1 × 10−5 mbar.

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005

G Model MASPEC-15061; No. of Pages 9

ARTICLE IN PRESS nska / International Journal of Mass Spectrometry xxx (2014) xxx–xxx M.M. Dawley, S. Ptasi´

7

for reaction (10) is lower than the onset observed in our experiments (Fig. 6, left), suggesting this pathway is more favorable. The thermodynamic stability of CO along with the low f H◦ of CH3 vs. CH2 strongly suggests that donation of H from HCO• to form NH2 − predominates as in reaction (10) with CO and • CH3 as the by-products. However, fragmentation via reaction (11) is accessible at 5.99 eV, which is also just above the onset of NH2 − formation in our experiments. These two competing pathways could be the reason for the two NH2 − resonances observed for m/z 16 from DEA to NMF at 6.6 and 7.9 eV. We note that electron-molecule scattering calculations of formamide by Goumans et al. have suggested that a fragmentation pathway leading to NH2 − + HCO (similar to reaction (11)) occurs via an indirect shape resonance at 3.7 eV and a direct high-energy resonance at 15 eV [46]; nevertheless, more detailed gas-phase scattering calculations for NMF are needed. The formation of O− at m/z 16 from DEA to NMF occurs at 10.5 eV (Fig. 6), which is in good agreement with O− formation induced by the DEA to formamide, which also occurs at ∼10 eV [28]. Contaminate production of O− can be ruled out because the O− production from DEA to CO yields a sharp peak near 9.6 eV [64], O− from DEA to CO2 yields peaks at 4.4, 8.2, and 13 eV [65], O− from DEA to O2 yields a peak at 6.5–7 eV [66], and the largest cross section of O− from H2 O occurs at ∼12 eV [67], all of which do not correspond with our observed resonance at 10.5 eV. We also performed blank control energy scans at ∼1 × 10−8 mbar at m/z 16 without NMF present in the chamber. These background scans do not show any peaks for O− or NH2 − and verify that O− is from NMF and not from background gases in the chamber, although the mechanism of O− formation from NMF cannot be sufficiently addressed without the assistance of computations. The observation of HCO− (m/z 29) as mentioned is not surprising and is shown as a function of electron energy in Fig. 6 (right), although the ion yield for this anion is relatively low suggesting that peptide bond cleavage has a low cross section. Two broad resonances with maxima at ∼6.8 and 10.6 eV are present for HCO− . This is also confirmed by d3 -NMF, in which both broad peaks are also observed. The 6.8 eV peak is very close to the Feshbach resonance calculated for formamide (6.86 eV) [7]. Some correlation in the dissociation between HCO− and O− may occur that results in similar resonances near 10.5 eV. We suggest the following dissociation channel leading to HCO− :

HCO (0.31 eV [55]), the predicted threshold of formation of HCO− is 4.06 eV. Alternatively, when the f H◦ values are tabulated for each species, the threshold is estimated at 7.27 eV. Both values are below the onset of the second 10.6 eV peak for m/z 29 shown in Fig. 6, whereas the 7.27 eV value lies near the maximum of the first peak (∼6.8 eV). The m/z 28 for CO− also displays two peaks at similar maxima (Fig. 6) at 7.4 and 10.3 eV, signifying a correlation with HCO− , O− , and CO− especially at the 10.5 eV resonance. Of note, Hamann et al. [28] reported that m/z 28 is instead HCNH− . However, we did not observe this anion (at m/z 28) in our d3 -NMF experiments, and HCND− is not expected from d3 -NMF because deuteration is only on the methyl group (not on the H bound to N). The presence of the N CH3 bond in the place of an N H bond may block this anion from forming. Thus, we conclude that CO− is induced by the electron attachment to NMF in contrast to the reported results on formamide [28].

e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → HCO− + NH• + • CH3

For the formation of NH− , two C N bonds in NMF must dissociate. The thermodynamic threshold for formation of NH− can be estimated at 7.20 eV given the previously mentioned f H◦ of each species involved and the electron affinity of NH (0.38 eV [55]. This 7.20 eV threshold is near the onset shown for m/z 15 for NH− from DEA to d3 -NMF (Fig. 7, right). However, the onset of this resonance

(12)

From the bond dissociation energy of 4.37 eV for C N dissociation reported for formamide [28] and the electron affinity of

3.5. NH− and CH3 − Finally, Fig. 7 displays the anion yields at m/z 15 and 18 for NMF (left) and d3 -NMF (right). Following the DEA to NMF, the observed resonance near 8.4 eV for m/z 15 could represent either NH− or CH3 − . However, in the case of d3 -NMF, we again observed a broad peak at m/z 15 with a maximum shifted near 9 eV that could only represent NH− due to the deuteration of the methyl group, and a resonance was also observed at 6.4 eV for m/z 18 that could only represent CD3 − . As a control, m/z 18 was monitored and showed no signal both without a sample and when NMF was dosed. The 6.4 eV peak could represent a core-excited shape resonance, based on recent formamide electron scattering calculations [7]. We conclude that both NH− and CH3 − are formed from DEA to NMF, based upon the isotopic experiments, and that the broad resonance for m/z 15 is a combination of both fragments. The following dissociation channels could explain the occurrence of these anions: e− + HC(O)NHCH3 → (HC(O)NHCH3 )−# → NH− + HCO• + • CH3 (13) −

−#

e + HC(O)NHCH3 → (HC(O)NHCH3 )

→ CH3 − + HCONH• (or HCO• + NH• )

(14)

Fig. 7. Anion signals as a function of electron energy below 15 eV for m/z 15 (black) and m/z 18 (red) induced by electron attachment to N-methylformamide (left) and N-methyl-d3 -formamide (right). The m/z 15 signal from NMF is a combination of NH− and CH3 − as shown by the d3 -NMF m/z 15 shift to higher energies (representing NH− ) and the m/z 18 (CD3 − ) peak for d3 -NMF that appears at ∼6.4 eV. The dosing pressure for m/z 15 and 18 for d3 -NMF was ∼8 × 10−6 mbar, while those for NMF were near 1 × 10−5 mbar.

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005

G Model MASPEC-15061; No. of Pages 9 8

ARTICLE IN PRESS M.M. Dawley, S. Ptasi´ nska / International Journal of Mass Spectrometry xxx (2014) xxx–xxx

is lower for m/z 15 following DEA to NMF (Fig. 7, left), possibly due to the contribution from CH3 − . The H(CH3 − ) with the HCONH• byproduct (reaction (14)) unfortunately cannot be estimated because the f H◦ of HCONH is not currently available in the literature. However, it is possible that this neutral product lowers the threshold for CH3 − formation. If instead the neutral by-products are HCO• and NH• , the H(CH3 − ) is estimated at 7.50 eV using the electron affinity of CH3 (0.08 eV [55]). This additional bond breaking appears unlikely because the estimated value is higher than the maximum (6.4 eV) observed in the CD3 − experiment for d3 -NMF (Fig. 7, right). However, recent electron stimulated desorption studies on acetamide report an energy maximum for CH3 − near 9 eV [42,43], which is both above our observed maximum (6.4 eV) and our estimated threshold (7.50 eV) for CH3 − formation induced by the electron attachment to NMF. 4. Conclusions We have measured the anion yields resulting from the DEA to Nmethylformamide and N-methyl-d3 -formamide using a QMS with an internal electron source producing an electron beam crossed with a molecular beam of gas-phase NMF at room temperature. The anion yields were measured as a function of electron energy from 0 to 15 eV. The experimentally observed resonances and proposed reaction channels were explained in terms of newly estimated thermodynamic thresholds for their formation. Some products are similar to those induced by the DEA to formamide, including CN− , OCN− , NH2 − , O− , HCO− , and NH− . However, new products observed for NMF are CH3 NHC(O)− /CH3 NC(O)H− , CH3 NH− , CO− , and CH3 − . Our results suggest that peptide bond cleavage is a minor dissociation channel compared to the formation of CN− and OCN− . In addition, the presence of the methyl group definitely affects some thermodynamic thresholds, and site selective dehydrogenation was observed. CO− was assigned here in contrast to the assignment of m/z 28 to HCNH− in the formamide DEA study [28]. Also, NH2 − production most likely resulted from a two-step process involving H abstraction from the HCO radical to form a neutral CO by-product. However, the fragmentation channels, nature of the anion resonances, and neutral by-products could be further elucidated with the much needed assistance of computations and experimental measurements of the radical neutrals. We also note briefly that other anions (H− and HCONH− ) were observed in our experiments, but the resonances could not be confirmed (reproduced sufficiently or confirmed with d3 -NMF), so they are not reported in this study. This first DEA to NMF investigation serves as a starting point for DEA studies of larger molecules, particularly longer chain molecules containing the peptide bond. One eventual aim is the understanding of the DEA to larger peptides and proteins, which is certainly valuable to many fields, specifically radiation therapy research. Acknowledgments The research described herein was supported by the Division of Chemical Sciences, Geosciences and Biosciences, Basic Energy Sciences, Office of Science, United States Department of Energy (DOE) through Grant No. DE-FC02-04ER15533. This is Radiation Laboratory contribution number NDRL-4992. The authors wish to thank Benjamin Puschnigg, a visiting Ph.D. student from the University of Innsbruck, Austria (advised by Paul Scheier) for his assistance with the design and construction of the Helmholtz coils for the instrument at NDRL described in this work. We also acknowledge William Alexander Cantrell, an undergraduate researcher at NDRL, for his assistance with building and testing the chamber and instrumentation used in this work.

References [1] J. Staehler, C. Gahl, M. Wolf, Dynamics and reactivity of trapped electrons on supported ice crystallites, Accounts of Chemical Research 45 (2012) 131–138. [2] C.R. Arumainayagam, H.L. Lee, R.B. Nelson, D.R. Haines, R.P. Gunawardane, Low-energy electron-induced reactions in condensed matter, Surface Science Reports 65 (2010) 1–44. [3] B. Boudaiffa, P. Cloutier, D. Hunting, M.A. Huels, L. Sanche, Resonant formation of DNA strand breaks by low-energy (3 to 20 eV) electrons, Science 287 (2000) 1658–1660. [4] F. Carelli, M. Satta, T. Grassi, F.A. Gianturco, Carbon-rich molecular chains in protoplanetary and planetary atmospheres: quantum mechanisms and electron attachment rates for anion formation, Astrophysical Journal 774 (2013). [5] A.A. Tuinman, A.S. Lahamer, R.N. Compton, Studies of low-energy electron attachment at surfaces, International Journal of Mass Spectrometry 205 (2001) 309–323. [6] I. Bald, J. Langer, P. Tegeder, O. Ingolfsson, From isolated molecules through clusters and condensates to the building blocks of life: a short tribute to Prof. Eugen Illenberger’s work in the field of negative ion chemistry, International Journal of Mass Spectrometry 277 (2008) 4–25. [7] Y.F. Wang, S.X. Tian, Low-energy electron collisions with formamide using the R-matrix method, Physical Review A 85 (2012). [8] E. Illenberger, J. Momigny, Gaseous Molecular Ions: An Introduction to Elementary Processes Induced by Ionization (Topics in Physical Chemistry), Springer, New York, 1992. [9] M.M. Dawley, A.M. Scott, F.C. Hill, J. Leszczynski, T.M. Orlando, Adsorption of formamide on kaolinite surfaces: a combined infrared experimental and theoretical study, Journal of Physical Chemistry C 116 (2012) 23981–23991. [10] S.N. Vinh, H.L. Abbott, M.M. Dawley, T.M. Orlando, J. Leszczynski, T.N. Minh, Theoretical study of formamide decomposition pathways, Journal of Physical Chemistry A 115 (2011) 841–851. [11] H.L. Barks, R. Buckley, G.A. Grieves, E. Di Mauro, N.V. Hud, T.M. Orlando, Guanine, adenine, and hypoxanthine production in UV-irradiated formamide solutions: relaxation of the requirements for prebiotic purine nucleobase formation, ChemBioChem 11 (2010) 1240–1243. [12] R. Saladino, G. Botta, S. Pino, G. Costanzo, E. Di Mauro, Materials for the onset. A story of necessity and chance, Frontiers in Bioscience (Landmark) 18 (2013) 1276–1290. [13] V.S. Nguyen, T.M. Orlando, J. Leszczynski, M.T. Nguyen, Theoretical Study of the decomposition of formamide in the presence of water molecules, Journal of Physical Chemistry A 117 (2013) 2543–2555. [14] R.V. Bennett, H.J. Cleaves, J.M. Davis, D.A. Sokolov, T.M. Orlando, J.L. Bada, F.M. Fernandez, Desorption electrospray ionization imaging mass spectrometry as a tool for investigating model prebiotic reactions on mineral surfaces, Analytical Chemistry 85 (2013) 1276–1279. [15] R. Saladino, C. Crestini, S. Pino, G. Costanzo, E. Di Mauro, Formamide and the origin of life, Physics of Life Reviews 9 (2012) 84–104. [16] H. Iijima, H.B. Patrzyc, E.E. Budzinski, J.B. Dawidzik, H.G. Freund, H.C. Box, The 1-carbamoyl-2-oxo-4,5-dihydroxyimidazolidine component of ROS-induced DNA damage in white blood cells, Radiation Research 174 (2010) 101–106. [17] B. Sivaraman, B.N.R. Sekhar, N.C. Jones, S.V. Hoffmann, N.J. Mason, VUV spectroscopy of formamide ices, Chemical Physics Letters 554 (2012) 57–59. [18] B.M. Jones, C.J. Bennett, R.I. Kaiser, Mechanistical studies on the production of formamide (H2 NCHO) within interstellar ice analogs, Astrophysical Journal 734 (2011). [19] D.T. Halfen, V. Ilyushin, L.M. Ziurys, Formation of peptide bonds in space: a comprehensive study of formamide and acetamide in Sgr B2(N), Astrophysical Journal 743 (2011). [20] P.A. Gerakines, M.H. Moore, R.L. Hudson, Ultraviolet photolysis and proton irradiation of astrophysical ice analogs containing hydrogen cyanide, Icarus 170 (2004) 202–213. [21] R. Saladino, G. Botta, S. Pino, G. Costanzo, E. Di Mauro, From the one-carbon amide formamide to RNA all the steps are prebiotically possible, Biochimie 94 (2012) 1451–1456. [22] J. Wang, J.D. Gu, M.T. Nguyen, G. Springsteen, J. Leszczynski, From formamide to purine: a self-catalyzed reaction pathway provides a feasible mechanism for the entire process, Journal of Physical Chemistry B 117 (2013) 9333–9342. [23] M. Ferus, S. Civis, A. Mladek, J. Sponer, L. Juha, J.E. Sponer, On the road from formamide ices to nucleobases: IR-spectroscopic observation of a direct reaction between cyano radicals and formamide in a high-energy impact event, Journal of the American Chemical Society 134 (2012) 20788–20796. [24] J.S. Hudson, J.F. Eberle, R.H. Vachhani, L.C. Rogers, J.H. Wade, R. Krishnamurthy, G. Springsteen, A unified mechanism for abiotic adenine and purine synthesis in formamide, Angewandte Chemie-International Edition 51 (2012) 5134–5137. [25] G.R. Adande, N.J. Woolf, L.M. Ziurys, Observations of interstellar formamide: availability of a prebiotic precursor in the galactic habitable zone, Astrobiology 13 (2013) 439–453. [26] K.H. Song, X. Wang, P. Qian, C. Zhang, Q. Zhang, Theoretical study of interaction of formamide with kaolinite, Computational and Theoretical Chemistry 1020 (2013) 72–80. [27] A.M. Scott, M.M. Dawley, T.M. Orlando, F.C. Hill, J. Leszczynski, Theoretical study of the roles of Na+ and water on the adsorption of formamide on kaolinite surfaces, Journal of Physical Chemistry C 116 (2012) 23992–24005. [28] T. Hamann, A. Edtbauer, F.F. da Silva, S. Denifl, P. Scheier, P. Swiderek, Dissociative electron attachment to gas-phase formamide, Physical Chemistry Chemical Physics 13 (2011) 12305–12313.

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005

G Model MASPEC-15061; No. of Pages 9

ARTICLE IN PRESS M.M. Dawley, S. Ptasi´ nska / International Journal of Mass Spectrometry xxx (2014) xxx–xxx

[29] Y. Kawashima, T. Usami, N. Ohashi, R.D. Suenram, J.T. Hougen, E. Hirota, Dynamical structure of peptide molecules, Accounts of Chemical Research 39 (2006) 216–220. [30] R. Crespo-Otero, A. Mardyukov, E. Sanchez-Garcia, M. Barbatti, W. Sander, Photochemistry of N-methylformamide: matrix isolation and nonadiabatic dynamics, ChemPhysChem 14 (2013) 827–836. [31] C. Desfrancois, V. Periquet, S. Carles, J.P. Schermann, L. Adamowicz, Neutral and negatively-charged formamide, N-methylformamide and dimethylformamide clusters, Chemical Physics 239 (1998) 475–483. [32] C. Desfrancois, V. Periquet, S. Carles, J.P. Schermann, D.M.A. Smith, L. Adamowicz, Experimental and ab initio theoretical studies of electron binding to formamide, N-methylformamide, and N,N-dimethylformamide, Journal of Chemical Physics 110 (1999) 4309–4314. ´ F. Blanco, G. García, B.P. Marinkovic, ´ A.R. Milosavljevic, ´ Absolute [33] J.B. Maljkovic, cross sections for elastic electron scattering from methylformamide, Physical Review A 85 (2012) 042723. [34] B. Puschnigg, S.E. Huber, M. Probst, K. Tanzer, V. Vizcaino, F.F. da Silva, P. Scheier, P. Limao-Vieira, S. Denifl, Electron attachment to the dipeptide dialanine: influence of methylation on site selective dissociation reactions, Physical Chemistry Chemical Physics 15 (2013) 3834–3840. [35] E. Alizadeh, A.G. Sanz, G. Garcia, L. Sanche, Radiation damage to DNA: the indirect effect of low-energy electrons, Journal of Physical Chemistry Letters 4 (2013) 820–825. [36] S. Ptasinska, Z.J. Li, N.J. Mason, L. Sanche, Damage to amino acid-nucleotide pairs induced by 1 eV electrons, Physical Chemistry Chemical Physics 12 (2010) 9367–9372. [37] P. Papp, P. Shchukin, S. Matejcik, Specific formation of negative ions from leucine and isoleucine molecules, Journal of Chemical Physics 132 (2010). [38] J. Kocisek, P. Papp, P. Mach, Y.V. Vasil’ev, M.L. Deinzer, S. Matejcik, Resonance electron capture by serine, Journal of Physical Chemistry A 114 (2010) 1677–1683. [39] S. Ptasinska, S. Denifl, P. Scheier, E. Illenberger, T.D. Mark, Bond- and siteselective loss of H atoms from nucleobases by very-low-energy electrons (<3 eV), Angewandte Chemie-International Edition 44 (2005) 6941–6943. [40] S. Ptasinska, S. Denifl, B. Mroz, M. Probst, V. Grill, E. Illenberger, P. Scheier, T.D. Mark, Bond selective dissociative electron attachment to thymine, Journal of Chemical Physics 123 (2005). [41] S. Ptasinska, S. Denifl, S. Gohlke, P. Scheier, E. Illenberger, T.D. Mark, Decomposition of thymidine by low-energy electrons: implications for the molecular mechanisms of single-strand breaks in DNA, Angewandte ChemieInternational Edition 45 (2006) 1893–1896. [42] H. Abdoul-Carime, S. Cecchini, L. Sanche, Alteration of protein structure induced by low-energy (<18 eV) electrons. I. The peptide and disulfide bridges, Radiation Research 158 (2002) 23–31. [43] P. Cloutier, C. Sicard-Roselli, E. Escher, L. Sanche, Low-energy (3–24 eV) electron damage to the peptide backbone, Journal of Physical Chemistry B 111 (2007) 1620–1624. [44] J.B. Bossa, F. Borget, F. Duvernay, G. Danger, P. Theule, T. Chiavassa, Solid-state methylamine VUV Irradiation study using carbon monoxide as an H radical scavenger, Australian Journal of Chemistry 65 (2012) 129–137. [45] M. Padovani, D. Galli, A.E. Glassgold, Cosmic-ray ionization of molecular clouds, Astronomy & Astrophysics 501 (2009) 619–631. [46] T.P.M. Goumans, F.A. Gianturco, F. Sebastianelli, I. Baccarelli, J.L. Rivail, Dissociative electron attachment to formamide: direct and indirect pathways from resonant intermediates, Journal of Chemical Theory and Computation 5 (2009) 217–221. [47] J. Zielkiewicz, (Vapour plus liquid) equilibria in (N-methylformamide plus methanol plus water) at the temperature 313.15 K, Journal of Chemical Thermodynamics 28 (1996) 887–894.

9

[48] F. Cataldo, G. Patane, G. Compagnini, Synthesis of HCN polymer from thermal decomposition of formamide, Journal of Macromolecular Science Part A: Pure and Applied Chemistry 46 (2009) 1039–1048. [49] C. Koenig-Lehmann, J. Kopyra, I. Dabkowska, J. Kocisek, E. Illenberger, Excision of CN(−) and OCN(−) from acetamide and some amide derivatives triggered by low energy electrons, Physical Chemistry Chemical Physics 10 (2008) 6954–6961. [50] M. Seydou, A. Modelli, B. Lucas, K. Konate, C. Desfrancois, J.P. Schermann, Electron attachment to strongly polar clusters – formamide molecule and clusters, European Physical Journal D 35 (2005) 199–205. [51] M. Sobczyk, W. Anusiewicz, J. Berdys-Kochanska, A. Sawicka, P. Skurski, J. Simons, Coulomb-assisted dissociative electron attachment: application to a model peptide, Journal of Physical Chemistry A 109 (2005) 250–258. [52] W. Sailer, A. Pelc, S. Matejcik, E. Illenberger, P. Scheier, T.D. Mark, Dissociative electron attachment study to nitromethane, Journal of Chemical Physics 117 (2002) 7989–7994. [53] D.R. Lide, CRC Handbook of Chemistry and Physics, Internet Version 2013, CRC Press, Boca Raton, FL, 2013 http://www.hbcpnetbase.com [54] I.I. Marochkin, O.V. Dorofeeva, Amide bond dissociation enthalpies: effect of substitution on N C bond strength, Computational and Theoretical Chemistry 991 (2012) 182–191. [55] P.J. Linstrom, W.G. Mallard, NIST Chemistry WebBook, NIST Standard Reference Database Number 69, National Institute of Standards and Technology, Gaithersburg, MD, 2013 http://webbook.nist.gov [56] H. Okabe, Photodissociation of HNCO in vacuum ultraviolet – production of NCO and NH, Journal of Chemical Physics 53 (1970) 3507–3515. [57] C. Leonard, H. Gritli, G. Chambaud, New study of the stability and of the spectroscopy of the molecular anions NCO− and CNO− , Journal of Chemical Physics 133 (2010). [58] F.F. da Silva, C. Matias, D. Almeida, G. Garcia, O. Ingolfsson, H.D. Flosadottir, B. Omarsson, S. Ptasinska, B. Puschnigg, P. Scheier, P. Limao-Vieira, S. Denifl, NCO(−), a key fragment upon dissociative electron attachment and electron transfer to pyrimidine bases: site selectivity for a slow decay process, Journal of the American Society for Mass Spectrometry 24 (2013) 1787–1797. [59] I. Carmichael, Unpublished DFT calculations using the B3LYP functional of the trans isomer of NMF, Personal communication. [60] S. Ptasinska, S. Denifl, V. Grill, T.D. Mark, E. Illenberger, P. Scheier, Bond- and site-selective loss of H− from pyrimidine bases, Physical Review Letters 95 (2005). [61] D. Radisic, S.J. Xu, K.H. Bowen, Photoelectron spectroscopy of the anions, CH3 NH− and (CH3 )2 N− and the anion complexes, H− (CH3 NH2 ) and (CH3 )2 N− [(CH3 )2 NH)], Chemical Physics Letters 354 (2002) 9–13. [62] M. Ruzi, D.T. Anderson, Photodissociation of N-methylformamide isolated in solid parahydrogen, Journal of Chemical Physics 137 (2012). [63] V.S. Ushakov, S.M. Sedov, B.A. Knyazev, B.I. Kuchkaev, The thermodynamic properties of N-methylformamide, Zhurnal Fizicheskoi Khimii 70 (1996) 1573–1577. [64] P. Cicman, G. Senn, G. Denifl, D. Muigg, J.D. Skalny, P. Lukac, A. Stamatovic, T.D. Mark, Dissociative electron attachment to CO2 , Czechoslovak Journal of Physics 48 (1998) 1135–1145. [65] O.J. Orient, S.K. Srivastava, Production of O− from CO2 by dissociative electronattachment, Chemical Physics Letters 96 (1983) 681–684. [66] R.J. van Brunt, L.J. Kieffer, Angular distribution of O− from dissociative electron attachment to O2 , Physical Review A 2 (1970) 1899–1905. [67] P. Rawat, V.S. Prabhudesai, G. Aravind, M.A. Rahman, E. Krishnakumar, Absolute cross sections for dissociative electron attachment to H2 O and D2 O, Journal of Physics B: Atomic Molecular and Optical Physics 40 (2007) 4625–4636.

´ Please cite this article in press as: M.M. Dawley, S. Ptasinska, Int. J. Mass Spectrom. (2014), http://dx.doi.org/10.1016/j.ijms.2013.12.005