- Email: [email protected]
ionized aniline molecule studied by electron impact
International Journal of Mass Spectrometry 351 (2013) 56–60
Contents lists available at ScienceDirect
International Journal of Mass Spectrometry journal homepage: www.elsevier.com/locate/ijms
Specific fragmentation of the K-shell excited/ionized aniline molecule studied by electron impact Kazumasa Okada ∗ , Yasuhiko Kashiwagi, Masamichi Sakai Department of Chemistry, Graduate School of Science, Hiroshima University, Higashi-Hiroshima 739-8526, Japan
a r t i c l e
i n f o
Article history: Received 7 February 2013 Received in revised form 26 July 2013 Accepted 26 July 2013 Available online 3 August 2013 Keywords: Site-specific fragmentation Inner-shell excitation Electron impact
a b s t r a c t Fragmentation of aniline upon K-shell excitation/ionization has been studied by the electron impact method. Ions detected as dominant fragments were C3 H3 + , CHNH+ /CNH2 + , C3 H2 + , etc. The C3 H3 + ion was produced characteristically at the N 1s edge. We found from the mass spectra of aniline-2,3,4,5,6d5 that CD3 + is formed independent of excitation energy, suggesting that the aniline molecular cation isomerizes prior to dissociation. The mass spectra of the fragment ions by the direct dissociation of aniline were deduced by subtracting the contribution from the dissociation of the isomer, 4-methylpyridine. Enhancement of the C3 H3 + , CHNH+ /CNH2 + , and NH+ yields at the N 1s edge shows that the cleavage of the C1 C2 bond of the initial benzene ring and the C N bond fission are likely to take place by the N 1s excitation and ionization. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Excitation of molecules at the K-shell level has attracted much attention for the last three decades, because site-specific fragmentation processes were observed in various molecular systems [1–15]. Inner-shell excited or ionized molecules dissociate into fragment ions following Auger electron emission. If the Auger decay process depends highly on the core-hole states, there is a high probability of producing fragment ions distinctive to the particular atomic site of excitation or ionization [16–18]. Recently, we have reported the site-specific fragmentation of the 2-, 3- and 4-methylpyridine molecules following the C 1s and N 1s excitation/ionization by electron impact [15]. Specific production of the C2 HN+ and C5 H5 + /C5 H6 + ions at the N K-edge reveals that the N C2 and C4 C5/C5 C6 bond breakings of the pyridine ring is likely to occur following the N 1s excitation and ionization. Ab initio molecular orbital calculations indicate that the site specificity observed can be reasonably understood by the dissociation of the molecular dications in the most probable Auger final states. This article reports the fragmentation of a structural isomer, aniline, with a nitrogen atom in the substituent group, induced by electron impact. Replacement of different positions in a target molecule with a heteroatom by changing molecules enables us to explore in more general the site-specific behavior of fragmentations of molecules. The main interest here lies in investigating if the site-specific fragmentations are observed in these molecular
∗ Corresponding author. Tel.: +81 82 424 7102; fax: +81 82 424 0727. E-mail address: [email protected] (K. Okada). 1387-3806/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijms.2013.07.013
systems and whether the bonds dissociate around the excitation site. To our knowledge, limited experimental studies have been reported on the spectroscopy of aniline in the K-shell region. Ohta and coworkers [19] observed the gas-phase X-ray photoelectron spectra of several mono-substituted derivatives of benzene including aniline. Oscillator strengths for the inner-shell excitation of aniline were reported by Hitchcock’s group using inner-shell electron energy-loss spectroscopy (ISEELS) [20]. About a decade later, Duflot et al. [21] measured the spectra with higher resolution and gave a reliable assignment of the bands with the aid of ab initio configuration interaction calculations. The X-ray photoabsorption of condensed aniline was published by Luo et al. [22]. However, the mass spectrometric study on the fragmentation of aniline in the K-shell region has not been reported so far. The article is organized as follows. Next section describes our experimental procedure. We then present the main results followed by the discussion. The mass spectra indicate that the C3 H3 + ion is produced characteristically at the N 1s edge, showing that the site-specific fragmentation occurs at this edge. The presence of the CD3 + peak for the aniline-2,3,4,5,6-d5 sample suggests that the ring permutation leading to the 4-methylpyridine cation is involved in the dissociation of the aniline cation. Finally, we summarize the conclusions of this study. 2. Experimental The experiments were performed in an electron-impact reaction chamber equipped with a quadrupole mass spectrometer (Hiden Analytical, HAL-4·EQS-300). Since the experimental setup and
K. Okada et al. / International Journal of Mass Spectrometry 351 (2013) 56–60
57
procedure have been described in detail previously [14], only a brief outline is given here. The electron beam for excitation was prepared by an electron gun (Yamamoto Shinku Lab., EG-50), which can emit electrons with the energy of 200–2000 eV. The energy resolution is stated as 0.7 eV. The electron beam current was monitored with a Faraday cup mounted downstream on the electron beam axis. An electrostatic potential of −50 V was applied at the extractor plate of the spectrometer in order to extract and detect fragment ions produced at the interaction region. The pressure in the chamber was kept at about 4 × 10−6 Torr (1 Torr = 133.2 Pa) during the measurements in which the effusive beam of sample gas was introduced. The measurements were also conducted for the sample where all the hydrogen atoms in the benzene ring have been replaced by deuterium (i.e., aniline-2,3,4,5,6-d5 ) in order to give unambiguous assignment among isobaric fragments for some mass-spectral peaks. Mass spectra of ions produced by the K-shell excitation/ionization were acquired by a spectral difference method: the mass spectra were first measured above the K-shell ionization threshold and then below the K-edge. Subtraction of the spectra between the above and below K-edges gives the mass spectrum of fragment ions produced at the energies of interest. The samples of aniline and aniline-2,3,4,5,6-d5 (minimum isotopic purity of 98% D) were obtained commercially (Aldrich Chemical Co., Inc.) and were carefully degassed under vacuum by repetitive freeze–pump–thaw cycles.
3. Results Mass spectra of fragment ions of aniline and aniline-2,3,4,5,6-d5 are displayed in Figs. 1 and 2, respectively. For each molecule the mass spectrum at the N 1s edge was obtained by subtracting the raw spectrum at 390 eV from the one at the 1000-eV electron energy, both of which had been normalized with electron beam intensities. On subtracting, efficiencies of ionization were taken into account by using the electron-impact ionization cross-sections. Since the cross-section data for aniline are not available, we assumed that they are approximated as those for benzene [23]. The spectrum at the C 1s edge was obtained in a similar way from the data at 390 eV and 270 eV. The spectrum for the valence ionization was acquired at the electron energy of 270 eV. The N 1s and C 1s ionization thresholds of aniline are 405.3 eV and 289.7–291.2 eV, respectively [19]. One finds from the mass spectra of normal aniline (Fig. 1) that the dominant fragments are, in order of abundance, the ions with m/e = 39, 28, 38, 27, 26, etc. Closer look at the spectra reveals the increase of the peak at m/e = 39 at the N 1s edge (Fig. 1a) compared with the data for the valence ionization (Fig. 1c). The prominent peaks in the mass spectra of deuterated aniline (Fig. 2) are m/e = 42, 28, 40, 29, and 30. Assignment of these peaks to fragment ions can be given from these spectral patterns. Let us begin with the peaks at m/e = 38 and 39. It is apparent that the trend in the increasing yield with m/e in 36–39 in Fig. 1 corresponds to that of m/e = 36, 38, 40, and 42 in Fig. 2. The peaks at m/e = 38 and 39 can thus be assigned mostly to C3 H2 + and C3 H3 + , respectively. Assignment of the peaks in m/e = 26–28 is, however, not as obvious as that of the peaks at m/e = 38 and 39. We employed a multiple linear regression (MLR) analysis [24] for the assignment of m/e = 26–28. Based on the fact that there are only trace amounts of CH+ in the fragmentation of deuterated aniline (Fig. 2), virtually no H/D exchange is presumed to occur: for aniline-2,3,4,5,6-d5 no D atoms make bonds with N throughout the process. We also assume that isotope effect is not crucial for the ion yields. For example, the peak at m/e = 27 in the mass spectra of normal aniline consists of CNH+ , CHN+ , and C2 H3 + , where the
Fig. 1. Mass spectra of aniline acquired at the N 1s and C 1s edges (panels a and b), and at the valence ionization (panel c). The spectra (a) and (b) were obtained by the difference between the above and below K-edges.
notation CHm NHn + represents that there are m hydrogen atoms bonded with carbon and n hydrogen atoms bonded with a nitrogen atom. That is, the yields of CNH+ are equivalent between normal and isotope-substituted anilines, and the intensity of CHN+ for normal aniline is the same as the CDN+ component for deuterated aniline. Such conditions formulate a set of simultaneous equations with the yields containing experimental errors. The solution for the N K-edge data indicates that C2 H2 + comprises 85% of the intensity at m/e = 26 and the peak m/e = 28 contains 66% CHNH+ and 21% CNH2 + . The intensity at m/e = 27 is not attributed to a particular fragment ion, consisting of CNH+ , CHN+ , and C2 H3 + in the ratio 34:25:41. Although the number of significant figures is one according to the statistical analysis, two digits are given here for the purpose of later discussion. Also, we are aware that the assumptions introduced above may bring errors in the MLR analysis due to the isomerization described in the next section, and such errors may be significant for the higher-mass peaks for which we did not employ the MLR analysis. The effect of the K-shell excitation/ionization on fragmentation can be clarified by an enhancement of the yield of the fragment ions relative to the mass spectral pattern obtained at the valence ionization. Enhancement of the yield of the fragment ions produced by the N 1s and C 1s excitation/ionization is shown in Fig. 3. The term “enhancement” here was defined in our previous studies [14,15]. That is, the spectrum of Fig. 3a is, for example, the result of the subtraction of Fig. 1c from Fig. 1a. Positive and negative values (as percentage points) mean an increase and decrease of the relative yields, respectively, compared with the yields for the valence
58
K. Okada et al. / International Journal of Mass Spectrometry 351 (2013) 56–60
Fig. 2. Mass spectra of aniline-2,3,4,5,6-d5 acquired at the N 1s and C 1s edges (panels a and b), and at the valence ionization (panel c).
ionization. One can inspect a promotion effect by the inner-shell excitation/ionization on the fragmentation, using the mass spectrum for the valence ionization as a reference. Ion yields were calculated from the peak areas of the mass spectra, with totals normalized to 100. The distinct production of a fragment ion, C3 H3 + , was observed at the N 1s edge (Fig. 3a). Only a weak enhancement was found of the peak at m/e = 27 for the C 1s edge data (Fig. 3b), which has a similar trend to the methylpyridine cases at the C 1s edge [15]. Therefore, in the next section we shall deal exclusively with the N 1s edge data. 4. Discussion Fragmentation of aniline is not as simple as it is expected. One can see a peak at m/e = 18 in Fig. 2, which indicates the formation of CD3 + being not produced by the direct dissociation of aniline. The CD3 + production is independent of the excitation energy. A clue to solve this mysterious peak can be found in a 193-nm laser photodissociation study of aniline reported by Tseng et al. [25]. They claimed that at least 23% of (neutral) aniline in the ground electronic state rearranges to a seven-membered ring. The ring can then rearomatize to yield 4-methylpyridine prior to dissociation. Such ring permutation is also presumed to occur in the aniline cation, as demonstrated by Choe et al. in a theoretical study on the reaction pathways for an H loss [26]. For the best of our knowledge, no experimental study has been reported so far on the isomerization and fragmentation of the aniline cation. Therefore, we assume in the following analysis that the dissociation of the 4-methylpyridine ion is contributed to 23% of the total ion yield for each spectrum.
Fig. 3. Enhancement of the yield of fragment ions produced by the N 1s and C 1s excitation/ionization of aniline, relative to the yield of ions at the valence ionization. The enhancement is defined by the difference from the valence ionization spectrum and depicted as a bar chart.
Mass spectra of fragment ions produced by the direct dissociation of aniline were obtained by subtracting the mass spectra of 4-methylpyridine [15] from the spectra given in Fig. 1. We cannot evaluate the extent of the contribution from 4-methylpyridine in this study, but we conclude that the value of 23% is reasonable because the peak at m/e = 15 gave a negative value for the spectra of the C 1s edge and valence ionization if the value >30% was used in the subtraction. Fig. 4 displays the mass spectra obtained by such evaluation. The mass spectra have been represented in terms of peak-area intensity. The white and gray bars indicate the intensities before and after subtracting the contribution from the 4-methylpyridine dissociation, respectively: the white bars (including the hidden area overlapped with gray bars) correspond to the mass spectra shown in Fig. 1. At the N 1s edge prominent peaks depicted as gray bars are m/e = 39, 38, 28, and 27. The peak at m/e = 15 is now identified as NH+ . The intensity of m/e = 39 is not so much affected, because this peak for 4-methylpyridine corresponds to C2 HN+ [15]. The intensity of m/e 28 becomes 56% after the subtraction, and the peak still has a contribution from both CHNH+ and CNH2 + : the peak counts up to 22% CHNH+ after subtracting the component of 4-methylpyridine dissociation. For the direct dissociation of aniline, the CHNH+ yield is comparable with the CNH2 + yield of 21%. We can conclude from the presence of the CHNH+ ion that the dissociation mechanism includes a hydrogen shift. Fig. 5 shows the enhancement of the yield by the direct dissociation of aniline following the N 1s excitation/ionization. The yields of the ions with m/e = 15, 28, and 39 are significant. Specific production of NH+ (m/e = 15) and CHNH+ /CNH2 + (28) at the N 1s edge demonstrates that the fission occurs at around the excited/ionized
K. Okada et al. / International Journal of Mass Spectrometry 351 (2013) 56–60
59
although we cannot totally exclude the mechanism which includes the cleavage of the C2 C3 and C5 C6 bonds. Here, the doublycharged molecular ion is formed through a normal Auger or resonant double Auger process, while the singly-charged ion is produced through a resonant Auger decay. Reaction (1) explains the enhancement of the peak at m/e = 54 as well. We cannot describe the production pathways to these fragment ions only from our results, but the formation of C5 H6 + (m/e = 66) in the gray bars of Fig. 4 supports the existence of a pathway to the cyclopentadiene ion (c-C5 H6 + ) in Ref. [26], although the enhancement of this peak is very small at the N 1s edge. The intermediate c-C5 H6 CNH+ ion can be the precursor of the fragments of CHNH+ , CNH+ , etc.: the pathway involves the C1 C2 scission of the initial six-membered ring also. Low yield of the NH+ and NH2 + ions, however, reveals that the probability of breaking the C N bond is small in this system. Finally, it is interesting to note that the specific fragment ion for m/e = 39 is C3 H3 + for aniline and C2 HN+ for 4-methylpyridine [15]. While the excitation site is the nitrogen situated in different positions between the two molecules, the same positions on the benzene or pyridine ring are likely to break to produce C3 H3 + or C2 HN+ . It is highly desirable that the Auger electron–ion coincidence spectra should be acquired for both aniline and 4-methylpyridine to get more insight into the fragmentation processes. In some practical applications either isobaric fragment can be available by just selecting the sample to dissociate. 5. Summary
Fig. 4. Mass spectra of aniline represented as peak-area intensity obtained at the N 1s edge (panel a), and at the valence ionization (panel b). The white and gray bars indicate the yields before and after subtracting the contribution from the 4methylpyridine dissociation, respectively.
nitrogen atom. The NH+ formation requires the cleavage of the C N bond and an N H bond, whereas the production of CHNH+ /CNH2 + includes scission of the second nearest C1 C2 and C1 C6 bonds of the aniline cation. Rupture of the C1 C2 and C4 C5 bonds yields the C3 H3 + ion at m/e = 39, i.e., C6 H5 NH2 2+/+ → C3 H3 + + C3 H2 NH2 (+) ,
(1)
Fragmentation of aniline induced by the K-shell excitation/ionization was studied by the electron impact method. Mass spectra of the fragment ions were acquired to discuss the fragmentation of aniline. Dominant peaks were observed at m/e = 39, 28, and 38, in order of abundance. Assignment of the fragment ions was given by the comparison with the mass spectra of aniline-2,3,4,5,6d5 and by the MLR analysis. We conclude that the peaks of m/e = 28, 38, and 39 are assigned to the CHNH+ /CNH2 + , C3 H2 + , and C3 H3 + ions, respectively. The presence of the CH3 + ion in the spectra suggests that the aniline molecular cation isomerizes to some extent prior to dissociation. The contribution from the fragmentation of the isomer, 4-methylpyridine, was subtracted in the mass spectra to discuss the specific fragmentation by the direct dissociation of aniline. The yields of the C3 H3 + , CHNH+ /CNH2 + , and NH+ fragments are then found to be enhanced at the N 1s edge, indicating that the C1 C2 bond of the ring and the C N bond are subject to break by the N 1s excitation/ionization. Although the enhancement is small at the N 1s edge, the formation of C5 H6 + suggests that there exists also a pathway to form the cyclopentadiene ion. Acknowledgments The authors wish to express sincere thanks to Dr. Yukiteru Katsumoto of Hiroshima University for his advice on the MLR analysis. This study was partially supported by a Grant-in-Aid for Scientific Research from Japan Society for the Promotion of Science. References
Fig. 5. Enhancement of the yield of fragment ions by the direct dissociation of aniline following the N 1s excitation/ionization, relative to the yield of ions at the valence ionization.
[1] W. Eberhardt, T.K. Sham, R. Carr, S. Krummacher, M. Strongin, S.L. Weng, D. Wesner, Physical Review Letters 50 (1983) 1038. [2] W. Habenicht, H. Baiter, K. Müller-Dethlefs, E.W. Schlag, Journal of Physical Chemistry 95 (1991) 6774. [3] I. Nenner, C. Reynaud, H.C. Schmelz, L. Ferrand-Tanaka, M. Simon, P. Morin, Zeitschrift für Physikalische Chemie 195 (1996) 43. [4] S. Nagaoka, T. Fujibuchi, J. Ohshita, M. Ishikawa, I. Koyano, International Journal of Mass Spectrometry and Ion Processes 171 (1997) 95. [5] S. Nagaoka, Y. Tamenori, M. Hino, T. Kakiuchi, J. Ohshita, K. Okada, T. Ibuki, I.H. Suzuki, Chemical Physics Letters 412 (2005) 459.
60
K. Okada et al. / International Journal of Mass Spectrometry 351 (2013) 56–60
[6] K. Okada, S. Tanimoto, T. Morita, K. Saito, T. Ibuki, T. Gejo, Journal of Physical Chemistry A 107 (2003) 8444. [7] T. Ibuki, Y. Shimada, R. Hashimoto, S. Nagaoka, M. Hino, K. Okada, I.H. Suzuki, Y. Morishita, Y. Tamenori, Chemical Physics 314 (2005) 119. [8] H. Fukuzawa, G. Prümper, S. Nagaoka, T. Ibuki, Y. Tamenori, J. Harries, X.-J. Liu, K. Ueda, Chemical Physics Letters 431 (2006) 253. [9] S. Nagaoka, G. Prümper, H. Fukuzawa, M. Hino, M. Takemoto, Y. Tamenori, J. Harries, I.H. Suzuki, O. Takahashi, K. Okada, K. Tabayashi, X.-J. Liu, T. Lischke, K. Ueda, Physical Review A 75 (2007) 020502R. [10] M.F. Erben, M. Geronés, R.M. Romano, C.O. Della Védova, Journal of Physical Chemistry A 110 (2006) 875. [11] A. Mocellin, M.S.P. Mundim, L.H. Coutinho, M.G.P. Homem, A. Naves de Brito, Journal of Electron Spectroscopy and Related Phenomena 156–158 (2007) 245. [12] Y. Baba, Low Temperature Physics 29 (2003) 228. [13] S. Wada, H. Kizaki, Y. Matsumoto, R. Sumii, K. Tanaka, Journal of Physics: Condensed Matter 18 (2006) S1629. [14] K. Okada, M. Sakai, K. Ohno, Bulletin of the Chemical Society of Japan 81 (2008) 1580. [15] M. Sakai, K. Okada, K. Ohno, K. Tabayashi, Journal of Mass Spectrometry 45 (2010) 306. [16] R. Murphy, W. Eberhardt, Journal of Chemical Physics 89 (1988) 4054. [17] C. Miron, M. Simon, N. Leclercq, D.L. Hansen, P. Morin, Physical Review Letters 81 (1998) 4104.
[18] I.H. Suzuki, A. Nitta, H. Fukuzawa, K. Ueda, O. Takahashi, Y. Tamenori, S. Nagaoka, Journal of Chemical Physics 131 (2009) 164309. [19] T. Ohta, T. Fujikawa, H. Kuroda, Bulletin of the Chemical Society of Japan 48 (1975) 2017. [20] C.C. Turci, S.G. Urquhart, A.P. Hitchcock, Canadian Journal of Chemistry 74 (1996) 851. [21] D. Duflot, J.-P. Flament, A. Giuliani, J. Heinesch, M. Grogna, M.-J. Hubin-Franskin, Physical Review A 75 (2007) 052719. [22] Y. Luo, H. Ågren, J. Guo, P. Skytt, N. Wassdahl, J. Nordgren, Physical Review A 52 (1995) 3730. [23] Y.-K. Kim, K.K. Irikura, Electron-impact ionization cross sections for polyatomic molecules, radicals, and ions, in: K.A. Berrington, K.L. Bell (Eds.), Atomic and Molecular Data and Their Applications, AIP Conf. Proc., 543, 2000, p. 220 (See also the database and the references therein at http://physics.nist.gov/PhysRefData/Ionization). [24] A.J. Dobson, A. Barnett, An Introduction to Generalized Linear Models, 3rd ed., Chapman & Hall/CRC Press, London/Boca Raton, 2008. [25] C.M. Tseng, Y.A. Dyakov, C.L. Huang, A.M. Mebel, S.H. Lin, Y.T. Lee, C.K. Ni, Journal of the American Chemical Society 126 (2004) 8760. [26] J.C. Choe, N.R. Cheong, S.M. Park, International Journal of Mass Spectrometry 279 (2009) 25.
Contents lists available at ScienceDirect
International Journal of Mass Spectrometry journal homepage: www.elsevier.com/locate/ijms
Specific fragmentation of the K-shell excited/ionized aniline molecule studied by electron impact Kazumasa Okada ∗ , Yasuhiko Kashiwagi, Masamichi Sakai Department of Chemistry, Graduate School of Science, Hiroshima University, Higashi-Hiroshima 739-8526, Japan
a r t i c l e
i n f o
Article history: Received 7 February 2013 Received in revised form 26 July 2013 Accepted 26 July 2013 Available online 3 August 2013 Keywords: Site-specific fragmentation Inner-shell excitation Electron impact
a b s t r a c t Fragmentation of aniline upon K-shell excitation/ionization has been studied by the electron impact method. Ions detected as dominant fragments were C3 H3 + , CHNH+ /CNH2 + , C3 H2 + , etc. The C3 H3 + ion was produced characteristically at the N 1s edge. We found from the mass spectra of aniline-2,3,4,5,6d5 that CD3 + is formed independent of excitation energy, suggesting that the aniline molecular cation isomerizes prior to dissociation. The mass spectra of the fragment ions by the direct dissociation of aniline were deduced by subtracting the contribution from the dissociation of the isomer, 4-methylpyridine. Enhancement of the C3 H3 + , CHNH+ /CNH2 + , and NH+ yields at the N 1s edge shows that the cleavage of the C1 C2 bond of the initial benzene ring and the C N bond fission are likely to take place by the N 1s excitation and ionization. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Excitation of molecules at the K-shell level has attracted much attention for the last three decades, because site-specific fragmentation processes were observed in various molecular systems [1–15]. Inner-shell excited or ionized molecules dissociate into fragment ions following Auger electron emission. If the Auger decay process depends highly on the core-hole states, there is a high probability of producing fragment ions distinctive to the particular atomic site of excitation or ionization [16–18]. Recently, we have reported the site-specific fragmentation of the 2-, 3- and 4-methylpyridine molecules following the C 1s and N 1s excitation/ionization by electron impact [15]. Specific production of the C2 HN+ and C5 H5 + /C5 H6 + ions at the N K-edge reveals that the N C2 and C4 C5/C5 C6 bond breakings of the pyridine ring is likely to occur following the N 1s excitation and ionization. Ab initio molecular orbital calculations indicate that the site specificity observed can be reasonably understood by the dissociation of the molecular dications in the most probable Auger final states. This article reports the fragmentation of a structural isomer, aniline, with a nitrogen atom in the substituent group, induced by electron impact. Replacement of different positions in a target molecule with a heteroatom by changing molecules enables us to explore in more general the site-specific behavior of fragmentations of molecules. The main interest here lies in investigating if the site-specific fragmentations are observed in these molecular
∗ Corresponding author. Tel.: +81 82 424 7102; fax: +81 82 424 0727. E-mail address: [email protected] (K. Okada). 1387-3806/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ijms.2013.07.013
systems and whether the bonds dissociate around the excitation site. To our knowledge, limited experimental studies have been reported on the spectroscopy of aniline in the K-shell region. Ohta and coworkers [19] observed the gas-phase X-ray photoelectron spectra of several mono-substituted derivatives of benzene including aniline. Oscillator strengths for the inner-shell excitation of aniline were reported by Hitchcock’s group using inner-shell electron energy-loss spectroscopy (ISEELS) [20]. About a decade later, Duflot et al. [21] measured the spectra with higher resolution and gave a reliable assignment of the bands with the aid of ab initio configuration interaction calculations. The X-ray photoabsorption of condensed aniline was published by Luo et al. [22]. However, the mass spectrometric study on the fragmentation of aniline in the K-shell region has not been reported so far. The article is organized as follows. Next section describes our experimental procedure. We then present the main results followed by the discussion. The mass spectra indicate that the C3 H3 + ion is produced characteristically at the N 1s edge, showing that the site-specific fragmentation occurs at this edge. The presence of the CD3 + peak for the aniline-2,3,4,5,6-d5 sample suggests that the ring permutation leading to the 4-methylpyridine cation is involved in the dissociation of the aniline cation. Finally, we summarize the conclusions of this study. 2. Experimental The experiments were performed in an electron-impact reaction chamber equipped with a quadrupole mass spectrometer (Hiden Analytical, HAL-4·EQS-300). Since the experimental setup and
K. Okada et al. / International Journal of Mass Spectrometry 351 (2013) 56–60
57
procedure have been described in detail previously [14], only a brief outline is given here. The electron beam for excitation was prepared by an electron gun (Yamamoto Shinku Lab., EG-50), which can emit electrons with the energy of 200–2000 eV. The energy resolution is stated as 0.7 eV. The electron beam current was monitored with a Faraday cup mounted downstream on the electron beam axis. An electrostatic potential of −50 V was applied at the extractor plate of the spectrometer in order to extract and detect fragment ions produced at the interaction region. The pressure in the chamber was kept at about 4 × 10−6 Torr (1 Torr = 133.2 Pa) during the measurements in which the effusive beam of sample gas was introduced. The measurements were also conducted for the sample where all the hydrogen atoms in the benzene ring have been replaced by deuterium (i.e., aniline-2,3,4,5,6-d5 ) in order to give unambiguous assignment among isobaric fragments for some mass-spectral peaks. Mass spectra of ions produced by the K-shell excitation/ionization were acquired by a spectral difference method: the mass spectra were first measured above the K-shell ionization threshold and then below the K-edge. Subtraction of the spectra between the above and below K-edges gives the mass spectrum of fragment ions produced at the energies of interest. The samples of aniline and aniline-2,3,4,5,6-d5 (minimum isotopic purity of 98% D) were obtained commercially (Aldrich Chemical Co., Inc.) and were carefully degassed under vacuum by repetitive freeze–pump–thaw cycles.
3. Results Mass spectra of fragment ions of aniline and aniline-2,3,4,5,6-d5 are displayed in Figs. 1 and 2, respectively. For each molecule the mass spectrum at the N 1s edge was obtained by subtracting the raw spectrum at 390 eV from the one at the 1000-eV electron energy, both of which had been normalized with electron beam intensities. On subtracting, efficiencies of ionization were taken into account by using the electron-impact ionization cross-sections. Since the cross-section data for aniline are not available, we assumed that they are approximated as those for benzene [23]. The spectrum at the C 1s edge was obtained in a similar way from the data at 390 eV and 270 eV. The spectrum for the valence ionization was acquired at the electron energy of 270 eV. The N 1s and C 1s ionization thresholds of aniline are 405.3 eV and 289.7–291.2 eV, respectively [19]. One finds from the mass spectra of normal aniline (Fig. 1) that the dominant fragments are, in order of abundance, the ions with m/e = 39, 28, 38, 27, 26, etc. Closer look at the spectra reveals the increase of the peak at m/e = 39 at the N 1s edge (Fig. 1a) compared with the data for the valence ionization (Fig. 1c). The prominent peaks in the mass spectra of deuterated aniline (Fig. 2) are m/e = 42, 28, 40, 29, and 30. Assignment of these peaks to fragment ions can be given from these spectral patterns. Let us begin with the peaks at m/e = 38 and 39. It is apparent that the trend in the increasing yield with m/e in 36–39 in Fig. 1 corresponds to that of m/e = 36, 38, 40, and 42 in Fig. 2. The peaks at m/e = 38 and 39 can thus be assigned mostly to C3 H2 + and C3 H3 + , respectively. Assignment of the peaks in m/e = 26–28 is, however, not as obvious as that of the peaks at m/e = 38 and 39. We employed a multiple linear regression (MLR) analysis [24] for the assignment of m/e = 26–28. Based on the fact that there are only trace amounts of CH+ in the fragmentation of deuterated aniline (Fig. 2), virtually no H/D exchange is presumed to occur: for aniline-2,3,4,5,6-d5 no D atoms make bonds with N throughout the process. We also assume that isotope effect is not crucial for the ion yields. For example, the peak at m/e = 27 in the mass spectra of normal aniline consists of CNH+ , CHN+ , and C2 H3 + , where the
Fig. 1. Mass spectra of aniline acquired at the N 1s and C 1s edges (panels a and b), and at the valence ionization (panel c). The spectra (a) and (b) were obtained by the difference between the above and below K-edges.
notation CHm NHn + represents that there are m hydrogen atoms bonded with carbon and n hydrogen atoms bonded with a nitrogen atom. That is, the yields of CNH+ are equivalent between normal and isotope-substituted anilines, and the intensity of CHN+ for normal aniline is the same as the CDN+ component for deuterated aniline. Such conditions formulate a set of simultaneous equations with the yields containing experimental errors. The solution for the N K-edge data indicates that C2 H2 + comprises 85% of the intensity at m/e = 26 and the peak m/e = 28 contains 66% CHNH+ and 21% CNH2 + . The intensity at m/e = 27 is not attributed to a particular fragment ion, consisting of CNH+ , CHN+ , and C2 H3 + in the ratio 34:25:41. Although the number of significant figures is one according to the statistical analysis, two digits are given here for the purpose of later discussion. Also, we are aware that the assumptions introduced above may bring errors in the MLR analysis due to the isomerization described in the next section, and such errors may be significant for the higher-mass peaks for which we did not employ the MLR analysis. The effect of the K-shell excitation/ionization on fragmentation can be clarified by an enhancement of the yield of the fragment ions relative to the mass spectral pattern obtained at the valence ionization. Enhancement of the yield of the fragment ions produced by the N 1s and C 1s excitation/ionization is shown in Fig. 3. The term “enhancement” here was defined in our previous studies [14,15]. That is, the spectrum of Fig. 3a is, for example, the result of the subtraction of Fig. 1c from Fig. 1a. Positive and negative values (as percentage points) mean an increase and decrease of the relative yields, respectively, compared with the yields for the valence
58
K. Okada et al. / International Journal of Mass Spectrometry 351 (2013) 56–60
Fig. 2. Mass spectra of aniline-2,3,4,5,6-d5 acquired at the N 1s and C 1s edges (panels a and b), and at the valence ionization (panel c).
ionization. One can inspect a promotion effect by the inner-shell excitation/ionization on the fragmentation, using the mass spectrum for the valence ionization as a reference. Ion yields were calculated from the peak areas of the mass spectra, with totals normalized to 100. The distinct production of a fragment ion, C3 H3 + , was observed at the N 1s edge (Fig. 3a). Only a weak enhancement was found of the peak at m/e = 27 for the C 1s edge data (Fig. 3b), which has a similar trend to the methylpyridine cases at the C 1s edge [15]. Therefore, in the next section we shall deal exclusively with the N 1s edge data. 4. Discussion Fragmentation of aniline is not as simple as it is expected. One can see a peak at m/e = 18 in Fig. 2, which indicates the formation of CD3 + being not produced by the direct dissociation of aniline. The CD3 + production is independent of the excitation energy. A clue to solve this mysterious peak can be found in a 193-nm laser photodissociation study of aniline reported by Tseng et al. [25]. They claimed that at least 23% of (neutral) aniline in the ground electronic state rearranges to a seven-membered ring. The ring can then rearomatize to yield 4-methylpyridine prior to dissociation. Such ring permutation is also presumed to occur in the aniline cation, as demonstrated by Choe et al. in a theoretical study on the reaction pathways for an H loss [26]. For the best of our knowledge, no experimental study has been reported so far on the isomerization and fragmentation of the aniline cation. Therefore, we assume in the following analysis that the dissociation of the 4-methylpyridine ion is contributed to 23% of the total ion yield for each spectrum.
Fig. 3. Enhancement of the yield of fragment ions produced by the N 1s and C 1s excitation/ionization of aniline, relative to the yield of ions at the valence ionization. The enhancement is defined by the difference from the valence ionization spectrum and depicted as a bar chart.
Mass spectra of fragment ions produced by the direct dissociation of aniline were obtained by subtracting the mass spectra of 4-methylpyridine [15] from the spectra given in Fig. 1. We cannot evaluate the extent of the contribution from 4-methylpyridine in this study, but we conclude that the value of 23% is reasonable because the peak at m/e = 15 gave a negative value for the spectra of the C 1s edge and valence ionization if the value >30% was used in the subtraction. Fig. 4 displays the mass spectra obtained by such evaluation. The mass spectra have been represented in terms of peak-area intensity. The white and gray bars indicate the intensities before and after subtracting the contribution from the 4-methylpyridine dissociation, respectively: the white bars (including the hidden area overlapped with gray bars) correspond to the mass spectra shown in Fig. 1. At the N 1s edge prominent peaks depicted as gray bars are m/e = 39, 38, 28, and 27. The peak at m/e = 15 is now identified as NH+ . The intensity of m/e = 39 is not so much affected, because this peak for 4-methylpyridine corresponds to C2 HN+ [15]. The intensity of m/e 28 becomes 56% after the subtraction, and the peak still has a contribution from both CHNH+ and CNH2 + : the peak counts up to 22% CHNH+ after subtracting the component of 4-methylpyridine dissociation. For the direct dissociation of aniline, the CHNH+ yield is comparable with the CNH2 + yield of 21%. We can conclude from the presence of the CHNH+ ion that the dissociation mechanism includes a hydrogen shift. Fig. 5 shows the enhancement of the yield by the direct dissociation of aniline following the N 1s excitation/ionization. The yields of the ions with m/e = 15, 28, and 39 are significant. Specific production of NH+ (m/e = 15) and CHNH+ /CNH2 + (28) at the N 1s edge demonstrates that the fission occurs at around the excited/ionized
K. Okada et al. / International Journal of Mass Spectrometry 351 (2013) 56–60
59
although we cannot totally exclude the mechanism which includes the cleavage of the C2 C3 and C5 C6 bonds. Here, the doublycharged molecular ion is formed through a normal Auger or resonant double Auger process, while the singly-charged ion is produced through a resonant Auger decay. Reaction (1) explains the enhancement of the peak at m/e = 54 as well. We cannot describe the production pathways to these fragment ions only from our results, but the formation of C5 H6 + (m/e = 66) in the gray bars of Fig. 4 supports the existence of a pathway to the cyclopentadiene ion (c-C5 H6 + ) in Ref. [26], although the enhancement of this peak is very small at the N 1s edge. The intermediate c-C5 H6 CNH+ ion can be the precursor of the fragments of CHNH+ , CNH+ , etc.: the pathway involves the C1 C2 scission of the initial six-membered ring also. Low yield of the NH+ and NH2 + ions, however, reveals that the probability of breaking the C N bond is small in this system. Finally, it is interesting to note that the specific fragment ion for m/e = 39 is C3 H3 + for aniline and C2 HN+ for 4-methylpyridine [15]. While the excitation site is the nitrogen situated in different positions between the two molecules, the same positions on the benzene or pyridine ring are likely to break to produce C3 H3 + or C2 HN+ . It is highly desirable that the Auger electron–ion coincidence spectra should be acquired for both aniline and 4-methylpyridine to get more insight into the fragmentation processes. In some practical applications either isobaric fragment can be available by just selecting the sample to dissociate. 5. Summary
Fig. 4. Mass spectra of aniline represented as peak-area intensity obtained at the N 1s edge (panel a), and at the valence ionization (panel b). The white and gray bars indicate the yields before and after subtracting the contribution from the 4methylpyridine dissociation, respectively.
nitrogen atom. The NH+ formation requires the cleavage of the C N bond and an N H bond, whereas the production of CHNH+ /CNH2 + includes scission of the second nearest C1 C2 and C1 C6 bonds of the aniline cation. Rupture of the C1 C2 and C4 C5 bonds yields the C3 H3 + ion at m/e = 39, i.e., C6 H5 NH2 2+/+ → C3 H3 + + C3 H2 NH2 (+) ,
(1)
Fragmentation of aniline induced by the K-shell excitation/ionization was studied by the electron impact method. Mass spectra of the fragment ions were acquired to discuss the fragmentation of aniline. Dominant peaks were observed at m/e = 39, 28, and 38, in order of abundance. Assignment of the fragment ions was given by the comparison with the mass spectra of aniline-2,3,4,5,6d5 and by the MLR analysis. We conclude that the peaks of m/e = 28, 38, and 39 are assigned to the CHNH+ /CNH2 + , C3 H2 + , and C3 H3 + ions, respectively. The presence of the CH3 + ion in the spectra suggests that the aniline molecular cation isomerizes to some extent prior to dissociation. The contribution from the fragmentation of the isomer, 4-methylpyridine, was subtracted in the mass spectra to discuss the specific fragmentation by the direct dissociation of aniline. The yields of the C3 H3 + , CHNH+ /CNH2 + , and NH+ fragments are then found to be enhanced at the N 1s edge, indicating that the C1 C2 bond of the ring and the C N bond are subject to break by the N 1s excitation/ionization. Although the enhancement is small at the N 1s edge, the formation of C5 H6 + suggests that there exists also a pathway to form the cyclopentadiene ion. Acknowledgments The authors wish to express sincere thanks to Dr. Yukiteru Katsumoto of Hiroshima University for his advice on the MLR analysis. This study was partially supported by a Grant-in-Aid for Scientific Research from Japan Society for the Promotion of Science. References
Fig. 5. Enhancement of the yield of fragment ions by the direct dissociation of aniline following the N 1s excitation/ionization, relative to the yield of ions at the valence ionization.
[1] W. Eberhardt, T.K. Sham, R. Carr, S. Krummacher, M. Strongin, S.L. Weng, D. Wesner, Physical Review Letters 50 (1983) 1038. [2] W. Habenicht, H. Baiter, K. Müller-Dethlefs, E.W. Schlag, Journal of Physical Chemistry 95 (1991) 6774. [3] I. Nenner, C. Reynaud, H.C. Schmelz, L. Ferrand-Tanaka, M. Simon, P. Morin, Zeitschrift für Physikalische Chemie 195 (1996) 43. [4] S. Nagaoka, T. Fujibuchi, J. Ohshita, M. Ishikawa, I. Koyano, International Journal of Mass Spectrometry and Ion Processes 171 (1997) 95. [5] S. Nagaoka, Y. Tamenori, M. Hino, T. Kakiuchi, J. Ohshita, K. Okada, T. Ibuki, I.H. Suzuki, Chemical Physics Letters 412 (2005) 459.
60
K. Okada et al. / International Journal of Mass Spectrometry 351 (2013) 56–60
[6] K. Okada, S. Tanimoto, T. Morita, K. Saito, T. Ibuki, T. Gejo, Journal of Physical Chemistry A 107 (2003) 8444. [7] T. Ibuki, Y. Shimada, R. Hashimoto, S. Nagaoka, M. Hino, K. Okada, I.H. Suzuki, Y. Morishita, Y. Tamenori, Chemical Physics 314 (2005) 119. [8] H. Fukuzawa, G. Prümper, S. Nagaoka, T. Ibuki, Y. Tamenori, J. Harries, X.-J. Liu, K. Ueda, Chemical Physics Letters 431 (2006) 253. [9] S. Nagaoka, G. Prümper, H. Fukuzawa, M. Hino, M. Takemoto, Y. Tamenori, J. Harries, I.H. Suzuki, O. Takahashi, K. Okada, K. Tabayashi, X.-J. Liu, T. Lischke, K. Ueda, Physical Review A 75 (2007) 020502R. [10] M.F. Erben, M. Geronés, R.M. Romano, C.O. Della Védova, Journal of Physical Chemistry A 110 (2006) 875. [11] A. Mocellin, M.S.P. Mundim, L.H. Coutinho, M.G.P. Homem, A. Naves de Brito, Journal of Electron Spectroscopy and Related Phenomena 156–158 (2007) 245. [12] Y. Baba, Low Temperature Physics 29 (2003) 228. [13] S. Wada, H. Kizaki, Y. Matsumoto, R. Sumii, K. Tanaka, Journal of Physics: Condensed Matter 18 (2006) S1629. [14] K. Okada, M. Sakai, K. Ohno, Bulletin of the Chemical Society of Japan 81 (2008) 1580. [15] M. Sakai, K. Okada, K. Ohno, K. Tabayashi, Journal of Mass Spectrometry 45 (2010) 306. [16] R. Murphy, W. Eberhardt, Journal of Chemical Physics 89 (1988) 4054. [17] C. Miron, M. Simon, N. Leclercq, D.L. Hansen, P. Morin, Physical Review Letters 81 (1998) 4104.
[18] I.H. Suzuki, A. Nitta, H. Fukuzawa, K. Ueda, O. Takahashi, Y. Tamenori, S. Nagaoka, Journal of Chemical Physics 131 (2009) 164309. [19] T. Ohta, T. Fujikawa, H. Kuroda, Bulletin of the Chemical Society of Japan 48 (1975) 2017. [20] C.C. Turci, S.G. Urquhart, A.P. Hitchcock, Canadian Journal of Chemistry 74 (1996) 851. [21] D. Duflot, J.-P. Flament, A. Giuliani, J. Heinesch, M. Grogna, M.-J. Hubin-Franskin, Physical Review A 75 (2007) 052719. [22] Y. Luo, H. Ågren, J. Guo, P. Skytt, N. Wassdahl, J. Nordgren, Physical Review A 52 (1995) 3730. [23] Y.-K. Kim, K.K. Irikura, Electron-impact ionization cross sections for polyatomic molecules, radicals, and ions, in: K.A. Berrington, K.L. Bell (Eds.), Atomic and Molecular Data and Their Applications, AIP Conf. Proc., 543, 2000, p. 220 (See also the database and the references therein at http://physics.nist.gov/PhysRefData/Ionization). [24] A.J. Dobson, A. Barnett, An Introduction to Generalized Linear Models, 3rd ed., Chapman & Hall/CRC Press, London/Boca Raton, 2008. [25] C.M. Tseng, Y.A. Dyakov, C.L. Huang, A.M. Mebel, S.H. Lin, Y.T. Lee, C.K. Ni, Journal of the American Chemical Society 126 (2004) 8760. [26] J.C. Choe, N.R. Cheong, S.M. Park, International Journal of Mass Spectrometry 279 (2009) 25.