Electron impact ionization of molecular ions: Time dependence of the kinetic energy release of propane fragment ions — a case study

The study of dissociative collisions of electrons with molecular ions has been dominated for many years by the investigation of dissociative recombination. Dissociative excitation, dissociative ionization and direct ionization of molecular ions by electron collisions have received less attention. Nevertheless, the incentives associated with controlled thermonuclear fusion have recently led to a number of efforts since the edge plasmas contain many molecular ions. In principle, the research literature abounds with studies of dissociative

18

or surface collisions. In some of these, the kinetic energy release (KER) was determined as well [54, 55]. In contrast, reports on electron-induced dissociative reactions of molecular ions are scarce. Most of them involved diatomic or other small ions with the goal of measuring absolute cross sections for fragmentation [56]. Cross sections for ionization and dissociation of fullerene ions were reported recently by Salzborn and coworkers [57]. Freiser and coworkers [58] explored dissociative electron capture by polyatomic ions to obtain structural information from the fragment pattern and McLafferty and coworkers [59] extended these studies to large multiply charged ions. However, except for recent work in our laboratory [60, 61], the KER released in electron-induced reactions of polyatomic ions has not been measured.

Induced reactions offer, however, the possibility of analyzing the KER for ions with very short lifetimes. There is considerable interest in the time dependence of the KER [62–64], because statistically driven decay reactions without a reverse activation energy feature a KER which increases with decreasing ion lifetime as — at the same time — the statistically distributed excitation energy in the parent ion increases [61, 65]. In contrast, decay reactions with a large reverse barrier are characterized by an essentially constant, or even decreasing, KER [66].

In previous studies on this subject, the time at which dissociation was measured (relative to the time of ion formation in the ion source) was varied by varying the acceleration voltages [64, 67]. The advent of ion traps has made it possible to greatly increase the ion lifetimes being sampled [65], but a lower limit of the order of 1 to 10 ms still exists. By recently positioning a high-performance electron gun near the intermediate slit of a double focusing magnetic mass spectrometer of reverse geometry we have been able to measure the KER for electron induced reactions occurring within less than 0.75 µs after electron-impact excitation of various propane derived ions. We observe a large fractional decrease of the average KER for decay of C3H8+ into C3H7+, but a much smaller effect for decay of C3H7+ into C3H5+ when comparing the KER of this electron induced decay with the KER of the spontaneous decay reaction at this position of the mass spectrometer system (occurring on the time scale of about 11 to 14 µs). As will become clear below the differences are attributed to the existence of a large reverse activation energy for decay of C3H7+ into C3H5+.

Details of the experimental set-up and data analysis have been published elsewhere [68, 69].

The apparatus consists of a high resolution double focusing tow sector field mass spectrometer of reversed Nier-Johnson type geometry as already shown in Fig. 2. Propane molecules are introduced via a capillary leak gas inlet system into the ion source where they are ionized by an electron beam of variable energy and current. Ensuing cations are extracted by a weak electric field and accelerated through a potential drop of Uac = 3 kV into the spectrometer. They pass through the first field free region (1ff), are then momentum-analyzed by a magnetic sector field, enter a second field-free region (2ff, length 33.3 cm), pass through a 90° electric sector field and are finally detected by a channeltron-type electron multiplier operated in the single-ion counting mode.

In order to analyze the decay of mass-selected molecular ions (MIKE scan technique [63]), the parent ions are selected with the magnetic sector field and decay reactions in the 2ff are analysed by varying the voltage of the electric sector field. The decay may be either spontaneous (metastable), or it may be induced in this region of the mass spectrometer by electrons from a high-performance, home built electron gun [69]. The electron gun is mounted just before the defining aperture between the magnetic sector and the electric sector and the electrons intersect the ion beam at 90º. The gun is an order of magnitude more

powerful than the electron gun used previously for post-ionization of fullerene cat- and anions [60, 70], typically parameters used are an electron energy of 150 eV which results in an electron beam current of 7 mA. This setting is a compromise as the electron beam current would increase further with increasing energy, but the ionization cross section of propane reaches its maximum near 75 eV [18].

Mass-analyzed ion kinetic energy (MIKE) spectra are usually recorded as follows: The magnet is tuned to the mass of the parent ion, mp, while the electric sector field voltage U is scanned (Fig. 6). Stable singly charged ions will have a kinetic energy of 3 keV and pass at the nominal sector field voltage of Up = 509 V. Daughter ions (mass md), formed in the 2ff, will then pass at a voltage

U

m U m

p

d

=

(4)

This equation relates the position of a daughter ion peak to the position of the parent ion peak in a MIKE spectrum. In practice, the parent ion peak will have a finite width and a distinct shape which will also be imposed on the daughter ion peak. Any kinetic energy release (KER) in the reaction will further modify the peak shape of the daughter. If the MIKE peak is strictly Gaussian, then the average kinetic energy release, can be extracted from its full width at half maximum, DU, [71]. If the peak in the MIKE spectrum is not a Gaussian, then the Kinetic Energy Release Distributions (KERD) has to be derived from its derivative with respect to the sector field voltage. Details, including the use of ion trajectory calculations to simulate MIKE peaks for our spectrometer and various corrections, have been described in detail in [68, 69].

As a MIKE scan will always sample decay reactions which occur between the magnetic and electric sector, there is an important difference in the time scale between spontaneous and induced decay reactions. A parent ion (here assumed to be C3H8+) formed in the ion source at t = 0, will traverse the 2ff at a time interval 11.2 £ t £ 14.2 µs. In contrast, if the reaction is electron-induced in the 2ff, it has to occur within t £ 0.75 µs in order to contribute to the MIKE spectrum.

0510 C₃H₈⁺®C₃H₇⁺ + H

spontaneous

Ion signal (arb.units)

496 497 498 508 509 510

0510 electron induced

Sector field voltage

Figure 6 Top: MIKE scan for spontaneous dissociation of C₃H₈⁺ into C₃H₇⁺ + H after [69]. The solid

20

Figure 6 (upper part) displays the MIKE scan for the spontaneous reaction (5).

C3H8+® C3H7+ + H

(5)

In this case the peak is Gaussian, except for a slightly enhanced background to the right of the daughter ion peak which is believed to stem from decay reactions in the electric sector field [72]. From the excellent non-linear least squares fit (solid line in Fig. 6) and after correcting for the width of the parent ion peak an average KER of 9.3 ± 1.5 meV can be derived in good agreement with an earlier values of 9.4 meV from our laboratory [68] and of 9.6 meV reported by Medved et al. [64]. In the lower part of Fig. 6 we show the MIKE spectrum for the electron-induced decay reaction (5). The corresponding average KER of = 13.2 ± 1.2 meV for the electron-induced reaction (5) thus is larger than the spontaneous reaction by 3.9 ± 1.8 meV. Moreover, the corresponding KERD are readily derived from the Gaussian fits to the MIKE spectra after deconvolution with the parent ion peaks. The shift of the distribution for the electron-induced reaction is obvious from the results shown in Fig. 7.

0 10 20 30 40

0.00.51.0 C₃H₈⁺ _® C₃ H₇⁺ + H

Probability (normalized)

Kinetic energy release (meV)

spontaneous induced

Figure 7. Kinetic energy release distributions (KERD) for spontaneous and electron-induced dissociation of C₃H₈⁺ into C₃H₇⁺ + H after [69].

Similar studies have been carried out for the reaction

C3H7+® C3H5+ + H2

(6)

In Fig. 8 upper panel we show the corresponding MIKE spectrum for the electron-induced reaction. The daughter peak is, like in the spectrum for the spontaneous reaction, flat-topped, characteristic of a reaction which features a sizeable reverse activation energy. The KERD derived from these data is displayed in Fig. 8 lower panel. A large threshold value in the KER is immediately apparent. The average KER for the electron induced reaction is found from this distribution to be 397 meV, while the average KER for the spontaneous reaction is obtained as 386 meV. The uncertainty of these values are probably on the order of 10 meV. In a previous study [68] we had derived an average KER of 400 meV for the spontaneous reaction, while Holmes et al. [71] reported a value of 440 meV.

0 200 400 600 800

01

KERD

Probability (normalized)

Kinetic energy release (meV)

482 484 486 488 490

0510

C₃H₇⁺ e® C₃ H₅⁺ + H₂

Ion signal (arb. units)

Sector field voltage (V)

Figure 8. Top: MIKE scan for electron-induced dissociation of C₃H₇⁺ into C₃H₅⁺ + H₂ after [69]. The solid line results from FFT-smoothing. Bottom: The KER distribution obtained from the smooth MIKE spectrum, after deconvolution with the parent ion peak.

Therefore, for reaction 5 we find a dramatic change of the KERD (see Fig. 6), corresponding to an increase of the average KER by 42 %, whereas the average KER for reaction 6, in contrast, is enhanced by a mere 3 %. It should be pointed out that the average KER measured for the induced reactions only represent lower limits, because the MIKE peak is actually a superposition of the spontaneous and electron induced signal. As discussed in detail in Ref.

[69] the strong enhancement in the case of reaction (5) is attributed to a change in the time scale over which the reaction is monitored. Furthermore, the occurrence of only a small change for reaction (6) is attributed to the existence of a large reverse activation energy.

It is interesting to note that we have recently extended our experimental set-up by adding a third field free region with a high performance electron gun and a third analyzer (electric sector field) thus allowing us to study in more detail electron induced dissociation and ionization of mass selected molecular ions including step by step ionisation mechanisms (see Fig. 9).

22

Figure 9. Schematic view of recently constructed three sector field mass spectrometer used for the study of electron induced dissociation and ionization reactions in Innsbruck.

4. Cross sections: Theoretical considerations