Dissociative electron attachment studies of gas-phase acetic acid using a velocity map imaging technique

Advancing instrumentation to explore dissociative electron attachment (DEA) studies allows previously unattainable information to be acquired. Using a newly constructed velocity map imaging spectrometer, we revisited a study on DEA to gas-phase acetic acid. We discuss possible fragmentation channels and compared the corresponding ion yields with previous high electron-energy resolution results. We focus on the channels occurring at higher energies, particularly near 10 eV, and calculate their thermodynamic thresholds. Moreover, we expand previous studies and perform time-sliced imaging near the 10 eV resonance to obtain the kinetic energy distribution of the fragment ions.


Introduction
Beyond its importance in the fundamental research of molecular physics, the development of advanced instrumentation to explore dissociative electron attachment (DEA) has benefited a wide range of other scientific and applied areas that involve low-energy electrons formed by high-energy irradiation [1].A technology recently growing in popularity is a spectrometer for velocity map imaging (VMI), although already invented decades ago and used for DEA studies [2][3][4][5][6][7].The VMI spectrometer can measure both the kinetic energy and angular distributions of negative ions formed in DEA to gas-phase Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence.Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.molecules [8,9].These measurements also provide information about the symmetries of resonant states, which are created before molecular dissociation.Therefore, the use of VMI on molecular targets previously investigated in typical DEA studies using mass spectrometry techniques is necessary to acquire a more comprehensive description of fragmentation.Because the DEA process is crucial in many areas, such as astrophysics, materials manufacturing, and radiotherapy, it has led to rich investigations of diverse molecular targets [10].Acetic acid (CH 3 COOH) is among several organic compounds that have sparked interest because, in addition to its fundamental nature, its dissociation is relevant to various biological and industrial processes and the formation of potential biomolecules in the interstellar medium (ISM).Detection of acetic acid as a component of the ISM raised the question of the formation mechanisms with a possible series of gas-phase reactions [11].Later, methyl formate (HCOOCH 3 ) and glycolaldehyde (HOCH 2 CHO), two structural isomers of acetic acid, were also reported to be present in the ISM [12].As determined by their characteristic spectra, the respective ratios of ISM abundance of these three isomers are 1:26:0.5(acetic acid: methyl formate: glycolaldehyde) [12]; and importantly, they can act as markers or precursors for larger biomolecules.For example, it has been suggested that simple organic acids can be an important precursor in the formation of biomolecules such as glycine (CH 3 COONH 2 ), the simplest biological amino acid potentially present in the ISM, formed by the reaction of acetic acid with the radical cation of ammonia [13].This suggestion motivated several laboratory studies and computational models to propose factors, also including electrons that can be produced by cosmic radiation [14], which play intimate roles in the chemical and physical pathways of formation and dissociation of organic compounds.Hence, the study of low-energy-electron (LEE)-induced dissociation of acetic acid can provide insight into the energetic pathways of its formation and abundance in space and shed light on the formation of more complex components of the ISM.
Among the primary LEE induced anionic dissociation products of carboxylic acids, which contain the most common organic acid functional group: -COOH, the carboxyl group, usually having a negative charge on its oxygen atoms when H is removed from it.According to previous studies, electron attachment to formic acid (HCOOH), the simplest carboxylic acid, yields three anionic fragments, HCOO − via H loss, OH − , and O − , with HCOO − being the dominant fragment resulting from cleavage of the O-H bond in the carboxyl group [15].Similarly, CH 3 COO − , the most prevalent DEA fragment from acetic acid, results from the cleavage of the O-H bond and thus H loss from the carboxyl group.Several previous experimental studies on LEE collisions with acetic acid have been reported, with a focus on DEA.For example, a study by Pelc et al [15,16] reported low-energy shape resonances, with the most dominant channel corresponding to CH 3 COO − and H-atom formation, as did experiments by Sailer et al [17] and Prabhudesai et al [18].In addition, some theoretical efforts focused on assigning these low-energy resonances to particular electronic transitions in acetic acid [17,19].However, no extensive report exists on the resonances involved at higher electron energies that can be assigned to coreexcited Feshbach resonances, lying slightly below the ionization threshold, which was calculated at 10.84 eV [20] for acetic acid.Such Feshbach resonances are formed when two lowest unoccupied molecular orbitals (LUMOs) become populated or one LUMO becomes doubly occupied.The core-excited resonance for acetic acid was discussed briefly by Sailer et al [17].Moreover, several anionic products were observed, i.e.CH 2 COO − , CHCO − , CCO − , OH − , O − , CH 2 − , and H − , at resonance energies near 10 eV.To provide additional information about the 10 eV Feshbach resonance present in DEA to acetic acid, we performed VMI studies of different fragment ions over the resonance energies.The time-sliced images recorded in this investigation allowed the kinetic energy and angular distribution of the fragment ions to be obtained.

Experimental and computational methodology
This investigation utilized the time-sliced VMI technique using a recently designed and constructed apparatus in the Notre Dame Radiation Laboratory.The custom-designed VMI spectrometer and the detector assembly from Photek USA are shown in figure 1.The experimental setup contains an electron gun, a Faraday cup, and the VMI spectrometer kept inside an ultrahigh vacuum (UHV) chamber.The spectrometer is shielded by mu-metal to minimize any effects caused by the external magnetic field.The electron gun (model No: ELG-2/EGPS-1022) was purchased from Kimball Physics.The electron gun comprises a resistively heated tantalum filament that produces an electron beam with a typical energy resolution of 0.6 eV and a minimum energy range of 2 eV.The electron beam is further collimated by the magnetic field produced by two coils in the Helmholtz configuration placed outside the vacuum chamber.This collimated electron beam is introduced at right angles to an effusive molecular beam produced by a capillary 550 µm in diameter along the axis of the VMI spectrometer.The spectrometer comprises repeller and extractor plates, an electrostatic lens assembly, a flight tube, and a detector.The stainless steel repeller and extractor plates with a 60 mm outer diameter and a 28 mm inner diameter are placed apart by 20 mm.The capillary needle is connected to the center of the repeller plate; however, they are electrically isolated from each other.Electron-molecule collisions occur between the repeller and extractor plates (figure 1).A set of four lens electrodes, with the same dimensions as the repeller/extractor plates, are placed approximately 13 mm from the extractor plate.The separation between each lens electrode is 10 mm, and these electrodes are used mainly to focus the negative ions and keep the electric field uniform along the VMI spectrometer.A 185.5 mm long flight tube is attached after the lens electrodes to provide a field-free expansion region for the negative ion sphere, a so-called Newton sphere.At the end of the flight tube, two microchannel plates (MCPs) are placed in chevron configurations.The effective diameter of these MCPs is 40 mm.The flight tube and front MCP are set at the same electric potential, while at the end of the MCPs, a phosphor screen is positioned and kept under UHV conditions.Behind the phosphor screen, a charge-coupled device camera is placed outside the UHV chamber behind an optical viewport to capture the image.
In the basic experimental procedure, a pulsed electron beam was applied for 200 ns with a repetition rate of 5 kHz.Electrons travel through the interaction region, where they collide with the molecular beam, resulting in various DEA fragments.After the electron beam pulse, a negative pulse with a 150 ns delay, 120 V amplitude, and 21 µs width is applied to the repeller plate that pushes negative ions from the interaction region to the VMI spectrometer.This delay allows for better kinetic resolution in the time-sliced images and prevents the electrons from reaching the detector [4].The spectrometer is designed to maintain the VMI condition, i.e., all the ions with a given directional velocity are mapped to a single point on the detector regardless of their origin.Several combinations of voltage values were tested using SIMION 8 to simulate the performance of the VMI spectrometer [21].In the simulations, the extractor plate was kept grounded, and a DC repeller voltage was applied.A representative outcome of the simulation with the specific voltage configuration is presented in figure 1.The kinetic energy distribution and angular distribution of the negative ions can be obtained from the projection of the Newton sphere onto the detector.Ions with higher kinetic energy fall onto the detector, forming an image with a larger diameter.The central slice of the Newton sphere represents the initial kinetic energy of the fragment ions.We recorded 50 ns time-sliced images from the central area of the Newton sphere, where the full width at half maximum of the Newton sphere is approximately 200 ns.To obtain the central slice, a gate module was used to pulse the rear MCP.A transistor-transistor logic pulse produced from a digital delay generator with the requisite delay was used to trigger the gate module.To calibrate the electron energy scale, we used the well-documented resonance peaks of O − ion yields from DEA to the oxygen molecule at 6.5 eV [22].Similarly, the kinetic energy distribution measurements were calibrated using the kinetic energy of the oxygen anions from O 2 at the same resonance energy [23].Additionally, both calibrations were confirmed by measuring the O − ion kinetic energy from DEA to CO 2 at 8.2 eV [24].
The signal detected on the screen was also used to record the ion yield curves of the negative fragments.First, a fast amplifier magnifies the AC-coupled screen signal and feeds it to a constant fraction discriminator (CFD).The CFD output is used as a stopping pulse to the Nuclear Instrumentation Module standard time-to-amplitude converter (TAC), and a starting pulse is generated by a master pulse generator (SRS DG645) that is synchronized with the electron gun pulse.The time difference between the starting and finishing pulses is the time-of-fight of the ions.The TAC's output is connected to a multichannel analyzer, which is then communicated to a computer via a USB 2.0 interface for data acquisition.Finally, we used a custom-built LabVIEW-based data acquisition system to construct ion yield curves for specific fragments.
The experiments were performed under UHV conditions with a base pressure as low as 10 −9 mbar and a working pressure of 5 × 10 −7 mbar with 99.7% pure acetic acid purchased from Sigma Aldrich, USA.
Quantum chemical calculations were performed using the GAUSSIAN 16 software [25].Thermodynamic threshold energies for each dissociation channel were obtained from the composite W1 method [26] using the B3LYP functional [27] and the flexible cc-pVTZ basis set [28] for geometry optimizations and basic thermochemistry, coupled with a series of correlation corrections [29] to ensure high-level energetic accuracy.The threshold energy of a particular dissociation channel of acetic acid was obtained by calculating the bond dissociation energies involved and the electron affinity of the fragment, which was detected mass spectrometrically in this study.

Results and discussion
Earlier reports [15][16][17][18] on DEA to gas-phase acetic acid indicate many different dissociation channels, as shown in reactions (1)-( 10): Out of these ten channels, three reactions (1), (3), and (4) exhibit low-energy resonant features in the range of 0-2 eV, while the remaining channels come from the higher-energy resonance over 5 eV.In this work, the ion yields as a function of electron energy from 2 to 16 eV were recorded, and VMI images were obtained for the fragment ions with sufficient signals.

Low-energy shape resonance
A shape resonance occurs when the incident electron is trapped in the molecule's electronic ground state potential and is named based on the shape of this potential [30].When an electron approaches a nonpolar molecule with nonzero polarizability, a temporary dipole is induced in the molecule, which weakly attracts the electron through a longrange attractive polarization potential.The electron also experiences a repulsive centrifugal potential associated with its angular momentum.The superposition of these two potentials results in the effective long-range interaction potential of the electron-molecule system.For all angular momentum quantum numbers larger than zero, a repulsive centrifugal potential barrier is formed, and the electron can tunnel through the barrier and be temporarily trapped therein.This type of resonance is called single-particle resonance.In the present case, the isolated gas-phase acetic acid molecule is computed to already have a dipole moment of magnitude 1.7 Debye, oriented parallel to the carbonyl bond, and this dipole can be reinforced by the polarizing field of the impinging electrons.Most of the shape resonances observed so far occurred at low energies (0-4 eV), exhibiting a lifetime of 10 −15 -10 −10 s or even longer, resulting in decay into vibrational and rotational states of molecules or fragmentation by DEA into anionic and neutral fragments [31].In the case of acetic acid, within the 0-2 eV region, the presence of two shape resonances was observed by Sailer et al [17].They recorded the ion yield curve of CH 3 COO − with a strong peak near 1.5 eV and a small hump near 0.75 eV [15][16][17].Formation of CH 3 COO − through the DEA process involves the direct cleavage of the O-H bond, and this fragment was observed to be the most dominant ion detected.Later, the absolute DEA cross section of the CH 3 COO − ions at the 1.5 eV resonance was measured by Prabhudesai et al [18] using the well-known relative flow technique [32,33].Previous experiments also indicate that the low-energy resonances are responsible for the formation of CH 2 O 2 − and HCOO − as well [17].Moreover, Sailer et al [17], on the basis of ab initio calculations, attempted to identify electronic transitions associated with the 0.75 eV and 1.5 eV resonances.The 0.75 eV resonance was assigned to the LUMO, while the 1.5 eV resonance was assigned to the next highest orbital (LUMO + 1).However, a theoretical study by Freitas et al [19] using the Schwinger multichannel method refuted this assignment, suggesting that the LUMO corresponds to the latter resonance that formed via electron capture into the π * state with A ′′ symmetry.In this work, two ion yield curves for CH 3 COO − and HCOO − show low-energy signals near 2 eV.However, they will not be discussed here because they represent only a partial feature of their resonances, which were observed in the previous study at 1.5 eV using an electron monochromator capable of reaching energies as low as 0 eV [17].In addition, we did not observe any measurable signal for the CH 2 O 2 − ions because they were reported to have a resonance at 0.75 eV [17], far below the minimum electron beam energy produced by our electron gun.

Core-excited Feshbach resonances
When an electron is captured and simultaneously electronically excites a neutral molecule, a so-called core-excited resonance is formed [31].The excitation of the orbital electrons causes less screening of the molecular nuclei; thus, the incident electron experiences a slightly positive charge and finds itself with insufficient energy to escape from the field of the excited molecule.The electron becoming temporarily bound to the molecule forms a temporary negative ion (TNI), which is energetically below the corresponding excited parent neutral state but above the ground neutral state [31].This type of resonance generally occurs slightly below the ionization energy of the parent molecule.Hence, the broad 10 eV resonant feature observed in DEA to acetic acid for several fragmentation channels can be assigned to be of the core-excited Feshbach type.Previous reports indicate that seven dissociation channels, i.e. reactions (2) and ( 5)- (10), occur near 10 eV; therefore, they can be associated with this type of resonance.Moreover, only two channels responsible for the formation of O − and H − , i.e. reactions ( 8) and (10), respectively, were observed at 5-8 eV [17,18].The 10 eV Feshbach resonances could be associated with a Rydberg excitation in the molecule, whereas the 5-6 eV region correlates to the first electronically excited state of neutral acetic acid, as suggested previously [17].In addition, several vacuum ultraviolet absorption spectroscopic studies of acetic acid have been reported, which indicate Rydberg transitions within this energy range [34,35].These Rydberg states could act as an initial molecular state for the Feshbach resonance present in the 10 eV region.On the other hand, the first and second ionization energies (IEs) of acetic acid are 10.65 and 12.1 eV and involve an excitation from the highest occupied molecular orbital (HOMO) 13a ′ and (HOMO-1) 3a ′′ , respectively [35].The relation between the IE of a molecule and the energy of a Feshbach resonance can be stated as EF = A × IE + B, where EF denotes the Feshbach resonance energy, A is the slope close to a value of unity, and B is a constant that depends on the resonance configuration involved in the transition [36].If the Feshbach resonance is composed of two Rydberg-like electrons bound to a positive ion core with an s 2 configuration, B is −3.9; if it has a p 2 configuration, B is −1.8.This relation indicates that the typical Feshbach resonance energy of a molecule can be predicted to be at least 2 eV below the IE of the corresponding molecular orbital, which is involved in the transition, i.e., in the present study, (HOMO-1) 3a ′′ is favored to be involved in the transition, and it is a p 2 -Feshbach.Further, the absorption spectroscopic studies of Leach et al [35] established the presence of 3a ′′ → 3 pa ′′ Rydberg transitions near the 9.268 eV range.This Rydberg state could also eventually act as a parent state for the 10 eV resonance observed in DEA to acetic acid.The following sections describe ion yields and VMI results of different DEA channels observed from the core-excited resonances.

CH 2 COO
− .Figure 2 shows the combined ion yield of CH 3 COO − and CH 2 COO − ions.As already discussed above, the 2 eV peak is due to the formation of CH 3 COO − ions, whereas the 9.6 eV peak is due to the CH 2 COO − ions [17].This CH 2 COO − ion can potentially be formed as three-or two-body dissociation, as shown by reactions (2a) and (2b), respectively.Reaction (2a) represents the dissociation channel, which includes the cleavage of O-H and C-H bonds.If the neutral counterparts are formed in the electronic ground state, the obtained threshold of this reaction channel is calculated to be 5.71 eV.In the case of reaction (2b), the two-body dissociation for this fragment would require the cleavage of O-H and C-H bonds and the simultaneous formation of an H-H bond.The calculated thermodynamic threshold for this channel is relatively low, 1.21 eV.Therefore, both reactions are energetically possible; thus, the 9.6 eV peak can be associated with the three-and/or two-body dissociation.However, the higher energy difference between the theoretical threshold and the experimentally observed appearance energy (7.5 eV) implies that reaction (2b) is less favorable unless this excess of available energy can be distributed as kinetic or rovibrational energy of the fragments.To measure the kinetic energy of the CH 2 COO − ions, time-sliced images were taken near the 9.6 eV resonance.Figure 3 shows the timesliced images of the CH 2 COO − ions at three incident electron energies.The time-sliced images show the maximum intensity at the center with an isotropic distribution.This result indicates that the kinetic energy of the fragment ions is near 0 eV.In a process where low-kinetic energy ions are formed, the flat time-sliced image does not accurately reflect the ion kinetic energy distribution, as it overestimates the abundance of lower kinetic energy ions compared to the higher kinetic energy ions [37].Hence, to extract an accurate kinetic energy value, we applied a weighted factor to the ion counts, i.e. we multiplied the ion counts by the square of their corresponding distance from the center.To verify the correct practicality of using a weighted factor, before the present study, we applied this methodology to the O − kinetic energy distribution of DEA to CO 2 and compared our results with previous measurements [37], and was also recently applied for our studies on DEA to ethanol [38].Figure 4 shows the weighted kinetic energy distribution showing a peak near 0.1 eV.The maximum peak of kinetic energy does not shift considerably with increasing incident electron energy between 9-10 eV.However, the kinetic energy distribution becomes broader with increasing electron energy.This observation indicates that only a fraction of the excess available energy of the dissociation process is deposited as the kinetic energy of the fragments, whereas the remaining energy is distributed as the rovibrational excitation.This lower kinetic energy of the CH 2 COO − ions near the 9.6 eV resonance also indicates that the involvement of three-body dissociation, as shown by reaction (2a), is more plausible.

HCCO − and CCO − .
Figure 5 shows the ion yield curve for HCCO − and CCO − combined signals with a broad resonance peak centered near 10 eV and a small hump within the 12-14 eV energy region.As mentioned previously, because of the limited mass resolution of the VMI spectrometer, we could not separate these two fragments.This observation agrees well with the previous study, which reported a peak at 10 eV for HCCO − and a slightly lower-intensity peak at 11 eV for CCO − [17].Because multiple bonds are cleaved to form these anions, an unambiguous assignment to the appropriate dissociation channel is challenging.Previous reports suggested the dissociation channels for these two  fragments, as stated by reactions ( 5) and ( 6) [17].Both these channels would require structural rearrangements of the TNI state.Formation of the HCCO − ion involves the cleavage of a C-OH bond and two C-H bonds and the formation of an O-H bond.Meanwhile, the formation of CCO − ions involves the cleavage of one C-OH bond and three C-H bonds and the formation of an O-H bond.Our theoretical calculations indicate that the thresholds for these two channels are 3.69 and 3.55 eV, respectively.Figure 5 shows that the experimental appearance energy is near 8 eV.This energy difference suggests that all the fragments produced in these two dissociation channels are energetically hot, i.e. possess relatively high kinetic energy and/or internal excitation.Therefore, the time-sliced images of these fragments over the resonance can be useful for understanding the dynamics of this process.In figure 6, the time-sliced images of the combined signal for HCCO − and CCO − fragments show an isotropic signal distribution with a maximum intensity at the center.Figure 7 shows the kinetic energy distributions that we extracted from these time-sliced images.A broad distribution with a kinetic energy peak at a maximum of 0.2 eV indicates that the fragments formed in the dissociation process are in rovibrational excited states.Figure 8 shows the ion yield curve with a broad resonance extending from 4 to 16 eV, which combines signals from O − and OH − ions.This wide spread of the resonance peak corresponds well to the previous DEA study by Sailer et al [17] who reported ion yields for both fragments separately.They found a broad resonance located at 10 eV for the OH − ions and two resonances near 5 and 9 eV, along with a smaller intensity peak near 7.5 eV for the O − ions [17].Due to their low cross-sections, these closely lying resonances cannot be well separated in our experiment and resulted as a broad ion signal.The dissociation channel for the OH − ions involved the cleavage of the C-OH bond, as shown in reaction (7).The theoretical threshold energy for this dissociation channel is 2.92 eV.In contrast, O − ion formation is ambiguous because the oxygen ion can originate from the hydroxyl group or the carbonyl group as two-body dissociation, as shown by reactions (8a) and (8b), or as  three-body dissociation involving subsequent H loss from the cleaved hydroxyl group (reaction (8c)).Dissociation from the carbonyl group involves C=O bond cleavage with a threshold energy of 6.09 eV.Dissociation from the hydroxyl group is more complex, as mentioned above, because it could be from two-body dissociation through rearrangements in the TNI by forming CH 3 C(=O)H as a neutral conjugate (reaction (8a)), or it could be from a three-body dissociation channel by forming CH 3 CO and H as neutral conjugates (reaction (8c)).The threshold energies for these two-and three-body dissociation channels are 3.88 eV and 7.74 eV, respectively.The two O − peaks near 5 and 9 eV observed in the previous DEA study could result from these two-and three-body dissociation processes, respectively.A detailed investigation using different isotopes would be necessary to validate this argument, which is beyond the aim of the present study.Because of the low cross sections for both channels, we could not perform VMI measurements with a sufficient signal-to-noise ratio.ions with a resonance peak near 10 eV, as also observed in the previous study [17].The formation of this ion can involve two possible dissociation channels.Reaction (9a) requires cleavage of a C-C bond and a C-H bond.This three-body dissociation process can occur with a thermodynamic threshold of 8.28 eV, which is very near the appearance energy of the CH 2 − signal.
In the previous DEA report, Sailer et al [17] identified the formation of CH 2 − as a two-body dissociation with HCOOH as a neutral counterpart.The threshold energy for this two-body dissociation process is computed to be 4.58 eV.This result suggests that the two-and three-body dissociation processes could have contributed to the formation of this fragment.As with O − and OH − ions, the low cross section prevented VMI measurements for this fragment.

Conclusion
We have investigated DEA to gas-phase acetic acid using a recently constructed VMI spectrometer.The ion yield curves of several fragment ions were recorded and agreed with previously reported results [15][16][17].Although two resonances near 2 and 10 eV were observed, we focused our discussion on the latter resonance and obtained time-sliced images for it.On the basis of these images, we obtained kinetic energy distributions of the fragment anions that showed a broad nearzero kinetic energy distribution.Our analysis indicated that acetic acid dissociates through a three-body (or potentially even higher-order) dissociation or through a two-body dissociation with considerably highly rovibrationally-excited fragments.However, to identify the exact dissociation channel, other techniques for neutral detection must be applied, which is challenging for molecules with low DEA cross sections [39].Nevertheless, to assist in assigning dissociation channels for each fragment, we performed thermodynamic threshold calculations.This study of DEA to acetic acid using VMI contributes toward reinforcing and extending our knowledge on this process and, more generally on the effects of LEEs that are valuable to many areas, encompassing radiation from natural phenomena to industrial processes.

Figure 1 .
Figure 1.Schematic of the VMI spectrometer: (right) a photo and (left) a SIMION screenshot of the simulated path of the O − ions with an initial kinetic energy of 2 eV.The voltage conditions are: repeller, −120 V, extractor, 0 V, lens assembly (25 V, 50 V, 75 V, and 100 V), flight tube, 550 V, and detector, 550 V.The red curves represent the electric field lines after applying the electrode voltages arranged to satisfy the VMI conditions and for 4π collection efficiency.The red dots represent positions of the fragment ions taken at 0.5 µs intervals.The effect of the magnetic field is not considered during the simulation.

Figure 2 .
Figure 2. Ion yield curve with a combined signal for CH 2 COO − and CH 3 COO − ions.The black circles are the experimental data points, and the blue solid line is to guide the eye.

Figure 3 .
Figure 3. Time-sliced images of CH 2 COO − ions at different incident electron energies.The electron beam direction is from bottom to top.

Figure 4 .
Figure 4. Weighted kinetic energy distribution of CH 2 COO − ions at different incident electron energies.

Figure 5 .
Figure 5. Ion yield curve with a combined signal for HCCO − and CCO − ions.The black circles are the experimental data points, and the blue solid line is to guide the eye.

Figure 6 .
Figure 6.Time-sliced images of HCCO − and CCO − ions at different incident electron energies.The electron beam direction is from bottom to top.

Figure 7 .
Figure 7. Weighted kinetic energy distribution of HCCO − and CCO − ions at different incident electron energies.

Figure 8 .
Figure 8. Ion yield curve with a combined signal for O − and OH − ions.The black circles are the experimental data points, and the blue solid line is to guide the eye.

Figure 9 .
Figure 9. Ion yield curve for CH 2 − ions.The black circles are the experimental data points, and the blue solid line is to guide the eye.