Fragmentation dynamics of CH₃Cl^q+ (q = 2,3): theory and experiment

Abstract

A combined theoretical and experimental study of the dissociation of the di- and trication of the CH₃Cl molecule has been performed. Experimentally, these multi-charged ions were produced after interactions of a CH₃Cl effusive jet with a mono-energetic beam of H⁺ or Ar⁹⁺ projectile ions. Theoretically, we mapped the multi-dimensional potential energy surfaces of CH₃Cl²⁺, H₂CClH²⁺ and CH₃Cl³⁺ species in their electronic ground and electronically excited states using post-Hartree–Fock configuration interaction methods. In addition to the obvious bond-breaking ionic fragments (i.e. H⁺ + CH₂Cl⁺, H⁺ + CH₂Cl²⁺ and CH₃⁺ + Cl⁺), the formation of H₂⁺ (+CHCl⁺ or CHCl²⁺), H₃⁺ (+CCl⁺) and HCl⁺ (+CH₂⁺) was observed upon bond rearrangement after ion impact of CH₃Cl. The interaction strength of the incident projectiles is found to affect the relative yields on the observed dissociation channels, however, it has no effect on the kinetic energy releases of the fragmentation pathways. For the observed dissociation channels, plausible formation mechanisms were proposed. These reaction pathways take place on the ground and/or electronic excited potential energy surfaces of the doubly and triply charged CH₃Cl ions, where spin–orbit and vibronic couplings are in action. Moreover, this work suggests that the mechanisms undertaken may depend on the multiply charged ion preparation by valence or inner-shell single photon photoionization, fast ion beam impact or ultrafast intense laser ionization.

---

I Introduction

The formation, dissociation and isomerization of hydrocarbons have been a subject of numerous studies for many decades.^1–3 Of these, the ionization and dissociation of chloromethane (CH₃Cl) has attracted wide attention as it is one of the most abundant chlorine containing molecules in Earth's atmosphere. Recently, CH₃Cl became the first organohalogen to be detected in the interstellar space.⁴

A number of studies have been performed on the chloromethane molecule under intense laser fields,^5–9 both valence and core photoionization,^2,10–17 ion impact,^3,18 electron impact^19,20 and double-charge-transfer spectroscopy.²¹ For instance, Ma et al.⁶ treated in 2017 the Coulomb explosion and fast H-atom rearrangement after intense laser field ionization of CH₃Cl. They have determined the kinetic energy releases (KERs) and the yields of two body dissociation channels of di- and tri-cations of CH₃Cl. Back in 1981, Monce et al.¹⁸ measured the dissociation energies for the fragments of CH₃Cl²⁺ under 1 MeV ion impact. Siegmann et al.³ reported the fragmentation patterns of CH₃Cl in collisions with swift ions and have measured the corresponding KERs. Aitken et al.¹⁹ have studied the electron impact ionization of oriented CH₃Cl molecules. Rejoub et al.²⁰ used electrons from threshold up to 1000 eV to measure the partial ionization yields of CH₃Cl using coincidence measurements. These works allowed identifying H⁺, H₂⁺, H₃⁺, C⁺, CH₂⁺, CH₃⁺, and Cl⁺ fragments and their respective kinetic energy spectra were recorded. Nevertheless, the relative yields of the different dissociation channels were not determined using projectile ion impact.

Concerning photoionization studies and the subsequent fragmentation of the doubly charged CH₃Cl species, several works were carried. Back in 1990, Ruhl et al.² have used single photon ionization of the CH₃X (X = Cl, OH, SH, NH₂, CN, CHO) molecules to study the dissociation of their doubly charged ions using the electron-ion-ion coincidence technique. They have determined the KERs for different fragmentation channels of CH₃Cl dication and their relative intensities. They were the first in reporting the production of H₃⁺ from methyl compounds and some evidences for rearrangement processes via ylide dications prior to charge separation. In 1994, Thissen et al.¹⁶ have used photoexcitation near Cl (2p) to measure the total photo-electron yield using charge transfer mass spectrometry. Later on, Miron et al.⁸ have used electron-ion coincidence technique in combination with high resolution spectroscopy to disentangle the complex competing decay processes leading either to HCl⁺ or to Cl⁺ after Cl (2p) CH₃Cl excitation. This work was complimented later by Céolin et al.¹⁵ who have measured the partial ion yields around the Cl (2p) and C (1s) core level thresholds in CH₃Cl. Interestingly, they have reported that the vibrational population of the intermediate state does not affect the H-atom isomerization processes. They have further showed that the formation of H₂⁺ and H₃⁺ is suppressed above C (1s) threshold; however, their formation increases above the Cl (2p) threshold. More recently, Kokkonen et al.^10,11 have recorded the photoions in coincidence with energy selected electrons to study the resonant Auger decay after removal of an electron from Cl (2p). They have also performed qualitative theoretical simulations to study the formation of HCl⁺. Using the photo-electron-photo-electron coincidence technique Hult Roos et al.¹³ have measured the vertical double ionization energies for methyl halides. For CH₃Cl, a vertical double ionization of 31.5 eV was determined.

Theoretically, Duflot et al.^22,23 have done a detailed study on the CH₃Cl²⁺ isomers and of its fragmentation channels, where they mapped the ground potential energy surface. Indeed, they characterized the stable isomers and the transition states leading to the major products at the Complete Active Space Self Consistent Field level of theory. Mainly, two isomers are found: a C_3v form (denoted as CH₃Cl²⁺) and a planar isomer (denoted as H₂CClH²⁺) (Fig. 1). They also determined the KERs associated with the fragmentations of CH₃Cl²⁺. Moreover, Walt et al.⁵ have done ab initio calculations to obtain the KERs for CH₃X²⁺ (X = F, Cl, Br, I) decomposition into CH₃⁺ + X⁺. Recently, Palaudoux and Hochlaf²⁴ have performed multi configuration interaction ab initio calculations to study the formation of H₃⁺ after dissociation of CH₃Cl²⁺. They have suggested a new multi-step mechanism for H₃⁺ formation where the three H-atoms in CH₃Cl approach each other along with singlet-triplet spin–orbit and internal conversions. This mechanism is in contrast to the H-migration and H₂ roaming mechanisms prior to H₃⁺ formation in alcohol molecules as suggested by Ekanayake et al.²⁵ Prior to that Grant et al.²¹ determined the double ionization energy of CH₃Cl using the second order algebraic diagrammatic construction Green's function method and Roithova et al.²⁶ investigated the potential energy surface of the reaction between CHCl²⁺ and H₂ using the Coupled Cluster approach. Nevertheless, none of previous theoretical works investigated the electronic excited states of CH₃Cl²⁺ and their possible role on the further evolution of this ion till dissociation.

Fig. 1 C _3v (left) and planar (right) stable forms of chloromethane doubly charged ions as computed by Duflot et al.^22,23 and confirmed presently at the RCCSD(T)/aug-cc-pVTZ level.

Despite of the large number of studies on the CH₃Cl²⁺ unimolecular decompositions, a detailed analysis of the fragmentation channels of CH₃Cl²⁺ considering the involved electronic states have not been performed so far. In particular, nothing is known on the possible role of the electronic excited states of CH₃Cl²⁺ since these previous works were treating only the corresponding ground state. The further evolution of the CH₃Cl³⁺ trication is also missing. In the present work, we have performed a systematic study on the fragmentation dynamics of CH₃Cl^q+ (q = 2,3) formed after energetic ion impact of CH₃Cl, where both ionic species resulting from the decomposition of CH₃Cl²⁺ or CH₃Cl³⁺ were detected in coincidence. Moreover, we mapped the potential energy surfaces of the electronic states of these ions along suitable coordinates using post Hartree–Fock configuration interaction approaches. Using the previous and present experimental data, our computational results are used to propose the unimolecular reaction pathways undertaken by the CH₃Cl²⁺ dication leading to the CH₃⁺, Cl⁺, H⁺, CH₂Cl⁺, H₃⁺, CCl⁺, CH₂⁺ and HCl⁺ products and by the CH₃Cl³⁺ trication forming the H⁺ + CH₂Cl²⁺ and the H₂⁺ + CHCl²⁺ ion pairs.

II. Methodologies

a. Experiment

The experiments were carried out at the low-energy ion-beam facility at Inter-University Accelerator Centre, New Delhi, India. The details of the experimental setup were published in ref. 27. Briefly, a mono-energetic beam of the projectile ions interacts with an effusive jet of a target gas kept at room temperature. The projectile ions (0.25 MeV H⁺, 0.45 MeV Ar⁹⁺, 2.25 MeV Ar⁹⁺) are generated using an electron-cyclotron-resonance ion source. The target gas used was CH₃Cl (purity > 99.5%, Sigma Aldrich). The experiments were performed in a high vacuum chamber to maintain single collision conditions. The pressures of the experimental chamber with and without gas load were 6 × 10⁻⁷ Torr and 8 × 10⁻⁸ Torr, respectively. The recoil ions and the ejected electrons produced in these collision experiments are guided using a homogeneous electric field (4.7 kV m⁻¹) toward their respective detectors. The mass resolution (Δm/m) of our TOF mass spectrometer is found to be about 0.0057. The recoil ions after passing through a time-of-flight spectrometer were detected using dual-micro-channel plates equipped with a delay-line position sensitive detector; the ejected electrons were detected using a channeltron. Then, the multi-ion coincidence technique was used to detect the recoil ions with reference to an ejected electron. The typical data acquisition time ranges from 22 hours to 48 hours in order to collect sufficient statistics. A typical TOF spectrum generated in collisions of 0.45 MeV Ar⁹⁺ with CH₃Cl is shown in Fig. 2 and a double ion coincidence spectrum is presented in Fig. 3.

Fig. 2 TOF spectrum (raw) of ions generated in 0.45 MeV Ar⁹⁺ collisions with CH₃Cl. Oxygenated species are originating from NO gas used in a trial run just before starting the experiments on CH₃Cl.

Fig. 3 Double ion coincidence spectrum (raw) in 0.45 MeV Ar⁹⁺collisions with CH₃Cl.

The detection efficiency is almost constant for all ions as the impact energy on the microchannel plate is 2.6 keV for singly charged ions, and even larger for higher charge states. Thus corrections for efficiency for the light ions used in this experiment are not envisaged.

After measuring the time and position information of the recoil ions in coincidence with an electron, the momenta and the kinetic energies of the recoil ions are obtained. The KER for a given dissociation channel is determined by summing the kinetic energies of the corresponding recoil (fragment) ions.

b. Theoretical computations

In order to deduce the energetics of the CH₃Cl^q+ (q = 2,3) molecular systems, we performed geometry optimizations of CH₃Cl, its di- and trications and the charged fragments observed presently. These optimizations were done without constraints (in C₁ point group) and followed by frequency computations to attest the minimal nature of the structures (all > 0 frequencies). These computations were done using the (partially spin restricted) coupled-cluster singles and doubles plus perturbative triples (R)CCSD(T) method^28–31 in conjunction with the aug-cc-pVXZ (X = T,D) basis sets^32–34 of Dunning and co-workers. For the dissociation energies, the (R)CCSD(T)/aug-cc-pVTZ and the (R)CCSD(T)/aug-cc-pVQZ methods lead to close results i.e. ∼0.1 eV energetics differences. Both set of data are accurate enough compared to the present measurements.

We also performed one dimensional cuts of the 10D potential energy surfaces (PESs) of CH₃Cl²⁺ and of CH₃Cl³⁺ by varying one internal coordinate, allowing giving an insight into the evolution of these electronic states not far from the Franck–Condon region accessed after either double or triple ionization of CH₃Cl. We thus computed the one dimensional evolutions of the lowest electronic states of CH₃Cl²⁺ of singlet and triplet spin multiplicities and the isomerization processes between CH₃Cl²⁺ and H₂CClH²⁺. We also calculated the lowest doublet and quartet potentials of CH₃Cl³⁺ by lengthening one CH or the CCl distances. These computations were carried out at the Complete Active Space Self Consistent field (CASSCF)^35,36 followed by the internally contracted Multi Reference Configuration Interaction (MRCI)^37–39 methods as implemented in the MOLPRO program suite.⁴⁰ For C and Cl(H) atoms, the spdf (spd) cc-pV5Z subset were used. The CASSCF active space comprised all valence orbitals and all electronic states having the same spin multiplicity were averaged together. At the MRCI level, all configurations with coefficients larger than 0.005 in the CI expansion of the CASSCF wave functions were taken into account as a reference resulting in more than 6 × 10⁸ uncontracted configuration state functions (CSFs) per C_s point group symmetry to be treated. All valence electrons were correlated. Moreover, we used the CASSCF wavefunctions to evaluate the spin–orbit coupling matrix elements in Cartesian coordinates, where these wavefunctions were used as the multi-electron basis for the two-step spin–orbit coupling calculations^41,42 at the level of Breit–Pauli Hamiltonian.⁴³

For the spin–orbit free energies, the accuracy of the relative energy position of the PESs and the dissociation limits is expected to be better than ∼0.1 eV. Nevertheless, the inclusion of spin–orbit splittings should lead to ∼0.1–0.2 eV levy for molecular species of interest and for > 0.3 eV for heavy atoms such as Cl and its ions. Obviously, this affects the relative energy positions of the PESs and the corresponding dissociation channels. But one should be aware that the error bars of the experimental determinations of the ionization energies, appearance energies and the KERs of the products issued from the CH₃Cl^q+ (q = 2,3) unimolecular decomposition reactions are as large as ±0.5–1 eV i.e. distinctly larger than the accuracy of the present computations. In the following, we will therefore rely on the spin–orbit free energies, except if explicitly specified.

III. Experimental results

We have studied the interaction of the energetic projectile ions (0.25 MeV H⁺, 0.45 MeV Ar⁹⁺, 2.25 MeV Ar⁹⁺) with a CH₃Cl molecule. The following channels are observed for two body dissociation of the CH₃Cl^q+ (q = 2,3) molecular ions formed in these collisions:

CH₃Cl²⁺ → CH₃⁺ + Cl⁺ (1)

CH₃Cl²⁺ → H⁺ + CH₂Cl⁺ (2)

CH₃Cl³⁺ → H⁺ + CH₂Cl²⁺ (3)

We also observed fragmentations that should occur obviously after bond rearrangements. They correspond to these channels:

CH₃Cl²⁺ → H₂⁺ + CHCl⁺ (4)

CH₃Cl³⁺ → H₂⁺ + CHCl²⁺ (5)

CH₃Cl²⁺ → H₃⁺ + CCl⁺ (6)

CH₃Cl²⁺ → CH₂⁺ + HCl⁺ (7)

a. Kinetic energy releases

We have measured the KERs for the dissociation channels of di- and tri-cations of CH₃Cl (see, eqn (1)–(7)). The KER distributions (KERD) for different dissociation channels obtained under the impact of all three projectile ions used in present experiments are similar. For a typical example, the KERDs for the dissociation channels produced under 0.45 MeV Ar⁹⁺ impact are presented in Fig. 4 and for the channels involving rearrangement of atoms are given in Fig. 5. These figures show that whatever the dissociation channel the KERs present a unique peak. This peak has an asymmetric shape with a sharp increase for low energy KERs and then more or less a long tail for high energy KERs. The latter are probably due to the steepness of the associated potential energy curves at small inter-nuclear distances (leading to a large spread in the KER). A detailed comparison of our experimental KERs with those earlier reported in the literature is presented in the discussion section.

Fig. 4 Present experimental KER distributions (black curve) for the dissociation channels of CH₃Cl^q+ (q = 2,3) under 0.45 MeV Ar⁹⁺ along with our theoretical calculations (red dashed lines) and those reported in the literature: Duflot et al.²³ (blue dashed line), Walt et al.⁵ (dashed dotted magenta line). See text for more details.

Fig. 5 Measured KER distributions (black curve) for the bond rearrangement channels observed in the dissociation of CH₃Cl^q+ (q = 2,3) under 0.45 MeV Ar⁹⁺ along with our theoretical calculations (red dashed line). The vertical lines correspond to the theoretical calculations for KERs reported in the literature: Palaudoux and Hochlaf²⁴ (red solid line), Duflot et al.²³ (blue dashed line), Walt et al.⁵ (dashed dotted magenta line). See text for more details.

b. Relative yields of dissociation channels

We have measured the relative yields of the dissociation channels of CH₃Cl^q+ (q = 2,3) by integrating the area under each coincidence island shown in Fig. 3 separately and dividing it by the total counts of all the two body dissociation channels observed in the present experiments (cf.Table 1). Coincidence islands for different two-body dissociation channels (Fig. 3) are extracted after correcting for false coincidences using the conservation of momentum in the centre of mass frame of the parent molecular ion. The statistical uncertainties in measuring these yields are ∼3% and ∼5% for the dissociation channels given in eqn (1)–(3) and for bond rearrangement channels (eqn (4)–(7)), respectively. These errors correspond to the standard deviations.

Table 1 Relative yields (%) of the two body dissociation channels observed in present measurements under the impact of different projectile ions with energy E, velocity v and charge state q. The atomic units [ℏ = e = m = 1 (a.u.)] are used. For velocity, 1 a.u. corresponds to 2.18 × 10⁸ cm s⁻¹

Dissociation channel	Experimental yield (%)
	Present			Literature
	H^+/E = 0.25 MeV/v = 3.16 a.u./q/v = 0.32	Ar^9+/E = 2.25 MeV/v = 1.50 a.u./q/v = 6.0	Ar^9+/E = 0.45 MeV/v = 0.67 a.u./q/v = 13.42	Intense laser^6		Photo-ionization at hν = 40.8 eV^2
				800 nm	400 nm
H^+ + CH2Cl^+	23.6	27.3	44.2	27.2	6.1	13
H2^+ + CHCl^+	2.0	1.7	2.8	4.6	1.3	6
H3^+ + CCl^+	9.1	7.1	4.3	10.8	1.3	9
CH2^+ + HCl^+	6.2	4.4	5.5	6.3	6.6	2
CH3^+ + Cl^+	59.1	59.5	43.1	51.1	84.6	70
H^+ + CH2Cl^2+	0	100.0	86.0	86.1	86.3	0
H2^+ + CHCl^2+	0	0	14.0	13.9	13.7	0

---

Table 1 shows that the most dominant dissociation channel at 0.25 MeV H⁺ and 2.25 MeV Ar⁹⁺ impact corresponds to the CCl bond breaking upon formation of CH₃Cl²⁺ (eqn (1)); the CH bond breaking of this dication (eqn (2)) is also observed abundantly. Interestingly, the relative yield between both channels depends on the interaction strengths (q/v) (Table 1). Indeed, the relative yields for the dominant dissociation channels (CH₃⁺ + Cl⁺ and H⁺ + CH₂Cl⁺) are comparable for q/v = 0.32, 6.0, whereas, the yields for these two dissociation channels are completely different at the highest interaction strength (q/v = 13.42) used here as compared to those observed at lower q/v values. Further, for 0.45 MeV Ar⁹⁺ impact, the relative yields for the channels H⁺ + CH₂Cl⁺ and CH₃⁺ + Cl⁺ are almost the same suggesting that the probabilities of breaking of the C–H and the C–Cl bonds are almost equal. Moreover, Table 1 shows that the other dissociation channels (eqn (3)–(7)) present relative yields of less than 10%, which also strongly depend on the experimental conditions. For instance, the H⁺ + CH₂Cl²⁺ (eqn (3)) ion pair is only observed while using Ar⁹⁺ as projectile. The H₂⁺ + CHCl²⁺ trication dissociation channel is not observed with H⁺ projectile and with low Ar⁹⁺ interaction (q/v = 6.0). Obviously this is due to different CH₃Cl^q+ (q = 2,3) formation/unimolecular decomposition mechanisms in each case.

The relative yields reported earlier in the literature under intense laser fields and photo-ionization of CH₃Cl have also been compiled in Table 1. It can be seen that the trend of the yields of the dissociation channels of CH₃Cl²⁺ is similar in case of 800 nm laser multiphoton ionization and our results at 0.25 MeV H⁺ and 2.25 MeV Ar⁹⁺. However, the relative yields under 400 nm laser multiphoton ionization, 40.8 eV photons ionization and our experiments are entirely different. Again, these results indicate that the formation of different dissociation channels depends on the mechanisms of excitation of the parent molecular ion and in turn on the electronic states populated under the given experimental conditions. For example, in case of intense laser fields experiments, Ma et al.⁶ have observed that the sequential ionization is the dominant source of excitation in CH₃Cl.

For ion impact where the velocity of the projectile ions is comparable to that of the orbital velocity of the electrons, the electron capture is the dominant source of excitation; however, if the velocity of the incident projectile is high enough the ionization becomes the dominant process. In classical limit (v = 1 a.u. = 2.18 × 10⁸ cm s⁻¹), it has been observed that the electron capture takes place dominantly; however, ionization is most prominent process for higher velocities of the projectile (v > 1 a.u.). In the present experiments, both electron capture and ionization phenomena are relevant for the formation of di- and trications of CH₃Cl under our energies of impact. Considering these details, we may suggest that the similar trend of relative yields at 0.25 MeV H⁺ and at 2.25 MeV Ar⁹⁺ is due to the ionization being the dominant process. However, in case of 0.45 MeV Ar⁹⁺, the process of electron capture from the continuum of CH₃Cl may be dominantly in action.

Further, to compare our results with those of Ma et al.,⁶ we have neglected a trication channel (CH₃⁺ + Cl²⁺) from Ma et al., which is not present in our experiments. It is interesting to note that the yields of the two observed channels are similar in case of trication dissociation under ion impact and under two different intense laser fields as reported by Ma et al.⁶ We suggest this is due to the increased Coulomb repulsion in case of CH₃Cl trication which may result in prompt dissociation of the trication. Our theoretical calculations also suggest that the potential energy curves of the trication are mostly repulsive (see below).

It is to be noted here that the CH₂⁺ + HCl⁺ channel is slightly merged with the CH₃⁺ + Cl⁺ channel due to the mass resolution of our spectrometer. However, other dissociation channels reported in this work are well resolved. The overlapping part is neglected in measuring the relative yield for CH₂⁺ + HCl⁺ channel. Therefore, the cross sections for this channel will be higher than the values reported in Table 1; the CH₃⁺ + Cl⁺ yield will reduce accordingly. It is not possible to determine exactly this quantity. Nevertheless, we have estimated crudely that the yields for the CH₂⁺ + HCl⁺ channel may get increased by about 10% at the most, while the corresponding decrease in the yield of the CH₃⁺ + Cl⁺ channel will be only 1%. Yields for the other channels would remain unchanged.

To explicitly observe the effect of the interaction strength on the bond rearrangement process, the relative yields of the dissociation channels involving bond rearrangement are separately reported in Table 2. It is observed, under all projectile ions used in the present experiments, that the yields for the formation of H₃⁺ (eqn (6)) and HCl⁺ (eqn (7)) are higher compared to that of H₂⁺ formation. Also, Table 2 shows that the yield for the H₃⁺ formation channel reduces at highest q/v used in present experiments, whereas the yields for those involving H₂⁺ or HCl⁺ formation increase. The role of electron capture and ionization processes in these dissociations channels will be an interesting study for our future experiments.

Table 2 Relative yields (%) of the dissociation channels involving rearrangement of atoms observed in present measurements under the impact of different projectile ions with energy E, velocity v and charge state q

Dissociation channel	Experimental yield (%)
	Present			Literature
	H^+/E = 0.25 MeV/v = 3.16 a.u./q/v = 0.32	Ar^9+/E = 2.25 MeV/v = 1.50 a.u./q/v = 6.0	Ar^9+/E = 0.45 MeV/v = 0.67 a.u./q/v = 13.42	Intense laser^6		Photoionization at hν = 40.8 eV^2
				800 nm	400 nm
H2^+ + CHCl^+	11.7	12.9	22.2	21.2	14.1	35.3
H3^+ + CCl^+	52.5	53.8	34.2	49.8	14.1	52.9
CH2^+ + HCl^+	35.7	33.3	43.6	29.0	71.7	11.8

---

Ma et al.⁶ have concluded for rearrangement of H-atoms, the height of the potential barriers is more important in determining the relative yields in comparison to the number of the potential barriers. In case of photoionization at 40.8 eV, Ruhl et al.² have proposed that the dissociation of CH₃Cl dications into the fragments in their ground states may occur statistically in respect of their initial energetic requirement.

IV. Theoretical results

We started our investigations by computing the energetics of the CH₃Cl²⁺ doubly charged ion. After geometry optimizations, minimal structures were located for CH₃Cl²⁺(X³A₂) and H₂CClH²⁺(X¹A′) on line with previous works. At the (R)CCSD(T)/aug-cc-pVQZ level, we deduce an adiabatic double ionization energy (ADIE) of CH₃Cl of ∼30.7 eV. This value agrees with the recent experimental ADIE determination of 30.6 eV by Hult Roos et al.¹³ The H₂CClH²⁺(X¹A′) isomer is computes at ∼2.7 eV lower than CH₃Cl²⁺(X³A₂). This energy difference is larger than that found by Duflot et al. (i.e. ∼1.8 eV).^22,23 Our value should be viewed as more accurate. For the fragmentation channels, we compute the CH₃⁺(X¹A′) + Cl⁺(³P) dissociation limit at ∼26.1 eV, CH₂⁺(X²A₁) + HCl⁺(X²Π) at ∼26.9 eV and CH₂Cl⁺(X¹A₁) + H⁺(¹S) at ∼26.6 eV with respect to neutral CH₃Cl(X¹A₁) minimal energy.

We also investigated the intramolecular isomerization of CH₃Cl²⁺ into H₂CClH²⁺. Fig. 6 gives the one dimensional evolutions of the lowest singlet and triplet electronic states of CH₃Cl²⁺ along the distance d, where we moved the H atom from the left to the right. For d close to 0 bohr, the CH₃Cl²⁺ can be found and for d close to 3 bohr the H₂CClH²⁺ isomer is formed. The curves on this figure do not correspond to the minimal isomerization energy paths but they connect the electronic states of the C_3v isomer to the trans planar form. Fig. 6 shows that the electronic ground state of CH₃Cl²⁺ is of X³A₂ nature and its of X¹A′ for H₂CClH²⁺, corresponding to the lowest ¹A₁ singlet state of CH₃Cl²⁺. Also, this figure reveals a high density of electronic states favoring the mixing of the wavefunctions of these electronic states by vibronic interactions (e.g. between ¹E and ¹A₁) and spin–orbit conversions between the singlets and the triplets. For instance, the spin–orbit coupling between the lowest ³A₂ and lowest ¹A₁ electronic states at their respective crossing (i.e. d ∼ 1.9 bohr) is calculated to be ∼290 cm⁻¹.

Fig. 6 spd(f) cc-pV5Z MRCI one-dimensional cuts of the PESs of the lowest singlet and triplet states of CH₃Cl²⁺ lying in the 0–6 eV internal energy range as function of d which is the position of the H moving from the methyl group to Cl atom. These curves are given relative to the CH₃Cl²⁺(X³A₂) minimum. The remaining coordinates are set to R_CCl = 3.453 bohr, R_CH = 2.084 bohr and θ_H–C–Cl = 116°.

The CH₃Cl²⁺(¹A′) → H₂CCl²⁺ (¹A′) isomerization process was already investigated by Duflot et al.²² while the reaction is taking place in the lowest singlet potential. Indeed, these authors identified a reaction coordinate as a combination of in-plane CClH bending mode and H migrating atom out of the plane of the molecule. At the CASSCF level, a transition state was located at ∼0.73 eV, which is in close agreement with the potential barrier we found along the ¹A₁ potential energy surface as depicted in Fig. 6. Nevertheless, Duflot et al.²² were not able to determine a reaction path and hence a transition state of the triplet reaction. This agrees with the present computations. For instance, Fig. 6 shows that the triplets have potential wells in the CH₃Cl²⁺ configurations and not in the H₂CClH²⁺ side. So no transition states can be found in the ground nor in the electronic excited potentials.

During the CH₃Cl²⁺ → H₂CClH²⁺ isomerization, the CH₃Cl²⁺(X³A₂) dication needs to be converted into the lowest singlet PES before reaching the H₂CClH²⁺(X¹A′) valley. The isomerization occurs along a coordinate, which involves the d distance and most likely a H₂CCl out-of-plane coordinates (leading to planarity of this part of the molecule as in H₂CClH²⁺). The reaction occurs along the triplet potential of CH₃Cl²⁺ and then the H₂CClH²⁺ isomer is formed after a spin–orbit conversion at the crossing between these triplet and singlet states i.e. for d distances of about 1.9 bohr. This is strongly favored since the corresponding spin–orbit integral is large (of ∼290 cm⁻¹). Moreover, the H₂CClH²⁺(X¹A′) minimum may be reached from the electronic excited states of CH₃Cl²⁺ (for internal energies > 0.7 eV). Multi step mechanisms are expected involving the vibronic and spin–orbit interactions between the electronic states of Fig. 6 mentioned above. For instance, large amplitude motions on the second ³E state of CH₃Cl²⁺ should induce isomerization to form the corresponding state of H₂CClH²⁺. After vibronic interaction between the first and second ³E states, the spin–orbit conversion of H₂CClH²⁺ (1³E) ions forms the H₂CClH²⁺ (X¹A′) ions.

The decomposition of the CH₃Cl²⁺ dication leads to the CH₃⁺ + Cl⁺ ions, which are produced after cleavage of the CCl bond. For illustration, Fig. 7 shows the spd(f) cc-pV5Z MRCI one dimensional cuts of the electronic states of CH₃Cl²⁺ along the CCl coordinate and where the remaining internal coordinates are kept fixed at their equilibrium values in CH₃Cl(X¹A₁) [R_CCl = 3.45 bohr, R_CH = 2.084 bohr, HCH = 116°]. This figure presents also the one dimensional evolution of the CH₃Cl²⁺ electronic states by stretching one CH bond whereas the other internal coordinates are kept fixed at their equilibrium values in CH₃Cl²⁺ ground state.

Fig. 7 spd(f) cc-pV5Z MRCI one-dimensional cuts of the PESs of the singlet and triplet states of CH₃Cl²⁺ lying in the 0–10 eV internal energy range. These curves are given relative to the CH₃Cl²⁺ (X³A₂) minimum. Left: These curves are given as function of r_CCl which is the distance between the Cl atom and the C atom. r_CH (=2.084 bohr) and θ_H–C–Cl (=116°) are set to their equilibrium values in the neutral molecule. Right: These curves are obtained by lengthening one CH distance (r_CH). Remaining r_CH distances are set to 1.910 bohr) and CCl distance to 3.375 bohr, and θ_H–C–Cl to 108°) i.e. their equilibrium values in CH₃Cl²⁺ ground state.

Several deep potential wells are computed for CH₃Cl²⁺ dication along the CH or the CCl distances (Fig. 7). For instance, this is the case for the X³A₂, ¹E, ¹A₁, ¹A₂, ³E and ³A₁ electronic states, for which large potential barriers are separating the dissociation asymptotes from the corresponding potential wells. Thus, long-lived CH₃Cl²⁺ dications may be found there. The other dicationic electronic states are repulsive in nature. The identification of these states corresponds to that already discussed by Grant et al.²¹ and by Hult Roos et al.¹³

Upon CH₃Cl²⁺ → H₂CClH²⁺ isomerization, the electronic states of the H₂CClH²⁺ species can be populated. Fig. 8 shows the evolution of the H₂CClH²⁺ electronic states along the central CCl bond, where the remaining internal coordinates are kept fixed at their equilibrium values in H₂CClH²⁺ ground state. By lengthening the CCl bond, we form the CH₂⁺ + HCl⁺ fragments. This figure reveals that the ground potential exhibits a potential well of >2 eV, which is separated from the lowest dissociation channel to which it correlates adiabatically by a potential barrier. We may locate long-lived H₂CClH²⁺ ions in this potential well. Besides, the upper states are mostly repulsive in nature or presenting shallow potential wells.

Fig. 8 spd(f) cc-pV5Z MRCI one-dimensional cuts of the PESs of the singlet and triplet states of H₂CClH²⁺ by lengthening the CCl distance (r_CCl). The r_CH distances are set to 2.112 bohr and the r_ClH distance to 2.647 bohr, and the angles are set to their equilibrium values in H₂CClH²⁺ ground state. These curves are given relative to the CH₃Cl²⁺ (X³A₂) minimum energy.

Fig. 9 shows the one dimensional cuts of the 10D-PESs of the lowest doublet and quartet states of the CH₃Cl³⁺ trication. Although some shallow potential wells can be observed, no stable structures were found in the lowest doublet or quartet PESs. Indeed, these potentials are mainly repulsive and lead to the [H + CH₂Cl]³⁺ and [Cl + CH₃]³⁺ fragmentation channels. Moreover, this figure reveals a high density of electronic states that favors their mutual interactions, which should lead to the population of the lowest doublet PES by vibronic (between states having the same spin-multiplicity) and/or spin–orbit (between states of different spin multiplicities) interactions.

Fig. 9 spd(f) cc-pV5Z MRCI one-dimensional cuts of the PESs of the doublet and quartet states of CH₃Cl³⁺ by lengthening the CCl distance (r_CCl) (top) and by lengthening the CH distance (r_CH) (bottom). The remaining coordinates were set to their equilibrium values in CH₃Cl (X¹A₁). These curves are given relative to the energy of the lowest CH₃Cl³⁺ doublet state at the CH₃Cl (X¹A₁) equilibrium.

We evaluated the vertical triple ionization energy (VTIE) of CH₃Cl as the difference between the energies of the tricationic lowest doublet state and that of the neutral CH₃Cl species where both species were taken at the CH₃Cl (X¹A₁) equilibrium geometry. At the (R)CCSD(T)/aug-cc-pVTZ level, VTIE is equal to 61.4 eV. At the same level of theory, we locate the H⁺ + CH₂Cl²⁺ channel at 46.2 eV, the H₂⁺ + CHCl²⁺ channel at 46.0 eV and the CH₃²⁺ + Cl⁺ channel at 51.2 eV, respectively. This results in KERs of ∼15.4 eV for both channels upon dissociating CH₃Cl³⁺ on the lowest doublet PES.

V. Discussion

The current theoretical computations are of higher level than those carried out previously. For instance, the computations by Duflot et al.^22,23 were done at the CASSCF and CIPSI levels of theory with relatively small basis sets for the description of the atoms. However, we performed extensive and much larger computations using the (R)CCSD(T) and MRCI techniques in conjunction with large basis sets. Also, Duflot et al.^22,23 computations consisted on the determination of stationary points on the lowest singlet and triplet potential energy surfaces, whereas we mapped a large number of potential energy surfaces of diverse di- and tricationic electronic states and probed their evolutions till reaching the dissociations, in particular their crossings and avoided crossings. Therefore, our calculations are comparatively more accurate and more informative to derive the appearance energies and the related KERs and the subsequent dissociation mechanisms.

In ion collisions with a molecule, we do not have control over the energy deposited into the molecular system and all energetically accessible states may be populated. This results a priori in more than one KER value. However, due to the experimental resolution of the present setup we record a broad distribution for the KERs (see Fig. 4 and 5). To compare our experimental data with the earlier reported experimental and theoretical results, the peak values of the KERDs of the present measurements are reported in Table 3. The errors involved in the measurement of KERs are ≤0.6 eV in case of dications and ≤2.3 eV in case of trications of CH₃Cl. Table 3 lists also the theoretical results from the present calculations and those earlier reported in the literature. These theoretical data are shown in Fig. 4 and 5 as vertical lines.

Table 3 Comparison of the measured KER (peak) values for the dissociation channels of CH₃Cl^q+ (q = 2,3) with our theoretical calculations and those reported in the literature

Dissociation channel	KER (eV)
	Experiment						Theory
	This work			Ref. 2	Ref. 5	Ref. 6
	0.25 MeV H^+	0.45 MeV Ar^9+	2.25 MeV Ar^9+				This work	Ref. 23	Ref. 5
a Ref. 24. b See text for details.
H^+ + CH2Cl^+	6.3	5.8	6.2	5.2		3.9	6.4	5.55
H2^+ + CHCl^+	5.8	5.2	5.6	2.8		3.6	—	5.52
H3^+ + CCl^+	5.7	5.6	5.8	3.4	4.27	3.6	5.2^a	5.55	5.87
CH2^+ + HCl^+	6.9	6.7	6.8	—		4.6	6.7	6.01
CH3^+ + Cl^+	7.9	7.7	7.7	5.4	4.3	4.6	6.4/7.7^b	5.96/6.38	5.0
H^+ + CH2Cl^2+	—	16.5	18.5			8.3	15.4
H2^+ + CHCl^2+	—	14.1	—			7.7	15.4
CH3^2+ + Cl^+	—	—	—			4.0	10.2

---

The dissociation pathway followed by a molecular ion is sensitive to the initial ionization process. In ion-molecule interaction, the velocity of the incident projectile plays a key role in determining the dominant mode of electron removal from the target molecule.⁴⁴ At 0.45 MeV Ar⁹⁺ (v = 0.67 a.u.), the dominant mode of electron removal from CH₃Cl is expected to be electron capture; however, ionization is the dominant source of electron removal by the other two projectiles (2.25 MeV Ar⁹⁺ (v = 1.5 a.u.) and 0.25 MeV H⁺ (v = 3.16 a.u.)).

a. CH₃Cl²⁺ → Cl⁺ + CH₃⁺

Fig. 7 shows that the electronic ground state of CH₃Cl²⁺ connects directly to the lowest dissociation limit . Along this potential, a potential barrier of about 1.8 eV is calculated. The energy difference between the energy at the maximum and this dissociation limit is ∼6.4 eV, which corresponds to our computed KER for this channel. In a rather crude approximation of measurement of dissociation energies of the Cl⁺ and CH₃⁺ fragment ions, Monce at al.¹⁸ have suggested a KER value of 6.6 eV for this dissociation channel under 1 MeV H⁺, He⁺ and O⁺ ion impact. This is in good agreement with our data. The KER of 5.4 ± 1.0 eV obtained after single photon ionizing CH₃Cl by Ruhl et al.² is also in reasonable agreement with our theoretical result. The theoretical KER values of Duflot et al.²² agree relatively well with our results (see Table 3). Nevertheless, these KERs are lower by ∼1 eV than the present experimental result for the KER distribution for this channel (of ∼7.7eV; Table 3). It is higher by ∼1 eV with that measured by Siegmann et al.³ in 200 keV He⁺ collisions with CH₃Cl. The KER values of 4.3 and 4.6 eV measured under intense laser fields^5,6 are comparatively off from our experimental and theoretical determinations.

Dealing with the dissociation of the CH₃Cl²⁺ electronic excited states, it may happen directly along the PESs of the corresponding electronic states and/or after vibronic couplings (occurring at the crossings and the avoid crossings of the electronic states having the same spin multiplicity), Jahn–Teller interactions (for the doubly degenerate electronic states) and spin–orbit conversions (between the singlets and the triplets), in competition with the isomerization processes mentioned in previous section. For illustration, the electronically excited CH₃Cl²⁺ (¹E and ¹A₁) dications may decompose directly forming the products or may be converted by spin–orbit interaction into the CH₃Cl²⁺(³E) dications (for energies >2.4 eV). Fig. 7 shows that the population of the first excited state (i.e. the lowest ¹E) followed by spin–orbit conversion to the lowest ³E state leads also to the CH₃⁺ + Cl⁺ products in their electronic ground states. The crossing between these singlet and triplet states occurs at ∼3.1 eV above CH₃Cl²⁺(X³A₂). This results in a KER of 7.7 eV in close agreement with the present 7.7–7.9 eV measured values as can be seen in Table 3. Therefore, we may assume that the singlet states of CH₃Cl²⁺ are populated first in the present experiment. Singlet CH₃Cl²⁺ are then converted into CH₃Cl²⁺(³E) ions. Afterwards, the formation of the Cl⁺ + CH₃⁺ fragments in their electronic ground states follows the ³E state potential till dissociation.

The appearance energy (AE) of the Cl⁺ + CH₃⁺ ion pair through the fragmentation of CH₃Cl²⁺ was measured 32.5 ± 0.5 eV.⁴⁵ By combining the CH₃Cl ADIE (of ∼30.7 eV) and the potential barrier we compute for CH₃Cl²⁺(X³A₂) (of ∼1.8 eV, Fig. 7), we obtain an AE of 32.5 eV for the CH₃Cl²⁺ → Cl⁺ + CH₃⁺ process while the reaction is occurring along the ground state potential. This value is in excellent agreement with the experimental value obtained by Thissen using photoelectron photoion photoion coincidence (PEPIPICO) technique.⁴⁵ Whereas we deduce a much larger AE (of 33.8 eV) when the reaction starts on the lowest singlet (i.e. the lowest ¹E state). The latter AE is also distinctly larger than the measured value. Again, this strongly suggests that the reaction takes place along the ground CH₃Cl²⁺ potential prior to fragmentations in Thissen's experiments.⁴⁵

b. CH₃Cl²⁺ → H⁺ + CH₂Cl⁺

The CH₃Cl²⁺ dication may lose a proton and produces the CH₂Cl⁺ and H⁺ ions. Fig. 7 shows that the ground state potential possesses a barrier of ∼2.3 eV, which results in a KER of 6.4 eV. The present computed KER is close to the presently measured ones of 6.3–6.2 eV. Thus, this assumes that the reaction occurs on the ground state potential, which is of triplet spin-multiplicity. Nevertheless, only the lowest singlet dicationic potential correlates adiabatically to the CH₂Cl⁺(X¹A₁) + H⁺(¹S) channel. Some spin–orbit conversion should occur at large separations between CH₂Cl⁺ and H⁺(¹S) converting triplet CH₃Cl²⁺ into singlet CH₃Cl²⁺. The KER measured by Ruhl et al.² is in reasonable agreement with our experimental value, within the experimental uncertainties of both sets of data. However, the KERs for this channel obtained under intense laser field by Ma et al.⁶ are considerably lower than our experimental ones (cf.Table 3). In addition, a second reaction pathway may be proposed where we assist to a cleavage of the ClH bond after CH₃Cl²⁺ → H₂CClH²⁺ intramolecular isomerization processes. Again all couplings mentioned above are expected to play a role. Nevertheless, we have no indication from experiments for its occurrence presently.

We deduce an AE of ∼33.0 eV for the CH₂Cl⁺ and H⁺ ion pair production assuming that the reaction takes place solely along the lowest triplet PES. This value is reduced to 31.0 eV if the reaction follows the lowest singlet potential first and then the lowest triplet leading to CH₂Cl⁺ and H⁺ in their electronic ground states. The latter computed AE is in excellent agreement with that measured by Thissen⁴⁵ of 31.0 eV ± 0.5 eV using PEPIPICO method. This strongly suggests thus the primary population of the singlet state of CH₃Cl²⁺ in these experiments.

c. CH₂⁺ + HCl⁺

The formation of the CH₂⁺ + HCl⁺ ion pair is observed. Obviously, we cannot propose a direct mechanism for the production of these species from the CH₃Cl²⁺ dication. Instead, this dication may isomerize first to the trans planar H₂CClH²⁺ form. Then the reaction proceeds by cleavage of the CCl bond of H₂CClH²⁺. Fig. 8 presents the one-dimensional cuts of the potentials of H₂CClH²⁺ along the CCl distance. This figure shows that the ground state of this species possesses a potential barrier towards dissociation to form CH₂⁺ + HCl⁺ of ∼2.9 eV. Therefore, these ions can be produced carrying ∼6.7 eV KER in excellent agreement with the present experimental measurements (of 6.7–6.9 eV). The associated AE is ∼32.6 eV i.e. close to the 32.0 ± 1.0 eV AE determined by Thissen.⁴⁵ This suggests that this reaction occurs on the lowest singlet potential upon double ionizing the CH₃Cl molecule.

Although previous works dealing with this channel suggested a multi-step mechanism for this channel, it is the first time that both calculated and experimental KERs coincide. Indeed, the calculations of Duflot et al.²³ imply that there are two possible pathways for the isomerization followed by fragmentation to produce HCl⁺. Their calculated KER value is however smaller than our experimental and theoretical determinations. Also, the measured KER for this channel by Ma et al.⁶ is lower than ours. In a photoionization study of CH₃Cl after fast resonant Auger decay in Cl by Kokkonen et al.,¹⁰ the formation of a stable HCl⁺ was also reported. However, their qualitative simulation could not find convergence to this isomerization channel.

d. H₂⁺ + CHCl⁺

The formation of H₂⁺ + CHCl⁺ is due to a bond arrangement upon double ionizing CH₃Cl. This ion pair can be produced either by dissociating CH₃Cl²⁺ or after isomerization from H₂CClH²⁺. In the former case, both H atoms approach each other to form the H₂⁺ ion, whereas in the latter we have a H atom migration in addition to this H₂⁺ formation. For this channel, we measure KERs of ∼5.2 to 5.6 eV in good agreement with Duflot et al.²³ computed KER. This validates thus the mechanism they proposed. The present experimental KER values are however higher than those reported under double photoionization² and intense laser field ionization.⁶

e. H₃⁺ + CCl⁺

Recently, a detailed theoretical study for the production of this channel was done by some of us.²⁴ Upon dissociating the CH₃Cl dication, a multistep mechanism is in action for the formation of H₃⁺ where the three H-atoms in CH₃Cl²⁺ approach each other simultaneously while CCl⁺ goes away. A KER value of 5.2 eV has been calculated in this multi-step process as the energy difference between the X³A₂/1¹A₁ crossing and the lowest dissociation limit of the dissociated fragments. This theoretical KER value is in good agreement with our experimental data. This confirms the occurrence of such multistep mechanism for the formation of H₃⁺ from CH₃Cl²⁺ in the present experiments. This also rules out the possibility of roaming of H₂ upon ionization in this case as proposed in Ekanayake et al.²⁵ for other organic compounds. Besides, the KER value measured by Walt et al.⁵ under intense laser field is lower than our theoretical and experimental values.

f. Trication channels

In the dissociation of CH₃Cl³⁺, we have experimentally observed H⁺ + CH₂Cl²⁺ and H₂⁺ + CHCl²⁺ channels. To the best of our knowledge, the KER distributions shown in Fig. 4 and 5 and those reported in Table 3 are the only available data on the dissociation of CH₃Cl³⁺. The experimental KERs for these two channels are in reasonable agreement, under the experimental uncertainties, with our theoretical calculations. Nevertheless, the KER values reported in the literature by Ma et al.⁶ are considerably lower than our experimental and theoretical values.

The CH₃Cl³⁺ ions once formed from neutral CH₃Cl by ion impact do not decompose by C–Cl bond breaking. Indeed, the CH₃²⁺ + Cl⁺ channel is not observed experimentally in spite of the fact that the CCl bond cleavage is the major channel for the dication decomposition. Our calculations also suggest that this channel is less favorable thermodynamically. Instead, we observe in majority the H⁺ + CH₂Cl²⁺ and the H₂⁺ + CHCl²⁺ products. This may suggest some state selectivity upon fragmenting CH₃Cl³⁺ ions as recently evidenced by some of us for SO₂²⁺ unimolecular decomposition forming the less favorable thermodynamically SO²⁺ + O⁴⁶ and O₂⁺ + S^+ 47 ion pairs in addition to the obvious SO⁺ + O⁺ ion pair.⁴⁶ Such experiments on polyatomic triply charged ions are still challenging in particular upon valence triple ionization because of the low abundance of triple ionization compared to single and double ionization processes.⁴⁸

In sum, the multiple ionization of CH₃Cl and the subsequent fragmentation of the multiply charged CH₃Cl^q+ (q = 2,3) ions was investigated by electron impact, ions impact, single photon valence and inner shell ionization and intense laser multiple photon ionization. All these processes are different and may lead to various population of the multicharged ions electronic sates and thus to different energetics. For instance, the experimental KERs reported by Ma et al.⁶ shown in Table 3 obtained after intense laser fields interaction with CH₃Cl are considerably smaller than our experimental values and those determined using single photon ionization.² This may be due to the effect of the intense laser field which causes the sequential ionization of the electrons during the dissociation and the fragments are produced from Coulomb explosion at a bond length larger than the equilibrium bond length of the neutral molecule. Generally, one needs to be aware of such considerations and the comparison of the theoretical data and the experimental ones should be done with care.

VI. Conclusion

We used a combined theoretical and experimental methodology to investigate the evolution of the CH₃Cl²⁺ and CH₃Cl³⁺ ions formed by ion collision of CH₃Cl with energetic H⁺ and Ar⁹⁺ ion beams. Several dissociation channels were characterized. Globally, we have good agreement between our computed KERs and AEs with those either measured presently or in previous experimental works. Therefore, our theoretical data allowed to elucidate the mechanisms of the formation of the ion pairs, where fragmentations occur on the ground potentials and/or from the electronic excited states of these multiply charged ions. Nevertheless, we observed some discrepancies between the experimental data to compare with. This is mainly related to the ionization processes in action to form the CH₃Cl²⁺ and CH₃Cl³⁺ ions. As expected, valence double or triple single photon ionization should lead to the population of a bench of CH₃Cl²⁺ and CH₃Cl³⁺ electronic states that may be different from those reached by intense lasers or inner shell ionization followed by Auger processes or by fast ion beams impact. Strictly speaking, the full understanding of the dynamics of CH₃Cl upon double or triple ionization requires further experiments to separate the relative yields in terms of ionization and electron capture processes. Such experiments can be done using threefold and fourfold electron–ion coincidence measurements, which fully connect the electrons to the ion pairs coming from the same events. This should give a full picturing of the state-to-state decomposition of the CH₃Cl²⁺ and CH₃Cl³⁺ ions. For other molecular systems, such experiments combined with advanced theoretical treatments allowed elucidating the unimolecular state-to-state reaction pathways undertaken by the doubly changed molecular ions formed by valence double photoionization of the corresponding neutral.^46,47,49,50

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The authors thank the Programme National “Physique et Chimie du Milieu Interstellaire” (PCMI) of Centre National de la Recherche Scientifique (CNRS)/Institut National des Sciences de l’Univers (INSU) with Institut de Chimie (INC)/Institut de Physique (INP) co-funded by Commissariat à l’Energie Atomique (CEA) and Centre National d’Etudes Spatiales (CNES). The authors also thank N. Das, A. Kumar and A. Duley for help during the experiments. The authors thank the staff of LEIBF at Inter University Accelerator Centre, New Delhi, India for smooth operation of the ion beam during the experiments. K. R. M. thanks the SRM Institute of Science and Technology (SRM IST) for the research fellowship. M. P. thanks the Department of Science and Technology (DST) – SERB India (Grant: ECR/2017/000891) for providing financial support. M. P. thanks also SRM-IST for providing the supercomputing facility.

References

1. J. H. D. Eland, Int. J. Mass Spectrom. Ion Phys., 1969, 2, 471.
2. E. Ruhl, S. D. Price, S. Leach and J. H. D. Eland, Int. J. Mass Spectrom. Ion Process., 1990, 97, 175.
3. B. Siegmann, U. Werner and R. Mann, Nucl. Instrum. Methods Phys. Res. B, 2005, 233, 182.
4. E. C. Fayolle, K. I. Öberg and J. K. Jørgensen, et al. , Nat. Astron., 2017, 1, 703.
5. S. G. Walt, N. Bhargava Ram, A. Von Conta, O. I. Tolstikhin, L. B. Madsen, F. Jensen and H. J. Worner, J. Phys. Chem. A, 2015, 119, 11772.
6. P. Ma, C. Wang, X. Li, X. Yu, X. Tian, W. Hu, J. Yu, S. Luo and D. Ding, J. Chem. Phys., 2017, 146, 244305.
7. P. D. Grugan, N. Ekanayake, S. White, S. Y. Luo, P. Ruan and B. C. Walker, J. Phys. B: At., Mol. Opt. Phys., 2018, 51, 165601.
8. C. Miron, P. Morin, D. Céolin, L. Journel and M. Simon, J. Chem. Phys., 2008, 128, 154314.
9. Handbook of High-resolution Spectroscopy, ed. M. Quack and F. Merkt, John Wiley & Sons, Ltd, 2011, ISBN: 978-0-470-74959-3.
10. E. Kokkonen, M. Vapa, K. Bucar, K. Jänkälä, W. Cao, M. Zitnik and M. Huttula, Phys. Rev. A, 2016, 94, 033409.
11. E. Kokkonen, K. Jänkälä, M. Patanen, W. Cao, M. Hrast, K. Bučar, M. Žitnik and M. Huttula, J. Chem. Phys., 2018, 148, 174301.
12. D. M. P. Holland, I. Powis, G. Öhrwall, L. Karlsson and W. von Niessen, Chem. Phys., 2006, 326, 535.
13. A. Hult Roos, J. H. D. Eland, D. Koulentianos, R. J. Squibb, L. Karlsson and R. Feifel, Chem. Phys., 2017, 491, 42.
14. J. Harvey, R. P. Tuckett and A. Bodi, J. Phys. Chem. A, 2012, 116, 9696.
15. D. Céolin, M. N. Piancastelli, R. Guillemin, W. C. Stolte, S.-W. Yu, O. Hemmers and D. W. Lindle, J. Chem. Phys., 2007, 126, 084309.
16. R. Thissen, M. Simon and M.-J. Hubin-Franskin, J. Chem. Phys., 1994, 101, 7548.
17. A. F. Lago, A. C. F. Santos and G. G. B. de Souza, J. Chem. Phys., 2004, 120, 9547.
18. M. N. Monce, A. K. Edwards, R. M. Wood, M. F. Steuer, A. V. Shah and P. Tabor, J. Chem. Phys., 1981, 74, 2860.
19. C. G. Aitken, D. A. Blunt and P. W. Harland, Int. J. Mass Spectrom. Ion Process., 1995, 149, 279.
20. R. Rejoub, B. G. Lindsay and R. F. Stebbings, J. Chem. Phys., 2002, 117, 6450.
21. P. Grant, F. M. Harris and D. E. Parry, Int. J. Mass Spectrom., 1999, 192, 111.
22. D. Duflot, J.-M. Robbe and J.-P. Flament, J. Chem. Phys., 1995, 103, 10571.
23. D. Duflot, J.-M. Robbe and J.-P. Flament, Int. J. Mass Spectrom. Ion Process., 1997, 171, 215.
24. J. Palaudoux and M. Hochlaf, ACS Earth Space Chem., 2019, 3, 980.
25. N. Ekanayake, T. Severt, M. Nairat, N. P. Weingartz, B. M. Farris, B. Kaderiya, P. Feizollah, B. Jochim, F. Ziaee, K. Borne, R. P. Kanaka, K. D. Carnes, D. Rolles, A. Rudenko, B. G. Levine, J. E. Jackson, I. Ben-Itzhak and M. Dantus, Nat. Commun., 2018, 9, 5186.
26. J. Roithova, J. Hrusak and Z. Herman, J. Phys. Chem. A, 2003, 107, 7355.
27. A. Kumar, J. Rajput, T. Sairam, M. Jana, L. Nair and C. Safvan, Int. J. Mass Spectrom., 2014, 374, 44.
28. C. Hampel, K. A. Peterson and H. J. Werner, Chem. Phys. Lett., 1992, 190, 1.
29. P. J. Knowles, C. Hampel and H.-J. Werner, J. Chem. Phys., 1993, 99, 5219.
30. P. J. Knowles, C. Hampel and H.-J. Werner, J. Chem. Phys., 2000, 112, 3106.
31. M. J. O. Deegan and P. J. Knowles, Chem. Phys. Lett., 1994, 227, 321.
32. D. E. Woon and T. H. Dunning, J. Chem. Phys., 1993, 98, 1358.
33. T. H. Dunning, J. Chem. Phys., 1989, 90, 1007.
34. R. A. Kendall, J. Chem. Phys., 1998, 96, 6796.
35. P. J. Knowles and H.-J. Werner, Chem. Phys. Lett., 1985, 115, 259.
36. H.-J. Werner and P. J. Knowles, J. Chem. Phys., 1998, 82, 5053.
37. H.-J. Werner and P. J. Knowles, J. Chem. Phys., 1998, 89, 5803.
38. P. J. Knowles and H.-J. Werner, Chem. Phys. Lett., 1988, 145, 514.
39. K. R. Shamasundar, G. Knizia and H.-J. Werner, J. Chem. Phys., 2011, 135, 054101.
40. MOLPRO (version 2015) is a ab initio package written by H. J. Werner and P. J. Knowles. Further details at https://www.molpro.net.
41. R. Llusar, M. Casarrubios, Z. Barandiarán and L. Seijo, J. Chem. Phys., 1996, 105, 5321.
42. T. Zeng, D. G. Fedorov, M. W. Schmidt and M. Klobukowski, J. Chem. Phys., 2011, 134, 214107.
43. A. Berning, M. Schweizer, H.-J. Werner, P. J. Knowles and P. Palmieri, Mol. Phys., 2000, 98, 1823.
44. C. Illescas and A. Riera, Phys. Rev. A, 1999, 60, 4546.
45. R. Thissen, unpublished results as given in ref. 23.
46. M. Jarraya, M. Wallner, G. Nyman, S. Ben Yaghlane, M. Hochlaf, J. H. D. Eland and R. Feifel, Sci. Rep., 2021, 11, 17137.
47. M. Wallner, M. Jarraya, E. Olsson, V. Ideböhn, R. J. Squibb, S. Ben Yaghlane, G. Nyman, J. H. D. Eland, R. Feifel and M. Hochlaf, Sci. Adv., 2022, 8, eabq5411.
48. A. Hult Roos, J. H. D. Eland, J. Andersson, R. J. Squibb, D. Koulentianos, O. Talaee and R. Feifel, Sci. Rep., 2018, 8, 16405.
49. R. Linguerri, E. Olsson, G. Nyman, M. Hochlaf, J. H. D. Eland and R. Feifel, Inorg. Chem., 2021, 60, 17966.
50. V. Ideböhn, A. J. Sterling, M. Wallner, E. Olsson, R. J. Squibb, U. Miniotaite, E. Forsmalm, M. Forsmalm, S. Stranges, J. M. Dyke, F. Duarte, J. H. D. Eland and R. Feifel, Phys. Chem. Chem. Phys., 2022, 24, 786.

---