Molecular photodissociation dynamics revealed by Coulomb explosion imaging †

Coulomb explosion imaging (CEI) methods are finding ever-growing use as a means of exploring and distinguishing the static stereo-configurations of small quantum systems (molecules, clusters, etc ). CEI experiments initiated by ultrafast (femtosecond-duration) laser pulses also allow opportunities to track the time-evolution of molecular structures, and thereby advance understanding of molecular fragmentation processes. This Perspective illustrates two emerging families of dynamical studies. ‘One-colour’ studies (employing strong field ionisation driven by intense near infrared or single X-ray or extreme ultraviolet laser pulses) aﬀord routes to preparing multiply charged molecular cations and exploring how their fragmentation progresses from valence-dominated to Coulomb-dominated dynamics with increasing charge and how this evolution varies with molecular size and composition. ‘Two-colour’ studies use one ultrashort laser pulse to create electronically excited neutral molecules (or monocations), whose structural evolution is then probed as a function of pump– probe delay using an ultrafast ionisation pulse along with time and position-sensitive detection methods. This latter type of experiment has the potential to return new insights into not just molecular fragmentation processes but also charge transfer processes between moieties separating with much better defined stereochemical control than in contemporary ion-atom and ion-molecule charge transfer studies.


I. Introduction
Photodissociation is a particular form of chemical reaction in which the absorption of a photon (or multiple photons) of light by a molecule leads to its fragmentation.Vital examples include the photodissociation of oxygen molecules (O 2 ) to oxygen (O) atoms, driven by the absorption of short wavelength photons at high altitudes in the Earth's atmosphere, and of ozone (O 3 ) to O 2 plus an O atom by longer wavelength ultraviolet (UV) photons lower in the stratosphere.Together, these two photodissociation processes are key to establishing the localisation (in altitude) of the ozone layer. 1,2hose interested in molecular photodissociation dynamics seek insights and understanding beyond simply determining the photoproducts.Such studies are often conducted as 'pump-probe' experiments, in the gas phase, at low pressures, mostly under jet-cooled molecular beam (i.e.collision-free) conditions, and seek to answer questions like: in what quantum (electronic, vibrational, rotational, spin-orbit, etc.) states are the products formed, and/or what are the various product recoil velocities, v? The photoexcitation event usually involves an electric dipole transition, the probability of which is maximised when the electric (e) vector of the exciting radiation is aligned parallel to the transition dipole moment (l).Thus, photoexcitation typically creates an anisotropic distribution of excited molecules.If the photoexcited molecules dissociate rapidly (i.e. on a timescale that is fast compared with the period of molecular rotation), this spatial anisotropy should be revealed as an angular anisotropy in the product yield, i.e. in a correlation between l and v, and thus between e and v.Other correlations, e.g. between l and the rotational angular momentum, J, of a photofragment, and even between the fragment v and J vectors, can be observed in favourable cases. 3All such information reports on the forces acting on the nuclei of the molecule in the act of dissociating.These forces are determined by the topography of the potential energy surface(s) (PES(s)) sampled during the dissociation process.Calculations of excited (and ground) state PESs for small and medium-sized neutral molecules, and of the non-adiabatic couplings between these PESs, are now amenable to contemporary ab initio electronic structure methods.][6][7] Until quite recently, experimental studies of molecular photofragmentation processes in the gas phase have tended to group into one of two classes. 8Both start with the 'pump' stepphotoexcitation, typically with a short duration pulse of reasonably monochromatic light from a laser.One class relies on 'probing' the (typically electronic) spectrum of one or more of the photofragments -which in this class of experiments are most usually diatomic species.The density of photofragments in such experiments is generally too low to allow use of absorption detection methods; favoured probe methods include laser induced fluorescence (LIF) and resonance enhanced multiphoton ionisation (REMPI) spectroscopy.Given a thorough understanding of the diatomic fragment spectroscopy, analysis of the spectral line intensities allows determination of the product quantum state population distributions, 9 while measurements of Doppler broadened spectral lineshapes (linewidths) can give information about the fragment recoil velocities. 10he second class of experiments report directly on the velocities (and thus the translational energies) of one or more of the photofragments.Early photofragment translational spectroscopy (PTS) experiments used electron impact ionisation and mass spectrometric detection methods after the photofragments had recoiled along the detection axis.Current methods of choice 'tag' the fragment of interest immediately after its creation.2][13][14][15] The image reports the velocity distribution of the probed fragment and, in favourable cases, shows structure which upon analysis reveals the quantum state population distribution in the partner fragment.A variant, that has been widely applied in studies of the photodissociation of hydride molecules that yield H Michael N. R. Ashfold This journal is © the Owner Societies 2023 (or D) atom fragments, involves two photon resonant excitation to Rydberg states of the H (D) atom lying at energies just below the ionisation limit. 16These Rydberg states are sufficiently long lived to allow measurement of their times-of-arrival at a distant detector.Such 'Rydberg tagging' experiments also return velocity information, i.e. the fragment recoil speeds and directions.Combining such data with the requirements of momentum and total energy conservation in the overall photodissociation process can yield parent bond dissociation energies and detailed insights into the partitioning of any excess energy (i.e.energy provided by the pump photon in excess of that required to break the bond) within the products and thus the photodissociation dynamics. 17,18Again, however, the data returned by the probe process generally requires quite detailed knowledge and understanding of the spectroscopy of the partner fragment.
The introduction of femtosecond lasers as a mainstream laboratory tool saw the growth of an alternative family of pump-probe experiments offering much higher time resolution but, in many cases, lower spectral resolution.The focus now turned to exploring molecular photodissociation dynamics by probing the evolution from excited parent molecule to dissociation products in real time, rather than interrogating the asymptotic fragments in the greatest possible detail. 19The first such experiments typically used gas phase samples and LIF or REMPI (i.e.spectroscopy based) probe methods, but interest was growing in the use of other probe techniques like timeresolved photoelectron spectroscopy [20][21][22][23][24] and techniques with the potential to yield structural information directly.Other ultrafast pump-probe studies have sought to explore the dynamical aspects of molecular photofragmentation processes in solution.Transient absorption probe methods offer a means of tracking the formation (and subsequent relaxation) of not just 'traditional' photofragments formed by bond fission, [25][26][27] but also of acyclic isomers formed by photoinduced bond fission in a cyclic parent molecule. 281][32][33] Coulomb explosion imaging (CEI) is another emerging technique, that is increasingly finding use both as a route to exploring the fragmentation of multiply charged cations and as a means of probing molecules in the act of dissociating.Removing several electrons from an isolated molecule yields a multiply charged cation that dissociates by Coulomb explosion (CE).The term CE was introduced to describe and rationalise the fragment ions and their kinetic energy distributions following X-ray induced core ionisation of molecules like HI and CH 3 I. 34 The earliest illustrations of CEI Fig. 1 (a) Schematic of a velocity-map imaging (VMI) experiment.A molecular beam of the target molecule (usually seeded in excess inert carrier gas) is crossed with a pump laser pulse, which initiates photodissociation.One (or more) chosen fragments are then ionised, using an appropriate probe laser pulse, and the charged atomic and/or molecular fragments are accelerated towards a position-sensitive detector by an electric field, yielding an (x, y, t) data point for each fragment ion.The ion arrival times correlate with their mass-to-charge (m/z) ratio, and the arrival positions correlate with their velocities.VMI studies of near infrared (NIR) strong field ionisation (SFI) or extreme UV (XUV) induced Coulomb explosions such as those described in Section II employ a similar set-up but with just one ultrafast laser pulse to induce CE.VMI studies of molecular photofragmentation processes using CE as the probe, such as those described in Section III, use ultrafast NIR SFI or XUV photons as the probe pulse.involved small molecular ions like CH 4 + , which were accelerated to high velocities and arranged to impact on, and pass through, a very thin (B3 nm thick) metal foil. 35,36This interaction led to the stripping of further valence electrons, and CE.Determining the speeds and the directions of the resulting fragment ions allowed reconstruction of the parent cation structure at the instant of CE.Short (femtosecond) duration laser pulses are now increasingly being used to initiate the CE of isolated molecules, [37][38][39][40] providing new insights into the structures of molecules and small clusters, [41][42][43][44] enabling the distinction of isomers 45,46 and even enantiomers 47,48 and the preparation of exotic cations. 49uch experiments have, for example, recently allowed real-time studies of the Jahn-Teller driven distortions of the CH 4 + cation. 50Fig. 2 provides an impressive recent illustration, where CE methods have been used to reveal the momenta (and thus, after due processing, the positions) of all atoms within the 2-iodopyridine molecule.Note the constrained C + ion that was originally bonded to the I atom in the neutral parent molecule (C2 in Fig. 2).This Perspective has another focus: how CEI methods are offering new ways of exploring the fragmentation dynamics of multiply charged cations and, when used as the probe in pump-probe experiments, of neutral molecules and singly charged molecular cations.
Current laser-induced CEI studies partition into two broad families, distinguished by the way in which the CE is initiated.2][53][54] The other employs one (or more) high-energy extreme UV (XUV) or soft X-ray photons, typically from a free-electron laser (XFEL). 55,56In SFI, the neutral species of interest is elevated to (normally a range of) multiply charged states of the parent cation by the removal of several valence electrons.In soft X-ray excitation, in contrast, the wavelength is typically tuned to maximise the probability of removing one (or more) innershell electrons from a target nucleus (i.e. the initial excitation is much more site-specific) and the multiply charged states of the cation are formed by subsequent (rapid) Auger-Meitner (AM) 57 cascade.In both cases, the multiply charged cations then 'explode' by virtue of the Coulombic repulsion between the constituent nuclei.Determining the velocities and momenta of the charged products using contemporary time-and positionsensitive detectors allows a direct mapping of the positions of the charged nuclei in coordinate space, i.e. a non-spectroscopic view of the fragmentation dynamics.
Laser-induced CEI studies further sub-divide according to the specifics of the time and position-sensitive detection method.Much of the data reported in this article was obtained using multi-mass imaging methods, high count rates (410 ions per laser shot) and covariance mapping techniques 58 to explore and establish possible correlations between different fragment ion pairs.An alternative approach, traditionally involving lower ion count rates and longer data accumulation times, relies on measuring coincidences between all fragment ionsas, for example, in the cold-target recoil-ion momentum spectroscopy (COLTRIMS) experiment used to generate the data shown in Fig. 2.
The illustrative examples that follow draw heavily on published and hitherto unpublished work performed in Bristol using NIR-SFI to induce CE but, when possible, the findings are compared and contrasted with those obtained when initiating CE by X-ray excitation (i.e.core ionisation).Note that some of the hitherto unpublished Bristol data are presented as ESI † to maintain a 'bigger-picture' narrative appropriate for a Perspective article.

II. Fragmentation of multiply charged molecular cations.
CF 3 I and alkyl iodides Fig. 3 shows time-of-flight (TOF) mass spectra of the fragment ions, after conversion to spectra of ion mass/charge (m/z) ratio, formed following 805 nm SFI of a jet-cooled CF 3 I sample recorded at three different incident intensities: I = 260, 650 and 1300 TW cm À2 .We note that an ultrafast laser pulse presents a spatially and temporally varying spread of intensities and that there are many ways to report a laser intensity; the values reported here are peak intensities, calculated using a priori knowledge of the input beam conditions and profile. 59he spectrum recorded at lowest I is dominated by peaks with m/z 127 (I + ) and 69 (CF and 31 (CF + ).These same peaks appear in the spectra recorded at higher I, along with peaks at m/z 42 (I 3+ ), 34.5 (CF 3 2+ ), 25 (CF 2 2+ ), 19 (F + ), 15.5 (CF 2+ ), 12 (C + ), 9.5 (F 2+ ) and 6 (C 2+ ).
The companion images show an obvious I 4+ component within the m/z 31 signal measured at high I. 60 Knowing both the time and the position of arrival of each ion at the detector allows determination of the I-dependent velocity (P(v)) (and momentum P(p)) distributions of each fragment ion and identification of correlations between the various product ions, i.e. products arising via a common fragmentation process.These data, the results of analogous studies of the fragment ions from NIR SFI of CH 3 I 59 and complementary ab initio trajectory simulations 61 all point to a continuum of behaviours for CF 3 I Z+ parent ions, ranging from simple, 'diatomic-like' C-I bond fission at low Z, through a complex pattern of fragmentation behaviours involving loss of one or more F atoms (or ions) at intermediate Z, before approaching limiting Coulomb-driven dissociation at high Z (4 N, where N is the number of atoms in the molecule), i.e.Z 4 5 in these cases.
Here we focus on the low I data and the insights they provide into the fragmentation of CF 3 I 2+ cations.Fig. 4 All images displayed in this work are plotted using a linear false-colour scale that spans the maximum to minimum intensity recorded in the pixels within each image.Image analysis yields the respective velocity distributions, which are plotted as P(p) distributions in Fig. 4(e).The low momentum features in the I + and CF 3 + fragment distributions are attributable to rival dissociative ionisation (DI) pathways from singly charged CF 3 I + parent ions and are not considered further. 60The higher momentum features in these P(p) distributions match well, consistent with 'diatomic-like' fragmentation of the CF 3 I 2+ dication, i.e.
This channel had been identified previously in coincidence studies following XUV photoexcitation of CF 3 I at discrete energies in the range 40-50 eV. 62Covariance map images derived from the present data provide another way of identifying correlations between given pairs of fragments.Fig. 4(f) and (g) show, respectively, the (CF 3 + , I + ) and (CF 2 + , I + ) covariance images.In these depictions, the I + ion has been selected as the reference ion (and its velocity fixed to be vertically upwards -defined as f = 01 and illustrated by the red arrow in these panels) and the correlated 2-D velocity distribution of the CF n + partner is then displayed in the frame defined by the reference ion.The correlations between the CF 3 + and I + products are tightly focussed along f = 1801, as required by momentum conservation in the case of a two-body dissociation.The correlation between the CF 2 + and I + products is also centred along f = 1801, but the covariance map image is more diffuse.This finding, and the evident similarities For ease of comparison, the spectra are displayed so that the m/z 69 (CF 3 + ) peak appears with the same peak intensity in each.between the various CF n + velocity distributions, suggest that these CF 2 + fragments arise by neutral F atom loss from internally excited primary CF 3 + photoproducts.Covariance map image analysis also identifies a (more diffuse) correlation between I + and CF + products (consistent with loss of two F atoms or of F 2 from primary CF 3 + products) and other charge symmetric fragmentation channels to CF 2 I + + F + and to CF 2 + + IF + product pairs, while the measured velocity distribution of the I 2+ fragments also implies operation of the charge asymmetric CF 3 + I 2+ dissociation pathway. 60s noted above, some but not all the above fragmentation channels for CF 3 I 2+ cations had been identified in a previous study using photoelectron-photoion-photoion and TOF photoelectron-photoelectron-photoion-photoion coincidence spectroscopy methods. 62Compared with CEI, such coincidence experiments, particularly when performed using tuneable XUV photons from a synchrotron, offer better control and definition of the excitation energy and the electronic state(s) in which the dication is prepared.But, as illustrated here, NIR SFI CEI measurements offer a substantial multiplex advantage (the present data sets reveal not just the various fragmentation channels for Z = 2 cations described here, but also provide information about the fragmentation of CF 3 I Z+ (Z 4 2) states, for which there is no prior experimental data, and can be undertaken in just a few hours using a table-top laser experiment.
Alkyl iodides, especially CH 3 I, have featured in many imaging studies designed to explore CE dynamics.NIR SFI CEI studies of CH 3 I 59,63-65 again illustrate the transition from 'diatomic-like' dissociation to CH 3 + + I + products in the case of CH 3 I 2+ cations, where valence bonding forces compete with Coulomb repulsion, 63 through to Coulomb dominated dissociations yielding I q+ (q r 6) fragment ions at the highest intensities used in the Bristol experiments 59 and, as an extreme example, complete explosion to H + , C p+ (p r 4) and I q+ (q r 15) fragments (i.e. from CH 3 I Z+ cations with Z r 22) following absorption of multiple 5.5 keV X-ray photons. 66any of the studies involving CH 3 I where CE has been induced by one or more X-ray photons are ultrafast pumpprobe studies, primarily designed to explore the ionisation and charge transfer (CT) dynamics as a function of time following an initial pump pulse that triggers C-I bond extension and progressive separation of the (neutral) I and CH 3 fragments.The results of an early experiment of this type, employing 1.5 keV probe photons, were interpreted in terms of (i) nearexclusive core ionisation of the iodine atom, (ii) subsequent fast (sub-femtosecond) AM decay leading to multiple positive charges localised on the iodine site and (iii) subsequent redistribution of this charge over the entire molecule, leading to fast Coulomb-repulsion dominated fragmentation. 67At early pump-probe time delays (corresponding to small H 3 CÁ Á ÁI bond extensions), the time-dependent I q+ yields show clear evidence of CT from the I q+ to the CH 3 moiety (or, more accurately, electron transfer from the CH 3 to the multiply charged I q+ centre).We revisit such pump-probe data in Section III.
Here, we note that such a two-stage picture -AM decay leading to localised charging within part of the molecule and subsequent redistribution of this charge -is also often assumed in the case of XUV ionisation of molecules at their equilibrium geometry.Such processes probably merit further study.The valence electrons released after AM decay following core-ionisation at the equilibrium geometry emanate from molecular orbitals (MOs) that, in many cases, will be delocalised over the nuclear framework.Would the sequential picture outlined above for an alkyl iodide, in which the least tightly bound valence electrons in the ground state molecule are in 'atomic-like' p-orbitals localised on the I atom, extrapolate to, for example, XFEL induced ionisation of a non-polar molecule in which the least tightly bound electrons are in delocalised p orbitals?
Related questions apply when considering CE induced by NIR SFI, which lacks the initial site-specificity achieved by coreionising a particular atom.Ionisation in this case is typically pictured as starting with the least tightly bound electrons; increasing the incident intensity leads to the progressive ejection of more of the less tightly bound valence electrons.NIR SFI might thus be expected to encourage simultaneous charging of multiple nuclei within a molecule.Again, however, alkyl iodides may not be 'general' test systems since, as noted above, the least tightly bound electrons in the ground state molecule are in orbitals localised on the I atom.So, even in the case of NIR SFI, the initial stages of the CE process in molecules like alkyl iodides, where the highest occupied MOs are localised non-bonding orbitals, may best be pictured as preferential charging of the I site and subsequent CT.

1,2-Dichloroethene
This example is chosen to illustrate the potential isomer selectivity afforded by CEI.Previous spectroscopic studies of the cations of both the cis-(or Z-) and trans-(or E-) isomers of 1,2-dichloroethene (DCE) have mainly focussed on the respective Z = 1 ions.1][72] Threshold photoelectronphotoion coincidence studies 73,74 have established the onset of the DI process HClCQCHCl + hn -H 2 C 2 Cl + + Cl + e À (2) at E B 12.10 eV for both isomers and ab initio calculations suggest rapid H atom migration following photoexcitation, so that the fragment ion in both cases is the C 2v symmetry H 2 CQCCl + species. 74The relative yields of H 2 C 2 Cl + fragment ions fall steeply at excitation energies E 416 eV.C 2 H 2 + fragments dominate at these higher excitation energies and the shape of the energy-dependent breakdown curves suggests that these ions are formed by Cl loss from highly internally excited H 2 C 2 Cl + species.A (weaker) rival DI process yielding HC 2 Cl + ions, by HCl elimination, has been observed at E 4 12.7 eV and a further increase in the relative yield of HC 2 Cl + fragment ions once E 417 eV has been attributed to H atom loss from the primary H 2 C 2 Cl + species. 74,75There are no documented DI channels for the Z = 1 parent cations yielding Cl + fragments.
This journal is © the Owner Societies 2023 The H x C 2 Cl + fragment ion images from both isomers are similar (Fig. 7(b)) and show similar trends with increasing I.
Each shows an intense central feature and an anisotropic ring at larger radius.The DI process ( 2) is likely to be the major contributor to the intense low velocity feature in the H x C 2 Cl + ion images.The relative magnitude of the higher velocity feature increases with increasing I.The mean momenta of these faster H x C 2 Cl + fragments match well with those of the most probable Cl + ions in the images recorded at low I, and this correlation is confirmed by the (H x C 2 Cl + , Cl + ) covariance map images shown in Fig. 7(d) which, again, are tightly focussed along f = 1801.As also deduced in recent XFEL enabled CE studies of these isomers following core-ionisation with 240 eV photons, 45,79 these fragments are attributable to a rival twobody dissociation of the Z = 2 parent cation, i.e.
Similar analyses of the H x C 2 Cl 2+ fragment ion images recorded at higher I (Fig. S1 in the ESI †) imply the operation of an analogous C-Cl bond fission from the Z = 3 parent cations, i.e.
The P(v) distributions derived from the H x C 2 Cl 2+ ions are sensibly consistent with the faster component in the Cl + velocity distributions observed at higher I (Fig. 7).This section devoted to 1,2-DCE ends with data for the C 2 H x + fragment ions (Fig. 8), which are henceforth assumed to be C 2 H 2 + .The images of these products from both parent isomers show an intense central feature, and an isotropic signal stretching to larger radius which becomes relatively more important at higher I.The central feature is attributable to C 2 H 2 + products from secondary decay (Cl atom loss) of internally excited H 2 C 2 Cl + products arising via DI channel (2). 74,75The P(v) distributions derived from the C 2 H 2 + images are shown in Fig. 8(b).The distributions obtained at lowest I both peak at v B 1400 m s À1 .
Excitation at this (relatively) low intensity will strongly favour formation of parent ions with low Z and the major source of these C 2 H 2 + ions is, again, secondary decay of the internally excited primary H 2 C 2 Cl + fragments formed, in this case, by the CE process (5).
Both P(v) distributions also show a 'tail' stretching to higher v, that gains in relative importance at higher I.The form of this 'tail' is isomer-specific, stretching to considerably higher v in the case of the cis-isomer.The (C 2 H 2 + , Cl + ) covariance plots derived from the I = 260 TW cm À2 data for both isomers show a view supported by photoion-photoion-photoion coincidence (PIPIPICO) measurements of these same three fragments arising via CE of both 1,2-DCE isomers induced by either 240 eV core-ionisation 45 or ultrafast NIR SFI. 80Fig. 9 depicts the derived fragment ion correlations from a NIR SFI study in socalled Newton plots, in which the momenta of two fragment ions (here C 2 H 2 + and Cl + ) are plotted with respect to the momentum of the third fragment (Cl + ).The Newton plot for trans-1,2-DCE (Fig. 9(b)) shows the two Cl + fragments recoiling in essentially opposite directions, with the C 2 H 2 + fragment remaining almost at rest, as would be predicted in the case of a near-instantaneous explosion from the undistorted equilibrium geometry where the centre of mass of the C 2 H 2 moiety is symmetrically positioned between the two Cl atoms.The corresponding plot for cis-1,2-DCE (Fig. 9(a)) is obviously different: the angle between the two Cl + fragment momenta is B1081 and the C 2 H 2 + product has significant momentum.Both Newton plots imply that the CE giving rise to these fragments The m/z spectra following NIR SFI of the two isomers illustrate another isomer specific difference relating to the weak peak at m/z B 71, due to Cl 2 + ions, that is not apparent when plotted with the intensity scale used in Fig. 5.The associated images are dominated by a central (i.e.zero kinetic  This journal is © the Owner Societies 2023 energy) feature, attributable to ionisation of trace Cl 2 impurity in the respective samples.But the P(v) distribution of the Cl 2 + fragments from SFI of cis-1,2-DCE at I = 260 TW cm À2 also shows some translationally excited fragments (peaking at v B 1000 m s À1 ) which are not present in the corresponding distributions from the trans-isomer.This difference is highlighted in the inset in Fig. 5 3), ( 6) and ( 8)).CEI also reveals that Z = 3 parent ions decay by CQC bond fission and by two triple dissociation processes.One, a near-concerted process ( 7), releases two Cl + fragments with clearly isomer specific kinematics.The other is attributable to secondary decay of internally excited H 2 C 2 Cl 2+ fragments formed via process ( 6).This represents just one of a rapidly growing number of triple dissociations investigated using ultrafast laser induced CEI, often in conjunction with so-called native-frame analysis wherein successive steps in a sequential three-body fragmentation are analysed in their own 'native' frame of reference. 81Such methods have recently been used to demonstrate that, for example, the three-body dissociation of the OCS 3+ cation proceeds sequentially, via rival channels yielding CO 2+ and CS 2+ intermediates that decay further over a longer time scale. 81As in the case of 1,2-DCE, the triple dissociation of C 2 H 4 Br 2 3+ cations to C 2 H 4 + + 2Br + products is best viewed as involving both concerted and non-concerted contributions, 82 as are the dissociations of the trications of both 1-and 2-iodopropane to CH (with a CHBr 2 2+ intermediate in the non-concerted pathway). 84mmary Table 1 illustrates many of the polyatomic molecules for which one or more of their polycation fragmentations have now been investigated by CEI methods.A critic could argue that many of these studies reveal active fragmentation channels rather than the fragmentation dynamics per se.It is striking that the P(v) (and the kinetic energy) distributions of the fragments arising via many of the charge symmetric binary dissociations are relatively narrow -see channels (1) (Fig. 4) and ( 3) (Fig. 6) and ( 5) (Fig. 7) in this work -notwithstanding the apparent energetic imprecision of the NIR SFI or XUV core-ionisation process that led to their formation.But the appearance of, for example, CF ) fragments attributable to secondary decay of CF 3 + (H 2 C 2 Cl + ) products in the examples reviewed here also implies that some of these primary fragments are formed with sufficient internal energy to enable subsequent unimolecular decay.More coincidence studies that provide energetic information about not just the fragment ions but also the photoelectrons formed upon multiple ionisation would surely be revealing.The growing number of studies of the three-body dissociation of parent trications are providing more dynamical insights, and the application of covariance and native frames analysis methods are allowing discrimination between concerted (or at least near concerted) and sequential fragmentations.However, a common finding is that a given trication displays a mix of both behaviours and there is still a way to go before one could determine, for example, whether these different outcomes reflect rival fragmentation dynamics from a common electronic state or dissociation from two (or more) different states of the photo-prepared trications.[87][88]

III. Using CEI probe methods to track the photofragmentation of neutral molecules
Alkyl iodides CEI methods are increasingly also finding use in two-colour studies of the fragmentation dynamics of photoexcited neutral molecules, and singly charged cations.As shown below, the associated velocities confirm that these DISS features are signatures of the well-characterised 267 nm photolysis of CF 3 I, yielding neutral CF 3 and I fragments, one or other of which is subsequent ionised (by NIR SFI) to a CF 3 + or I + ion.
These are the only features in these images that report directly on the field-free fragmentation of CF 3 I (i.e. are unperturbed by the probe NIR SFI process), but the CF 3 + , I + and I 2+ images recorded in the range 0 r dt r 600 fs each show additional features (two in the case of the CF 3 + images) labelled a UV = ultraviolet pump wavelength, X = XUV or X-ray, NIR = near infrared.(dt B 1 ns) all match well, support previous conclusions that most (490%) of the iodine atoms formed following excitation of CF 3 I molecules at 267 nm are in the spin-orbit excited ( 2 P 1/2 ) rather than ground ( 2 P 3/2 ) state (henceforth identified as I* and I, respectively) [201][202][203] and imply similar NIR SFI probabilities for both spin-orbit states of the I atom.
Several further aspects of these data merit note.First, the DISS feature is relatively much stronger in the I + than the CF 3 + map, implying that the probability of NIR strong field induced one photon ionisation of I is much greater than that of CF 3 under the prevailing experimental conditions.A weak DISS feature is even observable in the I 2+ map (particularly that recorded at lower I, Fig. S5(c), ESI †).Second, as can be seen from comparing the corresponding maps in Fig. 11 and Fig  . 187s with CF 3 I, non-coincident maps derived from multi-mass imaging experiments can show additional features attributable to processes other than the field-free fragmentation of CH 3 I that is of primary interest in the current context.The CH 3 + and I + maps recorded at lower I (e.g.Fig. S9 and, particularly, S10 in (ii) XUV/X-ray probing.Two-colour probe studies using FEL-based X-ray photons to induce ionisation have also been reported for CH 3 I 67,188,208 and for several other singly and doubly halogenated alkyl halides including 1-iodo-2-methylbutane, 192 2-iodopropane, 191 CH 2 ClI 185 and CH 2 BrI. 189,190These generally employ a probe photon energy above one of the inner shell ionisation thresholds of the iodine atom, which offers much greater initial site specificity and selectivity compared with NIR SFI probe methods.Fig. 15 illustrates the most obvious consequence of this difference by comparing 2D(v, dt) maps of the atomic iodine fragments formed in the 267 nm photolysis of CH 3 I, monitored via the I 3+ ion prepared by (a) B 120 fs, 108 eV XUV pulses tuned above the I 4d edge and (b) 805 nm SFI with I = 1300 TW cm À2 .Points to note regarding the XUV probe study 188 include: the I atom is estimated to present an order of magnitude larger photoabsorption cross-section than the CH 3 radical at this probe wavelength (11.5 nm); the observation of I q+ ions with q r 4 implies that multiple ionisations (involving the absorption of at least two XUV photons and subsequent AM decay) occurred under the prevailing experimental conditions; and the CH 3 I molecules were pre-aligned using another strong laser field, but this detail is not germane to the present discussion.
Fig. 15(a) is dominated by a constant velocity feature that grows after dt 4 0 and is attributable to site-selective ionisation of primary I photofragments, with no subsequent charge transfer to the geminate CH 3 radical.This signal, the equivalent of the DISS channel in Fig. 11 and 13, reports directly on the neutral C-I bond fission process and yields fragment translational energy distributions consistent with those derived in the earlier ion imaging studies. 204As Fig. 15(b) and Fig. S6 (ESI †) show, the yield of I 3+ ions when probing by 805 nm SFI is small, and the probability of forming I 3+ ions without also ionising the Careful inspection of the DISS signals associated with different I q+ products formed by core-ionisation (X-ray probing), like that shown in Fig. 15(a), reveals that their onsets are delayed relative to dt = 0 and that this delay increases with q. 67,183,188 Such trends can be explained, qualitatively at least, using a classical over-the-barrier model. 67,209At very early pump-probe delays, the selectively ionised fragment is too close to its geminate partner to prevent inter-fragment charge (electron) transfer and subsequent mutual Coulomb repulsion and acceleration.The barrier to charge transfer increases with increasing inter-fragment separation (i.e. with increasing dt) and, at some critical distance, charge transfer to the partner fragment is no longer possible and the DISS signal grows in.This critical distance increases with q, explaining the progressively delayed onset of the DISS feature when probing via higher I q+ states.

IV. Conclusions and future prospects
As noted at the outset, understanding of the dynamics of molecular photodissociation processes has advanced massively in the past few decades.The photodissociation dynamics of many families of small neutral molecules, including many of key atmospheric and astrochemical interest, have now been characterised experimentally in considerable detail.Theory has shown similarly impressive advances.Given sufficient demand, multi-dimensional excited state PESs can now be calculated with high accuracy for most small/medium sized neutral molecules, as can the non-adiabatic couplings between and the nuclear dynamics on such PESs.Ultrafast laser methods have encouraged the transitioning of such dynamical interests into condensed phase environments, and to studies of photoinduced intramolecular bond fission (e.g.ring-opening) reactions.
CEI methods have now entered mainstream contemporary atomic and molecular physics research and are finding ever increasing use as a means of exploring the static stereoconfigurations of small quantum systems (molecules, clusters, etc), and of tracking the time-evolution of such structures.Based on the examples featured in this Perspective one might argue that CEI methods have 'yet to deliver' penetrating new insights into the photofragmentation dynamics of neutral molecules, but this situation can surely be expected to improve as the technique is applied to a wider range of (larger) molecules, along with greater use of covariance and coincidence analysis methods. 210The data shown in Fig. 15(a) highlight some of the potential benefits (e.g.site selectivity, image simplification) of using X-ray core-ionisation methods to probe molecules in the act of dissociating.As noted in the Introduction, the dynamics of the UV photoinduced ring- opening of thiophenone has been explored recently by both timeresolved photoelectron spectroscopy 22 and ultrafast electron diffraction 32 methods.Simulated Newton plots of the H + or C + fragment momenta, in a plane defined by the O + and S + atom momenta, suggest that CE methods should provide another route to distinguishing ring-opened and ring-closed photoproducts.More widely, CEI methods are already providing access to an extraordinarily wide range of multiply charged molecular cations, thereby offering exciting new opportunities to explore the progression from valence-dominated to Coulomb-dominated fragmentation dynamics with increasing charge, and ways in which this varies with molecular size and composition.To date, the body of ab initio electronic structure and molecular dynamics calculations for molecular cations, particularly multiply charged molecular cations, is much sparser than for neutral systems.Modelling the diverse range of possible fragmentation pathways and dynamics available to multiply charged cations -formed in the presence of a strong laser field -is likely to remain a major challenge for some time to come.
Experimentally, image analyses return particle velocities, and momentum matching arguments built on such measurements are pivotal to understanding many of the fragmentation pathways reported for polyatomic parent cations.Thus far, much less attention has been given to understanding the absolute fragment velocities.One can anticipate that multiply charged parent cations prepared by NIR SFI might be formed in a range of electronic states, and with significant vibrational excitation (due to Raman excitations).NIR pump-probe studies of CH 3 I using moderate NIR field strengths 182 lend support to such expectations and for a 'ladder climbing' pumping mechanism 211 wherein higher charge states are accessed by successive promotions via parent states of lower charge.The overall time scale could be sufficient to allow some nuclear distortion from the starting geometry en route to the final higher charge state.But other NIR SFI studies of CH 3 I, at significantly higher field strengths, have returned quite sharp images for all I q+ (q r 5 at least) fragments, implying that each are formed with relatively narrow velocity distributions. 59Companion (field-free) ab initio electronic structure plus trajectory calculations consistently predict substantially higher I q+ (q 42) fragment ion velocities than those determined experimentally.This discrepancy has been rationalised by suggesting a role for non-adiabatic coupling and intramolecular charge transfer in the CE of more highly charged CH 3 I Z+ molecular ions. 59ore studies, preferably coincidence studies that track not just the fragment ion velocities but also those of the ejected photoelectrons would surely help in understanding the fragmentation dynamics of multiply charged parent cations.More generally, as also noted above, ultrafast pump -CEI probe experiments can also be expected to provide a wealth of new insights into charge transfer processes between moieties separating under much greater stereochemical control than achieved hitherto in charge transfer studies involving ion-atom and ionmolecule collisions.
Michael N. R. Ashfold obtained his PhD from the University of Birmingham in 1978.After postdoctoral research at Oxford, he was appointed to a Lectureship at the University of Bristol, where he was promoted to a Chair in Physical Chemistry in 1992.He retired from the Leverhulme Chair of Chemistry in late 2020 and is now an Emeritus Professor at Bristol.He was elected to the Fellowship of the Royal Society in 2009 and is a past winner of the Corday-Morgan, Tilden and Liversidge Prizes of the Royal Society of Chemistry.His research interests include gas phase molecular photochemistry and spectroscopy.
Fig.1(a) Schematic of a velocity-map imaging (VMI) experiment.A molecular beam of the target molecule (usually seeded in excess inert carrier gas) is crossed with a pump laser pulse, which initiates photodissociation.One (or more) chosen fragments are then ionised, using an appropriate probe laser pulse, and the charged atomic and/or molecular fragments are accelerated towards a position-sensitive detector by an electric field, yielding an (x, y, t) data point for each fragment ion.The ion arrival times correlate with their mass-to-charge (m/z) ratio, and the arrival positions correlate with their velocities.VMI studies of near infrared (NIR) strong field ionisation (SFI) or extreme UV (XUV) induced Coulomb explosions such as those described in Section II employ a similar set-up but with just one ultrafast laser pulse to induce CE.VMI studies of molecular photofragmentation processes using CE as the probe, such as those described in Section III, use ultrafast NIR SFI or XUV photons as the probe pulse.(b) Illustrative velocity map images of CF 2 + , I 2+ , CF 3 + and I + fragment ions from NIR SFI of CF 3 I at I = 650 TW cm À2 .

3 + ), with weaker features at m/z 177 (
Fig.3shows time-of-flight (TOF) mass spectra of the fragment ions, after conversion to spectra of ion mass/charge (m/z) ratio, formed following 805 nm SFI of a jet-cooled CF 3 I sample recorded at three different incident intensities: I = 260, 650 and 1300 TW cm À2 .We note that an ultrafast laser pulse presents a spatially and temporally varying spread of intensities and that there are many ways to report a laser intensity; the values reported here are peak intensities, calculated using a priori knowledge of the input beam conditions and profile.59The spectrum recorded at lowest I is dominated by peaks with m/z 127 (I + ) and 69 (CF 3 + ), with weaker features at m/z 177 (CF 2 I + ), 146 (IF + ), 139 (CI + ), 88.5 (CF 2 I 2+ ), 63.5 (I 2+ ), 50 (CF 2 + )

Fig. 2
Fig. 2 Newton plots showing the absolute momenta of (a) N + and H + and (b) N + and C + fragment ions following CE of 2-iodopyridine (C 5 NH 4 I, structure shown in the inset).All data sets contributing to these plots involved detection of the I + , N + and at least one H + and one C + ions, and the coordinate frame has been rotated so that in all cases the I + momentum points in the +z direction.(Adapted from Boll et al., ref. 42, with permission from Springer Nature, copyright 2022.) (a)-(d) show symmetrised images of the I + and CF n + (n = 1-3) fragments, recorded with the laser polarisation vector, e, aligned vertically in the plane of the image, as shown in the top left-hand panel.

Fig. 3
Fig. 3 TOF spectra of the fragment ions observed following 805 nm SFI of CF 3 I at I = (a) 260, (b) 650 and (c) 1300 TW cm À2 , displayed with a mass/ charge (m/z) scale, with the regions appropriate for CF 2 I 2+ , CI + , IF + and CF 2 I + ions shown on expanded vertical scales in the insets within panel (a).For ease of comparison, the spectra are displayed so that the m/z 69 (CF 3 + )

Fig. 4
Fig. 4 Symmetrised images of the (a) I + , (b) CF 3 + , (c) CF 2 + and (d) CF + fragments formed by 805 nm SFI of CF 3 I at I = 260 TW cm À2 .The ticks on the white horizontal scale on each image indicate 1000 m s À1 intervals.The e vector of the SFI laser radiation is shown by the double headed red arrow in (a) and the relative signal intensities are shown using the linear false-colour scale that spans the maximum to minimum signal levels in the regions of interest within each image displayed at the end of the first row.The momentum distributions derived from these images are shown in (e).(f and g) Show, respectively, the (CF 3 + , I + ) and (CF 2 + , I + ) covariance map images, with the velocity of the I + reference ion chosen to be vertically upwards (shown by a red arrow) and the covariance signal of interest in each case bounded by dashed white lines.

Fig. 5
Fig.5shows m/z spectra of the fragment ions formed by 805 nm SFI of cis-and trans-1,2-DCE diluted in He (red and black traces, respectively) at I = 260, 650 and 1300 TW cm À2 , plotted so that the Cl + (m/z B 35.5) feature appears with a common peak intensity.Except at low m/z, the resolution of the present experiment is not sufficient to distinguish features differing by a single mass unit, but peaks attributable to H x C 2 Cl + , H x CCl + , Cl + , C 2 H x + , CH x + (x = 0-2) and H + ions are readily identifiable in the spectra recorded at I = 260 TW cm À2 , as are additional peaks assigned to Cl 2+ , Cl 3+ and C 2+ ions when exciting at higher I. Contributions from the parent dication and from fragment cations like HCCl 2+ , etc., are not immediately assignable from the mass spectrum but are revealed by inspecting the m/z resolved images, as illustrated below.Isomer dependent differences are immediately obvious in these m/z spectra, including a weak signal due to translationally excited Cl 2 + products from cis-1,2-DCE (shown in the inset in Fig.5(a)).The I-dependent images associated with these various fragment ions provide much new data about the fragmentation of higher Z states of 1,2-DCE.Here we focus particularly on the parent ions with Z = 2 and 3. Fig.6shows data for the H x CCl + and H x CCl 2+ fragments (which, in both cases, are likely dominated by the x = 1 cations).To aid identification, all images and data pertaining to products from SFI of the cis-isomer are shown with a red frame.The images of the HCCl + (m/z B 48.5) ions from both isomers show an anisotropic ring of signal at large radius, the intensity of which peaks along the axis defined by e.The images measured at low I also show an obvious central feature (attributable to parent dications), the relative intensity of which declines with increasing I.The P(v) distributions of the HCCl + fragments from both isomers (Fig.6(c)) are dominated by a peak at v B 2700 m s À1 .A higher velocity feature becomes increasingly evident at higher I.The HCCl 2+ image measured for both isomers at low I is weak, but the images recorded following NIR SFI at I = 1300 TW cm À2 show an obvious anisotropic ring of signal.

Fig. 5
Fig. 5 TOF spectra of fragment ions observed following 805 nm SFI of cis-and trans-1,2-dichloroethene (red and black traces, respectively) at I = (a) 260, (b) 650 and (c) 1300 TW cm À2 , displayed with a mass/charge (m/z) scale.The various spectra have been plotted so that the Cl + peak is displayed with a common intensity and, for ease of display, the intensity of the H + peak (i.e. the m/z o 2 region) has been reduced by factors of 4 and 8 in panels (b and c), respectively.The inset in panel (a) highlights the relatively much greater yield of Cl 2 + ions with v 4 500 m s À1 from SFI of

Fig. 6
Fig. 6 Symmetrised images of the (a) H x CCl + and (b) H x CCl 2+ fragments observed following 805 nm SFI of cis-(red borders) and trans-1,2-DCE at I = 260, 650 and 1300 TW cm À2 .The laser polarisation vector, e, is vertical in the plane of the image, as shown in the top left-hand panel and the relative signal intensities are shown by the linear false-colour scale chosen to span the maximum to minimum signal levels in the regions of interest within each image at the end of row (b).The ticks on the white horizontal scale in the left-hand image in each row indicate 1000 m s À1 intervals.The P(v) distributions of the H x CCl + and H x CCl 2+ fragments from each parent isomer are shown in (c) with, in each case, the maximum signal scaled to an intensity of unity.The corresponding fragment kinetic energies are shown on the top x axis.(d and e) Show, respectively, the (H x CCl + , H x CCl + ) and (H x CCl 2+ , H x CCl + ) covariance map images from analysis of I = 260 TW cm À2 data from the cis-(left) and trans-isomers (right), selecting H x CCl + as the reference ion and fixing its velocity to be vertically upwards, as indicated by the red arrow.The covariance signal of interest in each case is bounded by dashed white lines, while the more intense feature along the f = 01 axis in both images in (d) is the autovariance signal.

Fig. 7 2 +,
Fig. 7 Symmetrised images of the (a) Cl + and (b) H x C 2 Cl + fragments observed following 805 nm SFI of cis-(red borders) and trans-1,2-DCE at I = 260, 650 and 1300 TW cm À2 .e is vertical in the plane of the image, as shown in the top left-hand panel, and the ticks on the white horizontal scale in the lefthand image in rows (a and b) indicate 1000 m s À1 intervals.The relative signal intensities in the images in row (b) are displayed using the same linear falsecolour scale as in row (a) and on a 100Â increased scale (left and right half images, respectively).(c) P(v) distributions of the Cl + and H x C 2 Cl + fragments from each parent isomer with, in each case, the maximum signal scaled to an intensity of unity.These confirm that the Cl + images recorded at higher I extend to (and beyond) the edge of the detector area.(d) and (e) (H x C 2 Cl + , Cl + ) covariance map images obtained following SFI of each isomer at I = 1300 and 650 TW cm À2 , respectively.In all cases, Cl + has been selected as the reference ion (and its velocity fixed to be vertically upwards, indicated by the red arrow) and the (weak) covariance signals of interest are bounded by dashed white lines.

Fig. 8
Fig. 8 (a) Symmetrised images of the C 2 H x + fragments observed following 805 nm SFI of cis-(red borders) and trans-1,2-DCE at I = 260, 650 and 1300 TW cm À2 .The data are plotted using the linear false-colour scale shown at the far right of this row (left halves) and on a 10Â expanded scale (right halves), wherein more of the central feature (attributed to C 2 H 2 + ions from secondary decay of H 2 C 2 Cl + products formed in the DI channel (2)) has deliberately been saturated to reveal the faster component of these P(v) distributions more clearly.e is vertical in the plane of the image, as shown in the top left-hand panel.(b) P(v) distributions of the C 2 H x + fragment ions derived from these images.(c) Covariance map images from analysis of I = 260 TW cm À2 data for cis-(left) and trans-1,2-DCE (right), selecting Cl + as the reference ion (and fixing its velocity to be vertically upwards, indicated by the red arrow) and displaying the correlated 2-D velocity distributions of the C 2 H x + fragments in the frame defined by the reference ion.The covariance signal of interest in each case (bounded by dashed white lines) is relatively 'fuzzy', but notably more extensive in the case of the cis-isomer.

Fig. 9
Fig. 9 Newton plots of the products arising from the triple coincidence channel (process (7)) following NIR SFI of (a) cis-and (b) trans-1,2-DCE.The momenta of the C 2 H 2 + fragment (upper half) and one of the Cl + fragments (lower half) are displayed in the frame of the momentum of the second Cl + fragment, which is shown as a horizontal black arrow.The momentum vectors of the C 2 H 2 + fragment and the first Cl + fragment are normalised to the length of the momentum vector of the second Cl + fragment (reproduced from ref. 80, with permission).

Fig. 10 compares
images of the CF 3 + , I + and I 2+ ions formed in a two-colour experiment performed in Bristol involving a jet-cooled sample of CF 3 I in He, 267 nm pump photons and 805 nm SFI (I B 1300 TW cm À2 ) at eight different pump probe time delays in the range À390 fs r dt r 1 ns.In each case, the raw two-colour image (and the corresponding NIR-only image) has been centred, symmetrised and 'cleaned' by re-setting any extreme signals associated with hot pixels.The difference images shown in Fig. 10 were then derived by subtracting the NIR-only signal from the two-colour image to minimise the one-colour contribution at each dt and, simply for the present illustration purposes, the intensity in any pixels returning a negative difference signal were then reset to zero.Each image is displayed in two halves, with the right half plotted on a 10Â expanded intensity scale to reveal weaker features induced by the 267 nm pump pulse more clearly.The original (symmetrised) images, without the NIR-only background subtraction, are shown in Fig. S3 in the ESI.† All I + images obtained at dt 4 0 show a ring displaying parallel recoil anisotropy (i.e. with v aligned preferentially parallel to e), labelled DISS (for 'dissociation').A much weaker DISS feature is discernible in the corresponding CF 3 + images.

Fig. 10
Fig. 10 Symmetrised images of the CF 3 + , I + and I 2+ fragments from ultrafast two-colour (267 nm pumpÀ805 nm SFI probe, I = 1300 TW cm À2 ) studies of CF 3 I at different time delays in the range À390 fs r dt r 1 ns after subtracting the one-colour contribution obtained with just the NIR pulses.The e vector of both laser beams is vertical, in the plane of the detector, as indicated by the double-headed arrow in the dt = 0 CF 3 + image and illustrative examples of the various channels are labelled with the acronyms defined in the text.The relative intensities in images displayed in any one row are depicted using a common linear false-colour scale spanning the maximum through minimum signal values shown at the far right (left halves) and on a 10Â expanded scale (right halves).

DCE 1 and
DCE 2 (for 'dynamic Coulomb explosion'), with radii that decrease with increasing dt.These features are better viewed via the false-colour 2-D maps showing the integrated (over all angles) ion yields as functions of velocity and dt (over the range À0.45 r dt r +10 ps) shown in Fig.11.The maps displayed here are derived from images such as those shown in Fig.10, i.e. after subtracting the optimum amount of the corresponding one-colour (NIR-only) image and resetting any resulting negative intensities to zero.The corresponding maps without the one-colour subtraction are shown in Fig. S4 in the ESI.† As Fig. 11(a) shows, this procedure leaves some residual noise at velocities associated with the one-colour NIR-only SFI signal (labelled ISFI) but clearly accentuates the two-colour features of current interest.Analogous CF 3 + , I + and I 2+ maps obtained using a lower NIR intensity (I B 650 TW cm À2 ) are displayed in Fig. S5 in the ESI.† The features labelled DCE 1 and DCE 2 in Fig. 11 and Fig. S5 (ESI †) arise from SFI of the CF 3 and I co-fragments from photodissociation of a common parent molecule.Post-SFI, the resulting CF 3 + + I + and CF 3 + + I 2+ ion pairs then Coulomb repel each other with a force that scales inversely with their separation, yielding the respective DCE 1 and DCE 2 signals.The velocities of these fragment ions are thus dt-dependent, as illustrated schematically in Fig. 12 (along with the mechanism responsible for the accompanying one-color ISFI signal).The CF 3 + products attributable to DCE 1 and DCE 2 merge at long dt with those arising via the DISS process, as shown in Fig. 11(a) and Fig. S5(a).Equivalent plots for the I + and I 2+ yields recorded in the same two-colour multi-mass imaging experiments are shown in Fig. 11(b) and (c).The I + plot shows the DISS feature much more clearly and confirms the carrier of the DCE 1 feature.As Fig. 11(d) shows, the momentum (P(p)) distributions of the CF 3 + , I + and I 2+ fragments sampled at long time delay . S5 (ESI †), the relative intensities of signals arising via the DISS and DCE channels are I-dependent.Such is to be expected, since the relative probability of ionising both partners (i.e. of generating DCE signals) increases with I. Third, the signal to noise (S/N) ratios in the data measured at I B 1300 TW cm À2 are sufficient

Fig. 11 2
Fig. 11 2-D false-colour maps showing (a) CF 3 + , (b) I + and (c) I 2+ ion velocities following 267 nm photolysis of CF 3 I as a function of dt obtained by angular integration of images like those shown in Fig. 10, recorded with 805 nm intensities I B 1300 TW cm À2 , with the various channels labelled using the acronyms introduced in the text.The relative intensities in each panel are depicted using the linear false-colour scale shown to the right of panel (b).(d) P(p) distributions of the CF 3 + , I + and I 2+ ions obtained by integrating data measured over the delay range 0.1 r dt r 1 ns, with the maximum p values

Fig. 12
Fig. 12 Illustration of the sources of important signatures evident in UV pump À805 nm SFI probe experiments at different time delays, dt, using CF 3 I as the exemplar and assuming removal of two electrons and a symmetric charge separation.(a) CF 3 I 2+ formation by 805 nm SFI alone, or with the UV and 805 nm pulses coincident in time (i.e.dt = 0).The subsequent CE of these dications results in the CF 3 + and I + products identified with the label ISFI in the respective 2-D maps.(b and c) CF 3 + and I + fragments formed by 805 nm SFI (at progressively later delays dt) of neutral CF 3 and I products formed by UV excitation at t = 0.The neutral photoproducts are formed with characteristic, thermochemically determined, recoil velocities (equal to that of the feature(s) labelled DISS in the 2-D maps).Subsequent SFI of these neutral fragments introduces an additional Coulombic repulsive contribution to the measured CF 3 + and I + velocities, the magnitude of which decreases with increasing dt (i.e.increasing inter-ion separation) as illustrated by the velocity vectors in (b and c).This process is responsible for the characteristic DCE signals, the velocities of which converge to that of the DISS component at long dt.

Fig. 13 2
Fig. 13 2-D false-colour maps showing (a) CH 3 + , (b) I + and (c) I 2+ ion velocities as a function of dt obtained by angular integration of the respective images (after appropriate subtraction of the NIR-only contribution) following 267 nm photodissociation of CH 3 I and 805 nm SFI probing at I B 1300 TW cm À2 .The data are plotted using the linear false-colour scale shown at the far right of panel (b), and the various channels labelled with the acronyms introduced in the text.(d) Comparison of the P(p) distributions of the CH 3 + , I + and I 2+ ions obtained by integrating data measured over the

Fig. 14 2
Fig. 14 2-D false-colour maps showing the velocities of (a) CH 3 + and (b) I + ions detected in coincidence, measured over the delay range À0.28 r dt r Cl bond formation in tandem with C-Cl bond rupture.Summarising, new fragmentation behaviour provided by CEI studies of cis-and trans-1,2-dichloroethene include the operation of rival charge symmetric CQC and C-Cl bond fission channels from the dication of both isomers, as well as Cl 2 + elimination from cis-1,2-DCE (processes (

Table 1
Triatomic and larger molecules for which the fragmentation of their polycations (with charge Z+) have been investigated by ultrafast laser induced CEI methods.Molecules are listed in order of increasing size.The numbers in parentheses indicate the range of Z values investigated in the various cited works.Ionisation in these studies was induced by NIR SFI, apart from in those references shown in bold font which employed XUV or X-ray core excitation This journal is © the Owner Societies 2023 Phys.Chem.Chem.Phys., 2023, 25, 16672-16698 | 16683

Table 2
Fragmentation studies involving isolated polyatomic molecules and singly-and doubly-charged molecular cations explored by ultrafast pump-probe experiments using CEI probe methods This journal is © the Owner Societies 2023 Phys.Chem.Chem.Phys., 25, 16672-16698 | 16689