Volume 100, Issue 6, 15 March 1994
Index of content:

Infrared diode laser spectroscopy of the ν_{3} fundamental and ν_{3}+ν_{5}−ν_{5} sequence bands of the C_{4} radical in a hollow cathode discharge
View Description Hide DescriptionThe infrared absorptionspectrum of the linear C_{4} radical has been studied in an extension of the original observation of gas‐phase C_{4} by Heath and Saykally [J. Chem. Phys. 94, 3271 (1991)]. The experiment was performed using a flowing mixture of acetylene and helium subjected to a hollow‐cathode discharge, which was probed in the 1525–1570 cm^{−1} spectral region using a tunable diode laser spectrometer. Transitions with N‐values up to 60 were measured. Their analysis yielded band origins, rotational, and centrifugal distortion parameters for the lower and upper vibrational states, and l‐type doubling parameters for the degenerate bending states ν_{5} and ν_{3}+ν_{5}. In particular, the ν_{3} origin was determined to be 1548.6128(4) cm^{−1}, the ground state rotational and centrifugal distortion parameters were B=4979.89(21) MHz and D=0.848(44) kHz, and the l‐doubling parameters for ν_{5} was q _{5}=10.98(13) MHz. This value for q _{5} was used to estimate the ν_{5} frequency of gas‐phase C_{4} to be 160±4 cm^{−1}. Both the l=0 and 2 components of the ν_{3}+2ν_{5}−2ν_{5}sequence band were also tentatively observed, but a detailed analysis was not yet possible. The results were completely consistent with a linear structure for the triplet ground state of C_{4}, and showed no effects of quasilinearity such as that exhibited by C_{3}.

On the static dielectric function of the sticky electrolyte model
View Description Hide DescriptionThe static dielectric function of the sticky electrolyte model of electrolyte solutions is studied. We consider a model with bonding distance L=1/2. It contains steric saturation for association at dimer level. At increasing polarity of the diatomic molecules the ordering in the system changes and is manifested in the shape of the static dielectric function ε(k). The short ranged order is characterized by the presence of negative branch of ε(k).

266 nm CH_{3}I photodissociation: CH_{3} spectra and population distributions by coherent Raman spectroscopy
View Description Hide DescriptionHigh resolution coherent anti‐Stokes Raman spectroscopy has been used to study the symmetric CH stretching mode of methyl radical formed by UV laserphotolysis of CH_{3}I cooled in a free jet expansion. Spectra obtained under near‐nascent conditions (∼3 to ∼6 collisions) show that most of the CH_{3} product is formed in the ground vibrational state, with little excitation seen in the ν_{2} out‐of‐plane bending coordinate [v _{2}=1/v _{2}=0, population ratio 0.27(10)]. This is in accord with recent theoretical calculations favoring slow, adiabatic CH_{3} relaxation from a pyramidal to planar structure as the C–I bond breaks. Extensive N,K rotational structure is resolved and the distributions obtained lend support to those deduced for nascent CH_{3} by Chandler and co‐workers from modeling of unresolved resonance enhanced multiphoton ionizationspectra. The results are consistent with conservation during dissociation of CH_{3} ortho‐para nuclear spin forms and of K spinning angular momentum about the symmetry axis. CH_{3} product tumbling motion is slightly greater (1–2 units of angular momentum) than predicted theoretically. Rapid collisional excitation of higher rotational levels is seen and the rich spectrum observed after ∼50 collisions has been analyzed to give improved or new vibrational‐rotational parameters for 1000 and 1100 states. The 1100 band origin is determined to be 2996.21(4) cm^{−1}, yielding −8.23(5) cm^{−1} for the x _{12} anharmonicity constant.

Missing mode effect in resonance Raman excitation profile
View Description Hide DescriptionVibrational structure in the resonance Raman excitation profile (REP) of adenosinetriphosphate (ATP) in aqueous solution has been observed for both the 1333 and 1482 cm^{−1} modes, even though the absorptionspectrum is structureless. The ∼2100 cm^{−1} spacing of the vibrational structure in both cases is much larger than the normal mode frequencies of the molecule. It is possible for such a vibrational structure to develop in a quasidiatomic (single mode) case, based on the reflection principle of the product wave functions ‖〈Q‖1〉〈0‖Q〉‖^{2} appropriate for the fundamental REP, if the (dimensionless) displacement Δ between excited and ground state potentials is sufficiently large (Δ≳1) and the absorptive component of the REP dominates. However, this is not the case for ATP, where we show that the observed vibrational structure is due to an interplay of multiple modes. This ‘‘missing mode effect’’ (MIME) for REP, as in the case of luminescencespectra of large molecules, can be explained using the time‐dependent correlation function picture of electronic transitions. The harmonic model is used and it is shown how the MIME frequency in the REP arises. The phenomenon is illustrated with ATP, where the harmonic parameters are derived from its resonanceRaman spectrum at an excitation frequency of 260 nm, which are then used to calculate the absorption and REPs to compare with the experimental results.

On the interpretation and rotational assignment of degenerate four‐wave mixing spectra: Four‐photon line strengths for crossover resonances in NO A ^{2}Σ^{+}–X ^{2}Π
View Description Hide DescriptionWe present here a set of equations specifically adapted to simulation of fully resonant, high‐resolution, phase‐conjugate degenerate four‐wave mixing (DFWM) in molecular gases. Signal‐intensity dependence on molecular wave functions, lifetimes, and laser beampolarizations is explicitly included in these equations. The emphasis of the presentation is on both physically intuitive interpretation and a practical, ‘‘cookbook’’ approach to spectral simulation. We present experimental verification of our calculations drawn from the spectrum of dilute NO in N_{2} at low pressures. Both degenerate two‐level and three‐level (crossover) resonances were observed. The experimental spectral intensities are accurately reproduced by the expressions presented here. We point out some of the subtleties of DFWMspectra that could be used as aids to interpretation, especially the use of laser polarization as a probe for spectral line assignments.

High resolution infrared spectroscopy of pyrazine and naphthalene in a molecular beam
View Description Hide DescriptionThe high resolution infrared spectrum of pyrazine and naphthalene were measured in a molecular beam in the vicinity of the C–H stretching transition. The rotational structure in the spectrum of pyrazine from 3065–3073 cm^{−1} reveals that the C–H stretch is coupled to one other vibrational mode in the molecule. The mode coupling is manifested in the spectrum as two overlapping vibrational bands. Each of these two bands are well modeled by an asymmetric top/rigid rotor Hamiltonian. The lack of any angular momentum dependence on the coupling indicates that the vibrations are coupled by an anharmonic mechanism. The magnitude of the coupling matrix element was determined to be 0.36 cm^{−1}. The rotational structure in the spectrum of naphthalene from 3063–3067 cm^{−1} reveals that except for several local perturbations, the spectrum is well modeled by an asymmetric top/rigid rotor Hamiltonian. The local perturbations include transitions that are split into doublets as well as transitions that have been shifted from their expected positions. The magnitude of the average coupling matrix element for the doublets was determined to be 0.0016 cm^{−1}. A comparison between the vibrational mode coupling in pyrazine and naphthalene indicates that mode coupling does not correlate with the density of states in the two molecules.

The σ* absorption peak at the oxygen 1s edge of O_{2}: Exchange splitting, ultrafast dissociation, and atomiclike Auger spectra
View Description Hide DescriptionThe x‐ray absorptionspectrum of solid O_{2} at the O 1sabsorption edge is analyzed, using its polarization dependence and the Auger de‐excitation spectra.Rydberg states are quenched in the solid, and the exchange splitting of the 1sionization threshold (1.1 eV) can be observed. Below the ionization threshold, core electrons can be excited into unoccupied antibonding π* and σ* orbitals. We conclude from the relative intensities and from the polarization dependence, that the exchange splitting of the σ* final state is small (<1 eV). This is confirmed by LDA calculations of core‐excited states. The calculated σ* potential surfaces are repulsive, which explains the large width of this absorption peak. Because of ultrafast dissociation (∼5 fs), core hole decay is likely to occur at large interatomic distances. Auger spectra at the σ* absorption of O_{2} do indeed show peaks that would be expected for free core‐excited oxygen atoms.

Hybrid simulations of solvation effects on electronic spectra: Indoles in water
View Description Hide DescriptionSolute–solvent interactions and dynamics are simulated with a fully molecular hybrid method consisting of a semiempirical quantum mechanical method with singly excited configurations for the solute and classical molecular dynamics (MD) for the solvent (H_{2}O). The interactions are purely electrostatic, with the solute being polarizable and sharing its charge information with the MD at 5 fs intervals. The solvent charges are fixed and the results are not sensitive to the point charges used. For the solute, the results depend on the dipole moment much more than on the point charge magnitudes leading to a given dipole. This method is applied to the spectral shifts, dynamics, linewidths, and free energies of indole and 3‐methylindole (3MI) in water at 300 K, including the effect of geometry changes and clarifications concerning vertical vs 0–0 transition predictions. Large fluorescence Stokes shifts are predicted, in fair agreement with observed values. The ^{1} L _{ a }excited state dipole is calculated to be about 12 D after solvent relaxation following excitation. This increase of about 5 D above that calculated in vacuum is caused by the solvent reaction field, and approximately doubles the calculated shift compared to that using the vacuum dipoles. There does not seem to be a need to invoke a solute–solvent excited state charge transfer complex (exciplex) to account for the large shifts. About 50% of the Stokes shift occurs in ∼15 fs with a Gaussian response function, and the remainder is approximately an exponential with τ=400 fs. The fast component is created by small rotational deviations in the trajectories of a few nearby waters. The change in free energy of solvation upon excitation is found to be half the sum of the absorption and fluorescence shifts.

SiC_{2}: A molecular pinwheel
View Description Hide DescriptionWe present the results of a combined experimental and theoretical study of the large‐amplitude motion in SiC_{2} in which the C_{2} fragment undergoes hindered internal rotation. Stimulated emission pumping (SEP) is used to obtain rovibrational term energies for levels with up to 14 quanta of excitation in the large‐amplitude vibration. We analyze the SEP data, as well as other available experimental data, using a semirigid bender model that allows for complete internal rotation within a triatomic molecule. From the least‐squares fitting of this model to the data, we determine the potential energy along the minimum energy path of the large‐amplitude vibration, the harmonic energies of the small‐amplitude vibrations, and the variations of these energies and of the molecular geometry with the large‐amplitude coordinate. The fitting is aided by results obtained from ab initio calculations we perform on the triangular and linear configurations of the molecule. The current data set is consistent with a large‐amplitude potential energy function in which the energy difference between the triangular and linear configurations is 1883 cm^{−1}. The statistical error on this energy difference is 22 cm^{−1}, but we estimate the physical uncertainty to be about 200 cm^{−1}. This result is in excellent agreement with the energy difference of 1819 cm^{−1} we obtain in our best ab initio calculations. The semirigid bender fitting and our best ab initio calculations are also both consistent with a potential energy function having no local minimum at linearity.

Determination of hyperfine interaction matrix principal values and principal axis orientations in an orientationally disordered solid: A multifrequency electron spin echo envelope modulation study of nitrogen‐15 in a copper(II)–^{15}N–imidazole complex
View Description Hide DescriptionA comprehensive experimental study of the magnetic field dependences of ESEEM (electron spin echo envelope modulation) for a spin one‐half nucleus in an orientationally disordered solid is presented. Modulation effects from the remote (unligated) nitrogen of ^{15}N‐labeled imidazole in a Cu(II)–diethylenetriamine–^{15}N‐imidazole complex were registered at electron spin excitation frequencies in the 4.3 to 11.5 GHz range, which encompasses the ‘‘match range’’ of the Cu(II)–^{15}N system under study. Field dependent trends in the ESEEM spectra—trends in spectral amplitudes, linewidths, and frequencies—are analyzed to obtain the magnitude and relative signs of the isotropic and axially symmetric hyperfine coupling constants (and to place an upper bound on the rhombic coupling). The relative utility of each of these trends for measuring hyperfine interactions in orientationally disordered solids is discussed. The orientation of the unique axis of the Cu(II)–^{15}N hyperfine interaction matrix within the Cu(II) electron spin magnetic axis system is obtained from the field dependence of the ESEEM frequencies observed with orientationally selective excitation. The results of this study are compared to those previously obtained for relevant Cu(II)–^{15}N systems by a variety of monofrequency ESEEM techniques.

Laser spectroscopy of the low‐lying electronic states of NbN: Electron spin and hyperfine effects in the states from the configurations σδ and δπ
View Description Hide DescriptionRotational and hyperfineanalyses have been carried out for the (0,0) bands of the C ^{3}Π–X ^{3}Δ, e ^{1}Π–X ^{3}Δ, and f ^{1}Φ–a ^{1}Δ transitions of gaseous NbN from laser excitation spectra taken at sub‐Doppler resolution. The δπ C ^{3}Π and e ^{1}Π states lie only 102 cm^{−1} apart in zero order but the spin–orbit matrix element between them, which is the sum of the spin–orbit constants for the δ and π electrons, is 698 cm^{−1}; as a result the ^{3}Π_{1} spin component lies below both the ^{3}Π_{0} and ^{3}Π_{2} components, and its hyperfine structure is highly irregular. This irregularity is an extreme example of how cross terms between the spin–orbit interaction and the Fermi contact hyperfine operator alter the apparent value of the hyperfinea constant, the coefficient of I⋅L in the magnetic hyperfine Hamiltonian. Molecular parameters for the C ^{3}Π and e ^{1}Π states have been obtained from a combined fit to the two of them. Including data for the B ^{3}Φ state recorded earlier [Azuma et al., J. Chem. Phys. 91, 1 (1989)], detailed information is now available for all six of the electronic states from the electron configurations σδ and δπ. It has been verified that the spin–orbit/Fermi contact cross terms cause roughly equal and opposite shifts in the hyperfinea constants for the singlet states and the Σ=0 components of the triplet states. After allowing for this effect, it has been possible to interpret the hyperfinea constants in terms of one‐electron parameters for the δ and π electrons, in similar fashion to spin–orbit parameters. Wavelength resolved fluorescence, following selective laser excitation of the C ^{3}Π, e ^{1}Π, and f ^{1}Φ states, has led to the discovery of three new electronic states, δ^{2} c ^{1}Γ, δ^{2} A ^{3}Σ^{−}, and σ^{2} b ^{1}Σ^{+}, besides giving the absolute position of a ^{1}Δ. Strong configuration interaction mixing is found to occur between the σ^{2} b ^{1}Σ^{+} and δ^{2} d ^{1}Σ^{+} states. The low‐lying electronic states of NbN are now well understood.

The B←X electronic spectra of N^{+} _{2}–Ne_{ n } (1≤n≤8)
View Description Hide DescriptionSpectra of ^{14}N^{+} _{2}–^{20}Ne, ^{14}N^{+} _{2}–^{22}Ne, and ^{15}N^{+} _{2}–^{20}Ne have been recorded in the region of the B ^{2}Σ^{+} _{ u }←X ^{2}Σ^{+} _{ g } origin transition of N^{+} _{2}. Measurements are made by mass selecting cooled ionic complexes and photodissociating them whilst monitoring the N^{+} _{2} fragment ion intensity as the laser wavelength is scanned. Various bands are assigned to transitions involving the stretching and bending motions of the Ne...Ne^{+} _{2}bond with their structure and spacings consistent with transitions between quasilinear geometries in the X and the B states. Spectra of complexes with up to eight neon atoms attached to a ^{14}N^{+} _{2} core have also been measured. Evidence from shifts of the band origins and analysis of the vibrational frequencies of N^{+} _{2}–Ne_{2} and N^{+} _{2}–Ne_{3} suggest a structure where the Ne ligands are sited at one end of the N^{+} _{2} chromophore.

Perturbative treatments of pump–probe laser‐molecule interactions with applications to azulene and trimethylazulene
View Description Hide DescriptionSemianalytic perturbative approaches for investigating the spectroscopy, and the underlying dynamics, associated with fast time‐resolved, two‐photon, two‐color, pump–probe, low intensity laser‐molecule interactions are developed and discussed. In particular the perturbation theory is developed with emphasis on molecular models associated with pump–probe experiments on the S _{0}→S _{1}→S _{2} two‐photon transitions in azulene and trimethylazulene. The experiments are discussed and the theory is used to determine the lifetimes of the intermediate S _{1} states by analyzing the experimental two‐photon fluorescence signals from the S _{2} states as a function of the time‐delay between the pump and probe lasers. The advantages of using this approach relative to the traditional methods for determining the lifetime of the S _{1} states are discussed. The dependence of the two‐photon fluorescence signals on the angle between the polarization vectors of the pump and probe lasers, for fixed time delay between the lasers, is also considered briefly.

Microwave spectroscopic investigation of the mixed rare gas van der Waals trimers Ne_{2}–Kr and Ne_{2}–Xe
View Description Hide DescriptionPure rotational spectra of six isotopomers of the rare gas trimer Ne_{2}–Kr and five isotopomers of the trimer Ne_{2}–Xe have been measured in their ground vibrational states using a Balle–Flygare‐type cavity pulsed microwave Fourier transform spectrometer. Rotational constants have been evaluated, from which the geometries of the complexes have been derived. In the case of Ne_{2}–Kr it has been possible to obtain centrifugal distortion constants and to carry out a force field analysis. The magnitudes of the induced dipole moments of these trimers have been estimated from the ‘‘π/2 condition.’’ The nuclear quadrupole hyperfine patterns due to ^{83}Kr and ^{131}Xe have been resolved, and the corresponding quadrupole coupling constants have been obtained. Observation of the spectra of these fundamental trimers has made it possible to compare their properties to those of their constituent rare gas dimers. The effects of three‐body nonpairwise additive forces have been discussed in light of the structures obtained, the estimated induced dipole moments, and the nuclear quadrupole coupling constants.

The long‐time behavior of reversible binary reactions: Theory, Brownian simulations and experiment
View Description Hide DescriptionMany‐body effects on reversible pseudo‐unimolecular reactions are investigated using a combination of theory, simulation, and experiment. Theoretically, we rederive the superposition approximation starting from the fundamental N‐particle equations. All the relations obtained are actually rigorous, except for a requirement that the concentration profile outside a vacant trap obeys a diffusion equation. Our derivation also yields a new numerical procedure for evaluating the superposition solution.Brownian dynamics simulations of one‐dimensional competitive binding are presented over an unprecedented time regime. Comparison with the superposition approximation shows that this mean‐field theory is exact at infinite dilution, but breaks down at high particle concentration. The main discrepancy is not at asymptotically long times as previously suspected, but rather at intermediate times, where a new power law‐phase emerges. This is reflected in a maximum in the logarithmic derivative of the survival probability, which is more pronounced in our simulation as compared with the approximate theory. Finally, we show that the transient fluorescence data from an excited dye molecule which transfers a proton reversibly to water, develops a similar maximum in its logarithmic derivative at low pH values.

Rotational resonance states of Ar–HCl(v=0) by finite range scattering wave function method
View Description Hide DescriptionThe low lying rotational resonance states of Ar–HCl van der Waals molecule in the vibrational ground state of HCl are calculated for several total angular momentum states within the model of a rigid rotor‐atom system. The necessary scattering calculations are done by the finite range scatteringwave function (FRSW) method [J. Chem. Phys. 99, 1057 (1993)] adopting energy independent auxiliary functions which makes the scattering calculations at many energies much more efficient. Discrete eigenvectors and eigenvalues of Hamiltonian matrix are calculated on a finite range via the successive diagonalization‐truncation scheme combined with the discrete variable representation (DVR). Analytical eigenfunctions of the asymptotic Hamiltonian operator in a body‐fixed frame excluding only the interaction potential, but including all the effective centrifugal potential terms, are used as asymptotic wave functions, which reduces the dynamical range required for the L^{2}scattering calculations. After a single diagonalization of the Hamiltonian in the finite range L^{2} representation, resonance parameters are extracted through the energy dependence of Smith’s lifetime matrix. A search algorithm for multiple resonances is used assuming constant background effects and isolated, simple resonances. Several recommendations for the choice of the translational basis functions and the basis set ranges which target specific kinds of resonances with various accuracies are given. Resonance energies and widths (lifetimes and partial decay probabilities) for many predissociating states of Ar–HCl are given with unprecedented accuracy.

Cluster effects in O_{3}/H_{2}O photochemistry: Dynamics of the O+H_{2}O→2OH reaction photoinitiated in the O_{3}⋅H_{2}O dimer
View Description Hide DescriptionThe dynamics of the 266 nm photoinitiated reaction of ^{16}O_{3} and H_{2} ^{18}O were studied using ^{16}O_{3}⋅H_{2} ^{18}O van der Waals dimers to orient the initial reagents. In the absence of perturbations, the geometry of the ^{16}O_{3}⋅H_{2} ^{18}O dimer is such that 266 nm photolysis of cluster‐bound ozone initiates glancing O+H_{2}O trajectories, with a 3 Å impact parameter. Laser induced fluorescence probes show that 81±7% (2σ) of the ‘‘new’’‐^{16}OH and essentially all of the ‘‘old’’‐^{18}OH products were formed with v=0, with a slight preference for the Π(A’) Λ doublets, and average rotational energies of 900±130 and 760±80 cm^{−1}, respectively. Approximately 19% of the ‘‘new’’‐^{16}OH products form with v=1 and average rotational energy of 930±210 cm^{−1}. No significant OH scattering anisotropy or other vector correlations were observed. Sub‐Doppler resolution experiments showed average kinetic energies for new‐^{16}OH(v=0) products about 19% higher than for old‐^{18}OH(v=0) products in the same rotational levels; increasing from values of about 500 cm^{−1} at low rotational levels, to about 1500 cm^{−1} at the highest rotational levels populated. Similar OH internal and kinetic energies were observed when the clusters were photolyzed at 281.5 nm. These dimer results are very different from those observed for the bimolecular O(^{1}D)+H_{2}O→2OH reaction, photoinitiated in gas phase mixtures of ^{16}O_{3} and H_{2} ^{18}O. The gas phase O(^{1}D)+H_{2}O→2 OH reaction produces OH with pronounced recoil anisotropy, these OH products carry far more internal energy than seen in the cluster products, and there is greater disparity between the internal energies of the gas phase ^{16}OH and ^{18}OH products. Evidently, cooperative effects in the cluster environment result in a significant change in reaction path.

Optimally controlled five‐laser infrared multiphoton dissociation of HF
View Description Hide DescriptionSimulations of the quantum dynamics of the HF molecule immersed in a field of five overlapping, intense, linearly polarized, infrared laser pulses of subpicosecond duration are performed. The HF molecule, initially in its ground state, is modeled as a rotating oscillator interacting with a classical laser field via electric dipole interaction. Realistic potential and dipole functions are used. Optimal overlaps of the five laser pulses, as well as the optimal carrier frequencies of the laser pulses, are found which maximize the HF dissociation yield. A maximal yield of 45% in a single combined pulse is achieved using the best available potential and dipole moment functions. The optimal infrared multiphotondissociation pathway for the HF molecule includes a series of the Δv=1 vibrational‐rotational transitions followed by a series of Δv≥2 vibrational‐rotational transitions. The latter is necessary as a consequence of the vanishing Δv=1 transition moment around v=12. In the Δv=1 regime, both P and R branch transitions are found to be important. The angular distribution of the dissociative flux is computed. Robustness of the results with respect to changes in the interatomic potentials, dipole functions and reduced mass, as well as to changes in laser pulse parameters (carrier frequencies, timings, phases, field amplitudes, and pulse durations) is investigated.

Ultralow temperature kinetics of neutral–neutral reactions. The technique and results for the reactions CN+O_{2} down to 13 K and CN+NH_{3} down to 25 K
View Description Hide DescriptionAn entirely new experimental method is described which enables the rate constants of neutral–neutral gas‐phase reactions to be measured at ultralow temperatures. The measurements are made by applying the pulsed laserphotolysis (PLP), laser‐induced fluorescence(LIF) technique of studying the kinetics of free radical reactions in the ultracold environment provided by the gas flow in a Cinétique de Réaction en Ecoulement Supersonique Uniforme (CRESU) apparatus. The experimental method is described in some detail and its application and limitations are discussed. Results are reported for the reactions of CN radicals with O_{2} and NH_{3}. For reaction (1) between CN and O_{2} data are reported for the temperature range T=13–295 K and the rate constants are well‐matched by the expression k _{1}(T)=(2.49±0.17)×10^{−11} (T/298)^{(−0.63±0.04)} cm^{3} molecule^{−1} s^{−1}. For reaction (2) between CN and NH_{3}, rate constants in the temperature range T=25–295 K fit the expression k _{2}(T)=(2.77±0.67)×10^{−11} (T/298)^{(−1.14±0.15)} cm^{3} molecule^{−1} s^{−1}. The kinetic data are discussed in terms of the latest quantum chemical and reaction rate theories for these systems.

Photodissociation of ICN in solid and in liquid Ar: Dynamics of the cage effect and of excited‐state isomerization
View Description Hide DescriptionPhotodissociation of ICN by UV excitation in solid and liquid Ar is studied by molecular dynamics simulations. The focus is on the differences between the cage effects on the CN photoproduct in the two phases, and on the excited stateisomerization ICN*→INC* dynamics in the solid matrix. Nonadiabatic transitions are neglected in this first study. The main results are: (1) No cage exit of the CN product is found in solid Ar, even in simulations at temperatures close to melting and for large excess energies. The result is in accord with recent experiments by Fraenkel and Haas. This should be contrasted with the large cage‐exit probabilities found in many systems for atomic photofragments. The result is interpreted in terms of geometric and energy transfer considerations. It is predicted that complete caging of diatomic and larger photofragments will be typically the case for photodissociation in rare‐gas matrices. (2) Almost 100% cage‐exit probability for the CN product is found for ICN photolysis on the ^{1}Π_{1}potential surface in liquid Ar. On the other hand, photolysis on ^{3}Π_{0+ }potential surface does not lead to cage exit on a time scale of 15 ps. The large differences between the reaction in the solid and in the liquid, and between the behavior of the process on the ^{3}Π_{0+ } and the ^{1}Π_{1} potentials, respectively in the liquid, are interpreted. (3) CN rotational dynamics and subsequent relaxation leads to isomerization in the excited electronic states. On the ^{3}Π_{0+ }potential surface one finds after t≳0.5 ps roughly equal amounts of the ICN and INC isomers. On the ^{1}Π_{1} surfaces only INC is found after t≳3.5 ps. This is explained in terms of the barriers for CN rotation in the two excited states, and in terms of the time scales for rotational relaxation. The results throw light on the differences between cage effects for photochemical reactions in solid and in liquidsolution, and on cage‐induced isomerizationdynamics in solid matrices.