Photodissociation dynamics of 2-chloro-6-nitrotoluene and nitrocyclopentane in gas phase: Laser-induced fluorescence detection of OH

Chemical Physics 443 (2014) 123–132

Contents lists available at ScienceDirect

Chemical Physics journal homepage: www.elsevier.com/locate/chemphys

Photodissociation dynamics of 2-chloro-6-nitrotoluene and nitrocyclopentane in gas phase: Laser-induced fluorescence detection of OH Monali N. Kawade, Ankur Saha, Hari P. Upadhyaya, Awadhesh Kumar, Prakash D. Naik ⇑ Radiation & Photochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, India

a r t i c l e

i n f o

Article history: Received 1 July 2014 In final form 26 August 2014 Available online 6 September 2014 Keywords: Photodissociation dynamics Laser induced fluorescence Nitro compounds Nascent OH radical detection Energy partitioning

a b s t r a c t Photodissociation of 2-chloro-6-nitrotoluene (ClNT) at 193, 248 and 266 nm and nitrocyclopentane (NCP) at 193 nm leads to the formation of OH, as detected by laser-induced fluorescence (LIF). The nascent OH produced from the photolysis of ClNT at all the wavelengths is vibrationally cold, with the Boltzmann type rotational state distributions. However, the nascent OH product from NCP is in the ground and vibrationally excited states with the measured average relative population in m00 ¼ 1 to that in m00 ¼ 0 of 0.12  0.03, and these levels are characterized by rotational temperatures of 650  180 K and 1570  90 K, respectively. The translational energy partitioned in the OH fragment has been measured for photodissociation of both ClNT and NCP. On the basis of both the experimental results and the ground state molecular orbital (MO) calculations, a plausible mechanism for the OH formation has been proposed. Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction The nitroaromatic and nitro cyclic compounds are common industrial chemicals with some of them being toxic, which raise safety and environmental-hazard concerns. They are either directly emitted into the atmosphere due to industrial activities or formed in situ by secondary photochemical processes in the atmosphere. The photodissociation dynamics of these compounds is currently an interesting topic due to their relevance to the atmospheric and combustion chemistry, and with their use as explosives or taggants in explosive compounds. Moreover, the low barriers for the dissociation of these compounds have motivated work on their unimolecular decay. The UV photodissociation of a nitrocompound in general leads to the formation of NO2, NO, O, C, etc. through various channels [1–3]. The NO2 and NO products can be produced in both the ground and electronically excited states. The detection of vibrationally excited NO is considered as an indicator for the presence of explosives [1]. In many cases, the UV/Visible fluorescence has been detected from electronically excited products [4–6]. The dissociation of a nitrocompound strongly depends on its structure and the photolysis conditions [2,3]. In addition to these products, an unexpected OH product is observed in several UV

⇑ Corresponding author. Tel.: +91 22 25595398; fax: +91 22 25505151. E-mail address: [email protected] (P.D. Naik). http://dx.doi.org/10.1016/j.chemphys.2014.08.007 0301-0104/Ó 2014 Elsevier B.V. All rights reserved.

photoexcitation studies on nitrocompounds [6–14]. Recent work on photo-oxidation processes of aromatic hydrocarbon performed in the presence of hydroxyl radical scavengers revealed that OH radicals are generated during the photolysis of nitrophenols [8]. However, the OH radical has been observed directly in the photodissociation of o-nitrophenol [9,10] and nitrotoluene [7]. The OH radical is an important radical in the chemistry of the atmosphere and combustion. It is a key reactive species in the interstellar medium and one of the most important oxidants in the troposphere acting as a ‘‘detergent’’, as it reacts with many pollutants. Even one of the major photoproducts, NO2 participates in various atmospheric reactions, including ozone generation in the troposphere. Hence, the photodissociation of nitrocompounds is very much relevant from the point of view of atmospheric chemistry, in addition to the basic understanding of the dynamics of dissociation. The relative complexities of small nitrocompounds, which have several different dissociation pathways at the energies of ultraviolet photons, make their photodissociation dynamics particularly interesting. Photodissociation of nitrocompounds involving roaming mechanism for NO formation has been demonstrated [15,16]. The dynamical information on the photodissociation process can be obtained by measuring the energy disposal in the photo fragmentation process, and the nature of the dissociative state potential energy surface can be deduced. Thus, due to importance in explosive, atmospheric and combustion chemistry, the

124

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

photochemistry and photodissociation dynamics of nitrocompounds have been extensively investigated by many researchers. Our group has carried out a detailed study of photodissociation dynamics of aliphatic and aromatic nitrocompounds by employing laser photolysis – laser induced fluorescence (LP–LIF) [6,7,17]. The present studies, on halogenated aromatic nitrocompound and cyclic nitroalkane, are an extension to the previous work. In our earlier work, it was observed that on UV absorption, among o-, m- and p-nitrotoluene, only o-nitrotoluene produces the OH photofragment [7]. Similarly, an aliphatic nitrocompound, having a methyl group adjacent to a nitro group, produces the OH fragment [6,17]. Thus, the presence of a methyl group adjacent to the nitro group is important in nitrocompounds for intramolecular rearrangement prior to the OH formation. Can the presence of a halogen atom in a nitrocompound modify the dynamics of the OH formation? Moreover, is the presence of a methyl group, at all, necessary in a nitrocompound for the OH radical generation? In the present work, we have selected two interesting molecules, 2-chloro-6-nitrotoluene (ClNT), an o-nitrotoluene with a Cl atom adjacent to the methyl group, and nitrocyclopentane (NCP), a cyclic nitroalkane having no methyl group, to address these queries. The photodissociation dynamics studies of ClNT, at 193, 248 and 266 nm, and NCP at 193 nm, have been carried out under a collision-free condition and the OH radicals detected using LIF. A dissociation mechanism has been proposed in conjunction with theoretical calculations.

2. Experimental In the present work, the LP–LIF set up was used to study the photodissociation dynamics of ClNT and NCP. The set up remains similar to our earlier work, and described in detail in the previous work [18,19]. Briefly, the photolysis laser employed was an excimer laser (Lambda Physik Model Compex-102, Fluorine version) and the probe laser was a Quantel dye laser (TDL90) with a frequency doubling and mixing module, working on DCM special dye (Lambdachrome, LC 6501, 590–640 nm) and pumped by a seeded Nd:YAG laser (model: Quantel, YG981 E 20). The reaction chamber was made of Pyrex glass with the probe and the photolysis arms at the right angles to each other. All the arms were equipped with baffles, and the windows were fixed at the Brewster angle to reduce the scattering. The detection system viewed the intersection volume of the photolysis and the probe lasers through the bottom arm window. The fluorescence was collected by a 38 mm diameter lens of the focal length 50 mm, and detected by a PMT (Hamamatsu model R 928P). A band-pass filter (k-center = 310 nm, fwhm = 10 nm) was introduced between the lens and the PMT to cut off the scattering from the photolysis laser. The signal from the PMT was sampled with a boxcar integrator (SRS250), with the gate width of 100 ns, averaged over 30 laser shots and fed to an interface (SRS245) for A/D conversion. A Pentium IV PC was used to control the scan of the dye laser via a RS232 interface and to collect data through a GPIB interface using a control and data acquisition program. To correct for the laser intensity fluctuations, both the pump and the probe lasers were monitored by photodiodes, and the fluorescence intensities were normalized. The experiments were carried out in a flow reactor, at a typical parent pressure of 20 mTorr, employing LP–LIF pump and probe technique. 2-Chloro-6-nitrotoluene being a solid compound, a special experimental arrangement was devised to flow its vapor through the reaction chamber, as described in the previous paper [18]. We have not used any carrier gas during these experiments. After loading the sample, the sample holder along with the reaction chamber was degassed thoroughly by an oil diffusion pump,

which is backed with a rotary pump. The ClNT vapor pressures in the chamber were maintained at 20 mTorr, which were measured with the help of a capacitance manometer. The solid ClNT, supplied by Aldrich with 99% purity, was used after degassing, without further purification. The 99% pure nitrocyclopentane (NCP) sample was obtained from Aldrich, and used without further purification, but after 4–5 freeze–pump–thaw cycles. Low pressure (20–40 mTorr) of the sample was maintained in a flow cell, and was measured by a capacitance gauge. The ClNT was photolyzed at 266, 248, and 193 nm and NCP at 193 nm. The OH fragments generated on photolysis were probed state selectively in the wavelength region 306–309 nm, by exciting the A2 R X2 P (0, 0) transition of OH, and monitoring the subsequent A ! X fluorescence. In case of NCP, the product OH (v 00 = 1, J00 ) was detected by exciting the A2 R X2 P (1, 1) transition. A circular aperture is used to select the homogeneous part of the rectangular excimer laser beam profile. In the present studies, all the photolysis experiments were carried out at laser intensities less than 1.0 mJ/cm2. Both the laser beams are unfocused and attenuated to prevent any saturation effect and multiphotonic events that occur at high laser intensities. This has been checked by recording the variation of LIF intensity of the OH radical on laser energy and the ratio of the main and satellite peak intensity. We have also measured emission spectra to detect the formation of electronically excited products. No emission was observed in photolysis of ClNT, however the emission from excited NO2 was observed in photolysis of NCP at 193 nm. At higher photolysis laser intensities, the emission from the excited OH radical was also observed, which is attributed to occurrence of a biphotonic process at higher intensities. The experimental set up for measuring time-resolved UV–vis emission consisted of a Bausch and Lomb monochromator, a PMT (Hamamatsu model R928), and a digital oscilloscope (Tektronix). The sample was allowed to flow in a stainless steel cell, equipped with four perpendicular MgF2 windows, at a pressure of 20–200 mTorr, and monitored by a capacitance gauge. During the measurement of the emission at wavelengths greater than 350 nm, suitable cut-off filters were used. 3. Results The OH radical was observed on photolysis of ClNT at 193, 248 and 266 nm, as expected. Photolysis of even NCP, which is devoid

Fig. 1. The fluorescence excitation spectra of the (0, 0) band of the A2 Rþ system of OH formed on photolysis of ClNT at 193 nm.

X2 P

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

of any methyl group, at 193 nm generates OH. The Fig. 1 shows the fluorescence excitation spectra of the (0, 0) band of the A 2 Rþ X2 P system of OH, formed on photolysis of ClNT at 193 nm. The excitation spectra of OH observed on photolysis at 248 and 266 nm are quite similar to that at 193 nm, and hence not shown. A similar spectrum is recorded for NCP at 193 nm, where OH is formed in both the vibrationally ground state, v00 = 0, and the excited state, v00 = 1. The Fig. 2 shows the fluorescence excitation spectra of the (1, 1) band of the A2 Pþ X2 P system of OH, formed on photolysis of NCP at 193 nm. The standard nomenclature is used for designating the observed transitions. The subscripts 1 and 2 represent the transitions RJ¼Nþ1=2 PJ¼Nþ1=2 and RJ¼N1=2 PJ¼N1=2 , respectively, whereas the subscripts 21 and 12 represent the transitions RJ¼N1=2 PJ¼Nþ1=2 and RJ¼Nþ1=2 PJ¼N1=2 , respectively. According to the parity selection rule (+ $ ), the Q branch arises from the P (A00 ) state, while the P and R branches originate from the Pþ (A0 ) state. The individual lines are broadened due to the Doppler effect, but are still well separated so that each individual line profile could be fitted by a Gaussian function. The relative populations of the OH fragment were determined by normalizing the respective peak areas of the rotational lines with respect to the pump and probe laser intensities, the pressure change, if any, and the respective Einstein absorption coefficients [20]. Spin–orbit and the K-doublet ratios were calculated from the relative populations of different states. The translational energy associated with the OH fragment was calculated from the Doppler profiles of the rotational lines. The detailed results are presented below.

125

Fig. 3. Typical Boltzmann plots of the distribution of rotational energy in the nascent OH(ðm00 ¼ 0Þ) from photolysis of ClNT at 193, 248, and 266 nm.

3.1. Rotational and vibrational energy distributions The nascent rotational state populations of the OH radicals, generated on photodissociation of ClNT at 193, 248 and 266 nm and NCP at 193 nm, are used to construct a Boltzmann plot for obtaining the rotational temperature of the nascent OH fragments. Typical Boltzmann plots based on this normalized population distribution are shown in Fig. 3 for ðm00 ¼ 0Þ states of OH at 193, 248 and 266 nm wavelengths for ClNT, while Fig. 4 shows the Boltzmann plots for ðm00 ¼ 0Þ and ðm00 ¼ 1Þ states of OH at 193 nm for NCP. As shown in Fig. 3, the Boltzmann plots for P and Q lines at 193 and 266 nm are having different rotational excitation, while at 248 nm P and Q lines are having comparable rotational excitation. The corresponding rotational temperatures using P lines (Q lines) at 193, 248 and 266 nm are 1000  60 K (750  50 K),

Fig. 2. The fluorescence excitation spectra of the (1, 1) band along with (0, 0) band of the A2 Rþ X2 P system of OH formed on photolysis of NCP at 193 nm. All the lines marked below correspond to the (1, 1) band and those marked above correspond to the (0,0) band.

Fig. 4. Typical Boltzmann plots of the distribution of rotational energy in OH(ðm00 ¼ 0Þ) and OHðm00 ¼ 1Þ from photolysis of NCP.

700  50 K (600  40 K), and 1200  80 K (650  50 K), respectively. This disparity arises due to (1) preferential population of the K(A0 ) state, which is probed by P lines, and (2) the pronounced effect of K splitting at higher N (rotational quantum number). The average rotational energy of OH, ER, in ClNT at all the three excitations is estimated after weighted summing over the different rotational quantum number, in the (m00 ¼ 0Þ vibrational state. The resulting rotational energies are 1.7, 1.3 and 1.8 kcal/mol at 193, 248 and 266 nm, respectively. In case of NCP, the OH fragment is formed in both the ðm00 ¼ 0Þ and ðm00 ¼ 1Þ vibrational states. The population distributions of the OH fragments in (ðm00 ¼ 0Þ) and ðm00 ¼ 1Þ levels are characterized by rotational temperatures of 1570  90 K and 650  180 K, respectively. By weighted summing up all the rotational populations in the ðm00 ¼ 1Þ and ðm00 ¼ 0Þ vibrational states, the ratio of the populations of the two vibrational states can be estimated. In the present experiments, populations in a few rotational states could not be determined very accurately due to overlapping of lines. Within this experimental limitation, the average ratio of population in the ðm00 ¼ 1Þ to that in ðm00 ¼ 0Þ vibrational state was estimated to be 0.12  0.03 from three different sets of rotational spectra. In view of a small fraction of the ðm00 ¼ 1Þ state, the presence of higher vibrational states is expected to be negligible. From this ratio, the mean vibrational energy of the OH fragments is calculated to be 4.9  1.4 kcal/mol. On considering the rotational

126

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

excitation in both the ground and the first vibrational states, the rotational energy of the OH fragment is estimated to be 3.5  1.3 kcal/mol. Thus, the total internal energy of the OH fragment is 8.4  1.9 kcal/mol. 3.2. Translational energy of the fragments From the OH Doppler profiles, the average kinetic energy of the OH radical in the laboratory frame, ET(OH), can be determined. All the rotational lines measured were found to have similar line widths, within the range of experimental errors. More than 45 rotational line profiles for each photolysis wavelength for ClNT and NCP were evaluated to estimate the average kinetic energy of the OH fragment in the laboratory frame, ET(OH). The Doppler profiles of the rotational lines were corrected for the probe laser line width. The average translational energy in OH, ET(OH), is given by the Eq. (1)

Elab T ðOHÞ ¼

3 mOH hm2Z iOH ; 2

ð1Þ

where vz represents the laboratory velocity distribution of the OH radical in the direction of laser propagation. In the present case, the Doppler profile is a Gaussian function, as depicted in Figs. 5a and 5b for the P1(5) line of the OH product for ClNT and NCP respectively, which suggests that the translational energy follows the Maxwell–Boltzmann distribution. The width and the shape of the Doppler-broadened LIF line include contributions from the fragment molecular velocity, the thermal motion of the parent molecule and the finite probe laser line width. Thus, the actual Doppler width is calculated using a deconvolution procedure using the width (FWHM) of the laser spectral profile of the probe laser beam, which is obtained from the OH Doppler profile measured in a thermalized condition. The product translational energy in the laboratory frame ET(OH) is obtained using Eq. (1), and is found to be 7.2  1.1, 8.7  1.0 and 11.1  1.1 kcal/mol for ClNT at 266, 248 and 193 nm excitation, respectively. In the case of NCP photodissociation at 193 nm, the similar analysis results in ET(OH) of 19.6  3.3 kcal/ mol.

Fig. 5b. Doppler profile of P1(5) line of the A2 Rþ radical produced in dissociation of NCP at 193 nm.

X2 P (0, 0) system of the OH

and P2(N) lines are used to obtain the spin–orbit population, as they are well separated from satellite peak. Fig. 6(a) and (b) show the spin–orbit (P3=2 /P1=2 Þ ratios multiplied by appropriate statistical weights (2J + 1) plotted versus the OH rotational quantum number (N), at 266, 248 and 193 nm for ClNT and at 193 nm for NCP, respectively. From the figure, it is evident that for NCP at 193 nm and for ClNT, at all the photolysis wavelengths studied, the average ratio shows the deviation from the statistical value of unity and preference for the P3=2 state. This preference for the lower state can be partly explained by the energy difference between these two spin–orbit states. Each spin–orbit state of OH has two K-doublet components, denoted as Pþ (or A0 ) and P (or A00 ), depending on the orientation

4. Population of spin–orbit and K-doublet states The ground electronic state of OH being 2P, there are two spin– orbit components, 2P3=2 and 2P1=2 . The LIF intensities of only P1(N)

Fig. 5a. Doppler profiles of P1(5) line of the A2 Rþ X2 P (0, 0) system of the OH radical produced in dissociation of ClNT at 193, 248, and 266 nm.

Fig. 6. The statistically weighted spin–orbit ratios of nascent OH (ðm00 ¼ 0Þ) as a function of the rotational quantum number (N00 ). (a) The red, blue and black symbols denote the ratios at 193, 248 and 266 nm photolysis, respectively, for ClNT. (b) The red circles are for NCP. (For interpretation of the references to colour in this figure caption, the reader is referred to the web version of this article.)

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

127

Fig. 8. (a) Fluorescence spectra recorded at 200 ns after the laser excitation of NCP. (b) Variation of emission intensity at 540 (black) and 310 (red) nm with intensity of the photolyzing laser (193 nm). (For interpretation of the references to colour in this figure caption, the reader is referred to the web version of this article.)

Fig. 7. K-doublet ratio, P (A00 )/ Pþ (A0 ), of nascent OH(ðm00 ¼ 0Þ) as a function of the rotational quantum number (N00 ). (a) The red, blue and black symbols denote the ratios at 193, 248 and 266 nm photolysis, respectively, for ClNT. (b) The red circles are for NCP. (For interpretation of the references to colour in this figure caption, the reader is referred to the web version of this article.)

4.2. Theoretical calculations on ground state dissociation of ClNT and NCP

of the p-lobe of the unpaired electron on OH with respect to the plane of rotation. In the high J limit, in the Pþ (A0 ) state, the p lobe lies in the plane of rotation, while in the P (A00 ) state, the p-lobe is perpendicular to the plane of rotation. The relative populations of the K-doublets provide the information about the exit channel dynamics, during the breaking of a chemical bond. In Fig. 7(a) and (b), the K-doublet ratio is plotted against the rotational number, N00 for ClNT and NCP, respectively. From both the figures it is evident that there is a clear preference for the Pþ (A0 ) state. The K-doublet ratios P (A00 )/Pþ (A0 ), less than unity suggests that the pp electronic orbital for the unpaired electron generated upon dissociation lies in the plane or it is parallel to the plane of rotation of OH, in the photolysis of both ClNT and NCP.

Both ClNT and NCP do not contain OH moiety in their molecular structures, and hence the first step for OH elimination should be formation of an intermediate with the OH group. The involvement of an intermediate suggests that probably the reaction pathways occur from the ground electronic state. Molecular orbital (MO) calculations were carried out to generate the relative potential energy diagram for the dissociation channels of ClNT and NCP on its ground state. We optimized geometries of various molecules at B3LYP level using 6-31+G⁄ basis set. The relative energy is then calculated at the MP4(sdq) level using 6-311++G⁄⁄ basis set, using Gaussian suite of program [21]. All the stationary points were calculated on the ground state, for various dissociation channels. We located various transition states involved on the PES, and characterized the same with frequency calculations and intrinsic reaction coordinate (IRC) analyses. We have reported the calculated energetics for various possible OH channels, occurring on the ground PES of ClNT and NCP.

4.1. Emission studies

5. Discussion

In nitro-compounds, cleavage of the CAN bond on photodissociation generally results in formation of electronically excited NO2 as a primary photoproduct, which gives prompt emission. We have observed a strong UV fluorescence from the dissociation products of o-nitrotoluene [7], but we were unable to measure any emission from ClNT. However, we could measure an emission from the photolysis of NCP at 193 nm, as shown in Fig. 8. The emission spectrum, with maximum around 540 nm is attributed to the excited NO2 photoproduct. At higher laser intensities, an intense emission band at 310 nm (not shown in the figure), due to excited OH product, is also observed. This emission at 310 nm could not be observed at lower laser energies used for all other measurements. A log–log plot of emission signal intensity at 540 and 310 nm with pump laser intensity is shown in Fig. 8 as an inset. The slopes of these plots are 1.1  0.1 at 540 nm and 1.9  0.2 at 310 nm, indicating that the origin of the emission at 540 nm is from a monophotonic process, while at 310 nm is due to a biphotonic process.

5.1. Nature of excited states The gas phase UV absorption spectra of ClNT and NCP were measured. The spectrum of ClNT consists of two bands; one strong band with kmax around 210 nm, while the other weak band around 250 nm. In addition, the spectrum shows increasing absorption at the wavelengths lower than 200 nm. Comparing these spectra with the absorption spectra measured for o-nitrotoluene [22] we can assign the strong band to a p ! p⁄ transition and the weaker one to a n ! p⁄ excitation of a nonbonding electron of O [23]. The UV absorption spectra for NCP are similar to that for nitromethane [5,24,25], consisting of a strong p ! p⁄ transition with kmax around 200 nm and a weak band around 280 nm due to the n ! p⁄ transitions. In another work [26], the weaker band at 280 nm was assigned to the rC—N ! p⁄ along with the n ! p⁄ transition. For further understanding the nature of the transitions involved in ClNT at 266, 248 and 193 nm, and NCP at 193 nm, ab initio

128

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

Fig. 9. The figure shows three different pathways (A, B and C) of OH formation from ClNT with all the optimized structures of stable molecules, intermediates and transition states. A few important bond lengths (in angstrom) and dihedral angles (in degree) are marked on the structures. Details are given in the text.

molecular orbital (MO) calculations were performed in detail. We carried out the excited electronic state time-dependent density functional theory (TD-DFT) calculations, using aug-cc-pVDZ basis sets, to calculate the vertical excitation energies of the low-lying electronic states, and MOs were analyzed to assign each electronic transition. Although the calculated vertical transition energies slightly differ as compared to the experimental results, the nature of transitions and that of the orbitals involved are accurately predicted, using this method. The orbitals participating in the different electronic transitions were visualized, for better understanding of the process. The detailed studies for o-NT by Abbott et al. [27] and theoretical calculations by Lin et al. [28] using TD-DFT at the B3LYP/ 6-311(d; pÞ level attribute the absorption band at 193 nm to a mixture of S7 and S8 states and the band around 248 nm to the S4 state. Comparing this with our results for ClNT, we assign the band at 193 nm to a mixture of S7 to S10 states, which mainly correspond to the pring ! pring transition with some contribution from the pNO ! pNO , transition. Similarly, the excitations at 248 and 266 nm can be assigned to S4 and S3 states, respectively. The S4 state arises from the mixture of nCl, O ! p⁄ and pring ! p⁄ transitions, while the S3 state from pure nCl, O ! p⁄ transition. From the excited state calculations for NCP, we observed that the S1(n ! p⁄ transition, excitation  300 nm) and S2 (a mixture of rC—N ! p⁄ and n ! p⁄ transitions, excitation  280 nm) states have almost zero oscillator strength. The S3 state (excitation  188 nm), or probably a mixture of the S3 and S4(excitation 181 nm) states, is populated after excitation of NCP at 193 nm, and corresponds to the p ! p⁄ transition. 5.2. OH formation channel Most of the nitrocompounds undergo photodissociation on UV excitation to generate OH among various products. Since these compounds are devoid of any OH group, the formation of OH without any rearrangement is difficult. An intramolecular H-atom transfer to an oxygen atom of the NO2 group can introduce an OH group in the molecule. Subsequently, the NAOH bond dissociation produces OH. In aliphatic nitrocompounds (RCH2NO2), the presence of an a-hydrogen atom to the nitro group results in its isomerization to an aciform [RCH@N(O) OH], whereas that of a b-hydrogen atom leads to elimination of HONO. Subsequently,

the HOANO bond dissociation produces OH. Greenblatt et al. [29] have proposed a five-membered ring intermediate for formation of the OH photofragment from nitroalkanes with b-hydrogen atom. Zabarnick et al. [30] have proposed OH formation from the aciform for nitromethane, which is devoid of any b-hydrogen atom but has only a-hydrogen atoms. On photodissociation of nitroalkanes [6] and halogenated nitroalkane, 2-bromo-2-nitropropane [17], studied in our laboratory, and also other nitroalkanes [31] with b-hydrogen atom, the OH photofragment formation is favoured via HONO elimination step. A nitrocompound with both a- and b-hydrogen atoms can dissociate to produce OH, involving both the mechanisms via the aciform and HONO. In an aromatic nitrocompound, an appropriate alkyl substitution facilitates hydrogen atom migration; a ring hydrogen atom transfer is expected to be a high energy channel. Hence, the OH product, involving aromatic ring hydrogen atom, is not reported on UV excitation of nitrobenzene [28]. However, the presence of an alkyl substitution at an appropriate position facilitates hydrogen atom migration, leading to formation of its aci-nitro tautomer. Thus, only o-nitrotoluene, and not m- and p-nitrotoluene, undergoes intramolecular rearrangement to produce the aci-nitro tautomer, and finally OH [7,28]. All the possible pathways for the OH formation from UV excitation of ClNT and NCP are discussed. 5.2.1. From 2-chloro-6-nitrotoluene (ClNT) In the photoexcitation of ClNT at 266, 248, and 193 nm, the observed OH product has been explained with the help of MO calculations. The energized ClNT molecule is expected to undergo intramolecular hydrogen atom migration from the methyl group to an oxygen atom of the nitro group to produce its aci-nitro tautomer, which can produce OH by the NAOH bond scission. We carried out MO calculations to find out the energetics of the expected pathways of the OH formation. All the optimized structures of three different pathways A, B and C are depicted in Fig. 9, and corresponding calculated relative energies (in kcal/mol) are marked in Fig. 10. The intramolecular hydrogen atom migration in ClNT (pathway A) is expected to produce its aci-nitro tautomer with trans HONO group, involving a six-centred cyclic TS (shown as A1-TS1 in Fig. 9) with an activation barrier of 52.3 kcal/mol. But, the structure of the aci-nitro tautomer with trans HONO (A1-ST1, not-optimized structure) could not be optimized, it gives either a negative frequency or converts to the cis isomer. However, the

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

Fig. 10. A relative potential energy curve depicting three different pathways (A, B and C) of OH formation from ClNT. All energies are in kcal/mol. The lowest excitation energy of 107 kcal/mol, corresponding to 266 nm excitation, is marked in the figure.

structures of its cis isomer (A1-ST2), along with the TS (A1-TS2), corresponding to isomerization of the trans to cis isomer, could be optimized. This implies that the energy barrier for conversion of the trans form to the cis form is quite low. Both the isomers of the aci-nitro tautomer can undergo the NAOH bond scission to produce OH with its cofragment (shown as A1-ST3). Alternatively, the cis aci-nitro tautomer can undergo 1,3-hydrogen atom migration from one oxygen atom to the other of the nitro group (A2-ST1) involving the corresponding four-centred cyclic TS structure (A2-TS1) with an activation barrier of 21.8 kcal/mol. The rearranged isomer (A2-ST1) with cis-HONO can isomerize to a structure with trans-HONO (A2-ST2, not-optimized), involving the TS structure A2-TS2 with an activation barrier of 9.3 kcal/ mol. Since the structure with trans-HONO could not be optimized, it implies that its energy is very close to the energy of the

129

corresponding transition state structure (A2-TS2, not shown in Fig. 9). These molecules can undergo the NAOH bond cleavage to produce OH with its cofragment (A2-ST3), whose structure is not shown in Fig. 9 because it differs in only the NO orientation. Both structures (A1-ST3 and A2-ST3) have almost the same energy within 1 kcal/mol (Fig. 10). ClNT can also isomerize to its more stable nitrite form (B-ST1, 6.3 kcal/mol), with an energy barrier of 59.7 kcal/mol (pathway B). The corresponding TS structure is shown as B-TS1 in Fig. 9. The nitrite form can undergo a chain of reactions similar as in nitro form, involving the intramolecular H-atom migration, to produce OH. But this high energy pathway is not expected to compete with the earlier mentioned pathways A1 and A2 for OH formation, since the intramolecular H-atom transfer is difficult through nitrite (pathway B). However, nitrite is responsible for the NO product formation with an energy requirement of about 22 kcal/mol. Two isomers of the nitrite form (B-ST1 and B-ST2) with corresponding TS structure (B-TS2) are shown in Fig. 9. In the pathway A of OH formation, intramolecular H atom migration to one O atom of the NO2 group involves the H atom of the CH3 group. In the pathway C, we have shown the H atom migration from the benzene ring, which is expected to be a higher energy pathway. The ring H atom migration produces HONO with its cofragment (C-ST1), involving a five-centred cyclic TS (C-TS1) having an activation barrier of 86.4 kcal/mol. Finally, HONO can undergo the OAN bond scission to produce OH (C-ST2). This pathway of OH formation is not possible with excitation of ClNT at 266 nm because of an energy constraint. Although excitation of ClNT at 248 and 193 nm can provide sufficient energy for this channel to be energetically possible, it cannot compete with lower energy pathways A and B. Thus, the pathway A, involving H atom migration from CH3 to an O atom of NO2 group, is the most favoured lowest energy pathway for the OH radical formation on UV excitation of ClNT. It implies that although ClNT has hydrogen atoms both in the ring (b-hydrogen) and side chain (c-hydrogen), the lowest energy pathway of OH formation involves the latter hydrogen. In addition to the OH channel, ClNT can have other dissociation pathways, such as a chlorine channel. Since we have focused on the

Fig. 11. The figure shows three different pathways (A, B and C) of OH formation from NCP with all the optimized structures of stable molecules, intermediates and transition states. A few important bond lengths (in angstrom) and dihedral angles (in degree) are marked on the structures. Details are given in the text.

130

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

cofragment (B2-ST2). The b-hydrogen atom migration to an oxygen atom of the NO2 group of the nitro form of NCP facilitates another OH pathway C. This pathway also leads to the formation of HONO and cyclopentene (B1-ST2), involving a five-centred cyclic TS (C-TS1) with an activation barrier of 45.5 kcal/mol. Finally, the HONO product can undergo the NAOH bond scission to produce OH (B1-ST3). Among all these energetically feasible OH pathways on excitation of NCP at 193 nm, the pathway C, involving the concerted HONO elimination from the nitro form of NCP, is the lowest energy pathway. Although NCP has both the a- and b-hydrogen atoms, OH formation via HONO elimination (involving b-hydrogen atom) is the lowest energy pathway. Thus, OH formation from NCP is mainly due to the pathway C, whereas that from ClNT is primarily due to the pathway A, involving H atom migration from the CH3 group to the NO2 group. 5.3. Partitioning of the available energy The partitioning of the available energy in different degrees of freedom of the products formed upon excitation of ClNT at 266, 248 and 193 nm and NCP at 193 nm was measured employing LIF. The energy pumped in molecules at 266, 248 and 193 nm is 107, 115 and 148 kcal/mol, respectively. The available energy, Eavl, is partitioned among various degrees of freedom of the photofragments with Fig. 12. A relative potential energy curve depicting three different pathways (A, B and C) of the OH formation from NCP. All energies are in kcal/mol.

OH channel, we did not search for the chlorine channel. But our energy calculations suggest that on UV excitation of ClNT, the chlorine channel cannot compete with the observed primary pathways. However, the halogen atom channel has been observed as a primary channel in UV photodissociation of aryl halides [32]. 5.2.2. From nitrocyclopentane (NCP) The NCP molecule is not an aromatic nitrocompound, and hence it is expected to behave more like aliphatic nitrocompounds with respect to the mechanism of OH formation on UV excitation at 193 nm. The molecule has both a- and b-hydrogen atoms with respect to the NO2 group. Thus, NCP is expected to produce OH both from its aci-nitro tautomer and the HONO product, after a- and b-hydrogen atom migration, respectively. Energetics of various OH pathways have been calculated, and optimized structures and energies are shown in Figs. 11 and 12, respectively. In NCP dissociation also, the a-hydrogen atom migration to an oxygen atom of the NO2 group produces its aci-nitro tautomer (A1-ST1) through a four-centred cyclic TS (A1-TS1) with an activation barrier of 69.4 kcal/mol. Like in ClNT, the structure of the expected trans aci-nitro tautomer (A1-ST1, non-optimized) in NCP could not be optimized. However, it is expected to be isomerized to its more stable cis aci-nitro tautomer (A1-ST2), involving the TS with structure shown as A1-TS2. The NAOH bond cleavage of the aci-form produces OH with its cofragment (A1-ST3). In addition, A1-ST1 can eliminate HONO with its cofragment (A2-ST1), and the HONO dissociation can produce OH (A2-ST2). NCP can also isomerize to its almost equally stable nitrite form (B1-ST1) involving a three-centred cyclic TS (B1-TS1) with an activation barrier of 61.7 kcal/mol. One b-hydrogen atom can migrate to the oxygen atom attached to the ring, producing HONO and cyclopentene (B1-ST2) involving a four-centred TS (B1-TS2) with an activation barrier of 60.0 kcal/mol. Finally, HONO can dissociate to produce OH (B1-ST3). Similarly, B1-ST2 can react to produce a bicyclic compound (B2-ST1), involving a three-centred TS (B2TS1) with an activation barrier of 45.1 kcal/mol. Finally, B2-ST1 can undergo the NAOH bond cleavage to produce OH with its

Eavl ¼ Eint ðOHÞ þ Eint ðcofragmentÞ þ ET ðOH þ cofragmentÞ;

ð2Þ

where Eint consists of both the rotational and vibrational excitations and ET is energy partitioned in translational energy of photofragment. On photodissociation of ClNT the energy partitioned in rotational degree of freedom at 266, 248 and 193 nm is 1.8, 1.3, and 1.7 kcal/mol, respectively. We did not observe any vibrationally excited OH photoproduct, in photodissociation of ClNT. However in our previous work on dissociation of o-nitrotoluene [7] at 248 and 193 nm, the OH photofragment was observed in both the vibrationally ground state, ðm00 ¼ 0Þ, and the excited state, ðm00 ¼ 1Þ. The translational energy imparted into the OH fragments at 266, 248 and 193 nm is found to be 7.2  1.1, 8.7  1.0 and 11.1  1.1 kcal/ mol, which shows an increasing trend with increasing excitation energy. However, the difference in translational energy of OH produced on photodissociation at 266, 248 and 193 nm is only about 4 kcal/mol (maximum), though the difference in the photon energies is 8 kcal/mol between 266 and 248 nm and 33 kcal/mol between 248 and 193 nm. The measured translation energy, almost independent of the excitation energy, suggests that the pathway of OH formation has an exit barrier. Unlike the photodissociation of ClNT, that of NCP at 193 nm produces OH in both the vibrationally ground state, ðm00 ¼ 0Þ, and the excited state, ðm00 ¼ 1Þ. The total internal energy of the OH fragment is 8.4  1.9 kcal/mol, as discussed in Section 3.1. The average translational energy of OH from NCP is estimated from the Doppler profile to be 19.6  3.3 kcal/mol. Comparing the partitioning of the available energy into various modes of NCP and ClNT after photolysis, it is observed that the fraction of energy partitioned into internal modes of the OH fragment from NCP is higher than in case of ClNT. Also, the energy partitioned into translational mode of OH is higher in case of NCP. The observed energy in OH from NCP being greater than that from ClNT can be explained based on following factors. (1) Our theoretical calculations predict that the energy required for OH formation in NCP is lower than that in ClNT, implying a greater available energy with the co-fragments of NCP than that of ClNT. (2) The number of atoms in co-fragment produced after OH formation from ClNT is more than that from NCP. Therefore, a large amount of energy will be taken by the co-fragment from ClNT. Also, the presence of heavy chlorine atom will enhance the effective statistical redistribution of energy. These will lead to

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

lower energy in the OH fragment from ClNT. This explains the observed higher fraction of energy in internal energy of the OH fragment from NCP than from ClNT. Thus, both the experimental and theoretical results suggest that the dynamics of the OH formation in ClNT differs from that in NCP. It seems that the presence of Cl atom in the molecule efficiently redistributes the local excitation in the excited state of the molecule, and reduces the internal energies in NO2 and OH products. Therefore, the presence of a Cl atom has a substantial influence on the excited state and exit channel dynamics of ClNT. The present results of photodissociation of ClNT at 266, 248 and 193 nm show that the rotational population distributions of the nascent OH fragments have not thermally equilibrated, as indicated by the difference in rotational temperatures calculated from Boltzmann’s plots for P and Q lines (Fig. 3). Similar dynamics of OH formation at three different wavelengths is attributed to involvement of a single mechanism in all three cases, with non-equilibrium distribution in the rotational states, a consequence of preferential population of Pþ ðA0 Þ. As orbital rotational interaction is more pronounced at higher N, the splitting is more pronounced at higher N and this leads to higher temperature of the preferred K-doublet state. The low energy in the internal modes of OH fragment formed on ClNT photolysis is very similar to that of some carboxylic acids [33]. This indicates the presence of a considerable barrier in the exit channels for OH formation. The presence of an exit barrier is supported by the fact that the energy partitioned into translational degree of freedom is almost equal, irrespective of excitation energies. The relative population of the K-doublets of OH, which provides the exit channel dynamics in the bond cleavage process, is similar at all the three wavelengths studied. This further supports that at all the wavelengths, the exit channel dynamics of OH formation is similar, and confirms that OH is formed with the same exit barrier. Therefore, we can conclude that at all the above wavelengths, the initially excited states cross over either to another excited dissociative state or the ground electronic state, from which subsequently the OH elimination takes place. As the OH formation proceeds via an intermediate, it is most likely to be taking place through the ground state. Our MO calculations support the OH formation path from the ground electronic state, involving mainly pathway C for NCP and pathway A for ClNT. Although the dissociation of these molecules is from the ground electronic state, the relative translational energy partitioned to the OH channel is somewhat greater than the statistical dissociation, because of the presence of an exit barrier to this channel. 5.4. UV/visible emission The UV excitation of a nitrocompound generally produces UV/ visible emission from electronically excited NO2 product. Mijoule et al. [34] have carried out detailed theoretical calculations for the nitromethane dissociation in its first excited state and predicted that NO2 molecules are produced, both in electronically excited and ground states. It was also observed by them that at shorter wavelengths, formation of NO2 molecules in the excited state increases. They have assigned the emissive state as 2B1 or 2 B2 excited state. A detailed excited state study for nitromethane carried out by Wade et al. [25] suggests that at 193 nm, NO2 formation does not occur on the ground electronic surface but rather on an excited electronic state, which is either purely repulsive or predissociative in nature. Further, it was suggested that the electronically excited state has a decay pathway to the ground state via a conical intersection. On UV excitation of ClNT at different wavelengths, we did not observe emission from the NO2 product. However, the emission was observed on UV excitation of o-NT at 193 nm [7]. This implies that the NO2 product in ClNT is produced mostly in the ground electronic state. Non-observation of the

131

excited state NO2 product from ClNT can be attributed to the presence of a chlorine atom in the molecule. Analogous to the influence of a heavy Cl atom on the OH formation dynamics, where a heavy atom enhances the non-radiative processes in an excited state, a similar effect is operating in the NO2 formation dynamics. Moreover, the cofragment of NO2 in ClNT is expected to have relatively greater internal energy than that in o-NT because of more number of atoms in the former. Thus, the internal energy in NO2 from ClNT is expected to be lower than that from o-NT. However, unlike in ClNT, emission could be observed on UV photodissociation of NCP at 193 nm. The cofragment of NO2 in NCP is expected to have relatively smaller internal energy than that in ClNT, implying relatively greater internal energy in NO2. In addition, the CANO2 bond dissociation energy is smaller by about 20 kcal/mol in NCP (56.2 kcal/mol) than that in ClNT (75.7 kcal/mol), leading to a higher available energy in the former. The emission spectra obtained on photoexcitation of NCP at 193 nm are similar to that obtained for previously studied nitroalkanes [6,17], with the maximum at around 540 nm. Similarity of spectra for aliphatic and cyclic nitroalkane indicates that there is not much intramolecular transfer of energy from the excited NO2 group to the alkyl group in ring during the CAN bond dissociation. This indicates that for cyclic nitroalkane also the NO2 moiety in the excited state is completely isolated and the molecule dissociates from the excited electronic state or the ground state after very fast radiationless cross over, through the conical intersections [25,35]. Even in an aromatic nitrocompound, for example nitrobenzene, vibrational energy is not transferred from the NO2 group, rather it is transferred from phenyl to the nitro group [36]. However, in o-nitrotoluene, vibrational energy is transferred from the NO2 group to the phenyl ring. This energy transfer (including other factors mentioned above) may be responsible for not observing UV–vis emission from the NO2 product in the case of ClNT. The effects of substituents in NT on energetics of dissociation are demonstrated theoretically [2]. One hydrogen atom migration from the CH3 group to the oxygen atom of the NO2 group is not much affected by a substituent in nitrotoluene. This reveals why the mechanism and energetics of OH formation on excitation of nitrotoluene [37] and ClNT at 193 nm remain similar. However, the CANO2 bond dissociation energy decreases with electron withdrawing group (NO2, CO2H) and increases with electron donating groups (NH2, OH). In addition to the nature, the position of a substituent also plays an important role in determining the dissociation energy. The substituent Cl being a mild electron withdrawing group (inductive effect), does not affect significantly the CANO2 bond dissociation energy in ClNT (75.7 kcal/mol) with respect to that in nitrotoluene (76.3 kcal/mol) [28]. Similarly, the CH3 group being mild electron donating group, the CANO2 bond dissociation energy of nitrobenzene (74.1 kcal/mol) [28] is not much affected in nitrotoluene. Although the CANO2 bond dissociation energy is similar, the dynamics appears to be different; NO2 is produced in an electronically excited state from NT, but in the ground state from ClNT.

6. Conclusions In summary, photodissociation of ClNT at 266, 248, and 193 nm and NCP at 193 nm generates the OH radical, as detected by laser induced fluorescence. We have measured the state distribution of the nascent OHðm00 ; J 00 Þ product, and found that it is produced mostly in the vibrational ground state ðm00 ¼ 0Þ in ClNT at all the excitation wavelengths and in both ðm00 ¼ 0Þ and ðm00 ¼ 1Þ states in NCP at 193 nm. The rotational population of ClNT is characterized by rotational energies of 1.7, 1.3 and 1.8 kcal/mol at 193, 248 and 266 nm, respectively. In case of NCP, the population distribu-

132

M.N. Kawade et al. / Chemical Physics 443 (2014) 123–132

tions of the OH fragment in (m00 ¼ 0Þ and ðm00 ¼ 1Þ levels are characterized by rotational temperatures of 1570  90 K and 650  180 K, respectively. The spin–orbit components of OH are not in equilibrium for NCP at 193 nm, and at all the photolysis wavelengths for ClNT; the average ratio P3=2 /P1=2 , seems to be more than one. The population distribution in the K-doublets has a preference for the Pþ (A0 ) state. The average translational energy partitioned into the OH photofragment for ClNT is found to be 7.2  1.1, 8.7  1.0 and 11.1  1.1 kcal/mol at 266, 248 and 193 nm excitation, respectively, while for NCP it is found to be 19.6  3.3 kcal/mol. From the experimental results for ClNT molecule, we have deduced that although photoexcitation energies at 193, 248 and 266 nm are quite different, the mechanism of OH formation is similar. It is mainly because the dissociation of ClNT at all three wavelengths takes place from the ground electronic state with an exit barrier. Based on our theoretical calculations of different paths of OH formation, it is concluded that the OH radical is formed mainly involving an aci-form from ClNT (pathway A), and via HONO elimination from NCP (pathway C). Emission could be detected from the NO2 product on UV excitation of NCP, but not from that of ClNT. Thus, NO2 is produced from an electronically excited state of NCP, and OH from the ground states of both NCP and ClNT. On comparing photodissociation dynamics of ClNT with o-nitrotoluene, it seems that the presence of Cl atom has a substantial influence on both the excited state and exit channel dynamics. Conflict of interest There is no conflict of interest. Acknowledgments We acknowledge University of Pune, for initial theoretical calculations. MNK sincerely acknowledges Homi Bhabha National Institute, Mumbai. References [1] [2] [3] [4] [5]

H. Diez-y-Riega, H. Eilers, Appl. Phys. B 108 (2012) 189. G. Fayet, L. Joubert, P. Rotureau, C. Adamo, J. Phys. Chem. A 113 (2009) 13621. A. Bhattacharya, Y. Guo, E.R. Bernstein, J. Chem. Phys. 136 (2012) 024321. Y.Q. Guo, A. Bhattacharya, E.R. Bernstein, J. Phys. Chem. A 113 (2009) 85. L.J. Butler, D. Krajnovich, Y.T. Lee, G. Ondrey, R. Bersohn, J. Chem. Phys. 79 (1983) 1708. [6] S. Sengupta, Y. Indulkar, A. Kumar, S. Dhanya, P.D. Naik, P.N. Bajaj, J. Phys. Chem. A 112 (2008) 12572.

[7] S. SenGupta, H.P. Upadhyaya, A. Kumar, S. Dhanya, P.D. Naik, P.N. Bajaj, Chem. Phy. Lett. 452 (2008) 239. [8] I. Bejan, Y.A.E. Aal, I. Barnes, T. Benter, B. Bohn, P. Wiesen, J. Kleffmann, Phys. Chem. Chem. Phys. 8 (2006) 2028. [9] Q. Wei, H.-M. Yin, J.-L. Sun, X.-F. Yue, K.-L. Han, Chem. Phys. Lett. 463 (2008) 340. [10] S.-B. Cheng, C.-H. Zhou, H.-M. Yin, J.-L. Sun, K.-L. Han, J. Chem. Phys. 130 (2009) 234311. [11] R.S. Booth, C.-S. Lam, M.D. Brynteson, L. Wang, L.J. Butler, J. Phys. Chem. A 117 (2013) 9531. [12] S.-B. Cheng, C.-H. Zhou, H.-M. Yin, J.-L. Sun, K.-L. Han, ChemPhysChem 10 (2009) 1135. [13] X.-F. Yue, J.-L. Sun, H.-M. Yin, Q. Wei, K.-L. Han, J. Phys. Chem. A 113 (2009) 3303. [14] C.-H. Zhou, S.-B. Cheng, J.-L. Sun, H.-M. Yin, K.-L. Han, G.-Z. He, J. Phys. Chem. A 113 (2009) 4923. [15] M.L. Hause, N. Herath, R. Zhu, M.C. Lin, A.G. Suits, Nature Chem. 3 (2011) 932. [16] Z. Homayoon, J.M. Bowman, A. Dey, C. Abeysekera, R. Fernando, A.G. Suits, Z. Phys. Chem. 227 (2013) 1267. [17] A. Saha, M. Kawade, H.P. Upadhyaya, A. Kumar, P.D. Naik, J. Chem. Phys. 134 (2011) 044316. [18] M.N. Kawade, A. Saha, H.P. Upadhyaya, A. Kumar, P.D. Naik, J. Phys. Chem. A 114 (2010) 12369. [19] M.N. Kawade, A. Saha, H.P. Upadhyaya, A. Kumar, P.D. Naik, J. Phys. Chem. A 117 (2013) 2415. [20] I.L. Chidsey, D.R. Crosley, J. Quant. Spectrosc. Radiat. Transfer 23 (1979) 187. [21] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, J.A. Montgomery, T.V. Jr., K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, A.L.G. Liu, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03. Gaussian Inc, Pittsburgh, PA, 2003. [22] K.J. Castle, J.E. Abbott, X. Peng, W. Kong, J. Phy. Chem. A 104 (2000) 10419. [23] N.S. Bayliss, E.G. McRae, J. Phy. Chem. 58 (1954) 1006. [24] S. Nagakura, Mol. Phys. 3 (1959) 152. [25] E.A. Wade, K.E. Reak, S.L. Li, S.M. Clegg, P. Zou, D.L. Osborn, J. Phy. Chem. A 110 (2006) 4405. [26] J.F. Arenas, J.C. Otero, D. Pelaez, J. Soto, J. Chem. Phys. 119 (2003) 7814. [27] J.E. Abbott, X. Peng, W. Kong, J. Chem. Phys. 117 (2002) 8670. [28] M.F. Lin, Y.T. Lee, C.-K. Ni, S. Xu, M.C. Lin, J. Chem. Phys. 126 (2007) 64310. [29] G.D. Greenblatt, H. Zuckermann, Y. Haas, Chem. Phys. Lett. 134 (1987) 593. [30] S. Zabarnick, J.W. Fleming, A.P. Baronavski, J. Chem. Phys. 85 (1986) 3395. [31] G. Radhakrishnan, T. Parr, C. Wittig, Chem. Phys. Lett. 111 (1984) 25. [32] K.-L. Han, G.-Z. He, J. Photochem. Photobiol. C: Photochem. Rev. 8 (2007) 55. [33] P.D. Naik, H.P. Upadhyaya, A. Kumar, A.V. Sapre, J.P. Mittal, J. Photochem. Photobiol. C: Photochem. Rev. 3 (2003) 165. [34] C. Mijoule, S. Odiot, S. Fliszar, J.M. Schnur, J. Mol. Struct.: Theochem 149 (1987) 311. [35] J.F. Arenas, J.C. Otero, D. Peláez, J. Soto, J. Chem. Phys. 122 (2005) 084324. [36] B.C. Pein, D.D. Dlott, J. Phys. Chem. A 118 (2014) 965. [37] S.C. Chen, S.C. Xu, E. Diau, M.C. Lin, J. Phys. Chem. A 110 (2006) 10130.