Article pubs.acs.org/JPCA
Resolved (v1, v2 = 1) Combination Vibrational States of CF3 Fragments in the Photofragment Translational Spectra of CF3I Dan Lin,†,§ Lili Hu,‡ Sheng Liu,†,§ Wenke Qi,†,§ Min Cheng,*,† Yikui Du,† and Qihe Zhu† †
Beijing National Laboratory of Molecular Sciences, State Key Laboratory of Molecular Reaction Dynamics, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China ‡ Institute of Physics, Chinese Academy of Sciences, Beijing 100190, China § University of Chinese Academy of Sciences, Beijing 100049, China ABSTRACT: The photodissociation of CF3I → CF3(v1,v2) + I*/I has been investigated at 248, 266, and 277 nm with our high resolution mini-TOF photofragment translational spectrometer. Based on the theoretical calculations of Clary and of Bowman et al., now in this manuscript, we assign 701 cm−1 to the CF symmetric stretch (breathing) ν1 mode, and 1086 cm−1 to the umbrella ν2 mode of the CF3 fragment. In the obtained TOF spectra of I+ from the I* channel, situated in the 701 cm−1 gaps between the original series of (v1, 0) vibrational peaks, a new series of weaker (v1, 1) vibrational peaks are partially resolved. These observed new peaks with 1086 cm−1 ν2 mode excitation have never been reported in previous literature. In the TOF spectra of I+ from the I channel, the new series of (v1, 1) peaks are also partially resolved. However, these spectra of I channel are less satisfactory, because for higher Eavl and higher ET, the higher resolution of PTS is required. The potential energy at the curve crossing point and the excitation of CF3 (v1, 2) and (v1, 3) vibrational states have been also analyzed. Van Veen et al.7 considered that, in their experiments on CF3I photodissociation at 248 nm, only the 701 cm−1 (ν1) mode of CF3 fragments had been excited, and they assigned the 701 cm−1 (ν1) to the umbrella mode. In the I* channel, the distribution was peaking around v1 = 6, higher than their theoretical result of peaking at v1 = 4 or 5. Many groups have performed on the CF3I photodissociation at different wavelengths.8−17,20−23 For the high resolution experiments with vibrational peaks resolved, Felder8 and Wang et al.9 studied the photodissociation of CF3I at 248 nm, and got the 701 cm−1 (ν1) states resolved TOF spectrum in the I* channel with the highest peak at v1 = 4 and 5, respectively. With the velocity map ion imaging technique, Pratt’s group got 701 cm−1 (ν1) state resolved images on the CF3I photodissociation between 304 and 277 nm10 and also at 266 nm.20 At 304 nm, Yu et al.11 got PTS of CF3I with 701 cm−1 (ν1) states of the CF3 fragment resolved. From the previous experimental reports on the photodissociation of CF3I, only the ν1 vibration (701 cm−1) of CF3 fragments have been observed to be excited, which is not in agreement with the theoretical prediction.3 Van Veen7 and Clary3 suggested the combinational excitation of ν1 and ν2 modes of CF3 during the photodissociation, due to the final state vibrational interaction on the excited potential energy surfaces. Using our high-resolution mini-TOF photofragment translational spectrometer, we intend to resolve and study the
1. INTRODUCTION As a typical simple polyatomic molecule, CF3I has attracted particular interest in the photodissociation dynamics for decades. In the A absorption band (220−350 nm) of the CF3I molecule, there are three transition states: one parallel transition state 3Q0 correlates adiabatically with the I*(2P1/2) channel, and two perpendicular transition states 1Q1 and 3Q1 correlate adiabatically with the I(2P3/2) channel.1,2 There is a conical crossing (c.c.) between the 3Q0 and 1Q1 surfaces, and the nonadiabatic coupling affects both the photodissociation branching ratios and the vibrational distributions of the photofragments.3−5 Previously, in the photodissociation of CF3I, only the 701 cm−1 vibrational states6 of the CF3 radical had been resolved in the PTS, and the 701 cm−1 vibrational mode was generally assigned to the ν2 umbrella vibration.7−17 While in the closecoupling calculation, Clary3 showed that both 701 and 1086 cm−1 modes6,18 contain two vibration characters, but the 701 cm−1 mode has more CF symmetric stretch character and the 1086 cm−1 mode has more umbrella character. He also suggested the preferential (v1, v2) combination states for the CF3 fragment. Recently Bowman et al.19 in their specific theoretical study, also showed that the 701 cm−1 mode corresponds mostly to the CF symmetric stretch (breathing) character and the 1086 cm−1 mode mostly to the umbrella character. In this manuscript, we follow the new assignment by Bowman et al. that the 701 cm−1 is the ν1 symmetric stretch (breathing) vibration and the 1086 cm−1 is the ν2 umbrella mode of the CF3 fragment. © XXXX American Chemical Society
Received: July 13, 2016 Revised: November 11, 2016 Published: November 15, 2016 A
DOI: 10.1021/acs.jpca.6b06988 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A excitation of both ν1 (701 cm−1) and ν2 (1086 cm−1) vibrational modes of CF3 fragments in the photodissociation of CF3I.
2. EXPERIMENTAL SET-UP The experimental apparatus has been described previously.24,25 A pulsed molecular beam valve (General Valve, 0.5 mm diameter orifice) is operated at 10 Hz with a pulsed width of about 260 μs. The sample CF3I (99% Aldrich) is seeded 10% in Kr gas with a stagnation pressure of 1 atm. The use of the heavy carrier gas Kr can lower the velocity and the velocity spread of the molecular beam in order to decrease the discrepancy of the experiments. The gas mixture from the pulsed valve expands supersonically into the source chamber. The expanded gas pulse is collimated by a skimmer (d = 0.5 mm, 20 mm from the orifice of the pulsed valve) to form a fine molecular beam in the reaction chamber. With the pulsed valve “on”, the pressure of the source chamber increases to 1.4 × 10−3 Pa, and the pressure of the reaction chamber can keep at about 1.5 × 10−5 Pa. The UV laser beams are frequency-doubled from two dye lasers (Sirah), pumped by the same Nd: YAG laser (SpectraPhysics, Pro. 230) at a repetition rate of 10 Hz. When the 248.00 nm laser is used for the photodissociation of CF3I, the 304.02 nm laser is used for the (2 + 1) REMPI detection of I* fragments. The polarization of the photodissociation laser is parallel to the detection axis, and the polarization of the ionization laser is perpendicular to the detection axis. The ionization laser crosses the photodissociation region 5 ns later than the photodissociation laser. At 266 and 277 nm, the photodissociation of CF3I and the state-selected ionization of iodine are achieved by a single laser beam with parallel polarization. Lasers of 266.60 and 277.38 nm are used to selectively ionize I*, while 266.51 and 277.87 nm are for the I fragment. The fragment ions I+ are accelerated in the weak electric field region (E ≈ 1 V/cm, l1 ≈ 11 mm) and then fly in the field-free region (L2 ≈ 30 mm). The total flight path is less than 50 mm. The ion signals are detected by a dual microchannel plate (MCP) and recorded by a multichannel scaler (P7888−1E).
Figure 1. TOF spectrum of I+ from CF3I → CF3 + I* at 248.00 nm (α = 0° for parallel transition) and 304.02 nm for I* REMPI.
now, between the original series of (v1, 0) stronger fine peaks, there are a new series of weaker fine peaks partially resolved. The TOF spectrum in Figure 2a is transformed into the photofragment translational spectrum (PTS) using the method described in our previous paper.26 The transformed PTS is shown in Figure 2b. From the conservation of energy, for the I* channel we have hv = D0(F3C−I) + Eso + E T + E int(CF3)
where hv is the photon energy (here the internal energy of the parent molecule CF3I has been neglected due to the strong cooling during supersonic expansion), D0(F3C−I) = 224.68 kJ/ mol is the C−I bond energy of CF3I,27 Eso = 0.943 eV is the spin−orbit splitting energy of the iodine atom, ET is the total translational energy of CF3 and I* fragments, and Eint(CF3) is the internal energy of the CF3 fragment. Then from the above equation and the ET of the fine peaks in Figure 2b, the Eint of CF3 corresponding to the fine peaks can be calculated. As the fragment CF3 is also in C3v symmetry, its rotation from dissociation can be neglected. Then Eint ≈ EV is obtained, and the appropriate vibrational state of the CF3 fragment can be assigned.24,28 In the TOF spectrum and the PTS of CF3I photodissociation at 248.00 nm as shown in Figure 2a,b, the original series of eight higher fine peaks are assigned to (v1 = 0, 1, 2, 3, 4, 5, 6, 7, v2 = 0) states of the CF3 fragment, with the top peak at (4, 0). The obtained average vibrational energy spacing ΔEV ≈ 700 cm−1 agrees well with the symmetric stretch ν1 mode of the CF3 radical. The new series of weaker fine peaks are situated near the gap center between the original (v1, 0) peaks, agreeable to ν2 − ν1 = 1086−701 = 385 cm−1. Therefore, these new peaks are assigned to (v1 = 0, 1, 2, 3, 4, 5, v2 = 1) states of the CF3 fragment. The profile of the whole spectrum reflects the Franck−Condon effect during X → 3Q0 excitation. 3.1.2. The Photodissociation of CF3I in the I* Channel at 266.60 nm. The experimental TOF spectrum of I+ from CF3I → CF3(v1, v2) + I* at 266.60 nm is shown in Figure 3a. The transformed PTS is shown in Figure 3b. There are many fine peaks resolved. Between the resolved six strong fine peaks (v1 = 0, 1, 2, 3, 4, 5, v2 = 0), there are four extra weak fine peaks (v1 = 0, 1, 2, 3, v2 = 1) partially resolved in the TOF spectrum and the PTS, as shown in Figure 3. The highest peak is assigned to the state (2, 0) of the CF3 fragment. In early photofragment vibrational distribution study of CF3I photodissociation at 266 nm, Aguirre and Pratt20 detected the
3. RESULTS AND DISCUSSION 3.1. The Photodissociation of CF3I in the I* Channel. 3.1.1. The Photodissociation of CF3I in the I* Channel at 248.00 nm. The whole TOF spectrum of I+ ions from CF3I → CF3 (v1, v2) + I* at 248.00 nm is shown in Figure 1. The I* atoms are ionized to I+ by the REMPI of 304.02 nm laser. There are four broad peaks in the whole spectrum. The first broad peak A corresponds to the forward I* fragments in the Newton spheres of CF3I photodissociation at 248 nm. The fourth broad peak D corresponds to the backward I* fragments in the Newton spheres. The I+ ions from backward I* in peak D are deaccelerated by the weak electric field, then turn-around and are accelerated in the forward direction. (The second B and the third C broad peaks correspond to the forward I* and backward I* in the smaller Newton spheres of CF3I photodissociation at 304.02 nm.) In the first broad peak A, corresponding to the forward I* from 248 nm, there are many fine peaks resolved. These fine peaks are very important and are enlarged as shown in Figure 2a. The profile of the broad peak A is quite similar to the TOF spectra in the papers of Felder8 and Wang et al.9 However, B
DOI: 10.1021/acs.jpca.6b06988 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
Figure 2. (a) Enlarged TOF spectrum of I+ from CF3I → CF3 (v1, v2) + I* at 248.00 nm, and (b) transformed PTS. (The open circles are the experimental data, the solid and dash lines in panel b are the simulated spectra.)
Figure 3. (a) TOF spectrum of I+ from CF3I → CF3 (v1, v2) + I* at 266.60 nm, and (b) transformed PTS.
Figure 4. (a) TOF spectrum of I+ from CF3I → CF3 (v1, v2) + I* at 277.38 nm, and (b) transformed PTS.
extra weaker fine peaks (v1 = 0, 1, 2, v2 = 1) partially resolved, as presented in Figure 4. The highest peak is assigned to the state (1, 0) of the CF3 fragment. 3.2. The Photodissociation of CF3I in the I Channel. To study the combinational excitation of CF3 fragments during the photodissociation of CF3I in the I channel, we performed the experiments with a single laser of 266.51 and 277.87 nm. Because of the higher ET, higher resolution is required. The quality of the TOF spectra in the I channel is less satisfactory,
CF3 and I fragments in the I* and I channels by using singlephoton ionization. In the I* channel, their total kinetic energy distribution extracted from the I+ image resolved six (v1, 0) states, which is quite consistent with our PTS. 3.1.3. The Photodissociation of CF3I in the I* Channel at 277.38 nm. The TOF and PTS of CF3I → CF3(v1, v2) + I* at 277.38 nm are shown in Figure 4a,b. Compared to 248.00 and 266.60 nm, the lower available energy at this wavelength results in the less excitation of internal energy. Between the resolved five fine peaks (v1 = 0, 1, 2, 3, 4, v2 = 0), there are only three C
DOI: 10.1021/acs.jpca.6b06988 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
Figure 5. (a) TOF spectrum of I+ from CF3I → CF3 (v1, v2) + I at 266.51 nm, and (b) transformed PTS.
Figure 6. (a) TOF spectrum of I+ from CF3I → CF3 (v1, v2) + I at 277.87 nm, and (b) transformed PTS.
is highly internally excited with the highest fine peak of (3, 0). In the present study, the excitation of internal energy in the I* and I channel at 277 nm agrees with the results of Furlan et al.17 and Pratt.10 As seen in Figure 6a,b, between the resolved eight fine peaks (v1 = 0, 1, 2, 3, 4, 5, 6, 7, v2 = 0), there are six extra weak fine peaks (v1 = 0, 1, 2, 3, 4, 5, v2 = 1) partially resolved. The recoil anisotropy is determined to be β(I*) = 1.86 and β(I) = 0.98 in this study. The contribution of the parallel
but the combinational vibrational excitation of CF3 is clearly shown in the spectra. 3.2.1. The Photodissociation of CF3I in the I Channel at 266.51 nm. The TOF spectrum of I+ from CF3I → CF3(v1, v2) + I at 266.51 nm is shown in Figure 5a. Because of 0.943 eV more available energy, the average vibrational energy of CF3 fragments for the I channel with the highest fine peak (5, 0) is higher than that with (2, 0) for the I* channel at 266.60 nm. As seen in Figure 5a,b, between the resolved 10 stronger fine peaks (v1 = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, v2 = 0), there are eight extra weaker fine peaks (v1 = 0, 1, 2, 3, 4, 5, 6, 7, v2 = 1) partially resolved. At 266 nm, the measured anisotropy parameter is β(I*) = 1.86 for the I* channel, and is β(I) = 1.03 for the I channel. The β(I*) = 1.86 is very close to the limiting value (β∥ = 2) for a parallel transition. However, the β(I) = 1.03 deviates from the limiting value (β⊥ = −1) for a perpendicular transition. This high deviation from β⊥ is due to the high contributions of the
c.c.
⎯ 1Q 1 → I is determined to be 66% at transition X → 3Q 0 ⎯→ 277.87 nm. 3.3. Assignment of the Central Top Peak. It is very important to assign the vibrational state of the top peak. This assignment will decide all peak assignments of the spectrum. At 248 nm Van Veen et al.7 assigned the top peak in their experimental TOF spectrum to v1 = 6. Felder8 got resolved peaks in the TOF spectra and assigned the top peak to v1 = 4. Wang et al.9 got TOF spectra with better resolution, and assigned the top peak to v1 = 5. Their assignments are different. We decide to verify this assignment by the turn-around time method. The turn-around time is the extra time (Δt = tb − tf) of the top backward fine peak in D compared with the top forward peak in A of the whole TOF spectrum (see Figure 1). From the turn-around time Δt, we can get the center of mass velocity VCM of the fragments I* and the translational energy ET of the fragments I* and CF3 from the following formulas,
c.c.
⎯ 1Q 1 → I (X∥), added to the parallel transition X → 3Q 0 ⎯→ ⊥
perpendicular transition X → 3Q 1 → I (X⊥). From our experimental data, we obtain X∥ = 0.68 and X⊥ = 0.32. This means the contribution from the parallel transition c.c.
X → 3Q 0 ⎯→ ⎯ 1Q 1 → I is 68% in the I channel at 266.51 nm. 3.2.2. The Photodissociation of CF3I in the I Channel at 277.87 nm. Figure 6a displays the TOF spectrum of I+ from CF3I → CF3(v1, v2) + I at 277.87 nm. Similar to the I channel at 266.51 nm, the CF3 fragment in the I channel at 277.87 nm D
DOI: 10.1021/acs.jpca.6b06988 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A Table 1. Normalized State Population P(v1, v2) of the Vibrational States of the CF3 Fragmenta P(v1, v2) I* channel (v1, v2) 0 1 0 2 1 3 2 4 3 5 4 6 5 7 6 8 7 9
0 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
Eavl (eV) ET (eV) Eint (eV) Eint/Eavl ∑P(v1, 1)/∑P(v1, 0) β
I channel
248.00 nm
266.60 nm
277.38 nm
266.51 nm
277.87 nm
0.005 0.026 0.032 0.070 0.064 0.128 0.108 * 0.155 0.102 0.105 0.073 0.067 0.033 0.031 1.73 1.35 0.38 0.220 0.41/0.59
0.048 0.142 0.118 * 0.171 0.139 0.160 0.088 0.084 0.034 0.014 1.38 1.18 0.20 0.145 0.38/0.62 1.86
0.171 * 0.274 0.154 0.188 0.101 0.069 0.023 0.020 1.20 1.07 0.13 0.108 0.28/0.72 1.86
0.007 0.007 0.011 0.011 0.019 0.041 0.065 0.099 0.105 * 0.124 0.088 0.085 0.082 0.074 0.059 0.052 0.039 0.034 2.32 1.85 0.47 0.202 0.47/0.53 1.03
0.009 0.024 0.030 0.039 0.091 * 0.167 0.126 0.137 0.102 0.079 0.071 0.062 0.038 0.023 2.14 1.81 0.33 0.154 0.46/0.54 0.98
a
Also the available energy Eavl, the translational energy ET, the internal energy of the CF3 fragment Eint, the energy partition fraction Eint/Eavl, the state partition ratio ∑P(v1, 1)/∑P(v1, 0), and the anisotropy parameter β from the CF3I photodissociation in the I* channel at 248.00, 266.60, and 277.38 nm and that in the I channel at 266.51 and 277.87 nm.
⎛ Δt ⎞ ⎛ qE ⎞⎛ Δt ⎞ (VCM)I * = a⎜ ⎟ = ⎜ ⎟⎜ ⎟ ⎝ 2 ⎠ ⎝ mI ⎠⎝ 2 ⎠
Therefore, Eint/Eavl increases with Eavl. The vibrational quantum number of the top peak increases with Eavl, and the partition ratio ∑P(v1, 1)/∑P(v1, 0) also increases with Eavl. Similarly, for the I channel at different hv, the energy partition fraction Eint/ Eavl, the quantum number of the top peak, and the partition ratio ∑P(v1, 1)/∑P(v1, 0) all increase with Eavl. The impulsive soft model29 for the photodissociation process of CF3I, in which only the C and I atoms are considered to participate in the initial dissociation step, predicts that EV (ν1 and ν2) will take up ∼75% of Eavl. While from the impulsive rigid model,29 the Eint would be 0 due to the C3v geometry of CF3I. In the present study, Eint/Eavl is obtained to be 10% ∼ 25% for the I* and I channel, which lies between the limiting values from the impulsive soft and rigid model. The photodissociation dynamics of CF3I should be considered under the quantum mechanism, and the excitation of the internal energy of CF3 fragments, which is an important dynamic parameter, would be predicted from the quantum calculation. The Eint/Eavl predicted from the classical models should be only for estimation. From the present experiments of CF3I photodissociation in the I* channel at 248.00, 266.60, and 277.38 nm, it is shown that the ν1 symmetric stretch mode (701 cm−1) is strongly excited, and the ν2 umbrella mode (1086 cm−1) is relatively weakly excited. From the I channel at 266.51 and 277.87 nm, the 1086 cm−1 ν2 umbrella mode is increasingly excited due to the higher available energy. The ν1 and ν2 vibration modes of CF3 have similar C3v vibrational behavior, but the excitations are quite different. From the experimental results, the ν1 vibration is excited strongly up to v1 = 9, while the v2 vibration is excited
1 1 2 mI(VCM)2 I * + mCF3(VCM)CF 3 2 2 ⎡ mCF + mI ⎤ 1 3 ⎥ = mI(VCM)2I * ⎢ 2 m ⎢⎣ ⎥⎦ CF3
ET =
from the turn-around time Δt and ET, we get Ev ≈ Eavl − ET. Then the assignment of top peaks can be ensured. In the calculation, we have considered the experimental conditions to make some modifications (e.g., the direction of the molecular beam). 3.4. State and Energy Distribution. The PTS in Figures 2−6b are simulated with two series of the resolved vibrational states. From the CF3I photodissociation in the I* channel at 248.00, 266.60, and 277.38 nm, and in the I channel at 266.51 and 277.87 nm, the vibrational state relative populations of CF3 fragments are given in Table 1. The corresponding available energy Eavl, the translational energy ET, the internal energy of the CF3 fragment Eint, the energy partition fraction Eint/Eavl, the state partition ratio ∑P(v1, 1)/∑P(v1, 0), and the anisotropy parameter β have been calculated and given in Table 1. The partitioning of the available energy Eavl into the translational energy ET and the CF3 fragment internal energy Eint (mainly EV) reveals some insight of the dissociation mechanism. As shown in Table 1, for the I* channel at different laser hv, when Eavl increases, Eint increases faster than ET. E
DOI: 10.1021/acs.jpca.6b06988 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A much less. This shows the stronger coupling between the ν1 vibration and the C−I reaction coordinate of the photodissociation process. 3.5. Quantum Yield Φ(I*) and the Energy at the Curve Crossing Point (Ec). The I* quantum yield Φ* (Φ* = [I*]/ ([I*] + [I])) is determined to be 0.87 at 277 nm, and is estimated to be about 0.90 at 266 nm. The obtained Φ* are consistent with other experimental values.8,12,15,20 At 277 nm, the quantum yield (Φ) for each product channel, the curve crossing possibility Pcc, and relative absorption strengths σ of each excited state are shown in Table 2. The relative absorption strengths σ are 96% and 4% for 3Q0 and 3Q1, in good agreement with the absorption spectra of gaseous CF3I obtained by Gedanken.30
(for vertical excitation energy calculation) and CASSCF (for geometrical optimization), under the basis sets of 4s3p2d1f (for C), 4s3p2d1f (for F), and 7s6p4d3f2g (for I), with the selected active space comprised of 8 electrons in 10 orbitals, the relative potential energy curves are calculated with Molcas7.5 software.40 The obtained vertical excitation energies for X → 1 Q1, 3Q0 and 3Q1 (photon absorption peaks for the three electronic states) are 4.97, 4.70, and 4.25 eV, respectively, in good agreement with Gedanken’s experimental results.30 In our calculation, the potential curves 3Q0 and 1Q1 cross at Ec = 31385 cm−1 at the RC−I = 2.358 Å. Ajitha et al.41 calculated the potential energy at the curve crossing is 28785 cm−1 at RC−I = 2.372 Å. While Yu42 calculated Ec = 30012 cm−1 at RC−I = 2.401 Å, and Alekseyev et al.43 calculated Ec near 27600 cm−1 at RC−I = 2.497 Å. The experimental measured Ec at the curve-crossing point in this work and by Li et al.,12 are quite close to the value obtained from our calculation. It can be seen that the theoretical data obtained by different groups deviate in a large range and are different from the experimental results, mainly due to the difficulty in calculating accurately for the curve-crossing region. The Ec will change a lot, while the C−I bond length changes a little at the crossing region.44 Ajitha et al.41 also suggested that many factors, (such as the zero-point corrections for spectator modes, the umbrella vibrational mode excitation), altogether make a difference of 3000 cm−1 between calculated potential curves and the experimentally ones are quite reasonable. 3.6. The Excitation of ν1 and ν2 Vibrational Modes. From the classical mechanism, It can be imagined that the ν1 and ν2 vibrations of CF3 fragments would be the most easily excited during photodissociation from the repulsive force on the bond breaking. However, from the quantum mechanism, the excited vibration should be also determined by the interplay between the geometric changes of the molecular system and the wavepacket evolution upon the excited potential surfaces. We have calculated the geometries of the CF3I molecule and CF3 free radical, as shown in Table 3. The CF3 group is pyramidal both in the CF3I molecule ground state and in the free radical. The geometric changes of C−F bond length and the FCF bond angle are slight. It can not be predicted which vibrational mode (ν1 symmetric stretch or ν2 umbrella) of the CF3 fragment would be more easily excited. This is different from the photodissociation of CH3I, in which the pyramidal CH3 moiety changes to planar in the fragment, and will cause the umbrella vibration of CH3 to be excited more than the symmetric stretch mode.
Table 2. Quantum Yield (Φ) for Each Product Channel, the Curve Crossing Possibility (Pcc), and the Absorption Strength (σ) at 277 nm excited state (CF3I) 3
Q0
3
Q1
product channel
Φ
I*(3Q0) I(1Q1← 3Q0) I(3Q1)
0.87 0.086 0.044
Pcc
σ 96%
9% 4%
At 277 nm, from the Pcc value, 9% of excited 3Q0 CF3I will undergo the curve crossing 1Q1 ← 3Q0, and then produce I instead of I*. We can get the potential energy Ec at the crossing point of the potential energy curves of the 3Q0 and 1Q1 states, according to the modified one-dimensional Landau−Zener formula.31,32 ⎛ ⎞ ζm ⎟ 1 − Pcc = exp⎜⎜ − 1/2 ⎟ ⎝ 2(Ehν − Ec) ⎠
where ζm is the modified Landau−Zener parameter, including a nonadiabatic coupling term and the difference in gradients at the crossing point of two potential surfaces. Ehv is the photon energy. By measuring the curve crossing possibility at two different wavelengths, the potential energy Ec at the curvecrossing point can be obtained. From the Pcc values at 277 and 304 nm33 obtained from our experiment, Ec is obtained to be 32871 cm−1, consistent with the value obtained experimentally by Li et al.12 (32854 cm−1). We have calculated the potential energy curves of CF3I involved in the photodissociation process, as shown in Figure 7. By employing the method34−39 of MS-CASPT2/CASSI-SO
Figure 7. Calculated potential curves of CF3I with spin−orbit effects. (a) Potential curves for ground and UV accessible electronic states of CF3I. (b) Zoom in of potential curves 3Q0 and 1Q1 in the crossing region. F
DOI: 10.1021/acs.jpca.6b06988 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
symmetric stretch mode and ν2 (1086 cm−1) the umbrella mode are closely coupled to the C−I bond fission. Here we get ∑P(v1, 1)/∑P(v1, 0) < 1, and this ratio increases with Eavl. We agree that the (v1, 2) and (v1, 3) states can be excited in lower populations, but they can not be resolved as independent peaks in PTS and would be mixed in (v1+3, 0) and (v1+3, 1) resolved peaks. The populations of (v1, 2) and (v1, 3) states have been estimated. The quantum yield Φ(I*) and the energy at curve crossing point have been also analyzed.
Table 3. Geometries of CF3I in the Ground State and CF3 Free Radical CF3I
RC−I (Å)
RC−F (Å)
∠FCF (deg.)
ref
2.13 2.14 2.131
1.332 1.329 1.337 1.312 1.32
108 108.42 108.2 111.3
2.358
1.325
45 46 this work this work 47 29 this work
CF3
crossing region
111.1 109.6
■
AUTHOR INFORMATION
Corresponding Author
Van Veen7 and Clary3 performed the theoretical calculation on the photodissociation dynamics of CF3I in the A band. They suggested that the combinational excitation of the ν1,ν2 vibrational modes of CF3 fragments is determined by the vibrational interaction on the excited state potential energy surfaces. Thus, in the CF3I photodissociation, the decisive factor for which vibrational modes would be excited is the dynamic process on the excited potential surfaces, not the slight change between the geometry of CF3I and CF3 fragment. The experimental observations presented here confirm the prediction of combinational excitation of ν1 and ν2 vibrations. 3.7. The Excitation of (v1, 2) and (v1, 3) Vibrational States of the CF3 Fragment. In the present PTS experiments on CF3I → CF3 (v1, v2) + I*/I, we resolved the main series of (v1, 0) peaks, and partially resolved a minor series of (v1, 1) peaks. How about the excitation of (v1, 2) states and (v1, 3) states? We agree that (v1, 2) and (v1, 3) states can be excited, but their populations are much lower than the (v1, 0) and (v1, 1) states. Therefore, they are difficult to be resolved. Moreover, comparing (0, 2) and (3, 0) two states, their vibrational energy difference is small.
*Postal address: State Key Laboratory of Molecular Reaction Dynamics, Institute of Chemistry, Chinese Academy of Sciences, Zhongguancun North First Street 2, Beijing 100190, China. E-mail address: chengmin@iccas.ac.cn. Phone: +86-1062563168. ORCID
Min Cheng: 0000-0001-5826-5149 Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work is supported by the National Natural Science Foundation of China under Grant Nos. 21203207 and 21173236. We thank Prof. Yajun Liu of Beijing Normal University for the help on the calculation of potential energy curves.
■
REFERENCES
(1) Mulliken, R. S. Intensities in Molecular Electronic Spectra X. Calculations on Mixed-Halogen, Hydrogen Halide, Alkyl Halide, and Hydroxyl Spectra. J. Chem. Phys. 1940, 8, 382−395. (2) Mulliken, R. S.; Teller, E. Interpretation of the Methyl Iodide Absorption Bands Near Lambda 2000. Phys. Rev. 1942, 61, 283−296. (3) Clary, D. C. A Theory for the Photodissociation of Polyatomic Molecules, with Application to CF3I. J. Chem. Phys. 1986, 84, 4288− 4298. (4) Hennig, S.; Engel, V.; Schinke, R. Vibrational State Distributions Following the Photodissociation of (Collinear) Triatomic Molecules: The Vibrational Reflection Principle in Model Calculations for CF3I. J. Chem. Phys. 1986, 84, 5444−5454. (5) Untch, A.; Hennig, S.; Schinke, R. The Vibrational Reflection Principle in Direct Photodissociation of Triatomic − Molecules - Test of Classical − Models. Chem. Phys. 1988, 126, 181−190. (6) Forney, D.; Jacox, M. E.; Irikura, K. K. Matrix Isolation Study of the Interaction of Excited Neon Atoms with CF4. Infrared Spectra of CF3+ and CF3−. J. Chem. Phys. 1994, 101, 8290−8296. (7) Van Veen, G. N. A.; Baller, T.; De Vries, A. E.; Shapiro, M. Photofragmentation of CF3I in the A Band. Chem. Phys. 1985, 93, 277−291. (8) Felder, P. Photodissociation of CF3I at 248 nm: Internal Energy Distribution of the CF3 Fragments. Chem. Phys. 1990, 143, 141−150. (9) Wang, X.; Tian, Z.-X.; Shi, T.-J.; Shi, X.-H.; Yang, D.-L.; Zhu, Q.H. High-Resolution Photofragment Translational Spectra of the Photodissociation of CF3I at 248 nm. Chem. Phys. Lett. 2003, 380, 600−603. (10) Aguirre, F.; Pratt, S. T. Velocity Map Imaging of the Photodissociation of CF3I: Vibrational Energy Dependence of the Recoil Anisotropy. J. Chem. Phys. 2003, 118, 1175−1183. (11) Yu, Z.-J.; Cheng, M.; Xu, X.-L.; Yu, D.; Du, Y.-K.; Zhu, Q.-H. Vibrational State Distributions Following the Photodissociation of CF3I Near 304 nm. Chem. Phys. Lett. 2010, 488, 158−161.
−ΔE T = ΔE V ≈ 2ν2 − 3ν1 ≈ 2 × 1086 − 3 × 701 = 69 cm−1
The 69 cm−1 energy difference between (0, 2) and (3, 0) states is too small to be resolved by the present PTS. So the excited weaker (0, 2) state will be mixed in the (3, 0) peak. Similarly, the excited weaker (v1, 2) states will be mixed in respective (v1+3, 0) peaks of the main series, and the excited weaker (v1, 3) states will be mixed in respective (v1+3, 1) peaks of the minor series. In our obtained PTS, the peaks of (v1, 2) and (v1, 3) states can not be resolved as independent new peaks. Let us try to estimate the approximate populations of (v1, 2) and (v1, 3) states. Assume the excitation probability ratio of ν2 1086 cm−1 and ν1 701 cm−1 states is about P(v2)/P(v1) = 0.5/1. Then we get the ratio for P(0, 2)/P(3, 0) = 0.25/1 approximately, i.e., the (0, 2) state is mixed in the (3, 0) peak with a ∼ 20% fraction of the peak. Similarly, (v1, 2) states are mixed in the main series (v1+3, 0) peaks with P(v1, 2)/ P(v1+3, 0) = 0.25/1, and (v1, 3) states are mixed in the minor series (v1+3, 1) peaks with P(v1, 3)/P(v1+3, 1) = 0.25/1 approximately. The above estimation is very crude, as the populations can be affected by various factors.
4. CONCLUSIONS The high resolution vibrational state resolved PTS of CF3I → CF3 (v1, v2) + I*/I are obtained at 248, 266, and 277 nm. Except the 701 cm−1 ν1 mode, the vibrational excitation of the 1086 cm−1 ν2 mode in the CF3 photofragment is also resolved for the first time. This proves that both ν1 (701 cm−1) the G
DOI: 10.1021/acs.jpca.6b06988 J. Phys. Chem. A XXXX, XXX, XXX−XXX
Article
The Journal of Physical Chemistry A
Populations of I (2P1/2) and I (2P2/3) and Dynamics of Curve Crossing. J. Chem. Phys. 1996, 104, 5815−5820. (33) Yu, Z.-J. Development of a Mini-TOF Photofragment Translational Spectrometer and Study on the Photodissociation of Perfluoroalkyl Iodides. Ph.D. Dissertation, Institute of Chemistry, Chinese Academy of Sciences, 2009. (34) Roos, B. O.; Taylor, P. R.; Siegbahn, P. E. M. A Complete Active Space SCF Method (CASSCF) Using a Density-Matrix Formulated Super-CI Approach. Chem. Phys. 1980, 48, 157−173. (35) Finley, J.; Malmqvist, P. A.; Roos, B. O.; Serrano-Andres, L. The Multi-State CASPT2 Method. Chem. Phys. Lett. 1998, 288, 299−306. (36) Forsberg, N.; Malmqvist, P. A. Multiconfiguration Perturbation Theory with Imaginary Level Shift. Chem. Phys. Lett. 1997, 274, 196− 204. (37) Hess, B. A. Relativistic Electronic-Structure Calculations Employing a 2-Component No-Pair Formalism with External-Field Projection Operators. Phys. Rev. A: At., Mol., Opt. Phys. 1986, 33, 3742−3748. (38) Hess, B. A.; Marian, C. M.; Wahlgren, U.; Gropen, O. A MeanField Spin-Orbit Method Applicable to Correlated Wavefuctions. Chem. Phys. Lett. 1996, 251, 365−371. (39) Roos, B. O.; Lindh, R.; Malmqvist, P. A.; Veryazov, V.; Widmark, P. O. Main Group Atoms and Dimers Studied with A New Relativistic and Basis Set. J. Phys. Chem. A 2004, 108, 2851−2858. (40) Karlstrom, G.; Lindh, R.; Malmqvist, P. A.; Roos, B. O.; Ryde, U.; Veryazov, V.; Widmark, P. O.; Cossi, M.; Schimmelpfennig, B.; Neogrady, P.; et al. MOLCAS: A Program Package for Computational Chemistry. Comput. Mater. Sci. 2003, 28, 222−239. (41) Ajitha, D.; Wierzbowska, M.; Lindh, R.; Malmqvist, P. A. Spinorbit ab initio Study of Alkyl Halide Dissociation via Electronic Curve Crossing. J. Chem. Phys. 2004, 121, 5761−5766. (42) Yu, L. Highly Accurate Extensive Theoretical Investigation on Small Molecular Reaction Systems Including Spin-Orbit Coupling. Ph.D. Dissertation, Institute of Chemistry, Chinese Academy of Sciences, 2012. (43) Alekseyev, A. B.; Liebermann, H. P.; Buenker, R. J. Potential Energy Surfaces for Ground and Excited Electronic States of the CF3I Molecule and Their Relevance to Its A-band Photodissociation. Phys. Chem. Chem. Phys. 2013, 15, 6660−6666. (44) Lao, K. Q.; Person, M. D.; Xayariboun, P.; Butler, L. J. Evolution of Molecular Dissociation through an Electronic Curve Crossing: Polarized Emission Spectroscopy of CH3I at 266 nm. J. Chem. Phys. 1990, 92, 823−841. (45) Herzberg, G. Molecular Spectra and Molecular Structure; Van Nostrand Press: Princeton, U.K., 1966. (46) Cox, A. P.; Duxbury, G.; Hardy, J. A.; Kawashima, Y. Microwave-Spectra of CF3Br and CF3I Structures and DipoleMoments. J. Chem. Soc., Faraday Trans. 2 1980, 76, 339−350. (47) Gedanken, A.; Rowe, M. D. Magnetic Circular-Dichroism Spectra of the Methyl Halides Resolution of the n-σ* Continuum. Chem. Phys. Lett. 1975, 34, 39−43.
(12) Li, G.-S.; Shin, Y. K.; Hwang, H. J. The Photodissociation Reaction Dynamics of CF3I at 304 nm (3Q0+, 1Q1 ← 3Q0+, and 3Q1). J. Chem. Phys. 2006, 125, 214312. (13) Kavita, K.; Das, P. K. Dynamics of I*(2P1/2) Production from Fluorinated Alkyl Iodides at 266, 280, and ∼ 305 nm. J. Chem. Phys. 2000, 112, 8426−8431. (14) Felder, P. Photodissociation of CF3I at 248 nm: Kinetic Energy Dependence of the Recoil Anisotropy. Chem. Phys. 1991, 155, 435− 445. (15) Kim, Y. S.; Kang, W. K.; Jung, K. State-Selective Photofragment Imaging of Iodine Atoms via Photodissociation of CF3I at 277 nm. J. Chem. Phys. 1996, 105, 551−557. (16) Felder, P. The Influence of the Molecular-Beam Temperature on the Photodissociation of CF3I at 308 nm. Chem. Phys. Lett. 1992, 197, 425−432. (17) Furlan, A.; Gejo, T.; Huber, J. R. Probing Curve Crossing by Wavelength-Dependent Recoil Anisotropy: The Photodissociation of CF3I at 275−303 nm Studied by Photofragment Translational Spectroscopy. J. Phys. Chem. 1996, 100, 7956−7961. (18) Carlson, G. A.; Pimentel, G. C. Infrared Detection of Gaseous Trifluoromethyl Radical. J. Chem. Phys. 1966, 44, 4053−4054. (19) Bowman, J. M.; Huang, X. C.; Harding, L. B.; Carter, S. The Determination of Molecular Properties from Multimode with an Application to the Calculation of Franck-Condon Factors for Photoionization of CF3 to CF3+. Mol. Phys. 2006, 104, 33−45. (20) Aguirre, F.; Pratt, S. T. Photoionization of Vibrationally Hot CH3 and CF3. J. Chem. Phys. 2005, 122, 234303. (21) Hancock, G.; Hutchinson, A.; Peverall, R.; Richmond, G.; Ritchie, G. A. D.; Taylor, S. 266 nm Photolysis of CF3I and C2F5I Studied by Diode Laser Gain FM Spectroscopy. Phys. Chem. Chem. Phys. 2007, 9, 2234−2239. (22) Person, M. D.; Kash, P. W.; Butler, L. J. The Influence of Parent Bending Motion on Branching at a Conical Intersection in the Photodissociation of CX3I (X = H,D,F). J. Chem. Phys. 1991, 94, 2557−2563. (23) Tian, Z.-X.; Bi, W.-B.; Deng, H.-D.; Wang, X.; Tang, Z.-C.; Zhu, Q.-H. A Mini-TOF Photofragment Translational Spectrometer Photofragmentation of CF3I at 281.73 nm. Chem. Phys. Lett. 2004, 400, 15−18. (24) Cheng, M.; Yu, Z.-J.; Hu, L.-L.; Yu, D.; Dong, C.-W.; Du, Y.-K.; Zhu, Q.-H. Vibrationally Resolved Photofragment Translational Spectroscopy of CH3I from 277 to 304 nm with Increasing Effect of the Hot Band. J. Phys. Chem. A 2011, 115, 1153−1160. (25) Hu, L.-L.; Zhou, Z.-M.; Dong, C.-W.; Zhang, L.-J.; Du, Y.-K.; Cheng, M.; Zhu, Q.-H. Photofragment Translational Spectroscopy of CH3I at 255 nmwith the High Excitation of the Symmetric Stretch Vibration of CH3 Fragment. J. Chem. Phys. 2012, 137, 144302. (26) Xu, X.-L.; Yu, Z.-J.; Bi, W.-B.; Xiao, D.-Q.; Yu, D.; Du, Y.-K.; Zhu, Q.-H. Photofragment Translational Spectroscopy of n-C3H7I and i-C3H7I Near 280 and 304 nm. J. Phys. Chem. A 2008, 112, 1857− 1863. (27) Asher, R. L.; Ruscic, B. On the Heats of Formation of Trifluoromethyl Radical CF3 and Its Cation CF3+. J. Chem. Phys. 1997, 106, 210−221. (28) Hu, L.-L.; Zhou, Z.-M.; Dong, C.-W.; Zhang, L.-J.; Du, Y.-K.; Cheng, M.; Zhu, Q.-H. Vibrationally Mediated Photodissociation of CH3I [v1 = 1] at 277.5 nm: The Vibrationally Adiabatic Process. J. Phys. Chem. A 2013, 117, 4352−4357. (29) Riley, S. J.; Wilson, K. R. Excited Fragments from Excited Molecules: Energy Partitioning in the Photodissociation of Alkyl Iodides. Faraday Discuss. Chem. Soc. 1972, 53, 132−146. (30) Gedanken, A. The Magnetic Circular Dichroism of the A band in CF3I, C2H5I and t-BuI. Chem. Phys. Lett. 1987, 137, 462−466. (31) Godwin, F. G.; Paterson, C.; Gorry, P. A. Photofragmentation Dynamics of n-C3H7I and i-C3H7I at 248 nm. Mol. Phys. 1987, 61, 827−848. (32) Kang, W. K.; Jung, K. W.; Kim, D.-C.; Jung, K.-H. Photodissociation of Alkyl Iodides and CF3I at 304 nm: Relative H
DOI: 10.1021/acs.jpca.6b06988 J. Phys. Chem. A XXXX, XXX, XXX−XXX