All-Electron Relativistic Multireference Configuration Interaction

Article pubs.acs.org/JPCA

All-Electron Relativistic Multireference Configuration Interaction Investigation of Fluoroiodo Carbene Erping Sun, Hang Lv, Dandan Shi, Changli Wei, Haifeng Xu,* and Bing Yan* Institute of Atomic and Molecular Physics, Jilin University, Changchun 130012, China ABSTRACT: We present herein the first all-electron relativistic internally contracted multireference configuration interaction with Davidson correction (icMRCI+Q) study on the low-lying states of fluoroiodo carbene, CFI, which contains the most electronegative element (fluorine) and the heavy halogen (iodine). The potential energy surface (PES) of the first excited singlet state (Ã 1A″) of CFI was carefully examined along the C−I bond distance at the icMRCI+Q/ANO-RCC level, while the other two geometric parameters were optimized at every C−I bond length in contrast to fixing them at the equilibrium values. A reliable barrier height of the Ã 1A″ state was determined to be 625 cm−1 by our high-level icMRCI+Q calculations with large ANO-RCC basis set and with inclusion of the spin−orbit coupling, core−valence correlation, and zero-point-energy. Finally, the electronic states of CFI with vertical transition energy up to 6 eV were studied. The calculation presented here will provide more comprehensive results about the structure and behavior of electronic states of CFI radical.

■

INTRODUCTION Halogenated carbenes are important reactive molecular species that play key roles in a wide variety of chemical reactions. For example, halogenated carbenes are possible photoproducts of chlorofluorocarbons and halons, which have significant contribution to ozone destruction; halogenated carbenes are also important intermediates in organic synthesis and in gasphase combustion.1,2 In particular, triatomic halogenated carbenes, CXY(X = H, F, Cl, Br, I; Y = F, Cl, Br, I), are viewed as model systems for understanding the spectroscopy, dynamics and chemistry of carbenes, and benchmarks for comparison between theoretical and experimental investigations (see ref 3 and references therein). During the past several decades, CXY carbenes have received continuous research interest regarding the structure, spectrum, and dynamics of the low-lying electronic states, with the aid of various spectroscopic techniques and computational methods (for examples, see refs 4−24). Despite the fact that the history of the study of halogenated carbenes spans more than a half century, our knowledge about the electronic states of fluoroiodo carbene, CFI, is rather limited. The lack of experimental and theoretical studies could be attributed to the most electronegative element, fluorine, and the heavy halogen, iodine, that are contained in CFI carbene, which can lead to unique properties and complicated interactions of electronic states (e.g., Renner−Teller effect, spin−orbit coupling, etc.). To the best of our knowledge, no experimental studies have been performed on the electronic states of CFI. However, ab initio calculations have led experimental investigations with a few studies on the ground singlet state, the first excited singlet state, and the lowest triplet state of CFI.9,13,14,20 However, the large numbers of electrons make a reliable calculation still a challenging work, which stimulates researchers to perform a high-level correlated © 2014 American Chemical Society

calculation method to retrieve structure and spectral data of CFI. As all the halocarbenes studied to date, the ground state of CFI (X̃ 1A′) is also a singlet state. The equilibrium geometry and the harmonic vibrational frequencies of the X̃ 1A′ state have been obtained by several studies,9,10,13,14 using CCSD(T), QCISD, CASSCF, CASPT2, and recently icMRCI+Q method20 with effective core potentials (ECPs) to represent the core electrons of iodine. The S−T gap of CFI was determined to be the smallest in the series of dihalogenated carbenes containing fluorine atom (CFX, X = F, Cl, Br, I), due to the effect of the electronegativity of the substituent.9,13,20,23 However, the S−T gaps are difficult to determine accurately for many of the halogenated carbenes, which have rather large variations using different levels of theory and basis sets. In case of CFI, the calculated S−T gap varies from 7940 to 10280 cm−1 using various theoretical methods.9,13,14,20,23 Regarding the singlet excited states, the first study was carried out by Drake et al.13 using the CASSCF, CASPT2, and CISD levels of theory to calculate the geometries and transition energies of the Ã 1A″ states of all iodine-containing carbenes. The Ã 1A″ state of CFI, along with other fluorine-containing carbenes, was investigated very recently by Sun et al. using a high-level icMRCI+Q/ccpV5Z theory.20 In their study, by taking CFBr as an example, Sun et al. also indicated that large basis sets and highly correlated methods are necessary to achieve reliable results for high-Z fluorine-containing carbenes. The only information about the potential energy surface (PES) of the Ã 1A″ state of CFI was provided by Standard and Quandt14 in 2003 in their CASPT2 calculation, which indicated that the state is a predissociative state. The barrier height was Received: December 6, 2013 Revised: March 7, 2014 Published: March 10, 2014 2447

dx.doi.org/10.1021/jp411967r | J. Phys. Chem. A 2014, 118, 2447−2452

The Journal of Physical Chemistry A

Article

RCC level. The barrier height for Ã 1A″ state was calculated with two different levels, all-electron and relativistic effective core potential (RECP). The all-electron calculations are performed at icMRCI+Q+DK3/ANO-RCC level. In RECP calculations, we use pseudopotential ECP28MDF in combination with aug-cc-pVXZ-PP (X = T(3),Q(4),5)34 basis sets describing outer electron for I atom, and aug-cc-pVXZ (X = T,Q,5)35 basis set series for C and F atoms. The RECP calculations are referred as MRCI+Q/RECP+VXZ (X = T, Q, 5). In RECP calculations, we extrapolated the energies to complete basis set (CBS) limit. The extrapolated energies including the zeroth-order reference energy (CASSCF energy, ECAS) and the dynamical correlation energy (energy difference between MRCI+Q energy and CASSCF energy, Ecorr). In the case of Dunning’s correlation-consistent cc-pVXZ basis sets, the zeroth order energies approaching their CBS limits are expressed as ECAS(X) = ECAS(CBS) + ao exp(−αX),36 and we have used the values for X = T, Q, 5 to determine the three unknowns, ECAS(CBS), ao, and α. The dynamical correlation energies approach their CBS limits according to the inverse power law, ΔECORR(X) = ΔECORR(CBS) + acX−3, and we have used the data for X = T,Q or Q,5 to determine the extrapolated value ΔECORR(CBS). The total energy is the sum of ECAS(CBS) and ΔECORR(CBS). The two-pronged CBS extrapolation procedure has indeed given reliable total CBS energies The one-dimensional potential energy cuts of 12 electronic states of CFI computed at icMRCI+Q+DK3/ANO-RCC level were given with respect to the angle of F−C−I, C−I distance, and C−F distance, respectively, with the other two parameters fixed at their respective equilibrium values. The oscillator strengths for different excited states to ground state are calculated at the same level of theory. All calculations were carried out using the MOLPRO37 software package.

calculated to be 475, 1043, and 1525 cm−1 with different basis sets of SBKJC(3d), SBKJC(3df), and Basis3 (SBKJC(3df) for I plus DZ(3df) for C and F), respectively. The authors argued that the dissociation barrier height of CFI could be larger than their CASPT2/Basis3 result (1525 cm−1). Such conclusion was deduced with the aid of comparing the calculated result of CFBr at the same theoretical levels with the experimental barrier reported by Knepp et al.24 Since no experimental results of the Ã 1A″ state of CFI is available to compare with the calculation results, further study of the PES at higher levels of theory should be carried out. In fact, in the case of CFBr, Knepp’s experimental results24 were debated by Truscott et al.17 in their recent high-resolution LIF spectroscopic study, which reassigned the T00 value of the Ã 1A″ state to 23271 cm−1 (2365 cm−1 higher than Knepp’s result24 and is consistent with several following theoretical studies18,20,22) and estimated the barrier height about 1000 cm−1 from fluorescence quantum yields (in contrast to Knepp’s result of 3360 cm−124). In addition, a recent study22 has demonstrated that the scalar relativistic effect, spin−orbit coupling (SOC), and core− valence (CV) effect should be considered in the calculation to obtain reliable energies of CFBr, which has not been investigated in CFI carbene. For the electronic states beyond the Ã 1A″ state of CFI that may play important roles in the photochemical processes in the UV region, there is unfortunately no information available in the literature. In the present work, we report the first all-electron relativistic icMRCI+Q study of CFI carbene. We focus particularly on the electronic excited states of CFI. The PES of the Ã 1A″ state was carefully examined along the C−I bond distance at the icMRCI +Q/ANO-RCC level, while the other two geometric parameters were optimized at every C−I bond length in contrast to fixing them at the equilibrium values. On this basis, we calculated the barrier height of the Ã 1A″ state with different basis sets, and including the SOC effect, CV correlation, and zero-point-energy (ZPE). Finally, the electronic states of CFI with transition energy up to 6 eV were studied. The calculation presented here will provide more comprehensive results about the structure and behavior of electronic states of CFI radical.

■

RESULT AND DISCUSSION A. All-Electron Relativistic icMRCI+Q Results of the Equilibrium Geometries of the X̃ 1A′, Ã 1A″, and ã3A″ States. The equilibrium bond lengths and bond angles of the ground singlet state (X̃ 1A′), the first excited singlet state (Ã 1A″), and the lowest triplet state (ã3A″) of CFI were calculated at a high-level all-electron relativistic icMRCI+Q/ ANO-RCC method. The results are presented in Table 1, along with available theoretical results in the literature using different levels of methods. There are no experimental determined geometries of CFI available for comparison. However, the geometries of the ground state calculated using different methods agree well with each other, with variations of 0.025 Å for the C−F bond length, 0.063 Å for the C−I bond length, and 1.2° for F−C−I angle. The calculated C−I bond length and F− C−I bond angle of the Ã 1A″ state at CASPT2/SBKJC(3d) level13 are 0.261−0.296 Å larger and 3.1−4.1° smaller than those of the other studies, respectively, and the calculated F− C−I bond angle of the ã3A″ state at QCISD/6-311G(d,p) level9 is 7.7−8.6° larger than the other calculated results. Except those, the geometries of the Ã 1A″and the ã3A″ state using different methods are quite consistent with variations less than 1.7%. There is little change of the C−F bond length or the C−I bond length as an electron is excited from the X̃ 1A′ state to the Ã 1A″ or ã3A″ state. The calculated F−C−I bond angle, however, is different for different electronic states. Our calculated bond angle of the ground X̃ 1A′ state is 107.4°, which is the smallest of the three electronic states, due to the

■

METHODS In our work, the electronic states of CFI were investigated using complete active space multiconfiguration self-consistent field (CASSCF)25 method followed by internally contracted singly and doubly excitation multireference configuration interaction (icMRCISD) method26,27 with Davidson correction (+Q)28 to account for higher order excitation configurations. The active space consists of 18 valence electrons and 12 valence orbitals corresponding to n = 2 atomic orbitals of C and F atoms and n = 5 orbital of I atom. The relativistic contracted atomic natural orbital basis sets, ANO-RCC,29 were used for F, C, and I atoms in ab initio calculations. The scalar relativistic corrections were considered with third order Douglas−Kroll (DK3)30−32 approximation with ANORCC basis sets. The SOC corrections33 are evaluated with state-interacting methods at icMRCI+Q+DK3/ANO-RCC level; four electronic states including singlet and triplet A′/A″ are considered in SOC calculations. The core and core−valence electrons correlations of n = 4 shell for I in FCI are calculated with entirely uncontracted ANO-RCC basis. The relaxed PES of the Ã 1A″ state along C−I bond was carefully calculated. We optimized other geometrical parameters of CFI at every C−I distance at icMRCI+Q+DK3/ANO2448



Article

other saddle point is found on PES of the Ã 1A″ state of CFI, which confirms the presence of a dissociation barrier along the C−I bond. The geometry of the saddle point on the Ã 1A″ state was optimized and is listed in Table 1. Although the equilibrium geometries of our all-electron relativistic icMRCI +Q results are consistent with those of Standard and Quandt’s CASPT2 results,14 as shown in Table 1, there is relatively large difference in the results of the geometries of the saddle point, particularly for the C−I bond length (about 0.2 Å difference). The harmonic vibrational frequencies at the saddle point were determined to be 226.80, 198.55i, and 1230.70 cm−1 at the icMRCI+Q/SBKJC level. With the above optimization of the structure along the C−I bond distance, we calculated the barrier height at the icMRCI +Q/ANO-RCC level including CV and SOC corrections as well as zero-point energy (ZPE). We also performed icMRCI +Q calculations using effective core potentials (ECPs) at different basis sets for comparison. The results are presented in Table 2. Our results indicate that the barrier height of the Ã 1A″

Table 1. Equilibrium Geometries of the X̃ 1A′, Ã 1A″, and a3̃ A″ States and the Saddle Point of the Ã 1A″ State of CFI RC−F (Å)

RC−I (Å)

X̃ 1A′ State icMRCI+Q/ANO-RCC 1.288 2.163 icMRCI+Q/ECP+cc-pV5Zb 1.298 2.155 CASPT2/SBKJC(3df), Basis3c 1.306, 1.306 2.208, 2.218 QCISD/6-311G(d,p)d 1.290 2.189 CCSD(T)/6-311+G(d,p)e 1.296 2.193 CASPT2/SBKJC(3d)f 1.313 2.217 Ã 1A″ State icMRCI+Q/ANO-RCCa 1.300 2.141 icMRCI+Q/ECP+cc-pV5Zb 1.298 2.154 CASPT2/SBKJC(3df), Basis3c 1.310, 1.303 2.165, 2.130 CASPT2/SBKJC(3d)f 1.308 2.426 ã3A″ State icMRCI+Q/ANO-RCCa 1.311 2.062 icMRCI+Q/ECP+cc-pV5Zb 1.309 2.065 QCISD/6-311G(d,p)d 1.315 2.096 CCSD(T)/6-311+G(d,p)e 1.318 2.087 Saddle Point of Ã 1A″ State icMRCI+Q/ANO-RCCa 1.289 2.354 CASPT2/SBKJC(3d), 1.300, 1.292, 2.557, 2.531, SBKJC(3df), Basis3c 1.283 2.522 a

a

b

c

d

e

∠F−C−I (deg) 107.4 108.2 107.0, 106.9 107.4 107.4 107.6 125.1 124.9 124.8, 125.8 121.7 125.2 125.2 132.9 124.3

Table 2. Barrier Height (cm−1) of the Ã 1A″ of CFI Calculated with Different Methods and Basis Sets barrier (cm−1)

121.9 121.4, 120.7, 120.5

icMRCI+Q /ECP+cc-pVTZ icMRCI+Q /ECP+cc-pVQZ icMRCI+Q/ECP+cc-pV5Z icMRCI+Q/CBS icMRCI+Q/ANO-RCC CV SOC ZPE TOTAL

f

This work. Ref 20. Ref 14. Ref 9. Ref 23. Ref 13.

in-plane sp2-like orbital of the two nonbonded electrons of carbon in the ground state. The bond angles of the Ã 1A″ and ã3A″ states are about 18° larger than that of the X̃ 1A′ state, which can be attributed to the transition of at least one nonbonding electron to an out-of-plane p-type orbital of carbon. B. Relaxed PES of the Ã 1A″ State and the Dissociation Barrier Height. The PES of the Ã 1A″ state of CFI radical was carefully examined along C−I bond distance with the high-level icMRCI+Q/ANO-RCC calculation, as shown in Figure 1. At

CFBr > CFI.11,12,17,24 The barrier height of CFCl was experimentally determined to be 4073 cm−1,12 which was about 1700 cm−1 smaller than the result calculated at the CASPT2(18,12)/cc-

Figure 1. Relaxed potential-energy surface scan along the C−I bond of the Ã 1A″ state of CFI. The inset figure shows dependence of the C−F bond length and F−C−I angle on the C−I bond length.

each step of the calculation we optimized the other geometric parameters as the C−I bond length was changing. As shown in the inset of Figure 1, the C−F bond length and the F−C−I angle change as the C−I bond length varies. The calculation step around the saddle point was set as small as 0.005 Å in order to retrieve accurate information about the saddle point. Similar as CFBr radical,22 only one maximum point while no 2449



Article

pVTZ level.11 For CFBr, the latest LIF spectroscopic study17 with sub-Doppler resolution gave a barrier height of about 1000 cm−1, more than 2300 cm−1 smaller than the result from the early LIF and photofragment excitation spectrum,24 which was attributed to the reassigned T00 value of the Ã 1A″ state. The newest experimental T00 value17 of CFBr (Ã 1A″) was verified by recent high-level MRCI18,20 and CASPT222 calculations. There are no experimental results of CFI for comparison with the calculated results. The barrier height in the present study, calculated at the icMRCI+Q/ANO-RCC level with the consideration of various corrections, is about 900 cm−1 smaller than previous CASPT2/Basis3 result.14 The present study was carried out at a high-level and highly correlated MRCI+Q method using large ANO-RCC basis sets, and with the inclusion of additional corrections (SOC, CV, and ZPE), which have not been considered in the previous study and thus could provide a more reliable barrier height of the Ã 1A″ state of CFI. C. PESs of High Electronic Excited States of CFI Radical. The PESs of 12 electronic states with vertical transition energy (VTE) up to 6 eV of CFI, which correlate to the lowest dissociation limit CF(X2Π) + I(2P3/2), were calculated at the icMRCI+Q/ANO-RCC level. To the best of our knowledge, to date there are no theoretical or experimental studies concerning the electronic states with VTE beyond the Ã 1A″ state. Figures 2−4 present the rigid one-dimensional

Figure 3. Potential energy curves of CFI with respect to the C−F bond; the C−I bond length and the F−C−I angle were fixed at their respective equilibrium values.

Figure 4. Potential energy curves of CFI with respect to the C−I bond; the C−F bond length and the F−C−I angle were fixed at their respective equilibrium values.

Figure 2. Potential energy curves of CFI with respect to the F−C−I angle; the C−F and C−I bond lengths were fixed at their respective equilibrium values.

length that is almost unchanged, the equilibrium bond angles change significantly for different bent states, which could be attributed to different electronic transitions. As all the triatomic halogenated carbenes, the lowest electronic state of CFI at the bent configuration is the singlet 11A′ state, and the S−T gap was calculated to be 8679 cm−1 at the icMRCI/ANO-RCC level with inclusion of ZPE, SOC, and CV corrections. At the linear configuration, however, the lowest electronic state turns out to be the first triplet state. The energy difference between the equilibrium and linear geometry of the Ã 1A″ (i.e., 11A″) state is 9400 cm−1, larger than the calculated dissociation barrier of 624.69 cm−1 as mentioned above, which supports that the photodynamics of the A-state is mainly the dissociation into

potential energy cuts along F−C−I angle, C−F bond, and C−I bond, respectively. In each figure, the other two geometric parameters were fixed at their respective equilibrium values. Table 3 lists our results of the VTE, the electron configuration, the oscillator strength, and the transition of each electronic state of CFI radical. While most of the electronic states of CFI are bent states, several excited states of CFI are considered to be linear (or quasi-linear) states with the global energy minimum at ∼180°, i.e., 13A′, 23A″, 21A′, and 31A′ (Figure 2). Unlike the bond 2450



Article

Table 3. VTE, Oscillator Strength, Electron Configuration, and Transition of Electronic States of CFI Calculated at the icMRCI +Q/ANO-RCC Level state ground state 13A″ 11A″ 23A″ 13A′ 21A″ 21A′ 23A′ 33A′ 33A″ 31A′ 31A″ a

VTE (eV) 0 1.51 2.68 3.46 3.49 3.51 3.90 4.07 5.08 5.27 5.34 5.69

main configuration

excitationa

(9a″) (24a ′) (25a′) (9a″)2(24a′)225a′10a″(9a″)224a′(25a′)210a″ (9a″)2(24a′)225a′10a″(9a″)224a′(25a′)210a″ (9a″)2(24a′)225a′10a″(9a″)224a′(25a′)210a″ (8a″)29a″(24a′-25a′)210a″ (9a″)2(24a′)225a′10a″(9a″)224a′(25a′)210a″ (8a″)29a″(24a′-25a′)210a″ (9a″)2(24a′)225a′26a′ (9a″)224a′(25a′)226a′ (9a″)2(24a′)225a′26a′(9a″)224a′ (25a′)226a′ (8a″)29a″(24a′-25a′)226a′ (9a″)2 (24a′)225a′26a′ (8a″)29a″(24a′-25a′)226a′

25a′ → 10a″(0.225) 24a′ → 10a″(0.651) 25a′ → 10a″(0.325) 24a′ → 10a″(0.584) 25a′ → 10a″(0.706) 24a′ → 10a″(0.226) 9a″ → 10a″(0.880) 25a′ → 10a″(0.608) 24a′ → 10a″(0.324) 9a″ → 10a″(0.775) 25a′ → 26a′(0.297) 24a′ → 26a′(0.534) 25a′ → 26a′(0.631) 24a′ → 26a′(0.306) 9a″ → 26a′(0.921) 25a′ → 26a′(0.869) 9a″ → 26a′(0.932)

oscillator strength 2

0.003027

0.0000413 0.007844

0.014827 0.009427

2

2

The value in parentheses refers to the coefficient of the corresponding configuration.

Finally, the PESs along three coordinates, the vertical transition energies, the electronic configurations, and the available oscillator strengths of total 12 states of CFI were calculated. It is indicated that there exists strong interactions between different states, leading to complicated dynamics of the electronic states, especially high excited states of CFI radical.

CF + I. Unlike other triatomic halogenated carbenes, such as CHBr,39 of which the X̃ 1A′ and the Ã 1A″ state are degenerate at linear configuration leading to Renner−Teller coupling between the states, the Ã 1A″ state of CFI is degenerate with the 21A′ state instead (with energy difference less than 0.002 eV) and is about 0.1 eV higher than the X̃ 1A′ state at linear configuration. As shown in Figure 3, all the calculated electronic states of CFI are bound along the C−F bond, with potential wells more than 1 eV for the excited states. The potential well decreases as the VTE increases. Dissociation barriers along the C−F bond are observed at the PESs of the states with VTE above 4 eV, which could be possible due to the coupling with some repulsive states at high energy region. On the contrary, all C−I potential energy cuts (Figure 4) are purely repulsive leading to I + CF fragments, except for the bound X̃ 1A′ (i.e., 11A′) and ã3A″ (i.e., 13A″) states and the predissociative Ã 1A″ (i.e., 11A″) state. These repulsive excited states should be involved in photodissociation dynamics of CFI radical in the UV region. Particularly, the 21A′ state at 3.90 eV, the 31A′ state at 5.34 eV, and the 31A″ state at 5.69 eV have larger oscillator strength than the Ã 1A″ state, which can be photoexcited at the corresponding wavelengths. However, photodissociation dynamics along these states could be rather complex processes due to unavoidable interactions with the nearby singlet state with low oscillator strength (e.g., 21A″) and spin-forbidden triplet states. Further insight into experimental and theoretical studies are necessary to retrieve the photodissociation dynamics of CFI in the UV region.

■

AUTHOR INFORMATION

Corresponding Authors

*(H.X.) Tel: 86-431-85168817. Fax: 86-431-85168816. E-mail: [email protected]. *(B.Y.) E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by National Basic Research Program of China (973 Program) (2013CB922200) and National Natural Science Foundation of China (11034003, 11074095, and 11274140).

■

REFERENCES

(1) Finlayson-Pitts, B. J.; Pitts, J. N. Atmospheric Chemistry: Fundamentals and Experimental Techniques; Wiley: New York, 1986. (2) Miller, W. T.; Kim, C. S. Y. Reactions of Alkyllithiums with Polyhalides. J. Am. Chem. Soc. 1959, 81, 5008−5009. (3) Kable, S. H.; Reid, S. A.; Sears, T. J. The Halocarbenes: Model Systems for Understanding The Spectroscopy, Dynamics and Chemistry of Carbenes. Int. Rev. Phys. Chem. 2009, 28, 435−480. (4) Cameron, M. R.; Kable, S. H.; Bacskay, G. B. The Electronic Spectroscopy of Jet-cooled Difluorocarbene (CF2): The Missing State Stretching Frequencies. J. Chem. Phys. 1995, 103, 4476−4483. (5) Irikura, K. K.; Hudgens, J. W.; Johnson, R. D. Spectroscopy of The Fluoromethylene Radicals HCF and DCF By 2 + 1 Resonance Enhanced Multiphoton Ionization Spectroscopy and by ab Initio Calculation. J. Chem. Phys. 1995, 103, 1303−1308. (6) Cheong, B.-S.; Cho, H.-G. Ab Initio Studies of Halogenated Methyl and Methylene Radicals: Molecular Structure,Vibrational Frequencies, and Enthalpies of Formation. J. Phys. Chem. A 1997, 101, 7901−7906. (7) Schmidt, W. T.; Bacskay, G. B.; Kable, S. H. Ab initio Potential Energy Surface and Vibrational Frequencies of Ã (1A″) HCF. Chem. Phys. Lett. 1998, 292, 80−86. (8) Knepp, P. T.; Kable, S. H. The Photodissociation Dynamics of CFBr Excited into The Ã (1A″) State. J. Chem. Phys. 1999, 110, 11789−11797.

■

CONCLUSIONS In conclusion, we have carried out all-electron relativistic icMRCI+Q calculations on the ground and excited states of CFI radical. The calculated geometric parameters of the ground X̃ 1A′ state, the first excited singlet Ã 1A″ state, and the lowest triplet ã3A″ state at the icMRCI+Q/ACC-RNO level were in good agreement with previous theoretical results. The Ã 1A″ state is demonstrated to be a predissociative state. The PES along the C−I bond of the Ã 1A″ state was carefully examined at the icMRCI+Q/ACC-RNO level, with optimization of the C− F bond and F−C−I angle at every C−I bond length in contrast to fix them at the equilibrium values. A reliable barrier height of the Ã 1A″ state was determined to be 625 cm−1 in our study using high-level icMRCI+Q method with large ANO-RCC basis set and including the SOC, CV and ZPE corrections. 2451



Article

(9) Schwartz, M.; Marshall, P. An ab Initio Investigation of Halocarbenes. J. Phys. Chem. A 1999, 103, 7900−7906. (10) Schwartz, R. L.; Davico, G. E.; Ramond, T. M.; Lineberger, W. C. Singlet-Triplet Splittings in CX2 (X = F, Cl, Br, I) Dihalocarbenes via Negative Ion Photoelectron Spectroscopy. J. Phys. Chem. A 1999, 103, 8213−8221. (11) Sendt, K.; Schmidt, T. W.; Bacskay, G. B. Quantum Chemical Studies of the Potential Energy Surfaces and Vibrational Frequencies of the X̃ (1A′), ã(3A″) and Ã (1A″) States of CHCl and CFCl. Int. J. Quantum Chem. 2000, 76, 297−305. (12) Guss, J. S.; Votava, O.; Kable, S. H. Electronic Spectroscopy of Jet-Cooled CFCl: Laser-Induced Fluorescence, Dispersed Fluorescence, Lifetimes, and C−Cl Dissociation Barrier. J. Chem. Phys. 2001, 115, 11118−11130. (13) Drake, S. A.; Standard, J. M.; Quandt, R. W. An ab Initio Investigation of the Ground and Excited Electronic State Properties of a Series of Bromine- and Iodine-Containing Singlet Carbenes. J. Phys. Chem. A 2002, 106, 1357−1364. (14) Standard, J. M.; Quandt, R. W. A CASPT2 Investigation of the Ground and First Excited Singlet States of Fluoroiodocarbene. J. Phys. Chem. A 2003, 107, 6877−6881. (15) Fan, H.; Ionescu, I.; Xin, J.; Reid, S. A. Polarization Quantum Beat Spectroscopy of HCF(Ã 1A″). I. 19F and 1H Hyperfine Structure and Zeeman Effect. J. Chem. Phys. 2004, 121, 8869−8873. (16) Fan, H.; Mukarakate, C.; Deselnicu, M.; Tao, C.; Reid, S. A. Dispersed Fluorescence Spectroscopy of Jet-Cooled HCF and DCF:Vibrational Structure of The State. J. Chem. Phys. 2005, 123, 014314. (17) Truscott, S. B.; Elliott, N. L.; Westerna, C. M. A Reanalysis of The Ã (1A″)-X̃ (1A′) Transition of CFBr. J. Chem. Phys. 2009, 130, 234301. (18) Standard, J. M.; Steidl, R. J.; Beecher, M. C.; Quandt, R. W. Multireference Configuration Interaction Study of Bromocarbenes. J. Phys. Chem. A 2011, 115, 1243−1249. (19) Ehara, M.; Oyagi, F.; Abe, Y.; Fukuda, R.; Nakatsuji, H. ExcitedState Geometries and Vibrational Frequencies Studied Using the Analytical Energy Gradients of the Direct Symmetry-Adapted ClusterConfiguration Interaction Method. I. HAX-type Molecules. J. Chem. Phys. 2011, 135, 044316. (20) Sun, E.; Zhang, J.; Li, R.; Sun, Q.; Wei, C.; Xu, H.; Yan, B. Geometries, Vibrational Frequencies and Excitation Energies of a Series of Fluorine-Substituted Carbenes, FCX (X = H, F, Cl, Br and I): a High-level Multireference Configuration Interaction Study. Int. J. Quantum Chem. 2014, 114, 66−73. (21) Schlachta, R.; Lask, G. M.; Bondybey, V. E. Fluorescence Spectroscopy of Supersonically Cooled CFCl. J. Chem. Soc., Faraday Trans. 1991, 87, 2407−2412. (22) Sun, E.; Li, R.; Sun, Q.; Wei, C.; Xu, H.; Yan, B. An ab Initio Investigation of Fluorobromo Carbene. J. Phys. Chem. A 2012, 116, 10435−10440. (23) Hajgato, B.; Nguyen, H. M. T.; Minh, T. N.; Nguyen, M. T. Triplet-Singlet Energy Gaps in Iodo-Carbenes (I-C-X): Remarkable Discrepancy Between Theory and Experiment. Phys. Chem. Chem. Phys. 2000, 2, 5041−5045. (24) Knepp, P. T.; Scalley, C. K.; Bacskay, G. B.; Kable, S. H. Electronic Spectroscopy and Ab Initio Quantum Chemical Study of The Ã (1A″)-X̃ (1A′) Transition of CFBr. J. Chem. Phys. 1998, 109, 2220−2232. (25) Werner, H.-J.; Knowles, P. J. A Second Order Multiconfiguration SCF Procedure with Optimum Convergence. J. Chem. Phys. 1985, 82, 5053−5063. (26) Knowles, P. J.; Werner, H.-J. An Efficient Method for The Evaluation of Coupling Coefficients in Configuration Interaction Calculations. Chem. Phys. Lett. 1988, 145, 514−522. (27) Werner, H.-J.; Knowles, P. J. An Efficient Internally Contracted Multiconfiguration-Reference Configuration Interaction Method. J. Chem. Phys. 1988, 89, 5803−5814.

(28) Langhoff, S. R.; Davidson, E. R. Configuration Interaction Calculations on the Nitrogen Molecule. Int. J. Quantum Chem. 1974, 8, 61−72. (29) Roos, B. O.; Lindh, R.; Malmqvist, P.-Å.; Veryazov, V.; Widmark, P.-O. Main group atoms and dimers studied with a new relativistic ANO basis set. J. Phys. Chem. A 2005, 108, 2851−2858. (30) Wolf, A.; Reiher, M.; Hess, B. A. The Generalized Douglas− Kroll Transformation. J. Chem. Phys. 2002, 117, 9215−9226. (31) Reiher, M.; Wolf, A. Exact Decoupling of The Dirac Hamiltonian. II. The Generalized Douglas−Kroll−Hess Transformation up to Arbitrary Order. J. Chem. Phys. 2004, 121, 10945−10956. (32) Reiher, M.; Wolf, A. Exact Decoupling of The Dirac Hamiltonian. I. General Theory. J. Chem. Phys. 2004, 121, 2037−2047. (33) Berning, A.; Schweizer, M.; Werner, H.-J.; Knowles, P. J.; Palmieri, P. Spin-Orbit Matrix Elements for Internally Contracted Multireference Configuration Interaction Wavefunctions. Mol. Phys. 2000, 98, 1823−1833. (34) Peterson, K. A.; Shepler, B. C.; Figgen, D.; Stoll, H. On the Spectroscopic and Thermochemical Properties of ClO, BrO, IO, and Their Anions. J. Phys. Chem. A 2006, 110, 13877−13833. (35) Dunning, T. H., Jr. Gaussian Basis Sets for Use in Correlated Molecular Calculations. I. The Atoms Boron Through Neon and Hydrogen. J. Chem. Phys. 1989, 90, 1007−1023. (36) Helgaker, T.; Klopper, W.; Koch, H.; Noga, J. Basis-Set Convergence of Correlated Calculations on Water. J. Chem. Phys. 1997, 106, 9639−9646. (37) Werner, H.-J.; Knowles, P. J.; Lindh, R.; Manby, F. R.; Schütz, M.; Celani, P.; Mitrushenkov, A.; Korona, T.; Rauhut, G.; Adler, T. B.MOLPRO, a package of ab initio programs; Cardiff University: Cardiff, U.K., 2009. (38) Fan, H.; Ionescu, I.; Annesley, C.; Cummins, J.; Bowers, M.; Xin, J.; Reid, S. A. On The Renner-Teller Effect and Barriers to Linearity and Dissociation in HCF Ã (1A″). J. Phys. Chem. A 2004, 108, 3732−3738. (39) Bacskay, G. B. Quantum Chemical Characterization of the X̃ (1A′), ã(3A″) and Ã (1A″) States of CHBr and CHI and Computed Heats of Formation for CHI and CI. J. Phys. Chem. A 2010, 114, 8625−8630.

2452


All-Electron Relativistic Multireference Configuration Interaction

Recommend Documents