Mechanistic Analysis of an Isoxazole–Oxazole Photoisomerization

Aug 31, 2015 - The mechanisms of the three reaction pathways for the photochemical transformation of 3,5-dimethylisoxazole (1) in its first singlet ex...
0 downloads 5 Views 613KB Size
Download PDF
Article pubs.acs.org/JPCA

Mechanistic Analysis of an Isoxazole−Oxazole Photoisomerization Reaction Using a Conical Intersection Ming-Der Su* Department of Applied Chemistry, National Chiayi University, Chiayi 60004, Taiwan Department of Medicinal and Applied Chemistry, Kaohsiung Medical University, Kaohsiung 80708, Taiwan

Downloaded by UNIV OF MANITOBA on September 7, 2015 | http://pubs.acs.org Publication Date (Web): September 3, 2015 | doi: 10.1021/acs.jpca.5b07312

S Supporting Information *

ABSTRACT: The mechanisms of the three reaction pathways for the photochemical transformation of 3,5-dimethylisoxazole (1) in its first singlet excited state (π→ π*)1 have been determined using the CASSCF (11-orbital/14-electron active space) and MP2-CAS methods with the 6-311G(d) basis set. These three reaction pathways are denoted as (i) the internal cyclization-isomerization path (path A), (ii) the ring contraction-ring expansion path (path B), and (iii) the direct path (path C). This work provides the first theoretical examinations of mechanisms for such photochemical rearrangements. The present theoretical findings suggest that the photoisomerization of 1 via path C should be much more favorable then either path A or path B. Nevertheless, the theoretical observations reveal that path B, which consists of a sequence of small geometric rearrangements, should be energetically feasible as well. Accordingly, the fleeting intermediate, acetyl nitrile ylide (4), which arises from the mechanism of path B, can be detected experimentally.

I. INTRODUCTION The photochemistry of the isomerization of five-membered heterocyclic molecules has been the subject of intense research for many decades.1−5 Traditionally, five mechanisms have been proposed to rationalize the experimental findings of such photochemical transformations: (1) the internal cyclization− isomerization mechanism; (2) the ring contraction−ring expansion mechanism; (3) the van Tamelen−Whitesides mechanism; (4) the zwitterion−tricycle mechanism; (5) the fragmentation−readdition mechanism (Scheme S1, Supporting Information).1 Nevertheless, these proffered mechanisms still retain many illogical problems, because these earlier mechanisms did not consider the photoreaction pathways from the excited state to the ground state. As a result, they cannot offer a good explanation of the available experimental observations.1−5 In recent years, the author has suggested an alternative mechanism, which is called (6) the direct mechanism (Supporting Information). The direct mechanism can profitably interpret the photochemical isomerization of organic heterocyclic compounds6−13 because this mechanism gives a proper answer that enables clarification of an efficient and barrier-free pathway via the conical intersection (CI) point to the final photoproducts.14−19 Recently, Nunes, Reva, and Fausto reported that the photolysis of 3,5-dimethylisoxazole (1) at λ = 222 nm yields two kinds of molecules, recognized as acetyl nitrile ylide (4) and 2,5-dimethyloxazole (5).20 In particular, they emphasized that species 4, which had been presumably considered as a fleeting molecule, was captured and characterized for the first time as an intermediate in the isoxazole−oxazole photoisomerization.20 As a result, they inferred a mechanism (Scheme 1) to describe the experimental results they found, which is quite similar to the so-called “ring contraction−ring expansion mechanism” shown in Scheme S1. However, the © XXXX American Chemical Society

mechanism they presented cannot powerfully explain the experimental findings they observed, because it did not reflect the photoreaction process. To obtain a better understanding of the photochemical behavior of 1, we have thus undertaken an investigation into the singlet excited state potential energy surfaces of 1 using the complete-active-space SCF (CASSCF) and the MP2-CAS levels of theory, which is implanted in the Gaussian 09 software package (Supporting Information).21 In this study, we focus on the mechanistic analysis of the photoformations of intermediate 4 and photoproduct 5.

II. GENERAL CONSIDERATIONS A basic pattern of the five p−π orbitals in 1 is schematically illustrated in Scheme 2, developing the foundation for the present study. As seen in Scheme 2, the lowest singlet π → π* excitation is the singlet π3 → π4* transition. The experimental data for 1 (near its maximum absorption) has been reported to be λ = 222 nm (=129 kcal/mol).20 Agreement between the reported experimental value and our computational data (127 kcal/mol) was good. It is thus believed that the reliability of this model computation should be dependable for the subsequent examination of the mechanisms of the photochemical rearrangement reactions of 1. On the basis of the p−π orbital model shown in Scheme 2, one may then utilize this MO model to search for the CI of “the direct mechanism” of 1 (Scheme S1). Figure 1 presents the qualitative potential energy curves for the S0 and S1 states of 1 as a function of rotation angle θ. As a result, twisting the NC π bond in 1 can lead to degeneration of the S0 and S1 energy Received: July 29, 2015 Revised: August 27, 2015

A

DOI: 10.1021/acs.jpca.5b07312 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A Scheme 1. Proposed Mechanism for the Photoisomerization Reaction of 1

surfaces at 95°. This strongly implies there exists a CI point nearby from which the excited state reactant is through the radiationless transition (i.e., conical intersection) before moving on the ground-state surface toward both reactant 1 and photoproduct 5.3

Scheme 2. Five p−π Orbitals of 1

Downloaded by UNIV OF MANITOBA on September 7, 2015 | http://pubs.acs.org Publication Date (Web): September 3, 2015 | doi: 10.1021/acs.jpca.5b07312

III. RESULTS AND DISCUSSION According to Scheme S1, as well as the available experimental results,20 three reaction routes were considered regarding the irradiation of 1, i.e., path A (the internal cyclization− isomerization path), path B (the ring contraction−ring expansion path), and path C (the direct path). These all lead to the same photoproduct, 5. Figure 2 displays all the relative energies of the various points with respect to the energy of 1. The structures of these points on the possible mechanistic pathways of Figure 2 are given in the Supporting Information. The reactant 1 is first irradiated to the first excited singlet state (FC) due to the Franck−Condon effect, as illustrated in the center of Figure 2. For path A, the computational results indicate that the mechanism for this route can be represented as follows: Path A: 1(S0) + hν → FC → CI‐A → Int‐1 → TS‐1 → Int‐2 → TS‐2 → Pro

However, according to the MP2-CAS results, it was found that the energy of TS-1 lies about 3.0 kcal/mol above that of FC. As a result, because the singlet excited energy of 1 (127 kcal/mol) is insufficient to surmount that of TS-1, this result demonstrates that it is unlikely that reactant 1 pursues the mechanism of path A (i.e., the internal cyclization−isomerization mechanism). On the contrary, after irradiating 1 to FC, the first step for 1 in path B is the formation of the intermediate, Int-3, on the singlet excited state. The transition state, TS-3, which connects Int-3 and CI-B, is computed to be 18 kcal/mol higher in energy than Int-3. After decay from the CI-B, reactant 1 can arrive at the intermediate, Int-4 (vinyl nitrene, 2) and then easily reach the other intermediate, Int-5 (2H-azirine, 3) with a three-membered ring containing a nitrogen atom. The MP2CAS results reveal that from Int-5, C−C bond breaking (i.e., TS-4) takes place, which results in two intermediates (Int-anti and Int-syn), with the activation energy of 37 kcal/mol. The MP2-CAS calculated data reveal that the energy difference between both intermediates is approximately 2.0 kcal/mol. Nevertheless, searching for the transition state between Int-anti and Int-syn has always failed using the present theoretical methods.22 It is noteworthy that 1 has an overall energy of 80 kcal/mol from FC to IC-B, which is much larger than the energy difference between Int-5 and TS-4. On the basis of the above theoretical data, one may see that this barrier can be readily surmounted so as to attain the final product, Pro, with the exothermicity of 24 kcal/mol. As a consequence, the mechanism for path B is given below:

Figure 1. Minimum-energy pathway of photochemistry of 1 along the torsion angle coordinate optimized for the S1 state at the CAS(14,11)/ 6-311G(d) level of theory.

B

DOI: 10.1021/acs.jpca.5b07312 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

Downloaded by UNIV OF MANITOBA on September 7, 2015 | http://pubs.acs.org Publication Date (Web): September 3, 2015 | doi: 10.1021/acs.jpca.5b07312

The Journal of Physical Chemistry A

Figure 2. Energy profiles for the photoisomerization modes of 1. The abbreviations FC and CI stand for Frank−Condon and conical intersection, respectively. The relative energies were obtained at the CAS(14,11)/6-311G(d) (in parentheses) and MP2-CAS(14,11)/6-311G(d)//CAS(14,11)/ 6-311G(d) levels of theory. All energies (in kcal/mol) are given with respect to 1. For more information see the text.

C−N ring bond via a conical intersection. Through this CI point, photoproduct 5 as well as the initial reactant 1 can be found on a ground-state relaxation path. That is to say, path C is a one-step and barrierless photochemical process. Nevertheless, the theoretical investigations reveal that the mechanism of path B cannot be ruled out during the photochemical transformations of 1, due to the reason that, in this case, path B was theoretically proved to be an energetically acceptable photochemical pathway. This could be the reason that the elusory intermediate 4 can be caught experimentally.20 Although the mechanism of path B is a feasible photochemical process from the energetic viewpoints, it consists of multiple steps, which involve small geometric rearrangements. This, in turn, makes path B less efficient than path C from a kinetic viewpoint. In other words, the mechanisms of the photopermutation of 1 should be included in both path B and path C. However, the latter is more efficient than the first owing to the kinetics. In consequence, one may then foresee that the quantum yield of photoproduct (5) through the mechanism of path C should be much larger than through that of path B. It is hoped that the present work can stimulate further research into this subject.

Path B: 1(S0) + hν → FC → Int‐3 → TS‐3 → CI‐B → Int‐4 → Int‐5 → TS‐4 → Int‐anti + Int‐syn → Pro

In path C, once the S0 → S1 vertical excitation happens, the reactant 1 relaxes from FC to the conical intersection CI-C, which is about 9.0 kcal/mol lower in energy. In reality, by following the gradient difference vector of CI-C and rotating a C−N bond (Supporting Information), 1 arrives at photoproduct 5 without any difficulty. Accordingly, the process for path C is given as follows: Path C: 1(S0) + hν → FC → CI‐C → Pro

IV. CONCLUSION The photochemical reaction mechanisms of 1 were investigated by utilizing three different kinds of reaction pathways (path A, path B, and path C). Considering both energetic and kinetic viewpoints, the theoretical findings suggest that the direct mechanism (path C) is more favorable than either path A or path B. Specifically, once 1 absorbs a photon to reach an excited singlet state through a 1π → 1π* transition, then 1 permutes the C

DOI: 10.1021/acs.jpca.5b07312 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

■

(14) Klessinger, M. Conical Intersections and the Mechanism of Singlet Photoreactions. Angew. Chem., Int. Ed. Engl. 1995, 34, 549− 551. (15) Bernardi, F.; Olivucci, M.; Robb, M. A. Modelling Photochemical Reactivity of Organic Systems  A New Challenge to Quantum Computational Chemistry. Isr. J. Chem. 1993, 33, 265−276. (16) Bernardi, F.; Olivucci, M.; Robb, M. A. The Role of Conical Intersections and Excited State Reaction Paths in Photochemical Pericyclic Reactions. J. Photochem. Photobiol., A 1997, 105, 365−371. (17) Bernardi, F.; Olivucci, M.; Robb, M. A. Potential Energy Surface Crossings in Organic Photochemistry. Chem. Soc. Rev. 1996, 25, 321− 328. (18) Klessinger, M. Theoretical Models for the Selectivity of Organic Singlet and Triplet Photoreactions. Pure Appl. Chem. 1997, 69, 773− 778. (19) Klessinger, M.; Michl, J. Excited States and Photochemistry of Organic Molecules; VCH Publishers: New York, 1995. (20) Nunes, C. M.; Reva, I.; Fausto, R. Capture of an Elusive Nitrile Ylide as an Intermediate in Isoxazole-Oxazole Photoisomerization. J. Org. Chem. 2013, 78, 10657−10665. (21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09; Gaussian, Inc.: Wallingford, CT, 2013. (22) However, using the B3LYP/6-311++G(d,p) method, Nunes, Reva, and Fausto estimated the barrier between Int-anti and Int-syn is about 13 kcal/mol. Moreover, it has to be mentioned here that structure 4 from Scheme 1 is Int-anti in Figure 2. See ref 20.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.5b07312. Photochemical mechanisms, orbital structure, optimized geometries and the relative energies for the ground states and first singlet excited states of the various points for 3,5-dimethylisoxazole at the CASSCF(14,11)/6-311G(d) and MP2-CAS(14,11)/6-311G(d)//CASSCF(14,11)/6311G(d) levels of theory, and computational methods (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +886-05-2717966. Fax: +886-05-2717901. E-mail: [email protected]. Downloaded by UNIV OF MANITOBA on September 7, 2015 | http://pubs.acs.org Publication Date (Web): September 3, 2015 | doi: 10.1021/acs.jpca.5b07312

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The author is grateful to the National Center for HighPerformance Computing of Taiwan for generous amounts of computing time, and the Ministry of Science and Technology of Taiwan for the financial support. The author also thanks Professor Michael A. Robb, Dr. S. Wilsey, Dr. Michael J. Bearpark (University of London, U.K.), and Professor Massimo Olivucci (Universita degli Sstudi di Siena, Italy) for their encouragement and support during his stay in London. Special thanks are also due to reviewers 1 and 2 for very help suggestions and comments.

■

REFERENCES

(1) Lablache-Combier, A. In Photochemistry of Heterocyclic Compounds; Buchardt, O., Ed.; Wiley: New York, 1976. (2) Padwa, A. In Rearrangements in Ground and Excited States; De Mayo, P., Ed.; Academic Press: New York, 1980; Vol. 3, p 501. (3) Buchardt, O. Photochemistry of Heterocyclic Compounds; Wiley: New York, 1976. (4) Kopecky, J. Organic Photochemistry: a Visual Approach; VCH Publishers: New York, 1992. (5) Lablache-Combier, A. In CRC Handbook of Organic Photochemistry and Photobiology; Horspool, W., Ed.; CRC: Boca Raton, FL, 1995; p 1063. (6) Su, M.-D. CASCSF Study on the Photochemical Tansposition Reactions of Pyrazines. J. Phys. Chem. A 2006, 110, 9420−9428. (7) Su, M.-D. Photochemical Isomerization Reactions of Cyanopyrroles: a Theoretical Study. J. Phys. Chem. A 2006, 110, 12653− 12661. (8) Su, M.-D. The Photochemical Rearrangement Pathways of Imidazoles: a Theoretical Study. J. Phys. Chem. A 2007, 111, 1567− 1574. (9) Su, M.-D. A Theoretical Insight Into the Reaction Mechanism of Photochemical Transposition From Pyrazole to Imidazole. J. Phys. Chem. A 2008, 112, 10420−10428. (10) Su, M.-D. Theoretical Investigations of the Photochemical Isomerizations of Indoxazene and Isoxazole. J. Org. Chem. 2009, 74, 6055−6063. (11) Su, M.-D. A Computational Study of Photochemical Isomerization Reactions of Thiophenes. J. Comput. Chem. 2010, 31, 43−56. (12) Su, M.-D. Model Study of the Photochemical Rearrangement Pathways of 1,2,4-Oxadiazole. ChemPhysChem 2014, 15, 2712−2722. (13) Su, M.-D. A Model Study on the Photochemical Isomerization of Isothiazoles and Thiazoles. Phys. Chem. Chem. Phys. 2014, 16, 17030−17042. D

DOI: 10.1021/acs.jpca.5b07312 J. Phys. Chem. A XXXX, XXX, XXX−XXX