Kinetic and Theoretical Studies of the [3 + 2] Cycloadditions of
Alkynyl Fischer Carbene Complexes with N-Alkyl Nitrones
Ming Lok Yeung,†Wai-Kee Li,†Hui-Jean Liu,‡Yu Wang,‡and Kin Shing Chan*,†
Department of Chemistry, The Chinese University of Hong Kong, Shatin, Hong Kong and Department of Chemistry, National Taiwan University, Taipei, Taiwan, ROC
Received March 2, 1998
The 1,3-dipolar cycloadditions of alkynyl Fischer carbene complexes with nitrones to give 2,3-dihydroisoxazole carbene complexes were found to undergo first-order kinetics both for the nitrones and alkynyl carbene complexes. The effects of metals, substitutents of the nitrones and complexes, and solvents were studied. The rates increased with the more electron rich nitrones and less electron rich metal complexes. Ab initio theoretical calculations supported that the HOMO of nitrones and the LUMO of carbene complexes were the interacting frontier molecular orbitals. Only little dependence of solvent was observed, and therefore a concerted pathway is more likely.
Introduction
The [3 + 2] cycloaddition reactions of 1,3-dipoles have been intensely investigated in the last two decades,1and
their importance in natural product synthesis has been thoroughly established.2 R,β-Unsaturated Fischer
car-bene complexes have recently been found to undergo rapid and highly regioselective [4 + 2]3 and [2 + 2]4
cycloaddition reactions with a variety of dienes and alkenes, respectively. However, there are only a few examples of [3 + 2] cycloaddition of Fischer carbene complexes with 1,3-dipoles.5 It has been shown that the
reaction of phenylethynylcarbene complexes with CH2N2
leads to competing reactions involving the formation of pyrazolylcarbene and N-pyrazolyl complexes in an overall low yield.6 However, with the same 1,3-dipole containing
a TMS group, Me3SiCHN2, phenylethynylcarbene
com-plexes give high yields of pyrazolylcarbene comcom-plexes in a highly regioselective and rate-enhancing manner com-pared to the analogues of organic esters with the carbene fragments replaced by an oxygen atom.7 Kalinin et al.
found that another 1,3-dipole, R,N-diphenyl nitrone,
undergoes [3 + 2] cycloaddition with trimethylsilyleth-ynylcarbene complexes to yield oxazoline carbene com-plexes in low to moderate yields (accompanied with ca. 30% of unreacted carbene complexes).8 Recently, we
reported the chemoselective and regioselective 1,3-dipolar cycloaddition of alkynyl Fischer carbene complexes with R-phenyl N-tert-butyl nitrones (PBN) to give essentially a quantitative yield of 2,3-dihydroisoxazole carbene complexes at room temperature.9
The detailed quantitative mechanistic aspects of this cycloaddition remains unaddressed. Since kinetic studies of cycloaddtions often yield important insights into the mechanism of the reactions with special regard to the concertedness of the reaction, we therefore have under-taken kinetic studies and theoretical calculations on the cycloaddtion reaction of nitrones with alknyl carbene complexes and now report our results here.
Results and Discussion
Kinetic experiments on reaction 1 were carried out spectrophotometrically at constant temperature within 0.2 °C. Reaction 1 was accompanied by characteristic changes in the MLCT bands and exhibited clean isos-bestic points for at least 3 half-lives. Typically, [complex]
) 6.0 × 10-5M, and an at least 10-fold excess of nitrone
was used in order to establish pseudo-first-order
condi-†The Chinese University of Hong Kong. ‡National Taiwan University.
(1) Padwa, A. 1,3 Dipolar Cycloaddition Chemistry; Wiley-Inter-science: New York, 1984; Vols. 1 and 4.
(2) Desimoni, G.; Tacconi, G.; Barco, A.; Pollini, G. P.; Eds. Natural
Products Synthesis Through Pericyclic Reactions; American Chemical
Society: Washington, DC, 1983.
(3) (a) Wulff, W. D.; Yang, D. C. J. Am. Chem. Soc. 1983, 105, 6746. (b) Wulff, W. D.; Yang, D. C. J. Am. Chem. Soc. 1984, 106, 7565. (c) Dotz, K. H.; Kuhn, W. J. Organomet. Chem. 1985, 486, C43. (d) Wulff, W. D.; Tang, P. C.; Chan, K. S. McCallum, J. S.; Yang, D. C.; Gilbertson, S. R. Tetrahedron 1985, 41, 5813. (e) Merlic, C. A.; Xu, D. J. Am. Chem.
Soc. 1991, 113, 7418. (f) Bao, J.; Dragisich, V.; Wenglowsky, S.; Wulff,
W. D. J. Am. Chem. Soc. 1991, 113, 9873.
(4) (a) Faron, K. L.; Wulff, W. D. J. Am. Chem. Soc. 1988, 110, 8747. (b) Camps, F.; Moreto, J. M.; Ricart, S.; Vinas, J. M.; Molins, E.; Miravitlles, C. J. Chem. Soc., Chem. Commun. 1989, 1560. (c) Faron, K. L.; Wulff, W. D. J. Am. Chem. Soc. 1990, 114, 6419. (d) Camps, F.; Llebaria, M. G.; Moreto, J. M.; Ricart, S.; Vinas, J. M. Tetrahedron
Lett. 1990, 31, 4479. (e) de Meijere, A.; Wessjohann, L. Synlett. 1990,
40. (f) Merlic, C. A.; Xu. D. J. Am. Chem. Soc. 1991, 113, 7418. (g) Pipoh, R.; van Eldik, R.; Wulff, W. D.; Wang, S. L. B. Organometallics
1994, 11, 490.
(5) For leading reference: Barluenga, J.; Fernandez-Mari, F.; Argimiro,.L. V.; Aguilar, E.; Olano, B. J. Chem. Soc., Perkin Trans. 1
1997, 2267.
(6) Kreissl, F. R.; Fischer, E. O.; Kreiter, C. G. J. Organomet. Chem.
1973, 57, C9.
(7) Chan, K. S.; Wulff, W. D. J. Am. Chem. Soc. 1986, 108, 5229.
7670 J. Org. Chem. 1998, 63, 7670-7673
10.1021/jo9803840 CCC: $15.00 © 1998 American Chemical Society Published on Web 10/08/1998
Downloaded by NATIONAL TAIWAN UNIV on August 20, 2009
tions. Wavelengths where maximum absorbance changes (λmax) occurred (λmaxfor 1 ) 502 nm andλmaxfor 2 and 3
) 486 nm) were monitored for at least 3-half-lives. Then the collected kinetic data were treated with Guggen-heim’s method.10 By plotting ln(A - At +
∆) vs t (where A is absorbance at time t, At +∆is the absorbance at time t +∆, and ∆ is a period two to three times as great as the half-life period of the reaction), a linear pseudo-first-order plot with slope equal to kobswas obtained for the
reaction (e.g., reaction of 2 with 4d), and hence the reaction was first-order with respect to complex (Figure 1).
Experiments were repeated at several nitrone concen-trations, and from the first-order plots, the pseudo-first-order rate constants were then determined. These data are shown in Table 1. The rate equation is now, rate ) kobs[complex], kobs) k[nitrone]n. The third column
of Table 1 lists the quotient kobs/[nitrone]initial, whose
constancy shows that n ) 1. Therefore, the cycloaddition is a bimolecular one, and the rate equation may be written as:
Similarly, the rate constants for other reactions were measured.
The activation parameters were determined by the temperature dependence of rate constants using the Erying plot encompassing the temperature range from 14.2 to 44.2 °C, and the data are listed in Table 2. The small values of enthalpies,∆Hq(between 19 and 34 kJ
mol-1) and large negative activation entropies,∆Sq(-146
to -174 J K-1mol-1) (Table 2) indicate the reaction
between a carbene complex and a nitrone goes through a highly structured or polar transition state.
Substituent Effect on Nitrone. The data shown in
Table 3 demonstrates that the reactions were accelerated by the electron rich substituents in nitrones. This observation can be explained according to the Frontier Molecular Orbital Theory (FMO).1 The electron rich
nitrone raises the HOMO energy level, decreases the energy gap of the cycloaddition between that of nitrone and carbene complex, and consequently accelerates the reaction. A Hammett linear free energy treatment of the data (Table 3) gives a F value of -3.15 (coefficient of correlation, R ) 0.9191) using σpo constants and a F+
value of -1.19 (Figure 2, R ) 0.9839) usingσp+constant
(Figure 2). It was found that R forσp+plot was larger
than that ofσpo plot. It implies that the reaction may
involve an enhanced resonance transition state. The magnitude of F+obtained from the Hammett plot (-1.19) is quite different from those observed for typical ionic processes (-3.3 to -4.3)12and is comparable with other
concerted [3 + 2] cycloaddition.13
Substituent Effect on Carbene Complex. The
substituent effect on the carbene complex was briefly investigated. The methyl-substituted W complex 3 (k ) 1.18× 10-2 M-1s-1)was found about 50% slower than
the unsubstituted W complex 2 (k ) 2.27 × 10-2 M-1 s-1)in the reactions with 4d (Table 2). The electron-donating methyl group may raise the LUMO energy level of the complex to enlarge the energy gap of cycloaddition and therefore slow down the reaction.
Solvent Effect. Reaction of 2 with 4d was examined
in several solvents (Table 2). The observed rate constants did not exhibit a large significant solvent dependence, at the most a factor of 4. This supports that there is not a larger charge separation in the activated complex than in the reactants, so the reaction probably goes through one-step, polar, concerted pathway as suggested in the Hammett plot of the substitutent effect of nitrones.
Metal Effect on Carbene Complex. The data in
Table 1 demonstrate that the reactivity of carbene complexes 2 (22.7× 10-2M-1s-1)is 2.9 times faster than 1 (7.7 × 10-2M-1 s-1) at 25 °C. This metal effect has also been observed in the [2 + 2] cycloaddition of alkynyl carbene complex with dihydropyran with the rate of W complex faster than that of Cr by 1.6 times.13
Theoretical Calculations. The HOMO and LUMO
energies of substituted R-phenyl N-methyl nitrones (PMN) as substitutes for PBN and methyl propiolate were calculated by ab initio method with basis set 3-21G using Gaussian 90 program14on an IBM-RS6000 workstation
(Table 4). It should be noted that kinetic studies of cycloaddition of tungsten complex 2 with PMN were not carried out since they were not clean enough to observe any isosbestic point, probably due to the fast decomposi-tion of the thermal labile cycloadduct.9b The HOMO and
LUMO energies of PMN increase systematically except for that of the dimethylamino group, which shows a
(8) Kalinin, V. N.; Shilova, O. S.; Kovredov, A. I.; Petrovskii, P. V.; Batsanov, A. S.; Struchkov, Y. T. Organomet. Chem. USSR 1989, 4 (3), 468.
(9) (a) Chan, K. S. J. Chem. Soc., Perkin Trans. 1 1991, 2602. (b) Chan, K. S. Yeung, M. L.; Chan, W.-K.; Wang, R.-J.; Mak, T. C. W. J.
Org. Chem. 1995, 60, 1741.
(10) Frost, A.; Pearson R. G. Kinetics and Mechanism; John Wiley & Sons: New York, 1961; p 49.
(11) Exner, O. Chapter 10 of Prog. Phys. Org. Chem. 1973, 10, 1. (12) Baldwin J. E.; Kapecki, J. A. J. Am. Chem. Soc. 1970, 92, 8 and references there in.
(13) Pipoh, R.; van Eldik, R.; Wulff, W. D.; Wang, S. L. B.
Organo-metallics 1992, 11, 490.
(14) Frisch, M. J.; Head-Gordon, M.; Trucks, G. W.; Foresman, J. B.; Schelgel, H. B.; Raghavachari, K.; Robb, M. A.; Binkley, J. S.; Gonzalez, C.; DeFrees, D. J.; Fox, D. J.; Whiteside, R. A.; Seeger, R.; Melius, C. F.; Baker, J.; Martin, R. L.; Kahn, L. R.; Stewart, J. J. P.; Topiol, S.; Pople, J. A. GAUSSIAN 90, Gaussian Inc., Pittsburgh, PA, 1990.
Figure 1. First-order plot of reaction of 2 with 4d in acteone at 34.2 °C.
Table 1. Determination of Reaction Order from Pseudo-First-Order Rate Constants using 2 with 4da
[4d]initial(10-4M) kobs(10-4s-1) k (M-1s-1) 6.3 1.33 0.21 12.0 2.86 0.24 30.0 6.93 0.23 60.0 13.70 0.23 a[2] initial) 6.0 × 10-5M in THF at 24.7 °C. rate ) k[complex][nitrone]
[3 + 2] Cycloadditions of Alkynyl Fischer Carbene Complexes J. Org. Chem., Vol. 63, No. 22, 1998 7671
Downloaded by NATIONAL TAIWAN UNIV on August 20, 2009
deviation from the linear correlation of Hammett con-stants. Since the rate of the cycloaddition increases with electron-donating nitrones, it is reasonable to conclude that the HOMO of the nitrones are the frontier orbitals. Scheme 1 demonstrates the results of theoretical calculations15,16of the three given compounds and the
cycloaddition processes of the chromium carbene com-plex16and its organic ester analogue15with the N-methyl
nitrone (PMN) in an attempt to understand the cycload-dition. Both of the dipolarophiles adapt their LUMO to
(15) Liu, H. J. M. Philos. Doctoral, Deformation Electron Density Study of Ficher Carbene Complex. National Taiwan University, 1991. (16) Wang, C.-C.; Wang, Y.; Liu, H.-J.; Lin, K.-J.; Chou, L.-K.; Chan, K. S. J. Phys. Chem. A 1997, 101 (47), 8887.
Table 2. Rate Constants and Activation Parameters of Cyclcoaddition Reactions
complex X-PBN temp (°C) solvent k (10-4M-1s-1) ∆Hq (kJ mol-1) ∆Sq (J mol-1K-1) ∆Gq 498 (kJ mol-1) 2 Me2N (4a) 14.2 THF 93700 19.5 -158.2 66.6 24.7 137000 34.2 154000 43.4 231000 MeO (4b) 14.8 6340 26.9 -154.7 73.0 24.7 10100 34.7 14400 44.2 20100 Me (4c) 14.7 2480 31.9 -145.1 75.1 24.7 4370 34.7 6340 44.2 9700 H (4d) 14.7 1370 31.2 -152.5 76.6 24.7 2270 34.9 3640 44.2 5030 Br (4e) 15.1 390 30.0 -167.3 79.9 24.7 673 34.0 960 43.7 1320 Cl (4f) 15.1 394 31.5 -161.8 79.7 24.7 684 34.0 969 43.5 1420 NO2(4g) 15.0 34 33.9 -174.3 85.8 24.7 57 33.8 88 43.5 123 H (4d) 15.2 hexane 1180 30.4 -157.1 77.2 24.7 1810 33.9 2660 43.5 4040 H (4d) 15.2 acetone 2210 31.8 -146.8 75.5 44.7 3510 34.4 5380 43.4 7930 H (4d) 15.3 CH3CN 1880 34.3 -146.5 76.0 24.7 3080 34.0 4460 43.7 7040 1 H (4d) 15.1 THF 499 32.8 -156.0 79.3 24.7 770 34.0 1220 43.7 1870 3 H (4d) 15.2 THF 726 31.8 -156.1 78.3 24.8 1180 33.8 1770 43.2 2570
Table 3. Substitutent Effect on Nitrone (reaction of 2 with 4a-g) nitrone X σpo a σp+ a k (10-3M-1s-1) 4a Me2N -0.32 -1.7 13700 4b MeO -0.12 -0.78 1010 4c Me -0.14 -0.3 440 4d H 0 0 230 4e Br 0.26 0.15 67 4f Cl 0.34 0.11 68 4g NO2 0.81 - 6 aFrom ref 11.
Figure 2. Plot of log(kX/kH) verseσp+for reaction of 2 with
substituted PBN.
7672 J. Org. Chem., Vol. 63, No. 22, 1998 Yeung et al.
Downloaded by NATIONAL TAIWAN UNIV on August 20, 2009
interact with HOMO of PMN, as the energy differences are smaller than the LUMO of the dipoles and the HOMO of the dipolarphiles. The energy gap between the carbene complex and PMN is 9.01 eV, while that between the organic ester and PMN is 10.80 eV. The energy gap for
carbene complex is 1.71 eV (equivalent to 154 kJ mol-1) smaller than that for the organic ester analogue. The smaller energy gap of cycloaddition of carbene complex with PMN supports its rate enhancement over the organic ester analogue.3,7,9
Conclusion
Alkynyl Fischer carbene complexes have been demon-strated to undergo polar, concerted [3 + 2] cycloaddition with N-alkyl nitrones to give 2,3-dihydroisoxazole car-bene complexes. The reactions have been found to be controlled by HOMO (nitrone)-LUMO (carbene complex) interaction both theoretically and experimentally.
Experimental Section
Unless otherwise noted, all materials were obtained from commercial suppliers and used without further purification. Kinetic measurements were performed on a spectrophotometer equipped with a temperature controller. The temperature was measured with a digital thermometer ((0.1 °C) with type K thermal couple wire. The solvents used, hexane (CaH2), THF
(Na/benzophenone ketyl), acetone (K2CO3) and acetonitrile
(CaH2), were distilled prior to use. All the solvents and stock
solutions were deoxygenated by the freeze-pump-thaw method (-195 to 45 °C, at least five cycles). Reactants were trans-ferred inside a drybox to a Schlenk type UV-cuvette. The progress of reaction was monitored spectrophotometrically. The rate constants were calculated using the computer pro-gram KaleidaGraph to find out the slopes of best linear straight lines. The rate constants and activation parameters were within 10% error.
Acknowledgment. We thank the Research Grants Council of Hong Kong (CUHK 88/93E) for financial support.
JO9803840 Table 4. HOMO and LUMO Energies of PMN,
Calculated by ab Initio Method (Basis Set: 3-21G) Using Gaussian 90 Program
X-PMN σp+ a HOMO (eV) LUMO (eV) Me2N (6a) -1.70 -8.073 2.339 MeO (6b) -0.78 -7.821 2.505 Me (6c) -0.30 -8.025 2.360 H (6d) 0.00 -8.198 2.275 Br (6e) 0.15 -8.347 1.968 Cl (6f) 0.11 -8.514 1.883 aFrom ref 11.
Scheme 1. Frontier Orbital Energies Relationships of Reactants
[3 + 2] Cycloadditions of Alkynyl Fischer Carbene Complexes J. Org. Chem., Vol. 63, No. 22, 1998 7673
Downloaded by NATIONAL TAIWAN UNIV on August 20, 2009
