Through-space 13Cndash;19F coupling can reveal conformations of modified BODIPY dyesdagger; Jiong Chen, Joe Reibenspies, Agnes Derecskei-Kovacs and Kevin Burgess* Department of Chemistry, Texas A amp; M University, PO Box 30012, College Station, TX 77842-3012, USA. E-mail: burgess@chemvx.tamu.edu Received (in Corvallis, OR, USA) 13th September 1999, Accepted 27th October 1999 The fact that only compounds 1a, 1b and 2a in the series 1ndash; 3 show long-range 13Cndash;19F coupling can be used to draw conclusions regarding the structures of these molecules.Through-space coupling between carbon and fluorine atoms that are forced together through geometric constraints is a wellknown phenomenon,1ndash;3 but it is not often used as a tool to gauge molecular conformations.4ndash;6 Here we describe such an application as applied to 3,5-diaryl-4,4-difluoro-4-bora-3a,4a-diaza-sindacene (BODIPYreg; ) dyes 1ndash;3.7 Briefly, the circumstances that led to this study were as follows. Our group is attempting to prepare new fluorescent dyes for biotechnological applications.8,9 An important part of these studies is to refine the fluorescence properties of the target dyes by relating them to molecular structure. As part of our investigations to determine molecular structure we observed that a through-space 13Cndash;19F coupling10 is observed only for compounds 1a, 1b and 2a in this series, and this parameter can be correlated with their molecular conformation.These compounds show a triplet in their 13C{1H} spectra (1a d 130.1, t, J = 11.0 Hz; 1b d 128.4, t, J = 11.2; 2a d = 126.7, t, J = 5.2 Hz).APT and HETCOR experiments11 established that the carbons responsible for these triplets were the ones marked a (henceforth referred to as Ca). A combination of two decoupling experiments was performed on 1a to prove that this splitting was due to coupling with the two fluorine atoms. First, a 13C NMR spectrum was recorded without any decoupling, and the Ca appeared as a doublet of triplets (J = 164 and 11 Hz).A proton-coupled, fluorine-decoupled 13C NMR spectrum of the same sample was then recorded, and the carbon of interest appeared as a doublet (J = 164 Hz). These experiments indicate that the triplet arises from coupling of the Ca to the two fluorine atoms. A comparison of spectral data for compounds 1 and 2 is shown in Table 1. The 19F and 11B NMR chemical shifts for these dyes decrease in the order 1 2a 2b, and the same trend was also observed in the 1H NMR spectra of these materials.The observed Cndash;F coupling constants in this series decrease in the same order. These observations imply that the factors that govern the 19F chemical shifts also impact Cndash;F coupling constants in this series. It is unlikely that Ca is coupled to the fluorine atoms via covalent bonds in the aromatic system, for several reasons. First, if it were to occur, the observed value of 11 Hz would be an exceptionally large J-parameter for a five-bond coupling.Second, other carbons in compounds 1 and 2a would most probably be affected in the same way if that coupling mechanism was operative, especially those that are less than five bonds away. Finally, similar five-bond coupling might then also be observed for compound 2b and 3, but they were not. A single crystal, X-ray diffraction analysis was performed on compound 1a to gain more insight into the nature of the Candash;F interactions.Dagger; Molecule 1a crystallizes in a bow-shaped conformation making the two fluorine atoms inequivalent (even though they are equivalent in the ambient solution phase 19F NMR spectrum of this compound, Fig. 1). Moreover, there are four Candash;F distances since the BF2 entity does not reside at a dagger; Spectral data for 1a,b and 2a,b and colour versions of Fig. 1 and 2 are available from the RSC web site, see http://www.rsc.org/suppdata/cc/ 1999/2501/ Table 1 Selected spectroscopic data for 1ndash;2 Chemical shift (d/ppm)a Compound 1Hb 19Fc 11Bd 1a 8.78 2137.5 9.29 1b 8.82 2137.3 3.98 2ae 8.75 2147.0 mdash; 2b 8.24 2150.6 0.41 a In CDCl3.b For Candash;H or the corresponding Cndash;H, at 300 MHz relative to a deuterium lock. c At 282 MHz relative to CFCl3 as external reference. d At 64 MHz relative to BF3 middot;OEt2 as external reference. e 11B NMR data could not be obtained for this compound since it is relatively insoluble in the common organic solvents used for NMR studies. Fig. 1 Chem3D representations of 1a from X-ray crystallographic data: (a) top view showing Candash;F distances; (b) side view illustrating curvature of the molecule.This journal is copy; The Royal Society of Chemistry 1999 Chem. Commun., 1999, 2501ndash;2502 2501position that is exactly equidistant to both aromatic rings in the crystal structure. The four Candash;F distances are similar; they vary between 2.97(1) and 3.14(1) Aring;, and average to 3.04(8) Aring;. More variance was observed for the Candash;Hmiddot;middot;middot;F distances. Electron density maps were used to locate the protons in this molecule, and their positions were refined (though of course, not to the same degree of accuracy as the heavier atoms).Nevertheless, the shorter Candash;Hmiddot;middot;middot;F distance is no less than 2.29(1) Aring;, the longer one is no more than 2.75(1) Aring;, and the average value was measured as 2.5(2) Aring;. Comparison of the average Cndash;F distance {3.04(8) Aring;} with the sum of the van der Waals radii for carbon and fluorine (3.19 Aring;)12 implies that an interaction between these atoms is likely.Similarly, the average Candash;Hmiddot;middot;middot;F distance {2.5(2) Aring;} is less than the sum of the van der Waals radii for hydrogen and fluorine (2.62 Aring;). It appeared from the solid state structural analyses, outlined above, that the Candash;F interaction observed in the 13C NMR spectra of compounds 1 is due to a close interaction of these atoms, through space. Crystallographic data was not obtained for compounds 2, but a series of calculations was performed to establish why the interaction was observed for 2a but not 2b. Briefly, these were performed in the Cerius2 collection of programs (Molecular Simulations, Inc.).Preliminary optimizations were performed via molecular mechanics using the Universal Force Field (designed to accommodate all the atoms in the periodic table).13215 Final geometry optimizations were obtained at the semi-empirical level using the AM1 method within MOPAC.16,17 Throughout these calculations, the iodine atoms in the real molecules were substituted by chlorine atoms in the virtual compounds.Calculations on the Cl-for-I analog of compound 1a gave a minimized structure obtained having the same curved conformation observed from the X-ray diffraction study (Fig. 1). Moreover, all the bond parameters were similar (data not shown). This result indicated that the AM1 calculations were reliable for this molecular type. Simulated structures of the Cl-for-I analogs of compounds 2 are shown in Fig. 2. Comparison of Fig. 1 and 2 shows that the latter are essentially planar: they do not have the bow-shaped structure that compound 1a has. Distances between the F- and Ca-atoms increase in the order 1a 2a 2b. They are no longer within the sum of the van der Waals radii for carbon and fluorine for compound 2b. These structural studies clarify why the trend on the F-to-Ca coupling constants should be as observed. The ethylene bridges of compound 1a force the benzene rings inwards towards the fluorine atoms, giving an exceptionally close interaction between the Ca and F atoms.Conversely, the S and O atoms in compounds 2 pull the benzene rings outward, giving a relatively open structure. However, this effect is less important for 2a than for 2b for two related reasons that expand the five-membered rings containing the sulfur atoms. First, the atomic size of sulfur is greater than that of oxygen. Second, Cndash;S bond lengths are longer than for Cndash;O bonds (typically 1.75 Aring; for sp2 Cndash;S bond vs. 1.34 Aring; for sp2 Cndash;O bond). It is surprising that couplings between the Candash;Hmiddot;middot;middot;F atoms were not observed for compounds 1 and 2a. It appears that the Candash;Hmiddot;middot;middot;F coupling is too small to be observed conveniently in the proton NMR spectra of these compounds. However, the Candash; H resonance for compound 1a was broader than other aromatic protons in this molecule (2.7 Hz at half-peak-height) and sharpened in a 1H{19F} experiment (2.1 Hz at half-peakheight).There are two possible rationales for the coupling observed between the Candash;F atoms in compounds 1 and 2a. One is a thermodynamically favorable H-bonding interaction between the aromatic hydrogen and the fluorine atoms. Similar interactions have been proposed before.18220 However, we believe that the coupling is really a consequence of crowding within the bay area that contains the BF2 entity in these molecules. Whatever the origin of these effects, itrsquo;s clear from the data presented here that they can be used as a tool to access molecular structure.We thank Ben Lane for running the 11B NMR spectra for this study, and Steve Silber for assistance with the decoupled NMR spectra. Financial support for this work was provided by the NIH (HG01745) and The Robert A. Welch Foundation. Notes and references Dagger; Crystal data for 1a: C34H28BCl2F2IN2O2, M = 743.19, a = 17.646(3), b = 11.457(3), c = 15.829(4) Aring;, b = 104.90(2)deg;, U = 3093(1) A3, T = 193(2) K, space group P21/c (No. 14), Z = 4, m(Mo-Ka) = 1.256 mm21, 5576 absorption correct reflections measured, 5538 unique (Rint = 0.0419) which were used in calculations. The final R(F) and wR(F2) were 0.0880 and 0.1749 respectively (all data). CH2Cl2 was modeled as disordered. CCDC 182/1472. 1 R. H. Contreras, C. G. Giribet, M. A. Natiello, J. Peacute;rez, I. D. Rae and J. A. Weigold, Aust. J. Chem., 1985, 38, 1779. 2 L. C. Hsee and D. J. Sardella, Magn. Reson. Chem., 1990, 28, 688. 3 L. Shimoni, H. L. Carrell, J. P. Glusker and M. M. Coombs, J. Am. Chem. Soc., 1994, 116, 8162. 4 P. Szczecinski and J. Zachara, J. Organomet. Chem., 1993, 447, 241. 5 T. Miyake and Y. Koyama, Carbohydr. Res., 1994, 258, 11. 6 K. Matsubara, A. Oba and Y. Usui, Magn. Reson. Chem., 1998, 36, 761. 7 A preparation of 3 was reported in ref. 8 and syntheses of 1 and 2 will be reported elsewhere. 8 L. H. Thoresen, H. Kim, M. B. Welch, A. Burghart and K. Burgess, Synlett, 1998, 1276. 9 H. Kim, A. Burghart, M. B. Welch, J. Reibenspies and K. Burgess, Chem. Commun., 1999, 1889. 10 F. B. Mallory and C. W. Mallory, Coupling Through Space in Organic Chemistry, in Encyclopedia of Nuclear Magnetic Resonance, Wiley, New York, 1996. 11 A. E. Derome, Modern NMR Techniques for Chemistry Research, Permagon, Oxford, 1987. 12 A. Bondi, Phys. Chem., 1964, 68, 441. 13 A. K. Rappe, C. J. Casewit, K. S. Colwell, W. A. Goddard-III and W. M. Skiff, J. Am. Chem. Soc., 1992, 114, 10 024. 14 L. A. Castonguay and A. K. Rappe, J. Am. Chem. Soc., 1992, 114, 5832. 15 A. K. Rappe, K. S. Colwell and C. J. Casewit, Inorg. Chem., 1993, 32, 3438. 16 M. J. S. Dewar, E. G. Zoebisch, E. F. Healy and J. J. P. Stewart, J. Am. Chem. Soc., 1985, 107, 3902. 17 M. J. S. Dewar, C. Jie and E. G. Zoebisch, Organometallics, 1988, 7, 513. 18 E. J. Corey, J. J. Rohde, A. Fischer and M. D. Azimioara, Tetrahedron Lett., 1997, 38, 33. 19 T. Steiner, Chem. Commun., 1997, 727. 20 A. Mele, B. Vergani, F. Viani, S. V. Meille, A. Farina and P. Bravo, Eur. J. Org. Chem., 1999, 187. 21 F. H. Allen and O. Kennard, Chem. Des. Automat. News, 1993, 8, 31. Communication 9/07559C Fig. 2 Chem3D representations of the Cl-for-I analogs from AM1 calculations: (a) top view of 2a showing Candash;F distances; (b) side view of 2a; (c) top view of 2b showing Candash;F distances; (d) side view of 2b. 2502 Chem. Commun., 1999, 2501ndash;2502
展开▼
