NMR Studies on [2 + 3] Cycloaddition of Nitrile Oxides to Polyunsaturated Medium Size Rings

Abstract

Site selectivity, regioselectivity and stereoselectivity of [2 + 3] cycloaddition of 4-trifluoromethylbenzonitrile oxide to polyunsaturated medium size rings including 1,5,9-cyclododecatriene, 11-membered sesquiterpenes, 1,3-cyclooctadiene and 5-vinyl-2-norbornene were examined. Site selectivity was correlated with electron charges of alkenyl carbon atoms. Structure of the products has been established by an extensive application of 1D and 2D ¹H and ¹³C nuclear magnetic resonance (NMR) spectroscopy and electrospray ionization mass spectrometry. Some of the obtained products showed moderate fungicidal activities.

Keywords:
[2 + 3] dipolar cycloaddition; sesquiterpenes; site selectivity; 2D NMR; fungicides

Introduction

Isoxazolines are one of major classes of five-membered nitrogen containing heterocycles, found in a large number of natural products and biologically active compounds. A variety of synthetic methods has been elaborated for preparation of 2-isoxazolines, of which the most convenient and attractive route is the [2 + 3] dipolar cycloaddition of nitrile oxides to alkenes.¹ 2-Isoxazolines can be easily reduced to several synthetically important compounds such as β-hydroxy ketones, β-hydroxy esters, α,β-unsaturated carbonyl compounds or iminoketones.²

The nitrile oxides can be formed either by Huisgen method from aldoximes by chlorination and base-induced dehydrochlorination¹ or by dehydration of primary nitro compounds by phenyl isocyanates (Mukayama method)³ or ethyl chloroformate (Shimizu method).⁴

A key feature of the cycloaddition is the cis-stereospecificity - from E-alkenes 4,5-anti isomers are produced and from Z-alkenes 4,5-syn products are obtained.⁵ Reactions of monosubstituted and 1,1-disubstituted alkenes are very regioselective favoring strongly 5-substituted 2-isoxazolines. On the other hand, 1,2-disubstituted olefins usually afford mixtures of regio- and stereoisomers. Two methods have been used to solve these problems. One approach was a substrate control, an application of appropriately functionalized cycloaddends. The second more effective approach relied on metal complexes acting as catalysts or on organocatalysts. Shortage of reports on metal assisted 1,3-dipolar cycloadditions of nitrile oxides was due to interference of catalyst with generation of these dipoles and formation of unreactive complexes.⁶

Site-selectivity of nitrile oxides cycloaddition to polyunsaturated alkenes was examined in several laboratories. In reactions of nitrile oxides with dimethyl 7-(diphenylmethylene)bicyclo[2.2.1]hept-2-ene-5,6-dicarboxylate only disubstituted norbornene double bond participated.⁷ Similarly, in recently examined cycloadditions of aryl nitrile oxides to norbornenes substituted with an acrylate-derived moiety, only adducts to norbornene system were formed with good site and exo selectivity.⁸ Site selectivity and regioselectivity of cycloaddition to cyclohexene derivatives was studied in our group.⁹

2,6-Dichlorobenzonitrile oxide reacts with isothiazolones at the ethylenic double bond. Mesitonitrile oxide, on the other hand, adds preferentially to the carbonyl double bond.¹⁰

No [2 + 3] cycloaddition reactions of nitrile oxides to medium ring cycloalkenes was reported before, to the best of our knowledge, presumably because of the low reactivity of these dipolarophiles. However, other types of cycloadditions to these systems are known. Diels-Alder cycloaddition of benzotropolone with humulene (11-membered sesquiterpene) was described.¹¹ Acetylacetonatoboron difluoride and oxalate undergo cycloaddition from their singlet excited state with cyclic olefins to give non-regiospecific products. A conjugated diene, 1,3-cyclooctadiene, reacted similarly but with slower rates.¹² Singlet excited dibenzoyl(methanato)boron difluoride underwent also cycloaddition to 1,3-cyclooctadiene.¹³ Reaction of SO with 1,3-cyclooctadiene gave the corresponding [1 + 4] adduct in low 2% yield.¹⁴ 1,3-Cyclooctadiene was unreactive in titanium complex catalyzed [2 + 6] cycloaddition reaction with cycloheptatriene.¹⁵ Reaction of unconjugated triene 1,5,9-cyclododecatriene with chlorosulfonyl isocyanate (CSI) gave an unsaturated β-lactam product, where only one double bond had reacted.¹⁶ Acetylacetone underwent a photochemical cycloaddition to 1,5,9-cyclododecatriene (the de Mayo reaction).¹⁷ The sterically controlled cycloaddition of dichloroketene to unreactive olefin, 5-methylene-2-norbornene, has been accomplished.¹⁸ It resulted in a spirocyclobutanone derivative formed by addition across the exocyclic double bond.

The sesquiterpenes possess fifteen carbons, derived from three isoprenoid units. This is a vast group of naturally occurring substances, containing an immense range of structural diversity which includes acyclic, monocyclic, bicyclic, tricyclic and tetracyclic compounds. Many sesquiterpenes display biological activity including antimicrobial, antitumor and cytotoxic properties.¹⁹ Some bisnor-sesquiterpenes exhibit allelopathic activity.²⁰

Herein we present results of our research concerning site selectivity, regioselectivity and stereoselectivity of 1,3-dipolar cycloaddition of 4-trifluoromethylbenzonitrile oxide to polyunsaturated medium rings including 11-membered sesquiterpenes 5, 8 and 10, 1,5,9-cyclododecatriene (13), 1,3-cyclooctadiene (18) and 5-vinyl-2-norbornene (21) envisaging to obtain new biologically active compounds. Structure of the products has been established by an extensive application of 1D and 2D ¹H and ¹³C nuclear magnetic resonance (NMR) spectroscopy and electrospray ionization mass spectrometry (ESI-MS).

Experimental

Materials and physical measurements

Reagent-grade chemicals were used without further purification unless otherwise noted. Compounds 5, 8, 10, 13, 18, 21 were purchased from Aldrich. Hydroximinoyl acid chlorides were prepared from the corresponding aryl aldehyde oximes and N-chlorosuccinimide (NCS) in N,N-dimethylformamide (DMF).²¹

Spectra were recorded as follows: Fourier transform (FT) infrared (IR) spectra on a JASCO FTIR-420 spectrometer; ¹H, ¹³C NMR, 2D correlation spectroscopy (COSY), 2D heteronuclear single quantum coherence (HSQC), 2D heteronuclear multiple bond correlation (HMBC) and 2D rotating frame nuclear Overhauser effect spectroscopy (ROESY) analyses on a Bruker AVANCE III 500 MHz and a Varian VNMRS 600 spectrometers in deuterated chloroform. The ¹H NMR spectra were recorded using single-pulse sequence and spectral width (SW) of 10000 Hz, 30° pulse width (pw) of 9.7 µs, an acquisition time (at) of ca. 3.3 s and 64 k complex points. The free induction decays (FIDs) were processed with zero filling. The ¹³C spectra were obtained using a spectral range of ca. 31250 Hz, 30° pulse width (3.3 µs), an acquisition time of ca. 1.05 s, a relaxation delay of 1.5 s and collecting 64 k complex points. Chemical shifts are given in ppm (δ) relative to tetramethylsilane (TMS) as an internal standard and coupling constants are reported in Hz. The 2D gradient selected (g)COSY and ROESY spectra were run using spectral width (ca. 6000 Hz) in both dimensions, at = 0.18 s, 2-4 (COSY) and 4-8 (ROESY) transients per 512 increments, relaxation delay (d1) of 2.0 s. Prior to FT the data were processed using squared sinebell (COSY) and gaussian (ROESY) multiple function. In the case of the ROESY experiments spinlock time ca. 200 ms was chosen. The echo-antiecho phase-sensitive ¹H/¹³C gHSQC correlation were obtained with an at of 0.18 s, spectral window of 5500 Hz (F2) and 25500 Hz (F1), 512 increments in the ¹³C dimension, d1 = 2.0 s and 2 transients per t₁ increment. Experiments were optimized for ¹J(C-H) = 145 Hz. The data were zero-filled to 2048 points and processed using cosine-squared window function in both dimensions prior to Fourier transformation. The proton and carbon 90° pulse lengths were 8 and 15 ms, respectively. The ¹H-¹³C HMBC experiments with pulse field gradient (PFG) coherence selection using two PFG pulses were recorded with the following parameters: an at of 0.18 s, spectral windows of 5500 Hz (F2) and 25500 Hz (F1), 512 increments in the ¹³C dimension, 4 transients per increment and d1 of 2 s. This kind of experiment was optimized for ⁿJ(C-H) = 10 Hz. The proton and carbon 90° pulse lengths were 8 and 15 ms, respectively. The ¹H-¹⁵N HMBC experiments with PFG coherence selection using two PFG pulses were recorded with the following parameters: an at of 0.25 s, spectral windows of 6500 Hz (F2) and 12000 Hz (F1); 2 × 256 increments in the ¹⁵N dimension, 16-32 transients per increment and d1 of 1.5 s, optimized for ⁿJ(N-H) = 5.0 Hz. The proton and nitrogen 90° pulse lengths were 7.2 and 30.0 ms, respectively.

Electron ionization (EI) MS were run on an AMD M-40 instrument, ESI-MS on an LCT (Micromass) apparatus. Flash chromatography was carried out using silica gel S 230-400 mesh (Merck) using hexane-ethyl acetate mixtures as an eluent. Calculations of electron charges on alkenyl carbon atoms, molecular modelling and substrate HOMO/LUMO energies were calculated with the Hyperchem 7.5 program using semiempirical AM1 method.²²

Cycloaddition reaction of dipolarophiles 5, 8, 10, 13, 18 and 21 with 4-trifluoromethylbenzonitrile oxide (4). A general procedure for preparation of 6, 7, 9, 11, 12, 14-17, 19, 20, and 22-27

4-Trifluoromethylbenzonitrile oxide (4) was generated as follows: a solution of the corresponding chloroxime (0.25 g, 1.12 mmol) in dry dichloromethane was passed through an Amberlyst-21 column and added dropwise over 30 min to the solution of a dipolarophile in dry dichloromethane, and the solution was stirred overnight at room temperature. Water was added, organic layer was separated and the aqueous one extracted with dichloromethane. The combined organic layers were dried (MgSO₄) and the product was purified by flash column chromatography.

3a,12a-trans-(6E,10E)-6,6,9,12a-Tetramethyl-3-[4-(trifluoromethyl)phenyl]-3aH,4H,5H,8H,9H,12H,12aH-cycloundeca[d][1,2]oxazole (6)

Celadon-yellow wax; yield: 35%; ¹H NMR (500 MHz, CDCl₃) δ 7.73 (d, 2H, J 8.3 Hz, H2', H6'), 7.67 (d, 2H, J 8.3 Hz, H3', H5'), 5.24 (ddd, 1H, J 16.5, 10.0, 5.0 Hz, H4), 5.18 (d, 1H, J 16.5 Hz, H5), 4.97 (bd, 1H, J 11.8 Hz, H8), 3.37 (dd, 1H, J 11.0, 2.0 Hz, H1), 2.32 (dq, 1H, J 11.8, 5.0 Hz, H3a), 2.20 (td, 1H, J 12.5, 5.0 Hz, H10a), 2.16-2.12 (m, 1H, H7a), 2.14-2.10 (m, 2H, H6b, H10b), 1.88 (d, 1H, J 12.5 Hz, H7b), 1.70-1.63 (m, 1H, H11a), 1.62 (s, 3H, H16), 1.61-1.54 (m, 1H, H11b), 1.53 (s, 3H, H13), 1.17 (s, 3H, H15), 1.15 (s, 3H, H14); ¹³C NMR (125.9 MHz, CDCl₃) δ 160.4 (C12), 143.4 (C5), 133.8 (C1'), 133.0 (C4), 131.4 (q, J 32.5 Hz, C4'), 131.1 (C9), 128.5 (C5), 127.4 (C6', C2'), 125.7 (q, J 3.9 Hz, C5', C3'), 123.9 (q, J 265.2 Hz, F₃C), 122.2 (C4), 90.8 (C2), 47.2 (C1), 42.8 (C3), 39.5 (C7), 37.7 (C6), 37.0 (C10), 30.2 (C15), 24.5 (C14), 23.9 (C11), 19.3 (C13), 15.8 (C16); ¹⁹F NMR (471 MHz, CDCl₃) δ -63.21 (s, F₃C); ¹⁵N NMR (from HMBC, 600 MHz, CDCl₃) δ -14.2; high resolution (HR)MS (FTMS + probe (p)ESI) calcd. for C₂₃H₂₈NOF₃Na [M]⁺: 414.2021; found: 414.2018.

3a,12a-trans-(5E,9E)-3-[4-(Trifluoromethyl)phenyl]-5,9,12,12-tetramethyl-3aH,4H,7H,8H,11H,12H,12aH-cycloundeca[d][1,2]oxazole (7)

Celadon-white wax; yield: 50%; ¹H NMR (500 MHz, CDCl₃) δ 7.83 (d, 2H, J 8.0 Hz, H2', H6'), 7.68 (d, 2H, J 8.0 Hz, H3', H5'), 5.00 (dd, 1H, J 11.5, 4.0 Hz, H1), 4.92 (d, 1H, J 10.9 Hz, H8), 4.57 (d, 1H, J 2.8 Hz, H5), 3.29 (ddd, 1H, J 10.5, 2.8, 1.8 Hz, H4), 2.32 (ddd, 1H, J 22.5, 11.5, 4.8 Hz, H11a), 2.24 (dm, 1H, J 12.3 Hz, H10a), 2.23-2.17 (m, 1H, H3a), 2.13 (dd, 1H, J 10.5, 1.8 Hz, H3b), 2.11-2.07 (m, 2H, H7a, H11b), 1.98 (dd, 1H, J 12.3, 4.8 Hz, H10b), 1.96 (dd, 1H, J 12.0, 4.8 Hz, H7b), 1.65 (s, 3H, H16), 1.55 (s, 3H, H15), 1.22 (s, 3H, H13), 0.74 (s, 3H, H14); ¹³C NMR (125.9 MHz, CDCl₃) δ 158.0 (C12), 133.7 (C9), 133.0 (C2), 132.4 (q, J 1.5 Hz, C1'), 131.4 (q, J 32.7 Hz, C4'), 127.7 (C1), 126.9 (C2', C6'), 125.9 (q, J 3.4 Hz, C3', C5'), 123.5 (q, J 272.4 Hz, F₃C), 123.0 (C8), 93.3 (C5), 46.0 (C4), 43.1 (C3), 38.9 (C10), 37.6 (C7), 25.9 (C11), 24.6 (C13), 22.7 (C14), 16.7 (C16), 16.5 (C15); ¹⁹F NMR (471 MHz, CDCl₃) δ -63.26 (s, F₃C); ¹⁵N NMR (from HMBC, 600 MHz, CDCl₃) d -9.57; HRMS (FTMS + pESI) calcd. for: C₂₃H₂₈NOF₃H [M]⁺: 392.2201; found: 392.2197.

3a,12a-trans-5,5,12a-Trimethyl-3,4'-bis[4-(trifluoromethyl)phenyl]-3a,4,5,8,10,11,12,12a-octahydro-3'H-spiro[cycloundeca[d][1,2]oxazole-9,2'-[1,5]oxazole] (9)

Celadon-brown oil, yield: 55%; ¹H NMR (500 MHz, CDCl₃) δ 7.80 (d, 2H, J 8.3 Hz, H2", H6"), 7.69 (d, 2H, J 8.0 Hz, H3', H5'), 7.67 (d, 2H, J 8.3 Hz, H3", H5"), 7.65 (d, 2H, J 8.0 Hz, H2', H6'), 5.58 (d, 1H, J 15.5 Hz, H5), 5.43 (ddd, 1H, J 15.5, 9.0, 5.0 Hz, H4), 3.32 (t, 1H, J 5.5 Hz, H8), 3.31 (d, 1H, J 16.5 Hz, H14a), 3.21 (d, 1H, J 16.5 Hz, H14b), 2.66-2.63 (m, 2H, H3a, H3b), 2.02-1.97 (m, 1H, H1a), 1.91-1.85 (m, 2H, H10a, H11a), 1.80 (dd, 1H, J 12.5, 6.5 Hz, H10b), 1.76 (dd, 1H, J 15.5, 6.8 Hz, H7a), 1.67-1.58 (m, 2H, H1b, H11b), 1.46 (d, 1H, J 15.5 Hz, H7b), 1.41 (s, 3H, H17), 1.17 (s, 3H, H16), 1.02 (s, 3H, H15); ¹³C NMR (125.9 MHz, CDCl₃) δ 160.8 (C12), 154.7 (C13), 141.7 (C5), 134.0 (C1"), 133.4 (C1'), 131.5 (q, J 32.5 Hz, C4'), 131.1 (q, J 32.1 Hz, C4"), 128.0 (C2', C6'), 126.7 (C2", C6"), 125.7 (q, J 3.8 Hz, C3', C5'), 125.5 (q, J 3.8 Hz, C3", C5"), 123.9 (q, J 272.1 Hz, F₃C-Ar'), 123.8 (q, J 272.1 Hz, F₃C-Ar"), 122.2 (C4), 91.4 (C9), 90.4 (C2), 47.5 (C8), 43.1 (C14), 42.8 (C3), 38.1 (C10), 38.05 (C1), 38.0 (C7), 37.3 (C6), 30.6 (C15), 24.8 (C16), 23.5 (C17), 16.9 (C11); MS (FTMS + pESI) 601 ([M + Na]⁺, 35), 577 ([M - H]⁺, 75); anal. calcd. for C₃₁H₃₂F₆N₂O₂ [M]⁺: C, 64.35%; H, 5.57%; found: C, 64.19%; H, 5.39%.

(1S,5E,9R)-6,10,10-Trimethyl-4'-[4-(trifluoromethyl)phenyl]-3'H-spiro[bicyclo[7.2.0]undecane-2,2'-[1,5]oxazole]-5-ene (11)

White wax; yield: 45%; ¹H NMR (500 MHz, CDCl₃) δ 7.83 (d, 2H, J 8.2 Hz, H2', H6'), 7.67 (d, 2H, J 8.2 Hz, H3', H5'), 5.25-5.16 (m, 1H, H4), 3.32 (d, 1H, J 16.0 Hz, H13a), 3.19 (d, 1H, J 16.0 Hz, H13b), 2.45-2.28 (m, 2H, H11, H3a), 2.14-2.08 (m, 1H, H6a), 2.09-2.03 (m, 1H, H2a), 2.01-1.96 (m, 1H, H3b), 1.99-1.93 (m, 2H, H2b, H6b), 1.81 (dd, 1H, J 12.0, 8.5 Hz, H10a), 1.71 (s, 3H, H14), 1.64-1.58 (m, 2H, H8, H7a), 1.58-1.50 (m, 1H, H7b), 1.33-1.24 (m, 1H, H10b), 0.97 (s, 3H, H15), 0.92 (s, 3H, H16); ¹³C NMR (125.9 MHz, CDCl₃) δ 154.1 (C12), 137.8 (C1'), 133.8 (C5), 131.1 (q, J 32.5 Hz, C4'), 126.6 (C2', C6'), 125.6 (q, J 3.9 Hz, C3', C5'), 123.9 (q, J 270.5 Hz, F₃C), 120.2 (C4), 95.1 (C1), 49.6 (C11), 48.5 (C8), 40.1 (C6), 38.7 (C2), 36.8 (C10), 36.0 (C13), 32.0 (C9), 30.5 (C15), 29.8 (C7), 23.5 (C16), 21.6 (C3), 16.0 (C14); HRMS (FTMS + pESI) calcd. for C₂₃H₂₈NOF₃Na [M]⁺: 414.2021; found: 414.2028.

(1'S,4'R,7'S,11'R)-1',5',5'-Trimethyl-4-12'-bis[4-(trifluoromethyl)phenyl]-3H-14'-oxa-13'-azaspiro[1,5-oxazole-2,8-tricyclo[9.3.0.0^4,7]tetradecan]-12'-ene (12)

Yellowish semisolid, yield: 30%; ¹H NMR (500 MHz, CDCl₃) δ 7.80 (d, 2H, J 8.3 Hz, H2", H6"), 7.71-7.70 (m, 2H, H3', H5'), 7.71 (d, 2H, J 8.3 Hz, H3", H5"), 7.70-7.69 (m, H2', H6'), 3.43 (d, 1H, J 16.0 Hz, H13a), 3.20 (d, 1H, J 16.0 Hz, H13b), 3.25 (dm, 1H, J 12.5 Hz, H4), 2.56 (app. q, 1H, J 9.3 Hz, H8), 2.23-2.17 (m, 1H, H6a), 2.17-2.09 (m, 2H, H2a, H6b), 1.97-1.89 (m, 1H, H11), 1.86-1.77 (m, 3H, H2b, H7a, H10a), 1.76-1.68 (m, 1H, H7b), 1.49 (s, 3H, H15), 1.33-1.26 (m, 3H, H3a, H3b, H10b), 1.03 (s, 3H, H16), 0.99 (s, 3H, H17); ¹³C NMR (125.9 MHz, CDCl₃) δ 158.9 (C14), 153.8 (C12), 133.7 (C1'), 133.2 (q, J 1.1 Hz, C1"), 131.6 (q, J 32.6 Hz, C4'), 131.5 (q, J 32.6 Hz, C4"), 127.4 (C2', C6'), 126.5 (C2", C6"), 125.7 (q, J 3.7 Hz, C3', C5'), 125.7 (q, J 3.7 Hz, C3", C5"), 123.8 (q, J 272.4 Hz, F₃C-Ar'), 123.7 (q, J 272.4 Hz, F₃C-Ar"), 92.9 (C1), 90.2 (C5), 50.0 (C4), 48.7 (C11), 46.8 (C8), 36.6 (C2), 36.5 (C6), 35.9 (C13), 35.1 (C10), 33.9 (C9), 29.9 (C16), 29.6 (C3), 22.3 (C7), 22.2 (C17), 19.6 (C15); MS (FTMS + pESI) 601 ([M + Na]⁺, 20), 577 ([M - H]⁺, 12); anal. calcd. for C₃₁H₃₂F₆N₂O₂: C, 64.35%; H, 5.57%; found: C, 64.58%; H, 5.42%.

(6E,10E)-3a,13a-cis-3-[4-(Trifluoromethyl)phenyl]-3aH,4H,5H,8H,9H,12H,13H,13aH-cyclododeca[d][1,2]oxazole (14, 15)

White-celadon dense oil (ratio of 14 and 15 isomers: 53:47), yield: 62%; ¹H and ¹³C NMR see Tables 1 and 2; MS (FTMS + pESI) 372 ([M + Na]⁺, 30), 350 ([M + H]⁺, 80), 227 ([M - C₆H₄CF₃]⁺, 99); anal. calcd. for C₂₀H₂₂F₃NO: C, 68.75%; H, 6.35%; found: C, 68.93%; H, 6.49%.

Table 1
¹H NMR signals of 14, 15, 16, and 17 (500 MHz, CDCl₃)

H	δ / ppm (J / Hz)
	14	15	16	17
1	5.09-4.98 (m)	5.53-5.45 (m)	3.37 (ddd, 10.5, 3.0, 1.5)	4.61 (ddd, 10.5, 4.5, 2.0)
2	5.43-5.30 (m)	5.41-5.27 (m)	4.66 (dm, 10.5)	3.46 (dm, 10.5)
3a	2.14-2.07 (m)	2.12-2.05 (m)	1.84-1.79 (m)	1.65-1.60 (m)
3b	2.05-1.99 (m)	2.02-1.96 (m)	1.68-1.62 (m)	1.57-1.51 (m)
4a	2.12-2.08 (m)	2.12-2.08 (m)	2.16-2.12 (m)	2.28-2.21 (m)
4b	2.20-2.14 (m)	2.02-1.96 (m)	2.05-1.98 (m)	2.20-2.18 (m)
5	5.23-5.12 (m)	5.38-5.30 (m)	5.52-5.42 (m)	5.30-5.22 (m)
6	5.43-5.30 (m)	5.23-5.12 (m)	5.40-5.35 (m)	5.52-5.42 (m)
7a	2.06-2.00 (m)	2.22-2.14 (m)	2.37-2.32 (m)	2.27-2.20 (m)
7b	2.02-1.95 (m)	2.11-2.05 (m)	2.15-2.10 (m)	2.16-2.10 (m)
8a	2.24-2.18(m)	1.31-1.22 (m)	2.29-2.20 (m)	2.15-2.10 (m)
8b	1.84-1.76 (m)	1.05 -1.00 (m)	2.15-2.10 (m)	2.12-2.06 (m)
9	4.24 (ddd, 12.5, 10.1, 1.8)	3.41 (dm, 11.2)	5.53-5.42 (m)	5.53-5.43 (m)
10	3.20 (dd, 12.5, 3.2)	4.35-4.30 (m)	5.29-5.22 (m)	5.40-5.34 (m)
11a	1.48-1.36 (m)	2.22-2.14 (m)	2.28-2.21 (m)	2.41-2.30 (m)
11b	1.07-0.98 (m)	1.82-1.77 (m)	2.15-2.10 (m)	2.15-2.10 (m)
12a	2.24-2.18 (m)	2.23-2.16 (m)	1.67-1.62 (m)	1.87-1.83 (m)
12b	2.02-1.95 (m)	2.02-1.95 (m)	1.59-1.53 (m)	1.74-1.68 (m)
2’, 6’	7.80 (d, 8.1)	8.09 (dm, 8.1)	7.83 (d, 8.0)	7.79 (d, 8.0)
3’, 5’	7.67 (dd, 8.1, 2.6)	7.71(dm, 8.1)	7.68 (d, 8.0)	7.68 (d, 8.0)

• δ: chemical shift; J: coupling constant.

Table 2
¹³C NMR signals of 14, 15, 16, and 17 (125.8 MHz, CDCl₃)

C	δ / ppm (J / Hz)
	14	15	16	17
1	130.5	130.9	50.4	85.9
2	131.8	129.3	86.0	49.7
3	32.3	31.6	31.4	28.4
4	31.5	31.7	31.8	30.2
5	129.3	131.5	132.2	129.1
6	130.9	129.0	129.7	130.0
7	27.1	31.2	24.9	32.0
8	29.7	28.6	22.4	24.9
9	87.2	53.9	131.3	130.9
10	56.9	87.4	129.2	129.7
11	30.6	29.7	22.4	24.9
12	31.1	27.0	28.4	32.5
13	159.2	159.4	159.5	159.7
1’	132.5 (q, 2.4)	132.8 (q, 2.4)	132.7 (q, 1.3)	132.8 (q, 1.3)
2’, 6’	127.3	127.1	127.2	127.2
3’, 5’	125.8 (q, 3.8)	125.7 (m)	125.8 (q, 3.7)	125.8 (q, 3.7)
4’	131.2 (q, 32.7)	131.1 (q, 32.6)	131.5 (m)	131.4 (m)
CF3	123.7 (q, 270.9)	124.1 (q, 272.0)	123.7 (q, 270.6)	123.7 (q, 270.6)

• δ: chemical shift; J: coupling constant.

(6E,10E)-3a,13a-trans-3-[4-(Trifluoromethyl)phenyl]-3a,4,5,8,9,12,13,13a-octahydrocyclododeca[d][1,2]oxazole (16, 17)

White wax (ratio of 16 and 17 isomers: 60:40), yield: 38%; ¹H and ¹³C NMR see Tables 1 and 2; anal. calcd. for C₂₀H₂₂F₃NO: C, 68.75%; H, 6.35%; found: C, 68.65%; H, 6.47%.

3a,9a-cis-3-[4-(Trifluoromethyl)phenyl]-3aH,4H,5H,6H,7H,9aH-cycloocta[d][1,2]oxazole (19); 3-[4-(trifluoromethyl)phenyl]-3aH,6H,7H,8H,9H,9aH-cycloocta[d][1,2]oxazole (20)

Celadon-yellow oil (ratio of 19/20 isomers 5/1), yield: 45%; 19: ¹H NMR (600 MHz, CDCl₃) δ 7.71 (d, 2H, J 8.3 Hz, H3', H5'), 7.66 (d, 2H, J 8.3 Hz, H2', H6'), 5.87-5.82 (m, 1H, H1), 5.60 (dm, J 11.8 Hz, 1H, H3), 5.55 (dd, 1H, J 11.8, 3.9 Hz, H2), 3.74 (td, 1H, J 10.1, 1.5 Hz, H8), 2.25-2.13 (m, 2H, H4), 1.90-1.80 (m, 2H, H7), 1.58-150 (m, 2H, H5); ¹³C NMR (from HMBC, 150.8 MHz, CDCl₃) δ 159.8 (C9), 133.2 C3), 133.2 (C1'), 131.4 (C4'), 127.4 (C-2', C-6'), 126.7 (C2), 125.7 (C3', C5'), 83.2 (C1), 53.0 (C8), 30.2 (C7), 26.6 (C5), 25.8 (C4), 24.6 (C16); ¹⁵N NMR (from HMBC, 600 MHz, CDCl₃) δ -14.10 (s, 1N, =N-O); 20: ¹H NMR (600 MHz, CDCl₃) δ 7.83 (d, 2H, J 8.0 Hz, H2', H6'), 7.64 (d, 2H, J 8.0 Hz, H3', H5'), 6.02-5.97 (m, 1H, H2), 5.16 (ddd, 1H, J 11.1, 7.0, 0.8 Hz, H3), 4.61 (ddd, 1H, J 11.1, 9.4, 3.3 Hz, H8), 4.41 (tm, 1H, J 8.0 Hz, H1), 2.28-2.20 (m, 2H, H4), 2.00-1.80 (m, H6, H7), 1.58-1.48 (m, 1H, H5); anal. calcd. for C₁₆H₁₆F₃NO: C, 65.08%; H, 5.46%; found: C, 65.36%; H, 5.67%.

9-Ethenyl-5-[4-(trifluoromethyl)phenyl]-3-oxa-4-azatricyclo[5.2.1.0^2,6]dec-4-ene (22)

White wax; yield: 12%; ¹H NMR (500 MHz, CDCl₃) δ 7.76 (d, 2H, J 8.5 Hz, H2', H6'), 7.64 (d, 2H, J 8.5 Hz, H3', H5'), 5.94 (ddd, 1H, J 17.0, 10.5, 1.5 Hz, H2), 5.21 (ddd, 1H, J 10.5, 1.5 Hz, H1a), 5.19 (dd, 1H, J 17.0, 1.5 Hz, H1b), 4.65 (dm, 1H, J 8.3 Hz, H7), 3.78 (dm, 1H, J 8.3 Hz, H6), 2.65 (d, 1H, J 5.5 Hz, H8), 2.67-2.61 (m, 1H, H3), 2.45 (d, 1H, J 3.0 Hz, H5), 1.85 (ddd, 1H, J 13.3, 11.0, 5.5 Hz, H4a), 1.57 (dm, 1H, J 9.3 Hz, H9a), 1.37 (dm, 1H, J 9.3 Hz, H9b), 1.03 (ddd, 1H, J 13.3, 5.5, 2.0 Hz, H4b); ¹³C NMR (125.9 MHz, CDCl₃) δ 156.4 (C10), 139.1 (C2), 132.8 (q, J 1.3 Hz, C1'), 131.3 (q, J 32.7 Hz, C4'), 126.9 (C2', C6'), 125.7 (q, J 3.8 Hz, C3', C5'), 123.8 (q, J 272.3 Hz, F₃C), 116.3 (C1), 88.4 (C7), 50.6 (C6), 44.6 (C5), 44.0 (C8), 41.9 (C3), 33.5 (C9), 27.8 (C4); anal. calcd. for C₁₇H₁₆F₃NO: C, 66.44%; H, 5.25%; found: C, 66.61%; H, 5.17%.

8-Ethenyl-5-[4-(trifluoromethyl)phenyl]-3-oxa-4-azatricyclo[5.2.1.0^2,6]dec-4-ene (23)

White-yelowish wax; yield: 30%; ¹H NMR (500 MHz, CDCl₃) δ 7.83 (d, 2H, J 8.0 Hz, H2', H6'), 7.65 (d, 2H, J 8.0 Hz, H3', H5'), 5.85 (ddd, 1H, J 17.3, 5.5, 1.5 Hz, H2), 5.13 (dd, 1H, J 17.3, 1.5 Hz, H1a), 5.12 (d, 1H, J 10.5, 1.5 Hz, H1b), 4.92 (d, 1H, J 8.5 Hz, H6), 3.51 (d, 1H, J 8.5 Hz, H7), 2.70 (d, 1H, J 4.8 Hz, H5), 2.69-2.63 (m, 1H, H3), 2.51 (d, 1H, J 4.5 Hz, H8), 1.91 (ddd, 1H, J 15.4, 11.0, 4.8 Hz, H4a), 1.56 (app. dq, 1H, J 10.5, 2.0 Hz, H9a), 1.35 (dt, 1H, J 10.5, 2.0 Hz, H9b), 1.30 (ddd, 1H, J 15.4, 4.8, 2.5 Hz, H4b); ¹³C NMR (125.9 MHz, CDCl₃) δ 156.0 (C10), 138.8 (C2), 132.8 (C1'), 131.2 (q, J 32.7 Hz, C4'), 127.0 (C2', C6'), 125.7 (q, J 3.9 Hz, C3', C5'), 123.9 (q, J 272.2 Hz, F₃C), 115.5 (C1), 85.0 (C6), 56.8 (C7), 48.2 (C8), 40.3 (C5), 40.0 (C3), 33.6 (C9), 32.4 (C4); anal. calcd. for C₁₇H₁₆F₃NO: C, 66.44%; H, 5.25%; found: C, 66.30%; H, 5.39%.

5-[4-(Trifluoromethyl)phenyl]-8-endo-{3-[4-(trifluoromethyl)phenyl]-4,5-dihydro-1,2-oxazol-5-yl}-3-oxa-4-azatricyclo[5.2.1.0^2,6]dec-4-ene (24)

Celadon wax; yield: 11%; ¹H and ¹³C NMR see Tables 3 and 4; anal. calcd. for C₂₅H₂₀F₆N₂O₂: C, 60.73%; H, 4.87%; found: C, 60.92%; H, 4.91%.

Table 3
¹H NMR signals of 24, 25, 26 and 27 (500 MHz, CDCl₃)

H	δ / ppm (J / Hz)
	24	25	26	27
1a	3.56 (dd, 16.3, 10.5)	3.54 (dd, 16.5, 10.5)	3.53 (dd, 16.5, 10.3)	3.53(dd, 16.3, 10.3)
1b	3.08 (dd, 16.3, 9.4)	3.08 (dd, 16.5, 9.0)	3.07 (dd, 16.5, 8.0)	3.07(dd, 16.3, 8.3)
2	5.01 (ddd, 10.5, 9.4, 4.8)	4.86 (ddd, 10.5, 9.0, 6.5)	4.73-4.65 (m)	4.73-4.65 (m)
3	2.29-2.22 (m)	2.30-2.24 (m)	2.25-2.19 (m)	2.28-2.19 (m)
4a	1.83 (ddd, 12.0, 12.0; 5.5)	1.92(ddd,12.0, 8.0, 4.0)	1.90 (ddd, 10.0, 7.5, 5.4)	1.90 (td, 12.0, 5.0 )
4b	1.18 (ddd,12.0, 5.8, 2.5)	1.40 (dm, 12.0)	0.79 (ddd, 10.0, 5.4, 3.5)	0.79 (dm, 12.0)
5	2.71 (d, 5.0)	2.77 (d, 4.0)	2.82 (bd, 3.5)	2.82 (d, 2.5)
6	4.80 (d, 8.3)	5.16 (d, 8.3)	3.99 (d, 8.3)	3.99 (d, 8.3)
7	4.20 (d, 8.3)	3.69 (d, 8.3)	4.70 (d, 8.3)	4.70 (d, 8.3)
8	2.54 (d, 3.5)	2.56 (bs)	2.70 (d, 5.0)	2.70 (d, 5.0)
9a	1.60 (dm, 11.0)	1.65-1.62 (m)	1.62 (dm, 11.0)	1.62 (dm, 11.0)
9b	1.36 (dm, 11.0)	1.37-1.32 (m)	1.33 (dm, 11.0)	1.33 (dm, 11.0)
2’, 6’	7.80 (d, 8.5)	7.83 (d, 8.5)	7.94 (d, 8.5)	7.93 (d, 8.0)
3’, 5’	7.69 (d, 8.5)	7.66 (d, 8.5)	7.68 (d, 8.5)	7.69 (d, 8.0)
2’’, 6’’	7.79 (d, 8.0)	7.78 (d, 8.3)	7.80 (d, 8.0)	7.80 (d, 8.0)
3’’, 5’’	7.68 (d, 8.0)	7.68 (d, 8.3)	7.69 (d, 8.0)	7.67 (d, 8.0)

• δ: chemical shift; J: coupling constant.

Table 4
¹³C NMR signals of 24, 25, 26 and 27 (125.8 MHz, CDCl₃)

C	δ / ppm (J / Hz)
	24	25	26	27
1	40.1	39.9	39.7	39.7
2	83.0	82.9	82.9	82.9
3	43.9	42.2	43.9	43.9
4	24.6	29.7	26.6	26.6
5	44.0	46.7	41.4	41.4
6	88.3	56.9	50.7	50.7
7	51.2	84.8	87.9	87.9
8	42.8	40.3	43.5	43.5
9	33.9	33.8	33.2	33.2
10	155.9	156.1	156.2	156.2
11	155.2	155.4	155.7	155.7
1’	132.8	132.8	132.8	132.8
2’, 6’	126.9	127.1	127.2	127.9
3’, 5’	125.8 (q, 3.8)	125.8 (q, 3.8)	125.8 (q, 3.8)	125.9 (q, 3.8)
4’	132.0 (q, 32.8)	131.9 (q, 32.8)	132.1 (q, 32.7)	131.5 (q, 32.6)
CF3’	123.8 (q, 272.4)	123.8 (q, 272.1)	124.9 (q, 272.2)	123.9 (q, 272.1)
1’’	132.7	132.7	132.3	132.3
2’’, 6’’	126.9	126.9	126.9	126.9
3’’, 5’’	125.8 (q, 3.8)	125.7 (q, 3.7)	125.8 (q, 3.8)	125.8 (q, 3.7)
CF3’’	123.8 (q, 272.4)	123.8 (q, 272.0)	124.8 (q, 272.4)	123.8 (q, 272.5)
4’’	131.5 (q, 32.7)	131.5 (q, 35.7)	131.3 (q, 32.6)	132.0 (q, 32.7)

• δ: chemical shift; J: coupling constant.

5-[4-(Trifluoromethyl)phenyl]-9-endo-{3-[4-(trifluoromethyl)phenyl]-4,5-dihydro-1,2-oxazol-5-yl}-3-oxa-4-azatricyclo[5.2.1.0^2,6]dec-4-ene (25)

Celadon wax; yield: 7%; ¹H and ¹³C NMR see Tables 3 and 4; anal. calcd. for C₂₅H₂₀F₆N₂O₂: C, 60.73%; H, 4.87%; found: C, 60.89%; H, 5.03%.

5-[4-(Trifluoromethyl)phenyl]-9-exo-{3-[4-(trifluoromethyl)phenyl]-4,5-dihydro-1,2-oxazol-5-yl}-3-oxa-4-azatricyclo[5.2.1.0^2,6]dec-4-ene (26)

Celadon wax; yield: 8%; ¹H and ¹³C NMR see Tables 3 and 4; anal. calcd. for C₂₅H₂₀F₆N₂O₂: C, 60.73%; H, 4.87%; found: C, 60.54%; H, 5.02%.

5-[4-(Trifluoromethyl)phenyl]-8-exo-{3-[4-(trifluoromethyl)phenyl]-4,5-dihydro-1,2-oxazol-5-yl}-3-oxa-4-azatricyclo[5.2.1.0^2,6]dec-4-ene (27)

Celadon wax; yield: 12%; ¹H and ¹³C NMR see Tables 3 and 4; anal. calcd. for C₂₅H₂₀F₆N₂O₂: C, 60.73%; H, 4.87%; found: C, 60.63%; H, 4.75%.

Fungicidal testing

The compounds were screened for fungicidal activity in vitro. The tests were carried out for Fusarium culmorum (F. c.), Phytophthora cactorum (P. c.), Rhizoctonia solani (Rh. s.), and Botrytis cinerea (B. c.), and involved determination of mycelial growth retardation in potato glucose agar (PGA). Stock solutions of test chemicals in acetone were added to agar medium to give a concentration of 200 mg L^-1 and dispersed into Petri dishes. Four discs containing the test fungus were placed at intervals on the surface of the solidified agar and the dishes were then inoculated for 4-8 days depending on the growth rate of the control samples, after which fungal growth was compared with that in untreated control samples. The fungicidal activity was expressed as the percentage of fungi linear growth inhibition compared to that of the control.

Results and Discussion

Structural analysis of the cycloadducts

We have examined [2 + 3] cycloaddition of 4-(trifluoromethyl)benzonitrile oxide (4) to polyunsaturated medium rings including 11-membered sesquiterpenes 5, 8, and 10, 1,5,9-cyclododecatriene (13), 1,3-cyclooctadiene (18) and 5-vinyl-2-norbornene (21). The compounds described in this work are presented in Schemes 1-6 and Figure 1. ¹H and ¹³C NMR chemical shifts and multiplicities of adducts 14-17 are shown, respectively, in Tables 1 and 2. ¹H and ¹³C NMR chemical shifts of adducts 24-27 are shown, respectively, in Tables 3 and 4. The spectroscopic data for the other products are presented in Experimental.

Scheme 1
Cycloaddition of dipole 4 to the dipolarophile sesquiterpene 5. (i) CH₂Cl₂, r.t. Non-systematic numbering of atoms has been used in adducts 6 and 7 for ease of comparison of spectral data.

Scheme 2
Cycloaddition of the dipole 4 to the C₁₅ sesquiterpene 8. (i) CH₂Cl₂, r.t.

Scheme 3
Cycloaddition of the dipole 4 to bicyclic C15 sesquiterpene 10. (i) CH₂Cl₂, r.t. Non-systematic numbering of atoms has been used in adducts 11 and 12 for ease of comparison of spectral data.

Scheme 4
Cycloaddition of the dipole 4 to trans,trans,cis-1,5,9-cyclododecatriene (13). (i) CH₂Cl₂, r.t. Non-systematic numbering of atoms has been used in adducts 14-17 for ease of comparison of spectral data.

Scheme 5
Cycloaddition of the dipole 4 to 1,3-cyclooctadiene (18). (i) CH₂Cl₂, r.t. Non-systematic numbering of atoms has been used in adducts 19 and 20 for ease of comparison of spectral data.

Scheme 6
Cycloaddition of the dipole 4 to 5-vinyl-2-norbornene (21). (i) CH₂Cl₂, r.t. Non-systematic numbering of atoms has been used in adducts 22-27 for ease of comparison of spectral data.

Figure 1
Selected COSYs, NOEs and HMBC for adducts 6, 9, 11, 14, and 25.

Cycloaddition of the sesquiterpene C₁₅trans,trans,trans-2,6,6,9-tetramethyl-1,4,8-cycloundecatriene (5) afforded two mono-adducts to different double bonds, 6 and 7, in 3:1 ratio (Scheme 1, Figure 1). Structure 6 was assigned to the major isomer based on analysis of 2D NMR HSQC and COSY spectra. Presence of three H-C correlations in an olefinic area in HSQC spectrum and presence of HC=HC group in the COSY spectrum indicated that an addition to one of two tri-substituted double bonds took place. Observation of a CH¹-CH₂¹¹-CH₂¹⁰ and not of a CH⁸-CH₂⁷ moiety in COSY spectrum showed that the addition occurred to C1-C2 double bond. Finally, regiochemistry of the cycloaddition was established as that shown in structure 6 by lack of H-C correlation in HSQC spectrum for carbon 2 connected to isoxazoline oxygen atom. This direction of addition was favored by the orbital factors. The reaction is LUMO-dipole controlled and oxygen atom of the dipole tends to attack the more substituted carbon atom of the dipolarophile.

Structure 7 was assigned to the minor isomer based on analysis of 2D NMR HSQC, COSY and HMBC spectra. Presence of two H-C correlations in an olefinic area in HSQC spectrum proved the addition to the di-substituted C4-C5 double bond. Correlations in the COSY spectrum between H5 and H13/H14 protons from the methyl groups and lack of correlations of these groups with H4, as well as a correlation of H3 with C12 in the HMBC spectrum, defined regiochemistry of the cycloaddition reaction where oxygen atom of the dipole attacked the less negative end of the olefinic bond.

Cycloaddition to the C₁₅ sesquiterpene trans,trans-8-methylene-1,4,4-trimethyl-1,5-cycloundecadiene (8) afforded a bis-adduct 9 in 55% yield (Scheme 2, Figure 1).

A complete assignment of ¹H and ¹³C NMR spectra of 9 was based on analysis of 1D and 2D ¹H and ¹³C NMR spectra (COSY, HSQC and ROESY) and is presented in Experimental. Presence of HC=HC unit in E configuration (a large coupling constant of 15.5 Hz in ¹H NMR spectrum) confirmed by the 2D COSY spectrum and two olefinic H-C correlations in HSQC spectrum indicated that C4-C5 double bond was preserved. Lack of H-C correlation in HSQC spectrum for carbon 9 connected to the isoxazoline oxygen atom and presence of H-C correlation for C8 have shown the same regiochemistry of the reaction as in case of compound 6. Oxygen atom of the second molecule of the dipole attacked the mono-adduct exocyclic double bond at the di-substituted carbon atom as expected. The 2D HSQC spectrum showed the presence of two H14a,b-C14 (43.1 ppm) correlations. Facial selectivity was proved to be as shown in the structure 9 by observation in the 2D ROESY NMR spectrum of a proximity between H3 and H14a/H14b.

Cycloaddition to the bicyclic C15 sesquiterpene trans-(1R,9S)-8-methylene-4,11,11-trimethylbicyclo[7.2.0]undec-4-ene (10) afforded a mixture of a mono-adduct 11 (45%) and a bis-adduct 12 (30%) (Scheme 3, Figure 1). Structure of the mono-adduct 11 was established based on analysis of 1D and 2D ¹H and ¹³C NMR spectra (COSY, HSQC and ROESY); a complete assignment of ¹H and ¹³C NMR spectra is presented in Experimental. Cycloaddition to the exocyclic double bond via attack of the oxygen atom of the dipole on the di-substituted carbon atom afforded a spiro system, where carbon atom C1 attached to oxygen did not show any H-C correlations in the 2D HSQC NMR spectrum. Facial selectivity was established by observation in the ROESY NMR spectrum of correlations between H13a,b at 3.32 and 3.19 ppm and H10a,b at 2.07 and 1.96 ppm, as well as between H13b and aryl H2'/H6' protons. The C4-C5 double bond was preserved as witnessed by observation in the ¹³C NMR spectrum signals of the quaternary C5 at 133.8 ppm and C4 at 120.2 ppm coupled in the HSQC spectrum to H4 at 5.21 ppm.

Mass spectrometry of the other product showed molecular weight of 578 mu, indicating a bis-adduct structure 12. It showed a similar regiochemistry and geometry of the spiro isoxazoline ring as the mono-adduct 11. The 2D ROESY NMR spectrum showed a proximity of H13a (3.43) and H11 (1.92), H13b (3.20) and H10a (1.81), as well as H13a,b and aryl H2'/H6' protons (7.71 ppm). Regiochemistry of the second cycloaddition step resulted from attack of the dipole oxygen atom at the more substituted carbon atom C5.

Cycloaddition to trans,trans,cis-1,5,9-cyclododecatriene (13) afforded a mixture of two inseparable regioisomeric mono-adducts to cis double bond 14 and 15 (53:47) as indicated by mass spectrometry giving a molecular weight of 349 mu, and two inseparable regioisomeric mono-adducts (60:40) to the trans double bond 16 and 17 in an overall 62:38 ratio (Scheme 4). A complete assignment of ¹H and ¹³C NMR spectra of the products was based on analysis of 1D and 2D ¹H and ¹³C NMR spectra (COSY, HSQC, HMBC, and ROESY NMR) and is presented in Tables 1 and 2.

Structures 14 and 15 were proposed to the major pair of adducts assuming a preferred addition to the cis-double bond. Calculations carried out with the molecular modelling program showed lower energies for these isomers and a much smaller difference of energy between endo and exo forms than for the adducts to the trans bonds of the dipolarophile 13. Larger coupling constants (12.5 Hz) were observed for the ring junction protons cis H9/H10 in these regioisomers than for the corresponding trans H1/H2 in isomers 16 and 17 (10.5 Hz). Calculations carried out with HNMR Predictor software v.12 confirmed higher J values for these protons in cis adducts 14 and 15 (9.5 Hz) than in trans adducts 16 and 17 (8.4 Hz). Smaller values of coupling constants of the corresponding critical H4/H5 protons in trans disubstituted isoxazolines than in cis disubstituted isomers were observed before.²³ Structure 14 was proposed to the endo adduct based on the ROESY 2D NMR spectrum showing a correlation of H10 (3.20 ppm) and H2' proton (7.80 ppm), while in the exo adduct 15 the corresponding proximity of H9 (3.41 ppm) and H2' (8.09) was absent. Model studies carried out with the Hyperchem program have confirmed this conclusion and showed spatial proximity of the relevant protons.

Similarly, endo configuration was proposed to the isomer 16 and exo configuration was ascribed to the isomer 17. 2D ROESY NMR spectrum of 16 showed a H1 (3.37 ppm)-H12a,b (1.64, 1.56 ppm) correlation, while in the other regioisomer 17 a corresponding correlation H2 (3.46 ppm)-H3a,b (1.63, 1.54 ppm) was missing.

Cycloaddition to 1,3-cyclooctadiene (18) afforded a mixture of regioisomeric mono-adducts 19 and 20 in 4:1 ratio (Scheme 5). Structure of the major isomer 19 was proposed as endo based on 2D ROESY NMR spectrum, where correlations of the bridgehead proton H8 at 3.74 ppm and the aliphatic H7 proton at 1.75 ppm with aromatic protons H2'/H6' at 7.71 ppm were observed, while no correlation in 2D HMBC NMR spectrum between the imine C9 carbon atom and the olefinic H2 proton was found, which would occur in isomer 20.

Cycloaddition to 5-vinyl-2-norbornene (21) (a mixture of exo/endo epimers, 78:22) afforded mono-adducts 22 and 23 in 11:30 ratio and regioisomeric bis-adducts 24-27 (Scheme 6). A complete assignment of ¹H and ¹³C NMR spectra of the products was based on analysis of 1D and 2D ¹H and ¹³C NMR spectra (COSY, HSQC, HMBC, and ROESY) and is presented in Tables 3 and 4 and Figure 1 for compounds 24-27. Assignments were facilitated by the published data on product of the [2 + 3] cycloaddition of nitrile oxides to the norbornenes.^9,24

In both monocycloadducts 22 and 23 reaction proceeded with a complete site-selectivity and exo-selectivity. The crucial isoxazoline H6 and H7 protons were in ¹H NMR spectra doublets coupled only to each other (J 8.3-8.5 Hz), while coupling constants with the bridgehead H5 and H8 are very small (0-1 Hz), which corresponds to the value of the respective dihedral angles close to 90º. Regiochemistry of the reaction was established by application of ¹H NMR and 2D NMR ROESY spectroscopy. In regioisomer 22 with syn relative position of the methylene bridge and the phenyl ring H7 proton at 4.65 ppm showed a cross peak with H8 at 2.65 ppm, and H6 at 3.78 ppm showed a cross peak with H5 at 2.45 ppm.

On the other hand, in regioisomer 23 with anti relative position of the methylene bridge and the phenyl ring, H6 proton (4.92 ppm) exhibited a cross peak with H5 (2.70 ppm), and H7 (3.51 ppm) exhibited a cross peak with H8 (2.51 ppm). Endo configuration of the vinyl side chain at C3 in regioisomer 22 was established by observance in the 2D ROESY NMR spectrum of a proximity of H3 (2.64) and H9 (1.57, 1.37 ppm) as well as H3 and H5. It was confirmed by finding a ⁴J W-type coupling between H5 (2.45) and H3. On the other hand, exo configuration at C3 was proposed for the regioisomer 23 based on 2D ROESY NMR spectrum, where correlations H3 (2.66 ppm)-H8 (2.51) and H1 (5.12 ppm)-H4b (1.30 ppm) were observed. Presence of anti configuration was corroborated by finding a correlation C10 (156.0 ppm)-H8 (2.51 ppm) in 2D HMBC NMR spectrum.

Spectral data and elemental analyses have proved that products 24-27 are bis-adducts. The first eluted bis-cycloadduct 24 was characterized by an anti relationship of the C10 aryl group and the methylene bridge as proved by 2D ROESY NMR spectrum, where spatial proximity of H6 (4.80 ppm) and H5 (2.71 ppm) as well as H7 (4.20 ppm ) and H8 (2.54 ppm) was found. Correlations in 2D ROESY NMR spectrum of H2 (5.01 ppm)-H7 and H2-H8 indicate that a C3 substituent is in the endo position.

The second eluted bis-cycloadduct 25 showed the opposite syn relationship of the C10 aryl group and the methylene bridge since in the 2D ROESY NMR spectrum correlations between H6 (5.16 ppm) and H5 (2.77 ppm) as well as between H7 (3.69 pm) and H8 (2.56 ppm) were observed. Endo configuration at C3 was proposed based also on the 2D ROESY NMR results, where spatial proximity of H1b (3.08 ppm) and H8, H2 (4.86 ppm) and H7, as well as H3 (2.27 ppm) and H9b (1.34 pm) were found, while no correlations H3-H7 and H3-H5 were observed.

The third bis-cycloadduct 26 showed also syn relationship of the C10 aryl group and the methylene bridge because in the 2D ROESY NMR spectrum correlations between H7 (4.70 ppm) and H8 (2.70 ppm), as well as between H6 (3.99 ppm) and H5 (2.82 ppm) were observed. On the other hand, exo position of the side substituent at C3 was established based on correlations H1b (3.07 ppm)-H3 (2.22 ppm) and H3-H7 in the 2D ROESY NMR spectrum. This assignment was corroborated by ¹³C NMR data, where larger chemical shift was recorded for C7 (87.9 ppm) than for the corresponding carbon atom in the endo isomer 25, 84.8 ppm because of the gauche γ-effect.

Finally, the fourth eluted bis-cycloadduct 27 showed anti relationship of the C10 aryl group and the methylene bridge because in the 2D ROESY NMR spectrum correlations H6 (4.70 ppm)-H5 (2.70 ppm) and H7 (3.99 ppm)-H8 (2.82 ppm) were found. Exo position of the C3 arylisoxazoline substituent was proposed based on the proximity of H3 (2.23 ppm) and H8. Carbon atoms C2 (82.9) and C7 (87.9) attached to the oxygen atoms were identified by ¹H NMR and the 2D HSQC and HMBC spectra. C2 showed cross peaks with the adjacent methylene group protons at 3.53 and 3.07 ppm, H3, and H4a,b protons (1.90 and 0.79 ppm). On the other hand, C7 exhibited correlations with H8, H9b (1.33 ppm), and H3.

Rationalization of the observed site selectivity and regioselectivity

Table 5 gives electron charges at the alkenyl carbon atoms of the dipolarophiles 5, 8, 10, 13, 18, 21, and the observed site-selectivity in the cycloaddition reaction.

Table 5
Electric charges at the alkenyl carbon atoms of the dipolarophiles 5, 8, 10, 13, 18, 21 and site-selectivity in the cycloaddition reaction

Dipolarophile/C atom	Electric charge
	C1	C2	C4/6	C3/C5	C8	C9/C12	C10	Site selectivity
5	-0.16	-0.11	-0.16	-0.144	-0.154	-0.106	-	C1/C2
8	-0.106	-0.159	-0.152	-0.154	-0.093	-0.234	-0.152	C1/C2
10	-	-	-0.104	-0.169	-0.086	-0.235	-	C8/C12
13	-0.159	-0.159	-0.159	-0.159	-	-0.159	-0.159	C9/C10
18	-0.150	-0.141	-0.150	-0.141	-	-	-	-
21 exo	-	-0.173	-	-0.176	-0.151	-0.113	-	C2/C3
21 endo	-	-0.172	-	-0.170	-0.145	-0.111	-	C2/C3

It was found that the amount of negative charges of both carbon atoms of the double bond correlated with the reactivity of the reaction. Cycloaddition to a-humulene (5) occurred mainly to the C1-C2 double bond with the greatest amount of electric charges (-0.303), and in the minor isomer the dipole added to one of the other two double bonds C4-C5 with a larger sum of electric charges (-0.263). Similarly, in cycloaddition to trans caryophylene (10) in the major adduct 11 more reactive was an exocyclic double bond C8-C12 with total electric charges of -0.321, while the other double bond had a charge of -0.273. The same trend was observed in case of 5-vinyl-2-norbornene (21), where higher electron densities were found in the C2-C3 bond (-0.349/-0.342, exo/endo isomers), than in the C8-C9 bond(-0.264/-0.256, exo/endo isomers). It correlated with a higher reactivity of the C2-C3 bond affording 41% of mono-adducts 22 and 23 to this bond, and 38% of bis-adducts 24-27 to C2-C3 and C8-C9 bonds.

Some discrepancy from this regularity was noticed for β-humulene (8), where the bis-adduct 9 was formed in reaction of the most electron-rich C8-C12 bond and the least electron rich C1-C2 double bond. It could be explained by the steric factors and a more difficult approach of the dipole to the hindered internal double bond of the dipolarophile flanked with two geminal methyl groups.

Biological activity of the products

Generally, only a modest or weak activity of the compounds synthesized in this project was recorded (Table 6). The highest fungicidal inhibitory potency was found against Botrytis cinerea strain of mono cycloadducts 6 and 19-20 pair. It was comparable to the activity of the reference compound, the commercial fungicide chlorothalonil.

Table 6
Fungicidal inhibitory activities of compounds 6, 7, 11, 12, 14, 15, 16, 17, 19, 20, 22, 23, 24, 25, 26, 27 at 200 mg L^-1

Compound	Linear growth inhibition / %
	A. a.	B. c.	F. c.	P. c.	Rh. s.
6	0	78.7	24.0	2.4	0
7	10	0	0	0	11
11	0	15.6	0	11.1	0
12	13	0	4.4	0	0
19 + 20	-	73.0	38.0	12.0	11.0
14 + 15	17.4	20.0	0	11.1	30.0
16 + 17	10.9	28.9	0	37.8	18.0
22	26.0	28.0	0	0	24.0
23	13.0	4.4	0	11.1	0
24	4.3	8.9	11.1	17.8	0
25	15.2	31.1	0	22.2	0
26	0	11.1	4.4	4.4	0
27	0	8.9	0	15.6	0
Chlorothalonila	-	80.0	38.0	61.0	88.0

• ^a
  Reference compound. A. a.: Alternaria alternate; B. c.: Botrytis cinerea; F. c.: Fusarium culmorum; P. c.: Phytophtora cactorum; Rh. s.: Rhizoctonia solani.

Analysis of the data from Table 6 allows concluding that generally exo regioisomers show a higher activity than the endo regioisomers. As an example isomer 6 was more active than isomer 7, and compounds 22 and 25 were more active than, respectively, cycloadducts 23 and 24. One can see very clearly the influence of the spatial structure on the fungicidal activity.

Conclusions

High regio- and site selectivity of [2 + 3] dipolar cycloaddition reaction of 4-trifluoromethylbenzonitrile oxide (4) to sesquiterpene 8 were observed. A correlation of double bond reactivity with electron charges was found in reactions of most dipolarophiles. The addition to the exocyclic double bond was favored. All diastereoisomeric products were fully characterized by ¹H and ¹³C NMR 1D and 2D spectroscopy. Some cycloadducts showed good fungistatic activity. Further research is in progress to analyze the biological potency of the new products, and to improve regioselectivity of the cycloaddition.

Supplementary Information

Supplementary data (¹H ROESY, COSY, ¹H-¹³C HSQC, ¹H-¹³C HMBC NMR spectra) are available free of charge at http://jbcs.sbq.org.br as PDF file.

https://minio.scielo.br/documentstore/1678-4790/9gcHpC6Gqmrg3dbJB3xTJYn/ab4fa84d76851ed1808dadcdc93950699de6e381.pdf

Acknowledgments

This work was supported in part by the Polish Ministry of Science and Higher Education (Research Grant EMC 902000031), which is gratefully acknowledged.

References

• ¹
  Bast, K.; Christl, M.; Huisgen, R.; Mack, W.; Sustman, R.; Chem. Ber 1973, 106, 3258.
• ²
  Caramella, P.; Grunanger, P. In 1,3-Dipolar Cycloaddition Chemistry; Padwa, A., ed.; Wiley: New York, 1984, pp. 177.
• ³
  Mukayama, T.; Hoshino, T.; J. Am. Chem. Soc. 1960, 82, 5339.
• ⁴
  Shimizu, T.; Hayashi, T.; Shibafuchi, H.; Teramura, K.; Bull. Chem. Soc. Jpn 1986, 59, 2827.
• ⁵
  Gothelf, K. V.; Jørgensen, K. A.; Chem. Rev 1998, 98, 863; Pellisier, H.; Tetrahedron 2007, 63, 3235; Kissane, M.; Maguire, A. R.; Chem. Soc. Rev 2010, 39, 845.
• ⁶
  Kanemasa, S.; Nishiuchi, M.; Kamimura, A.; Hori, K.; J. Am. Chem. Soc 1994, 116, 2324.
• ⁷
  Fisera, L.; Ondrus, V.; Timpe, H. J.; Collect. Czech. Chem. Commun 1990, 55, 512.
• ⁸
  Gucma, M.; Gołębiewski, W. M.; Krawczyk, M.; J. Braz. Chem. Soc 2013, 24, 805.
• ⁹
  Gucma, M.; Gołębiewski, W. M.; Michalczyk, A. K.; J. Mol. Struct 2014, 1060, 223.
• ¹⁰
  Coutouli-Argyropoulou, E.; Anastasopoulos, C.; J. Heterocycl. Chem 1996, 33, 731.
• ¹¹
  Baldwin, J. E.; Mayweg, A. V. W.; Neumann, K.; Pritchard, G. J.; Org. Lett 1999, 1933.
• ¹²
  Zhang, Y.-H.; Itoh, K.; Gao, D.; Chang, Y.; Cheng, J.-P.; Chow, Y. L.; Res. Chem. Intermed 2002, 28, 313.
• ¹³
  Chow, Y. L.; Wang, S.-S.; Johansson, C. I.; Liu, Z.-L.; J. Am. Chem. Soc 1996, 118, 11725.
• ¹⁴
  Nakayama, J.; Tajima, Y.; Xue-hua, P.; Sugihara, Y.; J. Am. Chem. Soc 2007, 129, 7250.
• ¹⁵
  Mach, K.; Antropiusová, H.; Petrusová, L.; Hanuš, V.; Tureček, F.; Sedmera, P.; Tetrahedron 1984, 40, 3295.
• ¹⁶
  Moriconi, E. J.; Hummel, C. F. J.; J. Org. Chem 1976, 41, 3583.
• ¹⁷
  Nozaki, H.; Kurita, M.; Mori, T.; Noyori, R.; Tetrahedron 1968, 24, 1821.
• ¹⁸
  Bak, D. A.; Brady, W. T. J.; J. Org. Chem 1979, 44, 107.
• ¹⁹
  Foley, A. D.; Maguire, A. R.; Tetrahedron 2010, 66, 1131.
• ²⁰
  Kikuchi, D.; Yoshida, M.; Shishido, K.; Synlett 2012, 4, 577.
• ²¹
  Liu, K. C.; Shelton, B. R.; Howe, R. K.; J. Org. Chem 1980, 45, 3916.
• ²²
  Hyperchem 7.5; Hypercube, Inc., Gainesville, USA, 2003.
• ²³
  Kanemasa, S.; Hayashi, T.; Yamamoto, H.; Wada, E.; Sakurai, T.; Bull. Chem. Soc. Jpn 1991, 64, 3274.
• ²⁴
  Mayo, P.; Hecnar, T.; Tam, W.; Tetrahedron 2001, 57, 5931.

Publication Dates

• Publication in this collection
  Nov 2016

History

• Received
  28 Jan 2016
• Accepted
  16 Mar 2016