¹³C NMR spectroscopy of heterocycles: 3,5-diaryl-4-bromoisoxazoles

Introduction

Historically, ¹³C NMR spectroscopy has been employed to study transmission in a number of aromatic and heteroaromatic systems [1]. In particular, a ¹³C NMR study of substituent effects [2] in 3,5-diarylisoxazoles has shown that substituents on 5-aryl groups substantially affect the chemical shift at C₄ of the isoxazole ring. However, transmission from substituents on the 3-aryl group of the same system has a greatly reduced effect (~1 order of magnitude) on the chemical shift of C₄ of the isoxazole ring. A subsequent ¹³C NMR study [3] of transmission in N-methylisoxazolium iodides, in which the effect of charge on transmission has been examined, concluded that transmission of substituent effects is essentially unchanged from that of the neutral isoxazole system, although the charged system shows a greater sensitivity. Typically, for single-parameter correlations, the best correlation is obtained with σ⁺ values for transmission of substituent effects. Dual substituent parameter approaches, in general, yield slightly better correlations with a greater dependence on the ρ resonance term than that of ρ inductive [2, 3]. These earlier studies have indicated that there are substantial resonance contributions to the transmission of substituent effects to the distal heterocyclic system. We were interested in evaluating distal transmission effects in a system that was no longer coplanar. In this study, we report the ¹³C NMR study for the 3,5-diarylisoxazole ring system containing a bulky bromo-substituent at position 4.

Results and discussion

The reaction of a series of 3,5-diarylisoxazoles 1–9 with N-bromosuccinimide in acetic acid [4] produced the corresponding 3,5-diaryl-4-bromoisoxazoles 1Br–9Brin low to moderate yield (Scheme 1). The compounds were purified by crystallization and characterized by spectral and physical methods.

Scheme 1

The structure of 3-(4-chlorophenyl)-5-phenyl-4-bromoisoxazole (2Br)was determined by X-ray diffraction. The compound was found to contain essentially three planar aromatic ring systems with torsion angles of ~40±2° and 36±2° between the 3- and 5-aryl groups and the central ring, respectively. This is consistent with expectations due to the steric bulk of the 4-bromo substituent. The structure is shown in Figure 1.

Figure 1

X-ray diffraction structure of 3-(4-chlorophenyl)-5-phenyl-4-bromoisoxazole (2Br).

(Top) Side view with ~40° tilt. (Bottom) End-on view.

Molecular modeling calculations on 3,5-diphenylisoxazole and 3,5-diphenyl-4-bromoisoxazole were carried out. As expected, the structure of 3,5-diphenylisoxazole (1) was calculated to be essentially coplanar with torsion angles between the components of ~0°. The structure of 3,5-diphenyl-4-bromoisoxazole (1Br) was calculated to be twisted with the 3-phenyl and 5-phenyl groups having torsion angles of ~50° and ~37° with the central ring, respectively. This is in general agreement with the X-ray diffraction structure (Figure 2).

Figure 2

Molecular modeling predicted structures for 1and 1Br.

(Top left) 3,5-Diphenylisoxazole (1)front view. (Bottom left) 3,5-Diphenylisoxazole (1) side view. (Top right) 3,5-Diphenyl-4-bromoisoxazole (1Br) front view. (Bottom right) 3,5-Diphenyl-4-bromoisoxazole (1Br) side view.

The ¹³C NMR spectra for the 4-bromo compounds, under standard conditions, are very similar to those for the corresponding isoxazoles with the major difference observed for the C₄ signal: upfield shift and loss of intensity. The chemical shifts for C₄ of the 3,5-diaryl-4-bromoisoxazoles with substituents (Y) only on the 3-aryl group are all well within experimental error of that found for the parent compound 1Br of the series (89.4 ppm). The signal for C₄ of the 3,5-diaryl-4-bromoisoxazoles with substituents X on the 5-aryl group is deshielded for electron-withdrawing groups and shielded for electron-donating groups. The ¹³C NMR data are listed in Table 1.

A plot of C₄ chemical shifts of 3,5-diaryl-4-bromoisoxazoles 1Br–9Br vs. those for C₄ of the corresponding isoxazoles 1–9 yields a good linear correlation (r² = 0.974) with a slope of 0.96 (Figure 3).

Figure 3

Plot of chemical shifts for δC₄ in 3,5-diarylisoxazoles vs. those for the 3,5-diaryl-4-bromoisoxazoles.

As the results showed that substituents on the 3-aryl group had essentially no effect on the chemical shift of C₄, the remaining data were also evaluated with a single-parameter approach using the constants for only the 5-side substituents. Despite the limited number (6) of data points, an excellent correlation (r= 0.994) was obtained vs. σ⁺ constants (ρ = 2.28) as opposed to σ constants (r= 0.94). Surprisingly, despite the substantial change in torsion angles for the 4-bromo derivatives vs. the isoxazole system, the 4-bromoisoxazole results still correlate with a large resonance contribution with an apparent ρ value ~92% of that for the isoxazole system [2]. As overlap should be decreased in the 4-bromo derivatives, the changes in C4 chemical shifts are likely due to subtle changes in bond angle(s) and/or bond length(s). Chemical shift calculations will be undertaken to gain additional insights into the mechanism(s) of distal transmissions.

Experimental

Chemicals were purchased from commercial sources and used without additional purification. Melting points were determined with a Stuart SMP10 apparatus and are uncorrected. ¹H NMR spectra were obtained at 23°C in CDCl₃ at 60 MHz on a Nanalysis 60 MHz FT spectrometer. ¹³C NMR spectra were obtained in DMSO-d₆ at 100 MHz on a Bruker 400 spectrometer at 50°C and referenced to the solvent at 39.50 ppm. Chemical shift data error is ±0.15 ppm. ¹³C NMR chemical shift data for C₄ position of the isoxazole system are reported to two decimal places to reduce rounding errors in correlations. Molecular modeling studies were conducted using the Spartan ’14 MMFF force field. X-ray structure determination of 2Br (single crystal, colorless monoclinic prisms – space group P2₁/c) was performed at Emory University on a Bruker APEX-11 CCD diffractometer; solved as a four-component system with one-fourth occupancy, each with all four C-Br bonds in the same direction (two-component twin in which the molecular stack within each twin had the C-Br in the same direction, but the 3-5 and 5-3 aryl groups were stacked alternatively). Exact mass data (ESI, M⁺+1) were obtained at Georgia State University.

The 3,5-diarylisoxazoles 1–9 were prepared from the corresponding chalcone dibromides by the published method [2, 5]. The experimental data for 1–7 are in good agreement with the published values (¹³C chemical shifts for C₄ at 50° for 1 = 98.45; for 2 = 98.36; for 3 = 98.19; for 4 = 98.09; for 5 = 98.93; for 6 = 96.74; for 7 = 96.79; for 8 = 97.97; for 9 = 97.77). Isoxazoles 8 and 9 have not been previously reported.