939 Research Article Received: 14 May 2008 Revised: 3 June 2008 Accepted: 11 June 2008 Published online in Wiley Interscience: 5 August 2008 (www.interscience.com) DOI 10.1002/mrc.2283 Selective HOESY experiments for stereochemical determinations Mehdi Yemloul, a Sabine Bouguet-Bonnet, a Lalla Aï cha Ba, b Gilbert Kirsch b and Daniel Canet a* Heteronuclear Overhauser effect spectroscopy (HOESY) is a powerful method for tracking geometrical proximities between two heteronuclei (for instance, 1 H and 13 C, as this will be the case here). The method is based on cross-relaxation arising from dipolar interactions. Sensitivity permitting, it is applied in the 2D mode yielding all spatial correlations in a single experiment. Whenever sensitivity is not sufficient, it can be applied in the one-dimensional mode by selectively inverting one particular proton. In that case, it yields, from the carbon-13 spectrum, remote spatial correlations. The method has been employed here for the discrimination between two possible E or Z isomers in a medium-size molecule. Copyright c  2008 John Wiley & Sons, Ltd. Keywords: carbon-13 NMR; nuclear Overhauser effect; selective HOESY; Solomon equations; build-up curves; heterocyclic chemistry Introduction Cross-relaxation between two spins indicates on an exclusive basis that they interact through a dipolar mechanism. Cross-relaxation measurements, which lead to the so-called nuclear Overhauser effect (nOe [1] ), can be performed in a two-dimensional mode in the homonuclear case (the well-known nuclear Overhauser effect spectroscopy (NOESY) experiment) or in the heteronuclear case (the less employed heteronuclear Overhauser effect spectroscopy (HOESY) experiment [2–4] ). An interesting application of the method concerns location of water with respect to molecules or systems in aqueous solutions. [5–8] For sensitivity reasons, it might be convenient to resort to selective HOESY experiments, which, in the one-dimensional mode, consist in perturbing the sole water protons while observing the responses arising from carbons in the molecule of interest. [6–8] However, the method does not seem to have been used for an arbitrary proton in a medium-size molecule so as to detect the proximities of carbons in the molecule under investigation. This article aims to detect correlations between a well-chosen proton and remote carbons so as to obtain stereochemical information. Of course, two-dimensional HOESY would provide the required information in a single experiment, but the lack of sensitivity might make it unsuccessful, either because a high-field spectrometer is not available or because concentration is not sufficient. Indirect detection [9,10] would not be helpful in that case since remote correlations imply necessarily small J-couplings. The total correlation spectroscopy (TOCSY)- HOESY concatenation advocated before [11] for the measurement of such weak responses has been disregarded here because of its lack of quantitative character. Indeed, we shall deal here with experimental methods, applied to medium-size molecules, which provide accurate cross-relaxation rates thus meaningful ratios of interatomic distances. These procedures will be shown to be of general applicability provided that the resonance of the selected proton does not belong to a crowded part of the spectrum (otherwise TOCSY-HOESY or two-dimensional spectroscopy has to be thought of). Here, with the help of this methodology, we shall tackle a stereochemical problem involving the discrimination between two E/Z isomers. Experimental Methods The sequences we have employed are sketched in Fig. 1. The first one (1a) is identical to an experiment proposed a long time ago [6] for measuring nOe enhancements arising from water protons. Here, as we selectively invert a normal proton resonance, the relevant proton is necessarily directly bonded to a carbon-12 and we will observe all remote nOe enhancements arising from this proton. Saturating carbon-13 prior to the mixing time and setting a recycle time longer than five times the proton T 1 insure that initial conditions are perfectly defined and that build-up curves start effectively from zero. The second step of the phase cycling applied to the delays alternating with nutations for tailored excitation (DANTE) pulse train is such that protons are not perturbed. As the result is subtracted from that of the first step, the effect of specific carbon-13 relaxation would be removed. This is true, however, only at short mixing times as discussed below. The second sequence (1b) is the classical non-selective one- dimensional HOESY experiment, the DANTE pulse train being substituted by a non-selective inversion or no inversion at all for the phase cycle second step. The purpose of the phase cycling is therefore the same as for sequence (1a). This sequence allows us to measure, at the level of the carbon-13 directly bonded to ∗ Correspondence to: Daniel Canet, M´ ethodologie RMN (UMR CNRS-UHP 7565), Nancy-Universit´ e, Facult´ e des Sciences et Techniques, B.P. 239, 54506 Vandoeuvre-l` es-Nancy (cedex), France. E-mail: Daniel.Canet@rmn.uhp-nancy.fr a M´ ethodologie RMN (UMR CNRS-UHP 7565), Nancy-Universit´ e, Facult´ e des Sciences et Techniques, B.P. 239, 54506 Vandoeuvre-l` es-Nancy (cedex), France b LIMBP (EA3472), Universit´ e de Metz, 1, Bd Arago, 57078 Metz-Technopˆ ole (cedex), France Magn. Reson. Chem. 2008, 46, 939–942 Copyright c  2008 John Wiley & Sons, Ltd.