Pure Shift F1 Proton-Coupled J-Scaled BIRD HSQC to Measure RDCs with Higher Accuracy and Sensitivity: Features and Limitations

ERICH HELLEMANN; MANUELA E. GARCIA; CHAKICHERLA GAYATHRI; ROBERTO R. GIL

Atlanta Georgia

Congreso; SMASH 2014, Small Molecules Are Still Hot NMR Conference; 2014

SMASH Organization

Due to their global orientation information content, residual dipolar couplings (RDCs) have shown significant impact on the constitutional, configurational and conformational structure determination of small organic compounds.[1,2] Among several alignment media compatible with organic solvents, we prefer to use compressed PMMA gels to partially align smalls organic molecules, as previously described by our group.[3] One of the most common problems when measuring RDCs is the significant signal broadening in F2 of HSQC experiments due to proton-proton dipolar couplings. This is the major reason why we avoid the use of F2 1H-coupled HSQC experiments to measure one-bond 1H,13C RDCs (1DCH). Instead, in 2011, we have started to measure these RDCs with the F1 1H-coupled version of HSQC experiments.[4] We later found that the J Scaled BIRD version of this experiment generated RDCs with higher accuracy.[5] However, the proton-proton signal broadening is still present in the experiment. In order to circumvent this problem we implemented proton homonuclear broadband decoupling capabilities (pure shift, PS) to the J Scaled (JS) BIRD-HSQC to boost resolution in F2 and overall experimental sensitivity. Pure shift is achieved during the acquisition of a single FID, without any special data processing, using trains of BIRD-based homonuclear decoupling.[6] We used strychnine (Figure 1) as model molecule and PMMA gels as alignment medium, to measure RDCs in order to demonstrate the advantages of implementing PS capabilities to the JS-BIRD-HSQC experiment. JS-BIRD-HSQC experiments with and without PS were collected on a sample of strychnine dissolved in CDCl3, in both isotropic and anisotropic condition. The addition of PS has introduced better resolution in F2 and better signal to noise (S/N) ratio in most of the cross-peaks. The increase in S/N is shown in Figure 2 and the resolution improvement in F2 is clearly shown in Figure 3. It is important to note, as shown in the Figure 3, that signal multiplicities collapse efficiently in F2 and allow easy extraction of data in anisotropic media without multiplets overlapping. This would be particularly useful in complex molecules with many proton-carbon correlation peaks. Due to ring tension limitations, only eight geometrically possible 3D diastereomeric structures of strychnine were generated and energy minimized by DFT. SVD fitting of the RDC data to these eight structures permitted the selection of the correct configuration (7R,8S,12S,13R,14R,16S). RDC data collected with the PS experiments produced a slightly better quality factor (Q) for the correct structure (0.058 vs 0.069), while the Q factor for RDCs collected using F2 1H-coupled CLIP-HSQC was much higher (0.247). Comparison of the structural selection power of RDCs measured in F1 with and without PS capabilities, as well as RDCs measured in F2 with the CLIP-HSQC, are shown in Figure 4. Being configuration 6 the correct one, the graph shows the advantage of PS over non-PS JS-BIRD-HSQC experiments, and a significant improvement in the Q factor respect to RDCs collected in F2 using the CLIP-HSQC experiment. In a realistic view of the outcome of this work, considering its features and limitations, it is important to highlight the significant resolution introduced by the PS capabilities in many signals, as well as a significant increase in S/N of the cross-peaks. Unfortunately, not all of the signals are benefited by the addition of PS; i.e., the BIRD decoupling cannot decoupled geminal protons, since both protons are attached to a 13C carbon, because the filter decouples protons attached to 12C from protons attached to 13C. In addition, the BIRD filter is tuned to decouple protons with J values within a normal range of 0-18 Hz, while the alignment medium introduces total proton-proton signal-splittings (nTHH=nJHH+nDHH) that can be much larger than what the BIRD filter can remove. We particularly appreciate the increase in S/N since the amount of sample that can be loaded in the gel is limited.