Direct detection is the process of recording chemical shift information by directly detecting the nucleus of interest. Indirect detection involves transferring magnetisation from a less sensitive nucleus to a more sensitive one to enhance sensitivity. Using indirect detection, sensitivity is increased by the ratio of the gyromagnetic ratios of the nuclei involved. In the case of 1H and 13C this means a nearly fourfold increase in sensitivity.
γ1H = 2.6752 x 108 T-1 s-1
γ13C = 6.728 x 107 T-1 s-1
γ1H/γ13C = 3.976
To obtain the greatest sensitivity increase the directly detected nucleus should be the highest frequency available, which in practice means detecting 1H (3H has a higher gyromagnetic ratio, but it is radioactive). Since detection will nearly always be via 1H, all our probes are constructed with the 1H coil inside the X nucleus coil to maximise sensitivity. This is another reason why I recommend recording 2Ds to obtain 13C chemical shifts.
To demonstrate the increase in sensitivity provided by indirect detection I recorded a pair of indirectly detected 2D 1H-13C correlation experiments in the same time used to record a 1D 13C direct detect experiment. All three spectra were recorded at 600MHz using a 1.7mm TXI probe and a 2mM standard sample of sucrose in 10% D2O (27.4µg of sucrose). The 1D 1H-13C spectrum was acquired with a 30o pulse, a relaxation delay of 0.5s and 4500 scans for a total time of 87 minutes. The HSQC was acquired with 4 scans, 160 t1 increments and a relaxation delay of 2 seconds for a total acquisition time of 25.5 minutes. And the HMBC was acquired using 8 scans, 250 t1 increments and a relaxation delay of 2s for a total acquisition time of 60.5 minutes.
The 1D 13C spectrum shows no signals from the sucrose, while the 2D spectra show all the expected signals.
The HSQC experiment correlates hydrogen atoms with their directly attached carbon atoms, while the HMBC experiment correlates hydrogen atoms with carbon atoms within two, three or four bonds. Since the HSQC spectrum only shows signals from hydrogens directly attached to carbons, the chemical shifts of quaternary carbons must be obtained from the HMBC spectrum. For example, the quaternary 2' carbon in the fructose ring does not show any signals in the HSQC, but in the HMBC shows two correlations at δ13C = 103.4ppm. The HMBC peak at 5.40,103.4 correlates the fructose C2' with the glucose H1 resonance at δ1H = 5.40ppm, thus identifying the quarternary carbon and confirming the linkage of the two rings.
Recording the indirectly-detected HSQC and HMBC is not only more sensitive than recording a 1D directly-detected 13C spectrum, it also provides more information. The 2D spectra correlate the hydrogens and carbons facilitating assignment of the resonances. It is also possible to record the HSQC spectrum in such a way that the CH2 crosspeaks are of opposite phase to the others. In the HSQC spectrum above the negative CH2 peaks are red.