Abstract
When performing volumeβlocalized spectroscopy measurements, the amount of spatial contamination is an important quality criterion. With the ISIS localization technique contamination cannot only arise from the transition regions around the volume of interest, but also from remote regions of the sample. The latter contamination component is a consequence of inhomogeneous excitation pulses, if short repetition times TR are used. Its severity depends both on the order of the eight phase cycling experiments needed for an ISIS measurement, and on the ratio T~R~/T~1~. Here it is theoretically discussed from which regions of the sample contamination can arise for a specific phase cycling order. For the worst orders the contaminating regions are almost three times as large as for the optimal orders. The ratio for the effectively measured contamination, however, can be moderated in real experiments, because cancellation effects occur due to the phase distribution of the contaminating signals. ^31^P phantom experiments clearly demonstrate that contamination is present even if adiabatic excitation pulses are applied and that spatial contamination can be reduced to about a third by an optimal choice of the phase cycling order.