High resolution NMR probes for liquids may contain parts near the coil consiting of the nuclei being observed. The parts give rise to background signals which can severely affect the NMR data. When observing11B, there is a background signal from boron containing parts near the coil and also the borosilicate glass in the NMR tube containing the sample.
Cory and Ritchey* introduced a very simple, clever method to suppress background signals in 1988. Their method uses a composite pulse, consisting of a 90° and two 180° pulses with appropriate phase cycling, in place of a conventional 90° pulse. The phase cycled composite pulse is essentially a 90° pulse for all spins inside the coil and 0° for all spins outside of the coil. An example of its implementation is shown in the figure below. The bottom traces show the 11B [1H] NMR spectra for a dilute sample of NaBH4 and a "real" synthetic sample on the left and right, respectively. One can see an enormous background signal from both the NMR probe and the NMR tube. In the case of the "real" synthetic sample, the information from the spectrum is difficult or impossible to recover. The top traces show similar spectra acquired using the composite pulse. The only background signal remaining is that from the portion of the NMR tube inside the coil. This pulse sequence (without proton decoupling) is in the Bruker pulse program library called "zgbs". It is not exclusive to 11B.
D.G. Cory and W.M. Ritchey. Journal of Magnetic Resonance, 80, 128 (1988).