DANTE excitation sidebands

Selective excitation was accomplished by using a delays alternating with nutations for tailored excitation (DANTE) sequence of 20 short rf pulses of x =0.6 is duration [4]. The pulse separation used was 20 ps resulting in the total duration of the DANTE sequence of x,=400 [iS and the corresponding spectral excitation width A of the center band of 2.5 kHz. The excitation sidebands were separated from the centerband by (Xp) = 50 kHz and did not affect the NMR spectrum. After a recovery period,x of at least 35 ps a final 90° pulse was applied followed by the detection of the free induction decay. 

The constant amplitude pulses are usually easier to implement and they do not require special hardware. A typical example and also the most frequently used kind of constant amplitude, frequency selective pulse is the DANTE pulse train [4]. In the simplest case it consists of a sequence of square pulses interleaved with periods of free precession. Unfortimately, the excitation profile of the DANTE sequence has extensive sidelobes. It also produces sidebands at frequencies ... 

An early alternative to soft pulses was the DANTE Delays Alternating with Nutation for Tailored Excitation) experiment, which used a sequence of short, hard pulses of angle a <3C 90°, followed by a fixed delay t to achieve selective excitation. Thus, the pulse sequence is (a-T), ]. Nuclei that are on resonance are eventually driven to the y axis and hence are selected, whereas those more removed from the frequency range are not affected. The sequence of hard pulses can achieve a result similar to that of soft pulses and even can be shaped by modulating the duration of the pulse lengths, but DANTE pulses lead to spectral artifacts not created by soft pulses, such as unwanted sidebands. 

The major difference between soft shaped pulses and DANTE methods is the occurrence of strong sideband excitation windows either side of the principal window with DANTE. These occur at offsets from the transmitter at multiples of the hard-pulse frequency, 1/x. They arise from magnetisation vectors that are far from resonance and which process full circle during the x period. Since this behaviour is precisely equivalent to no precession, they are excited as if on-resonance. Further sidebands at 2/x, 3/x and so on also occur by virtue of trajectories completing multiple full circles during x. Such multisite excitation can at times be desirable [50,51] but if only a single excitation window is required, the hard pulse repetition frequency must be adjusted by varying x to ensure the sideband excitations do not coincide with other resonances. 

---