Reduction of Sideband Intensities in Adiabatic Decoupling Using Modulation Generated through Adiabatic R-Variation (MGAR)
✍ Scribed by Tsang-Lin Hwang; Michael Garwood; Alberto Tannus; Peter C.M. van Zul
- Publisher
- Elsevier Science
- Year
- 1996
- Tongue
- English
- Weight
- 143 KB
- Volume
- 121
- Category
- Article
- ISSN
- 1064-1858
No coin nor oath required. For personal study only.
✦ Synopsis
Recent reports on decoupling have shown that major im-Dykstra achieved sideband reduction by varying the lengths of the composite pulses in the decoupling sequence (12). provements in inversion quality (1, 2) and decoupling bandwidth (3-5) can be achieved when using adiabatic pulses. One problem with this latter approach is that the composite pulse lengths can only be varied within a few percent of the When compared to square-pulse decoupling, adiabatic wideband decoupling can achieve substantially reduced power nominal values, to avoid flip angle errors and changes in bandwidth (12,13). In this Communication, we show that deposition because the bandwidth of an adiabatic inversion pulse is proportional to the square of the RF amplitude func-this limitation does not apply to adiabatic decoupling and that sidebands can be effectively dispersed by incrementing tion B 1 (t), whereas this dependence is linear for a square pulse (4). It was recently shown that the complete 13 C band-the lengths of the adiabatic inversion pulses, while keeping the bandwidth and flip angle constant. The method is then width at 750 MHz ( 1 H) can easily be decoupled (3) when using so-called hyperbolic secant (HS) adiabatic pulses demonstrated by the application of HS pulses to 1 H-detected 13 C decoupling. The principle is however general and applies (6, 7). In addition to the average power requirements, other factors to be considered for decoupling are the availability to all types of adiabatic decoupling methods.
The inversion (and decoupling) bandwidth attainable for of peak power and the size of the J coupling. The interplay among these parameters and the adiabatic condition sets lim-adiabatic pulses is determined by the amplitude of the adiabatic frequency sweep, Dv max , which is an adjustable param-itations on the adiabatic pulse length. In a decoupling experiment, the pulsewidth T p is usually set equal to or smaller eter. For a HS pulse, the amplitude-modulated function B 1 (t) and frequency-modulated function Dv(t) are (7) than 1/(4J), to avoid the occurrence of large sidebands. In addition, sideband intensities can become appreciable when, in order to avoid sample heating, the gB 1 level is set low
[1] relative to the gB 1 level necessary for complete decoupling.
An approach to improve decoupling performance is repeti-
[2] tive application of so-called supercycle schemes (8). However, sidebands generally still occur because ideal decou-For most modern spectrometers, it is not straightforward to pling cannot be maintained within the supercycle scheme sweep the frequency, and the sweep is accomplished by (9-11). When multiple data points are acquired during this modulating the pulse phase, which is given by the integral interval, modulation of these data points results in sidebands of the frequency modulation. The time integral of Eq. [2] at multiple frequencies related to the cycling rate. To reduce yields the phase-modulation function sideband intensities, it is necessary to break down the repetitive pattern of supercycle schemes within a single scan. To accomplish this in composite decoupling sequences, Shaka
] et al. suggested cyclic permutation of the phases of a cycling scheme (10). Further reduction of sideband intensities can where 0 £ t £ T p , b Å 5.3, and f 0 Å Rp/2b. The parameter be achieved by time averaging. Asynchronous sampling was R is a convenient unitless constant to describe adiabatic inalso used to cancel sidebands which have random phases at version pulses since different scans (13). Starc ˇuk et al. (2) demonstrated the ''accordion'' approach to reduce the sideband intensities by inserting different delay times between the inversion pulses. R Å (Dv max )T p /p Å (bw)T p , [4] 221