核磁共振定量NMR 實驗

　　實驗考慮：

　　Good signal to noise 好的信燥比

　　90° pulse sufficiently short to excite the entire region uniformly

　　Delay between pulses of at least 5 T1’s 脈沖延遲至少T1的5倍

　　Flat base line to start with and baseline correction 基線平坦和基線校正

　　Filter bandwidth set to at least twice the sw

　　Enough data points to give 15 to 20 points per peak - remember that you lose half the points to the imaginaries in Fourier transformation

　　Acquire at least 32k and zero fill several times – memory is cheap

　　Exception to 32k for other nuclei requiring 1H decoupling - it is important to limit the acquisition time to as short as possible to avoid NOE build up during the acquisition so < 32k may be necessary with zero filling to make up the difference

　　Zero filling should only be done after dc correction and line broadening otherwise it introduces baseline artifacts

　　Gated decoupling for non-hydrogen analysis

　　Careful phasing on a spectrum that has been greatly expanded in the vertical direction

　　Careful integration

　　Place a single integral over the spectral limits of interest

　　Adjust bias and slope on the full integral

　　Cut integral into smaller regions as close as possible to the peak

　　Consistently include or exclude satellites or account for them when they overlap