The acquisition time

In an NMR experiment the acquisition time is the period used to record the signal. Changing this value can affect the quality and appearance of your spectrum. Typically, larger acquisition times are used for one-dimensional spectra, with shorter values used in multidimensional spectra. The impact of using different acquisition times on 1D ¹H spectra are shown in this post.

To record the spectra a halogenated cholestane in CDCl₃ was used. Five spectra were collected with the acquisition time initially set to 0.143 seconds and then doubled for each successive spectrum. The spectra were processed identically with exponential line broadening of 0.3 Hz and zero-filling to 64K points. A stackplot of the spectra showing the upfield region with the methyl peaks is shown below.

The first thing to notice is that at shorter acquisition times the peaks are broader and resolution is reduced. The doublets at 0.88 ppm are only resolved when the acquisition time is at least 0.570 s. Other multiplets (0.75, 1.30, 1.54 ppm) also show reduced resolution at the shorter acquisition times. When longer acquisition times are used the signal has time to decay to near zero, but with shorter acquisition times the signal is still significant at the end. During processing, a window function is used to scale the data so that it is close to zero at the end of the acquisition time. This reduces truncation artifacts, but it makes the signals appear to decay more rapidly than they actually do. The increased apparent decay rate translates into broader peaks and a loss of resolution.

The other effect of the reduced acquisition time is the artifacts in the baseline, particularly obvious above 1.00 ppm. These "sinc wiggles" are due to the presence of significant signal at the end of the acquisition time. The fourier transform converts the intensity difference between the signal and the zeros used for zero filling into a sinc function in the baseline.

Based on these examples one might assume that the longer the acquisition time the better, but in fact this is not true. The figure below shows the signal-to-noise ratio, measured using the peak at 0.68 ppm, plotted against the acquisition time. The signal-to-noise rapidly increases to a maximum and then slowly decreases.

Signal-to-noise in NMR experiments has long been known to be maximal near 1.26T₂¹. This is because the noise is constant, and will continue to increase if recorded for longer, but the signal decays. After a certain point in the acquisition time, 1.26T₂, the signal has decayed sufficiently that the noise is greater. Increasing the acquisition time beyond 1.26T₂ will provide increased resolution but at the expense of sensitivity and the cost of increased experimental time. 

For quick 1D experiments, where resolution of multiplets is important, longer acquisition times are typically used. For multi-dimensional experiments where crosspeak fine structure is not important, but total experiment time is, acquisition times are kept short. This also has the benefit of reducing the size of the processed multidimensional data. At the Skaggs NMR Facility the acquisition time is typically set to 1.140s for 1D ¹H spectra, 0.865s for 1D ¹³C spectra, and 0.285s for 2D and 3D spectra.

References
1. Rovnyak D.
The past, present, and future of 1.26T₂.
Concepts Magn Reson Part A. 2019; 47A:e21473

Acknowledgments
Thanks to Prof Ted Molinski for providing the cholestane sample.