sw: 500 Hz
o1: above the frequency of the CHCl₃ signal
p1, p2: 90° ¹H transmitter pulse
d1: 20 s
d2: 1 ms
rg: receiver gain for correct ADC input
transmitter attenuation [3 dB]
ns: 1

Record one spectrum and check on all parameters. Use an automation routine which performs this experiment 64 times in sequence in order to have enough data for statistics.

5. Processing

Use standard 1D processing (see Exp. 3.1) applying an exponential window with a line-broadening factor lb = 1 Hz. Adjust the phase of the first spectrum roughly for dispersion and always use the same digital phase correction. The stability can be estimated from the standard deviations of the positive and negative peak heights.

6. Result

The figure shows the 64 dispersion signals obtained on an AMX-500 spectrometer with a 5 mm multinuclear probe-head. Note the severe dropout which occurred in spectrum #21, probably due to an external influence, since this was not reproducible. Neglecting this dropout, a statistical evaluation of the positive and negative intensities gave a standard deviation of 0.7% in amplitude, which corresponds to a phase error of 0.1°.

7. Comments

The first 90° pulse p1 aligns the magnetization towards the −y axis. Since the offset of the signal is 37 Hz from the transmitter and d2 was chosen to be 1 ms, the magnetization vector rotates about 13° from this axis. The second 90° pulse p2, if it also comes exactly from the x-direction, will leave 22.5% (sin 13°) of this signal in the x,y-plane, which is detected during the acquisition time. In the region of 13° the sine function is rather “linear”, and therefore phase deviations between the two r.f. pulses are faithfully transformed into an amplitude variation of the signal. This is best observed if the signals are displayed in dispersion mode.

Further variations and different processing possibilities for this experiment are given in the literature [2, 3].