First the ¹H and ¹³C offsets of the sample have to be determined. The spectrometer is set up for ¹H observation with ¹³C decoupling (inverse mode on older instruments). The desired shape of the decoupler pulse and its length have to be selected. For this experiment a Gaussian shaped pulse of 10 ms length was chosen. You have to set:

decoupler attenuation for soft pulse, initial value 80 dB, to be varied
td	4 k
sw	500 Hz
o1	on ¹H resonance
o2	on ¹³C resonance
p1	90° ¹H transmitter pulse
p2	¹H spin-lock pulse for ¹H decoupling, same length as p4, typical attenuation 12 dB, effective phase must be the same as for the hard pulse p1, adjust if necessary.
p3	leave at zero for the determination of the correct attenuation of the shaped pulse. For phase determination, set to hard 90° ¹³C decoupler pulse.
p4	shaped ¹³C decoupler pulse, 10 ms
d1	10 s
d2	1/[2J(C,H)] = 2.33 ms, calculated from ¹J(C,H) = 215 Hz
rg	receiver gain for correct ADC input
ns	1

In order not to change the selectivity you have to determine the effect of the shaped pulse as a function of the decoupler attenuation. With a very high decoupler attenuation (ca. 80 dB) adjust the phase of the C,H doublet to an antiphase pattern. Then repeat the experiment with different decoupler attenuations until the satellites disappear, which corresponds to the soft 90° decoupler pulse. To obtain the correct phase of the soft decoupler pulse one introduces another hard 90° decoupler pulse before the soft pulse in the pulse program. If both have the same phase, they are additive and yield an antiphase pattern with opposite phases as adjusted before.

5. Processing

Use standard 1D processing as described in Experiment 3.1.

6. Result

The figure shows spectra obtained on an AMX-500 spectrometer. a is the initial spectrum with high decoupler attenuation and in b the effect of a shaped 90° decoupler pulse is shown. In c p3 was set to 90° and the phase of the shaped pulse was adjusted correctly in order to form, in combination with p3, a 180° pulse.

7. Comments

This calibration experiment follows the description given in a recent publication [1]. In earlier work, due to phase problems, it was found very difficult to obtain a correct calibration. With the spin-lock pulse p2 these problems are removed. It serves two