The probe-head must be tuned to the sample used. Record a normal ¹H NMR spectrum and center the offset at the water resonance. Load the DPFGSE pulse program. You have to set:

td: 32k sw: 10 ppm o1: on water resonance p1: 90° ¹H transmitter pulse d2: 500 μs g1–g4: sinusoidal-shaped field gradients, 1 ms duration and 5% truncation, with gradient loop counters, ring-down delays (100 μs), lock blanking and gradient coil blanking switches according to the actual instrumentation used. Gradient strength ratio: 40 : 40: 7: 7. ds: 4

5. Processing

Use standard 1D processing as described in Experiment 3.1 with exponential multiplication (lb = 0.5 Hz) and a base-line correction.

6. Result

The figure shows the result obtained on an AMX-500 spectrometer with a BGU (10 A) gradient amplifier and a 5 mm inverse z-gradient probe-head. Compared with Experi- ment 11.16 the method shown here is superior in all respects.

7. Comments

The essence of the water suppression scheme is the same as described in Experiment 11.16. By an elegant matrix treatment given in [1] it can be shown that, by applying the WATERGATE sequence twice, all base-line distortions and phasing problems are eliminated as long as g1 and g2 are not correlated to g3 and g4. As described in Experiment 11.16, the length of the delay d2 decides the frequency positions of zero excitation. As an additional exercise you may try replacing the binomial excitation parts of the sequence by other schemes, such as the jump-and-return technique (Exp. 6.19), or using a selective 180° pulse on the water resonance together with a hard 180° pulse.