decoupler attenuation and 90° pulse for CPD [28 dB, 100 μs]
p1	90° ¹H decoupler pulse [3 dB], phase adjustment with respect to spin-lock pulse required
p2	waltz-16 ¹H decoupler spin-lock sequence with 50 μs 90° pulse [16 dB]
p3	waltz-16 ²⁹Si transmitter spin-lock sequence with 50 μs 90° pulse [11 dB]; length of both spin-lock sequences p2 = p3 = 1/J(Si,H) = 152 ms, corresponding to ²J(Si,H) = 7 Hz, determined by loop parameter of waltz-16 spin-lock sequence [l = 32]
d1	4 s
ns	1

5. Processing

Use standard 1D processing as described in Experiment 3.2 with exponential multiplication (lb = 1 Hz).

6. Result

The figure shows the 99.36 MHz ²⁹Si NMR spectrum obtained for TMS with an AMX-500 spectrometer using a multinuclear inverse probe-head. Spectrum a is a normal spectrum, obtained with one 90° pulse and ¹H CPD decoupling. Spectrum b was taken with the spin-lock polarization sequence under otherwise identical conditions. Note the improvement in the signal-to-noise ratio.

7. Comments

For liquids and using pulsed spin-locks it is especially easy to obtain the Hartmann–Hahn condition [Eq. (1)], because one simply has to adjust the power levels on both the proton and the X channel so that the 90° pulses have identical lengths. The corresponding equations are outlined in Experiment 2.9.

γ_HB₁ = γ_XB₂(1)

With the 90° pulse duration of 50 μs used here, a γB₁ is about 5 kHz, which seems to be sufficient in most ²⁹Si applications. The merits of the technique shown here are its insensitivity to the number of protons causing the polarization transfer, and thus, once adjusted, the method is more robust than INEPT or DEPT and yields better results in routine use. For silanes with directly attached protons the DEPT sequence should be applied.