Refocused INEPT is useful for nuclides with low natural abundance and low gyromagnetic ratio. The enhancement factors η for different nuclides in comparison with the NOE enhancement factors are as follows:

Nuclide	¹³C	¹⁵N	²⁹Si	⁵⁷Fe	¹⁰³Rh	¹⁰⁹Ag	¹¹⁹Sn	¹⁸³W
η(NOE)	2.99	−3.94	−1.52	16.48	−16.89	−9.75	−0.41	13.02
η(INEPT)	3.98	9.87	5.03	30.95	31.77	21.50	2.81	24.04

For this reason, INEPT has often been used for the observation of ¹⁵N and ²⁹Si, however compare Exp. 9.4. Especially in the case of ²⁹Si special attention has to be paid to the last delay d3, which controls the optimum polarization transfer.

The optimum value of this delay is a function of the scalar coupling constant J and the number of coupled nuclei n that are responsible for the polarization transfer [1] and is given by Equation (1):

d3_opt = (πJ)⁻¹ arcsin(n)^−1/2(1)

As n increases the value of d3_opt decreases and the enhancement factor E becomes more sensitive to a variation in d3.

Number of protons n	1	2	3	6	9	12
Enhancement E	5.03	5.03	5.82	7.83	9.44	10.82
Delay d3_opt (in units of J⁻¹)	0.5	0.25	0.196	0.134	0.108	0.093

The advantage of INEPT, like all other polarization transfer methods, is that the pulse repetition time of the experiment is dictated by the spin–lattice relaxation time of the protons rather than that of the nuclide under observation, here ¹³C. The disadvantage of the INEPT sequence is its sensitivity towards both delays d2 and d3. Carbon nuclei with widely different C,H spin coupling constants can give signals with lower intensity or even the wrong sign. Therefore the DEPT sequence is more often used, since the choice of the delays is not as critical (see Exp. 6.9).

8. Own Observations