“…Remaining phases are cycled according to φ 2 = y; φ 3 = 2(x), 2(−x); φ 4 = x, −x; φ 5 = y; φ 6 = x; φ receiver = x, 2(−x), x. Gradients are sine-bell shaped and have the following durations, peak amplitudes, and directions: G 1 , 0.5 ms, 7.5 G/cm, x; G 2 , 0.5 ms, 6 G/cm, y; G 3 , 0.5 ms, 5 G/cm, z; G 4 , 0.4 ms, 39.45 G/cm, xyz; G 5 , 0.3 ms, −4 G/cm, xy; G 6 , 0.3 ms, 5.5 G/cm, xy; G 7 , 0.2 ms, 16 G/cm, xyz. For each t 1 increment N-and P-type transients are collected alternately by inverting the polarity of G 4 along with pulse phases φ 5 and φ 6 . The two transients are stored separately and then added and subtracted to form the real and imaginary parts of a complex FID with a 90 • zero-order phase shift being added to one of the components.…”