td2: 2 k data points in F, td: 256 data points in F;

sw2: 1.5 ppm swl: 1.5 ppm o1: middle of ¹H NMR spectrum p!: 90° ¹H transmitter pulse p2: 45° ¹H transmitter pulse dl:2s initial value for ¢, evolution: 3 μs increment for ¢,; evolution: 1/[2-sw1]

5. Processing

Apply zero-filling in F, to 1 k real data points to obtain a symmetrical matrix of 1024x1024 real points. Use Gaussian windows in both dimensions. Apply real Fourier transformation corresponding to the TPPI mode of data acquisition in F\. Phase correction as described in Experiment 10.5 for phase-sensitive COSY spectra must be performed.

6. Result

The figure shows an expansion of the phase-sensitive COSY-45 spectrum obtained on an AC-300 spectrometer for the same two cross-peaks connecting H-3b with H-2 and H-3b with H-3a as given in Experiment 10.5. Note that specific contours are missing, and the tilt of the cross-peaks has a different slope, depending on the sign of the coupling: the cross-peak at 6, = 3.85 is caused by a (positive) vicinal spin coupling whereas the cross-peak at 5 = 3.35 is caused by a (negative) geminal spin coupling.

7. Comments

In a two-spin system the intensity of the cross-peaks is proportional to sin’, where B is the pulse angle of the second pulse in the COSY sequence. In multi-spin systems, however, the situation is more complex. In the three-spin system of the example shown, the cross-peaks are reduced from a 4x4 matrix to two 2x2 matrices, which are offset both in F, and F2 by the passive coupling. Thus, if the sign of active and passive couplings is the same, one obtains a different slope of the cross-peak compared with the situation when the signs differ.

The method shown here is also known as #-COSY, since the pulse angle of the second pulse may be varied widely. Compare the results with the E.COSY technique demonstrated in Experiment 10.7.