Phasing in 3D spectra



Generally phasing is set up to be have 0 0 phase corrections for the indirect dimension

Alternatively, it can be 90 -180, if the sampling delay corresponds to exactly half the dwell time.

In most experiments the pulseprogramme will make this so.

However, for some of the Bruker canned experiments, particularly the 15N edited NOESY and TOCSY spectra this is not the case.
You can calculate the expected phases.


For the indirect 1H dimension in the NOESY experiment the sequence for evolution in t1 is:

        90(1H)---d0---180(15N)---d0----90(1H)

the effective sampling delay tau is therefore given by


4*90(1H)/pi  + 2*d0  + 180(15N)

in our syntax (brukers)

4*p1/pi   + 2*d0 + p22   


The linear phase is then given by (for states-tppi):
phase(1) = (tau/in0)  *180

As an example: for my NOESY experiment p1 = 16.67; p22 = 70; d0 = 3; in0 = 79.34

so tau = (4*16.67/pi) + 6 + 70
thus tau = 97.22

so the linear phase correction is  97.22/79.34 * 180

= 220.56

the zero order phase correction is half this: = -110.28

so input phc0 = -110.28
phc1 = 220.56

NOTE: you do not need to scale the first point

CAVEAT: the bruker spectrometer sometimes applies a phase correction (usually 90), it should not affect this unless you are processing in the echo anti-echo mode.
When I find the inclination I will change the setup so the phase correction is 0 0, but in the meantime this works.


For more info see:
Bax et al.. JMR 91, 174-178 (1991)
or me.