Data Processing

wft2da is the command to process the full 2D for phase sensitive spectra (almost all 2D spectra are phase sensitive except gHMBC and gCOSY), but often data is massaged before full processing. wft2d is the command to process gHMBC and gCOSY. wft(1) processes the first fid for the purpose of setting phasing and weighting.

Relevant Processing Parameters:

fn, fn1, sb, sbs, gf, gfs, lb, awc, sb1, sbs1, gf1, gfs1, lb1, awc1, proc, proc1, rp, lp, rp1, lp1, ssfilter


Zero Fill-

The parameters fn and fn1 are parameters that affect how many points are added on that were not acquired (zero filling). Fn should be set to equal or twice np parameter. fn = np or fn = 2*np. fn = 2*np will give a very slight improvement in resolution over fn = np. fn1 should be set to at least twice ni, four times ni will improve resolution a little more. If you want to have a square 2D matrix, fn = fn1. Note that processing will take alot longer if fn and fn1 are large (2048 or more).


Linear Prediction-

Linear prediction can improve resolution slightly if set correctly, however, the data may not process at all if set incorrectly. The macro setLP1 will correctly set linear prediction parameters for t1 dimension. Note that the macro might increase fn and fn1 and make the data slow to process; they can be reset to original values.


Apodization (Weighting)-

2D data is often multiplied by mathematical functions before Fourier transformation to improve signal to noise or resolution. In vnmr this is called weighting the fid (hence wft rather than ft). Weighting functions should be
adjusted to make the data look as good as possible. They can be adjusted by clicking on the menu button "adjust weighting". Weighting the first part of the fid results in greater signal to noise at the expense of resolution, weighting the right side of the fid results in greater resolution but might lead to negative contours on the edge of the positive contours and also poorer signal to noise. Weighting functions should be adjusted in both dimensions. Probably the most important part is being sure the weighting function goes to zero at the end of the fid, if it does not, there will be noise in the 2D. Different weighting functions are optimal for different experiments, especially magnitude spectra such as gCOSY, and gHMBC.

For phase sensitive experiements (all except gCOSY and gHMBC), a good way to set weighting is:
sb = -at and sbs = sb for F2
sb1 = -ni/sw1 and sbs1 = sb1 for F1 (note that ni is the amount of acquired points in the t1 dimension so if you set ni = 128 but abort the experiment after 64 increments sb1 should be set to 64/sw1 not 128/sw1.

For gCOSY or gHMBC set sb = -at/2 sb1 = -ni/(sw1*2) sbs = sbs1 = 'n'

There are macros to appropriately set these weighting functions:
sqcosine will set a cosine squared function in both dimensions (for all spectra except gCOSY and gHMBC)
sqsine will set a sine squared function in both dimensions (for gCOSY and gHMBC)


Phasing-

The detect dimension (t2) can be phased by phasing the 1st block (wft(1)) then adjusting the rp and lp (right phase and left phase), click on the menu button phase to and adjust with the mouse

both the indirect detect and the direct detect dimension can be phased after processing the full 2D; by putting the cursor on a peak and typing ds (display slice) when the t1 dimension (or now F1 dimension) is along the horizontal axis. then a 1D slice will appear and can be phased by clicking the phase button. if there is an error such as cannot phase this data, then check pmode parameter; pmode should be set to 'full' rather than partial. to make the horizontal axis the t1 dimension type trace ='f1'. (the reverse is trace = 'f2'). Also, t2 can be rephased if necessary as well by the same process as t1 so trace would have to be set to f2 (trace='f2')


Solvent Suppression-

Solvent can be suppressed during processing if necessary. set ssfilter = 10 then wft2da, if that is not enough try ssfilter up to 500. Normally solvent suppression is unnecessary so ssfilter should be set equal to no (ssfilter =
'n')


Baseline Correction

There are two types of baseline correction dc (drift correct) and bc (baseline correct)

first try drift correct: type dc2d('f2')

if that does not work, reprocess the data, type wft(1) then set the integrals as you normally would then type wft2da then bc('f2',0) will correct baseline shifts across a spectrum in 'f2'