Chemical shift referencing

From NESG Wiki
Jump to navigation Jump to search

This page describes UIPAC-Recommended referencing procedures. See and references therein. Also check the information at the CABM website

Proper chemical shift referencing is of key importance for matching resonances in various spectra. DSS serves as the direct 1H reference standard in liquid state NMR and other spins are referenced indirectly through gyromagnetic ratios. Common referencing methods are internal (~20 uM of DSS is added to the protein sample) and substitution (a separate sample with DSS is used). For substitution method a standard sucrose sample with trace DSS amount can be used.

Measuring zero frequency

To calibrate the spectra you need to know the frequency of the methyl peak of DSS in a 1D spectrum - this is the proton frequency at 0 ppm w0. Zero frequency w0 remains constant for a given spectrometer until its lock frequency (lockfreq on Varian spectrometers) is reset due to B0 drift. If you are not sure whether w0 has been determined for the current spectrometer setup measure in a 1D 1H spectrum of DSS.

  1. Buffalo.VNMR (Varian)
    • Record a standard 1H 1D spectrum of DSS sample and locate the DSS methyl peak.
    • Expand the region around it and use nl and movetof to shift 1H carrier to the peak maximum.
    • Invoke spcfrq to report w0 with a 7-digit precision
  1. Topspin (Bruker)
    • Record a standard 1H 1D spectrum of DSS sample and locate the DSS methyl peak.
    • Expand the region around it and use 'calibrate' to set this point to 0 ppm.
    • In the 'ProcPar' tab SF parameter will contain the w0 frequency.

To calculate zero frequencies for 13C and 15N spins use indirect chemical shift referencing ratios from BMRB

Temperature effects

The chemical shift of DSS is largely temperature-independent - after all that in in part why it was chosen as the referencing standard.

However, the exact B0 field will be different for the same sample and spectrometer, but at different temperatures. It happens because spectrometers are locked by adjusting z0, to make the deuterium line of the solvent match a certain predefined frequency. The chemical shift of water, in particular, has a very strong temperature dependence, which leads to B0 discrepancy.

Therefore, it is crucial that the zero frequency w0 is measured for a DSS sample at the same temperature as the target sample in the substitution method.

For temperature-scan experiments it may be convenient to have the spectra referenced to the water line and not to the DSS peak. Provide a table or formula for 1H chemical shift of water vs. temperature.

Calibrating Spectra in Various Software

Topspin (Bruker)

Copy zero frequencies for 1H, 13C and 15N into the SF fields of the 'ProcPar' tab.

PROSA and XEASY (manual calibration)

Varian Data

Get sfrq, dfrq and dfrq2 parameters with 7 decimal digits either by invoking spcfrq or parsing the ~/acqfil/procpar file.

Spectral widths in ppm are calculated as sw/w0. PROSA scripts use the parameter delta = 1/sw.

Maximum ppm shift of the direct dimension is calculated as

max. ppm = ( sfrq(MHz)*1000000 + sw/2 - w0(MHz)*1000000 ) / w0(MHz)

Indirect dimensions are calibrated similarly.

Bruker Data

Get SFO1, SFO2 and SFO3 parameters with 7 decimal digits from Topspin or by parsing the acqus file. Spectral width in the direct dimension in Hz can be read from the SW_h parameter.

Indirect spectral widths are determined by the incremented delays. Typically IN0 is used for the first indirect dimension and IN10 is used for the second. Usually 2 or 4 incremented delays (D0 and d10) used to sample indirect dimensions determined by parameters ND0 and ND10, respectively. See the pulse sequence file for the particular setup.

Thus the spectral width of the first indirect dimension in ppm is calculated as sw(ppm) = 1/(ND0*IN0). In PROSA delta = ND0*IN0.

Maximum ppm shift of the direct dimension is calculated as

max. ppm = ( w0(MHz)*1000000 + SW_h/2 - w0(MHz)*1000000 ) / w0(MHz)

Maximum ppm shift of the first indirect dimension is calculated as

max. ppm = ( w0(MHz)*1000000 + 1/(2*ND0*IN0) - w0(MHz)*1000000 ) / w0(MHz)


-- Main.GaohuaLiu - 24 Jan 2007