ELIMINATION OF FREQUENCY DRIFT FROM FOURIER-TRANSFORM ION-CYCLOTRON RESONANCE MASS-SPECTRA BY DIGITAL QUADRATURE HETERODYNING - ULTRAHIGH MASS RESOLVING POWER FOR LASER-DESORBED MOLECULES
Sh. Guan et al., ELIMINATION OF FREQUENCY DRIFT FROM FOURIER-TRANSFORM ION-CYCLOTRON RESONANCE MASS-SPECTRA BY DIGITAL QUADRATURE HETERODYNING - ULTRAHIGH MASS RESOLVING POWER FOR LASER-DESORBED MOLECULES, Analytical chemistry, 65(24), 1993, pp. 3647-3653
At sufficiently low pressure, FT-ICR mass resolving power is no longer
pressure-limited. Rather, the observed spectral peaks are broadened b
y ion cyclotron frequency drift during the detection period, due to ch
ange in shape of the coherently orbiting ion packet during detection.
The frequency drift may be quantitated by Fourier transformation of ea
ch of a series of consecutive segments of the time-domain ICR signal,
followed by fitting the frequency vs time behavior to a polynomial in
time. Correction for that frequency drift is then achieved by a digita
l quadrature procedure, followed by multiplication by a weight factor
which removes the frequency drift. We demonstrate a 750-fold reduction
in FT-ICR mass spectral peak width for pseudomolecular (M + K)+ ions
of laser-desorbed leucine enkephalin (m/DELTAm = 1 300 000)! Moreover,
correction based on the frequency drift of ions of a given m/z also c
orrects for frequency drift of ions of other m/z values, as demonstrat
ed for isotopic peaks from (M + K)+ from gramicidin S (m/z 1179). Narr
owing of the FT-ICR mass spectral peaks results in a corresponding inc
rease in peak height-to-noise ratio as well. In addition, we propose a
theoretical model for frequency drift during detection of the ion cyc
lotron resonance signal. Simultaneous relaxation of coherent cyclotron
motion and compression of the axial distribution of an initially radi
ally coherent ion packet account for ion cyclotron frequency drift dur
ing detection. The potential energy generated by mutual ion-ion Coulom
b repulsions varies with ion cyclotron orbital radius as ions undergo
collisional damping. Excellent agreement with experimental frequency d
rift vs time is achieved with the four parameters (only one of which i
s adjustable) of the model: number of ions, initial cyclotron radius,
initial ion z-distribution length, and collisional damping time consta
nt.