*** Outline ***
The C program harminv-1.3.1 and the two Java applications harmonic_inversion.jar and nmrsvd2.jar are already described in harmonic inversion for improving spectrum resolution
We extend their applications to NMR probe head dead time correction.
79Br MAS FID signal and its spectrum are observed for setting the magic angle of the NMR probe.
- D. L. Bryce, G. M. Bernard, M. Gee, M. D. Lumsden, K. Eichele, and R. Wasylishen
Practical aspects of modern routine solid-state multinuclear magnetic resonance spectroscopy: one-dimensional experiments,
Can. J. Anal. Sci. Spectrosc. 46, 46-82 (2001).
- J. S. Frye and G. E. Maciel
Setting the magic angle using a quadrupolar nuclide,
J. Magn. Reson. 48, 125-131 (1982).
The acquisition parameters for one-pulse MAS experiment on KBr powder: 10-kHz rotor spinning speed; SWH = 500 kHz; dead time DE = 0.01 msec; DW (TopSpin definition) = 0.001 msec; TD (TopSpin definition) = 16384; TDeff = 16248; qsim AQ_mode. (Its TopSpin file 20.zip)
Figure 1: 79Br MAS spectrum baseline is distorted due to the DE = 0.01 msec dead time of NMR probe head, duration between the end of the RF excitation pulse and the beginning of the FID signal sampling.
We decrease the noise of the FID signal with SVD (singular value decomposition) method implemented in the Java application nmrsvd2.jar, using 1500 complex points and 26 singular values. As a result the FID is denoised but truncated.
Figure 2: Fourier transform of denoised 79Br MAS FID. Due to truncation of the FID by SVD treatment, outer spinning sidebands are missing.
We submit the truncated, denoised FID to the C program harminv-1.3.1.
Figure 3: Result provided by harminv-1.3.1 on truncated, denoised 79Br MAS FID.
We simulate the FID with data provided by harminv-1.3.1.
Figure 4: Fourier transform of simulated 79Br MAS FID from truncated, denoised one. Outer missing spinning sidebands are not recovered
We simulate a new FID with previous data provided by harminv-1.3.1, including 0.01 msec (or five complex points) dead time correction.
Figure 5: Harmonic_inversion.jar panel in dead time correction mode.
After dead time correction, the spectrum has a flat baseline.
Figure 6: Fourier transform of simulated 79Br MAS FID after dead time correction.
We apply backward linear prediction on the one-pulse MAS FID simulated with harminv-1.3.1, whose spectrum is shown in Figure 4.
Bruker: Linear prediction
Figure 7: Backward linear prediction parameters applied to simulated 79Br MAS FID for dead time correction.
Figure 8: Fourier transform of simulated 79Br MAS FID after dead time correction with backward linear prediction.