Fine-structure enhancement - assessment of a simple method to resolve overlapping bands in spectra

A simple mathematical procedure - fine-structure enhancement - has been ass essed on its ability to resolve overlapping bands in spectra. Its advantage s and limitations have been explored using synthetic and experimental spect ra. Fine-structure enhancement involves smoothing the original spectrum, mu ltiplying the smoothed spectrum with a weighting factor and subtracting thi s spectrum from the original spectrum. As a result, the fine-structure of t he original spectrum is enhanced in the processed spectrum and bands that o verlap in the original spectrum appear as distinct bands in the processed s pectrum. To be resolved by fine-structure enhancement, Lorentzian lines hav e to be separated by more than their quarter width at half maximum, Gaussia n lines by more than their half width at half maximum. A comparison of fine -structure enhancement and Fourier self-deconvolution shows that Fourier se lf-de convolution has in theory a higher potential to resolve overlapping b ands. However, this depends crucially on the correct choice of the paramete rs. In practice, when parameters commonly used are chosen for Fourier self- deconvolution, fine-structure enhancement leads to similar results. This is demonstrated at the example of the infrared absorbance spectrum of the pro tein papain, where the amide I band components could be resolved similarly with both methods. Thus, fine-structure enhancement seems to be a simple al ternative to Fourier self-deconvolution that does not require specialised s oftware. (C) 2000 Elsevier Science B.V. All rights reserved.