Abstract: In order to solve the mutual interference problem among homogeneous peaks in Time-of-flight and Mass Spectrometry (TOF-MS) quantitative analysis, a spectral peak separation method was presented. First, according to the technical characteristics that those spectral peaks with close mass number have close shape, the peak positions were determined by the correlation algorithm between Gaussian curve and the original mass spectrum. Then, the individual single peaks were selected by the principle of “high amplitudes kept and low amplitudes left”. The characteristic peak was obtained by stacking and normalization of the single peaks. The power spectral density function was constructed by the characteristic peak, and positions and amplitudes of the overlapped peak were determined in accordance with the minimum principle, thus achieving the overlapped peak separation. The result obtained by the above characteristic peak matching method was compared with the results of Gaussian matching method and Lorentzian matching method respectively. When dealing with symmetric peaks with δ= 0.2, these results are basically consistent. However, when dealing with asymmetric peaks with δ= 1.6, the effect of characteristic peak matching method is much better than the other two methods. Experimental results show that the presented method can be used to separate the ²⁹Si and ²⁸Si¹H overlapped peak, the separation resolution is improved from 0.371 to 0.519, while the original shape information of the peeks is maintained.

Key words: information processing, time-of-flight mass spectrometry(TOF-MS), characteristic peak, peak separation