Monte Carlo Simulation of Low-Count Signals in Time-of-Flight Mass Spectrometry and Its Application to Single-Particle Detection

Many modern time-of-flight mass spectrometry (TOFMS) instruments use fast analog-to-digital conversion (ADC) with high-speed digitizers to record mass spectra with extended dynamic range (compared to time-to-digital conversion). The extended dynamic range offered by ADC detection is critical for accurate measurement of transient events. However, the use of ADC also increases the variance of the measurements by sampling the gain statistics of electron multipliers (EMs) used for detection. The influence of gain statistics on the shape of TOF signal distributions is especially pronounced at low count rates and is a major contributor to measurement variance. Here, we use Monte Carlo methods to simulate low-ion-count TOFMS signals as a function of Poisson statistics and the measured pulse-height distribution (PHD) of the EM detection system. We find that a compound Poisson distribution calculated via Monte Carlo simulation effectively describes the shape of measured TOFMS signals. Additionally, we apply Monte Carlo simulation results to single-particle inductively coupled plasma (sp-ICP) TOFMS analysis. We demonstrate that subtraction of modeled TOFMS signals can be used to quantitatively uncover particle-signal distributions buried beneath dissolved-signal backgrounds. On the basis of simulated signal distributions, we also calculate new critical values (L_C) that are used as decision thresholds for the detection of discrete particles. This new detection criterion better accounts for the shape of dissolved signal distributions and therefore provides more robust identification of single particles with ICP-TOFMS.