The first step for determining particle size distribution was to sepa- rate the nanoparticle pulse from background signal. An iterative algo- rithm based on 5σ criterion (σ was the standard deviation of dataset) was used to discriminate particle pulse and background signal. The
Table 3
Possibilities of particle coincidence for different nanoparticle number concentration.
steps of the iterative algorithm were described as below: the average in- tensity (μ) and standard deviation (σ) of the whole dataset was calcu- lated, the data points exceeding (μ + 5σ) were considered nanoparticle signals and were removed. Calculation of (μ + 5σ) was then repeated multiple times, and the data above the new (μ + 5σ) value were removed until no further data points were removed. The re- maining dataset was considered as background signal. Tuoriniemi et al. pointed out that 5σ iterative algorithm criterion can ensure that contri- bution of false positives is less than 0.1% of the actual particle events and retain a sufficiently high particle count over a broad concentration range