A large number of pulses of the silver and gold nanoparticles
larger than the baseline were observed. This is because that a
cluster of metal ions were generated and analyzed by the mass spectrometer
when the NPs in diluted solution were introduced into the
plasma. However, the signal distribution was small and homogeneous
after filtration; few strong pluses were observed in Fig. 7 and Fig. 8.
The results in Table 6 showed that low recoveries of AgNPs and AuNPs
were obtained with these five types of filters. It indicated that the NPs
could not pass through these filters, although the NPs diameter was
smaller than the pore size of the nominal filter membrane. The NPs
were blocked to the membrane surfaces due to electrostatic attraction
(Hassellov et al., 2008; Mitrano et al., 2012). The nanoparticles which
were smaller than the pore size transported through the membrane
more slowly than the liquid did. This was because that the electrostatic
repulsion could lead to higher collision rates between particles and aggregation.
As a result high-efficiency trapping of nanoparticles and aggregates
occurred on the membranes (Hassellov et al., 2008). This
study indicated that filtration might be a problematic pre-treatment
method and it can't be applied in the sample preparation for the detection
of nanoparticle in natural water.