Single particle inductively coupled plasma mass spectrometry (SP-ICP-MS) is a method that allows to obtain size, size distribution, and particle number concentration of nanoparticles (NPs) in suspensions only after few minutes of measurement. However, several challenges of commercially available SP-ICP-MS instruments currently exist that limit the analytical performance of the method. This thesis reports novel developments and improvements of SP-ICP-MS with the ultimate goal to be able to use it as a routine NP analysis method in the future.
First, the state-of-the-art in SP-ICP-MS is reviewed on a step-by-step basis, from the sample introduction system to the detector. Applications of the method for the analysis of nanomaterials are critically discussed and current challenges are highlighted. Necessary improvements and directions for further developments are identified.
Second, capillary electrophoresis (CE) is coupled to SP-ICP-MS and used for the first time with a data aquistion system (DAQ) that provides microsecond time resolution (5 µs dwell time, µsDAQ) for the separation and characterization of mixtures of Ag NPs. An online preconcentration approach is implemented to decrease the detection limits to the sub-microgram-per-liter range. In addition, it is demonstrated for the first time that the optimized CE-SP-ICP-MS method can be successfully used to separate NPs with similar sizes but different surface coatings. Each component in a complex mixture of 20 nm, 40 nm, 60 nm sized citrate-coated and 40 nm, 60 nm sized PVP-coated Ag NPs can be distinguished.
Finally, a novel data processing algorithm for SP-ICP-MS with the µsDAQ is developed to extract NP signals from a continuous background signal. The method is based on Poisson statistics and allows to distinguish and quantify both NPs and dissolved elements. It is demonstrated that Ag NPs (20 nm – 100 nm) can be identified on a particle-by-particle level even in the presence of a significant concentration of ionic background (Ag+ up to 7.5 µg L-1, 107Ag+ up to 1 000 000 counts per second).