This was from a paper I recently wrote about ICP-OES.

This generally explains how an ICP-OES works (Inductively Coupled Plasma Optical Emission Spectroscopy).

The sample is mixed with the solvent gas (Ar) in the nebulizer which was where the sample is atomized. This then goes through a spray chamber to remove big droplets. The sample then moved to the ICP torch.

A torch consists of a coolant tube, which is the outer most tube, with the highest flow rate of Ar. The Auxiliary gas tube is the middle tube, both of these flow Argon gas for the plasma. The nebuliser tube carried the sample being measured. Most ICPs use radio frequencies from around 1.1 kW to 1.4 kW to produce a magnetic field to shape the plasma. The power is transferred to the ICP using a water-cooled Copper load coil (Hill, 2006).

The ICP is coupled with an OES.

This OES is a solid-state CID chip. CID are Charge-Injection Devices which are a type of Charge-Transfer Devices (CTD). CTD fundamentally work in three steps, first photons of the desired wavelength hit a photosensitive area and are converted to photoelectrons, these then move into a ‘storage area’ where they are kept for the integration time. Once integrated, they move to an output register and a readout is produced. CIDs consist of a two-dimensional array of detectors which, coupled with an Echelle spectrometer, can analyse multiple spectra lines simultaneously over the range 170-800 nm. (Hill, 2006).

Reference: