This thesis presents a fast semi-quantitative analysis method for ICP-OES (inductively coupled plasma optical emission spectroscopy) using only one calibration standard with one element and a plasma model. The same plasma model was used to mathematically correct ICP-OES matrix effects which can degrade concentration accuracy. Also, efforts are described to improve: 1) detection limits by increasing the amount of sample delivered to the plasma per time (using a heated close coupled spray chamber (CCSC) and an electronic nebulizer (eNeb)) and 2) precision by evaporating droplets before they enter the plasma (CCSC) or introducing a consistent flow of small droplets with a narrow size distribution (eNeb). An ICP generated using flat plates was compared to an ICP generated by a helical load coil. The feasibility of two ICPs with separate RF generators was assessed for potential future use with a very high sample transport rate.
ICP-OES typically requires the measurement of calibration standards that contain all of the elements of interest over a range of concentrations. A fast, semi-quantitative analysis method is described based on one calibration standard containing only one element and a pLTE (partial local thermodynamic equilibrium) plasma model for determining the concentrations of 66 elements by ICP-OES. The concentration accuracy is within a factor of three for 85% of 227 emission lines from 66 elements studied. The pLTE model can also be used to improve the accuracy of ICP-OES analysis by correcting for analyte sensitivity changes that result from changes in plasma temperature due to differences in chemical composition (matrix) between the sample and standards.
Sensitivity and detection limits could be improved by increasing the amount of sample delivered to the plasma per time (sample transport rate) if there is no change in plasma conditions or the ICP background. Two systems were evaluated for increasing the sample transport rate compared to a conventional sample introduction system: a heated close coupled spray chamber (CCSC) and a prototype vibrating mesh electronic nebulizer (eNeb).
The CCSC was used to increase the sample transport rate by a factor of four which resulted in improvements in sensitivities of four to nine times and detection limits of an average of a factor of five. Short term precision was improved by an average of a factor of two, perhaps due to the evaporation of large droplets before they enter the plasma. An increase in the velocity of the sample prior to entering the plasma was proposed to increase the sample transport rate able to enter the plasma without it being extinguished. At the highest sample transport rates the plasma was cooled and sensitivities decreased. A second plasma source is suggested to overcome this limitation. Initial results with a conventional sample introduction system showed a factor of three to seven times increase in end on sensitivity/background ratios but the same side on sensitivity to background ratios using the dual plasmas compared to a single plasma. It is hypothesized that it may be possible to increase the sample transport rate significantly compared to a conventional sample introduction system without cooling of the center of the plasma and decreasing sensitivities using dual plasmas.
The eNeb generates droplets by ejecting sample through a vibrating mesh of regularly spaced holes. Detection limits were up to four times improved compared to a conventional sample introduction system by electronically increasing the sample transport rate with the eNeb. Perhaps due to the consistent production of small droplets with a narrow size distribution the short term RSDs of analyte emission intensity averaged 0.21% (a factor of four to five improvement over a conventional sample introduction system).
Water droplets (generated using the eNeb) and water vapor with dried analyte particles (generated using the CCSC) were separately introduced into the plasma to compare the effect of each as the sample transport rate is increased. The maximum sample transport rate able to enter the plasma without it being extinguished was 75 mg/min of water vapor and 57 mg/min of water droplets. The smaller tolerated transport rate of water droplets tolerated may be due to increased cooling and movement downstream of the bottom of the plasma. Water droplets were also found to result in more cooling of the plasma center than the same transport rate of water vapor. However, sample introduced as water droplets resulted in a higher plasma temperature, sensitivity, and background intensity in the normal analytical zone.
Flat plates were investigated as an alternative to the helical load coil used to generate the plasma. The plasma generated using the flat plates can be operated at approximately 2/3 the argon gas flow and was found to tolerate more solvent. While the sensitivities for the plasma generated using the flat plates are the same to 9 times lower (likely due to the lower plasma temperature) the standard deviation in blank emission intensity is smaller for the flat plates generated plasma. This leads to detection limits that range from 6 times improved to three times worse.