Determination of lead and iodine species by capillary electrophoresis inductively coupled plasma mass spectrometry.

Thesis (Ph. D.)--University of Texas at Austin, 2000. / Vita. Includes bibliographical references. Available also in a digital version from Dissertation Abstracts.

published_or_final_version / Chemistry / Doctoral / Doctor of Philosophy

The research in this thesis aims to characterize the signal-production processes of single particles in the ICP using time-resolved ICP-AES. Signal-production processes, including desolvation, vaporization, atomization, ionization, and diffusion, determine the temporal emission intensity of a single particle. Bimetallic nanoparticles of BaTiO3 (average diameter = 115 nm) were used as test particles. The particles were introduced into the ICP by nebulization of the suspension of the particles in water. As the ion plume of a particle moves up in central channel of the ICP, a temporal emission peak of the analyte atoms in the plume is produced. The emission intensity at any point of time in the temporal profile is related to the degree of vaporization and excitation of the particle at the corresponding vertical position of the ICP. The signal-production processes can, in principle, be studied by measuring the temporal emission profiles. However, the emission intensity of single particles is typically low. Continuous integration of the entire ICP central channel further reduces the signal-to-background ratio (SBR). A novel double-slit method has been developed to measure the temporal emission intensity of a single particle at two pre-defined ICP vertical positions. Two horizontal slits of slit height of 1 mm were placed in front of the monochromator. As the ion plume passes through the double-slit, two peaks in the temporal emission profile are produced. The configuration of the double-slit (slit height and distance between the two slits) was optimized for maximum signal-to-noise ratio (SNR) and temporal resolution of the double-peaks. Fast data sampling rate (50,000 Hz) was used in proper sampling of the temporal emission peaks. Large data sets were obtained. Custom programs were developed to extract the relatively weak double-peaks from the temporal emission profiles. The data treatment strategy includes smoothing of the temporal profile to increase SNR and automated peak extraction based on the characteristics of the double-peaks (peak height, peak width, time-difference of the peak pair, and SNR). Four smoothing methods, including Moving Average Filtering, Savitzky-Golay Filtering, Fast Fourier Transform (FFT) and Wavelet Transform, were tested. FFT was adopted because the method requires only one parameter (the cutoff frequency) and is relatively easy to optimize. Hundreds of double-peaks were obtained in a typical temporal profile of time duration of approximately 120 s. The emission intensity and peak width of the peak pair are correlated to determine the degree of vaporization of the analyte atoms, the extent of diffusion of the analyte atoms and the plume size, and the velocity of the plume in the ICP. Two types of double-peaks are identified. The relative peak height and peak width of the double-peaks in each type are related to the degree of vaporization of the single particles. Simulation of the evaporation rate of water droplets that enclose the single BaTiO3 particles shows that the time required for complete evaporation of water is a major factor that determines the degree of vaporization of BaTiO3 particles at the double-slit. Aggregation of BaTiO3 particles in the suspension was also investigated. / published_or_final_version / Chemistry / Master / Master of Philosophy

The technique of single-cell analysis using time-resolved inductively coupled plasma-mass spectrometry has been characterized and optimized. Determination of the metal contents of individual cells provides data on the natural metal contents of the cells and the corresponding distributions in the population. The distribution is a useful indicator of the health and the state of development of the cells. The contents of sorbed metals of individual cells over a duration of time are required to understand the dynamics of metal-cell interactions. A green alga, Chlorella vulgaris, was used as a model biological cell in this study. The criteria and procedures for proper sampling of the cells into the ICP will be discussed. Ideally, each ICP-MS spike corresponds to one cell, but cell overlapping occurs because the cells enter the ICP randomly. Selection of cell number density and sample uptake rate to minimize spike overlapping will be discussed. A cell counting method based on the frequency of the spikes has been developed. The distribution of the metal contents of cells was determined by measuring large number of spikes. The minimum number of spikes required was determined by statistical analysis. The spike intensity distribution was correlated with the size distribution of the cells. The peak maximum of the spike intensity distribution was used for the determination of the average metal content of the cells. The use of the peak maximum reduces errors due to spike overlapping in the measurement. Quantitative determination of the metal contents was achieved using standard particles for calibration. Errors in calibration using standard solution nebulization were discussed. The technique was applied in the study of metal-cell interactions. Sorption of heavy metal ions (as environmental pollutants) by Chlorella vulgaris, and uptake of biometal (as nutrient) and metallodrug (as toxin) by Helicobacter pylori were studied. The technique requires simple sample preparation of removing the culture medium by filtration or centrifugation. The health state of the cells in the presence of toxic metals was related to the change in cell number density. The ratio of the FWHM of the spike intensity distributions of the sorbed metals to the natural metal contents of the cells is identified as a possible indicator of the location of the sorbed metals. The kinetics of metal sorption by the cells can be studied using a single cell culture. The method reduces errors due to uncertainties in cell number density and metal concentration in multiple samples that are required in conventional methods. The optimal ICP-MS sampling depth of 17 elements, introduced into the ICP by conventional solution nebulization of aqueous standard solutions, has been determined. The elements were selected to represent a wide range of boiling points and ionization potentials. Boiling point of the dried residues and ionization potential of the analyte element were identified as the major factors that determine the optimal sampling position. Since dried sample solution aerosols are effectively nanoparticles, the study provides useful insight on the optimization of the operation conditions and calibration strategies for single-particle analysis using ICP-MS. / published_or_final_version / Chemistry / Doctoral / Doctor of Philosophy

Time-resolved Inductively Coupled Plasma –Mass Spectrometry (ICP-MS) is a versatile tool for the analysis of single particles such as air particles, nanoparticles, and biological cells. In this study, the processes of particle vaporization and analyte atom diffusion and ionization in the ICP were investigated using computer simulation. Gold nanoparticles of particle diameter 10 to 250 nm were used as the model particle. The parameters of the model were optimized with respect to the experimental data. The relative importance of these parameters was investigated. Simulated ICP-MS intensity versus sampling depth for different particle size was calculated. Two models of particle vaporization, namely heat-transfer-limited and mass-transfer-limited, were adopted to describe the kinetics of vaporization of the gold nanoparticles. The rate of particle vaporization of the limiting model in each 5-µs time step was used in the simulation. The heat-transfer-limited process dominates at lower position of the ICP. The mass-transfer-limited process takes over at sampling depth of 4mm or above where the ICP temperature is higher than 4000K. The simulation assumed that the gold atoms vaporized from the particle in each time step diffuse independently. The number density of the gold atoms was calculated using the Chapman-Enskog diffusion theory for each subsequent time step. The degree of ionization of the gold atoms was estimated using Saha equation and was assumed to be dependent on the plasma temperature only. The simulated ICP-MS intensity at any instant was the sum of the gold ions in the ion plumes from all previous time steps that pass through a 1-mm sampler cone. The effects of several simulation parameters on the calculated ICP-MS intensity were investigated. The simulation depth profile of ICP-MS intensity of 100-nm gold nanoparticle was compared to the experimental ICP-MS depth profile. The ICP-MS intensity depends strongly on the ionization temperature of the plasma and the evaporation coefficient of the analyte. The ICP temperature profile, gas velocity, ionization temperature and evaporation coefficient were optimized for the best fit of simulated results to the experimental data. Simulated calibration curves of gold nanoparticles of nominal diameter of 10 nm to 250 nm are non-linear at any sampling depth. The calibration curve rolls off at high mass due to incomplete vaporization of the larger particles in the ICP. The calibration curve at high sampling depth concaves upward in the low mass range because of significant diffusion loss of the analyte atoms for the small particles. / published_or_final_version / Chemistry / Master / Master of Philosophy

abstract / Chemistry / Doctoral / Doctor of Philosophy

Transient emission of a particle in inductively coupled plasma-atomic emission spectrometry (ICP-AES) depends on the fundamental processes of aerosol desolvation, particle vaporization and atomization, ionization, excitation and diffusion of the analyte. Ideally, the rate of the above processes can be determined from the evolution of the transient emission as the ion plume moves along the central channel of the ICP. However, the dimension of the ion plume is significantly smaller than the central channel. The signal-to-background and signal-to-noise ratios suffer when the entire channel is imaged. Deconvolution of the temporal profile is required to determine the emission intensity of the ion plume versus observation height. Small aperture can be used to locate the vertical emission position accurately, but the evolution of the plume emission is lost. In this study, a double-slit method has been developed to pin-point two vertical positions of an ion plume. An ion plume travelling along the ICP central channel produces two peaks in the temporal emission profile. The temporal evolution of emission intensity can be correlated to delineate the degree of particle vaporization at the two positions. The relative widths and separation of the two peaks in a double-peak can be used to determine the analyte diffusion rate and particle velocity in the ICP, respectively. An unicellular green algae, chlorella vulgaris, was used as the test particles. The average Mg content of the algae is equivalent to MgO particles of diameter of 265nm. The strong ionic emission at wavelength of 279.55nm was monitored using a ¼ -m monochromator equipped with a PMT detector. Method of curve fitting was used to filter out the noise with minimum distortion of the peak shape for accurate determination of peak height and peak width. The merits of curve fitting versus methods of smoothing such as moving average and Savitzky-Golay filtering will be discussed. All transient emissions from the algal cells were detected with sufficient signal-to-noise ratio using a single-slit setup with slit height of 1mm at observation height of 18 mm above the load coil and ICP forward power of 1400 W. However, using the double-slit setup, less than half of the expected double-peaks were observed. One of the peaks in the double-peak can be below the detection limit and the double-peak is lost. An innovative development of this study is that the relative sensitivity corresponding to the 2 slits can be varied to enhance the intensity of the weaker emission peak. The peak with insufficient signal-to-noise ratio for detection can be enhanced to a level above the limit of detection. The number of observed double-peaks in increased and the observed particles are more representative of the population. Two types of double-peaks are categorized according to the relative intensity of the first peak to the second peak. A computer model was used to estimate the intensity ratio of the two emission peaks at different observation position of the ICP. The experimental and theoretical ratios agree generally. The theoretical ratio also shows the bias in the population sampled by the double-slit setup. / published_or_final_version / Chemistry / Master / Master of Philosophy

This research focuses on electrothermal vaporization (ETV) as a sample introduction source for inductively coupled plasma mass spectrometry (ICPMS). ETV creates a dry plasma that causes problems when used at the high applied powers (e.g., 1.2 kW) typically employed for nebulizer-based sample introduction for certain ICPMS instruments. A secondary discharge forms in the sampling region of the spectrometer, but this effect was removed by reducing the applied power (e.g., 0.7 kW). A novel, steady state, dry aerosol introduction system was developed to permit optimization of the ICPMS settings. The device used solid NbF₅, SnBr₄ and a W filament plated with Pb to generate dry aerosols which produced ²⁸Si⁺, ⁷⁹Br⁺, ¹²⁰Sn⁺, ¹⁸⁴W⁺, and ²⁰⁸Pb⁺ that were used to optimize the ICPMS. When compared to an ICPMS optimized using a nebulizer, the dry optimized plasma produced an average enhancement of 4.5(±0.4) for 26 elements when using ETV sample introduction. The ETV produces a short (0.5-2 s) transient pulse once every 2-3 min, resulting in a sample throughput of 20-30 samples/h. To increase this throughput, a low power, low background multiplexed ETV device using carbon braids as vaporizers was developed. Oxygen ashing was demonstrated with recoveries of 96±17% for a suite of elements. Refractory elements (e.g., V and U) showed precision of greater than 25% while medium to high volatility elements were in the 10-20% range. The lifetime of the braids was limited to ca. 30 vaporizations when heated to 2,800°C. A device for coupling an ETV in parallel with a nebulizer was also designed and characterized. The device was designed to minimize the impact of the ETV's presence on nebulizer-based ICPMS performance. The ETV could be easily switched on line to provide complimentary information that may be unavailable with nebulizer sample introduction because of isobaric interference problems. For example, in a 1% HCl matrix, the detection limits for ⁵¹V⁺ (⁵¹ClO⁺ interference), ⁷⁵As⁺ (⁷⁵ArO⁺ interference) and ⁷⁸Se⁺ were found to be 0.008 ppb, 0.088 ppb, 0.063 ppb, respectively. By contrast, the nebulizer detection limits in 1% HCl for ⁵¹V⁺, ⁷⁵As and ⁷⁸Se were found to be 0.593 ppb 1.488 ppb and 1.158 ppb, respectively. / text

Fundamental and applied aspects of ICP-MS have been investigated to gain an increased understanding of the technique and improve on its analytical capabilities. Dissociation temperatures of polyatomic ions were calculated using a double-focusing sector instrument, to obtain more reliable mass spectral data with controlled vapour introduction via a Dreschel bottle to allow accurate calculation of the ingredients in the plasma. The equilibrium temperature for the plasma, operated at 1280 W calculated using CO*, and as the thermometric probes, was c.a. 5800 - 7400 K, while using ArO* and ArC* as the thermometric probes the temperature calculated was c.a 2000 - 7000 K. Calculated dissociation temperatures were used to elucidate the site of formation of these ions. Results confirmed that strongly bound ions such as CO* and C2* were formed in the plasma whereas weakly bound ions such as ArO* and ArC* were formed in the interface region due to gross deviation of the calculated temperatures from those expected for a system in thermal equilibrium. The use of helium gas in a hexapole collision cell Attenuated the signals of ArH* Ar* ArO*, Arc*, ArCl* and Ara* allowing improved determination of ^^K*, *'Ca*, ^^e* ^^Cr*, ''As* and ^°Se*in standard solutions. The use of the hexapole collision cell also resulted in an enhancement of analyte signals due to the thermalisation of the ion beam. The ion kinetic energy of ions sampled from the plasma and those sampled from the skimmer cone were determined using a modified lens stack to assess the significance for memory effects of material deposited on the skimmer cone. The most probable kinetic energy of Be* ions sampled from the skimmer cone was found to be 2A eV, which was considerably lower than the most probable kinetic energy of Be* ions sampled from the plasma, which was found to be 9.5 eV. The low kinetic energy of the ions deposited on the skimmer cone means they will only contribute to the analytical signal under certain instrumental operating conditions. The feasibility of liquid sample introduction into a LP-ICP-MS system designed for gaseous sample introduction was investigated using a particle beam separator. The low signal was attributed to the low gas kinetic temperature of the plasma which was confirmed by the fact that the signal increased rapidly with increasing temperature of the transfer line between the particle beam separator and the LP-ICP torch. This was also supported by the fact that more volatile compounds gave mass spectra whereas less volatile compounds did not. A limit of detection of 30 mg 1'^ for chlorobenzene was achieved. Finally, silicon and phosphorus speciation was performed by HPLC coupled to sectorfield ICP-MS. Silicones ranging in molecular weight from 162 g mol'^ - 16500 g mol"^ were extracted from spiked human plasma and separated by size exclusion chromatography. Limits of detection ranged from 12 ng ml"' Si* for the 162 gmol'^ silicone to 30 ng ml'' Si* for the 16500 g mol' silicone. Organophosphate pesticides were extracted from spiked plasma and separated by reversed phase chromatography. Recoveries were between 55 - 81 %. Limits of detection were 0.9 ng ml'' P* 1.8 ng ml'' P* 1.6 ng ml"' P* and 3.0 ng ml'' P* for dichlorvos, methyl parathion, malathion and quinolphos respectively. Phosphates were extracted from various food products and separated by ion-exchange chromatography. Limits of detection were 1.0 ng ml"' P* 2.3 ng ml"' P*, and 39 ng ml"' P* for P04^", PaOy"^ and PsOio^" respectively.