Abstract: An inductively coupled plasma, connected to the sampling cone of a mass spectrometer, is computationally investigated. The occurrence of rotational motion of the auxiliary and carrier gas flows is studied. The effects of operating parameters, i.e., applied power and gas flow rates, as well as geometrical parameters, i.e., sampler orifice diameter and injector inlet diameter, are investigated. Our calculations predict that at higher applied power the auxiliary and carrier gas flows inside the torch move more forward to the sampling cone, which is validated experimentally for the auxiliary gas flow, by means of an Elan 6000 ICP-MS. Furthermore, an increase of the gas flow rates can also modify the occurrence of rotational motion. This is especially true for the carrier gas flow rate, which has a more pronounced effect to reduce the backward motion than the flow rates of the auxiliary and cooling gas. Moreover, a larger sampler orifice (e.g., 2 mm instead of 1 mm) reduces the backward flow of the auxiliary gas path lines. Finally, according to our model, an injector inlet of 2 mm diameter causes more rotations in the carrier gas flow than an injector inlet diameter of 1.5 mm, which can be avoided again by changing the operating parameters.

Abstract: An inductively coupled plasma, connected to a mass spectrometer interface, is computationally investigated. The effect of pressure behind the sampler, injector gas flow rate, auxiliary gas flow rate, and applied power is studied. There seems to be an optimum range of injector gas flow rate for each setup which guaranties the presence and also a proper length of the central channel in the torch. Moreover, our modeling results show that for any specific purpose, it is possible to control that either only the central gas flow passes through the sampler orifice or that it is accompanied by the auxiliary gas flow. It was also found that depending on geometry, the variation of outgoing gas flow rate is much less than the variation of the injector gas flow rate and this causes a slightly higher pressure inside the torch. The general effect of increasing the applied power is a rise in the plasma temperature, which results in a higher ionization in the coil region. However, the negative effect is reducing the length of the cool central channel which is important to transfer the sample substances to the sampler. Using a proper applied power can enhance the efficiency of the system. Indeed, by changing the gas path lines, the power can control which flow (i.e., only from injector gas or also from the auxiliary gas) goes to the sampler orifice. Finally, as also reported from experiments in literature, the pressure behind the sampler has no dramatic effect on the plasma characteristics.

Abstract: An inductively coupled plasma (ICP) is analyzed by means of experiments and numerical simulation. Important plasma properties are analyzed, namely, the effective temperature inside the central channel and the mean flow velocity inside the plasma. Furthermore, the effect of torches with different injector diameters is studied by the model. The temperature inside the central channel is determined from the end-on collected line-to-background ratio in dependence of the injector gas flow rates. Within the limits of 3% deviation, the results of the simulation and the experiments are in good agreement in the range of flow rates relevant for the analysis of relatively large droplets, i.e., 50 μm. The deviation increases for higher gas flow rates but stays below 6% for all flow rates studied. The velocity of the gas inside the coil region was determined by side-on analyte emission measurements with single monodisperse droplet introduction and by the analysis of the injector gas path lines in the simulation. In the downstream region significantly higher velocities were found than in the upstream region in both the simulation and the experiment. The quantitative values show good agreement in the downstream region. In the upstream region, deviations were found in the absolute values which can be attributed to the flow conditions in that region and because the methods used for velocity determination are not fully consistent. Eddy structures are found in the simulated flow lines. These affect strongly the way taken by the path lines of the injector gas and they can explain the very long analytical signals found in the experiments at low flow rates. Simulations were performed for different injector diameters in order to find conditions where good analyte transport and optimum signals can be expected. The results clearly show the existence of a transition flow rate which marks the lower limit for effective analyte transport conditions through the plasma. A rule-of-thumb equation was extracted from the results from which the transition flow rate can be estimated for different injector diameters and different injector gas compositions.

Abstract: The present study analyses a specific laser ablation cell, the High Efficiency Aerosol Dispersion (HEAD) cell (see J. Pisonero et al., J. Anal. At. Spectrom., 2006, 21, 922931), by means of computational fluid dynamics (CFD) simulations. However, this cell consists of different modular parts, therefore, the results are probably of interest for the further development of other ablation cells. In the HEAD cell, the ablation spot is positioned below an orifice in the ceiling of the sample chamber. The particle transport through this orifice has been analysed for a ceiling height of 0.8 mm. The critical velocity for the onset of particle losses was found to be independent on the ejection angle at the crater spot. The deceleration of the particles can be described as the stopping in an effectively steady gas. Particle losses were negligible in this modular part of the cell at the evaluated laser ablation conditions. The transport efficiency through the Venturi chamber was investigated for different sample gas flow rates. In this case, it was found that small particles were predominantly lost at low flow rates, the large particles at higher flow rates. Making use of the simulation results, it was possible to design a modification of the HEAD cell that results in extremely short calculated washout times. The simulations yielded a signal of less than 10 ms, which was produced by more than 99% of the introduced mass.