Abstract: A packed bed dielectric barrier discharge (DBD) is widely used for plasma catalysis applications, but the exact plasma characteristics in between the packing beads are far from understood. Therefore, we study here these plasma characteristics by means of fluid modelling and experimental observations using ICCD imaging, for packing materials with different dielectric constants. Our study reveals that a packed bed DBD reactor in dry air at atmospheric pressure may show three types of discharges, i.e. positive restrikes, filamentary microdischarges, which can also be localized between two packing beads, and surface discharges (so-called surface ionization
waves). Restrikes between the dielectric surfaces result in the formation of filamentary microdischarges, while surface charging creates electric field components parallel to the dielectric surfaces, leading to the formation of surface discharges. A transition in discharge mode occurs from surface discharges to local filamentary discharges between the packing beads when the dielectric constant of the packing rises from 5 to 1000. This may have implications for the efficiency of plasma catalytic gas treatment, because the catalyst activation may be limited by constraining the discharge to the contact points of the beads. The production of reactive species occurs most in the positive restrikes, the surface discharges and the local microdischarges in between the beads, and is less significant in the longer filamentary microdischarges. The faster streamer propagation and discharge development with higher dielectric constant of the packing beads leads to a faster production of reactive species. This study is of great interest for plasma catalysis, where packing beads with different dielectric constants are often used as supports for the catalytic materials. It allows us to better understand how different packing materials can influence the performance of packed bed plasma reactors for environmental applications.

Abstract: In recent years there has been growing interest in the use of plasma technology for CO2 conversion. To improve this application, a good insight into the underlying mechanisms is of great importance. This can be obtained from modeling the detailed plasma chemistry in order to understand the chemical reaction pathways leading to CO2 conversion (either in pure form or mixed with another gas). Moreover, in practice, several plasma reactor types are being investigated for CO2 conversion, so in addition it is essential to be able to model these reactor geometries so that their design can be improved, and the most energy efficient CO2 conversion can be achieved. Modeling the detailed plasma chemistry of CO2 conversion in complex reactors is, however, very time-consuming. This problem can be overcome by using a combination of two different types of model: 0D chemical reaction kinetics models are very suitable for describing the detailed plasma chemistry, while the characteristic features of different reactor geometries can be studied by 2D or 3D fluid models. In the first instance the latter can be developed in argon or helium with a simple chemistry to limit the calculation time; however, the ultimate aim is to implement the more complex CO2 chemistry in these models. In the present paper, examples will be given of both the 0D plasma chemistry models and the 2D and 3D fluid models for the most common plasma reactors used for CO2 conversion in order to emphasize the complementarity of both approaches. Furthermore, based on the modeling insights, the paper discusses the possibilities and limitations of plasma-based CO2 conversion in different types of plasma reactors, as well as what is needed to make further progress in this field.

Abstract: A packed bed dielectric barrier discharge plasma reactor is computationally studied with a fluid model. Two different complementary axisymmetric 2D geometries are used to mimic the intrinsic 3D problem. It is found that a packing enhances the electric field strength and electron temperature at the contact points of the dielectric material due to polarization of the beads by the applied potential. As a result, these contact points prove to be of direct importance to initiate the plasma. At low applied potential, the discharge stays at the contact points, and shows the properties of a Townsend discharge. When a high enough potential is applied, the plasma will be able to travel through the gaps in between the beads from wall to wall, forming a kind of glow discharge. Therefore, the inclusion of a so-called ‘channel of voids’ is indispensable in any type of packed bed modelling.

Abstract: The use of plasma technology for CO2 splitting is gaining increasing interest, but one of the major obstacles to date for industrial implementation is the considerable energy cost. We demonstrate that the introduction of a packing of dielectric zirconia (ZrO2) beads into a dielectric barrier discharge (DBD) plasma reactor can enhance the CO2 conversion and energy efficiency up to a factor 1.9 and 2.2, respectively, compared to that in a normal (unpacked) DBD reactor. We obtained a maximum conversion of 42 % and a maximum energy efficiency of 9.6 %. However, it is the ability of the packing to almost double both the conversion and the energy efficiency simultaneously at certain input parameters that makes it very promising. The improved conversion and energy efficiency can be explained by the higher values of the local electric field and electron energy near the contact points of the beads and the lower breakdown voltage, demonstrated by 2 D fluid modeling.

Abstract: In this paper, we investigate the etch process of so-called low-k organic material by means of a N2/H2 capacitively coupled plasma, as applied in the micro-electronics industry for the manufacturing of computer chips. In recent years, such an organic material has emerged as a possible alternative for replacing bulk SiO2 as a dielectric material in the back-end-of-line, because of the smaller parasitic capacity between adjacent conducting lines, and thus a faster propagation of the electrical signals throughout the chip. Numerical simulations with a hybrid plasma model, using an extensive plasma and surface chemistry set, as well as experiments are performed, focusing on the plasma properties as well as the actual etching process, to obtain a better insight into the underlying mechanisms. Furthermore, the effects of gas pressure, applied power and gas composition are investigated to try to optimize the etch process. In general, the plasma density reaches a maximum near the wafer edge due to the so-called 'edge effect'. As a result, the etch rate is not uniform but will also reach its maximum near the wafer edge. The pressure seems not to have a big effect. A higher power increases the etch rate, but the uniformity becomes (slightly) worse. The gas mixing ratio has no significant effect on the etch process, except when a pure H2 or N2 plasma is used, illustrating the synergistic effects of a N2/H2 plasma. In fact, our calculations reveal that the N2/H2 plasma entails an ion-enhanced etch process. The simulation results are in reasonable agreement with the experimental values. The microscopic etch profile shows the desired anisotropic shape under all conditions under study.