Development of a dimensionless parameter for characterization of dielectric barrier discharge devices with respect to geometrical features

Transcription

1 2017 Hefei Institutes of Physical Science, Chinese Academy of Sciences and IOP Publishing Printed in China and the UK Plasma Science and Technology Plasma Sci. Technol. 19 (2017) (10pp) Development of a dimensionless parameter for characterization of dielectric barrier discharge devices with respect to geometrical features Mook Tzeng LIM ( 林木森 ) 1,3, Ahmad Zulazlan SHAH ZULKIFLI 1, Kanesh Kumar JAYAPALAN 2 and Oihoong CHIN 2 1 Fuels and Combustion, Generation Unit, TNB Research Sdn. Bhd., Malaysia 2 Plasma Technology Research Center, Physics Department, Universiti Malaya, Malaysia mook.tzeng@tnb.com.my and mooktzeng.lim@live.com Received 17 January 2017, revised 12 May 2017 Accepted for publication 16 May 2017 Published 26 July 2017 Abstract Non-thermal plasma (NTP) devices produce excited and radical species that have higher energy levels than their ground state and are utilized for various applications. There are various types of NTP devices, with dielectric barrier discharge (DBD) reactors being widely used. These DBD devices vary in geometrical configuration and operating parameters, making a comparison of their performance in terms of discharge power characteristics difficult. Therefore, this study proposes a dimensionless parameter that is related to the geometrical features, and is a function of the discharge power with respect to the frequency, voltage, and capacitance of a DBD. The dimensionless parameter, in the form of a ratio of the discharge energy per cycle to the gap capacitive energy, will be useful for engineers and designers to compare the energy characteristics of devices systematically, and could also be used for scaling up DBD devices. From the results in this experiment and from the literature, different DBD devices are categorized into three separate groups according to different levels of the energy ratio. The larger DBD devices have lower energy ratios due to their lower estimated surface discharge areas and capacitive reactance. Therefore, the devices can be categorized according to the energy ratio due to the effects of the geometrical features of the DBD devices, since it affects the surface discharge area and capacitance of the DBD. The DBD devices are also categorized into three separate groups using the Kriegseis factor, but the categorization is different from that of the energy ratio. Keywords: low temperature plasma, discharge power, energy ratio, dimensionless parameter, scaling factor, dielectric barrier discharge, geometrical effects (Some figures may appear in colour only in the online journal) 1. Introduction Non-thermal plasma (NTP) devices produce excited and radical species that have higher energy levels than their ground state [1, 2] and are utilized in various fields, including combustion [3 5], gasification [6], surface modification 3 Author to whom any correspondence should be addressed. [7 10], particulate matter mitigation [11 14] and wastewater treatment [15 18]. There are various types of NTP devices, with dielectric barrier discharge (DBD) reactors being widely used. The solid dielectric barrier is placed in the gap between the live and the ground electrode and the electrical discharge is maintained in the gas gap. When a sufficiently high electric field is applied between the two electrodes, the gas atoms/ molecules participate in collisional processes resulting in /17/ $

2 Figure 1. (a) Schematic diagram of the a NTP test rig; (b) sketch of the front view of an NCC NTP device; (c) sketch of the CC DBD reactor; (d) sketch of the TCC DBD reactor. The TCC reactor consists of three CC reactors. ionization, excitation, and/or dissociation to produce charged and excited chemical species. When the charged and excited chemical species recombine and de-excite, respectively, energy is released in the form of radiation ranging from ultraviolet to the near infrared wavelength region. Currently, there are various designs for DBD reactors. The discharge power from these reactors characterizes their ability and energy efficiency to produce charged and excited chemical species. The discharge power or energy is affected by various factors such as the voltage, frequency, geometrical features and capacitance of the device. The voltages used in these devices have been reported to vary from 10 to 20 kv. Similarly, the discharge gap varies from 1 to 5 mm, while the size of the electrodes are also different, all of which affect the capacitance of the system [19 22]. The type of the power source, either in the form of alternating current (AC) or pulsed direct current (DC), are known to affect the discharge and hence the energy characteristics of the DBD devices [23 27]. Due to the varying parameters and power sources, a direct comparison of the DBDs is usually not possible, making it difficult to discern their energy characteristics. There are existing parameters that can be used to describe the discharge characteristics and therefore differentiate one DBD device from another, such as the electron number density and the electron temperature [28 32], but these parameters are not available during the design stage of the NTP, when the relation between the diameter, discharge gap and the discharge power are required. Thus, the present study proposes a dimensionless parameter that relates the discharge power to the frequency, voltage, and capacitance of a DBD. The dimensionless parameter could be useful for engineers and designers to compare the energy characteristics of DBD devices systematically, and could also be used for scaling up of the devices. Such dimensionless parameters have been used in other fields of engineering, e.g. in fluid mechanics, heat transfer, and the scaling up of fluidized beds [33 35]. The dimensionless parameter for the devices in this study is then used for characterizing other DBD devices in the literature. The current study encompasses DBD devices that are connected to AC power supplies and sinusoidal power sources, so as to focus on the effects of geometrical features on the dimensionless parameter. The dimensionless parameter is also correlated with another factor in the literature (proposed by Kriegseis et al [36]) for further comparison, although the Kriegseis factor is a scaling factor and is not dimensionless. The Reynolds, Archimedes, Biot and Fourier numbers are just some examples of dimensionless parameters that describe the phenomena in fluid mechanics and heat transfer. The factor developed by Kriegseis et al does not include the effects of the geometrical features of the DBD devices, which can affect the capacitance of the system, as will be discussed further in the manuscript. 2. Methodology The present study investigates the discharge power from three different DBD set-ups: concentric (CC), tri-concentric (TCC) and non-concentric (NCC) DBD reactors. Operating data and geometrical features from other DBD devices in the literature are used for comparison. 2

3 Figure 2. Lissajous diagram of a discharge from the DBD device Dielectric barrier discharge reactor system set-up Figures 1(a) and (b) show a schematic diagram of a DBD device and the system set-up. The input power into the DBD device is supplied by a high voltage (HV) supply. The HV supply consists of a 2 kw VARIAC that controls the amount of power that is imparted to the load, while the duration of the power pulses is controlled by the duty cycle, which was limited to 60%. The frequency is tuned to the resonance operating frequency (27 36 khz) of each set-up. The discharge power for the various set-ups, voltages, and current parameters are determined from the Lissajous diagram (figure 2), which is a plot of the charge transferred across the electrodes, Q with respect to the input voltage, V in. A high voltage probe and oscilloscope were used to measure the input voltage V in, and its frequency f, to the DBD systems. The charge can be determined by measuring the voltage across a capacitor, V c, which is connected in series with the DBD device at the ground electrode. From V c, the charge accumulated in the capacitor is determined via Q = C c V c, where C c is the capacitance of the capacitor. The capacitance used for the CC and TCC set-ups were 0.1 μf, while the NCC set-up used a capacitance of 0.22 μf. The area under the curve of the Lissajous diagram gives the discharge energy per cycle. The respective slopes of the Lissajous diagram represent the total and dielectric capacitance, C T and C d (assuming the discharge fills up the entire gap), respectively (refer figure 2). The total capacitance of the set-up also consists of the gap capacitance, C g, and is determined according the following equation: C T = (1/C d + 1/C g ) DBD reactor configuration Figures 1(c) and (d) show the configuration of the CC and TCC reactors. The CC reactor is essentially a wire-cylinder DBD reactor [37]. It has an outer quartz tube (open-ended) with an outer diameter of 6 mm and thickness of 1 mm, and the inner tube (close-ended) has an outer diameter of 2 mm and a thickness of 0.5 mm. The outer electrode has a copper layer that is connected to the ground, while the inner electrode is a copper wire of 0.5 mm in diameter and is connected to the high voltage simulator. The TCC reactor consists of three CC reactors of similar dimensions. Figure 1(b) shows the configuration of the NCC reactor, which has a stainless steel outer shell (110 mm inner diameter) that acts as the ground electrode. Quartz tubes (with a thickness of 2 mm) were used as the dielectric barrier, and they were held in place by a Teflon holder. Inner quartz tubes of three different outer diameters, 2r oiq (14, 30 and 44 mm), were used with each diameter having a different number of tubes (10, 6 and 3, respectively). The inner surface of the quartz tubes were layered with copper sheets that were connected to the high voltage supply. The diameter of the live electrodes was taken to be the same as the inner diameter of the quartz tubes. The air gap between quartz tube and the outer shell was 1.0 mm. The length of the NCC was 200 mm, while the length for the CC and TCC set-ups were similar at 100 mm Dimensionless parameter energy ratio The following dimensionless parameter, energy ratio δ, ispro- posed for comparing the power characteristics of different DBDs: Pd f d = ( 1) 1 2C V eg in 2 where P d, is the discharge power; f is the frequency; V in is the voltage input into the DBD reactor; C eg is the estimated gap capacitance of the system. For simplicity, the estimated gap capacitance for concentric DBD reactors is estimated through the following equation: 2pLeo Ceg = ( 2) r ioq ln roiq where L is the length of the device, ε o is the permittivity in free space ( Fm 1 ), r ioq is the inner radii of the outer quartz tube and r oiq is the outer radii of the inner quartz tube of the DBD device. The data from this study and from the literature are mostly gas mixtures that are nitrogen dominant. Therefore the dielectric constant for the gas mixtures was assumed to have a value of unity. The estimated gap capacitance C eg is not to be confused with the gap capacitance C g that can be obtained from the Lissajous diagram with experimental measurements of the charge and voltage (refer to figure 2). During the design stage, when such measurements are unavailable, the gap capacitance could be estimated using equation (2) to determine the energy ratio. For the NCC and TCC DBD reactors, the gap capacitances are obtained from the Lissajous diagrams (refer to figure 2). The parameters Pd f and 1 2C eg V 2 in are related to the respective energy dissipated in the discharge per cycle and capacitive energy of the NTP test rig or DBD device. Therefore, the proposed energy ratio is indicative of the ability of each DBD device to dispense energy to produce charged and excited 3

4 Figure 3. Variation of the energy ratio with respect to the discharge gap d. species, as compared to the total amount of energy that could be stored within the capacitive gap of the DBD. Kriegseis et al [36] proposed the following factor for a plasma actuator: Pd L Q A = ( 3) 3.5 V f1.5 in where L is the length of the device. A dimensional analysis of equation (3) shows that the unit of factor Θ is W s 1.5 m 1 V 3.5 and it is not dimensionless, as the parameters used to describe phenomena in fluid mechanics and heat transfer mentioned earlier. Equation (3) is still applicable as a scaling factor, but is limited in its function to describe the characteristics of DBD devices. A dimensionless analysis of equation (1) shows that it is dimensionless and thus could be used for both scaling up and to describe the characteristics. Equations (1) and (3) were used to determine the energy ratio and the Kriegseis factor for the DBD devices in this study and those in the literature [14, 38 40]. A comparison was then made between the two factors. Table 1 shows a summary of the operating conditions of the DBD devices in this study and in the literature. Figure 4. (a) Variation of the energy ratio with respect to the inner diameter of the outer quartz tube 2r ioq. (b) Variation of the energy ratio with respect to the outer diameter of the inner quartz tube 2r oiq. 3. Results 3.1. Dimensionless parameter energy ratio Figures 3 to 9 show that the energy ratio is not a strong function of the input voltage, frequency, capacitance, discharge gap, and diameter for the DBD devices in this study and from the literature. For the same discharge gap, diameter and length (figures 3 5, respectively), the energy ratio varies significantly due to the frequency and the strength of the electric field. Figure 6 shows that for the same frequency, the energy ratio still varies due to the variation in the strength of Figure 5. Variation of the energy ratio with respect to the length L. 4

5 5 Discharge gap, d (mm) Table 1. Operating parameters of different DBD devices in this study and from literature. Inner diameter of outer quartz tube, 2r ioq (m) Outer diameter of inner quartz tube, 2r oiq (m) Length, L (m) Frequency, f (khz) Gap capacitance, C g (pf) Peak-to-peak voltage, V pp (kv) Non concentric a 0.014, 0.030, a (NCC) DBD a Concentric (CC) DBD Tri-concentric b (TCC) DBD b Wang et al [40] c Fang et al [38] c Xia et al [39] Babaie et al [14] Discharge power, P d (W) a The NCC set-up does not have an outer quartz tube, only a ground electrode that is an outer stainless steel shell with an inner diameter of m. The gap between the inner quartz tube and the outer ground electrode is 1.0 mm. The 0.014, and m system has ten, six and three live electrodes and dielectric barriers respectively that are arranged in a circular manner around the circumference of the outer shell. b The TCC is a bundle of three CC DBD reactors, each positioned separately at 120 in a full circle. Only the outer diameter of one live electrode is taken. c Diameter of inner electrode when only one dielectric barrier layer is used. Plasma Sci. Technol. 19 (2017)

6 Figure 6. Variation of the energy ratio with respect to the frequency f. Figure 8. Variation of the energy ratio with respect to the discharge power. Figure 7. Variation of the energy ratio with respect to the input voltage V in. the electric field that is controlled by the voltage. When the energy ratio is plotted against the voltage, as shown in figure 7, the different DBD devices can be categorized into three different ranges. The distinction between the three different groups of DBD devices is also evident with respect to the discharge power, as shown in figure 8. Figure 8 shows that in group 1, the DBD devices from [38, 39] have energy ratios ranging from 16 to 58. In group 2, the CC and TCC reactors, and the DBD devices from [14, 40] have energy ratios that are lower, ranging from 0.01 to In group 3, the DBDs with NCC construction have the lowest energy ratios, which ranged from to The same categorization could also be observed in figure 7, where the energy ratio is different for similar voltages for different DBD devices. The energy ratios of group 1 devices (and from group 2) exceeds a value of unity, but does not indicate that more energy is discharged compared to the energy stored. This is because the estimated gap capacitance is used to estimate the energy stored in Figure 9. Variation of the energy ratio with respect to the effective discharge area. the gap, and not the total capacitance of the system. The total capacitance of the system was not considered in formulating equations (1) and (2) since information such as the capacitance of the electrical circuits is usually not available in the literature or during the design stage. The data in all three groups are correlated as power functions (in the form of y = ax b ),withgroups3and2showing correlation coefficients (R 2 ) of and , respectively. There is a significant scatter of data in group 1, although their energy ratio was in the same range. Additional data for DBD devicesingroup1inthefuturemayimprovethecorrelation coefficient. The slopes of the curve (exponent b) of groups 3 and 2 are and , respectively, while the two devices in group 1 from [38, 39] have values of and , respectively. The value of the slope indicates the variation rate of the energy ratio with respect to the discharge power, and is related to the 6

7 Figure 10. Variation of discharge power of various DBD devices with respect to the capacitive reactance X c. capacitance of the system. Group 3 devices have higher gap capacitances ( pf, see table 1), thus allowing the capacity for the device to have a higher rate of increase in energy ratio as the discharge power increases. Group 1 and 2 devices have capacitances that are lower than 85 pf, and the rate of energy ratio increase is slower. Figure 8 shows that group 1 DBD devices have the lowest energy ratio increase with regards to the discharge power, especially the device of Xia et al [38, 39] where the slope is only Exponent a of the power function increases from and to more than 9 for groups 3, 2 and 1, respectively. The increase in exponent a indicates that the energy ratio δ is increasing from groups 3 to 1. In group 1 the exponent a for [38, 39] was 9.73 and 14.11, respectively. The variation in exponent a with respect to each group is possibly due to the surface areas (as shown in figure 9 based on the length and inner diameter of the outer quartz tube, 2r ioq ) and capacitances. Group 1 DBD devices have solid surface electrodes that are wrapped, foiled or coated on the outer quartz tube, allowing the whole cylindrical surface area of the outer electrodes to have a larger discharge area, hence the higher energy ratios. By contrast, group 2 DBD devices consisting of those from [14] and [40] have wire meshes on the outer quartz tubes, and the cylindrical surface area is only fractional, thus the potential discharge area is reduced. The factor of reduction is estimated to be lower than 0.5, based on comparisons with the energy ratio and surface areas of the DBD devices in group 1 (see table 2). The CC the TCC reactors have copper foils as outer electrodes, but have lower surface areas compared to the DBD devices in group 1, therefore their respective energy ratios are lower in group 2. Figure 9 shows where group 3 devices have smaller areas (4 to 7 cm 2 ) compared to those in group 1 (207 to 327 cm 2 ).As the surface discharge area increases, the energy ratio δ increases and so does exponent a. The TCC reactor has a larger diameter and surface area compared to the CC, but the TCC reactor has three sets of parallel quartz tubes and electrodes tubes instead of one set in the CC reactor. The higher number of parallel tubes increases the capacitance, and with the capacitance being the denominator in equation (1) the increase in capacitance offsets the higher surface area of the TCC reactor, and the resulting energy ratio δ is not significantly higher than that of the CC reactor. The DBD devices in group 3 (the NCC reactors) are larger in diameter, with a 2r ioq of m and inner quartz tube outer diameters (2r oiq ) of 0.014, and m. Although the size and surface areas are larger, the NCC reactors have a non-concentric configuration, thus restricting the potential surface area for discharge to the proximities between the inner quartz tube and the outer ground electrode along the length of the reactor (see figure 1(b)). Based on the energy and surface areas, the resulting reduction factors are <0.006, , and , respectively for 2r oiq of 0.014, and m for the NCC reactors. In addition, the NCC reactors consist of 3 10 quartz tubes, resulting in higher gap capacitances than the DBD devices in groups 1 and 2. Figure 9 shows that the energy ratio for the NCC reactors decreases with an increase in surface area, which is similar to that of the TCC reactor. As the diameter increases, the NCC reactors capacitances increase and offset the increase in the surface area, decreasing the energy ratio δ. Figure 10 shows that the variation of the discharge power with respect to the capacitive reactance X c (= 1/2πfC eg ) for the DBD devices was not significant, as different capacitive reactance would yield similar discharge powers. Figure 11 shows a plot of the energy ratio with respect to the capacitive reactance, where the DBD devices are categorized into three groups, similar to that in figures 8 and 9. Thus, the energy ratio can be used to categorize DBD devices based on the surface area and capacitive reactance. Being a dimensionless parameter, the energy ratio is more descriptive than the factor developed by Kriegseis et al (equation (3)), which is not dimensionless (as shown in section 2.3). Nevertheless the parameter is still useful as a scaling factor Comparison with the factor of Kriegseis et al Figure 12 shows that the Kriegseis factor categorizes the DBD devices into three different groups. Group 1 consists of only the device of Fang et al [38] with Kriegseis factors of , while group 2 consists of the device of Xia et al [39] with Kriegseis factors of to ,andtherestof the devices are in group 3, with values ranging from to TheNCCset-upisalsoingroup3,indicating that the non-concentric DBD devices have similar characteristics to the devices that have concentric live, ground electrodes and quartz tubes. However, the Kriegseis factor does not account for the larger gap capacitance of the NCC set-up, which has a larger diameter (see table 1). Figure9 also shows that group 1 (the device of Fang et al) has significantly higher Kriegseis factors due to its lower operating frequency at khz (see figure 6) compared to group 2. The energy ratio did not made this distinction, as the device of Fang et al had similar energy ratios to those of Xia et al [39] (see figure 8). 7

8 8 Table 2. Type of electrodes and the respective effective discharge areas. Group Type of electrodes DBD device Reduction factor 1 Wrapped, foiled or coated on outer quartz tubes, with a full cylindrical area for discharge. Fang et al [38] 327 Xia et al [39] Wire meshes on the outer quartz tubes with reduced effective discharge areas. The CC and TCC Babaie et al [14] <0.5 <138 reactors have smaller diameters, thus have smaller effective discharge areas, even though the outer quartz tubes are wrapped with copper. Wang et al [40] <0.5 <170 CCs 13 3 Effective discharge area is limited to the proximity of the ground and live electrodes (reference Figure 4). TCC 35 NCC (0.014 m 10 <0.006 <4 tubes) NCC (0.030 m 6 tubes) NCC (0.044 m 3 tubes) Estimated surface area for discharge (cm 2 ) Plasma Sci. Technol. 19 (2017)

9 Figure 11. Variation of the energy ratio of various DBD devices with respect to the capacitive reactance X c. Figure 13. Variation of the Kriegseis factor with respect to the energy ratio. and exponent b. The figure shows that the power law correlation is more affected by constant a and estimates the Kriegseis factor with better accuracy compared to exponent b. Thus, it is possible that constant a is more indicative of the difference in capacitance of the DBD systems. The power law correlation also shows that the energy ratio can be correlated to the Kriegseis factor, even though each factor categorizes the DBD devices differently. Thus, both factors can be used interchangeably. 4. Conclusions Figure 12. Variation of the Kriegseis factor with respect to the discharge power. As both the energy ratio and the Kriegseis factor categorized the devices into different groups, both factors were plotted against each other in figure 13 to determine if a correlation exists. Figure 13 shows that the DBD devices from Fang et al [38] and the NCC set-up both have higher Kriegseis factors compared to the rest of the devices reported in the literature. This is mainly because both devices have higher gap capacitances, which ranged from 85 to 730 pf. The rest of the devices have significantly lower capacitances, from 2 to 58 pf. Both the energy ratio and the Kriegseis factor are correlated in the form of a power law, which has the form y = ax b. The correlation coefficients for the DBD systems with high and low capacitances are and , as shown in figure 13. Figure 14 shows a variation in the power law correlation with respect to different values of the constant a The study has proposed a dimensionless parameter δ in the form of an energy ratio that can be used for comparing different DBD devices, and can be used for designers and engineers to scale up their DBD systems. Three different groups of DBD devices were categorized according to the energy ratio, which indicated the efficiency of the DBD devices in discharging energy as compared to the capacitive energy in the gaps. The energy ratio was the lowest for the larger DBD device, which has a lower estimated effective surface discharge area and capacitive reactance (but higher capacitance). Thus, the energy ratio is able to describe the discharge power characteristics with respect to the geometrical features of the DBD devices since it affects the surface discharge area and capacitance of the DBD. Being a dimensionless parameter, the energy ratio is more descriptive than the factor developed by Kriegseis et al which is not dimensionless but is still useful as a scaling factor. Three separate and distinctively different groupings of the DBD devices were obtained with the Kriegseis factor. However, both the energy ratio and the Kriegseis factor are strongly correlated to each other, mainly defined by the capacitance of the DBD systems. 9

10 The correlation between both factors means that they can be used interchangeably. Acknowledgments The authors would like to thank Tenaga Nasional Berhad (Malaysia) for funding of this research (TNBR/SF195/2015 and TNBR/SF240/2016) and would like to acknowledge those who have contributed directly or indirectly towards the project. References Figure 14. Variation of the factors from equations (1) and (3) as a result of a variation in exponent a with (a) exponent b = 0.7. (b) b = 1.1. (c) b =1.5. [1] Sjöberg M et al 2003 J. Electrostat [2] Starikovskiy A and Aleksandrov N 2013 Prog. Energy Combust. Sci [3] Ombrello T et al 2010 Combust. Flame [4] Ombrello T et al 2010 Combust. Flame [5] Hu H B et al 2013 J. Therm. Sci [6] Du C M et al 2015 Int. J. Hydrogen Energy [7] Wascher R et al 2014 Surf. Coat. Technol [8] Vaswani S, Koskinen J and Hess D W 2005 Surf. Coat. Technol [9] Podgorski L et al 2000 Int. J. Adhes. Adhes [10] Han Y et al 2011 Carbohydr. Polym [11] Ye D et al 2005 J. Hazard. Mater [12] Wang P et al 2015 Appl. Therm. Eng [13] Talebizadeh P et al 2014 Renew. Sustain. Energy Rev [14] Babaie M et al 2015 Chem. Eng. J [15] Wang X Y, Zhou M H and Jin X H 2012 Electrochim. Acta [16] Wang N, Chen D Z and Zou L S 2015 Appl. Therm. Eng [17] Tichonovas M et al 2013 Chem. Eng. J [18] Jiang B et al 2014 Chem. Eng. J [19] Petitpas G et al 2007 Int. J. Hydrogen Energy [20] Nozaki T and Okazaki K 2013 Catal. Today [21] Stepanyan S A, Soloviev V R and Starikovskaia S M 2014 J. Phys. D: Appl. Phys [22] Nunnally T et al 2014 Int. J. Hydrogen Energy [23] Laroussi M et al 2004 J. Appl. Phys [24] Liu S H and Neiger M 2001 J. Phys. D: Appl. Phys [25] Lu X P and Laroussi M 2005 J. Appl. Phys [26] Liu S H and Neiger M 2003 J. Phys. D: Appl. Phys [27] Fang Z et al 2009 J. Phys. D: Appl. Phys [28] Kühn S et al 2010 Plasma Sources Sci. Technol [29] Rajasekaran P 2011 Atmospheric-pressure Dielectric barrier discharge (DBD) in air: plasma characterization for skin therapy PhD Thesis Institute for Electrical Engineering and Plasma Technology, Ruhr-Universität Bochum [30] Rajasekaran P, Bibinov N and Awakowicz P 2012 Meas. Sci. Technol [31] Rajasekaran P et al 2009 J. Phys. D: Appl. Phys [32] Tendero C et al 2006 Spectrochim. Acta Part B 61 2 [33] Glicksman L R, Hyre M and Woloshun K 1993 Powder Technol [34] Horio M et al 1989 J. Chem. Eng. Jpn [35] Kehlenbeck R et al 2001 AIChE J [36] Kriegseis J et al 2011 J. Electrostat [37] Zulazlan A et al 2016 J. Telecommunic. Electron. Comput. Eng [38] Fang Z et al 2008 J. Electrostat [39] Xia L Y et al 2008 J. Hazard. Mater [40] Wang T and Sun B M 2016 Fuel Process. Technol