Scholarly article on topic 'Dielectric Barrier Discharges: Progress on Plasma Sources and on the Understanding of Regimes and Single Filaments'

Dielectric Barrier Discharges: Progress on Plasma Sources and on the Understanding of Regimes and Single Filaments Academic research paper on "Materials engineering"

Share paper
Academic journal
Plasma Sources Sci. Technol.
OECD Field of science

Academic research paper on topic "Dielectric Barrier Discharges: Progress on Plasma Sources and on the Understanding of Regimes and Single Filaments"



Home Search Collections Journals About Contact us My lOPscience

Dielectric Barrier Discharges: Progress on Plasma Sources and on the Understanding of Regimes and Single Filaments

This content has been downloaded from lOPscience. Please scroll down to see the full text.

Download details: IP Address:

This content was downloaded on 03/03/2017 at 15:01

Manuscript version: Accepted Manuscript Brandenburg

To cite this article before publication: Brandenburg, 2017, Plasma Sources Sci. Technol., at press:

This Accepted Manuscript is: © 2017 IOP Publishing Ltd

As the Version of Record of this article is going to be / has been published on a gold open access basis under a CC BY 3.0 licence, this Accepted Manuscript is available for reuse under a CC BY 3.0 licence immediately.

Everyone is permitted to use all or part of the original content in this article, provided that they adhere to all the terms of the licence

Although reasonable endeavours have been taken to obtain all necessary permissions from third parties to include their copyrighted content within this article, their full citation and copyright line may not be present in this Accepted Manuscript version. Before using any content from this article, please refer to the Version of Record on lOPscience once published for full citation and copyright details, as permissions will likely be required. All third party content is fully copyright protected, unless specifically stated otherwise in the figure caption in the Version of Record.

When available, you can view the Version of Record for this article at:

Dielectric Barrier Discharges: Progress on Plasma Sources and on the Understanding of Regimes and Single Filaments

10 11 12

17 Ronny Brandenburg

18 Leibniz Institute for Plasma Science and Technology

19 Felix-Hausdorff-Strasse 2, D-17489 Greifswald, Ger:

21 E-mail:

23 Abstract. Dielectric Barrier Discharges (DBDs) are plasmas generated in

24 configurations with an insulating (dielectric) material between the electrodes which

is responsible for a self-pulsing operation. DBDs are a typical example of nonthermal atmospheric or normal pressure gas discharges. Initially used for the generation of ozone, they have opened up many other fields of application. Therefore DBDs are a relevant tool in current plasma technology as well as an object for fundamental

30 studies. Another motivation for further research is the fact, that so-called partial

31 discharges in insulated high voltage systems are special types of DBDs. The breakdown

32 processes, the formation of structures, and the role of surface processes are currently

33 under investigation. This review is intended to give an update to the already existing

34 literature on DBDs considering the research and development within the last two

35 decades.

The main principles and different modes of discharge generation are summarized. A

collection of known as well as special electrode configurations and reactor designs will be presented. This shall demonstrate the different and broad possibilities, but also the

40 similarities and common aspects of devices for different fields of applications explored

41 within the last years. The main part is devoted to the progress on the investigation

42 of different aspects of breakdown and plasma formation with the focus on single

43 filaments or microdischarges. This includes a summary of the current knowledge on the

44 electrical characterization of filamentary DBDs. In particular, the recent new insights

on the elementary volume and surface memory mechanisms in these discharges will be discussed. An outlook for the forthcoming challenges on research and development will be given.

Progress on DBD Sources and Filaments

1. Introduction

Dielectric Barrier Discharges (DBDs) are self-sustaining electrical discharges in electrode configurations containing an insulating material in the discharge path. This so-called dielectric barrier is responsible for a self-pulsing plasma operation and thus, the formation of a nonthermal plasma at normal pressure. They are also known as "silent discharges", "barrier discharges" or "ozonizer discharges". The partial gas discharges occurring in insulated high voltage systems can also be considered as DBDs. First introduced for the generation of ozone in 1857 [1], they have opened up many other fields of application, e.g. surface treatment, degradation of pollutant molecules in gases, pumping of gas lasers, plasma displays, and generation of excimer radiation. Since the 1990's they have been explored for the biological decontamination of medical devices, air flows, and tissues. DBDs are also considered as sources for the electric wind in aerodynamic control systems or used in novel analytical detection devices.

Due to their high technological relevance its principle and applications have already been described in detail in comprehensive textbooks and review papers, e.g. [2, 3, 4, 5, 6]. This review is intended to update this information by giving an overview on the progress in the field within the last two decades. Therefore, in the first and second part, the main basic configurations and principles of DBDs as well as the spectrum and characteristics of the different discharge regimes will be summarized, respectively. The third part will give a broad overview on DBD reactors and plasma sources for all current fields of application. It will include already known concepts but extent this to new arrangements which were developed for the exploration of new applications, e.g for treatment of biological samples or liquids. The technical details will not be given as this part aims to demonstrate the different and broad possibilities of DBDs. These can be found in the cited literature. The intention of this part is also to show the similarities and common aspects of devices for the different application areas. The fourth part will be devoted to the progress in the understanding of the discharge physics with the focus is on single filaments as the elementary object of the most common discharge mode of DBDs. Due to the presence of insulating walls a high complexity of plasma-surface interaction and breakdown processes has to be considered. The new insights on the fundamental volume and surface memory mechanisms in these discharges, obtained by new diagnostic approaches, modeling and theory will be summarized. The contribution will conclude with an outlook to future research.

It has to be mentioned, that a considerable progress was also obtained on DBD-based plasma chemistry and structure formation. The review will not cover this and the reader is referred to other topical reviews and books, e.g. [2, 4, 5, 7, 8, 9, 10, 11, 12].

10 11 12

20 21 22

Progress on DBD Sources and Filaments 2. Principles and configurations 2.1. Basic configurations


DBDs configurations and principles are characterized by the presence of material in the discharge path. Typically, dielectric materials such as glass, qUartz," ceramics, enamel, mica, plastics, silicon rubber or teflon are used. To operate a plasma at normal pressure with moderate high voltage amplitudes the discharge ga\ istypically in the range of 0.1 to 10 mm. Because of the capacitive character of the discharge arrangement alternating or pulsed high voltage is required. The high voltage amplitude is in the range of 1 to 100 kVrms. Based on the general description different configurations can be realized. The main basic configurations in planar geome^^^re sketched in Fig. 1.

I Mill

-e-1 Lo-

^■(-^o High Voltage Source i i Insulator ^ Plasma Region E|ectrode

Figure 1. Bpfic planar configurations of DBDs: (a) volume DBD (1- symmetric, 2- asymmetric, 3- floated dielectric); (b) surface DBD (1- symmetric, 2- asymmetric ctuator" design); (c) coplanar discharge

flWtha voume DBDs one or both electrodes are covered by the dielectric barrier(s) (ai Bnd a2) or a dielectric layer is separating the gas gap in two sections (a3) [13]. In configuration (a1) both metallic electrodes are protected from the reactive species that might be formed in the plasma. In configurations (a2,3) one or both metallic electrodes are exposed to the plasma which can lead to its erosion or corrosion, but the plasma can be operated at lower high voltage amplitude than in (a1) since electrons and ions are more easily released from metals than from dielectrics. Since the total capacity is

Progress on DBD Sources and Filaments ^/4

doubled compared to (ai), a higher amount of charge per electrode area element can be dissipated. In (a3) the dielectric is floated. This has no influence on the general operation principle. It is just an option for the treatment of gases by DBDs. In the surface DBDs (b1,2) both electrodes are in direct contact with the barrier [14]. HeW| the plasma is formed in the gas at the exposed surface electrode and propagate^along1 the dielectric surface while the sheet or counter electrode is embedded in an additional dielectric layer (not shown in the figures). Such configuration has been realize« by mesh wire electrodes mounted on an insulating plate with the second sheet electrode below the plate [15], by printing of structured metal films on the insulating plates, or the structured etching of circuit boards [16, 17, 18]. The asymmetric configuration (b2) is well described in the literature on plasma actuators e.g. [19, 20]. In the coplanar discharges (d) both electrodes are embedded in the insulator and the discharge appears in the gas above the dielectric surface [21, 22, 23, 24].

2.2. Special configurations

The research and adaptation of DBDs on different problems has lead to several novel configurations. A few examples are given in Fig. 2.

-o ^^ O <>

Ô 1 i

Voltage Source I I Insulator

Plasma Region


Piezoelectric crystal

Figure 2. Special configurations of DBDs: (a) sliding discharge [19]; (b) capillary plasma electrode discharge [26]; (c) microcavity plasma array [27, 28, 29]; (d) piezoelectric operated DBD [37] (cross sectional views)

The so-called sliding discharge (Fig. 2 a) has been developed for plasma actuation id catalysis [19, 25]. The idea is to enhance the area of plasma-surface interaction by

21 22 23

Progress on DBD Sources and Filaments 5

5 enforcing the plasma to develop along the gas-dielectric interface. The configuration is

6 based on the asymmetric surface DBD. The third electrode (sliding electrode) is biased

8 with an additional potential that accelerates the charge carriers along the interface.

9 Finally, the plasma slides over the surface. i n

11 A special form of DBDs is the capillary plasma electrode discharge (Fig. 2 b)

12 [26]. The dielectric is perforated which leads to more more uniform and higher density

13 plasmas than in classical DBDs (see section 3). A relatively new type of DBDs are the microplasma arrays. They are produced by micro-machining techniques (e.g. anisotropic

16 etching) on silicon layers [27, 28]. The cavities can have a rectangular cross section as

shown in Fig. 2 c or of an inverted square pyramid. The silicon-electrode is coated with a ^m thick film of silicon nitride as the barrier and a nickel array as the second electrode. 20 The use of polymer-based replica molding processes allows fabrication of flexible

arrays of microcavity plasma devices. Linear plasmas with trapezoidal or parabolic spot profiles, generated within nanoporous alumina microchannels were studied more

24 recently. Cavity dimensions as small as 3 ^m with aspect ratios (length : width) of

26 1,000 : 1 can now be realized. Microplasmas have attracted considerable attention in

27 various applications, such as mask-less micro-fabrication processes, synthesis of nano-

scale particles, in life-science (medicine) as well as meta-materials (wave-propagating media with extraordinary properties) [29, 30].

31 Beside insulators other barrier materials has been used. In the so-called Resistive

Barrier Discharge (RBD) a highly resistive sheet (e.g. a top-wetted ceramic or silicate layer with a resistivity of a few MQ/cm or higher) is covering one or both electrodes. It 35 acts as a distributed resistive load [31, 32], playing the same role as the dielectric barrier,

37 but, it enables plasma operation also by DC high voltage. The current is self-pulsing

38 with a repetition rate of a few kilo-Hertz and pulse durations of a few microseconds.

39 Systems with semiconductor electrodes at cryogenic temperatures were studied, e.g. in

41 [33, 34, 35]. In order to control the discharge inception the semiconductor is illuminated

42 externally. The illumination increases the specific conductivity and thus, the capacity per electrode area unit (also called effective capacity). Consequently, the voltage drop across the gas gap increases and leads to electrical breakdown. This research is not only

46 motivated by studying the plasma dynamics but also the development of infrared image

converters with high response speed [36].

As mentioned above, DBDs and other atmospheric plasmas sources require high 50 voltages for discharge operation. The reduction of the voltage amplitude is possible by

52 the utilization of piezoelectric crystals. Mechanical, thermal, or electrical forces alter

53 the crystal structure and change the magnitude of its polarization. Thus, the surface

54 potential can be increased to values which allow gaseous breakdown [37, 38, 39]. The

56 generation of DBD-based plasma jets by means of piezoelectric transformers has been

57 demonstrated and many applications like lighting and ozone formation are possible [37, 38, 40]. As shown in Fig. 2 d the DBD appears between a grounded dielectric covered electrode and the ceramic surface of the piezoelectric crystal. The high surface potential is generated via resonant electrical-mechanical vibrations induced by an AC

10 11 12

20 21 22

Progress on DBD Sources and Filaments

voltage with low amplitude. Recently, thermally excited pyroelectric crystals were used to generate corona-like discharges [41]. Despite the power of the plasma is limited in these concepts, the compactness and low-voltage driving makes them promising for many industrial applications.

A multitude of atmospheric pressure plasma jets which implement a DBD in the broader sense exist [42, 43, 44, 45]. Most of them are coaxial, i.e. the barrier is a tube of dielectric material and thus, forms a gas nozzle (see Fig. 3 a and b). The gas flowing through the arrangement extends the plasma as an effluent into the surrounding gas. Beside coaxial arrangements planar or linear plasma jets are possible and can treat surfaces with several meters length [46, 47]. As an example, in Fig. 3 c, a double slit volume DBD with a ribbed grounded electrode and a non-equidistant discharge gap is shown. Two parallel effluents are formed by the gas flow and merge into one [46, 48]. Whether the plasma jets can be considered as classical DBDs or not is mainly determined by the frequency of the applied voltage (see section 3). Since this review is not intended to focus on plasma jets, the reader is referred to recent comprehensive review articles in this field, e.g. [43, 44, 45, 49]. Indeed, some of the principles and diagnostic approaches described in the following can be applied to plasma jets, in particular the characterization of its core plasma inside the nozzle.

h Voltage Source i i Insulator

Plasma Region



Figure 3. DBD-based plasma jet configurations: (a) core plasma generated inside dielectric tube with two metal ring electrodes outside the tube; (b) configuration with one metal electrode outside and a concentric needle electrode inside [44]; (c) double slit linear DBD-based plasma jet ©2015 IOPP. Reproduced with permission from [48]

Progress on DBD Sources and Filaments

5 3. General operation and discharge regimes

67 In most studies and applications DBDs are operated with AC high voltage in the kHz-

9 range. Within the last two decades more and more studies have extended this to pulsed

10 or radio-frequency (i.e. frequencies of 1 MHz and more) driven discharges. In any

12 case the DBD arrangement has to be interpreted as a capacitive element. Thus, the

13 displacement current is determined by the total capacity of the DBD arrangement as

14 well as the time derivative of the applied voltage. Indeed, the total capacity is given by

16 the dielectric constant and the thickness of the barrier(s) as well as the geometry of the

17 DBD arrangement.

20 21 22

3.1. The classical operation regime and filamentary plasmas

In case of kHz-operated DBDs (AC or pulsed) the electrical breakdown leads to the 23 charging of the insulator surfaces. The surface charges induce an electric field opposed

to the applied electric field. Thus, the total field is decreased and the discharge extincts.

26 The insulator electrically acts like a load which limits the amount of charge and the

27 average current density in the gas. This keeps the plasma in the nonthermal regime,

29 i.e. without such limitation, a spark or arc discharge would be formed. In case of AC

30 high voltage a breakdown occurs during the increase of the value of the applied voltage

(i.e twice per high voltage period) and the discharge activity stops, when the voltage maximum is reached. In case of unipolar pulsed operated DBDs the surface charges

34 induce an electric field at the falling slope of the high voltage pulse which causes the

second breakdown in the same high voltage period (also referred as back discharge) [50, 51, 52, 53].

38 At medium, normal and even higher pressures, gas discharges tend to constrict due

40 to the streamer breakdown mechanism [4, 54, 55]. Electron avalanches create a space

41 charge and thus an additional electric field which enhances the growth of secondary

42 electron avalanches locally. Consequently, the ionized region and the perturbation

44 of the electric field grows rapidly and forms distinct plasma channels. Due to the

45 repetitive generation these so-called microdischarges visually appear as tiny discharge channels, often named filaments. DBDs in most molecular gases but also in argon or mixtures of noble gases with molecular gases are typical examples of such filamentary

49 plasmas. Filamentary plasma generation was also obtained in radio frequency operated

asymmetric surface DBDs [56]. The frequency (13.56MHz) is high enough to trap electrons in the discharge region across successive cycles. It leads to an increase in the 53 length, intensity and degree of branching of the discharge channels.

The microdischarge channels are spreading on the barrier surface covering a region

56 much larger than the original channel diameter. An increase of the voltage amplitude

in a AC operated, plane parallel discharge gap DBD leads to a higher number of microdischarge per active phase, but will not change the amount of charge transferred 60 to a single microdischarge [57, 58, 59, 60, 61]. The latter depends - apart from the

gas type and pressure - on the design parameters of the DBD. It is proportional to

Progress on DBD Sources and Filaments 8

5 the discharge gap g [62]. In [63, 64] it was shown that the values are independent on

6 electrode material but on the width of the discharge gap and the specific capacitance 8 of the barrier (Cd/A). In [65] it scaled with while in [66] it scaled roughly with

9 (1 - Cd/Ccell).

11 The duration of microdischarges is determined by the gas as well as the discharge

12 arrangement (gas gap, type, and thickness of barriers). In air at 1 bar and 1mm

13 discharge gap the microdischarges have a duration of about 10 to 100 ns with a total

15 transferred charge of 0.1 to 10 nC. Microdischarges in argon can have a duration of a

16 few p.s. Although maximum current density can reach up to 1,000 A/cm2 significant gas

17 heating in the remaining channel does not happen because of the short duration.

20 3.2. Diffuse DBDs

22 For more than 100 years, DBDs were known to operate only in the filamentary regime.

24 In the 1970s it was shown that external pre-ionisation by UV-photons or X-rays enables

25 the generation of uniform DBDs in gas laser devices [5, 67]. Uniform or diffuse means,

26 that the surface cross section is covered by the discharge more or less entirely, although

28 the plasma is not necessarily homogeneous in axial direction. Already in 1969 the

29 first AC operated uniform DBD in helium was described [68]. In the 1990s further investigations of uniform DBDs were presented in reproducible manner and verified [69, 70, 71, 72, 73, 74, 75]. They are also known to be generated in plasma display panel

33 cells [72, 76]. In all these examples discharge uniformity is possible without external pre-ionisation but requires specific conditions, in particular on the gas composition or purity. Furthermore, it depend on the frequency of applied voltage and the properties

37 of the external electrical circuit limiting the dissipated power [77, 78, 79, 80]. It must be

noted that a filamentary DBD can appear to be uniform to the naked eye or slow cameras

40 if the number density of microdischarges per cycle is high [80, 81, 82, 83, 84]. There

can also be transition regimes, where uniform and filamentary mode exists within the same half period or stagger between the two successive half periods (e.g. [69, 85]. The

44 uniform character can be proved by short exposure time photographs combined with fast

discharge current measurements. In case of a filamentary DBD individual and erratic appearing current "spikes" are observed, while in a diffuse DBD much longer current

48 pulses which are in phase with the applied voltage, are obtained. Diffuse DBDs were

described either as Atmospheric Pressure Townsend Discharge (APTD) or Atmospheric Pressure Glow Discharge (APGD) in the literature (see e.g. [80] and references therein).

52 They have been generated in pure gases (helium, neon, nitrogen) as well as gas mixtures

(argon with ammonia, air with precursor molecules) [71, 80, 83, 84, 86].

55 An APGD is typically observed in rare gases like helium and neon with discharge

gaps of a few millimeters [73, 80, 87, 88, 89]. These gases are characterized by a relatively high gas ionization at comparatively low electric field strength which can

59 lead to a considerable "slow down" of the breakdown and avoid the formation of rapid

60 electric field gradients as it is obtained with the streamer mechanism [5, 13, 90, 91].

21 22 23

Progress on DBD Sources and Filaments 9

5 Furthermore, the metastable states of noble gases are considered to induce a variety o

6 multistage ionization processes. In particular when impurities or admixtures of nitrogen 8 are present in the gas, Penning ionisation (chemi-ionisation, A* + B ^ A + B+ + e,

9 with A* representing an excited (in particular metastable) specie), stepwise ionisation

11 and charge transfer collisions play an important role in the overall ionisation dynamics

12 [72, 80, 92]. The plasma display cells are a typical example of DBDs operated in so-called

13 Penning gas mixtures (in this case Neon/Xenon) [76, 93]. An essential feature of the

15 APGD is the formation of a cathode fall [80, 88, 89, 94, 95, 96, 97] and the discharge is

16 characterized as a subnormal transient glow discharge with one or several narrow current

17 pulses per half period. The current density per pulse is in the range of 1 - 100 mA/cm2 and thus, orders of lower than in a microdischarge [86]. An APTD is typically obtained

20 in nitrogen and noble gases in discharge gaps of about 1 mm [80, 98, 99, 100]. It is

characterized by an exponential increase of the electron density towards the anode and a low space charge production [80, 101]. The current pulse is significantly longer than 24 in an APGD and the current density is lower (0.01 - 10 mA/cm2).

27 3.3. Operation at radio frequency

29 At frequencies in the MHz-range (radio frequency, RF) the current limitation by the

dielectric becomes less effective, the breakdown voltage is decreased [26], the discharge operation changes significantly and impedances must be considered. The charge carriers 33 in the volume do not completely diminish between two subsequent voltage half periods

and the mobility of the ions is too small to follow the variation of the applied electric field. Thus, ions are trapped in the gap and do not charge up the barriers. Strictly speaking the 37 discharge is not a DBD as the role of barrier is not induce the self pulsing character. The

discharge operates in steady state and its properties are comparable with RF discharges

ent pulse is í

40 or capacitively coupled plasmas at low and medium pressure [102, 103, 104, 105, 106]. As

the discharge sustainment is without secondary emission the main electron production and heating likely happens in the sheath region. This discharge regime is called a-mode 44 (in contrast to y-mode) [107, 108, 109, 110]. The discharge mode has significant effects

on the plasma parameters [103]. The role of the dielectric here is mainly the protection of the electrode material from plasma species.

3.4. Capillary jet mode

Special discharge modes are also investigated in the capillary plasma electrodes where

53 perforated dielectric barrier(s) or dielectric capillaries from 0.01 to 1 mm and length-

54 to-diameter ratios of about 10:1 cover one or both electrodes (see Fig. 2 b). This

56 arrangement exhibits the so-called "capillary jet mode", i.e. the capillaries serve

57 as plasma sources producing jets of high intensity plasma under certain operating

58 conditions [111]. These jets emerge to the gas gap and support breakdown in the gap 60 leading to a more uniform and stable plasma. This mode also shows a higher electron

density than usual filamentary DBDs.

Progress on DBD Sources and Filaments

4. Designs of DBD-based plasma sources

Due to the numerous application areas many different DBD reactor or device designs exist. This section aims to give an overview on the current status of the DBD-based plasma source technology. The following subsections show selected examples for the treatment of gas streams, the generation of light, surface/object treatment, plasma medicine and decontamination, and liquid treatment, respectively. The last subsection gives some remarks about the development of desired high voltage power supplies. This overview makes no claim to be complete. The different concepts and possibilities for the design of plasma sources should be shown and inspire new developments.

The classical reactors for gas treatment (e.g. ozon< n from oxygen or

air, pollutant degradation, synthesis) are cylindrical devices with coaxial electrode arrangement. Tubes of insulating materials are used either as the outer electrode (Fig. 4 a) and/or covering the inner electrode (Fig. 4 b). The electrode can be a metallic film deposited or pasted on the inner/outer dielectric tube wall as the inner/outer electrode, instead of a solid metal electrode. In laboratory reactors metal meshes are often used, but then parasitic discharges outside the tube must be considered. They can lead to an overestimation of the dissipated power [112]. Large scale reactors, e.g. ozone generators, consists of several DBD tubes mounted in a generator tank. In modern ozone generators the inner electrode is coated with enamel [113]. The electrode arrangement also enables the cooling of the discharge tube. These discharge reactors are still the most robust and reliable DBD sources for gas treatment.

In packed bed reactors (Fig. 4 c) pellets or spheres made of dielectric or ferroelectric material are filling the space between the electrodes. The polarization of the pellet material generates regions with high electrical fields and gas discharge appears in the void spaces between the pellets and on their surfaces [114, 115, 116, 117]. The field refraction and enhancement at the pellets depends on their shape, porosity, and the permittivity of the filling material. Porous ceramic foams can also be used instead of pellets beds [118]. The advantage of such reactors is that the filling material can have catalytic properties or can be prepared to have a catalytic surface which increases the efficiency and selectivity of the plasma chemical processes [7, 9, 112]. Therefore, they were mainly investigated for the cleaning and conversion of gases.

The stacked volume DBD reactor (Fig. 4 d) consists of a stack of dielectric plates and electrodes at alternating potential. Metal plates as well as metal grids can be used [119, 120] and spacers of dielectric material adjust the desired gap distances. Such compact and well up-scalable DBD reactors are often used for the treatment of exhaust gases and its deodorization [119, 121, 122, 123]. A similar arrangement can be realized based on the surface DBD principle (Fig. 4 e). Here, the space between the electrodes can be increased without the need for increasing the voltage amplitude as the discharge is generated only on the dielectric surface. However, in such a case only the

4.1. Gas treatment with DBD-reactors

10 11 12

20 21 22

Progress on DBD Sources and Filaments

(a) (b)

o^« High Voltage Source i i Insulator (££)

Plasma Region



Figure 4. DBD reactors or reactor modules for the treatment of gases (cross sectional sketches): cylindrical or coaxial volume DBD with outer (a) and inner (b) dielectric electrode [5, 112]^^^cked-bed DBD [112]; (d) stacked volume DBD [119, 120]; (e) stacked surface DBD [120, 124]; (f) open microplasma array type system (§2006 John Wiley and Sons. Reproduced with permission from [29]

gas fraction flowing directly along the electrodes is subjected to the active plasma. With an additional electrode an ionic wind can be expelled from the plasma. These ions are charging and deflecting pjrticulate matter similar as in electrostatic precipitators [124]. These arrangements do not increase the back pressure significantly since the gas flows along the electrodes [125]. They are also more easy to construct and more compact than the cylindrical device^mentioned before. However the extra cooling of the electrodes is not possible and in case of a defect of one of the barriers sheets the repair takes more efforts.

The arrangement in Fig. 4 f is based on a two insulator coated metal plates [29, 126, 127, 128, 129]. A sandwich of at least two plates with about 200 |am wide holes enable the set-up of a microplasma array which can be penetrated by the gas flow. This concept is even more compact than all other mentioned ones and since both electrodes are embedded in the dielectric the oxidation of metallic electrodes is not limiting the lifetime of reactor modules as in all the previous discussed examples. Disadvantages are the higher back pressure and the short retention time of gas particles in the active plasma region. A setup for the purification of inert gases using a DBD plasma together with a

10 11 12

20 21 22

Progress on DBD Sources and Filaments

rotating sacrificial electrode is presented in [130]. This approach benefits from the fact that the plasma treatment of metallic substrates under inert gases has been found to mainly result in its oxidation due to oxygen impurities. Thus, oxygen impurities could be removed - similar as in gettering processes - by plasma-driven chemisorption on the sacrificial metal electrode. Since surface properties change the plasma the electrode is under rotation during operation.

4-2. Light generation with DBDs

the electi

The generation of light is based on the formation of plasma in dedicated gas mixtures sealed in silica glass vessels [5, 131, 132]. The gas mixture usually contains a buffer gas (mixture of noble gases) to enable low ignition voltage. Planar light panels (Fig. 5 a) and cylindrical lamps (b) have been realized. The vessel glass also serves as the barrier. The outer or front electrodes are metallic wire meshes, perforated metallic films or transparent metallic layers. The rear electrode can be utilized as a mirror. Depending on the application UV or VUV-photons emitted from the plasma are used (excimer lamps for special applications, e.g curring) or these photons are converted to visible light due to the excitation of phosphors coated on the interior surface of the vessel. Flat light panels can also be constructed with coplanar electrodes (Fig. 5 c). Such lamps are manufactured by means of thick film printing [5]. In commercial plasma display panels each cell consists of a microscopic DBD with dimension of about 0.1 mm x 0.1 mm. The VUV radiation of the generated rare gas DBD plasma is used to excite phosphors emitting in red, green or blue, respectively [76]. Microplasma arrays as mentioned above are also considered as novel light sources and on the way of commercialization [28].

Mirror Electrode

I I Insulator

] Transparent Electrode

Figure 5. DBD modules for the generation of light (cross sectional sketches): (a) sealed planar and (b) cylindrical excimer lamp configurations; (c) coplanar excimer lamp configuration ©(all subfigures) 2006 Elsevier. Reproduced with permission from [132]

10 11 12

20 21 22

Progress on DBD Sources and Filaments 4-3. Surface treatment

The most important DBD application in surface treatment today is the activation of plastic foils in order to increase its surface energy prior to bonding, ^printing, coating, adhesion, or, for cleaning it from organic compounds. The deposition or functionalisation of surfaces is also possible. The scheme of a typical DBD treatment of fabric material, e.g. foils, wool or textiles, is shown in Fig. 6 a. The material is rolled off and up from storage coils and thus, the surface to be treated is continuously moved through the DBD [2, 5, 133, 134]. It is formed between a dielectric covered electrode (mostly water-cooled) and the surface which is transported over a metallic roll. The rotating drum can also be covered with a dielectric layer. Speeds of up to 10 m/s are reached and large facilities can handle plastic webs of 10 m width or more [2, 5, 133].

The cascaded DBD shown in Fig. 6 b combines the treatment of surfaces and goods by plasma-generated (V)UV radiation and reactive species in one device. It was developed for the biological decontamination of packagings [135]. The discharge consists of a high voltage powered electrode, a sealed glass chamber filled with excimer-forming gas for efficient UV emission, an air flushed gas gap for the generation of the reactive species, and the grounded electrode below the object.

Storage coil # lUlrdl

o^o High Voltage Source n Insulator

Plasma Region





Figure 6. Schemes for DBD-based surface or object treatment: (a) fabric material treatment [5]; (b) cascaded DBD [135]; (c) coplanar DBD arrangement (DCSBD) for fabric materials [22]; (d) microplasma stamp [136]; (e) in-package DBD treatment [139]; (f) contacted plasma [141]

21 22 23

Progress on DBD Sources and Filaments 14

5 Another principle for the treatment of web and fabrics is based on the so-called

6 Diffuse Coplanar Surface Barrier Discharge (DCSBD, see Fig. 6 c) [22]. It generates a 8 visually diffuse plasma with a high power density in the immediate vicinity of the textile

9 surface. This arrangement allows more uniform plasma treatment without pinholing or

11 other critical changes in the tensile strength of the fabrics.

12 The microplasma stamps schematically shown in Fig. 6 d enable an area-selective

13 treatment of surfaces [136, 137, 138]. Therefore, the dielectric is patterned in order to

15 form cavities with a size in the range of 100 - 500 ^m when the stamp is pressed against

16 the substrate. Since the plasma is only ignited in the voids a lateral microstructuring and surface treatment is achieved in a single process step.

A new possibility for the treatment of packed goods is the local application of 20 a surface DBD. The electrode arrangement is contained in a label bonded inside the

airtight flexible or rigid package (so-called "plasma label") as demonstrated in Fig. 6 e [139, 140]. The arrangement sketched in Fig. 6 f is an example for DBD device which can 24 be designed as a hand-held device which allows to bring the plasma in closed proximity

to the surface. The process is supported by a gas flow through the arrangement but

27 the substrate is not integrated ("contacted plasma") [141, 142]. The gas flow allows the

28 reduction of operating high voltage amplitude if noble gases are used and can influence

30 the compostion of the plasma, similar as in plasma jets. It is probably an interesting

31 device for the treatment of curved objects. DBDs are also considered for the enhancement of other material processes. Novel

hybrid laser-plasma methods were presented in [143, 144]. The method combines an 35 laser with DBD-based plasma jet for the chemical reduction of glass surfaces or micro-

37 structuring purposes. The combination reduces the ablation threshold and improves the

38 machining quality. Furthermore, the DBD was found to act as a thermal lens for the

39 the laser beam causing a defocussing by several millimeters [144].

42 4-4- DBDs for plasma medicine and decontamination

44 Many activities in the exploration of plasma life-science applications have inspired new DBD-based plasma sources. The first activities were devoted to the biological decontamination of surfaces, medical devices and other sensitive goods [145, 146]. The

48 cascaded DBD, the in-package surface DBD and the contacted plasma device discussed

50 in the previous subsection has also been mainly considered for such applications [147].

51 A commercial device for the cleaning and sterilization of liquid transfer devices such

52 as pipette tips, cannulae and pin tools is available [148]. In [149] a DBD for the decontamination of rotating cutting tools is presented. The metallic cutting tool acts

55 as grounded electrode and two equal dielectric isolated electrodes realized the plasma

56 treatment from both sides. The in-package treatment and decontamination of food can

58 also be realized in a large-gap volume DBD in air with up to 4 cm discharge gap. High

59 voltage amplitude between 50 and 70kVrms is required under such conditions [150].

60 The usage of plasmas in therapeutic applications requires plasma sources which

10 11 12

20 21 22

Progress on DBD Sources and Filaments

o^o High Voltage Source n Insulator

Plasma Region

Figure 7. Schemes electrode volume DBD [153]; based surface DBDI142, 164]

>r plasma medical applications (a) floated mer film-based surface DBD [160]; (c) fabric-

are flexible to use (e.g. as a handheld device) as well as safe for patients and doctors [151, 152]. One of the first DBD devices used for the treatment of animals and humans is the floating electrode DBDs shown in Fig. 7 a [153, 154, 155, 156, 157]. The treated object itself serves as the grounded electrode. The discharge device consists of a dielectric-protected powered electrode and is operated by high voltage pulses. The plasma is in direct contact with the treated surface. The electrical shock is avoided due to the limitedLurrent of the DBD. Decontamination of skin, coagulation of blood and other biological effects have been demonstrated with such sources. There are also commercial devices available and already used in clinical trials [158, 159].

New developments use flexible materials or arrays of movable electrodes. Such principles allow the treatment of objects with a complex geometry as extremities or other bodily parts [142, 160, 161, 162, 163]. All these devices are usually operated with classical AC high voltage in the range of several kHz. The example shown in Fig. 7 b is made of a printed circuit board [160]. The dielectric barrier is a polymer film equipped with a thin, uniform powered electrode on one side and a thin, meshed grounded electrode on the other side. The technology enables the construction of formats as much large as required by the wound or skin area to be treated. The advantage of such device is that no current is flowing through the treated sample which increases the electrical

Progress on DBD Sources and Filaments 16

5 safety. The flexible surface DBD plasma strip is able to quickly inactivate significant

6 numbers of bacteria on and in skin. Another example for such kind of flexible DBD is o the use of insulated conductor wires in a woven arrangement (fabric-type electrode), as

9 shown in Fig. 7 c [142, 164]. Such arrangements can even be constructed as a sticking

11 plaster device. Fabric-based discharge devices could be produced with reasonable efforts

12 and cost as required for one-way adjutants in medicine.

13 Other developments of DBD based plasma sources are dedicated for the inner

15 treatment of long flexible tubes, as can be found in complex medical devices like

16 endoscopes. Among DBDs with an outer and an inner-tube electrode or the usage of afterglow plasmas [165, 166, 167] a bifilar helix electrode configuration enables plasma operation in tubes with length of several meters and an inner diameter of 2 mm [168].

20 The powered and grounded wire electrodes are equidistantly twisted around the tube.

21 The electrodes are embedded in an additional dielectric layer to avoid discharges outside

23 the tube.

4-5. Treatment of liquids

An increasing number of research on the treatment of liquids has been done. Some DBD-29 based concepts for the treatment of liquids by DBDs are schematically shown in Fig. 8.

They offer low power density plasma generation compared to pulsed corona discharges. Fig. 8 a shows a surface DBD device for the indirect treatment of water, i.e. the

33 liquid is in vicinity to the plasma and the liquid is not an active part of the discharge

generation. Although there is no direct impact this arrangement leads to significant effects, e.g. acidification, or formation of reactive nitrogen and oxygen species. Such 37 studies are important for the understanding of plasma-cell interaction [169] and offers

3o a wide range of applications such as degradation of pollutants and microorganisms in

40 water, the analysis of trace compounds in solutions, or the synthesis of active liquids,

species or nano-materials.

Many liquids are conductive and can be used as an active electrode for the plasma

44 generation (e.g. [174, 170, 171, 172, 173]). Examples for DBDs are shown in Figs. 8

b-d. The liquid is covering the grounded electrode to avoid its charging. Using liquid electrodes can also be utilized for the treatment of outer and inner surfaces of tubes and

48 other hollow dielectric bodies as demonstrated in [172].

The water falling film reactor, shown in Fig. 8 c consists of a sealed cylinder of dielectric with a concentric hollow metal tube electrode inside and a metal electrode

52 outside [175]. Water or other liquids are pumped through the inner tube electrode and

form a laminar flowing layer along the inner metallic tube. This reactor has has been 55 demonstrated to be capable of decomposing pollutants in the liquid phase [176, 177]

as well as for gas cleaning [123, 178]. A similar concept is realized in the rotating drum reactor (Fig. 8 d). A homogeneous, thin film (10 ^m, the thickness depends on

59 rotational speed) of bacterial suspension is produced on the rota-table hollow cylinder

60 made of metal and treated by the plasma. Plasma-specific stress responses of bacteria to

10 11 12

20 21 22

Progress on DBD Sources and Filaments


o^^o High Voltage Source Vft Insulator m Liquid

Plasma Region ■ Electrode Flow

Figure 8. Schemes for DBD-based treatment of liquids: (a) indirect treatment by means of surface barrier discharge [169]; (b) direct treatment, i.e. liquid serving as one electrode [174]; (c) water falling film reactor [175]; (d) rotating drum reactor [179]; (e) liquid electrode DBD [181]

argon and air plasmas were studied in this set-up [179, 180]. The DBD-liquid electrode setup in Fig. 8 e was developed for aqueous analysis with low sample volumes [181, 182]. The analyte is supplied in a dielectric capillary covered with metal layers serving as high voltage electrode. The grounded electrode is a metal wire connected over a high ohmic resistance. The concept has been demonstrated for emission based analysis of liquid solutions and is supposed to be combined with micro-separation devices in the future.

4.6. Powers

Along with th^improvement of industrial DBD installations and the development of new discharge geometry concepts there has been a significant progress in the field of high voltage power supplies. AC high voltage power supplies, in particular sinusoidal, are still the most common in industry. State-of-the-art in surface processing is the use of variable-frequency drives in connection with high voltage transformers. Modern power supplies of surface treatment stations are microcomputer-controlled and automatically adjust its operation frequency to the resonance point of the system (i.e. capacitance of electrode configuration and inductance of high voltage transformer). The effective ability

Progress on DBD Sources and Filaments

5 to control power transfers by means of current magnitude and frequency is highlighted

6 in [183]. Special power supplies have been developed to generate repetitive pulse train: 8 resulting in an improved distribution of the energy, e.g. [184].

9 A new trend are solid-state devices that can eliminate bulky hi-voltage

transformers or realize resonant power supply system topologies for higher energetic

12 efficiencies [5, 185]. Pulsed power supplies and topologies enabling a variety of pulse

13 waveforms (unipolar, bipolar) has been investigated and developed, in particular for

15 ozone synthesis and lighting. A classification of novel and state-of-the-art power

16 electronic pulse inverter topologies for driving DBD-based lamps can be found in [186]. The novel topologies are based on pure resonant operation or a combination of resonant operation and inductor pre-charge. Silicon carbide power semiconductors were

20 implemented in transformer-less topologies and increased the electrical power efficiency

21 of excimer lamps significantly. The piezo-transformer based DBD sources have already

23 been mentioned in the subsection 2.2.

26 5. Discharge development in filamentary DBDs

28 5.1. Methods and techniques

30 The diagnostics and modeling of DBDs and other plasmas at elevated pressures made

32 a significant progress within the last twenty years. This was possible by the technical

33 developments and the increase of computational power, but has been motivated by the

34 growing technological interest on atmospheric pressure plasmas in general. The diagnostics of DBDs was for a long time dominated by electrical measurements,

37 optical emission spectroscopy and chemical analysis. In the meantime methods for the

spatio-temporally resolved study of the discharge development, laser diagnostics for the determination of active species densities and for the quantitative measurement of 41 surface charges are available. Basic plasma parameters (reduced electric field strength,

electron density) has been determined by emission based methods and densities of metastable states have been measured by laser induced fluorescence. Reviews on 45 the methods for the investigation of microdischarges and microplasmas are given

elsewhere [187, 188, 189, 190] and recently optical diagnostics on atmospheric pressure plasmas are reviewed in [49, 191, 192]. The time resolution and high sensitivity of 49 laser induced fluorescence, two-photon absorption laser-induced fluorescence or broad-

51 band absorption spectroscopy allows the study of the development of key species like

52 metastable molecules, nitrogen atoms or radicals (e.g. OH) [188, 193, 194, 195, 196, 197]. Molecular beam mass spectrometry enables the in-situ detection of negative and positive ions from DBDs and the study of the ionic chemistry. In case of water vapor admixture

56 or impurities cluster ions are measured [198, 199].

58 A special approach is the investigation of single filament discharges. The overall

59 yield of plasma processes in DBDs is determined by the multitude of discharges, but

60 the properties of the single discharges determine the main processes. Thus, the rate

21 22 23

Progress on DBD Sources and Filaments 19

5 coefficients of excitation and ionisation processes as well as plasma chemistry depend

6 on the basic parameters in the microdischarge [14, 200]. The investigation of single 8 filaments is possible in special electrode arrangements. Two semi-spherical electrodes

9 (symmetric and asymmetric volume DBDs) [201, 202, 203, 204, 205, 206], pin-to-plate

11 electrode configurations (asymmetric volume DBD and surface DBD) [207, 208, 209] or

12 two pins embedded in an isolator (coplanar DBD) [210, 211, 212, 213] enable sufficient

13 localization and stabilization for dedicated studies.

15 The short duration of the microdischarges makes current measurements challenging.

16 Current pulses of single microdischarges are measured by well-shielded and fast current probes (Rogowski coils) [62, 214] or via the voltage drop across a non-inductive shunt resistor inserted between the grounded electrode and grounding lead [213] on the special

20 DBD arrangements mentioned above.

The spatial resolution of imaging and spectroscopic methods could be increased up to 10 |o.m and a technical time resolution of about 12 ps is reached today [215]. State-

24 of-the-art streak cameras reach sufficient time resolution of up to 50 ps and can also

25 analyze erratically appearing microdischarges [215, 216, 217]. A considerable progress

27 was also achieved in the analysis of cathode directed streamers in corona and surface

discharges [218, 219, 220, 221, 222].

The behavior and amount of charge carrier on the dielectric barriers can be 31 investigated by the electro-optic Pockels-effect [223, 224, 225, 226, 227, 228, 229]. A

crystal serves as the barrier or is directly attached to the discharge channel. The crystal changes its polarization properties with the electric field applied on it, which is 35 determined by the accumulated surface charges. The first applications of this principle

37 were devoted to the study of dynamics and formation of discharge pattern [229, 230, 231]

38 but are now also applied to measure the surface charge during the breakdown. Therefore,

39 the Pockels-method was extended for phase-resolved and thus, time resolved studies [232] and the time-resolution was significantly improved to 200 ns [233]. This method has been

42 directly combined with electrical characterization and LIF diagnostics for an identical

discharge arrangement in [195, 232, 234, 235]. Another approach for the measurement of surface charges is the fit of discharge delay time from electrical measurements [105]. 46 It is only applicable for very stable discharges.

It must be noted that some electro-optic crystals have distinct differences from the classical dielectric barrier materials. The dielectric constant is considerable higher, the

50 secondary electron emission coefficient is assumed to be smaller than for glass [98, 236],

and photo-induced transport of charges on the minutes-scale has been obtained for bismuth silicon oxide (BSO) crystals [98, 215, 237].

54 Accompanying the progress on the experimental studies there is an increasing

56 number of publications on the modeling of DBDs and other plasmas at normal

57 pressure, in particular corona discharges and plasma jets. These models calculate the

58 temporal behavior of the electric field distribution in the discharge gap coupled with the

60 dynamics of the charged particles. Different kinds of modeling approaches have been reported, including fluid models, Monte Carlo simulations and particle-in-cell models. A

Progress on DBD Sources and Filaments 20

5 combination of several models is referred to as hybrid models. Comprehensive overviews

6 can be found, e.g. in [238, 239, 240]. At elevated pressures, the energy gained by

8 the electrons from the electric field will be more or less balanced by the energy lost

9 due to collisions. Thus, the calculation of the average electron energy with an energy i n

11 balance equation in the fluid model is considered as a reasonable approach [241]. In

12 particular, microscopic fluid models are able to describe the rapid dynamics and behavior

13 of streamers [242, 243, 244, 245, 246, 247, 248, 249]. Many of these contributions are not only relevant for plasma technology but also related to the understanding of lightning

16 and transient luminous events in the upper-atmosphere.

In the fluid models the continuity equations for all considered species coupled with the Poisson equation are solved numerically [5, 14, 63, 64, 76, 239, 250, 251, 252].

20 In general the models also incorporate the plasma chemistry, with respective necessary

21 22 23

complexity. Rather extensive reaction systems including free radical reactions with more than 1000 elementary reactions have been investigated. Most of the models are one-

24 dimensional (1D, i.e. resolving along the discharge gap axis) but there are also 2D and

25 even 3D simulations [14, 252, 253, 254]. Normally the local mean energy approximation

27 is used. Thus, all electron transport parameters and rate coefficients can be described

28 as a sole function of the mean electron energy. The electron energy distribution function

30 is calculated by solving the stationary Boltzmann equation for prescribed electric field

31 strength [255]. Full kinetic models provide the electron energy distribution function as a function of space and time. Spatially one-dimensional simulations are indeed suitable for the analysis of diffuse DBDs. It has also been used for the analysis of

35 single microdischarges [256, 257, 258]. It is justified if the time scales which are inherent

37 in the discharge plasma are much shorter than the residence time of a unit gas volume in

38 the active plasma zone. In these models the electron energy balance equation is solved

39 to determine its mean energy. A new input in these activities are studies describing the plasma-surface interaction.

42 First approaches to model the surface electrons were via phenomenological rate

equations characterized by sticking coefficients, residence times, and recombination coefficients [259]. Other models consider very complex surfaces as plasma boundaries, 46 e.g. liquid layers [260] or catalysts [261]. A new approach is the theoretical study

of the microphysics of plasma-surface interaction [262]. The aim is that the plasma walls are not treated as perfect absorbers for electrons and ions but to describe the 50 charge transfer across the interface by an electron surface layer approach. It considers

52 the accumulation, distribution and release of surface electrons from the surface using a

53 quantum-mechanical surface physics approach. Electron impinging on a solid surface is

54 either reflected, inelastically scattered, or temporarily deposited to the surface. Possible

56 trapping states sites depend on their energy, the inelastic coupling to the elementary

57 excitations of the surface driving energy relaxation, and the electron affinity of the

58 material. The electron affinity is the energy gap between the vacuum energy and the

60 lower border of the conduction band and thus, it is related to the work function of the material. The properties of the trapping sites of electrons determine their corresponding

21 22 23

Progress on DBD Sources and Filaments 21

5 residence times and penetration depths. For simplicity a floating wall exposed to a quasi-

6 stationary plasma is assumed in the electron surface layer approach, i.e. complications

8 due to the presence of individual discharge channels are not considered, yet. Up

9 to now surface electron sticking coefficients and desorption times were calculated for i n

11 various uncharged dielectrics [263, 264, 265] and the distribution of the wall charge

12 across the interface between a plasma and a floating dielectric surface was determined

13 [266]. Furthermore, the release of electrons from the dielectric surfaces via de-exciting of metastable molecules was investigated [267, 268]. Recently, the absorption of electrons

16 by dielectric walls with positive electron affinity (e.g. alumina, glass) smaller than

the band gap was studied and energy-dependent sticking probabilities were calculated [269]. These investigations will provide a deeper insight on the discharge physics at 20 insulating surfaces and the calculation of basic parameters describing the plasma-surface


All these activities, but also dedicated studies on other atmospheric pressure 24 plasmas (corona discharges, plasma jets) increased the knowledge on the physics of

filamentary and diffuse DBDs significantly. In the following subsections only some

27 aspects will be emphasized further, namely the electrical description, the physics of

28 single filaments or microdischarges and the mechanisms determining distinct DBD

30 regimes.

32 5.2. Electrical behavior and characterization

35 For many purposes it is sufficient to characterize DBDs by overall electrical quantities,

36 such as the frequency and amplitude of the applied voltage, an average discharge voltage,

37 and the dissipated plasma energy or power. Besides the control of the discharge, these

38 measures are for example useful for the determination of discharge modes. This approach

40 was introduced in 1943 by Manley for ozone generators and is still well accepted in

research and technology [270, 271, 272]. However, it is not considering the processes at the insulator surfaces or the non-uniform breakdown. It has to be interpreted as a

44 purely macroscopic characterization which is limited in many situations, e.g. when the

density of filaments per period is low. Recent activities were addressed to extend this approach in order to include the elementary processes of discharge physics. The topic

48 thus has been developed from a purely macroscopic characterization to an approach that

can even be used to characterize single filaments [273] or structured discharge regimes [48, 232].

52 The simplest electrical approach of a DBD is the lumped-element equivalent electric

54 circuit as shown in Fig. 9 b [86, 274, 275, 276]. In the non-ignited case the equivalent

55 circuit is purely capacitive and consists of two capacitances, one is resembling the barrier

Cd and the other is the gas gap Cg (A in the equations (1) and (2) is the area of the electrode). Both together in linear arrangement form a total capacitance Cce11. The sum 59 of the voltages across the gas gap Ug and the barrier Ud equals to the applied voltage

V(t). The voltage Ug is too low to cause discharge ignition as long as the amplitude

10 11 12

20 21 22

Progress on DBD Sources and Filaments

Voltage '



Ccell "V0-

-C„n wV„

Inception Termination


œ < C

Figure 9. The classical electrical characterization on a asymmetric, one-sided volume DBD configuration (a) with the most simplest equivalent circuit (b), the schematic development of applied voltage V(t), gap voltage Ug, burning voltage Ub, mean current i and its capacitive component icap (c) and the corresponding voltage-charge plot (d). The circled (i) and (t) stands for "inception" and "termination".

of the applied voltage is below a certain threshold Vmin and the current i is only the capacitive displacement current icap through the arrangement (amplitude: Ccea-dV/dt).

Cd = £o£r A/d (1)

Cg = A/g (2)

Ccell = (Cg Cd)/(Cg + Cd) (3)

The inception of the discharge is obtained when the applied voltage magnitude Vo ceeds the threshold Vmin at which the instantaneous gap voltage reaches a certain


Progress on DBD Sources and Filaments

5 threshold, the burning or discharge voltage Ub. It is a macroscopic parameter which is

6 determined by the gas composition, its pressure and the discharge gap (comp. Paschen 8 law) and it enables the determination of the spatially and temporally averaged reduced

9 electric field strength. Figure 9 (c) schematically shows the voltage and current in an

II AC operated DBD when the discharge is ignited. An additional current is obtained

12 in the active phases between inception ("i") and termination ("t"). In these phases

13 Cg is bypassed and charge is transfered into the gap. This is described by the parallel

15 connection of a variable resistor RP(t) and gas gap capacitor Cg in the equivalent circuit.

Q (t) are as follows.

20 1 21 ip 22

The equations for the plasma current iP (t), the gap voltage Ug (t) and the total charge

1 — Ccell /Cd

i(t) - Ccell ^ y (4)

24 U = V(t) - ^Q(t) ^ 7 (5)

Q(t) = (l - Cg^ Qp(t) + CcellV(t) (6)

QP is the charge which is transferred conductively during the discharge

30 (iP =dQP/dt). The variable resistor RP (t) can also be symbolized by a variable current

31 source which would make no difference on the considerations about uniform DBDs as

33 discussed in [271, 277, 278]. In this classical approach Ug is considered to be almost

34 constant during the active phase with Ug ~ Ub (also shown in Fig. 9 c). However, this is not entirely the case for pulsed operated DBDs. Here the applied voltage changes on the typical timescale of discharge inception and thus, the gas gap voltage is not constant

38 during the active discharge [279]. However, the equivalent circuit in Fig. 9 b covers this

40 aspect as well as the change of effective capacitance (capacity divided by area) in case

41 of non-uniform discharging (see below). It has to be mentioned that the current slope

42 shown in Figure 9 (c) does not resemble the measured curve in a filamentary DBD with

44 a limited number of discharge events which is obtained when the voltage amplitude V0

45 is just slightly exceeding Vmin. The discharge current then persists of individual peaks

46 and the analysis is more complicated but can also give information about the individual

48 events [57].

49 The investigation of voltage-vs.-charge (V-Q) plots is an established approach since the charge measurement do not require high bandwidth probes and oscilloscopes, in contrast to the recording of current waveforms [272]. A measuring capacitor CM is

53 inserted between the grounded electrode and grounding lead and records the overall

55 charge. The value of CM must be chosen to be significantly larger than Ccea (as a rule

56 of thumb a factor 1,000). In case of sinusoidal high voltage the V-Q plot is a straight

57 line with slope Cceu when the discharge is not ignited (and if there are no phase errors on

59 the high voltage probe) [272]. When the plasma is ignited and if Ug is constant during

60 each active discharge phase the V-Q plot is a parallelogram, as schematically depicted in Fig. 9 d. The four sides represent the different phases during one high voltage

Progress on DBD Sources and Filaments 24

5 period T = 1/frep (frep is the frequency). The parameters Cce11, Cd, Vmin and Ub as well

6 as the total plasma energy per high voltage cycle Ee1 and total power P = frep ■ Ee1

can be extracted directly from the parallelogram [50, 52, 53, 214, 272, 275]. Recently,

9 measurements of Ee1 were discussed in the context with thermodynamic measurements i n

11 (performed by fiber-optic thermometers) and calorimetric investigations [280]. Fair

12 agreement between both methods was archived. Energy per cycle or power can be

13 related to the volume of the plasma or the gas flow resulting in specific energies or

15 power densities. To give the values of such similarity measures should become a standard

16 in publications since they enable a reliable comparison of different DBDs and plasma


However, a large variety of V-Q plots, also referred to as Lissajous-figures, diverging

20 from parallelograms is reported in the literature. Round edges at the discharge

21 termination points indicate parasitic resistances Rpar which can occur when the barrier

23 material is heated and thus, becomes conductive. In such cases the V-Q plot can even

24 be deformed to an ellipse and its enclosed area is not necessarily a measure for the

26 dissipated plasma energy [280]. In case of pulsed operated DBDs, asymmetric discharge

27 arrangements, or, unstable high voltage power supplies the shape of the V-Q plot can also differ significantly from a parallelogram (Ug not constant) and one needs to analyze the obtained plots very carefully. If the filaments are ignited in an organized

31 regime (pulsing action or simultaneously ignited groups of microdischarges) the slopes

corresponding to the plasma phase are stair-like [272, 281]. To obtain internal discharge parameters such as the discharge current, Ub, and instantaneous power P(t) and energy 35 E(t) in these situations requires the exact knowledge of the capacities Cce11 and Cd.

37 These values are then not necessarily determined from the slopes and its calculation is

38 not straightforward. The overall capacity Cce11 can be determined from the comparison

39 of the measured current with dV/dt when the discharge device is not ignited. The slope

41 of Q0 — V0 points (obtained from a passel of V-Q plots measured at several voltage

42 amplitudes) can be used to determine Cd [279]. The influence of experimental errors of the values for Cce11 and Cd was discussed in detail in [282] in particular for pulsed operated DBDs.

46 In real DBD devices stray capacitances CStray are caused by surrounding

capacitances (e.g. cables, edge effects, high voltage throughputs). The resistivity of the secondary side of the high voltage power supply (e.g. transformer series resistance)

50 RS and the parallel losses of energy due to the barrier material and reactor construction

(symbolized by resistivity Rpar) can significantly influence the current and charge that are to be measured [283]. Fig. 10 (b) shows how these parasitic elements can be

54 implemented in the equivalent circuit. The technique of impedance characterization

utilized in [186] also allows the quantification of parasitic electrical elements of DBD

57 lamps and its circuits.

There has been many studies on the construction of equivalent circuit for DBDs. For example, a circuit for characterization of diffuse DBDs presented in [77] incorporates Zener diodes in order to model the Townsend breakdown and an additional capacitance

10 11 12

20 21 22

Progress on DBD Sources and Filaments

' C. Cd

a + ß= 1

°Tr P1 "T

Cgi + Cg2 = const.

Figure 10. Advanced equivalent circuits for DBDs: (a) circuit including stray or parasitic elements; (b) partial discharged volume DBD ©2015 IOP Publishing. Reprinted, with permission, from [48]; (c) surface DBD actuator [288]; (b) and (c) can also be affected by external elements (see outside the dashed box in part (a))

is introduced to describe the memory effect from one half period to the following. This capacity represents the space-charge cloud formed by charge carrier separation during plasma generation in pulsed operated DBDs [186, 284, 285].

Multiple microdischarges can penetrate the entire discharge area cross section in case of high overvoltages, i.e. V0 ^ Vmin and in plane-parallel discharge arrangements with minor electrode edge effects [279, 286]. In case of DBD arrangements with a low number of microdischarges per half period the individual discharge events correspond with distinct steps in the "wt-lines" of the parallelogram in Fig. 9 d [58, 287]. The mean slopes of this line are then depending on V0 which is attributed to the partial discharging. The modified equivalent circuit in Fig. 10 b considers this by splitting the apacitances Cd and Cg into two sections in order to describe a non-discharging areal a and a discharging areal fraction 3 (only this part has to implement RP (t)) 48]. The slope of the wt-lines (also called effective or apparent dielectric capacitance) is then smaller or equal to Cd. Its value saturates to Cd with sufficient increase of V0

capacita] fraction [48]. Th

21 22 23

Progress on DBD Sources and Filaments 26

5 in a plane-parallel arrangement [286]. In case of the ribbed design of one electrode as

6 shown in Fig. 3 c a sharp increase of the effective dielectric capacitance with the voltage 8 amplitude is obtained [48] which can be described as a linear combination of Cce11 and

9 Cd. This analysis finally leads to the conclusion that the effect of partial coverage of

11 the dielectric by spreading surface charges is analogous to implementing Cstray in the

12 circuit which significantly affects the determination of Ub, the net charge density across

13 the gap during discharging or the resistivity of the plasma averaged over the discharge ^ area [48].

16 Surface and coplanar DBDs require even more complicated equivalent circuits which

were mostly developed for plasma actuators. A typical example is given in Fig. 10 c. The edges at point "i" of V-Q-plots are often not sharp as in Fig. 9 d which indicates 20 a non-equal voltage threshold in non-plane discharge gaps or surface DBDs [288]. In

surface DBDs the exposed surface of the barrier acts as a third, virtual electrode as it collects charge. Therefore, three barrier capacitances Cd1, Cd2, Cd3 and two gas 24 gap capacitances Cg1 and Cg2 must be considered. The capacitor Cg1 represents the

26 capacitance between the exposed physical electrode (compare with Fig. 1 b) and the

27 virtual electrode. The capacitor Cd1 represents the capacitance between the virtual

28 electrode and the encapsulated or covered physical electrode. These capacitors are

30 variable with V0 since the area covered by plasma is not constant. This could be

31 confirmed by studies on surface DBD reactors where the total capacitance of the plasma reactor increases when the operating high voltage amplitude is increasing [289]. The capacitor Cd3 is constant as it is describing the constant capacity due to the direct

35 electric field between the physical electrodes. It induces an additional displacement

current but does not affect the discharge itself, i.e. the effect is similar as that of Cstray [288, 290]. Another approach to describe the filamentary, non-regular character of DBDs 39 is the parallel distribution of a certain number of serial connections of finite gas gap and

dielectric barrier capacitances [291].

5.3. Single filaments in volume and on surfaces

In air and other DBDs containing nitrogen gas the emission is dominated by molecular bands of nitrogen, namely the second positive system of molecular nitrogen and the first 48 negative system of molecular nitrogen ions. The spatio-temporally resolved development

50 of filaments in a sinusoidal voltage operated volume DBD generated between two semi-

51 spherical electrodes in synthetic air is shown in Fig. 11. An overview spectrum is

52 shown at the right bottom part of the figure and the investigated spectral ranges

54 are marked. The band heads of the vibrational 0-0 transitions of the second positive

55 system of nitrogen at AC = 337.1 nm and of the first negative system at AB = 391.5 nm are

investigated. The method used to obtain the temporal development is cross-correlation spectroscopy, the most sensitive method for the investigation of irregular appearing

59 microdischarges [201, 203]. The discussed results have also been used to calculate the

60 axially and radially resolved evolution of the reduced electric field strength [292].

10 11 12

20 21 22

Progress on DBD Sources and Filaments

Xc= 337 nm (2nd pos. system)

(a) t= 0.4 ns (b)t= 1.5 ns (c)t= 1.7 ns

(d)t= 2.0 ns (e)t= 2.1ns (f)t= 2.2 ns

XB= 391 nm (1st neg. system)

(a) t= 0 4 ns (b) t= 1.5 ns (c)Ë 1.7 ns

r+-------- ---------

■0.3 ' Ô ' o'î (d) t= 2.0 ns (e) t= 2.1ns (f)t= 2.2 ns


(g) t= 3.2 ns (h) t= 3 8 ns (i) t= 6.6 ns

t= 3.2 ns (h) t= 3.8 ns (i)t= 6.6 ns

Relative Intensity .....

100 50 25 12 6 3 1.5 0.7 0.3 0.15 0 08 time


100 59 33 20 11 6 .4 2 1.3 08 0.4

N2(c_b)o-o \ Emission

spectrum in air



wavelength A.

Figure 11. Evolution of light-intensity distributions for 0-0 vibrational transition of the second positive system (AC = 337.1 nm) and the first negative system (A b = 391.4nm) of nitrogen (see emission spectrum for spectral range A= 290... 450 nm) of microdischarges in AC-operated volume DBD in synthetic air at atmospheric pressure (discharge gap 0.9 mm; voltage amplitude of 12kVpp; frequency of 6.9 kHz; accumulation over about 109 microdischarges in total) ©2008 IEEE. Reprinted, with permission, from [294].

For each transition the observed spatio-temporal distributions of the microdischarge luminosities are determined by the population of the upper electronic states via direct electron excitation and collisional quenching. The excitation rates are directly proportional to the electron density and the rate coefficients [293]. According to the corresponding excitation energies (11 eV vs. 18.7eV) the signal at AB represents the development of E/N, while the signal at AC can be attributed as a convolution of the electric field and the electron density. From the results in Fig. 11 the microdischarge development can be described as follows.

21 22 23

Progress on DBD Sources and Filaments

5 (1) The (Townsend) pre-phase is characterized by a weak, localized light spot at

6 the anode (see Fig. 11 (AC, a). It is caused by a slight distortion of the electric field 8 due to charge accumulation in front of the anode by means of electron avalanches. This

9 phase can last for several 100 ns [201, 204].

11 (2) The propagation phase of a cathode-directed ionisation wave (positive streamer

12 mechanism) starts when the local electric field strength of the accumulated positive

13 space charge reaches a critical value. As shown in Figs. 11 b-f it transits the discharge

15 gap within about 2 ns.

16 (3) The two local maxima in Fig. 11 g and h of AC (left series of pictures) correspond to two active zones of the microdischarge with different properties [201]. The anode glow is caused by electrons drifting and accelerated behind the ionizing wave

20 towards the anode. Near the cathode, a high electric field is formed. Due to charge

accumulation on the dielectric surfaces, considerable broadening of the microdischarge channel, branching, and expansion on the insulator surfaces is obtained. 24 (4) The decay phase is initiated by the accumulation of charge carriers on the

26 barrier surfaces. The axial electric field is reduced within about 10 ns (Fig. 11 h and

27 i) and the plasma current drops down. Charge carriers and long lived excited species

28 from the discharge channel can remain in the discharge gap. Ions are drifting to the

30 electrodes on timescales of 1 |xs, the most important chemical reactions occur on the

31 micro- to millisecond time scale [2, 4, 200]. The anode glow phase can show an even more complex behavior. A striated or

stratified structure as known from low and medium pressure glow discharges has been 35 investigated for single microdischarges in argon at atmospheric pressure under certain

37 conditions [295, 296].

38 The analysis of the single filament volume DBD in synthetic or dry air

39 along the microdischarge axis (without radial resolution) for asymmetrical discharge configurations [204] and for a coplanar DBD arrangements [210, 211] have shown

42 qualitatively the same mechanism of electrical breakdown and discharge development,

but, the experimental findings reveal that residual surface charges on the dielectric covered electrode cause local enhancement of the electric field in the pre-phase and 46 promote the emission of electrons. This is in agreement with simulations [252, 297].

The decay phase of the microdischarge development is determined by the properties of the anode. The accumulation of negative charge on the surface of the dielectric covered

50 anode results in a faster decrease of the electric field strength within the discharge

channel as for the case when the anode is metallic. Above the cathode surface a long-lasting (up to 150 ns) and about 30 ^m thin layer of enhanced electric field strength is

54 observed. These values correspond to simulation results describing the cathode layer on

a metal cathode in asymmetric volume DBDs [63, 64].

57 The pulsed operation with rectangular high voltage pulse with a gradient of

58 about 250V/ns of a similar single filament volume DBD as discussed above revealed 60 a comparable discharge development as in sinusoidal operated DBDs. But, due to the

three orders of magnitude higher voltage steepness in case of the pulsed DBD the pre-

10 11 12

20 21 22

Progress on DBD Sources and Filaments

phase is limited by the pulse rise time. Thus, less space charge can accumulate before the breakdown. This leads to a shorter pre-phase but to an increase of the burning voltage and, consequently, to a larger amount of transferred charge and a higher amount of dissipated energy [216, 298]. The significant rise of the pulsed applied voltage during the pre- and propagation phase enhances the charge carrier generation.

0.150 0.050 0.020 0.010 0.002 0.001

Figure 12. Spatio-temporal development of the emission at AC = 337.1 nm at the falling slope of a rectangular high voltage for tpuise= 10 |is (top diagrams) and 1 |is (bottom diagrams) showing the 2D structure (ICCD) during the pre-phase (left) for a time window of 5 ns, indicated by the grey dashed lines in the streak image (right). (Vo=10kV; T= 100 |is; gas: nitrogen with an admixture of 0.1vol.% oxygen) [217]. With permission from the author.

The pulsed operation mode enabled the study of the development of a microdischarge evolving in the decay phase of a preceding discharge event. Microdischarges occur at the rising and the falling slope of an unipolar high voltage pulse, while the latter is initiated by the electric field of the remaining surface charges (back discharge) [257, 299]. In case of a symmetric high voltage pulse shape the microdischarges have the same properties. In case of shorter pulse width of the high voltage (tpuiae) the discharge at the falling slope can be significantly influenced by the remaining volume charges [257]. The fig. 12 presents the effect of short pulse widths (pause times) of 10 |is and 1 |is on the spatio-temporally resolved development of single filament volume DBD.

The ICCD photos (left) visualize the pre-phase of the corresponding microdischarges. For tpulse = 10 |is, an isolated, temporally limited SPS emission occurs in front of the anode. The space charge maximum and the inception point of the cathode-directed ionisation front is shifted into the volume. For tpulse= 1 |is the emission during

10 11 12

20 21 22

Progress on DBD Sources and Filaments

Surface DBD

exposed needle electrode


hidden needle electrode

Side view ICCD photos

12 14 16 18 20 22 24 26 26 30 %ie (ns)

Figure 13. Surface DBD electrode arrangement with gated ICCD-photos in top view and side view of microdischarges in dry air for different amplitudes of the applied sinusoidal voltage and development of the intensity at AC=337.1 nm with 6.4kVpp (Solid line represents exposed anode tip and dashed line tip position of hidden cathode) [209, 235]. ©2009 Wiley. Reprinted, with permission, from [235].

the pre-phase starts near to the cathode and from this region simultaneous cathode-and anode-directed waves of luminosity originate. The remaining positive ions in the gap create a "virtual" anode which displace the inception point of the ionisation wave close to the cathode. For shorter tpuise < 500 ns this effect becomes even more evident since no propagation phase is obtained at all.

The experimental and modeling results for pulsed-operated DBDs in nitrogen-oxygen gas mixtures suggest a strong influence of the gas composition on the recombination rate of the different positive molecular ions of nitrogen and oxygen [257, 300, 301]. As shown in [302] the convective transport of particles by gas flow also affects the residence time of charge carriers and thus, the amount of residual volume charges. Its role was also emphasized in [60, 303] for a multi-filamentary DBDs.

Any discharge in contact with an insulator causes discharge channels along the dielectric-gas interface. The discharge cell shown in Fig. 13 enabled the study of microdischarges by optimizing the electrostatic field geometry (see the ICCD photos) 209, 304]. It consists of two needle electrodes placed on the opposite sides of an alumina he strong asymmetric geometry leads to different microdischarge properties in two half periods of AC high voltage [250, 305].

If the exposed needle is at positive potential the microdischarge starts with a short

U= 4.3 kV

U= 6.4 kV

10 11 12

20 21 22

Progress on DBD Sources and Filaments

Operation start

Re-ign ition

approx. 1 min of plasma operation

Re-ignition voltage V

2 4 6 8 10 12

Time between re-ignitions (hours)

ignition minimum 0 voltage

Figure 14. Difference of ignition and minimum sustainment voltage Vmin on a DBD (semi-spherical electrodes; discharge gap 1.2mm; frequency 6.7 kHz) in nitrogen (left). Development of Vmin as a function of time between re-ignitions after one minute of discharge operation (right). ©2009 IOP Publishing. Reprinted, with permission, from [307].

pre-phase at the tip of the anode followed by the cathode directed ionisation wave. This is similar as observed in volume DBD and coplanar discharges [204]. But, the maximum velocity of the ionisation wave is about one order of magnitude lower and decreases with increasing distance from the anode. This was also suggested in other surface DBD arrangements [305, 306]. An anode glow is also not obtained [209]. With an increase of the voltage amplitude additional discharge channels develop [304]. While the first microdischarge in the half cycle takes the direct path between the exposed electrode tip and the opposite position of the electrode beneath the dielectric, the following plasma channels evade the region of the preceding breakdown thus, taking a curved and thus longer path. Obviously, the surface charges lead to this specific discharge patterns [304].

The long time stability of surface charges manifests in the fact that the high voltage amplitude for the very first ignition (also called ignition voltage Vign) is much higher then the minimum external voltage for sustaining a once operating DBD (Vmin). This is shown in Fig. 14 for a volume DBD (semi-spherical glass covered electrodes) operated by sinusoidal high voltage in nitrogen [307]. In the left diagram the difference between Vign and Vmin is demonstrated. About 28 kVpp are necessary to ignite the discharge for the first time. This value has a large spread since the first ignition is a statistical phenomenon. After one minute of plasma operation the applied voltage amplitude can be significantly decreased without extinguishing the plasma. Obviously, a residual bias potential is formed. The right diagram demonstrates the decay of this bias. The voltage amplitude necessary for re-igniting the discharge after a defined time without plasma operation increases slowly and it takes about 14hrs to reach the original conditions.

Progress on DBD Sources and Filaments

5 5.4. Filamentary behavior and memory effects

7 In a DBD reactor with a filamentary discharge many microdischarges are

generated. In general, the higher the amplitude of the applied voltage, the higher number of filaments per cycle. The higher the number of filaments per dielectric surface

11 area, the larger is the cross section of electrode surface penetrated by microdischarges.

In plane parallel arrangements only the number of discharge events, but not the

14 charge transfered per event are affected by the amplitude of the applied voltage

15 [57, 59, 60]. The structure and behavior of DBDs can be manifold. A large variety

16 of filamentary static or dynamic structures were obtained and they were summarized, 18 e.g. in [13, 87, 308, 309, 310]. The stationary behavior of the filaments was investigated

and described by Monte Carlo based simulations [6, 311]. The regime and dynamics of several microdischarges and their pattern are determined not only by the surface 22 properties and the geometry of the electrode arrangement (which also determines the

amount of charge dissipated in one channel), but also by the parameters of the high voltage, namely frequency and amplitude [200]. The properties of the gas has also to be

26 considered. More recent results obtained with semi-conductive barriers are presented in

[12, 35, 312].

29 The most general behavior, namely a filamentary DBD with eratically appearing

and distributed microdischarges in a plane parallel DBD is shown in Fig. 15. It presents the result of the time-resolved measurement of the surface charges via the Pockels-effect 33 in a helium-nitrogen gas mixture together with typical voltage and current waveforms.

The current is without the capacitive component. It has to be mentioned that the value of transferred charge determined from electrical measurements nearly equals the value 37 of measured surface charges in many different discharges [98, 232, 234, 314].

Before current pulse appears in the first half period (t= 0-500 |xs) surface charges from the previous breakdown (with the opposite polarity) are measured on the electro-

41 optic crystal. In this example the spots of negative charge are obtained, i.e. the

electrode covered by the crystal was the anode in the foregoing half period. During

44 the rise of the high voltage slope current pulses are generated and the spots change

from negative to a positive surface charge density since the crystal is now the cathodic dielectric. Obviously, the microdischarge inceptions are preferred at positions where

48 discharge events happened in the foregoing half period because the surface charges

enhanced the applied field in this situation. The surface charges are also inhibiting the formation of microdischarges at the same positions during the same half period. This 52 is apparent from the fact, that (1) the spots do not exceed a certain maximum (about

54 15 C/cm2) and (2) the number of positive spots increases during the rise of high voltage.

55 The surface charge pattern remain constant and thus determine the distribution in the

56 following active phase (t= 1000-1500 |xs). A detailed analysis of the surface charge

58 density profiles show a Gaussian distribution with a constant offset value [98, 232, 233].

59 Both contributions of surface charges are inverted during each following half period.

60 New filaments can occur in the discharge free areas. Sometimes, a filament will not

10 11 12

20 21 22

Progress on DBD Sources and Filaments

\ I 1 \

-15WI-5 0 5 10 15 \ o [nC/cm2]

Figure 15. Voltage and net current characteristics of a filamentary DBD in helium-nitrogen gas mixture (He/N2=9/1) operated at 0.5kHz frequency and 1.75kV amplitude (top). Grey bars show the time windows (50 |is) for phase-resolved surface charge measurements over a full period of applied high voltage (bottom) [313], with permission from the author

re-ignite at the former position in the following one or more (up to three) periods of the applied voltage. In this situation positive as well as negative surface charge spots are obtained simultaneously but at different locations [232].

Negative surface charges are supposed to be electrons while positive surface charges are assumed to be defect electrons in the dielectric resulting from electron-ion-recombination at the surface [265]. The electron surface layer model in [262, 266] reveals that the axial location of the surface charge along the plasma-dielectric interface region is determined by the electron affinity of the material. The defect electrons are situated in its valence band. The plasma electrons get either trapped in low-energy image potential sites just in front of the crystallographic material boundary (in case of negative electron

21 22 23

Progress on DBD Sources and Filaments 34

5 affinity materials like magnesium oxide) or in the conduction band inside the dielectric

6 with subsequent energy relaxation (in case of positive electron affinity materials like

8 quartz or alumina). The electron affinity of electro-optic crystal has not yet been

9 calculated but, the the observed stability of the surface charge profiles suggest positive i n

11 affinity [98]. During the discharge channel formation, the initially accumulated surface

12 charges sidetrack the following charge carriers of the same polarity and thus, the lateral

13 extent of the surface charge spots is determined by the mobility of the responsible charge carriers [229]. The important role of surface processes is also known from investigations

16 of partial discharges where they propagate along the gas-dielectric interface [315]. Beside

electron impact ionization charges can be produced via surface ionisation due to de-trapping, ion impact and photo effect and attachment of charge carriers at the surface 20 has also to be considered.

The electro-optic measurements discussed above suggest that the surface charges are more or less static. The existence of a constant Ub in case of sinusoidal operation

24 throughout a discharge half period in the equivalent circuit model implies that the

26 microdischarge inception starts at approximately the same critical voltage, irrespective

27 to the instantaneous value of the applied voltage at the inception time. This behavior

28 is attributed to the existence of the surface charge which is somewhat contradictory

30 to the explanation of multi-filament DBDs [58]. Dedicated electrical measurements in

31 a DBD arrangement with relatively low number of distinguishable filaments combined with the analysis of surface coverage and the statistical distribution of the transferred charge per microdischarge leads to the conclusion, that the larger part of the fraction of

35 the deposited charge is mobile on time-scales of hundreds of ns. Such a plasma-induced

37 surface conductivity could explain the offset values of the surface charge profiles [232]

38 as well as the charge transport in case of inhomogeneously charged surfaces obtained

39 in [237]. It would also fit with the proposed deposition of plasma electrons in the conduction band of the dielectric material [262, 266]. However, these electrons can

42 become trapped in localized impurities or intrinsic trap states below the conduction

band resulting in the observed "memory charges" at the filament spots in case of alumina as the barrier material [316, 317, 318]. 46 Humidity is known as an important factor in the physics and chemistry of DBDs.

For example it tends to increase the current pulse amplitude of a microdischarge [200, 214, 319] induced by a higher surface conductivity [5, 320]. Experimental results on 50 plane-parallel DBDs in helium combined with surface charge measurement have shown a

52 correlation between humidity in the system and the dynamic behavior of filaments [228].

53 The filaments are non-stationary under humid conditions. The trajectories formed by

54 subsequent filaments resemble a random walk motion. Obviously, the plasma channels

56 evaporate the water film on the dielectric locally. As the water vapor concentration

57 increases the ignition voltage, the filament is not ignited in vapor cloud and the next

58 microdischarge incepts in the adjacent area. This behavior can be regarded as the

60 motion of self-propelled quasi-particles due to mutual interaction of the plasma channel footprints and the surface covered with adsorbed water vapors. The motion stops and a

21 22 23

Progress on DBD Sources and Filaments 35

5 stationary filament is formed when the discharge inception takes place at a location

6 with less humidity at the surface, i.e. where the water film was evaporated by a

8 previous discharge event. Taking this complex and still not fully explored interaction

9 into account, the analysis of the electrical behavior in arrangements with a low number of i n

11 microdischarges similar as in [58] under controlled gas atmospheres would be beneficial

12 to understand the role of humidity and surface conductivity.

13 A collective behavior of filamentary volume and surface DBDs in time, i.e. the simultaneous inception of individual microdischarges can be confirmed from

16 synchronized ICCD camera recordings in combination with current measurements

[305, 310, 321, 322, 323, 324, 325]. From these studies an important role of radiative processes is argued. The VUV emissions from the first microdischarges may lead to 20 photoionisation in the vicinity of the electrodes or to photodesorption of electrons from

the charged insulator surface [321, 324]. More details on the role of photoionisation are available from the investigation of corona discharges [301, 326]. In nitrogen-oxygen 24 gas mixtures the number of VUV photons is determined by the nitrogen concentration

26 (atomic lines) while the oxygen concentration determines the absorption length of these

27 photons. Thus, direct photoionisation mechanism is not considered as the dominant

28 source of free electrons, in particular at low oxygen concentrations. Although the

30 mentioned investigations have been performed in air which is known to absorb VUV

31 photons efficiently (mainly by molecular oxygen), there exist photoabsorption windows in the VUV region [321]. A much more pronounced effect could be expected for discharges in noble gases. Here, VUV photons are significantly generated by excimers

35 and less absorbed in the carrier gas [67]. Studies on simultaneous breakdown in noble

37 gases has not yet been performed, but could clarify the role of radiative processes.

38 The collective behavior of filamentary and patterned DBDs can be influenced

39 by residual species in the volume [60, 303]. At certain overvoltage the formation of microdischarge in the discharge channel from a previous one in the same half period

42 was obtained in coplanar DBDs [327]. It could be confirmed by single microdischarge

studies in [328], where a complex branching of discharge channels against the formation of new channels was observed. In [329] the effect of residual heat was considered to 46 lead to the spatial stabilization of the filaments. The increase of temperature and thus

decrease of gas density can lead to a significant reduction of breakdown voltage which favor microdischarge inception at the previous position.

55 mode are quite specific and determined by the portion between secondary processes

56 at the electrodes and ionisation processes in the discharge volume. The APGD is

58 usually formed in gases with a relatively high gas ionization at comparatively low

59 electric field strength and the ionisation dynamics is further delayed by multistage

60 ionization processes. The content of metastable queching molecules in the gas will

5.5. Transition between filamentary and diffuse DBDs

As discussed in section 3.2 the conditions for the operation of DBDs in the diffuse

21 22 23

Progress on DBD Sources and Filaments 36

5 therefore determine whether the DBD will operate in diffuse regime or not

6 The main feature of the APTD is that the ratio of secondary electron emission at the

8 cathode to the ionization in the discharge volume is relatively high and the metastable

9 N2(A) species are considered to enhance secondary electron emission from the cathode i n

11 (i.e. the previous anode with adsorbed electrons on the surface) [80, 195, 262, 268, 330].

12 The gas composition also affects the portion of these processes since the effective lifetime

13 of the N2(A) metastables can be significantly reduced by collisional quenching [100, 101]. Another example where secondary electron emission is important are plasma display

16 panel cells [76, 93, 250]. Here, a layer of magnesium oxide which has a high secondary

electron emission coefficient is covering the dielectric surface.

A certain overvoltage on APGD or APTD also leads to filamentation. Systematic 20 measurements in [98, 331] have shown that an increase in the voltage slew rate

(dV/dt) results in higher power input and thus higher volume ionization which favors filamentation [332]. Filamentary and diffuse mode can coexists in such cases (also

24 called hybrid mode). Due to different secondary electron emission coefficients of

26 materials hybrid modes are obtained in asymmetric discharge cell configurations where

27 filamentation occurs when the electrode with the smaller secondary electron emission coefficient acts as the cathode [98, 236]. Furthermore, the diffuse mode can also show a radial development which is supposed to be correlated with surface processes

31 [88, 98, 237]. In particular in the APGD regime a variety of complex self-organized

static or dynamical structures of discharge spots can be observed. Annihilation, motion and self-organization of discharge filaments (typically of reaction-diffusion systems, e.g. 35 hexagons or stripes) can be obtained [13, 332, 333, 334]. Experiments and models

37 conclude that filaments are always associated with low current, "side discharges" that

38 develops in its vicinity beyond the inhibition region. The properties of these side

39 discharges can explain many aspects of filament interactions, dynamics, and pattern

41 formation [333].

42 Still an open question in DBDs is the role of negative ions on the development and chemistry in gas mixtures containing electronegative compounds. Laser photodetach-ment of O-, O-, and O- ions performed on an APGD in helium with admixtures of

46 oxygen shows an influence on the breakdown characteristics which indicates an enhanced

pre-ionization aser detached electrons.[335].

Progress on DBD Sources and Filaments

5 6. Conclusion and perspectives of DBD research

Although used since more than hundred years DBDs are still one of the most important technological plasma sources for the generation of nonthermal gas discharges at

10 atmospheric pressure. Researchers have been able to adapt the DBD principle on many

12 different tasks and objects. The presented collection of DBD arrangements in this review

13 is aimed to inspire further novel designs and solutions toward the handling of technical

14 challenges in the future. These are mostly in the field of chemistry (e.g. removal of

16 pollutants, plasma synthesis), life-science (e.g. decontamination, would healing and

17 other therapies) and flow control.

DBDs are well-scalable, robust and good-controllable plasma sources. Thus, the interest on the development of new sources and understanding its fundamentals will

21 continue. The further miniaturization of plasmas towards higher surface-to-volume

ratios could enable new and more efficient processes. The evolving opportunities of nano-technology will enable the development of new diagnostics and more sophisticated 25 plasma sources. Novel power sources give the chance to control plasma parameters and

27 to enhance the efficiency.

28 The development of new DBD reactors or treatment devices should always be

29 accompanied by a reliable characterization. In particular the electrical characterization

31 sketched here and in the cited literature should be mandatory. In connection with other

32 diagnostics and simulation this relatively simple technique allows not only a macroscopic characterization, but, also further insights on the plasma properties and dynamics. Understanding the microdischarges as the most fundamental elements of a DBD process

36 is one important key for the optimization of existing processes and the development

37 of new applications. Nowadays, several techniques enable the desired sensitivity and

39 time resolution for the study of the development of these transient discharges and the

40 determination of the basic plasma parameters. Following and intensifying the discussed

42 work will open up more possibilities for application and research. In particular the

43 merging of single microdischarge studies with multi-filament DBDs is still an open

question and requires different approaches, including theory, modeling and experimental work. Other open questions regard the role of radiative processes, surface properties,

47 and humidity in the system. Dedicated experiments under well-defined conditions are

required for these studies.

The relevant processes in DBDs and other nonthermal plasmas are characterized

51 by different time scales. The "volume memory processes" are connected with the

53 recombination of charge carriers in the discharge gap and thus, typically proceed in

54 the range of microseconds. The "surface memory processes" are related to the amount

55 of surface charges and the mechanisms of charge trapping and transport on dielectric

57 surfaces. The typical timescales are in the range of seconds or higher. But, even

58 faster surface processes are under discussion and the understanding of plasma-surface interaction in these discharges deeper is still a big scientific challenge. This is also the case for the investigation of different discharge modes and operation regimes. A

Progress on DBD Sources and Filaments

5 much larger diversity of this is known today than twenty years ago and researchers are

6 capable to control and reproduce patterned and diffuse DBD modes. The control of 8 DBD uniformity is not only important for its exploitation in surface processing but also

9 a fundamental question for gas discharge physics. Today much more is known about

11 the microphysical processes in these plasmas and at the plasma-surface interface, but it

12 should be extended to more complex boundaries, e.g. catalytic surfaces or liquid layers.

13 To remember the situation in the ozone generator research in the 1970's at this

15 time, it was not expected that basic research on a discharge known about for more than

16 100 years could have far-reaching technological and economic consequences. However, the efficiency of such generations was significantly improved and DBDs have opened up many new applications in the following years. It seems that this evolution is still going

20 on.

23 Acknowledgments

With deepest gratitude and respect I dedicate this review to Dr. Ulrich Kogelschatz (1937-2016). Our community will miss his input and support.

28 The author expresses his gratefulness to all colleagues and cooperation partners.

In particular Hans Höft, Marc Bogaczyk, Tomas Hoder, Manfred Kettlitz, Wolfgang Reich, Michael Schmidt, Andrei Pipa, Torsten Gerling, Detlef Loffhagen, Markus

32 M. Becker, Jan Schöfer, Lars Stollenwerk, Franz X. Bronold, Robert Tschiersch,

34 Sebastian Nemschokmichal, Robert Wild, Peter J. Bruggeman, Hans-Erich Wagner,

35 Kirill V. Kozlov , JUrgen F. Kolb, JUrgen Meichsner and Klaus-Dieter Weltmann are

36 acknowledged for fruitful cooperation and support.

38 Part of the work shown here was supported by Deutsche Forschungsgemeinschaft

39 (DFG, TRR 24 "Fundamentals of Complex Plasmas"), the Federal German Ministry of Education and Research (BMBF, grants 03FO1072 and 13N11188) and the Ministry of Education, Research and Culture of the State of Mecklenburg-Vorpommern (grant AU

43 07139).

46 References

48 [1] Siemens W. 1857 Poggendorfs Ann. Phys. Chem. 102 66-122

49 [2] Becker KH, Kogelschatz U, Schoenbach KH and Barker RJ (editors) 2005 Non-equilibrium air plasmas at atmospheric pressure (Bristol: Institute of Physics Publishing)

[3] Samoilovich V G, Gibalov V I and Kozlov K V 1997 Physical chemistry of the barrier discharge 2nd edition (Diisseldorf: Verlag DVS)

[4] Fridman A 2008 Plasma Chemistry (Cambridge University Press)

55 [5] Kogelschatz U 2003 Plasma Chem. Plasma Proc. 23 1 1-46

56 [6] Fridman A, Chirokov A and Gutsol A 2005 J. Phys. D: Appl. Phys. 38, R1-R24

57 [7] Parvulescu VI, Margureanu M and Lukes P (editors) 2012 Plasma chemistry and catalysis in gases

58 and liquids (Weinheim: Wiley-VCH)

59 [8] Chu PK and Lu XP (editors) 2014 Low temperature plasma technology, Methods and applications (Boca Raton: CRC Press, Taylor & Francis Group)

10 [11

13 [12

14 [13

15 [14

16 [15

17 [16

19 [17

20 [18

22 [19

23 [20

24 [21

26 [22

28 [23

30 [24

32 [25

33 [26

34 [27

37 [28

38 [29

39 [30

40 [31

42 [32

43 [33

44 [34

45 [35

46 [36

48 [37

49 [38

50 [39

52 [40

53 [41

54 [42

55 [43

56 [44

57 [45

58 [46

59 [47

Kawai Y, Ikegami H, Sato N, Matsuda A, Uchino K, Kuzuya M and Mizuno A 2010 Industrial plasma technology, Applications from environmental to energy tec (Weinheim: Wiley-VCH) Meichsner J, Schmidt M, Schneider R and Wagner H-E (Editors) 2012 Nonthe

Chemistry and Physics (Boca Raton: CRC Press Taylor & Francis Group) Thomas M and Mittal KL (Editors) 2013 Atmospheric pressure plasma treatm

relevance to adhesion (London: Wiley) Raizer Y P and Mokrov M S 2013 Phys. Plasmas 20, 101604 Kogelschatz U 2010 J. Phys: Conf. Ser. 25 7 012015 Pietsch GJ and Gibalov VI 2012 Plasma Sources Sci. Technol. 21 2 024 Morfill G E, Shimizu T, Steffes B and Schmidt H-U 2009 New J. Phys. 11 115019 Shimizu S, Barczyk S, Rettberg P, Shimizu T, Klaempfl T, Zimmermann J I, Hoeschen T, Linsmeier

C, Weber P, Morfill G E and Thomas H M 2014 Planetary and Space Scie. 90 60-71 Hähnel M, Brüser V and Kersten H 2007 Plasma Process. Polym. 4 629-637 Hahnel M, von Woedtke T and Weltmann K-D 2010 Plasma Process. Polym. 7 244-249 Moreau E, Sosa R and Artana G 2008 J. Phys. D: Appl. Phys. 41 11 115204 Corke TC, Enloe CL and Wilkinson SP 2010 Annu. Rev. Fluid Mech. 42 505-529 Masuda S, Akutsu K, Kuroda M, Awatsu Y and Shibuya Y 1988 IEEE Trans. Ind. Appl. 24 223-231

Cernak M, Kovacik D, Rahel J, Zahoranova A, Kubincova J, Toth A and Cernakova 2011 Plasma,

Phys. Control. Fusion 53 124031 Cernak M, Cernakova L , Hudec I, Kovacik D and Zahoranova A 2009 Eur. Phys. J. Appl. Phys. 47 2 22806

Li Y F, Shimizu T, Zimmermann J L and Morfill GE 2012 Plasma Process. Polym. 9 9 585-589 Sato S, Hensel K, Hayashi H and Mizuno A 2009 Journal of Electrostatics 67 (2-3) 77-83 Kunhardt EE 2000 IEEE Trans. Plasma Sci. 28 1 189-200

Eden J G, Park S-J, Cho J H, Kim H M, Houlahan Jr. T H, Li B, Kim E S, Kim T L, Lee S K, Kim K S, Yoon J K, Sung S H, Sun P, Herring C M and Wagner C J 2013 IEEE Trans. Plasma Sci. 41 661

Becker K H, Schoenbach K H and Eden J G 2006 J. Phys. D: Appl. Phys. 39 R55-R70 Tachibana K 2006 Trans. Electrical and Electronic Eng. 1 145-155 Sakai O and Tachibana K 2012 Plasma Sources Sci. Technol. 21 013001 Laroussi M, Alexeff I, Richardson J P and Dyer F F 2002 IEEE Trans. Plasma Sci. 30 158 Wang X, Li C, Lu M and Pu Y 2003 Plasma Sources Sci. Technol. 12 358 Radehaus Ch, Dirksmeyer T, Willebrand H and Purwins H G 1987 Phys. Lett. A 125 92-4 Purwins H-G, Amiranshvili Sh and Boedeker H U 2010 Adv. Phys. 59 485-701 Wild R, Schumann T and Stollenwerk L 2014 Plasma Sources Sci. Technol. 23 054004 Portsel L M, Astrov Yu A, Reimann I, Ammelt E and Purwins H-G 1999 J. Appl. Phys. 85 3960-3965

Itoh H, Teranishi K and Suzuki S 2006 Plasma Sources Sci. Technol. 15 S51-S61 Teschke M and Engemann J 2009 Contrib. Plasma Phys. 49 614-623 Johnson M J and Go D B 2015 J. Appl. Phys. 118 24330

Teranishi K, Shimomura N, Suzuki S and Itoh H 2009 Plasma Sources Sci. Technol. 18 045011

Johnson M J, Linczer J and Go D B 2014 Plas. Sources Sci. Technol. 23, 065018

Laroussi M and Akan A 2007 Plas. Proc. Polym. 4 9 777-788

Lu X, Laroussi M and Puech V 2012 Plasma Sources Sci. Technol. 21 034005

Winter J, Brandenburg R and Weltmann K-D 2015 Plasma Sources Sci. Technol. 24 064001

Schoenbach K H and Becker K 2016 Eur. Phys. J. D 70 29

Panousis E, Clement F, Loiseau J-F, Spyrou N, Held B, Thomachot M and Marlin L 2006 Plasma, Sources Sci. Technol. 15 828-839

10 [52

12 [53

14 [54

15 [55

16 [56

18 [57

19 [58

22 [59

23 [60

25 [61

26 [62

27 [63

28 [64

29 [65

30 [66

34 [67

35 [68

36 [69

37 [70

38 [71

40 [72

42 [73

44 [74

46 [75

47 [76

48 [77

49 [78

50 [79

52 [80

53 [81

54 [82

55 [83

56 [84

58 [85

59 [86

60 [87

Progress on DBD Sources and Filaments

Peeters F J J and van de Sanden M C M 2015 Plasma Sources Sci. Technol. 24 015016 Lu X, Naidis G V, Laroussi M, Reuter S, Graves D B and Ostrikov K 2016 Physics R 1-84

Laroussi M, Lu X, Kolobov V and Arslanbekov R 2004 J. Appl. Phys. 96 5 3028-Walsh J L, Liu D X, Iza F, Rong M Z and Kong M G J. Phys. D: Appl. Phys. 4 Shao T, Zhang C, Yu Z, Niu Z, Jiang H, Xu J, Li W, Yan P and Zhou Y 2012 V Shao T, Jiang H, Zhang C, Yan P, Lomaev M I and Tarasenko V F 2013 Euro 45002

Raizer Yu P 1991 Gas Discharge Physics (Springer)

Loeb L B 1939 Fundamental Processes of Electrical Discharge in Gases (John W

urces Sci. Technol. 25

J. Phys. B 38 159-165 ov A and Trushkin N 2011

& Sons)

Dedrick J, Boswell R W, Audier P, Rabat H, Hong D and Charles C 2011 J. Phys. D: Appl. Phys. 44 205202

Jidenko N, Petit M and Borra J P 2006 J. Phys. D: Appl. Phys. 39 281-293 Peeters F J J , Rumphorst R F and van de Sanden M C M 201 03LT03

Drimal J, Kozlov K V, Gibalov V I and Samoylovich V G Akishev Y, Aponin G, Balakirev A, Grushin M, Karalnik V, Pe

Plasma Sources Sci. Technol. 20 2 024005 Peeters F J J , Yang R and van de Sanden M C M 2015 Plasma Sources Sci. Technol. 24 045006 Eliasson B, Hirth M and Kogelschatz U 1987 J. Phys. D: Appl. Phys. 20 1421-1437 Braun D, Gibalov V I and Pietsch G J 1992 Plasma Sources Sci. Technol. 1 166 74-89 Gibalov V I and Pietsch G J 2000 J. Phys. D: Appl. Phys. 33 2618-2636 Heise M, Lierfeld T, Franken O and Neff W 2004 Plasma Sources Sci. Technol. 13 351-358 Peeters F J J 2015 The Electrical Dynamics of Dielectric Barrier Discharges Dissertation, Eindhoven University of Technology (available at: Kogelschatz U 2012 J. Opt. Technol. 79 8 484-493 Bartnikas R 1969 J. Appl. Phys. 40 1974-1976

Okazaki S, Kogoma M, Uehara M and Kimura Y 1993 J. Phys. D: Appl. Phys. 26 5 889-892 Roth J R, Rahel J, Dai X and Sherman D M 2005 J. Phys. D: Appl. Phys. 38 555-567 Starostin S A, Premkumar P A, Creatore H, de Vries H, Paffen R M J and van de Sanden M C

M 2010 Appl. Phys. Lett. 96 061502 Müller S and Zahn RJ 1996 Contrib. Plasma Phys. 36 6 697-709

Massines F, Rabehi A, Decomps Ph, Ben Gadri R, Segur P and Mayox C 1998 J. Appl. Phys. 28 2950

Miralai S F, Monette E, Bartnikas R, Czeresmuszkin G, Latreche M and Wertheimer M R 2005

Plasmas Polym. 5 63-77 Nersisyan G and Graham W G 2004 Plasma Sources Sci. Technol. 13 582-587 Boeuf JP 2003 J. Phys. D: Appl. Phys. 36 R53-R79

Naude N, Cambronne J P, Gherardi N and Massines F 2005 J. Phys. D: Appl. Phys. 38 4 530-538

Aldea E, Peeters P, de Vries H and van de Sanden MCM 2005 Surface Coating Technol. 200 46-50

Malik D A, Orlov K E, Miroshnikov I V and Smirnov A S 2008 J. Appl. Phys. 103 033303

Massines F, Gherardi N, Naude N and Segur P 2009 Eur. Phys. J. Appl. Phys. 47 22805

Rahel J and Sherman D M 2005 J. Phys. D: Appl. Phys. 38 547-554

Guikema J, Miller N, Niehof J, Klein M and Walhout M 2000 Phys. Rev. Lett. 85 3817

Trunec D, Brablec A and Buchta J 2001 J. Phys. D: Appl. Phys. 34 1697-1699

Starostin S A, Premkumar A, Creatore M, van Veldhuizen E M , de Vries H, Paffen R M J and

van de Sanden MCM 2009 Plasma Sources Sci. Technol. 18 045021 Rahel J, Sira M, Stahel P and Trunec D 2007 Contrib. Plasma Phys. 47 1-2, 34-39 Massines F, Gherardi N, Naude N and Segur P 2005 Plasma Phys. Control. Fusion 47 B577-B588 Radu I, Bartnikas R and Wertheimer M R 2004 J. Phys. D: Appl. Phys. 37 449-462

2 3 Pro

5 [88]

7 [89]

9 [90]

10 L J [91]

12 [92]

13 [93]

15 [94]

16 [95]

18 [96]

19 [97]

20 L J

22 [98

23 [99

24 [100

26 [101

28 [102

29 [103

30 [104

32 [105

33 [106

34 [107

36 [108

37 [109

38 [110

39 [111

42 [112

43 [113

44 [114

45 [115

46 [116

48 [117

49 [118

50 [119

52 [120

53 [121

54 [122

55 [123

57 [124

58 [125

59 [126

60 [127

Progress on DBD Sources and Filaments

Mangolini L, Orlov K, Kortshagen U, Heberlein J and Kogelschatz U 2002 Appl. Phys. Lett. 80 1722-4

Navratil Z, Brandenburg R, Trunec D, Brablec A, Stahel P, Wagner H-E, Kopecky Z 2006 Plasma

Sources Sci. Technol. 15 8 Brenning N, Axnas I, Nilsson O and Eninger JE 1997 IEEE Trans. Plasma Sci. 25 1 83-88 Tochikubo F, Chiba T and Watanabe T 1999 Jpn. J. Appl. Phys. 38 5244-5250 Adler F and Muller 2000 J. Phys. D: Appl. Phys. 33 1705-1715

Tachibana K, Kawai S, Asai H, Kikuchi N and Sakamoto S 2005 J. Phys. D: Appl. Phys. 38 1739-1749

Obradovic BM, Ivkovic SS and Kuraica MM 2008 Appl. Phys. Lett. 92 191501 Ivkovic SS, Obradovic BM, Cvetanovic N, Kuraica MM and J Puric 2009 J. Phys. D: Appl. Phys. 42 225206

Ivkovic SS, Obradovic BM and Kuraica MM 2012 J. Phys. D: Appl. Phys. 45 275204 Ivkovic SS, Sretenovic G B, Obradovic BM, Cvetanovic N and Kuraica MM 2014 J. Phys. D: Appl. Phys. 47 055204

Tschiersch R, Bogaczyk M and Wagner H-E 2014 J. Phys. D: Appl. Phys. 47 365204 Gherardi N, Gouda G, Gat E, Ricard A, Massines F 2000 Plasma Sources Sci. Technol. 9 340-346 Kozlov KV, Brandenburg R, Wagner H-E, Morozov AM, Michel P 2005 J. Phys. D: Appl. Phys. 38 518-529

Brandenburg R, Maiorov V A, Golubovskii Yu B, Wagner H-E, Behnke J and Behnke JF 2005

J. Phys. D: Appl. Phys. 38 2187 Moravej M, Yang Y, Hicks RF, Penelon J and Babayan SE 2006 J. Appl. Phys. 99 093305 Bazinette R, Subileau R, Paillol J and Massines F 2014 Plasma Sources Sci. Technol. 23 035008 Bazinette R, Paillol J and Massines F 2015 Plasma Sources Sci. Technol. 24 055021 Meyer C, Franzke J and Gurevich EL 2012 J. Phys. D: Appl. Phys. 45 35 355205 Balcon N, Anesland A and Boswell R 2007 Plasma Sources Sci. Technol. 16 217-225 Lieberman M A and Lichtenberg A J 1997 Principles of Plasma Discharges and Material

Processing (New York: Wiley) Iza P, Lee JK and Kong M G 2007 Phys. Rev. Lett. 99 075004 Liu D W, Iza F and Kong M G 2009 Pias. Proc. Polym. 6 446-450

Farouk T, Farouk B, Gutsol A and Fridman A 2008 Plasma Sources Sci. Technol. 17 035015 Becker K, Koutsospyros A, Yin S-M, Christodoulatos C, Abramzon N, Joaquin J C and Brelles-

Mari G 2005 Plasma Phys. Control. Fusion 47 7 B513-B523 Kim H H 2004 Plas. Proc. Polym. 1 91-110

Vezzu G, Lopez J L, Freilich and A. Becker K.H. 2009 IEEE Trans. Pias. Sci. 37 6 890-896 Holzer F, Roland U and Kopinke F-D 2002 Appl. Cat. B: Environ. 38 163-181 Roland U, Holzer F and Kopinke F-D 2002 Catalysis Today 73 315-323

Kim HH, Teramoto Y, Sano T, Negishi N and Ogata A 2015 Appl. Cat. B: Environ. 166-167 9-17

Van Durme J, Dewulf J, Leys Ch and Van Langenhove H 2008 Appl. Cat. B: Environ. 78 324-333.

Kraus M, Eliasson B, Kogelschatz U and Wokaun A 2001 Phys. Chem. Chem. Phys. 3 294-300

Rafflenbeul R 1998 Mull und Abfall 1, 38-44 (in german)

Müller S and Zahn R J 2007 Contrib. Plas. Phys 47 520-529

Mizuno A 2007 Plasma Phys. Control. Fusion 49 A1-A15

Chang J S 2008 Pias. Sources Sci. Technol. 17 045004

Brandenburg R, Kovacevic, Schmidt M, Basner R, Kettlitz M, Sretenovic G B, Obradovic B M, Kuraica M M and Weltmann K-D 2014 Contrib. Plasma Phys. 54 2 202-214 Müller, Zahn R-J and Grundmann J 2007 Plasma Process. Polym. 4 S1004-S1008 Schmidt M, Basner R and Brandenburg R 2013 Plasma Chem. Plasma Process.33 1, 323-335 Park S-J, Kim K S and Eden J G 2005 Appl. Phys. Lett. 86 221501-1 to 221501-3 Tachibana K, Kishimoto Y, Kawai S, Sakaguchi T and Sakai O 2005 Plasma Phys. Control.

10 11 12

20 21 22

Progress on DBD Sources and Filaments


Fusion 47 A167-A177 Blajan M and Shimizu K 2012 Appl. Phys. Lett. 101 104101 Kogelschatz U 2007 Contrib. Plasma Phys. 47 80-88

Dahle S, Hirschberg J, Viül W and Maus-Friedrichs W 2015 Plas. Sources Sci. Techn Tanaka Y 1955 J. Opt. Soc. Am. 45 710

Kogelschatz U, Esrom H, Zhang J-Y, Boyd I W 2000 Appl. Surf. Sa. 168 29-Pochner K, Neff W and Leber R 1995 Surf. Coat. Technol. 74/75 394-398 Starostin S, Aldea E, de Vries H, Creatore M and van de Sanden MCM 2007 Plas. 4 S440-S444

Heise M, Neff W, Franken O, Muranyi P and Wunderlich 2004 Plas. Polym. 9 1, 23-33 Lucas N, Franke R, Hinze A, Klages CP, Frank R and Buüttgenbach S 2008 Plas. Proc. Polym. 6 S370-S374

Buüttgenbach S, Lucas N and Sichler P 2009 Contrib. Plasma Phys. 49 624-630 Hinze A, Marchesseault A, Büttgenbach S, Thomas M and Klages CP 2013 In: Atmospheric pressure plasma treatment of polymers - relevance to adhesion, Thomas M, Mittal KL (editors) (London: Wiley)

Schwabedissen A, Lacinski P, Chen X and Engemann J 2007 Contrib. Plasma Phys. 47 551-558

Eto H, Ono Y, Ogino A and Nagatsu M 2008 Appl. Phys. Lett. 93 221502

Weltmann K-D, Kindel E, von Woedtke T, Hühnel M, Stieber M and Brandenburg R 2010 Pure

Appl. Chem. 82 6 1223-1237 Weltmann K-D, Fricke K, Stieber M, Brandenburg R, von Woedtke T and Schnabel U 2012 IEEE

Trans. Plasma Sci. 40 11 2963-2969 Brückner S, Rüsner S, Gerhard C, Wieneke S and Viol W 2011 Mater. Test. 53 639 Hoffmeister J, Bruückner S, Gerhard C, Wieneke S and Viüol W 2014 Plasma Sources Sci. Technol. 23 064008

Kelly-Wintenberg K, Hodge A, Montie T C, Deleanu L, Sherman D, Roth J R, Tsai P and Wadsworth L 1999 Journal of Vacuum Science & Technology A: Vacuum, Surfaces, and Films 17 4, 1539-1544

Ehlbeck J, Schnabel U, Polak M, Winter J, von Woedtke T, Brandenburg R, von dem Hagen T and Weltmann K-D 2011 J. Phys. D: Appl. Phys. 44 1 013002 Matthes R, Bender C, Schluüter R, Koban I, Bussiahn R, Reuter S, Lademann J, Weltmann K-D and Kramer A 2013 PLoS ONE 8 7, e70462 Schwartz G 2011 Plasma-based Cleansing Techniques for Biopharmaceutical Research: An Introduction to and Explanation of the Technology Conducted at (accessed July 2016)

Leipold F, Kusano Y, Hansen F and Jacobsen T 2010 Food Control 21 1194-1198

Moiseev T, Misra N N, Patil S, Cullen P J, Bourke P, Keener KM and Mosnier J P 2014 Plas.

Sources Sci. Technol. 23 065033 Park G Y, Park S J, Choi M Y, Koo I G, Byun J H, Hong J W, Sim J Y, Collins G J and Lee J K 2012 Plas. Sources Sci. Technol. 21 043001 von Woedtke T, Reuter S, Masur K and Weltmann K-D 2013 Phys. Rep. 530 7 291-320 Fridman G, Peddinghaus M, Ayan H, Fridman A, Balasubramanian M, Gutsol A, Brooks A and

Friedman G 2006 Plasma Chem. Plasma Process. 26 425-442 Fridman G, Friedman G, Gutsol A, Shekhter A B, Vasilets V N and Fridman A 2008 Plasma

Process. Polym. 5 6 503-533 Kuchenbecker M, Bibinov N, Kaemlimg A, Wandke D, Awakowicz P and Viüol W 2009 J. Phys.

D: Appl. Phys. 42 4 045212 Liu C, Dobrynin D and Fridman A 2014 J. Phys. D: Appl. Phys. 47 252003 Tiede R, Hirschberg J, Viül W and Emmert S 2015 Plasma Process. Polym. 13 775-787 Emmert S, Brehmer F, Hüanszle H, Helmke A, Mertens N, Ahmed R, Simon D, Wandke D, Maus-Friedrichs W, Düschlein G, Schün M P and Viül W 2013 Clinical Pias. Med. 1 24-29

2 3 Pro

5 [159

7 [160

9 [161

13 [163

15 [164

16 [165

17 [166

18 [167

20 [168

23 [169

25 [170

26 [171

27 [172

29 [173

30 [174

32 [175

33 [176

35 [177

37 [178

38 [179

40 [180

43 [181

44 [182

45 [183

47 [184

49 [185

50 [186

54 [187

55 [188

56 [189

57 [190

58 [191

59 [192

60 [193

Progress on DBD Sources and Filaments

Brehmer F, Hanszle H, Daschlein G, Ahmed R, Pfeiffer S, Görlitz A, Simon D, Schon M P,

Wandke D and Emmert S 2015 J. Eur. Acad. Dermatol. Venereol. 29 148 - 155 Boekema B K H L, Vlig M, Guijt D, Hijnen K, Hofmann S, Smits P, Sobota A, van Veldhuizen E M, Bruggeman P and Middelkoop E 2016 J. Phys. D: Appl. Phys. 49 044001 Shimizu T, Zimmermann J L and Morfill G E 2011 New J. Phys. 13 023026 Zimmermann J L, Shimizu T, Boxhammer V and Morfill G E 2012 Plasma Process. Polym. 9 9 792-798

Weltmann K-D, Polak M, Masur K, von Woedtke T, Winter J and Reuter S 2012 Contrib. Plasma Phys. 52 7 644-654 Sakai O and Tachibana K 2007 J. Phys.: Conf. Ser. 86 012015 Eto H, Ono Y, Ogino A and Nagatsu M 2008 Plasma Process. Polym. 5 269 Pointu A-M, Ricard A, Odic E and Ganciu M 2008 Plasma Process. Polym. 5 559 Robert E, Barbosa E, Dozias S, Vandamme M, Cachoncinlle C, Viladrosa R and Pouvesle JM 2009 Plasma Proc. Polym. 6, 12, 795-802 Polak M, Winter J, Schnabel U, Ehlbeck J and Weltmann K-D 2012 Plasma Proc. Polym. 9 67-76

Oehmigen K, Hahnel M, Brandenburg R, Wilke C, Weltmann K-D and von Woedtke T 2010

Plasma Proc. Polym. 7 3-4 250-257 Rumbach P, Bartels D M, Sankaran R M and Go D B 2015 J. Phys. D: Appl. Phys. 48 424001 Malik M A, Ghaffar A and Malik S A 2001 Plasma Sources Sci. Technol. 10 82-91 Pavlinak D, Galmiz O, Zemanek M, Brablec A, Cech, J and Cernak M 2014 Appl. Phys. Lett. 105 154102

Liu W and Li Ch 2014 Pias. Sci. Technol. 16 26-31

Baroch P, Saito N and Takai O 2008 J. Phys. D: Appl. Phys. 41 085207

Kuraica M M, Obradovic B M, Manojlovic D, Ostojic D R and Puric J 2004 Vacuum 73 705-708 Manojlovic D, Ostojic D R, Obradovic B M, Kuraica M M, Krsmanovic V D and Puric J 2007 Desalination 213 116

Dojcinovic B P, Roglic G M, Obradovic B M, Kuraica M M, Kostic M M, Nesic J and Manojlovic D 2011 J. Hazard. Mat. 192 763-771 Obradovic B M, Sretenovic G B and Kuraica M M 2011 J. Hazard. Mat. 185 1280-1286 Winter T, Winter J, Polak M, Kusch K, Mäder U, Sietmann R, Ehlbeck J, van Hijum S, Weltmann

K-D, Hecker M and Kusch H 2011 Proteomics 11 3518-3530 Winter T, Bernhardt J, Winter J, Mäder U, Schläter R, Weltmann K-D, Hecker M and Kusch H

2013 Proteomics 13 2608-2621 Meyer C, M'uller S, Gurevich E L and Franzke J 2011 Analyst 136 2427-2440 Krähling T, Mäller S, Meyer C, Stark A-K and Franzke J 2011 J. Anal. At. Spectrom. 26 1974 Bonnin X, Piquet H, Naude, Bouzidi M C and Gherardi N 2013 Eur. Phys. J. Appl. Phys. 64 1 10901

Bonnin X, Brandelero J, Piquet H and Meynard 2014 IEEE Trans. Power Electron. 29 8 42614269

Jakubowski T, Holub and Kalisiak S 2013 Eur. Phys. J. Appl. Phys. 61 24304 Meißer S 2013 Resonant Behaviour of Pulse Generators for the Efficient Drive of Optical Radiation Sources Based on Dielectric Barrier Discharges Dissertation, Karlsruhe Institute for Technologie (KIT) ( Wagner H-E, Kozlov KV and Brandenburg R 2004 J. Adv. Oxid. Technol 7, 11-19 Dilecce G and De Benedictis S 2011 Plasma Phys. Control. Fusion 53 124006 Dilecce G 2014 Plasma Sources Sci. Technol. 23 015011 Bruggeman P J and Brandenburg R 2013 J. Phys. D: Appl. Phys. 46 464001 Simek M 2014 J. Phys. D: Appl. Phys. 47 463001 Ono R 2016 J. Phys. D: Appl. Phys. 49 083001

Dilecce G, Ambrico P F and De Benedictis S 2007 Plasma Sources Sci. Technol. 16 511-522

10 11 12

20 21 22

Progress on DBD Sources and Filaments

194] Es-Sebbar E, Gherardi N and Massines F 2013 J. Phys. D: Appl. Phys. 46 015202

195] Nemschokmichal S and Meichsner J 2013 Plasma Sources Sci. Technol. 22, 015005

196] Nemschokmichal S and Meichsner J 2013 Plasma Sources Sci. Technol. 22, 015006

197] Dilecce G, Ambrico P F, Simek M and De Benedictis S 2012 J. Phys. D: Appl. Phys. 45

198] Abd-Allah Z, Sawtell D A G , McKay K, West G T, Kelly P J and Bradley J W 2015 J. Phys.

D: Appl. Phys. 48 085202

199] Abd-Allah Z, Sawtell DAG, West G T , Kelly P J and Bradley J W 2016 Plasma Process.

Polym. 13 649-653

200] Kogelschatz U, Eliasson B and Egli W 1997 J. de Physique IV 7, C4 47-66

201] Kozlov K V, Wagner H-E, Brandenburg R and Michel P 2001 J. Phys. D: Appl. Phys. 34 3164

202] Merbahi N, Sewraj N, Marchal F, Salamero Y and Millet P 2004 J. Phys. D: Appl. Phys. 37 1664-78

203] Brandenburg R, Wagner H-E, Morozov A M and Kozlov K V 2005 J. Phys. D: Appl. Phys. 38

204] Hoder T, Brandenburg R, Basner R, Weltmann K-D, Kozlov KV and Wagner H-E 2010 J. Phys. D: Appl. Phys. 43 124009

205] Kloc P, Wagner H-E, Trunec D, Navratil Z and Fedoseev G 2010 J. Phys. D: Appl. Phys. 43 345205

206] Sewraj N, Merbahi N, Gardou J P, Rodriguez Akerreta P and Marchal F 2011 J. Phys. D: Appl.

Phys. 44 145201

207] Lukas C, Spaan M, Schulz-von der Gathen V, Thomson M, Wegst R, Dobele HF and Neiger M 2001 Plasma Sources Sci. Technol. 10 445

208] Rajasekaran P, Mertmann P, Bibinov N, Wandke D, Vioül W and Awakowicz P 2009 J. Phys. D: Appl. Phys. 42 225201

209] Grosch H, Hoder T, Weltmann K-D and Brandenburg R 2011 Eur. Phys. J. D 60 3 547-553

210] Hoder T, Sira M, Kozlov K V and Wagner H-E 2008 J. Phys. D: Appl. Phys. 41 035212

211] Hoder T, Sira M, Kozlov KV and Wagner H-E 2009 J. Phys. D: Appl. Phys. 42 4 049802

212] Simek M, Ambrico PF and Prukner V 2011 Plasma Sources Sci. Technol. 20 025010

213] Simek M, Prukner V and Schmidt J 2011 Plasma Sources Sci. Technol. 20 025009

214] Falkenstein Z and Coogan JJ 1997 J. Phys. D: Appl. Phys. 30 817-825

215] Brandenburg R, Bogaczyk M, Hoft, Nemschokmichal S, Tschiersch R, Kettlitz M, Stollenwerk L, Hoder T, Wild R, Weltmann K-D, Meichsner J and Wagner H-E 2013 J. Phys. D: Appl. Phys. 46 464015

216] Kettlitz M, Hoft H, Hoder T, Reuter S, Weltmann K-D and Brandenburg R 2012 J. Phys. D: Appl. Phys. 45 245201

217] Hüft H 2015 Characteristics of pulsed operated dielectric barrier discharges in molecular gas mixtures Dissertation, University of Greifswald

218] Hoder T, Cernak M, Paillol J, Loffhagen D and Brandenburg R 2012 Phys. Rev. E 86 055401

219] Hoder T, Simek M, Bonaventura Z, Prukner V and Gordillo-Vazquez J F 2016 Plasma Sources Sci. Technol. 25 045021

220] Starikovskaia SM, Allegraud K, Guaitella O and Rousseau A 2010 J. Phys. D: Appl. Phys. 43 124007

221] Stepanyan S A, Starikovskiy A Yu, Popov N A and Starikovskaia S M 2014 Plasma Sources Sci. Technol. 23 045003

222] Stepanyan S A, Soloviev V R and Starikovskaia S M 2014 J. Phys. D: Appl. Phys. 47 485201

223] Kawasaki T, Terashima T, Suzuki S and Takada T 1994 J. Appl. Phys. 76 3274

224] Zhu Y C, Takada T and Tu DM 1995 J. Phys. D: Appl. Phys 28 1468-77

225] Kumada A, Chiba M and Hidaka K 1998 J. Phys. D: Appl. Phys. 69 3059

226] Sugimoto K, Takahashi H, Shimomura O and Sakurai T 2003 J. Phys. D: Appl. Phys. 36 23


[227] Jeong DC, Bae HS and Whang KW 2005 J. Appl. Phys. 97 1 013304

10 11 12

20 21 22

Progress on DBD Sources and Filaments

[228 [229 [2ЗО [2З1 [2З2 [2ЗЗ [234

[2З6 [2З7 [2З8

[242 [24З [244 [245 [246

[248 [249

[251 [252 [25З [254 [255 [256 [257

[259 [26О

[26I [262 [26З [264 [265 [266

2 З10-З20 J and Wagner

Stollenwerk L, Amiranashvili S, Boeuf JP and Purwins HG 2006 Phys. Rev. Lett. 96 25 Stollenwerk L, Laven J G and Purwins HG 2007 Phys. Rev. Lett. 98 255001 Stollenwerk L 2009 New Journal of Physics 11 10З0З4 Stollenwerk L 2010 Plasma Phys. Controlled Fusion 52 12 124017 Bogaczyk M, Wild R, Stollenwerk L and Wagner H-E 2012 J. Phys. D: Appl. P, Wild R and Stollenwerk 2014 New J. Phys. 16 11З040

Bogaczyk M, Wild R, Stollenwerk L and Wagner H-E 2012 J. Adv. Oxid.. Techn Bogaczyk M, Nemschokmichal S, Wild R, Stollenwerk L, Brandenburg R H-E 2012 Contrib. Plasma Phys. 52 10 847-855 Stollenwerk L and Stroth U 2011 Contrib. Plasma Phys. 51 61-67 Wild R, Benduhn J and Stollenwerk 2014 J. Phys. D: Appl. Phys. Ishihare, Candler G, Laux C O, Napartovich A P, Pitchford L C, Boeuf J P and Verboncoeur J 2005 In: Non-equilibrium air plasmas at atmospheric pressure, Becker KH, Kogelschatz U, Schoenbach KH and Barker RJ (editors) (Bristol: Institute of Physics Publishing) Georghiou G E, Papadakis A P, Morrow R and Metaxas A C 2005 J. Phys. D: Appl. Phys. 38 R303-R328

Bogaerts A, De Bleecker K, Georgieva V, Kolev I, Madani M and Neyts E 2006 Plasma Process. Polym. 3 110-119

Bogaerts A, De Bie C, Eckert M, Georgieva V, Martens T, Neyts E and Tinck S 2010 Pure Appl.

Chem. 82 1283-1299 Kulikovsky A A 1995 J. Phys. D: Appl. Phys. 28 2483-2493 Babaeva N Yu and Naidis G V 1996 J. Phys. D: Appl. Phys. 29 2423 Morrow R and Lowke J J 1997 J. Phys. D: Appl. Phys. 30 614-27 Naidis G V 2005 J. Phys. D: Appl. Phys. 38 З889-З89З

Ebert U, Montijn C, Briels TMP, Hundsdorfer W, Meulenbroek B, Rocco A and van Veldhuizen

E M 2006 Plasma Sources Sci. Technol. 15 S118-S129 Bourdon A, Pasko V P, Liu N Y, Celestin S, Segur P and Marode E 2007 Plasma Sources Sci.

Technol. 16 656-678 Naidis G V 2009 Phys. Rev. E 79 5 057401

Chanrion O, Bonaventura Z, Cinar D, Bourdon A and Neubert T 2014 Environ. Res. Lett. 9 5, 5500З

Boeuf J P, Lagmich Y, Unfer T, Callegari T and Pitchford L C 2007 J. Phys. D: Appl. Phys. 40 652-662

Eliasson B, Egli W and Kogelschatz U 1994 Pure Appl. Chem. 66 6 1275-1286 Yurgelenas Y V and Wagner H-E 2006 J. Phys. D: Appl. Phys. 39 4031-4043 Xu X P and Kushner M 1998 J. Appl. Phys. 83 7522-7532 Van Laer K and Bogaerts A 2016 Plasma Sources Sci. Technol. 25 015002 Hagelaar G J M and Pitchford L C 2005 Plasma Sources Sci. Technol. 14 722-7ЗЗ Becker M M, Hoder T, Brandenburg R and Loffhagen D 2013 J. Phys. D: Appl. Phys. 46 355203 Hoft H, Kettlitz M, Becker M M, Hoder T, Loffhagen D, Brandenburg R and Weltmann K-D 2014 J. Phys. D: Appl. Phys. 47 465206 Ponduri S, Becker M M, Welzel S, van de Sanden M C M, Loffhagen D and Engeln R 2016 J. Appl. Phys 119 093301

Golubovskii YB, Maiorov VA, Behnke J and Behnke JF 2002 J. Phys. D: Appl. Phys. 35 751 Tian W and Kushner M 2014 J. Phys. D: Appl. Phys. 47 165201 Neyts E C and Bogaerts A 2014 J. Phys. D: Appl. Phys. 47 224010

Bronold F X, Fehske H, Heinisch R L and Marbach J 2012 Contrib. Plasma Phys. 52 856-863

Heinisch R L, Bronold F X and Fehske H 2010 Phys. Rev. B 81 155420

Heinisch R L, Bronold F X and Fehske H 2010 Phys. Rev. B 82 125408

Heinisch R L, Bronold F X and Fehske H 2011 Phys. Rev. B 83 195407

Heinisch R L, Bronold F X and Fehske H 2012 Phys. Rev. B 85 075323

2 3 Pro

5 [267

6 [268

7 [269

8 [270

9 [271

13 [273

15 [274

16 [275

17 [276

18 [277

19 [278

20 [279

22 [280

24 [281

25 [282

26 [283

27 [284

29 [285

32 [286

33 [287

34 [288

36 [289

37 [290

38 [291

39 [292

40 [293

43 [294

44 [295

45 [296

47 [297

48 [298

50 [299

52 [300

53 [301

55 [302

56 [303

58 [304

59 [305

60 [306

Progress on DBD Sources and Filaments

Marbach J, Bronold F X and Fehske H 2011 Phys. Rev. B 84, 085443 Marbach J, Bronold F X and Fehske H 2012 Eur. Phys. J. D 66 106 Bronold F X and Fehske H Phys. Rev. Lett. 115 225001 Manley T C 1943 Trans. Electrochem. Soc. 84 83-96 Samoilovic V G, Vendillo V P and Filippov Yu 1962 Russ. J. Physic. Chem. 36 517


: Appl. Phys.


Kogelschatz U 1998 Advanced Ozone Generation in Process Technologies for Water Trea

ed. by Stucki S (Earlier Brown Boveri Symposia, Springer US) 87-118 Hoft H, Kettlitz M, Hoder T, Weltmann K-D and Brandenburg R 2013 J. 46 095202

Lomaev M I 2001 Atmos.Oceanic Opt. 14 1005-1008

Pipa A, Koskulics J, Brandenburg R and Hoder T 2012 Rev. Sci. Instrum. Diez R, Salanne J P 2007 Eur. Phys. J. - Appl. Phys. 37 3 307-313 Liu S and Neiger M 2001 J. Phys. D: Appl. Phys. 34 1632-1638 Liu S and Neiger M 2003 J.Phys.D: Appl. Phys. 36 3144-3150

Pipa A, Hoder T, Koskulics J, Schmidt M and Brandenburg R 2012 Rev. Sci. Instrum. 83 075111 Archambault-Caron M, Gagnon H, Nisol B, Piyakis K and Wertheimer M R 2015 Plasma S'ources

Sci. Technol. 24 045004 Kozlov K V, Michel P and Wagner H-E 2000 Plasmas Polym. 5 129-49 Pipa A, Hoder T and Brandenburg R 2013 Contrib. Plasma Phys. 53 469-480 Holub M 2012 Int. J. Appl. Electromagnetics and Mechanics 39 81-87

Trampert K 2009 Ladungstransportmodell dielektrisch behinderter Entladungen Dissertation, University of Karlsruhe

Paravia M 2010 Effizienter Betrieb von Xenon-Excimer-Entladungen bei hoher Leistungsdichte

Dissertation, Karlsruhe Institute for Technologie (KIT) Reichen P, Sonnenfeld A and Rudolf von Rohr Ph 2010 J. Phys. D: Appl. Phys. 43 025207 Gobrecht H, Meinhardt O and Hein F 1964 Ber. Bunsengesellschaft 66 1 55-63 Enloe C L, McLaughlin T E, VanDyken R D, Kachner K D, Jumper E J and Corke T C 2004 AIAA 42 589-594

Schmidt M, Schiorlin M, Brandenburg R 2015 Open Chem. 3 1 477-483

Kriegseis J, Duchmann A, Tropea C and Grundmann S 2013 J. Appl. Phys. 114 053301

Holub M 2014 Arch. Elec. Eng. 63 2 263-272

Kozlov K V and Wagner H-E 2007 Contrib. Plasma Phys. 47 26-33

Paris P, Aints M, Valk F, Plank T, Haljaste H, Kozlov K V and Wagner HE 2005 J. Phys. D: Appl. Phys. 38 3894

Brandenburg R, Hoder T and Wagner H-E 2008 IEEE Trans. Plasma Sci. 36 1318

Hoder T, Wilke C, Loffhagen D and Brandenburg R 2011 IEEE Trans. Plasma Sci. 39 2158

Hoder T, Loffhagen D, Wilke C, Grosch H, Schaefer J, Weltmann KD and Brandenburg R 2011

Phys. Rev. E 84 4 046404 Yurgelenas Yu V and Leeva M A 2009 IEEE Trans. Plasma Sci. 37 809-15 Kettlitz M, Hoft H, Hoder T, Weltmann K-D and Brandenburg R 2013 Plasma Sources Sci. Technol. 22 025033

Hoder T, Höft H, Kettlitz M, Weltmann K-D and Brandenburg R 2012 Phys. Plas. 19 070701 Hoft H, Kettlitz M, Weltmann K-D and Brandenburg R 2014 J. Phys. D: Appl. Phys. 47 455202 Nijdam S, Takahashi E, Markosyan A H and Ebert U 2014 Plasma Sources Sci. Technol. 23 025008

Höft H, Kettlitz M and Becker M M 2016 Phys. Pias. 23 033504

Akishev Y, Aponin G, Balakirev A, Grushin M, Karalnik V, Petryakov A and Trushkin N 2011

Eur. Phys. J. D 61 421-429 Brandenburg R, Grosch H, Hoder T and Weltmann K-D 2011 Eur. Phys. J. Appl. Phys. 55 13813 Allegraud K, Guaitella O. and Rousseau A. 2007 J. Phys. D: Appl. Phys. 40 7698-7706 Starikovskii A Yu, Nikipelov A A, Nudnova M M and Roupassov D V 2009 Plasma Sources Sci.

10 11 12

20 21 22

Sources Sci.

Progress on DBD Sources and Filaments Technol. 18 034015

[307] Brandenburg R, Navratil Z, Jansky J, Stahel P, Trunec D and Wagner H-E 2009 J.

Appl. Phys. 42 085208

[308] Purwins H G, Bödeker H U and Amiranashvili S 2010 Adv. Phys. 59 485

[309] Callegari T, Bernecker B and Boeuf J P 2013 Plasma Sources Sci. Technol. 23

[310] Klein M, Miller N and Walhout M 2001 Phys. Rev. E 64 026402

[311] Chirokov A, Gutsol A, Fridman A, Sieber K D, Grace J M and Robinson

Sources Sci. Technol. 13 623-635

[312] Stauss S, Muneoka H, Ebato N, Oshima F, Pai DZ and Terashima K 2013 Plasm

Technol. 22 025021

[313] Bogaczyk M 2014 Akkumulation von Oberflächenladungen und Entladungsentwicklung in

Barrierenentladungen in Helium, Stickstoff und deren Gemischen Dissertation, University of Greifswald

[314] Dosoudilova L, Tschiersch R, Bogaczyk M, Navratil Z, Wagner H-E and Trunec D 2015 J. Phys.

D: Appl. Phys. 48 355204

[315] Niemeyer L 1995 IEEE Trans. Dielect. El. Ins. 4 510-528

[316] Pickard P S and Davis M V 1970 J. Appl. Phys. 41 2636

[317] Ambrico P F, Ambrico M, Colaianni A, Schiavulli L, Dilecce G and De Benedictis S 2010 J. Phys.

D: Appl. Phys. 43 325201

[318] Ambrico P F, Ambrico M, Schiavulli L, Ligonzo T and Augelli V 2009 Appl. Phys. Lett. 94


[319] Falkenstein Z 1997 J. Appl. Phys. 81 5975-5979

[320] Kogelschatz U private communication

[321] Kashiwagi Y and Itoh H 2006 J. Phys. D: Appl. Phys. 39 113

[322] Guaitella O, Thevenet F, Guillard C and Rousseau A 2006 J. Phys. D: Appl. Phys. 39 2964-2972

[323] Celestin S, Canes-Boussard G, Guaitella O, Bourdon A and Rousseau A 2008 J. Phys. D: Appl.

Phys. 41 205214

[324] Guaitella O, Marinov I and Rousseau A 2011 Appl. Phys. Lett. 98 071502

[325] Dong L, Yin Z, Li X, Chai Z and Wang L 2003 J. Electrostat. 57 3-4, 243-250

[326] Nijdam S, van de Wetering F M J H, Blanc R, van Veldhuizen E M and Ebert U 2010 J. Phys.

D: Appl. Phys. 43 145204

[327] Cech J, Hanusova J, St'ahel P and Cernak M 2015 Open Chemistry 13 528-540

[328] Hoder T, Synek P, Chorvat, Rahel J, Brandenburg R and Cernak M 2017 Plas. Phys. Contr.

Fus. submitted

[329] Rahel J, Szalay Z, Cech J and Moravek 2016 Eur. Phys. J. D 70 92

[330] Nemschokmichal S and Meichsner J 2015 J. Phys. D: Appl. Phys. 48 405203

[331] Bogaczyk M, Sretenovic G B and Wagner H-E Eur. Phys. J. D 67 212-222

[332] Maiorov VA and Golubovskii YB 2007 Plasma Sources Sci. Technol. 16 S67-S75

[333] Boeuf J P, Bernecker B, Callegari Th, Blanco S and Fournier R 2012 Appl. Phys. Lett. 100 244108

[334] Bernecker B, Callegari T and Boeuf J P 2011 J. Phys. D: Appl. Phys. 44 262002

[335] Tschiersch R, Nemschokmichal S and Meichsner J 2016 Plasma Sources Sci. Technol. 25 02500

hiersch R,