Advertisement

Towards universal plasma-enabled platform for the advanced nanofabrication: plasma physics level approach

  • O. Baranov
  • S. Xu
  • K. Ostrikov
  • B. B. Wang
  • U. Cvelbar
  • K. Bazaka
  • I. Levchenko
Review Paper

Abstract

Growing demand for efficient, high-resolution surface processing has led to the emergence of a rich variety of plasma-based technologies underpinned by an equally wide range of technological setups, each optimized for a specific task, e.g. highly-selective removal of surface layers, precision surface functionalization and nanoscale structuring, or deposition of thin films and nanostructures. However, with increasing device integration and miniaturization, flexible processing technologies capable of delivering complex treatments, such as the growth of complex hierarchical single- and multi-component nanostructures within a single processing environment are highly desired. Yet, such systems are difficult to achieve due to the necessity of the in-process float, and limiting technological capabilities of individual plasma sources and treatment setups. Using a novel flexible platform as an example, this review presents a careful analysis of the physical principles, capabilities and limitations of existing plasma technologies with an ultimate aim to define key principles for the development of prospective flexible platforms for complex plasma-enabled material synthesis and processing. Such a platform would have a significant potential to increase the effectiveness of plasma technology with respect to productivity, material and energy consumption, cost and turnaround time.

Keywords

Plasma technology Advanced materials Deposition Sputtering 

1 Introduction

Emerging from the advancements in one of the most exciting fields of physics—that of the fourth state of matter, i.e. the plasma—over the last century, plasma processing has evolved from a mere deposition of sputtered metal atoms, to a family of sophisticated, fundamental science-driven high-tech approaches for manipulation of complex material systems, nanostructures and metamaterials (Anders 2000; Levchenko et al. 2018b). Growing demand for high-resolution processing spurred the development of a wide variety of structurally- and functionally-diverse plasma-enabled techniques, and penetration of plasma-based processing technologies into many areas of science and industry, from structural materials (Boxman et al. 1995; Baranov et al. 2017a) to the space propulsion applications (Charles et al. 2016; Levchenko et al. 2018a, c; Mazouffre 2016) and miniaturized space plasma devices (Lemmer 2017; Charles 2009). Importantly, these technologies enabled efficient synthesis of an enormous variety of remarkably diverse nanostructures (Jacob et al. 2015, 2017), metamaterials, and other material systems not found in nature.

A major reason for the diverse and pervasive applications of plasmas in material processing is the unique features afforded by the charged matter when compared to its neutral state equivalent (Ichimaru 2017; Ostrikov et al. 2013; Tanaka et al. 2017). However, it should be mentioned that being an ionized (and sometimes strongly ionized) medium, plasma environment can be rich in highly energetic species and thus excessively harsh for the processing of finely-resolved and temperature-sensitive nanostructures, introducing undesirable defects into lattices, and changing the structure of complex metamaterials. Thus, a careful selection of plasma parameters for nanofabrication is needed.

After the transition from neutral gas to plasma, ionized atoms can be affected not only by a pressure gradient, but also by Coulomb and Lorentz forces to gain the specified energy and moving direction. Moreover, matter in the plasma state is not in thermal equilibrium and hence is highly chemically-reactive, thus allowing for various chemical reactions which cannot occur at thermal equilibrium (Lieberman and Lichtenberg 2005), and facilitating enhanced transport mechanisms (Zhang et al. 2016; Han et al. 2013; Kumar et al. 2012).

However, a vast variety of plasma-based technologies and distinct methods for their realizations make it quite challenging to match the technological needs with available capabilities (Han et al. 2011). Hence,
  • In-depth examination of physics that governs plasma generation coupled with the critical examination of the available schematic solutions is urgently needed to realize a flexible, multifunctional, rapidly-adjustable universal technological platform for the advanced nanofabrication.

Yet, a practical realization of such a platform, namely its general composition, architecture and philosophy is not trivial. It is apparent that an excessive conglomeration of sub-systems may result in a platform that is too complex and expensive to assemble and operate. On the other hand, the lack of some not presently evident but nevertheless essential components would significantly compromise the total functionality of the entire platform.

Clearly, a systematic, plasma physics-based critical approach is needed to address the aforementioned challenge.

This paper intends to outline an approach to designing such a platform, which could in principle be composed of already approbated schematic and technical solutions, and should be adaptable and flexible enough to satisfy quite different requirements for a wide variety of processes and materials being processed.

We approach this aim by structuring the paper as follows:

  1. 1.

    First, a general concept of a flexible multifunctional universal technological platform capable of replacing a rich variety of plasma technologies by delivering complex material treatment processes is introduced; key elements of the composition and capabilities of such a platform, as well as the background physical principles that govern its operation will be briefly examined (Sect. 2);

     
  2. 2.

    Then, in order to critically examine presently-available suitable solutions, the basic physical principles of plasma generation in technological setups are summarized for the three major groups of technological processes, which could be subdivided by introducing the rate V Σ of the surface geometry change, i.e.

     
  • Processes based on removal of material, i.e. with V Σ  < 0;

  • Processes based on size-invariant modification, i.e. V Σ  ≈ 0, and

  • Additive processes based on material deposition with the apparent V Σ  > 0.

Figure 1 shows the characteristic examples of these three kinds of technological processes, namely the formation of trenches (a process of material removal), then nitriding of a structural alloy (an example of a process without thickness change), and deposition of an artificial diamond from plasma (an additive process resulting in material deposition); this is outlined in Sect. 3.
Fig. 1

Illustrative examples of the three major groups of technological processes, clearly requiring notably different plasma-based approaches and schematic realizations. a A process based on removal of material is illustrated by an SEM image of reactive ion-etched trenches of varying widths and aspect ratio. Reprinted with permission from Abdolvand et al. (2008), Copyright 2008 Elsevier. b A process based on size-invariant treatment—nitriding the structural alloy in nitrogen plasma environment; as a result of the efficient plasma processing, the microhardness of the nitrided layer increased from an initial value of 600 HV0.1 to 1600 HV0.1, i.e. nearly threefold (Chang et al. 2014). Reprinted with permission from Chang et al. Copyright 2014 Elsevier. c An additive process of material deposition—a diamond is grown atop of an Al sublayer on a Si substrate using a microwave plasma technique. Reprinted with permission from Schelling et al. (2005) Copyright 2005 Elsevier

Finally, using experimental works of a large number of researchers, Sect. 4 is intended to:
  1. 3.

    Analyze specific technical solutions from the perspective of their realization; consider the opportunities and limitations for combining of the most typical processes, and their possible implementation in the proposed perspective multifunctional platform.

     

It should be noted that the primary focus of this work is on the relatively energetic processes featuring strongly ionized plasmas with the electron density reaching 1019 m−3, electron temperature ranging from several to about 10 eV, and gas/metal ion energy reaching 1000 eV, with several examples of the use of moderately ionized plasmas also presented for the benefit of the reader. Although such energetic environments are not generally used for the synthesis of intricate or delicate carbon nanostructures, e.g. carbon nanotubes and graphene, their suitability for the synthesis and processing of other valuable nanomaterials merits their inclusion in our article. Indeed, where other review papers are focused primarily on the conditions conducive to the synthesis and growth of most nanocarbon materials, e.g. weakly-ionized plasmas without external magnetic fields, with the density up to 1014 m−3, and electron energy not exceeding 1 eV (Hatakeyama 2017), this paper critically reviews a body of literature covering a much more broader range of parameters, which may complement and extend the concepts discussed in other review papers.

2 General concept of a flexible multifunctional plasma treatment platform

Figure 2 shows a generalized concept for a flexible plasma treatment platform suitable to perform a wide range of technological processes by implementing multiple adjustable interchangeable sub-modules capable of producing plasma environment with quite different parameters and characteristics (plasma density, ion and electron temperatures, etc. (a); and an indicative set of plasma-generating and plasma-conditioning sub-modules (b–h).
Fig. 2

Concept of perspective multifunctional setup. Schematic of a plasma reactor suitable to conduct a wide range of technological processes, and the technological stages: schematic of the setup with optional set of plasma sources and facilities (a); treatment with gas ions extracted from the external plasma source (helicon, ICP, ECR) (b); treatment with gas ions in neutral loop discharge plasma (c) or in dual-frequency CCP reactor (d); surface heating and cleaning by use of hollow cathode magnetron plasma (e); plasma nitriding (f); duplex treatment (g); coating deposition by use of external sources of metal plasma (vacuum arc or HIPIMS) (h). Additional functionality is achieved with the use of a set of magnetic coils mounted at the plasma inlet and under the processed substrate is used (b); the external coils are winded around the chamber to guide the plasma ICP torus (c); the dual-frequency powered electrodes are applied (d); the double-walled cathode structure is moved up and the coil under the substrate is powered (e); the lid with orifices is added (f) or removed (g); the set of coils under the substrate is coupled with the coils of the plasma sources (h)

This conceptual platform (Fig. 2a) features several principal sub-concepts, which are expected to satisfy the general need in presenting various plasma parameters, including the basic ones such as plasma density np, ion temperature Ti, electron temperature Te, energy density and other key parameters. In this view, the conceptual platform should incorporate a sophisticated selection of submodules with different physical mechanisms of plasma production.
  • A set of plasma-generating modules should produce a full range of plasma environments needed for the efficient plasma-based treatment, thus covering the plasma density range np ≈ 1012–1020 m−3, electron temperature range Te ≈ 0.1–10 eV, incident ion energy to in the range 0–104 V, and ambient gas pressures in the range of 10−3–105 Pa.

It is evident that a simple aggregation of different sub-units could hardly cover such wide ranges, and hence, we have proposed a novel concept of the active plasma-generation system. While it builds on a basic chamber, this system is equipped with a carefully-chosen plasma generating sub-systems, and is therefore capable of active generation of large plasma bulks in relatively large volumes, and with relatively high (density, temperatures) parameters. Specifically, the basic chamber should represent an active system made, e.g., as a hollow cathode system (Amemiya and Ogawa 1997) capable of producing plasma in a quite wide range of parameters (Hagelaar et al. 2010). Besides, a range of plasma management and conditioning sub-systems could be involved for selective manipulation of plasma parameters, thus the whole system will comprise (Fig. 2):
  1. 1.

    Hollow-cathode basic manifold (chamber) capable of producing and maintaining relatively dense plasma by holding the electron population electrostatically;

     
  2. 2.

    A system of external coils embracing the basic manifold and installed beneath, for active plasma management and conditioning plasma parameters (heating electron population, directing ions and quasi-neutral plasma, etc.);

     
  3. 3.

    Lower plasma-generating sub-units, e.g. magnetrons, for ensuring three-dimensional plasma exposure for the treated objects; and

     
  4. 4.

    A set of interchangeable and co-operating plasma generating modules capable of sustaining the required plasma parameters and other treatment needs; apparently, the sub-unit set could be specified and diversified for the specific tasks.

     
The set of coils and electrodes forms an entire systems that, when sophisticatedly designed, will allow for a wide, flexible and selective modulation of a range of plasma parameters throughout the entire process volume. Below, let us briefly define the basic sub-modules from the point of view of the fundamental physical mechanisms that govern plasma generation; interestingly, the combination of various physical mechanisms allows for unique possibilities of shaping the plasma (Levchenko et al. 2004a). A characteristic example is shown in Fig. 3, where plasmas of complex structure are generated in a system with the two planar magnetrons and inductively-coupled plasma (ICP) systems.
Fig. 3

Towards flexible plasma-enabled materials treatment platform: a successful attempt to design a multifunctional flexible setup comprising three sub-systems: two planar magnetrons generating metal plasma and working together with the ICP radio-frequency (13.56 MHz) system (a); cylindrical magnetron generating plasma in the system with biased substrate and chamber walls used as anode (b); photograph of plasma with complex structure, generated in a system with the two planar magnetrons and ICP system (c).

Reprinted from Yick et al. (2012)

It is quite logical to subdivide all the plasma-generating sub-systems by the primary physical mechanism which governs ionization of the neutral media (both gas and solids, such as carbon and metals, can be used for the plasma production on technological systems). Hence, we will distinguish the three groups of the plasma generating modules:
  1. 1.

    The plasma generating modules which use the resonant electron impact ionization, i.e. the systems based on radio, high frequency, microwave and other similar types of energy injected to the neutral media for ionization;

     
  2. 2.

    The modules which use electron impact ionization in systems with electron trapping for longer electron trajectories; as a rule, direct current magnetrons and Hall-type systems fall under this type; and finally,

     
  3. 3.

    The systems where electric field and thermionic emission in the burst-like processes in the cathode spots provide the ionization; various arc systems of both pulsed and direct current types are the representative of this class.

     

The gas plasma systems of type 1 (radio, high frequency, and microwave energy) are often built as helicon (Chen 2015; Chen and Torreblanca 2009, ICP (Kim et al. 2004; Vasenkov and Kushner 2004), or microwave (Thomassen 1965) plasma generating sub-system, see schematics in Fig. 2b. The typical helicon or ICP plasma generating module usually incorporates the radio-transparent body made of, e.g., quarts, and coils to supply the energy via electromagnetic waves into the gas. Such modules, as a rule, cannot effectively evaporate solids and thus require the plasma-forming gas atmosphere, mainly Ar. They produce quite dense plasmas, with the density reaching 1018 m−3, with low electron density of several eV (Kim et al. 2013).

Another type of the plasma generating sub-system (type 2) shown in Fig. 2c allows for processing of the substrate with gas ions by the use of a neutral loop discharge (Uchida and Hamaguchi 2008; Sakurai and Osaga 2011. Typically, such a system generates plasmas in the entire volume of the main chamber, in a gas environment. As a rule, less dense plasmas of 1017 m−3 but in larger volumes are created. Importantly,
  • Highest possible plasma density is not a common aim, since the damage and overheating of the surfaces being processed may be an issue!

To process the substrate with highly tunable process environment and plasma parameters, the setup is transformed to the dual frequency (Sung et al. 2012; Perret et al. 2003) capacitive coupled plasma reactor shown in Fig. 2d. If the substrate has a particular design, a low-frequency ICP enhanced by a ferromagnetic core (FMICP) can be generated with a purpose of the uniform processing (Meziani et al. 2001; Godyak 2013). At the same time, the simplest way to heat the workpieces and to conduct the ion cleaning of the surface is achieved by applying the magnetron discharge as shown in Fig. 2e. Here, a double-walled hollow cathode structure (HCS) is pushed up by a drive to ensure the discharge generation.

While the double wall configuration is not required to ignite the magnetron discharge (a single-walled structure will be sufficient), it is still necessary for the processes where the bias is required at the treated part (e.g. for high-energy active cathode nitriding). For this aim, a configuration with a lid with holes (accelerating mesh) should be placed on a top of the double-walled HCS, as shown in Fig. 2f. High voltage applied to the mesh (it may be pulsed) ensures high energy of the ions and hence, deep enough penetration of the ligands into the material.

By removal of the lid, the duplex treatment configuration shown in Fig. 2g could be obtained, thus the nitriding and magnetron coating deposition can be accomplished. Finally, material deposition and growth of complex nanostructures and metamaterials can be achieved by the use of external sources of metal plasma such as vacuum arc (Baranov et al. 2017b; Anders and Brown 2011) or high power impulse magnetron sputtering (HIPIMS) (Gudmundsson et al. 2012; Bohlmark et al. 2006) setups mounted on the chamber flanges. These systems use primary the type 3 ionization mechanism (in the cathode spots) and thus they are capable of producing very dense (reaching 1020 m−3), strongly ionized plasma which could further be focused into dense beams by magnetic lenses (Levchenko et al. 2016a).

Table 1 summarizes the desirable plasma parameters and other characteristics of the plasma-generating modules, as outlined by reviewing a number of relevant publications listed in Table 1.
Table 1

Nominal characteristics of plasma setups for plasma processing using different operating steps

Operating step/rate of geometry change

Plasma discharge

Scheme of substrate electric circuit

Electron density, m−3

Gas pressure, Torr

Power input, W

Modulation of flow of plasma species and energy to the substrate

Ion cleaning and etching

V Σ  < 0

Helicon, ECR (b)

ESS

1018–1019

0.01–0.1

500–2000

Array of helicon plasma sources or set of magnetic coils (Chen 2015)

Neutral loop discharge plasma (c)

10−3–0.1

Neutral loop configuration of magnetic field (Uchida and Hamaguchi 2008)

Capacitive coupled plasma (d)

ISS

1017

0.05–1

200–500

Driving frequency, RF power (dual-frequency CCP etcher) (Sung et al. 2012)

Glow discharge with magnetic confinement (e)

1016

0.1–5

100–1000

Electrostatic and (double-walled hollow cathode structure) magnetic confinement (arc-shaped magnetic field) (Baranov et al. 2011) of plasma

Ion modification and alloying

V Σ  ≈ 0

Glow discharge with (f)

ESS

200–1000

Duplex treatment

V Σ  ≈ 0; V Σ  > 0

Glow discharge (g)

1–500

Material deposition

V Σ  > 0

Vacuum arc (h)

1017–1018

10−4–10−2

500–5000

Set of coils under substrate and magnetic mirror at a plasma generation region (Baranov et al. 2017b)

HiPIMS (h)

1017–1018

10−3–10−2

200–600

Guiding magnetic field of open drift configuration (Bohlmark et al. 2006)

Below we examine the fundamentals of plasma generation physics with respect to the aforementioned schematic solutions, and then consider in more detail the above-mentioned schematic solutions.

3 Basic principles of plasma generation

Ionization of neutral particles by electron impact is the basic principle implemented in plasma technology; thus, an effective technological setup for plasma-enhanced treatment should support the following capabilities:
  1. (a)

    An effective transfer of electric field energy to electrons;

     
  2. (b)

    Full utilization of the energy gained by the electron in the electric field;

     
  3. (c)

    A wide range of powers absorbed by plasma to obtain a wide range of plasma density;

     
  4. (d)

    Decrease plasma loss on the reactor walls, which means an effective transportation of the plasma from a plasma-generating module to a processed substrate;

     
  5. (e)

    A possibility of operating with gas as well as with metal plasmas to conduct a wide range of technological steps: from removal of surface layers to deposition of coatings.

     
To realize these capabilities, the following physical principles that govern plasma behavior need to be considered (Fig. 4):
Fig. 4

Basic principles that govern plasma behavior. More details on the application of these principles to real-life processing setups and the examples of their use could be found in the text

  1. 1.
    A constant electric field with one-time transfer of energy to an ‘igniting’ electron beam (primary electrons) by use of:
    1. 1.1.

      An electron avalanche in the constant electric field caused by the primary high energy electrons.

       
    2. 1.2.

      Magnetic confinement and reflection of plasmas from the elements of the setup structure, with penetration of the electric field into the plasma.

       
    3. 1.3.

      Multiple electrostatic reflections for full utilization of the primary electrons energy.

       
     
  2. 2.
    Alternating electric field and multiple transfer of energy to the igniting electrons from the electric field.
    1. 2.1.

      Oscillatory motion of electrons at their interaction with the electric field of the sheath, when electromagnetic waves excite the plasma surface areas, and then the energy is transferred to the plasma.

       
    2. 2.2.

      Oscillatory motion of the electrons, when the waves act within the whole volume of the plasma.

       
     
In order for the ionization of the background gas atoms by the electron impact, the electrons must gain the energy by interaction with an electric field. The cheapest method to create the necessary field is to apply the electric potential difference between two electrodes (the cathode and anode) to ignite the glow discharge (Braithwaite 2000). The electron avalanche is generated in the discharge gap to create ions in a quantity sufficient to generate secondary electrons from the negatively grounded electrode (cathode) by means of ion bombardment, and to meet the condition of the discharge self-sustaining (Lieberman and Lichtenberg 2005).
$$\alpha d = \ln \left( {1 + \frac{1}{{\gamma_{\text{see}} }}} \right),$$
(1)
where α is the first Townsend coefficient given by the formula:
$$\alpha = \frac{\text{const}}{{\lambda_{\text{inel}} }}\exp \left( { - \frac{{\varepsilon_{\text{iz}} }}{{E\lambda_{\text{inel}} }}} \right),$$
(2)
where λinel is the mean free path of electron for inelastic collisions, E is the electric field, εiz is the energy of ionization.

However, this method has a limitation due to a rather large background gas pressure necessary to effectively utilize the electron energy for the ionizing collisions on their way from the cathode to the anode. Since for technological reasons, the gas pressure should be up to a thousand times less than that used in common glow discharges (from 100 Pa to about 10−1 Pa), various techniques are applied to overcome the pressure limitation either by the use of electric or magnetic fields (Helander 2014; Boeuf 2017; Knapp 2015; Kolobov and Metel 2015).

All of them are based on the idea that a significant increase in the electron path from the cathode to the anode will lead to an increase in the probability of the ionizing collisions and thus will result in a more efficient utilization of the electron energy. However, where the applied electric fields are strong enough to affect both types of the charged particles (i.e. both ions and electrons) in plasmas, for the magnetic fields to directly influence the motion of ions (‘magnetize’), they require powerful sources capable to generate the fields up to a few Teslas, which makes them not cost effective. Hence, the magnetic fields in the technological setup usually are of a few tens of milliteslas.

The particle balance for the entire plasma volume is limited by the density of the background gas pressure na, electron temperature (Kiz term, which is the ionization rate depending on the temperature), and flow of plasma species from the plasma to the reactor walls φw (Chabert and Braithwaite 2011):
$$\frac{{{\text{d}}\bar{n}_{\text{e}} }}{{{\text{d}}t}} = \bar{n}_{\text{e}} n_{\text{a}} K_{\text{iz}} - \frac{{2\varphi_{\text{w}} }}{l}.$$
(3)
A flow of charged particles leaving the plasma in a planar geometry is given by
$$\varphi_{\text{w}} = h_{\text{l}} n_{0} u_{\text{B}} ,$$
(4)
where n0 is the plasma density at the discharge centre, uB is the Bohm velocity, uB = (kTe/M)1/2, and hl is the edge-to-centre density ratio, which depends on the pressure:
$$h_{\text{l}} \approx 0.86\left[ {3 + \frac{l}{{2\lambda_{\text{i}} }} + \frac{{T_{\text{i}} }}{{5T_{\text{e}} }}\left( {\frac{l}{{\lambda_{\text{i}} }}} \right)^{2} } \right]^{ - 1/2} ,$$
(5)
where l is the distance between the electrodes; M is the ion mass; Ti, Te are the ion and electron temperatures, respectively, and λi is the ion-neutral mean free path. The first term dominates at low pressure (λi ≫ l), the second term dominates at intermediate pressure (λi ≈ l), and the last term dominates at high pressure (λi ≪ l).
Thus, in low-temperature plasmas used in technological setups, the behavior of electrons is of a special importance. In a classical approach, the motion of electrons in the plasma can be described as (Chen 1984):
$$mn\frac{{{\text{d}}\vec{v}}}{{{\text{d}}t}} = - en\left( {\vec{E} + \vec{v} \times \vec{B}} \right) - \nabla p - mn\nu \vec{v},$$
(6)
where \(\vec{v}\) is the electron velocity, \(\vec{E}\) and \(\vec{B}\) are the electric and magnetic fields, n is the density of the electrons, ∇p is the pressure gradient, e and m are the electron charge and mass, respectively, ν is a frequency of collisions of electrons with neutrals.
In a stationary mode, Eq. (6) is simplified, and the following expression for the electron velocity across the magnetic field is deduced:
$$0 = - en\left( {\vec{E} + \vec{v} \times \vec{B}} \right) - \nabla p - mn\nu \vec{v},$$
(7)
$$\vec{v}_{ \bot } = - \mu_{ \bot } \vec{E} - D_{ \bot } \frac{\nabla n}{n} + \left( {\vec{v}_{\text{E}} + \vec{v}_{\text{D}} } \right)\left[ {1 + \left( {\frac{m\nu }{eB}} \right)^{2} } \right]^{ - 1} ,$$
(8)
where μ and D are the electron mobility and coefficient diffusion transverse to the magnetic field, \(\vec{v}_{\text{E}} = \frac{{\vec{E} \times \vec{B}}}{{B^{2} }}\) and \(\vec{v}_{\text{D}} = - \frac{{\nabla p \times \vec{B}}}{{enB^{2} }}\) are the electrical and diamagnetic drift velocities, respectively; eB/m = ωce is the electron cyclotron frequency. After the simplification, the following expression is obtained for the velocity of the electron drift across the magnetic field, and it can be seen that this motion is limited by the magnetic field B:
$$\vec{v}_{ \bot } = \left[ {1 + \left( {\frac{eB}{m\nu }} \right)^{2} } \right]^{ - 1} \left( { - \mu \vec{E} - D\frac{\nabla n}{n} + \left( {\frac{e}{m\nu }} \right)^{2} \left[ {\left( {\vec{E} \times \vec{B}} \right) + \frac{{\nabla p \times \vec{B}}}{en}} \right]} \right),$$
(9)
where μ and D are the electron mobility and diffusion coefficient at the absence of the magnetic field.

Magnetic bottle is one of the configurations which explore the possibilities deduced from the above expression. Although the classical approach does not work well in real world plasmas, it is used to describe plasma in a magnetic field; yet, the frequency ν is treated as a frequency of the electron collisions with the fluctuation of the electrical field in plasmas (Bohm conductivity), collisions with the walls of the plasma vessel (near wall conductivity), or collisions with the charged particles neutralized on the walls (Morozov and Savelyev 2000).

Magnetic mirrors usually are a part of the magnetic trap with the magnetic lines parallel to the reactor walls to confine plasmas and to decrease the plasma loss on the reactor walls. To make a mirror, a gradient of the magnetic field, which is parallel to the direction of the field, should be created. The simplest case of the configuration is a magnetic bottle that is generated between two simultaneously powered coils, and magnetic mirrors are created near the bottle ‘necks’ (Baranov and Romanov 2008). The magnetic field is axisymmetric, and a component Br directed along the radius, is connected with an axial component by equation (Chen 1984):
$$\frac{1}{r}\frac{\partial }{\partial r}\left( {rB_{\text{r}} } \right) + \frac{{\partial B_{z} }}{\partial z} = 0.$$
(10)
By analysing the components of the Lorentz force, an axial component of the force is deduced:
$$F_{\text{z}} = \frac{{qv_{\theta } r}}{2}\frac{{\partial B_{\text{z}} }}{\partial z},$$
(11)
where v θ is a axial component of the charge q motion.
For a particle with a guiding centre on the axis of symmetry of the magnetic bottle, v θ is considered to be constant, v θ  = v . At that r = rL (Larmour radius), and the average force is
$$F_{\text{z}} = \mp \frac{{qv_{ \bot } r_{\text{L}} }}{2}\frac{{\partial B_{\text{z}} }}{\partial z} = - \frac{{mv_{ \bot }^{2} }}{2B}\frac{{\partial B_{\text{z}} }}{\partial z} = - \mu \frac{{\partial B_{\text{z}} }}{\partial z},$$
(12)
or, in general
$$\vec{F}_{\parallel } = - \mu_{\text{m}} \nabla_{ \bot } B,$$
(13)
where µm is a magnetic moment of the gyrating particle.
Considering that B0 and Bm are the maximum and minimum of the magnetic field in the trap, a ratio of these parameters determines the smallest angle θm in velocity space for the confined particle, and particles with the angle θ > θm can leave the magnetic trap along the magnetic lines:
$$\frac{{B_{0} }}{{B_{\text{m}} }} = \sin^{2} \theta_{\text{m}} .$$
(14)

Hollow cathode discharge is a concept that is widely applied in the plasma processing and is an alternative for the magnetic confinement (Hagelaar et al. 2010; Gallo et al. 2011). The plasma discharge is confined in the cathode designed to reflect the electrons from the electrostatic sheath associated with the cathode walls. This concept allows for obtaining a significantly larger plasma density compared to the common glow discharge, yet this enhancement is not because of the magnetic confinement as in previous case, but is rather attributed to the electrostatic confinement due to the electrostatic mirroring of the secondary electron on its path to the anode. The most used shape of the hollow cathode is a thin-walled long cylinder; however, other configurations are used in plasma processing, e.g. two long plates or two coaxial large-diameter cylinders with the discharge between the walls.

Due to the design features, no sheath multiplication is required as in common glow discharge, and the electron energy gained in the plasma sheath is utilized much more effectively. The electrons emitted as a result of the ion bombardment of the cathode (or due to the thermionic emission from the heated cathode), are described as hot electrons, and they bounce between the charged sheathes, thus generating the plasma, or cold, electrons. The simple model of the hollow cathode discharge for a small-diameter long cylindrical cathode is developed by Lieberman, and the present model for the planar geometry is based on the same assumptions (Lieberman and Lichtenberg 2005).

Here, the hot electrons are generated by secondary emission due to ion bombardment of the two plates. The electrons are accelerated by the electric field of the plasma sheath and go to another wall, where they are reflected, thus oscillating between the walls. The discharge gap L between the walls should be lower than the electron energy relaxation length (Chabert and Braithwaite 2011; Godyak 2006):
$$\lambda_{\text{er}} = \lambda_{\text{el}} \left[ {\frac{2m}{M} + \frac{{\nu_{\text{ee}} }}{{\nu_{\text{m}} }} + \frac{2}{3}\left( {\frac{{e\varepsilon_{\text{ex}} }}{{kT_{\text{e}} }}} \right)\frac{{\nu_{\text{ex}} }}{{\nu_{\text{m}} }} + \frac{2}{3}\left( {\frac{{e\varepsilon_{\text{iz}} }}{{kT_{\text{e}} }}} \right)\frac{{\nu_{\text{iz}} }}{{\nu_{\text{m}} }} + 3\frac{{\nu_{\text{iz}} }}{{\nu_{\text{m}} }}} \right]^{ - 1/2} ,$$
(15)
where λel is the mean free path for electron–atom elastic collisions; 2m/M term presents the electron energy loss due to elastic collisions and the corresponding heating of the background gas; νee/νm term reflects the electron energy loss by electron–electron coulomb collisions; the third and fourth terms present the electron energy loss by ionization and excitation, respectively, and the fifth term accounts for the energy loss at the boundaries (νee, νm, νex, νiz are the electron–electron collision, electron-atom elastic collision, excitation, and ionization frequencies, respectively).
The ambipolar diffusion is a driving force for the plasma electrons and ions in the hollow cathode discharge, and the ion motion is described by the expression (Lieberman and Lichtenberg 2005):
$$- D_{\text{a}} \frac{{{\text{d}}^{2} n_{\text{i}} }}{{{\text{d}}x^{2} }} = k_{\text{iz}} n_{\text{a}} n_{{{\text{h}}0}} ,$$
(16)
where nh0 and Kiz are the density and ionization rate for the hot electrons, respectively; na is a density of the background gas, and Da is the ambipolar diffusion coefficient. By integrating one may obtain:
$$\frac{{{\text{d}}n_{\text{i}} }}{{{\text{d}}x}} = - \frac{{k_{\text{iz}} n_{\text{a}} n_{{{\text{h}}0}} }}{{D_{\text{a}} }}x + C_{1} .$$
(17)
The flow of ions is described by expression
$$\varphi_{\text{i}} = - D_{\text{a}} \frac{{{\text{d}}n_{\text{i}} }}{{{\text{d}}x}} = k_{\text{iz}} n_{\text{a}} n_{\text{h0}} x - D_{\text{a}} C_{1} ,$$
(18)
and it equals zero for x = 0 since the distribution of the ion density is supposed to be symmetric relative to the discharge gap; hence, C 1  = 0.
After repeated integration, we obtain
$$n_{\text{i}} = - \frac{{k_{\text{iz}} n_{\text{a}} n_{\text{h0}} }}{{D_{\text{a}} }}\frac{{x^{2} }}{2} + C_{0} .$$
(19)
Since ni = nh0 at x = L/2, and ni = nh0 + ne0 at x = 0 (where ne0 is the density of the plasma electrons in the discharge centre):
$$n_{\text{h0}} = - \frac{{k_{\text{iz}} n_{a} n_{\text{h0}} }}{{D_{\text{a}} }}\frac{{L^{2} }}{8} + n_{\text{h0}} + n_{\text{e0}} ,$$
(20)
we obtain a relation between ne0 and nh0:
$$n_{\text{e0}} = \frac{{k_{\text{iz}} n_{\text{a}} }}{{8D_{\text{a}} }}L^{2} n_{\text{h0}} .$$
(21)

Here, the density nh0 is much lower than ne0, and the plasma density in the discharge centre approximately equates to ne0. The plasma density is inversely proportional to the diffusion coefficient Da that can be affected by applying the transverse magnetic field to decrease the coefficient, thus making an additional circuit to modulate the plasma density due to the combination of the electrostatic and magnetic confinement.

As an alternative to the electro- and magnetostatic confinement, alternating electric field is applied resulting in generation of radio-frequency (RF) plasmas (Chen and Chang 2002). Here, the electron travelling from the cathode to the anode is limited by the alternating fields; thus the electron oscillates and undergoes multiple ionizing collisions with the background gas before reaching the electrode or walls of the processing chamber. RF domain is located within the range of 1–200 MHz (Perret et al. 2005), thus affecting the ions at the lower end, and the electrons throughout the entire frequency range. While in common DC electropositive plasmas, a negatively biased electrode is used to modulate the energy of ions extracted from the plasma to the treated surface while the floating potential is not large enough (a few tens of volts), the RF sheath modulation allows for obtaining the floating potential that is significantly more negative (Chabert and Braithwaite 2011; Kawata et al. 2008; Rahman and Dewan 2015).

Various models of RF sheaths are developed, and the fixed ion density model that ignores all ion dynamics is the simplest of them (Chabert and Braithwaite 2011).

The electric field of a small-amplitude electromagnetic wave propagating in the z-direction may be defined as (Chabert and Braithwaite 2011)
$$E_{\text{x}} = {\text{Re}}\left[ {\tilde{E}_{\text{x}} \exp \left\{ {i\omega \left( {t - n_{\text{ref}} z/c} \right)} \right\}} \right],$$
(22)
where nref is the refractive index, and c is the speed of light.
Since the refractive index is a complex number nref = nr + inim, two domains for frequency ω can be distinguished:
  1. (i)
    ω > ωpe, where ωpe = (ne2/ 0 )1/2 and ωpi = (ne2/ 0 )1/2 are the plasma electron and ion densities, respectively; typically, the electron frequency lies in the GHz range, and since the pressures are below 100 Pa, the collision frequency is usually small compared to the wave frequency, νm/ω ≪ 1, and
    $$n_{\text{im}} \approx 0$$
    (23)
    $$n_{\text{r}} \approx \left( {1 - \frac{{\omega_{\text{pe}}^{2} }}{{\omega^{2} }}} \right)^{1/2} .$$
    (24)

    As such, the electromagnetic waves with a frequency ω propagate in the plasmas with very weak attenuation.

     
  2. (ii)
    (ii) The domain ω < ωpe at low pressure νm/ω ≪ 1, where
    $$n_{\text{im}} \approx \frac{{\omega_{\text{pe}} }}{\omega }.$$
    (25)
    $$n_{\text{r}} \approx 0.$$
    (26)
     
Here, the waves decay with a characteristic scale called the inertial skin depth given by
$$\delta \approx \frac{c}{{\omega n_{\text{im}} }} = \frac{c}{{\omega_{\text{pe}} }}.$$
(27)
For high-pressure limit of νm/ω ≫ 1, the collisions are important and result in high resistivity, and the refractive index and resistive skin depth are
$$n_{\text{im}} \approx \frac{{\omega_{\text{pe}} }}{\omega }\left( {\frac{\omega }{{2\nu_{\text{m}} \omega }}} \right)^{1/2} .$$
(28)
$$\delta \approx \frac{c}{{\omega n_{\text{im}} }} = \frac{c}{{\omega_{\text{pe}} }}\left( {\frac{{2\nu_{\text{m}} }}{\omega }} \right)^{1/2} .$$
(29)

Three modes are distinguished to describe the coupling between the electrons and the power of the external RF generator: the electrostatic (E) mode, the evanescent electromagnetic (H) mode, and the propagating wave (W) mode.

Capacitively coupled plasmas (CCPs) consist of two parallel electrodes (with a radius of a few tens cm) separated by a distance of a few cm, and biased by RF power supply, typically operating at 13.56 MHz (Lieberman and Lichtenberg 2005; Rauf et al. 2010). Generally, CCPs operate in the E-mode.

According to the basic model of a symmetrical capacitively coupled plasma discharge (Chabert and Braithwaite 2011), when the sheath thickness is small compared to the distance between the electrodes (sm ≪ L), and stochastic heating (collisionless) dominates in the power absorption (which is valid at a pressure less than a few Pa), the plasma density is:
$$n_{0} = \left[ {\frac{{\varepsilon_{0} K_{\text{stoc}} k_{\text{cap}} \left( {\text{mM}} \right)^{1/2} }}{{4eh_{\text{l}} \varepsilon_{\text{T}} \left( {T_{\text{e}} } \right)}}} \right]\omega^{2} V_{0} ,$$
(30)
and it increases linearly with the applied RF voltage V0, and exhibits a quadratic dependence on the applied frequency. Similar result is obtained for high pressure mode, when ohmic heating (collisional) dominates in the plasma. In the latter expression k is Boltzmann constant, Kstoc, Kcap are constants (Kcap = 0.613, Kstoc = 0.72 for collisionless sheath, and Kcap = 0.715, Kstoc = 0.8 for collisional sheath), and εT(Te) is described by the equation:
$$\varepsilon_{\text{T}} \left( {T_{\text{e}} } \right) = \varepsilon_{\text{iz}} + \frac{{K_{\text{ex}} }}{{K_{{{\text{i}}z}} }}\varepsilon_{\text{ex}} + \frac{3m}{M}\frac{{K_{\text{el}} }}{{K_{\text{iz}} }}kT_{\text{e}} + 2kT_{\text{e}} + eV_{\text{w}} .$$
(31)
where εiz, εex are the energies of atom ionization and excitation, respectively, and Vw is a potential drop near the walls.

The plasma density and the sheath voltage are determined by the RF power and cannot be governed independently in a single-frequency CCP (Diomede and Economou 2014; Wilczek et al. 2015). To overcome this limitation, dual-frequency RF plasma discharges are used, were one frequency is responsible for the plasma generation, while another determines the ion energy (Lieberman et al. 2002).

Magnetic field directed parallel to the electrodes can be applied in CCP plasma reactor to enhance its effectiveness, as demonstrated in the magnetically enhanced reactive ion etchers (MERIE) (Kushner 2003; Yang et al. 2017).

Inductive discharges are generated by RF current in a coil separated from the plasma by a dielectric window. In ICP reactors, a substrate is powered independently from the plasma RF generator, so it is possible to separately regulate the energy and the flow of ions, because the plasma is generated by a coil and the substrate holder is biased independently (Chabert and Braithwaite 2011; Godyak 2013).

For the generation of ICP discharges in a cylindrical tube, N turns are winded on it, and RF sinusoidal current is applied:
$$I_{\text{RF}} \left( t \right) = {\text{Re}}\left[ {\tilde{I}_{\text{RF}} \exp \left\{ {i\omega t} \right\}} \right].$$
(32)
For the low-pressure and high-frequency limit (νm/ω ≪ 1) at the high plasma density, the electron density is:
$$n_{\text{e}} = \left[ {\frac{{\pi r_{0} N^{2} \nu_{\text{m}} \left( {m/\varepsilon_{0} } \right)^{1/2} }}{{4u_{\text{B}} \left( {h_{\text{l}} \pi r_{0}^{2} + h_{\text{r0}} \pi r_{0} l} \right)e\varepsilon_{\text{T}} \left( {T_{\text{e}} } \right)lc}}} \right]^{2/3} I_{\text{coil}}^{ 4 / 3} .$$
(33)

Thus, for a given current, the electron density increases with the number of turns; the density does not depend on the frequency.

For the high-pressure and low-frequency limit (νm/ω ≫ 1) at high plasma densities, the electron density is described by the expression:
$$n_{\text{e}} = \left[ {\frac{{\pi r_{0} N^{2} \left( {2\omega \nu_{\text{m}} } \right)^{1/2} \left( {m/\varepsilon_{0} } \right)^{1/2} }}{{4u_{\text{B}} \left( {h_{\text{l}} \pi r_{0}^{2} + h_{\text{r0}} \pi r_{0} l} \right)e\varepsilon_{\text{T}} \left( {T_{\text{e}} } \right)lc}}} \right]^{2/3} I_{\text{coil}}^{ 4 / 3} ,$$
(34)
and it increases with the frequency.

ICP reactors can operate both in E-and in H-modes (Filipič et al. 2014).

Helicon plasmas are generated when a static magnetic field is applied along the length of the cylindrical tube intended for the RF plasma generation, and enables propagation of electromagnetic waves at the condition of ω ≪ ωpe (Stenzel and Urrutia 2015; Chen 2012; Furukawa et al. 2017). In this case, the right-hand polarized (RHP) wave can propagate along the magnetic field lines with the dispersion relation
$$n_{\text{ref - R}}^{ 2} = 1 + \frac{{\omega_{\text{pe}}^{ 2} }}{{\omega \omega_{\text{ce}} \left( {1 + \frac{{\omega_{\text{ci}} }}{\omega } - \frac{\omega }{{\omega_{\text{ce}} }}} \right)}},$$
(35)
which is transformed at the condition of ωce ≪ ωpe and ω ≪ ωpe:
$$n_{\text{ref}} = \frac{{\omega_{\text{pe}} }}{{\left( {\omega \omega_{\text{ce}} } \right)^{1/2} \left( {1 + \frac{{\omega_{\text{ci}} }}{\omega } - \frac{\omega }{{\omega_{\text{ce}} }}} \right)^{1/2} }}.$$
(36)
At the condition ω < ωce, the dispersion relation with ωci ≪ ω is
$$n_{\text{ref}} = \frac{{\omega_{\text{pe}} }}{{\left( {\omega \omega_{\text{ce}} } \right)^{1/2} \left( {1 - \frac{\omega }{{\omega_{\text{ce}} }}} \right)^{1/2} }} .$$
(37)
When ωci ≪ ω < ωce, the expression is transformed to the dispersion relation for helicon waves
$$n_{\text{ref}} = \frac{{\omega_{\text{pe}} }}{{\left( {\omega \omega_{\text{ce}} } \right)^{1/2} }}.$$
(38)

Helicon reactors operate in the W-mode, and are designed to allow helicon wave propagation with a frequency of 13.56 MHz and the typical conditions of operation in argon are ne = 1018 m−3 and B0 ≈ 10 mT.

For frequencies close to electron cyclotron frequency ω = ωce a resonance occurs, when the electrons are rotating synchronous to the electric field wave, which results in an effective resonant heating. This process is fundamental for the electron cyclotron resonance (ECR) plasma processing reactors, with the applied frequency of 2.45 GHz, and a corresponding magnetic field of 87.5 mT.

4 Detailed examination of specific schematic solutions

Here we will examine in more detail several specific schematic solutions and their operational modes, with respect to their potential for implementation and integration into a conceptual multifunctional plasma treatment platform.

We can outline the following technical solutions in terms of their ability to realize the aforementioned key physical principles that should preferentially govern processes within a complex treatment platform (Fig. 5).
Fig. 5

Key capabilities of the effective technological setup, technical solutions required for their realization, and physical principles that govern the desired plasma behavior

  • It is evident that a single technical solution cannot support realization of the entire set of capabilities necessary to make an efficient plasma setup.

  • A flexible, multi-purpose reactor, where all of the physical solutions can be realized in response to the needs of the specific steps along the technological cycle is needed.

Together, the flows of ions and electrons determine the power of the heat transfer to the treated surface and, in conjunction with other factors, the temperature of the surface and the balance between incident neutral and ion species. The latter, in turn, determines the processes of nucleation and concomitant growth of the new phase (Hovsepian et al. 2014). Fundamentally, the plasma technologies can be divided into three main groups with respect to the rate of change in surface geometry: (i) those with the negative rate (sputtering and reactive etching), (ii) those with the rate close to zero (modification, implantation and alloying), and (iii) those with the positive rate (thin film growth and coating deposition). Depending on the nature of materials and application environment, plasma treated workpieces may exhibit dramatic increase in operational capability and mechanical properties due to the change in chemical composition, structure and morphology of the treated surface layers (Shimizu et al. 2015; Ferreira et al. 2016; Wang et al. 2015; She et al. 2013; Zhang et al. 2015; Baranov et al. 2013). An example of such changes is illustrated in Fig. 1.

Increasing complexity, miniaturization, and degree of integration of modern devices demand technologies with the capacity to deliver desired plasma-generated species to the surface of workpieces in a highly-defined manner, yet current individual technologies are close to reaching their capacity. On the other hand, by combining the distinct technologies, it may be possible to not only achieve a more efficient and precise delivery of the desired species to the surfaces but also circumvent the costly, multi-environment processing, such as that involving setup rearrangement or in-process transportation between the plasma setups, which may negatively affect the quality of the produced structures (Anders 2014; Yazdi et al. 2014; Luo et al. 2013; Sun et al. 2012). Yet, integration of distinct technologies setups is not trivial, and generally relies on expensive and lengthy efforts of trial and error.
  • A combination of multiple adjustable and interchangeable plasma-generating sub-modules, and additional modules for finely-resolved manipulation of individual plasma parameters, such as constant or alternating electric fields and alternating (with a varied in time space configuration) magnetic field realize the 6th technical solution for the concept of the effective multifunctional plasma platform.

The next part of this paper describes the typical schematics for plasma generation and manipulation, and their operational modes, which may be selectively combined in the proposed flexible setup to cover the technological needs of modern industry. We would like to stress that the emphasis of this paper is on the presently-available processes and their realization in the existing plasma setups. We also acknowledge that in addition to providing significant opportunities for workflow optimization, combining distinct solutions and thereby increasing the complexity of the setup is fraught with new challenges. However, these challenges should not discourage the reader from pursuing these novel technological setups. Indeed, the aim of this paper is to equip the reader with a fundamental understanding of the key principles, advantages and disadvantages of typical modules used for plasma generation and manipulation, and thus assist them with the appropriate choice of such modules for the flexible processing system.

We would also like to note that the perspective multifunctional setup shown in Fig. 2 is just an example of a single complex flexible manufacturing system that serves to illustrate the complexity and flexibility of the approach, and possible opportunities for the different modes of operation, with many other combinations possible. For this reason, in this paper, we chose to not only describe the modules used in this particular example, but also other promising modules and examples that can be integrated to realize the most effective plasma treatment for a given purpose.

4.1 Etching and deposition

Plasma etching and deposition processes are cornerstones of modern semiconducting industry, where they are used in a multitude of applications. In particular, plasma etching represents a group of processes with negative rate of the surface geometry change (i.e. with removal of the surface layers) and is used as an alternative to the conventional wet-chemical hydrofluoric acid dip for solar cell processing (Tang et al. 2014). In nanophotonics, the GaAs/AlGaAs inductively coupled plasma (ICP) etching has been used as a single-step process for production of semiconductor nano-waveguides (Volatier et al. 2010). It was shown that surface passivation in BCl3/N2ICP etching of GaAs-based photonic crystals (PhCs) is crucial for achieving cylindrical, vertical PhC holes (Atlasov et al. 2009).

Of particular value is the ability of plasma treatment to selectively remove material, creating deep yet narrow surface features. Deep etching of silicon has been investigated in an ICP etch reactor using short SiCl4/O2 plasma steps to passivate the sidewalls (Duluard et al. 2011). Interestingly, when Stillahn et al. (2011) compared the etching and passivation during etching of Si and SiO2 in SF6/O2 plasmas, the etch rate of Si was found to increase monotonically with power, in contrast to the etch rate of SiO2 which remained insensitive to this change in plasma conditions.

Critical to the success of these processes is the ability to selectively modulate spatio-temporal distribution of plasma-generated effects, including the motion and characteristics of ions and electrons within the processing setup. To some extent, these processes can be defined through a careful selection of processing gas, energy used to generate and sustain plasma, and a temperature distribution within the reactor. Yet, for more complex treatment processes where the distribution of specific plasma-generated species and their delivery towards the surface of the substrate need to be altered and switched quickly during processing, the realization of these processes necessitate the use of other mechanisms. This is particularly true for large-scale substrates where multiple workpieces are to be processed and for multi-step processes that involve etching and deposition.

Deposition is a process with positive rate of the surface geometry change, since a new layer is formed on a workpiece. The precise delivery of the matter from plasma to the surface of the substrate can be enabled and indeed further enhanced by the use of magnetic field (Fig. 6). For instance, application of a multi-polar magnet field to the ICP plasma-generating module during the deposition of silicon film on a large-area substrate using SiH4/H2was used to increase the deposition rates, which transformed the structure of the silicon films deposited on the glass substrates from amorphous to nanocrystalline (Kim et al. 2009). Furthermore, the magnetic confinement resulted in a higher deposition uniformity, and thus higher quality of the film.
Fig. 6

Plasma etching and passivation of surface layers (including MEMS production), or plasma nitriding. a ICP and Helicon plasma-generating modules for gas plasma generation are attached to the chamber flanges to make an array while a plasma-modulating sub-system (i.e. a set of magnetic coils) is installed under the substrate for additional manipulation and enhancement of specific plasma parameters. Some of the external plasma-generating modules can be the sources of metal plasma thus allowing etching and deposition of the surface layers. The coils under the substrate govern the motion of species extracted from the plasma discharges generated by external plasma modules. Thus-generating gas plasma discharge in a self-sustainment mode (magnetron discharge) is used to treat the surface in a DC or pulse-bias modes; b distribution of the ion current density, when only centre coil is powered under the substrate; the distribution is defined by the discharge power and configuration of the magnetic coil core; c distribution of the ion current densities along the substrate, when any one of the coils is powered (flat) or two side coils are powered in line (thus a magnetic cusp is formed between them), which results in focusing of the flow of plasma species from the external plasma-generating module to the substrate centre

The utility of the magnetic field confinement is not restricted to deposition of smooth, defect-free coatings. As was demonstrated by Luo et al. (2015), with assistance of a strong magnetic field (up to 0.5 T in the centre) during chemical vapor deposition (CVD), growth of carbon nanofibers (CNFs) resulted in a notable reduction in the diameter and improvement in uniformity of CNFs with an increase in the magnetic field. This resulted in the transformation of the disordered CNFs into bamboo-liked carbon nanotubes.

The ability of magnetic field to tune specific plasma parameters can be successfully applied in areas other than semiconductor manufacturing. The combined magnetron discharge and high-frequency inductive discharge located in the external magnetic field showed good performance for magnetron deposition of coatings with ion assistance, where the magnetron discharge is used to supply the materials for the coatings while the ICP discharge is a source of ions used for the ion assistance. The magnetic field is used to optimize the power input to the plasma, and to regulate the motion of ions and their distribution along the treated substrate. The uniform distribution is also obtainable for the hybrid system, such as in a system with the diameter of 150 mm at 0.7 Pa of gas pressure, where the magnetic fields near the substrate is about 2–8 mT and 0.5–25 mT at the RF input supply (Vavilin et al. 2016).

Magnetically-enhanced external arc plasma-generating modules are widely applied for thin film deposition and surface modification, such as nitriding. The utility of this approach has been demonstrated by Zhao and colleagues, who modified AISI 304 stainless steel by low-pressure arc plasma-assisted nitriding (PAN) process. The treatment was performed for 60 min using pure nitrogen at a pressure of 0.6 Pa, temperatures in a range of 380–580 °C. A pulsed bias of 700 V was applied to the samples with a frequency of 40 kHz. The gas-arc-discharge plasma was generated by a hot-cathode gas arc, and the current density on the sample surface reaches the value of 32 A/m2, which caused a surface hardening up to the value of 1300 HV0.05 at a higher nitrided rate of about 30–50 μm/h (Zhao et al. 2012).

Additional processing flexibility can be attained through the use of several distinct plasma-generating modules within a single setup, where magnetic field is used to enhance and manipulate the flow of plasma species as they are directed toward the processed surface. The time-averaged distribution of ions within the chamber volume is the result of the superposition of the individual distributions, the statistical weights of which are derived from the relative duration of the plasma treatment under each magnetic field configuration.

Krysina and colleagues developed a more complex system with two independent plasma-generating modules, namely, arc and hot filament, both containing modules for generation of magnetic fields, which was then used for the synthesis of nitride coatings. In this system, the arc evaporator and a hot filament gas-plasma module with hollow cathode can be operated independently or simultaneously (Krysina et al. 2016). Hot biased filament and biased substrate were used by Wei and colleagues for nitriding of AerMet 100 material at a temperature of 427–480 °C for the treatment time up to 100 h. The surface hardness was found to increase to 800–900 HK and a 120 mm nitride layer was generated with a purpose of gear applications (Wei et al. 2007).

Based on the key mechanisms discussed in the above reports, the following operation mode is proposed for the flexible treatment with gas plasma ions, where the magnetically-enhanced gas plasma can be used for broad-brush or selective etching, or plasma nitriding of the surface layer, while the magnetic traps are used to shape the plasma distribution along the substrate (Fig. 6).

When plasma is generated near the substrate by the use of the centre or side coils under the substrate, the plasma production is limited by a power supply of the plasma generating circuit. As such, very intense plasma can be generated with enhanced plasma consumption due to low plasma loss on the chamber walls. When plasma is transported from the external plasma-generating module, the coils under the substrate ensure the uniform distribution of the desired plasma species over the substrate up to 300 mm in diameter. Alternatively, they can be used to focus the flow of such species to a selected area of 100 mm in diameter (Baranov et al. 2011). Indeed, placement of the coils and their independent operation provide for extended diversity of the processes that can take place within a single plasma-enhanced setup.

Complex treatment of workpieces is possible in such an operation mode of the proposed setup as shown in Fig. 7. When the magnetic coils are mounted on the plasma-generating module, it affects greatly not only the process of plasma generation [helicon (Chen 2015), ECR (Bowles et al. 1996), vacuum arc (Anders 2008), magnetron sources (Lieberman and Lichtenberg 2005)], but also the propagation of the plasma towards the substrate, by creating a magnetic mirror in the region of the plasma generation. As a result, plasma expands from the source (Fietzke et al. 2009), and the plasma losses on the chamber walls are decreased due to the magnetic field, as was demonstrated by Window and Savvides (unbalanced magnetrons), Bohlmark and Gudmundsson (HIPIMS) (Gudmundsson et al. 2012), Anders and Boxman (arc sources) (Boxman et al. 1995), and others. When more than one coil is engaged to modulate plasma propagation, various configurations of the resulting magnetic field are generated, with their shape and magnitude dependent on the powering of the coils. When only two coils are powered in-line, a trap with magnetic bottle configuration occurs between the coils, which results in the generation of dense plasmas in the trap. When the coils are powered opposite, a magnetic cusp (circled mirror) is generated between the coils (Chen 1984).
Fig. 7

Low voltage plasma immersion ion implantation and deposition of mono- and multilayer coatings on extended and small-sized workpieces (including complex-shaped parts) using gas or metal plasma modules with guiding magnetic field. Small-sized workpieces are arranged in a chamber between the poles of two electromagnets (a), which are used in two operation modes (b): gas plasma generation in a self-sustainment mode by applying the negative DC or pulsed bias to the workpieces and generating a magnetic bottle between the coils (yellow distributions in b), and interaction with an external plasma-generating modules and scanning with the plasma-generated species along the workpieces by various powering of the magnetic coils (blue distributions in b). Complex intense treatment of the workpieces is possible when magnetic bottle trap is generated between two in-line powered coils: without any external plasma source (the distributions of ion flux is shown in yellow and marked “1”), or with the plasma sources (the distributions, shown in blue and marked “2–6”, are the ion fluxes between coils powered in-line, while the distributions marked “7–8” are the fluxes extracted from two opposite powered coils) in gas plasma. Rapid ion heating and cleaning, lower background gas, large area of plasma treatment, and a possibility to obtain multilayer structure with complex composition are some of the advantages of this setup. No complex plasma-generating modules for gas plasma production are necessary in the setup since the magnetron discharge generated around the workpieces allows to enhance the productivity by 10 times comparing to glow discharge

In the latter configuration, the plasma density is lower compared to the case of the magnetic bottle. However, the discharge is more stable due to facilitated transport of the plasma electrons to the chamber walls (anode).

It is evident that switching of different combinations of magnetic coils yields a number of distinct configurations of the magnetic field, and, consequently, generation of unique processing environments. If two powered coils are parts of a magnetron source and arranged angularly relative to each other, plasmas can be directed from the plasma-generating module to the workpieces, as was shown by Musil (2015) in the dual magnetron setup. When powering the coils arranged on the opposite sides of the chamber, where the workpieces are located in the trap and negative bias is applied to them, the magnetic bottle ensures plasma generation in the magnetic trap formed between the coils via PIII technique. In a case when one of the opposite-located coils is mounted on the operating plasma-generating module and serves as plasma duct, the plasma focusing to the treated parts is possible, as was demonstrated by Petrov (Shin et al. 2002) and others. In a configuration where other coils are arranged along the perimeter of the chamber, these coils can also be powered to enhance the electron drift to the chamber walls and prevent the negative influence of the bottle configuration on the operation of plasma sources (Baranov et al. 2012; Bilek et al. 2001).

4.2 Large area processing

Plasma treatment of large-area substrates is an important requirement for industrial applications, since the plasma setup productivity is determined by the number of workpieces which can be treated within a single technological run. Many current and emerging applications rely on the use of large-area sheet materials, such as metal plates, glass panes, etc. Since a typical technological cycle includes operations of ion heating, cleaning, and etching performed usually with gas ions, and operations of material deposition, which typically requires the presence of metal ions, two types of plasma-generating modules are involved in the treatment: gas and metal plasma sources. It is strongly desirable that both of these module types are mounted on the flanges of the same setup, thus minimizing the time and effort lost for pumping, transportation, and possible cleaning, as well as preventing possible poisoning of the semi-manufactured workpieces due to the influence of atmosphere.

Addition processing flexibility can be attained by using magnetic coils with the physical movement of the workpieces within the volume of the processing setup, either through substrate rotation or through roll-to-roll approach. The schematic depicted in Fig. 8 may be useful for operation modes of treatment of items with large surface area. Figure 8a presents a setup configuration with the capacity for treatment of sheet workpieces of approximately 0.5–1 m in diameter. The workpieces are fixed on a circular substrate that can rotate by means of a driving shaft, thus exposing the workpieces to spatially separated gas and metal plasma-generating modules. Gas plasma module can be presented by a planar magnetron provided with a hollow cathode structure (HCS) above the substrate to lower the background gas pressure and to ensure the discharge ignition. Depending on the HCS bias, the substrate can be treated with gas atoms from the plasma ignited within the confinement region, or with metal atoms sputtered from the HCS surface. The external plasma-generating module mounted on a flange of the chamber can also act as a source of gas or metal plasma. In the latter case, the additional magnetic coils mounted under the substrate opposite to the plasma-generating module may facilitate shaping of the distribution of the ion current along the substrate surface for the purpose of uniform treatment.
Fig. 8

Schematic of setup configuration for low voltage plasma immersion ion implantation and deposition on disk sheet items (a) and distribution of the ion current along the substrate: a magnetic trap of arc configuration is used to generate and confine gas plasma near a substrate with a purpose of ion cleaning, heating, and functionalization; a set of magnetic traps of bottle and cusp configurations is generated between an external plasma-generating module and the substrate at the coating deposition with a purpose to modulate the ion current density distribution along the substrate surface; then a drive shaft supplies the treated area under the external plasma-generating module; when treating the substrate with gas ions (yellow distribution in b, c), the coil under the substrate is used without ferromagnetic core but with a hollow cathode structure above the substrate to ensure the closed drift of the plasma electrons; when treating the substrate with ions from the external plasma-generating module, the distribution (blue distribution in b, c) is governed by the powering of the coils

Figure 9 presents an alternative plasma setup configuration for the treatment of items, where roll-to-roll in-process float (Yamada et al. 2012) of the treated parts is used in place of rotating substrate to move the workpieces between distinct processing zones. Various plasma-generating modules are also spatially separated in the setup, and can also be sources of gas or metal ions. The magnetic coils mounted under the substrate opposite to the external plasma-generating module are aligned to defocus the flow of plasma species and create a ribbon-like distribution of the ion current along the substrate width (Baranov et al. 2009a). Rapid ion cleaning and heating, lower pressure of background gas, absence of gas plasma sources, a possibility to treat sheet items with diameter up to 1 m or width of about 0.5 m, a possibility to obtain the surface layers of complex composition and various structure by use of different cathodes of the external plasma-generating modules and by utilization of metal atoms sputtered from the hollow cathode structure are among the advantages of this setup.
Fig. 9

Low voltage plasma immersion ion implantation and deposition of sheet items with roll-to-roll in-process float of the treated parts: a front view and b side view of the equipment and the flow of ions. For treatment of the items with gas ions, a planar magnetron system provided with the hollow cathode structure is applied to generate an axisymmetric distribution (c, d); in contrast, the coils coupled with the external plasma-generating module are under the substrate in such a way to generate a ribbon-like distribution by use of magnetic fields to treat the passing items uniformly

Let us now consider an example of a common industrial process that typically requires the use of several distinct processing steps within independent environments, and for which a single processing environment can offer distinct economic and time savings. Wear-proof, decorative, and functional (e.g. anti-fog) coatings are widely applied in modern industry for machine parts and consumer goods (Kawasaki et al. 2011; Di Mundo et al. 2014; Martinu and Poitras 2000; Takikawa and Tanoue 2007; Chu 2013). These items need to be processed in large quantities while keeping the cost and time associated with such processing low. Yet, the current methods of processing require multiple steps, such as surface preparation and cleaning, followed by coating deposition, each of these demanding a specific plasma environment, e.g. gas plasma processing for cleaning and metal plasma-enabled coating stages. A schematic of atypical system for complex treatment of workpieces with gas and metal plasma is shown in Fig. 10. By use of holders, the treated parts are positioned in the region of the magnetic trap generated by the powering of the coils. Moreover, different configurations of magnetic coils enable specific shaping of the magnetic fields and thus, shaping of plasmas and modulation of their parameters (Baranov and Romanov 2008).
Fig. 10

Low voltage plasma immersion ion implantation and deposition on rings; production accessories are used to deliver the workpieces to a magnetic trap between the plasma-generating modules: schematic of the system (a) and well-defined physically-separated distributions of metal (blue) or gas (yellow) plasmas over the workpieces (b). A bottle configuration of the magnetic field generated by the additional upper and lower coils enables generation of a plasma discharge, when the workpieces are biased. Similar schematic is proposed by Pillaca et al. (2012) and de Mariano et al. (2014) to deposit DLC films directly over the surface of the stainless steel alloy. When operated using the external plasma-generating modules, the axis of the magnetic bottle is turned by 90°, and the additional upper or lower coils modulate the distribution of the ion current from the plasma injected through the magnetic mirrors located in the plasma generating region. This configuration is widely applied to deposit various vacuum arc and magnetron coatings (see, e.g., Aksenov 2005; Gudmundsson et al. 2012). When using a tube with a coil moved along the tube axis (c), the arc magnetic field of the coil modulates the distribution and scanning of the treated surface with the plasma-generated ions (Levchenko et al. 2004b)

Deposition of coatings on surfaces of small parts and cutting tools is another important application of plasma processing technology. The following setups (Figs. 11, 12) allow for the treatment of the tools with gas and metal plasmas, where the plasma is produced with the use of the substrate, or by means of an external plasma-generating module. While using the substrate, DC or pulsed bias is supplied to the workpieces in presence of arc configuration of the magnetic field, thus generating plasma in the magnetic trap. In the case of RF bias supply, the known MERIE (Kim et al. 2004; Vasenkov and Kushner 2004) schematic is obtained by powering the centre coil or the set of peripheral coils under the substrate, which is useful for heating and etching with gas ions. Both gas and metal ions can be effectively delivered from the external plasma-generating module by use of the magnetic fields, as was shown in experiments with arc or ICP plasma to grow wear-proof TiN (Lang 2012) and Al2O3 coatings (Li et al. 2011). When ICP external plasma-generating module is coupled with magnetized capacitive discharge generated by use of RF driven substrate, the schematic can be used to enhance the performance of RF capacitive discharges for semiconductor processing, similar to the approach proposed by Carter et al. (2006).
Fig. 11

Deposition of vacuum arc coatings on cutting tools by use of PIII&D technique combined with magnetic modulation of plasma parameters: a drills or carbide cutting inserts; b schematics of the flow of plasma-generated ions when treating with gas ions (yellow) and metal ions (blue). Typical operation modes are (i) ion energy Ei = 1200 eV, density of gas ion current Jg = 20 A/м2, background gas pressure P = 1 Pa for ion cleaning and heating; (ii) ion energy Ei = 250 eV, density of metal ion current Jm = 40 A/м2, background gas pressure P = 0.1 Pa for ion deposition of the wear proof coatings

Fig. 12

Deposition of magnetron coatings onto drills, reamers, taps. Magnetic arc trap is used to generate gas (nitrogen) plasma and sputter the metal atoms from a surface of the hollow cathode; the relative flow of the gas and metal ions is determined by separate powering the magnetic coil and the cathode; an additional circuit for modulation of plasma parameters is provided by biasing the accessories with attached cutting tools: a schematic of a setup and b distributions of gas ions (yellow) along the cathode surface when powering the disc part of the cathode, and directed flow of metal ions from the cathode when powering the whole hollow cathode structure (blue)

Next setup allows for the deposition of magnetron coatings on the workpieces (Fig. 12). The use of magnetic field enhances the setup productivity, as was shown by Rodil et al. (2007) in their investigation of the effects of a magnetic field configuration on the respective delivery of ion and metal species (expressed as a ratio Ji/Ja) and the microstructure of NbN films grown with a variable magnetron system. Within the same setup, with the increased gas pressure it is possible to treat the workpieces with gas ions through the use of the magnetron discharge.

The modification of the above setup can be used for nitriding and carbonitriding of the workpieces, including parts of impeller machines made of titanium alloys, where the use of the magnetic field opens up a wide range of treatment possibilities. Surface nitriding and carbonitriding are among the most-widely used industrial processes to reduce friction and wear in load-bearing materials, from parts of machinery to implantable devices, and they are very suitable to investigate the key challenges characteristic of the plasma-enabled processes with zero rate of a workpiece geometry change.

Without magnetic field, conventional glow discharge is ignited between the grounded chamber walls and negatively biased workpieces, which is the base for the conventional plasma nitriding process. Numerous investigators contributed to the development of the plasma nitriding technique as well as studied the key plasma parameters that underpinned this process (Sharma et al. 2008).

The DC plasma nitriding of 304 austenitic steel was conducted by Wang et al. 2013 at gas (N2:H2 = 24:76) pressures ranged between 80 and 400 Pa and at the temperature of 400 °C for 8 h. The thickness of the nitride layer is significantly improved by using low gas pressure, and the thickest cast layer (51.7 μm) was obtained at gas pressure of 100 Pa. Pellets of AISI 316 stainless steel were nitrided by Alves et al. (2001)in a nitrogen–hydrogen mixture at temperatures 673, 773 and 843 K and pressures of 100 and 500 Pa. C38 carbon steel was also successfully plasma nitrided by Bouanis et al. (2011) using RF nitrogen plasma discharge on non-heated substrates at a pressure of 16.8 Pa, where the thickness of the nitride layer depends on the treatment time and is approximately 14 µm for 10 h when the temperatures are in the range of 350–450 °C.

Another industrially-important API5LX-70 steel used in the transmission lines for transfer and transport of oil and natural gas was plasma nitrided by da Cesconetto et al. (2015) in pulsed DC plasma (up to 1 kV) experiments performed in 10% N2 + 90% H2 gas mixture at temperatures of 410, 440 and 470 °C, and for times of 1, 3 and 5 h. The compound layer with large needle-like structure was formed, increasing the wear resistance. The experiments show the possibility of plasma nitriding of a steel pipe under vacuum 1 up to 10 Torr. The glow-discharge nitriding of AISI 316L steel is also reported by Fossati et al. (2006) at 703 K for times ranging from 0 (cast thickness of 3 µm) to 5 h (10 µm). The current density was 26 A/m2, and the voltage drop was 175 V. In the experiment, the surface hardness of treated samples increases from 220 HK0.1 up to a value of about 1450 HK0.1. The pitting corrosion resistance and the hardness of nitrided samples increase as the treatment time increases.

In addition to the gas flow through the holes, the outcomes of the nitriding can be determined by the use of two-walled active screen nitriding with or without bias to samples. To evaluate role of substrate bias voltages in the nitriding process, experiments on AISI 5140 low alloy steel have been conducted by Li et al. (2010a) by using a hollow cathode discharge when the plasma was generated by a pulsed dc glow discharge, and the screen consisted of two low carbon steel cylinders with different diameter. Under these conditions, some samples were nitrided with substrate bias of − 50 V while other samples were nitrided at floating potential in NH3 atmosphere at pressure of 150 Pa and at temperatures of 450–550 °C for 2, 4, and 6 h. It was also demonstrated that the microstructure of the nitride layer strongly depends on the bias voltage, and the better corrosion resistance with a thicker cast layer is obtained on the biased sample: 1 µm against 2 µm for 2 h treatment, and 5.5 µm against 7 µm for 6 h. The discharge current and voltage between the two-cylinder active screen and the anode (chamber walls) was 4–6 A and 500–670 V.

In another study, active plasma nitriding at anodic and cathodic potential with two cylinders was investigated (Li et al. 2010b). The cylinders made up a hollow cathode in a discharge system for nitriding of 42CrMo low alloy steels. The difference in diameter of the two hollow cathode cylinders is about 8–10 mm to maintain strong discharge between them. Using the hollow cathode effect that exists between two cylinders, the temperature of the working space was easily increased to the required value. While the discharge current and voltage between the two-cylinder active screen and the anode were the same, the working pressure was maintained at 500–600 Pa, and the process was performed at temperatures of 450 and 540 °C for 4 h. While the thickness, phase composition and hardness profile of cast layer does not depend on the potential, the disadvantages intrinsic to general plasma nitriding are avoided.

Active plasma nitriding at floating and cathodic potentials was conducted by Shen et al. (2013), who investigated the utilization of controllable hollow cathode effect for heating. In this study, the hollow cathode effect between two coaxial cylindrical cathodes arises between the cylinders in a range of pressures of 100–3000 Pa. AISI 304 steel was efficiently heated and nitrided at 420–500 °C; still, precipitation of chromium nitride was present at 520 °C. The operating pressure was 400–420 Pa, and the voltage applied on the cathodic cylinders was 600–700 V.

Another interesting solution to use the hollow cathode effect was proposed by Benda et al. (1997) who describe an auxiliary cathode with holes of diameter B and depth d. The auxiliary cathode intensifies the discharge and strongly sputters the cathode material. The sputtered material is transferred onto the surface of nitrided steel sample, thus the nitrided substrates can be doped with Mo, Cr, Ti, V, etc. The nitriding was carried out at a pressure of 1067 Pa and a temperature of 550 °C for a time of 2.5 h.

To enhance the effect of doping of the treated workpieces with additives, deposition of CN-based films for surface modification of polymers was investigated by Crespi et al. (2011). In this study, pure graphite was used as the grid material to act both as an active screen and as a source of carbon atoms to grow the CN layer with columnar nanostructured morphology. As for the transport of the species sputtered from the active screen, the sputtering and redeposition theory is the most accepted. From the study conducted by Gallo and Dong (2010), it is concluded that the sputtered material was transferred from the active screen, and deposited onto the treated surfaces, and it plays an important role of in the hardening effect.

Active screen plasma nitriding was introduced by Yazdani et al. (2011) as a novel approach for deposition of nanosized titanium nitride. H11 tool steel samples were placed in a titanium screen with 0.8 mm thickness and plasma nitrided in a pulsed DC reactor with 60% duty cycle, 10 kHz frequency, 500 ± 100 Pa pressure, and three different gas mixtures of H2/N2 = 3, 1, 1/3 were supplied at a temperature of 550 °C. Most of the coatings were less than 1 micron thick, but by increasing the nitriding period both the coatings thickness and the grain size increase.

Other notable attempts made to improve the quality and efficacy of the nitriding process include pre-nitriding shot peening conducted by Manfridini et al. (2014). In their work, triode plasma nitriding was applied to Ti-stabilized interstitial free steels. The measurements indicated that the shot peening pre-treatment did not have any beneficial effect on nitriding kinetics. In contrast, the application of a magnetic field in the nitriding process strongly affects the substrate properties and the nitriding kinetics across a wide range of the magnetic field strengths.

Kovaci et al. (2017) propose a novel pre-magnetization process, which enables easy diffusion of nitrogen during plasma nitriding of AISI 4140 steel. Before nitriding, magnetic fields with intensities of 0.15 and 0.25 T were applied to the samples. After the pre-magnetization, the untreated and pre-magnetized samples were plasma nitrided for 4 h in a gas mixture of 50% N2–50% H2 under a pressure of 500 Pa at temperatures of 500 and 600 °C. The thickness of the modified layer increased with pre-magnetization process, and higher surface hardness was obtained, i.e. the initial hardness was 200–220 HV0.1, and is increased up to 1380–1400 HV0.1 for the layer thickness up to 20 µm.

The effect of the magnetic field on the nitriding phase was reported by Shigarev and Dmitrieva (1978), when they described the brief gas nitriding conducted in various gas mixtures. According to their findings, the magnetic field was found to accelerate gas nitriding in pure ammonia by a factor of 3–4, eliminate brittleness of the nitrided ease, increase the wear resistance and seizing resistance of the case, and increase the fatigue strength. The nitriding process could be accelerated to the maximum with a magnetic field strength of 2.5–3 mT induced at frequency of 50 Hz. For strong magnetic fields (up to 12 T) imposed by the superconductor magnet located in the middle part of the tube furnace, the influence on the nitriding was investigated by Tong et al. (2010) in a mixture of ammonia and hydrogen. It was found that high magnetic field shifts the equilibrium of nitriding reaction.

For less strong magnetic fields, the results were obtained by Vafin et al. (2012) who placed a planar magnetron with magnetic field of 30 mT in a vacuum chamber to nitride tool steels in a mixture of nitrogen, argon, and acetylene (75% N2 + 20% Ar + 5% C2H2) at the pressure varied in the range of 5–200 Pa and the temperature of 490–510 °C. The thickness of the nitride layer with the application of the magnetic field during the nitriding for 4 h was estimated to be at 10 and 80 μm, and the microhardness of 2100 and 1950 HV at a load of 0.49 N is obtained for different tool materials.

From the performed analysis of the literature, the main facilities of the setup for plasma nitriding can be deduced. Evidently, the setup should be provided with an assembly of the substrate and a two-walled hollow cathode structure made of coaxial cylinders with a possibility to mount a lid provided with a holes to manage the gas flow to the treated workpieces mounted inside the assembly. In addition, the setup should be provided with independent power supplies to feed the hollow cathode and the workpieces with electrical power in DC or pulsed mode, as well as the magnetic coil installed under the substrate to affect the current–voltage relations of the discharge. As for the typical parameters of the power supply, working gas mixture and pressure, and substrate temperature, these can be selected from the above discussion depending on the material of the workpieces and intended modification. A possible schematic of the setup is shown in Fig. 13.
Fig. 13

Plasma nitriding of workpieces in presence of magnetic field in pulsed or DC modes of the substrate bias where the arc magnetic field is used to ignite the magnetron discharge and to modulate the ion energy at a given ion current density: a Schematic of a setup and b distribution of the ion current density. Presence of the magnetic field modulates the breakdown voltage thus determining the ion energy and sputtering processes; very intense flows of ions at low ion energy can be obtained. Typical operation modes are (i) ion energy Ei = 1200 eV, density of gas ion current Jg = 10 A/м2, background gas pressure P = 1 Pa for ion cleaning and heating under the environment of magnetron discharge; (ii) ion energy Ei = 250–300 eV, density of gas ion current Jg = 20 A/m2, background gas pressure P = 10 Pa for preferential treatment with gas ions (nitriding)

Modification of the surface by formation of hardened layer proved its usefulness in many applications. Still, the advantages of the technology are yet to be fully exhausted. The technique is very beneficial for the following coating deposition in a treatment that is called a duplex treatment. For instance, plasma nitriding can be followed by a closed field unbalanced magnetron sputter deposition, as demonstrated by Ge et al. (2013) who used it to increase the adhesion of CrTiAlN coatings to H13 steels and wear resistance for duplex-treated coatings. H13 steel samples were heated at 520 °C for 7 h in the N2/H2 = 1/4 mixture under the pressure of 260 Pa to form a nitride layer with a thickness up to 200 µm, and the surface hardness is increased from 550 to 1300 HV. Prior to the coating deposition, cleaning of the workpiece was performed at a bias of 500 V for 20 min. The CrTiAlN coating of 4 μm in thickness was deposited for 1 h with four unbalanced magnetron sat bias potential of − 70 V. After the Rockwell indentation of workpieces, the number of the cracks near the indentation decreased greatly for nitrided samples compared to the samples without the nitriding.

Grigore et al. (2009) investigated the plasma nitriding treatment and deposition of nitrogen alloyed austenitic stainless steel coatings by a magnetron sputtering assisted by a high voltage pulse discharge. Before deposition, a DC sputter etching was applied for 15 min. Deposition was conducted for 2 h at the substrate temperature of about 330 °C, gas pressure of 6.6 × 10−1 Pa, and a DC negative bias of 80 V applied on the substrate. High voltage pulses of 40 kV amplitude and 20 ms width were applied at a frequency of 25 Hz on the substrate simultaneously with the DC bias. As for the plasma nitriding, it was performed for 15 h in a N2:H2 = 25:75 mixture at a pressure of 400 Pa and a temperature of about 430 °C. The deposition performed in the nitrogen environment resulted in increase in microhardness from 700 to 1400 HV0.1.

CrN and CrNiN coatings were deposited by Jin et al. (2017) on the non-nitrided and nitrided Ti6Al4V alloys with the unbalanced magnetron ion plating. For the nitriding, the gas mixture of N2:Ar = 1:1 was supplied for 3 h with the pressure of 45 Pa at temperature of 850 °C. Prior to the deposition, the substrates were sputter cleaned in argon plasma for 30 min at a bias voltage of − 450 V. Then a thin Cr metallic layer was deposited for 6 min to enhance adhesion, and a CrN layer was deposited as an interfacial layer followed by deposition of a CrN or CrNiN layer (2 × 10−6 Torr, − 80 V, 2 h for each step). The indentation tests indicated that the hardness was increased by 8 times, and the wear volume and wear rate decreased by 50 times.

The biocompatibility properties of the specimens made of AISI 316L stainless steel were investigated after the duplex plasma treatment in a research conducted by Kao et al. (2016). First, the specimens were nitrided at a temperature of 350 °C and a pressure of 100 Pa in a nitrogen–hydrogen gas mixture (N2:H2 = 1:5) for 20 min, then the temperature was then increased to 390 °C and nitriding was performed for 23 h under a pressure of 50 Pa. Prior to the deposition process, the specimens were ion cleaned for 20 min in argon plasma at a DC voltage of 340 V and a pulse frequency of 150 kHz. Then a pure Ti interlayer was deposited to serve as an adhesion layer, and Ti–C:H coatings were then deposited for 100 min at a pressure of 10−3 Pa, bias of − 42 V for both operations. The hardness was increased from 3 up to 18 GPa for the thickness up to 4 µm, the wear depth and coefficient of friction are decreased by 6–7 times and 3–8 times, respectively. The results showed that plasma nitriding followed by Ti–C:H coating deposition is a powerful tool to improve the biocompatibility performance of AISI 316L stainless steel.

Review of the literature suggests that the prospective setup for duplex treatment should combine the facilities for plasma treatment in gas discharge modulated by the magnetic field, and facilities for the coating deposition by use of external (ESS) or incorporated substrate scheme (ISS). To create such an environment, all of the means to conduct the nitriding under various operating parameters discussed above should be included in this system. One of the possible configurations is presented in Fig. 14.
Fig. 14

Plasma nitriding of small-sized gears, bearings, lobes at the direct and inverse polarity, followed by deposition of wear-proof coatings at generation of plasma by use of external plasma-generating modules (duplex treatment): a Plasma setup where the workpieces fixed on the holder that is a cathode are biased negatively relative to the HCS that is an anode (nitriding at direct polarity), or HCS is biased negatively relative to the workpieces on the holder that is an anode (nitriding at inverse polarity), or deposition of magnetron coatings by use of HCS; b Distribution of the ion current density on the workpieces at the inverse polarity nitriding (red) and deposition of coating by use of either magnetron (yellow) or external plasma source (blue). Applied magnetic field governs the current–voltage relations thus the ion energy and flow. Typical operation modes are (i) ion energy Ei = 300–350 eV, density of gas ion current Jg = 30 A/м2, background gas pressure P = 10–100 Pa for the nitriding; (ii) ion energy Ei = 500 eV, density of gas ion current Jg = 20 A/м2, background gas pressure P = 1 Pa for deposition of magnetron coatings; (iii) ion energy Ei = 250 eV, density of metal ion current Jm = 40 A/м2, background gas pressure P = 0.1 Pa for vacuum arc ion deposition (external source)

If the nitriding or deposition is carried out by use of the external plasma-generating module, the setup can take form of that shown in Fig. 14a, where the upper lead with orifices is removed to make an access point for the gas or metal plasma generated in the external plasma-generating module. The centre coil under the substrate is used to generate the plasma in ISS mode, or to focus the flow of plasma in ESS mode. Side coils can also be present as in Fig. 11 to modulate the distribution of the ion current in the ESS mode or to generate the plasma rings in ISS mode. To apply the ion flux extracted from the external plasma source, the setup should be modified as shown in Fig. 14b.

Another approach was used by Kharkov et al. (2017). to modify the properties of titanium alloy samples. In contrast to the aforementioned approach, two external plasma-generating modules (ICP and magnetron) were applied to conduct the duplex treatment. Prior to the nitriding, the samples were cleaned for 10 min in argon plasma at 1 Pa and RF power of 1.5 kW at the bias of − 300 V. Then plasma nitriding was performed in the ICP facility at a pressure of 0.33 Pa (N2:Ar = 1:2.3) or pure N2, 1.5 kW of RF power, negative bias of 100–300 V, temperature < 900 °C, and processing time of 30–240 min for N2:Ar mixture. Bias of 1500 V, temperature < 700 °C, and processing time of 120 min for N2 were used. The nitriding was followed by magnetron deposition of TiN coatings, prior to which the samples were sputter-cleaned for 10 min in glow discharge in argon at 1.7 Pa at − 1.5 kV. The magnetron deposition was conducted at gas pressure of 0.62 Pa, discharge power of 300 W, and bias voltage of − 300 V in gas mixture of N2:Ar (4.50:0.30 or 4.64:0.16) for 60 or 120 min, with the resultant coating thickness of 3 or 6 µm. While the initial hardness is 360 HV0.1, the hardness of TiN coatings deposited after the nitriding procedures reaches the value of 1600–1700 HV0.1 for the hardened layer of up to 25 µm.

Duplex treatment (nitriding and vacuum arc deposition) system enhanced by the use of magnetic field during ion nitriding process in glow discharge was reported by Ramazanov et al. (2014). The nitriding was conducted for 4 h in glow discharge with magnetic field in a mixture of gases (N2, Ar, C2H2) at P = 50 Pa, and the temperature not exceeding 500 °C. This was followed by ion cleaning for 5–20 min at 1 kV in high-density argon plasma created by magnetic field at the substrate temperature not exceeding 350 °C. After that, TiN–TiAlN coating was deposited for 3 h at the substrate voltage of 200 V at a pressure of 0.11–0.13 Pa. The hardness of the surface layer was increased from the initial 4.2–19.7 GPa after the ion nitriding, then to 33.2 GPa after the ion nitriding followed by Ti–TiN coating, and to 48.3 GPa after the ion nitriding followed by Ti–TiAlN coating. By analyzing the results, it was assumed that the magnetic field increases the discharge current and thus promotes growth at the maximum surface temperature of 500 °C.

Titanium specimens were also modified by Tong et al. (2011) with DC plasma nitriding and arc deposition of TiN film. At first, the specimens were etched by argon sputtering for 15 min at a pressure of 2.5 Pa. Then DC plasma nitriding was carried out at 700 °C for 2 h in atmosphere of ammonia, which was followed by deposition of TiN films at 75 A, 100 V for 1 h at the nitrogen pressure of 0.15 Pa and the substrate temperature of 300 °C. While the initial microhardness of titanium was about 200 HV, the value increased to about 480 and 770 HV after depositing TiN coating and duplex treatment, respectively. The friction coefficient of the untreated titanium was 0.571, and the value decreased to 0.393 after TiN coating, yet the lowest friction coefficient 0.272 was obtained after the duplex treatment. The wear depth value was about 5 μm for the untreated samples, 3.3 μm for the TiN coated samples, and 1.6 μm for the duplex treated samples. In addition to the results, the nitrided layer exhibited excellent properties as a supporting layer for the hard TiN coating.

In the setup shown in Fig. 15, the workpiece with large inner diameter is arranged around the cage which is the hollow cathode structure, to be treated with gas or metal plasma generated in HCS magnetron.
Fig. 15

Plasma nitriding of large-diameter internal toothing in magnetron discharge: a schematic of plasma setup and b distribution of ion current density to the gear surface. A magnetron discharge is ignited in the internal space of the gear by use of the magnetic coil

Another configuration of the setup for duplex treatment with the external plasma-generating module is introduced in Fig. 16. It is based on the schematic proposed by Wei et al. (2004) for magnetic field enhanced plasma (MFEP) deposition of various coatings on inner diameter of tubes or pipes. The technique utilizes a magnetic field to generate hollow cathode glow discharge inside a long tube and enhance the plasma production, where the working gas is introduced at the entrance of the tube. Using this method, diamond-like and SiC films have been deposited on the inner walls of tubes with small diameters and high aspect ratios (0.9–2.5 cm in diameter, up to 71 cm long) at the uniform coating thickness.
Fig. 16

Duplex treatment of tubes with plasma nitriding of internal surfaces in magnetron discharge and deposition of coating by use of external plasma-generating module: a schematic of plasma setup and b distribution of ion current density to surface

For the benefit of the reader, we have summarized a number of the aforementioned examples of plasma applications in two overview tables (Tables 2, 3).
Table 2

Characteristics of a plasma nitriding treatment

Discharge type

Key plasma and process parameters

DC glow

304 austenitic steel—gas (N2:H2 = 24:76) pressures between 80 and 400 Pa and at the temperature of 400 °C, for 8 h (Wang et al. 2013)

AISI 316 stainless steel—nitrogen–hydrogen mixture at temperatures 673, 773 and 843 K and pressures of 100 and 500 Pa (Alves et al. 2001)

AISI 316L steel—703 K 5 h; the current density was 26 A/m2, and the voltage drop was 175 V (Fossati et al. 2006)

RF plasma

C38 carbon steel—pressure of 16.8 Pa for 10 h when the temperatures are in the range of 350–450 °C (Bouanis et al. 2011)

Pulsed DC plasma (up to 1 kV)

API5LX-70 steel—10% N2 + 90% H2 gas mixture at temperatures of 410, 440 and 470 °C, and for times of 1, 3 and 5 h. The experiments show the possibility of plasma nitriding of a steel pipe under vacuum 1 up to 10 Torr (da Cesconetto et al. 2015)

Hollow cathode discharge

AISI 5140 low alloy steel—substrate bias of − 50 V or at floating potential in NH3 atmosphere at pressure of 150 Pa and at temperatures of 450–550 °C for 2–6 h. The discharge current and voltage between the two-cylinder active screen and the anode (chamber walls) was 4–6A and 500–670 V (Li et al. 2010)

42CrMo low alloy steels—the difference in diameter of the two hollow cathode cylinders is about 8–10 mm, the pressure was 500–600 Pa, and the temperatures are of 450 and 540 °C for 4 h (Li et al. 2010)

AISI 304 steel—discharge between the cylinders in a range of pressures of 100–3000 Pa at 420–500 °C; The operating pressure was 400–420 Pa, and the voltage applied on the cathodic cylinders was 600–700 V (Shen et al. 2013)

The auxiliary cathode strongly sputters the cathode material, which is transferred onto the surface of nitrided steel sample, thus the nitrided substrates can be doped with Mo, Cr, Ti, V, etc. The nitriding was carried out at a pressure of 1067 Pa and a temperature of 550 °C for a time of 2.5 h (Benda et al. 1997)

Deposition of nanosized titanium nitride on H11 tool steel samples placed in a titanium screen with 0.8 mm thickness and plasma nitrided in a pulsed DC reactor with 60% duty cycle, 10 kHz frequency, 500 ± 100 Pa pressure, and three different gas mixtures of H2/N2 = 3, 1, 1/3 were supplied at a temperature of 550 °C (Yazdani et al. 2011)

Glow discharge and magnetron deposition

Alloying of nitrided layer. Magnetic field of 5–30 mT in a mixture of nitrogen, argon, and acetylene (75% N2 + 20% Ar + 5% C2H2, e.g.) at the pressure varied in the range of 5–200 Pa and the temperature of 490–510 °C. Time of nitriding of different tool materials is of 4 h (Vafin et al. 2012)

Table 3

Characteristics of a plasma duplex treatment

Discharge type

Process and plasma parameters, effect on the treated material

Glow discharge and magnetron deposition

Increase in adhesion wear resistance of CrTiAlN coatings to H13 steels. The samples were heated at 520 °C for 7 h in the N2/H2 = 1/4 mixture under the pressure of 260 Pa. Prior to the coating deposition, cleaning of the workpiece was performed at a bias of 500 V for 20 min. The CrTiAlN coating of 4 μm in thickness was deposited for 1 h with four unbalanced magnetron sat bias potential of − 70 V (Ge et al. 2013)

Glow discharge and magnetron deposition assisted by a high voltage pulse discharge

Nitriding was performed for 15 h in a N2:H2 = 25:75 mixture at a pressure of 400 Pa and a temperature of about 430 °C. Before deposition, a DC sputter etching was applied for 15 min. Deposition of nitrogen alloyed austenitic stainless steel coatings was conducted for 2 h at the substrate temperature of about 330 °C, gas pressure of 6.6 × 10−1 Pa, and a DC negative bias of 80 V applied on the substrate. High voltage pulses of 40 kV amplitude and 20 ms width were applied at a frequency of 25 Hz on the substrate simultaneously with the DC bias (Grigore et al. 2009)

CrN and CrNiN coatings were deposited on Ti6Al4V alloys with the unbalanced magnetron ion plating. For the nitriding, the gas mixture of N2:Ar = 1:1 was supplied for 3 h with the pressure of 45 Pa at temperature of 850 °C. Prior to the deposition, the substrates were sputter cleaned in argon plasma for 30 min at a bias voltage of − 450 V. Then a thin Cr metallic layer was deposited for 6 min to enhance adhesion, and a CrN layer was deposited as an interfacial layer followed by deposition of a CrN or CrNiN layer (2 × 10−6 Torr, − 80 V, 2 h for each step) (Jin et al. 2017)

AISI 316L stainless steel specimens were nitrided at a temperature of 350 °C and a pressure of 100 Pa in a nitrogen–hydrogen gas mixture (N2:H2 = 1:5) for 20 min, then the temperature was then increased to 390 °C and nitriding was performed for 23 h under a pressure of 50 Pa. Prior to the deposition process, the specimens were ion cleaned for 20 min in argon plasma at a DC voltage of 340 V and a pulse frequency of 150 kHz. Then a pure Ti interlayer was deposited to serve as an adhesion layer, and Ti–C:H coatings were then deposited for 100 min at a pressure of 10−3 Pa, bias of − 42 V for both operations (Kao et al. 2016)

Two external plasma-generating modules (ICP and magnetron)

Titanium alloy samples. Prior to the nitriding, the samples were cleaned for 10 min in argon plasma at 1 Pa and RF power of 1.5 kW at the bias of − 300 V. Then plasma nitriding was performed in the ICP facility at a pressure of 0.33 Pa (N2:Ar = 1:2.3 or pure N2, 1.5 kW of RF power, negative bias of 100–300 V, temperature < 900 °C, and processing time of 30–240 min for N2:Ar mixture. Bias of 1500 V, temperature < 700 °C, and processing time of 120 min for N2 were used. The nitriding was followed by magnetron deposition of TiN coatings, prior to which the samples were sputter-cleaned for 10 min in glow discharge in argon at 1.7 Pa at − 1.5 kV. The magnetron deposition was conducted at gas pressure of 0.62 Pa, discharge power of 300 W, and bias voltage of − 300 V in gas mixture of N2:Ar (4.50:0.30 or 4.64:0.16) for 60 or 120 min (Kharkov et al. 2017)

Glow discharge nitriding and vacuum arc deposition

The nitriding was conducted for 4 h in glow discharge with magnetic field in a mixture of gases (N2, Ar, C2H2) at P = 50 Pa, and the temperature not exceeding 500 °C. This was followed by ion cleaning for 5–20 min at 1 kV in high-density argon plasma created by magnetic field at the substrate temperature not exceeding 350 °C. After that, TiN–TiAlN coating was deposited for 3 h at the substrate voltage of 200 V at a pressure of 0.11–0.13 Pa. It was assumed that the magnetic field increases the discharge current and thus promotes growth at the maximum surface temperature of 500 °C (Ramazanov et al. 2014)

Titanium specimens were etched by argon sputtering for 15 min at a pressure of 2.5 Pa. Then DC plasma nitriding was carried out at 700 °C for 2 h in atmosphere of ammonia, which was followed by deposition of TiN films at 75 A, 100 V for 1 h at the nitrogen pressure of 0.15 Pa and the substrate temperature of 300 °C (Tong et al. 2011)

5 Perspectives and trends

Over more than a century of plasma technology development, a large number of plasma setups has been developed, evolving from the very simple apparatus designed for the cathodic sputtering application to a versatile, sophisticated setup able to support a broad range of research activity, such as Magnetized Plasma Interaction Experiment (MAGPIE) helicon plasma source (Blackwell et al. 2012) or a neutral loop discharges in the VINETA device (Fig. 17). Yet, there are a number of technological challenges that need to be overcome to enable further development and application of plasma-based processing systems. The literature is rich with various engineering decisions that are proposed to address specific processing needs, such as high-resolution shaping of the surface geometry and morphology, enhancement of hardness and wear resistance, changing the optical properties and surface biocompatibility (Levchenko et al. 2016b), production of complex metamaterials (Levchenko et al. 2016c), as well as manufacturing demands, i.e. to enhance the productivity of these setups.
Fig. 17

Top panel: diagram of the VINETA experiment. Additionally, the total magnetic field (color coded) and its field lines (blue) are shown. Bottom panel: diagram of the 3 RF antenna configurations in the plasma vessel, including the magnetic field coils and the resulting separatrices (blue). Reprinted with permission from von Stechowa et al. (2012)

The development of high-productivity plasma-based setups to meet the requirement of modern industry is limited by the efficacy with which the energy and flow of ions can be modulated. Although the use of energy gained from the electric field by electrons is the most widespread method to generate plasma in the plasma setups, the directed transport of thus-generated plasma species is a much more complex problem. Mutual alignment of workpieces and plasma-generating modules, well-defined distribution of the flow of ions along the treated surface, time sequence of the treatment and necessity to separate the treated areas, and other factors contribute to the complexity of the setup design. For the vacuum plasma technology, magnetic fields of proper configuration are believed to be the most universal tool to modulate the energy as well as the distribution of the plasma-generated ions on the treated surface. Still, the configuration of the magnetic fields depends greatly on the specifics of the technological operation of the plasma technology. In this review, a wide range of commonly used technologies for surface modification and deposition have been critically reviewed with a purpose to deduce the most essential physical processes and technical features of the corresponding plasma setups, and to outline the possible engineering solutions for the future plasma setups.

Presently, complex multistep processes are widely utilized in the modern technologies, with the techniques of quite different nature and chemistry involved at various technological steps. Cluster systems that allow for the efficient sample manipulation in the course of production or research activities is one such approach of significant importance. Moreover, the complexity and sensitivity of modern materials and metamaterials to their environment demand careful handling to prevent cross-contamination and other damage to the specimens. This is not a trivial task for large, general-purpose plasma-processing setups and chambers, whereas multifunctional cluster-like systems present a viable solution for a number of specific technologies.

Figures 18 and 19 present two possible variants of a transformable multifunctional cluster-like setup that have the potential to address the aforementioned challenges. Furthermore, careful selection of the materials and specific methods is needed to appropriately designate the specific technique.
Fig. 18

Schematic of a clustered setup proposed for the synthesis of microscale structures on the surface of the substrate. (1) treatment with gas ions (doping, modification, heating) extracted from the plasma over a wide range of applied power, combined with the PIII treatment provided by the substrate biasing and an independent control of the ion flux distribution with a set of magnetic coils (Baranov et al. 2009b; Liu 2013; Pillaca et al. 2012); (2, 3) a duplex treatment with gas ions generated by the ionization of the background gas, and metal atoms sputtered from the cathode via the ion bombardment; in this case, independent control of ion-to-gas ratio is ensured by a system of magnetic fields combined with the PIII technique (Baranov et al. 2014a; Boromei et al. 2013); (4) deposition of coatings from various plasma sources in reactive gases; the system of magnetic fields is also combined with the substrate biasing to control the morphology and chemical structure of the coatings (Chen et al. 2013; Baranov et al. 2014b). Panel (2) and micrograph at the bottom of panel (4) are reprinted with permission from Baranov et al. (2017b)

Fig. 19

Schematic of a transformable setup for the synthesis of nanostructures with various dimensionalities: (1) preliminary treatment and generation of nanodots in CCP plasma; (2) deposition of metal thin film to generate nanoislands which are the seeds for the sequential growth of nanotubes as shown in (3) or vertical graphene structures (Fang et al. 2014) in ICP or an atmospheric pressure plasma reactor (Altaweel et al. 2014) are shown in (4); (5) growth of the oxide nanowires in ICP plasma reactor(Filipič et al. 2015). Reprinted with permission from Ostrikov et al. (2010 panels 1 and 5); Yick et al. (2015 panels 2 and 3)

By the analysis of the plasma setups (Fig. 18), three main steps of plasma processing have been identified as involved in the technologies with respect to the rate V Σ of the geometry change: (i) sputtering and reactive etching, when the surface layers are removed (V Σ  < 0); (ii) modification and alloying, when the change of the chemical composition or morphology of the existing surface is the main goal of the processing (V Σ  ≈ 0); and (iii) deposition of coating (including the passivation of a surface), when the change in the surface properties is obtained by creation of a new layer on the surface (V Σ  > 0). Both the metal and gas ions are used to carry out the necessary reactions, and two kind of interactions are applied to generate the gas or metal plasma, namely, the incorporated substrate scheme (ISS, when a substrate with the workpieces is a part of the plasma generating circuit), and the external substrate scheme (ESS, when the substrate is exposed to gas or metal plasma extracted from an external source) (Baranov et al. 2017b).

When sputtering and etching, both the ISS and ESS schemes can be applied to the biased substrate. The simplest way is to treat the substrate in a glow discharge (ISS). However, the absence of the means to selectively modulate plasma parameters is an essential disadvantage of this approach, which can be overcome by applying the magnetic field in the discharge area to split the current–voltage relations for a given gas pressure. Another very promising approach is applying the double-frequency capacitive coupled plasma, where the distribution of energies and flows of plasma-generated species can be modulated independently. To optimize the parameters of the delivery of plasma species from the external plasma source (ESS), inductively coupled plasma (ICP) sources with magnetic modules are the most suitable (helicon and neutral loop discharge). To carry out the plasma nitriding, the double-walled hollow cathode structure (HCS) is suitable for pure nitriding but has some disadvantages for duplex treatment because of re-deposition of sputtered atoms due to high gas pressure. The energy is also low due to collision sheath. The magnetic field applied within the hollow cathode area allows decreasing the pressure, thus changing the sheath to the collisionless and increasing greatly the mean free path for the sputtered atoms. In a case of ISS, application of the HCS with changeable geometry coupled with an electromagnet under the substrate with changeable configuration of magnetic core is desirable. In the case of ESS (Baranov et al. 2017b), the guiding magnetic field with desired configuration of magnetic traps for plasma electrons should be used to shape the distribution.

The setup proposed in Fig. 19 comprises a set of process-oriented chambers for the purposes of nanotechnological synthesis. Activation and functionalization of the surfaces is conducted in chamber (1); then, substrates may be transferred to chamber (2) to deposit e.g. layers of thin metal films required to catalyze the subsequent growth of nanostructures and metamaterials. After deposition, the substrates could be exposed to gas plasma in chamber (3) to grow e.g. carbon nanotubes, or in chamber (4) to grow vertical graphene structures, or in chamber (5) to grow arrays of oxide nanowires.

6 Concluding remarks

Modern industry continually places more exacting and stringent demands on the quality of plasma processed workpieces and the cost associated with such processing. Complex interaction between various factors within each plasma-enabled setup, and concomitant efforts to optimize and reduce the cost of processing within individual setups designed for a specified case, have led to the development of a large variety of plasma setups, technological processes, methods for their modulation and enhancement. Yet, at the same time, such specialization of individual setups limited their flexibility and restricted possible outcomes that can be achieved using these systems, e.g. characteristics of the processed surface layer.

Using an example schematic of a prospective plasma setup, we put forward a set of guiding principles and key physical considerations for the development of flexible, rapidly-adjustable and economical processing setups capable of conducting a large variety of industrially-relevant operations of the plasma processing.

By analyzing the existing technologies of plasma processing, the essential technical features of the existing plasma setups were determined with respect to the plasma generation and modulation with a purpose of a surface sputtering, reactive etching, modification, alloying, and coating deposition (Bazaka et al. 2016; Hundt et al. 2011). A possibility of these distinct techniques in the same plasma setup to enhance its effectiveness in terms of productivity, energy and material consumption, and processing uniformity was considered.

Notes

Acknowledgements

This work was supported by EDB (OSTIn) and National Research Foundation, Singapore. O.B. acknowledges the support from European Union’s Horizon 2020 research and innovation programme under Grant agreement no. 766894. I. L. acknowledges the support from the School of Chemistry, Physics and Mechanical Engineering, Science and Engineering Faculty, Queensland University of Technology. K.B. acknowledges the funding from the Australian Research Council (DE130101550, DP160103116, DP180101254).

References

  1. R. Abdolvand, F. Ayazi, Sens. Actuator A. A 144, 109–116 (2008)CrossRefGoogle Scholar
  2. I.I. Aksenov, in Vacuum Arc in Erosive Plasma Sources (In Russian) (Kharkov: National Scientific Centre “Kharkov Institute of Physics and Technology”, 2005)Google Scholar
  3. A. Altaweel, G. Filipič, T. Gries, T. Belmonte, J. Cryst. Growth 407, 17–24 (2014)ADSCrossRefGoogle Scholar
  4. C. Alves Jr., E.F. da Silva, A.E. Martinelli, Surf. Coat. Technol. 139, 1–5 (2001)CrossRefGoogle Scholar
  5. H. Amemiya, K. Ogawa, J. Phys. D Appl. Phys. 30, 879 (1997)ADSCrossRefGoogle Scholar
  6. A. Anders, Handbook of Plasma Immersion Ion Implantation and Deposition (Wiley, New York, 2000)Google Scholar
  7. A. Anders, Cathodic Arcs: From Fractal Spots to Energetic Condensation (Springer, New York, 2008)CrossRefGoogle Scholar
  8. A. Anders, Surf. Coat. Technol. 257, 308–325 (2014)CrossRefGoogle Scholar
  9. A. Anders, J. Brown, IEEE Trans. Plasma Sci. 36(11), 2528–25291 (2011)ADSCrossRefGoogle Scholar
  10. K.A. Atlasov, P. Gallo, A. Rudra, B. Dwir, E. Kapon, J. Vac. Sci. Technol. B 27(5), L21–L24 (2009)CrossRefGoogle Scholar
  11. O. Baranov, M. Romanov, Plasma Process. Polym. 5, 256 (2008)CrossRefGoogle Scholar
  12. O. Baranov, M. Romanov, K. Ostrikov, Phys. Plasmas 16, 053505 (2009a)ADSCrossRefGoogle Scholar
  13. O. Baranov, M. Romanov, K. Ostrikov, Phys. Plasmas 16, 063505 (2009b)ADSCrossRefGoogle Scholar
  14. O. Baranov, M. Romanov, S. Kumar, X. Zhong, K. Ostrikov, J. Appl. Phys. 109(6), 063304 (2011)ADSCrossRefGoogle Scholar
  15. O. Baranov, M. Romanov, J. Fang, U. Cvelbar, K. Ostrikov, J. Appl. Phys. 112(7), 073302 (2012)ADSCrossRefGoogle Scholar
  16. O. Baranov, J. Fang, A. Rider, S. Kumar, K. Ostrikov, I.E.E.E. Trans, Plasma Sci. 41(12), 3640 (2013)CrossRefGoogle Scholar
  17. O. Baranov, X. Zhong, J. Fang, S. Kumar, S. Xu, U. Cvelbar, D. Mariotti, K. Ostrikov, I.E.E.E. Trans, Plasma Sci. 42(10), 2518 (2014a)CrossRefGoogle Scholar
  18. O. Baranov, J. Fang, M. Keidar, X. Lu, U. Cvelbar, K. Ostrikov, I.E.E.E. Trans, Plasma Sci. 42(10), 2464 (2014b)CrossRefGoogle Scholar
  19. O. Baranov, K. Bazaka, H. Kersten, M. Keidar, U. Cvelbar, S. Xu, I. Levchenko, Plasma under control: advanced solutions and perspectives for plasma flux management in material treatment and nanosynthesis. Appl. Phys. Rev. 4, 041302 (2017a)ADSCrossRefGoogle Scholar
  20. O. Baranov, J. Fang, K. Ostrikov, U. Cvelbar, Mater. Chem. Phys. 188, 143–153 (2017b)CrossRefGoogle Scholar
  21. K. Bazaka, M. V. Jacob, K. Ostrikov, Sustainable Life Cycles of Natural-Precursor-Derived Nanocarbons, Chem. Rev. 116, 163–214 (2016)CrossRefGoogle Scholar
  22. M. Benda, J. Vlček, V. Cibulka, J. Musil, J. Vac. Sci. Technol. A 15(5), 2636–2643 (1997)ADSCrossRefGoogle Scholar
  23. M.M.M. Bilek, A. Anders, I.G. Brown, Plasma Sources Sci. Technol. 10, 606 (2001)ADSCrossRefGoogle Scholar
  24. B.D. Blackwell, J.F. Caneses, C.M. Samuell, J. Wach, J. Howard, C. Corr, Plasma Sources Sci. Technol. 21, 055033 (2012)ADSCrossRefGoogle Scholar
  25. J.-P. Boeuf, J. Appl. Phys. 121, 011101 (2017)ADSCrossRefGoogle Scholar
  26. J. Bohlmark, M. Östbye, M. Lattemann, H. Ljungcrantz, T. Rosell, U. Helmersson, Thin Solid Films 515, 1928 (2006)ADSCrossRefGoogle Scholar
  27. I. Boromei, L. Ceschini, A. Marconi, C. Martini, Wear 302(1–2), 899–908 (2013)CrossRefGoogle Scholar
  28. F.Z. Bouanis, F. Bentiss, S. Bellayera, J.B. Vogta, C. Jama, Mater. Chem. Phys. 127, 329–334 (2011)CrossRefGoogle Scholar
  29. J.H. Bowles, D. Duncan, D.N. Walker, W.E. Amatucci, J.A. Antoniades, Rev. Sci. Instrum. 67(2), 455–461 (1996)ADSCrossRefGoogle Scholar
  30. R.L. Boxman, D. Sanders, P. Martin (eds.), Handbook of Vacuum Arc Science & Technology: Fundamentals and Applications (Noyes Publications, Park Ridge, 1995)Google Scholar
  31. NStJ Braithwaite, Plasma Sources Sci. Technol. 9, 517 (2000)ADSCrossRefGoogle Scholar
  32. P. Chabert, N.St.J. Braithwaite, in Physics of Radio-Frequency Plasmas (Cambridge University Press, New York, 2011)Google Scholar
  33. C. Charles, R. Boswell, A. Bish, et al., Direct measurement of axial momentum imparted by an electrothermal radiofrequency plasma micro-thruster. Front. Phys. 4, Article 19 (2016). http://doi.org/10.3389/fphy.2016.00019
  34. M.D. Carter, P.M. Ryan, D. Hoffman, W.S. Lee, D. Buchberger, V. Godyak, J. Appl. Phys. 100, 073305 (2006)ADSCrossRefGoogle Scholar
  35. K.M. Chang, C.C. Kuo, Y.W. Chang, C.G. Chao, T.F. Liu, Surf. Coat. Technol. 254, 313–318 (2014)CrossRefGoogle Scholar
  36. C. Charles, Plasmas for spacecraft propulsion. J. Phys. D Appl. Phys. 42(16), 163001 (2009)ADSCrossRefGoogle Scholar
  37. F. Chen, Introduction to Plasma Physics and Controlled Fusion (Plenum, New York, 1984)CrossRefGoogle Scholar
  38. F.F. Chen, Phys. Plasmas 19, 093509 (2012)ADSCrossRefGoogle Scholar
  39. F.F. Chen, Plasma Sources Sci. Technol. 24, 014001 (2015)ADSCrossRefGoogle Scholar
  40. F.F. Chen, J.P. Chang, Lecture notes on principles of plasma processing (Plenum/Kluwer Publishers, New York, 2002)Google Scholar
  41. F.F. Chen, H. Torreblanca, Phys. Plasmas 16, 057102 (2009)ADSCrossRefGoogle Scholar
  42. S. Chen, D. Luo, G. Zhao, Phys. Proc. 50, 163–168 (2013)ADSCrossRefGoogle Scholar
  43. P.K. Chu, Surf. Coat. Technol. 229, 2–11 (2013)CrossRefGoogle Scholar
  44. A.E. Crespi, M.E.H.M. da Costa, C.A. Figueroa, M.E.R. Dotto, A.P. Kauling, G.V. Soares, I.J.R. Baumvol, C. Giacomelli, Mater. Lett. 65, 2985–2988 (2011)CrossRefGoogle Scholar
  45. M.C.R.L. da Cesconetto, A.R. Franco Jr., E.A. Vieira, Mater. Res. 18(2), 334–340 (2015)CrossRefGoogle Scholar
  46. S.F.M. de Mariano, E.J.D.M. de Pillaca, M. Ueda, R.M. de Oliveira, Surf. Coat. Technol. 256, 47–51 (2014)CrossRefGoogle Scholar
  47. R. Di Mundo, R. d’Agostino, F. Palumbo, ACS Appl. Mater. Interfaces. 6, 17059–17066 (2014)CrossRefGoogle Scholar
  48. P. Diomede, D.J. Economou, J. Appl. Phys. 115, 233302 (2014)ADSCrossRefGoogle Scholar
  49. C.Y. Duluard, P. Ranson, L.E. Pichon, J. Pereira, E.H. Oubensaid, P. Lefaucheux, M. Puechand, R. Dussart, J. Micromech. Microeng. 21, 065015 (2011)ADSCrossRefGoogle Scholar
  50. J. Fang, I. Levchenko, S. Kumar, D. Seo, K. Ostrikov, Sci. Technol. Adv. Mater. 15, 055009 (2014)CrossRefGoogle Scholar
  51. F. Ferreira, J.C. Oliveira, A. Cavaleiro, Surf. Coat. Technol. 291, 365–375 (2016)CrossRefGoogle Scholar
  52. F. Fietzke, H. Morgner, S. Gunther, Plasma Process. Polym. 6, S242–S246 (2009)CrossRefGoogle Scholar
  53. G. Filipič, O. Baranov, M. Mozetič, K. Ostrikov, U. Cvelbar, Phys. Plasmas 21, 113506 (2014)CrossRefGoogle Scholar
  54. G. Filipič, O. Baranov, M. Mozetič, U. Cvelbar, J. Appl. Phys. 17, 043304 (2015)ADSCrossRefGoogle Scholar
  55. A. Fossati, F. Borgioli, E. Galvanetto, T. Bacci, Surf. Coat. Technol. 200, 3511–3517 (2006)CrossRefGoogle Scholar
  56. T. Furukawa, K. Takizawa, D. Kuwahara, S. Shinohara, Phys. Plasmas 24, 043505 (2017)ADSCrossRefGoogle Scholar
  57. S.C. Gallo, H. Dong, Vacuum 84, 321–325 (2010)CrossRefGoogle Scholar
  58. S.C. Gallo, A.E. Crespi, F. Cemin, C.A. Figueroa, I.J.R. Baumvol, IEEE Trans. Plasma Sci. 39(11), 3028–3029 (2011)ADSCrossRefGoogle Scholar
  59. P.L. Ge, M.D. Bao, H.J. Zhang, K. You, X.P. Liu, Surf. Coat. Technol. 229, 146–150 (2013)CrossRefGoogle Scholar
  60. V.A. Godyak, IEEE Trans. Plasma Sci. 34(3), 755–766 (2006)ADSCrossRefGoogle Scholar
  61. V. Godyak, J. Phys. D Appl. Phys. 46, 283001 (2013)CrossRefGoogle Scholar
  62. E. Grigore, C. Ruset, X. Li, H. Dong, Plasma Process. Polym. 6, S321–S325 (2009)CrossRefGoogle Scholar
  63. J.T. Gudmundsson, N. Brenning, D. Lundin, U. Helmersson, J. Vac. Sci. Technol. A 30(3), 030801 (2012)CrossRefGoogle Scholar
  64. G.J.M. Hagelaar, D.B. Mihailova, J. van Dijk, Analytical model of a longitudinal hollow cathode discharge. J. Phys. D Appl. Phys. 43, 465204 (2010)CrossRefGoogle Scholar
  65. Z.J. Han, S. Yick, I. Levchenko, E. Tam, M.M.A. Yajadda, S. Kumar, P.J. Martin, S. Furman, K. Ostrikov, Controlled synthesis of a large fraction of metallic single-walled carbon nanotube and semiconducting carbon nanowire networks. Nanoscale 3(8), 3214–3220 (2011)ADSCrossRefGoogle Scholar
  66. Z. J. Han, A. E. Rider, M. Ishaq et al., Carbon nanostructures for hard tissue engineering, RSC Advances 3(28), 11058–11072 (2013)CrossRefGoogle Scholar
  67. R. Hatakeyama, Rev. Mod. Plasma Phys. 1, 7 (2017).  https://doi.org/10.1007/s41614-017-0009-y ADSCrossRefGoogle Scholar
  68. P. Helander, Rep. Prog. Phys. 77, 087001 (2014)ADSCrossRefGoogle Scholar
  69. PEh Hovsepian, A.P. Ehiasarian, I. Petrov, Surf. Coat. Technol. 257, 38–47 (2014)CrossRefGoogle Scholar
  70. M. Hundt, P. Sadler, I. Levchenko et al., Real-time monitoring of nucleation-growth cycle of carbon nanoparticles in acetylene plasmas. J. Appl. Phys. 109(12), 123305 (2011)ADSCrossRefGoogle Scholar
  71. S. Ichimaru, Phase transitions, interparticle correlations, and elementary processes in dense plasmas. Rev. Mod. Plasma Phys. 1, 6 (2017).  https://doi.org/10.1007/s41614-017-0008-z ADSCrossRefGoogle Scholar
  72. M. V. Jacob, R. S. Rawat, B. Ouyang, K. Bazaka, D. S. Kumar, D. Taguchi, M. Iwamoto, R. Neupane, O. K. Varghese, Catalyst-free plasma enhanced growth of graphene from sustainable sources, Nano Letters 15(9), 5702–5708 (2015)ADSCrossRefGoogle Scholar
  73. M. V. Jacob, D. Taguchi, M. Iwamoto, K. Bazaka, R. S. Rawat, Resistive switching in graphene-organic device: Charge transport properties of graphene-organic device through electric field induced optical second harmonic generation and charge modulation spectroscopy, Carbon 112, 111–116 (2017)CrossRefGoogle Scholar
  74. J. Jin, H. Duan, X. Li, Vacuum 136, 112–120 (2017)ADSCrossRefGoogle Scholar
  75. W.H. Kao, Y.L. Su, J.H. Horng, K.X. Zhang, J. Biomater. Appl. 31(2), 215–229 (2016)CrossRefGoogle Scholar
  76. H. Kawasaki, T. Ohshima, Y. Yagyu, Y. Suda, Trans. Mater. Res. Soc. Japan 36(3), 459–498 (2011)CrossRefGoogle Scholar
  77. H. Kawata, M. Yasuda, Y. Hirai, Jpn. J. Appl. Phys. 47(8), 6914–6916 (2008)ADSCrossRefGoogle Scholar
  78. M.M. Kharkov, A.V. Kaziev, A.V. Tumarkin, V.E. Drobinin, T.V. Stepanova, A.A. Pisarev, I.O.P. Conf, Ser. J. Phys. Conf. Ser. 789, 012026 (2017)CrossRefGoogle Scholar
  79. D.W. Kim, H.Y. Lee, S.J. Kyoung, H.S. Kim, Y.J. Sung, S.H. Chae, G.Y. Yeom, IEEE Trans. Plasma Sci. 32(3), 1362–1366 (2004)ADSCrossRefGoogle Scholar
  80. T.H. Kim, et al., Plasma characteristics of inductively coupled plasma using dual-frequency antennas. Jpn. J. Appl. Phys. 52, 05EA02 (2013)Google Scholar
  81. H.B. Kim, H.C. Lee, K.N. Kim, S.K. Kang, G.Y. Yeom, J. Nanosci. Nanotechnol. 9(12), 7440–7445 (2009)Google Scholar
  82. D.R. Knapp, J. Phys. Conf. Ser. 591, 012018 (2015)CrossRefGoogle Scholar
  83. V.I. Kolobov, A.S. Metel, J. Phys. D Appl. Phys. 48, 233001 (2015)ADSCrossRefGoogle Scholar
  84. H. Kovaci, A.F. Yetim, Y.B. Bozkurt, A. Celik, Surf. Rev. Lett. 24(1), 1750005 (2017)CrossRefGoogle Scholar
  85. O.V. Krysina, N.N. Koval, I.V. Lopatin, V.V. Shugurov, S.S. Kovalsky, J. Phys. Conf. Ser. 669, 012032 (2016)CrossRefGoogle Scholar
  86. S. Kumar, I. Levchenko, K. K. Ostrikov et al., Plasma-enabled, catalyst-free growth of carbon nanotubes on mechanically-written Si features with arbitrary shape, Carbon 50(1), 325–329 (2012)CrossRefGoogle Scholar
  87. M.J. Kushner, J. Appl. Phys. 94(3), 1436–1447 (2003)ADSCrossRefGoogle Scholar
  88. W. Lang, Adv. Mater. Res. 399–401, 2018–2025 (2012)Google Scholar
  89. K. Lemmer, Propulsion for CubeSats. Acta Astronaut. 134, 231–243 (2017)ADSCrossRefGoogle Scholar
  90. Levchenko et al., IEEE Trans. Plasma Sci. 32, 2139 (2004a)ADSCrossRefGoogle Scholar
  91. I. Levchenko, M. Romanov, O. Baranov, M. Keidar, Vacuum 72, 335–344 (2004b)ADSCrossRefGoogle Scholar
  92. I. Levchenko, K.K. Ostrikov, J. Zheng, X. Li, M. Keidar, K.B.K. Teo, Scalable graphene production: perspectives and challenges of plasma applications. Nanoscale 8, 10511 (2016a)ADSCrossRefGoogle Scholar
  93. I. Levchenko, M. Keidar, U. Cvelbar, D. Mariotti, A. Mai-Prochnow, J. Fang, K. Ostrikov, Novel biomaterials: plasma-enabled nanostructures and functions. J. Phys. D Appl. Phys. 49(27), 273001 (2016b)CrossRefGoogle Scholar
  94. I. Levchenko, I.I. Beilis, M. Keidar, Nanoscaled metamaterial as an advanced heat pump and cooling media. Adv. Mater. Technol. 1(2), 1600008 (2016c)CrossRefGoogle Scholar
  95. I. Levchenko, K. Bazaka, Y. Ding, Y. Raitses, S. Mazouffre, T. Henning, P. J. Klar, S. Shinohara, J. Schein, L. Garrigues, M. Kim, D. Lev, F. Taccogna, R. W. Boswell, C. Charles, H. Koizumi, S. Yan, C. Scharlemann, M. Keidar, S. Xu, Space micropropulsion systems for cubesats and small satellites: from proximate targets to furthermost frontiers. Appl. Phys. Rev. 5, 011104 (2018a).  https://doi.org/10.1063/1.5007734 ADSCrossRefGoogle Scholar
  96. I. Levchenko, K. Bazaka, M. Keidar, S. Xu, J. Fang, Hierarchical multicomponent inorganic metamaterials: intrinsically driven self-assembly at the nanoscale. Adv. Mater. 30, 1702226 (2018b)CrossRefGoogle Scholar
  97. I. Levchenko, S. Xu, G. Teel, D. Mariotti, M.L.R. Walker, M. Keidar, Smart nanomaterials in space: recent progress in electric propulsion systems for small satellites. Nat. Commun. 9, 879 (2018c).  https://doi.org/10.1038/s41467-017-02269-7 ADSCrossRefGoogle Scholar
  98. Y. Li, L. Wang, D. Zhang, L. Shen, J. Alloy. Compd. 497, 285–289 (2010a)CrossRefGoogle Scholar
  99. Y. Li, L. Wang, L. Shen, D. Zhang, C. Wang, Surf. Coat. Technol. 204, 2337–2342 (2010b)CrossRefGoogle Scholar
  100. X. Li, Q. Chen, L. Sang, L. Yang, Z. Liu, Z. Wang, Phys. Proc. 18, 100–106 (2011)ADSCrossRefGoogle Scholar
  101. M.A. Lieberman, A.J. Lichtenberg, Principles of Plasma Discharges for Materials Processing (Wiley, New York, 2005)CrossRefGoogle Scholar
  102. M.A. Lieberman, J.P. Booth, P. Chabert, J.M. Rax, M.M. Turner, Plasma Sources Sci. Technol. 11, 283 (2002)ADSCrossRefGoogle Scholar
  103. B. Liu, Energy Proc. 38, 289–296 (2013)CrossRefGoogle Scholar
  104. Q. Luo, S. Yang, K.E. Cooke, Surf. Coat. Technol. 236, 13–21 (2013)CrossRefGoogle Scholar
  105. C. Luo, Q. Fu, C. Pan, Sci. Rep. 5, 9062 (2015)ADSCrossRefGoogle Scholar
  106. A.P.A. Manfridini, C. Godoy, J.C.A.-B. Wilson, M.V. Auad, Surf. Coat. Technol. 260, 168–178 (2014)CrossRefGoogle Scholar
  107. L. Martinu, D. Poitras, J. Vac. Sci. Technol. A 18(6), 2619–2645 (2000)ADSCrossRefGoogle Scholar
  108. S. Mazouffre, Electric propulsion for satellites and spacecraft: established technologies and novel approaches. Plasma Sources Sci. Technol. 25(3), 033002 (2016)ADSCrossRefGoogle Scholar
  109. T. Meziani, P. Colpo, F. Rossi, Plasma Sources Sci. Technol. 10, 276–283 (2001)ADSCrossRefGoogle Scholar
  110. A.I. Morozov, V.V. Savelyev, in Review of Plasma Physics, ed. by B.B. Kadomtsev, Shafranov (Consultant Bureau, New York, 2000)Google Scholar
  111. J. Musil, RSC Adv. 5, 60482–60495 (2015)CrossRefGoogle Scholar
  112. K. Ostrikov, I. Levchenko, U. Cvelbar, M. Sunkara, M. Mozetic, Nanoscale 2, 2012–2027 (2010)ADSCrossRefGoogle Scholar
  113. K. Ostrikov, E.C. Neyts, M. Meyyappan, Adv. Phys. 62(2), 113 (2013)ADSCrossRefGoogle Scholar
  114. A. Perret, P. Chabert, J.-P. Booth, J. Jolly, J. Guillon, Ph Auvray, Appl. Phys. Lett. 83(2), 243–245 (2003)ADSCrossRefGoogle Scholar
  115. A. Perret, P. Chabert, J. Jolly, J.-P. Booth, Appl. Phys. Lett. 86, 021501 (2005)ADSCrossRefGoogle Scholar
  116. E.J.D.M. Pillaca, M. Ueda, K.G. Kostov, H. Reuther, Appl. Surf. Sci. 258(24), 9564–9569 (2012)ADSCrossRefGoogle Scholar
  117. M.T. Rahman, M.N.A. Dewan, Plasma Sci. Tech. 17(2), 141–146 (2015)ADSCrossRefGoogle Scholar
  118. K. Ramazanov, V. Budilov, R. Vafin, I. Zolotov, XXVI Int. Symp. on Discharges and Electrical Insulation in Vacuum, 2014Google Scholar
  119. S. Rauf, K. Bera, K. Collins, Plasma Sources Sci. Technol. 19, 015014 (2010)ADSCrossRefGoogle Scholar
  120. S.E. Rodil, J.J. Olaya, S. Muhl, B. Bhushan, G. Wei, Surf. Coat. Technol. 201, 6117–6121 (2007)CrossRefGoogle Scholar
  121. Y. Sakurai, T. Osaga, IEEE Trans. Plasma Sci. 39(11 Pt. 1), 2550–2551 (2011)Google Scholar
  122. P.K. Schelling, L. Shi, K.E. Goodson, Mater. Today 8(6), 30–35 (2005)CrossRefGoogle Scholar
  123. M.K. Sharma, B.K. Saikia, S. Bujarbarua, Surf. Coat. Technol. 203, 229–233 (2008)CrossRefGoogle Scholar
  124. D. She, W. Yue, Z. Fu, Y. Gu, C. Wang, J. Liu, Mater. Des. 49, 392–399 (2013)CrossRefGoogle Scholar
  125. L. Shen, L. Wang, J.J. Xu, Surf. Coat. Technol. 228, S456–S459 (2013)CrossRefGoogle Scholar
  126. A.S. Shigarev, L.P. Dmitrieva, Met. Sci. Heat Treat. 20, 213 (1978)ADSGoogle Scholar
  127. T. Shimizu, Y. Teranishi, K. Morikawa, H. Komiya, T. Watanabe, H. Nagasaka, M. Yang, Thin Solid Films 581, 39–47 (2015)ADSCrossRefGoogle Scholar
  128. C.-S. Shin, D. Gall, Y.-W. Kim, N. Hellgren, I. Petrov, J.E. Greene, J. Appl. Phys. 92(9), 5084 (2002)ADSCrossRefGoogle Scholar
  129. R.L. Stenzel, J.M. Urrutia, Phys. Rev. Lett. 114, 205005 (2015)ADSCrossRefGoogle Scholar
  130. J.M. Stillahn, J. Zhang, E.R. Fisher, J. Vac. Sci. Technol. A 29(1), 011014 (2011)CrossRefGoogle Scholar
  131. J. Sun, W.P. Tong, H. Zhang, L. Zuo, Z.B. Wang, Surf. Coat. Technol. 213, 247–252 (2012)CrossRefGoogle Scholar
  132. D. Sung, V. Volynets, W. Hwang, Y. Sung, S. Lee, M. Choi, G.-H. Kim, J. Vac. Sci. Technol. A 30(6), 061301 (2012)CrossRefGoogle Scholar
  133. H. Takikawa, H. Tanoue, IEEE Trans. Plasma Sci. 35(4), 992–999 (2007)ADSCrossRefGoogle Scholar
  134. H. Tanaka, K. Ishikawa, M. Mizuno et al., State of the art in medical applications using non-thermal atmospheric pressure plasma. Rev. Mod. Plasma Phys. 1, 3 (2017).  https://doi.org/10.1007/s41614-017-0004-3 ADSCrossRefGoogle Scholar
  135. M. Tang, J. Ge, J. Wong, Z. Peng Ling, T. Dippell, Z. Zhang, M. Huber, M. Doerr, O. Hohn, P. Wohlfart, A.G. Aberle, T. Mueller, Phys. Status Solidi 9(1), 47–52 (2014)Google Scholar
  136. K.I. Thomassen, Microwave plasma density measurements. J. Appl. Phys. 36, 3642 (1965)ADSCrossRefGoogle Scholar
  137. W.P. Tong, H. Zhang, J. Sun, L. Zuo, J.C. He, J. Mater. Res. 25(11), 2082–2085 (2010)ADSCrossRefGoogle Scholar
  138. Y. Tong, T.-W. Guo, J. Wang, H.-F. Liang, Q. Mi, Adv. Mater. Res. 217–218, 1050–1055 (2011)CrossRefGoogle Scholar
  139. T. Uchida, S. Hamaguchi, J. Phys. D Appl. Phys. 41, 083001 (2008)CrossRefGoogle Scholar
  140. K. Vafin, K.N. Ramazanov, V.V. Budilov, XXVth Int. Symp. on Discharges and Electrical Insulation in Vacuum, Tomsk, 2012Google Scholar
  141. A.V. Vasenkov, M.J. Kushner, J. Appl. Phys. 95(3), 834–845 (2004)ADSCrossRefGoogle Scholar
  142. K.V. Vavilin, E.A. Kralkina, P.A. Nekludova, A.K. Petrov, A.M. Nikonov, V.B. Pavlov, A.A. Airapetov, V.V. Odinokov, G.Ya. Pavlov, V.A. Sologub, J. Phys. Conf. Ser. 669, 012033 (2016)CrossRefGoogle Scholar
  143. M. Volatier, D. Duchesne, R. Morandotti, R. Ares, V. Aimez, Nanotechnology 21, 134014 (2010)ADSCrossRefGoogle Scholar
  144. A. von Stechowa, O. Grulke, T. Klinger, Characterization of neutral loop discharges in the VINETA device. Eur. Phys. J. D 66, 64 (2012)ADSCrossRefGoogle Scholar
  145. S. Wang, W. Cai, J. Li, W. Wei, J. Hu, Mater. Lett. 105, 47–49 (2013)CrossRefGoogle Scholar
  146. Z. Wang, D. Zhang, P. Ke, X. Liu, A. Wang, J. Mater. Sci. Technol. 31(1), 37–42 (2015)CrossRefGoogle Scholar
  147. R. Wei, C. Rincon, T.L. Booker, J.H. Arps, Surf. Coat. Technol. 188–189, 691–696 (2004)CrossRefGoogle Scholar
  148. R. Wei, R.C. Benn, C.V. Cooper, Plasma Process. Polym. 4, S700–S706 (2007)CrossRefGoogle Scholar
  149. S. Wilczek, J. Trieschmann, J. Schulze, E. Schuengel, R.P. Brinkmann, A. Derzsi, I. Korolov, Z. Donko, T. Mussenbrock, Plasma Sources Sci. Technol. 24, 024002 (2015)ADSCrossRefGoogle Scholar
  150. T. Yamada, M. Ishihara, J. Kim, M. Hasegawa, S. Iijima, Carbon 50, 2615 (2012)CrossRefGoogle Scholar
  151. S. Yang, Y. Zhang, H.-Y. Wang, S. Wang, W. Jiang, Plasma Sources Sci. Technol. 26, 065011 (2017)ADSCrossRefGoogle Scholar
  152. A. Yazdani, M.R. Soltanieh, H. Aghajani, S. Rastegari, Vacuum 86, 131–139 (2011)ADSCrossRefGoogle Scholar
  153. M.A.P. Yazdi, F. Lomello, J. Wang, F. Sanchette, Z. Dong, T. White, Y. Wouters, F. Schuster, A. Billard, Vacuum 109, 43–51 (2014)ADSCrossRefGoogle Scholar
  154. S. Yick, I. Levchenko, S. Kumar, Z.J. Han, M.M.A. Yajadda, K. Ostrikov, Current control in the magnetron systems for nanofabrication: a comparison. IEEE Trans. Plasma Sci. 40(4), 1094–1097 (2012)ADSCrossRefGoogle Scholar
  155. S. Yick, A. Mai-Prochnow, I. Levchenko, J. Fang, M.K. Bull, M. Bradbury, A.B. Murphy, K. Ostrikov, RSC Adv. 5, 5142–5148 (2015)CrossRefGoogle Scholar
  156. Z. Zhang, X. Li, H. Dong, Acta Mater. 86, 341–351 (2015)CrossRefGoogle Scholar
  157. Y. Zhang, C. Charles, R. Boswell, Cross-field transport of electrons at the magnetic throat in an annular plasma reactor. J. Phys. D Appl. Phys. 50, 015205 (2016)ADSCrossRefGoogle Scholar
  158. Y. Zhao, B. Yu, L. Dong, H. Du, J. Xiao, Surf. Coat. Technol. 210, 90–96 (2012)CrossRefGoogle Scholar

Copyright information

© Division of Plasma Physics, Association of Asia Pacific Physical Societies 2018

Authors and Affiliations

  1. 1.National Aerospace UniversityKharkovUkraine
  2. 2.Department for Surface Engineering and OptoelectronicsJožef Stefan InstituteLjubljanaSlovenia
  3. 3.Plasma Sources and Applications Centre, NIENanyang Technological UniversitySingaporeSingapore
  4. 4.School of Chemistry, Physics, and Mechanical EngineeringQueensland University of TechnologyBrisbaneAustralia
  5. 5.CSIRO-QUT Joint Sustainable Processes and Devices LaboratoryLindfieldAustralia
  6. 6.College of Chemistry and Chemical EngineeringChongqing University of TechnologyChongqingPeople’s Republic of China

Personalised recommendations