0% found this document useful (0 votes)

20 views17 pages

A Tutorial On Theoretical and Computational Techniques For Gas Breakdown in Microscale Gaps

Uploaded by

Martin Martinec

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

20 views17 pages

A Tutorial On Theoretical and Computational Techniques For Gas Breakdown in Microscale Gaps

Uploaded by

Martin Martinec

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 17

808 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO.

4, APRIL 2020

A Tutorial on Theoretical and Computational

Techniques for Gas Breakdown in Microscale Gaps
Allen L. Garner , Senior Member, IEEE, Amanda M. Loveless, Student Member, IEEE,
Jiba Nath Dahal, Member, IEEE, and Ayyaswamy Venkattraman, Member, IEEE

Abstract— Paschen’s law (PL), derived based on the Townsend I. I NTRODUCTION

avalanche (TA) condition, is commonly used to predict gas
breakdown. For microscale gaps near atmospheric pressure, TA
is insufficient to drive breakdown and ion-enhanced field emis-
sion (FE) dominates. Accurately predicting breakdown voltages
G AS breakdown plays a major role in device reliability
across high voltage [1], pulsed power [2], switch [3],
gas insulated switchgear [4], high power microwave [5], [6],
for these gaps is critical for numerous applications, including directed energy [5], fusion [7]–[11], and accelerator appli-
environmental remediation, medicine, combustion, and propul- cations [12]. In other areas, such as combustion [13]–[16]
sion. This tutorial summarizes theoretical and computational
approaches for predicting this behavior and demonstrating the
and biomedical applications [17]–[21], scientists and engineers
transition between the TA and FE mechanisms. It focuses on the seek to induce gas breakdown to create plasmas. Both scenar-
derivation of closed-form solutions from a theory that accounts ios require accurately predicting the gas breakdown conditions
for the generation of additional positive space charge at the cath- to optimize device design.
ode due to electrons generated by the strong FE-induced electric Breakdown is traditionally predicted by Paschen’s law (PL)
fields. Appropriate simplifications using a matched asymptotic
analysis in terms of the total ionization in the gap agree well with
[22], which is derived from the Townsend avalanche (TA)
simulations and experiments and show the transition from FE for condition where the number of ions generated in the volume
small gaps to PL at larger gaps. Specifically, this theory shows by a single electron emitted from the cathode is sufficient to
that the breakdown voltage varies linearly with gap distance emit a new electron from the cathode to replace it. This is
when FE dominates, agreeing with both the experimental and mathematically given by
simulation results. The particle-in-cell Monte Carlo collision
−1
(PIC/MCC) simulations used to predict the ionization coefficient exp(αd) − 1 = γSE (1)
provide additional insight into the mechanisms involved. Future
benefits of extended theoretical and computational research for where γSE is the secondary emission coefficient and α is the
examining microscale and nanoscale breakdown and electron ionization coefficient defined as the number of ions generated
emission, particularly assessing the impact of electrode surface per electron per unit length and is given by [23]
structure and device design and coupling with additional emission
mechanisms, will be discussed. α = A p p exp(−B p p/E) (2)
Index Terms— Electric breakdown, electron emission, gas dis- where E is the electric field, p is the gas pressure, and A p
charge, mathematical models, Monte Carlo methods, plasmas.
and B p are gas-dependent parameters tabulated in [23]. Sub-
stituting (2) into (1) yields PL for the macroscale breakdown
Manuscript received October 24, 2019; revised December 27, 2019;
accepted March 2, 2020. Date of publication March 23, 2020; date of voltage Vb as
current version April 10, 2020. This work was supported in part by the B p pd
Office of Naval Research under Grant N00014-17-1-2702, in part by the Vb = −1
(3)
Air Force Office of Scientific Research under Grant FA9550-18-1-0218, and ln(A p pd) − ln ln 1 + γSE
in part by a Directed Energy Professional Scholarship. The review of this
article was arranged by Senior Editor S. J. Gitomer. (Corresponding author: where Vb = Ed for a parallel plate geometry with gap distance
Allen L. Garner.) d. This geometric assumption is important because many
Allen L. Garner is with the School of Nuclear Engineering, Purdue experimental setups use other geometries, most notably pin-
University, West Lafayette, IN 47906 USA, also with the Department of
Agricultural and Biological Engineering, Purdue University, West Lafayette, to-plate geometries, where E varies with position x across the
IN 47907 USA, and also with the School of Electrical and Computer gap [23], [24], which also makes α and Vb functions of x.
Engineering, Purdue University, West Lafayette, IN 47907 USA (e-mail: Noting from (3) that Vb is a function of pd, we can set
[email protected]).
Amanda M. Loveless is with the School of Nuclear Engineering, Purdue dVb /d(pd) = 0 and show that the minimum Vb , Vb,min , occurs
−1
University, West Lafayette, IN 47906 USA. when pd = exp(1) ln(1 + γSE )/A p , yielding
Jiba Nath Dahal is with the Department of Physics, Truman State University,
−1
Kirksville, MO 63501 USA. B p exp(1) ln(1 + γSE )
Ayyaswamy Venkattraman is with the Department of Mechanical Engi- Vb,min = . (4)
neering, University of California Merced, Merced, CA 95343 USA (e-mail: Ap
[email protected]). Gas breakdown is well-characterized in many applications
Color versions of one or more of the figures in this article are available
online at http://ieeexplore.ieee.org. within a certain range of ionization given by αd. Accord-
Digital Object Identifier 10.1109/TPS.2020.2979707 ing to the Meek criterion, when αd is sufficiently high
0093-3813 © 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
GARNER et al.: TUTORIAL ON THEORETICAL AND COMPUTATIONAL TECHNIQUES 809

to E s to give an enhanced FN current density of [27], [32]

jFN = CFN (E s + E + )2 exp[−DFN /(E s + E + )2 ]. (8)

This positive space charge also modifies the TA criterion from

(1) by adding an ion-enhancement coefficient γ to γ , yielding

exp(αd) − 1 = (γSE + γ )−1 . (9)

Go and Venkattraman [28] previously reviewed ion-

enhanced FE-driven microscale gas breakdown and pointed
out multiple instances of this deviation from the classical
PL at microscale. While many plasma and high-voltage sys-
tems operate outside of this parameter regime, the continued
reduction in electronic device sizes to microscale for micro-
electromechanical systems (MEMS) [33]–[35] and the further
reduction to nanoelectromechanical systems (NEMS) [36]
Fig. 1. Schematic of the mechanisms involved in microscale gas breakdown.
(a) Impact ionization and collisional process leading to TA under certain
necessitates characterizing breakdown voltages at even smaller
conditions. (b) Ion-induced secondary emission, where ions left from the gap distances to ensure device reliability. Moreover, many
impact ionization collide with the cathode to release the additional electron(s). applications require intentionally creating microdischarges
(c) FE tunneling process where electrons enter the gap from the cathode due
to intense electric fields lowering the potential barrier.
[17]–[21]. For instance, microplasma-assisted combustion [37]
and micropropulsion [38] require the predictable generation
of microplasmas. Biomedical and environmental applications
frequently use plasma jets, where the plasma is initially formed
(≈20 for air at atmospheric pressure for a 1-cm gap) [25],
by applying a voltage sufficient to induce breakdown across
[26], gas breakdown is driven by the streamer process. For
a micrometer-sized gap [17]–[20]. Many of these applications
sufficiently small αd, the number of ionizations is inadequate
occur at atmospheric pressure, leading to the conditions first
to drive breakdown by the traditional TA process [27]–[29].
stated by Boyle and Kisliuk [27], making the mechanism
This becomes particularly evident for microscale gaps at
different from both the classical PL [19] and vacuum break-
atmospheric pressure, where d is too small for the avalanche
down [39].
condition to occur. This causes a deviation from PL character-
The increasing importance of this operating regime for
ized by either the absence of the classical Vb,min at a certain
these applications has motivated further computational and
pd or an extended plateau prior to Vb decreasing linearly with
theoretical studies to understand this deviation from PL.
decreasing gap distance [28], [30]. Boyle and Kisliuk [27]
Several of these studies use simulations, such as fluid
proposed that this linear decrease in V arose because field
[40]–[45] and particle-in-cell (PIC) [46]–[56]. Others have
emission (FE) generated electrons that would interact with
derived equations based on first principles and then used
the neutral gas near the cathode to induce an ion-enhanced
either numerical [28], [32] or semiempirical approaches [47]
secondary emission that would combine with the standard
to account for FE to predict breakdown voltage. More recent
secondary emission involved in PL. Fig. 1 shows a schematic
studies have attempted to elucidate the impact of surface
describing the phenomena involved.
roughness [42]–[45], [57]–[59] and nonuniform geometry (i.e.,
To describe this, Boyle and Kisliuk [27] proposed modifying
nonplanar) [24] on breakdown voltage. Another approach
the Fowler–Nordheim (FN) equation, which mathematically
involves deriving equations from first principles to directly
describes the FE current density. One formulation of the FN
unify FE with TA [32].
equation is given by [31]
While using simulations or numerically solving these
jFN = CFN E s2 exp(−DFN /E s2 ) (5) equations yields insight on the overall behavior [32], these
approaches do not provide closed-form solutions to elucidate
where the behavior at extreme gap distance, such as analytically
predicting the scaling of Vb as d → 0. This motivated
AFN β 2 (3.79×10−4 )2 BFN multiple studies applying matched asymptotic analyses to
CFN = 2 exp (6)
φt (y) φ 1/2 these equations to derive universal breakdown equations valid
DFN = 0.95BFN φ /β
3/2
(7) for any gas [29], [57], [60]–[65]. This permits the assessment
of the behavior in limiting cases and the quantification of
where β is the field enhancement factor, φ is the electrode the relative contribution of each mechanism on breakdown
work function, BFN = 6.85 × 107 V·cm−1 ·eV−3/2 , AFN = voltage.
6.2 × 10−6 A·eV·V−2 , t 2 (y) ≈ 1.1, and E s is the surface This tutorial will briefly summarize several approaches used
electric field [E s = E(0)], which is simply E = V /d for to model microscale gas breakdown before focusing on ana-
a parallel plate geometry. Boyle and Kisliuk proposed that lytic approaches and PIC Monte Carlo collision (PIC/MCC)
the positive space charge that the electrons created near the techniques. Section II reviews various analytic and semiana-
cathode created an additional electric field E + that is added lytic approaches used. Section III derives analytic breakdown

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
810 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO. 4, APRIL 2020

equations based on [32] and applies them to various experi- Equation (12) provides utility and agrees well with experi-
ments. Section IV summarizes the applications of PIC/MCC mental data in the FE regime [68], while also demonstrating
to examine breakdown from a particle perspective. We make the transition to the traditional PL. This approach was also
concluding remarks in Section V. applied to assess experimental measurements of breakdown in
argon [69] and air [70] using the appropriate definitions of α
II. B RIEF R EVIEW OF A NALYTIC AND S EMIANALYTIC in (15)–(17). However, this method does not directly lead to PL
A PPROACHES mathematically and it introduces two new material-dependent
constants (K and D) that are not commonly tabulated in the
While more details on various analytic and semianalytic plasma physics literature. One potential alternative would be to
approaches for predicting microscale gas breakdown may be couple it with experimental measurements of an effective sec-
found in an earlier review [28], we provide a brief overview to ondary electron emission coefficient γeff , as reported as a func-
highlight the differences between these approaches and the one tion of E and E/P for various gap distances from 1 to 100 μm
discussed in detail in Section III. One point we make is that all in argon [71]. Others have modified this approach to more
approaches for theoretically assessing these phenomena suffer directly couple the contribution of FE to PL to obtain [72], [73]
from some degree of “semiempirical” nature in the sense that
many parameters involved in the theory, particularly α, depend (γSE + K e−d D/V )(exp[ Apd exp(−Bpd/V )] − 1) = 1 (13)
on semiempirical expressions depending on the gas and the which still requires the two additional constants introduced in
electric field [23]. The main distinction between the theories (10), but considers γ from (10) as an FE correction to γSE
discussed in this section and the one discussed in more detail in such that (13) will reduce to PL when FE effects diminish.
Section III is the proximity to the base equations from FN and One may also anticipate the transition to space-charge
PL. Section III uses analytic techniques to relate a full equation limited emission (SCLE) at smaller gaps [74], although the
based on these phenomena without introducing any additional gap distance for the transition in emission mechanisms is
assumptions or parameters. In this way, the equations derived most likely on the order of hundreds of nanometers to a
there are more directly related to the accepted fundamental few microns based on a theoretical assessment of electron
FN and PL equations. emission that incorporated collisions with neutral gas atoms
Tirumala and Go [66] used the base equation of (9) for into the energy balance equation for an emitted electron [75].
the fundamental microscale PL condition and linked it to a Darr et al. [75] proposed that electron emission would transi-
modified FN relationship to determine the ion-enhanced field tion from FN to Mott–Gurney (MG) [76] as one reduces gap
emission as a function of distance along the cathode as a func- distance for a given pressure. Further reduction in gap distance
tion of time. They then derived a relationship for the modified will eventually satisfy the Child–Langmuir (CL) condition for
secondary electron coefficient γ by integrating over radius and SCLE in vacuum [77]–[79] for sufficiently small gap distance,
distance. While a complete derivation, the resulting equations regardless of gap pressure. Ongoing experiments are assessing
did not yield a closed-form solution and required numerical electron emission at atmospheric pressure for various gap
solution, prohibiting assessment of limiting behavior. distances.
Another theoretical study started with a relationship similar One may also incorporate thermionic and thermo-FE into
to (9), but modified it by introducing a single effective microscale gas breakdown [80]–[83], similar to coupling FE
secondary emission term that accounts for ion-enhanced FE and TA for microscale gas breakdown. Other studies have
by [27] incorporated temperature from a quantum perspective by eval-
uating the ion impact on cold FE and thermo-FE at room and
γ = K e−D/E (10)
higher temperature [84]. Another approach enhanced classic
where K and D are material- and gas-dependent constants FE by determining the impact of a gaseous ion on electron
[67]. As we shall show later, some studies have considered (10) tunneling by solving the 1-D Schrödinger equation with an
as strictly the correction for FE-enhanced breakdown and add ion present [85]. This suggested that the resonant emission
it to γSE . The resulting coupled avalanche condition becomes current may be controlled by manipulating the cathode supply
[67] function [85].
Thus, multiple theoretical approaches have examined the
K e−d D/V (exp[ Apd exp(−Bpd/V )] − 1) = 1 (11) coupled TA–FE regime, but most suffer from lacking
closed-form solutions to demonstrate limiting behavior and
assuming E = V /d for a planar geometry. While straight-
the transition between FE and PL, or they require addi-
forward to derive, this equation is transcendental and also
tional parameters. This motivated the effort summarized in
requires numerical techniques to solve. Rearranging (11) to
Section III to derive closed-form solutions to characterize this
obtain exp[ Apd exp(−Bpd/V )] = exp(Dd/V )/K +1, taking
behavior without introducing additional parameters (besides
the natural logarithm of both sides, and then keeping the first
those already present in FN and PL).
term of the resulting Taylor series expansion of the left-hand
side since exp(Dd/V )/K 1 because Dd/V 1 in the FE III. D ERIVATION OF A NALYTIC E QUATIONS
regime yields an analytic solution for the FE regime as [68]
This tutorial focuses on analyzing the equations derived
d(D + Bp) by Venkattraman and Alexeenko [32] that incorporated the
V = . (12)
ln(Apd K ) positive space charge arising from the FE-generated electrons

ionizing neutral gas near the cathode. They assessed these Substituting (19) into (20) yields
effects using the modified FN equation of (8) and the modified + DFN E +
TA condition of (9). jFN 1 + 2 EE exp E2 exp(αd)
jtot = . (21)
Deriving the unified FE–TA breakdown condition requires 1 − γSE (exp(αd) − 1)
first relating the total steady-state current density in the gap j
Alternatively, we may write jtot in terms of the corrected γ as
to the current density of the cathode source j0 [23] by
jFN exp(αd)
j0 exp(αd) jtot = . (22)
j= . (14) 1 − (γSE + γ )(exp(αd) − 1)
1 − γSE (exp(αd) − 1)
Equating (21) and (22) and solving for γ gives
The PIC simulations indicate that j is approximately constant
across the gap, while the electron and ion current densities, 1 − [1 − γSE (exp(αd) − 1)] exp(−DFN E + /E 2 )
γ = − γSE .
je and ji , respectively, vary with position [32]. The ioniza- (exp(αd) − 1)(1 + 2E + /E)
tion coefficient α is typically determined semiempirically for (23)
macroscale gaps by either
To determine E + , Venkattraman and Alexeenko [32] started

α p 1/2 from the 1-D Poisson’s equation in Cartesian coordinates,
= C exp −D (15) given by [32]
p Macroscale E
dE ρ
with C = 29.2 cm−1 ·torr−1 and D = 26.6 = (24)
dx ε0
V1/2 ·cm−1/2 ·torr−1/2 for argon or
where ρ is the charge density and ε0 is the permittivity of
α p free space. For a planar geometry with minimal space charge,
= A p exp −B p (16)
p Macroscale E one may assume that the electric field at the center of the gap
is the nominal electric field (E = V /d) and that the charge
where A p and B p are the constants for inert gases. One
density is constant from d/2 to d, so [32]
limitation of these approaches is that these constants are
often only valid for certain ranges of E/ p and may not be ρd
E+ = (25)
valid for microscale gaps [23], [86]; however, they provide 2ε0
reasonable first-order estimates. Because the applied voltages where
are comparable to the ionization potential Vip in units of V
jion
and the microscale gaps are on the order of the electron mean ρ= (26)
free path, (15) and (16) require correction. The PIC/MCC vd
simulations yield the semiempirical correction of with the ion drift velocity given by

α α V /Vip − 1.0 0.8 2ekT E
= 1 − exp − (17) vd = (27)
p p Macroscale 3.1 πmpσCE

with either (15) or (16) used for (α/ p)Macroscale , as appropriate, where k is the Boltzmann’s constant, m is the mass of the
and Vip is the ionization potential [32]. We point out that α gas atom, T is the neutral gas temperature, and σCE is the
is taken as a constant here because the electric field across charge exchange cross section, which typically depends on
the gap is always the nominal electric field (E = V /d) for energy. This assumption generally predicts breakdown volt-
a planar geometry. In other geometries, such as pin-to-plate, age well compared with the PIC simulations for E/P from
one must integrate across the gap to determine the overall 600 to 850 V/cm/torr for gap distances from 1 to 3 μm with
ionization [24]. divergence between the simulation and theory occurring while
Alternatively, one may consider the total current as the sum approaching breakdown [32]. This approach also assumes
of the current due to FE, ionization, and ion enhancement and sufficiently negligible buildup of space charge in the gap,
recast (14) in terms of these phenomena. This gives the current which will make E nonconstant across the gap. This approach
density due to the combination of FE and ionization as focuses on the FE regime, which generally does not consider
space charge itself, although the microscale gas breakdown
jFN+ion = jFN exp(αd). (18) theory considers the influence of the positive ions that generate
E + . Further reductions in the gap size or increases in V will
Because E + E, we may rewrite (8) for the enhanced FN ultimately induce sufficient space charge to cause transition
current density as from FE to SCLE, as discussed in more detail elsewhere and
estimated to occur at ∼250 nm for nitrogen at atmospheric
E+ DFN E + DFN
jFN = CFN E s2 1 + 2 exp 2
exp − . (19) pressure [75].
Es Es Es Experiments typically use either pin-to-plate [57] or pin-
to-pin [62], [64], [65] geometries. For either case, E will
Considering j0 = jFN allows us to rewrite jtot from (14) as
not be uniform across the gap, making the assumption of

jFN exp(αd) the nominal electric field problematic. Recent simulations
jtot = . (20)
1 − γSE (exp(αd) − 1) coupled FE and TA numerically to account specifically for the

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
812 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO. 4, APRIL 2020

more common pin-to-plate geometry [24] and future analytic A. Concept of the Matched Asymptotic Analysis
work may further adapt the equations that follow based on
the assumption from (25) to more accurately account for Before further assessing the breakdown behavior, we first
these geometries. Another option may be to apply variational outline the philosophy for the matched asymptotic analysis
calculus (VC), as done recently to derive a coordinate system and the path followed in these studies.
invariant equation for the CL law for SCLE in vacuum [87]. In many cases across a range of scientific disciplines, one
The VC derivation of CL yielded the first exact analytic can derive equations for fundamental phenomena that will
solutions for SCLE for cylindrical and spherical geometries not yield exact, closed-form solutions. While this is not a
and could be valuable here. For now, we have assumed the complete hindrance because these equations may be amenable
nominal electric field and demonstrated the effectiveness of to numerical and computational approaches, it does raise some
this assumption for matching the experimental results in its challenges. First, depending on the nature of the equations
current state [57], [62], [64], [65], although its utility as a under study, the numerical and computational approaches
purely predictive tool is hampered. may be complicated and computationally expensive, requiring
The ion current density may then be written as extensive resources to solve. One such example involves solv-
ing stiff differential equations or partial differential equations
jtot (exp(αd) − 1)
jion = . (28) in which the desired solution varies slowly even though nearby
exp(αd) solutions vary rapidly. This necessitates using very small
Substituting (21) into (28) yields step sizes, which dramatically increases the computational
+ DFN E + expense. In other cases, one may note fundamental scaling
jFN 1 + 2 EE exp E2 (exp (αd) − 1)
jion = (29) behavior under various limits from computational or numerical
1 − γSE (exp(αd) − 1) solutions; however, noting this from a trend of numerical
which can be combined with (25) and (26) to give solutions does not necessarily “prove” or demonstrate the
physics behind this behavior. In both cases, deriving closed-
2v d ε0 E 2 [1 − γSE (exp(αd) − 1)]
E+ = . (30) form solutions that provide a more rapid way to predict the
d DFN jFN (exp(αd) − 1) solutions under appropriate limits rather than solely relying
Multiplying both sides of (30) by DFN /E 2 and defining E ∗ = on exact computational or numerical approaches may be more
E/DFN , x = DFN E + /E 2 , and convenient and/or more illuminating.
One common approach for achieving these goals involves
2v d ε0 E 2 [1 − γSE (exp(αd) − 1)]
Fbr = (31) performing a matched asymptotic analysis. In its simplest
DFN d jFN (exp(αd) − 1) form, this requires identifying the independent variable of most
gives interest for examining the fundamental behavior of the desired
exp(x)(1 + 2E ∗ x) equation and then deriving closed-form solutions under the
g(x) = = 1. (32) limits of this variable becoming small and large without bound.
Fbr x
If conducted correctly, the solutions for the matched asymp-
Because v d is a function of E, it will also vary with position
totic analysis will agree well with the exact solution in these
for nonplanar geometries. Moreover, the dependence of α, v d ,
limits and often more poorly for intermediate values. Because
and x on E makes (32) transcendental, preventing a closed- the equations derived will be analytic, they can be easily cal-
form solution.
culated using typical mathematical software or spreadsheets.
The coupled FE–TA breakdown criterion arises from solv- The general approach of the matched asymptotic analysis
ing (32) for E + , which only has a solution for a prebreakdown
of taking limits with respect to an appropriate variable as
steady state. The minimum of (32) occurs when
√ it becomes small and large without bound usually motivates
1 + 8E ∗ − 1 nondimensionalizing the relevant equations. Eliminating the
x = x0 = (33)
4E ∗ dimensions or units makes each independent variable “equiva-
and no solution arises when g(x 0 ) > 1; therefore, the break- lent” when taking its limit as large or small. When the units are
down condition is g(x 0 ) = 1 and we can combine (27), (31), included, the meaning of “large” or “small” may be obscured
and (33) to obtain and relative. For instance, if one is considering Ohm’s law,
a resistance R ∼ O(M ) may be significant under certain
2v d ε0 E 2 [1 − γSE (exp(αd) − 1)] circumstances, while R ∼ O( ) may be quite important
Fbr =
DFN d jFN (exp(αd) − 1) when dealing with losses. Taking the limit of large R with
exp(x 0 )(1 + 2E ∗ x 0 ) the dimensions involved may become somewhat arbitrary.
= . (34)
x0 Once nondimensionalized, the meaning of “large” or “small”
Equation (34) cannot be solved analytically and must be involves comparing the appropriate term to unity, eliminat-
solved numerically. These solutions suggest that breakdown ing the concerns of the units and permitting the application
voltage decreases linearly with decreasing gap distance once of series expansions under the appropriate conditions for
FE drives microscale breakdown; however, this cannot be convergence.
stated definitively without an exact analytic solution. This Because (34) is complicated with several dependences on
motivated the application of a matched asymptotic analysis E, we performed the matched asymptotic analyses in order of
to derive closed-form solutions under appropriate conditions. increasing complexity, starting with the simplest case of argon

at atmospheric pressure [60], then adding in gas dependence TABLE I

at atmospheric pressure [61], and finally deriving equations PARAMETERS U SED IN C ALCULATIONS FOR A IR
for any gas at any pressure (as long as FE and TA drive
breakdown) [29]. We also performed these studies in order
of increasing consideration of electrode effects.
One of the critical aspects of the matched asymptotic
analysis involves defining the nondimensional terms with the
goal of eliminating (or at least minimizing) the number of
nondimensional terms in the equation that are not independent
variables. Under this motivation, our initial study of argon at
atmospheric pressure included cathode effects, represented by
φ and β, within our scaling terms such that the breakdown
voltage was not a function of these variables [60]. This gave
the benefit of making the solution universal as a function of
electrode behavior, or true for any electrode design, material,
or surface roughness.
Our subsequent studies [29], [61]–[65] did not incorporate
φ and β into the scaling terms, but instead included them
as separate independent variables. The resulting equations
were no longer universal for any electrode but depended on
electrode material and structure. We did this because we
wanted to directly understand the impact of the changes in
electrode design and structure on gas breakdown. Practically,
each approach is mathematically valid for comparing to exper-
imental behavior because they yield the same results upon Table I reports the common parameters used for calculations
making the equations dimensional. The choice depends on the with air. Coupling (35) and (36) with (34) yields
focus of the given study.

exp[φ̄ 3/2 /(β Ē)] T̄ Ē {1 − γSE [exp(ᾱ d̄) − 1]}
B. Applying the Matched Asymptotic Analysis to Microscale β φ̄ 1/2 exp(φ̄ −1/2 ) p̄d̄ 2 [exp(ᾱ d̄) − 1]
Gas Breakdown exp(x o )(1 + 2 Ē x o )
= (37)
As mentioned above, we first applied the matched asymp- xo
totic analysis to argon at atmospheric pressure because the
resulting closed-form solutions were straightforward and we which is a fully nondimensionalized breakdown condition
could simplify critical electric field dependences, particularly valid for all gases and pressures with αd defined by (15) and
on α [60]. We next applied a semiempirical correction factor (16) and nondimensionalized using (35) and (36).
to relate pressure and electric field for different gases at For the matched asymptotic analysis of (37), we noted the
atmospheric pressure [61]. Finally, we derived a universal dependence on ᾱ d̄ and considered the limits for ᾱ d̄ 1 and
theory that was true for any gas at any pressure where ᾱ d̄ 1. We then determined the resulting limits as
FE or TA drove breakdown [29]. It is this theory that we
summarize here with full details provided elsewhere [29]. ⎡ ⎤
We first define the scaling terms relating the dimensionless − 1 − 2
1 −2 1 (φ̄
3/2 /β + γSE p̄2 d̄ + p̄)
V̄ = d̄ ⎣ ⎦ (38)
and dimensional quantities as 1

Ē = E E ∗−1 , T̄ = T T∗−1 , p̄ = pp∗−1

d̄ = d L −1 , j̄FN = jFN j0−1 , ᾱ = αL, φ̄ = φφ∗−1 (35) for ᾱ d̄ 1, where

−2
where Ē, T̄ , p̄, d̄, j̄FN , ᾱ, and φ̄ are the dimensionless electric ln[T̄ p̄−1 d̄ ] ln[ 1 ]
field (∼0.04), temperature (∼0.03), pressure (∼0.01), gap 1 = − ln[β φ̄ 1/2 ] − φ̄ −1/2 −
2 2
distance (∼500), FN current (∼10–13), ionization coefficient 3
(∼0.01), and work function (∼0.05), respectively, with the − ln[ p̄d̄] − γSE p̄ d̄ − (39)
2
scaling terms given by
2 and = 4 × 104 , and
p∗ = E ∗ B −1 −1 −1 1
p , L = p∗ A p , j0 = A FN E ∗ /(t (y)φ∗ )
2

φ∗ = [(3.79 × 10−4 )2 BFN ]2 , E ∗ = 0.95BFN φ∗3/2 ⎡ ⎤

2 − 2 − 2
2 −2 2 φ̄
3/2 /β
πmσCE B p AFN V̄ = d̄ ⎣ ⎦ (40)
T∗ = −4 2
. (36)
o A p t (y)[(3.79 × 10 ) BFN ]
8ek 2 2 2

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
814 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO. 4, APRIL 2020

for ᾱ d̄ 1, where
−2
ln[T̄ p̄−1 d̄ ]
2 = − ln[β φ̄ 1/2 ]
2
−1/2 ln[ 2 ] p̄d̄
− φ̄ − + ln 1 − γSE exp
2 exp(1)

p̄d̄ 3
− ln exp −1 − (41)
exp(1) 2
and 2 = 105 . The detailed mathematical manipulations
necessary to achieve (38)–(41) are provided elsewhere [29].
We can also recast (39) and (41) in terms of the contri-
butions of FE and TA by isolating the terms based on their
physical origin. Defining a = −[μ̃a + ν̃a ] and noting that
a < 0 (with a = 1 or 2 as appropriate), we define the FE
component of (39) as [29]
ln( 1 ) 3
μ̃1 = + ln(β φ̄ 1/2 ) + φ̄ −1/2 + (42) Fig. 2. Breakdown parameter 1 as a function of dimensionless gap distance
d̄ for dimensionless pressures of p̄ = 4.47 × 10−6 and p̄ = 5.89 × 10−7 .
2 2
The vertical dotted line denotes the approximate regime where breakdown
and the TA component as transitions from FE-driven to TA-driven.
ln[T̄ /( p̄d̄ 2 )]
ν̃1 = ln( p̄d̄) + γSE p̄d̄ − . (43)
2
Using an analogous definition for 2 yields
μ̃2 = ln[β φ̄ 1/2 ] + φ̄ −1/2 (44)
and
−2
ν̃2 = − ln[τ̄ p̄−1 d̄ ]/2 + ln[ 2 ]/2
− ln{1 − γSE [exp( p̄d̄ exp(−1))]}
+ ln[exp( p̄d̄ exp(−1)) − 1] + 3/2 (45)
for the FE and TA components, respectively. When μ̃a > ν̃a ,
FE dominates; μ̃a = ν̃a corresponds to when FE and TA
contribute equally, or the transition to the TA condition; μ̃a <
ν̃a corresponds to the TA regime. Prior experience indicates
that the specific experimental results determine whether ᾱ d̄
1 or ᾱ d̄ 1 better fits the data.
Experiments indicate that ᾱ d̄ generally does not become Fig. 3. Breakdown parameter 1 as a function of dimensionless pressure p̄
for dimensionless gap distances of d̄ = 2.25 × 106 and d̄ = 5.10 × 106 .
much larger than unity in practice in the combined FE–TA
regime because one transitions rapidly to the PL. Returning
to (37), we note that once ᾱ d̄ becomes sufficiently large upon Solving (46) in dimensionless units yields a universal form
transition to the TA regime, we recover a universal condition of PL (UPL), which is true for any gas, of
(true for any gas) for PL, given by [29]
p̄d̄
1 − γSE [exp(ᾱ d̄) − 1] = 0. (46) V̄ = −1
. (48)
ln[ p̄d̄] − ln ln 1 + γSE
Given the importance of |ᾱ d̄| for the transition to FE and PL, One important point is that the FE-driven regime scales with
it is interesting to note that the PL condition is satisfied when ᾱ d̄, while the UPL regime scales as p̄d̄. Thus, the two regimes

γSE + 1 do not have the same universal scales [65].
ᾱ d̄ = ln ≈ − ln γSE (47) Specifically, further reducing d̄ into the FE-driven regime
γSE
ultimately yields V̄ ∝ d̄. We can show this for (38) for
where γSE 1. For γSE = 10−3 , ᾱ d̄ = 6.91, which ᾱ d̄ 1 by factoring | 1 | from the radical in (38) and applying
agrees with the transition to PL from applying this theory the binomial expansion to the resulting expression inside the
to experimental data [65]. Given the nature of the matched radical to obtain
asymptotic analysis at predicting the analytic solution for
d̄ φ̄ 3/2
values on the order of unity, this suggests that the solution V̄ = + γSE p̄2 d̄δ + p̄ (49)
β
using ᾱ d̄ 1 may often adequately predict behavior until the 1

transition to PL, especially if γSE is sufficiently large to make where δ = 1 −exp{−[(d̄/λ̄−1)/3.1]0.8 } is a correction to α as
ᾱ d̄ sufficiently small. gap distance approaches the mean free path [32]. Because the

Fig. 4. Dimensionless breakdown voltage V̄ as a function of dimensionless gap distance d̄ for dimensionless pressures of p̄ = 4.47×10−6 and p̄ = 5.89×10−7
on (a) and (c) log–log and (b) and (d) linear scales.

first term in parentheses dominates in the FE regime, further the limiting expression for FE for low ionization, and the
simplifying (49) gives UPL for large ionization, we can apply the full theory to
3/2 various conditions to illustrate the transition between the
φ̄
V̄ ≈ d̄. (50) mechanisms. We have previously demonstrated the utility of
β 1 this theory to experimental data for both pin-to-pin and pin-
Following a similar process for ᾱ d̄ 1 yields [65] to-plate geometries [57], [62], [64], [65], so we focus here on
3/2 the theoretical implications under different conditions.
φ̄ Fig. 4 demonstrates the transition from FE-driven break-
V̄ ≈ d̄. (51)
β 2 down at small gap distance to the UPL at larger gap distance.
As an example, Fig. 2 shows 1 as a function of d̄ for While the limiting (50) does not pick up the transition to the
two values of p̄. Over approximately 2 orders of magnitude UPL because it assumed ᾱ d̄ 1, solving the exact equation
change of d̄ in the FE regime, 1 decreases from 20 to (37) numerically appropriately accounts for this transition
15 with decreasing gap distance, demonstrating that it changes upon satisfying (46), after which it satisfies the UPL. Plotting
by less than a factor of ∼0.8. Thus, the change in d̄ will the exact numerical solution of (37) and the limiting equation
clearly dominate the overall ratio. Similarly, Fig. 3 shows 1 (50) on a linear scale in Fig. 4(b) and (d) demonstrates the
as a function of p̄ for two values of d̄. Over 3 orders of transition to the FE-dominated regime, where V̄ ∝ d̄. As
magnitude of pressure, 1 only varies by a factor of 1.5. This discussed above and shown in Fig. 2, 1 changes much more
indicates that 1 is relatively insensitive to both gap distance slowly than d̄. This causes V̄ in (46) to decrease linearly with
and pressure, so (50) indicates V̄ ∝ d̄ in the FE regime. d̄ in the FE regime in Fig. 4(b) and (d).
Another point discussed in previous studies concerns the
dominant mechanisms with respect to ᾱ d̄. Fig. 5 shows ᾱ d̄ as
C. Applying the Matched Asymptotic Solutions to Microscale a function of d̄ based on the exact numerical solution of (37)
Gas Breakdown Experiments for two different pressures. In the FE regime, ᾱ d̄ 1. As we
Now that we have derived the fundamental universal equa- approach the transition to the TA regime, indicated by μ̃1 = ν̃1
tions for the unified FE–TA breakdown, the matched asymp- in Fig. 6 (d̄ 106 for p̄ = 4.47 × 10−6 and d̄ 107 for p̄ =
totic solutions under the limits of small and large ionization, 5.89 × 10−7 ), ᾱ d̄ > 1. In this case, very quickly upon
Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
816 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO. 4, APRIL 2020

distances studied. Thus, even in the smaller gap case when PL

does not occur, the transition from FE to TA still occurs.
At sufficiently high ᾱ d̄, the exact numerical solution of (37)
satisfies (47), as shown in Fig. 9 for the larger gap distance,
and V̄ subsequently follows the UPL in Fig. 7(b) once ᾱ d̄
remains constant. For the smaller gap distance where V̄ does
not follow the UPL, ᾱ d̄ never satisfies (47) (peak ᾱ d̄ = 6.79
compared with ᾱ d̄ = 6.91 for the UPL) and then decreases.
Although the reason for this behavior is unclear at this point,
it is most likely based on the assumptions made in deriving
the theory, notably that FE and TA are the main drivers of gas
breakdown, γSE is constant, and ᾱ d̄ follows a relatively simple
semiempirical model. As gap distance decreases to nanoscale,
electron emission transitions from FE to SCLE, which may
add an extra component to the space-charge component in
deriving the breakdown condition. Second, γSE may not
Fig. 5. Product of the ionization coefficient and dimensionless gap distance, be constant as some theories propose semiempirical models
ᾱ d̄, as a function of dimensionless gap distance d̄ for dimensionless pressures
of p̄ = 4.47 × 10−6 and p̄ = 5.89 × 10−7 . where it depends on the electric field [91]. Finally, ᾱ d̄ may not
follow the actual simple semiempirical models typically used,
particularly at microscale. Section IV outlines the PIC/MCC
simulations for better characterizing this behavior.
One explanation is that the numerical solution upon satisfy-
ing the UPL condition for the small gap distance may require
additional sensitivity to pick up the UPL solution. These
results highlight the importance of continuing experiments
at smaller gap distances and lower pressures to characterize
breakdown phenomena and perform more detailed simulations
of the ionization coefficient, field enhancement factor, and sec-
ondary emission coefficient to incorporate into these theories
to make them more predictive.

IV. PARTICLE - IN -C ELL /M ONTE C ARLO C OLLISION

(PIC/MCC) S IMULATIONS OF I ONIZATION C OEFFICIENT
As described in detail earlier, the ionization coefficient
Fig. 6. FE component μ̃ and Townsend component ν̃ as a function of
describes the volume processes contributing to the electron
dimensionless gap distance d̄ for dimensionless pressures of p̄ = 4.47 × 10−6 avalanche and eventual breakdown. Therefore, it is an impor-
and p̄ = 5.89 × 10−7 . tant input parameter for describing discharge formation and
microdischarges are no exception. Despite the unique oper-
ating regimes of FE-driven microdischarges, approaches to
transition to the TA regime, ᾱ d̄ ≈ 6.9, corresponding to the the UPL using empirical data for the ionization coefficient
transition to the UPL calculated based on (48). Interestingly, were largely based on macroscale measurements performed in
this indicates that ᾱ d̄ is no longer a function of Ē in the centimeter-scale gaps filled with gas at low pressure. While
UPL regime unless γSE is independently a function of Ē or a correction that would account for microscale gaps was
d̄ [68]–[70], [88]–[90]. proposed recently and has been used with some success in
We observe similar behavior when assessing V̄ as a function various UPL approaches, it still uses information from the
of p̄, as shown in Fig. 7. As discussed above and demonstrated empirical correlations in (15) and (16) based on macroscale
in Fig. 3, 1 is relatively constant with respect to p̄ in the measurements as the baseline. Specifically, while the correc-
FE regime, so while V̄ varies wth p̄ in the FE regime, its tion in (17) accounts for the effective decrease in the ionization
dependence is weaker than that on d̄. Fig. 7(a) shows that coefficient at applied voltages that are on the same order as
the breakdown voltage does not follow PL regardless of p̄ for the ionization potential, it does not account for the macroscale
d̄ = 2.25 × 106 ; at d̄ = 5.10 × 106 , the standard transition to ionization coefficient correlations in (15) and (16) failing at
PL occurs. This indicates that some minimum d̄ is required high reduced electric fields. For example, the classical text
for the transition to PL to occur. We can better understand this by Raizer [23] specifies a limit of about 700 V/cm/torr for
behavior by examining the transition from FE to TA and the the validity of the empirical correlation of the type in (15)
behavior of ᾱ d̄. and notes that the ionization coefficient would decrease with
Fig. 8 shows μ̃1 and ν̃1 as functions of p̄ with the transition further increase in E/ p beyond 2000 V/cm/torr. However,
from FE to TA occurring for 10−6 < p̄ < 10−5 for both gap this was deemed unimportant because traditional electrical

Fig. 7. Dimensionless breakdown voltage V̄ as a function of dimensionless pressure p̄ for dimensionless gap distances of (a) d̄ = 2.25 × 106 and
(b) d̄ = 5.10 × 106 .

Fig. 8. FE component, μ̃, and Townsend component, ν̃, as a function of Fig. 9. Product of ionization coefficient and dimensionless gap distance, ᾱd̄,
dimensionless pressure, p̄, for dimensionless gap distances of d̄ = 2.55 × 106 as a function of dimensionless pressure, p̄, for dimensionless gap distances
and d̄ = 5.10 × 106 . of d̄ = 2.55 × 106 and d̄ = 5.10 × 106 .

discharges would seldom operate at such extreme conditions. This example clearly demonstrates the importance of study-
Microdischarges routinely violate this assumption by operating ing the microscale ionization coefficient in greater detail either
at E/P above this limit for the validity of the empirical experimentally or numerically rather than making predictions
correlations for the ionization coefficient. based on empirical correlations that may be invalid in con-
Fig. 10 shows the microscale breakdown data for argon ditions where microdischarges may operate. In this tutorial,
measured over a range of operating conditions including we focus on 1-D PIC/MCC simulations similar to those used to
pressure and gap size. This clearly indicates the unique oper- propose the microscale correction to the macroscale ionization
ating regimes of microdischarges that make it necessary to coefficient in (17) and to determine the ionization coefficients
revisit our knowledge and understanding of the ionization at high E/ p as reported in Fig. 11.
coefficients. The 1-D PIC/MCC simulations performed for Computational techniques used to determine the ionization
argon confirm that the ionization coefficient decreases with coefficient to serve as an input to various analytical models
increasing E/ p beyond a certain point with the results sum- (including breakdown predictors and plasma sheath models)
marized in Fig. 11. The results of the PIC–MCC simulations can be classified based on their dimensionality and the method
were used to determine the values of gas-specific constants used to determine the electron energy distribution function
C and D from (15) for three different ranges of E/P to (EEDF), which, in turn, determines the ionization coefficient.
augment the traditionally used values of C and D for argon. One of the most common assumptions used to determine
We emphasize that the traditional values of C and D, indeed, the ionization coefficient numerically is the zero-dimensional
fit the ionization coefficient obtained using the PIC–MCC approximation, in which the electrons are subjected to a
simulations for sufficiently small E/ p. uniform electric field (and, consequently, reduced electric

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
818 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO. 4, APRIL 2020

On the other hand, the ionization coefficient may also be

computed using 1-D PIC/MCC simulations where a known
electron source at the cathode introduces a certain current
density of electrons into the gap [32]. The evolution of these
electrons is tracked as in a standard PIC/MCC simulation
including a move step, a field solve step, and a collisions step.
Once the simulation reaches a steady state (confirmed using
the number of electrons in the computational domain), the total
current density in the gap (same as the electron current density
at the anode) is used to determine the ionization coefficient in
the gap as

1 je,anode
α = ln (52)
d je,cathode
where je,anode and je,cathode are the electron current density at
the anode and the cathode, respectively.
With both zero-dimensional approaches, the EEDF is
uniquely determined by the specified value of the reduced
Fig. 10. Summary of microscale gas breakdown data for argon gas at various electric field E/N with no dependence on the applied volt-
gap sizes and pressures and comparison to the limit for the validity of the
classical empirical calculation based on [23].
age. However, as discussed earlier, the microscale ionization
coefficients depend on the applied voltage, which forms the
basis for the empirical correction in (17) to estimate them.
Clearly, the ability of an electron to ionize an atom/molecule
depends on the potential difference it experiences during its
journey from the boundary (from sources such as secondary
emission and FE). This indicates that the EEDF also depends
on the potential difference apart from the reduced electric field.
Here, it is worth reiterating that (17) was based on the average
ionization coefficient in the entire gap.
While the average ionization coefficient in the gap may well
be sufficient for the purposes of the UPL, it is interesting
to study the variation in the EEDF and hence the ionization
coefficient at various locations in the gap with the goal of
eventually eliminating the empiricism in (17). We selected a
gap size d = 200 μm filled with argon gas for this study.
While a standard PIC–MCC algorithm (using the XPDP1 code
[95], [96]) was used for the move and collide step, the field
solve was not performed and the electric field was hard-coded
as E = V /d with V the applied potential at the anode. This
ensured that the electric field was not affected even by the
small space charge that developed in the gap that can cause
small variations in the electric field across the gap. Omitting
Fig. 11. Ionization coefficient α/ p computed as a function of E/ p using
1-D PIC/MCC simulations and the corresponding empirical parameters. The
the field solve step allowed us to eliminate even these small
experimental data are obtained indirectly using the breakdown data for variations in electric field across the gap. We used a cell size
100 μm based on γSE = 0.005. of 2 μm and a time step of 20 fs for all the simulations reported
here. The ratio of real to simulated particles was chosen such
that the total number of simulated particles at steady state
field E/N where N is the background gas number density) was about 150 000, yielding approximately 1500 particles per
with no spatial dependence for any of the properties. This cell. The source of electrons at the cathode (right boundary
approximate EEDF is determined either using a two-term corresponding to X = 200 μm) was set to a current density
approximation such as in BOLSIG+ [92] or using a Monte of 1000 A/m2 . Other boundary processes such as secondary
Carlo technique such as in MAGBOLTZ [93] or METHES emission and FE were turned OFF because the primary goal
[94]. While the predictions of the two methods certainly differ of these simulations was to study the volume process of
under certain conditions, the two-term approximation and the electron multiplication (ionization). Multiple simulations were
more accurate Monte Carlo approach generally agree well, performed for a given value of E/N achieved by suitably
making BOLSIG+ a popular choice for determining the EEDF varying the applied voltage and the background gas pressure
because of its favorable computational cost compared with the to study the dependence of the ionization coefficient on the
Monte Carlo simulations. applied voltage at a fixed E/N. Once the simulations reached

Fig. 12. PIC/MCC results for the spatial profiles of the ionization coefficient
α/N for various values of applied voltage with E/N fixed at 200 Td. Note Fig. 13. PIC/MCC results for the local ionization coefficient α/N as a
that X = 200 μm corresponds to the cathode. function of the local potential for various values of applied voltage with E/N
fixed at 200 Td.

steady state, the electron current density profile was used to

obtain the local ionization coefficient as
1 d je
α= . (53)
je d x
The first set of simulations were performed for
E/N = 200 Td with applied voltages ranging from 20 to
400 V and the corresponding pressures ranging from 15.53 to
310.53 torr. It should be mentioned that low operating
voltages are particularly interesting for applications such as
gas ionization sensors [97] that involve the gas breakdown
of microgaps. Fig. 12 shows the variation in the ionization
coefficient across the gap at different applied voltages. The
ionization coefficient profiles at low applied voltages oscillate
with local maxima and minima in the ionization profiles.
The amplitudes of these oscillations decrease with increasing
distance from the cathode. As the applied voltage is increased,
the ionization coefficient reaches a constant value in almost
the entire gap with the oscillations restricted to a small
region near the cathode. Plotting the ionization coefficient Fig. 14. PIC/MCC results for the spatial profiles of the electron temperature
as a function of the local potential for each applied voltage for various values of applied voltage with E/N fixed at 200 Td.
in Fig. 13 further elucidates this behavior. The ionization
coefficient profiles overlap with each other except for a
small region near the anode. This is not very surprising It is interesting to compare how the findings of these
and confirms that the EEDF at a given location depends on PIC/MCC simulations compare with other existing tools com-
the potential difference through which the electrons have monly used for determining the ionization coefficient. Using
traveled apart from the applied reduced electric field E/N. BOLSIG+, α/N = 3.45 × 10−21 m2 at E/N = 200 Td is
This is a direct consequence of the number of collisions in obtained as 3.45 × 10−21 m2 , which agrees well with the
which the electrons participate. Once the electrons undergo PIC/MCC simulations once the electrons have achieved a
a sufficient number of collisions (which is equivalent to sufficient number of collisions. The PIC/MCC simulations
reaching a sufficiently large value of potential at a given were repeated for E/N = 2500 Td with Fig. 15 showing the
E/N), the EEDF reaches a steady state and stops changing ionization profile as a function of the local potential. Unlike
with location. This is confirmed by the electron temperature the results for E/N = 200 Td, the ionization profiles do
(two-third of the mean electron energy) profiles shown not oscillate and achieve the constant value monotonically.
in Fig. 14. In addition, the constant value of α/N occurs at a higher

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
820 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO. 4, APRIL 2020

Fig. 15. PIC/MCC results for the local ionization coefficient α/N as a Fig. 16. Variation in the ratio of microscale to macroscale α/N as a function
function of the local potential for various levels of applied voltage with E/N of applied voltage for all simulations considered in this work. Also included
fixed at 2500 Td. for reference is the variation proposed by the empirical correlation in (54).

value of the local potential because higher E/N corresponds that the simulations used to fit the empirical correlation were
to a lower background number density for a given applied not based on a single value of E/N with N fixed at a value
potential. This requires the electrons to travel longer distances corresponding to the atmospheric pressure and the voltage
before achieving a constant α/N. Using BOLSIG+, α/N = varied. Therefore, it is not surprising that the correlation has
39.65 × 10−21 m2 at E/N = 2500 Td, which agrees with the limited accuracy for pressures that are significantly different
constant value obtained from the PIC/MCC simulations. from those used to fit the data. Nevertheless, it still has
The oscillations in α/N at lower E/N arise due to the reasonable accuracy for both E/N values considered in this
differences in the distance that the electrons travel under these article, which makes it a good correlation to use in the absence
conditions. Increasing E/N is equivalent to increasing E for of additional data and a reasonable first step for the analytic
a given background N, which fixes the mean free path of studies performed earlier. However, it is clear that more work
electrons colliding with the background gas. To participate is required to fine-tune the models of the ionization coefficient
in ionizing collisions, the electrons must gain an energy that would work well for low applied voltages and high E/N.
equivalent to the ionization potential of the background gas The PIC/MCC simulations such as the ones reported here are
(∼16 eV here) for which they must travel through an equiva- excellent candidates to achieve this goal and will certainly help
lent potential difference. At low E/N, the electrons must travel improve the predictive capability of models such as the UPL.
a certain distance x to gain an energy equal to the ionization
potential; however, this distance is almost negligible at high
E/N. As a result, the EEDF at high E/N is almost stationary V. C ONCLUSION
(spatially), while the EEDF at low E/N clearly manifests
This tutorial has highlighted the application of theoretical
the initial overpopulation of energetic electrons (high α/N)
and computational techniques to predict gas breakdown for
followed by loss of the energy through the ionizing collisions
microscale gaps. We have summarized the existing approaches,
(low α/N).
predominantly focusing on the application of a matched
Fig. 16 shows the variation in the average value of the
asymptotic analysis to a theory derived from the coupling of
ionization coefficient in the gap (normalized with respect to
FE and TA phenomena. This analysis permitted the derivation
the constant values) with the applied potential for both E/N
of closed-form solutions to demonstrate the transition from FE
values reported in this article and the ratio of the corrected
to TA to the UPL and characterize the relative contribution
α/ p to the empirical (α/ p)Marcoscale from (17), given by
from each mechanism.

α α V /Vip −1.0 0.8 One challenge with these theoretical approaches is that
/ = 1 − exp − . certain parameters, particularly the ionization coefficient, use
p p Macroscale 3.1
semiempirical models that are not strictly valid under the
(54)
high electric field conditions present during microscale gas
The comparison clearly shows that the empirical correlation breakdown. The PIC/MCC simulations provide a way to
could deviate significantly from the values obtained from the determine the behavior of this parameter for various pressures
PIC/MCC simulations in this article. One potential reason is and electric fields to potentially update the semiempirical

approaches typically found in the literature to incorporate into yield CL at atmospheric pressure (not accounting for break-
this theory. down) [75]. Performing these experiments at various pressures
In addition to improving the characterization of the and gap distances will further elucidate these mechanisms and
ionization coefficient α, the current theoretical approach potentially show transition to the quantum-CL [106]–[109] at
to modeling microscale gas breakdown requires determin- sufficiently small gap sizes. Using a common set of scaling
ing β by fitting to the FN law or breakdown equation parameters could permit the derivation of a group of coupled
[29], [57], [61]–[65], [98]. This hinders the application of theories describing electron emission and breakdown across
this theory for predicting breakdown a priori. Several studies these regimes. Similarly, one could incorporate an updated
have indicated that β varies with electric field [58], [59], model of FN [110] or unified theory for field and thermionic
[65], which may at least simplify the fitting by providing a emission into the microscale gas breakdown theory [111]
single constant (say amplitude) rather than fitting β at each to include more physics and fully demonstrate the coupling
gap distance studied. Predicting β may also be simplified for between these mechanisms. Again, care must be taken to
a given geometry using the analytic equation for β [24] and balance physical completeness with the utility provided by
then adjusting accordingly. This approach may be incorporated being able to derive closed-form solutions to understand
directly into (6) and (15)–(17), which will then propagate to limiting behavior. Future studies will continue incrementally
(34) and (37); however, this approach may complicate the introducing more physics into the theories and exploring the
subsequent derivation of closed-form solutions for Vb due to implications numerically, theoretically, and compared with
the additional dependence on E. Experiments assessing the experiments to determine the appropriate balance between
impact of different aspect ratios under different microscale complexity and simplicity to elucidate the underlying physics
(and smaller) gap distances and pressures may further eluci- and provide a predictive tool for engineers and scientists.
date the impact of β and permit direct assessment of other
factors that may influence the effective β observed from the
R EFERENCES
fits [99]. One potential contributory effect could be space
charge, which could influence the electric field profile in [1] S. Li et al., “Synergic effect of adsorbed gas and charging on surface
the gap and potentially cause electron emission to transition flashover,” Sci. Rep., vol. 9, no. 1, Dec. 2019, Art. no. 5464.
[2] A. L. Garner et al., “Design, characterization and experimental valida-
FE to SCLE at sufficiently small gap size [75]. Molecular tion of a compact, flexible pulsed power architecture for ex vivo platelet
dynamics simulations [100] assessing the relevant geometries activation,” PLoS ONE, vol. 12, no. 7, 2017, Art. no. e0181214.
may provide additional valuable input to this phenomenon for [3] W. Zhong, G. Zhang, and A. Xu, “Detachment of microparticles from
electrode erosion and the subsequent breakdown,” AIP Adv., vol. 9,
the analysis, particularly for understanding the electric field no. 4, Apr. 2019, Art. no. 045023.
variation in a small gap with nonplanar geometry. [4] A. Darwish, S. S. Refaat, H. A. Toliyat, and H. Abu-Rub, “On the
Similarly, γSE is often defined as a constant determined electromagnetic wave behavior due to partial discharge in gas
insulated switchgears: State-of-art review,” IEEE Access, vol. 7,
by fitting to experimental data; however, in principle, it pp. 75822–75836, 2019.
may depend on E, d, p and the cathode surface conditions [5] P. Zhao and L. Guo, “Effect of air breakdown in near-field region
[67]–[70], [86]–[88]. Again, the balance between the com- on maximum power radiated from aperture antenna,” J. Electromagn.
Waves Appl., vol. 30, no. 6, pp. 795–804, Apr. 2016.
pleteness of the theory and the ability to extract useful [6] S. Yan, A. D. Greenwood, and J.-M. Jin, “Simulation of high-power
information from the theory must be considered. Incorporating microwave air breakdown modeled by a coupled Maxwell–Euler sys-
these effects could be valuable, particularly when one starts to tem with a non-Maxwellian EEDF,” IEEE Trans. Antennas Propag.,
vol. 66, no. 4, pp. 1882–1893, Apr. 2018.
explore the impact of surface roughness in more detail [57]. [7] D. R. Welch, M. E. Cuneo, C. L. Olson, and T. A. Mehlhorn, “Gas
Surface roughness may also impact φ [101], which would also breakdown effects in the generation and transport of light ion beams
impact the behavior in the FE regime, so these effects could for fusion,” Phys. Plasmas, vol. 3, no. 5, pp. 2113–2121, May 1996.
[8] X. Song et al., “Experimental results of plasma breakdown and
complicate predictions in both FE and TA regimes. Again, flux optimization on HL-2A tokamak,” Fusion Eng. Des., vol. 125,
performing a matched asymptotic analysis could elucidate pp. 195–198, Dec. 2017.
the significance of these effects under each condition and [9] K. C. Hammond, R. Raman, and F. A. Volpe, “Application of townsend
avalanche theory to tokamak startup by coaxial helicity injection,” Nucl.
potentially yield simplifications. Fusion, vol. 58, no. 1, Jan. 2018, Art. no. 016013.
Finally, reducing the gap distance or changing pressure will [10] B. Lloyd, G. L. Jackson, T. S. Taylor, E. A. Lazarus, T. C. Luce, and
impact the phenomena involved in driving the mechanisms R. Prater, “Low voltage ohmic and electron cyclotron heating assisted
startup in DIII-D,” Nucl. Fusion, vol. 31, no. 11, pp. 2031–2053,
involved. Previous theoretical [102], simulation [103], and Nov. 1991.
experimental [104] results have demonstrated the transition [11] Y. Peng et al., “On the breakdown modes and parameter space of
from FN to CL in vacuum as one reduces the gap size. Pre- ohmic tokamak start-up,” J. Plasma Phys., vol. 84, no. 5, Oct. 2018,
vious theoretical studies have demonstrated similar behavior Art. no. 905840505.
[12] K. A. Rezvykh and V. A. Romanov, “Gases breakdown voltage
when collisions were included to demonstrate the transition calculation for the case of accelerator nonuniform fields by the method
from CL to MG with increasing pressure [105]. A more recent of base,” Nucl. Instrum. Methods Phys. Res. A, Accel. Spectrom. Detect.
study unified FN, CL, and MG and defined a “nexus” where all Assoc. Equip., vol. 423, nos. 2–3, pp. 203–212, Mar. 1999.
[13] A. M. R. N. Alrashidi, N. M. Adam, A. A. Hairuddin, and
three asymptotic solutions matched [75], which was typically L. C. Abdullah, “A review on plasma combustion of fuel in internal
on the order of several hundred nanometers at atmospheric combustion engines,” Int. J. Energy Res., vol. 42, no. 5, pp. 1813–1833,
pressure. This suggests that further reducing the gap size at Apr. 2018.
[14] J. Kolbeck, A. Anders, I. I. Beilis, and M. Keidar, “Micro-propulsion
atmospheric pressure could lead to transitions across these based on vacuum arcs,” J. Appl. Phys., vol. 125, no. 22, Jun. 2019,
mechanisms and, for a sufficiently high voltage, potentially Art. no. 220902.

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
822 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO. 4, APRIL 2020

[15] D. Lev, R. M. Myers, K. M. Lemmer, J. Kolbeck, H. Koizumi, and [42] Y. Fu, P. Zhang, and J. P. Verboncoeur, “Paschen’s curve in microgaps
K. Polzin, “The technological and commercial expansion of electric with an electrode surface protrusion,” Appl. Phys. Lett., vol. 113, no. 5,
propulsion,” Acta Astronautica, vol. 159, pp. 213–227, Jun. 2019. Jul. 2018, Art. no. 054102.
[16] Y. Ju and W. Sun, “Plasma assisted combustion: Dynamics and chem- [43] Y. Fu, P. Zhang, J. P. Verboncoeur, A. J. Christlieb, and X. Wang,
istry,” Progr. Energy Combustion Sci., vol. 48, pp. 21–83, Jun. 2015. “Effect of surface protrusion on plasma sheath properties in
[17] K. H. Becker, K. H. Schoenbach, and J. G. Eden, “Microplasmas and atmospheric microdischarges,” Phys. Plasmas, vol. 25, no. 1, Jan. 2018,
applications,” J. Phys. D: Appl. Phys., vol. 39, no. 3, pp. R55–R70, Art. no. 013530.
Feb. 2006. [44] Y. Fu, P. Zhang, and J. P. Verboncoeur, “Gas breakdown in atmospheric
[18] K. H. Schoenbach and K. Becker, “20 years of microplasma research: pressure microgaps with a surface protrusion on the cathode,” Appl.
A status report,” Eur. Phys. J. D, vol. 70, no. 2, Feb. 2016, Art. no. 29. Phys. Lett., vol. 112, no. 25, Jun. 2018, Art. no. 254102.
[19] K. Becker et al., “Environmental and biological applications of [45] Y. Fu, P. Zhang, J. Krek, and J. P. Verboncoeur, “Gas breakdown and its
microplasmas,” Plasma Phys. Controlled Fusion, vol. 47, no. 12B, scaling law in microgaps with multiple concentric cathode protrusions,”
pp. B513–B523, Dec. 2005. Appl. Phys. Lett., vol. 114, no. 1, Jan. 2019, Art. no. 014102.
[20] M. G. Kong et al., “Plasma medicine: An introductory review,” New [46] M. Radmilovic-Radjenovic, J. K. Lee, F. Iza, and G. Y. Park, “Particle-
J. Phys., vol. 11, no. 11, 2009, Art. no. 115012. in-cell simulation of gas breakdown in microgaps,” J. Phys. D: Appl.
[21] M. Laroussi, “Low-temperature plasma jet for biomedical applications: Phys., vol. 38, no. 6, pp. 950–954, Mar. 2005.
A review,” IEEE Trans. Plasma Sci., vol. 43, no. 3, pp. 703–712, [47] M. Radmilovic-Radjenovic and B. Radjenovic, “A particle-in-cell sim-
Mar. 2015. ulation of the high-field effect in devices with micrometer gaps,” IEEE
[22] F. Paschen, “Ueber die zum Funkenübergang in Luft, Wasserstoff Trans. Plasma Sci., vol. 35, no. 5, pp. 1223–1228, Oct. 2007.
und Kohlensäure bei verschiedenen Drucken erforderliche Potentiald- [48] M. Radmilovic-Radjenovic and B. Radjenovic, “A particle-in-cell
ifferenz,” Annalen der Physik, vol. 273, no. 5, pp. 69–96, 1889. simulation of the breakdown mechanism in microdischarges with an
[23] Y. P. Raizer, Gas Discharge Physics. Berlin, Germany: Springer, 1991. improved secondary emission model,” Contrib. Plasma Phys., vol. 47,
[24] Y. Fu, J. Krek, P. Zhang, and J. P. Verboncoeur, “Evaluating microgap no. 3, pp. 165–172, May 2007.
breakdown mode transition with electric field non-uniformity,” Plasma [49] Y. Li, R. Tirumala, P. Rumbach, and D. B. Go, “The coupling
Sources Sci. Technol., vol. 27, no. 9, 2018, Art. no. 095014. of ion-enhanced field emission and the discharge during microscale
[25] M. Keidar and I. I. Beilis, Plasma Engineering: Applications from breakdown at moderately high pressures,” IEEE Trans. Plasma Sci.,
Aerospace to Bio and Nanotechnology. New York, NY, USA: vol. 41, no. 1, pp. 24–35, Jan. 2013.
Academic, 2013, p. 109. [50] A. Venkattraman, A. Garg, D. Peroulis, and A. A. Alexeenko,
[26] L. B. Loeb and J. M. Meek, Mechanism of the Electric Spark. Stanford, “Direct measurements and numerical simulations of gas charging in
microelectromechanical system capacitive switches,” Appl. Phys. Lett.,
CA, USA: Stanford Univ. Press, 1941.
vol. 100, no. 8, Feb. 2012, Art. no. 083503.
[27] W. S. Boyle and P. Kisliuk, “Departure from Paschen’s law of break-
[51] D. Levko, “Electron kinetics in a microdischarge in nitrogen at
down in gases,” Phys. Rev., vol. 97, no. 2, pp. 255–259, Jan. 1955.
atmospheric pressure,” J. Appl. Phys., vol. 114, no. 22, Dec. 2013,
[28] D. B. Go and A. Venkattraman, “Microscale gas breakdown: Ion- Art. no. 223302.
enhanced field emission and the modified Paschen’s curve,” J. Phys. D:
[52] D. Levko, S. Yatom, and Y. E. Krasik, “Particle-in-cell modeling of the
Appl. Phys., vol. 47, no. 50, Dec. 2014, Art. no. 503001.
nanosecond field emission driven discharge in pressurized hydrogen,”
[29] A. M. Loveless and A. L. Garner, “A universal theory for gas break- J. Appl. Phys., vol. 123, no. 8, Feb. 2018, Art. no. 083303.
down from microscale to the classical paschen law,” Phys. Plasmas,
[53] D. Levko and L. L. Raja, “High-voltage microdischarge as a source
vol. 24, no. 11, Nov. 2017, Art. no. 113522.
of extreme density plasma,” Phys. Plasmas, vol. 25, no. 1, 2018,
[30] J.-M. Torres and R. S. Dhariwal, “Electric field breakdown at Art. no. 013509.
micrometre separations,” Nanotechnology, vol. 10, no. 1, pp. 102–107,
[54] L. Kong, W. Wang, A. B. Murphy, and G. Xia, “Numerical analysis
Mar. 1999.
of direct-current microdischarge for space propulsion applications
[31] R. H. Fowler and L. Nordheim, “Electron emission in intense electric using the particle-in-cell/Monte Carlo collision (PIC/MCC) method,”
fields,” Proc. Roy. Soc. London A, Containing Papers Math. Phys. J. Phys. D: Appl. Phys., vol. 50, no. 16, Apr. 2017, Art. no. 165203.
Character, vol. 119, no. 781, pp. 173–181, May 1928.
[55] S. S. Tholeti, A. Semnani, D. Peroulis, and A. A. Alexeenko, “Dark-
[32] A. Venkattraman and A. A. Alexeenko, “Scaling law for direct cur- to-arc transition in field emission dominated atmospheric microdis-
rent field emission-driven microscale gas breakdown,” Phys. Plasmas, charges,” Phys. Plasmas, vol. 22, no. 8, Aug. 2015, Art. no. 083508.
vol. 19, no. 12, Dec. 2012, Art. no. 123515.
[56] G. J. Kim, F. Iza, and J. K. Lee, “Field emission and lifetime
[33] R. Bogue, “MEMS sensors: Past, present and future,” Sensor Rev., of microcavity plasma,” Phys. Plasmas, vol. 16, no. 1, Jan. 2009,
vol. 27, no. 1, pp. 7–13, Jan. 2007. Art. no. 013502.
[34] C. B. Ru et al., “Design and fabrication of MEMS-based solid [57] R. S. Brayfield et al., “The impact of cathode surface roughness
propellant microthrusters array,” Appl. Mech. Mater., vols. 490–491, and multiple breakdown events on microscale gas breakdown at
pp. 1042–1046, Jan. 2014. atmospheric pressure,” J. Appl. Phys., vol. 125, no. 20, May 2019,
[35] H. Gensler, R. Sheybani, P.-Y. Li, R. L. Mann, and E. Meng, Art. no. 203302.
“An implantable MEMS micropump system for drug delivery in small [58] A. Venkattraman, “Electric field enhancement due to a saw-tooth
animals,” Biomed. Microdevices, vol. 14, no. 3, pp. 483–496, Jun. 2012. asperity in a channel and implications on microscale gas breakdown,”
[36] L. Boodhoo et al., “Fabrication and characterisation of suspended nar- J. Phys. D: Appl. Phys., vol. 47, no. 42, Oct. 2014, Art. no. 425205.
row silicon nanowire channels for low-power nano-electro-mechanical [59] J. A. Buendia and A. Venkattraman, “Field enhancement factor depen-
(NEM) switch applications,” Microelectron. Eng., vol. 145, pp. 66–70, dence on electric field and implications on microscale gas breakdown:
Sep. 2015. Theory and experimental interpretation,” Europhys. Lett., vol. 112,
[37] G. Shivkumar, L. Qiao, and A. A. Alexeenko, “Plasma-flow interactions no. 5, Dec. 2015, Art. no. 55002.
in field-emission discharges with applications in microcombustion,” [60] A. M. Loveless and A. L. Garner, “Scaling laws for gas breakdown
J. Phys. D: Appl. Phys., vol. 52, no. 38, Sep. 2019, Art. no. 384001. for nanoscale to microscale gaps at atmospheric pressure,” Appl. Phys.
[38] T. Takahashi, D. Mori, T. Kawanabe, Y. Takao, K. Eriguchi, and Lett., vol. 108, no. 23, Jun. 2016, Art. no. 234103.
K. Ono, “Microplasma thruster powered by X-band microwaves,” [61] A. M. Loveless and A. L. Garner, “Generalization of microdischarge
J. Appl. Phys., vol. 125, no. 8, Feb. 2019, Art. no. 083301. scaling laws for all gases at atmospheric pressure,” IEEE Trans. Plasma
[39] L. Cranberg, “The initiation of electrical breakdown in vacuum,” Sci., vol. 45, no. 4, pp. 574–583, Apr. 2017.
J. Appl. Phys., vol. 23, no. 5, pp. 518–522, May 1952. [62] G. Meng et al., “Demonstration of field emission driven microscale
[40] A. K. Verma, A. Alamatsaz, and A. Venkattraman, “A comparison gas breakdown for pulsed voltages usingin-situoptical imaging,” Phys.
of continuum and kinetic simulations of moderatepd microplasmas Plasmas, vol. 25, no. 8, Aug. 2018, Art. no. 082116.
integrated with high secondary yield cathodes,” Plasma Processes [63] S. D. Dynako, A. M. Loveless, and A. L. Garner, “Sensitivity of mod-
Polym., vol. 14, nos. 4–5, Apr. 2017, Art. no. 1600130. eled microscale gas breakdown voltage due to parametric variation,”
[41] D. Levko and L. L. Raja, “Fluid modeling of a high-voltage nanosecond Phys. Plasmas, vol. 25, no. 10, Oct. 2018, Art. no. 103505.
pulsed xenon microdischarge,” Phys. Plasmas, vol. 23, no. 7, Jul. 2016, [64] G. Meng et al., “Spatio-temporal dynamics of pulsed gas breakdown in
Art. no. 073513. microgaps,” Phys. Plasmas, vol. 26, no. 1, Jan. 2019, Art. no. 014506.

[65] A. M. Loveless et al., “The transition to Paschen’s law for microscale [89] V. P. Nagorny and P. J. Drallos, “Effective secondary emission coef-
gas breakdown at subatmospheric pressure,” Sci. Rep., vol. 9, no. 1, ficient in a high-pressure noble gas,” Plasma Sources Sci. Technol.,
Dec. 2019, Art. no. 5669. vol. 6, no. 2, pp. 212–219, May 1997.
[66] R. Tirumala and D. B. Go, “An analytical formulation for the modified [90] J. R. M. Vaughan, “A new formula for secondary emission yield,” IEEE
Paschen’s curve,” Appl. Phys. Lett., vol. 97, no. 15, Oct. 2010, Trans. Electron Devices, vol. 36, no. 9, pp. 1963–1967, Sep. 1989.
Art. no. 151502. [91] M. Radmilović-Radjenović, B. Radjenović, Š. Matejčík, and M. Klas,
[67] M. Radmilović-Radjenović and B. Radjenović, “Theoretical study of “The breakdown phenomena in micrometer scale direct-current gas
the electron field emission phenomena in the generation of a microm- discharges,” Plasma Chem. Plasma Process., vol. 34, no. 1, pp. 55–64,
eter scale discharge,” Plasma Sources Sci. Technol., vol. 17, no. 2, Jan. 2014.
May 2008, Art. no. 024005. [92] G. J. M. Hagelaar and L. C. Pitchford, “Solving the Boltzmann equation
[68] M. Radmilović-Radjenović and B. Radjenović, “An analytical rela- to obtain electron transport coefficients and rate coefficients for fluid
tion describing the dramatic reduction of the breakdown voltage for models,” Plasma Sources Sci. Technol., vol. 14, no. 4, pp. 722–733,
the microgap devices,” Europhys. Lett., vol. 83, no. 2, Jul. 2008, Nov. 2005.
Art. no. 25001. [93] G. W. Fraser and E. Mathieson, “Monte Carlo calculation of electron
[69] M. Klas, Š. Matejc̆ík, B. Radjenović, and M. Radmilović-Radjenović, transport coefficients in counting gas mixtures,” Nucl. Instrum. Methods
“Experimental and theoretical studies of the direct-current breakdown Phys. Res. A, Accel. Spectrom. Detect. Assoc. Equip., vol. 257, no. 2,
voltage in argon at micrometer separations,” Phys. Scripta, vol. 83, pp. 339–345, 1986.
no. 4, Apr. 2011, Art. no. 045503. [94] M. Rabie and C. M. Franck, “METHES: A Monte Carlo collision code
[70] M. Klas, Š. Matejc̆ík, B. Radjenović, and M. Radmilović-Radjenović, for the simulation of electron transport in low temperature plasmas,”
“Experimental and theoretical studies of the breakdown voltage char- Comput. Phys. Commun., vol. 203, pp. 268–277, Jun. 2016.
acteristics at micrometre separations in air,” Europhys. Lett., vol. 95, [95] J. P. Verboncoeur, M. V. Alves, V. Vahedi, and C. K. Birdsall,
no. 3, Aug. 2011, Art. no. 35002. “Simultaneous potential and circuit solution for 1D bounded plasma
[71] M. Radmilović-Radjenović, Š. Matejc̆ík, M. Klas, and B. Radjenović, particle simulation codes,” J. Comput. Phys., vol. 104, pp. 321–328,
“The role of the field emission effect in direct-current argon discharges Feb. 1993.
for the gaps ranging from 1 to 100 μm,” J. Phys. D: Appl. Phys., [96] J. Verboncoeur and C. K. Birdsall. The Plasma Theory and Simula-
vol. 46, no. 1, Jan. 2013, Art. no. 015302. tion Group. Accessed: Oct. 21, 2019. [Online]. Available: http://ptsg.
[72] D. B. Go and D. A. Pohlman, “A mathematical model of the modified egr.msu.edu/
Paschen’s curve for breakdown in microscale gaps,” J. Appl. Phys., [97] R. Savari, H. Savaloni, S. Abbasi, and F. Placido, “Design and engi-
vol. 107, no. 10, May 2010, Art. no. 103303. neering of ionization gas sensor based on Mn nano-flower sculptured
[73] A. Semnani, A. Venkattraman, A. A. Alexeenko, and D. Peroulis, “Pre- thin film as cathode and a stainless steel ball as anode,” Sens. Actuators
breakdown evaluation of gas discharge mechanisms in microgaps,” B, Chem., vol. 266, pp. 620–636, Aug. 2018.
Appl. Phys. Lett., vol. 102, no. 17, Apr. 2013, Art. no. 174102. [98] M. A. Bilici, J. R. Haase, C. R. Boyle, D. B. Go, and
[74] Š. Matejc̆ík, B. Radjenović, M. Klas, and M. Radmilović-Radjenović, R. M. Sankaran, “The smooth transition from field emission
“Field emission driven direct current argon discharges and electrical to a self-sustained plasma in microscale electrode gaps at
breakdown mechanism across micron scale gaps,” Eur. Phys. J. D, atmospheric pressure,” J. Appl. Phys., vol. 119, no. 22, Jun. 2016,
vol. 69, no. 11, Nov. 2015, Art. no. 251. Art. no. 223301.
[75] A. M. Darr, A. M. Loveless, and A. L. Garner, “Unification of field [99] J. Lin, P. Y. Wong, P. Yang, Y. Y. Lau, W. Tang, and P. Zhang,
emission and space charge limited emission with collisions,” Appl. “Electric field distribution and current emission in a miniaturized
Phys. Lett., vol. 114, no. 1, Jan. 2019, Art. no. 014103. geometrical diode,” J. Appl. Phys., vol. 121, no. 24, Jun. 2017,
Art. no. 244301.
[76] N. F. Mott and R. W. Gurney, Electronic Processes in Ionic Crystals.
[100] K. Torfason, A. Valfells, and A. Manolescu, “Molecular dynamics
Oxford, U.K.: Clarendon Press, 1940.
simulations of field emission from a prolate spheroidal tip,” Phys.
[77] I. Langmuir, “Convection and conduction of heat in gases,” Phys. Rev.,
Plasmas, vol. 23, no. 12, Dec. 2016, Art. no. 123119.
vol. 34, no. 6, pp. 401–423, 1912.
[101] W. Li and D. Y. Li, “On the correlation between surface roughness and
[78] I. Langmuir, “The effect of space charge and residual gases on work function in copper,” J. Chem. Phys., vol. 122, no. 6, Feb. 2005,
thermionic currents in high vacuum,” Phys. Rev., vol. 2, no. 6, Art. no. 064708.
pp. 450–486, Dec. 1913. [102] Y. Y. Lau, Y. Liu, and R. K. Parker, “Electron emission: From the
[79] C. D. Child, “Discharge from hot CaO,” Phys. Rev., vol. 32, no. 5, Fowler–Nordheim relation to the Child–Langmuir law,” Phys. Plasmas,
pp. 492–511, May 1911. vol. 1, pp. 2082–2085, Feb. 1994.
[80] D. B. Go, “Theoretical analysis of ion-enhanced thermionic emission [103] Y. Feng and J. P. Verboncoeur, “Transition from Fowler-Nordheim
for low-temperature, non-equilibrium gas discharges,” J. Phys. D: Appl. field emission to space charge limited current density,” Phys. Plasmas,
Phys., vol. 46, no. 3, Jan. 2013, Art. no. 035202. vol. 13, no. 7, Jul. 2006, Art. no. 073105.
[81] J. R. Haase and D. B. Go, “Analysis of thermionic and thermo-field [104] S. Bhattacharjee and T. Chowdhury, “Experimental investigation of
emission in microscale gas discharges,” J. Phys. D: Appl. Phys., vol. 49, transition from Fowler–Nordheim field emission to space-charge-
no. 5, Feb. 2016, Art. no. 055206. limited flows in a nanogap,” Appl. Phys. Lett., vol. 95, no. 6, Aug. 2009,
[82] M. U. Lee, J. Lee, J. K. Lee, and G. S. Yun, “Extended scaling Art. no. 061501.
and paschen law for micro-sized radiofrequency plasma breakdown,” [105] M. S. Benilov, “The Child–Langmuir law and analytical theory of
Plasma Sources Sci. Technol., vol. 26, no. 3, 2017, Art. no. 034003. collisionless to collision-dominated sheaths,” Plasma Sources Sci.
[83] M. U. Lee, J. Lee, G. S. Yun, and J. K. Lee, “Scalings and universality Technol., vol. 18, no. 1, Feb. 2009, Art. no. 014005.
for high-frequency excited high-pressure argon microplasma,” Eur. [106] Y. Y. Lau, D. Chernin, D. G. Colombant, and P.-T. Ho, “Quantum
Phys. J. D, vol. 71, no. 4, Apr. 2017, Art. no. 94. extension of Child-Langmuir law,” Phys. Rev. Lett., vol. 66, no. 11,
[84] Y. Li and D. B. Go, “The quantum mechanics of ion-enhanced field pp. 1446–1449, Mar. 1991.
emission and how it influences microscale gas breakdown,” J. Appl. [107] L. K. Ang, T. J. T. Kwan, and Y. Y. Lau, “New scaling of Child-
Phys., vol. 116, no. 10, Sep. 2014, Art. no. 103306. Langmuir law in the quantum regime,” Phys. Rev. Lett., vol. 91, no. 20,
[85] X. Tan, P. Rumbach, N. Griggs, K. L. Jensen, and D. B. Go, “The- Nov. 2003, Art. no. 208303.
oretical analysis of 1D resonant tunneling behavior in ion-enhanced [108] L. K. Ang, Y. Y. Lau, and T. J. T. Kwan, “Simple derivation of quantum
cold field and thermo-field emission,” J. Appl. Phys., vol. 120, no. 21, scaling in Child-Langmuir law,” IEEE Trans. Plasma Sci., vol. 32,
Dec. 2016, Art. no. 213301. no. 2, pp. 410–412, Jun. 2004.
[86] L. H. G. Huxley, R. W. Crompton, and M. T. Elford, “Use of the [109] S. Bhattacharjee, A. Vartak, and V. Mukherjee, “Experimental study of
parameter E/N,” Brit. J. Appl. Phys., vol. 17, pp. 1237–1238, Jul. 1966. space-charge-limited flows in a nanogap,” Appl. Phys. Lett., vol. 92,
[87] A. M. Darr and A. L. Garner, “A coordinate system invariant formu- no. 19, May 2008, Art. no. 191503.
lation for space-charge limited current in vacuum,” Appl. Phys. Lett., [110] R. G. Forbes and J. H. B. Deane, “Correction for forbes and deane,
vol. 115, no. 5, Jul. 2019, Art. no. 054101. reformulation of the standard theory of Fowler–Nordheim tunnelling
[88] A. V. Phelps and Z. L. Petrovic, “Cold-cathode discharges and and cold field electron emission,” Proc. Roy. Soc. A, Math., Phys. Eng.
breakdown in argon: Surface and gas phase production of secondary Sci., vol. 464, no. 2100, p. 3378, Dec. 2008.
electrons,” Plasma Sources Sci. Technol., vol. 8, no. 3, pp. R21–R44, [111] K. L. Jensen, “A reformulated general thermal-field emission equation,”
Aug. 1999. J. Appl. Phys., vol. 126, no. 6, 2019, Art. no. 065302.

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.
824 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO. 4, APRIL 2020

Allen L. Garner (Senior Member, IEEE) received Jiba Nath Dahal (Member, IEEE) received the
the B.S. degree (Hons.) in nuclear engineering from B.S. and M.S. degrees in physics from Tribhuvan
the University of Illinois at Urbana–Champaign, University, Nepal, in 2004 and 2006, respectively,
Champaign, IL, USA, in 1996, the M.S.E. degree in the M.S. degree in physics and materials science
nuclear engineering from the University of Michi- from the University of Memphis, Memphis, TN,
gan, Ann Arbor, MI, USA, in 1997, the M.S. degree USA, in 2013, and the Ph.D. degree in physics from
in electrical engineering from Old Dominion Uni- the University of California Merced, Merced, CA,
versity, Norfolk, VA, USA, in 2003, and the Ph.D. USA, in 2018.
degree in nuclear engineering from the University of He joined Georgia Southern University, Savannah,
Michigan in 2006. GA, USA, in 2018, as a Visiting Assistant Professor.
From 1997 to 2003, he was an Active Duty Naval He is currently an Assistant Professor of physics
Officer serving onboard the USS Pasadena (SSN 752) and as an Instructor with Truman State University, Kirksville, MO, USA. His research interest
of the Prospective Nuclear Engineering Officer course at Submarine Training includes theoretical calculations of gas breakdown and electron emission
Facility, Norfolk. From 2006 to 2012, he was an Electromagnetic Physicist in microscale plasmas and experimental studies of magnetic nanomaterials,
with GE Global Research Center, Niskayuna, NY, USA. In 2012, he joined oxide-based Sr/Ba ferrites, and rare-earth-based permanent magnets.
the School of Nuclear Engineering, Purdue University, West Lafayette, IN, Dr. Dahal received the Best Graduate Student Award in 2014 from the
USA, where he is currently an Associate Professor and the Undergraduate Department of Physics, University of California Merced.
Program Chair. He is currently a Captain in the Navy Reserves assigned as
Commanding Officer of the ONR S&T 102 Unit, Washington, DC, USA.
His research interests include electron emission, gas breakdown, high-power
microwaves, and biomedical applications of pulsed power and plasmas.
Dr. Garner received the University of Michigan Reagents’ Fellowship,
the National Defense Science and Engineering Graduate Fellowship, two
Meritorious Service Medals, the Navy and Marine Corps Commendation
Medal, and five Navy and Marcine Corps Achievement Medals. He is a
licensed Professional Engineer in Michigan.

Amanda M. Loveless (Student Member, IEEE)

received the B.S. and M.S. degrees in nuclear engi-
neering from Purdue University, West Lafayette, IN, Ayyaswamy Venkattraman (Member, IEEE)
USA, in 2015 and 2017, respectively. received the B.S. degree in aerospace engineering
She is currently a Graduate Student in nuclear from the IIT Madras, Chennai, India, in 2007,
engineering with Purdue University. Her research and the M.S. and Ph.D. degrees in aeronautics
interests include theoretical studies of gas break- and astronautics from Purdue University, West
down and electron emission. Lafayette, IN, USA, in 2009 and 2012, respectively.
Ms. Loveless received the 2016 Krauss Schol- In 2014, he joined the Department of Mechanical
arship from Purdue University, the 2016 IEEE Engineering, University of California Merced,
Dielectrics and Electrical Insulation Society Grad- Merced, CA, USA, as an Assistant Professor. His
uate Fellowship, the 2017 IEEE Nuclear and Plasma Sciences Society research interests include computational methods
(NPSS) Graduate Scholarship, the 2018 Igor Alexeff Outstanding Student for low-temperature plasmas with an emphasis in
in Plasma Science Award, and three Directed Energy Professional Society microplasmas.
(DEPS) Scholarships. She also won the 1st place for “Best Student Paper” Dr. Venkattraman was a recipient of the Hellman Faculty Fellowship,
at the 2016 International Conference on Plasma Science (ICOPS) and the Magoon Award for Excellence in Teaching and was a nominee for the
2016 Electrostatic Society of America (ESA) Annual Meeting. Chorafas Foundation Awards.

Authorized licensed use limited to: CZECH TECHNICAL UNIVERSITY. Downloaded on May 22,2024 at 15:30:52 UTC from IEEE Xplore. Restrictions apply.

Grounding and Bonding For The Radio Amateur
100% (3)
Grounding and Bonding For The Radio Amateur
260 pages
Bonaventure Creation
100% (3)
Bonaventure Creation
352 pages
Solution Chapter 2 2
100% (2)
Solution Chapter 2 2
43 pages
Pachin PHD Thesis Novel Environmentally Friendly Insulating Gas Mixtures
No ratings yet
Pachin PHD Thesis Novel Environmentally Friendly Insulating Gas Mixtures
322 pages
Breakdown in Gases
100% (2)
Breakdown in Gases
49 pages
Breakdown Mechanism in Gases
0% (1)
Breakdown Mechanism in Gases
19 pages
Pedersen
No ratings yet
Pedersen
20 pages
Institute of Aeronautical Engineering: Electrical and Electronics Engineering
No ratings yet
Institute of Aeronautical Engineering: Electrical and Electronics Engineering
194 pages
High Voltage Module 1
No ratings yet
High Voltage Module 1
55 pages
LECTURE-3 & 4 (A) Mechanism of Spark BD in Gases
50% (2)
LECTURE-3 & 4 (A) Mechanism of Spark BD in Gases
80 pages
Modified Paschen S Law
No ratings yet
Modified Paschen S Law
6 pages
2009 JAP Go Pohlman Revision2
No ratings yet
2009 JAP Go Pohlman Revision2
32 pages
Solution Chapter 2
100% (1)
Solution Chapter 2
45 pages
EE HVE MCQ
No ratings yet
EE HVE MCQ
33 pages
3074 PDF
No ratings yet
3074 PDF
18 pages
02 Gaseous, Solid - Liquid Breakdown
No ratings yet
02 Gaseous, Solid - Liquid Breakdown
103 pages
Paschen's Law
No ratings yet
Paschen's Law
8 pages
Paschen's Law PDF
No ratings yet
Paschen's Law PDF
6 pages
ELEN4003 Notes-Conduction and Breakdown in Gases & Liquids
No ratings yet
ELEN4003 Notes-Conduction and Breakdown in Gases & Liquids
21 pages
KEE 073 Solution
No ratings yet
KEE 073 Solution
24 pages
Unit II
No ratings yet
Unit II
41 pages
Hve Notes
No ratings yet
Hve Notes
31 pages
Plasma Application Library Manual
No ratings yet
Plasma Application Library Manual
404 pages
HVE Imp Que. Ans Bank
No ratings yet
HVE Imp Que. Ans Bank
26 pages
4-Deriving and Obtaining The Minimum Breakdown Voltage & Product of Minimum PD Based On Paschen's Law-18-Dec-2019Material - I - 18-Dec-2019 - FALLSEM2019-20 - EE
No ratings yet
4-Deriving and Obtaining The Minimum Breakdown Voltage & Product of Minimum PD Based On Paschen's Law-18-Dec-2019Material - I - 18-Dec-2019 - FALLSEM2019-20 - EE
7 pages
Conduction and Breakdown in Gases: Faculty of Electrical Engineering Department of Power Engineering
No ratings yet
Conduction and Breakdown in Gases: Faculty of Electrical Engineering Department of Power Engineering
51 pages
Streamer Mechanism and Paschens Law
No ratings yet
Streamer Mechanism and Paschens Law
16 pages
IEC 60870-5-104 - The New Solution For Communication in Substations
No ratings yet
IEC 60870-5-104 - The New Solution For Communication in Substations
5 pages
AJNSA - Volume 51 - Issue 2 - Pages 1-9
No ratings yet
AJNSA - Volume 51 - Issue 2 - Pages 1-9
9 pages
2-Conduction and Breakdown in Gases Updated
No ratings yet
2-Conduction and Breakdown in Gases Updated
46 pages
R W Crompton 1956 Proc. Phys. Soc. B 69 2
No ratings yet
R W Crompton 1956 Proc. Phys. Soc. B 69 2
13 pages
Gas Dielectric
No ratings yet
Gas Dielectric
28 pages
AO No.30. (ASF) 12-14-21
No ratings yet
AO No.30. (ASF) 12-14-21
28 pages
Models - Plasma.corona Discharge Air 1d
No ratings yet
Models - Plasma.corona Discharge Air 1d
26 pages
2006-Chen Electrical Breakdown Phenomena For Devices With Micron Separations
No ratings yet
2006-Chen Electrical Breakdown Phenomena For Devices With Micron Separations
9 pages
PhysRev 50 650
No ratings yet
PhysRev 50 650
9 pages
Static Breakdown in Gas. Part I - LinkedIn
No ratings yet
Static Breakdown in Gas. Part I - LinkedIn
14 pages
U2d5 2 5 Avoiding Bias
No ratings yet
U2d5 2 5 Avoiding Bias
11 pages
1,2 Chapters
No ratings yet
1,2 Chapters
14 pages
The Transition To Paschens Law For Microscale Gas
No ratings yet
The Transition To Paschens Law For Microscale Gas
8 pages
HV 2
No ratings yet
HV 2
80 pages
Lec# 09 Paschen's Law
No ratings yet
Lec# 09 Paschen's Law
15 pages
Ia3 HV
No ratings yet
Ia3 HV
7 pages
KE47503 HIGH VOLTAGE CHAPTER 2 - BD of Gasses and Vacuum (Autosaved)
No ratings yet
KE47503 HIGH VOLTAGE CHAPTER 2 - BD of Gasses and Vacuum (Autosaved)
48 pages
Electron Acceleration and Betatron Radiation Drive
No ratings yet
Electron Acceleration and Betatron Radiation Drive
223 pages
Breakdown in Electronegative Gases, V-T Characteristics & Post Breakdown Phenomena Post Breakdown Phenomena
No ratings yet
Breakdown in Electronegative Gases, V-T Characteristics & Post Breakdown Phenomena Post Breakdown Phenomena
22 pages
Guru Puja - The Declaration of Shri Mataji - Nirmala Vidya Amruta
No ratings yet
Guru Puja - The Declaration of Shri Mataji - Nirmala Vidya Amruta
38 pages
MODUL AJAR GTT PKWU BUDIDAYA SEMESTER Ganjil
No ratings yet
MODUL AJAR GTT PKWU BUDIDAYA SEMESTER Ganjil
11 pages
هندسة كهربية 2
No ratings yet
هندسة كهربية 2
38 pages
New Part 1
No ratings yet
New Part 1
9 pages
F4 D 2023 Frank Milan DP 20 Fi Frank M
No ratings yet
F4 D 2023 Frank Milan DP 20 Fi Frank M
153 pages
HVT Lecture 16
No ratings yet
HVT Lecture 16
28 pages
Hutcherson-1987-Triggered Spark Gap Recovery
No ratings yet
Hutcherson-1987-Triggered Spark Gap Recovery
5 pages
Formula For Steel Quantity
No ratings yet
Formula For Steel Quantity
19 pages
A Dielectric Barrier Discharge (DBD) Plasma Reactor: An Efficient Tool To Measure The Sustainability of Non-Thermal Plasmas Through The Electrical Breakdown of Gases
No ratings yet
A Dielectric Barrier Discharge (DBD) Plasma Reactor: An Efficient Tool To Measure The Sustainability of Non-Thermal Plasmas Through The Electrical Breakdown of Gases
13 pages
2002 - Filamentary Patterned and Diffuse Barrier Discharges - Kogelschatz
No ratings yet
2002 - Filamentary Patterned and Diffuse Barrier Discharges - Kogelschatz
9 pages
Garner-2020-Transition Between Electron Emission and Gas Breakdown Mechanism
No ratings yet
Garner-2020-Transition Between Electron Emission and Gas Breakdown Mechanism
91 pages
Breakdown Phenomena
No ratings yet
Breakdown Phenomena
30 pages
Collocations Challenge
No ratings yet
Collocations Challenge
2 pages
Coaching Model - The Hexagon of Life - : Quasylatic
No ratings yet
Coaching Model - The Hexagon of Life - : Quasylatic
8 pages
2019 EE 74 Lab 6
No ratings yet
2019 EE 74 Lab 6
6 pages
4 Breakdown Theory1
No ratings yet
4 Breakdown Theory1
61 pages
Pedersen
No ratings yet
Pedersen
7 pages
Collision and Diffusion in Microwave Breakdown of Nitrogen Gas in and Around Microgaps
No ratings yet
Collision and Diffusion in Microwave Breakdown of Nitrogen Gas in and Around Microgaps
10 pages
Electric Field Features and Its Application For Air Gap Breakdown Voltage Prediction
No ratings yet
Electric Field Features and Its Application For Air Gap Breakdown Voltage Prediction
24 pages
At NTP Gases Are Very Good Insulators. Conduction Current Is About 10 A/cm
No ratings yet
At NTP Gases Are Very Good Insulators. Conduction Current Is About 10 A/cm
13 pages
High Voltage Engineering : Gas, Liquid and Solid Materials As Insulating Media
No ratings yet
High Voltage Engineering : Gas, Liquid and Solid Materials As Insulating Media
18 pages
Ex: For Parallel Plate Capacitor F 1
No ratings yet
Ex: For Parallel Plate Capacitor F 1
14 pages
M3L1 Creating Inclusive Cultures 1
No ratings yet
M3L1 Creating Inclusive Cultures 1
16 pages
Roenthal-20XX-Air Hydrodynamics of The UltafastLaser Triggered Spark Gap
No ratings yet
Roenthal-20XX-Air Hydrodynamics of The UltafastLaser Triggered Spark Gap
19 pages
Applsci 12 01989 v2
No ratings yet
Applsci 12 01989 v2
19 pages
Experimental Study On Spatial Characteristics of Filamentation and Propagation of Ultrashort Pulse Laser
No ratings yet
Experimental Study On Spatial Characteristics of Filamentation and Propagation of Ultrashort Pulse Laser
7 pages
Project Report: Hospital Management System
No ratings yet
Project Report: Hospital Management System
10 pages
Complaint Under Section 2 (D) of The CR.P.C.: Hidayatullah National Law University
No ratings yet
Complaint Under Section 2 (D) of The CR.P.C.: Hidayatullah National Law University
16 pages
Geophysical Research Letters - 2022 - Fu - Observations of The Breakthrough Phase of Attachment Proce
No ratings yet
Geophysical Research Letters - 2022 - Fu - Observations of The Breakthrough Phase of Attachment Proce
9 pages
CXV 0 BF5 Ycd OAXNqb AA4 W LDF BPM 0653
No ratings yet
CXV 0 BF5 Ycd OAXNqb AA4 W LDF BPM 0653
26 pages
Jenny Stifflemire Resume
No ratings yet
Jenny Stifflemire Resume
1 page
Triggered Vacuum and Gas Spark Gaps: IEEE Transactions On Power Delivery May 1996
No ratings yet
Triggered Vacuum and Gas Spark Gaps: IEEE Transactions On Power Delivery May 1996
8 pages
Influence of Gas Temperature On Electrical Breakdown in Cylindrical Electrodes
No ratings yet
Influence of Gas Temperature On Electrical Breakdown in Cylindrical Electrodes
5 pages
Ozone Therapy: A Clinical Review
No ratings yet
Ozone Therapy: A Clinical Review
5 pages
Electrical Characteristics and Plasma Diagnostics of (Ar/O) Gas Mixture Glow Discharge
No ratings yet
Electrical Characteristics and Plasma Diagnostics of (Ar/O) Gas Mixture Glow Discharge
7 pages
HV Assign
No ratings yet
HV Assign
7 pages
HV Assign
No ratings yet
HV Assign
7 pages
Paschen's Law
No ratings yet
Paschen's Law
6 pages
1 Online
No ratings yet
1 Online
10 pages
5 Mother Goose Rhymes
No ratings yet
5 Mother Goose Rhymes
11 pages
334 Gss 2010-01
No ratings yet
334 Gss 2010-01
8 pages
Visualisation of Plasma Induced Processes in Inter
No ratings yet
Visualisation of Plasma Induced Processes in Inter
7 pages
Principles of Marketing Term Paper On Knorr Pasta
100% (4)
Principles of Marketing Term Paper On Knorr Pasta
27 pages
Boroumand-2022-J. Phys. B At. Mol. Opt. Phys. 55 213001
No ratings yet
Boroumand-2022-J. Phys. B At. Mol. Opt. Phys. 55 213001
11 pages
Metis 201781
No ratings yet
Metis 201781
5 pages
Arantchouk-20xx-Simple HV HC Spark Gap With Subnanosecond Jitter Triggered by FS Laser Filamentation
No ratings yet
Arantchouk-20xx-Simple HV HC Spark Gap With Subnanosecond Jitter Triggered by FS Laser Filamentation
6 pages
Ispirli, Mehmet Murat Kalenderli, Özcan Seifert, Florian Rock, Michael Oral, Bülent
No ratings yet
Ispirli, Mehmet Murat Kalenderli, Özcan Seifert, Florian Rock, Michael Oral, Bülent
10 pages
Three-Step Delivery - Odoo 17.0 Documentation
No ratings yet
Three-Step Delivery - Odoo 17.0 Documentation
5 pages
Sci Rep 2008 Laser Ionization Discharge
No ratings yet
Sci Rep 2008 Laser Ionization Discharge
11 pages
Use of Trigger D Spark Gaps
No ratings yet
Use of Trigger D Spark Gaps
3 pages
Generate Radio-Frequency Capacitively Coupled Plasma 4 MHZ and Study of Breakdown Voltage Curves in Low-Pressure
No ratings yet
Generate Radio-Frequency Capacitively Coupled Plasma 4 MHZ and Study of Breakdown Voltage Curves in Low-Pressure
5 pages
Lesson 28 Gender and Other Cross-Cutting Issues
No ratings yet
Lesson 28 Gender and Other Cross-Cutting Issues
5 pages
Answers To Practice Test Questions 2 Molecular Orbital Theory: Homonuclear Diatomic Molecules
No ratings yet
Answers To Practice Test Questions 2 Molecular Orbital Theory: Homonuclear Diatomic Molecules
5 pages
Larrson-20XX-Laser Triggering of Spark Gap Switches
No ratings yet
Larrson-20XX-Laser Triggering of Spark Gap Switches
5 pages
Sci Rep 2007 Laser Ionization Discharge
No ratings yet
Sci Rep 2007 Laser Ionization Discharge
7 pages
Sci Rep 2013 Laser Ionization Discharge
No ratings yet
Sci Rep 2013 Laser Ionization Discharge
3 pages
Event - NepalSedimentWorkshopBrochure - Contact Us - Jan10
No ratings yet
Event - NepalSedimentWorkshopBrochure - Contact Us - Jan10
17 pages
Locating Information: Scanning: Exploring Skills
No ratings yet
Locating Information: Scanning: Exploring Skills
3 pages
Laser Triggering of Gas Fiu - Ed Spark Gaps
No ratings yet
Laser Triggering of Gas Fiu - Ed Spark Gaps
2 pages
Self-Velocity Manipulation - Superpower Wiki - Fandom
No ratings yet
Self-Velocity Manipulation - Superpower Wiki - Fandom
8 pages
Mindfulness and Awareness - Tai Situ Rinpoche
No ratings yet
Mindfulness and Awareness - Tai Situ Rinpoche
1 page
Tutorial Questions On Matlab
No ratings yet
Tutorial Questions On Matlab
2 pages
Essay Writing 12 May 2023
No ratings yet
Essay Writing 12 May 2023
2 pages
CLASS IX Durgapur SYLLABUS Physics
No ratings yet
CLASS IX Durgapur SYLLABUS Physics
2 pages
Introduction to Electromagnetic Theory
From Everand
Introduction to Electromagnetic Theory
George E. Owen
No ratings yet
Fundamental Concepts in Heterogeneous Catalysis
From Everand
Fundamental Concepts in Heterogeneous Catalysis
Jens K. Nørskov
No ratings yet

A Tutorial On Theoretical and Computational Techniques For Gas Breakdown in Microscale Gaps

Uploaded by

A Tutorial On Theoretical and Computational Techniques For Gas Breakdown in Microscale Gaps

Uploaded by

808 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 48, NO.

A Tutorial on Theoretical and Computational

Abstract— Paschen’s law (PL), derived based on the Townsend I. I NTRODUCTION

to E s to give an enhanced FN current density of [27], [32]

This positive space charge also modifies the TA criterion from

Go and Venkattraman [28] previously reviewed ion-

at atmospheric pressure [60], then adding in gas dependence TABLE I

Ē = E E ∗−1 , T̄ = T T∗−1 , p̄ = pp∗−1

φ∗ = [(3.79 × 10−4 )2 BFN ]2 , E ∗ = 0.95BFN φ∗3/2 ⎡  ⎤

distances studied. Thus, even in the smaller gap case when PL

IV. PARTICLE - IN -C ELL /M ONTE C ARLO C OLLISION

On the other hand, the ionization coefficient may also be

steady state, the electron current density profile was used to

Amanda M. Loveless (Student Member, IEEE)

You might also like

φ∗ = [(3.79 × 10−4 )2 BFN ]2 , E ∗ = 0.95BFN φ∗3/2 ⎡ ⎤