RESEARCH ARTICLE | JULY 17 2017

Numerical modeling for investigating the optical breakdown

threshold of laser-induced air plasmas at different laser
characteristics 
Kholoud A. Hamam; Laila H. Gaabour; Yosr E. E. D. Gamal

Phys. Plasmas 24, 073515 (2017)

https://doi.org/10.1063/1.4993062


View Export
Online Citation

19 May 2024 19:12:08

 PHYSICS OF PLASMAS 24, 073515 (2017)

Numerical modeling for investigating the optical breakdown threshold

of laser-induced air plasmas at different laser characteristics
Kholoud A. Hamam,1,a) Laila H. Gaabour,1 and Yosr E. E. D. Gamal2
1
Department of Physics, Faculty of Science, King Abdulaziz University, Jeddah, Kingdom of Saudi Arabia
2
National Institute of Laser Enhanced Science, Cairo University, El-Giza, Egypt
(Received 2 April 2017; accepted 21 June 2017; published online 17 July 2017)
In this work, we report a numerical investigation of two sets of experimental measurements that
were previously carried out to study the breakdown threshold dependence on laser characteristics
(wavelength, pulse width, and spot size) in the breakdown of laboratory air at different pressures.
The study aimed to inspect the significance of the physical mechanisms in air breakdown as related
to the applied experimental conditions. In doing so, we adopted a simple theoretical formulation
relying on the numerical solution of a rate equation that describes the growth of electron density
due to the joined effect of multi-photon and avalanche ionization processes given in our earlier
work [Gaabour et al., J. Mod. Phys. 3, 1683–1691 (2012)]. Here, the rate equation is adapted to
include the effect of electron loss due to attachment processes. This equation is then solved numeri-
cally using the Runge–Kutta fourth order technique. The influence of electron gain and loss pro-
cesses on the breakdown threshold is studied by calculating the breakdown threshold intensity and
RMS electric field for atmospheric air using different laser parameters (wavelength, pulse widths,
and focal length lenses), in correspondence to the experimental conditions given by Tambay and
Thareja [J. Appl. Phys. 70(5), 2890 (1991)]. To validate the model, a comparison is made between
those calculated thresholds and the experimentally measured ones. Moreover, the effective contri-
bution of each of the considered physical processes to the breakdown phenomenon is examined by
studying the effect of laser wavelength and spot diameter on the threshold intensities, as well as on
the temporal variation of the electron density. The correlation between the threshold intensity and
gas pressure is tested in relation to the measurements of Tambay et al. [Pramana–J. Phys. 37(2),

19 May 2024 19:12:08

163 (1991)]. Calculations are also carried out to depict the impact of pulse width on the threshold
intensity. Published by AIP Publishing. [http://dx.doi.org/10.1063/1.4993062]

I. INTRODUCTION Laser-induced breakdown is a complicated process

involving coupling between particles, electromagnetic wave
During the last decades of the previous century, advan-
interactions, chemical reactions, and fluid dynamic effects.
ces in the significant development of high-intensity laser sys-
Ionization of the air in the path of a high power laser is found
tems have revived a keen interest in the laser-induced optical
to be a limit on the maximum intensity which can be propa-
breakdown (LIOB) study. These laser sources are operated at
gated through the atmosphere.14 Therefore, the crucial goal of
a wavelength range extended from the near infrared to the
this work is to analyze the results of the measurements that
ultraviolet regions. Consequently, extensive experimental
were carried out to determine the air breakdown threshold
and theoretical research works have been performed to
intensity and electric field in correspondence to wide operat-
explain the physical phenomena involved in the LIOB of
air.2–9 Some of these studies fascinated the dependence of ing experimental conditions, namely, laser wavelength, pulse
LIOB on the electric field since the latter determines the width, and spot size for laboratory air at different pressures.
onset of ionization. However, for applications, the depen- These quantities are critical in the validation of laser break-
dence of LIOB threshold on the laser fluence was found to down models that are applied to examine the effect of such
be more meaningful. discharge parameters on air spark ignition. To our knowledge,
The interest in this phenomenon was due to its critical most of the recent measurements were concerned to the study
applications in different fields.10–12 In particular, those where of plasma formation in atmospheric air for pulse width in the
air cannot be avoided owing to its natural abundance and the pico- or nanosecond scale.15,16 Moreover, femtosecond laser
important role that interaction of air with laser radiation sources are also used to study the effect of gas pressure on the
plays. While it is widely accepted that infrared laser pulses threshold intensity for only one laser wavelength, at a specific
with small widths, reaching few nanoseconds, can produce spot size.17
air breakdown via collisional processes under atmospheric Accordingly, the objective of the present investigation is
conditions,13 the dynamics underlying this highly nonlinear to understand the physical mechanisms involved in air break-
phenomenon for ultraviolet lasers is yet not fully understood. down induced by various laser characteristics and pressures.
To accomplish this, we applied a simple numerical model
given in our past work.1 The model relies on the numerical
a)
Author to whom correspondence should be addressed: khamam@kau.edu.sa solution of a rate equation that describes the net gain of the

1070-664X/2017/24(7)/073515/9/$30.00 24, 073515-1 Published by AIP Publishing.

073515-2 Hamam, Gaabour, and Gamal Phys. Plasmas 24, 073515 (2017)

electron density due to the joined effect of multi-photon and II. THEORETICAL FORMULATION
avalanche ionization processes,18 besides the loss processes
Using the Runge–Kutta fourth order technique, numeri-
which are expected to take place during the interaction. These
cal solution is undertaken from the adapted rate equation that
losses include energy loss through the elastic collision between
describes the net gain of the electron density as a result of
an electron and a neutral gas molecule as well as the effect of
the considered physical processes.
electron loss due to diffusion out of the focal volume and
The solution is based on more realistic relations for the
recombination. In the present study, to assemble the case of
rates and rate coefficients of the electron gain and loss pro-
air breakdown, an extra term is added to the rate equation to
cesses accounted in the rate equation. The model is applied
account for the effect of electron loss due to attachment pro-
cesses. The model is applied to investigate the experimental to study the phenomenon of the optical breakdown of air
measurements that were carried out in Refs. 2 and 3. These using various laser characteristics (wavelength, different
experimental measurements are deliberately chosen since focal spot diameters, and pulse width) and gas pressure. The
they have been operated under different laser characteristics findings of their measurements showed an increase in the
and air pressure ranges below the atmospheric pressure. This threshold intensity and the RMS electric field (Erms) for
pressure range has a great interest for applications involving smaller spot size. Dependence of the size of the visual spark
aircraft since the static pressures involved are undoubtedly sig- as a function of laser parameters is also studied from the
nificantly sub-atmospheric. Moreover, the nanosecond pulse time evolution of the electron density.
widths used in this experiment are appropriate for investigating
the role played by the loss processes associated with air break-
A. Basic equation
down, in particular, at the ultraviolet wavelengths.
In modeling ionization by a strong electromagnetic The adapted form of the rate equation that describes the
wave, the Keldysh theory19 is assumed to be a key parameter net gain of the electron density in air breakdown under the
c in characterizing the ionization mechanism responsible for influence of the considered electron generation and loss pro-
the generation of the seed electrons. This parameter reads20 cesses is written as
 12 
ei m dn dn 
c¼x ; (1) ¼  þ i n  D n  R n2  gn: (2)
I Z0 e2 dt dt MPI

19 May 2024 19:12:08

where x is the laser angular frequency, ei is the ionization Here, n is the free electron density.  i, D, R, and g represent
potential, I is the laser intensity, m and e are the electron the ionization, diffusion, recombination, and attachment rates,
mass and charge, respectively, and Z0¼377 X is the vacuum respectively. In Eq. (2), the first two terms on the right-hand
wave impedance. All parameters in Eq. (1) are expressed in side refer to electron generation through multiphoton ioniza-
international system (SI) of units. tion and electron cascade ionization. The remaining terms
Keldysh derived approximate expressions for the proba- account for the loss processes (electron diffusion out of the
bility of ionization by the electric field in both gases and con- focal volume, recombination, and attachments).
densed media. A multiphoton ionization regime was taken For accurate computation, we take into consideration
for c  1, while a strong field tunneling process was consid- the depletion of air molecules due to ionization by subtract-
ered for c  1. ing the density of the generated electrons from the molecular
To decide the ionization mechanism accountable for the density. The evaluation of Eq. (2) requires information on
breakdown of air corresponding to the experimental condi- the air data, the shape of the focal volume, as well as the
tions under investigation, Table I shows the calculated values temporal variation of the laser intensity in this volume.
of the Keldysh parameter c corresponding to the threshold
intensity measured for three focused beam diameters for
each value of the experimentally studied laser wavelengths. B. Rates and rate coefficients
From Table I, it is clear that the parameter c exceeds
1. Multiphoton ionization rate
unity even for the longest wavelength of 1060 nm. This
in turn confirms that ionization proceeds via multi-photon To ionize a medium through multiphoton absorption, an
processes rather than tunneling ionization for the experimen- effective ionization potential e0i is defined, which depends on
tal conditions under investigation. the electric field amplitude E and laser frequency x, such as
TABLE I. The experimentally measured thresholds at different values of the spot diameter (d) for each laser wavelength and the corresponding calculated
value of Keldysh parameter c.

d ¼ 30 lm Keldysh parameter d ¼ 50 lm Keldysh parameter d ¼ 60 lm Keldysh parameter

k (nm) Ith exp. (W/cm2) c Ith exp. (W/cm2) c Ith exp. (W/cm2) c

266 5.2  1011 47.56 1.7  1011 83.189 1  1011 108.47

355 5.4  1011 34.97 1.4  1011 68.688 9  1010 85.669
532 4.8  1011 24.75 1.4  1011 45.834 9  1010 57.165
1060 1.8  1012 6.415 5.8  1011 11.3 3  1011 15.713
073515-3 Hamam, Gaabour, and Gamal Phys. Plasmas 24, 073515 (2017)

 
e2 E2 1 low energy electrons. The cascade ionization rate per elec-
e0i ¼ ei þ 0 2 ¼ ei 1 þ 2 : (3) tron is given in Ref. 23 and is written as
4m x 4c
 
The number of photons K required to ionize a molecule is 1 e2 sm m x2 sm
i ¼ 2 2 I ; (7)
given by h1þe0i =
hxi; where the symbol hi denotes the integer x sm þ 1 c nr m e0 ei M
of the number. K is sometimes known as the degree of non-
linearity of multiphoton ionization. Consequently, the where sm is the mean free time between electron-air mole-
expression for the multiphoton ionization probability takes cule collisions, M is the mass of an air molecule, and m is
the form in Ref. 19, the electron mass. The first term of this equation is related to
the energy gain by electrons from the laser electric field. The
 32 !K second term refers to the energy transfer from electrons to
2x m0 x e2
W  0 0 I exp ð2 K Þ /ðzÞ; air molecules during elastic collisions. For atmospheric air,
9p h 16m ei x2 c e0 nr sm is turned to be 300.0 fs.17
(4)
3. Electron diffusion rate
where
The rate of electron diffusion along the radial and axial
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 e0 distances corresponding to the assumed cylindrical shape of
Z ¼ 2K  i the focal volume of radius w0 and length zR is given by Refs.
h x
24 and 25 as
with 1ðxÞ representing Dawson’s integral21 "   2 #
ðx sm eav 2:4 2 p
 D¼ þ ; (8)
1ð xÞ ¼ expðx2 Þ exp y2 dy : (5) 3m w0 ZR
0

Here, x is the angular frequency of the laser radiation, h is where w0 is the beam waist, zR¼ npw02/k is the Rayleigh
Planck’s constant/2p, e0 is the permittivity of free space, c is length, and eav is the average kinetic energy of the free elec-
the vacuum speed of light, nr is the refractive index of the trons. Studies concerned with the electron average energy on
gases26 realized that under breakdown conditions near thresh-

19 May 2024 19:12:08

medium at frequency x, and e0i is the effective ionization
old, the average electron energy is one third to one fourth the
energy for air. In this formula, e is the electron charge, m0 is
ionization energy. Therefore, for the ionization of air (15.6 eV),
the reduced mass (it is assumed to be m/2, where m is the
taking the average electron energy as 5 eV seems a reason-
electron mass), and I is the laser irradiance in W/cm2.
able approximation for calculations of the breakdown parame-
Equation (4) has a quite complex reliance on the optical
ters using the simple model applied in the present work.
field E. Therefore, this equation is greatly simplified by using
the limiting condition, c  1 (Ref. 22) which reduces Eq. (3) to
4. Recombination rate coefficient
16 4c12  1, e0i ¼ ei , K¼h1þei =hxi, and z  ½2K  2ei =hx].
With these changes, the field dependence is confined to Another undesirable process limiting the electron density
the second bracket of Eq. (4), and the rate takes the formal, growth is charged particle recombination between electrons
simple form: W¼ AIK and positive ions. We mainly concentrate on the two body
dissociative recombination process that follows the reaction.
 32  K
K 2x m0 x e2
AI  0
I Aþ  
2 þ e ! A þ A;
9p 16m ei x2 c e0 nr
h
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi !
2 ei where the asterisk denotes an electronically excited state. In
 exp ð2 K Þ / 2K  ; (6) the present work, we consider that air is a mixture of nitrogen
h x
and oxygen only at a ratio of 79% and 21%, respectively.
with ei being the ionization potential of air molecules which Therefore, we applied for the electron-ion recombination
is taken as 15.6 eV.17 coefficient, R, of N2 and O2, expressions based on the meas-
urements that were carried out in Ref. 27 and given in Ref.
28. These expressions vary with the electron temperature
2. Cascade ionization rate (energy), expressed in eV, such as
The occurrence of this process requires the presence of
few free electrons in the focal volume during the initial RN ðm3 =sÞ ¼ 2:035  1012 T0:39
e ; ðe  Nþ 2 Þ;
stages of the laser pulse. These electrons are assumed to be RO ðm3 =sÞ ¼ 1:138  1011 T0:70 ; ðe  Oþ 2 Þ:
e
generated by multiphoton ionization and gain energy from
the laser field through inverse Bremsstrahlung absorption The present work is devoted to the study of atmospheric air
(IBA) in colliding with the air molecules. In undergoing sev- plasma using a wide range of laser wavelengths. At this condi-
eral IBA events, their kinetic energy could exceed the ioniza- tion, the elastic collision rate is considerably high, and the
tion energy ei. Then, they can lose this energy through electron energy, Te, is assumed to thermalize at the same
inelastic collisions leading ultimately to the creation of new ambient air temperature as the gas molecules.28 Therefore, in
073515-4 Hamam, Gaabour, and Gamal Phys. Plasmas 24, 073515 (2017)

these calculations, we adopted the assumption made in the of the neutral gas molecules in the focal volume by a ratio
measurements that electrons possess an average energy of d ¼ 1.0%).
7.452  1019 J (4.65 eV). This energy (the lower elec-
tronic excitation energy of the oxygen molecule) corresponds III. RESULTS AND DISCUSSION
to an electron temperature of 5.4  104 K.27 Accordingly, the Applying the modified rate equation, calculations are
electron-ion recombination coefficient of air is obtained from first devoted to investigate the experimental measurements
the relationship: findings given by Ref. 2. In this experiment, the measure-
ments are carried out to study the effect of focal length of
Rðm3 =sÞ ¼ 0:79 RN þ 0:21 RO : (9)
lens on the breakdown threshold intensity and RMS electric
field (Erms) for laboratory air at the atmospheric pressure.
5. Attachment rate The breakdown was induced by a Nd: YAG laser operating
at wavelengths covering a range 1060 nm down to 266 nm
At normal electric fields and atmospheric pressure, the and pulse width varying from 4.0 ns to 8.0 ns at spot sizes
two and three body attachment rates are expressed as29,30 varying between 30 to 100 lm. The beam divergence was
  less than 0.5 mrad. The spot size was determined under dif-
N 42:3
g2 ðs1 Þ ¼ 1:22  108 e E0 ; (10) fraction limited beam assumption. For laboratory air break-
N0
down studies, the problem of particle contaminants in the air
where E is the electric field expressed in V/cm, N is the air is neglected since only minimum humidity and temperature
density expressed in cm3, and N0 ¼ 2.7  1019 cm3. The fluctuations are allowed in this experiment.
three body attachment rate is given by Ref. 30,
A. Effect of the focal spot size
 2
N 0:62 þ 800 E20 An extensive and methodical study is undertaken to pro-
g3 ðs1 Þ ¼ 108  1 :
N0 3
1 þ 10 E20 E20 1 þ 0:03 E20 3 vide a numerical investigation on the effect of laser wave-
length and beam spot size (the focal length lens) on the
(11)
breakdown threshold of laboratory air induced by the first
Here, four harmonics of the Nd:YAG laser. For a realistic compari-

19 May 2024 19:12:08

son, the calculated peak threshold intensities (solid line) and
E the measured ones (symbol) are plotted and shown in Fig. 1
E0 ¼ 3:34  105  ; (12)
N at (a) k ¼ 1060 nm, s ¼ 8.0 ns, (b) k ¼ 532 nm, s ¼ 6.0 ns,
N0 (c) k ¼ 355 nm, s ¼ 4.0 ns, and (d) k ¼ 266 nm, s ¼ 4.0 ns.
A reasonable agreement is indicated for the four wavelengths
where E represents the electric field expressed in V/cm. The in relation to their trend over the spot diameter range consid-
total attachment rate is then written as ered in the experiment. This agreement gives evidence for
the validity of the model. The results showed, high threshold
g ¼ g2 þ g3 ) ga  g2 ¼ 1:22  108 s1 : (13) intensities at the smallest spot diameter (3  103 cm3).
This is shown from Fig. 1 where for 1060 nm [(Fig. 1(a)], the
C. Method of calculation
threshold intensity is found to be 1.4566  1012 W/cm2,
while for 532 nm [Fig. 1(b)], its value is 3.8  1011 W/cm2 .
The rate equation is solved numerically using the For the shorter wavelengths 355 nm [Fig. 1(c)] and 266 nm
Runge–Kutta fourth order technique with adaptive time step. [Fig. 1(d)], the obtained threshold intensities are 6.86  1011
In these calculations, we adopted the experimental assump- W/cm2 and 6.9  1011 W/cm2, respectively. The highest
tion that intensity distribution remains Gaussian in the focal threshold corresponds to the longer wavelength 1060 nm
plane of the cylindrical focal volume and takes the form while the lowest one corresponds to the wavelength 532 nm.
  This result may explain the experimentally observed slope
ðt  sÞ2 which is found to be changing with the laser wavelength.
I ðt; r; zÞ ¼ I0 C1 exp   C2
 2 4s2  Despite the visual detection of the spark formation and the
r z2 high probability of diffusion losses at small spot diameters,
 exp  þ ; (14) the above result assured that the onset of breakdown with
w 0 2 zR 2
visible and ultraviolet wavelengths is initiated with multi-
where C1 and C2 are constants. r and z represent the radial photon absorption. Ionization by this process requires the
and axial distances at any time t, during the laser pulse, absorption of 7, 5, and 4 photons, which correspond to pho-
respectively. So that, the generated plasma is assumed to be ton energies of 2.32 eV, 3.5 eV, and 4.66 eV, respectively.
positioned at a cylindrical volume with the axial distance zR Electron growth then proceeds via the cascade ionization
and the radial distance w0. In the present work, calculations process. Whereas due to the low multiphoton ionization rate
are undertaken at the central focal point. associated with the 1060 nm wavelength (14 photons each of
The determination of the threshold intensity, Ith, or the energy 1.16 eV required being absorbed to ionizing air),
ERMS for air breakdown is grounded on the attainment of elec- collisional ionization dominates. The gradual decrease in the
tron density that satisfies the breakdown criterion (ionization threshold intensity shown over the remaining spot diameter
073515-5 Hamam, Gaabour, and Gamal Phys. Plasmas 24, 073515 (2017)

FIG. 1. Comparison between the cal-

culated threshold intensity for atmo-
spheric air (solid lines) and measured
ones (symbols) of Ref. 2, as a function
of the spot diameter for the wave-
lengths (a) 1060 nm, (b) 532 nm, (c)
355 nm, and (d) 266 nm.

range clarifies that this parameter has only a strong effect on 10.0 cm, and 12.0 cm, respectively.2 It is noticed here that
the threshold intensity over the short region as compared to the calculated threshold intensity as a function of the laser
the longer one. The close values shown for the threshold wavelength (solid lines) is consistent with the experimentally

19 May 2024 19:12:08

intensity, corresponding to 355 and 266 nm at the small spot measured values (symbols). Moreover, curves 1–3 display
diameter, are attributed to the enhancement of recombination almost the same tendency, where over the wavelength range
and attachment losses which overcome the development of of 266–532 nm, the threshold intensities showed low values,
electrons through the multiphoton ionization mechanism. while for the longer wavelength (1060 nm), a noticeable
These losses revealed a pronounced effect, in particular, at increase is observed. Additionally, the threshold intensity is
the small spot diameter as observed from the steep slope found to decrease as the spot size increases. This behavior
shown in Figs. 1(c) and 1(d). favors the fact that the onset of breakdown with wavelengths
To confirm this result, Figs. 2(a) and 2(d) present the in visible and ultraviolet spectral regions (h is about 2.33 eV
comparison between the calculated RMS electric field (solid and 4.7 eV, respectively) is initiated with high electron density
lines) and the measured ones2 (symbols) as a function of diffu- generated mainly by multiphoton absorption and then via
sion length K (K is related to the lens focal length as shown in the cascade process. Whereas breakdown at the 1060 nm is
Sec. II B 3). A close agreement is perceived for the four wave- mostly produced by collisional ionization, multiphoton ioniza-
lengths; both showed an observable slope dependence over tion is merely responsible for the generation of the seed elec-
the tested diffusion length range (3  103 cm–6  103 cm). trons required to ignite the avalanche ionization mechanism.
For the wavelength 1060 nm and pulse width 8.0 ns [Fig. The highest intensities are obtained for the smallest spot size
2(a)], the slope is found to be 1.24, while for 532 nm and shown by curve (1). This result is referred to for the increase
6 ns [Fig. 2(b)], the slope is 1.1. For the shorter wave- in electron loss rate by diffusion of the focal volume with the
lengths (355 and 266 nm) of 4 ns, the slope is 0.94. This decrease in the spot diameter. On the other hand, the influence
result clarifies the dynamical effect of diffusion length (focal of electron losses through recombination and attachment pro-
length) on the RMS electric field in relation to the laser wave- cesses, which compensate the high ionization rate associated
length. This behavior proved the strong action of electron loss with the shorter wavelength (266 nm) for the spot diameter
for the short wavelengths due to diffusion, recombination, and 60 lm (curve 3), explains the slight increase in the threshold
attachments, which decrease the effect of fast growth of elec- intensity shown at this wavelength.
tron density through the multiphoton ionization process. This is not the case for the smallest spot diameter (curve
1), where the electron loss rate shown at the wavelength
B. Threshold intensity dependence on laser 355 nm is found to exceed those corresponding to the wave-
wavelength lengths 266 and 532 nm.
Figure 3 shows an investigation of the measurements that
C. Time evolution of the electron density
were carried out to relate the atmospheric air threshold inten-
sity and the laser wavelength for three spot diameters, namely, For a deeper understanding of the effect of the spot size
30 lm (curve 1), 50 lm (curve 2), and 60 lm (curve 3). These on the air spark ignition mechanism, a relation is plotted
spot diameters correspond to lens’ focal lengths 6.0 cm, to present a comparative study of the time development of
073515-6 Hamam, Gaabour, and Gamal Phys. Plasmas 24, 073515 (2017)

FIG. 2. Plot comparing the calculated and measured values of the RSM electric field for atmospheric air as a function of diffusion length for laser wavelengths:

19 May 2024 19:12:08

(a) 1060 nm, (b) 532 nm, (c) 355 nm, and (d) 266 nm.

the electron density obtained for the four laser wavelengths The rate of growth is found to decrease as the spot diameter
at three values of the spot diameter. This relation is shown increases, as shown by curves 2 and 3. Further, with the
in Figs. 4(a)–4(d), for the wavelengths 1060 nm down to increase in time, the electron density revealed a gradual
266 nm, at spot diameters d ¼ 30 lm (curve 1), d¼ 50 lm increase followed by an almost leveling off behavior. This
(curve 2), and d¼ 60 lm (curve 3). This figure demonstrates result is attributed to the high rate of energy gain released
the effect of laser wavelength and spot size on the gain and through the collisional ionization mechanism, which is offset
loss processes which control the breakdown of atmospheric by the electron loss processes. The decrease observed near
air. It is noticed here that for the visible and ultravoilet range the end of the pulse for the small spot diameter shown by
of the laser wavelength, the electron density undergoes a curve (1) in Figs. 4(a)–4(d) reflects the high loss rate of elec-
remarkable growth during the early stages of the pulse. This trons through diffusion of the focal volume. This process is
growth increases with the increase in the photon energy, also observed for 1060 nm at the spot diameter of 50 lm as
which proves the domination of photoionization processes. shown by curve (2) in Fig. 4(a). The decrease in electrons
shown near the end of the pulse for the 266 nm in Fig. 4(d),
curves (1–3), clarifies the effective contribution of electron
recombination and attachments at the three spot diameter
values. For the laser wavelengths 532 and 355 nm [Figs. 4(b)
and 4(c)], no decrease is observed at spot diameters 50 and
60 lm (curves 2 and 3, respectively). On the contrary, these
curves showed a slight increase in the electron density during
the descending phase of the pulse. This result shows that the
increase in the spot diameter reduces the effect of electron
loss.

D. Effect of photon energy on the initial electron

density
From the above study, we noticed that the electron growth
rate during the early stages of the pulse depends mainly on the
FIG. 3. Variation of the calculated threshold intensity (solid lines) with the
laser wavelength at three spot diameters: (1) d ¼ 30 lm, (2) d¼ 50 lm, and laser wavelength (photon energy). Therefore, Fig. 5 shows the
(3) d¼ 60 lm. The symbols illustrate the experimentally measured values.2 variation of both the initial electron density, within the first
073515-7 Hamam, Gaabour, and Gamal Phys. Plasmas 24, 073515 (2017)

19 May 2024 19:12:08

FIG. 4. Effect of the spot diameter on the time evolution of the electron density calculated at atmospheric pressure for the four laser wavelengths.

nanosecond of the laser pulse, (curve I) and the corresponding showed a reliable agreement between the calculated thresholds
threshold intensity (curve II), as a function of the photon and the measured ones over the whole pressure regime for the
energy calculated at the spot diameter of 30 lm. From this fig- three wavelengths. It is observed that the threshold intensity
ure, it is clear that as the photon energy increases, the initial increases with the decrease in pressure. This result could be
electron density undergoes a noticeable increase, as shown by attributed to the fact that, at low pressures, the collision fre-
curve (I). A reverse situation occurs for the threshold intensity, quency is low and sufficient ionization can be maintained by
where it suffers a decrease of about one order of magnitude as increasing the probability of ionization at each collision (i.e.,
the photon energy increases from 1.16 eV to 2.32 eV, as the electron energy), and thus, the electric field associated
shown by curve (II). Beyond this energy, curve(II) showed an with laser radiation should be high. Hence, the threshold inten-
unobservable variation as the photon energy increases. This sity increases as the pressure decreases. An observable
increase in the threshold intensity shown at the small photon decrease in the threshold intensity is shown for the 532 nm at
energy (1.16 eV) is attributed to the low photoionization rate atmospheric pressures. This result referred to the less effect of
which cannot easily overcome the effect of electron diffusion
out of the focal volume. This process showed a constructive
role, in particular, for small spot sizes.

E. Effect of pressure
Second, the program was run to investigate the experi-
mental data given in Ref. 3 to study the dependence of thresh-
old intensity on air pressure and pulse width. The calculations
are performed for three laser wavelengths, namely, 355, 532,
and 1060 nm, which correspond to laser pulse widths of 1.3,
1.8, and 2.5 ns, respectively. The laser was focused with a lens
of 12 cm focal length (60 mm spot diameter). The tested air
pressure covered a range of 24 Torr–760 Torr. The calculated
thresholds as a function of pressure are shown in Fig. 6. For
an easy comparison, the measured thresholds are also shown
FIG. 5. Variation in the initial electron density during the early stages of the
in the same figure (symbols) and we adopted the same scale laser pulse and the corresponding threshold intensity as a function of the
applied to represent the experimental results. This relation laser frequency.
073515-8 Hamam, Gaabour, and Gamal Phys. Plasmas 24, 073515 (2017)

FIG. 7. Dependence of air breakdown threshold intensity on the laser wave-

length at pressures of 380 Torr (curve 1) and 760 Torr (curve 2).

760 Torr) is displayed. From this figure, it is clear that at the

lower pressure (curve 1), the threshold intensity decreases
with the increase in wavelength. However, at atmospheric
pressure of air, it is observed that the threshold intensity first
decreases to a minimum at 532 nm, and then increases with
the increase in the wavelength reaching its highest value at
1060 nm. This result contradicts that obtained by Ref. 31 in
their study of the breakdown threshold intensity of rare gases
at wavelengths 1060, 693, 530, and 350 nm. Their measure-
ments showed that the threshold intensity for each rare gas

19 May 2024 19:12:08

studied first increases to a maximum and then decreases with
decreasing wavelength.

F. Effect of pulse width

The effect of pulse width on the threshold intensity is
presented in Fig. 8. In this figure, the threshold intensity as a
function of the laser wavelength is plotted for two sets of
pulse widths for atmospheric air. Curve (1) represents the
threshold intensity calculated for three laser wavelengths,
355 nm (s ¼ 1.3 ns), 532 nm (s ¼ 1.8 ns), and 1060 nm
(s ¼ 2.5 ns). Curve (2) is plotted for the same wavelengths
but with different pulse widths, namely, 4.0 ns, 6.0, and
8.0 ns, respectively. It is clear from this figure that as the
FIG. 6. Comparison between the calculated threshold intensity and the mea- pulse width is reduced to about one-third of its value, the
sured ones as a function of gas pressure at the focal length ¼12.0 cm for the
laser wavelengths: (a) 1060 nm, (b) 532 nm, and (c) 355 nm.

losses. The observed slope of the curve is found to lie between

0.8 and 0.4. For 1060 nm [Fig. 6(a)], the breakdown
threshold intensity shows a pronounced pressure dependence
which assures that the onset of breakdown is typically a colli-
sional ionization. On the other hand, the onset of breakdown
with visible [Fig. 6(b)] and UV [Fig. 6(c)] wavelengths is initi-
ated with multiphoton ionization and then continued by the
cascade mechanism. It is noticed at 355 nm, the threshold
intensity showed a gradual decrease with the increase in the
gas pressure, where it attained a higher value at the atmo-
spheric pressure. This could be attributed to the effect of elec-
tron losses.
In Fig. 7, the dependence of breakdown threshold inten- FIG. 8. Threshold intensity variation with both the laser wavelength and
sity on laser wavelength at two air pressure values (380 and pulse width for air at the atmospheric pressure.
073515-9 Hamam, Gaabour, and Gamal Phys. Plasmas 24, 073515 (2017)

behavior of the threshold intensity showed totally different breakdown threshold dependence on air pressure for both
dependence. It is shown that for 355 nm, decreasing the pulse the tested parameters (laser wavelength and pulse width).
width from 4.0 ns to 1.3 ns results in an increase in the thresh-
old intensity by a factor of about 20. While for 532 nm, ACKNOWLEDGMENTS
changing the pulse width from 6.0 ns down to 1.8 ns results in
an increase in the threshold intensity by a factor of about 5. This work was funded by the Deanship of Scientific
This result showed much weaker dependence on pulse width Research (DSR), King Abdulaziz University, Jeddah, under
for the longer wavelength where reducing the pulse width Grant No. (363–028–D1434). The authors, therefore,
from 8.0 ns down to 2.5 ns resulted only in a slight increase in gratefully acknowledge the DSR technical and financial
the threshold intensity by a factor of 2. From this result, we support.
may conclude that the threshold intensity is sensitive to the
1
pulse width for the short wavelengths. This effect is less pro- L. H. Gaabour, Y. E. El Din Gamal, and G. Abdellatif, J. Mod. Phys. 3,
1683–1691 (2012).
nounced as the laser wavelength increases. 2
R. Tambay and F. L. K. Thareja, J. Appl. Phys. 70(5), 2890 (1991).
Moreover, the high threshold intensity shown for the 3
R. Tambay, D. S. Muthu, V. Kumar, and R. K. Thareja, Pramana – J.
short wavelength at small pulse width could be attributed to 4
Phys. 37(2), 163 (1991).
the effect of electron losses even at 60 lm spot diameter H. Y. Zhang, Y. Q. Xia, D. Y. Chen, P. Xiao, R. W. Fan, and Z. G. Ma,
Sci. China Ser. G 48, 78 (2005).
(12 cm focal length lens). 5
N. Glumac and G. Elliott, Opt. Lasers Eng. 45, 27–35 (2007).
6
T. B. Petrova, H. D. Ladouceur, and A. P. Baronavski, Phys. Rev. E 76,
IV. CONCLUSION 066405 (2007).
7
J. J. Camacho, M. Santos, L. Dıaz, L. J. Juan, and J. M. L. Poyato, Appl.
We investigated the threshold intensity (RMS electric Phys. A 99, 159–175 (2010).
8
field) dependence on spot diameter (diffusion length) for dif- M. Thiyagarajan and S. Thompson, J. Appl. Phys. 111, 073302 (2012).
9
A. E. Hussein, P. K. Diwakar, S. S. Harilal, and A. Hassanein, J. Appl.
ferent laser wavelengths: in laser-induced plasma in atmo- Phys. 113, 143305 (2013).
spheric air measured by Ref. 2. The results showed the 10
C. A. David, Lasers-Induced Plasmas and Applications (CRC, Boca
following findings: Raton, 1989).
11
J. P. Chen, X. W. Ni, J. Lu, and B. M. Bian, Opt. Commun. 176, 437–440
• For all wavelengths, the calculated threshold intensity (2000).
12
(RMS electric field) is found to be in reasonable agree- N. Khan, N. Mariun, I. Aris, and J. Yeak, New J. Phys. 4, 61.1–61.20

19 May 2024 19:12:08

(2002).
ment with the measured values. Both undergo increase for 13
J. Stricker and J. G. Parker, J. Appl. Phys. 53, 851–856 (1982).
small spot diameters (diffusion lengths) as compared to 14
D. C. Smith, Opt. Eng. 20(6), 206962 (1981).
the large ones. 15
A. Schmitt-Sody, H. G. Kurz, L. Berge, S. Skupin, and P. Polynkin, New
• The executed time evolution of the electron density during J. Phys. 18, 093005 (2016).
16
N. A. Popov, Plasma Sources Sci. Technol. 25(4), 044003 (2016).
the laser pulse provided a comprehensive view of different 17
M. Thiyagarajan and J. Scharer, J. Appl. Phys. 104, 013303 (2008).
electron gain and loss processes involved in laser-plasma 18
R. Buffa, Opt. Lett. 20, 204 (1995).
19
generation in atmospheric air, and their dependence on L. V. Keldysh, Sov. Phys. JETP 20(5), 1307 (1965); available at https://
spot diameter (diffusion length) and laser wavelength. doi.org/SPHJAR.
20
Z. Mics, F. Kadlec, P. Kuzel, P. Jungwirth, S. E. Bradforth, and V. Ara
• Moreover, at short wavelengths, the change of the spot Apkarian, J. Chem. Phys. 123, 104310 (2005).
diameter showed less effect on the variation of the elec- 21
Handbook of Mathematical Functions with Formulas, Graphs, and
tron growth rate. Mathematical Tables, edited by M. Abramowitz and I. A. Stegun (U.S.
• The fast growth of the electron density during the early Government Printing Office, Washington, DC, 1964), p. 319.
22
P. K. Kennedy, IEEE J. Quantum Electron. 31(12), 2241 (1995).
times (<1.0 ns) of the laser pulse with the increase in the 23
Y. R. Shen, The Principle of Nonlinear Optics (Wiley, New York, 1984).
photon energy explored the role played by the photoioni- 24
Y. Zeldovich and Y. Raizer, “Cascade ionization of a gas by a light pulse,”
zation mechanism in the breakdown of atmospheric air. 25
Sov. Phys. JETP 20(3), 772–780 (1965).
This instantaneous electron growth acts to compete with N. Kroll and K. Watson, Phys. Rev. A 5(4), 1883–1905 (1972).
26
G. M. Weyl, in Laser Induced Plasma and Applications, edited by L. J.
electron loss. Hence, there is a lower threshold intensity Radziemski, D. A. Cremers (Marcel Dekker, New York,1989)
for the wavelength range 532 nm–266 nm. 27
M. A. Biondi, “Recombination,” in Principles of Laser Plasmas, edited by
• These results supported the enhancement of ionization G. Bekefi (Wiley, New York, 1976), pp. 125–157.
28
X. M. Zhao, J.-C. Diels, C. Y. Wang, and J. M. Elizondo, IEEE J.
through the multiphoton absorption process for short
Quantum Electron. 31, 599–612 (1995).
wavelengths while inverse Bremsstrahlung is the domi- 29
D. K. Davies and P. J. Chantry, “Air chemistry measurements II,”
nant absorption mechanism at 1060 nm. Technical Report No. 352, Westing-House Electric Co., Pitlsburgh, PA,
• Second study of air breakdown over a pressure range of 1984.
30
C. L. Longmire, R. L. Gardner, J. L. Gilbert, and M. H. Frese, “A physical
24–760 Torr induced at 355, 532, and 1060 nm with pulse model of nuclear lightning,” Lightning Phenomenology Notes, Technical
widths varying from 1.3 ns up to 2.5 ns and spot diameter Report No. 4, Mission Research Corporation, Santa Barbara, CA (1982).
31
60 lm (Ref. 3) showed different behaviors regarding H. Buscher, R. Tomlinson, and E. Daman, Phys. Rev. Lett. 15, 847 (1965).