Spectrochimica Acta Part B 122 (2016) 1–8

Contents lists available at ScienceDirect

Spectrochimica Acta Part B

journal homepage: www.elsevier.com/locate/sab

Evaluation of explosive sublimation as the mechanism of nanosecond

laser ablation of tungsten under vacuum conditions
Hassan Yousefi Oderji, Nazar Farid, Liying Sun, Cailong Fu, Hongbin Ding ⁎
School of Physics and Optical Engineering, Key Laboratory of Materials Modification by Laser, Ion and Electron Beams, Chinese Ministry of Education, Dalian University of Technology, Dalian
116024, PR China

a r t i c l e i n f o a b s t r a c t

Article history: A non-equilibrium mechanism for nanosecond laser ablation is suggested herein, and its predictions are com-
Received 28 February 2015 pared to the results of W experiments performed under vacuum conditions. A mechanism of particle formation
12 April 2016 is explained via this model, with partial sublimation of the superheated irradiated zone of the target considered
Accepted 22 May 2016
to be the mechanism of laser ablation. In this study, a mixture of vapor and particles was explosively generated
Available online 24 May 2016
and subsequently prevented the rest of a laser pulse from reaching its intended target. This mechanism was
Keywords:
found to play an essential role in the ablation of W under vacuum conditions, and it provides a theoretical justi-
Laser-induced breakdown spectroscopy fication for particle formation. Moreover, special considerations were taken into account for the expansion of
LIBS plasma into a vacuum. The model was evaluated by measuring the mass of ablated particles using a quartz crystal
Laser ablation modeling deposition monitor and time-resolved optical emission spectroscopy. The results of this model were found to be
Laser ablation mechanism in good agreement with experimental values.
ITER © 2016 Elsevier B.V. All rights reserved.
Tungsten

1. Introduction
Laser ablation and resulting laser-produced plasma have been found
to have numerous applications in a variety of scientific and technologi-
cal fields, ranging from materials, chemical analytics, and biological sci-
ences to the exploration of life on other planets [1–4]. Among these
applications are laser-induced ablation spectroscopy (LIAS), laser-in-
duced desorption spectroscopy (LIDS), and laser-induced breakdown
spectroscopy (LIBS), in which laser-based methods are used in combi-
nation with spectroscopy for analytical purposes. Recently, these three
methods have been considered for use as in-situ diagnostic tools for
plasma-facing materials within the International Thermonuclear Exper-
imental Reactor (ITER), which is presently under construction [5–11].
Due to the occurrence of several interaction processes between the
edge plasmas and inner walls of fusion device chambers and divertors,
the wall materials are subjected to harmful mechanisms, such as ero-
sion, re-deposition and fuel retention. These can affect overall system
performance and can also raise safety issues, thus necessitating strict
control of steady-state and transient wall power loads to within techni-
cally acceptable limits, as well as the control of in-vessel tritium inven-
tories. Hence, the ejected wall materials should be properly monitored
via the methods consisting with the harsh conditions expected in ITER
as well as in other fusion tokamaks. Although the suggested laser-
based methods of LIAS, LIDS, and LIBS seem to be capable for this
⁎ Corresponding author.
E-mail address: hding@dlut.edu.cn (H. Ding).
purpose, Huber et al. [12] advocated that additional research on the
basic processes that support the special conditions inherent to fusion
devices is required before applying these methods in ITER.
Among these laser-based diagnostic methods that may be used in
ITER, LIBS can be performed between discharges, in which the compli-
cated interactions between the edge plasma, laser-produced plasma,
and magnetic field effects do not need to be considered. To avoid
these complexities, our research team focused its study on LIBS, in
which only the vacuum condition makes it compatible with the special
conditions found in fusion devices.
Due to its hardness and high melting point, W is a suitable element
for the ITER divertor upper baffle and dome and thus could be used in
either of two ways in high-power divertor areas [12–16]. Therefore,
knowledge of ablation due to laser irradiation, plasma plume generation
and expansion, and light emission is crucial to the design and optimal
setup of LIBS for its in-situ diagnostics. Moreover, an analogy has recent-
ly been found in regimes of W erosion within ITER-like reactors and ero-
sion due to laser ablation [17]. This investigation was thus motivated by
the quest of investigative findings related to ITER plasma–wall interac-
tions associated with environments characterized by W laser-plasma
generation and expansion (in a vacuum) at different laser–pulse irradi-
ation levels.
There have been several model-based studies on laser ablation (and
its related processes) over the years for a number of candidate metallic
targets. While many problems such as heat conduction through the tar-
get, as well as plasma formation, expansion, shielding, and emission
must be considered, one of the most crucial challenges is the actual

http://dx.doi.org/10.1016/j.sab.2016.05.008
0584-8547/© 2016 Elsevier B.V. All rights reserved.
2 H.Y. Oderji et al. / Spectrochimica Acta Part B 122 (2016) 1–8
phenomenon of ablation itself. Different mechanisms associated with
surface evaporation and phase explosion (also known as explosive boil-
ing or volumetric boiling) have been considered for laser ablation. In
1968, Anisimov [18] developed a model for evaporation based on
treating the Knudsen layer region as a “gas dynamic discontinuity”
across which certain jump conditions applied; as a result, a full expres-
sion of conservation of mass, momentum, and energy principles follow-
ed. Anisimov employed this model only for strong vaporization limit
scenarios, wherein the flow just outside the Knudsen layer was sonic.
Over subsequent years, jump conditions for subsonic expansions and
the Knudsen layer with recondensation boundary conditions were de-
veloped by many researchers [19–24], and particularly through the
more recent method proposed by Gusarov and Smurov [25]. Several
other researchers have also used the evaporation mechanism and the
jump conditions of the Knudsen layer to model the nanosecond laser
ablation process (for examples, see Refs. [26–32]). At higher laser
fluences, phase explosion (or explosive boiling) has regularly been con-
sidered the primary mechanism behind the ablation process. This
mechanism, proposed by Martynyuk [33], has likewise also been inves-
tigated by numerous researchers. Experimentally, a sudden increase in
crater depth or dramatic changes in other physical properties have
been attributed to phase explosion (for example, see Refs. [34–37]). Ex-
periments with different materials have yielded different delay times
for phase explosion [35,36,38], many of which have lasted up to the
ends of the given laser pulse periods and some of which have
transcended down into the microsecond level. The delay time has con-
ventionally been attributed to the time lag necessary for the formation
of homogeneous nuclei [19]. Recently, a multiphase collisional radiative
model was developed by Autrique et al. in which the entire target and
plume system was described by a one-dimensional hydrodynamic
model that adequately accounts for energy conservation, mass conser-
vation, and pressure relaxation [20,39]. A review on radiative models
was provided by Gornushkin and Panne [40].
For the ablation of W in vacuum, however, equilibrium explosive
boiling and normal evaporation are minimally present when
superheating occurs. As the surrounding pressure is very low, atoms
that are released from lattices expand freely. Moreover, the formation
of large particles, ejected outward from a target surface at constant ve-
locities on the order of a few meters per second, can properly be ex-
plained using a partially superheated sublimation mechanism. Note
that due to equilibrium phase transitions such as homogeneous boiling,
the ejected liquid (in its entirety) is expected to be converted into vapor.
Although some small clusters are likely to be formed by equilibrium
boiling, the generation of stable particles on the order of a micrometer
in size (as what one can typically expect from nanosecond laser ablation
of W in a vacuum), is not probable. Similar to heterogeneous and homo-
geneous boiling, normal evaporation thus cannot be considered as an
ablation mechanism. Moreover, the commonly used Clausius–
Clapeyron relation for the calculation of saturated pressure on target
surfaces leads to a value that is almost four times greater than the crit-
ical pressure at the critical temperature of W. Therefore, even at low
laser fluences, the formulations for evaporation and the Knudsen layer
developed in Refs. [18–24] are not adequate for W. Hence, in this
paper, the partial sublimation of the superheated zone is proposed as
the mechanism of the nanosecond laser ablation of W under vacuum
conditions.
For the modeling of ablation in vacuum conditions, the most conse-
quential issues relate to the boundary conditions of the plume. At the
vacuum edge, the media is not in a continuum-mode; thus, associated
hydrodynamic equations cannot be established. Any temptation to use
a very low concentration ambient gas for the vacuum boundary and to
solve the subject equations using the Godunov scheme or the Riemann
solvers leads to a strong fictitious shockwave, which in reality can never
occur in a vacuum [41].
In this investigation, a model was developed and evaluated for the
nanosecond laser ablation of W based on the more likely phenomenon
of explosive partial sublimation of a superheated irradiated zone and
resulting formation of a rarefaction wave that propagates through the
vacuum. Plasma shielding and target heating by a laser beam tail were
also evaluated. However, if targets heat to temperatures higher than
material melting points during this process, the pressure in front of
the target surface (generated from explosive sublimation) will prevent
atoms from flying freely, which will then likely result in a homogenous
boiling (volumetric vaporization) condition. Target heating by interac-
tions with adjacent plasma, and via irradiation by plasma, were
overlooked. The research team used both experimental and computa-
tional methods for this study (1) to verify the applied model and (2)
to achieve a clear representation of the involved complex phenomena.
Section 2 below is devoted to explaining the implemented model
and to reviewing the basic theories behind the modeling of laser abla-
tion. Section 3 explains the experiments performed for checking the ac-
curacy of the subject model. The associated results are accordingly
presented and discussed in Section 4, with final conclusions being pro-
vided in Section 5.
2. Model and theory
We constructed its model on the basis of experimental observations
through which the laser ablation of W under vacuum conditions pro-
duces many large particles (Fig. 1). Panel (a) of Fig. 1 shows that at a
delay time of 1 μs, the plume is split into two parts. The front part de-
picts the expanding plasma, while the hot part near the surface pro-
duces particles at later delay times (panel (b)). The particle speed is
ultimately dependent on the laser fluence and typically falls within
the range of a fraction of a meter per second to a few meters per second.
This phenomenon can be explained easily if one supposes that an irradi-
ated target zone is explosively fragmented into many atoms and parti-
cles. The number of atoms is always expected to be much higher than
the number of particles. Atoms experience many collisions and form a
high-pressure vapor plume, while particles typically only undergo a
few collisions that scatter them into a solid angle of 2π in front of the tar-
get surface. The pressure of the plume may also be increased by absorp-
tion of laser energy through an inverse Bremsstrahlung phenomenon.
Therefore, the driving force for atoms is generally much higher than
the driving force for particles. While vapor plumes can expand to ap-
proximately 1 cm away from a target surface, particles with a constant
speed of 1 nm/ns can only travel 1 μm during 1 μs (see Fig. 1(a)). As a
result, the plume is split into two parts, as is evident in Fig. 1(a).
These phenomena can be modeled by a partial sublimation of a su-
perheated irradiated zone adjacent to a vacuum. Fig. 2 schematically
represents a one-dimensional hydrodynamic model. A target zone
reaches temperatures higher than the sublimation temperature of W.
Subsequently, this superheated zone is explosively fragmented for
Fig. 1. Laser ablation of W under vacuum conditions by Nd:YAG laser, λ = 1064 nm,
FWHM = 7 ns, and laser fluence of 131 J cm−2 at (a) 1 μs delay time and (b) 700 μs
delay time.
 H.Y. Oderji et al. / Spectrochimica Acta Part B 122 (2016) 1–8 3

the surface at time t, and x is the depth inside the target from the surface
at x = 0. The value of Rf has been found to be dependent upon the laser
fluence as well as the type of target metal being used; the value for W
varies between 0.6 and 0.12 for fluences of 0.1–100 J cm−2 via ablation
by Nd:YAG laser pulses [46]. In this investigation, I(t) was assumed to
have a temporal Gaussian distribution with τ = 7 ns (FWHM) centered
at 10.5 ns, and a maximum value related to the fluence, F, by
IMax ≈0.94F/τ.
Neglecting radiative loss from the surface and heat exchange be-
tween the surface and vapor, laser beam energy at the surface is bal-
anced by the heat flux (into the target) and the energy transferred by
atoms vaporized into the plume. Hence, the boundary conditions for
Eq. (1) at x = 0 are


  ∂T ðt; xÞ
1−R f Iðt Þ ¼ −λðt; xÞ þ ρðt Þurec ðt ÞΔH vap ; ð2Þ
Fig. 2. Explosive partial sublimation model for nanosecond laser ablation of W under
∂x x¼0
vacuum conditions. Superheated zone is explosively fragmented into atoms and
particles. Atoms form vapor plume expanding into vacuum, and particles freely fly into
where ΔHvap is the heat of vaporization. The rate of surface recession,
vacuum at constant velocities.
urec(t), and laser irradiation at the target surface, I(t), are strongly
coupled to the equations governing plasma plume expansion and plas-
ma shielding.
particles and atoms through a non-equilibrium pathway. The reason for
Plume expansion can be described using time–dependent Euler
fragmentation of this superheated zone into a mixture of particles and
equations, which are a system of nonlinear hyperbolic conservation
atoms (and not into a completely atomized plume) is that the value of
laws that govern the dynamics of a gas for which the effects of body
superheating is less than the enthalpy of sublimation and is thus not
forces, viscous stresses, and heat flux are neglected. In one dimension,
high enough to cause complete atomization of the superheated zone.
including source terms, the Euler equations in compact form are as fol-
Due to the physical explosion, the produced particles and atoms are ac-
lows [41]:
celerated outward from the surface of the target. Moreover, due to the
numerous atomic collisions between atoms, a vapor plume is ultimately
formed, which expands into the vacuum while the particles travel at Ut þ FðUÞx ¼ SðUÞ ð3Þ
constant velocities.
In the course of irradiating metals by nanosecond laser pulses, con- and
duction band electrons absorb the energy of photons through inverse
Bremsstrahlung and transfer it to the lattice through numerous colli- 32 2 3 2 3
ρ ρv 0
sions with one another and with lattice phonons [24,31]. Therefore, 4 5 4 2 5
U ¼ ρv ; F ¼ ρv þ p ; S ¼ 4 0 5;
for nanosecond laser–pulse interactions with metal targets, it is reason- E vðE þ pÞ aIB I laser −ε rad
able to assume that the electromagnetic energy from laser light is im-
mediately converted into heat and that a local thermodynamic
where ρ is the vapor density, v is the vapor stream velocity, E is the total
equilibrium state is hence established at the absorption penetration
energy, and p is the pressure at every localized part of plume. For the en-
depth (lp), which is the numerical reciprocal of the absorption coeffi-
ergy source term, only absorption of laser beam energy through the in-
cient (α) [42]. α is dependent upon the extinction coefficient of the tar-
verse Bremsstrahlung of electrons in the fields of neutral particles and
get metal and the wavelength of the laser beam; the absorption
ions (aIBIlaser) is considered, as well as the loss of energy due to the
coefficient of W by a laser beam with a wavelength of 1064 nm is
Bremsstrahlung process (εrad).
about 4.45 × 107 m−1 [43].
A caloric equation of state is also required to close the Euler equa-
The temporal evolution and spatial evolution of the temperature in-
tions. In this discussion, it is assumed that the vapor behaves like an
side the target are governed by the heat conduction equation; in one di-
ideal gas, and only ionizations up to the second level are considered;
mension [24,44],
thus, the quantity of total energy is given by the summation of the kinet-

 ic, internal, and ionization energies:
∂T ðt; xÞ ∂T ðt; xÞ ∂ ∂T ðt; xÞ
cp ðT ÞρðT Þ −urec ðt Þ ¼ λðx; T Þ ; ð1Þ
∂t ∂x ∂x ∂x
1 2 p ρxþ ρx2þ
E¼ ρv þ þ IP 1 þ ðIP 1 þ IP 2 Þ þ ⋯; ð4Þ
where x is the direction normal to the target surface and T, Cp, ρ, λ, and 2 ðγ−1Þ MW MW
urec(t) are, respectively, the temperature, specific heat, mass density,
thermal conductivity, and rate of surface recession. This equation can where γ is the ratio of specific heat; IP1 and IP2 are the first and second
be derived from Fourier's law, which states that local heat flux is propor- ionization potentials, respectively; MW is the molar mass of W; and x+
tional to the negative of the temperature gradient [44]. However, and x2+ are the fractions of W+ and W2+, respectively [41,47].
Fourier's law is subjected to some restrictions with regard to tempera- Prior to the termination of the laser pulse, if the vapor temperature
ture gradient. According to Harrington [43], the temperature gradient increases enough to ionize a few W atoms (which means that the
should be constant over at least 10 mean free paths of heat carriers. fluence should be higher than a threshold value), the resulting free elec-
Peterlongo et al. [24] checked the validity of Fourier's law under the trons absorb photons from the laser beam through inverse Bremsstrah-
conditions of laser ablations and confirmed its applicability within the lung, which is known as plasma shielding. This absorption ultimately
nanosecond regime. supplies more energy to atoms undergoing ionization, with a plasma
The absorption of laser light by a target may be added to the right formed as a result. The inverse Bremsstrahlung process for an electron
side of Eq. (1) as a source term following the Beer–Lambert law [45], needs to occur in a field of heavy particles (neutral particles or ions)
(1 − Rf)αI(t) exp (−αx), or as a boundary property (1 − Rf)I(t), where to conserve momentum. aIB in Eq. (3) is thus given by the summation
Rf is the surface reflectivity, I(t) is the incident laser irradiation level at of the electron–neutral and electron–ion inverse Bremsstrahlung
4 H.Y. Oderji et al. / Spectrochimica Acta Part B 122 (2016) 1–8

absorption coefficients [48]: environment. This is therefore a non-equilibrium phenomenon, and

  2 there is no requirement for the system to wait for an equilibrium pro-
hc ρ cess (i.e., homogeneous or volumetric vaporization) to change its
α IB;e‐n ¼ 1− exp − Q xe x0 ; ð5Þ
λkB T m2W phase. When superheating exceeds a threshold, the irradiated zone dis-
integrates or fragments into a mixture of atoms and particles of different
and sizes. For metals with low melting temperatures, most of the particles

  0 6 3  1=2 2 may not be stable at even the melting temperature. It has been found
hc 4e λ xe 2π ρ that a common characteristic of nanoparticles is that their melting tem-
α IB;e‐i ¼ 1− exp − ðxþ þ 4x2þ Þ; ð6Þ
λkB T 4
3hc me 3me kT m2W peratures are considerably lower than the melting temperatures of bulk
materials (for example, see Refs. [50,51]). For W, as the results of this
where Q is the cross-section for photon absorption by an electron dur- study show (see Fig. 1), laser ablation produces many particles that
ing the collision with neutral particles; λ is the laser wavelength; e0 ¼ e eject from the target. To calculate the total mass of these ejected parti-
pffiffiffiffiffiffiffiffiffiffiffi
= 4πε0 ; me and e are the mass and charge of an electron, respectively; cles, a Quartz Crystal deposition Monitor (QCM) was used in experi-
ε0, kB, c, and h are the vacuum permittivity, Boltzmann constant, velocity ments and discussed in the following section.
of light, and Planck's constant, respectively; mW is the mass of one W As discussed previously, the superheated zone explosively changes
atom; and xe,x0, and x+/x2+ are the fractions of electrons, neutral parti- to a mixture of particles and vapor. At the start of ejection, particles ex-
cles, and ions, respectively. perience a few collisions with each other and subsequently scatter at
Conversely, a loss of energy (εrad) occurs in the plasma before and random angles relative to the target surface. Moreover, when numerous
after the pulse duration due to the Bremsstrahlung process: atoms experience collisions, a vapor plume is formed as a result. If the
temperature of the vapor plume is high or if it absorbs energy from a
 1
2πkB T 2 32πe0 6 ρ2 laser beam, a plasma plume is ultimately formed. The temperatures,
εrad ¼ xe ðxþ þ 4x2þ Þ; ð7Þ pressures, and densities of vapor plumes can be approximated using
3me 3hme c3 m2W
the ideal gas law. The mass of vapor can be related to the value or excess
where the symbols are as defined above. Temperature can be related to energy of superheating, Esuper = Mc(Tsuper − Tsub), where M is the mass
the partial pressures of vapor and electrons using the ideal gas law, of the superheated zone, c is the specific heat, and Tsuper and Tsub are
the average and sublimation temperatures of the superheated zone, re-
mW p spectively. Esuper supplies the required energy for the partial sublima-
T¼ : ð8Þ
ð1 þ xe ÞkB ρ tion of the superheated zone and the kinetic energies of the involved
atoms and particles. Therefore, Esuper/(ΔHsub + EK , a + EK , p) portion of
Furthermore, assuming local thermal equilibrium has been the superheated zone are converted to vapor, where EK,a is the kinetic
established, the electron and ions fractions can be calculated by the energy of the subject atoms, and EK,p is the kinetic energy of the subject
Saha equations [49] (Eqs. (9) and (10)) in combination with the conser- particles. The density of the vapor is ρ = Nama/V, where ma is the mass of
vation of matter (Eq. (11)) and conservation of electric charge (Eq. one atom, and Na is the number of atoms in the superheated zone vol-
(12)): ume, V. The pressure is separately calculated via the ideal gas law.
 3
 When the vapor temperature is high, it can hence be ionized and ul-
xþ xe mW 2Z þ ðT Þ 2πme kB T 2 −kIPB1T timately generate plasma. The nonlinear Eqs. (9)–(12) above can be
¼ e ; ð9Þ
x0 ρ Z 0 ðT Þ h
2
solved by a numerical method, such as the Newton–Raphson method,
to calculate the characteristics of the plasma. Subsequently, the plasma
 3

x2þ xe mW 2Z 2þ ðT Þ 2πme kB T 2 −kIPB2T plume expands into the vacuum and generates a rarefaction wave with
¼ e ; ð10Þ pffiffiffiffiffiffiffiffiffiffiffiffi
xþ ρ Z þ ðT Þ h
2 a front velocity of vf = 2a/(γ − 1), where a ¼ γp=ρ is the speed of
sound [41]. The physical variables of the rarefaction wave at time t
x0 þ xþ þ x2þ ¼ 1; ð11Þ and location x can be calculated using the following equations [41]:
8
and < WL if x=t ≤−a
W ðx; t Þ ¼ W fan if −abx=tbv f ; ð14Þ
xþ þ 2x2þ ¼ xe ; ð12Þ :
W0 if x=t ≥v f

where Z0(T), Z+(T), and Z2+(T) are the electronic partition functions for where WL = {ρL, vL = 0,pL} and W0 = {ρ = 0,u = vf, p = 0} are the physi-
neutral W, W+, and W2+, given by cal variables of the plume and vacuum, and Wfan represents the rarefac-
X tion fan.
Z ðT Þ ¼ g i e−Ei =T ; ð13Þ
8
2
ðγ−1Þ
x γ−1
i
>
> 2
>
> ρ ¼ ρ −
>
> L
ðγ−1Þ ðγ þ 1Þa t
where Ei is the energy of an electronic state with the degeneracy or sta- >
>
tistical weight of gi. The values of Ei and gi can be obtained from the U.S.
< 2 h xi
W fan ¼ u¼ aþ ð15Þ
National Institute of Standards and Technology. >
> ðγ þ 1Þ t
>
>
 2γ
However, an appropriate solution method and proper boundary
>
>
> 2 ðγ−1Þ
x γ−1
>
: p ¼ p L −
conditions are required to solve these nonlinear equations. For this γ þ 1 ðγ þ 1Þa t
model, it is firstly assumed that superheating might occur at high laser
fluences such that the thermodynamic properties in Eq. (1) are correlat- The head of the rarefaction wave cannot propagate beyond t =
ed to the solid phase. This assumption is reasonable at the high irradia- xsup ⋅ a, where xsup is the length of the superheated zone, as it reaches
tion levels of nanosecond laser beams in which the heating is very rapid. the target surface at this point. Subsequently, the velocity of the wave
The second assumption is explosive partial sublimation. In the vacuum head remains zero while its tail propagates via the front-end constant
conditions associated with this model, the external pressure is zero, velocity. For the fan of the rarefaction wave, the velocity linearly in-
and thus the superheated zone has already obtained enough energy to creases from zero to the highest potential value at the front. Therefore,
break some bonds in the metallic lattice; as soon as the bonds between the velocity at a specified location of the plume,vx = vfx/xp, tends para-
some atoms or grains are broken, they fly freely into the vacuum bolically toward zero as the plume length approaches infinity. The
 H.Y. Oderji et al. / Spectrochimica Acta Part B 122 (2016) 1–8 5

wave is discretized into a constant number of computational cells at the arctan (a/R), θ0 is the sensor angle relative to the surface normal, a is
time it reached the target surface and ultimately calculated the conser- the radius of the sensor, and R is the distance of the sensor from the tar-
vatives properties of ρ,ρv, and E (Eq. (3)) in each cell. Because the medi- get. The mass deposited on a ring is given by the mass measured by the
um is not continuous at the vacuum boundary, it is not possible to keep sensor located on that ring multiplied by the ratio of the solid angle of
the length of cells fixed and to use the Godunov scheme and Riemann the sensor to the solid angle of the ring.
solvers to calculate the fluxes on the walls between the cells. If one For the temporal study of optical emissions, a 0.75 m triple grating
did so, a fictitious shockwave would appear and would subsequently (3600 lines/mm, 1200 lines/mm, and 600 lines/mm) spectrograph (SR
propagate at the boundary conditions, ultimately skewing all of the re- 750, Andor) was used. The emission from the plasma plume was collect-
sults. Instead, one can expand the size of cells in cases in which no fluxes ed normal to the direction of expansion using a combination of two
can occur between the cells themselves. By increasing the length of lenses. The timing between the incident laser pulse and the detector
plume to xp + Δxp, all computational cells grow equivalently by Δxp/N, system was controlled by a delay generator (DG645).
where N is the number of cells. Therefore, in every time step, the conser-
vatives can be rescaled in terms of the volumetric increment of the cell, 4. Results and discussion
with plasma formation and shielding ultimately being evaluated by
solving the nonlinear Eqs. (9)–(12). The simulations commenced by assuming a predetermined
superheating temperature (Ts) for all the fluences (i.e. Ts = 0.8Tc,
where Tc is the critical temperature). The reason for this assumption is
3. Experiments
due to the dependence of the temperature of superheating on the pres-
sure that tends to almost 0.8Tc at very low pressures [52]. The incre-
To determine the particle masses and vapor densities, experimental
ments of W surface temperature per exposure duration are provided
results were obtained by performing laser ablations using an Nd:YAG
in Fig. 4 for different laser fluences. The total duration of the Gaussian
laser (Quantal) with a fundamental wavelength of 1064 nm, operating
laser pulses was 21 ns by the highest laser irradiance at 10.5 ns. The
at 20 Hz, with a full-width-at-half-maximum (FWHM) of 7 ns. All abla-
only process for cooling a surface prior to sublimation is assumed to
tion operations were performed in a high vacuum with a base pressure
be the thermal conductivity of W. Competing processes of laser heating
of 1.3 × 10−5 Torr. The laser beam was delivered through the quartz
and heat conductivity cooling (from the surface of W to its inner layers)
window of the apparatus chamber and was focused perpendicular to
results in increasing surface temperatures up to the delay time of 13 ns
the target surface using a plano-convex lens of focal length equal to
(rising time) for fluences lower than 3.5 J cm−2. Subsequently, the
40 cm. The estimated spot radius at the target surface was 300 μm.
cooling process dominates the laser heating process, with the surface
The laser fluence at the target surface was varied by modifying the
temperature ultimately decreasing. Moreover, Fig. 4 shows that the
laser pulse energy while the spot size was kept constant. To increase
temperature of the heated zone cannot reach the superheating temper-
the pulse-to-pulse energy stability, a combination half-wave plate and
ature for fluences lower than 3.5 J cm−2. The threshold fluence for W
cube polarizer were used to vary the laser energy. In order to avoid a po-
damage can be obtained by assuming the value of 0.8Tc as the
tential crater effect due to successive laser pulses, the target was
superheating temperature. If superheating occurs at lower tempera-
mounted on a computer-controlled rotational target holder, which pro-
tures, the threshold fluence may therefore also occur at lower fluences.
vided a fresh surface for each measurement. Two QCMs at different an-
For fluences higher than the threshold value, Fig. 4 shows that
gles (12° and 45°) at a 15 cm distance from the target surface via an
superheating occurs at earlier times by increasing the fluence level.
imaginary spherical surface were employed, as depicted in Fig. 3. For
For example, superheating occurs at delay times shorter than 3 ns
each experimental trial, the target exposed to 2000 laser pulses and sub-
when the W surface is exposed to fluences greater than 500 J cm−2.
sequently averaged the results in order to ascertain the net deposition
When superheating occurs during a very short time period, heat can
on each sensor due to a single pulse. The angular distribution of particles
penetrate to a lesser extent within the target via heat conductivity, ulti-
ejected from metals via the nanosecond laser ablation process has been
mately affecting the thickness of the superheated layer and the ablated
found to be of Gaussian character, in which the density is maximized at
mass.
0° (relative to the target surface normal), while it decreases with in-
The masses of the superheated layer and the resulting vapor are pre-
creasing angle (and is accordingly ~10% of the maximum at 90°) [50].
sented in Fig. 5. For simplicity, the vapor mass was calculated by
Hence, initially the data was fit to a Gaussian function to find the angular
distribution of particles on the plane at which the QCM sensors were lo-
cated. Next, the function was integrated over all solid angles (rings in
Fig. 3) to find the net mass deposition of W due to a single laser pulse.
The solid angles of a sensor and a ring are respectively ΩS ¼ 2πð1−R=
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
R2 þ a 2 Þ and Ωr = 2π(cos(θ+) − cos (θ−)), where θ± = θ0 ±

Fig. 3. Imaginary hemispherical surface surrounding target to calculate total mass ablation Fig. 4. Increment in temperature of target surface irradiated by Gaussian laser beam with
of particles due to laser beam. Sensors ‘A’ and ‘B’ are located at angles of 12° and 45°, λ = 1064 nm and FWHM of 7 ns at different fluences. For fluences higher than 3.6 J cm−2,
respectively. Radius of sphere is 15 cm. superheating occurs.
6 H.Y. Oderji et al. / Spectrochimica Acta Part B 122 (2016) 1–8
Fig. 5. Superheated mass due to nanosecond laser pulse with λ = 1064 nm and FWHM of
7 ns at different fluences (right), and initial mass of vapor formed from superheated mass
due to superheating (left).
considering only the enthalpy of sublimation, with the potential contri-
butions of the kinetic energies of atoms and particles disregarded. Fig. 5
shows that the thickness of the superheated zone decreases with in-
creasing laser fluence, which is primarily due to the time scale of heat
conduction. At higher laser fluences, surface temperatures rapidly
reach superheating temperatures so that heat cannot effectively be con-
ducted into the inner layers of the target. However, the ratio of excess
superheating energy to the superheated zone mass is minimal at
lower laser fluences. This characteristic implies that although particles
are likely generated at lower fluences, their associated kinetic energies
are still quite low; thus, they cannot escape from the crater and the tar-
get focus area of the laser beam. Therefore, they can still absorb energy
from the laser beam to facilitate potential evaporation, or they can re-
join and ultimately dissipate their elevated temperatures due to the
thermal conductivity of W. It is also probable that superheating temper-
atures vary by laser fluence levels. The difference in the trends of Figs. 5
and 6 could be due to the following reasons; (a) as the model is 1D, the
lateral expansion of the plasma is not considered, this could lead to a re-
duction of plasma shielding, particularly at the higher laser fluences,
allowing for more of the target to be ablated. (b) As the lens to target
distance was kept constant for the experiments, the spatial profile of
the laser pulse might affect the comparison to the model. The area of
Fig. 6. Average deposited mass of ablated W generated by nanosecond laser pulse,
Nd:YAG, λ = 1064 nm, FWHM = 7 ns, at different laser fluences.
the laser pulse of sufficient intensity to ablate the target could vary by
several orders of magnitude from low to high fluences (e.g. at the higher
fluencies the full 300 μm could ablate, but at the lower fluences only 25–
30 μm near the peak of the spatial profile of the laser pulse ablated).
Moreover, the neglecting of plasma target interactions that becomes in-
creasingly important by increasing the laser fluences could also be the
reason. To better understand the dependence of ablation on laser
fluence, the mass of released particles from the W target was measured
by using a QCM at different fluences. To perform these measurements,
we analyzed the frequency variation of the quartz sensor in the QCM
during a time interval of at least 100 s (2000 pulses) in determined
laser fluences at two different angles (12° and 45°) at a 15 cm target dis-
tance. Fig. 7 displays the sensor responses to a deposited mass at a
fluence of 148.5 J cm− 2. The time interval of data acquisition by the
QCM varied between 0.64 s and 1.00 s. At several time intervals, the sen-
sor frequency values remained constant, indicating the absence of par-
ticles or possibly the inability of the sensor to detect lighter particles
due to sensitivity limitations. To take into account the accumulation of
light particles, we also calculated the variation of sensor frequency dur-
ing the entire measurement period. However, no significant differences
were discovered between the results of the two separate methods.
Therefore, it can be deduced that the deposition of ablated vapor on
the subject sensors is negligible and that the measured mass is repre-
sentative only for the particles. Moreover, the deposited mass on the
sensor located at 12° was much higher than that deposited on the one
at 45°.
The total mass of particles ablated due to a single laser pulse was cal-
culated by the summation over all solid angles of the hemispherical
imaginary surface surrounding the target (Fig. 3). The results at differ-
ent fluences are given in Fig. 6 for a sample of W irradiated by the
Nd:YAG laser with wavelength of 1064 nm, operating at 20 Hz, and a
FWHM of 7 ns. Fig. 6 shows that the particle ejection from the W in-
creases with fluence. This behavior is likely due to the higher kinetic en-
ergies of subject particles produced by higher laser fluences. Fast (i.e.,
highly energetic) particles can escape from the crater without signifi-
cant interaction with the laser pulse. As a result, in order to better eval-
uate particle and vapor mass ratio variations, the contributions of the
kinetic energies of the particles and atoms must be taken into account.
To evaluate the plume dynamics, we studied the time-resolved opti-
cal emissions of two spectral lines for W. For this purpose, a 0.75 m triple
grating (3600 lines/mm, 1200 lines/mm, and 600 lines/mm) spectro-
graph (SR 750, Andor) coupled to an intensified CCD camera (iStar)
was used to collect the wavelength-dispersed spectral features of the
plasma emissions. The emissions from the expanding plasma were
Fig. 7. Deposited masses on sensors shown in Fig. 3 averaged over time intervals of 1 s
during 180 s exposure by nanosecond laser beam at fluence of 148.5 J cm−2.
 H.Y. Oderji et al. / Spectrochimica Acta Part B 122 (2016) 1–8
Fig. 8. Temporal variation of emission intensities of WI 429.4 nm and WII 434.8 nm from the laser-produced W plasma with 131 J cm−2 at 60 μm and 1360 μm. A gate width of 30 ns and a
slit width of 30 μm were used to acquire emission intensities.
collected in the direction perpendicular to plume expansion and were
sent to the entrance slit (30 μm) of the spectrograph. A gate width of
30 ns was used to collect such emissions. Fig. 8 shows the time-resolved
emission intensity of the neutral (WI 429.4 nm) and ionic (WII
434.8 nm) lines at 60 μm and 1360 μm, respectively, from the target. It
should be mentioned that the same camera gain was used for each spec-
trum recording and that the provided intensities are actually the inte-
grated line intensities (to reduce effects of spectral broadening) after
subtracting the background continuum. In Fig. 8, the intensities of the
WI 429.4 nm and WII 434.81 nm lines are normalized to their respective
maximum values. The predicted behavior in Fig. 8 can be readily
interpreted via the rarefaction wave of Eq. (14). Due to expansion into
the vacuum, the physical variables of density, pressure, and tempera-
ture exhibit typical behaviors, as shown in Fig. 9. A typical location (at
1360 au) has been marked in Fig. 9 by an arrow to check the variations
Fig. 9. The evolution of a rarefaction wave at three consecutive times of t1, t2, and t3. The
value at the position demarcated by the arrow is almost zero at time t1, then increases
at time t2, and then again decreases at time t3. This behavior is in agreement with the
experimental time evolution presented in Fig. 8.
7
of the physical variables at three consecutive times (t1 to t3) at this
point. At t1, the density at the marked location is almost zero. When
the rarefaction wave reaches the subject location, however, the density
starts to increase. It ultimately reaches a maximum value at t2 and then
begins to decrease back to zero immediately afterwards. All of the
graphs in Fig. 8 confirm the behavior of the rarefaction wave (Eq.
(14)); for the line WII 434.8 nm at 60 μm, the rising part of the curve oc-
curred prior to the delay time at 100 ns.
5. Conclusion
W is an important material in fusion reactors such as the ITER. It is
subjected to many destructive reactions and has to be inspected re-
motely. LIBS has been envisaged for the in-situ diagnostics of W wall
material. Therefore, this investigation was devoted to the modeling of
the laser ablation of W under vacuum conditions, which are similar to
the conditions that regularly exist within fusion reactor environments.
The results generally showed that nanosecond laser ablation of W pro-
duces many relatively large particles that eject from the material sur-
face. No equilibrium mechanisms can fully predict particle formations
of this size. Therefore, we modeled the non-equilibrium mechanism of
explosive partial sublimation in the superheated zone. According to
this model, a W irradiated zone receives a level of energy or heat from
superheating that is not large enough to cause sublimation of “whole
materials” in the superheated zone. Hence, only a portion of lattice
bonds can therefore be broken in this scenario. This phenomenon hap-
pens explosively and generates a mixture of atoms and particles as a re-
sult. Subsequently, such atoms form a vapor plume; the subject
particles, in contrast, fly freely. We measured the mass of particles ex-
perimentally using two QCM sensors located at different angles to the
normal of the target surface and also theoretically employed a heat con-
duction equation to confirm the results of this experiment. The theoret-
ical results showed that the mass of the superheated zone was higher at
lower laser fluences. It was moreover concluded that the expansion of
vapor plumes in a vacuum can only produce rarefaction waves. This re-
sult was confirmed by performing numerous related experiments via
time-resolved emission spectroscopy. In conclusion, the explosive par-
tial sublimation mechanism is capable of explaining particle formation
8 H.Y. Oderji et al. / Spectrochimica Acta Part B 122 (2016) 1–8
and other features of the nanosecond laser ablation of W under vacuum
conditions.
Acknowledgments
This work was supported by the National Natural Science Founda-
tion of China (Nos. 11175035 and 11475039), the National Magnetic
Confinement Fusion Science Program of China (Nos. 2013GB109005),
Chinesisch-Deutsches Forschungsprojekt (GZ768), and the Fundamen-
tal Research Funds for the Central Universities (DUT15TD44).
References
[1] R.E. Russo, X. Mao, J.J. Gonzalez, V. Zorba, J. Yoo, Laser ablation in analytical chemis-
try, Anal. Chem. 85 (2013) 6162–6177.
[2] C.R. Phipps, Laser Ablation and its Applications, Springer, 2007.
[3] J.C. Miller, Laser Ablation: Principles and Applications, first ed. Springer-Verlag, Ber-
lin, Germany, 1994.
[4] J.N. Walsh, Laser-ablation-ICPMS in the earth sciences. Principles and applications,
Mineral. Mag. 66 (2002) 375.
[5] S. Almaviva, L. Caneve, F. Colao, R. Fantoni, G. Maddaluno, Remote-LIBS characteriza-
tion of ITER-like plasma facing materials, J. Nucl. Mater. 421 (2012) 73–79.
[6] L. Mercadier, A. Semerok, P.A. Kizub, A.V. Leontyev, J. Hermann, C. Grisolia, P.Y. Thro,
In-depth analysis of ITER-like samples composition using laser-induced breakdown
spectroscopy, J. Nucl. Mater. 414 (2011) 485–491.
[7] V. Philipps, A. Malaquias, A. Hakola, J. Karhunen, G. Maddaluno, S. Almaviva, L.
Caneve, F. Colao, E. Fortuna, P. Gasior, M. Kubkowska, A. Czarnecka, M. Laan, A.
Lissovski, P. Paris, H.J. Van Der Meiden, P. Petersson, M. Rubel, A. Huber, M.
Zlobinski, B. Schweer, N. Gierse, Q. Xiao, G. Sergienko, Development of laser-based
techniques for in situ characterization of the first wall in ITER and future fusion de-
vices, Nucl. Fusion 53 (2013) 093002.
[8] N. Gierse, B. Schweer, A. Huber, O. Karger, V. Philipps, U. Samm, G. Sergienko, In situ
characterisation of hydrocarbon layers in TEXTOR by laser induced ablation and
laser induced breakdown spectroscopy, J. Nucl. Mater. 415 (2011) S1195–S1198.
[9] C. Grisolia, A. Semerok, J.M. Weulersse, F. Le Guern, S. Fomichev, F. Brygo, P. Fichet,
P.Y. Thro, P. Coad, N. Bekris, M. Stamp, S. Rosanvallon, G. Piazza, In-situ tokamak
laser applications for detritiation and co-deposited layers studies, J. Nucl. Mater.
363–365 (2007) 1138–1147.
[10] Q. Xiao, A. Huber, G. Sergienko, B. Schweer, P. Mertens, A. Kubina, V. Philipps, H.
Ding, Application of laser-induced breakdown spectroscopy for characterization of
material deposits and tritium retention in fusion devices, Fusion Eng. Des. 88
(2013) 1813–1817.
[11] R. Hai, Q. Xiao, L. Zhang, H. Ding, Y. Zhou, L. Yan, Characterization and removal of co-
deposition on the first mirror of HL-2A by excimer laser cleaning, J. Nucl. Mater. 436
(2013) 118–122.
[12] R.A. Pitts, S. Carpentier, F. Escourbiac, T. Hirai, V. Komarov, A.S. Kukushkin, S. Lisgo, A.
Loarte, M. Merola, R. Mitteau, A.R. Raffray, M. Shimada, P.C. Stangeby, Physics basis
and design of the ITER plasma-facing components, J. Nucl. Mater. 415 (2011)
S957–S964.
[13] G. Federici, C.H. Skinner, J.N. Brooks, J.P. Coad, C. Grisolia, A.A. Haasz, A. Hassanein, V.
Philipps, C.S. Pitcher, J. Roth, W.R. Wampler, D.G. Whyte, Plasma–material interac-
tions in current tokamaks and their implications for next step fusion reactors,
Nucl. Fusion 41 (2001) 1967.
[14] V. Philipps, Tungsten as material for plasma-facing components in fusion devices, J.
Nucl. Mater. 415 (2011) S2–S9.
[15] M. Merola, D. Loesser, A. Martin, P. Chappuis, R. Mitteau, V. Komarov, R.A. Pitts, S.
Gicquel, V. Barabash, L. Giancarli, J. Palmer, M. Nakahira, A. Loarte, D. Campbell, R.
Eaton, A. Kukushkin, M. Sugihara, F. Zhang, C.S. Kim, R. Raffray, L. Ferrand, D. Yao,
S. Sadakov, A. Furmanek, V. Rozov, T. Hirai, F. Escourbiac, T. Jokinen, B. Calcagno, S.
Mori, ITER plasma-facing components, Fusion Eng. Des. 85 (2010) 2312–2322.
[16] R.A. Pitts, S. Carpentier, F. Escourbiac, T. Hirai, V. Komarov, S. Lisgo, A.S. Kukushkin, A.
Loarte, M. Merola, A. Sashala Naik, R. Mitteau, M. Sugihara, B. Bazylev, P.C. Stangeby,
A full tungsten divertor for ITER: physics issues and design status, J. Nucl. Mater. 438
(2013) S48–S56 (Supplement).
[17] S. Tatyana, H. Ahmed, Scaling mechanisms of vapour/plasma shielding from laser-
produced plasmas to magnetic fusion regimes, Nucl. Fusion 54 (2014) 023004.
[18] S. Anisimov, Vaporization of metal absorbing laser radiation, Sov. Phys. - JETP 27
(1968) 182–183.
[19] C.J. Knight, Transient vaporization from a surface into vacuum, AIAA J. 20 (1982)
950–954.
[20] C.J. Knight, Theoretical modeling of rapid surface vaporization with back pressure,
AIAA J. 17 (1979) 519–523.
[21] T. Ytrehus, Condensation and evaporation: theory and experiments on gas kinetics
in evaporation, Rarefied Gas Dynamics, Parts I and II, American Institute of Aeronau-
tics and Astronautics 1977, pp. 1197–1212.
[22] C. Cercignani, Strong Evaporation of a Polyatomic Gas, Rarefied gas dynamics, parts I
and II, American Institute of Aeronautics and Astronautics 1981, pp. 305–320.
[23] A. Miotello, A. Peterlongo, R. Kelly, Laser-pulse sputtering of aluminium: gas-dy-
namic effects with recondensation and reflection conditions at the Knudsen layer,
Nucl. Instrum. Methods Phys. Res., Sect. B 101 (1995) 148–155.
[24] A. Peterlongo, A. Miotello, R. Kelly, Laser-pulse sputtering of aluminum: vaporiza-
tion, boiling, superheating, and gas-dynamic effects, Phys. Rev. E 50 (1994)
4716–4727.
[25] A.V. Gusarov, I. Smurov, Thermal model of nanosecond pulsed laser ablation: anal-
ysis of energy and mass transfer, J. Appl. Phys. 97 (2005).
[26] Z. Chen, A. Bogaerts, Laser ablation of Cu and plume expansion into 1 atm ambient
gas, J. Appl. Phys. 97 (2005) 063305.
[27] B. Wu, Y.C. Shin, Modeling of nanosecond laser ablation with vapor plasma forma-
tion, J. Appl. Phys. 99 (2006).
[28] A. Bogaerts, Z. Chen, D. Autrique, Double pulse laser ablation and laser induced
breakdown spectroscopy: a modeling investigation, Spectrochim. Acta B At.
Spectrosc. 63 (2008) 746–754.
[29] Y. Zhou, S. Tao, B. Wu, Backward growth of plasma induced by long nanosecond
laser pulse ablation, Appl. Phys. Lett. 99 (2011) 051106.
[30] F. Rezaei, S.H. Tavassoli, Numerical and experimental investigation of laser induced
plasma spectrum of aluminum in the presence of a noble gas, Spectrochim. Acta B
At. Spectrosc. 78 (2012) 29–36.
[31] M. Stafe, Theoretical photo-thermo-hydrodynamic approach to the laser ablation of
metals, J. Appl. Phys. 112 (2012) 123112.
[32] A.H.A. Lutey, An improved model for nanosecond pulsed laser ablation of metals, J.
Appl. Phys. 114 (2013).
[33] M.M. Martynyuk, Sov. Phys. Tech. Phys. 19 (1974) 793.
[34] J.H. Yoo, O.V. Borisov, X. Mao, R.E. Russo, Existence of phase explosion during laser
ablation and its effects on inductively coupled plasma-mass spectroscopy, Anal.
Chem. 73 (2001) 2288–2293.
[35] C. Porneala, D.A. Willis, Observation of nanosecond laser-induced phase explosion in
aluminum, Appl. Phys. Lett. 89 (2006).
[36] J.H. Yoo, S.H. Jeong, R. Greif, R.E. Russo, J. Appl. Phys. 88 (2000) 1638.
[37] P. Cristian, A.W. David, Time-resolved dynamics of nanosecond laser-induced phase
explosion, J. Phys. D. Appl. Phys. 42 (2009) 155503.
[38] A.V. Pakhomov, M.S. Thompson, D.A. Gregory, J. Phys. D 36 (2003) 2067.
[39] D. Autrique, Z. Chen, V. Alexiades, A. Bogaerts, B. Rethfeld, A multiphase model for
pulsed ns-laser ablation of copper in an ambient gas, AIP Conf. Proc. 1464 (2012)
648–659.
[40] I.B. Gornushkin, U. Panne, Radiative models of laser-induced plasma and pump-
probe diagnostics relevant to laser-induced breakdown spectroscopy, Spectrochim.
Acta B At. Spectrosc. 65 (2010) 345–359.
[41] E.F. Toro, Riemann Solvers and Numerical Methods for Fluid Dynamics, Springer-
Verlag, Berlin Heidelberg, 2009.
[42] J.R. Ho, C.P. Grigoropoulos, J.A.C. Humphrey, Computational study of heat transfer
and gas dynamics in the pulsed laser evaporation of metals, J. Appl. Phys. 78
(1995) 4696–4709.
[43] R.E. Harrington, Application of the theory of heat conduction to the absorption of
blackbody radiation, J. Appl. Phys. 38 (1967) 3266–3270.
[44] Bergman, Introduction to Heat Transfer, 2011.
[45] Telle, Laser Chemistry, 2007.
[46] O. Benavides, O. Lebedeva, V. Golikov, Reflection of nanosecond Nd:YAG laser pulses
in ablation of metals, Opt. Express 19 (2011) 21842–21848.
[47] A. Bogaerts, Z. Chen, R. Gijbels, A. Vertes, Laser ablation for analytical sampling: what
can we learn from modeling? Spectrochim. Acta B At. Spectrosc. 58 (2003)
1867–1893.
[48] L.J. Rakziemski, D.A. Cremers, Laser-induced Plasmas and Applications, Marcel Dek-
ker Inc., New York, 1989.
[49] D. Salzmann, Atomic Physics in Hot Plasmas, Oxford University Press, New York,
1998.
[50] B. Verhoff, S.S. Harilal, A. Hassanein, Angular emission of ions and mass deposition
from femtosecond and nanosecond laser-produced plasmas, J. Appl. Phys. 111
(2012) 123304.
[51] N. Farid, H. Wang, C. Li, X. Wu, H.Y. Oderji, H. Ding, G.-N. Luo, Effect of background
gases at reduced pressures on the laser treated surface morphology, spectral emis-
sion and characteristics parameters of laser produced Mo plasmas, J. Nucl. Mater.
438 (2013) 183–189.
[52] M.N. Berberan-Santos, E.N. Bodunov, L. Pogliani, The van der Waals equation: ana-
lytical and approximate solutions, J. Math. Chem. 43 (2007) 1437–1457.