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

Download details:

IP Address: 37.239.59.26
This content was downloaded on 24/11/2023 at 16:19

Please note that terms and conditions apply.

You may also like:

Full Field Optical Metrology and Applications

Principles of Plasma Spectroscopy

A.L. Osterheld

Golf Balls, Boomerangs and Asteroids: The impact of missiles on society

Steve Allman

Automotive Control Systems: For Engine, Driveline, and Vehicle

U Kiencke and L Nielsen

Introduction to Optical Engineering

Chris Creasey

Nonlinear System Identification

Oliver Nelles
 IOP Publishing

Nonlinear Phenomena in the Radiation from Plasmas

Eugene Oks

Appendix A
The effects of various electric fields on the 5D,
5F, and 5G energy levels of helium

The simplest three-level scheme employed by Baranger and Mozer [1] to describe the
satellites of dipole-forbidden spectral lines under a quasimonochromatic electric
field (QEF) of frequency ω and amplitude E0 was applied by Kunze and Griem [2] to
observe the satellites of the 492.2 nm (singlet) and 447.1 nm (triplet) helium lines.
For these lines, the allowed transition is 4D–2P and the forbidden transition is 4F–
2P. In order for the QEF to sufficiently intermix the levels 4D and 4F, thus causing
the satellites, the QEF amplitude required to practically observe the satellites of
these two helium lines is relatively high: ~10 kV cm−1.
For many practical purposes, such as the spectroscopic diagnostic of the QEF at
the lower hybrid frequency in tokamaks, in which the QEF amplitude is expected to
be ~(1–4) kV cm−1, it is necessary to increase the sensitivity of the helium atoms to
the QEF. The following dimensionless parameter α controls the relative intensities of
the satellite, the total number of satellites that can be observed, and most
importantly, the sensitivity of this method:
α = 2dE 0 /(ħΔ). (A.1)
Here, d is the dipole matrix element coupling the levels, which originates the allowed
spectral line and the satellites of the forbidden spectral line and Δ is the separation
(in the angular frequency scale) between these levels. Situations for which α ≪ 1 can
be described by the perturbation theory (as used by Baranger and Mozer [1]).
However, for situations in which α is greater than or of the order of unity, the
perturbation theory becomes inapplicable.
In order to increase the sensitivity parameter α at a fixed QEF amplitude E0, the
only way is to proceed to spectral lines originating from levels of the principal
quantum number n > 4. Indeed, α is proportional to the ratio d/Δ. As n increases, d
also increases (proportionally to n2), while the separation Δ between closely lying
levels decreases.

doi:10.1088/978-0-7503-5552-0ch9 A-1 ª IOP Publishing Ltd 2023

 Nonlinear Phenomena in the Radiation from Plasmas

Selecting prospective spectral lines of helium involves finding an optimal balance

between several competing, counter-acting factors, such as the sensitivity to the
anticipated QEF, the required spectral resolution, and sufficient absolute intensity.
Indeed, on the one hand, the higher the principal quantum number n of the upper
level, the better the sensitivity with respect to the electric fields. On the other hand,
the higher the value of n, the smaller the separation between the forbidden and
allowed components (or between satellites and the allowed component) of
the corresponding spectral lines, thus requiring a higher spectral resolution.
Also, as n increases, the absolute intensity of the spectral lines decreases. It should
be also emphasized that a spectral line of helium, sensitive enough to the QEF ~
(1–4) kV cm−1, should be chosen such that the effect of the QEF is strong enough
that it cannot be described by the perturbation theory.
The three-level scheme theoretically employed by Baranger and Mozer [1] (and
later by Cooper and Ringler [3]) was relatively simple because the QEF was
supposed to intermix only two levels: 4D and 4F. (We recall that there are no levels
4G, 4H. etc.) Since enhancing the sensitivity requires proceeding to helium spectral
lines originating from levels of n > 4, this means that the QEF will couple more than
two levels. For example, if one were to use helium spectral lines in the vicinity of the
allowed transition 5D–2P, then there would be two forbidden transitions: 5F–2P and
5G–2P. In other words, in this case, more sophisticated analytical approaches—
going beyond the perturbation theory—should be employed for the four-level
scheme whose three upper levels are 5D, 5F, and 5G.
The need may also arise to use helium spectral lines in the vicinity of the allowed
transition 6D–2P. There would then be three forbidden transitions: 6F–2P, 6G–2P,
and 6H–2P. In other words, in this case, the more sophisticated analytical
approaches (beyond the perturbation theory) should be implemented for the five-
level scheme whose four upper levels are 6D, 6F, 6G, and 6H.
Specifically, the two following analytical theories presented in [4] are relevant for
this purpose. Both theories consider satellites of dipole-forbidden spectral lines
under a linearly-polarized electric field E(t) = E0 cos(ωt). Their regions of validity,
while differing, actually complement each other. For exactness, we provide below
details for helium spectral lines corresponding to the radiative transitions to the
lower level 2P from the upper levels (or sublevels) nD, nF, … . The total number of
upper sublevels is n − 2, where n is the principal quantum number of the upper
sublevels. Atomic units are used in the subsequent formulas.
The first theory is called the adiabatic theory of satellites. Its region of validity can
be expressed as follows:
(zE 0 / Δ)(ω / Δ) ≪ 1. (A.2)
Here, z is the average of the matrix elements of the z-coordinate (chosen along the
oscillatory electric field) between sublevels nD and nF, nF, and nG, etc.; Δ is the
separation between the highest and the lowest of the sublevels nD, nF, … . For
example, the adiabatic theory of satellites allows situations in which the amplitude
E0 is relatively high (so that the standard perturbation theory fails) to be treated
analytically as long as the frequency ω is relatively low.

A-2
 Nonlinear Phenomena in the Radiation from Plasmas

The second theory is called the high-frequency or strong-field theory of satellites.

As follows from its name, it is valid where either the amplitude E0 or the frequency ω
is sufficiently large. Its region of validity can be expressed as follows:
(zE 0 / Δ)2 + (ω / Δ)2 ≫ 1. (A.3)
Figure A.1 schematically illustrates the regions of validity of the adiabatic theory of
satellites and the high-frequency or strong-field theory of satellites. The former
theory is valid below the solid curve, while the latter theory is valid above the dashed
curve.
We now specifically consider the situation in which the three upper energy levels
are 5D, 5F, and 5G. In figure A.2, these levels are denoted by numbers from one to
three for the brevity of the subsequent equations and the analytical results.
In the following, we consider the subset of the energy levels corresponding to the
zero projection of the angular momentum in the direction of the electric field: m = 0.
They are coupled by the matrix elements of the z-coordinate, where (in atomic units)
z12 = 11.34, z13 = 15.21. (A.4)

Figure A.1. Regions of validity of the adiabatic theory of satellites and the high-frequency or strong-field
theory of satellites. The former theory is valid below the solid curve, while the latter theory is valid above the
dashed curve.

Figure A.2. The situation in which the three upper energy levels of helium in the four-level scheme are 5D, 5F,
and 5G.

A-3
 Nonlinear Phenomena in the Radiation from Plasmas

The energy differences in atomic units are the same as the frequency differences,
which are as follows:

ω21 = 2.237 × 10−6 , ω31 = −1.294 × 10−5,

(A.5)
ω23 = ω21 − ω31 = 1.518 × 10−5.
We introduce the following notations:
1/2
α ≡ B sin ωt , B ≡ ξE 0 / ω, ξ ≡ (z122 + z132) ,
z˜12 ≡ z12 / ξ, z˜13 ≡ z13/ ξ, (A.6)
V≡ ω21z˜122 + ω31z˜132, W= ω21z˜132 + ω31z˜122 .

The primary parameter controlling the quasienergies is the quantity B from equation
(A.6). The practical formula for the parameter B is:
B = 24.28E 0 / ω, (A.7)
where the field amplitude E0 is in kV cm−1 and the field frequency ω is in GHz.
Using the Fourier expansions
∞
sin α = 2 ∑ J2k−1(B )sin(2k − 1)ωt , cos α = J0(B )
k=1
∞ (A.8)
+ 2 ∑ J2k(B )sin 2kωt ,
k=1

where Jq(B) are the Bessel functions, the quasienergies μ1, μ2, and μ3 can be obtained
in the following form [4]:
2 1/2
μ1 = a11, μ2,3 = (a22 + a33)/2 ± [(a22 − a33)2 /4 + a 23 ] . (A.9)

In equation (A.9), the quantities a11, a22, a33, and a23 are as follows:
a11 = [1 − J0(2B )]V /2;
a23 = a32 = z˜12z˜13{(z˜132 − z˜122 )ω23[J0(B ) − 1] + [J0(2B ) − 1]V /2} ;
(A.10)
app = ωp1 + ( −1) p2z˜122z˜132ω23[J0(B ) − 1]
+ z˜12p[J0(2B ) − 1]V /2, (p = 2,3).

ED denotes the scaled dimensionless energy density of the oscillatory field:

ED = B 2. (A.11)
Figure A.3 presents the quasienergies of the energy levels from figure A.2 versus the
scaled energy density ED of the oscillatory field.
From figure A.3, one can see that the dependence of the quasienergies on the
scaled energy density of the oscillatory field is highly nonlinear.

A-4
 Nonlinear Phenomena in the Radiation from Plasmas

Figure A.3. Quasienergies of the energy levels from figure A.2 versus the scaled energy density ED of the
oscillatory field.

Figure A.4. The same as figure A.2 but with a different numeration of the energy levels.

For comparison with the above situation, we consider below the same system
under a uniform electric field F. We slightly change the numeration of these energy
levels, as shown in figure A.4.
The energy is counted from level one. Its eigenvalues are denoted by w. We
introduce the following notations:
e = −E2 > 0, u = w/e, a = z12F / e , b = z13F / e , c = E3/ e < 1. (A.12)
The eigenvalues are the roots of the cubic equation:
u 3 + (1 − c )u 2 − (a 2 + b 2 + c )u − (b 2 − ca 2 ) = 0. (A.13)
The strong-field approximation corresponds to the case in which
(a 2 + b 2 )1/2 ≫ 1. (A.14)
Under this condition, the roots of the cubic equation (A.13) are as follows:
u1 = −(b 2 − ca 2 )/(a 2 + b 2 ), (A.15)

u2 = (a 2 + b 2 )1/2 − d , (A.16)

A-5
 Nonlinear Phenomena in the Radiation from Plasmas

u3 = −(a 2 + b 2 )1/2 − d , (A.17)

where
d = (1 − c − u1)/2. (A.18)
It should be noted that in the above results, obtained in the strong-field approx-
imation, the quantities u1 and d do not depend on the field (they are constants), while
u2 and u3 linearly depend on the field. For comparison, figure A.5 presents the exact
results for the energies versus the scaled field b for a = (9/5)1/2b and c = 1/8 (as is the
case for the levels from figure A.4). It can be seen that as the scaled field b increases,
the exact results approach the above asymptotic results obtained in the strong-field
approximation.
It is instructive to represent the exact results for the energies versus b2, that is,
versus the scaled energy density of the field. This is shown in figure A.6. It can seen

Figure A.5. The exact results for the energies versus the scaled field b for a = (9/5)1/2b and c = 1/8 (as is the case
for the levels from figure A.4).

Figure A.6. The same results as those of figure A.5, but versus b2, that is, versus the scaled energy density of the
field.

A-6
 Nonlinear Phenomena in the Radiation from Plasmas

that the dependence of the energies is not as highly nonlinear as in the case of the
quasienergies presented in figure A.3.

Bibliography
[1] Baranger M and Mozer B 1961 Phys. Rev. 123 25
[2] Kunze H J and Griem H R 1968 Phys. Rev. Lett. 21 1048
[3] Cooper W S and Ringler H 1969 Phys. Rev. 179 226
[4] Oks E 1995 Plasma Spectroscopy: The Influence of Microwave and Laser Fields (Berlin:
Springer)

A-7
 IOP Publishing

Nonlinear Phenomena in the Radiation from Plasmas

Eugene Oks

Appendix B
Examples of the use of spectral line radiation to
perform diagnostics of oscillatory electric fields
in laser-produced plasmas

The authors of [1] were the first to use the spectral line radiation (specifically, in the
x-ray range) to perform diagnostics of the oscillatory electric fields in laser-produced
plasmas. The experiment was conducted at the nanosecond Nd:glass laser facility at
LULI (France). The laser intensity at the target was relatively moderate: 2 × 1014 W
cm−2. The target was structured: it consisted of a (20–60) μm thick powdered
aluminum carbide Al4C3 strip placed between substrates made of plastic or
magnesium. The plasma parameters (estimated by hydrodynamic simulations) at
the surface of the target were an electron density of Ne ~ 5 × 1022 cm−3 and a
temperature of T ~ 300 eV.
The Vertical Johann Spectrometer produced two sets of spatially resolved spectra
simultaneously. The two sets were symmetrically located in relation to the
unperturbed wavelength of the spectral line: the right set and the left set. This
feature very significantly enhanced the possibility of identifying tiny details
(Langmuir ‘dips’) within the experimental line profiles.
Figure B.1 shows the experimental profile of the Lyman-gamma line of Al XIII.
In this figure, the right set and the left set are superimposed. This demonstrates the
reproducibility of the Langmuir dip position (marked b2) in the blue part of the
experimental profile. The Langmuir-wave-induced ‘bump-dip-bump’ structures,
being imposed on the inclined unperturbed profile, can create a secondary minimum
(in addition to the primary minimum, whose location is controlled by the electron
density) of no practical importance. The secondary minimum is marked by the
arrow in the blue part of the experimental profile. The structures marked in the red
part of the experimental profiles are the charge-exchange-induced dips—irrelevant
to the phenomenon of Langmuir-wave-induced dips.

doi:10.1088/978-0-7503-5552-0ch10 B-1 ª IOP Publishing Ltd 2023

 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.1. The experimental profile of the Lyman-gamma line of Al XIII. In this figure, the right set and the
left set are superimposed. This demonstrates the reproducibility of the Langmuir dip position (marked b2) in
the blue part of the experimental profile—see the text above the figure for more details. Reprinted from [1],
Copyright (2006), with permission from Elsevier.

From the experimental position of the Langmuir dip, the electron density was
determined to be 2.2 × 1022 cm−3. From the width of the experimental Langmuir
dip, the amplitude of the Langmuir wave was estimated to be 0.1 GV cm−1.
The authors of [2] analyzed the experimental profiles of the Al XII beta line
emitted from the aluminum plasma. The experiment was conducted at the Friedrich
Schiller University Jena laser system, JETi.
The plasma was produced by a picosecond laser beam and then exposed to the
beam of another picosecond laser. The second laser was turned on after the pulse of
the first laser had terminated.
The first laser beam had an intensity of 5 × 1015 W cm−2. The second laser beam
had an intensity of 1.2 × 1016 W cm−2. Both beams overlapped in space and were
synchronized in time with a precision exceeding 1 ps.
Figure B.2 demonstrates the reproducibility of the structures in the experimental
profiles of the Al XII beta line.
Simulations employing the Floquet–Liouville formalism [3] were conducted for
electron densities of 2.5 × 1020 cm−3 ⩽ Ne ⩽ 1.5 × 1021 cm−3 at a temperature of
Te = 150 eV for a variety of field strengths of the quasimonochromatic electric field:
E = 0, 0.1, 0.3, and 0.6 in units of E0 = E0 = 5.146 GV cm−1, the latter being the
atomic unit of the electric field. The results are presented in figure B.3 (for the
‘parallel’ profiles, i.e. those polarized along the oscillatory electric field) and in

B-2
 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.2. A demonstration of the reproducibility of the structures in the experimental profiles of the Al XII
beta line. Reproduced from [2]. © IOP Publishing Ltd. All rights reserved.

figure B.4 (for the ‘perpendicular’ profiles, i.e. those polarized perpendicular to the
oscillatory electric field).
Figure B.5 displays the outcome of kinetic PIC simulations of the distribution (in
space) of the transverse averaged oscillatory electric field inside the plasma for
Ne = 0.1 Nc, Ne = 0.2 Nc, and Ne = 0.3 Nc. Here, Nc = 1.7 × 1021 cm−3 is the critical
density. The nonlinear enhancement of the incident laser field occurs at the surface
of the critical density.
Figure B.6 shows a comparison of the experimental profiles with the profiles
simulated for the following four values of the initial electron density: 0.05 Nc, 0.1 Nc,
0.2 Nc, and 0.35 Nc.
Figure B.7 presents a comparison between the experimental profile and a profile
simulated for an average electron density of N¯e = 0.2 Nc = 3.3 × 1020 cm−3 and a time
interval ΔTunpert = 8.9 ps. The comparison shows good agreement in the central part
of the profile as well as in the first two satellites in the red and blue parts of the

B-3
 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.3. Simulated profiles of the Al XII beta line at Te = 150 eV for the indicated range of electron density
and the indicated amplitude range E of the oscillatory electric field, the latter being in units of E0 = 5.146 GV
cm−1. The profiles are polarized along the oscillatory electric field. Reproduced from [2]. © IOP Publishing
Ltd. All rights reserved.

profile. More distant satellites are in the far wings: there, the experimental profile
blends with the noise. An additional maximum marked by the arrow corresponds to
the dipole forbidden transition 1s3s 1S0 — 1s2 1S0 caused by the quadrupole
interaction with the plasma ion field.
The authors of [4] analyzed the experimental hydrogenic spectral lines of Si XIV
and of Al XIII produced during the interaction between relativistic laser radiation
(at an intensity of ~1021 W cm−2) and thin foils of silicon. As a result, the authors of
[4] revealed the development of ‘ionic sound’ (i.e. the ion acoustic turbulence). This
marked the discovery of ionic sound in in spectroscopic studies of laser-produced
plasmas. Ionic sound developed at the surface of the critical density due to the
parametric decay instability in the plasma. This experiment was conducted at the
Vulcan petawatt laser facility located at the Rutherford Appleton Laboratory [5].
Parametric decay instability is a nonlinear process in plasmas. In this process, the
electromagnetic wave is converted into an ionic sound wave and a Langmuir wave,
both of which represent electrostatic plasma turbulence. The electrostatic plasma
turbulence corresponds to the collective degrees of freedom in those plasmas.
The ion acoustic turbulence strongly influences transport phenomena in variety of
plasmas. In particular, the resistivity of the plasma becomes anomalous: the primary
cause of resistivity becomes the scattering of the conductivity electrons on the ion

B-4
 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.4. The same profiles as those of figure B.3 but polarized perpendicular to the oscillatory electric field.
Reproduced from [2]. © IOP Publishing Ltd. All rights reserved.

acoustic wave—rather than the scattering on individual plasma ions (the latter being
the cause of the ‘normal’ resistivity of plasmas).
Figure B.8 displays the experimental setup and the experimental profiles of the
Ly-beta and Ly-gamma lines of Si XIV. In the insets, the locations of the Langmuir
dips are indicated by vertical line segments. The distance between the experimental
Langmuir dips is either 2λpe or 4λpe, where
λ pe = [λ 02 /(2πc )]ωpe . (B.1)

In equation (B.1), λ0 is the unperturbed wavelength of the corresponding line and

ωpe is the plasma electron frequency:
ωpe = (4πe 2Ne / m e )1/2 . (B.2)

Figure B.9 presents a comparison between the experimental and theoretical

profiles of the Ly-beta and Ly-gamma lines of Si XIV. The theoretical profiles were
calculated by allowing both for the effect of the ion acoustic turbulence and for the
Langmuir dips.
Figure B.10 displays the experimental profile of the Ly-beta line of Al XIII
obtained when the spectrometer was directed at the front surface of the laser-
irradiated target—in contrast to the experimental profiles in figures B.8 and B.9,

B-5
 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.5. The outcome of the kinetic PIC simulations of the distribution (in space) of the transverse
averaged oscillatory electric field inside the plasma: (a) Ne = 0.1Nc; (b) Ne = 0.2Nc; (c) Ne = 0.3Nc. Here,
Nc = 1.7 × 1021 cm−3 is the critical density. Reproduced from [2]. © IOP Publishing Ltd. All rights reserved.

which were obtained by directing the spectrometer at the back surface of the laser-
irradiated target. The target was aluminum foil coated by a layer of CH. The
corresponding theoretical profile is also presented. The experimental spectrum
exhibits two pairs of Langmuir dips: one pair at a distance of ±33.6 mÅ from the
line center and the second pair at a distance of ±16.8 mÅ from the line center.
The authors of [6] presented an in-depth spectroscopic study of the simultaneous
production of Langmuir waves and the ionic sound at the surface of the relativistic
critical density. This experiment was also conducted at the Vulcan petawatt laser
facility located at the Rutherford Appleton Laboratory [5]. The focus of the paper
was the Langmuir dips in the experimental profiles of the Ly-beta line of Si XIV. The
laser intensity at the target ranged from 1 to 3 × 1020 W cm−2 in different shots.

B-6
 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.6. A comparison between the experimental and simulated profiles for the following four values of the
initial electron density: 0.05 Nc, 0.1 Nc, 0.2 Nc, and 0.35 Nc. Reproduced from [2]. © IOP Publishing Ltd. All
rights reserved.

Figure B.7. A comparison between the experimental profile and a profile simulated for an average electron
density of N¯e = 0.2 Nc = 3.3 × 1020 cm−3 and a time interval of ΔTunpert = 8.9 ps. Reproduced from [2]. © IOP
Publishing Ltd. All rights reserved.

B-7
 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.8. (a) The experimental setup and (b) the experimental profiles of the Ly-beta and Ly-gamma lines of
Si XIV. In the insets, the locations of the Langmuir dips are indicated by the vertical line segments.
Reproduced from [4]. © The Optical Society.

Figure B.9. A comparison between the experimental and theoretical profiles of the Ly-beta and Ly-gamma
lines of Si XIV. The theoretical profiles were calculated by allowing both for the effect of the ion acoustic
turbulence and for the Langmuir dips. The theoretical profiles are calculated for the electron density Ne = 1.74
× 1022 cm−3. Reproduced from [4]. © The Optical Society.

B-8
 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.10. The experimental profile of the Ly-beta line of Al XIII obtained by directing the spectrometer at
the front surface of the laser-irradiated target—in contrast to the experimental profiles in figures B.8 and B.9,
which were obtained by directing the spectrometer at the back surface of the laser-irradiated target. The target
was aluminum foil coated by a layer of CH. The corresponding theoretical profile is also presented.
Reproduced from [4]. © The Optical Society.

Figure B.11. Experimental profiles of the Si XIV Ly-beta line obtained in three shots. Also shown are the
corresponding profiles simulated by the FLYCHK code. The best fit produced by this code yielded an
unrealistically high electron density of Ne = 6 × 1023 cm−3. Reproduced from [6]. © IOP Publishing Ltd. All
rights reserved.

Figure B.11 displays experimental profiles obtained in three different shots. There
are very distinct ‘bump-dip-bump’ structures in the experimental profiles. Actually,
these are Langmuir superdips, that is, the superposition of two Langmuir dips at the
same location in the profile. Also shown are theoretical profiles calculated using the
FLYCHK code, which does not take into account the phenomenon of the Langmuir
dips and the broadening due to ion acoustic turbulence. The best fit by this code
yielded an unrealistically high electron density of Ne = 6 × 1023 cm−3. This was one
of the indications that ion acoustic turbulence had developed and that it has to
be taken into account.

B-9
 Nonlinear Phenomena in the Radiation from Plasmas

Figure B.12 shows the experimental profiles obtained in the shots marked A, B,
and C. Also shown are the corresponding theoretical profiles calculated by more
advanced code that took account of the phenomenon of the Langmuir dips and the
broadening due to ion acoustic turbulence—in addition to all other broadening
mechanisms. The deduced electron density was Ne = 2.2 × 1022 cm−3. The root-mean-
square strengths of the electric field of the ion acoustic turbulence were deduced to be
4.8 GV cm−1, 4.4 GV cm−1, and 4.9 GV cm−1 for shots A, B, and C, respectively.
Shot D was obtained at a significantly lower laser intensity at the target than shots
A, B, and C. Therefore, in shot D the electron density was significantly smaller than
in shots A, B, and C. The consequence was a significantly smaller damping of the
Langmuir waves, so that they attained a considerably greater amplitude.
The deduced parameters for shot D were an electron density of Ne = 6.6 × 1021
cm−3, a temperature of T = 550 eV, a root-mean-square electric field of the ion
acoustic turbulence of Ft,rms = 2.0 GV cm−1, and a Langmuir wave amplitude of

Figure B.12. Experimental profiles of the Si XIV Ly-beta line obtained in four different shots and their
comparison with the profiles calculated by more advanced code that took into account the phenomenon of the
Langmuir dips and the broadening due to ion acoustic turbulence (for shots A, B, and C)—in addition to all
other broadening mechanisms. The deduced electron density was Ne = 2.2 × 1022 cm−3. The root-mean-square
strengths of the electric field of the ion acoustic turbulence were deduced to be 4.8 GV cm−1, 4.4 GV cm−1, and
4.9 GV cm−1 for shots A, B, and C, respectively. Reproduced from [6]. © IOP Publishing Ltd. All rights
reserved.

B-10
 Nonlinear Phenomena in the Radiation from Plasmas

E0 = 2.0 GV cm−1. The theoretical and experimental profiles in shot D do not

exhibit bump-dip-bump structures. This is because in the situation in which
E0/Ft,rms > 0.5, Langmuir dips cannot form, as explained in [7].

Bibliography
[1] Renner O, Dalimier E, Oks E, Krasniqi F, Dufour E, Schott R and Förster E 2006 J. Quant.
Spectrosc. Radiat. Transfer 99 439
[2] Sauvan P, Dalimier E, Oks E, Renner O, Weber S and Riconda C 2009 J. Phys. B: At. Mol.
Opt. Phys. 42 195001
[3] Sauvan P and Dalimier E 2009 Phys. Rev. E 79 036405
[4] Oks E et al 2017 Opt. Express 25 1958
[5] Danson C N et al 2005 Laser Part. Beams 23 87
[6] Oks E et al 2017 J. Phys. B: At. Mol. Opt. Phys. 50 245006
[7] Oks E 1995 Plasma Spectroscopy: The Influence of Microwave and Laser Fields, (Springer
Series on Atoms and Plasmas) 9 (New York: Springer)

B-11
 IOP Publishing

Nonlinear Phenomena in the Radiation from Plasmas

Eugene Oks

Appendix C
The validity of using the analytical method
based on separating rapid and slow subsystems
for a more accurate analysis of the Stark
broadening of hydrogenlike spectral lines
by plasma electrons

The frequency of the spectral line is the frequency of the revolution of the atomic
electron around the nucleus:
Z 2UH ⎛ 1 1
Ω= ⎜ − 2 ⎞⎟ , (C.1)
ℏ ⎝ n b2 na ⎠

where UH is the ionization potential of hydrogen.

The typical frequency characterizing the motion of the perturbing electron
around the radiating ion, leading to the Stark broadening of spectral lines, is the
Weisskopf frequency (see, e.g. [1]):
vT Z m vT2 Z T
ω We = ∼ 2 2
∼ 2 2
. (C.2)
ρ We (n a − n b )ℏ (n a − n b )ℏ
Consequently, the ratio of the above frequencies is given by

ω We T ⎞⎡ n a2n b2 ⎤
∼ ⎛ ⎜ ⎟ . (C.3)
⎢ 2 2 2⎥
Ω ⎝ Z UH ⎠⎣ (n a − n b ) ⎦
In the situation in which na ≫ nb, this ratio simplifies as follows:

ω We ⎛ T ⎞⎟ ≪ 1
∼⎜ 2
(C.4)
Ω ⎝ Z n a UH ⎠

doi:10.1088/978-0-7503-5552-0ch11 C-1 ª IOP Publishing Ltd 2023

 Nonlinear Phenomena in the Radiation from Plasmas

if
T (eV) ≪ (13.6eV)Zn a2 . (C.5)

Thus, under condition (C.5), it is legitimate to apply the method of separating rapid
and slow subsystems to the problem under consideration.
For instance, for the case of ionized helium (Z = 2), the above validity condition
(C.5) becomes
T (eV) ≪ (27.2eV) n a2 . (C.6)

The inequality (C.6) is met for a sufficiently large range of temperatures at which the
spectral line radiation of ionized helium is observed in various plasmas.

Bibliography
[1] Oks E 2006 Stark Broadening of Hydrogen and Hydrogenlike Spectral Lines in Plasmas: The
Physical Insight (Oxford: Alpha Science International)

C-2
 IOP Publishing

Nonlinear Phenomena in the Radiation from Plasmas

Eugene Oks

Appendix D
A brief outline of the impact approximation in
the conventional theory of the Stark broadening
of spectral lines in plasmas

For the dynamical Stark broadening of spectral lines by plasma electrons to be

efficient, it is required that the formally calculated number νWe of perturbing
electrons located within the sphere of the electron Weisskopf radius ρWe should be
much smaller than unity (see, e.g. [1]):
νWe = 4πNe ρ We3 /3 ≪ 1. (D.1)

In equation (D.1), ρWe ~ n2 ħ/(mevTe ), where n is the principal quantum number and
vTe is the mean thermal velocity of the perturbing electrons in a plasma. When
inequality (D.1) is satisfied, the instantaneous Stark splitting in the electron micro-
field is much smaller than the typical frequency of the variation in the electron
microfield Ωe ~ vTe /ρWe (for the overwhelming majority of perturbing electrons).
Physically, the electron Weisskopf radius is associated with values of the impact
parameters ρ ~ ρWe, which provide the primary contribution to the electron-induced
dynamical Stark broadening of spectral lines in plasmas [2, 3].
The essence of the electron-induced dynamical Stark broadening is as follows:
virtual transitions within the lower (n′) and upper (n) multiplets occur in the process
of the radiative transition n ↔ n′. Their principal effect is to shorten the lifetimes of
the states n′ and/or n; this causes the broadening of the line.
The assumption that virtual transitions take place mostly within the upper and
lower multiplets is called the no-quenching approximation. Within this approxima-
tion, virtual transitions between states that differ by the principal quantum numbers
are not taken into account. This approximation makes it possible to use the concept
of the line space. The latter is the direct product of the Hilbert space spanned on the
basis vectors of the n-subspace with the Hilbert space spanned on the (complex-
-conjugated) basis vectors of the n′-subspace.

doi:10.1088/978-0-7503-5552-0ch12 D-1 ª IOP Publishing Ltd 2023

 Nonlinear Phenomena in the Radiation from Plasmas

Not only the impact formalism (introduced in [2] and further developed in [4]), but
also the primary properties of the unified theory of the electron-induced dynamical
Stark broadening [5] can stem from the same succession of mathematical operations,
as shown in [6]. The main difference between these two formalisms is as follows.
The impact formalism treats all collisions as being completed. In contrast, the
unified theory of the dynamical Stark broadening takes into account the fact that
that not all collisions are completed. Another difference between the two formalisms
is that within the unified theory, the calculated Stark profile of the spectra line
contains a transition to the quasistatic shape in the wings; however, the unified
formalism does not always produce this transition in the correct way.
The crucial feature of the impact formalism is that it introduces a coarse-grained
timescale Δt defined through the following condition:
−1
ρ / ve ≪ Δt ≪ ⎡ max(γ , Δω, ωpe )⎤ . (D.2)
⎣ ⎦
In equation (D.2), ωpe = (4πNee2/me)1/2 is the plasma electron frequency, Δω is the
detuning from the unperturbed frequency of the spectral line, and γ is the inverse of the
lifetime of the radiator (the latter being of the order of the impact Stark width of the line).
The physical meaning of this coarse-grained timescale is as follows. The focal
point is the radiator evolution in time during the intervals ~Δt defined by equation
(D.2), while the details of the time evolution during the intervals ~ρ/ve (i.e. during
intervals of the order of the typical time of the individual collision) are of no interest.
The validity limits of the impact formalism are controlled by whether or not it is
possible to introduce the coarse-grained timescale defined by the inequality (D.2). It
is worth mentioning that in the unified theory, in the left part of the condition (D.2),
the strong inequality ρ/ve ≪ Δt is lessened to ρ/ve ~ Δt.
The operator of the dynamical electron broadening Φab is defined as follows:
Φab = [Ua(t , t + Δt )Ub*(t , t + Δt ) − 1]/ Δt . (D.3)
In equation (D.3), Ua and Ub are the operators of the evolution in time for the upper
and lower subspaces in the line space, respectively. If Δt ≫ ρ/ve (i.e. under the impact
approximation), these operators are replaced by the respective scattering matrices Sa
and Sb. The utilization of the two scattering matrices (rather than one) is associated
with the concept of the line space.
A final note: for any hydrogenic spectral line of any spectral series, dynamical
Stark broadening takes place in the overlapping situation—that is, there are overlaps
of the Stark components of the line. This is the case even for the Lyman-alpha line
(the simplest line) of hydrogenic atoms/ions—see [7].
Further details on the impact approximation and its comparison with the unified
theory can be found in [8–11]. Also, [6] is quite illuminating on this issue.

Bibliography
[1] Lisitsa V S 1977 Sov. Phys. Usp. 122 603
[2] Baranger M 1958 Phys. Rev. 111 481
[3] Baranger M 1958 Phys. Rev. 112 1958 855

D-2
 Nonlinear Phenomena in the Radiation from Plasmas

[4] Kolb A C and Griem H R 1958 Phys. Rev. 111 514

[5] Vidal C R, Cooper J and Smith E W 1970 J. Quant. Spectrosc. Radiat. Transfer 10 1011
Vidal C R, Cooper J and Smith E W 1970 J. Quant. Spectrosc. Radiat. Transfer 11 263
[6] Sahal-Brechot S 1969 Astron. Astrophys. 1 91
[7] Strekalov M L and Burshtein A I 1972 Sov. Phys. JETP 34 53 http://jetp.ras.ru/cgi-bin/e/
index/e/34/1/p53?a=list
[8] Griem H R 1964 Plasma Spectroscopy (New York: McGraw-Hill)
[9] Griem H R 1974 Spectral Line Broadening by Plasmas (Cambridge, MA: Academic) https://
shop.elsevier.com/books/spectral-line-broadening-by-plasmas/griem/978-0-12-302850-1
[10] Oks E 2006 Stark Broadening of Hydrogen and Hydrogenlike Spectral Lines in Plasmas: The
Physical Insight (Oxford: Alpha Science International)
[11] Oks E 2017 Diagnostics of Laboratory and Astrophysical Plasmas Using Spectral Lines of
One-, Two-, and Three-Electron Systems (Hackensack, NJ: World Scientific)

D-3