Ion acoustic solitary structures in a magnetized nonthermal dusty plasma

Debdatta Debnath, Anup Bandyopadhyay, and K. P. Das

Citation: Physics of Plasmas 25, 033704 (2018); doi: 10.1063/1.5021127

View online: https://doi.org/10.1063/1.5021127
View Table of Contents: http://aip.scitation.org/toc/php/25/3
Published by the American Institute of Physics

Articles you may be interested in

Magnetosonic cnoidal waves and solitons in a magnetized dusty plasma
Physics of Plasmas 25, 043704 (2018); 10.1063/1.5011948

Coupled ion acoustic and drift solitons in a magnetized bi-ion plasma with pseudo-potential approach
Physics of Plasmas 25, 022113 (2018); 10.1063/1.5018228

Electron-acoustic solitary waves in the Earth's inner magnetosphere

Physics of Plasmas 25, 022905 (2018); 10.1063/1.5007907

Effect of polarization force on head-on collision between multi-solitons in dusty plasma

Physics of Plasmas 25, 033705 (2018); 10.1063/1.5020194

New method for rekindling the nonlinear solitary waves in Maxwellian complex space plasma
Physics of Plasmas 25, 043703 (2018); 10.1063/1.5011317

A study of the non-Maxwellian pair-ion and pair-ion-electron plasmas

Physics of Plasmas 25, 022105 (2018); 10.1063/1.5000572
 PHYSICS OF PLASMAS 25, 033704 (2018)

Ion acoustic solitary structures in a magnetized nonthermal dusty plasma

Debdatta Debnath,1 Anup Bandyopadhyay,1,a) and K. P. Das2
1
Department of Mathematics, Jadavpur University, Kolkata 700032, India
2
Department of Applied Mathematics, University of Calcutta, 92 A. P. C. Road, Kolkata 700009, India
(Received 1 January 2018; accepted 1 March 2018; published online 20 March 2018)
The Sagdeev potential technique has been used to investigate the arbitrary amplitude ion acoustic
solitary structures in a collisionless magnetized dusty plasma consisting of negatively charged
static dust grains, adiabatic warm ions and nonthermal electrons. The present system supports both
positive and negative potential solitary waves, the coexistence of solitary waves of both polarities,
and negative potential double layers. The system does not support any positive potential double
layer. Although the system supports negative potential double layers, these double layer solutions
cannot restrict the occurrence of all solitary structures of same polarity. In fact, there exists a
parameter regime for which the negative potential double layer is unable to restrict the occurrence
of negative potential solitary waves, and in this region of the parameter space, there exist negative
potential solitary waves after the formation of a negative potential double layer. Consequently,
negative potential supersolitons have been observed and the Mach number M corresponding to a
negative potential supersoliton is restricted by the inequality MNPDL < M < Mcr, but this
supersoliton structure reduces to a conventional solitary wave of same polarity if M  Mcr, where
MNPDL is the Mach number corresponding to a negative potential double layer and Mcr is a critical
value of M. Thus, we have seen a transition process of negative potential solitary structures, viz.,
soliton ! double layer ! supersoliton ! soliton. Different solitary structures have been
investigated with the help of compositional parameter spaces and the phase portraits of the
dynamical system describing the nonlinear behaviour of ion acoustic waves. The mechanism of
transition of a negative potential supersoliton to a conventional soliton after the formation of a
double layer of same polarity has been discussed with the help of phase portraits. Published by AIP
Publishing. https://doi.org/10.1063/1.5021127

I. INTRODUCTION dust particulates. In the present paper, we have investigated

IA solitary structures in a collisionless magnetized dusty
Magnetized dusty plasma naturally occurs in a number
plasma system consisting of warm adiabatic ions, static neg-
of astrophysical environments such as the planetary rings,
atively charged dust grains and nonthermal electrons,
asteroid zones, comets, the interstellar medium, Earth’s iono-
immersed in a uniform static magnetic field directed along a
sphere and Earth’s magnetosphere.1–8 Acoustic wave modes
fixed direction.
in an unmagnetized or a magnetized dusty plasma have
Yinhua and Yu10 have investigated ion acoustic (IA)
received a great deal of attention since the last few decades.
solitary waves propagating obliquely into an external uni-
There can exist two or more acoustic waves in a typical
form static magnetic field in the presence of heavy dust par-
dusty plasma depending on different time scales. Dust acous-
tic3 (DA) and dust ion acoustic9 (DIA) waves are two such ticulates. They have found the existence of small amplitude
acoustic waves in a collisionless unmagnetized plasma con- IA solitary waves. Choi et al.7 have studied IA solitary
taining electrons, ions, and charged dust grains. For the first waves obliquely propagating into an external magnetic field
time, Shukla and Silin9 reported that a collisionless unmag- in a collisionless magnetized dusty plasma. Later, Choi
netized dusty plasma supports low-frequency DIA waves et al.11 have studied the effect of ion temperature on the IA
with the phase velocity much smaller (larger) than the elec- solitary waves by considering the isothermal ion pressure.
tron (ion) thermal velocity. For long wavelength plane wave Maitra and Roychoudhury12 have studied the IA solitary
perturbation, the dispersion relation of DIA waves is similar waves in a collisionless magnetized dusty plasma, but they
to that of ion acoustic (IA) waves for a plasma with Ti  Te have used the Poisson equation instead of taking the quasi-
and ne0/ni0 is of moderate magnitude for the dusty plasma neutrality condition. In the above-mentioned papers, the
system, where Ti (Te) is the average ion (electron) tempera- authors have used Sagdeev potential formalism. But, finally,
ture and ne0 (ni0) is the unperturbed number density of elec- they have expanded the Sagdeev potential in a Taylor series
trons (ions). For the case of usual two-component electron— up to a finite order of an appropriate dependent variable to
ion plasma, the above two conditions are modified as fol- investigate small but finite amplitude solitary structures.
lows: Ti  Te and ne0 ¼ ni0. Thus, IA waves in dusty plasmas Several authors13–30 have studied DIA/IA solitary structures
are basically IA waves modified by the presence of heavy in different unmagnetized/magnetized dusty plasmas. Some
authors19,22,26,28–30 considered the nonthermal electrons in
a)
Electronic mail: [email protected] different unmagnetized or magnetized dusty plasmas.

1070-664X/2018/25(3)/033704/13/$30.00 25, 033704-1 Published by AIP Publishing.

033704-2 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

Nonthermally distributed energetic particles in dusty plas- can produce more energetic particles in comparison with
mas are observed in a number of astrophysical environ- the Maxwell-Boltzmann velocity distribution function
ments.31–47 Specifically, the observations of electric field without changing the qualitative behaviour, i.e., the distri-
structures by the FAST38–40,43,47 Satellite, Viking Satellite,33,35 bution function has only one maximum. So, we see that
Freja Satellite36 and GEOTAIL37 and POLAR41,42,47 missions the nonthermal distribution of Cairns et al.48 describes the
in the Earth’s magnetosphere, indicate the existence of fast nonisothermal distribution having a deviation from the
energetic electrons. The electrostatic wave structures observed Maxwell-Boltzmann distribution if be is restricted by
by the Freja Satellite36 can be described by Cairns48 distributed the inequality 0  be  4/7. So, the effective range of be is
nonthermal electrons. Following Cairns et al.,48 the number 0  be  4/7. As 4/7 0.571429, in the present paper, we
density of nonthermal energetic electrons can be written as take 0  be  0.57.
ð1   In the present paper, we have considered a collisionless
2e magnetized dusty plasma system consisting of warm adia-
ne ¼ fe0 v2  / dv; (1)
1 me batic ions, static negatively charged dust grains and nonther-
mal electrons, immersed in a uniform static magnetic field
where –e is the charge of an electron, me is the mass of an directed along a fixed direction. The nonlinear behaviour of
electron, / is the electrostatic potential and fe0(v2) is given IA waves can be described by the continuity equation of
by ions, the equation of motion of ion fluids and the pressure
"  4 # equation of ions. But, instead of the Poisson equation, we
v " # have used the quasi-neutrality condition to make a closed
ne0 1 þ ae
2
vte v2 and consistent system of equations. The quasi-neutrality con-
fe0 ðv Þ ¼ pffiffiffiffiffiffi exp  2 ; (2)
ð1 þ 3ae Þ 2pvte 2vte dition is considered on the basis of the assumption that the
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi length scale of the solitary structure is greater than the
with vte ¼ ðKB Te Þ=me ; ae  0 and KB is the Boltzmann Debye length or the gyroradius.7,11 In this plasma system, IA
constant. solitary structures including double layers and supersolitons
From (1) and (2), we get have been investigated with special emphasis on the follow-
!   ing points:
ne e/ e2 /2 e/
¼ 1  be þ be 2 2 exp : (3) (i) Lifting the equation of continuity, the equation of
ne0 KB Te KB Te KB Te motion and the pressure equation for ion fluids in the
wave frame moving with a constant velocity M nor-
Here, be ¼ 4ae/(1 þ 3ae) with ae  0 and for ae ¼ 0 malized by Cs (ion acoustic speed) along a direction
(ae ¼ 0 () be ¼ 0), we get the usual isothermal distribu- having direction cosines (lx, ly, lz), using appropriate
tion of electrons. It is simple to check that be is an increasing boundary conditions for the solitary structures
function of ae. Here, be (or ae) is the nonthermal parameter together with the condition that the electrostatic
that determines the proportion of the fast energetic particles. potential vanishes at infinity, an energy integral is
Using the inequality ae  0, it is simple to check that derived. The Sagdeev pseudo potential function50
0  be < 4/3. However, we cannot take the entire region of associated with this energy integral is analysed to find
be (0  be < 4/3). In fact, if we plot the nonthermal velocity the nature of existence of different solitary structures
distribution (fe0(v2)) against its velocity (v) in phase space, with respect to the basic parameters of the system.
then we see that for increasing be, the distribution function (ii) For the first time, in the magnetized plasma, an ana-
develops wings which become stronger as be increases. At lytic theory is presented to find the upper bound
the same time, the center density in phase space drops, con- Mpmax (Mnmax) of M for the existence of all positive
sequently, we should not take values of be > 4/7, since that (negative) potential solitary structures, i.e., one can
stage might stretch the credibility of the Cairns model too get positive (negative) potential solitary structures for
far.49 In fact, for be > 4/7, the nonthermal velocity distribu- Mc < M < Mpmax (Mc < M < Mnmax), where Mc is the
tion function of Cairns et al.48 attains three maximum values lower bound of M for the existence of solitary struc-
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
at three different points v ¼ vte  2 þ 7  4ðbe Þ1 , tures, i.e., solitary structures start to exist for M > Mc.
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi (iii) For the first time, in the magnetized plasma, an ana-
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 lytic theory is presented to find the Mach number
v ¼ 0 and v ¼ vte  2þ 7  4ðbe Þ in phase space, and MPPDL (MNPDL) corresponding to a positive (negative)
consequently this property of the nonthermal velocity dis- potential double layer [PPDL (NPDL)] solution of the
tribution function of Cairns et al.48 is qualitatively differ- energy integral, i.e., one can get a PPDL (NPDL)
ent from the Maxwellian velocity distribution function solution of the energy integral for M ¼ MPPDL (M
because the Maxwell-Boltzmann velocity distribution ¼ MNPDL).
function has only one maximum at v ¼ 0. For example, for (iv) With the help of different bounds of M for the exis-
be ¼ 0.58 > 4/7(0.571429), the nonthermal velocity dis- tence of different solitary structures, we draw qualita-
tribution function of Cairns et al.48 has three maxima at tively different compositional parameter spaces which
points v ¼ –1.52369  vte, v ¼ 0 and v ¼ 1.52369  vte. But, show the nature of existence of different solitary struc-
here, our aim is to construct a distribution function which tures with respect to any parameter of the system.
033704-3 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

(v) The regions of existence of different solitary struc- ni ¼ ne þ 1  l: (7)

tures in a compositional parameter space have been
clearly indicated. The compositional parameter space Equations (4)–(6) are supplemented by the following equilib-
shows the existence of negative potential solitons rium charge neutrality condition:
after the formation of NPDL and consequently, the Zd nd0
formation of negative potential supersolitons (NPSSs) ¼ 1  l: (8)
ni0
is confirmed. The description of solitary structures
after the formation of a double layer of same polarity Here, ni, ne, ~ u i ¼ ðuix ; uiy ; uiz Þ, pi, /, (x, y, z), and t are,
through the compositional parameter space is helpful respectively, the ion number density, the electron number
for investigating the existence of supersoliton struc- density, the ion fluid velocity, the ion fluid pressure, the elec-
tures in comparison with the study of supersoliton trostatic potential, the spatial variables and the time, and
structures at some discrete points of the compositional they have been nomalizedby ni0 (unperturbed ion number
parameter spaces. But, still now, supersoliton struc- qffiffiffiffiffiffiffiffi
tures in a magnetized plasma have been discussed at density), ni0 ; Cs ¼ KmB Ti e ; ni0 KB Ti ; KBeTe ; rg ¼ Cs =xc and
some discrete points of the compositional parameter (xc)1, respectively. It is important to note that each spatial
spaces without giving a proper reason for the exis- variable is normalized by the ion gyroradius (rg) and time is
tence of supersoliton structures at those points of the normalized by the inverse of ion gyrofrequency [(xc)1].
compositional parameter spaces. This is purely based Here, mi is the mass of an ion, nd0 and Zd are, respectively,
on the trial and error method. But, if we can draw the the unperturbed dust number density and the
existence domains for different solitary structures,  number
 of elec-
then these existence domains clearly indicate the trons residing on the dust grain surface, c ¼ 53 is the adia-
region of existence of solitons after the formation of batic index, and rie ¼ TTei ; l ¼ nne0i0 .
NPDL, which gives an idea regarding the existence Under the above-mentioned normalization of the field
domain of NPSSs in the compositional parameter variables, the number density (ne) as given by Eq. (3) of non-
spaces. But, of course, this compositional parameter thermal electrons48 can be written as
space cannot clearly indicate the upper bound of M
for the existence of NPSSs. ne ¼ lð1  be / þ be /2 Þe/ : (9)
(vi) We have analyzed the formation of supersoliton struc-
tures and their limitations with the help of phase por- The linear dispersion relation for low frequency
traits of the dynamical system describing the (x  xc) IA waves obtained from the original system of
nonlinear behaviour of IA waves. This analysis deter- equations (without any normalization of independent and
mines the exact bound of M for the existence of dependent variables) corresponding to Eqs. (4)–(9) can be
NPSSs. written as
(vii) For the first time, in the magnetized plasma, the
ðx=xc Þ
mechanism of transition from the NPSS to a negative ¼ fMs2 þ ðrg k? Þ2 g1=2 ; (10)
potential conventional soliton (soliton before the for- rg k z
mation of double layer of same polarity) after the for-
where
mation of a double layer of same polarity has been
investigated through the phase portrait analysis. sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1
Ms ¼ crie þ ; (11)
lð1  be Þ
II. BASIC EQUATIONS
2
Here, we consider a dusty plasma system consisting of k? ¼ kx2 þ ky2 , and we have assumed that all the perturbed
warm adiabatic ions, static negatively charged dust grains dependent variables vary as exp ½iðkx x þ ky y þ kz z  xtÞ .
and nonthermal electrons immersed in a uniform (space If we use the normalized variables, i.e., if we use the
independent) static (time independent) magnetic field system of Eqs. (4)–(9), the dimensionless linear dispersion
~ ¼ B0 ^
ðB z Þ directed along the z-axis. The nonlinear behaviour relation for low frequency (x/xc  1) IA waves can be writ-
of IA waves can be described by the continuity equation, the ten as
equation of motion and the pressure equation of ion fluids
1=2
together with the quasi-neutrality condition x lð1  be Þ 2
¼ þ k? : (12)
kz 1 þ crie lð1  be Þ
@ni ~
þ r  ðni~u i Þ ¼ 0; (4) Here, x is normalized by xc and kx, ky, kz are normalized by
@t
  1/rg, i.e., replacing x/xc by x, rgkz by kz and rgk? by k? in
@ ~ rie ~ ~ þ~ Eq. (10), we get Eq. (12). The dimensionless linear disper-
þ~
ui  r ~ ui þ rpi ¼ r/ u i  ^z ; (5)
@t ni sion relations (5) of Yinhua and Yu,10 (9) of Choi et al.7 and
(10) of Choi et al.11 can be easily obtained from the dimen-
@pi ~ i þ cpi ðr
~ ~ sionless linear dispersion relation (12) of this paper by con-
þ ð~
u i  rÞp u i Þ ¼ 0; (6)
@t sidering appropriate values of parameters. For example, the
033704-4 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

dispersion relation (12) is exactly the same as the dispersion where

relation (10) of Choi et al.,11 if we take be ¼ 0, c ¼ 1, rie ¼ r ð/
c
and ky ¼ 0. G ¼ Gð/Þ ¼ rie fðni Þ  1g þ ni d/: (24)
0
Using (4), the pressure equation of ion fluids (6) can be
written as pi ¼ nci . Using this expression of pi, Eq. (5) can be From Eqs. (19), (21) and (23), we get the following expres-
simplified as sion for lxuix þ lyuiy:
 
@ M l2z
þ~ ~
ui  r ~ u i þ crie ðni Þc2 rn ~ þ~
~ i ¼ r/ u i  ^z : (13) lx uix þ ly uiy ¼ M   Gð/Þ: (25)
@t ni M
Again, using (9), Eq. (7) can be written as Substituting the expression of W as given by (21) in Eqs.
2 / (16) and (17), we get
ni ¼ lð1  be / þ be / Þe þ 1  l: (14)
M duix dH
We have used Eqs. (4), (13) and (14) to derive the energy  þ lx  uiy ¼ 0; (26)
ni dn dn
integral.
M duiy dH
 þ ly þ uix ¼ 0: (27)
III. ENERGY INTEGRAL ni dn dn
To study the arbitrary amplitude time independent IA Multiplying (26) by lx and (27) by ly, and finally, adding the
solitary structures, we suppose that all the dependent varia- resulting equations, we get
bles depend only on a single variable n ¼ lxx þ lyy þ lzz – Mt
with l2x þ l2y þ l2z ¼ 1, where M is independent of x, y, z and t. dP
ly uix  lx uiy ¼  ; (28)
Therefore, lifting Eq. (4) and each component of the momen- dn
tum equation (13) in the waveframe moving with a constant
where we have used (25) and
velocity M normalized by Cs along a direction having direc-
tion cosines (lx, ly, lz), we get the following equations: M2 crie
P¼ 2
þ ðni Þc1 þ /: (29)
d 2ðni Þ c1
ðni WÞ ¼ 0; (15)
dn
Solving Eqs. (25) and (28) for the unknowns uix and uiy, we
duix dH get
W þ lx  uiy ¼ 0; (16)
dn dn

M l2z dP
duiy dH lx M   Gð/Þ  ly
W þ ly þ uix ¼ 0; (17) ni M dn
dn dn uix ¼ ; (30)
l2x þ l2y
duiz dH

W þ lz ¼ 0; (18) M l2z dP
dn dn ly M   Gð/Þ þ lx
ni M dn
uiy ¼ : (31)
where we have used the following notations: l2x þ l2y
W ¼ M þ lx uix þ ly uiy þ lz uiz ; (19) Substituting these two values of uix and uiy in (26) or in (27),
crie we get the following equation:
H ¼ Hð/Þ ¼ ðni Þc1 þ /: (20)
c1
d2 P l2
¼ f ð/Þ ¼ ni  1  z2 ni Gð/Þ: (32)
From (15), we get the following expression of W: dn 2 M
M
W¼ ; (21) This equation can be written as
ni
where we have used the following boundary conditions: dQ 2 dR 2
þ Q ¼ f ð/Þ; (33)
  d/ R d/ R
d/ 2
ni ; uix ; uiy ; uiz ; /; ! ð1; 0; 0; 0; 0; 0Þ as jnj ! 1: where Q ¼ ðd/ dP
dn dn Þ and R ¼ d/.
Equation (33) is a first order and first degree linear dif-
(22)
ferential equation in Q and the general solution of (33) can
Substituting this expression of W as given by (21) in Eq. be written as
(18), integrating the resulting equation and finally using the  2  2 ð
boundary condition (22) to find the integration constant, we d/ dP dP
 ¼ 2 f ð/Þd/ þ C1 ; (34)
get the following equation: dn d/ d/

lz where C1 is an integration constant. Using the boundary con-

uiz ¼ Gð/Þ; (23) dition (22), we get
M
033704-5 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

ð
dP any value of / and for any set of values of the parameters of
0¼ 2 f ð/Þd/j/¼0 þ C1 : (35)
d/ the system. Again, one can easily check that K(0) ¼ 1.
It is simple to check that Vð0Þ ¼ 0; V 0 ð0Þ ¼ 0 and
Using Eq. (35), Eq. (34) can be put in the following 00
V ð0Þ < 0 give the following bounds for M:
form:
 2 Mc < M < M s ; (40)
1 d/
þ Vð/Þ ¼ 0; (36) where Mc ¼ lzMs, and we have assumed the condition
2 dn dP
d/ j/¼0 6¼ 0.
where Therefore, the solitary structures start to exist for
ð/ M > Mc. Although M < Ms for the existence of solitary struc-
dP tures, Ms is not necessarily the exact upper bound of M for
f ð/Þd/
d/ the existence of solitary structures. To find the exact upper
Vð/Þ ¼ VðM; /Þ ¼  0   : (37) dP
dP 2 bound of M, we consider the equation d/ ¼ 0 along with the
d/ inequality M < Ms. In fact, the inequality M < Ms states that
dP
d/ 6¼ 0 at / ¼ 0, but to consider the exact bounds of M, it is
Equation (36) is the well-known energy integral with dP
necessary to consider d/ 6¼ 0 throughout the entire possible
Sagdeev pseudopotential V(/)( ¼ V(M, /)) given in (37).
According to the mechanical analogy of Sagdeev,50 the sys- range of /, where V(/) is well defined.
dP dP
tem supports a positive (negative) potential solitary wave Instead of considering d/ 6¼ 0, one can consider d/ ¼ 0,
[PPSW] [(NPSW)] solution of (36), if (i) / ¼ 0 is the posi- and this equation gives
tion of unstable equilibrium of a particle of unit mass associ- M2
¼ Kð/Þ: (41)
ated with the energy integral (36), i.e., Vð0Þ ¼ V 0 ð0Þ ¼ 0 Ms2
and V 00 ð0Þ < 0; (ii) the condition for the oscillation of the
particle within the interval minf0; /m g < / < maxf0; /m g Obviously Eq. (41) is well defined, only when K(/) > 0 for
holds good when the particle is slightly displaced from all /. From Eq. (39), using (7) and (9), we see that K(/) is
dP
its unstable position of equilibrium (/ ¼ 0), i.e., Vð/m Þ independent of M. Therefore, d/ ¼ 0 holds good only when
pffiffiffiffiffiffiffiffiffiffiffi
¼ 0; V 0 ð/m Þ > 0 ðV 0 ð/m Þ < 0Þ for some /m > 0 (/m < 0); M assumes the value Ms Kð/Þ and consequently, for the
and (iii) the energy integral (36) is well defined within the existence of positive (negative) potential solitary structures,
pffiffiffiffiffiffiffiffiffiffiffiffiffi
interval minf0; /m g < / < maxf0; /m g, i.e., V(/) < 0 for M cannot exceed the value Ms Kð/m Þ, where K(/) attains
all minf0; /m g < / < maxf0; /m g. On the other hand, the its global minimum at / ¼ /m in the positive (negative) side
system supports a PPDL (NPDL) solution of (36) if the second of the /-axis, i.e., for /m  0 (/m  0). Now, we explain the
condition is replaced by Vð/m Þ ¼ 0; V 0 ð/m Þ ¼ 0; V 00 ð/m Þ phrase “global minimum” with the help of Fig. 1. We have
< 0 for some /m > 0 (/m < 0), which states that the particle already stated that K(/) > 0 for any value of / and for any
cannot be reflected back from the point / ¼ /m to the point
/ ¼ 0. The physical interpretation for the existence of differ-
ent solitary structures has been given in detail in the paper of
Das et al.51 Using this mechanical analogy of Sagdeev,50 we
present analytical theories to determine different bounds of M
for the existence of different solitary structures, and also to
find the Mach numbers MPPDL and MNPDL corresponding to a
PPDL and a NPDL solution, respectively, in Subsections III A
and III B.

A. Different bounds of M: Mc, Mpmax and Mnmax

Differentiating Eq. (29) with respect to /, we get
 " 2 #
dP dni M
¼ Ms2 n3
i  Kð/Þ ; (38)
d/ d/ Ms2

where
"  1 #
1 cþ1 3 dni
Kð/Þ ¼ 2 crie ni þ ni : (39)
Ms d/
dni FIG. 1. K(/) is plotted against / for c ¼ 5/3 with (a) rie ¼ 0.01, l ¼ 0.5,
Now, it is simple to check that ni, d/ and K(/) are strictly be ¼ 0.18, (b) rie ¼ 0.01, l ¼ 0.5, be ¼ 0.2, (c) rie ¼ 0.1, l ¼ 0.6, be ¼ 0.2,
dni
positive functions of /, i.e., ni > 0, d/ > 0 and K(/) > 0 for (d) rie ¼ 0.01, l ¼ 0.6, be ¼ 0.2, and (e) rie ¼ 0.01, l ¼ 0.6, be ¼ 0.3.
033704-6 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

set of physically admissible values of parameters. Plotting (Mnmax ¼ Ms). Again, there does not exist any /(0)
pffiffiffiffiffiffiffiffiffiffiffi
K(/) against / for different values of parameters of the sys- (/  0), for which M assumes the value Ms Kð/Þ
tem, it is simple to check that K(/) > 0 for all /. In Figs. pffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffi
because M ¼ Ms Kð/Þ ) MMs ¼ Kð/Þ > 1 ) M
1(a)–1(e), K(/) is plotted against / for different values of
> Ms which contradicts the fact Mc < M < Ms and
parameters. These figures indicate five qualitatively different dP
curves of K(/) against /. Figure 1(a) shows that the mini- consequently, d/ 6¼ 0 for / 6¼ 0. So, V(/) is well
mum value of K(/) on the positive side of the /-axis (/  0) defined as a real valued function of / if it is well
as well as on the negative side of the /-axis (/  0) is equal defined at / ¼ 0. Therefore, the upper bound of M for
to K(0). Figure 1(b) shows that K(/) has only one minimum the existence of the positive (negative) potential soli-
at / ¼ /m1 ð> 0Þ such that 0 < Kð/m1 Þ < 1, and the mini- tary structures is given by the inequality (40).
mum value of K(/) on the negative side of the /-axis (II) If 0  K(/)  1 for some / then let K(/) attains its
(/  0) is equal to K(0). Figure 1(c) shows that K(/) global minimum at p ¼ /m such that 0  K(/m)  1.
/ ffiffiffiffiffiffiffiffiffiffiffiffiffi
has only one minimum at / ¼ /m2 ð< 0Þ such that 0 Therefore, M ¼ Ms Kð/m Þ is a minimum value of M
< Kð/m2 Þ < 1 and the minimum value of K(/) on the posi- at which there is a singularity p V/) at / ¼ /m, i.e.,
of ffiffiffiffiffiffiffiffiffiffiffiffiffi
dP
tive side of /-axis (/  0) is equal to K(0). Figure 1(d) j
d/ /¼/m ¼ 0 when M ¼ M s  Kð/ m Þ.

shows that K(/) has two minima at / ¼ /m2 ð< 0Þ and The us to take Mpmax ¼ Ms
ffi above discussionspffiffiffiffiffiffi
pffiffiffiffiffiffi guide
ffi
/ ¼ /m3 ð< 0Þ such that 0 < Kð/m2 Þ; Kð/m3 Þ < 1, and the  K þ and Mnmax ¼ Ms  K  .
minimum value of K(/) on the positive side of /-axis
(/  0) is equal to K(0). Figure 1(e) shows that K(/) has two B. Mach numbers MPPDL and MNPDL corresponding to a
minima at / ¼ /m1 ð> 0Þ and / ¼ /m2 ð< 0Þ such that /m1 PPDL and a NPDL
lies on the positive side of the /-axis (/  0), /m2 lies in Let us use the notation V(M, /) instead of V(/). Using
the negative potential side of the /-axis (/  0) and this notation, the conditions for the existence of a PPDL
0 < Kð/m1 Þ; Kð/m2 Þ < 1. Now, our aim is to minimize (NPDL) solution of the energy integral (36) are (i) VðM; 0Þ
Kð/m1 Þ and/or Kð/m2 Þ in order to obtain the global minimum ¼ 0; V 0 ðM;0Þ ¼ 0; V 00 ðM;0Þ < 0; (ii) VðM;/m Þ ¼ 0; V 0 ðM;/m Þ
on the positive and/or negative potential side(s). By the ¼ 0; V 00 ðM;/m Þ < 0 for some /m >0 (/m <0); and (iii) V(M,
phrase “global minimum,” we would like to mean the mini- /)<0 for all 0</</m (/m </<0). We have used condi-
mum of the minimum values of K(/). For example, for Fig. tion (i) to find the lower bound and get a rough idea about
1(d), the global minimum of K(/) is minfKð/m2 Þ; Kð/m3 Þg the upper bound for the existence of any solitary structure
¼ Kð/m2 Þ on the negative side of the /-axis (/  0), including double layers of any polarity. Condition (iii) is
whereas the global minimum of K(/) is K(0) on the positive necessary to define the energy integral (36) within the inter-
side of the /-axis (/  0), i.e., we can say that K(/) attains val minf0;/m g < / < maxf0;/m g. Now, we use condition
the global minimum value at / ¼ /m ¼ /m2 on the negative (ii) to find the Mach number corresponding to a PPDL and/
side of /-axis (/  0), whereas K(/) attains the global mini- or NPDL solution of the energy integral (36). Let Mdl and
mum value at / ¼ /m ¼ 0 on the positive side of the /-axis /dl 6¼ 0 are the solutions of the equations V(M, /m)¼0 and
(/  0). On either side of the /-axis, including / ¼ 0, there V 0 ðM;/m Þ ¼ 0 for the unknowns M and /m. It is important to
always exists a /m such that K(/) attains the global mini- note that /¼0 is a solution of the equations V(M, /m)¼0
mum value at / ¼ /m. If Kþ and K–, respectively, denote the and V 0 ðM;/m Þ ¼ 0 for any value of M in view of condition
global minimum of K(/) on the positive and negative poten- (1). Therefore, we have V(Mdl, /dl)¼0 and V 0 ðMdl ;/dl Þ ¼ 0
tial sides, then Kþ ¼ K(/m) with /m  0 and K– ¼ K(/m) for non-zero /dl and consequently, Mdl corresponds to a
with /m  0. For example, Kþ ¼ K(/m) with /m ¼ /m1 and Mach number MPPDL (MNPDL) for a PPDL (NPDL) solution
K– ¼ K(/m) with /m ¼ 0 for Fig. 1(b), whereas Kþ ¼ K(/m) of the energy integral (36) if /dl >0 (/dl <0) and /dl is the
with /m ¼ 0 and K– ¼ K(/m) with /m ¼ /m2 for Fig. 1(d). smallest (largest) real, satisfying the conditions V(Mdl,
From the above discussions, it is clear that both Kþ and K– /dl)¼0 and V 0 ðMdl ;/dl Þ ¼ 0.
exist finitely and consequently, for the existence of a positive Equations V(Mdl, /dl) ¼ 0 and V 0 ðMdl ; /dl Þ ¼ 0 give the
(negative) potential solitary structure, M is restricted by the following expression of Mdl:
pffiffiffiffiffiffiffi pffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
inequality Mc < M < Ms  K þ (Mc < M < Ms  K  ).
l2z ni ð/dl ÞGð/dl Þ
The above analysis can be illustrated in the following Mdl ¼ ; (42)
paragraphs. ni ð/dl Þ  1
Let Mpmax (Mnmax) is the upper bound of M for the exis-
where G(/) is given by Eq. (24). The equation V(Mdl,
tence of positive (negative) potential solitary structures, i.e.,
/dl) ¼ 0 gives the following equation:
for the existence of positive (negative) potential solitary struc- ð /dl
tures, we must have Mc < M < Mpmax (Mc < M < Mnmax). P0 ðMdl ; /Þf ðMdl ; /Þd/ ¼ 0; (43)
Now, we have the following two cases: (I) K(/) > 1 and (II) 0
0  K(/)  1. 0 dP
where P ¼ d/.
Substituting (42) in (43), we get an equation
(I) If K(/) > 1 for all /  0 (/  0), then the upper for the unknown /dl. Solving this equation for the unknown
bound of M for the existence of positive (negative) /dl at least numerically and putting this solution in Eq. (42),
potential solitary structures is given by Mpmax ¼ Ms one can get the value of Mdl. If V 00 ðMdl ; /dl Þ < 0, then this
033704-7 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

Mdl corresponds to the Mach number MPPDL (MNPDL) of a Based on the above-mentioned notations and terminolo-
PPDL (NPDL). gies, the existence domains are self-explanatory. For exam-
Now, Mnmax (Mpmax) can restrict the existence of all ple, Fig. 2 shows the existence domain or the compositional
NPSWs (PPSWs). But, if both Mnmax (Mpmax) and parameter space with respect to be. From this figure, we have
M ¼ MNPDL (M ¼ MPPDL) exist for fixed values of the param- the following observations. (i) The system supports PPSWs
eters, then we must have Mc < MNPDL < Mnmax (NPSWs) for admissible values of be whenever M lies within
(Mc < MPPDL < Mpmax). For this case, we can split the entire the interval Mc < M < Mpmax (Mc < M < Mnmax). (ii) The sys-
range of M into two disjoint sub-intervals Mc < M < MNPDL tem supports the coexistence of solitary waves of both polar-
(Mc < M < MPPDL) and MNPDL < M < Mnmax (MPPDL < M ities and the region of coexistence is bounded by the curves
< Mpmax). For Mc < M < MNPDL (Mc < M < MPPDL), we get a M ¼ Mc and M ¼ Mnmax. This figure also shows that the exis-
sequence of NPSWs (PPSWs) converging to the NPDL tence region of negative potential solitary structures is a sub-
(PPDL) solution at M ¼ MNPDL (M ¼ MPPDL). In other words, set of the existence region of positive potential solitary
the NPDL (PPDL) solution at M ¼ MNPDL (M ¼ MPPDL) can structures. (iii) The system starts to support NPDL along
restrict the occurrence of all NPSWs (PPSWs) for all M lying the curve M ¼ MNPDL whenever be  bð1Þ e . (iv) Now, for
within the interval Mc < M < MNPDL (Mc < M < MPPDL), be  bð1Þ
e , we observe that M NPDL  M nmax. Consequently,
whereas for MNPDL < M < Mnmax (MPPDL < M < Mpmax), we ð1Þ
for be  be , the curve M ¼ MNPDL cannot bound the exis-
get NPSWs (PPSWs) after the formation of NPDL (PPDL) tence region of NPSWs and there exist two types of NPSWs.
at M ¼ MNPDL (M ¼ MPPDL) and consequently, the existence The first type of NPSWs is restricted by Mc < M < MNPDL
of supersolitons52 or dias-type solitons22 is confirmed. These and the second type of NPSWs is restricted by
types of solitons have been recently reported by many MNPDL < M < Mnmax, and there exists a jump type disconti-
authors.28–30,53–59 nuity in the amplitudes of the NPSWs just before and just
after the formation of the double layer. In other words, the
IV. EXISTENCE DOMAINS NPDL solution is unable to restrict the occurrence of
NPSWs and there exist NPSWs after the formation of the
Here, we draw different regions bounded by the curves
double layer and such NPSWs are bounded by the curves
M ¼ Mc, M ¼ Ms, M ¼ Mpmax, M ¼ Mnmax, M ¼ MNPDL, and
M ¼ MNPDL and M ¼ Mnmax. Therefore, the existence of
M ¼ MPPDL with respect to any parameter of the system.
NPSWs after the formation of NPDL confirms the existence
These regions essentially describe the existence domains of
of a sequence of NPSSs. (v) The system does not support
different solitary structures bounded by different curves. For
any PPDL and consequently, it does not support any positive
example, the region bounded by the curves M ¼ Mc and
potential supersoliton (PPSS).
M ¼ Mpmax (M ¼ Mnmax) represents the domain of existence
Figures 3–5 are self-explanatory. From Figs. 2 and 3, we
of all PPSWs (NPSWs) including all PPDLs (NPDLs). At have the following observations. (i) There is no effective
each point along the curve M ¼ MNPDL (M ¼ MPPDL), one change in the area of the existence region of PPSWs for
can get a NPDL (PPDL). The region bounded by the curves increasing l. (ii) The existence region of NPSWs decreases
M ¼ Mc and M ¼ MNPDL (M ¼ MPPDL) represents the domain as l increases. (iii) With the increment in l, the value of bð1Þ
e
of existence of all NPSWs (PPSWs) before the formation of increases, i.e., with the increment in l, the system starts to
NPDL (PPDL). If Mc < MNPDL < Mnmax, then the region
bounded by the curves M ¼ MNPDL and M ¼ Mnmax represents
the domain of existence of all NPSWs after the formation of
NPDL. We have used the following notations and terminolo-
gies: C – Region of coexistence of both NPSWs and PPSWs,
N – Region of existence of NPSWs, P – Region of existence
of PPSWs and NS – Region of existence of negative poten-
tial solitons after the formation of NPDL. The curves
M ¼ Mc, M ¼ Ms, M ¼ Mpmax, M ¼ Mnmax and M ¼ MNPDL
have been shown as the heavy magenta line, the heavy black
line, the blue line, the red line and the heavy green line,
respectively.
Now, we define the following three cut-off values of be
which help us to study the different existence regions. (i) bð1Þ
e
is a cut-off value of be such that NPDL solutions start to
exist for be  bð1Þ ð2Þ
e . (ii) be is a cut-off value of be such that
M ¼ Mpmax coincides with Ms for 0  be  bð2Þ e , i.e., for
be > be ; M ¼ Mpmax is different from Ms. (iii) bð3Þ
ð2Þ
e is a cut- FIG. 2. Mc, Ms, Mpmax, Mnmax and MNPDL are plotted against be. The
off value of be such that M ¼ Mnmax starts to differ from Mc magenta line, the black line, the blue line, the red line and the green line cor-
respond to the curves M ¼ Mc, M ¼ Ms, M ¼ Mpmax, M ¼ Mnmax and
for be > bð3Þ ð3Þ
e , i.e., for 0  be  be ; M ¼ Mnmax coincides M ¼ MNPDL, respectively. This figure shows the existence of NPSWs after
with Mc. the formation of NPDL along with other solitary structures.
033704-8 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

FIG. 3. Mc, Ms, Mpmax, Mnmax and MNPDL are plotted against be. The
magenta line, the black line, the blue line, the red line and the green line cor-
respond to the curves M ¼ Mc, M ¼ Ms, M ¼ Mpmax, M ¼ Mnmax and M
¼ MNPDL, respectively. This figure shows that there does not exist any FIG. 5. Mc, Ms, Mpmax, Mnmax and MNPDL are plotted against be for different
NPSW after the formation of NPDL. lz. The magenta line, the black line, the blue line, the red line and the green
line correspond to the curves M ¼ Mc, M ¼ Ms, M ¼ Mpmax, M ¼ Mnmax and
support NPDLs as well as NPSSs for higher nonthermality M ¼ MNPDL, respectively. There is a qualitative difference between (a) and
(b). For (a), the system supports the coexistence of PPSWs and NPSWs after
of electrons. (iv) The existence region of NS decreases for the formation of NPDL as indicated by P þ NS, whereas figure (b) shows
increasing values of l and ultimately, the system does not that the system does not support any NPSW after NPDL.
support any NS.
Figure 4 shows the existence domain with respect to l magnetic field. In fact, Fig. 5(a) shows the coexistence of
for isothermal electrons, i.e., for be ¼ 0. From this figure, we PPSWs and NPSWs after the formation of NPDL, whereas
have the following observations. (i) Mnmax coincides with Ms Fig. 5(b) shows that the coexistence of PPSWs and NPSWs
for 0 < l  l(1), i.e., Mnmax starts to differ from Ms for after the formation of NPDL is impossible.
l > l(1). (ii) The system supports the coexistence of both In Fig. 6, Mpmax and Mc are plotted against be for two
NPSWs and PPSWs. (iii) The system does not support the different values of l. This figure shows that there is no effec-
double layer of any polarity. (iv) This figure also shows that tive change in the area of the existence region of PPSWs for
the existence region of NPSWs is smaller than the existence increasing l. In Fig. 7, Mnmax and Mc are plotted against be
region of PPSWs. for two different values of l. This figure shows that the
Figure 5 shows that the qualitative behaviour of IA soli- region of existence of NPSWs decreases for increasing l. In
tary structures is very much sensible on the angle of propaga- Fig. 8, MNPDL is plotted against be for different values of l.
tion of the solitary structures with the uniform static In this figure, we see that the interval of be for the existence
of NPDLs increases for increasing values of l if l  l(cr),
and if l exceeds l(cr), the interval of be for the existence of
NPDLs decreases for increasing values of l, where l(cr) is a
critical value of l. On the other hand, the region of existence

FIG. 4. Mc, Ms, Mpmax, Mnmax and MNPDL are plotted against l for be ¼ 0.
The magenta line, the black line, the blue line, the red line and the green line
correspond to the curves M ¼ Mc, M ¼ Ms, M ¼ Mpmax, M ¼ Mnmax and FIG. 6. Mc and Mpmax are plotted against be for different values of l. There
M ¼ MNPDL, respectively. As there is no green line corresponding to the is no considerable change in the area of the existence regions of PPSWs for
curve M ¼ MNPDL, then there does not exist any NPDL solution for be ¼ 0. increasing l.
033704-9 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

value (Mcr) of M such that the phase portraits of the solitons

after the formation of the double layer [M > Mcr(> MNPDL)]
are qualitatively the same as the conventional solitons before
the formation of the double layer (Mc < M < MNPDL).
Consequently, for the existence of NPSSs, we have
MNPDL < M < Mcr. So, there must be a transition of solitary
structures, viz., soliton ! double layer ! supersoliton
! soliton after the formation of the double layer of same
polarity. To discuss this transition process, it is important to
know the phase portraits of the dynamical system describing
different IA solitary structures.
Now, differentiating the energy integral (36) with
respect to /, we can write the resulting equation in the fol-
lowing two coupled differential equations:
FIG. 7. Mc and Mnmax are plotted against be for different values of l. The
region of existence of NPSWs decreases for increasing l. d/1 d/2
¼ /2 ; ¼ V 0 ð/1 Þ; where /1 ¼ /: (44)
dn dn
of NPSWs before the formation of NPDLs increases for
increasing values of l whenever l  l(cr), whereas the region The equilibrium point of (44) is ð/
1 ; /
2 Þ, where /
2 ¼ 0
of existence of NPSWs before the formation of NPDLs and /
1 is given by V 0 ð/
1 Þ ¼ 0. The equation V 0 ð/
1 Þ ¼ 0
decreases for increasing values of l for l > l(cr). For the val- gives the value(s) of /
1 as a function of the parameters of the
ues of the parameters as mentioned in Fig. 8, the value of system along with M, i.e., /
1 ¼ /
1 ðM; c; l; be ; rie ; lz Þ.
l(cr) is approximately equal to 0.6. Therefore, the values of the parameters c, l, be, rie and lz are
not sufficient to determine /
1 . To get /
1 , it is necessary to
V. PHASE PORTRAITS know the value of M. But, one can get the value of M by con-
sidering the existence domain with respect to any parameter
In Sec. IV, we have seen that the system supports the of the system. For example, if the values of c, l, rie and lz are
following IA solitary structures: (i) NPSWs, (ii) PPSWs, (iii) given, then by drawing the existence domain with respect to
the coexistence of NPSWs and PPSWs (iii) NPDLs, (iv) the be, one can get the value of be and the corresponding value of
coexistence of PPSWs and NPDLs, (v) NPSWs after the for- M for the existence of the required solitary structure.
mation of NPDLs, and (vi) the coexistence of PPSWs and To describe the existence and the shape of NPSSs, and
NPSWs after the formation of NPDLs. From the composi- the coexistence of solitons of both polarities, we consider
tional parameter spaces or existence domains of different IA Figs. 9–15, where we have used the existence domain as
solitary structures, it is impossible to find the upper bound of shown in Fig. 2 to determine the Mach numbers for the for-
the Mach number for the existence of NPSSs, although it is mation of different solitary structures including NPSSs.
simple to check that there is a finite jump between the ampli- Figures 9(b)–15(b) are, respectively, the phase portraits
tudes of solitons just before and just after the formation of of PPSW, NPSW, the coexistence of solitary waves of both
the double layer which confirms the existence of NPSSs. So, polarities, NPDL, NPSW just before the formation of NPDL,
it is not possible to investigate the existence of a critical

FIG. 9. (a) V(/) and (b) the phase portrait of the system have been drawn on
the same /-axis. This figure shows that the curve V(/) of the upper panel (a)
corresponds to the separatrix of the phase portrait of the dynamical system
FIG. 8. MNPDL is plotted against be for different values of l. This figure (44) as shown by the heavy blue line. Each maximum (minimum) point of
shows that for increasing l, NPDLs start to exist for lower values of be up to V(/) corresponds to a saddle (non-saddle) fixed point of the dynamical sys-
l ¼ 0.6, and after this value of l, NPDLs start to exist for higher values of be. tem (44). This figure confirms the existence of PPSW.
033704-10 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

FIG. 10. (a) V(/) and (b) the phase portrait of the system have been drawn
on the same /-axis. This figure also shows that the curve V(/) of the upper
panel (a) corresponds to the separatrix of the phase portrait of the dynamical
system (44), as shown by heavy blue line. Each maximum (minimum) point FIG. 12. (a) V(/) and (b) the phase portrait have been drawn on the same /
of V(/) corresponds to a saddle (non-saddle) fixed point of the dynamical for M ¼ MNPDL. The portion of V(/) included within –1.94  /  0 has been
system (44). This figure confirms the existence of NPSW. shown in the inset. We see that the separatrix (as shown by the heavy blue
line) of the phase portrait appears to pass through two saddle points corre-
sponding to two maximum values of V(/) at / ¼ 0 and / ¼ –1.94 (approxi-
NPSW just before the formation of NPDL (NPSS), and mately). This figure confirms the existence of NPDL.
NPSW after the formation of double layer. In the upper panel
[or marked (a)] of each figure of Figs. 9–15, V(/) is plotted solitary structure appears to start and end at the saddle point
against /, whereas the lower panel [or marked (b)] of each (0, 0). The separatrix corresponding to a solitary structure is
figure shows the phase portrait of the system (44). The curve shown with a heavy bold blue line, whereas other separatri-
V(/) and the phase portrait have been drawn on the same ces (if any) are shown by heavy bold green lines. The phase
horizontal axis /(¼/1). The small solid circle and the small portrait of the dynamical system corresponding to a conven-
solid star correspond to a saddle point and an equilibrium tional soliton contains only one separatrix that appears to
point other than the saddle point of system (44), respectively. start and end at the origin enclosing only one non-saddle
From Figs. 9–15, we see that each maximum (minimum) fixed point, and there does not exist any other separatrix
point of V(/) corresponds to a saddle point (an equilibrium within the first one. Again, Paul et al.30 reported that the tra-
point other than a saddle point) of system (44). jectory corresponding to the separatrix approaches the origin
The one–one correspondence between the separatrix of as n ! 61, and a separatrix corresponding to a solitary
the phase portrait (as shown with a heavy bold line) in the structure does not correspond to a periodic solution because
lower panel with the curve V(/) against / of the upper panel for this case, the trajectory takes forever trying to reach a
has been explained elaborately in the recently published saddle point which proves an important result that a pseudo-
paper of Paul et al.30 The origin (0, 0) is always a saddle particle associated with the energy integral (36) takes an infi-
point of system (44), and the separatrix corresponding to a nite long time to move away from its unstable position of
equilibrium and it continues its motion until / takes the

FIG. 11. (a) V(/) and (b) the phase portrait of the system have been drawn FIG. 13. (a) V(/) and (b) the phase portrait of the system have been drawn
on the same /-axis. This figure confirms the coexistence of both PPSW and on the same /-axis for M ¼ MNPDL – 0.002. The existence of NPSW is con-
NPSW. firmed before the formation of the double layer.
033704-11 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

the saddle point (0, 0), but it encloses two non-saddle fixed
points—one lies on the positive side of /-axis, viz., (0.7, 0)
and the other fixed point lies on the negative side of /-axis,
viz., (–0.43, 0). As one separatrix that appears to start
and end at the origin (0, 0) (saddle point) encloses two
non-saddle fixed points lying on the opposite direction of
/-axis, we have a coexistence of solitary waves of both
polarities.
Figure 12(b) shows the phase portrait of the dynamical
system corresponding to a NPDL, and this figure shows that
the separatrix corresponding to the double layer solution
appears to start and end at the saddle (0, 0), and again, it
appears to pass through the saddle point at (–1.94, 0)
enclosing the non-saddle fixed point (–0.91, 0). In Fig.
12(a), V(/) is plotted against /. Figures 12(a) and 12(b)
together give a one–one correspondence between the sepa-
ratrix of the phase portrait as shown with a heavy blue line
in the lower panel with the curve V(/) against / of the
FIG. 14. (a) V(/) and (b) the phase portrait of the system have been drawn upper panel. This mechanism holds good for the formation
on the same /-axis for M ¼ MNPDL þ 0.002. Here, we see that the separatrix of PPSWs and also for the formation of NPSWs. Therefore,
of the phase portrait of the dynamical system (44) which appears to pass
Fig. 12(b) shows that the separatrix corresponding to the
through the saddle (0, 0) contains more than one non-saddle fixed point and
another separatrix through the non-zero saddle (–1.66, 0). Consequently, the double layer solution appears to pass through two saddle
separatrix that appears to pass through the saddle (0, 0) corresponds to a points and it encloses other two fixed points. If both the
supersoliton. Therefore, this figure confirms the existence of supersolitons saddle points exist after a small increment of M from
after the formation of the double layer.
M ¼ MNPDL, then the separatrix through the origin encloses
an inner separatrix through a non-zero saddle and at least
value /m and again, it takes an infinite long time to come two equilibrium points as shown in Fig. 14(b) for M
back to its unstable position of equilibrium,44,51 where ¼ MNPDL þ 0.002. So, according to the definition of super-
V(/m) ¼ 0 and V 0 ð/m Þ > 0 ðV 0 ð/m Þ < 0Þ for /m > 0 soliton as prescribed by Dubinov and Kolotkov,52 Fig.
(/m < 0). The closed curve about an equilibrium point (other 14(b) corresponds to a NPSS. But, for M ¼ MNPDL þ 0.01,
than a saddle point) contained in at least one separatrix indi- we get Fig. 15.
cates the possibility of the periodic wave solution about that From Figs. 10(b), 11(b) (part of the phase portrait corre-
fixed point. sponding to NPSW), 13(b), and 15(b), we observe that there
Figure 11(b) shows the phase portrait of the dynamical is no qualitative difference between these four phase por-
system (44) corresponding to the coexistence of solitary traits. For all four phase portraits, there exist a saddle at the
waves of both polarities, and in Fig. 11(a), V(/) is plotted origin, a non-saddle fixed point lying on the negative /-axis,
against / which shows the coexistence of solitary waves of and we have only one separatrix which appears to start and
both polarities. Here, from the phase portrait, we see that end at the saddle point (0,0) enclosing the non-saddle fixed
there is only one separatrix that appears to start and end at point lying on the negative /-axis. Therefore, Fig. 15 indi-
cates the existence of a NPSW after the formation of NPDL,
and the phase portrait corresponding to this NPSW looks
like a phase portrait corresponding to a conventional
NPSW. But, the blue solid curve and the blue dashed curve
of Fig. 16 show that there is a finite jump between the ampli-
tudes of solitons for M ¼ MNPDL – 0.01 and M ¼ MNPDL
þ 0.01. This is not only true for M ¼ MNPDL – 0.01 and
M ¼ MNPDL þ 0.01, this result is also true for M ¼ MNPDL – 
and M ¼ MNPDL þ , i.e., there is a finite jump between the
amplitudes of solitons for M ¼ MNPDL –  and M ¼ MNPDL
þ , and this has been shown by the black solid curve and the
black dashed curve of Fig. 16 for  ¼ 0.0001. Thus, we get a
sequence of supersolitons for increasing values of
M(>MNPDL), but if M exceeds a cut-off value Mcr, then the
solitons after the formation of the double layer simply reduce
to conventional solitons, i.e., there exists a transition from a
supersoliton to a soliton with the increment of M after the
FIG. 15. (a) V(/) and (b) the phase portrait of the system have been drawn
formation of the double layer. To understand this transition
on the same /-axis for M ¼ MNPDL þ 0.01. The separatrix of the lower panel process, we draw the saddle and other equilibrium points of
of this figure corresponds to a conventional soliton. system (44) on the /(¼/1)-axis for increasing values of M
033704-12 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

shows the transition from supersoliton to soliton structures

after the formation of a double layer.

VI. CONCLUSIONS
IA solitary structures have been investigated in a colli-
sionless magnetized dusty plasma system consisting of warm
adiabatic ions, static negatively charged dust grains and non-
thermal electrons. A complete set of equation of continuity
in three dimensions, the equation of motion in three dimen-
sions and the pressure equation in three dimensions for ion
fluids have been considered to describe nonlinear behaviour
of IA waves. Instead of the Poisson equation, the quasi-
neutrality condition is assumed to close the system of
equations.
A complete analytic theory is presented to find the upper
FIG. 16. / is plotted against n for (a) M ¼ MNPDL þ 0.0001 (solid black bounds Mpmax (Mnmax) of the Mach number M for the exis-
curve), (b) M ¼ MNPDL – 0.0001 (dashed black curve), (c) M ¼ MNPDL tence of all PPSWs (NPSWs). Again, we have also investi-
þ 0.01 (solid blue curve), and (d) M ¼ MNPDL – 0.01 (dashed blue curve).
gated an analytic theory to find the Mach numbers MPPDL
This figure clearly indicates the jump type discontinuity between the ampli-
tudes of solitons before and after the formation of the double layer. (MNPDL) corresponding to a PPDL (NPDL) solution of the
energy integral. With the help of these two analytic theories,
starting from M ¼ MNPDL þ 0.0001 in Fig. 17. This figure one can easily develop appropriate algorithms to determine
shows that the distance between the non-zero saddle and the Mpmax, Mnmax, MPPDL and MNPDL. These Mach numbers help
non-saddle fixed point nearest to it decreases for increasing us to draw the compositional parameter space which shows
values of M, and ultimately both of them disappear from the the nature of existence of different solitary structures with
system. Finally, the system contains only one saddle at the respect to any parameter of a system.
origin and a non-saddle equilibrium point. Consequently, We have seen that the description of solitary structures
only one separatrix enclosing the non-saddle fixed point is after the formation of the double layer of same polarity is
possible that appears to pass through the saddle at the origin. helpful for investigating the existence of supersoliton struc-
So, the existence of a soliton after the formation of a double tures of same polarity in comparison with the study of super-
layer confirms the existence of a sequence of supersolitons, soliton structures at some discrete points of the compositional
and there exists a critical value Mcr of M such that for the parameter spaces. The mechanism of transition from the
existence of supersolitons, we must have MNPDL < M < Mcr, NPSS to a negative potential soliton after the formation of a
whereas for Mcr  M < Mnmax, we get soliton like structures double layer of same polarity has been investigated through
after the formation of a double layer. Thus, Fig. 17 clearly the phase portrait analysis of the dynamical system describing
the nonlinear behaviour of IA waves. We have analyzed the
formation of supersoliton structures and their limitations with
the help of phase portraits. This is another important aspect of
this paper. In fact, this analysis determines the exact bound M
for the existence of NPSSs.
It is observed that the system supports PPSWs, NPSWs,
the coexistence of solitary waves of both polarities, NPDLs,
and NPSWs after the formation of NPDLs and NPSSs. The
qualitative behaviour of IA solitary structures is independent
of the intensity of the magnetic field because the Sagdeev
potential V(/) is free from the intensity of the magnetic field.
But, the qualitative behaviour of IA solitary structures is
very much sensible on the angle of propagation of the soli-
tary structures with the uniform static magnetic field. In fact,
V(/) depends on lz ¼ cos d, but does not depend on lx and ly,
where d is the angle between the direction of the solitary
structure and the direction of the external uniform static
magnetic field.
The region of existence of PPSWs increases for increas-
ing values of l, whereas the region of existence of NPSWs
FIG. 17. Saddle points (small solid circles) and equilibrium points other decreases for increasing values of l. The interval of be for
than saddle points (small solid stars) for the dynamical system (44) have
the existence of NPDLs increases for increasing values of l,
been drawn on the /-axis for different values of M ¼ MNPDL þ . This figure
clearly shows the transition from supersoliton structures to conventional if l  l(cr), whereas if l exceeds l(cr), the interval of be for
solitons. the existence of NPDLs decreases for increasing values of l,
033704-13 Debnath, Bandyopadhyay, and Das Phys. Plasmas 25, 033704 (2018)

where l(cr) is a critical value of l. The region of existence of 32

W. Feldman, R. Anderson, S. Bame, S. Gary, J. Gosling, D. McComas, M.
Thomsen, G. Paschmann, and M. Hoppe, J. Geophys. Res. 88, 96, https://
NPSWs before the formation of NPDLs increases for
doi.org/10.1029/JA088iA01p00096 (1983).
increasing values of l whenever l  l(cr), whereas the region 33
R. Bostr€om, G. Gustafsson, B. Holback, G. Holmgren, H. Koskinen, and
of existence of NPSWs before the formation of NPDLs P. Kintner, Phys. Rev. Lett. 61, 82 (1988).
34
decreases for increasing values of l for l > l(cr). R. Lundin, A. Zakharov, R. Pellinen, H. Borg, B. Hultqvist, N. Pissarenko,
E. Dubinin, S. Barabash, I. Liede, and H. Koskinen, Nature 341, 609
(1989).
ACKNOWLEDGMENTS 35
R. Bostr€om, IEEE Trans. Plasma Sci. 20, 756 (1992).
36
P. O. Dovner, A. I. Eriksson, R. Bostr€ om, and B. Holback, Geophys. Res.
The authors are grateful to the referee for extremely Lett. 21, 1827, https://doi.org/10.1029/94GL00886 (1994).
helpful comments, without which this paper could not have 37
H. Matsumoto, H. Kojima, T. Miyatake, Y. Omura, M. Okada, I. Nagano,
been written in its present form. One of the authors and M. Tsutsui, Geophys. Res. Lett. 21, 2915, https://doi.org/10.1029/
94GL01284 (1994).
(Debdatta Debnath) is thankful to the Department of Science 38
R. E. Ergun, C. W. Carlson, J. P. McFadden, F. S. Mozer, G. T. Delory,
and Technology, Government of India, INSPIRE Fellowship W. Peria, C. C. Chaston, M. Temerin, R. Elphic, R. Strangeway et al.,
Scheme, for financial support. 39
Geophys. Res. Lett. 25, 2025, https://doi.org/10.1029/98GL00635 (1998).
R. E. Ergun, C. W. Carlson, J. P. McFadden, F. S. Mozer, G. T. Delory,
1 W. Peria, C. C. Chaston, M. Temerin, R. Elphic, R. Strangeway et al.,
E. C. Whipple, T. G. Northrop, and D. A. Mendis, J. Geophys. Res. 90, Geophys. Res. Lett. 25, 2061, https://doi.org/10.1029/98GL00570 (1998).
7405, https://doi.org/10.1029/JA090iA08p07405 (1985). 40
G. T. Delory, R. E. Ergun, C. W. Carlson, L. Muschietti, C. C. Chaston,
2
C. K. Goertz, Rev. Geophys. 27, 271, https://doi.org/10.1029/RG027i002p00271 W. Peria, J. P. McFadden, and R. Strangeway, Geophys. Res. Lett. 25,
(1989). 2069, https://doi.org/10.1029/98GL00705 (1998).
3
N. N. Rao, P. K. Shukla, and M. Y. Yu, Planet. Space Sci. 38, 543 (1990). 41
J. R. Franz, P. M. Kintner, and J. S. Pickett, Geophys. Res. Lett. 25, 1277,
4
F. Verheest, Planet. Space Sci. 40, 1 (1992). https://doi.org/10.1029/98GL50870 (1998).
5
N. N. Rao, J. Plasma Phys. 53, 317 (1995). 42
C. A. Cattell, J. Dombeck, J. R. Wygant, M. K. Hudson, F. S. Mozer, M.
6
P. K. Shukla, Phys. Plasmas 8, 1791 (2001). A. Temerin, W. K. Peterson, C. A. Kletzing, C. T. Russell, and R. F. Pfaff,
7
C. R. Choi, C.-M. Ryu, N. C. Lee, and D.-Y. Lee, Phys. Plasmas 12, Geophys. Res. Lett. 26, 425, https://doi.org/10.1029/1998GL900304
022304 (2005). (1999).
8
R. L. Merlino, Plasma Phys. Appl. 81, 73 (2006). 43
R. Pottelette, R. E. Ergun, R. A. Treumann, M. Berthomier, C. W.
9
P. K. Shukla and V. P. Silin, Phys. Scr. 45, 508 (1992). Carlson, J. P. McFadden, and I. Roth, Geophys. Res. Lett. 26, 2629,
10
C. Yinhua and M. Y. Yu, Phys. Plasmas 1, 1868 (1994). https://doi.org/10.1029/1999GL900462 (1999).
11
C. R. Choi, C.-M. Ryu, N. C. Lee, D.-Y. Lee, and Y. Kim, Phys. Plasmas 44
F. Verheest, Waves in Dusty Space Plasmas (Kluwer Academic,
12, 072301 (2005). Dordrecht, Netherlands, 2000).
12
S. Maitra and R. Roychoudhury, Phys. Plasmas 13, 112302 (2006). 45
P. K. Shukla and A. A. Mamun, Introduction to Dusty Plasma Physics
13
S. Ghosh, S. Sarkar, M. Khan, and M. R. Gupta, Phys. Scr. 63, 395 (2001). (IoP, Bristol, UK, 2002).
14
P. K. Shukla and A. A. Mamun, New J. Phys. 5, 17 (2003). 46
Y. Futaana, S. Machida, Y. Saito, A. Matsuoka, and H. Hayakawa,
15
F. Verheest, T. Cattaert, and M. A. Hellberg, Phys. Plasmas 12, 082308 J. Geophys. Res. 108, 18729, https://doi.org/10.1029/2002JA009366 (2003).
(2005). 47
J. P. McFadden, C. W. Carlson, R. E. Ergun, F. S. Mozer, L. Muschietti, I.
16
C. R. Choi, C.-M. Ryu, D.-Y. Lee, N. C. Lee, and Y.-H. Kim, Phys. Roth, and E. Moebius, J. Geophys. Res. 108, 8018, https://doi.org/
Lett. A 364, 297 (2007). 10.1029/2002JA009485 (2003).
17
S. Mahmood and S. Hussain, Phys. Plasmas 14, 082303 (2007). 48
R. A. Cairns, A. A. Mamum, R. Bingham, R. Bostr€ om, R. O. Dendy, C.
18
P. Chatterjee, T. Saha, and C.-M. Ryu, Phys. Plasmas 15, 123702 (2008). M. C. Nairn, and P. K. Shukla, Geophys. Res. Lett. 22, 2709, https://
19
T. Saha and P. Chatterjee, Phys. Plasmas 16, 013707 (2009). doi.org/10.1029/95GL02781 (1995).
20
S. K. El-Labany, E. F. El-Shamy, and S. A. El-Warraki, Phys. Plasmas 16, 49
F. Verheest and S. R. Pillay, Phys. Plasmas 15, 013703 (2008).
013703 (2009). 50
R. Z. Sagdeev, in Reviews of Plasma Physics, edited by M. A. Leontovich
21
S. K. El-Labany, M. Shalaby, E. F. El-Shamy, and L. S. El-Sherif, Planet. (Consultant Bureau, New York, NY, 1966), Vol. 4.
Space Sci. 57, 1246 (2009). 51
A. Das, A. Bandyopadhyay, and K. P. Das, Phys. Plasmas 16, 073703
22
A. Das, A. Bandyopadhyay, and K. P. Das, J. Plasma Phys. 78, 149 (2012). (2009).
23 52
M. Shahmansouri and H. Alinejad, Phys. Plasmas 19, 123701 (2012). A. E. Dubinov and D. Y. Kolotkov, Plasma Phys. Rep. 38, 909 (2012).
24 53
M. Adnan, S. Mahmood, and A. Qamar, Contrib. Plasma Phys. 54, 724 F. Verheest, M. A. Hellberg, and I. Kourakis, Phys. Plasmas 20, 012302
(2014). (2013).
25 54
B. Sahu, A. Sinha, and R. Roychoudhury, Phys. Plasmas 21, 103701 (2014). F. Verheest, M. A. Hellberg, and I. Kourakis, Phys. Rev. E 87, 043107
26
M. Sayyar, H. Zahed, S. J. Pestehe, and S. Sobhanian, Phys. Plasmas 23, (2013).
073704 (2016). 55
F. Verheest, G. S. Lakhina, and M. A. Hellberg, Phys. Plasmas 21, 062303
27
S. A. El-Tantawy, Astrophys. Space Sci. 361, 249 (2016). (2014).
28 56
A. Paul and A. Bandyopadhyay, Astrophys. Space Sci. 361, 172 (2016). O. R. Rufai, R. Bharuthram, S. V. Singh, and G. S. Lakhina, Phys.
29
A. Paul, A. Das, and A. Bandyopadhyay, Plasma Phys. Rep. 43, 218 Plasmas 21, 082304 (2014).
57
(2017). O. R. Rufai, R. Bharuthram, S. V. Singh, and G. S. Lakhina, Phys.
30
A. Paul, A. Bandyopadhyay, and K. P. Das, Phys. Plasmas 24, 013707 Plasmas 22, 102305 (2015).
58
(2017). O. R. Rufai, R. Bharuthram, S. V. Singh, and G. S. Lakhina, Phys.
31
J. R. Asbridge, S. J. Bame, and I. B. Strong, J. Geophys. Res. 73, 5777, Plasmas 23, 032309 (2016).
59
https://doi.org/10.1029/JA073i017p05777 (1968). F. Verheest and M. A. Hellberg, Phys. Plasmas 24, 022306 (2017).