See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/372200883

The effect of single-doping engineering on ferroelectric properties of α-

In2Se2X: A density functional theory study

Article in Journal of Applied Physics · July 2023

DOI: 10.1063/5.0156566

CITATION READS

1 74

8 authors, including:

Gui-lin Zhuang
Anhui Normal University
257 PUBLICATIONS 8,427 CITATIONS

SEE PROFILE

All content following this page was uploaded by Gui-lin Zhuang on 29 December 2023.

The user has requested enhancement of the downloaded file.

RESEARCH ARTICLE | JULY 07 2023

The effect of single-doping engineering on ferroelectric

properties of α-In2Se2X: A density functional theory study 
Jun Wen ; Yi-Jie Chen; Fu-Li Sun; Cun-Biao Lin ; Xian-Jie Zhang; Jia-Hao Wang; Wen-Xian Chen;
Gui-Lin Zhuang 

Journal of Applied Physics 134, 014102 (2023)

https://doi.org/10.1063/5.0156566

CrossMark

 
View Export
Online Citation

07 July 2023 15:01:08

 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

The effect of single-doping engineering on

ferroelectric properties of α-In2Se2X: A density
functional theory study
Cite as: J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566
Submitted: 5 May 2023 · Accepted: 20 June 2023 · View Online Export Citation CrossMark
Published Online: 7 July 2023

Jun Wen, Yi-Jie Chen, Fu-Li Sun, Cun-Biao Lin, Xian-Jie Zhang, Jia-Hao Wang, Wen-Xian Chen,
and Gui-Lin Zhuanga)

AFFILIATIONS
H-PSI Computational Chemistry Lab, Institute of Industrial Catalysis, College of Chemical Engineering, Zhejiang University of
Technology, Hangzhou 310032, People’s Republic of China

a)
Author to whom correspondence should be addressed: [email protected]

ABSTRACT
Tailoring the specific properties to practical applications by structural modification is of vital importance for the envisioned development of
two-dimensional ferroelectric materials. Herein, a comprehensive investigation on the effects of single doping on the ferroelectric properties

07 July 2023 15:01:08

and electronic transport in a monolayer of α-In2Se3 was carried out via the combination of first-principles density functional theory calcula-
tions and electron–phonon coupling simulations. Our results show that single-doping in In2Se3 can reduce effective mass of carriers and
thereby enhance the high carrier mobility potential of the material. Moreover, the ferroelectric phonon mode in single-doped In2Se2X fea-
tures a lower scattering rate, associating with the single-doping atom, and indicates reduced hindrance to carrier transport during ferroelec-
tric switching. Compared to pristine In2Se3, the obtained smaller ferroelectric barriers (<1 eV) of single-doped ones promote more
promising ferroelectricity from the analysis of the ferroelectric soft mode. Interestingly, the observed variations in ferroelectric behaviors
resulting from doping of different elements highlight the significance of single-doping in modifying the ferroelectric properties of monolay-
ers. Furthermore, strain engineering results reveal that single doping obviously affects the dependence of gap on strains: linear relationship
for doping ones and nonlinearity for pristine one. Our study provides valuable insights into achieving higher carrier mobility in these criti-
cal materials.

Published under an exclusive license by AIP Publishing. https://doi.org/10.1063/5.0156566

I. INTRODUCTION the electric field with their surface-charge layers with a significant
enhancement of the charge separation rate. Furthermore, Vasić and
The latest decades had witnessed a surge of tremendous pro-
Gajić20 successfully doped Co atoms into an organic trimeric acid
gress regarding two-dimensional (2D) ferroelectric materials,
(TMA) ligand and magnetized this multifunctional molecule by
largely stimulated by envisioned promising applications, e.g.,
changing the substrate temperature, providing a foundation for the
memory storage devices, optoelectronics and energy storage, as well
further research of TMA in spintronics. By applying these
as several advantages over their 3D counterparts, including stable methods, improved performance can be achieved in applications
polarization and clean surfaces.1–13 Although significant progress such as optoelectronics,21 sensing,20 and energy storage.22,23
has been made in the synthesis of 2D ferroelectric materials, the Therefore, 2D ferroelectric materials have enormous potential for
rational synthesis of these materials still presents huge challenges future technological advancements, and defect engineering is a
due to limited suitable synthesis methods and material stability promising strategy for improving their properties and applications.
issues.14–17 To this end, various approaches, such as introducing Extensive research has demonstrated that structural doping
vacancies, doping atoms, modifying grain boundaries or edges, can can significantly impact the mechanical, electronic, magnetic, and
be systematically developed.18 Lu et al.19 employed oxygen plasma optical properties of ferroelectrics.24–26 For instance, structural
to tailor the defect density inside ZnO nanowire, thereby regulating doping is commonly used to adjust the effective mass (Ef ) values

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-1

Published under an exclusive license by AIP Publishing
 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

and drive insulators with wide bandgaps toward metallic states, cutoff of plane-wave basis was set to 400 eV, and the Γ-centered
thereby enabling multiferroic systems to exhibit both ferroelectricity 10 × 10 × 1 k-point mesh was used for Brillion zone sampling by the
and metallicity.27,28 In addition to its influence on electronic prop- convergence test. For self-consistent calculations, the convergence
erties, doping has also been found to induce ferromagnetism, a criterion for energy was set to 10−7 eV, and all the structures were
state in which polarization order coexists with magnetic order.29 relaxed by the conjugated gradient method until the interaction
Therefore, it is urgent and important to reveal how doping affects force each atom less than 0.02 eV/Å. In addition, more than 25 Å
the intrinsic properties of two-dimensional ferroelectric nanosheets of the vacuum region was introduced to prevent the interaction
and heterostructures. Furthermore, understanding the carrier scat- between layers due to the periodic condition. The advanced
tering mechanism in any material provides an excellent foundation hybrid functional HSE06 was used to evaluate the bandgap.42,43
for improving its transport properties, revealing new physical High symmetric points, involving M(0 1/2 0)-Γ(0 0 0)-K(−1/3 2/
effects, and ultimately applications. A typical example is the perov- 3 0)-M(0 1/2 0), were sampled in the first Brillouin zone for the
skite SrTiO3, which exhibits a widely studied 2D electron gas, band structure calculations. For the Wannier band structure inter-
superconductivity,30 soft phonon mode-induced phase transi- polation,44,45 the s orbitals and the p orbitals of all elements were
tion,31,32 etc. More interestingly, Li et al.33 showed that the local chosen for initial projections. The climbing image nudged elastic
microelectric field generated by the polarization of the ferroelectric band (CI-NEB) method was employed to calculate ferroelectric
body promotes the fast transport and diffusion of ions, thereby transition barriers.
improving the magnification performance of the anode. Recently, it Both lattice-dynamical properties and the electron–phonon
has been reported that the semiconductive and out-of-plane electric coupling (EPC) were performed by using the QUANTUM
polarization properties of α-In2Se3 monolayers lead to piezoelectric ESPRESSO (7.0) software.46 Optimized norm-conserving Vanderbilt
properties and good carrier separation.34,35 From rational design of pseudopotential from the PseudoDojo library were used to describe
viewpoint, the structural modification (e.g., doping) of α-In2Se3 core-valence interactions.47 The plane-wave energy and charge
monolayers is an effective tool to tailor their properties to specific density cutoffs were set to 80 and 320 Ry, respectively. For structural
applications and improve their stability and scalability. However, relaxations, a 10−9 Ry convergence threshold and a 16 × 16 × 1
the effect of doping on the ferroelectric behavior of α-In2Se3 has k-point grid were sufficient to achieve the computational accuracy.
been rarely reported until now. The minimum lattice parameter of α-In2Se3 monolayer (4.105 Å)
Herein, via the congener-element substitution strategy was obtained at an atomic pressure level less than 0.01 GP, in con-
involved in ferroelectricity-related atoms, we designed single- sistent with previous reports.48 By the means of density functional
doping monolayer In2Se2X (X = Se, S, Te) and focused on the effect perturbation theory (DFPT),49 the phonon dispersions were calcu-

07 July 2023 15:01:08

of doping atom on the ferroelectric properties. Computational lated on an 8 × 8 × 1 q-point grid. To obtain the relaxation time and
results show that the band structures of all three doping In2Se2X the scattering of electrons, the EPW code50 was used for the calcula-
(X = Se, S, Te) exhibit high anisotropic dispersion in the CBM with tion the electron–phonon coupling matrix elements. We initially
small effective masses, benefiting for the efficient transport of 2D performed the calculation on a 16 × 16 × 1 electron grid and an
semiconductor charge carriers. In comparison to In2Se3, the doped 8 × 8 × 1 phonon grid. Subsequently, the fine k and q grids of
variant In2Se2S retains its original ferroelectric behavior while 160 × 160 × 1 were interpolated by Wannier–Fourier technique.51
requiring a lower energy barrier to overcome. It is important to All these parameters were carefully tested to ensure the convergence
highlight that among the three-step path, In2Se2Te stands as the and accuracy. Then, the carrier scattering lifetime τnk associated
only exception, as its ferroelectric behavior is slightly compromised. with the electron–phonon interaction (EPI) for each electronic state
Moreover, both biaxial tensile strains can narrow their bandgaps, was computed as given by the standard Fermi golden rule:52
thus further enhancing electron transport. More in-depth carrier
scattering studies show that the mobility of all materials at room ! ð
temperature is limited by the lattice scattering of the transverse 2 X 2π X dq  2

1
τ nk ¼ Im ¼ gmn ðk; qÞ
optical transverse mode ZO mode out of the highest energy plane. h nk

h m ΩBZ
Surprisingly, the scattering rate of the ferroelectric optical trans- BZ
(
verse (TO) mode of In2Se2S is the smallest among the three materi-    
als investigated. The possibility of using methods similar to  nq þ f m; kþq δ Em; kþq þ hωq
En;k
homologous substitution to reduce lattice scattering is an intriguing )
direction for future work.    
þ 1 þ nq
f m; kþq δ Em; kþq
En;k
hωq ; (1)

II. COMPUTATIONAL METHODS

The calculations of geometrical optimization and electronic where Im(∑nk) denotes the imaginary part of electronic self-
properties were carried out utilizing the Vienna ab initio simulation energies, gmn(k,q) represents the element of the matrix element of
package (VASP 6.3.1).36,37 Perdew–Burke–Ernzerhof (PBE) func- the electron–phonon coupling, m and n are the electron band
tional under the generalized gradient approximation (GGA) was indices, k and q are the electronic wavevectors and phonon wavevec-
used to describe the exchange-correlation potential of electrons in tors, respectively, Enk and ωq represent the Kohn–Shame energy and
the Kohn–Shame equation.38 The van der Waals (vdW) interactions phonon frequency, respectively, fnk is the Fermi–Dirac occupation
were corrected by using the DFT-D3 (BJ) strategy.39–41 The energy function at equilibrium, while nq is also the distribution function of

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-2

Published under an exclusive license by AIP Publishing
 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

ωq, the two terms in the square brackets indicate the phonon emis- Ef (In2 Se2 X) ¼ E(In2 Se2 X)
E(In2 Se3 )
μ(X) þ μ(Se), (3)
sion and absorption.
where E(In2Se3) and E(In2Se2X) represent the total energy of In2Se3
III. RESULTS AND DISCUSSION monolayer and single-doping In2Se2X (X = S, Te) monolayers, E
(In) and E(Se) denote the total energy per unit of the In bulk, and
The bulk phase of the ferroelectric α-In2Se3, consisting of
Se molecular crystal, respectively. And, μ(X) and μ(Se) label the
individual five-member (Se–In–Se–In–Se) layers with weak van der
chemical potential for single-doping X (X = S, Te) and single Se
Waals interaction between each monolayer, can be easily exfoli-
atoms, respectively. The calculated formation energies of In2Se2X
ated.53 Now, based on the 2D α-In2Se3 geometry obtained by this
monolayers (−3.12 eV for Se, −2.93 eV for S, −2.41 eV for Te)
approach, In2Se2X (X = Se, S, Te) monolayers were formed through
demonstrate their thermodynamic stability.
the substitution of the middle Se2 atom. As a consequence, the
First, the electrical transport properties of In2Se2X monolayers
original periodicity is locally broken. The resulting optimized struc-
as a 2D ferroelectric material were investigated. As shown in
ture, depicted in Fig. 1(a), is associated with the R3m space group
Figs. 2(a)–2(c), the band structures and PDOS of isolated In2Se2X
(No. 160). It is evident to find that each subatomic layer of In2Se2X
monolayer comprises only a single element, where two inequivalent
In1 and In2 atoms reside in the hole sites of the Se atom’s upper
and lower layers, respectively. The middle X atom is vertically con-
nected to the In2 atom, leading to a tetrahedral structure.
Meanwhile, the oblique In1–X bond on the other side creates a dis-
torted octahedron around the In1 atom shown in Fig. 1(b).
In general, the initial point group C3v of In2Se3 remains
unchanged after doping, indicating that the room-temperature fer-
roelectric properties of In2Se3 are shared by In2Se2S and In2Se2Te.
The lattice parameters of these compounds show minor variations:
In2Se2Te has a slightly higher lattice parameter of 4.180 Å, while
In2Se2S has a slightly reduced value of 4.061 Å compared to
In2Se30 s lattice parameter of 4.105 Å. Such lattice parameters are
almost insensitive to the type of X atom, and have a negligible

07 July 2023 15:01:08

impact on some inherent properties of In2Se3. Additionally, in
order to investigate the structural stability, the formation energies
of In2Se2X (X = Se, S, or Te) monolayers were calculated using the
following formulas:54,55

Ef (In2 Se3 ) ¼ E(In2 Se3 )
2E(In)
3E(Se), (2)

FIG. 1. (a) Top and side views of In2Se2X (X = Se, S, Te) monolayers. The
black solid lines outline the trigonal unit cell. Purple and cyan balls represent FIG. 2. (a)–(c) Electronic structures and partial density of states (PDOS) of
two inequivalent In atoms, while the yellow and olive ones denote surface Se In2Se2X (X = Se, S, Te) monolayers calculated with the HSE06 functional. The
atoms, X atoms, respectively. (b) The distorted octahedron of the In1 atoms and red and blue lines represent CBM with In-s orbitals and VBM with surface Se-p
the tetrahedron of the In2 atoms. (c) Illustration of the reciprocal space path for orbitals, respectively. (d) and (e) Zoomed-in schematic of the partial density of
In2Se2X (X = Se, S, Te) monolayers employed in the calculations of electron states (PDOS) for In2Se2X (X = Se, S, Te) monolayers. The image shows the
and phonon band structures. perfect overlap between the px and py orbitals.

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-3

Published under an exclusive license by AIP Publishing
 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

monolayers by the HSE-06 hybrid functional showed strong quali- level than those of Se, stemming from its stronger electron affinity.
tative similarities, especially in the case of doping with S and Se. Finally, the strong coupling between the In-p orbitals and the
Such similarity can also be observed in electronic structures with surface Se-p orbitals induces an evident separation between their
the PBE functional, as shown in Fig. S1 in the supplementary mate- bonding and antibonding states, however, is suppressed by the Te-p
rial. The semiconductors doped with different chalcogenides have orbitals.
bandgaps of 1.27, 1.44, and 0.58 eV, respectively, without altering When subjected to in-plane biaxial tensile strain, the interac-
the symmetry of their structures or the hexagonal Bravais shape. tion between orbitals is significantly weakened, thereby accomplish-
This consistency is reflected in their corresponding reciprocal space ing the decline of the bandgap. As illustrated in Fig. 3, the gradient
paths, as shown in Fig. 1(c). The In2Se2Te monolayer features an of In2Se2S begins to increase beyond a threshold of 2% strain,
indirect gap with the minimum conduction band (CBM) located at implying the obvious transition from a direct bandgap to an indi-
the Γ point and the maximum valence band (VBM) located along rect bandgap. Such a variation renders the bandgap more sensitive
the M-Γ direction, similar to In2Se3. On the other hand, In2Se2S to tensile strain, thus the enhancement of transport efficiency can
monolayer is a direct-gap semiconductor with both the VBM and be predicted. With respect to In2Se3 and In2Se2Te, the gap
CBM located at the Γ point. Interestingly, all three materials exhibit decreases without any obvious change. Interestingly, differing from
pronounced energy dispersion characteristics near the CBM, sug- the biaxial strain, the feasibility of modulating the bandgap by uni-
gesting the efficient transport of nearly free electrons. However, axial strain is inferior.57 Since the p orbitals at the VBM are con-
their bands near VBM are too flat to make it favorable for hole tributed by the perfectly overlapping px and py states, as shown in
transfer. The electronic structures of these monolayers are highly Figs. 2(d)–2(f ), there is a robust platform of modulation for biaxial
anisotropic due to their distinct bonding properties along different strain, analogous to the previous studies.58 Specifically, we further
directions near the CBM.56 analyze the structural parameters under different strains given in
To quantify previous phenomenon, the effective masses for Tables S2 and S3 in the supplementary material. Under biaxial
three materials was identified as presented in Table I. Specifically, tensile strain, the increased bond length of In-Se (surface) weakens
for In2Se2Te, its electron has a relatively low effective mass of the coupling between In-p and surface Se-p orbitals, thus reducing
approximately 0.19 times the mass of a free electron (m0). the splitting of bonding and antibonding states. For uniaxial strain,
However, for In2Se2S and In2Se3, the electron is slightly heavierwith even when the tensile strain reaches 6%, the In-Se (surface) dis-
an effective mass of around 0.21m0. These findings suggest that the tances are only modified slightly, implying that the coupling
effective mass of electrons in these materials is similar, with only strength hardly altered.
slight variations between them. More detailed analysis of the band Furthermore, we utilized the maximum localized Wannier

07 July 2023 15:01:08

structures indicates that the CBM at the Γ point has a comparable functions (MLWF) approximation59 to derive a tight-binding (TB)
mass tensor with the symmetry of a small group of wave vector at model of In2Se2X monolayers as depicted in Fig. 4. The accuracy of
the extremum. Inspection of the partial density of state (PDOS) of the model was verified by comparing it to the band structure
In2Se2X monolayers reveals following conclusions. First, the CBM obtained from fully density functional theory (DFT) calculations,
is mainly contributed by the In-s states, despite the Se-p orbitals of
In2Se3 nearly coinciding with the In-s orbitals, while the surface
Se-p orbitals dominate the VBM for In2Se3 and In2Se2S. Second,
the Te-p orbitals in the VBM are found to be closer to the Fermi

TABLE I. Bandgaps EPBE HSE

g (eV) from PBE calculations, HSE bandgaps Eg (eV)
from HSE-06 calculations, effective masses m* (m0) along different direction for
In2Se2X (X = Se, S, Te) monolayers, where the Λ point and the p point represent
the VBM of In2Se3 and In2Se2Te, respectively.

Structures Band type EPBE

g EHSE
g Carrier type Direction m*
In2Se3 Indirect 0.79 1.27 Electron Λ-Γ 0.21
Λ-M 0.21
Hole Λ-Γ 2.52
Λ-M 3.42
In2Se2S Direct 0.89 1.48 Electron Γ-M 0.21
Γ-K 0.21
Hole Γ-M 2.15
Γ-K 2.44
In2Se2Te Indirect 0.07 0.56 Electron P-Γ 0.19
P-M 0.19
Hole P-Γ 1.41 FIG. 3. Bandgap variation of the In2Se2X (X = Se, S, Te) monolayers under the
P-M 2.83 biaxial strain calculated using HSE functional calculations.

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-4

Published under an exclusive license by AIP Publishing
 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

FIG. 4. Comparison of the DFT band structures (red lines) and those from the
Wannierization (blue dashed lines) of In2Se2X (X = Se, S, Te) monolayers. The
energy axis is shown relative to the Fermi level (Ef ) defined as the reference
energy level with zero potential energy.

and it showed an almost perfect match within the internal energy

window. Specifically, the description of the states around the Fermi
level demonstrated the precision of the MLWF-TB method. To
simplify the process, the disentanglement step was skipped by
using relevant parameters in Wannier90 to ignore some bands. The
essential conduction band at the Γ point was unambiguously
described, despite a relatively large Wannier center and spread that
occurred in all projected orbitals, as shown in Tables S4–S6 in the FIG. 5. Ferroelectric switching paths and barriers of In2Se2X (X = Se, S, Te)
supplementary material. Although such high spread might suggest monolayers. (a) and (b) Barriers and switching path of path-1. In path-1, only X
atoms move. (c) and (d) Barriers and switching path of path-2. In path-2, the

07 July 2023 15:01:08

a failure in Wannierization, it was not likely to impact the bands upper-layer Se-In-X atoms move first, then the central X atoms rotate around
near the Fermi level. The study provided an accurate electronic the centrosymmetric position by 60°.
structure for the entire Brillouin zone that was consistent with ab
initio simulations. Additionally, the researchers used appropriate
initial projections and energy windows to facilitate the subsequent
simulation of EPI using the Wannier-based EPW code. According to previous reports, the ideal range of switching barrier
The stability of ferroelectricity in In2Se2X monolayers is for ferroelectric material should approximately range from 0.1 to
mainly determined by their unique covalent bonding structure.35 1 eV, rendering negligible probability of quantum tunneling.62
The generation of in-plane and out-of-plane polarization is set The NEB method is used to compute the ferroelectric switching
apart from that in traditional ferroelectric materials with the long- barrier, which is desirable within a range of only 1 eV, as shown in
range Coulomb interactions. To eliminate or even reverse the Fig. 5(a). Such a modest switching barrier is high enough against
out-of-plane dipole, the middle X atoms must shift laterally, thermal fluctuations while low enough to be reversed by a small
accompanied by the breaking and healing of the In-X covalent voltage, giving rise to a favorable Curie temperature. The initial
bond. As alluded to before, both the In1 and In2 atom have a In2Se3 has a switching barrier of 0.91 eV, consistent with previous
unique local coordination. Thus, the significant difference in the studies.34 However, In2Se2S has a higher barrier of 0.82 eV, while
interlayer spacing between the middle X layer and the two In layers In2Se2Te has a lower barrier of 0.66 eV. The unstable paraelectric
effectively break the inversion symmetry. Moderate charge compen- phase serves as an intermediate state, with the X layer atoms in a
sation also lays a potential foundation for the emergence of sponta- centrosymmetric position. After accounting for the temperature
neous out-of-plane polarization.60 Figure 5(b) plotted the effect and zero-point energy correction, the energy barriers
ferroelectric switching pathway of In2Se2X monolayers in detail. remained below 1 eV and the original order of energy barriers was
Achieving a positive polarization in the In2Se2X ferroelectric struc- preserved (see Table S1 in the supplementary material). In contrast,
ture requires aligning the middle X layer atoms perpendicular to systems that completely replace X atoms have relatively higher bar-
the lower In2 layer. Conversely, aligning them with the upper In1 riers, confirming the feasibility of single-doping (as shown in
layer results in a negative polarization. The existence of a robust Fig. S3 in the supplementary material). Considering the existence
dipole indicates the energetically preferred ground states consist of of the metastable structure [see Fig. 5(d)], an alternative ferroelec-
two states with opposite polarization.61 Two switching paths were tric switching path (path-2) involving the synergistic movement of
taken into account. First, the ferroelectric switching process the upper-layer Se-In-X atoms and the 60° rotation of the central-
(path-1) can be simplified as the exchange of positions between the layer X atoms, was considered as shown in Fig. 5(c). After doping
octahedra of the upper In2 and tetrahedra of the lower In1. single S atom, the ferroelectric barrier was obviously reduced, while

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-5

Published under an exclusive license by AIP Publishing
 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

the material still maintained its ferroelectric behavior. However,

after the introduction of single Te atom, the middle well drops to a
position lower than the ferroelectric state, suggesting that the origi-
nal ferroelectric properties may be altered. Electronically, the M1
state of In2Se2Te with a modest bandgap of 0.70 eV was easier to
obtain, while both the M1 state of In2Se2S and In2Se3 were metallic
and unstable. Nevertheless, it should be noted that manipulating
the ferroelectric domain of In2Se2Te by an electric field may lead to
some unexpected behaviors, such as the emergence of dead layer.
Hence, in order to reverse through readily accessible kinetic path-
ways, path-2 should be more suitable. Moreover, it is not difficult
to find that the non-metallic paraelectric phase of In2Se2Te is more
interesting to the related research of polarization.
Furthermore, the electric-phonon interaction (EPI) in In2Se2X
monolayers was identified. The phonon dispersions along the high-
symmetry path in the Brillouin zone reveal the dynamic stability of
the In2Se2X monolayers, as shown in Figs. 6(c)–6(e). The emer-
gence of a few imaginary frequencies near the Γ point involves the
out-of-plane acoustic transverse (ZA) mode peculiar to 2D materi-
als. It is noteworthy that each monolayer has 15 vibrational modes,
but we only discuss some specific relevant phonons. The maximum
frequencies for the three spectra are 7.3, 10.7, and 6.0 THz, respec-
tively, all corresponding to the ZO mode with the A1 symmetry.
The only difference between the phonons is their hardness due to
the different mass of the doped atoms. Furthermore, the
longitudinal-optical (LO) and transverse-optical (TO) modes with
E symmetry have multiple degeneracies near the center of the
Brillouin zone, associated with the R3m space group of the crystal.

07 July 2023 15:01:08

The ferroelectricity of the three different structures is induced by
the phonon modes shown in Fig. 6(a), all of which belong to the
TO mode with E symmetry. The ferroelectric mode frequency of
In2Se3 is about 2.5 THz, but higher than In2Se2Te, 1.9 THz. In
terms of In2Se2S, this mode blue-shifts to 3.4 THz, signaling that
In2Se2S is more likely to have a reliable ferroelectric reversal at
higher temperatures. The phonon spectra of paraelectric phases
were investigated as shown in Fig. S7 in the supplementary mate-
rial. Specially, by combination of the analysis of vibration modes, FIG. 6. (a) and (b) Schematics of vibrational modes at points marked in (c)–(e)
Phonon dispersions of In2Se2X (X = Se, S, Te) monolayers. The blue arrows
the soft optical modes below zero frequency indicate the instability
represent the vibration direction of atoms in In2Se2X. (f )–(h) Mode resolved
of X atoms in the paraelectric states, which intrinsically drives electron-phonon scattering rate of the conducting electron of In2Se2X (X = Se,
spontaneous symmetry breaking to achieve the ferroelectric prop- S, Te) monolayers. The energy zero is set at the CBM.
erty after non-overlap of positive and negative charges.
Moreover, the effect of phonon scattering on carriers was eval-
uated. To this end, based on Fröhlich EPC matrix elements,63 the
band structures mapping the electron relaxation time are presented Generally speaking, at room temperature, electron mobility far
in Figs. 7(a)–7(c). When Te is doped, the maximum relaxation from the CBM remains largely unaffected by their scattering rate,64
time (c.a. 6.7 × 10−13 s) in the CBM is slightly ahead of the other as long as the energy is above 100 meV. This observation is sup-
twos. This is mainly related to the fact that the electrons in the ported by the distribution of relaxation time. On the basis of
CBM of In2Se2Te are lighter so that it preserves a larger group Fermi’s golden rule in quantum mechanics,52 eigenstates can trans-
velocity, rather than the magnitude of EPC matrix elements at form with each other. Similarly, the relaxation time mapped onto
CBM. According to the temperature-dependent scattering rates band structures can be calculated as the scattering rates, while the
shown in Figs. 7(d)–7(f ), it is apparent that the conclusions can be energy can be converted into the density of states (DOS).
drawn for all three structures: the scattering of charge carriers grad- Therefore, we can assume that the scattering rate (1/τnk) is propor-
ually raises with the increasing temperature, due to the thermal tional to the DOS. Taking the few occupied electronic states near
vibrations of the lattice itself. The scattering rate sharply increases the CBM as an example, the scattering rate around here is relatively
at 60–70 meV, indicating the starting of the relaxation process low. This fact sufficiently confirms that the phase space of scatter-
when carriers are transitioning to another drift velocity. ing is predominantly determined by the DOS. In addition, we
Meanwhile, this implies that they will suffer stronger scattering. plotted the spatial decay of the Hamiltonian and the dynamical

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-6

Published under an exclusive license by AIP Publishing
 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

vibrational modes were depicted in Fig. S6 in the supplementary

material. This choice is a key prerequisite to achieve the goal of
intuitively visualizing the different effects of the dominant phonon
modes on the EPC. It should be clarified that among the 15 modes,
there are multiple phonon modes of the same type. Thus, for better
identification, the TO mode that can induce the ferroelectric phase
transition is called the TO2 mode, with the subscript 2 representing
the second TO mode. As regards In2Se3, the TO2 mode covers
roughly the same scattering rate range as the dominant ZO mode,
with a maximum around 30 ps−1. Note that the scattering rate
range of the TO3 mode with E symmetry is much smaller than the
TO2 mode. After Te doping, the scattering of the TO2 mode
becomes modest, featuring a narrower distribution in the range of
0–22 ps−1 compared to pure In2Se3. Unexpectedly, the scattering
rate of the TO2 mode in In2Se2S is reduced to below 20 ps−1, exhib-
iting a smaller distribution range than both In2Se3 and In2Se2Te.
Moreover, the finding shown in Fig. 6(b) firmly establishes the
out-of-plane relative stretching of the In–X bond as the main
source of the largest scattering contribution.
To verify the phenomenon of reduced carrier scattering by
single-atom doping, the analyses were conducted on the raw data.
In traditional semiconductors, the conduction band typically con-
tains s-like orbital. This symmetry can lead to a small or near-zero
scattering matrix element of long-wavelength optical transverse
phonons. In other words,65,66 the higher the symmetry of the
s-orbitals in the conduction band, the less significant the role of
TO phonons in electron scattering. We excluded the possibility that
the symmetry of p-orbitals at the VBM interacts with the holes,

07 July 2023 15:01:08

since the holes themselves are already heavy.67 Referring back to
the previously obtained PDOS, it is clear that the s-orbitals at the
FIG. 7. (a–c) Electron relaxation time mapping on the band structures of CBM of In2Se2S have relatively high symmetry. In contrast, the
In2Se2X (X = Se, S, Te) monolayers. The maximum relaxation time in the CBM s-orbitals of the other two monolayers either overlap with the
of In2Se3, In2Se2S, In2Se2Te is 4.74 × 10−13 s, 3.59 × 10−13 s, 6.72 × 10−13 s, p-orbitals or exhibit bad-symmetric distributions. This may explain
respectively. The CBM is set to zero. (d)-(e) Temperature resolved why the ferroelectric phonon modes of In2Se2S insignificantly limit
electron-phonon scattering rate of the conducting electron of In2Se2X (X = Se, carrier transport.
S, Te) monolayers. The energy zero is set at the CBM.

IV. CONCLUSIONS
matrix elements during the process of Wannier-interpolation in In this study, we conducted a systematic investigation of the
Figs. S4 in the supplementary material and S5 to verify the conver- impact of single doping on the ferroelectric properties and elec-
gence of denser q grids. tronic transport in α-In2Se3 monolayer by the combination of first-
To gain a deeper insight into the mechanism of scattering, principles DFT calculations and the EPI simulations. It is demon-
the phonon mode-resolved scattering rate is described in strated that the doped In2Se2X retains its high carrier mobility
Figs. 6(f )–6(h). The contribution of each phonon mode to trans- potential as evidenced by the calculated electronic structure and
port is calculated by Eq. (1). However, we did not present the scat- effective mass. Moreover, strain engineering profiles reveals that
tering rates for all 15 phonon modes of each structure, since any single-doping effectively modifies band structures to further
individual phonon modes exhibit a scattering rate one order of enhance carrier transport properties. The achieved small ferroelec-
magnitude lower than the total scattering rate at a given tempera- tric barriers (<1 eV) indicate that In2Se2S still hold promise as non-
ture. Although the ZO mode with the A1 symmetry in each struc- volatile memory candidates. Notably, in comparison to In2Se3,
ture has a relatively high scattering rate at low energies (below In2Se2Te achieves more stable intermediate non-ferroelectric state.
60 meV), it does not significantly scatter electrons due to the lack Additionally, the low scattering rate of the ferroelectric phonon
of scattering phase space near the CBM. As a result, it makes negli- mode from In2Se2S suggests a reduced hindrance to carrier trans-
gible contributions to carrier transport. port during ferroelectric switching. This observation, along with
We further retained the vibration modes with the highest or the understanding of how doping affects carrier transport obtained
lowest contributions in each structure, as well as the scattering dis- in this study, provides valuable insights for achieving higher carrier
tribution of the most representative ferroelectric TO mode. Other mobility in these critical materials.

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-7

Published under an exclusive license by AIP Publishing
 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

8
SUPPLEMENTARY MATERIAL J. Igo, S. Zhou, Z.-G. Yu, O. P. Amnuayphol, F. Zhao, and Y. Gu, J. Chem.
Phys. C 122(40), 22849 (2018).
See the supplementary material for the band structures calcu- 9
P. Hohenberg and W. Kohn, Phys. Rev. 136(3B), B864 (1964).
lated with the PBE, the bandgap variation under the uniaxial strain, 10
V. Barone and J. E. Peralta, Nano Lett. 8(8), 2210 (2008).
the ferroelectric switching barriers of In2X3 (X = Se, S, Te) mono- 11
A. R. Botello-Méndez, F. López-Urías, M. Terrones, and H. Terrones,
layers, the spatial decay of Hamiltonian and dynamical matrix ele- Nanotechnology 20(32), 325703 (2009).
12
ments, the other vibrational modes, the zero-pint energy correction S. Lebègue and O. Eriksson, Phys. Rev. B 79(11), 115409 (2009).
13
data, and the structure parameters under different types of strain. L. Ju, X. Tan, X. Mao, Y. Gu, S. Smith, A. Du, Z. Chen, C. Chen, and L. Kou,
Nat. Commun. 12(1), 5128 (2021).
14
J. Junquera and P. Ghosez, Nature 422(6931), 506 (2003).
ACKNOWLEDGMENTS 15
Z. Guan, H. Hu, X. Shen, P. Xiang, N. Zhong, J. Chu, and C. Duan,
This work was supported by the National Natural Science Adv. Electron. Mater. 6(1), 1900818 (2020).
16
Foundation of China (NNSFC) (Grant Nos. 22022108, 22072135, A. Belianinov, Q. He, A. Dziaugys, P. Maksymovych, E. Eliseev, A. Borisevich,
A. Morozovska, J. Banys, Y. Vysochanskii, and S. V. Kalinin, Nano Lett. 15(6),
and 21671172), the Zhejiang Provincial Natural Science
3808 (2015).
Foundation of China (Grant No. LR19B010001), and the Natural 17
H. Wang, Z. R. Liu, H. Y. Yoong, T. R. Paudel, J. X. Xiao, R. Guo, W. N. Lin,
Science Foundation of Shandong Province (No. ZR2020ZD35). P. Yang, J. Wang, G. M. Chow, T. Venkatesan, E. Y. Tsymbal, H. Tian, and
J. S. Chen, Nat. Commun. 9(1), 3319 (2018).
18
AUTHOR DECLARATIONS Z. Hu, Z. Wu, C. Han, J. He, Z. Ni, and W. Chen, Chem. Soc. Rev. 47(9),
3100 (2018).
Conflict of Interest 19
M.-P. Lu, C.-W. Chen, and M.-Y. Lu, Phys. Rev. Appl. 6(5), 054018 (2016).
20
The authors have no conflicts to disclose. B. Vasić and R. Gajić, Phys. Rev. Appl. 4(2), 024007 (2015).
21
F. Tong, X. S. Miao, Y. Wu, Z. P. Chen, H. Tong, and X. M. Cheng,
Appl. Phys. Lett. 97(26), 261904 (2010).
Author Contributions 22
Y. Zhang, W. Li, W. Cao, Y. Feng, Y. Qiao, T. Zhang, and W. Fei, Appl. Phys.
Lett. 110(24), 243901 (2017).
Jun Wen: Data curation (equal); Formal analysis (equal); 23
X. Gao, W. Ouyang, O. Hod, and M. Urbakh, Phys. Rev. B 103(4), 045418
Investigation (equal); Validation (equal); Writing – original draft (2021).
(equal); Writing – review & editing (equal). Yi-Jie Chen: Formal 24
Z. Cheng, X. Wang, S. Dou, H. Kimura, and K. Ozawa, Phys. Rev. B 77(9),
analysis (equal); Validation (equal); Writing – review & editing 092101 (2008).
(equal). Fu-Li Sun: Validation (equal); Writing – review & 25
J. Li, J. Lv, D. Zhang, L. Zhang, X. Hao, M. Wu, B.-X. Xu, M. Otonicar,
editing (equal). Cun-Biao Lin: Validation (equal); Writing –

07 July 2023 15:01:08

T. Lookman, B. Dkhil, and X. Lou, Phys. Rev. Appl. 16(1), 014033
review & editing (equal). Xian-Jie Zhang: Validation (equal). (2021).
26
Jia-Hao Wang: Formal analysis (equal); Validation (equal); Z. Liu, X. Wang, X. Ma, Y. Yang, and D. Wu, Appl. Phys. Lett. 122(12),
Writing – review & editing (equal). Wen-Xian Chen: Validation 122901 (2023).
27
(equal); Writing – original draft (equal). Gui-Lin Zhuang: J.-x. Gu, K.-j. Jin, C. Ma, Q.-h. Zhang, L. Gu, C. Ge, J.-s. Wang, C. Wang,
H.-z. Guo, and G.-z. Yang, Phys. Rev. B 96(16), 165206 (2017).
Conceptualization (equal); Formal analysis (equal); Funding 28
C. Ke, J. Huang, and S. Liu, Mater. Horiz. 8(12), 3387 (2021).
acquisition (equal); Project administration (equal); Supervision 29
C. Huang, Y. Du, H. Wu, H. Xiang, K. Deng, and E. Kan, Phys. Rev. Lett.
(equal); Validation (equal); Visualization (equal); Writing – 120(14), 147601 (2018).
review & editing (equal). 30
J. Appel, Phys. Rev. 180(2), 508 (1969).
31
A. Verma, S. Raghavan, S. Stemmer, and D. Jena, Appl. Phys. Lett. 107(19),
DATA AVAILABILITY 192908 (2015).
32
R. A. Evarestov, E. Blokhin, D. Gryaznov, E. A. Kotomin, and J. Maier,
The data that support the findings of this study are available Phys. Rev. B 83(13), 134108 (2011).
33
from the corresponding author upon reasonable request. R. Li, G. Zhang, Y. Wang, Z. Lin, C. He, Y. Li, X. Ren, P. Zhang, and H. Mi,
Nano Energy 90, 106591 (2021).
34
W. Ding, J. Zhu, Z. Wang, Y. Gao, D. Xiao, Y. Gu, Z. Zhang, and W. Zhu,
REFERENCES
Nat. Commun. 8, 14956 (2017).
1 35
R. C. G. Naber, C. Tanase, P. W. M. Blom, G. H. Gelinck, A. W. Marsman, J. Xiao, H. Zhu, Y. Wang, W. Feng, Y. Hu, A. Dasgupta, Y. Han, Y. Wang,
F. J. Touwslager, S. Setayesh, and D. M. de Leeuw, Nat. Mater. 4(3), 243 (2005). D. A. Muller, L. W. Martin, P. Hu, and X. Zhang, Phys. Rev. Lett. 120(22),
2
A. Q. Jiang, C. Wang, K. J. Jin, X. B. Liu, J. F. Scott, C. S. Hwang, T. A. Tang, 227601 (2018).
36
H. B. Lu, and G. Z. Yang, Adv. Mater. 23(10), 1277 (2011). G. Kresse and J. Hafner, Phys. Rev. B 47(1), 558 (1993).
3 37
R. C. G. Naber, K. Asadi, P. W. M. Blom, D. M. de Leeuw, and B. de Boer, G. Kresse and J. Hafner, Phys. Rev. B 49(20), 14251 (1994).
38
Adv. Mater. 22(9), 933 (2010). J. P. Perdew, K. Burke, and M. Ernzerhof, Phys. Rev. Lett. 77(18), 3865 (1996).
4 39
J. F. Scott, Science 315(5814), 954 (2007). S. Grimme, J. Comput. Chem. 27(15), 1787 (2006).
5 40
D. D. Fong, G. Brian Stephenson, S. K. Streiffer, J. A. Eastman, O. Auciello, S. Grimme, J. Antony, S. Ehrlich, and H. Krieg, J. Chem. Phys. 132(15),
P. H. Fuoss, and C. Thompson, Science 304(5677), 1650 (2004). 154104 (2010).
6 41
M. Si, P.-Y. Liao, G. Qiu, Y. Duan, and P. D. Ye, ACS Nano 12(7), 6700 (2018). S. Grimme, S. Ehrlich, and L. Goerigk, J. Comput. Chem. 32(7), 1456 (2011).
7 42
J. Jiang, L. Zhang, C. Ming, H. Zhou, P. Bose, Y. Guo, Y. Hu, B. Wang, J. Heyd, G. E. Scuseria, and M. Ernzerhof, J. Chem. Phys. 118(18), 8207
Z. Chen, R. Jia, S. Pendse, Y. Xiang, Y. Xia, Z. Lu, X. Wen, Y. Cai, C. Sun, (2003).
43
G. C. Wang, T. M. Lu, D. Gall, Y. Y. Sun, N. Koratkar, E. Fohtung, Y. Shi, and A. V. Krukau, O. A. Vydrov, A. F. Izmaylov, and G. E. Scuseria, J. Chem.
J. Shi, Nature 607(7919), 480 (2022). Phys. 125(22), 224106 (2006).

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-8

Published under an exclusive license by AIP Publishing
 Journal of
ARTICLE pubs.aip.org/aip/jap
Applied Physics

44 54
A. A. Mostofi, J. R. Yates, G. Pizzi, Y.-S. Lee, I. Souza, D. Vanderbilt, and C. Tang, L. Zhang, D. Wijethunge, K. K. Ostrikov, and A. Du, J. Chem. Phys. C
N. Marzari, Comput. Phys. Commun. 185(8), 2309 (2014). 125(44), 24648 (2021).
45 55
I. Souza, N. Marzari, and D. Vanderbilt, Phys. Rev. B 65(3), 035109 (2001). S.-J. Lv, G.-X. Yin, H.-L. Cui, and H.-Y. Wang, Solid State Commun. 325,
46
P. Giannozzi, O. Andreussi, T. Brumme, O. Bunau, M. Buongiorno Nardelli, 114159 (2021).
56
M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, M. Cococcioni, N. Colonna, L. Hu and X. Huang, RSC Adv. 7(87), 55034 (2017).
57
I. Carnimeo, A. Dal Corso, S. de Gironcoli, P. Delugas, R. A. DiStasio, Jr., C.-H. Chang, X. Fan, S.-H. Lin, and J.-L. Kuo, Phys. Rev. B 88(19), 195420
A. Ferretti, A. Floris, G. Fratesi, G. Fugallo, R. Gebauer, U. Gerstmann, (2013).
58
F. Giustino, T. Gorni, J. Jia, M. Kawamura, H. Y. Ko, A. Kokalj, E. Küçükbenli, X. Jiang, Y. Feng, K. Q. Chen, and L. M. Tang, J. Phys.: Condens. Matter
M. Lazzeri, M. Marsili, N. Marzari, F. Mauri, N. L. Nguyen, H. V. Nguyen, 32(10), 105501 (2020).
59
A. Otero-de-la-Roza, L. Paulatto, S. Poncé, D. Rocca, R. Sabatini, B. Santra, A. A. Mostofi, J. R. Yates, Y.-S. Lee, I. Souza, D. Vanderbilt, and N. Marzari,
M. Schlipf, A. P. Seitsonen, A. Smogunov, I. Timrov, T. Thonhauser, P. Umari, Comput. Phys. Commun. 178(9), 685 (2008).
60
N. Vast, X. Wu, and S. Baroni, J. Phys.: Condens. Matter 29(46), 465901 (2017). I. Efe, N. A. Spaldin, and C. Gattinoni, J. Chem. Phys. 154(2), 024702 (2021).
47 61
M. J. van Setten, M. Giantomassi, E. Bousquet, M. J. Verstraete, Y.-T. Huang, N.-K. Chen, Z.-Z. Li, X.-P. Wang, H.-B. Sun, S. Zhang, and
D. R. Hamann, X. Gonze, and G. M. Rignanese, Comput. Phys. Commun. 226, X.-B. Li, InfoMat 4(8), e12341 (2022).
62
39 (2018). Q. Yang, T. Zhong, Z. Tu, L. Zhu, M. Wu, and X. C. Zeng, Adv. Sci. 6(1),
48
M. Küpers, P. M. Konze, A. Meledin, J. Mayer, U. Englert, M. Wuttig, and 1801572 (2019).
63
R. Dronskowski, Inorg. Chem. 57(18), 11775 (2018). C. Verdi and F. Giustino, Phys. Rev. Lett. 115(17), 176401 (2015).
49 64
S. Baroni, S. de Gironcoli, A. Dal Corso, and P. Giannozzi, Rev. Mod. Phys. Z. Zhu, X. Yao, S. Zhao, X. Lin, and W. Li, J. Am. Chem. Soc. 144(10), 4541
73(2), 515 (2001). (2022).
50 65
S. Poncé, E. R. Margine, C. Verdi, and F. Giustino, Comput. Phys. Commun. C. Hamaguchi, Basic Semiconductor Physics, 2nd ed. (Springer, New York, 2010).
66
209, 116 (2016). P. Y. Yu and M. Cardona, Fundamentals of Semiconductors, 4th ed. (Springer,
51
F. Giustino, M. L. Cohen, and S. G. Louie, Phys. Rev. B 76(16), 165108 (2007). New York, 2010).
52 67
F. Giustino, Rev. Mod. Phys. 89(1), 015003 (2017). A. Verma, A. P. Kajdos, T. A. Cain, S. Stemmer, and D. Jena, Phys. Rev. Lett.
53
X. Tao and Y. Gu, Nano Lett. 13(8), 3501 (2013). 112(21), 216601 (2014).

07 July 2023 15:01:08

J. Appl. Phys. 134, 014102 (2023); doi: 10.1063/5.0156566 134, 014102-9

Published under an exclusive license by AIP Publishing

View publication stats