Electronic properties of 2H-stacking bilayer MoS2 measured by terahertz time-domain spectroscopy

Xingjia Cheng, Wen Xu, Hua Wen, Jing Zhang, Heng Zhang, Haowen Li, Francois M. Peeters, Qingqing Chen

Front. Phys. ›› 2023, Vol. 18 ›› Issue (5) : 53303.

PDF(5767 KB)
Front. Phys. All Journals
PDF(5767 KB)
Front. Phys. ›› 2023, Vol. 18 ›› Issue (5) : 53303. DOI: 10.1007/s11467-023-1295-1
RESEARCH ARTICLE
RESEARCH ARTICLE

Electronic properties of 2H-stacking bilayer MoS2 measured by terahertz time-domain spectroscopy

Author information +
History +

Abstract

Bilayer (BL) molybdenum disulfide (MoS2) is one of the most important electronic structures not only in valleytronics but also in realizing twistronic systems on the basis of the topological mosaics in moiré superlattices. In this work, BL MoS2 on sapphire substrate with 2H-stacking structure is fabricated. We apply the terahertz (THz) time-domain spectroscopy (TDS) for examining the basic optoelectronic properties of this kind of BL MoS2. The optical conductivity of BL MoS2 is obtained in temperature regime from 80 K to 280 K. Through fitting the experimental data with the theoretical formula, the key sample parameters of BL MoS2 can be determined, such as the electron density, the electronic relaxation time and the electronic localization factor. The temperature dependence of these parameters is examined and analyzed. We find that, similar to monolayer (ML) MoS2, BL MoS2 with 2H-stacking can respond strongly to THz radiation field and show semiconductor-like optoelectronic features. The theoretical calculations using density functional theory (DFT) can help us to further understand why the THz optoelectronic properties of BL MoS2 differ from those observed for ML MoS2. The results obtained from this study indicate that the THz TDS can be applied suitably to study the optoelectronic properties of BL MoS2 based twistronic systems for novel applications as optical and optoelectronic materials and devices.

Graphical abstract

Keywords

terahertz time-domain spectroscopy / bilayer MoS2 / optoelectronics

Cite this article

Download citation ▾
Xingjia Cheng, Wen Xu, Hua Wen, Jing Zhang, Heng Zhang, Haowen Li, Francois M. Peeters, Qingqing Chen. Electronic properties of 2H-stacking bilayer MoS2 measured by terahertz time-domain spectroscopy. Front. Phys., 2023, 18(5): 53303 https://doi.org/10.1007/s11467-023-1295-1

1 Introduction

The ability to freely manipulate electromagnetic waves or light waves, an important carrier of energy and information, has always been a hot research topic. However, most optical media and systems exhibit coupled transmission and the reflection performances, and their amplitude and phase of transmission/reflection waves are also interdependent. Tuning one or more parameters of an optical medium or configuration usually leads to simultaneous changes in the power, amplitude and phase of the transmission and reflection waves, which limits the degree of freedom in optical manipulation and the information capacity of light. Recently, to overcome this bottleneck, some composite metasurfaces have been developed, which consist of arrays of artificial microstructures that can manipulate light in both transmission and reflection spaces [14], and even achieve simultaneous control of phase and amplitude [5, 6].However, these metasurfaces are usually composed of large numbers of complex microstructures. To achieve specific functions, it is inevitable to adjust numerous unit cells. Moreover, the characteristics of most metasurfaces are geometrically dependent, which pose a challenge for integration.
On the other hand, epsilon-near-zero media (ENZ) have attracted much interest due to their robust optical characteristics against geometric transformations [710]. In ENZ, the vanishing permittivity leads to a near-zero wavenumber, a wavelength much greater than its scale and the out-of-plane magnetic fields static in space. These features enable the control of the radiating direction [11, 12], phase distribution [13, 14], transmission performance [15, 16], and energy distribution [17] of light. More interestingly, ENZs exhibit extremely stable transmission performance irrespective of their geometric shape for TM polarized waves [15]. The spatially static magnetic fields also enable the photonic doping effect of ENZ such that the optical properties of the whole structure are tunable through one or few macroscopic dopants embedded in any position of the ENZ host [1836]. These unconventional features make doped ENZs useful for engineering applications enabling the design of various doped-ENZ-based devices [37], such as power divider [38], antenna [12, 39], and filter [40]. Recently, increasing researchers are interested in non-Hermitian photonic doping ENZ with tailored loss and/or gain. With the introduction of optical non-Hermiticity in some components, the reconfigurability of doped ENZ is significantly improved [41], which brings more interesting effects, such as the enhancement of optical attenuation or amplification [42], coherent absorption [43, 44], and nonreciprocal transmission [42, 45]. Doping of ENZ provides a more convenient way for optical manipulation. Optical devices based on doped ENZ are convenient for integration and are expected to be applied in more fields, for their relatively simple structures and geometrically insensitive optical properties. However, out-of-plane fields with a constant amplitude in ENZ further intensify the correlation among the power, amplitude and phase of transmission and reflection waves.
Inspired by composite metasurfaces, we propose a novel method to break the correlation between the transmission and reflection waves of non-Hermitian doped ENZ by inserting a narrow dielectric slit. We demonstrate that both the amplitude and phase of both the reflection and transmission waves can be tuned independently through few dopants. Utilizing the amplification effect of loss or gain of dopants induced by resonance [42, 43], the adjustable range increases significantly. Moreover, we show that our method can be exploited to design a highly reconfigurable reflectionless signal distributor and generator based on a three-port doped ENZ waveguide structure.

2 Theories and analyses

We investigate the scattering properties of an arbitrarily shaped ENZ structure connecting two ports, as shown in Fig.1(a). The structure consists of two ENZ regions (region 1 and region 2 with cross areas A1 and A2 respectively) separated by a dielectric slit with width ws, thickness d and relative permittivity εs. The ENZ regions are connected to two ports with the same width wp and relative permittivity εp. Each ENZ region contains a cylindrical dopant j (j=1,2) with radius Rj and relative permittivity εj to control the reflection and transmission performance of the structure. The dopants are nonmagnetic and may exhibit dielectric loss or gain. The structure is bounded by perfect electric conductor (PEC) walls. We assume that a transverse magnetic polarized (TM) plane wave with angular frequency ω normally incident from the input port (the left port). If the time factor eiωτ is omitted, the magnetic field inside the input port, the slit and the output port (the right port) are expressed as
Fig.1 (a) Diagrammatic sketch of the doped ENZ structure with two ports. (b) Diagram of the solutions of m1 on the complex plane for a given reflectance R when the transmission coefficient is fixed as t0. (c) Diagram of the solutions of m1 and corresponding m2 on the complex plane for a given transmittance T when the reflection coefficient is fixed as r0.

Full size|PPT slide

Hi=H0(eikpx+reikpx)ez,
Hs=H0[aeiks(xl1)+beiks(xl1)]ez,
Ho=H0teiko(xl1dl2)ez,
where kp=εpk0 and ks=εsk0 are wavenumbers in the two ports and the slit, respectively; k0=ω/c is the angular wavenumber in vacuum; and c is the velocity of light in vacuum; H0 is the complex amplitude of the incident wave. t and r are the transmission coefficient and the reflection coefficient of the doped ENZ respectively. a and b are the complex amplitudes of the forward and backward waves in the slit normalized by H0. Correspondingly, the electric fields in these regions are derived from iωεE=×H as
Ei=z0zpH0(eikpxreikpx)ey,
Es=z0zsH0[aeiks(xl1)beiks(xl1)]ey,
Eo=z0zpH0teiko(xl1dl2)ey.
zp=1/εp and zs=1/εs are the relative impedances of the two ports and the slit, respectively, and z0 is the impedance in vacuum. We can calculate r, a, b, t from continuous boundary condition and the integral form of Faraday’s law of electromagnetic induction as M(r,a,b,t)T=(1,1,0,0)T where
M=(11101αim1αim100D+D10αD+αD1im2);
α=zsws/zpwp, D±=e±ksd and mj=k0μeffjAj/(zpwp). Here the homogenized effective magnetic permeability μeffj is calculated with μeffj=1πRj2/Aj+2πRjJ1(εdjk0Rj)/[Ajεdjk0J0(εdjk0Rj)] according to the theory of photonic doping [18], which is commonly used to describe the transmission and reflection properties of doped ENZs. However, μeffj alone is not enough to capture the effects of other physical quantities that influence the transmission and reflection properties, such as Aj, wp, zp, and k0. Therefore, we adopt the dimensionless parameter mj that combines μeffj and these quantities in a simple way. This parameter can characterize the ENZ region j more comprehensively and simplify our analysis. It is easy to see that mj dominates both the transmission and reflection performance of the ENZ in region j. In this work, we name mj as the transmission-reflection main control (TRMC) parameter of the ENZ in region j, and use mj as the variable to describe the ENZ in region j. We can obtain expressions for the reflection coefficient r and the transmission coefficient t as a function of m1 and m2, by inverting matrix M. The correlation between r and t depends on their Jacobian determinant calculated by
det[(r,t)(m1,m2)]=4γ2(zpwpzsws)3[m2(ηi)]det3(M),
where det(M)=2iαγ[η(m1+m2+2i)(m1+i)(m2+i)+α2], γ=sin(ksd)/α and η=αcot(ksd). When ksdnπ (n=0,±1,±2,), m2ηi and det(M) is convergent or both |det(M)|3 and |γ2(m2ηi)| are small quantities of the same order, the Jacobian determinant is nonzero, which means r and t are independent. When t is fixed as t0, r can be expressed as a univalent complex function of m1 or m2
r=m1(η+γ1t0+i)m1(ηi)=γt0[m2(ηi)]1.
On the complex plane of m1, there are a zero point at ηγ1t0+i and a pole point at ηi of the reflection coefficient r. This means that the reflection wave disappears when m1=ηγ1t0+i and diverges when m1=ηi. The transmission coefficient t0 determines the location of the reflection zero point, but not the unique value of reflection coefficient r. Fig.1(b) shows the diagram of how r varies with m1. We represent r as the negative ratio of the vector rn and rd, where rn is the vector from the zero point to m1 and rd is the vector from the pole point to m1. Any reflectance R=|r|2 from near zero to near infinity is achievable by tuning m1. Accordingly, to ensure the permanent t0, m2 should also be tuned to meet Eq. (5). On the complex plane of m2, the zero point maps to η+1/γt0i which is also determined by t0. The pole point maps to the point of infinity. Based on the resonance properties of dopants, a wide range of TRMC parameters m1 and m2, including near-infinite values, is available by doping the proper materials with low loss or gain. In this way, without affecting the transmission coefficient t0, the reflectance is broadly tunable from near zero to near infinity by adjusting the parameters of the dopants embedded in both ENZ regions unless the zero point and the pole point coincide or both are at infinity. Both the transmission and reflection waves of the doped ENZ structure can be controlled independently. Fig.1(b) also illustrates that the contour of the amplitude of any reflection coefficient |r| (0 or ) forms an Apollonius circle around the zero point or the pole point of r. The reflection phase φr=θnθd is represented as the angle between vectors rn and rd. When m1 moves along a contour of |r|, φr changes from π to π. We can achieve any reflection phase for any given |r|. The argument principle also supports this interpretation. Thus, the correlation between the amplitude and phase of the reflection wave is broken.
Similarly, for a determined reflection coefficient r=r0, t is a univalent complex function of m1 or m2
t=γ(1+r0)[m1(η+κi)]=1+r0γ[m2(ηi)],
where κ=(1r0)/(1+r0). On the complex plane of m1, the transmission coefficient t vanishes at η+κi and diverges at the point of infinity. While, if t is expressed as a function of m2, it vanishes at the point of infinity and diverges at ηi. The reflection coefficient r0 affect the location of the transmission zero point but it does not determine t uniquely. As illustrated in Fig.1(c), on the plane of m1, the contour of constant amplitude |t| (0 or ) forms a circle with the zero point as the center and a radius proportional to |t|. On the plane of m2, the corresponding contour mapped by Eq. (6) forms another circle with the pole point as the center and a radius inversely proportional to |t|. Based on the resonance effect, it is not difficult to obtain the TRMC parameter m1 or m2 with a value close to infinity by doping. Hence, when the reflection coefficient is fixed, the transmittance T=|t|2 can also be tailored in a wide range from near zero to near infinity in theory. Meanwhile, the argument principle also enables the full angle tunability of the transmission phase φt for any given non-zero and non-divergent |t|. From the perspective of vector geometry in Fig.1(c), φt is related to φv(φv), which is the angle of the vector from the zero (pole) point to the corresponding contour of |t|. They satisfy φv=π+φtArg(1+r0). When r0 and |t| are determined, φt can be tuned in any angle by changing φv. Therefore, the amplitude and the phase of transmission wave are no longer interrelated.
We further discuss the physical mechanism behind the decoupling of transmission and reflection. The incident and reflection waves superpose in ENZ region 1, which means the reflected waves depend on the magnetic field in ENZ region 1. Moreover, the transmitted wave is determined by the magnetic field in ENZ region 2. The magnetic fields in the two ENZ regions interact through the dielectric slit between them. The modes of magnetic fields in the dielectric slit paly a critical role in the correlation between the transmitted and reflected waves. If the single-pass phase shift in the slit satisfies ksd=nπ (n=0,±1,±2,), the magnetic fields in the slit and two ENZ regions are symmetric or antisymmetric with respect to the central axis of the slit, which cannot be changed by any photonic doping. Such symmetry of magnetic modes requires the transmission and reflection coefficients to be rigidly constrained by 1+r=±t. Of course, this covers the case of n=0 in which no slits are introduced in the doped ENZ. Thus, the coupling of transmission and reflection waves of a doped ENZ mainly originates from the symmetry of magnetic fields between the input and output ports. Reasonably, introducing the dielectric slit whose parameters satisfy ksdnπ is an effective way to break this symmetry of magnetic fields. In addition, non-Hermitian dopants embedded in ENZ regions also play an important role in the decoupling of transmission and reflection, because they can overcome the other limitation for transmission and reflection waves brought by the law of energy conservation. Furthermore, these non-Hermitian dopants provide enough degrees of freedom to tune both the amplitude and phase of both transmission and reflection waves. Hence, we achieve the decoupling of transmission and reflection in the ENZ structure by introducing a proper dielectric slit and non-Hermitian dopants. The desired transmission and reflection waves can be obtained simultaneously through in proposed structure as demonstrated in Fig.1(b) and (c).
Nevertheless, there are some noteworthy cases in which the decoupling of transmission and reflection is not valid, other than ksd=nπ. Firstly, when m2=ηi, the zero point and the pole point of the reflection coefficient coincide. In this case, the transmission and reflection coefficients are fixed as 2iγ and −1 respectively regardless of m1. This means that the magnetic field always vanishes in ENZ region 1 and the photonic doping effect in ENZ region 1 is completely invalid. This strange phenomenon is caused by the reflection features of gain-based ENZ in region 2. The reflection wave from ENZ region 2 is amplified to match the amplitude of the corresponding incident wave. Coincidentally, its reflection phase cancels out the round-trip phase shift in the slit and produces an extra phase shift of π. These lead to destructive interference of light in ENZ region 1, irrespective of the properties of ENZ region 1. Secondly, the transmission-reflection decoupling makes sense only when both r and t are convergent. From Eqs. (3), (5), and (6) we can easily prove that divergent r will result in the divergent t0, and divergent t will lead to the divergent r0. In fact, the divergence of r and t corresponds precisely to the laser mode induced by the constructive interference of multiple reflected and amplified waves in non-Hermitian optical media [4650].

3 Results and discussion

3.1 Reconfigurability of reflection wave without transmission

We numerically analyze two specific cases in the section: near-zero transmission and near-zero reflection. First, we investigate how reflection wave of the structure can be reconfigured when the transmittance vanishes. According to Eq. (5), the reflection coefficient r depends on the divergence of TRMC parameters m1 and m2 when t0 tends to zero. If m2 is convergent, m1 must be divergent and r=1. ENZ region 1 acts as a perfect magnetic conductor. If m2 diverges, the reflection zero point η+i and the pole point ηi are symmetrical with respect to the real axis. We assume the following parameters: wp=λ0, ws=2λ0, d=0.1λ0, A1=2.5048λ02, A2=2.4411λ02, R1=R2=0.35λ0, and εs=1. To block the transmission wave, the permittivity of dopant 2 is set to close to a state of resonance reflection. In our calculation, εd2=1.196 gives a large m2=12122. Fig.2(a) and (b) show the calculation results of the reflectance and the transmittance as a function of the real part εd1 and the imaginary part εd1 (the loss factor) of the relative permittivity of dopant 1. The reflectance varies continuously from 6.11×109 to 1.64×108 with εd1 and εd1. In our calculation domain, there is a reflection zero point and a reflection pole point symmetrically distributed on both sides of the real axis. In the most cases, waves are prevented from transmitting through the output port. When εd1 approaches the reflection pole point, the transmittance increases to 12.89. As parameters get close to satisfying the laser mode condition, the resonant dopant 2 cannot block the transmission wave effectively, but the transmittance is still negligible compared with the reflectance. Hence, we obtain a wide range of reflectance from near-zero to near-infinity by adjusting dopant 2 with the near-zero transmittance. Moreover, Fig.2(c) shows the distribution of the reflection phase near some specific reflectance values including 103,102,101,1,101,102 and 103, which demonstrate the reconfigurability of the reflection phase. Any determined reflectance with any reflection phase is achievable by choosing the proper combination of εd1 and εd1. We chose two values of the relative permittivity of dopant 1 close to the zero point εd1=6.485+0.017i and the pole point εd1=6.4850.017i to conduct simulations using COMSOL Mutiphysics, as shown in Fig.3. When the transmission wave is blocked, the zero reflection means that the energy of the incident wave is totally absorbed, as shown by the magnetic field distribution illustrated in Fig.3(a). Fig.3(b) plots the amplitude of the magnetic fields distributed along the gray dashed line in (a). Clearly, in the input port, the incident wave with amplitude H0 is almost unaffected, which means that there are almost no reflection waves in the doped ENZ structure. Meanwhile, in the other port, the amplitude of the transmission wave is less than 0.001H0. When εd1=6.4850.017i, the reflection wave is amplified to about 101H0, which interferes with the incident wave. Because the amplitude of the reflection wave is much greater than that of the incident wave, the interference effect is not significant [Fig.3(c) and (d)]. Although the amplitude of the transmission wave is enhanced to 0.028H0, the impact is still trivial. By changing the permittivity of dopant 1, we achieved perfect absorption and amplified reflection waves with near-zero transmission.
Fig.2 Calculation results of (a) reflectance R and (b) transmittance T versus the real part of the relative permittivity of dopant 1 εd1 and the imaginary part of the relative permittivity of dopant 1 εd1 when εd2=1.196. (c) Calculation results of the reflection phase for R=103±4×105, R=102±2×104, R=101±1×103, R=1±1×102, R=101±1×101, R=102±2 and R=103±4×101.

Full size|PPT slide

Fig.3 (a) Simulation results of the distribution of magnetic fields in the ENZ structure for perfect absorption. (b) The corresponding distribution of the amplitude of magnetic fields along the gray dashed line in (a). (c) Simulation results of the distribution of magnetic fields in the ENZ structure for amplified reflection with extremely low transmission. (d) The corresponding distribution of the amplitude of magnetic fields along the gray dashed line in (c).

Full size|PPT slide

3.2 Reconfigurability of transmission wave without reflection

Next, we numerically analyze how the transmission wave can be reconfigured without reflection. To obtain the near-zero reflectance, dopants in both regions should be tuned to satisfy [m1(η+i)][m2(ηi)]=γ2, which is different from the case of near-zero transmission. In Fig.4(a), we calculated the transmittance of the doped ENZ structure as a function of εd1, T(εd1), when the reflectance is close to zero. We use the same parameters as in Fig.2 and Fig.3 except for the permittivity of dopants. In the calculation domain, T(εd1) is tunable in a wide range. The transmission zero point is near εd1=6.485+0.017i, which agrees with the above calculation for the case of εd2=1.196. Due to the resonance effect of dopant 1, the transmission pole point is mapped from the infinity point of m1 to a point on the real axis of εd1 near 6.301. Correspondingly, if we express the reflectionless transmittance as a function of εd2, the transmission zero point is mapped to near εd2=6.301 and the pole point is mapped to near εd1=6.4920.019i, as shown in Fig.4(b). Near the transmission pole point, T(εd2) varies more sharply than T(εd1), which requires a higher accuracy for εd2. In contrast, near the transmission zero point, T(εd2) varies more gently than T(εd1), which means that εd1 must be more accurate. Fig.5(a) and (b) show the simulation results of the amplification of the transmission wave with extremely low reflection. We choose εd1=6.2020.046i to obtain T(εd1)=71.24. Based on T(εd1)=T(εd2) and φt(εd1)=φt(εd2), we find the corresponding dopant 2 with εd2=6.5000.0175i from the calculation results. The simulated amplitude of the transmission wave is approximately 8.45H0, which matches the calculation results. The reflection wave is effectively suppressed. We implement a cascadable signal amplifier that can be useful for signal processing using the doped ENZ structure. As a result, the transmittance and reflectance are decoupled, which allows the transmittance to be tuned in a very broad range by adjusting two dopants while maintaining a negligible reflectance.
Fig.4 Calculation results of the transmittance T versus (a) the real part and imaginary part of the relative permittivity of dopant 1 and (b) that of dopant 2 for reflectance R=0. (c) Distributions of phase as a function of the real part and imaginary part of the relative permittivity of dopant 1 depicted by the red coordinate system and that of dopant 2 depicted by the blue coordinate system when the transmittance T=1±0.01.

Full size|PPT slide

Fig.5 (a) Simulation results of the distribution of magnetic fields in the ENZ structure for the amplified transmission with extremely low reflection. (b) The corresponding distribution of the amplitude of magnetic fields along the gray dashed line in (a). Simulation results of the distribution of magnetic fields in the ENZ structure for the reflectionless total transmission with (c) near-zero phase advance and (d) a π-phase advance.

Full size|PPT slide

In addition, for any given transmittance other than the zero point and the pole point in the calculation domain, all the conforming εd1 and corresponding εd2 are respectively on a curve enclosing the zero or the pole point. On any curve with a constant T, different values of εd1 or εd2 indicate different transmission phases. In Fig.4(c), we show the calculation results of the transmission phase as a function of εd1, φt(εd1), and that of εd2, φt(εd2), when T1 with R0. One can obtain any transmission phase ranging from π to π by simultaneously tuning εd1 and εd2. For example, when we want the transmission wave to be in phase with the incident wave, we find from the calculation results that we need εd1=6.441+0.0099i and εd2=6.4460.0105i. Conversely, if we want the transmission wave to be out of phase with the incident wave, we need εd1=6.568+0.036i and εd2=6.5830.040i. Fig.5(c) and (d) demonstrate the simulation results of magnetic field of the ENZ structure when the transmission wave is in and out of phase with the incident wave respectively. By proper doping, reflectionless total transmission with the expected phase is obtained. These results show that the amplitude and the phase of the transmission wave are also decoupled.
Specifically, in our calculation domain, we notice that there are two time-reversal symmetric reflectionless total transmission modes where both dopants are pure dielectric with εd1=6.444, εd2=6.448 (m1=m2=0.47) or εd1=6.567, εd2=6.582 (m1=m2=5.99). They are two discrete resonance total transmission modes formed by destructive interference of multiple reflection waves [35]. Other reflectionless total transmission modes without time-reversal symmetry, in addition to the interference effect, involves energy exchanges due to the balanced loss and gain of dopants, which requires m1 and m2 to be PT symmetrical [42, 45, 46]. In our calculations, the difference between εd1 and εd2 is compensated by the geometric asymmetry. The total transmission modes induced by resonance can be regarded as some special cases of the PT symmetrical total transmission. In the non-Hermitian system, reflectionless total transmission modes degenerated by their transmission phases are continuous, which can be extended to other reflectionless modes with any fixed transmittance. Due to the periodicity of the resonance states of electromagnetic fields in dopants, these continuous modes will recur outside our calculation domains.

4 Applications in reflectionless signal distributor and generator

One can generalize the principle of transmission-reflection decoupling and the reconfigurability of amplitudes and phases to structures of doped ENZ with multiple ports. Since the number of ports does not affect its physical essence, we expect that the S-parameters of non-Hermitian doped ENZ with multiple ports will be uncorrelated and reconfigurable in their amplitude and phase when several slits are introduced. To demonstrate this, we propose a scheme to realize a highly reconfigurable signal distributor and generator using a structure of non-Hermitian doped ENZ with slits as illustrated in Fig.6(a). The structure is composed of three ENZ regions labeled by j=1,2,3. They are separated by two slits. Each ENZ region contains a dopant and connects to a port. Port 1 connecting to region 1 is the input port. Ports 2 and 3 connecting to regions 2 and 3 are two output ports. S11 denotes the reflection coefficient. S21 and S31 are the transmission coefficients of ports 2 and 3 respectively. According to the universal expression of S-parameters of doped ENZ with multiple ports and slits [36]. We find that when the reflection coefficient S11=0, the two transmission coefficients can be simplified into a very concise form similar to Eq. (6):
Fig.6 Plane (a) and three-dimensional (b) diagrammatic sketch of the highly reconfigurable reflectionless signal distributor and generator. Simulation results of the magnetic field distribution in the signal distributor and generator when (c) the power of the incident wave is equally distributed into two output ports with opposite transmission phases, (d) when the incident wave is totally distributed into one output port with φ21=π/2 and (e) when the incident wave is amplified in one output port and decayed in another output port with the same phase.

Full size|PPT slide

Sj1=1γj[mj(ηji)],j=2,3,
with
m1=ij=23[(mj+i)cos(ksjdj)]+zsjwsjzp1wp1sin(ksjdj),
where mj=k0μeffjAj/(zpjwpj); μeffj and Aj are the effective permeability and the cross-sectional area of ENZ region j, respectively; zpj and wpj are the relative impedance and the width of port j, respectively; zsj, ksj, wsj and dj are the relative impedance, wavenumber, width and thickness of the slit between regions 1 and j, respectively; γj=zpjwpjsin(ksjdj)/(zsjwsj); η=zsjwsjcot(ksjdj)/(zpjwpj). Since these ports and slits are nonmagnetic, their relative permittivites are εpj=zpj2 and εsj=zsj2, respectively. The dopant in region j, dopant j with relative permittivity εdj, exhibits a circular cross section with the radius Rj. As shown in Eq. (7a), the transmission coefficient Sj1 is expressed as a univalent function of mj. One can obtain any designed transmission coefficients S21 and S31 by choosing suitable dopants 2 and 3 respectively, and then find a matched dopant 1 to eliminate the reflection wave according to Eq. (7b). The incident wave can be split, amplified, decayed and phase shifted by any means with near-zero reflection through the signal distributor and generator.
Further, we construct this ENZ-based signal distributor and generator through several interconnected three-dimensional metal waveguides that work at TE10 mode [18, 3234], as shown in Fig.6(b). The top metal wall of the waveguide structure is removed. The effective relative permittivity for TE10 mode is εeff=εf(cπ/ωh)2, where εf is the relative permittivity of the filler and h is the height of the waveguide. We fill the three irregular-shaped regions with air (εf=1) and set the height and frequency to satisfy cπ/(ωh)=1, creating lossless ENZ host media at the cutoff frequency of TE10 mode. We also fill media with relative permittivities εpj+1, εsj+1 and εdj+1 in corresponding regions of ports, slits and dopants, respectively, to achieve the desired effective relative permittivities εpj, εsj and εdj for TE10 mode. To prevent higher order modes from interfering, some thin metal rods are inserted near interfaces between different media.
We numerically simulate three cases with different transmission requirements of the signal distributor and generator constructed by the TE10 waveguide structure using the following parameters: wp1=wp2=wp3=λ0, where λ0 is the working wavelength in vacuum; ws2=ws3=1.2λ0; d1=d2=0.1λ0; h=cπ/ω=0.5λ0; R1=R2=R3=0.4λ0; A1=2.986λ02; A2=A3=1.667λ02; zp1=zp2=zp3=zs2=zs3=1; ks2=ks3=k0. We use PECs to simulate metal sidewalls of the waveguide structure and introduce a total of 104 thin PEC rods with cross section area 7.854×105λ02 near interfaces between different media. In the first case, we split the incident wave into two output ports with equal power and opposite phases. We achieve this by using εd2=4.9910.014i and εd3=5.1400.0507i for dopants 2 and 3, and εd1=4.951+0.0081i for dopant 1 to eliminate reflection. Fig.6(c) shows the simulated magnetic field distribution for this case. The transmission waves in ports 2 and 3 have the same amplitude 2H0 where H0 is the incident wave amplitude. The phase difference between port 1 and port 2 is zero, while the phase difference between port 1 and port 3 is π. Meanwhile, there almost no reflection from port 1. In the second case, we want total transmission through port 2 with a phase shift φ21=π/2, no transmission through port 3 and no reflection through port 1. We achieve this by using εd1=4.9510.0084i, εd2=5.042+0.0251i and εd3=4.824 for dopants 1, 2 and 3. Such function is demonstrated by the simulated magnetic field distribution in Fig.6(d). Interestingly, by choosing proper dopants, the structure can even generate an amplified and a decayed reflectionless transmission signal simultaneously. In the third case, we amplify the incident wave in port 2 and decay it in port 3. By using dopants 1, 2 and 3 with εd1=4.9230.0049i, εd2=5.0150.0183i and εd3=4.9930.0143i, the amplitude of the transmission wave in port 2 is amplified to 1.5H0 and that in port 3 decays to 0.75H0 [Fig.6(e)]. In port 1, the reflection wave is also blocked effectively. The designed signal distributor and generator exhibits ultrahigh reconfigurability and potential applications in transmission and processing of optical signals.

5 Conclusion

We have demonstrated a novel way of manipulating optical waves by exploiting the properties of non-Hermitian doped ENZ structures with slits. We have shown that a dielectric slit in the ENZ structure can break the correlation between the transmission and reflection waves and allows us to control their amplitude and phase independently by adjusting the permittivity of few dopants. We have achieved various optical effects such as perfect absorption, high-gain reflection without transmission, reflectionless high-gain transmission and total transmission with phase modulation, by choosing appropriate values of permittivity with tailored low loss or gain for dopants. We have also extended our approach to doped ENZ structures with multiple slits and ports, and designed a highly reconfigurable and reflectionless signal distributor and generator based on a three-port ENZ waveguide structure. This device can distribute, amplify and phase-shift the incident wave in any desired way, which is useful for various applications such as optical communication and signal processing. Our work overcomes the restrictions of doped ENZ imposed by the correlation between the transmission and reflection waves as well as that between amplitudes and phases for freer optical manipulation, opening up a new avenue for designing novel optical devices based on non-Hermitian doped ENZ structures.

References

[1]
K. S. Novoselov , A. K. Geim , S. V. Morozov , D. Jiang , Y. Zhang , S. V. Dubonos , I. V. Grigorieva , A. A. Firsov . Electric field effect in atomically thin carbon films. Science, 2004, 306(5696): 666
CrossRef ADS Google scholar
[2]
C. Castellani , C. DiCastro , P. A. Lee . Metallic phase and metal-insulator transition in two-dimensional electronic systems. Phys. Rev. B, 1998, 57(16): R9381
CrossRef ADS Google scholar
[3]
K. F. Mak , D. Xiao , J. Shan . Light-valley interactions in 2D semiconductors. Nat. Photonics, 2018, 12(8): 451
CrossRef ADS Google scholar
[4]
J. W. Jiang . Graphene versus MoS2: A short review. Front. Phys., 2015, 10(3): 287
CrossRef ADS Google scholar
[5]
H. M. Hill , A. F. Rigosi , C. Roquelet , A. Chernikov , T. C. Berkelbach , D. R. Reichman , M. S. Hybertsen , L. E. Brus , T. F. Heinz . Observation of excitonic Rydberg states in mono-layer MoS2 and WS2 by photoluminescence excitation spectroscopy. Nano Lett., 2015, 15(5): 2992
CrossRef ADS Google scholar
[6]
D. Xiao , G. B. Liu , W. Feng , X. Xu , W. Yao . Coupled spin and valley physics in monolayers of MoS2 and other group-VI dichalcogenides. Phys. Rev. Lett., 2012, 108(19): 196802
CrossRef ADS Google scholar
[7]
Q. Tong , H. Yu , Q. Zhu , Y. Wang , X. Xu , W. Yao . Topological mosaics in moiré superlattices of van der Waals heterobilayers. Nat. Phys., 2017, 13(4): 356
CrossRef ADS Google scholar
[8]
Y. Cao , V. Fatemi , S. Fang , K. Watanabe , T. Taniguchi , E. Kaxiras , P. Jarillo-Herrero . Unconventional superconductivity in magic-angle graphene superlattices. Nature, 2018, 556(7699): 43
CrossRef ADS Google scholar
[9]
K. Seyler , P. Rivera , H. Yu , N. Wilson , E. Ray , D. Mandrus , J. Yan , W. Yao , X. Xu . Signatures of moiré-trapped valley excitons in MoSe2/WSe2 heterobilayers. Nature, 2019, 567(7746): 66
CrossRef ADS Google scholar
[10]
T. Cai , S. A. Yang , X. Li , F. Zhang , J. Shi , W. Yao , Q. Niu . Magnetic control of the valley degree of freedom of massive Dirac fermions with application to transition metal dichalcogenides. Phys. Rev. B, 2013, 88(11): 115140
CrossRef ADS Google scholar
[11]
K. F. Mak , C. Lee , J. Hone , J. Shan , T. F. Heinz . Atomically thin MoS2: A new direct-gap semiconductor. Phys. Rev. Lett., 2010, 105(13): 136805
CrossRef ADS Google scholar
[12]
C. Lee , H. Yan , L. E. Brus , T. F. Heinz , J. Hone , S. Ryu . Anomalous lattice vibrations of single- and few-layer MoS2. ACS Nano, 2010, 4(5): 2695
CrossRef ADS Google scholar
[13]
J.E. PadilhaH.PeelaersA.JanottiC.G. Van de Walle, Nature and evolution of the band-edge states in MoS2: From monolayer to bulk, Phys. Rev. B 90(20), 205420 (2014)
[14]
H. M. Dong , S. D. Guo , Y. F. Duan , F. Huang , W. Xu , J. Zhang . Electronic and optical properties of single- layer MoS2. Front. Phys., 2018, 13(4): 137307
CrossRef ADS Google scholar
[15]
X. Z. Zhang , R. Y. Zhang , Y. Zhang , T. Jiang , C. Y. Deng , X. A. Zhang , S. Q. Qin . Tunable photoluminescence of bilayer MoS2 via interlayer twist. Opt. Mater., 2019, 94: 213
CrossRef ADS Google scholar
[16]
M. Xia , B. Li , K. Yin , G. Capellini , G. Niu , Y. Gong , W. Zhou , P. M. Ajayan , Y. H. Xie . Spectroscopic signatures of AA′ and AB stacking of chemical vapor deposited bilayer MoS2. ACS Nano, 2015, 9(12): 12246
CrossRef ADS Google scholar
[17]
A. M. van der Zande , J. Kunstmann , A. Chernikov , D. A. Chenet , Y. M. You , X. X. Zhang , P. Y. Huang , T. C. Berkelbach , L. Wang , F. Zhang , M. S. Hybertsen , D. A. Muller , D. R. Reichman , T. F. Heinz , J. C. Hone . Tailoring the electronic structure in bilayer molybdenum disulfide via interlayer twist. Nano Lett., 2014, 14(7): 3869
CrossRef ADS Google scholar
[18]
C. Wang , W. Xu , H. Mei , H. Qin , X. Zhao , C. Zhang , H. F. Yuan , J. Zhang , Y. Xu , P. Li , M. Li . Substrate-induced electronic localization in monolayer MoS2 measured via terahertz spectroscopy. Opt. Lett., 2019, 44(17): 4139
CrossRef ADS Google scholar
[19]
H. Wen , W. Xu , C. Wang , D. Song , H. Y. Mei , J. Zhang , L. Ding . Magneto-optical properties of monolayer MoS2-SiO2/Si structure measured via terahertz time-domain spectroscopy. Nano Select, 2021, 2(1): 90
CrossRef ADS Google scholar
[20]
S. Kumar , A. Singh , S. Kumar , A. Nivedan , M. Tondusson , J. Degert , J. Oberle , S. J. Yun , Y. H. Lee , E. Freysz . Enhancement in optically induced ultrafast THz response of MoSe2−MoS2 heterobilayer. Opt. Express, 2021, 29(3): 4181
CrossRef ADS Google scholar
[21]
M. Bala Murali Krishna , J. Madéo , J. P. Urquizo , X. Zhu , S. Vinod , C. S. Tiwary , P. M. Ajayan , K. M. Dani . Terahertz photoconductivity and photocarrier dynamics in few-layer hBN/WS2 van der Waals heterostructure laminates. Semicond. Sci. Technol., 2018, 33(8): 084001
CrossRef ADS Google scholar
[22]
M. Liao , Z. Wei , L. Du , Q. Wang , J. Tang , H. Yu , F. Wu , J. Zhao , X. Xu , B. Han , K. Liu , P. Gao , T. Polcar , Z. Sun , D. Shi , R. Yang , G. Zhang . Precise control of the interlayer twist angle in large scale MoS2 homostructures. Nat. Commun., 2020, 11(1): 2153
CrossRef ADS Google scholar
[23]
H. Yu , M. Liao , W. Zhao , G. Liu , X. J. Zhou , Z. Wei , X. Xu , K. Liu , Z. Hu , K. Deng , S. Zhou , J. A. Shi , L. Gu , C. Shen , T. Zhang , L. Du , L. Xie , J. Zhu , W. Chen , R. Yang , D. Shi , G. Zhang . Wafer-scale growth and transfer of highly-oriented monolayer MoS2 continuous films. ACS Nano, 2017, 11(12): 12001
CrossRef ADS Google scholar
[24]
Y. Yu , C. Li , Y. Liu , L. Su , Y. Zhang , L. Cao . Controlled scalable synthesis of uniform, high-quality monolayer and few-layer MoS2 films. Sci. Rep., 2013, 3(1): 1866
CrossRef ADS Google scholar
[25]
X. Wang , H. Feng , Y. Wu , L. Jiao . Controlled synthesis of highly crystalline MoS2 flakes by chemical vapor deposition. J. Am. Chem. Soc., 2013, 135(14): 5304
CrossRef ADS Google scholar
[26]
Q. Q. Wang , J. Tang , X. M. Li , J. P. Tian , J. Liang , N. Li , D. P. Ji , L. D. Xian , Y. T. Guo , L. Li , Q. H. Zhang , Y. B. Chu , Z. Wei , Y. C. Zhao , L. J. Du , H. Yu , X. D. Bai , L. Gu , K. H. Liu , W. Yang , R. Yang , D. X. Shi , G. Y. Zhang . Layer-by-layer epitaxy of multi-layer MoS2 wafers. Natl. Sci. Rev., 2022, 9(6): nwac077
CrossRef ADS Google scholar
[27]
H. Sajjad , A. S. Muhmmad , V. Dhanasekaran , Z. I. Muhmmad , S. Jai , F. K. Muhmmad , E. Jonghwa , S. Yongho , J. Jongwan . Controlled synthesis and optical properties of polycrystalline molybdenum disulfide atomic layers grown by chemical vapor deposition. J. Alloys Compd., 2015, 653(25): 369
[28]
C. R. Zhu , G. Wang , B. L. Liu , X. Marie , X. F. Qiao , X. Zhang , X. X. Wu , H. Fan , P. H. Tan , T. Amand , B. Urbaszek . Strain tuning of optical emission energy and polarization in monolayer and bilayer MoS2. Phys. Rev. B, 2013, 88(12): 121301
CrossRef ADS Google scholar
[29]
F. Ullah , J. H. Lee , Z. Tahir , A. Samad , C. T. Le , J. Kim , D. Kim , M. U. Rashid , S. Lee , K. Kim , H. Cheong , J. I. Jang , M. J. Seong , Y. S. Kim . Selective growth and robust valley polarization of bilayer 3R-MoS2, ACS Appl. Mater. & Inter., 2021, 13(48): 57588
[30]
J. K. Ellis , M. J. Lucero , G. E. Scuseria . The indirect to direct band gap transition in multilayered MoS2 as predicted by screened hybrid density functional theory. Appl. Phys. Lett., 2011, 99(26): 261908
CrossRef ADS Google scholar
[31]
S. Bhattacharyya , A. K. Singh . Semiconductor-metal transition in semiconducting bilayer sheets of transition- metal dichalcogenides. Phys. Rev. B, 2012, 86(7): 075454
CrossRef ADS Google scholar
[32]
P.L. ChristiansenM.P. SrensenA.C. Scott, Nonlinear Science at the Dawn of the 21st Century, Berlin, Heidelberg: Springer, 2000
[33]
M. Hangyo , T. Nagashima , S. Nashima . Spectroscopy by pulsed terahertz radiation. Meas. Sci. Technol., 2002, 13(11): 1727
CrossRef ADS Google scholar
[34]
L.DuvillaretF.GaretJ.L. Coutaz, A reliable method for extraction of material parameters in terahertz time-domain spectroscopy, IEEE J. Sel. Top. Quantum Electron. 2(3), 739 (1996)
[35]
S.NudelmanS.S. Mitra, Optical Properties of Solids, Springer, 1969
[36]
M. Tinkham . Energy gap interpretation of experiments on infrared transmission through superconducting films. Phys. Rev., 1956, 104(3): 845
CrossRef ADS Google scholar
[37]
J.D. Jackson, Classical Electrodynamics, 3rd Ed., Wiley, 1998
[38]
P. Drude . Bestimmung der optischen Constanten der Metalle. Annalen der Physik, 1890, 275(4): 481
CrossRef ADS Google scholar
[39]
N. V. Smith . Classical generalization of the Drude formula for the optical conductivity. Phys. Rev. B, 2001, 64(15): 155106
CrossRef ADS Google scholar
[40]
F.W. HanW.XuL.L. LiC.Zhang, A generalization of the Drude-Smith formula for magneto-optical conductivities in Faraday geometry, J. Appl. Phys. 119(24), 245706 (2016)
[41]
A. Mukhopadhyay , S. Kanungo , H. Rahaman . The effect of the stacking arrangement on the device behavior of bilayer MoS2 FETs. J. Comput. Electron., 2021, 20(1): 161
CrossRef ADS Google scholar
[42]
B. Baugher , H. Churchill , Y. Yang , P. Jarillo-Herrero . Intrinsic electronic transport properties of high-quality monolayer and bilayer MoS2. Nano Lett., 2013, 13(9): 4212
CrossRef ADS Google scholar
[43]
D. Valerini , A. Cretí , M. Lomascolo , L. Manna , R. Cingolani , M. Anni . Temperature dependence of the photo-luminescence properties of colloidal CdSe/ZnS core/shell quantum dots embedded in a polystyrene matrix. Phys. Rev. B, 2005, 71(23): 235409
CrossRef ADS Google scholar
[44]
W. Xu , F. M. Peeters , T. C. Lu . Dependence of resistivity on electron density and temperature in graphene. Phys. Rev. B, 2009, 79(7): 073403
CrossRef ADS Google scholar
[45]
S. Kim , A. Konar , W. S. Hwang , J. H. Lee , J. Lee , J. Yang , C. Jung , H. Kim , J. B. Yoo , J. Y. Choi , Y. W. Jin , S. Y. Lee , D. Jena , W. Choi , K. Kim . High-mobility and low-power thin-film transistors based on multilayer MoS2 crystals. Nat. Commun., 2012, 3(1): 1011
CrossRef ADS Google scholar
[46]
H.Schwarz, Laser Interaction and Related Plasma Phenomena, US: Springer, 1972
[47]
W. Xu , H. M. Dong , L. L. Li , J. Q. Yao , P. Vasilopoulos , F. M. Peeters . Optoelectronic properties of graphene in the presence of optical phonon scattering. Phys. Rev. B, 2010, 82(12): 125304
CrossRef ADS Google scholar
[48]
W. S. Yun , S. W. Han , S. C. Hong , I. G. Kim , J. D. Lee . Thickness and strain effects on electronic structures of transition metal dichalcogenides: 2H-MX2 semiconductors (M=Mo, W; X=S, Se, Te). Phys. Rev. B, 2012, 85(3): 033305
CrossRef ADS Google scholar
[49]
E. Scalise , M. Houssa , G. Pourtois , V. V. Afanas’ev , A. Stesmans . First-principles study of strained 2D MoS2. Physica E, 2014, 56: 416
CrossRef ADS Google scholar
[50]
H. M. Dong , Z. H. Tao , L. L. Li , F. Huang , W. Xu , F. M. Peeters . Substrate dependent terahertz response of monolayer WS2. Appl. Phys. Lett., 2020, 116(20): 203108
CrossRef ADS Google scholar
[51]
H. M. Dong , W. Xu , Z. Zeng , T. C. Lu , F. M. Peeters . Quantum and transport conductivities in monolayer graphene. Phys. Rev. B, 2008, 77(23): 235402
CrossRef ADS Google scholar
[52]
S. H. Zhang , W. Xu , S. M. Badalyan , F. M. Peeters . Piezoelectric surface acoustical phonon limited mobility of electrons in graphene on a GaAs substrate. Phys. Rev. B, 2013, 87(7): 075443
CrossRef ADS Google scholar

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (NSFC) (Grant Nos. U2230122 and U2067207) and Shenzhen Science and Technology Program (No. KQTD20190929173954826). The numerical calculations in this work were conducted at Hefei advanced computing center.

RIGHTS & PERMISSIONS

2023 Higher Education Press
AI Summary AI Mindmap
PDF(5767 KB)

Accesses

Citations

Detail

Sections
Recommended

/