Open Access
Add to BibSonomy
Abstract
Electrically tunable TiN/SiO2/TiN epsilon-near-zero photonic structures with various parameters were fabricated using the reactive DC magnetron sputtering approach. Effective medium approximation was used to predict the optical permittivity of a multilayered TiN/SiO2 metamaterial and guide the design/fabrication. Experimental reflectance measurements for tunable TiN/SiO2/TiN structures were obtained using the ellipsometer technique in the visible and near-infrared spectral ranges. Results show that reflectance for biased (12 V) and un-biased bulk TiN/SiO2/TiN structure changes up to ∼ 2% with the spectral shift at the ENZ spectral point ∼ 10 nm for samples with an optimal SiO2 dielectric layer (thickness d=10 nm). Reflectance measurements for multilayered tunable TiN/SiO2/TiN structures show strong variation in reflectance change for s- polarized light at epsilon-near-zero wavelengths due to applied voltage (12 V). We expect that the results of this research study of the tunable TiN/SiO2/TiN epsilon-near-zero photonic structures will potentially be useful for the photonic density of states engineering, surface sensing, and metamaterial-based super-resolution imaging.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The recent development of nanofabrication techniques has enabled the creation [1] of smaller, more compact, photonic structures, which are used in aircraft, medical and telecommunications equipment. Achieving device miniaturization due to enhanced light-matter interaction is possible via tightly confining light using epsilon-near-zero (ENZ) effects [2,3]. Light can be squeezed deeply into sub-wavelength spaces and reach high field intensities at a material’s ENZ wavelength [4]. Therefore, photonic devices based on the ENZ metamaterials show potential for precise control of linear and nonlinear optical interactions [5,6].
Titanium nitride (TiN), one of the most common transition metal nitrides, has been shown to be a promising material for plasmonics [7]. The ENZ wavelength, which can be tuned (510-645 nm) by varying fabrication parameters, is in the visible spectral range [8]. TiN films have been experimentally demonstrated to withstand high temperatures and optical intensities making them useful for applications in nonlinear optics. Nanofabrication methods, such as DC magnetron sputtering and pulsed laser deposition (PLD), can be utilized to fabricate high quality thin TiN films [9,10]. As a result, the TiN films have been shown to be a promising material for linear and nonlinear plasmonic photonic devices for high-temperature applications due to the high melting point of titanium nitride and a relatively high damage threshold compared to metal films.
Designing optical properties of the multilayered stacks of transparent conductive oxides (TCOs) and dielectrics has been used to tune the broadband optical permittivity. Optical permittivity for multilayered metamaterials based on TCO often exhibit dramatically different behavior for in-plane and out-plane components in near-IR and visible spectral range. For example, an indium-tin-oxide (ITO)/SiO2 multilayered structure was designed and analyzed for tuning of the ENZ wavelength in near IR spectral range [11]. Variation in PLD and atomic layer deposition (ALD) fabrication parameters were utilized to tune ENZ wavelength for out-of-plane optical permittivity for aluminum-doped zinc oxide multilayered metamaterials [12–14]. The permittivity of the conductor-dielectric metamaterials with uniaxial permittivity tensor can be extracted using either polarized reflectometry or spectroscopic ellipsometry techniques [15]. Overall sample thickness for aluminum-doped zinc oxide multilayered metamaterials can potentially dramatically affect optical properties and spectral position of the ENZ wavelength [13]. While tunability of these structures via thickness variation and/or tailoring TCO optical properties has been suggested, the dynamic tunability has not been widely explored for multilayered TCO-based structures.
Optical properties of a multilayered ENZ material are fixed due to fabrication conditions [13,14,11] and did not allow ordinarily for dynamic tunability of reflected and transmitted wave amplitude which limits their use in a wide variety of applications. Voltage control could offer a necessary degree of freedom to tailor the optical properties of the multilayered epsilon-near-zero metamaterial in “real time”. The general approach in modification of the refractive index of optical materials via applying electric fields (e.g., using the Pockels and Kerr electro-optic effects) has been used in photonics for years [16]. However, only recently electrical switching has been suggested for dynamic control of phased array indium tin oxide (ITO) metasurfaces [17]. Novel design approaches for metasurfaces with electrical tunability were proposed based on nanowires-based metamaterial with metal-oxide semiconductor and metal-insulator-metal configurations [18]. Using ITO as the active metasurface material and a composite hafnia/alumina gate dielectric, a dual-gated active metasurface device was demonstrated [19]. The significant benefits of using metal-insulator plasmonic cavities as the building blocks for the actively tuned metasurfaces were demonstrated [20].
Changing the carrier density, and as a result, refractive index, is a promising way of refractive index control in conductive oxides [21]. An applied voltage can modify the carrier density of the TCO layer across the insulator/TCO interface. As the carrier density increases, the TCO layer becomes more metallic like. According to an estimate based on effective medium calculations, this carrier density increase will cause a shift of the topological transition point. This wavelength shift can be extracted using the ellipsometry technique. For example, a recent theoretical study of ENZ metamaterial consisting of two 20 nm Ag layers separated by a 15 nm layer of ITO showed that the epsilon-near-pole wavelength can be blue-shifted under applied bias by more than 60 nm [22]. The tunability of the plasma frequency of an ITO thin film by changing its carrier concentration via gating with an ionic liquid has been demonstrated experimentally [23]. The exact control of the change in carrier density requires detailed theoretical model and experimental analysis both for TCO and multilayered TCO/dielectric structures.
In this work we investigate active tuning of epsilon-near-zero TiN/SiO2 photonic structures in visible spectral range. The effect of carrier density changes in TiN accumulation layer on the optical parameters is investigated. We study potential tuning mechanisms such as modifying complex optical permittivity via field effect gating to electrically modulate the permittivity in conductive TiN layers. This approach has potential to provide power-efficient ultrafast tunable optical devices.
2. Concept and theory
2.1 Field effect for tuning
Charge carrier injection can be modeled via a uniform transparent conductive oxide layer with increased carrier concentration. Based on an electrostatic calculation on a dielectric/conductor interface the applied voltage V and thickness of the dielectric d are related to the breakdown field Ebr in dielectric via [22]:
where kdiel is the DC dielectric constant of the dielectric, ɛo is the vacuum permittivity, e is the charge of the electron, l is the Debye length, and N is the carrier density in the accumulation layer. To achieve tunability of the optical permittivity within the visible regime, the accumulation layer incorporated into the photonic device must have a plasma frequency within the visible spectrum which leads to a high background carrier concentration (e.g., initial carrier concentration for un-biased TiN layer is ∼ 1022 cm −3). As a result, the dielectric materials with high electrical breakdown fields must be used. For our design, SiO2 material was chosen which has electrical breakdown field Ebr = (30−40) MV/cm [24]. Based on Eq. (1), the breakdown field Ebr = 30 MV/cm corresponds to applied voltage V = 30 V for SiO2 layer with thickness d = 10 nm. The maximum achievable carrier concentration for TiN layer N for reported value of breakdown field Ebr =30 MV/cm is estimated ∼1.29 × 1021 cm −3 (for SiO2 layer kdiel = 3.9 [22], Debye length l = 0.5 nm) using Eq. (1). For the experimental implementation applied voltage was selected below the breakdown voltage (V = 12 V). According to Eq. (1), maximum achievable carrier concentration in TiN layer will be proportional to the applied voltage V.Design and fabrication of a tunable TiN/SiO2 photonic structure using DC reactive magnetron sputtering was performed as described in Sec. 3. The tunable TiN/SiO2 photonic structure consists of the ENZ multilayered TiN/SiO2 metamaterial on top (Fig. 1) and conductive TiN layer at the bottom. Pulsed DC reactive magnetron sputtering was used to deposit a thin insulating layer of SiO2 (Fig. 1) of thickness d between the top and the bottom layer. Bulk TiN layer was used as a top layer instead of the multilayered structure to investigate performance of the device for various SiO2 thicknesses d (Fig. 1(a)). The TiN fill fraction for the multilayered metamaterial (Fig. 1(b)) was varied (40-60%) to adjust initial ENZ spectral point to specific wavelength range. The structure presented in Fig. 1 was deposited on silicon or quartz substrates.
Fig. 1. Sketch of electrically tunable bulk TiN (a) and multilayered TiN/SiO2 (b) photonic structure.
2.2 Optical permittivity model for multilayered TiN/SiO2
We used first-order Effective Medium Approximation (EMA) to model the optical permittivity of the TiN/SiO2 nano-layered structure. The multilayered TiN/SiO2 is assumed to be a uniaxial material with different values for the in-plane dielectric permittivity in the parallel direction, ɛ||, and the out-of-plane permittivity in the perpendicular direction, ɛ⊥. According to EMA, in-plane permittivity, ɛ||, and out-of-plane permittivity, ɛ⊥, can be described:
2.3 Spectral shift in reflectance due to change in carrier concentrations for the bulk TiN layer
A theoretical approach to the description of the spectral shift in reflectance for the electrically tunable bulk TiN layer was developed by following several steps. First, TiN optical permittivity model was used (Eq. (3)) to calculate ENZ wavelength for unbiased bulk TiN film sample and initial carrier concentration. As the carrier density increases due to applied voltage, we expect a shift of the ENZ spectral point toward higher energies. Therefore, the optical permittivity for the biased sample was calculated next. Finally, the reflectance for s- and p- polarized light was calculated based on optical permittivity for the biased and un-biased sample.
Optical permittivity for fabricated bulk TiN sample was modeled using the Drude-Lorentz model with one oscillator [27–29]:
The calculated optical permittivity for TiN layer based on Eq. (2) and various carrier’s concentrations due to applied voltage is presented in Fig. 2. The calculated initial (unbiased) ENZ wavelength was ∼547 nm. As expected, as the carrier concentration increases\decreases due to applied bias, the ENZ wavelength shifts to shorter/longer wavelength. E.g., injection of the 2.4% of the initial carrier concentration N0 can lead to up to ∼6 nm ENZ wavelength shift (Table 1). The optical permittivity imaginary component remains approximately the same Im(ɛ)∼5.37. This resulting ENZ wavelength shift resulted in reflectance shift for s-polarized light and p-polarized light (Fig. 3) as described below.
Fig. 2. Calculated permittivity (real (left) and imaginary (right) parts) vs wavelength for various resulting spectral shifts in s-polarized reflectance of the ENZ wavelength (carrier concentrations indicated in Table 1).
Fig. 3. Calculated s-polarized (left) and p- polarized (right) light reflectance spectra (angle of incidence: 60° for various resulting spectral shifts of the ENZ wavelength (carrier concentrations indicated in Table 1 and Table 2). The position of calculated initial (unbiased) ENZ wavelength for top TiN layer (∼547 nm) is indicated.
Table 1. Carrier concentration (N0), plasma frequency (ωp), ENZ wavelength (λ ENZ) for permittivity, and carrier concentration change (ΔN) for corresponding spectral shifts in reflectance for s-polarized light.
Table 2. Carrier concentration (N0), plasma frequency (ωp), ENZ wavelength (λ ENZ) for permittivity, and carrier concentration change (ΔN) for corresponding spectral shifts in reflectance for p-polarized light.
The reflectance for s- and p- polarized light was calculated using Fresnel equations [16] for TiN top layer based on the permittivity data modeled using Eq. (3). The angle of incidence (60°) was chosen based on the experimental parameters for the ellipsometer. Then using best-fit Drude-Lorentz parameters (Eq. (3)), we found the necessary increase in carrier concentration for TiN layer (and corresponding “new” plasma frequency) for corresponding spectral shift in reflectance (Table 1 and Table 2). We found that reflectance for p-polarization exhibits spectral shift at lower carrier concentration. E.g., for the spectral shift of 10 nm in s- polarized light reflectance the change in carrier concentration ΔN is increased by ∼ 20% compared to p-polarized light (Table 1, Table 2).
The corresponding calculated reflectance spectra for each spectral shift due to carrier concentration change are shown in Fig. 3. Both s-polarized and p-polarized light reflectance curves have characteristic minimum at shorter wavelengths (see inserts in Fig. 3) due to sign change in real part of optical permittivity. In general, the magnitude for the s-polarized light reflectance is higher across the investigated wavelength range (400-2000 nm).
2.4 Spectral shift in reflectance for the multilayered TiN/SiO2 photonic structure
The spectral shift in reflectance for the electrically tunable multilayered TiN/SiO2 photonic structures was obtained by following several steps. First, in-plane, ɛ||, and out-of-plane, ɛ⊥, optical permittivity for the un-biased multilayered structure with various TiN fill fraction were calculated using TiN optical permittivity model for unbiased bulk TiN film sample (Eq. (3)) and effective medium approximation (Eq. (2)). Next, the optical permittivity for the biased multilayered TiN/SiO2 sample was calculated using optical permittivity for biased bulk TiN layer with adjusted plasma frequency (ωp), given in Table 1. Finally, the reflectance for s- polarized light was calculated based on optical permittivity for the biased and un-biased multilayered TiN/SiO2 photonic structures.
The calculated in-plane, ɛ||, (solid line) and out-of-plane, ɛ⊥, (dashed line) real part of optical permittivity of TiN/SiO2 multilayered films with various TiN fill fractions based on effective medium approximation (Eq. (1)) is presented in Fig. 4. Optical permittivity for TiN individual layer was modeled using the Drude-Lorentz model with parameters described above (Eq. (3)). As expected, the calculated initial (unbiased) ENZ wavelength for in-plane optical permittivity shifts toward shorter wavelengths as the TiN fill fraction increases. E.g., ENZ wavelength for in-plane permittivity Re(ɛ||) is calculated 681, 636, and 601 nm for multilayered TiN/SiO2 films with 33%, 50%, and 60% TiN fill fractions respectively (Fig. 4). The calculated imaginary part of the in-plane optical permittivity at ENZ wavelength is increased as well: Im(ɛ||) = 4.93 (33%), Im(ɛ||) = 5.49 (50%), and Im(ɛ||) = 6.43 (60%).
Fig. 4. Calculated in-plane (solid line) and out-of-plane (dashed line) real part of optical permittivity vs wavelength for multilayered TiN/SiO2 multilayered films with various indicated TiN fill fractions.
The s- polarized light reflectance for multilayered TiN/SiO2 photonic structures was calculated using Fresnel equations [16]. The insulating layer of SiO2 (thickness 10 nm) and quartz substrate were both included in the calculations. The unbiased permittivity of TiN individual layer in multilayered structure was modeled using Eq. (3). using best-fit Drude-Lorentz parameters (Fig. 4). The maximum increase in carrier concentration (ΔN = 0.124 × 1022 cm−3) for TiN individual layer in multilayered TiN/SiO2 structure (and corresponding “new” plasma frequency) was used to model corresponding spectral shift in reflectance for biased multilayered TiN/SiO2 structure (Table 1). As expected, as the carrier concentration increases due to applied bias, the ENZ wavelength shifts to shorter wavelengths. E.g., injection of the additional carrier concentration ΔN leads to 6.5 nm, 6.4 nm, and 6.3 nm ENZ wavelength shift toward shorter wavelengths for multilayered TiN/SiO2 structurers with TiN fill fraction 33%, 50%, and 60% (Fig. 5). While the ENZ spectral point shifts toward shorter wavelengths as the carrier concentration in individual TiN layer increases, the reflectance at ENZ wavelength and spectral shift due to applied bias is higher for multilayered TiN/SiO2 structures with lower TiN fill fraction (Fig. 5(a)).
Fig. 5. Calculated s-polarized light reflectance spectra (angle of incidence: 60°) for resulting spectral shifts of the ENZ wavelength (maximum carrier concentration change indicated in Table 1) for multilayered TiN/SiO2 samples with various fill fractions (number of TiN/SiO2 periods N = 15). Insert: close-up of the main reflection curve.
The results of the calculations for s-polarized light reflectance for multilayered TiN/SiO2 structures (TiN fill fraction 50%) with various number of layers N are presented in Fig. 6. The results show that the overall shape for the reflectance curve remained similar (Fig. 6(insert)) for 400-1600 nm wavelength range. The magnitude for the s-polarized light reflectance for the un-biased samples at ENZ wavelength remains relatively similar (∼ 3% variation between samples with N = 10 and N = 30). The expected spectral shift in reflectance curve at the ENZ spectral point (636 nm) due to applied bias has slight dependence on the number of the layers (∼ 5 nm for N = 10, ∼ 6 nm for N = 20, and ∼ 7 nm for N = 30).
Fig. 6. Calculated s-polarized light reflectance spectra (angle of incidence: 60°) for resulting spectral shifts of the ENZ wavelength (maximum carrier concentration change indicated in Table 1) for multilayered TiN/SiO2 samples with various number of TiN layers indicated (TiN fill fraction 50%). Insert: calculated reflectance curve for the un-biased sample (N = 20). The position of calculated initial (unbiased) ENZ wavelength for multilayered top layer (636 nm) is indicated.
3. Fabrication methods and experimental characterization
3.1 Fabrication and deposition approaches
DC reactive magnetron sputtering was used to deposit the multilayered TiN/SiO2 metamaterial on quartz and silicon substrates. The fabrication environment was controlled to allow precise tailoring of the film's properties. The TiN layers were deposited at room temperature, with the only surface heating resulting from the magnetron sputtering plasma. The chamber base pressure before beginning deposition was ∼1.5 × 10−7 Torr. Two inch diameter Ti and Si magnetron sputtering targets were used in clamp style magnetron sputtering guns from Kurt J. Lesker. For TiN deposition a MKS RPDG-50 power supply was used in a DC power mode, ramping the output power to 150W, with a pressure of ∼2.8 mTorr, 7 SCCM N2 mixed with 13 SCCM Ar before entering the chamber. The pressure was controlled using a VAT valve having a fixed position to minimize pressure oscillations during growth. The Ti voltage was ∼433 Vdc, and a current of ∼ 0.34 A.
For SiO2 layer an Advanced Energy Pinnacle Plus Pulsed DC power supply was used to reactively sputter Si with a constant current mode of 0.4A, 80 kHz, 4.5 µsec; gas flow of 10 SCCM Ar through the Si sputtering gun, 10 SCCM Ar through the Ti sputtering gun, and 4 SCCM O2 into the chamber with the goal being to minimize Si target poisoning and minimizing Si deposition onto the Ti target. The Si target voltage was ∼359 V, ∼144 W, ∼0.4 A. A pneumatic shutter was actuated between each layer’s growth while the reactive sputtering process was ramped to steady state conditions over ∼ 3 minutes. The in-situ J.A. Woollam M-2000 spectroscopic ellipsometer (370-1670 nm detection range) was used to optimize deposition parameters. Ellipsometry measurements were continuously made during the deposition process (Fig. 7).
Fig. 7. Ellipsometry amplitude parameter Ψ vs. deposition time for indicated wavelengths for multilayered TiN/SiO2 sample with N = 15 periods.
During the growth, an initial layer of TiN was deposited first. The sample was then removed through a load-lock into a 6-sample cassette, and the load-lock vented to remove the cassette, and a mask was applied to establish ground contact for the structure’s capacitors. After re-pumping the load-lock chamber, the sample was inserted back into the chamber to grow the dielectric SiO2 layer and top TiN layer (Fig. 1), which established the structure’s top electrode. Once the growth was completed, the samples were cleaned to prepare for characterization. The total targeted thickness of the samples (100-300 nm), the TiN/SiO2 deposition ratio, and the number of the layers (10-30) were varied to investigate the optimal parameters for a tunable metamaterial.
3.2 Transmission electron microscopy characterization
Transmission electron microscopy characterization of tunable TiN/SiO2 photonic structures on quartz substrate were made using a lamella prepared using focused ion beam (FIB) technique. The resulting sample has received a 5 kV Ga ion beam clean. After the initial imaging and EDS maps the lamella was cleaned in the Fischione Nanomill low energy argon ion mill with 1 kV beam. Reduction in the amorphous layer resulted in good quality images presented in Fig. 8.
Fig. 8. (a) Cross-sectional TEM images of the bulk TiN/SiO2 photonic structure. Low (b) and high (c) magnification cross-sectional STEM images of the multilayered TiN/SiO2 photonic structure (TiN number of layers: 16; SiO2 number of layers: 15; 43% TiN fill fraction). Energy dispersive X-ray spectroscopy (EDS) elemental mapping images of the multilayered TiN/SiO2 photonic structure (d). The corresponding STEM image of the Si, Ti, O, N mapping area (e).
Results of the transmission electron microscopy (TEM) characterization of bulk TiN/SiO2 photonic structures are shown in Fig. 8(a). The structure (see sketch in Fig. 1(a)) was designed for 30 nm TiN bottom layer, 10 nm SiO2 insulating middle layer, and 20 nm TiN top layer. The measurements show clear 29.3 nm thickness for the bottom TiN layer and 10-11.9 nm thickness of the SiO2 insulating middle layer which is consistent with target thicknesses.
The results (Fig. 8(c)) of the scanning transmission electron microscopy (STEM) characterization of the multilayered TiN/SiO2 photonic structures (see sketch in Fig. 1(b)) show that the top layer has a clear multilayered structure. The uniform bottom layer of TiN has a measured thickness of 27.2 nm and insulating layer of SiO2 has thickness d ∼10 nm. Individual layer thickness measurements for top multilayered TiN/SiO2 structure show a clear structure for the TiN layers of 6.43 nm thickness and SiO2 layers of 4.92 nm. Based on those measurements, the calculated TiN fill fraction for the top multilayered film is ∼43% as originally designed. The overall thickness of the top multilayered TiN/SiO2 film was measured 189 nm.
High spatial resolution compositional maps and line profiles were obtained in the STEM mode by collecting the EDS signals point-by-point across the TiN/SiO2 photonic structures (Fig. 8(d,e)). An elemental mapping using energy dispersive X-ray spectroscopy (EDS) clearly show the periodic spatial distribution of titanium (Ti), nitrogen (N), silicon (Si), and oxygen (O) elements, indicating the periodicity of the top layer and uniformity of the bottom TiN layer and SiO2 insulating layer.
3.3 Electron energy loss spectra (EELS) measurements
The analysis of the plasmon loss and core loss for the sample’s area shaded red (Fig. 9(a)) are presented in Fig. 9. Cross-sectional specimens (Fig. 9(a)) of the TiN film were prepared using a focused ion-beam technique. Measurements have been performed using TITAN TEM 80-300 equipped with an image aberration corrector. Images were acquired using the high angle annular dark field (HAADF) scanning transmission electron microscope technique with an energy resolution of ∼ 1eV and spatial resolution of ∼ 1.5 Angstroms. The electron energy loss (EEL) spectra were acquired using 300 kV primary energy.
Fig. 9. (a) Cross-sectional high-angle annular dark-field image and core-loss (b) EELSs for bulk TiN sample of 30 nm thickness on Si substrate. Experimental energy-loss near-edge structure (ELNES) for titanium (c) and nitrogen (d).
The elemental composition of the sample was investigated using core-loss EELS of the Ti-L and N-K edges. The peaks are located at approximately 461.15 eV and 398 eV respectively (Fig. 9(c-d)) which is consistent with current literature [33,34]. The sample composition at investigated location was found to be 47 at. % Ti and 53 at. % N. Only a very weak oxygen peak was detected in core-loss EELS (Fig. 9(b)).
3.4 Ellipsometry and reflectance measurements
The RC2 J.A. Woollam spectroscopic ellipsometer was used to characterize linear optical properties for TiN/SiO2 photonic structures. To investigate active tunability, the ellipsometry and reflectance measurements were done first for un-biased samples. Then, an electrical probe was placed on the ground and top contacts. The DC square pulse voltage was supplied from a function generator via connected probes. A simple electrostatic calculation of the conductor/dielectric interface predicts that applied voltage V must satisfy: V/t∼Ebd (t is thickness of the dielectric layer). Typical breakdown field Ebd for SiO2 dielectric layer is 30−40 MV/cm [22]. Therefore, we limited investigated voltages to 12 V to avoid damaging the samples. The results for bulk and multilayered TiN/SiO2 photonic structures are presented in Sec. 4 and Sec. 5.
A typical complex optical permittivity deduced from ellipsometry measurements for TiN film (52 nm thickness) deposited on quartz substrates is presented in Fig. 10(a). The ENZ wavelength is achieved at 549 nm with Im(ɛ) = 5.2. Several samples using deposition parameters described in Sec. 3.1. were fabricated to ensure reproducibility. The four-point probe measurements show the impedance of the TiN film (52 nm thickness) ∼ 33 Ω/sq. The reflectance measurements for light incident angles (30°-85°) are shown in Fig. 10(b). One can see that the reflectance curve experiences a minimum with sharp increase of reflectivity in visible spectral range. Reflectance curves have similar shape for incident angles (30°-60°). The incident angle 60° was chosen for bulk and multilayered TiN/SiO2 tunable photonic structures to investigate anisotropy in s-polarized and p-polarized light reflectance.
Fig. 10. Complex optical permittivity deduced from ellipsometry measurements (a) and measured reflectance at various angles of incidence (b) for TiN film (52 nm thickness) deposited on quartz substrates.
4. Bulk TiN/SiO2/TiN tuning: results and discussion
The bulk TiN/SiO2/TiN ENZ photonic structure was used to investigate feasibility of the gate tuning and optimal parameters. The tunable bulk TiN/SiO2/TiN photonic structure consists of the conductive bulk TiN material on top (thickness 30 nm) and at the bottom (thickness 20 nm) with an insulating SiO2 layer in-between (Fig. 1(a)). The thickness of the insulating SiO2 layer was varied to find optimal conditions for maximum spectral shift in reflectance. Based on Eq. (1), we expect that decreasing the thickness of the insulating SiO2 layer for the fixed applied voltage (12 V) would lead to enhanced spectral shift for the ENZ wavelength.
Measured reflectance spectra (angle of incidence: 60 deg, non-polarized light) for biased (12 V) and un-biased bulk TiN/SiO2/TiN structure for various SiO2 thicknesses (d) are shown in Fig. 11. Calculations from the reflection spectra show that the reflectance changes up to ∼1.7% with spectral shift at the ENZ spectral point 9.5 ± 0.5 nm for samples with SiO2 dielectric layer thickness d = 10 nm (Fig. 11(b)). Increasing thickness of the dielectric SiO2 layer to d = 15 nm leads to a decrease of the magnitude of the reflectance change 0.7% as expected for fixed applied voltage (Fig. 11(c)) with corresponding spectral shift at the ENZ spectral point 4.0 ± 0.5 nm. We found that decreasing the thickness of the dielectric SiO2 layer to d = 5 nm leads to decrease of the magnitude of the reflectance change ∼0.5% because of the thickness non-uniformity of the SiO2 insulating layer due to reactive magnetron sputtering fabrication limitations. The corresponding spectral shift at the ENZ spectral point for samples with d = 5 nm is 4.5 ± 0.5 nm (Fig. 11(a)). Therefore, thickness of the dielectric SiO2 layer d = 10 nm was found to be optimal for future device design.
Fig. 11. Measured reflectance spectra (angle of incidence: 60 deg, non-polarized light) for biased (12 V) and un-biased bulk TiN/SiO2/TiN structure for various indicated SiO2 thicknesses (d). The position of initial (unbiased) ENZ wavelength for top TiN layer deduced from ellipsometry is indicated.
Carrier concentration in accumulation layer for TiN/SiO2/TiN structure was estimated based on the ENZ wavelength spectral position shift using Eq. (4). Spectral shift due to applied voltage leads to a dramatic change in carrier concentration for TiN/SiO2/TiN structures of various thicknesses d (Table 1 and Table 2).
5. Design and ellipsometry measurements of a tunable multilayered TiN/SiO2 photonic structure
A multilayered TiN/SiO2 photonic structure for voltage control of ENZ transition was grown using reactive DC magnetron sputtering (Fig. 1(b)). As the carrier density of the TiN layer increases due to applied voltage across the dielectric/TiN interface, the TiN/SiO2 metamaterial becomes more “metallic like”. According to an estimate based on the permittivity model, this carrier density increase will cause a shift of the ENZ spectral point toward higher energies (Fig. 5, Fig. 6). Various multilayered TiN/SiO2 metamaterials were fabricated using reactive DC magnetron sputtering. The number of layers (N = 4-30) and TiN fill fraction (30-60%) were varied. The contacts were implemented to the structure the same as described in Sec. 3.1.
The RC2 J.A. Woollam spectroscopic ellipsometer was used to obtain reflectance measurements. Measured reflectance spectra (angle of incidence: 60 deg, s-polarized light) for biased (12 V) and un-biased multilayered TiN/SiO2 structures with various indicated TiN fill fractions are shown in Fig. 12. Reflectance measurements show variation in reflectance change for s- polarized light at ENZ wavelengths due to applied voltage (12 V) (Fig. 12(a)). The spectral shift in s-polarized light reflectance at ENZ spectral point in case of the multilayered TiN/SiO2 samples decreases as the TiN fill fraction increases to ∼60% (Fig. 12(b)). This behavior is related to change of the slope of the reflectance curves for samples with various reflectance. The reflectance magnitude for samples with higher TiN fill fraction is increased as expected from theoretical model (Fig. 5).
Fig. 12. (a) Measured reflectance spectra (angle of incidence: 60 deg, s-polarized light) for biased (12 V) and un-biased multilayered TiN/SiO2 structures with various indicated TiN fill fractions. (b) Measured spectral shift in s-polarized reflectance due to applied bias (12 V) at ENZ wavelength for each sample vs. TiN fill fraction.
Measured reflectance spectra (angle of incidence: 60 deg, s-polarized light) for biased (12 V) and un-biased multilayered TiN/SiO2 structures with various indicated numbers of TiN/SiO2 periods are shown in Fig. 13. As expected from the theoretical model (Fig. 6), reflectance magnitude remains similar as the number of the TiN/SiO2 periods varied N = 20-30 (Fig. 13(a)). The expected spectral shift for s-polarized light reflectance due to applied bias is ∼ 5-7 nm based on the model (Fig. 6). The experimental reflectance measurements show spectral shift of ∼ 6 nm for samples N = 10-15 periods. However, the spectral shift is significantly reduced for multilayered TiN/SiO2 structures with higher number of periods (N = 20-30). This drop in magnitude for spectral shift is related to a decrease in efficiency for carrier injection for samples with higher number of layers and total thickness. A small discontinuity in the reflection measurements at ∼ 1000 nm (Fig. 12(a), Fig. 13(a)) is related to the detector switch at ∼1000 nm between main spectral range and near-IR extension for RC2 Woolam ellipsometer.
Fig. 13. (a) Measured reflectance spectra (angle of incidence: 60 deg, s-polarized light) for biased (12 V) and un-biased multilayered TiN/SiO2 structures with various indicated numbers of TiN/SiO2 periods. (b) Measured spectral shift in s-polarized reflectance due to applied bias (12 V) at ENZ wavelength (636 nm) vs the number of TiN/SiO2 periods.
6. Conclusions
In conclusion, high quality TiN-based ENZ metamaterial was fabricated using reactive DC magnetron sputtering approach. Electrically tunable TiN/SiO2/TiN epsilon-near-zero photonic structures with various parameters were designed. Theoretical model based on effective medium approximation predicts ENZ wavelength for optical permittivity of the multilayered TiN/SiO2 metamaterials. Spectral reflectance dependence based on this model exhibits ∼10 nm spectral shift due to applied bias. Experimental reflectance measurements for tunable TiN/SiO2/TiN structures, obtained by an ellipsometry technique in the visible and near-infrared spectral ranges, show that reflectance for biased (12 V) and un-biased bulk TiN/SiO2/TiN structure changes up to ∼ 2% with spectral shift at the ENZ spectral point ∼ 10 nm for samples with optimal SiO2 dielectric layer (thickness t = 10 nm). Reflectance measurements for multilayered tunable TiN/SiO2/TiN structures show strong variation in reflectance change for s- polarized light at near-IR and ENZ wavelengths due to applied voltage (12 V). Multilayered ENZ TiN/SiO2/TiN structures with TiN fill fraction ∼ 50% and period number N = 15 have the most spectral shift (∼ 6 nm). We expect that the results of this research study of the tunable TiN/SiO2/TiN epsilon-near-zero photonic structures will potentially be useful for the photonic density of states engineering, surface sensing, metamaterial-based gate-tunable optical filters, and ultrafast optical applications.
Funding
Lowell Burnett Foundation (Student Fellowship); Air Force Office of Scientific Research (Summer Faculty Fellowship Program).
Disclosures
The authors declare no conflict of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. A.F. Koenderink, A. Alù, and A. Polman, “Nanophotonics: shrinking light-based technology,” Science 348(6234), 516–521 (2015). [CrossRef]
2. M. Silveirinha and N. Engheta, “Tunneling of electromagnetic energy through subwavelength channels and bends using ɛ-near-zero materials,” Phys. Rev. Lett. 97(15), 157403 (2006). [CrossRef]
3. V. J. Sorger, N. D. Lanzillotti-Kimura, R.-M. Ma, et al., “Ultra-compact silicon nanophotonic modulator with broadband response,” Nanophotonics 1(1), 17–22 (2012). [CrossRef]
4. S. Campione, D. de Ceglia, M. A. Vincenti, et al., “Electric field enhancement in ɛ-near-zero slabs under TM-polarized oblique incidence,” Phys. Rev. B 87(3), 035120 (2013). [CrossRef]
5. A. Capretti, Y. Wang, N. Engheta, et al., “Enhanced third-harmonic generation in Si-compatible epsilon-near-zero indium tin oxide nanolayers,” Opt. Lett. 40(7), 1500 (2015). [CrossRef]
6. M. A. Vincenti, M. Kamandi, D. de Ceglia, et al., “Second-harmonic generation in longitudinal epsilon-near-zero materials,” Phys. Rev. B 96(4), 045438 (2017). [CrossRef]
7. G. V. Naik, V. M. Shalaev, and A. Boltasseva, “Alternative plasmonic materials: beyond gold and silver,” Adv. Mater. 25(24), 3264–3294 (2013). [CrossRef]
8. A. Capretti, Y. Wang, N. Engheta, et al., “Comparative study of second-harmonic generation from epsilon-near-zero indium tin oxide and titanium nitride nanolayers excited in the near-infrared spectral range,” ACS Photonics 2(11), 1584–1591 (2015). [CrossRef]
9. B. Majora, W. Mrózb, T. Wierzchońc, et al., “Pulsed laser deposition of advanced titanium nitride thin layers,” Surf. Coat. Technol. 180-181, 580–584 (2004). [CrossRef]
10. S. Grosso, L. Latu-Romain, G. Berthomé, et al., “Titanium and titanium nitride thin films grown by dc reactive magnetron sputtering,” Surf. Coat. Technol. 324, 318–327 (2017). [CrossRef]
11. A. Pianelli, V. Caligiuri, M. Dudek, et al., “Active control of dielectric singularities in indium-tin-oxides hyperbolic metamaterials,” Sci. Rep. 12(1), 16961 (2022). [CrossRef]
12. P. Kelly, M. Liu, and L. Kuznetsova, “Designing optical metamaterial with hyperbolic dispersion based on an Al:ZnO/ZnO nano-layered structure using the atomic layer deposition technique,” Appl. Opt. 55(11), 2993–2997 (2016). [CrossRef]
13. P. Kelly, W. Zhang, M. Liu, et al., “Engineering the structural, plasmonic, and optical properties of multilayered aluminum-doped zinc oxide metamaterial grown by pulsed laser deposition,” Appl. Opt. 58(21), 5681–5686 (2019). [CrossRef]
14. C. T. Riley, T. A. Kieu, J. S. T. Smalley, et al., “Plasmonic tuning of aluminum doped zinc oxide nanostructures by atomic layer deposition,” Phys. Status Solidi RRL 8(11), 948–952 (2014). [CrossRef]
15. T. Tumkur, Y. Barnakov, S. T. Kee, et al., “Permittivity evaluation of multilayered hyperbolic metamaterials: ellipsometry vs. reflectometry,” J. Appl. Phys. 117(10), 103104 (2015). [CrossRef]
16. E. Hecht, Optics (Addison-Wesley, 2002).
17. Y.-W. Huang, H. W. Howard Lee, R. Sokhoyan, et al., “Gate-tunable conducting oxide metasurfaces,” Nano Lett. 16(9), 5319–5325 (2016). [CrossRef]
18. M. Mahdi Salary and Hossein Mosallaei, “Electrically tunable metamaterials based on multimaterial nanowires incorporating transparent conductive oxides,” Sci. Rep. 7(1), 10055 (2017). [CrossRef]
19. G. Shirmanesh, R. Sokhoyan, R. A. Pala, et al., “Dual-gated active metasurface at 1550 nm with wide (>300°) phase tunability,” Nano Lett. 18(5), 2957–2963 (2018). [CrossRef]
20. J. Park, J.-H. Kang, X. Liu, et al., “Electrically tunable epsilon-near zero metafilm absorbers,” Sci. Rep. 5(1), 15754 (2015). [CrossRef]
21. E. Feigenbaum, K. Diest, and H. A. Atwater, “Unity-order index change in transparent conducting oxides at visible frequencies,” Nano Lett. 10(6), 2111–2116 (2010). [CrossRef]
22. G. T. Papadakis and H. A. Atwater, “Field-effect induced tunability in hyperbolic metamaterials,” Phys. Rev. B 92(18), 184101 (2015). [CrossRef]
23. X. Liu, J.-H. Kang, H. Yuan, et al., “Tuning of plasmons in transparent conductive oxides by carrier accumulation,” ACS Photonics 5(4), 1493–1498 (2018). [CrossRef]
24. C. Sire, S. Blonkowski, M. J. Gordon, et al., “Statistics of electrical breakdown field in HfO2 and SiO2 films from millimeter to nanometer length scales,” Appl. Phys. Lett. 91(24), 242905 (2007). [CrossRef]
25. M. Schubert, Infrared Ellipsometry on Semiconductor Layer Structures (Springer, 2004).
26. H. Fujiwara, Spectroscopic Ellipsometry: Principles and Applications (Wiley, 2007).
27. S. R. Loria, B. R. Bricchi, A. Schirato, et al., “Unfolding the origin of the ultrafast optical response of titanium nitride,” Adv. Opt. Mater. 11, 2300333 (2023). [CrossRef]
28. D. Gall, I. Petrov, and J. E. Greene, “Epitaxial Sc1-xTixN (001): Optical and electronic transport properties,” J. Appl. Phys. 89(1), 401–409 (2001). [CrossRef]
29. L. Roux, J. Hanus, J. C. Francois, et al., “The optical properties of titanium nitrides and carbides: spectral selectivity and photothermal conversion of solar energy,” Sol. Energy Mater. 7(3), 299–312 (1982). [CrossRef]
30. C.G.H. Walker, J.A.D. Matthew, C.A. Anderson, et al., “An estimate of the electron effective mass in titanium nitride using UPS and EELS,” Surf. Sci. 412-413, 405–414 (1998). [CrossRef]
31. P. Patsalas and S. Logothetidis, “Optical, electronic, and transport properties of nanocrystalline titanium nitride thin films,” J. Appl. Phys. 90(9), 4725–4734 (2001). [CrossRef]
32. D. Shah, A. Catellani, H. Reddy, et al., “controlling the plasmonic properties of ultrathin tin films at the atomic level,” ACS Photonics 5(7), 2816–2824 (2018). [CrossRef]
33. Y. Kihn, C. Mirguet, and L. Calmels, “EELS studies of Ti-bearing materials and ab initio calculations,” J. Electron Spectrosc. Relat. Phenom. 143(2-3), 117–127 (2005). [CrossRef]
34. A. A. Herzing, U. Guler, X. Zhou, et al., “Electron energy loss spectroscopy of plasmon resonances in titanium nitride thin films,” Appl. Phys. Lett. 108(17), 171107 (2016). [CrossRef]
