Search

Mode size converter enables to match two waveguides with different mode sizes. It has already become an indispensable component in modern integrated photonics. Here, we propose a strategy for implementing an on-chip mode size converter consisted of a gradient geometry metasurface and two cascaded waveguide tapers on the lithium-niobate-on-insulator platform. The metasurface, which is formed by a series of length-graded silica nanostrips embedded into LN waveguide, functions as focusing of incident wave and exaltation of conversion efficiency. A mode size converter that enables to match a fundamental TE 00 mode in a 10-μm-wide waveguide with that in a 0.8-μm-wide one was simulated to verify the feasibility of the strategy. We show that the device is not only compact with a length of 22.9 μm, but also shows an ultra-broad operation band as much as 1150 nm (from 1100 to 2250 nm), over which the conversion efficiency keeps >90 %. In addition, we show that the device has a good robustness to fabrication deviations. The strategy proposed here can be flexibly extended to other waveguide cross-section sizes and other guided modes. • A strategy is proposed for realizing an on-chip ultra-broadband mode size converter. • It uses gradient geometry metasurface and cascaded tapers on LNOI platform. • It is proved by simulating a device connecting a wide waveguide with a narrow one. • The device is valid for on-chip use and has an ultra-broad band as much as 1150 nm. • The strategy has a good scalability, the device has an excellent fabrication tolerance. [ABSTRACT FROM AUTHOR]

Er3+/Yb3+ doped LiNbO 3 (LN) single crystal films with varying Yb3+ doping level were grown on z -cut LiTaO 3 bulk single crystals via using e-beam evaporation method. The crystallographic orientation, elemental distribution, fluorescence properties, ionic valence states, Er3+ concentration, Yb3+ concentration and Li 2 O content of films were carefully studied. Results from XRD, SEM and Raman scattering conclude that the predominant phase of these films is LN and they are oriented along (006). The results of energy dispersive spectroscopy indicate that constituent elements are uniformly distributed over the entirety of the films. The visible and near-infrared luminescence intensities rise with increasing Yb3+-doping concentration. The Li 2 O content of films remains constant at 48.8 mol% with growing Yb3+-doping level based on the optical birefringence method, which is consistent with the Raman results. XPS results suggest that the Er3+ and Yb3+ concentrations in the films are lower than that in the polycrystalline powder since e-beam evaporation has a low efficiency of heavy Er3+ and Yb3+ ions. • The Er3+, Yb3+ doped LiNbO 3 single crystal thin films were grown by the e-beam evaporation method. • The fluorescence properties enhance clearly with a rise in Yb3+-doping concentration. • The accurate Li 2 O content have been evaluated by the birefringence method. • The Er3+ and Yb3+ concentration in the films have been calculated by the XPS. [ABSTRACT FROM AUTHOR]

A post-growth Li-rich vapor transport equilibration (VTE) technique was used to prepare near-stoichiometric Er3+ doped LiNbO 3 thin films. Er3+ fluorescence property, defect feature and Li 2 O content of the films were characterized. The results show that the visible and near-infrared fluorescence intensity increases remarkably by ∼10 times the VTE time is increased from 0 to 5 h. The enhancement can be explained from viewpoint of Nb4+ defects concentration. X-ray photoelectron spectrometer study confirms presence of antisite Nb Li 4+ in the films and Nb Li 4+ is major defect. Its concentration decreases with increased VTE time. The Li 2 O content in the films is evaluated to be 48.8–49.7 mol% from the birefringence results. It increases with a rise in VTE time, matching with the characteristic of antisite defects. As defects concentration decreases, the non-radiative transition weakens , which is beneficial to improve fluorescence properties. • A Li-rich vapor transport equilibration technique was used to prepare near-stoichiometric Er3+ doped LiNbO 3 films. • Visible and near-infrared fluorescence intensity increases by ∼10 times as VTE time is increased to 5 h. • Li 2 O contents in the grown films were evaluated from measured birefringence to be from 48.8 to near-stoichimetric 49.7 mol%. • The enhancement can be explained from viewpoint of Nb4+ defects concentration measured by XPS. [ABSTRACT FROM AUTHOR]

Abstract: An X-cut and a Z-cut Li-deficient, non-congruent LiNbO3 crystal were prepared by carrying out Li-poor vapor transport equilibration (VTE) treatment on one mm thick congruent plates at 1100°C for 170h. From the measured optical absorption edges (OAE), the Li2O content is evaluated to be 47.74±0.13mol% in the X-cut plate and 47.61±0.13mol% in the Z-cut VTE crystal. The composition homogeneity is verified by measuring the OAE as a function of the crystal thickness. The visible optical absorption and powder X-ray diffraction studies allow to conclude that both VTE crystals retain still the LiNbO3 phase. OH− absorption characteristics of the two VTE crystals were investigated in comparison with those of the as-grown crystals. The results show that the Li-poor VTE treatment only slightly modifies the area, spectral width, spectral shape and background absorption of the OH− band. Comparison shows that the Li-poor VTE effects on the OH− absorption have large differences from the Li-rich VTE effects in bandwidth and areas of three component peaks at 3466, 3481 and 3488cm−1. These spectral differences are explained on the basis of the H+ site occupation and crystal defect models taking into account the VTE effect on crystal defect concentration. The Li-poor VTE effects are attributed to the VTE inducing the redistribution of OH− ions in different sites. [Copyright &y& Elsevier]

Abstract: Influence of thermal reduction on intrinsic (bipolarons), extrinsic (Er3+) defects and OH− groups in Er:LiNbO3 crystals, which were as-grown and VTE-treated (VTE: vapor transport equilibration) before being reduced, was studied by measuring the polarised or unpolarised optical absorption in visible and near infrared regions. A wide and strong band extending from the optical absorption edge up to the infrared region and peaking around 500nm (∼2.5eV), resulting from the absorption of reduction-induced bipolarons, is observed. Meanwhile, the thermal reduction also induces an additional, relatively much narrow absorption band around 370nm in a crystal whether it is Er-doped or undoped and whether it is congruent or originally VTE-treated. Both the 500nm and the 370nm bands show similar polarisation dependence. The thermal reduction treatment hardly influences Er3+ spectroscopic properties such as absorption amplitude, linewidth, peaking position and polarisation dependence. The original VTE effects on the spectroscopic properties of Er:LiNbO3 crystal are essentially retained still. The thermal reduction has a similar effect on the OH− absorption to a strong VTE treatment: the removal of the OH− groups contained in the crystal. [Copyright &y& Elsevier]

Visible and near infrared absorption and emission (488 nm excitation) characteristics of ErNbO4 powder, which were prepared by calcining the Er2O3 (50 mol%) and Nb2O5 (50 mol%) powder mixture at 1100 and 1600 °C for different durations, have been investigated at room temperature. The absorption and emission characteristics of these calcined ErNbO4 powder were summarized and discussed in comparison with those of Er2O3. Weak emission of Er2O3 relative to the calcined ErNbO4 is mainly conducted with absorption difference at the excitation wavelength 488 nm. The obvious spectral changes from Er2O3 to calcined ErNbO4 samples are related to an elevated-temperature-assisted phase transformation according to the solid-state chemical reaction equation: Er2O3 + Nb2O5 ⇌ 2ErNbO4, which results in the changes of the ion environment of Er3+ and hence changes of the Stark levels of Er3+. The further spectral change as the strengthened calcination results from the improvement of ErNbO4 purity in the calcined mixture. The borders between two green transitions and between two near infrared transitions in the emission spectra of both calcined samples and Er2O3 were tentatively identified by referencing earlier reported emission spectra of the precipitated Z-cut VTE Er(2.0 mol%):LiNbO3 crystal and the match relation between absorption and emission spectra of the ErNbO4 powder. A comparison was performed on the spectra of calcined ErNbO4 powder and those of VTE Er(2.0 mol%):LiNbO3 crystals. The results allow to preliminarily deem the contribution of ErNbO4 precipitates, generated inside these crystals by the VTE procedure, to the spectra of these crystals. [Copyright &y& Elsevier]

Generation of multiple photonic hooks (PHs) at THz region in reflection mode by a multi-dielectric structure that consists of periodical or aperiodical scatterer units on top surface of a reflection screen has been systematically studied. The study focuses on effects of geometric and optical parameters of the multi-dielectric structure on characteristics of the PHs generated, especially the bending angle. These geometric and optical parameters include spacing between adjacent scatterer units, length of reflection screen, as well as the size, refractive index and number of scatterer units. The results show that these geometric and optical parameters are key factors in tailoring the performance of the multiple PHs generated. The spacing determines bending direction of PHs and there exists a critical spacing value. Smaller than this critical value, the PHs generated bend outwards. Larger than this critical value, the PHs bend inwards. In addition, the spacing can be also used to effectively adjust the intensity and bending angle of the PHs. Screen length affects mainly PH's bending. The size and refractive index differences between scatterer units introduce additional optical inhomogeneity that leads to the generation of asymmetrical multiple PHs. Finally, the possibility of generation of an arrayed PHs in the reflection mode is demonstrated. • Generation of photonic hooks (PHs) array in reflection mode is studied. • PHs formation is ascribed to optical inhomogeneity of multi-dielectric structure. • PHs performance can be effectively tailored by geometric and optical parameters. • PHs generated at smaller or larger spacing bend in opposite directions. • Optical-axis asymmetric structure generates asymmetrical PHs array. [ABSTRACT FROM AUTHOR]

We measured fluorescence spectra of visible and infrared electronic transitions at 530 (2H 11/2 → 4I 15/2), 560 (4S 3/2 → 4I 15/2), 670 (4F 9/2 → 4I 15/2), 980 (4I 11/2 → 4I 15/2), 1530 (4I 13/2 → 4I 15/2) and 2700 nm (4I 11/2 → 4I 13/2) of Er3+ in Gd 3 (Al 0.4 Ga 0.6) 5 O 12 garnet single-crystal excited at the wavelength 520 nm (for the 980 nm emission) or 980 nm (for other emissions). From measured fluorescence spectrum, Er3+ emission cross-section spectrum was calibrated for each transition. Absorption cross-section spectra were further derived on the basis of McCumber equation, and compared with those spectra obtained from measured absorption spectrum on the basis of the Beer-Lambert law. The comparison showed that the results given by two methods can be regarded as same within the error. The transient features of visible and infrared emissions were also studied under pulse excitation. The results showed that, while the three IR emissions at 980, 1530 and 2700 nm all follow a mono-exponential function, both green and red 980-nm-upconverted emissions at 560 and 670 nm exhibit non-mono-exponentially decaying feature, and each decaying trace can be fitted by a sum of two exponential functions. The non-mono-exponential feature is related to cross relaxation caused by non-radiative short-distance energy transfer between clustered Er3+ sites. • Cross-section spectra of Er3+ ions in Gd 3 (Al,Ga) 5 O 12 garnet crystal were studied. • Emission cross-section spectra are computed from measured fluorescence spectra. • Absorption cross-section spectra are derived by two ways and results are same. • 980-nm-upconverted visible emissions show non-mono-exponential transient feature. • It is caused by non-radiative energy transfer between clustered Er3+ sites. [ABSTRACT FROM AUTHOR]

A previous paper [J. Lumin. 177 (2016) 219] has studied Judd-Ofelt spectroscopic properties of Er3+ (0.5 at%)-doped Gd 3 (Al, Ga) 5 O 12 garnet single-crystal, and reported the Judd-Ofelt intensity parameters Ω 2 = 1.47 × 10−20 cm2, Ω 4 = 1.38 × 10−20 cm2 and Ω 6 = 1.19 × 10−20 cm2. We have performed a re-study on the Judd-Ofelt spectroscopic properties of a similar garnet single-crystal Er3+(20 at%)-doped Gd 3 (Al 0.4 Ga 0.6) 5 O 12 , and obtained quite different values of Judd-Ofelt intensity parameters Ω 2 = (0.34 ± 0.11) × 10−20 cm2, Ω 4 = (0.48 ± 0.16) × 10−20 cm2 and Ω 6 = (1.05 ± 0.17) × 10−20 cm2. The quite different Ω t values given by two studies are mainly due to remarkably different absorption characteristics of Er3+ in the crystals employed in two studies, and factors that lead to quite different absorption characteristics are discussed. Another reason is that the hypersensitive transition at 380 nm was excluded from the Judd-Ofelt calculation in that paper, resulting in deviation of their result. Instead, all of the Er3+ absorption bands were included in our calculation, allowing to obtain more accurate Ω t values. Quite different Ω t values lead to considerable differences of relevant spectroscopic parameters, such as radiative rates and lifetimes of typical Er3+ emissions, for which the difference from each other is as much as three times. In addition, visible, near-infrared and mid-infrared fluorescence spectra and lifetimes of Er3+ in the studied crystal were measured, and the emission characteristics are discussed. • Er:Gd 3 (Al,Ga) 5 O 12 crystal was grown and its spectroscopic properties are studied. • Judd-Ofelt analysis was done and the result is compared with previously reported. • The Judd-Ofelt data obtained show large differences from the previously reported. • Main reason is large difference of Er3+ absorption features in two studies. • Sellmeier equation, lattice constant and segregation coefficient are presented. [ABSTRACT FROM AUTHOR]

Dependences of two factors on Judd-Ofelt (JO) intensity parameters Ω t (t = 2,4,6) have been studied theoretically by exemplifying the widely studied Er3+ ion, which is doped into Gd 3 Ga 5 O 12 , LiNbO 3 and SrGdGa 3 O 7 bulk single-crystals and β-NaYF 4 powder. The two factors include 1) selection of induced electric-dipole transitions for JO calculation and 2) integrated absorption coefficient (IAC) or integrated absorption cross-section (IACS) of individual electric-dipole transition of Er3+. Absolute and relative root-mean-square methods were considered in the JO fits and consistent results were obtained. Ω t depends mainly on strong and medium-intensity transitions. These transitions must be all included in the fit. The exclusion of an individual transition from the fit leads to deviation of Ω t. From the matrix elements [ U (t)]2 values, it is possible to rate how the inclusion or not of a transition affects the Ω t. The study also shows that the Ω t changes almost linearly with the IAC or IACS of an individual transition, and the relevant linear slope may have a positive or negative sign, depending on the [ U (t)]2 value. Moreover, the larger the [ U (t)]2 value is, the larger the corresponding Ω t value is. In addition, the Ω 2 , Ω 4 and Ω 6 are dependent of each other because it is the summation ∑ t Ω t [ U (t) ] 2 instead of its component term Ω t [ U (t) ] 2 that has a linear relationship to the IAC or IACS. [ABSTRACT FROM AUTHOR]

A simple photonic crystal is a periodic and regular nanostructure. Breaking the periodicity of a photonic crystal by a defect layer can be an interesting pattern. We study the spectral statistics of eigenfrequency spacings (the so-called level spacing statistics) for TM-polarized waves in the one-dimensional (1D) defective photonic crystal containing graphene-based hyperbolic metamaterial as a defect layer. The localization of light in the proposed structure has been investigated using the transfer matrix method. Moreover, the effects of different parameters, such as the number of graphene cells and the defect layer's optical axis orientation, are discussed. The performance of the ratios of nearest neighbor level spacings has become a Wigner distribution at the localized phases which are introduced by inserting a defect layer. The spectral statistics are found to approach the Wigner distribution for different repeat numbers of graphene layers. In contrast, in the absence of graphene layers, the level spacing statistics follow the Poisson distribution. - 1D photonic crystal with anisotropic defect layer containing graphene-based hyperbolic metamaterial is proposed. - The optical properties of the structure such as localization of light has been investigated using the well-known transfer matrix method. - Chaotic behavior has been found for different repeat numbers of graphene nanolayers. - In the absence of graphene layers, the level spacing statistics follow the Poisson distribution. [ABSTRACT FROM AUTHOR]

Dependences of Judd-Ofelt (JO) intensity parameters Ω t (t = 2,4,6) and their relative ratios (Ω 4 /Ω 2 and Ω 6 /Ω 2) on overall changing factor (q) of integrated absorption bands of Er3+ have been studied theoretically. These dependences can help us to use relevant method to correctly extract Ω t of a powder material. The study was done by exemplifying the widely studied Er3+ system, which is doped into Gd 3 Ga 5 O 12 , LiNbO 3 and SrGdGa 3 O 7 bulk single-crystals and β-NaYF 4 crystalline powder. The results show that the Ω t changes linearly with the overall changing factor q : Ω t = γ t q + ξ t. The linear feature changes from one material to another. The magnetic-dipole oscillator of the 1.5 μm transition results in non-zero nodal increment ξ t and hence dramatical changes of the relative ratios Ω 4 /Ω 2 and Ω 6 /Ω 2 (with q) in smaller q region. The dramatical feature limits application of the relevant method for Ω t calibration of a luminescent powder. To get the precise Ω t values using the relevant method, the q value must be large enough. This can be realized by adopting a thin powder plate with a low Er3+ concentration for spectral measurement. • Ω t is studied against overall varying factor q of all Er3+ absorption band areas. • Ω t has a linear relation to q and the linear feature relies on the host material. • Ω 4 /Ω 2 and Ω 6 /Ω 2 change dramatically with the q parameter in smaller q region. • The feature is caused by magnetic-dipole oscillator of the 1.5 μm transition. • The change concerns validity of related method for Ω t calibration of a phosphor. [ABSTRACT FROM AUTHOR]

Group velocity dispersion characteristics of rib-type photonic wire(PW) based on LiNbO 3 on insulator(LNOI) were studied theoretically. The study focuses on flexibility in dispersion tailoring, zero dispersion, dispersion flatness and zero modal birefringence(ZMB). The results show that the LNOI PW is more flexible in dispersion tailoring than the conventional Ti-diffused LiNbO 3 waveguide and Si-on-Insulator (SOI) PW. Zero dispersion at 1.55 μm wavelength can be realized for both TE and TM modes. Zero dispersion wavelength shows definite polarization dependence. It can be tailored by adjusting the rib height. For a TE mode guided in a appropriately designed LNOI PW, zero dispersion can be realized in a broad wavelength range of 1.3–1.7 μm. The TE mode is thus more suitable for dispersion flatness application. The dispersion flatness of the LNOI PW is one order of magnitude better than that of the SOI PW. In contrast, the TM mode is more suitable for dispersion engineering as it shows more significant dispersion feature. ZMB can be realized by adopting an X-cut/Y-propagation scheme. A PW geometry for simultaneous realization of single-mode operation and ZMB is presented. • Dispersion characteristics of rib-type LNOI photonic wire (PW) have been studied. • LNOI PW has a larger flexibility in dispersion tailoring than the SOI one. • Zero dispersion wavelength can be effectively tailored via rib height H rib. • Zero dispersion in the wavelength range of 1.3–1.7 μm can be realized for TE mode. • PW geometry for monomode operation and zero mode birefringence at 1.5 μm is given. [ABSTRACT FROM AUTHOR]

Electro-optic (EO) property of titanium-diffused LiNbO 3 optical waveguide was studied using prism coupling technique. The EO-effect-induced increment ΔN of effective refractive index (at the wavelength 1311 nm) was measured as a function of externally applied electric field E for TM 0 and TM 1 modes guided in the titanium-diffused LiNbO 3 planar waveguide. By using the linearity between ΔN and E, the EO coefficient r 33 is determined. For reference, the measurement was also performed on a bulk material of congruent LiNbO 3 , yielding r 33 = 24.9 ± 0.8 p.m./V, which is in qualitative agreement with that obtained by extrapolating previously reported result to the 1311 nm wavelength, ~24 p.m./V, confirming the feasibility of the measurement method. Measurements on the titanium-diffused planar waveguide give r 33 = 25.7 ± 0.6 p.m./V for the TM 0 mode and 25.7 ± 0.4 p.m./V for the TM 1 mode. Both values can be considered as identical within the error, and are the same as that of the bulk material, indicating that the EO property is well preserved in the titanium-diffused LiNbO 3 waveguide. The main reason is that the LiNbO 3 crystalline phase is well preserved in the titanium-diffused waveguide layer. • Electro-optic (EO) property of Ti:LiNbO 3 planar waveguide is studied. • The study uses the linearity between refractive index change and electric field. • The study shows that EO property is well preserved in Ti:LiNbO 3 waveguide. • Little effect of Ti-diffusion doping on EO property is qualitatively explained. [ABSTRACT FROM AUTHOR]

An application manner of Judd-Ofelt (J-O) theory to a luminescent powder is studied by exemplifying an Er3+-doped Li 5 La 3 Nb 2 O 12 nanophosphor, on which J-O analysis has been performed previously using the approach based on measured relative absorption (optical density) spectrum and Er3+ 1.5 μm fluorescence lifetime. Present study focuses on two key issues that are closely related to the reliability and application manner of the approach but not clarified in the previous study. First, whether or not a strong or medium-intensity oscillator, such as the 4I 15/2 ↔4I 13/2 (1.5 μm) one, can be excluded from the J-O least-square fit. Second, whether the relative ratio between the J-O intensity parameters Ω i (i = 2, 4, 6) remains unchanged as the integrated optical densities (IODs) of all transitions change all together by a factor x. Our studies show that the strong or medium-intensity transition oscillator, such as the 4I 15/2 ↔4I 13/2 one, must be included in the J-O fit. The relative ratios between Ω i change remarkably with the collectively varying factor x of the IODs for smaller x values because of magnetic dipole contribution of the 1.5 μm oscillator. The change is undesired as it may result in the deviation of Ω i values, and hence restricts the application manner of the related approach. A more reliable application manner of the approach is proposed. • Ω i are determined mainly by strong and medium-intensity transitions. • 1.5 μm transition affects Ω 4 and Ω 6 strongly, it must be included in J-O fit. • Ω i shows a linear relationship to integrated absorption cross-section. • Ω 4 /Ω 2 and Ω 6 /Ω 2 change remarkably with the relative absorption cross-section. • A more convincing and reliable method for calibration of Ω i has been suggested. [ABSTRACT FROM AUTHOR]