GIPO newsletter

A 3D electrode design for fast-response VA-FFS liquid crystal mode

Professor Wing-Kit Choi

Graduate Institute of Photonics and Optoelectronics, National Taiwan University

In this work, a 3D electrode design is proposed [1] for a Vertically-Aligned Fringe-Field Switching (VA-FFS or Dual-FFS) liquid crystal mode which is known for its intrinsic sub-millisecond fast response time without using thin cell gap or other liquid crystal phases. Compared to the previously proposed 2D electrode designs [2-3], we found that, by using this new 3D electrode design, it is possible to improve the potential maximum transmission due to reduced disclination lines or deadzones (without using a double-sided electrode structure) and also to improve the potential response speed due to smaller effective domain size of liquid crystals. A major mechanism for causing such a fast response time is due to the formation of self-imposed boundaries [3] (or so-called virtual walls) which can be viewed as liquid crystals having very small “effective” cell gap. This proposed approach of achieving fast response time in liquid crystals can be very attractive since it doesn’t require more complicated liquid crystal phases (or materials) such as Blue Phase or SmC* Ferroelectrics. This approach may simply use standard nematic liquid crystal phase or materials such as E7. Moreover, this approach doesn’t require the use of very thin cell gap (e.g. 2μm or below) which is less practical for large area fabrication. This proposed technique of achieving fast response time in liquid crystals can therefore be attractive for the future development of fast response liquid crystal displays, intensity or phase modulators.

[1] Wing-Kit Choi, Chia-Hsiang Tung and Bo-Kai Tseng, “Fast Response VA-FFS Liquid Crystal Mode using 3D electrode design”, SID 2017 Digest, pp.1838-1840, May 2017

[2] M. Jiao, Z. Ge, S. T. Wu, and W. K. Choi, “Sub-millisecond response liquid crystal modulators using dual field switching in a vertically aligned cell,” Appl. Phys. Lett., 92, 111101, Mar. 2008

[3] W.K. Choi & S.T. Wu, “Fast response liquid crystal mode,” US Patent 7298445 B1, Nov. 2007

Fig. 1 shows electrode design and molecular orientation of 2D VA-FFS (left) and the new 3D VA-FFS (right) at voltage-on state.

Fig. 2 shows the top-view transmission of 2D VA-FFS (left) and 3D VA-FFS (right). The figure on the right shows that disclination lines (or deadzones) are indeed reduced along the transverse (or y) direction in the 3D VA-FFS.

Fig. 3 shows the improvement of the transmission (figure on left, with circular polarizer) and also improvement of response speed (Table on right) of 3D VA-FFS compared to 2D VA-FFS liquid crystal mode.

Threshold Voltage Improvement Through Sidewall Control of InGaAs Fin-structured High Electron Mobility Transistors (Fin-HEMTs)

Professor Chao-Hsin Wu’s laboratory

Graduate Institute of Photonics and Optoelectronics, National Taiwan University

The fabrication process was started with fin formation by inductively coupled plasma dry etch. The source and drain metal were deposited by e-beam evaporator and followed by rapid thermal annealing to form Ohmic contacts. Then, the recess region was etched by diluted sulfuric acid followed by self-aligned gate metal of Ni/Au. Fin widths were finally defined by recess etch and varied from 396 to 54 nm. The schematic cross section and structure are shown in Fig. 1.

Figure 2 shows I_D-V_G characteristics of FinHEMTs with different fin widths (W_fin). Threshold voltage (V_T) shifts toward positive direction from -1.41 V to 0.56 V as W_fin decreases. It can be observed that compared to planar devices, V_T of Fin-HEMT moves to +V_G as W_fin decreases. As W_fin is reduced down to about 90 nm, the device starts to work at enhancement mode operation. The mechanism is proposed in Fig.3. When V_G decreases from V_G1 (large) to V_G2 (smaller), top gate still cannot turn off the channel. So the planar devices are still at on-state operation. On the other hand, for fin-shaped devices, though top gate is not capable of turning off the channel, the depletion region from surface of sidewalls is large enough to fully deplete the channel.

Fig. 1. Process flow, schematic cross section and device structure of InGaAs Fin-HEMT

Fig. 2. Transfer characteristics of InGaAs FinHEMTs with different fin width at V_D = 1.9 V

Fig. 3. Schematic mechanism of top gate and sidewall gates proposed in this work

Multi-Band Infrared Thermal Emitter with Selectable Polarization

Professor Si-Chen Lee’s laboratory

Graduate Institute of Photonics and Optoelectronics, National Taiwan University

This study demonstrated a multi-band infrared thermal emitter featuring a narrow bandwidth emission and polarization characteristics, which is quite suitable to be applied to the non-dispersive infrared (NDIR) detection system. NDIR system has been widely used in monitoring the hazardous and harmful substances in the human living environment. The target molecules can be detected by measuring the transmission optical energy at specific wavelengths due to their particular molecular bond vibration, which is inversely proportional to the molecular concentration in the NDIR system. Because the biomolecules usually exhibit multiple absorption peaks in the infrared regions, the multi-wavelength infrared light source plays an indispensable role in the NDIR system. The device is constructed by embedding the metallic grating strips within the resonant cavity of a metal/dielectric/metal (MDM) structure, as shown in Fig. 1(a) and 1(b). This arrangement makes it possible to generate waveguide resonances with mutually orthogonal polarization, thereby providing an additional degree of freedom to vary the resonant wavelengths and polarizations in the medium infrared region. The measured reflection spectra and the finite-difference time-domain simulation indicated that the electric fields of the waveguide modes with two orthogonal polarizations are distributed in different regions of the cavity, as shown in Fig. 1(c) and (d). Resonant wavelengths in different polarizations can be adjusted by altering the period, the metallic line width, or the position of the embedded gold strips. The ratio of the full width at half maximum (FWHM) to the peak wavelength was achieved to be smaller than 0.035. This work has been published in AIP Advances 7, 085122 (2017).

FIG. 1. (a) A photograph of the multi-band infrared thermal emitter with polarized waveguide modes. (b) Schematic diagrams of polarized waveguide thermal emitter. (c) The experimental and FDTD simulated reflection spectra in x- and y-polarizations. (d) The experimental thermal radiation spectra of the device in x- and y-polarizations.

High-resolution Analysis of Leaky Modes in Surface Plasmon Stripe Waveguides

Professor Hung-chun Chang

Graduate Institute of Photonics and Optoelectronics, National Taiwan University

The surface plasmon polariton (SPP) existing at an interface between metal and dielectric material is a basic phenomenon in the area of “plasmonics” and has been a well-known wave mode particularly in the visible and infrared wavelength ranges. With the rapid progress of the plasmonics research in recent years, various SPP waveguides have also been proposed and investigated. Here, we report our recent studies of the stripe plasmonic waveguides with a gold stripe having a silica substrate. Such simple structure was one of the earlier studied SPP waveguides. Weeber et al. reported their experimental measurement results of waveguide mode characteristics of such structure in 2003 [Phys. Rev. B 68, 115401 (2003)]. Zia et al. in 2005 [Phys. Rev. B 71, 165431 (2005)] presented finite-difference numerical analysis of the same waveguides, obtained their leaky and bound modes, and concluded that calculated leaky mode characteristics could be consistent with the observed SPP-propagation ones by Weeber et al. Zia et al. calculated the real and imaginary parts of the complex effective refractive indices, n_eff’s, of the leaky, quasi-transverse-magnetic (quasi-TM) SPP modes for different stripe widths, where n_eff is defined as the modal propagation constant divided by the free-space wavenumber. We have recently conducted high-resolution finite-element numerical analysis of the same stripe waveguide using an in-house developed full-vector finite-element imaginary-distance beam propagation method (FV-FE-ID-BPM) and discovered some subtle mode-field characteristics in the leaky modes so that the n_eff versus the stripe width curves appear to be of non-monotonic variation [H. H. Liu and H. C. Chang, J. Lightwave Technol. 34, pp. 2752–2757 (2016)]

Figure 1 Mode-field profiles of the TM1 mode of the stripe waveguide when the stripe width is 3.7 μm: (a) Re[E_x], (b) Re[H_x], (c) Re[E_y], (d) Re[H_y], (e) Re[E_z], and (f) Re[H_z].

Fig. 1 shows the magnitude profiles of the real parts of the six field-component phasors of the lowest TM1 mode for the stripe width of 3.7 μm (the gold stripe thickness is 55 nm). In each panel, the horizontal thin black line shows the boundary of the substrate, and the outline of the boundary of the stripe cross-section appears to be the thicker portion of the thin black line. In the analysis we used 638,347 unknowns for the computational domain of 8 μm x 11.5 μm size, which covers the right or left half of the waveguide cross-section making use of the structure symmetry. Such high spatial resolution and large coverage of the substrate region reveals the details of the leaky fields in the substrate including the interference feature. It is interesting to notice that the two stripe edges provide two sources for the propagating leaky fields. Note that H_x (Fig. 1(b)) and E_y (Fig. 1(c)) are the major field components. For further discussions, please refer to [J. Lightwave Technol. 34, pp. 2752–2757 (2016)].

A method for enhancing the favored transverse-electric-polarized emission of an AlGaN deep-ultraviolet quantum well

Professor C. C. Yang’s Lab.

Graduate Institute of Photonics and Optoelectronics, National Taiwan University

An AlGaN quantum well (QW) structure of a deep-ultraviolet (UV) light-emitting diode (LED) needs to be well designed for controlling its band structure such that the heavy-hole (HH) band edge becomes lower than the split-off (SO) band edge and hence the transverse-electric (TE) polarization dominates the emission for achieving a higher light extraction efficiency. Here, we report the discovery of un-intentionally formed high-Al AlGaN nano-layers right above and below such a QW and their effects on the QW for changing the relative energy levels of the HH and SO bands. The comparison between the results of simulation study and polarization-resolved photoluminescence measurement confirms that the high-Al layers (HALs) represent the key to the observation of the dominating TE-polarized emission. By applying a stress onto a sample along its c-axis to produce a tensile strain in the c-plane for counteracting the HAL effects in changing the band structure, we can further understand the effectiveness of the HALs. The formation of the HALs is attributed to the hydrogen back-etching of Ga atoms during the temperature transition from quantum barrier growth into QW growth and vice versa. The Al filling in the etched vacancies results in the formation of an HAL. This discovery brings us with a simple method for enhancing the favored TE-polarized emission in an AlGaN deep-UV QW LED.

Fig. 1 (upper) TEM image to show the HALs (dark stripes). (lower) Geometric phase analysis image.

Fig. 2 Schematic drawings to show the formation mechanism of the HALs.

Surface plasmon coupling for suppressing p-GaN absorption and TM-polarized emission in a deep-UV light-emitting diode

Professor Y. W. Kiang’s laboratory

Graduate Institute of Photonics and Optoelectronics, National Taiwan University

The radiated power enhancement (suppression) of an in- (out-of-) plane-oriented radiating dipole at a desired emission wavelength in the deep-ultraviolet (UV) range when it is coupled with a surface plasmon (SP) resonance mode induced on a nearby Al nanoparticle (NP) is demonstrated. Also, it is found that the enhanced radiated power propagates mainly in the direction from the Al NP toward the dipole. Such SP coupling behaviors can be used for suppressing the transverse-magnetic (TM)-polarized emission, enhancing the transverse-electric (TE)-polarized emission, and reducing the UV absorption of the p-GaN layer in an AlGaN-based deep-UV light-emitting diode by embedding a sphere-like Al NP in its p-AlGaN layer.

Fig. 1. (a)-(d) Schematic demonstrations of simulation structures A-D.

Fig. 2. Spectra of the normalized radiated power in structures B-D when an x-dipole is used (d = 50 nm). The vertical arrow marks 270 nm.

Fig. 3. Similar to Fig. 2 except when a z-dipole is used (d = 50 nm).