The group of zincblende III-V compound semiconductors, especially (100)-grown AlGaAs and GaAs, have recently been presented as promising materials for second harmonic generation (SHG) at the nanoscale. However, major obstacles to push the technology towards practical applications are the limited control over directionality of the SH emission and especially zero forward/backward radiation. In this work we provide both theoretically and experimentally a solution to these problems by presenting the first SHG nanoantennas made from (111)-GaAs embedded in a low index material. These nanoantennas show superior forward directionality compared to their (100)-counterparts. Most importantly, it is possible to manipulate the SHG radiation pattern of the nanoantennas by changing the pump polarization without affecting the linear properties and the total nonlinear conversion efficiency.
Optical nanoantennas possess great potential for controlling the spatial distribution of light in the linear regime as well as for frequency conversion of the incoming light in the nonlinear regime. However, the usually used plasmonic nanostructures are highly restricted by Ohmic losses and heat resistance. Dielectric nanoparticles like silicon and germanium can overcome these constrains [1,2], however second harmonic signal cannot be generated in these materials due to their centrosymmetric nature. GaAs-based III-V semiconductors, with non-centrosymmetric crystallinity, can produce second harmonic generation (SHG) [3]. Unfortunately, generating and studying SHG by AlGaAs nanocrystals in both backward and forward directions is very challenging due to difficulties to fabricate III-V semiconductors on low-refractive index substrate, like glass. Here, for the first time to our knowledge, we designed and fabricated AlGaAs nanoantennas on a glass substrate. This novel design allows the excitation, control and detection of backwards and forwards SHG nonlinear signals. Different complex spatial distribution in the SHG signal, including radial and azimuthal polarization originated from the excitation of electric and magnetic multipoles were observed. We have demonstrated an unprecedented SHG conversion efficiency of 10-4; a breakthrough that can open new opportunities for enhancing the performance of light emission and sensing [4].
References
[1] A. S. Shorokhov et al. Nano Letters 16, 4857 (2016).
[2] G. Grinblat et al. Nano Letters 16, 4635 (2016).
[3] S. Liu et al. Nano Letters 16, 7191 (2016).
[4] R. Camacho et al. Nano Lett. 16, 7191 (2016).
Metallic nanoantenna possess versatile scattering properties enabling to engineer the emission directionality at the nanoscale. However, due to their Ohmic losses and low heat resistance they cannot be practically applied in nonlinear optical processes for optical frequency conversion. Dielectric nanoparticles, e.g. silicon and germanium, are good candidates to overcome these limitations [1, 2]. Nevertheless, the centrosymmetric nature of these materials have voided the second-harmonic generation (SHG). Alternatively, the use of GaAs-based III-V semiconductors, with non-centrosymmetric structures, can overcome this difficulty [3,4]. However, fabrication of III-V semiconductor nanoantennas on low refractive index substrates remains very challenging, blocking the possibility to explore the SHG directionality in both forward and backward direction. Here, for the first time to our knowledge, we design and fabricate high-quality AlGaAs nanostructures on a glass substrate. Through this novel platform, we manage to excite, control and detect backward and forward nonlinear signals by SHG in AlGaAs nanodisks [5,6]. In particular, we observe that for certain size of nanoantenna, the SHG emission has a complex spatial distribution polarization state corresponding to radial polarization in the forward direction and a polarization state of a more general nature in the backward direction. Furthermore, we demonstrate an unprecedented SHG conversion efficiency of 10-4. Our breakthrough can open new avenues for enhancing the performance of photodetection, light emission and sensing.v
Nanostencil Lithography (NStL), while comparatively still in infant stages, is proving to be a viable option for low-cost and high resolution fabrication. An ideal stencil for NStL consists of a low-stressed silicon nitride membrane supported on a silicon chip with required patterned features in nanometer range that become apertures. The stencil is used as a shadow mask and placed in close contact on top of a substrate/wafer. This pair is then ready for either depositing metal through the apertures in the stencil using variety of deposition techniques or etching the substrate using dry etching techniques with stencil acting as a mask. The nanostencils were fabricated using focused ion beam writing on a silicon nitride window/membrane. We made well-ordered array of 700 nm diameter and 15 nm thick gold and chromium nanodots on III-V substrate. Metal layers were deposited using e-beam evaporator. The formed gold nanodots can be used for vapor-liquid-solid nanowire growth (bottom-up), while the chromium nanodots were used as a mask for reactive ion etching of GaAs structures, for instance, fabricating nanowires (top-down approach). We used the nanostencil directly as a mask for dry etching of InP substrate for making nanoholes array. Making these types of nanoholes in silicon oxide layer deposited on the top of III-V substrate opens the possibility to use in selective area growth of nanowires. Additionally, we fabricated optical nanoantenna structures to demonstrate other possible usage of NStL.
X-ray lithographic conditions for high-aspect-ratio SU-8 resist structures are characterized for potential application in x-ray optics and bioMEMS. The effects of the main process parameters, such as exposure dose, postexposure bake, development time, and the packing density of the microfabricated features, on the increase in feature size at the top portion of the resist (as compared to that in masks) are investigated. We find that lower postexposure bake and exposure dose leads toward minimizing dimensional errors. Further improvements in reducing the dimensional errors can be achieved by overdeveloping the structures. Using overdevelopment, we demonstrate an improvement in dimensional error for a given structure from 5 to 3.3%.
We investigate the development time for densely packed high-aspect-ratio SU-8 structures. We find that a simple notch model as used by other workers to predict development time for isolated structures is inadequate for our case. We develop a theoretical model that allows the effects of mass transport of the resist in the developer to be included. By calibrating the process, we find that we are able to predict development time for our structures.
X-ray lithographic conditions for high aspect ratio SU-8 resist structures are characterized for potential application in x-ray optics and bioMEMS. The effects of the main process parameters such as exposure dose, post exposure bake, development time and the packing density of the microfabricated features on the development depth and increase in feature size at the top portion of the resist (as compared to that in mask) were investigated. As test samples, we fabricated 1mm high, densely-packed SU-8 structures comprising of 30μm square pillars with spacings of 36μm, 12μm and 6μm. Dissolution rates are found to be longer for densely packed structures than predicted by simple physical models based on isolated structures. We examine the effect on dissolution rates of the density of features in our structures. We also optimised our process with respect to the parameters described above using the Taguchi method. We find that optimisation of the development time with exposure dose and post-exposure bake time can reduce the dimensional error to ~ 3% for certain densely-packed structures.
Carbon Tetraflouride (CF4) plasma etching condition for SU-8 negative photoresist is characterized for its potential applications in photonics and bioMEMS. The effects of main plasma etching parameters such as rf power, gas flow rate, chamber pressure and time were systematically studied and the parameters were optimized by a three-level, L9 orthogonal array of the Taguchi method. By optimization, the optimal parameter range and the weighted percent of each parameter on the final results i.e. depth, surface roughness and wall angle were determined. Photoresist & metal were used and compared as masks for plasma etching. The minimum feature size was 1µm in both cases. Results indicated that with the increase of rf power, etch rate and roughness increases almost linearly. With increase in gas flow rate, etch rate increases while roughness decreases non-linearly. Etch rate is linear with time but roughness is significantly dependent on time initially. The side-wall angle of the samples with metal mask was found to be nearly 90° whereas samples with photoresist as the mask showed poor side-wall angle and surface roughness mainly due to poor mask-resist selectivity. Optimized values of rf power, gas flow rate, time and pressure were found to be 200W, 240sccm, 20minutes and 1Torr respectively, which yielded high etch rate (80nm/min), low surface roughness (5nm) and nearly vertical side-walls (89°).
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.