SPIE Membership Get updates from SPIE Newsroom
  • Newsroom Home
  • Astronomy
  • Biomedical Optics & Medical Imaging
  • Defense & Security
  • Electronic Imaging & Signal Processing
  • Illumination & Displays
  • Lasers & Sources
  • Micro/Nano Lithography
  • Nanotechnology
  • Optical Design & Engineering
  • Optoelectronics & Communications
  • Remote Sensing
  • Sensing & Measurement
  • Solar & Alternative Energy
  • Sign up for Newsroom E-Alerts
  • Information for:
SPIE Photonics West 2018 | Call for Papers




Print PageEmail PageView PDF


Atomic-layer-deposited thin films for silicon nanophotonics

Improved performance for silicon slot waveguides and photonic crystal cavities is achieved with atomic-layer-deposited thin films.
8 May 2012, SPIE Newsroom. DOI: 10.1117/2.1201204.004218

During the last decade, silicon photonics has gained tremendous interest for two main reasons: the large-scale manufacturing capacity of the vast silicon microelectronics industry, and the special optical properties of silicon. Its high refractive index enables very efficient light confinement and dispersion engineering of waveguides, and the optical nonlinearity—X(3)—in silicon is very high, about 200× higher than in silica. These properties have made it possible to demonstrate silicon-based all-optical devices with unprecedented performance.1 However, silicon is not the best material for everything required by optical telecommunications systems. Therefore, it is necessary to integrate other materials onto silicon photonic chips. In this article, we briefly review our recent advances in the integration of atomic-layer-deposited thin films with silicon nanophotonic devices.

We have improved silicon photonic devices with thin films grown by atomic layer deposition (ALD).2–4 Figure 1 shows the principle of ALD. Gas phase precursors are introduced into the growth chamber one at a time, and these growth cycles are separated by purge phases. During each growth pulse, precursor molecules react with the molecules on the substrate. Typically, a fraction of an atomic layer is deposited in each growth cycle. ALD films are highly conformal—they are even and pinhole-free—and have excellent thickness accuracy, which makes them attractive for coating complex surfaces.

Figure 1. Principle of atomic layer deposition (ALD) growth. Al(CH3)3: Trimethyl aluminum.

The slot waveguide makes possible efficient confinement of light in a low-index material on a silicon photonic circuit. The optimal slot width is below 100nm, making fabrication a challenge even with deep-UV (DUV) lithography, the state of the art in large-scale silicon chip processing. A conformal high-index film can be used to produce narrow nano-slots, even when a coarser linewidth is used for lithography, making slot waveguides more practical for large-scale fabrication.5 Conformal thin films also reduce nanoscale roughness due to lithography. We demonstrated significant propagation loss reduction of DUV-patterned slot waveguides using ALD-grown thin films.2 The propagation loss was reduced from 70dB/cm for an uncoated slot waveguide to 7dB/cm by deposition of 50nm of titanium dioxide, significantly increasing the usefulness of these waveguides for photonic devices. The slot in the resulting waveguide is still open from above after deposition (shown in Figure 2), making it feasible for use as a sensor, or for filling with another material for added functionality or passivation. We have also achieved propagation loss as low as 5dB/cm for a slot waveguide patterned with 248nm DUV and completely filled with ALD-grown aluminum oxide.3

Silicon-based photonic crystal nanocavities present a world record in terms of Q/V, i.e., quality factor divided by mode volume.6 The quality factor describes the capacity of the resonator to store light, and mode volume gives the volume of the cavity in which the light is confined. This superlative Q/V means that the cavity provides the highest concentration of optical energy, which makes these devices the ultimate tools for experimenting and using nonlinear optical phenomena. We demonstrated Q value improvement for these cavities with ALD films (see Figure 3).4 This is further proof of the high optical quality of ALD-grown aluminum oxide. The ability to grow material on the high-Q cavity without compromising its performance is also an excellent starting point for future research. With ALD, it is very easy to introduce dopants into the films and to control their doping profile. By incorporating luminescent material into the silicon nanobeam cavities, we can study light-matter interaction in higher Q/V cavities than before, and even search for the ‘holy grail’ of silicon photonics: an integrated emitter.7

Figure 2. (a) Scanning electron microscope (SEM) image of a silicon slot waveguide clad with a titanium dioxide thin film of thickness 30nm. (b) Loss in slot waveguides with different dimensions and film thicknesses.2

Figure 3. Quality (Q) factors for silicon nanobeam cavities before and after deposition of 20nm of aluminum oxide (Al2O3) by ALD.4 The Q factor—the capacity of the resonator to store light—was improved in every cavity in the experiment.

Our work has shown that the ALD films are well applicable to silicon nanophotonic devices. In other studies, ALD growth of dozens of different materials have been demonstrated, with their refractive indices ranging from 1.3 to 2.8.7 The layer-by-layer growth of ALD makes it possible to engineer the layer structure at the nanoscale (shown in Figure 4) and to control the crystallinity of the film as it grows to optimize its optical properties, such as nonlinearity.

Figure 4. SEM image of an ALD-grown nanolaminate.8

ALD is already being used in the microelectronics industry and thus is particularly attractive for silicon photonics. The growth temperatures are typically low, which makes the technique applicable to processing silicon wafers that already have electronic components attached. Therefore, ALD offers great potential for nanophotonics, as well as unexplored opportunities for future research. The wide range of materials, the ability to engineer films with such precision, and the possibility of doping the films in a controlled way all open up numerous opportunities for optimizing the optical properties of the material. Our next steps will be to study films with engineered nanocrystal structures, as well as doped thin films, with the goal of engineering thin films for optimal nonlinear properties or added functionality, e.g., light emission. Silicon nanophotonics gives us a unique toolbox for designing the ideal optical mode properties for many functions. A range of devices, such as strip- and slot-based photonic crystal waveguides, and microring or photonic crystal-based resonators are possible with this platform.

Antti Säynätjoki
Aalto University
Espoo, Finland

Antti Säynäjoki is a post-doctoral researcher at Aalto University in Finland. He is currently a visiting researcher at the College of Optical Sciences, University of Arizona. His research interests include hybrid integration for silicon nanophotonics, and optical nonlinearities in silicon and metal nanostructures.