Silicon nitride on-chip C-band spontaneous emission generation based on lanthanide doped microparticles

This paper presents a hybrid fabrication method that integrates monodisperse lanthanide-doped microparticles into silicon nitride waveguide wells to generate broadband C-band spontaneous emission, achieving a 0.25% coupling efficiency and offering a scalable route for on-chip active light sources.

The Gist

Imagine you are trying to build a super-fast, microscopic internet highway made of glass (specifically, silicon nitride). This highway is incredibly smooth and efficient at carrying light signals, which is perfect for sending data across the globe. However, there's a major problem: this highway is "passive." It's like a perfectly paved road with no cars, no traffic lights, and no gas stations. It can't create the light signals itself; it can only guide them if someone else puts them there.

Usually, to get light onto this road, engineers have to glue in a separate, complex engine (like a laser made of different materials). But these engines are picky—they hate the high heat needed to build the glass road, and they don't fit well together. It's like trying to glue a heavy, hot brick onto a delicate ice sculpture.

The Big Idea: The "Magic Dust" Solution
Instead of trying to build a complex engine directly onto the road, the researchers in this paper came up with a clever workaround. They decided to use "magic dust"—tiny, glowing particles made of special crystals (doped with elements like Erbium and Ytterbium).

Think of these particles like tiny, self-contained solar-powered lanterns.

1. The Fuel: When you shine a specific type of light (a 950 nm laser, which is invisible to the human eye) onto these particles, they absorb the energy.
2. The Transformation: The particles act like a translator. They take that invisible energy and spit it out as a different color of light: the "C-band." This is the golden frequency used for almost all modern internet and phone communications (around 1530 nm).
3. The Placement: The tricky part is getting these lanterns to shine into the glass highway. If you just sprinkle them on top, the light goes everywhere, like a lightbulb in a dark room.

The Innovation: The "Funnel" and the "Pocket"
The team created a smart way to catch this light and force it onto the highway.

• The Pocket: They etched tiny, shallow wells (pockets) into the glass road. They then carefully "printed" the glowing particles into these pockets, ensuring they sit exactly where they need to be.
• The Funnel: The glass road isn't just a straight line; it has a special section shaped like a wide, circular fan (a taper). As the light travels from the wide part of the fan toward the narrow road, it gets squeezed and guided.

Imagine the glowing particles sitting in a wide, shallow bowl. The bowl is shaped so that any light the particles emit naturally rolls down the sides and funnels directly into the narrow pipe (the waveguide) at the bottom.

The Results: A Working Prototype
They built this device and tested it:

• The Test: They shined their "fuel" laser on the particles.
• The Output: The particles glowed, and the light successfully traveled down the glass highway.
• The Quality: The light they produced was exactly the right color for telecommunications (covering the 1500–1600 nm range). It was a broad spectrum, meaning it could carry a lot of different types of data at once.
• The Efficiency: Currently, about 0.25% of the light generated makes it into the highway. While that sounds small, think of it like catching rain in a bucket during a light drizzle. It proves the concept works! The researchers believe that by tweaking the shape of the "funnel" and the "pocket," they can catch much more of the light in the future.

Why Does This Matter?
This is a game-changer for the future of the internet.

• Simplicity: It avoids the messy, expensive process of gluing different materials together.
• Scalability: You can print these "lanterns" onto chips just like you print ink on paper, making it easy to mass-produce.
• Versatility: These glowing particles could act as the light source for internet routers, signal boosters, or even tiny lasers for future quantum computers.

In a Nutshell
The researchers figured out how to turn a passive glass road into an active light highway by embedding tiny, glowing "lanterns" into custom-made pockets and using a funnel shape to guide their light directly into the system. It's a simpler, cleaner, and more scalable way to build the next generation of optical internet.

Technical Summary

Here is a detailed technical summary of the paper "Silicon nitride on-chip C-band spontaneous emission generation based on lanthanide doped microparticles."

1. Problem Statement

The integration of active light-emitting elements into planar photonic circuits based on silicon nitride (SiN) is a significant technological challenge. While SiN offers low intrinsic optical losses and is compatible with CMOS processes, it faces two major hurdles when integrating active components for the telecommunications C-band (1530–1565 nm):

• Material Incompatibility: SiN is incompatible with standard active materials (like InP or GaAs) used in the near-infrared band.
• Thermal Constraints: The fabrication of high-quality SiN waveguides requires high-temperature annealing (approx. 1300°C), which destroys most active semiconductor materials.
• Current Limitations: Existing solutions typically rely on complex hybrid assembly methods (bonding separate chips), which are difficult to scale and integrate monolithically.

2. Methodology

The authors propose a hybrid integration method that embeds monodisperse luminescent microparticles into lithographically defined wells on a SiN chip, avoiding high-temperature processing of the active material itself.

• Active Material: The device utilizes NaYF$_4$ doped with Ytterbium (Yb) and Erbium (Er) microparticles. These particles act as upconversion phosphors, absorbing pump light in the 900–1000 nm range and emitting in the C-band (1530–1565 nm).
• Device Architecture:
  • A circular-sector-shaped taper coupler (radius 48 µm, opening angle 40°) transitions into a 1.4 µm-wide SiN waveguide.
  • The taper is designed to collect emission from particles deposited above it and couple it into the fundamental waveguide mode.
  • A 500 µm-long output taper narrows to 0.5 µm at the chip facet to enhance collection efficiency.
• Fabrication Process:
  1. Substrate Prep: Thermal oxidation of a silicon wafer to create a 4 µm SiO$_2$ buffer layer.
  2. Waveguide Formation: Deposition of a 200 nm SiN film (PECVD) followed by high-temperature annealing (1150°C).
  3. Patterning: Electron-beam lithography (EBL) and reactive ion etching to define the waveguide and taper.
  4. Well Creation: A top SiO$_2$ layer is deposited and planarized. Photoresist is used to define "wells" (1 µm deep) where particles will sit.
  5. Particle Deposition: Monodisperse NaYF$_4$:Yb,Er particles (1.5–2 µm lateral size) are deposited into the wells using a "printing" method based on hexagonal close-packing.
  6. Passivation: The particles are covered with a SiO$_2$ layer for protection and planarization.
  7. Facet Formation: Deep etching using a nickel hardmask to create the chip facets for optical access.

3. Key Contributions

• Novel Integration Strategy: Demonstrated a scalable method to integrate lanthanide-doped microparticles directly onto SiN photonic circuits without requiring hybrid bonding of separate active chips.
• Process Development: Developed a specific fabrication workflow involving well etching, particle deposition, and planarization that maintains waveguide integrity while ensuring precise particle placement.
• C-Band Source: Successfully realized an on-chip spontaneous emission source covering the optical telecommunications C-band (1500–1600 nm) using a 950 nm pump source.

4. Experimental Results

• Luminescence Characteristics:
  • Under 975 nm laser excitation, the particles exhibit strong emission in the near-infrared.
  • Peak Wavelength: 1532 nm.
  • Spectral Width: 60 nm (Full Width at Half Maximum, FWHM), covering the C-band.
• Optical Performance:
  • Pump Source: 950 nm diode laser.
  • Output Power: The total integrated power at the chip output was measured at 48 ± 2 pW.
  • Peak Spectral Power: 0.8 pW/nm at ~1530 nm.
  • Saturation Behavior: The luminescence power follows a saturation model ($P_{lum} = P_{max} P_{pump} / (P_{pump} + P_{sat})$), with a saturation pump power ($P_{sat}$) of 51 ± 5 mW.
• Coupling Efficiency:
  • Numerical simulations using the Finite-Difference Time-Domain (FDTD) method modeled the particles as incoherent dipole sources.
  • The average coupling efficiency into the fundamental TE waveguide mode was calculated to be 0.25% (-26 dB).
  • Simulations indicated that coupling efficiency decays exponentially with the distance between the dipole source and the waveguide surface, highlighting the importance of particle proximity to the taper.

5. Significance and Future Outlook

• Scalability: This approach offers a route to mass-produce active elements on SiN platforms, which are otherwise passive. It avoids the complexity and cost of hybrid assembly.
• Applications: The technology enables the creation of:
  • Broadband emission sources for telecommunications.
  • Planar optical amplifiers.
  • Gain media for microlasers.
• Optimization Potential: The authors suggest that coupling efficiency can be significantly improved by:
  • Tailoring the taper geometry (e.g., multi-stage tapers).
  • Reducing surface roughness to minimize scattering losses.
  • Engineering particle density and orientation.
  • Integrating resonant cavities or photonic crystals to enhance emission via the Purcell effect.

In conclusion, the paper successfully demonstrates a viable pathway for integrating active, broadband light sources into silicon nitride photonic circuits using lanthanide-doped microparticles, overcoming traditional material incompatibility issues.