Hybrid integration of a SESAM with a silicon nitride photonic integrated circuit

summary

The gist

The realization of high-performance femtosecond lasers on photonic integrated circuits (PICs) requires robust…

In short

This work integrates a semiconductor saturable absorber mirror (SESAM) with a silicon nitride (SiN) photonic integrated circuit via micro-assembly to create high-performance femtosecond lasers. A key innovation is using inverse waveguide tapers to manage coupling loss and control the effective mode area at the interface. Characterization shows that controlling taper tip width allows precise tuning of SESAM fluence, enabling design for chip-scale, self-starting ultrafast lasers.

Key concepts

SESAM
A semiconductor saturable absorber mirror is a material used in lasers that selectively absorbs light based on its intensity. In this work, an indium phosphide-based SESAM is bonded to the SiN waveguide facet. Its non-linear properties are what allow it to act as a saturable absorber, crucial for pulse shaping and lasing.
Inverse Waveguide Taper
This is a specially designed shape for the coupling structure between the SESAM and the waveguide. Standard waveguides confine light tightly, causing divergence at the bonding edge. The inverse taper expands this mode at that facet, which significantly reduces coupling loss by improving overlap and minimizing diffraction losses during bonding.
Saturation Fluence
This parameter describes how much optical power is needed to fully saturate the SESAM's absorption effect. By controlling the geometry of the interface using different taper tip widths (e.g., 500 nm vs. 700 nm), researchers can precisely adjust this saturation fluence, which directly dictates the required pulse energy for mode locking.
Hybrid Integration
This strategy involves physically bonding two different optical components—the non-linear SESAM and the SiN PIC—using a UV-curable epoxy. This micro-assembly technique allows the non-linear functionality of the absorber to be directly incorporated onto a standard photonic chip, creating an integrated laser source.

Terminology used across episodes

This episode discusses

The paper

Hybrid integration of a SESAM with a silicon nitride photonic integrated circuit · Read on arXiv

Photonics Research Group, Ghent University imec · RefleKron

Transcript

Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.

Kai: We're looking at the paper: "Hybrid integration of a SESAM with a silicon nitride photonic integrated circuit".

Mira: Here's the gist of the paper. The realization of high-performance femtosecond lasers on photonic integrated circuits (PICs) requires robust, low-loss coupling between a saturable absorber and a photonic waveguide <ref:2611.08057#pg2>.

Kai: This work demonstrates the hybrid integration of a semiconductor saturable absorber mirror (SESAM) with a silicon nitride (SiN) PIC by micro-assembly <ref:2611.08057#pg3>.

Mira: Hybrid Integration Strategy The core of the work involves a micro-assembly strategy that transfers the non-linear functionality of the absorber onto the PIC <ref:2611.08057#pg3>.

Kai: Specifically, an indium phosphide (InP)-based SESAM is bonded directly to the polished edge facet of a SiN PIC using a UV-curable epoxy <ref:2611.08057#pg3>.

Title and authors: Kai: This paper by Amaryllis Verstraete and her team summarizes their characterization of both an unbonded SESAM and a SESAM that’s actually bonded to those inverse-tapered silicon nitride waveguides <ref:2611.08057#pg2>. They are testing linear and non-linear performance across a wavelength range from one thousand five hundred forty to one thousand six hundred nanometers.

Mira: That characterization includes measuring things like the waveguide coupled modulation depth, which is how effectively the SESAM modulates the light in the waveguide, and they also measure maximum reflectivity and saturation fluence <ref:2611.08057#pg1>. It's about getting concrete performance numbers on this bonded device.

Lev: It’s telling us that this approach isn't just theoretical; they are actually measuring concrete performance numbers on a bonded device <ref:2611.08057#pg2>. They are using two SESAMs from the same wafer to compare before and after bonding, which adds a layer of experimental rigor to their results.

Kai: And the main result they highlight is that at one thousand five hundred sixty nanometers, for this bonded device, they achieved a waveguide coupled modulation depth of five point two percent and a maximum reflectivity of thirty-seven point two percent <ref:2611.08057#pg1>. That's the linear performance baseline they established on this integrated setup.

Mira: That saturation fluence number is also important; they reported it as forty seven microjoules per square centimeter <ref:2611.08057#pg1>. That tells us the energy level required to actually saturate the absorber in this configuration, which is a key metric for pulse energy requirements later on.

Lev: Those numbers are useful, but we need to be careful about what those numbers mean when you try to move them onto actual hardware; they're operating at one thousand five hundred sixty nanometers, which is right in that silica window where many of these lasers operate <ref:2611.08057#pg2>.

The paper's summary: Kai: The authors suggest a few ways to improve this setup, starting with the waveguide tapers themselves. They designed tapers with tip widths of seven hundred sixty nanometers and five hundred sixty nanometers, which resulted in fabricated tips of seven hundred nanometers and five hundred nanometers when they measured them using scanning electron microscopy <ref:2611.08057#pg3>.

Mira: What that actually means is that narrowing the taper tip changes the effective mode area at the facet; they found that this narrows the confinement of the fundamental mode but increases its effective mode area from one point six six square micrometers when the width was one thousand ten nanometers to one point eight four square micrometers when it was seven hundred nanometers <ref:2611.08057#pg3>.

Lev: That change in effective mode area is what gives them control over the fluence on that SESAM, which is a big deal because you want to adjust how much pulse energy you need before saturation occurs <ref:2611.08057#pg3>.

Kai: And they also found that a narrower tip couples more light back into the waveguide, but its larger mode area lowers the fluence and raises the pulse energy needed to saturate the absorber; that’s a direct trade-off they found <ref:2611.08057#pg3>.

Mira: They also noted that simulations suggest even if you use a thinner bonding gap, it would raise the coupled reflectivity, so you have this design degree of freedom to tune both the mode area and the fluence by controlling the taper tip width <ref:2611.08057#pg3>.

Lev: If you can reliably engineer the mode area and the fluence this way, you have a pathway toward fully integrated, self-starting chip-scale femtosecond lasers that might replace those table-top systems <ref:2611.08057#pg6>.

The paper's improvements: Kai: To wrap up on this paper, "Hybrid integration of a SESAM with a silicon nitride photonic integrated circuit," they show that by using micro-assembly and specific inverse-tapered waveguides, you can integrate the non-linear functionality onto the PIC <ref:2611.08057#pg1>. They showed concrete figures at one thousand five hundred sixty nanometers with a saturation fluence of forty seven microjoules per square centimeter <ref:2611.08057#pg1>.

Mira: The implication is that this method gives engineers a way to control loss and saturation energy by engineering the silicon nitride waveguide facet itself, which is very useful for designing integrated laser systems <ref:2611.08057#pg6>. It shows SESAMs can be scaled up with SiN photonics.

Lev: From where I look at it, if you can reliably engineer the mode area and the fluence this way, you have a pathway toward fully integrated, self-starting chip-scale femtosecond lasers that might replace those table-top systems <ref:2611.08057#pg6>. The overhead for real hardware would depend on how well that coupling loss is managed in practice.

Kai: Exactly. The paper shows the non-linear performance they got is compatible with thirty picojoules of intracavity pulse energy, which is what you see in some silica waveguide mode-locked lasers <ref:2611.08057#pg8>. That's a solid starting point for chip-scale designs.

Mira: It gives us a route toward those integrated systems that are self-starting and operate on a chip; we need to keep looking at how this micro-assembly works with different materials and structures next time <ref:2611.08057#pg6>.

Lev: I just think the real test is whether that coupling loss remains low enough when you move it from simulation to actual fabricated devices <ref:2611.08057#pg8>.

Kai: Yeah, that's the next hurdle. We’ll see if we can push those pulse energies even higher with this integrated approach in future work.

Conclusion: Kai: So, this paper on "Hybrid integration of a SESAM with a silicon nitride photonic integrated circuit," it boils down to showing you can take the non-linear function of an absorber and physically bond it onto a silicon nitride waveguide using micro-assembly.

Mira: It means they're not just keeping the absorber separate; they are putting that non-linear functionality right where you need it on the chip, using UV-curable epoxy to stick it to the facet.

Lev: From an engineering viewpoint, that combination is tough because you are dealing with a semiconductor component and a high-index dielectric waveguide at the same time, right? The title points to solving that integration challenge.

Kai: Right. And they show concrete numbers at one thousand five hundred sixty nanometers where they got a waveguide coupled modulation depth of five point two percent and a maximum reflectivity of thirty-seven point two percent on this bonded device.

Mira: That saturation fluence number is also important because it's forty seven microjoules per square centimeter, which tells us the energy level you need to actually saturate the absorber in that configuration.

Lev: Those numbers are useful for context, but we have to be careful about what those numbers mean when you try to move them onto actual hardware; they are operating at one thousand five hundred sixty nanometers, which is right in that silica window where many of these lasers operate.

Kai: Exactly. The paper shows the non-linear performance they got is compatible with thirty picojoules of intracavity pulse energy, which is what you see in some silica waveguide mode-locked lasers. That’s a solid starting point for chip-scale designs.

Mira: It gives us a route toward those integrated systems that are self-starting and operate on a chip. We need to keep looking at how this micro-assembly works with different materials and structures next time.

Lev: If you can reliably engineer the mode area and the fluence this way, you have a pathway toward fully integrated, self-starting chip-scale femtosecond lasers that might replace those table-top systems. That’s where the real overhead lands for running it on a real system.

Kai: So, to wrap up on "Hybrid integration of a SESAM with a silicon nitride photonic integrated circuit," they show you can integrate non-linear function onto the PIC by bonding an SESAM using micro-assembly and inverse-tapered waveguides.

Mira: The implication is that this method gives engineers a way to control loss and saturation energy by engineering the silicon nitride waveguide facet itself, which is very useful for designing integrated laser systems. It shows SESAMs can be scaled up with silicon nitride photonics.

Lev: I just think the real test is whether that coupling loss remains low enough when you move it from simulation to actual fabricated devices.

Kai: Next time we look at papers, we want to see how these integration techniques hold up when you push the pulse energy even higher in a real system.

More episodes

← Home