This article has been written by Alex who hosts this years JePPIX course with Photon Design; essentially I explain Photon Design’s software for a living and one of the best places I get to do this is at the JePPIX yearly training program in Eindhoven.

For me, many demo sessions can feel somewhat prescriptive, moving from one basic device to the next: click here, next here, well done, you’ve finished… I’m often asking ‘but why am I doing all of this!?’ So this year I’ve put together a more exciting set of demos, many of which build to an impressively involved simulation, plus plenty of context along the way about the photonics industry and the simulation landscape.

We have this fantastic new example in our circuit simulator PICWave’s last update, a laser built in Si + III-V hybrid epi. When our laser team showed me how their design brings together work from many pieces of our software I thought: ‘THIS! This is what a demo session should be!’

Let me share some of what the folks at JePPIX got to work through:

Left: Diagram of hybrid laser including tapers. SOI in blue/ purple, III-V in red.

Right: Modes evolving over hybrid laser taper (simulated in FIMMWAVE)

In a hybrid epi, we need to move modes confined in the lower silicon waveguides up into the III-V material to experience gain. To do this we taper the waveguides; the preferred method for this is FIMMPROP’s EME simulations.

Why?

• EME simulations are invariant to length. The taper is 40 um or 400um long, the simulation time is the same.
• FIMMPROP’s taper algorithm excels in the market - I’m not sure if without FIMMPROP’s dynamic step sizing this taper would be viable in a short demo session (reach out to us to learn more about how this works, it could save you a great deal of time).
• Once a taper simulation is done you can relatively instantly perform a ‘length sweep’ to find what length is needed to make your taper adiabatic.

If you’re using popular FDTD simulations for tapers then you will have plenty of time (while waiting for your simulations) to think if there’s a faster way of doing things! Read more on comparing these methods here: LINK

To make a laser we’ll need reflective gratings so next attendees will further investigate FIMMPROP’s ability to simulate periodic structures with ease. I really like this demo example where the instructions pull away a little and give attendees a chance to apply what they’ve learned to design the gratings the task asks for.

Again, FIMMPROP is a great method here:

• Periodic simulations are incredibly fast, simulating just one period and re-using the result.
• Once again, fast scans make a parameter sweep with thousands of points possible for the attendee’s laptop.

I should add here that we run all the simulations locally at these sessions so attendees see a realistic example of how fast their simulations are running.

These results all come together in an active laser simulation in PICWave where we pull together results from the previous examples tuning Bragg gratings and investigating tapers. As well as this, the epitaxy has already been set up with thorough simulations of the epi’s gain spectra from our software Harold; this is Photon Design operating on full force! 

A huge thanks again to the folks at JePPIX for creating a place where we get to talk about the most exciting topics in the industry and hopefully help new engineers save hours of their time by picking the right simulation methods.