Constance Chang-Hasnain. [Image: © Noah Berger]
For its May 2020 print article “The Laser at 60,” OPN interviewed a range of OSA Fellows to get their insights on some particularly interesting horizons in laser research today. We’re presenting a selection of those interviews online. Below is an edited version of our interview with Constance Chang-Hasnain, University of California, Berkeley, USA. A pioneer in the development of second-generation vertical-cavity surface-emitting lasers (VCSELs), semiconductor optoelectronic devices and other areas, Chang-Hasnain is will serve as President of The Optical Society in 2021.
VCSELs have been a tremendously successful technology, and seem to be taking off into new markets. At a very high level, what has made VCSELs so successful?
Two reasons, I think, come mind as the most important. One is the user-friendliness of the output beam—it’s a circular beam that is very easy to guide it, easy to couple it to, for example, optical fiber or free space.
The other reason is that VCSELs are extremely scalable. Once you are able to master the fabrication of one, you can make tons of them. The fabrication is very similar to that of an LED, with the same characteristics of wafer-scale processing and testing and so on, and they can be processed in an array. And yet at the same time, they have the characteristics of a laser: the provide coherent light, can be turned on and off, modulated very fast—at rates of tens of gigahertz—and so on.
How did you get first get involved in working with VCSELs?
It was really being in the right place at the right time. I worked at Bellcore [New Jersey, USA] in the very beginning of what we call the “second generation” of VCSEL development. The first generation, of course, was pioneered by Professor Kenichi Iga at the Tokyo Institute of Technology. But the particular processing and manufacturing or fabrication techniques that he was using at the time, the tools that were available to him, did not make it particularly easy for people to jump in.
And then, in 1988 or 1989, there was a lot of development by lots of different people, putting different pieces of processing tools together and making them available. So it was a combination of advances in MBE [molecular-beam epitaxy] and etching and other tools, with an improved understanding of semiconductor and quantum-well semiconductor physics adds up together. And there was breakthrough work at Bell Labs and Bellcore, led by people like Jack Jewell and Jim Harbison.
I was at Bellcore at that time. There was a need to understand semiconductor laser physics—not just making the lasers, but understanding how to actually better design them, how to modulate them, how to predict the mode pattern and so on. That's how I got involved.
VCSELs are now making a big push into consumer electronics, such as 3D smartphone sensors. What is making that expansion possible?
Well, two things, I think. One is that the market has finally arrived. You now have a smartphone in everybody’s hands, and thanks to Moore’s law it has the capability to do a lot of signal processing.
The other thing is that, due to smartphones and the internet generally, bandwidth needs have gotten much, much higher. And that’s meant that the fiber communications backbone, and network data centers, for example, have already been utilizing VCSELs as their primary vehicle for shorter-distance interconnects inside the data center. The VCSELs had to be massively produced already for these uses. So, quite a few companies have developed the manufacturing know-how and technology, and proved that the devices could be reliable or ready for that application.
So when the market or application needs started to push VCSELs for a little bit more—for larger arrays and cheaper prices—the technology was ready.
You have had a few conference presentations recently on high-contrast-grating (HCG) VCSELs. Could you talk about those?
The high-contrast grating is one single layer of ultrathin material that is roughly subwavelength in period. Using one single layer like that, we can replace, the entire very thick stack of DBR [distributed Bragg reflector] mirrors in the VCSEL. That means that it can reduce the thickness of a VCSEL roughly to half of its present size.
As a result, you can make, once again, the processing and the manufacturing much easier, and cut the epitaxy needed by about one half. That’s important, because the epitaxy still requires about eight to nine hours to grow the material—it is actually a bottleneck in mass production. If you reduce that by half, that’s a huge amount of savings in time and cost.
On top of that, you can make lasers much denser. Again, every dimension is reduced by a factor of two, so you reduce the material by a factor of eight. So that’s number one—larger, denser arrays at lower costs.
But the HCG VCSEL has some other properties as well. For example, it can control the emission mode, to make it single mode. It can help if you build it into a MEMS structure, because it’s only about 200 nanometers thick. So you can create a very flexible MEMS to make a very fast, wavelength-swept source.
In fact, this particular aspect is extremely exciting. If you look at the means to control the wavelength of a laser, there’s no other structure, aside from a MEMS VCSEL, that can give you single, pure wavelength sweeping, at rates close to a megahertz—a few hundred kilohertz rate. And over a very wide range—Δλ/λ on the order of 15% in some demonstrations.
This is already making headway in terms optometry, for examining and 3D imaging of the retina in real time. Many people are exploring other medical applications. In real-time artery surgery, like surgery in the heart, for example, surgeons would like to know as they cut through your tissue how deep the cut is, in real time, to allow for precision cutting in surgery. This is the only technique that allows you real-time imaging of treatment with micron resolution.
So it’s very exciting—this ability to combine VCSELs with MEMS.
Are there other developments in VCSELs on the horizon that we should be looking for in the next few years?
Yes, I think so. For example, the current commercial applications we mentioned—3D imaging and 3D sensing, particularly in the smartphone, consumer electronics—those VCSELs are still using older technology, probably developed over the past 20 years. Now, 20 years ago that was really state-of-the-art research, but it took time to become mass-producible, high-yield products.
But now, looking forward, what will be really interesting is to explore, for example, very-high-power, coherent VCSELs. I think that will be a very interesting area of research that will enable many more imaging applications, over longer distances, such as lidar for autonomous vehicles, for example.
Another area that will be very interesting will be exploring different wavelengths. So far, the communications lasers are still in the near-infrared. There have already been publications of UV VCSELs, as well red VCSELs; some people are working on green lasers. So extending the wavelength range, covering visible, UV—and the far-infrared, too, for spectroscopy applications, ranging from food safety to medical applications to many others in communications. Extending wavelength range will, I think, be tremendous.
And then finally, the “final frontier” from a laser physics point of view is VCSELs with really tiny cavities. Typically, the VCSEL it has really poor linewidths. But to get very high coherence, you need the linewidth to be very, very narrow. Semiconductor lasers can get to the kilohertz range, and solid-state lasers can go even to millihertz, which means you’re exceedingly coherent. But VCSELs thus far have not been too coherent—typically the best results are between 5 to 10 megahertz. Can you go narrower when the cavity is small, and the gain region is small? That would be very, very interesting to find out.
A final question. The laser has made such a difference, on so many fronts. Are there areas, beyond the ones that we’ve talked about, that you find particularly exciting or interesting for the future?
I find integrated optics and integrated optoelectronics fascinating. All the silicon substrates for photonics—silicon photonics; silicon nitride, alternative methods to build on a silicon substrate to make integrated optics. Silicon is a really great platform. There are some challenges, certainly, and a lot of funding has already gone toward building foundries to do research on this. But even so, more research, particularly on materials suitable to be leveraged on silicon platforms, is an area that I would keep an eye on.
And then I guess, at the systems level of research, new ways of using lasers for distance ranging—lidar, but much more—is very exciting. Some people, for example, are working on the frequency domain, in what’s called frequency-modulated, continuous-wave lidar, or FM CW lidar.
And then, of course, there’s the quantum domain. Quantum optics and quantum physics are really very exciting areas, and worth watching.