Bell Labs Claims Laser First

Scientists at Bell Labs have invented a new type of laser that might one day slash the cost of building DWDM systems (see Bell Labs Lights Ultra-Laser).

It's called a "supercontinuum" laser, and it delivers a broad band of wavelengths simultaneously. This should not be confused with widely tunable lasers, which also offer a broad band of wavelengths, but only make one of them available at a time. Tunable lasers promise to reduce inventory costs by making it possible to use one type of laser instead of many different types throughout a DWDM system, whereas a single supercontinuum source could be used to replace all the lasers in a system.

And this idea isn't just pie in the sky. Last year, researchers from NTT Electronics Corp. (NEL) reported an experiment that used a single light source to supply all 106 channels in a DWDM system (see CLEO Report: Pick of the Papers).

However, NEL's light source was a complicated configuration involving a pulsed laser diode, an erbium-doped fiber amplifier (EDFA), and a piece of polarization-maintaining fiber. In the case of Bell Labs' laser, only the semiconductor laser diode is required, and it emits continuous – rather than pulsed – light.

Claire Gmachl and her co-workers report their results in today's issue of Nature (issue 415, p.883).

Gmachl has already made a bit of a name for herself for her work on the so-called "quantum cascade" laser. Invented in 1994 by Bell Labs researchers Federico Capasso (also an author on today's Nature paper) and Jerome Faist, quantum cascade (QC) lasers are designed in a completely different way from ordinary diode lasers. Indeed, the new laser unveiled today is a variation on a QC laser.

To understand how Gmachl's team made the new device, it's worth looking briefly at how an ordinary laser works. Ordinary lasers require both negatively-charged electrons and positively-charged "holes" (places in the crystal where electrons should be, but aren't). They attract each other, and when the electron pops into the hole, it gives up surplus energy as a photon of very specific energy. That energy determines its wavelength.

QC lasers work with electrons only. The semiconductor layers are designed in such a way that the electron can drop energy in a series of small amounts, giving up a photon each time it does so. "Consider it an electronic waterfall," says Bell Labs' description of a QC laser at http://www.bell-labs.com/org/physicalsciences/projects/qcl/qcl.html. The cascade of light generated in this way makes the QC laser much more powerful than existing lasers.

This waterfall effect is key to making the device broadband. Instead of creating all the small steps in the waterfall the same size, they can be made slightly different, so that each successive photon emitted has a slightly different wavelength. The laser reported today had 36 "steps" in the waterfall, each one created by growing thin layers of several different semiconductor materials.

The fact that all the different materials are grown on top of each other is significant because, with a bit of tweaking to the design of the layers, it is possible to create lasing at every wavelength in the band, not just at discrete wavelengths corresponding to the peak gain at each "step." This differentiates the device from so-called multiwavelength lasers, which are made by growing different materials side by side on a semiconductor wafer, and have wavelength spikes at the output.

But there is a catch. Since the electrons only make small energy jumps, the photons tend to have long wavelengths. So far, Bell Labs has only managed to make devices that emit light in the region 4 to 24 microns. (The device reported in Nature actually emitted all wavelengths in the range 6 to 8 microns.) It has been Bell Labs' stated intention for a quite a while to make a device that emits at the shorter telecom wavelengths of 1.3 and 1.55 microns, but so far no progress has been reported in this respect.

— Pauline Rigby, Senior Editor, Light Reading
Page 1 / 2   >   >>
Dr.Q 12/4/2012 | 10:52:16 PM
re: Bell Labs Claims Laser First >> "Anyone know the internal corporate spiel
>> on why Lucent still even does material
>> research?"

The reason that Lucent continues to do this kind of research is that at the Lucent/Agere split the Bell Labs president (Bill Brinkman) pulled the wool over Lucent president Rich McGinn, telling him that Lucent needed to keep *all* of the Bell Labs work intact in order for Lucent to be at the forefront of new components as well as systems.
Trouble with that rationalization is that soon Rich McGinn got tossed overboard, and Henry Schadt came back, took a look at the revenue stream and understood he couldn't afford thousands of people whose work was not contributing to his business. That is a key reason why Bell Labs lost thousands of people in the 2001 time frame.
fiber_in_my_diet 12/4/2012 | 10:52:25 PM
re: Bell Labs Claims Laser First In the spirit of the editor bashing thread, the article didn't mention the bunch of guys who did a wide spectrum light source well before Lucent. It was a continuous spectral source that, in theory, allows a gazillion lambdas for optical comms.

I think the inventors were in Menlo Park.....

Seriously, for those readers lacking in imagination, the Lucent invention is relevant in future when the group in cubes next to the one that invented the waterfall laser comes up with a low loss fiber that can do a transcontinental transfer of the photon energy from a single electron transition. If you're skeptical, look at what they did for multimode fiber at 10G. You saw the vision here on Lighreading first folks.
Big-dog 12/4/2012 | 10:53:47 PM
re: Bell Labs Claims Laser First It is very difficult to make a QC laser for telecom wavelengths due to current material limitations. A wavelength of 1.55 microns corresponds to an energy difference (either conduction band to valence band or between quantum confined states within a conducton band) of about 0.8 eV. For GaInAsP lattice matched to InP, one has <0.4 eV available in the conduction band. For AlGaAs on GaAs, the number is about 0.3 eV. It is highly doubtful if current III-V materials will ever be useful for QC telecom unless one goes to flouride fibers which require longer wavelengths and promise lower loss compared with silica fibers.

Aside from these, the practicalities of cryogenic cooled devices within the standard telecom scenario should also not be overlooked. My view of the QC devices is the same as the earlier hype from Bell Labs about cleaved coupled cavity lasers-nice science but-----!
spookydoodle 12/4/2012 | 10:53:59 PM
re: Bell Labs Claims Laser First Hi,

Depending on teh output power per wavelength, this laser may be used to chracterized DWDM components, such as EDFA's. For telecom, each emitted wavelength needs to be modulated to carry useful information. To do that one would have to first separate wavelengths and then direct them to a modulator. This may become an expensive scenario. However the advantage may be that since all wabvelengths are generated by a single laser, wavelength locker may be simple to design.

melao 12/4/2012 | 10:54:02 PM
re: Bell Labs Claims Laser First "Melao,

Bell Labs has been harping on about using QC lasers for telecoms for ever. I quote from the press release this week:

"We picked the range of 6 to 8 micrometer for laser action as a good range for a convincing demonstration of the idea. In the future, we may be able to custom tailor the laser to the specific needs of individual applications, including fiber optics."

Obviously, it's difficult to do make the laser at telecoms wavelengths or they would have done it by now.

[email protected]"

Yeah, i went right after that to the Nature site and i found that quote. Sorry the negative post.

Anyway, as i-¦ve seen the difficulty will be to match the quantum wells to the specific wavelength. And that-¦s a major issue i think, so probably it will take a while...

opticalwatcher 12/4/2012 | 10:54:07 PM
re: Bell Labs Claims Laser First Well, I am one of those "non-experts" who likes to keep up on up-and-coming technologies. So I appreciate the simple explanations as provided by Light Reading.

For those in the industry who like to come here speaking in technical terms while spewing their ever so important opinions, all I have to say is GET LOST! This is a non-expert forum. If your so darn smart, go forth and use your genius where its most needed.

Why am I so rude to these people? Here's a simple explanation. Invention comes from needs and ideas. People that spew negativity over new discoveries are backward thinkers (as opposed to forward thinkers), followers, and bean counters. Such important words like "bottom line" and "applicability" have more meaning to these people than words such as development, research, advancement. People like this remind me of those that believed the world was flat because that's all they had ever known, or those that believed the earth is the center of the universe.

Here's a question for you (you in general that is), should all scientific research be directed by ability to make a profit? I can hear the answers and sarcasm now. Your answers will most likely demonstrate that we are only a few inches in mental development from apes.
Pauline Rigby 12/4/2012 | 10:54:09 PM
re: Bell Labs Claims Laser First Melao,

Bell Labs has been harping on about using QC lasers for telecoms for ever. I quote from the press release this week:

"We picked the range of 6 to 8 micrometer for laser action as a good range for a convincing demonstration of the idea. In the future, we may be able to custom tailor the laser to the specific needs of individual applications, including fiber optics."

Obviously, it's difficult to do make the laser at telecoms wavelengths or they would have done it by now.

[email protected]
mu-law 12/4/2012 | 10:54:12 PM
re: Bell Labs Claims Laser First "This has to include citing the possible long term impact on telecom equipment."

No disagreement here... I think the issue is that the commercial prospects that can be inferred for this technology are neither overt or compelling. If you can present a different case, I would encourage you to do so...
bongwang 12/4/2012 | 10:54:14 PM
re: Bell Labs Claims Laser First Peter:

I often don't resonate with you, but in this case I heartily agree. LR does an outstanding job of making it possible for non-experts to undertand optical networking and its technologies.

As to this thread, more emotion than thought has been the basis for the posts.

Keep these articles coming, LR has become an important source of information for many thousands of readers, and we do look forward to the daily and weekly updates.

gea 12/4/2012 | 10:54:15 PM
re: Bell Labs Claims Laser First Does Lucent (part of Bell Labs) still plan to manufacture components? I am slightly suprised that this kind of research is done there, as opposed to Agere (the recently spun off coponent division).
Anyone know the internal coporate spiel on why Lucent still even does materials research?
Page 1 / 2   >   >>
Sign In