Engineers grow nanolasers on silicon, pave way for on-chip photonics

Feb 06, 2011
The unique structure of the nanopillars grown by UC Berkeley researchers strongly confines light in a tiny volume to enable subwavelength nanolasers. Images on the left and top right show simulated electric field intensities that describe how light circulates helically inside the nanopillars. On the bottom right is an experimental camera image of laser light from a single nanolaser. Credit: Connie Chang-Hasnain Group

Engineers at the University of California, Berkeley, have found a way to grow nanolasers directly onto a silicon surface, an achievement that could lead to a new class of faster, more efficient microprocessors, as well as to powerful biochemical sensors that use optoelectronic chips.

They describe their work in a paper to be published Feb. 6 in an advanced online issue of the journal .

"Our results impact a broad spectrum of scientific fields, including materials science, transistor technology, science, optoelectronics and ," said the study's principal investigator, Connie Chang-Hasnain, UC Berkeley professor of electrical engineering and computer sciences.

The increasing performance demands of electronics have sent researchers in search of better ways to harness the inherent ability of to carry far more data than can. Optical interconnects are seen as a solution to overcoming the communications bottleneck within and between .

Because , the material that forms the foundation of modern electronics, is extremely deficient at generating light, engineers have turned to another class of materials known as III-V (pronounced "three-five") semiconductors to create light-based components such as light-emitting diodes (LEDs) and lasers.

But the researchers pointed out that marrying III-V with silicon to create a single optoelectronic chip has been problematic. For one, the atomic structures of the two materials are mismatched.

"Growing III-V semiconductor films on silicon is like forcing two incongruent puzzle pieces together," said study lead author Roger Chen, a UC Berkeley graduate student in electrical engineering and computer sciences. "It can be done, but the material gets damaged in the process."

Moreover, the manufacturing industry is set up for the production of silicon-based materials, so for practical reasons, the goal has been to integrate the fabrication of III-V devices into the existing infrastructure, the researchers said.

"Today's massive silicon electronics infrastructure is extremely difficult to change for both economic and technological reasons, so compatibility with silicon fabrication is critical," said Chang-Hasnain. "One problem is that growth of III-V semiconductors has traditionally involved high temperatures – 700 degrees Celsius or more – that would destroy the electronics. Meanwhile, other integration approaches have not been scalable."

Shown is a schematic (left) and various scanning electron microscope images of nanolasers grown directly on a silicon surface. The achievement could lead to a new class of optoelectronic chips. Credit: Connie Chang-Hasnain Group

The UC Berkeley researchers overcame this limitation by finding a way to grow nanopillars made of indium gallium arsenide, a III-V material, onto a at the relatively cool temperature of 400 degrees Celsius.

"Working at nanoscale levels has enabled us to grow high quality III-V materials at low temperatures such that silicon electronics can retain their functionality," said Chen.

The researchers used metal-organic chemical vapor deposition to grow the nanopillars on the silicon. "This technique is potentially mass manufacturable, since such a system is already used commercially to make thin film solar cells and light emitting diodes," said Chang-Hasnain.

Once the nanopillar was made, the researchers showed that it could generate near infrared laser light – a wavelength of about 950 nanometers – at room temperature. The hexagonal geometry dictated by the crystal structure of the nanopillars creates a new, efficient, light-trapping optical cavity. Light circulates up and down the structure in a helical fashion and amplifies via this optical feedback mechanism.

The unique approach of growing nanolasers directly onto silicon could lead to highly efficient silicon photonics, the researchers said. They noted that the miniscule dimensions of the nanopillars – smaller than one wavelength on each side, in some cases – make it possible to pack them into small spaces with the added benefit of consuming very little energy

"Ultimately, this technique may provide a powerful and new avenue for engineering on-chip nanophotonic devices such as lasers, photodetectors, modulators and solar cells," said Chen.

"This is the first bottom-up integration of III-V nanolasers onto silicon chips using a growth process compatible with the CMOS (complementary metal oxide semiconductor) technology now used to make integrated circuits," said Chang-Hasnain. "This research has the potential to catalyze an optoelectronics revolution in computing, communications, displays and optical signal processing. In the future, we expect to improve the characteristics of these lasers and ultimately control them electronically for a powerful marriage between photonic and electronic devices."

Explore further: New imaging technique shows how cocaine shuts down blood flow in mouse brains

Related Stories

Ultrathin alternative to silicon for future electronics

Nov 22, 2010

There's good news in the search for the next generation of semiconductors. Researchers with the U.S. Department of Energy's Lawrence Berkeley National Laboratory and the University of California Berkeley, ...

A new twist for nanopillar light collectors

Nov 16, 2010

Sunlight represents the cleanest, greenest and far and away most abundant of all energy sources, and yet its potential remains woefully under-utilized. High costs have been a major deterrant to the large-scale ...

Photoluminescence in nano-needles

Apr 22, 2008

Silicon is the workhorse among semiconductors in electronics. But in opto-electronics, where light signals are processed along with electronic signals, a semiconductor that is capable of emitting light is needed, which silicon ...

Recommended for you

What time is it in the universe?

1 hour ago

Flavor Flav knows what time it is. At least he does for Flavor Flav. Even with all his moving and accelerating, with the planet, the solar system, getting on planes, taking elevators, and perhaps even some ...

Breakthrough in light sources for new quantum technology

8 hours ago

One of the most promising technologies for future quantum circuits are photonic circuits, i.e. circuits based on light (photons) instead of electrons (electronic circuits). First, it is necessary to create ...

A new, tunable device for spintronics

19 hours ago

Recently, the research group of Professor Jairo Sinova from the Institute of Physics at Johannes Gutenberg University Mainz in collaboration with researchers from the UK, Prague, and Japan, has for the first time realised ...

User comments : 14

Adjust slider to filter visible comments by rank

Display comments: newest first

Quantum_Conundrum
5 / 5 (4) Feb 06, 2011
Well, if the laser itself is "sub-wavelength", then it's smaller than 950nm. This is incredible because it means it's at least 50nm smaller than the smallest thing seen recently in plasmonic lasers.

Lol. Bring on 3d optical circuitry.

Once mass production with this technology gets implemented the implications are staggering. Intel, IBM, AMD, or whoever, will probably be able to instantly double, triple, or quadruple the number of processor cores in a computer without even much effort. This just as a "trial run" product line to fill time.

Then the real fun will begin with a totally new architecture that actually makes the best use of the technology. We're talking hundreds of terahertz processors in a PC...maybe a decade...
Dummy
not rated yet Feb 06, 2011
A lot of people in this world are gonna get left behind...
Sanescience
5 / 5 (1) Feb 06, 2011
All this pales compared to the paradigm shift that quantum programming will bring. That is some brain bending stuff!
Bob_Kob
3 / 5 (1) Feb 07, 2011
All this pales compared to the paradigm shift that quantum programming will bring. That is some brain bending stuff!


Yeah but will most likely take more than 10 - 15 years as this would.
Beard
not rated yet Feb 07, 2011
Hexagons are such a ubiquitous polygon.

Snowflakes, honeycombs, compound eyes, carbon allotropes and Saturn's polar hexagon. Now this too.
claudiu_andone
not rated yet Feb 07, 2011
How about a revolution in imagery and displays. I can see nanolasers taking resolution and color of a display to perfection.
Royale
not rated yet Feb 07, 2011
aren't we just about at the color shade limit the human eye can see already anyway?
wealthychef
not rated yet Feb 07, 2011
I most wonder about the potential ability to speed up communications to memory, I/O, and GPUs with this. These devices all seem bandwidth limited by the bus currently.
sender
not rated yet Feb 07, 2011
Adding graphene nanocoils on these would hopefully increase polarization strength via direct induction.
Quantum_Conundrum
5 / 5 (1) Feb 08, 2011
aren't we just about at the color shade limit the human eye can see already anyway?


We were past that a very long time ago, excluding a few exceptional people.

Color depth in computers is actually so high that even video games do not use a fraction of what the video cards are capable of for color depth. They use the polygons, pixel resolution, and frame rate. Increasing color depth has long been pointless. Increasing framerate beyond about 40 is mostly pointless, except in weird situations where it allows a player to abuse the game engine to "break" some abilities.

The top 5% of Starcraft pro gamers might get some benefits out of having higher frame rates, because they have the eyes and the actions per minute to abuse it, but normal people cannot even tell the difference.
Origen
not rated yet Feb 09, 2011
If these nano-lasers are 'subspectrum' does that mean that they can can be used to detect the speed and position of objects that are smaller than the shortest wavelength of light? If that's the case than this is a momentous achievement!
quasi44
not rated yet Feb 09, 2011
I most wonder about the potential ability to speed up communications to memory, I/O, and GPUs with this. These devices all seem bandwidth limited by the bus currently.

But if we are already above the board with our 'laserboard' then we are no longer bound by the strict physical limit of the bus. A series of deflectors right to memory buffer. Seems like it gives a way for optical runlevels to me. Liberal use of high quality films, optical gels, fiber optics and high efficiency heatsinks and this is a done deal in a year; check out GlobalSpec, they have made to order optical transfer tech that could make this fly now.
Quantum_Conundrum
not rated yet Feb 12, 2011
I most wonder about the potential ability to speed up communications to memory, I/O, and GPUs with this. These devices all seem bandwidth limited by the bus currently.


Actually, these devices will eventually become photonic "wireless buses", which will enable much more compact processors which can also make use of the advantages of true 3-d circuitry.
Shane_Hammer
not rated yet Feb 28, 2011
whell, It looks like they are getting closer and closer to making some thing like the human brain, the next few years will bring some awsome things i think