Transistor Moment

Flying back home from San Jose I couldn't help wonder with excitement if our field is on the verge of a "transistor moment." Maybe it was just my CLEO conference euphoria coupled with high-end caffeine from Cafe Frascati still in my system. However, I feel like something big is going to happen, particularly in the field of photonic circuits and nanophotonics.

 The explosion of work in this subarea is impressive and CLEO hosted a number of talks from the leaders and pioneers in this field. You can still watch a handful of these on the CLEO On Demand video  such as Yurii Valsov's plenary talk on fundamentals and applications of silicon nanophotonics, Larry Coldren's tutorial, CWK1.1, on single-chip transmitters and receivers, and Dave Welch's tutorial, JM4.I.1 on semiconductor photonic integrated circuits, just to name a few. Cutting edge science is interfacing with better fabrication processes- repeatability and low cost. At the poster session on Wednesday night it seemed every group was using some kind of micro ring resonator. Simple photonic circuits are becoming standard. Will our children have the nanophotonic equivalent of a Heathkit radio- something like "My first Fab." It seems a sure thing to me that my daughters will be using optical/electrical hybrid computers in their lifetime. And it seems even more certain to me that nanophotonics is the future of our field.

However, will something even bigger, more profound, and unexpected happen like when Walter Brattain dumped his amplifier experiment in a thermos of water in 1947 to successfully demonstrate electrical gain of what was to become the transistor? The same little amplifier that gave birth to a small startup company named Sony and then later to Texas Instrument, Intel and the entire business of integrated circuits and computation as we know it. The transistor was at first a "mere" amplifier. Later it became the foundation for all computer logic and a new era of technology. I wonder what is within our grasp that we don't realize.

Yurii Vlasov used imagery from the Wizard of Oz in his plenary talk of a road to follow to the Emerald City (our goals of nanophotonics and computation and the windy road we will take). However, I wonder what ruby slippers we are wearing right now. What  "transistor potential" is waiting to be unlocked. It's a good time to be in the field of photonics. We will be the leaders of the new information age and the technology that drives and supports it.

Limber-up for the Postdeadline Session Tonight

From Postdeadline paper CTh5D.1 "Wavelength-Size Silicon Modulator." Scanning electron micrograph of the silicon integrated waveguide modulator.

Make sure to stretch your legs if you want to move from session to session in this frenzy of fantastic photonics research (say that five times fast). Tonight from 8:00-10:00 pm marks the crème de la crème of contributed papers to CLEO. I haven't quite made up my mind of which to attend, but found a number of them particularly exciting:

CTh5D.1, "Wavelength-size Silicon Modulator"  from V.J. Sorger et al.

 This is work out of the Zhang Lab from Berkely showing an optical modulator with 1-dB/micron extinction (a 20 micron long device gives 20 dB extinction). The modulator is based upon tuning the carrier concentration of an active nm-thin layer of Indium Tin Oxide sandwiched between a MOS structure. Just yesterday, Larry Coldren from UCSB was gently ribbing the silicon folk in his tutorial CW1k.1, "Single-chip Integrated Transmitters and Receivers" for the dearth of practical active components such as a modulator. Coldren sees InP based photonic circuits as the more robust platform for photonic integrated circuits. However, great work like this from the Zhang group will be pushing silicon to the forefront.

CTh5C.4, "In Vivo Three-Photon Microscopy of Subcoritical Structures wihtin an Intact Mouse Brain" from N. Horton et al.

 This work from the Xu Group from Cornell University uses a clever choice of longer-excitation wavelength coupled to the improved localization of three photon fluorescence in order to image deep through intact tissue. Even though longer wavelengths are more readily absorbed in tissue, they are significantly less scattered. The overall effect is higher throughput and deeper penetration. Combine that with a 1/z⁴ fall-off in three-photon fluorescence signal (tighter localization), and now you can make beautiful images of intact tissue. The Xu group shows 1.2 mm stack of brain tissue taken in 4 micron increments. The broader goal will be to eventually use this for optical biopsy in humans. I would prefer to have my tissue scanned with a laser rather than excised from my body with a knife by a surgeon, wouldn't you?  


CTh5C.1, "Demonstration of a Bright 50 Hz Repetition Rate Table-Top Soft X-Ray Laser Driven by a Diode-Pumped Laser" from B. Reagan et al.

This work from the Rocca Group of Colorado State University and the Research Center for Extreme Ultraviolet Science and Technology shows a significant improvement of table-top soft x-ray lasers. To see how quickly this group is improving these  systems, just look at a March 2012 Laser Focus World feature article highlighting their work- now outdated. The aim of table-top soft x-ray research is to bring systems that are typically found at a shared national lab facility to the many optics tables of university labs and industry. Applications for coherent soft x-rays include laser-induced materials processing at the nanoscale level as well as ultrafast characterization of nanoscale motion. Spectra Physics or Coherent may not be selling ultrfast soft x-ray lasers just yet, but this paper shows a  5-fold increase in repetition rate (important for higher average power applications) and a 20-fold increase in pulse energy from previous best efforts.

ATh5A.4 "Highly Efficient GaAs Solar Cells with Dual Layer of Quantum Dots and a Flexible PDMS Film"  from C. Lin et al.

In this paper a Taiwanese collobaration from the Institute of Photonic Systems, National Chiao Tung University, and the Research Center for Applied Sciences has shown a 22 % enhanced efficiency in a GaAs solar cell by spraying a coating of UV absorptive quantum dots onto a polydimethylsiloxane film at the top surface of the cell. This collaboration has found a clever way to not let so many UV photons from the sun go to waste.

Protecting Troops and Civilians with Light

From Joint IED Defeat Organization (JIEDDO) https://www.jieddo.dod.mil. Soldiers from the 713th Engineer Company, out of Valparaiso, Ind., conducted counter improvised explosive device training at Camp Atterbury Joint Maneuver Training Center Aug. 20. Photo by Staff Sgt. Matthew Scotten.

The panelists from Tuesday's 2:00 pm Market Focus, Defense: Laser Interrogation for Standoff Detection of Hazardous Materials, presented the audience with a difficult problem to which the U.S. Department of Defense is allocating many resources and substantial funding:

How can you accurately detect threats from chemical, biological, radiological, nuclear, or high-yield explosives (CBRNE) from a safe stand-off distance to protect or warn those in harms way?

Laser spectroscopy is the short answer, be it UV Raman, NIR Raman, Long Wave Absorption Spectroscopy, Laser-induced Breakdown Spectroscopy (LIBS), Photoacoustice Spectroscopy, Ultrafast Spectroscopy, just to name a few. However, what kind of spectroscopy you use to identify a threat is just the beginning to making a system that can function in rugged battlefield environments and accurately deliver the information you need in the time you need it.

Panelist Scott Robertson,  Research Senior Manager at Lockheed Martin, posed just how difficult this can be with some specific targets of the type of systems needed in the field. One project whose objective was to analyze threats by the vapors and residues from vehicles needed a stand-off detection distance of 400 m, an entire scan, detect and process time of 1.0 second, with a false alarm rate of only 1 in one million, and packaged in a volume of 1 cubic meter. Another specification target was to be able to scan an area of 2,700 square meters per second while searching a road 100 m wide, while traveling 60 mph.

There are other constraints as well. Tom Stark (no relation to Tony from the Iron Man series), from Landmark Technologies Joint IED Defeat Organization, reminded the audience that 99.9% of the people in an area you want to scan are not the threat. You can't and  don't want to blatently scan a crowd with a potentially dangerous high-power laser system. Another constraint therefore is laser safety, particularly eye safety. Add this to the checklist of specification targets and you start bumping up against fundamental limits for power needed to detect a spectroscopic signature of a threat, as well as selectivity and sensitivity for identification of molecules.

Augustus Fountain, Senior Research Scientist in Chemistry at Edgewood Chemical Biological Center, spoke to some of these issues. Fountain spoke about choosing the wavelength/spectroscopic for your method. In the UV you gain in sensitivity but loose in selectivity. The opposite is true as you move into the IR. Another problem to consider in system design is 1/r² loss and atmospheric attenuation. What kind of time window do you have available for scanning? Is the analyte a mixture of compounds- harder to detect spectroscopically, or something simple?  Scott Roberston echoed many of these remarks. Do you want to identify the threat or do you just want to know if it is going to kill you? The specific use and system dictate different constraints on what you design. Robertson also argued most users want the latter- "just give me a green or red indicator  light," not a beautiful Raman spectrum that requires interpretation. More often you just want to know "threat or no threat" for fast decision making in an environment of potential threats.

Much of the panel discussion centered around the do's and don'ts of collaborating with companies for defense money and contracts or even directly submitting proposals to broad agency announcements from DoD. If you are a small business or researcher trying to connect with defense contractors or apply directly for money the advice was to follow the rules, connect with partners and collaborators early, ask lots of questions early, and once again follow the rules.

The panel did offer some specific areas where there is need for technology. Fountain spoke how the 785 nm laser has been inappropriately the workhorse for Raman. This wavelength region has many problems. He would like sources further into the IR or deeper into the UV, particularly solid state sources. Edwin Dottery, President of Alakai Defense Systems, pleaded for a UV laser source less than 250 nm. Specifically between 220-240 nm will be ideal for UV Raman.

The difficult obstacles to overcome for practical stand-off detection are worth the effort. The end-user is particularly important and worth the time, soldiers continually putting their lives in harms way as well as civilians who want to carry on a normal life and provide for their families without fear of attacks. Lasers just may make this possible.

Interest in 2.0 micron Light is Growing

Wavelength Modulation Spectrum using tunable 2.0 micron VCSEL; From JTh1L.6, A. Kahn et al,. "Open-Path Green House Gas Sensor for UAV Applications"

Today at CLEO I spent a large amount of time at the expo hunting down which companies were selling 2.0 micron wavelength products and why. In the technical program, there are a large number of contributed talks regarding 2.0 micron lasers, pulsed and continuous wave. On Monday I attended session CM1B, Ultrafast Mid-IR in which 5 out of 8 papers demonstrated ultrafast pulses about 2.0 microns. Today there was a session titled CTu2D, 1.5 to 5 micron Lasers which also had 5 talks out of 8 showing laser systems operating near 2.0 microns. There have been and will be a handful of talks not pinned down to these topic categories as well:

-CM2B.2, "A Broadband 1850-nm 40-Gb/s Receiver Based on Four-Wave Mixing in Silicon Waveguides"

-CTu3M.7, "All-fiber 10-GHz Picosecond Pulse Generation at 1.9 microns without Mode-locking"

-JTh1L.6, "Open-Path Green House Gas Sensor for UAV Applications"

-CF1K.1, "Single-Frequency kHz-Linewidth 2-μm GaSb-Based Semiconductor Disk Lasers With Multiple-Watt Output Power"   

-CF1N.4, "Double-wall carbon nanotube Q-switched and Mode-locked Two-micron Fiber Lasers" 

However, what we like to research and what we can actually bring to market are often two very different things. I am therefore excited that it is not just 2.0 micron papers that are cropping up at this years conference, but 2.0 micron products at the expo as well.

So why is anyone interested in light in the 2.0 micron region? My personal interest stems from a research talk I saw by analytical chemist, Mark Arnold, at University of Iowa. Arnold is trying to perform some hard analytical  chemistry on 2.0 micron light shone through the skin on the back of one's hand. He hopes that by looking at the absorption spectra, he can measure blood glucose levels without having to draw blood. This noninvasive testing would be a boon to diabetics who are not thrilled about pricking their fingers regularly. Wavelengths that are helpful for pinning down glucose, but that are not absorbed as readily by tissue are 2.13 microns, 2.27 microns, and 2.33 microns.

In short, there are some interesting molecules around 2.0 microns on which to perform spectroscopy. For environmental sensing, there is 1877 nm, a well defined water absorption line, and 2004 nm, a good line for carbon dioxide detection, and many more.

Many of the companies I spoke with selling 2.0 micron components and sources confirmed such spectroscopic applications of their customers:

-Oz Optics now sells passive fiber components at 2.0 microns as well as DFB sources.

-Sacher Lasertchnik and Nanoplus make DFB lasers extending through the 2.0 micron region depending on your molecule of interest.

-Advalue Photonics makes thulium-based fiber laser systems and sells passive 2.0 micron products.

-New Focus will be developing tunable laser diodes about 2.0 microns in the next few months.

-Nufern and CorActive are selling Tm- and Ho-doped fiber for 2.0 micron amplification and for fiber sources.

-IPG sells a number of lasers from 2.0-2.8 microns based Cr:ZnSe as well as 2.0 micron fiber lasers using thulium doped fibers.

There are other advantages to 2.0 micron light as well. 2.2 microns is where the two-photon absorption coefficient in silicon drops to nothing. If you are interested in confining light to a silicon waveguide, doing so at 1550 nm could be the worst choice since it coincides with the peak two-photon absorption. However, above 2.2 microns allows higher throughput as well as access to other nonlinear effects like parametric amplification. 

This is one of the pursuits of Thorlab startup, PicoLuz. Among other optical instrumentation, PicoLuz is developing 2.0 micron amplifiers which will eventually support its endeavors of a chip scale optical parametric amplification. 

-Thorlab's Quantum electronics division is also selling a laser that is tunable about the gain bandwidth of thulium 1800-2000 nm, an FTIR spectrometer that goes out to 2500 nm,  a handful of moderate speed long-wavelength detectors, passive fiber components for 2.0 microns, as well as pumps for Tm-amplifiers.

Another reason for generating 2.0 micron light is for opening up new spectral bandwidth for telecommunications or signal processing, whether on fiber or on silicon. To that end,

-Eospace is now offering 20 Gb/s speed LiNbO3 intensity and phase modulators at 2.0 microns.

-Electro-Optics Technology is pushing past 1 GHz speed for 2.0 micron detectors.

Malcom Minty, a project manager from New Focus told me that New Focus was actively persuing thulium based laser systems in the 90's. The thought was that bandwidth would all be used up in the C- and L-band during the telecom boom, requiring expansion. Thulium has wide efficient gain and was a natural choice. Nufocus dumped the project as the the telecom bubble burst. Now they will be rejuvenating it, but more likely to sell to customers interested in spectrscopic pursuits.

Minty conjectured that 2.0 microns is becoming an interesting color to customers, vendors and researchers because it is in a region (or getting close to a region) of spectrscopic and biomedical interest. However, because it is close enough to the L-band, much telecom technology can still be used. It just squeaks by with some efficiency for detection on InGaAs-based detectors where as detection methods above 3.0 microns start getting tricky.

I think Minty likely has this right. If so I will be interested to see what new research we can carryout with 2.0 micron light while leveraging what we already know from fiber systems and telecom.