The big news from Tokushima is that Shuji Nakamura announced the new Nichia laser results. They have demonstrated a 10,000 hour lifetime (Actually it was 1000 hrs at 50°C, but this extrapolates to 10,000 hours at room temperature in "standard" reliability models. 10,000 real hours is 1 year, 51 days. See the clarification from Shuji Nakamura) for a blue laser made with the new lateral-epitaxy overgrowth process. The other innovation he reported was the use of modulation-doped strained layer superlattices in the cladding to reduce cracking. Reports in the Japanese press have indicated that Nichia intends to commercialize this laser by the end of 1998.
The lateral overgrowth was reported by Dr. Nakamura at the ISCS in SanDiego, but to me the interesting thing here is that 10 microns of GaN is being overgrown. The lasers are then made in what amounts to small, elongated crystallites of GaN. A strain relief mechanism is inherent at the joins of the structure. The modulation-doped SLS is another intriguing thing. The alternating compression and tension in a strained layer superlattice is well known to result in physical toughening, but I can't think of a good reason to use modulation doping in the cladding layer. The doping in the Nakamura's laser is not conventional modulation doping, where there is transfer of carriers into the low bandgap region, but rather it is doping of the wells only. Is there something more there that we should be thinking about? Perhaps it's just something like an interaction of the doping with surface morphology.
Further discussion of the modulation doping is below.
Also at Tokushima, Xerox announced their laser. The Xerox announcement got the American press rather excited, although it tells you something that there was no comparable national press interest when Cree announced their laser. The use of a SiC substrate in that work was an important technical contribution.
The other newcomer to the nitride laser party was Sony. Their paper was presented by Fumihiko Nakamura (No relation to Nichia's Nakamura.) Sony is not a newcomer to the blue laser business, as their II-VI laser program has been among the world's most successful.
MIJ-NSR readers who were at the Tokushima Meeting are encouraged to send a note with their impressions of what was exciting to journalmaster@nsr.mij.mrs.org. Your colleagues will really appreciate your efforts!
The Tokushima region is famous for production of indigo dye. (The color indigo is between blue and violet, made famous by the reknowned but mythical spectroscopist Roy G. Biv.) Perhaps we should be calling the new devices "indigo" lasers!
Eric HellmanBack to News Page Back to Top
The meeting was kicked off by plenary talks by Hisashi Yamada on DVD optical storage, its applications in both computers and home entertainment, and the role blue lasers (either nitride-based or second harmonic structures) might be expected to play and by Bo Monemar on the basic physics of nitrides. Prof. Monemar emphasized the need for better material to obtain consistent values for parameters such as donor binding energies. He drew comparisons with SiC where the best material now has a residual doping level in the 1014 range compared to ~1016 for GaN to offer both optimism and a challenge to the growers. This general theme was echoed by a number of other spectroscopists and theorists throughout the week.
The highlight for most participants was a session on Thursday afternoon on laser diodes. Shuji Nakamura gave a keynote address in which he discussed their most recent laser structures. These are grown on laterally overgrown layers of GaN formed by first growing ~2 µm of GaN then masking off stripes with SiO2 and using stripes of GaN as "seeds" for the growth of ~10 µm layers of GaN. The regions over the SiO2 have greatly reduced defect densities and are used for the lasers. The LDs also feature modulation doped strained layer superlattices as cladding layers which eliminated cracking and lowered the operating voltage to about 5 V. These improvements resulted in 3000 hr. lifetimes at room temperature (estimated to be as long as 10,000 hr. by accelerated testing). John Edmond of Cree Research presented and demonstrated their most recent Nitride on SiC laser, the best of which have operating voltages of about 22 V. He indicated that they were also working on lateral overgrowth and that reducing the operating voltage was achievable. Akito Kuramata of Fujitsu also presented their most recent nitride on SiC laser results and M.P Mack (UCSB), Fumihiko Nakamura (Sony) and M. Kneissl (Xerox) presented results for their sapphire-based LDs. Hiramitsu Sakai (Meijo) presented results of optically pumped SE from MQWs on Sapphire.
The topic of lateral overgrowth was naturally a hot topic with a number of reports on growth techniques and characterization of such layers. One concern with such a technique is the length of time needed to form the thick overgrowth layer by MOCVD. C. Sasaoka (NEC) reported on using HVPE to grow this layer at a much faster rate (as high as 100 µm/hr.) which gave layers with defect densities of 6X107 cm-2. Approximately ten other papers addressed various aspects of lateral overgrowth.
In addition to the opto-electronic devices and physics, electronic applications were also discussed. Steve Binari (NRL) presented an overview of the accomplishments of a number of groups in this area as well as some of the problems which still must be resolved. There was also a rump session on electronic applications (as well as ones on cubic GaN and phase separation) in which GaN-based devices were compared to GaAs-based microwave devices and SiC-based high power/temperature devices.
Although the conference was clearly dominated by light emitting devices, there was nothing on detectors which has certainly been considered an important area of potential applications.
Finally, the conference was not entirely business, Nichia hosted an excellent reception and at the banquet we learned that Shuji Nakamura (and several others) should not consider dancing as a second career.
W.E. Carlos
Back to News Page Back to Top
Translated by E. Hellman.
This is a newly established biennial conference. It reconvened in Tokoshima with close to 500 participants of whom 400 were from Japan.
There were 250 presentations, 150 of them from Japanese labs.
7 laboratories have successfully made blue lasers.
CREE on SiC (they also mentioned ELOG), Fujitsu, Toshiba, UCSB, Xerox ... and Sony ! (All the laser structures were made by MOVPE).
All sorts of hypotheses were discussed: composition fluctuations, phase decomposition (3 papers on calculations of the thermodynamic approach to spinodal decomposition of GaN), formation of quantum boxes. An object of controversy is the mechanism responsible for laser emission in the MQW. Emission from quantum boxes, localization of excitons by composition fluctuations or excitonic emission (would also explain the "Stokes Shift" (biexciton) and also e-h !). Nothing seems definitive despite the number of theoretical approaches.
Pierre GIBART Director of Research Centre de Recherche sur l'Hétéroépitaxie et ses Applications (CRHEA) CNRS Sophia-Antipolis F-06560 Valbonne
Back to News Page Back to Top
At 0:17 PM 97.11.6 -0500, Eric S. Hellman wrote: >There's been some confusion about the exact lifetime you've reported for >your new laser. > >I've heard 1000 hours at 50C, 3000 hours and 10,000 hours extrapolated.Dear Eric
The lifetimes of our best LDs are 700 hour at 60°C, 1000 hours at 50°C, 3000 hours at 40°C, 10000 hours at 20°C. The activation energy is about 0.5 eV.
The 10,000 hours is extrapolated.
Shuji----------------------------- Shuji Nakamura R & D Department Nichia Chemical Industries Ltd. 491, Oka, Kaminaka, Anan Tokushima 774, Japan ------------------------------
Back to News Page Back to Top
Modulation doping can push the free (minority) carrier density higher in the laser active region thereby reducing the injection current required to reach transparency and subsequent inversion.
This effect was posited, and proven effective in reducing threshold currents in GaAs/AlGaAs lasers, and published by Gary Wicks of Rochester University, Institute of Optics. I recommend you talk with him about the subtleties.
I did not go to Japan, so am not exactly sure of the structure, and details of the modulation doping . (thickness, offset, sheet-density, or even donors, acceptors, or both).
Inevitably there is a modulation doping effect at most heterostructures, and most good lasers unwittingly take advantage of it.
Modulation doping can also allow deep(ish) centers to empty into wells. Specifically, if the acceptor / donor energy lies within or close to the valence / conduction band discontinuity, then the effective binding energy is zero or close to zero. Carriers can then transfer with ease, and avoid relying on (insufficient) ambient kT to overcome the high binding energies.
Colin Wood Electronics, Code 312 Office of Naval Research 800 N Quincy St. Arlington VA 22217
Back to News Page Back to Top
Here's my 2 cents on the latest Nakamura laser result.
First of all, the lifetime is 1000 hrs. at 50 C, not 10000 hrs. This result extrapolates to 10000 at room temperature. Nakamura's results are good - but not that good.
The modulation doping was not the type that I published in AlGaAs/GaAs lasers, or at least not only that type. When Nakamura's uses the term "modulation doping", he doesn't mean it in the normal sense where there is charge transfer from a doped high bandgap layer into a low bandgap layer, thereby forming a 2DEG. He constructs the cladding layers in the form of a superlattice that consists of alternating layers of undoped AlGaN and doped GaN - and calls this modulation doping. Apparently this improves the carrier transport through the clad. The AlGaN layers are very thin, so it looks like the carriers can tunnel through the AlGaN layers from one GaN layer to the next GaN. Of course, if he could get high enough carrier concentrations in the AlGaN by direct doping, this "modulation doping" would be unnecessary.
The type of modulation doping that we published in AlGaAs-based lasers was doping the barriers near the active quantum well to cause carriers to transfer into the well. Among other things, this improves the threshold. Interestingly, Nakamura does this also -- sort of -- but doesn't make a big deal out of it. Actually what he does is dope both the barriers and the active quantum wells, so he has carriers in the wells by directly doping the wells in addition to carrier transfer from the doped barriers.
Finally, most references on color define the wavelength region from 400 to 450 nm as "violet" - indigo isn't used (sorry Mr. Biv). This latest work by Nakamura is 401 nm, so they should be called violet lasers -- but they're only a couple of nanometers from being UV.
Gary W. Wicks, Professor The Institute of Optics, Univ. of Rochester, Rochester NY 14627 wicks@optics.rochester.edu
Back to News Page Back to Top
We developed InGaN MQW laser diodes grown on SiC substrates using a low-pressure metalorganic vapor phase epitaxy. Changing the MQW structure from five 2.5-nm thick wells to three 4-nm thick wells improved both the threshold current and the slope efficiency. HR facet coating also effectively reduced the threshold current. The following is the latest characteristics of our LDs.
Room temperature pulsed operation (maximum duty ratio : 1%) threshold current : 500 mA (12 kA/cm2) threshold voltage : 22 V maximum light output : 80 mW wavelength : 405 - 425 nm Akito Kuramata Fujitsu Laboratories Ltd. 10-1 Morinosato-Wakamiya Atsugi 243-01, Japan e-mail : akurama@flab.fujitsu.co.jp
Back to News Page Back to Top
On October 14, 1997, researchers at the Xerox Palo Alto Research Center demonstrated a blue laser diode in the InGaAlN materials system. The diodes lased at wavelengths from 422 to 432 nm under pulsed current-injection conditions (pulse duration 400ns, repetition frequency 1kHz) from a 10 quantum well heterostructure that was grown on a sapphire substrate by MOCVD. The mirrors were fabricated by chemically assisted ion beam etching. Both spectral line narrowing and a far-field interference pattern were observed. The threshold current density was 25kA/cm2, and the beam was TE polarized. Lasing was demonstrated in a series of test diodes with metal stripes from 4 to 20 microns wide and cavity lengths ranging from 300 to 1000 microns. The output power was greater than 10 mW.
The work at Xerox PARC was partially supported by the DOC Advanced Technology Program (agreement # 70NANB2H1241 for growth and device fabrication) and by DARPA (agreement # MDA972-96-3-0014 for materials and device characterization and for CAIBE process development) through the Blue BAND II Consortium. The DOC ATP teamed Xerox Corporation and SDL Inc. to develop multi-wavelength laser arrays at visible wavelengths. The Blue BAND II Consortium teams the Xerox Corporation, SDL Inc., and the Hewlett Packard Company along with five universities (Boston U., MIT, U. of New Mexico, U. of Texas at Austin, and U. of Utah). The program targets the development of commercially-viable single-mode, ridge-waveguide laser diodes and demonstration of vertical-cavity surface emitting laser (VCSEL)diodes. The work ranges from basic theoretical and experimental growth and material studies to demonstrations of advanced devices.
Michael Kneissl Electronic Materials Laboratory Xerox Palo Alto Research Center 3333 Coyote Hill Road Palo Alto, CA 94304 E-mail: kneissl@parc.xerox.com
Michael P. Mack
Graduate Student Researcher
ECE, Box #233
University of California
Santa Barbara, CA 93106
Note: The UCSB laser has been described in a MIJ-NSR paper.
Back to News Page Back to Top
We presented a paper at Tokushima, entitled "Room-temperature Pulsed Operation of a GaInN Multiple-Quantum-Well Laser Diode with Optimized Well Number"
The well-number dependence of the optical pumping threshold power for stimulated emission of GaInN multiple quantum-well (MQW) laser structures was investigated. The pumping threshold power for a three GaInN MQW sample was found to be as low as 33kW/cm2 at room temperature. The room-temperature pulsed operation of a five GaInN MQW laser diode (LD), whose number of wells was determined based on the optical pumping experiment, was also demonstrated. A 1.8mm-long cavity was formed by cleaving along the (11-20) plane of the GaN epitaxial film. No facet coating was employed. Lasing operation under pulsed current injection of 500ns and 1kHz at room temperature was achieved. The lowest threshold current density was 9.5kA/cm2. The lasing wavelength was 417.5nm with a full width at half maximum (FWHM) less than the spectrum resolution of 0.2nm.
Fumihiko Nakamura ------------------------------------- Laboratories for Materials and Devices Sony Corporation Research Center 174, Fujitsuka-cho, Hodogaya-ku, Yokohama-shi, 240 Japan E-mail: fnakamra@src.sony.co.jp -------------------------------------
Back to News Page Back to Top
I read somewhere that Newton, who introduced the ROYGBIV nomenclature for prismatic colours, had an uncommon congenital sensitivity to the different shades of blue light. IMHO, anything is better than "purple", but never having had the chance to examine one of these beasts.......hey, I guess that's all changed now!
Kevin Peter O'DonnellBack to News Page Back to Top
| blue | indigo | violet |
Back to News Page Back to Top
last updated November 17, 1997 11:07:14 PM.
© 1997 The Materials Research Society
