Chapter 4. Semiconductor Lasers: Physics and Applications

Chapter 4. Semiconductor Lasers: Physics and Applications


Academic and Research Staff

Professor Rajeev J. Ram, Dr. Jianyao Chen

Visiting Scientists and Research Affiliates

Dieter Graef

Graduate Students

Ravindra V Dalal, Steven G. Patterson, Farhan Rana

4.1 Introduction

We are developing novel semiconductor lasers toimprove the performance of communications sys-tems. Our approach is to build an understanding ofthese devices that is solidly rooted in electromag-netic modeling, quantum mechanics and condensedmatter physics. In the past, we have developeddevices that exploit physical phenomena such as (1)Bose-Einstein condensation and coherent populationtransients, (2) novel fabrication processes such aswafer fusion and x-ray lithography, and (3) noveldevice designs such as surface-emitting microcavi-ties and quantum cascade lasers to improve devices.

In the last year, several significant advances havebeen made. These include,

* the development of efficient numerical algorithmsfor energy band structure calculations

* the development of x-ray diffraction for character-izing the interfaces between optical thin films suchas distributed Bragg reflectors

* the prediction and demonstration of low distortionsignal transmission achieved by cancellation ofspatial-hole burning and gain compression nonlin-earities in DFB lasers, and

* the construction of an analog fiber link utilizingultracompact surface emitting laser sources.

These are among the results discussed in the follow-ing sections. Extensive discussion will be presentedin several forthcoming journal publications.

4.2 Band Structure and Optical Gain inStrained Layer Quantum Wells

Sponsor

MIT Lincoln Laboratory

Contract BX-6558

Project Staff

Farhan Rana, Professor Rajeev J. Ram

Design of high-performance semiconductor lasersrequires a careful and detailed consideration of thephysical processes involved. In the case of lasersoperating at high modulation frequencies and highoutput power levels, non-equilibrium effects associ-ated with carrier transport, carrier relaxation in quan-tum wells and carrier heating effects are important.Analytical models currently available to describelaser physics have little accuracy. Design of novelhigh-performance lasers require efficient and power-ful computer simulation tools. Unlike passivedevices, active optical devices have a rich variety ofphysics ranging from the microscopic to the macro-scopic level. In order to design semiconductor lasersone needs to know the band structure of strainedquantum wells, laser gain, radiative and Augerrecombination rates, carrier leakage rates, non-equi-librium carrier dynamics. Careful design of opticalwaveguiding structures and gratings (in the case ofDFB lasers) is also crucial. In addition, microwavepropagation characteristics of metal contacts are alsoof importance especially at high frequencies. Wehave developed a suite of modeling tools for theaccurate simulation of various laser structures:

1. Band Structure Solver: We have implementedan efficient finite difference technique to solveband structure of strained quantum wells using

an eight-band k - p approach. This schemealso gives us the k-dependent wavefunctions


for the conduction and the valence bandstates. An efficient algorithm is developed by anovel implementation of the axial approxima-tion which accurately treats the conduction andsplit-off valence bands. Figure 1 shows the cal-culated valence band structure in the quantumwell active region.

2. Carrier Dynamics and Gain: Using results from(1) above, this program calculates laser gain,radiative recombination rates, many bodyeffects related to carrier density dependentbandgap renormalization and index change. Asample gain spectrum is shown in figure 2.Non-equilibrium effects are simulated by solv-ing a system of coupled differential equationsthat describe carrier scattering from acousticphonons, optical phonons and with other carri-ers, and carrier recombination via stimulatedand spontaneous emission of photons and viaAuger processes. The solution of these equa-tions give the non-equilibrium hot carrier distri-bution in k-space. Our model currently treatscarrier-carrier scattering only phenomenologi-cally and cannot simulate carrier transport incladding layers. Future work is planned to rem-edy these shortcomings.

3. Auger Recombination Rates: The Auger solvercalculates Auger recombination rates in quan-tum wells using the actual band structure andwavefunctions found by (1) above and the non-equilibrium carrier distribution found in (2)above. This will be the first accurate estimateof Auger recombination rates; Auger is thedominant nonradiative recombination processin lasers used for long-distances communica-tions.

4. Inter-Valence Band Absorption Rates: Opticalloss in semiconductor lasers caused by inter-valence band absorption in p-doped layers canreach as high as 40 cm -1 for 101 cm -3 dopinglevel. Using results from (1) and (2) this pro-gram calculates these optical losses from firstprinciples.

5. Optical Mode Solver: We have implemented a2D finite element scheme to calculate opticalmodes in semiconductor laser waveguides.The mode solver can find modes for the scalarwave equation and also for the semi-vectorialwave-equation.

Models of heat flow and carrier transport will com-plete the tools in this development suite.

> -60

E-80- HH3

-e 100

LH1-120-

-140-

-160-

- 1 8 0 I ' ' ' L ' '0 1 2 3 4 5 6 7 8 9 10

transverse wavevector (m- ) x 108

Valence band structure for a 1000'A 1% compressively

Figure 1. Valence band structure for a 100 A 1%compressively strained InGaAsP quantum welloptimized for emission at 1.55 ptm

3000[

2500

2000

1500

N=1 1 cm- 3

1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7wavelength (mn) x 10

-6

Figure 2. GainFigure 1.

spectrum the same quantum well as

156 RLE Progress Report Number 140

4.3 Band Gap Engineering ofDistributed Bragg Reflectors

Sponsor


Project Staff

Steven G. Patterson, Professor Leslie A. Kolodzie-jski, Professor Rajeev J. Ram

Vertical cavity surface emitting lasers (VCSEL) pos-sess a number of properties which make them moreattractive than conventional edge emitting laserdiodes. They may be two-dimensionally packed atvery high densities. It is possible to integrate themonto chips with transistor logic, drivers, and detec-tors. Furthermore, while edge emitting lasers typi-cally have divergent, elliptical beams, VCSEL maybe designed to have minimally divergent, circularbeams, making them ideal for coupling into fibers.Finally, the end mirrors of the optical resonator do notneed to be cleaved thus eliminating a costly andunreliable manufacturing step.

The physical structure of the VCSEL is similar to a p-i-n diode where the p-type and n-type regions arecoincident with Distributed Bragg Reflectors (DBR).The DBR is a stack of alternating thin films of twomaterials with different indices of refraction, eachmaterial a quarter wavelength thick. For this work,pairs of AIo. 1Ga0.9As and AIo0 .9Gao.1As films form the

DBR stack. Abrupt heterojunctions between thesetwo materials result in large conduction barriers forholes and electrons (as large as 300 meV). Heavyuniform doping is an unacceptable solution as thisgreatly increases optical losses due to free carrierabsorption. Hence, recent efforts have been focusedon the bandgap engineering of the Alo. 1Ga0.9As/Al0. 9Gao.1As transition in order to reduce this barrier

to current transport.

One can consider linearly grading the aluminum con-centration at the junction as well as the dopant con-centration. The electric field resulting from theionized donors can be used to compensate for theheterojunction potential barrier. Figure 3 shows theconduction and valence band energies with a) onlythe aluminum concentration gradient and b) with boththe dopant and aluminum concentration appropri-ately graded. A well-engineered interface does notpresent any barrier to carrier transport while main-taining the desired optical index change.


Achieving the appropriate control of the aluminumconcentration during epitaxial growth of the DBRsrequires careful characterization of the growth pro-cesses. We are able to verify that the DBR mirror hasthe desired composition gradients by x-ray scatteringmeasurements. Figure 4 shows the measured andsimulated x-ray diffraction spectra from a 15-periodDBR. For the first time, this technique allows directverification of the DBR alloy composition. Tools arecurrently being developed to study the dopant profilein these laser structures.

-3

-3.5

ES-4 c

>, -4.5 Graded doping0)

----- Uniform doping

LU -5

E-5.5 v

-6-

0.06 0.08 0.1 0.12 0.14

Grid Pos (microns)

Figure 3. Calculated plots for: an Alo01Gao0 9As/AI0.9Gao01As heterojunction, with a 300 angstromgrading layer and modulation doping designed toflatten the valence band.

0.0001

o1000 -500 0 oo500 1000

Angle (arcsec)

Figure 4. X-ray data and simulation for a 15-periodnominally flatband DBR. The difference in location ofthe higher order peaks is due to a slight period driftduring growth.

157


4.4 Thermal Oxide for Ridge-WaveguideSemiconductor Lasers

Sponsor

MIT Lincoln LaboratoriesContract BX-6558

Project Staff

Dieter Graef, Steven G. Patterson, Professor RajeevJ. Ram

The basic idea of this work is to realize an indexguided GaAs-lnGaAs quantum well heterostructurelaser with a DFB structure to achieve single modelasing. The lateral index guiding and the periodicindex change in the propagation direction are real-ized by oxidation of large Al-content AIxGal-xAs.

In the last few years, the technique of wet oxidationof high composition AIxGalxAs has been developedmaking it posible to fabricate high performance opto-electronic devices providing current as well as opticalconfinement. The native oxide posseses excellentinsulating properties and a low refractive index(n~1.6). The formation of square-wall oxidation pro-files was demonstrated by P.W. Evans and N. Holon-yak. This technique is an exceptional method formaking integrated optoelectronic devices combiningthe high refractive index step realized by reactive ionetching (RIE) or wet etching and the planar geometryof ion implantation. Additionally it is possible to use asingle process step for both imprinting the DFB struc-ture and providing the light and current confinement.

Several groups have previously established that theoxidation rate is exponentially dependent on the alu-minum concentration (Figure 5). In the proposeddevice, thermal oxidation is initiated at the wafer sur-face and proceeds both horizontally and verticallyinto the semiconductor. A vertical oxide boundarycan be established by linearly increasing the alumi-num content with the distance into the semiconduc-tor. Figure 5 shows the lateral profile of the ridge-waveguide as a function of oxidation time.

Oxidation

0.05 .1E

0 O-5

e. 0.15 80 5, so 95 10W) Al-concentration [%]0

0.2

0.25 t1=19_4 min

t2=19 8mi 13=20.2min t4=20.5min0.3

stoppng layer0.35

0 0.02 0.04 0.06 0.08 0.1 0.12

Width [urn]

Figure 5. Lateral profile development with time forexponential rate approximation. Inset: Oxidation ratefor various aluminum content.

Figure 6 uses this technique to construct an indexgrating along the length of the laser cavity. First a 0.5p.m GaAs (n+) buffer layer covers a (100) GaAs (n+)substrate. Then a 1.5 pm Al0o.95Gao.05As(n) lowerconfining layer is grown followed by a 0.16 umundoped GaAs waveguide with two 80 AIn0.2Gao.8As QW in the center separated by a 120 AGaAs wall. Next an 0.3 pm upper confining layer ofAIxGal-xAs(p) with a graded composition 0.88 < x <0.95) is grown and a 20 nm low gap AIxGal-xAs(n+)layer (stop layer for upper oxidation). The secondoxidation layer(p+) is grown on the top (0.3 pm withgraded composition: 0.84 < x < 0.99).


2Tm_,

0.23 pm 0.23 pm

20 nm 0.3 pm

0.3 pm

300 pme

Figure 6. Basic laser structure: light and currentconfining oxide walls in the in the cladding layer andthe DFB structure

While the buffer layer guarantees a smooth surface(on atomic scale) for the following layers the lowercladding layer is made thick enough to neglect anyinfluence of the buffer layer on the waveguiding. TheAl content of this layer is 93% to assure index sym-metry in the transverse direction. The active region isformed by two InGaAs QW embedded in GaAs. Thestrength of the waveguide was chosen well below cutoff of the second order mode to enable only the fun-damental transverse mode. The upper cladding layerwill be oxidized to give lateral confinement, andtherefore its grading and its thickness are crucial forlaser fabrication. The GaAs guiding layer acts asstopping layer for this oxidation. In order to use onlyone oxidation step the two graded layers can not bedesigned independently. They have to be grown soas to get square walls of different aspect ratios duringthe same oxidation step. The goal is to get a squarewall periodic index change with a given duty cycle inthe top layer, while making a compact confining layerin the lateral direction by oxidizing the lower gradedAIGaAs layer at a higher rate. The calculated oxida-tion time is ~0.3 hours.

The laser fabrication itself begins with patterning 2.4pm wide lines on the sample using standard lithogra-phy techniques. These stripes serve as a mask for


the fast etch of the two uppermost layers, exposingthe lower AIGaAs for subsequent oxidation. The finalridge width will be 2 gm where the additional 0.4 gmaccount for the lateral oxidation of the lower layer.Now the protecting layer is removed and a differentline pattern (0.04 lim) is deposited perpendicular tothe ridge. The width of this fine line pattern is deter-mined by the duty cycle of the DFB structure in thesecond order mode and the lateral oxidation rate.The sample is then immediatly placed in the wet oxi-dation furnace (4250C; supplied with N2 bubbled

through H20 at ~950C). When the sample is intro-

duced in the furnace the exposed AIGaAs (both lay-ers start oxidizing vertically and horizontally with arate depending on the Al-content (Figure 5). The con-centration difference in the ridge will provide asquare wall profile (Figure 5 and Figure 6) equal tothe second order Bragg condition. At the same timethe cladding layer will oxidize with a higher velocitydue to its higher Al concentration resulting in twicethe aspect ratio of the upper layer. Thus the materialunderneath the lines is also oxidized resulting in acompact insulating and light confining wall in lasingdirection. Because the oxidation is isotropic in the lat-eral direction we get oxide (square walls) underneatha part of the ridge (Figure 6) To take account of thiseffect the initial ridge width was choosen larger thanused for calculations.

Planar anisotropic oxidation of graded AIxGalxAs

can be used for fabricating high performance laserdiodes. Once the exact Al-gradings for the two layersin question are determined, the laser structure canbe formed by a single process step resulting in pla-nar structures with high index contrast. MBE makes itpossible to grow well determined grading functions.This possibility together with the strong oxidation ratedependence on the Al-content gives a very powerfultool for engineering the lateral index profile whilekeeping the surface planar. In addition to squarewalls, it is possible to produce closed loops with dif-ferent curvatures, e.g., circles resulting in cylindricalwires.


4.5 High-Speed Semiconductor LaserDevelopment

Sponsor

MIT Lincoln LaboratoryContract BX-6558

Project Staff

Farhan Rana, Michael H. Lim, Elisabeth Marley, Pro-fessor Rajeev J. Ram, Professor Henry I. Smith, Pro-fessor Leslie A. Kolodziejski

Lasers capable of being modulated at high frequen-cies are a key component for high speed opticalcommunications. Semiconductor lasers can be inten-sity or frequency modulated up to very high frequen-cies by directly varying the drive current. Directmodulation schemes are much simpler to implementthan those involving external modulators. The goal ofthis research is to develop, design, and fabricate veryhigh speed semiconductor lasers operating at wave-lengths suitable for fiber optic communication sys-tems.

Design requirements for high speed lasers demandthat careful attention be paid to both the microscopicand the macroscopic physics of these devices. Themicroscopic physics includes proper design of semi-conductor heterostructures. Important laser parame-ters such as gain, differential gain, radiativerecombination rate, Auger recombination rate, fre-quency chirp, carrier leakage, and gain compressionfactor can be significantly improved by optimizing thedesign of semiconductor quantum wells, barriers,and cladding regions. At high modulation frequenciesnon-equilibrium effects such as carrier transport andrelaxation in heterostructures and carrier heatingalso become increasingly important. At the macro-scopic level, attention must be paid to the properdesign of optical waveguides. For DFB (and DBR)lasers, designing optical grating structures that mini-mize spatial hole burning and reduce chirp, distortionand noise pose a significant challenge. The electricalcharacteristics of a laser including its series resis-tance and capacitance can also become a potentialbottleneck for high speed operation. In addition, athigh frequencies the distributed nature of semicon-ductor laser structures cannot be ignored. Laserdevices must also be properly impedance matchedwith the electrical driving circuitry. Our research pro-gram includes looking at all the above issues fordesigning better and faster lasers.

We have developed techniques for fabricating poly-imide-planarized ridge-waveguide laser structuresthat are ideally suited for high frequency operationbecause of their very low capacitance and low leak-age currents. The low capacitance is a result of thelarge separation between the top metal contact andthe substrate (see Figure 7). In addition, this tech-nique allows us to make ohmic contacts to ridgeswithout having to open contact vias lithographically.Thus, very narrow ridge structures can be realized.Crystal growth for these laser structures is also doneat MIT The grown material shows good electrical andoptical characteristics.

Figure 7. Scanning electron micrograph of an InPridge waveguide with a polyimide planarization layer.

Techniques for fabricating optical gratings for DFB(and DBR) lasers using combinations of interferomet-ric, X-ray, and spatially phase-locked e-beam lithog-raphy have also been recently developed at theNanostructures Laboratory at MIT These techniquesallow us to make optical gratings with spatially vary-ing feedback characteristics with long range spatialcoherence which has never been done before. Weplan to explore a variety of grating based semicon-ductor laser structures for high speed as well as forlow noise operation. In addition, the high speed laserprocesses developed will also enable us to realizehigh-performance mode-locked semiconductorlasers.


4.6 Nonlinear Response of DFB Lasers

Sponsor


Project Staff

Dr. Jianyao Chen. Professor Rajeev J Ram

With the outstanding features of dynamic single fre-quency operation and wide modulation bandwidth,DFB lasers are the primary candidates for variousanalog optical fiber communication systems. Forsuch applications, the laser light intensity is requiredto be a linear function of the drive current. The char-acteristics of nonlinearly generated harmonic andintermodulation distortions become the quantities ofconcern. Due to the nature of distributed feedback,various spatial features of DFB lasers, such as theasymmetric facet conditions, abrupt phase shift andnonuniform carrier injection, have strong influence onthe uniformity of photon distribution inside the lasercavity. This causes the carriers in the active region torecombine at different rates at different locations.Therefore, a spatially resolved analysis of nonlinearresponses in DFB lasers becomes significant.

To include the various spatial features of DFB lasersaccurately and to retain the advantage of closed formanalysis, the spatially dependent rate equationsbased on the Greens function method are developedto study the harmonic and intermodulation distortionsin DFB lasers. The self-consistent analysis showsthat spatial hole burning, nonlinear gain compressionand relaxation oscillation are three major sources ofsignal distortion and that their effects can interactwith each other both constructively and destructively.

Detailed frequency analysis reveals that the nonlin-ear distortion at low frequency depends on the unifor-mity of the optical field distribution significantly. Theeffect, however, is limited by the carrier damping ratein DFB laser. The induced distortion rolls off at rangeof a few hundred megahertz. Numerical results showthat, as the coupling strength of DFB grating kL devi-ates away from the optimum value, the harmonic andintermodulation distortions can increase rapidly in thelaser. On the other hand the relaxation oscillationeffect is dominated in the intermodulation distortionat high frequency. which is independent of kL. There-fore, choosing a proper grating with the optimumcoupling coefficient and working away from the reso-nance frequency are important


Facet reflection determines the optical field distribu-tion in DFB laser significantly. As the grating phase atfacet is cleaved randomly, the selected lasing modecan appear on either short or long wavelength sidesof the Bragg stopband. Our analysis indicates that,although their optical field distributions can be simi-lar, the induced distortions give different phases. Onthe other hand. gain compression and relaxationoscillation always give 0 and 7 phases in third orderintermodulation distortion. Therefore, as we choosedifferent grating structure and facet conditions, it ispossible to make spatial hole burning effect to cancelwith gain compression and relaxation oscillationeffects to get very low total distortion at low and highfrequencies, respectively. Such properties are veryuseful to the applications in high fidelity communica-tion systems.

0.32

0.31

0.3

0.29

0.28

0.27

0.26 L10 15 20 25 30 35 40 45 50

Current [mA]

Figure 8. Slope efficiency of HR/AR coated DFBlaser with grating phase -= 0 at HR facet, whereinset gives the structure of ridge waveguidedistributed feedback laser.


4

2- 0--

-2--

1- -4

0 SN=4X0-17 = o cm-

Si - =3x10"

cm

c -30 -=2x10 cm

o 3... =0 m'

.2 -60 !S-90

S-120-1 0 0 102

Modulation Frequency [GHz]

Figure 9. Frequency response of third orderintermodulation distortion in of HR/AR coated DFBlaser with grating phase 0-0 at HR facet, where1=Il+g35 imA. OMD=30%, fi-f 2 =1 MHz.

4.7 High-Fidelity, High-Dynamic RangeFiber Communications

Sponsor


Project Staff

Ravindra V. Dalal, Dr. Jianyao Chen,2Helkey,' Harold V. Roussell, Professor

Ram

Dr. RogerRajeev J.

Cable television providers and telecommunicationscompanies are competing to create broad-band sub-scriber networks that are upgradeable, flexible, andable to carry large bandwidths of various types ofdata. The wide bandwidths of semiconductor lasersand the accompanying fiber optic components arecost-competitive and compatible with existing coaxialcable systems, thus making them attractive to cableproviders as well as telecommunications companies.Through subcarrier multiplexing (SCM) a single net-work can handle multiple data formats, includingvoice, data, video, digital audio, high definition video,and any combination of these services. Subcarriermultiplexing has also stimulated great interest in nar-

row band analog optical applications, which includephase array radar and personal communications sys-tems (PCS). As their name implies, these applica-tions require a relatively narrow bandwidth ofchannels.

The performance of laser diodes for optical commu-nications is limited by application bandwidth, distor-tion, and noise. Distortion can be caused by manyfactors. Static distortion occurs when the nonlineari-ties of the light-versus current (LI) curve are present.These nonlinearities are caused by spatial hole burn-ing, gain compression, finite carrier transport times,and leakage currents. Dynamic distortion occurswhen the nonlinearities of the device cause differentfrequency components to mix together through theinteraction of photons and electrons in the laser cav-ity. As a result of these nonlinearities in the device, itis possible for frequencies from separate channels tointerfere and mix with each other to produce new fre-quency components.

The application bandwidth determines the type ofdistortion that will affect the system. A narrow bandsystem will be primarily affected by third order inter-modulation distortion (IMD3). The key figure of meritfor such narrow band applications is the spuriousfree dynamic range (SFDR), which is the range ofinputs over which the output signal is unaffected byeither noise or distortion.

Intermodulation distortion and dynamic range mea-surements were performed for a 1.3 ptm packagedFujitsu DFB laser (model FLD130F3ACH-AL), whichis made for CATV signal distribution. The laser isdesigned to minimize second order intermodulationdistortion. A standard two-tone electrical modulation(Figure 10) is used to study the nonlinear responseof the laser.

1 MIT Lincoln Laboratories, Lexington, Massachusetts.

2 Center for Compound Semiconductor Technology, Sandia National Laboratory, Albuquerque, New Mexico.



PackagedFujitsuDFB laser

Detector

DC

Synthesizer I I I If2=cf + Af

Figure 10. Experimental set-up for two-tone modulation test of an optical communications link.

We found that the SFDR peaks because of a mini-mum in the third order intermodulation distortionrather than because of a minimum in the noise, ascan be seen in Figure 11. We believe that the distor-tion minimum occurs due to a cancellation in spatialhole burning and gain compression, two nonlineareffects which are dependent on the photon density inopposite ways for this particular structure for particu-lar operating conditions. This suggests that animproved laser design may be realized by adjustingthe spatial hole burning and gain compression suchthat the distortion minimum coincides with the mini-mum in the noise, thus leading to a larger SFDR. Ithas been shown that improving the SFDR by as littleas 4 dB Hz2 /3 can double the maximum number ofchannels that a cellular telephone system can handlewhile ensuring a certain call blocking probability. Amaximum SFDR of 125.5 dB Hz2/3 was measured forthe Fujitsu device, which is significantly greater thanthe dynamic range required for PCS (72-83 dB Hz21/3).

-20

-40

-60

C.

-80 3

100S dB Attenuation -100100/

95 -12015 20 25 30 35 40 45

Current, mA

Figure 11. Spurious intermodultion free dynamicrange and noise figure versus bisa current for asystem employing a high-linearity DFB laser.


4.8 Device Level Modeling ofCommunication Systems

Sponsor


Contract BX-6558

Project Staff

Dr. Jianyao Chen, Ravindra V. Dalal,Rajeev, J. Ram, Dr. Roger Helkey

Professor

Due to the nature of distributed feedback, various

spatial features in DFB lasers have great influence

on device characteristics and hence determines the

performance for the fiber optic system significantly.

To achieve the optimum design for adequate signal

quality, a detailed investigation of the influence of

various structures of DFB lasers to the system per-

formance of analog fiber optic links is therefore

important. We present the first such simulation and

reveal that cancellations between different nonlinear-

ities are significant enough to suppress the signal

distortions in the system.

In general, link gain, noise figure and intermodulation

distortion free dynamic range are the most important

quantities to concern for the high fidelity, narrow

band SCM fiber optic systems. Since nonuniform

photon distribution in DFB lasers gives a complex

influence to the laser power output, noise perfor-

mance and linearity of laser response under direct

modulation, we extensively study the dependence of

those parameters on various structure details of DFB

grating, such as complex grating, facet coating,abrupt phase shift and etc. The revealed knowledge

is very helpful to the optimization of system perfor-

mance.

A strong grating feedback confines the optical field

inside the DFB laser and hence reduces the laser's

external quantum efficiency, which directly leads to

decrease link gain for signal transmission in the sys-

tem. On the other hand, weaker grating feedback

makes the laser have a higher threshold which

increases the intensity noise level in the laser output

and hence lowers the signal-noise-ratio of the sys-

tem. Theoretical study helps understand that the can-

cellation between spatial hole burning and gain

compression effects is very useful to achieve a very

high IM free dynamic range in DFB lasers.

4.9 Analog Signal Transmission UsingSurface Emitting Lasers

Sponsor

U.S. Navy - Office of Naval Research/MURI

Project Staff

Ravindra V. Dalal, Dr. Roger Helkey, Harold V. Rous-

sell, Dr. Kent D. Choquette, Professor Rajeev J. Ram

Vertical cavity surface emitting lasers (VCSEL) have

potential to be low cost sources for optical communi-

cations. For digital applications, transmission rates of

up to 10 Gb/s for short wavelength lasers have been

demonstrated, and 2.5 Gb/s at long wavelength

lasers. However, the performance of VCSEL under

analog modulation has not been investigated. These

applications require a large carrier to noise ratio, high

bandwidths, low intermodulation distortion, and wide

dynamic range. For the first time, analog signal trans-

mission using a short wavelength VCSEL is studied.

We demonstrate a spurious free dynamic range suit-

able for personal communications systems.

The measured lasers were 20 pm x 20 pm oxide-

aperture top-emitting devices fabricated at Sandia

National Laboratories (Figure 12). The parasitic

capacitance and resistance of the devices were mea-

sured to be 50 fF and 30Q, respectively. The nonlin-

ear response of the VCSEL was measured using

conventional two-tone modulation.

Figure 12. Oxide-apertured surface emitting laserused for signal transmission experiments.


As expected, the nonlinear response shows a powerlaw dependence on the input power with measuredexponents of 1.214, 2.201, and 3.187. The noisefloor corresponds to a channel bandwidth of 1 MHz.The spurious free dynamic range (SFDR) is the ratioof the fundamental to the intermodulation product atthe point where the intermodulation product crossesthe noise floor. Because of the relatively low outputpower for the VCSEL (350 pW), the noisefloor for themeasured link is limited by receiver noise rather thanby RIN from the laser. Figure 13 shows the depen-dence of the optical power and the dynamic rangecurve on bias current. The detuning of the gain curvefrom the cavity resonance due to lattice heatingresults in a roll-off of the optical power from theVCSEL at high current bias. This roll-off leads to astrong nonlinear response that is still observable as aplateau in the low frequency dynamic range versuscurrent bias.

-O--20


improved dynamic range for VCSEL. At higher fre-quencies, performance will not be influenced by ther-mal nonlinearities, but instead will be limited by therelatively high relaxation oscillation frequencies. Inour future work, we will present the dependence ofdynamic range on oxide aperture diameter illustratingthe effects of spatial hole burning.

75

350

250

150

50 ,0 10

7o3070 -

O60

65 0

T

60

Laser current, mA

Figure 13. Spurious intermodulation free dynamicrange and noise figure versus bias current for asystem employing a high-speed surface emittinglaser.

The maximum dynamic range was measured at 85dB Hz2 3 between 100 MHz and 1 GHz using a higherbandwidth detector. Although this is lower than themaximum achievable dynamic range for DFB lasers(SFDR=125 dB Hz213), these results indicate thatVCSELs satisfy the dynamic range requirements forpersonal communication systems (PCS), which arein the range of 72-83 dB Hz2/3 . Increasing the opticalpower coupled into the detector and operating athigher modulation frequencies may result in

Chapter 4. Semiconductor Lasers: Physics and Applications

Documents

Transcript of Chapter 4. Semiconductor Lasers: Physics and Applications