highly strained mid-infrared type-i diode lasers on gasb

IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 21, NO. 6, NOVEMBER/DECEMBER 2015 1502410

Highly Strained Mid-Infrared Type-IDiode Lasers on GaSb

Scott D. Sifferman, Student Member, IEEE, Hari P. Nair, Rodolfo Salas, Nathanial T. Sheehan,Scott J. Maddox, Student Member, IEEE, Adam M. Crook, and Seth R. Bank, Senior Member, IEEE

(Invited Paper)

Abstract—We describe how growth at low temperatures can en-able increased active layer strain in GaSb-based type-I quantum-well diode lasers, with emphasis on extending the emissionwavelength. Critical thickness and roughening limitations typi-cally restrict the number of quantum wells that can be grown at agiven wavelength, limiting device performance through gain satu-ration and related parasitic processes. Using growth at a reducedsubstrate temperature of 350 °C, compressive strains of up to 2.8%have been incorporated into GaInAsSb quantum wells with GaSbbarriers; these structures exhibited peak room-temperature pho-toluminescence out to 3.96 μm. Using this growth method, low-threshold ridge waveguide lasers operating at 20 °C and emittingat 3.4 μm in pulsed mode were demonstrated using 2.45% com-pressively strained GaInAsSb/GaSb quantum wells. These devicesexhibited a characteristic temperature of threshold current of 50 K,one of the highest values reported for type-I quantum-well laserdiodes operating in this wavelength range. This temperature stabil-ity is attributable to the increased valence band offset afforded bythe high strain values, due to the simultaneously high quantum wellindium and antimony mole fractions. Exploratory experiments us-ing bismuth both as a surfactant during quantum well growth,as well as in dilute amounts incorporated into the crystal werealso studied. Both methods appear to be promising avenues to sur-mount current strain-related limitations to laser performance andemission wavelength.

Index Terms—GaInAsSb, GaSb, mid-IR, semiconductor lasers.

I. INTRODUCTION

THE mid-infrared (2 to 10 μm) spectral region has greatpotential for applications ranging from science and tech-

Manuscript received February 28, 2015; revised April 13, 2015; acceptedApril 24, 2015. This work was supported by the Army Research Office(W911NF-11-1-0452, W911NF-11-1-0521) and the National Science Foun-dation (ECCS-0954732) under a CAREER award. This work was performed atthe Microelectronics Research Center at the University of Texas at Austin, amember of the National Nanotechnology Infrastructure Network, supported bythe National Science Foundation.

S. D. Sifferman, R. Salas, N. T. Sheehan, S. J. Maddox, and S. R. Bankare with the Microelectronics Research Center, The University of Texas atAustin, Austin, TX 78758 USA. (e-mail: [email protected];[email protected]; [email protected]; [email protected];[email protected]).

H. P. Nair was with the Microelectronics Research Center, The Universityof Texas at Austin, Austin, TX 78758 USA. He is currently with the MaterialsScience and Engineering Department, Cornell University, Ithaca, NY 14853USA. (e-mail: [email protected]).

A. M. Crook was with the Microelectronics Research Center, The Universityof Texas at Austin, Austin, TX 78758 USA. He is currently with LockheedMartin, Goleta, CA 93117 USA. (e-mail: [email protected]).

Color versions of one or more of the figures in this paper are available onlineat http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/JSTQE.2015.2427742

nology to security and industrial needs. This region contains alarge number of fundamental molecular rotational and vibra-tional resonances, including methane (3.3 μm), carbon dioxide(4.2 μm), and carbon monoxide (4.6 μm), making this regioncrucial for spectroscopic, atmospheric, and environmental stud-ies. The absorption strength at the fundamental resonances canbe several orders of magnitude higher than at overtones in thenear-infrared [1]. Therefore detection of trace amounts of gasor chemicals can be much more sensitive. There also exists anatmospheric transmission window between 3.5 and 4 μm thatcan be utilized for applications such as free-space optical com-munications, military countermeasures, and remote explosivedetection. These applications demand compact, high-efficiency,room temperature-operable laser sources.

While visible and near-infrared optical sources have evolvedinto very mature and varied technology bases, the mid-infraredregion has suffered from a relative scarcity of similarly well-developed technology. Several physical limitations have heldback the progress of mid-infrared photonics. The transparenciesof many materials are hampered by multi-phonon relaxation inthis region, leading to large losses for lasers and other pho-tonic devices. Additionally, materials that have suitable trans-parencies also often have poor mechanical strength, chemicalstability, and thermal conductivity.

Mid-infrared semiconductor laser sources can be generallyclassified into three main types: (1) intersubband quantumcascade lasers (QCLs), (2) interband cascade lasers (ICLs), typ-ically with type-II active regions, and (3) diode lasers, typicallywith type-I active regions. State-of-the-art performance dataof representative devices for each of these classes operating atroom temperature is summarized in Fig. 1. The next severalparagraphs summarize the development, advantages, anddrawbacks of these different approaches to mid-infrared lasersources.

The first QCL, emitting at 4.2 μm, was reported in 1994 [2].Quantum cascade lasers offer high output powers, tunability, andcan operate in continuous wave (cw) above room temperature[3]–[9]. Two features of QCLs set them apart from interbanddevices. First, the emission wavelength is due to the devicestructure design, the composition and spacing of the quantumwell (QW) layers and barriers, and not to the material bandgapas with interband devices [10]. This has allowed QCL fabrica-tion in well-developed substrate systems such as InP and GaAs.The second is that QCLs are unipolar, with electrons being theonly carrier. This allows for a single electron to undergo mul-tiple lasing transitions in multiple cascades of QWs, leading to

1077-260X © 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.See http://www.ieee.org/publications standards/publications/rights/index.html for more information.

1502410 IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 21, NO. 6, NOVEMBER/DECEMBER 2015

Fig. 1. State-of-the-art performance data for solid-state mid-infrared lasers versus lasing wavelength for room temperature (a) output power [8], [11]–[37], (b)threshold current density [11]-[28], [30]–[34], [36]–[41], (c) wallplug efficiency [12]–[20], [22]–[25], [29]–[38], and (d) characteristic temperature of thresholdcurrent [11], [12], [14], [18], [21]–[23], [27], [33], [38], [39], [41], [42].

high output powers (5.1 W cw, 120 W pulsed) and wallplugefficiencies (53% at 40 K, 21% at room temperature) [5], [22],[43]. These devices also have good threshold current tempera-ture stability since the intersubband nature of the devices freesthem from the effects of Auger recombination. Due to the shortupper state lifetime, however, these devices have much higherpower consumption at threshold, which potentially limits theirutility in portable applications.

Interband cascade lasers were first proposed by Yang et al.in 1994 and demonstrated with both type-I and type-II activeregions [44], [45]. Type-II devices progressed more quicklydue to the reduced Auger recombination rates stemming fromthe spatial separation of electrons and holes in the stag-gered type-II band alignments [46]. ICLs lend themselveswell to cascaded designs, much like QCLs, the difference be-ing that the optical transitions occur band-to-band [40], [47],[48]. The development of these devices has been rapid. Con-tinuous wave ICLs were limited to cryogenic operation aslate as 2004 [49]. Current devices now operate continuouswave at room temperature over much of the 3–5 μm range[29], [30], [48], [50]–[52], [55]–[57]. However, the stabil-ity of device performance with temperature remains poor, asevidenced by the relatively low characteristic temperatures

(T0), suggesting significant Auger recombination effects at longwavelengths.

Below 3 μm, type-I laser diodes grown on GaSb exhibithigh powers with high wall plug efficiencies [12], [13], [19],[33], [53]. Room temperature threshold current densities areeven comparable to the room temperature values reported fornear-infrared laser diodes [19], [53]–[57]. As the wavelengthof type-I laser diodes is extended beyond 3 μm, however, sev-eral problems limit the quest for efficient, reliable sources. Thephysical processes involved include Auger recombination due toinsufficient spin-orbit splitting, free-carrier absorption, and va-lence band offsets. Moreover, Auger recombination [58], [59]and free carrier absorption [60], [61], particularly by holes,become much more severe at longer wavelengths, further re-stricting efficient mid-infrared type-I diode laser sources. Themore urgent challenge that has historically plagued the wave-length extension of these sources is that the standard GaInAsSbQW alloy, surrounded by AlGaAsSb barriers, begins to have avery low valence band offset as the alloy composition is ad-justed to achieve longer wavelength emission. This reducesthe carrier confinement of holes in the QW region, allow-ing barrier states to become populated, reducing internal effi-ciency. These challenges serve to limit output powers and power

SIFFERMAN et al.: HIGHLY STRAINED MID-INFRARED TYPE-I DIODE LASERS ON GaSb 1502410

Fig. 2. Illustration of a typical approach to extending the laser emission wave-length in GaSb-based type-I QW diode lasers. (a) Increasing the indium contentof the GaInAsSb QWs decreases the bandgap, as well as increases the compres-sive strain in the QW. In order to mitigate the strain limitations, (b) the arseniccontent of the QWs must also be increased. This has the unintended consequenceof reducing the valence band offset, thereby reducing the hole confinement in theQW. (c) By adding indium and increasing the arsenic contents in the AlGaAsSbbarrier layers, the conduction and valence band offsets to be independentlytuned, allowing sufficient confinement for holes while avoiding uneven electrondistribution in the conduction band.

efficiency, increase threshold currents, and reduce the charac-teristic temperature as the emission wavelength is increased.

In this paper, we will focus on type-I QW diode lasers andhow increased compressive strain could potentially overcomethe challenges associated with decreasing hole confinement,though it likely also mitigates Auger recombination. We willdescribe some of the challenges that must be addressed toincorporate additional strain and some of our ongoing efforts tosurmount them. Section II describes conventional approachestaken to extend wavelength emission in GaSb-based type-Idiode lasers. Section III discusses how low temperature growthcan achieve the high strains necessary to extend efficientdiode laser emission to 4 μm and potentially beyond. SectionIV presents some potential alternative methods of extendingthe wavelength in GaSb-based materials using bismuth as asurfactant during active region growth as well as the use ofdilute applications of bismuth to extend the critical thicknesslimitations.

II. DESIGNING GASB-BASED LASER MATERIALS

TO EXTEND EMISSION WAVELENGTH

The typical QW alloy for GaSb-based laser devices isGaInAsSb. Using this alloy, watt-level output powers have beendemonstrated in lasers emitting up to about 2.5 μm [12], [14]and hundreds of milliwatts from 2.5 to past 3 μm [18], [23]at room temperature. These devices commonly employ barrierlayers consisting of alloys of AlGaAsSb or AlGaInAsSb lattice-matched to GaSb. Type-I GaSb-based lasers with AlGaAsSbbarriers have been demonstrated out to 3 μm [62] As shown inFig. 1, these devices exhibit excellent operating characteristics:low operating voltage, low threshold current, and high stabilityover operating temperature.

The typical approach to extend the emission wavelength issummarized in Fig. 2. The emission wavelength of GaInAsSbQWs can be increased by increasing the indium content of thealloy, as shown in Fig. 2(a). Additional indium lowers the QWbandgap as well as increases the compressive strain, improving

the overall valence band offset and the hole confinement. Toohigh indium content, however, pushes the strain past the limitwhere roughening occurs, or relaxation and misfit dislocationsbegin to form, which degrade luminescence efficiency. In orderto avoid the onset of relaxation or roughening, the number ofQWs must either be reduced, hampering laser performance dueto gain saturation, or the QW strain must be reduced. In order todecrease the compressive strain in the QW, the arsenic fraction istypically increased (see Fig. 2(b)). The increase in QW arseniccontent has the effect of lowering the valence band in thesealloys [24]. This reduces the barrier for hole confinement inthe QW and severely affects laser performance at and beyond3 μm [63].

The effect of increased arsenic content can be exploited inthe barriers. As illustrated in Fig. 2(c), the valence band off-set can be improved by increasing the arsenic content of thebarrier, lowering its valence band. In order to maintain latticematching, the aluminum fraction of AlGaAsSb barriers mustalso be increased. The higher aluminum content increases theconduction band offset, leading to uneven distributions of elec-trons in multi-QW active regions [63]. The increased aluminumcontent also reduces the index contrast between the waveguideand cladding layers, allowing the optical mode to leak out ofthe waveguide into the doped cladding layers and leading toincreased free carrier absorption losses.

Switching to an AlGaInAsSb quinternary alloy barriers pre-sented a novel path to engineering the active region in GaSb-based lasers, as shown in Fig. 2(c) [63]. The valence band andconduction band energies can be independently adjusted whilemaintaining lattice matching to the substrate. This allowed forbarriers that provide sufficient confinement for holes in the va-lence band, while also providing more moderate confinementin the conduction band to create more uniform carrier distribu-tions within the multi-QW active regions. This design strategyhas enabled emission wavelengths out to 3.44 μm [20] and, morerecently, 3.73 μm [27]. However, this approach is still subjectto strain and critical thickness limitations, both in the tensilely-strained barriers and compressively-strained QWs, ultimatelyleading to the degraded performance at longer wavelengths seenin Fig. 1.

An alternative design for the laser active region, barriers, andwaveguide is to employ GaInAsSb QWs with GaSb barriers[62], [64]. Aluminum-free active regions offer several advan-tages, including simpler materials fabrication and growth, andbetter index contrast with the cladding. The latter advantageleads to better optical mode confinement, improving differen-tial gain and reducing losses due to free carrier absorption in thedoped cladding layers [65]. Devices with GaSb barriers havebeen demonstrated out to about 3 μm [62]; however, they toosuffer from small valence band offsets and poor hole confine-ment, leading to reduced luminescence efficiency.

III. EXTENDING LASER EMISSION USING HIGHLY-STRAINED

QUANTUM WELLS AND AL-FREE ACTIVE REGIONS

By increasing the compressive strain in the QW, the va-lence band offset can be increased allowing for better holeconfinement and lower lasing thresholds. Previously reported


Fig. 3. Calculated valence band offset of a Ga0 .45 In0 .55 Asx Sb1−x /GaSbQW, as a function of increasing compressive strain, illustrating the increasedoffset with increased strain. Calculated with [67]. Note that significant va-lence band offsets are accessible, despite the GaSb barriers, due to the highQW indium and antimony contents that are achievable with growth at reducedtemperatures.

work has restricted QW strain to less than 2.0% in order to avoidstrain relaxation in the QW region. [27], [62], [66] Using param-eters from Krijn [67], we calculated that material strains beyond2.0% would be required for sufficient valence band offset inGa0.45In0.55AsxSb1−x QWs (see Fig. 3). To test the feasibilityof this approach, we grew 10-nm thick Ga0.45In0.55AsxSb1−xQWs with GaSb barriers on undoped (1 0 0)-oriented GaSbsubstrates in a Varian Gen. II solid source molecular beamepitaxy (MBE) system at a substrate temperature of 410 °C,which is typical for GaSb-based diode lasers. As shown inFig. 4(a), room-temperature photoluminescence (PL) measure-ments showed initial improved luminescence efficiency with in-creasing compressive strain. For further increases in QW strain(>2.1%), however, the luminescence significantly degraded.High-resolution X-ray diffraction (HR-XRD) ω-2θ symmetricscans, shown in Fig. 4(b), indicated a loss of structural qual-ity as the cause for this loss of luminescence. In particular, thehigh-frequency Pendellosung fringes in the HR-XRD scans be-came progressively less pronounced with increasing compres-sive strain. These fringes were absent for the highest strain levelstudied, suggestive of a loss of interface quality between theQW and the upper GaSb barrier. This is likely due to indiumsurface segregation during the QW growth [62].

By further reducing the growth temperature to 350 °C, thestructural quality of the QWs improved, with the QW strainincreasing up to 2.45% with strong luminescence efficiency, asshown in Fig. 5(a). Using a low substrate temperature duringthe growth of the GaInAsSb QWs, the indium surface segre-gation can be kinetically suppressed, leading to higher indiumincorporation in the wells, as well as superior structural qualityof the grown structures seen in Fig. 5(b). Room-temperature PLmeasurements of these structures also improved with increas-ing strain, attributable to the higher hole confinement from theincreased valence band offset. This offset was calculated to beabout 80 meV, comparable to the calculated valence band off-set in the more complex quinternary AlGaInAsSb barriers usedelsewhere [68].

To test the limits of using heavy compressive strain toextend the emission wavelength of GaInAsSb/GaSb QWs,

Fig. 4. (a) Room-temperature photoluminescence of otherwise identicalGaInAsSb/GaSb QWs of varying strains grown at 410 °C. Peak PL initiallyincreased with increasing strain before degrading rapidly. (b) HR-XRD ω-2θ symmetric scans about the GaSb (0 0 4) peak showed the Pendellosungfringes washed out as strain increased, suggesting a loss of layer interfacemorphology.

we grew proof-of-concept test structures consisting ofGa0.35In0.65AsSb/GaSb single QWs, varying the incorporatedgroup-V ratio. The compressive strain in the QW was increasedto 2.8% to maintain sufficient valence band offset (calculatedto be ∼70 meV). Room temperature PL spectra shown inFig. 6 demonstrate that the peak emission wavelength canbe extended to 3.9 μm, albeit with a ∼50% decrease in PLintensity. Further exploration of MBE growth space can likelyreduce this degradation in luminescence efficiency; however,these findings illustrate the potential of growth at reducedsubstrate temperatures as a viable method for extending theemission wavelength of GaSb-based diode lasers further into themid-infrared.

The concern with increasing strain is the limit it places onthe number of QWs that can be successfully grown in the laseractive region before the accumulated strain begins to degradethe structural quality of the material. Increasing the number ofQWs in the laser active region increases the differential gain,reducing the threshold for lasing [69]. Concentrating on the


Fig. 5. The effect of reducing the substrate growth temperature from 410 °C to350 °C. (a) Room-temperature PL redshifted and the peak emission increased by50%. (b) HR-XRD scans showed increased QW strain with strong Pendellosungfringes, indicative of high-quality interfaces.

Fig. 6. Room temperature PL emission of two GaInAsSb/GaSb QW samplesvarying the indium content, and hence the QW strain. By increasing the indiumcontent from 55% to 65% the peak photoluminescence emission shifted to3.9 μm.

3.4 μm active region design from Fig. 6, multi-QW structureswith up to five highly-strained GaInAsSb QWs, each separatedby 20-nm barriers of GaSb, were grown at 350 °C. These struc-tures all exhibited excellent structural quality, as evidenced byHR-XRD symmetrical ω-2θ scans showing them to be coher-ently strained (not shown). Power-dependent PL spectroscopywas performed at room temperature, under excitation densi-

Fig. 7. Peak room temperature PL emission intensity as a function of excita-tion intensity. illustrating that the peak PL intensity saturates for a single-QWactive region but multi-QW structures show less saturation effects.

Fig. 8. Band diagram of the highly-strained GaInAsSb type-I QW diode laserwith GaSb barriers, emitting at 3.4 μm. The band offsets were calculated using[67].

ties comparable to pumping near laser threshold. As illustratedin Fig. 7, under low excitation densities, the single-QW struc-ture exhibited higher peak PL efficiency than the multi-QWstructures; however, the single QW quickly saturated with in-creasing pump power, despite the high QW strain. Conversely,the multi-QW structures exhibited significantly less saturationeffects.

Devices with fewer than four QWs failed to lase due to thegain saturation implied from the power-dependent PL measure-ments discussed earlier. The active regions were also restrictedto five or fewer QWs due to critical thickness limitations. Dueto these bounds, a nominally-identical four-QW active region,shown in Fig. 8, was integrated in an edge-emitting laser struc-ture and processed into ridge waveguide lasers. The optical con-finement factor for this device structure as calculated by 2D finitedifference techniques was 0.91% per QW [70], [71]. Followinglaser stripe definition by photolithography and metallization us-ing electron-beam evaporation, an inductively-coupled plasmawas employed to etch most of the epitaxial layer stack, withthe metal contacts serving as the mask for the self-aligned etch.Silicon nitride was then blanket-deposited on the epi-side of the


Fig. 9. Light-output versus current input and (inset) lasing spectrum from aridge waveguide quadruple QW laser under pulsed mode operation at 10 °C.The threshold current density was 590 A/cm2.

Fig. 10. Natural logarithm of pulsed threshold current with device temper-ature. This laser exhibits a characteristic temperature of threshold current of50 K, one of the highest reported values for a laser operating near 3.4 μm.

sample. Contact windows above the laser stripes were definedby photolithography. The silicon nitride layer was then etchedwith a SF6 plasma to open up windows to the contact layer.Contact pads above the laser stripe contacts were defined usingconventional photolithography methods, and the contacts weredeposited by electron-beam evaporation. The GaSb substratewas lapped down to approximately 100 μm to aid the cleav-ing process. A bottom metal contact was then deposited usingelectron-beam evaporation. Following fabrication, lasers werecleaved into 2-mm long laser bars and mounted epi-side up oncopper submounts. The facets were left as-cleaved and no anti-reflection/high reflection coatings were applied. Devices weretested in pulsed mode using 200-ns current pulses at a repetitionrate of 10 kHz. As shown in Fig. 9, laser emission was observedat 3.4 μm, the longest emission wavelength from a GaSb-basedtype-I QW diode laser operating with an aluminum-free activeregion [23]. These devices exhibited a threshold current densityat 10 °C of 0.59 kA/cm2 and, as shown in Fig. 10, the char-acteristic temperature of threshold current, T0 , for this devicewas measured to be 50 K. The threshold current density perQW of 150 A/cm2 is comparable with the threshold densities

reported around this emission wavelength using more compli-cated quinternary active regions [64], [66], [68]. These findingsare promising for two reasons. First, they illustrate that furtherincreasing the QW strain is a viable path to longer wavelengthlasers with improved T0 and low threshold current density. Sec-ond, they offer the potential for further wavelength extensionusing the methods described in Section II of introducing alu-minum, indium, and arsenic into the barriers, but starting froma significantly longer lasing wavelength.

IV. THE USE OF BISMUTH TOWARD HIGHER

STRAINED MATERIALS

Based on the success extending the laser emission wave-length by increasing the compressive strain in the QW, we haveexplored the use of surfactant bismuth to assist the growth atelevated strain levels. Surfactant-assisted MBE growth wasintroduced by Copel et al. in 1989 as a method of suppressingisland formation of strained layers of germanium grown onsilicon [72]. The application of surfactant-mediated growth waslater applied to III-V materials by Massies et al. [73]. By addinga surface-segregating species during crystal growth, the growthkinetics are altered significantly; in the case of bismuth, a lowflux can promote layer-by-layer growth in metastable III-Vmaterial systems by kinetically-limiting the growth process,partially avoiding the limits imposed by thermodynamicconstraints [74]–[77].

As discussed in Section III, the key issue associated withgrowing highly-strained GaInAsSb QWs is the strain-drivensurface segregation of indium [62], which can cause rougheningand the inevitable introduction of defects that degrade opticalquality. While this can be mitigated to some degree by growingat reduced substrate temperatures, employing bismuth asa reactive surfactant could enable still higher compressivestrains, while maintaining or improving structural and opticalquality.

We have investigated the effects of low levels of bismuth fluxduring the growth of highly-strained GaInAsSb QWs. Samplescontaining varying numbers of GaInAsSb/GaSb QWs (rangingfrom one to seven) with 2.8% compressive strain were grownboth with and without bismuth, under conditions otherwisenominally identical to those shown in Fig. 6. The bismuth beamequivalent pressure ranged from 0 to 1.3 × 10−7 Torr. As shownin Fig. 11(a), a moderate bismuth flux of ∼5 × 10−8 Torr beamequivalent pressure improved the peak PL efficiency of themulti-QW structures increased by 25% compared to samplesgrown without bismuth. Equivalent single-QW structures didnot exhibit the marked difference in PL as the multi-QWstructures. Both single and multi-QW structures exhibitedwell-resolved peaks and Pendellosung fringes in HR-XRDω-2θ scans, shown in Fig. 11(b), indicative of high-qualityinterfaces. However, nominally-identical structures grown withhigher bismuth fluxes (∼1.2 × 10−7 Torr) exhibited no PL (seeFig. 11(a)) and significantly degraded structural quality (seeFig. 11(b)) suggesting an optimal growth window where bis-muth affords significant benefits. This window is currently beingexplored to optimize the growth of these materials; however,the increased luminescence efficiency from these active regions


Fig. 11. GaInAsSb/GaSb multi-QW structures strained to 2.8% compressive strain. (a) For low levels of bismuth flux (∼5×10−8 Torr beam equivalent pressure)during QW growth, the peak room-temperature PL emission increased by 25%. Emission degraded rapidly with an increase in bismuth flux during growth. (b)HR-XRD ω-2θ scans of multi-QW structures showing well-defined layer interfaces and QW superlattice peaks for lower fluxes of bismuth. Samples grown withhigher bismuth fluxes exhibited significantly degraded QW interfaces.

Fig. 12. Bandgap versus lattice parameter for GaSb, InAs, InSb, and theiralloys. Note the decrease in bandgap for dilute amounts of incorporated bismuth.

suggests that surfactant-mediated epitaxy of type-I GaSb-baseddiode laser active regions offers significant potential toextend their emission to even longer wavelengths, with highperformance.

An additional potential benefit to this approach is that bycombining a bismuth flux and low substrate growth tempera-tures, bismuth can incorporate into the crystal. In addition tothe well-documented bandgap reduction with alloying of bis-muthide compounds into conventional III-V materials, as illus-trated in Fig. 12 [78], the introduction of bismuth can also in-crease the spin-orbit splitting sufficiently [79] to mitigate Augerrecombination transitions involving the split-off band in diodelasers [80]. Indeed, spin-orbit splittings that exceed the tran-sition energy have been demonstrated in GaAsBi/GaAs [81]and GaInAsBi/InP [82]. However, our focus here is on a less-recognized benefit of incorporating bismuth, specifically thatthe critical thickness can be extended beyond the typical lim-its of traditional strained-layer epitaxy [83]–[85]. In particu-lar, the critical thickness of mid-IR materials can be extended

by greater than 10×, which could greatly alleviate the criticalthickness limitations on extending the lasing wavelength dis-cussed in Section II. These findings are consistent with whathas been observed for near-IR bandgap GaInAsBi materials onInP [86].

By applying recent advances from the MBE growth ofGaAsBi [87] to the growth of InAsBi and GaInAsBi on InAs,we identified a growth space free of bismuth droplet and al-ternate phase formation in which the optical quality is signifi-cantly enhanced. Samples were grown on n-type, sulfur-doped(100)-oriented InAs substrates and a range of substrate tem-peratures, bismuth beam equivalent pressures, and V/III fluxratios were explored. At substrate temperatures below ∼330°Cwith nearly stoichiometric (V/III) flux ratios of ∼1.15, bismuthincorporation became linear with bismuth flux, signifying near-unity sticking. In addition to preventing the formation of bis-muth surface droplets, this “unity sticking” growth regime [88]is extremely important for achieving high optical quality inhighly mismatched alloys, such as dilute-nitrides (for example,see [89]).

Under these growth conditions, InAsBi layers 150 nm thickwith bismuth concentrations up to 5.25%, as determined byHR-XRD ω-2θ measurements shown in Fig. 13(a), were grownwithout droplet formation, as confirmed by Nomarski phasecontrast microscopy and atomic force microscopy. As shown inthe (2 2 4) reciprocal space maps of Fig. 13(b), these films werefully coherent to the underlying InAs substrate, despite the layerthickness being as much as ∼12× the Matthews-Blakeslee crit-ical thickness. Certainly, further study is required to (1) clarifythe mechanisms mitigating strain relaxation and (2) integratethese materials into the active region of mid-IR diode lasers.However, these results suggest that bismuth-mediated growthcould greatly alleviate the design constraints imposed by strainrelaxation considerations, potentially greatly improving laserperformance (e.g. by more highly strained QWs with largervalence band offsets) or enabling still longer wavelength activeregions.


Fig. 13. (a) (0 0 4) ω-2θ XRD and (b) (2 2 4) reciprocal space maps of InAsBifilms with varying bismuth contents. The Mathews-Blakeslee critical thicknesscan be substantially exceeded by >10×, without evidence of lattice-relaxation.

V. CONCLUSION

A key challenge in extending the emission wavelength oftype-I mid-IR diode lasers is managing strain, along with the re-lated issues of band offsets, critical thickness limitation, androughening during growth. Using growth at a reduced sub-strate temperature of 350 °C, compressive strains of up to 2.8%have been incorporated into GaInAsSb QWs with GaSb bar-riers; these structures exhibited peak room-temperature photo-luminescence out to 3.96 μm. Using this growth method, low-threshold ridge waveguide lasers operating at 10 °C and emittingat 3.4 μm in pulsed mode were also demonstrated using 2.45%compressively strained GaInAsSb/GaSb QWs. These devicesoffer the potential for further wavelength extension using themethods described in Section II of introducing aluminum, in-dium, and arsenic into the barriers, but starting from a signifi-cantly longer lasing wavelength than previous devices.

We also described alternative approaches to managing strainissues in type-I diode laser active regions using surfactant-mediated epitaxy and incorporation of bismuth. In particular,this approach can increase the critical thickness to greater than10× the Matthews-Blakeslee limit, which could greatly alleviatethe critical thickness limitations on extending the lasing wave-length discussed in Section II.

These approaches to the growth of type-I GaSb-based diodelaser active regions offer significant potential to extend theiremission to longer wavelengths and improve performance.

REFERENCES

[1] L. S. Rothman, I. E. Gordon, Y. Babikov, A. Barbe, D. Chris Benner, P. F.Bernath, M. Birk, L. Bizzocchi, V. Boudon, L. R. Brown, A. Campargue,K. Chance, E. A. Cohen, L. H. Coudert, V. M. Devi, B. J. Drouin, A. Fayt,J.-M. Flaud, R. R. Gamache, J. J. Harrison, J.-M. Hartmann, C. Hill, J. T.Hodges, D. Jacquemart, A. Jolly, J. Lamouroux, R. J. Le Roy, G. Li, D. A.Long, O. M. Lyulin, C. J. Mackie, S. T. Massie, S. Mikhailenko, H. S. P.Mller, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. Perevalov, A. Perrin,E. R. Polovtseva, C. Richard, M. A. H. Smith, E. Starikova, K. Sung, S.Tashkun, J. Tennyson, G. C. Toon, Vl. G. Tyuterev, and G. Wagner., “TheHITRAN 2012 molecular spectroscopic database,” J. Quantum Spectrosc.Radiat. Transfer, vol. 130, pp. 4–50, Nov. 2013.

[2] J. Faist, F. Capasso, D. L. Sivco, C. Sirtori, and A. L. Hutchinson, “Quan-tum cascade laser,” Science, vol. 264, no. 5158, pp. 553–556, Apr. 1994.

[3] M. Razeghi, “High-performance InP-based mid-IR quantum cascadelasers,” IEEE J. Sel. Topics Quantum Electron., vol. 15, no. 3, pp. 941–951,May/Jun. 2009.

[4] O. Cathabard, R. Teissier, J. Devenson, and A. N. Baranov, “InAs-baseddistributed feedback quantum cascade lasers,” Electron. Lett., vol. 45, no.20, pp. 1028–1030, Sep. 2009.

[5] Y. Bai, S. Slivken, S. Kuboya, S.R. Darvish, and M. Razeghi, “Quan-tum cascade lasers that emit more light than heat,” Nat. Photon., vol. 4,pp. 99–102, Feb. 2010.

[6] O. Cathabard, R. Teissier, J. Devenson, J. C. Moreno, and A. N. Baranov,“Quantum cascade lasers emitting near 2.6 μm,” Appl. Phys. Lett., vol.96, no. 14, pp. 141110-1–141110-3, Apr. 2010.

[7] J. P. Commin, D. G. Revin, S. Y. Zhang, A. B. Krysa, K. Kennedy, andJ. W. Cockburn, “High peak power λ ∼ 3.3 and 3.5 μm InGaAs/AlAs(Sb)quantum cascade lasers operating up to 400 K,” Appl. Phys. Lett., vol. 97,no. 3, , pp. 031108-1–031108-3, Jul. 2010.

[8] N. Bandyopadhyay et al., “Watt level performance of quantum cas-cade lasers in room temperature continuous wave operation at λ ∼3.76 μm,” Appl. Phys. Lett., vol. 97, no. 13, , pp. 131117-1–131117-3,Oct. 2010.

[9] A. Bismuto, M. Beck, and J. Faist, “High power Sb-free quantum cascadelaser emitting at 3.3 μm above 350 K,” Appl. Phys. Lett., vol. 98, no. 19,pp. 191104-1–191104-3, May 2011.

[10] H. K. Choi, “Coherent semiconductor sources in the long-wavelengthinfrared spectrum,” in Long-Wavelength Infrared Semiconductor Lasers,H. K. Choi, Ed. Hoboken, NJ, USA: Wiley, 2004, pp. 1–17.

[11] H. K. Choi, G. W. Turner, and S. J. Eglash, “High-power GaInAsSb-AlGaAsSb multiple-quantum-well diode lasers emitting at 1.9 μm,” IEEEPhoton. Technol. Lett., vol. 6, no. 1, pp. 7–9, Jan. 1994.

[12] J. G. Kim et al., “Room-temperature 2.5 μm InGaAsSb/AlGaAsSb diodelasers emitting 1 W continuous waves,” Appl. Phys. Lett., vol. 81, no. 17,pp. 3146–3148, Oct. 2002.

[13] D. Donetsky, G. Kipshidze, L. Shterengas, T. Hosoda, and G. Belenky,“2.3 μm type-I quantum well GaInAsSb/AlGaAsSb/GaSb laser diodeswith quasi-CW output power of 1.4W,” Electron. Lett., vol. 43, no. 15,pp. 810–811, Jul. 2007.

[14] L. Shterengas, G. Belenky, M. V. Kisin, and D. Donetsky, “High power2.4 μm heavily strained type-I quantum well GaSb-based diode laserswith more than 1W of continuous wave output power and a maximumpower-conversion efficiency of 17.5%,” Appl. Phys. Lett., vol. 90, no. 1, ,pp. 011119-1–011119-3, Jan. 2007.

[15] M. Kim et al., “Interband cascade laser emitting at λ = 3.75 μm incontinuous wave above room temperature,” Appl. Phys. Lett., vol. 92, no.19, pp. 191110-1–191110-3, May 2008.

[16] Y. Bai, S. Slivken, S. R. Darvish, and M. Razeghi, “Room temperature con-tinuous wave operation of quantum cascade lasers with 12.5% wall plugefficiency,” Appl. Phys. Lett., vol. 93, no. 2, pp. 021103-1–021103-3, Jul.2008.

[17] A. Lyakh et al., “1.6W high wall plug efficiency, continuous-wave roomtemperature quantum cascade laser emitting at 4.6μm,” Appl. Phys. Lett.,vol. 92, no. 11, pp. 111110-1–111110-3, Mar. 2008.

[18] J. Chen et al., “2.7-μm GaSb-based diode lasers with quinary waveguide,”IEEE Photon. Technol. Lett., vol. 21, no. 16, pp. 1112–1114, Aug. 2009.

[19] M. T. Kelemen et al., “Diode laser arrays for 1.8 to 2.3 μm wavelengthrange,” Proc. SPIE, vol. 7483, p. 74830C, Sep. 2009.

[20] T. Hosoda, G. Kipshidze, L. Shterengas, and G. Belenky, “Diode lasersemitting near 3.44 μm in continuous-wave regime at 300K,” Electron.Lett., vol. 46, no. 21, pp. 1455–1457, Oct. 2010.

[21] K. Vizbaras and M.-C. Amann, “3.6 μm GaSb-based type-I lasers withquinternary barriers, operating at room temperature,” Electron. Lett.,vol. 47, no. 17, pp. 980–981, Aug. 2011.


[22] Y. Bai, N. Bandyopadhyay, S. Tsao, S. Slivken, and M. Razeghi, “Roomtemperature quantum cascade lasers with 27% wall plug efficiency,” Appl.Phys. Lett., vol. 98, no. 18, pp. 181102-1–181102-3, May 2011.

[23] T. Hosoda, “GaSb-based type-I diode lasers operating at 3 μm and above,”Ph.D. dissertation, Dept. Elect. Eng., Stony Brook University, StonyBrook, NY, USA, Dec. 2011.

[24] G. Belenky, L. Shterengas, G. Kipshidze, and T. Hosoda, “Type-I diodelasers for spectral region above 3 μm,” IEEE J. Sel. Topics QuantumElectron., vol. 17, no. 5, pp. 1426–1434, Sep.-Oct. 2011.

[25] Q. Y. Lu, Y. Bai, N. Bandyopadhyay, S. Slivken, and M. Razeghi, “2.4 Wroom temperature continuous wave operation of distributed feedbackquantum cascade lasers,” Appl. Phys. Lett., vol. 98, no. 18, pp. 181106-1–181106-3 181106, May 2011.

[26] J. D. Bruno et al., “High-performance interband cascade lasers emittingbetween 3.3 and 3.5 microns,” Proc. SPIE, vol. 8374, p. 83740L, May2012.

[27] K. Vizbaras and M.-C. Amann, “Room-temperature 3.73 μm GaSb-basedtype-I quantum-well lasers with quinternary barriers,” Semicond. Sci.Technol., vol. 27, no. 3, pp. 032001-1–032001-4, Jan. 2012.

[28] G. Ryu et al., “Gain and losses and room-temperature operation in inter-band cascade lasers,” IEEE Photon. J., vol. 4, no. 1, pp. 133–142, Jan.2012.

[29] W. Bewley et al., “High-power room-temperature continuous-wave mid-infrared interband cascade lasers,” Opt. Exp., vol. 20, no. 19, pp. 20894–20901, Aug. 2012.

[30] W. Bewley et al., “Continuous-wave interband cascade lasers operatingabove room temperature at λ = 4.7–5.6 μm,” Opt. Exp., vol. 20, no. 3,pp. 3235–3240, Mar. 2012.

[31] N. Bandyopadhyay, Y. Bai, S. Tsao, S. Nida, S. Slivken and M. Razeghi,“Room temperature continuous wave operation of λ ∼ 3–3.2 μm quan-tum cascade lasers,” Appl. Phys. Lett., vol. 101, no. 24, p. 241110,Dec. 2012.

[32] N. Bandyopadhyay, S. Slivken, Y. Bai, and M. Razeghi, “High power,continuous wave, room temperature operation of λ ∼ 3.4 μm and λ ∼3.55 μm InP-based quantum cascade lasers,” Appl. Phys. Lett., vol. 100,no. 21, p. 212104, May 2012.

[33] L. Shterengas et al., “Cascade pumping of GaSb-based type-I quantumwell diode lasers,” Proc. SPIE, vol. 9002, p. 900213, Feb. 2014.

[34] L. Shterengas et al., “Cascade type-I quantum well diode lasers emitting960 mW near 3 μm,” Appl. Phys. Lett., vol. 105, no. 16, p. 161112, Oct.2014.

[35] R. Liang et al., “Diffraction limited 3.15 μm cascade diode lasers,” Semi-cond. Sci. Technol., vol. 29, no. 11, p. 115016, Sep. 2014.

[36] C. Canedy et al., “Pulsed and CW performance of 7-stage interband cas-cade lasers,” Opt. Exp., vol. 22, no. 7, pp. 7702–7710, Mar. 2014.

[37] Y. Jiang et al., “Type-I interband cascade lasers near 3.2 μm,” Appl. Phys.Lett., vol. 106, no. 4, p. 041117, Jan. 2015.

[38] S. Simanowski et al., “Growth and layer structure optimization of 2.26 μm(AlGaIn)(AsSb) diode lasers for room temperature operation,” J. Cryst.Growth, vols. 227–228, pp. 595–599, Jul. 2001.

[39] T. Hosoda, G. Belenky, L. Shterengas, G. Kipshidze, and M. V. Kisin,“Continuous-wave room temperature operated 3.0μm type I GaSb-basedlasers with quinternary AlInGaAsSb barriers,” Appl. Phys. Lett., vol. 92,no. 9, p. 091106, Mar. 2008.

[40] I. Vurgaftman et al., “Mid-infrared interband cascade lasers operating atambient temperatures,” New J. Phys., vol. 11, p. 125015, Dec. 2009.

[41] H. P. Nair, R. Salas, N. T. Sheehan, S. J. Maddox, and S. R. Bank,“3.4 μm diode lasers employing Al-free GaInAsSb/GaSb MQW activeregions at 20 °C,” in Proc. 71st Annual Device Res. Conf., Jun. 2013,pp. 187–188.

[42] R. Liang et al., “Novel cascade diode lasers based on type-I quantumwells,” Int. J. High Speed Electron. Syst., vol. 23, p. 1450022, Jul. 2014.

[43] Y. Bai et al., “High power broad area quantum cascade lasers,” Appl. Phys.Lett., vol. 95, no. 22, p. 221104, Dec. 2009.

[44] R. Q. Yang, “Infrared laser based on intersubband transitions in quantumwells,” Superlattices Microstruct., vol. 17, no. 1, pp. 77–83, Jan. 1995.

[45] J. R. Meyer, I. Vurgaftman, R. Q. Yang, and L. R. Ram-Mohan, “Type-II and type-I interband cascade lasers,” Electron. Lett., vol. 32, no. 1,pp. 45–46, Jan. 1996.

[46] W. W. Bewley et al., “Lifetimes and Auger coefficients in type-II Winterband cascade lasers,” Appl. Phys. Lett., vol. 93, no. 4, p. 041118, Jan.2008.

[47] W. W. Bewley et al., “Interband cascade laser operating CW to 257 K atλ = 3.7 μm,” Appl. Phys. Lett., vol. 89, no. 16, p. 161106, Oct. 2006.

[48] I. Vurgaftman et al., “Interband cascade lasers with low threshold powersand high output powers,” IEEE J. Sel. Topics Quantum Electron., vol. 19,no. 4, p. 1200210, Jul.–Aug. 2013.

[49] C. L. Canedy et al., “High-power continuous-wave midinfrared type-II‘W’ diode lasers,” Appl. Phys. Lett., vol. 85, no. 21, p. 211105, May 2005.

[50] S. Forouhar et al., “Reliable mid-infrared laterally-coupled distributed-feedback interband cascade lasers,” Appl. Phys. Lett., vol. 105, no. 5,p. 051110, Aug. 2014.

[51] R. Weih, A. Bauer, M. Kamp, and S. Hofling, “Interband cascade laserswith AlGaAsSb bulk cladding layers,” Opt. Mater. Exp., vol. 3, no. 10,pp. 1624–1631, Sep. 2013.

[52] W. W. Bewley et al., “Mid-IR interband cascade laser operating with <30mW of input power,” Proc. SPIE, vol. 8374, p. 83740H, May 2012.

[53] A. Garnache et al., “2–2.7μm single frequency tunable Sb-based lasersoperating in CW at RT: microcavity and external cavity VCSELs, DFB,”Proc. SPIE, vol. 6184, p. 61840N, Apr. 2006.

[54] Y. Roulillard, J. Angellier, A. Salhi, P. Grech, and F. Chevrier,“GaInAsSb/AlGaAsSb laser diodes for the 2- to 3-μm range,” Proc. SPIE,vol. 5738, p. 120, Apr. 2005.

[55] C. Lin, M. Grau, O. Dier, and M.-C. Amann, “Low thresh-old room-temperature continuous-wave operation of 2.24–3.04 μmGaInAsSb/AlGaAsSb quantum-well lasers,” Appl. Phys. Lett., vol. 84,no 25, pp. 5088–5090, Jul. 2004.

[56] G. Turner, H. Choi, and M. Manfra, “Ultralow-threshold (50 A/cm2)strained single-quantum-well GaInAsSb/AlGaAsSb lasers emitting at2.05 μm,” Appl. Phys. Lett., vol. 72, no. 8, p. 876, Jan. 1998.

[57] A. Salhi, Y. Rouillard, J. Angellier, P. Grech, and A. Vicet, “2.61 μmGaInAsSb/AlGaAsSb type I quantum well laser diodes with low thresh-old,” Electron. Lett., vol. 40, no. 7, pp. 424–425, Apr. 2004.

[58] A. R. Adams, “Band-structure engineering for low-threshold high-efficiency semiconductor lasers,” Electron. Lett.,vol. 22, no. 5, pp. 249–250, Feb. 1986.

[59] E. Yablonovitch and E. O. Kane, “Reduction of lasing threshold currentdensity by the lowering of valence band effective mass,” J. Lightw. Tech-nol., vol. 4, no. 5, pp. 504–506, May 1986.

[60] A. R. Adams, M. Asada, Y. Suematsu, and S. Arai, “The temperaturedependence of the efficiency and threshold current of In1−x Gax Asy P1−ylasers related to intervalence band absorption,” Jpn. J. Appl. Phys.,vol. 19, no. 10, p. L621, Oct. 1980.

[61] A. Chandola, R. Pino and P. S. Dutta, “Below bandgap optical absorp-tion in tellurium-doped GaSb,” Semicond. Sci. Technol., vol. 20, no. 8,pp. 886–893, Jul 2005.

[62] T. Lehnhardt et al., “Continuous wave single mode operation ofGaInAsSb/GaSb quantum well lasers emitting beyond 3 μm,” Appl. Phys.Lett., vol. 92, no. 18, p. 183508, May 2008.

[63] M. Grau, C. Lin, O. Dier, C. Lauer, and M.-C. Amann, “Room-temperatureoperation of 3.26 μm GaSb-based type-I lasers with quinternary Al-GaInAsSb barriers,” Appl. Phys. Lett., vol. 87, no. 24, pp. 241104, Dec.2005.

[64] N. Bertru et al., “Long-wavelength (Ga, In)Sb/GaSb strained quantumwell lasers grown by molecular beam epitaxy,” Semicond. Sci. Technol.,vol. 13, no. 8, p. 936, Aug. 1998.

[65] T. Lehnhardt et al., “Influence of GaSb and AlGaInAsSb as barrier materialon ∼2.8-μm GaSb-based diode laser Properties,” IEEE Photon. Technol.Lett., vol. 23, no. 6, pp. 371–373, Jan. 2011.

[66] J. G. Kim, L. Shterengas, R. U. Martinelli, and G. L. Belenky, “High-power room-temperature continuous wave operation of 2.7 and 2.8 μmIn(Al)GaAsSb/GaSb diode lasers,” Appl. Phys. Lett., vol. 83, no. 10,p. 1926, Sep. 2003.

[67] M. Krijn, “Heterojunction band offsets and effective masses in III-V qua-ternary alloys,” Semicond. Sci. Technol., vol. 6, no. 1, p. 27, Jan. 1991.

[68] K. Vizbaras et al., “Room-temperature type-I GaSb-based lasers in the3.0 - 3.7 μm wavelength range,” Proc. SPIE, vol. 8277, p. 82771B, Feb.2012.

[69] L. Shterengas, G. Belenky, T. Hosoda, G. Kipshidze, and S. Suchalkin,“Continuous wave operation of diode lasers at 3.36 μm at 12 °C,” Appl.Phys. Lett., vol. 93, no. 1, p. 011103, Jul. 2008.

[70] V. A. Sabnis, “Optically-controlled electroabsorption modulators for fu-ture generation optical networks,” Ph.D dissertation, Dept. Elect. Eng.,Stanford University, Stanford, CA, USA, Dec. 2003.

[71] S. Adachi, “Band gaps and refractive indices of AlGaAsSb, GaInAsSb,and InPAsSb: Key properties for a variety of the 2–4 μm optoelec-tronic device applications,” J. Appl. Phys., vol. 61, pp. 4869–4876, May1987.

[72] W. Li, J. B. Heroux and W. I. Wang, “InGaAsSbN: A dilute nitride com-pound for midinfrared optoelectronic devices,” J. Appl. Phys., vol. 94,no. 7, p. 4248, Sep. 2003.

[73] J. Massies, N. Grandjean, and V. H. Etgens, “Surfactant-mediated epitaxialgrowth of Inx Ga1−x As on GaAs (001),” Appl. Phys. Lett., vol. 61,no. 1, p. 99, Apr. 1992.


[74] E. Tournie, and K. H. Ploog. “Surfactant-mediated molecular beam epi-taxy of strained layer semiconductor heterostructures.” Thin Solid Films,vol. 231, nos. 1/2, pp. 43–60, Aug. 1993.

[75] M. R. Pillai, S.-S. Kim, S. T. Ho, and S. A. Barnett, “Growth ofInx Ga1−x As/GaAs heterostructures using Bi as a surfactant,” J. Vac. Sci.Technol. B, vol. 18, no. 3, pp. 1232–1236, Mar. 2000.

[76] S. Tixier, M. Adamcyk, E.C. Young, J.H. Schmid, and T. Tiedje, “Surfac-tant enhanced growth of GaNAs and InGaNAs using bismuth,” J. Cryst.Growth, vol. 251, nos. 1–4, pp. 449–454, Apr. 2003.

[77] E. C. Young, S. Tixier, and T. Tiedje, “Bismuth surfactant growth ofthe dilute nitride GaNx As1−x ,” J. Cryst. Growth, vol. 279, nos. 3–4,pp. 316–320, Jun. 2005.

[78] K. Oe, S. Ando, and K. Sugiyama, “InSb1−x Bix Films Grown by Molec-ular Beam Epitaxy,” Jpn. J. Appl. Phys., vol. 20, no. 4, p. L303, Mar.1981.

[79] B. Fluegel et al., “Giant Spin-Orbit Bowing in GaAs1−x Bix ,” Phys. Rev.Lett., vol. 97, no. 6, p. 067205, Aug. 2006.

[80] S. J. Sweeney and S. R. Jin, “Bismide-nitride alloys: Promising for effi-cient light emitting devices in the near- and mid-infrared,” J. Appl. Phys.,vol. 113, no. 4, p. 043110, Jan. 2013.

[81] Z. Batool et al., “The electronic band structure of GaBiAs/GaAs lay-ers: Influence of strain and band anti-crossing,” J. Appl. Phys., vol. 111,no. 11, p. 113108, Jun. 2012.

[82] I. P. Marko et al., “Temperature and Bi-concentration dependence of thebandgap and spin-orbit splitting in InGaBiAs/InP semiconductors for mid-infrared applications,” Appl. Phys. Lett., vol. 101, no. 22, p. 221108, Nov.2012.

[83] J. W. Matthews and A. E. Blakeslee, “Defects in epitaxial multilayers: I.Misfit dislocations,” J. Cryst. Growth, vol. 27, pp. 118–125, Dec. 1974.

[84] J. W. Matthews and A. E. Blakeslee, “Defects in epitaxial multilayers:II. Dislocation pile-ups, threading dislocations, slip lines and cracks,” J.Cryst. Growth, vol. 29, no. 3, pp. 273–280, Jul. 1975.

[85] J. W. Matthews and A. E. Blakeslee, “Defects in epitaxial multilayers: III.Preparation of almost perfect multilayers,” J. Cryst. Growth, vol. 32, no.2, pp. 265–273, Feb. 1976.

[86] Y. Zhong, P. B. Dongmo, J. P. Petropoulos, and J. M. O. Zide, “Effectsof molecular beam epitaxy growth conditions on composition and opticalproperties of Inx Ga1−x Biy As1−y ,” Appl. Phys. Lett, vol. 100, no. 11,p. 112110, Mar. 2012.

[87] A. J. Ptak et al., “Kinetically limited growth of GaAsBi by molecular-beamepitaxy,” J. Cryst. Growth, vol. 338, no. 1, pp. 107–110, Jan. 2012.

[88] S. G. Spruytte et al., “Incorporation of nitrogen in nitride-arsenides: Originof improved luminescence efficiency after anneal,” J. Appl. Phys., vol. 89,no. 8, pp. 4401–4406, Apr. 2001.

[89] H. B. Yuen, S. R. Bank, M. A. Wistey, J. S. Harris, and A. Moto, “Compar-ison of GaNAsSb and GaNAs as quantum-well barriers for GaInNAsSboptoelectronic devices operating at 1.3—1.55 μm,” J. Appl. Phys., vol. 96,no. 11, pp. 6375–6381, Nov. 2004.

Scott D. Sifferman (S’01–M’06–S’12) received theB.S.E. degree from Arizona State University, Tempe,AZ, USA, in 2004, and the M.S. degree from Stan-ford University, Stanford, CA, USA, in 2006, bothin electrical engineering. He is currently working to-ward the Ph.D. degree in electrical engineering atThe University of Texas at Austin, Austin, TX, USA,working on MBE growth, fabrication, and character-ization of mid-infrared GaSb-based lasers. He pre-viously worked in industry as a Systems Engineerdeveloping space-based photonics communications

systems and as a Research and Development Engineer for an industrial automa-tion and machine vision manufacturer.

Hari P. Nair received the B.Tech degree in engineer-ing physics from the Indian Institute of TechnologyMadras, India, in 2006. He received the M.S. andPh.D. degrees from The University of Texas at Austin,Austin, TX, USA, both in electrical engineering, in2009 and 2013, respectively. His doctoral researchled to a new approach to extend the emission wave-length of mid-infrared diode lasers. He is currently aPostdoctoral Scholar with the Materials Science andEngineering Department, Cornell University, Ithaca,NY, USA.

Rodolfo Salas received the B.S. and M.S. degreesfrom the University of Oklahoma, Norman, OK,USA, both in electrical engineering, in 1999 and2006, respectively. He previously worked in indus-try as a Research and Development Engineer work-ing on mid-infrared lasers and tunable diode laserabsorption spectroscopy. He is currently working to-ward the Ph.D. degree at The University of Texas atAustin, Austin, TX, USA, studying MBE growth andproperties of new photoconductive materials for THzgeneration.

Nathanial T. Sheehan received the B.S. degree fromthe University of California, Santa Barbara, SantaBarbara, CA, USA, in 2012, and the M.S. degreefrom the University of Texas at Austin, Austin, TX,USA, in 2014, and is currently working toward thePh.D. degree at The University of Texas at Austin, allin electrical engineering. His current research inter-ests include novel III–V materials grown by molecu-lar beam epitaxy, and the fabrication of mid-infraredoptical devices from these epitaxial materials.

Scott J. Maddox (S’12) received the B.S. and M.S.degrees from The University of Texas at Austin,Austin, TX, USA, in 2009 and 2011, respectively,where he is currently working toward the Ph.D. de-gree, all in electrical engineering. His current researchinterests include design, MBE growth, and characteri-zation of novel III–V materials and optoelectronic de-vices including epitaxial plasmonic materials, dilute-bismuthide alloys for photodetector applications, andlow-noise InAs avalanche photodiodes.

Adam M. Crook received the B.S. degree from theUniversity of California, Santa Barbara, CA, USA, in2007, and the M.S. and Ph.D. degrees from The Uni-versity of Texas at Austin (UT), Austin, TX, USA, allin electrical engineering, in 2009 and 2012, respec-tively. While at UT, he studied the MBE growth andoptical properties of mid-infrared compound semi-conductor and rare-earth nanocomposite optical ma-terials. In 2012, he joined the R&D division at Lock-heed Martin, Goleta, CA.

Seth R. Bank (S’95–M’06–SM’11) received the B.S.degree from the University of Illinois at Urbana-Champaign (UIUC), Urbana, IL, USA, in 1999 andthe M.S. and Ph.D. degrees from Stanford University,Stanford, CA, USA, all in electrical engineering, in2003 and 2006, respectively.

While at UIUC, he studied the fabrication ofInGaP–GaAs and InGaAs–InP HBTs. His Ph.D.research focused upon the MBE growth, fabrica-tion, and device physics of long-wavelength VC-SELs and low-threshold edge-emitting lasers in the

GaInNAs(Sb)–GaAs material system. In 2006, he was a Postdoctoral Scholarat the University of California, Santa Barbara, CA where his research centeredon the growth of metal–semiconductor hetero- and nanostructures (e.g., ErAsnanoparticles in GaAs). In 2007, he joined The University of Texas at Austin,Austin, TX, USA, where he is currently an Associate Professor of Electricaland Computer Engineering and holder of a Temple Foundation Endowed Fac-ulty Fellowship. His current research interests include the MBE growth of novelheterostructures and nanocomposites and their device applications. He has coau-thored more than 200 papers and presentations in these areas.

Dr. Bank received the 2010 Young Investigator Program Award from ONR,a 2010 NSF CAREER Award, a 2009 Presidential Early Career Award for Sci-entists and Engineers nominated by ARO, a 2009 Young Investigator ProgramAward from AFOSR, the 2009 Young Scientist Award from the InternationalSymposium on Compound Semiconductors, a 2008 DARPA Young FacultyAward, the 2008 Young Investigator Award from the North American MBEMeeting, and several best paper awards.

highly strained mid-infrared type-i diode lasers on gasb

Documents