IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 14, NO. 4, JULY/AUGUST 2008
1239
Quantum Dash Intermixing Hery Susanto Djie, Member, IEEE, Yang Wang, Student Member, IEEE, Yun-Hsiang Ding, Dong-Ning Wang, James. C. M. Hwang, Fellow, IEEE, Xiao-Ming Fang, Ying Wu, Joel M. Fastenau, Amy W. K. Liu, Senior Member, IEEE, Gerard T. Dang, Wayne H. Chang, Member, IEEE, and Boon S. Ooi, Senior Member, IEEE
Abstract—We investigate the intermixing effect in InAs/ InAlGaAs quantum-dash-in-well structures grown on InP substrate. Both impurity-free vacancy disordering (IFVD) via dielectric cap annealing, and impurity-induced disordering (IID) using nitrogen ion-implantation techniques have been employed to spatially control the group-III intermixing in the quantum-dash (Qdash) system. Differential bandgap shifts of up to 80 nm and 112 nm have been observed from the IFVD and IID processes, respectively. Compared to the control (nonintermixed) lasers, the light–current characteristics for the 125 nm wavelength shifted Qdash lasers are not significantly changed, suggesting that the quality of the intermixed material is well preserved. The intermixed lasers exhibit a narrower linewidth as compared to the asgrown laser due to the improved dash homogeneity. The integrity of the material is retained after intermixing, suggesting the potential application for the planar integration of multiple active/passive Qdash-based devices on a single InP chip. Index Terms—Disordering, interdiffusion, ion-implantation, photonic integrated circuits (PICs), quantum-dash (Qdash), quantum-dot intermixing, quantum-well (QW) intermixing.
I. INTRODUCTION HE SPATIALLY selective bandgap engineering using intermixing methods (or sometimes referred as interdiffusion or disordering) has been successfully implemented for the fabrication of multiple-section photonic integrated circuits (PICs) [Fig. 1(a)] [1]–[9]. This technology enables various benefits such as excellent alignment, negligible reflection losses, and intrinsic mode matching. These unique features provide a very enticing vision for future monolithic integration of photonics/optoelectronics. Further advances of quantum-well (QW) intermixing has enabled improved photonic device performances
T
Manuscript received December 2007. This work was supported in part by the National Science Foundation under Grant 0725647, in part by the United States Army Research Laboratory through Lehigh-Army Cooperation Agreement, and in part by the Commonwealth of Pennsylvania, Department of Community and Economic Development. H. S. Djie was with the Department of Electrical and Computer Engineering, Lehigh University, Bethlehem, PA 18015 USA. He is currently with JDS Uniphase Corporation, San Jose, CA 95134 USA (e-mail:
[email protected]). Y. Wang, Y.-H. Ding, D. N. Wang, J. C. M. Hwang, and B. S. Ooi are with the Department of Electrical and Computer Engineering and the Center for Optical Technologies, Lehigh University, Bethlehem, PA 18015 USA (e-mail:
[email protected];
[email protected];
[email protected];
[email protected];
[email protected]). X.-M. Fang, Y. Wu, J. M. Fastenau, and A. W. K. Liu are with the IQE Inc., Bethlehem, PA 18015 USA (e-mail:
[email protected];
[email protected];
[email protected];
[email protected]). G. T. Dang and W. H. Chang are with the Army Research Laboratory, Adelphi, MD 20783 USA (e-mail:
[email protected];
[email protected]). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JSTQE.2008.921396
Fig. 1. (a) Illustration of photonic integration circuits that combines various functional active and passive devices across single chip including laser diode, photodetector, electroabsorption (EA) modulator, waveguide, and combiner. (b) Blue-shift phenomenon in the interband transition of quantum-confined energy levels from conduction E c to valence band E v that can be utilized for the fabrication of multiwavelength laser and superluminescent diodes.
with cost-effective fabrication process such as wavelengthshifted laser, multiple-wavelength laser array, broadband superluminescent diodes, etc. [Fig. 1(b)]. A typical intermixing process involves the introduction of beneficial defects such as vacancies and interstitials to the QW material. During hightemperature annealing, diffusion of point defects enhance the atomic interdiffusion rate between the quantum nanostructure and the quantum barrier; hence, promote intermixing as illustrated in the inset of Fig. 1(b). As a result, the bandgap of the quantum nanostructure is being converted from abrupt to parabolic profile and leads to a blueshift of bandgap energy. Current advances in III–V compound, semiconductor quantum-confined nanostructures utilizing quantum-dots (Qdot) and quantum-dash (Qdash) have gradually placed this material system en route to succeed their QW counterpart for high-performance photonic devices [10]. Meanwhile, there has been an increased interest to spatially modify the Qdot and Qdash properties in the same way as QW structures [11], [12]. This effort stems primarily from the combined advantages of postgrowth bandgap engineering processes and the attractive properties of Qdot or Qdash devices. These unique properties include low-threshold current, low-temperature sensitivity, high-quantum efficiency, low chirp, and high-modulation frequency. Our early studies on both theoretical [13]–[15] and experimental [16]–[19] works suggest that high annealing temperature will induce a significant interdiffusion effect in Qdot, since the interface area to volume ratio between Qdot and the surrounding barriers is large when compared to QW. While the preliminary results seem encouraging [11]–[19], the demonstration of active Qdot/Qdash devices subjected to intermixing process is still lacking. Moderate temperature annealing causes
1077-260X/$25.00 © 2008 IEEE
1240
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 14, NO. 4, JULY/AUGUST 2008
Fig. 2. (a) Illustration of one-dimensional (1-D) diffusion of group-III atoms in QW structure and 3-D diffusion in Qdash structure due to the composition gradient of In and Ga across the InAlGaAs system. (b) Conduction band edge profile of a single QW after interdiffusion across the growth direction z with varying interdiffusion degree. Inset shows half cross section of the intermixed QW band edge profile from the middle of the well (c) Amount of bandgap shift from the QW energy transition as a function of Indium diffusion length calculated for different group-III diffusion rate: 1) D In < D G a ; 2) D In < D G a ; and 3) D In > D G a .
degradation in the material quality and optical properties of the self-assembled nanostructure materials due to the high strain around the dot/dash interface. The degradation becomes more severe since the intermixing usually employs the introduction of defects to enhance the interdiffusion rate to a specific region. These will pose significant roadblocks for the usefulness of intermixed materials in the fabrication of high quality photonic devices. Therefore, the demonstration of intermixed devices is important as it can provide valuable insight into the capabilities of the process in regards to the monolithic integration of active–passive Qdot/Qdash-based devices at a postgrowth level. In this paper, we report the development of postgrowth bandgap trimming of InAs/InAlGaAs Qdash grown on InP substrate. The theoretical model of group-III interdiffusion in InAs/InAlGaAs system has been developed to further gain insight of the interdiffusion process. The material properties and the modification subjected to the heat treatment are studied and applied to the device fabrication process. The impurity-free induced disordering (IFVD) with dissimilar dielectric cap annealing and the nitrogen ion-implantation induced disordering (NIID) techniques were used to spatially control the group-III intermixing rate in InAs/InAlGaAs Qdash material. The bandgaptuned laser is demonstrated using the optimized processes. II. THEORETICAL MODEL OF GROUP-III INTERDIFFUSION The intermixing in the InAlGaAs quarternary system is complex as it involves simultaneous diffusion of multiple group-III atoms (In, Ga, and Al), as depicted in Fig. 2(a). The understanding of multiple species diffusion processes is important to the interpretation of interdiffusion induced compositional change and the band structure modification related to the experimental works presented in the next section. In a multiple species system, the diffusion flux of any species does not depend only on its own concentration or chemical potential gradient, but also on those of all other diffusing species. Assuming that only the first spatial derivatives of concentrations
are important for the flux of any species, Fick’s law can be extended to a system of n species in the linear domain [19]. Because of the small Qdash height to base ratio of ∼0.1 or less [19], we assume that the intermixing of the quantized energy level is dominated by the diffusion in the vertical direction. With the assumption that the diffusion is 1-D (i.e., there is a concentration gradient only along the growth direction z), and the effective diffusion coefficient is assumed to be independent of position, we model the general solution for a single Qdash diffusion problem as B Ci − CiW B Ci (z) = Ci − 2 LZ − 2z LZ + 2z + erf (1) × erf 4Ld,i 4Ld,i where CiW and CiB are the initial concentrations of atom i in the well and in the barrier, respectively, Lz is the initial well width before interdiffusion, Ld,i = Di t is the effective diffusion length of each atom i in a multiple species system, and the QW is centered at z = 0. Here, we assume different diffusion rates of In and Ga in the modeling of this multiple cations intermixing system, which leads to different effective diffusion lengths of In and Ga. The change in the compositional profiles after interdiffusion modifies the carrier confinement potential. The electronics states of interdiffused QW structure can be obtained by solving the BenDaniel–Duke’s equation using finite difference method [20], which has been applied for numerous studies on the group-III (i.e., In–Ga, Al–Ga) and group-V (As–P, As–N) intermixing effect earlier [21]–[24]. Fig. 2(b) depicts the conduction band edge profile of an arbitrary QW structure embedded in a barrier and separate confinement heterostructure (SCH) as a superposition of the interdiffusion of three-independent layers. While in the InAlGaAs QW diffusion, the intermixing occurs in 1-D, in InAs Qdash diffusion, the atomic exchange occurs in 3-D, as illustrated in Fig. 2(a), due to the atomic gradient of In and Ga atoms across the heterointerfaces. As can be seen from Fig. 2(c), the disparity between the In and Ga diffusion rate will result in the different amount of wavelength shift. If Ga diffuses faster than In (i.e., DIn < DGa ), the intermixing will produce a smaller wavelength blueshift as compared with the case of In diffuses faster than Ga (i.e., DIn > DGa ). The similar results can be expected in the Qdash diffusion problem, as the typical self-assembled Qdash has a small height-to-base ratio (∼0.2 or less); therefore, the main intermixing effect results from 1-D diffusion across growth direction. III. EXPERIMENTS The InAs/InAlGaAs Qdash material used in this study was grown by molecular beam epitaxy (MBE) on (1 0 0) oriented InP substrate [19], [25]. The laser is a p-i-n structure with active region consisting of four-sheet of InAs Qdashes, and each Qdash layer is embedded in an asymmetric InAlGaAs QW. The QWs are then sandwiched between two sets of SCHs, as schematically illustrated in
SUSANTO DJIE et al.: QUANTUM DASH INTERMIXING
1241
Fig. 4. State-filling PL spectra of the following measured at 77 K under different excitation powers from 3 to 1500 W/cm2 . (a) InAs/InAlGaAs Qdash. (b) InAs/InP Qdot.
Fig. 3 (a) Schematic bandgap diagram of 4-stack InAs/InAlGaAs Qdash active region. (b) Plane-view AFM image (area of 0.5 × 0.5 µm2 ; height contrast of 8 nm). (c) Cross-sectional TEM images of InAs/InAlGaAs Qdash-in-well structure.
Fig. 3(a). The quantum-dash-in-well structure consists of a 1.3-nm-thick compressively strained In0.64 Ga0.16 Al0.2 As layer, a five-monolayer- (ML) thick InAs dash layer, and a 6.3nm-thick compressively strained In0.64 Ga0.16 Al0.2 As layer. Each dash-in-well stack is separated by a 30-nm-thick tensilestrained In0.50 Ga0.32 Al0.18 As layer that acts as the straincompensating barrier. The lower cladding consists of a 200nm-thick In0.52 Al0.48 As layer doped with Si at 1 × 1018 cm−3 , which is lattice matched to the InP substrate. The upper cladding and contact layers are 1700-nm-thick In0.52 Al0.48 As and 150nm -hick In0.53 Ga0.47 As, respectively. Both layers are doped with Be at 2 × 1018 cm−3 . Fig. 3(b) and (c) present the atomic force microscope (AFM) image and the cross-sectional transmission electron microscopy (TEM) images taken under the (0 0 2) dark field projection from the as-grown sample, respectively. The composition-sensitive TEM images reveal dislocation-free flat InAs elongated islands (the brightest area), shaped line truncated pyramids with a relatively homogeneous height. From an AFM, the dash or shortwire is preferentially elongated along the [0 1 1] direction with an average height of 3.2 nm, and the base varied from 30 to 60 nm. The linewidth of dash is in the 15–22 nm range extracted from the TEM image along the [1 1 0] direction. To probe the lateral carrier confinement effect in InAs Qdash nanostructure, cryogenic photoluminescence (PL) was performed at 77 K using a 532 nm diode pumped solid-state laser as an excitation source with varying optical density. As comparison, InAs Qdot embedded in InP matrix was grown using similar system, and the PL spectra were also measured. The ground state (GS) PL peak emission is longer in Qdot due to the InP matrix with larger bandgap energy than InAlGaAs confining layers in Qdash. As shown in Fig. 4(a), the Qdash shows less-resolved quantized states (E0 to E4 ) with a narrower energy separation between E0 and E1 (∆E = 30 meV), if compared to Qdot characteristics (up to E4 with ∆E = 34 meV)
in Fig. 4(b). At a similar optical excitation density, fewer states are excited in dash than dot as a manifestation of the reduced density of states (DOS) in Qdash subbands. At high-excitation density (1500 W/cm2 ), a large number of minima in Qdash spectra are populated, resulting a broad emission line, while in PL spectra of Qdot, the individual minima are more apparent. These properties corroborate the quasi-continuous interband transition characteristics in Qdash. We performed the dielectric cap annealing technique to induce selective intermixing using 200-nm-thick SiO2 and Six Ny layers deposited by using plasma enhanced chemical vapor deposition system. For the ion-implantation process, instead of utilizing As and P ions for implantation [16], the neutral nitrogen species was used in our experiment. Being the lightest and electrical neutral species in group-V elements, nitrogen is expected to generate minimum residual damage in the material, compared with other heavier group-V species, and enable the introduction of implantation-induced defects spatially close or around the active region in the typical p-i-n laser structure to reduce the requirement of annealing temperature Ta . Further, the light damage introduced by nitrogen ions can prevent the generation of loop clusters that usually produces extended defects after intermixing, as for the case of implantation using large atomic size species [9]. N implantations were performed at room temperature using a dynamitron accelerator. N ions were accelerated to 1500 keV with the ion angle tilted by 7◦ to introduce the peak defect density center at the active region of the unmasked (bare) Qdash laser structure. The samples were annealed in nitrogen ambient for 2 min at temperatures from 650 ◦ C to 850 ◦ C using a rapid thermal processor. Broad area lasers with 50-µm-wide oxide stripes were fabricated from the as-grown, annealed, and intermixed Qdash samples. In order to maximize the gain [25], the optical cavity of both lasers is aligned along [0 1 1] orientation and is perpendicular to the dash direction. The current injection was performed at room temperature under pulsed operation (10 µs pulse width and 1% duty cycle). IV. RESULTS AND DISCUSSION A. Thermal Stability and Defect Annealing of Qdash The thermal annealing process significantly affects the bandgap energy of quantum-confined heterostructures related to the interdiffusion of constituent atoms in QWs [1]–[9] as well as in Qdot [11]–[19]. The latter effect becomes more prominent
1242
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 14, NO. 4, JULY/AUGUST 2008
Fig. 5. (a) PL spectrum and (b) PL peak shift at 77 K taken under an excitation density of 3 W/cm2 , when the PL signal from ground state of Qdash samples is not saturated, versus annealing temperature.
Fig. 6. (a) Excitation power-density-dependent PL spectra at 77 K. (b) L–I characteristics taken from the as grown (dashed line) and annealed (solid line) Qdash structures at 700 ◦ C for 2 min. The inset shows the schematic illustration of 50 × 1000-µm-wide oxide stripe lasers with [011] cavity orientated perpendicular to the dash direction.
due to the 3-D nature of nanostructures even at moderate temperature. In the intermixing process, the thermal treatment plays fundamental roles to: 1) promote the interdiffusion spatially and 2) anneal the crystal damage resulting from the introduction of localized point defects in the case of ion implantation. The study of annealing behavior gives information about the Qdash thermal stability, which affects the thermal budget of subsequent device processing, and a baseline annealing parameters from which the effects of spatially enhanced intermixing can be implemented. Fig. 5(a) gives the evolution of PL spectra from annealed Qdash sample, and Fig. 5(b) summarizes the PL peak shift from GS transition at various annealing temperatures. Above 700 ◦ C, the PL peak shift increases linearly as the temperature rises, indicating increased of thermal-induced group-III intermixing effect. The PL linewidth narrowing effect, due to the improved inhomogeneity, has also been observed from the intermixed samples. At 850 ◦ C, the blueshift as large as 180 nm is observed from the Qdash sample. The annealing temperature of 700 ◦ C is selected as the optimum intermixing condition as minimum bandgap shift is produced (less than 20 nm) at the control (nonintermixed) area. The evolution of PL spectra at different excitation densities is depicted in Fig. 6(a) for as-grown and annealed samples. The PL spectra are broadened with increasing optical excitation densities for both samples as a result of significant emission contributed from the excited states. Compared
Fig. 7. (a) Threshold current density Jth of as-grown and annealed Qdash lasers against the inverse of cavity length 1/L. The line is a linear fit yielding Jth at the infinite cavity length J∞ . (b) Reciprocal external efficiency 1/η e x t as a function of the cavity length L for both lasers.
to the as-grown samples, the annealed samples give 30%–40% stronger integrated PL intensity. On the contrary, annealing of the AlInGaAs QW structure (that has similar structure as Qdash active region but absence of the dash layer) results in a reduction in the PL intensity that might be associated with the optical quality degradation of the QW (not shown here). The increase in radiative efficiency as indicated by the improvement of the integrated PL signal is expected to have a significant impact on the device characteristics [25]. Fig. 6(b) compares the light–current (L–I) curves at 20 ◦ C from the asgrown and annealed Qdash lasers. The annealed laser exhibits a comparable slope of efficiency (∼0.12 W/A) and a lower threshold current density Jth at 1.28 kA/cm2 than the as-grown laser (Jth = 1.8 kA/cm2 ). In Fig. 7(a) and (b), the laser characteristics with different cavity lengths are summarized. From the linear fit of Jth versus 1/L, the current density at infinite length (J∞ ) can be extracted to be 1.32 and 0.76 kA/cm2 for as-grown and annealed Qdash lasers, respectively. This corresponds to a significant reduction in J∞ of 73% after the annealing process. The internal quantum efficiency ηint and the internal optical loss αi are obtained from the y-intercept and the slope of the 1/ηext versus L plot in Fig. 7(a), respectively. The linear extrapolation using the equation 1/ηext = 1/ηint (1 − αi L/ln R) yields an improved ηint from 91% (as-grown) to 93% (annealed laser), while the slope of the curves result in the equal internal absorption loss (αi ) of 10.5 cm−1 for both lasers. The reflectivity (R) of the cleaved facets is assumed to be 0.33. The comparable internal loss implies that the scattering loss of photons due to local undulation of refractive index at the dash/well interface and the doping profile during heating are not altered significantly by the annealing at 700 ◦ C. Fig. 8 compares the laser emission spectra from both lasers at an injection level of 1.1 × Jth at different cavity lengths. Both lasers exhibit GS state lasing for device with >1 mm cavity length. The lasing line progressively shift toward a shorter wavelength as the cavity length decreases as a result of the band-filling effect in the localized subband of the GS state energy. The spectral linewidth of the as-grown laser varied from 30 to 60 nm and contains multiple modes. However, the
SUSANTO DJIE et al.: QUANTUM DASH INTERMIXING
Fig. 8. Normalized laser spectra for the following with different cavity lengths at 1.1 × Jth . (a) As-grown. (b) Annealed Qdash lasers.
spectral linewidth is substantially narrowed and the number of lasing modes is lower for the annealed laser. A small wavelength blueshift of ∼15 nm, attributed to the group-III intermixing due to the Ga vacancies outdiffusion from Qdash surface to SiO2 layer [19], is observed from the annealed QDash laser. Correlating the internal loss and the laser emission at variable lengths in Fig. 7(b), we can deduce the total modal gain gm o d of GS level as high as 31 cm−1 from these four-stack Qdash lasers with short cavity length of 0.54 mm. The high gain might be attributed to the role of strain-compensated InAlGaAs barriers that enhances the material gain from large overlaps in the electron and hole wavefunctions [26]. The high density of grown-in defects, which is below our TEM resolution limit, might be present near dash-in-well interfaces due to the low growth temperature in MBE [27] and the increased stress field around the high-density islands with the insertion of QW and barrier [28]. These defects serve as effective channels for nonradiative carrier recombination and reduce the PL efficiency. Upon annealing, the defect density is reduced; hence, resulting in an increase in the internal quantum efficiency of the annealed Qdash lasers in addition to the substantial reduction of the threshold current density and laser linewidth [25]. B. Optical and Material Properties of Intermixed Qdash In this section, we investigate two major intermixing techniques that have been implemented successfully in QW materials. These two techniques are: 1) IFVD using dielectric cap annealing [4], [21] and 2) IID using neutral species [9]. The first technique involves the deposition of a dielectric cap material on the QW materials, and subsequent high-temperature annealing to promote the injection of vacancies from the dielectric cap to materials, and hence, enhance the intermixing at selected area. For instance, in GaAs/AlGaAs QWs and InGaAs/GaAs QWs structures, SiO2 induces out-diffusion of Ga atoms during annealing and generates group-III vacancies in the system [14]. Fig. 9 depicts clearly the role of dielectric caps to obtain different degrees of intermixing at the selected area for SiO2 and Six Ny caps. For typical GaAs/AlGaAs QWs and InGaAs/GaAs QWs, SiO2 cap enhances the intermixing degree, while the Six Ny cap suppresses the intermixing that the combination of those layers can be used to create two-bandgap QWs across a
1243
Fig. 9. PL spectra measured at 77 K from SiO2 capped and Six Ny capped (a) GaAs/AlGaAs QWs, (b) InGaAs/GaAs QW s, and (c) In0 . 6 4 Al0 . 2 Ga0 . 1 6 As/In0 . 5 0 Al0 . 1 8 Ga0 . 3 2 As QWs after annealing at 900 ◦ C (2 min) except at 800 ◦ C (0.5 min) for InAlGaAs QWs. The annealing conditions were chosen at the highest possible annealing temperature to produce a high selectivity, which a given QW structure, its bandgap experiences a minimum thermal shift (i.e., less than 5 meV).
semiconductor sample with a single annealing step. However, we note an opposite observation for the case of InAlGaAs QWs, which SiO2 acts as intermixing suppressor in the samples. The differential bandgap shift obtained between SiO2 and Six Ny capped samples is ∼13 nm. The peculiar behavior might be attributed to the disparity of group-III atomic diffusion in InAlGaAs system, which involves the exchange of In, Ga, and Al species. As Al atoms have the lowest diffusivity in the system and the largest bonding strength (AlAs: 62 kcal/mol, Ga–As: 47.7 kcal/mol, and InAs: 36 kcal/mol), the Al profiles are not significantly altered during interdiffusion [29]. Ga atoms are known to have a much higher solubility in SiO2 than in Six Ny , leading to an efficient Ga vacancy injection VGa to the GaAs/AlGaAs QW material to promote atomic intermixing [14]. Other factors such as strain/stress, charge state of point defects, grown-in defects, type of semiconductor surface material, dielectric/semiconductor interface conditions. etc., might also play significant roles in promoting atomic intermixing in III–V material systems [4]. By considering only the dominant effect of Ga vacancy diffusion, we postulate that In has a smaller diffusion coefficient than Ga (i.e., DGa > DIn ) in the SiO2 -capped sample [24]. By contrast, the injection of Ga vacancies is negligible under the Six Ny cap due to the less solubility of Ga atoms in the cap layer. The thermal energy from annealing will solely drive the dominant In diffusion (DGa < DIn ) in the Six Ny -capped sample. Using our model reported earlier [19], we therefore propose that the enhanced intermixing/larger blueshift under Six Ny capping layer could be related to the dominant In diffusion with respect to the Ga diffusion [Fig. 2(c)]. The intermixing suppression in bare (uncapped) samples indicates the possible precipitation of group-III atoms from the sample surface during annealing with GaAs proximity capping We further performed the same IFVD process to Qdash samples, and the PL shift versus the annealing temperature is summarized in Fig. 10. Both the control and the SiO2 capped samples begin to experience a thermal shift (over 10 nm) at an annealing temperature of 700 ◦ C for 2 min. The intermixing enhancement with Six Ny cap is well reproduced in the InAs/InAlGaAs dash-in-well system with larger degree of
1244
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 14, NO. 4, JULY/AUGUST 2008
Fig. 10. Peak shift of 77 K PL against annealing temperature T a using the following. (a) Dielectric cap annealing: SiO2 cap ( ) and Six Ny cap ( ). (b) N implantation at doses of 5 × 101 2 cm−3 (), 5 × 101 3 cm−3 (䉬), and 5 × 101 4 cm−3 (•). The dotted line indicates the activation temperature T a required to initiate the spatially selective intermixing.
intermixing, i.e., the differential bandgap shift of 78 nm at the intermixing activation temperature Ta = 700◦ C (2 min), than the QW system. The low Ta requirement is advantageous as a large bandgap selective process can be developed without inducing significant thermal shift in the control region. We further compare the IFVD process with NIID process to the same Qdash structure. Fig. 10 presents the PL peak shift as a function of the annealing temperature after N implantation to bare surface samples at doses of 5 × 1012 , 5 × 1013 , and 5 × 1013 ions/cm2 , respectively. The annealing activation is lower at 650 ◦ C for NIID cases, and the corresponding blueshifts at Ta are 112 nm for dose of 5 × 1012 and 5 × 1013 ions/cm2 , respectively. The bandgap shift reduces slightly to 108 nm for Qdash sample implanted with 5 × 1014 ions/cm2 . The shift reduction at high-dose implantation could be attributed to the formation of clusters that trapped point defects during annealing treatment, and thus effectively reduced the defect concentration. At any doses of NIID, the differential bandgap shift is larger than the previous IFVD result. To correlate the defect migration range and the intermixing efficiency, we investigate the NIID effect from Qdash samples related to the placement of implantation-induced damage into semiconductor samples. Fig. 11(a) compares the integrated PL intensity using the direct NIID (implantation to unmasked samples) and the indirect NIID (implantation to the 300-nm-thick SiO2 protected samples) processes with varying doses to spatially situate the damage to the center of the active region and above the active region, respectively. The SiO2 mask was removed using buffered oxide (BOE) etch solution after implantation to eliminate the IFVD effect during annealing. In the indirect NIID, the damage is separated from the active region. Only a negligible concentration of vacancy (i.e., 1.15 × 1017 cm−3 ) is located in the active region, which corresponds to less than 1% of calculated vacancies using direct NIID. This is further confirmed by the presence of weak PL signals exhibiting similar shape and peak positions to the as-grown samples immediately after implantation. On the other hand, the PL signals after direct implantation could not be detected.
Fig. 11. (a) Integrated PL intensity as a function of the annealing temperature after NIID from the direct (uncapped; solid lines) and the indirect implantation (capped with 200-nm-thick SiO2 layer; dashed lines) at different implant doses. The dashed line represents the integrated PL intensity from the as-grown sample. (b) PL linewidth against the annealing temperature for both direct and indirect implanted Qdash. The dashed and dotted lines represent the linewidth from as-grown and annealed-only samples, respectively.
Upon annealing, the PL signal from indirect NIID recovers to at least a similar level to the as-grown sample for samples implanted with all doses of nitrogen, while the PL signal from direct NIID exhibits weaker PL intensity. This observation validates that the following annealing, the residual defect in the form of nonradiative recombination centers generated by using the indirect NIID is minimum, and the high optical quality is retained. It should be noted that PL peak shifts for indirect NIID are ∼112 nm at annealing temperature of 650 ◦ C and 700 ◦ C, and these blueshifts are independent of the implant dose in the interested range (not shown here). By introducing the remote damage, the blueshift at a dose of 5 × 1014 ions/cm2 from the indirect NIID is considerably larger than that of the earlier result from the direct NIID in Fig. 10(b). These facts further suggest that: 1) the Qdash intermixing is governed solely by the indiffusion of defects placed further away from the active region and 2) the indirect NIID can be utilized to effectively minimize the formation of damage clusters and complexes in the Qdash region that might degrade the Qdash optical quality after intermixing. The linewidth of both direct and indirect implanted Qdash remain at about the same level as the as-grown sample after annealing at up to Ta = 800 ◦ C [Fig. 11(b)]. In all cases, the linewidth of annealed-only samples is wider than the intermixed samples. The intermixing effect effectively compensates the PL narrowing effect due to the reduced size dispersion in QW as intermixing proceeds, causing the invariant PL linewidth after intermixing at annealing below 800 ◦ C. Fig. 12(a) depicts the intermixing degree and the concentration of created vacancies as a function of implant mask thickness, hence the vacancy peak distance from Qdash active region, after NIID process at a dose of 5 × 1012 ions/cm2 . A SiO2 layer was chosen as the implant mask up to 1 µm in thickness to control the amount of defect concentration in the samples, and the photoresist (PR) layer was subsequently applied to increase the total thickness of the mask. The distance of the vacancy peak from the active region increases linearly with the increasing of implant mask thickness from 0 to 3 µm. Upon annealing at 700 ◦ C, blueshifts from 108 to 104 nm are observed for
SUSANTO DJIE et al.: QUANTUM DASH INTERMIXING
Fig. 12. (a) PL peak shift from the intermixed Qdash (dose = 5 × 101 2 ions/cm2 , T a = 700 ◦ C) and the calculated vacancies against the implant mask thickness, hence the depth of vacancy peak. (b) PL spectra of InAlGaAs QWs and InAs/InAlGaAs Qdash partial structures at 77 K from the as-grown, control, or annealed-only under Six Ny cap, and intermixed samples. The NIID process was carried out using the N implant dose of 5 × 101 2 cm−2 to the 1-µm-thick SiO2 protected samples and annealing temperature of 700 ◦ C. The insets show the QW and QDs partial epilayer structures and the corresponding bandgap energy.
the intermixed samples with implant thickness varying from 0 to 2 µm. The result substantiates that the interdiffusion is enhanced with the presence of implantation-induced defects, and these created defects from N implantation are highly mobile in the In(AlGa)As material as opposed in (Al)GaAs system [17], while the location of defects is not critical. It is worth noting that the PL intensity level of different N doses in indirect NIID remains comparable [Fig. 12(a)]. Fig. 12(b) compares the intermixing effectivity between partial QWs and Qdash structures after NIID at a dose of 5 × 1012 ions/cm2 and Ta at 700 ◦ C (2 min). In these partial structures, all samples were masked with a 1000-nm-thick SiO2 layer prior to implantation to allocate the damage peak around Qdash region. At Ta = 700 ◦ C, the annealed-only bare samples exhibited a negligible thermal shift, while the implanted samples are appreciably blueshifted. The differential wavelength shifts between annealed-only and implanted samples are 18 nm and 160 nm for QW and Qdash samples, respectively. As expected, this pronounced Qdash intermixing results from the large surface to volume ratio, if compared to QW structure. In addition, the intermixed Qdash exhibits the reduced PL linewidth, indicating the improved homogeneity in dash nanostructures. To evaluate the microstructural changes after NIID, we perform the cross-sectional TEM measurement for intermixed Qdash laser structures under (0 0 2) dark-field projection. Fig. 13 compares the (0 0 2) dark-field TEM images along [0 1 1] direction for as-grown and intermixed samples. Qdash remain well separated with improvement in the uniformity of the effective size distribution, especially the dash height that plays significant role in determining the Qdash transition energy. The dash interface becomes less distinctive after intermixing due to the enhanced group-III diffusion in a 3-D direction. No dislocation was observed from both samples, noting the absence of dislocations introduced during N implantation subjected to annealing process. Consistent with this picture, the indirect NIID is favorable as it allows a nearly complete recovery of the optical efficiency with the absence of dislocation under our TEM reso-
1245
Fig. 13. [0 1 1] cross-sectional TEM images under (0 0 2) dark-field contrast revealing the chemical contrast of InAs/InAlGaAs dash-in-well structure using NIID (dose = 5 × 101 2 ions/cm2 , T a = 700 ◦ C).
Fig. 14. (a) LI characteristics and (b) the laser spectra at injection level of 1.1 × Jth from: 1) as-grown 2) control, and bandgap-tuned samples by 3) IFVD and 4) indirect NIID at injection level of J = 1.1 × Jth . The inset shows the schematic illustration of 50 × 1000 µm2 broad area lasers and the energy diagram of intermixed and nonintermixed Qdash.
lution while giving an adequate intermixing degree suitable for active device fabrication. C. Bandgap-Tuned Qdash Lasers Bandgap-tuned lasers were fabricated and characteristics from samples intermixed with both IFVD and NIID. The broad area lasers with similar fabrication process, oxide stripe geometry, and cavity orientation with respect to the dash orientation as in Section IV-A were fabricated and tested under pulse condition (10 µs pulse width and 1% duty cycle) from the as-grown, control (SiO2 capped), bandgap-tuned samples by IFVD (Six Ny cap annealing), and NIID (N dose: 5 × 1012 cm−2 , 1-µm-thick SiO2 implant mask, annealing: 700 ◦ C) [30], [31]. Fig. 14(a) and (b) depicts the L–I characteristics and the lasing spectra from fabricated Qdash lasers measured at 20 ◦ C. The as-grown and control lasers exhibit a threshold current density Jth of 1.80 and 1.28 kA/cm2 , while the bandgap-tuned lasers by IFVD and NIID processes start to lase at 1.28 and 1.16 kA/cm2 , respectively. All lasers show comparable slope efficiency (∼0.12 W/A). The summary of laser performances is tabulated in Table I. The notable reduction in Jth from the control lasers with respect to the as-grown lasers results from
1246
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 14, NO. 4, JULY/AUGUST 2008
TABLE I SUMMARY OF FABRICATED LASER PERFORMANCE FROM AS-GROWN, CONTROL, IFVD TUNED-, AND NIID TUNED- QDASH LASERS WITH CAVITY LENGTH OF 1 MM
of the intermixed material can be well preserved; hence, the realization of high performance Qdash devices is attainable, and it can provide an alternative cost-effective technology for future monolithic integration of Qdash devices on a single chip. V. CONCLUSION
the removal of grown-in defects that act as nonradiatiave recombination traps/centers during thermal treatment [25]. While lasers intermixed by IFVD using Six Ny cap exhibits a comparable Jth , the laser intermixed by NIID shows a lower Jth . The measured serial resistance from control and intermixed lasers by IFVD and NIID are 0.75 and 0.94, and 0.63 Ω, respectively, while the as-grown operates with a serial resistance of 0.56 Ω. The increase in serial resistance from processed lasers with respect to the as-grown laser is related to the thermal annealing and/or ion implantation process. The increase from both dielectric capped samples is attributed to the dopant depletion at the highly doped contact layer due to the outdiffusion process across semiconductor-dielectric interface at an elevated temperature. Further increased resistance in IFVD-tuned laser, which is fabricated from the Six Ny capped sample, is ascribed to the imperfect removal of the densified Six Ny layer using the BOE solution. This issue can be addressed by utilizing the combination of low-damage plasma etch and short BOE dip to completely remove the annealed Six Ny layer; hence, preserving the surface morphology for device fabrication process. This deleterious effect associated with the dopant profile alteration is minimized in the case of indirect NIID as the desirable defects responsible for intermixing are situated near the Qdash active region, and the outdiffusion to GaAs proximity capping is negligible if compared to the dielectric cap induced intermixing. With the minimum degradation in electrical properties of laser, the intermixed laser by NIID is expected to exhibit an improved performance if compared to the intermixed device by IFVD. This argument is consistent with our tested results that the intermixed laser by NIID has a lower threshold current if compared to the control and IFVD-tuned lasers. The reduced threshold in the intermixed laser by NIID is, therefore, ascribed to the enhanced Qdash gain characteristics from the improved dash inhomogeneity as corroborated by PL and TEM analysis [30]. The lasing peak emissions from GS level are at 1635, 1618, 1518, and 1506 nm for as-grown, control, and intermixed lasers by IFVD and NIID, respectively. This corresponds to the wavelength tuning of 117 nm and 129 nm for IFVD and NIID processes, respectively. The larger shift attained from NIID process is consistent with the PL measurement earlier [Fig. 10(b)] from the test samples, which NIID gives an enhanced intermixing rate if compared to IFVD process with similar annealing condition. Our results strongly suggest that, with proper optimization of intermixing process, the quality
In summary, N implantation has been performed to promote group-III intermixing in InAs/InAlGaAs dash-in-well structures. Significant improvement in laser performance has been observed from the annealed Qdash lasers. The annealed lasers exhibit lower threshold current densities, narrower laser spectrum, and higher internal quantum efficiency. A lower intermixing activation (i.e., 650 ◦ C) than the IFVD process has been observed from the NIID process. Differential bandgap shift of 112 nm (65 meV) has been measured from the sample implanted with 5 × 1012 cm−2 , and subsequent annealing at 700 ◦ C for 2 min. The indirect or shallow implantation induced damage to Qdash active region promotes an efficient group-III atomic exchange based on long defect migration upon annealing with the absence of residual damage as evidenced from the excellent PL signal. High quality Qdash lasers with bandgap tuned by 125 nm have fabricated from sample intermixed using NIID. Our results indicate a highly attractive wavelength trimming and selective bandgap tuning method, well suited for planar, monolithic Qdash integration of optoelectronics components. REFERENCES [1] J. H. Marsh, “Quantum well intermixing,” Semicond. Sci. Technol., vol. 8, pp. 1136–1155, 1993. [2] S. Charbonneau, E. S. Koteles, P. J. Poole, J. J. He, G. C. Aers, J. Haysom, M. Buchanan, Y. Feng, A. Delage, F. Yang, M. Davies, R. D. Goldberg, P. G. Piva, and I. V. Mitchell, “Photonic integrated circuits fabricated using ion implantation,” IEEE J. Sel. Topics Quantum Electron., vol. 4, no. 4, pp. 772–793, Jul./Aug. 1998. [3] E. J. Skogen, J. S. Barton, S. P. Denbaars, and L. A. Coldren, “A quantumwell-intermixing process for wavelength-agile photonic integrated circuits,” IEEE J. Sel. Topics Quantum Electron., vol. 8, no. 4, pp. 863–869, Jul./Aug. 2002. [4] B. S. Ooi, K. McIlvaney, M. W. Street, A. Helmy, S. G. Ayling, A. C. Bryce, J. H. Marsh, and J. S. Roberts, “Selective quantum well intermixing in GaAs/AlGaAs structures using impurity-free vacancy diffusion,” IEEE J. Quantum Electron., vol. 33, no. 10, pp. 1784–1793, Oct. 1997. [5] S. Yuan, Y. Kim, C. Jagadish, P. T. Burke, M. Gal, J. Zou, D. Q. Cai, D. J. H. Cockayne, and R. M. Cohen, “Novel impurity-free interdiffusion in GaAs/AlGaAs quantum wells by anodization and rapid thermal annealing,” Appl. Phys. Lett., vol. 70, pp. 1269–1271, 1997. [6] O. P. Kowalski, C. J. Hamilton, S. D. McDougall, J. H. Marsh, A. C. Bryce, R. M. De La Rue, B. Vogele, and C. R. Stanley, “A universal damage induced technique for quantum well intermixing,” Appl. Phys. Lett., vol. 72, pp. 581–583, 1998. [7] B. S. Ooi, T. K. Ong, and O. Gunawan, “Multiple-wavelength integration of InGaAs/InGaAsP structures using pulsed laser irradiation induced quantum well intermixing,” IEEE J. Quantum Electron, vol. 40, no. 5, pp. 481–490, May 2004. [8] H. S. Djie and T. Mei, “Plasma induced quantum well intermixing for monolithic photonic integration,” IEEE J. Select. Topics Quantum Electron., vol. 11, no. 2, pp. 373–382, Mar.-Apr. 2005. [9] V. Aimez, J. Beauvais, J. Beerens, D. Morris, H. S. Lim, and B. S. Ooi, “Low-energy ion-implantation induced quantum well intermixing,” IEEE J. Sel. Topics Quantum Electron., vol. 8, no. 4, pp. 870–879, Jul./Aug. 2002.
SUSANTO DJIE et al.: QUANTUM DASH INTERMIXING
[10] D. Bimberg, M. Grundmann, and N. N. Ledentsov, Quantum Dot Heterostructures. Chichester, U.K.: Wiley, 1998. [11] R. Leon, Y. Kim, C. Jagadish, M. Gal, J. Zou, and D. J. H. Cockayne, “Effects of interdiffusion on the luminescence of InGaAs/GaAs quantum dots,” Appl. Phys. Lett., vol. 69, pp. 1888–1890, 1996. [12] D. Bhattacharyya, A. S. Helmy, A. C. Bryce, E. A. Avrutin, and J. H. Marsh, “Selective control of self-organized In0.5Ga0.5As/GaAs quantum dot properties: Quantum dot intermixing,” J. Appl. Phys., vol. 88, pp. 4619–4622, 2000. [13] O. Gunawan, H. S. Djie, and B. S. Ooi, “Electronics states of interdiffused quantum dots,” Phys. Rev. B, vol. 71, pp. 205319-1–205319-10, 2005. [14] H. S. Djie, O. Gunawan, D.-N. Wang, B. S. Ooi, and J. C. M. Hwang, “Group-III vacancy induced InGaAs quantum-dot interdiffusion,” Phys. Rev. B, vol. 73, pp. 155324-1–155324-5, 2006. [15] Y. Wang, H. S. Djie, and B. S. Ooi, “Quantum-confined stark effect in interdiffused quantum dots,” Appl. Phys. Lett., vol. 89, pp. 151104-1– 151104-3, 2006. [16] H. S. Djie, B. S. Ooi, and V. Aimez, “Neutral ion implantation induced selective quantum dot intermixing,” Appl. Phys. Lett., vol. 87, pp. 2611021–261102-3, 2005. [17] H. S. Djie, B. S. Ooi, and O. Gunawan, “Quantum dot intermixing using excimer laser irradiation,” Appl. Phys. Lett., vol. 89, pp. 081901-1– 081901-3, 2006. [18] H. S. Djie, D.-N. Wang, B. S. Ooi, J. C. M. Hwang, X.-M. Fang, Y. Wu, J. M. Fastenau, and W. K. Liu, “Intermixing of InGaAs quantum-dots grown by cycled monolayer deposition,” J. Appl. Phys., vol. 100, pp. 030527-1–030527-6, 2006. [19] Y. Wang, H. S. Djie, and B. S. Ooi, “Group III intermixing in InAs/InAlGaAs quantum dots-in-well,” Appl. Phys. Lett., vol. 88, pp. 111110-1–111110-3, 2006. [20] H. S. Djie, S. L. Ng, O. Gunawan, P. Dowd, V. Aimez, J. Beauvais, and J. Beerens, “Analysis of strain-induced polarisation insensitive integratedwaveguides fabricated using ion implantation-induced intermixing,” in Proc. Inst. Elect. Eng. Optoelectron., 2002, vol. 149, pp. 138–144. [21] H. S. Djie, C. K. F. Ho, T. Mei, and B. S. Ooi, “Quantum well intermixing enhancement using Ge-doped sol-gel derived SiO2 encapsulant layer in InGaAs/InP laser structure,” Appl. Phys. Lett., vol. 8, pp. 081106-1– 081106-3, 2005. [22] H. S. Djie, T. Mei, and J. Arokiaraj, “GaAs/AlGaAs quantum well intermixing using high-density Argon plasma,” Semicond. Sci. Technol., vol. 20, pp. 244–249, 2005. [23] T. K. Ng, H. S. Djie, S. F. Yoon, and T. Mei, “Thermally induced diffusion study of GaInNAs/GaAs and GaInAs/GaAs quantum well structure grown by solid source molecular beam epitaxy,” J. Appl. Phys., vol. 97, pp. 013506-1–013506-8, 2005. [24] Y. Wang, H. S. Djie, and B. S. Ooi, “Interdiffusion in InGaAsSb/AlGaAsSb quantum wells,” J. Appl. Phys., vol. 98, pp. 073508-1–073508-7, 2005. [25] H. S. Djie, Y. Wang, B. S. Ooi, D. N. Wang, J. C. M. Hwang, G. T. Dang, and W. H. Chang, “Defect annealing of InAs/InAlGaAs quantum-dashin-asymmetric-well laser,” IEEE Photon. Technol. Lett., vol. 18, no. 22, pp. 2329–2331, Nov. 2006. [26] Y. Matsui, H. Murai, S. Arahira, Y. Ogawa, and A. Suzuki, “Novel design scheme for high-speed MQW lasers with enhanced differential gain and reduced carrier transport effect,” IEEE J. Quantum Electron., vol. 34, no. 12, pp. 2340–2349, Dec. 1998. [27] T. Kawai, H. Yonezu, Y. Ogaswara, D. Saito, and L. Park, “Segregation and interdiffusion of In atoms in GaAs/InAs/GaAs heterostructures,” Appl. Phys. Lett., vol. 74, pp. 1770–1772, 1993. [28] H. Y. Liu, I. R. Sellers, T. J. Badcock, D. J. Mowbray, M. S. Skolnick, K. M. Groom, M. Gutierrez, M. Hopkinson, J. S. Ng, J. P. R. David, and R. Beanland, “Improved performance of 1.3 µm multilayer InAs quantumdots lasers using a high-growth-temperature GaAs spacer layer,” Appl. Phys. Lett., vol. 85, pp. 704–706, 2004. [29] R. J. Baird, T. J. Potter, G. P. Kothiyal, and P. K. Bhattacharya, “Indium diffusion in the chemical potential gradient at an In0.53Ga0.47As/In0.52Al0.48As interface,” Appl. Phys. Lett., vol. 52, pp. 2055–2054, 1988. [30] H. S. Djie, Y. Wang, B. S. Ooi, D.-N. Wang, J. C. M. Hwang, Y. Wu, X.-M. Fang, J. M. Fastenau, W. K. Liu, G. T. Dang, and W. H. Chang, “Wavelength tuning of InAs/InAlGaAs quantum-dash-in-well laser using postgrowth intermixing,” Electron. Lett., vol. 43, pp. 33–35, 2007. [31] H. S. Djie, Y. Wang, D. Negro, and B. S. Ooi, “Postgrowth wavelength engineering of InAs/InAlGaAs quantum-dash laser,” Appl. Phys. Lett., vol. 90, no. 3, pp. 031101-1–031101-3, 2007.
1247
Hery Susanto Djie (S’00–M’04) received the B.Eng. degree (first class honors) in electrical engineering from Pelita Harapan University, Banten, Indonesia, in 1999, and the Ph.D. degree from the Electrical Engineering Department, Nanyang Technological University, Singapore, in 2004. From 2002 to 2003, he was a Research Staff Member of the Photonics Research Center, Nanyang Technological University, where he was engaged in developing the widely tunable semiconductor laser for the phased-array-antenna-based true time delay system. He later joined Agilent Technologies, Singapore, where he was in charge of the development of high-brightness light-emitting diodes. From 2004 to 2007, he was a Research Scientist with the Center for Optical Technologies, Lehigh University, Bethlehem, PA, where he was a key contributor in the technology and development of quantum-dot bandgap engineering and semiconductor broadband emitter. Since February 2007, he has been with JDS Uniphase Corporation, San Jose, CA, where he is engaged in high-volume manufacturing of high-power laser diode, high-speed semiconductor laser, widely tunable laser, optical adddrop multiplexer, and photonic integrated circuits. He is the author or coauthor of more than 150 international technical papers and several U.S. patents pending in the field of monolithic integration of photonic devices, quantum-dot and quantum-well intermixing, semiconductor laser, and broadband semiconductor light source. Dr. Djie is a member of the International Society for Optical Engineering (SPIE), the Material Research Society (MRS), and Sigma Xi Scientific Research Society.
Yang Wang received the B.S. degree from the University of Science and Technology of China, Hefei, China, in 2002, the M.S. degree from the National University of Singapore, Singapore, in 2003, and the Ph.D. degree in electrical engineering from Lehigh University, Bethlehem, PA, in 2007. Since September 2003, he has been a Research Associate in the Center for Optical Technologies, Lehigh University, Bethlehem, PA. His research focused on the realization of monolithic photonic integration in semiconductor nanostructures using various postgrowth intermixing techniques, and also included design, fabrication, and characterization of novel InP-based quantum dash photonic devices. He is currently an Optoelectronic Device Engineer with OptiComp Corporation, Zephyr Cove, NV. He has authored or coauthored over 25 international technical papers. Dr. Yang is a member of the International Society for Optical Engineering (SPIE), the Material Research Society (MRS), and Sigma Xi.
Yun-Hsiang Ding received the B.S. degree in electrophysics from National Chiao Tung University, Hsinchu, Taiwan, in 2000, and the M.S. degree in electrical engineering from The George Washington University, Washington, DC, in 2005. She is currently working toward the Ph.D. degree in electrical engineering at Lehigh University, Bethlehem, PA. From 2000 to 2002, she was a full-time Teaching Assistant at the National Chiao Tung University. Her current research interests include fabrication and characterization of high-power semiconductor lasers, quantum well, and quantum dash intermixing process. Miss Ding has been an officer of the International Society for Optical Engineering Lehigh Student Chapter since 2006.
1248
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 14, NO. 4, JULY/AUGUST 2008
Dong-Ning Wang received the Ph.D. degree in materials science and engineering from Lehigh University, Bethlehem, PA, in 1998. She worked as a Characterization and Failure Model Analysis Engineer for quality control and yield improvement in Lucent/Agere Optoelectronic Center, Breinigsville, PA. She is currently a Research Scientist at the Compound Semiconductor Technology Laboratory, Lehigh University. Her current research interest includes implementing transmission electron microscopy (TEM), low-voltage scanning electron microscopy (SEM), and focus ion beam (FIB) to characterize compound and Si semiconductors in microelectromechanical systems (MEMs) and integrated photonic devices. She specializes on combining various analytical tools including low-voltage SEM, FIB, TEM, Auger electron spectroscopy (AES), security information management (SIMS), and XML paper specification (XPS) on III–V materials, devices, and subsystems to evaluate the quality and to conduct failure mode analysis.
James C. M. Hwang (M’81–SM’82–F’94) received the B.S. degree in physics from the National Taiwan University, Taipei, Taiwan, R.O.C., in 1970, and the M.S. and Ph.D. degrees in material science and engineering from Cornell University, Ithaca, NY, in 1973 and 1976, respectively. After 12 years of industrial experience at IBM, AT&T, GE, and GAIN, he joined the faculty of Lehigh University, Bethlehem, PA, in 1988, as a Professor of electrical engineering and the Director of the Compound Semiconductor Technology Laboratory. In 2002, he helped establish the Center for Optical Technologies between Lehigh University and Pennsylvania State University and served as its interim Director. In 2006, he helped initiated the Investigate Multi-Physics Modeling and Performance Assessment-Driven Characterization and Computation Technology (IMPACT) Center for Micro-Electromechanical Systems/Nano-Electromechanical Systems (MEMS/NEMS) Very Large-Scale Integration (VLSI) between Lehigh University, University of Illinois, Purdue University, and Georgia Institute of Technology, and led one of the major tasks. He has been a Nanyang Professor at Nanyang Technological University, Singapore, and an Advisory Professor at Shanghai Jiao Tong University, Shanghai, China. He cofounded GAIN and QED. He has authored or coauthored more than 200 technical papers and holds four U.S. patents.
Xiao-Ming Fang received the B.S. degree in physics from the University of Hangzhou, Hangzhou, China, in 1981, and the M.S. and Ph.D. degrees in semiconductor physics from Shanghai Institute of Technical Physics, Chinese Academy of Sciences, Shanghai, China, in 1986 and 1990, respectively. In 1992, he was with the University of Pittsburgh, Pittsburgh, PA, as a Post-doctoral Researcher. In 1994, he was with the University of Oklahoma, Norman. Since 1999, he has been with IQE, Inc. (then QED), Bethlehem, PA, as a Senior Research Engineer.
Ying Wu (M’97–SM’07) received the B.S. degree in electrical engineering from Southeast University, Nanjing, China, in 1985, and the M.S. degrees in electrical engineering from Nanjing Electronic Devices Institute (NEDI), Nanjing, Jiangsu, China, and the University of Cincinnati, OH, in 1988 and 1996, respectively. From 1988 to 1994, she worked on design and development of AlGaAs/GaAs High Electron Mobility Transistor (HEMT) and Heterojunction Bipolar Transistor (HBT) Devices and Circuits at NEDI, as a Research Engineer then Group Leader for HBT Projects. In 1996, she joined IQE Inc. (then QED), Bethlehem, PA, as a Senior Research Engineer, where she established quick turn-around device processing capability. She is currently a Material and Device Engineering Manager and in charge of Material Characterization Department at IQE Inc. (then QED). She is the coauthor of more than 50 refereed journal and conference publications.
Joel M. Fastenau received the B.S. and M.S. degrees from Iowa State University, Ames, in 1991 and 1993, respectively, and the Ph.D. degree from Colorado State University, Ft. Collins, in 1999, all in electrical engineering. He was a Postdoctoral Researcher at Colorado State University for one year. In 2000, he joined the Engineering Staff, IQE, Inc., Bethlehem, PA, where he is engaged in the commercial III–V semiconductor epiwafer production by molecular beam epitaxy (MBE), and is currently an MBE Engineering Manager and directs the engineers responsible for all growth processes including both high-volume products and specialized R&D programs. His current research interests include MBE growth of quantum dot optoelectronic structures, high-speed transistors in InP, and next-generation integration of III–V’s with Si architecture. He has authored or coauthored more than 60 refereed journal and conference publications.
Amy W. K. Liu (S’87–M’91–SM’03) received the B.Sc. degree in electrical engineering (with honors) and the Ph.D. degree in physics from the Imperial College, London, U.K., in 1986 and 1991, respectively, and the M.Sc. degree in microwave and modern optics from the University College, University of London, London, in 1987. She was a Postdoctoral Research Assistant at the U.K. Interdisciplinary Research Centre for Semiconductor Materials. In 1996, she became a member of the Research Staff of the University of Oklahoma, where she was a Postdoctoral Fellow for two years and worked on narrow-gap semiconductor materials. In 1997, she joined IQE, Inc. (then QED), Bethlehem, PA, as a Senior Research Engineer and is now the Director of Research and Development. She has 20 years of experience in the molecular beam epitaxy (MBE) growth and characterization of semiconductor materials. She has authored or coauthored more than 200 journal and conference publications, and has also coedited one book. Her current research interests include MBE growth of low-dimensional structures such as quantum dots and dashes for emitter and detector applications, InP-based transistor structures for high-speed electronic applications, and integration of III–V compound semiconductor with Si. Dr. Liu is a Member of the American Institute of Physics and the Institute of Physics (U.K.), and a Chartered Physicist (U.K.).
SUSANTO DJIE et al.: QUANTUM DASH INTERMIXING
Gerard T. Dang received the B.S. degree from the University of California, Los Angeles, and the Ph.D. degree from the University of Florida, Gainesville, in 1998 and 2001, respectively, both in chemical engineering. In 2000, he was a Visiting Scientist at Bell Laboratories, Lucent Technologies, Murray Hill, NJ, where he designed, developed, and characterized high speed vertical-cavity surface-emitting lasers. He has experience in the design, fabrication, and characterization of gallium nitride electronic devices and myriad arsenide-based emitters, and detectors. Since 2002, he has been a Scientist in the Sensors and Electron Devices Directorate at the U.S. Army Research Laboratory, Adel-phi, MD, where he has been currently developing gallium arsenide-based photonic bandgap waveguide devices with microelectromechanical systems (MEMs) features.
Wayne H. Chang (M’80) received the B.S. degree from the National Taiwan University, Taipei, Taiwan, R.O.C., in 1965, and the M.S. degree from the University of Maryland, College Park, in 1976, both in chemical engineering. In 1976, he joined Comsat Laboratories, Clarksburg, MD, to work on the GaAs devices, as well as microwave integrated circuits, monolithic microwave integrated circuits, and optical integrated circuits. He joined the Adelphi Laboratory Center, U.S. Army Research, Adelphi, MD, in 1989, where he has been working on optoelectronic devices and device fabrication technology. He is currently the Team Leader of vertical cavity surface-emitting laser (VCSEL) and Optical Interconnect, with responsibility for research and development on the related projects.
1249
Boon S. Ooi (M’94–SM’04) received the B.Eng. and Ph.D. degrees in electronics and electrical engineering from the University of Glasgow, Glasgow, U.K., in 1992 and 1994, respectively. From 1996 to 2000, he was an Assistant Professor at Nanyang Technological University, Singapore. During 2000–2003, he was the Vice President of Technology of Phosistor Technologies, Inc. In August 2003, he joined Lehigh University, Bethlehem, PA, as an Associate Professor. His current research interests primarily include the development of monolithic integration processes for semiconductor photonic integrated circuits using quantum-well/dash/dot intermixing, and the development of broadband emitters such as superluminescent diode and broadband diode laser for communications, sensing, and imaging applications. He has authored or coauthored over 200 journal and conference papers. Dr. Ooi is a Fellow of the Institute of Physics (U.K.).