ZnSe/ZnS Quantum-Dot Semiconductor Optical Amplifiers

Transcription

1 Optics and Photonics Journal, 011, 1, doi:10.436/opj Published Onle June 011 ( ZnSe/ZnS Quantum-Dot Semiconductor Optical Amplifiers Abstract K. H. Al-Mossawi Scholarships & Cultural Affairs, Mistry of Higher Education & Scientific Research, Iraq Received April 5, 011; revised May 7, 011; accepted May 17, 011 ZnSe quantum dot (QD) semiconductor optical amplifier (SOA) is studied theoretically usg net ga for lear Absorption coefficient and lear emission and these results are used to calculate noise figure and small-signal ga. Keywords: Lear Absorption Coefficient, Lear Emission Coefficient, Rate Equations (Res), Noise Figure 1. Introduction Quantum dot photonic materials have attracted much attention recent years as they have the potential to deliver the stability and coherence of atomic sources with a compact and efficient semiconductor device. characteristics such as reduced sensitivity to optical feedback makes such materials attractive as laser sources. In addition, suppression of pattern effects QD semiconductor optical amplifiers (SOAs) shown to be promises for high speed application. The understandg of the high speed carrier dynamics of these materials is crucial for their optimization and exploitation. To address this issue, time resolved spectroscopy has been used to vestigate the fundamental carrier decay time scales of SOA structures and determe their suitability for high-speed applications. Such pump-probe studies are usually performed usg pulse width of a few hundred femtoseconds to picoseconds order to sufficiently resolve the relaxation dynamics of high-speed devices. [1] Quantum dot (QD) semiconductor optical amplifiers (SOAs) demonstrate best features when compared with other SOAs based on bulk or quantum well materials. As a result, QD SOAs are very promisg for applications high-speed optical communications. One of the most important features of QD materials results these best performances is the discrete structure of their energy levels []. In this work, ZnSe/ZnS QD-SOAs are studied. First, energy levels are calculated usg quantum box model where the dot is assumed the form a cube with quantum dot size is nm 3. net ga is then calculated by usg lear Absorption coefficient and lear emission coefficient components the QD.. Ga of QDs Ga and threshold current models are calculated with quasi-fermi levels energies E f c and E fv, which are directly controlled by the dopg concentration or jected current. Given that w x, w y and w z are the lengths each dimension of a quantum box, see Figure 1, and E cnml and E vnml are allowed quantized electron and hole energies, respectively, the correspondg carrier densities, n and p(~n) may be expressed as [3] n p fc Ecnml (1) nml nml Ecnml Ef c 1 exp kt 1 fvevnml () nml nml Ef E v cnml 1 exp kt w z w x barrier Figure 1. Diagram of dot/barrier quantum box. w y dot Copyright 011 SciRes.

2 66 K. H. AL-MOSSAWI ET AL. Where f c and f v are the correspondg Fermi functions for electrons the conduction and valence bands. The factor of two takes to account the electron sp. The summation is over all the confed energy states, which depends on the QD material and the dot/barrier valence and conduction band offsets (ΔE c and ΔE v ) that enable electron and hole confement. Solvg known E f c and E fv are ed for directly and applied to fd the lear absorption coefficient α(w) and lear emission coefficient e(w) components the QD [3] h g fv E11 fc E w o aw R de (3) nr o nml Eg h E hw h g fc E1 fv E1 w o ew R de nr o nml Eg h E hw Where R is the dipole moment, n r is the refractive dex and g is the density of states for the QD, given by [3] g E Ecnml Eg (4) (5) Where τ the traband relaxation time. Equations (3) and (4) signify the rates of absorption and emission per unit length with E 1 and E representg the hole energy level the valence band and the electron energy level the conduction band, respectively. The net ga of the system is then [3] Gwew aw (6) To determe the optimal operatg region for an amplifier system, the ga coefficient, expressed as [3] Gw ga coefficient (7) P 3. QD-SOA Based on the rate equations used [4] for the wettg layer (WL) carrier density N w, occupation probabilities for ground and exited states (f and h), respectively one can write. N w 1 w J N h Nwh Nw (8) t ql w w w wr 1 N L h 1 1 h N L h f h f h t N N w w w w Q w Q w 1 1 (9) 1 f h f 1h gl (10) f f p c f 1Sav t R NQ s Where J is the current density, q is the electron charge, L w is the effective thickness of the active layer, τ w is the carrier relaxation time from the two-dimensional WL to the ES, τ w is the carrier escape time from the ES to the WL, τ wr is the spontaneous radiative lifetime WL, τ 1R is the spontaneous radiative lifetime the QDs, τ 1 is the carrier relaxation time from ES to GS, τ 1 is the carrier relaxation time from the GS to the ES, N Q is the surface density of QDs, ε s is the dielectric constant of the QD material and Г is the optical confement factor. g p is the peak material ga. Accordg to Pauli's exclusion prciple, the occupation probability the quantum-dot ground state relates to the carrier density N the QD by f = N/(N Q /w x ) [5] where w x is the size of the QD. The average signal photon density S av is given by [6] S av G Rf RbG Pn g (11) RRG RRG hwplwdlg pc f b 4 f b s Where P is the put signal power, n g is the group refractive dex, ћ is the normalized Plank s constant, w p is the peak frequency, D is the width of the active layer, L is the cavity length, c is the free space light speed. Note that the Fibry-Perot amplifier ga G is given by [7] G 1 1 R R G f b S 1 f b s 4 f b ss RRG RRG (1) R f is the front mirror reflectivity, R b the back mirror reflectivity. Φ is the phase angle while G S is the sgle-pass ga of the structure, given by [7] Gs exp gpat L (13) Note that α t is the loss coefficient. The carrier dynamics described by the rate Equations are related only to electrons the conduction band while hole dynamics can be neglected due to their larger effective mass [4]. In addition, typical values of the material parameters Table 1 [5] for ZnSe/ZnS system where (E g =.8 ev, m e = 0.17 m o and m v = 0.60 m o ) for ZnSe and (E g = 3.68 ev, m e = 0.39 m o and m v = 0.49 m o ) for ZnS. Rate equations model are solved at steady state case usg the expression of the average signal photon density S av, Equation (11). The noise added to the signal durg the amplification process is a fundamental property for every kd of amplifiers. The noise characteristics of an amplifier are quantified by a parameter called noise figure, when the shot noise part is taken to account, the amplifier noise figure can be written as [7]. Copyright 011 SciRes.

3 K. H. AL-MOSSAWI ET AL. 67 Table 1. Parameters used the Calculations [5]. Parameter Value Unit L 00 μm L w 0. μm D 10 μm N Q cm - τ w 3 ps τ w 1 ns τ 1 1. ps τ ps τ 1R 0.4 ns τ wr 1 ns Φ 0 R f = R b 10 4 α t 3 1/cm Γ L b 0 nm L c 75 nm P 1 μw F n G1 / G 1/ G (14) n sp The the spontaneous emission factor n sp is given by [8] n sp ew at G w (15) thus one can use this property to work with the same structure at higher ga by choosg the subbands that occur between them. Also, one can refers to the somewhat crease between the ground and excited peaks with raband relaxation time. Rate equations are then solved at steady state, usg the parameters listed Table 1, the QD-SOA characteristics are examed. Small-signal ga spectrum of this QD-SOA is shown Figure 6. These ga values are comparable to that obtaed by Ben-Azra [9]. Note that it is obtaed here at a lower current value. So, this is a good modification. The effect of QD layer is also examed where the ga creases with QD layers as shown Figure 7. Figure 8 shows the output power calculated as a function of carrier density with put power is taken as a parameter. Includg the shot noise part, the noise figure spectrum is shown Figure 9, while the noise figure curve at different carrier densities is shown Figure 10 from the last two figures one can refers to the lower noise accompanied with this amplifier while the last three figures shows the properties of ZnSe/ZnS QD-SOA which is a good results to use it applications. 4. Results and Discussion Usg quantum box model we calculate QD energy levels for conduction and valence bands and then it is used to calculate lear absorption coefficient, lear emission coefficient and net ga. The barrier layer thickness is 0 nm while the thickness of the clad layer is 75 nm, see Figure. All calculations are done at τ = 0.1 ps [3] and carrier density 10 4 m 3. Also we calculate the small signal ga for structures conta one and six layers. Ga coefficients are calculated usg Equation (7) for several values of put power. Figure 3 shows a comparison between the lear absorption and emission coefficients where a multi-peaks appear for ground and excited state transitions. One can refers to the higher emission value compared to that of absorption which results from lower transparency value for this structure. The net ga is shown Figure 4, where a double peak can also be seen. This is very important some applications. For example, this can be used long haul optical transmission. The reason for this double emission is reasoned to fite ground state relaxation time, which brgs ground state emission to a constant value after the excited state threshold [10]. Figure 5 shows the effect of traband relaxation time τ where a higher ga is obtaed at longer traband time. Ga of the highest peak creases by ~ 4.5 time when τ creases from 0.1 ps to 0.5 ps. Really, traband relaxation time controlled the transition between tersubband Figure. QD system conta six active layers. Figure 3. Comparison between lear absorption and emission coefficients. Copyright 011 SciRes.

4 68 K. H. AL-MOSSAWI ET AL. Figure 4. Ga of ZnSe/ZnS QDs. Figure 7. Small-signal ga spectra with QD layer as a parameter. Figure 5. Ga versus wavelength at several values of traband relaxation time. Figure 8. Output power versus current at values several from put power. Figure 6. Small signal ga spectrum. Figure 9. Noise figure with shot noise cluded the calculations. Copyright 011 SciRes.

5 K. H. AL-MOSSAWI ET AL. 69 Figure 10. noise figure with shot noise versus carrier density. 5. References [1] T. Piwonski, I. O Driscoll, J. Houlihan, G. Huyet and J. Manng, Carrier Capture Dynamics of InAs/GaAs Quantum Dot, Applied Physics letters, Vol. 90, No. 1, 007, pp.. [] T. Erneux, E. A. Viktorov, P. Mandel, T. Piwonski, G. Huyet and J. Houlihan, The Fast Recovery Dynamics of a Quantum Dot Semiconductor Optical Amplifier, Applied Physics Letters, Vol. 94, No. 11, 009, pp doi: / [3] C.-J. Wang, L. Y. L and B. A. Parviz, Modelg and Simulation for a Nano-Photonic Quantum Dot Wave- guide Fabricated by DNA-Directed Self-Assembly, IEEE Journal of Selected Topics Quantum Electronics, Vol. 11, No., 005, pp doi: /jstqe [4] Y. Ben-Ezra, M. Haridim and B. I. Lembrikov, Theoretical Analysis Of Ga-Recovery Time and Chirp QD-SOA, IEEE Journal of Selected Topics Quantum Electronics, Vol. 17, No. 9, 005, pp [5] M. A. Al-Mossawi, A. H. Al-Khursan, R. A. Al-Ansari, ZnO-MgZnO Quantum-Dot Semiconductor Optical Amplifiers, Recent Patents on Electrical Engeerg, Vol., No. 3, 009, pp doi:10.174/ [6] M. Vasileiadis, D. Alexandropoulos, M. J. Adams, H. Simos and D. Syvridis, Potential of InGaAs/GaAs Quantum Dots for Applications Vertical Cavity Semiconductor Optical Amplifiers IEEE Journal of Selected Topics Quantum Electronics, Vol. 14, No. 4, 008, pp doi: /jstqe [7] E. Forestieri, Optical Communication Theory and Techniques Ch. 8, Sprger, Pisa, 005. [8] Y. Ben-Ezra, B. I. Lembrikov and M. Haridim, Acceleration of Ga Recovery and Dynamics of Electrons QD-SOA, IEEE Journal of Selected Topics Quantum Electronics, Vol. 41, No. 10, 005, pp doi: /jqe [9] K. Veselv, F. Grillot, C. Cornet, J. Even, A. Bekiarski, M. Gianni and S. Loualiche, Analysis of the Double Laser Emission Occurrg 1.55 mm InAs-InP (133) B Quantum-Dot Lasers, IEEE Journal of Selected Topics Quantum Electronics, Vol. 43, No., 007, pp Copyright 011 SciRes.