Both ge and gh are linearly proportional to the excitation power of the corresponding laser, P0 and PIR, respectively, as given by: ge = eNadGaAsP0/hex and gh = hNDLdGaAsPIR/hIR. Here dGaAs = 200 nm is the thickness of the GaAs barriers, e (h) is the acceptor (DL) optical cross section for electrons (holes) and Na (NDL) is the concentration of ionized acceptors (empty DLs) in the GaAs. To get an assessment of ge and gh, typical employed values are: NDL = 5 · 1012 cm-3 , Na = 6 · 1013 cm-3 , h = 6 · 10-18 cm2 (the value for the EL2 DL in GaAs at 1.233 eV ) and e = 3.5 · 10-15 cm2 for the carbon acceptor.1 Thus, for the lowest value of P0 (17 nW) and the highest value of PIR (100 µW), we obtain ge gh 3 · 105 s-1. Since gad is determined by the lower value of ge or gh, one can consider gad = gh for all laser powers used in this study.
The increase of IWL is relatively small due to the fact that PL the excess electrons and holes are generated by two different Figure 3. The symbols show the ratio RQD of the spectrally photons, i. e., in two different moments separated by a time integrated QD PL intensities measured under dual and single interval of the order of the lower value of g-1 or g-1. This laser excitation, respectively, at T = 5K with hex = 1.503 eV e h interval is at shortest 5.7 ns for the highest P0 used here as a function of the power ratio of the two lasers, PIR/P0, as follows: circles — PIR = 50 µW and P0 in the range 17 to (10 µW), i. e., much longer than the time needed by the 10 000 nW; triangles — PIR = 100 µW and P0 in the range photo-excited carrier for migration, capture and relaxation to 10 000 nW; and squares — P0 = 17 nW and PIR in the range to the lowest QD energy level (< 1ns ). Consequently, 0.55 to 50 µW. The solid line is calculated, based on the expression the excess electron and hole will likely be captured and -RQD = 1 + a(PIR/P0) with a = 1.1 · 1010.
recombine in the QD rather than in the WL. Accordingly, in the expressions for ge and gh of our simplified estimate, it is assumed that all excess carriers are efficiently collected into the QD, as earlier demonstrated for excess electrons .
In order to understand the increase of IQD, the rate RQD PL is introduced, defined as the ratio between IQD measured PL under dual and single laser excitation, respectively, i. e., RQD = IQD(PIR)/IQD(PIR = 0). According to the abovePL PL described model, RQD should be expressed as gQD + gad gad RQD = = 1 +. (1) gQD gQD The dependence of RQD on the powers ratio of the two lasers, PIR/P0, wirh either P0 or PIR fixed is shown in Fig. 3. For high P0’s and/or low PIR’s, RQD is close to unity, since gad/gQD 1 (cf. Eq. (1)). This Figure 4. Micro-PL spectra of a single InAs/GaAs QD under fact is in agreement with the experimental observation single (solid line) and dual (dotted line) laser excitation measured that no redistribution of the µPL spectrum occurs (not at 5 K with hex = 1.508 eV, P0 = 20 nW, and PIR = 50 µW.
shown here). RQD increases for low P0’s and/or high The spot of the IR laser is positioned on the QD, while the PIR’s, because gad/gQD > 1 in this case (cf. Eq. (1)).
spot of the primary laser is moved 2.5 µm aside of the QD, as The experimental points are well fitted by the linear exschematically shown in the inset.
pression RQD = 1 + a(PIR/P0) with a = 1.1 · 10-3 (Fig. 3).
According to Eq. (1) and the expressions for gh and gQD with wc(hex = 1.503 eV) =1.7 · 10-3, a should be hNDLdGaAshex/(wcWLdWLhIR) =0.9 · 10-3. This is in by a distance of 2.5 µm. This distance and P0 were chosen good correspondence with the experimental value, which in such a way that the PL signal rendered undetectable further supports the proposed model.
under single laser excitation (see Fig. 4). The e-h pairs In a complementary experiment, the spot of the primary generated in this case can not reach the QD because of laser (with hex = 1.508 eV) was moved aside of the QD the long distance, but will rather recombine in the WL.
1 However the excess electrons are efficiently collected into Calculated for hex = 1.503 eV by means of the formulae given in W.P. Dumke, Phys. Rev. 132 (5), 1998 (1963). the QD . Indeed, when the IR laser, focussed on the Физика твердого тела, 2006, том 48, вып. Enhancement of the photoluminescence intensity of a single InAs/GaAs quantum dot... from the valence band in GaAs. On the one hand, this should increase the number of the optically generated excess electrons via the ionized acceptors. On the other hand, the number of thermally generated holes from the acceptor levels in GaAs grows quickly for T > 30 K2 and becomes comparable to the number of optically generated excess electrons. The appearance of free holes in the valence band leads to an efficient band-to-band recombination, which decreases the number of the excess electrons and consequently results in a reduced charging of the QD (see the discussion in ). Indeed, with increasing temperature, the single laser excited µPL spectrum is gradually redistributed from the double charged (X2-) toward the single charged exciton line (X-) and at T = 55 K, the X- line already dominates the spectrum, as shown in Fig. 1, b. This temperature induced redistribution of the PL spectrum under single laser excitation is illustrated in Fig. 5, c. The figure displays the relative contribution Ws of the neutral together with the single and double charged excitons to the total QD PL intensity, calculated from the areas under the X, Xand X2- spectral lines, respectively. It is seen that the contribution of X2 decreases with temperature, while that of X- increases and becomes dominant at T > 40 K.
Another effect of the increased temperature is the thermally enhanced exciton and carrier transport in the WL , resulting in a higher collection efficiency of excitons and photo-carriers from the WL to the QD. This effect explains the gradual increase of the single laser excited IQD (and PL consequently gQD) in the range from 5 to 55 K, shown in Fig. 5, a. At T = 55 K, IQD is about 10 times larger with PL respect to T = 5 K. The decrease of IQD for T > 55 K can PL be due to thermal activation of non-radiative recombination channels .
Figure 5. The temperature evolution of (a) the intergrated QD Concerning the IR laser generation of excess holes, we PL intensity, IQD, measured under single laser excitation; (b) the PL can assume that gh does not change significantly with inratio RQD of the integrated QD PL intensities measured under creasing temperature. This is justified by the fact that in the dual and single laser excitation, respectively and compared with temperature range under study, the thermal ionization of the the calculated values (given by stars); (c) and (d) the relative DLs is negligible, because the DL ionization energy (of the contributions of the neutral (done triangles), single (uptriangles) GaAs order 0.5Eg ) is much larger than kT. No increase of gh is and double (circles) charged excitons to the total QD PL intensity in the case of (c) single and (d) dual laser excitation. expected, but rather a slight decrease as a result of electron hex = 1.503 eV, P0 = 40 nW, and PIR = 100 µW. The lines in (c) trapping from the conduction band into the DL (which and (d) are only guides for the eyes.
reduces the number of empty DLs). It is also assumed that for the experimental conditions (P0, PIR and T ) employed, gh remains below ge, i. e., gad = gh (as estimated above for T = 5K). Taking these facts into account together with QD, was added, a well-defined µPL spectrum consisting of the temperature increase of gQD (Fig. 5, a) and Eq. (1), one the neutral exciton X appeared (Fig. 4). This phenomenon expects a decrease of RQD as a function of temperature. This can be explained with the assumption that each laser has been experimentally confirmed, as shown in Fig. 5, b.
supplies the QD with just one type of carriers, electrons It is seen that RQD gradually decreases with temperature or holes, respectively. This experimental observation can be to become nearly 1 at 55 K. Moreover, according to regarded as a direct proof of the effect of separate generation Eq. (1) and the above considerations, the dependence of of excess electrons and holes by the primary and the IR RQD(T ) on the temperature should follow the expression laser, respectively.
1 + const/IQD(T ). The so-derived prediction of RQD(T ) Finally, the temperature dependence of the described PL effect with dual laser excited PL spectra has been studied Calculated in dark using Fermi statistics and the carbon ionization (with P0 = 40 nW and PIR = 100 µW). The increase of energy (26 meV) in the charge neutrality equation with an assumed total the temperature enhances the acceptor ionization processes acceptor concentration of 1 · 1014 cm-3.
Физика твердого тела, 2006, том 48, вып. 1882 V. Donchev, E.S. Moskalenko, K.F. Karlsson, P.O. Holtz, B. Monemar, W.V. Schoenfeld...
(employing the values of IQD from Fig. 5, a) is represented pairs generated by the primary laser in the WL are more PL in Fig. 5, b by the stars. As expected, the two sets of data and more efficiently collected into the QD, due to the points in Fig. 5, b coincide very well. This finding gives thermally enhanced transport in the WL, and consequently further support to the proposed model. As the temperature the contribution of excess electrons and holes becomes less increases, the relative contribution of the excess electrons important. As a result the effect of the QD PL intensity and holes supplied into the QD becomes less important increase gradually weakens with temperature, while the with respect to the increased number of e-h pairs created redistribution effect exhibits a decreasing tendency above 40 K, to get quenched above 55 K.
by the primary laser in the WL and subsequently collected into the QD.
While RQD qradually decreases with temperature, the References second IR laser induced effect, the altered QD charge configuration in favor of a more neutral exciton state is  S. Raymond, K. Hinzer, S. Fafard, J.L. Merz. Phys. Rev. B 61, clearly demonstrated even at 55 K. This is illustrated in R16331 (2000).
Fig. 1, b, where the doubly excited µPL spectrum is again  U. Jahn, R. Ntzel, J. Ringling, H.-P. Schnherr, H.T. Grahn, K.H. Ploog, E. Runge. Phys. Rev. B 60, 11 038 (1999).
dominated by the neutral exciton line X (as was the case at  S. Marcinkeviius, R. Leon. Appl. Phys. Lett. 76, 2406 (2000).
5 K, Fig. 1, a). The temperature evolution of this effect of  M. Grundmann, D. Bimberg. Phys. Stat. Sol. (a) 164, the IR laser is illustrated in Fig. 5, d. The figure displays the (1997).
relative contributions Wd of the neutral, single and double  D.A. Mazurenko, A.V. Scherbakov, A.V. Akimov, A.J. Kent, charged excitons to the total QD PL intensity in the case of M. Henini. Semicond. Sci. Technol. 14, 1132 (1999).
dual laser excitation. It is seen that the contribution of X is  A.V. Akimov, V.G. Shofman. J. Lumin. 53, 335 (1992).
more than 3 times larger than the other two components for  M. Lomascolo, A. Vergine, T.K. Johal, E. Rimaldi, A. Pasall temperatures in the range 5 T 55 K. While the X sarelo, R. Cingolani, S. Patan, M. Labardi, M. Allegrini, exciton is under the detection limit at any temperature F. Troiani, E. Molinari. Phys. Rev. B 66, 041 302 (2002).
with single laser excitation (see Fig. 5, c), the IR laser will  J.J. Finley, A.D. Ashmore, A. Lematre, D.J. Mowbray, M.S. Skolnik, I.E. Itskevich, P.A. Maksym, M. Hopkinson, change the QD charge configuration in favor of a more T. Krauss. Phys. Rev. B 63, 073 307 (2001).
neutral exciton state in whole the temperature range under  E.S. Moskalenko, V. Donchev, K.F. Karlsson, P.O. Holtz, study. Nevertheless, a decreasing tendency for the X exciton B. Monemar, W.V. Schoenfeld, J.M. Garcia, P.M. Petroff. Phys.
intensity is observed for T > 40 K, where the contributions Rev. B 68, 155 317 (2003).
of X and X- are approaching (see Fig. 5, d). This behavior  E.S. Moskalenko, K.F. Karlsson, P.O. Holtz, B. Monemar, is consistent with the low values (close to 1) of RQD in this W.V. Schoenfeld, J.M. Garcia, P.M. Petroff. Phys. Rev. B 64, range (see Fig. 5, b). As explained above, for T > 40 K, the 085 302 (2001).
photo carriers generated as a result of the absorption of the  E.S. Moskalenko, K.F. Karlsson, P.O. Holtz, B. Monemar, primary laser in the WL are efficiently collected from the W.V. Schoenfeld, J.M. Garcia, P.M. Petroff. Phys. Rev. B 66, WL into the QD (see Fig. 5, a) and their relative contribution 19, 195 332 (2002).
 K.F. Karlsson, E.S. Moskalenko, P.O. Holtz, B. Monemar, to the QD PL (gQD) becomes more and more dominant W.V. Schoenfeld, J.M. Garcia, P.M. Petroff. Appl Phys. Lett.
with respect to the PL originating from the excess electrons 78, 19, 2952 (2001).
and holes (gad).
 A.M. Wite, P.J. Dean, D.J. Ashen, G.B. Mullin, B. Webb, B. Day, P.D. Greene. J. Phys. C 6, L243 (1973).
 S. Adachi. J. Appl. Phys. 66, 12, 6030 (1989).
By means of dual laser excitation of a single InAs/GaAs  P. Blood, J.J. Harris. J. Appl. Phys. 56, 4, 993 (1984).
QD with an appropriate choice of the excitation energies  M. Heiblum, E.E. Mendez, L. Osterling. J. Vac. Sci. Techand powers, the QD is supplied with separately generated nol. B 2, 2, 233 (1984); I.H. Goodridge. Properties of Gallium excess electrons and holes from shallow acceptors and Arsenide. Inspec, London (1986).
DLs, respectively, in the GaAs barriers. Although the  P. Silverberg, P. Omling, L. Samuelson. Appl. Phys. Lett. 52, generation rates of excess carriers are much lower than 20, 1689 (1988).
the e-h pair generation rate of the primary laser in the  C. Lobo, R. Leon, S. Marcinkeviius, W. Yang, P. Sercel, X.Z. Liao, J. Zou, D.J.H. Cockayne. Phys. Rev. B 60, 24, WL, the resulting effect is quite strong at 5 K. The QD 16 647 (1999).
PL intensity increases several times compared to single  K. Mukai, M. Sugawara. In: Self-Assembled InGaAs/GaAs laser excitation and the µPL spectrum is redistributed in Quantum Dots. / Ed. by M. Sugawara. Semiconductors and favor of a more neutral exciton configuration. This is due Semimetals. Academic Press, San Diego (1999). Vol. 60.
to the fact that for the excitation powers employed, the P. 183.
Материалы этого сайта размещены для ознакомления, все права принадлежат их авторам.
Если Вы не согласны с тем, что Ваш материал размещён на этом сайте, пожалуйста, напишите нам, мы в течении 1-2 рабочих дней удалим его.