WWW.DISSERS.RU


...
    !

Pages:     || 2 | 3 |
, 2005, 47, . 11 Effect of the electric field on the carrier collection efficiency of InAs quantum dots, E.S. Moskalenko, K.F. Karlsson, V. Donchev, P.O. Holtz, W.V. Schoenfeld, P.M. Petroff Department of Physics and Measurement Technology, Linkping University, S-58183 Linkping, Sweden A.F. Ioffe Physical-Technical Institute, Russian Academy of Sciences, 194021 St. Petersburg, Russia Faculty of Physics, Sofia University, 1164 Sofia, Bulgaria Materials Department, University of California, Santa Barbara, 93106 California, USA E-mail: evgenii.moskalenko@mail.ioffe.ru (Received January 17, 2005) Individual and multi quantum dots of InAs are studied by means of microphotoluminescence in case when, in addition to the principal laser exciting photoluminescence, second infrared laser is used. It is demonstrated that the absorption of the infrared photons effectively creates free holes in the sample, which leads to both change in the charge state of a quantum dot and to the considerable reduction of their photoluminescence signal. The later effect is explained in terms of an effective screening of the internal electric field, facilitating the carrier transport along the plane of a wetting layer, by the surplus holes from the infrared laser. It is shown that the effect of quenching of quantum dot photoluminescence gradually disappears at increased sample temperature (T ) and / or dot density.

This fact is due to the essentially increased value of quantum dot collection efficiency which could be achieved at elevated sample temperatures for the individual quantum dots or even at low T for the case of multi quantum dots.

It is suggested that the observed phenomena can be widely used in practice to effectively manipulate the collection efficiency and the charge state of quantum dot-based optical devices.

E.S.M. gratefully acknowledges financial support from the WennerGren Foundations and partial support from the program Low-Dimensional Quantum Structures of the Russian Academy of Sciences. V.D. is thankful to the Swedish Foundation for International Cooperation in Research and Higher Education (STINT) for financial support.

1. Introduction carrier hopping between QDs [13], by trapping of migrating particles into localized states of the WL [14] or into nonSemiconductor quantum dots (QDs) effectively confine radiative centers [15] in the surrounding media. A more efficient carrier transfer from the WL into the QDs via electrons (es) and holes (hs) on the nanometer length radiation-induced defects in the WL has been reported [16].

scale in all three dimensions and hence may be considered A magnetic field directed perpendicular to the plane of as artificial atoms [1]. Unlike real atoms, QDs can be the structure was observed to limit the lateral transport of manipulated in different ways, which opens the possibility carriers [17].

to tailor their shape, size and composition [2] in order to It has also been suggested [8] that the carrier drift achieve the desired properties. Consequently, QDs are could be considerably influenced by a long-range attractive potential candidates for various optoelectronic (electronic) potential caused by the strain field surrounding the QD.

applications such as QD lasers [3], QD infrared detectors [4], On the other hand, strain-induced potential barriers in the QD memory devices [5] and single-electron transistors [6].

barrier/QD [18] and in the WL/QD interface [19] were For a majority of these devices the QDs become considered to limit the carrier capture into the QD. The populated with carriers, which are primarily created outside important role of an electric field directed in the growth the QDs somewhere in the sample (in the barriers or in the direction of the sample on the carrier capture into and wetting layer (WL), on which QDs are normally grown [7]) escape out of the QD was demonstrated by the studies of by means of electrical or optical excitation. Consequently, the electric current passing through the QDs [20].

excited carriers undergo a transport in the WL/barriers prior In our previous study [21] we pointed out another to the capture into the QDs. This circumstance highlights mechanism of the carrier transfer from the WL into the the crucial role of the carrier capture processes into the QD QDs, which has not been previously considered. A built-in for the performance and operation of the QD-based devices.

electric field (F) directed in the plane of a WL to The carrier capture mechanisms intensively studied in facilitate the lateral carrier transport. However, in these the last decade reveal optical phonon assisted [8,9], Augermeasurements individual QDs were studied only at a fixed like [10], shake-up [11] processes and carrier relaxation sample temperature (T ) of 5 K. In the present paper the through the band tail states of the WL with a subsequent suggested mechanism on the carrier capture into the QDs is emission of localized phonons [12]. The lateral carrier investigated at increased sample temperatures (up to 70 K) transport (in the plane of the WL) could be affected by as well as increased dot density.

Effect of the electric field on the carrier collection efficiency of InAs quantum dots In our experiments we use an additional infrared (IR) surface down to a spot diameter of 2 m. The main laser laser to influence the field F. The excitation energy of (L0) was used to excite the PL of the WL and the QDs.

the IR laser, hIR = 1.240 eV, is considerably less than The exitation energy (hex) was tuned in the range from the lowest transition energy of the sample studied and, 1.410 to 1.480 eV with a maximum excitation power (P0) accordingly, can not simultaneously excite both electrons of 20 W. The other laser, LIR, operating at a fixed excitation (es) and holes (hs), but can generate solely either es or energy, hIR = 1.240 eV has its maximum output power hs by excitation of deep level (DL) defects positioned in (PIR) of 100 W. It is important to note that the hIR is well the band gap of the CaAs barriers [22]. According to our below the value of the QDs related emission and accordingly model, these extra carriers, excited by the IR laser, will no signal neither from the WL nor from the QDs was effectively screen the field F and will consequently slow detected with excitation solely with LIR. The sample was down the carrier transport in the plane of the WL. Due positioned inside a continuous-flow cryostat operating in a to this effect, a considerable reduction (up to 10 times) of temperature (T ) range of 5-70 K.

QD photoluminescence (PL) signal (IQD) is experimentally To find the particular QD to study, a laser beam was observed when the sample is exposed to dual excitation of scanned across the sample surface. Once the desired QD an IR laser and a main laser. was found, special marks (grids) were burnt into the sample To the best of our knowledge, there are very few earlier surface around the QD with a high power laser beam. The publications [23,24] devoted to studies of IR laser induced average distance between the adjacent QDs in the low dot changes in IQD. In contrast to our findings, it was found [23] density area of the sample was around 10 m. To control that the IR laser induces an increase of the PL from the the exact position of the laser spot on the sample surface, QDs by up to 40%. This phenomenon was explained in the image of the interesting region was projected by a video terms of an IR laser induced release of carriers, which camera, which made it easy to find the desired QD marked were trapped into deep defects from the QDs. The by the grid. In addition, this arrangement allowed us to considerable changes in the fluctuations of the IQD during effectively correct the laser position on the sample, if the the time interval of the measurement were detected, when sample was moved due to the thermal drift. It should be the sample was illuminated with an additional near-IR laser noted that the method to locate the exact QD position by irradiation [24]. Carriers trapped at deep localized centers using the described grids is favourable in several respects in the vicinity of the QDs were suggested to be responsible compared to alternative methods, e. g. employing a metal for the observed phenomenon [24]. mask with small holes deposited on top of the sample which Our present results demonstrate that the strength of the may produce an electric field in the near-surface region observed quenching effect of IQD progressively decreases and, consequently, may influence the carrier transport in with an increasing temperature as well as dot density. the plane of the WL. In addition, the metal mask may act This is explained in terms of an essentially increased QDs as a stressor, which could spoil the quality of the QDs.

collection efficiency (), i. e. the ability of QDs to collect Eight SQDs located at different spatial positions of the photoexcited carriers from illuminated area. At these sample, all with an analogous behavior with respect to experimental conditions the role of F, which facilitates the LIR, were examined in this study. In this report we the carrier transport at a lower values of, becomes present data measured on low density structure with one diminishing. SQD within the laser spot together with a high dot density area with a varying number of QDs within the area of the laser spot.

2. Sample and experimental setup The sample studied was grown by molecular beam 3. Experimental results and discussion epitaxy on the GaAs (100) substrate. It consists of lensshaped InAs QDs developed on a InAs WL from about The low-temperature PL spectrum of a sample spot with 1.7 monolayer InAs deposited in StranskiKrastanov growth only one QD (SQD) within the laser illumination obtained mode. The WL and dot layer was sandwiched between under excitation with a single laser L0 at an excitation two 100 nm thick GaAs barriers. The sample was grown energy hex = 1.471 eV is shown in Fig. 1 (dotted lines).

without rotation of the substrate, resulting in a gradual The PL spectrum of the sample is dominated by the WL variation of In flux across the wafer and consequently emission at an energy around 1.445 eV. A weak emission a gradient in the QDs density. The QDs were studied line, peaking around 1.340 eV is attributed to the PL from a by means of a diffraction-limited micro-PL (PL) setup SQD. The spectrally integrated PL signal from a SQD (IQD) (detailed description of the setup and the sample growth is approximately 1% with respect to the corresponding PL procedure is given in [25]). The PL technique empoyed intensity of the WL (IWL). This ratio is much higher than in the present experiments allowed us to excite and study the corresponding volume ratio ( 10-3) of the SQD vs single QD (SQD). the excited WL volume. This fact is a direct evidence that To excite the sample we used two TiSp lasers, which the PL signal from a SQD is not entirely determined by beams were focused on the same position of the sample the number of carriers excited in the SQD as a result of 10 , 2005, 47, . 2068 E.S. Moskalenko, K.F. Karlsson, V. Donchev, P.O. Holtz, W.V. Schoenfeld, P.M. Petroff The increase of the IWL could be explained in terms of a release of carriers from the centers of non-radiative recombination as a result of the infrared absorption by LIR (arrow IR1 in the inset in Fig. 1). Such an enhancement of the radiative efficiency, induced by an additional below band gap excitation, has been reported earlier (see e. g. [26]).

If this was the only operating mechanism caused by LIR, an increase of IQD would also be expected, since the WL serves as a reservoir of carriers shich could be trapped into the SQD, as explained above. Consequently, a different mechanism of the LIR influence on the sample has to be considered in order to explain the observed decrease of IQD.

To understand the origin of infrared laser effect on the PL intensity we studied the temperature (T ) dependence of the observed effect. Fig. 2 a shows a number of pairs of SQD PL spectra recorded with single laser L0 excitation (dotted lines) and dual laser excitaion (solid lines) at different T.

Figure 1. PL spectra of a SQD and the WL measured at Two emission lines marked as X and X- have earlier T = 5K and hex = 1.471 eV with single (dotted lines) and dual been interpreted [25] as excitonic lines with different charge (solid lines) laser excitation, respectively, at an excitation power of configurations. The neutral and single negatively charged P0 = 200 nW and PIR = 100 W. The vertical solid arrows show excitons, i. e. 1e1h and 2e1h charge states, respectively, the excitation energies of L0 used in the experiment. The inset could be detected. The progressive redistribution of the shows the principle transitions involved in the energy scheme of the sample together with the positions of the GaAs conduction PL spectra in favour of the X- line, detected with band (CB), valence band (VB) and deep level (DL). The vertical single laser L0 excitation with increasing T (dotted lines and curved arrows are explained in the text.

in Fig. 2, a) was explained [27] in terms of a temperatureinduced increased electron diffusivity in the WL plane. This leads to a faster diffusion and, hence, capture of es into the dot with respect to the capture of hs. The PL spectra absorption of photons in the dot volume, but rather to a significant extent by carriers excited in the WL. In other words, the WL serves as a reservoir which supplies the QD with carriers. Indeed, carriers excited in the WL (hex in the inset in Fig. 1) can undergo alternative trapping processes:

a) a relaxation down to the localized WL states, followed by a radiative recombination, b) a trapping at centers of nonradiative recombination (CNRs) or c) capture into a SQD (processes shown by the arrows r, nr nad c, respectively, in the inset in Fig. 1). The interplay between these processes determines the values of IQD and IWL measured in the experiment.

We have solved a simple set of rate equations in steady-state conditions (not shown here) which resulted in IQD/IWL = c/r = at any value of nr, where the parameter is denoted the collection efficiency of the SQD.

The physical meaning of the introduced parameter could be explained by the following example. If the capture probability c exceeds the value of r, the PL of the SQD would dominate the PL spectrum of the sample at the expense of the WL PL, i. e. the dot would efficiently collect a majority of carriers created by the laser absorption within the laser spot.

Pages:     || 2 | 3 |



2011 www.dissers.ru -

, .
, , , , 1-2 .