Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.lssm.inorg.chem.msu.ru/blinking.pdf
Дата изменения: Mon Oct 25 18:54:10 2010
Дата индексирования: Mon Oct 1 19:30:03 2012
Кодировка:
Blinking as a Tool of Quantum Dot OLEDs Studies
Panos Argyrakis1, Vladimir Lidsky2, Roman Vasiliev3, Dmitry Dirin3, Vasily Zalunin2, 4, Alexei Vitukhnovsky2 1 Aristotle University of Thessaloniki, Thessaloniki, 541 24, Greece, Email: panos@physics.auth.gr 2 Lebedev Physical Institute RAS, 53 Leninsky pr., Moscow, 119991, Russia, Email: alexei@sci.lebedev.ru 3 Lomonosov Moscow State University, 1-3 Leninsky Gory, Moscow, 119991, Russia, Email: ddirin@gmail.com 4 Lomonosov Moscow State University, 1-2 Leninsky Gory, Moscow, 119991, Russia, Email: vzalunin@mail.ru Abstract: Synthesis conditions for different QDs types were suggested and original experimental set-up for luminescence studies of prepared QDs was used. The luminescence time dependence for two types of core/shell QDs (CdSe/CdS and CdTe/CdSe ) were investigated and were compared with behavior of naked ones (CdSe). QDs application for effective OLEDS is discussed. Key Words: quantum dots, blinking, OLED.

1 Introduction
Semiconductor nanocrystals (quantum dots, QDs) demonstrate novel quantum-size properties that differ from bulk material properties. In this connection colloidal QDs capped with long-chain organic molecule are of interest because of their manipulation simplicity. Size-depended optical properties, high extinction coefficient and photostability make QDs are very attractive for optoelectronic devices production, e.g. QLEDs (Organic Light Emitting Diodes) [1]. However there are any difficulties caused by fluorescence intermittency. Under excitation semiconductor quantum dots exhibit the so-called "blinking" behavior: the time-trace of the luminescence of each dot consists of rather long (up to tens of seconds) "on" and "off" intervals, i.e. time intervals at which the QD produce intensive radiation and intervals with nearly no luminescence at all, respectively. Optical dynamics such as QDs blinking are now ubiquitous in the literature of single chromophore systems [2].This effect has a serious impact on the design of the new OLEDs as it can reduce the efficiency of the diodes. Every QD shows a similar powerlaw behavior for the offtime distribution regardless of temperature, excitation intensity, surface morphology or size [3]. Despite numerous studies, uncertainty in the influence of the surface morphology modification to the on-time distribution remains. One of the QDs blinking describing models [4] assumes the long-live traps absence in the QD environment and postulate that holes localize in the deep surface traps. One of the most significant predictions of this model consists in blinking statistic distribution independency on shell thickness for core/shell nanocrystals. Such behavior was found by Nienhaus et al. [5] in the CdSe/ZnS system. Tendency to long time interval quantity increasing was observed at the same time. Contrariwise, direct dependence of QDs blinking statistics on ZnS shell thickness was found in work [3, 6]. Our experiments mission was to shed light on surface morphology modification influence to QDs fluorescence intermittency. For that we observe a pronounced correlation between QDs electronic structure and blinking events. Three different types of QDs were used in this work: simple core CdSe QDs, core/shell CdSe/CdS QDs and type II core/shell CdTe/CdSe heterostructures in charge carrier division regime (Fig.1).

Figure 1. (a)-simple core CdSe QDs, (b)-core/shell CdSe/CdS QDs, (c)-core/shell CdTe/CdSe QDs; band structures are depicted below

2 Experimental
CdSe and CdTe QDs were synthesized by the colloidal chemistry methods from supersaturated solution in highboiling solvent (diphenylether) [7]. Cadmium oleate was used as starting precursor and oleic acid as stabilized agent. For CdSe quantum dots synthesis 1M trioctylphosphine selenide solution in trioctylphosphine was quickly injected in the reaction mixture under 200. After injection the particle growth was carried out during 5 min, and then the reaction mixture was cooled. Equal volume of acetone was added for precipitation of the quantum dots. Coagulated quantum dots were separated by centrifugation, twice washed with acetone and dissolved in hexane. CdSe/CdS core/shell QDs were prepared by epitaxial growth of CdS shell on the CdSe nanocrystal surface. For that 1M trioctylphosphine sulfide solution in trioctylphosphine was injected in mixture of characterized QDs redispersed into cadmium oleate solution. CdTe/CdSe type II QDs were synthesized by combined slow injection of cadmium oleate and trioctylphosphine selenide in CdTe QDs solution under 200. Synthesis course control was carried out by constant sampling and testing aliquots by spectroscopy absorption method. Nanocrystals growth stopped by cooling reaction mixture on reaching desired shell thickness. The optical absorption spectra have been registered using Perkin-Elmer Lambda-35 spectrometer. CdSe QDs size and size of CdTe core of CdTe/CdSe heterostructures was estimated using data from [8] and found to be 3.5 nm and 3.6


nm corresponding with a weak size distribution (<8%). Shell thickness in CdTe/CdSe heterostructures was estimated with use earlier developed quantum-mechanical model. QDs doped polystyrene films were prepared on thoroughly dried glasses. For that QDs-hexane sol was diluted by hexane to desired concentration (~104-1010 QDs/ml) and dried. Then QDs precipitate was redispersed in solution of polystyrene dissolved in toluene and drop of this solution was deposited on glass and dried. Polystyrene concentration chose on purpose that film thickness be equal about 3-5 m. QDs concentrations in films vary from 104-1010 NC/cm2. QDs blinking studies were conducted on luminescence microscope with continuous wavelength 473 nm diode laser. Microscope optical scheme is shown in Fig.2. Excite laser beam was weakened by available filters system to reduce pumping power. Luminescence registration realized by liquid nitrogen cooled CCD-camera.

Photofluorescence and absorption of films with QDs were investigated by normal spectroscopic way. Some data for samples under investigation are in Table I
0,50 0,45 0,40 1,0 0,9 0,8

absorbance, a.u.

0,30 0,25 0,20 0,15 0,10 0,05 0,00 500 600 700 800

0,6 0,5 0,4 0,3 0,2 0,1 0,0 900

wavelength, nm

Figure 2. Set-up scheme

3 Results and Discussion

Figure 3. Photoluminescence spectra and absorption spectra of QDs: CdSe (solid lines), CdSe/CdS (dashed lines), CdTe/CdSe (dotted lines). The essential sample drift in microscope observation field was detected during investigation time of QD blinking. The drift scale was different, but the average nanoobject displacement was 4 µm/h (comparable with average distance between QDs ­ about 5 µm). Thus the drift gives some errors for time-dependent fluorescence measurements of QDs. By this reason, the special algorithm of determination of single nanoobjects displacement as time function was evolved. The sample moving map was obtained from QD luminescence microscope movie (see Fig.4) Nanoobejct coordinates were determinate independently in each averaged image, and then were connected by polynomial method. Thus, the trace of each nanoobject was obtained for time dependent fluorescence changes. One frame duration of fluorescence movie was 40 ms. If QD is in "ON" position, during one frame time detector records up to 104 photons. So, recorded fluorescence intensity depends not only of "ON" or "OFF" QD position, but of QD fluorescence quantum yield. For example, CdSe QDs were detected by best way with QD concentration 1 nanoobject per 1010 µ2 (or 106 QDs/cm2, film thickness c.a.5 µm, depth of focus c.a.4 µm). CdTe/CdSe QDs have a higher fluorescence quantum yield, so QDs concentration estimated c.a 4*104 QDs/2. Images of CdSe QDs films with high and low concentration are depicted on Fig.5

Fig.4. The sample drift (left part ­ y-displacement, right part x- displacement) as time function (scale in minutes) Black color ­ the luminescence maximum.

luminescence, a.u.

0,35

0,7


Table I. Operated QDs properties. Compound Exciton absobtion max exc nm CdSe 560 nm CdSe/CdS 605 nm CdTe/CdSe -

Fluorescence max, fl, nm 595 nm 615 mn 772 nm

Core size Rcore, е 35 35 36

Shell thickness Hshell, е 0 10 12

Overlap integral 2 0,99 0,98 0,4

Pair lifetime ns ~0,1 ~1 ~10

,

QDs quantity 6*1016 6*1016 ~1014

Fig.6. (top) ­ typical CdSe sample fluorescence intensity as function of time; (bottom) ­ the same after background noise substation, 4s average.

Fig.5. CdSe QD film with different QDs concentration: () ­ high (b) ­ low. The typical fluorescence intensity for CdSe/CdS QDs ("ON" and "OFF" time domains) is depicted on Fig.6. Three types QDs were analyzed. We integrated of fluorescence intensity by 5x5 pixels, because the single QD image is approximately 10 pixels. The CCD noise background was took into account (see Fig.6) The band gap of CdS is larger than CdSe one (see Fig.1), as result the CdS shell of CdSe/CdS QD has the function of potential barrier for electrons and holes, isolating these species from QD's surface. The CdSe shell of CdTe/CdSe QDs has other function. The band edge shift in conjugated polymer systems CdTe CdSe one relative the other one leads to electron localization in the shell mostly, if the shell thickness has an enough value. From other side, the hole is localized within core, thus the space charge carriers separation takes place. At the electron transfer into shell (type I/type II switching) the overlap integral of wave functions of electron and hole is decreasing in several times and determinate the considerable changes of optical properties of QD-system (increasing the lifetime of photoexcited charge carrier pair, depressing of excitonic absorption, increasing of luminescence quantum yield) with comparing the simple core/shell QDs. We compare mean integral luminescence intensity on long time scales and found the significant difference for various QDs types. So core/shell nanocrystals demonstrate intermediate behavior between CdSe core-type QDs and CdTe/CdSe core/shell heterostructures. Some of them demonstrate small angle of slope the same as simple QDs, while another have a higher angle of slope which typical for core/shell heterostructures. It could be marked that core-type QDs curves have a curly structure while core/shell heterostructures curves are rather smooth. We suspect that these differences are connected with different traps allocation in nearest QDs neighborhoods, which may be caused by different average size of various QDs types. Different energyband structure of various QDs types also should be taken in account.

Fig.7 Typical mean integral luminescence intensity for CdSe QDs (a) and CdTe/CdSe heterostructures. Y-axis scales are different for these draws. Also shown angle of slope on right edge of curves.

4 Conclusion
Synthesis conditions for different QDs types were suggested and original experimental set-up for luminescence studies of prepared QDs was used. The luminescence time dependence for two types of core/shell QDs (CdSe/CdS and CdTe/CdSe ) were investigated and were compared with behavior of naked ones (CdSe). We suspect that slope variation on long-time scale is connected with different traps allocation in nearest QDs neighborhoods, which may be caused by different average size of various QDs types. Detail analysis of mentioned variation needs more long time experiments.These experiments are in progress. Nevertheless, the obtained results open door for initial properties of semiconductor QDs extremely perspective for new optoelectronic device fabrication, like organic light emitting diodes of new generation..


5 Acknowledgement
This work was supported by the Russian Foundation for Basic Research (Projects No 06-02-16399, 06-02-17089, 07-

02-00873, 07-02-00656, and 06-02-08120-OFI), and by the Federal Agency on Science and Innovations (contracts No 02.513.11.3062, 02.513.11.3065, and No 02.513.11.3249 ­ "Support of Leading Scientific Schools of the Russian Federation").

References [1] S. Coe, W -K. Woo, M. Bawendi, and V. Bulovic, Nature (London) 420, 800, 2002; T. Tsutsui, Nature, 420, 752, 2002. [2] R. G. Neuhauser, K. T. Shimizu, W. K. Woo, S. A. Empedocles, M. G. Bawendi, Phys.Rev.Lett., 85, 3301, 2000. [3] K. T. Shimizu, R. G. Neuhauser, C. A. Leatherdale, S. A. Empedocles, W. K. Woo, M. G. Bawendi, Phys.Rev.B, 63, 205316, 2001. [4] P. A. Frantsuzov, R. A. Marcus, Phys.Rev.B, 72, 155321, 2005. [5] C. D. Heyes, A. Yu. Kobitski, V. V. Breus, G. U. Nienhaus, Phys.Rev.B, 75, 125431, 2007. [6] M. Nirmal, L. Brus, Acc.Chem.Res., 32, 407, 1999. [7] R.B.Vasiliev, S.G.Dorofeev, D.N.Dirin, D.A.Belov, T.A.Kuznetsova. Mendeleev Commun., 14(4), 169, 2004. [8] W. W. Yu, Y. A. Wang, X. Peng, Chem.Mater., 15, 4300, 2003.