School Work

A Fluorescence Correlation Spectroscopy, Steady-State, and TimeResolved Fluorescence Study of the Modulation of Photophysical Properties of Mercaptopropionic Acid Capped CdTe Quantum Dots upon Exposure to Light

A Fluorescence Correlation Spectroscopy, Steady-State, and TimeResolved Fluorescence Study of the Modulation of Photophysical Properties of Mercaptopropionic Acid Capped CdTe Quantum Dots upon Exposure to Light
of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  A Fluorescence Correlation Spectroscopy, Steady-State, and Time-Resolved Fluorescence Study of the Modulation of PhotophysicalProperties of Mercaptopropionic Acid Capped CdTe Quantum Dotsupon Exposure to Light Satyajit Patra and Anunay Samanta * School of Chemistry, University of Hyderabad, Hyderabad 500046, Andhra Pradesh, India * S  Supporting Information  ABSTRACT:  Light-induced modulation of the  󿬂 uorescence behavior of mercaptopropionic acid (MPA) capped CdTequantum dots (QDs) in aqueous solution is studied by acombination of   󿬂 uorescence correlation spectroscopy (FCS)and steady-state and time-resolved  󿬂 uorescence techniques.These investigations reveal a dramatic variation in the 󿬂 uorescence properties of the QDs under exposure to light.In the FCS measurement, a large decrease in amplitude andchange in shape of the correlation curves are observed withincreasing excitation power. The change in the shape of the correlation curves, particularly at short lag time, e.g., a fasterrelaxation at high excitation power, is attributed to the increasing contribution of the o ff   state of the QDs. Interestingly, despitethis increasing contribution of the o ff   state, which reduces the e ff  ective number of emitters in the observation volume and henceshould increase the amplitude of the correlation curve, the latter actually decreases at high excitation power. This apparentcontradiction is resolved by considering light-induced transformation of the dark QDs to bright QDs due to surface passivation of the QDs with increasing excitation power. Enhancement of the steady-state  󿬂 uorescence intensity under light irradiation, both inaerated and deaerated environments, supports the mechanism of passivation of the surface trap states by photoadsorption of  water molecules. Fluorescence lifetime data is also shown to be consistent with this light-induced surface passivation mechanism. 1. INTRODUCTION Semiconductor quantum dots (QDs) have been receivingincreasing attention in recent years because of their interesting size-dependent optical and electronic properties, 1 − 4  which isdue to quantum con 󿬁 nement of both electron and hole(produced because of electronic excitation) in all three dimensions. 1 ,4 − 7 Emission covering the entire visible andinfrared region can be obtained simply by tuning the size of  the QDs. 1 ,8  As the QDs exhibit broad absorption and narrow emission pro 󿬁 le in comparison to organic  󿬂 uorophores, 9 it ispossible to excite multiple QDs at a single excitation wavelength. 10 Furthermore, the narrow  emission pro 󿬁 le of the QDs provides easy spectral separation, 9 thus making theman e ffi cient system for multicolor imaging of biologicalsamples. 10 QDs also exhibit a higher photostability and superioroptical properties compared to conventional organic  󿬂 uoro-phores. 9  All these excellent properties make QDs idealcandidates for applications ranging from biological imaging tolasing to optoelectronics. 3 ,10 − 17 Given the huge potentialscienti 󿬁 c and technological impact of QDs in variousapplications, the  󿬂 uorescence response of QDs has remainedan intriguing topic of research since its discovery.The small size of Q  Ds results in a high surface-to-volumeratio of the substances, 1 ,8 and it has long been believed that thesurface of the QD plays an important role in determining itsluminescence properties. 5,18 − 25 Uncoordinated atoms on thesurface disrupt the crystalline periodicity and leave behind one or more dangling orbitals on each atom. 1 ,23 These danglingorbitals form the mid band gap states and reduce theluminescence e ffi ciency of the QDs b y providing additional nonradiative deactivation pathways. 1 ,8,26 ,27 Surface passivationis a crucial parameter for the preparation of  QDs with high 󿬂 uorescence quantum yield and photostability. 1 ,8 ,23,28 ,29 One of the major obstacles to the progress of the development of highly luminescent and stable QDs is the poorly understoodQD surface chemistry and the relation of the surface chemistry  to QD photophysical properties. 30 − 32 It is reported that the 󿬂 uorescence e ffi ciency  of  the QDs is enhanced under irradiation with light. 22 ,32 − 41 This enhancement of emissionon exposure to light, which is termed photoactivation, isattributed to photoinduced passivation of the surface trapstates. However, no general consensus on the mechanism of photoactivation has yet been reached because of thecomplicated photophysics and photochemistry of QDs.Suggested mechanisms include elimination of the topologicalsurface defects (i.e., smoothing of the surface during the Received:  July 18, 2013 Revised:  October 9, 2013 Published:  October 10, 2013 © 2013 American Chemical Society  23313  |  J. Phys. Chem. C   2013, 117, 23313 − 23321  process of photocorrosion), 26,32 ,34 ,42 − 44 passivation of thesurface trap states by photoadsorbed molecules, 22 ,32 ,35 ,38 photoinduced rearrangement of the surface stabilizing agents, 32,33 ,36,44 − 47 photoneutralization of  the local chargedcenters inside and outside the QDs, etc. 48 Solvents also have been found to play an important role in the photoactivation of  QDs. 32,35 ,38 ,48 Photoluminescence enhancement of QDs oftenstrongly depends on the amount of water vapor present in the atmosphere. 35 ,38,48 This is ascribed to photoadsorption of the water molecules on the QD surface leading to the passivation of  the nonradiative surface trap states. 32 ,35,38 ,48,49 Polarity of thesolvent also a ff  ects the photoactivation. 22 ,32,36 It is observedthat addition of methanol to trioctylphosphine oxide (TOPO)-hexane or TOPO-toluene solution shows an acceleratedincrease of the  󿬂 uorescence e ffi ciency compared to TOPO-toluene only or TOPO-hexane only QD solutions. 36 Hence, aclear understanding of the photoactivation mechanism isabsolutely essential to understand the role of surface statesand surface reaction on the luminescence yields and photo-stability of QDs. We have studied light-induced changes in the 󿬂 uorescence behavior of MPA-coated CdTe QDs in water by acombination of FCS, steady-state, and time-resolved spectro-scopic techniques. CdTe QDs are chosen as the subject matterof this investigation because the e ff  ect of light irradiation on theQD optical properties is not as widely investigated as that onCdSe QDs.It is known that FCS is a highly sensitive and powerfultechnique in which the  󿬂 uorescence  󿬂 uctuations arising from asmall observation volume (on the order of a femtoliter) iscorrelated to obtain the temporal evolution of the system aboutits equilibrium state. 50 This technique has been successfully applied to study the  󿬂 uorescence blinking dynamics and thekinetics of bimolecular reactions, binding of ligands to aprotein, protein − protein interaction, DNA hybridization, andconformational  󿬂 uctuation and translational di ff  usion of the 󿬂 uorescent pro bes in polymer, lipid vesicles, micelles, ionicliquids etc. 51 − 61 Considering that QDs are superior  󿬂 uorescentprobes compared to organic  󿬂 uorophores, the FCS techniquehas been used for the characterization of QDs; speci 󿬁 cally, ithas been used to determine their hydrodynamic radius,concentration, monodispersity, and aggregation ten- dency. 12 ,37 ,62 − 68 Interestingly, only a handful of FCS studieson QDs photophysics have been made despite the potential of this technique to determine the blinking kinetics at faster timescales. 51 ,52,69 − 74  Widely distributed kinetics of   󿬂 uorescence blinking 75 of  QDs limited its study using the FCStechnique. 70 ,71,73 The blinking kinetics of QDs complicatesthe analysis of the FCS data as it distorts the shape of thecorrelation curve, especially at shorter correlation times, 73 thusmaking it di ffi cult to model QD blinking dynamics. Doose et al.attempted to simulate the anomalous shape of the correlationcurve by employing Monte Carlo calculations of the di ff  usingand blinking dots. 73 However, no unique set of blinkingparameters for a given data set could be found. As theamplitude of the correlation curve,  G (0), provides informationon the number of emitters in the observation volume, acomparison of the  G (0) value as a function of the laser power isexpected to provide a comprehensive understanding of thelight-induced changes in QDs. Larson et al. observed a decreasein the  G (0) value with increasing excitation power by two-photon excitation FCS and attributed this to the broadening of the observation volume on excitation saturation. 12 Doose et al.compared the photophysical and colloidal properties of some biocompatible QDs using FCS. 73 They examined the  G (0) values at di ff  erent excitation power and compared them withthose obtained for the  󿬂 uorescence beads and rhodamine 6G.The decrease of the  G (0) value for the QDs is found to bemuch higher than that of the beads and rhodamine 6G and isattributed to excitation saturation and change in the blinkingstatistics at high excitation power, which increases theconcentration of the emitters in the observation volume.However, they could not model the blinking statistics of theQDs as a function of the excitation power. Miyasaka and co- workers could  󿬁 t the blinking dynamics of water-soluble CdTeQDs by introducing a stretched exponential term in thecorrelation function. 52 This stretched exponential term takescare of the widely distributed kinetics of the blinking of QDs.These authors also observed a decrease in the  G (0) value and afaster decay of the autocorrelation with increasing excitationpower. They agreed to the point that excitation saturation alonecould not explain the changes in the shape and amplitude of thecorrelation curves and ascribed the observation to a fasterrelaxation from the dark state at higher excitation power. Theseauthors, however, did not compare the fraction of the dark stateas a function of the excitation power, which would haveprovided more information on the changes in the shape of thecorrelation curves. Dong et al. attributed the decrease of   G (0) value with irradiation times to the photoactivation of theQDs. 37  According to them, photoactivation occurred because of laser-induced aggregation of the QDs and consequentmodi 󿬁 cation of the QD surface structure, which turns thepermanently dark QDs to bright QDs. 37 It is therefore evidentthat no single mechanism can explain light-induced variation of the luminescence e ffi ciency of the QDs and determination of the mechanism of photoactivation requires a thorough anddetailed investigation. This explains the motivation for thepresent study. We have also carried out steady-state and time-resolved conventional emission studies to supplement theresults of the FCS measurements and also to obtain a clearunderstanding of the mechanism of this light-induced change inthe luminescence properties of the QDs. To the best of ourknowledge, this is the  󿬁 rst study in which the origin of decreasing amplitude of the correlation curve with increasingexcitation power is explained clearly. We have also successfully monitored the evolution of the blinking parameters as afunction of the excitation power to provide support for themechanism involved in the photoactivation process. 2. EXPERIMENTAL SECTION 2.1. Materials.  Cadmium acetate dihydrate((CH 3 COO) 2 Cd · 2H 2 O) and tellurium (Te) powder for thesynthesis of QDs were obtained from local suppliers.Hexadecylamine (HDA), trioctylphosphine (TOP), 3-mercap-topropionic acid (MPA), and rhodamine 123 (R123) werepurchased from Sigma Aldrich. Potassium hydroxide (as  󿬂 akes) was purchased from Merck. Solvents such as methanol andchloroform were procured from Merck and dried following areported procedure prior to their use in sample preparation andspectral measurements. 76  All the experiments were carried outat 25  ° C. Milli-Q water was used in the present study. 2.2. Synthesis of CdTe/HDA QDs.  CdTe/HDA QDs wereprepared following a reported procedure. 77 Brie 󿬂  y, 5 g of HDA and 3 mL of TOP were taken in a two-necked, round-bottom(RB)  󿬂 ask and heated at 80  ° C for 15 min to bring the mixtureto a liquid state. In the mean time, a separate solution wasprepared in a reagent bottle containing 0.41 g of  The Journal of Physical Chemistry C  Article  |  J. Phys. Chem. C   2013, 117, 23313 − 23321 23314  (CH 3 COO) 2 Cd · 2H 2 O and 0.16 g of Te powder in 4 mL of TOP. The mixture was sonicated for 1 h to obtain an almostclear solution. This solution was then quickly injected into a RB 󿬂 ask containing HDA and TOP. The reaction was carried outat 180  ° C for several minutes so that the desired size of theCdTe core is reached. Because the luminescence of the QDdepends on its size, the growth of the particle was monitored by its  󿬂 uorescence at regular intervals of the reaction time. Aftercompletion of the reaction, excess starting materials wereremoved by washing the QDs with methanol followed by repeated precipitation and centrifugation. Then the QDs weredissolved in nonpolar solvents like CHCl 3. 2.3. Synthesis and Characterization of Water-SolubleCdTe/MPA QDs.  Water-soluble CdTe/MPA QDs wereprepared from CdTe/HD A  by a ligand exchange method asreported in the literature. 78  A 0.5 M methanolic solution of MPA-KOH (20 mol % excess KOH) was added dropwise to astirring solution of CdTe/HDA QDs in CHCl 3  until the QDs 󿬂 occulated out of solution. The solution was then centrifugedto separate the precipitate, which was easily soluble in water asthe exchange of HDA with MPA made the outer layer of theQDs polar. This transfer of the QDs from CHCl 3  to water wascarried out under a nitrogen environment. The stronger binding capability of the thiol with surface Cd atoms helpedthe replacement of HDA by MPA. Thus, synthesized CdTe/MPA QDs, which exhibited  󿬁 rst excitonic absorption maximumat 600 nm and emission peak at 635 nm, were characterized by UV  −  vis absorption,  󿬂 uorescence, and transmission electronmicroscopy (TEM) measurements. The core size of this QD was estimated to be 4.7 nm from the TEM measurements (seeFigure S1 of the Supporting Information for TEM images). 2.4. Instrumentation.  FCS measurements were performedusing a time-resolved confocal  󿬂 uorescence microscope(MicroTime 200, PicoQuant). An inverted microscope(Olympus IX71) equipped with a water immersion objective(UPlansApo NA 1.2, 60 × ) served as the microscope body. QDs were excited at 485 nm from a pulsed diode laser (PDL 828 SSepia II, PicoQuant) with full width at half-maximum of 144 ps.The laser repetition rate was 20 MHz. The output of thispulsed diode laser was coupled to the main optical unit by apolarization maintaining single-mode optical  󿬁  ber. In the mainoptical unit the excitation light was guided through a dichroicmirror, which re 󿬂 ected the excitation light onto the entranceport of the microscope. The water immersion objective of themicroscope focused the collimated laser beam onto the drop of CdTe/MPA QDs sample in water placed on a coverslip.Fluorescence from the QD sample was collected by the sameobjective, passed through a dichroic mirror and 510 nm longpass  󿬁 lter to block the excitation light, and then focused onto a50  μ m pinhole to cut o ff   the out-of-focus light. Then therecollimated  󿬂 uorescence signal was directed onto a (50/50) beam splitter prior to entering two single-photon avalanchephotodiodes (SPADs). By cross-correlating the signals from thetwo detectors,  󿬂 uorescence correlation traces were generated.Data acquisition was performed with a PicoHarp 300 TCSPCmodule in a time-tagged time-resolved (TTTR) mode.Correlation f unction for the  󿬂 uorescence intensity   󿬂 uctuationsis given by  50 τ  δ δ τ  = ⟨ + ⟩⟨ ⟩ GF t F t F t  ( ) ( ) ( )( )  2 (1)  Where  ⟨ F  ( t  ) ⟩  is the average  󿬂 uorescence intensity.  δ  F  ( t  ) and δ  F  ( t   +  τ  ) are the  󿬂 uctuations from the average  󿬂 uorescenceintensity at time  t   and  t   +  τ   ,respecetively, and are given by  δ τ τ  = − + = + − F t F t F t F t F t  ( ) ( ) F(t) , ( ) ( ) ( ) (2)  All the correlation curves were  󿬁 tted to a three-dimensional(3D) di ff  usion model with a stretched exponential term of  which the analytical expression is represented by  τ  τ τ τ τ τ κ τ  = +−− ++  β   −− ⎡⎣⎢⎢⎛⎝⎜⎞⎠⎟⎤⎦⎥⎥⎛⎝⎜⎞⎠⎟⎛⎝⎜⎞⎠⎟ G T T N  ( ) 11exp 111 T D12D1/2 (3)  where  τ  D  is the QD di ff  usion time,  N   the average number of molecules undergoing reversible transition between on and o ff  state in the observation volume,  T   the fraction of the o ff   state,and  τ  T  the dark-state relaxation time.  β   is the stretchingexponent with a value between 0 and 1 and is related to thedistribution of   τ  T .  κ   (which is equal to  ω  z / ω xy ) is the structureparameter of the observation volume;  ω  z  and  ω xy  are thelongitudinal and transverse radii, respectively. The structureparameter of the observation volume is calibrated usingrhodamine 123 in water of known di ff  usion coe ffi cient 460  ± 40  μ m 2 /s. 79 FCS measurements were carried out for di ff  erentconcentrations (30, 55, 64, and 270 nM) of the QDs. Theconcentrations were calculated by   󿬁 rst estimating the molarextinction coe ffi cient from the measured size of the QD usingthe empirical formula of Peng and co-workers 80 and then usingthe measured optical density of the solution. Excitation power was varied between 6  μ  W and 392  μ  W. For each excitationpower, a fresh drop of CdTe/MPA solution in water was kepton the coverslip in order to avoid evaporation of water at higherexcitation power. Each measurement was repeated at least 20times. The size of the error was calculated from the deviationfrom the mean value.Steady-state absorption and emission spectra were recordedon a Cary 100, Varian UV  −  vis spectrophotometer and aFluorolog 3, Horiba Jobin Yvon spectro 󿬂 uorimeter, respec-tively. Fluorescence lifetimes of the QDs at di ff  erent lightirradiation times were measured using a time-correlated single-photon counting (TCSPC) spectrometer (Horiba Jobin YvonIBH). Nano LED (  λ ex   = 439 nm) was used to excite the sampleand an MCP photomultiplier (Hammatsu R3809U-50) wasused as the detector. The pulse repetition rate and pulse widthof the excitation source were 1 MHz and 150 ps, respectively.The lamp pro 󿬁 le was recorded by placing a dilute solution of Ludox in water as a scatterer in the sample chamber. The 󿬂 uorescence decay curves were  󿬁 tted to a multiexponentialequation using the IBH DAS6 (Version 2.2) decay analysissoftware. QDs were irradiated with FL8 D daylight 8 W  󿬂 uorescent tube lamp (Toshiba) for di ff  erent exposure timesfor steady-state absorption, emission, and time-resolvedemission studies. The concentration of the QDs for steady-state absorption, emission, and  󿬂 uorescence lifetime studies wasmaintained at 3.3  ×  10 − 7 M. The samples for TEMmeasurements were prepared by placing a drop of clearaqueous solution of the QDs on carbon-coated copper gridsfollowed by removal of the solvent under high vacuum. Thesize of the QDs was determined using a Tecnai G2 FE1 F12 The Journal of Physical Chemistry C  Article  |  J. Phys. Chem. C   2013, 117, 23313 − 23321 23315  transmission electron microscope at an accelerating voltage of 200 kV. 3. RESULTS AND DISCUSSION 3.1. FCS Study of CdTe/MPA QDs in AqueousSolution.  Figure 1 shows the FCS curves of the 55 nMCdTe QDs in water and  G (0) values at di ff  erent excitationpower. The correlation curves could be best  󿬁 tted (asdetermined by the residuals) to a stretched exponential witha 1-component di ff  usion model (eq  3; the quality of the  󿬁 ts toother models is provided in Figure S2 of the SupportingInformation). This stretched exponential  󿬁 t is in accordance with the distributed kinetics of   󿬂 uorescence blinking of theQDs, as reported in the literature. 52 ,70,75 It is important to notethat the amplitude of the correlation curve decreases sharply  with increasing excitation power (Figure 1). The  G (0) valuedecreases from 0.98  ±  0.13 at 6  μ  W to 0.30  ±  0.12 at 186  μ  W. A further increase in power, however, slightly increases the G (0) value (0.45  ±  0.12 at 392  μ  W). This decrease in  G (0) value with increasing excitation power for CdTe/MPA is muchlarger than that for R123 in water (Figure S3 of the SupportingInformation). Figure 2 shows the normalized correlation curves and the dependence of the di ff  usion time on increasingexcitation power. The shape of the correlation curves is strongly dependent on the excitation power especially at shortcorrelation times. 73 ,76 Figure 3 shows the dependence of theo ff  -state fraction ( T  ) and the blinking time ( τ  t ) on increasingexcitation power.  T   increases and  τ  t  decreases with increasingexcitation power and attains a saturation value at around 186  μ  W.These results suggest that the blinking kinetics becomesfaster and dominant with increasing excitation power. At highexcitation power blinking kinetics dominates the entirecorrelation curve. This is why the correlation curves are shiftedto shorter correlation time with increasing excitation power. Asthe amplitude of the correlation curve depends on the numberof   󿬂 uorescent molecules in the observation volume,  G (0) = 1/  N  (1  −  T), 50  where,  N   is the average number of molecules inthe observation volume undergoing reversible transition between  󿬂 uorescent on and o ff   state, an increase in  T   impliesa decrease in the number of   󿬂 uorescent molecules  N  (1 − T  ) inthe confocal volume. Hence, the amplitude of the correlationcurves should increase with increasing excitation power. Figure 1. ( a) Correlation curves of the 55 nM CdTe/MPA QDs inaqueous media at di ff  erent excitation power. Points are the data andlines represent  󿬁 t to eq  3 and (b) change of the  G (0) value withexcitation power. Excitation wavelength is 485 nm. Figure 2.  (a) Normalized  󿬁 tted correlation curves of the 55 nMCdTe/MPA QDs in aqueous media for di ff  erent excitation power. (b) A plot of the measured di ff  usion time of the QDs against excitationpower. Figure 3.  Variation with excitation power of the estimated (a) dark-state fraction ( T  ) and (b) blinking time ( τ  t ), obtained from the  󿬁 t toeq  3. The Journal of Physical Chemistry C  Article  |  J. Phys. Chem. C   2013, 117, 23313 − 23321 23316  However, a completely opposite observation is made in theexperiment. A decrease in the  G (0) value is commonly rationalized considering the broadening of the observation  volume due to excitation saturation that increases  N  . 12 ,70 ,73 However, the changes in di ff  usion time in the low excitationpower regime (Figure 2) is almost negligible, and in this regimethe decrease of the  G (0) value is huge. A decrease in the  G (0) value after 100  μ  W excitation power can be rationalized as dueto excitation saturation. However, the decrease of the  G (0) value in the low excitation regime is certainly not due toexcitation saturation and hence some other processes must beinvolved. Because the decrease of the  G (0) value is much largerin the case of QDs compared to R123, an increase of  background intensity with increasing excitation power is notresponsible for this observation (Figure S3 of the SupportingInformation). As stated in the Introduction , Doose et al. assigned a decrease in the amplitude of the correlation curve toexcitation saturation and blinking 73  but could not model theexcitation power-dependent blinking of the QDs. We havesuccessfully modeled the excitation power-dependent blinkingkinetics by   󿬁 tting the data to a stretched exponential model andfound out that this decrease in the  G (0) value is not due to blinking (as it increases the fraction of the o ff   state) (Figure 3).Under these circumstances, this decrease of the  G (0) value can be explained only if   N   increases by exposure to light. Thissuggests light-induced brightening of the otherwise dark particles (which does not  󿬂 uoresce), and this process isenhanced with an increase in e xcitation power, a processcommonly termed photoactivation. 35 In this context, it is to benoted that Dong et al. attributed a decrease of the  G (0) valueunder laser irradiation of di ff  erent times to photoactivation of the quantum dots 37 resulting from facile aggregation of theQDs that led to a shift of the correlation curves to longercorrelation times. 37 However, in our study we did not observeany increase in the di ff  usion time in the lower excitation powerregion where the decrease of   G (0) value is huge (Figures 1 and2). Hence, even though photoactivation is responsible for thedecrease of the  G (0) value upon increasing excitation power, itis clearly not due to laser-induced aggregation. The increase inthe  G (0) value from 0.30  ±  0.12 (at 186  μ  W) to 0.45  ±  0.12(at 392  μ  W) (Figure 1) is clearly due to photobleaching, whichdecreases the number of   󿬂 uorescent molecules in theobservation volume. This fact is also supported by the decreasein the di ff  usion time from 2.01  ±  0.16 ms (at 186  μ  W) to 1.76 ±  0.15 ms (at 392  μ  W) (Figure 2) due to a decrease in theresidence time of the QDs because of photobleaching. An important point to note here is that for low excitationpower, the number of QDs in the observation volume (10 − 15 L) estimated from the FCS measurement (  N  FCS ) is much lowerthan the actual number of QDs (  N  actual ) present in this volume.For example, for a concentration of 55 nM, the actual numberof QDs in the observation volume is 33. However, the numberof QDs calculated from the measured  G (0) value of 0.98  ±  0.13and an o ff  -state fraction of 0.35 is only 1.56  ±  0.25, indicatingthat a large fraction of the QDs remain in their dark state(absorb light but do not emit). The experiments carried out for various QD concentrations, the results of which are collected inTable 1 , show a very similar behavior. That the concentrationmismatch in the case of the QDs is indeed due to the dark fraction is further substantiated by the fact that for themolecular system, R123, no such mismatch is observed (FigureS11 of the Supporting Information). With increasing excitationpower, as the dark fractions are turned into bright ones, the G (0) value decreases despite an increase in the o ff  -state fractionunder this condition. At even higher power, when almost all theQDs are turned bright and participate in the on − o ff   transition,the saturation of the  G (0) value is observed. 3.2. Steady-State Behavior of the QDs at Di ff  erentIllumination Times.  To identify the process(es) responsiblefor the change in luminescence behavior of the QDs, theexperiments have also been performed under steady-stateconditions for di ff  erent irradiation periods using deoxygenatedsolutions of QDs (prepared by nitrogen bubbling of thesolution). The e ff  ect of exposure to light on the UV  −  visabsorption spectrum also was studied. A noticeable decrease of the absorbance throughout the entire absorption band isobserved with increasing exposure time (Figure S12 of theSupporting Information). As the absorption peak position isnot well-resolved, it is di ffi cult to  󿬁 gure out from the data whether light irradiation leads to any shift of the absorptionmaximum. Figure 4 depicts the e ff  ect of illumination of anaqueous solution of the QDs for di ff  erent exposure times onthe integrated emission intensity (  I  em ) and wavelengthcorresponding to the emission peak (  λ emmax  ). The illuminationleads to a nearly 3-fold enhancement of   I  em  up to a certain time(350 min under experimental conditions) beyond which theemission intensity decreases rapidly. During this irradiationprocess,  λ emmax  decreases slightly at the early stages, but itdecreases sharply after a certain time (which coincides with thetime when the  I  em  value starts dropping). To examine any possible role of oxygen on these results, the aqueous solution of the QDs in the cuvette was purged with nitrogen and theexperiments were repeated in an oxygen-free environment.These results are also shown in Figure 4. As can be seen, also indeaerated condition the initial luminescence enhancement onexposure to light is observed, but the subsequent rapidreduction of   I  em  that is observed in aerated condition is notpresent. In the N 2 -purged solution, the  λ emmax   value, unlike that inthe other case, also remains more or less constant (emissionspectra of CdTe/MPA in aerated and deaerated solutions atdi ff  erent exposure times are given in Figure S13 of theSupporting Information). These  󿬁 ndings clearly exhibit thedistinct role of oxygen in the rapid reduction of   I  em  and the blueshift of   λ emmax  observed following post-maximization of the Table 1. Comparison of the  G (0) Value, O ff  -State Fraction ( T  ), Number of Particles in the Observation Volume (10 − 15 L)Obtained from the FCS Measurement (  N  FCS ), and Actual Number of Particles in this Volume Determined from theConcentration (  N  actual ) and Dark-State Fraction at Di ff  erent Concentrations of the QDs in an Aqueous Solution for anExcitation Power of 6  μ  W  concentration (nM)  G (0)  N  (1  −  T  )  T N  FCS  N  actual  dark fraction30 3.0770  ±  0.3100 0.3250  ±  0.030 0.42  ±  0.04 0.56  ±  0.05 18  ±  1 0.970  ±  0.00255 0.9800  ±  0.1300 1.020  ±  0.150 0.35  ±  0.04 1.56  ±  0.25 33  ±  2 0.950  ±  0.00264 0.8293  ±  0.0800 1.201  ±  0.080 0.34  ±  0.05 1.83  ±  0.13 38  ±  2 0.950  ±  0.002270 0.1462  ±  0.0100 6.840  ±  0.440 0.32  ±  0.04 10.04  ±  0.69 162  ±  6 0.940  ±  0.004 The Journal of Physical Chemistry C  Article  |  J. Phys. Chem. C   2013, 117, 23313 − 23321 23317
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks