Effect of shell thickness and composition on blinking suppression and the blinking mechanism in ‘giant’ CdSe/CdS nanocrystal quantum dots

FULL ARTICLE

Effect of shell thickness and composition on blinkingsuppression and the blinking mechanism in ‘giant’CdSe/CdS nanocrystal quantum dots

Javier Vela1, Han Htoon2, Yongfen Chen3, Young-Shin Park2, Yagnaseni Ghosh2,Peter M. Goodwin4, James H. Werner4, Nathan P. Wells5, Joanna L. Casson2,Jennifer A. Hollingsworth*; 2

1 Department of Chemistry, Iowa State University, Ames, Iowa 50011, USA2 Chemistry Division and the Center for Integrated Nanotechnologies, Los Alamos National Laboratory, Los Alamos,

New Mexico 87545, USA3 Life Technologies, Eugene, Oregon 97402, USA4 Materials Physics and Applications Division, Center for Integrated Nanotechnologies, Los Alamos National Laboratory,

Los Alamos, NM 87545, USA5 LIDAR and Atomic Clocks, Photonics Technology Department, Aerospace Corporation, El Segundo, CA 90245, USA

Received 20 April 2010, revised 11 June 2010, accepted 19 June 2010Published online 12 July 2010

Key words: nanocrystalline materials, quantum dots, fluorescence, giant nanocrystal quantum dot, CdSe/CdS, blinkingsuppression, single particle tracking, bioimaging and microscopy, advanced optical microscopy, advanced spectroscopy

Æ Supporting information for this article is available free of charge under http://dx.doi.org/10.1002/jbio.201000058

# 2010 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Journal of

BIOPHOTONICS

We recently developed an inorganic shell approach forsuppressing blinking in nanocrystal quantum dots(NQDs) that has the potential to dramatically improvethe utility of these fluorophores for single-NQD trackingof individual molecules in cell biology. Here, we consid-er in detail the effect of shell thickness and compositionon blinking suppression, focusing on the CdSe/CdS core/shell system. We also discuss the blinking mechanism asunderstood through profoundly altered blinking statis-tics. We clarify the dependence of blinking behavior andphotostability on shell thickness, as well as on interro-gation times. We show that, while the thickest-shell sys-tems afford the greatest advantages in terms of en-hanced optical properties, thinner-shell NQDs may beadequate for certain applications requiring relativelyshorter interrogation times. Shell thickness also deter-mines the sensitivity of the NQD optical properties toaqueous-phase transfer, a critical step in renderingNQDs compatible with bioimaging applications. Lastly,we provide a proof-of-concept demonstration of the uti-lity of these unique NQDs for fluorescent particle track-ing.

High-resolution image of an ultra-thick-shell ‘giant’ na-nocrystal quantum dot (left). Suppressed blinking behav-ior afforded by this class of semiconductor nanocrystalyields new statistical relationships in the probabilitydensities of fluorescence on- and off-time distributions(right).

* Corresponding author: e-mail: [email protected], Phone: +01 505 665 1246, Fax: +01 505 665 0743

J. Biophotonics 3, No. 10–11, 706–717 (2010) / DOI 10.1002/jbio.201000058

1. Introduction

In many respects, semiconductor nanocrystal quan-tum dots (NQDs) are near-ideal light emitters forapplications as fluorescent molecular probes and inadvanced bioimaging. Their optically excited emis-sion is efficient (quantum efficiencies can approachunity), narrow-band, and NQD-size-tunable, i.e.,light-output colors are precisely tunable from theultraviolet through the visible and into the infrareddepending on NQD composition and size. Absorp-tion is quasi-broadband, making excitation facile andallowing for the possibility of using a single exci-tation source to excite multiple NQDs for a rangeof emission colors (e.g., for optical barcoding [1]).Furthermore, NQDs are synthesized using solution-phase approaches that render NQDs with “mole-cule-like” properties, including solubility and che-mical manipulability. Despite these enabling charac-teristics, conventional NQD optical properties aresensitive to NQD surface chemistry and chemical en-vironment, which has three important results: (1)high “as-prepared” NQD photoluminescence quan-tum yields are often not maintained when nonpolar-soluble NQDs are transferred into water, (2) NQDs“bleach” (emission intensity degrades over time,though significantly less readily than organic fluoro-phores), and (3) NQDs at the single-particle level“blink” (exhibit fluorescence intermittency), all lim-iting total NQD “brightness”. Importantly, NQDblinking affords difficulties in using conventionalNQDs for single-particle/single-molecule trackingapplications.

We recently reported that these key NQD opticalproperties – quantum yield, photobleaching andblinking – can be rendered independent of NQDsurface chemistry and chemical environment, withphotobleaching and blinking significantly suppressed,by growth of a very thick, defect-free inorganic shell[2–4]. Importantly, this ability to suppress blinkinghas been specifically recognized as a key advance inthe application of NQDs as probes for biologicallyrelevant single-molecule tracking studies [5, 6]. Toachieve this advance, starting with 3–4 nm CdSeNQD cores, we grew the particles to a size of �10–20 nm by sequentially applying monolayers of inor-ganic shells. The shell monolayers – typically CdS,but in one case CdxZnyS alloys followed by ZnS(for enhanced relevance to biological imaging ap-plications, separating the Cd-containing region fromthe surface environment) – were grown onto CdSecores using modified literature procedures basedon a successive ion layer absorption and reaction(SILAR) method. Effectively, we isolated the wave-function of the NQD core from its surface, creatinga colloidal NQD that is structurally more akin tophysically grown epitaxial QDs. We named this newfunctional class of NQD the “giant” NQD (g-NQD).

Notably, our prior analyses showed that g-NQDs donot photobleach over long observation times and arecharacterized by significantly suppressed blinking.Here, we detail the effect of shell thickness on blink-ing behavior and provide an in-depth analysis ofblinking statistics. We also provide a biologically re-levant demonstration of the new utility afforded bylargely non-blinking and robust NQD fluorophores.

2. Experimental

2.1 General synthesis approach

Recently, core/shell growth techniques have been re-fined to allow for precise control over shell thicknessand shell monolayer additions. Specifically, a techni-que developed originally for the deposition of thin-films onto solid substrates – known as successiveion layer adsorption and reaction (SILAR) – wasadapted for NQD shell growth [7, 8]. By this meth-od, homogenous nucleation of the shell compositionis largely avoided and higher shell-growth tempera-tures are tolerated because the cationic and anionicspecies do not coexist in the growth solution. Thismethod has allowed for growth of thick shells, com-prising many shell monolayers, without loss of NQDsize monodispersity and with superior shell crystal-line quality. Originally demonstrated for a single-composition shell (CdS over CdSe) up to five mono-layers thick [8, 9], the approach has been extendedto multishell architectures [10], and now to thick(�7–9 shell monolayers) [11, 12] and ‘ultra-thick’shell systems (>10 monolayers) [2, 3]. The resultingnanocrystals are highly crystalline, uniform in shape,and electronically well passivated [2, 3, 8–12].

The specific synthetic methods followed for thepreparation and characterization of ‘giant’ CdSe/CdSand CdSe/alloyed-shell NQDs with suppressed blink-ing and remarkably enhanced stability are as fol-lows:

Synthesis of CdSe-based giant-shell NQDsMaterials: Cadmium oxide, oleic acid (90%), 1-octa-decene (ODE, 90%), dioctylamine (95%), octa-decylamine, zinc oxide, sulfur, selenium pellet(�99.999%), trioctyl-phosphine (TOP), mercapto-succinic acid, and tetramethylammonium hydroxidewere purchased from Aldrich and used withoutfurther purification. Trioctylphosphine oxide (TOPO)(90%) was purchased from Strem and used withoutfurther purification.

The synthesis of giant CdSe/multishell NQDs wasbased on a SILAR approach with modifications [8,10]. The CdSe core was prepared by injection of1 mL 1.5 M Se-TOP solution into a hot solution con-taining 1.5 g octadecylamine, 0.5 g TOPO, 5 g octa-

J. Biophotonics 3, No. 10–11 (2010) 707

FULLFULLARTICLEARTICLE

# 2010 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheimwww.biophotonics-journal.org

decene, and 0.2 mmol Cd-oleate under standard air-free condition. After injection of Se-TOP at 290 �C,the temperature was set at 250 �C for CdSe NQDgrowth. After ten minutes, the solution was cooleddown to room temperature, and CdSe NQDs werecollected by precipitation with acetone and centrifu-gation. CdSe core NQDs were re-dispersed in hex-ane and their concentration determined from theiroptical density [13].

1.5 � 10�7 mol CdSe core NQDs in hexane wereput into a 100 mL flask with 3 g ODE and 3 gdiooctylamine. 0.2 M elemental sulfur dissolved inODE, 0.2 M Cd-oleate in ODE and 0.2 M ZnxCd1–x-oleate (x ¼ 0.13, 0.49, 0.78, respectively) were usedas precursors for shell growth. The quantity of pre-cursors for each monolayer of shell was calculatedaccording to the volume increment of each mono-layer shell, considering the changing total NQD sizewith each successive monolayer grown. The reactiontemperature was set at 240 �C. Growth times were1 h for sulfur and 3 h for the cation precursors.

CdSe g-NQDs were transferred into water bystirring purified NQDs (�5 � 10�9 mol) in hexanewith 1 mmol mercaptosuccinic acid in 5 mL deionizedwater overnight. Mercaptosuccinic acid was neutra-lized by tetramethylammonium hydroxide in water.The pH of the water was �7. Mercaptosuccinic acid-capped g-NQDs were collected by centrifugation, andwere then re-dispersed in a small amount of waterand precipitated again using an excess of methanolto remove excess mercaptosuccinic acid. Finally, thepurified mercatosuccinic acid-capped g-NQDs weredispersed in deionized water to form optically clearsolutions.

2.2 Ensemble and single-NQDcharacterization

Ensemble NQD absorption and emission spectrawere recorded on a CARY UV-VIS-NIR spectro-photometer and a NanoLog fluorometer, respec-tively. Quantum yields (QYs) for the g-NQDs andthe various NQD control samples in hexane weremeasured by comparing the NQD emission with thatof an organic dye (Rhodamine 590 in methanol).The excitation wavelength was 505 nm and emissionwas recorded from 520 nm to 750 nm. The QY ofRhodamine 590 was taken to be 95%, and those forthe NQD samples were calculated by comparing theemission peak areas of the NQDs with the knowndye solution [14]. Specifically, the NQD QYs werecalculated using the formula:

QYNQDs ¼Absdye /AbsNQDs * Peak area of NQDs/Peak area of dye * QYdye* (RIdye

2/RINQDs2) (1)

RIdye – refractive index of dye solution in methanol,¼ 1.3284 (2)

RINQDs – refractive index of CdSe NQD solution inhexane, ¼ 1.3749 (3)

The absorbance of the dye and the CdSe NQD solu-tions were controlled between 0.01 and 0.05 opticaldensity. The absorbance and emission for each sam-ple were measured at least twice at no less than twodifferent concentrations.

For blinking studies, freshly diluted (�0.1–50 pM)g-NQDs in either HPLC-grade toluene or deionizedwater were dispersed onto a clean quartz slide.“Slow” blinking data (temporal resolution �200 ms¼ 100 ms integration time þ 100 ms readout time)were obtained using a standard wide-field micro-photoluminescence (micro-PL) experiment, wherethe data set was collected over an area of 40 � 40 mmfrom multiple NQDs simultaneously. The excitationsource was a 532 nm, 180 mW cw laser focused to�50 mm diameter spot, and emission was monitoredusing a liquid-nitrogen-cooled charged-coupled de-vice (CCD).

Importantly, for all analyses here, we set the “on/off” threshold of the NQDs at 10 counts above thereadout noise level of the CCD detector. We notethat this level is arbitrarily defined based on the sig-nal level of the CCD detector and does not corre-spond to the real “off level.” In fact, our recent stu-dies using a more sensitive time-correlated single-photon system have shown that more than 80% theg-NQDs with shell thickness >12 monolayers –although exhibiting wide fluctuations in PL emissionintensity – never turn into a completely off-state,even at faster time scales of 10 ms (Supporting Infor-mation Figure S1). Thus, the “off-levels” of the CCDstudy described here likely correspond, instead, to aweaker emitting state in the case of very-thick-shellg-NQDs (>12 monolayers), rendering our analyseshighly conservative in our estimates of “non-blink-ing” g-NQD fractions.

2.3 Correlated single-NQD atomic-forcemicroscopy (AFM) and photoluminescence(PL) study

A dilute solution of NQDs was drop-cast onto aclean coverslip, and the solvent was allowed to eva-porate. The sample coverslip was mounted on thestage of an inverted optical microscope (OlympusIX71) equipped with a piezoelectric scanner (PhysikInstrumente P773.3CD XYZ) used to raster scanthe sample for photoluminescence imaging. Excita-tion was provided by a 440 nm pulsed diode laser(PicoQuant LDH-P-C 440) operating at a pulse re-

J. Vela et al.: Shell thickness and composition effects on blinking in ‘giant’ nanocrystal quantum dots708

Journal of

BIOPHOTONICS

# 2010 by WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.biophotonics-journal.org

petition rate of 40 MHz and pulse width of �100 pi-coseconds. The average power over the observationarea (10 �10 mm) was 1 mW. Emitted fluorescencewas collected by the same microscope objective anddirected onto a single-photon counting avalanchephotodiode detector (APD) (Perkin Elmer SPCM-14AQR). The emission was spatially filtered using a75 micron diameter pinhole located in the imageplane of the microscope and spectrally filtered usinga bandpass filter centered at 605 nm before reach-ing the detector. The output of the APD was di-rected to a time-correlated single-photon countingmodule (PicoQuant PicoHarp 300) that controlledthe piezoelectric stage and recorded the time-corre-lated single-photon data. The photon data was pro-cessed using vendor-supplied software (PicoquantSymphotime) to generate the photoluminescence im-ages. An atomic force microscope (Veeco Instru-ments Bioscope SZ), mounted on the stage of theinverted optical microscope, was used to record nan-ometer scale topography images of the g-NQDs. Theimages were recorded using silicon nanoprobes(Veeco RTESP) operated in tapping mode at a reso-nance frequency of approximately 300 kHz. Regis-tration of the AFM height image with the photolu-minescence image was accomplished by aligning theAFM tip at the center of its scan range with the op-tical probe region. This was done by monitoring theexcitation laser light scattered from the tip using asecond APD.

3. Results and discussion

Thick-shell (7–11 CdS shell monolayers) and ultra-thick-shell (�12 CdS shell monolayers) CdSe/CdScore/shell NQDs were synthesized by the SILARmethod (Experimental section). This method al-lowed relatively facile control over NQD size andsize distribution even for the ultra-thick systems(Figure 1), although the approach was sometimesfound to lead to bimodal size distributions for the

thickest shells (see Discussion below). Quantumyields (QYs) in emission were typically 40–60% forcore/shell systems comprising from 4–16 CdS mono-layers. QYs typically dropped to 20–30% for thethickest-shell g-NQDs (>17 CdS monolayers), possi-bly correlated with the tendency toward a non-idealbimodal distribution of particles and/or the introduc-tion of defects within the shell structure.

Blinking as a function of shell thickness for differ-ent on-time fractions and shell composition. Our ap-proach for assessing the blinking behavior of CdSe/CdS core/shell NQDs is presented in Figure 2a.(Note: we previously confirmed the single-particlenature of our samples [2] and provide additional ver-ification here in Supporting Information.) From suchdata, we determined the percentages of the totalNQD populations that exhibited ‘on-times’ of �20,50, 80 or 99%, respectively (Figure 2b). In this case,%-on-times represent fractions of the total observa-tion time of 54 minutes (18,000 collected frames). Inother words, to be considered as having been ‘on’for �50% of the observation time, a g-NQD wouldhave to have been in a bright or ‘on’ state for atleast 27 minutes (non-continuously). NQDs thatwere on for �99% of the observation time were con-sidered to be “non-blinking”, while those that at-tained an overall on-time of �80% were viewed as“largely non-blinking” for the nearly one hour ofinterrogation by laser excitation. As evident in Fig-ure 2, half or more of a population of the thickest-shell g-NQDs (�12 monolayers of CdS), were eitherlargely or essentially completely non-blinking. Weemphasize this point due to the extraordinarily longobservation time – the robustness of these ultra-thick g-NQDs to blinking withstands significantly ex-tended interrogation times.

In contrast, CdSe/4CdS samples only occasionallycontained even small non-blinking or largely non-blinking fractions. Furthermore, blinking analysesfor these thinner-shell systems were more sensitiveto sample preparation (e.g., differences in NQD sur-face chemistry or charging as influenced by variedsample cleaning or dilution protocol, as well as sub-strate chemistry), sample age, PL thresholding pro-cedures (see Experimental), and experimental setup,where the first two strongly influence photobleach-ing behavior (see below discussion). The addition ofat least 7 monolayers provided clear evidence ofsuppressed blinking, and further additions of up to12 monolayers revealed continued advantages tofurther shell addition. With respect to simple blink-ing behavior alone, addition of shells beyond 12would appear from this data to be superfluous. How-ever, the thickest shells have been shown to provideadditional advantages in terms of more substantiallysuppressed photobleaching [3] and suppressed non-radiative Auger recombination [4], as well as the re-lated, unprecedented multiexcitonic emission [15].

Figure 1 Lower (a) and higher (b) resolution transmis-sion electron microscopy (TEM) images for CdSe/16CdSg-NQDs.




Thus, depending upon the application requirements,‘thick’ vs. ‘ultra-thick’-shell systems offer distinct ad-vantages.

With respect to photobleaching behavior, thethickest shell samples (�16 CdS shell monolayers)

were found not to photobleach even after severalhours of interrogation repeated each of several days(Figure 2c). In contrast, CdSe/4 CdS and CdSe/5 CdSsystems photobleached (complete absence of PL)with a t1/2 � 5–15 min, comparable to commercialNQDs (Qdot655ITKTM) [2]. Thick-shell g-NQDsfrom 10–12 monolayers of CdS photobleached with at1/2 � 1.5–3 h, while even thicker 15-shell systems dis-played some degree of decay, resulting in t1/2 � 5 h.Again, depending upon the application, simple mul-tishell NQDs or thick or ultra-thick-shell systems willbe preferred.

We reported previously that for biological im-aging or other applications for which a ‘less toxic’surface chemistry was desired, the outermost CdSmonolayers could be replaced with non-heavy-metal-containing ZnS layers [2]. These mixed-ultrathick-shell systems, e.g., CdSe/11 CdS/6 CdZnS/2 ZnS be-haved similarly to their CdS-shell counterparts, withblinking behavior apparently mostly determined bythe number of CdS layers. In other words, the CdSe/11 CdS/6 CdZnS/2 ZnS system behaved like a CdSe/11 CdS system [3]. To help better understand thiseffect of shell composition on blinking behavior,here, we directly compare CdS and ZnS in thinner-shell systems with respect to their tendencies towardblinking suppression (Figure 3). We observe thateven the application of as few as four CdS shells re-sults in a significant fraction of NQDs occupying anon-state for �20% of a long observation time (54–60 min) (Figure 3a), or even occasionally comprisinga small non-blinking fraction (Figure 2b). A 4-shellsystem containing 2 CdS and 2 ZnS monolayers alsoshowed similar results (Figure 3b), but further inves-

Figure 2 (online color at: www.biophotonics-journal.org)Single-NQD photoluminescence studies. (a) On-time histo-gram of a CdSe/19 CdS g-NQD population constructedfrom analysis of typically >100 individual g-NQDs. Anexample fluorescence time trace (used to prepare a histo-gram) for an individual CdSe/19 CdS g-NQD is shown inthe inset to (a). Plot of ‘percent NQD population’ versusthe number of CdS shell monolayers for different on-times(b). Two preparations/analyses are plotted for the 10-, 16-,and 19-shell systems, providing an indication of experimen-tal variability in (b). Photobleaching behavior: plots ofemitting NQD fractions over time are presented for CdSe/5 CdS (top left), CdSe/10 CdS (top right), CdSe/15 CdS(bottom left), and CdSe/19 CdS (bottom right) core/shellNQDs (c).

Figure 3 (online color at: www.biophotonics-journal.org)On-time histograms for a series of 4-monolayer shell sys-tems: CdSe/4 CdS (a), CdSe/2 CdS,2 ZnS (b), CdSe/1 CdS,3 ZnS (c) and CdSe/4 ZnS (d). �99%, �80%, �50%, and�20% on-time fractions for each system, respectively, are0%, 0%, 0%, 22% (a), 0%, 0%, 1%, 18% (b), 0%, 0%,0%, 2.8% (c), and 0%, 0%, 0%, 0% (d).


Journal of

BIOPHOTONICS


tigations of other 4-shell-monolayer systems – 1 CdS/3 ZnS and 4 ZnS – revealed that shell thicknessalone was not the determining factor in controllingblinking behavior. Rather, the replacement of CdSmonolayers with ZnS monolayers was found to becritical in this regard, with increasingly reduced on-time fractions resulting in the case of majority or ex-clusively ZnS shell systems (Figure 4c–d). The poorperformance of ZnS shells in providing suppressedblinking in the case of CdSe-core NQDs may resultfrom several factors, including (i) increased systemstrain for the more poorly lattice matched CdSe/ZnSpair (12.7% vs. 3.9% mismatch) and (ii) differencesin core/shell conduction and valence energy-bandoffsets. The former (i) may result in the increasedpresence of trap states and related photo-inducedcharging events, where charging is related to the‘off’ (dark) events in blinking [16]. The latter (ii) im-plies differences in electron-hole separation for thetwo core/shell systems [17]. In the case of CdSe/CdScore/shell NQDs, the hole is largely confined to the

core, while the electronic wavefunction is able toleak into the shell. In contrast, the electron and thehole are more-or-less equally well confined to thecore in the case of CdSe/ZnS. Irrespective of the me-chanism, the utility of CdS shells for blinking sup-pression in the CdSe/CdS system is indicated in theblinking data as early as 4 shells, allowing these thin-ner-shell systems to serve as ‘predictors’ of thick-shell behavior.

Decline in non-blinking population as a functionof observation time. The non-blinking fractions as afunction of shell thickness and observation time areplotted in Figure 4a. Thin, thick, and ultra-thick-shellsamples are largely comparable at very short obser-vation times. The fraction of the NQD populationthat is non-blinking over a period of 0.5 minute isonly approximately doubled for the thickest shellsamples compared to their thinner-shell counter-parts. However, after 54 minutes of observation time,the thickest-shell samples (�12 monolayers of CdS)comprise �10–15 times more non-blinking NQDs(versus in the ‘best-case’ scenario for the CdSe/4 CdS system, i.e., wherein the fraction of non-blink-ing NQDs is non-zero for the observation time of54 minutes – compare Figure 2b with Figure 3a). Forlimited observation times (e.g., �5 minutes), thinnershell systems may be acceptable [11], dependingupon the application, as the percentage of largelynon-blinking NQDs (‘on’ �80% of the observationtime) is �50% (Figure 4b). In the case of the thick-est shell systems, the decrease in the fractions ofnon-blinking and largely non-blinking NQDs as afunction of time is understood in terms of the ‘intrin-sic’ non-blinking population for the specific times,i.e., the actual fraction of NQDs that can sustain an‘on’ state for the observation period. In contrast, inthe case of the thinner-shell systems, photobleachingcontributes significantly to the poorer statistics atlonger observation times. In other words, as a resultof photobleaching (non-reversible transition to an‘off’ state), a substantial fraction of the NQDs thatcontributed to an initial, relatively large on-timepercentage is eliminated from the population ofNQDs that are emitting (either blinking or non-blinking). In this way, the thinner-shell NQDs showthe largest changes in on-time fractions over time(Figure 4).

Ability to tolerate ligand exchange procedures as afunction of shell thickness. The photoluminescenceefficiency (i.e., the QY in emission) of thinner shell,multi-shell NQDs is known to be susceptible tolosses when the initial, protective ligand monolayeris compromised. As-prepared, colloidal NQDs com-prise the semiconductor nanocrystal and a layer oforganic ligands that datively couple to the nanocrys-tal and render the NQD soluble in nonpolar sol-vents. These ligands help to passivate so-called sur-face ‘dangling bonds’ that otherwise serve as ‘trap

Figure 4 (online color at: www.biophotonics-journal.org)Comparison of percent-NQD population versus shellthickness as a function of the total observation time (0.5,5, and 54 minutes) for NQDs ‘on’ �99% of the observa-tion time (‘non-blinking’ population) (a) and �80% of theobservation time (‘largely non-blinking’ population) (b).




states’ that provide non-radiative or inefficient, non-band-edge electron-hole recombination pathways. Itis commonly observed that partial or complete re-moval and/or replacement of the as-prepared ligandsreduces the effectiveness of the ligand layer to servethis purpose, with lower QYs the result [18]. Ligandexchange is a common method for rendering nonpo-lar NQDs water soluble, whereby the alkyl-termi-nated ligands are replaced with polar-group and/orcharge-terminated ligands. As non-blinking behaviorin NQD fluorophores is of particular interest forbioimaging and biomolecular tracking applications,we compared photoluminescence and blinking behav-ior for as-prepared and ligand-exchanged core/shellNQDs as a function of shell thickness.

The ensemble photoluminescence showed a cleardifference between CdSe/4 CdS and CdSe/12 CdSwhen the nonaqueous and the aqueous solutions ofeach were compared (Figure 5a). This dramatic re-sult is consistent with our previous observation thatthe thinner shell systems are more susceptible to PLdecline in response to simple NQD ‘washing’ proce-dures (i.e., removal of excess ligand) compared tothe thicker shell systems [2]. Interestingly, blinkingbehavior was not dramatically affected, even in thecase of the thinner shell systems (Figure 5b). Thisapparently unexpected result may have several ex-planations. First, the 4-shell system that showednearly compete PL suppression, but on-time frac-tions comparable to many of our nonpolar 4-shellsystem analyses, likely comprised a high fraction ofnon-emitting or rapidly photobleached NQDs. TheseNQDs would contribute to low ensemble QYs andemission but would not be ‘on’ long enough (at least180 ms, the temporal resolution of our blinking ex-periment) to be counted in the blinking statistics.

The relatively fewer, ‘active’ 4-shell NQDs couldpossess similar stability to blinking and photobleach-ing to their nonpolar counterparts and would domi-nate the single-NQD measurements. Second, it isknown that the effects of ligand exchange withsmall-molecule reducing agents, as used here (i.e.,mercaptosuccinic acid), are strongly ligand concen-tration, time, and process dependent [18], with for-tuitous exchanges even possibly resulting in blinkingsuppression [19].

Correlated AFM-PL. Recent work has demon-strated the benefit of “ultra-thick” shells (�12 mono-layers) for realizing significantly suppressed non-radiative recombination processes that otherwiseprevent conversion of charged and multiexcitonicstates into usable photons by way of efficient nonra-diative Auger recombination [4, 15]. Unfortunately,to date, we have observed that complications duringshell growth (including incomplete particle solubilityas more and more shells are added, especially be-yond 12 CdS monolayers) can result in product inho-mogeneity. When present, the various sub-popula-tions likely comprise the desired core/shell materials,as well as shell-only particles and core/shell materialswith untargeted shell thicknesses and/or defectiveshells. We used correlated AFM-PL measurementsto determine the relationship between g-NQD sizeand the ability to emit light under laser excitation.We further used the approach to assess our ability toseparate “dark” NQDs from emitting NQDs by wayof size-selective precipitation.

Based on transmission electron microscopy(TEM) measurements (Supporting InformationFigure S2), we suspected a CdSe/19 CdS sample ofcomprising particles that were highly dispersed insize. Correlated AFM/PL studies allowed us to con-firm the bimodal size distribution of this sample, aswell as to assess the relationship between NQD sizeand the ability to photoluminesce. In Figure 6 weshow the correlated AFM-PL image for the suspectCdSe/19 CdS sample. Similar to TEM data, g-NQDheights were found to be bimodal, with �60% of theNQDs being large (�20 nm) and �40% being smal-ler (�10 nm). All of the smaller g-NQDs were ob-served to be non-emitting (36 NQDs: 8.3 3.0 nm),while many of the large g-NQDs were also dark(42 NQDs: 23.6 3.2 nm). The 10% of the batchthat was found to be emitting was all ‘large’, but,interestingly, overall smaller in size (9 NQDs: 19.9 3.7 nm) compared to the ‘non-emitting large’NQDs.

Using size-selective precipitation we were able toseparate both populations of dark g-NQDs from thebright g-NQD fraction, as evident in subsequentcorrelated AFM-PL images (Figure 6b). After size-selection, >75% of the sample was found to be emit-ting (23 of 30 NQDs) and of similar size to the emit-ting fraction in the ‘crude’ sample (20.4 1.7 nm),

Figure 5 (online color at: www.biophotonics-journal.org)CdSe/4 CdS (left) and CdSe/16 CdS (right) NQDs photolu-minescing under UV lamp excitation in hexane solvent (a)and in water (b). Plot of ‘percent NQD population’ versusthe number of CdS shell monolayers for different on-timesfor mercaptosuccinic acid capped NQDs in water (c).


Journal of

BIOPHOTONICS


with most of the remaining dark g-NQDs beingsmall (5 NQDs: 7.6 4.4 nm). The observed �20 nmsize for the emitting fraction in this CdSe/19 CdSsystem is consistent with that expected for an‘ideal’ CdSe/19 CdS NQD [3.8 nm diameter core þ(19 shell monolayers � 0.35 nm � 2) þ (2 ligandlayers � �1 nm) ¼ 19.1 nm]. Thus, the bright-NQDfraction does appear to be closest in size to the de-sired CdSe/19 CdS system. (We note, however, thatwithout deliberate calibration of the AFM, a trulyquantitative comparison is currently not possible.)The undesired distribution in sizes may result fromseveral factors including nucleation of shell materialto form CdS NQDs, as well as unequal addition ofshell material onto existing core/shell particles.Either process could result in the observed small-NQD fraction. The latter process is perhaps ex-pected in the case of a partially precipitated system,where the phase-heterogeneity compromises shellgrowth of the relatively less-well solvated particles.The dark, but very-large fraction may constitute par-ticles that incorporated shells in a relatively less wellcontrolled manner, resulting in defective or distortedshells (Supporting Information Figure S3). Signifi-

cantly, the bimodal distribution is generally not ob-served for systems possessing <16 CdS monolayers(Figure 1). Modifications to the synthesis to improvethe batch uniformity for the thickest shell systemsare being investigated and will be reported else-where.

Novel blinking statistics as a function of shellthickness. As described in the Experimental section,our recent studies using a more sensitive time-cor-related single-photon counting (TCSPC) systemhave indicated that more than 80% of our ultra-thickg-NQDs (shell thicknesses >12 CdS monolayers)never turn into a completely off-state, although exhi-biting wide fluctuations in PL emission intensity(Supporting Information Figure S1). (It should benoted that our TCSPC system is capable of detectingoff-intervals down to very fast 1 ms timescales, with10–100 ms resolution shown in Figure S1 and 1 msresolution demonstrated previously [3]). Thus, weconsider the “off-levels” of the CCD study describedhere to correspond, instead, to a weaker emittingstate and the analyses to be highly conservative inour estimates of “non-blinking” g-NQD fractions.Nevertheless, we use this more traditional approachof on/off time-distribution analysis to study PL inten-sity fluctuation behavior of our g-NQDs, as it hasbeen applied previously by others to study blinkingin conventional NQD systems.

The probability density of on/off time distribu-tions for conventional NQDs,

PðtiÞ ¼ 2 Ni=½ðtiþ1 � tiÞ þ ðti � ti�1Þ ð1Þ

where Ni represents the number of events with on/off time ti, and the denominator is the average timeduration between the closest preceding and follow-ing events [20], exhibits the well known power-lawbehavior,

PðtÞ / 1=ðton=offÞm ð2Þ

where mon/off is �1.5 (Figure 7a) [21, 22]. We ob-served that CdSe/CdS NQDs with shell thicknesses<7 monolayers showed power-law behavior withmon/off �1.5, as well. Furthermore, initially, the ultra-thick-shell systems also appeared to follow thistrend. This was the case when the entire ensemblewas analyzed, although the mon/off was found to beslightly larger compared to conventional and thin-ner-shell systems – �1.7 (Figure 7b). Significantly,however, the ultra-thick-shell systems revealed a un-ique trend in the P(ton). While conventional NQDsshow a cut-off (departure) from the power-law de-cay behavior (Figure 7a), our g-NQD on-times donot cut-off and even exhibit a spreading of datapoints at long on-times (Figure 7b). This spreadingof data points resulted from the sub-ensemble of g-NQDs characterized by strongly suppressed blink-ing.

Figure 6 (online color at: www.biophotonics-journal.org)Correlated AFM/photoluminescence images for ‘crude’(a) and size-selected (c) CdSe/19CdS core/shell g-NQDs.NQD heights were determined by difference between theNQD and background heights, with examples shown forthe crude (b) and size-selected (d) samples. Two examplesof the remaining ‘dark’, small-NQD fraction in the size-se-lected sample are outlined in (c) (blue squares surroundinga 3 and an 11 nm NQD).




The blinking behavior of the sub-ensemble ofg-NQDs with suppressed blinking was further ana-lyzed, separate from that contributed by the blinkingfraction. We achieved this by extracting and analyz-ing the P(ton/off) only for the sub-ensemble of theg-NQD population that had total on-time fractionsgreater than 75%. As an example, Figure 7c showsP(ton/off)‘suppressed-blinkingfraction’ of the 16-shell system(also: Supp Information Figure S4). The data show adramatic deviation from the usual power-law be-havior. The P(ton) can no longer be described by asimple power-law decay. Instead, it decays with twodistinct exponents characterized by different time-scales (mon ¼ 0.8 and mon ¼ 1.8 for time <2.0 s andtime >2.0 s, respectively). Furthermore, althoughP(toff) still exhibits a power-law decay, the decay inP(toff) becomes very rapid, with exponent m varyingfrom 2.2 to 3.2 with increasing shell thickness (Fig-ure 7d). Interestingly, the variation of m as a func-tion of shell thickness mirrors the variation of theg-NQD population that exhibits on-time fractions of>99% (Figure 2b). This important result indicatesthat ‘m’ can be utilized as a criterion for understand-ing non-blinking behavior.

Furthermore, the dramatic increase in the value ofm suggests that the mechanism responsible for theremaining blinking in g-NQD systems could be dif-ferent from that of conventional NQDs. In conven-tional NQDs, very slow decay of P(toff) with m < 2leads to the divergence of the statistical average,

htoffðtexpÞi ¼Ðtexp

t¼ti

tPðtÞ dt=Ðtexp

t¼ti

PðtÞ dt ð3Þ

with total experiment time texp. This observation ledto the conclusion that the ionization/re-neutraliza-tion processes responsible for blinking [16, 21] occurover (i) a distribution of tunneling distances betweenthe NQD core and interface states or (ii) an expo-nential distribution in the depths of the trap states.In contrast, studies of blinking in InP self-assembledquantum dots (QDs) have revealed that, althoughthe trapping of charges in the vicinity of the QD isresponsible for the blinking process (similar toNQDs), the average toff and the switching rate canbe described by a thermal activation model involvinga single activation barrier [23], i.e., unlike the casefor colloidal NQDs. The large m values (>3) that weobserve for g-NQDs cause the convergence of htoffito a finite value for this unusual type of colloidalNQD (Supporting Information Figure S4). Thus, theblinking behavior of our g-NQDs appears to be moreakin to the behavior observed for InP self-assembledQDs, as these, too, exhibit a finite htoffi. Also, in con-trast with the mechanism proposed for conventio-nal colloidal NQDs, the finite htoffi observed for theg-NQDs suggests a thermal activation model invol-ving a single activation barrier. Thus, not only isblinking suppressed in the case of g-NQDs, but theblinking mechanism of any remaining blinking behav-ior appears distinctly altered compared to other col-loidal NQDs and similar to self-assembled QDs.

Utility of ultra-thick-shell CdSe/CdS NQDs for3-D particle tracking. The ability to follow the tra-jectory of single molecules (e.g., proteins) as theyperform their biological functions in vivo is an im-portant goal for advanced bioimaging. To date, single-particle tracking has led to an improved understand-ing of bacteria motility [24], membrane dynamics[25–27], and motor proteins [28], revealing detailsthat are otherwise hidden in ensemble measurements.Members of our team have previously reported on anovel approach to three-dimensional (3-D) trackingof small, dim particles, such as a single proteintagged with a single fluorophore [29]. The geometryof this 3-D single-particle tracking instrument is avariation on that of a confocal microscope and in-volves tracking freely diffusing nanoparticles in threedimensions using four-point confocal detection andpositional feedback. Its principle has been describedin detail elsewhere [29, 30]. Significantly, the approachuses a low illumination power that makes it ulti-

Figure 7 (online color at: www.biophotonics-journal.org)P(ton/off) of CdSe/4 CdS NQDs showing typical powerlaw behavior with m ¼ 1.5. P(ton) exhibits cut-off result-ing from the power-law behavior at on-times longer than1 s (a). P(ton/off) of the whole ensemble of CdSe/16 CdSg-NQDs (b). P(ton/off) of the sub-ensemble of CdSe/16 CdSg-NQDs that exhibits a total on-time fraction of 75% (c).[a–c: open squares ¼ P(ton), solid circles ¼ P(toff)]. Plot ofmoff versus shell thickness (d), where open squares ¼ thewhole ensemble of CdSe/4CdS and CdSe/5CdS NQDs andsolid circles ¼ the sub-ensemble of g-NQDs with total on-time fractions >75%.


Journal of

BIOPHOTONICS


mately compatible with live-cell investigation [29].Here, we compare non-blinking g-NQDs (CdSe/19 CdS) with conventional NQDs for their ability tobe tracked over time and space in a biologically rele-vant medium (Figure 8). We find that the reducedblinking of the g-NQD fluorophores compared tothat of conventional NQDs is a significant advantagefor 3-D tracking. On average, we are able to trackg-NQDs over much longer intervals than conven-tional NQDs. In the case of conventional NQDs,blinking limits our ability to track. In contrast, withg-NQDs, our ability to track is more typically limitedby the range-of-travel of our sample stage (30 � 30� 10 microns), i.e., the experimental limitation ofour setup rather than any fundamental limit impliedby the fluorophore.

4. Conclusion

Shell thickness in the CdSe/CdS core/shell NQDsystem is clearly a key criterion to be used in con-

trolling blinking behavior, improving NQD photo-stability, altering the blinking mechanism, and pro-viding significantly enhanced utility for single-particle tracking applications, for example, in cellbiology. Significantly, performance advantages wereobserved for ‘ultra-thick-shell’ variations, especiallyin the case of long interrogation times and non-idealchemical environments. Shell composition was alsofound to be an important consideration in realizingthe unusual blinking behaviors, where physical andor electronic-structure relationships between thecore and the shell materials are likely controllingfactors. Studies are underway, with some initial re-sults already reported [4], to better understand theunderlying mechanism(s) for these remarkable prop-erties and to extend the approach to other core/shellsystems, and, ideally, to more ‘bio-compatible’ NQDcompositions. To this end, we are investigating giant-shell systems based on InP and ZnSe.

Acknowledgements This work was conducted in part inthe Center for Integrated Nanotechnologies (CINT)jointly operated by Los Alamos and Sandia National

Figure 8 (online color at: www.biophotonics-journal.org) Fluorescence intensity trajectory from a conventional CdSe/ ZnScore/shell NQD system [Invitrogen 605 ITK amino (PEG)] diffusing in 80/20 (v/v) glycerol-water (a)-bottom. Data are binnedin 1 ms intervals. Three-dimensional trajectory of this nanoparticle recorded in 5 ms intervals is shown in (a)-top left. Thedimensions of the cube containing the particle are 3 � 3 � 3 mm3. The rainbow color scheme is correlated with that used forthe fluorescence intensity trajectory to show the time dependence. The mean-squared displacement (MSD) of the nanoparticleas a function of measurement time interval (Dt) is shown in (a)-top right. In the case of pure diffusive motion in three dimen-sions (as is the case here), MSD and Dt are linearly related with a slope of six times the diffusion coefficient (D). Fluorescenceintensity trajectory for CdSe/19 CdS g-NQDs diffusing in 60/40 (v/v) glycerol-water is shown in (b)-bottom. A g-NQDtrajectory (cube dimensions: 11 � 11 � 11 mm3) and g-NQD MSD are shown in (b)-top left and top-right, respectively.




Laboratories (LANL and SNL) for the U.S. Departmentof Energy (DOE). H.H. and J.A.H. acknowledge partialsupport by NIH-NIGMS Grant 1R01GM084702–01. J.V.acknowledges support by Los Alamos National Labora-tory Directed Research and Development Funds.

References

[1] M. Han, X. Gao, J. Z. Su, and S. Nie, Nat. Biotechnol.19, 631 (2001).

[2] Y. Chen, J. Vela, H. Htoon, J. L. Casson, D. J. Werder,D. A. Bussian, V. I. Klimov, and J. A. Hollingsworth,J. Am. Chem. Soc. 130, 5026 (2008).

[3] J. A. Hollingsworth, J. Vela, Y. Chen, H. Htoon, V. I.Klimov, and A. R. Casson, Proc. SPIE Int. Soc. Opt.Eng. 7189, 718904 (2009).

[4] F. Garcıa-Santamarıa, Y. Chen, J. Vela, R. D. Schaller,J. A. Hollingsworth, and V. I. Klimov, Nano Lett. 9,3482 (2009).

[5] F. Pinaud, S. Clarke, A. Sittner, and M. Dahan, Nat.Methods 7, 275 (2010).

[6] T. Pons and H. Mattoussi, Ann. of Biomed. Eng. 37,1934 (2009).

[7] S. Park, B. L. Clark, D. A. Keszler, J. P. Bender, J. F.Wager, T. A. Reynolds, and G. S. Herman, Science297, 65 (2002).

[8] J. J. Li, Y. A. Wang, W. Guo, J. C. Keay, T. D. Mishi-ma, M. B. Johnson, and X. Peng, J. Am. Chem. Soc.125, 12567 (2003).

[9] R. Xie, U. Kolb, J. Li, T. Basche, and A. Mews, J. Am.Chem. Soc. 127, 7480 (2005).

[10] J. van Embden, J. Jasieniak, and P. Mulvaney, J. Am.Chem. Soc. 131, 14299 (2009).

[11] B. Mahler, P. Spinicelli, S. Buil, X. Quelin, J.-P. Her-mier, and B. Dubertret, Nature Mater. 7, 659 (2008).

[12] B. Mahler, N. Lequeux, and B. Dubertret, J. Am.Chem. Soc. 132, 953 (2010).

[13] W. W. Yu, L. Qu, W. Guo, and X. Peng, Chem. Mater.15, 2854 (2003); 16, 560 (2004).

[14] S. Fery-Forgues and D. Lavabre, Chem. Mater. 76,1260 (1999).

[15] H. Htoon, A. Malko, D. Bussian, J. Vela, Y. Chen, J. A.Hollingsworth, and V. I. Klimov, Nano Lett. ArticleASAP. DOI: 10.1021/nl1004652.

[16] T. D. Krauss and L. E. Brus, Phys. Rev. Lett. 83, 4840(1999).

[17] G. E. Cragg and A. L. Efros, Nano Lett. 10, 313 (2010).[18] S. Jeong, M. Achermann, J. Nanda, S. Ivanov, V. I.

Klimov, and J. A. Hollingsworth, J. Am. Chem. Soc.127, 10126 (2005).

[19] S. Hohng and T. Ha, J. Am. Chem. Soc. 126, 1324(2004).

[20] J. J. Peterson and D. J. Nesbitt, Nano Lett. 9, 338(2009).

Javier Vela received a B.Sc.from the National Univer-sity of Mexico in 2001 and aPh.D. in chemistry from theUniversity of Rochester in2005. He was the recipientof the Hooker and Messer-smith graduate fellowshipsin 2003–2004, and the ACSInorganic Young Investiga-

tor Award in 2006. After postdoctoral work at the Uni-versity of Chicago and at Los Alamos National Labora-tory, he joined the faculty at Iowa State University in2009. His research lies at the interface between nano-technology and molecular chemistry and their applica-tions in bio-imaging, renewable energy, and catalysis.

Han Htoon is a technicalstaff member and a formerDirector’s Postdoctoral Fel-low at Los Alamos NationalLaboratory. He received aB.Sc. (Honors) and an M.S.in physics from Universityof Yangon, Yangon, Myan-mar and Western IllinoisUniversity, Macomb, IL, re-

spectively. He obtained his Ph.D. in physics from Uni-versity of Texas at Austin in 2001. He currently directsresearch efforts in advanced optical characterization atthe scale of individual nanostructures, making pioneer-ing contributions in spectroscopic technique develop-ment and in fundamental photophysics of semiconduc-tor nanocrysals.

Jennifer Hollingsworth is atechnical staff member anda former Director’s Postdoc-toral Fellow at Los AlamosNational Laboratory. Shereceived a B.A. (Phi BetaKappa) in chemistry fromGrinnell College, Grinnell,Iowa. She obtained herPh.D. in materials inorganic

chemistry from Washington University in St. Louis in1999 as a NASA Research Fellow. She currently directsresearch efforts in the advancement of novel syntheticmethods for functional semiconductor nanocrystal het-erostructures, focusing on their applications as biomole-cular probes and in energy-conversion materials.


Journal of

BIOPHOTONICS


[21] M. Kuno, D. P. Fromm, H. F. Hamann, A. Gallagher,and D. J. Nesbitt, J. Chem. Phys. 112, 3117 (2000).

[22] F. D. Stefani, J. P. Hoogeboom, and E. Barkai, Phys.Today 62, 34 (2009).

[23] M. Sugisaki, R. Hong-Wen, K. Nishi, and Y. Masu-moto, Phys. Rev. Lett. 86, 4883 (2001).

[24] H. C. Berg, Rev. Sci. Instrum. 42, 868 (1971).[25] L. S. Barak and W. W. Webb, J. Cell Biol. 90, 595

(1981).[26] A. Kusumi, Y. Sako, and M. Yamamoto, Biophys. J.

65, 2021 (1993).

[27] T. Schmidt, G. J. Schutz, W. Baumgartner, H. J. Gru-ber, and H. Schindler, Proc. Natl. Acad. Sci. USA 93,2926 (1996).

[28] A. Yildiz, J. N. Forkey, S. A. McKinney, T. Ha, Y. E.Goldman, and P. R. Selvin, Science 300, 2061 (2003).

[29] G. A. Lessard, P. M. Goodwin, and J. H. WernerAppl. Phys. Lett. 91, 224106 (2007).

[30] G. A. Lessard, P. M. Goodwin, and J. H. Werner, inUltrasensitive and Single-Molecule Detection Tech-nologies, edited by J. Enderlein and Z. K. Gryczynski,(SPIE, 2006), Vol. 6092, p. 609205.

>>> communicates important findings witha high degree of novelty and need forexpress publication, as well as otherresults of immediate interest to thesolid state physics and materials sciencecommunity.

>>> offers extremely fast publicationtimes: typically less than 14 days fromsubmission to online publication. Thisis definitely a world record for Letterjournals in solid state physics! Peerreview by two independent refereesguarantees strict quality standards.

physica status solidi RRL –Rapid Research Letters,

the fastest peer-reviewedpublication medium in

solid state physics!

2010. Volume 4, 12 issues.Print ISSN: 1862-6254

Online ISSN: 1862-6270

The latest memberof the physica status solidi journals

For more information please contact Wiley Customer Service:

601101008_bu

With an Impact Factor of * pss RRL isthe most highly cited journal publishing exclusively Letter

articles in condensed matter and applied physics.

www.pss-rapid.com

2.560

[email protected](Germany, Austria, Switzerland)

[email protected](Japan)

[email protected](Americas, Europe, Middle East and Africa, Asia Pacific)

*2009 Journal Citation Reports® (Thomson Reuters, 2010)

wileyonlinelibrary.com




Documents

Effect of shell thickness and composition on blinking suppression and the blinking mechanism in ‘giant’ CdSe/CdS nanocrystal quantum dots