`
`9463
`
`(CdSe)ZnS Core-Shell Quantum Dots: Synthesis and Characterization of a Size Series of
`Highly Luminescent Nanocrystallites
`
`B. O. Dabbousi,† J. Rodriguez-Viejo,‡ F. V. Mikulec,† J. R. Heine,§ H. Mattoussi,§ R. Ober,⊥
`K. F. Jensen,‡,§ and M. G. Bawendi*,†
`Departments of Chemistry, Chemical Engineering, and Materials Science and Engineering,
`Massachusetts Institute of Technology, 77 Massachusetts AVe., Cambridge, Massachusetts 02139, and
`Laboratoire de Physique de la Matie`re Condense´e, Colle`ge de France, 11 Place Marcellin Berthelot,
`75231 Paris Cedex 05, France
`ReceiVed: March 27, 1997; In Final Form: June 26, 1997X
`
`We report a synthesis of highly luminescent (CdSe)ZnS composite quantum dots with CdSe cores ranging in
`diameter from 23 to 55 Å. The narrow photoluminescence (fwhm e 40 nm) from these composite dots
`spans most of the visible spectrum from blue through red with quantum yields of 30-50% at room temperature.
`We characterize these materials using a range of optical and structural techniques. Optical absorption and
`photoluminescence spectroscopies probe the effect of ZnS passivation on the electronic structure of the dots.
`We use a combination of wavelength dispersive X-ray spectroscopy, X-ray photoelectron spectroscopy, small
`and wide angle X-ray scattering, and transmission electron microscopy to analyze the composite dots and
`determine their chemical composition, average size, size distribution, shape, and internal structure. Using a
`simple effective mass theory, we model the energy shift for the first excited state for (CdSe)ZnS and (CdSe)-
`CdS dots with varying shell thickness. Finally, we characterize the growth of ZnS on CdSe cores as locally
`epitaxial and determine how the structure of the ZnS shell influences the photoluminescence properties.
`
`I. Introduction
`
`Semiconductor nanocrystallites (quantum dots) whose radii
`are smaller than the bulk exciton Bohr radius constitute a class
`of materials intermediate between molecular and bulk forms of
`matter.1 Quantum confinement of both the electron and hole
`in all three dimensions leads to an increase in the effective band
`gap of the material with decreasing crystallite size. Conse-
`quently, both the optical absorption and emission of quantum
`dots shift to the blue (higher energies) as the size of the dots
`gets smaller. Although nanocrystallites have not yet completed
`their evolution into bulk solids, structural studies indicate that
`they retain the bulk crystal structure and lattice parameter.2
`Recent advances in the synthesis of highly monodisperse
`nanocrystallites3-5 have paved the way for numerous spectro-
`scopic studies6-11 assigning the quantum dot electronic states
`and mapping out their evolution as a function of size.
`Core-shell
`type composite quantum dots exhibit novel
`properties making them attractive from both an experimental
`and a practical point of view.12-19 Overcoating nanocrystallites
`with higher band gap inorganic materials has been shown to
`improve the photoluminescence quantum yields by passivating
`surface nonradiative recombination sites. Particles passivated
`with inorganic shell structures are more robust than organically
`passivated dots and have greater tolerance to processing
`conditions necessary for incorporation into solid state structures.
`Some examples of core-shell quantum dot structures reported
`earlier include CdS on CdSe and CdSe on CdS,12 ZnS grown
`on CdS,13 ZnS on CdSe and the inverse structure,14 CdS/HgS/
`CdS quantum dot quantum wells,15 ZnSe overcoated CdSe,16
`and SiO2 on Si.17,18 Recently, Hines and Guyot-Sionnest
`
`reported making (CdSe)ZnS nanocrystallites whose room tem-
`perature fluorescence quantum yield was 50%.19
`This paper describes the synthesis and characterization of a
`series of room-temperature high quantum yield (30%-50%)
`core-shell (CdSe)ZnS nanocrystallites with narrow band edge
`luminescence spanning most of the visible spectrum from 470
`to 625 nm. These particles are produced using a two-step
`synthesis that is a modification of the methods of Danek et al.16
`and Hines et al.19 ZnS overcoated dots are characterized
`spectroscopically and structurally using a variety of techniques.
`The optical absorption and photoluminescence spectra of the
`composite dots are measured, and the lowest energy optical
`transition is modeled using a simplified theoretical approach.
`Wavelength dispersive X-ray spectroscopy and X-ray photo-
`electron spectroscopy are used to determine the elemental and
`spatial composition of ZnS overcoated dots. Small-angle X-ray
`scattering in solution and in polymer films and high-resolution
`transmission electron microscopy measurements help to deter-
`mine the size, shape, and size distribution of the composite dots.
`Finally, the internal structure of the composite quantum dots
`and the lattice parameters of the core and shell are determined
`using wide-angle X-ray scattering.
`In addition to having higher efficiencies, ZnS overcoated
`particles are more robust than organically passivated dots and
`potentially more useful for optoelectronic device structures.
`Electroluminescent devices (LED’s) incorporating (CdSe)ZnS
`dots into heterostructure organic/semiconductor nanocrystallite
`light-emitting devices may show greater stability.20 Thin films
`incorporating (CdSe)ZnS dots into a matrix of ZnS using
`electrospray organometallic chemical vapor deposition (ES-
`OMCVD) demonstrate more than 2 orders of magnitude
`improvement in the PL quantum yields (∼10%) relative to
`identical structures based on bare CdSe dots.21 In addition, these
`structures exhibit cathodoluminescence21 upon excitation with
`high-energy electrons and may potentially be useful in the
`Nanoco Technologies, Ltd
`S1089-5647(97)01091-2 CCC: $14.00 © 1997 American Chemical Society
`EXHIBIT 1016
`1 of 13
`
`* To whom correspondence should be addressed.
`† Department of Chemistry, MIT.
`‡ Department of Chemical Engineering, MIT.
`§ Department of Materials Science and Engineering, MIT.
`⊥ Colle`ge de France.
`X Abstract published in AdVance ACS Abstracts, September 1, 1997.
`
`
`
`9464 J. Phys. Chem. B, Vol. 101, No. 46, 1997
`
`production of alternating current thin film electroluminescent
`devices (ACTFELD).
`
`II. Experimental Section
`Materials. Trioctylphosphine oxide (TOPO, 90% pure) and
`trioctylphosphine (TOP, 95% pure) were obtained from Strem
`and Fluka, respectively. Dimethylcadmium (CdMe2) and di-
`ethylzinc (ZnEt2) were purchased from Alfa and Fluka, respec-
`tively, and both materials were filtered separately through a 0.2
`(cid:237)m filter in an inert atmosphere box. Trioctylphosphine selenide
`was prepared by dissolving 0.1 mol of Se shot in 100 mL of
`TOP, thus producing a 1 M solution of TOPSe. Hexamethyl-
`disilathiane ((TMS)2S) was used as purchased from Aldrich.
`HPLC grade n-hexane, methanol, pyridine, and 1-butanol were
`purchased from EM Sciences.
`Synthesis of Composite Quantum Dots. (CdSe)ZnS. Nearly
`monodisperse CdSe quantum dots ranging from 23 to 55 Å in
`diameter were synthesized via the pyrolysis of the organome-
`tallic precursors, dimethylcadmium and trioctylphosphine se-
`lenide,
`in a coordinating solvent,
`trioctylphosphine oxide
`(TOPO), as described previously.3 The precursors were injected
`at temperatures ranging from 340 to 360 (cid:176)C, and the initially
`formed small (d ) 23 Å) dots were grown at temperatures
`between 290 and 300 (cid:176)C. The dots were collected as powders
`using size-selective precipitation3 with methanol and then
`redispersed in hexane.
`A flask containing 5 g ofTOPO was heated to 190 (cid:176)C under
`vacuum for several hours and then cooled to 60 (cid:176)C after which
`0.5 mL of trioctylphosphine (TOP) was added. Roughly 0.1-
`0.4 (cid:237)mol of CdSe dots dispersed in hexane was transferred into
`the reaction vessel via syringe, and the solvent was pumped
`off.
`Diethylzinc (ZnEt2) and hexamethyldisilathiane ((TMS)2S)
`were used as the Zn and S precursors. The amounts of Zn and
`S precursors needed to grow a ZnS shell of desired thickness
`for each CdSe sample were determined as follows: First, the
`average radius of the CdSe dots was estimated from TEM or
`SAXS measurements. Next, the ratio of ZnS to CdSe necessary
`to form a shell of desired thickness was calculated based on
`the ratio of the shell volume to that of the core assuming a
`spherical core and shell and taking into account the bulk lattice
`parameters of CdSe and ZnS. For larger particles the ratio of
`Zn to Cd necessary to achieve the same thickness shell is less
`than for the smaller dots. The actual amount of ZnS that grows
`onto the CdSe cores was generally less than the amount added
`due to incomplete reaction of the precursors and to loss of some
`material on the walls of the flask during the addition.
`Equimolar amounts of the precursors were dissolved in 2-4
`mL of TOP inside an inert atmosphere glovebox. The precursor
`solution was loaded into a syringe and transferred to an addition
`funnel attached to the reaction flask. The reaction flask
`containing CdSe dots dispersed in TOPO and TOP was heated
`under an atmosphere of N2. The temperature at which the
`precursors were added ranged from 140 (cid:176)C for 23 Å diameter
`dots to 220 (cid:176)C for 55 Å diameter dots. 22 When the desired
`temperature was reached, the Zn and S precursors were added
`dropwise to the vigorously stirring reaction mixture over a period
`of 5-10 min.
`After the addition was complete, the mixture was cooled to
`90 (cid:176)C and left stirring for several hours. A 5 mL aliquot of
`butanol was added to the mixture to prevent the TOPO from
`solidifying upon cooling to room temperature. The overcoated
`particles were stored in their growth solution to ensure that the
`surface of the dots remained passivated with TOPO. They were
`later recovered in powder form by precipitating with methanol
`and redispersed into a variety of solvents including hexane,
`chloroform, toluene, THF, and pyridine.
`
`Dabbousi et al.
`
`(CdSe)CdS. Cadmium selenide nanocrystallites with diam-
`eters between 33.5 and 35 Å were overcoated with CdS to
`varying thickness using the same basic procedure as that outlined
`for the ZnS overcoating. The CdS precursors used were Me2-
`Cd and (TMS)2S. The precursor solution was dripped into the
`reaction vessel containing the dots at a temperature of 180 (cid:176)C
`and a rate of (cid:24)1 mL/min. The solution became noticeably
`darker as the overcoat precursors were added. Absorption
`spectra taken just after addition of precursors showed a
`significant shift in the absorption peak to the red. To store these
`samples, it was necessary to add equal amounts of hexane and
`butanol since the butanol by itself appeared to flocculate the
`particles.
`Optical Characterization. UV-vis absorption spectra were
`acquired on an HP 8452 diode array spectrophotometer. Dilute
`solutions of dots in hexane were placed in 1 cm quartz cuvettes,
`and their absorption and corresponding fluorescence were
`measured. The photoluminescence spectra were taken on a
`SPEX Fluorolog-2 spectrometer in front face collection mode.
`The room-temperature quantum yields were determined by
`comparing the integrated emission of the dots in solution to
`the emission of a solution of rhodamine 590 or rhodamine 640
`of identical optical density at the excitation wavelength.
`Wavelength Dispersive X-ray Spectroscopy. A JEOL SEM
`733 electron microprobe operated at 15 kV was used to
`determine the chemical composition of the composite quantum
`dots using wavelength dispersive X-ray (WDS) spectroscopy.
`One micrometer thick films of (CdSe)ZnS quantum dots were
`cast from concentrated pyridine solutions onto Si(100) wafers,
`and after the solvent had completely evaporated the films were
`coated with a thin layer of amorphous carbon to prevent
`charging.
`X-ray Photoelectron Spectroscopy. XPS was performed
`using a Physical Electronics 5200C spectrometer equipped with
`a dual X-ray anode (Mg and Al) and a concentric hemispherical
`analyzer (CHA). Data were obtained with Mg KR radiation
`(1253.6 eV) at 300 W (15 keV, 20 mA). Survey scans were
`collected over the range 0-1100 eV with a 179 eV pass energy
`detection, corresponding to a resolution of 2 eV. Close-up scans
`were collected on the peaks of interest for the different elements
`with a 71.5 eV pass energy detection and a resolution of 1 eV.
`A base pressure of 10-8 Torr was maintained during the
`experiments. All samples were exchanged with pyridine and
`spin-cast onto Si substrates, forming a thin film several
`monolayers thick.
`Transmission Electron Microscopy. A Topcon EM002B
`transmission electron microscope (TEM) was operated at 200
`kV to obtain high-resolution images of individual quantum dots.
`An objective aperture was used to selectively image the (100),
`(002), and (101) wurtzite lattice planes. The samples were
`prepared by placing one drop of a dilute solution of dots in
`octane onto a copper grid supporting a thin film of amorphous
`carbon and then wicking off the remaining solvent after 30 s.
`A second thin layer of amorphous carbon was evaporated onto
`the samples in order to minimize charging and reduce damage
`to the particles caused by the electron beam.
`Small-Angle X-ray Scattering (SAXS) in Polymer Films.
`Small-angle X-ray scattering (SAXS) samples were prepared
`using either poly(vinyl butyral) (PVB) or a phosphine-func-
`tionalized diblock copolymer [methyltetracyclododecene]300-
`[norbornene-CH2O(CH2)5P(oct)2]20, abbreviated as (MTD300P20),
`as the matrix.23 Approximately 5 mg of nanocrystallites of
`dispersed in 1 mL of toluene, added to 0.5 mL of a solution
`containing 10 wt % PVB in toluene, concentrated under vacuum
`to give a viscous solution, and then cast onto a silicon wafer.
`The procedure is the same for MTD300P20, except THF is used
`
`2 of 13
`
`
`
`(CdSe)ZnS Core-Shell Quantum Dots
`
`J. Phys. Chem. B, Vol. 101, No. 46, 1997 9465
`
`as the solvent for both nanocrystallites and polymer. The
`resulting (cid:24)200 (cid:237)m thick film is clear to slightly opaque. X-ray
`diffraction spectra were collected on a Rigaku 300 Rotaflex
`diffractometer operating in the Bragg configuration using Cu
`KR radiation. The accelerating voltage was set at 60 kV with
`a 300 mA flux. Scatter and diffraction slits of 1/6(cid:176) and a 0.3
`mm collection slit were used.
`Small-Angle X-ray Scattering in Dilute Solutions. The
`X-ray source was a rotating copper anode operated at 40 kV
`and 25 mA. The apparent point source (electron beam irradiated
`area on the anode) was about 10-2 mm2. The beam was
`collimated onto a position sensitive detector, PSPE (ELPHYSE).
`A thin slit, placed before the filter, selects a beam with the
`dimensions of 3 (cid:2) 0.3 mm2 on the detector. The position
`sensitive linear detector has a useful length of 50 mm, placed
`at a distance D ) 370 mm from the detector. The spatial
`resolution on the detector is 200 (cid:237)m. This setup allows a
`continuous scan of scattering wavevectors between 6 (cid:2) 10-3
`and 0.40 Å-1, with a resolution of about 3 (cid:2) 10-3 Å-1.
`The samples used were quartz capillary tubes with about 1
`mm optical path, filled with the desired dispersion, and then
`flame-sealed after filling. The intensity from the reference, Iref,
`is collected first, and then the intensity from the sample, Is. The
`intensity used in the data analysis is the difference: I ) Is -
`Iref.
`Wide-Angle X-ray Scattering (WAXS). The wide-angle
`X-ray powder diffraction patterns were measured on the same
`setup as the SAXS in polymer dispersions. The TOPO/TOP
`capped nanocrystals were precipitated with methanol and
`exchanged with pyridine. The samples were prepared by
`dropping a heavily concentrated solution of nanocrystals
`dispersed in pyridine onto silicon wafers. A slow evaporation
`of the pyridine leads to the formation of glassy thin films which
`were used for the diffraction experiments.
`
`III. Results and Analysis
`A. Synthesis of Core-Shell Composite Quantum Dots.
`We use a two-step synthetic procedure similar to that of Danek
`et al.16 and Hines et al.19 to produce (CdSe)ZnS core-shell
`quantum dots.
`In the first step we synthesize nearly mono-
`disperse CdSe nanocrystallites ranging in size from 23 to 55 Å
`via a high-temperature colloidal growth followed by size
`selective precipitation.3 These dots are referred to as “bare”
`dots in the remainder of the text, although their outermost
`surface is passivated with organic TOPO/TOP capping groups.
`Next, we overcoat the CdSe particles in TOPO by adding the
`intermediate temperatures.22 The
`Zn and S precursors at
`resulting composite particles are also passivated with TOPO/
`TOP on their outermost surface.
`The temperature at which the dots are overcoated is very
`critical. At higher temperatures the CdSe seeds begin to grow
`via Ostwald ripening, and their size distribution deteriorates,
`leading to broader spectral line widths. Overcoating the particles
`at relatively low temperatures could lead to incomplete decom-
`position of the precursors or to reduced crystallinity of the ZnS
`shell. An ideal growth temperature is determined independently
`for each CdSe core size to ensure that the size distribution of
`the cores remains constant and that shells with a high degree
`of crystallinity are formed.22
`The concentration of the ZnS precursor solution and the rate
`at which it is added are also critical. Slow addition of the
`precursors at low concentrations ensures that most of the ZnS
`grows heterogeneously onto existing CdSe nuclei instead of
`undergoing homogeneous nucleation. This probably does not
`eliminate the formation of small ZnS particles completely so a
`final purification step in which the overcoated dots are subjected
`
`Figure 1. Absorption spectra for bare (dashed lines) and 1-2
`monolayer ZnS overcoated (solid lines) CdSe dots with diameters
`measuring (a) 23, (b) 42, (c) 48, and (d) 55 Å. The absorption spectra
`for the (CdSe)ZnS dots are broader and slightly red-shifted from their
`respective bare dot spectra.
`
`Figure 2. Photoluminescence (PL) spectra for bare (dashed lines) and
`ZnS overcoated (solid lines) dots with the following core sizes: (a)
`23, (b) 42, (c) 48, and (d) 55 Å in diameter. The PL spectra for the
`overcoated dots are much more intense owing to their higher quantum
`yields: (a) 40, (b) 50, (c) 35, and (d) 30.
`
`to size selective precipitation provides further assurance that
`mainly (CdSe)ZnS particles are present in the final powders.
`B. Optical Characterization. The synthesis presented
`above produces ZnS overcoated dots with a range of core and
`shell sizes. Figure 1 shows the absorption spectra of CdSe dots
`ranging from 23 to 55 Å in diameter before (dashed lines) and
`after (solid lines) overcoating with 1-2 monolayers of ZnS.
`The definition of a monolayer here is a shell of ZnS that
`measures 3.1 Å (the distance between consecutive planes along
`the [002] axis in bulk wurtzite ZnS) along the major axis of
`the prolate-shaped dots. We observe a small shift
`in the
`absorption spectra to the red (lower energies) after overcoating
`due to partial leakage of the exciton into the ZnS matrix. This
`red shift is more pronounced in smaller dots where the leakage
`of the exciton into the ZnS shell has a more dramatic effect on
`the confinement energies of the charge carriers. Figure 2 shows
`the room-temperature photoluminescence spectra (PL) of these
`
`3 of 13
`
`
`
`9466 J. Phys. Chem. B, Vol. 101, No. 46, 1997
`
`Dabbousi et al.
`
`Figure 3. Color photograph demonstrating the wide spectral range of bright fluorescence from different size samples of (CdSe)ZnS. Their PL
`peaks occur at (going from left to right) 470, 480, 520, 560, 594, and 620 nm (quartz cuvettes courtesy of Spectrocell Inc., photography by F.
`Frankel).
`
`same samples before (dashed lines) and after (solid lines)
`overcoating with ZnS. The PL quantum yield increases from
`5 to 15% for bare dots to values ranging from 30 to 50% for
`dots passivated with ZnS.
`In smaller CdSe dots the surface-
`to-volume ratio is very high, and the PL for TOPO capped dots
`is dominated by broad deep trap emission due to incomplete
`surface passivation. Overcoating with ZnS suppresses deep trap
`emission by passivating most of the vacancies and trap sites on
`the crystallite surface, resulting in PL which is dominated by
`band-edge recombination.
`Figure 3 (color photograph) displays the wide spectral range
`of luminescence from (CdSe)ZnS composite quantum dots. The
`photograph shows six different samples of ZnS overcoated CdSe
`dots dispersed in dilute hexane solutions and placed in identical
`quartz cuvettes. The samples are irradiated with 365 nm
`ultraviolet light from a UV lamp in order to observe lumines-
`cence from all the solutions at once. As the size of the CdSe
`core increases, the color of the luminescence shows a continuous
`progression from blue through green, yellow, orange, to red. In
`the smallest sizes of TOPO capped dots the color of the PL is
`normally dominated by broad deep trap emission and appears
`as faint white light. After overcoating the samples with ZnS
`the deep trap emission is nearly eliminated, giving rise to intense
`blue band-edge fluorescence.
`To understand the effect of ZnS passivation on the optical
`and structural properties of CdSe dots, we synthesized a large
`quantity of (cid:24)40 Å diameter CdSe dots. We divided this sample
`into multiple fractions and added varying amounts of Zn and S
`precursors to each fraction at identical temperatures and addition
`times. The result was a series of samples with similar CdSe
`cores but with varying ZnS shell thickness. Figure 4 shows
`the progression of the absorption spectrum for these samples
`with ZnS coverages of approximately 0 (bare TOPO capped
`CdSe), 0.65, 1.3, 2.6, and 5.3 monolayers. (See beginning of
`this section for definition of number of monolayers.) The
`spectra reflect a constant area under the lowest energy 1S3/2-
`1Se absorption peak (constant oscillator strength) for the samples
`with varying ZnS coverage. As the thickness of the ZnS shell
`increases, there is a shift in the 1S3/2-1Se absorption to the red,
`
`Figure 4. Absorption spectra for a series of ZnS overcoated samples
`grown on identical 42 Å ( 10% CdSe seed particles. The samples
`displayed have the following coverage: (a) bare TOPO capped, (b)
`0.65 monolayers, (c) 1.3 monolayers, (d) 2.6 monolayers, and (e) 5.3
`monolayers (see definition for monolayers in text). The right-hand side
`shows the long wavelength region of the absorption spectra showing
`the lowest energy optical transitions. The spectra demonstrate an
`increased red-shift with thicker ZnS shells as well as a broadening of
`the first peak as a result of increased polydispersity. The left-hand side
`highlights the ultraviolet region of the spectra showing an increased
`absorption at higher energies with increasing coverage due to direct
`absorption into the ZnS shell.
`
`reflecting an increased leakage of the exciton into the shell, as
`well as a broadening of the absorption peak, indicating a
`distribution of shell thickness. The left-hand side of Figure 4
`shows an increased absorption in the ultraviolet with increasing
`ZnS coverage as a result of direct absorption into the higher
`band gap ZnS shell.
`The evolution of the PL for the same (cid:24)40 Å diameter dots
`with ZnS coverage is displayed in Figure 5. As the coverage
`of ZnS on the CdSe surface increases, we see a dramatic increase
`in the fluorescence quantum yield followed by a steady decline
`
`4 of 13
`
`
`
`(CdSe)ZnS Core-Shell Quantum Dots
`
`J. Phys. Chem. B, Vol. 101, No. 46, 1997 9467
`
`TABLE 1: Summary of the Results Obtained from WDS, TEM, SAXS, and WAXS Detailing the Zn/Cd Ratio, Average Size,
`Size Distribution, and Aspect Ratio for a Series of (CdSe)ZnS Samples with a (cid:24)40 Å Diameter CdSe Cores and Varying ZnS
`Coverage
`ZnS coverage
`(TEM)
`
`measd Zn/Cd
`ratio (WDS)
`
`calcd Zn/Cd ratio
`(SAXS in polymer)
`
`calcd Zn/Cd ratio
`(WAXS)
`
`0.46
`1.50
`3.60
`6.80
`
`0.58
`1.32
`
`0.7
`1.4
`2.9
`6.8
`
`bare
`0.65 monolayers
`1.3 monolayers
`2.6 monolayers
`5.3 monolayers
`
`measd TEM size
`39 Å ( 8.2%
`43 Å ( 11%
`47 Å ( 10%
`55 Å ( 13%
`72 Å ( 19%
`
`measd average
`aspect ratio
`1.12
`1.16
`1.16
`1.23
`1.23
`
`calcd size (SAXS
`in polymer)
`42 Å ( 10%
`46 Å ( 13%
`50 Å ( 18%
`
`Figure 6. (A) Survey spectra of (a) (cid:24)40 Å diameter bare CdSe dots
`and (b) the same dots overcoated with ZnS showing the photoelectron
`and Auger transitions from the different elements present in the quantum
`dots. (B) Enlargement of the low-energy side of the survey spectra,
`emphasizing the transitions with low binding energy.
`shows the survey spectra of (cid:24)40 Å diameter bare dots and of
`the same sample overcoated with (cid:24)1.3 monolayers of ZnS. The
`presence of C and O comes mainly from atmospheric contami-
`nation during the brief exposure of the samples to air (typically
`around 15 min). The positions of both C and O lines correspond
`to standard values for adsorbed species, showing the absence
`of significant charging.24 As expected, we detect XPS lines
`from Zn and S in addition to the Cd and Se lines. Although
`the samples were exchanged with pyridine before the XPS
`measurements, small amounts of phosphorus could be detected
`on both the bare and ZnS overcoated CdSe dots, indicating the
`presence of residual TOPO/TOP molecules bound to Cd or Zn
`on the nanocrystal surfaces.25 The relative concentrations of
`Cd and Se are calculated by dividing the area of the XPS lines
`by their respective sensitivity factors.24
`In the case of nano-
`crystals the sensitivity factor must be corrected by the integral
`s
`e-z/(cid:236) dz to account for the similarity between the size of the
`nanocrystals and the escape depths of the electrons.26 The
`integral must be evaluated over a sphere to obtain the Se/Cd
`ratios in CdSe dots. In the bare CdSe nanocrystals the Se/Cd
`ratio was around 0.87, corresponding to 46% Se and 54% Cd.
`This value agrees with the WDS results.
`We use the Auger parameter, defined as the difference in
`binding energy between the photoelectron and Auger peaks, to
`identify the nature of the bond in the different samples.24 This
`difference can be accurately determined because static charge
`corrections cancel. The Auger parameter of Cd in the bare and
`
`0d
`
`Figure 5. PL spectra for a series of ZnS overcoated dots with 42 (
`10% Å diameter CdSe cores. The spectra are for (a) 0, (b) 0.65, (c)
`1.3, (d) 2.6, and (e) 5.3 monolayers ZnS coverage. The position of the
`maximum in the PL spectrum shifts to the red, and the spectrum
`broadens with increasing ZnS coverage. (inset) The PL quantum yield
`is charted as a function of ZnS coverage. The PL intensity increases
`with the addition of ZnS reaching, 50% at (cid:24)1.3 monolayers, and then
`declines steadily at higher coverage. The line is simply a guide to the
`eye.
`after (cid:24)1.3 monolayers of ZnS. The spectra are red-shifted
`(slightly more than the shift in the absorption spectra) and show
`an increased broadening at higher coverage. The inset to Figure
`5 charts the evolution of the quantum yield for these dots as a
`function of the ZnS shell thickness. For this particular sample
`the quantum yield starts at 15% for the bare TOPO capped CdSe
`dots and increases with the addition of ZnS, approaching a
`maximum value of 50% at approximately (cid:24)1.3 monolayer
`coverage. At higher coverage the quantum yield begins to
`decrease steadily until it reaches a value of 30% at about (cid:24)5
`monolayer coverage. In the following sections we explain the
`trends in PL quantum yield based on the structural characteriza-
`tion of ZnS overcoated samples.
`C. Structural Characterization. WaVelength DispersiVe
`X-ray Spectroscopy. We analyze the elemental composition of
`the ZnS overcoated samples using wavelength dispersive X-ray
`spectroscopy (WDS). This method provides a quantitative
`analysis of the elemental composition with an uncertainty of
`less than (5%. We focus on obtaining a Zn/Cd ratio for the
`ZnS overcoated samples of interest. Analysis of the series of
`samples with a (cid:24)40 Å diameter core and varying ZnS coverage
`gives the Zn/Cd ratios which appear in Table 1. The WDS
`analysis confirms that the Zn-to-Cd ratio in the composite dots
`increases as more ZnS is added. We also use this technique to
`measure the Se/Cd ratio in the bare dots. We consistently
`measure a Se/Cd ratio of (cid:24)0.8-0.9/1,
`indicating Cd-rich
`nanoparticles.
`X-ray Photoelectron Spectroscopy. Multiple samples of
`(cid:24)33 and (cid:24)40 Å diameter CdSe quantum dots overcoated with
`variable amounts of ZnS were examined by XPS. Figure 6
`
`5 of 13
`
`
`
`9468 J. Phys. Chem. B, Vol. 101, No. 46, 1997
`
`Dabbousi et al.
`
`Figure 8. Transmission electron micrographs of (A) one “bare” CdSe
`nanocrystallite and (B) one CdSe nanocrystallite with a 2.6 monolayer
`ZnS shell.
`
`of the XPS and Auger intensities of the Cd photoelectrons for
`bare and overcoated samples.14,28 The depth dependence of the
`observed intensity for the Auger and XPS photoemitted electrons
`is
`
`I ) J0N(z)i(cid:243)iYi,nF(KE)e-z/(cid:236)(KE)
`
`(1)
`
`where J0 is the X-ray flux, N(z)i is the number of i atoms, (cid:243)i is
`the absorption cross section for atoms i, Yi,n is the emission
`quantum yield of Auger or XPS for atoms i, F(KE) is the
`energy-dependent instrument response function, and (cid:236)(KE) is
`the energy-dependent escape depth. Taking the ratio of the
`intensities of the XPS and Auger lines from the same atom, Cd
`or Zn, it is possible to eliminate the X-ray flux, number of atoms,
`and absorption cross sections from the intensity equations for
`the Auger and the primary X-ray photoelectrons. The value of
`the intensity ratio I ) iovercoated(Cd)/ibare(Cd), where i ) iXPS-
`(Cd)/iAuger(Cd), is only a function of the relative escape depths
`of the electrons. Therefore, due to the smaller escape depths
`of the Cd Auger electrons in both ZnS (13.2 Å) and CdSe (10
`Å) compared to the Cd XPS photoelectron (23.7 Å in ZnS and
`15 Å in CdSe), the intensity I should increase with the amount
`of ZnS on the CdSe surface. Calculated values of 1.28 and
`1.60 for the 0.65 and 2.6 monolayer, respectively, confirm the
`growth of ZnS on the surface of the CdSe dots.
`Transmission Electron Microscopy. High-resolution TEM
`allows us to qualitatively probe the internal structure of the
`composite quantum dots and determine the average size, size
`distribution, and aspect ratio of overcoated particles as a function
`of ZnS coverage. We image the series of (CdSe)ZnS samples
`described earlier. Figure 8 shows two dots from that series,
`one with (A) no ZnS overcoating (bare) and one with (B) 2.6
`monolayers of ZnS. The particles in the micrographs show well-
`resolved lattice fringes with a measured lattice spacing in the
`bare dots similar to bulk CdSe. For the 2.6 monolayer sample
`these lattice fringes are continuous throughout the entire particle;
`the growth of the ZnS shell appears to be epitaxial. A well-
`defined interface between CdSe core and ZnS shell was not
`observed in any of the samples, although the “bending” of the
`lattice fringes in Figure 8Bsthe lower third of this particle is
`slightly askew compared with the upper partsmay be suggestive
`of some sort of strain in the material. This bending is somewhat
`anomalous, however, as the lattice fringes in most particles were
`straight. Some patchy growth is observed for the highest
`coverage samples, giving rise to misshapen particles, but we
`do not observe discrete nucleation of tethered ZnS particles on
`the surface of existing CdSe particles. We analyze over 150
`crystallites in each sample to obtain statistical values for the
`length of the major axis, the aspect ratio, and the distribution
`of lengths and aspect ratios for all the samples. Figure 9 shows
`histograms of size distributions and aspect ratio from these same
`samples. This figure shows the measured histograms for (A)
`
`Figure 7. X-ray photoelectron spectra highlighting the Se 3d core
`transitions from (cid:24)40 Å bare and ZnS overcoated CdSe dots: (a) bare
`CdSe, (b) 0.65 monolayers, (c) 1.3 monolayers, and (d) 2.6 monolayers
`of ZnS. The peak at 59 eV indicates the formation of selenium oxide
`upon exposure to air when surface selenium atoms are exposed.
`
`overcoated samples is 466.8 ( 0.2 eV and corresponds exactly
`to the expected value for bulk CdSe.
`In the case of ZnS the
`Auger parameter for Zn in the 1.3 and 2.6 monolayer ZnS
`samples is 757.5 eV, which is also very close to the expected
`value of 758.0 eV.
`The degree of passivation of the CdSe surface with ZnS is
`examined by exposing the nanocrystal surface to air for extended
`periods of time and studying the evolution of the Se peak.
`The oxidation of CdSe quantum dots leads to the formation
`of a selenium oxide peak at higher energies than the main Se
`peak.27 Figure 7 shows the formation of a SeO2 peak at 59 eV
`after an 80 h exposure to air in both the bare, TOPO capped,
`CdSe and 0.65 monolayer ZnS overcoat