Double-crowned 2D semiconductor nanoplatelets with bicolor power-tunable emission | Panda Anku


Octadecene (ODE, Alpha Aesar, 90%), cadmium acetate dihydrate (Cd(OAc)2.2H2O,Sigma-Aldrich, 98%), cadmium oxide (CdO, Strem Chemicals, 99.99%), myristic acid (Aldrich, >99%), oleic acid (OA, Alpha Aesar 90%), trioctylphosphine (TOP, Alpha Aesar, 90%), selenium (Strem Chemicals 99.99%), sulfur (Sigma-Aldrich), tellurium (Alpha Aesar, 18 + 60 mesh, 99.999%), hexane (VWR Chemicals), ethanol absolute (VWR Chemicals), methanol (VWR Chemicals)

1M TOP:S precursor

In a glovebox, 20 mL of TOP are mixed with 0.64 mg of S powder. The mixture is stirred for a whole night and is then stored in the glovebox for further use.

1M TOP:Se precursor

In a glove box, 20 mL of TOP are mixed with 1.58 mg of Se powder. The mixture is stirred for a whole night and is then stored in the glovebox for further use.

1M TOP:Te precursor

2.54 g of Te powder are mixed in 20 mL of TOP in a three-neck flask. The flask is kept under vacuum at room temperature for 5 min and then the temperature is raised to 100 °C. Furthermore, degassing of the flask is conducted for the next 20 min. The atmosphere is switched to nitrogen and the temperature is raised to 275 °C. The solution is stirred until a clear orange coloration is obtained. The flask is cooled down to room temperature and the color changes to yellow. Finally, this solution is transferred to a nitrogen-filled glove box for storage.

Cd(Myr)2 precursor

In a 50 mL three necks flask, 2.56 g of cadmium oxide and 11 g of myristic acid are mixed and degassed at 80 °C for 30 min. The atmosphere is then switched to argon and the temperature is set at 200 °C. The mixture is heated for 40 min until the solution becomes colorless. The solution is then cooled down and 30 mL of methanol are added at 60 °C. The formed cadmium myristate is washed five times by centrifugation using methanol. The final solid is dried under vacuum at 70 °C for a whole night.

CdSe 4ML core

In a 50 mL three necks flask, 340 mg of Cd(Myr)2, 24 mg of Se powder and 25 mL of ODE are added. After 20 min of degassing at room temperature, the atmosphere is switched to argon and the temperature is set at 230 °C. When the temperature reaches ~203 °C, 110 mg of Cd(OAc)2.2H2O are swiftly added. The solution is heated for 20 min and then cooled down to room temperature. At 150 °C, 500 µL of oleic acid is added. Then the obtained solution is precipitated with 20 mL of hexane and 30 mL of ethanol. The obtained pellets are washed a second time using less ethanol. The NPLs are finally redispersed in 10 mL of ODE.

CdSe-CdTe core-crown

In a 25 mL three-neck flask, 92 mg of dried Cd(OAc)2, 180 µL of oleic acid and 1 mL of CdSe 4 ML core NPLs (O.D.: 1 at 512 nm for 50 µL in 3 mL of hexane) redispersed in 5 mL of ODE are degassed for 30 min at room temperature then 30 min at 80 °C. The atmosphere is switched to argon and the temperature is set at 205 °C. When the temperature is stabilized, 0.25 mL (1eq) of a solution of TOP:Te (1 M) in ODE (final concentration 0.01 M) is added at a 2 mL h−1 rate. After the injection, the mixture is further heated for 10 min.

CdSe external crown

After the growth of the CdTe crown, the temperature is set at 215 °C. When the temperature is stabilized, 2 mL of TOP:Se (1 M) in ODE (final concentration 0.1 M) is added at a 1 mL h−1 rate. After the injection, the mixture is cooled down to room temperature and the NPLs are precipitated with hexane and ethanol for 5 min. The final pellets are redispersed in hexane.

Transmission electron microscopy on NCs

A drop of the NC solution is drop-casted onto a copper grid covered with an amorphous carbon film. The grid is degassed overnight to reduce future contamination. A JEOL 2010F is used for the acquisition of pictures and operated at 200 kV. Complementarily, TEM/STEM observations were made on a Titan Themis 200 microscope (FEI/Thermo Fischer Scientific) equipped with a geometric aberration corrector on the probe. The microscope was also equipped with the “Super-X” systems for EDX analysis with a detection angle of 0.9 steradian. The observations were made at 200 kV with a probe current of about 35 pA and a half-angle of convergence of 17 mrad. HAADF-STEM images were acquired with a camera length of 110 mm (inner/outer collection angles were respectively 69 and 200 mrad).

k·p modeling

Electron and heavy hole states are calculated with single-band k·p Hamiltonians. These include self-energy terms arising from the dielectric mismatch with the organic medium, as well as strain arising from the lattice mismatch between the core and the crown. The latter is obtained within the continuum elastic approximation. Excitonic interactions are obtained with a self-consistent calculation, where Coulomb integrals are obtained by integrating the Poisson equation in a dielectrically inhomogeneous system, see SI of ref. 15. for details of the model and material parameters. Many-hole calculations (Supplementary Note 11) are carried out using full configuration routines. The basis set is then formed by all possible combinations of the eight lowest spin-orbitals corresponding to the Ag, B1g, B2u, and B3u irreducible representations of the NPL point group (D2h) group, which provide the top-most hole states in CdTe crowns42. The calculations in Fig. 2 correspond to CdSe/CdTe/CdSe NPLs with 4.5 monolayer thickness. Following TEM data for a typical reference sample, the CdSe core is taken with dimensions 7 × 30 nm2, the CdTe crown 9 × 32 nm2 and the external CdSe crown 42 × 65 nm2. While experimental samples may have even larger external crowns, the dimensions we choose are already in the asymptotical limit for the description of the ground state (red peak), as shown in Supplementary Note 10.

Optical spectroscopy

Absorption/PL measurements

UV-visible spectra are acquired with a Cary 5000 spectrometer. PL and excitation spectra are obtained with an Edinburgh instrument spectrometer. During the measurements, the NPLs are dispersed in hexane.


In order to compare the weight of each component in the PL spectra (green vs red emission) and be able to identify the processes that govern the PL power dependence, only light coming from a homogeneously excited volume should be collected (to probe equally each excitation-relaxation channel). This ‘small’ volume that roughly corresponds to the waist region of the incident beam was addressed with a confocal-like setup (afocal configuration) using the spectrometer slit (Princeton Instruments, Acton SP2750) as a spatial filter. To reach strong rejection and select the waist emission the latter was coupled to an infinity-corrected microscope objective (NA ≈ 0.6, equivalent focal length ≈6 mm) used to excite and collect light in reflection configuration. A cooled CCD (Spec10, PI) was used as a detector at the exit of the spectrometer. The excitation was supplied by a laser diode (Alphalas) operating at 407 nm (Δt ≈ 70 ps). The repetition rate was adjusted at ≈300 kHz to allow complete inter-pulse relaxation of the long lifetime species responsible for the red emission. A long-pass edge filter (LP03-458RE-25, λcutoff ≈ 458 nm) from Semrock company was also placed along the detection path to suppress scattered light from the excitation beam.

Time-resolved PL

PL time-resolved measurements were performed using the same ‘confocal’ configuration; with two different methods. Relatively ‘long’ lifetime decays (associated with the red emission) were characterized through TCSPC using a correlator board from Picoquant (TimeHarp 260), an avalanche photodiode for the detection (MPD company, PDM module accommodating a dark count rate of ≈25 counts/s) and an Alphalas laser diode to excite the material (λ ≈ 407 nm, Δt ≈ 70 ps). The setup IRF is then measured to be ≈220 ps. Fast decays (green emission) were measured with a streak-camera (C5680 model from Hamamatsu incorporating an M5675 synchroscan unit) coupled to our Acton SP2750 spectrometer. In this configuration, the excitation (pulses of ≈2 ps duration) is the second harmonic of a Titanium-sapphire laser operating at 82 MHz and a temporal resolution of ≈ 15 ps is typically obtained depending on the dispersion of the PL through the spectrometer. Different filters combinations are used throughout time-resolved experiments: colored filters—from Corning and Thorlabs—in order to extract the spectrum part of interest as well as an additional highly selective Semrock filter (to reject photons from the laser). Due to the high emission yield of the system, a great attention was drawn during TCSPC to keep the ratio of the count rate to the excitation frequency below 2%, in order to avoid pile-up effects deleterious to the counting statistics.

All the μ-PL experiments were carried on NPLs dissolved in hexane at a relatively low concentration (O.D. ≈ 0.7 at 510 nm for a 10 mm beam path), chosen to keep a satisfying signal to noise ratio under low power excitation (nW range). It was carefully ensured that inter-NPLs effects could be discarded in the so-defined experimental conditions.

Transient absorption

Transient absorption measurements were performed by splitting the 800 nm fundamental of a 2 kHz 35 fs Ti:Sapphire laser (SpectraPhysics) into two branches. One branch, the pump, was frequency-doubled to 400 nm, chopped to 1 kHz, and focused on the sample. The other branch, the probe, was focused into a 2 mm sapphire disk to generate a white light supercontinuum and then focused onto the sample. The beams were overlapped spatially on the sample and the pump-probe delay was controlled by a delay stage. Spectra of the white light supercontinuum were collected under pump-on and pump-off conditions to generate ΔA data using Helios software (Ultrafast systems).

Amplified spontaneous emission

ASE measurements were performed using a frequency-doubled Ti:Sapphire pump excitation (400 nm, ≈35 fs, 100 Hz) focused as a 4 mm stripe on a thin-film sample of nanoplatelets. The pump power was controlled with continuous optical density wheels. Emitted light was collected normal to the pump excitation direction, focused into a fiber, and directed to a spectrometer and CCD. Films for measurements were prepared by drop-casting 9:1 hexane: octane solutions of nanoplatelets onto clean glass slides, to form smooth, reflective films. A similar configuration (using a circular lens with front-face collection) was employed to perform power-dependent PL experiments on dilute samples in cuvettes.

Single particle measurements

In the experiments the PL was collected with a 0.6 NA, infinity corrected objective (producing a <1 μm diameter spot) and analyzed using a 2750 Acton Spectrometer from Princeton Instruments, keeping the confocal configuration described for the TRPL. The NPLs were dispersed on ≈120 microns thick coverslips that were ‘stuck’ to the cold finger of a He-flow micro-PL cryostat (Oxford Instruments) with silver particles-based varnish to ensure a good thermal contact42,43. The excitation was tuned at 390 nm (SHG of a Ti-Sapphire laser delivering ps duration pulses). A dichroic filter (from Semrock company) was placed along the optical path (FF01-430/LP-25) to suppress scattering from the excitation beam; note that the scattered light from the Ti-Sapphire pump laser (THG of a diode-pumped Nd:YAG laser) could not be totally suppressed and is always present in the shown spectra as a sharp line peaking at 532 nm. Due to the poor sample stability the pump intensity was limited to a fraction of microwatt and the spectra were recorded with relatively low integration times (4–10 s typically). It is finally important noting that the strong and ≈1 s scale operating spectral diffusion was clearly identified as an important source of spectral broadening. The phenomenon was not further investigated in the course of the present project.

Photoemission spectroscopy

Sample preparation for photoemission

Silicon wafers are rinsed with acetone, sonicated in acetone for 5 min. They are rinsed again with acetone and isopropanol and dried with N2 gun. A 5 nm layer of Cr and an 80 nm layer of Au are deposited using thermal evaporation. A diluted solution of NPLs in a mix of hexane octane (9:1) is drop-casted on the prepared substrate. After drying, the film is dipped in a solution of EDT (1% in acetonitrile) for 1 min. The procedure is repeated 3 times. The film final is stored under inert atmosphere before its introduction into the preparation chamber, where it is degassed for at least two hours and then transferred to the analysis chamber.

Photoemission data acquisition

XPS experiments are carried out on the TEMPO beamline from SOLEIL French synchrotron facility. The photon sources were HU80 and HU44 Apple II undulators set to deliver linearly polarized light. The photon energy is selected using a high-resolution plane grating monochromator. During the XPS measurements, the photoelectrons are detected at 0° from the sample surface normal (vec{{{{{{bf{n}}}}}}}) and at 44° from the polarization vector (vec{{{{{{bf{E}}}}}}}). The spot size is 100 × 80 μm². The signal is acquired onto a MBS A-1 photoelectron analyzer equipped with a delay line detector developed by Elettra44.

Energy calibration

Valence and secondary electron cut-off measurements are conducted at 150 eV which corresponds to a surface sensitive condition, while core levels are typically acquired using a 700 eV photon energy. The photon energy is precisely measured using the first and second order of a given core level using the formula: (h{nu }_{{{exp}}}={{{{{{rm{KE}}}}}}}_{{2}^{{{{{{rm{nd}}}}}}}}-{{{{{{rm{KE}}}}}}}_{{1}^{{{{{{rm{st}}}}}}}}), where KE stands for kinetic energy. The work function of the analyzer (({{{{{{rm{WF}}}}}}}_{A})) is determined by measuring the kinetic energy of electrons at the Fermi level from a gold reference sample. The binding energy of the Fermi level is set to 0 eV. ({{{{{{rm{WF}}}}}}}_{A}=h{nu }_{{{exp}}}-{{{{{{rm{KE}}}}}}}_{{{{{{rm{Fermi}}}}}}})

Valence band measurement

We determine the value of ({V}_{B}-{E}_{F})by looking at high KE electrons. We measure the highest kinetic energy available (({{{{{{rm{KE}}}}}}}_{{{{{{rm{VB}}}}}}})) and extract ({V}_{B}-{E}_{F}) with the formula: ({V}_{B}-{E}_{F}=h{nu }_{{{exp }}}-{{{{{{rm{KE}}}}}}}_{{{{{{rm{VB}}}}}}}-{{{{{{rm{WF}}}}}}}_{A}).

Core level

All spectra are calibrated in energy by shifting them so that the Fermi level of metallic samples presents null binding energy. A Shirley background is subtracted in all core level spectra. The core level is then fitted using a Voigt curve, which displays a typical 0.8 eV full width at half maximum.

Work function measurement

In order to measure the work function, which is the difference in energy between vacuum level and Fermi level, we look for the cut-off of secondary electrons (({{{{{{rm{KE}}}}}}}_{{{{{{rm{Cut; off}}}}}}})). We start by polarizing the sample using an 18 V (({{{{{{rm{Pol}}}}}}}_{{{{{{rm{Bias}}}}}}})) voltage supply (TDK lambda) and we look for the energy edge of the lowest kinetic energy photoelectrons. The work function is deduced with the formula: ({{{{{{rm{WF}}}}}}}_{{{{{{rm{Sample}}}}}}}={{{{{{rm{KE}}}}}}}_{{{{{{rm{Cut; off}}}}}}}-{{{{{{rm{Pol}}}}}}}_{{{{{{rm{Bias}}}}}}}).

LED fabrication

Materials for LED

PEDOT:PSS (poly(3,4- ethylenedioxythiophene) polystyrene sulfonate, Al 4083, M121, Ossila), Poly-TPD (Poly(N,N’-bis-4- butylphenyl-N,N’-bisphenyl)benzidine, Ossila), PVK (Poly(9-vinyl) carbazole, average Mn 25,000–50,000, Aldrich), PMMA (polymethyl metacrylate, Arkema), chlorobenzene (anhydrous, 99.8%, Sigma-Aldrich), m-xylene (anhydrous, ≥99%, Sigma-Aldrich), epoxy-glue (Ossila), zinc acetate dihydrate (<97%, Alfa Aesar), tetramethylammonium hydroxide pentahydrate (TMAOH, 98%, Alfa Aesar), dimethyl sulfoxide (DMSO, ≥99.9%, Sigma-Aldrich), ethyl acetate (VMR), ethanol absolute anhydrous (VMR), octane (VWR, technical) and acetone (VWR). All the materials were used as received.

Synthesis of ZnO nanoparticles

The procedure is taken from ref. 41. In flask A, 3 mmol of zinc acetate are dissolved in 30 mL of DMSO by vigorous stirring. At the same time, 5.5 mmol of TMAOH are dissolved with 10 mL of ethanol in flask B. Then the contents of the two flasks are mixed and stirred for 24 h under ambient conditions. The reaction mixture turns whitish during the first few seconds and becomes clear soon after. ZnO particles are precipitated by ethyl acetate and redisperse in ethanol. 160 µL of 2-ethanolamine are added to stabilize the nanoparticles before they are precipitated and redispersed with ethyl acetate and ethanol respectively again. Finally, the ZnO nanoparticles in ethanol are filtered using a 0.22 µm PTFE filter.

ITO substrate patterning

ITO substrates (30 Ω/sq) are cut into 15 mm × 15 mm pieces and cleaned by sonication in acetone for 5 min. After sonication, the substrates are rinsed with acetone and isopropanol before being dried completely with N2 flow. The substrates are further cleaned with O2 plasma for 5 min to remove organic residuals on the surface. After cleaning, TI-Prime and AZ 5214E photoresist are sequentially spin-coated on the surface of ITO substrates at the rate of 4000 rpm for 30 s and baked at 110 °C for 120 s and 90 s, respectively. In the next stage, a mask aligner is used to expose the substrates to UV light for 20 s through a lithography mask (1 mm width). A photoresist is then developed using AZ 726 developer for 20 s before rinsing with deionized water and drying with N2 flux. After another 5-minute plasma cleaning, the substrates are etched in a 25% HCl (in water) bath for 10 min at 40 °C before they are dipped immediately in deionized water. Finally, the lift-off is conducted in an acetone bath. Before being used, the patterned ITO substrates are cleaned with acetone and isopropanol first and put under plasma for 10 min.

LED Fabrication

PEDOT:PSS solution (filtered through 0.45 µm filter) is spin-coated on a patterned ITO glass electrode at 4000 rpm for 60 s and annealed at 140 °C for 10 min in air. Inside a nitrogen-filled glovebox, Poly-TPD (8 mg mL−1 in chlorobenzene), PVK (1.5 mg mL−1 in m-xylene), NPLs (in a mix of hexane/octane (9:1)), PMMA (5 mg mL−1 in acetone) and ZnO nanoparticles are successively spin-coated at 2000 rpm for 45 s on the PEDOT: PSS-coated substrate. After the deposition of Poly-TPD, the sample is annealed at 110 °C for 20 min, and for PVK the annealing is at 170 °C for 30 min. Finally, 80 nm of Ag is deposited on top of the ZnO using a shadow mask by thermal evaporation. The thickness of NPL and ZnO layers are 18 nm and 80 nm respectively, as obtained by profilometry. The devices are encapsulated inside the glove box with a piece of glass by epoxy-glue. The size of the pixel is 1 mm2 which is the overlap of ITO and Ag electrodes.

LED characterization

The EQE of the device is determined according to the method from ref. 45. Considering the Lambertian emission of LED device, the flux leaving the device directly can be described as ({F}_{{{{{{rm{ext}}}}}}}={int }_{0}^{pi /2}2pi {L}_{0}{{cos}}theta {{sin}}theta dtheta=pi {L}_{0}), with ({L}_{0}) the flux per solid angle of light leaving the device in the forward direction. Since the solid angle from the photodetector to the light source is (Omega=frac{{S}_{1}}{{l}^{2}}) with ({S}_{1}) the area of the detector and (l) the distance between the light source and detector, then ({L}_{0}=frac{{P}_{{rm {det }}}}{Omega }=frac{{P}_{{{{{{rm{d}}}}}}{{{{{rm{et}}}}}}}{l}^{2}}{{s}_{1}}) and ({F}_{{{{{{rm{ext}}}}}}}=frac{{pi P}_{{{{{{rm{d}}}}}}{{{{{rm{et}}}}}}}{l}^{2}}{{s}_{1}}) . The number of photons emitted per second to the forward direction can then be calculated by ({N}_{P}=frac{{F}_{{{{{{rm{ext}}}}}}}}{hnu }=frac{{pi P}_{{{{{{rm{d}}}}}}{{{{{rm{et}}}}}}}{l}^{2}lambda }{{s}_{1}{hc}}), with (lambda) the wavelength of electroluminescence, (h) the Plank’s constant and (c) the speed of light. The number of electrons injected per second can be obtained by ({N}_{P}=frac{I}{e}), with (I) the current flow of the device. Thus, the EQE can be calculated as ({{{{{rm{E}}}}}}{{{{{rm{QE}}}}}}=frac{{N}_{p}}{{N}_{e}}=frac{{pi P}_{d{et}}{l}^{2}lambda e}{{s}_{1}{hcI}}) . The irradiance of the device is (R=frac{{F}_{{{{{{rm{ext}}}}}}}}{{s}_{2}}=frac{{pi P}_{{{{{{rm{d}}}}}}{{{{{rm{et}}}}}}}{l}^{2}}{{s}_{1}{s}_{2}}), with ({S}_{2}) the area of the pixel. The luminance L of the device is (L=frac{683.Vleft(lambda right).{F}_{{{{{{rm{ext}}}}}}}}{{pi .S}_{2}}), with (Vleft(lambda right)) the function of photonic eye sensitivity. For the characterization, we collected current-voltage-luminance characteristics with a Keithley K2634B sourcemeter unit and a PM100A powermeter coupled with the S120C Si detector from Thorlabs. Knowing that the working diameter of the detector area is 9.5 mm and assuming the distance between detector and device to be 6.5 mm, the geometry-related value is (frac{{l}^{2}}{{s}_{1}}approx, 0.6).

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Leave a Comment