Toward Bright Mid-Infrared Emitters: Thick-Shell n-Type HgSe/CdS Nanocrystals

A procedure is developed for the growth of thick, conformal CdS shells that preserve the optical properties of 5 nm HgSe cores. The n-doping of the HgSe/CdS core/shell particles is quantitatively tuned through a simple postsynthetic Cd treatment, while the doping is monitored via the intraband optical absorption at 5 μm wavelength. Photoluminescence lifetime and quantum yield measurements show that the CdS shell greatly increases the intraband emission intensity. This indicates that decoupling the excitation from the environment reduces the nonradiative recombination. We find that weakly n-type HgSe/CdS are the brightest solution-phase mid-infrared chromophores reported to date at room temperature, achieving intraband photoluminescence quantum yields of 2%. Such photoluminescence corresponds to intraband lifetimes in excess of 10 ns, raising important questions about the fundamental limits to achievable slow intraband relaxation in quantum dots.


Table of contents
. The mixture was sonicated at room temperature to dissolve, and was warmed to ~40°C before loading to a syringe pump.

Cd(acetate)2 doping solution:
A 0.05M solution of cadmium acetate was prepared by adding 13mg (0.05mmol) cadmium acetate dihydrate to 0.9mL hexadecane and 0.1mL oleylamine in a test tube. The mixture was heated in air to around ~100°C to dissolve the cadmium acetate. The solution formed a gel on cooling, but it could be dissolved on gentle warming.
I2 solution for doping modification: A 0.02M solution of I2 in TCE was prepared by adding 10mg I2 to 2mL TCE and sonicating for 15 mins in a glass vial. A 0.004M solution was prepared by adding 0.4mL of this solution to 1.6mL TCE in another vial.

1C: HgSe core QD synthesis
HgSe core QDs were synthesized following the procedure by Melnychuk et. al. 6 Briefly, 108mg of HgCl2 was added to a 3-neck flask with 10mL oleylamine. The flask was equipped with a stir bar, rubber sleeve stoppers with a thermocouple attached, and connected to a Schlenk line manifold. Three vacuum (p~1 torr) -Argon flush cycles were performed at room temperature, and the flask was heated at 100 0 C for 45 minutes. The temperature was then set to 90 0 C. A TMSSe solution (50uL in 0.95mL oleylamine) was prepared in a N2-purged glovebox and injected swiftly into the flask. The temperature dropped to 88 0 C and was kept at this temperature for 2 minutes. The flask was then cooled to room temperature using compressed air, and 10mL octane was added.
Purification of the QDs was done in air. On cooling, the solution was centrifuged to obtain the excess unreacted Hg-complex as a grey pellet. The supernatant containing the HgSe QDs was collected and precipitated by addition of ethanol, and centrifuged. The precipitate was dispersed in 0.1mL oleylamine and 4mL TCE. The solution might appear foamy at this stage due to presence of excess Hg 2+ species in solution. A second purification was done by addition of 0.8mL of 0.1M DDAB/TCE and precipitated with IPA. After centrifugation, the precipitate was dispersed in 0.1mL oleylamine and 2mL TCE and stored as a stock solution in the freezer. Concentration of the HgSe stock solution was determined by measuring the absorbance at 700 nm using the cross-section of HgSe as 3.5 × 10 -18 cm 2 per Hg atom (calculated from the reported cross-section at 415 nm 6 and the measured HgSe absorption spectrum). Around 30mg of HgSe QDs were obtained from the reaction, giving a yield of ~50% with respect to the Selenium added.

1D: Thin shell HgSe/CdS QD synthesis by cALD
HgSe/CdS QDs were synthesized by cALD adapting the work by Shen et. al. 7 with notable changes. The reaction was done in a scintillation vial heated at 50-55°C, since this helped prevent the formation of an emulsion. No difference was observed between syntheses in air or in glovebox, so the reaction was done in air. The reaction was performed in dark to prevent photo-oxidation of sulfide to sulfates.
The (NH4)2S stock solution was prepared by adding 1 mL of ammonium sulfide solution (40% in water) to 4 mL of formamide (FA) in the glovebox in a scintillation vial. The vial was capped with a rubber stopper and brought to the fume hood.
15mg of purified HgSe QDs was added to a scintillation vial with 4 mL of TCE, 0.4 mL OAm and equipped with a stir bar. 4 mL of formamide (FA) was added, and the vial was heated to 50-55°C with gentle stirring for 10 mins. The FA layer was discarded (to remove residual Hg 2+ ). For the first Cd layer, 4 mL FA was added with 30 mg cadmium acetate, and stirred at 50-55°C for 10 mins. The FA layer was then discarded, 4 mL FA was added and stirred for 2 mins. The FA layer was discarded to remove excess Cd 2+ and the washing was performed a second time. The washings are an essential part of the procedure to prevent independent nucleation of CdS.
For the first S-layer, 4 mL FA was added, and 0.5 mL of the (NH4)2S stock solution was added dropwise. The vial was stirred at 50-55°C for 2 mins. The FA layer was then discarded, 4 mL FA was added and stirred for 2 mins, and then discarded. The washing was done for a second time. Three more Cd-and S-layers were similarly performed using 32 mg, 35 mg and 40 mg cadmium acetate respectively, and 0.5 mL (NH4)2S stock solution. A fifth Cd-layer using 45 mg cadmium acetate was performed to n-dope the QDs. The stirring was done gently to avoid formation of an emulsion. If an emulsion was formed, few drops of isopropanol (IPA) were added to aid in the separation of the layers.
After the synthesis of the thin CdS shell, the QDs were purified twice by precipitationdispersion using IPA and TCE.
It is important to note that the cALD reaction is sensitive to the stirring speed. Excessive stirring leads to formation of an emulsion and incomplete removal of excess Cd 2+ and S 2-, leading to independent nucleation of CdS.

1E: Thin shell HgSe/CdS QD synthesis using Cd(PDTC)2
Unless otherwise noted, all syntheses of thin shell HgSe/CdS QDs were performed using this method. Cadmium bis(phenyldithiocarbamate) (Cd(PDTC)2) was used as highly reactive single-source precursor for deposition of a thin shell on HgSe at low temperatures, motivated by the work by Buhro and coworkers. 2 30mg of purified HgSe core QDs was added to a 3-neck flask, and 15mL of a 20% oleylamine -hexadecane solution was added. Three vacuum (p ~ 1 torr) -argon evacuation cycles were performed at room temperature. The temperature was then set to 80°C. When the temperature reached 40°C, 1.17mL of 0.1M Cd(PDTC)2 solution (2 monolayers (ML) of CdS) was injected. The solution was kept at 80 0 C for 5 minutes. The flask was then cooled to 50°C and 0.82mL of 0.1M Cd(PDTC)2 solution (1 ML of CdS) was injected, and the solution was heated at 80°C for 5 minutes. The flask was then connected to vacuum at 80°C to remove the H2S formed during the shell growth. When the pressure dropped below 1 torr, the flask was returned to Argon and cooled to room temperature.
The HgSe/CdS QDs showed more colloidal stability than the HgSe cores, and did not require the addition of DDAB during precipitations. The HgSe/CdS QDs were purified in air by precipitation using methyl acetate. Some oily residue was observed after synthesis which is difficult to remove using precipitation / dissolution. 2 mL TCE was added to the black precipitate, and the solution was stored as a stock solution in a vial in the freezer. TEM imaging showed that the QD diameter increased from 4.8 nm to 6.1 nm after the shell growth.
No homogenous nucleation was observed at the above reaction conditions, though homogenous nucleation was seen when the Cd(PDTC)2 precursor volume was doubled.
The concentration of the QDs was determined using the absorbance at 700nm using the cross-section of HgSe, under the assumption that the shell growth leads to negligible change in absorbance at this wavelength.

1F: Testing the thermal stability of thin shell HgSe/CdS QDs
Prior to the thick CdS shell synthesis, the thin shell HgSe/CdS QDs were annealed in solution to test for thermal stability. Thin shell HgSe/CdS QDs (containing 3 mg of HgSe cores) was purified by precipitation with methyl acetate. 1.5 mL of saturated amine solution was used to dissolve the QDs, and transferred to a 3-neck flask. 8.2mg Cd(oleate)2 (1.5 surface equivalents) was then added. Three vacuum-Ar cycles were performed, and the solution was heated and kept at 220°C for a desired duration. The solution was cooled to room temperature and precipitated with methyl acetate. The pellet was dispersed in 0.9mL of TCE and 0.1mL oleylamine. The solution was heated to boil (~110°C -120°C) for a few seconds to dissolve the saturated amines and passivate the QDs with oleylamine. After cooling to room temperature, the solution was purified twice by precipitation-dispersion using methyl acetate and TCE. The pellet after the second precipitation was dried under vacuum before dispersing in TCE for FTIR measurements. Apart from an increase in the n-doping, no observable change was seen in the thin HgSe/CdS QDs after annealing at 220°C (Fig. S1F). HgSe/CdS QDs before and after annealing at 220°C. Due to the low-temperature synthesis of the thin shell, the interband PLQY is relatively low at 0.6% (black), but rises to ~2-3% on annealing (red and blue). The intraband PL is too weak to be resolved at this scale.

1G: Thick shell HgSe/CdS QD synthesis
A thick CdS shell was overgrown on the thin shell HgSe/CdS QDs using Cd(DEDTC)2 as a single-source precursor, with amines and Cd(oleate)2 as ligands. The Cd(oleate)2 helped prevent the growth of wurtzite arms during shell growth, 8,9 The concentration of Cd(oleate)2 is crucial, as the sulfur-containing side products from decomposition of Cd(DEDTC)2 reacted with excess Cd(oleate)2 and led to undesired homogenous nucleation of CdS. We found that a 1:5 molar ratio of Cd(oleate)2 : Cd(DEDTC)2 provided a balance to prevent growth of wurtzite arms, while avoiding homogenous nucleation. Cadmium dodecanethiolate (Cd(DDT)2) was observed to give similar results as cadmium oleate when used as the ligand.
The synthesis was done in the absence of unsaturated solvents, as they led to the formation of a thick oil /gel after synthesis which was difficult to separate from the QDs. We hence employed a 2:2:1 mixture of hexadecane-dodecylamine-hexadecylamine as the solvent, where the mixture aided in entropic dispersion of the QDs. 10 Thin shell HgSe/CdS QDs (containing 8mg of HgSe cores) was precipitated with methyl acetate to remove leftover oleylamine. The pellet was dispersed in 4mL of warmed saturated amine solution, and 1.35mL of 0.024M Cd(oleate)2 solution (1.5 surface equivalents) was added. The mixture was added to a three-neck flask equipped with rubber sleeves and a thermocouple, and connected to a Schlenk line. The 0.02M Cd(DEDTC)2 precursor solution was kept in a plastic syringe connected to a syringe pump. Three vacuum (p ~ 1 torr) -Argon cycles were performed, and the solution was heated rapidly to 220°C (average heating rate is around 50°C/min). When the temperature reached 215°C, the precursor solution was injected at rate 0.396 mL/min for 10 mins (equivalent to 3MLs of CdS). The temperature was maintained at 220°C and was not let to exceed 225°C. The flask developed a yellow tinge after a few minutes, indicating the growth of CdS. Depending on the reaction scale and heating rate, a reddish tinge might be observed, indicating dissolution of a fraction of the QDs and deposition of a HgCdSSe shell on the larger QDs.
After 10 mins, the injection was stopped and the syringe was removed. The solution was cooled to 120°C and evacuated to remove the H2S formed during the reaction. Removal of the H2S was necessary to prevent CdS homogenous nucleation during post-synthetic ndoping or during further CdS shell growth.
For a further growth of 3MLs of CdS, the flask was heated to 220°C and 0.02M CdS precursor injection was resumed at 0.650 mL/min for 10 mins (equivalent to 3MLs of CdS). The solution becomes cloudy with time as the dots grow larger. After the injection is complete, the flask was then cooled to 120°C and evacuated. The solution was then cooled to room temperature.
All purifications were done in air. On centrifuging the cooled reaction mixture, the nanocrystals typically precipitated as a thick bulky precipitate. Remaining QDs in the solution were precipitated using methyl acetate and centrifuged. After decanting the supernatant, the precipitate was dispersed in ~4mL TCE and 0.4mL OAm was added. The solution was brought to a boil (~110°C -120°C) for a few seconds to dissolve the saturated amine solvent, and let to cool. The solution was then purified twice by precipitation with methyl acetate and dispersion in TCE. After the second precipitation, the pellet was dried under vacuum, and then dispersed in TCE for measurements.
The shape of the thick shell HgSe/CdS QDs was sensitive to the reaction scale, where a 4 mg reaction scale led to more compact QDs than an 8 mg scale. interband absorptions of HgSe confirms that the integrity of the core is preserved after the thick shell growth. The data is obtained by stitching spectra using an FTIR spectrometer (<5000 cm -1 ) and a dispersive NIR spectrophotometer (>5000 cm -1 ).

1H: Doping control in HgSe/CdS QDs
The as-synthesized HgSe/CdS QDs typically showed a partial n-doping, as seen by the prominent 1Sh -1Se interband absorption and small 1Se -1Pe intraband absorption. To achieve doping control, we first completely undoped the QDs by performing a cALD with (NH4)2S, and then treated with cadmium acetate to achieve the desired n-doping level. 11 Reaction of cadmium acetate with thick shell HgSe/CdS QDs at room temperature led to incomplete n-doping, and it was necessary to perform the surface treatment at an elevated temperature to ensure a complete reaction.
The undoping was performed adapting a reported procedure. 7 At ambient conditions, purified 11nm HgSe/CdS QDs (core diameter 4.8nm, HgSe mass = 6mg) was added to a scintillation vial with 0.3 mL oleylamine and 2.7 mL octane. Separately, 0.16 mL of 40% (NH4)2S solution (10x surface equivalents to HgSe/CdS) was mixed with 3mL formamide (FA), and added to the HgSe/CdS solution to form a biphasic mixture. The vial was heated at 40°C with stirring for 5 minutes. The FA layer was removed, and washed with 3 mL of FA to remove excess (NH4)2S. The washing was repeated another time, after which the QDs were purified twice by precipitation-dispersion using methyl acetate and TCE. After treatment with (NH4)2S, the intraband absorption was completely bleached (Fig. S1H).
To n-dope the QDs, the undoped HgSe/CdS QDs (containing 6 mg of HgSe cores) were added to a three-neck flask with 0.6 mL of oleylamine and 5.4 mL hexadecane. After adding a calculated volume of 0.05M cadmium acetate solution (see Fig. 2D of main text. 1.0 mL of cadmium acetate corresponded to 1x surface saturation), the flask was connected to a Schlenk line and three vacuum-Argon cycles were performed. The flask was then heated to 180°C and kept at this temperature for 10 mins.
The reaction mixture was cooled to room temperature, and purified twice by precipitationdispersion using methyl acetate and TCE. After the second precipitation, the pellet was dried under vacuum before dispersing in TCE for measurements.

Synthesis at 160°C with Cd(oleate)2:
Reaction solvent was a 10% solution of OAm in ODE.
Thin shell HgSe/CdS QDs (6 nm diameter, synthesized by cALD) containing 3.3 mg of HgSe cores were purified and added to a 3-neck flask with 2.5mL of 10% OAm-ODE, and 4mg of Cd(DEDTC)2 (corresponding to 1ML of CdS) was added. The flask was evacuated at room temperature in a Schlenk line for 10 minutes, and then filled with Argon. The flask was then heated to 140°C and kept at this temperature for 1 hour. The flask was then heated to 160°C and the CdS precursor was injected using a syringe pump at 1.31 mL/hour. Aliquots were taken and purified twice by precipitation with IPA and dissolution in TCE for TEM imaging.

Synthesis at 220°C without Cd(oleate)2:
Reaction solvent was the saturated amine mixture: a 2:2:1 mixture of hexadecanedodecylamine-hexadecylamine by mass. Thin shell HgSe/CdS QDs (6 nm diameter, synthesized by using Cd(PDTC)2) containing 4 mg of HgSe cores were purified and added to a 3-neck flask with 2 mL of saturated amine mixture, and 0.14 mL of cadmium oleate solution was added. Three vacuum -argon evacuation cycles were performed at room temperature, with evacuation till 1 torr. The flask was then heated to 220°C. When the temperature reached 220°C, the CdS precursor solution was injected at a rate 0.26 mL/min (3ML CdS in 10 mins). A 2.4 mL aliquot was taken at 10 mins, 0.05 mL of cadmium oleate was added and the CdS precursor was injected at 0.25 mL/min for 13 mins. The aliquots were purified twice by precipitation with IPA and dissolution in TCE for TEM imaging. The shell grew as uniform spheres at 10 mins of injection ( Fig. S1I(E)), with some faceting visible at 23 mins (Fig. S1I(F)). The addition of cadmium oleate leads to a reduced (103) wurtzite feature in the SAED during transmission electron microscopy ( Fig. S1I(K)).

Synthesis at 220°C with excess Cd(oleate)2:
Reaction solvent was the saturated amine mixture: a 2:2:1 mixture of hexadecanedodecylamine-hexadecylamine by mass. Thin shell HgSe/CdS QDs (6 nm diameter, synthesized by using Cd(PDTC)2) containing 4 mg of HgSe cores were purified and added to a 3-neck flask with 2 mL of saturated amine mixture, and 0.54 mL of cadmium oleate solution was added. Three vacuum -argon evacuation cycles were performed at room temperature, with evacuation till 1 torr. The flask was then heated to 220°C. When the temperature reached 220°C, the CdS precursor solution was injected at a rate 0.27 mL/min (3ML CdS in 10 mins). After 10 mins, the solution was cooled to room temperature, divided into two parts and precipitated with IPA. One part was purified twice by precipitation -dispersion using IPA and TCE for TEM imaging. The other part was precipitated once with IPA, dispersed in 2.6 mL saturated amine mixture, and moved to the 3-neck flask with 0.55 mL of 0.1M cadmium oleate. After three vacuum -Argon cycles, the flask was heated to 220°C and the CdS precursor solution was injected at 0.139 mL/min for 10 mins. The solution was purified twice and imaged by TEM.

1J: Kinetics of decomposition of Cd(PDTC)2
To determine to temperature for growth of a thin CdS shell on HgSe using Cd(PDTC)2, we performed a kinetics study on the decomposition of Cd(PDTC)2. A solution of 0.1 M Cd(PDTC)2 in 20% OAm-ODE was prepared.
Thin HgSe/CdS QDs (synthesized using Cd(PDTC)2) containing 4 mg of HgSe was added to a 3-neck flask with 2 mL of 20% OAm-ODE solution. Three vacuum-argon cycles were performed, and the solution was heated to 60°C. 0.33mL of Cd(PDTC)2 solution was injected.
Aliquots were taken immediately (0 mins), 2 mins, 5 mins, 10 mins and 30 mins after injection. The aliquots were stored in a freezer to prevent further reaction.
Absorption spectra of the aliquots were recorded in a quartz cuvette by dilution in TCE, and normalized to absorbance at 1000 nm. The kinetics study was also performed for reaction at 80°C. As shown in Fig. S1J, the reaction is nearly complete in less than 2 mins at 80°C but took 10 mins at 60°C. The reaction temperature for growth of thin shell HgSe/CdS QDs using Cd(PDTC)2 was thus set at 80°C.

2A: Infrared absorption and photoluminescence measurements
All measurements were done on QDs dispersed in TCE in a cell with CaF2 windows and a path length of 0.5mm. Absorption measurements were performed on a ThermoNicolet Nexus 670 FTIR spectrometer, a Nicolet Magna 550 FTIR spectrometer and an Agilent Cary 5000 UV-vis-NIR spectrometer.
Photoluminescence spectra were recorded using a step-scan FTIR spectrometer with an MCT detector and a gated integrator. The samples were excited with a 150mW 808 nm diode laser, modulated at 90 kHz. A Si wafer was placed in front of the detector to block the excitation light. The transmittance of the solution at 808nm was measured using a Si diode detector behind the sample cell. A 3.4µm long-pass filter was placed in front of the detector to block the 1.7 µm interband emission and preferentially measure the 5 µm intraband emission. The PL spectrum was corrected for detector response and atmospheric absorption following Melnychuk et. al. 6 The PL spectrum was normalized by the measured 808 nm absorbance, and was also corrected for the infrared absorption by TCE.

2C: Measurement of PLQY and 1Se occupancy
Absolute PL quantum efficiency measurements (PLQE) were done following previous reports 6 using a Spectralon integrating sphere for a reference partially doped HgSe/CdS QD solution. The area under the PL spectrum of HgSe/CdS QDs (normalized to 808 nm absorption and corrected for detector response) was taken between 1600 -2800 cm -1 for intraband, and 4000 -7000 cm -1 for interband spectra. The ranges were taken as 1600 -3000 cm -1 and 5000 -8000 cm -1 for the HgSe cores. The PLQE per unit area was then calculated by dividing the measured PLQE by sum of areas under the intraband and interband PL spectra. The PLQE of other samples were calculated by recording the PL spectra and using the PLQE per area calculated above.
The average occupancy of the 1Se state (Ne) in the HgSe/CdS QDs was determined by undoping the QDs on treatment with I2. FTIR absorption spectra of the QDs in TCE were measured before and after injection of I2 in TCE (0.02M or 0.004M). Difference between subsequent spectra were used to calculate the 1Se -1Pe intraband absorption (see Ne was then calculated as twice the ratio of intraband peak absorbance to the maximum intraband absorbance (Y-intercept in Fig. 2(C) in main text). The 1Se(0), 1Se(1) and 1Se (2) populations were calculated from Ne using a binomial distribution (SI Section 3D). The fraction of doped QDs was taken as sum of 1Se(1) and 1Se(2) populations, and undoped fraction was 1Se(0). The intraband and interband PLQYs were calculated by normalizing the measured PLQEs to the fraction of doped and undoped QDs respectively.  The interband shape of the Ne = 0.07 sample deviates significantly due to scattering artefacts during FTIR measurements.

2D: Particle Size Characterizations
Transmission Electron Microscopy (TEM) images were recorded using an FEI Spirit 120kV electron microscope and an FEI Tecnai F30 300kV microscope. pXRD spectra were recorded on QDs films on Si using a Bruker D8 Powder X-Ray Diffractometer.

3A: Dissolution of thin shell HgSe/CdS QDs and deposition of HgCdSSe shell
During the initial stages of the thick shell HgSe/CdS synthesis, a fraction of the thin shell HgSe/CdS QDs dissolve and deposit on the remaining QDs as a HgCdSSe shell. This leads to a red tail in the visible spectrum beyond the CdS band edge, and larger core/shell sizes than calculated from the precursors added. A batch-to-batch variability is observed in the extent of dissolution, and it also depends on the heating rate. Two synthetic trials are described below that show a substantial QD dissolution (synthesis 1) and a negligible dissolution As shown in Fig. S3A(A, B), the core/shell sizes in synthesis 1 match well with the diameter expected from the Cd(DEDTC)2 precursor added, but the synthesis 2 leads to much larger core/shell diameters. From the size discrepancy in synthesis 2, the fraction of dissolved QDs can be calculated:

Calculation of fraction of dissolved QDs
Denote the initial diameter as di , the expected final diameter as dexp , measured final diameter as dm , and initial number of nanocrystals as N. The initial volume of QDs is Despite the significant dissolution of QDs in synthesis 2, the FTIR spectra of HgSe/CdS QDs from synthesis 2 is similar to those from synthesis 1 (Fig. S3A(D)). This is expected, since 220°C is much higher than the thermal stability window of HgSe, and the ripening dots dissolve nearly instantaneously. Hence none of the cores of the dissolving QDs survive, and the infrared properties of the HgSe/CdS QDs remain unaffected by the dissolution occurring during the synthesis.

3B: Radiative and nonradiative lifetimes of HgSe and HgSe/CdS QDs
The radiative lifetime W of a nanoparticle dipole transition is 12-15 Here is the angular frequency, c is the vacuum light speed, ℏ is the reduced Planck's constant, p is the transition dipole matrix element and ] is the vacuum permittivity. The   Growth of the CdS shell leads to a change in the dielectric screening and a decrease in the radiative lifetime. To calculate for this system, we performed an electrostatic analysis of a core/shell structure modeled as nested dielectric spheres in a dielectric matrix. 21,22 For a core/shell nanocrystal with core radius of R1 and total radius of R2, the radiative rate is found to be: where • is the shell optical dielectric constant. The radiative lifetime W as a function of core/shell diameter is plotted in Fig. S3B. The radiative lifetime approaches a value of 700 ns at diameters greater than 8 nm. Taking into account the uncertainties in frequency and ϵ1, we get the core/shell lifetime to be (700 ± 160). We can therefore use this radiative lifetime for all thick shell HgSe/CdS QDs. From this calculated lifetime and the measured intraband PLQY of 2.1 × 10 -2 for thick shell HgSe/CdS QDs, the nonradiative lifetime is 15 ns, with a one-standard deviation range of 7.7 -28 ns.

3C: Measurement of average diameter of bullet-shaped HgSe/CdS QDs
HgSe/CdS QDs develop an anisotropic bullet-shape for HgSe/CdS QDs at large shell thicknesses. To estimate the 'average' diameter, we assumed the nanocrystals to have the same volume as an ellipsoid with two diameters as the short length of the bullet (Fig. S3C(B)) and the third diameter as the long axis of the bullet (Fig. S3C(A)). The volume of the nanobullet was then approximated as   When the QD is photoexcited, the filled defect state could introduce a nonradiative decay channel either by (B) hole trapping or by (C) near-field energy transfer (FRET). Bulk CdS is known to absorb infrared light by transitions between defect states. 23,24 This absorption might be too weak to measure by FTIR, but the close proximity of the core to the shell can lead to a significant nonradiative decay.