Observation of ordered organic capping ligands on semiconducting quantum dots via powder X-ray diffraction

Powder X-ray diffraction is one of the key techniques used to characterize the inorganic structure of colloidal nanocrystals. The comparatively low scattering factor of nuclei of the organic capping ligands and their propensity to be disordered has led investigators to typically consider them effectively invisible to this technique. In this report, we demonstrate that a commonly observed powder X-ray diffraction peak around \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$q=1.4{\AA}^{-1}$$\end{document}q=1.4Å−1 observed in many small, colloidal quantum dots can be assigned to well-ordered aliphatic ligands bound to and capping the nanocrystals. This conclusion differs from a variety of explanations ascribed by previous sources, the majority of which propose an excess of organic material. Additionally, we demonstrate that the observed ligand peak is a sensitive probe of ligand shell ordering. Changes as a function of ligand length, geometry, and temperature can all be readily observed by X-ray diffraction and manipulated to achieve desired outcomes for the final colloidal system.


Supplementary Discussion
Supplemental Figure 1. Powder X-ray diffraction spectra of synthesized indium myristate. a, Drop cast from solution. b, Recrystallized from acetone. Cu K-( = 1.5418 Å) is the X-ray source for the powder X-ray diffraction spectra.
One of the most common explanations for the ligand peak is unreacted organic precursor. 1,2 Based on the spectra of drop cast and recrystallized indium myristate around = 1.4 Å −1 , where is the scattering length vector (Supplemental Figure 1), it is apparent why many would assume that this precursor is the source of the previously unassigned peak.
Interestingly, the sharp peak at larger values present in the recrystallized indium myristate is not present in the drop cast indium myristate, indicating greater ordering and decreased spacing between ligands in the former, which would be somewhat expected in a sample that was recrystallized rather than drop cast. This same ordering and corresponding peak sharpness are also observed in the stearate and octadecylphosphonate capped indium phosphide particles in Figure 3. This is consistent with the hypothesis that the stearate and octadecylphosphonate ligands are in a more crystalline phase on the surface of the quantum dots. However, the 1 H NMR spectra (Supplemental Figure 2) of the solutions used to drop cast the quantum dots for measurements depicted in Figure 3a and d (stearate capped and oleylamine exchanged indium phosphide) indicates otherwise. The spectra show the methyl and vinyl peaks of the bound capping ligands. Unbound ligands correspond to sharp 1 H NMR peaks upshifted from the broad peak attributed to ligands bound to the nanocrystal. 3,4 These measurements indicate little unbound precursor after significant washing steps, corresponding to approximately 5 free stearates per nanocrystal for the sample capped with stearate. Considering the low scattering factors of the organic ligands, 5 unbound precursor cannot be the source of the peak in the powder X-ray diffraction patterns.  Figure 3. Characterization of small cadmium selenide quantum dots. a, TEM image of cadmium selenide quantum dots. b, Absorption and emission spectra of cadmium selenide quantum dots. c, Experimental powder X-ray diffraction pattern of cadmium selenide quantum dots with reference peaks. Cu K-( = 1.5418 Å) is the X-ray source for the powder X-ray diffraction spectrum.

Supplemental
The observation of the previously unassigned peak has been observed in other systems besides indium phosphide. Supplemental Figure 3 shows a powder X-ray diffraction pattern of cadmium sulfide quantum dots (4.3 nm diameter) capped with oleates where the ligand peak also appears (ICSD 1604283). and emission spectra of cadmium selenide quantum dots. c, Experimental powder X-ray diffraction pattern of cadmium selenide quantum dots with reference peaks and peak deconvolution. Inset shows region where the oleate ligand powder X-ray diffraction peak appears. Cu K-( = 1.5418 Å) is the X-ray source for the powder X-ray diffraction spectrum.

Supplemental
When particles become much larger, such as in the case of the cadmium selenide particles (approximately 10 nm diameter), the ligand peak becomes much less apparent when viewing the entire powder X-ray diffraction pattern. However, while diminished, the peak can still be deconvoluted with careful baseline subtraction, shown in Supplemental Figure 4  The absorbance spectra of the indium phosphide quantum dots synthesized with different length capping ligands have previously been reported, but these measurements (Supplemental Figure 5) show that the QDs are similarly sized and thus isolating ligand length as the experimental variable in Figure 3 is reasonable. 3 Supplemental Figure 6. Experimental powder X-ray diffraction background spectrum of silicon wafer. Cu K-( = 1.5418 Å) is the X-ray source for the powder X-ray diffraction spectrum.
The previously unassigned peak also cannot be attributed to any background signal as it does not appear in the background powder X-ray diffraction pattern (Supplemental Figure 6).

Note on powder X-ray diffraction Instruments
The measurements in Figures 4E and F were taken on different instruments to optimize different variables. The instrument used for Figure 4E's measurements provides better signal to noise ratio allowing us to clearly see nuances in the ligand peak shape at different temperatures.
Whereas, the diffractometer used for Figure 4F's measurements could accommodate a wider range of temperatures allowing us to observe ligand shell behavior at temperature extremes. to an argon glovebox. In centrifuge tubes, 60 mL of methyl acetate was added to precipitate the particles. The solution was centrifuged at 5702 RCF for 5 minutes, forming a pellet. The supernatant was discarded, and the particles were resuspended in 8 mL of hexanes. An additional 24 mL of methyl acetate was added, and the solution was centrifuged at 5702 RCF for 5 minutes.

Additional
After redispersing in 8 mL of hexanes, an additional 12 mL of methyl acetate was added followed by centrifuging at 5702 RCF for 5 minutes and the pellet was redispersed in 8 mL of hexanes.
Indium Myristate Synthesis: The synthesis of indium myristate used has previously been reported and characterized by FTIR, CHNS elemental analysis, and 1 H NMR. 4 For synthesis, 1.000 gram centrifuged again, the supernatant discarded, and washed again with 20 mL of acetone. This process was repeated three times before the solid was added to a glass vial with 20 mL of acetone. After capping, the solution was heated on a hot plate until all of the precipitate dissolved. The solution was allowed to slowly cool to room temperature, and the indium myristate precipitated from solution. The solution was centrifuged one last time, the supernatant was discarded, and the precipitate was dried under a vacuum line overnight to remove residual acetone.
1 H NMR: 1 H NMR of indium phosphide quantum dots synthesized without hydrogen capped with stearate, palmitate, and myristate has previously been reported in tetrahydrofuran-d8. 3 In an argon glovebox, samples were evaporated from the stock solution and redispersed in toluene-d8.
Internal mesitylene standard was 1% by volume. Measurements were performed on a Bruker Avance 700 instrument in oven-dried NMR tubes. Chemical shifts were referenced to the residual toluene signal. For quantitative measurements, the 90 pulse was calibrated, and samples were allowed to relax for 20 seconds between pulses. Uncertainty in the integration of NMR peaks was estimated at 5%.
Transmission Electron Microscopy: TEM imaging was performed on a FEI Tecnai T20 S-TWIN TEM operating at 200 kV with a LaB6 filament using holey carbon TEM grids.
Additional Optical Characterization: Luminescence measurements were taken on a Horiba Jobin Yvon TRIAX 320 Fluorolog. Excitation wavelength was 437 nm with excitation and emission slit widths of 2.5 nm. The resolution was 1.0 nm, and each wavelength was integrated for 2 seconds. Table 1. List of sample references where this peak is observed in powder Xray diffraction along with the explanation given (if any) and the material system.