The Chemistry of Cu3N and Cu3PdN Nanocrystals

Abstract The precursor conversion chemistry and surface chemistry of Cu3N and Cu3PdN nanocrystals are unknown or contested. Here, we first obtain phase‐pure, colloidally stable nanocubes. Second, we elucidate the pathway by which copper(II) nitrate and oleylamine form Cu3N. We find that oleylamine is both a reductant and a nitrogen source. Oleylamine is oxidized by nitrate to a primary aldimine, which reacts further with excess oleylamine to a secondary aldimine, eliminating ammonia. Ammonia reacts with CuI to form Cu3N. Third, we investigated the surface chemistry and find a mixed ligand shell of aliphatic amines and carboxylates (formed in situ). While the carboxylates appear tightly bound, the amines are easily desorbed from the surface. Finally, we show that doping with palladium decreases the band gap and the material becomes semi‐metallic. These results bring insight into the chemistry of metal nitrides and might help the development of other metal nitride nanocrystals.

Synthesis of Cu3N. The synthesis for Cu3N nanocrystals was adapted from the literature. [1] Copper nitrate (0.24 mmol, 60 mg, 1 eq) was dissolved in 7.5 mL of hexadecane in a three-neck flask and stirred. Distilled oleylamine (7.57 mmol, 2.5 mL, 31 eq) was added and the mixture was degassed for 30 minutes at 50 °C. Under argon, the temperature was increased to 260 °C at a rate of 10 °C per minute. The mixture was left to react for 15 minutes. The flask was then removed from the heating mantle and left to cool to room temperature over the course of an hour. The 10 mL solution of nanoparticles were transferred to centrifuge tubes. A total of 15 mL of acetone was added and the particles were centrifuged at 5 000 rpm for 3 minutes. The supernatant was removed and the particles were redispersed in 4 mL of cyclohexane. A 10% by volume stock solution (SS) of purified oleylamine in cyclohexane was prepared. Oleylamine (1 mL SS) was added to the redispersed particles, so the total redispersion volume was 5 mL.
The particles were placed in the ultrasonic bath until fully redispersed. The particles were washed once more with acetone (15 mL) and this time redispersed in 3 mL of cyclohexane. 2 mL of the oleylamine SS was added, so the volume was still 5 mL. The particles were placed in the ultrasonic bath. 15 mL of ethanol was added for a final wash and the particles were centrifuged at 8'000 rpm for 10 minutes. The supernatant was removed, and the particles were redispersed in 2 mL of cyclohexane.
Synthesis of Cu3PdN. The synthesis for Cu3PdN nanocrystals was adapted from the literature. [1] Copper nitrate (0.24 mmol, 60 mg, 1 eq) was dissolved in 7.5 mL of hexadecane in a three-neck flask and stirred. Palladium(II) 2,4-pentadionate (0.08 mmol, 25.2 mg, 0.33 eq) was added and stirred. Distilled oleylamine (7.57 mmol, 2.5 mL, 31 eq) was added and the mixture was degassed for 30 minutes at 50 °C. Under argon, the temperature was increased to 240°C at a rate of 10°C per minute. The mixture was left to react for 15 minutes. The flask was removed from the heating mantle and left to cool to room temperature over the course of an hour. The nanocrystals were purified as described previously. However, contrary to the Cu3N nanoparticles, the final centrifugation after the ethanol wash required 10 -15 minutes at 10,000 rpm to precipitate the nanoparticles. Synchrotron X-ray total scattering experiments. Samples were prepared in 1mm polyamide kapton tube. The samples were measured at beamline P21.1 at DESY in Hamburg, Germany.
X-ray total scattering data were collected at room temperature in rapid acquisition mode, using a Perkin Elmer digital X-ray flat panel amorphous silicon detector (2048 × 2048 pixels and 200 × 200 μm pixel size) with a sample-to-detector distance of 380 mm. The incident wavelength of the X-rays was λ = 0.1220 Å (101.62 keV). Calibration of the experimental setup was performed using a Ni standard.
Raw 2D data were corrected for geometrical effects and polarization, then azimuthally integrated to produce 1D scattering intensities versus the magnitude of the momentum transfer Q (where Q = 4π sin θ/λ for elastic scattering) using pyFAI and xpdtools. [2] The program xPDFsuite with PDFgetX3 was used to perform the background subtraction, further corrections, and normalization to obtain the reduced total scattering structure function F(Q), and Fourier transformation to obtain the pair distribution function (PDF), G(r). [3] For data reduction, the following parameters were used after proper background subtraction: Qmin = 0.8 Å -1 , Qmax = 19 Å -1 , Rpoly = 0.9 Å. Modeling and fitting were carried out using Diffpy-CMI or PDFgui. [4] XPS and DFT. X-ray photoelectron spectroscopy (XPS) measurements were performed on a Thermo Scientific K-Alpha spectrometer with a monochromated microfocused Al Kα X-ray source (hυ = 1486.7 eV) and a spot size of 400 µm. The X-ray source was operated at 6 mA emission current and 12 kV anode bias and a flood gun was used for charge compensation. A pass energy of 20 eV was used for all core level and valence spectra. The Thermo Scientific Avantage software package was used for all data analysis. All XPS core level spectra were normalized to the peak height of the main feature in the Cu 2p3/2 spectra.
The details of the density functional theory (DFT) calculations for Cu3N, including density of states, have been previously reported. [5] [6] [7] To enable the direct comparison of the theory with the experimental valence spectra, the calculated projected density of states (PDOS) was broadened with a Gaussian to match the experimental broadening of 600 meV. The PDOS was weighted by one-electron photoionisation cross sections for the respective orbitals of Cu and N using the Galore software package, based on Scofield cross section values. [8] NMR measurements. NMR measurements were recorded at 298K on Bruker UltraShield 500 MHz spectrometer or a 600 MHz Bruker Avance III spectrometer. Regular 1 H NMR measurement were acquired with a 30 degree pulse with a recycle delay of 1.5 sec. Quantitative 1 H NMR measurements were acquired with a 90 degree pulse, 64k data points, 20 ppm spectral width, and a recycle delay of 30 sec. DOSY measurements were performed with a double stimulated echo and bipolar gradient pulses (dstebpgp2s). The gradient strength was varied quadratically from 2-95% of the probe's maximum value in 16 or 64 steps. The gradient pulse duration and diffusion delay were optimized to ensure a final attenuation of the signal in the final increment of less than 10% relative to the first increment. The diffusion coefficients were obtained by fitting a modified Stejskal-Tanner equation to the signal intensity decay: IIe g 2 D  I are the signal intensities, D are the linear diffusion coefficients,  is the gyromagnetic ratio of the studied nucleus, g is the gradient strength,  is the pulsed field gradient duration and  is the diffusion delay. A correction factor of 0.6 is applied for  due to the smoothed squared pulse shape used for the gradient pulses. [9] Other instrumentation. TEM imaging was done using a JEOL JEM2800 field emission gun microscope operated at 200 kV equipped with a TVIPS XF416ES TEM camera. DLS measurements were conducted on a Malvern Zetasizer Ultra in backscattering mode (173°) in a glass cuvette. All measurements were performed at 25°C after equilibrating inside the system for 240 seconds, sample concentration was tuned to achieve system attenuator values between 9-10. UV-VIS spectra were recorded on a PerkinElmer Lambda 365. FTIR spectra were recorded on a Perkin Elmer Spectrum Two spectrometer (attenuated total reflection, ATR). Figure S1. Powder XRD of Cu3N was obtained after 15 min at the different reaction temperatures. Bulk

Supporting figures
Cu3N reference is presented in blue.      Since Cu3N is being fully decomposed to Cu 0 at the end of the thermal process, the nitrogen loss should also be taken into account.

+ → 3
We believe that the two first decomposition steps belong to the organics while the last decomposition starting at 356°C belongs to Cu3N decomposition (bulk Cu3N decomposes around 400 °C). When starting with Cu3N + organics, we recover, 65.9 % of the mass as we assume that at the end of the heating process the end product is Cu 0 and Pd 0 in a 3:1 ratio.   Table S1. (B) Cu3PdN nanocrystals with Cu3PdN (Pm-3m). The refined values are given in Table S2.