Cd implantation in $\alpha \text{−}{\mathrm{MoO}}_3$: An atomic scale study

Lamellar $\alpha \text{−}{\mathrm{MoO}}_3$ crystals were implanted with low fluence of radioactive ${}^{111\mathrm{m}}\mathrm{Cd}$ ions at ISOLDE-CERN. Subsequently, we have probed the interaction of the Cd impurity in the lattice with native point defects, such as oxygen vacancies, as a function of annealing temperature using the time differential perturbed angular correlations nanoscopic technique. The experimental data were complemented and interpreted by modeling different Cd-defect configurations in $\alpha \text{−}{\mathrm{MoO}}_3$ with first-principles density functional theory (DFT). The agreement between experiments and DFT simulations shows that only the interstitial Cd (${\mathrm{Cd}}_I$) prevails in the van der Waals gap, by inducing a polaron effect. Upon raising the annealing temperature, ${\mathrm{Cd}}_I$ is able to trap hole charge carriers resultant from the oxygen vacancies $V_{\mathrm{O}}$. Oxygen vacancies were found to form most commonly at two-fold coordinated (O2) atoms. According to comparison DFT results with the experimental electric field gradient values ($V_{zz}$ and $\eta$) and the calculated formation energies for different defect complexes, the configuration of ${\mathrm{Cd}}_I$ with two (O2) vacancies ($V_{\mathrm{O}2}$), located at different planes, is found to be more favorable and stable than the other defect configurations. The electron-polaron formation around the Cd impurity at an interstitial site is enhanced by inducing (O2) vacancies with the creation of hole polaron states.

Figure 1

Artistic plot of the $\alpha \text{−}{\mathrm{MoO}}_3$ structure identifying the relevant planes, the nonequivalent oxygen atomic sites, and the interstitial positions of interest. The green, yellow, and red spheres represent the O3, O2, and O1, respectively.

Figure 2

Artistic representation of the TD-PAC experimental concept (left) and of the TD-PAC data observable (right).

Figure 3

Left: the $R\left(t\right)$ TD-PAC observable (black points) with the respective fit (red line) as a function of time. Right: the respective Fourier transforms. The blue lines in the Fourier spectra mark the triplets of observable frequencies characteristic of every EFG. The experimental points result from merging similar data obtained with two implanted samples that were annealed at $300{}^{\circ }\mathrm{C}$ and $320{}^{\circ }\mathrm{C}$, respectively. The measurement was performed at RT.

Figure 4

Left: the $R\left(t\right)$ TD-PAC observable (black points) with the corresponding fit (red line) as a function of time. Right: the respective Fourier transforms. The blue continuous lines at the Fourier spectra highlight the triplets of the observable frequencies characteristic of every EFG. The figure shows data of implanted samples that were subsequently annealed in air for 10 min at $450{}^{\circ }\mathrm{C}$. The measurements were then performed at $24{}^{\circ }\mathrm{C}, 100{}^{\circ }\mathrm{C}, 120{}^{\circ }\mathrm{C}, 200{}^{\circ }\mathrm{C}$, and $300{}^{\circ }\mathrm{C}$. Similar results were obtained at $100{}^{\circ }\mathrm{C}$ and $120{}^{\circ }\mathrm{C}$ and the data were merged on a single spectrum.

Figure 5

Structure of Cd-doped $\alpha \text{−}{\mathrm{MoO}}_3$ with a nominal concentration of 4.16% [Cd] in substitutional site (left) and interstitial site (right).

Figure 6

Plot of comparison data between the calculated hyperfine parameters (|$V_{zz}$|, $\eta$) and experimental results for the different configuration of Cd in $\alpha \text{−}{\mathrm{MoO}}_3$. The dashed lines and colored lines are corresponding to the calculated DFT results and the circle points are corresponding to the experimental results for different configurations. The calculated |$V_{zz}$| and $\eta$ that could be matched to the experiments are plotted in color.

Figure 7

Left: the calculated $V_{zz}$ (left axis, shown by the black circles), and $\eta$ (right axis represented by blue squares), for the ${}^{V\mathrm{O}2}\mathrm{Cd}_{IV\mathrm{O}2}$ configuration calculated by considering the experimental lattice constant at different temperature values, as presented in Ref. [60]. The calculations are done by employing $\mathrm{PBE}+U_{\text{Mo}}=6\mathrm{eV}$ and the dispersion corrections [DFT-D3(BJ)]. Right: the experimental temperature dependence of $V_{zz}$ and $\eta$ of the main EFG parameters (EFG4) (samples after annealing at $450{}^{\circ }\mathrm{C}$) for temperatures ranging from $24{}^{\circ }\mathrm{C}$ up to $300{}^{\circ }\mathrm{C}$.

Figure 8

Left: partial density of states of the pristine $\alpha \text{−}{\mathrm{MoO}}_3$ system. Right: corresponding isosurface of the charge densities projected around the Fermi energy (addition of partial charge densities of the VBM and CBM).

Figure 9

Left: partial density of states of interstitial Cd-doped $\alpha \text{−}{\mathrm{MoO}}_3$ system. Right: corresponding isosurface of the charge densities projected around the Fermi energy.

Figure 10

Left: partial density of states of interstitial Cd-doped $\alpha \text{−}{\mathrm{MoO}}_3$ system and with one O2 vacancy, ${\mathrm{Cd}}_{IV\mathrm{O}2}$. Right: corresponding isosurface of the charge densities projected around the Fermi energy.

Figure 11

Left: partial density of states of interstitial Cd-doped $\alpha \text{−}{\mathrm{MoO}}_3$ system and with two O2 vacancies, ${}^{V\mathrm{O}2}\mathrm{Cd}_{IV\mathrm{O}2}$. Right: corresponding isosurface of the charge densities projected around the Fermi energy.

Figure 12

Isosurfaces of the charge density differences of the Cd-doped $\alpha \text{−}{\mathrm{MoO}}_3$ structure where the left plot is the unit-cell representation, and the right plot is the contour map along the (0 0 1) axis and centered at the Cd site. The silver, purple, and red spheres represent the Mo, Cd, and O ions, respectively. The colors of the isosurfaces are attributed to the different charges: blue isosurfaces represent the negative charges and the yellow are the positive charges. The color of the contour plot is assigned with absolute values, where the light blue areas with circular contour lines refer to regions with high-density electrons and the dark blue regions with circular contour lines are of high hole density.

Figure 13

Isosurfaces of the charge density differences of the Cd-doped $\alpha \text{−}{\mathrm{MoO}}_3$ structure, and one O2 vacancy, where the left plot is the unit-cell representations, and the right plot is the contour map along the (0 0 1) axis and centered at the Cd site. The silver, purple, and red spheres represent the Mo, Cd, and O ions, respectively. The colors of the isosurfaces are attributed to the different charges: blue isosurfaces represent the negative charges and the yellow are the positive charges. The color of the contour plot is assigned with absolute values, where the light blue areas with circular contour lines refer to regions with high-density electrons and the dark blue regions with circular contour lines are of high hole density.

Figure 14

Isosurfaces of the charge density differences of the Cd-doped $\alpha \text{−}{\mathrm{MoO}}_3$ structure, and two O2 vacancies, where the left plot is the unit-cell representations, and the right plot is the contour map along the (0 0 1) axis and centered at the Cd site. The silver, purple, and red spheres represent the Mo, Cd, and O ions, respectively. The colors of the isosurfaces are attributed to the different charges: blue isosurfaces represent the negative charges and the yellow are the positive charges. The color of the contour plot is assigned with absolute values, where the light blue areas with circular contour lines refer to regions with high-density electrons and the dark blue regions with circular contour lines are of high hole density.