Metastability of ${\mathbf{Mn}}^{3+}$ in ZnO driven by strong $d$(Mn) intrashell Coulomb repulsion: Experiment and theory

Depopulation of the ${\mathrm{Mn}}^{2+}$ state in ZnO:Mn upon illumination, monitored by quenching of the ${\mathrm{Mn}}^{2+}$ EPR signal intensity, was observed at temperatures below 80 K. ${\mathrm{Mn}}^{2+}$ photoquenching is shown to result from the ${\mathrm{Mn}}^{2+}\to {\mathrm{Mn}}^{3+}$ ionization transition, promoting one electron to the conduction band. Temperature dependence of this process indicates the existence of an energy barrier for electron recapture of the order of 1 meV. $\mathrm{GGA}+U$ calculations show that after ionization of ${\mathrm{Mn}}^{2+}$ a moderate breathing lattice relaxation in the 3+ charge state occurs, which increases energies of $d$(Mn) levels. At its equilibrium atomic configuration, ${\mathrm{Mn}}^{3+}$ is metastable since the direct capture of photoelectron is not possible. The metastability is mainly driven by the strong intrashell Coulomb repulsion between $d$(Mn) electrons. Both the estimated barrier for electron capture and the photoionization energy are in good agreement with the experimental values.

Figure 1

(a) EPR spectrum of ${\mathrm{Mn}}^{2+}$ in ZnO with $B$ perpendicular to the $c$ axis. Black line shows the signals in the dark, green, and red traces are recorded under illumination with 532 nm (50 mW) and 850 nm (660 mW) laser lines, respectively. (b) EPR signal of (a) in the magnetic field range 4060–4160 G showing the shift of the line position and the change of the lineshape from Gaussian (in the dark) to Dysonian under illumination. (c),(d) EPR signal in the dark and under illumination measured in the thin ($100\mu \mathrm{m}$) sample. The illumination conditions are the same as in Figs. 1 and 1.

Figure 2

(a) Quenching of ${\mathrm{Mn}}^{2+}$ EPR intensity, $\mathrm{\Delta }{I}_{EPR}$, at $T=2.8$ K as a function of excitation wavelength measured in the thin sample. (b) Spectral dependence of the absorption coefficient for both pure ZnO and ZnO doped with 0.5% Mn at 300 K.

Figure 3

Shift of the ${\mathrm{Mn}}^{2+}g$ factor under excitation with 532 nm light (left axis, solid squares) and ${\mathrm{Mn}}^{2+}$ EPR signal photoquenching, $\mathrm{\Delta }{I}_{\mathit{EPR}}$, (right axis, circles) vs excitation power ($I$). The measurements were performed at 3.3 K with the magnetic field directed perpendicular to the ZnO $c$ axis. The error bar shows the difference between the power value measured before and after recording of the EPR spectra.

Figure 4

(a) Ratio $\mathrm{\Delta }{I}_{\mathit{EPR}}$(thick)/$\mathrm{\Delta }{I}_{\mathit{EPR}}$(thin) in thick and thin samples vs excitation wavelength. (b) Spectral dependence of absorbance at $T=16\mathrm{K}$. Note the different wavelength window in (a) and (b).

Figure 5

Temperature dependencies of ${\mathrm{Mn}}^{2+}$ EPR signal photoquenching ($\mathrm{\Delta }{I}_{\mathit{EPR}}$) in the as-grown thin sample (squares) and in the same sample after hydrogenation (triangles) illuminated with 532 nm light at 50 mW. The dashed lines are illustrations of an $(1-exp(-E/kT\left)\right)$ dependence with, $E=0.6$ and 1.25 meV for the as-grown and hydrogenated sample, respectively.

Figure 6

Dots: kinetics of ${\mathrm{Mn}}^{2+}$ photoquenching under 445 nm excitation at a power of 190 (upper trace) and 20 mW (lower trace) at 3 K. Red lines are calculated for an exponential rise and decay with the same rise/decay time of $\tau =63$ ms.

Figure 7

Energy bands (left panels) and DOS (right panels) of (a) ZnO:${\mathrm{Mn}}^{2+}$ and (b) ZnO:${\mathrm{Mn}}^{3+}$. Gray area and black lines (blue lines in the online version) in DOS display the total DOS and DOS projected on $d$(Mn) orbitals, respectively. Horizontal (red) lines denote the band gap of ZnO.

Figure 8

(a) Total energy change of ${\mathrm{Mn}}^{2+}$ and (${\mathrm{Mn}}^{3+},e_{CB}$) as a function of configuration coordinate $Q$. $Q_2$ and $Q_3$ are equilibrium atomic configurations of ${\mathrm{Mn}}^{2+}$ and (${\mathrm{Mn}}^{3+},e_{CB}$) charge states, respectively, and $Q_B$ is the configuration coordinate of the barrier. (b) Single particle energy of the $t_{2\uparrow }$ level for both ${\mathrm{Mn}}^{2+}$ (red symbols) and (${\mathrm{Mn}}^{3+},e_{CB}$) (blue symbols); note the strong dependence of the $t_{2\uparrow }$ energy on the charge state. $U\left(\text{Mn}\right)=0$ is assumed.

Figure 9

Calculated diagrams of Mn levels for (a) ${\mathrm{Mn}}^{2+}$ at equilibrium configuration $Q_2$, (b) ${\mathrm{Mn}}^{3+}$ with a photoelectron $e_{CB}$ in the CB in the same atomic configuration $Q_2$, (c) ${\mathrm{Mn}}^{3+}$ with $e_{CB}$ in the relaxed configuration $Q_3$, (d) ${\mathrm{Mn}}^{3+}$ with $e_{CB}$ in the barrier configuration $Q_B$, (e) ${\mathrm{Mn}}^{2+}$ (when $e_{CB}$ is captured by Mn) in the barrier configuration $Q_B$, and (f) ${\mathrm{Mn}}^{2+}$ relaxed to $Q_2$; this is the end step of the recombination process, the configuration is the same as in (a). $U\left(\text{Mn}\right)=0$ is assumed.

Figure 10

(a) Mn-O bonds. The atomic positions in cartesian coordinates are $\mathrm{Mn}(0,0,0),\mathrm{O}1(0,0,d_{1z}),\mathrm{O}2(0,d_{2y},d_{2z})$. (b) Two paths between $Q_2$ and $Q_3$. For $d_{2y}^B,{\mathrm{Mn}}^{2+}$ becomes unstable, i.e., the $t_{2\uparrow }$ level of ${\mathrm{Mn}}^{2+}$ is above CBM. $Q_B$ is the estimated barrier configuration.

Figure 11

Dependence of the $t_{2\uparrow }$ energy level on $U$(Mn) for ${\mathrm{Mn}}^{2+}$ and (${\mathrm{Mn}}^{3+},e_{CB}$) in the configurations $Q_2$ and $Q_3$.

Figure 12

The $Q$ dependence of (a) the total energy and (b) single particle levels of ${\mathrm{Mn}}^{2+}$ and (${\mathrm{Mn}}^{3+},e_{CB}$) for $U\left(\text{Mn}\right)=4$ eV.