Vibronic effects in the $1.4\text{−}\mathrm{eV}$ optical center in diamond

We report optical absorption and luminescence measurements on the $1.4\text{−}\mathrm{eV}$ center in diamond. We show that the zero-phonon lines have a temperature-dependent Ni-isotope shift, that the isotopic shifts induced by carbon and nickel are opposite in sign, and that a local vibronic mode is present in the absorption spectrum but not in luminescence. The microscopic properties of the center are successfully analyzed with the Ludwig-Woodbury theory (LWT), revealing that the ${\mathrm{Ni}}^{+}$ ion in the $1.4\text{−}\mathrm{eV}$ center only weakly interacts with the diamond lattice. The importance of vibronic effects in the LWT analysis is experimentally demonstrated. It is believed that similar effects can account for the discrepancies previously encountered in modeling other $3d^9$ impurities in semiconductors.

Figure 1

Schematic diagram of a $3d^9$ ion in a cubic crystal undergoing different perturbations from left to the right: $T_d$ crystal field, trigonal field, spin-orbit coupling, and vibronic coupling. The latter is illustrated for a specific vibrational mode, coupling to which affects differently the involved electronic states. The vertical (energy) scale is severely distorted for presentation purposes.

Figure 2

Schematic drawing of a (110) cross section of a typical HPHT diamond crystals showing the main crystallographic sectors and indicating the sites from which samples A and B were cut.

Figure 3

Optical absorption, CL and PLE spectra from the $1.4\text{−}\mathrm{eV}$ center in the HPHT diamond sample B, all recorded at $77\mathrm{K}$. The CL spectrum is mirror reflected relatively to the $1.404\mathrm{eV}$ energy. Numbers indicate the prominent spectral features discussed in the text.

Figure 4

Optical absorption spectra measured at 12 and $7\mathrm{K}$ from sample B, and their difference multiplied by 4. Panel (b) shows the details of the $1.4\text{−}\mathrm{eV}$ doublet for the wider range spectrum (a). Note that the vertical scale of panel (b) is $\sim 10$ times coarser than that of panel (a). The narrow multiplet structure in panel (b) originates from the Ni isotopic structure (natural abundance). The curves for 7 and $12\mathrm{K}$ in panel (a) and right part of panel (b) are undistinguishable in this presentation.

Figure 5

Optical absorption spectra measured at indicated temperatures from samples C (natural carbon abundance) and D (100% ${}^{13}\mathrm{C}$). The multiplet structure originates from the Ni isotopic structure (natural abundance), and the vertical dashed line merely assist the observation of its temperature-induced shift. The spectra are offset for clarity. Spectra for samples C and D are labeled as ${}^{12}\mathrm{C}$ and ${}^{13}\mathrm{C}$, respectively.

Figure 6

Absorption spectrum from the sample B (1.1% ${}^{13}\mathrm{C}$) and PLE spectrum sample D (100% ${}^{13}\mathrm{C}$), both measured at $77\mathrm{K}$.

Figure 7

Absorption spectra measured at $77\mathrm{K}$ with linearly polarized light in different polarization arrangements. Curve A—sample B, [110] view, $\left[1\overline{1}1\right]$ polarization (along the growth direction); curve B—sample B, [110] view, $\left[\overline{1}12\right]$ polarization (perpendicular to the growth direction); curve C—sample A, $\left[1\overline{1}1\right]$ view, any polarization in the $\left(1\overline{1}1\right)$ plane (perpendicular to the growth direction). Panel (b) shows the details of the $1.4\text{−}\mathrm{eV}$ doublet for the wider range spectrum (a)