Point defects in ZnO

 

Alexey A. Sokol, Samuel A. French, Stefan T. Bromley, C. Richard A. Catlow

Huub J. J. van Dam and Paul Sherwood

 

ZnO is widely used in catalysis, electrical devices, optoelectronics, and pharmaceuticals, which often crucially depend on the defect properties of this versatile material. The nature of the intrinsic defects in ZnO, however, remains elusive, and, so far, there is no unambiguous assignment of experimental data to particular defect species (with the possible exception of the positron annihilation spectroscopic evidence for Zn vacancies - see references [[1],[2]-[3]] and references therein). Theoretical work is therefore essential to complement the extensive body of experimental data.

We have investigated first the intrinsic point defects in ZnO and secondly, H, N, P, Li, Fe, Cu, Al and In impurity centres [[4]]. Atomic and electronic structures as well as defect formation energies have been obtained for the main oxidation states of all the defects using our embedded cluster, hybrid Quantum Mechanical / Molecular Mechanical approach to the treatment of localised states in ionic solids [[5],[6]]. In contrast to a large number of recent periodic density functional calculations, we are able to employ in these studies, significantly more accurate QM methods, for example, those based on hybrid exchange-correlation density functionals, which allows us to approach the limit of chemical accuracy in the energetics of defect formation, which are given in Table 1), where we also report energies calculated using the classical Mott-Littleton approach.

Table 1: Calculated energies for the defect pair formation in ZnO. Pure MM energies obtained using Mott-Littleton approach and interatomic potentials are shown for comparison for relevant charged defects.

The results show that oxygen Frenkel pairs have low energies of formation; moreover, we note that O interstitials have the lowest energies of formation among all intrinsic defects, which suggests their dominance under oxidising conditions. However, both Zn interstitials (2.2 eV) and O interstitials (1.7 eV) have similar, relatively low energies of formation. Hence, the dominant defect species should be decided upon by the sample history and working conditions.

Next, based on the calculated values of the vertical ionisation potential and electron affinity, we have determined defect levels in the band gap of the ZnO; the results are illustrated in the figure. With these calculations we have been able to explain the following experimentally observed phenomena:

·         We propose that the neutral and singly positively charged Zn interstitial defect is responsible for E1 and E3 (majority) donor bands from electric measurements.

·         Zn vacancies are proposed as the majority acceptor in agreement with experimental assignment based on positron annihilation spectroscopic and other studies. This defect is found to be stable in five charge states. Exciton recombination at this defect species is proposed as a source for main photoluminescence bands: ultraviolet (an acceptor level at 3.2 eV in a Donor to Acceptor Photoluminescence, DAP, transition); green (a triplet level at 2.5 eV), and red (at 1.9-2.0 eV).

·         The neutral O interstitial in a split-interstitial peroxy configuration (at 2.8 eV) should also contribute to blue and green luminescence by an exciton recombination and DAP transition from donor Zn interstitials.

·         O vacancies could not be a source of green luminescence, but could contribute to near-gap (UV) and red-orange luminescence bands (at 2.1 eV and below) via the exciton recombination mechanism.

·         Our calculations confirm that Cu, which is stable in ZnO in two charge states, is an efficient electron scavenger. The singly negatively charged Cu impurity is proposed as an E4 donor. Calculated defect levels (at 2.7 and 0.55 eV) are in good agreement with experiment, which established Cu as a distinct source of green luminescence from ZnO.

Calculated energy levels of intrinsic and extrinsic defects in ZnO (in eV)

Our calculations on impurities show their donor and acceptor properties in agreement with experiment where available. Curiously, our calculations suggest that a number of such extrinsic defects, for example, Li and N, can also contribute to the green luminescence band, the origins of which caused much controversy in the literature.

These studies are currently continuing on other extrinsic defects and defect complexes.



[1] Ü. Özgür, Ya. I. Alivov, C. Liu, A. Teke, M. A. Reshchikov, S. Doğan, V. Avrutin, S.-J. Cho, H. Markoç,  J. Appl. Phys. 2005, 98, 041301-1.

[2] S. J. Pearton, D. P. Norton, K. Ip, Y. W. Heo, T. Steiner, Progress Materials Science 2005, 50, 293.

[3] D. C. Look, B. Claflin, Phys. Stat. Sol. (b) 2004, 241, 624.

[4] A. A. Sokol, S. A. French, S. T. Bromley, C. R. A. Catlow, H. J. J. van Dam, P. Sherwood, Faraday Discussions 2007, 134, 267-282.

[5] A. A. Sokol, S. T. Bromley, S. A. French, C. R. A. Catlow, P. Sherwood, Int. J. Quant. Chem. 2004, 99, 695-712.

[6] P. Sherwood, A. H. de Vries, M. F. Guest, G. Schreckenbach, C. R. A. Catlow, S. A. French, A. A. Sokol, S. T. Bromley, W. Thiel, A. J. Turner, S. Billeter, F. Terstegen, S. Thiel, J. Kendrick, S. C. Rogers, J. Casci, M. Watson, F. King, E. Karlsen, M. Sjĝvoll, A. Fahmi, A. Schäfer, Ch. Lennartz, J. Mol. Struct. (Theochem.) 2003, 632, 1.