We study the influence of nitrogen, a potential acceptor in ZnO, on the lattice dynamics of ZnO. A series of samples grown by chemical vapor deposition ͑CVD͒ containing different nitrogen concentrations, as determined by secondary ion mass spectroscopy ͑SIMS͒, was investigated. The Raman spectra revealed vibrational modes at 275, 510, 582, 643, and 856 cm Ϫ1 in addition to the host phonons of ZnO. The intensity of these additional modes correlates linearly with the nitrogen concentration and can be used as a quantitative measure of nitrogen in ZnO. These modes are interpreted as local vibrational modes. Furthermore, SIMS showed a correlation between the concentration of incorporated nitrogen and unintentional hydrogen, similar to the incorporation of the p-dopant magnesium and hydrogen in GaN during metalorganic CVD.There is increasing interest in investigating the properties of ZnO epitaxial films with a direct gap of 3.37 eV at room temperature. 1 The material is a potential competitor for GaN-based light-emitting devices in the ultraviolet and blue spectral range. There are reports of superior ZnO properties such as a high exciton binding energy combined with a low lasing threshold density 2 and a good resistance to bombardment with high-energy particles. 3,4 For other wide-band-gap semiconductors as GaN ͑Ref. 5͒ and ZnSe ͑Ref. 6͒ controlled p-type doping is problematic. As-grown ZnO typically has n-type conductivity with background concentrations between 10 16 and 10 17 cm Ϫ3 . However, there have been reports on the synthesis of p-conducting ZnO doped with As ͑Ref. 7͒ and a Ga/N codoping 8 as well as the fabrication of a p-n-junction by excimer-laser doping. 9 In this letter, we report on doping experiments with nitrogen as a potential acceptor and its influence on the lattice dynamics of ZnO.The ZnO thin films under investigation were grown by chemical vapor deposition ͑CVD͒ using a home built epitaxy system which consists of a horizontal quartz reactor and a resistance heating with different temperature zones. Metallic zinc was kept in one zone at a temperature of 470°C the growth temperature was 650°C. We used NO 2 as oxygen precursor and NH 3 as nitrogen source for the doping experiments. The epitaxial films were deposited on GaN/sapphire templates which offers the advantage of a lattice parameter similar to ZnO. We investigated samples containing different nitrogen concentrations. Secondary ion mass spectroscopy ͑SIMS͒ was applied to determine the concentration of nitrogen and unintentional dopants such as hydrogen. The primary ion species was cesium. Nitrogen was detected as 14 N 16 O Ϫ and hydrogen as 64 Zn 1 H Ϫ clusters. The given abso-lute concentrations are accurate to within half an order of magnitude. Despite this accuracy the relative error is less than 10%. The Raman-scattering experiments were carried out in backscattering geometry with a triple-grating spectrometer equipped with a cooled charge-coupled device detector. The lines at 488 and 514.5 nm of an Ar ϩ /Kr ϩ mixedgas laser were used...

PACS: 61.72.Vv; 78.47.+p; 78.55.Et Similar to other wide-band gap semiconductors such as GaN [1] and ZnSe [2] p-type doping in ZnO has been a challenge for many years. As-grown ZnO typically shows n-type conductivity with background concentrations between 10 16 and 10 17 cm --3 . These residual donor concentrations have to be overcome by a suitable acceptor impurity. Suitable means absence of site competition, e.g. substitutional Li (acceptor) versus interstitial Li (donor) [3], a high enough solubility, small size mismatch and hence no lattice relaxation, and avoiding if possible the formation of compensating/passivating centres. Nitrogen on an oxygen site can be considered the best candidate as demonstrated by its success in p-type doping ZnSe [2] and ZnS [4]. In this communication we present the optical properties of the shallow nitrogen acceptor in ZnO and give an estimate of its binding energy.The epitaxial films were deposited in a home-built epitaxy system which consists of a horizontal quartz reactor, and a resistance heating with different temperature zones. Metallic Zn with 6N purity was kept in one zone at a temperature of 470 C. The growth temperature was between 600 and 650 C. As oxygen precursor we used NO 2 and for nitrogen doping ammonia. We grew on GaN-templates which were grown by MOCVD on (0001) sapphire substrates. The films were investigated by time-integrated and time-resolved photoluminescence (PL). The 325 nm line of a HeCd laser and a pulsed dye laser (pulse width l ¼ 292 nm) were used as excitation source.In order to get information on the energy level of the nitrogen acceptor in the band gap of ZnO we performed low temperature photoluminescence experiments. In Fig. 1 we compare undoped and N-doped ZnO layers. The undoped films show predominantly bound exciton emission (Fig. 1a), in the inset the energy range around 3.36 eV is shown on an expanded scale. The two recombinations occur at 3.362 and 3.361 eV (note that the line positions may depend on strain in the films). They are commonly attributed to neutral donor bound excitons having localisation energies between 14 and 15 meV. From the observation of two-electron satellites (TES) we could determine the donor binding energies to between 52 and 54 meV. Characteristic changes occured in the nitrogen doped film (see Fig. 1b). The excitonic recombinations can no longer be distinguished due to the increase in linewidth. The recombination at 3.33 eV was related by spatially resolved cathodoluminescence experiments to structural defects. At lower energies pronounced donor-acceptor pair transitions with an intensity comparable to the excitonic recombination are observed. Its zero phonon line (ZPL) peaks at 3.235 eV and it is repeated by longitudinal optical phonon replicas with an energy of 73 meV. The intensity of the ZPL with respect to 1LO, 2LO, . . . replicas is described by a Poisson distribution (weak-coupling regime)

In order to realize controlled p-type doping in ZnO the role of extrinsic and intrinsic donors has to be clarified. The extrinsic n-type dopants Al, Ga and In are commonly found in bulk ZnO crystals, but hydrogen also appears in relevant concentrations eventually controlling the residual n-type carrier concentrations in nominally undoped ZnO. The optical properties of excitonic recombinations in bulk, n-type ZnO are investigated by photoluminescence (PL). At liquid helium temperature the neutral donor-bound excitons dominate in the PL spectrum. Two electron satellite (TES) transitions of the donor-bound excitons allow us to determine the donor binding energies ranging from 46 to 73 meV. In the as-grown crystals a shallow donor with an activation energy of 30 meV controls the conductivity. Annealing annihilates this shallow donor which has a bound exciton recombination at 3.3628 eV. Correlated by magnetic resonance experiments we attribute this particular donor to hydrogen. These results are in line with the temperature-dependent Hall-effect measurements. The Al, Ga and In donor-bound exciton recombinations are identified based on doping and diffusion experiments, and using secondary ion mass spectroscopy. We report on the optical properties of the shallow nitrogen acceptor in ZnO incorporated by diffusion, by ion implantation and by in situ doping in epitaxial films.

We present results of magneto-optical measurements and theoretical analysis of shallow bound exciton complexes in bulk ZnO. Polarization and angular dependencies of magneto-photoluminescence spectra at 5 T suggest that the upper valence band has Γ7 symmetry. Nitrogen doping leads to the formation of an acceptor center that compensates shallow donors. This is confirmed by the observation of excitons bound to ionized donors in nitrogen doped ZnO. The strongest transition in the ZnO:N (I9 transition) is associated with a donor bound exciton. This conclusion is based on its thermalization behavior in temperature-dependent magneto-transmission measurements and is supported by comparison of the thermalization properties of the I9 and I4 emission lines in temperature-dependent magneto-photoluminescence investigations.

No abstract