Phosphorous-implanted in ZnO thin films

So far, there have been several research groups proposed the formation of p-type ZnO by various dopants and doping method [103,106]. However, no detailed studies were made to investigate the effect of group V-dopants on the structural change and photoluminescence properties of ZnO films. Figure 4.1 shows the XRD patterns of

the phosphorous-implanted (fluence of 5×₁₀¹²_{, 1 × 10}¹⁴, and 5 × 10¹⁵ cm^-2) and non-implanted ZnO thin films annealed at 850˚C in nitrogen atmospheres. With an increase of phosphorus concentration, the (002)-peak intensity decreased obviously and a weak diffraction peak was observed at the fluence of 5× 10¹⁵ cm^-2 for the phosphorous-implanted ZnO films annealed at 850^oC in nitrogen. Similar behavior was also observed in the case of oxygen atmosphere. It was believed that the phenomenon is strongly dependent on the solubility limit of the implanted phosphorus in ZnO films. In addition, the (002) diffraction peak of the ZnO films was shifted towards the direction of smaller 2θ angle with the increase of fluence from 5× 10¹² to 5× 10¹⁵ cm^-2. According to Bragg Law, the shift toward smaller 2θ direction indicates an increase of the lattice constant that was considered due to the incorporation of the phosphorus into ZnO matrix to form antisite PZn or phosphide (PO4-) compound.

Figure 4.2 shows the depth profile of phosphorus-implanted ZnO films with various fluences. As it can be seen, the secondary ion counts of phosphorus abruptly increase near the side of Si substrate regions. It was believed due to the original implanted phosphorus and incomplete diffusion in annealing process. This concentration corresponds to the solubility [106] of phosphorus in ZnO film that is determined as 2.5× 10¹⁷, 1.5× 10¹⁸, and 8.5× 10¹⁹ ions/cm³ for the fluence of 5× 10¹², 1× 10¹⁴, and 5× 10¹⁵ ions/cm², respectively.

Figure 4.1: XRD patterns of ZnO films implanted with different phosphorus fluences and annealed at 850˚C in nitrogen.

Figure 4.2: SIMS depth profile of ZnO films implanted with different phosphorus fluences after annealed at 850˚C in nitrogen atmospheres.

Figure 4.3: SEM images of ZnO films doped with phosphorus at fluences of (a) 5×10¹² and (b) 5×10¹⁵ ions/cm² and then annealed at 850˚C in nitrogen atmosphere.

Figure 4.3 shows the SEM images of ZnO films doped with phosphorus and then annealed at 850˚C in nitrogen atmosphere. As shown in Figure 4.3(a), both ZnO films with non-implanted and implanted with 5 × 10¹² ions/cm² exhibit similar surface morphology (r.m.s: ~2.5 nm). Above that concentration, i.e., 5×10¹⁵ ions/cm², Figure 4.3(b) illustrates that several ridge regions are formed in ZnO films (r.m.s.:~5.4 nm).

It was postulated that the formation of the glass-like ridge structure may be related to the excess doping of phosphorus. In order to investigate the phosphorus doping effect on the crystalline of ZnO films, TEM analysis were performed. As shown in Figure 4.4 for ZnO films annealed at 850˚C in nitrogen atmosphere, it was found that the cross-sectional microstructure was clearly divided into two regions: crystalline (columnar shape) and interlayer (flat-belt) structure. The interlayer could be considered as a buffer layer to reduce the stress due to lattice mismatch between ZnO and Si. However, for ZnO films with 1× 10¹⁴ ions/cm² implanted, several small clusters were observed in the ZnO interlayer in Figure 4.4(b) as marked with arrows.

According to energy-dispersive spectrometry (EDS) measurement, those clusters were primarily composed of phosphorus, zinc and oxygen elements that may be related to the formation of glass-like ridge structure. Figure 4.5 shows the room-temperature PL spectra of non-implanted and phosphorus-implanted ZnO films annealed at 850˚C in nitrogen atmosphere The PL behavior for deep-level emission of ZnO films implanted with various fluences of phosphorus is also illustrated in the inset of Figure 4.5 for comparison. The inset is the deep-level emission of ZnO films implanted with various fluences. It was observed that the peak intensity of the UV emission varies with the concentration of fluence. A very stronger UV peak (378 nm) and a relatively low deep-level emission (545 nm) were obtained for the non-implanted sample. However, as the ZnO films were implanted with different phosphorus fluences and annealed in nitrogen atmosphere, the UV emission peak of the ZnO films becomes weaker and

presents slightly red shift as compared to the non-implanted one. This PL result along with the XRD analysis and surface morphology implies that there should be a solubility limit for phosphorus incorporated into ZnO films. If the implanted concentration is close to the solubility limit, both crystal structure and NBE emission would be strongly influenced and become poor.

Figure 4.4: Cross-sectional TEM images of ZnO films annealed at 850˚C in nitrogen atmospheres. (a) without phosphorus-implanted, (b) with phosphorus-implanted (fluence: 1× 10¹⁴ ions/cm²).

Figure 4.5: Dependence of fluences conditions on room temperature PL spectra for the phosphorus-implanted ZnO films annealed at 850˚C in nitrogen atmospheres. The inset is the deep-level emission of ZnO films implanted with various fluences of phosphorus.

On the other hand, as the phosphorus-implanted ZnO films were annealed in O2

atmosphere (not shown here), the UV peak intensity was remarkably decreased compared to that annealed in N2 atmosphere. In addition, with increasing phosphorus-implanted concentration up to 1× 10¹⁴ fluence, a weaker UV peak accompanied with a stronger deep-level emission around 545 nm was detected in O2

atmosphere than that in N2 atmosphere. It implies that more defects were probably induced in ZnO film annealed in oxygen atmosphere than that annealed in nitrogen atmosphere. Therefore, according to above discussion, it was believed that the property deterioration in the phosphorus-implanted ZnO films is correlated closely with the formation of defects and some glassy phase (MPO4-) as evidenced from Figure 4.4(b) of TEM. In addition, the resistivity and carrier type of non-implanted and phosphorus-implanted ZnO films were further investigated by Hall measurement.

The non-implanted ZnO film has the 52 Ωcm and exhibits n-type (-1.65×10¹⁶ cm^-3) characteristics. However, with phosphorus-implanted fluence of 5×10¹² ions/cm^-2, the resistivity of phosphorus-doping ZnO films increases up to 256Ωcm but the carrier concentration approaches to 1.65×10¹⁵ cm^-3. Thus, the conversion of carrier type was never observed that is probably attributed to the formation of phosphide even though the phosphorus element was successfully incorporated into ZnO films.