X-ray techniques investigate copper doping processes in zinc-oxide nanowires

The development of wurtzite ZnO nanowires as biocompatible and sustainable semiconductors at the nanoscale has attracted increasing interest in electronics and energy applications [1]. A widely explored piezoelectric device involves vertically integrating ZnO nanowires between two electrodes to harvest ambient mechanical energy as a micro-source of power. However, the screening of the piezoelectric potential generated under mechanical stress remains a significant challenge, hindering performance.

This screening effect results from the high density of free electrons in ZnO nanowires, primarily due to the incorporation of hydrogen-related defects during their growth via chemical bath deposition [2]. Efforts to mitigate this issue have focused on reducing hydrogen-related defects through post-deposition treatments [3], but their impact on piezoelectric performance has been limited.

A more promising approach involves introducing Cu dopants as acceptors into ZnO nanowires, employing compensatory doping to reduce the free electron density. The effectiveness of this strategy depends critically on the local positions and environments of the Cu dopants, which were previously not fully understood.   

This study examined the incorporation of Cu into ZnO nanowires by combining X-ray absorption near-edge structure (XANES) spectroscopy with mass spectrometry and optical spectroscopy. ZnO nanowires were grown by chemical bath deposition with varying Cu dopant concentrations and subsequently subjected to thermal annealing at 500°C for one hour in an O2 atmosphere. The samples were analyzed using X-ray linear dichroism (XLD) at beamline ID12 at the Zn and Cu K-edges to probe the local positions and environments of the Cu dopants in the ZnO nanowires.

The experimental Zn K-edge XANES spectra for Cu-doped ZnO nanowires with a [Cu]/[Zn]_bath ratio of 5% are presented in Figures 1a and 1b. The spectra were recorded with the polar c-axis oriented perpendicular (E⊥c) and parallel (E//c) to the electric field of the linearly polarized X-ray beam, with their difference represented as the XLD spectrum. The XANES and XLD spectral shapes confirm that Zn atoms maintain a 2+ oxidation state and adopt the characteristic wurtzite configuration of ZnO nanowires, where Zn²⁺ cations are tetrahedrally coordinated by O^2- anions.

Click figure to enlarge

Fig. 1: a) Zn K-edge XANES and (b) corresponding XLD spectra of Cu-doped ZnO nanowires grown by chemical bath deposition (CBD) with a [Cu]/[Zn]_bath ratio of 5%. c) Normalized Cu K-edge XANES and (d) corresponding XLD spectra of ZnO nanowires grown by CBD with [Cu]/[Zn]_bath ratios of 1%, 3%, 5% and 10%. e) Normalized Cu K-edge XANES and (f) corresponding XLD spectra of ZnO nanowires grown by CBD with a [Cu]/[Zn]_bath ratio of 5%, recorded before and after thermal annealing at 500°C for one hour under an O₂ atmosphere.

The normalized isotropic Cu K-edge XANES and corresponding XLD spectra of ZnO nanowires grown at different [Cu]/[Zn]_bath ratios are shown in Figures 1c and 1d. The isotropic XANES spectra maintain a consistent shape across all [Cu]/[Zn]_bath ratios, featuring a leading-edge peak followed by a prominent feature approximately 11 eV higher. No pre-edge peak is observed above the noise level. The XLD spectra at the Cu K-edge closely resemble those at the Zn K-edge, exhibiting minimal dependence on the [Cu]/[Zn]_bath ratio. This similarity indicates that only a minority of Cu atoms adopt a wurtzite configuration by substituting Zn sites as Cu_Zn dopants. By comparing the XLD intensities at the Cu and Zn K-edges, it was determined that only 15 ± 5% of the total Cu is incorporated as Cu_Zn, while the majority (~85%) exists in a Cu₂O-like environment, primarily located on the nanowire surface.

Following thermal annealing of the Cu-doped ZnO nanowires, notable changes were observed in the isotropic XANES and XLD spectra at the Cu K-edge, as shown in Figures 1e and 1f. The isotropic XANES spectra show the disappearance of the sharp peak at the leading edge, accompanied by broadening of the subsequent prominent peak. In the XLD spectra, a moderate peak emerges around 9890 eV, while the most intense peak at ~9000 eV shifts slightly (~2 eV) to higher energies. These spectral modifications indicate a change in the electronic structure of the Cu dopants, suggesting a more oxidizing local environment and the formation of V_Zn defects near Cu atoms.

The combination of XANES and XLD with mass spectrometry, optical spectroscopy, and electrical measurements facilitated the estimation of CuZn dopant concentrations ranging from 1.3 to 12.7 x 10¹⁷ at/cm³. This approach also enabled the correlation of the zero-phonon line emission, strongly coupled with phonons at 2.86 eV, to Cu_Zn-V_Zn complexes. Additionally, the electrical activity of the Cu_Zn-V_Zn defect complexes was identified, as illustrated in Figure 2.

Click figure to enlarge

Fig. 2: Schematic representation of the adsorption and incorporation processes of Cu atoms in ZnO nanowires grown by CBD in a high-pH environment. The figure also illustrates the migration of Cu_Zn acceptors, facilitated by the extensive formation and diffusion of V_Zn defects following thermal annealing at 500°C under an O₂ atmosphere.

These findings offer new insights into the incorporation of Cu into ZnO nanowires and highlight the importance of thermal activation in forming electrically active Cu_Zn-V_Zn defect complexes. The next step will focus on increasing the content of Cu dopants, effectively promoting Cu_Zn formation rather than having the majority of Cu accumulate on the surfaces of ZnO nanowires. This work opens new perspectives for drastically reducing the detrimental screening of piezoelectric potential through compensatory doping and, consequently, for improving the power density of ZnO nanowire-based piezoelectric devices.

Principal publication and authors
Incorporation and Compensatory Doping Processes of Cu into ZnO Nanowires Investigated at the Local Scale, M. Manrique (a,b,c), B. Salem (b), E. Sarigiannidou (a), H. Roussel (a), F. Wilhelm (d), F. Donatini (e), V. Jacob (f), G. Le Rhun (c), V. Consonni (a), Small Struct., 2400534 (2024); https://doi.org/10.1002/sstr.202400534
(a) Université Grenoble Alpes, CNRS, Grenoble INP, LMGP, Grenoble (France)
(b) Université Grenoble Alpes, CNRS, CEA, Grenoble INP, LTM, Grenoble (France)
(c) Université Grenoble Alpes, CEA, LETI, Grenoble (France)
(d) ESRF
(e) Université Grenoble Alpes, CNRS, Grenoble INP, Institut NEEL, Grenoble (France)
(f) Université Grenoble Alpes, CNRS, IRD, Grenoble INP, IGE, Grenoble (France)

References
[1] R.K. Pandey et al., J. Phys. D: Appl. Phys. 4, 044011 (2021).
[2] J. Villafuerte et al., J. Phys. Chem. C 124, 16652-16662 (2020).
[3] J. Villafuerte et al., Phys. Rev. Mater. 5, 056001 (2022).