PHYSICAL REVIEW MATERIALS 5, 065402 (2021) n- to p-type conductivity transition and band-gap renormalization in ZnO:(Cu+Te) codoped films Alfredo Beristain-Bautista, 1 Daniel Olguín, 2 and Sergio Jiménez-Sandoval 1, * 1 Centro de Investigación y de Estudios Avanzados del I. P. N., Unidad Querétaro, Libramiento Norponiente No. 2000, Fracc. Real de Juriquilla, C. P. 76230, Querétaro, Qro., Mexico 2 Centro de Investigación y de Estudios Avanzados del I. P. N., Departamento de Física, Av. Instituto Politécnico Nacional No. 2508, Col. San Pedro Zacatenco, C.P. 07360, Ciudad de México, Mexico (Received 2 October 2020; accepted 19 May 2021; published 3 June 2021) The natural conductivity of as-grown ZnO is n-type. It has been challenging to produce stable p-type material, which has delayed possible technological applications. In this work, the conductivity transformation from n- to p-type ZnO films deposited by sputtering was followed as a function of copper and tellurium concentrations and of the substrate temperature during growth (room temperature, 150 ◦ C, 250 ◦ C, and 350 ◦ C). The nominal codopant concentrations ranged from 1 to 12 at %. For the minimum concentration, compensation effects yielded highly resistive ZnO. Nonetheless, by tailoring the concentration of Cu and Te, it was possible to vary the resistivity (10 3 –10 −2  cm), mobility (∼10 −2 –10 ◦ cm 2 /V s), and free-hole density (10 15 –10 20 cm −3 ) of p-type ZnO grown at 250 ◦ C. Besides modifying the electrical properties, codoping changed the host band structure significantly, producing a band-gap renormalization from 3.2 eV (UV) to 1.8 eV (red). This control over the band gap is advantageous for applications where controllable photon absorption or emission are sought. Experimentally, films were stable for a period of at least six months. Band-gap engineered p-type ZnO:(Cu+Te) films open the possibility for the fabrication of all-ZnO optoelectronic devices such as homojunction solar cells and/or light-emitting diodes. DOI: 10.1103/PhysRevMaterials.5.065402 Zinc oxide (ZnO) is an earth-abundant, low-cost, wide- band-gap semiconductor whose physical properties make it appropriate for a large variety of applications [1,2]. As in most wide-band-gap materials, n-type doping is easily or un- intentionally achieved, while p-type doping has proven to be a challenging task. The reasons for this difficulty arise mainly from the following causes: (a) low solubility of dopants in the host, (b) deep impurity energy levels that do not contribute effectively to conductivity, and (c) the formation of low- energy compensating defects (such as Zn i ,V O , complexes, and hydrogen, a common contaminant in growth chambers). Different approaches have been undertaken to overcome these difficulties, which include (i) codoping aimed to improve the solubility of shallow acceptors and reduce the defect ioniza- tion level, (ii) the theoretical and experimental search for ap- propriate group-I and group-V dopants, and (iii) modification of the ZnO band structure through appropriate defect bands inside the band gap [3]. The most common codoping approach has focused on improving the low solubility of nitrogen (an appropriate acceptor candidate due to its low ionization en- ergy) in ZnO using acceptor-donor pairs (e.g., N and group-III elements) [4]. Although this route produced samples with p-type characteristics, some issues remain unsolved such as the position of the N and group-III levels within the band gap, and the long-term stability [3]. Yan et al. proposed reducing the dopants ionization energies by introducing acceptor-donor passivated impurity bands, complemented by an effective dop- ing made by increasing the acceptor concentration [5]. * Correspondence author: sergio.jimenez@cinvestav.mx Copper is a well-known p-type dopant in ZnO; however, it forms deep acceptors (ionization energy of Cu Zn = 0.35 eV) that contribute to conductivity but not efficiently [6]. Through density functional theory calculations, it was shown that Cu affects the ZnO band structure in such a way that the top of the valence band shifts upwards and the bottom of the con- duction band downwards, creating a reduced effective band gap [7]. In the case of tellurium, it has been reported that the incorporation of Te in the ZnO lattice may be beneficial in two ways. First, it suppresses the formation of Zn i , a common n-type defect; second, in the alloy regime ZnO:Te modifies the host band structure, producing shallow acceptors [8]. When used as codopant with nitrogen, it has been reported that Te may act as a surfactant to decrease the formation energy of N O defects, thus improving nitrogen solubility [9]. Moreover, it was observed that Te induced a small blue shift in the band gap of ZnO [10,11]. The approach followed in this work has focused on pur- suing the use of alternative codopants. Cu is a well-known p-type dopant, which indicates that it enters as a substitu- tional impurity (Cu Zn ) rather than interstitial (Cu i is a n-type dopant). The experimental work undertaken here shows that Te, when used as codopant with Cu, produces conductivity with holes as majority carriers. It is pointed out, however, that the experimental findings reveal that the formation of p-type material depends upon the concentration of Cu+Te and on the thermal energy supplied to the atomic species during growth, i.e., the substrate temperature. Additionally, in the al- loy regime Cu and Te favorably modify the host band structure for hole-based conductivity. The deep acceptor characteristic of Cu, when single dopant, is lifted. Codoping ZnO with Cu 2475-9953/2021/5(6)/065402(10) 065402-1 ©2021 American Physical Society