|Home | About | Journals | Submit | Contact Us | Français|
La2O3 films were grown on Si substrates by atomic layer deposition technique with different thickness. Crystallization characteristics of the La2O3 films were analyzed by grazing incidence X-ray diffraction after post-deposition rapid thermal annealing treatments at several annealing temperatures. It was found that the crystallization behaviors of the La2O3 films are affected by the film thickness and annealing temperatures as a relationship with the diffusion of Si substrate. Compared with the amorphous La2O3 films, the crystallized films were observed to be more unstable due to the hygroscopicity of La2O3. Besides, the impacts of crystallization characteristics on the bandgap and refractive index of the La2O3 films were also investigated by X-ray photoelectron spectroscopy and spectroscopic ellipsometry, respectively.
During the past decades, lanthanum oxide (La2O3) has raised great research interests due to its remarkable chemical, thermal, optical, and electrical properties [1–3]. On the one hand, featuring with high dielectric constant (approximately 27) and large band offsets with silicon (over 2 eV), La2O3 is one among the most promising high-k dielectric materials to replace SiO2 and Si3N4 in advanced metal-oxide gate stack in semiconductor devices . Up to now, benefiting from the approach of surface passivation prior to oxide deposition, high-quality ceria/lanthana gate stack suitable for high-k integration in a gate-last process has been accomplished . On the other hand, La2O3 is usually used as a kind of effective dopant in thermionic emitters , ferroelectric ceramics , and oxide catalysts , in order to improve properties such as emission capability, effective dielectric constant, and catalytic activity. Besides, La2O3 thin films have also received increasing attentions for the various applications in glass ceramic , gas sensor , supercapacitor , etc.
La2O3 thin films have been prepared by various physical and chemical deposition methods, such as electron beam evaporation , vacuum evaporation , chemical vapor deposition , atomic layer deposition (ALD) , and molecular beam epitaxy . Among the deposition methods mentioned above, due to the nature of the self-limited reaction, ALD has been considered as one of the most promising deposition techniques to produce high quality La2O3 thin films with atomic scale thickness controllability, fine uniformity, and excellent conformality . La2O3 thin films can be found in several crystalline phases, namely, hexagonal (h-La2O3), cubic (c-La2O3), amorphous (a-La2O3), or a mixture of the phases depending on the film deposition method and post-deposition heat treatment . It is well known that the structural properties of La2O3 thin film are determined, to a large extent, by its crystallization and microscopic morphology . Therefore, the study of the crystallization and structure of La2O3 thin film is of great significance for the compatibility of the film application into advanced electronic devices. In this article, the structural properties of La2O3 thin films prepared by ALD technique were investigated by means of a variety of measurements. Attentions were focused on the crystallization conditions of La2O3 film and the structural properties characterized by the crystalline states.
La2O3 films were deposited on p-type Si (100) wafers in an atomic layer deposition reactor (Picosun R-150) using La(i-PrCp)3 as the La precursor while O3 was used as the oxidant. Prior to deposition of the films, native SiO2 was removed in a diluted HF solution (1:50). At the deposition temperature of 300 °C, a steady-state growth rate of ~0.85 Å/cycle is obtained by optimizing the process parameters (0.1 s La(i-PrCp)3 pulse/4 s purge with N2/0.3 s O3 pulse/10 s purge with N2). Ten and twenty nanometer La2O3 films were prepared by varying the number of ALD cycles. For both the 10 and 20 nm La2O3 films, post-deposition rapid thermal annealing (RTA) was carried out at 400, 600, and 800 °C for 60 s in vacuum ambient (~1 mbar). The ellipsometric spectra of La2O3 films were measured before and after annealing by spectroscopic ellipsometry (SE) system (J.A.Woollam Co. M2000U, Lincoln, NE, USA) over the wavelength range from 245 to 1000 nm. In order to address the evolution of the crystallographic structure, grazing incidence X-ray diffraction (GIXRD) measurements were carried out at an angle of incidence of 1° on both the as-grown and annealed La2O3 films. Cross-sectional high-resolution transmission electron microscopy (HRTEM) and energy-dispersive X-ray spectroscopy (EDX) line scan measurements were performed with  direction of the Si substrate to observe the microstructures and atomic compositions of the La2O3 films. X-ray photoelectron spectroscopy (XPS) analysis on a Theta 300 XPS system from Thermo Fisher was employed to investigate the bandgaps of the deposited films. After being exposed to air in clean room environment with a relative humidity of 50% for 48 h, GIXRD and HRTEM measurements were carried out on the as-grown and annealed La2O3 films again for further analysis.
Figure 1 illustrates the GIXRD analysis performed on the as-grown and annealed La2O3 films. The powder patterns of h-La2O3  and h-La(OH)3  are added for comparison. As the GIXRD measurements were carried out immediately after the deposition and annealing process, no peaks attributed to La(OH)3 exist in the GIXRD diffractograms. The 10 nm La2O3 film (as shown in Fig. 1a) shows no diffraction features before and after a 400 °C annealing treatment, suggesting an amorphous disordered structure of the film. After being annealed at 600 and 800 °C, only weak crystalline planes such as hexagonal (101) appear [22, 23], indicating the impossibility of converting the 10 nm La2O3 film into complete crystalline phase. The very small and broad peak around 50° in the diffractogram of the 10 nm La2O3 film annealed at 800 °C does not fit to the h-La2O3 or h-La(OH)3 patterns. We think it may be formed under the influence of several crystalline planes of h-La2O3 around 50°. However, for the 20 nm La2O3, the as-grown film already shows a small degree of crystallinity with a couple of peaks attributed to h-La2O3 (as shown in Fig. 1b). After being annealing treated, the intensities of the GIXRD peaks increase, which means the enhancement in the degree of crystallinity. After annealing at 600 °C, except for the weak cubic (332) plane , the film was mainly crystallized to hexagonal phase as the GIXRD diffractograms exhibit strong hexagonal planes such as (101), (102), (103), and (112). Besides, further increase in the annealing temperature up to 800 °C does not seem to significantly affect the GIXRD diffractograms of the film. That is, upon 600 °C, the increase in the annealing temperature does not enhance the crystallinity of the film. Consequently, when annealed upon 600 °C, an almost complete crystallization could be accomplished for the 20 nm La2O3 film.
Additional structural information at the La2O3/Si interface after RTA treatment at 600 °C is provided by HRTEM-EDX analysis as shown in Fig. 2. For both 10 nm (Fig. 2a) and 20 nm (Fig. 2b) La2O3 films, the lanthanum and oxygen in-depth distributions in the EDX elemental ratio profiles show a parallel profile and the La/O ratio is close to 2:3 which meets well with the stoichiometry of La2O3. In the HRTEM images, an amorphous region between the Si substrate and the fabricated film, corresponding to an interfacial layer (IL) formed during the ALD growth and RTA process , could be found in both Fig. 2a, b. After the amorphous IL, it is possible to identify a region containing nanometer-sized crystals in the 10 nm La2O3 film, indicating the existence of an incomplete structural conversion (from amorphous to crystallographic structure) during the RTA treatment. However, the structure of the 20 nm La2O3 film is a little complicated. With the guidance of dotted lines, an amorphous region, a nanometer-sized crystal transition region, and a long-range ordered crystal region could be observed in the HRTEM image of Fig. 2b. The presence of long-range ordered crystals manifests, in accordance with the GIXRD results shown in Fig. 1, that RTA process upon 600 °C induces an almost complete crystallization of the 20 nm La2O3 film.
It is worth noting that upon the same annealing condition of at 600 °C for 60 s in vacuum ambient (~1 mbar), the 10 and 20 nm La2O3 films show different crystalline characteristics. We attribute this difference to the RTA-induced Si diffusion from the substrate into the La2O3 layer . As we know, La2O3 exhibits the highest affinity for Si atoms among the rare-earth oxide films due to the so called “lanthanide contraction” property of rare-earth elements . Even in the as-deposited La2O3 film grown by ALD method, substrate silicon atoms diffuse moderately and distribute in gradient from Si substrate to the upper layer, causing the presence of an IL about 1 nm [27, 28]. Besides, part of the as-deposited La2O3 film close to the IL could be considered as Si-riched and difficult to crystallize as Si rich help to prevent the formation of crystalline La2O3 precipitates . Furthermore, post-deposition annealing causes extra silicon out diffusion and reaction with excess oxygen in the film. Consequently, in thin La2O3 film with the thickness of 10 nm or less, during the annealing process, the substrate Si atoms would diffuse deep easily to the upper layer before the film is crystallized. However, for the 20 nm as-deposited La2O3, since Si atoms distribute in gradient from Si substrate to the upper layer, a great part of the film relatively far away from Si substrate is pure. We think that this part of La2O3 film could be crystallized at appropriate post-deposition treatment such as RTA carried out at 600 and 800 °C for 60 s in vacuum ambient (~1 mbar) in this work. Crystallization of the film brings in an aggressive enhancement in the packing density and thermodynamic stability. Thus, to a certain extent, the diffusion of Si atoms from substrate into the upper layer would be restrained. As a result, the silicate layer of the 20 nm La2O3 film is slightly thinner than what could be observed in the 10 nm La2O3 film. Besides, in the 20 nm La2O3 film, only 3~4 nm La2O3 closed to the IL was converted into nanometer-sized crystals under the influence of Si diffusion during the annealing process. Complete crystallization of the as-grown film into the h-La2O3 phase is achieved in the region not affected by Si diffusion.
The bandgaps of the as-grown and annealed (at 600 °C) La2O3 films were measured by examining the energy loss of the O 1s core levels as shown in Fig. 3. As we know, the bandgap equals the energy distance between the photoemission peak centroid and the onset of the features due to single particle excitations, and it is usually obtained from the inelastic energy loss features observed on the high binding energy side of the core level photoemission peaks . The onset of O 1s loss spectrum was determined by linearly extrapolating the segment of maximum negative slope to the back ground level . The bandgaps of the as-grown 10 and 20 nm La2O3 films are determined to be 5.55 and 5.45 eV, respectively. These values are in fairly good agreement with Ohmi et al. , who have reported a bandgap of 5.50 eV for non-crystallized La2O3 on Si substrate. The bandgap of the annealed 20 nm La2O3 film is determined to be 5.20 eV, which agrees well with the bandgap of 5.30 eV for crystallized La2O3 reported by Zhao et al. . However, the diffusion of Si during the annealing process brings in large mounts of La-O-Si bonds for the 10 nm La2O3, leading to the increase of the inelastic energy loss during the transition from valence band to conduction band, which means the increment of bandgap . As a result, the bandgap of the annealed 10 nm La2O3 is figured out as 6.0 eV, which is evidently larger than the bandgap of crystallized La2O3.
Figure 4 illustrates the annealing temperature dependence of refractive indexes for the as-grown and annealed La2O3 films revealed by SE fitting. The refractive indexes of the La2O3 films were determined by fitting the ellipsometry data using the well-known Tauc-Lorentz dispersion mode, which was proposed by Jellison and Modine and has been successfully applied to a variety of amorphous and crystallized materials [35–37]. As revealed in Fig. 4, the refractive indexes of the as-grown La2O3 films increase with varying degrees after being annealed at different temperatures. It was reported that the refractive index is closely related to the density of materials, being lower at lower density. Consequently, the increase in the refractive index is caused by the stress release and densification during the annealing process [38, 39]. Furthermore, for the 20 nm La2O3 film, an abrupt increase in the refractive index could be observed when the annealing temperature increased from 400 to 600 °C, indicating an aggressive enhancement in the packing density upon crystallization. As a result, after being annealed at 600 °C, the 20 nm La2O3 film shows an index of refraction of 1.943 at the wavelength of 632.8 nm, which is much higher than that of the as-grown film (1.838). The refractive indexes obtained in this work are of good comparability with the results reported by Armelao et al.  and Kukli et al. .
Figure 5 illustrates the GIXRD diffractograms for the as-grown and annealed La2O3 films after being exposed to air in clean room environment with a relative humidity of 50% for 48 h. Compared with the GIXRD diffractograms obtained before the air exposure as displayed in Fig. 1, almost all the GIXRD peaks attributed to h-La2O3 disappear, whereas new peaks attributed to h-La(OH)3 appear due to the hygroscopicity of La2O3 [22, 23, 41]. It is noteworthy that strong h-La(OH)3 phase peaks are only found in the well crystallized samples such as the 20 nm La2O3 films annealed at 600 and 800 °C, while few weak peaks are observed in the amorphous disordered and nanometer-sized crystallographic samples. Besides, it seems that the air exposure has a much heavier effect on the 20 nm La2O3 than that on the 10 nm La2O3. For clarity, cross-sectional HRTEM measurements on the annealed 10 and 20 nm La2O3 films after the air exposure were performed. The cross-sectional HRTEM image of the annealed 20 nm La2O3 after being exposed to air is shown in Fig. 6b, in which much more uneven interface and surface are observed than what can be found in the 10 nm La2O3. The deteriorations in the interface and surface properties are attributed to the degradation in the film density caused by the conversion from h-La2O3 to h-La(OH)3. With the time exposed to air, the amount of La(OH)3 in La2O3 film increases and then the density of the film is degraded, resulting in the changes of the surface and interfacial morphologies . However, the existence of large mounts of LaSiO in the 10 nm La2O3 enhances the stability of the film structure, providing a high immunity against moisture ambient.
The crystallization of La2O3 film grown by atomic layer deposition on Si substrate is restricted by the thickness of the film and the post-deposition annealing temperature. For thin (~10 nm) La2O3 film, only nanometer-sized crystals are formed after the annealing treatment due to the diffusion of Si substrate. For thick (~20 nm) La2O3, films can be mainly crystallized into h-La2O3 upon RTA performed in vacuum environment at 600 °C. After being crystallized, the refractive index of La2O3 film increases dramatically, while the bandgap is slightly decreased. After an exposure to air for 48 h, the h-La2O3 films are converted into h-La(OH)3 due to the hygroscopicity of La2O3.
The authors gratefully acknowledge the financial supports for this work from the National Natural Science Foundation of China (grant nos. 61376099 and 61434007) and the Foundation for Fundamental Research of China (grant no. JSZL2016110B003). The National Natural Science Foundation of China and the Foundation for Fundamental Research of China did neither participate in the design of the study nor in the collection, analysis, and interpretation of data or the writing of the manuscript. We are very grateful to Dr. Yanlin Pan for the valuable assistance and excellent technical support for the HRTEM-EDX measurements.
XW generated the research idea, analyzed the data, and wrote the paper. XW and LZ carried out the experiments and measurements. XyF and CxF participated in the discussions. SpC and YtW gave kind suggestions about the experiments and measurements. HxL has given final approval of the version to be published. All authors read and approved the final manuscript.
XW, LZ, and CxF are PhD students in Xidian University. HxL is a professor in Xidian University. XyF and YtW are Master students in Xidian University.
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Xing Wang, Email: moc.361@dxgnawx.
Hongxia Liu, Email: nc.ude.naidix.liam@uilxh.
Lu Zhao, Email: moc.361@dxoahzl.
Chenxi Fei, Email: moc.621@8891ixnehcief.
Xingyao Feng, Email: moc.361@33278425131.
Shupeng Chen, Email: moc.621@999gnepuhsnehc.
Yongte Wang, Email: moc.361@tywekim.