Next Article in Journal
Impact Abrasive Wear Resistance of CrN and CrAlN Coatings
Next Article in Special Issue
Evaluation of Transparent ITO/Nano-Ag/ITO Electrode Grown on Flexible Electrochromic Devices by Roll-to-Roll Sputtering Technology
Previous Article in Journal
Effects of Magnetohydrodynamics Flow on Multilayer Coatings of Newtonian and Non-Newtonian Fluids through Porous Inclined Rotating Channel
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Coatings
Volume 12
Issue 4
10.3390/coatings12040431
coatings-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Structure and Electrical Properties of Zirconium-Aluminum-Oxide Films Engineered by Atomic Layer Deposition

by
Joonas Merisalu
*,
Taivo Jõgiaas
,
Toomas Daniel Viskus
,
Aarne Kasikov
,
Peeter Ritslaid
,
Tanel Käämbre
,
Aivar Tarre
,
Jekaterina Kozlova
,
Hugo Mändar
,
Aile Tamm
,
Jaan Aarik
and
Kaupo Kukli
Institute of Physics, University of Tartu, W. Ostwaldi 1, 50411 Tartu, Estonia
*
Author to whom correspondence should be addressed.
Coatings 2022, 12(4), 431; https://doi.org/10.3390/coatings12040431
Submission received: 21 February 2022 / Revised: 17 March 2022 / Accepted: 21 March 2022 / Published: 23 March 2022
(This article belongs to the Special Issue Atomic Layer Deposition for the Synthesis and Application of Thin Films)
Browse Figures
Versions Notes

Abstract

:
Thin films containing either multilayer ZrO2:Al2O3 structures or ZrO2 deposited on ZrxAlyOz buffer layers were characterized. The films were grown by atomic layer deposition (ALD) at 300 °C from ZrCl4, Al(CH3)3, and H2O. The multilayer ZrO2:Al2O3 structures were grown repeating different combinations of ZrO2 and Al2O3 ALD cycles while the ZrxAlyOz layers were obtained in a novel process using ALD cycles based on successive adsorption of ZrCl4 and Al(CH3)3, followed by surface reaction with H2O. The films were grown on TiN electrodes, and supplied with Ti top electrodes, whereby ZrxAlyOz films were exploited as thin buffer layers between TiN and ZrO2. The as-deposited ZrO2 films and ZrO2:Al2O3 structures with sufficiently low concentrations of Al2O3 were crystallized in the form of cubic or tetragonal ZrO2 polymorph possessing relative permittivities reaching 35. Notably, multilayered ZrO2:Al2O3 films could exhibit resistive switching behavior with ratios between low- and high-resistive-state current values, extending up to five orders of magnitude. Implications of multilevel switching were recorded. In the double-layered ZrxAlyOz-ZrO2 stacks, the ON/OFF current ratios remained below 40, but the endurance could become extended over 3000 cycles. Remarkably, instabilities, when detected in endurance behavior expressed by reduction in an ON/OFF current ratio could be compensated and the current values restored by real time readjustment of the programming voltage amplitude.

1. Introduction

Atomic layer deposition (ALD) has enabled the development of ZrO2-based capacitor structures for dynamic random-access memories (DRAM) scalable according to the industrial requirements [1,2,3,4]. In regard with nonvolatile memory technologies, ZrO2 has been mentioned among numerous metal oxides investigated as resistively switching (RS) media in recent reviews [5,6,7]. Several studies have been carried out using ZrO2 as a single RS medium of memristors [8,9,10,11,12]. An alternative approach, supported with first-principles calculations, has been based on nitrogen-doped ZrO2 serving as a multilevel RS medium containing oxygen vacancies coupled with nitrogen in ZrO2 between Pt and TiN electrodes [13]. For analogous reasons, double layers consisting of oxygen deficient ZrO2-x on top of ZrO2 have been grown and characterized [14]. ZrO2 films have been also stacked with other metal oxides, such as TiO2 [10,15], to usefully tune the RS parameters. Notably, in pulsed laser deposited ZrO2 films embedding Cu nanoparticles, the RS memory window, expressed by the low- to high-resistivity (denoted also as ON/OFF) current ratio (ILRS/IHRS), has been extended over four orders of magnitude [16].
A few decades ago, ALD was successfully used for the development of trenched DRAMs based on single Al2O3 thin insulating dielectric films grown conformally on three-dimensional substrates [17]. Later, amorphous Al2O3 as a layered additive in hosting ZrO2 film turned out to be a useful dopant stabilizing cubic and/or tetragonal polymorphs of ZrO2 with dielectric permittivity values markedly exceeding that of the monoclinic ZrO2. In addition, doping has allowed for the deposition of films with smoother surfaces and concurrent reduction of the leakage currents through the dielectric layers. Such advantages have enabled the development of ZrO2-Al2O3 stacked structures for DRAMs [1,2,3,4].
Some studies have reported RS in oxide films comprising an oxygen-deficient Al2O3-x witching layer deposited on stoichiometric Al2O3 [18,19]. In a paper by Huang et al. [19], sputtering and ALD techniques were complementarily applied to fabricate AlOx/Al2O3 stacks, whereby Al(CH3)3 and H2O were exploited as ALD precursors. Multilevel RS was observed in a double-layer AlOx/Al2O3 film, both layers grown by ALD from Al(CH3)3 and H2O [18]. High ILRS/IHRS ratios, exceeding four orders of magnitude, were achieved in the case of multilayer devices consisting of thermally grown Al metal and electron beam evaporated Al2O3 films [20]. Comparably large ILRS/IHRS ratios have been obtained even earlier in devices based on Al2O3 films fabricated by plasma oxidation of Al electrodes [21].
In many cases, multilayered RS media have been built up on metal oxide layers characterized either by different defect densities or different valences of metals forming the compounds layered alternately. For example, triple-layered ALD structures consisting of Al2O3 films between bottom and top HfO2 films became functional due to the differences in the oxidation states leading to the formation of energetic barriers, which promoted multilevel switching phenomenon [22]. However, aside of the proven relevance of Al2O3- and ZrO2-based DRAM media, reports on RS media built on combinations of ZrO2 and Al2O3 are scarce in general. Only a few research papers have reported the appearance of RS in ZrO2:Al2O3 mixtures [23] and periodical multilayers [24] grown by ALD, and in ZrO2/Al2O3 double layers sputtered on flexible transparent substrates [25]. It has also been reported that larger ILRS/IHRS ratios could be achieved with ZrO2/AlON double layers, compared to the devices based on single ZrO2 films [26].
In the present study, RS was investigated in dielectrics containing, first, periodical ZrO2:Al2O3 structures and, second, ZrO2 films deposited on buffer ZrxAlyOz layers in order to obtain new knowledge concerning the influence of the dielectric layer structure on the dielectric properties and RS performance of ZrO2-based dielectrics. The RS media were grown to thicknesses of 12–15 nm by ALD at 300 °C, using Al(CH3)3, ZrCl4 and H2O as precursors. Dependence of the composition, crystalline structure, RS polarity and stability of ILRS/IHRS ratios on the ALD cycle ratios and sequences used for the deposition of ZrO2:Al2O3 structures, and application of a novel ALD process for deposition of ZrxAlyOz were addressed.

2. Materials and Methods

The RS films were deposited in an in-house built hot-wall flow-type ALD reactor schematically described in an earlier paper by Arroval et al. [27]. The films were grown at 300 °C on crystalline TiN-covered substrates cut out of Si(100) wafers with a resistivity of 0.014–0.020 Ω⋅cm. The Si wafers were boron-doped to concentrations ranging from 5 × 1018 to 1 × 1019 cm–3. A TiN layer of 10 nm was pre-grown by pulsed chemical vapor deposition using a batch TiCl4/NH3 process [28,29] at temperatures of 450–500 °C in an ASM A412 Large Batch 300 mm reactor (Advanced Semiconductor Materials, Almere, The Netherlands). at Fraunhofer IPMS-CNT (Dresden, Germany).
The precursors used for ALD were ZrCl4, Al(CH3)3 (TMA) and H2O. Two different sample sets were prepared with different distributions of Al2O3 in the RS layers. The RS media of the first set were grown as nominally periodical ZrO2:Al2O3 multilayers (Figure 1a) using ALD cycle sequences 30 × (1 × Al2O3 + 4 × ZrO2) and 6 × (1 × Al2O3 + 24 × ZrO2). This means that one single Al2O3 deposition cycle consisting of sequential exposures of surfaces to Al(CH3)3 and H2O pulses was performed alternately with 4 or 24 consecutive ZrO2 deposition cycles consisting of exposures to ZrCl4 and H2O. In these samples, the formation of continuous Al2O3 layers, intermediating ZrO2, was not expected because of the inability of a single ALD cycle to form a full monolayer due to the low nucleation density during the related incubation period [30]. By application of intermediate single Al2O3 cycles, doped ZrO2 layers rather than nanolaminates were targeted. Since the insertion of Al in the ZrO2 host was realized via surface reactions between Al(CH3)3 and H2O, the solid film materials eventually formed are to be denoted as ZrO2:Al2O3 instead of ZrxAlyOz. In addition, binary Al2O3 and ZrO2 films were deposited as reference dielectrics.
In the second set of samples, asymmetric double-layer structures were devised by depositing, at first, a ZrxAlyOz buffer layer on the bottom TiN electrode, creating a chemically mixed and, presumably, structurally more defective starting layer between the bottom TiN electrode and the host ZrO2 medium (Figure 1b). The ZrxAlyOz buffer layers were deposited using ALD cycles consisting of consecutive exposures of the surface to ZrCl4, Al(CH3)3, and H2O pulses, that is, without the application of a water pulse after the ZrCl4 pulse. The amount of such cycles, applied successively, varied from 1 to 5. The deposition of ZrxAlyOz buffer layers was always followed by the deposition of the main medium consisting of ZrO2. The ZrO2 layer was deposited applying 80 ALD cycles consisting of sequential exposure of the surface to ZrCl4 and H2O pulses. Additionally, a 11 nm thick ZrxAlyOz reference film was deposited applying 80 ALD cycles consisting of successive ZrCl4, Al(CH3)3, and H2O pulses. Unfortunately, the results of comparative composition analysis performed for the reference film and the ZrxAlyOz-ZrO2 stacks indicated that, probably because of an incubation period typical for chloride-based ALD processes [31], the composition of the ZrxAlyOz buffer layers obtained after 1–2 ALD cycles could not be controlled with sufficient accuracy. Therefore, these samples were not used in the further studies described in this paper.
Useful knowledge, gained before the present experiments on either Al-doped ZrO2, or sandwiched Al2O3 and ZrO2 films, allowed one to expect the formation of metastable cubic or tetragonal phase of ZrO2, highly stable against heat-treatments [32,33]. Overall, an insulating material with bandgap likely lower than that characteristic of stable bulk ZrO2 would form [34,35]. It has also been found earlier that defect clusters, consisting of Al atoms associated with oxygen vacancies, should form as a result of Al doping in the cubic phase of ZrO2, whereas the concentration of such clusters should remain negligible in the monoclinic, i.e., stable ZrO2 [36]. Thus, the ZrxAlyOz interfacial layer could perform as a reservoir of defects below the host ZrO2 and, possibly, be a source of ionic current. Application of a similar interface layer has recently been reported in studies of TiN/HfO2:Al/HfO2/Pt [37] and Pt/Al2O3/HfO2/HfAlOx/TiN [38] RRAM stacks. In the present study, the content of aluminum in the ZrxAlyOz buffer layer was expected to be lower than that of zirconium, because in the successive exposures of surface to ZrCl4 and Al(CH3)3 without intermittent supply of oxygen precursor, the incorporation of zirconium into the film material was more likely than that of aluminum. In order to enable convenient elemental analysis of the film material, a reference 11 nm thick ZrxAlyOz film was deposited applying 80 deposition cycles consisting of ZrCl4, Al(CH3)3, and H2O consecutive pulses.
In order to determine the film thickness, optical measurements of the structures were performed with a spectroscopic ellipsometer GES-5E (Semilab Co., Budapest, Hungary) using a microspot option. The spot size was about 0.35 × 0.8 mm at an incidence angle of 65°, and the converging angle of a beam was about 4°. Fitting was performed using SEA software. The fitting quality was characterized using a correlation function R2 reaching unity for ideal correspondence between the measured and computed spectra.
Elemental composition of the films was, in general, measured by wavelength dispersive X-ray fluorescence spectroscopy (XRF) using a spectrometer ZSX-400 (Rigaku, Tokyo, Japan). The growth cycle sequences, thicknesses and elemental compositions of both ZrO2:Al2O3 and selected ZrxAlyOz-ZrO2 sample sets are presented in Table 1 and Table 2. In addition, X-ray photoelectron spectra (XPS) were collected from selected ZrxAlyOz, ZrO2 and Al2O3 reference samples using a Gammadata/Scienta SES100 hemispherical photoelectron energy analyser (in constant transmission mode) and a non-monochromatic Mg-Kα (hv = 1253.6 eV) X-ray excitation source, a Thermo VG Scientific XR3E2 non-monochromatic dual Al/Mg anode X-ray gun (East Grinstead, West Sussex, UK) at an overall spectral resolution of approximately 0.8 eV. The minor sample charging was corrected for by adjusting the adventitious carbon (assumed to contain mainly sp3-hybridised carbon species, typically present in ex situ measured samples) C 1s peak to 285.1 eV. Spectral components were fitted, and elemental content from survey spectra was estimated using CasaXPS software (version 2.3.19) [39], taking into account the photoionization cross-sections at the excitation energy used and the measurement geometry. To estimate the (kinetic energy dependent) depth range of origin of detected photoelectrons, their inelastic mean free path (IMFP) was estimated using the TPP2M formula [40] and also weighed into the XPS sensitivity factor of each core level.
The crystal structure was evaluated by grazing incidence X-ray diffraction (GIXRD) method using an X-ray diffractometer SmartLab (Rigaku, Tokyo, Japan) and the CuKα radiation with a wavelength of 0.15406 nm. Scanning transmission electron microscopy (STEM) and elemental mapping of the ZrO2:Al2O3 films in cross-sectional orientation were performed in a Cs-corrected Titan Themis 200 microscope (FEI, Hillsboro, OR, USA) equipped with a Super-X energy dispersive X-ray spectrometry (EDX) system (FEI/Bruker) at 200 kV. EDX maps were acquired using Esprit software version 1.9 (Bruker, Billerica, MA, USA). Thin cross-sectional samples for STEM observations were prepared using the in situ lift-out technique using Helios Nanolab 600 scanning electron microscope/focused ion beam system (FEI, Hillsboro, OR, USA). In order to protect the surface from ion milling during the preparation of STEM samples, the area of interest was covered with a platinum protection layer.
For the electrical measurements, the samples were supplied with titanium top electrodes, electron-beam evaporated (EBE) to a thickness of 70 nm through a shadow mask at 230 °C. The titanium electrodes of 6 nm for periodical ZrO2:Al2O3 film structures were additionally covered with 100 nm of gold in the same EBE process to ensure sufficiently low contact resistance between a measurement probe and the Ti layer remaining in a direct contact to ZrO2. The top electrodes used in the measurements had diameters of 50 and 250 μm, and areas of 0.002 and 0.052 mm2, respectively. Electrical measurements were carried out in a light-proof and electrically shielded box on Cascade Microtech MPS150 probe station (Beaverton, OR, USA), using Keithley 2636A source-meter (Keithley Instruments, Cleveland, OH, USA) and Agilent E4980A LCR analyzer (Agilent Technologies, Palo Alto, CA, USA).

3. Results and Discussion

3.1. Composition

The XRF analysis revealed that the contents of aluminum and zirconium were expectedly correlated to the ZrO2:Al2O3 cycle ratios used for growing the films with cycle sequences providing homogeneous distribution of Al2O3 intermediate layers throughout the film thickness (Table 1). The zirconium content was naturally higher in the film grown using a higher relative amount of ZrO2 deposition cycles. The contents of oxygen and metal atoms approximately corresponded to that in oxygen-deficient zirconium dioxide, influenced by the presence of Al2O3. In addition, the presence of residual chlorine, originating from the ZrCl4 precursor, was detected in the films. The content of chlorine that did not depend on the Zr to Al ratio (Table 1) was well comparable to the contents measured in the ZrO2 films grown from ZrCl4 and H2O under similar conditions earlier [41].
The presence of rather low contents of aluminum was detected in the films deposited directly on the TiN electrode surface by applying successive ZrCl4, Al(CH3)3 and H2O pulses, followed by the deposition of the main ZrO2 layer, thus creating asymmetric solid media. The XRF results (Table 2) indicated that the relative amounts of aluminum, as measured, did not correlate with the amounts of ZrCl4, Al(CH3)3 and H2O pulse sequences. This could be explained by the small thickness of the ZrxAlyOz buffer layer between TiN and ZrO2. The host ZrO2 has somewhat variable thicknesses because of either experimental uncertainty or lateral thickness profile characteristic of chloride-based ALD of ZrO2 or other chloride-based processes in research-scale flow-type reactors, which has been known over decades [42,43].
Expectedly the reference ZrxAlyOz film, grown by applying 80 ALD cycles, each consisting of successive ZrCl4, Al(CH3)3 and H2O pulses, contained relatively small amounts of aluminum; 7.4 at.% of aluminum against 26.8 at.% of zirconium (Table 2). Analogous phenomenon has been studied earlier in an ALD process, where the Al deposition rate was inhibited in the case on consecutive TiCl4-Al(CH3)3-H2O pulses, compared to that in the case of conventional TiCl4-H2O-Al(CH3)3-H2O cycle sequence [44]. This effect was due to the occupation and screening of adsorption sites for Al(CH3)3 by species formed during the preceding TiCl4 pulse. The reason for why Al(CH3)3 can be adsorbed on the surface saturated by a metal chloride is the saturation of chloride adsorption because of steric effects rather than due to the occupation of all possible adsorption sites. It appeared that steric effects, which were able to stop further adsorption of TiCl4, did not avoid adsorption of Al(CH3)3 [44].
It has been studied and recognized earlier that, in thermal ALD process of ZrO2 from ZrCl4 and H2O, submonolayer growth per cycle is to be considered [45] and obtained because of steric effects limiting the adsorption of ZrCl4 [46]. As ALD of Al2O3 from Al(CH3)3 and H2O should be more feasible process on hydroxylated surfaces, compared to the growth of ZrO2 from ZrCl4 and H2O [47], Al(CH3)3 can be adsorbed even on sites that are not able to adsorb ZrCl4 due to the steric limitations. However, because of the occupation and screening of adsorption sites by the metal precursor first adsorbing in the sequence, the amount of precursor adsorbing after the first one is limited.
Within the accuracy limits of XPS analysis, the composition estimates derived from the XPS survey spectra (Figure 2a) revealed the stoichiometric zirconium to oxygen (Zr:O) atomic ratio of 1:2 within the 2–3 nm analysis depth from the reference zirconium oxide film surface. At the same time for the reference aluminum oxide, the Al:O atomic ratio estimated was approximately 2.0:4.5, instead of the nominal 2:3 for the stoichiometric Al2O3, indicating the presence of excess oxygen. This can be accepted as somewhat expected result, because formation of aluminum hydroxides can be considered as a possible result of surface reactions between Al precursors and H2O. This result is consistent, for instance, with that earlier obtained for Al2O3 films grown by chemical vapor deposition in the work of Rawat et al. [48] where the results of XPS analysis were interpreted assuming dominance of surface hydroxyls over the fully oxidized aluminum.
Similarly, from the O 1s spectrum measured in the present study (Figure 2b), one could recognize the presence of a strong hydroxyl signal. Even the Al 2p band (Figure 2c) could suggest the hydroxide being even more dominant, with a narrow single feature spectrum at ~0.6 eV higher than that detected in the mixed ZrxAlyOz reference film. The spectra from the latter film well converged with the values earlier reported for crystalline alumina [49,50] and also for ZrO2-Al2O3 composite oxides [51].
The Zr 3d binding energy spectra (Figure 2d) allowed one to consider full Zr4+ oxidation state in both ZrO2 and ZrxAlyOz reference films. One could also note a visible, but numerically negligible, broadening of the Zr 3d lines without appearance of the second component (doublet) in the ZrxAlyOz spectrum, suggesting all the Zr being accommodated in equivalent sites.
Comparison of the O 1s spectra (Figure 2b) allowed one to recognize surface hydroxyls present in similar minor amounts in both ZrxAlyOz and ZrO2 reference films, but in much lesser amounts than in the Al2O3 reference film. The binding energy of the hydroxyl groups, however, coincides with that attributed, in an earlier study, to oxygen vacancies at neighboring sites [51]. There with the possible presence or absence of such defects in different samples can be rationalized to a certain degree by looking at elemental abundancies. At first, the feature just below 532 eV binding energy (Figure 2b, an orange dashed component line) is strong in the O 1s spectrum of the Al2O3 sample, where the content of oxygen is above that required for Al2O3 stoichiometry, which therefore suggests the presence of OH groups rather than oxygen vacancies. In the ZrxAlyOz, the oxygen content, which was estimated from the survey spectrum (Figure 2a), was lower than that in the corresponding stoichiometric oxide. The latter observation could indicate the likely presence of oxygen vacancies, considered to be also present in the alumina-stabilized zirconia studied in an earlier work [51].

3.2. Structure

Figure 3 depicts X-ray diffraction patterns of films deposited as multilayer ZrO2:Al2O3 structures. One can see that the films were crystallized in their as-deposited state. The low intensity and large broadening of the XRD reflections indicated that the films could be regarded as nanocrystalline, as it was possible to expect, taking into account the doping effect and low thickness of the films. The patterns indicated the formation of metastable tetragonal or cubic polymorphs in ZrO2. It became also obvious that the application of decreasing amounts of successive ZrO2 deposition cycles between single Al2O3 cycles, has led to the growth of more homogeneously mixed oxide layers, inhibiting the structural ordering down to the complete disordering, that is, inducing growth of amorphous matter, as observed in the cases of ZrO2:Al2O3 films grown using a cycle ratio of 4:1 (Figure 3).
Figure 4 demonstrates cross-sectional bright-field (BF) STEM images from a representative ZrO2:Al2O3 film grown on TiN bottom electrode and supplied with Au/Ti top electrodes. Crystal growth in the ZrO2:Al2O3 film has become obvious according to the STEM images, supporting the XRD results. The interface between ZrO2:Al2O3 film and TiN electrode appeared somewhat diffuse due to certain roughness of the TiN layer, but the whole medium was clearly crystallized throughout all layers.
Figure 5 presents the results of STEM-EDX elemental composition mapping carried out on a representative ZrO2:Al2O3 film grown on TiN bottom electrode and supplied with Au/Ti top electrodes. One can clearly see that the constituent films have formed with defined interfaces between the switching medium and mounting electrodes. Quite expectedly, no separate continuous Al2O3 layers have been formed during single Al2O3 deposition cycles, applied periodically between multiple ZrO2 layers, rather a diffuse, mixed-oxide film containing both Zr and Al oxides has been formed.
Figure 6 represents X-ray diffraction patterns of the double-layer ZrxAlyOz-ZrO2 structures. Quite expectedly, these films showed higher crystallinity in comparison to those deposited as multilayer ZrO2:Al2O3 structures. The 10–15 nm thick ZrO2 layers obtained in the ZrxAlyOz-ZrO2 structures after applying 80 ALD cycles were ordered in the form of metastable cubic and/or tetragonal polymorphs. Similarity of diffraction patterns of these polymorphs did not enable distinction between tetragonal and cubic phases unambiguously. Nevertheless, domination of the tetragonal phase in the films seems to be more likely considering the results presented in Figure 6. A possible effect of the nominal thickness of interfacial ZrxAlyOz layer between TiN electrode and ZrO2 was not clarified in terms of its correlation to the structural ordering. The interfacial layer itself must have grown as amorphous, as was concluded on the basis of XRD results obtained for the reference film grown using 80 ALD cycles (the bottom pattern in Figure 6).
Figure 7 demonstrates a cross-section image from the ZrxAlyOz-ZrO2 stack deposited on a TiN bottom electrode. One can see, that both the main ZrO2 film and the TiN electrode layer contain crystallites with hardly distinguishable boundaries between crystallites with different orientations and between crystallites and amorphous regions. The ZrxAlyOz buffer layer has, quite expectedly, also remained too thin for the unambiguous detection by energy-dispersive X-ray elemental analysis (not shown), as accumulation of aluminum could not be recognized.

3.3. Dielectric Properties

Figure 8 depicts results of the capacitance dispersion measurements, transferred to the permittivity-frequency dependences. Since the TiN and Ti electrodes of the stacks could unlikely cause formation of interface layers, which would have been insulating and of lower permittivity compared to the main medium, simple parallel-plate capacitor formula based on a single dielectric material could be used for the estimation of the relative permittivity (k) in a good approximation. Herewith the k versus frequency curves (Figure 8) were averaged over at least ten measurement points on different dot electrodes arbitrarily selected over the electrode matrix.
The k values of dielectric layers with different compositions expectedly appeared as clearly distinguishable. The amorphous Al2O3 film possessed the lowest k value in the sample set, not exceeding 11 (Figure 8a). The nanocrystalline ZrO2 film, possibly containing also some amount of amorphous phase in addition to metastable cubic/tetragonal phases, exhibited k slightly above 30. Remarkably, in the ZrO2:Al2O3 film grown using one Al2O3 ALD cycle alternately with 24 ZrO2 cycles, k values exceeding 35 were measured. In the latter film, the medium was also crystallized dominantly in the form of cubic or tetragonal polymorph (Figure 3). Higher content of tetragonal phase in these films compared to that in the undoped ZrO2 films is a possible reason for the increase of permittivity with this kind of doping with Al2O3.
Expectedly, the amorphous ZrO2:Al2O3 film grown using one Al2O3 cycle alternately with four ZrO2 ALD cycles (Figure 3) exhibited k values close to 20, that is, values which are between those of Al2O3 and ZrO2 (Figure 8a). At the same time, the capacitance values could not be measured reliably at relatively low measurement frequencies, i.e., below 5 kHz, for the same sample. This might be indicative of an increasing role for the space charge polarization or, at least, a tendency towards conductivity enhancing stepwise when closer steady state potentials. The polarization stability influenced by the internal distribution of Al and Zr in the solid medium evidently affected the electrical characteristics even further. Notably, the ZrO2-Al2O3 films behaved differently under direct current measurements in resistive switching regime. The films grown using ZrO2:Al2O3 cycle ratio of 24:1 enabled defined bipolar wide-window resistive switching, whereas the films grown using ZrO2:Al2O3 cycle ratio of 4:1 allowed one to observe certain destabilization of the window with the concurrent appearance of multilevel switching, as will be described below.
The experimental k values measured for different materials in the present study are approximately consistent with those reported in the literature. In the case of crystalline ZrO2, calculations have allowed one to expect orientation-dependent k values up to 10–17, 32, and over 40 for the monoclinic, cubic and tetragonal ZrO2 polymorphs, respectively [52]. High permittivity values have also been forecasted for amorphous ZrO2. According to the first principles calculations of Vanderbilt et al. [53], the electronic density distribution indicated that there are no defect states in the band gap of ZrO2, despite the amorphous nature of the material and accompanying fairly broad distribution of coordination numbers of the atoms. The atomic- and coordination-number-dependent Born effective charges in amorphous material, naturally smaller than those in the crystal lattice, would yield somewhat lower dielectric activities for the phonon modes, and an orientation-averaged lattice k may not exceed 17.6. Nonetheless, it has been concluded earlier that the average k of disordered ZrO2 as high as 22.2 can be considered [53]. Such permittivity values for crystalline and disordered ZrO2 have been supported by a number of experimental studies and may be regarded as reference values for ZrO2 films synthesized by various methods [54].
The k values of single-crystal Al2O3, as reported, should not rise over a value of 9.5 [55] while those ranging from 6 to 7 have been measured for 12 nm thick amorphous Al2O3 films grown by ALD from Al(CH3)3 and H2O in the temperature range of 125–425 °C, without clear dependence on the growth temperature [56]. Higher k values up to 8.9 may have been achieved and reported for ALD-grown 3–5 nm thick Al2O3 films after specific substrate electrode surface treatments [57].
For comparison, studies on composites, or sandwiched and multilayered stack structures consisting of ZrO2 and Al2O3, have reported permittivity values correlated to the relative content of constituents. For instance, in ZrO2-strengthened Al2O3 powders containing 15 wt.% ZrO2, divided to 9 wt.% of tetragonal and 6 wt.% of monoclinic ZrO2 phases, k reached 14 at room temperature [58]. A k value of 20.1 was measured in electron beam evaporated ZrO2–Al2O3 composite films, annealed and partially crystallized as tetragonal ZrO2 [59]. In sol-gel processed crystallized ZrO2–Al2O3 nanolaminates, consisting of 7–40 nm thick ZrO2 and Al2O3 constituent layers sandwiched to total thicknesses of 70 nm, k correlated to the relative content of ZrO2, with edge values of 6.9 and 20.3, characteristic of reference Al2O3 and ZrO2, respectively [60]. Similar behavior in sol-gel deposited ZrO2–Al2O3 nanolaminates was observed and described in another, more recent, paper, as well [61]. Although the alternately layered sandwich structures of Al2O3 and ZrO2 are not necessarily to be regarded as homogeneous mixtures, the effective permittivity values should at least partially follow certain mixing rules relevant to composites of distinct compounds [62,63].
Waggoner et al. [64] have measured a k value of 15 for ZrO2:Al2O3 nanolaminates, which were grown by ALD from Zr alkylamide and Al(CH3)3, consisted of 1.25–15 nm thick Al2O3 layered alternately with ZrO2 to total thicknesses of 40–200 nm, and contained 75% of ZrO2, whereas in the reference Al2O3 and ZrO2 films, the k values were 7 and 25, respectively. In another study, triple layers consisting of 2 nm-thick Al2O3, 20-nm-thick ZrO2 and 2 nm-thick Al2O3, sequentially grown by ALD from Al(CH3)3 and Zr[(CH3)2N]4, possessed k up to 41.9 after microwave annealing [65]. Yet in even earlier study, k values up to 38 were reported as characteristic of industrially relevant memory dielectric stacks consisting of sandwiched 0.3–5.5 nm-thick tetragonal ZrO2 and amorphous Al2O3 layers grown from Zr[N(CH3)C2H5]4, Al(CH3)3, and O3 at 300 °C by ALD [2].
The averaged k values measured for the ZrO2, ZrO2:Al2O3, and ZrxAlyOz–ZrO2 films in the present study were, in a good approximation, consistent with the values either predicted or measured earlier for the amorphous, metastable, or appreciably crystallized media composed of the same elements. It is reasonable to assume that higher permittivity values accompany better-defined phase composition with higher concentration of tetragonal phase in the films. Furthermore, defined phase composition could ensure a clearer distinction between different conductivity states, provided that the capacitance (permittivity) dispersion behavior remains stable, without fluctuations in insulating properties at low frequencies.

3.4. Resistive Switching Performance

Resistive switching was observed and recorded in the films, which contained ZrO2 constituent layers, but could not be initiated in reference Al2O3 as well as in reference ZrxAlyOz, whereby the latter was deposited via successive adsorption steps and reactions between ZrCl4, Al(CH3)3 and H2O.
In the 14.6-nm-thick multilayered film grown using the deposition cycle sequence of 6 × (1 × Al2O3 + 24 × ZrO2), clear bipolar switching behavior was achieved (Figure 9a). The SET and RESET voltages were around –1.6 and 2.2 V, respectively. Remarkably, the ILRS/IHRS ratios reached values up to 105. However, despite an initially markedly uniform and wide switching window, the endurance of such devices did not reach acceptable ranges, as the switching collapsed after 25 switching cycles (Figure 9b). For comparison, somewhat analogous switching current–voltage curves have earlier been recorded for Au/HfO2/TiN devices [2,66], whereby the endurance tests allowed monitoring of stable switching during about 102 cycles with an ILRS/IHRS ratio of 104 [66]. Clarification of the reasons behind high, but unstable, current ratios would require further detailed studies, including evaluation after structural modifications. In general, higher switching stability could be a result of higher chemical stability and structural uniformity of the medium in terms of the phase composition. Speculatively, one could expect higher switching stability in the HfO2-based medium, due to its tendency to preferably grow in the stable monoclinic phase, differently from ZrO2 that can more easily occur as a multiphase material comprising stable and metastable phases.
In a film grown using the deposition cycle sequence of 30 × (1 × Al2O3 + 4 × ZrO2), the maximum ILRS/IHRS ratios also reached values close to 105 (Figure 10). In the case of this sample, differently from the previously described one, clear indications of multistate switching were observed, as implied by two appreciably distinct RESET voltages established nearby 1 V and above 2 V (denoted as RESET I and RESET II in Figure 10a). The sequential SET and RESET I events, followed by SET and RESET II operations carried out, can be better followed in Figure 10b, where rather minor, but still distinct differences can be noticed between LRS from different sequences denoted as LRS1 and LRS2 although in general the difference is not sufficient for practical application. The difference between LRS1 and LRS2 is probably caused because of the lower resistance in HRS1 and quicker current rise during the SET operation hence longer duration of higher current being applied. The implications of multilevel performance were, thus, observed in the sample deposited by layering Al2O3 into ZrO2 with markedly shorter period, that is, in the sample containing more Al2O3. Thus, the multilevel RS might be related to the lower electrical stability of highly disordered, X-ray amorphous medium, although high ILRS/IHRS ratios were initially observed.
Multilevel RS has, probably intentionally, been initiated and recorded in ZrO2-based devices earlier, whereby the appearance of multistate behavior was explained by the effects of dopant atoms or layers of variable composition embedded by the host material. For instance, multilevel switching was observed in Ti/ZrO2/n+-Si cells where Cu atoms from a Cu layer deposited between top Ti electrode and the host ZrO2 medium were thermally diffused as dopants into a sputtered ZrO2 layer [67]. In the latter study, the rise of multilevel switching was attributed to the filamentary conduction while the formation and destruction of otherwise parallel filaments were affected by the distribution of dopants, occurring at different current levels and voltages. Further, multilevel switching was initiated also in sputtered ZrO2 films mounted between TiN electrodes, with nonstoichiometric ZrO2-x layer embedded in the host ZrO2 [68]. In the latter study, the multistate performance was supposedly due to the transformation of interface type switching mechanism to the filamentary conduction, affected by the embedded layer with modified defective stoichiometry.
Regardless of the instabilities and implications of multistate conductivity levels, the switching voltage amplitudes of the structures in the present study exceeded 1–2 V. Yildirim and Pachter [69] have proposed after using an ab initio simulation, that dopants favoring interstitial doping, such as aluminum in ZrO2, could facilitate an electrochemical metallization mechanism in RS cells, whereby the switching becomes manifested by conductive bridge formation due to dissolution of cations of the active electrode. This means that metallic filament is formed by the cations of an active electrode, instead of vacancy driven channel formation. This might, partially, explain the necessity for rather high switching voltages required for such media. Regarding the samples devised in the present study, the unambiguous distinction between dominant switching mechanisms, either metallic or oxygen vacancy driven filamentary or interfacial, may not be solved at this stage. However, most studies carried out earlier on switching oxide media seem to presume and trust in filamentary conduction pathways. Domination of interfacial conduction requires detailed specific studies on the interdependences between electrode materials, areas, and switching directions. Herewith it is worth mentioning, that in very few cases, direct proofs of the physical existence of filaments have been reported, realized using an in situ TEM approach, e.g., in HfO2 [70], TiO2 [71], SiO2 [72], or WOx [73].
Current–voltage curves measured from a 12.2 nm thick reference ZrO2 films grown without aluminum additives revealed appreciably low switching voltage amplitudes remaining between 0.5 and 1.0 V under both voltage polarities (Figure 11a). Analogously, all the devices built on asymmetric ZrxAlyOz-ZrO2 films with ZrxAlyOz buffer layers grown between the bottom TiN electrodes and ZrO2 films (Figure 1b) demonstrated RS behavior (Figure 11c,e). Notably, in all these samples the switching voltages were below 1 V, that is, lower than those needed for ZrO2:Al2O3 media (Figure 1a) described above.
The endurance tests on the 12.2 nm thick reference ZrO2 films revealed marked instability, expressed by a decrease in the ILRS/IHRS values during the first 300 switching cycles (Figure 11b), mainly due to an increase in the conduction currents in the high resistance state during consecutive measurement cycles. However, it was possible to recover the high resistance state value and improve the RS stability during the following 1700 switching cycles (Figure 11b) changing the programming RESET voltage from the starting value of –0.65 V to –0.85 V. After reaching 2000 switching cycles, an abrupt change in the properties of the switching media occurred, as implied by moderate, but clear decrease in the ILRS/IHRS ratio. The decrease could, however, be compensated and the switching ratio restored setting the programming voltage at the RESET event to –1.0 V. The necessity for intentional drift in programming voltage toward higher values was noticed earlier, and the procedure may be regarded as work in process while seeking optimized parameters for stable switching performance. Somewhat speculatively, it may give a hint about that more conductive pathways or filaments can start to form after multiple switching cycles. Alternatively, and possibly, just one relatively large filament might also form, which in any case would require higher electric field strengths for disruption during RESET events. Specific kinetics analysis would require in depth investigations by high speed and precisely time controlled short pulses that would allow one to estimate and model the process [74].
Regardless of the obviously tunable endurance, the ILRS/IHRS values of the un-doped ZrO2 films remained rather low, not exceeding 5. This observation referred to the necessity for combining ZrO2 with another, a higher-band gap oxide such as Al2O3, in order to reduce the conduction currents, in particular, in the high resistivity states. Since the ZrO2:Al2O3 structures with periodical distribution of Al2O3 throughout the RS medium (Figure 1a) enabled switching between dramatically differing resistivity states (Figure 9a and Figure 10) but without acceptable endurance (Figure 9b) and contact yield, the further measurements were conducted on stacks where just the interface between TiN and ZrO2 was modified by deposition of defective ZrxAlyOz buffer layers, presumably able to ensure higher stability of switching.
The samples based on ZrxAlyOz-ZrO2 stacks that contained the interfacial layers grown by applying 3 ZrxAlyOz cycles exhibited more stable counterclockwise switching (Figure 11c) with ILRS/IHRS ratios extending to values of 30–40. Such a difference in the switching direction compared to that obtained in multilayer ZrO2:Al2O3 structures might indicate that the deposition of the ZrxAlyOz layer caused an increase in defect (vacancy) density at the interface between TiN and ZrO2, whereby the Al-rich interface served as a source of oxygen vacancies and conduction electrons. Relatively uniform switching could, after initial device stabilization over about a thousand cycles, easily become extended over 3000 cycles (Figure 11d), similarly to that observed in the reference ZrO2 (Figure 11b). The latter is indicative of the necessity for the application of work-in cycle sequences after formation pulses, before the reliable switching process can be induced. It is, thus, worth underlining, that counterclockwise switching direction dominated in all ZrxAlyOz-ZrO2 stacks as well as in reference ZrO2 film (Figure 11), whereas in the case of periodically layered ZrO2:Al2O3 films the clockwise switching was established (Figure 9 and Figure 10). Reasons for the changes in the switching direction may not yet be unambiguously determined and explained at this stage of studies. Nevertheless, one might suppose that, in the case of ZrxAlyOz-ZrO2 stacks, the ZrxAlyOz buffer layers have indeed served as reservoirs for defects, promoting channeling electronic currents under positive bias first, whereas in the case of periodically layered ZrO2:Al2O3 media, the surface of the deposited film below the top metal electrode contact might have remained and regarded as the most defective junction of the whole structure.
Studies of devices containing 12–15 nm thick ZrxAlyOz-ZrO2 dielectrics, based on interfacial ZrxAlyOz layers deposited using 5 ZrCl4–Al(CH3)3–H2O cycles on TiN followed by 80 ALD cycles of ZrO2, revealed increased initial stability of the switching voltages (Figure 11e). The ILRS/IHRS values reaching 7–10 were recorded for these devices during the first tens of the switching cycles. Along with the further endurance tests, the ILRS/IHRS ratio fluently decreased down to two because of increase in the high-resistive state current during the following 3500 switching cycles (Figure 11f). After changing the programming voltage of the RESET event from –1.0 to –1.2 V, the initial ILRS/IHRS ratio was restored and even stabilized for further switching cycles, analogously to the performance of ILRS/IHRS observed in the ZrO2-based reference device (Figure 11b).
In regard with the literature reporting the capacitive parameters of ZrO2-based thin films, a reader could attend a recent review paper by Xie et al. [54] and references therein. The results reported to date allow one to conclude that the permittivity of capacitive ZrO2-based cells can vary between 7 and 37, depending on the processing parameters, thickness, structural order, and dopant metal in ZrO2, together with the electrode materials selected to complete the cell structure. Regarding the parameters characteristic of ZrO2-based RRAM cells, including doped media, Panda and Tseng [8] have listed a considerable number of case studies, although without aluminum as an alternative dopant metal. Effects of different ionic dopants in ZrO2 based RS media have been studied theoretically [75,76], without consideration of Al. Recently, the influence of aluminum dopant in ZrO2 has also been investigated [77] in accord with density functional theory, revealing that the defect formation energy in ZrO2 should decrease due to the doping effect. Studies on RS behavior in aluminum (oxide)-doped ZrO2 media have so far evidently been rather scarce. In Table 3, nevertheless, some basic RS parameters are represented, concerning switching cell structures built on ZrO2 films supported by Al2O3 layers, together with selected devices containing un-doped ZrO2.
The results have implied that the low- to high-resistivity state ratio in ZrO2-based dielectrics along with the decrease in the thickness of the thin solid films, could be increased dramatically after alternate deposition of ZrO2 and Al2O3 constituent layers. At the same time, the repeatability of the switching, expressed by the endurance cycle number, may inversely relate to the switching reliability between highly insulating and highly conducting states of the solid medium. Therefore, trade-off between endurance and feasibly achievable composition of the structured medium should further be targeted.
Notably, the results obtained from the ZrO2 films grown on interfacial ZrxAlyOz layers between bottom TiN electrodes and the host media, implied possibilities to improve the switching behavior in terms of lowering the conductivity in the high-resistive states, compared to that in the ZrO2 films deposited directly on TiN, and providing stabilized switching, compared to both periodically layered ZrO2:Al2O3 structures and reference ZrO2 films. Notably, the ILRS/IHRS instability and decrease during the endurance test could be suppressed by modifications of the programming voltage in real time.

4. Conclusions

In the present study, zirconium–aluminum oxide thin films were grown by ALD and mounted between bottom TiN and top Ti electrode layers. Multilayered ZrO2:Al2O3 films, containing aluminum oxide up to 2 at.%, were ordered as nanocrystalline metastable tetragonal or cubic polymorph of ZrO2, in their as-deposited states. In double-layered ZrxAlyOz-ZrO2 stacks, where the ZrxAlyOz buffer layers were grown by applying 3–5 ALD cycles comprising successive ZrCl4 and Al(CH3)3 pulses followed by an H2O pulse, the 10–15 nm thick ZrO2 layers deposited on top of ZrxAlyOz were all crystallized in the tetragonal or cubic phase. Notably, the crystallinity of ZrO2 implied a tendency to enhance with increasing number of ALD cycles used for the deposition of the ZrxAlyOz buffer layer. The relative permittivity increased up to 36 and its instability at low measurement frequencies due to the dispersion effects decreased with increasing crystallinity in both multilayer ZrO2:Al2O3 and double-layer ZrxAlyOz-ZrO2 stacks.
Resistive switching behavior was awoken in both Ti/ZrO2:Al2O3/TiN and Ti/ZrO2-ZrxAlyOz/TiN devices. In the periodically layered ZrO2:Al2O3, the conduction currents in the initial high resistance state could remain at the nanoampere level, resulting in remarkably high ILRS/IHRS ratios reaching 100,000. Although the endurance of such a device reached few tens of cycles, the stability of the switching was appreciably uniform before collapsing. Significantly, upon engineering the structure of the medium by shortening the period of alternating metal oxides, thus increasing the number of Al2O3 dopant layers and aluminum content together with structural disorder in the RS medium, multilevel switching could be caused. In the double-layered ZrxAlyOz-ZrO2 films with ILRS/IHRS ratios in the range of 10–40, the endurance characteristics were markedly improved, extending over 5000 cycles. Interestingly, in the case of the ILRS/IHRS ratios either fluctuating or fluently decreasing during the endurance tests, one could restore and stabilize the low to high-resistivity state ratio by increasing the programming voltage amplitude in real time. An important result of this work was the demonstration that the RS direction depends on the Al distribution in the dielectric. Further studies could be devoted to the optimization of dopant concentration and related endurance limits, also after application of post-deposition heat treatments with optimum duration at selected temperatures.

Author Contributions

Conceptualization, J.M. and K.K.; methodology, J.M., T.J., A.K. and T.K.; software, J.M.; investigation, J.M., A.T. (Aivar Tarre), T.D.V., T.K., J.K., H.M., A.K. and P.R.; resources, K.K. and A.T. (Aile Tamm); data curation, J.M. and T.D.V.; writing—original draft preparation, J.M., K.K., T.K. and J.K.; writing—review and editing, J.A.; supervision, A.T. (Aile Tamm), K.K. and J.A.; project administration, K.K.; funding acquisition, A.T. (Aile Tamm) and K.K. All authors have read and agreed to the published version of the manuscript.

Funding

The work was partially supported by the European Regional Development Fund projects “Emerging orders in quantum and nanomaterials” (TK134) and “Advanced materials and high-technology devices for sustainable energetics, sensorics and nanoelectronics” (TK141), and Estonian Research Agency (PRG753).

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Data is contained within the article.

Acknowledgments

Chinedu Henry Ofoegbu is thanked for his assistance in capacitance measurements. The authors are indebted to the European Regional Development Fund project “Centre of nanomaterials technologies and research” (NAMUR+, Project No. 2014–2020.4.01.16-0123).

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Kil, D.-S.; Song, H.-S.; Lee, K.-J.; Hong, K.; Kim, J.-H.; Park, K.-S.; Yeom, S.-J.; Roh, J.-S.; Kwak, N.-J.; Sohn, H.-C.; et al. Development of new TiN/ZrO2/Al2O3/ZrO2/TiN capacitors extendable to 45 nm generation DRAMs replacing HfO2 based dielectrics. In Proceedings of the IEEE 2006 Symposium on VLSI Technology, Honolulu, HI, USA, 13–15 June 2006. [Google Scholar] [CrossRef]
  2. Cho, H.J.; Kim, Y.D.; Park, D.S.; Lee, E.; Park, C.H.; Jang, J.S.; Lee, K.B.; Kim, H.W.; Ki, Y.J.; Han, I.K.; et al. New TIT capacitor with ZrO2/Al2O3/ZrO2 dielectrics for 60 nm and below DRAMs. Solid State Electr. 2007, 51, 1529–1533. [Google Scholar] [CrossRef]
  3. Martin, D.; Grube, M.; Weinreich, W.; Müller, J.; Weber, W.M.; Schröder, U.; Riechert, H.; Mikolajick, T. Mesoscopic analysis of leakage current suppression in ZrO2/Al2O3/ZrO2 nano-laminates. J. Appl. Phys. 2013, 113, 194103. [Google Scholar] [CrossRef]
  4. Pešić, M.; Knebel, S.; Cho, K.; Jung, C.; Chang, J.; Lim, H.; Kolomiiets, N.; Afanas’ev, V.V.; Mikolajick, T.; Schroeder, U. Conduction barrier offset engineering for DRAM capacitor scaling. Solid-State Electr. 2016, 115, 133–139. [Google Scholar] [CrossRef]
  5. Zahoor, F.; Zainal, T.; Zulkifli, A.; Khanday, F.A. Resistive random access memory (RRAM): An overview of materials, switching mechanism, performance, multilevel cell (mlc) storage, modeling, and applications. Nanoscale Res. Lett. 2020, 15, 90. [Google Scholar] [CrossRef] [PubMed]
  6. Sun, K.; Chen, J.; Yan, X. The future of memristors: Materials engineering and neural networks. Adv. Funct. Mater. 2021, 31, 2006773. [Google Scholar] [CrossRef]
  7. Shi, T.; Wang, R.; Wu, Z.; Sun, Y.; An, J.; Liu, Q. A review of resistive switching devices: Performance, improvement, characterization, and applications. Small Struct. 2021, 2, 2000109. [Google Scholar] [CrossRef]
  8. Panda, D.; Tseng, T.-Y. Growth, dielectric properties, and memory device applications of ZrO2 thin films. Thin Solid Films 2013, 531, 1–20. [Google Scholar] [CrossRef]
  9. Awais, M.N.; Kim, H.C.; Doh, Y.H.; Choi, K.H. ZrO2 flexible printed resistive (memristive) switch through electrohydrodynamic printing process. Thin Solid Films 2013, 536, 308–312. [Google Scholar] [CrossRef]
  10. Kärkkänen, I.; Shkabko, A.; Heikkilä, M.; Niinistö, J.; Ritala, M.; Leskelä, M.; Hoffmann-Eifert, S.; Waser, R. Study of atomic layer deposited ZrO2 and ZrO2/TiO2 films for resistive switching application. Phys. Status Sol. A 2014, 211, 301–309. [Google Scholar] [CrossRef]
  11. Parreira, P.; Paterson, G.W.; McVitie, S.; MacLaren, D.A. Stability, bistability and instability of amorphous ZrO2 resistive memory devices. J. Phys. D Appl. Phys. 2016, 49, 095111. [Google Scholar] [CrossRef]
  12. Abbas, Y.; Han, I.S.; Sokolov, A.S.; Jeon, Y.-R.; Choi, C. Rapid thermal annealing on the atomic layer deposited zirconia thin film to enhance resistive switching characteristics. J. Mater. Sci. Mater. Electron. 2020, 31, 903–909. [Google Scholar] [CrossRef]
  13. Wei, X.; Huang, H.; Ye, C.; Wei, W.; Zhou, H.; Chen, Y.; Zhang, R.; Zhang, L.; Xia, Q. Exploring the role of nitrogen incorporation in ZrO2 resistive switching film for enhancing the device performance. J. Alloys Compd. 2019, 775, 1301–1306. [Google Scholar] [CrossRef]
  14. Huang, R.; Yan, X.; Morgan, K.A.; Charlton, M.D.B.; de Groot, C.H. Selection by current compliance of negative and positive bipolar resistive switching behaviour in ZrO2−x/ZrO2 bilayer memory. J. Phys. D Appl. Phys. 2017, 50, 175101. [Google Scholar] [CrossRef]
  15. Liu, H.-C.; Tang, X.-G.; Liu, Q.-X.; Jiang, Y.-P.; Li, W.-H.; Guo, X.-B.; Tang, Z.-H. Bipolar resistive switching behavior and conduction mechanisms of composite nanostructured TiO2 /ZrO2 thin film. Ceram. Int. 2020, 46, 21196. [Google Scholar] [CrossRef]
  16. Fu, L.; Li, Y.; Han, G.; Gao, X.; Chen, C.; Yuan, P. Stable resistive switching characteristics of ZrO2 -based memory device with low-cost. Microelectron. Eng. 2017, 172, 26–29. [Google Scholar] [CrossRef]
  17. Seidl, H.; Gutsche, M.; Schroeder, U.; Birner, A.; Hecht, T.; Jakschik, S.; Luetzen, J.; Kerber, M.; Kudelka, S.; Popp, T.; et al. A fully integrated Al2O3 trench capacitor DRAM for sub-100 nm technology. In Proceedings of the IEEE Digest: International Electron Devices Meeting, San Francisco, CA, USA, 8–11 December 2002. [Google Scholar] [CrossRef]
  18. Vishwanath, S.K.; Woo, H.; Jeon, S. Enhancement of resistive switching properties in Al2O3 bilayer-based atomic switches: Multilevel resistive switching. Nanotechnology 2018, 29, 235202. [Google Scholar] [CrossRef]
  19. Huang, X.-D.; Li, Y.; Li, H.-Y.; Xue, K.-H.; Wang, X.; Miao, X.-S. Forming-free, fast, uniform, and high endurance resistive switching from cryogenic to high temperatures in W/AlOx/Al2O3/Pt bilayer memristor. IEEE Electron Device Lett. 2020, 41, 549–552. [Google Scholar] [CrossRef]
  20. Jang, J.; Choi, H.-H.; Paik, S.H.; Kim, J.K.; Chung, S.; Park, J.H. Highly improved switching properties in flexible aluminum oxide resistive memories based on a multilayer device structure. Adv. Electron. Mater. 2018, 4, 1800355. [Google Scholar] [CrossRef]
  21. Kim, S.; Choi, Y.-K. Resistive switching of aluminum oxide for flexible memory. Appl. Phys. Lett. 2008, 92, 223508. [Google Scholar] [CrossRef]
  22. Mahata, C.; Kang, M.; Kim, C. Multi-level analog resistive switching characteristics in tri-layer HfO2/Al2O3/HfO2 based memristor on ITO electrode. Nanomaterials 2020, 10, 2069. [Google Scholar] [CrossRef]
  23. Castán, H.; Dueñas, S.; Kukli, K.; Kemell, M.; Ritala, M.; Leskelä, M. Study of the influence of the dielectric composition of Al/Ti/ZrO2:Al2O3/TiN/Si/Al structures on the resistive switching behavior for memory applications. ECS Trans. 2018, 85, 143. [Google Scholar] [CrossRef]
  24. Kukli, K.; Kemell, M.; Castán, H.; Dueñas, S.; Seemen, H.; Rähn, M.; Link, J.; Stern, R.; Heikkilä, M.J.; Ritala, M.; et al. Atomic layer deposition and performance of ZrO2-Al2O3 thin films. ECS J. Solid State Sci. Technol. 2018, 7, P287–P294. [Google Scholar] [CrossRef]
  25. Lin, C.-C.; Su, C.-T.; Chang, C.-L.; Wu, H.-Y. Resistive switching behavior of Al/Al2O3/ZrO2/Al structural device for flexible nonvolatile memory application. IEEE Trans. Magn. 2014, 50, 3000704. [Google Scholar] [CrossRef]
  26. Tsai, T.-L.; Chang, H.-Y.; Lou, J.J.-C.; Tseng, T.-Y. A high performance transparent resistive switching memory made from ZrO2/AlON bilayer structure. Appl. Phys. Lett. 2016, 108, 153505. [Google Scholar] [CrossRef]
  27. Arroval, T.; Aarik, L.; Rammula, R.; Kruusla, V.; Aarik, J. Effect of substrate-enhanced and inhibited growth on atomic layer deposition and properties of aluminum–titanium oxide films. Thin Solid Films 2016, 600, 119–125. [Google Scholar] [CrossRef]
  28. Granneman, E.; Fischer, P.; Pierreux, D.; Terhorst, H.; Zagwijn, P. Batch ALD: Characteristics, comparison with single wafer ALD, and examples. Surf. Coat. Technol. 2017, 201, 8899–8907. [Google Scholar] [CrossRef]
  29. Zagwijn, P.; Verweij, W.; Pierreux, D.; Adjeroud, N.; Bankras, R.; Oosterlaken, E.; Snijders, G.J.; van den Hout, M.; Fischer, P.; Wilhelm, R.; et al. Novel batch titanium nitride CVD process for advanced metal electrodes. ECS Trans. 2008, 13, 459–464. [Google Scholar] [CrossRef]
  30. Lee, S.S.; Baik, J.Y.; An, K.-S.; Suh, Y.D.; Oh, J.-H.; Kim, Y. Reduction of incubation period by employing OH-terminated Si(001) substrates in the atomic layer deposition of Al2O3. J. Phys. Chem. B 2004, 108, 15128–15132. [Google Scholar] [CrossRef]
  31. Aarik, J.; Aidla, A.; Kikas, A.; Käämbre, T.; Rammula, R.; Ritslaid, P.; Uustare, T.; Sammelselg, V. Effects of precursors on nucleation in atomic layer deposition of HfO2. Appl. Surf. Sci. 2013, 230, 292–300. [Google Scholar] [CrossRef]
  32. Zhao, C.; Richard, O.; Bender, H.; Caymax, M.; De Gendt, S.; Heyns, M.; Young, E.; Roebben, G.; Van Der Biest, O.; Haukka, S. Miscibility of amorphous ZrO2–Al2O3 binary alloy. Appl. Phys. Lett. 2002, 80, 2374–2376. [Google Scholar] [CrossRef]
  33. Gehensel, R.J.; Zierold, R.; Schaan, G.; Shang, G.; Petrov, A.Y.; Eich, M.; Blick, R.; Krekeler, T.; Janssen, R.; Furlan, K.P. Improved thermal stability of zirconia macroporous structures via homogeneous aluminum oxide doping and nanostructuring using atomic layer deposition. J. Eur. Ceram. Soc. 2021, 41, 4302–4312. [Google Scholar] [CrossRef]
  34. Ravichandran, A.T.; Pushpa, K.C.S.; Ravichandran, K.; Karthika, K.; Nagabhushana, B.M.; Mantha, S.; Swaminathan, K. Effect of Al doping on the structural and optical properties of ZrO2 nanopowders synthesized using solution combustion method. Superlattices Microstruct. 2014, 75, 533–542. [Google Scholar] [CrossRef]
  35. Balakrishnan, G.; Sairam, T.N.; Reddy, V.R.; Kuppusami, P.; Song, J.I. Microstructure and optical properties of Al2O3/ZrO2 nano multilayer thin films prepared by pulsed laser deposition. Mater. Chem. Phys. 2013, 140, 60–65. [Google Scholar] [CrossRef]
  36. Århammar, C.; Araújo, C.M.; Ahuja, R. Energetics of Al doping and intrinsic defects in monoclinic and cubic zirconia: First-principles calculations. Phys. Rev. B 2009, 80, 115208. [Google Scholar] [CrossRef]
  37. Zhu, Y.-L.; Xue, K.-H.; Cheng, X.-M.; Qiao, C.; Yuan, J.-H.; Li, L.-H.; Miao, X.-S. Uniform and robust TiN/HfO2/Pt memristor through interfacial Al-doping engineering. Appl. Surf. Sci. 2021, 550, 149274. [Google Scholar] [CrossRef]
  38. Ismail, M.; Mahata, C.; Kim, S. Forming-free Pt/Al2O3/HfO2/HfAlOx/TiN memristor with controllable multilevel resistive switching and neuromorphic characteristics for artificial synapse. J. Alloys Compd. 2021, 892, 162141. [Google Scholar] [CrossRef]
  39. Fairley, N. CasaXPS: Processing Software for XPS, AES, SIMS and More (Casa Software Ltd., 2018). Available online: http://www.casaxps.com (accessed on 28 January 2021).
  40. Tanuma, S.; Powell, C.J.; Penn, D.R. Calculations of electron inelastic mean free paths. V. Data for 14 organic compounds over the 50–2000 eV range. Surf. Interf. Anal. 1994, 21, 165–176. [Google Scholar] [CrossRef]
  41. Kukli, K.; Ritala, M.; Aarik, J.; Uustare, T.; Leskelä, M. Influence of growth temperature on properties of zirconium dioxide films grown by atomic layer deposition. J. Appl. Phys. 2002, 92, 1833–1840. [Google Scholar] [CrossRef]
  42. Ritala, M.; Leskelä, M. Zirconium dioxide thin films deposited by ALE using zirconium tetrachloride as precursor. Appl. Surf. Sci. 1994, 75, 333–340. [Google Scholar] [CrossRef]
  43. Siimon, H.; Aarik, J. Thickness profiles of thin films caused by secondary reactions in flow-type atomic layer deposition reactors. J. Phys. D Appl. Phys. 1997, 30, 1725–1728. [Google Scholar] [CrossRef]
  44. Arroval, T.; Aarik, L.; Rammula, R.; Aarik, J. Growth of TixAl1−xOy films by atomic layer deposition using successive supply of metal precursors. Thin Solid Films 2015, 591, 276–284. [Google Scholar] [CrossRef]
  45. Aarik, J.; Aidla, A.; Mändar, H.; Uustare, T.; Sammelselg, V. Growth kinetics and structure formation of ZrO2 thin films in chloride-based atomic layer deposition process. Thin Solid Films 2002, 408, 97–103. [Google Scholar] [CrossRef]
  46. Han, J.H.; Gao, G.; Widjaja, Y.; Garfunkel, E.; Musgrave, C.B. A quantum chemical study of ZrO2 atomic layer deposition growth reactions on the SiO2 surface. Surf. Sci. 2004, 550, 199–212. [Google Scholar] [CrossRef]
  47. Jeloaica, L.; Estève, A.; Rouhani, M.D.; Estève, D. Density functional theory study of HfCl4, ZrCl4, and Al(CH3)3 decomposition on hydroxylated SiO2: Initial stage of high-k atomic layer deposition. Appl. Phys. Lett. 2003, 83, 542–544. [Google Scholar] [CrossRef]
  48. Rawat, A.; Meer, M.; Surana, V.K.; Pendem, V.; Garigapati, N.S.; Yadav, Y.; Ganguly, S.; Saha, D. Thermally grown TiO2 and Al2O3 for GaN-based MOS-HEMTs. IEEE Trans. Electron Devices 2018, 65, 3725–3731. [Google Scholar] [CrossRef]
  49. Lee, J.S.; Kim, H.S.; Park, N.-K.; Lee, T.J.; Kang, M. Low temperature synthesis of α-alumina from aluminum hydroxide hydrothermally synthesized using [Al(C2O4)x(OH)y] complexes. Chem. Eng. J. 2013, 230, 351–360. [Google Scholar] [CrossRef]
  50. Crist, B.V. Handbooks of Monochromatic XPS Spectra, Volume 2—Commercially Pure Binary Oxides and a few Common Carbonates and Hydroxides; XPS International LLC.: Mountain View, CA, USA, 2005; Available online: https://xpslibrary.com/wp-content/uploads/----PDFs/Vol-2-XPS-of-Binary-Oxides.pdf (accessed on 12 January 2022).
  51. Tao, N.; Liu, J.; Xu, Y.; Feng, Y.; Wang, Y.; Liu, W.; Wang, H.; Lv, J. Highly selective and stable ZrO2–Al2O3 for synthesis of dimethyl carbonate in reactive distillation. Chem. Pap. 2020, 74, 3503–3515. [Google Scholar] [CrossRef]
  52. Zhao, X.; Vanderbilt, D. Phonons and lattice dielectric properties of zirconia. Phys. Rev. B 2002, 65, 075105. [Google Scholar] [CrossRef] [Green Version]
  53. Vanderbilt, D.; Zhao, X.; Ceresoli, D. Structural and dielectric properties of crystalline and amorphous ZrO2. Thin Solid Films 2005, 486, 125–128. [Google Scholar] [CrossRef]
  54. Xie, J.; Zhu, Z.; Tao, H.; Zhou, S.; Liang, Z.; Li, Z.; Yao, R.; Wang, Y.; Ning, H.; Peng, J. Research progress of high dielectric constant zirconia-based materials for gate dielectric application. Coatings 2020, 10, 698. [Google Scholar] [CrossRef]
  55. Krupka, J.; Geyer, R.G.; Kuhn, M.; Hinken, J.H. Dielectric properties of single crystals of Al2O3, LaAlO3, NdGaO3, SrTiO3, and MgO at cryogenic temperatures. IEEE Trans. Microw. Theor. Tech. 1994, 42, 1886–1890. [Google Scholar] [CrossRef]
  56. Groner, M.D.; Elam, J.W.; Fabreguette, F.H.; George, S.M. Electrical characterization of thin Al2O3 films grown by atomic layer deposition on silicon and various metal substrates. Thin Solid Films 2002, 413, 186–197. [Google Scholar] [CrossRef]
  57. Wu, J.Z.; Acharya, J.; Goul, R. In vacuo atomic layer deposition and electron tunneling characterization of ultrathin dielectric films for metal/insulator/metal tunnel junctions. J. Vac. Sci. Technol. A 2020, 38, 040802. [Google Scholar] [CrossRef]
  58. Link, G.; Heidinger, R. Dielectric property measurements of ZrO2-strengthened Al2O3. J. Appl. Phys. 1997, 81, 3257–3262. [Google Scholar] [CrossRef]
  59. Zhu, M.; Chen, P.; Fu, R.K.Y.; Liu, W.; Lin, C.; Chu, P.K. Microstructure and electrical properties of Al2O3–ZrO2 composite films for gate dielectric applications. Thin Solid Films 2005, 476, 312–316. [Google Scholar] [CrossRef]
  60. Peng, J.; Wei, J.; Zhu, Z.; Ning, H.; Cai, W.; Lu, K.; Yao, R.; Tao, H.; Zheng, Y.; Lu, X. Properties-adjustable alumina-zirconia nanolaminate dielectric fabricated by spin-coating. Nanomaterials 2017, 7, 419. [Google Scholar] [CrossRef] [Green Version]
  61. Liang, Z.; Zhou, S.; Cai, W.; Fu, X.; Ning, H.; Chen, J.; Yuan, W.; Zhu, Z.; Yao, R.; Peng, J. Zirconium-aluminum-oxide dielectric layer with high dielectric and relatively low leakage prepared by spin-coating and the application in thin-film transistor. Coatings 2020, 10, 282. [Google Scholar] [CrossRef] [Green Version]
  62. Chen, X.; Cheng, Y.; Wu, K.; Meng, Y.; Wu, S. Calculation of dielectric constant of two-phase disordered composites by using FEM. In Proceedings of the Conference Record of the 2008 IEEE International Symposium on Electrical Insulation, Vancouver, BC, Canada, 9–12 June 2008. [Google Scholar] [CrossRef]
  63. Sihvola, A. Homogenization principles and effect of mixing on dielectric behavior. Photonics Nanostruct. Fundam. Appl. 2013, 11, 364–373. [Google Scholar] [CrossRef] [Green Version]
  64. Waggoner, T.; Triska, J.; Hoshino, K.; Wager, J.F.; Conley, J.F., Jr. Zirconium oxide-aluminum oxide nanolaminate gate dielectrics for amorphous oxide semiconductor thin-film transistors. J. Vac. Sci. Technol. B 2011, 29, 04D115. [Google Scholar] [CrossRef] [Green Version]
  65. Zhu, B.; Wu, X.; Liu, W.-J.; Ding, S.-J.; Zhang, D.W.; Fan, Z. Dielectric enhancement of atomic layer-deposited Al2O3/ZrO2/Al2O3 MIM capacitors by microwave annealing. Nanoscale Res. Lett. 2019, 14, 53. [Google Scholar] [CrossRef]
  66. Walczyk, C.; Wenger, C.; Sohal, R.; Lukosius, M.; Fox, A.; Dąbrowski, J.; Wolansky, D.; Tillack, B.; Müssig, H.-J.; Schroeder, T. Pulse-induced low-power resistive switching in HfO2 metal-insulator-metal diodes for nonvolatile memory applications. J. Appl. Phys. 2009, 105, 114103. [Google Scholar] [CrossRef]
  67. Liu, M.; Abid, Z.; Wang, W.; He, X.; Liu, Q.; Guan, W. Multilevel resistive switching with ionic and metallic filaments. Appl. Phys. Lett. 2009, 94, 233106. [Google Scholar] [CrossRef] [Green Version]
  68. Huang, R.; Yan, X.; Ye, S.; Kashtiban, R.; Beanland, R.; Morgan, K.A.; Charlton, M.D.B.; de Groot, C.H. Compliance-free ZrO2/ZrO2 − x/ZrO2 resistive memory with controllable interfacial multistate switching behaviour. Nanoscale Res. Lett. 2017, 12, 384. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  69. Yildirim, H.; Pachter, R. Extrinsic dopant effects on oxygen vacancy formation energies in ZrO2 with implication for memristive device performance. ACS Appl. Electron. Mater. 2019, 1, 467–477. [Google Scholar] [CrossRef]
  70. Pey, K.L.; Thamankar, R.; Sen, M.; Bosman, M.; Raghavan, N.; Shubhakar, K. Understanding the switching mechanism in RRAM using in-situ TEM. In Proceedings of the 2016 IEEE Silicon Nanoelectronics Workshop (SNW), Honolulu, HI, USA, 12–13 June 2016. [Google Scholar] [CrossRef]
  71. Kwon, D.-H.; Kim, K.M.; Jang, J.H.; Jeon, J.M.; Lee, M.H.; Kim, G.H.; Li, X.-S.; Park, G.-S.; Lee, B.; Han, S.; et al. Atomic structure of conducting nanofilaments in TiO2 resistive switching memory. Nat. Nanotechnol. 2010, 5, 148–153. [Google Scholar] [CrossRef]
  72. Yang, Y.; Gao, P.; Gaba, S.; Chang, T.; Pan, X.; Lu, W. Observation of conducting filament growth in nanoscale resistive memories. Nat. Commun. 2012, 3, 732. [Google Scholar] [CrossRef]
  73. Arita, M.; Takahashi, A.; Ohno, Y.; Nakane, A.; Tsurumaki-Fukuchi, A.; Takahashi, Y. Switching operation and degradation of resistive random access memory composed of tungsten oxide and copper investigated using in-situ TEM. Sci. Rep. 2015, 5, 17103. [Google Scholar] [CrossRef] [Green Version]
  74. Rodriguez-Fernandez, A.; Cagli, C.; Perniola, L.; Suñé, J.; Miranda, E. Identification of the generation/rupture mechanism of filamentary conductive paths in ReRAM devices using oxide failure analysis. Microelectron. Reliab. 2017, 76–77, 178–183. [Google Scholar] [CrossRef]
  75. Yan, H.; Li, J.; Guo, Y.; Song, Q.; Han, J.; Yang, F. A study on dopant selection for ZrO2 based RRAM from density functional theory. Phys. B 2021, 612, 412915. [Google Scholar] [CrossRef]
  76. Zhang, H.; Gao, B.; Sun, B.; Chen, G.; Zeng, L.; Liu, L.; Liu, X.; Lu, J.; Han, R.; Kang, J.; et al. Ionic doping effect in ZrO2 resistive switching memory. Appl. Phys. Lett. 2010, 96, 123502. [Google Scholar] [CrossRef]
  77. Hussain, F.; Imran, M.; Arif Khalil, R.M.A.; Sattar, M.A.; Niaz, N.A.; Rana, A.M.; Ismail, M.; Khera, E.A.; Rasheed, U.; Mumtaz, F.; et al. A first-principles study of Cu and Al doping in ZrO2 for RRAM device applications. Vacuum 2019, 168, 108842. [Google Scholar] [CrossRef]
Coatings 12 00431 g001 550
Figure 1. Schematic representation of (a) periodical ZrO2:Al2O3 structures and (b) ZrO2 films grown on buffer ZrxAlyOz layers on bottom TiN electrodes. The component-material layer thicknesses are not to scale.
Figure 1. Schematic representation of (a) periodical ZrO2:Al2O3 structures and (b) ZrO2 films grown on buffer ZrxAlyOz layers on bottom TiN electrodes. The component-material layer thicknesses are not to scale.
Coatings 12 00431 g001
Coatings 12 00431 g002 550
Figure 2. (a) Survey, (b) O1s, (c) Al 2p and (d) Zr 3d, X-ray photoelectron spectra of a ZrxAlyOz film grown using cycle sequence of 80 × (ZrCl4 + Al(CH3)3 + H2O), and reference ZrO2 and Al2O3 films.
Figure 2. (a) Survey, (b) O1s, (c) Al 2p and (d) Zr 3d, X-ray photoelectron spectra of a ZrxAlyOz film grown using cycle sequence of 80 × (ZrCl4 + Al(CH3)3 + H2O), and reference ZrO2 and Al2O3 films.
Coatings 12 00431 g002
Coatings 12 00431 g003 550
Figure 3. GIXRD patterns from (1) pure ZrO2 and (2,3) periodically layered ZrO2:Al2O3 films deposited by applying single Al2O3 deposition cycles between successive ZrO2 cycles in ratios of (2) 24:1 and (3) 4:1. Patterns were measured from the films supplied with top Au/Ti electrodes. Miller indexes are assigned to the cubic phase of ZrO2, in addition to the reflections attributed to top metal and bottom TiN electrode films.
Figure 3. GIXRD patterns from (1) pure ZrO2 and (2,3) periodically layered ZrO2:Al2O3 films deposited by applying single Al2O3 deposition cycles between successive ZrO2 cycles in ratios of (2) 24:1 and (3) 4:1. Patterns were measured from the films supplied with top Au/Ti electrodes. Miller indexes are assigned to the cubic phase of ZrO2, in addition to the reflections attributed to top metal and bottom TiN electrode films.
Coatings 12 00431 g003
Coatings 12 00431 g004 550
Figure 4. BF STEM images of the ZrO2:Al2O3 film deposited using the ALD cycle sequence of 6 × (1 × Al2O3 + 24 × ZrO2). The images were taken under different magnifications implied by the scale bars. The locations of substrates, electrodes and ZrO2:Al2O3 medium are indicated by corresponding labels.
Figure 4. BF STEM images of the ZrO2:Al2O3 film deposited using the ALD cycle sequence of 6 × (1 × Al2O3 + 24 × ZrO2). The images were taken under different magnifications implied by the scale bars. The locations of substrates, electrodes and ZrO2:Al2O3 medium are indicated by corresponding labels.
Coatings 12 00431 g004
Coatings 12 00431 g005 550
Figure 5. Elemental composition mapping of an Au/Ti/ZrO2:Al2O3/TiN stack containing ZrO2:Al2O3 film deposited using the ALD cycle sequence of 6 × (1 × Al2O3 + 24 × ZrO2). The chemical elements (Au [orange], Ti [green], Zr [blue], Al [red] and O [magenta]) as detected and mapped are indicated by the corresponding labels. Note that the high intensity of the Al signal in the Au contact region is due to the X-ray continuum. The intensity of the X-ray continuum is higher in regions with high atomic numbers. Although the X-ray continuum contributes to all maps, the effect is especially pronounced in the Al map due to low concentration of Al and, as a result, low signal/background ratio.
Figure 5. Elemental composition mapping of an Au/Ti/ZrO2:Al2O3/TiN stack containing ZrO2:Al2O3 film deposited using the ALD cycle sequence of 6 × (1 × Al2O3 + 24 × ZrO2). The chemical elements (Au [orange], Ti [green], Zr [blue], Al [red] and O [magenta]) as detected and mapped are indicated by the corresponding labels. Note that the high intensity of the Al signal in the Au contact region is due to the X-ray continuum. The intensity of the X-ray continuum is higher in regions with high atomic numbers. Although the X-ray continuum contributes to all maps, the effect is especially pronounced in the Al map due to low concentration of Al and, as a result, low signal/background ratio.
Coatings 12 00431 g005
Coatings 12 00431 g006 550
Figure 6. Grazing incidence X-ray diffraction patterns from asymmetrically deposited ZrxAlyOz-ZrO2 films, comprising successive pulses of ZrCl4, Al(CH3)3 and H2O, applied N times (N indicated by labels) directly on TiN bottom electrode surface and followed by 80 ZrO2 deposition cycles. Labels also indicate the Miller indexes attributable to either cubic or tetragonal phases of ZrO2, as well as those of the cubic TiN substrate layer. The bottom pattern, enclosed for comparison, represents that from X-ray amorphous reference film grown with 80 ALD cycles including successive pulses of ZrCl4, Al(CH3)3 and H2O.
Figure 6. Grazing incidence X-ray diffraction patterns from asymmetrically deposited ZrxAlyOz-ZrO2 films, comprising successive pulses of ZrCl4, Al(CH3)3 and H2O, applied N times (N indicated by labels) directly on TiN bottom electrode surface and followed by 80 ZrO2 deposition cycles. Labels also indicate the Miller indexes attributable to either cubic or tetragonal phases of ZrO2, as well as those of the cubic TiN substrate layer. The bottom pattern, enclosed for comparison, represents that from X-ray amorphous reference film grown with 80 ALD cycles including successive pulses of ZrCl4, Al(CH3)3 and H2O.
Coatings 12 00431 g006
Coatings 12 00431 g007 550
Figure 7. BF STEM image of ZrxAlyOz-ZrO2 film deposited using the cycle sequence of 5 × (ZrCl4 + Al(CH3)3 + H2O) + 80 × ZrO2. Substrate, bottom electrode and component layers are denoted by labels.
Figure 7. BF STEM image of ZrxAlyOz-ZrO2 film deposited using the cycle sequence of 5 × (ZrCl4 + Al(CH3)3 + H2O) + 80 × ZrO2. Substrate, bottom electrode and component layers are denoted by labels.
Coatings 12 00431 g007
Coatings 12 00431 g008 550
Figure 8. Relative dielectric permittivity versus measurement frequency curves for (a) periodically layered ZrO2:Al2O3 stacks compared to those of reference ZrO2 and Al2O3 films and (b) stacks deposited by applying 3, 4 or 5 ALD cycles of interfacial ZrxAlyOz followed by 80 ALD cycles of ZrO2. The growth cycle sequences are indicated by the labels pointing via arrows to the corresponding dispersion curves. For the thicknesses and composition data, see Table 1 and Table 2.
Figure 8. Relative dielectric permittivity versus measurement frequency curves for (a) periodically layered ZrO2:Al2O3 stacks compared to those of reference ZrO2 and Al2O3 films and (b) stacks deposited by applying 3, 4 or 5 ALD cycles of interfacial ZrxAlyOz followed by 80 ALD cycles of ZrO2. The growth cycle sequences are indicated by the labels pointing via arrows to the corresponding dispersion curves. For the thicknesses and composition data, see Table 1 and Table 2.
Coatings 12 00431 g008
Coatings 12 00431 g009 550
Figure 9. (a) Current–voltage and (b) endurance characteristics measured in resistive switching regime on an Au/Ti/ZrO2:Al2O3/TiN structure containing ZrO2:Al2O3 films grown to a thickness of 14.3 nm using the deposition cycle sequence of 6 × (1 × Al2O3 + 24 × ZrO2).
Figure 9. (a) Current–voltage and (b) endurance characteristics measured in resistive switching regime on an Au/Ti/ZrO2:Al2O3/TiN structure containing ZrO2:Al2O3 films grown to a thickness of 14.3 nm using the deposition cycle sequence of 6 × (1 × Al2O3 + 24 × ZrO2).
Coatings 12 00431 g009
Coatings 12 00431 g010 550
Figure 10. (a) Current–voltage characteristics and (b) endurance characteristic of resistive switching sequences measured in Au/Ti/ZrO2:Al2O3/TiN structure containing 14.3 nm thick ZrO2:Al2O3 film grown using the deposition cycle sequence of 30 × (1 × Al2O3+ 4 × ZrO2). Numbered LRS and HRS assign low and high resistivity levels, respectively, recorded at 0.2 V during two-level switching cycles before RESET events depicted in panel (a). Arrows in panel (b) are guides to eye, following the order of recording currents at different resistivity levels.
Figure 10. (a) Current–voltage characteristics and (b) endurance characteristic of resistive switching sequences measured in Au/Ti/ZrO2:Al2O3/TiN structure containing 14.3 nm thick ZrO2:Al2O3 film grown using the deposition cycle sequence of 30 × (1 × Al2O3+ 4 × ZrO2). Numbered LRS and HRS assign low and high resistivity levels, respectively, recorded at 0.2 V during two-level switching cycles before RESET events depicted in panel (a). Arrows in panel (b) are guides to eye, following the order of recording currents at different resistivity levels.
Coatings 12 00431 g010
Coatings 12 00431 g011 550
Figure 11. (a) Current–voltage switching and (b) endurance characteristics of a ZrO2 reference film, compared to (c) current–voltage and (d) endurance behavior of a ZrxAlyOz-ZrO2 stack deposited using 3 ZrxAlyOz and 80 ZrO2 ALD cycles, and (e) current–voltage and (f) endurance test results for a ZrxAlyOz-ZrO2 stack deposited using 3 ZrxAlyOz and 80 ZrO2 ALD cycles. The deposition cycle sequences are indicated by labels.
Figure 11. (a) Current–voltage switching and (b) endurance characteristics of a ZrO2 reference film, compared to (c) current–voltage and (d) endurance behavior of a ZrxAlyOz-ZrO2 stack deposited using 3 ZrxAlyOz and 80 ZrO2 ALD cycles, and (e) current–voltage and (f) endurance test results for a ZrxAlyOz-ZrO2 stack deposited using 3 ZrxAlyOz and 80 ZrO2 ALD cycles. The deposition cycle sequences are indicated by labels.
Coatings 12 00431 g011
Table
Table 1. ALD cycle sequence, thickness, and elemental composition measured by XRF for ZrO2:Al2O3 films designed as periodical multilayers with uniform distribution of Al2O3 layers throughout the thickness of the host ZrO2 medium. The measurement uncertainty ranges for atomic percentages express the differences between measurement results obtained on different locations of samples on substrate holder, due to the lateral thickness and composition profile.
Table 1. ALD cycle sequence, thickness, and elemental composition measured by XRF for ZrO2:Al2O3 films designed as periodical multilayers with uniform distribution of Al2O3 layers throughout the thickness of the host ZrO2 medium. The measurement uncertainty ranges for atomic percentages express the differences between measurement results obtained on different locations of samples on substrate holder, due to the lateral thickness and composition profile.
ALD Cycle SequenceThickness, nmAl, at.%Zr, at.%O, at.%Cl, at.%
6 × (1 × Al2O3 + 24 × ZrO2)14.6 1.8 ± 0.137.1 ± 1.360.8 ± 1.40.39 ± 0.02
30 × (1 × Al2O3 + 4 × ZrO2)14.3 8.9 ± 0.2 28.7 ± 2.862.1 ± 2.60.43 ± 0.02
150 × Al2O320.0 39.9 ± 0.2-60.1 ± 0.2-
120 × ZrO212.2 -37.1 ± 2.162.5 ± 2.00.44 ± 0.0
Table
Table 2. ALD cycle sequence, together with thickness and elemental composition, measured by XRF for ZrxAlyOz-ZrO2 films deposited on TiN substrates, applying first consecutive exposures of substrate surface to metal precursors without oxidizing steps between ZrCl4 and Al(CH3)3 pulses in order to form ZrxAlyOz buffer layer prior to the main ZrO2 layer.
Table 2. ALD cycle sequence, together with thickness and elemental composition, measured by XRF for ZrxAlyOz-ZrO2 films deposited on TiN substrates, applying first consecutive exposures of substrate surface to metal precursors without oxidizing steps between ZrCl4 and Al(CH3)3 pulses in order to form ZrxAlyOz buffer layer prior to the main ZrO2 layer.
ALD Cycle SequencesThickness, nmAl, at.%Zr, at.%O, at.%Cl, at.%
3 × (ZrCl4 + Al(CH3)3 + H2O)
+ 80 × ZrO2		13.00.5 ± 0.427.4 ± 1.471.3 ± 1.00.8 ± 0.4
4 × (ZrCl4 + Al(CH3)3 + H2O)
+ 80 × ZrO2		12.11.3 ± 0.432.1 ± 1.465.8 ± 1.00.8 ± 0.4
5 × (ZrCl4 + Al(CH3)3 + H2O)
+ 80 × ZrO2		14.61.2 ± 0.431.0 ± 1.467.0 ± 1.00.8 ± 0.4
80 × (ZrCl4 + Al(CH3)3 + H2O)10.77.4 ± 0.426.8 ± 1.464.8 ± 1.01.0 ± 0.4
Table
Table 3. Comparison of RS stacks containing ZrO2 or different combinations of ZrO2 with Al2O3 as RS media. Given are also the film growth techniques, thicknesses, and maximum switching voltage range from negative to positive polarity in relation to bottom electrode. For low and high resistivity state ratios as well as amounts of endurance cycles, order of magnitudes implied in the reference studies are indicated. Data not revealed in the reference papers, and thus not available, are denoted by “NA”.
Table 3. Comparison of RS stacks containing ZrO2 or different combinations of ZrO2 with Al2O3 as RS media. Given are also the film growth techniques, thicknesses, and maximum switching voltage range from negative to positive polarity in relation to bottom electrode. For low and high resistivity state ratios as well as amounts of endurance cycles, order of magnitudes implied in the reference studies are indicated. Data not revealed in the reference papers, and thus not available, are denoted by “NA”.
Switching Cell StructureDeposition TechniqueFilm Thickness, nmSwitching Voltage Range, VLRS:HRS Current RatioEndurance, Number of Cycles Reference
Al/Ti/ZrO2-Al2O3/TiN/Si/AlALD276 −1.5–1.5 2NA[23]
Al/Ti/ZrO2-Al2O3/TiN/Si/AlALD41 –1.5–1.2 7.3NA[24]
TiN/ZrO2-Al2O3/Pt/TiSputtering50 −1.5–1.510104[76]
Al/ZrO2/Cu/ZrO2/AlEBE43 −2–1.5 103102[16]
TiN/ZrO2/ZrO2-x/ZrO2/TiNSputtering45 −6–5 102102[68]
ITO/ZrO2/AlON/ITOALD/Sputtering10 −4–2 10104[26]
Al/Al2O3/ZrO2/AlSputteringNA−4.5–2.5 102102[25]
Au/Ti/ZrO2:Al2O3/TiNALD14.6 −2–3 10520This study
Ti/ZrO2/ZrxAlyOz/TiNALD14.6 −1.5–1 103 × 103This study
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Merisalu, J.; Jõgiaas, T.; Viskus, T.D.; Kasikov, A.; Ritslaid, P.; Käämbre, T.; Tarre, A.; Kozlova, J.; Mändar, H.; Tamm, A.; et al. Structure and Electrical Properties of Zirconium-Aluminum-Oxide Films Engineered by Atomic Layer Deposition. Coatings 2022, 12, 431. https://doi.org/10.3390/coatings12040431

AMA Style

Merisalu J, Jõgiaas T, Viskus TD, Kasikov A, Ritslaid P, Käämbre T, Tarre A, Kozlova J, Mändar H, Tamm A, et al. Structure and Electrical Properties of Zirconium-Aluminum-Oxide Films Engineered by Atomic Layer Deposition. Coatings. 2022; 12(4):431. https://doi.org/10.3390/coatings12040431

Chicago/Turabian Style

Merisalu, Joonas, Taivo Jõgiaas, Toomas Daniel Viskus, Aarne Kasikov, Peeter Ritslaid, Tanel Käämbre, Aivar Tarre, Jekaterina Kozlova, Hugo Mändar, Aile Tamm, and et al. 2022. "Structure and Electrical Properties of Zirconium-Aluminum-Oxide Films Engineered by Atomic Layer Deposition" Coatings 12, no. 4: 431. https://doi.org/10.3390/coatings12040431

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop