Subscriber access provided by University of Newcastle, Australia
Article
In situ Control of Oxygen Vacancies in TaOx Thin Films via Plasma-Enhanced Atomic Layer Deposition for Resistive Switching Memory Applications Konstantin Viktorovich Egorov, Dmitry S. Kuzmichev, Pavel Sergeevich Chizhov, Yury Yu. Lebedinskii, Cheol Seong Hwang, and Andrey M. Markeev ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b00778 • Publication Date (Web): 28 Mar 2017 Downloaded from http://pubs.acs.org on March 29, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
In situ Control of Oxygen Vacancies in TaOx Thin Films via Plasma-Enhanced Atomic Layer Deposition for Resistive Switching Memory Applications
Konstantin V. Egorov1, Dmitry S. Kuzmichev1, Pavel S. Chizhov2, Yuri Yu. Lebedinskii1,3, Cheol Seong Hwang4,#, and Andrey M. Markeev1,* 1
Moscow Institute of Physics and Technology, Institutskii Lane 9,141700 Dolgoprudny, Russian Federation
2
Chemistry Department, Moscow State University, Leninskie Gory 1, 119992 Moscow, Russian Federation 3
4
National Research Nuclear University, Moscow Engineering Physics Institute, Kashirskoye Shosse 31, 115409 Moscow, Russian Federation
Department of Materials Science and Engineering and Inter-University Semiconductor Research Center, Seoul National University, Seoul 08826, Republic of Korea
# electronic mail:
[email protected] * electronic mail:
[email protected] Key words: PEALD; TaOx; ReRAM; Hydrogen plasma; Reliability
Abstract The plasma-enhanced atomic layer deposition (PEALD) process using Ta(OC2H5)5 as a Ta precursor and plasma-activated hydrogen as a reactant for the deposition of TaOx films with a controllable concentration of oxygen vacancies (VO) is reported herein. The VO concentration control was achieved by varying the hydrogen volume fraction of the hydrogen-argon mixture in the plasma, allowing the control of the leakage current density in the tantalum oxide films within the range of five orders of magnitude compared with the Ta2O5 film grown via thermal ALD using the identical Ta precursor and H2O. Temperature-dependent current-voltage measurements combined with Poole-Frenkel emission modelling demonstrated that the bulk trap depth decreases with the increasing hydrogen volume fraction, which could be attributed to the increase of the VO concentration. The possible chemical change in the PEALD TaOx films grown under different hydrogen volume fractions was confirmed by the in situ X-ray photoelectron spectroscopy (XPS) measurements of the Ta4f core and valence band spectra. The comparison of the XPS-measured non-stoichiometry and the secondary ion mass spectrometry analysis of the 1 ACS Paragon Plus Environment
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 25
hydrogen content allowed this study to conclude that the non-stoichiometry is largely related to the formation of Ta-VO sites rather than of Ta-H sites. Such oxygen-deficient TaOx layers were studied for application as an oxygen deficient layer in a resistance switching random access memory stack (Ta2O5/TaOx) where the actual switching occurred within the stoichiometric Ta2O5 layer. The bilayer memory stack showed reliable resistance switching up to ~106 switching cycles, whereas the single-layer Ta2O5 memory showed only several hundred switching cycles.
1. Introduction Atomic layer deposition (ALD) enables the thickness control of thin films with feasible atomic accuracy due to its self-saturating nature, and is now a widely accepted thin-film growth process in microelectronics technology, especially for the high-k oxides for the metal oxide semiconductor field effect transistor and for the three-dimensional (3D) capacitors in the dynamic random access memory [1, 2]. The latter application of ALD is inspired by the unprecedented deposition conformality over the 3D structures [3-5]. It is noteworthy that metal oxides grown via ALD are usually highly oxygen-stoichiometric [6-12]. There are some applications, however, where the involvement of non-stoichiometry is desirable, such as oxidebased high-performance sensors, energy storage and conversion, electro- and photocatalysts [1316], and resistance switching random access memory (ReRAM) [17-20]. There are three classes in redox-based ReRAM– electrochemical metallization mechanism (ECM), valence change mechanism (VCM), and thermochemical mechanism (TCM) [21]. ECM type is based on cation drifting from electrochemically active electrode (Ag, Cu) to inert electrode through ion conductor layer (GeSex, Ta2O5, SiO2, WOx). In contrast, VCM is triggered by a field induced migration of anions such as oxygen in oxide layer. Redistribution of oxygen/oxygen vacancies lead to stoichiometry and resistance changing. In TCM, stoichiometry is changed mainly by current-induced increase of the temperature. Among them, VCM-type ReRAM is particularly interesting because it demonstrated highly promising properties, such as sub-1µA programming currents, sub-nanosecond switching, good scalability (down to 10x10 nm2), and high endurance up to 1012 switching cycles [17-24]. While numerous oxide materials have been explored for use as a resistance switching (RS) layer in the past decades, the Ta2O5-based material has become prominent recently due to its high reliability [22-24]. High reliability in terms of RS uniformity, endurance, and data retention is of utmost concern in the fabrication of reliable memory devices. As the RS operation is intimately related with the migration of VO and their aggregation/percolation to form the conducting filament (CF), the control of the VO source and 2 ACS Paragon Plus Environment
Page 3 of 25
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
its interaction with the insulating region of the device are primarily important for the fluent ReRAM operation. Along with this idea, it is highly notable that the most promising ReRAM MIM stacks based on the RS of the Ta2O5 layer included the oxygen-deficient layer of metal oxide (e.g., TaOx or TaO2-x) [22-24], which played the role of a VO source. The oxygen-deficient TaOx layer was obtained via reactive ion sputtering or RF sputtering using a metal tantalum [24] target under an appropriately controlled O2-Ar gas environment to achieve the desired Ta/O ratio of the film [24]. Stoichiometric Ta2O5 may either be deposited on top of the oxygen-deficient layer or achieved by oxidizing the non-stoichiometric layer [23, 24]. It is expected, however, that ReRAM will have a 3D structure with a vertical integration scheme so that it can eventually compete with the already-3D-configured vertical-NAND flash memory [25]. Therefore, the proposed 3D ReRAM architectures require the highly conformal deposition of metal oxide on high-aspect-ratio substrates [26-28], making the sputtering-based deposition techniques, which have low conformality, undesirable, and ALD process is indispensable in this regard. Therefore, there is great interest in finding a method of direct oxygen content control in metal oxides during their ALD growth. A few attempts to achieve oxygen-deficient TiOx and TaOx have been made in previous works [29, 30] through the use of plasma-enhanced ALD (PEALD) with active oxygen as a reactant by varying its partial pressure and/or pulse duration. In such (active) oxygen-based PEALD processes, however, the active oxygen has to carry out rather contradictory tasks: (1) removing the organic ligands; and (2) achieving oxygen deficiency in the growing metal oxide films, which make the precise process control tricky. Thus, in this work, an alternative approach was suggested to achieve a robust ALD process for depositing an oxygen-deficient TaOx film based on plasma-activated hydrogen as the reactant and the alkoxide compound Ta(OC2H5)5 as the Ta precursor, which already has a Ta-O bond in it. In this approach, the critical idea is to remove the C2H5 group by forming volatile C2H6 or C2H5(OH) molecules via the reaction with plasma-activated hydrogen without breaking the relatively stable Ta-O bond in the chemisorbed Ta precursor molecules, whereby the oxide film can be grown even when hydrogen is adopted as the reactant for oxide ALD. Another critical feature of this type of novel reaction pathway is that it is possible to partly reduce the growing oxide in a controllable manner due to the attack of active H to Ta-O bonds, whereby the TaOx film can be grown in a well-controlled manner. The film was combined with a stoichiometric Ta2O5 film, which was grown through another ALD process using the same Ta precursor and H2O as the oxygen source in the thermal ALD mode. The RS performance of the stacked film was compared with that of the single-layer Ta2O5 layer. 3 ACS Paragon Plus Environment
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 25
2. Experimental Procedure Tantalum oxide films were grown in a Picosun R200Adv PEALD tool equipped with a remote RF inductively coupled plasma source. Silicon wafers coated with 20-nm-thick ALD TiN films (TiCl4/NH3 process at T=400oC in thermal ALD mode) were used as the substrates for the tantalum oxide deposition. Tantalum ethoxide (Ta(OC2H5)5) and plasma-activated hydrogen gas were employed as the Ta source and reactant, respectively. Ar and H2 gases were injected into the plasma source region to obtain activated hydrogen atoms. The sum of the H2 (fH2) and Ar (fAr) flow rates was fixed at 130 standard cubic centimeters per min (sccm). The hydrogen volume
fraction
of
the
plasma
gas
mixture
was
characterized
by
the
value
R=(fH2/(fH2+fAr))·100%. The plasma pulse discharge power was fixed at 2500 W. During the ALD, the substrate was maintained at 300oC, and the chamber pressure was ∼5 Torr. The optimized Ta precursor pulse and purge time were 0.5 and 8 sec, respectively, and the H2-plasma pulse and purge time were 12 and 8 sec, respectively, while the R was the major parameter for controlling the VO concentration in the film. Thermal-ALD Ta2O5 films were also grown as the reference samples via the Ta(OC2H5)5/H2O process, at the identical substrate temperature. For thermal ALD, the Ta precursor pulse-purge-H2O pulse-purge time was 0.5-8-0.1-8, respectively, which was confirmed to correspond to a well-saturated ALD condition. The thicknesses of all types of tantalum oxide film were estimated via ellipsometry (SENTECH SE 500adv) using film that were grown on Si for each ALD process. The chemical states of the as-grown tantalum oxide films were analyzed using an in situ Xray photoelectron spectroscopy (XPS, Theta Probe, Thermo Scientific spectrometer) connected through a vacuum transport system with the PEALD tool. Spectra were taken at a pass energy of 50 eV providing a 0.6 eV resolution, estimated from the full width at half maximum (FWHM) of the Au4f7/2 core level. The energy scale of the XPS spectra was calibrated with respect to the Au 4f7/2 line (84.0 eV). The hydrogen contents of the films were evaluated via time-of-flight secondary ion mass spectrometry (TOF-SIMS, 5 IonTOF GmbH spectrometer) using tantalum oxide films grown on Si with known thickness determined by ellipsometry. To evaluate the performances of the films as RS layers, the current density-voltage (J-V) curves and pulse-switching performance were estimated. For these tests, Pt top electrodes (300 µm in diameter, 50 nm in thickness) were deposited through a shadow mask on the oxide films, via magnetron sputtering at room temperature. The J-V data were acquired at temperatures ranging from 20 to 100oC using an Agilent B1500A semiconductor parameter analyzer. The optimized ReRAM-stack structure was TiN (bottom electrode)/TaOx(12nm)/Ta2O5(7nm)/ Pt(top 4 ACS Paragon Plus Environment
Page 5 of 25
electrode), where the TaOx layer was grown via PEALD with R=70%, and the Ta2O5 layer was grown via thermal ALD. The weak oxidation potential of H2O, which was adopted as the oxygen source for thermal ALD, did not adversely interfere with the oxygen non-stoichiometry of the previously grown TaOx layer. Voltage was applied to the Pt top electrode while the TiN bottom electrode was grounded.
3. Results and Discussion Figure 1a shows the log10 (J) vs. electric field (E) curves, where E was calculated from E=V/d, and d is the film thickness of the thermal-ALD Ta2O5 and PEALD TaOx films with R=7, 14, 28 and 70%, respectively, measured at room temperature. For all the growth conditions, d was commonly 15 nm, and the Pt top electrode was negatively biased. B x (R=70%) TaO F x (R=28%) TaO H x (R=14%) TaO J x (R=7%) TaO TaN 2O5
(b)
(c)
-5
E=1.3MV/cm
-6
φP-F 0.8eV
-7 -8 -9
-10
0.6
0.8
1.0
1.2
2.6
1.4
2.8
3.0
3.2
1000/T (K-1)
Electric Field (MV/cm)
-2
0.5eV
-3 -4 -5
E=0.35MV/cm
-8
-13
0.4
E=0.8MV/cm φ P-F
-7
E=0.8MV/cm
-12
0.2
-1
-6
-11
0.0
0
ln (J)
1 0 -1 -2 -3 -4 -5 -6 -7 -8
ln (J)
(a)
log10(J)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
3.4
2.6
2.8
3.0
3.2
3.4
1000/T (K-1)
Figure 1. (a) Leakage current density for the thermal-ALD film (green right-pointed triangle symbol), and PEALD films with different hydrogen-flows (R=7%, blue-triangle symbol; R=14%, circle purple symbol; R=28% diamond yellow symbol and R=70%, red-square symbol) MIM stacks as a function of the applied electric field. Temperature dependences of the leakage current density for R=7% (b) and R=70% (c), fitted by the Poole-Frenkel emission model.
All PEALD films showed good-fitting results in relation to the P-F mechanism (black lines). It is seen that in the case of the PEALD film obtained with R=70%, the P-F emission well described the current density in the whole measured range of E, while the J-E behavior of the samples grown with R=7%, 14% and 28% has a deviation from P-F emission at low-field region, that can be ascribed to the involvement of other leakage mechanisms [30,31], which was not of concern in this work, or current detection limit of the equipment. The leakage current mechanism of the thermal-ALD film was not examined in detail because it will eventually be the RS layer in the ReRAM cell, meaning that the electrical properties after electrical switching will have low relevance to the pristine leakage current property. As expected, the thermal-ALD film showed a very low J level (
