Study of the Resistive Switching Effect in Chromium Oxide Thin Films by Use of Conductive Atomic Force Microscopy KIM NGOC PHAM,1MINSU CHOI,2CAO VINH TRAN,3 TRUNG DO NGUYEN,1VAN HIEU LE,
Trang 1Study of the Resistive Switching Effect in Chromium Oxide Thin Films by Use of Conductive Atomic Force Microscopy
KIM NGOC PHAM,1MINSU CHOI,2CAO VINH TRAN,3 TRUNG DO NGUYEN,1VAN HIEU LE,1TAEKJIB CHOI,4 JAICHAN LEE,2and BACH THANG PHAN1,3,5
1.—Faculty of Materials Science, University of Science, Vietnam National University, Ho Chi Minh City, Vietnam 2.—School of Advanced Materials Science and Engineering, Sungkyunkwan University, Suwon, Republic of Korea 3.—Laboratory of Advanced Materials, University of Sci-ence, Vietnam National University, Ho Chi Minh City, Vietnam 4.—Hybrid Materials Research Center and Faculty/Institute of Nanotechnology and Advanced Materials Engineering, Sejong University, Seoul, Republic of Korea 5.—e-mail: pbthang@hcmus.edu.vn
Reversible resistive switching of Cr2O3films was studied by use of conductive atomic force microscopy Resistive switching in Cr2O3films occurs as a result
of Ag filament paths formed during electrochemical redox reactions A large memory density of 100 Tbit/sq inch was achieved with a small filament diameter of 2.9 nm under the action of a compliance current of 10 nA A fast switching speed of 10 ns, high scalability, and low set/reset currents suggest that Cr2O3-based resistive memory is suitable for nanoscale devices
Key words: Chromium oxide, resistive switching, electrochemical redox
reactions, C-AFM, Ag filament
Recent research has shown that resistance
switching random access memory (ReRAM) is a
promising candidate for nanoscale nonvolatile
mem-ory applications Oxide-based ReRAM structures
exploit the functionality of capacitor structures in
which the oxide materials, for example perovskite
(Cr-doped SrTiO3, Cr-doped SrZrO3, Pr0.7Ca0.3MnO3,
etc.),1 8chalcogenide materials (GeSbTe),9transition
metal oxides (TMOs), or binary oxides (NiO, TiO2,
CuOx, HfO2, ZrOx, ZnO, Nb2O5, Al2O3, WOx,
CrOx)10–18 are sandwiched between two metal
elec-trodes Choosing a material compatible with CMOS
processes is a crucial challenge in current research on
ReRAM Among the different materials used, TMOs
have the major advantages of simple fabrication and
compatibility with CMOS processes.19–22 We have
focused on correlation of the switching behavior of
oxide films (SrTiO3, ZnO, TiO2, WO3and CrOx) with
crystallinity and electrode material.5 8,14–17From the
perspective of application, the basic requirement for
next-generation non-volatile memory is high
scala-bility Because it has recently been shown that
switchable conducting nano-filaments may have
potential for realizing high-density devices, filamen-tary switching in nanoscale devices has attracted much attention.20–23 Further physical insights into geometrical aspects of conducting filaments, for example their number, size, and location, can be obtained by use of conductive atomic force microscopy (C-AFM) Recently, we reported the switching behavior of CrOxthin films, and that the mechanism
of switching was an electrochemical redox reaction.17
To complement previous work, in this paper we report the progressive appearance of conducting fil-aments in CrOxthin films during resistance switch-ing, studied by use of C-AFM
Silver and chromium oxide films were fabricated,
by use of the DC sputtering technique at room temperature, from metallic Ag and Cr targets, on commercial Pt substrates Deposition of 100-nm-thick chromium oxides was performed in a gaseous mixture of 6% oxygen in argon with the total pres-sure kept at 7 9 10 3Torr During deposition of the
Ag top electrode, in an argon environment at
7 9 10 3 Torr, a mask was used for top electrode patterning X-ray photoelectron spectroscopy (XPS) was used to investigate the chemical state of the films Current–voltage (I–V) measurements were obtained by use of a semiconductor-characterization system (Keithley 4200 SCS) and probe station The (Received March 12, 2015; accepted June 4, 2015)
2015 The Minerals, Metals & Materials Society
Trang 2voltage profile for I–V measurements was 0 V fi
(+)Vmax fi 0 V fi + ( )Vmax fi 0 V The Pt
bottom electrode was biased and the top electrode
was grounded For C-AFM measurement,
10-nm-thick chromium oxide was deposited on the Ag
bottom electrode C-AFM measurements were
con-ducted under ambient conditions by use of a Veeco
Dimension D3100 atomic force microscope with Pt
conductive tips as the top electrode
Figure1shows the Cr 2p and O 1s core level XPS
spectra of CrOxfilms prepared at room temperature
As shown in Fig.1a, the Cr 2p3/2core level spectrum
was deconvolved into three peaks with binding
energies of 576.1 eV, 577.5 eV, and 579.2 eV The
576.1 eV-peak was attributed to Cr3+in Cr2O3 The
two peaks at higher binding energies (577.5 eV
and 579.2 eV) were assigned to Cr3+ and Cr6+,
cor-responding toCrO(OH)/Cr(OH)3 and CrO3,
respec-tively The relative amounts of Cr2O3, CrO(OH)/
Cr(OH)3, and CrO3, estimated by
Gaussian–Lor-entzian curve fitting, were 51.63%, 36.7%, and
11.5%, respectively It is clearly apparent that the
Cr2O3phase is predominant
Deconvolution of the O 1s spectrum in Fig 1
resulted in three peaks centered at 530.2 eV,
532 eV, and 533.6 eV The highest-intensity peak of 530.2 eV was assigned to lattice oxygen or a stoi-chiometric Cr2O3phase The lower-intensity peak at higher binding energy was assigned to non-lattice oxygen or non-stoichiometric phases The binding energy of 532 eV corresponds to absorbed oxygen species (O ,O22 ) on the surface of the film The lowest-intensity peak centered at 533.6 eV was attributed to the presence of CrO(OH)/Cr(OH)3 phases in the CrOxfilm
Figure2 shows typical current–voltage charac-teristics of the Ag/100 nm-Cr2O3/Pt structure investigated by use of a dc sweeping voltage with electric pulses applied to Pt bottom electrode As is apparent from Fig.2a, the pristine Ag/Cr2O3/Pt structure has a high-resistance state (HRS) A negative bias voltage applied to the Pt bottom elec-trode switched the structure to the low-resistance state (LRS) Subsequently, on sweeping the positive voltage up to +2 V, the structure was converted back to the HRS The hysteresis I–V curve fol-lows bipolar resistance switching The resistance switching described above can also realized by applying electric pulses with a width of 10 ns, as
Fig 1 XPS spectra of the (a) Cr 2p3/2 and (b) O 1s core levels for
chromium oxide film.
Fig 2 (a) Typical bipolar current–voltage characteristics of Ag/
Cr 2 O 3 /Pt structures and (b) endurance of Ag/Cr 2 O 3 /Pt devices under the action of cycling pulses 10 ns wide.
Pham, Choi, Tran, Nguyen, Hieu Le, Choi, Lee, and Phan
Trang 3shown in Fig.2b The switching was a relatively
fast process Therefore, RRAMs as universal
mem-ories should match DRAMs in terms of switching
speed (DRAM write/erase time 10 ns/10 ns) The
resistance ratio of the HRS and LRS is >30 and
both the HRS and LRS are quite stable after 103
cycles, indicative of good endurance of the Ag/Cr2O3/
Pt structure
Because the I–V curve of the LRS on the log–log scale is indicative of a linear relationship between current and voltage (not shown here), in addition to the nature of the electrodes, a reactive Ag electrode
Fig 3 (a) Schematic diagram of the device used for C-AFM measurements (b) Local I–V hysteresis curve obtained by C-AFM at a compliance current of 10 nA (c–f) Current mapping images in 2D and 3D for the writing and erasing processes for Cr 2 O 3 thin films (g) Statistical distribution
of the size of silver conductive filaments during the writing process at a compliance current of 10 nA.
Trang 4and an inert Pt electrode, and the switching
direc-tion, it is suggested that the mechanism of
switch-ing of Cr2O3 thin films involves electrochemical
redox reactions, which are explained as follows On
application of a negative voltage to the Pt bottom
electrode (positive voltage to the Ag top electrode),
an electrochemical reaction occurs at the anode
(Ag), which oxidizes the Ag metal atoms to Ag ions
These Ag+ions start from the top interface and drift
through the Cr2O3films to connect with the bottom
electrode At the Pt cathode, electrochemical
reduction and electro-crystallization of Ag occur
This electro-crystallization process results in the
formation of an Ag filament, which grows toward
the Ag electrode As a result, the Ag filaments grow
and connect the Ag top electrode, leading to HRS to
LRS switching To reset the cell, a positive voltage is
applied to the Pt bottom electrode (negative
switching voltage to the Ag top electrode), which
leads to dissolution of the Ag filament, and LRS to
HRS switching occurs
XPS analysis shows the presence of oxygen
vacancies Vo2+in the Cr2O3thin films These oxygen
vacancies can affect resistive switching of the Cr2O3
thin films To check the effect of these oxygen
vacancies and of the Ag filaments, we replaced Ag
by Ti as top electrode The Ti/Cr2O3/Pt structure
had no resistive switching behavior Therefore, the
oxygen vacancies do not make a major contribution,
if any, to the switching mechanisms It can again be
concluded that Ag filament paths mediated by
electrochemical redox reactions are responsible for
resistive switching in the Cr2O3thin films
Local I–V hysteresis measurements for the Ag/
10 nm-Cr2O3/Pt structure were conducted by use of
C-AFM A conductive Pt-coated C-AFM tip was used
as the top electrode, as shown schematically in
Fig.3a The voltage was applied to the bottom
electrode during the C-AFM scan The sweeping
voltage followed the sequence 0 fi + 2 V fi
0 fi 2 V fi 0, repeatedly A compliance current
was used during measurements, to protect the C-AFM probe and the structure Hysteresis in the I–V curve at a compliance current Ic= 10 nA is clearly observed in Fig.3b In process 1, the initial HRS was switched to the LRS at an applied voltage (Vset) of approximately +1.7 V In the subsequent voltage sweep, process 3, a negatively applied volt-age (Vreset) of 1.5 V resulted in reversion of the structure back to the HRS It is noted that the large set and reset currents hinder the application of TMOs to integrated RRAM devices However, our Ag/Cr2O3/Pt structure can switch repeatedly with low set and reset currents of 10 nA, leading to very low power consumption
Conductivity mapping results for the writing and erasing processes with Cr2O3films are shown in 2D and 3D images in Fig.3c–f In the writing process, a positive voltage of +0.5 V was applied to the Ag bottom electrode leading to the random presence of bright spots on a dark background; these represent conducting spots or multiple filaments In the erasing process, application of a negative voltage of 0.5 V to the Ag bottom electrode, deletes the cur-rent spots completely, resulting in a uniform dark background The presence of the conducting spots qualitatively confirms the filament model of resis-tive switching
Figure3g shows the statistical distribution of the size of conductive filaments obtained from the writing process at a compliance current of 10 nA The lateral size of the bright spots ranged from 2.9 nm to 30 nm The spot shape also indicates the spots contain multiple filaments The predominant size is <15 nm The small size of the filaments suggests that memory cell size can be scaled down to nanometers
The ability to store multiple resistance states in a single memory cell is one of the most important requirements for non-volatile RRAM, because this can enable dramatic enhancement of memory den-sity Compliance current dependence should also be tested, because different compliance currents are believed to result in multiple resistance states or multiple logical bits
Local I–V hysteresis, current mapping results, and the statistical distribution of filament sizes obtained from the writing process at a compliance current of 500 nA are shown in Fig.4a–d In com-parison with results obtained by use of a compliance current of 10 nA, conductivity mapping shows con-ducting spots with a greater density in the Cr2O3 films The lateral size of the bright spots ranged from 8.8 nm to 100 nm, with the size predominantly below 20 nm The statistical distribution of filament sizes shows the minimum diameter of the filaments
in the Cr2O3 are 2.9 nm and 8.8 nm for Ic= 10 nA and 500 nA, respectively A larger physical diame-ter is induced by use of the larger compliance cur-rent; this results in a lower resistance The different compliance currents clearly indicate that filaments with different resistances are formed Therefore, Fig 3 continued.
Pham, Choi, Tran, Nguyen, Hieu Le, Choi, Lee, and Phan
Trang 5controlling the compliance current modulates the
size of the filament and thus the resistance In
addition, the corresponding memory densities are
100 Tbit/sq and 10.6 Tbit/sq for filament diameters
of 2.9 nm and 8.8 nm, respectively
In this study, reversible resistive switching of
Cr2O3 films was performed by use of C-AFM
Fila-ment-controlled bipolar resistance switching were
clearly apparent from local I–V hysteresis,
conduc-tivity, and current mapping Our study revealed the
correlation between compliance current and
fila-ment size, and the multilevel capability and
mem-ory density of Cr2O3-based RRAM devices The
small compliance current results in small filament
sizes, higher memory density, and low power
con-sumption, suggesting that memory cell size can be
scaled down to tens of nanometers
ACKNOWLEDGEMENTS
This work was funded by the National Foundation
of Science and Technology Development of Vietnam
(NAFOSTED—103.02-2012.50), The Exchange
Fel-lowship Programme under ASEAN-ROK Academic
Exchange Programme 2014, and the Basic Science Research Program through the National Research Foundation of Korea (2009-0092809)
REFERENCES
1 A Beck, J.U Bednorz, C Gerber, Ch Rossel, and D Windmer, Appl Phys Lett 77, 139 (2000).
2 S.Q Liu, N.J Wii, and A Ignatiev, Appl Phys Lett 76,
2749 (2000).
3 Y Watanabe, J.G Bednorz, A Bietsch, C Gerber, D Widmer, A Beck, and S.J Wind, Appl Phys Lett 78, 3738 (2001).
4 R Waser and M Aono, Nat Mater 6, 833 (2007).
5 B.T Phan and J Lee, Appl Phys Lett 93, 222906 (2008).
6 B.T Phan and J Lee, Appl Phys Lett 94, 232102 (2009).
7 B.T Phan, N.C Kim, and J Lee, J Kor Phys Soc 54, 873 (2009).
8 B.T Phan, T Choi, A Romanenko, and J Lee, Solid-State Electronics 75, 43 (2012).
9 Y-C Chen, C.F Chen, C.T Chen, J.Y Yu, S Wu, S.L Lung,
R Liu, and C-Y Lu, IEDM Tech Dig (2003).
10 B.J Choi, D.S Jeong, S.K Kim, S Choi, J.H Oh, C Rohde, H.J Kim, C.S Hwang, K Szot, R Waser, B Reichenberg, and S Tiedke, J Appl Phys 98, 033715 (2005).
Fig 4 (a) Local current behavior investigated by C-AFM (b, c) Current mapping images in 2D and 3D for the writing process (d) Statistical distribution of the size of conductive filaments for a compliance current of 500 nA.
Trang 611 K Jung, H Seo, Y Kim, H Im, J.P Hong, J.W Park, and
J.K Lee, Appl Phys Lett 90, 052104 (2007).
12 A Chen, S Haddad, Y.C Wu, Z Lan, T.N Fang, and S.
Kaza, Appl Phys Lett 91, 123517 (2007).
13 C.Y Lin, C.Y Wu, C Hu, and T.Y Tseng, J Electrochem.
Soc 154, G189 (2007).
14 T Le, H.C.S Tran, V.H Le, T Tran, C.V Tran, T.T Vo,
M.C Dang, S.S Kim, J Lee, and B.T Phan, J Korean Phys.
Soc 60, 1087 (2012).
15 K.N Pham, T.D Nguyen, T.K.H Ta, K.L.D Thuy, V.H Le,
D.P Pham, C.V Tran, D Mott, S Maenosono, S.S Kim, J.
Lee, D.T Pham, and B.T Phan, Eur Phys J Appl Phys.
64, 30102 (2013).
16 J.B Park, K.P Biju, S.J Jung, W.T Lee, J.M Lee, S.H.
Kim, S.S Park, J.H Shin, and H.S Hwang, IEEE Electron
Device Lett 32, 476 (2011).
17 N.K Pham, D.T Nguyen, B.T.T Dao, K.T.H Ta, V.C Tran, V.H Nguyen, S.S Kim, S Maenosono, and T.B Phan,
J Electron Mater 43, 2747 (2014).
18 T.B.T Dao, K.N Pham, Y.L Cheng, S.S Kim, and B.T Phan, Curr Appl Phys 14, 1707 (2014).
19 J.W Seo, Doctor Thesis, Transparent resistive random ac-cess memory for transparent electronics, Korean Advanced Institute of Science and Technology (2011).
20 S.Z Rahaman, S Maikap, T.C Tien, H.Y Lee, W.S Chen, F.T Chen, M.J Kao, and M.J Tsai, Nanoscale Res Lett 7, 345 (2012).
21 Q Liu, J Sun, H.B Lv, S.B Long, K.B Yin, N Wan, Y.T.
Li, L.T Sun, and M Liu, Adv Mater 24, 1844 (2012).
22 A Prakash, S Maikap, H.C Chiu, T.C Tien, and C.S Lai, Nanoscale Res Lett 8, 288 (2013).
23 A.A Sharma, M Noman, M Abdelmoula, M Skowronski, and J.A Bain, Adv Funct Mater 24, 5522 (2014). Pham, Choi, Tran, Nguyen, Hieu Le, Choi, Lee, and Phan