Gate-tunable gigantic lattice deformation in VO2D. Okuyama, M. Nakano, S. Takeshita, H. Ohsumi, S. Tardif, K. Shibuya, T. Hatano, H. Yumoto, T. Koyama, H.Ohashi, M. Takata, M. Kawasaki, T. Arima, Y. Tokura, and Y. Iwasa Citation: Applied Physics Letters 104, 023507 (2014); doi: 10.1063/1.4861901 View online: http://dx.doi.org/10.1063/1.4861901 View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/104/2?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Anomalous lattice deformation in GaN/SiC(0001) measured by high-speed in situ synchrotron X-ray diffraction Appl. Phys. Lett. 108, 012102 (2016); 10.1063/1.4939450 Porous silicon-VO2 based hybrids as possible optical temperature sensor: Wavelength-dependent opticalswitching from visible to near-infrared range J. Appl. Phys. 118, 134503 (2015); 10.1063/1.4932023 Field-effect modulation of conductance in VO2 nanobeam transistors with HfO2 as the gate dielectric Appl. Phys. Lett. 99, 062114 (2011); 10.1063/1.3624896 Negative thermal expansion and spontaneous magnetostriction of Tb 2 Fe 16.5 Cr 0.5 compound J. Appl. Phys. 97, 116102 (2005); 10.1063/1.1921334 Microhardness and structural analysis of (Ti,Al)N, (Ti,Cr)N, (Ti,Zr)N and (Ti,V)N films J. Vac. Sci. Technol. A 18, 1038 (2000); 10.1116/1.582296

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 132.168.8.237 On: Thu, 30 Jun 2016

15:09:43

Gate-tunable gigantic lattice deformation in VO2

D. Okuyama,1,a),b) M. Nakano,1,2,a),b) S. Takeshita,3 H. Ohsumi,3 S. Tardif,3 K. Shibuya,4,c)

T. Hatano,1 H. Yumoto,5 T. Koyama,5 H. Ohashi,5 M. Takata,3,5 M. Kawasaki,1,6

T. Arima,1,3,7 Y. Tokura,1,6 and Y. Iwasa1,6,b)

1RIKEN Center for Emergent Matter Science (CEMS), Wako 351-0198, Japan2Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan3RIKEN SPring-8 Center, Hyogo 679-5148, Japan4National Institute of Advanced Industrial Science and Technology, Tsukuba 305-8562, Japan5Japan Synchrotron Radiation Research Institute, SPring-8, Hyogo 679-5198, Japan6Quantum-Phase Electronics Center and Department of Applied Physics, University of Tokyo,Tokyo 113-8656, Japan7Department of Advanced Materials Science, University of Tokyo, Kashiwa 277-8561, Japan

(Received 23 October 2013; accepted 29 December 2013; published online 15 January 2014)

We examined the impact of electric field on crystal lattice of vanadium dioxide (VO2) in a

field-effect transistor geometry by in-situ synchrotron x-ray diffraction measurements. Whereas

the c-axis lattice parameter of VO2 decreases through the thermally induced insulator-to-metal

phase transition, the gate-induced metallization was found to result in a significant increase of

the c-axis length by almost 1% from that of the thermally stabilized insulating state. We also

found that this gate-induced gigantic lattice deformation occurs even at the thermally stabilized

metallic state, enabling dynamic control of c-axis lattice parameter by more than 1% at room

temperature. VC 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4861901]

The field-effect transistor (FET) is a key building block

for modern information technology, enabling electrical

switching of current flowing through a channel surface by

external voltages. The essential feature of FET is to tune the

number of electrons in a channel material electrostatically,

and thus FET technique has been widely employed as a

powerful tool for controlling electronic properties of con-

densed matters over the past few decades.1 More recently,

the idea of utilizing significantly large capacitance of

electric-double layer (EDL) formed at a solid/electrolyte

interface as a gate dielectric layer of FET, namely, EDL tran-

sistor (EDLT), plays a key role because of its great potential

for high-density carrier accumulation up to 1015 cm�2 at a

surface of solids. This amount of charge is high enough for

triggering electronic phase transitions only by voltages, as

has been practically demonstrated by electric-field induced

superconductivity,2–4 ferromagnetism,5–7 and metal-insulator

transition (MIT)8–10 for the last five to six years.

We recently reported that a combination of EDLT and

vanadium dioxide (VO2) provides remarkable features rele-

vant to the first-order MIT.11 VO2 is a classic strongly corre-

lated oxide showing characteristic MIT above room

temperature with a resistance jump of more than four orders

of magnitude.12 This MIT is of the first order, and hence, elec-

tronic states are strongly coupled with lattice structures; a

half-filled metallic state is thermally stabilized in a tetragonal

phase above the transition temperature (TMI), whereas a

low-temperature insulating state is stabilized in a monoclinic

phase with dimerized V4þ ion pairs.13–15 We fabricated

EDLT with VO2 thin films, and found that the

low-temperature insulating state can be completely switched

to the metallic state over the first-order MIT by application of

gate voltages (VG). Moreover, it turned out that an electrically

induced conducting channel is extended to an entire region of

the 70-nm thick film along c-axis direction beyond the funda-

mental electrostatic screening length, which is in marked con-

trast to conventional FETs that have a two-dimensional

conducting channel. We have attributed these phenomena to

electric-field induced collective phase transitions; electrostatic

charge accumulation at a surface triggers long-range bulk lat-

tice deformation to minimize interface energy, and subse-

quently, this bulk structural transformation drives a cascade of

electronic phase transition inside bulk region of a film beyond

the screening length. This distinguished function of

electric-field induced bulk phase transition is particularly ad-

vantageous for optoelectronic device applications such as

energy-saving smart windows as we recently demonstrated.16

Due to its fundamental interest regarding the operation

mechanism as well as its practical importance toward future

ultralow-energy consumption electronics based on correlated

oxides, VO2-EDLT has attracted growing attention.17–21

Among a number of recent reports, Jeong et al. has also real-

ized gate-induced metallization in VO2 thin films by EDLT

technique using ionic liquid, but they have proposed a differ-

ent mechanism; electric-field induced oxygen vacancy for-

mation.21 As direct evidence, they presented secondary ion

mass spectrometry data using oxygen isotope, which indi-

cates that in their case oxygen fills into a gate-induced metal-

lic state at a surface of VO2 film within 5 to 10 nm length

scale upon applying negative VG under oxygen atmosphere.

In addition, they noted that a gate-induced metallic state

remains even after removal of ionic liquid, suggesting irre-

versible chemical reactions. These results may not be able to

explain our results, where a gate-induced metallic state is not

a)D. Okuyama and M. Nakano contributed equally to this work.b)Electronic addresses: okuyama@riken.jp, nakano@imr.tohoku.ac.jp, and

iwasa@ap.t.u-tokyo.ac.jpc)This research was performed while K. Shibuya was at RIKEN, Wako 351-

0198, Japan.

0003-6951/2014/104(2)/023507/5/$30.00 VC 2014 AIP Publishing LLC104, 023507-1

APPLIED PHYSICS LETTERS 104, 023507 (2014)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 132.168.8.237 On: Thu, 30 Jun 2016

15:09:43

limited to a surface but extended to an entire region of the

70-nm thick film, and it can be reversibly switched on and

off by application of VG. This implies that the governing

mechanism can differ depending on sample variations and/or

experimental conditions, even though a metallic ground state

can be induced by application of VG for both cases. There

are even other mechanisms suggested by other groups; one

group insists that hydrogen doping should play a major

role,19 and another group claims that gating effect is limited

only at a surface.20 Accumulation of experimental evidence

from different aspects is highly desired to understand what is

going on in VO2-EDLT. In this report, we directly elucidate

by in-situ x-ray diffraction (XRD) measurements that crystal

lattice of VO2 significantly deforms under the presence of

electric field throughout a whole film over the screening

length. We also show that this bulk lattice deformation can

be dynamically controlled at room temperature by VG in a re-

versible isothermal process.

We prepared two VO2-based (001) epitaxial thin films

having different thicknesses of 10 nm (sample #1) and 40 nm

(sample #2), both of which were grown on TiO2 (001) single

crystal substrates by pulsed laser deposition. The growth con-

dition of the films and the device fabrication process were

nominally the same as the previous reports.11,22 Sample speci-

fications are listed in Fig. 1(a). The 40-nm thick film

(sample #2) was doped with 2% of tungsten in order to reduce

TMI below 300 K due to limitation of the maximum tempera-

ture available with the measurement systems used in this

study. Both films were grown at around 400 �C under the ox-

ygen pressure of 10 mTorr, and then patterned into a standard

Hall-bar geometry by combining conventional photolithogra-

phy and Ar-ion etching techniques. Figure 1(b) shows an opti-

cal micrograph of an actual device with measurement

configurations. Source/drain electrodes, voltage probes, and a

side gate electrode were made with Ti/Au deposited by

electron-beam evaporation. The dimensions of a channel were

60 lm in width and 500 lm in length. Both a channel and a

gate electrode were finally covered by an organic ionic liquid,

N,N-diethyl-N-(2-methoxyethyl)-N-methylammonium bis-(tri-

fluoromethylsulfonyl)-imide (DEME-TFSI). The sample #1

was then loaded in a standard cryostat and its electronic and

structural properties were characterized in high vacuum

(P< 10�5 Pa), whereas those properties of the sample #2 were

evaluated under N2 flow. The XRD measurements on the sam-

ple #1 and #2 were performed by a commercially-available

four-circle diffractometer and the home-made x-ray microp-

robe system, respectively, both of which are installed on the

FIG. 1. (a) Specifications of the samples used in this study. (b) An optical

micrograph of a typical device with measurement circuits prepared for in-situ x-ray diffraction (XRD) measurements. X-ray beams were focused on

the channel region (yellow/orange semicircle area) so that we could selec-

tively characterize the local crystal structure during device operation. The

green region indicates the channel that had a direct contact with ionic liquid

and therefore was subjected for EDLT operation. (c) The temperature de-

pendence of the sheet resistance (Rs) for the samples #1 and #2 both at the

ungated OFF states and at the gated ON states. The gate voltages (VG) at

OFF and ON states of the sample #1 were 0.0 V and 1.5 V, respectively,

while those of the sample #2 were �3.0 V and 4.4 V, respectively. Symbols

A, B, C, D, E, and F are the points where XRD measurements were

performed.

FIG. 2. (a) The XRD patterns for the (002) Bragg reflection peak collected

at the channel region of the sample #1 at three different conditions, where

the horizontal axis 2d is twice the spacing between the (002) planes. Upturns

seen in each profile at higher 2d side were due to strong (002) reflection

from TiO2 substrate, whose 2d value is 2.959 A. A small sub-peak seen in

the spectrum C at around the 2d of 2.90 A was from Laue oscillations. (b)

The same set of data for the sample #2.

023507-2 Okuyama et al. Appl. Phys. Lett. 104, 023507 (2014)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 132.168.8.237 On: Thu, 30 Jun 2016

15:09:43

Beam Line 19 LXU in SPring-8.23 The energy of the incident

x-ray was tuned to 12.4 keV. The sheet resistance (Rs) was

simultaneously recorded during XRD measurements using

semiconductor parameter analyzer (Agilent Technologies,

4156C).

Figure 1(c) shows the temperature dependence of Rs for

the samples #1 and #2 both at the ungated OFF states and at

the gated ON states evaluated under dark condition when the

x-ray beams were off. At OFF states, clear MITs were

observed for both samples around room temperature with ab-

rupt resistance jumps of a few orders of magnitude accompa-

nied by thermal hysteresis, whereas MITs were completely

suppressed at ON states in spite of the presence or absence of

tungsten. Rs of the sample #2 at ON state (red data) was much

lower than that of the sample #1 (green data), indicating that

the gate-induced conducting channels were extended to the

whole films in three dimensions for both cases irrespective of

the thickness as we previously reported.11 In order to directly

verify that the whole film underwent phase transitions in the

light of crystal structure, we next performed XRD measure-

ments at different conditions labeled as A to F in Fig. 1(c).

VO2 transforms its crystal structure across the

thermally-driven MIT from the low-temperature insulating

monoclinic phase to the high-temperature metallic tetragonal

phase with a dramatic change in a c-axis length by almost 1%

in a bulk single crystalline form.24 This significant lattice de-

formation along c-axis direction is also seen in a strained VO2

(001) thin film grown on a lattice-matched TiO2 (001) sub-

strate. Figure 2(a) displays the (002) XRD patterns taken

exactly at the channel region of the sample #1 at three differ-

ent conditions labeled as A, B, and C on the Rs-T curves in

Fig. 1(c), where the horizontal axis 2d is twice the spacing

between the (002) planes. The black open and filled symbols

correspond to the data taken at OFF states (VG¼ 0.0 V) below

(B) and above (C) TMI, respectively, showing nearly 1%

change between the 2d values across the thermally-driven

MIT, where the high-temperature metallic state has a smaller

c-axis length as is the case with bulk VO2.

We found that applying electric field yields an opposite

effect. The filled green symbol in Fig. 2(a) denotes the XRD

pattern taken at ON state (VG¼ 1.5 V) well below the origi-

nal TMI (A), indicating more than 1% increase of the 2d

FIG. 3. The temporal evolution of (a) VG and (b) Rs while measuring the XRD pattern at OFF state (“Measure” region), applying positive VG to move on to

ON state (“Set” region), and applying negative VG to get back to OFF state (“Reset/Measure” region), characterized at room temperature (T¼ 305 K) for the

sample #2. A gray area corresponds to the period used for optical alignment and preparation for the XRD measurements (“Alignment” region). (c) The tempo-

ral change of the XRD pattern collected during resetting VO2 from ON to OFF states. Small sub-peaks seen in the spectra G, H, and I at around the 2d of

2.86 A were most likely from ungated VO2 underneath voltage probes, which were gate-insensitive. (d) The relationship between the 2d(002) value and Rs,

which is time-averaged over 10 min required for each XRD measurement.

023507-3 Okuyama et al. Appl. Phys. Lett. 104, 023507 (2014)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 132.168.8.237 On: Thu, 30 Jun 2016

15:09:43

value from the low-temperature insulating monoclinic phase

(B). The obtained profile is constituted with a single

Gaussian distribution function without any shoulder compo-

nents at the position of the thermally-stable insulating state

(B). This suggests that a whole film has a single crystal struc-

ture, directly evidencing from structural viewpoint that the

gate-induced MIT occurs throughout an entire film as we

concluded previously in the lights of transport and optical

spectroscopy measurements.11,16 This bulk lattice deforma-

tion upon electrically driven MIT is also confirmed in the

sample #2, which consists of the 40-nm thick film, as shown

in Fig. 2(b). We note that the linear thermal expansion coef-

ficients of VO2 below and above TMI and that of TiO2 are

approximately 6, 17, and 9� 10�6 K�1, respectively,25,26

suggesting negligible contribution of these thermal effects to

the obtained data on the length scale discussed here.

Surprisingly, we found that this gigantic change in a c-

axis lattice parameter can be induced even above TMI, where

the high-temperature “metallic” state is originally stabilized at

OFF state. Figure 3 demonstrates dynamic control of lattice

deformation in VO2 by electric field at room temperature

(T¼ 305 K) examined for the sample #2. From our previous

studies, we have known that the application of VG above TMI

results in a slight increase of Rs, which is essential for a

decrease of TMI and eventual emergence of the gate-induced

three-dimensional metallic ground state.11 We focus on this

behavior as a measure of the change in the electronic state of

VO2 during gating process. Starting from the

high-temperature metallic state (F), positive VG was firstly

applied while monitoring Rs in order to get to the

gate-induced “metallic” state [“Set” region in Figs. 3(a) and

3(b)]. Rs monitoring was then interrupted to do optical align-

ment for the XRD experiments [gray “Alignment” region in

Figs. 3(a) and 3(b)]. After that, simultaneous characterizations

of the XRD patterns and Rs were started while VG was gradu-

ally altered from positive to negative so as to release electrical

charges accumulated at the surface of VO2 [“Reset/Measure”

region in Figs. 3(a) and 3(b)]. Figure 3(c) illustrates the snap-

shots of the XRD patterns collected during this “reset” pro-

cess, showing significant shift of the 2d value by more than

1% from the start point (G) to the end point (M) according to

the change in VG and corresponding change in Rs [see Figs.

3(a) and 3(b)]. There was a linear relationship between the

2d(002) value and Rs as shown in Fig. 3(d), implying that the

electronic state and the structural phase are strongly coupled

with each other at the gated state. The XRD pattern and Rs

eventually returned to those of the high-temperature metallic

state, verifying the reversible device operation.

We note that the obtained results on the gating effects at

the thermally stable metallic state, in particular, increase of

Rs upon applying VG is unusual and cannot be understood

within the framework of a standard parallel capacitor model.

One of possible interpretations is that a gate-induced metallic

state has a larger resistivity than that of a thermally stable

metallic state due to strong deformation in lattice structure

throughout an entire film under the presence of extremely

large electric field. On the length scale of this gate-induced

lattice deformation, we speculate that it should be considered

separately from the electrostatic screening length; it should

not be limited by the screening length, but rather governed

by the size of the metal/insulator domains of two different

structures, which is known to be fairly large (in the order of

a few tens micrometers) for VO2 (001) epitaxial thin films

grown on lattice-matched TiO2 (001) substrates.27

In summary, we revealed by in-situ synchrotron XRD

experiments that the gate-induced metallic states in

VO2-EDLT have larger c-axis lattice parameters than those

of the thermally stable insulating and metallic states, and

that this gate-induced lattice deformation is extended to a

whole film. A gate sweep measurement above TMI enabled

dynamic control of lattice deformation at room temperature

in a reversible process, where a remarkable shift of a c-axis

lattice parameter by more than 1% was achieved. The

obtained results suggest the formation of unidentified phases

in VO2 under the presence of electric field both below and

above TMI. Further structural analysis including in-plane

reflection measurements in a wider VG-T parameter space as

well as x-ray spectroscopic measurements will give us fur-

ther insights into the origin of these phases, and eventually,

into the device operation mechanism and physics behind.

This work was supported by the Japan Society for the

Promotion of Science (JSPS) through its “Funding Program

for World-Leading Innovative R&D on Science and

Technology (FIRST Program).” This work was partly sup-

ported by Grants-in-Aid for Scientific Research (Grant Nos.

21224009, 24224009, 24226002, 25000003, 25790051, and

25708040) from JSPS. M.N. was partly supported by The

Murata Science Foundation and by Grant for Basic Science

Research Projects from The Sumitomo Foundation. Y.I. was

supported by Strategic International Collaborative Research

Program (SICORP-LEMSUPER) from Japan Science and

Technology Agency (JST). The synchrotron x-ray diffraction

experiments were performed at BL19LXU in SPring-8 with

approval of RIKEN (Proposal Nos. 20110091 and 20120084).

1C. H. Ahn, A. Bhattacharya, M. Di Ventra, J. N. Eckstein, C. Daniel

Frisbie, M. E. Gershenson, A. M. Goldman, I. H. Inoue, J. Mannhart, A. J.

Millis, A. F. Morpurgo, D. Natelson, and J.-M. Triscone, Rev. Mod. Phys.

78, 1185 (2006).2K. Ueno, S. Nakamura, H. Shimotani, A. Ohtomo, N. Kimura, T. Nojima,

H. Aoki, Y. Iwasa, and M. Kawasaki, Nature Mater. 7, 855 (2008).3A. T. Bollinger, G. Dubuis, J. Yoon, D. Pavuna, J. Misewich, and I.

Bo�zovic, Nature 472, 458 (2011).4J. T. Ye, Y. J. Zhang, R. Akashi, M. S. Bahramy, R. Arita, and Y. Iwasa,

Science 338, 1193 (2012).5M. Weisheit, S. F€ahler, A. Marty, Y. Souche, C. Poinsignon, and D.

Givord, Science 315, 349 (2007).6Y. Yamada, K. Ueno, T. Fukumura, H. T. Yuan, H. Shimotani, Y. Iwasa,

L. Gu, S. Tsukimoto, Y. Ikuhara, and M. Kawasaki, Science 332, 1065

(2011).7K. Shimamura, D. Chiba, S. Ono, S. Fukami, N. Ishiwata, M. Kawaguchi,

K. Kobayashi, and T. Ono, Appl. Phys. Lett. 100, 122402 (2012).8S. Asanuma, P.-H. Xiang, H. Yamada, H. Sato, I. H. Inoue, H. Akoh, A.

Sawa, K. Ueno, H. Shimotani, H. Yuan, M. Kawasaki, and Y. Iwasa,

Appl. Phys. Lett. 97, 142110 (2010).9R. Scherwitzl, P. Zubko, I. G. Lezama, S. Ono, A. F. Morpurgo, G.

Catalan, and J.-M. Triscone, Adv. Mater. 22, 5517 (2010).10T. Hatano, Y. Ogimoto, N. Ogawa, M. Nakano, S. Ono, Y. Tomioka, K.

Miyano, Y. Iwasa, and Y. Tokura, Sci. Rep. 3, 2904 (2013).11M. Nakano, K. Shibuya, D. Okuyama, T. Hatano, S. Ono, M. Kawasaki,

Y. Iwasa, and Y. Tokura, Nature 487, 459 (2012).12F. J. Morin, Phys. Rev. Lett. 3, 34 (1959).13A. S. Barker, H. W. Verleur, and H. J. Guggenheim, Phys. Rev. Lett. 17,

1286 (1966).14C. N. Berglund and H. J. Guggenheim, Phys. Rev. 185, 1022 (1969).

023507-4 Okuyama et al. Appl. Phys. Lett. 104, 023507 (2014)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 132.168.8.237 On: Thu, 30 Jun 2016

15:09:43

15J. B. Goodenough, J. Solid State Chem. 3, 490 (1971).16M. Nakano, K. Shibuya, N. Ogawa, T. Hatano, M. Kawasaki, Y. Iwasa,

and Y. Tokura, Appl. Phys. Lett. 103, 153503 (2013).17Z. Yang, Y. Zhou, and S. Ramanathan, J. Appl. Phys. 111, 014506 (2012).18Y. Zhou and S. Ramanathan, J. Appl. Phys. 111, 084508 (2012).19H. Ji, J. Wei, and D. Natelson, Nano Lett. 12, 2988 (2012).20K. Liu, D. Fu, J. Cao, J. Suh, K. X. Wang, C. Cheng, D. F. Ogletree, H.

Guo, S. Sengupta, A. Khan, C. W. Yeung, S. Salahuddin, M. M.

Deshmukh, and J. Wu, Nano Lett. 12, 6272 (2012).21J. Jeong, N. Aetukuri, T. Graf, T. D. Schladt, M. G. Samant, and S. S. P.

Parkin, Science 339, 1402 (2013).

22K. Shibuya, M. Kawasaki, and Y. Tokura, Appl. Phys. Lett. 96, 022102

(2010).23M. Yabashi, T. Mochizuki, H. Yamazaki, S. Goto, H. Ohashi, K. Takeshita,

T. Ohata, T. Matsushita, K. Tamasaku, Y. Tanaka, and T. Ishikawa, Nucl.

Instrum. Methods Phys. Res., Sect. A 467–468, 678 (2001).24D. Kucharczyx and T. Niklewski, J. Appl. Crystallogr. 12, 370 (1979).25T. Kawakubo and T. Nakagawa, J. Phys. Soc. Jpn. 19, 517 (1964).26K. V. K. Rao, S. V. Nagender Naidu, and L. Iyengar, J. Am. Ceram. Soc.

53, 124 (1970).27K. Kawatani, H. Takami, T. Kanki, and H. Tanaka, Appl. Phys. Lett. 100,

173112 (2012).

023507-5 Okuyama et al. Appl. Phys. Lett. 104, 023507 (2014)

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. Download to IP: 132.168.8.237 On: Thu, 30 Jun 2016

15:09:43