Applied Surface Science 399 (2017) 346–350

Contents lists available at ScienceDirect

Applied Surface Science

journal homepage: www.elsevier.com/locate/apsusc

Full Length Article

Metal-insulator transition in epitaxial NdNiO3 thin film: A structural,

electrical and optical study
Tao Shao, Zeming Qi ∗ , Yuyin Wang, Yuanyuan Li, Mei Yang, Chuansheng Hu
National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei, Anhui, 230029, China

a r t i c l e i n f o a b s t r a c t

Article history: NdNiO3 thin film has been prepared by pulsed laser deposition on LaAlO3 (001) single crystalline sub-
Received 31 July 2016 strate. Temperature-dependent resistivity measurement shows a sharp metal-insulator transition in such
Received in revised form 8 December 2016 thin film. The phase transition temperature can be tuned from 90 K to 121 K by changing the thick-
Accepted 13 December 2016
ness of thin film. The structure evolution during phase transition is studied by Raman spectroscopy.
Available online 14 December 2016
Optical conductivity reveals that the variation carrier density in the process of phase transition. The
results of structural, electrical and optical studies provide useful insights to understand the mechanism
Keywords:
of metal-insulator transition of NdNiO3 thin film.
Metal-insulator transition
Pulsed laser deposition
© 2016 Elsevier B.V. All rights reserved.
Raman spectrum
Optical conductivity
NiO6 octahedra

1. Introduction
Materials with tunable metal–insulator transition (MIT) attract
growing interest in various applications, such as sensors, elec-
tronic switches, thermochromic coatings, non-volatile memory, etc
[1–4]. Among several MIT materials, correlated oxides such as rare
earth nickelates (RNiO3 , R = Rare Earth) that undergo temperature-
induced first order MIT, where the MIT temperature (TMI ) increases
with decreasing cation radius (increasing R atomic number) [3],
has been intensively studied in recent years. As more scientific
and technological importance create increasing interests in the
research of RNiO3 , it is difficult to be prepared this material in
bulk form because it requires very high temperatures and high
oxygen pressure in synthesis process [5]. Thin film deposition pro-
vides an opportunity to stabilize the metastable phases and enables
a near single crystal films under epitaxial stabilization. Besides,
certain strain states can be achieved in thin films by choosing
the substrate of compatible lattice parameters [6–8], which may
induce the desired compressive or tensile strain in the films [9,10].
Furthermore, the stain can also be manipulated by conveniently
adjusting thin film thickness[11]. RNiO3 thin film is, therefore, an
ideal candidate to investigate the behavior of MIT when subjected
to confinement, lattice mismatch and strain. It has been reported
that different strain states can change the phase transition prop-
∗ Corresponding author.
E-mail address: zmqi@ustc.edu.cn (Z. Qi).
erties of RNiO3 thin film. For example, the transport properties
of NdNiO3 (NNO) thin films are altered by the changing of lattice
mismatch of the film with substrates [12,13]. Keimer’s et al. have
carried out a series of depth resolved Raman scattering measure-
ments to investigate the lattice dynamics of ultra-thin films of
LaNiO3 [14]. Catalan et al. found the thickness dependence of the
MIT of NNO film on LaAlO3 (LAO) substrate [15,16].
In this paper, a group of NNO thin films with different
thicknesses were prepared on LaAlO3 substrate by pulsed laser
deposition (PLD) method. The structural, electrical and optical
properties are studied by combining X-ray diffraction, resistivity,
Raman and infrared spectroscopy. The results provide insights into
understanding the temperature and thickness effects on the metal-
insulator transition of NNO thin film.
2. Experiment
To prepare the target material for PLD deposition [17], the pre-
cursor compounds, Nd2 O3 (99.99%) and NiO (99.99%) were weighed
in the correct proportions and mixed together. The mixture was
firstly wet ground in ethanol and then dry ground. Both grindings
were for 20 min using an alumina mortar and pestle. The ground
powders were pressed into pellets and were annealed in two differ-
ent steps [18]. The first annealing step was at 1000 ◦ C for 12 h and
the second step was at 1200 ◦ C for 24 h. Between each annealing
step, the pellets were re-ground and re-pressed as described above.
A KrF-Excimer Laser (LPX200, Lambda Physik) with a wavelength of
248 nm and an energy density of 2 J/cm2 was used for the thin film

http://dx.doi.org/10.1016/j.apsusc.2016.12.102
0169-4332/© 2016 Elsevier B.V. All rights reserved.
 T. Shao et al. / Applied Surface Science 399 (2017) 346–350
Fig. 1. (a) (002) diffraction peak and rocking curve(insert) of NNO thin film
deposited on LAO substrate, (b) The ␸-scan curves of the NNO film at 80 nm thick-
ness.
deposition. The different thickness of NNO films was achieved by
alternating the deposition time. The chamber was turbo-molecular
pumped to a vacuum of 2.4 × 10−4 Pa, then backfilled with O2 to
a pressure of 3 mbar. During the growth, the substrate was kept
at a temperature of 710 ◦ C. After deposition, the pressure of O2
was increased to 300 mbar and the thin film was kept at 710 ◦ C
for 15 min before being cooling to room temperature at a rate of
20 ◦ C/min. The above optimized conditions ensure us to obtain the
best stabilized the NNO thin film samples.
The structural properties of the different thickness samples
were characterized by X-ray diffraction (XRD). The thin film thick-
ness was determined to be 40 nm, 80 nm and 160 nm by measuring
the cross-section of the thin film using scanning electron micro-
scope (SEM), corresponding to the deposition time of 2.5 min, 5 min
and 10 min, respectively. The temperature-dependent resistivity of
the samples was measured by the four-probe technique using a
Physical Property Measurement System (PPMS). The temperature-
dependent Raman spectra were recorded with an inVia-Reflex
spectrometer with triple monochromatoron. The excitation source
was 532 nm line of an Ar+ laser. The temperature-dependent
infrared reflectivity was measured using a Bruker IFS 66 v FTIR
spectrometer on an infrared beamline (BL01B) at the National
Synchrotron Radiation Laboratory(NSRL), China. The sample was
mounted on the cold finger of a continuous flow cryostat, and the
measurement temperatures from the room temperature to 78 K,
which included the metal-insulator transition region.
3. Result and discussion
Fig. 1(a) shows the XRD patterns of (002) diffraction plane for
NNO/LAO thin films at three kinds of thickness. For the bulk NNO
and LAO, both (002) diffraction peaks appear at 2␪∼48◦ [19,20].
For the thin film, the (002) peaks of NNO film and LAO substrate
do not overlap, indicateing a small lattice mismatch between the
film and the substrate [21,22]. This mismatch causes an in-plane
compressive strain (∼0.75%) in thin film and results in the NNO
diffraction peak appears at a lower 2␪angle. The peak shifts to lower
347
Fig. 2. Measured resistivity ␳ (a) and −d(ln␳)/dT as a function of temperature (b) for
NNO films at different thickness. Inset: d␳/dT as a function of increasing temperature
for NNO films at different thickness.
angle when the film thickness is increased, indicating an increase of
out-of-plane lattice constant. The rocking curves are shown in the
insert of Fig. 1(a). It is clear that the thicker film has broader peak
suggesting higher mosaicity. This is consistent with strain relax-
ation causing increasing defects as the symmetry of parts of the
film returns to bulk-like O Ni O bond angles. The in-plane epitax-
ial relationship between the NNO film and the LAO substrate was
further examined through XRD ␸-scans of the NNO(101) planes
as shown in Fig. 1(b). Four sharp peaks are observed in ␸-scan
results,which confirms that the NNO film has four-fold symmetry
corresponding to a cube-on-cube growth direction, suggesting the
epitaxial nature of the NNO film on the LAO substrate [23].
In order to evaluate the phase transition property of the NNO
thin film, temperature-dependent resistivity for different thick-
ness NNO thin films was measured and shown in Fig. 2. It is clear
that a sharp insulator-metal transition occurs from a high resis-
tivity insulate state to a metallic state with a hysteresis behavior
across the phase transition boundary. To quantitatively compare
the difference of these samples, the resistivity-temperature curves
were further analyzed, and the following parameters such as

TMI TMI ,ransition sharpness, transition hysteresis, and quality of the
metallic state were obtained and listed in Table 1. The definition
of these parameters adopts the same convention, as shown in Ref
[24], where TMI corresponds to the temperature associated with
the −d(ln␳)/dT maximum on heating, which is shown in Fig. 2(b).

TMI s the temperature at which the sign of d␳/dT changed, which
is shown in the insert of Fig. 2(a). Transition sharpness is defined
as the magnitude of the peak height of −d(ln␳)/dT on heating. The
quality of the metallic state is quantified by normalized resistiv-
ity slope at high temperature (e.g.-d(ln␳)/dT at 300 K). Transition
hysteresis represents the TMI difference between heating and cool-
ing processes. According to the results, the TMI increases from 90 K
to 121 K and the hysteresis between heating and cooling curves
changes from 7 K to 19 K when the thin film thickness changes
348 T. Shao et al. / Applied Surface Science 399 (2017) 346–350

Table 1
Properties of structure, metal-insulator transition, and metallic state for NNO films at different thickness.

NdNiO3 thickness ac (Å) Statistical error bar Mismatch (%) TMI (K) – TMI K) Hysteresis (K) Transition sharpness −d(ln␳)/dT Metallic quality(1/K)

40 nm 3.817 0.003 −0.236 90 131 7 0.094 2.59 × 10−3

80 nm 3.821 0.003 −0.341 100 143 15 0.069 2.21 × 10−3
160 nm 3.825 0.003 −0.440 121 163 19 0.059 1.97 × 10−3

from 40 nm to 160 nm. More specifically, the transition sharpness

is smaller and the metallic quality is also inferior as the film thick-
ness is increased, indicating decreased metallic states in the thicker
sample. When the thin film thickness is raised, the lattice param-
eter is also increased and is close to the bulk value. The previous
study has shown that the rhombohedral symmetry of LAO substrate
increases the in-plane symmetry in the NNO thin film, which causes
Ni O Ni bond angles changed. When the thickness of NNO thin
film decreases, it suffers more influence from LAO substrate and
the Ni O Ni bond is straightened [24]. Therefore, more distortion
of NiO6 octahedra in the thicker film make the TMI increasing from
90 K to 121 K.
Temperature-dependent Raman spectra are shown in Fig. 3.
At high temperatures, NNO thin film possesses orthorhom-
Fig. 4. The Raman peak at 420 cm−1 changed with temperature for different thick-
bic distorted perovskite structure with space group Pbnm. ness NNO films. The arrows show the temperatures corresponding to the largest
peak shift.

According to group theory, there are 24 Raman active modes

Fig. 3. Temperature-dependent Raman spectra of NNO films at different thickness
(a) 40 nm (b)80 nm and (c)160 nm. (For interpretation of the references to colour in
the text, the reader is referred to the web version of this article.)
(Raman = 7A1g + 7B1g + 5B2g + 5B3g ) [25], but only part of them are
observed in the measured wavelength range due to poor intensity
or peak overlap. It clearly shows that the Raman spectra of the stud-
ied NNO films at the room temperature have five main Raman peaks
at 304 cm−1 (B1g ), 420 cm−1 (Ag ), 436 cm−1 (B2g ), 459 cm−1 (B3g ) and
486 cm−1 (Ag ), which matches with the literature. It directly proves
the formation of well-crystallized NNO thin films [26]. Some signif-
icant changes can be found in the Raman spectra when heating the
samples: (i) The intensity of peaks at 294 cm−1 decreases while the
intensity of peak around 304 cm−1 increases. The most dramatic
change of the two peaks occurs near TMI , which is indicated by a
red line in Fig. 3. At room temperature, 294 cm−1 peak almost dis-
appears and only 304 cm−1 peak is observed. (ii) The intensity of
the peak at 436 cm−1 becomes weaker, suggesting a smaller dis-
tortion of the NiO6 octahedra after phase transition [27]. (iii) The
peak at 420 cm−1 moves to higher wavenumber and the inten-
sity decreases. (iv) The peak at 459 cm−1 gradually weakens and
disappears finally.
All the changes on the spectroscopic features provide direct evi-
dence for a structural phase transition across TMI . Next, we further
discuss the temperature- and thickness- dependent signatures of
420 cm−1 peak. This mode is assigned to the Ni-O stretching mode
[28]. and is usually associated with vibrations of the distortions of
the NiO6 octahedra due to significant changes in the Ni O bond-
ing. Fig. 4 shows the peak position change with temperature for
different thickness NNO films. The mode hardens with increasing
temperature might be the signature of an enhancement of the Ni-O
stretching and decrease of the tilt level of NiO6 octahedra. On the
other hand, the peak shifts to lower frequency when the film thick-
ness increases. This is because the stress relaxation phenomenon
between the substrate and the thin film causes more tilt of NiO6
octahedra and weaken the metal property of thin film. The arrows
in Fig. 4 indicate the temperature corresponding to the largest
peak shift at different thickness,which is consistent with the TMI
obtained from the resistivity measurement result.
The MIT characteristic is also clearly presented in the
temperature-dependent infrared reflection measurement. Fig. 5
shows the real part of the optical conductivity 1 (ω)xtracted from
the infrared spectroscopy. At low temperature, the low inten-
sity of 1 (ω)uggests that the thin film is in an insulating state.
 T. Shao et al. / Applied Surface Science 399 (2017) 346–350
Fig. 5. Temperature-dependent real part of the optical conductivity of NNO films at
different thickness (a)40 nm (b)80 nm and (c)160 nm.
As the temperature raised, the 1 (ω)ntensity increases indicat-
ing that the thin film enters the metal state. It is interesting that
the 1 (ω)ntensity reaches the highest value when the tempera-
 the MIT of a NNO film.
ture is up to TMI After that, the 1 (ω)ntensity continously decreases
with the temperature increasing. This phenomenon is consistent
with resistivity-temperature measurement, where the resistivity

has a slight increase after TMI For further studying the dependance
of 1 (ω) temperature and thin film thickness, we integrated the
1 (ω)p to 3000 cm−1 to obtain the spectrum weight (SW) given
ω
by SW (␻) = 1 (ω) dω which represents the effective number of
0
carriers contributing to absorption below the cutoff frequency ␻.
As shown in Fig. 6, the SW value increases with the sample heated
 
to TMI nd then slowly decreases back to TMI This indicates that more
and more effective number of carriers exist in the sample when it
transforms from insulating state to metal state driven by temper-
ature. The effective number of carriers reaches the largest value at
 
TMI nd then slightly reduces above TMI in consistent with the change
of 1 (ω)ntensity. As the resistivity  has an inverse relationship
to the effective number of carriers, we can plot the reciprocal of
349
Fig. 6. The temperature-dependent spectrum weight (SW) and reciprocal of spec-
trum weight(SW−1 ) (insert) for the NNO films at different thickness.
spectrum weight (SW−1 ) (as shown in the insert of Fig. 6) and com-
pare it with resistivity measurement. It is clear that the SW−1 as
a function of temperature is in agreement with the result of the
temperature-dependent resistivity. Moreover, the influence of thin
film thickness on SW−1 is also as same as that of resistivity. This
consistence between spectroscopic and electrical results confirms
the fluctuation of carriers during the process of insulator-metal
transition.
In conclusion, we synthesized epitaxial NNO thin films on LAO
(001) using PLD method. Temperature-dependent resistivity mea-
surement shows a sharp metal-insulator transition in the thin
films. The phase transition temperature can be tuned from 90 K
to 121 K by changing the thin film thickness. XRD shows that the
thin films have good orientation and the out-of-plane lattice con-
stant increases with thickness. The level of NiO6 distortion changed
by thin film thickness is revealed by Raman spectroscopy, and
results in the TMI modification. In addition, optical conductivity
study shows the carrier variation in the process of phase transi-
tion which further confirms the resistivity result. The combination
of structural, electrical and optical studies provides a deep insight
to understand the influences of temperature and film thickness on
Acknowledgments
This work was supported by the National Natural Science Foun-
dation of China (Nos. 11275203, U1232128, 11275205), National
Key Scientific Instrument and Equipment Development Project
(2011YQ130018),Technological Development Grant of Hefei Sci-
ence Center of CAS(2014TDG-HSC002).
References
[1] G. Catalan, Phase Transit. 81 (2008) 729–749.
[2] Chadol Oh, Seungyang Heo, Hyun M. Jang, Junwoo Son, Appl. Phys. Lett. 108
(2016) 122106.
[3] Z. Yang, C. Ko, S. Ramanathan, Annu. Rev. Mater. Res. 41 (2011) 337–367.
[4] M.M. Qazilbash, M. Brehm, Byung-Gyu Chae, P.-C. Ho, G.O. Andreev, Bong-Jun
Kim, D.N. Basov, Science 318 (2007) 1750–1753.
[5] P. Lacorre, J.B. Torrance, J. Pannetier, A.I. Nazzal, P.W. Wang, T.C. Huang, J.
Solid State Chem. 91 (1991) 225.
350 T. Shao et al. / Applied Surface Science 399 (2017) 346–350
[6] P.-H. Xiang, N. Zhong, C.-G. Duan, X.D. Tang, Z.G. Hu, P.X. Yang, Z.Q. Zhu, J.H.
Chu, J. Appl. Phys. 114 (2013) 243713.
[7] X.K. Lian, F. Chen, X.L. Tan, L.F. Wang, X.F. Xuan, G.Y. Gao, S.W. Jin, W.B. Wu,
AIP Adv. 3 (2013) 062133.
[8] M. Liu, M. Kargarian, B. Kareev, P.J. Gray, W. Ryan, G.A. Freeland, J. Fiete, Nat.
Commun. 4 (2013) 2714.
[9] G. Catalan, R.M. Bowman, J.M. Gregg, Phys. Rev. B 62 (2000) 7892.
[10] F.Y. Bruno, K.Z. Rushchanskii, S. Valencia, Y. Dumont, C. Carrétéro, E. Jacquet,
R. Abrudan, S. Blügel, M. Ležaić, M. Bibes, Phys. Rev. B 88 (2013).
[11] D. Kaur, J. Jesudasan, P. Raychaudhuri, Solid State Commun. 136 (2005)
369–374.
[12] C. Girardot, F. Conchon, A. Boulle, P. Chaudouet, N. Caillault, J. Kreisel, R.
Guinebretière, F. Weiss, S. Pignard, Appl. Phys. Lett. 91 (2007),
192110–192110.
[13] M.K. Hooda, C.S. Yadav, B. Physica, Condens. Matter 491 (2016) 31–36.
[14] M. Hepting, D. Kukuruznyak, E. Benckiser, M. Le Tacon, B. Keimer, B. Physica,
Condens. Matter 460 (2015) 196–198.
[15] A. Kumar, P. Singh, D. Kaur, J. Phys. D: Appl. Phys. 39 (2006) 5310.
[16] R. Scherwitzl, Pavlo Zubko, I. Gutierrez Lezama, Shimpei Ono, Alberto F.
Morpurgo, Adv. Mater. 22 (2010) 5517–5520.
[17] S. Lafanea, T. Kerdjaa, B.D. Ngomb, S. Abdelli-Messaci, S. Malek, Appl. Surf. Sci.
269 (2013) 120–124.
[18] G. Catalan, R.M. Bowman, J.M. Gregg, J. Appl. Phys. 87 (2010) 606–608.
[19] A.E. Lee, C.E. Platt, J.F. Burch, R.W. Simon, J.P. Goral, M.M. Al-Jassim, Appl.
Phys. Lett. 57 (1999) 2019.
[20] M.T. Escote, A.M.L. da Silva, J. Solid State Chem. 151 (2000) 298–307.
[21] Adam J. Hauser, Evgeny Hauser, Nelson E. Mikheev, Jinwoo Hwang, Jack Y.
Zhang, Susanne Stemmer, Appl. Phys. Lett. 106 (2015) 092104.
[22] Patrick M. McBride, Anderson Janotti, Cyrus E. Dreyer, Burak Himmetoglu,
Chris G. Van de Walle, Appl. Phys. Lett. 107 (2015) 261901.
[23] A. Gupta, J. Narayan, T. Dutta, Appl. Phys. Lett. 97 (2010) 151912.
[24] C. Girardot, J. Kreisel, S. Pignard, N. Caillault, F. Weiss, Phys. Rev. B 78 (2008)
104101.
[25] H.C. Gupta, M.K. Singh, L.M. Tiwari, J. Phys. Chem. Solids 64 (2003) 531–533.
[26] N.D. Todorov, M.V. Abrashev, V.G. Ivanov, G.G. Tsutsumanova, V. Marinova,
Y.-Q. Wang, M.N. Iliev, Phys. Rev. B 83 (2011) 224303.
[27] P. Laffez, O.I. Lebedev, P. Ruello, R. Desfeux, G. Banerjee, F. Capon, Eur. Phys. J.
Appl. Phys. 25 (2004) 25–31.
[28] M.K. Stewart, J. Liu, M. Kareev, J. Chakhalian, D.N. Basov, Phys. Rev. Lett. 107
(2011) 176401.