Estimation of Residual Stress

Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Estimation of residual stress in Pb(Zr0.52Ti0.

48)O3/BiFeO3 multilayers deposited on


silicon
Shankar Dutta, Akhilesh Pandey, O. P. Thakur, Ramjay Pal, and Ratnamala Chatterjee

Citation: Journal of Applied Physics 114, 174103 (2013); doi: 10.1063/1.4828874


View online: http://dx.doi.org/10.1063/1.4828874
View Table of Contents: http://scitation.aip.org/content/aip/journal/jap/114/17?ver=pdfcov
Published by the AIP Publishing

Articles you may be interested in


Enhanced dielectric properties of Pb0.92La0.08 Zr0.52Ti0.48O3 films with compressive stress
J. Appl. Phys. 112, 114117 (2012); 10.1063/1.4768926

Improved electrical properties of PbZrTiO3/BiFeO3 multilayers with ZnO buffer layer


J. Appl. Phys. 112, 084101 (2012); 10.1063/1.4759123

Role of Pb(Zr0.52Ti0.48)O3 substitution in multiferroic properties of polycrystalline BiFeO3 thin films


J. Appl. Phys. 110, 114116 (2011); 10.1063/1.3668123

Residual stress and magnetic behavior of multiferroic CoFe 2 O 4 / Pb ( Zr 0.52 Ti 0.48 ) O 3 thin films
J. Appl. Phys. 105, 084113 (2009); 10.1063/1.3115452

Lattice strain and lattice expansion of the SrRuO 3 layers in SrRuO 3 / PbZr 0.52 Ti 0.48 O 3 / SrRuO 3
multilayer thin films
J. Appl. Phys. 92, 101 (2002); 10.1063/1.1483369

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. IP: 200.130.19.131 On: Mon, 22 Feb 2016 18:08:51
JOURNAL OF APPLIED PHYSICS 114, 174103 (2013)

Estimation of residual stress in Pb(Zr0.52Ti0.48)O3/BiFeO3 multilayers


deposited on silicon
Shankar Dutta,1,a) Akhilesh Pandey,1 O. P. Thakur,1 Ramjay Pal,1
and Ratnamala Chatterjee2
1
Solid State Physics Laboratory, DRDO, Lucknow Road, Timarpur, Delhi–110054, India
2
Department of Physics, IIT Delhi, Hauz Khas, New Delhi–110016, India
(Received 7 July 2013; accepted 21 October 2013; published online 4 November 2013)
Thin multilayer films possess residual stress components which vary from microscopic to
macroscopic scale. In this study, Pb(Zr0.52Ti0.48)O3/BiFeO3 (PZT-BFO) multilayer thin film is
deposited via chemical solution deposition technique on silicon substrate. The microscopic and
macroscopic residual stress components of the multilayer films are investigated. The average
microscopic residual stress is estimated to be 791.15 MPa (tensile) by using x-ray diffraction
technique; on the other hand, the average macroscopic stress is found to be 774.23 MPa (tensile)
by using wafer curvature measurement technique. As the thermally grown SiO2 layer possesses
compressive stress, the combined residual stress of the PZT-BFO multilayer and SiO2 will almost
cancel each other. This is reasonably encouraging for integration of the multilayer in MEMS
C 2013 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4828874]
structures. V

I. INTRODUCTION Residual stresses can be discussed in terms of length


over which the stresses equilibrate.16 The microscopic resid-
In the recent years, multiferroic materials have attracted
ual stresses equilibrate over atomic dimensions and balance
significant attention for their concurrent ferroelectric and mag-
within a grain or even over a number of grain dimensions.
netic properties. Among them, bismuth ferrite (BiFeO3) has
The micro-stresses can be estimated by using x-ray diffrac-
been widely studied due to its multiferroic properties at room
tion (XRD) technique.16–18 Long range (macroscopic)
temperature.1–4 But the application of BiFeO3 (BFO) is seri-
stresses equilibrate over dimensions of the scale of the
ously hindered due to its small remnant polarization and high
structure.16,19 Such stresses often calculated using wafer cur-
leakage current (even at low electric field). Significant efforts
vature measurement technique which ignores the polycrys-
have been made to prepare high quality multiferroics (low
talline or multiphase nature of the material.
leakage current, enhanced dielectric constant, etc.) either by
In this report, the PZT-BFO multilayer thin film is depos-
substitution at A/B site into the perovskite crystal (ABO3)5–8
ited on the silicon substrate. The microscopic stress associated
or by forming solid solution with other ferroelectric
with the PZT-BFO multilayer thin film is estimated by using
materials.9–11 This can also be obtained “artificially” by com-
XRD technique; on the other hand, macroscopic residual
bining two or more ferroelectric/multiferroic materials.6–8,12,13
stress of the film is estimated by wafer curvature techniques.
The combination of PZT (PbZrTiO3) - BFO multilayer
thin film is one of the most promising “artificial multiferroic”
II. EXPERIMENTAL
and is found to possess superior dielectric and ferroelectric
properties.6–8 Different techniques, like pulsed laser deposi- In this work, PZT-BFO multilayer thin film is deposited
tion (PLD)5 and sol–gel,6–9,14,15 are being used to deposit the on the 50 mm diameter oxidized silicon wafer (n-type, h100i
PZT-BFO multilayer thin films. All of the depositions were orientation) for estimation of the residual stress. A buffer/seed
done on platinized silicon wafers. But integration of the layer of ZrO2 is introduced to deposit crystalline PZT-BFO
PZT-BFO multilayer thin films with non-platinum coated sil- multilayer on amorphous SiO2 layer. The 0.3 M ZrO2 pre-
icon wafers is equally important especially for realization of cursor20 sol was prepared from zirconium n-propoxide
variety of multiferroic based smart MEMS devices. Keeping Zr[O(CH2)2CH3]4 (Alfa Aesar) in 2-methoxyethanol (Sigma
in mind the MEMS application, deposition/integration of Aldrich). Details of the preparation of PZT and BFO precur-
PZT-BFO multilayer thin films on SiO2 coated silicon wafers sors have been reported elsewhere.14 The ZrO2 buffer layer
is being reported first time. was spin-coated (Karl Suss make spin-coater) on a SiO2
Although microstructure and ferroelectric/dielectric coated Si (100) wafer at 5000 rpm for 45 s and then baked on
properties of the PZT-BFO multilayer films have been exten- hot plate at 250  C for 5 min. Subsequently, the ZrO2 film was
sively studied, associated residual stress generated due to the annealed at 700  C in a preheated tube furnace for 30 min in
deposition of the multilayer thin films have not been system- air. After depositing the ZrO2 buffer layer, PZT sol was spun
atically carried out. As the generation of residual stress may on the substrate at 5000 rpm for 45 s and dried at 250  C for
affect the performance of the multiferroic based MEMS de- 5 min. Thereafter, BFO sol was spin coated and dried under
vice, estimation of the residual stress is very important. the identical conditions. This procedure was repeated number
of times, resulting three layers of PZT and two layers of BFO.
a)
Electronic mail: [email protected] Finally, the multilayer was annealed at 650  C in the tube

0021-8979/2013/114(17)/174103/5/$30.00 114, 174103-1 C 2013 AIP Publishing LLC


V

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. IP: 200.130.19.131 On: Mon, 22 Feb 2016 18:08:51
174103-2 Dutta et al. J. Appl. Phys. 114, 174103 (2013)

furnace for 30 min in air. To check the ferroelectric nature of From literature, Elastic modulus and Poisson’s ratio for the
the multilayer film, electric field induced polarization (P-E PZT thin film are found to be 130 GPa and 0.3, respec-
loop) measurement has been performed. Since the PZT-BFO tively;17 and for BFO thin film 133 GPa and 0.25.21
multilayer was deposited on SiO2/Si, there is no bottom elec- The macroscopic/average in-plane residual stress in the
trode. Therefore, the electrical properties of the multilayer thin films is determined by ex-situ wafer-curvature measure-
thin film were studied in-plane by generating inter-digitated ments using TOHO FLX-2320S laser reflectance system. An
terminal (IDT) patterns on top of the film. The multilayer film accurate fit for the radius of curvature of the substrate was
exhibited remnant polarization (Pr) of 30 lC/cm2 and coer- determined using a line scan consisting of 50 points. To av-
cive field of (Ec) 214 kV/cm. Dielectric constant (room tem- erage out the residual stress value, two curvature measure-
perature) of the multilayer films is estimated to be 650 at ments are taken along the diameter—first, perpendicular to
1 KHz. Details on the study of electrical properties of these the wafer’s primary flat and second, parallel to the primary
films are being communicated separately. flat. Initially, radius of curvature (R0) of the oxidized silicon
Phase and crystalline structure of the PZT-BFO film was substrate is measured as the reference point. Subsequently,
analyzed using high resolution XRD (PANalytical PW the wafer curvatures are measured after ZrO2, PZT-BFO
3050/65 X-Pert Pro MRD HRXRD) in h–2h geometry as multilayer depositions on the oxidized silicon. The changes
shown in Fig. 1. It is well known that changes in the crystal- in radius of curvatures (R) after each film deposition can
lographic inter-planer spacing (dhkl) can be used with Bragg then be used to calculate the residual stress (r) in that partic-
equation to detect elastic strain in the crystal. To estimate ular film via the famous Stoney’s equation16,19
the changes in the crystallographic inter-planer spacing   2  
(dhkl), XRD data are recorded around the corresponding dif- Es ts 1 1
r¼  : (3)
fraction angle (2h) of the plane at different incident angles 1   s 6tf R R0
(w). Then strain (e) in the crystal can be found out by using
Eq. (1). Residual stress (r) of the sample can be estimated This well-known equation is valid only in the thin- film re-
from the slope of e and sin2w and by using Eq. (2).16–18 gime (ts  tf). Hence, the calculation of residual stress in the
film only requires knowledge of substrate Young’s modulus
d0ðhklÞ  dðhklÞ (Es) and Poisson’s ratio ( s) and does not rely on the material
eðhkl:wÞ ¼ ; (1)
d0ðhklÞ properties of the film. For Silicon (100) wafer, Es ¼ 130 GPa
  and  s ¼ 0.28.22
eðhkl:wÞ Ef
r¼ : (2)
Sin2 w ð1 þ  f Þ
III. RESULTS AND DISCUSSIONS
Here, d(hkl) and d0(hkl), respectively, indicate the d spacing of Fig. 1 shows the XRD patterns of the buffer ZrO2 and
the (hkl) plane in the film and that in the bulk. Elastic modu- PZT-BFO multilayer films. The diffraction pattern shows
lus is Ef and  f is Poisson’s ratio. For the present PZT-BFO that the deposited ZrO2 films are polycrystalline in nature
(3:2) multilayer film, the effective elastic modulus and having orientations along (110), (211), (220), (003), and
Poisson’s ratio are estimated by weighted average (310) planes. From the XRD peaks at 30 , 50 , and multiple
3x EfðPZTÞ þ2x EfðBFOÞ peaks around 55 , one can visualize that the ZrO2 film de-
EfðPZT=BFOÞ ¼ and posited on the Si sample is monoclinic.20 The diffraction pat-
5
3x  fðPZTÞ þ2x  fðBFOÞ tern of PZT-BFO multilayer shows polycrystalline
 fðPZT=BFOÞ ¼ : perovskite structure having contribution of several orienta-
5
tions, namely (100), (110), (111), (200), (210), (211), (220),
and (310). The XRD data also show a sharp peak corre-
sponding to the Si (004) contributed from silicon substrate.
It is also well known that for the estimation of residual
stress by XRD, higher angle peak should be chosen and at
the same time the peaks should have sufficient intensity.
From the XRD pattern in Fig. 1, one can see that the peaks
corresponding to (200) and (211) planes have sufficient in-
tensity for this analysis. Hence, XRD data are recorded
around these planes at 8 different X-ray incident angles w
(0 , 15.50 , 22.21 , 27.58 , 32.31 , 36.70 , 40.89 , and 45 )
as shown in Fig. 2. The values of w are chosen to get equidis-
tant Sin2w values during d-spacing vs. sin2w plot. Fig. 3
shows the variation of d-spacing of (200) and (211) planes
with Sin2w. Corresponding residual stress values for the mul-
tilayer thin film are estimated by using XRD peaks and
PANalytical X0 Pert-stressTM software based on Eqs. (1) and
(2). The residual stress values are found to be tensile in na-
FIG. 1. XRD pattern of PZT-BFO multilayer thin film and ZrO2 buffer layer. ture. The estimated stress values are 825.4 MPa and 756.9

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. IP: 200.130.19.131 On: Mon, 22 Feb 2016 18:08:51
174103-3 Dutta et al. J. Appl. Phys. 114, 174103 (2013)

FIG. 3. Plot of d-spacing of (a) (200) and (b) (211) planes vs. Sin2w.

FIG. 2. XRD peaks of (a) (200) and (b) (211) planes at different incident Individual PZT and BFO layers are not distinguishable in
angles (w).
SEM cross-section. This is because the interfaces of the
layers are not abrupt. As all the layers are annealed at the
MPa corresponding to the (200) and (211) planes, respec- final stage, inter-diffusion of elements might have occurred
tively. The average residual stress of the PZT-BFO multi- across the individual layers. From Tables I and II, the ZrO2
layer estimated from the XRD technique is therefore buffer layer is found to possess 279.88MPa (average) macro-
791.15MPa. scopic stress. The macroscopic residual stress (tensile) val-
Residual stress generated due to deposition of ZrO2, ues are found to be 803.63 MPa and 744.84 MPa
PZT-BFO multilayer thin films are calculated from the corresponding to wafer curvature measured perpendicular
changes in radius of curvature of the wafer measured perpen- and parallel to primary flat, respectively. The average macro-
dicular and parallel to primary flat as shown in Figs. 4(a) and scopic residual stress is found to be 774.23MPa.
4(b), respectively. The sol-gel deposited PZT-BFO multilayer thin film
The measured radii of curvature of the sample along possesses residual stress components, which varies from mi-
perpendicular and parallel to primary wafer flat are reported croscopic to macroscopic range. The microscopic stress
in Tables I and II, respectively. Thickness of the ZrO2, arises due to lattice mismatch between PZT and BFO layers,
PZT-BFO multilayer thin films is measured from SEM constrained densification and shrinkage during annealing,
cross-section as shown in Fig. 5. Thickness of the ZrO2 seed formation of inter-granular stresses as anisotropic grains
layer is found to be 40 nm; on the other hand, thickness of grow, as well as the stress caused by the transformation from
the PZT-BFO multilayer thin film is found to be 130 nm. one phase to another. The macroscopic component of the

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. IP: 200.130.19.131 On: Mon, 22 Feb 2016 18:08:51
174103-4 Dutta et al. J. Appl. Phys. 114, 174103 (2013)

FIG. 5. SEM cross-section of the PZT-BFO multilayer.

residual stress arises from mismatches of thermo-elastic


coefficients and/or crystal lattice parameters of the ZrO2,
PZT, and BFO layers and substrate (SiO2/Si). The high
dielectric constant (650) may be attributed to the large ionic
displacement in the distorted perovskite structure due to the
residual stress in the multilayer thin film. Differential ther-
mal contraction stresses are an unavoidable consequence of
processing/annealing at elevated temperature (650  C) and
can become quite large because coating and substrate often
have very different thermal expansion characteristics.
It is well known that the thermally grown SiO2 layer on
Si possesses a compressive stress 1000 MPa.22 Therefore,
residual stress of PZT-BFO multilayer and SiO2 will almost
cancel each other (slightly compressive). This is quite favor-
able during integration of PZT-BFO multilayer and Si for
FIG. 4. Cuvature of the oxidized silicon wafer before and after the multi- MEMS devices.
layer deposition: (a) Perpendicular to primary flat; (b) parallel to primary
flat.
IV. CONCLUSIONS
In this work, PZT-BFO multilayer thin film is deposited
TABLE I. Wafer curvature measured perpendicular to primary flat. via sol-gel technique on SiO2/Si substrate. The deposited
multilayer thin film possesses tensile residual stress values,
Residual both microscopic and macroscopic, as estimated by using
Radius of Substrate/ stress in XRD and wafer curvature measurement techniques. The av-
Bow curvature layer thin film erage microscopic residual stress is found to be 791.15 MPa;
Substrate/layer (lm) (m) thickness (MPa)
on the other hand, the average macroscopic stress is esti-
Oxidized Si substrate 13.88 15.263 280 lm – mated to be 774.23 MPa. Thus the residual stress of the
Buffer layer (ZrO2) 16.20 14.177 40 nm 296.01 PZT-BFO multilayer thin film and that of SiO2 almost can-
PZT-BFO multilayer 25.82 8.102 130 nm 803.63 celled each other. This is quite favorable for future integra-
tion of PZT-BFO multilayer thin film in MEMS devices.

ACKNOWLEDGMENTS
TABLE II. Wafer curvature measured parallel to primary flat.
The authors acknowledge Director SSPL for his continu-
Residual
Radius of Substrate/ stress in
ous support and for the permission to publish this work. Help
Bow curvature layer thin film from other colleagues is also acknowledged.
Substrate/layer (lm) (m) thickness (MPa)
1
N. A. Spaldin, S. W. Cheong, and R. Ramesh, Phys. Today 63(10), 38
Oxidized Si substrate 3.42 62.314 280 lm – (2010).
Buffer layer (ZrO2) 7.28 48.733 40 nm 263.76 2
A. K. Pradhan, K. Zhang, D. Hunter, J. B. Dadson, G. B. Loutts, P.
PZT-BFO multilayer 14.31 14.343 130 nm 744.84 Bhattacharya, R. Katiyar, J. Zhang, D. J. Sellmyer, U. N. Roy, Y. Cui, and
A. Burger, J. Appl. Phys. 97, 093903 (2005).

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. IP: 200.130.19.131 On: Mon, 22 Feb 2016 18:08:51
174103-5 Dutta et al. J. Appl. Phys. 114, 174103 (2013)

3 13
S. K. Singh, Y. K. Kim, H. Funakubo, and H. Ishiwara, Appl. Phys. Lett. J. Li, P. Li, G. Zhang, J. Yu, Y. Wu, and X. Wen, Thin Solid Films 519,
88, 162904 (2006). 6021 (2011).
4 14
K. Y. Yun, M. Noda, M. Okuyama, H. Saeki, H. Tabata, and K. Saito, S. Dutta, A. Pandey, I. Yadav, O. P. Thakur, R. Laishram, R. Pal, and R.
J. Appl. Phys. 96, 3399 (2004). Chatterjee, J. Appl. Phys. 112, 084101 (2012).
5 15
F. Yan, T. J. Zhu, M. O. Lai, and L. Lu, J. Appl. Phys. 110, 114116 Y. W. Li, J. L. Sun, J. Chen, X. J. Meng, and J. H. Chu, Appl. Phys. Lett.
(2011). 87, 182902 (2005).
6 16
S. Iakovlev, C.-H. Solterbeck, M. Kuhnke, and M. Es-Souni, J. Appl. P. J. Withers and H. K. D. H. Bhadeshia, Mater. Sci. Technol. 17, 355
Phys. 97, 094901 (2005). (2001).
7 17
S. H. Jo, S. G. Lee, and Y. H. Lee, Nanoscale Res. Lett. 7, 54 (2012). X. Zheng, J. Li, and Y. Zhou, Acta Mater. 52, 3313 (2004).
8 18
S. H. Jo, S. P. Nam, S. G. Lee, S. H. Lee, Y. H. Lee, and Young-Gon Kim, R. J. Ong, T. A. Berfield, N. R. Sottos, and D. A. Payne, J. Eur. Ceram.
Trans. Electr. Electron. Mater. 12, 193 (2011). Soc. 25, 2247 (2005).
9 19
V. Stancu, C. Dragoi, V. Kuncser, G. Schinteie, L. Trupina, E. Vasile, and K. Fujito, N. Wakiya, N. Mizutani, and K. Shinozaki, Jpn. J. Appl. Phys.
L. Pintilie, Thin Solid Films 519, 6269 (2011). Part 1 44, 6900 (2005).
10 20
P. Murugavel, M. P. Singh, W. Prellier, B. Mercey, Ch. Simon, and B. S. Dutta, A. Pandey, I. Yadav, O. P. Thakur, A. Kumar, R. Pal, and R.
Raveau, J. Appl. Phys. 97, 103914 (2005). Chatterjee, J. Appl. Phys. 114, 014105 (2013).
11 21
D. A. Sanchez, N. Ortega, A. Kumar, R. Roque-Malherbe, R. Polanco, S. L. Shang, G. Sheng, Y. Wang, L. Q. Chen, and Z. K. Liu, Phys. Rev. B
J. F. Scott, and R. S. Katiyar, AIP Adv. 1, 042169 (2011). 80, 052102 (2009).
12 22
N. Wang, J. Cheng, A. Pyatakov, A. K. Zvezdin, J. Li, L. E. Cross, and D. M. J. Madau, Fundamental of Microfabrication–The Science of
Viehland, Phys. Rev. B 72, 104434 (2005). Miniaturization (CRC Press, 2002).

Reuse of AIP Publishing content is subject to the terms at: https://publishing.aip.org/authors/rights-and-permissions. IP: 200.130.19.131 On: Mon, 22 Feb 2016 18:08:51

You might also like