Skip to content


  • Nano Express
  • Open Access

Nano-embossing technology on ferroelectric thin film Pb(Zr0.3,Ti0.7)O3 for multi-bit storage application

Nanoscale Research Letters20116:474

  • Received: 24 December 2010
  • Accepted: 27 July 2011
  • Published:


In this work, we apply nano-embossing technique to form a stagger structure in ferroelectric lead zirconate titanate [Pb(Zr0.3, Ti0.7)O3 (PZT)] films and investigate the ferroelectric and electrical characterizations of the embossed and un-embossed regions, respectively, of the same films by using piezoresponse force microscopy (PFM) and Radiant Technologies Precision Material Analyzer. Attributed to the different layer thickness of the patterned ferroelectric thin film, two distinctive coercive voltages have been obtained, thereby, allowing for a single ferroelectric memory cell to contain more than one bit of data.


  • Remnant Polarization
  • Ferroelectric Thin Film
  • Ferroelectric Film
  • Emboss Process
  • Piezoresponse Force Microscopy


The development of miniaturized ferroelectric field effect transistors (FeFETs) and random access memories (FeRAMs) [1, 2] has called for fabrication of high-quality ferroelectric nanostructures. How to retain excellent ferroelectricity in nanoscale patterned structures posts a great challenge, as the small thickness of ultra thin films as well as the damages and defects introduced by the conventional photo lithography [36] could drastically degrade the ferroelectric properties. Better controlling the quality of the ultra-thin ferroelectric films and alternative patterning techniques are, therefore, highly required.

It is widely understood that multi-bit operation could be one of the most efficient approaches to increase storage densities. In recent years, a great deal of efforts has been made on realizing multi-value storage through circuit design. One of the drawbacks is the additional budget of densities in circuit integration. There have been rarely reports on the research tackling the improvement of fabrication processes and device structures. Nano-embossing technology has emerged as a fast and cost effective technique suitable for patterning structures with feature size down to 20 nm, well below the limit of other lithography techniques used for mass production [711]. In this article, we report our initial progress in developing a nano-embossing technique to achieve large arrays of ferroelectric PZT cells, which have potential application in multi-bit storage based on ferroelectric nanostructures.

Since the principle of FeRAM is based on the polarization reversal by an externally applied electric field of metal-ferroelectric-metal capacitors, the computational '0'and '1' are represented by the nonvolatile storage of the negative or positive remnant polarization state, respectively [12]. The ferroelectric films with different thickness need different coercive voltages. A staggered structure with two distinct thickness layers on a PZT thin film can be readily created by a one-step embossing process. In principle, the thinner layer should give rise to the lower switch voltage and the thicker layer to the higher switch voltage [13]. The voltage magnitude corresponds to one of the two polarization charge levels stored in the ferroelectric memory cell. In this way, multiple bits of data can be obtained from a single ferroelectric memory cell by applying different voltages.

Scanning atomic force microscopy (AFM) was used to study the morphology of the embossed arrays, whereas the PFM, which is proved to be one of the most effective method for the nanoscale study and control of ferroelectric domains in bulk crystals and thin films [14, 15], was applied for the study of polarization switching behavior of patterned regions on a PZT film. A Radiant Technologies Precision Material Analyzer was also used for electrical characterizations. The same measurements were also performed on un-patterned regions for comparison.

Results and discussion

Figure 1a illustrates the nano-embossing process of ferroelectric PZT. Figure 1b displays the embossed PZT film profiles measured by AFM; showing an embossed depth of about 160 nm on a 450-nm thick PZT film. It could be seen that well quadrate patterned profiles of the PZT film has been obtained by the nano-embossing process. After crystallization, the morphology of the embossed region remained stable and no collapse was found even after several months. The embossed PZT films were found to be with [111] preferred orientation and a tetragonal structure previously by X-ray diffraction and Raman spectroscopy (not presented here) [13, 16], respectively (Additional files 1 and 2). Figure 2 shows the hysteretic dependence of out-of-plane piezoresponse (OPP) phase under a bias from -10 to 10 V. The blue and red curves represent the hysteresis loops of embossed top and bottom areas, respectively. The polarization phase contrast is nearly 180° and the switch voltage of embossed bottom area (thinner layer) is about 1.5 V, approximately 1.5 V smaller than the coercive voltage of embossed top area (thicker layer). Rapid saturations are found at around 3 and 4 V, respectively. This provides convincing evidence of excellent ferroelectricity property of the embossed regions and demonstrates that two distinct coercive voltages can be generated by a one-step embossing process on ferroelectric thin films.
Figure 1
Figure 1

PZT films nano-embossing process and results of embossed profiles. (a) A schematic diagram illustrating a one-step embossing process to form a stagger shape in a ferroelectric PZT film with two different thicknesses. (b) AFM images of embossed stagger like profiles of ferroelectric PZT film.

Figure 2
Figure 2

Hysteretic dependence of OPP phase with applied voltage from -10 to 10 V for embossed top (blue) and bottom regions (red).

We propose a method for operation of multi-bit storage in embossed regions, which is schematically illustrated in Figure 3a. For instance, after a minus bias -V2, which is larger than the coercive voltage of the thicker film, is added both on the embossed and un-embossed regions, the polarization of the ferroelectric film under electrode pads could be wholly switched upward. Then with a plus voltage V1 applied on, which is smaller than the coercive voltage of the thicker film and larger than the coercive voltage of the thinner film, the polarization of the embossed bottom area can be switched downward. While for the embossed top area and un-embossed region, the polarization still keeps upward. In this way, an additional storage state for embossed region is apparently achieved. Indeed, Figure 3b plots the ratio of the remnant polarization of the embossed region to that of the un-embossed region on the same PZT film under voltages from 1 to 10 V. In the range of small voltages from 1 to 3 V, the polarization of the embossed area is about four times as large as that of the un-embossed region, and, with the voltage increasing to 5 V, the ratio decreases close to 1. This is attributed to the fact that the thinner layer in the embossed area could be more switched than the un-embossed region (coercive voltage is approx. 3 V) under low voltages such as 1 to 3 V. With the voltage rising to 5 V and above, both the embossed region and un-embossed region could be almost switched, and thus the remnant polarization for both regions with the same electrode pad areas (100 × 100 μm square) approaches a similar value.
Figure 3
Figure 3

Sketch map of multi-bit storage operation and remnant polarization comparison between an embossed and an un-embossed region. (a) Schematically illustration of multi-bit storage operation for embossed regions on a PZT film. (b) Remnant polarization ratio of the embossed and the un-embossed regions in a PZT film in the voltage range from 1 to 10 V.

Figure 4 shows the hysteresis loops of the embossed region under external voltages of 5 and 3 V, respectively, as well as that of the un-embossed region on the same PZT thin film under 3 V for comparison. All of the hysteresis loops under varies voltages were measured from a pre-polarized state. The remnant polarization under 3 V for embossed region is 5.04 μc/cm2, while for un-embossed region is only 1.24 μc/cm2. This suggests two remnant polarization states can be created for '10' and '01' states by switching the polarization of thinner layer at a lower voltage. Correspondingly, when the polarization of whole layer was switched at a higher voltage, another two storage states '11' and '00' are obtained. As illustrated schematically in Figure 4, four storage states can be achieved in the embossed region.
Figure 4
Figure 4

Hysteresis loops of the embossed region in the voltages of 3 and 5 V, respectively. The same loops obtained from the un-embossed region of the same PZT thin film is obtained under 3 V for comparison.

Fatigue measurements were done at room temperature using a Radiant Technologies Precision Tester. Both embossed and un-embossed regions were subjected to bipolar square wave voltage cycling with a width of 0.5 μs and period of 1 μs. Figure 5 displays the change in switchable polarization as a function of the number of switching cycles for both embossed and un-embossed regions on the same PZT thin film. It can be seen that un-embossed region is nearly fatigue free through 107 cycles under a voltage of 5 V, while the switchable polarization undergoes a slow decay starting at around 106 switching cycles for the embossed region under the same voltage. This could be attributed to the fact that the polarization of the thinner region in the embossed area was much more switched than the un-embossed region (coercive voltage is approx. 3 V) under the applied voltage, and thus underwent an earlier decay. Indeed, the switchable polarization of the embossed region reached as many as 107 switching cycles with no noticeable decay under a smaller voltage of 3 V. Further study on the fatigue [1721] of the patterned ferroelectric film capacitor structure is underway. The retention characteristics of an embossed region have been obtained by a PUND method with the pulse width and amplitude of 1 ms and ± 8 V, respectively. Figure 6 shows the time dependence of the polarization, where P s is the switched polarization between the two opposite pulses and P r is the remnant polarization. As it can be seen that there is no polarization descending trend observed even after retention time of 3200 s, clearly confirming the excellent polarization phase maintained after the removal of external fields for a long time.
Figure 5
Figure 5

Fatigue measurements from both embossed and un-embossed areas. Change in switchable polarization as a function of the number of switching cycles for both embossed region with the biases of 5 and 3 V, respectively. Also shown in the figure is the switching behavior from the un-embossed region on the same PZT thin film under 5 V.

Figure 6
Figure 6

Time dependence of the polarization on an embossed region from PUND measurements at room temperature.


In summary, we have successfully demonstrated a new method to fabricate multi-bit memory devices by embossing on thin PZT films at room temperature to form a stagger structure. More than one bit of data is obtained from a single ferroelectric memory cell with such a stagger configuration. Our process has the advantages of high throughput and low cost with a prospect for fabricating multi-bit memory devices by the developed embossing technique.

Experiment details

Ferroelectric PZT thin films were prepared on Pt/Ti/SiO2/Si substrates by the sol-gel method. The raw materials were lead acetate trihydrate [Pb(OCOCH3)2.3H2O, 99.5%], zirconium tetra n-propoxide (Zr(OC3H8)4, 70%), titanium (IV) butoxide (Ti(OC4H9)4, 98%) as precursor material and Methanol/Acetic acid mixed solvent as a solvent. Figure 1a illustrates the embossing process of the PZT film. After spin-on, the precursor film was first baked on hotplate at 60°C in air for 5 min. Then, an embossing process was carried out at room temperature under a pressure of 9 Mpa for 15 min using a silicon template, which has a grating 500 nm lines/spaces. The template used was first coated with an anti-stick layer on its surface in order to reduce its adhesion to the embossed gels and make it easier in later de-mold procedure. After embossing, the gel layers were first pyrolyzed in air on a hotplate at 350°C for 5 min and then crystallized by conventional thermal annealing in air at 650°C for 15 min.

The PFM measurements of ferroelectric characteristics of structured PZT films were carried out at room temperature by a commercial multimode AFM with a Pt coated cantilever (force constant 0.03 to 0.2 N/m, and resonant frequency 14 to 28 kHz). An ac voltage of 1 V and frequency about 200 kHz was applied to measure the in-field nanoscale hysteresis loops.

Further electrical characterization of the embossed and un-embossed PZT ferroelectric films were performed using a Radiant Technologies Precision Material Analyzer with a triangular wave form at 1 kHz after forming Au/Cr electrode pads (100 × 100 μm square) on the films (Figure 3a).



atomic force microscopy


ferroelectric field effect transistors


ferroelectric random access memories


out-of-plane piezoresponse


piezoresponse force microscopy.



This work was financially supported by National Basic Research Program of China (2011CBA00603) and the 985" Micro/nanoelectronics Science and Technology Innovation Platform at Fudan University.

Authors’ Affiliations

ASIC & System State Key Laboratory, Department of Microelectronics, Fudan University, Shanghai, 200433, China
Department of Material Science, Fudan University, Shanghai, 200433, China
Rutherford Appleton Laboratory, Chilton, Didcot, Oxfordshire, OX11 0QX, UK


  1. Scott JF: Application of modern ferroelectrics. Science 2007, 315: 954–959. 10.1126/science.1129564View ArticleGoogle Scholar
  2. Waser R, Rüdiger A: Ferroelectrics: pushing towards the digital storage limit. Nat Mater 2004, 3: 81–82. 10.1038/nmat1067View ArticleGoogle Scholar
  3. Lee W, Han H, Lotnyk A, Schubert MA, Senz S, Alexe M, Hesse D, Baik S, Gösele U: Indiviually addressable epitaxial ferroelectric nanocapacitor with near tb inch -2 density. Nat Nanotechnol 2008, 3: 402–407. 10.1038/nnano.2008.161View ArticleGoogle Scholar
  4. Cofer A, Rajora P, Deornellas S, Keil D: Plasma etch processing of advanced ferroelectric devices. Integr Ferroelectr 1997, 16: 53–61. 10.1080/10584589708013029View ArticleGoogle Scholar
  5. Deornellas S, Rajora P, Cofer A: Challenges for plasma etch integration of ferroelectric capacitors in FeRAM's and DRAM's. Integr Ferroelectr 1997, 17: 395–402. 10.1080/10584589708013014View ArticleGoogle Scholar
  6. Ishihara K, Ishikawa T, Hamada K, Onishi S, Kudo J, Sakiyama K: The degradation of ferroelectric properties of pzt thin films due to plasma damage. Integr Ferroelectr 1995, 6: 301–307. 10.1080/10584589508019373View ArticleGoogle Scholar
  7. Schift H: Nanoimprint lithography: an old story in modern times? A review. J Vac Sci Technol B 2008, 26: 458–480. 10.1116/1.2890972View ArticleGoogle Scholar
  8. Guo LJ: Nanoimprint lithography: methods and material requirements. Adv Mater 2007, 19: 495.View ArticleGoogle Scholar
  9. Hu ZJ, Tian M, Nysten B, Jonas AM: Regular arrays of highly ordered ferroelectric polymer nanostructrues for non-volatile low-voltage memories. Nat Mater 2009, 8: 62–67. 10.1038/nmat2339View ArticleGoogle Scholar
  10. Zhang L, Ducharme S, Li JY: Microimprinting and ferroelectric properties of poly(vinylidene fluoride-trifluoroethylene) copolymer films. Appl Phys Lett 2007, 91: 172906–172908. 10.1063/1.2800803View ArticleGoogle Scholar
  11. Liu Y, Weiss DN, Li J: Rapid nanoimprinting and excellent piezorepsonse of polymeric ferroelectric nanostructures. ACS Nano 2010, 4: 83–90. 10.1021/nn901397rView ArticleGoogle Scholar
  12. Setter N, Damjanovic D, Eng L, Fox G, Gevorgian S, Hong S, Kingon A, Kohlstedt H, Park NY, Stephenson GB, Stolitchnov I, Taganstev AK, Taylor DV, Yamada T, Streiffer S: Ferroelectric thin films: review of materials, properties, and applications. J Appl Phys 2006, 100: 051606–051651. 10.1063/1.2336999View ArticleGoogle Scholar
  13. Shen ZK, Chen ZH, Qiu ZJ, Lu BR, Wang J, Deng SR, Jiang AQ, Qu XP, Liu R, Chen YF: Influences of embossing technology on Pb(Zr 0.3 ,Ti 0.7 )O 3 ferroelectric thin film. Micorelectron Eng 2010, 87: 869–871. 10.1016/j.mee.2009.12.049View ArticleGoogle Scholar
  14. Gruverman A, Kalinin SV: Piezoresponse force microscopy and recent advances in nanoscale studies of ferroelectrics. J Mater Sci 2006, 41: 107–116. 10.1007/s10853-005-5946-0View ArticleGoogle Scholar
  15. Ivry Y, Chu DP, Durkan C: Nanometer resolution piezorepsonse force microscopy to study deep submicron ferroelectric and ferroelastic domains. Appl Phys Lett 2009, 94: 162903–162905. 10.1063/1.3105942View ArticleGoogle Scholar
  16. Shen Z, Chen Z, Lu Q, Jiang A, Qiu Z, Qu X, Chen Y, Liu R: Characterizations of nanoembossed Pb(Zr 0.3 ,Ti 0.7 )O 3 ferroelectric films. J Vac Sci Technol B 2010, 28: C6M28-C6M31. 10.1116/1.3497018Google Scholar
  17. Brazier M, McElfresh M, Mansour S: Origin of anomalous polarization offsets in compositonally graded Pb(Zr, Ti)O 3 thin films. Appl Phys Lett 1999, 74: 299–301. 10.1063/1.123004View ArticleGoogle Scholar
  18. Scott JF, Dawber M: Oxygen-vacancy ordering as a fatigue mechanism in perovskite ferroelectrics. Appl Phys Lett 2000, 76: 3801–3803. 10.1063/1.126786View ArticleGoogle Scholar
  19. Warren WL, Tuttle BA, Dimos DB: Ferroelectric fatigue in perovskite oxides. Appl Phys Lett 1995, 67: 1426–1428. 10.1063/1.114515View ArticleGoogle Scholar
  20. Jiang AQ, Lin YY, Tang TA: Coexisting depinning effect of domain walls during the fatigue in ferroelectric thin films. Appl Phys Lett 2006, 89: 032906–032906. 10.1063/1.2227626View ArticleGoogle Scholar
  21. Jiang AQ, Lin YY, Tang TA: Charge injection and polarization fatigue in ferroelectric thin films. J Appl Phys 2007, 102: 074109–074117. 10.1063/1.2786886View ArticleGoogle Scholar


© Shen et al; licensee Springer. 2011

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.