Ferroelectric Random Access Memories (FeRAMs/FRAMs)
- Practical Electron Microscopy and Database -
- An Online Book -


This book (Practical Electron Microscopy and Database) is a reference for TEM and SEM students, operators, engineers, technicians, managers, and researchers.


One of the main disadvantages of Dynamic Random Access Memory (DRAM) and Static Random Access Memory (SRAM) is their volatility, that is, the stored data is lost upon power failure. Commercial non-volatile memories have been Flash memory, Complementary Metal Oxide Semiconductors (CMOS) using battery as backup, and Electrically Erasable Programmable Read Only Memory (EEPROM).

Based on the spontaneous polarization and its reversibility with the application of an external electric field in ferroelectric films, the ferroelectric random access memories (FRAMs), which are a type of non-volatile emerging memories, have been developed. In the FRAMs, the data is stored in a ferroelectric thin-film capacitor by localized polarization switching in the microscopic regions of ferroelectric thin films. The FRAMs are non-volatile because the polarization remains in the same state when the power is off. The upward and downward polarizations in FRAM's unit cells are referred to “0” and “1” state or vice versa. FRAMs also have other interesting features, e.g. a high-speed rewriting capacity. In practice, the macroscopic dipole moment, which induces the spontaneous polarization, in actual ferroelectric crystals is compensated by free surface charges, and thus there is no electric field outside the crystal. However, ferroelectrics insulators can physically be “switched” by applying an external electric field, resulting in an electrical current pulse between the two surfaces of the crystal.

FRAM technology has practically been developed mainly for producing low-density products in which well-known large-feature-size CMOS technology is applied. Most FRAM technologies are based on 1T1C (1 transistor and 1 capacitor) [1-3] and 2T2C (2 transistors and 2 capacitors) architectures, in combination with COB (capacitor over bitline) and CUB (capacitor under bit-line) cell structures. Figure 3534 shows the FRAM 1T1C or 2T2C architectures.

FRAM 1T1C and 2T2C cell designs

Figure 3534. FRAM 1T1C and 2T2C cell designs. BL: Bit Line; WL: Word Line; PL: Plate Line, BLB: Complementary Bit Line; FE: Ferroelectric capacitor.

There are still some challenges in applications of FRAM in industry:
        i) Polarization fatigue.
            In FRAM applications, the loss of remanent polarization is a key issue after a continuous electric field cycling of negative and positive biasing (cyclic read and write operations). [4,5]
        ii) Retention loss.
            In practice, a ferroelectric material cannot retain the polarization for a long time due to a reduced difference between the switching and non switching charges.
        iii) Imprint.
            This is the tendency of one polarization state to be more stable than the opposite one, which results in the loss of polarization. The imprint affects the ferroelectric nature by shifting the ferroelectric hysteresis loop, which makes it difficult to distinguish and address the write and read modes.
        iv) Vanishing of ferroelectricity in thin films.
             Ferroelectric behavior vanishes, for instance, if the thickness of PbTiO3 film is equal to or less than 1.2 nm [6], and 4 nm for PZT (PbZrO3)[7].
        v) Compatibility problem of ferroelectric materials with CMOS technology [4,5].

In practice, it is necessary that ferroelectric thin films for FRAMs have the characteristics below:
        i) A large amount of polarization reversal charge (Pr > 10 μC/cm2).
        ii) An excellent data retention (≥ 10 years).
        iii) A small coercive electric field and thus low-voltage operation.
        iv) A strong resistance to polarization fatigue.
        v) A high polarization reversal speed.
        vi) A small leakage current (≤ 10−6 A/cm2).
        vii) A small dielectric constant and thus a large S/N (signal to noise) ratio.






[1] D. J. Jung, N. S. Kang, S. Y. Lee, B. J. Koo, J. W. Park, Y. S. Chun, M. H. Lee, B. G. Lee, S. I. Lee, T. E. Shim and C. G. Hwang, VLSI Tech. Symp., p. 139 (1997).
[2] J. W. Lee, S. Y. Lee, D. J. Jung, B. J. Koo and K. N. Kim, J. Korean Phys. Soc. 33, 208 (1998).
[3] K. N. Kim, Technology Perspective for 1T/1C FRAM, Integrated Ferroelectrics 25, 149 (1999).
[4] H. Ishiwara, M. Okuyama, Y. Arimoto (Eds.): Ferroelectric Random Access Memories, Topics Appl. Phys. 93, 139–149, (2004).
[5] H. Kohlstedt, Y. Mustafa, A. Gerber, A. Petraru,M. Fitsilis, R. Meyer, U. Böttger and R Waser., Microelec. Engg, 80, 296–304, (2005).
[6] D. D. Fong, G. B. Stephenson, S. K. Streiffer, J. A. Eastman, O. Auciello, P. H. Fuoss, C. Thompson, Science, 304, 1650, (2004).
[7] T. Tybell, C. H. Ahn, J.-M. Triscone, Appl. Phys. Lett. 75, 856, (1999).





The book author (Yougui Liao) welcomes your comments, suggestions, and corrections, please click here for submission. If you let book author know once you have cited this book, the brief information of your publication will appear on the “Times Cited” page.