Active solid-state devices (e.g. – transistors – solid-state diode – Field effect device – Having insulated electrode
Reexamination Certificate
2000-03-07
2001-11-27
Flynn, Nathan (Department: 2826)
Active solid-state devices (e.g., transistors, solid-state diode
Field effect device
Having insulated electrode
C257S296000, C257S298000, C438S003000, C438S238000, C438S239000, C438S386000, C438S399000
Reexamination Certificate
active
06323512
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a nonvolatile ferroelectric capacitor for nonvolatile semiconductor memory. More specifically, the present invention relates to a nonvolatile ferroelectric capacitor comprising layered perovskite ferroelectric thin film made of Bi
4−x
A
x
Ti
3
O
12
for FeRAM (ferroelectric random access memory) application.
2. Description of the Related Art
DRAM (Dynamic Random Access Memory) typically used in computer main memory systems can provide a low cost RAM solution with high integration density and, particularly, has substantially no limitation on the number of write operations that can be performed. But DRAM is susceptible to damage from radiation, needs periodic refreshing to retain stored data, and is volatile, that is, it loses data in the absence of power.
In contrast, conventional nonvolatile memories such as EPROM, EEPROM and Flash Memory can maintain stored data even in the absence of power. However, these nonvolatile memories are relatively costly, have low integration density, require extremely high voltages for relatively long time periods to write and erase data, and, most undesirably, allow very limited cycles of write and erase operations compared to DRAM. Therefore, conventional nonvolatile memories are generally used in read-only or read-mostly applications.
Recently, a new type of nonvolatile memory, so called the ferroclectric RAM (FeRAM), is getting attention in the semiconductor industry. Since FeRAM stores digital data as two stable polarization states of the ferroelectric material, the polarization states being maintained when power is removed from FeRAM, it can maintain stored data even in the absence of power. In other words, FeRAM has nonvolatility. Moreover, as a change of polarization states occurs in substantially under 100 ns, read/ write operations of FeRAM can be performed as fast as those of DRAM. In addition, FeRAM is highly resistant to radiation damages and requires a low operation voltage. Therefore, FeRAM has been recognized as a next generation mainstream memory selection.
However, several challenges still remain in order to provide commercially practicable FeRAM. The ferroelectric thin film used in a FeRAM should maintain high remnant polarization, and be substantially free of fatigue (a reliability failure caused by the decrease of the magnitude of remnant polarization under repeated polarization switchings). In addition, the processing temperature of the ferroelectric material should be low enough to be compatible with the conventional semiconductor fabrication process.
For example, ferroelectric capacitors constructed with perovskite-family ferroelectric materials such as PZT (PbTiO
3
—PbZrO
3
) are well known in the art. However, when a ferroelectric capacitor is fabricated by depositing a PZT film on a conventional Pt electrode, the magnitude of the remnant polarization of the ferroelectric thin film decreases with the number of times that the direction of polarization is switched, which is so called fatigue. Therefore, the FeRAM with PZT film can provide only a limited number of read/write cycles, failing to overcome the problems of conventional nonvolatile memories such as flash memory.
It has been reported that the fatigue failure originates from movement of oxygen vacancies and their entrapment at the electrode/ferroelectric interface. Under an external electric field, the oxygen vacancies generated in the ferroelectric film during the processing move towards the electrode/ferroelectric interface and get entrapped at the interface, which results in the loss of polarization.
Two possible approaches have been suggested to overcome the fatigue problem. One of them is to reduce the tendency for entrapment of oxygen vacancies by employing a multilayer electrode structure having conductive oxide electrodes such as RuO
2
, as disclosed in U.S. Pat. No 5,491,102 issued to Desu et al. on Feb. 13, 1996.
Another approach is to use ferroelectric materials other than PZT without changing the conventional electrode structure. Such approach is disclosed in U.S. Pat. No. 5,519,234 issued to Paz de Araujo et al. entitled “Ferroelectric dielectric memory cell can switch at least Giga cycles and has low fatigue-has high dielectric constant and low leakage current”. U.S. Pat. No. 5,519,234 discloses a memory cell capacitor with extremely low fatigue comprising a layered superlattice material having formula A1
w1
+a1
A2
w2
a2
Aj
wj
+aj
S1
x1
+s1
S2
x2
+s2
Sk
xk
+sk
B1
y1
+b1
B2
y2
+b2
Bj
yj
bj
Q
a
−2
, wherein A1, A2, , Aj represent A-site elements in a perovskite-like structure, B1, B2, , Bj represent B-site elements in a perovskite-like structure, S1, S2, , Sk represent superlattice generator elements, and Q represents an anion. One or more perovskite ferroelectric layers which have a rigid crystal lattice and a non-ferroelectric layer which has a less rigid structure alternate with each other throughout the crystal of the layered superlattice material. According to the U.S. Pat. No. 5,519,234, the non-ferroelectric layers between the perovskite ferroelectric layers absorb the shock generated in the perovskite ferroelectric layers by repeated switching of polarization and allow the ferroelectric thin film to maintain its high polarizable state. SBT, an exemplary layered superlattice material, maintains relatively high remnant polarization and low fatigue after 10
12
switching cycles. It should be noted, however, that excellent ferroelectric properties of SBT in bulk had already been reported in various publications (see Solid State 3, 651(1961), G. A. Smolenski et al.; J. Am. Ceram. Soc. 45, 166(1962), E. C. Subbarao; J, Phys. Chem. Solids 23, 655(1962), E. C. Subbarao). The significance of U.S. Pat. No. 5,519,234 lies in that the layered superlattice materials such as SBT were found to exhibit extremely low fatigue even in the form of thin film and were used in fabricating FeRAM.
Meanwhile, Bi
4
Ti
3
O
12
(BTO) is another Bi-layered perovskite ferroclectric material which is known to show good ferroelectricity in bulk. However, BTO thin film has not been considered suitable for non-volatile ferroelectric memory, since the BTO thin film has a serious fatigue problem and Ti ions in the BTO thin film are known to diffuse into a silicon substrate to form conductive titanium silicide during heat treatment. The U.S. Pat. No. 5,519,234 solved these problems by sandwiching BTO thin film between buffer layers made of SrTiO
3
.
Even though SBT thin film on the metal electrode exhibits extremely low fatigue, it has two disadvantages. First, SBT has lower remnant polarization (2Pr≈20 mC/) than that of PZT (2Pr≈35 mC/), which makes it difficult for the change of the polarization states to be sensed by a sense amplifier. Second, the existence of the intermediate inetastable nonferroelectric fluorite phase (Appl. Phys. Lett. 73, 2518 (1998), S. J. Hyun et al.) requires SBT thin film to be annealed at high temperature ranging from 750 to 850 for extended periods in order to transform SBT material as deposited to have the layered perovskite phase exhibiting ferroelectricity. The high temperature annealing imposes serious constraints on the back-end process such as interconnect formation and contact metalization process which usually require a relatively low thermal budget.
Moreover, the buffer layers sandwiching the BTO film complicate the fabrication process and increase the memory size, which results in increases the operating voltage and power consumption.
SUMMARY OF THE INVENTION
The object of the present invention is to provide a nonvolatile ferroelectric capacitor comprising a thin film of layered perovskite ferroelectric material which is substantially free of fatigue on the conventional metal electrode and has a large value of remnant polarization and low processing temperature.
The present invention provides a nonvolatile ferroelectric capacitor employing layered perovskite ferroelectric material which is obtained by s
Bu Sang-Don
Kang Bo-Soo
Noh Tae-Won
Park Bae-ho
Flynn Nathan
Quinto Kevin
Senniger Powers Leavitt & Roedel
Tae-Won Noh
LandOfFree
Nonvolatile ferroelectric capacitor and nonvolatile... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Nonvolatile ferroelectric capacitor and nonvolatile..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Nonvolatile ferroelectric capacitor and nonvolatile... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2607204