JP2948836B2 - Ferroelectric element - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to an electronic device utilizing characteristics of a ferroelectric. In particular, the present invention relates to a ferroelectric element which can be applied as a non-volatile and non-destructive memory or as a sensor, transducer, actuator, or the like which operates stably.

[Conventional technology]

As is well known, a ferroelectric can take two remanent polarization states having different polarities. Attempts have been made to realize a large-capacity, non-volatile, high-speed ferroelectric memory utilizing this property.

For example, a ferroelectric memory using a simple matrix electrode structure as shown in FIG. 10 has been proposed for a relatively long time. In the figure, 1... Are X-direction address lines.
These address lines 1 are connected to an X-direction address circuit 2. Reference numeral 3 denotes a Y-direction address line, which is connected to the Y-direction address circuit 4. X direction address line 1
.. And Y-direction address lines 3 are memory cells. A ferroelectric memory element is formed in each memory cell as shown. Such a ferroelectric memory can achieve a large capacity, non-volatility and high speed, but has the following problems.

Memory retention is unstable (poor durability).

At the time of reading, even cells other than the target cell are destroyed (crosstalk problem).

Since the read cell changes to the non-write state (destruction of the memory), rewrite is required after the read.

 Hereinafter, the above three problems will be described.

The problem of <durability> is a phenomenon characteristic of ferroelectric memories, and has not yet been sufficiently solved. One of the reasons is depolarization due to a demagnetizing electric field. That is, the charge distribution (space charge polarization electric field) generated inside the ferroelectric thin film by polarization generates a reverse electric field in the direction opposite to the polarization. Due to the effect of this anti-electric field, remanent polarization
Degrades Both the P _r and the coercive voltage V _c. FIG. 11 is a diagram for explaining this depolarization. As shown in the figure, the hysteresis curve 5 at the beginning of polarization changes to a curve 6 due to depolarization.

Another reason may be that the space charge polarization electric field is biased by the externally applied electric field, so that the symmetry with respect to the polarity of the applied electric field is deteriorated. As a result, as shown in FIG. 12, the residual polarization Pr when 0 V is applied substantially changes. In the figure, 7 is a hysteresis curve before the change, and 8 is a hysteresis curve after the change.

Still another reason is considered to be that internal structural defects increase due to internal distortion when the domain wall moves.

However, at present, no proof has been obtained as to whether the instability of the memory holding state is due to any of the above-mentioned reasons or to another reason. However, an anti-electric field should be reduced, an ideal crystal in which internal structural defects are unlikely to be generated during the movement of the domain wall, and a hard strong enough not to be depolarized by a certain threshold internal electric field (space charge polarization electric field). It can be easily inferred that if the use of the dielectric material is performed at the same time, an improvement path is opened. However, until now, it has been considered difficult to realize this concrete means.

Regarding the problem of <Crosstalk>, the present applicant has proposed a structure capable of avoiding the problem of crosstalk by using a switch element such as diac (Japanese Patent Application No. 63-170471). With this structure, a pseudo-square hysteresis curve can be obtained, and reading can be performed with a voltage of about 1/2 Vc without destroying cells other than the intended cells. However, there is a new structural problem that the structure becomes extremely complicated.

Another solution to the crosstalk problem in
This is to take the SRAM structure as shown in FIG. The memory cell in this structure includes a flip-flop circuit 9, two ferroelectric elements 10 and 11, and two transistors 12 and 13. The ferroelectric elements 10 and 11 are transistors 12,
Since they are connected to the X and Y address lines via 13, no problem of crosstalk occurs and it is easy to operate stably. However, this method also has the disadvantage that it is difficult to increase the memory capacity and that the manufacturing process is complicated.

<Memory Destruction by Reading> In conventional reading, a method of completely reversing the polarity of the remanent polarization and reading the recording state with the current value flowing at this time is used. In this read method, the read memory cell is inevitably a destructive memory, and the problem of inevitably being avoided. It is to be noted that a method of performing rewriting when the current value is large is generally used, but has a drawback that the reading time is lengthened correspondingly.

As described above, the problems of the conventional ferroelectric memory have been described, but the problem is not limited to the memory, and is a common problem in devices such as a piezoelectric sensor, a transducer, and a pyroelectric sensor using a ferroelectric. . That is, with respect to devices other than these memories, the problem is a factor leading to a change in characteristics over time, and thus improvement is desired.

[Problems to be solved by the invention]

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and an object thereof is to provide a ferroelectric element capable of simultaneously solving the problems of the conventional example.

[Means and actions for solving the problem]

A ferroelectric element according to the present invention is a ferroelectric element in which a lower electrode, a ferroelectric thin film, and an upper electrode are sequentially stacked on a substrate, wherein a polarization axis of the ferroelectric thin film is in a direction perpendicular to the two electrode surfaces. It is characterized by being inclined with respect to it.

 Hereinafter, details of the present invention will be described.

In order to solve the above-mentioned problems, the present inventor has made the following study on the polarization phenomenon of a ferroelectric substance, and has arrived at the above-mentioned invention based on this study.

As shown in the hysteresis curve of FIG.
A ferroelectric has a positive and negative remanent polarization value (± Pr). These two remanent polarization values correspond to two stable polarization states (crystal displacement states) T ₁ and T ₂ as shown in the potential diagram of FIG. 1 (B). The transition between the two polarization states T ₁ and T ₂ is induced by an external electric field.

The amount of polarization P is the volume integral of the cosine value in the direction of the electric field of the permanent dipole existing in the unit cell. This value depends on whether the crystal is a single crystal or a polycrystal, and is determined by the symmetric orientation direction of the crystal, the polarization axis direction, and the like. How to change the amount of polarization also depends on these situations.

Therefore, a case where an electric field is applied to one permanent dipole will be considered. FIG. 2A shows a case where the direction of the applied electric field 22 matches the direction of the polarization (permanent dipole) 21. As shown, the polarization is in the + Z direction and the applied electric field 22
Is the -Z direction. If the applied electric field 22 is in the opposite direction to the direction of the dipole 21 in this way, the dipole is inverted by 180 ° in the opposite direction and aligned in the electric field direction (21 ′).

On the other hand, FIG. 2B shows a case where the direction of the polarization 21 and the direction of the applied electric field 22 do not match, and have a gradient θ. In this case, the dipole looks for another stable direction to approach θ = 0, and rotates to that direction (2
1 '). This figure shows the case of rotating 90 °,
It is easily presumed that rotations other than 90 ° may occur depending on the symmetry and orientation of the crystal. Thus, the applied electric field
When there is no polarization axis in the direction of 22, the direction of the polarization 21 'after the transition is not aligned with the electric field direction.

As mentioned above, the transition between the stable states of the dipole is
It can be seen that there are two ways, either by inversion or by rotation. Comparing these two types of transition, in the case of inversion, since the symmetry is maintained before and after the transition, distortion generated around the transition is small. On the other hand, in the case of rotation, since the symmetry is broken by the transition, there is a large difference that a large distortion is induced in the periphery. Also, it can be said that polarization due to inversion occurs abruptly, whereas polarization due to rotation occurs relatively slowly. From another viewpoint, when applying an electric field having the same amplitude and the same pulse width, it leads to the idea that the rotational transition has a greater resilience than the inversion transition. The present inventors have focused on this point.

That is, if such a transition between the polarization state takes place only by the rotation causes a slight polarization rotates by applying an electric field having a constant amplitude E _A and the pulse width T _P, and removal of the electric field pulse This should return you to the original stable point again. Then, it was considered that only the polarization rotation could be used if the polarization axis did not coincide with the direction of the electric field and was oriented with a certain angle with respect to the direction of the electric field. More specifically, a ferroelectric layer oriented between the upper electrode and the lower electrode such that the polarization axis has a certain angle with respect to the direction perpendicular to the electrode surface is provided. Then, if a voltage _Vap of a certain value or more is applied between both electrodes, + Pr or -P
r can be set to any of the polarization states. Further, by applying a considerably smaller 1 / 2Vc voltage of about from V _ap, can be read by the displacement current is accompanied polarization rotation flows, then returns to its original state when released voltage. That is, non-destructive reading becomes possible. As a result, the problem of crosstalk is solved and the durability is improved, so that all the conventional problems described above are solved.

As described above, two means are available for orienting the polarization axis so that the polarization axis does not coincide with the direction of the electric field and has a certain angle with respect to the direction of the electric field.

One method is that the polarization axis is the crystal axis (100), (010),
In the crystal that matches (001), the crystal axis (100),
This is a method of (110) or (111) crystal orientation, for example, so that (010) and (001) have a certain angle with respect to the direction perpendicular to the electrode surface. Another method is to orient a crystal whose polarization axis deviates from the crystal axis along the crystal axis.

In any of these methods, there are only two polarization stable states corresponding to + Pr and -Pr, and if there is no intermediate stable state, it is easy to return to the original state after releasing the read voltage. An intermediate stable state has a large number of polarization axes,
It is easily caused when the crystal orientation is random (polycrystalline). Therefore, it is desirable to use a single crystal having a small polarization axis and good orientation in at least a certain direction. Such a condition can be satisfied by having an ABO ₃ type perovskite structure (Pb _{1 + α-x} A _x ) (Zr _1-Yz Ti
In the case of using the _Y B _Z) O ₃ + _βMeO based compounds with tetragonal, (B
This corresponds to the case where the compound represented by i ₂ O ₂ ) ²⁺ (A _m-1 B _m O _{3m + 1} ) ^2- is used with its c-axis oriented.

FIG. 3A shows a unit cell having an ABO ₃ perovskite structure. The X-axis, Y-axis, and Z-axis in FIG.
0), (010), and (001) directions. Curie point (T
c) At the above temperatures, the B ⁴⁺ ion is located at the center of gravity of the six O ²⁻ ions (indicated by a cross in the figure). However, at temperatures below Tc, the B ⁴⁺ ions are located at any of the six stable points slightly offset from the center of gravity. Fig. 3 (A)
^{Shows a case} where the B ⁴⁺ ion is located at a stable point in the + Z direction. The B ⁴⁺ ion forms a permanent dipole with the negative virtual ion at the X position. I have. As shown in FIG. 3B, the electric field in the −Z direction is in this state.
When 22 is added, the B ⁴⁺ ion transitions to another stable point in the −Z direction, and the dipole 23 is inverted by 180 °. FIG. 3 (C)
When an electric field 22 in the + Y direction is applied as shown in (1), the B ⁴⁺ ions move to another stable point in the + Y direction, and the dipole 23 rotates 90 °.

On the other hand, when an electric field is applied in the (111) direction in FIG. 3 (A), the permanent dipole rotates by α ° <90 °. This situation will be described with reference to FIGS. In FIG. 1A, reference numeral 24 denotes an upper electrode, and 25 denotes a lower electrode.
As shown in the figure, a ferroelectric substance having a perovskite structure oriented so that the (111) axis is perpendicular to the electrode surface is provided between the two electrodes. When an electric field in the − (111) direction is applied by both electrodes 24 and 25, as shown in FIG.
The dipole 23 facing in the direction rotates in the + Y direction. This rotation is 90 ° in the dipole plane of rotation. However, the rotation angle α ° projected on (111) is smaller than 90 °. As described above, the relationship between the direction of the electric field and the direction of the polarization axis is the third.
It is understood that by setting as shown in the figure, only dipole rotation of α ° <90 ° can be caused.

Although dipoles have been described above, polarization is a collective expression of these dipoles, and there is no essential difference in concept. On the other hand, when an electric field is applied in the (110) direction, there are three stable directions of the dipole projected on the (110) axis. For this reason, although it is better than applying an electric field in the (100) direction, it cannot be denied that the resilience is smaller than when an electric field is applied in the (111) direction.

Next, referring to FIGS. 5A, 6B and 6, (Bi
_The case where the compound represented by ₂ O ₂ ) ²⁺ (A _m-1 B _m O _{3m + 1} ) ^2- is used with its c-axis oriented is used. This compound has a crystal structure as shown in FIGS. 5 (A) and (B). The polarization axis is in a direction ± several degrees from the a axis in a plane including the a axis and the c axis. Therefore, the sixth
As shown in the figure, polarization rotation can be promoted by applying an electric field to the c-axis oriented crystal in the c-axis direction.

The above is the concept on which the present invention is based. In summary, the present invention utilizes rotation instead of polarization reversal for both writing and reading. In addition, when the projection is performed in the direction in which the electric field is applied, the smaller the number of stable directions due to rotation, the better. More preferably, the number of stable directions is only two. To achieve this, the tetragonal perovskite crystal is oriented (111),
An electric field is applied in this direction. Alternatively, a method is employed in which a crystal whose polarization direction is shifted from the crystal axis is oriented in the crystal axis direction. In order to detect these polarization states, the magnitude of the current when a voltage smaller than the coercive electric field is applied may be checked. If the detected electric field is much smaller than the coercive electric field, it is similar to the other case where the ferroelectric is used for a small signal. Therefore, it can be said that the present invention is also useful for improving the stability of sensors, transducers, communication components, and the like used for small signals.
In the present invention, the non-destructive readout is possible, the stability is good, and any of the features of avoiding Christoke are caused by the strong reversion of the polarization rotation as described above. Embodiment 1 A first embodiment according to the present invention will be described with reference to FIG.

For example, quartz glass, NA-40 (non-alkali high heat-resistant glass: manufactured by HOYA), magnesium oxide (MgO), sapphire, strontium titanate (SrTiO ₃ ), silicon (S
i), a substrate 31 made of gallium arsenide (GaAs), gallium aluminum arsenide (GaAlAs) or the like is used. When the substrate 31 is conductive (Si, GaAs, GaAlAs), an insulating layer 32 made of, for example, silicon oxide (SiO ₂ ) or silicon nitride (Si ₃ N ₄ ) is coated by means such as sputtering or CVD. After that, the lower electrode 33 is attached thereon by means such as sputtering, CVD, vacuum deposition, ion plating and the like. On the other hand, when the substrate 31 has no conductivity (quartz glass, NA-40, MgO, sapphire, SrT
On iO ₃ ), a lower electrode 65 is directly attached on the substrate.

In this embodiment, [100] plane Si is used for the substrate 31, and the surface is oxidized to form a SiO ₂ film 32, and platinum (Pt) is sputtered while heating to 300 ° C. to form a (111) orientation. The formed Pt lower electrode 33 was formed. A metal alkoxide solution was spin-coated thereon. This solution is treated with (Pb _{1 + α-x} A _x ) (Zr _1-Yz T
i _Y B _z ) O ₃ + βMeO
It has a composition containing these elements. However, in the above chemical formula, α = 0 to 0.2, A: any of Ca, Sr, Ba, x = 0 to 0.3, y ≧ 0.58, B: any of Hf, Sn, z = 0 to 0.3, Me : La, Y, Sm, Dy, Ce, Bi, Sb, Nb, Ta, W, Mo, Cr, Co, Ni, Fe, Cu, T
Any one of h, U, Sc, Si, Ge, and Al, or a combination of a plurality of them, β = 0 to 0.05.

After spin coating the above metal alkoxide solution,
Dried at 250 ° C. Further, by irradiating infrared rays at intervals of 10 seconds while heating at 250 ° C., heat treatment was performed at 450 ° C. to 800 ° C. for 10 minutes to 120 minutes in air or oxygen gas. This was defined as one cycle, and the cycle was repeated 5 to 10 times to obtain a ferroelectric thin film 34 having a thickness of about 1 μm. The obtained ferroelectric thin film 34 had a perovskite structure and, as shown in FIG. 4A, had a crystal orientation in which the (111) axis was perpendicular to the lower electrode surface.

In the above chemical formula, α = 0 may be used.
When the composition is set to be 0 to 0.2, liquid phase sintering easily occurs during heat treatment, and a dense film without pinholes can be obtained at a relatively low heat treatment temperature. When α> 0.2, free lead oxide is precipitated, and good characteristics cannot be obtained.

Ca, Sr, and Ba used as A increase the dielectric constant so that evaporation of Pb does not have to be a problem. Furthermore, even if the Zr / Ti ratio is relatively large, it can be made tetragonal by substituting with A, and the characteristics (particularly the dielectric constant, remanent polarization value, and temperature characteristics thereof) can be changed in a wide range. It is possible.

When x> 0.3 or more, the Curie point Tc is 100 ° C. or less.
For this reason, the temperature characteristic is rather deteriorated, and depolarization is liable to occur.

The condition of y ≧ 0.58 differs depending on the balance with x. That is, when x = 0, y ≧ 0.5, and when x = 0.15 to 0.3, y ≧ 0.58.

B (Sn, Hf) is intended to obtain a larger variation in characteristics by replacing the Zr and Ti sites as in A.

The value of z is determined in consideration of x and y. In particular, x + z <
0.3 so that Tc does not become too low.

Me is an additive, and La, Y, Sm, Dy, Ce, Bi, Sb, Nb, Ta, W, Mo, and Th are used for the purpose of softening (Pr increase, Vc decrease). If the purpose is to harden (Pr decrease, Vc increase), Cr, Ca, Ni, Co, Fe, U, Cu, Sc are used. In order to control the crystal grain size in the plane direction, and to accelerate liquid phase sintering with Pb to obtain a dense film, Si, A
l, Ge may be added. If the softening additive and the hardening additive are appropriately mixed, it becomes easier to control the domain generation process during polarization rotation.

When β ≧ 0.05, favorable characteristics cannot be obtained due to generation of a separated phase.

In this example, a metal alkoxide was used as a spin-coating starting solution. However, a ferroelectric thin film 34 is similarly formed by spin coating using an acetylacetone metal salt or a naphthenic acid or octylate metal soap as a starting solution. be able to.

As described above, the ferroelectric thin film 34 thus obtained is spin-coated, dried, by devising all steps of heat treatment,
Like the Pt electrode, it can be oriented in the (111) direction. However, it is polycrystalline in the plane direction.

On this oriented ferroelectric thin film 34, any one of Ti, Mo, Cr, W is vacuum-deposited to improve adhesion, and after undercoating by sputtering, Au, Pt, Ag, Either Ni or Al is deposited by vacuum evaporation sputtering or the like to form the upper electrode 35. Here, the base coat may be omitted if the adhesion of the upper electrode 35 does not matter.

In the ferroelectric cell region 36 obtained as described above, by applying a voltage between the lower electrode 33 and the upper electrode 35,
An electric field can be applied at a fixed angle with respect to the polarization axes 37 and 38. For example, when a voltage higher than the coercive voltage from the upper electrode 35 to the lower electrode 33 is applied, remanent polarization in the direction of 37 (broken line) is obtained. Also, the lower electrode 33 is connected to the upper electrode 35.
When a voltage equal to or higher than the coercive voltage is applied, remanent polarization oriented in the direction of 70 (solid line) is obtained. By associating these two states with the “1” and “0” states, respectively, the memory can be written. On the other hand, by applying a voltage considerably smaller than the coercive voltage Vc to rotate the remanent polarization and detecting the magnitude of the displacement current generated by the rotation,
It is possible to detect whether the state is "1" or "0", that is, read out.

Embodiment 2 A second embodiment of the present invention will be described with reference to FIG.

For example, quartz glass, NA-40 (non-alkali high heat-resistant glass: manufactured by HOYA), magnesium oxide (MgO), sapphire, strontium titanate (SrTiO ₃ ), silicon (S
i), a substrate 41 made of gallium arsenide (GaAs), gallium aluminum arsenide (GaAlAs) or the like is used. When the substrate 41 is conductive (Si, GaAs, GaAlAs), an insulating layer 42 made of, for example, silicon oxide (SiO ₂ ) or silicon nitride (Si ₃ N ₄ ) is coated by means such as sputtering or CVD. After that, the lower electrode 43 is adhered thereon by means such as sputtering, CVD, vacuum deposition, ion plating and the like. On the other hand, when the substrate 41 has no conductivity (quartz glass, NA-40, MgO, sapphire, SrT
On iO ₃ ), the lower electrode 43 is directly attached on the substrate.

In this embodiment, a [100] plane Si is used as a substrate, and the surface is oxidized to form a SiO ₂ film 42, and furthermore, platinum (Pt) is sputtered while being heated to 700 ° C. to be (100) oriented. A Pt lower electrode 43 was formed. A metal alkoxide solution was spin-coated thereon. This solution has a composition represented by the following chemical formula by heat treatment after coating.

(Bi ₂ O ₂ ) ²⁺ (A _m-1 B _m O _{3m + 1} ) ^2- However, in the above chemical formula, A: Any of Bi, Pb, Ba, Sr, Ca, Na, K, B: Any of Ti, Nb, Ta, W, Mo, Fe, Co, Cr, m: A, 1, 2, 3, 4, 5 according to the valence of B, after spin coating the above metal alkoxide solution,
Dried at 250 ° C. Further, heat treatment was performed at 450 ° C. to 800 ° C. for 10 to 120 minutes in air or oxygen gas. This process was defined as one cycle, and the treatment cycle was repeated 5 to 10 times to obtain a ferroelectric thin film having a thickness of about 1 μm.

Although a metal alkoxide was used as a starting solution for spin coating in this example, the ferroelectric thin film 44 was similarly formed by spin coating using a metal salt of acetylacetone or a metal soap of naphthenic acid or octylic acid. Can be membrane.

The ferroelectric thin film 44 obtained in this manner is spin-coated, dried,
The c-axis can be oriented in the same direction as the Pr electrode. However, it is polycrystalline in the plane direction.

On the oriented ferroelectric thin film 44, any one of Ti, Mo, Cr, W is vacuum-deposited to improve adhesion, and after undercoating by sputtering or the like, Au, Pt, Ag, The upper electrode 45 is formed by depositing either Ni or Al by vacuum evaporation or sputtering. If the adhesion of the upper electrode 45 does not matter, the base coat may be omitted.

The ferroelectric cell 46 obtained as described above can apply an electric field at a fixed angle to the polarization axes 47 and 48 by the voltage applied between the lower electrode 43 and the upper electrode 45. When a voltage higher than the coercive voltage from the upper electrode 45 to the lower electrode 43 is applied, remanent polarization in the direction of 47 (broken line) is obtained. When a voltage equal to or higher than the coercive voltage from the lower electrode 43 to the upper electrode 45 is applied, remanent polarization in the direction of 70 (solid line) is obtained. Each of these two states is “1”,
By associating with the state of “0”, writing to the memory can be performed. On the other hand, by applying a voltage considerably smaller than the coercive voltage Vc to rotate the remanent polarization and detecting the magnitude of the displacement current generated by the rotation, it is possible to detect whether the state is "1" or "0", that is, read out. Can be performed.

Embodiment 3 A third embodiment according to the present invention will be described with reference to FIG.

In this embodiment, the lower striped electrodes 53 ₁ , 53 ₂ are arranged such that the single cells of the embodiment 1 or 2 are two-dimensionally arranged.
, And upper stripe-shaped electrodes 55 ₁ , 55 _2, ... 51 is a substrate, 52 is an insulating film, and 54 is a ferroelectric layer. In order to use the polarization rotation of the ferroelectric layer 54 for writing and reading, the ferroelectric thin film layer was used in Example 1.
Or they are oriented as described in the second embodiment.

The control circuit for the ferroelectric memory shown in FIG. 9 is the same as that described in the invention (Japanese Patent Application No. 63-321639) filed by the present applicant (FIG. 23 to FIG. 23). (Figure 28).

Basically, an access circuit is used to select one of the lower striped electrode groups 53 ₁ , 53 ₂ ... And one of the upper striped electrode groups 55 ₁ , 55 ₂ . ,
One cell is selected from the ferroelectric two-dimensional array elements. For this selected cell, a voltage V having a polarity and magnitude such that the cell is in either a positive or negative remanent polarization state.
_{Apply ap} . That is, by applying + V _ap / 2, -V _ap / 2, or -V _ap / 2, + V _ap / 2 to each of a selected set of stripe-shaped electrodes, any of two values can be obtained. Set (write) in such a way. In addition, in order to detect the state, the access circuit is used as in the case of writing, a target cell is selected, and a voltage lower than the cell coercive voltage Vc is applied. That is, for each of a pair of stripe-shaped electrodes that are selected, + V _c / 2 or less, applying a -V _c / 2 or less voltage.
At this time, reading is performed by monitoring the magnitude of the current flowing between both stripe electrodes.

〔The invention's effect〕

As described in detail above, in the ferroelectric element of the present invention, the polarization axis of the ferroelectric thin film is inclined with respect to the direction perpendicular to the upper and lower electrode surfaces, and the electric field is applied at a fixed angle with respect to the polarization axis. And writing and reading are performed by rotation of polarization. As a result, the storage state is unstable (poor in durability). The problem of crosstalk becomes a destructive memory.

Problems that have hindered the practical use of ferroelectric memories, such as non-destructive, non-volatile, large capacity,
In addition, a high-speed switching ferroelectric memory can be realized.

In addition, the improvement described above is not limited to a memory element, but also leads to an improvement in durability (aging) of sensors, transducers, and communication components using ordinary small signals.

[Brief description of the drawings]

1 to 6 are views for explaining the technical concept of the present invention, FIG. 7 is a sectional view showing a ferroelectric memory according to an embodiment of the present invention, and FIG. FIG. 9 is a sectional view showing a ferroelectric memory according to an embodiment, FIG. 9 is a perspective view showing a partial cross section of a ferroelectric two-dimensional array memory according to still another embodiment of the present invention, and FIG. FIG. 11 is a circuit diagram showing a dielectric memory, FIGS. 11 and 12 are diagrams for explaining the problem, and FIG. 13 is a circuit diagram showing another conventional ferroelectric memory. 21,21 '… Polarization, 22… External electric field, 23… Permanent dipole, 24…
Upper electrode, 25 ... Lower electrode, 31, 41, 51 ... Substrate, 32, 42, 52 ...
Insulating film, 33, 43, 53 ... lower electrode, 34, 44, 54 ... ferroelectric thin film, 35, 45, 55 ... upper electrode, 36, 46 ... ferroelectric memory cell, 37, 38, 47, 48 ... Polarization axis

──────────────────────────────────────────────────の Continued on the front page (72) Inventor Yoshiki Ota 2-43-2 Hatagaya, Shibuya-ku, Tokyo Inside O-Limpus Optical Industry Co., Ltd. (72) Inventor Jun Funasaki 2-34-2 Hatagaya, Shibuya-ku, Tokyo No. Olympus Optical Co., Ltd. (72) Inventor Atsushi Yusa 2-43-2 Hatagaya, Shibuya-ku, Tokyo In Olympus Optical Co., Ltd. (56) References JP-A-48-88500 (JP, A) ( 58) Field surveyed (Int.Cl. ⁶ , DB name) H01L 27/10 451 H01L 29/788 H01L 29/792 H01L 21/8247 H01L 27/115 H01L 41/187

Claims (5)

(57) [Claims] 1. A ferroelectric device comprising a substrate, on which a lower electrode, a ferroelectric thin film and an upper electrode are sequentially laminated, wherein the ferroelectric thin film is arranged such that its polarization axis is perpendicular to the two electrode surfaces. A ferroelectric element characterized in that the crystal is oriented so as to be inclined. 2. The ferroelectric device according to claim 1, wherein the lower electrode has the same crystal orientation direction as the crystal orientation direction of the ferroelectric thin film. 3. The substrate is made of glass, single crystal of magnesium oxide (MgO), single crystal of strontium titanate (SrTi
O ₃ ), silicon single crystal (Si), gallium arsenide single crystal (Ga
As) selected from the group consisting of gallium aluminum arsenide single crystal (GaAlAs) and surface-oxidized silicon (SiO ₂ / Si), and the lower electrode is made of platinum (Pt) oriented in the (111) direction And the ferroelectric thin film is (Pb _{1 + α-x} A _x ) (Zr _{1 -Yz} Ti _Y
B _z ) O ₃ + βMeO α = 0-0.2, y ≧ 0.58, A: Any of Ca, Sr, Ba, x = 0-0.3, B: Any of Hf, Sn, z = 0-0.3, Me: La, Y, Sm, Dy, Ce, Bi, Sb, Nb, Ta, W, Mo, Cr, Co, Ni, Fe, U, S
Any one of c, Th, Cu, Si, Ge, and Al or a combination of a plurality thereof, has a chemical formula of β = 0 to 0.05, and has a tetragonal crystal structure (11
1) that it is oriented, the upper electrode is a single-layer structure made of any of Pt, Au, Al, Ni, or an underlayer made of any of Cr, Ti, Mo, W and Pt, Au,
2. The ferroelectric element according to claim 1, wherein the ferroelectric element has an upper two-layer structure made of any of Al and Ni. 4. The substrate is made of glass, single crystal of magnesium oxide (MgO), single crystal of strontium titanate (SrTi
O ₃ ), silicon single crystal (Si), gallium arsenide single crystal (Ga
As) being selected from the group consisting of gallium aluminum arsenide single crystal (GaAlAs) and surface oxidized silicon (SiO ₂ / Si); the lower electrode being made of (100) oriented Pt; The thin film is (Bi ₂ O ₂ ) ²⁺ (A _m-1 B _m O _{3m + 1} ) ^2- A = Bi, Pb, Ba, Sr, Ca, Na, K, Rare element B = Ti, Nb , Ta, W, Mo, Fe, Co, Cr m = 1,2,3,4,5 and has c-axis orientation, and the upper electrode is composed of Pt, Au, Al , Ni, a single layer structure, or Cr, Ti, Mo, W underlayer and Pt, Au,
2. The ferroelectric element according to claim 1, wherein the ferroelectric element has an upper two-layer structure made of any of Al and Ni. 5. The device according to claim 1, wherein both the lower electrode and the upper electrode are composed of a plurality of stripe-shaped electrodes, and have a structure in which the lower electrode and the upper electrode are arranged orthogonal to each other. Ferroelectric two-dimensional array element.