JP7572058B2 - NONLINEAR RESISTANCE ELEMENT, SWITCHING ELEMENT, AND METHOD FOR MANUFACTURING NONLINEAR RESISTANCE ELEMENT - Google Patents

Description

This disclosure relates to a nonlinear resistance element, a switching element, and a method for manufacturing a nonlinear resistance element.

In semiconductor integrated circuits, miniaturization (scaling law) has led to increased integration and lower power consumption of devices, improving their performance. In recent years, the rising cost of lithography processes and physical limitations of device dimensions have created a demand for improving device performance using approaches other than the scaling law used up to now.

In recent years, rewritable programmable logic devices called FPGAs (Field Programmable Gate Arrays) have been developed as an intermediate between gate arrays and standard cells. FPGAs allow customers to freely set the circuit configuration of their chips. FPGAs have resistance change elements inside a multi-layer wiring structure, allowing the electrical connections of the wiring to be changed. By using semiconductor devices equipped with such FPGAs, it is possible to improve the flexibility of the circuit.

Examples of resistance change elements include MRAM (Magneto-resistive Random Access Memory), PRAM (Phase Change Random Access Memory), ReRAM (Resistive Random Access Memory), and CBRAM (Conductive Bridging Random Access Memory, a RAM that forms a conductive path using ions in a solid electrolyte).

ReRAM utilizes the characteristic that the resistance value changes depending on the voltage and current applied from the outside. In a ReRAM cell, a structure having a resistance change film sandwiched between two electrodes is used. For example, an electric field is applied between the two electrodes to generate a filament inside the resistance change film made of metal oxide, or a conductive path is formed between the two electrodes, thereby turning the device into an ON state. Then, an electric field is applied in the opposite direction between the two electrodes to eliminate the filament, or to eliminate the conductive path formed between the two electrodes, thereby turning the device into an OFF state. By reversing the direction of the applied electric field, the resistance value between the two electrodes is greatly different, and switching between the ON state and the OFF state is performed. Data is stored by utilizing the fact that the current flowing through this resistance change element is different depending on the difference in resistance value between the ON state and the OFF state. When writing data, a voltage value, a current value, and a pulse width for causing a transition from the OFF state to the ON state or from the ON state to the OFF state are selected according to the data to be stored, and a filament for data storage is generated or eliminated, or a conductive path is formed or eliminated.

Non-Patent Document 1 discloses a non-volatile resistance change element as an example of a resistance change element that can be used in such a ReRAM. The non-volatile resistance change element disclosed in Non-Patent Document 1 has a solid electrolyte made of an ion conductor, and a first electrode and a second electrode provided in contact with each of two surfaces of the solid electrolyte. The first electrode is made of a first metal, and the second electrode is made of a second metal. The first metal and the second metal have different standard Gibbs energies of formation ΔG in the process of oxidizing the metal to generate metal ions.

The resistance change element of Non-Patent Document 1 is a non-volatile switching element that performs switching by reversibly changing the resistance value between electrodes sandwiching a resistance change film, utilizing the movement of metal ions in an ion conductor and the electrochemical reactions of "metal deposition due to reduction of metal ions" and "production of metal ions due to oxidation of metal." This non-volatile switching element is a non-volatile metal bridge type resistance change element that can transition between an on state and an off state by forming and dissolving a metal bridge structure.

In the resistance change element of Non-Patent Document 1, in the transition process from the off state to the on state (set process), when the second electrode, which is an inactive electrode, is grounded and a positive voltage is applied to the first electrode, which is an active electrode, the metal of the first electrode becomes metal ions at the interface between the first electrode and the solid electrolyte and dissolves in the solid electrolyte. Meanwhile, on the second electrode side, the metal ions in the solid electrolyte become metal and precipitate in the solid electrolyte using electrons supplied from the second electrode. A metal bridge structure is formed by the metal precipitated in the solid electrolyte, and finally a metal bridge is formed that connects the first electrode and the second electrode. The first electrode and the second electrode are electrically connected via the metal bridge, and the switch is turned on.

On the other hand, in the transition process from the on state to the off state (reset process), when the second electrode of the switch is in the on state and a negative voltage is applied to the first electrode while grounding the second electrode, the metal that constitutes the metal bridge becomes metal ions and dissolves into the solid electrolyte. As the dissolution of the metal progresses, part of the metal bridge structure that constitutes the metal bridge breaks. Finally, when the metal bridge connecting the first electrode and the second electrode is cut, the electrical connection is broken and the switch turns off.

As the dissolution of the metal progresses, the metal bridge structure that constitutes the conductive path becomes thinner, increasing the resistance between the first electrode and the second electrode, and at the interface between the first electrode and the solid electrolyte, the dissolved metal ions are reduced and precipitate as metal. As a result, the concentration of metal ions contained in the solid electrolyte decreases and the relative dielectric constant changes, causing changes in the interelectrode capacitance and other electrical properties before the electrical connection is completely severed, and ultimately the electrical connection is severed.

Furthermore, after the metal-bridged variable resistance element is transitioned (reset) from the on state to the off state, when the second electrode is again grounded and a positive voltage is applied to the first electrode, the transition process from the off state to the on state (set process) proceeds. In other words, the metal-bridged variable resistance element can reversibly perform the transition process from the off state to the on state (set process) and the transition process from the on state to the off state (reset process).

Such solid electrolyte layer type variable resistance elements are characterized by being smaller in size and having a smaller on-resistance than semiconductor switches such as MOSFETs (Metal Oxide Semiconductor Field Effect Transistors). Therefore, solid electrolyte layer type variable resistance elements are promising for application as switching elements in programmable logic devices.

In addition, the conductive state (on or off) of the resistance change element is maintained even when the applied voltage is turned off. Therefore, the resistance change element can also be used as a non-volatile memory element. For example, a memory element having a resistance change element can be configured such that a memory cell including one selection element such as a transistor and one resistance change element is used as a basic unit, and multiple memory cells are arranged vertically and horizontally. By arranging the memory cells in this manner, it is possible to select any memory cell from multiple memory cells using word lines and bit lines. Then, a non-volatile memory can be realized that can sense the conductive state of the resistance change element of the selected memory cell and read whether information "1" or "0" is stored based on the on or off state of the resistance change element.

When the above-mentioned two-terminal nonvolatile resistance change element is mounted on a semiconductor device and this nonvolatile resistance change element is programmed, a configuration is used in which one selection transistor (access transistor) is provided for each nonvolatile resistance change element. This configuration is generally called 1T1R. The problem with 1T1R is that the area occupied by the selection transistor is large, so the overall area is effectively large. Therefore, it has been considered to reduce the mounting area of the circuit by replacing the selection transistor with a two-terminal selection element (selector) such as a nonlinear resistance element. This configuration is generally called 1S1R. For example, Patent Document 1 describes a technology for forming a two-terminal selection element (selector) on the top of a resistance change element to form a 1S1R. Patent Document 2 also describes an example of forming a 1S1R by using an Ovonic Memory Switch (OMS) as a resistance change element and an Ovonic Threshold Switch (OTS) as a two-terminal selection element (selector). A chalcogenide material is used for the Ovonic Memory Switch and Ovonic Threshold Switch described in Patent Document 2.

Japanese Patent No. 5380612 Japanese Patent Publication No. 2006-086526

M. Tada, K. Okamoto, T. Sakamoto, M. Miyamura, N. Banno, and H. Hada, "Polymer Solid-Electrolyte Switch Embedded on CMOS for Nonvolatile Crossbar Switch", IEEE TRANSACTIONS ON ELECTRON DE VICES, Vol. 58, No. 12, pp. 4398-4406, (2011). K. Tanaka, "Structual phase transitions in chalcogenide glasses", PHYSICAL REVIEW B, VOLUME 39, NUMBER 2, pp. 1270-1279, (1988).

As described in Patent Document 2, a nonlinear resistance element using a chalcogenide material is known as a two-terminal selection element (selector) that is an Ovonic threshold switch. An example of a nonlinear resistance element is an element with an MSM (Metal-Semiconductor-Metal) structure that includes an amorphous chalcogenide layer. In such a nonlinear resistance element, the electrical characteristics may deteriorate due to diffusion or desorption of the elements that make up the amorphous chalcogenide layer. However, despite the fact that various technologies have been investigated in the past, including the prior art described above, there is currently a lack of technology to obtain a nonlinear resistance element with excellent electrical characteristics.

The present disclosure has been made in consideration of the above circumstances, and an object of one embodiment of the present disclosure is to provide a nonlinear resistance element having excellent electrical characteristics.
Another problem to be solved by another embodiment of the present disclosure is to provide a switching element including the above nonlinear resistance element.
Another problem to be solved by another embodiment of the present disclosure is to provide a method for manufacturing the above nonlinear resistance element.

The present disclosure includes the following aspects.
<1> A first electrode,
a first nonlinear resistance layer made of an amorphous chalcogenide thin film;
The middle class,
a second nonlinear resistance layer made of an amorphous chalcogenide thin film;
a second electrode;
a first electrode, a first nonlinear resistance layer, an intermediate layer, a second nonlinear resistance layer, and a second electrode laminated in this order;
A nonlinear resistive element.
<2> The nonlinear resistance element according to <1>, wherein the amorphous chalcogenide thin film contains a first group element consisting of at least one element of Ge and Si, and a second group element consisting of at least one element selected from the group consisting of Se, Te, and S.
<3> The composition of the second group elements constituting the amorphous chalcogenide thin film of the first nonlinear resistance layer changes from the interface between the first nonlinear resistance layer and the first electrode toward the interface between the first nonlinear resistance layer and the intermediate layer,
the composition of the second group elements constituting the amorphous chalcogenide thin film of the second nonlinear resistance layer changes from the interface between the second nonlinear resistance layer and the second electrode toward the interface between the second nonlinear resistance layer and the intermediate layer;
The nonlinear resistance element according to <2>.
<4> The nonlinear resistance element according to any one of <1> to <3>, wherein the amorphous chalcogenide thin film contains a third group element consisting of at least one element selected from the group consisting of As, Sb, and N.
<5> The composition of the third group elements constituting the amorphous chalcogenide thin film of the first nonlinear resistance layer changes from the interface between the first nonlinear resistance layer and the first electrode toward the interface between the first nonlinear resistance layer and the intermediate layer,
the composition of the third group elements constituting the amorphous chalcogenide thin film of the second nonlinear resistance layer changes from the interface between the second nonlinear resistance layer and the second electrode toward the interface between the second nonlinear resistance layer and the intermediate layer;
The nonlinear resistance element according to <4>.
<6> The nonlinear resistance element according to any one of <1> to <5>, wherein the intermediate layer contains at least one element selected from the group consisting of Ti, Zr, Ta, Nb, Mo, W, Hf, V, Si, and C.
<7> The nonlinear resistance element according to any one of <1> to <6>, wherein the intermediate layer has a thickness of 2 nm or less.
<8> The nonlinear resistance element according to any one of <1> to <7>, which changes to a low resistance state by applying a voltage having an absolute value equal to or greater than a first threshold voltage between the first electrode and the second electrode, and changes to a high resistance state by applying a voltage having an absolute value equal to or less than a second threshold voltage that is less than the first threshold voltage.
<9> A switching element provided in a signal path of a logic circuit,
Two two-terminal nonlinear resistance elements;
Two two-terminal nonvolatile resistance change elements;
one terminal of each of the two two-terminal nonvolatile resistance change elements is connected to each other and to one electrode of each of the two two-terminal nonlinear resistance elements;
The two two-terminal nonlinear resistance elements are the nonlinear resistance elements according to any one of <1> to <8>.
Switching element.
<10> A method for manufacturing a nonlinear resistance element according to any one of <1> to <8>,
A first electrode;
A first nonlinear resistance layer;
The middle class,
a second nonlinear resistance layer made of an amorphous chalcogenide thin film;
a first electrode and a second electrode are laminated in this order in the same vacuum.
A method for manufacturing a nonlinear resistance element.

According to one embodiment of the present disclosure, a nonlinear resistance element having excellent electrical characteristics is provided.
According to another embodiment of the present disclosure, there is provided a switching element including the above nonlinear resistance element.
According to another embodiment of the present disclosure, there is provided a method for manufacturing the above nonlinear resistance element.

FIG. 1 is a cross-sectional view showing an example of the configuration of a nonlinear resistance element according to the first embodiment. FIG. 2 is a diagram for explaining an example of an experiment for investigating and analyzing the functional effect of the intermediate layer. FIG. 3 is a diagram showing an example of the current-voltage characteristic (IV characteristic) of the nonlinear resistance element of the first embodiment. FIG. 4 is a cross-sectional view showing an example of the configuration of a nonlinear resistance element of a comparative example. FIG. 5 is a diagram showing an example of the IV characteristic of a nonlinear resistance element of a comparative example. FIG. 6 is a cross-sectional view showing a configuration example of a modified nonlinear resistance element. FIG. 7 is a cross-sectional view showing an example of a nonlinear resistance element according to the second embodiment fabricated in an integrated circuit. FIG. 8 is an equivalent circuit diagram illustrating an example of the configuration of the crossbar switch according to the third embodiment. FIG. 9 is a cross-sectional view showing an example of an element structure of a semiconductor device according to the fourth embodiment.

The nonlinear resistance element, switching element, and method for manufacturing the nonlinear resistance element according to the present disclosure are described in detail below.

The drawings referred to in the following description are illustrative and schematic, and the present disclosure is not limited to these drawings. The same reference numerals indicate the same components. Also, reference numerals in the drawings may be omitted.

First Embodiment
The first embodiment relates to a nonlinear resistance element having excellent electrical characteristics as a first example of the present disclosure. An example of the configuration of the nonlinear resistance element of the first embodiment will be described with reference to FIG.

As shown in FIG. 1, the nonlinear resistance element 100 of the first embodiment includes a first electrode 101, a first nonlinear resistance layer 102 made of an amorphous chalcogenide thin film, an intermediate layer 103, a second nonlinear resistance layer 104 made of an amorphous chalcogenide thin film, and a second electrode 105, and these electrodes and layers are stacked in this order.

In this manner, the intermediate layer 103 is provided between the first nonlinear resistance layer 102 and the second nonlinear resistance layer 104 .
The intermediate layer 103 functions as a gettering material for the second group elements or the third group elements (e.g., As, Se, etc.) constituting the amorphous chalcogenide thin film of the first nonlinear resistance layer 102 and the second nonlinear resistance layer 104. The elements constituting the amorphous chalcogenide thin film diffuse and move toward the intermediate layer 103, and the concentration becomes high near the interface between the first nonlinear resistance layer 102 and the intermediate layer 103 and near the interface between the second nonlinear resistance layer 104 and the intermediate layer 103. Therefore, a composition gradient occurs across the first nonlinear resistance layer 102, the intermediate layer 103, and the second nonlinear resistance layer 104.
This prevents the elements constituting the amorphous chalcogenide thin film from diffusing to the element surface or from being desorbed to the outside of the element, and allows them to be stably present in the first nonlinear resistance layer 102 and the second nonlinear resistance layer 104. As a result, a nonlinear resistance element with excellent electrical characteristics can be obtained.
Furthermore, by providing intermediate layer 103, it is possible to suppress the desorption of elements constituting the amorphous chalcogenide thin film during film formation and after element formation, and as described above, it is possible to form a region having a high concentration of elements constituting the amorphous chalcogenide thin film.

The nonlinear resistance element 100 may be changed to a low resistance state by applying a voltage whose absolute value is equal to or greater than a first threshold voltage between the first and second electrodes, and may be changed to a high resistance state by applying a voltage whose absolute value is equal to or less than a second threshold voltage, the absolute value of which is less than the first threshold voltage.

The first nonlinear resistance layer 102 and the second nonlinear resistance layer 104 are made of an amorphous chalcogenide thin film. The material of the amorphous chalcogenide thin film can be, for example, an amorphous chalcogenide containing a first group element consisting of at least one element of Ge and Si.

The amorphous chalcogenide thin film may include a second group element consisting of at least one element selected from the group consisting of Se, Te, and S, for example.

In an embodiment in which the amorphous chalcogenide thin film contains a second group element, the composition of the second group elements constituting the amorphous chalcogenide thin film of the first nonlinear resistance layer 102 may change from the interface between the first nonlinear resistance layer 102 and the first electrode 101 toward the interface between the first nonlinear resistance layer 102 and the intermediate layer 103.
In the same embodiment, the composition of the second group elements constituting the amorphous chalcogenide thin film of the second nonlinear resistance layer 104 may change from the interface between the second nonlinear resistance layer 104 and the second electrode 105 toward the interface between the second nonlinear resistance layer 104 and the intermediate layer 103.
The above-mentioned compositional changes contribute to the above-mentioned compositional gradient, and therefore facilitate improvement of the electrical characteristics of the nonlinear resistance element.

The amorphous chalcogenide thin film may contain a third group element consisting of at least one element selected from the group consisting of As, Sb, and N.

In an embodiment in which the amorphous chalcogenide thin film contains a third group element, the composition of the third group elements constituting the amorphous chalcogenide thin film of the first nonlinear resistance layer 102 may change from the interface between the first nonlinear resistance layer 102 and the first electrode 101 toward the interface between the first nonlinear resistance layer 102 and the intermediate layer 103.
In the same embodiment, the composition of the third group elements constituting the amorphous chalcogenide thin film of the second nonlinear resistance layer 104 may change from the interface between the second nonlinear resistance layer 104 and the second electrode 105 toward the interface between the second nonlinear resistance layer 104 and the intermediate layer 103.
The above-mentioned compositional changes contribute to the above-mentioned compositional gradient, and therefore facilitate improvement of the electrical characteristics of the nonlinear resistance element.

The composition of the first nonlinear resistance layer 102 may be the same as or different from the composition of the second nonlinear resistance layer 104.

The film thickness of the first nonlinear resistance layer 102 and the second nonlinear resistance layer 104 is not particularly limited. The thicker the film thickness, the higher the first threshold voltage at which the element resistance becomes low, and the smaller the leakage current in the high resistance state tends to be. Therefore, the film thickness of the first nonlinear resistance layer 102 and the second nonlinear resistance layer 104 is adjusted according to the characteristics determined by the materials such as the composition of the first nonlinear resistance layer 102 and the second nonlinear resistance layer 104 and the required electrical characteristics of the nonlinear resistance element. The film thickness of the first nonlinear resistance layer 102 and the second nonlinear resistance layer 104 may be set to, for example, 5 nm to 50 nm.

The film thickness of the first nonlinear resistance layer 102 may be the same as or different from the film thickness of the second nonlinear resistance layer 104.

The intermediate layer 103 is made of a metal containing one of the elements Ti, Zr, Ta, Nb, Mo, W, Hf, and V, for example.
The intermediate layer 103 may also contain Si or C, for example.
Thus, the intermediate layer 103 may contain at least one element selected from the group consisting of Ti, Zr, Ta, Nb, Mo, W, Hf, V, Si, and C, for example.

The thickness of the intermediate layer 103 is not particularly limited, but is preferably 0.5 nm or more, for example. The thickness of the intermediate layer 103 is preferably 2 nm or less, and more preferably 1 nm or less.

Materials for the first electrode 101 and the second electrode 105 can be, for example, Cu, Al, Ti, Zr, Ta, Nb, Mo, W, Hf, and V, as well as nitrides thereof.

The material of the first electrode 101 may be the same as or different from the material of the second electrode 105.

A lower wiring may be provided below the first electrode 101 (i.e., on the side of the first electrode 101 opposite to the side having the first nonlinear resistance layer 102). The material of the lower wiring may be a metal containing one of the elements Cu, Al, or W.

A specific method for forming the nonlinear resistance element 100 of the first embodiment will be described below.
As the first electrode 101, for example, a TiN film having a thickness of 10 nm is deposited on the substrate.
The TiN film can be formed by DC sputtering. For example, argon (Ar) gas and nitrogen (N ₂ ) gas are introduced into a sputtering chamber depressurized to about 5×10 ⁻⁶ Pa, and a direct current (DC) power of 500 W to 2 kW is applied to a titanium (Ti) target to deposit the TiN film.

Next, the first nonlinear resistance layer 102 is formed.
As the first nonlinear resistance layer 102, for example, a Ge-Si-As-Se based amorphous chalcogenide film (amorphous chalcogenide thin film) having a thickness of 20 nm is deposited.
The amorphous chalcogenide thin film can be formed by simultaneous sputtering using a Si target and, for example, a Ge-As-Se compound target (chalcogenide target). For example, argon (Ar) gas is introduced into a sputtering chamber decompressed to about 5×10 ⁻⁶ Pa, DC power or RF power of 10 W to 500 W is applied to the Si target, and RF power of 10 W to 2 kW is applied to the chalcogenide target. The composition of the first nonlinear resistance layer 102 can be adjusted by adjusting the composition of the chalcogenide target or the power applied to the Si target and the chalcogenide target during simultaneous sputtering. Although a normal sintered target can be used as the chalcogenide target, chalcogenide glass having a composition such as Ge ₂₂ As ₂₀ Se ₅₈ can also be used as the target. When chalcogenide glass is used as a target, the uniformity of the target density and composition is superior to that of a sintered body, so that the first nonlinear resistance layer 102 having excellent uniformity can be obtained. In addition, the nonlinear resistance layer 102 can be formed by normal single sputtering using only a chalcogenide target, instead of simultaneous sputtering with Si as described above, to form an amorphous chalcogenide thin film that functions as the first nonlinear resistance layer 102. Furthermore, when nitrogen (N ₂ ) gas is used in addition to argon (Ar) gas during sputtering, nitrogen can be added to the first nonlinear resistance layer 102.

The first electrode 101 and the first nonlinear resistance layer 102 can be formed consecutively in the same vacuum by switching the gas and target.

Next, the intermediate layer 103 is formed.
As the intermediate layer 103, for example, a Ti film having a thickness of 0.5 nm is deposited.
The Ti film can be formed by DC sputtering. For example, only argon (Ar) gas is introduced into a sputtering chamber depressurized to about 5×10 ⁻⁶ Pa, and a direct current (DC) power of 500 W to 2 kW is applied to a titanium (Ti) target to deposit a Ti film on the first nonlinear resistance layer 102.

Next, the second nonlinear resistance layer 104 is formed.
As the second nonlinear resistance layer 104, for example, a GeSiAsSe-based amorphous chalcogenide thin film having a thickness of 20 nm is deposited in the same manner as the first nonlinear resistance layer 102.
The amorphous chalcogenide thin film can be formed by simultaneous sputtering using a Si target and, for example, a GeAsSe chalcogenide target. For example, argon (Ar) gas is introduced into a sputtering chamber decompressed to about 5×10 ⁻⁶ Pa, and DC or RF power of 10 W to 500 W is applied to the Si target, and RF power of 10 W to 2 kW is applied to the chalcogenide target. The composition of the second nonlinear resistance layer 102 can be adjusted by adjusting the composition of the chalcogenide target or the power applied to the Si target and the chalcogenide target during simultaneous sputtering. The other features of the first nonlinear resistance layer 102 are as described above.

It is preferable to deposit the first nonlinear resistance layer 102, the intermediate layer 103, and the second nonlinear resistance layer 104 successively in the same vacuum. This keeps the interface between the first nonlinear resistance layer 102 and the intermediate layer 103, and the interface between the intermediate layer 103 and the second nonlinear resistance layer 104 clean, allowing the effect of the intermediate layer 103 to be maximized.

Next, the second electrode 105 is formed.
As the second electrode 105, for example, a TiN film having a thickness of 25 nm is deposited.
The second electrode 105 is formed by DC sputtering using argon (Ar) gas and nitrogen (N ₂ ) gas, similarly to the first electrode 101 .

The second nonlinear resistance layer 104 and the second electrode 105 can be formed consecutively in the same vacuum by switching the gas and target.

In order to fabricate a device from the laminate obtained by the above-mentioned forming method, it is preferable to use a microfabrication technique using so-called photolithography and dry etching.
Furthermore, a lower wiring buried in an insulating film may be formed in advance before forming the first electrode 101. The lower wiring is further covered with an insulating film and is electrically connected to the first electrode 101 through the opening.

(Experimental Example)
Here, an experimental example in which the functional effects of the intermediate layer 103 of the nonlinear resistance element 100 were investigated and analyzed will be described.

In this example, as shown in FIG. 2, a sample was fabricated by laminating, from the bottom, a 10 nm thick TiN electrode film, a 2 nm thick Ti film, and a 20 nm thick amorphous Ge-As-Se chalcogenide film, and then forming a 2 nm thick Ti film and a 25 nm thick TiN electrode film. The 2 nm thick Ti film provided at the interface between the upper and lower TiN electrodes and the Ge-As-Se chalcogenide amorphous film has a different layered structure from that of the first embodiment, but has the same function as the intermediate layer 103 of the nonlinear resistance element 100. In other words, it has a gettering effect on the second group elements or the third group elements.

The cross section of this sample was observed with a transmission electron microscope (TEM), and an example of the results of analyzing the composition of the amorphous chalcogenide thin film is shown in Table 1. As shown in Figure 2, the composition of the amorphous chalcogenide thin film was analyzed at five points (positions a to e) spaced approximately equally in the film thickness direction, with the interface between the upper Ti film and the amorphous chalcogenide as a and the interface between the amorphous chalcogenide thin film and the lower Ti as e.

As shown in Table 1, the composition of the amorphous chalcogenide thin film for the nonlinear resistance element, which is an example of this experimental example, changed in the film thickness direction. To facilitate understanding of the composition change, the relative compositions of As and Se with respect to Ge are also shown in Table 1. The relative compositions of As and Se were larger at the interface a between the upper Ti film and the amorphous chalcogenide thin film, and at the interface e between the amorphous chalcogenide thin film and the lower Ti film, compared to positions b to d. This shows that As, a third group element, and Se, a second group element, migrate toward the interface with the Ti film. In particular, As also migrates from the center c of the amorphous chalcogenide thin film toward the interface with the Ti film.

From the above, it is shown that the Ti film provided at the interface between the TiN electrode and the amorphous chalcogenide thin film functions as a gettering material for As and Se, which are elements that make up the amorphous chalcogenide thin film of the nonlinear resistance layer, and can strongly adsorb and fix As and Se.
Furthermore, by placing a thin film having a gettering effect on As and Se adjacent to the amorphous chalcogenide thin film, it is possible to concentrate As and Se, forming a composition region having higher As and Se concentrations than the composition at the time of film formation.
When a thin film having a gettering effect is inserted as the intermediate layer 103 in the center of the amorphous chalcogenide thin film as in the first embodiment, a region with high As and Se concentrations can be provided in the center of the amorphous chalcogenide thin film. This allows As and Se, which are prone to diffusion, to be firmly fixed in the film and prevented from diffusing to the element surface. In addition, As and Se, which have high vapor pressure and are prone to desorb outside the element, can be stably retained within the element. As a result, the nonlinear resistance element of the first embodiment exhibits stable electrical characteristics.
For the intermediate layer 103, Ti, Zr, Ta, Nb, Mo, W, Hf, and V are suitable as materials that exhibit a gettering effect on As, Se, and the like.
In addition, the vapor pressure of Se, As, etc., which constitute the second or third group, is high, and in the sputtering deposition of the amorphous chalcogenide thin film, Se, As, etc., are easily volatilized, and it is difficult to increase the concentration in the film. As described above, by arranging Ti, etc., which has a gettering effect on Se, As, etc., which constitute the second or third group, in contact with the amorphous chalcogenide thin film, it is possible to concentrate Se, As, etc. Therefore, it is possible to make the Se concentration and As concentration higher than when the amorphous chalcogenide thin film is deposited, and the composition adjustment range of the amorphous chalcogenide thin film can be widened. As a result, a nonlinear element with excellent electrical characteristics can be obtained.

Figure 3 is a diagram showing an example of the current-voltage characteristics (IV characteristics) of the nonlinear resistance element 100 of the first embodiment. In this example, the first electrode 101 is a 10 nm-thickness TiN film, the second electrode 105 is a 25 nm-thickness TiN film, and the intermediate layer 103 is a 0.5 nm-thickness Ti film. The first nonlinear resistance layer 102 and the second nonlinear resistance layer 104 are GeAsSe-based amorphous chalcogenide thin films with a thickness of 25 nm.

The I-V characteristics were measured by applying a voltage of 0V to -2V to the second electrode of the nonlinear resistance element. In order to prevent damage to the element, the current was limited by the measuring device to not exceed 100μA.

The nonlinear resistance element, which is an example of the first embodiment, exhibited strong nonlinearity, and the current increased rapidly especially near -1.2V. This indicates that the state changed to a low resistance state with a first threshold voltage of about 1.2V. The current value in the low resistance state was limited to 100 μA, and it can be seen that a current of 100 μA or more actually flows. The current ratio of this nonlinear resistance element when about 0.6V was applied to it and when it was in the low resistance state of 1.25V was about 1.5×10 ^6. Thus, the nonlinear resistance element, which is an example of the first embodiment, has excellent electrical characteristics.

Comparative Example
As a comparative example, an example of a nonlinear resistance element according to the related art will be described with reference to Fig. 4. As shown in Fig. 4, a nonlinear resistance element 300 according to the related art includes a first electrode 301, a nonlinear resistance layer 303, and a second electrode 305, and these electrodes and layers are laminated in this order.

As the material for the first electrode 301 and the second electrode 305, for example, a metal nitride containing any one of the elements Ti, V, Zr, Nb, Ta, Hf, and W can be used.
The nonlinear resistance layer 303 is made of an amorphous chalcogenide thin film. The material of the amorphous chalcogenide thin film may be an amorphous chalcogenide containing at least one element selected from the group consisting of Ge and Si, and at least one element selected from the group consisting of Se, Te, and S. The amorphous chalcogenide thin film may further contain at least one element selected from the group consisting of As, Sb, and N.

A specific method for forming the nonlinear resistance element 300 according to the related art will be described below.
As the first electrode 301, for example, a TiN film having a thickness of 10 nm is deposited on the substrate.
The TiN film can be formed by DC sputtering. For example, argon (Ar) gas and nitrogen (N ₂ ) gas are introduced into a sputtering chamber depressurized to about 1×10 ⁻⁶ Pa, and a direct current (DC) power of 500 W to 2 kW is applied to a titanium (Ti) target to deposit the TiN film.

Next, the nonlinear resistance layer 303 is formed.
As the nonlinear resistance layer 303, for example, a Ge--Se based amorphous chalcogenide film (amorphous chalcogenide thin film) having a thickness of 50 nm is deposited.
As a method for forming an amorphous chalcogenide thin film, a simultaneous sputtering method using a Ge target and a Ge-Se compound target having a composition of, for example, Ge ₂₀ Se ₈₀ can be used. For example, argon (Ar) gas is introduced into a sputtering chamber decompressed to about 1×10 ⁻⁶ Pa, DC power or RF power of 10 W to 500 W is applied to the Ge target, and RF power of 100 W to 2 kW is applied to the Ge-Se compound target. The composition of the nonlinear resistance layer 303 can be adjusted by adjusting the composition of the Ge-Se compound target or the power applied to the Ge target and the Ge-Se compound target during simultaneous sputtering. In addition, when nitrogen (N ₂ ) gas is used in addition to argon (Ar) gas during simultaneous sputtering, nitrogen can be added to the nonlinear resistance layer 303.
In this comparative example, for example, an amorphous chalcogenide thin film having a composition of Ge ₄₇ Se ₅₃ and a thickness of about 50 nm is formed.

Next, the second electrode 305 is formed.
As the second electrode 305, for example, a TiN film having a thickness of 25 nm is deposited.
The second electrode 305 is formed by DC sputtering using argon (Ar) gas and nitrogen (N ₂ ) gas, similarly to the first electrode 301 .

5 is a diagram showing an example of the current-voltage characteristics (IV characteristics) of the comparative nonlinear resistance element 300. In this example, the first electrode 301 is a TiN film with a thickness of 10 nm, the second electrode 305 is a TiN film with a thickness of 25 nm, and the nonlinear resistance layer 303 is an amorphous chalcogenide thin film with a composition of Ge ₄₇ Se ₅₃ and a thickness of 50 nm.

As shown in Figure 5, the nonlinear resistance element of the comparative example had low nonlinearity in its electrical characteristics. Thus, the nonlinear resistance element of the comparative example has inferior electrical characteristics.

(Modification)
A configuration example of the modified nonlinear resistance element will be described with reference to Fig. 6. As shown in Fig. 6, the modified nonlinear resistance element 500 includes a first electrode 501, a first diffusion barrier layer 502, a nonlinear resistance layer 503 made of an amorphous chalcogenide thin film, a second diffusion barrier layer 504, and a second electrode 505, and these electrodes and layers are laminated in this order.

In this manner, the nonlinear resistance layer 503 is provided between the first diffusion barrier layer 502 and the second diffusion barrier layer 504 .
The first diffusion barrier layer 502 and the second diffusion barrier layer 504 function as diffusion prevention layers for elements (e.g., As, Se, etc.) constituting the amorphous chalcogenide thin film of the nonlinear resistance layer 503. The first diffusion barrier layer 502 and the second diffusion barrier layer 504 suppress the elements constituting the amorphous chalcogenide thin film of the nonlinear resistance layer 503 from diffusing and moving to the first electrode 501 and the second electrode 505, and maintain the composition of the nonlinear resistance layer 503 as a switch layer. Therefore, a nonlinear resistance element with excellent electrical characteristics can be obtained.

The first diffusion barrier layer 502 and the second diffusion barrier layer 504 are preferably made of a metal containing one of the elements Ru, Ir, Pt, Au, or Rh. The first diffusion barrier layer 502 and the second diffusion barrier layer 504 are preferably made of a nitride such as Ti, Zr, Ta, Nb, Mo, W, or Hf.

The composition of the first diffusion barrier layer 502 may be the same as or different from the composition of the second diffusion barrier layer 504.

The film thickness of the first diffusion barrier layer 502 and the second diffusion barrier layer 504 is not particularly limited, but is preferably, for example, 3 nm or less.

The thickness of the first diffusion barrier layer 502 may be the same as or different from the thickness of the second diffusion barrier layer 504.

The nonlinear resistance layer 503 is made of an amorphous chalcogenide thin film. The material of the amorphous chalcogenide thin film can be, for example, an amorphous chalcogenide containing a first group element consisting of at least one element of Ge and Si.

The amorphous chalcogenide thin film may include a second group element consisting of at least one element selected from the group consisting of Se, Te, and S, for example.

The amorphous chalcogenide thin film may contain a third group element consisting of at least one element selected from the group consisting of As, Sb, and N.

The thickness of the nonlinear resistance layer 503 is not particularly limited. As the thickness increases, the first threshold voltage at which the element resistance becomes low tends to increase, and the leakage current in the high resistance state tends to decrease. Therefore, the thickness of the nonlinear resistance layer 503 is adjusted according to the characteristics determined by the material, such as the composition of the nonlinear resistance layer 503, and the required electrical characteristics of the nonlinear resistance element. The thickness of the nonlinear resistance layer 503 may be set to, for example, 5 nm to 50 nm.

Materials for the first electrode 501 and the second electrode 505 can be, for example, Cu, Al, Ti, Zr, Ta, Nb, Mo, W, Hf, and V, as well as nitrides thereof.

The material of the first electrode 501 may be the same as or different from the material of the second electrode 505.

A lower wiring may be provided below the first electrode 501 (i.e., on the side of the first electrode 501 opposite to the side having the first diffusion barrier layer 502). The material of the lower wiring may be a metal mainly composed of one of Cu, Al, or W.

A specific method for forming the modified nonlinear resistance element 500 will be described below.
As the first electrode 501, for example, a TiN film having a thickness of 10 nm is deposited on the substrate.
The TiN film can be formed by DC sputtering. For example, argon (Ar) gas and nitrogen (N ₂ ) gas are introduced into a sputtering chamber depressurized to about 5×10 ⁻⁶ Pa, and a direct current (DC) power of 500 W to 2 kW is applied to a titanium (Ti) target to deposit the TiN film.

Next, a first diffusion barrier layer 502 is formed.
As the first diffusion barrier layer 502, for example, a Ru film having a thickness of 2 nm is deposited.
The Ti film can be formed by DC sputtering. For example, only argon (Ar) gas is introduced into a sputtering chamber depressurized to about 5×10 ⁻⁶ Pa, and a direct current (DC) power of 500 W to 2 kW is applied to a ruthenium (Ru) target to deposit a Ru film on the first electrode 501.

The first electrode 501 and the first diffusion barrier layer 502 can be formed consecutively in the same vacuum by switching the gas and target.

Next, the nonlinear resistance layer 503 is formed.
As the nonlinear resistance layer 503, for example, a Ge--As--Se based amorphous chalcogenide film (amorphous chalcogenide thin film) having a thickness of 20 nm is deposited.
As a method for forming an amorphous chalcogenide thin film, a simultaneous sputtering method using a Si target and, for example, a Ge-As-Se compound target (chalcogenide target) can be used. For example, argon (Ar) gas is introduced into a sputtering chamber decompressed to about 1×10 ⁻⁶ Pa, DC power or RF power of 10 W to 500 W is applied to the Si target, and RF power of 100 W to 2 kW is applied to the chalcogenide target. The composition of the chalcogenide target compound target or the power applied to the Si target and the chalcogenide target during simultaneous sputtering can be adjusted to adjust the composition of the nonlinear resistance layer 503. In addition, when nitrogen (N ₂ ) gas is used in addition to argon (Ar) gas during simultaneous sputtering, nitrogen can be added to the nonlinear resistance layer 503.

The first diffusion barrier layer 502 and the nonlinear resistance layer 503 can be formed consecutively in the same vacuum by switching the gas and target.

Next, a second diffusion barrier layer 504 is formed.
As the first diffusion barrier layer 504, for example, a Ru film having a thickness of 2 nm is deposited.
The Ti film can be formed by DC sputtering. For example, only argon (Ar) gas is introduced into a sputtering chamber depressurized to about 5×10 ⁻⁶ Pa, and a direct current (DC) power of 500 W to 2 kW is applied to a ruthenium (Ru) target to deposit a Ru film on the nonlinear resistance layer 503.

The nonlinear resistance layer 503 and the second diffusion barrier layer 504 can be formed consecutively in the same vacuum by switching the gas and target.

Next, the second electrode 505 is formed.
As the second electrode 505, for example, a TiN film having a thickness of 25 nm is deposited.
The second electrode 505 is formed by DC sputtering using argon (Ar) gas and nitrogen (N ₂ ) gas, similarly to the first electrode 501 .

The second diffusion barrier layer 504 and the second electrode 505 can be formed consecutively in the same vacuum by switching the gas and target.

In order to fabricate a device from the laminate obtained by the above-mentioned forming method, it is preferable to use a microfabrication technique using so-called photolithography and dry etching.
Furthermore, a lower wiring buried in an insulating film may be formed in advance before forming the first electrode 501. The lower wiring is further covered with an insulating film and is electrically connected to the first electrode 501 through the opening.

Below, aspects related to the modified examples will be additionally described.
(Appendix 1)
A first electrode;
a first diffusion barrier layer;
a nonlinear resistance layer made of an amorphous chalcogenide thin film;
a second diffusion barrier layer;
a second electrode;
a first electrode, a first diffusion barrier layer, a nonlinear resistance layer, a second diffusion barrier layer, and a second electrode laminated in this order;
A nonlinear resistive element.
(Appendix 2)
A nonlinear resistance element as described in Appendix 1, wherein the amorphous chalcogenide thin film includes a first group element consisting of at least one element of Ge and Si, and a second group element consisting of at least one element selected from the group consisting of Se, Te, and S.
(Appendix 3)
The nonlinear resistance element according to claim 1 or 2, wherein the amorphous chalcogenide thin film contains a third group element consisting of at least one element selected from the group consisting of As, Sb, and N.
(Appendix 4)
A method for manufacturing a nonlinear resistance element according to any one of claims 1 to 3, comprising the steps of:
A first electrode;
a first diffusion barrier layer;
a nonlinear resistance layer made of an amorphous chalcogenide thin film;
a second diffusion barrier layer;
a first electrode and a second electrode are laminated in this order in the same vacuum.
A method for manufacturing a nonlinear resistance element.

Second Embodiment
The second embodiment, as a second example of the present disclosure, relates to a nonlinear resistance element fabricated in an integrated circuit.

When fabricating a nonlinear resistance element as described above into an integrated circuit, it is necessary to ensure compatibility with general integrated circuit processes. In general, transition metal nitrides are more resistant to oxidation, corrosive gases, and their plasma than simple metals, and are highly compatible with lithography and dry etching processes. Therefore, when fabricating a nonlinear resistance element into an integrated circuit, it is preferable to cover the upper and lower surfaces of the nonlinear resistance element with metal nitride in order to ensure compatibility with the integrated circuit process.

FIG. 7 is a cross-sectional view showing a nonlinear resistance element 600 fabricated in an integrated circuit.
As shown in FIG. 7, the nonlinear resistance element 600 has a first electrode 601, a first nonlinear resistance layer 602, an intermediate layer 603, a second nonlinear resistance layer 604, a second electrode 605, a first insulating film 606, a second insulating film 607, a Cu electrode 608, and a substrate 609.

A Cu electrode 608 is formed on a substrate 609, and the Cu electrode 608 is embedded in a first insulating film 606 and a second insulating film 607. An opening is provided in the second insulating film 607, and a TiN film is formed as a first electrode 601 so as to cover the opening. Since the Cu electrode 608 is covered with the TiN film, the Cu electrode 608 can be prevented from being oxidized or corroded during the formation of the first nonlinear resistance layer 602, the intermediate layer 603, the second nonlinear resistance layer 604, the second electrode 605, and the like. An amorphous chalcogenide thin film is formed as the first nonlinear resistance layer 602 on the first electrode 601, and a Ti film, for example, is formed as the intermediate layer 603 on the first nonlinear resistance layer 602. An amorphous chalcogenide thin film is formed as the second nonlinear resistance layer 604 on the intermediate layer 603. In addition, a TiN film is formed on the second nonlinear resistance layer 604 as a second electrode 605.

Details of the materials of the first electrode 601, the first nonlinear resistance layer 602, the intermediate layer 603, the second nonlinear resistance layer 604, and the second electrode 605 are as described above in the first embodiment for the first electrode, the first nonlinear resistance layer, the intermediate layer, the second nonlinear resistance layer, and the second electrode.

The nonlinear resistance element 600 is further embedded in an insulating film (not shown), and a through hole (not shown) is formed in the insulating film to reach the second electrode 605. Metal nitrides such as TiN are excellent in terms of process compatibility and electrical properties as an etching stopper and electrical contact material when forming the through hole. In addition, if the nitride materials constituting the first electrode 601 and the second electrode 605 are each the metal nitride of the electrode material constituting the intermediate layer 603, this is preferable in terms of the configuration of the film formation device and the configuration of an efficient manufacturing process.

Third Embodiment
The third embodiment relates to a crossbar switch having switching elements including the nonlinear resistance elements and nonvolatile resistance change elements described in the first embodiment, as a third example of the present disclosure.

The switching element is provided in the signal path of the logic circuit, and has two two-terminal type nonlinear resistance elements according to the present disclosure and two two-terminal type nonvolatile resistance change elements. In addition, one terminal of each of the two two-terminal type nonvolatile resistance change elements is connected to each other and to one electrode of each of the two two-terminal type nonlinear resistance elements.

The nonvolatile resistance change element is a two-terminal element consisting of an active electrode, an inactive electrode, and a resistance change film sandwiched between the active electrode and the inactive electrode. Details are disclosed in JP-A-2016-203751 and other publications.

FIG. 8 is an equivalent circuit diagram showing an example of the configuration of the crossbar switch 700 according to the third embodiment.
8, the crossbar switch 700 of the third embodiment has a plurality of switching elements 730 arranged in an array. The switching elements 730 correspond to switching elements provided in the signal paths of a logic circuit.
The switching element 730 has nonvolatile variable resistance elements 731 and 732, and nonlinear resistance elements 721 and 722. The nonlinear resistance elements 721 and 722 are each a two-terminal element corresponding to the nonlinear resistance element 100 shown in the first embodiment.

The nonvolatile resistance change elements 731 and 732 have their inactive electrodes connected to each other. The active electrode of the nonvolatile resistance change element 731 is connected to a first wiring 741.
The active electrode of the nonvolatile resistance change element 732 is connected to the second wiring 742 .
Of the two electrodes of the non-linear resistance element 721 , one electrode is connected to an inactive electrode of the non-volatile resistance change element 731 , and the other electrode is connected to a third wiring 743 .
Of the two electrodes of the non-linear resistance element 722 , one electrode is connected to the inactive electrode of the non-volatile resistance change element 732 , and the other electrode is connected to a fourth wiring 744 .
8, the first wiring 741 and the third wiring 743 are arranged in parallel, and the second wiring 742 and the fourth wiring 744 are arranged in parallel. The first wiring 741 and the third wiring 743 are perpendicular to the other two wirings (the second wiring 742 and the fourth wiring 744).

Next, a method of programming the switching element 730 shown in FIG. 8 will be described.
When the nonvolatile resistance change element 731 is transitioned to the ON state (low resistance state), the third wiring 743 is grounded, and a positive voltage equal to or greater than the threshold voltage (set voltage) is applied to the first wiring 741 .
On the other hand, when the nonvolatile resistance change element 731 is transitioned from the ON state to the OFF state (high resistance state), the first wiring 741 is grounded and a positive voltage equal to or greater than the threshold voltage (reset voltage) is applied to the third wiring 743 .
Moreover, when the nonvolatile resistance change element 732 is transitioned to the ON state, the fourth wiring 744 is grounded, and a positive voltage equal to or greater than the threshold voltage (set voltage) is applied to the second wiring 742 .
On the other hand, when the nonvolatile resistance change element 732 is changed from the ON state to the OFF state, the second wiring 742 is grounded, and a positive voltage equal to or higher than the threshold voltage (reset voltage) is applied to the fourth wiring 744 .
In this manner, the nonvolatile resistance change element 731 can be programmed via the nonlinear resistance element 721 , and the nonvolatile resistance change element 732 can be programmed via the nonlinear resistance element 722 .

As described above, the nonlinear resistance element of the first embodiment of the present disclosure has excellent nonlinearity. Therefore, by using the nonlinear resistance element of the first embodiment as a two-terminal selection element for selecting the nonvolatile resistance change element to be programmed, it is possible to prevent erroneous writing and malfunction of the switching element. As a result, a nonvolatile switching element that does not use a selection transistor can be realized, and a crossbar switch with excellent integration density and high reliability can be provided. This makes it possible to manufacture an excellent FPGA.

Fourth Embodiment
The fourth embodiment relates to an element structure for realizing a switching element including the nonlinear resistance element and the nonvolatile resistance change element described in the third embodiment, as a fourth example of the present disclosure.

FIG. 9 is a cross-sectional view showing an example of an element structure of a semiconductor device 800. As shown in FIG.
A semiconductor device 800 shown in FIG. 9 includes switching elements 822a and 822b.
The switching element 822 a has an active electrode (first wiring 805 a ) of the nonvolatile resistance change element, a resistance change film 809 of the nonvolatile resistance change element, a first electrode 810 , a nonlinear resistance layer 811 , and a second electrode 812 .
The first electrode 810 is a multilayer film that functions as both an inactive electrode of the nonvolatile resistance change element and a first electrode of the nonlinear resistance element (corresponding to the first electrode 101 shown in FIG. 1).
The second electrode 812 is a second electrode of the nonlinear resistance element (corresponding to the second electrode 105 shown in FIG. 1).
In addition, the nonlinear resistance layer 811 is a multilayer film that combines the first nonlinear resistance layer of the nonlinear resistance element (corresponding to the first nonlinear resistance layer 102 shown in Figure 1), the intermediate layer of the nonlinear resistance element (corresponding to the intermediate layer 103 shown in Figure 1), and the second nonlinear resistance layer of the nonlinear resistance element (corresponding to the first nonlinear resistance layer 104 shown in Figure 1).

A nonvolatile resistance change element is formed from the active electrode (first wiring 805a), the resistance change film 809, and the first electrode 810, and a nonlinear resistance element is formed from the first electrode 810, the nonlinear resistance layer 811, and the second electrode 812. A stacked body 840 shown in FIG. 9 corresponds to the resistance change film 809, the first electrode 810, the nonlinear resistance layer 811, and the second electrode 812.
The switching element 822 b has an active electrode (first wiring 805 b ), a resistance change film 809 , a first electrode 810 , a nonlinear resistance layer 811 , and a second electrode 812 , similar to the switching element 822 a .

The switching elements 822a and 822b share the resistance change film 809, the first electrode 810, and the nonlinear resistance layer 811. In addition, the switching elements 822a and 822b each have a second electrode 812 that serves as a control electrode. The second electrode 812 of the switching element 822a is connected to the second wiring 818a via the barrier metal 820a and the plug 819a. The second wiring 818a is integrated with the plug 819a. The second electrode 812 of the switching element 822b is connected to the second wiring 818b via the barrier metal 820b and the plug 819b. The second wiring 818b is integrated with the plug 819b.

9, the multilayer wiring structure has an insulating laminate in which an interlayer insulating film 802, a barrier insulating film 803, an interlayer insulating film 804, an insulating barrier film 807, a protective insulating film 814, an interlayer insulating film 817, a hard mask film 816, and a barrier insulating film 821 are laminated in this order on a semiconductor substrate (not shown). The multilayer wiring structure has first wirings 805a and 805b and second wirings 818a and 818b. The first wirings 805a and 805b are embedded in wiring grooves formed in the interlayer insulating film 804 and the barrier insulating film 803 via barrier metals 806a and 806b. The second wirings 818a and 818b and plugs 819a and 819b are embedded in wiring grooves formed in the interlayer insulating film 817 and the hard mask film 816. The second wiring 818a and the plug 819a are integrated, and the side and bottom surfaces of the second wiring 818a and the plug 819a are covered with a barrier metal 820a.

A portion of the upper surface of the first wirings 805a and 805b, which serve as the active electrodes of the nonvolatile resistance change elements of the switching elements 822a and 822b, is exposed in an opening formed in the insulating barrier film 807. A resistance change film 809, a first electrode 810, a nonlinear resistance layer 811, and a second electrode 812 are laminated in this order on the wall surface of the opening in the insulating barrier film 807 and on the insulating barrier film 807. The switching elements 822a and 822b are complementary resistance change elements with a nonlinear resistance element.

After processing the second electrode 812 into a desired shape, a protective insulating film 814 is formed. In the element structure shown in FIG. 9, the side of the laminate 840 consisting of the resistance change film 809, the first electrode 810, the nonlinear resistance layer 811, and the second electrode 812 is covered with a protective insulating film 814. The first wirings 805a and 805b also serve as active electrodes of the nonvolatile resistance change element, which simplifies the number of manufacturing steps and reduces the electrode resistance. By simply fabricating at least two mask sets as an additional step to the normal Cu damascene wiring process, the nonvolatile resistance change element can be mounted, and low resistance and low cost of the element can be achieved at the same time.

In the fourth embodiment, the variable resistance non-volatile elements used for the switching elements 822a and 822b utilize metal ion migration and electrochemical reactions in an ion conductor. The non-volatile variable resistance elements are controlled to be on/off by applying a voltage or passing a current. The non-volatile variable resistance elements are controlled to be on/off by, for example, electrodiffusion of metal related to the first wirings 805a and 805b into the variable resistance film 809.

The structure of each film shown in FIG. 9 will be described.
9 is a substrate on which a semiconductor element is formed. For example, a silicon substrate, a single crystal substrate, a silicon on insulator (SOI) substrate, a thin film transistor (TFT) substrate, a substrate for manufacturing liquid crystal, or the like can be used as the semiconductor substrate.

The interlayer insulating film 802 is an insulating film formed on a semiconductor substrate. For example, a silicon oxide film or a low dielectric constant film (e.g., a SiOCH film) having a lower relative dielectric constant than a silicon oxide film can be used for the interlayer insulating film 802. The interlayer insulating film 802 may be a laminate of a plurality of insulating films.
The interlayer insulating film 804 may be made of the same type of film as the interlayer insulating film 802 .

The barrier insulating film 803 is an insulating film having barrier properties and provided between the interlayer insulating film 802 and the interlayer insulating film 804. The barrier insulating film 803 serves as an etching stop layer when the first wirings 805a and 805b are formed in the wiring trenches. As the barrier insulating film 803, for example, a SiN film, a SiC film, a SiCN film, or the like can be used.

The insulating barrier film 807 is an insulating film formed on the interlayer insulating film 804. For example, a silicon oxide film or a low dielectric constant film (e.g., a SiOCH film) having a lower relative dielectric constant than a silicon oxide film can be used for the insulating barrier film 807. The insulating barrier film 807 may be a laminate of multiple insulating films.

In the interlayer insulating film 804 and the barrier insulating film 803, wiring trenches for burying first wirings 805a and 805b are formed, and the first wirings 805a and 805b are buried in the wiring trenches via barrier metals 806a and 806b.
The first wirings 805a and 805b are wirings embedded in wiring grooves formed in the interlayer insulating film 804 and the barrier insulating film 803 via barrier metals 806a and 806b. The first wirings 805a and 805b also serve as active electrodes of the nonvolatile resistance change elements of the switching elements 822a and 822b, and are in contact with the resistance change film 809. An electrode layer or the like may be inserted between the first wirings 805a and 805b and the resistance change film 809. When an electrode layer is formed, the electrode layer and the resistance change film 809 are deposited in a continuous process and processed in a continuous process. In addition, the lower part of the resistance change film 809 is not connected to the lower wiring via a contact plug. For the first wirings 805a and 805b, a metal capable of diffusion and ion conduction in the resistance change film 809 is used, and for example, Cu or the like can be used. The first wirings 805a and 805b may be alloyed with Al, Mn, or the like.

The barrier metals 806a and 806b are conductive films having barrier properties that cover the side and bottom surfaces of the first wirings 805a and 805b in order to prevent the metal of the first wirings 805a and 805b from diffusing into the interlayer insulating film 802 and the lower layers. For example, when the first wirings 805a and 805b are made of metal elements mainly composed of Cu, the barrier metals 806a and 806b can be made of high-melting point metals such as tantalum (Ta), tantalum nitride (TaN), titanium nitride (TiN), and tungsten carbonitride (WCN) or their nitrides, or a laminate film thereof.

The insulating barrier film 807 is an insulating film formed on the interlayer insulating film 804 including the first wirings 805a and 805b. The insulating barrier film 807 prevents oxidation of (e.g., Cu) constituting the first wirings 805a and 805b, prevents diffusion of the metal constituting the first wirings 805a and 805b into the interlayer insulating film 804, and serves as an etching stop layer when processing the second electrode 812, the nonlinear resistance layer 811, the first electrode 810, and the resistance change film 809. For the insulating barrier film 807, for example, a silicon oxide film (SiO _x ), a low dielectric constant film having a lower relative dielectric constant than a silicon oxide film (e.g., a SiOCH film), or the like can be used. For the insulating barrier film 807, for example, a SiC film, a SiCN film, a SiN film, and a laminated structure thereof can be used. It is preferable that the insulating barrier film 807 is made of the same material as the protective insulating film 814 and the hard mask film 816. The insulating barrier film 807 may be a laminate of a plurality of insulating films.

The insulating barrier film 807 has an opening on the first wiring 805a and 805b. In the opening of the insulating barrier film 807, the first wiring 805a and 805b are in contact with the resistance change film 809. The opening of the insulating barrier film 807 is formed within the region of the first wiring 805a and 805b. In this way, it becomes possible to form a non-volatile resistance change element on the surface of the first wiring 805a and 805b, which has small unevenness. The wall surface of the opening of the insulating barrier film 807 is a tapered surface that becomes wider as it moves away from the first wiring 805a and 805b. The tapered surface of the opening of the insulating barrier film 807 is set to 85° or less with respect to the upper surface of the first wiring 805a and 805b. In this way, electric field concentration on the outer periphery of the connection between the first wirings 805a and 805b and the resistance change film 809 (near the outer periphery of the opening of the insulating barrier film 807) is alleviated, and the insulation resistance can be improved.

The resistance change film 809 is a film whose resistance changes. For the resistance change film 809, a material whose resistance changes due to the action (diffusion, ion conduction, etc.) of the metal related to the first wiring 805a and 805b (lower electrode (active electrode) of the nonvolatile resistance change element) can be used. When the resistance change of the nonvolatile resistance change element is performed by deposition of metal ions, an ion-conductive film is used as the resistance change film 809. For example, Ta ₂ O ₅ , TaSiO, etc., which are oxide insulating films containing Ta, can be used. In addition, the resistance change film 809 can have a laminated structure in which Ta ₂ O ₅ and TaSiO are laminated in this order from the bottom. By making such a laminated structure, when the resistance change film 809 is used as a solid electrolyte, the bridge formed by the metal ion (for example, copper ion) formed inside the ion conduction layer at the time of low resistance (on time) can be broken by the Ta ₂ O ₅ layer, and the metal ion can be easily collected at the time of off, thereby improving the switching characteristics. The resistance change film 809 is formed on the first wirings 805a and 805b, the tapered surfaces of the openings of the insulating barrier film 807, and the insulating barrier film 807. The outer periphery of the resistance change film 809 at the connection portion between the first wirings 805a and 805b and the resistance change film 809 is provided at least along the tapered surfaces of the openings of the insulating barrier film 807.

Of the first electrode 810 having a multi-layer structure, the lower layer in direct contact with the resistance change film 809 is preferably made of a metal that is less likely to ionize than the metal of the first wiring 805a and 805b and is less likely to diffuse and conduct ions in the resistance change film 809. For example, Pt, Ru, etc. can be used. RuTa, RuTi, etc., whose main component is a metal material such as Pt or Ru, may also be used. The first electrode 810 is in direct contact with the resistance change film 809 on its lower surface and in contact with the nonlinear resistance layer 811 on its upper surface. Of the first electrode 810 having a multi-layer structure, the upper layer in direct contact with the nonlinear resistance layer is the first electrode of the nonlinear resistance element (corresponding to the first electrode 101 shown in FIG. 1). Details of the material of the upper layer of the first electrode 810 are as described above for the first electrode in the first embodiment.

The nonlinear resistance layer 811 is a multilayer film consisting of a first nonlinear resistance layer, an intermediate layer, and a second nonlinear resistance layer (corresponding to the first nonlinear resistance layer 102, the intermediate layer 103, and the second nonlinear resistance layer 104 shown in FIG. 1). Details of the material of the nonlinear resistance layer 811 are as described above for the first nonlinear resistance layer, the intermediate layer, and the second nonlinear resistance layer in the first embodiment.

The second electrode 812 is the second electrode of the nonlinear resistance element (corresponding to the second electrode 105 shown in FIG. 1). The upper surface of the second electrode 812 is in contact with the barrier metals 820a and 820b. The upper layer of the second electrode 812 functions as an etching stopper when processing via holes to form the plugs 819a and 819b. Details of the material of the second electrode 812 are as described above for the second electrode in the first embodiment.

The second electrode 812 is electrically connected to the plugs 819a and 819b via the barrier metals 820a and 820b. The diameter R2 (or the area of the area) of the circle where the second electrode 812 and the plugs 819a and 819b (strictly speaking, the barrier metals 820a and 820b) contact each other is set to be smaller than the diameter R1 (or the area of the area) of the circle where the first wirings 805a and 805b contact the resistance change film 809. By setting it in this way, poor filling of the pilot holes formed in the interlayer insulating film 817, which are the connection parts between the second electrode 812 and the plugs 819a and 819b, with plating (e.g., copper plating) is suppressed, and the occurrence of voids can be suppressed.

It is preferable that the protective insulating film 814 and the insulating barrier film 807 are made of the same material. In other words, by surrounding the switching elements 822a and 822b entirely with the same material, the material interfaces are integrated, preventing the intrusion of moisture and the like from the outside and preventing oxygen from being released from the switching elements 822a and 822b themselves.

The protective insulating film 814 is an insulating film that prevents oxygen from being released from the resistance change film 809 without damaging the switching elements 822a and 822b. For example, a SiN film, a SiCN film, or the like can be used for the protective insulating film 814. It is preferable that the protective insulating film 814 is made of the same material as the hard mask film 816 and the insulating barrier film 807. When these films are made of the same material, the protective insulating film 814 is integrated with the insulating barrier film 807 and the hard mask film 816, improving the adhesion at the interface and providing better protection for the switching elements 822a and 822b.

The interlayer insulating film 817 is an insulating film formed on the protective insulating film 814. For example, a silicon oxide film (SiO _x ), a SiOC film, a low dielectric constant film (e.g., a SiOCH film) having a lower relative dielectric constant than a silicon oxide film, or the like can be used for the interlayer insulating film 817. The interlayer insulating film 817 may be a laminate of a plurality of insulating films. In the interlayer insulating film 817, pilot holes for embedding plugs 819a and 819b and wiring grooves for embedding second wirings 818a and 818b are formed. The plugs 819a and 819b and the second wirings 818a and 818b are embedded in the pilot holes and wiring grooves via barrier metals 820a and 820b.

The second wiring 818a and 818b are wirings embedded in wiring grooves formed in the interlayer insulating film 817 via barrier metals 820a and 820b. The second wiring 818a is integrated with a plug 819a. The plug 819a is embedded in a pilot hole formed in the interlayer insulating film 817 and the hard mask film 816 via a barrier metal 820a. The plug 819a is electrically connected to the switching element 822a via the barrier metal 820a. The second wiring 818a and the plug 819a can be made of, for example, Cu. The second wiring 818b and the plug 819b have the same configuration as the second wiring 818a and the plug 819a.

The barrier metals 820a and 820b are conductive films having barrier properties that cover the side and bottom surfaces of the second wirings 818a and 818b and the plugs 819a and 819b in order to prevent the metal constituting the second wirings 818a and 818b (including the plugs 819a and 819b) from diffusing into the interlayer insulating film 817 and the lower layers. For example, when the second wirings 818a and 818b and the plugs 819a and 819b are made of metal elements mainly composed of Cu, the barrier metals 820a and 820b can be made of high-melting point metals such as tantalum (Ta), tantalum nitride (TaN), titanium nitride (TiN), and tungsten carbonitride (WCN), their nitrides, or laminated films thereof.

The barrier insulating film 821 is an insulating film formed on the hard mask film 816 including the second wirings 818a and 818b, and serves to prevent oxidation of the metal (e.g., Cu) constituting the second wirings 818a and 818b and to prevent the diffusion of the metal constituting the second wirings 818a and 818b to upper layers. For the barrier insulating film 821, for example, a SiC film, a SiCN film, a SiN film, or a laminated structure thereof can be used.

The above element structure makes it possible to realize a non-volatile switching element that does not use a selection transistor, and to provide a crossbar switch with excellent integration density and high reliability. This makes it possible to manufacture excellent FPGAs.

The above description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure, and are merely an example of the technology of the present disclosure. For example, the above explanation of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above description and illustrations, within the scope of the gist of the technology of the present disclosure. Also, in order to avoid confusion and to facilitate understanding of the parts related to the technology of the present disclosure, the above description and illustrations omit explanations of technical common knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

100 Nonlinear resistance element 101 First electrode 102 First nonlinear resistance layer 103 Intermediate layer 104 Second nonlinear resistance layer 105 Second electrode 300 Nonlinear resistance element 301 First electrode 303 Nonlinear resistance layer 305 Second electrode 500 Nonlinear resistance element 501 First electrode 502 First diffusion barrier layer 503 Nonlinear resistance layer 504 Second diffusion barrier layer 505 Second electrode 600 Nonlinear resistance element 601 First electrode 602 First nonlinear resistance layer 603 Intermediate layer 604 Second nonlinear resistance layer 605 Second electrode 606 First insulating film 607 Second insulating film 608 Cu electrode 609 Substrate 700 Crossbar switch 721, 722 Nonlinear resistance element 730 Switching element 731, 732 Nonvolatile resistance change element 741 First wiring 742 Second wiring 743 Third wiring 744 Fourth wiring 800 Semiconductor device 802 Interlayer insulating film 803 Barrier insulating film 804 Interlayer insulating films 805a, 805b First wirings 806a, 806b Barrier metal 807 Insulating barrier film 809 Resistance change film 810 First electrode 811 Non-linear resistance layer 812 Second electrode 814 Protective insulating film 816 Hard mask film 817 Interlayer insulating films 818a, 818b Second wirings 819a, 819b Plugs 820a, 820b Barrier metal 821 Barrier insulating films 822a, 822b Switching element 840 Stacked body

Claims (8)

A first electrode;
a first nonlinear resistance layer made of an amorphous chalcogenide thin film;
The middle class,
a second nonlinear resistance layer made of an amorphous chalcogenide thin film;
a second electrode;
the first electrode, the first nonlinear resistance layer, the intermediate layer, the second nonlinear resistance layer, and the second electrode are laminated in this order ;
the amorphous chalcogenide thin film includes a first group element consisting of at least one element of Ge and Si, and a second group element consisting of at least one element selected from the group consisting of Se, Te, and S;
a composition of the second group elements constituting the amorphous chalcogenide thin film of the first nonlinear resistance layer changes from an interface between the first nonlinear resistance layer and the first electrode toward an interface between the first nonlinear resistance layer and the intermediate layer,
a composition of the second group elements constituting the amorphous chalcogenide thin film of the second nonlinear resistance layer changes from an interface between the second nonlinear resistance layer and the second electrode toward an interface between the second nonlinear resistance layer and the intermediate layer,
the second group element constituting the amorphous chalcogenide thin film has a high concentration in the vicinity of the interface between the first nonlinear resistance layer and the intermediate layer and in the vicinity of the interface between the second nonlinear resistance layer and the intermediate layer;
A nonlinear resistive element. The nonlinear resistance element according to claim 1, wherein the amorphous chalcogenide thin film contains a third group element consisting of at least one element selected from the group consisting of As, Sb, and N. a composition of the third group elements constituting the amorphous chalcogenide thin film of the first nonlinear resistance layer changes from an interface between the first nonlinear resistance layer and the first electrode toward an interface between the first nonlinear resistance layer and the intermediate layer,
a composition of the third group elements constituting the amorphous chalcogenide thin film of the second nonlinear resistance layer changes from an interface between the second nonlinear resistance layer and the second electrode toward an interface between the second nonlinear resistance layer and the intermediate layer;
3. The nonlinear resistance element according to claim 2 . The nonlinear resistance element according to any one of claims 1 to 3 , wherein the intermediate layer contains at least one element selected from the group consisting of Ti, Zr, Ta, Nb, Mo, W, Hf, V, Si, and C. 5. The nonlinear resistance element according to claim 1 , wherein the intermediate layer has a thickness of 2 nm or less. A nonlinear resistance element according to any one of claims 1 to 5, which changes to a low resistance state by applying a voltage having an absolute value equal to or greater than a first threshold voltage between the first electrode and the second electrode, and changes to a high resistance state by applying a voltage having an absolute value equal to or less than a second threshold voltage which is less than the first threshold voltage. A switching element provided in a signal path of a logic circuit,
Two two-terminal nonlinear resistance elements;
Two two-terminal nonvolatile resistance change elements;
one terminal of each of the two two-terminal nonvolatile resistance change elements is connected to each other and to one electrode of each of the two two-terminal nonlinear resistance elements;
The two two-terminal nonlinear resistance elements are the nonlinear resistance elements according to any one of claims 1 to 6 .
Switching element. A method for manufacturing a nonlinear resistance element according to any one of claims 1 to 6 , comprising the steps of:
The first electrode;
the first nonlinear resistance layer made of the amorphous chalcogenide thin film;
The intermediate layer;
the second nonlinear resistance layer made of the amorphous chalcogenide thin film;
the second electrode and the first electrode are laminated in this order in the same vacuum.
A method for manufacturing a nonlinear resistance element.