JP2010114231A - Variable resistance element and method for manufacturing the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a variable resistance element which is rewritable and nonvolatile, and to provide a method for manufacturing the same. <P>SOLUTION: The variable resistance element X includes: an oxide part 4 having P-type semiconductive property; a pair of electrodes 1 joined to the oxide part 4 while separating from each other; a second electrode 2 joined to the oxide part 4 between the pair of electrodes 1; and an electrode 3 joined to the oxide part 4 while having a portion opposed to the electrode 2 through the oxide part 4. The method for manufacturing a variable resistance element includes, for example, steps of: forming a conductive material film on a substrate S; forming an oxide film having P-type semiconductive property on the conductive material film; and forming a pair of electrodes 1 separated from each other and an electrode 2 located between the electrodes 1 on the oxide film. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a resistance variable element having a portion whose resistance state can be changed, and a method of manufacturing such a resistance variable element.

  A field effect transistor is known as one of elements having a portion whose resistance state can be changed. A field effect transistor is, for example, a three-terminal element having a source electrode, a drain electrode, and a gate electrode, and a current passing through a region (channel) located between the source and drain by changing a voltage applied to the gate electrode. It is an element that can control. In a field effect transistor, an electric field corresponding to a voltage applied to a gate electrode is generated in a channel, and the resistance of the channel is changed by the generation and change of the electric field. Such field effect transistors are described in, for example, Patent Documents 1 to 3 below.

JP 2006-186336 A JP 2006-237304 A JP 2006-295653 A

  However, in order to keep the resistance state of the channel constant in the field effect transistor, a constant voltage must be continuously applied to the gate electrode. That is, the field effect transistor is a volatile element with respect to the resistance state.

  The present invention has been conceived under the circumstances as described above, and an object thereof is to provide a rewritable and nonvolatile variable resistance element and a method for manufacturing the same.

  According to the first aspect of the present invention, a resistance variable element is provided. This resistance variable element includes an oxide portion, a pair of first electrodes, a second electrode, and a third electrode. The oxide part has P-type semiconductivity. The pair of first electrodes are separated from each other and joined to the oxide portion. The second electrode is joined to the oxide portion between the pair of first electrodes. The third electrode has a portion facing the second electrode through the oxide portion and is joined to the oxide portion.

  The resistance variable element having such a configuration is configured such that the oxide between the pair of first electrodes is subjected to a change in voltage applied between the second and third electrodes after receiving a predetermined initialization process. The resistance state of the part can change.

  In this device, when the second electrode is used as a positive electrode and the third electrode is used as a negative electrode and an initialization process is performed by applying a predetermined voltage between the electrodes for a predetermined time, carrier movement between the pair of first electrodes in the device is performed. It is possible to limit the route within the oxide part. Specifically, in the initialization process, a relatively large amount of oxygen ions are generated in the oxide portion by the electric field action between the second electrode (positive electrode) and the third electrode (negative electrode), and the oxygen ions are It is possible to draw electrons from the oxygen ions to the second electrode by attracting the two electrodes. Oxygen that has been deprived of electrons oxidizes the end face on the oxide portion side of the second electrode. At the same time, a relatively large amount of oxygen vacancies with positive charges are unevenly distributed in the vicinity of the interface with the second electrode inside the oxide portion from which a certain amount of constituent oxygen has been released to the second electrode side. On the other hand, the electrons that have moved from the oxide portion to the second electrode flow into the third electrode through the external circuit. In the third electrode, the electrons move to a partial region that forms an electric field in cooperation with the second electrode, but at the interface between the third electrode and the P-type semiconducting oxide portion, the third electrode The exchange of electrons and holes does not occur sufficiently so that electrons flow from the electrode to the oxide part. Therefore, a relatively large amount of electrons that have flowed into the third electrode accumulate in a partial region that forms an electric field in cooperation with the second electrode in the third electrode. The portion where the electrons are locally accumulated in the third electrode has an excessively high electron density, and the degree of freedom of electron movement is considerably reduced. Therefore, the resistance is considerably increased. . On the other hand, in the oxide portion, as described above, oxygen vacancies with positive charges are unevenly distributed near the interface with the second electrode. Therefore, the route of carrier movement between the first electrodes in this element when a voltage is applied between the pair of first electrodes is substantially limited within the oxide portion (P-type semiconductivity). The main carrier in the oxide part is a hole). By the initialization process as described above, a channel or a current path is formed in the oxide portion, and the device can be operated. Even if the applied voltage for the initialization process is extinguished, the channel formation state, that is, the local electron accumulation state in the third electrode is maintained. An electrostatic attractive force is generated between the accumulated electrons in the third electrode and the above-described oxygen vacancies existing in the oxide part with a positive charge, and the action of the electrostatic attractive force causes the electrostatic force in the third electrode. This is because the diffusion of the stored electrons is prevented.

  In this element that has undergone initialization processing, when a predetermined voltage (high resistance voltage) smaller than the initialization voltage is applied between the electrodes with the second electrode as the positive electrode and the third electrode as the negative electrode, It is possible to increase the resistance between the first electrodes. Specifically, by applying a high resistance voltage, an electrostatic repulsive force is generated between oxygen vacancies (with positive charges) that are unevenly distributed in the vicinity of the second electrode in the oxide portion and the second electrode (positive electrode). And oxygen vacancies can be dispersed to the third electrode side to a predetermined extent. That is, a region where the density of oxygen vacancies accompanied by positive charges is a predetermined value or more can be expanded to a predetermined degree toward the third electrode side by applying a high resistance voltage. Due to such relaxation of the uneven distribution of oxygen vacancies, the channel in the oxide portion (the route of carrier movement) is partially narrowed, and the resistance between the first electrodes in this element becomes higher (higher resistance). ). The degree of uneven relaxation of oxygen vacancies, that is, the degree of resistance increase, can be adjusted by adjusting the magnitude of the high resistance voltage and the application time. Moreover, even if the high resistance voltage is extinguished, the high resistance state is maintained. An electrostatic attractive force is generated between the accumulated electrons in the third electrode and the above-described oxygen vacancies existing in the oxide part with a positive charge, and the action of the electrostatic attractive force causes This is because the diffusion of oxygen vacancies is blocked and the distribution of oxygen vacancies is fixed.

  In this element having undergone high resistance, when a second electrode is used as a positive electrode and a third electrode is used as a negative electrode, a predetermined voltage that is lower than the initialization voltage and higher than the already applied high resistance voltage is applied between the electrodes for a predetermined time. The resistance between the first electrodes in the element can be further increased. Further, in this element having undergone high resistance, the second electrode is used as a positive electrode and the third electrode is used as a negative electrode, and a voltage smaller than the initialization voltage and having the same magnitude as the already applied high resistance voltage is set between the electrodes. When applied over time, it may be possible to further increase the resistance between the first electrodes in the device. Specifically, by the application of the further increased resistance voltage, between the oxygen vacancies (with positive charge) unevenly distributed on the second electrode side in the oxide portion and the second electrode (positive electrode). It is possible to further disperse oxygen vacancies to the third electrode side by generating electrostatic repulsion. That is, it is possible to further bulge the region where the density of oxygen vacancies accompanied by positive charges is a predetermined value or more to the third electrode side by applying a higher resistance voltage. By further mitigating the uneven distribution of oxygen vacancies, the channel (carrier transfer route) in the oxide portion is partially further narrowed, and the resistance between the first electrodes in the device is further increased (further. Higher resistance). The degree of further uneven relaxation of oxygen vacancies, that is, the degree of further increase in resistance can be adjusted by adjusting the magnitude of the further higher resistance voltage and the application time. Moreover, even if the further high resistance voltage is extinguished, the high resistance state is maintained. An electrostatic attractive force is generated between the accumulated electrons in the third electrode and the above-described oxygen vacancies existing in the oxide part with a positive charge, and the action of the electrostatic attractive force causes This is because the diffusion of oxygen vacancies is blocked and the distribution of oxygen vacancies is fixed.

  In this element that has undergone high resistance (including further increase in resistance), when a predetermined voltage (low resistance voltage) is applied between the electrodes with the second electrode as a negative electrode and the third electrode as a positive electrode, It is possible to reduce the resistance between the first electrodes in the element. Specifically, the application of a low resistance voltage generates an electrostatic attractive force between oxygen vacancies (with positive charges) dispersed to a predetermined degree in the oxide portion and the second electrode (negative electrode). Thus, it is possible to draw oxygen vacancies to the second electrode side to a predetermined extent and make them unevenly distributed. That is, it is possible to reduce the region where the density of oxygen vacancies accompanied by positive charges is equal to or higher than a predetermined level to the second electrode side by applying a low resistance voltage. Due to the uneven distribution of oxygen vacancies, at least a part of the channel (route of carrier movement) in the oxide portion is expanded, and the resistance between the first electrodes in this element becomes lower (lower resistance). . The degree of uneven distribution of oxygen vacancies, that is, the degree of resistance reduction, can be adjusted by adjusting the magnitude of the resistance reduction voltage and the application time. Further, even if the low resistance voltage is extinguished, the low resistance state is maintained. An electrostatic attractive force is generated between the accumulated electrons in the third electrode and the above-described oxygen vacancies existing in the oxide part with a positive charge, and the action of the electrostatic attractive force causes This is because the diffusion of oxygen vacancies is blocked and the distribution of oxygen vacancies is fixed. This element in such a low resistance state can be increased in resistance again through the above-described process of increasing resistance.

  When the second electrode is used as a negative electrode and the third electrode is used as a positive electrode and a predetermined voltage greater than the already applied low resistance voltage is applied between the electrodes for a predetermined time in the present element that has undergone resistance reduction, It is possible to further reduce the resistance between the first electrodes. In addition, in this element having undergone the resistance reduction, the second electrode is used as the negative electrode and the third electrode is used as the positive electrode, and a voltage having the same magnitude as the already applied low resistance voltage is applied between the electrodes for a predetermined time. Then, it may be possible to further reduce the resistance between the first electrodes in the element. Specifically, by applying the further resistance-reducing voltage, an electrostatic charge is generated between the oxygen vacancies (with positive charges) and the second electrode (negative electrode) that are unevenly distributed in the oxide portion to a predetermined extent. It is possible to generate an attractive force and further draw oxygen vacancies toward the second electrode to make them unevenly distributed. That is, it is possible to further reduce the region where the density of oxygen vacancies accompanied by positive charges is equal to or higher than a predetermined level to the second electrode side by applying a further low resistance voltage. Due to the further uneven distribution of oxygen vacancies, at least a part of the channel (carrier transfer route) in the oxide portion is further expanded, and the resistance between the first electrodes in the device is further reduced (further. Low resistance). The degree of further uneven distribution of oxygen vacancies, that is, the degree of further lowering the resistance, can be adjusted by adjusting the magnitude of the further lowering resistance voltage and the application time. Moreover, even if the further low resistance voltage is extinguished, the low resistance state is maintained. An electrostatic attractive force is generated between the accumulated electrons in the third electrode and the above-described oxygen vacancies existing in the oxide part with a positive charge, and the action of the electrostatic attractive force causes This is because the diffusion of oxygen vacancies is blocked and the distribution of oxygen vacancies is fixed. The device in such a further low resistance state can be increased in resistance again through the above-described process of increasing resistance.

  As described above, in the resistance variable element, the channel or current path in the oxide portion between the pair of first electrodes is partially changed by changing the voltage applied between the second and third electrodes. By scaling, the resistance of the channel can be reversibly changed. That is, this element can rewrite the resistance state of the channel by controlling the voltage applied between the second and third electrodes. In such an element, it is possible to switch between a relatively high resistance state and a relatively low resistance state. Further, in this element, it is possible to rewrite between a relatively high resistance state and a relatively low resistance state. Furthermore, in this element, it is possible to rewrite between three or more resistance states (that is, this element has multivalued resistance states).

  In addition, in the resistance variable element, the resistance state generated in the oxide portion by applying a predetermined voltage can be maintained even when the voltage application is stopped. This is because, as described above, the accumulated electrons in the third electrode interact with oxygen vacancies present with a positive charge in the oxide portion.

  As described above, the resistance variable element is rewritable and nonvolatile. Such a resistance variable element can be used as a switching element, a memory element, or a programmable element (an element included in a logic device), and is suitable for reducing power consumption.

  In addition, the variable resistance element can be used as a three-terminal element in which the second electrode is a voltage control electrode and the pair of first electrodes is a current readout electrode. In such an element, voltage control that causes a resistance change and current reading in a predetermined resistance state can be performed using different electrodes. Therefore, this element tends to have a high degree of freedom in switching or rewriting the resistance state by applying a voltage.

  According to a second aspect of the present invention, a resistance variable element manufacturing method is provided. This method includes a conductive material film forming step, an oxide film forming step, and an electrode forming step. In the conductive material film forming step, a conductive material film is formed on the substrate. In the oxide film forming step, an oxide film having P-type semiconductivity is formed on the conductive material film. In the electrode formation step, a pair of first electrodes spaced apart from each other and a second electrode positioned between the first electrodes are formed on the oxide film. In the electrode forming step, the first and second electrodes may be formed through different processes included in the step. According to such a method, the resistance variable element according to the first aspect of the present invention can be appropriately manufactured.

  According to a third aspect of the present invention, a resistance variable element manufacturing method is provided. This method includes an electrode forming step, an oxide film forming step, and a conductive material film forming step. In the electrode forming step, a pair of first electrodes spaced apart from each other and a second electrode positioned between the first electrodes are formed on the substrate. In the electrode forming step, the first and second electrodes may be formed through different processes included in the step. In the oxide film forming step, an oxide film having P-type semiconductivity is formed so as to cover the pair of first electrode and second electrode. In the conductive material film forming step, a conductive material film is formed on the oxide film. According to such a method, the resistance variable element according to the first aspect of the present invention can be appropriately manufactured.

  1 and 2 show a resistance variable element X according to the present invention. FIG. 1 is a cross-sectional view of the resistance variable element X. FIG. 2 is a plan view of the resistance variable element X, and corresponds to an arrow view along the line II-II in FIG.

  The resistance variable element X includes a substrate S, a pair of electrodes 1, electrodes 2 and 3, and an oxide layer 4, and the oxide layer 4 between the electrodes 1 can be rewritten between a plurality of resistance states. It is possible.

The substrate S is, for example, a silicon substrate or an oxide substrate. A thermal oxide film may be formed on the surface of the silicon substrate. Examples of the oxide substrate include a MgO substrate, a SrTiO ₃ substrate, an Al ₂ O ₃ substrate, a quartz substrate, and a glass substrate.

The pair of electrodes 1 are joined to the oxide layer 4 while being separated from each other. Each electrode 1 is a first electrode in the present invention, and is made of a non-oxidizing metal material that is not easily oxidized by oxygen. Such an electrode 1 includes, for example, a conductive material selected from the group consisting of Pt, Au, and SrRuO ₃ . The thickness of the electrode 1 is, for example, 20 to 100 nm.

  The electrodes 2 are joined to the oxide layer 4 between the electrodes 1. The electrode 2 is the second electrode in the present invention, and is made of an oxidizing metal material that is easily oxidized by oxygen. Preferably, the electrode 2 is made of a material that is more easily oxidized than the electrode 1 and / or the electrode 3. Such an electrode 2 comprises, for example, a conductive material selected from the group consisting of Ta, Ti, Al, Fe, Co, and Ni. The thickness of the electrode 2 is, for example, 20 to 100 nm.

The electrode 3 is the third electrode in the present invention, and is joined to the oxide layer 4 such that the oxide layer 4 is interposed between the electrodes 1 and the oxide layer 4 is interposed between the electrodes 2. . That is, the electrode 3 has a portion that faces each electrode 1 via the oxide layer 4 and a portion that faces the electrode 2 via the oxide layer 4. Such an electrode 3 comprises, for example, a conductive material selected from the group consisting of Pt, Au, and SrRuO ₃ . The thickness of the electrode 3 is, for example, 5 to 10 nm.

The oxide layer 4 is an oxide part in the present invention, is located between the electrodes 1 and 2 and the electrode 3 and has P-type semiconductivity. The oxide layer 4 has a first surface 4a, a second surface 4b opposite to the first surface 4a, and a side surface 4c. The electrodes 1 and 2 are bonded to the first surface 4a side, and the electrode 3 is bonded to the second surface 4b side. In the present embodiment, the oxide layer 4 is made of an oxygen ion conductor, for example, a fluorite structure type oxide, a perovskite structure type oxide, a pyrochlore structure type oxide, a tungsten bronze structure type oxide, or a brown miralite structure. It consists of a type oxide. As the fluorite structure type oxide, ZrO ₂ and Y ₂ O ₃ can be employed. As the perovskite structure type oxide, PrCaMnO ₃ , SrTiO ₃ , LaSrCoO _{3 and the} like can be adopted. As the pyrochlore structure type oxide, Nd ₂ Mo ₂ O ₇ or the like can be used. As the tungsten bronze structure type oxide, CuWO ₃ or the like can be adopted. Sr ₂ Fe ₂ O ₅ or the like can be adopted as the brown mirrorite structure type oxide. Preferably, the oxide layer 4 is made of a transition metal oxide having P-type semiconductivity, and the transition metal oxide includes a transition metal capable of changing the valence. Examples of such transition metals capable of changing the valence include Mn, Ti, and Co. The thickness of such an oxide layer 4 is, for example, 50 to 150 nm.

As shown in FIG. 2, in this embodiment, in the D ₂ direction orthogonal to the D ₁ direction in which the electrodes 1 and 2 are arranged, the length L ₁ at which each electrode 1 is joined to the oxide layer 4 is 2 is equal to the length L ₂ bonded to the oxide layer 4. Preferably, the length L ₁ at which the electrode ₁ is bonded to the oxide layer 4 is equal to or shorter than the length L ₂ at which the electrode 2 is bonded to the oxide layer 4. Further, the length of the electrode 3 in the D ₂ direction is, for example, 1 μm or less, and the length of the oxide layer 4 in the D ₂ direction is, for example, 1 μm or less.

  FIG. 3 shows a method for manufacturing the resistance variable element X. In this method, first, the material film 11 and the material film 12 are sequentially stacked on the substrate S as shown in FIG. For example, a sputtering method, a vacuum evaporation method, a CVD method, or a PLD (Pulsed Laser Deposition) method can be employed as the film formation method for the material films 11 and 12 ( Can be used). The material film 11 is made of the material described above with respect to the electrode 3. The material film 12 is made of the material described above with respect to the oxide layer 4.

  Next, the electrode 2 is formed on the material film 12 as shown in FIG. In forming the electrode 2, for example, after the material described above with respect to the electrode 2 is formed on the material film 12, the material film is patterned using a predetermined resist pattern as a mask.

  Next, as shown in FIG. 3C, a pair of electrodes 1 is formed on the material film 12. In forming the electrode 1, for example, after the material described above with respect to the electrode 1 is formed on the material film 12, the material film is patterned using a predetermined resist pattern as a mask.

  Next, as shown in FIG. 3D, the material films 11 and 12 are patterned to form the electrode 3 and the oxide layer 4. As described above, the resistance variable element X can be appropriately manufactured by forming the electrode 3, the oxide layer 4, and the electrodes 1 and 2 on the substrate S. In this method, as long as the resistance variable element X can be manufactured, the order of performing the above-described steps may be changed.

  FIG. 4 shows an initialization process of the resistance variable element X. In the initialization process, the electrode 2 is the positive electrode and the electrode 3 is the negative electrode, a predetermined voltage is applied between the electrodes 2 and 3 for a predetermined time, and the carrier movement route between the pair of electrodes 1 in the resistance variable element X is oxidized. Restricted to the physical layer 4. The applied voltage in the initialization process is, for example, 8 to 12 V, and the application time of the initialization voltage is, for example, 1 to 10 msec.

  Specifically, in the initialization process, a relatively large amount of oxygen ions in the oxide layer 4 as shown in FIG. 4A due to the electric field effect between the electrode 2 (positive electrode) and the electrode 3 (negative electrode). 5 and oxygen ions 5 are attracted to the electrode 2 as shown in FIG. In the initialization process, electrons are taken away from the oxygen ions 5 attracted to the electrode 2 to the electrode 2 side. The oxygen deprived of the electrons oxidizes the end face of the electrode 2 on the oxide layer 4 side. At the same time, within the oxide layer 4 from which a certain amount of constituent oxygen has been released to the electrode 2 side, a relatively large amount of oxygen vacancies 6 accompanied by positive charges are formed with the electrode 2 as shown in FIG. It is unevenly distributed near the interface.

  On the other hand, electrons moved from the oxide layer 4 to the electrode 2 flow into the electrode 3 through an external circuit. In the electrode 3, the electrons move to the partial region 3 a that forms an electric field in cooperation with the electrode 2, but at the interface between the electrode 3 and the P-type semiconducting oxide layer 4, The exchange of electrons and holes does not occur sufficiently so that electrons flow into the oxide layer 4. Therefore, a relatively large amount of electrons 7 that have flowed into the electrode 3 accumulate in the partial region 3 a in the electrode 3. The portion where the electrons 7 are locally accumulated in the electrode 3 has an excessively high electron density, and the degree of freedom of electron movement is considerably reduced. Therefore, the resistance is considerably increased. . On the other hand, in the oxide layer 4, the oxygen vacancies 6 accompanied by positive charges are unevenly distributed in the vicinity of the interface with the electrode 2 as described above. Therefore, the route of carrier movement between the electrodes 1 in the resistance variable element X when a voltage is applied between the pair of electrodes 1 is substantially limited within the oxide layer 4 (P-type half The main carriers in the conductive oxide layer 4 are holes).

  By the initialization process as described above, a channel or a current path is formed in the oxide layer 4 and the variable resistance element X can be operated. Even when the applied voltage for the initialization process is extinguished, the channel formation state, that is, the local electron accumulation state in the electrode 3 is maintained. An electrostatic attraction is generated between the accumulated electrons 7 in the electrode 3 and the oxygen vacancies 6 present in the oxide layer 4 with a positive charge. Due to the action of the electrostatic attraction, the electrode 3 This is because the diffusion of the stored electrons 7 is prevented.

  5 and 6 show the operation principle of the resistance variable element X. FIG. In the resistance variable element X that has undergone the initialization process, as shown in FIG. 5A, the electrode 2 is the positive electrode and the electrode 3 is the negative electrode, and a predetermined voltage (high resistance voltage) is applied between the electrodes 2 and 3 over a predetermined time. When applied, the resistance between the electrodes 1 in the resistance variable element X can be increased. The high resistance voltage is smaller than the initialization voltage, for example, 2 to 3V. The application time of the high resistance voltage is, for example, 50 to 100 nsec.

  Specifically, by applying a high resistance voltage, for example, as shown in FIG. 4C, between the oxygen vacancies 6 (with positive charges) unevenly distributed in the vicinity of the electrode 2 and the electrode 2 (positive electrode). For example, as shown in FIG. 5A, the oxygen vacancies 6 can be dispersed toward the electrode 3 by generating an electrostatic repulsive force. That is, the region 6A in which the existence density of oxygen vacancies 6 accompanied by positive charges is not less than a predetermined value can be expanded to a predetermined degree toward the electrode 3 side by applying a high resistance voltage. Due to the relaxation of the uneven distribution of the oxygen vacancies 6, the channel (carrier transfer route) in the oxide layer 4 is partially narrowed, and the resistance between the electrodes 1 in the resistance variable element X is higher. (High resistance). The degree of relaxation of uneven distribution of the oxygen vacancies 6, that is, the degree of increasing the resistance, can be adjusted by adjusting the magnitude of the increasing resistance voltage and the application time. Moreover, even if the high resistance voltage is extinguished, the high resistance state is maintained. An electrostatic attractive force is generated between the accumulated electrons 7 in the electrode 3 and the above-described oxygen vacancies 6 present in the oxide layer 4 with a positive charge. This is because the diffusion of the oxygen vacancies 6 in the layer 4 is prevented and the distribution of the oxygen vacancies 6 is fixed.

  In the resistance variable element X that has undergone high resistance, when a predetermined voltage (further high resistance voltage) is applied between the electrodes 2 and 3 with the electrode 2 as a positive electrode and the electrode 3 as a negative electrode over a predetermined time, the resistance variable element The resistance between the electrodes 1 in X can be further increased. The further high resistance voltage is smaller than the initialization voltage and larger than the above high resistance voltage, for example, 3 to 4V. In this case, the application time of the further high resistance voltage is, for example, 50 to 100 nsec. Alternatively, the further resistance increase voltage is smaller than the initialization voltage and the same magnitude as the above-described resistance increase voltage, for example, 2 to 3V. In this case, the application time of the further high resistance voltage is, for example, 500 to 1000 nsec.

  Specifically, by applying a higher resistance voltage, for example, as shown in FIG. 5A, an electrostatic repulsive force is generated between the oxygen vacancies 6 unevenly distributed on the electrode 2 side and the electrode 2. Thus, for example, as shown in FIG. 5B, the oxygen vacancies 6 can be further dispersed toward the electrode 3 side. That is, the region 6A in which the existence density of oxygen vacancies 6 accompanied by positive charges is not less than a predetermined value can be further expanded to the electrode 3 side by applying a higher resistance voltage. By further mitigating the uneven distribution of the oxygen vacancies 6, the channel (carrier transfer route) in the oxide layer 4 becomes partially narrower, and the resistance between the electrodes 1 in the resistance variable element X is as follows. It becomes higher (further high resistance). The degree of further uneven relaxation of the oxygen vacancies 6, that is, the degree of higher resistance can be adjusted by adjusting the magnitude of the further higher resistance voltage and the application time. Moreover, even if the further high resistance voltage is extinguished, the high resistance state is maintained. An electrostatic attractive force is generated between the accumulated electrons 7 in the electrode 3 and the above-described oxygen vacancies 6 present in the oxide layer 4 with a positive charge. This is because the diffusion of the oxygen vacancies 6 in the layer 4 is prevented and the distribution of the oxygen vacancies 6 is fixed.

  In the resistance variable element X that has undergone higher resistance (including further higher resistance), the electrode 2 is the negative electrode and the electrode 3 is the positive electrode, and a predetermined voltage (low resistance voltage) is applied between the electrodes 2 and 3 for a predetermined time. When applied, the resistance between the electrodes 1 in the resistance variable element X can be lowered. The low resistance voltage is set to such a magnitude that a significant amount of oxygen present on the end face (partial oxidation portion) on the oxide layer 4 side of the electrode 2 is not reduced by electrons. Further, the low resistance voltage is, for example, 3 to 4 V, and the application time of the low resistance voltage is, for example, 100 to 150 nsec.

  Specifically, by applying a low resistance voltage, for example, between the oxygen vacancies 6 dispersed in the oxide layer 4 and the electrode 2 as shown in FIG. It is possible to generate an electrostatic attraction force to attract the oxygen vacancies 6 toward the electrode 2 and make them unevenly distributed as shown in FIG. That is, the region 6A in which the existence density of oxygen vacancies 6 accompanied by positive charges is not less than a predetermined value can be reduced to a predetermined level toward the electrode 2 by applying a low resistance voltage. Due to the uneven distribution of the oxygen vacancies 6, at least a part of the channel (carrier movement route) in the oxide layer 4 is expanded, and the resistance between the electrodes 1 in the resistance variable element X becomes lower. (Low resistance). The degree of uneven distribution of the oxygen vacancies 6, that is, the degree of resistance reduction, can be adjusted by adjusting the magnitude of the resistance reduction voltage and the application time. Further, even if the low resistance voltage is extinguished, the low resistance state is maintained. An electrostatic attractive force is generated between the accumulated electrons 7 in the electrode 3 and the above-described oxygen vacancies 6 present in the oxide layer 4 with a positive charge. This is because the diffusion of the oxygen vacancies 6 in the layer 4 is prevented and the distribution of the oxygen vacancies 6 is fixed. The resistance variable element X in such a low resistance state can be increased in resistance again through the above-described process of increasing resistance.

  When a predetermined voltage is applied between the electrodes 2 and 3 with the electrode 2 as a negative electrode and the electrode 3 as a positive electrode over a predetermined time in the resistance variable element X that has undergone a reduction in resistance, the resistance between the electrodes 1 in the resistance variable element X is reduced. The resistance can be further reduced. The further resistance reduction voltage is set to such a magnitude that a significant amount of oxygen present on the end face (partial oxidation portion) on the oxide layer 4 side of the electrode 2 is not reduced by electrons. Such a further resistance reduction voltage is larger than the above-described resistance reduction voltage, for example, 4 to 5V. In this case, the application time of the further low resistance voltage is, for example, 100 to 150 nsec. Alternatively, the further resistance reduction voltage is the same magnitude as the above-described resistance reduction voltage, for example, 3 to 4V. In this case, the application time of the further low resistance voltage is, for example, 1 to 5 μsec.

  Specifically, by applying the further low resistance voltage, for example, as shown in FIG. 6A, an electrostatic attractive force is generated between the oxygen vacancy 6 and the electrode 2 that are unevenly distributed to a predetermined degree. Thus, for example, as shown in FIG. 6B, the oxygen vacancies 6 can be further drawn toward the electrode 2 to be unevenly distributed. That is, the region 6A in which the existence density of oxygen vacancies 6 accompanied by positive charges is not less than a predetermined value can be further reduced to the electrode 2 side by applying a further low resistance voltage. Due to the further uneven distribution of the oxygen vacancies 6, at least a part of the channel (carrier transfer route) in the oxide layer 4 is further expanded, and the resistance between the electrodes 1 in the resistance variable element X is further increased. Lower (further lower resistance). The degree of further uneven distribution of the oxygen vacancies 6, that is, the degree of further lowering the resistance, can be adjusted by adjusting the magnitude of the further lowering resistance voltage and the application time. Moreover, even if the further low resistance voltage is extinguished, the low resistance state is maintained. An electrostatic attractive force is generated between the accumulated electrons 7 in the electrode 3 and the above-described oxygen vacancies 6 present in the oxide layer 4 with a positive charge. This is because the diffusion of the oxygen vacancies 6 in the layer 4 is prevented and the distribution of the oxygen vacancies 6 is fixed. The resistance variable element X in such a further low resistance state can be increased in resistance again through the above-described process of increasing resistance.

  As described above, in the resistance variable element X, the channel or current path in the oxide layer 4 between the pair of electrodes 1 is partially expanded or contracted by changing the voltage applied between the electrodes 2 and 3. Thus, the resistance of the channel can be reversibly changed. That is, the resistance variable element X can rewrite the resistance state of the channel by controlling the voltage applied between the electrodes 2 and 3. Such a resistance variable element X can be used by switching between a relatively high resistance state and a relatively low resistance state. The resistance variable element X can be used by rewriting between a relatively high resistance state and a relatively low resistance state. Furthermore, the resistance variable element X can be used by rewriting between a plurality of resistance states of three or more (that is, the resistance variable element X has multivalued resistance states).

  In addition, in the resistance variable element X, the resistance state generated in the oxide layer 4 by applying a predetermined voltage can be maintained even when the voltage application is stopped. This is because, as described above, the accumulated electrons 7 in the electrode 3 interact with the oxygen vacancies 6 existing in the oxide layer 4 with a positive charge.

  As described above, the resistance variable element X is rewritable and nonvolatile. Such a resistance variable element X can be used as a switching element, a memory element, or a programmable element (a resistance variable element included in a logic device), and is suitable for reducing power consumption.

  Further, the resistance variable element X can be used as a three-terminal element in which the electrode 2 is a voltage control electrode and the pair of electrodes 1 are current readout electrodes. In such a resistance variable element X, voltage control that causes a resistance change and current reading in a predetermined resistance state can be performed using different electrodes. Therefore, the resistance variable element X tends to have a high degree of freedom in switching or rewriting the resistance state by applying a voltage.

In the resistance variable element X, as described above, the length L ₁ at which each electrode 1 is bonded to the oxide layer 4 in the direction D ₂ shown in FIG. 2 is the length at which the electrode 2 is bonded to the oxide layer 4. L ₂ or less is preferred. Such a configuration appropriately distributes and unevenly distributes the oxygen vacancies 6 in the oxide layer 4 (channels for carrier movement) in the oxide layer 4 to be read by the pair of electrodes 1. This is suitable for limiting the area to be used.

  The electrode 2 in the resistance variable element X is preferably made of a material that is more easily oxidized than the electrode 1 and / or the electrode 3. Such a configuration is suitable for attracting the oxygen ions 5 generated in the oxide layer 4 to the electrode 2 and taking the electrons from the oxygen ions 5 in the initialization process described above.

  The thickness of the electrode 3 in the resistance variable element X is preferably 10 nm or less. It is preferable that the thickness of the electrode 3 is 10 nm or less and that the electrode 3 is sufficiently thin in order to efficiently increase the resistance of the electrode 3 due to the inflow of electrons (blocking the route of carrier movement) in the initialization process described above. is there.

The length of the electrode 3 in the resistance variable element X in the direction D ₂ shown in FIG. 2 is preferably 1 μm or less. Such a configuration is suitable for suppressing variation in the distribution in the D ₂ direction of the electrons 7 accumulated in the electrode 3 by the initialization process.

  As described above, the oxide layer 4 in the resistance variable element X is preferably made of a transition metal oxide having P-type semiconductivity, and the transition metal oxide includes a transition metal capable of changing the valence. . Such a configuration maintains the crystal structure of the oxide layer 4 even if oxygen vacancies 6 (with positive charges) are generated and increased in the oxide layer 4 through the above-described initialization process. It is suitable. The transition metal capable of changing the valence in the oxide layer 4 can change in valence so as to electrically compensate for the positive charge generated and increased in the oxide layer 4 by the initialization process.

  The thickness of the oxide layer 4 in the resistance variable element X is preferably 50 nm or more. In order to ensure a sufficient resistance change width of the channel in the oxide layer 4, the thickness of the oxide layer 4 is preferably 50 nm or more. The thickness of the oxide layer 4 is preferably 150 nm or less. After the initialization process, a sufficient interaction (electrostatic attractive force) is ensured between the accumulated electrons 7 in the electrode 3 and the oxygen vacancies (with positive charges) in the oxide layer 4 so that the resistance variable element X In order to ensure the non-volatility, the thickness of the oxide layer 4 is preferably 150 nm or less.

The length in the direction D ₂ shown in FIG. 2 of the oxide layer 4 in the resistance variable element X is preferably 1 μm or less. Such a configuration is suitable for suppressing variation in the distribution in the D ₂ direction of the oxygen vacancies 6 generated in the oxide layer 4 by the initialization process.

  FIG. 7 is a cross-sectional view of a first modification of the resistance variable element X. As shown in FIG. 7, in the resistance variable element X, each electrode 1 is not positioned on the first surface 4 a of the oxide layer 4 but is positioned on the side surface 4 c. You may join to. Even in such a configuration, the pair of electrodes 1 can function as current readout electrodes of the resistance variable element X1. Further, even when each of the electrodes 1 is bonded to the oxide layer 4 on the side surface 4c of the oxide layer 4, the pair of electrodes 1 functions as a current reading electrode of the resistance variable element X1. It is possible.

  8 and 9 show a second modification of the resistance variable element X. FIG. FIG. 8 is a cross-sectional view of a second modification. FIG. 9 is a plan view of the second modification, and corresponds to a view taken along the line IX-IX in FIG. In the resistance variable element X, as shown in FIGS. 8 and 9, the electrodes 1 and 2 are positioned on the substrate S, and the oxide layer 4 is formed on the substrate S so as to partially cover the electrodes 1 and 2. The electrode 3 may be located on the oxide layer 4. The variable resistance element X having such a structure can be subjected to the initialization process described above with reference to FIG. 4 and operates on the operation principle described above with reference to FIGS. It is possible to make it.

  FIG. 10 shows a manufacturing method of the second modified example of the resistance variable element X. In this method, first, an electrode 2 is formed on a substrate S as shown in FIG. In forming the electrode 2, for example, the material described above with respect to the electrode 2 is formed on the substrate S, and then the material film is patterned using a predetermined resist pattern as a mask.

  Next, as illustrated in FIG. 10B, a pair of electrodes 1 is formed on the substrate S. In the formation of the electrode 1, for example, after the material described above with respect to the electrode 1 is formed on the substrate S, the material film is patterned using a predetermined resist pattern as a mask.

  Next, as shown in FIG. 10C, the material film 12 and the material film 11 are sequentially formed so as to cover the electrodes 1 and 2. The material film 12 is made of the material described above with respect to the oxide layer 4. The material film 11 is made of the material described above with respect to the electrode 3.

  Next, as shown in FIG. 10D, the material films 11 and 12 are patterned to form the electrode 3 and the oxide layer 4. As described above, the electrodes 1 and 2, the oxide layer 4, and the electrode 3 can be formed on the substrate S to appropriately manufacture the second modification of the resistance variable element X. In this method, as long as the resistance variable element X can be manufactured, the order of performing the above-described steps may be changed.

A variable resistance element having the configuration shown in FIG. 11 was produced as an example of the variable resistance element X described above. The variable resistance element according to the present embodiment includes a substrate S that is an MgO single crystal substrate, a pair of electrodes 1 made of Au, an electrode 2 made of Ta, an electrode 3 made of Pt, and an oxide layer made of PrCaMnO _3. 4.

  In the manufacture of the resistance variable element of this example, first, Pt was formed to a thickness of 10 nm on the (100) plane of the MgO single crystal substrate by a sputtering method performed using a sputtering apparatus. In this sputtering, Ar gas (0.5 Pa) was used as the sputtering gas, a Pt target was used, DC discharge was performed, the input power was 0.8 kW, and the temperature condition was 400 ° C.

Next, PrCaMnO ₃ (Pr _0.7 Ca _0.3 MnO ₃ ) was formed to a thickness of 80 nm on the Pt film by sputtering. In this sputtering, a mixed gas of Ar and O ₂ (0.5 Pa, oxygen concentration 20 vol%) is used as a sputtering gas, a Pr _0.7 Ca _0.3 MnO ₃ target is used, an RF discharge is performed, an input power is set to 0.8 kW, and a temperature is set. The condition was 400 ° C.

Next, Ta was deposited to a thickness of 50 nm on the PrCaMnO ₃ film by sputtering. In this sputtering, Ar gas (0.5 Pa) was used as the sputtering gas, a Ta target was used, DC discharge was performed, the input power was 0.8 kW, and the temperature condition was room temperature.

Next, the Ta film was patterned using a predetermined resist pattern as a mask to form an electrode 2 (a square with a side length of 1000 nm) on the PrCaMnO ₃ film.

Next, Au was deposited to a thickness of 50 nm on the PrCaMnO ₃ film by sputtering. In this sputtering, Ar gas (0.5 Pa) was used as the sputtering gas, an Au target was used, DC discharge was performed, the input power was 0.8 kW, and the temperature condition was room temperature.

Next, the Au film was patterned using a predetermined resist pattern as a mask to form a pair of electrodes 1 (a square with a side length of 1000 nm) on the PrCaMnO ₃ film.

Next, the PrCaMnO ₃ film was patterned using a predetermined resist pattern as a mask to form an oxide layer 4 (length or width in the direction D _{2 in} FIG. 2 is 1000 nm).

Next, the Pt film was patterned using a predetermined resist pattern as a mask to form an electrode 3 (the length or width in the direction D _{2 in} FIG. 2 is 1000 nm).

As described above, a plurality of variable resistance elements as examples of the variable resistance element X were manufactured. And about each resistance variable element of a present Example, the resistance value between the electrodes 1 was measured, the initialization process was performed, and the resistance value between the electrodes 1 was again measured after that. In the resistance value measurement before the initialization process, when a DC voltage of 400 mV was applied between the pair of electrodes 1 of each resistance variable element, each element exhibited a resistance value of 1 to 5 × 10 ⁴ Ω. In the initialization process, the electrode 2 was used as the positive electrode and the electrode 3 was used as the negative electrode, and a pulse voltage with an applied intensity of 10 V and an application time of 1 msec was applied between the electrodes 2 and 3. In the resistance value measurement after the initialization process, when a DC voltage of 400 mV between the pair of electrodes 1 was applied to each resistance variable element, each element exhibited a resistance value of 1 to 50 × 10 ⁹ Ω. Since the resistance value between the electrodes 1 has greatly increased in this way, the carrier (electron) movement route existing in the electrode 3 in each element is substantially closed by the initialization process, and the electrode in each element It can be seen that the route of carrier movement between 1 is restricted within the oxide layer 4. In each resistance variable element in which the route of carrier movement between the electrodes 1 is restricted in the oxide layer 4 in this way, as described above with reference to FIGS. It was possible to change the resistance state of the channel (the route of carrier movement in the oxide layer 4) by voltage control.

  As a summary of the above, the configurations of the present invention and variations thereof are listed below as supplementary notes.

(Appendix 1) P-type semiconducting oxide part;
A pair of first electrodes that are spaced apart from each other and bonded to the oxide portion;
A second electrode joined to the oxide portion between the pair of first electrodes;
And a third electrode that has a portion facing the second electrode through the oxide portion and is joined to the oxide portion.
(Supplementary Note 2) The oxide portion is an oxide layer having a first surface and a second surface opposite to the first surface, and the pair of first electrodes and the second electrode includes the oxide layer. The resistance variable element according to appendix 1, wherein the third electrode is bonded to the second surface side of the oxide layer.
(Supplementary note 3) The resistance variable element according to Supplementary note 1 or 2, wherein the second electrode is made of a material that is more easily oxidized than the first electrode and / or the third electrode.
(Supplementary note 4) The resistance variable element according to any one of supplementary notes 1 to 3, wherein the second electrode includes Ta, Ti, Al, Fe, Co, or Ni.
(Supplementary Note 5) The resistance variable element according to any one of Supplementary notes 1 to 4, wherein the first electrode and / or the third electrode includes Pt, Au, or SrRuO ₃ .
(Supplementary note 6) The resistance variable element according to any one of supplementary notes 1 to 5, wherein the third electrode has a thickness of 10 nm or less.
(Appendix 7) The oxide portion is made of a transition metal oxide having P-type semiconductivity, and the transition metal oxide includes any one of transition metals capable of changing the valence. The resistance variable element according to 1.
(Supplementary note 8) The resistance variable element according to any one of supplementary notes 1 to 7, wherein the oxide portion includes PrCaMnO ₃ .
(Supplementary note 9) The resistance variable element according to any one of supplementary notes 1 to 8, wherein the oxide portion has a thickness of 50 to 150 nm.
(Supplementary Note 10) In the second direction orthogonal to the first direction in which the pair of first electrodes and the second electrode are arranged, the length of the first electrode joined to the oxide portion is the second length The resistance variable element according to any one of appendices 1 to 9, wherein the electrode has a length equal to or shorter than a length at which the electrode is bonded to the oxide portion.
(Supplementary note 11) The resistance variable element according to supplementary note 10, wherein a length of the oxide portion in the second direction is 1 μm or less.
(Supplementary note 12) The resistance variable element according to supplementary note 10 or 11, wherein a length of the third electrode in the second direction is 1 μm or less.
(Appendix 13) A step of forming a conductive material film on a substrate;
Forming an oxide film having P-type semiconductivity on the conductive material film;
Forming a pair of first electrodes spaced apart from each other and a second electrode positioned between the first electrodes on the oxide film.
(Supplementary Note 14) A step of forming a pair of first electrodes spaced apart from each other on a substrate and a second electrode positioned between the first electrodes;
Forming an oxide film having P-type semiconductivity so as to cover the pair of first electrodes and the second electrode;
Forming a conductive material film on the oxide film.

It is sectional drawing of the resistance variable element which concerns on this invention. It is a top view of the resistance variable element which concerns on this invention, and is equivalent to the arrow view along line II-II of FIG. The manufacturing method of the resistance variable element shown in FIG. The initialization process of the resistance variable element according to the present invention is shown. The operation principle of the resistance variable element according to the present invention is shown (high resistance). The operation principle of the resistance variable element according to the present invention is expressed (low resistance). The 1st modification of the resistance variable element shown in FIG. 1 is represented. The 2nd modification of the resistance variable element shown in FIG. 1 is represented. It is a top view of the resistance variable element which concerns on a 2nd modification, and is equivalent to the arrow line view along line IX-IX of FIG. The manufacturing method of the resistance variable element shown in FIG. The laminated structure in the sample element of an Example is represented.

Explanation of symbols

X variable resistance element S substrate 1, 2, 3 electrode 4 oxide layer 5 oxygen ion 6 oxygen vacancy 7 electron 11, 12 material film

Claims (6)

An oxide portion having P-type semiconductivity;
A pair of first electrodes that are spaced apart from each other and bonded to the oxide portion;
A second electrode joined to the oxide portion between the pair of first electrodes;
And a third electrode that has a portion facing the second electrode through the oxide portion and is joined to the oxide portion.   The variable resistance element according to claim 1, wherein the second electrode is made of a material that is more easily oxidized than the first electrode and / or the third electrode.   The resistance change element according to claim 1, wherein the oxide portion has a thickness of 50 to 150 nm.   In the second direction orthogonal to the first direction in which the pair of first electrodes and the second electrode are arranged, the length of the first electrode joined to the oxide portion is such that the second electrode is the oxide portion. The resistance change element according to claim 1, wherein the resistance change element has a length equal to or shorter than a length to be bonded to the electrode. Forming a conductive material film on the substrate;
Forming an oxide film having P-type semiconductivity on the conductive material film;
Forming a pair of first electrodes spaced apart from each other and a second electrode positioned between the first electrodes on the oxide film. Forming a pair of first electrodes spaced apart from each other on the substrate and a second electrode positioned between the first electrodes;
Forming an oxide film having P-type semiconductivity so as to cover the pair of first electrodes and the second electrode;
Forming a conductive material film on the oxide film.

Images