storage device

Abstract

A novel storage device is provided. An inspection bit generation section has a function of generating an inspection series from an information series, a storage section has a function of storing a code word constituted by the information series and the inspection series in a flip-flop circuit and outputting the same as a received word, an error detection and correction section has a function of outputting an error vector indicating an error position of the received word and a decoded word in which an error of the received word is corrected, the storage section has a function of storing the decoded word in the flip-flop circuit based on the error vector, a function of writing the decoded word to a holding circuit, and a function of writing the decoded word written to the holding circuit back to the flip-flop circuit, the flip-flop circuit includes a first transistor, the holding circuit includes a second transistor, and an off-state current of the second transistor is smaller than an off-state current of the first transistor.

Description

Technical Field

[0001] One aspect of the present invention relates to a storage device.

[0002] Note that one aspect of the present invention is not limited to the aforementioned technical fields. The technical field of one aspect of the invention disclosed in this specification relates to an object, method, driving method, or manufacturing method. Furthermore, one aspect of the present invention relates to a process, machine, manufacture, or composition of matter. More specifically, as examples of the technical field of one aspect of the present invention disclosed in this specification, examples include semiconductor devices, display devices, light-emitting devices, energy storage devices, optical devices, imaging devices, lighting devices, projection devices, electro-optical devices, light-receiving devices, detection devices, power supply devices, communication devices, information processing devices, computing devices, control devices, storage devices, input devices, output devices, input / output devices, signal processing devices, arithmetic processing devices, electronic computers, electronic devices, systems including them, driving methods for them, or manufacturing methods for them. Background Technology

[0003] It is known that OS transistors (transistors containing oxide semiconductors in the channel formation region) have extremely small off-state currents. For example, Patent Document 1 discloses a low-power computing device (e.g., a CPU) that utilizes the low off-state current characteristic of OS transistors. Specifically, it discloses a technique for power gating a flip-flop by assembling a holding circuit including an OS transistor and a capacitor into a flip-flop on an integrated circuit.

[0004] [Preliminary Technology Documents] [Patent Literature] [Patent Document 1] Japanese Patent Application Publication No. 2016-82593 [Patent Document 2] Japanese Patent Application Publication No. 2017-135698 [Patent Document 3] International Patent Application Publication No. 2023 / 180849. Summary of the Invention

[0005] The technical problem that the invention aims to solve In computing devices, a defect known as a soft error sometimes occurs, which is a failure caused by the unintentional reversal of a portion of the data held by triggers or similar devices. Soft errors are particularly prone to occur in environments with high radiation, such as outer space, leading to a decrease in reliability.

[0006] Here, it is known that OS transistors exhibit minimal changes in electrical characteristics due to irradiation. For example, Patent Document 2 discloses the following evaluation results: when irradiated with radiation, both a memory circuit using OS transistors and an inverter loop using Si transistors (transistors containing silicon in the channel formation region), the memory circuit using OS transistors is less prone to soft errors compared to the inverter loop using Si transistors. Furthermore, for example, Patent Document 3 discloses the following evaluation results: when OS transistors are irradiated with radiation, the changes in subthreshold slope are small, the changes in field-effect mobility are small, and the off-state current remains extremely small.

[0007] Therefore, for example, in a structure disclosed in Patent Document 1, where a holding circuit using an OS transistor is assembled into a flip-flop using a Si transistor, the reliability can be improved by holding the data through this holding circuit. However, in this structure, the data held by the flip-flop may be unintentionally reversed before the holding circuit holds the data. In this case, the holding circuit holds the unintentionally reversed data, thus causing a soft error.

[0008] One objective of this invention is to provide a highly reliable storage device. Another objective is to provide a storage device that achieves miniaturization. Another objective is to provide a storage device with reduced power consumption. Another objective is to provide a storage device with increased operating speed. Finally, one objective is to provide a novel storage device.

[0009] Furthermore, one objective of the present invention is to provide a semiconductor device capable of being used in the aforementioned memory device. Another objective of the present invention is to provide a computational processing apparatus using the aforementioned memory device.

[0010] Note that the description of the above objectives does not preclude the existence of other objectives. Those skilled in the art can naturally derive and extract other objectives from the description in this specification, drawings, claims, etc. Note that one embodiment of the invention does not necessarily need to achieve all objectives (the above-described objectives and others).

[0011] means of solving technical problems (1) One aspect of the present invention is a storage device, comprising: a check bit generation unit; a storage unit; and an error detection and correction unit, wherein the storage unit includes a trigger circuit and a holding circuit, the check bit generation unit has the function of generating a check sequence from an information sequence, the storage unit has the function of storing a codeword composed of the information sequence and the check sequence in the trigger circuit and outputting it as a received word, the error detection and correction unit has the function of outputting an error vector indicating the error position of the received word and a decoded word indicating that the error of the received word has been corrected, and the storage unit has the function of storing the decoded word in the trigger circuit according to the error vector, writing the decoded word to the holding circuit, and writing the decoded word written to the holding circuit back to the trigger circuit, the trigger circuit including a first transistor, the holding circuit including a second transistor, and the off-state current of the second transistor being less than the off-state current of the first transistor. (2) Furthermore, in (1) above, the codeword can be Hamming code. (3) Furthermore, in (1) or (2) above, the change in electrical characteristics of the second transistor due to the irradiated radiation can be smaller than the change in electrical characteristics of the first transistor due to the irradiated radiation. (4) Furthermore, in (1) or (2) above, the second transistor may be stacked on the layer on which the first transistor is disposed. (5) Furthermore, in (1) or (2) above, the first transistor may contain silicon in the channel forming region, and the second transistor may contain oxide semiconductor in the channel forming region.

[0016] Invention Effects According to one aspect of the present invention, a highly reliable storage device can be provided. Furthermore, according to one aspect of the present invention, a storage device that achieves miniaturization can be provided. Furthermore, according to one aspect of the present invention, a storage device with reduced power consumption can be provided. Furthermore, according to one aspect of the present invention, a storage device capable of improving operating speed can be provided. Furthermore, according to one aspect of the present invention, a novel storage device can be provided.

[0017] Furthermore, according to one aspect of the present invention, a semiconductor device capable of being used in the aforementioned memory device can be provided. Additionally, according to another aspect of the present invention, an arithmetic processing apparatus using the aforementioned memory device can be provided.

[0018] Furthermore, the description of the above effects does not preclude the existence of other effects. Those skilled in the art can naturally derive and extract other effects from the description in this specification, drawings, claims, etc. Note that one embodiment of the invention does not necessarily need to have all the effects (the above-described effects and others).

[0019] Brief description of the attached figures Figure 1 This is a circuit diagram illustrating an example of the structure of a storage device.

[0020] Figure 2A and Figure 2B This is a circuit diagram illustrating an example of the structure of a storage device.

[0021] Figure 3 This is a timing diagram illustrating an example of how a storage device works.

[0022] Figure 4 This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0023] Figure 5 This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0024] Figure 6A This is a top view showing an example of the structure of a semiconductor device. Figure 6B and Figure 6C This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0025] Figure 7 This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0026] Figure 8 This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0027] Figure 9A This is a top view showing an example of the structure of a semiconductor device. Figures 9B to 9D This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0028] Figure 10 This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0029] Figure 11A This is a top view showing an example of the structure of a semiconductor device. Figure 11B This is a three-dimensional schematic diagram showing an example of the structure of a semiconductor device. Figure 11C and Figure 11D This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0030] Figure 12A and Figure 12B This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0031] Figure 13 This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0032] Figures 14A to 14C This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0033] Figure 15A and Figure 15B This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0034] Figure 16A This is a top view showing an example of the structure of a semiconductor device. Figure 16B This is a three-dimensional schematic diagram showing an example of the structure of a semiconductor device. Figures 16C to 16E This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0035] Figure 17A and Figure 17B This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0036] Figure 18A This is a top view showing an example of the structure of a semiconductor device. Figure 18B This is a three-dimensional schematic diagram showing an example of the structure of a semiconductor device. Figures 18C to 18E This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0037] Figure 19 This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0038] Figure 20A This is a top view showing an example of the structure of a semiconductor device. Figure 20B This is a three-dimensional schematic diagram showing an example of the structure of a semiconductor device. Figures 20C to 20E This is a cross-sectional view showing an example of the structure of a semiconductor device.

[0039] Figure 21 This is a diagram that shows various storage devices in a hierarchical manner.

[0040] Figure 22 This is a schematic diagram illustrating an example of the structure of a storage device.

[0041] Figures 23A to 23H This is a circuit diagram illustrating an example of the structure of a memory cell.

[0042] Figure 24A and Figure 24B This is a schematic diagram illustrating an example of the structure of a display device.

[0043] Figure 25 This is a cross-sectional view illustrating an example of the structure of a display device.

[0044] Figure 26This is a cross-sectional view illustrating an example of the structure of a display device.

[0045] Figure 27A and Figure 27B This is a circuit diagram illustrating an example of the structure of a semiconductor device.

[0046] Figure 28A and Figure 28B This is a diagram showing an example of an electronic component.

[0047] Figure 29A and Figure 29B This is a diagram illustrating an example of an electronic device. Figures 29C to 29E This is a diagram illustrating an example of a large computer.

[0048] Figure 30A This is a diagram illustrating an example of a space device. Figure 30B This is a diagram illustrating an example of a storage system that can be used in a data center. Figure 30C and Figure 30D This is a diagram illustrating an example of an electronic device.

[0049] Figures 31A1 to 31A7 and Figures 31B1 to 31B6 It is a diagram illustrating electrical connections.

[0050] Methods of implementing the invention In this specification and the like, a semiconductor device refers to a device that utilizes the properties of a semiconductor, such as a circuit that includes semiconductor elements (e.g., transistors or diodes) or a device that includes such a circuit. Furthermore, a semiconductor device refers to any device capable of functioning by utilizing the properties of a semiconductor. Examples of semiconductor devices include electronic circuits that include semiconductor elements, chips that include electronic circuits, electronic components that store chips in packages, or electronic devices that mount electronic components. Additionally, devices such as display devices, light-emitting devices, energy storage devices, optical devices, imaging devices, lighting devices, projection devices, electro-optical devices, light-receiving devices, detection devices, power supply devices, communication devices, information processing devices, arithmetic devices, control devices, storage devices, input devices, output devices, input / output devices, signal processing devices, arithmetic processing devices, electronic computers, or electronic devices are themselves semiconductor devices, and sometimes include semiconductor devices.

[0051] The embodiments are described with reference to the accompanying drawings. Note that the embodiments can be implemented in many different ways. Therefore, those skilled in the art will readily understand that the embodiments and their details can be varied in many forms without departing from the spirit and scope of the invention. Therefore, the invention should not be construed as being limited to the contents described in the following embodiments.

[0052] Furthermore, the structures shown in each embodiment in this specification and the like can be appropriately combined with the structures shown in other embodiments to constitute a mode of the present invention. Additionally, when multiple structures are shown in one embodiment, these structures can be appropriately combined to constitute a mode of the present invention.

[0053] Note that, regarding the accompanying drawings illustrating the embodiments, in the structure of the invention, the same symbols are sometimes used in different drawings to represent the same parts or parts having the same function, thereby omitting repeated descriptions. Furthermore, in the drawings, when parts having the same function are represented, the same shading lines are sometimes used, for example, without specifically adding additional symbols. For example, in perspective views or top views (also called "plan views"), for clarity, illustrations of some constituent elements are sometimes omitted. For example, descriptions of some hidden lines in the drawings are sometimes omitted. Furthermore, for example, descriptions of shading lines, etc., in the drawings are sometimes omitted.

[0054] In the accompanying drawings, dimensions, layer thicknesses, or areas are sometimes exaggerated for clarity. Therefore, the drawings are not limited to the dimensions or aspect ratios shown. Furthermore, the drawings are schematic for ease of understanding and are not limited to the shapes or values ​​shown. For example, in actual manufacturing processes, layers or resist masks are sometimes unintentionally thinned due to processes such as etching, but this is sometimes not reflected in the drawings for ease of understanding. Similarly, in actual circuit operation, voltage or current non-uniformity is sometimes caused by noise or timing deviations, but these are sometimes not reflected in the drawings for ease of understanding.

[0055] In this specification and accompanying drawings, the constituent elements are categorized according to function and represented as independent constituent elements. However, it is difficult to categorize constituent elements according to function, as sometimes a constituent element involves multiple functions or multiple constituent elements involve a single function. Therefore, the constituent elements shown in this specification and accompanying drawings are not limited to their descriptions, and appropriate wording may be used.

[0056] In this specification and accompanying drawings, when multiple constituent elements use the same symbol and it is necessary to distinguish them, sometimes an identification symbol such as "A", "b", "_1", "[n]", or "[m,n]" is added to the symbol. Furthermore, when describing the common content among multiple constituent elements for which an identification symbol is added, or when it is not necessary to distinguish them, sometimes the identification symbol is not added and the description is presented.

[0057] Note that in this specification, the term "on state" or "turn-on state" of a transistor refers, for example, to a state in which the source and drain of the transistor are electrically short-circuited, or to a state in which current can flow between the source and drain (also called a current-flowable state). For example, the following states are sometimes referred to as "on state" or "turn-on state": in an n-channel transistor, the voltage between the gate and source is higher than the threshold voltage; or in a p-channel transistor, the voltage between the gate and source is lower than the threshold voltage, etc. Furthermore, the term "off state," "disconnected state," or "turn-off state" of a transistor refers to a state in which the source and drain of the transistor are electrically disconnected. For example, the following states are sometimes referred to as "off state," "disconnected state," or "turn-off state": in an n-channel transistor, the voltage between the gate and source is lower than the threshold voltage; or in a p-channel transistor, the voltage between the gate and source is higher than the threshold voltage, etc.

[0058] Furthermore, in this specification, the voltage between the gate and source (gate-source, unless otherwise specified, based on the source) is sometimes referred to as "gate voltage," the voltage between the drain and source (drain-source, unless otherwise specified, based on the source) is sometimes referred to as "drain voltage," and the voltage between the back gate and source (back gate-source, unless otherwise specified, based on the source) is sometimes referred to as "back gate voltage." Additionally, the current flowing between the drain and source (unless otherwise specified, the direction from the drain to the source is considered positive) is sometimes referred to as "drain current." Note that descriptions of high gate voltage, high drain voltage, and high back gate voltage for n-channel transistors can be appropriately converted to descriptions of low gate voltage, low drain voltage, and low back gate voltage for p-channel transistors. Furthermore, descriptions of low gate voltage, low drain voltage, and low back gate voltage for n-channel transistors can be appropriately converted to descriptions of high gate voltage, high drain voltage, and high back gate voltage for p-channel transistors.

[0059] Furthermore, in this specification and other documents, unless otherwise specified, the "off-state current" of a transistor refers to the drain current when the transistor is in the off state. Note that in this specification and other documents, the off-state current and the current flowing between the gate and the source and drain (also called gate leakage current) are sometimes referred to as leakage current.

[0060] Furthermore, in this specification, etc., one of the source and drain of a transistor (also referred to as the two input / output terminals) is sometimes called the first terminal, and the other of the source and drain of a transistor is called the second terminal. That is, a transistor includes at least a gate (also called the gate terminal), a first terminal, and a second terminal. Additionally, sometimes one terminal of a capacitor (also referred to as one of a pair of terminals) is called the first terminal, and the other terminal of the capacitor (also referred to as the other of a pair of terminals) is called the second terminal. Additionally, sometimes one terminal of a display element is called the first terminal, and the other terminal of the display element is called the second terminal. Additionally, sometimes one terminal of a liquid crystal element is called the first terminal, and the other terminal of the liquid crystal element is called the second terminal. Additionally, sometimes one terminal of a light-emitting element is called the first terminal, and the other terminal of the light-emitting element is called the second terminal. Additionally, sometimes one terminal of a light-receiving element is called the first terminal, and the other terminal of the light-receiving element is called the second terminal. Additionally, sometimes one of the anode and cathode of a diode (also referred to as one of a pair of terminals) is called the first terminal, and the other of the anode and cathode of a diode (also referred to as the other of a pair of terminals) is called the second terminal.

[0061] (Implementation Method 1) A storage device according to one aspect of the present invention is described with reference to the accompanying drawings. At least a portion of the storage device according to one aspect of the present invention can be used, for example, in an arithmetic processing apparatus. In particular, it can be used in registers included in an arithmetic processing apparatus.

[0062] One aspect of the present invention provides a storage device that has the function of storing a series of information (also called information bits) of k bits (where k is an integer greater than or equal to 1). For example, the storage device may function as a register with a width of k bits.

[0063] In addition to the k-bit information sequence, this storage device also has the function of storing an m-bit check sequence (also called a check bit) where m is an integer greater than or equal to 1. That is, this storage device has the function of storing an n-bit (n=k+m) codeword (also called a code position) composed of the k-bit information sequence and the m-bit check sequence. Therefore, this storage device can have the function of detecting errors in the stored codewords. Furthermore, it can also have the function of correcting erroneous codewords.

[0064] A structural example of a storage device with this function will be described later.

[0065] As a codeword, error correction codes can be used. For example, Hamming codes can be used as error correction codes. By using Hamming codes, an error in any bit of the codeword can be detected and corrected. Thus, one-bit error correction can be performed using Hamming codes.

[0066] In an n-bit Hamming codeword, there are n+1 combinations (i.e., k+m+1 combinations) of no error and any error occurring in any bit. Furthermore, the information content of the m-bit check sequence is 2 to the power of m. Therefore, the number of bits k in the information sequence and the number of bits m in the check sequence must satisfy the following equation (1).

[0067] [Equation 1] For example, when the information sequence is 4 bits (k=4), the check sequence is 3 bits (m=3), and the codeword is 7 bits (n=7). Furthermore, for example, when the information sequence is 8 bits (k=8), the check sequence is 4 bits (m=4), and the codeword is 12 bits (n=12). Furthermore, for example, when the information sequence is 16 bits (k=16), the check sequence is 5 bits (m=5), and the codeword is 21 bits (n=21). Furthermore, for example, when the information sequence is 32 bits (k=32), the check sequence is 6 bits (m=6), and the codeword is 38 bits (n=38). Furthermore, for example, when the information sequence is 64 bits (k=64), the check sequence is 7 bits (m=7), and the codeword is 71 bits (n=71).

[0068] Note that extended Hamming codes, which are Hamming codes with one additional check bit, can also be used as error correction codes. By using extended Hamming codes, one-bit error correction (detecting an error in any bit of the codeword and correcting the erroneous bit) and two-bit error detection (detecting an error in any two bits of the codeword) can be performed.

[0069] In addition, Bose-Chaudhuri-Hocquenghem (BCH) codes and Reed-Solomon (RS) codes can also be used as error correction codes. By using these codes, errors of two or more bits can be corrected in addition to one-bit errors.

[0070] Thus, when a storage device according to one aspect of the present invention has the function of storing codewords using error correction codes, a storage device that is less prone to soft errors can be realized. Therefore, the reliability of the storage device and the arithmetic processing device using the storage device can be improved.

[0071] Here, as a technique for implementing a storage device less prone to soft errors, Triple Modular Redundancy (TMR) is cited as an example. In a storage device utilizing TMR, three storage circuits store the same information and output their majority vote results. Thus, even if an error occurs in one of the three storage circuits, a correct result can still be obtained through majority voting. Note that a correct result cannot be obtained if errors occur in two or more of the three storage circuits. To achieve this effect in a storage device utilizing TMR, a storage circuit with a bit depth equivalent to three times the number of bits of information is required.

[0072] On the other hand, in a storage device according to one aspect of the present invention, as described above, in addition to the storage circuitry corresponding to a k-bit information sequence, a storage circuitry corresponding to an m-bit check sequence is also required. Therefore, for example, the storage circuitry required to store a 4-bit information sequence is 12 bits in the TMR, while it is 7 bits in a storage device according to one aspect of the present invention. Furthermore, for example, the storage circuitry required to store an 8-bit information sequence is 24 bits in the TMR, while it is 12 bits in a storage device according to one aspect of the present invention. Furthermore, for example, the storage circuitry required to store a 16-bit information sequence is 48 bits in the TMR, while it is 21 bits in a storage device according to one aspect of the present invention. Furthermore, for example, the storage circuitry required to store a 32-bit information sequence is 96 bits in the TMR, while it is 38 bits in a storage device according to one aspect of the present invention. Furthermore, for example, the storage circuitry required to store a 64-bit information sequence is 192 bits in the TMR, while it is 71 bits in a storage device according to one aspect of the present invention. Thus, the storage device according to one aspect of the present invention requires fewer bits than when using a TMR, thereby reducing the circuit size. This enables the miniaturization and low power consumption of storage devices and computing devices that use them.

[0073] <Example of storage device structure> Next, an example of the structure of a storage device having the above-mentioned functions will be described.

[0074] Figure 1 This is a circuit diagram illustrating an example structure of a storage device 100 according to one aspect of the present invention. (See diagram for example.) Figure 1 As shown, the storage device 100 includes a storage unit 101, a check bit generation unit 102, and an error detection and correction unit 103. Note that although not shown, the storage device 100 may also include a control unit that controls the operation of each of the storage unit 101, the check bit generation unit 102, and the error detection and correction unit 103.

[0075] The check bit generation unit 102 has the function of generating an m-bit check sequence from a k-bit information sequence. For example, the m-bit check sequence can be generated in such a way that an n-bit codeword composed of a k-bit information sequence and an m-bit check sequence becomes an error correction code such as a Hamming code.

[0076] The storage unit 101 has the function of storing and holding an n-bit codeword and outputting it as an n-bit receive word. Furthermore, the storage unit 101 has the function of storing and holding an n-bit decoded word corrected by the error detection and correction unit 103 based on an n-bit error vector (also called an error pattern) output from the error detection and correction unit 103 and outputting it as an n-bit receive word. Additionally, the storage unit 101 may also have the function of backing up an n-bit decoded word.

[0077] The error detection and correction unit 103 has the following functions: detecting an error in at least one bit of an n-bit received word, correcting the erroneous bit, and outputting it as an n-bit decoded word. Furthermore, it has the function of outputting an n-bit error vector, indicating which bit of the n-bit received word is corrected. For example, when using Hamming code as an error correction code, it can detect an error in any bit of the n-bit received word and correct any bit that has erroneously occurred.

[0078] Wiring Xc[1] to Xc[k], which has the function of transmitting a k-bit information sequence, is connected to the storage unit 101 and the check bit generation unit 102. Furthermore, wiring Pc[1] to Pc[m], which has the function of transmitting an m-bit check sequence, is connected to the check bit generation unit 102 and the storage unit 101. Furthermore, wiring Yc[1] to Yc[n], which has the function of transmitting an n-bit receive word, is connected to the storage unit 101 and the error detection and correction unit 103. Furthermore, wiring Wc[1] to Wc[n], which has the function of transmitting an n-bit decoded word, is connected to the error detection and correction unit 103 and the storage unit 101. Furthermore, wiring Ec[1] to Ec[n], which has the function of transmitting an n-bit error vector, is connected to the error detection and correction unit 103 and the storage unit 101.

[0079] Therefore, a k-bit information sequence supplied from outside the storage device 100 is input to the storage unit 101 and the check bit generation unit 102 via wiring Xc[1] to Xc[k]. Furthermore, an m-bit check sequence output from the check bit generation unit 102 is input to the storage unit 101 via wiring Pc[1] to Pc[m]. In other words, an n-bit codeword composed of the k-bit information sequence and the m-bit check sequence is input to the storage unit 101. Furthermore, an n-bit received word output from the storage unit 101 is input to the error detection and correction unit 103 via wiring Yc[1] to Yc[n]. Furthermore, an n-bit decoded word output from the error detection and correction unit 103 is input to the storage unit 101 via wiring Wc[1] to Wc[n]. Furthermore, an n-bit error vector output from the error detection and correction unit 103 is input to the storage unit 101 via wiring Ec[1] to Ec[n]. Note that the k-bit decoding sequence corresponding to the k-bit information sequence in the n-bit decoding word can also be output to the outside of the storage device 100 through wiring Wc[1] to wiring Wc[k].

[0080] Note that in this embodiment, the codewords transmitted via wiring Xc[1] to Xc[k] and wiring Pc[1] to Pc[m] are sometimes referred to as encoded data. Furthermore, the received word transmitted via wiring Yc[1] to Yc[n] is sometimes referred to as hold data. Furthermore, the decoded word transmitted via wiring Wc[1] to Wc[n] is sometimes referred to as decoded data. The error vector transmitted via wiring Ec[1] to Ec[n] is sometimes referred to as an error signal.

[0081] Storage unit 101 includes n unit storage circuits 110 (unit storage circuit 110[1] to unit storage circuit 110[n]).

[0082] Each of the n unit storage circuits 110 includes wiring D1 connected to wiring Xc[1] to Xc[k] and wiring Pc[1] to Pc[m]. Each of the n unit storage circuits 110 includes wiring Q connected to wiring Yc[1] to Yc[n]. Each of the n unit storage circuits 110 includes wiring D2 connected to wiring Wc[1] to Wc[n]. Each of the n unit storage circuits 110 includes wiring Ec connected to wiring Ec[1] to Ec[n].

[0083] Therefore, wiring D1 is supplied with encoded data, wiring Q is supplied with held data, wiring D2 is supplied with decoded data, and wiring Ec is supplied with error signals.

[0084] The unit storage circuit 110 includes a scan trigger circuit 120 and a backup circuit 130.

[0085] Backup circuit 130 is connected to scan trigger circuit 120 via wiring D2. Additionally, backup circuit 130 is connected to scan trigger circuit 120 via wiring SD.

[0086] The unit storage circuit 110 has the function of storing 1 bit of information. Specifically, it has the function of storing and holding encoded data in the scan trigger circuit 120 as a hold data output. In addition, it has the function of storing and holding decoded data according to an error signal as a hold data output.

[0087] Furthermore, the unit storage circuit 110 may also have the function of storing 1 bit of information even when the power is off (power supply is stopped). Specifically, it may also have the function of writing the decoded data to the backup circuit 130 for retention and writing the retained data back to the scan trigger circuit 120.

[0088] The scan trigger circuit 120 includes at least a trigger circuit (corresponding to the trigger circuit 122 described later). For example, a Si transistor is used as the transistor constituting the scan trigger circuit 120.

[0089] The backup circuit 130 includes at least a holding circuit (corresponding to the holding circuit 131 described later), which includes at least a switching element sw and a storage element me. The switching element sw has the function of controlling the writing of decoded data. The storage element me has the function of holding the written decoded data. As the switching element sw, various transistors can be used. For example, a MOS field-effect transistor, a junction field-effect transistor, or a bipolar transistor can be used. As the storage element me, for example, a capacitor can be used. In addition, for example, storage elements utilizing ferroelectric materials, storage elements utilizing magnetoresistive effects, storage elements utilizing phase-change recording technology, or storage elements utilizing electric field-induced resistance changes can be used.

[0090] like Figure 1 As shown, for example, backup circuit 130 may include at least transistor M31 and capacitor C31.

[0091] One of the source and drain terminals of transistor M31 is connected to one terminal of capacitor C31. The other of the source and drain terminals of transistor M31 is connected to wiring D2. The other terminal of capacitor C31 is connected to wiring supplied with an arbitrary potential.

[0092] Furthermore, a signal controlling the conduction state of transistor M31 is supplied from a control unit (not shown) to the gate of transistor M31. Based on the signal supplied to the gate of transistor M31, the potential of the decoded data supplied by wiring D2 is stored and maintained at one terminal of capacitor C31. Additionally, the potential based on this maintained potential is supplied to wiring SD.

[0093] A specific structural example of the unit storage circuit 110 will be described later.

[0094] In one embodiment of the invention, the transistor constituting the backup circuit 130 is preferably a transistor whose off-state current is smaller than that of the transistor constituting the scan trigger circuit 120. This allows for a longer retention time of data written to the backup circuit 130, enabling, for example, power gating of the storage device 100. Consequently, power consumption of the storage device and the processing unit using the storage device can be reduced.

[0095] Furthermore, the transistors constituting the backup circuit 130 are preferably transistors with higher resistance to radiation than the transistors constituting the scan trigger circuit 120. For example, transistors whose electrical characteristics change less due to radiation exposure than those constituting the scan trigger circuit 120 are preferably used. For example, transistors whose off-state current changes less due to radiation exposure than those constituting the scan trigger circuit 120 are preferably used. For example, transistors whose subthreshold slope changes less due to radiation exposure than those constituting the scan trigger circuit 120 are preferably used. For example, transistors whose field-effect mobility changes less due to radiation exposure than those constituting the scan trigger circuit 120 are preferably used. For example, transistors whose threshold voltage changes less due to radiation exposure than those constituting the scan trigger circuit 120 are preferably used. Thus, the data held in the backup circuit 130 can be suppressed from being intentionally reversed due to the influence of radiation. Therefore, the reliability of the storage device and the processing device using the storage device can be improved.

[0096] Furthermore, the transistors constituting the backup circuit 130 are preferably transistors that can be stacked on the same layer as the transistors constituting the scan trigger circuit 120. This reduces the area overhead associated with the backup circuit 130. Consequently, miniaturization of the storage device and the processing unit using the storage device is possible. Furthermore, the wiring distance connecting the backup circuit 130 and the scan trigger circuit 120 can be shortened, thus reducing the energy required for data delivery (access energy). Therefore, power consumption of the storage device and the processing unit using the storage device can be reduced.

[0097] For example, an OS transistor can be used as the transistor constituting the backup circuit 130.

[0098] Because the band gap of the oxide semiconductor forming the channel is above 2 eV, the OS transistor has a very small off-state current. At room temperature, the off-state current of an OS transistor with a channel width of 1 μm can be 1 aA (1 × 10⁻⁶). -18 A) Below, 1zA (1×10-21 A) or below or 1yA (1×10 -24 A) Below. Note that in a Si transistor (a transistor containing silicon in the channel formation region), the off-state current per 1 μm channel width at room temperature is 1 fA (1 × 10⁻⁶). -15 A) and above and 1pA (1×10) -12 A) Below. Therefore, it can also be said that the off-state current of an OS transistor is about 10 bits smaller than that of a Si transistor. Therefore, for example, when a wiring connected to one of the source and drain terminals of an OS transistor is in a floating state, the charge stored in that wiring can be retained for a long time. Therefore, for example, by using OS transistors to construct memory cells, data written to those memory cells can be stored for a long time.

[0099] Furthermore, even at high temperatures, the off-state current of the OS transistor hardly increases. Specifically, even in environments above room temperature but below 200°C, the off-state current hardly increases. Moreover, even at high temperatures, the on-state current of the OS transistor does not easily decrease. On the other hand, the on-state current of the Si transistor decreases at high temperatures. In other words, at high temperatures, the on-state current of the OS transistor is larger than that of the Si transistor. Furthermore, even in environments above 125°C but below 150°C, the ratio of on-state current to off-state current of the OS transistor is large, thus enabling good switching operation. Therefore, semiconductor devices using OS transistors operate stably and reliably even at high temperatures.

[0100] Furthermore, OS transistors have a high breakdown voltage between their source and drain (also known as drain breakdown voltage). Therefore, semiconductor devices using OS transistors operate stably and reliably even when driven at high voltages.

[0101] Furthermore, the electrical characteristics of the OS transistor change little due to radiation exposure. For example, as shown in Patent Document 3, the subthreshold slope of the OS transistor changes little due to radiation exposure. Furthermore, the field-effect mobility changes little due to radiation exposure. Moreover, as shown in Patent Document 3, the OS transistor can maintain a very small off-state current even after radiation exposure. For example, the off-state current of an OS transistor with a channel width of 1 μm after radiation exposure can be 1 fA (1 × 10⁻⁶). -15 A) Below, 1aA (1×10 -18 A) Below, 1zA (1×10 -21 A) or below or 1yA (1×10 -24 A) The following. Therefore, semiconductor devices using OS transistors can operate stably and reliably even in environments with strong radiation, such as outer space.

[0102] Furthermore, OS transistors can be freely configured on silicon substrates on which Si transistors are disposed, thus facilitating integration. Additionally, OS transistors can be manufactured using the same fabrication equipment as Si transistors, enabling low-cost manufacturing.

[0103] Therefore, by using OS transistors as transistors constituting backup circuit 130, it is possible to achieve power saving, high reliability, miniaturization, integration, and low cost in storage devices and computing devices using storage devices.

[0104] In one aspect of the present invention, transistors of various structures can be used as the transistors constituting the memory device 100. For example, transistors with various structures such as top-gate (e.g., planar and interleaved), bottom-gate (e.g., anti-planar and anti-interleaved), dual-gate (a structure in which gates are arranged on both sides (e.g., top and bottom) sandwiching the channel forming region), FIN type, TRI-GATE type, and GAA type can be used. In addition, for example, vertical transistors (transistors in which the channel length direction has a longitudinal component (also called the height direction or the direction perpendicular to the formed surface) can be used.

[0105] An example of the structure of a transistor that can be used in the storage device 100 will be described in Embodiment 2, which will be explained later.

[0106] Note that, for example, when the storage device 100 is used in outer space, radiation randomly reaches each transistor constituting the storage device 100. Therefore, for example, by making the distance between the unit storage circuits 110 constituting the storage device 100 greater than the range affected by radiation, the possibility of simultaneous errors occurring in two or more unit storage circuits 110 can be reduced. Thus, for example, in the storage device 100, it is preferable to arrange the unit storage circuits 110 with a distance of 1 μm or more, 2 μm or more, 3 μm or more, 5 μm or more, or 10 μm or more. In this case, the unit storage circuits 110 can also be arranged randomly. With such an arrangement, it is less likely that two or more unit storage circuits 110 will experience simultaneous errors, so, as an error correction for the storage device 100, for example, a Hamming code capable of 1-bit error correction can be used.

[0107] [Work Example] Next, the error correction in the storage device 100 will be explained using the (7,4) Hamming code as an example. In the (7,4) Hamming code, the codeword is 7 bits (n=7) and the information sequence is 4 bits (k=4). In addition, the check sequence is 3 bits (m=3).

[0108] In the storage device 100 using (7,4) Hamming code, the storage unit 101 includes 7 unit storage circuits 110. Therefore, the storage unit 101 stores and holds a 7-bit codeword (x1, x2, x3, x4, p1, p2, p3) with 3 bits of check sequence (p1, p2, p3) added to a 4-bit information sequence (x1, x2, x3, x4), and outputs it as a 7-bit receive word (y1, y2, y3, y4, y5, y6, y7).

[0109] The check bit generation unit 102 can generate a check series (p1, p2, p3) from the information series (x1, x2, x3, x4).

[0110] At this point, in the check bit generation unit 102, for example, the check row and column (p1, p2, p3) can be calculated by the following operation. Note that "mod 2" means the remainder when divided by 2.

[0111] p1 = (x1 + x2 + x3) mod 2 p2 = (x2 + x3 + x4) mod 2 p3 = (x1 + x2 + x4) mod 2 Even if any bit in the 7-bit received word (y1, y2, y3, y4, y5, y6, y7) is erroneous, the error detection and correction unit 103 can output a 7-bit decoded word (w1, w2, w3, w4, w5, w6, w7) in which the bit that was erroneous is corrected, and a 7-bit error vector (e1, e2, e3, e4, e5, e6, e7) indicating which bit was corrected.

[0112] At this time, in the error detection and correction unit 103, the error position is detected from the received word (y1, y2, y3, y4, y5, y6, y7), and the detected error is corrected by calculation.

[0113] First, for example, the 3-bit corrector (s1, s2, s3) can be obtained through the following operation.

[0114] s1 = (y1 + y2 + y3 + y5) mod 2 s² = (y² + y³ + y⁴ + y⁶) mod 2 s3 = (y1 + y2 + y4 + y7) mod 2 Next, for example, the error vectors (e1, e2, e3, e4, e5, e6, e7) can be obtained through the following correspondence. Note that "S(s1, s2, s3)" represents the corrector (s1, s2, s3), and "E(e1, e2, e3, e4, e5, e6, e7)" represents the error vectors (e1, e2, e3, e4, e5, e6, e7).

[0115] S(1,0,1)=E(1,0,0,0,0,0,0) S(1,1,1)=E(0,1,0,0,0,0,0) S(1,1,0)=E(0,0,1,0,0,0,0) S(0,1,1)=E(0,0,0,1,0,0,0) S(1,0,0)=E(0,0,0,0,1,0,0) S(0,1,0)=E(0,0,0,0,0,1,0) S(0,0,1)=E(0,0,0,0,0,0,1) S(0,0,0)=E(0,0,0,0,0,0,0) At this point, the bits that are "1" in the error vector (e1, e2, e3, e4, e5, e6, e7) indicate that an error has occurred in the corresponding bit in the received word (y1, y2, y3, y4, y5, y6, y7).

[0116] Furthermore, the decoded words (w1, w2, w3, w4, w5, w6, w7) can be obtained through the following operations.

[0117] w1 = (y1 + e1) mod 2 w2 = (y2 + e2) mod 2 w3 = (y3 + e3) mod 2 w4 = (y4 + e4) mod 2 w5 = (y5 + e5) mod 2 w6 = (y6 + e6) mod 2 ( ) If there is a 1-bit error in the received word (y1, y2, y3, y4, y5, y6, y7), the decoded word (w1, w2, w3, w4, w5, w6, w7) obtained by the above operation is equivalent to the codeword (x1, x2, x3, x4, p1, p2, p3). Therefore, the same information as the information series (x1, x2, x3, x4) can be obtained as the decoded sequence (w1, w2, w3, w4).

[0118] The storage unit 101 can correct errors in the received word (y1, y2, y3, y4, y5, y6, y7) using the decoded words (w1, w2, w3, w4, w5, w6, w7) and the error vector (e1, e2, e3, e4, e5, e6, e7).

[0119] At this time, in the storage unit 101, the decoded words (w1, w2, w3, w4, w5, w6, w7) corresponding to the bits of the error vector (e1, e2, e3, e4, e5, e6, e7) that are "1" can be stored and held in the unit storage circuit 110 corresponding to the bits of the error vector that are "1". As a result, the error of the received word (y1, y2, y3, y4, y5, y6, y7) can be corrected.

[0120] Here, error correction in storage device 100 is illustrated using the case of a supplied information series (0, 1, 0, 1) as an example.

[0121] In the check bit generation unit 102, a check row (1,0,0) is generated from the information series (0,1,0,1). Therefore, the codeword (0,1,0,1,1,0,0) is stored and maintained in the storage unit 101.

[0122] For example, if no error occurs in the received word, the error detection and correction unit 103 can derive the decoded word (0, 1, 0, 1, 1, 0, 0) and the error vector (0, 0, 0, 0, 0, 0, 0) from the received word (0, 1, 0, 1, 1, 0, 0). In this case, e1 to e7 in the error vector are all "0", indicating that no error occurred in the received word. Furthermore, the decoded word is equivalent to the codeword and the received word.

[0123] On the other hand, for example, in the case where an error occurs where the received word y3 is inverted from "0" to "1", the error detection and correction unit 103 can calculate the decoded word (0, 1, 0, 1, 1, 0, 0) and the error vector (0, 0, 1, 0, 0, 0, 0) from the received word (0, 1, 1, 1, 1, 0, 0). In this case, the error vector e3 is "1", indicating that an error has occurred in the received word y3. Furthermore, the decoded word is equivalent to the codeword. That is, the decoded word w3 is the same as the codeword x3, which is "0".

[0124] In storage unit 101, the "0" of the decoded word w3 is stored and held in unit storage circuit 110 corresponding to error vector e3. As a result, the received word y3 is corrected from "1" to "0".

[0125] <Example of a unit storage circuit structure> This section describes a specific structural example of the unit storage circuit 110.

[0126] Figure 2A and Figure 2B This is a circuit diagram illustrating an example of the structure of the unit storage circuit 110.

[0127] like Figure 2A As shown, the unit storage circuit 110 includes a scan trigger circuit 120 and a backup circuit 130.

[0128] The scan trigger circuit 120 includes a selector circuit 121 and a trigger circuit 122. The backup circuit 130 includes a hold circuit 131 and a transistor M33. The hold circuit 131 includes transistor M31, transistor M32 and capacitor C31.

[0129] Various signals controlling the operation of the storage circuit 110 are supplied to wiring BK, wiring RV, wiring SE, wiring PCK, wiring GBK, and wiring Ec.

[0130] Note that these signals can be generated in the control unit (not shown) included in the storage device 100, or they can be generated outside the storage device 100.

[0131] The unit storage circuit 110 can output to the circuit Q by storing and holding data input from the circuit D1 or the circuit SD in the flip-flop circuit 122 within the scan flip-flop circuit 120, synchronized with the clock signal supplied to the circuit PCK. Furthermore, data input from the circuit D2 can be stored and held in the flip-flop circuit 122 according to the signal supplied to the circuit Ec. Additionally, data input from the circuit D2 can be written to and held in the holding circuit 131 within the backup circuit 130 according to the signal supplied to the circuit BK. This operation is sometimes referred to as saving, retaining, storing, or backing up. Furthermore, data held in the holding circuit 131 can be written back to the flip-flop circuit 122 via the circuit SD according to the signal supplied to the circuit RV. This operation is sometimes referred to as loading, regenerating, restoring, or recovering.

[0132] Here, if an error occurs in the data held in the flip-flop circuit 122 due to the error detection and correction unit 103 (the logic value of the held data is inverted), data whose logic value is inverted and output to wiring Q is supplied to wiring D2, and a signal indicating that an error has occurred is supplied to wiring Ec. Furthermore, if no error occurs in the data held in the flip-flop circuit 122 (the logic value of the held data is not inverted), data with the same logic value as the data output to wiring Q is supplied to wiring D2, and a signal indicating that no error has occurred is supplied to wiring Ec.

[0133] The trigger circuit 122 includes an input terminal D1f connected to wiring D1, an input terminal D2f connected to wiring D2, an output terminal Qf connected to wiring Q, and a control terminal Ecf connected to wiring Ec. The trigger circuit 122 has the function of storing and holding data input from the input terminal D1f in sync with the clock signal supplied to wiring PCK, and then outputting it to the output terminal Qf. Furthermore, the trigger circuit 122 has the function of storing and holding data input from the input terminal D2f according to the signal supplied to the control terminal Ecf, and then outputting it to the output terminal Qf.

[0134] A specific structural example of the trigger circuit 122 will be explained later.

[0135] Selector circuit 121 has the function of transmitting data supplied to wiring D1 or wiring SD to trigger circuit 122 according to the signal supplied to wiring SE. Wiring D1 is supplied with encoded data. Wiring SD is supplied with data held by holding circuit 131 within backup circuit 130 or data input from wiring SD_IN. Wiring SD_IN is supplied with data for scan testing.

[0136] As the selector circuit 121, for example, a selector circuit prepared in a standard circuit library can be used.

[0137] When performing power gating, the backup circuit 130 can maintain the state of the scan trigger circuit 120 in the holding circuit 131.

[0138] like Figure 2A As shown, the holding circuit 131 is connected to wiring D2 and wiring SD. In the holding circuit 131, the terminal (wiring) connected to wiring D2 is an input terminal and the terminal (wiring) connected to wiring SD is an output terminal. That is, in the unit storage circuit 110, the input terminal D2f of the trigger circuit 122 is connected to the input terminal of the holding circuit 131, and the input terminal D1f of the trigger circuit 122 is connected to the output terminal of the holding circuit 131 through the selector circuit 121.

[0139] In the holding circuit 131, one of the source and drain of transistor M31 is connected to one terminal of capacitor C31. One of the source and drain of transistor M32 is connected to one terminal of capacitor C31. The other terminal of capacitor C31 is connected to a wiring supplied with a constant potential. The other of the source and drain of transistor M31 is connected to the input terminal of holding circuit 131 (i.e., wiring D2). The other of the source and drain of transistor M32 is connected to the output terminal of holding circuit 131 (i.e., wiring SD). The gate of transistor M31 is connected to wiring BK. The gate of transistor M32 is connected to wiring RV.

[0140] Note that in the holding circuit 131, the wiring connecting one of the source and drain terminals of transistor M31, one of the source and drain terminals of transistor M32, and one terminal of capacitor C31 to each other is sometimes referred to as wiring SN.

[0141] In backup circuit 130, one of the source and drain of transistor M33 is connected to wiring SD. The other of the source and drain of transistor M33 is connected to wiring SD_IN.

[0142] The gate of transistor M33 is connected to wiring GBK. Wiring GBK is supplied with a signal that controls whether a scan test is performed.

[0143] In one embodiment of the invention, transistors M31, M32, and M33 may, for example, be OS transistors. OS transistors have the characteristic of extremely low off-state current. Furthermore, they have the characteristic that the off-state current hardly increases even at high temperatures and the on-state current does not easily decrease.

[0144] Therefore, by deactivating transistors M31 and M32, the holding circuit 131 can retain the data written to the wiring SN for an extended period. For example, by power gating, data can be retained even when power to the scan trigger circuit 120 is stopped. In other words, the holding circuit 131 can be used as a non-volatile memory.

[0145] Here, in the unit storage circuit 110, when data held in the wiring SN is written back to the flip-flop circuit 122, the potential of the data sometimes changes due to the parasitic capacitance of the wiring SD. Therefore, the electrostatic capacitance of the capacitor C31 can be made greater than the parasitic capacitance of the wiring SD in such a way that the change in the potential of the data is, for example, less than the logic threshold of the flip-flop circuit 122.

[0146] Note that, as another example of the structure of the unit storage circuit 110, a structure using a Si transistor as transistor M33 may be adopted.

[0147] In one embodiment of the unit storage circuit 110 of the present invention, a backup circuit 130 can be provided without changing the circuit structure and layout of the scan trigger circuit 120. That is to say, the backup circuit 130 is a highly versatile circuit.

[0148] Furthermore, since the backup circuit 130 is stacked on top of the scan trigger circuit 120 in the unit storage circuit 110, the wiring distance connecting them can be shortened. Therefore, the energy required for data storage and loading (access energy) can be suppressed. Thus, the power consumption of the unit storage circuit 110, which is accompanied by power gating, can be reduced.

[0149] Note that in the unit storage circuit 110, the backup circuit 130 may also include multiple holding circuits 131. Thus, for example, in a computing device, when processing multiple tasks while switching between them, each of the multiple holding circuits 131 can maintain the state of the scan trigger circuit 120 for each task. By adopting this structure, there is no need to transfer data for executing each task between the unit storage circuit 110 and a memory located outside the unit storage circuit 110 (e.g., a cache memory or main memory). Therefore, the time required for switching tasks can be shortened and the access energy associated with data transfer can be reduced. This results in increased operating speed and reduced power consumption of the computing device.

[0150] Next, a specific structural example of the trigger circuit 122 will be described.

[0151] like Figure 2B As shown, the trigger circuit 122 may also include a switch circuit X21, a NOT gate circuit X22, a NOT circuit X23, a switch circuit X24, a switch circuit X25, a NOT circuit X26, a NOT circuit X27, a switch circuit X28, and a buffer circuit X29. Furthermore, the trigger circuit 122 may also include a selector circuit 123.

[0152] Data input from input terminal D1f is stored in a first latch, which includes a first inverter loop consisting of NOT circuits X22 and X23, a switching circuit X21, and a switching circuit X24. Data stored in the first latch is stored in a second latch, which includes a second inverter loop consisting of NOT circuits X26 and X27, a switching circuit X25, and a switching circuit X28. Data stored in the second latch is output to output terminal Qf via buffer circuit X29.

[0153] The inverting control terminal of switch circuit X21, the non-inverting control terminal of switch circuit X24, the non-inverting control terminal of switch circuit X25, and the inverting control terminal of switch circuit X28 are connected to wiring PCKA. The non-inverting control terminals of switch circuit X21, X24, X25, and X28 are connected to wiring PCKB. Wiring PCKA is supplied with a signal having the same logic value as the signal supplied to wiring PCK, and wiring PCKB is supplied with a signal that inverts the logic value of the signal supplied to wiring PCK.

[0154] By employing this structure, the trigger circuit 122 can store and retain the data input from the input terminal D1f in sync with the rising edge of the signal supplied to the wiring PCK, and output the data to the output terminal Qf.

[0155] Furthermore, the trigger circuit 122 includes a selector circuit 123 in the second latch. The selector circuit 123 has the function of selecting and performing either of the following operations based on the signal supplied to the control terminal Ecf: holding the data in the first latch and the second latch; or transmitting the data supplied to the wiring D2f to the first latch and the second latch.

[0156] In other words, for example, if a signal indicating an error has occurred in the data held in the trigger circuit 122 is supplied to the control terminal Ecf, the data input from the input terminal D2f is stored in the trigger circuit 122. Furthermore, if a signal indicating no error has occurred in the data held in the trigger circuit 122 is supplied to the control terminal Ecf, the data stored in the trigger circuit 122 continues to be held.

[0157] Note that the switching circuit included in the trigger circuit 122 can be, for example, an analog switching circuit prepared in a standard circuit library. Furthermore, the NOT circuit can be, for example, an inverter circuit prepared in a standard circuit library. Additionally, the buffer circuit can be, for example, a buffer circuit prepared in a standard circuit library. Furthermore, the selector circuit 123 can be, for example, a selector circuit prepared in a standard circuit library.

[0158] Additionally, one embodiment of the present invention includes a structure in which at least one of the gate, source, and drain of one or more transistors is either not connected to any wiring or is connected to any wiring. Furthermore, another embodiment of the present invention includes a structure in which no content is input to one or more wirings, or any signal or potential is input.

[0159] [Work Example] Next, an example of the operation of the unit storage circuit 110 will be described.

[0160] Figure 3 This is an explanation Figure 2A and Figure 2B Timing diagram of an example operation of the unit storage circuit 110 shown. Figure 3 This illustrates a working example where the power supply to the unit storage circuit 110 is turned off due to power gating.

[0161] In the following description, the signal potential is referred to as potential H or potential L. Potential H is the potential at which an n-channel transistor is turned on when supplied to its gate and is turned off when supplied to its gate. Potential L is the potential at which an n-channel transistor is turned off when supplied to its gate and is turned on when supplied to its gate.

[0162] Note that potentials H and L do not necessarily need to be the same across multiple signals. Each of the multiple signals may also have different potentials H and L depending on the threshold voltage of the transistor to which the signal is supplied. For example, the potentials H and L may be different between a signal supplied to the gate of a transistor included in the scan trigger circuit 120 and a signal supplied to the gate of a transistor included in the backup circuit 130.

[0163] Furthermore, in the following operating instructions, wiring D2 is supplied with the same potential as wiring Q, and wiring Ec is supplied with potential L. Additionally, wiring GBK is supplied with potential L.

[0164] Figure 3 The timing diagram shown illustrates the status (potential H or potential L) of the signals supplied to cabling PCK, cabling BK, cabling RV, and cabling SE during each period of operation (period T11 to period T14). Additionally, the status of the data supplied to each of cabling D1, cabling Q, cabling SD, and cabling SN is shown.

[0165] The period preceding T11 is the period for routine work (Run). T11 is the period for stopping routine work and performing backup. T12 is the period for power gating (PG). T13 is the period for recovery. T14 is the period for resuming routine work (Run).

[0166] Assume that the clock signal is supplied to wiring PCK just before entering period T11. Additionally, assume that wiring BK, wiring RV, and wiring SE are all supplied with a potential L.

[0167] Therefore, synchronously with the rising edge of the clock signal supplied to wiring PCK, the data da supplied to wiring D1 is stored in the scan flip-flop circuit 120 and output to wiring Q. Furthermore, wiring D2 is also supplied with data da. Additionally, the states of the data supplied to wiring SD and wiring SN are undefined. Note that in the following description, unless otherwise specified, the previous states are maintained.

[0168] During period T11, the clock signal supplied to the wiring PCK is stopped.

[0169] Next, wiring BK is supplied with a potential H. Thus, the data da supplied to wiring D2 is stored in wiring SN of holding circuit 131. Then, wiring BK is supplied with a potential L. Thus, the data da stored in wiring SN is held.

[0170] During period T12, power supply to the scan trigger circuit 120 is stopped. In other words, the power supply is turned off.

[0171] At this point, the data da stored in the scan trigger circuit 120 disappears. Furthermore, the data da stored in the wiring SN of the holding circuit 131 is held.

[0172] During period T13, power is supplied to the scan trigger circuit 120 again. In other words, the power is turned on.

[0173] Next, wiring RV is supplied with potential H. Thus, the data da held by wiring SN in holding circuit 131 is supplied to wiring SD. Furthermore, wiring SE is supplied with potential H. Therefore, in selector circuit 121, wiring SD is selected.

[0174] Next, a pulse signal is supplied to wiring PCK. Then, synchronously with the rising edge of this pulse signal, data da supplied to wiring SD is stored in the scan trigger circuit 120 and output to wiring Q. Additionally, data da is also supplied to wiring D2. Then, potential L is supplied to wiring RV and wiring SE.

[0175] During T14, the clock signal to the wiring PCK is supplied again.

[0176] Then, assume that wiring D1 is supplied with data db. Synchronously with the rising edge of the clock signal supplied to wiring PCK, the data db supplied to wiring D1 is stored in the scan flip-flop circuit 120 and output to wiring Q. Furthermore, wiring D2 is also supplied with data db.

[0177] Note that one aspect of the present invention is not limited to the structural and operational examples described in this embodiment. The contents described in this embodiment can be appropriately combined and implemented. Furthermore, the contents described in this embodiment can be appropriately combined and implemented with the contents described in other embodiments, etc.

[0178] (Implementation Method 2) In this embodiment, examples of the structure of transistors and capacitors that can be used in the memory devices described in the above embodiments will be described.

[0179] <Example 1 of transistor and capacitor structure> One aspect of the storage device of the present invention can utilize transistors having various structures. Furthermore, stacked structures of transistors having various structures can be used.

[0180] Figure 4 This is a cross-sectional view of a semiconductor device including transistor 550, transistor 500, and capacitor 590. Furthermore, Figure 4 A cross-sectional view of the channel length direction (shown as X direction) of transistor 550 is shown.

[0181] like Figure 4As shown, transistor 500 is positioned above transistor 550. Capacitor 590 is positioned above transistor 500.

[0182] Furthermore, the conductor 328 on the transistor 550 can also be connected to conductors 591 and 594 on the transistor 550 via conductors 330, 356_1, 356_2, 356_3, 356_4, 518, and 546. Additionally, these conductors can also be formed using conductors that function as plugs or wiring.

[0183] In this specification and other materials, the same symbol is sometimes used to represent multiple conductors that function as plugs or wiring. Furthermore, wiring and plugs can also be a single component. That is, a portion of a conductor is sometimes used as wiring, and another portion of the conductor is sometimes used as a plug.

[0184] For the materials used in the plugs and wiring, single or multilayer conductive materials such as metals, alloys, metal nitrides, or metal oxides can be used.

[0185] In particular, high-melting-point materials that combine heat resistance and electrical conductivity are preferably used for each plug or wiring. Examples of such materials include tungsten or molybdenum. Furthermore, low-resistance conductive materials that reduce wiring resistance are preferred for each plug or wiring. Examples of such materials include aluminum or copper.

[0186] [Transistor 550] The transistor 550 will be described.

[0187] like Figure 4 As shown, transistor 550 is disposed on substrate 311 and includes a conductor 316 serving as a gate electrode, an insulator 315 serving as a gate insulating film, a semiconductor region 313 serving as a channel forming region, a low-resistance region 314a serving as one of the source region and drain region, and a low-resistance region 314b serving as the other of the source region and drain region.

[0188] Furthermore, transistor 550 includes a component separation layer 312 embedded in substrate 311. The component separation layer 312 is disposed between two adjacent transistors 550.

[0189] Furthermore, transistor 550 can be either a p-channel transistor or an n-channel transistor. For example, by connecting the gate of the n-channel transistor 550 to the gate of the p-channel transistor 550, a CMOS circuit (e.g., a circuit in which transistors operate complementaryly, a CMOS logic gate, or a CMOS logic circuit, etc.) can be constructed.

[0190] Therefore, for example, when transistor 550 is used in a storage device, it can also be used to form a transistor or the like that used to construct a peripheral circuit for operating the storage device. Furthermore, for example, in the storage device 100 shown in Embodiment 1 above, transistors can also be used to form the check bit generation unit 102, the error detection and correction unit 103, and the scan trigger circuit 120 included in the unit storage circuit 110.

[0191] Furthermore, transistor 550 may, for example, have the following characteristics: Figure 5 The semiconductor region 313, formed by a portion of substrate 311, has a so-called Fin-type structure where the top surface and the side surfaces in the channel width direction are covered by an insulator 315 and a conductor 316. This increases the effective channel width, thereby improving the on-state characteristics of the transistor 550. Furthermore, since the effect of the electric field at the gate electrode can be enhanced, the off-state characteristics of the transistor 550 can be improved.

[0192] For example, transistor 550 preferably contains a semiconductor such as silicon, preferably monocrystalline silicon, in the channel formation region of semiconductor region 313, the region theren, a low-resistance region 314a serving as one of the source and drain regions, and a low-resistance region 314b serving as the other of the source and drain regions. Alternatively, transistor 550 may also be formed using materials such as germanium, silicon germanium, gallium arsenide, or gallium aluminum arsenide. Alternatively, transistor 550 may also use silicon, where stress is applied to the crystal lattice to change the interplanar spacing and control the effective quality. Alternatively, transistor 550 may also be a high electron mobility transistor (HEMT) using materials such as gallium arsenide and gallium aluminum arsenide.

[0193] For example, in the low resistance regions 314a and 314b, in addition to the semiconductor material applied to the semiconductor region 313, elements such as arsenic and phosphorus that impart n-type conductivity or elements such as boron that impart p-type conductivity are also included.

[0194] As the conductor 316, semiconductor materials such as silicon, which contain elements that impart n-type conductivity, such as arsenic or phosphorus, or elements that impart p-type conductivity, such as boron, can be used. Alternatively, conductive materials such as metallic materials, alloy materials, and metal oxide materials can be used.

[0195] Furthermore, since the work function is determined by the material of the conductor, the threshold voltage of the transistor can be adjusted by selecting the material of the conductor.

[0196] As the conductor 316, materials such as titanium nitride or tantalum nitride are preferably used, for example. Furthermore, in order to combine conductivity and embedding properties, a stack of metallic materials such as tungsten or aluminum is preferably used, for example. In particular, from the perspective of heat resistance, a stack of tungsten is preferred, for example.

[0197] Insulators 320, 322, 324 and 326 are stacked sequentially in a manner that covers transistor 550.

[0198] As insulators 320, 322, 324, and 326, silicon oxide, silicon oxynitride, silicon oxynitride, silicon nitride, aluminum oxide, aluminum oxynitride, aluminum oxynitride, and aluminum nitride can be used, for example. In particular, from the viewpoint of thermal stability, silicon oxide or silicon oxynitride is preferred.

[0199] Note that in this specification, silicon oxynitride refers to a material in which the oxygen content is greater than the nitrogen content, while silicon oxynitride refers to a material in which the nitrogen content is greater than the oxygen content. Note that in this specification, aluminum oxynitride refers to a material in which the oxygen content is greater than the nitrogen content, and aluminum oxynitride refers to a material in which the nitrogen content is greater than the oxygen content.

[0200] The insulator 322 may also have a planarization film function to flatten the steps caused by the transistor 550 and the like disposed below it. For example, in order to improve the flatness of the top surface of the insulator 322, its top surface may also be planarized by a planarization process such as chemical mechanical polishing (CMP).

[0201] As the insulator 324, it is preferable to use a barrier insulator that can prevent impurities such as hydrogen from diffusing from the substrate 311 or transistor 550 located below the insulator 324 to the region located above the insulator 324.

[0202] Silicon nitride, formed by chemical vapor deposition (CVD), can be used as an insulator that blocks hydrogen. Alternatively, metal oxides such as aluminum oxide, hafnium oxide, and tantalum oxide can be used.

[0203] Here, for example, hydrogen may sometimes diffuse into semiconductor devices having oxide semiconductors, such as transistor 500 described later, causing a deterioration in the characteristics of the semiconductor device. Therefore, it is preferable to provide an insulator that suppresses hydrogen diffusion between the region where transistor 500 is disposed and the region where transistor 550 is disposed. Specifically, the insulator that suppresses hydrogen diffusion refers to an insulator with a low amount of hydrogen detachment.

[0204] The dielectric constant of insulator 326 is preferably lower than that of insulator 324. For example, the relative dielectric constant of insulator 326 is preferably less than 4, more preferably less than 3. For example, the relative dielectric constant of insulator 326 is preferably less than 0.7 times that of insulator 324, more preferably less than 0.6 times. By using a material with a low dielectric constant for the interlayer film, parasitic capacitance generated between wirings can be reduced.

[0205] Conductor 328 is embedded in insulators 320 and 322. In addition, conductor 330 is embedded in insulators 324 and 326.

[0206] Both conductor 328 and conductor 330 have the function of plug or wiring.

[0207] Alternatively, a wiring layer can be provided on the insulator 326 and the conductor 330. For example, in Figure 4 In the middle, insulators 350_1, 352_1 and 354_1 are stacked in sequence. In addition, conductors 356_1 are formed in insulators 350_1, 352_1 and 354_1.

[0208] Conductor 356_1 functions as a plug or wiring. Furthermore, conductor 356_1 can be formed, for example, using the same material as conductors 328 and 330. In particular, a conductor that blocks hydrogen is preferred.

[0209] Furthermore, the same materials as those used for insulators 324, 322, and 326 can be used as insulators 350_1, 352_1, and 354_1. In particular, insulators that block hydrogen are preferred.

[0210] Here, a hydrogen-blocking conductor is formed within the opening of the hydrogen-blocking insulator 350_1. By employing this structure, the hydrogen-blocking conductor can be used to separate transistor 550 from transistor 500. Therefore, hydrogen diffusion from transistor 550 to transistor 500 can be suppressed.

[0211] Tantalum nitride, for example, can be used as a conductor that blocks hydrogen. Alternatively, a stack of tantalum nitride and highly conductive tungsten can be used. By using a stack of tantalum nitride and tungsten as the conductor, the conductor can maintain its conductivity as a wiring device while also suppressing hydrogen diffusion.

[0212] In other words, by using tantalum nitride and tungsten as conductors 356_1, not only can the conductivity of the wiring be maintained, but hydrogen diffusion from the transistor 550 can also be suppressed. In this case, the tantalum nitride layer in the hydrogen-blocking conductor 356_1 is preferably in contact with the hydrogen-blocking insulator 350_1.

[0213] Alternatively, a wiring layer can be provided on the insulator 354_1 and the conductor 356_1. For example, in Figure 4 In the middle, insulators 350_2, 352_2 and 354_2 are stacked in sequence. In addition, conductors 356_2 are formed in insulators 350_2, 352_2 and 354_2.

[0214] Conductor 356_2 functions as a plug or wiring. Furthermore, conductor 356_2 can be formed, for example, using the same material as conductor 356_1. In particular, it is preferable to include a conductor that blocks hydrogen.

[0215] Furthermore, the same materials as those used for insulators 350_2, 352_2, and 354_2 can be used as insulators 350_1, 352_1, and 354_1, for example. In particular, insulators that block hydrogen are preferred.

[0216] Alternatively, a wiring layer can be provided on the insulator 354_2 and the conductor 356_2. For example, in Figure 4 In the middle, insulators 350_3, 352_3 and 354_3 are stacked in sequence. In addition, conductors 356_3 are formed in insulators 350_3, 352_3 and 354_3.

[0217] Conductor 356_3 functions as a plug or wiring. Furthermore, conductor 356_3 can be formed, for example, using the same material as conductor 356_1. In particular, it is preferable to include a conductor that blocks hydrogen.

[0218] Furthermore, the same materials as those used for insulators 350_3, 352_3, and 354_3 can be used as insulators 350_1, 352_1, and 354_1. In particular, insulators that block hydrogen are preferred.

[0219] Alternatively, a wiring layer can be provided on the insulator 354_3 and the conductor 356_3. For example, in Figure 4 In the middle, insulators 350_4, 352_4 and 354_4 are stacked in sequence. In addition, conductors 356_4 are formed in insulators 350_4, 352_4 and 354_4.

[0220] Conductor 356_4 functions as a plug or wiring. Furthermore, conductor 356_4 can be formed, for example, using the same material as conductor 356_1. In particular, it is preferable to include a conductor that blocks hydrogen.

[0221] Furthermore, the same materials as those used for insulators 350_4, 352_4, and 354_4 can be used as insulators 350_1, 352_1, and 354_1. In particular, insulators that block hydrogen are preferred.

[0222] Note that this example illustrates a four-layer wiring layer, which is the same as the wiring layer including conductor 356_1, but it is not limited to this. Alternatively, a wiring layer including conductor 356_1 may not be provided, and the number of wiring layers is three or fewer, or five or more.

[0223] Notice, Figure 4 The transistor 550 shown is just an example and is not limited to this structure.

[0224] [Transistor 500] The transistor 500 will be described.

[0225] Figure 6A This is a top view of the transistor 500. Figure 6B It is along Figure 6A The section shown by the dotted lines A1-A2 in the diagram is a cross-sectional view, which is also a cross-sectional view of the channel length of transistor 500 (represented as the X direction). Figure 6C It is along Figure 6A The cross-sectional view shown by the dashed lines A3-A4 is also a cross-sectional view of the channel width of transistor 500 (represented as the Y direction). Note that in Figure 6A In the top view, some constituent elements have been omitted for ease of understanding.

[0226] Transistor 500 is a so-called planar transistor, and compared to vertical transistors such as transistor 600 (described later), its channel length can be easily increased. Therefore, for example, it is easier to reduce short-channel effects such as drain-induced barrier lowering (DIBL). In other words, it is easier to achieve transistors with high saturation (small changes in drain current relative to drain voltage in the transistor's saturation region).

[0227] Therefore, for example, when transistor 500 is used in a storage device, it can also be used to form a transistor for a readout amplifier used to read data from the storage cells included in the storage device. Furthermore, for example, in the storage device 100 shown in Embodiment 1 above, transistor 500 can also be used to form a transistor for a backup circuit 130 included in the unit storage circuit 110.

[0228] like Figure 6B and Figure 6C As shown, insulators 514 and 516 are stacked sequentially on insulator 512.

[0229] As any one of insulator 512, insulator 514 and insulator 516, an insulator that is resistant to oxygen, hydrogen and the like is preferred.

[0230] As the insulator 514, it is preferable to use a barrier insulator that can prevent impurities such as hydrogen from diffusing from the outside of the region where the transistor 500 is disposed into the region where the transistor 500 is disposed. As the insulator 514, for example, the same material as the insulator 324 described above can be used.

[0231] For example, metal oxides such as aluminum oxide, hafnium oxide, and tantalum oxide are preferred as insulators that block hydrogen.

[0232] In particular, aluminum oxide has a high barrier property against both impurities such as oxygen, hydrogen, and water. Therefore, during and after the transistor manufacturing process, aluminum oxide can prevent impurities such as hydrogen and water from entering the transistor 500, and can suppress the release of oxygen from the oxides constituting the transistor 500. Therefore, aluminum oxide is suitable for use as a protective film for the transistor 500.

[0233] Insulators 512 and 516 are made of materials with low dielectric constants, thereby reducing parasitic capacitance generated between wirings. For example, the same material as insulator 326 described above can be used for insulators 512 and 516.

[0234] For example, conductors 518 and the like are embedded in insulators 510, 512, 514 and 516.

[0235] Conductor 518 functions as a plug or wiring. Conductor 518 can be made of the same material as conductors 328 and 330 described above.

[0236] In particular, in conductor 518, the region in contact with insulator 510 and insulator 514 is preferably a conductor that is resistant to both oxygen and impurities such as hydrogen and water. By adopting this structure, transistor 550 can be separated from transistor 500 using a conductor that is resistant to both oxygen and impurities such as hydrogen and water, thereby suppressing the diffusion of hydrogen from transistor 550 into transistor 500.

[0237] like Figure 6B and Figure 6C As shown, transistor 500 includes a conductor 503 disposed in an insulator 514 and an insulator 516, an insulator 522 disposed on the insulator 516 and the conductor 503, an insulator 524 disposed on the insulator 522, an oxide 530a disposed on the insulator 524, an oxide 530b disposed on the oxide 530a, conductors 542a and 542b disposed separately on the oxide 530b, an insulator 580 disposed on the conductors 542a and 542b and having an opening formed therein, overlapping between the conductors 542a and 542b, an insulator 545 disposed along the opening, and a conductor 560 disposed on the forming surface of the insulator 545.

[0238] Note that oxides 530a and 530b are sometimes referred to collectively as oxide 530.

[0239] The oxide 530 has the function of a semiconductor film including the channel formation region of the transistor 500.

[0240] Oxide semiconductors can be used as oxide 530. Note that the semiconductors that can be used as oxide 530 are not limited to oxide semiconductors. Therefore, in this specification and the like, "oxide" may be appropriately replaced with "semiconductor", "semiconductor layer" or "semiconductor film".

[0241] Conductor 503 is arranged to overlap with oxide 530 and conductor 560.

[0242] Here, the conductor 503 preferably includes a conductor 503a disposed in contact with the insulators 514 and 516, and a conductor 503b disposed embedded inside the conductor 503a. Furthermore, it is preferable that an insulator 544 is disposed between the oxides 530a and 530b, the conductors 542a and 542b, and the insulator 580. Furthermore, the conductor 560 preferably includes a conductor 560a disposed inside the insulator 545 and a conductor 560b disposed embedded inside the conductor 560a. Furthermore, it is preferable that an insulator 582 is disposed on the insulator 580, the conductor 560, and the insulator 545.

[0243] Note that in Figure 6B and Figure 6C In the transistor 500 shown, two layers, conductor 503a and conductor 503b, are stacked as conductors 503, but the transistor is not limited to this. For example, it may have a single-layer structure or a stacked structure of three or more layers.

[0244] Furthermore, a structure is shown in which the ends of conductors 542a and 542b coincide with the ends of oxide 530, but the structure is not limited thereto. For example, conductors 542a and 542b may extend beyond the ends of oxide 530.

[0245] Furthermore, the oxide 530 is shown to have a two-layer stacked structure of oxide 530a and oxide 530b, but it is not limited thereto. For example, it may also have a single-layer structure or a stacked structure of three or more layers.

[0246] Furthermore, the conductor 560 is shown as having a two-layer structure with conductors 560a and 560b stacked together, but it is not limited to this. For example, it may also have a single-layer structure or a stacked structure with three or more layers.

[0247] Here, in transistor 500, conductor 560 functions as a gate electrode, insulator 545 functions as a gate insulating film, and conductors 542a and 542b function as one of the source electrode and the other of the drain electrode.

[0248] As described above, the conductor 560 is formed by embedding the insulator 580 in an opening (sometimes referred to as the opening of the insulator 580) in the region sandwiched between the conductors 542a and 542b. Therefore, the conductors 560, 542a, and 542b are self-aligned relative to the opening of the insulator 580. That is, in the transistor 500, the gate electrode can be self-aligned between the source and drain electrodes. By employing this structure, the conductor 560 can be formed without requiring any alignment space. Therefore, the occupied area of ​​the transistor 500 can be reduced. This allows for miniaturization or high integration of semiconductor devices.

[0249] Furthermore, the conductor 560 is self-aligned and disposed in the region between conductors 542a and 542b, so the conductor 560 does not include a region overlapping with conductors 542a or 542b. This reduces the parasitic capacitance formed between the conductor 560 and conductors 542a and 542b. Therefore, the switching speed of the transistor 500 can be improved. Consequently, the frequency characteristics of the semiconductor device can be improved.

[0250] Furthermore, when miniaturizing semiconductor devices, it is necessary to shorten the gate length of the transistor 500, but at this time, it is necessary to prevent a decrease in the conductivity of the conductor 560. To this end, by increasing the thickness of the conductor 560, the conductor 560 may have a shape with a high aspect ratio. Therefore, by providing the conductor 560 in such a way that it is embedded in the opening of the insulator 580, a conductor 560 with a high aspect ratio can be formed without collapsing during the process.

[0251] Here, conductor 560 sometimes functions as a first gate (also simply called a gate) electrode, and conductor 503 sometimes functions as a second gate (also called a back gate) electrode. In this case, insulator 545 functions as a first gate insulating film, and insulators 522 and 524 function as second gate insulating films.

[0252] As described above, the conductor 503 is arranged to overlap with the oxide 530 and the conductor 560. Therefore, when a potential is supplied to the conductor 560 and the conductor 503, the electric field generated from the conductor 560 and the electric field generated from the conductor 503 are connected and can cover the channel formation region formed in the oxide 530.

[0253] In this case, in transistor 500, the threshold voltage of transistor 500 can be controlled by changing the potential supplied to conductor 503 independently of the potential supplied to conductor 560. In particular, by supplying a negative potential to conductor 503, the threshold voltage of transistor 500 can be increased and the off-state current can be decreased. Therefore, for example, by supplying a negative potential to conductor 503, the drain current (sometimes called the cutoff current) when the potential supplied to conductor 560 is 0V can be reduced.

[0254] Note that in this specification, the structure of a transistor in which the electric field of the gate electrode forms a region surrounding the channel is referred to as a surrounded channel (S-channel) structure. Furthermore, it can be said that the S-channel structure disclosed in this specification differs from Fin-type and planar structures. On the other hand, the S-channel structure disclosed in this specification can also be considered as either a Fin-type or planar structure. Furthermore, in this specification, a Fin-type structure refers to a structure in which the gate electrode is arranged with at least two or more surfaces surrounding the channel (specifically, two, three, or four surfaces, etc.). By employing Fin-type and S-channel structures, transistors with improved tolerance to short-channel effects can be realized. In other words, transistors that are less prone to short-channel effects can be realized.

[0255] By employing a transistor with the aforementioned S-channel structure, a region can be formed around the channel by the electric field of the gate electrode. Because the S-channel structure forms a region around the channel by the electric field of the gate electrode, it can be said that this structure is essentially the same as a GAA (Gate All Around) structure or a LGAA (Lateral Gate All Around) structure. By equipping the transistor with an S-channel, GAA, or LGAA structure, the channel formation region formed at or near the interface between the semiconductor film and the gate insulating film can be disposed throughout the entire bulk of the semiconductor film. Therefore, the current density flowing through the transistor can be increased, thereby improving the transistor's on-state current or field-effect mobility.

[0256] As described above, in the conductor 503, a conductor 503a is formed in contact with the insulator 514 and the insulator 516, and a conductor 503b is formed on its inner side.

[0257] As the conductor 503a, a conductive material that has barrier properties (i.e., at least one of hydrogen atoms and hydrogen molecules), water, and copper, is preferably used. Furthermore, as the conductor 503a, a conductive material that has barrier properties (i.e., has the function of inhibiting the diffusion of the impurity, i.e., the impurity does not easily permeate) is preferably used. In other words, the conductor 503a preferably has barrier properties against one or all of the aforementioned impurities and oxygen.

[0258] For example, by making conductor 503a oxygen-barrier, the decrease in conductivity caused by oxidation of conductor 503b can be suppressed. Therefore, conductor 503 can also function as a wiring conductor.

[0259] In this case, a conductive material with high conductivity is preferably used as the conductor 503b. For example, a conductive material with tungsten, copper, or aluminum as the main components can be used.

[0260] In addition, transistor 500 may also exclude conductor 503 (i.e., exclude back gate).

[0261] Here, an oxide semiconductor that can be used with oxide 530 is described. This oxide semiconductor contains a metal oxide.

[0262] The metal oxide preferably contains at least one of indium and zinc. For example, it preferably contains indium, M (M is one or more selected from gallium, aluminum, yttrium, tin, silicon, boron, copper, vanadium, beryllium, titanium, iron, nickel, germanium, zirconium, molybdenum, lanthanum, cerium, neodymium, hafnium, tantalum, tungsten, magnesium, and cobalt), and zinc. In particular, M is preferably one or more selected from gallium, aluminum, yttrium, and tin.

[0263] In particular, oxides containing indium (In), gallium (Ga), and zinc (Zn) (also denoted as IGZO) are preferred as metal oxides. Alternatively, oxides containing indium (In), aluminum (Al), and zinc (Zn) (also denoted as IAZO) may be used. Alternatively, oxides containing indium (In), aluminum (Al), gallium (Ga), and zinc (Zn) (also denoted as IAGZO) may be used. Alternatively, oxides containing indium (In), tin (Sn), and zinc (Zn) (also denoted as "ITZO (registered trademark)") may be used. Alternatively, oxides containing indium (In), gallium (Ga), zinc (Zn), and tin (Sn) (also denoted as "IGZTO") may be used.

[0264] By increasing the proportion of indium atoms in the total number of atoms of all metal elements in the metal oxide, excellent properties such as high on-state current, high field-effect mobility, and high frequency characteristics can be obtained in transistors in which the metal oxide is used as a semiconductor film including the channel formation region.

[0265] When the metal oxide is an In-M-Zn oxide, the ratio of the number of In atoms in the In-M-Zn oxide is preferably greater than or equal to the number of M atoms. Examples of possible atomic ratios of the metal elements in this In-M-Zn oxide include In:M:Zn = 1:1:1 or similar, In:M:Zn = 1:1:1.2 or similar, In:M:Zn = 2:1:3 or similar, In:M:Zn = 3:1:2 or similar, In:M:Zn = 4:2:3 or similar, In:M:Zn = 4:2:4.1 or similar, In:M:Zn = 5:1:3 or similar, In:M:Zn = 5:1:6 or similar, In:M:Zn = 5:1:7 or similar, In:M:Zn = 5:1:8 or similar, In:M:Zn = 6:1:6 or similar, or In:M:Zn = 5:2:5 or similar. Furthermore, sometimes the atomic ratio of In in this In-M-Zn oxide can be less than the atomic ratio of M. Examples of suitable atomic ratios for the metal elements in this In-M-Zn oxide include In:M:Zn = 1:3:2 or similar compositions, In:M:Zn = 1:3:4 or similar compositions, etc. Note that "similar compositions" include a range of ±30% of the desired atomic ratio.

[0266] For example, when the atomic ratio of the metallic elements is recorded as In:Ga:Zn = 4:2:3 or a similar composition, the content ratio of each element includes the following cases: when In is 4, Ga is 1 or more and 3 or less, and Zn is 2 or more and 4 or less. Furthermore, when the atomic ratio of the metallic elements is recorded as In:Ga:Zn = 5:1:6 or a similar composition, the content ratio of each element includes the following cases: when In is 5, Ga is greater than 0.1 and 2 or less, and Zn is 5 or more and 7 or less. Moreover, when the atomic ratio of the metallic elements is recorded as In:Ga:Zn = 1:1:1 or a similar composition, the content ratio of each element includes the following cases: when In is 1, Ga is greater than 0.1 and 2 or less, and Zn is greater than 0.1 and 2 or less.

[0267] Furthermore, when stacking metal oxides, a three-layer structure can be adopted, for example: a first layer is a metal oxide with an atomic ratio of In:Ga:Zn = 1:1:1; a second layer is a metal oxide with an atomic ratio of In:Zn = 4:1; and a third layer is a metal oxide with an atomic ratio of In:Ga:Zn = 1:1:1. Preferably, the band gaps of the first and third metal oxide layers are larger than the band gap of the second metal oxide layer. By adopting this structure, the second metal oxide layer can be used as the main current path, thus realizing a so-called embedded channel structure.

[0268] Analysis of the composition of metal oxides can be performed using methods such as secondary ion mass spectrometry (SIMS), energy-dispersive X-ray spectrometry (EDX), X-ray photoelectron spectrometry (XPS), inductively coupled plasma mass spectrometry (ICP-MS), or inductively coupled plasma atomic emission spectrometry (ICP-AES). Alternatively, multiple methods can be combined. Note that the actual content of elements with low concentrations can differ from the analytically obtained content due to variations in analytical precision. For example, when the content of element M is low, the analytically obtained content of element M may sometimes be lower than the actual content.

[0269] Metal oxides can be formed using sputtering or atomic layer deposition (ALD). Note that when metal oxides are formed using sputtering, the composition of the resulting metal oxide sometimes differs from that of the sputtering target. In particular, the zinc content of the formed metal oxide can sometimes be reduced to about 50% of the zinc content in the sputtering target.

[0270] Oxide semiconductors are preferably crystalline. Examples of crystalline oxide semiconductors include CAAC-OS (c-axis aligned crystalline oxide semiconductor), nc-OS (nanocrystalline oxide semiconductor), polycrystalline oxide semiconductors, and single-crystal oxide semiconductors. CAAC-OS or nc-OS are preferred as oxide semiconductors, and CAAC-OS is particularly preferred.

[0271] CAAC-OS preferably has multiple layered crystal regions with its c-axis oriented in the normal direction of the formed surface. For example, the oxide semiconductor preferably has layered crystals parallel to the formed surface. By adopting this structure, the layered crystals of the oxide semiconductor are parallel to the channel length direction of the transistor, thus increasing the on-state current of the transistor.

[0272] Furthermore, in one embodiment of the oxide semiconductor deposition method of the present invention, the crystallinity of the oxide semiconductors formed above and below the CAAC-OS can be improved by using a CAAC-OS as a nucleus or seed. This improves the overall crystallinity of the oxide semiconductor. In other words, by using the CAAC-OS as a nucleus or seed to grow the oxide semiconductors above and below in a solid phase, a highly crystalline oxide semiconductor can be formed. The oxide semiconductor formed using this deposition method can be referred to as axially grown CAAC (AG CAAC).

[0273] By improving the crystallinity of oxide semiconductors, transistors with excellent characteristics (e.g., transistors with high on-state current, high field-effect mobility, low S-value, high frequency characteristics (also known as f-characteristics), and high reliability) can be realized when the oxide semiconductor is used in a semiconductor film including a channel formation region.

[0274] Furthermore, it is preferable to perform a process to improve the crystallinity of the oxide semiconductor during or after deposition. Examples of such processes for improving the crystallinity of the oxide semiconductor include heat treatment, plasma treatment, microwave (typically 2.45 GHz) treatment, microwave plasma treatment, and light (e.g., ultraviolet light) irradiation. Note that multiple of these processes can be performed simultaneously or sequentially. For example, heat treatment and microwave plasma treatment can be performed simultaneously. Alternatively, microwave plasma treatment can be performed after heat treatment.

[0275] In this specification and the like, microwaves refer to electromagnetic waves with frequencies of 300 MHz or higher and 300 GHz or lower. Microwave plasma treatment, for example, refers to a process using a device that includes a power source for generating high-density plasma using microwaves. Microwave plasma treatment may also be referred to as microwave-excited high-density plasma treatment.

[0276] Furthermore, it is preferable to perform multiple treatments to improve the crystallinity of the oxide semiconductor during deposition. For example, when forming an oxide semiconductor film by ALD, it is preferable to perform microwave plasma treatment after forming one atomic layer at a time. Alternatively, performing a treatment to improve crystallinity after forming an oxide semiconductor layer of a predetermined thickness each time is preferable, as this can improve productivity. Specifically, it is preferable to form a first oxide semiconductor film of 1 nm or more and 10 nm or less and perform a first microwave plasma treatment, and then form a second oxide semiconductor film of 1 nm or more and 10 nm or less and perform a second microwave plasma treatment. Note that there are no particular limitations on the deposition methods of the first and second oxide semiconductor films; either ALD or sputtering can be used, respectively. In particular, by using ALD to deposit the first oxide semiconductor film, it is preferable to prevent elements constituting the layer of the formed surface from mixing into the first and second oxide semiconductor films (also known as mixing). In particular, it is suitable for cases where the element in the layer constituting the formed surface blocks the crystallization of the oxide semiconductor (e.g., cases containing silicon, carbon, etc.). Furthermore, the first and second oxide semiconductor films may have different compositions. Furthermore, a stacked structure of a first oxide semiconductor film and a second oxide semiconductor film is shown here, but it is not limited to this. The same treatment can also be used when the oxide semiconductor film has a single-layer structure or a stacked structure of three or more layers.

[0277] Alternatively, a process to improve the crystallinity of the oxide semiconductor can be performed after deposition. Specifically, this process can be performed directly on the deposited oxide semiconductor or through an insulating film or other film deposited on the oxide semiconductor. For example, microwave plasma treatment can be performed after oxide semiconductor deposition, or an insulating film (e.g., silicon nitride film, silicon oxide film, aluminum oxide film, etc.) can be deposited after oxide semiconductor deposition, and then the oxide semiconductor can be subjected to heat treatment or microwave plasma treatment through the insulating film.

[0278] Note that the above-described treatment for improving the crystallinity of oxide semiconductors can also serve as a treatment for removing impurities from the oxide semiconductor. For example, carbon, hydrogen, nitrogen, etc., can be appropriately removed from the oxide semiconductor. Alternatively, by performing the treatment for improving the crystallinity of the oxide semiconductor under an oxygen gas atmosphere, the number of oxygen vacancies (also known as V vacancies) in the oxide semiconductor can be reduced. o (oxygen vacancy).

[0279] When performing a process to improve the crystallinity of oxide semiconductors, it is preferable to set the substrate temperature to room temperature (e.g., 25°C) or higher, 100°C or higher and 600°C or lower, or 300°C or higher and 450°C or lower. Furthermore, the temperature for heat treatment is preferably 100°C or higher and 700°C or lower, or 300°C or higher and 450°C or lower.

[0280] By increasing the crystallinity of oxide semiconductors, transistors with high reliability can be achieved.

[0281] The crystallinity of oxide semiconductors can be analyzed, for example, by X-ray diffraction (XRD) patterns, transmission electron microscopy (TEM) images, or electron diffraction (ED) patterns. Alternatively, multiple of the above methods can be combined for analysis.

[0282] Furthermore, it is preferable to minimize the amount of hydrogen in the oxide semiconductor. Hydrogen in oxide semiconductors bonds with oxygen vacancies, forming defects (also known as V0) where hydrogen enters oxygen vacancies. O Therefore, transistor characteristics (e.g., the initial Id-Vg characteristic of the transistor or the Id-Vg characteristic in long-term reliability testing) may deteriorate. Thus, materials with low hydrogen release are preferred as the surrounding material of the oxide semiconductor, such as the material used as an insulator in contact with the oxide semiconductor. Examples of materials with low hydrogen release include silicon nitride, silicon oxynitride, aluminum oxide, and hafnium oxide. This suppresses hydrogen incorporation into the oxide semiconductor. In particular, when silicon nitride is used in at least one of the insulators in contact with the oxide semiconductor, transistor reliability can be improved. Furthermore, materials with low hydrogen release sometimes have the function of trapping or fixing (also known as gettering) hydrogen inside the insulator.

[0283] Furthermore, in oxide semiconductors, sometimes V OH is used as a donor to generate electrons as charge carriers. Furthermore, sometimes electrons are generated as charge carriers due to partial bonding of hydrogen with oxygen atoms bonded to metal atoms. Therefore, transistors using oxide semiconductors containing a large amount of hydrogen tend to have always-on characteristics. However, because hydrogen in oxide semiconductors is easily transferred due to heat, electric fields, etc., a large amount of hydrogen in the oxide semiconductor may lead to a decrease in transistor reliability. In one aspect of the invention, it is preferable to minimize Vo in the oxide semiconductor. O H thus becomes a high-purity intrinsic or substantially high-purity intrinsic.

[0284] Thus, in order to obtain V O For oxide semiconductors where H is sufficiently reduced, it is important to: remove impurities such as hydrogen and water from the oxide semiconductor (sometimes referred to as "dehydration" or "dehydrogenation treatment"); and supply oxygen to the oxide semiconductor to fill oxygen vacancies (sometimes referred to as "oxidation treatment"). By increasing V... O Oxide semiconductors with sufficiently reduced impurities such as hydrogen (H) can impart stable electrical characteristics when used in the channel formation region of transistors.

[0285] Here, the insulator in contact with oxide 530 is preferably an insulator containing oxygen in excess of its stoichiometric composition. This oxygen is readily released from the insulator by heating. In this specification and the like, the oxygen released by heating is sometimes referred to as "excess oxygen".

[0286] Insulator 524 is in contact with oxide 530. Therefore, it is preferable to form a region containing excess oxygen (also called "excess oxygen region") in insulator 524.

[0287] By providing an insulator containing excess oxygen in a manner that contacts the oxide 530, oxygen vacancies in the oxide 530 can be reduced, thereby improving the reliability of the transistor 500.

[0288] When insulator 524 includes a region of excess oxygen, insulator 522 preferably has oxygen-blocking properties. When insulator 522 has oxygen-blocking properties, for example, the diffusion of oxygen contained in oxide 530 to the insulator 516 side can be suppressed. In addition, the reaction between conductor 503 and oxygen contained in insulator 524 or oxide 530 can be suppressed.

[0289] Here, in the channel formation region of a transistor using an oxide semiconductor as a semiconductor film, it is preferable that there are fewer oxygen vacancies or a lower impurity concentration (e.g., concentrations of hydrogen, nitrogen, and metal elements) compared to the source and drain regions. Furthermore, sometimes Vo is formed from hydrogen near the oxygen vacancies. O H generates electrons that become charge carriers, therefore V OH is also preferably less. Thus, the channel formation region of this transistor is a high-resistivity region with low carrier concentration. Therefore, the channel formation region of this transistor can be described as i-type (intrinsic) or substantially i-type.

[0290] Furthermore, compared to the channel formation region of the transistor, the source and drain regions preferably have more oxygen vacancies and more Vo. O H or a higher impurity concentration. Thus, compared to the channel region of the transistor, the source and drain regions are n-type regions with higher carrier concentration and lower resistance.

[0291] The band gap of the metal oxide used as an oxide semiconductor is preferably 2 eV or more, more preferably 2.5 eV or more. By using a metal oxide with a large band gap as an oxide semiconductor, the off-state current of the transistor can be reduced.

[0292] As the insulator 522, it is preferable to use an insulator made of a high-k material.

[0293] With the miniaturization and high integration of transistors, problems such as gate leakage current sometimes occur due to the thinning of the gate insulating film. By using high-k materials as the insulator for the gate insulating film, the gate potential during transistor operation can be reduced while maintaining the physical thickness.

[0294] As the insulator used as the gate insulating film, a single layer or stack of insulators containing aluminum oxide, hafnium oxide, oxides containing aluminum and hafnium (hafnium aluminate), tantalum oxide, zirconium oxide, lead zirconate titanate (PZT), strontium titanate (SrTiO3) or barium strontium titanate (BST) is preferably used.

[0295] In particular, as an insulator that blocks oxygen and impurities, it is preferable to use an insulator containing an oxide of one or both of aluminum and hafnium. For example, aluminum oxide, hafnium oxide, and oxides containing aluminum and hafnium (hafnium aluminate) are preferred as such insulators.

[0296] In addition, aluminum oxide, bismuth oxide, germanium oxide, niobium oxide, silicon oxide, titanium oxide, tungsten oxide, yttrium oxide, or zirconium oxide can be added to the insulator. Alternatively, the insulator can be nitrided. Furthermore, silicon oxide, silicon oxynitride, or silicon nitride can be laminated onto the insulator.

[0297] When this material is used to form insulator 522, insulator 522 can be used as an insulator to suppress the release of oxygen from oxide 530 and the entry of impurities such as hydrogen from the periphery of transistor 500 into oxide 530.

[0298] Note that in Figure 6B and Figure 6CIn the transistor 500 shown, two layers of insulator 522 and insulator 524 are stacked as the second gate insulating film, but the transistor is not limited to this. For example, it may have a single-layer structure or a stacked structure of three or more layers. In this case, a structure of stacked materials of the same material or different materials may be used.

[0299] In transistor 500, a single layer or stack of metal oxide used as an oxide semiconductor is used as oxide 530, which includes the channel formation region. For example, a semiconductor suitable for oxide 650 (described later) can also be used as oxide 530. Note that the semiconductor that can be used for oxide 530 is not limited to metal oxide.

[0300] In oxide 530, when oxide 530a is disposed below oxide 530b, impurities can be suppressed from diffusing from the structure formed below oxide 530a to oxide 530b.

[0301] Furthermore, oxide 530 preferably has a structure comprising multiple oxide layers in which the atomic ratios of each metal atom are different. Specifically, the atomic ratio of element M in the metal oxide of oxide 530a is preferably greater than that in the metal oxide of oxide 530b. Furthermore, the atomic ratio of element M relative to In in the metal oxide of oxide 530a is preferably greater than that in the metal oxide of oxide 530b. Furthermore, the atomic ratio of In relative to element M in the metal oxide of oxide 530b is preferably greater than that in the metal oxide of oxide 530a.

[0302] Preferably, the energy of the conduction band bottom of oxide 530a is higher than that of oxide 530b. In other words, the electron affinity of oxide 530a is preferably less than that of oxide 530b.

[0303] Here, at the junction of oxides 530a and 530b, the energy level at the bottom of the conduction band changes gradually. In other words, the above situation can also be expressed as the energy level at the bottom of the conduction band at the junction of oxides 530a and 530b changing continuously or continuously joining. For this reason, the defect state density of the mixed layer formed at the interface of oxides 530a and 530b can also be reduced.

[0304] Specifically, by including common elements (as the main components) in addition to oxygen in oxides 530a and 530b, a mixed layer with low defect state density can be formed. For example, when oxide 530b is an In-Ga-Zn oxide, In-Ga-Zn oxide, Ga-Zn oxide, and gallium oxide can be used as oxide 530a.

[0305] At this point, the primary pathway for charge carriers is oxide 530b. By giving oxide 530a the aforementioned structure, the defect state density at the interface between oxide 530a and oxide 530b can be reduced. Therefore, the influence of interface scattering on charge carrier conduction is reduced, which can increase the on-state current of transistor 500.

[0306] The oxide 530b is provided with conductors 542a and 542b, which serve as source and drain electrodes, respectively.

[0307] As conductors 542a and 542b, for example, metallic elements selected from aluminum, chromium, copper, silver, gold, platinum, tantalum, nickel, titanium, molybdenum, tungsten, hafnium, vanadium, niobium, manganese, magnesium, zirconium, beryllium, indium, ruthenium, iridium, strontium, and lanthanum, alloys containing the aforementioned metallic elements, or alloys combining the aforementioned metallic elements can be used. In particular, as conductive materials that are not easily oxidized or that maintain their conductivity even when absorbing oxygen, tantalum nitride, titanium nitride, nitrides containing titanium and aluminum, nitrides containing tantalum and aluminum, ruthenium oxide, ruthenium nitride, oxides containing strontium and ruthenium, and oxides containing lanthanum and nickel are preferred. Furthermore, from the viewpoint of blocking oxygen and hydrogen, metal nitride films such as tantalum nitride are preferred, for example.

[0308] Note that in Figure 6B and Figure 6C In the transistor 500 shown, conductors 542a and 542b have a single-layer structure, but are not limited to this. Conductors 542a and 542b may also have a stacked structure of two or more layers, for example.

[0309] As conductors 542a and 542b, for example, a stacked structure of tantalum nitride film and tungsten film, a stacked structure of titanium film and aluminum film, a structure of aluminum film stacked on tungsten film, a structure of copper film stacked on copper-magnesium-aluminum alloy film, a structure of copper film stacked on titanium film, or a structure of copper film stacked on tungsten film can also be used.

[0310] In addition, for example, a three-layer structure can also be adopted, in which an aluminum film or a copper film is stacked on a titanium film or a titanium nitride film and a titanium film or a titanium nitride film is stacked on top of it; or a three-layer structure in which an aluminum film or a copper film is stacked on a molybdenum film or a molybdenum nitride film and a molybdenum film or a molybdenum nitride film is stacked on top of it, etc.

[0311] In addition, conductors 542a and 542b may also be made of transparent conductive materials containing indium oxide, tin oxide or zinc oxide.

[0312] In addition, such as Figure 6B As shown, region 543a is sometimes formed as a low-resistance region at and near the interface between oxide 530 and conductor 542a. Similarly, region 543b is sometimes formed as a low-resistance region at and near the interface between oxide 530 and conductor 542b. In this case, region 543a is used as one of the source region and drain region, and region 543b is used as the other of the source region and drain region. Furthermore, a channel forming region is formed in the region sandwiched between region 543a and region 543b.

[0313] Thus, by placing conductors 542a and 542b in contact with oxide 530, the oxygen concentration in regions 543a and 543b sometimes decreases. Furthermore, a metal compound layer containing components of the metal in conductors 542a and 542b and oxide 530 sometimes forms in regions 543a and 543b. In this case, the carrier concentration in regions 543a and 543b increases, and regions 543a and 543b become low-resistance regions.

[0314] The insulator 544 is provided in such a way that it covers the conductors 542a and 542b to suppress the oxidation of the conductors 542a and 542b. In this case, the insulator 544 may also be provided in such a way that it covers the oxide 530 and the respective sides of the insulator 524 and is in contact with the insulator 522.

[0315] As the insulator 544, one or more metal oxides selected from hafnium, aluminum, gallium, yttrium, zirconium, tungsten, titanium, tantalum, nickel, germanium, neodymium, lanthanum, or magnesium can be used. Alternatively, silicon oxynitride or silicon nitride can also be used, for example.

[0316] Alternatively, insulators containing oxides of one or both of aluminum and hafnium can be used. For example, alumina, hafnium oxide, and oxides containing aluminum and hafnium (hafnium aluminate) can be used. In particular, hafnium aluminate is preferred, for example, because of its high heat resistance and resistance to crystallization during subsequent heat treatment processes.

[0317] Furthermore, if conductors 542a and 542b are made of materials that are resistant to oxidation or materials whose conductivity does not decrease significantly when absorbing oxygen, it is not necessarily necessary to provide insulator 544.

[0318] Furthermore, by including insulator 544, the diffusion of impurities such as hydrogen and water contained in insulator 580 to oxide 530b can be suppressed. Additionally, the oxidation of conductors 542a and 542b by excess oxygen contained in insulator 580 can be suppressed.

[0319] As the insulator 545, it is preferable to use an insulator containing excess oxygen and releasing oxygen by heating, similar to the insulator 524 described above. Therefore, oxygen can be effectively supplied from the insulator 545 to the channel forming region of the oxide 530b.

[0320] Specifically, as the insulator 545, silicon oxide with excess oxygen, silicon oxynitride, silicon oxynitride, silicon nitride, silicon oxide with added fluorine, silicon oxide with added carbon, silicon oxide with added carbon and nitrogen, and silicon oxide with pores can be used. In particular, from the viewpoint of thermal stability, silicon oxide or silicon oxynitride is preferred.

[0321] Furthermore, similar to the insulator 524 described above, the insulator 545 preferably uses an insulator with reduced concentrations of impurities such as hydrogen and water. The thickness of the insulator 545 is preferably 1 nm or more and 20 nm or less.

[0322] To efficiently supply excess oxygen contained in the insulator 545 to the oxide 530, a metal oxide may be disposed between the insulator 545 and the conductor 560. This metal oxide preferably has oxygen-blocking properties. This suppresses the diffusion of excess oxygen from the insulator 545 to the conductor 560. Therefore, the reduction of excess oxygen supplied to the oxide 530 can be suppressed. Furthermore, oxidation of the conductor 560 due to excess oxygen can be suppressed. As this metal oxide, a material suitable for the insulator 544 can also be used.

[0323] exist Figure 6B and Figure 6C In the transistor 500 shown, the insulator 545 has a single-layer structure, but is not limited to this. For example, similar to the insulators 522 and 524 used as the second gate insulating film, the insulator 545 used as the first gate insulating film can also have a stacked structure of two or more layers. For example, the insulator 545 can also have a stacked structure of a high-k material and a thermally stable material. Thus, the gate voltage of the transistor 500 during operation can be reduced while maintaining the physical thickness of the insulator 545.

[0324] As the conductor 560a included in the conductor 560, a conductive material that blocks impurities such as hydrogen, water, nitrogen, nitrogen oxides (e.g., N2O, NO, or NO2), and copper is preferably used. Furthermore, a conductive material that blocks oxygen is preferably used. When the conductor 560a blocks oxygen, the oxidation of the conductor 560b by the oxygen contained in the insulator 545 can be prevented, thus suppressing a decrease in conductivity.

[0325] As a conductive material that blocks oxygen, tantalum, tantalum nitride, ruthenium, or ruthenium oxide are preferred.

[0326] Furthermore, an oxide semiconductor applicable to oxide 530 can be used as conductor 560a. In this case, by depositing conductor 560b using a sputtering method, the resistance of conductor 560a can be reduced to make it a conductor. It can be referred to as an OC (Oxide Conductor) electrode.

[0327] Conductor 560 can also function as a wiring device. Therefore, similar to conductor 503b, a conductive material with high conductivity is preferably used as conductor 560b. For example, a conductive material with tungsten, copper, or aluminum as its main components can be used.

[0328] Furthermore, the conductor 560b may also have a structure in which different materials are stacked. For example, it may have a stacked structure of titanium or titanium nitride with the aforementioned conductive materials.

[0329] The insulator 580 is preferably disposed on the conductors 542a and 542b, with the insulator 544 in between.

[0330] Insulator 580 preferably includes an excess oxygen region.

[0331] As the insulator 580, for example, silicon oxide, silicon oxynitride, silicon oxynitride, silicon nitride, fluorine-added silicon oxide, carbon-added silicon oxide, silicon oxide with both carbon and nitrogen added, porous silicon oxide, or resin can be used. In particular, from the viewpoint of thermal stability, silicon oxide or silicon oxynitride is preferred. Furthermore, from the viewpoint of preventing the formation of excess oxygen regions in subsequent processes, silicon oxide or porous silicon oxide is preferred.

[0332] By incorporating an insulator 580 that includes an excess oxygen region, oxygen can be released upon heating, thereby efficiently supplying oxygen from the insulator 580 to the oxide 530. Furthermore, it is preferable to reduce the concentration of impurities such as hydrogen and water in the insulator 580.

[0333] The insulator 582 is preferably disposed in contact with the top surface of the insulator 580, the top surface of the conductor 560, and the top surface of the insulator 545. By depositing the insulator 582 using a sputtering method, excess oxygen regions can be formed in the insulator 545 and the insulator 580. Oxygen can then be supplied from these excess oxygen regions to the oxide 530.

[0334] As an insulator 582, for example, one or more metal oxides selected from hafnium, aluminum, gallium, yttrium, zirconium, tungsten, titanium, tantalum, nickel, germanium and magnesium can be used.

[0335] In particular, alumina exhibits high barrier properties against impurities such as hydrogen, suppressing the diffusion of these impurities even in thin films with a thickness of 0.5 nm to 3.0 nm. Therefore, alumina deposited using sputtering can function as both an oxygen supply source and an insulator that blocks impurities such as hydrogen.

[0336] Preferably, an insulator 584, which serves as an interlayer film, is provided on the insulator 582. Similar to the insulator 524, it is preferable to reduce the concentration of impurities such as hydrogen and water in the insulator 584.

[0337] Furthermore, conductors 540a and 540b are disposed in the openings formed in insulators 584, 582, 580, and 544. In this case, conductors 540a and 540b are disposed opposite each other with conductor 560 sandwiched between them. Conductors 540a and 540b have the same structure as conductor 546, which will be described later.

[0338] An insulator 586 is provided on insulator 584.

[0339] As insulator 586, it is preferable to use an insulating material that blocks oxygen and hydrogen. As insulator 586, for example, the same material as insulator 514 can be used.

[0340] Insulator 588 is provided on insulator 586.

[0341] By using a material with a low dielectric constant as insulator 588, the parasitic capacitance generated between wirings can be reduced. For example, the same material as insulator 512 or insulator 516 can be used as insulator 588.

[0342] For example, conductor 546 is embedded in insulator 580, insulator 582, insulator 584, insulator 586 and insulator 588.

[0343] Conductor 546 has the function of a plug or wiring.

[0344] Alternatively, after forming the transistor 500, an opening can be formed around the transistor 500, and an insulator with high barrier properties against hydrogen or water can be formed to cover the opening. By enclosing the transistor 500 with the aforementioned high barrier insulator, hydrogen and water can be prevented from entering from the outside. Alternatively, multiple transistors 500 can be enclosed by an insulator with high barrier properties against hydrogen or water. Furthermore, when an opening is formed around the transistor 500, for example, an opening can be formed that reaches the insulator 522 or the insulator 514 and the aforementioned high barrier insulator can be formed in contact with the insulator 522 or the insulator 514. Thus, it can also be part of the manufacturing process of the transistor 500. Furthermore, as the insulator with high barrier properties against hydrogen or water, for example, the same material as the insulator 522 or the insulator 514 can be used.

[0345] Notice, Figures 6A to 6C The transistor 500 shown is just an example and is not limited to the structure described above.

[0346] [Capacitor 590] like Figure 4 As shown, a capacitor 590 is disposed above the transistor 500. The capacitor 590 includes a conductor 591 on a conductor 546, an insulator 592 on the conductor 591, and a conductor 593 on the insulator 592. The conductor 591 is used as one of a pair of electrodes (sometimes referred to as the upper electrode), the conductor 593 is used as the other of a pair of electrodes (sometimes referred to as the lower electrode), and the insulator 592 is used as a dielectric. That is, the capacitor 590 constitutes a MIM (Metal-Insulator-Metal) capacitor.

[0347] Alternatively, a conductor 594 may be provided on the conductor 546. The conductor 594 functions as a plug or wiring. The conductor 591 functions as an electrode of the capacitor 590. Furthermore, the conductors 594 and 591 can be formed in the same process.

[0348] As conductors 594 and 591, for example, metal films containing elements selected from molybdenum, titanium, tantalum, tungsten, aluminum, copper, chromium, neodymium, and scandium, or metal nitride films (tantalum nitride film, titanium nitride film, molybdenum nitride film, or tungsten nitride film) composed of the aforementioned elements can be used. Furthermore, indium tin oxide, materials containing tungsten oxide and indium oxide, indium zinc oxide containing tungsten oxide, materials containing titanium oxide and indium oxide, indium tin oxide containing titanium oxide, indium zinc oxide, or indium tin oxide containing silicon oxide can also be used.

[0349] Note that in Figure 4In the capacitor 590 shown, both conductors 594 and 591 have a single-layer structure, but this is not a limitation. For example, both conductors 594 and 591 may have a stacked structure with two or more layers.

[0350] For example, the conductor can be a structure in which a conductor with high adhesion between the two is placed between a conductor that blocks hydrogen and a conductor with high conductivity.

[0351] The insulator 592 is preferably made of a high-k material with a high relative permittivity. By using a high-k material as the insulator 592, the insulator can be thickened to a degree that can suppress gate leakage current and the electrostatic capacitance of the capacitor including the insulator can be adequately ensured.

[0352] For example, as an insulator made of high-k materials, oxides, oxynitrides, oxynitrides, or nitrides containing one or more metallic elements selected from aluminum, hafnium, zirconium, and gallium can be used. Furthermore, silicon may also be included in these materials. Additionally, insulators made of the above-mentioned materials may be laminated.

[0353] In addition, as an insulator made of high-k materials, for example, aluminum oxide, hafnium oxide, zirconium oxide, oxides containing aluminum and hafnium, oxynitrides containing aluminum and hafnium, oxides containing silicon and hafnium, oxynitrides containing silicon and hafnium, oxides containing silicon and zirconium, oxynitrides containing silicon and zirconium, oxides containing hafnium and zirconium, or oxynitrides containing hafnium and zirconium, etc.

[0354] Alternatively, a laminate of insulators made of the aforementioned materials can be used. A laminate structure of a high-k material and a material with a dielectric strength greater than that of the high-k material is preferred.

[0355] As the aforementioned insulator, for example, an insulator sequentially stacked with zirconium oxide, aluminum oxide, and zirconium oxide can be used. Furthermore, for example, an insulator sequentially stacked with zirconium oxide, aluminum oxide, zirconium oxide, and aluminum oxide can be used. Furthermore, for example, an insulator sequentially stacked with hafnium zirconium oxide, aluminum oxide, hafnium zirconium oxide, and aluminum oxide can be used. By stacking an insulator with relatively high dielectric strength, such as aluminum oxide, as the insulator, the dielectric strength can be increased to suppress electrostatic breakdown of the capacitor having this insulator.

[0356] The conductor 593 is provided in such a way that it overlaps the conductor 591 with the insulator 592 in between.

[0357] As a conductor 593, conductive materials such as metallic materials, alloy materials, or metal oxide materials can be used.

[0358] As the conductor 593, a high-melting-point material that combines heat resistance and conductivity is preferably used. Such a material can be, for example, tungsten or molybdenum, with tungsten being particularly preferred. Alternatively, when the conductor 593 is formed in the same process as other conductors, from the viewpoint of low-resistance metal materials, copper or aluminum can also be used.

[0359] An insulator 595 is provided on the conductor 593 and the insulator 592. The insulator 595 can be made of the same material as the insulator 320, for example. Alternatively, the insulator 595 can also be used as a planarization film covering the uneven shape beneath it.

[0360] [Example of a variation of the transistor 500] The transistor that can be used in one aspect of the present invention is not limited to the transistor 500 shown in FIG. 6. For example, it can also be used Figure 7 The transistor 500 is shown in the diagram. Figure 7 The difference between transistor 500 shown and transistor 6 is that: Figure 7 The transistor 500 shown uses an insulator 555; and conductors 542a and 542b have a stacked structure.

[0361] like Figure 7 As shown, conductor 542a has a stacked structure of conductor 542a1 and conductor 542a2 on conductor 542a1. Conductor 542b has a stacked structure of conductor 542b1 and conductor 542b2 on conductor 542b1.

[0362] The conductors 542a1 and 542b1 in contact with the oxide 530b are preferably conductors that are not easily oxidized, such as metal nitrides. This prevents over-oxidation of the conductors 542a and 542b due to oxygen contained in the oxide 530b.

[0363] Furthermore, conductors 542a2 and 542b2 are preferably conductors with higher conductivity than conductors 542a1 and 542b1. Thus, conductors 542a and 542b can be used as highly conductive wiring or electrodes.

[0364] Thus, conductors 542a and 542b, which have the function of wiring or electrodes, can be provided in a manner that contacts the top surface of oxide 530.

[0365] Metal nitrides are preferably used as conductors 542a1 and 542b1. In particular, conductive materials that are not easily oxidized or that maintain conductivity even when absorbing oxygen are preferred.

[0366] As conductors 542a1 and 542b1, nitrides containing tantalum, nitrides containing titanium, nitrides containing molybdenum, nitrides containing tungsten, nitrides containing tantalum and aluminum, or nitrides containing titanium and aluminum are preferably used. In particular, nitrides containing tantalum are preferred. In addition, ruthenium, ruthenium oxide, ruthenium nitride, oxides containing strontium and ruthenium, or oxides containing lanthanum and nickel may also be used, for example.

[0367] The conductivity of conductors 542a2 and 542b2 is preferably higher than that of conductors 542a1 and 542b1. For example, the thickness of conductors 542a2 and 542b2 is preferably greater than that of conductors 542a1 and 542b1. Conductors that can be used in the aforementioned conductor 560b can also be used as conductors 542a2 and 542b2. By adopting the above structure, the resistance value of conductors 542a2 and 542b2 can be reduced.

[0368] For example, tantalum nitride or titanium nitride can be used as conductors 542a1 and 542b1, and tungsten can be used as conductors 542a2 and 542b2.

[0369] In addition, such as Figure 7 As shown, in transistor 500, insulator 555 is formed in such a way that it contacts the side surfaces of openings formed in insulators 580 and 544, and conductors 542a1 and 542b1 are separated using a mask. Here, the openings overlap the region between conductors 542a2 and 542b2. Furthermore, a portion of conductors 542a1 and 542b1 protrudes into the openings. Therefore, insulator 555 contacts the top surface of conductor 542a1, the top surface of conductor 542b1, the side surface of conductor 542a2, and the side surface of conductor 542b2 within the openings. Additionally, insulator 545 contacts the top surface of oxide 530 in the region between conductors 542a1 and 542b1.

[0370] Therefore, when viewed in cross-section along the channel length of transistor 500, the distance between conductors 542a1 and 542b1 is smaller than the distance between conductors 542a2 and 542b2. By employing this structure, the distance between the source and drain can be further shortened, and correspondingly, the channel length can be shortened. Therefore, the frequency characteristics of transistor 500 can be improved. Thus, a semiconductor device with increased operating speed can be provided.

[0371] For example, insulator 555 is preferably an insulator that is not easily oxidized, such as a nitride. Insulator 555 is formed in contact with the sides of conductor 542a2 and conductor 542b2, and functions to protect conductor 542a2 and conductor 542b2. Since insulator 555 is exposed to an oxidizing atmosphere, it is preferable to use an inorganic insulator that is not easily oxidized. Furthermore, since it is in contact with conductor 542a2 and conductor 542b2, it is preferable to use an inorganic insulator that does not easily oxidize conductor 542a2 and conductor 542b2. Therefore, insulator 555 is preferably an insulator that blocks oxygen.

[0372] Silicon nitride can be used as an insulator, for example.

[0373] An opening is formed in insulators 580 and 544, and insulator 555 is formed in such a way that it contacts the side of the opening. A mask is used to separate conductors 542a1 and 542b1, thereby forming Figure 7 The transistor 500 is shown. Here, the opening overlaps the region between conductors 542a1 and 542b2. Furthermore, portions of conductors 542a1 and 542b1 protrude into the opening. Therefore, insulator 555 contacts the top surface of conductor 542a1, the top surface of conductor 542b1, the side surface of conductor 542a2, and the side surface of conductor 542b2 within the opening. Additionally, insulator 545 contacts the top surface of oxide 530 in the region between conductors 542a1 and 542b1.

[0374] Preferably, after separating conductors 542a1 and 542b1, and before depositing insulator 545, heat treatment is performed in an oxygen-containing atmosphere. This supplies oxygen to oxides 530a and 530b, thereby reducing oxygen vacancies. Furthermore, by forming insulator 555 in contact with the sides of conductors 542a2 and 542b2, excessive oxidation of conductors 542a2 and 542b2 can be prevented. This improves the electrical characteristics and reliability of the transistor. In addition, it suppresses the inhomogeneity of electrical characteristics among multiple transistors formed on the same substrate.

[0375] Notice, Figure 4 , Figures 6A to 6C The transistor 500 shown and Figure 7 The transistor 500 shown is just one example and is not limited to this structure. Furthermore, Figure 4 The capacitor 590 shown is just one example and is not limited to this structure.

[0376] [Example of a variation of capacitor 590] The capacitor that can be used in one aspect of the present invention is not limited to Figure 4 The capacitor 590 shown is an example. For instance, capacitors with the following characteristics can also be used: Figure 8 The capacitor 590 is shown in the diagram. Note that... Figure 8 The transistors 550 and 500 shown are Figure 4 Similarly, therefore, only this explanation is provided here. Figure 8 The capacitor shown is 590.

[0377] Figure 8 Insulator 596 on insulator 588, capacitor 590 on insulator 596, insulator 598 on capacitor 590, and insulator 595 on insulator 598 are shown.

[0378] like Figure 8 As shown, an insulator 596 with a conductor 546 embedded in it is provided on an insulator 588. An insulator 597 and conductors 591 and 594 embedded in the insulator 597 are provided on the insulator 596. Furthermore, an insulator 592 is provided on the conductor 591 such that it has a region overlapping with the insulator 597. Furthermore, a conductor 593 is provided on the insulator 592 such that its end is aligned with the end of the insulator 592. Furthermore, an insulator 598 is provided such that it covers the top surface of the insulator 596, the top surface of the conductor 594, the side surface of the insulator 592, the side surface of the conductor 593, and the top surface of the conductor 593. Furthermore, an insulator 595 is provided on the insulator 598.

[0379] Insulators 596 and 598 can also be made of insulating materials that block oxygen and hydrogen. For example, the same materials as insulator 586 can be used for insulators 596 and 598. Insulator 597 can also be made of a material with a relatively low permittivity. This reduces parasitic capacitance generated between wirings. Insulator 597 can also be made of the same material as insulator 588. Furthermore, the area of ​​conductor 591 in contact with insulators 596 and 597 can also be made of a conductive material that blocks oxygen and hydrogen.

[0380] Here, as Figure 8As shown, by embedding a conductor 591, which serves as one of the pairs of electrodes of capacitor 590, into insulator 597, the top surfaces of conductor 591 and insulator 597 can be planarized and their heights aligned. Therefore, an insulator 592, serving as the dielectric of capacitor 590, and a conductor 593, serving as the other of the pairs of electrodes of capacitor 590, can be formed with high flatness on the top surfaces of conductor 591 and insulator 597, which have good flatness. Furthermore, insulator 592 and conductor 593 are arranged to cover conductor 591. By employing this structure, electric field concentration on insulator 592 can be suppressed. Therefore, leakage current between the pairs of electrodes of capacitor 590 (between conductor 591 and conductor 593) can be prevented. Thus, a highly reliable semiconductor device can be realized.

[0381] <Example 2 of transistor and capacitor structure> Other examples of transistor and capacitor structures of the present invention, which differ from the above description, will be described.

[0382] Figures 9A to 9D It is a top view and cross-sectional view of a semiconductor device including transistor 500A and capacitor 590A. Figure 9A This is a top view of the semiconductor device. Furthermore, Figures 9B to 9D This is a cross-sectional view of the semiconductor device. Here, Figure 9B It is along Figure 9A The cross-sectional view shown by the dashed lines A1-A2 is also a cross-sectional view of the channel length of transistor 500A (represented by the X direction). Furthermore, Figure 9C It is along Figure 9A The cross-sectional view shown by the dashed lines A3-A4 is also a cross-sectional view of the channel width direction (represented as the Y direction) of the transistor 500A. Furthermore, Figure 9D It is along Figure 9A The cross-sectional view of the section marked with dotted lines A5-A6 is also the cross-sectional view of the 590A capacitor in the Y direction. Note that in Figure 9A In the top view, some constituent elements are omitted for clarity.

[0383] Figures 9A to 9D The illustrated semiconductor device includes an insulator 514, a transistor 500A and a capacitor 590A on the insulator 514, an insulator 580 disposed on an insulator 544 in the transistor 500A, an insulator 582 disposed on the insulator 580, an insulator 584 disposed on the capacitor 590A and on the insulator 582, and conductors 546 (conductors 546a and 546b). Insulators 514, 580, 582, and 584 are used as interlayer films. Figure 9BAs shown, at least a portion of transistor 500A and capacitor 590A are configured to be embedded in insulator 580.

[0384] Note that insulators 582 and 522 both have the function of capturing or fixing hydrogen. Thus, for example, hydrogen including insulators 580, 524, 545a, and 545b can be captured or fixed in insulators 582 or 522.

[0385] In addition, conductors 546 (conductors 546a and 546b) are used as plugs (also referred to as connection electrodes) by being connected to transistor 500A.

[0386] Conductor 546 is disposed, for example, in an opening 568 formed in an insulator 580 or the like. Conductor 546 has a region that contacts a portion of the top surface and a portion of the side surface of conductor 542a.

[0387] Here, transistor 500A includes an oxide 530 serving as a semiconductor film including a channel formation region, a conductor 560 serving as a first gate (also called a single gate) electrode, a conductor 503 serving as a second gate (also called a back gate) electrode, a conductor 542b serving as one of the source and drain electrodes, and a conductor 542a serving as the other of the source and drain electrodes. Additionally, it includes insulators 545a and 545b serving as a first gate insulating film. Furthermore, it includes insulators 522 and 524 serving as a second gate insulating film.

[0388] The first gate electrode and the first gate insulating film are disposed in the opening 558 formed in the insulator 580 and the insulator 544. That is, the conductor 560, the insulator 545b and the insulator 545a are disposed in the opening 558.

[0389] Capacitor 590A includes a conductor 574 serving as the lower electrode, an insulator 572 serving as the dielectric, and a conductor 570 serving as the upper electrode. In other words, capacitor 590A constitutes a MIM (Metal-Insulator-Metal) capacitor.

[0390] The upper electrode, dielectric, and a portion of the lower electrode of capacitor 590A are disposed in an opening 578 formed in insulators 582, 580, and 544. In other words, conductor 570, insulator 572, and conductor 574 are disposed in the opening 578.

[0391] [Transistor 500A] like Figures 9A to 9CAs shown, transistor 500A includes an insulator 516 on an insulator 514, conductors 503 (conductors 503a and 503b) disposed in the insulator 516, an insulator 522 on the insulator 516 and conductors 503, an insulator 524 on the insulator 522, an oxide 530a on the insulator 524, an oxide 530b on the oxide 530a, and conductors 542a (conductors 542a1 and 542a2) on the oxide 530b. The insulator 542b (conductor 542b1 and conductor 542b2), the insulator 545a on the oxide 530b, the insulator 545b on the insulator 545a, the conductor 560 (conductor 560a and conductor 560b) located on the insulator 545b and overlapping a portion of the oxide 530b, and the insulator 544 disposed on the insulator 522, the insulator 524, the oxide 530a, the oxide 530b, the conductor 542a and the conductor 542b.

[0392] Note that oxides 530a and 530b are sometimes collectively referred to as oxide 530. Additionally, conductors 542a and 542b are sometimes collectively referred to as conductor 542.

[0393] An opening 558 is formed in insulators 580 and 544, reaching oxide 530b. That is, opening 558 can be said to have a region overlapping with oxide 530b. Furthermore, insulator 544 can be said to have an opening overlapping with the opening of insulator 580. In other words, opening 558 has both the opening of insulator 580 and the opening of insulator 544.

[0394] An insulator 545a, an insulator 545b, and a conductor 560 are disposed within the opening 558. Specifically, the conductor 560 has a region overlapping the oxide 530b with the insulators 545a and 545b. Furthermore, along the channel length direction of the transistor 500A, the conductor 560, the insulator 545a, and the insulator 545b are disposed between the conductors 542a and 542b. The insulator 545b has a region that contacts the side surface of the conductor 560 and a region that contacts the bottom surface of the conductor 560.

[0395] Note that, as Figure 9C As shown, in the region of opening 558 that does not overlap with oxide 530, the top surface of insulator 522 is exposed.

[0396] Furthermore, materials with high hydrogen trapping or fixation capabilities can be used as insulator 545a, and materials with high hydrogen barrier properties can be used as insulator 545b. This suppresses the diffusion of impurities such as water and hydrogen into oxide 530. For example, aluminum oxide can be used as insulator 545a, and silicon nitride can be used as insulator 545b.

[0397] Oxide 530 preferably includes oxide 530a on insulator 524 and oxide 530b on oxide 530a. When oxide 530a is included below oxide 530b, the diffusion of impurities from the structure formed below oxide 530a to oxide 530b can be suppressed.

[0398] In transistor 500A, an oxide 530 is shown to have a stacked structure with two layers, oxide 530a and oxide 530b. However, the present invention is not limited to this. For example, oxide 530 may also be a single-layer structure of oxide 530b. Alternatively, a stacked structure of three or more layers may be used. Alternatively, both oxide 530a and oxide 530b may also have a stacked structure.

[0399] Conductor 560 is used as the first gate electrode, and conductor 503 is used as the second gate electrode. Furthermore, insulators 545a and 545b are used as the first gate insulating film, and insulators 522 and 524 are used as the second gate insulating film. Additionally, conductor 542b is used as one of the source and drain electrodes, and conductor 542a is used as the other of the source and drain electrodes. Furthermore, at least a portion of the region of oxide 530 overlapping with conductor 560 is used as a channel forming region.

[0400] [Capacitor 590A] like Figure 9A , Figure 9B and Figure 9D As shown, capacitor 590A includes conductor 574, insulator 572, and conductors 570 (conductors 570a and 570b). Conductor 574 is used as one of a pair of electrodes of capacitor 590A (also called the lower electrode), conductor 570 is used as the other of a pair of electrodes of capacitor 590A (also called the upper electrode), and insulator 572 is used as the dielectric of capacitor 590A.

[0401] At least a portion of conductor 574, insulator 572, conductor 570a, and conductor 570b are disposed in openings 578 provided in insulators 544, 580, and 582. Conductor 574 is disposed on conductor 542b, insulator 572 is disposed on conductor 574, conductor 570a is disposed on insulator 572, and conductor 570b is disposed on conductor 570a.

[0402] The conductor 574 is disposed along an opening 578 formed in insulators 544, 580, and 582. The height of a portion of the top surface of the conductor 574 is preferably the height of the top surface of the insulator 582. Furthermore, the bottom surface of the conductor 574 contacts the top surface of the conductor 542b.

[0403] Conductor 574 is preferably deposited using a deposition method with good coverage, such as ALD or CVD.

[0404] As conductor 574, materials suitable for conductors 503, 560, or 542 described above can also be used. For example, by using the same conductive material as conductor 542b as conductor 574, the contact resistance between conductor 574 and conductor 542b can be reduced. As conductor 574, titanium nitride or tantalum nitride deposited using the ALD method can be used, for example.

[0405] Insulator 572 is configured to cover a portion of conductor 574 and insulator 582.

[0406] Insulator 572 is preferably deposited by a high-coverage deposition method such as ALD or CVD.

[0407] A material with a high relative permittivity (high-k) is preferably used as insulator 572. For example, a material suitable for the aforementioned insulator 592 can also be used as insulator 572.

[0408] The conductor 570 is disposed in an opening 578 formed in the insulator 544, the insulator 580 and the insulator 582.

[0409] The conductor 570 is preferably deposited using methods such as ALD or CVD.

[0410] As conductor 570, materials suitable for conductors 503 or 560 described above can also be used. For example, titanium nitride deposited using the ALD method can be used as conductor 570a, and tungsten deposited using the CVD method can be used as conductor 570b. Furthermore, when the adhesion of tungsten to insulator 572 is sufficiently high, a monolayer of tungsten deposited using the CVD method can also be used as conductor 570.

[0411] The opening 578 is provided in such a way that it reaches the conductor 542b. That is to say, the opening 578 can be said to have an area that overlaps with the conductor 542b. The conductor 542b is one of the source electrode and drain electrode of the transistor 500A, and the transistor 500A can be connected to the capacitor 590A by contacting the bottom surface of the conductor 574 provided in the opening 578.

[0412] Preferably, the opening 578 is close to the oxide 530 when viewed from above. This structure reduces the area occupied by the capacitor 590A and the transistor 500A. Furthermore, the shape of the opening 578 when viewed from above can be a square, a polygon other than a square, a polygon with curved corners, or a circular shape including ellipses.

[0413] like Figure 9B and Figure 9D As shown, a conductor 574 is disposed along the opening 578. Therefore, the conductor 574 is in contact with the side surfaces of the insulators 544, 580, and 582, the side surface of the conductor 542b1, the side surface and top surface of the conductor 542b2, and the top surface of the insulator 522. Furthermore, an insulator 572 is disposed in contact with the top surface of the conductor 574, a conductor 570a is disposed in contact with the top surface of the insulator 572, and a conductor 570b is disposed in contact with the top surface of the conductor 570a.

[0414] Capacitor 590A has the above structure, such as Figure 9B and Figure 9D As shown, a capacitor 590A can be formed in the opening 578 such that the conductor 574 and the conductor 570 are positioned opposite each other with an insulator 572 in between. Therefore, by increasing the depth of the opening 578 (or the thickness of the insulator 580), the electrostatic capacitance of the capacitor 590A can be increased.

[0415] like Figure 9B As shown, a portion of conductor 574, a portion of insulator 572, and a portion of conductor 570 are disposed in such a way that they are exposed from opening 578. In other words, a portion of conductor 574, a portion of insulator 572, and a portion of conductor 570 are formed above the top surface of conductor 560 or above the top surface of insulator 582.

[0416] A portion of the conductor 574 and a portion of the insulator 572 are in contact with the top surface of the insulator 582. That is, the side end of the conductor 574 is covered by the insulator 572. Furthermore, the conductor 570 preferably has an area where it overlaps with the insulator 582, separated by the insulator 572. Here, as... Figure 9B As shown, the side end of conductor 570 is aligned with the side end of insulator 572. By adopting this structure, conductor 570 and conductor 574 can be separated by insulator 572, thus suppressing short circuits between conductor 570 and conductor 574.

[0417] Alternatively, the portion above the insulator 582 of the conductor 570 can be wound into a wiring shape. For example, as... Figure 9CAs shown, the conductor 570 can be extended along the channel width direction of the transistor 500A. Therefore, when multiple transistors 500A and capacitors 590A are provided, the conductor 570 can also be used as wiring. Furthermore, the conductor 570 can also be extended together with the insulator 572.

[0418] [Examples of variations of transistor 500A and capacitor 590A] Figure 10 This is an explanation Figure 9B Cross-sectional views of other structural examples of the semiconductor device shown, including transistor 500A and capacitor 590A. Here, the main focus is on... Figure 10 The semiconductor device shown is Figure 9B The differences shown are in the semiconductor devices.

[0419] Figure 10 The semiconductor device shown is Figure 9B The difference in the semiconductor device shown is that... Figure 10 The semiconductor device shown is formed with the side ends of insulator 524, oxide 530a, oxide 530b, conductor 542a1 and conductor 542a2 aligned.

[0420] also, Figure 10 The semiconductor device shown, in addition to Figure 9B In addition to the semiconductor device shown, it also includes insulators 552a, 552b, 554, 556, 521, 583a, and 583b.

[0421] like Figure 10 As shown, insulator 552a is disposed between conductor 542a2 and insulator 544. Furthermore, insulator 552b is disposed between conductor 542b2 and insulator 544. Here, insulators 552a and 552b may also function as etch stop layers protecting conductors 542a2 and 542b2 during simultaneous processing of insulator 524, oxide 530a, oxide 530b, conductor 542a1, conductor 542a2, insulator 552a, and insulator 552b.

[0422] like Figure 10As shown, in the opening 558, the insulator 554 is disposed between the insulator 545a and the conductors 542a2, 542b2, 552a, 552b, 544, and 580, and in contact with a portion of the top surface of the conductors 542a1 and 542b1. In other words, the insulator 554 can be described as being formed as a sidewall in contact with the side of the opening 558. Here, the insulator 554 may also have the function of providing a protective film to prevent the conductors 542a2 and 542b2 from being over-oxidized when heat treatment is performed in an oxygen-containing atmosphere after the conductors 542a1 and 542b1 have been separated.

[0423] like Figure 10 As shown, insulator 521 is disposed in contact with the bottom surface of insulator 522. Furthermore, insulator 583a is disposed in contact with the top surface of insulator 582. Here, insulator 521 may also have the function of suppressing the diffusion of impurities such as water and hydrogen from the interlayer insulator disposed below insulator 521 to transistor 500A. Furthermore, insulator 583a may also have the function of suppressing the diffusion of impurities such as water and hydrogen from the interlayer insulator disposed above insulator 583a to transistor 500A.

[0424] In addition, such as Figure 10 As shown, insulator 583b is disposed between insulator 583a and insulator 584. Here, insulator 583b has the function of changing the electrostatic capacitance of capacitor 590A according to its thickness. That is, by increasing the depth of opening 578 in capacitor 590A (for example, by increasing the thickness of insulator 583b), the electrostatic capacitance of capacitor 590A can be increased.

[0425] In addition, such as Figure 10 As shown, insulator 556 is disposed in contact with the side of conductor 546. Specifically, insulator 556 is disposed with respect to the openings of insulators 516, 521, 522, 544, 580, 582, 583a, 583b, and 584 (equivalent to...). Figure 9B The opening 568 shown is formed in a manner that makes contact with the side of the opening. Furthermore, an insulator 556 is formed in a protruding manner within this opening. Additionally, an insulator 556 is also formed on the side of the insulator 524, oxide 530, and conductor 542a. Here, at least a portion of the conductor 542a protrudes from the insulator 556 and contacts the conductor 542a. That is, the conductor 546 is formed in a manner that embeds itself within the opening through the insulator 556.

[0426] The uppermost part of the insulator 556 formed below the conductor 542a is preferably located below the top surface of the conductor 542a. With this structure, the conductor 546 can contact at least a portion of the side end of the conductor 542a. Furthermore, the insulator 556 formed below the conductor 542a preferably has a region that contacts the side surface of the oxide 530. With this structure, for example, impurities such as water or hydrogen contained in the insulator 580 can be suppressed from mixing into the oxide 530 through the conductor 546. Furthermore, the insulator 556 may also have a stacked structure of two or more layers.

[0427] Notice, Figures 9A to 9D The transistor 500A and capacitor 590A shown are as follows: Figure 10 The transistor 500A and capacitor 590A shown are just examples and are not limited to this structure.

[0428] <Example 3 of transistor and capacitor structure> Other examples of transistor and capacitor structures of the present invention, which differ from the above description, will be described.

[0429] Figures 11A to 11D It is a top view, a three-dimensional schematic diagram and a cross-sectional view of a semiconductor device including a transistor 600 and a capacitor 690. Figure 11A This is a top view of the semiconductor device. Additionally, Figure 11B This is a three-dimensional schematic diagram of the semiconductor device. Figure 11C and Figure 11D This is a cross-sectional view of the semiconductor device. Here, Figure 11C It is along Figure 11A A cross-sectional view of the section marked with dashed lines A1-A2. Furthermore, Figure 11D It is along Figure 11A The cross-sectional view of the section marked with dashed lines A3-A4. Note that in... Figure 11A Top view and Figure 11B In the three-dimensional diagram, some constituent elements are omitted for clarity.

[0430] Figures 11A to 11D The diagram shows an insulator 612, a conductor 610 on the insulator 612, a transistor 600 and a capacitor 690 on the conductor 610, an insulator 620 on the conductor 610, an insulator 640 on the insulator 620, and an insulator 678 on the transistor 600 and the capacitor 690. Insulators 612, 620, 640, and 678 are used as interlayer films. The conductor 610 is used as wiring.

[0431] like Figures 11A to 11DAs shown, in transistor 600, oxide 650, serving as a semiconductor including a channel formation region, is disposed along an opening 648 provided in insulator 640. Furthermore, in capacitor 690, insulator 632, serving as a dielectric, is disposed along an opening 628 provided in insulator 620. Moreover, transistor 600 is disposed in an overlapping manner with capacitor 690. Furthermore, the opening 648, which is a part of the structure of transistor 600, has a region overlapping with the opening 628, which is a part of the structure of capacitor 690. In particular, conductor 630 is used as one of the source and drain electrodes of transistor 600 and as one of a pair of electrodes of capacitor 690, so transistor 600 and capacitor 690 share a portion of the structure. By employing this structure, transistor 600 and capacitor 690 can be disposed in a manner that does not significantly increase the occupied area when viewed from above.

[0432] Therefore, for example, when transistor 600 and capacitor 690 are used in a storage device, they can also be used in the storage cells included in that storage device. Furthermore, for example, in the storage device 100 shown in Embodiment 1 above, transistors and capacitors that constitute the backup circuit 130 included in the unit storage circuit 110 can also be used.

[0433] [Capacitor 690] Capacitor 690 includes a conductor 634 on conductor 610, an insulator 632 on conductor 634, and a conductor 630 on insulator 632. Conductor 630 is used as one of a pair of electrodes (sometimes referred to as the upper electrode), conductor 634 is used as the other of a pair of electrodes (sometimes referred to as the lower electrode), and insulator 632 is used as the dielectric. That is, capacitor 690 constitutes a MIM (Metal-Insulator-Metal) capacitor.

[0434] like Figure 11C and Figure 11D As shown, an opening 628 is provided in the insulator 620 to reach the conductor 610. At least a portion of the conductor 634 is disposed in the opening 628. Note that the conductor 634 has a region in the opening 628 that contacts the top surface of the conductor 610, a region in the opening 628 that contacts the side surface of the insulator 620, and a region that contacts at least a portion of the top surface of the insulator 620. The insulator 632 is disposed such that at least a portion of it is located in the opening 628. The conductor 630 is disposed such that at least a portion of it is located in the opening 628. Furthermore, as... Figure 11C and Figure 11D As shown, the conductor 630 is preferably provided in a manner that embeds the opening 628.

[0435] The capacitor 690 has a structure in which the upper and lower electrodes are separated by a dielectric material not only at the bottom but also on the side of the opening 628, thus increasing the electrostatic capacitance per unit area. Therefore, the deeper the opening 628, the larger the electrostatic capacitance of the capacitor 690 can be.

[0436] The side of the opening 628 (sometimes referred to as the side of the opening 628 in the insulator 620) is preferably perpendicular to the top surface of the conductor 610. In other words, the insulator 620 may also be said to include an opening 628 extending in a direction perpendicular to the top surface of the conductor 610. In this case, the opening 628 has a cylindrical shape.

[0437] Note that this embodiment shows an example where the opening 628 is circular in top view, but the invention is not limited to this. For example, the shape of the opening 628 in top view can also be an ellipse or a roughly circular shape, a quadrilateral or a polygon, or a shape with rounded corners. In this case, the maximum width of the opening 628 can be appropriately calculated based on the top view shape of the uppermost part of the opening 628.

[0438] For example, when the opening 628 is a quadrilateral when viewed from above, the maximum width of the opening 628 can also be the length of the diagonal of the quadrilateral. Alternatively, for example, when the opening 628 has a generally circular shape such as an ellipse, a polygonal shape, or a shape that makes the corners of the polygonal shape rounded when viewed from above, the maximum width of the opening 628 can also be the maximum width of the shape when viewed from above.

[0439] The portion of the conductor 634, insulator 632, and conductor 630 disposed in the opening 628 reflects the shape of the opening 628. Therefore, the conductor 634 is disposed along the opening 628, the insulator 632 is disposed in such a way that it covers the conductor 634, and the conductor 630 is disposed in such a way that it is embedded in the recess of the insulator 632 that reflects the shape of the opening 628.

[0440] In other words, a portion of the dielectric (equivalent to the insulator 632) of the capacitor 690 is disposed along the side of the opening 628. In other words, it is disposed in a direction perpendicular to the top surface of the conductor 610. That is to say, both the surface of the upper electrode of the capacitor 690 that contacts the dielectric and the surface of the lower electrode that contacts the dielectric have a component in a direction perpendicular to the top surface of the conductor 610.

[0441] Note that in Figure 11C and Figure 11D The opening 628 is provided such that the side of the opening 628 is perpendicular to the top surface of the conductor 610, but one aspect of the invention is not limited to this. For example, the side of the opening 628 may also have a tapered shape.

[0442] In this specification, a conical shape refers to a shape in which at least a portion of the side surface of a constituent element is inclined relative to the substrate surface. Furthermore, the angle formed by the inclined side surface and the substrate surface is called a conical angle. In particular, in this specification, a conical shape having a conical angle greater than 0° and less than 90° is sometimes referred to as a positive conical shape, and a conical shape having a conical angle greater than 90° and less than 180° is sometimes referred to as a negative conical shape.

[0443] A conductor 634 and an insulator 632 are stacked along the side of the opening 628 and the top surface of the conductor 610. Furthermore, a conductor 630 is provided on the insulator 632 in a manner that embeds into the opening 628. In this specification and the like, the capacitor 690 having this structure is sometimes referred to as a trench capacitor, a trench capacitor, or a deep-hole multilayer capacitor, etc.

[0444] An insulator 640 is disposed on the capacitor 690. That is, the insulator 640 is disposed above the conductor 634, the insulator 632, and the conductor 630. In other words, the conductor 630 is disposed below the insulator 640.

[0445] A conductor 610 is disposed below the conductor 634. The conductor 634 has an area that contacts the conductor 610.

[0446] The conductor 610 is disposed on the insulator 612. The conductor 610 may be, for example, a planar shape.

[0447] As insulator 612, for example, a material that can be used for the aforementioned insulator 514 can also be used.

[0448] The conductor 610 is preferably made of a conductive material with high conductivity. Furthermore, the conductor 610 can have a single-layer structure or a structure in which different materials are stacked. For example, materials suitable for the conductors 503 and 560 described above can be used as the conductor 610. For example, tungsten can be used.

[0449] Furthermore, as the conductor 634, it is preferable to use a conductive material that is not easily oxidized, either as a single layer or in a multilayer, or a conductive material that has the function of inhibiting oxygen diffusion. Therefore, when an oxide insulator is used as the insulator 632, oxidation of the conductor 634 due to the insulator 632 can be suppressed. Furthermore, when an oxide insulator is used as the insulator 620, oxidation of the conductor 634 due to the insulator 620 can be suppressed.

[0450] As conductor 634, materials suitable for conductors 503 or 560 described above can also be used. For example, titanium nitride or indium tin oxide with added silicon can also be used. Furthermore, for example, a structure in which titanium nitride is stacked on tungsten can also be used. Alternatively, for example, a structure in which tungsten is stacked on a first titanium nitride and a second titanium nitride is stacked on the tungsten can also be used.

[0451] An insulator 632 is disposed on the conductor 634. The insulator 632 is disposed in contact with the top and side surfaces of the conductor 634. That is, the insulator 632 preferably covers the side ends of the conductor 634. This prevents a short circuit between the conductor 634 and the conductor 630.

[0452] In addition, such as Figure 11C and Figure 11D As shown, insulator 632 can also extend in contact with the top surface of insulator 620.

[0453] Alternatively, a structure can be adopted in which the side ends of the insulator 632 are aligned with the side ends of the conductor 634. By adopting this structure, the insulator 632 and the conductor 634 can be formed using the same mask, thereby simplifying the manufacturing process.

[0454] A high-k material is preferably used as the insulator 632. By using a high-k material as the insulator 632, the thickness of the insulator 632 can be increased to a level that can suppress gate leakage current and sufficiently ensure the electrostatic capacitance of the capacitor 690.

[0455] As insulator 632, for example, materials that can be used for insulator 592 or insulator 572 described above can also be used.

[0456] The conductor 630 is disposed in contact with a portion of the top surface of the insulator 632. Furthermore, the side end of the conductor 630 is preferably located inside the side end of the conductor 634 in both the X and Y directions. Note that in a structure where the insulator 632 covers the side end of the conductor 634, the side end of the conductor 630 may also be located outside the side end of the conductor 634.

[0457] The conductor 630 can be a single layer or a stack of the aforementioned conductive materials. Preferably, the conductor 630 is a conductive material that is not easily oxidized or has the function of inhibiting oxygen diffusion.

[0458] As conductor 630, materials suitable for conductors 503, 560, or 542 described above can also be used. For example, titanium nitride or tantalum nitride can be used.

[0459] Insulator 620 is used as an interlayer film, so its relative permittivity is preferably low. By using a material with a low relative permittivity in the interlayer film, parasitic capacitance generated between wirings can be reduced. Insulator 620 can be a single layer or a stack of insulators containing a material with a low relative permittivity.

[0460] As insulator 620, materials suitable for the aforementioned insulator 516 can also be used, for example. From the viewpoint of thermal stability, silicon oxide or silicon oxynitride is preferred.

[0461] [Transistor 600] Transistor 600 includes a conductor 630, a conductor 660 on an insulator 640, an oxide 650, an insulator 672 on the oxide 650, and a conductor 670 on the insulator 672. The oxide 650 is used as a semiconductor film including a channel formation region, the conductor 670 is used as a gate electrode, the insulator 672 is used as a gate insulating film, the conductor 630 is used as one of the source and drain electrodes, and the conductor 660 is used as the other of the source and drain electrodes.

[0462] In transistor 600, a metal oxide used as an oxide semiconductor is used as oxide 650, which includes the channel formation region. For example, a metal oxide that can be used for oxide 530 described above can also be used as oxide 650.

[0463] like Figure 11C and Figure 11D As shown, an opening 648 is provided in the insulator 640 and the conductor 660, leading to the conductor 630. At least a portion of the oxide 650 is disposed in the opening 648. Note that the oxide 650 has a region in the opening 648 that contacts the top surface of the conductor 630, a region in the opening 648 that contacts the side surface of the conductor 660, and a region that contacts at least a portion of the top surface of the conductor 660. The insulator 672 is disposed such that at least a portion of it is located in the opening 648. The conductor 670 is disposed such that at least a portion of it is located in the opening 648. Furthermore, as... Figure 11C and Figure 11D As shown, the conductor 670 is preferably provided in a manner that embeds the opening 648.

[0464] As the conductor 630, a structure in which tantalum nitride is stacked on titanium nitride can be used, for example. In this case, the structure can be used such that the titanium nitride is in contact with the insulator 632 and the tantalum nitride is in contact with the oxide 650. By using this structure, excessive oxidation of the conductor 630 due to the oxide 650 can be suppressed. Furthermore, when an oxide insulator is used as the insulator 632, excessive oxidation of the conductor 630 due to the insulator 632 can be suppressed. Alternatively, as the conductor 630, a structure in which tungsten is stacked on titanium nitride can also be used, for example.

[0465] Furthermore, since the conductor 630 has a region in contact with the oxide 650, it is preferable to use a conductive material containing oxygen. With this structure, conductivity can be maintained even if the conductor 630 absorbs oxygen. Additionally, if the insulator 632 uses a material containing oxygen, the conductivity of the conductor 630 can also be maintained.

[0466] As a conductor 630, indium tin oxide (also known as ITO), indium tin oxide containing silicon oxide (also known as ITSO), indium zinc oxide (also known as IZO (registered trademark)) and the like can be used, for example, in a single layer or in a stack.

[0467] The oxide 650 has a region in the opening 648 that contacts the side surface of the conductor 660 and a region that contacts a portion of the top surface of the conductor 660. Thus, by having the oxide 650 contact the top surface of the conductor 660 in addition to its side surface, the contact area between the oxide 650 and the conductor 660 can be increased.

[0468] also, Figure 11D The diagram shows a structure where the side end of oxide 650 is located inside the side end of conductor 660. Note that the invention is not limited to this. For example, a structure in which the side end of oxide 650 is aligned with the side end of conductor 660 in the Y direction may also be used. Alternatively, a structure in which the side end of oxide 650 is located outside the side end of conductor 660 may also be used.

[0469] In addition, such as Figures 11A to 11D As shown, preferably, conductor 670 extends in the Y direction and conductor 660 extends in the X direction. By employing this structure, conductors 670 and 660 intersect each other. Furthermore, in Figure 11A The conductor 610 is arranged in a planar shape, but the present invention is not limited to this. For example, the conductor 610 may be parallel to either the conductor 670 or the conductor 660.

[0470] The side of the opening 648 (sometimes referred to as the side of the opening 648 in the insulator 640) is preferably perpendicular to the top surface of the conductor 610. In other words, the insulator 640 can also be said to include an opening 648 extending in a direction perpendicular to the top surface of the conductor 610. In this case, the opening 648 has a cylindrical shape.

[0471] Note that this embodiment shows an example where the opening 648 is circular in top view, but the invention is not limited to this. For example, the opening 648 in top view can also be an ellipse or a roughly circular shape, a quadrilateral or a polygon, or a shape with rounded corners. In this case, the maximum width of the opening 648 can be appropriately calculated based on the top view shape of the uppermost part of the opening 648.

[0472] For example, when the opening 648 is a quadrilateral when viewed from above, the maximum width of the opening 648 can also be the length of the diagonal of the quadrilateral. Alternatively, for example, when the opening 648 has a generally circular shape such as an ellipse, a polygonal shape, or a shape that makes the corners of the polygonal shape rounded when viewed from above, the maximum width of the opening 648 can also be the maximum width of the shape when viewed from above.

[0473] The arrangement of oxide 650, insulator 672, and conductor 670 in the portion of opening 648 reflects the shape of opening 648. Therefore, oxide 650 is provided along opening 648, insulator 672 is provided in a manner that covers oxide 650, and conductor 670 is provided in a manner that embeds into a recess of insulator 672 that reflects the shape of opening 648.

[0474] In other words, a portion of the semiconductor film (equivalent to oxide 650) of transistor 600, including the channel formation region, is disposed along the side of opening 648. In other words, it is disposed in a direction perpendicular to the top surface of conductor 610. That is to say, the channel length direction of transistor 600 has a directional component perpendicular to the top surface of conductor 610. That is to say, the channel length direction has a longitudinal direction (in... Figures 11A to 11D The component in the Z direction (also known as the height direction or the direction perpendicular to the surface being formed) is the source electrode and drain electrode located at different heights, with the drain current flowing in the longitudinal direction. Therefore, one aspect of the transistor of the present invention is a transistor whose channel length direction has a longitudinal component (i.e., a transistor with drain current flowing in the longitudinal direction), which may be referred to as a VFET (Vertical Field Effect Transistor), a vertical transistor, a vertical channel transistor, etc.

[0475] Here, the conductor 660 can be, for example, a material suitable for conductor 630. Furthermore, the conductor 670 can be, for example, a material suitable for conductor 630. Furthermore, the insulator 672 can be, for example, a material suitable for insulators 522, 524, or 545 described above. Furthermore, the insulator 640 can also be, for example, a material suitable for insulator 620. Additionally, the insulator 678 can also be, for example, a material suitable for insulator 514 or insulator 612 described above.

[0476] Note that in Figure 11C and Figure 11D The opening 648 is provided such that the side of the opening 648 is perpendicular to the top surface of the conductor 610, but the present invention is not limited thereto. For example, the side of the opening 648 may also have a tapered shape.

[0477] In addition, Figure 11C and Figure 11D The oxide 650 is shown as a single-layer structure, but this is not the only aspect of the invention. The oxide 650 may also have a stacked structure of multiple oxide layers with different chemical compositions.

[0478] Here, Figure 12A Show Figure 11C A magnified view of oxide 650 and its vicinity. Furthermore, Figure 12B A cross-sectional view of the XY plane including conductor 660 is shown.

[0479] like Figure 12A As shown, oxide 650 has region 650i, region 650na and region 650nb arranged in a manner that clamps region 650i.

[0480] Region 650na is the region of oxide 650 in contact with conductor 630. At least a portion of region 650na is used as one of the source and drain regions of transistor 600. Region 650nb is the region of oxide 650 in contact with conductor 660. At least a portion of region 650nb is used as the other of the source and drain regions of transistor 600. Figure 12B As shown, the conductor 660 is in contact with the entire outer periphery of the oxide 650. Therefore, another portion of the source and drain regions of the transistor 600 may be formed on the outer periphery of the portion of the oxide 650 formed in the same layer as the conductor 660.

[0481] Region 650i is the region between region 650na and region 650nb of oxide 650. At least a portion of region 650i is used as the channel forming region of transistor 600. That is, the channel forming region of transistor 600 is located in the region between conductor 630 and conductor 660 of oxide 650. Alternatively, the channel forming region of transistor 600 can also be described as being located in or near the region of oxide 650 that contacts insulator 640.

[0482] The channel length of transistor 600 is the distance between the source and drain regions. In other words, the channel length of transistor 600 is determined by the thickness of the insulator 640 on the conductor 630. Figure 12A In the diagram, the channel length Lch of transistor 600 is indicated by a dashed double arrow. In cross-section, the channel length Lch is the distance between the end of the region where oxide 650 and conductor 630 contact and the end of the region where oxide 650 and conductor 660 contact. That is, the channel length Lch is equivalent to the length of the side of the insulator 640 on the side of the opening 648 in cross-section.

[0483] In planar transistors, the channel length is limited by the exposure limit of photolithography, making further miniaturization difficult. However, in this invention, the channel length can be set according to the thickness of the insulator 640. Therefore, the channel length of the transistor 600 can be set to a very fine structure below the exposure limit of photolithography (e.g., below 60 nm, 50 nm, 40 nm, 30 nm, 20 nm, or 10 nm and above 1 nm or 5 nm and above). This increases the on-state current of the transistor 600, thereby improving its frequency characteristics.

[0484] Furthermore, as described above, a channel formation region, a source region, and a drain region can be formed within the opening 648. Therefore, compared to a planar transistor where the channel formation region, source region, and drain region are respectively provided on the XY plane, the occupied area of ​​the transistor 600 can be reduced.

[0485] In addition, with Figure 12B Similarly, on the XY plane including the channel formation region of oxide 650, oxide 650, insulator 672, and conductor 670 are also arranged in a concentric circle. Therefore, the side of the conductor 670 located at the center faces the side of oxide 650 across the insulator 672. In other words, when viewed from above, the entire outer periphery of oxide 650 forms the channel formation region. At this time, for example, the channel width of transistor 600 is determined by the length of the outer periphery of oxide 650. That is to say, the channel width of transistor 600 is determined by the size of the maximum width of opening 648 (the maximum diameter if the shape of opening 648 is circular when viewed from above). Figure 12A and Figure 12B In the diagram, a double-headed arrow with a double-dotted line represents the maximum width Dia of the opening 648. Figure 12B In the diagram, the channel width Wch of transistor 600 is represented by a double-headed dotted arrow. By increasing the maximum width Dia of opening 648, the channel width per unit area can be increased, thereby increasing the on-state current.

[0486] When forming the opening 648 using photolithography, the maximum width Dia of the opening 648 is limited by the exposure limit of the photolithography, making further miniaturization difficult. Furthermore, the maximum width Dia of the opening 648 is set based on the thickness of the oxide 650, insulator 672, and conductor 670 disposed within the opening 648. The maximum width Dia of the opening 648 is preferably, for example, 5 nm or more, 10 nm or more, or 20 nm or more but less than 100 nm, less than 60 nm, less than 50 nm, less than 40 nm, or less than 30 nm. Note that when the shape of the opening 648 in top view is circular, the maximum width Dia of the opening 648 is equivalent to the diameter of the opening 648, and the channel width Wch can be calculated as "Dia × π".

[0487] Furthermore, in one embodiment of the present invention, the channel length Lch of the transistor 600 is preferably at least less than the channel width Wch of the transistor 600. The channel length Lch of the transistor 600 in one embodiment of the present invention is at least 0.1 times and less than 0.99 times the channel width Wch of the transistor 600, preferably at least 0.5 times and less than 0.8 times. By employing this structure, a transistor with good electrical characteristics and high reliability can be realized.

[0488] Furthermore, by forming the opening 648 in a circular shape when viewed from above, the oxide 650, insulator 672, and conductor 670 are arranged in a concentric circle. As a result, the distance between the conductor 670 and the oxide 650 is uniform, so a gate electric field can be uniformly applied to the oxide 650.

[0489] Notice, Figures 11A to 11D The transistor 600 and capacitor 690 shown are just examples and are not limited to this structure.

[0490] [Transistor 600B] In one embodiment of the present invention, the transistor 600 described above may also include a back gate.

[0491] As one aspect of the present invention Figure 13 This is a cross-sectional view illustrating an example of a transistor structure with a back gate. Figure 13 The transistor 600B shown is a variation of the transistor 600 described above. Transistor 600B includes two gates (gate and back gate) corresponding to the channel forming regions.

[0492] The difference between transistor 600B and transistor 600 is that transistor 600B includes a conductor 680 and an insulator 682. In transistor 600B, an insulator 682 is disposed between the insulator 640 and the oxide 650 on the side of the opening 648, and a conductor 680 is disposed in a portion of the insulator 640 in a manner that surrounds the outer periphery of the oxide 650 with the insulator 682 in between.

[0493] Here, conductor 670 is sometimes used as the first gate (also simply referred to as the gate) electrode, and conductor 680 is sometimes used as the second gate (also referred to as the back gate) electrode. In this case, insulator 672 is used as the first gate insulating film, and insulator 682 is used as the second gate insulating film.

[0494] The difference between transistor 600B and transistor 600 is that in transistor 600B, the conductor 630 has a recess at the position overlapping with the opening 648. In transistor 600B, a portion of oxide 650 and a portion of insulator 682 are disposed in the recess of the conductor 630. At this time, the bottom surface of oxide 650 is located below the bottom surface of insulator 682.

[0495] By adopting this structure, the contact area between the oxide 650 and the conductor 630 can be increased. Therefore, the contact resistance between the oxide 650 and the conductor 630 can be reduced.

[0496] Here, as conductor 680, for example, a material suitable for conductor 670 can also be used. Furthermore, as insulator 682, for example, a material suitable for insulator 672 can also be used.

[0497] Here, in a transistor with a back gate, the threshold voltage drifts according to the back gate voltage. Furthermore, the back gate of this transistor can be connected to the gate, or to one of the source and drain, or the other of the source and drain.

[0498] Figure 14A This is a cross-sectional view showing an example of the structure in which the back gate (equivalent to conductor 680) of transistor 600B is connected to the gate (equivalent to conductor 670) of transistor 600B through conductor 684 embedded in insulator 672 and insulator 640. Figure 14B This is a cross-sectional view showing an example of a structure in which the back gate (equivalent to conductor 680) of transistor 600B is connected to another (equivalent to conductor 660) of the source and drain of transistor 600B through conductor 684 embedded in insulator 640. Figure 14CThis is a cross-sectional view showing an example of a structure in which the back gate (equivalent to conductor 680) of transistor 600B is connected to one of the source and drain terminals (equivalent to conductor 630) of transistor 600B via conductor 684 embedded in insulator 640.

[0499] In addition, conductor 684 has the function of plug or wiring.

[0500] [Examples of variations of transistor 600 and capacitor 690] The transistors and capacitors that can be used in one aspect of the present invention are not limited to those described above. Figures 11A to 11D The transistor 600 and capacitor 690 are shown. For example, a transistor with... Figure 15A The transistor 600 with the structure shown and having Figure 15B The capacitor 690 is shown in the diagram.

[0501] exist Figure 15A and Figure 15B In the diagram, a stacked structure of conductor 630_1 and conductor 630_2 on conductor 630_1 is shown as conductor 630. Conductor 630_1 is used as one of a pair of electrodes of capacitor 690, and conductor 630_2 is used as one of the source electrode and drain electrode of transistor 600.

[0502] Furthermore, a stacked structure of conductor 660_1 and conductor 660_2 on conductor 660_1 is shown as conductor 660. For example, conductor 660_1 can be used as wiring, and conductor 660_2 can be used as another of the source and drain electrodes of transistor 600.

[0503] exist Figure 15A In the transistor 600 shown, the conductor 630_2 has a recess at a position overlapping with the opening 648. A portion of the oxide 650, a portion of the insulator 672, and a portion of the conductor 670 are disposed in the recess of the conductor 630_2. At this time, the bottom surface of the conductor 670 in the recess can be located below the top surface of the conductor 630_2 outside the recess.

[0504] By providing oxide 650 in the recess of conductor 630_2, the contact area between oxide 650 and conductor 630_2 can be increased. Therefore, the contact resistance between oxide 650 and conductor 630_2 can be reduced.

[0505] Furthermore, by reducing the height of the bottom surface of conductor 670, a gate electric field can be easily applied to the channel formation region of oxide 650. This improves the electrical characteristics of transistor 600. Additionally, the region of oxide 650 in contact with conductor 630_2 can also easily have a gate electric field applied. This increases the on-state current of transistor 600. Furthermore, regardless of whether conductor 630 or conductor 660 is used as the drain electrode, transistor 600 can exhibit good electrical characteristics.

[0506] Here, as Figure 15A As shown, oxide 650 may also have a region 650p with a corner bend within the recess of conductor 630_2. Therefore, for example, compared to the case where region 650p is a right angle or acute angle (having a corner), electric field concentration near insulator 672 can be suppressed. Thus, by suppressing electric field concentration on insulator 672, insulation breakdown of insulator 672 can be suppressed, thereby providing a highly reliable semiconductor device.

[0507] exist Figure 15B In the capacitor 690 shown, the conductor 610 has a recess at a position overlapping with the opening 628, and a portion of the conductor 634, a portion of the insulator 632, and a portion of the conductor 630_1 are disposed in the recess of the conductor 610. At this time, the bottom surface of the conductor 630_1 in the recess can be located below the top surface of the conductor 610 outside the recess.

[0508] By placing the conductor 634 in the recess of the conductor 610, the contact area between the conductor 634 and the conductor 610 can be increased. Therefore, the contact resistance between the conductor 634 and the conductor 610 can be reduced.

[0509] Here, as Figure 15B As shown, the conductor 634 may also have a region 634p with a corner bend within the recess of the conductor 610. Therefore, for example, compared to the case where region 634p is a right angle or an acute angle (having a corner), the electric field concentration near the insulator 632 can be suppressed. Furthermore, the end 634q of the conductor 634 may also be located below the top surface of the insulator 620. Therefore, compared to the case where the end 634q is located on the insulator 620, the electric field concentration near the insulator 632 can be suppressed. Thus, by suppressing the electric field concentration on the insulator 632, insulation breakdown of the insulator 632 can be suppressed, thereby providing a highly reliable semiconductor device.

[0510] <Example 4 of transistor structure> Other examples of transistor structures of one aspect of the present invention, which differ from the above description, will be described.

[0511] Figures 16A to 20E This is a diagram illustrating a structural example of a transistor 500F that has a different structure from the transistor 500 described above.

[0512] [Transistor 500F] Figure 16A This is a top view of the transistor 500F. Figure 16B This is a 3D schematic diagram of the 500F transistor. Furthermore, Figures 16C to 16E This is a cross-sectional view of transistor 500F. Here, Figure 16C It is along Figure 16A The section shown by the dotted lines A1-A2 in the figure is a cross-sectional view, which is also a cross-sectional view of the channel width direction (here, represented by the Y direction) of the transistor 500F. Figure 16D It is along Figure 16A The cross-sectional view shown by the dashed lines A3-A4 is also a cross-sectional view along the channel width direction of the transistor 500F. Furthermore, Figure 16E It is along Figure 16A The cross-sectional view shown by the dashed lines A5-A6 is also a cross-sectional view along the channel length direction (here, represented by the X direction) of the transistor 500F. Here, dashed lines A5-A6 are orthogonal to dashed lines A1-A2 and A3-A4, and dashed lines A1-A2 are parallel to dashed lines A3-A4. Note that in... Figure 16A Top view and Figure 16B The three-dimensional diagram omits descriptions of some constituent elements. Furthermore, Figure 17A Show Figure 16E A magnified view of the conductor near 560. Furthermore, Figure 17B Show Figure 16C A magnified view of the oxide near 530.

[0513] Compared to the transistor 500, the transistor 500F has a structure that allows for increased channel width without significantly increasing the occupied area. In other words, it allows for increased on-state current without significantly increasing the occupied area.

[0514] Therefore, for example, by using transistor 500F in a storage device, the operating speed of the storage device can be improved, such as the read speed and write speed of data in the storage cells included in the storage device. Furthermore, for example, in the storage device 100 shown in Embodiment 1 above, transistor 500F can also be used as a transistor constituting the backup circuit 130, etc., included in the unit storage circuit 110.

[0515] Transistor 500F includes an insulator 514 on a substrate (not shown), an insulator 516 on the insulator 514, an insulator 521 on the insulator 516, an insulator 522 on the insulator 521, an oxide 530 on the insulator 522, conductors 542a and 542b on the oxide 530 and the insulator 522, an insulator 545 on the oxide 530, and conductors 560 (conductors 560a and 560b) on the insulator 545. Furthermore, in this specification and the like, conductors 542a and 542b are sometimes collectively referred to as conductor 542.

[0516] An insulator 544 is provided on the conductor 542, and an insulator 580 is provided on the insulator 544. An insulator 545 and a conductor 560 are disposed inside a first opening in the oxide 530, passing through the insulator 580 and the insulator 544. The first opening, when viewed from above, includes a region overlapping the oxide 530 and a region extending along the width direction of the channel beyond the oxide 530. Therefore, the insulator 545 and the conductor 560 disposed inside the first opening also have a region overlapping the oxide 530 and a region extending along the width direction of the channel beyond the oxide 530 when viewed from above. The conductor 560 is also used for wiring. The insulator 545 has a region in the first opening that contacts the oxide 530. Furthermore, an insulator 582 is provided on the insulator 580 and the conductor 560. Furthermore, an insulator 584 is provided on the insulator 582.

[0517] Furthermore, an insulator 541a is provided in such a way that it contacts the side of the second opening that passes through insulators 584, 582, 580, and 544 to reach conductor 542a, and a conductor 540a is provided in such a way that it contacts insulator 541a. Conductor 540a has a region that contacts conductor 542a at the bottom of the first opening.

[0518] Furthermore, an insulator 541b is provided in such a way that it contacts the side of the third opening that passes through insulators 584, 582, 580, and 544 to reach conductor 542b, and a conductor 540b is provided in such a way that it contacts insulator 541b. Conductor 540b has a region at the bottom of the second opening that contacts conductor 542b.

[0519] In this specification and other materials, conductor 540a and conductor 540b are sometimes collectively referred to as conductor 540. Additionally, insulator 541a and insulator 541b are sometimes collectively referred to as insulator 541.

[0520] Oxide 530 includes the channel forming region of transistor 500F. Conductor 560 has a region serving as the gate electrode of transistor 500F. Insulator 545 has a region serving as the gate insulating film of transistor 500F. In transistor 500F, the region of oxide 530 overlapping with conductor 560 is used as the channel forming region. Furthermore, the region of conductor 560 overlapping with oxide 530 is used as the gate electrode. Additionally, the region of insulator 545 where insulator 545 overlaps with oxide 530 and where insulator 545 overlaps with conductor 560 is used as the gate insulating film.

[0521] Conductor 542a has a region that serves as one of the source and drain electrodes of transistor 500F. Conductor 540a is used as a connector to connect conductor 542a to a conductor (not shown) on insulator 584. Conductor 542b has a region that serves as the other of the source and drain electrodes of transistor 500F. Conductor 540b is used as a connector to connect conductor 542b to a conductor (not shown) on insulator 584.

[0522] Oxide 530 is formed on insulator 522. For example... Figure 17B As shown, oxide 530 has a high aspect ratio shape when viewed in cross-section along the channel width direction. Therefore, it can be said that oxide 530 has a fin-like shape.

[0523] In this specification, the maximum length of the oxide 530 in the channel width direction within the channel forming region is defined as length Lx, and the maximum length of the oxide 530 in the channel forming region in the direction perpendicular to the formed surface (e.g., the top surface of the insulator 522) (here, the Z direction) is defined as length Lh. The ratio of length Lh to length Lx is referred to as the aspect ratio of the oxide 530. Furthermore, fin-shaped refers to a shape in which the oxide 530 has a high aspect ratio in a cross-sectional view along the channel width direction (a shape where length Lh is large relative to length Lx). Here, transistors in which the semiconductor layer including the channel forming region is fin-shaped are sometimes referred to as fin transistors, Fin-type transistors, Fin transistors, etc.

[0524] Note that the length Lx can also be described as the maximum width of the oxide 530 in the channel formation region. Therefore, "length Lx" can be replaced with "width Lx". Furthermore, the length Lh can also be described as the maximum height of the oxide 530 in the channel formation region. Therefore, "length Lh" can be replaced with "height Lh".

[0525] The aspect ratio of oxide 530 is preferably as high as possible within a range that will not cause oxide 530 to collapse during the manufacturing process of transistor 500F. The aspect ratio of oxide 530 may also be greater than 1 and less than 400, preferably more than 2 and less than 100, more preferably more than 5 and less than 40, and even more preferably more than 10 and less than 20. That is, in the channel formation region of oxide 530, the height Lh of oxide 530 is preferably at least longer than the length Lx of oxide 530. The height Lh of oxide 530 may also be greater than 1 times and less than 400 times the length Lx of oxide 530, preferably more than 2 times and less than 100 times, more preferably more than 5 times and less than 40 times, and even more preferably more than 10 times and less than 20 times. Furthermore, for example, the height Lh may also be more than 2 times and less than 10 times the length Lx. For example, the length Lx may also be more than 5 nm and less than 100 nm, preferably more than 5 nm and less than 50 nm, and even more preferably more than 10 nm and less than 30 nm. Furthermore, for example, the height Lh can be 50 nm or more and 2000 nm or less, preferably 100 nm or more and 1000 nm or less. Furthermore, for example, the height Lh can be 50 nm or more and 100 nm or less.

[0526] In addition, such as Figure 17B As shown, in a cross-sectional view along the channel width, the angle θ formed by the forming surface of the oxide 530 on the insulator 522 and the side surface of the oxide 530 is preferably perpendicular.

[0527] The insulator 545, conductor 560, and conductor 542 are disposed in a manner that covers the oxide 530 with a high aspect ratio. In transistor 500F, as... Figure 17B As shown, the insulator 545 and a portion of the conductor 560 are disposed in a folded manner, sandwiching the oxide 530. Thus, in a cross-sectional view along the channel width direction, the oxide 530 and conductor 560 are disposed with the insulator 545 sandwiched between them on the upper part, the side surface on the A1 side, and the side surface on the A2 side of the oxide 530. In other words, the upper part, the side surface on the A1 side, and the side surface on the A2 side of the oxide 530 are all used as channel forming regions. Therefore, compared to the case where the oxide 530 is formed as a planar shape, the channel width of the transistor 500F is increased by increasing the size of the side surfaces on the A1 and A2 sides of the oxide 530.

[0528] As described above, increasing the channel width can improve the on-state current, transconductance, and frequency characteristics of transistor 500F. This allows for the provision of a semiconductor device with high operating speed. Furthermore, in the structure of transistor 500F, the channel width can be increased without increasing the area occupied by the oxide 530. This enables miniaturization or high integration of semiconductor devices.

[0529] In addition, such as Figure 17B As shown, the upper part of oxide 530 can also have a curved shape. By having this curved shape, defects such as voids can be prevented from forming in the insulator 545 and conductor 542 near the upper part of oxide 530. Note that although in Figure 17B While the oxide 530 may have a curved, symmetrical structure on both sides (A1 and A2) of its upper portion, one aspect of the invention is not limited to this. For example, sometimes an asymmetrical structure may be used, where one of the A1 or A2 sides of the oxide 530 has a curved shape.

[0530] Here, a structural example is shown of oxide 530 including oxide 530a, oxide 530b in contact with oxide 530a, and oxide 530c in contact with oxide 530b.

[0531] At this time, for example, films that will become oxides 530a and 530c can be formed by atomic layer deposition (ALD), and films that will become oxide 530b can be formed by sputtering. Specifically, films that will become oxide 530a can be deposited with an In:Zn ratio of 2:1 or similar. Alternatively, indium oxide can be used for films that will become oxide 530a. Furthermore, films that will become oxide 530b can be deposited using an oxide target with an In:Sn:Zn ratio of 4:0.1:1 or similar. Furthermore, films that will become oxide 530c can be deposited with an In:Zn ratio of 2:1 or similar. Alternatively, indium oxide can be used for films that will become oxide 530c.

[0532] Next, a heat treatment is preferably performed. The heat treatment is preferably performed within a temperature range in which the oxide 530 does not become polycrystalline.

[0533] For example, as a heat treatment, it can be carried out for 1 hour at a nitrogen to oxygen gas flow ratio of 4:1 and a temperature of 450°C.

[0534] By depositing oxide 530 using the above method and then subjecting it to heat treatment, oxide 530 can be made into AGCAAC. This improves the on-state current, S-value, field-effect mobility, and frequency characteristics of transistor 500F, thereby providing a semiconductor device with excellent electrical characteristics. Furthermore, a highly reliable semiconductor device can be provided.

[0535] Furthermore, when oxide semiconductors are used as oxide 530, such as Figure 17A and Figure 17BAs shown, the insulator 545 preferably has a stacked structure consisting of an insulator 545a in contact with the oxide 530, an insulator 545b on the insulator 545a, an insulator 545c on the insulator 545b, and an insulator 545d on the insulator 545c. In this case, the insulators 545a and 545c preferably have the function of capturing or fixing hydrogen.

[0536] Metal oxides with amorphous structures can be cited as insulators capable of trapping or fixing hydrogen. For example, magnesium oxide or oxides containing one or both of aluminum and hafnium are preferred as insulators 545a and 545c. These metal oxides with amorphous structures sometimes possess the property that oxygen atoms have dangling bonds, which trap or fix hydrogen. In other words, metal oxides with amorphous structures have a high ability to trap or fix hydrogen.

[0537] Insulators 545a and 545c are preferably made of high-k materials with a high relative permittivity. Furthermore, an example of a high-k material is an oxide containing one or both of aluminum and hafnium. When a high-k material is used as insulator 545a and 545c, the gate potential applied during transistor operation can be reduced while maintaining the physical thickness of the gate insulating film. Additionally, the equivalent oxide thickness (EOT) of the insulator used as the gate insulating film can be reduced.

[0538] As insulators 545a and 545c, it is preferable to use an oxide containing one or both of aluminum and hafnium, and more preferably an oxide having an amorphous structure and containing one or both of aluminum and hafnium.

[0539] As the insulator 545a, an alumina film can be used, for example. Furthermore, the alumina preferably has an amorphous structure. Here, by providing the insulator 545a in contact with the oxide 530, the insulator 545a can more effectively trap and fix hydrogen contained in the oxide 530, etc.

[0540] Hafnium oxide can be used, for example, as an insulator 545c. Here, by providing an insulator 545c between insulators 545b and 545d, hydrogen contained in insulators 545b and the like can be captured and fixed more effectively.

[0541] Next, a thermally stable insulator such as silicon oxide or silicon oxynitride is preferably used as the insulator 545b. The silicon oxide film used as the insulator 545b is preferably formed using the PEALD method.

[0542] Furthermore, in order to suppress the oxidation of conductors 542a, 542b, and 560, it is preferable to provide an oxygen barrier insulator near each of conductors 542a, 542b, and 560. For example, an oxygen barrier insulator may also be provided in insulators 545a, 545d, 545c, and 544.

[0543] Note that in this specification, etc., a barrier insulator refers to an insulator that has barrier properties. In this specification, "barrier property" means having the property of hindering the transmission of the corresponding substance (also called low permeability). For example, a barrier insulator has the property that the corresponding substance does not easily diffuse into the interior of the insulator. For example, a barrier insulator has the function of trapping or fixing (also called gettering) the corresponding substance inside the insulator.

[0544] Examples of oxygen barrier insulators include oxides containing one or both of aluminum and hafnium, magnesium oxide, gallium oxide, silicon nitride, and silicon oxynitride. Examples of oxides containing one or both of aluminum and hafnium include aluminum oxide, hafnium oxide, oxides containing aluminum and hafnium (hafnium aluminate), and oxides containing hafnium and silicon (hafnium silicate). For example, insulators 545a, 545c, 545d, and 544 preferably employ a single-layer or multilayer structure of the aforementioned oxygen barrier insulators.

[0545] Insulator 545a preferably has oxygen-barrier properties. Insulator 545a is preferably less permeable to oxygen than insulator 580. Insulator 545a has regions that contact the sides of conductor 542a and the sides of conductor 542b. When insulator 545a has oxygen-barrier properties, oxidation of the sides of conductors 542a and 542b, resulting in the formation of an oxide film on those sides, can be suppressed. Therefore, a decrease in the on-state current or field-effect mobility of transistor 500F can be suppressed.

[0546] Furthermore, the insulator 545a is provided in contact with the top and side surfaces of the oxide 530 and the top surface of the insulator 522. When the insulator 545a is oxygen-blocking, it can suppress the detachment of oxygen from the channel formation region of the oxide 530 during heat treatment or similar processes. Therefore, the formation of oxygen vacancies in the oxide 530 can be reduced.

[0547] Furthermore, by providing insulator 545a, excessive oxygen supply from insulator 580 to oxide 530 is suppressed, thereby allowing an appropriate amount of oxygen to be supplied to oxide 530. Therefore, the decrease in on-state current or field-effect mobility of transistor 500F due to excessive oxidation of the source and drain regions can be suppressed.

[0548] Because oxides containing one or both of aluminum and hafnium are oxygen-blocking, they can be suitably used as insulators 545a.

[0549] The insulator 545d preferably also has oxygen-barrier properties. The insulator 545d is disposed between the channel formation region of the oxide 530 and the conductor 560, and between the insulator 580 and the conductor 560. By employing this structure, the diffusion of oxygen from the channel formation region of the oxide 530 to the conductor 560, thus preventing the formation of oxygen vacancies in the channel formation region of the oxide 530, can be suppressed. Furthermore, the diffusion of oxygen from the oxide 530 and the insulator 580 to the conductor 560, thus preventing oxidation of the conductor 560, can be suppressed. The insulator 545d preferably allows oxygen to permeate at least less than the insulator 580. For example, a silicon nitride film is preferably used as the insulator 545d. In this case, the insulator 545d is an insulator that at least contains nitrogen and silicon.

[0550] Furthermore, the insulator 545d preferably has hydrogen-blocking properties. This prevents impurities such as hydrogen contained in the conductor 560 from diffusing into the oxide 530.

[0551] The insulator 544 preferably also has oxygen-blocking properties. The insulator 544 is disposed between the insulator 580 and the conductor 542a, and between the insulator 580 and the conductor 542b. The insulator 544 is disposed in contact with the side surface of the conductor 542, the side surface of the oxide 530, and the top surface of the insulator 522. By employing this structure, the diffusion of oxygen contained in the insulator 580 into the conductor 542 can be suppressed. Therefore, the oxidation of the conductor 542 caused by oxygen contained in the insulator 580, which would increase the resistivity, can be suppressed. The insulator 544 preferably does not allow oxygen to permeate as much as the insulator 580. For example, silicon nitride is preferably used as the insulator 544. In this case, the insulator 544 is an insulator containing at least nitrogen and silicon.

[0552] To suppress the decrease in hydrogen concentration in the source and drain regions of oxide 530, it is preferable to provide hydrogen barrier insulators near the source and drain regions. For example, a hydrogen barrier insulator is used as insulator 544.

[0553] Examples of hydrogen barrier insulators include oxides such as aluminum oxide, hafnium oxide, and tantalum oxide, as well as nitrides such as silicon nitride. For example, the insulator 544 preferably employs a single-layer structure or a multilayer structure of the aforementioned hydrogen barrier insulators.

[0554] By incorporating insulator 544, hydrogen diffusion from the source and drain regions to the outside can be reduced, thus suppressing the decrease in hydrogen concentration in the source and drain regions. Therefore, the source and drain regions can be n-type.

[0555] By adopting the above structure, the channel formation region can be i-type or substantially i-type, and the source and drain regions can be n-type, providing a semiconductor device with excellent electrical characteristics. By adopting the above structure, even when the semiconductor device is miniaturized or highly integrated, it can still possess excellent electrical characteristics. Furthermore, miniaturizing the transistor 500F can improve high-frequency characteristics. Specifically, the cutoff frequency can be increased.

[0556] Insulators 545a to 545d are used as part of the gate insulating film. Insulators 545a to 545d are disposed together with conductor 560 in an opening formed in insulator 580. To achieve miniaturization of transistor 500F, the thickness of insulators 545a to 545d is preferably small. The thickness of each of insulators 545a to 545d is preferably 0.1 nm or more and 10 nm or less, more preferably 0.1 nm or more and 5.0 nm or less, further preferably 0.5 nm or more and 5.0 nm or less, even more preferably 1.0 nm or more and less than 5.0 nm, and even more preferably 1.0 nm or more and 3.0 nm or less. Furthermore, at least a portion of each of insulators 545a to 545d may include a region having the aforementioned thickness.

[0557] Furthermore, the thickness of the silicon oxide film used as insulator 545 is preferably 0.7 nm or more and 3 nm or less.

[0558] To reduce the thickness of insulators 545a to 545d as described above, deposition using the ALD method is preferred. Furthermore, to provide insulators 545a to 545d in openings in insulators 580, etc., deposition using the ALD method is preferred. By depositing insulator 545 using the ALD method, high coverage deposition can be achieved on the side surfaces of the first opening formed in insulator 580, the side ends of conductor 542a, and the side ends of conductor 542b, etc.

[0559] Note that while the insulator 545 is described above as having a four-layer structure consisting of insulators 545a to 545d, this is not the only embodiment of the invention. The insulator 545 may have a structure including at least one of insulators 545a to 545d. By comprising one, two, or three layers of insulators 545a to 545d, the manufacturing process of the transistor 500F can be simplified, thereby increasing the productivity of semiconductor devices including the transistor 500F.

[0560] like Figure 16AAs shown, the oxide 530, when viewed from above, is preferably annular (or, more accurately, frame-shaped, ring-shaped, or closed-curve-shaped). That is, the oxide 530 preferably has multiple portions extending in the channel width direction and multiple portions extending in the channel length direction. Therefore, when the oxide 530 has a high aspect ratio structure, it is possible to suppress the oxide 530 from collapsing during the transistor manufacturing process. Furthermore, it can also be said that… Figure 16A The oxide 530 shown has an opening in the central portion. Figure 16A In the diagram, the oxide 530 has a linearly symmetrical shape when viewed from above, centered on A1-A2. However, this invention is not limited to this. For example, the oxide 530 may also have an asymmetrical shape when viewed from above.

[0561] Figure 16A The structure shown is a structure in which two ring-shaped oxides 530 are formed in the width direction of the channel. For example... Figure 16A As shown, the oxide 530 preferably overlaps with the conductor 560 in two or more places when viewed from above. Therefore, the conductor 560 preferably has two or more regions that overlap with the oxide 530. In other words, it is preferable to have two or more regions where the oxide 530 and the conductor 560 overlap with each other.

[0562] By adopting this structure, such as Figure 16B As shown, in a cross-sectional view along the channel width direction, multiple fin-shaped oxides 530 are formed. Each of the multiple fin-shaped oxides 530 includes a channel forming region. In other words, the transistor 500F is used as a multi-channel transistor. Therefore, in the transistor 500F, the channel width can be further increased, thus increasing the on-state current. Therefore, the operating speed of the semiconductor device including the transistor 500F can be improved.

[0563] Here, a structure with two annular oxides 530 is described, but the invention is not limited to this. For example, a structure with one or more annular oxides 530 may also be used. Furthermore, the annular oxides 530 may be connected to form an oxide 530 having a shape including multiple openings. Additionally, an oxide 530 with a grid-like shape when viewed from above may also be used.

[0564] Insulators 584, 582, 522, and 521 preferably all include insulators that suppress the diffusion of impurities such as water and hydrogen, as well as oxygen. Examples of suitable materials include aluminum oxide, magnesium oxide, hafnium oxide, zirconium oxide, oxides containing aluminum and hafnium (hafnium aluminate), oxides containing hafnium and zirconium (hafnium zirconium oxide), gallium oxide, silicon nitride, or silicon oxynitride. For example, insulators 584 and 521 preferably use silicon nitride, which has higher hydrogen barrier properties. Furthermore, insulator 582 preferably uses aluminum oxide, which has a high hydrogen trapping or fixing ability. Similarly, insulator 522 preferably uses hafnium oxide, which has a high hydrogen trapping or fixing ability and is a high-k material.

[0565] Note that, as at least one of insulators 521 and 522, a multilayer structure of silicon oxide or silicon oxynitride may also be used in addition to the materials described above. For example, a multilayer structure of silicon nitride and silicon oxide may also be used as insulator 521. Furthermore, for example, a multilayer structure of hafnium oxide and silicon oxide may also be used as insulator 522.

[0566] By employing this structure, the diffusion of impurities such as water and hydrogen from the interlayer insulating film disposed above insulator 584 to transistor 500F and the like can be suppressed. Furthermore, the diffusion of impurities such as water and hydrogen from the interlayer insulating film disposed below insulator 521 to transistor 500F and the like can be suppressed. Additionally, hydrogen in insulators 580 and 545 can be captured or fixed to insulator 582 or insulator 522. Furthermore, by providing insulators 582 and 584, the diffusion of oxygen in insulator 580 to the top of transistor 500F and the like can be suppressed. Furthermore, by providing insulators 522 and 521, the diffusion of oxygen in oxide 530 to the bottom of transistor 500F and the like can be suppressed. Thus, by employing a structure in which insulators with the function of suppressing the diffusion of impurities such as water and hydrogen, as well as oxygen, surround the top and bottom of transistor 500F, excess oxygen and hydrogen diffusion into the oxide semiconductor can be reduced. This results in improved electrical characteristics and reliability of the semiconductor device.

[0567] The dielectric constants of insulators 516 and 580 are preferably lower than those of insulator 522. By using materials with low dielectric constants for the interlayer films, parasitic capacitances generated between wirings can be reduced.

[0568] For example, insulator 516 and insulator 580 preferably each contain one or more of silicon oxide, silicon oxynitride, fluorine-added silicon oxide, carbon-added silicon oxide, carbon and nitrogen-added silicon oxide, and porous silicon oxide.

[0569] In particular, silicon oxide and silicon oxynitride are preferred due to their thermal stability. Especially, materials such as silicon oxide, silicon oxynitride, and porous silicon oxide are preferred because they readily form regions containing oxygen released upon heating.

[0570] In addition, the top surfaces of insulators 516 and 580 can also be planarized.

[0571] The concentration of impurities such as water and hydrogen in the insulator 580 is preferably reduced. For example, silicon oxides such as silicon oxynitride are preferably used as the insulator 580.

[0572] In addition, such as Figures 18A to 18E As shown, in transistor 500F, an insulator 524 can also be provided under oxide 530. The planar shape of insulator 524 (here, the shape viewed from the Z direction) is the same as that of oxide 530, and in top view, insulator 524 overlaps with oxide 530. The bottom surface of insulator 524 contacts insulator 522, the side surface of insulator 524 contacts insulator 545 and conductor 542a, and the top surface of insulator 524 contacts the bottom surface of oxide 530. Furthermore, insulating material suitable for insulator 545b can also be used for insulator 524. For example, silicon oxide can be used as insulator 524. Figures 18A to 18E Corresponding to Figures 16A to 16E .in addition, Figure 19 Corresponding to Figure 17B Regarding Figures 18A to 18E and Figure 19 For matters not addressed below within the structure, please refer to the above section regarding... Figures 16A to 16E and Figure 17B Explanation, etc.

[0573] Here, as Figure 19 As shown, the thickness t2 of the insulator 545 at the bottom of the first opening is preferably thinner than the thickness t1 (length in the direction perpendicular to the surface of the insulator 524) of the insulator 524. By adopting this structure, the bottom surface of the conductor 560 (conductor 560a) located at the first opening can be lower than the bottom surface of the oxide 530 by the difference between thickness t1 and thickness t2 (t1-t2).

[0574] By positioning the bottom surface of the conductor 560 below the bottom surface of the oxide 530, a sufficient gate electric field can be applied to the upper and lower ends of the oxide 530. In other words, within the opening of the insulator 580, the entire oxide 530 can be electrically surrounded by the electric field of the conductor 560 and used as a channel formation region. This structure prevents the lower end of the oxide 530 from being used as a parasitic channel, thereby reducing the off-state current between the source and drain electrodes. Furthermore, it suppresses the always-on behavior of the transistor caused by this parasitic channel. In other words, the electrical characteristics of the transistor 500F can be improved.

[0575] As described above, by using the upper to lower ends of oxide 530 as the channel formation region, the channel width can be increased. This, in turn, improves the on-state current, transconductance, and frequency characteristics of transistor 500F.

[0576] Note that the transistor structure described above, where the electric field of the gate electrode forms a region around the channel, is called an S-channel structure. In an S-channel structure, the gate electrode is arranged with at least two or more surfaces surrounding the channel (specifically, two, three, or four surfaces, etc.). By employing an S-channel structure, the tolerance to short-channel effects can be improved; in other words, transistors that are less prone to short-channel effects can be realized.

[0577] Note that the S-channel structure is a structure in which electricity forms a region around the channel; therefore, it can be said that this structure is substantially the same as the GAA or LGAA structure. By giving the transistor 500F an S-channel, GAA, or LGAA structure, the channel formation region formed at or near the interface between the oxide 530 and the insulator 545 used as a gate insulating film can be disposed throughout the bulk of the oxide 530. This increases the current density flowing through the transistor, thus improving the transistor's on-state current or field-effect mobility. Furthermore, in one embodiment of the invention, the oxide 530 has a CAAC structure and a fin shape. By employing this structure, it is possible for the current path flowing between the source and drain of the transistor to be parallel to the ab plane of the crystal axis. In other words, the oxide semiconductor with the CAAC structure and fin structure has a conduction path that appears to be equivalent to that of a two-dimensional semiconductor material. Furthermore, by using such an oxide semiconductor, devices with two-dimensional conductivity can be fabricated.

[0578] In addition, such as Figures 20A to 20E As shown, in transistor 500F, conductor 503 can also be placed under insulator 521. Note that... Figures 20A to 20E Corresponding to Figures 16A to 16E Regarding Figures 20A to 20E For matters not addressed below within the structure, please refer to the above section regarding... Figures 16A to 16E Explanation, etc.

[0579] Similar to conductor 560, conductor 503 has a region that serves as a gate electrode. Conductor 560 is sometimes referred to as the first gate electrode (upper gate electrode) of transistor 500F, and conductor 503 is sometimes referred to as the second gate electrode (lower gate electrode) of transistor 500F. Furthermore, conductor 560 is sometimes referred to as the gate electrode of transistor 500F, and conductor 503 is sometimes referred to as the back gate electrode of transistor 500F.

[0580] When transistor 500F includes a conductor 503 under insulator 521, similar to insulator 545, both insulators 522 and 521 include regions that serve as gate insulating films. Specifically, the regions overlapping with the conductors 503 of each of insulators 522 and 521 are used as gate insulating films. Furthermore, insulator 545 is sometimes referred to as the first gate insulating film (the upper gate insulating film), and insulators 522 and 521 are sometimes referred to as the second gate insulating film (the lower gate insulating film).

[0581] In transistor 500F, conductor 503 is arranged to overlap with oxide 530 and conductor 560. Figure 20C and Figure 20E In this configuration, a conductor 503 is disposed inside the fourth opening that passes through the insulator 516 to reach the insulator 514. Furthermore, the fourth opening, when viewed from above, includes a region overlapping with the oxide 530 and a region extending along the channel width direction beyond the oxide 530. Therefore, the conductor 503 disposed inside the fourth opening also has, when viewed from above, a region overlapping with the oxide 530 and a region extending along the channel width direction beyond the oxide 530. The conductor 503 is also used for wiring.

[0582] like Figure 20C and Figure 20E As shown, conductor 503 preferably includes conductor 503a and conductor 503b. Conductor 503a is disposed in contact with the bottom and side of the fourth opening. Conductor 503b is disposed in a recess formed along the bottom and side of conductor 503a of the fourth opening. Here, the height of the top surface of conductor 503 is aligned with the height of the top surface of insulator 516.

[0583] Here, the conductor 503a preferably comprises a conductive material that has the function of suppressing the diffusion of impurities such as hydrogen atoms, hydrogen molecules, water molecules, nitrogen atoms, nitrogen molecules, nitrogen oxide molecules (N2O, NO, NO2, etc.), and copper atoms. Alternatively, it is preferable to include a conductive material that has the function of suppressing the diffusion of oxygen (e.g., at least one of oxygen atoms, oxygen molecules, etc.).

[0584] By using a conductive material that inhibits hydrogen diffusion as conductor 503a, impurities such as hydrogen contained in conductor 503b can be prevented from diffusing to oxide 530 through insulator 516, etc. Furthermore, by using a conductive material that inhibits oxygen diffusion as conductor 503a, oxidation of conductor 503b and subsequent decrease in conductivity can be prevented. Examples of conductive materials that inhibit oxygen diffusion include titanium, titanium nitride, tantalum, tantalum nitride, ruthenium, and ruthenium oxide. Conductor 503a can have a single-layer structure or a multilayer structure of the aforementioned conductive materials. For example, conductor 503a preferably contains titanium nitride.

[0585] Furthermore, the conductor 503b preferably uses a conductive material whose main component is tungsten, copper, or aluminum. For example, the conductor 503b preferably contains tungsten.

[0586] As described above, conductor 503 can be used as a second gate electrode. In this case, the threshold voltage of transistor 500F can be controlled by independently changing the potential supplied to conductor 503 without linking it to the potential supplied to conductor 560. In particular, by applying a negative potential to conductor 503, the threshold voltage of transistor 500F can be increased and the off-state current can be reduced. Therefore, compared with not applying a negative potential to conductor 503, applying a negative potential to conductor 503 can reduce the drain current when the potential of conductor 560 is 0V.

[0587] Furthermore, the resistivity of the conductor 503 is designed based on the potential applied to the conductor 503, and the thickness of the conductor 503 is set according to this resistivity. Additionally, the thickness of the insulator 516 is approximately the same as that of the conductor 503. Here, it is preferable to reduce the thickness of both the conductor 503 and the insulator 516 within the design limits of the conductor 503. By reducing the thickness of the insulator 516, the absolute amount of impurities such as hydrogen contained in the insulator 516 can be reduced, thus suppressing the diffusion of these impurities into the oxide 530.

[0588] Note that a stacked structure of conductors 503a and 503b is shown here, but the present invention is not limited thereto; conductor 503 can have a single-layer structure or a stacked structure of three or more layers. For example, when conductor 503 has a three-layer stacked structure, a stacked structure of conductors 503a and 503b can be used, and a conductor containing the same material as conductor 503a can be provided on conductor 503b. In this case, the conductor can also be formed such that the top surface of conductor 503b is lower than the uppermost part of conductor 503a and is embedded in the recess formed by conductors 503a and 503b.

[0589] <Constructing Materials> The constituent materials used in semiconductor devices, including transistors and capacitors, are not limited to the structural examples described above. In one aspect of the invention, in addition to the constituent materials described above, the constituent materials shown below may also be suitably used.

[0590] [Substrate] As substrates for semiconductor devices and memory devices including such semiconductor devices, which can be provided according to one aspect of the present invention, examples of substrates that can be used include glass substrates, quartz substrates, sapphire substrates, ceramic substrates, metal substrates (e.g., stainless steel substrates, substrates containing stainless steel foil, tungsten substrates, substrates containing tungsten foil, etc.), semiconductor substrates (e.g., single-crystal semiconductor substrates, polycrystalline semiconductor substrates, or compound semiconductor substrates), or SOI (Silicon on Insulator) substrates. Furthermore, heat-resistant plastic substrates can also be used as substrates. Examples of glass substrates include barium borosilicate glass, aluminosilicate glass, aluminoborosilicate glass, or soda-lime glass. In addition, crystallized glass can also be used as glass substrates.

[0591] Furthermore, as a substrate, flexible substrates, laminated films, paper containing fibrous materials, or substrate films can be used, for example. Examples of flexible substrates, laminated films, or substrate films include plastics such as polyethylene terephthalate (PET), polyethylene naphthalate (PEN), polyethersulfone (PES), and polytetrafluoroethylene (PTFE). Alternatively, synthetic resins such as acrylic resins can be used. Alternatively, polypropylene, polyester, ethylene fluoride, or polyvinyl chloride can be used. Alternatively, polyamide, polyimide, aromatic polyamide resin, epoxy resin, inorganic vapor-deposited films, and paper can be used. In particular, by using semiconductor substrates, single-crystal substrates, or SOI substrates to manufacture transistors, it is possible to produce transistors with small deviations in characteristics, size, or shape, high current capability, and small size. When circuits are constructed using the aforementioned transistors, low power consumption or high integration of the circuit can be achieved.

[0592] Alternatively, a flexible substrate can be used as the substrate, and one or more transistors, resistors, and capacitors can be directly formed on the flexible substrate. Alternatively, a release layer can be provided between the substrate and one or more transistors, resistors, and capacitors. This release layer can be used when a part or all of a semiconductor device is manufactured on the release layer, and then separated from the substrate and transferred to another substrate. In this case, one or more transistors, resistors, and capacitors can also be transferred to a substrate with low heat resistance or a flexible substrate. Furthermore, as the aforementioned release layer, for example, a stacked structure of inorganic films such as tungsten film and silicon oxide film, a structure in which an organic resin film such as polyimide is formed on the substrate, or a hydrogen-containing silicon film can be used.

[0593] In other words, a semiconductor device can be formed on one substrate and then transferred to another substrate. The substrate used for transferring the semiconductor device can be not only the substrates mentioned above that can form transistors, but also paper substrates, cellophane substrates, aramid film substrates, polyimide film substrates, stone substrates, wood substrates, cloth substrates (including natural fibers (silk, cotton, hemp), synthetic fibers (nylon, polyurethane, polyester) or recycled fibers (acetate fiber, cupro fiber, artificial fiber, recycled polyester), etc.), leather substrates, rubber substrates, etc. By using these substrates, it is possible to manufacture flexible semiconductor devices or semiconductor devices that are not easily damaged. Furthermore, heat resistance can be imparted to the semiconductor device. In addition, it is possible to achieve lightweight and thin semiconductor devices.

[0594] By placing semiconductor devices on a flexible substrate, it is possible to provide semiconductor devices that suppress weight gain and are not easily damaged.

[0595] Ferroelectric materials As an insulator (such as insulator 632) used as a dielectric in a semiconductor device that can be used in one aspect of the present invention, a material that can be ferroelectric can also be used. Examples of materials that can be ferroelectric include hafnium oxide, zirconium oxide, and HfZrO. X (X is a real number greater than 0) and other metal oxides. Furthermore, as materials that can exhibit ferroelectric properties, examples include materials in which element J1 (here, element J1 is selected from one or more of zirconium, silicon, aluminum, gadolinium, yttrium, lanthanum, strontium, etc.) is added. Furthermore, the atomic ratio of hafnium to element J1 can be appropriately set. For example, the atomic ratio of hafnium to element J1 can be 1:1 or close to it. Furthermore, as materials that can exhibit ferroelectric properties, examples include materials in which element J2 (here, element J2 is selected from one or more of hafnium, silicon, aluminum, gadolinium, yttrium, lanthanum, strontium, etc.) is added. Furthermore, the atomic ratio of zirconium to element J2 can be appropriately set. For example, the atomic ratio of zirconium to element J2 can be 1:1 or close to it. Furthermore, as materials that can exhibit ferroelectric properties, lead titanate (PbTiO2) can also be used. X Piezoelectric ceramics with perovskite structure include barium strontium titanate (BST), strontium titanate, lead zirconate titanate (PZT), strontium bismuth tantalate (SBT), bismuth ferrite (BFO), and barium titanate.

[0596] Furthermore, metal nitrides containing elements M1, M2, and nitrogen can be cited as materials that can exhibit ferroelectric properties. Here, element M1 is, for example, one or more selected from aluminum, gallium, indium, etc. Furthermore, element M2 is, for example, one or more selected from boron, scandium, yttrium, lanthanum, cerium, neodymium, europium, titanium, zirconium, hafnium, vanadium, niobium, tantalum, chromium, etc. Furthermore, the ratio of the number of atoms of element M1 to the number of atoms of element M2 can be appropriately set. Furthermore, metal nitrides containing element M1 and nitrogen sometimes exhibit ferroelectric properties even without element M2. Furthermore, materials that can exhibit ferroelectric properties include those to which element M3 is added. Here, element M3 is, for example, one or more selected from magnesium, calcium, strontium, zinc, cadmium, etc. Here, the ratio of the number of atoms of element M1, element M2, and element M3 can be appropriately set.

[0597] In addition, perovskite-type oxynitrides such as SrTaO2N and BaTaO2N, and κ-type alumina such as GaFeO3 can be cited as materials that can exhibit ferroelectric properties.

[0598] Note that while examples of metal oxides and metal nitrides are shown in the above description, the invention is not limited to these. For example, metal oxynitrides with nitrogen added to the aforementioned metal oxides or metal oxynitrides with oxygen added to the aforementioned metal nitrides may also be used.

[0599] Furthermore, as a material that can exhibit ferroelectricity, for example, a mixture or compound composed of multiple materials selected from the above-mentioned materials can be used. Additionally, as an insulator using a material that can exhibit ferroelectricity, a layered structure composed of multiple materials selected from the above-mentioned materials can be employed. The crystal structure (properties) of the materials, etc., exemplified above may vary not only depending on the deposition conditions but also depending on various processes, etc. Therefore, in this specification, etc., materials exhibiting ferroelectricity are sometimes referred to not only as ferroelectric materials but also as materials that can exhibit ferroelectricity.

[0600] Metal oxides containing one or both of hafnium and zirconium can exhibit ferroelectricity even when processed into thin films of a few nm, and are therefore preferred. Here, the thickness of the insulator using a ferroelectric material can be 100 nm or less, preferably 50 nm or less, more preferably 20 nm or less, and even more preferably 10 nm or less (typically, 2 nm or more and 9 nm or less). For example, the thickness of this insulator is preferably 8 nm or more and 12 nm or less. By using a ferroelectric layer that can be thinned as an insulator for use as a dielectric in a capacitor, the capacitor can be combined with miniaturized semiconductor elements such as transistors to form a semiconductor device. In this specification, the ferroelectric material formed in a layered manner is sometimes referred to as a ferroelectric layer, a metal oxide film, or a metal nitride film. Furthermore, in this specification, devices comprising ferroelectric layers, metal oxide films, or metal nitride films are sometimes referred to as ferroelectric devices.

[0601] Furthermore, metal oxides containing one or both of hafnium and zirconium can exhibit ferroelectricity even when their area is extremely small, and are therefore preferred. For example, the ferroelectric layer can have an area (occupied area) of 100 μm when viewed from above. 2 Below, 10μm 2 Below, 1μm 2 Below or 0.1μm 2 The following can also exhibit ferroelectric properties. Furthermore, sometimes ferroelectric layers can exist even with a top-view area (occupied area) of 10000 nm. 2 Below or 1000nm 2 The following also exhibit ferroelectric properties. By forming a small ferroelectric layer, the area occupied by the capacitor can be reduced.

[0602] Ferroelectric materials are insulators that exhibit the property of becoming polarized internally under the influence of an applied electric field and maintaining that polarization even when the electric field is zero. Therefore, non-volatile storage elements can be formed by using capacitors (hereinafter sometimes called ferroelectric capacitors) with this material as the dielectric. Non-volatile storage elements using ferroelectric capacitors are sometimes referred to as FeRAM (Ferroelectric Random Access Memory), ferroelectric memory, etc. For example, a ferroelectric memory includes a transistor and a ferroelectric capacitor, with one of the source and drain terminals of the transistor connected to a terminal of the ferroelectric capacitor.

[0603] Furthermore, ferroelectricity is thought to arise from the displacement of oxygen or nitrogen in the crystal contained within the ferroelectric layer under the influence of an applied electric field. Moreover, the presence of ferroelectricity is presumed to depend on the structure of the crystal contained within the ferroelectric layer. Therefore, for an insulator made of a material capable of exhibiting ferroelectricity to possess ferroelectricity, the insulator needs to contain a crystal. In particular, the insulator preferably contains a crystal with an orthorhombic crystal structure, thereby exhibiting ferroelectricity. The crystal structure of the crystal contained in the insulator can be selected from one or more of the isometric, tetragonal, orthorhombic, monoclinic, and hexagonal crystal systems. Furthermore, the insulator can also have an amorphous structure. In this case, the insulator can also have a composite structure of amorphous and crystalline structures.

[0604] One embodiment of the present invention may employ transistors that appropriately combine the structures of transistors 500, 500A, 600, 600B, and 500F. Additionally, capacitors that appropriately combine the structures of capacitors 590, 590A, and 690 may be employed.

[0605] Furthermore, in one embodiment of the present invention, the transistors 500, 500A, 600, 600B, 500F, and appropriately combined transistors can be used in the storage device 100 shown in Embodiment 1. For example, they can be used as transistors constituting the backup circuit included in the unit storage circuit 110. Additionally, the capacitors 590, 590A, 690, and appropriately combined capacitors can be used in the storage device 100 shown in Embodiment 1. For example, they can be used as capacitors constituting the backup circuit included in the unit storage circuit 110. Furthermore, the transistor 550 can be used in the storage device 100 shown in Embodiment 1. For example, it can be used as a transistor constituting the check bit generation unit 102, the error detection and correction unit 103, and the scan trigger circuit 120 included in the unit storage circuit 110.

[0606] Note that one aspect of the present invention is not limited to the structural and operational examples described in this embodiment. The contents described in this embodiment can be appropriately combined and implemented. Furthermore, the contents described in this embodiment can be appropriately combined and implemented with the contents described in other embodiments, etc.

[0607] (Implementation Method 3) In this embodiment, an oxide semiconductor transistor (OS transistor) is described. Furthermore, in the description of the OS transistor, a simple comparison is given with a transistor containing silicon in the channel formation region (also known as a Si transistor).

[0608] [OS transistor] Oxide semiconductors with low carrier concentrations are preferably used in OS transistors. For example, the carrier concentration in the channel formation region of an oxide semiconductor is 1 × 10⁻⁶. 18 cm -3 Below, it is preferred to be less than 1×10 17 cm -3 More preferably, less than 1×10 16 cm -3 Further optimization of less than 1×10 13 cm -3 Furthermore, it is preferred to select those with a density lower than 1×10 10 cm -3 And it is 1×10 -9 cm -3 The above applies. When aiming to reduce the carrier concentration in an oxide semiconductor, the impurity concentration in the oxide semiconductor can be reduced to decrease the defect state density. In this specification, a state with both low impurity concentration and low defect state density is referred to as high-purity intrinsic or substantially high-purity intrinsic. Furthermore, oxide semiconductors with low carrier concentrations are sometimes referred to as high-purity intrinsic or substantially high-purity intrinsic oxide semiconductors.

[0609] Because high-purity intrinsic or substantially high-purity intrinsic oxide semiconductors have a low defect state density, they sometimes have a low trapped state density. Furthermore, the charge trapped in the trapped states of oxide semiconductors takes a long time to dissipate, sometimes acting like a fixed charge. Therefore, transistors forming channel formation regions in oxide semiconductors with high trapped state densities sometimes exhibit unstable electrical characteristics.

[0610] Therefore, reducing the impurity concentration in the oxide semiconductor is effective in stabilizing the electrical characteristics of the transistor. Furthermore, to reduce the impurity concentration in the oxide semiconductor, it is preferable to also reduce the impurity concentration in the adjacent film. Examples of impurities include hydrogen or nitrogen. Note that impurities in an oxide semiconductor refer to elements other than the main components constituting the oxide semiconductor. For example, elements with a concentration of less than 0.1 atomic percent can be considered impurities.

[0611] In OS transistors, the presence of impurities or oxygen vacancies in the channel formation region of the oxide semiconductor can easily alter electrical characteristics, potentially leading to decreased reliability. Furthermore, in OS transistors, hydrogen can enter oxygen vacancies in the oxide semiconductor, forming defects (sometimes referred to below as V0). O H), which may generate electrons that become charge carriers. Furthermore, in an OS transistor, when V is formed in the channel formation region... OAt H, the donor concentration in the channel formation region sometimes increases. Therefore, in an OS transistor, as the donor concentration in the channel formation region increases, the threshold voltage sometimes becomes uneven. Thus, when oxygen vacancies are present in the channel formation region of an oxide semiconductor, the OS transistor exhibits always-on characteristics (the characteristic that drain current flows when the gate voltage is 0V). Therefore, in the channel formation region of an oxide semiconductor, it is preferable to minimize impurities, oxygen vacancies, and V. O H.

[0612] Furthermore, the band gap of the oxide semiconductor is preferably larger than that of silicon (typically 1.1 eV), preferably 2 eV or more, more preferably 2.5 eV or more, and even more preferably 3.0 eV or more. By using an oxide semiconductor with a larger band gap than silicon, the off-state current (also known as Ioff) of the transistor can be reduced.

[0613] For example, in Si transistors, the short-channel effect (SCE) emerges as transistors become miniaturized. Therefore, miniaturization of Si transistors is difficult. One reason for the short-channel effect is the small band gap of silicon. On the other hand, in OS transistors, oxide semiconductors, which are semiconductor materials with large band gaps, are used, thus suppressing the short-channel effect. In other words, OS transistors are transistors with no or very little short-channel effect.

[0614] The short-channel effect refers to the decline in electrical characteristics that occurs as transistors are miniaturized (channel length is reduced). Specific examples of the short-channel effect include a decrease in threshold voltage, an increase in subthreshold swing (sometimes referred to as the S-value), and an increase in leakage current. Here, the S-value refers to the change in gate voltage in the subthreshold region where the drain current changes by a one-digit percentage while maintaining a fixed drain voltage.

[0615] Characteristic length is widely used as an indicator of tolerance to short-channeling effects. Characteristic length refers to the tortuosity of the potential in the channeling region. The smaller the characteristic length, the more rapidly the potential rises, thus indicating higher resistance to short-channeling effects.

[0616] OS transistors are accumulation-type transistors, while Si transistors are inversion-type transistors. Therefore, compared to Si transistors, OS transistors have smaller characteristic lengths between the source and channel formation regions, and smaller characteristic lengths between the drain and channel formation regions. Consequently, OS transistors have higher resistance to short-channel effects than Si transistors. In other words, OS transistors are more suitable than Si transistors when it is desirable to manufacture transistors with short channel lengths.

[0617] Even when the carrier concentration in the oxide semiconductor is reduced to the point where the channel formation region is i-typed or substantially i-typed, the conduction band bottom of the channel formation region in a short-channel transistor decreases due to the Conduction-Band-Lowering (CBL) effect. Therefore, the energy difference between the conduction band bottom of the source or drain region and the channel formation region can potentially decrease to above 0.1 eV and below 0.2 eV. Thus, an OS transistor can be considered as having n + / n - / n + Accumulation-type junction-less transistor structure or n + / n - / n + An accumulation-type non-junction transistor structure, wherein the channel formation region is n - The source and drain regions are all n-type. + Type of area.

[0618] When the above structure is used as an OS transistor, good electrical characteristics can be achieved even with miniaturization or high integration. For example, good electrical characteristics can be obtained even with gate lengths of 20nm or less, 15nm or less, 10nm or less, 7nm or less, or 6nm or less but 1nm or more, 3nm or more, or 5nm or more. On the other hand, in Si transistors, it is sometimes difficult to achieve gate lengths of 20nm or less or 15nm or less due to the short-channel effect. Therefore, compared with Si transistors, OS transistors are suitable for use as transistors with small channel lengths. Gate length refers to the length of the gate electrode in the direction of carrier migration within the channel formation region when the transistor is operating; it is the width of the bottom surface of the gate electrode when the transistor is viewed from above.

[0619] Furthermore, miniaturizing the OS transistor can improve its high-frequency characteristics. Specifically, the transistor's cutoff frequency can be increased. When the gate length of the OS transistor is within the aforementioned range, for example at room temperature, the transistor's cutoff frequency can be 50 GHz or higher, preferably 100 GHz or higher, and more preferably 150 GHz or higher.

[0620] As explained above, OS transistors have advantages over Si transistors, such as lower off-state current and the ability to manufacture transistors with shorter channel lengths.

[0621] Note that the contents described in this embodiment can be appropriately combined and implemented. Furthermore, the contents described in this embodiment can be appropriately combined and implemented in combination with contents described in other embodiments, etc.

[0622] (Implementation Method 4) This embodiment illustrates an application example of a storage device according to one aspect of the present invention. The storage device according to one aspect of the present invention is, for example, a storage device using oxide semiconductors.

[0623] <Example of storage device hierarchy> Generally speaking, in computers and other similar devices, various storage devices are used depending on their purpose. Figure 21 The various storage devices are shown in a hierarchy. Higher-level storage devices require faster operating speeds, while lower-level storage devices require larger storage capacities and higher storage densities. Figure 21 In the diagram, from top to bottom, registers, cache memory, main memory, and storage are shown in sequence. Furthermore, cache memory can also include L1, L2, and L3 caches, etc., from top to bottom. Note that while this example shows at most a L3 cache, it can also include caches at lower levels. Sometimes the lowest-level cache is referred to as LLC (Last Level Cache) or FLC (Final Level Cache). Additionally, for example, storage-level memory can be included between main memory and storage.

[0624] Registers are integrated into computing devices (also known as processors) such as CPUs (Central Processing Units), GPUs (Graphics Processing Units), NPUs (Neural Processing Units), and TPUs (Tensor Processing Units) to temporarily store the results of core operations. In addition, registers also serve to maintain settings information for the computing devices. Therefore, they are accessed frequently by the computing devices. Consequently, registers are required to operate at high speeds.

[0625] As a cache memory, for example, SRAM (Static Random Access Memory) is used. A cache memory has the function of copying and maintaining a portion of the data that is held in main memory. By copying frequently used data, the speed of data access can be improved. Cache memories are required to operate at a faster speed than main memory.

[0626] As main memory, for example, DRAM (Dynamic Random Access Memory) is used. Main memory has the function of holding programs and data read from storage. Compared with cache memory, main memory requires larger storage capacity and higher storage density.

[0627] Storage serves to hold data that needs to be preserved long-term and various programs used by computing devices. Therefore, storage requires large capacity and high density. Examples of storage devices include HDDs (Hard Disk Drives) and SSDs (Solid State Drives) located above HDDs. SSDs can be large-capacity, non-volatile storage devices such as NAND flash memory (e.g., 3D NAND).

[0628] The advantages of a storage device according to one aspect of the present invention (e.g., a storage device using oxide semiconductors) are as follows: high operating speed, ability to retain data for a long time, high rewrite tolerance, and ability to operate at low voltage.

[0629] From the viewpoint of being able to retain data for a long time, a storage device according to one aspect of the present invention is suitable for use as a storage device located in region target 1, which includes the layer where the cache memory is located, the layer where the main memory is located, and the layer where the storage is located. That is, it is preferable to use a storage device according to one aspect of the present invention in region target 1, which includes the region where the main memory is located, the boundary region between the main memory and the storage, and the boundary region between the main memory and the cache memory.

[0630] Therefore, for example, it is preferable to use a storage device according to one aspect of the invention instead of DRAM for main memory. Here, DRAM is a storage device that requires refresh operations and destructive reads, and therefore has high power consumption compared to other storage devices. Therefore, by not using DRAM, power consumption can be reduced. For example, it is also preferable to use a storage device according to one aspect of the invention instead of a portion of SRAM for cache memory and a portion of 3D NAND for storage.

[0631] Furthermore, from the viewpoint of achieving fast and excellent write and read speeds, a storage device according to one aspect of the present invention is suitable for use as a storage device in region target 2, which includes the level where the cache memory is located and the level where the registers are located. That is, it is preferable to use a storage device according to one aspect of the present invention in region target 2, which includes the region where the cache memory is located and the region where the registers are located.

[0632] Therefore, for example, it is preferable to use the storage device according to one aspect of the present invention as at least a portion of the registers included in a CPU, GPU, and NPU. Furthermore, for example, it is preferable to use the storage device according to one aspect of the present invention as at least a portion of a cache memory (L1, L2, L3, LLC, and FLC, etc.).

[0633] One aspect of the present invention eliminates the need for DRAM, which is conventionally used for main memory, etc. In this case, a storage device according to one aspect of the present invention can be used instead of DRAM. By adopting this structure, power consumption can be significantly reduced (e.g., less than 1 / 100th or 1 / 1000th of its original value). Therefore, by promoting information processing devices including supercomputers (also known as HPCs, computers, servers, etc.) with this structure globally, global warming can be mitigated.

[0634] As one aspect of the present invention, for example, at least a portion of the storage device 100 shown in Embodiment 1 above can be used to install in the registers of computing devices such as CPU, GPU, NPU, and TPU.

[0635] <Example of storage device structure> A storage device 700 according to one aspect of the present invention will be described. For example, at least a portion of the storage device 700 can be used as a storage device, the storage device being located in the above-described... Figure 21 The region taeget1 includes the level where the cache memory is located, the level where the main memory is located, and the storage level. Furthermore, the transistors shown in Embodiment 2 can be used in at least a portion of the storage device 700.

[0636] Figure 22 This is a block diagram illustrating an example of the structure of the storage device 700. Figure 22 The storage device 700 shown includes a storage array 721 and a drive circuit 722.

[0637] The storage array 721 includes a plurality of storage cells 741. The plurality of storage cells 741 are configured as a matrix of M rows × N columns. Here, M is an integer greater than or equal to 1. N is an integer greater than or equal to 1.

[0638] In addition, Figure 22 Typically shown are storage cells 741[1,1] in the first row and first column, storage cells 741[1,N] in the first row and Nth column, storage cells 741[M,1] in the first row and first column, and storage cells 741[M,N] in the first row and Nth column.

[0639] also, Figure 22Wiring WL[1] connected to N storage cells 741 arranged in the first row, wiring WL[M] connected to N storage cells 741 arranged in the M row, wiring BL[1] connected to M storage cells 741 arranged in the first column, and wiring BL[N] connected to M storage cells 741 arranged in the N column are shown.

[0640] The drive circuit 722 includes power switches 761 and 762 and peripheral circuits 771. The peripheral circuits 771 include peripheral circuits 781, control circuits 772 and voltage generation circuits 773.

[0641] In one aspect of the invention, for example, a Si transistor (a transistor in which silicon is contained in the channel forming region) can be used as the transistor constituting the driving circuit 722. Therefore, for example, a CMOS circuit (e.g., a circuit that operates complementaryly, a CMOS logic gate, or a CMOS logic circuit, etc.) can be used as the driving circuit 722 by connecting the gate of an n-channel Si transistor and the gate of a p-channel Si transistor.

[0642] As the Si transistor, at least a portion of the transistor 550 shown in Embodiment 2 above can be used, for example.

[0643] Here, for example, by using OS transistors (transistors with oxide semiconductors in the channel formation region) as transistors constituting memory cells 741, the memory array 721 can be stacked on the drive circuit 722 using Si transistors. This allows for miniaturization of the memory device 700. Furthermore, the wiring distance between the drive circuit 722 and the memory array 721 can be shortened. Therefore, improvements in read and write speeds of the memory device 700 can be achieved.

[0644] As the OS transistor, at least a portion of transistors 500, 500A, 600, 600B, and 500F shown in Embodiment 2 above can be used.

[0645] Furthermore, although not illustrated, a different structure can be adopted: in the storage device 700, the storage array 721 includes multiple readout amplifiers configured in a matrix, with multiple storage cells 741 stacked on top of these readout amplifiers. By employing this structure, multiple readout amplifiers can be accessed simultaneously, thereby enabling large-scale parallel reading of data stored in the storage array 721.

[0646] Terminals BW, CE, GW, MCK, WAKE, ADDR, WDA, PON1, and PON2 are supplied with signals from outside the storage device 700, for example. Additionally, signals are output to the outside of the storage device 700, for example, from terminal RDA.

[0647] For example, terminal MCK is supplied with a clock signal. Additionally, terminals BW, CE, and GW are supplied with control signals. Terminal CE is supplied with a chip enable signal. Terminal GW is supplied with a global write enable signal. Terminal BW is supplied with a byte write enable signal. Terminal ADDR is supplied with an address signal. Terminal WDA is supplied with write data. Terminal RDA is supplied with read data. Terminals PON1 and PON2 are supplied with power gating control signals. Furthermore, the signals supplied to terminals PON1 and PON2 can also be generated in the control circuit 772, for example.

[0648] The control circuit 772 has the function of controlling the operation of the storage device 700. For example, the control circuit 772 has the function of performing logical operations on the signals supplied to each of the terminals CE, GW, and BW to determine the operating mode of the storage device 700 (e.g., write operation or read operation). Furthermore, it also has the function of generating signals to control the peripheral circuit 781 to execute the operating mode.

[0649] The voltage generation circuit 773 has the function of generating an arbitrary potential to enable the drive circuit 722 to operate. For example, the voltage generation circuit 773 has the function of generating an arbitrary potential based on a signal supplied to the WAKE terminal and a clock signal input to the MCK terminal. The WAKE terminal is, for example, supplied with a signal that controls whether the clock signal supplied to the MCK terminal is input to the voltage generation circuit 773.

[0650] The peripheral circuit 781 has the function of writing and reading data from the storage unit 741. For example, the peripheral circuit 781 has the function of generating various signals to control the operation of the storage unit 741, etc. The peripheral circuit 781 includes a row decoder 782, a column decoder 784, a row driver 783, a column driver 785, a data driver 786, an input circuit 787, and an output circuit 788.

[0651] Row decoder 782 and column decoder 784 are functionally designed to decode address signals supplied to terminal ADDR. Row decoder 782 is functionally designed to specify the row to be accessed. Column decoder 784 is functionally designed to specify the column to be accessed. Row driver 783 is functionally designed to select the row specified by row decoder 782 and supply the desired signal, for example, to the corresponding memory cell 741. Column driver 785 is functionally designed to select the column specified by column decoder 784 and supply the desired signal, for example, to the corresponding memory cell 741.

[0652] Data driver 786 has the function of writing and reading data to memory cell 741 selected by row driver and column driver. Input circuit 787 has the function of holding data supplied from outside the storage device 700 to terminal WDA. Data held in input circuit 787 (data Din) is written to memory cell 741 via data driver 786. Data stored in memory cell 741 is read from data cell 741 to output circuit 788 via data driver 786. Output circuit 788 has the function of holding the read data (data Dout). In addition, it also has the function of outputting the held data from terminal RDA to outside the storage device 700.

[0653] Figure 22 The storage device 700 shown, for example, has the function of supplying desired signals to wirings WL[1] to WL[M] via row driver 783, and has the function of exchanging data with wirings BL[1] to BL[N] via column driver 785 and data driver 786.

[0654] Power switch 761 has the function of controlling whether the potential supplied to terminal VMD is supplied to peripheral circuit 771. Power switch 762 has the function of controlling whether the potential supplied to terminal VMH is supplied to row driver 783. Here, for example, terminal VMD is supplied with a high power supply potential (e.g., potential VDD) to operate drive circuit 722, and terminal VMS is supplied with a low power supply potential (e.g., potential VSS). Furthermore, for example, terminal VMH is supplied with a high power supply potential (e.g., a potential higher than potential VDD) to operate memory cell 741, etc. The on / off state of power switch 761 is controlled according to the signal supplied to terminal PON1. The on / off state of power switch 762 is controlled according to the signal supplied to terminal PON2.

[0655] In the drive circuit 722, the aforementioned circuits and terminals can be appropriately selected or omitted. Furthermore, other circuits and terminals can be added as appropriate.

[0656] <Example of a storage unit structure> An example of a structure that can be used in memory cell 741 will be described. As for the transistors included in the memory cell described below, OS transistors can be used. Furthermore, for example, at least a portion of transistors 500, 500A, 600, 600B, and 500F shown in Embodiment 2 above can be used.

[0657] Figure 23A The memory cell 950a shown includes a transistor M911 and a capacitor C911. One of the source and drain terminals of transistor M911 is connected to one terminal of capacitor C911. The other of the source and drain terminals of transistor M911 is connected to wiring BL, which serves as a bit line. The gate of transistor M911 is connected to wiring WL, which serves as a word line. The other terminal of capacitor C911 is connected to wiring CL. Note that sometimes the wiring connecting one of the source and drain terminals of transistor M911 to one terminal of capacitor C911 is referred to as wiring MN.

[0658] In memory cell 950a, data with two values ​​can be stored by setting the potential corresponding to the amount of charge stored in capacitor C911, i.e., the amount of charge held in wiring MN, to "1" or "0". Note that, for example, data with three or more values ​​can also be stored. Furthermore, when writing data to memory cell 950a, by controlling the conduction state of transistor M911, a potential corresponding to the data is supplied from wiring BL to wiring MN, thereby maintaining the charge corresponding to that potential. Furthermore, when reading data from memory cell 950a, by controlling the conduction state of transistor M911, the charge held in wiring MN can be extracted to wiring BL.

[0659] By reading data from storage cell 950a, the charge held in wiring MN is extracted to wiring BL, thereby changing the potential of wiring MN. In other words, reading data from storage cell 950a destroys the stored data. That is, when data is read from storage cell 950a, a destructive read is performed. Therefore, after reading data from storage cell 950a, a data write-back (refresh) is required.

[0660] In one aspect of the invention, for example, an n-channel OS transistor can be used as transistor M911.

[0661] Figure 23AThe shown memory cell 950a is a DRAM (Dynamic Random Access Memory) memory cell; specifically, the structure using OS transistors as transistor M911 is sometimes referred to as DOSRAM (registered trademark). DOSRAM uses OS transistors with extremely low off-state current, allowing for long-term data storage. Furthermore, it can store multi-valued or analog data. Additionally, since written data can be stored for extended periods, the data refresh rate can be reduced. Moreover, the cell size can be reduced because the electrostatic capacitance of the cell capacitor (capacitor C911) can be minimized. Therefore, by using DOSRAM, power consumption of semiconductor devices and memory devices can be reduced, while storage density can be increased.

[0662] Figure 23B The storage cell 950b shown is Figure 23A The storage cell 950a shown is a variant example, the difference being that the storage cell 950b does not include the capacitor C911.

[0663] exist Figure 23B In the storage cell 950b shown, charge can be stored in the parasitic capacitance (the electrostatic capacitance between the source and drain of transistor M911 and the gate), indicated by the dashed line. By adopting this structure, for example, the cell size can be reduced, thereby increasing the storage density of the semiconductor device and the storage device.

[0664] Figure 23C The memory cell 950c shown includes transistors M921 and M922, and capacitor C921. One of the source and drain terminals of transistor M921 is connected to the gate of transistor M922 and one terminal of capacitor C921. The other of the source and drain terminals of transistor M921 is connected to wiring WBL, used as a write bit line. The gate of transistor M921 is connected to wiring WWL, used as a write word line. One of the source and drain terminals of transistor M922 is connected to wiring RBL, used as a read bit line. The other of the source and drain terminals of transistor M922 is connected to wiring PL. The other terminal of capacitor C921 is connected to wiring RWL, used as a read word line. Note that sometimes the wiring connecting one of the source and drain terminals of transistor M921, the gate of transistor M922, and one terminal of capacitor C921 to each other is referred to as wiring MN.

[0665] In memory cell 950c, data with two values ​​can be stored by setting the potential corresponding to the amount of charge stored in capacitor C921, i.e., the amount of charge held in wiring MN, to "1" or "0". Note that, for example, data with three or more values ​​can also be stored. Furthermore, when writing data to memory cell 950c, by controlling the conduction state of transistor M921, a potential corresponding to the data is supplied from wiring WBL to wiring MN, thereby maintaining the charge corresponding to that potential. Furthermore, when reading data from memory cell 950c, by turning transistor M922 on or off according to the potential of wiring MN, the potential corresponding to that data can be extracted to wiring RBL.

[0666] In one embodiment of the invention, an n-channel OS transistor can be used as transistor M921, for example. Alternatively, an n-channel transistor (e.g., an OS transistor or a Si transistor, etc.) can be used as transistor M922, for example.

[0667] Figure 23C The shown memory cell 950c is a gain-cell type memory cell. In particular, the structure using an OS transistor as transistor M921 is sometimes referred to as NOSRAM (registered trademark). NOSRAM is short for Nonvolatile Oxide Semiconductor RAM. NOSRAM uses OS transistors with extremely low off-state current, allowing for long-term data storage. Furthermore, it can store multi-valued or analog data. Moreover, since the write transistor (transistor M921) and the read transistor (transistor M922) are different, non-destructive readout is performed. Therefore, it can be used, for example, as a non-volatile memory.

[0668] Figure 23D The storage cell 950d shown is Figure 23C The variations of memory cell 950c shown differ in that memory cell 950d does not include capacitor C921. Furthermore, in memory cell 950d, one of the source and drain terminals of transistor M922 is connected to wiring RWL.

[0669] exist Figure 23D In the shown memory cell 950d, charge can be stored in the parasitic capacitance attached to the wiring MN. By employing this structure, for example, the cell size can be reduced, thereby increasing the storage density of the semiconductor device and the memory device.

[0670] Figure 23E The storage cell 950e shown is Figure 23CThe variations of memory cell 950c shown differ in that, in memory cell 950e, one of the source and drain of transistor M921 is connected to wiring BL, and one of the source and drain of transistor M922 is connected to wiring BL.

[0671] exist Figure 23E In the memory cell 950e shown, the wiring BL can be used as both a write bit line and a read bit line. By adopting this structure, for example, the cell size can be reduced, thereby increasing the storage density of the semiconductor device and the memory device.

[0672] Figure 23F The storage cell 950f shown is Figure 23C The illustrated variation of memory cell 950c differs in that memory cell 950f includes transistor M922p instead of transistor M922. For example, a p-channel Si transistor can be used as transistor M922p.

[0673] exist Figure 23F In the memory cell 950f shown, since p-channel transistors are used to read data, the structure and operation of the readout amplifier can be simplified, for example. By adopting this structure, the layout area of ​​the drive circuit can be reduced, thereby enabling miniaturization of the semiconductor device and the memory device.

[0674] Figure 23G The storage unit shown is 950g. Figure 23C The illustrated variation of memory cell 950c differs in that memory cell 950g also includes transistor M923. One of the source and drain of transistor M922 is connected to one of the source and drain of transistor M923; the other of the source and drain of transistor M922 is connected to wiring PL; the other of the source and drain of transistor M923 is connected to wiring RBL; and the gate of transistor M923 is connected to wiring RWL. Furthermore, the other terminal of capacitor C921 is not connected to wiring RWL but to wiring CL. For example, an n-channel transistor (e.g., an OS transistor or a Si transistor, etc.) can be used as transistor M923.

[0675] exist Figure 23G In the memory cell 950g shown, the parasitic capacitance between wiring MN and wiring RBL can be reduced. By adopting this structure, for example, noise can be suppressed from entering wiring MN through the gate capacitance of transistor M922, thereby improving the reliability of the semiconductor device and the memory device.

[0676] Figure 23HThe shown memory cell 950h includes transistors M931, M932, M933, and M934, capacitors C931 and C932, inverter X931, and inverter X932. One of the source and drain terminals of transistor M931 is connected to one of the source and drain terminals of transistor M933, the input terminal of inverter X931, and the output terminal of inverter X932. One of the source and drain terminals of transistor M932 is connected to one of the source and drain terminals of transistor M934, the output terminal of inverter X931, and the input terminal of inverter X932. The other of the source and drain terminals of transistor M933 is connected to one terminal of capacitor C931. The other of the source and drain terminals of transistor M934 is connected to one terminal of capacitor C932. The other of the source and drain terminals of transistor M931 is connected to wiring BL, which serves as one of a pair of bit lines. One of the source and drain terminals of transistor M932 is connected to wiring BLB, which serves as the other of a pair of bit lines. The gates of transistors M931 and M932 are connected to wiring WL, which serves as a word line. The gates of transistors M933 and M934 are connected to wiring BRL. The other terminal of capacitor C931 and the other terminal of capacitor C932 are connected to wiring CL.

[0677] The storage cell 950h can store binary data ("1" or "0") in an inverter loop composed of inverters X931 and X932. Furthermore, when writing data to the storage cell 950h, by controlling the on / off states of transistors M931 and M932, a potential corresponding to the data can be supplied to the inverter loop from each of wirings BL and BLB. Furthermore, when reading data from the storage cell 950h, by controlling the on / off states of transistors M931 and M932, a potential corresponding to the data stored in the inverter loop can be extracted to each of wirings BL and BLB.

[0678] Furthermore, in the storage unit 950h, by controlling the conduction states of transistors M933 and M934, potentials corresponding to the data stored in the inverter loop are supplied to one terminal of capacitor C931 and one terminal of capacitor C932, respectively, thereby maintaining the charge corresponding to that potential. In other words, data backup is possible. Additionally, in the storage unit 950h, by controlling the conduction states of transistors M933 and M934, the charges held at one terminal of capacitor C931 and one terminal of capacitor C932, respectively, can be extracted into the inverter loop. In other words, data recovery is possible.

[0679] In one embodiment of the invention, for example, n-channel OS transistors can be used as transistors M931, M932, M933, and M934. Furthermore, inverter circuits prepared in standard circuit libraries can be used as inverters X931 and X932. That is, for example, n-channel and p-channel Si transistors can be used as the transistors constituting inverters X931 and X932.

[0680] Figure 23H The storage cell 950h shown is a storage cell of SRAM (Static Random Access Memory) that can be backed up. In particular, the structure that uses OS transistors as transistors M931, M932, M933 and M934 is sometimes called OS-SRAM (Oxide Semiconductor-SRAM).

[0681] Note that one aspect of the present invention is not limited to storage cells 950a to 950h, but may also be a combination of storage cells with appropriate structures.

[0682] Note that the contents described in this embodiment can be appropriately combined and implemented. Furthermore, the contents described in this embodiment can be appropriately combined and implemented in combination with contents described in other embodiments, etc.

[0683] (Implementation Method 5) In this embodiment, an example of the structure of a display module that can use a storage device according to one aspect of the present invention is described.

[0684] [Display Module] Figure 24A A perspective view of display module 880 is shown. Display module 880 includes display device 800A and FPC 890. Note that the display panel included in display module 880 is not limited to display device 800A, but may also be display device 800B, which will be described later.

[0685] The display module 880 includes a substrate 891 and a substrate 892. The display module 880 includes a display area 881. The display area 881 is an area for displaying images.

[0686] Figure 24BA perspective view of one side of the structure of substrate 891 is shown. A circuit section 882, a display section 883 on the circuit section 882, and a pixel section 884 on the display section 883 are stacked on the substrate 891. Furthermore, a terminal section 885 for connecting to an FPC 890 is provided on a portion of the substrate 891 that does not overlap with the pixel section 884. The terminal section 885 is connected to the circuit section 882 via a wiring section 886 composed of multiple wirings.

[0687] The pixel unit 884 includes a plurality of pixels 884a arranged periodically. Figure 24B The right side shows an enlarged view of pixel 884a. Pixel 884a includes a light-emitting element 810R that emits red light, a light-emitting element 810G that emits green light, and a light-emitting element 810B that emits blue light.

[0688] The display unit 883 includes a plurality of pixel circuits 883a arranged periodically. Each pixel circuit 883a controls the emission of light from three light-emitting elements included in a pixel 884a. A pixel circuit 883a may include three circuits controlling the emission of light from a single light-emitting element. For example, the pixel circuit 883a may employ a structure having at least one selection transistor, one current control transistor (driving transistor), and a capacitor for each light-emitting element. In this case, the gate of the selection transistor receives a gate signal, and the source receives a source signal. Thus, an active matrix display panel can be realized.

[0689] The circuit section 882 includes circuitry for driving each pixel circuit 883a of the display section 883. For example, it preferably includes one or both of a gate line driving circuit and a source line driving circuit. Furthermore, it may include at least one of an arithmetic circuit, a storage circuit, and a power supply circuit. Additionally, the transistors disposed in the circuit section 882 may also constitute part of the pixel circuit 883a. That is, the pixel circuit 883a may also be composed of transistors included in the display section 883 and transistors included in the circuit section 882.

[0690] For example, at least a portion of the storage device 100 shown in Embodiment 1 can be provided in the circuit section 882.

[0691] The FPC890 is used for wiring to supply video signals and power potentials to the circuit section 882 from the outside. Additionally, ICs can be mounted on the FPC890.

[0692] The display module 880 can have a structure in which a layer including the display section 883 and a layer including the circuit section 882 are stacked on the lower side of the pixel section 884, so that the display area 881 can have an extremely high aperture ratio (effective display area ratio). For example, the aperture ratio of the display area 881 can be 40% or more and less than 100%, preferably 50% or more and less than 95%, more preferably 60% or more and less than 95%. In addition, the pixels 884a can be arranged in an extremely high density, thereby enabling the display area 881 to have extremely high resolution. For example, the display area 881 preferably has a resolution of 2000 ppi or more, more preferably 3000 ppi or more, further preferably 5000 ppi or more, and even more preferably 6000 ppi or more and less than 20000 ppi or less or less than 30000 ppi.

[0693] This display module 880 has extremely high resolution, making it suitable for use in devices with spatial computing capabilities (also known as spatial computing devices, space computers, etc.), VR devices such as head-mounted displays, or AR devices such as glasses. For example, because the display module 880 has an extremely high-resolution display area 881, even when the display is magnified by a lens, the user cannot see any pixels when viewing the display section of the display module 880 through a lens, thus achieving a highly immersive display. Furthermore, the display module 880 is not limited to this; it can also be applied to electronic devices with relatively small display sections. For example, it is suitable for use in the display sections of wearable electronic devices such as watch-type devices.

[0694] [Display Device 800A] Figure 25 This is a cross-sectional view illustrating an example of the structure of the display device 800A. The display device 800A has a structure in which transistors 550 having channels formed in a substrate 311, transistors 500 having a semiconductor layer containing metal oxide forming the channels, capacitors 590, and light-emitting elements 810 (light-emitting element 810R, light-emitting element 810G, and light-emitting element 810B) are stacked.

[0695] Substrate 311 is equivalent to Figure 24A and Figure 24B Substrate 891 in the middle.

[0696] Notice, Figure 25 The structure shown is equivalent to that in embodiment 2 described above. Figure 4 The structure shown has a light-emitting element 810 positioned above it. Therefore, the above description can be used as appropriate, and detailed descriptions of transistor 550, transistor 500, and capacitor 590 are sometimes omitted.

[0697] An insulator 595 is provided to cover the capacitor 590, an insulator 854 is provided on the insulator 595, and an insulator 855 is provided on the insulator 854.

[0698] Insulators 595, 854, and 855 are suitable for using inorganic insulating films. For example, it is preferable to use a silicon oxide film as insulators 595 and 855, and a silicon nitride film as insulator 854. Thus, insulator 854 can be used as an etching protective film. Although an example is shown here where a portion of insulator 855 is etched to form a recess, it is also possible not to provide a recess in insulator 855.

[0699] The insulator 855 is provided with a light-emitting element 810R that emits red light, a light-emitting element 810G that emits green light, and a light-emitting element 810B that emits blue light.

[0700] Light-emitting element 810R includes a pixel electrode 811R on an insulator 855, an EL layer 812R covering the pixel electrode 811R, a portion of a common layer 814 on the EL layer 812R, and a portion of a common electrode 813 on the common layer 814. The EL layer 812R includes a light-emitting layer that emits red light. Light-emitting element 810G includes a pixel electrode 811G on an insulator 855, an EL layer 812G covering the pixel electrode 811G, a portion of a common layer 814 on the EL layer 812G, and a portion of a common electrode 813 on the common layer 814. The EL layer 812G includes a light-emitting layer that emits green light. Light-emitting element 810B includes a pixel electrode 811B on an insulator 855, an EL layer 812B covering the pixel electrode 811B, a portion of a common layer 814 on the EL layer 812B, and a portion of a common electrode 813 on the common layer 814. The EL layer 812B includes a light-emitting layer that emits blue light.

[0701] The display device 800A forms light-emitting elements separately for each light-emitting color, resulting in minimal chromaticity variation between low-brightness and high-brightness light emission. Furthermore, the EL layers 812R, 812G, and 812B are separated from each other, thus suppressing crosstalk between adjacent sub-pixels even when using a high-definition display panel. Therefore, a high-definition display panel with high display quality can be achieved.

[0702] Insulators 825 and 826 are provided in the area between adjacent light-emitting elements.

[0703] Inorganic insulating films can be used as insulators 825. Insulator 825 preferably has a barrier function against at least one of water and oxygen. When insulator 825 has the barrier function, it can have a structure that inhibits the entry of impurities (typically at least one of water and oxygen) that may diffuse from the outside into each light-emitting element. By adopting this structure, a highly reliable light-emitting element and a highly reliable display device can be provided.

[0704] The insulator 826 can be an insulator containing organic materials. As the organic material, a photosensitive organic resin is preferred, for example, a photosensitive resin composition including acrylic resin is preferred. Furthermore, a material that absorbs visible light can also be used as the insulator 826. By absorbing the light emitted from the light-emitting element through the insulator 826, light leakage from the light-emitting element to adjacent light-emitting elements (stray light) can be suppressed. Therefore, the display quality of the display device can be improved. In addition, even without using a polarizer in the display device, the display quality can be improved, thus enabling the display device to be lightweight and thin.

[0705] Pixel electrodes 811R, 811G, and 811B are connected to conductor 591 via a conductor 856, which serves as a connector, embedded in insulators 592, 595, 854, and 855. They are also connected to one of the source and drain terminals of transistor 500 via a conductor 540, which also serves as a connector. The height of the top surface of insulator 855 is aligned with the height of the top surface of conductor 856. Sometimes, pixel electrodes 811R, 811G, and 811B are collectively referred to as pixel electrode 811.

[0706] No insulator covering the top end of the pixel electrode 811 is provided between two adjacent pixel electrodes 811. Therefore, the spacing between adjacent light-emitting elements can be very small. Thus, a high-definition or high-resolution display device can be realized.

[0707] A protective layer 821 is provided on the light-emitting elements 810R, 810G, and 810B. A substrate 870 is attached to the protective layer 821 by an adhesive layer 871.

[0708] Substrate 870 is equivalent to Figure 24A The substrate 892 in the middle.

[0709] In the display device 800A, for example, transistor 550 can be used in the circuit included in the circuit section 882, and transistor 500 can be used in the circuit included in the display section 883. Note that transistor 500 can also be used in a part of the circuit included in the circuit section 882.

[0710] Alternatively, transistors of various structures, such as transistors 500A, 600, 600B, and 500F shown in Embodiment 2 above, can be used to replace transistor 500.

[0711] By adopting this structure, a display section 883 including pixel circuitry and a circuit section 882 including driving circuitry can be formed directly below the pixel section 884, which includes light-emitting elements. Therefore, compared to the case where driving circuitry is arranged around the periphery of the display area, miniaturization of the display panel can be achieved.

[0712] [Display device 800B] Figure 26 This is a cross-sectional view illustrating an example of the structure of display device 800B. In display device 800B, transistors 500_1 and 500_2, both containing metal oxide in the semiconductor layer forming a channel, are stacked. By employing this structure, different transistors can be formed using the constituent elements of the metal oxide in the semiconductor layer. Therefore, display devices using OS transistors with different transistor characteristics can be obtained.

[0713] Note that in the accompanying drawings, the structure of transistor 500 is shown as transistor 500_1 and transistor 500_2, but it is not limited to this. Various transistor structures can be used, such as those shown in Embodiment 2 above for transistors 500A, 600, 600B, and 500F. Furthermore, transistors 500_1 and 500_2 can each use transistors with different structures.

[0714] In the display device 800B, for example, transistors 550 and 500_1 are used in the circuit included in the circuit section 882, and transistor 500_2 is used in the circuit included in the display section 883.

[0715] By adopting this structure, the circuitry (including the pixel circuitry included in the display section 883 and the driving circuitry included in the circuit section 882, etc.) located directly below the pixel section 884 including the light-emitting element can be configured at a higher density, thereby enabling the display panel to be miniaturized compared to the case where the driving circuitry is located near the display area.

[0716] Note that the contents described in this embodiment can be appropriately combined and implemented. Furthermore, the contents described in this embodiment can be appropriately combined and implemented in combination with contents described in other embodiments, etc.

[0717] (Implementation Method 6) In this embodiment, an example of a circuit structure including an OS transistor according to one aspect of the present invention is described.

[0718] <Circuit Structure Example> Figure 27A and Figure 27B An example of a circuit structure of an OS transistor including one aspect of the present invention is shown. Figure 27A The circuit diagram shown illustrates the structure of an inverter circuit constructed using so-called CMOS (Complementary Metal Oxide Semiconductor) circuitry, where n-channel transistor 3102 and p-channel transistor 3104 are connected in series with their gates connected. Additionally, Figure 27B The circuit diagram shown illustrates the structure of a circuit used as a so-called analog switch, with the source and drain connections of the n-channel transistor 3102 and the p-channel transistor 3104.

[0719] <Transistor Structure> As Figure 27A and Figure 27B The n-channel transistor 3102 and p-channel transistor 3104 shown can be made of various types of transistors. Specifically, planar transistors, vertical field-effect transistors (VFETs), fin transistors, and GAA (Gate All Around) transistors can be used as the n-channel transistor 3102 and p-channel transistor 3104. Alternatively, a CFET (Complementary Field Effect Transistor) combining the n-channel transistor 3102 and the p-channel transistor 3104 can also be used.

[0720] In this specification, a planar transistor refers to a structure where the source and drain electrodes are at the same height and the current flowing through the semiconductor has a lateral component. Furthermore, in this specification, a VFET refers to a structure where the source and drain electrodes are at different heights and the current flowing through the semiconductor has a height-oriented component. Because a VFET can stack two or more of the source, semiconductor, and drain electrodes, its footprint can be significantly reduced compared to a planar transistor.

[0721] Furthermore, in this specification, a Fin-type transistor refers to a structure in which, in a cross-sectional view along the channel width direction, the gate electrode covers two or more sides of the channel through a gate insulating film. In particular, in a cross-sectional view along the channel width direction, when the channel height (H) is greater than the channel width (W), the channel width per unit area can be increased, which is therefore preferred. Furthermore, in this specification, a GAA-type transistor refers to a structure in which, in a cross-sectional view along the channel width direction, the gate electrode covers all four sides of the channel through a gate insulating film. [0...

Claims

1. A storage device, comprising: Check the bit generation unit; Storage Department; as well as Error Detection and Correction Department The storage unit includes a trigger circuit and a holding circuit. The check bit generation unit has the function of generating a check series from the information series. The storage unit has the function of storing the codeword composed of the information series and the check series in the trigger circuit and outputting it as a receive word. The error detection and correction unit has the function of outputting an error vector representing the error position of the received word and a decoded word in which the error of the received word has been corrected. The storage unit has the functions of storing the decoded word in the trigger circuit according to the error vector, writing the decoded word to the holding circuit, and writing the decoded word written to the holding circuit back to the trigger circuit. The trigger circuit includes a first transistor. The holding circuit includes a second transistor. Furthermore, the off-state current of the second transistor is less than the off-state current of the first transistor.

2. The storage device according to claim 1, The codewords mentioned are Hamming codes.

3. The storage device according to claim 1 or 2, The change in electrical characteristics of the second transistor due to radiation is smaller than that of the first transistor due to radiation.

4. The storage device according to claim 1 or 2, The second transistor is stacked on the layer on which the first transistor is disposed.

5. The storage device according to claim 1 or 2, The first transistor contains silicon in the channel formation region. Furthermore, the second transistor contains an oxide semiconductor in the channel formation region.

Images