A CMOS inverter device structure and fabrication method based on zigzag nanowires

By employing a CMOS inverter device structure with closely packed zigzag nanowires and combining PECVD and CVD technologies, p-type and n-type nanowires are fabricated in situ, solving the problems of single-component nanowires and high vertical height in existing technologies, and realizing low-cost and high-efficiency CMOS inverter fabrication.

CN123124437APending Publication Date: 2026-11-17SUZHOU VOCATIONAL UNIVERSITY (SUZHOU OPEN UNIVERSITY)
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202610315035.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-16
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In existing CMOS inverter devices, nanowires are a single component, making it difficult to realize CMOS inverter devices with the same projected area. Furthermore, the existing structures have a high vertical height, making etching difficult and the fabrication cost high.

Method used

A CMOS inverter device structure based on close-packed zigzag nanowires is adopted. The nanowires are zigzag structures, combining p-type and n-type nanowires into one unit. They are prepared in situ by PECVD or CVD system, and p-type and n-type nanowires are grown by H2 plasma etching and vacuum annealing technology, which simplifies the preparation process.

Benefits of technology

This reduces the vertical height of the device, makes etching more difficult, saves on fabrication costs, and increases the on-state current of the device, thus achieving high uniformity and low cost in nanowire fabrication.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123124437A_ABST
    Figure CN123124437A_ABST
Patent Text Reader

Abstract

This invention discloses a CMOS inverter device structure and its fabrication method based on zigzag nanowires. The device includes a dielectric layer, electrodes, nanowires, a gate dielectric, and a gate electrode. The nanowires have a zigzag structure, specifically p-type and n-type nanowires, which together form the zigzag structure. Addressing the issues of existing technologies requiring separate fabrication of PMOS and NMOS, resulting in high fabrication costs, and the limited performance of nanowire-based CFET (Complementary Field-Effect Transistor) devices with only two conductive channels, this invention proposes a CMOS inverter device structure based on closely packed zigzag nanowires. In this structure, one NMOS and one PMOS are vertical channel structures, while the other is a planar device structure. This reduces the vertical height, simplifies etching, and allows the pn nanowires to be fabricated in a single step, significantly saving fabrication costs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of semiconductor technology and relates to a CMOS inverter device structure based on zigzag nanowires and its fabrication method. It can reduce the vertical height, reduce the etching difficulty, and the pn nanowires can be fabricated in one step, which greatly saves the fabrication cost. Background Technology

[0002] CMOS (Complementary Metal-Oxide-Semiconductor) technology is a core technology in current integrated circuit manufacturing. Its core advantages include extremely low power consumption, high noise immunity, and ultra-high integration density, enabling the integration of analog and digital circuits onto a single chip to achieve complex system-level functions. CMOS technology achieves efficient logic operations through the complementary structure of P-type and N-type semiconductors and is widely used in microprocessors, memory chips, radio frequency circuits, image sensors, and other fields. Existing technology provides a field-effect transistor structure based on zigzag nanowires. In this technology, silicon nanowires grown from bottom to top by nano-metal droplets can be arranged along a predetermined path, and the nanowire portion perpendicular to the substrate can serve as the conductive channel of the field-effect transistor, reducing the vertical projected area of ​​the transistor. However, in this technology, the nanowires are a single component, either p-type or n-type, making it difficult to realize CMOS inverter devices within the same projected area. Furthermore, this technology proposes to use stacked cycles of silicon nitride and silicon oxide on the sidewalls to prepare multiple zigzag nanowire conductive channels. If transistors are fabricated based on this structure of closely packed nanowires, the channel length of each nanowire on one side is different, and the source and drain electrodes are difficult to fabricate. Existing technologies provide a vertical-channel CFET device structure that enables the fabrication of small-sized CFET devices without relying on high-precision photolithography. However, this technology uses a thin-film channel structure, which has poorer electrostatic control compared to nanowire channel structures. Furthermore, the entire device has a vertical height equivalent to five thin-film layers, three of which are metal films, making etching difficult. Existing technologies also provide a nanowire-based vertical-channel CFET device structure, but in a single CFET device, both PMOS and NMOS have only two nanowires (distributed horizontally), limiting the on-state current of the CFET device. The overall device unit height is also relatively high, making etching difficult. In summary, in existing technologies, the vertical-channel CFET structure, with its thin-film conductive channel, has weaker gate control capability compared to nanowire structures. Moreover, the existing structures have a high vertical depth, making etching difficult in small-sized devices. Therefore, new technologies are needed to reduce etching difficulty and save fabrication costs. Summary of the Invention

[0003] To address the issues that existing technologies such as PMOS and NMOS require separate fabrication, resulting in high fabrication costs, and that CFET (Complementary Field-Effect Transistor) devices based on nanowire structures have only two conductive channels, limiting device performance, this invention proposes a CMOS inverter device structure based on closely packed zigzag nanowires. This structure can reduce vertical height, decrease etching difficulty, and allow the pn nanowires to be fabricated in a single step, significantly saving fabrication costs.

[0004] The present invention adopts the following technical solution.

[0005] A CMOS inverter device structure based on close-packed zigzag nanowires includes a dielectric layer, electrodes, nanowires, a gate dielectric, and a gate electrode; the nanowires have a zigzag structure.

[0006] In this invention, a zigzag structure of nanowires means that the nanowires are not vertical or horizontal, that is, the nanowires have an angle (neither 0° nor 180°), preferably, the nanowires have right angles.

[0007] In this invention, the nanowires are p-type nanowires and n-type nanowires, which together form a zigzag structure. Preferably, each of the p-type and n-type nanowires is a straight nanowire, and the two are combined to form a zigzag structure, that is, the nanowire as a whole is a zigzag structure. More preferably, the p-type and n-type nanowires are an integral structure.

[0008] In this invention, the dielectric layer includes a second dielectric layer and a first dielectric layer arranged vertically; preferably, the first dielectric layer has an L-shaped structure. More preferably, the L-shaped structure has an opening that fills the electrode.

[0009] In this invention, the first dielectric layer and the second dielectric layer are arranged vertically with the second dielectric layer on top and the first dielectric layer on the bottom. Furthermore, the first dielectric layer and the second dielectric layer arranged vertically may or may not be in contact, but it is preferred that they are in contact.

[0010] In this invention, the dielectric layer and the electrode are located inside the nanowire.

[0011] In this invention, the thickness of the first dielectric layer and the second dielectric layer is in the nanometer dimension or can be in the micrometer dimension. The thicknesses of the two dielectric layers can be the same or different. The materials of the first dielectric layer and the second dielectric layer are independently selected from insulating dielectric materials such as silicon nitride, silicon oxide, silicon oxynitride, aluminum oxide, hafnium oxide, zirconium oxide, hafnium zirconium oxide, hafnium lanthanum oxide, and quartz, and the two dielectric layers are made of different materials. Preferably, the widths of the thicknesses of the first dielectric layer and the second dielectric layer are different.

[0012] In this invention, electrodes are respectively provided on the upper and lower parts of the dielectric layer; preferably, electrodes are provided on both sides of the second dielectric layer and on one side of the first dielectric layer. As an example, the first dielectric layer has an L-shaped structure, and the opening of the L-shaped structure is filled with electrodes. The electrodes are source / drain contact electrodes of n-type nanowires and p-type nanowires, with a thickness at the nanometer or micrometer level. Their materials can be single metal materials such as titanium, platinum, gold, copper, aluminum, nickel, and shale, or mixed metal materials such as platinum, nickel-gold, nickel-aluminum, and titanium-gold, or compound conductive materials such as titanium nitride, polycrystalline silicon, and heavily doped polycrystalline silicon, or a stacked combination of the above materials; wherein, different conductive electrodes can be made of the same material or different materials.

[0013] In this invention, a gate dielectric and a gate electrode are provided on the outer side of the nanowire.

[0014] In this invention, the gate includes a Fin structure or a GAA structure. The thickness of the gate is at the nanometer or micrometer level, and its material can be a single metal material such as titanium, platinum, gold, copper, aluminum, nickel, or shale, or a mixed metal material such as platinum, nickel-gold, nickel-aluminum, or titanium-gold, or a compound conductive material such as titanium nitride, polycrystalline silicon, or heavily doped polycrystalline silicon, or a stacked combination of the above materials.

[0015] In this invention, the gate dielectric material can be insulating dielectric materials such as silicon nitride, silicon oxide, silicon oxynitride, aluminum oxide, hafnium oxide, zirconium oxide, hafnium zirconium oxide, hafnium lanthanum oxide, hafnium aluminum oxide, or their laminated materials.

[0016] In this invention, there is a plurality of nanowires, the plurality of which is 4n, where n is 1 or an integer greater than 1, such as 4 or more nanowires, and further, 8 or more nanowires; preferably, there is a gap between adjacent nanowires; for example, there is a first dielectric layer or a second dielectric layer between adjacent nanowires, that is, adjacent nanowires are separated by a first dielectric layer or a second dielectric layer, or there may be a gate between adjacent nanowires.

[0017] In this invention, the nanowires are made of silicon, but can also be III-V group materials.

[0018] In this invention, the CMOS inverter device structure based on closely packed zigzag nanowires further includes a third dielectric layer, on both sides of which nanowires are disposed.

[0019] In this invention, p-type nanowires and n-type nanowires correspond to PMOS (P-type metal-oxide-semiconductor field-effect transistor) and NMOS (N-type metal-oxide-semiconductor field-effect transistor), respectively. This CMOS device structure based on closely packed zigzag nanowires has one NMOS with a vertical channel structure and the other with a planar device structure. The conductive channel length of the vertical channel structure depends on the thickness of the first dielectric layer, while the conductive channel length of the planar device structure depends on the time (formation width) of the second dielectric layer obtained by wet etching. Both MOS structures can be realized without high-precision photolithography.

[0020] In this invention, the number of close-packed nanowires is determined by the number of cycles of depositing dielectric layer 1 (first dielectric layer) and dielectric layer 2 (second dielectric layer) on the sidewalls. For example, if dielectric layer 1 and dielectric layer 2 are deposited only once on the sidewalls of the third dielectric layer, the number of channels is 4. If dielectric layer 1 and dielectric layer 2 are deposited twice, the number of channels is 8. If dielectric layer 1 and dielectric layer 2 are deposited n times, the number of channels is 4n, forming close-packed nanowires. Wet etching of dielectric layer 2 forms nanowire channels. The morphology or diameter of the nanowires is determined by the width and thickness of dielectric layer 2. The nanowires have columnar or nanosheet structures and high uniformity.

[0021] This invention first deposits n-type amorphous silicon. Taking advantage of the lower etching rate of the n-type amorphous silicon film in the vertical groove compared to the n-type amorphous silicon film in the planar groove, the n-type amorphous silicon film in the top groove can be etched in situ using H2 plasma in PECVD. Then, intrinsic amorphous silicon is deposited, followed by annealing in a vacuum environment to achieve the growth of pn junction nanowires. This saves on preparation costs and reduces the difficulty of growth. In actual production, this growth method can also be used to prepare nanowires of other materials.

[0022] This invention discloses a method for fabricating the above-mentioned CMOS inverter device structure based on close-packed zigzag nanowires, comprising the following steps: (1) A lower first dielectric layer, a second dielectric layer and an upper first dielectric layer are sequentially disposed on the lower conductive layer, wherein the width of the second dielectric layer is smaller than that of the first dielectric layer; (2) Conductive layers are respectively disposed on both sides of the second dielectric layer between the upper first dielectric layer and the lower first dielectric layer; (3) A vertical first dielectric layer is provided on the sidewall of the lower conductive layer below the lower first dielectric layer, and the upper first dielectric layer is removed; (4) A third dielectric layer is set in the middle of the structure obtained in step (3), and then the first dielectric layer is deposited and etched back in sequence, the second dielectric layer is deposited and etched back; the structure with the third dielectric layer in the middle and the first and second dielectric layers on both sides are arranged alternately is repeated; then the second dielectric layer is etched to form a channel; (5) Deposit catalytic metal and form metal droplets on the channel wall, then grow p-type nanowires and n-type nanowires; then prepare the gate dielectric and gate to obtain the above-mentioned CMOS inverter device structure based on close-packed zigzag nanowires; or, (6) Deposit catalytic metal and form metal droplets on the channel wall, then grow p-type nanowires and n-type nanowires; then etch part of the first dielectric layer and the second dielectric layer, and then prepare the gate dielectric and the gate to obtain the above-mentioned CMOS inverter device structure based on close-packed zigzag nanowires.

[0023] In this invention, the above structure is achieved by deposition and etching (photolithography, wet etching, etch-back, etc.), which are conventional techniques in the field.

[0024] In this invention, a third dielectric layer is provided in the middle of the structure obtained in step (3). The middle refers to the middle of the sidewall and the middle of the upper surface of the structure. As an example, the third dielectric layer is framed in the middle of the structure obtained in step (3) on three sides. The first dielectric layer and the second dielectric layer subsequently deposited also wrap the structure obtained in step (3) on three sides.

[0025] In this invention, in step (5), the catalytic metal is indium, tin, or their alloy; a metal droplet is formed by plasma treatment. The catalytic metal is deposited and a metal droplet is formed on the channel wall. Then, a doped amorphous silicon precursor is deposited and etched back, then intrinsic amorphous silicon is deposited, and then vacuum annealing is performed to achieve the growth of p-type nanowires and n-type nanowires. This invention can directly and in-situ prepare p-type and n-type nanowires in a growth system (e.g., in PECVD or CVD). The process is simple, the preparation cost is low, and the uniformity is good.

[0026] A memory comprising the aforementioned CMOS inverter device structure based on close-packed zigzag nanowires.

[0027] An electronic device comprising the aforementioned CMOS inverter device structure based on closely packed zigzag nanowires.

[0028] In this invention, the PMOS and NMOS devices are respectively a vertical channel structure and a planar structure. The conductive channel length of the planar structure is determined by the time required for wet etching of the second dielectric layer, which does not rely on high-precision photolithography. This invention increases the number of transistor channels and improves the on-state current of the device without significantly increasing the projected area. Moreover, it can directly achieve the fabrication of p-type and n-type nanowires in one step using a simple growth technique, resulting in a simple process and low fabrication cost. In particular, the overall device height is lower than that of existing technologies, and the etching difficulty is reduced. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of a nanowire structure.

[0030] Figure 2 This is a schematic diagram of the CMOS inverter device structure in Example 1.

[0031] Figure 3 This is a schematic diagram of the internal structure of a CMOS inverter device in Example 1, omitting the gate dielectric and gate.

[0032] Figure 4 The circuit schematic for the structure is shown.

[0033] Figure 5 This is a flowchart illustrating the fabrication process of the CMOS inverter device structure in Example 1.

[0034] Figure 6 This is a schematic diagram of the CMOS inverter device structure in Example 2.

[0035] Figure 7 This is a schematic diagram of the internal structure of the CMOS inverter device in Example 2, omitting the gate dielectric and gate.

[0036] Figure 8 This is a flowchart of the fabrication process of the CMOS inverter device structure in Example 2 (last two steps). Detailed Implementation

[0037] Inverters are frequently used in electronic circuit design. An inverter can reverse the phase of an input signal by 180 degrees. This type of circuit is applied in analog circuits, such as audio amplifiers and clock oscillators. A CMOS inverter circuit consists of two enhancement-mode MOS field-effect transistors. In existing CMOS inverter designs, the conductive channels of vertical and thin-film structures have weaker gate control capabilities compared to nanowires. Furthermore, the existing structures have a high vertical depth, making etching difficult in small-size devices. This invention proposes a CMOS inverter device structure based on close-packed zigzag nanowires, which can reduce the vertical height, reduce etching difficulty, and allow for the fabrication of pn nanowires in a single step. Moreover, this invention can directly achieve in-situ fabrication of p-type and n-type nanowires in a growth system (e.g., PECVD or CVD), resulting in a simple process with good uniformity and significantly reduced fabrication costs.

[0038] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. The accompanying drawings are for illustrative purposes only and should not be construed as limiting the invention. To better illustrate this embodiment, some components in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings. The positional relationships described in the accompanying drawings are for illustrative purposes only and should not be construed as limiting the invention. In the description of the invention, it should be understood that the indicated orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings, with the inner and outer orientations defined by the center point of the device. This is only for the convenience of describing the invention and simplifying the description. Those skilled in the art can form a device and achieve the technical effects described in the present invention based on the layout of the present invention and combined with conventional electrode connection methods.

[0039] The innovation of this invention lies in proposing a CMOS inverter device structure based on close-packed zigzag nanowires. The material selection and specific fabrication operations of each component (layer) are conventional techniques and do not affect the understanding of the technical effects of this invention by those skilled in the art.

[0040] This invention discloses a CMOS inverter device structure based on closely packed zigzag nanowires, including a dielectric layer, electrodes, nanowires, a gate dielectric, and a gate electrode; the nanowires are zigzag structures. A zigzag structure means that the nanowires are not vertical or horizontal, i.e., the nanowires have an angle (neither 0° nor 180°), preferably a right angle. The nanowires are p-type and n-type nanowires, forming a zigzag structure. Each p-type and n-type nanowire is a straight nanowire, and their combination constitutes the zigzag structure. Furthermore, the p-type and n-type nanowires are an integral structure. This invention allows for the direct in-situ fabrication of p-type and n-type nanowires in a growth system (e.g., PECVD or CVD), resulting in a simple process with good uniformity. Specifically, in a plasma-enhanced chemical deposition system, a metal catalyst is treated with H2 plasma to form a metal catalyst droplet at 200°C. The temperature is then lowered to 100-150°C for in-situ deposition of doped amorphous silicon. The temperature is then lowered to below 100°C, and the doped amorphous silicon is etched with H2 plasma. Due to the shielding effect of the grooves, the doped amorphous silicon on the sidewalls is etched, while the doped amorphous silicon within the grooves is retained. The temperature is then raised to 100-150°C to deposit intrinsic amorphous silicon. After deposition, at 350-400°C in a vacuum or N2 environment, the metal catalyst droplet melts and gradually absorbs the precursor, resulting in the growth of crystalline silicon nanowires. Therefore, silicon nanowires with the same doping type as the precursor grow in regions containing the doped precursor, while silicon nanowires with the same doping type as the metal doped precursor grow in regions without the precursor, thus preparing regionally doped silicon nanowires.

[0041] In this invention, the gate control structure of NMOS and PMOS can be a Fin-like structure or a GAA structure, such as... Figure 2 and Figure 6 As shown.

[0042] This invention discloses a method for fabricating the above-mentioned CMOS inverter device structure based on close-packed zigzag nanowires, comprising the following steps: (1) A lower first dielectric layer, a second dielectric layer and an upper first dielectric layer are sequentially disposed on the lower conductive layer, wherein the width of the second dielectric layer is smaller than that of the first dielectric layer; (2) Conductive layers are respectively disposed on both sides of the second dielectric layer between the upper first dielectric layer and the lower first dielectric layer; (3) A vertical first dielectric layer is provided on the sidewall of the lower conductive layer below the lower first dielectric layer, and the upper first dielectric layer is removed; (4) A third dielectric layer is set in the middle of the structure obtained in step (3), and then the first dielectric layer is deposited and etched back in sequence, the second dielectric layer is deposited and etched back; the structure with the third dielectric layer in the middle and the first and second dielectric layers on both sides are arranged alternately is repeated; then the second dielectric layer is etched to form a channel; (5) Deposit catalytic metal and form metal droplets on the channel wall, then grow p-type nanowires and n-type nanowires; then prepare the gate dielectric and gate to obtain the above-mentioned CMOS inverter device structure based on close-packed zigzag nanowires, with the gate control structure being a Fin structure; or, (6) Deposit catalytic metal and form metal droplets on the channel wall, then grow p-type nanowires and n-type nanowires; then etch part of the first dielectric layer and the second dielectric layer, and then prepare the gate dielectric and the gate to obtain the above-mentioned CMOS inverter device structure based on close-packed zigzag nanowires, with the gate control structure being a GAA structure.

[0043] In this invention, conductive layers are respectively disposed on both sides of the second dielectric layer to form a left conductive layer and a right conductive layer. The dimensions of the left and right conductive layers can be the same or different. The materials of the lower conductive layer, the left conductive layer, and the right conductive layer can be the same or different materials. They can be single metal materials such as titanium, platinum, gold, copper, aluminum, nickel, and shale, or mixed metal materials such as platinum, nickel-gold, nickel-aluminum, and titanium-gold, or compound conductive materials such as titanium nitride, polycrystalline silicon, and heavily doped polycrystalline silicon, or a stacked combination of the above materials.

[0044] In this invention, the width of the second dielectric layer determines the conductive channel length of the planar device, and the height of the first dielectric layer determines the channel length of the vertical device. The range is nanometer-sized or micrometer-sized. The width of the second dielectric layer and the height of the first dielectric layer can be the same or different, and can be precisely controlled according to actual needs. The material can be insulating dielectric materials such as silicon nitride, silicon oxide, silicon carbide, silicon oxynitride, aluminum oxide, hafnium oxide, zirconium oxide, hafnium zirconium oxide, hafnium lanthanum oxide, quartz, etc. The first dielectric layer and the second dielectric layer are different dielectric materials.

[0045] The n-type and p-type nanowires of this invention form a zigzag morphology, and their cross-section can be circular or square, such as... Figure 1 As shown, they form nanocylindrical and nanostrip structures, respectively; the materials can be silicon nanowires or group III-V nanowire materials.

[0046] The gate dielectric of the present invention may be made of insulating dielectric materials such as silicon nitride, silicon oxide, silicon oxynitride, aluminum oxide, hafnium oxide, zirconium oxide, hafnium zirconium oxide, hafnium lanthanum oxide, hafnium aluminum oxide, or stacked materials thereof.

[0047] The gate metal of the present invention has a thickness at the nanometer or micrometer level, and its material can be a single metal material such as titanium, platinum, gold, copper, aluminum, nickel, or shale, or a mixed metal material such as platinum, nickel-gold, nickel-aluminum, or titanium-gold, or a compound conductive material such as titanium nitride, polycrystalline silicon, or heavily doped polycrystalline silicon, or a stacked combination of the above materials.

[0048] The catalytic metal of this invention includes indium, tin, Bi, or their alloys, and the catalytic metal is p-type or n-type doped in the nanowires. Metal droplets are formed using plasma treatment. In the above steps, the catalytic metal is deposited and metal droplets are formed on the channel wall. Then, a doped amorphous silicon precursor is deposited (the doping type of the precursor is opposite to the metal doping type; for example, if the metal catalyst is p-type doped in the nanowire, then an n-type doped amorphous precursor is deposited) and etched back. Then, an intrinsic amorphous silicon precursor is deposited again, followed by vacuum annealing. The region where the doped amorphous silicon precursor is deposited will cause the original p-type nanowires to transform into n-type nanowires, thereby achieving in-situ growth of p-type and n-type nanowires.

[0049] In this invention, in steps (5) / (6) above, a catalytic metal is deposited and a metal droplet is formed on the channel wall. Then, a doped amorphous silicon precursor is deposited and etched back. Intrinsic amorphous silicon is then deposited and vacuum annealed to achieve the growth of p-type nanowires and n-type nanowires. This invention can directly and in-situ realize the preparation of p-type and n-type nanowires in a growth system (e.g., in PECVD or CVD). The process is simple, the preparation cost is low, and the uniformity is good.

[0050] As a specific example, in a plasma-enhanced chemical deposition system, a metal catalyst is treated with H2 plasma to form metal catalyst droplets at a temperature of 200°C. Then, doped amorphous silicon is deposited in situ at 100-150°C. Next, the doped amorphous silicon is etched with H2 plasma at a temperature below 100°C. Due to the shielding effect of the grooves, the doped amorphous silicon on the sidewalls is etched, while the doped amorphous silicon in the grooves is retained. Then, intrinsic amorphous silicon is deposited at 100-150°C. After deposition, at around 400°C in a vacuum or N2 environment, the metal catalyst droplets melt and gradually absorb the precursor, and crystalline silicon nanowires grow behind it. Therefore, silicon nanowires with the same doping type as the doped precursor will grow in the region containing the doped precursor, while the silicon nanowires grown in the region without the doped precursor will have the same doping type as the metal doping, thus preparing regionally doped silicon nanowires.

[0051] In this invention, the above structure is achieved by deposition and etching (photolithography, wet etching, etch-back, etc.), which are conventional techniques in the field; the thickness of each layer (component) in the device is selected according to conventional needs and does not affect the realization of the technical effect of this invention. Example 1

[0052] See Figure 2 The figure shows the overall structure of the CMOS inverter device of the present invention. Figure 200 shows the device, where 210 is a p-type transistor, 220 is an n-type transistor, and 201 and 202 are the gate dielectric and gate of the CMOS inverter device, respectively.

[0053] See Figure 3 In this diagram, 100 represents the internal structure of the CMOS device based on closely packed zigzag nanowires as conductive channels. 102 and 101 are dielectric layer 1 (first dielectric layer) and dielectric layer 2 (second dielectric layer), respectively. The first dielectric layer 102 has an L-shaped structure, and 101 has a cuboid structure. 103, 104, and 105 are the metal electrodes of the CMOS device. 103 serves as the source / drain contact electrode of the n-type nanowire, connected to the GND (low level) of the CMOS inverter device. 104 serves as the common source / drain electrode of the n-type FET and p-type FET, connected to the Vout of the CMOS inverter device. 105 serves as the source / drain contact electrode of the p-type nanowire, connected to the Vdd (high level) of the CMOS inverter device. 106 and 107 are the n-type nanowire and p-type nanowire, respectively.

[0054] Figure 4The circuit schematic for the entire structure described above shows that when Vin is connected to a high level, the NMOS transistor is turned on, the PMOS transistor is turned off, and Vout outputs GND (low level). When Vin is connected to a low level, the NMOS transistor is turned off, the PMOS transistor is turned on, and Vout outputs Vdd (high level).

[0055] As shown in the figure, dielectric layers 101 and 102 are made of silicon nitride and silicon oxide, respectively; source and drain contact electrodes 103, 104, and 105 are n-type and p-type nanowires, respectively, and are made of platinum; gate dielectric 201 is made of hafnium oxide; and gate metal 202 is made of aluminum.

[0056] As shown in the figure, 106 and 107 are n-type nanowires and p-type nanowires, respectively, with a zigzag morphology, specifically a right-angled three-sided frame structure, and a circular cross-section, forming a nanocylinder. In this embodiment, silicon nanowires are selected.

[0057] The fabrication process of the above CMOS inverter device is described in [reference needed]. Figure 5 The following steps are performed sequentially: (1) A metal layer (lower conductive layer), dielectric layer 1, dielectric layer 2 and dielectric layer 1 are sequentially deposited on a conventional substrate; (2) The first vertical groove is patterned and etched, and then the dielectric layer 2 is wet-etched to shrink it between the two dielectric layers 1, forming the structure shown in the figure. (3) ALD deposits metal (left conductive layer) and etches it back; (4) A medium layer 3 is filled inside the first vertical groove as groove protection; (5) Pattern and etch to form the second vertical groove, then wet etch metal layer 1 in sequence, fill the etched vacancy with dielectric layer 1, etch back, wet etch dielectric layer 2, fill the etched vacancy with metal (right conductive layer), etch back, to form the structure shown in the figure. (6) Etching to remove the groove protective dielectric layer 3 and the top dielectric layer 1; (7) Deposit medium layer 4 (third medium layer), pattern, etch, the third medium layer is like a three-sided frame in the middle of the structure obtained in step (6) to form the structure shown in the figure; (8) Dielectric layer 1 is deposited on both sides of dielectric layer 4 in sequence, etched back, dielectric layer 2 is deposited, and etched back. This process is repeated once per cycle, resulting in 4 more conductive channels for PMOS and NMOS. (9) Patterning and etching are performed to make the dielectric layer 1, dielectric layer 2 and dielectric layer 4 deposited in step (8) flush with each other, forming the structure shown in the figure; (10) Wet etching of dielectric layer 2 forms trenches; (11) Deposit catalytic metal In and process it into In droplets by plasma in PECVD; (12) n-type doped amorphous silicon was deposited in situ in PECVD and etched in situ in PECVD with H2 plasma to form the structure shown in the figure, where the purple film is doped amorphous silicon. (13) In situ deposition of intrinsic amorphous silicon in PECVD; (14) Nanowires are annealed and grown in a vacuum environment at 350℃. The nanowires that tend to grow in the region containing n-type doped amorphous silicon are n-type, and the nanowires that grow in the region containing only intrinsic amorphous silicon are p-type. (15) Deposit HK medium and metal gate, the gate control structure is Fin structure, also known as Ω structure; A CMOS inverter device was obtained. Example 2

[0058] The CMOS inverter device structure is the same as in Example 1, except that the gate structure is a GAA structure, which has better gate control capability. See [link to example]. Figure 6 The gate dielectric is wrapped around the surface of the gate dielectric. Figure 400 shows the device, where 410 is a p-type transistor, 420 is an n-type transistor, 401 is the gate dielectric, and 402 is the gate metal.

[0059] See Figure 7 300 is the internal structure of the above-mentioned CMOS inverter device structure, wherein 302 and 301 are dielectric layer 1 (first dielectric layer) and dielectric layer 2 (second dielectric layer), respectively. The structure of the first dielectric layer 302 is an L-shaped structure, and the structure of 301 is a cuboid structure. 303, 304, and 305 are the metal electrodes of the CMOS device, respectively. 306 and 307 are n-type nanowires and p-type nanowires, respectively.

[0060] The fabrication method of this device is as described in Example 1. In step (15), wet etching is used to remove part of dielectric layer 1 and dielectric layer 2, i.e., to reduce dielectric layer 1 and dielectric layer 2, so that the nanowire is suspended. Then, ALD is deposited to deposit HK dielectric and metal gate and etched back to obtain the CMOS inverter device. See [link to example]. Figure 8 .

[0061] This invention discloses a CMOS device structure based on closely packed zigzag nanowires. One NMOS is a vertical channel structure and the other is a planar device structure. The conductive channel length of the vertical channel structure depends on the thickness of the dielectric layer 1, and the conductive channel length of the planar device structure depends on the wet etching time of the dielectric layer 2. Both MOS structures can be realized without high-precision photolithography. The number of close-packed nanowires in the device is determined by the number of cycles of depositing dielectric layers 1 and 2 on the sidewalls. With n cycles of depositing dielectric layers 1 and 2, the number of channels is 4n. The morphology or diameter of the nanowires is determined by the width and thickness of dielectric layer 2. The nanowires have high uniformity and are generally columnar or nanosheet structures. The gate structure of NMOS and PMOS can be a Fin-like structure or a GAA structure. Importantly, this invention first deposits n-type amorphous silicon. Taking advantage of the fact that the etching rate of the n-type amorphous silicon film in the vertical groove is lower than that in the planar groove, the n-type amorphous silicon film in the top groove can be etched in situ using H2 plasma in PECVD. Then, intrinsic amorphous silicon is deposited and annealed in a vacuum environment to achieve the growth of pn junction nanowires, saving fabrication costs and reducing the difficulty of growth.

Claims

1. A CMOS inverter device structure based on close-packed zigzag nanowires comprising a dielectric layer, an electrode, a nanowire, and a gate dielectric and a gate, characterized in that, The nanowires have a zigzag structure.

2. The CMOS inverter device structure based on close-packed zigzag-shaped nanowires according to claim 1, wherein, The nanowires are p-type and n-type nanowires, which together form a zigzag structure.

3. The CMOS inverter device structure based on close-packed zigzag-shaped nanowires according to claim 1, wherein, The dielectric layer includes a second dielectric layer and a first dielectric layer arranged vertically; the first dielectric layer has an L-shaped structure; the thickness of the first dielectric layer and the second dielectric layer is nanometer or micrometer; the materials of the first dielectric layer and the second dielectric layer are independently selected from insulating dielectric materials such as silicon nitride, silicon oxide, silicon oxynitride, aluminum oxide, hafnium oxide, zirconium oxide, hafnium zirconium oxide, hafnium lanthanum oxide, and quartz, and the two dielectric layers are made of different materials.

4. The CMOS inverter device structure based on close-packed zigzag-shaped nanowires according to claim 1, wherein, The dielectric layer and electrodes are located within the nanowires; electrodes are respectively provided on the upper and lower parts of the dielectric layer; the thickness of the electrodes is at the nanometer or micrometer level, and their materials can be single metal materials such as titanium, platinum, gold, copper, aluminum, nickel, and shale, or mixed metal materials such as platinum, nickel-gold, nickel-aluminum, and titanium-gold, or compound conductive materials such as titanium nitride, polycrystalline silicon, and heavily doped polycrystalline silicon, or a stacked combination of the above materials.

5. The CMOS inverter device structure based on close-packed zigzag-shaped nanowires according to claim 1, wherein, The nanowire has a gate dielectric and a gate on its outer side; the gate includes a Fin structure or a GAA structure, and the thickness of the gate is at the nanometer or micrometer level. Its material can be a single metal material such as titanium, platinum, gold, copper, aluminum, nickel, or shale, or a mixed metal material such as platinum, nickel-gold, nickel-aluminum, or titanium-gold, or a compound conductive material such as titanium nitride, polycrystalline silicon, or heavily doped polycrystalline silicon, or a stacked combination of the above materials; the gate dielectric material can be an insulating dielectric material such as silicon nitride, silicon oxide, silicon oxynitride, aluminum oxide, hafnium oxide, zirconium oxide, hafnium zirconium oxide, hafnium lanthanum oxide, or hafnium aluminum oxide, or a stacked material of them.

6. The CMOS inverter device structure based on close-packed zigzag-shaped nanowires according to claim 1, wherein, The CMOS inverter device structure based on close-packed zigzag nanowires further includes a third dielectric layer, on both sides of which nanowires are arranged; there are multiple nanowires, with spacing between adjacent nanowires; the nanowires are made of silicon or a III-V group material.

7. The method for fabricating the CMOS inverter device structure based on close-packed zigzag nanowires as described in claim 1, comprising the following steps: (1) A lower first dielectric layer, a second dielectric layer and an upper first dielectric layer are sequentially disposed on the lower conductive layer, wherein the width of the second dielectric layer is smaller than that of the first dielectric layer; (2) Conductive layers are respectively disposed on both sides of the second dielectric layer between the upper first dielectric layer and the lower first dielectric layer; (3) A vertical first dielectric layer is provided on the sidewall of the lower conductive layer below the lower first dielectric layer, and the upper first dielectric layer is removed; (4) A third dielectric layer is set in the middle of the structure obtained in step (3), and then the first dielectric layer is deposited and etched back in sequence, the second dielectric layer is deposited and etched back; the structure with the third dielectric layer in the middle and the first and second dielectric layers on both sides are arranged alternately is repeated; then the second dielectric layer is etched to form a channel; (5) Deposit catalytic metal and form metal droplets on the channel wall, then grow p-type nanowires and n-type nanowires; then prepare the gate dielectric and gate to obtain the above-mentioned CMOS inverter device structure based on close-packed zigzag nanowires; or, (6) Deposit catalytic metal and form metal droplets on the channel wall, then grow p-type nanowires and n-type nanowires; then etch part of the first dielectric layer and the second dielectric layer, and then prepare the gate dielectric and the gate to obtain the above-mentioned CMOS inverter device structure based on close-packed zigzag nanowires.

8. The method of claim 7, wherein the method further comprises: In step (5), the catalytic metal is indium, tin, or their alloys; and the metal droplets are formed by plasma treatment.

9. A memory comprising the CMOS inverter device structure based on close-packed zigzag nanowires as described in claim 1.

10. An electronic device comprising the CMOS inverter device structure based on close-packed zigzag nanowires as described in claim 1.