WO2015007108A1 - 相变存储单元及其制备方法 - Google Patents

相变存储单元及其制备方法 Download PDF

Info

Publication number
WO2015007108A1
WO2015007108A1 PCT/CN2014/075276 CN2014075276W WO2015007108A1 WO 2015007108 A1 WO2015007108 A1 WO 2015007108A1 CN 2014075276 W CN2014075276 W CN 2014075276W WO 2015007108 A1 WO2015007108 A1 WO 2015007108A1
Authority
WO
WIPO (PCT)
Prior art keywords
phase change
material layer
change memory
change material
memory cell
Prior art date
Application number
PCT/CN2014/075276
Other languages
English (en)
French (fr)
Inventor
宋志棠
任堃
饶峰
宋三年
陈邦明
Original Assignee
中国科学院上海微系统与信息技术研究所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中国科学院上海微系统与信息技术研究所 filed Critical 中国科学院上海微系统与信息技术研究所
Priority to US15/739,900 priority Critical patent/US20180269388A1/en
Publication of WO2015007108A1 publication Critical patent/WO2015007108A1/zh

Links

Classifications

    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/801Constructional details of multistable switching devices
    • H10N70/821Device geometry
    • H10N70/823Device geometry adapted for essentially horizontal current flow, e.g. bridge type devices
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/20Multistable switching devices, e.g. memristors
    • H10N70/231Multistable switching devices, e.g. memristors based on solid-state phase change, e.g. between amorphous and crystalline phases, Ovshinsky effect
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/011Manufacture or treatment of multistable switching devices
    • H10N70/021Formation of switching materials, e.g. deposition of layers
    • H10N70/023Formation of switching materials, e.g. deposition of layers by chemical vapor deposition, e.g. MOCVD, ALD
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/011Manufacture or treatment of multistable switching devices
    • H10N70/061Shaping switching materials
    • H10N70/063Shaping switching materials by etching of pre-deposited switching material layers, e.g. lithography
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/801Constructional details of multistable switching devices
    • H10N70/821Device geometry
    • H10N70/826Device geometry adapted for essentially vertical current flow, e.g. sandwich or pillar type devices
    • H10N70/8265Device geometry adapted for essentially vertical current flow, e.g. sandwich or pillar type devices on sidewalls of dielectric structures, e.g. mesa-shaped or cup-shaped devices
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/801Constructional details of multistable switching devices
    • H10N70/841Electrodes
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/801Constructional details of multistable switching devices
    • H10N70/841Electrodes
    • H10N70/8413Electrodes adapted for resistive heating
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/801Constructional details of multistable switching devices
    • H10N70/881Switching materials
    • H10N70/882Compounds of sulfur, selenium or tellurium, e.g. chalcogenides
    • H10N70/8828Tellurides, e.g. GeSbTe
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10NELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10N70/00Solid-state devices having no potential barriers, and specially adapted for rectifying, amplifying, oscillating or switching
    • H10N70/801Constructional details of multistable switching devices
    • H10N70/881Switching materials
    • H10N70/884Switching materials based on at least one element of group IIIA, IVA or VA, e.g. elemental or compound semiconductors

Definitions

  • the invention belongs to the field of microelectronics, relates to a phase change memory unit and a preparation method thereof, and particularly relates to a two-dimensional phase change memory unit having high density, low power consumption and high speed and a preparation method thereof. Background technique
  • Phase Change Random Access Memory uses the operating signal to generate Joule heat to operate the phase change material to change between different phases, thereby reflecting the difference between high and low resistance values and completing the storage of information.
  • Phase change memory is considered to be the most promising next generation non-volatile due to its fast operation speed, good data retention, strong cycle operation capability, compatibility with traditional CMOS processes, and its ability to maintain its operation at small sizes.
  • One of the memories One of the memories.
  • Phase change material is the information storage medium of phase change memory. Its thermal stability, solid phase stability, crystallization speed, melting point and other characteristics directly affect the data retention of the phase change memory, cycle operation life, operation speed and operation power consumption. Therefore, choosing an excellent phase change material can directly improve the performance of the phase change memory.
  • GeSbTe material is the most widely used phase change material, and its biggest feature is balanced performance in all aspects.
  • GeSbTe is a phase change material based on nucleation crystals. It exhibits a uniform crystal phase with a data retention of 90 degrees Celsius for 10 years, a melting point of 630 degrees Celsius, and a crystallization rate of 50 nanoseconds.
  • the GeTe phase change material is a phase change material mainly composed of growth crystals, which exhibits a uniform and stable crystal phase.
  • the data retention force is 100 degrees Celsius for ten years, and the crystallization speed can reach 1 nanosecond.
  • the disadvantage is higher melting point. 730 degrees Celsius.
  • TiSbTe is a new type of phase change material, which is mainly composed of growth crystals. It exhibits a uniform and stable crystal phase.
  • the data retention can reach 110 degrees Celsius for ten years.
  • the crystallization rate can reach 6 nanoseconds and the melting point is 540 degrees Celsius.
  • Potential phase change materials are made of phase change memory devices with good electrical operation performance, and are the first choice for phase change memory manufacturing.
  • phase change materials exhibit a rhombohedral lattice structure in a crystalline state, and the distance between adjacent atoms is about 6 angstroms.
  • the literature ( nature doi: 10.1038/nmatl215 ) reports that the phase transition of GST involves only the jump of Ge atoms between different positions, and the unit where Ge atoms are located is a cube with three sides of side length and a side length of about 6 angstroms. It can be considered that the size of this cube is the smallest dimension of the phase change, and this cube is the smallest phase change unit. In order to complete the phase change, the phase change material must have a dimension of more than 6 angstroms in any dimension.
  • the effective part of the operating power consumption is the energy that is used to achieve the phase transition of the phase change material.
  • the limited structure phase change memory reduces the operating power consumption of the device by reducing the phase change region.
  • the preparation of small-sized electrodes such as a blade structure, a ring structure, and the like is also intended to reduce the phase change region, thereby reducing power consumption.
  • the device resistances of the several structures described are primarily determined by the resistance of the phase change material film.
  • Contact resistance is the resistance generated at the interface of the phase change material in contact with the metal electrode, and its magnitude is proportional to the contact resistivity and inversely proportional to the contact area.
  • the contact resistivity is determined by the material on both sides of the interface and is the essential property of the interface; The smaller the area, the greater the contact resistance.
  • the literature (APPLIED PHYSICS LETTERS 102, 213503 (2013)) reports that the contact resistivity of the amorphous phase change material GST and TiN interface is 1.58 ⁇ 10 7 ⁇ ⁇ ⁇ m 2 , which is almost the contact resistivity of the crystalline GST and TiN interface. It is 1000 times that of 1.74 X 10 4 ⁇ ⁇ ⁇ m 2 .
  • the contact resistance between the phase change material and the metal electrode is small in the overall resistance of the conventional phase change memory cell, which is much smaller than the ratio of the resistance provided by the phase change material film.
  • the literature (APPLIED PHYSICS LETTERS 102, 213503 (2013)) indicates that the small proportion of contact resistance is due to the residual crystalline phase change material layer at the interface. Since the RESET state of the conventional phase change memory only needs to form an amorphous region in the phase change material to block the crystalline low resistance path between the upper and lower electrodes, the phase change material at the interface during the RESET operation is due to the high thermal conductivity of the metal electrode. Rapid heat dissipation, the temperature is always lower than the melting point, so a layer of crystalline phase change material remains at the interface.
  • the interface between the crystalline phase change material and the metal electrode is maintained at a low resistance during the SET and RESET operations, which has a very small effect on the total resistance of the device unit.
  • the literature (APPLIED PHYSICS LETTERS 102, 213503 (2013)) also indicates that the contact resistance between the amorphous phase change material and the metal electrode is about 1000 times the contact resistance between the crystalline phase change material and the metal resistance, which can significantly affect The resistance to the device and the desired operating signal strength.
  • the main cause of failure of the phase change memory is the decrease in material uniformity due to elemental segregation of the phase change material.
  • the elemental diffusion mainly occurs under the high temperature conditions generated by the current during operation, and the longer the high temperature duration, the more severe the element segregation. Therefore, long-term high-power operation of the phase change material will promote element segregation, accelerate device failure, and reduce the number of cycles that the device can be cycled.
  • an object of the present invention is to provide a phase change memory unit and a method for fabricating the same, which are used to solve the phase change memory of the prior art phase change memory unit due to high power consumption and slow operation speed.
  • the problem of failure is to provide a phase change memory unit and a method for fabricating the same, which are used to solve the phase change memory of the prior art phase change memory unit due to high power consumption and slow operation speed.
  • the present invention provides a method of fabricating a phase change memory cell, the method of fabricating comprising at least the following steps:
  • step 3 depositing a phase change material on the surface of the structure obtained in step 2) to form a phase change material layer having a first thickness
  • phase change material layer on the lower electrode layer removing a portion of the phase change material layer on the lower electrode layer to divide the phase change material layer into two parts, thereby respectively providing a phase change material layer for a pair of phase change memory cells;
  • step 4 depositing a third dielectric material layer on the surface of the structure obtained in step 4), filling the window while isolating the phase change material layer divided into two parts in step 4);
  • the first dielectric material layer has a thickness ranging from 2 to 10 nanometers.
  • the window has an opening width ranging from 10 to 100 nanometers.
  • the first thickness ranges from 6 to 60 angstroms for the phase change memory cell to store information using a difference in interface resistance of the phase change material layer.
  • the first thickness ranges from 6 to 20 angstroms.
  • the phase change material layer has a length ranging from 50 to 100 angstroms, and the phase change material layer has a width ranging from 50 to 100 angstroms.
  • the phase change material comprises at least any one of Ge-Sb-Te, Ge-Te or Ti-Sb-Te.
  • the cell driving device for implementing the phase change memory cell read/write erase function comprises a transistor or a diode, wherein the single driving device is a transistor to form a 1T1R structure, and the cell driving device is a diode to form a 1D1R structure.
  • the present invention also provides a phase change memory unit, the phase change memory unit including at least:
  • phase change material has a first thickness
  • the layer separates the second dielectric material layer from the third dielectric material layer
  • An upper electrode layer in contact with the phase change material layer.
  • the cross-section of the phase change material layer is two opposite L-types separated by a third dielectric material layer, wherein One side of the L-shaped contact with the lower electrode layer is a first side, and the other side of the L-shape perpendicular to the first side is a second side, and the thickness of the first side and the second side are both First thickness.
  • the first thickness ranges from 6 to 60 angstroms for the phase change memory cell to store information using a difference in interface resistance of the phase change material layer.
  • the first thickness ranges from 6 to 20 angstroms.
  • the invention also provides a method for preparing a phase change memory unit, the preparation method comprising at least the following steps:
  • step 2) depositing a phase change material on the surface of the structure obtained in step 1) to form a phase change material layer having a first thickness;
  • phase change material layer etching the phase change material layer to form a phase change material layer having a width equal to or less than a width of the electrode pair;
  • the first thickness ranges from 6 to 60 angstroms for the phase change memory cell to store information using a difference in interface resistance of the phase change material layer.
  • the first thickness ranges from 6 to 20 angstroms.
  • the first distance ranges from 10 to 100 nanometers.
  • the third dielectric material layer has a thickness of 20 to 100 nanometers.
  • the material of the electrode pair is a metal, a metal alloy, a metal nitride or a graphene;
  • the metal is, for example, Ti, W, Pt, etc., the alloy of the metal alloy, such as Ti, W, Pt, etc., Metal nitrides such as TiN, TaN, and the like.
  • the present invention also provides a phase change memory unit, the phase change memory unit including at least:
  • phase change material layer formed on the surface of the electrode pair and the second dielectric material layer, having a width less than or equal to the width of the electrode pair, and having a first thickness
  • a third dielectric material layer formed on the surface of the phase change material layer and the electrode pair, or formed on the surface of the phase change material layer.
  • the first thickness ranges from 6 to 60 angstroms for the phase change memory cell to store information using a difference in interface resistance of the phase change material layer.
  • the first thickness ranges from 6 to 20 angstroms.
  • the first distance ranges from 10 to 100 nanometers.
  • the third dielectric material layer has a thickness of 20 to 100 nanometers.
  • the material of the electrode pair is a metal, a metal alloy, a metal nitride or a graphene;
  • the metal is, for example, Ti, W, Pt, etc., the alloy of the metal alloy, such as Ti, W, Pt, etc., Metal nitrides such as TiN, TaN, and the like.
  • the phase change memory cell of the present invention and the method of fabricating the same have the following beneficial effects:
  • the thickness of the phase change material layer used is equivalent to a single unit cell or a plurality of cell scales, on the one hand, the phase change region is reduced, and on the other hand, the bulk material of the phase change material layer is weakened.
  • the operation power consumption of the phase change memory cell is reduced and the operation time is shortened, thereby reducing the damage of the phase change material in each operation process.
  • the effect of element segregation on the material is reduced for each operation, and the maximum operable number of the phase change memory unit is increased, thereby facilitating the ability to increase the number of cycle operations of the device;
  • the graphene electrode pair used in the present invention has a signal response Fast, mechanical strength, and low energy loss make the phase change memory cell based on graphene electrode have the advantages of high speed, low power consumption and long life.
  • phase change material layer cell of the invention The reversible phase transition behavior of a small amount of phase change material layer cell of the invention, the similarity and similar behavior of amorphous and polycrystalline basic units, the behavior of interface defects, and the behavior of metal and phase change materials differing greatly between amorphous and polycrystalline resistance It is compatible with the new CMOS, and it has tremendous capabilities in technology nodes, high speed, and low power consumption below 10 nanometers.
  • FIG. 1A to 1G are schematic views showing the structure of a phase change memory cell of the present invention in a first embodiment
  • FIG. 1G is a schematic structural view of a pair of phase change memory cells in the first embodiment.
  • FIG. 2A to 2D are structural diagrams showing the corresponding steps of the phase change memory cell and the method for fabricating the same according to the second embodiment of the present invention, wherein FIG. 2C is a top view, and FIG. 2D is also a structure of a phase change memory cell in the second embodiment. schematic diagram. Component label description
  • the main cause of failure of the phase change memory is the decrease in material uniformity due to elemental segregation of the phase change material.
  • the elemental diffusion mainly occurs under the high temperature conditions generated by the current during operation, and the longer the high temperature duration, the more severe the element segregation. Therefore, long-term high-power operation of the phase change material will promote element segregation, accelerate device failure, and reduce the number of cycles that the device can be cycled.
  • a low-power, fast-operating phase-change memory has a reduced operating element time for each operation due to a short operating time, which facilitates the ability to increase the number of cycle operations of the device.
  • the phase change memory unit provided by the invention and the preparation method thereof reduce the size in one dimension of the three-dimensional phase change material size, so that the phase change material substantially exhibits interface characteristics, and weakens the material properties
  • the thickness of the phase change material layer used is equivalent to a single unit cell or a plurality of unit cell scales, on the one hand, the phase change region is reduced, and on the other hand, the bulk material properties of the phase change material layer are weakened.
  • the reversible phase change behavior of the phase change material layer is mainly interface characteristics, thereby preparing a high-density, low-power, high-speed two-dimensional phase change memory cell that utilizes interface resistance change to store information.
  • the phase change memory list is promoted.
  • the reduction of the power consumption of the meta-operation and the shortening of the operation time reduce the damage of the phase change material in each operation process, so that the element segregation effect of the material is reduced for each operation, and the maximum operable number of the phase change memory unit is increased, thereby
  • the graphene electrode pair used in the present invention has the characteristics of fast signal response, high mechanical strength, and low energy loss, so that the phase change memory unit based on the graphene electrode has high speed and low speed. Power consumption, long life advantages.
  • phase change material layer cell of the invention The reversible phase transition behavior of a small amount of phase change material layer cell of the invention, the similarity and similar behavior of amorphous and polycrystalline basic units, the behavior of interface defects, and the behavior of metal and phase change materials differing greatly between amorphous and polycrystalline resistance , achievable and new
  • Embodiment 1 CMOS compatible, and showing great capabilities in technology nodes, high speed, low power consumption below 10 nanometers.
  • Embodiment 1 Embodiment 1
  • the present invention provides a phase change memory cell, comprising at least: a Si substrate 1, a first dielectric material layer 21, a lower electrode layer 31, a second dielectric material layer 22, a phase change material layer 4, and a third The dielectric material layer 23 and the upper electrode layer 32.
  • the first dielectric material layer 21 is formed on a surface of the Si substrate; and the lower electrode layer 31 is formed on a surface of the first dielectric material layer 21.
  • the upper surfaces of the second dielectric material layer 22, the phase change material layer 4 and the third dielectric material layer 23 are all located in the same plane, and the second dielectric material layer 22, the phase change material layer 4 and the third dielectric material Layers 23 are all formed on the lower electrode layer 31 and are both in contact with the lower electrode layer 31.
  • the phase change material layer 4 has a first thickness D, and the phase change material layer 4 isolates the second dielectric material layer 22 from the third dielectric material layer 23, further, as shown in FIG. 1G, in this embodiment
  • the cross-section of the phase change material layer 4 is two opposite L-shapes separated by the third dielectric material layer 23 to divide the phase change material layer 4 into two parts, thereby forming a pair in a one-to-one correspondence manner.
  • the phase change memory cell provides a phase change material layer, respectively. Wherein, one side of the L-shaped contact with the lower electrode layer 31 is a first side, and the other side of the L-shape perpendicular to the first side is a second side, and the first side and the second side are The thickness is the first thickness D.
  • Fig. 1G shows a schematic structural view of a pair of phase change memory cells.
  • the first thickness D ranges from 6 to 60 angstroms for the phase change memory unit to store information by using the interface resistance difference of the phase change material layer. Further, the first thickness D ranges from 6 to 20 In the present embodiment, preferably, the first thickness D is 15 angstroms.
  • the upper electrode layer 32 is in contact with the phase change material layer 4.
  • the present invention further provides a method for fabricating the above phase change memory unit, comprising at least the following steps:
  • step 1) is performed. As shown in FIG. 1A, a Si substrate 1 having a first dielectric material layer 21 formed thereon is provided, and then a lower electrode layer 31 is deposited on the first dielectric material layer 21 in order from bottom to top. And a second dielectric material layer 22.
  • the first dielectric material layer 21 is an insulating dielectric material commonly used in a semiconductor process, and includes at least one of silicon oxide, hafnium oxide, gallium oxide, silicon nitride, tantalum nitride or gallium nitride; The thickness of the first dielectric material layer 21 ranges from 2 to 10 nanometers; the material of the lower electrode layer 31 is selected from a good conductor, and at least includes any one of Cu, TiN, W, Ta, Ti, and Pt, or Any one of the above-mentioned good conductor alloys; the second dielectric material layer 22 is an oxygen-free insulating dielectric material commonly used in semiconductor processes, and includes at least any one of gallium nitride, tantalum nitride or silicon nitride. .
  • the first dielectric material layer 21 is preferably 6 nm thick silicon oxide; the lower electrode layer 31 is preferably TiN; and the second dielectric material layer 22 is preferably silicon nitride. Then proceed to step 2).
  • step 2) as shown in FIG. 1B, a portion of the second dielectric material layer 22 is photolithographically and etched by a process such as gluing, exposing, etching, de-glue, or the like until the lower electrode layer 31 is exposed.
  • the window A is formed, wherein the opening width W1 of the window A ranges from 10 to 100 nm. In the present embodiment, it is preferable that the opening width W1 of the window A is 60 nm. Then proceed to step 3).
  • a phase change material is deposited on the surface of the structure obtained in step 3 by using a chemical vapor deposition (CVD) or an atomic layer deposition (ALD) method.
  • a phase change material layer 4 having a first thickness D.
  • the phase change material comprises at least any one of Ge-Sb-Te, Ge-Te or Ti-Sb-Te, and the phase change memory cells made of the above three materials all have good electrical operation performance;
  • the first thickness D ranges from 6 to 60 angstroms for the phase change memory unit to store information by using the interface resistance difference of the phase change material layer. Further, the first thickness D ranges from 6 to 20 angstroms.
  • the phase change material layer 4 has a length ranging from 50 to 100 angstroms, and the phase change material layer 4 has a width ranging from 50 to 100 angstroms. .
  • the phase change material is preferably Ti-Sb-Te; preferably, the first thickness D is 15 angstroms; preferably, the length of the phase change material layer 4 is 80 angstroms, and the width of the phase change material layer 4 It is 80 angstroms, but is not limited to the case where the length and width of the phase change material layer 4 are equal. In other embodiments, the length and width of the phase change material layer may also be unequal.
  • the conventional phase change material layer has a three-dimensional material size of length, width and thickness.
  • the phase change material layer 4 reduces the size in one dimension of the three-dimensional material size, in this embodiment.
  • the dimension of the phase change material layer 4 corresponds to a thickness, and specifically, the phase change material layer 4 is precisely controlled by adjusting the deposition time of the phase change material.
  • the thickness of the phase change material layer 4 is reduced to a first thickness corresponding to a single unit cell or a plurality of unit cell scales, so that in the case of a very thin thickness, the thickness of the phase change material layer 4 can be ignored.
  • the variable material layer is considered to be a two-dimensional phase change material layer.
  • the small first thickness suppresses the formation of large crystal grains, thereby reducing the phase change region on the one hand, significantly reducing the power consumption of the RESET operation of the phase change memory cell, and on the other hand, weakening the bulk material properties of the phase change material layer, And to ensure that the reversible phase change behavior of the phase change material layer is mainly interface characteristics, at this time, the interface characteristics are the key to storing information in the phase change memory unit. Further, since the phase change material has a small size, it is advantageous for the phase change memory cell to be downsized, so that the phase change memory cell of the present invention has the potential of ultra-high density information storage.
  • the thickness of the phase change material layer 4 (i.e., the first thickness) must be not less than 6 angstroms.
  • the thickness of the phase change material layer 4 (i.e., the first thickness) should be less than the size of ten minimum cells, i.e., 60 angstroms. Therefore, the thickness (i.e., the first thickness) of the phase change material layer 4 in the present invention is controlled to be between 6 and 60 angstroms.
  • the effect of the size effect on the thermal stability of the phase change material is as follows: When the thickness of the phase change material layer is above 10 nm, the crystallization temperature of the phase change material changes very little with thickness. When the thickness of the phase change material layer is less than 10 nm, the crystal temperature of the phase change material increases with the decrease of the thickness.
  • the thickness (i.e., the first thickness) of the phase change material layer 4 of the present embodiment ranges from 6 to 60 angstroms, more preferably from 6 to 20 angstroms, so that the crystallization temperature of the phase change material layer 4 follows the first thickness. The reduction has been improved to varying degrees.
  • the effect of the size effect on the crystallization rate of the phase change material is as follows:
  • the thickness of the phase change material is thinned, the specific surface area of the material increases, and the interface of the phase change material is likely to form a unit cell due to defects.
  • the presence of the unit cell is that the crystallization process of the phase change material shortens the cell formation time and reduces the time required for the crystallization process, thereby increasing the operating speed of the phase change memory.
  • the cell formation time is shortened, grain growth becomes a major factor affecting the crystallization time, and the grain growth time becomes shorter as the size is reduced, which ensures a faster phase transition speed of the small-sized device. Then proceed to step 4).
  • step 4 a portion of the lower electrode layer is removed by a photolithography and etching process including at least a process of gluing, exposing, etching, and stripping, or by using a focused ion beam FIB.
  • the phase change material layer 4 on 31 is exposed until the lower electrode layer 31 located thereunder, that is, the phase change material layer 4 partially located in the window A and in contact with the lower electrode layer 31 is removed and exposed partially under
  • the electrode layer 31 divides the phase change material layer 4 into two parts, thereby respectively providing a phase change material layer for a pair of phase change memory cells in a one-to-one correspondence manner.
  • the symmetry axis of the phase change material layer of the removed portion is the center line of the window A to halve the phase change material layer 4. Then perform step 5).
  • step 5 as shown in FIG. 1E, using low temperature chemical vapor deposition or low temperature atomic layer deposition method, in step Step 4)
  • the obtained structural surface deposits a third dielectric material layer 23, which is filled with the window A while isolating the phase change material layer 4 divided into two parts in the step 4), wherein the third dielectric material layer 23 is
  • the oxygen-free insulating dielectric material commonly used in the semiconductor process includes at least any one of gallium nitride, tantalum nitride or silicon nitride.
  • the third dielectric material layer 23 is preferably tantalum nitride. .
  • the reason why the second dielectric material layer 22 and the third dielectric material layer 23 do not contain oxygen is that the material contacting the phase change material is not suitable for the oxygen-containing material, and because of the second dielectric material layer 22 and The third dielectric material layer 23 is in contact with the phase change material layer 4, and therefore, the second dielectric material layer 22 and the third dielectric material layer 23 are oxygen-free insulating dielectric materials commonly used in semiconductor processes.
  • first dielectric material layer 21, the second dielectric material layer 22, and the third dielectric material layer 23 can select a uniform dielectric material, that is, the three can be the same. They can be different, or they can be the same.
  • the third dielectric material layer 23 when depositing the third dielectric material layer 23 in the step 5), it is preferable to suppress the formation of defects at the interface between the third dielectric material layer 23 and the phase change material layer 4 to ensure the interface. Smoothing, providing favorable conditions for lattice matching of the phase change material layer 4 and the third dielectric material layer 23, that is, the phase change material layer 4 and the third dielectric material layer 23 form a good interface, that is, a small amount of the interface exists. defect.
  • phase change material layer has a good interface, the stability of a small number of atoms in the phase change material layer in the reversible phase transition process is ensured, thereby ensuring the composition of the phase change material in the reversible phase transition process.
  • step 6 the structure obtained in step 5) is planarized by a chemical mechanical polishing (CMP) process until the first dielectric material layer 21 and a portion of the phase change material layer 4 are exposed to The cross-section of the phase change material layer 4 is made of two opposing L-shapes that are not in contact with each other. Then proceed to step 7).
  • CMP chemical mechanical polishing
  • an upper electrode layer 32 overlying the exposed phase change material layer 4 is formed to complete the preparation of two phase change memory cells.
  • the upper electrode layer 32 is photolithographically and etched by physical vapor deposition (PVD), low temperature chemical vapor deposition or low temperature atomic layer deposition, and the upper electrode layer 32 is formed to cover the upper electrode layer 32.
  • PVD physical vapor deposition
  • low temperature chemical vapor deposition or low temperature atomic layer deposition low temperature chemical vapor deposition or low temperature atomic layer deposition
  • the unit driving device for realizing the phase change memory cell read/write erase function includes a transistor or a diode, so that the phase change memory cell preparation process can be fully compatible with the CMOS process, wherein the single driving device is a transistor A 1T1R structure is formed, and when the unit driving device is a diode, a 1D1R structure is formed. This is well known to those skilled in the art and will not be repeated here. It should be noted that the reason for suppressing defects at the interface of the phase change material layer 4 and forming the good interface between the phase change material layer 4 and the third dielectric material layer 23 is as follows:
  • phase change material has defects in the unit cell, the atomic structure is broken or the bond energy is small, and it is easy to ionize under the action of external force.
  • the specific surface area is large, so many atoms are ionized, and the ions deviate from the original position. It causes rapid changes in structure; at the same time, the energy required for structural changes is also reduced by the presence of defects, thereby reducing the power consumption of phase change material amorphization.
  • the presence of excessive defects will make nucleation easier, and the crystallization of the phase change material will be carried out at a lower temperature, and the thermal stability of the amorphous phase change material will decrease, which will result in data retention of the phase change memory cell.
  • the size of the phase change material is as small as one or several unit cell scales, the presence of too many interfacial defects will affect the formation of the normal crystalline atomic structure in the phase change material, resulting in complete atomic structure dysfunction, thereby making the phase change material The ability to lose reversible phase transitions.
  • the interface defect density of the phase change memory cell of the present invention should be controlled to a small extent.
  • the total resistance of the phase change memory cell of the present invention is mainly caused by the thin thickness of the phase change material layer.
  • the contact resistance at the interface ie, the interface resistance
  • the phase change memory cell controls the crystalline and amorphous states of the phase change material to make a huge difference in the contact resistance at the interface, so that the phase change memory cell exhibits two distinct high and low resistances, thereby realizing the storage of data. .
  • the phase change material in the phase change memory cell is operated into an amorphous state by using a sub-nanosecond electric pulse signal, and the interface between the electrode and the phase change material is in a high resistance state, which is called a RESET state;
  • the pulse signal operates the phase change material in the phase change memory cell into a crystalline state, and the interface between the electrode and the phase change material is in a low resistance state, which is called a SET state.
  • the contact resistance at the interface between the amorphous phase change material and the electrode material is the first contact resistance
  • the contact resistance at the interface between the crystalline phase change material and the electrode material is the second contact resistance
  • the first contact resistance ranges from 10 3 to 10 5 , wherein the electrode material includes an upper electrode layer and a lower electrode layer.
  • the ratio of the total resistance of the RESET state phase change memory cell to the total resistance of the SET state phase change memory cell ranges from 10 to 10 5 times.
  • the thickness of the phase change material is equivalent to a single unit cell or a plurality of unit cell scales, the amorphous to crystalline state transition When only a small number of unit cells are formed, and the formation of large crystal grains is suppressed, the distance required for the atoms to be arranged from the disordered state to the ordered state is short, and it takes less time to make the phase change memory cell of the present invention.
  • the operation has the advantage of high speed.
  • the thickness of the phase change material layer is thinned to reduce the phase change region, thereby reducing the operating power consumption, and the thickness of the phase change material layer is thin and a small number of defects are present at the interface of the phase change material layer, thereby improving the operation speed and thus the operation.
  • the shortening of time reduces the damage of the phase change material in each operation process, so that the segregation effect of the element on the material is reduced by each operation, and the maximum operable number of the phase change memory unit is increased, thereby facilitating the number of cycle operations of the device. ability.
  • the reversible phase transition behavior of a small number of phase change material layer cells of the present invention the similarity and similar behavior of amorphous and polycrystalline basic units, the behavior of interface defects, and the difference between amorphous and polycrystalline resistors of metal and phase change materials.
  • the large behavior can be compatible with the new CMOS, and it shows great power in the technology node below 10 nanometers, high speed and low power consumption.
  • the present invention provides a phase change memory cell comprising at least: a Si substrate 1, a second dielectric material layer 22, an electrode pair 5, a phase change material layer 4, and a third dielectric material layer 23.
  • 2D is a schematic structural diagram of a phase change memory unit in the embodiment.
  • the second dielectric material layer 22 is formed on the surface of the Si substrate 1.
  • the second dielectric material layer 22 is an oxygen-free semiconductor dielectric material, and includes at least one of gallium nitride, tantalum nitride or silicon nitride. In this embodiment, it is silicon nitride.
  • the electrode pair 5 is formed on the surface of the second dielectric material layer 22, and the two electrodes of the electrode pair 5 have a first distance and are isolated from each other.
  • the first distance ranges from 10 to 100 nanometers. In the embodiment, the first distance is preferably 60 nanometers.
  • the material of the electrode pair 5 includes metal or graphene, and the metal includes at least Cu. Any one of TiN, W, Ta, Ti, and Pt, or any of the above metal alloys. In the present embodiment, it is preferable that the material of the electrode pair 5 is graphene.
  • the phase change material layer 4 has a first thickness D formed on the surface of the electrode pair 5 and the second dielectric material layer 22, and the width W4 of the phase change material layer 4 is less than or equal to the width W5 of the electrode pair 5, wherein The width W4 of the phase change material layer 4 and the width W5 of the electrode pair 5 are as shown in FIG. 2C. In the embodiment, it is preferable that the width of the phase change material layer 4 is less than or equal to the width of the electrode pair 5;
  • the first thickness D ranges from 6 to 60 angstroms for the phase change memory unit to store information using the interface resistance difference of the phase change material layer. Further, the first thickness D ranges from 6 to 20 angstroms. In the embodiment, the first thickness D is preferably 15 angstroms.
  • the third dielectric material layer 23 is formed on the surface of the phase change material layer 4 and the electrode pair 5, or when the width of the phase change material layer 4 is W4 is equal to the width of the electrode pair 5 At W5, the third dielectric material layer 23 is formed on the surface of the phase change material layer 4.
  • the third dielectric material layer 23 is formed on the surface of the phase change material layer 4 and the electrode pair 5.
  • the third dielectric material layer 23 has a thickness of 20 to 100 nanometers, that is, a range between the surface of the third dielectric material layer 23 and the surface of the phase change material layer 4 is 20 to 100 nanometers.
  • the third dielectric material layer 23 has a thickness of 60 nanometers; the third dielectric material layer 23 is an oxygen-free semiconductor dielectric material, including at least gallium nitride, tantalum nitride or silicon nitride. In any of the embodiments, in the embodiment, the third dielectric material layer 23 is preferably tantalum nitride.
  • the present invention further provides a method for fabricating the above phase change memory unit, comprising at least the following steps:
  • step 1) is performed to provide a Si substrate 1 having a second dielectric material layer 22 formed thereon, and an electrode pair 5 is formed on the second dielectric material layer 22, wherein the electrode pair 5 is The distance is a first distance, and the first distance ranges from 10 to 100 nanometers;
  • the material of the electrode pair 5 includes metal or graphene, wherein the metal includes at least Cu, TiN, W, Ta, Ti, and Pt Any one of the above, or any of the above metal alloys;
  • the second dielectric material layer 22 is an oxygen-free insulating dielectric material commonly used in semiconductor processes, including at least gallium nitride, tantalum nitride or silicon nitride Any of them.
  • the first distance is preferably 60 nm; the material of the electrode pair 5 is preferably graphene, but is not limited thereto. In another embodiment, the material of the electrode pair 5 It may also preferably be TiN; the second dielectric material layer 22 is preferably silicon nitride.
  • graphene is a new material of a single-layer sheet structure composed of carbon atoms, which has extremely low electrical resistivity, extremely fast electron migration, very stable structure of an olefin, and extraordinary electrical conductivity. More than ten times the strength of steel and excellent light transmission. At normal temperature, the electron mobility of graphene exceeds 15000 cm 2 /Vs, and electrons can migrate extremely efficiently.
  • the current computer chip wastes 72%-81% of electricity in this way.
  • Graphene is different, its electron energy is not lost, which makes it extraordinary. Excellent characteristics.
  • the graphene used in the present invention is used as an electrode Yes, it has the characteristics of fast signal response, high mechanical strength and low energy loss. Then proceed to step 2).
  • step 2) as shown in FIG. 2B, a phase change material is deposited on the surface of the structure obtained in step 1 by chemical vapor deposition (CVD) or atomic layer deposition (ALD) to form a phase change material layer 4 having a first thickness D.
  • the first thickness D ranges from 6 to 60 angstroms for the phase change memory unit to store information by using the interface resistance difference of the phase change material layer. Further, the first thickness D ranges from 6 to 20 In the present embodiment, preferably, the first thickness D is 15 angstroms.
  • step 2) of the present embodiment regarding the material type, the size, and the beneficial effects of the phase change material layer 4 is substantially the same as that of the first embodiment, so the description of the first embodiment is the same. It will not be described again, except that the phase change material layer 4 in this embodiment is different from the first embodiment in length and width dimensions. Continue with step 3).
  • step 3 the phase change material layer 4 is photolithographically and etched by a process such as gluing, exposing, etching, degumming, etc., to form a phase having a width less than or equal to the width of the electrode pair 5.
  • a third dielectric material layer 23 is deposited on the surface of the structure obtained in the step 3) by low temperature chemical vapor deposition or low temperature atomic layer deposition, and fills the region between the electrode pairs 5.
  • the third dielectric material layer 23 has a thickness of 20 nm to 100 nm. In the embodiment, the third dielectric material layer 23 has a thickness of 60 nm.
  • the third dielectric material layer 23 is not included.
  • the oxygen semiconductor dielectric material includes at least any one of gallium nitride, tantalum nitride or silicon nitride. In the present embodiment, the third dielectric material layer 23 is preferably tantalum nitride.
  • the reason why the second dielectric material layer 22 and the third dielectric material layer 23 are oxygen-free dielectric materials is as described in the first embodiment; further, the present invention does not limit the second medium.
  • the material selected by the material layer 22 and the third dielectric material layer 23 may be identical.
  • the third dielectric material layer 23 is deposited in step 4) of the embodiment, it is preferable to suppress the formation of defects at the interface between the third dielectric material layer 23 and the phase change material layer 4 to ensure the flatness of the interface.
  • the lattice matching of the phase change material layer 4 and the third dielectric material layer 23 provides an advantageous condition, that is, the phase change material layer 4 forms a good interface with the third dielectric material layer 23, that is, the interface has a small amount of defects.
  • an operation signal is applied to the electrode pair 5 through a wire to realize operation of the phase change memory cell;
  • a cell driving device for implementing the phase change memory cell read/write erase function includes a transistor or a diode to make a phase change
  • the preparation process of the memory cell can be fully compatible with the CMOS process, wherein the single-drive device forms a 1T1R structure when the transistor is a transistor, and the 1D1R structure is formed when the cell drive device is a diode. This is well known to those skilled in the art. I will not repeat them here. This is well known to those skilled in the art and will not be repeated here.
  • the operation power consumption of the phase change memory cell is reduced and the operation time is shortened, thereby reducing the damage of the phase change material in each operation process.
  • the effect of element segregation on the material is reduced for each operation, and the maximum operable number of the phase change memory unit is increased, thereby facilitating the ability to increase the number of cycle operations of the device;
  • the graphene electrode pair used in the present invention has a signal response Fast, mechanical strength, and low energy loss make the phase change memory cell based on graphene electrode have the advantages of high speed, low power consumption and long life.
  • phase change material layer cell of the invention The reversible phase transition behavior of a small amount of phase change material layer cell of the invention, the similarity and similar behavior of amorphous and polycrystalline basic units, the behavior of interface defects, and the behavior of metal and phase change materials differing greatly between amorphous and polycrystalline resistance It is compatible with the new CMOS, and it has tremendous capabilities in technology nodes, high speed, and low power consumption below 10 nanometers. Therefore, the present invention effectively overcomes various shortcomings in the prior art and has high industrial utilization value.

Landscapes

  • Engineering & Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Semiconductor Memories (AREA)

Abstract

提供了一种相变存储单元及其制备方法。采用厚度与单个晶胞尺度相当的相变材料层,使相变材料基本上体现出界面特性,而弱化体材料特性,以制备利用界面电阻变化来存储信息的高密度、低功耗、高速二维相变存储单元。相变材料层的厚度薄及相变材料层上存在少量的缺陷,促使相变存储单元操作功耗降低和操作时间缩短,减少了每次操作过程对相变材料的损害,使每次操作对材料的元素偏析效果降低,增加了相变存储单元的最大可操作次数,从而有利于提高器件循环操作次数的能力。

Description

相变存储单元及其制备方法 技术领域
本发明属于微电子技术领域, 涉及相变存储单元及其制备方法, 特别是涉及具有高密 度、 低功耗、 高速的二维相变存储单元及其制备方法。 背景技术
相变存储器 (Phase Change Random Access Memory, PCRAM) 利用操作信号产生焦耳 热对相变材料进行操作, 使其在不同的相之间进行转变, 从而体现出高低电阻值差异, 完成 对信息的存储。 相变存储器由于其操作速度快, 数据保持力好, 循环操作能力强, 与传统 CMOS工艺兼容, 并且在小尺寸时仍能保持其操作性能, 所以被认为是最有希望的下一代非 挥发性存储器之一。
相变材料是相变存储器的信息存储介质, 其热稳定性, 固相稳定性, 结晶速度, 熔点等 特性直接影响到相变存储器的数据保持力, 循环操作寿命, 操作速度和操作功耗。 所以选择 优秀的相变材料能最直接地提升相变存储器的性能。 GeSbTe 材料是被最广泛应用的相变材 料, 其最大特点是各方面性能均衡。 GeSbTe 是一种以成核结晶为主的相变材料, 体现出均 一稳定的晶相, 数据保持力十年为 90摄氏度, 熔点为 630摄氏度, 结晶速度为 50纳秒左 右。 而 GeTe相变材料是一种以成生长结晶为主的相变材料, 体现出均一稳定的晶相, 数据 保持力十年为 100摄氏度, 结晶速度可达到 1纳秒, 缺点为较高的熔点 730摄氏度。 TiSbTe 是一种新型相变材料, 以成生长结晶为主, 体现出均一稳定的晶相, 数据保持力十年可达 110摄氏度, 结晶速度可达到 6纳秒, 熔点 540摄氏度, 是一种很具潜力的相变材料。 以上 三种材料做成相变存储器件均具有很好的电学操作表现, 是相变存储器制造首选的几种材 料。
以上三种相变材料晶态时均呈现菱形的晶格结构, 相邻原子之间的距离约为 6埃。 文献 ( nature doi: 10.1038/nmatl215 ) 报道称 GST 的相变只涉及 Ge 原子在不同位置之间的跳 跃, 而 Ge原子所处的这个单元是三个原子为边长的立方体, 边长约 6埃, 可以认为这个立 方体的尺寸是相变进行的最小尺寸, 这个立方体是最小的相变单元。 相变材料为了完成相 变, 其任何一个维度的尺寸必须在 6埃以上。
操作功耗的有效部分为实现相变材料相转变部分的能量。 相变区域越小, 所需能量越 小, 器件功耗降低。 而限制型结构相变存储器正是通过减小相变区域降低了器件操作功耗。 刀片结构, 环形结构等小尺寸电极的制备其目的也是减小相变区域, 从而降低功耗。 而以上 述的几种结构的器件电阻都是主要由相变材料薄膜的电阻决定的。
接触电阻是相变材料与金属电极接触的界面处产生的电阻, 其大小和接触电阻率成正比 和接触面积成反比, 其中, 接触电阻率由界面两边的材料决定, 是界面的本质属性; 接触面 积越小, 接触电阻越大。 文献 (APPLIED PHYSICS LETTERS 102, 213503 (2013 ) ) 报 道非晶态相变材料 GST与 TiN界面的接触电阻率为 1.58 Χ 107 Ω · μ m 2, 几乎是晶态 GST 与 TiN界面的接触电阻率为 1.74 X 104 Ω · μ m 2的 1000倍。
相变材料与金属电极之间的接触电阻在传统的相变存储单元的整体电阻中所占比例很 小, 远小于相变材料薄膜提供的电阻所占的比例。 文献 (APPLIED PHYSICS LETTERS 102, 213503 ( 2013 ) ) 指出接触电阻所占比例小的原因是由于界面处残留了晶态相变材 料层。 由于传统相变存储器的 RESET 状态只需要在相变材料中形成一个非晶区域隔断上下 电极之间的晶态低电阻通路, 在 RESET 操作过程中界面处的相变材料由于受到高热导的金 属电极快速散热, 温度一直低于熔点, 因此在界面残留了一层晶态相变材料层。 晶态相变材 料与金属电极界面在 SET与 RESET操作过程中一直维持在低电阻状态, 对器件单元的总电 阻影响非常小。 同时文献 (APPLIED PHYSICS LETTERS 102, 213503 (2013 ) ) 也指出 非晶态相变材料与金属电极之间的接触电阻是晶态相变材料与金属电阻之间的接触电阻约 1000倍, 能明显影响到器件的电阻值及所需的操作信号强度。
相变存储器的主要失效原因是由于相变材料的元素偏析导致的材料均匀性降低。 而元素 扩散主要发生在操作时电流产生的高温条件下, 高温持续时间越长元素偏析越严重。 所以对 相变材料长时间高功率操作会促使元素偏析, 加快器件失效, 降低器件可循环操作次数。
相反地, 由于低功耗快速操作的相变存储器在操作时由于操作时间短, 每次操作对材料 的元素偏析效果降低, 有利于提高器件循环操作次数的能力, 因此, 如何制备具有低功耗快 速操作特性的相变存储器是亟需解决的技术问题。 发明内容
鉴于以上所述现有技术的缺点, 本发明的目的在于提供相变存储单元及其制备方法, 用 于解决现有技术的相变存储单元由于功耗高和操作速度慢等引发的相变存储器失效的问题。
为实现上述目的及其他相关目的, 本发明提供一种相变存储单元的制备方法, 所述制备 方法至少包括以下步骤:
1 ) 提供一表面形成有第一介质材料层的 Si衬底, 自下而上依次在所述第一介质材料层 上形成下电极层和第二介质材料层; 2) 光刻、 刻蚀部分所述第二介质材料层直至暴露所述下电极层以形成窗口;
3 ) 在步骤 2) 获得的结构表面沉积相变材料以形成具有第一厚度的相变材料层;
4) 去除部分位于所述下电极层上的相变材料层, 以将相变材料层分割为两部分, 从而 为一对相变存储单元分别提供相变材料层;
5 ) 在步骤 4) 获得的结构表面沉积第三介质材料层, 填充满所述窗口的同时隔离步骤 4) 中分割为两部分的相变材料层;
6) 利用化学机械抛光工艺平坦化处理步骤 5 ) 获得的结构, 直至暴露所述第一介质材 料层及部分相变材料层, 以使所述相变材料层的横截面为两个相对的 L型;
7) 形成覆盖于所述被暴露的相变材料层上的上电极层。
可选地, 所述第一介质材料层的厚度范围为 2~10纳米。
可选地, 所述窗口的开口宽度范围为 10~100纳米。
可选地, 所述第一厚度的范围为 6~60 埃之间, 以供相变存储单元利用相变材料层的界 面电阻差异存储信息。
可选地, 所述第一厚度的范围为 6~20埃之间。
可选地, 所述相变材料层的长度范围为 50~100 埃之间, 所述相变材料层的宽度范围为 50~100埃之间。
可选地, 所述相变材料至少包括 Ge-Sb-Te、 Ge-Te或 Ti-Sb-Te中的任意一种。
可选地, 用于实现相变存储单元读写擦功能的单元驱动器件包括晶体管或二极管, 其 中, 所述单驱动器件为晶体管时形成 1T1R 结构, 所述单元驱动器件为二极管时形成 1D1R 结构。 本发明还提供一种相变存储单元, 所述相变存储单元至少包括:
Si衬底;
形成于所述 Si衬底表面的第一介质材料层;
形成于所述第一介质材料层表面的下电极层;
上表面均位于同一平面、 均形成于所述下电极层上且均与其接触的第二介质材料层、 相 变材料层和第三介质材料层, 其中, 具有第一厚度的所述相变材料层将第二介质材料层和第 三介质材料层隔离;
与所述相变材料层相接触的上电极层。
可选地, 所述相变材料层的横截面为被第三介质材料层隔离的两个相对的 L型, 其中, 与所述下电极层相接触的为 L型的一边为第一边, 与所述第一边相垂直的 L型另一边为第 二边, 所述第一边和第二边的厚度均为第一厚度。
可选地, 所述第一厚度的范围为 6~60 埃之间, 以供相变存储单元利用相变材料层的界 面电阻差异存储信息。
可选地, 所述第一厚度的范围为 6~20埃之间。 本发明还提供一种相变存储单元的制备方法, 所述制备方法至少包括以下步骤:
1 ) 提供一表面形成有第二介质材料层的 Si衬底, 并在所述第二介质材料层上制备一电 极对, 其中, 所述电极对之间的间距为第一距离;
2) 在步骤 1 ) 获得的结构表面沉积相变材料以形成具有第一厚度的相变材料层;
3 ) 光刻、 刻蚀所述相变材料层, 形成宽度小于等于电极对宽度的相变材料层;
4) 在步骤 3 ) 获得的结构表面沉积第三介质材料层, 并填充满所述电极对之间区域。 可选地, 所述第一厚度的范围为 6~60 埃之间, 以供相变存储单元利用相变材料层的界 面电阻差异存储信息。
可选地, 所述第一厚度的范围为 6~20埃之间。
可选地, 所述第一距离的范围为 10~100纳米。
可选地, 所述第三介质材料层的厚度为 20~100纳米。
可选地, 所述电极对的材料为金属、 金属合金、 金属氮化物或者石墨烯; 所述金属例如 Ti, W, Pt等, 所述金属合金例如 Ti, W, Pt等组成的合金, 所述金属氮化物例如 TiN, TaN等。 本发明还提供一种相变存储单元, 所述相变存储单元至少包括:
Si衬底;
形成于所述 Si衬底表面的第二介质材料层;
形成于所述第二介质材料层表面、 且之间具有第一距离的电极对;
形成于所述电极对及第二介质材料层的表面、 宽度小于等于电极对宽度、 且具有第一厚 度的相变材料层;
形成于相变材料层及电极对的表面上、 或形成于相变材料层表面上的第三介质材料层。 可选地, 所述第一厚度的范围为 6~60 埃之间, 以供相变存储单元利用相变材料层的界 面电阻差异存储信息。
可选地, 所述第一厚度的范围为 6~20埃之间。 可选地, 所述第一距离的范围为 10~100纳米。
可选地, 所述第三介质材料层的厚度为 20~100纳米。
可选地, 所述电极对的材料为金属、 金属合金、 金属氮化物或者石墨烯; 所述金属例如 Ti, W, Pt等, 所述金属合金例如 Ti, W, Pt等组成的合金, 所述金属氮化物例如 TiN, TaN等。 如上所述, 本发明的相变存储单元及其制备方法, 具有以下有益效果:
本发明的相变存储单元中, 采用的相变材料层的厚度与单个晶胞或者多个晶胞尺度相 当, 一方面使相变区域减小, 另一方面, 弱化相变材料层的体材料特性, 且保证相变材料层 的可逆相变行为以界面特性为主, 从而制备出利用界面电阻变化来存储信息的高密度、 低功 耗、 高速二维相变存储单元。 本发明中, 由于相变材料层薄及相变材料层界面上存在少量的 缺陷, 促使相变存储单元操作功耗的降低和操作时间的缩短, 减少了每次操作过程对相变材 料的损害, 使每次操作对材料的元素偏析效果降低, 增加了相变存储单元的最大可操作次 数, 从而有利于提高器件循环操作次数的能力; 进一步, 本发明中采用的石墨烯电极对具有 信号响应快、 机械强度大、 能量损耗少等特点, 使基于石墨烯电极的相变存储单元具有高 速, 低功耗, 长寿命的优点。
本发明的少量相变材料层晶胞的可逆相变行为、 非晶与多晶基本单元的相近与相似行 为、 界面缺陷行为、 金属与相变材料在非晶与多晶电阻差别较大的行为, 可实现与新型 CMOS的兼容, 并在 10纳米以下的技术节点、 高速、 低功耗方面显现出巨大的能力。 附图说明
图 1A至 1G显示为本发明的相变存储单元及其制备方法在实施例一中相应步骤的结构 示意图, 其中, 图 1G也为实施例一中一对相变存储单元的结构示意图。
图 2A至 2D显示为本发明的相变存储单元及其制备方法在实施例二中相应步骤的结构 示意图, 其中, 图 2C为俯视图, 图 2D也为实施例二中一个相变存储单元的结构示意图。 元件标号说明
1 Si衬底
21 第一介质材料层
22 第二介质材料层
23 第三介质材料层 31 下电极层
32 上电极层
4 相变材料层
5 电极对
A 窗口
W1 窗口的开口宽度
D 第一厚度
W4 相变材料层的宽度
W5 电极对的宽度 具体实施方式
以下通过特定的具体实例说明本发明的实施方式, 本领域技术人员可由本说明书所揭露 的内容轻易地了解本发明的其他优点与功效。 本发明还可以通过另外不同的具体实施方式加 以实施或应用, 本说明书中的各项细节也可以基于不同观点与应用, 在没有背离本发明的精 神下进行各种修饰或改变。
请参阅图 1A 至图 2D。 需要说明的是, 以下具体实施例中所提供的图示仅以示意方式 说明本发明的基本构想, 遂图式中仅显示与本发明中有关的组件而非按照实际实施时的组件 数目、 形状及尺寸绘制, 其实际实施时各组件的型态、 数量及比例可为一种随意的改变, 且 其组件布局型态也可能更为复杂。
相变存储器的主要失效原因是由于相变材料的元素偏析导致的材料均匀性降低。 而元素 扩散主要发生在操作时电流产生的高温条件下, 高温持续时间越长元素偏析越严重。 所以对 相变材料长时间高功率操作会促使元素偏析, 加快器件失效, 降低器件可循环操作次数。 相 反地, 低功耗快速操作的相变存储器在操作时由于操作时间短, 则每次操作对材料的元素偏 析效果降低, 有利于提高器件循环操作次数的能力。
有鉴于此, 本发明提供的相变存储单元及其制备方法, 将三维相变材料尺寸中的一个维 度上的尺寸缩小, 使相变材料基本上体现出界面特性, 而弱化体材料特性, 本发明的相变存 储单元中, 采用的相变材料层的厚度与单个晶胞或者多个晶胞尺度相当, 一方面使相变区域 减小, 另一方面, 弱化相变材料层的体材料特性, 且保证相变材料层的可逆相变行为以界面 特性为主, 从而制备出利用界面电阻变化来存储信息的高密度、 低功耗、 高速二维相变存储 单元。 本发明中, 由于相变材料层薄及相变材料层界面上存在少量的缺陷, 促使相变存储单 元操作功耗的降低和操作时间的缩短, 减少了每次操作过程对相变材料的损害, 使每次操作 对材料的元素偏析效果降低, 增加了相变存储单元的最大可操作次数, 从而有利于提高器件 循环操作次数的能力; 进一步, 本发明中采用的石墨烯电极对具有信号响应快、 机械强度 大、 能量损耗少等特点, 使基于石墨烯电极的相变存储单元具有高速, 低功耗, 长寿命的优 点。 本发明的少量相变材料层晶胞的可逆相变行为、 非晶与多晶基本单元的相近与相似行 为、 界面缺陷行为、 金属与相变材料在非晶与多晶电阻差别较大的行为, 可实现与新型
CMOS 的兼容, 并在 10纳米以下的技术节点、 高速、 低功耗方面显现出巨大的能力。 以下 将详细阐述本发明的相变存储单元及其制备方法的实施方式, 使本领域技术人员不需要创造 性劳动即可理解本发明的相变存储单元及其制备方法。 实施例一
请参阅图 1G, 本发明提供一种相变存储单元, 至少包括: Si衬底 1、 第一介质材料层 21、 下电极层 31、 第二介质材料层 22、 相变材料层 4、 第三介质材料层 23及上电极层 32。
所述第一介质材料层 21形成于所述 Si衬底表面; 所述下电极层 31形成于所述第一介 质材料层 21的表面。
所述第二介质材料层 22、 相变材料层 4 及第三介质材料层 23 的上表面均位于同一平 面, 同时, 所述第二介质材料层 22、 相变材料层 4及第三介质材料层 23均形成于所述下电 极层 31上且均与下电极层 31接触。
所述相变材料层 4具有第一厚度 D, 且所述相变材料层 4将第二介质材料层 22和第三 介质材料层 23 隔离, 进一步, 如图 1G所示, 在本实施例中, 所述相变材料层 4的横截面 为被第三介质材料层 23隔离的两个相对的 L型, 以将相变材料层 4分割为两部分, 从而以 一一对应的方式为一对相变存储单元分别提供相变材料层。 其中, 与所述下电极层 31 相接 触的为 L型的一边为第一边, 与所述第一边相垂直的 L型另一边为第二边, 所述第一边和 第二边的厚度均为第一厚度 D。 从而, 图 1G显示为一对相变存储单元的结构示意图。
其中, 所述第一厚度 D的范围为 6~60埃之间, 以供相变存储单元利用相变材料层的界 面电阻差异存储信息, 进一步, 所述第一厚度 D的范围为 6~20埃之间, 本实施例中, 优选 所述第一厚度 D为 15埃。
所述上电极层 32与所述相变材料层 4相接触。
需要指出的是, 通过引线对所述下电极层 31和上电极层 32施加操作信号以实现对相变 存储单元的操作。 此为本领域技术人员所熟知的内容, 在此不再一一赘述。 如图 1A至图 1G所示, 本发明还提供上述相变存储单元的制备方法, 至少包括以下步 骤:
首先执行步骤 1 ), 如图 1A所示, 提供一表面形成有第一介质材料层 21的 Si衬底 1, 而后自下而上依次在所述第一介质材料层 21 上沉积下电极层 31 和第二介质材料层 22。 其 中, 所述第一介质材料层 21 为半导体工艺中常用的绝缘介质材料, 至少包括氧化硅、 氧化 锗、 氧化镓、 氮化硅、 氮化锗或氮化镓等中的任意一种; 所述第一介质材料层 21 的厚度范 围为 2~10纳米; 所述下电极层 31的材料均选自良导体, 至少包括 Cu、 TiN、 W、 Ta、 Ti和 Pt中的任意一种、 或上述良导体合金的任意一种; 所述第二介质材料层 22为半导体工艺中 常用的不含氧的绝缘介质材料, 至少包括氮化镓、 氮化锗或氮化硅等中的任意一种。
在本实施例中, 所述第一介质材料层 21优选 6纳米厚的氧化硅; 所述下电极层 31的材 料优选 TiN; 所述第二介质材料层 22优选氮化硅。 接着执行步骤 2)。
在步骤 2) 中, 如图 1B 所示, 利用涂胶、 曝光、 刻蚀、 去胶等工艺对部分所述第二介 质材料层 22进行光刻、 刻蚀, 直至暴露所述下电极层 31以形成窗口 A, 其中, 所述窗口 A 的开口宽度 W1的范围为 10~100纳米, 在本实施中, 优选窗口 A的开口宽度 W1为 60纳 米。 接着执行步骤 3 )。
在步骤 3 ) 中, 如图 1C所示, 采用化学气相沉积 (Chemical Vapor Deposition, CVD) 或原子层沉积 (Atomic Layer Deposition, ALD) 方法, 在步骤 3 ) 获得的结构表面沉积相变 材料以形成具有第一厚度 D的相变材料层 4。 其中, 所述相变材料至少包括 Ge-Sb-Te、 Ge- Te或 Ti-Sb-Te 中的任意一种, 上述三种材料做成的相变存储单元均具有很好的电学操作表 现; 所述第一厚度 D的范围为 6~60埃之间, 以供相变存储单元利用相变材料层的界面电阻 差异存储信息, 进一步, 所述第一厚度 D的范围为 6~20埃之间; 在相变材料层 4的另外二 维方向上, 所述相变材料层 4的长度范围为 50~100埃之间, 所述相变材料层 4的宽度范围 为 50~100埃之间。
在本实施例中, 相变材料优选 Ti-Sb-Te; 优选所述第一厚度 D为 15埃; 优选所述相变 材料层 4的长度为 80埃, 所述相变材料层 4的宽度为 80埃, 但并不局限于所述相变材料层 4的长度和宽度相等的情况, 在其他实施例中, 所述相变材料层的长度和宽度也可以不等。
需要说明的是, 传统的相变材料层具有长度、 宽度和厚度三维材料尺寸, 在本发明中, 所述相变材料层 4将三维材料尺寸中的一个维度上的尺寸缩小, 本实施例中, 相变材料层 4 小尺寸对应的维度为厚度, 具体地, 通过调节相变材料沉积时间实现精确控制相变材料层 4 的厚度, 使相变材料层 4在厚度缩小至与单个晶胞或者多个晶胞尺度相当的第一厚度, 从而 在厚度很薄的情况下, 可以忽略相变材料层 4 的厚度将三维相变材料层视为二维相变材料 层。 该很小的第一厚度抑制了大晶粒的形成, 从而一方面使相变区域减小, 显著降低相变存 储单元 RESET 操作功耗, 另一方面, 弱化相变材料层的体材料特性, 且保证相变材料层的 可逆相变行为以界面特性为主, 此时, 界面特性为相变存储单元存储信息的关键。 进一步, 由于相变材料尺寸小, 有利于相变存储单元尺寸缩小, 使本发明的相变存储单元具备超高密 度信息存储的潜力。
需要进一步说明的是, 由于相变材料在相变前后可实现 Ge 原子跳跃的最小单元为边长 约为 6埃的立方体。 因此, 为了保证相变材料反复相变的能力, 相变材料层 4的厚度 (即第 一厚度) 必须不小于 6埃。 同时为了保证相变材料层 4体现出界面效应占主导的特性, 相变 材料层 4 的厚度 (即第一厚度) 应小于十个最小单元的尺寸, 即 60埃。 从而, 本发明中相 变材料层 4的厚度 (即第一厚度) 控制在 6~60埃之间。
尺寸效应对相变材料热稳定性的影响如下: 当相变材料层厚度在 10 纳米以上时, 相变 材料的结晶温度随厚度的变化非常微弱。 当相变材料层厚度低于 10 纳米时, 相变材料的结 晶温度随着厚度的减小有不同程度的提高。 本实施例的相变材料层 4的厚度 (即第一厚度) 范围为 6~60埃之间, 进一步优选为 6~20埃之间, 从而相变材料层 4的结晶温度随着第一厚 度的减小有不同程度的提高。
同时, 尺寸效应对相变材料的结晶速度的影响如下: 当相变材料厚度减薄, 材料的比表 面积增加, 而相变材料的界面由于存在缺陷而容易形成晶胞。 而晶胞的存在为相变材料的结 晶过程缩短了晶胞形成时间, 减少了结晶过程所需时间, 进而提升了相变存储器的操作速 度。 当晶胞形成时间缩短, 晶粒成长变成为影响结晶时间的主要因素, 而晶粒生长时间随着 尺寸缩小而变短, 这就保证了小尺寸器件更快的相变速度。 接着执行步骤 4) 。
在步骤 4) 中, 如图 1D所示, 利用至少包括涂胶、 曝光、 刻蚀及去胶等工艺的光刻及 刻蚀工艺、 或利用聚焦离子束 FIB, 去除部分位于所述下电极层 31 上的相变材料层 4直至 暴露位于其下的下电极层 31, 亦即去除部分位于所述窗口 A中且与所述下电极层 31相接触 的相变材料层 4并暴露出部分下电极层 31, 以将相变材料层 4分割为两部分, 从而以一一 对应的方式为一对相变存储单元分别提供相变材料层。 如图 1D所示, 在本实施例中, 优选 被去除部分的相变材料层的对称轴为窗口 A 的中心线, 以将所述相变材料层 4 二等分。 接 着执行步骤 5) 。
在步骤 5) 中, 如图 1E所示, 利用低温化学气相沉积或者低温原子层沉积方法, 在步 骤 4) 获得的结构表面沉积第三介质材料层 23, 填充满所述窗口 A的同时隔离步骤 4) 中分 割为两部分的相变材料层 4, 其中, 所述第三介质材料层 23 为半导体工艺中常用的不含氧 的绝缘介质材料, 至少包括氮化镓、 氮化锗或氮化硅中的任意一种, 在本实施例中, 所述第 三介质材料层 23优选氮化锗。
需要说明的是, 第二介质材料层 22和第三介质材料层 23之所以不含氧原因在于与相变 材料接触的材料不宜用含氧的材料, 且由于所述第二介质材料层 22和第三介质材料层 23均 与相变材料层 4相接触, 因此, 所述第二介质材料层 22和第三介质材料层 23为半导体工艺 中常用的不含氧的绝缘介质材料。
需要进一步说明的是, 在本发明中, 并未限制第一介质材料层 21、 第二介质材料层 22 及第三介质材料层 23 可否选择一致的介质材料, 亦即三者之间可以相同也可以各不相同, 或者可以两两相同。
需要指出的是, 在所述步骤 5) 中沉积所述第三介质材料层 23 时, 优选抑制所述第三 介质材料层 23与相变材料层 4界面处缺陷的形成, 以保证该界面的平整, 为相变材料层 4 与第三介质材料层 23晶格匹配提供有利条件, 亦即所述相变材料层 4与第三介质材料层 23 形成良好的界面, 亦即该界面存在少量的缺陷。
需要说明的是, 由于相变材料层具有很好的界面, 保证相变材料层中为数较少的原子在 可逆相变过程中的稳定性, 进而保证可逆相变过程中相变材料的组份与原子数目保持稳定, 不被氧化, 提高相变材料层的热稳定性, 使相变存储单元能在更高的工作温度可靠工作; 由 于相变材料层界面上存在少量的缺陷, 一方面可使多晶时相变材料的晶胞晶格有畸变, 在电 脉冲产生的热冲击下, 有利于多晶向非晶转化, 同时也可使非晶时相变材料的晶胞的形成容 易, 从而对结晶速度提升的效果变得更加明显, 进而有效提高相变存储单元的 SET 操作速 度。 接着执行步骤 6) 。
在步骤 6) 中, 如图 1F所示, 利用化学机械抛光 (CMP) 工艺, 平坦化处理步骤 5) 获得的结构, 直至暴露所述第一介质材料层 21及部分相变材料层 4, 以使所述相变材料层 4 的横截面为两个相对的且互不接触的 L型。 接着执行步骤 7) 。
在步骤 7 ) 中, 如图 1G所示, 形成覆盖于所述被暴露的相变材料层 4 上的上电极层 32, 以完成两个相变存储单元的制备。 具体地, 在本实施例中, 利用物理气相沉积 (Physical Vapor Deposition, PVD) , 低温化学气相沉积或低温原子层沉积上电极层 32, 光 刻、 刻蚀所述上电极层 32, 形成覆盖于所述被暴露的相变材料层 4上的上电极层 32。
而后, 通过引线对所述下电极层 31和上电极层 32施加操作信号以实现对相变存储单元 的操作; 用于实现相变存储单元读写擦功能的单元驱动器件包括晶体管或二极管, 以使相变 存储单元的制备工艺与 CMOS 工艺可以达到完全兼容, 其中, 所述单驱动器件为晶体管时 形成 1T1R结构, 所述单元驱动器件为二极管时形成 1D1R结构。 此为本领域技术人员所熟 知的内容, 在此不再一一赘述。 需要说明的是, 抑制相变材料层 4界面处缺陷, 使相变材料层 4与第三介质材料层 23 形成该良好的界面的原因在于:
界面处由于晶格失配等原因导致在相变材料层界面上存在缺陷, 缺陷的存在可以降低结 晶所需的能量, 缩短晶核形成的时间, 从而提高相变材料的结晶速度; 在界面的相变材料的 晶胞存在缺陷, 原子结构破裂或者键能很小, 在外力的作用下容易电离, 尤其当晶粒尺寸很 小时, 比表面积很大, 因此很多原子电离, 离子偏离原来的位置, 引起结构的快速变化; 同 时结构变化所需的能量也因缺陷的存在降低, 从而降低了相变材料非晶化的功耗。
但是, 过多缺陷的存在将使成核更容易, 相变材料的结晶在更低的温度下进行, 则相变 材料非晶态的热稳定性降低, 将导致相变存储单元的数据保持力降低; 当相变材料尺寸小到 一个或者几个晶胞的尺度时, 界面过多缺陷的存在将影响相变材料中正常晶态原子结构的形 成, 导致原子结构完全失常, 从而使相变材料失去可逆相变的能力。 由于以上原因, 则本发 明制备相变存储单元的界面缺陷密度应控制在较小的程度。 为了使本领域技术人员进一步理解本发明, 以下将详述相变存储单元的相关原理: 在本发明中, 由于相变材料层厚度薄, 使本发明的相变存储单元存储的总电阻主要由界 面处的接触电阻 (即界面电阻) 决定。 相变存储单元通过对相变材料进行晶态与非晶态的控 制, 使界面处的接触电阻产生巨大差异, 从而使相变存储单元体现出两种区别明显的高低电 阻, 实现对数据的存储。 其中, 利用亚纳秒量级电脉冲信号将相变存储单元中的相变材料操 作成非晶态, 电极与相变材料界面呈高阻态, 称为 RESET 态; 利用亚纳秒量级电脉冲信号 将相变存储单元中的相变材料操作成晶态, 电极与相变材料界面呈低阻态, 称为 SET态。
在本实施例中, 非晶态相变材料与电极材料界面处的接触电阻为第一接触电阻, 晶态相 变材料与电极材料界面处的接触电阻为第二接触电阻, 且第一接触电阻与第二接触电阻的比 值范围为 103~105, 其中, 所述电极材料包括上电极层和下电极层。
针对相变存储单元存储的总电阻而言, 本实施例中, RESET 态相变存储单元的总电阻 与 SET态相变存储单元的总电阻的比值范围为 10~105倍。
本发明中, 由于相变材料的厚度与单个晶胞或者多个晶胞尺度相当, 因此非晶到晶态转 变只形成小尺寸的数个晶胞, 而抑制了大晶粒的形成, 则原子从无序状态排列成有序状态所 需迁移的距离短, 花费时间少, 使本发明的相变存储单元的操作具备高速的优点。 本发明 中, 相变材料层厚度薄减小了相变区域从而使操作功耗的降低, 同时相变材料层厚度薄及相 变材料层界面上存在少量的缺陷, 提高了操作速度从而使操作时间缩短, 都减少了每次操作 过程对相变材料的损害, 使每次操作对材料的元素偏析效果降低, 增加了相变存储单元的最 大可操作次数, 从而有利于提高器件循环操作次数的能力。
综上, 本发明的少量相变材料层晶胞的可逆相变行为、 非晶与多晶基本单元的相近与相 似行为、 界面缺陷行为、 金属与相变材料在非晶与多晶电阻差别较大的行为, 可实现与新型 CMOS的兼容, 并在 10纳米以下的技术节点、 高速、 低功耗方面显现出巨大的能力。 实施例二
如图 2D所示, 本发明提供一种相变存储单元, 至少包括: Si衬底 1、 第二介质材料层 22、 电极对 5、 相变材料层 4及第三介质材料层 23。 其中, 图 2D为本实施例中一个相变存 储单元的结构示意图。
所述第二介质材料层 22形成于所述 Si衬底 1表面。 其中, 所述第二介质材料层 22为 不含氧的半导体介质材料, 至少包括氮化镓、 氮化锗或氮化硅中的任意一种。 在本实施例中 为氮化硅。
所述电极对 5形成于所述第二介质材料层 22表面, 所述电极对 5中的两个电极之间具 有第一距离且相互隔离。 其中, 所述第一距离的范围为 10~100 纳米, 在本实施例中, 优选 所述第一距离为 60纳米; 所述电极对 5的材料包括金属或石墨烯, 所述金属至少包括 Cu、 TiN、 W、 Ta、 Ti和 Pt中的任意一种、 或上述金属合金的任意一种, 在本实施例中, 优选所 述电极对 5的材料为石墨烯。
所述相变材料层 4 具有第一厚度 D, 形成于所述电极对 5 及第二介质材料层 22 的表 面, 同时, 相变材料层 4的宽度 W4小于等于电极对 5的宽度 W5, 其中, 所述相变材料层 4的宽度 W4和电极对 5的宽度 W5如图 2C中所示, 在本实施例中, 优选所述相变材料层 4 的宽度小于等于电极对 5的宽度; 所述第一厚度 D的范围为 6~60埃之间, 以供相变存储单 元利用相变材料层的界面电阻差异存储信息, 进一步, 所述第一厚度 D的范围为 6~20埃之 间, 本实施例中, 优选所述第一厚度 D为 15埃。
当相变材料层 4的宽度 W4小于电极对 5的宽度 W5时, 所述第三介质材料层 23形成 于相变材料层 4及电极对 5的表面上, 或者当相变材料层 4的宽度 W4等于电极对 5的宽度 W5 时, 所述第三介质材料层 23 形成于相变材料层 4 的表面上。 在本实施例中, 所述第三 介质材料层 23形成于相变材料层 4及电极对 5的表面上。 其中, 所述第三介质材料层 23的 厚度为 20~100纳米, 亦即所述第三介质材料层 23的表面与相变材料层 4的表面之间的范围 为 20~100纳米, 在本实施例中, 优选所述第三介质材料层 23的厚度为 60纳米; 所述第三 介质材料层 23 为不含氧的半导体介质材料, 至少包括氮化镓、 氮化锗或氮化硅中的任意一 种, 在本实施例中, 所述第三介质材料层 23优选氮化锗。
需要指出的是, 通过引线对所述电极对 5施加操作信号以实现对相变存储单元的操作。 此为本领域技术人员所熟知的内容, 在此不再一一赘述。 如图 2A至图 2D所示, 本发明还提供上述相变存储单元的制备方法, 至少包括以下步 骤:
首先执行步骤 1 ) , 提供一表面形成有第二介质材料层 22的 Si衬底 1, 并在所述第二 介质材料层 22上制备一电极对 5, 其中, 所述电极对 5之间的间距为第一距离, 所述第一 距离的范围为 10~100纳米; 所述电极对 5 的材料包括金属或石墨烯, 其中, 所述金属至少 包括 Cu、 TiN、 W、 Ta、 Ti和 Pt 中的任意一种、 或上述金属合金的任意一种; 所述第二介 质材料层 22 为半导体工艺中常用的不含氧的绝缘介质材料, 至少包括氮化镓、 氮化锗或氮 化硅中的任意一种。
在本实施例中, 优选所述第一距离的范围为 60 纳米; 所述电极对 5 的材料优选石墨 烯, 但并不局限于此, 在另一实施例中, 所述电极对 5的材料还可以优选为 TiN; 所述第二 介质材料层 22优选氮化硅。
需要指出的是, 石墨烯是一种由碳原子构成的单层片状结构的新材料, 具有电阻率极 低, 电子迁移的速度极快, 烯的结构非常稳定、 非同寻常的导电性能、 超出钢铁数十倍的强 度和极好的透光性。 常温下石墨烯的电子迁移率超过 15000 cm2/Vs, 电子能够极为高效地 迁移, 而传统的半导体和导体, 例如硅、 纳米碳管或铜, 均没有石墨烯的电子迁移率高; 石 墨烯的电阻率只约 10_6 Ω -cm, 且比铜或银的电阻率更低, 为目前世上电阻率最小的材料; 石墨烯的结构非常稳定, 石墨烯内部的碳原子之间的连接很柔韧, 当施加外力于石墨烯时, 碳原子面会弯曲变形, 使得碳原子不必重新排列来适应外力, 从而保持结构稳定, 体现出牢 固坚硬的特点; 同时由于电子和原子的碰撞, 传统的半导体和导体用热的形式释放了一些能 量, 目前一般的电脑芯片以这种方式浪费了 72%-81%的电能, 石墨烯则不同, 它的电子能 量不会被损耗, 这使它具有了非比寻常的优良特性。 从而, 本发明中采用的石墨烯作为电极 对, 具有信号响应快, 机械强度大, 能量损耗少等特点。 接着执行步骤 2) 。
步骤 2) 中, 如图 2B所示, 采用化学气相沉积 (CVD) 或原子层沉积 (ALD) 在步骤 1 ) 获得的结构表面沉积相变材料以形成具有第一厚度 D 的相变材料层 4。 其中, 所述第一 厚度 D 的范围为 6~60 埃之间, 以供相变存储单元利用相变材料层的界面电阻差异存储信 息, 进一步, 所述第一厚度 D的范围为 6~20埃之间, 本实施例中, 优选所述第一厚度 D为 15埃。
需要指出的是, 本实施例步骤 2) 关于相变材料层 4材料类别、 尺寸及产生的有益效果 的相关描述与实施例一的基本相同, 因此相同之处请参阅实施例一的描述在此不再一一赘 述, 不同之处仅在于, 本实施例中的相变材料层 4在长度和宽度尺寸上不同于实施例一。 继 续执行步骤 3 ) 。
在步骤 3 ) 中, 如图 2C 所示, 利用涂胶、 曝光、 刻蚀、 去胶等工艺对所述相变材料层 4进行光刻和刻蚀, 形成宽度小于等于电极对 5宽度的相变材料层 4, 其中, 所述相变材料 层 4的宽度 W4和电极对 5的宽度 W5如图 2C中所示。 接着执行步骤 4) 。
步骤 4) 中, 如图 2D所示, 利用低温化学气相沉积或者低温原子层沉积, 在步骤 3 ) 获得的结构表面沉积第三介质材料层 23, 并填充满所述电极对 5 之间区域。 其中, 所述第 三介质材料层 23的厚度为 20~100纳米, 在本实施例中, 优选所述第三介质材料层 23的厚 度为 60纳米; 所述第三介质材料层 23为不含氧的半导体介质材料, 至少包括氮化镓、 氮化 锗或氮化硅中的任意一种, 在本实施例中, 所述第三介质材料层 23优选氮化锗。
需要说明的是, 本实施例中, 第二介质材料层 22和第三介质材料层 23为不含氧的介质 材料的原因请参阅实施例一的相关描述; 进一步, 本发明未限制第二介质材料层 22和第三 介质材料层 23选择的材料可否一致。
需要进一步说明的是, 本实施例步骤 4) 沉积第三介质材料层 23 时, 优选抑制所述第 三介质材料层 23与相变材料层 4界面处缺陷的形成, 以保证该界面的平整, 为相变材料层 4与第三介质材料层 23 晶格匹配提供有利条件, 亦即所述相变材料层 4与第三介质材料层 23 形成良好的界面, 亦即该界面存在少量的缺陷。 关于形成该良好的界面特性的原因及带 来的有益效果的相关描述, 请参阅实施例一中的相应的具体内容。
需要指出的是, 通过引线对所述电极对 5施加操作信号以实现对相变存储单元的操作; 用于实现相变存储单元读写擦功能的单元驱动器件包括晶体管或二极管, 以使相变存储单元 的制备工艺与 CMOS工艺可以达到完全兼容, 其中, 所述单驱动器件为晶体管时形成 1T1R 结构, 所述单元驱动器件为二极管时形成 1D1R结构。 此为本领域技术人员所熟知的内容, 在此不再一一赘述。 此为本领域技术人员所熟知的内容, 在此不再一一赘述。
关于相变存储单元的相关原理及产生的有益效果详情请参阅实施例一的相关描述。 综上所述, 本发明的相变存储单元及其制备方法, 在相变存储单元中, 采用的相变材料 层的厚度与单个晶胞或者多个晶胞尺度相当, 一方面使相变区域减小, 另一方面, 弱化相变 材料层的体材料特性, 且保证相变材料层的可逆相变行为以界面特性为主, 从而制备出利用 界面电阻变化来存储信息的高密度、 低功耗、 高速二维相变存储单元。 本发明中, 由于相变 材料层薄及相变材料层界面上存在少量的缺陷, 促使相变存储单元操作功耗的降低和操作时 间的缩短, 减少了每次操作过程对相变材料的损害, 使每次操作对材料的元素偏析效果降 低, 增加了相变存储单元的最大可操作次数, 从而有利于提高器件循环操作次数的能力; 进 一步, 本发明中采用的石墨烯电极对具有信号响应快、 机械强度大、 能量损耗少等特点, 使 基于石墨烯电极的相变存储单元具有高速, 低功耗, 长寿命的优点。 本发明的少量相变材料 层晶胞的可逆相变行为、 非晶与多晶基本单元的相近与相似行为、 界面缺陷行为、 金属与相 变材料在非晶与多晶电阻差别较大的行为, 可实现与新型 CMOS的兼容, 并在 10纳米以下 的技术节点、 高速、 低功耗方面显现出巨大的能力。 所以, 本发明有效克服了现有技术中的 种种缺点而具高度产业利用价值。
上述实施例仅例示性说明本发明的原理及其功效, 而非用于限制本发明。 任何熟悉此技 术的人士皆可在不违背本发明的精神及范畴下, 对上述实施例进行修饰或改变。 因此, 举凡 所属技术领域中具有通常知识者在未脱离本发明所揭示的精神与技术思想下所完成的一切等 效修饰或改变, 仍应由本发明的权利要求所涵盖。

Claims

权利要求书 、 一种相变存储单元的制备方法, 其特征在于, 所述制备方法至少包括以下步骤:
1 ) 提供一表面形成有第一介质材料层的 Si 衬底, 自下而上依次在所述第一介质材 料层上形成下电极层和第二介质材料层;
2) 光刻、 刻蚀部分所述第二介质材料层直至暴露所述下电极层以形成窗口;
3 ) 在步骤 2) 获得的结构表面沉积相变材料以形成具有第一厚度的相变材料层;
4) 去除部分位于所述下电极层上的相变材料层, 以将相变材料层分割为两部分, 从 而为一对相变存储单元分别提供相变材料层;
5) 在步骤 4) 获得的结构表面沉积第三介质材料层, 填充满所述窗口的同时隔离步 骤 4) 中分割为两部分的相变材料层;
6) 利用化学机械抛光工艺平坦化处理步骤 5) 获得的结构, 直至暴露所述第一介质 材料层及部分相变材料层, 以使所述相变材料层的横截面为两个相对的 L型;
7) 形成覆盖于所述被暴露的相变材料层上的上电极层。 、 根据权利要求 1 所述的相变存储单元的制备方法, 其特征在于: 所述第一介质材料层的 厚度范围为 2~10纳米。 、 根据权利要求 1 所述的相变存储单元的制备方法, 其特征在于: 所述窗口的开口宽度范 围为 10~100纳米。 、 根据权利要求 1 所述的相变存储单元的制备方法, 其特征在于: 所述第一厚度的范围为 6-60埃之间, 以供相变存储单元利用相变材料层的界面电阻差异存储信息。 、 根据权利要求 4所述的相变存储单元的制备方法, 其特征在于: 所述第一厚度的范围为 6~20埃之间。 、 根据权利要求 1 所述的相变存储单元的制备方法, 其特征在于: 所述相变材料层的长度 范围为 50~100埃之间, 所述相变材料层的宽度范围为 50~100埃之间。 、 根据权利要求 1 所述的相变存储单元的制备方法, 其特征在于: 所述相变材料至少包括 Ge-Sb-Te、 Ge-Te或 Ti-Sb-Te中的任意一种。 、 根据权利要求 1 所述的相变存储单元的制备方法, 其特征在于: 用于实现相变存储单元 读写擦功能的单元驱动器件包括晶体管或二极管, 其中, 所述单驱动器件为晶体管时形 成 1T1R结构, 所述单元驱动器件为二极管时形成 1D1R结构。 、 一种相变存储单元, 其特征在于, 所述相变存储单元包括:
Si衬底;
形成于所述 Si衬底表面的第一介质材料层;
形成于所述第一介质材料层表面的下电极层;
上表面均位于同一平面、 均形成于所述下电极层上且均与其接触的第二介质材料 层、 相变材料层和第三介质材料层, 其中, 具有第一厚度的所述相变材料层将第二介质 材料层和第三介质材料层隔离;
与所述相变材料层相接触的上电极层。 、 根据权利要求 9 所述的相变存储单元, 其特征在于: 所述相变材料层的横截面为被 第三介质材料层隔离的两个相对的 L型, 其中, 与所述下电极层相接触的为 L型的一边 为第一边, 与所述第一边相垂直的 L型另一边为第二边, 所述第一边和第二边的厚度均 为第一厚度。 1、 根据权利要求 9或 10所述的相变存储单元, 其特征在于: 所述第一厚度的范围为 6~60 埃之间, 以供相变存储单元利用相变材料层的界面电阻差异存储信息。 、 根据权利要求 11 所述的相变存储单元, 其特征在于: 所述第一厚度的范围为 6~20埃 之间。 3、 一种相变存储单元的制备方法, 其特征在于, 所述制备方法至少包括以下步骤:
1 ) 提供一表面形成有第二介质材料层的 Si 衬底, 并在所述第二介质材料层上制备 一电极对, 其中, 所述电极对之间的间距为第一距离;
2) 在步骤 1 ) 获得的结构表面沉积相变材料以形成具有第一厚度的相变材料层;
3 ) 光刻、 刻蚀所述相变材料层, 形成宽度小于等于电极对宽度的相变材料层; 4) 在步骤 3 ) 获得的结构表面沉积第三介质材料层, 并填充满所述电极对之间区 域。 、 根据权利要求 13 所述的相变存储单元的制备方法, 其特征在于: 所述第一厚度的范围 为 6~60埃之间, 以供相变存储单元利用相变材料层的界面电阻差异存储信息。 、 根据权利要求 14所述的相变存储单元的制备方法, 其特征在于: 所述第一厚度的范围 为 6~20埃之间。 、 根据权利要求 13所述的相变存储单元的制备方法, 其特征在于: 所述第一距离的范 围为 10~100纳米。 、 根据权利要求 13所述的相变存储单元的制备方法, 其特征在于: 所述第三介质材料 层的厚度为 20~100纳米。 、 根据权利要求 13所述的相变存储单元的制备方法, 其特征在于: 所述电极对的材料 为金属、 金属合金、 金属氮化物或者石墨烯。 、 一种相变存储单元, 其特征在于, 所述相变存储单元包括:
Si衬底;
形成于所述 Si衬底表面的第二介质材料层;
形成于所述第二介质材料层表面、 且之间具有第一距离的电极对;
形成于所述电极对及第二介质材料层的表面、 宽度小于等于电极对宽度、 且具有第 一厚度的相变材料层;
形成于相变材料层及电极对的表面上、 或形成于相变材料层表面上的第三介质材料 层。
、 根据权利要求 19所述的相变存储单元, 其特征在于: 所述第一厚度的范围为 6~60埃 之间, 以供相变存储单元利用相变材料层的界面电阻差异存储信息。 、 根据权利要求 20所述的相变存储单元, 其特征在于: 所述第一厚度的范围为 6~20埃 之间。 、 根据权利要求 19 所述的相变存储单元, 其特征在于: 所述第一距离的范围为
10-100纳米。 、 根据权利要求 19所述的相变存储单元, 其特征在于: 所述第三介质材料层的厚度为 20-100纳米。 、 根据权利要求 19所述的相变存储单元, 其特征在于: 所述电极对的材料为金属、 金 属合金、 金属氮化物或者石墨烯。
PCT/CN2014/075276 2013-07-19 2014-04-14 相变存储单元及其制备方法 WO2015007108A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US15/739,900 US20180269388A1 (en) 2013-07-19 2014-04-14 Phase change memory unit and preparation method therefor

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310306025.X 2013-07-19
CN201310306025.XA CN103346258B (zh) 2013-07-19 2013-07-19 相变存储单元及其制备方法

Publications (1)

Publication Number Publication Date
WO2015007108A1 true WO2015007108A1 (zh) 2015-01-22

Family

ID=49281042

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/075276 WO2015007108A1 (zh) 2013-07-19 2014-04-14 相变存储单元及其制备方法

Country Status (3)

Country Link
US (1) US20180269388A1 (zh)
CN (1) CN103346258B (zh)
WO (1) WO2015007108A1 (zh)

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103346258B (zh) * 2013-07-19 2015-08-26 中国科学院上海微系统与信息技术研究所 相变存储单元及其制备方法
CN103515535A (zh) * 2013-10-10 2014-01-15 中国科学院苏州纳米技术与纳米仿生研究所 相变存储器接触电极的制备方法及相变存储器接触电极
CN103904214B (zh) * 2014-03-03 2017-06-16 上海新储集成电路有限公司 一种二维相变存储器单元结构及其制造方法
CN105322090B (zh) * 2014-06-13 2018-09-25 中国科学院上海微系统与信息技术研究所 一种存储器及其制作方法
CN105393375B (zh) * 2014-06-26 2018-12-14 华为技术有限公司 一种金属掺杂的Ge-Sb-Te基多值存储相变材料及相变存储器
US10141503B1 (en) * 2017-11-03 2018-11-27 International Business Machines Corporation Selective phase change material growth in high aspect ratio dielectric pores for semiconductor device fabrication
US10808316B2 (en) * 2018-05-10 2020-10-20 International Business Machines Corporation Composition control of chemical vapor deposition nitrogen doped germanium antimony tellurium
CN109560104A (zh) * 2018-12-20 2019-04-02 上海集成电路研发中心有限公司 一种相变存储器及其制备方法
CN110752292A (zh) * 2019-09-24 2020-02-04 华中科技大学 用覆盖层的界面效应调控相变材料晶化阈值的方法
CN112015367B (zh) * 2020-08-26 2024-05-31 上海新氦类脑智能科技有限公司 非易失性布尔逻辑操作单元及方法、装置
CN113437213A (zh) * 2021-06-02 2021-09-24 长江先进存储产业创新中心有限责任公司 相变存储器及相变存储器的制作方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1554125A (zh) * 2001-09-07 2004-12-08 ض� 相变材料存储器装置
CN101136453A (zh) * 2006-08-29 2008-03-05 国际商业机器公司 薄膜相变存储单元及其制造方法
CN103346258A (zh) * 2013-07-19 2013-10-09 中国科学院上海微系统与信息技术研究所 相变存储单元及其制备方法

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100639206B1 (ko) * 2004-06-30 2006-10-30 주식회사 하이닉스반도체 상변환 기억 소자 및 그 제조방법
EP1844501A1 (en) * 2005-01-25 2007-10-17 Nxp B.V. Fabrication of a phase-change resistor using a backend process
ATE420440T1 (de) * 2005-05-19 2009-01-15 Koninkl Philips Electronics Nv Verfahren zum regeln des zuerst schmelzenden bereiches einer pcm-zelle und damit hergestellte vorrichtungen
JP4492816B2 (ja) * 2006-10-03 2010-06-30 株式会社半導体理工学研究センター 多値記録相変化メモリ素子、多値記録相変化チャンネルトランジスタおよびメモリセルアレイ
TWI347670B (en) * 2007-02-01 2011-08-21 Promos Technologies Inc Phase-change memory and fabrication method thereof
EP2156481B1 (en) * 2007-05-31 2012-10-31 Nxp B.V. An electronic device comprising a convertible structure, and a method of manufacturing an electronic device
KR101458953B1 (ko) * 2007-10-11 2014-11-07 삼성전자주식회사 Ge(Ⅱ)소오스를 사용한 상변화 물질막 형성 방법 및상변화 메모리 소자 제조 방법
JP2010027835A (ja) * 2008-07-18 2010-02-04 Renesas Technology Corp 不揮発性記憶装置およびその製造方法
US8685291B2 (en) * 2009-10-13 2014-04-01 Ovonyx, Inc. Variable resistance materials with superior data retention characteristics

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1554125A (zh) * 2001-09-07 2004-12-08 ض� 相变材料存储器装置
CN101136453A (zh) * 2006-08-29 2008-03-05 国际商业机器公司 薄膜相变存储单元及其制造方法
CN103346258A (zh) * 2013-07-19 2013-10-09 中国科学院上海微系统与信息技术研究所 相变存储单元及其制备方法

Also Published As

Publication number Publication date
US20180269388A1 (en) 2018-09-20
CN103346258B (zh) 2015-08-26
CN103346258A (zh) 2013-10-09

Similar Documents

Publication Publication Date Title
WO2015007108A1 (zh) 相变存储单元及其制备方法
US7033856B2 (en) Spacer chalcogenide memory method
CN109860387A (zh) 具有选择器件的pcram结构
TW202036892A (zh) 相變化隨機存取記憶體裝置及其製造方法
US20070164266A1 (en) Semiconductor device and method of manufacturing the same
US11031435B2 (en) Memory device containing ovonic threshold switch material thermal isolation and method of making the same
US8344347B2 (en) Multi-layer electrode structure
US7811933B2 (en) CMOS-process-compatible programmable via device
US8828785B2 (en) Single-crystal phase change material on insulator for reduced cell variability
WO2014040358A1 (zh) 含有TiSiN材料层的相变存储单元及其制备方法
JP2009141369A (ja) 相変化物質を用いた電子素子及び相変化メモリ素子並びにその製造方法
CN107086269A (zh) 存储器结构
JP2013511154A (ja) 相変化メモリ・セルおよびこれを操作する方法
JP2011082316A (ja) 半導体記憶装置
CN102832340A (zh) 一种相变存储器单元及其制备方法
JP6270600B2 (ja) 相変化メモリ
US20130292629A1 (en) Phase change memory cell and fabrication method thereof
CN103427022A (zh) 包含三明治型电极的相变存储结构及其制备方法
CN114079005A (zh) 包括数据存储材料图案和选择器材料图案的半导体装置
US20090101885A1 (en) Method of producing phase change memory device
CN102637821B (zh) 一种相变存储单元及其形成方法
CN103794722A (zh) 新型相变存储单元结构及其制备方法
CN106185800B (zh) 一种GeTe/Ge类超晶格纳米相变薄膜材料及其制备方法和应用
US11380842B2 (en) Phase change memory cell with second conductive layer
CN112968037A (zh) 相变存储器及其制作方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14826127

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14826127

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 15739900

Country of ref document: US