• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    PURPOSE: A method of forming a DRAM on a silicon chip is provided, where an NMOSFET of a memory cell is provided in the center region of the silicon chip, a CMOSFET of a support circuit is provided in the peripheral region of the silicon chip. CONSTITUTION: A support circuit (100B) is masked with an SiO2 film (20), and a polyside film (22) in a memory (100A) is doped with N-type impurities. An SiN cap layer (26) is deposited thereon and covered with a patterned mask, the laminate is successively etched up to a gate insulating oxide film (12), a substrate (10) is doped with N-type impurities, and the source and drain region (32) of the memory (100A) are formed. A sidewall dielectric spacer (34) is formed, and after an NMOSFET memory is nearly completed, a CMOSFET is formed in the supper circuit (100B).
    公开号:KR20010006849A
    申请号:KR1020000014498
    申请日:2000-03-22
    公开日:2001-01-26
    发明作者:디르크 토벤；요한 알스마이어
    申请人:인피니언 테크놀로지스 노쓰 아메리카 코포레이션；
    IPC主号:
    专利说明:

    High performance DDR and manufacturing method {HIGH PERFORMANCE DRAM AND METHOD OF MANUFACTURE}
    DETAILED DESCRIPTION The present invention relates to dynamic random access memory (DRAM), particularly DRAM, wherein the complementary transistors of an auxiliary circuit comprise a double work function gate.
    At present, the technical level of DRAM is a silicon chip in which an array of memory cells arranged in columns and rows is placed in the center (array) and an auxiliary circuit for reading and writing binary digits (bits) stored in the memory cell is located at the periphery. Include. In general, at the current state of the art, each memory cell comprises an N-channel metal oxide semiconductor field effect transistor (N-MOSFET), and the auxiliary circuit generally comprises complementary metal oxide semiconductor field effect transistors (C-MOSFETs). It includes both known N-MOSFETs and P-channel metal oxide semiconductor field effect transistors (P-MOSFETs). In the most recent technological level of DRAM, both array transistors and auxiliary transistors use a polycide (eg, WSiX) gate formed of essentially the same stacks as all transistors. This method reduces costs, eases manufacturing and reduces manufacturing costs since all gates can be deposited and patterned at the same time. Generally, the stacks are sequential layers of n type of doped polysilicon, polyside and silicon nitride caps. However, this is a performance limitation for DRAMs, especially for P-MOSFETs in auxiliary circuits. This becomes even worse in the future, which will address the improved performance demands that will require the improved performance of auxiliary circuits in the future.
    In addition, DRAM technology generally tends to shrink cells in arrays for higher density and increased productivity. One of the most effective ways to reduce array cell size is to use borderless contacts in the source / drain regions of transistors of memory cells (typically transistor outputs coupled to the bit lines of DRAM). This requires a relatively thick SiN cap on top of the conductive portion of the gate stack. Such a cap, which is also necessary for the auxiliary circuit, makes it difficult to control the line width during etching, which in turn makes it difficult to control the gate width, another important factor of device performance.
    High performance logic circuits, on the other hand, are usually fabricated of n and p type doped polysilicon as part of the gate stack for each of the N-MOSFETs and P-MOSFETs, which is generally described as a dual work function stack. have. These are fabricated by a so-called salicide (auto-aligned silicide) method that simultaneously dope the gate polysilicon and silicon substrate to form source and drain diffusion regions. In this way, the need for SiN caps facilitates line width control. The disadvantage of this method is the need for additional masks. Because of the large thermal diffusion constant of the p-type dopant (usually boron), the method makes the process more complicated and reduces in the allowed thermal budget. This factor limits the use of annealing steps commonly used to reflux typical boroPhosPho-silicate glasses (BPSGs) because of the high aspect ratio of the space between the gates in the circuit.
    The present invention provides a method of fabricating high performance DRAMs using conventional techniques in a new manner to provide dual work function gates for the C-MOSFET transistors of the auxiliary circuit to enable improved performance for the auxiliary circuit.
    1-17 illustrate a portion of a silicon chip in which cross-sectional views of various successive fabrication steps, both N-MOSFETs and complementary pairs N-MOSFETs and P-MOSFETs suitable for memory cell array transistors are formed.
    * Description of the symbols for the main parts of the drawings *
    10 substrate 12 silicon oxide layer
    14: undoped silicon layer 22: polycide layer
    26 silicon nitride cap layer 34 sidewall dielectric spacer
    36A: BPSG layer 40A, 40B: gate area mesa
    42: spacer 57, 58, 59: salicide connection
    60A: flat layer
    In particular, the new process is used in salicide gated auxiliary circuits that are common in the CMOS art, where the use for high speed low power logic circuits is generally limited.
    Especially. The related process flow is not compromised with the ground rules applicable to the array transistors of each cell, but includes a thermal budget that provides for the desired dual work function gates for CMOS auxiliary circuit transistors and is consistent with the high electrical performance of the final product. do.
    Moreover, the method of the present invention initially comprises an etch stop layer of silicon oxide in a multilayer stack used to form auxiliary circuit transistors for patterning and gap filling steps in the fabrication of array transistors. It effectively separates the above steps in the preparation of the. In addition, since the polycide layer is part of a gate stack of array transistors only, the auxiliary circuit transistors can be easily made using conventional salicide techniques. Finally, the process steps are arranged such that the p-type dopants in which the thermal budget is critical are critical to the fabrication of only the array transistors that have not yet been implanted into the silicon chip. In addition, the process allows the space between auxiliary circuit gates to have a reduced aspect ratio so that the space is charged at a lower temperature than otherwise.
    Thus, the present invention can be widely regarded as a process for providing DRAM to a semiconductor body such as a silicon chip, where memory cells are arrayed in the center of the chip and auxiliary circuits are formed in the periphery thereof. Depending on the characteristics of the DRAM technology level, the memory cell transistors are all N-MOSFETs and the auxiliary circuit includes N-MOSFETs and P-MOSFETs. An important feature is that each array N-MOSFETs contain N-doped polyside gate contacts and each MOSFET in the auxiliary circuit comprises a suitably doped polysilicon gate. A related feature is that the source, drain and gate contacts of the auxiliary circuit transistors are all formed by an automatic alignment technique that forms salicide contacts with no boundary.
    In terms of devices, the present invention includes a silicon chip in which a memory cell array each containing an N-MOSFET is formed in a central portion of a DRAM and an auxiliary circuit including both N-MOSFETs and P-MOSFETs is formed in a periphery thereof. DRAM, wherein the N-MOSFETs of the memory cell use N-doped polyside gates, the N-MOSFETs of the auxiliary circuit use N-doped polysilicon gates, and the P-MOSFETs of the auxiliary circuit are P-doped Polysilicon gates.
    In terms of method, the present invention relates to a method for forming a DRAM consisting of a silicon chip comprising a memory cell array using N-MOSFETs in the center of the chip and an auxiliary circuit using C-MOSFETs in the periphery of the chip. It is about. The method includes the following steps: forming a masking layer of silicon oxide on the surface of the chip region and selectively removing the layer from the center where memory cell arrays are to be included, but suitable for the periphery where the auxiliary circuit should be included. Leaving the layer to be positioned such that it is located in a slate; Forming gate N-MOSFETs of the memory cells in the central portion and including gate conductors in the N-MOSFET including a lower polysilicon layer doped with a donor atom and an upper layer that is a metal silicide; Covering said chip region with a masking layer and selectively removing said masking layer from said center portion of said chip; Removing the silicon oxide layer from the periphery of the chip region; Covering the peripheral portion with a masking layer and removing the peripheral portion at the portion where N-MOSFETs should be formed; Forming gated N-MOSFETs of the auxiliary circuitry in the periphery and including gate conductors in the N-MOSFET including a lower polysilicon layer doped with donor atoms and an upper layer that is a metal silicide; Covering the periphery with a masking layer and removing the masking layer at a portion where a P-MOSFET is to be formed; Forming the P-MOSFET of the auxiliary circuit in the periphery and including the P-MOSFET in a gate conductor including a lower silicon layer doped with acceptor atoms and an upper layer that is a metal silicide.
    The invention will be more readily understood from the following more detailed description, which is set forth in conjunction with the accompanying drawings. The drawings do not depend on accumulation.
    As is known, in the manufacturing process of silicon integrated circuits, it is common for most of the process to take place on a relatively large silicon wafer after the wafer is diced into individual silicon chips containing the desired integrated circuit. If the description is primarily for a single chip on which a single DRAM is formed, it will facilitate the description of the process of the present invention. There will be little difficulty in understanding the description of the wafer scale process.
    1 is a multilayer comprising an array portion 100A in which an N-MOSFET is used for a memory cell array of a DRAM and a portion 100B in which complementary N-MOSFETs and P-MOSFETs are used for an auxiliary circuit of a DRAM. A portion of a silicon wafer to be a product composed of the stack 100 is shown. In general, the auxiliary circuit is concentrated in regions that are in contact with the regions where the memory cells are concentrated. Stack 100 includes a single crystal silicon substrate 10 that will include, for example, various diffusion wells, trenches, and other regions needed for the silicon substrate to form and insulate the various circuit devices required for DRAM. They are not shown because they can be formed into shapes suitable for a particular type of desired DRAM.
    Stack 100 is generally provided as follows. The gate oxide layer 12 is grown on top of the single crystal p-type silicon substrate 10, and then the undoped silicon layer 14 and the silicon oxide layer 20 are blanket deposited. As used herein, "undoped" means that no dopants are normally added to control conductivity and conductivity type.
    After proper masking, the silicon oxide 20 is etched so that the silicon oxide remains only in a portion approximately corresponding to the auxiliary portion 100B. Next, the mask is peeled off and the top of the stack 100 is covered with a suitable polycide layer (silicon nitride), such as, for example, tungsten silicide (WSi), for example by sputtering or chemical vapor deposition. All. In some cases, if a thin polysilicon layer (not shown) is first deposited on the silicon oxide, it can improve the adhesion of the deposited polycide. The resulting stack is shown in FIG.
    Chemical mechanical polishing (CMP) is then used to selectively etch the polycide layer 22 with respect to the silicon oxide layer 20 comprising the silicon oxide 20 serving as an etch stop. Hard polishing pads are preferably used to prevent excessive dishing in region 100A. Thus, as shown in FIG. 3, the top surface 21 of the stack partially formed by the polycide 22 and the silicon oxide 22 is flat and the polycide layer 22 is approximately in the array portion 100A. And the silicon oxide layer 20 is approximately limited to the auxiliary portion 100B. Preferably, the polycide layer 22 slightly erodes the secondary portion 100B of the stack, as shown.
    As shown in FIG. 4, the planar surface 21 of the stack is in turn ordered to a cap layer 26 of silicon nitride and a mask layer 28 of a suitable photoresist, for example a mask layer usable in the deep ultraviolet range of a photolithography pattern. Covered. At this time, the mask layer 28 is patterned to leave a stack, as shown in FIG.
    As shown later, this pattern in the mask layer 28 ultimately determines the region of the silicon substrate 10 to be doped, in which the source of the transistors of both the memory cells and the auxiliary circuits to be formed on the silicon substrate. N type doped in array portion 100A and p type or n type doped in auxiliary portion 100B to form a drain.
    Thereafter, several etching steps are followed and the result is shown in FIG. 6. First, silicon nitride 26 is etched to stop the polycide layer 22 of the array portion 100A and the oxide layer 20 of the auxiliary portion 100B. Then, photoresist mask 28 is stripped and timely polycide etching is performed to etch through exposed polycide layer 22. This etching is chosen so as not to etch silicon oxide so that the polyside 22 of the array portion 100A is patterned, while the silicon oxide layer 20 of the auxiliary portion 100B of the stack hardly penetrates the pattern. Remain unsettled The opening 100C between the silicon oxide layer 20 and the first polycide segment 22 shown in FIG. 6 is determined by the polycide layer 22 eroding the secondary portion of the stack, as mentioned above.
    As schematically shown in FIG. 6, the stack structure is radiated with n type dopant ions 30 to implant these ions into the exposed portions of the polysilicon layer 14 of the array portion 100A. Thermal annealing is then performed, so that adjacent polysilicon region layer 14 that is not directly radiated due to the upper portions of polycide regions 22 and nitride regions 26 in the doped exposed polysilicon region layer 14. Spread n-type dopants laterally.
    It is then etched up to the silicon oxide layer 12 of the exposed polysilicon region array portion 100A of layer 14 and both of the narrow region 100C between the array portion 100A and the auxiliary portion 100B. . The stack is then radiated back with n-type ions to form a local n-type region 32 that will function as source / drain regions of N-MOSFETs in the memory cell array portion. Inject in the exposed areas of (10). Thereafter, sidewall dielectric spacers (silicon nitride), typically made of silicon nitride, are formed on the sidewalls of the various stacks located on the chip top surface, as shown in FIG. These spacers may be formed in a known shape to form sidewall spacers. These will later function as portions of the silicon nitride layer 26 for selective borderless contact etching in the array portion of the stack as part of the salicide process.
    And, as shown in FIG. 8, a suitable capping dielectric layer 36, such as borophosposilicate glass (BPSG), is deposited throughout the stack. In general, in order to prevent auto-aligned contact etching in the array region 100A from making holes in the gate oxide layer 12, it is first necessary to lower the thin line layer, which is generally made of silicon nitride or silicon nitrate. Thermal reflux is used to fill the gaps so that the stack surface is nearly flat again. Now, the processing of the memory cell array portion of the stack structure is almost completed. Next, as shown in FIG. 8, deposition of the photoresist masking layer 38 over the entire top surface of the stack structure is followed.
    A photoresist masking layer 38 is then formed by standard photolithography, and generally by reactive ion etching (RIE), the BPSG layer 36 is sequentially etched from the secondary portion 100B of the stack. The results of the stack structure are shown in FIG. 9, where the BPSG layer 36 and photoresist layer 38 are limited to an array of stack structures. The reactive ion etching used preferably is selected to etch both silicon nitride and polysilicon such that the exposed portions of the silicon oxide layer 20 are also etched. As a result, as shown in FIG. 9, the pattern of the SiN cap nitride layer and the sidewall spacer 34 is transferred to the silicon oxide layer 20.
    Photoresist 38 is removed from memory array portion 100A. Thereafter, in auxiliary portion 100B, SiN cap layer 26 is removed by etching and the etching is completed under oxide gate 12 to reach the result shown in FIG. As shown, the undoped polysilicon layer 14 is also patterned such that the undoped polysilicon mesas 40A, 40B remain on the silicon substrate 14. The oxide layer formed during the previous process on top of the polysilicon layer 14 serves as a hard mask to protect the mesa portions 40A, 40B of the layer 14. After mesa formation, the oxide layer 20 is removed by wet etching, for example HF. Because of the initial removal of the photoresist layer 38, the BPSG layer 36 left in the portion 100C and the array portion 100A is also thin during this etching process, as shown in the structure of FIG. 10 (layer 36A). Shown). The gate oxide layer 12 is not etched because it is protected by a thin silicon nitride layer (described above but not shown) deposited before the film 36.
    Silicon nitride spacers 42 are formed on the top and sides of mesas 40A and 40B using conventional standard techniques. It is necessary to make oxide sidewalls before forming the spacers. This can be done by exposing the stack structure shown in FIG. 10 to an oxidizing environment at elevated temperatures. This forms a thin oxide layer on the sides and top of the mesas 40A and 40B.
    The formation of the sidewall spacers 42 results in the removal of the nitride layer on top of the gate oxide layer 12 described above.
    The structure is then patterned as shown in FIG. 12 to expose only the region comprising the polysilicon mesa 40A of the auxiliary portion 100B where the N-MOSFETs of complementary pairs of auxiliary circuits should be formed. Covered with resist layer 44.
    As schematically shown in FIG. 12, the stack provides donor ions for implanting donor ions into the polysilicon mesa 40A and the exposed regions 47A, 47B of the silicon substrate 10 on both sides of the mesa 40A. 46). After annealing, these regions 47A and 47B become the source and drain of the N-MOSFET of the auxiliary circuit, respectively. Polysilicon mesa 40A is also similarly doped to n type to function better as a gate conductor of N-MOSFET.
    Next, all photoresist masks 44 are removed and another photoresist mask 48 patterned to expose only the minor partial regions including mesas 40B on which the P-MOSFETs should be formed, as shown in FIG. Replaced by). As schematically shown in FIG. 13, acceptor ions 50 for doping p-type the exposed portions 49A, 49B of the silicon substrate 14 and the mesa itself on both sides of the polysilicon mesa 40B. Radiates to the gate conductor of the P-MOSFET of the auxiliary circuit. After annealing, portions 49A and 49B become the source and drain of the P-MOSFET of the auxiliary circuit, respectively, and the doped mesa 40B becomes the gate conductor.
    Mask 48 is removed and an HF etch is performed to remove oxide gate 12 between mesas 40A, 40B and residual oxide on top of mesas 40A, 40B. Thereafter, a suitable metal film 52 (generally titanium or cobalt) is usually deposited by sputtering on the entire stack, as shown in FIG. The film 52 is used to form self-aligned silicides (salicides) on the surface of the polysilicon mesas 40A and 40B. After thermal annealing, the salicide resistance contacts may be exposed to silicon on the film 52 such as the exposed top surface of the mesas 40A, 40B and the implanted surface regions 47A, 47B, 49A, 49B of the silicon substrate 10. Will be formed in direct contact with it. Thus, these contacts, which will each function as source and drain electrodes, will be automatically aligned to the source and drain regions.
    After such deposition and annealing, as shown in FIG. 15, the wet etching is performed with the salicide contacts 56A and 56B fused with the polysilicon gate conductors 40A and 40B, respectively, and the implanted source of the silicon substrate 10. It is used to remove excess metal film 52 that does not react with the exposed silicon to insulate the salicide contacts 57, 58, 59 fused with / drain regions 47A, 47B, 49A, 49B. As shown in FIG. 15, if implanted regions 47A and 47B are needed in the CMOS circuit, a single salicide contact can be formed. However, as will be appreciated, other devices may also be electrically insulated from one another by techniques such as shell trench isolation or LOCOS oxidation. For purposes of simplicity, these particular shapes are not shown. The invention can be applied to devices including insulated devices and a single salicide contact 58, as shown in FIG.
    As shown in FIG. 16, after a suitable liner (not shown) is preferably attached on the surface of the stack, a new capping layer 60 of a dielectric such as BPSG is deposited on the stack.
    For example, since the height of the gate region mesas (silicon nitride A, 40B) in the auxiliary portion 100B may be relatively lower, the aspect ratio of the gap of the auxiliary portion 100B between the N-MOSFET and the P-MOSFET being charged These are not particularly stringent compared to the description of the previous technical structure. Thus, filling of these gaps can be done at relatively lower process temperatures than the prior art structure. For a while, it may be desirable to slightly increase the density of the BPSG 60 by heating it at 600 degrees to allow the CMP rates for the BPSG layer 60 and the rest of the BPSG layer to facilitate planarization.
    Finally, as shown in FIG. 17, the BPSG layer 60 on the stack is preferably flattened by CMP to leave the flat layer 60A.
    The remaining process of supplying various contacts to array transistors and auxiliary transistors and conductive interconnects such as bit lines and word lines is done conventionally. Similarly, the capacitors used to supply the storage nodes of the memory cells may be disposed in a conventional manner, preferably at a suitable location on the silicon chip surface as a trench or a suitable multilayer stack. By appropriate modification of the process, trench capacitors may also be used.
    Various modifications in the process described as one example of the invention may be devised without departing from the spirit and scope of the invention. In particular, modifications may be made to the specific metal or dielectric used as described above, for example, in forming polysides and salicides. Similarly, other possible changes include relocation to replace gas phase diffusion instead of ion implantation in the order of steps made or in some doping steps. In particular, if there is a need to improve the quality of a DRAM transistor, the portions of the silicon oxide layer 12 formed on the single crystal silicon substrate 10 at the beginning of the process may be used in later processes, for example to form source and drain regions. Just before the ion implantation process, it can be removed and replaced with a new dielectric layer for use as the gate dielectric layer. Other regions of the substrate 10 could be formed on substrates of various thicknesses of the silicon oxide layer 12.
    The present invention uses P-MOSFETs in place of N-MOSFETs in the memory cell array portion of a chip, and forms an integrated circuit other than DRAM with a large area of one type of transistors surrounded by an area containing two types of transistors. May be used in any suitable modification for
    The present invention has the effect of improving the performance of the auxiliary circuit by providing a double work function gate to the C-MOSFET transistors of the auxiliary circuit.
    权利要求:
    Claims (9)
    [1" claim-type="Currently amended] A DRAM comprising a silicon chip in which a memory cell array each including an N-MOSFET is formed in a central portion thereof, and a subsidiary circuit including both the N-MOSFETs and the P-MOSFETs is formed in a peripheral portion thereof.
    The N-MOSFETs of the memory cell use N-doped polyside gates, the N-MOSFETs of the auxiliary circuit use N-doped polysilicon gates, and the P-MOSFETs of the auxiliary circuit are P-doped DRAM using polysilicon gates.
    [2" claim-type="Currently amended] 2. The DRAM of claim 1 wherein the transistors of the auxiliary circuit use salicide source and drain contacts.
    [3" claim-type="Currently amended] 3. The DRAM of claim 2 wherein the source and drain contacts are borderless.
    [4" claim-type="Currently amended] A method for forming a DRAM consisting of a silicon chip comprising a central portion comprising arrays of memory cells using N-MOSFETs and a peripheral portion comprising an auxiliary circuit using C-MOSFETs, the method comprising:
    Forming a masking layer of silicon oxide on the surface of the chip region and selectively removing the layer from the center where memory cell arrays are to be included, but leaving the layer properly positioned at the periphery where the auxiliary circuit should be included;
    Forming gate N-MOSFETs of the memory cells in the central portion and including gate conductors in the N-MOSFET including a lower polysilicon layer doped with a donor atom and an upper layer that is a metal silicide;
    Covering said chip region with a masking layer and selectively removing said masking layer from said center portion of said chip;
    Removing the silicon oxide layer from the periphery of the chip region;
    Covering the peripheral portion with a masking layer and removing the peripheral portion at the portion where N-MOSFETs should be formed;
    Forming gated N-MOSFETs of the auxiliary circuitry in the periphery and including gate conductors in the N-MOSFET including a lower polysilicon layer doped with donor atoms and an upper layer that is a metal silicide;
    Covering the periphery with a masking layer and removing the masking layer at a portion where a P-MOSFET is to be formed;
    Forming the P-MOSFET of the auxiliary circuit in the periphery and including the P-MOSFET in a gate conductor comprising a lower silicon layer doped with acceptor atoms and an upper layer that is a metal silicide. How to.
    [5" claim-type="Currently amended] 5. The method of claim 4, wherein an undoped polysilicon layer is patterned at the periphery of the chip to form a plurality of mesas on the silicon substrate, and the mesas that should be gate conductors of the N-MOSFETs are implanted with donor ions, Mesas that should be gate conductors of a P-MOSFET are implanted with acceptor ions, and a gate dielectric is included between the mesas and the silicon substrate.
    [6" claim-type="Currently amended] 6. The method of claim 5, wherein the source and drain regions of the transistors are formed on both sides of the mesas by the ion implantation.
    [7" claim-type="Currently amended] 7. The method of claim 6, wherein after ion implantation, the mesas and the source and drain regions are covered with a silicide layer, and after the chip is heated, salicide contacts are selectively formed in the source and drain regions. How to.
    [8" claim-type="Currently amended] In a method of manufacturing a dynamic random access memory,
    Forming a first oxide layer on the top surface of the silicon substrate layer to which both a suitable complementary N-MOSFET and P-MOSFET pair should be transformed into suitable N-MOSFETs and an auxiliary circuit with memory cell transistors;
    Stacking a second layer of undoped polysilicon over the first oxide;
    Stacking a third layer of silicon oxide over the undoped polysilicon second layer;
    Patterning the silicon oxide third layer to remove portions where array transistors should be formed;
    Depositing a fourth layer of polycide on the stack;
    Planarizing the stack structure to the level of the silicon oxide third layer;
    Forming a fifth layer of silicon nitride on the planarized stack;
    A fourth layer and a fourth layer to form gaps in the fourth and fifth layers to expose regions of the polysilicon second layer into which dopants are implanted into the silicon substrate to form source and drain regions in the silicon substrate; Patterning the fifth layers;
    implanting n-type ions into an exposed region of the undoped polysilicon second layer below the gaps of the fourth and fifth layers;
    Etching the exposed regions of the polysilicon second layer not covered by the fourth and fifth layers to remove the silicon oxide first layer, and forming gaps in the first layer;
    Implanting the stack with n type dopants to selectively form n type source / drain regions in the silicon substrate layer in the gaps not covered by the fourth and fifth layers;
    Forming dielectric spacer layers on sidewalls of the gaps formed in the fourth and fifth layers;
    Covering the stack with a glass layer and filling the gaps using thermal reflux;
    Patterning the glass to expose an auxiliary portion of the stack where the complementary circuits are to be formed and to expose regions of the silicon oxide third layer that are not covered by the sidewall space layer of the silicon nitride fifth layer and fifth layer. Doing;
    Patterning said silicon oxide third layer and said undoped polysilicon second layer of said auxiliary portion not masked by said remaining silicon nitride fifth layer;
    Etching the exposed portions of the remaining silicon nitride fifth layer, the lower silicon oxide third layer, and the undoped polysilicon second layer in the auxiliary portion;
    Forming dielectric sidewall spacers on the exposed and undoped polysilicon second layer;
    Covering the stack with a patterned masking layer to open a window over the regions where N-MOSFETs should be formed in the auxiliary portion;
    The previously of the silicon substrate and the auxiliary portions not covered by the previously undoped polysilicon second layer of the array to form the source / drain regions of the N-MOSFETs in the auxiliary portions. Implanting n-type dopant ions into the undoped polysilicon second layer;
    Covering the stack with a patterned masking layer to open a window on the region where P-MOSFETs are to be formed in the auxiliary portions;
    Implanting P-type ions into the exposed silicon substrate and previously undoped polysilicon second layer of the auxiliary region to form the source / drain regions of the P-MOSFETs in the auxiliary portions;
    Removing the masking layer from the stack structure;
    Depositing a metal layer on the stack structure suitable for forming self-aligned polysilicon contacts in the exposed source / drain regions and the exposed P and n type doped polysilicon second layers in the stack structure. ;
    Heating the stack structure to form a salicide contact in which the metal layer is in contact with silicon;
    Etching the stack to remove the metal layer that does not react to form the salicide contact;
    Depositing a capping layer on the stack structure; And
    Planarizing the capping layer.
    [9" claim-type="Currently amended] A method of forming a stack comprising a single crystal silicon p-type substrate in which N-MOSFETs serving as array transistors of a dynamic random access memory are formed in a central portion and C-MOSFETs serving as an auxiliary circuit are formed in the peripheral portion.
    Forming a dielectric first layer on an upper surface of the single crystal silicon substrate;
    Stacking an undoped polysilicon second layer on the dielectric first layer;
    Stacking a silicon oxide third layer on the undoped polysilicon second layer;
    Patterning the silicon oxide third layer to expose the polysilicon second layer portion to which the array transistors are to be modified;
    Stacking a silicide fourth layer on the exposed portion of the polysilicon second layer and the remaining portion of the silicon oxide third layer;
    Planarizing the stack to the level of the remaining portion of the silicon oxide third layer;
    Stacking a silicon nitride fifth layer on the planarized stack;
    The silicon nitride fifth layer to expose the polysilicon second layer portions of the array region of the stack structure and the silicon nitride fifth layer portions of the auxiliary circuit region of the stack structure to form the gaps in the stack Patterning the lower silicide fourth layer;
    Spinning onto the stack with donor ions to implant donor ions into the exposed portion of the undoped polysilicon second layer of the array region of the stack structure to diffuse throughout the polysilicon array region of the stack ;
    Etching the exposed portion of the polysilicon second layer to the gate dielectric first layer;
    Donor ions in the lower single crystal silicon substrate layer through the exposed portions of the gate dielectric first layer to form n type regions in the silicon p type substrate suitable for use as source and drain regions of the N-MOSFETs. Spinning back onto the stack with donor ions for implantation;
    Forming dielectric spacers on sidewalls of the gaps in the stack;
    Forming a masking layer on the stack and removing the masking layer from the periphery where the C-MOSFETs of the auxiliary circuit should be formed;
    The exposed silicon oxide third layer, the undoped polysilicon second layer, and the mesa regions of the undoped polysilicon second layer and the region of the lower gate dielectric first layer to expose at the periphery. Etching the exposed gate dielectric first layer in turn at the periphery;
    Forming dielectric spacer layers on the sidewalls of the mesa regions of undoped silicon;
    Spinning the stack to implant donor ions into the undoped polysilicon substrate where the exposed portions of the single crystal silicon substrate and the N-MOSFETs should be formed in the auxiliary circuit region of the stack structure;
    Covering portions of the stack where the N-MOSFETs are to be modified and exposing portions of the stack where the P-MOSFETs are to be modified; And
    Emitting P-MOSFETs into the stack with acceptor ions to implant acceptor ions into the exposed silicon single crystal substrate and the exposed and undoped polysilicon to be formed in the stack. How to feature.
    类似技术:
    公开号 | 公开日 | 专利标题
    US9514976B2|2016-12-06|Trench isolation implantation
    KR101989273B1|2019-06-13|Structure and method for single gate non-volatile memory device
    US8067286B2|2011-11-29|Methods of forming recessed access devices associated with semiconductor constructions
    KR100533511B1|2005-12-06|Modified gate processing for optimized definition of array and logic devices on same chip
    US6238967B1|2001-05-29|Method of forming embedded DRAM structure
    JP4860022B2|2012-01-25|Manufacturing method of semiconductor integrated circuit device
    JP3648376B2|2005-05-18|Manufacturing method of semiconductor device
    US5733809A|1998-03-31|Split-polysilicon CMOS process for multi-megabit dynamic memories incorporating stacked container capacitor cells
    US7087956B2|2006-08-08|Semiconductor device and it's manufacturing method
    DE10324491B4|2005-07-14|Manufacturing Method for Dual Workfunction Logic Devices in Vertical DRAM Processes
    US5821160A|1998-10-13|Method for forming a laser alterable fuse area of a memory cell using an etch stop layer
    KR100621714B1|2006-09-06|Trench capacitor with epi buried layer
    US7682896B2|2010-03-23|Trench metal-insulator-metal | capacitors integrated with middle-of-line metal contacts, and method of fabricating same
    US6498062B2|2002-12-24|DRAM access transistor
    KR100391985B1|2003-07-22|Method of making a scalable two transistor memory device
    US7425740B2|2008-09-16|Method and structure for a 1T-RAM bit cell and macro
    US6555450B2|2003-04-29|Contact forming method for semiconductor device
    US6767789B1|2004-07-27|Method for interconnection between transfer devices and storage capacitors in memory cells and device formed thereby
    US6448618B1|2002-09-10|Semiconductor device and method for manufacturing the same
    KR100598301B1|2006-07-07|A trench capacitor with isolation collar
    US6096595A|2000-08-01|Integration of a salicide process for MOS logic devices, and a self-aligned contact process for MOS memory devices
    US6190979B1|2001-02-20|Method for fabricating dual workfunction devices on a semiconductor substrate using counter-doping and gapfill
    US5607879A|1997-03-04|Method for forming buried plug contacts on semiconductor integrated circuits
    US6620674B1|2003-09-16|Semiconductor device with self-aligned contact and its manufacture
    US5780338A|1998-07-14|Method for manufacturing crown-shaped capacitors for dynamic random access memory integrated circuits
    同族专利:
    公开号 | 公开日
    TW448543B|2001-08-01|
    EP1039533A3|2001-04-04|
    CN1268772A|2000-10-04|
    KR100690121B1|2007-03-08|
    CN1156914C|2004-07-07|
    JP2000311991A|2000-11-07|
    US6235574B1|2001-05-22|
    EP1039533A2|2000-09-27|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    法律状态:
    1999-03-22|Priority to US27340299A
    1999-03-22|Priority to US09/273,402
    2000-03-22|Application filed by 인피니언 테크놀로지스 노쓰 아메리카 코포레이션
    2001-01-26|Publication of KR20010006849A
    2007-03-08|Application granted
    2007-03-08|Publication of KR100690121B1
    优先权:
    申请号 | 申请日 | 专利标题
    US27340299A| true| 1999-03-22|1999-03-22|
    US09/273,402|1999-03-22|
    [返回顶部]