This application is based upon and claims the benefit of priority from the Japanese Patent Application No. 2020-050773, filed Mar. 23, 2020, the entire contents of which are incorporated herein by reference.
Embodiments described herein relate generally to a semiconductor memory device.
A semiconductor memory device in which memory cells are three-dimensionally arranged is known.
In general, according to one embodiment, a semiconductor memory device comprises: a first semiconductor layer extending in a first direction; a second semiconductor layer stacked above the first semiconductor layer in a second direction intersecting the first direction, the second semiconductor layer extending in the first direction; and a first conductive layer intersecting the first semiconductor layer and the second semiconductor layer and extending in the second direction, wherein the first conductive layer includes a first portion intersecting the first semiconductor layer and a second portion intersecting the second semiconductor layer, a first width of the first portion in the first direction is smaller than a second width of the second portion in the first direction, and a first thickness of the first semiconductor layer in the second direction is larger than a second thickness of the second semiconductor layer in the second direction.
Hereinafter, embodiments will be described with reference to the accompanying drawings. In the description below, constituent elements having substantially the same function and configuration will be assigned a common reference symbol. Each of the embodiments described below merely indicates an exemplary apparatus or method for embodying the technical idea of the embodiment. The material, shape, structure, arrangement, etc. of the technical idea of each embodiment are not limited to the ones described below. The technical ideas of the embodiments may be variously modified within the range of the claims. Hereinafter, a three-dimensionally stacked type NAND flash memory, in which memory cell transistors are three-dimensionally stacked above a semiconductor substrate, will be described as an example of the semiconductor memory device of each embodiment.
A semiconductor memory device according to a first embodiment will be described.
1.1 Configuration of Semiconductor Memory Device
1.1.1 Circuit Configuration of Semiconductor Memory Device
First, a circuit configuration of the semiconductor memory device will be described with reference to
As shown in
The input/output circuit 10 controls input and output of a signal DQ to and from an external controller 2. The signal DQ includes, for example, data DAT, an address ADD, and a command CMD. Specifically, the input/output circuit 10 transmits the data DAT, address ADD, and command CMD received from the controller 2 to the data register 21, address register 13, and command register 14, respectively. The input/output circuit 10 also transmits status information STS received from the status register 12, the data DAT received from the data register 21, the address ADD received from the address register 13, and the like to the controller 2.
The logic controller 11 receives various control signals from the controller 2. The logic controller 11 controls the input/output circuit 10 and the sequencer 15 in accordance with the received control signals.
The status register 12 temporarily stores, for example, status information STS in a write operation, a read operation, or an erase operation, and notifies the controller 2 whether or not the operation has been normally completed.
The address register 13 temporarily stores the received address ADD. The address ADD includes a row address RADD and a column address CADD. The address register 13 transfers the row address RADD to the row decoder 19, and transfers the column address CADD to the column decoder 22.
The command register 14 temporarily stores the received command CMD, and transfers it to the sequencer 15.
The sequencer 15 controls the overall operation of the semiconductor memory device 1. Specifically, in accordance with the received command CMD, the sequencer 15 controls, for example, the status register 12, the ready/busy circuit 16, the voltage generator 17, the row decoder 19, the sense amplifier 20, the data register 21, the column decoder 22, and the like to execute a write operation, a read operation, an erase operation, and the like.
The ready/busy circuit 16 transmits a ready/busy signal RBn to the controller 2 in accordance with the operation status of the sequencer 15.
The voltage generator 17 generates various voltages used in a write operation, a read operation, and an erase operation under control of the sequencer 15, and supplies the generated voltages to, for example, the memory cell array 18, the row decoder 19, the sense amplifier 20, and the like. The row decoder 19 and the sense amplifier 20 apply the voltages supplied from the voltage generator 17 to memory cell transistors in the memory cell array 18.
The memory cell array 18 includes a plurality of blocks BLK0, BLK1, BLK2, . . . , and BLKm (where m is an integer larger than or equal to 1), each including a plurality of nonvolatile memory cell transistors (hereinafter also referred to as “memory cells”) each associated with a row and a column. Hereinafter, when the blocks BLK0 to BLKm are not distinguished from one another, each block will be referred to as a block BLK.
Each block BLK includes a plurality of memory units MU. Each memory unit MU includes a plurality of memory groups MG. The number of blocks BLK in the memory cell array 18, the number of memory units MU in each block BLK, and the number of memory groups MG in each memory unit MU may be any number. The memory cell array 18 will be described in detail later.
The row decoder 19 decodes the row address RADD. Based on a result of the decoding, the row decoder 19 applies a control voltage supplied from the voltage generator 17 to transistors in the memory cell array 18.
In a read operation, the sense amplifier 20 senses data read from the memory cell array 18. The sense amplifier 20 transmits the read data to the data register 21. In a write operation, the sense amplifier 20 transmits write data to the memory cell array 18.
The data register 21 includes a plurality of latch circuits. The latch circuits each temporarily store write data or read data.
In a write operation, a read operation, an erase operation, and the like, the column decoder 22 decodes the column address CADD, and selects a latch circuit in the data register 21 based on a result of the decoding.
1.1.2 Overall Configuration of Memory Cell Array
Next, an overall configuration of the memory cell array 18 will be described with reference to
As shown in
Each of the semiconductor layers 31 corresponds to one memory group MG to be described later, and functions as an active area in which channel layers of a plurality of memory cell transistors MC and select transistors ST1 and ST2 are formed. The semiconductor layers 31 extend in the X direction, which is parallel to the semiconductor substrate, and are stacked in the Z direction, which is perpendicular to the semiconductor substrate, apart from one another (with insulating layers (not shown) interposed therebetween). Sets of the semiconductor layers 31 stacked in the Z direction are aligned in the Y direction, which is parallel to the semiconductor substrate and is orthogonal to (or intersects) the X direction.
Between the sets of the semiconductor layers 31 aligned in the Y direction, a plurality of word line pillars WLP extending in the Z direction are aligned in the X direction. In other words, a set of the word line pillars WLP aligned in the X direction and a set of the semiconductor layers 31 stacked in the Z direction are alternately aligned in the Y direction. Word lines WL extending in the Y direction are provided above the word line pillars WLP. Each of the word line pillars WLP includes a contact plug CWL (hereinafter also referred to as an “interconnect CWL”) electrically coupled to a word line WL provided above the contact plug CWL, and a block insulating layer provided in the side surfaces of the word line pillar WLP. In the layer at the same level as each of the semiconductor layers 31 from the semiconductor substrate, a charge storage layer and a tunnel insulating layer are provided between each word line pillar WLP and the semiconductor layer 31.
One memory cell transistor MC is provided at the intersection of one word line pillar WLP and semiconductor layer 31. Accordingly, a plurality of memory cell transistors MC are coupled in the X direction via a semiconductor layer 31. In other words, the channels of the memory cell transistors MC are coupled in the X direction.
A contact plug CBL is provided in the vicinity of one X-directional end of each set of the semiconductor layers 31 stacked in the Z direction. The contact plug CBL passes through the semiconductor layers 31 stacked in the Z direction, and is coupled to those semiconductor layers 31. A plurality of contact plugs CBL are provided in correspondence with the sets of the semiconductor layers 31 aligned in the Y direction. A bit line BL extending in the X direction is provided above each contact plug CBL. The contact plugs CBL are coupled to different bit lines BL.
A contact plug CSL is provided in the vicinity of the other X-directional end of each set of the semiconductor layers 31 stacked in the Z direction. The contact plug CSL passes through the semiconductor layers 31 stacked in the Z direction, and is coupled to those semiconductor layers 31. A plurality of contact plugs CSL are provided in correspondence with the sets of the semiconductor layers 31 aligned in the Y direction. A source line SL extending in the Y direction is provided above the contact plugs CSL. The contact plugs CSL are coupled in common to the source line SL.
One X-directional end each of the ends of the semiconductor layers 31 aligned in the Y direction in each layer is in contact with one of different insulating layers which are in contact with a select gate line SGD extending in the Y direction. Accordingly, the semiconductor layers 31 are not electrically coupled to the select gate line SGD. Similarly, the other X-directional ends of the semiconductor layers 31 aligned in the Y direction in each layer are in contact with different insulating layers which are in contact with a select gate line SGS extending in the Y direction. Accordingly, the semiconductor layers 31 are not electrically coupled to the select gate line SGS. The select gate lines SGD and SGS are stacked in such a manner that one select gate line SGD and one select gate line SGS are arranged in the same layer as each of the semiconductor layers 31 stacked in the Z direction.
Below the select gate lines SGD, a plurality of global select gate lines GSGD are formed on the X-Y plane.
On each of the global select gate lines GSGD, a contact plug CSGD extending in the Z direction is provided. The contact plugs CSGD are aligned in the Y direction. Each of the contact plugs CSGD includes a coupling portion which is electrically coupled to one of the select gate lines SGD. Namely, each contact plug CSGD electrically couples a global select gate line GSGD and a select gate line SGD. In the example of
Below the select gate lines SGS, a plurality of global select gate lines GSGS (not shown) are formed on the X-Y plane.
On each of the global select gate lines GSGS, a contact plug CSGS extending in the Z direction is provided. The contact plugs CSGS are aligned in the Y direction. Each of the contact plugs CSGS includes a coupling portion which is electrically coupled to one of the select gate lines SGS. Namely, each contact plug CSGS electrically couples a global select gate line GSGS and a select gate line SGS.
A plurality of memory groups MG (semiconductor layers 31) aligned in the Y direction and corresponding to one select gate line SGD and one select gate line SGS are included in one memory unit MU. Furthermore, a plurality of memory units MU coupled in common to a word line pillar WLP are included in one block. BLK.
1.1.3 Circuit Configuration of Memory Cell Array
Next, a circuit configuration of the memory cell array 18 will be described with reference to
As shown in
The memory string MSa includes, for example, four memory cell transistors MCa0 to MCa3. Similarly, the memory string MSb includes, for example, four memory cell transistors MCb0 to MCb3. Hereinafter, the memory cell transistors MCa0 to MCa3 and MCb0 to MCb3 will each be referred to as a memory cell transistor MC when they are not distinguished from one another.
Each memory cell transistor MC includes a control gate and a charge storage layer, and nonvolatilely stores data. The memory cell transistor MC may be of a floating gate (FG) type which uses a conductive layer as the charge storage layer, or of a metal-oxide-nitride-oxide-silicon (MONOS) type which uses an insulating layer as the charge storage layer. The number of memory cell transistors MC included in each memory string MS may be 8, 16, 32, 48, 64, 96, 128, etc., and is not limited to a particular number.
The current paths of the memory cell transistors MCa0 to MCa3 included in the memory string MSa are coupled in series. Similarly, the current paths of the memory cell transistors MCb0 to MCb3 included in the memory string MSb are coupled in series. The drains of the memory cell transistors MCa0 and MCb0 are coupled in common to the source of the select transistor ST1. The sources of the memory cell transistors MCa3 and MCb3 are coupled in common to the drain of the select transistor ST2. The numbers of select transistors ST1 and ST2 included in each memory group MG may be any number larger than or equal to one.
The gates of the memory cell transistors MC of a plurality of memory groups MG aligned in the Z direction are coupled in common to one word line WL. Specifically, the gates of the memory cell transistors MCa0 aligned in the Z direction are coupled in common to a word line WLa0. Similarly, the gates of the memory cell transistors MCa1, MCa2 and MCa3 are respectively coupled to word lines WLa1, WLa2 and WLa3. The gates of the memory cell transistors MCb0 to MCb3 are respectively coupled to word lines WLb0 to WLb3.
The drains of the select transistors ST1 of the memory groups MG aligned in the Z direction are coupled in common to one bit line BL. The gates of the select transistors ST1 of the memory groups MG aligned in the Z direction are coupled to different select gate lines SGD. Specifically, the gate of the select transistor ST1 corresponding to the memory group MG arranged in the lowermost layer is coupled to the select gate line SGD1. The gate of the select transistor ST1 corresponding to the memory group MG arranged in the uppermost layer is coupled to the select gate line SGDn.
The sources of the select transistors ST2 of the memory groups MG aligned in the Z direction are coupled in common to one source line SL. The gates of the select transistors ST2 of the memory groups MG aligned in the Z direction are coupled to different select gate lines SGS. Specifically, the gate of the select transistor ST2 corresponding to the memory group MG arranged in the lowermost layer is coupled to the select gate line SGS1, and the gate of the select transistor ST2 corresponding to the memory group MG arranged in the uppermost layer is coupled to the select gate line SGSn.
Next, couplings between the select gate lines SGD, the contact plugs CSGD, and the global select gate lines GSGD will be described with reference to
As shown in
Next, couplings between the select gate lines SGS, the contact plugs CSGS, and the global select gate lines GSGS will be described with reference to
As shown in
1.1.4 Layout Configuration of Memory Cell Array
Next, an example of the layout configuration of the memory cell array 18 will be described with reference to
As shown in
The memory cell region is a region in which the memory cell transistors MC are provided, i.e., a region in which the word line pillars WLP are arranged. The drain-side ST region is adjacent to the memory cell region in the X direction, and is a region in which the semiconductor layers 31 are coupled to the contact plugs CBL and in which the select transistors ST1 are provided. The source-side ST region is adjacent to the memory cell region in the X direction, and is a region in which the semiconductor layers 31 are coupled to the contact plugs CSL and in which the select transistors ST2 are provided.
The select gate contact region corresponding to the select gate lines SGD is adjacent to the drain-side ST region in the X direction, and is a region in which the contact plugs CSGD are coupled to the select gate lines SGD. The select gate contact region corresponding to the select gate lines SGS is adjacent to the source-side ST region in the X direction, and is a region in which the contact plugs CSGS are coupled to the select gate lines SGS.
The number of memory cell regions, drain-side ST regions, source-side ST regions, select gate contact regions corresponding to the select gate lines SGD, and select gate contact regions corresponding to the select gate lines SGS included in the memory cell array 18 may be any number.
In the memory cell region, a plurality of (12 in the example of
1.1.4.1 Layout Configuration of Memory Cell Region
A layout configuration of the memory cell region in the memory cell array 18 will be described with reference to
As shown in
On each of the Y-directional side surfaces of each semiconductor layer 31, an insulating layer 32 is provided. The insulating layer 32 functions as an etching stopper when a tunnel insulating layer 36 and charge storage layer 35 (to be described later) are formed. The insulating layer 32 contains, for example, silicon nitride.
A memory trench MT is provided between two semiconductor layers 31 aligned in the Y direction. The memory trench MT is a trench provided to divide the semiconductor layers 31 in the Y direction, and an insulating layer (such as silicon oxide) 55 is embedded in the memory trench MT. In the memory trench MT, a plurality of word line pillars WLP are aligned in the X direction at predetermined intervals to divide the insulating layer 55 in the X direction.
Each word line pillar WLP includes a conductive layer 33 and an insulating layer 34 extending in the Z direction. The conductive layer 33 is electrically coupled to a word line WL provided thereabove, and functions as the control gate (or word line) of the memory cell transistor MC. A conductive material is used for the conductive layer 33. The conductive material may be a metal material, a metal compound, or a semiconductor material doped with impurities. The conductive layer 33 contains, for example, tungsten (W) and titanium nitride (TiN). TiN is used as a barrier metal when W is formed by chemical vapor deposition (CVD).
The insulating layer 34 is provided on the side surfaces of the conductive layer 33 so as to surround the periphery of the conductive layer 33 in the X and Y directions. The insulating layer 34 functions as a block insulating layer that prevents the charge stored in the charge storage layer 35 from diffusing into the conductive layer 33 (word line WL). An insulating material is used for the insulating layer 34. The insulating material may be, for example, a stack structure of Hf(Si)Ox/SiO2/Hf(Si)Ox using hafnium (Hf) and silicon oxide (SiO2), a single layer of Hf(Si)Ox or SiO2, or a single layer of zirconium oxide (Zro), aluminum oxide (AlO), or the like. Hf (Si)Ox may be HfOx or HfSiOx.
The charge storage layer 35 and tunnel insulating layer 36 are provided between each word line pillar WLP and corresponding semiconductor layer 31 in order from the word line pillar WLP side. Specifically, the charge storage layer 35 is arranged on each Y-directional side surface of the insulating layer 34, and the tunnel insulating layer 36 is arranged on the side surface of the charge storage layer 35.
The charge storage layer 35 has a function of storing a charge injected from the semiconductor layer 31 in the memory cell transistor MC. As mentioned above, the charge storage layer 35 may be of, for example, the FG type, which uses a conductive layer, or the MONOS type, which uses an insulating layer. Charge storage layers 35 of both the FG type, which uses a conductive layer, and the MONOS type, which uses an insulating layer, may be used. Hereinafter, the case where the charge storage layers 35 are of the FG type will be described.
The conductive layer used for each charge storage layer 35 contains, for example, polysilicon. The conductive layer of the charge storage layer 35 may contain a metal such as tantalum nitride (TaN), titanium nitride (TiN), tungsten (W), or ruthenium (Ru). The tunnel insulating layer 36 contains, for example, silicon oxide or silicon oxynitride (SiON).
1.1.5 Cross-Sectional Structures of Memory Cell Region
Next, cross-sectional structures of the memory cell region in the memory cell array 18 will be described with reference to
As shown in
The memory cell array 18 is provided on the insulating layer 51. Specifically, an insulating layer 52 is provided on the insulating layer 51. The insulating layer 52 functions as an etching stopper when the memory trenches MT and holes used for various contact plugs and the like are processed. The insulating layer 52 may be any insulating material that can achieve a sufficient etching selectivity with respect to insulating layers 53 provided in upper layers. The insulating layer 52 contains, for example, silicon nitride (SiN) or a metal oxide such as aluminum oxide (AlO).
An insulating layer 53 is provided on the insulating layer 52. The insulating layer 53 contains, for example, silicon oxide.
On the insulating layer 53, for example seven semiconductor layers 31 are stacked with insulating layers 53 interposed therebetween. Namely, seven insulating layers 53 and seven semiconductor layers 31 are alternately stacked on the insulating layer 52. The number of stacked semiconductor layers 31 may be any number. An insulating layer 54 is provided on the uppermost semiconductor layer 31. The insulating layer 54 contains, for example, silicon oxide.
Each word line pillar WLP extends in the Z direction to intersect the insulating layer 54, the alternately-stacked seven semiconductor layers 31 and seven insulating layers 53. The word line pillar WLP passes from above the insulating layer 54 in the Z direction between a stack of the insulating layer 54, semiconductor layers 31 and insulating layers 53, and another stack of the insulating layer 54, semiconductor layers 31 and insulating layers 53 adjacent to the aforementioned stack in the Y direction, to reach the insulating layer 52.
As described above, the word line pillar WLP includes the conductive layer 33 and insulating layer 34 extending in the Z direction. Specifically, a hole AH is formed in the Z direction to intersect the insulating layer 54, semiconductor layers 31, and insulating layers 53. An insulating layer 34 is formed on the side and bottom surfaces of the hole AH. On the side surfaces of the insulating layer 34, the conductive layer 33 is formed to fill the inside of the hole AH.
The charge storage layer 35 and tunnel insulating layer 36 are provided between the insulating layer 34 and each semiconductor layer 31 in order from the insulating layer 34 side. Specifically, the charge storage layer 35 is provided between the insulating layer 34 and each semiconductor layer 31 so as to be in contact with the side surface of the insulating layer 34. Furthermore, the tunnel insulating layer 36 is provided between the charge storage layer 35 and the semiconductor layer 31.
The memory trench MT passes from above the insulating layer 54 in the Z direction between a stack of the insulating layer 54, semiconductor layers 31, and insulating layers 53 and another stack of the insulating layer 54, semiconductor layers 31, and insulating layers 53 adjacent to the aforementioned stack in the Y direction, to reach the insulating layer 52. The insulating layer 55 is provided in the memory trench MT. The insulating layer 32 is further provided between the insulating layer 55 and each semiconductor layer 31.
As shown in
As shown in
In contrast to the word line lengths Lg, the X-directional distance between adjacent word line pillars WLP gradually decreases from between word line pillars WLP corresponding to the lowermost semiconductor layer 31 to between word line pillars WLP corresponding to the uppermost semiconductor layer 31. Namely, in the cross-sectional structure shown in
As shown in
The Y-directional width of the conductive layer 33 in the word line pillar WLP (hereinafter referred to as a “word line width”) is approximately the same from the portion of the conductive layer 33 corresponding to the lowermost semiconductor layer 31 to the portion of the conductive layer 33 corresponding to the uppermost semiconductor layer 31. Namely, in the cross-sectional structure shown in
As described above, the memory cell region of the first embodiment has a configuration in which the thickness h1 of a semiconductor layer 31 on the lower layer side is larger than the thickness hn of a semiconductor layer 31 on the upper layer side (h1>hm>hn), the X-directional width (word line length) Lg1 of the conductive layer 33 on the lower layer side is smaller than the X-directional width Lgn of the conductive layer 33 on the upper layer side (Lg1<Lgm<Lgn), and the distance Ld1 between word line pillars WLP on the lower layer side is larger than the distance Ldn between word line pillars WLP on the upper layer side (Ld1>Ldm>Ldn)
1.2 Method for Manufacturing Semiconductor Memory Device
Next, a method for manufacturing the memory cell region in the semiconductor memory device of the first embodiment will be described with reference to
First, as shown in
Next, as shown in
Next, as shown in
Specifically, the semiconductor layers 31 are each partly etched from the side surface of the open memory trench MT by, for example, wet etching, thereby forming a recess region. Then, the insulating layer 32 is formed on the side and bottom surfaces of the memory trench MT, which include the recess regions, and on the insulating layer 54. The extra insulating layer 32 on the side and bottom surfaces of the memory trench MT and on the insulating layer 54 is removed by etching back, while leaving the insulating layer 32 in the recess regions. Next, the insulating layer 55 is embedded in the memory trench MT by, for example, CVD.
As shown in
As shown in
Specifically, the insulating layer 32 around the hole AH is removed by, for example, wet etching. Thereafter, the side surfaces of each semiconductor layer 31 exposed to the hole AH are oxidized by oxidation treatment, thereby forming the tunnel insulating layer 36. The charge storage layer 35 is then formed on the side and bottom surfaces of the hole AH and on the insulating layer 54. The extra charge storage layer 35 on the side and bottom surfaces of the hole AH and on the insulating layer 54 is removed by etching back, while leaving the charge storage layer 35 on the side surface of the tunnel insulating layer 36.
As shown in
Specifically, the insulating layer 34 is formed on the bottom and side surfaces of the hole AH by, for example, CVD. Next, for example, TiN and W as the conductive layer 33 are sequentially formed by CVD. Thereafter, TiN and W on the insulating layer 54 are removed by, for example, chemical mechanical polishing (CMP). The memory cell region shown in
1.3 Advantages of First Embodiment
The semiconductor memory device according to the first embodiment can improve reliability of the read operation. Advantages of the first embodiment will be described below.
A structure in which a plurality of semiconductor layers 31 extending in the X direction are stacked above the semiconductor substrate 50, and word line pillars WLP extending in the Z direction to intersect these semiconductor layers 31 are provided may be formed by embedding a material that forms the word line pillars WLP into holes AH. Due to the processing condition of the holes AH, the X-directional width (word line length) Lg of the conductive layer 33 in the word line pillar WLP tends to be smaller on the lower layer side close to the semiconductor substrate 50 than on the upper layer side far from the semiconductor substrate 50, and the X-directional distance Ld between word line pillars WLP tends to be larger on the lower layer side than on the upper layer side. In other words, the gate length Lg of a memory cell transistor MC tends to be smaller on the lower layer side close to the semiconductor substrate 50 than on the upper layer side far from the semiconductor substrate 50, and the distance Ld between gates of memory cell transistors MC tends to be larger on the lower layer side than on the upper layer side. Accordingly, the gate length Lg of a memory cell transistor MC runs counter to the distance Ld between gates of memory cell transistors MC and, as the gate length Lg of a memory cell transistor MC on the lower layer side decreases, the distance Ld between gates of memory cell transistors MC increases, and the resistance between gates increases. As a result, there arises a concern about a decrease in the drain current that flows through a memory cell transistor MC on the lower layer side when a word line voltage (control gate voltage) is applied thereto. The drain current corresponds to, for example, a current that flows through a memory cell transistor MC in a read operation (hereinafter referred to as a cell current). Therefore, reliability of the read operation in the semiconductor memory device may deteriorate.
The semiconductor memory device of the first embodiment has a configuration in which: the thickness of a first semiconductor layer 31 on the lower layer side is larger than the thickness of a second semiconductor layer 31 on the upper layer side, i.e., h1>hm>hn is satisfied; and the word line pillar WLP has a first portion intersecting the first semiconductor layer 31 on the lower layer side and a second portion intersecting the second semiconductor layer 31 on the upper layer side, and an X-directional first width Lg1 of the first portion is smaller than an X-directional second width Lgn of the second portion, i.e., Lg1<Lgm<Lgn is satisfied.
The semiconductor memory device also has a configuration in which: the thickness of the first semiconductor layer 31 on the lower layer side is larger than the thickness of the second semiconductor layer 31 on the upper layer side, i.e., h1>hm>hn is satisfied; and a first distance Ld1 between a first conductive layer 33 and a second conductive layer 33 adjacent to each other in the X direction at a portion (level) where each word line pillar WLP intersects the first semiconductor layer 31 on the lower layer side is larger than a second distance Ldn between the first conductive layer 33 and the second conductive layer 33 at a portion (level) where each word line pillar WLP intersects the second semiconductor layer 31 on the upper layer side, i.e., Ld1>Ldm>Ldn.
The above configurations can make the Z-directional length of the channel layer formed in the semiconductor layer 31 on the lower layer side larger than that in the semiconductor layer 31 on the upper layer side, and the thickness of the semiconductor layer 31 between gates of memory cell transistors MC on the lower layer side larger than that between gates of memory cell transistors MC on the upper layer side. Accordingly, the resistance between gates of memory cell transistors MC on the lower layer side can be lowered, and a decrease in the cell current in each memory cell transistor MC on the lower layer side can be suppressed. As a result, reliability of the read operation can be improved. In a write operation and erase operation, a write verification operation and erase verification operation, which correspond to the read operation, are respectively performed. Therefore, in the first embodiment, reliability of the write operation and erase operation can also be improved.
As shown in
Next, a semiconductor memory device according to a second embodiment will be described. Described in the second embodiment is a case where the Y-directional dimension of each word line pillar WLP is small in a lower layer portion, and large in an upper layer portion. The other circuit configuration, overall configuration, layout configuration, and cross-sectional structure are the same as those of the first embodiment. In the second embodiment, points different from the first embodiment will be mainly described.
2.1 Cross-Sectional Structure of Memory Cell Region
A cross-sectional structure of the memory cell region in the memory cell array 18 will be described with reference to
As shown in
As in the first embodiment shown in
The relationship between the distance Ld1 between word line pillars WLP corresponding to the lowermost semiconductor layer 31, the distance Ldm between word line pillars WLP corresponding to the middle semiconductor layer 31, and the distance Ldn between word line pillars WLP corresponding to the uppermost semiconductor layer 31 is expressed as Ld1>Ldm>Ldn.
As shown in
The relationship between the Y-directional width (word line width) tgate1 of the conductive layer 33 corresponding to the lowermost semiconductor layer 31, the word line width tgatem of the conductive layer 33 corresponding to the middle semiconductor layer 31, and the word line width tgaten of the conductive layer 33 corresponding to the uppermost semiconductor layer 31 is expressed as tgate1<tgatem<tgaten.
As described above, the memory cell region of the second embodiment has a configuration in which: the thickness h1 of a semiconductor layer 31 on the lower layer side is larger than the thickness hn of a semiconductor layer 31 on the upper layer side (h1>hm>hn), the X-directional width (word line length) Lg1 of the conductive layer 33 on the lower layer side is smaller than the X-directional width Lgn of the conductive layer 33 on the upper layer side (Lg1<Lgm<Lgn), the distance Ld1 between word line pillars WLP on the lower layer side is larger than the distance Ldn between word line pillars WLP on the upper layer side (Ld1>Ldm>Ldn), and the Y-directional length tpoly1 of a semiconductor layer 31 on the lower layer side is larger than the Y-directional length tpolyn of a semiconductor layer 31 on the upper layer side (tpoly1>tpolym>tpolyn).
2.2 Method for Manufacturing Semiconductor Memory Device
Next, a method for manufacturing the memory cell region in the semiconductor memory device of the second embodiment will be described with reference to
As shown in
Next, as shown in
The subsequent manufacturing steps shown in
Specifically, as shown in
As shown in
As shown in
2.3 Advantages of Second Embodiment
The semiconductor memory device of the second embodiment can improve reliability of the read operation as in the first embodiment. Advantages of the second embodiment will be described below.
In the structure in which a plurality of semiconductor layers 31 extending in the X direction and word line pillars WLP extending in the Z direction to intersect the semiconductor layers 31 are provided above a semiconductor substrate 50, as in the first embodiment, the X-directional width (word line length) Lg of the conductive layer 33 in each word line pillar WLP tends to be smaller on the lower layer side than on the upper layer side, and the X-directional distance Ld between word line pillars WLP tends to be larger on the lower layer side than on the upper layer side, depending on the processing condition of the holes AH. In addition, when memory trenches MT are formed in a structure in which a plurality of semiconductor layers 31 are stacked with insulating layers 53 interposed therebetween, the Y-directional lengths tpoly of the semiconductor layers 31 may be larger on the lower layer side than on the upper layer side, depending on the processing condition. As the Y-directional length tpoly of a semiconductor layer 31 increases, the distance between word line pillars WLP on the both sides of the semiconductor layer 31 in the Y direction increases, and the back-gate effect from the word line pillar WLP to which the read voltage is applied and the word line pillar WLP arranged on the other side of the semiconductor layer 31 decreases. Therefore, there arises a concern about a change in the threshold voltages of a memory cell transistor MC, and a decrease in the cell current that flows through a memory cell transistor MC on the lower layer side when a word line voltage (control gate voltage) is applied thereto. This may cause deterioration in reliability of the read operation in the semiconductor memory device.
In the above-described first embodiment, the decrease in the cell current is suppressed by adjusting the thickness of each semiconductor layer 31 in accordance with the X-directional width (gate length of each memory cell transistor MC) Lg of the conductive layer 33. In the second embodiment, the thickness of each semiconductor layer 31 is adjusted in consideration of the Y-directional length tpoly of the semiconductor layer 31 or the Y-directional width tgate of the conductive layer 33 which runs counter to the length tpoly, in addition to the X-directional width Lg of the conductive layer 33. Namely, the decrease in the cell current is suppressed by adjusting the thickness of each semiconductor layer 31 in accordance with the X-directional width Lg of the conductive layer 33 and the Y-directional length tpoly of the semiconductor layer 31. Specifically, as Lg/tpoly decreases, i.e., as Lg×tgate decreases, the thickness of the semiconductor layer 31 is increased, which can increase the Z-directional length of the channel layer formed in the semiconductor layer 31, and can lower the resistance between gates of memory cell transistors MC. Accordingly, a decrease in the cell current in each memory cell transistor MC on the lower layer side can be suppressed.
The semiconductor memory device of the second embodiment has a configuration in which: the thickness h1 of a first semiconductor layer 31 on the lower layer side is larger than the thickness hn of a second semiconductor layer 31 on the upper layer side, i.e., h1>hn is satisfied; and each word line pillar WLP has a first portion intersecting the first semiconductor layer 31 on the lower layer side and a second portion intersecting the second semiconductor layer 31 on the upper layer side, and Lg1/tpoly1<Lgn/tpolyn is satisfied, where an X-directional first width of the first portion is Lg1(L1), an X-directional second width of the second portion is Lgn(L2), the Y-directional length of the first semiconductor layer 31 on the lower layer side is tpoly1(P1), and the Y-directional length of the second semiconductor layer 31 on the upper layer side is tpolyn(P2). In addition, the semiconductor memory device of the second embodiment has a configuration in which Lg1×tgate1<Lgn×tgaten is satisfied, where the Y-directional width of the conductive layer 33 in the first portion is tgate1 and the Y-directional width of the conductive layer 33 in the second portion is tgaten.
The above configurations can make the Z-directional length of the channel layer formed in the semiconductor layer 31 on the lower layer side larger than that in the semiconductor layer 31 on the upper layer side, and can lower the resistance between gates of memory cell transistors MC on the lower layer side. Accordingly, a decrease in the cell current in each memory cell transistor MC on the lower layer side can be suppressed, and reliability of the read operation can be improved. Furthermore, reliability of the write operation and erase operation can be improved.
Next, a semiconductor memory device according to a third embodiment will be described. Described in the third embodiment is a case where a plurality of semiconductor layers 31 stacked above the semiconductor substrate 50 constitute a first group on the lower layer side and a second group on the upper layer side, and the thickness of the semiconductor layers 31 in the first group is larger than the thickness of the semiconductor layers 31 in the second group. The other circuit configuration, overall configuration, layout configuration, and cross-sectional structure are the same as those of the first embodiment. In the third embodiment, points different from the first embodiment will be mainly described.
3.1 Cross-Sectional Structure of Memory Cell Region
A cross-sectional structure of the memory cell region in the memory cell array 18 will be described with reference to
In the third embodiment, as shown in
As shown in
As described above, the memory cell region of the third embodiment has a configuration in which: the thickness h1 of the semiconductor layers 31 in the first group G1 on the lower layer side is larger than the thickness h2 of the semiconductor layers 31 in the second group G2 on the upper layer side (h1>h2), the X-directional width (word line length) Lg1 of the conductive layer 33 on the lower layer side is smaller than the X-directional width Lgn of the conductive layer 33 on the upper layer side (Lg1<Lgm<Lgn), the distance Ld1 between word line pillars WLP on the lower layer side is larger than the distance Ldn between word line pillars WLP on the upper layer side (Ld1>Ldm>Ldn), and the Y-directional length tpoly1 of a semiconductor layer 31 on the lower layer side is larger than the Y-directional length tpolyn of a semiconductor layer 31 on the upper layer side (tpoly1>tpolym>tpolyn).
3.2 Method for Manufacturing Semiconductor Memory Device
Next, a method for manufacturing the memory cell region in the semiconductor memory device of the third embodiment will be described.
First, as shown in
The subsequent manufacturing steps are the same as those shown in
3.3 Advantages of Third Embodiment
The semiconductor memory device according to the third embodiment can improve reliability of the read operation. Advantages of the third embodiment will be described below.
In the third embodiment, the thickness h1 of the semiconductor layers 31 in the first group G1 on the lower layer side is larger than the thickness h2 of the semiconductor layers 31 in the second group G2 on the upper layer side, i.e., h1>h2 is satisfied. This configuration can make the Z-directional length of the channel layer formed in each semiconductor layer 31 in the first group G1 on the lower layer side larger than that in each semiconductor layer 31 in the second group G2 on the upper layer side, and can make the resistance between gates of memory cell transistors MC in the first group G1 on the lower layer side lower than that between gates of memory cell transistors MC in the second group G2 on the upper layer side. Accordingly, a decrease in the cell current in each memory cell transistor MC in the first group G1 on the lower layer side can be suppressed, and reliability of the read operation can be improved. Furthermore, reliability of the write operation and erase operation can be improved.
Described in the third embodiment is the case where there are two groups, i.e., the first group G1 and the second group G2; however, a plurality of semiconductor layers 31 may be grouped into any number of groups.
Next, a semiconductor memory device according to a fourth embodiment will be described. Described in the first to third embodiments are cases where the dimensions of each word line pillar WLP gradually decrease from the upper layer portion to the lower layer portion. Described in the fourth embodiment is a case where each word line pillar WLP is the thickest in a portion lower than the portion corresponding to the uppermost semiconductor layer 31 and gradually becomes thinner from this portion to the lower layer portion. In the fourth embodiment, points different from the second embodiment will be mainly described.
4.1 Cross-Sectional Structure of Memory Cell Region
A cross-sectional structure of the memory cell region in the memory cell array 18 will be described with reference to
As shown in
As shown in
The X-directional width Lg of the conductive layer 33A gradually increases from the lower layer portion with Lg1 through the middle portion with Lgm to the largest portion with Lgmx, and gradually decreases from the largest portion to the upper layer portion with Lgn. Namely, the relationship between the X-directional widths Lg1, Lgm, Lgmx, and Lgn of the conductive layer 33A is expressed as Lg1<Lgm<Lgmx and Lgmx>Lgn.
In addition, Ld1 is assigned to the distance between word line pillars WLP adjacent to each other in the X direction corresponding to the lower layer portion, Ldm is assigned to the distance between word line pillars WLP adjacent to each other in the X direction corresponding to the middle portion, Ldmn is assigned to the distance between word line pillars WLP adjacent to each other in the X direction corresponding to the largest portion, and Ldn is assigned to the distance between word line pillars WLP adjacent to each other in the X direction corresponding to the upper layer portion in the cross-sectional structure shown in
The X-directional distance Ld between word line pillars WLP gradually decreases from the distance Ld1 corresponding to the lower layer portion through the distance Ldm corresponding to the middle portion to the distance Ldmn corresponding to the largest portion, and gradually increases from the distance Ldmn to the distance Ldn corresponding to the upper layer portion. Namely, the relationship between the X-directional distances Ld1, Ldm, Ldmn, and Ldn between word line pillars WLP is expressed as Ld1>Ldm>Ldmn and Ldmn<Ldn.
Moreover, as shown in
The Y-directional lengths tpoly of the semiconductor layers 31 gradually decrease from the length tpoly1 corresponding to the lower layer portion through the length tpolym corresponding to the middle portion to the length tpolymn corresponding to the largest portion, and gradually increase from the length tpolymn to the length tpolyn corresponding to the upper layer portion. The relationship between the Y-directional lengths tpoly1, tpolym, tpolymn, and tpolyn of the semiconductor layers 31 is expressed as tpoly1>tpolym>tpolymn and tpolymn<tpolyn.
As shown in
The Y-directional width tgate of the conductive layer 33A gradually increases from the lower layer portion with tgate1 through the middle portion with tgatem to the largest portion with tgatemx, and gradually decreases from the largest portion to the upper layer portion with tgaten. Namely, the relationship between the Y-directional widths tgate1, tgatem, tgatemx, and tgaten of the conductive layer 33A is expressed as tgate1<tgatem<tgatemx and tgatemx>tgaten.
Described herein is the case where the conductive layer 33A has the largest X-directional width Lgmx and the largest Y-directional width tgatemx at approximately the same level from the semiconductor substrate 50; however, the levels of the largest width portions may not be the same.
4.2 Method for Manufacturing Semiconductor Memory Device
Next, a method for manufacturing the memory cell region in the semiconductor memory device of the fourth embodiment will be described.
First, as shown in
Next, as shown in
Next, as shown in
Next, as shown in
Shown in
The subsequent manufacturing steps shown in
Specifically, as shown in
4.3 Advantages of Fourth Embodiment
The semiconductor memory device according to the fourth embodiment can improve reliability of the read operation. Advantages of the fourth embodiment will be described below.
In the fourth embodiment, as shown in
In the semiconductor memory device of the fourth embodiment, the thickness hmn of the semiconductor layer 31 corresponding to the largest portion is smaller than the thicknesses hm and h1 of the semiconductor layers 31 corresponding to the middle portion and the lower layer portion, and smaller than the thickness hn of the semiconductor layer 31 corresponding to the upper layer portion, i.e., hmn<hm<h1 and hmn<hn are satisfied. In addition, in the conductive layer 33A, the X-directional width Lgmx of the largest portion is larger than the X-directional widths Lgm and Lg1 of the middle portion and the lower layer portion, and is larger than the X-directional width Lgn of the upper layer portion, i.e., Lgmx>Lgm>Lg1 and Lgmx>Lgn are satisfied. In a dimensionally opposite manner, the distance Ldmn between word pillars WLP corresponding to the largest portion is smaller than the distances Ldm and Ld1 between word line pillars WLP corresponding to the middle portion and the lower layer portion, and is smaller than the distance Ldn between word line pillars WLP corresponding to the upper layer portion, i.e., Ldmn<Ldm<Ld1 and Ldmn<Ldn are satisfied. In addition, where the X-directional width of the conductive layer 33A is Lg, and the Y-directional length of each semiconductor layer 31 is tpoly, Lgmx/tpolymn of the largest portion is larger than Lgm/tpolym of the middle portion and Lg1/tpoly1 of the lower layer portion, and is larger than Lgn/tpolyn of the upper layer portion, i.e., Lgmx/tpolymn>Lgm/tpolym>Lg1/tpoly1, and Lgmx/tpolymn>Lgn/tpolyn are satisfied.
In the fourth embodiment, when the X-directional widths of the conductive layer 33A have the relationship Lgmx>Lgm>Lg1 and Lgmx>Lgn and the distances between word line pillars WLP adjacent to each other in the X direction has the relationship Ldmn<Ldm<Ld1 and Ldmn<Ldn, the thicknesses of the semiconductor layers 31 have the relationship hmn<hm<h1 and hmn<hn. Alternatively, when the X-directional widths Lg of the conductive layer 33A and the Y-directional lengths tpoly of the semiconductor layers 31 have the relationship Lgmx/tpolymn>Lgm/tpolym>Lg1/tpoly1 and Lgmx/tpolymn>Lgn/tpolyn, the thicknesses of the semiconductor layers 31 have the relationship hmn<hm<h1 and hmn<hn.
For example, when the X-directional width Lgmx of the largest portion of the conductive layer 33A is larger than the X-directional width Lgm of the middle portion, and is larger than the X-directional width Lgn of the upper portion, or the distance Ldmn between word line pillars WLP corresponding to the largest portion is smaller than the distance Ldm between word line pillars WLP corresponding to the middle portion, and is smaller than the distance Ldn between word line pillars WLP corresponding to the upper layer portion, even if the Y-directional length tpolymn of the semiconductor layer 31 corresponding to the largest portion is smaller than the Y-directional length tpolym of the semiconductor layer 31 corresponding to the middle portion, and is smaller than the Y-directional length tpolyn of the semiconductor layer 31 corresponding to the upper layer portion, the thickness hm of the semiconductor layer 31 corresponding to the middle portion and the thickness hn of the semiconductor layer 31 corresponding to the upper layer portion are larger than the thickness hmn of the semiconductor layer 31 corresponding to the largest portion.
The above configuration can make the Z-directional lengths of the channel layers formed in the semiconductor layers 31 corresponding to the middle portion and the upper layer portion larger than that in the semiconductor layer 31 corresponding to the largest portion, and can make the resistances between gates of memory cell transistors MC corresponding to the middle portion and the upper layer portion lower than that between gates of memory cell transistors MC corresponding to the largest portion. Accordingly, a decrease in the cell currents not only in the memory cell transistors MC on the lower layer side, but also in the memory cell transistors MC corresponding to the middle portion and the upper portion, can be suppressed, and reliability of the read operation can be improved. Furthermore, reliability of the write operation and erase operation can be improved.
The structure of the semiconductor memory device of the fourth embodiment is not limited to the structure in which the dimension of the word line pillar gradually changes in the Z direction depending on the processing condition. For example, the present embodiment can be applied to a structure in which a word line pillar is thicker in a middle portion between an upper layer portion and a lower layer portion than in the upper layer portion and the lower layer portion in a semiconductor memory device in which after a first word line pillar is formed, a second word line pillar is further formed above the first word line pillar.
In the third embodiment, a plurality of semiconductor layers 31 on the lower layer side constitute a first group, and a plurality of semiconductor layers 31 on the upper layer side constitute a second group, and the thicknesses of the semiconductor layers 31 are different between the groups; however, this configuration can be applied to the structure of the first embodiment in which only the X-directional dimension of the word line pillar WLP gradually increases from the lower layer side to the upper layer side.
The above configuration of the third embodiment can be applied to the fourth embodiment. In this case, for example, a plurality of semiconductor layers 31 on the lower layer side and the upper layer side including the semiconductor layer 31 corresponding to the largest portion of the conductive layer 33A constitute a first group, and a plurality of semiconductor layers 31 below the first group constitute a second group, and a plurality of semiconductor layers 31 above the first group constitute a third group, and the thicknesses of the semiconductor layers 31 are different among the first, second, and third groups.
In the fourth embodiment, the word line pillar WLP in both the X-directional and Y-directional dimensions has the largest portion between the lower layer portion and the upper layer portion, and the thickness of the semiconductor layer 31 corresponding to the largest portion is the smallest; however, even when the word line pillar WLP in only the X-directional dimension has the largest portion between the lower layer portion and the upper layer portion, and has approximately the same Y-directional dimension, the thickness of the semiconductor layer 31 corresponding to the largest portion may be the smallest.
While some embodiments have been described, those embodiments have been presented as examples, and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. The embodiments and their modifications are included in the scope and spirit of the invention and are included in the scope of the claimed inventions and their equivalents.
Number | Date | Country | Kind |
---|---|---|---|
2020-050773 | Mar 2020 | JP | national |