US20170271364A1 - Nonvolatile semiconductor memory device and method of manufacturing the same - Google Patents

Abstract

This nonvolatile semiconductor memory device includes: a memory cell array including a memory cell; a wiring part connecting the memory cell array to an external circuit; and a transistor that connects the wiring part and the external circuit, the transistor including: a first insulating layer including a first region, a second region, and a third region, the second and third regions being disposed on both sides of the first region, and a height of an upper surface of the first region being lower than those of the second region and the third region; a semiconductor layer disposed along upper surfaces of the first region, the second region, and the third region; and a gate electrode layer disposed via the semiconductor layer and a gate insulating film, on an upper part of the second region.

Description

CROSS REFERENCE TO RELATED APPLICATIONS
• This application is based on and claims the benefit of priority from prior Japanese Patent Application No. 2016-054724, filed on Mar. 18, 2016, the entire contents of which are incorporated herein by reference.
BACKGROUND
• Field
• Embodiments described below relate to a nonvolatile semiconductor memory device.
• Description of the Related Art
• In recent years, a three-dimensional type nonvolatile semiconductor memory device has been gathering attention as a nonvolatile semiconductor memory device for achieving raising of integration level, without being confined to a limit of resolution of lithography technology.
• Such a three-dimensional type nonvolatile semiconductor memory device includes a wiring part for connection to an external peripheral circuit. The wiring part is connected to the external peripheral circuit via the likes of a contact plug and upper layer wiring, via a switch transistor.
• However, along with an increase in memory capacity, the number of switch transistors has ended up increasing, and this is a factor that hinders raising of the integration level.
BRIEF DESCRIPTION OF THE DRAWINGS
• FIG. 1 is a block diagram showing a schematic configuration of a NAND type flash memory according to a first embodiment.
• FIG. 2 is a perspective view showing a schematic configuration of a memory cell array MA and a stepped wiring part SR.
• FIG. 3 is a circuit diagram describing a circuit configuration of the memory cell array MA.
• FIG. 4 is a schematic perspective view of the memory cell array MA.
• FIG. 5 is a cross-sectional view of the memory cell array MA and the stepped wiring part SR.
• FIG. 6 is a cross-sectional view showing details of a structure of amemory transistor layer30.
• FIG. 7 is a perspective view showing a structure of a transistor SWTr included in a word line connection circuit SW formed in an upper part of the stepped wiring part SR, of the first embodiment.
• FIG. 8 is a cross-sectional view of the transistor SWTr formed in the upper part of the stepped wiring part SR.
• FIGS. 9A to 9H show a method of manufacturing the transistor SWTr according to the first embodiment.
• FIG. 10 shows a modified example of the transistor SWTr according to the first embodiment.
• FIG. 11 shows a modified example of the transistor SWTr according to the first embodiment.
• FIG. 12 is a cross-sectional view showing a structure of a transistor SWTr according to a second embodiment.
• FIGS. 13A to 13F show a method of manufacturing the transistor SWTr according to the second embodiment.
• FIG. 14 shows a modified example of the transistor SWTr according to the second embodiment.
DETAILED DESCRIPTION
• A nonvolatile semiconductor memory device according to an embodiment includes: a memory cell array including a memory cell; a wiring part connecting the memory cell array to an external circuit; and a transistor that connects the wiring part and the external circuit, the transistor including: a first insulating layer including a first region, a second region, and a third region, the second and third regions being disposed on both sides of the first region, and a height of an upper surface of the first region being lower than those of the second region and the third region; a semiconductor layer disposed along upper surfaces of the first region, the second region, and the third region; and a gate electrode layer disposed via the semiconductor layer and a gate insulating film, on an upper part of the second region.
• Next, nonvolatile semiconductor memory devices according to embodiments will be described in detail with reference to the drawings.
First Embodiment
• First, a NAND type flash memory according to a first embodiment will be described with reference toFIG. 1. As shown inFIG. 1, this NAND type flash memory of the first embodiment includes a memory cell array MA.
• In addition, this NAND type flash memory includes a row decoder RD, a word line connection circuit SW, a bit line connection circuit BLHU, a sense amplifier circuit S/A, and a peripheral circuit PERI, in a periphery of the memory cell array MA.
• As will be described later, the memory cell array MA has memory cells arranged three-dimensionally therein. Moreover, this memory cell array MA includes: a plurality of word lines WL that extend longitudinally in an X direction ofFIG. 1; and a plurality of bit lines BL and a source line SL that extend longitudinally in a Y direction ofFIG. 1. The plurality of word lines WL are stacked in a stacking direction (Z direction) in the memory cell array MA. The plurality of word lines WL are each connected to different memory cells MC arranged in the stacking direction in the memory cell array MA.
• In addition, a stepped wiring part SR for connecting the word line WL and an external circuit is formed in the periphery of this memory cell array MA. As shown inFIG. 2, this stepped wiring part SR includes a lead-out wiring connected to an identical layer to aconductive layer31 acting as the word line WL, and this lead-out wiring is formed in a stepped shape. The stepped wiring part SR is formed by etching stackedconductive layers31 and inter-layerinsulating films32 while performing slimming processing isotropically on a resist, by a well-known method. Therefore, as shown inFIGS. 1 and 2, the stepped wiring part SR is usually formed so as to surround all sides of the memory cell array MA.
• The row decoder RD selects the plurality of word lines WL and supplies a voltage required in an operation. Moreover, the word line connection circuit SW is a switching circuit for connecting the word line WL and the row decoder RD, and includes many transistors SWTr that connect the word line WL and the row decoder RD. As will be described later, the transistor SWTr configuring this word line connection circuit SW is disposed so as to be superposed on the stepped wiring part SR in an XY plane, upwardly of the stepped wiring part SR.
• Furthermore, the sense amplifier circuit S/A and the bit line connection circuit BLHU are disposed in the Y direction of the stepped wiring part SR. The sense amplifier circuit S/A is connected to the bit line BL via the bit line connection circuit BLHU, and functions to provide the bit line BL with a voltage for write and to detect and amplify a potential appearing in the bit line BL during read. The bit line connection circuit BLHU includes a transistor that controls connection of the bit line BL and the sense amplifier circuit S/A. The peripheral circuit PERI includes a circuit other than the above-described, for example, a power supply circuit, a charge pump circuit (booster circuit), a data register, and so on.
• Next, a circuit configuration of the memory cell array MA will be described.FIG. 3 is a circuit diagram of the memory cell array MA. Note that this structure of the memory cell array MA shown inFIG. 3 is merely an example. It goes without saying that a later-described structure of the word line connection circuit SW may be applied also to a variety of three-dimensional type memory cell arrays other than that shown.
• As shown inFIG. 3, the memory cell array MA includes a plurality of memory blocks MB. The memory blocks MB are arranged in the Y direction on a semiconductor substrate Ba (not shown).
• The memory block MB includes a plurality of memory strings MS, a source side select transistor SSTr, and a drain side select transistor SDTr. The memory string MS is configured by memory transistors (memory cells) MTr1-MTr4 connected in series. To simplify description, the example shown describes the case where one memory string MS includes four memory transistors MTr, but it goes without saying that the present embodiment is not limited to this, and one memory string MS may include more memory transistors.
• The drain side select transistor SDTr is connected to one end of the memory string MS (the memory transistor MTr4). The source side select transistor SSTr is connected to the other end of the memory string MS (the memory transistor MTr1). For example, the memory strings MS are provided in a matrix in the XY plane over a plurality of rows and a plurality of columns, in each one of the memory blocks MB.
• In the memory block MB, control gates of the memory transistors MTr1 arranged in a matrix are commonly connected to a word line WL1. Similarly, control gates of the memory transistors MTr2 are commonly connected to a word line WL2. Control gates of the memory transistors MTr3 are commonly connected to a word line WL3. Control gates of the memory transistors MTr4 are commonly connected to a word line WL4.
• Moreover, in the memory block MB, control gates of each of the drain side select transistors SDTr arranged in a line in the X direction are commonly connected to a drain side select gate line SGD. A plurality of the drain side select gate lines SGD are provided at a certain pitch in the Y direction in one memory block MB. Moreover, other ends of the drain side select transistors SDTr arranged in a line in the Y direction are commonly connected to the bit line BL. The bit line BL is formed so as to extend in the Y direction straddling the memory block MB. A plurality of the bit lines BL are provided in the X direction.
• In one memory block MB, control gates of all of the source side select transistors SSTr are commonly connected to a source side select gate line SGS. Moreover, other ends of the source side select transistors SSTr arranged in the Y direction are commonly connected to a source line SL.
• The above-described kind of circuit configuration of the memory cell array MA is achieved by a stacked structure shown inFIGS. 4 and 5.FIG. 4 is a schematic perspective view of the memory cell array MA.FIG. 5 is a cross-sectional view of the memory cell array MA and the stepped wiring part SR.
• As shown inFIGS. 4 and 5, the memory cell array MA includes a source sideselect transistor layer20, amemory transistor layer30, a drain sideselect transistor layer40, and awiring layer50 that are stacked sequentially on the semiconductor substrate Ba in each of the memory blocks MB.
• The source sideselect transistor layer20 is a layer functioning as the source side select transistor SSTr. Thememory transistor layer30 is a layer functioning as the memory string MS (memory transistors MTr1-MTr4). The drain sideselect transistor layer40 is a layer functioning as the drain side select transistor SDTr. Thewiring layer50 is a layer functioning as various kinds of wirings.
• As shown inFIGS. 4 and 5, the source sideselect transistor layer20 includes a source side first insulatinglayer21, a source sideconductive layer22, and a source side second insulatinglayer23 that are formed sequentially above the semiconductor substrate Ba. The source sideconductive layer22 is formed so as to extend two-dimensionally (in a plate-like shape) in the X direction and the Y direction, over the memory block MB.
• The source side first insulatinglayer21 and the source side second insulatinglayer23 are configured by silicon oxide (SiO₂), for example. The source sideconductive layer22 is configured by polysilicon (p-Si), for example. Moreover, as shown inFIG. 5, the source sideselect transistor layer20 includes asource side hole24 formed so as to penetrate the source side first insulatinglayer21, source sideconductive layer22, and source side second insulatinglayer23. The source side holes24 are formed in a matrix in the X direction and the Y direction.
• Furthermore, as shown inFIG. 5, the source sideselect transistor layer20 includes a source sidegate insulating layer25 and a source sidecolumnar semiconductor layer26 that are formed sequentially on a sidewall facing thesource side hole24. The source sidegate insulating layer25 is formed with a certain thickness on the sidewall facing thesource side hole24. The source sidecolumnar semiconductor layer26 is formed so as to fill thesource side hole24. The source sidecolumnar semiconductor layer26 is formed in a column shape extending in the stacking direction. An upper surface of the source sidecolumnar semiconductor layer26 is formed so as to contact a lower surface of a later-describedcolumnar semiconductor layer35. The source sidecolumnar semiconductor layer26 is formed on a source diffusion layer Ba1 on the semiconductor substrate Ba. The source diffusion layer Ba1 functions as the source line SL.
• The source sidegate insulating layer25 is configured by silicon oxide (SiO₂), for example. The source sidecolumnar semiconductor layer26 is configured by polysilicon (p-Si), for example.
• In the above-described configuration of the source sideselect transistor layer20, the source sideconductive layer22 functions as the control gate of the source side select transistor SSTr and as the source side select gate line SGS.
• In addition, as shown inFIGS. 4 and 5, thememory transistor layer30 includes first through fourth word lineconductive layers31a-31dand first through fourth inter-wordline insulating films32a-32dthat are stacked sequentially on the source sideselect transistor layer20. The first through fourth word lineconductive layers31a-31dand the first through fourth inter-wordline insulating films32a-32dare formed so as to extend two-dimensionally (in a plate-like shape) in the X direction and the Y direction. The first through fourth word lineconductive layers31a-31dand the first through fourth inter-wordline insulating films32a-32dare divided for each memory block MB.
• Moreover, as shown inFIG. 5, thememory transistor layer30 includes amemory hole33 formed so as to penetrate the first through fourth word lineconductive layers31a-31dand the first through fourth inter-wordline insulating films32a-32d. Thememory holes33 are formed in a matrix in the X direction and the Y direction. Thememory hole33 is formed at a position aligned with thesource side hole24.
• Furthermore, as shown inFIG. 6, thememory transistor layer30 includes ablock insulating layer34a, acharge accumulation layer34b, atunnel insulating layer34c, and thecolumnar semiconductor layer35 that are formed sequentially on a sidewall facing thememory hole33.
• As shown inFIG. 6, theblock insulating layer34ais formed with a certain thickness on the sidewall facing thememory hole33. Thecharge accumulation layer34bis formed with a certain thickness on a sidewall of theblock insulating layer34a. Thetunnel insulating layer34cis formed with a certain thickness on a sidewall of thecharge accumulation layer34b. Thecolumnar semiconductor layer35 is formed so as to fill thememory hole33. Thecolumnar semiconductor layer35 is formed in a column shape so as to extend in the stacking direction. The lower surface of thecolumnar semiconductor layer35 is formed so as to contact the upper surface of the source sidecolumnar semiconductor layer26. Moreover, an upper surface of thecolumnar semiconductor layer35 is formed so as to contact a lower surface of a later-described drain sidecolumnar semiconductor layer44. Note that thecolumnar semiconductor layer35 may also be configured having an insulating film core at its center. Note that theblock insulating layer34aand thetunnel insulating layer34care configured by silicon oxide (SiO₂), for example. Thecharge accumulation layer34bis configured by silicon nitride (SiN), for example. Thecolumnar semiconductor layer35 is configured by polysilicon (p-Si), for example.
• In the above-described configuration of thememory transistor layer30, the first through fourth word lineconductive layers31a-31dfunction as the control gates of the memory transistors MTr1-MTr4 and as the word lines WL1-WL4.
• As shown inFIGS. 4 and 5, the drain sideselect transistor layer40 includes a drain sideconductive layer41 stacked on thememory transistor layer30. The drain sideconductive layer41 is formed directly above where thecolumnar semiconductor layer35 is formed. The drain sideconductive layers41 extend having longitudinally the X direction, and are formed in stripes with a certain pitch in the Y direction. The drain sideconductive layer41 is configured by polysilicon (p-Si), for example.
• Moreover, as shown inFIG. 5, the drain sideselect transistor layer40 includes adrain side hole42 formed so as to penetrate the drain sideconductive layer41. The drain side holes42 are formed in a matrix in the X direction and the Y direction. Thedrain side hole42 is formed at a position aligned with thememory hole33.
• Furthermore, as shown inFIG. 5, the drain sideselect transistor layer40 includes a drain sidegate insulating layer43 and a drain sidecolumnar semiconductor layer44 that are formed sequentially on a sidewall facing thedrain side hole42. The drain sidegate insulating layer43 is formed with a certain thickness on the sidewall facing thedrain side hole42. The drain sidecolumnar semiconductor layer44 is formed so as to fill thedrain side hole42. The drain sidecolumnar semiconductor layer44 is formed in a column shape so as to extend in the stacking direction. The lower surface of the drain sidecolumnar semiconductor layer44 is formed so as to contact the upper surface of thecolumnar semiconductor layer35. Note that the drain sidegate insulating layer43 is configured by silicon oxide (SiO₂), for example. The drain sidecolumnar semiconductor layer44 is configured by polysilicon (p-Si), for example. The drain sideconductive layer41 functions as the control gate of the drain side select transistor SDTr and as the drain side select gate line SGD.
• As shown inFIG. 5, thewiring layer50 includes the likes of afirst wiring layer51 in a region including the memory cell array MA, and includes the likes of a channel semiconductor layer CR, a gate insulating film GI, and a gate electrode layer GE in a region including the stepped wiring part SR.
• Thefirst wiring layer51 is formed so as to contact an upper surface of the drain sidecolumnar semiconductor layer44. The first wiring layers51 are formed with a certain pitch in the X direction so as to extend in the Y direction. Thefirst wiring layer51 functions as the bit line BL.
• Moreover, the channel semiconductor layer CR, the gate insulating film GI, and the gate electrode layer GE are members configuring the previously described transistor SWTr in the word line connection circuit SW. The channel semiconductor layer CR functions as a channel region of these transistors SWTr. As will be described later, the gate insulating film GI is formed on a surface of the channel semiconductor layer CR, and the gate electrode layer GE is formed on the surface of the channel semiconductor layer CR via this gate insulating film GI.
• The channel semiconductor layer CR may be configured by, for example, polysilicon, monocrystalline silicon, TiO₂, or a semiconductor oxide (for example, InGaZnO, ZnO, InOx, and so on).
• The gate insulating film GI is configured from a silicon oxide (SiO_x), for example. The gate insulating film GI, besides being configured from a silicon oxide, may also be configured from at least one of silicon nitride (SiN), silicon oxynitride (SiON), aluminum oxide (Al₂O₃), aluminum oxynitride (AlON), hafnia (HfO₂), hafnium aluminate (HfAlO₃), hafnium oxynitride (HfON), hafnium aluminate nitride (HfAlON), hafnium silicate (HfSiO), hafnium silicate nitride (HfSiON), lanthanum oxide (La₂O₃), and lanthanum aluminate (LaAlO₃).
• The gate electrode layer GE may be formed from, for example, polysilicon or polysilicon to which an impurity has been added. The following may be included instead of polysilicon, namely a metal compound such as tantalum nitride (TaN), tantalum carbide (TaC), and titanium nitride (TiN), or the following that show metallic electrical conductivity characteristics, namely Ni, V, Cr, Mn, Y, Mo, Ru, Rh, Hf, Ta, W, Ir, Co, Ti, Er, Pt, Pd, Zr, Gd, Dy, Ho, Er, and silicides of these.
• As shown inFIG. 5, the stepped wiring part SR includesconductive layers31a′-31d′ formed by extending the first through fourth word lineconductive layers31a-31d. That is, theconductive layers31a′-31d′ are formed in identical layers to, and are electrically and physically connected to the first through fourth word lineconductive layers31a-31d. Theconductive layers31a′-31d′ and first through fourth inter-wordline insulating films32a′-32d′ sandwiched between theconductive layers31a′-31d′ are formed in a stepped shape such that positions of their ends in the X direction differ, and configure a stepped part ST. The stepped part ST (ST1-ST4) inFIG. 5 configures part of the stepped wiring part SR of the kind shown inFIG. 2. The stepped part ST shown specifically inFIG. 5 includes the stepped parts ST1-ST4 formed by theconductive layers31a′-31d′ and the first through fourth inter-wordline insulating films32a′-32d′ the positions of whose ends differ in the X direction.
• A contact plug C1 having longitudinally the stacking direction (Z direction) extends from each of these stepped parts ST1-ST4, so as to penetrate the first through fourth inter-wordline insulating films32a′-32d′.
• As shown inFIG. 5, an inter-layer insulatinglayer60 is formed so as to fill a periphery of the source sideselect transistor layer20, thememory transistor layer30, and the drain sideselect transistor layer40.
• A structure of the transistor SWTr according to the present embodiment will be described with reference toFIGS. 7 and 8.
• FIG. 7 is a perspective view showing the structure of the transistor SWTr included in the word line connection circuit SW formed in an upper part of the stepped wiring part SR.FIG. 8 is a cross-sectional view in a YZ plane of the transistor SWTr formed in the upper part of the stepped wiring part SR. Note that to simplify illustration,FIG. 7 does not show the inter-layer insulatinglayer60.
• As shown inFIGS. 7 and 8, the transistor SWTr included in the word line connection circuit SW includes a base insulating layer BI (first insulating layer), the channel semiconductor layer CR (semiconductor layer), the gate insulating film GI, the gate electrode layer GE, an embedded insulating layer EI (second insulating layer), and a diffusion layer DL (impurity layer). These channel semiconductor layer CR, gate insulating film GI, and gate electrode layer GE form a MOS structure, and configure a thin film transistor (TFT) structure in which a current flowing between the diffusion layers DL is controlled by applying a voltage to the gate electrode layer GE.
• As shown inFIG. 7, the channel semiconductor layer CR is formed along a longitudinal direction of the stepped parts ST1-ST4. In the case of such a configuration, many channel semiconductor layers CR can be disposed along a shape of the stepped part ST, and it becomes easy for more transistors SWTr to be formed.
• In the example shown inFIG. 7, the gate electrode layer GE of this embodiment is commonly (continuously) connected over a plurality of the channel semiconductor layers CR. As a result, one gate electrode layer GE (of plate-like shape) is connected to a plurality of the transistors SWTr, and the number of upper layer wirings or contact plugs can be reduced. The gate electrode layer GE is not limited to such a configuration, and may be provided independently to each different transistor SWTr.
• Moreover, as shown inFIGS. 7 and 8, an upper end of the contact plug C1 is connected to one end (a back surface) of the channel semiconductor layer CR where the diffusion layer DL is formed. On the other hand, a contact plug C2 separate from the contact plug C1 is connected to the other end (a front surface) of the channel semiconductor layer CR. An upper layer wiring M1 (not shown) is connected to an upper end of the contact plug C2. This upper layer wiring M1 is connected to the row decoder RD via another contact plug or wiring in an unshown region.
• In the example shown inFIG. 8, the base insulating layer BI includes two trenches T1 (concave parts), and the channel semiconductor layer CR is deposited so as to cover a surface of the base insulating layer BI. A trench T2 is formed by an upper surface of the channel semiconductor layer CR following the trench T1. The case where cross sections of the trenches T1 and T2 are rectangular is shown, but another shape may also be adopted for the trenches T1 and T2. For example, the cross sections of the trenches T1 and T2 may be configured as the likes of a trapezoid, a triangle, a polygon, a semi-ellipse, or a semi-circle. Moreover,FIG. 8 shows the case where the two trenches T1 have a similar shape, but they may be configured such that their depths in the stacking direction or shapes are different.
• By thus forming the trench T1 in the base insulating layer BI and forming the channel semiconductor layer CR on the surface of the base insulating layer BI so as to also follow the trench T1, a channel length of the transistor SWTr can be made larger than when the trench T1 is not formed. Note that the number or size and shape of the trenches T1 (first regions) may be appropriately changed, and are of course not limited to those shown.
• As shown inFIG. 8, in a region (convex part) sandwiched by the two trenches T1, the gate electrode layer GE is formed via the channel semiconductor layer CR and the gate insulating film GI, above the base insulating layer BI. The trench T1 formed in the base insulating layer BI is embedded by the channel semiconductor layer CR and the embedded insulating layer EI formed on the channel semiconductor layer CR. Now, a region where the trench T1 is formed, of the base insulating layer BI, is assumed to be a first region; a region where the gate electrode layer GE is formed, of the base insulating layer BI, is assumed to be a second region; and a region other than the first region and the second region, of the base insulating layer BI, is assumed to be a third region. That is, the trench T1 is formed by the first region having a height of its upper surface configured lower than those of the second region and the third region. In the example shown, the upper surface of the second region and the upper surface of the third region have the same height. As shown, the embedded insulating layer EI is planarized so as to attain the same height as the upper surfaces of the channel semiconductor layer CR on the second region and the third region.
• In the example shown inFIG. 8, the gate electrode layer GE is formed in an upper part of the second region sandwiched by inner sidewall surfaces of the two trenches T1, and the diffusion layer DL is formed in each of the third regions adjacent to outer sidewall surfaces of the two trenches T1. In other words, the diffusion layer DL is formed at both ends of the channel semiconductor layer CR. In other words, the transistor SWTr includes: a first insulating layer having on its surface a concave part and a convex part adjacent to the concave part; a semiconductor layer disposed along the surface including the concave part and the convex part of the first insulating layer; and a gate electrode layer disposed via the semiconductor layer and a gate insulating film, on the convex part. By adopting such a structure, the channel length of the transistor SWTr can be made larger than when the trench T1 is not formed. A method of manufacturing such a transistor structure will be described later.
• The diffusion layer DL is formed by an impurity being implanted in part of the channel semiconductor layer CR, and functions as a source/drain region of the transistor SWTr. A region between the second region where the gate electrode layer GE is formed and the diffusion layer DL, of the channel semiconductor layer CR, that is, the channel semiconductor layer CR along a sidewall surface and upper surface of the first region is not implanted with an impurity, or is implanted with an impurity of a concentration 10 times or more less compared to that of the diffusion layer DL. This region is called an offset region of the diffusion layer DL. For example, a large offset region is required as a means for improving withstand voltage when a high voltage is applied to the gate electrode layer GE or drain.
• As shown inFIG. 8, the transistor SWTr of the present embodiment is formed such that part of the offset region follows a bottom surface and sidewall surface of the trench T1 formed in the base insulating layer BI, and has a partly folded kind of shape. Forming in this way makes it possible for a transistor SWTr provided with a large offset region to be made small on a plane, hence a reduction of chip area can be achieved. Now, when an offset length in a plane indicated by d inFIG. 8 is small, respective potentials interfere with each other between the gate electrode and diffusion layers, hence an offset effect ends up weakening. Therefore, it is desirable that an offset length d is to a certain extent large. Specifically, the offset length d is desirably 600 nm or more.
• Moreover, in the present embodiment, the trench T1 formed in the base insulating layer BI has an aspect ratio in cross section whose value is close to 1.
• Next, manufacturing steps of the transistor according to the first embodiment will be described with reference toFIGS. 9A to 9H.
• Upper portions ofFIGS. 9A to 9E,FIG. 9F, and upper portions ofFIGS. 9G and 9H are cross-sectional views in the YZ plane; and lower portions ofFIGS. 9A to 9E, 9G, and 9H are arrow views of the XY plane as seen from above.
• First, as shown inFIG. 9A, a base insulating layer BI′ which will become the base insulating layer BI is formed by depositing silicon oxide (SiO₂), for example. Next, as shown inFIG. 9B, the two trenches T1 are formed in part of the base insulating layer BI′ by implementing anisotropic dry etching after patterning by photolithography, for example, whereby the base insulating layer BI is formed. The base insulating layer BI may be formed by silicon nitride (SiN), for example, instead of silicon oxide.
• Next, as shown inFIG. 9C, polysilicon, for example, is deposited so as to follow a surface of the base insulating layer BI including a bottom surface and sidewall surface of the trench T1, and a channel semiconductor layer CR′ which will become the channel semiconductor layer CR is deposited. As a result, a trench T2′ which will become the trench T2 in a later step, is formed.
• Next, as shown inFIG. 9D, in order that the channel semiconductor layer CR′ on parts of the second region and the third region and part of the channel semiconductor layer CR′ in the trench T1 are continuous, another part of the channel semiconductor layer CR′ is removed by implementing anisotropic dry etching after patterning by photolithography, for example, whereby the channel semiconductor layer CR′ undergoes division processing in the X direction. As a result, an upper surface of the channel semiconductor layer CR′ divided in the trench T1 becomes the trench T2. Moreover, division of the channel semiconductor layer CR′ results in part of the trench T1 being exposed. At this time, the channel semiconductor layer CR′, by undergoing division processing in the X direction without being divided in the Y direction, attains a shape extending as a whole in the Y direction so as to follow surfaces of the second region and the third region and a bottom surface and sidewall surface of the trench T1. Note that channel semiconductor layers CR and CR′ formed with a uniform thickness along the trench T1 are described for reference, but the likes of shape, size or thickness may be arbitrarily changed.
• Next, as shown inFIG. 9E, the embedded insulating layer EI (for example, silicon oxide) is deposited on the trenches T1 and T2, and planarization processing by CMP (Chemical Mechanical Polishing) is further implemented, whereby the trenches T1 and T2 are embedded. The embedded insulating layer EI may be formed by silicon nitride instead of silicon oxide.
• Next, as shown inFIG. 9F, for example, polysilicon and polysilicon to which an impurity has been added are deposited in this order, and a gate insulating film GI′ which will become the gate insulating film GI and a gate electrode layer GE′ which will become the gate electrode layer GE are formed on surfaces of the embedded insulating layer EI and a convex part of the channel semiconductor layer CR′.
• Next, as shown inFIG. 9G, the gate electrode layer GE′ undergoes division processing in the X direction by implementing anisotropic dry etching after patterning by photolithography, for example. At this time, the gate insulating film GI′ need not be divided. Moreover, an end of the gate electrode GE is formed so as to overlap in the Z direction the channel semiconductor layer CR′ formed along a sidewall surface on a gate electrode layer GE side of the trench T1.
• Next, as shown inFIG. 9H, an impurity is implanted in an entire surface of the base insulating layer BI by ion implantation, for example, and the diffusion layer DL is formed at both ends of the channel semiconductor layer CR, whereby the same structure as inFIG. 8 is formed. At this time, the gate electrode layer GE is formed so as to overlap the channel semiconductor layer CR. That is, a length in the X direction of the gate electrode layer GE is formed so as to be larger than a length in the X direction of the channel semiconductor layer CR on the second region. As a result, due to the gate electrode layer GE or the embedded insulating layer EI, it is difficult for the impurity to be implanted in, respectively, the second region directly below the gate electrode layer GE or a region where the trench T1 has been embedded (first region). Therefore, the diffusion layer DL can be formed by self-alignment by a method implanting the entire surface without performing patterning by photolithography, and a low cost can be achieved. Possible materials of the impurity of the diffusion layer DL and the offset region when implanted with the impurity at low concentration, are an impurity configuring an n type semiconductor, for example, a pentavalent element such as arsenic (As) and phosphorus (P), an impurity configuring a p type semiconductor, for example, a trivalent element such as boron (B) and indium (In), or a combination of these materials.
• Modified examples of the transistor SWTr according to the first embodiment are shown with reference toFIGS. 10 and 11. In the case ofFIG. 10, the gate electrode layer GE does not overlap the channel semiconductor layer CR. Even in this case, a similar structure is obtained by performing patterning by photolithography, for example.
• In the case ofFIG. 11, only one trench T1 is formed in the base insulating layer BI. That is, the base insulating layer BI includes the following in its B-B cross section, sequentially from the left side ofFIG. 11, namely: the third region where the diffusion layer DL is formed; the first region where the trench T1 is formed; the second region having the gate electrode layer GE formed in its upper part; and the third region having the diffusion layer DL formed at its end. Even in this case, the first region has a height of its upper surface configured lower than those of the second region and the third region. By making the trench T1 only on one side of the gate electrode layer GE in this way, it is also possible to make a different offset length and to form a transistor SWTr in which a current is easily led out only from the diffusion layer DL on one side configuring a short offset.
Second Embodiment
• Next, a second embodiment will be described with reference toFIGS. 12 to 14. An overall configuration of a nonvolatile semiconductor memory device according to the second embodiment is similar to that of the first embodiment (FIGS. 1 to 6), hence a detailed description thereof will be omitted. In the second embodiment, the structure of the transistor SWTr is different from in the first embodiment.
• FIG. 12 is a cross-sectional view in the YZ plane of the transistor SWTr according to the second embodiment. Similarly to in the first embodiment, the transistor SWTr includes the base insulating layer BI (first insulating layer), the channel semiconductor layer CR (semiconductor layer), the gate insulating film GI, the gate electrode layer GE, and the diffusion layer DL. In addition, part of the channel semiconductor layer CR is formed so as to follow a bottom surface and sidewall surface of a trench T3 formed in the base insulating layer BI, and has a folded kind of shape. A trench T4 is formed by an upper surface of the channel semiconductor layer CR along the trench T3. The case where cross sections of the trenches T3 and T4 are trapezoids is shown, but another shape may also be adopted for the trenches T3 and T4. The example shown shows the case where each of the trenches T3 and T4 respectively all have similar shapes, but the present embodiment is not limited to this, and each of the trenches T3 and T4 may also be given different shapes. Moreover, the diffusion layer DL is formed at both ends of the channel semiconductor layer CR. Now, similarly to in the first embodiment, a region where the trench T3 is formed, of the base insulating layer BI, is assumed to be a first region; a region where the gate electrode layer GE is formed, of the base insulating layer BI, is assumed to be a second region; and a region other than the first region and the second region, of the base insulating layer BI, is assumed to be a third region. Even in the second embodiment, the first region has a height of its upper surface configured lower than those of the second region and the third region.
• In the present embodiment, contrary to in the first embodiment, a plurality (inFIG. 12, three each on either side, a total of six) of the trenches T3 are formed in a region between the second region where the gate electrode layer GE is formed and the diffusion layer DL. That is, the base insulating layer BI includes three first regions on both sides sandwiching the second region where the gate electrode GE is formed, and has a structure in its B-B cross section that includes the following, sequentially from the left side ofFIG. 12, namely: three each alternately of the third regions and the first regions; the second region; and three each alternately of the first regions and the third regions. Moreover, contrary to in the first embodiment, aspect ratios in cross section of each of the trenches T3 are formed so as to be larger than 1. By thus forming a plurality of long-and-thin trenches T3 respectively on both sides sandwiching the gate electrode layer GE, it is possible that while the offset length in the plane indicated by d inFIGS. 8 and 12 is maintained unchanged, an effective offset length is further increased. Moreover, the trench T3 formed in the base insulating layer BI is embedded simultaneously to when the gate insulating film GI is formed on the channel semiconductor layer CR. Details of this embedding will be described in the following manufacturing steps.
• Manufacturing steps of the transistor according to the second embodiment will be described with reference toFIGS. 13A to 13F. Upper portions ofFIGS. 13A to 13F are cross-sectional views in the YZ plane; and lower portions ofFIGS. 13A to 13F are arrow views of the XY plane as seen from above.
• First, as shown inFIG. 13A, the base insulating layer BI formed by depositing silicon oxide (SiO₂) has the trench T3 formed therein, by performing patterning by photolithography and then implementing anisotropic dry etching. At this time, contrary to in the case ofFIG. 9B, patterning is performed such that a plurality of the trenches T3 are respectively formed on both sides of the region where the gate electrode layer GE is formed (second region). At this time, contrary to in the first embodiment, aspect ratios of each of the trenches T3 are formed so as to be larger than 1. By forming the trench T3 in this way, even if a length of d indicated inFIG. 12 is about the same as in the case where there is one trench T3 (is about the same as a length of d indicated inFIG. 8), a plurality of longer-and-thinner trenches T3 are formed in the same region, whereby the effective offset length can be further increased. Note that the number and size of the trenches T3, and intervals between each of the trenches T3 are not limited to those shown, and may be appropriately changed.
• Next, as shown inFIG. 13B, polysilicon, for example, is deposited so as to follow a surface of the base insulating layer BI including a bottom surface and sidewall surface of the trench T3, whereby a channel semiconductor layer CR′ is formed. At this time, by preventing the inside of the trench T3 from being completely filled, a trench T4′ which will become the trench T4 in a later step, can be formed.
• Next, as shown inFIG. 13C, in order that the channel semiconductor layer CR′ on parts of the second region and the third region and part of the channel semiconductor layer CR′ in the trench T3 are continuous, another part of the channel semiconductor layer CR′ is removed by implementing anisotropic dry etching after patterning by photolithography, whereby the channel semiconductor layer CR′ undergoes division processing in the X direction. As a result, an upper surface of the channel semiconductor layer CR′ divided in the trench T3 becomes the trench T4. Moreover, division of the channel semiconductor layer CR′ results in part of the trench T3 being exposed. At this time, the channel semiconductor layer CR′, by undergoing division processing in the X direction without being divided in the Y direction, attains a shape extending as a whole in the Y direction so as to follow upper surfaces of the second region and the third region and a bottom surface and sidewall surface of the trench T3.
• Next, the trenches T3 and T4 are embedded by an insulating layer similarly to inFIG. 9, but in the present embodiment, as shown inFIG. 13D, silicon oxide (SiO₂) which will become the gate insulating film GI′ can be deposited on the channel semiconductor layer CR′ so as to embed the trenches T3 and T4. By thus adjusting a width in the Y direction of the trench T3 in the step shown inFIG. 13A and a width in the Y direction of the trench T3 in the step shown inFIG. 13C, embedding by the embedded insulating layer EI and planarization processing can be omitted, and manufacturing costs can be lowered.
• Specifically, in order for embedding by the embedded insulating layer EI to be omitted and for the trenches T3 and T4 to be embedded simultaneously to deposition of the gate insulating film GI′, it becomes a condition that, if a width in the Y direction of an opening of the trench T3 is assumed to be a, a thickness of the channel semiconductor layer CR in a direction perpendicular to a surface of the base insulating layer BI including the trench T3 is assumed to be b, a distance between an end in the X direction of the opening of the trench T3 and an end in the X direction of that of the trench T4 shown in the lower portion ofFIG. 13C is assumed to be c, and a thickness in the stacking direction of the gate insulating film GI shown inFIG. 13D is assumed to be e, then a≦2 (b+e) and c≦2e are satisfied. In the case that this condition cannot be satisfied, it is also possible for the trench T3 to be embedded by the embedded insulating layer EI and for planarization processing to be performed, similarly to in the first embodiment. The example shown inFIG. 13D shows a state where the trench T3 has been embedded simultaneously to deposition of the gate insulating film GI.
• Next, as shown inFIG. 13E, for example, polysilicon to which an impurity has been added is deposited on the gate insulating film GI′, and the gate electrode layer GE is formed.
• Implementing anisotropic dry etching after patterning by photolithography and performing division processing in the Y direction of the gate electrode layer GE′ and the gate insulating film GI′ results in a structure in which the channel semiconductor layer CR′, the gate insulating film GI, and the gate electrode layer GE are disposed in this order on the second region. At this time, the gate insulating film GI′ need not be divided. Moreover, the gate electrode layer GE is formed so as to overlap the channel semiconductor layer CR of a sidewall surface of the trench T4 adjacent to the second region. That is, a length in the X direction of the gate electrode layer GE is formed so as to be larger than a length in the X direction of the channel semiconductor layer CR on the second region.
• Next, as shown inFIG. 13F, after performing patterning in advance, an impurity is implanted in entire surfaces of the base insulating layer BI and the channel semiconductor layer CR′ by the likes of ion implantation, and the diffusion layer DL is formed at both ends of the channel semiconductor layer CR, whereby the same structure as inFIG. 12 is formed. At this time, in a method where the impurity is implanted in entire surfaces, the impurity ends up being implanted also in part of the offset region, hence the impurity must be implanted after patterning has been performed in advance by photolithography, for example. Possible materials of the impurity are an impurity configuring an n type semiconductor, for example, a pentavalent element such as arsenic (As) and phosphorus (P), an impurity configuring a p type semiconductor, for example, a trivalent element such as boron (B) and indium (In), or a combination of these materials.
• FIG. 14 is a cross-sectional view showing a modified example of the transistor SWTr according to the second embodiment. In this example, a plurality of the trenches T3 are respectively formed on both sides sandwiching the gate electrode layer GE, but the number of trenches T3 formed on both sides is different on the left and on the right. By changing the number of trenches T3 in this way, it is also possible to make a different offset length and, similarly to in the modified example shown inFIG. 11, to form a transistor SWTr in which a current is easily led out only from the diffusion layer DL on one side configuring a short offset.
• As described above, even in the second embodiment, part of the channel semiconductor layer CR is formed so as to follow bottom surfaces and sidewall surfaces of the plurality of trenches T3 formed in the base insulating layer BI, and has a partly folded kind of shape. Forming in this way makes it possible for a transistor SWTr provided with a large offset region to be made small on a plane, and makes it possible to suppress an increase in area occupied by the word line connection circuit SW and reduce chip area as a whole.
• While certain embodiments have been described, these embodiments have been presented by way of example only, and are not intended to limit the scope of the inventions. Indeed, the novel methods and systems described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the methods and systems described herein may be made without departing from the spirit of the inventions. The accompanying claims and their equivalents are intended to cover such forms or modifications as would fall within the scope and spirit of the inventions.
• For example, in the above-described embodiments, the memory string of the NAND type flash memory adopts thesemiconductor layer35 extending linearly in the stacking direction. However, it is also possible to adopt a configuration of the kind where the offset region of the channel semiconductor layer CR of the transistor SWTr is folded, like that of the above-described embodiments, in a NAND type flash memory that has a structure in which thesemiconductor layer35 is instead folded back in a U shape, for example. Moreover, the configurations of the above-described embodiments are not limited to a three-dimensional type NAND type flash memory, and may also be applied to another three-dimensional memory, for example, a resistance varying memory, and so on. That is, the configurations of the above-described embodiments may be applied to various forms of three-dimensionally disposed memories.

Claims (19)

What is claimed is:

1. A nonvolatile semiconductor memory device, comprising:

a memory cell array including a memory cell;

a wiring part connecting the memory cell array to an external circuit; and

a transistor that connects the wiring part and the external circuit,

the transistor comprising:

a first insulating layer including a first region, a second region, and a third region, the second and third regions being disposed on both sides of the first region, and a height of an upper surface of the first region being lower than those of the second region and the third region;

a semiconductor layer disposed along upper surfaces of the first region, the second region, and the third region; and

a gate electrode layer disposed via the semiconductor layer and a gate insulating film, on an upper part of the second region.

2. The memory device according toclaim 1, further comprising

a second insulating layer disposed on the first region via the semiconductor layer.

3. The memory device according toclaim 1, wherein

the first insulating layer includes a plurality of the first regions whose upper surfaces are in a lower position than that of the second region is, on at least one side of the second region, and the third region is disposed between a plurality of the first regions.

4. The memory device according toclaim 2, wherein

the first insulating layer includes a plurality of the first regions whose upper surfaces are in a lower position than that of the second region is, on at least one side of the second region, and the third region is disposed between a plurality of the first regions.

5. The memory device according toclaim 1, wherein

the first insulating layer includes a plurality of the first regions whose upper surfaces are in a lower position than that of the second region is, on each of both sides of the second region, and the third region is disposed between a plurality of the first regions.

6. The memory device according toclaim 2, wherein

the first insulating layer includes a plurality of the first regions whose upper surfaces are in a lower position than that of the second region is, on each of both sides of the second region, and the third region is disposed between a plurality of the first regions.

7. The memory device according toclaim 3, wherein

a plurality of the first regions have an aspect ratio of a sidewall surface and the upper surface which is larger than 1.

8. The memory device according toclaim 5, wherein

a plurality of the first regions have an aspect ratio of a sidewall surface and the upper surface which is larger than 1.

9. The memory device according toclaim 1, wherein

a length of the gate electrode layer in a first direction from the second region to the third region along the upper surface of the second region is larger than a length in the first direction of the semiconductor layer on the second region.

10. The memory device according toclaim 2, wherein

a length of the gate electrode layer in a first direction from the second region to the third region along the upper surface of the second region is larger than a length in the first direction of the semiconductor layer on the second region.

11. The memory device according toclaim 3, wherein

a length of the gate electrode layer in a first direction from the second region to the third region along the upper surface of the second region is larger than a length in the first direction of the semiconductor layer on the second region.

12. The memory device according toclaim 1, wherein

the semiconductor layer comprises an impurity layer at both ends of the semiconductor layer.

13. The memory device according toclaim 2, wherein

the semiconductor layer comprises an impurity layer at both ends of the semiconductor layer.

14. The memory device according toclaim 3, wherein

the semiconductor layer comprises an impurity layer at both ends of the semiconductor layer.

15. The memory device according toclaim 12, wherein

the semiconductor layer has an impurity concentration between the impurity layers which is 10 times or more less than that of the impurity layer.

16. A method of manufacturing a nonvolatile semiconductor memory device, comprising:

forming a first insulating layer;

removing part of the first insulating layer to a certain depth from an upper surface of the first insulating layer to define a first region;

forming a semiconductor layer along a surface of the first insulating layer;

embedding a second insulating layer above an upper part of the first region where the semiconductor layer has been formed;

forming a gate insulating film above the semiconductor layer, and then forming a gate electrode layer above the gate insulating film; and

introducing an impurity to form an impurity layer in part of the semiconductor layer.

17. A method of manufacturing a nonvolatile semiconductor memory device, comprising:

forming a first insulating layer;

removing part of the first insulating layer to a certain depth from an upper surface of the first insulating layer to define a first region;

forming a semiconductor layer along follow a surface of the first insulating layer including the first region;

forming a gate insulating film above an upper surface of the semiconductor layer and forming the gate insulating film above an upper part of the first region;

forming a gate electrode layer above the gate insulating film in a second region different from the first region; and

introducing an impurity to form an impurity layer in part of the semiconductor layer.

18. The method according toclaim 16, wherein

the first region has an aspect ratio of a sidewall surface and the upper surface which is larger than 1.

19. The method according toclaim 17, wherein

the first region has an aspect ratio of a sidewall surface and the upper surface which is larger than 1.

Images