Semiconductor memory device and method for manufacturing the same

This application is based upon and claims the benefit of priority from U.S Provisional Patent Application 62/189,029, filed on Jul. 6, 2015; the entire contents of which are incorporated herein by reference.

Embodiments described herein relate generally to a semiconductor memory device and a method for manufacturing the same.

In a NAND flash memory, an integration degree of memory cells has been increased through shrinking of a plane structure to reduce bit cost. In recent years, in order to further improve the integration degree of memory cells, a technique for stacking the memory cells in an up-down direction has been proposed. However, even in such a stacked type semiconductor memory device, it is desired to highly integrate a plane structure.

FIG. 1 is a plan view showing a semiconductor memory device according to a first embodiment;

FIG. 2 is a sectional view taken along line A-A′ in FIG. 1 ;

FIG. 3 is a sectional view taken along line B-B′ in FIG. 1 ;

FIGS. 4A and 4B are sectional views showing the semiconductor memory device according to the first embodiment;

FIGS. 5A to 13B are sectional views showing a manufacturing method for the semiconductor memory device according to the first embodiment;

FIG. 14 is a plan view showing a semiconductor memory device according to a second embodiment;

FIG. 15 is a sectional view taken along line E-E′ in FIG. 14 ; and

FIG. 16 is a sectional view showing a semiconductor memory device according to a third embodiment.

A semiconductor memory device according to one embodiment, includes a plurality of first interconnects extending in a first direction and arrayed along a second direction crossing the first direction, a plurality of semiconductor pillars arrayed in a row along the first direction in each of spaces among the first interconnects and extending in a third direction crossing the first direction and the second direction, a first electrode disposed between one of the semiconductor pillars and one of the first interconnects, a first insulating film disposed between the first electrode and one of the first interconnects, and a first insulating member disposed between the semiconductor pillars in the first direction, extending in the third direction and opposed the first interconnects not via the first insulating film.

Embodiments of the invention are described below with reference to the drawings.

First Embodiment

First, a first embodiment is described.

FIG. 1 is a plan view showing a semiconductor memory device according to the embodiment.

FIG. 2 is a sectional view taken along line A-A′ in FIG. 1 .

FIG. 3 is a sectional view taken along line B-B′ in FIG. 1 .

FIGS. 4A and 4B are sectional views showing the semiconductor memory device according to the embodiment. FIG. 4B shows a cross section taken along line C-C′ in FIG. 4A . FIG. 4A shows a cross section taken along line D-D′ in FIG. 4B .

Note that, in FIGS. 1 to 3 , for convenience of illustration, only a part of a plurality of members is shown and the remaining part is omitted. For example, in FIG. 1 , in an upper part of the figure, components in an upper layer (e.g., upper layer interconnects 44 and 48 ) are shown on the top surface. In the middle of the figure, the components in the upper layer are omitted and components in an intermediate layer (e.g., intermediate interconnects 42 and 46 ) are shown on the top surface. In a lower part of the figure, the components in the upper layer and the intermediate layer are omitted and components in a lower layer (e.g., a stacked body 20 ) are shown on the top surface. Components seen from the upward direction are indicated by solid lines. Components in the immediately below stage hidden by the components are indicated by broken lines. Components below the components are omitted. The same applies to FIGS. 14 and 15 referred to below.

First, the schematic configuration of the semiconductor memory device according to the embodiment is described.

As shown in FIGS. 1 to 3 , a silicon substrate 10 is provided in a semiconductor memory device 1 according to the embodiment.

In the following description, for convenience of description, in the specification, an XYZ orthogonal coordinate system is adopted. Two directions parallel to the upper surface of the silicon substrate 10 and orthogonal to each other are referred to as “X-direction” and “Y-direction”. A direction perpendicular to the upper surface is referred to as “Z-direction”.

In the semiconductor memory device 1 , a memory region Rm, interconnect lead regions Rd, and peripheral circuit regions Rc are set. In the memory region Rm, a large number of memory cells are disposed three-dimensionally. Note that, in FIG. 1 , illustration of the memory region Rm is omitted. The interconnect lead regions Rd are disposed on both sides in the X-direction of the memory region Rm. The peripheral circuit regions Rc are disposed around the memory region Rm and the interconnect lead regions Rd. Peripheral circuits are formed in the peripheral circuit regions Rc. The peripheral circuits input signals to and output signals from memory cells to thereby operate the memory cells. The interconnect lead regions Rd are regions for drawing out, from the memory region Rm, interconnects extending in the X-direction and connecting the interconnects to the peripheral circuits.

The memory region Rm and the interconnect lead regions Rd are sectioned into a plurality of blocks arrayed along the Y-direction. In FIG. 1 , boundary lines IB among the blocks are indicated by alternate long and two short dashes lines. The same applies to FIG. 14 referred to below.

On the silicon substrate 10 , for example, one lower selection gate interconnect layer 21 , a plurality of word line interconnect layers 22 , and one or more, for example, two upper selection gate interconnect layers 23 are spaced from one another via interlayer insulating films 31 and stacked to configure the stacked body 20 . In the memory region Rm, all layers of the stacked body 20 are disposed. In the interconnect lead regions Rd, the shape of the stacked body 20 is a step shape. The stacked body 20 is drawn further to the outer side toward a lower stage. That is, the word line interconnect layers 22 and the upper selection gate interconnect layers 23 are not disposed in regions right above both end portions in the X-direction of the lower selection gate interconnect layer 21 . In regions right above both end portions in the X-direction of a certain word line interconnect layer 22 , the word line interconnect layers 22 in upper stages of the word line interconnect layer 22 and the upper selection gate interconnect layers 23 are not disposed. In regions right above both end portions in the X-direction of the word line interconnect layers 22 in the top stage, the upper selection gate interconnect layers 23 are not disposed.

In the stacked body 20 , a plurality of linear memory trenches MT extending in the X-direction are formed. The memory trenches MT pierce through the stacked body 20 in the Z-direction and reach the silicon substrate 10 . The memory trenches MT also pierce through the stacked body 20 in the X-direction. Therefore, the memory trenches MT divide the upper selection gate interconnect layers 23 , the word line interconnect layers 22 , and the lower selection gate interconnect layer 21 .

Portions sandwiched by the memory trenches MT adjacent to each other in the upper selection gate interconnect layers 23 are upper selection gate lines 23 a extending in the X-direction. Concerning the word line interconnect layers 22 and the lower selection gate interconnect layer 21 , similarly, portions sandwiched by the memory trenches MT are respectively word lines 22 a and lower selection gate lines 21 a.

In the memory trenches MT, a plurality of silicon pillars 26 are arrayed in rows along the X-direction. Therefore, the lower selection gate lines 21 a, the word lines 22 a, and the upper selection gate lines 23 a are disposed on both sides in the Y-direction of the silicon pillars 26 . The shape of the silicon pillars 26 is, for example, a quadratic prism shape extending in the Z-direction. The lower ends of the silicon pillars 26 are connected to the silicon substrate 10 . Insulating members 32 made of, for example, silicon oxide are provided among the silicon pillars 26 in the memory trenches MT.

In the memory trenches MT, through-holes SH extending in the Z-direction are discontinuously formed in the X-direction. For example, through-hole SH in one place is formed for every ten silicon pillars 26 . The through-holes SH pierce through the stacked body 20 in the Z-direction and reach the silicon substrate 10 . In the through-holes SH, insulating members 33 made of, for example, silicon oxide are embedded in the through-holes SH.

The memory region Rm is described.

Bit line plugs 27 are provided on the silicon pillars 26 . A plurality of bit lines 28 extending in the Y-direction are provided on the bit line plugs 27 . The silicon pillars 26 are connected to the bit lines 28 via the bit line plugs 27 . The bit lines 28 are connected to, for example, a sense amplifier of the peripheral circuit. On the stacked body 20 , an interlayer insulating film 30 made of, for example, silicon oxide is provided. The bit line plugs 27 and the bit lines 28 are embedded in the interlayer insulating film 30 .

FIGS. 4A and 4B are partial sectional views showing small portions including one silicon pillar 26 and the through-hole SH in one place in th semiconductor memory device 1 .

As shown in FIGS. 4A and 4B , a tunnel insulating film 36 is provided around the silicon pillar 26 when viewed from the Z-direction. The tunnel insulating film 36 is a film that feeds a tunnel current when a predetermined voltage within a range of a driving voltage of the semiconductor memory device 1 is applied to the tunnel insulating film 36 . The tunnel insulating film 36 is, for example, a silicon oxide film of a single layer or a three-layer film consisting of a silicon oxide layer, a silicon nitride layer, and a silicon oxide layer.

A floating gate electrode 29 is provided between the tunnel insulating film 36 and the word line 22 a. The floating gate electrode 29 is a conductive member that accumulates electric charges and is formed of, for example, polysilicon (Si).

On a region excluding a region in contact with the tunnel insulating film 36 on the surface of the floating gate electrode 29 , a high dielectric constant layer 37 a made of a high dielectric constant material such as silicon nitride (SiN), hafnium oxide (HfO 2 ), or aluminum oxide (Al 2 O 3 ) is provided. In the high dielectric constant layer 37 a, metal such as ruthenium (Ru) may be contained. The shape of the high dielectric constant layer 37 a is a cup shape surrounding the floating gate electrode 29 .

On the other hand, on a side surface on the floating gate electrode 29 side in the word line 22 a and the upper surface and the lower surface of the word line 22 a, a high dielectric constant layer 37 c made of a high dielectric constant material is formed. On the high dielectric constant layer 37 c, a silicon oxide layer 37 b made of silicon oxide (SiO 2 ) is formed. The shape of the silicon oxide layer 37 b and the high dielectric constant layer 37 c in a YZ cross section is a C shape including the word line 22 a.

Consequently, the high dielectric constant layer 37 a, the silicon oxide layer 37 b, and the high dielectric constant layer 37 c are stacked in this order between the floating gate electrode 29 and the word line 22 a. A block insulating film 37 is configured by the high dielectric constant layer 37 a, the silicon oxide layer 37 b, and the high dielectric constant layer 37 c. The block insulating film 37 is a film that does not substantially feed an electric current even if a voltage within the range of the driving voltage of the semiconductor memory device 1 is applied to the block insulating film 37 .

In the word line 22 a, a main body section 24 made of, for example, tungsten (W) is provided. On the surface of the main body section 24 , a barrier metal layer 25 made of, for example, titanium nitride (TiN) is provided. The same applies to the upper selection gate line 23 a and the lower selection gate lines 21 a. The insulating member 33 is opposed the word lines 22 a not via the block insulating film 37 . The insulating member 33 is opposed the insulating member 32 via the silicon oxide layer 37 b and the high dielectric constant layer 37 c, and not via the high dielectric constant layer 37 a.

In the semiconductor memory device 1 , a transistor including one floating gate electrode 29 is formed in each of intersection portions of the silicon pillars 26 and the word lines 22 a. The transistor functions as a memory cell. A NAND string in which a plurality of memory cells are connected in series is connected between the bit lines 28 and the silicon substrate 10 .

The interconnect lead region Rd is described.

As shown in FIGS. 1 and 2 , contacts 41 are provided on end portions of the upper selection gate lines 23 a of the upper selection gate interconnect layers 23 . Intermediate interconnects 42 are provided on the contacts 41 . Vias 43 are provided on the intermediate interconnects 42 . Upper layer interconnects 44 extending in the X-direction are provided on the vias 43 . Consequently, the upper selection gate lines 23 a are connected to the upper layer interconnects 44 via the contacts 41 , the intermediate interconnects 42 , and the vias 43 . In this way, after being drawn out upward (in the Z-direction) in a route of (the contacts 41 —the intermediate interconnects 42 —the vias 43 ), the upper selection gate lines 23 a are drawn out in the X-direction by the upper layer interconnects 44 and connected to the peripheral circuits. Therefore, the upper selection gate lines 23 a can be driven independently from one another.

Contacts 45 are provided on one end portions of the word lines 22 a of the word line interconnect layers 22 . End portions where the contacts 45 are provided among both end portions in the X-direction of the word lines 22 a are alternate in a continuously arrayed plurality of word lines 22 a. That is, concerning two word lines 22 a adjacent to each other, the contacts 45 are provided on end portions different from each other in the X-direction. Intermediate interconnects 46 extending in the Y-direction are provided on the contacts 45 . Consequently, every other word lines 22 a are connected to the intermediate interconnects 46 via the contacts 45 in common. The same applies to the lower selection gate interconnect layer 21 . The intermediate interconnects 46 extend in the entire length in the Y-direction in one block and are connected to every other word lines 22 a among the word lines 22 a belonging to the word line interconnect layers 22 . Vias 47 are provided on the intermediate interconnects 46 . Upper layer interconnects 48 extending in the X-direction are provided on the vias 47 . Consequently, in the blocks, all the word lines 22 a belonging to the word line interconnect layers 22 are connected to one upper layer interconnect 48 via the contacts 45 , the intermediate interconnects 46 , and the vias 47 . In this way, after being respectively drawn out upward (in the Z-direction) by the contacts 45 , every other word lines 22 a belonging to a certain word line interconnect layer 22 are bound into one by the intermediate interconnects 46 extending in the Y-direction. Consequently, a plurality of word lines 22 a belonging to a certain word line interconnect layer 22 are connected in a pair of comb shapes disposed in positions where the comb shapes mesh with each other. After being further drawn out upward by the vias 47 , the intermediate interconnects 46 are drawn out in the X-direction by the upper layer interconnects 48 and connected to the peripheral circuits. Similarly, in the blocks, every other lower selection gate lines 21 a are bound by the contacts 45 and the intermediate interconnects 46 and then drawn out in the X-direction by one upper layer interconnect 48 and connected to the peripheral circuits.

On the other hand, the intermediate interconnects 46 do not extend exceeding the boundary lines IB of the blocks. The word lines 22 a belonging to different blocks are not connected. The lower selection gate lines 21 a belonging to different blocks are not connected. Therefore, the lower selection gate interconnect layer 21 and the word line interconnect layers 22 can be driven for each of the blocks.

As described above, the shape of both end portions in the X-direction of the stacked body 20 is the step shape. Therefore, the contacts 45 connected to end portions in the X-direction of the word line interconnect layers 22 are present further on the outer side, that is, a side farther from the memory region Rm than the contacts 41 connected to end portions in the X-direction of the upper selection gate interconnect layers 23 . The contacts 45 connected to the end portions in the X-direction of the word line interconnect layers 22 , in lower layers among the word line interconnect layers 22 , are located further on the outer side. A contact connected to an end portion in the X-direction of the lower selection gate interconnect layer 21 is present further on the outer side than the contacts 45 . In other words, the distance between an end portion in the X-direction of the word line interconnect layer 22 , in the upper layer of the word line interconnect layers 22 in certain two layers and the silicon pillar 26 closest to the end portion is shorter than the distance between an end portion in the X-direction of the word line interconnect layer 22 in the lower layer and the silicon pillar 26 closest to the end portion.

A manufacturing method for the semiconductor memory device according to the embodiment is described.

FIGS. 5A to 13B are sectional views showing the manufacturing method for the semiconductor memory device according to the embodiment.

FIG. 5B shows a cross section taken along line C-C′ in FIG. 5A . FIG. 5A shows a cross section taken along line D-D′ in FIG. 5B . The same applies to FIGS. 6A to 13B .

First, as shown in FIGS. 5A and 5B , the interlayer insulating films 31 and a sacrificial film 51 are alternately formed on the silicon substrate 10 (see FIG. 2 ) to form the stacked body 20 . For example, the interlayer insulating films 31 are formed of silicon oxide. The sacrificial film 51 is formed of silicon nitride.

Subsequently, as shown in FIGS. 6A and 6B , a hard mask for MT (not shown in the figure) in a line-and-space shape extending in the X-direction is formed on the stacked body 20 by a lithography method. By applying anisotropic etching such as RIE (Reactive Ion Etching) using the hard mask for MT as a mask, the memory trench MT extending in the X-direction is formed in the stacked body 20 . The memory trench MT is made to pierce through the stacked body 20 in the Z-direction to reach the silicon substrate 10 (see FIG. 2 ). Subsequently, for example, silicon oxide is deposited in the memory trench MT and the insulating member 32 is embedded in the memory trench MT.

Subsequently, as shown in FIGS. 7A and 7B , a hard mask for AA (not shown in the figure) in a line-and-space shape extending in the Y-direction is formed on the stacked body 20 . When viewed from the Z-direction, the hard mask for AA and the hard mask for MT (not shown in the figure) are arrayed in a lattice shape. By applying RIE using the hard mask for AA (not shown in the figure) and the hard mask for MT (not shown in the figure) as masks, a plurality of memory holes MH are formed in the insulating member 32 in a matrix shape. The memory holes MH are made to pierce through the stacked body 20 in the Z-direction to reach the silicon substrate 10 .

Subsequently, as shown in FIGS. 8A and 8B , by applying wet etching using, for example, hot phosphoric acid via the memory hole MH, the sacrificial film 51 exposed on the side surface of the memory hole MH is etched back to form recessed portions 57 on the side surface of the memory hole MH. At this point, since the insulating member 32 is substantially not etched, the recessed portions 57 are not formed in both sides in the X-direction of the memory hole MH and are formed only on both sides in the Y-direction. The recessed portions 57 in a plurality of stages are arrayed along the Z-direction. Subsequently, by performing oxidization treatment, the sacrificial film 51 exposed on the side surfaces of the recessed portions 57 is oxidized to form cover oxidized films 59 .

Subsequently, as shown in FIGS. 9A and 9B , by depositing a high dielectric constant material containing metal, high dielectric constant layers 37 a are formed on the inner surfaces of the memory holes MH and the recessed portions 57 . Subsequently, by depositing silicon, silicon films are formed on the high dielectric constant layers 37 a. Subsequently, by applying wet etching using TMY (choline water solution) or isotropic etching such as CDE (chemical dry etching), the silicon films and the high dielectric constant layers 37 a are recessed and portions deposited on the outside of the recessed portions 57 in the silicon films and the high dielectric constant layers 37 a are removed to leave silicon films and the high dielectric constant layers 37 a on the insides of the recessed portions 57 . Consequently, the high dielectric constant layers 37 a are formed on the inner surfaces of the recessed portions 57 and the floating gate electrodes 29 made of silicon are formed in the recessed portions 57 .

Subsequently, as shown in FIGS. 10A and 10B , for example, silicon oxide is deposited on the inner surfaces of the memory holes MH by, for example, an ALD (Atomic Layer Deposition) method to form the tunnel insulating films 36 . Subsequently, cover silicon films are formed on the side surfaces of the tunnel insulating films 36 . Subsequently, the cover silicon film and the tunnel insulating film 36 deposited on the bottom surface of the memory hole MH are removed by the RIE method to expose the silicon substrate 10 (see FIG. 2 ). Subsequently, silicon is deposited in the memory trench MT to embed a body silicon film. The body silicon film is in contact with the silicon substrate 10 . The silicon pillar 26 is formed by the cover silicon films and the body silicon film. Subsequently, heat treatment for crystallizing the silicon is applied. In this way, the tunnel insulating films 36 are formed on the inner side surfaces of the memory hole MH. The silicon pillar 26 is formed to be surrounded by the tunnel insulating films 36 .

Subsequently, as shown in FIGS. 11A and 11B , the through-hole SH extending in the Z-direction and piercing through the stacked body 20 is formed by the lithography method in a part of a portion where the silicon pillar 26 is not formed in the memory trench MT. The through-hole SH is formed for each fixed number of silicon pillars 26 in the memory trenches MT. The length in the Y-direction of the through-hole SH is set larger than the width of the memory trench MT to protrude both end portions in the Y-direction of the through-hole SH from the memory trench MT. Consequently, on the inner surface of the through-hole SH, the inter-layer insulating films 31 and the sacrificial film 51 are exposed besides the insulating member 32 .

Subsequently, as shown in FIGS. 12A and 12B , wet etching by, for example, hot phosphoric acid is applied using the cover oxide films 59 as etching stoppers. Consequently, the sacrificial film 51 (see FIGS. 11A and 11B ) is removed via the through-hole SH. At this point, the interlayer insulating film 31 , the insulating member 32 , and the cover oxide films 59 made of silicon oxide are substantially not etched. Consequently, a hollow 60 is formed in a space after the removal of the sacrificial film 51 . The upper surface and the lower surface of the hollow 60 are configured by the interlayer insulating film 31 . The side surface of the hollow 60 is configured by the insulating member 32 and the cover oxide films 59 . The hollow 60 is mainly supported by the insulating member 32 and the silicon pillar 26 that pierce through the hollow 60 in the Z-direction. Subsequently, the cover oxide films 59 are removed.

Subsequently, as shown in FIGS. 13A and 13B , by depositing silicon oxide via the through-hole SH with, for example, the ALD method, a silicon oxide layer 37 b is formed on the inner surface of the hollow 60 . Subsequently, by depositing a high dielectric constant material such as silicon nitride (SiN), hafnium oxide (HfO 2 ), or aluminum oxide (Al 2 O 3 ) via the through-hole SH, a high dielectric constant layer 37 c is formed on the silicon oxide layer 37 b. Subsequently, by depositing titanium nitride (TIN) with, for example, the CVD method, a barrier metal layer 25 is formed on the high dielectric constant layer 37 c. Subsequently, by depositing tungsten (W) with, for example, the CVD method, the main body section 24 is embedded in the hollow 60 .

Subsequently, as shown in FIGS. 4A and 4B , the main body section 24 and the barrier metal layer 25 are etched back via the through-hole SH to remove the main body section 24 and the barrier metal layer 25 from the inside of the through-hole SH. As a result, the main body section 24 and the barrier metal layer 25 are divided in each of spaces among the memory trenches MT. Consequently, the upper selection gate lines 23 a, the word lines 22 a, and the lower selection gate lines 21 a are formed by the main body section 24 and the barrier metal layer 25 disposed among the memory trenches MT. Subsequently, silicon oxide is deposited in the through-hole SH to embed the insulating member 33 .

Subsequently, as shown in FIGS. 1 to 3 , both the end portions in the X-direction of the stacked body 20 are processed in a step shape to expose end portions of the upper selection gate lines 23 a, end portions of the word lines 22 a, and end portions of the lower selection gate lines 21 a. Subsequently, for example, silicon oxide is deposited to form the interlayer insulating film 30 that embeds the entire stacked body 20 and, at the same time, form the contacts 41 and 45 , the intermediate interconnects 42 and 46 , the vias 43 , the bit line plugs 27 , the upper layer interconnects 44 , and the bit lines 28 . In this way, the semiconductor memory device 1 according to the embodiment is manufactured.

Effects of the embodiment are described.

In the embodiment, the memory trench MT is formed, the floating gate electrodes 29 are formed on the sides of the memory trench MT, and the tunnel insulating film 36 and the silicon pillar 26 are formed in the memory trench MT. Thereafter, the through-hole SH is formed to overlap the memory trench MT, the sacrificial film 51 is removed via the through-hole SH, and the silicon oxide layer 37 b, the high dielectric constant layer 37 c, the word line 22 a, and the like are formed in the formed hollow 60 . Therefore, compared with the case where the through-hole SH is formed in a position spaced from the memory trench MT, it is possible to reduce an interval among the memory trenches MT and increase the array density of the memory cells in the Y-direction.

In the embodiment, the memory trenches MT divide the upper selection gate interconnect layer 23 into the plurality of upper selection gate lines 23 a. The upper selection gate lines 23 a are connected to the upper layer interconnects 44 via the contacts 41 , the intermediate interconnects 42 , and the vias 43 . Therefore, it is possible to drive the upper selection gate lines 23 a independently from one another.

On the other hand, the memory trenches MT divide the word line interconnect layer 22 into the plurality of word lines 22 a and divide the lower selection gate interconnect layer 21 into the plurality of lower selection gate lines 21 a. However, in the block, the word lines 22 a are connected to one another and the lower selection gate lines 21 a are connected to one another by the intermediate interconnects 46 . Consequently, it is possible to drive the word lines 22 a and the lower selection gate lines 21 a in block units.

Note that, as the array interval of the through-holes SH is set longer, it is possible to increase the number of silicon pillars 26 in the memory trenches MT. It is possible to improve an integration degree of the memory cells. On the other hand, when the array interval of the through-holes SH is too long, it is difficult to remove the sacrificial film 51 via the through-holes SH and form the silicon oxide layer 37 b, the high dielectric layer 37 c, and the barrier metal layer 25 . It is also difficult to embed the main body section 24 . Therefore, the array interval of the through-holes SH is determined taking into account a requested integration degree of the memory cells and requested productivity.

Second Embodiment

A second embodiment is described.

FIG. 14 is a plan view showing a semiconductor memory device according to the embodiment.

FIG. 15 is a sectional view taken along line E-E′ in FIG. 14 .

As shown in FIGS. 14 and 15 , in a semiconductor memory device 2 according to the embodiment, compared with the semiconductor memory device 1 (see FIG. 1 ) according to the first embodiment, the memory trenches MT in the blocks are short. Therefore, the memory trenches MT pierce through the stacked body 20 in the Z-direction but do not completely pierce through the stacked body 20 in the X-direction.

As described above, in the stacked body 20 , the length in the X-direction of the upper selection gate interconnect layers 23 is smaller than the length in the X-direction of the word line interconnect layers 22 and the lower selection gate interconnect layer 21 which are lower layers than the upper selection gate interconnect layers 23 . Therefore, in the embodiment, the memory trenches MT pierce through two upper selection gate interconnect layers 23 in the X-directions but do not pierce through the word line interconnect layers 22 and the lower selection gate interconnect layer 21 in the X-direction. Therefore, the upper selection gate interconnect layers 23 are completely divided by the memory trenches MT. However, the respective end portions in the X-direction of the word line interconnect layers 22 and the lower selection gate interconnect layer 21 are not divided by the memory trenches MT. However, in the boundary lines IB among the blocks, the memory tranches MT completely divide the word line interconnect layers 22 and the lower selection gate interconnect layer 21 .

Therefore, the word interconnect layers 22 and the lower selection gate interconnect layer 21 are processed in a ladder shape in the blocks. That is, in the word line interconnect layers 22 , both end portions of the plurality of word lines 22 a extending in the X-direction are integrally coupled to coupling sections 22 b extending in the Y-direction. The coupling sections 22 b are end portions in the X-direction remaining without being divided by the memory trenches MT in the word line interconnect layers 22 . Similarly, in the lower selection gate interconnect layer 21 , both end portions of the plurality of lower selection gate lines 21 a extending in the X-direction are integrally coupled to the coupling sections 21 b extending in the Y-direction.

In this way, in the embodiment, in the blocks, the plurality of word lines 22 a are connected to one another by the coupling sections 22 b. The plurality of lower selection gate lines 21 a are connected to one another by the coupling sections 21 b. Therefore, it is unnecessary to provide the contacts 45 and the intermediate interconnects 46 for each of the word lines 22 a and the lower selection gate lines 21 a. One contact 61 and one intermediate interconnect 62 only have to be provided for each of the blocks. In the embodiment, the contact 61 is provided in one of two coupling sections 22 a of the word line interconnect layers 22 . The intermediate interconnect 62 is provided on the contact 61 . Lead interconnects 63 are provided at the same height as that of the intermediate interconnect 62 . Consequently, the word line interconnect layers 22 are connected to the lead interconnects 63 via the contact 61 and the intermediate interconnect 62 . The lead interconnects 63 once extend in the Y-direction starting from connection points to the intermediate interconnect 62 and reach above an adjacent block. Thereafter, the lead interconnects 63 bend at a right angle, extend in the X-direction, and move away from the stacked body 20 . The same applies to the lower selection gate interconnect layer 21 . In this way, in the two blocks adjacent to each other, the lower selection gate interconnect layer 21 and the word line interconnect layers 22 are drawn around to the adjacent block by the lead interconnects 63 and then drawn out in a direction away from the stacked body 20 .

In the semiconductor memory device 2 , columns 65 extending in the Z-direction are provided on the silicon substrate 10 to pierce through the coupling section 21 b of the lower selection gate interconnect layer 21 . The columns 65 are formed of a material different from that of the sacrificial film 51 , for example, silicon oxide or metal. The word line interconnect layers 22 and the upper selection gate interconnect layers 23 are not disposed in regions right above the columns 65 .

Further, in the semiconductor memory device 2 , in the blocks of the interconnect lead regions Rd, the through-holes SH are also formed on extended lines of the memory trenches MT, that is, on the outside of the memory trenches MT. The insulating members 33 are also provided in the through-holes SH. In the interconnect lead regions Rd, none of the high dielectric constant layer 37 a, the silicon oxide layer 37 b, the high dielectric constant layer 37 c, and the barrier metal layer 25 is disposed between the insulating member 33 and the main body section 24 of the coupling sections 21 b and 22 b. The insulating member 33 is in contact with the main body section 24 .

According to the embodiment, the memory trenches MT do not completely divide the word line interconnect layers 22 . The end portions in the X-direction of the word lines 22 a are coupled by the coupling section 22 b. Consequently, it is unnecessary to provide the contacts 45 and the intermediate interconnects 46 . As a result, there is room in a space and a degree of freedom of layout of other interconnects is improved.

According to the embodiment, by providing the columns 65 formed of the material different from that of the sacrificial film 51 , as shown in FIGS. 12A and 12B , in a process for removing the sacrificial film 51 via the through-hole SH, the columns 65 are not removed. The columns 65 support the interlayer insulating film 31 . Consequently, it is possible to prevent the hollow 60 from collapsing.

Further, in the interconnect lead regions Rd, by forming the through-holes SH in regions where the memory trenches MT are not formed, it is easy to remove the sacrificial film 51 and form the silicon oxide layer 37 b, the high dielectric constant layer 37 c, the barrier metal layer 25 , and the main body section 24 .

Note that, when the columns 65 are formed of an insulating material, even after the end portion of the stacked body 20 are processed in the step shape, the columns 65 may be disposed in positions where the columns 65 pierce through the coupling sections 22 b of the word line interconnect layers 22 . On the other hand, when the columns 65 are formed of a conductive material, after the end portions of the stacked body 20 are processed in the step shape, it is necessary to dispose the columns 65 in positions where the columns 65 are not in contact with the word line interconnect layers 22 .

A configuration, a manufacturing method, an operation, and an effect in the embodiment other than those described above are the same as those in the first embodiment.

Third Embodiment

A third embodiment is described.

FIG. 16 is a sectional view showing a semiconductor memory device according to the embodiment.

Note that, in FIG. 16 , illustration of the barrier metal layer 25 is omitted.

As shown in FIG. 16 , in a semiconductor memory device 3 according to the embodiment, the silicon pillars 26 are provided not only in the memory region Rm but also in the interconnect lead region Rd. Annular floating gate electrodes 29 a surrounding the silicon pillars 26 are provided around the silicon pillars 26 disposed in the interconnect lead region Rd. On the other hand, in the memory region Rm, the floating gate electrodes 29 are provided in both sides in the Y-direction of the silicon pillars 26 .

In the semiconductor memory device 3 , as in the second embodiment, the through-holes SH are also formed in the interconnect lead region Rd. The insulating members 33 are provided in the through-holes SH. However, unlike the second embodiment, the through-holes SH are disposed in positions deviating from extended lines of the memory trenches MT. Further, the contacts 61 are connected to regions where the silicon pillars 26 and the insulating members 33 are not provided in the coupling sections 22 b of the word line interconnect layers 22 .

In the first embodiment, after the memory trenches MT (see FIG. 6A ) extending in the X-direction are formed and the memory trenches MT are filled with the insulating members 32 , the memory holes MH are formed by applying the etching using the hard mask for AA and the hard mask for MT (not shown in the figure) disposed in the lattice shape as a mask. Consequently, the memory holes MH in a rectangular shape viewed from the Z-direction are formed.

On the other hand, in the embodiment, as shown in FIG. 16 , after the memory trenches MT are formed and the insulating members 32 are embedded in the memory trenches MT, the memory holes MH in a circular shape viewed from the Z-direction are formed by applying the etching using a hard mask for AA (not shown in the figure) in which a plurality of circular openings are formed. At this point, the openings of the hard mask for AA are also formed on the extended lines of the memory trenches MT in the interconnect lead regions Rd. Consequently, the semiconductor memory device 3 according to the embodiment is manufactured.

According to the embodiment, by providing the silicon pillars 26 in the interconnect lead region Rd as well, in the process for forming the hollow 60 shown in FIGS. 12A and 12B , it is possible to support the hollow 60 with the silicon pillars 26 . Therefore, unlike the second embodiment, the columns 65 are unnecessary. Note that the silicon pillars 26 and the floating gate electrodes 29 a formed in the interconnect lead region Rd do not configure memory cells.

In the semiconductor memory device 3 , contacts 67 are provided in the memory trenches MT provided along the boundary lines IB of the blocks. The contacts 67 pierce through the insulating members 32 . Lower ends of the contacts 67 are connected to the silicon substrate 10 (see FIG. 2 ). A common source line 68 extending in the X-direction is provided on the contacts 67 and connected to the upper ends of the contacts 67 . Consequently, it is possible to apply source potential to the silicon substrate 10 via the common source line 68 and the contacts 67 .

In the embodiment, compared with the second embodiment, it is possible to omit a process for forming the columns 65 . Therefore, productivity is high.

A configuration, a manufacturing method, an operation, and an effect other than those in the embodiment are the same as those in the second embodiment.

According to the embodiments described above, it is possible to realize a semiconductor memory device having a high integration degree and a manufacturing method for the semiconductor memory device.

Note that, in the first to third embodiments, the example is described in which the memory holes MH are formed, the tunnel insulating films 36 are formed on the inner surfaces of the memory holes MH, and the silicon pillars 26 are formed on the insides of the memory holes MH, that is, on the side surfaces of the tunnel insulating films 36 . However, not only this, but, for example, in the first and second embodiments, tunnel insulating films and silicon films may be formed on both side surfaces of the memory trenches MT. Thereafter, silicon pillars may be formed by dividing the tunnel insulating films and the silicon films along the X-direction.

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 embodiments described herein may be embodied in a variety of other forms; furthermore, various omissions, substitutions and changes in the form of the embodiments 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 invention. Additionally, the embodiments described above can be combined mutually.