Method of manufacturing semiconductor device

[0001] 1. Field of the Invention

[0002] The present invention relates to a method of manufacturing a semiconductor device including a memory area and a logic circuit area. More specifically the invention pertains to a method of manufacturing a semiconductor device, on which each of non-volatile memory devices formed in the memory area has two charge accumulation regions relative to one word gate.

[0003] 2. Description of the Related Art

[0004] One type of non-volatile semiconductor memory devices is MONOS (Metal Oxide Nitride Oxide Semiconductor) or SONOS (Silicon Oxide Nitride Oxide Silicon), in which a gate insulating layer between a channel area and a control gate is a multi-layered body of a silicon oxide layer and a silicon nitride layer and charges are trapped by the nitride silicon layer.

[0005] FIG. 20 shows a known MONOS non-volatile semiconductor memory device (refer to: Y. Hayashi, et al., 2000 Symposium on VLSI Technology Digest of Technical Papers p.122-123).

[0006] Each MONOS memory cell 100 has a word gate 14, which is formed on a semiconductor substrate 10 via a first gate insulating layer 12. A first control gate 20 and a second control gate 30 are formed as side walls on both sides of the word gate 14. A second gate insulating layer 22 is present between the bottom of the first control gate 20 and the semiconductor substrate 10. An insulating layer 24 is present between the side face of the first control gate 20 and the word gate 14. Similarly the second gate insulating layer 22 is present between the bottom of the second control gate 30 and the semiconductor substrate 10. The insulating layer 24 is present between the side face of the second control gate 30 and the word gate 14. Impurity layers 16 and 18, each of which constitutes either a source area or a drain area, are formed in the semiconductor substrate 10 to be located between the control gate 20 and the control gate 30 of adjoining memory cells.

[0007] Each memory cell 100 accordingly has two MONOS memory elements on the side faces of the word gate 14. These two MONOS memory elements are controlled independently. Namely each memory cell 100 is capable of storing 2-bit information.

[0008] A memory area including such MONOS memory cells and a logic circuit area including peripheral circuits of memories are formed on an identical semiconductor substrate in a semiconductor device. A prior art method of manufacturing such a semiconductor device first forms memory cells in the memory area and subsequently forms peripheral circuits in the logic circuit area. The manufacturing method forms diverse wiring layers via an insulating layer, after formation of the memory area and the logic circuit area.

[0009] The manufacturing method forms an insulating layer of, for example, silicon oxide, and polishes the insulating layer by CMP (chemical mechanical polishing) technique.

[0010] It is desirable that the upper face of the insulating layer after the polishing process is sufficiently flat and even, in order to carry out subsequent processes with high accuracy, for example, in order to form a wiring layer above the insulating layer with high accuracy. The polishing rate of the insulating layer is, however, not constant but is varied. The insulating layer in the logic circuit area is often polished relatively faster than the insulating layer in the memory area. This may cause unevenness on the upper face of the polished insulating layer.

[0011] The object of the present invention is thus to provide a technique of making surface of an insulating layer sufficiently flat and even after a polishing process in a manufacturing method of a semiconductor device including a memory area and a logic circuit area.

[0012] In order to attain at least part of the above and the other related objects, the present invention is directed to a first method of manufacturing a semiconductor device, which includes a memory area having a non-volatile memory device and a logic circuit area having a peripheral circuit of the non-volatile memory device. The first manufacturing method includes the steps of: providing a semiconductor substrate, which has an element separating region formed on surface of a semiconductor layer to attain insulation between semiconductor elements, a first conductive layer formed above the semiconductor layer and patterned to give a word gate of the non-volatile memory device, a stopper layer formed above the first conductive layer, and control gates formed as side walls via an ONO membrane on both side faces of the first conductive layer in the memory area; patterning the first conductive layer in the logic circuit area to create a gate electrode of an insulated gate field effect transistor, which constructs the peripheral circuit, in the logic circuit area and to create a dummy gate electrode above the element separating region in the logic circuit area; forming an insulating layer over whole surface of the memory area and the logic circuit area on the semiconductor substrate; and polishing the insulating layer to make the stopper layer in the memory area exposed.

[0013] The first manufacturing method of the invention creates the dummy gate electrodes above the element separating region in the logic circuit area and thereby makes the formation density of the gate electrodes and the dummy gate electrodes in the logic circuit area approximate to the formation density of the first conductive layers in the memory area. The density of occurrence of irregularities on the surface of the insulating layer in the logic circuit area thus approaches to the density of occurrence of irregularities on the surface of the insulating layer in the memory area. There is accordingly a less difference in height of the surface of the insulating layer between the memory area and the logic circuit area. This arrangement effectively reduces the unevenness on the surface of the insulating layer, which may arise in the process of polishing the insulating layer, and makes the polished insulating layer sufficiently flat and even.

[0014] The present invention is also directed to a second method of manufacturing a semiconductor device, which includes a memory area having a non-volatile memory device and a logic circuit area having a peripheral circuit of the non-volatile memory device. The second manufacturing method includes the steps of: forming an element separating region on surface of a semiconductor layer to attain insulation between semiconductor elements; forming a first insulating layer above the semiconductor layer; forming a first conductive layer above the first insulating layer; forming a stopper layer above the first conductive layer; patterning the stopper layer and the first conductive layer in the memory area; forming an ONO membrane over whole surface of the memory area and the logic circuit area; forming a second conductive layer above the ONO membrane; and carrying out anisotropic etching of the second conductive layer, so as to form control gates as side walls via the ONO membrane on both side faces of the first conductive layer in at least the memory area. The second manufacturing method further includes the steps of: etching out the stopper layer in the logic circuit area; patterning the first conductive layer in the logic circuit area to create a gate electrode of an insulated gate field effect transistor, which constructs the peripheral circuit, in the logic circuit area and to create a dummy gate electrode above the element separating region in the logic circuit area; forming side wall insulating layers on at least both side faces of the gate electrode; forming a first impurity layer as either one of a source area and a drain area of the non-volatile memory device and a second impurity layer as either one of a source area and a drain area of the insulated gate field effect transistor; forming a silicide layer on surface of the first impurity layer, the second impurity layer, and the gate electrode; forming a second insulating layer over whole surface of the memory area and the logic circuit area; polishing the second insulating layer to make the stopper layer in the memory area exposed; etching out the stopper layer in the memory area; and patterning the first conductive layer in the memory area, so as to create a word gate of the non-volatile memory device in the memory area.

[0015] Like the first manufacturing method, the second manufacturing method of the invention effectively reduces the unevenness on the surface of the second insulating layer, which may arise in the process of polishing the second insulating layer, and makes the surface of the polished second insulating layer sufficiently flat and even.

[0016] The above and other objects, features, aspects, and advantages of the present invention will become more apparent from the following detailed description of the preferred embodiment with the accompanying drawings.

[0017] FIG. 1 is a plan view schematically illustrating the layout of a memory area in a semiconductor device;

[0018] FIG. 2 is another plan view schematically illustrating the layout of the memory area in the semiconductor device;

[0019] FIG. 3 is a plan view schematically illustrating a main part of the semiconductor memory device;

[0020] FIG. 4 is a sectional view taken on the line A-A in FIG. 3;

[0021] FIG. 5 is a sectional view illustrating one process in a manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0022] FIG. 6 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0023] FIG. 7 is a plan view showing one process in the manufacturing method of the semiconductor device shown in FIG. 6;

[0024] FIG. 8 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0025] FIG. 9 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0026] FIG. 10 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0027] FIG. 11 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0028] FIG. 12 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0029] FIG. 13 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0030] FIG. 14 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0031] FIG. 15 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0032] FIG. 16 is a sectional view illustrating one process in the manufacturing method of the semiconductor device shown in FIGS. 1 through 4;

[0033] FIGS. 17(A) and 17(B) show the characteristics of the polishing process by CMP technique;

[0034] FIG. 18 is a sectional view illustrating one process in a manufacturing method of a semiconductor device in one embodiment of the present invention;

[0035] FIG. 19 is a sectional view illustrating one process in the manufacturing method of the semiconductor device in the embodiment of the present invention; and

[0036] FIG. 20 is a sectional view illustrating a known MONOS memory cell.

[0037] FIGS. 1 and 2 are plan views illustrating the layout of a memory area in a semiconductor device manufactured according to a manufacturing method in one embodiment of the present invention. FIG. 3 is a plan view illustrating part of the semiconductor device manufactured according to the manufacturing method in the embodiment of the present invention. FIG. 4 is a sectional view taken on the line A-A in FIG. 3.

[0038] The semiconductor device shown in FIGS. 1 through 4 has a memory area 1000 and a logic circuit area 2000. The memory area 1000 has a memory cell array, in which MONOS non-volatile memory devices (hereafter referred to as memory cells) 100 are arranged in lattice of multiple rows and multiple columns. The logic circuit area 2000 includes peripheral circuits of memories.

[0039] A. Structure of Device

[0040] The layout of the memory area 1000 is discussed first with reference to FIGS. 1 and 2.

[0041] FIG. 1 shows a first block B1 and a second block B2 adjoining to the first block B1 as part of the memory area 1000. FIG. 2 shows the contact structure of the first block B1 with the second block B2.

[0042] An element separating region 300 is formed in a partial area between the first block B1 and the second block B2. Multiple word lines 50 (WL) extending in a direction X (in a row direction) and multiple bit lines 60 (BL) extending in a direction Y (in a column direction) are arrayed in each block B1 or B2. Each of the word lines 50 is connected to multiple word gates 14 arranged in the direction X. The bit lines 60 are composed of impurity layers 16 and 18.

[0043] Conductive layers 40 are formed to surround the respective impurity layers 16 and 18 and constitute first and second control gates 20 and 30. The first and the second control gates 20 and 30 respectively extend in the direction Y. The respective one ends of each pair of the first and the second control gates 20 and 30 are connected with each other via the conductive layer 40 extending in the direction X. The respective other ends of each pair of the first and the second control gates 20 and 30 are linked with one common contact element 200. The first and the second control gates 20 and 30 accordingly have general functions as the control gate of the memory cell and wiring functions of connecting the paired control gates arranged in the direction Y.

[0044] Each memory cell 100 has one word gate 14, the first and the second control gates 20 and 30 arranged on both sides of the word gate 14, and the impurity layers 16 and 18 that are formed in the semiconductor substrate and located outside these control gates 20 and 30. The impurity layers 16 and 18 are shared by the adjoining memory cells 100.

[0045] The two impurity layers 16 adjoining to each other in the direction Y, that is, the impurity layer 16 formed in the block B1 and the impurity layer 16 formed in the adjoining block B2, are electrically connected with each other via a contact impurity layer 400 formed in the semiconductor substrate. The contact impurity layer 400 is located opposite to the common contact element 200 of the control gates across the impurity layer 16.

[0046] A contact 350 is formed on each contact impurity layer 400. The bit lines 60 of the impurity layers 16 are electrically linked with an upper wiring layer via the contacts 350.

[0047] Similarly, the two impurity layers 18 adjoining to each other in the direction Y are electrically connected with each other via the contact impurity layer 400 on the side without the common contact element 200 (see FIG. 2).

[0048] As shown in FIG. 1, the planar layout of the multiple common contact elements 200 in each block has a zigzag pattern, where the common contact elements 200 are arranged alternately on different sides of the impurity layers 16 and 18. Similarly, as shown in FIG. 2, the planar layout of the multiple contact impurity layers 400 in each block has a zigzag pattern, where the contact impurity layers 400 are arranged alternately on different sides of the impurity layers 16 and 18.

[0049] The planar structure and the sectional structure of the semiconductor device are discussed with reference to FIGS. 3 and 4. The logic circuit area 2000 including peripheral circuits of memories is formed adjacent to the memory area 1000. The memory area 1000 is electrically separated from the logic circuit area 2000 by means of the element separating region 300. The memory area 1000 includes at least the multiple memory cells 100. The logic circuit area 2000 includes at least insulated gate field effect transistors (hereafter referred to as MOS transistors) 500 constructing logic circuits.

[0050] The description first regards the memory area 1000.

[0051] As shown in FIG. 4, each memory cell 100 includes the word gate 14 that is formed on a semiconductor substrate 10 via a first gate insulating layer 12, the impurity layers 16 and 18 that are formed in the semiconductor substrate 10 to constitute either a source area or a drain area, and the first and the second control gates 20 and 30 that are formed as side walls along both sides of the word gate 14. Silicide layers 92 are arranged on the top of the impurity layers 16 and 18.

[0052] The first control gate 20 is arranged on the semiconductor substrate 10 via a second gate insulating layer 22 and on one side face of the word gate 14 via a side insulating layer 24. Similarly the second control gate 30 is arranged on the semiconductor substrate 10 via the second gate insulating layer 22 and on the other side face of the word gate 14 via the side insulating layer 24.

[0053] The second gate insulating layer 22 and the side insulating layer 24 are ONO membranes. More specifically, the second gate insulating layer 22 and the side insulating layer 24 are multi-layered membranes including a silicon oxide bottom layer (first silicon oxide layer (O)), a silicon nitride layer (N), and a silicon oxide top layer (second silicon oxide layer (O)).

[0054] The first silicon oxide layer of the second gate insulating layer 22 makes a potential barrier between a channel area and a charge accumulation region.

[0055] The silicon nitride layer of the second gate insulating layer 22 functions as a charge accumulation region for trapping carriers (for example, electrons).

[0056] The second silicon oxide layer of the second gate insulating layer 22 makes a potential barrier between the control gate and the charge accumulation region.

[0057] The side insulating layer 24 electrically separates the word gate 14 from the control gates 20 and 30. In order to prevent a short circuit between the word gate 14 and the first and the second control gates 20 and 30, the upper end of the side insulating layer 24 is located above the upper ends of the control gates 20 and 30 relative to the semiconductor substrate 10.

[0058] The side insulating layer 24 and the second gate insulating layer 22 are produced by the same film forming process and have the identical layer structure.

[0059] An embedded insulating layer 70 is disposed between the first control gate 20 and the second control gate 30 of the adjoining memory cells 100. The embedded insulating layer 70 covers over at least the control gates 20 and 30 to prevent exposure thereof In the concrete structure, the upper face of the embedded insulating layer 70 is located above the upper end of the side insulating layer 24 relative to the semiconductor substrate 10. Such arrangement of the embedded insulating layer 70 ensures the electrical separation of the first and the second control gates 20 and 30 from the word gates 14 and the word lines 50.

[0060] A conductive layer is formed on the common contact element 200 to apply a predetermined potential to the control gates 20 and 30. The common contact element 200 includes a first contact insulating layer 212, a second contact insulating layer 210, a first contact conductive layer 214, a second contact conductive layer 232, a third contact insulating layer 252, and a third contact conductive layer 260.

[0061] The first contact insulating layer 212 is produced by the same manufacturing process as that of the first gate insulating layer 12.

[0062] The second contact insulating layer 210 is produced by the same manufacturing process as that of the second gate insulating layer 22 and the side insulating layer 24. The second contact insulating layer 210 is a multi-layered body including a first silicon oxide layer, a silicon nitride layer, and a second silicon oxide layer.

[0063] The first contact conductive layer 214 is produced by the same manufacturing process as that of the word gate 14. The first contact conductive layer 214 is formed outside the second contact insulating layer 210.

[0064] The second contact conductive layer 232 is formed inside the second contact insulating layer 210. The second contact conductive layer 232 is produced by the same manufacturing process as that of the first and the second control gates 20 and 30 to be integrated with these control gates 20 and 30. Namely the second contact conductive layer 232 and the control gates 20 and 30 are made of an identical material.

[0065] The third contact insulating layer 252 is formed inside the second contact conductive layer 232. The third contact insulating layer 252 is produced by the same manufacturing process as that of side wall insulating layers 152 discussed later.

[0066] The third contact conductive layer 260 is produced by the same manufacturing process as that of the word line 50 and is linked with the first contact conductive layer 214 and the second contact conductive layer 232.

[0067] The MOS transistors 500 are created in the logic circuit area 2000. Each MOS transistor 500 includes a gate electrode 142 that is created on the semiconductor substrate 10 via a third gate insulating film 122, impurity layers 162 and 182 that are formed in the semiconductor substrate 10 to constitute either a source area or a drain area, and side wall insulating layers 152 that are formed along both sides of the gate electrode 142. Silicide layers 192 are arranged on the top of the impurity layers 162 and 182, and a silicide layer 194 is arranged on the top of the gate electrode 142.

[0068] In the logic circuit area 2000, the MOS transistors 500 are covered with an insulating layer 270. This insulating layer 270 is produced by the same manufacturing process as that of the embedded insulating layer 70.

[0069] As shown in FIGS. 3 and 4, a boundary element 140c, which is composed of the same material as that of the word gate 14 and the gate electrode 142, is arranged in a boundary region between the memory area 1000 and the logic circuit area 2000. The boundary element 140c is produced by the same film forming process as that of the word gate 14 and the gate electrode 142. At least part of the boundary element 140c is formed above the element separating region 300.

[0070] A side wall conductive layer 20a, which is composed of the same material as that of the control gates 20 and 30, is arranged on one side face of the boundary element 140c (on the side of the memory area 1000). The side wall conductive layer 20a extends in the direction Y and is electrically connected with the adjoining control gate 30 via the common contact element 200. The side wall conductive layer 20a is not used as the control gate of the memory cell. The electrical connection of the side wall conductive layer 20a with the adjoining control gate 30 causes the electric properties of the control gate 30 adjacent to the side wall conductive layer 20a to be substantially equal to the electric properties of the other control gates.

[0071] A side wall insulating layer 152, which is produced by the same manufacturing process as that of the side wall insulating layers 152 of the MOS transistor 500, is arranged on the other side face of the boundary element 140c (on the side of the logic circuit area 2000).

[0072] An inter-layer insulating layer 72 is formed on the semiconductor substrate 10 with the memory cells 100 and the MOS transistors 500. The inter-layer insulating layer 72 has contact holes, which run to, for example, the third contact conductive layer 260 of the common contact element 200. Each contact hole is filled with a conductive layer 82 of, for example, a tungsten plug or a copper plug, which is connected to a wiring layer 80 located above the inter-layer insulating layer 72.

[0073] B. Basic Manufacturing Method of Semiconductor Device

[0074] Prior to description of a manufacturing method of a semiconductor device in one embodiment of the present invention, a basic manufacturing method is discussed with reference to FIGS. 5 through 16. The respective sectional views of FIGS. 5 through 16 correspond to the part taken on the line A-A in FIG. 3. In the drawings of FIGS. 5 through 16, like elements to those shown in FIGS. 1 through 4 are expressed by like numerals and are not specifically described here.

[0075] (1) Referring to FIG. 5, the method first forms the element separating regions 300 on the surface of the semiconductor substrate 10 by the technique of trench isolation. The method then forms the contact impurity layers 400 (see FIG. 1) in the semiconductor substrate 10 by the technique of ion implantation.

[0076] The method subsequently forms an insulating layer 120, which constructs the gate insulating layers, on the surface of the semiconductor substrate 10, and makes a gate layer 140, which constructs the word gates 14 and the gate electrodes 142, deposit on the insulating layer 120. The insulating layer 120 corresponds to the first insulating layer of the present invention. The gate layer 140 is made of doped polysilicon. A stopper layer S100, which works as an indication of the end of polishing in a later CMP (Chemical Mechanical Polishing) process is further formed on the gate layer 140. The stopper layer S100 is composed of the silicon nitride layer.

[0077] (2) Referring to FIG. 6, a patterned gate layer 140a is formed in the memory area 1000. One concrete procedure for formation of the patterned gate layer 140a forms a resist layer (not shown) on the stopper layer S100 (see FIG. 5) to cover over the whole logic circuit area 2000 and to be extended to part of the memory area 1000. The procedure then patterns the stopper layer S100 with the resist layer as the mask, and etches the gate layer 140 with the patterned stopper layer as the mask. This results in patterning the gate layer 140 to give the patterned gate layer 140a shown in FIG. 6. In this process, the gate layer 140 in the logic circuit area 2000 is not patterned. As a matter of convenience, hereafter the gate layer 140 in the logic circuit area 2000 is called the gate layer 140b.

[0078] FIG. 7 is a plan view showing the state of the memory area 1000 after patterning. The patterning makes openings 160 and 180 in the multi-layered body of the gate layer 140 and the stopper layer S100 in the memory area 1000. The openings 160 and 180 substantially correspond to the regions in which the impurity layers 16 and 18 are formed by a later ion implantation process. The side insulating layers and the control gates are formed afterwards along the side faces of the openings 160 and 180.

[0079] (3) Referring to FIG. 8, an ONO membrane 220 is formed over the whole face of the semiconductor substrate 10. The ONO membrane 220 is obtained by successive deposition of a first silicon oxide layer (O), a silicon nitride layer (N), and a second silicon oxide layer (O). The first silicon oxide layer is formed, for example, by thermal oxidation technique or CVD technique. The silicon nitride layer is formed, for example, by CVD technique. The second silicon oxide layer is formed, for example, by CVD technique or more specifically by high temperature oxidation (HTO) technique. The preferable procedure carries out annealing treatment to densify the respective layers, after formation of these layers.

[0080] A later patterning process of the ONO membrane 220 makes the second gate insulating layer 22, the side insulating layer 24, and the second contact insulating layer 210 (see FIG. 4).

[0081] (4) Referring to FIG. 9, a doped polysilicon layer 230 is made to deposit over the whole surface of the ONO membrane 220. A later etching process of the doped polysilicon layer 230 gives the conductive layer 40 (see FIG. 1) of the control gates 20 and 30 and the second conductive layer 232 (see FIG. 3) of the common contact element 200.

[0082] A resist layer R100 is then formed in the region for the common contact element 200.

[0083] (5) Referring to FIG. 10, anisotropic etching of the whole doped polysilicon layer 230 (see FIG. 9) with the resist layer R100 as the mask gives the first and the second control gates 20 and 30 and the second contact conductive layer 232.

[0084] This etching process makes the control gates 20 and 30 as the side walls along the side faces of the openings 160 and 180 (see FIG. 7) in the memory area 1000. Simultaneously, the second contact conductive layers 232 are formed in the masked parts with the resist layer R100 (see FIG. 9). The doped polysilicon layer 230 depositing in the logic circuit area 2000 is completely removed. In the boundary region, however, the doped polysilicon layer 230 remains as a side wall on the side face of one end of the gate layer 140b (on the side of the memory area 1000). The resist layer R100 is then removed.

[0085] (6) Referring to FIG. 11, a resist layer R200 is then formed to cover over the whole memory area 1000 and to be extended to part of the logic circuit area 2000. The ONO membrane 220 and the stopper layer S100 in the logic circuit area 2000 are removed with the resist layer R200 as the mask. This etching process removes all the stopper layer S100 in the logic circuit area 2000 except the boundary region.

[0086] The part of the gate layer 140b located in the boundary region between the memory area 1000 and the logic circuit area 2000 and covered with both the resist layer used in the etching process (2) and the resist layer R200 used in the etching process (6) forms the boundary element 140c (see FIG. 4) in a later process. A stopper layer S100a remaining through this patterning process has a greater width than the width of the remaining stopper layers S100 in the memory area 1000. The resist layer R200 is removed subsequently.

[0087] (7) Referring to FIG. 12, a resist layer R300 is formed for creation of the gate electrodes 142. The resist layer R300 is patterned to cover over the whole memory area 1000 and a predetermined part in the logic circuit area 2000. Etching of the gate layer 140b (see FIG. 11) with the resist layer R300 as the mask gives the gate electrodes 142 in the logic circuit area 2000. This etching process also gives the boundary element 140c in the boundary region in a self aligning manner with the resist layer R300 and the stopper layer S100a as the mask.

[0088] The resist layer R300 is then removed. Subsequent doping of an N-type impurity creates extension layers 161 and 181 of the source areas and the drain areas in the logic circuit area 2000.

[0089] (8) Referring to FIG. 13, an insulating layer 250 of silicon oxide or silicon oxide nitride is formed over the memory area 1000 and the logic circuit area 2000.

[0090] (9) Referring to FIG. 14, anisotropic etching of the whole insulating layer 250 (see FIG. 13) gives the side wall insulating layers 152 on both sides of each gate electrode 142 in the logic circuit area 2000. Simultaneously, the anisotropic etching gives the side wall insulating layer 152 on one side face of the boundary element 140c facing the logic circuit area 2000. This etching process also makes insulating layers 152a remain on the control gates 20 and 30, and forms the third contact insulating layer 252 covering over the second contact conductive layer 232. The etching process removes the insulating layers depositing on specified regions for formation of silicide layers in a later process and on the gate electrodes 142 in the logic circuit area 2000 to expose the semiconductor substrate 10. The specific regions include, for example, regions for formation of the impurity layers 16 and 18 in the memory area 1000 and regions for formation of the impurity layers 162 and 182 in the logic circuit area 2000 in a later ion implantation process.

[0091] Subsequent implantation of an N-type impurity ion forms the impurity layers 16 and 18, each of which constitutes either a source area or a drain area in the memory area 1000, and the impurity layers 162 and 182, each of which constitutes either a source area or a drain area in the logic circuit area 2000, in the semiconductor substrate 10.

[0092] A subsequent process makes a metal for formation of a silicide deposit on the whole surface. Typical examples of the metal for formation of the silicide are titanium and cobalt. The metal depositing on the impurity layers 16, 18, 162, and 182 and the gate electrodes 142 is subjected to a silicidation reaction. This forms the silicide layers 92 on the top of the impurity layers 16 and 18, the silicide layers 192 on the top of the impurity layers 162 and 182, and the silicide layer 194 on the top of the gate electrodes 142. This silicidation process silicidates the gate electrodes and either the source areas or the drain areas of the MOS transistors 500 (see FIG. 4) in a self aligning manner in the logic circuit area 2000. Simultaneously, the silicidation process silicidates the surface of either the source areas or the drain areas of the memory cells 100 (see FIG. 4) in a self aligning manner in the memory area 1000.

[0093] The insulating layer 270 of silicon oxide or silicon oxide nitride is formed over the whole surface of the memory area 1000 and the logic circuit area 2000. The insulating layer 270 is formed to cover over the stopper layers S100 and S100a.

[0094] (10) Referring to FIG. 15, the insulating layer 270 is polished by the CMP technique to exposure of the stopper layers S100 and S100a and is leveled off. The polishing makes the insulating layer 270 remain between the two side insulating layers 24 facing each other across the control gates 20 and 30 to define the embedded insulating layer 70.

[0095] The upper ends of the side insulating layers 24 formed on the side faces of the gate layer 140a and the stopper layer S100 are located above the upper ends of the first and the second control gates 20 and 30 relative to the semiconductor substrate 10 in the memory area 1000. The MOS transistors 500 are completely covered with the insulating layer 270 in the logic circuit area 2000.

[0096] On completion of this polishing process, the stopper layers S100 and S100a are accordingly present on the gate layer 140a, which constructs the word gates 14, and the boundary element 140c, respectively. No stopper layer is present on the gate electrodes 142, but the gate electrodes 142 are covered with the insulating layer 270.

[0097] (11) The stopper layers S100 and S100a (see FIG. 15) are removed with hot phosphoric acid. This results in exposure of at least the upper faces of the gate layer 140a and the boundary element 140c. A doped polysilicon layer is then made to deposit on the whole surface.

[0098] Referring to FIG. 16, a patterned resist layer R400 is subsequently formed on the depositing doped polysilicon layer. Patterning of the doped polysilicon layer with the resist layer R400 as the mask gives the word lines 50 and the third contact conductive layer 260.

[0099] The gate layer 140a (see FIG. 15) is etched with the resist layer R400 as the mask. The etching removes part of the gate layer 140a where the word lines 50 are not formed thereon. This gives the word gates 14 arranged in an array. The removed part of the gate layer 140a corresponds to the region of a P-type impurity layer (element separating impurity layer) 15 created in a later process (see FIG. 3).

[0100] The conductive layer 40, which constructs the first and the second control gates 20 and 30, is covered with the embedded insulating layer 70 and is thus not etched but remains by this etching process. The MOS transistors 500 in the logic circuit area 2000 are not affected by this etching process, since the MOS transistors 500 are completely covered with the insulating layer 270.

[0101] The whole semiconductor substrate 10 is then doped with a P-type impurity. The P-type impurity layer (element separating impurity layer) 15 (see FIG. 3) is accordingly formed between each pair of the word gates 14 adjoining to each other in the direction Y. The P-type impurity layer 15 ensures separation between the adjoining memory cells 100.

[0102] (12) The process subsequently forms a first inter-layer insulating layer, makes contact holes by any known method, and creates a conductive layer in each contact hole and a first wiring layer. For example, as shown in FIG. 4, the process forms the inter-layer insulating layer 72, makes contact holes in the inter-layer insulating layer 72, and creates the conductive layer 82 and the wiring layer 80 connecting with each contact element 200. This process simultaneously creates contact elements and a wiring layer in the logic circuit area 2000.

[0103] The series of processes discussed above manufactures the semiconductor device shown in FIGS. 1 through 4.

[0104] C. Characteristics of Polishing Process of Insulating Layer by CMP Technique

[0105] FIG. 17 shows the characteristics of the polishing process by the CMP technique discussed above in the process (10). FIG. 17 is a sectional view schematically illustrating the main part of the memory area and the logic circuit area in the semiconductor device. As shown in FIG. 17(A), there are irregularities on the top face of the insulating layer 270, due to varying heights of the elements under the insulating layer 270, that is, the gate layers 140a in the memory area 1000 and the gate electrodes 142 in the logic circuit area 2000. The gate layers 140a are formed at a relatively high density in the memory area 1000, while the gate electrodes 142 are formed at a relatively low density in the logic circuit area 2000. The density of the irregularities on the top face of the insulating layer 270 is thus relatively high in the memory area 1000 and is relatively low in the logic circuit area 2000. Especially the region of the memory area 1000 with the array of multiple memory cells 100 has a higher density of irregularities, compared with the logic circuit area 2000. In the polishing process by the CMP technique, there may be a variation in polishing rate of the insulating layer 270, due to this varying density of the irregularities. More specifically, the insulating layer 270 in the logic circuit area 2000 having a relatively low density of the irregularities is polished faster than the insulating layer 270 in the memory area 1000 having a relatively high density of the irregularities. This causes exposure of the gate electrodes 142 in the logic circuit area 2000, prior to exposure of the stopper layer S100a in the memory area 1000 as shown in FIG. 17(B).

[0106] Exposure of the gate electrodes 142 causes the MOS transistors 500 to be exposed to the etching gas, which may affect the properties of the MOS transistors 500, in the process (11) of patterning the word gates 14 of the memory cells 100.

[0107] The height of the top face of the insulating layer 270 in the region of the memory area 1000 with the array of the multiple memory cells 100, because of its relatively high density of the irregularities, tends to be greater than the height of the top face of the insulating layer in the logic circuit area 2000 having the relatively low density of the irregularities. This makes a difference in height of the surface of the insulating layer 270. The height difference and the variation in polishing rate cause the top face of the insulating layer 270 to be not sufficiently planarized but to be uneven after the polishing process by the CMP technique.

[0108] The non-flat and uneven top face of the insulating layer 270 often leads to difficulties in accurate micro-fabrication of wiring above the insulating layer 270.

[0109] As discussed above, in the polishing process (10) by the CMP technique, the MOS transistors 500 are often exposed to the etching gas, which may affect the properties of the MOS transistors 500. Another problem is difficulties in micro-fabrication of wiring above the insulating layer 270.

[0110] D. Manufacturing Method of Semiconductor Device in Embodiment

[0111] A manufacturing method of a semiconductor device in one embodiment of the present invention carries out the process (7) as shown in FIG. 18, after execution of the processes (1) through (6). The manufacturing method then carries out the process (9) to form the insulating layer 270 as shown in FIG. 19, after execution of the process (8). The manufacturing method subsequently performs the process (10) of polishing the insulating layer 270 by the CMP technique.

[0112] Referring to FIG. 18, the process (7) forms a resist layer R300a to cover not only areas corresponding to the gate electrodes 142 of the MOS transistors 500 in the logic circuit area 2000 but specified areas of the element separating region 300 to create gate electrodes (hereafter referred to as dummy gate electrodes) 142a as dummy elements 600 on the element separating region 300.

[0113] The process etches the gate layer 140b (see FIG. 11) with the resist layer R300a as the mask, so as to create not only the gate electrodes 142 of the MOS transistors 500 but the dummy gate electrodes 142a, which have an identical size with that of the gate electrodes 142, in the areas of the element separating region 300 corresponding to the dummy elements 600. The dummy gate electrodes 142a are formed to make the density of occurrence of irregularities on the surface of the insulating layer 270 in the logic circuit area 2000 approximate to the density of occurrence of irregularities on the surface of the insulating layer 270 in the memory area 1000. More specifically, the dummy gate electrodes 142a are created and arrayed on the element separating region 300, such that the formation density of the gate electrodes 142 of the MOS transistors 500 and the dummy gate electrodes 142a in the logic circuit area 2000 approaches to the formation density of the gate layers 140a in the memory area 1000.

[0114] It is not essential that the dummy gate electrode 142a has the identical size with that of the gate electrode 142 of the MOS transistor 500. For example, the dummy gate electrode 142a may have an identical size with that of the word gate 14 in the memory area 1000. The dummy gate electrodes 142a of the dummy elements 600 may have any size and shape and may be arranged in any arbitrary array, as long as the dummy gate electrodes 142a are formed to make the density of occurrence of irregularities on the surface of the insulating layer 270 in the logic circuit area 2000 approximate to the density of occurrence of irregularities on the surface of the insulating layer 270 in the memory area 1000.

[0115] Referring to FIG. 19, the subsequent processes (7) through (9) complete the dummy elements 600 in the same manner as the MOS transistors 500. Namely each of the dummy elements 600 includes the dummy gate electrode 142a, side wall insulating layers 152a formed on both side faces of the dummy gate electrode 142a, and a silicide layer 194a formed on the dummy gate electrode 142a. The insulating layer 270 is formed over the whole surface of the memory area 1000 and the logic circuit area 2000 including the dummy elements 600 thus produced.

[0116] As discussed above, the dummy gate electrodes 142a are arranged to make the formation density of the gate electrodes 142 of the MOS transistors 500 and the dummy gate electrodes 142a in the logic circuit area 2000 approximate to the formation density of the gate layers 140a in the memory area 1000. This causes the density of occurrence of irregularities on the surface of the insulating layer 270 in the logic circuit area 2000 to be similar to the density of occurrence of irregularities on the surface of the insulating layer 270 in the memory area 1000. This also reduces the difference in height of the surface of the insulating layer 270 between the memory area 1000 and the logic circuit area 2000.

[0117] This arrangement makes the rate of polishing in the logic circuit area 2000 approximate to the rate of polishing in the memory area 1000 in the process (10) of polishing the insulating layer 270 by the CMP technique.

[0118] The manufacturing method of the semiconductor device in the embodiment effectively prevents exposure of the gate electrodes 142 in the logic circuit area 2000, prior to exposure of the stopper layer S100a in the memory area 1000, in the polishing process by the CMP technique.

[0119] The surface of the insulating layer 270 is thus made sufficiently flat and even with higher accuracy in the polishing process (10) by the CMP technique. This ensures accurate micro-fabrication of wiring over the insulating layer 270.

[0120] The above embodiment is to be considered in all aspects as illustrative and not restrictive. There may be many modifications, changes, and alterations without departing from the scope or spirit of the main characteristics of the present invention. Although the embodiment regards the bulk semiconductor substrate, the principle of the invention is also applicable to a semiconductor layer of an SOI substrate.

[0121] The scope and spirit of the present invention are indicated by the appended claims, rather than by the foregoing description.