JP2013110436A - Semiconductor device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve electric characteristics in a semiconductor device employing a split-gate memory cell structure and having a nonvolatile memory using a nitride film as a charge storage layer.SOLUTION: A semiconductor device manufacturing method comprises: forming an n-type semiconductor region 6 on a principal surface of a semiconductor substrate 1Sub; subsequently, forming a memory gate electrode MG and a charge storage layer CSL of a split-gate memory cell on the semiconductor region 6; subsequently, forming side walls 8 on lateral faces of the memory gate electrode MG; subsequently, forming a photoresist pattern PR2 on the principal surface of the semiconductor substrate 1Sub; subsequently, removing a part of the semiconductor substrate 1Sub by etching by using the photoresist pattern PR2 as an etching mask to form a dent 13; removing the n-type semiconductor region 6 in a formation region of the dent 13; and subsequently, forming a p-type semiconductor region for forming a channel of an nMIS for memory cell selection in the formation region of the dent 13.

Description

  The present invention relates to a method of manufacturing a semiconductor device and a semiconductor device technology, and more particularly to a method of manufacturing a semiconductor device having a nonvolatile memory and a technology effective when applied to the semiconductor device.

  Electrically rewritable non-volatile memory such as EEPROM (Electrically Erasable Programmable Read Only Memory) and flash memory can be rewritten on-board, which shortens the development period and improves development efficiency. In addition, the application is expanding to various applications such as small volume, high-mix production, tuning by destination, and program update after shipment. In particular, in recent years, there is a great need for a microcomputer incorporating an MPU (Micro Processing Unit) and an EEPROM (or flash memory). As an electrically rewritable nonvolatile memory, an EEPROM using a normal polysilicon as a floating electrode is mainly used. However, in the EEPROM with this structure, if there is a defect in any part of the oxide film surrounding the floating gate electrode, the charge storage layer is a conductor, so all charges stored in the storage node are lost due to abnormal leakage. May end up. In particular, it is considered that this problem will become more prominent when miniaturization progresses and the degree of integration improves.

Therefore, in recent years, attention has been paid to a MNOS (Metal Nitride Oxide Semiconductor) structure or a MONOS (Metal Oxide Nitride Oxide Semiconductor) structure using a nitride film (Si ₃ N ₄ or the like) as a charge storage layer. In this case, the charge that contributes to data storage is accumulated in the discrete trap of the nitride film, which is an insulator. Therefore, even if a defect occurs in some part of the oxide film surrounding the accumulation node and an abnormal leak occurs, the charge Since all the charges in the accumulation layer are not lost, the reliability of data retention can be improved.

  As a configuration of the memory cell, a memory cell having a single transistor structure has been proposed. As a writing / erasing method, writing is performed by whole surface FN (Fowler Nordheim) tunneling injection from a semiconductor substrate, erasing by FN tunneling current to the semiconductor substrate, writing by hot electron injection, semiconductor substrate or source, drain A method of performing erasing by FN tunneling current to the region has been proposed. Further, since the MONOS type single transistor cell structure is more susceptible to disturbance than the EEPROM cell structure, a split gate type memory cell structure having a two-transistor structure having a control gate electrode has also been proposed. This type of split gate type memory cell structure has a memory gate using a control gate electrode mount type, a memory gate electrode mount type, and a sidewall method, depending on the process of which one of the gate electrodes runs over the other gate electrode. An electrode-mounted type can be realized.

  For example, in Japanese Patent Laid-Open No. 6-85251, after forming a gate electrode of a MOS / FET that is not a memory cell, a surface layer portion of a semiconductor substrate located on both sides of the gate electrode is etched to form a channel region and a source A technique for forming a step with the drain region is disclosed (for example, Patent Document 1).

  Further, for example, in Japanese Laid-Open Patent Publication No. 5-267250, a side wall is formed on the side surface of an adjacent gate electrode (floating gate) in a manufacturing process of a flash memory, and then the lower field A technique for forming a groove by partially etching an insulating film and a semiconductor substrate is disclosed (for example, Patent Document 2).

  Further, for example, in Japanese Patent Laid-Open No. 2000-91452, data is not written to or read from a non-selected memory cell transistor adjacent to a selected memory cell transistor during operation of a non-volatile memory having a MONOS structure. Thus, a configuration is disclosed in which an offset sidewall is provided on the channel side surface of the insulating film pattern on the source and drain buried regions of the memory cell transistor so that the channel of the memory cell transistor is difficult to expand ( For example, Patent Document 3).

  For example, JP-A-6-125094 discloses a configuration in which an offset region is provided between an ONO film and a gate electrode of a memory cell transistor and a source region (for example, Patent Document 4).

  For example, Japanese Patent Laid-Open No. 2001-168219 discloses a configuration in which a step is provided in the channel of a MONOS type memory cell transistor and channel hot electron injection is performed through this step. Also, paragraph number 0038 of this publication discloses that the gate electrode of a memory cell transistor which performs erasing by extracting charges using FN tunnel current is made conductive by introducing a high concentration of p-type or n-type impurities. (For example, Patent Document 5).

  Further, for example, in Japanese Patent Application Laid-Open No. 2002-298591, in order to suppress or prevent the destruction of data by increasing the threshold voltage of non-selected memory cells by repeating the read operation of the EEPROM, A configuration is disclosed in which the threshold voltage of a cell is set higher than the lower one of the voltages applied to the source and drain during data read and lower than the voltage applied to the gate electrode during data read. (For example, Patent Document 6).

JP-A-6-85251 Japanese Patent Laid-Open No. 5-267250 JP 2000-91452 A JP-A-6-125094 JP 2001-168219 A Japanese Patent Laid-Open No. 2002-298591

  However, a nonvolatile memory using a split gate type memory cell structure as described above and using a nitride film as a charge storage layer is a new structure, and in a semiconductor device having the nonvolatile memory, how are the electrical characteristics? One of the important issues is how to improve this.

  An object of the present invention is to provide a technique capable of improving electrical characteristics in a semiconductor device having a nonvolatile memory that employs a split gate type memory cell structure and uses a nitride film as a charge storage layer.

  The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  Of the inventions disclosed in the present application, the outline of typical ones will be briefly described as follows.

  That is, according to the present invention, in a semiconductor substrate having a nonvolatile memory cell having first and second field effect transistors arranged adjacent to each other, the first field effect already formed on the semiconductor substrate. A portion of the first conductivity type first semiconductor region of the transistor that is disposed in the formation region of the second field effect transistor is etched, and then the first conductivity type of the second field effect transistor is etched in the etched region. The second semiconductor region of the second conductivity type opposite to the above is formed.

  According to another aspect of the present invention, there is provided a non-volatile memory cell having first and second field effect transistors arranged adjacent to each other on a semiconductor substrate, wherein a first gate electrode under the first gate electrode of the first field effect transistor is provided. In the first semiconductor region of one conductivity type, the end on the second field effect transistor side protrudes under the second gate electrode of the second field effect transistor or coincides with the end of the second gate electrode. In addition, a second semiconductor region of a second conductivity type opposite to the first conductivity type is formed under the second gate electrode of the second field effect transistor.

  Further, the present invention includes a nonvolatile memory cell having a gate electrode and a charge storage layer, and after processing the gate electrode, the end of the charge storage layer is the end side of the gate electrode. Over-etching is performed so as to be positioned below the side surface of the end.

  The present invention also relates to a nonvolatile memory having a structure in which one of the first and second gate electrodes of the first and second field effect transistors arranged adjacent to each other runs over the other gate electrode. In the case where a cell is provided on a semiconductor substrate, before depositing the conductive film for forming the one gate electrode, a side surface of the other gate electrode on the side where the one gate electrode does not run A protective film is formed on the main surface.

  Further, the present invention includes a nonvolatile memory cell that erases data by extracting charges stored in the charge storage layer to the n-type gate electrode side, and the charge storage in the n-type gate electrode The n-type impurity concentration in the first region on the layer side is made lower than the n-type impurity concentration in the second region other than the first region in the n-type gate electrode.

  Further, the present invention has a nonvolatile memory cell that erases data by extracting the charge stored in the charge storage layer to the gate electrode side, and the state with the lowest write level is the initial state of the nonvolatile memory cell. It is set to a state higher than the threshold voltage.

  The effects obtained by the representative ones of the inventions disclosed by the present application will be briefly described as follows.

  That is, it is possible to improve the electrical characteristics of a semiconductor device having a nonvolatile memory cell having first and second field effect transistors arranged adjacent to each other on a semiconductor substrate.

1 is a circuit diagram of a memory cell of a semiconductor device according to an embodiment of the present invention. FIG. 2 is a cross-sectional view of a main part of a device basic structure example of the memory cell of FIG. 1. FIG. 3 is an explanatory diagram illustrating an example of a voltage value applied to each unit during a data read operation, an erase operation, and a write operation of the memory cell of FIGS. 1 and 2; FIG. 7 is a cross-sectional view of a principal part of another example of the device basic structure of the memory cell of FIG. 1. FIG. 10 is a cross-sectional view of a principal part of still another example of the basic device structure of the memory cell of FIG. 1. FIG. 7 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step for illustrating a problem occurring in the memory cell formation step of FIG. 1; FIG. 7 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step for illustrating a problem occurring in the memory cell formation step of FIG. 1; It is principal part sectional drawing in the manufacturing process of the semiconductor device which is one embodiment of this invention. FIG. 9 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 8; FIG. 10 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 9; FIG. 11 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 10; FIG. 12 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step following that of FIG. 11; FIG. 13 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step following that of FIG. 12; FIG. 14 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 13; FIG. 15 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 14; FIG. 16 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step following that of FIG. 15; FIG. 17 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 16; FIG. 18 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step following that of FIG. 17; FIG. 19 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 18; It is principal part sectional drawing in the manufacturing process of the semiconductor device which is other embodiment of this invention. FIG. 21 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 20; FIG. 22 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 21; It is principal part sectional drawing in the manufacturing process of the semiconductor device which is other embodiment of this invention. FIG. 24 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 23; FIG. 25 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 24; FIG. 26 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 25; FIG. 27 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 26; FIG. 28 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 27; FIG. 29 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 28; FIG. 30 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 29; FIG. 31 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 30; FIG. 10 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step for illustrating another problem that occurs during the memory cell formation step of FIG. 1; It is an expanded sectional view of the area | region B of FIG. It is principal part sectional drawing in the manufacturing process of the semiconductor device which is other embodiment of this invention. FIG. 35 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 34; FIG. 36 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 35; FIG. 37 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 36; FIG. 38 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 37; It is an expanded sectional view of the area | region D of FIG. It is a principal part expanded sectional view of the semiconductor device which is other embodiment of this invention. FIG. 10 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step for illustrating another problem that occurs during the memory cell formation step of FIG. 1; FIG. 42 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 41; FIG. 43 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 42; It is principal part sectional drawing in the manufacturing process of the semiconductor device which is other embodiment of this invention. FIG. 45 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 44; FIG. 46 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 45; FIG. 47 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 46; FIG. 48 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 47; FIG. 49 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 48; FIG. 50 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 49; FIG. 51 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 50; FIG. 52 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 51; It is principal part sectional drawing in the manufacturing process of the semiconductor device which is other embodiment of this invention. FIG. 54 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 53; FIG. 55 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 54; FIG. 56 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 55; FIG. 57 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 56; FIG. 58 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 57; FIG. 59 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 58; FIG. 60 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 59; FIG. 61 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 60; It is a principal part top view of an example of the memory area | region of the semiconductor device which is other embodiment of this invention. FIG. 10 is a fragmentary cross-sectional view of the semiconductor device during a manufacturing step for illustrating another problem that occurs during the memory cell formation step of FIG. 1; FIG. 64 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 63; FIG. 65 is an enlarged cross-sectional view of a main part of FIG. It is principal part sectional drawing in the manufacturing process of the semiconductor device which is other embodiment of this invention. FIG. 67 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 66; FIG. 68 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 67; FIG. 69 is an essential part enlarged cross-sectional view of a memory cell of the semiconductor device of FIG. 68. It is a principal part expanded sectional view of the memory cell of the semiconductor device which is other embodiment of this invention. It is principal part sectional drawing in the manufacturing process of the semiconductor device which is other embodiment of this invention. FIG. 72 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 71; FIG. 73 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 72; FIG. 74 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 73; It is principal part sectional drawing in the manufacturing process of the semiconductor device which is other embodiment of this invention. FIG. 76 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 75; FIG. 77 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 76; FIG. 78 is an essential part cross sectional view of the semiconductor device during a manufacturing step following FIG. 77; FIG. 10 is a diagram for explaining another problem that occurs in the process of forming the memory cell of FIG. 1 and is a graph showing the dependency of the erase operation on the impurity concentration in the memory gate electrode. FIG. 4 is an energy band diagram of a memory gate electrode and a charge storage layer in the vicinity thereof. It is principal part sectional drawing of the semiconductor device which is other embodiment of this invention. It is principal part sectional drawing of the semiconductor device which is other embodiment of this invention. It is principal part sectional drawing of the semiconductor device which is other embodiment of this invention. It is principal part sectional drawing of the semiconductor device which is other embodiment of this invention. It is explanatory drawing of the write-in and erased state of the semiconductor device which is further another embodiment of this invention.

  Before describing embodiments of the present invention in detail, the meaning of terms in the present embodiment will be described as follows.

When silicon nitride, silicon nitride, or silicon nitride is used, not only Si ₃ N ₄ but also silicon nitride and an insulating film having a similar composition are included.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the following embodiment, when necessary for the sake of convenience, the description will be divided into a plurality of sections or embodiments. However, unless otherwise specified, they are not irrelevant to each other. Is related to some or all of the other modifications, details, supplementary explanations, and the like. Further, in the following embodiments, when referring to the number of elements (including the number, numerical value, quantity, range, etc.), especially when clearly indicated and when clearly limited to a specific number in principle, etc. Except, it is not limited to the specific number, and may be more or less than the specific number. Further, in the following embodiments, the constituent elements (including element steps and the like) are not necessarily indispensable unless otherwise specified and apparently essential in principle. Needless to say. Similarly, in the following embodiments, when referring to the shapes, positional relationships, etc. of the components, etc., the shapes are substantially the same unless otherwise specified, or otherwise apparent in principle. And the like are included. The same applies to the above numerical values and ranges. Also, components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiments, and the repetitive description thereof is omitted. In the drawings used in the present embodiment, hatching may be added even in a plan view for easy understanding of the drawings. In the following embodiments, MIS • FETs (Metal Insulator Semiconductor Field Effect Transistors) representing field effect transistors are abbreviated as MIS, p-channel type MIS • FETs are abbreviated as pMIS, and n-channel type MIS • FETs. FET is abbreviated as nMIS. Note that a MOS • FET (Metal Oxide Semiconductor FET) is a field effect transistor having a structure in which a gate insulating film is a silicon oxide (SiO ₂ or the like) film, and is included in the subordinate concept of the MIS.

(Embodiment 1)
The inventor has studied a semiconductor device having a nonvolatile memory such as an EEPROM or a flash memory, and each of a plurality of memory cells of the nonvolatile memory has, for example, a split gate electrode structure having a two-transistor structure. And a MONOS (Metal Oxide Nitride Oxide Semiconductor) structure having a nitride film (such as a silicon nitride film) as a charge storage layer for data storage on one transistor side. In the case of the MONOS structure, the single transistor memory cell is more susceptible to disturbance than in the case of the EEPROM structure. Therefore, in order to prevent this, a split gate electrode structure having a two-transistor structure is adopted. Further, in any of the above-described split gate type memory cell structures, a negative voltage is applied to the memory gate electrode in order to inject holes from the semiconductor substrate into the charge storage layer or to extract electrons from the charge storage layer to the semiconductor substrate at the time of erasing. A power supply circuit is required, and the memory occupancy rate is reduced, resulting in a decrease in manufacturing yield. In addition, when writing is performed with the FN tunneling current, there is a problem that it is difficult to increase the writing speed. Therefore, when writing data, hot electrons are injected to inject electrons from the semiconductor substrate into the charge storage nitride film, thereby speeding up the writing time, while erasing data, a positive voltage is applied to the memory gate electrode side. Thus, the power supply circuit can be simplified by performing the operation of extracting the electrons in the charge storage nitride film toward the memory gate electrode. However, the memory cell structure that performs the circuit operation as described above has not been studied in detail. Therefore, in this embodiment, a case where the present invention is applied to a memory cell configuration that performs the circuit operation as described above will be described. However, the following embodiment is applied to the memory cell structure of the circuit operation as described above. The present invention is not limited and includes those that can be applied to memory cells having various circuit operations.

  FIG. 1 shows a circuit diagram of the memory cell MC. The memory cell MC includes two transistors, for example, an nMISQnc for memory cell selection (hereinafter simply referred to as nMISQnc for selection) and an nMISQnm for memory between the drain electrode D and the source electrode S of the memory cell MC. have. The nMISQnc for selection has a control gate electrode CG, and the nMISQnm for memory has a memory gate electrode MG and the charge storage layer CSL. The symbol Vd indicates a drain voltage, Vcg indicates a control gate voltage, Vmg indicates a memory gate voltage, Vs indicates a source voltage, and Vsub indicates a substrate voltage.

FIG. 2 shows an example of a basic device section of the memory cell MC. Here, a control gate electrode riding-on type memory cell MC1 (MC) is illustrated. A semiconductor substrate (hereinafter simply referred to as a substrate) 1Sub is made of, for example, p-type silicon (Si) single crystal, and its main surface (device formation surface) includes nMISQnc for selecting the memory cell MC1, nMISQnm for memory, and the like. Is arranged. The drain region Drm and the source region Srm of the memory cell MC1 have, for example, an n ⁻ type semiconductor region 2a and an n ⁺ type semiconductor region 2b having an impurity concentration higher than that of the semiconductor region 2a (LDD ( Lightly Doped Drain) structure). The n ⁻ type semiconductor region 2a is arranged on the channel region side of the memory cell MC, and the n ⁺ type semiconductor region 2b is arranged at a position separated from the channel region side of the memory cell MC by the n ⁻ type semiconductor region 2a. Has been. On the main surface of the substrate 1Sub between the drain region Drm and the source region Srm, the control gate electrode CG of the selection nMISQnc and the memory gate electrode MG of the memory nMISQnm are disposed adjacent to each other. ing.

The control gate electrode CG and the memory gate electrode MG are made of, for example, n-type low resistance polysilicon. A part of the control gate electrode CG rides on the memory gate electrode MG side. Between the control gate electrode CG and the main surface of the substrate 1Sub, a gate insulating film 3 made of, for example, thin silicon oxide (SiO ₂ or the like) having a thickness of about ₂ to 3 nm is provided. A p-type semiconductor region 4 is formed on the main surface of the substrate 1Sub below the gate insulating film 3. The semiconductor region 4 is a semiconductor region for forming a channel for selection nMISQnc, and the threshold voltage of the selection nMISQnc is set to a predetermined value by the semiconductor region 4. The semiconductor region 4 has a function of increasing the threshold voltage of the selection nMISQnc. Boron (B) is introduced into the semiconductor region 4.

On the other hand, between the memory gate electrode MG and the main surface of the substrate 1Sub, the charge storage layer CSL is provided between the upper and lower insulating films 5t and 5b. The charge storage layer CSL is made of, for example, silicon nitride and has a thickness of, for example, 50 nm or less. However, the charge storage layer CSL is not limited to silicon nitride and can be variously changed. For example, a material capable of forming an insulating trap level such as alumina (Al ₂ O ₃ ) is used. it can. The insulating films 5b and 5t are made of, for example, silicon oxide. The insulating film 5b has a thickness of about 2 to 6 nm, for example, and the insulating film 5t has a thickness of about 10 to 17 nm, for example. The insulating film 5t may be formed of silicon oxynitride (SiON).

  The insulating films 5b and 5t can be formed as silicon oxide films containing nitrogen. This is because a silicon oxide film is formed as a gate insulating film of nMISQnc for selection or MIS formed in the MIS formation region around the memory cell after forming nMISQnm for memory. There is a problem that the memory gate electrode MG made of polycrystalline silicon is oxidized, thereby increasing the film thickness at the end portions of the insulating films 5b and 5t. In the nMISQnm for memory as in the present embodiment, a write operation is performed by injecting electrons from the substrate 1Sub to the charge storage layer CSL, and an erase operation is performed by extracting electrons from the charge storage layer CSL to the memory gate electrode MG. Therefore, such an increase in the thickness of the insulating films 5b and 5t at the end of the memory gate electrode MG causes erroneous writing / erase. Therefore, by forming the insulating films 5b and 5t as silicon oxide films containing nitrogen, it becomes possible to prevent the formation of excessive silicon oxide films at the end portions of the insulating films 5b and 5t.

  As a specific implementation method, after forming the insulating film 5b, for example, oxynitridation (NO) annealing is performed, and nitrogen is introduced into the substrate 1Sub side of the insulating film 5b. As a result, it is possible to prevent an increase in the thickness of the end portion of the insulating film 5b during another oxidation process. Further, nitrogen is introduced to the conductive film side of the insulating film 5t by performing nitrogen plasma treatment after forming the insulating film 5t or by performing oxynitridation (NO) annealing after depositing the conductive film to be the memory gate electrode MG. Can do. As a result, it is possible to prevent an increase in the thickness of the end portion of the insulating film 5t during another oxidation process.

  Furthermore, since the gate insulating film (silicon oxide film) of the nMISQnc for selection or the MIS around the memory cell is formed by the CVD method, the film forming temperature can be lowered as compared with the thermal oxidation process, so that the insulating film 5b It becomes possible to further inhibit the formation of an excessive silicon oxide film at the end of 5t.

  An n-type semiconductor region 6 is formed below the insulating film 5b and on the main surface of the substrate 1Sub between the p-type semiconductor region 4 and the source region Srm. The semiconductor region 6 is a semiconductor region for forming an nMISQnm channel for memory, and the threshold voltage of the nMISQnm for memory is set to a predetermined value by the semiconductor region 6. The semiconductor region 6 has a function of lowering the threshold voltage of nMISQnm for memory. For example, arsenic (As) or phosphorus (P) is introduced into the semiconductor region 6. An insulating film 7 made of, for example, silicon oxide is provided on the upper surface of the memory gate electrode MG.

Further, sidewalls 8 made of, for example, silicon oxide are formed on the side surfaces of the memory gate electrode MG, the insulating films 5t and 5b, and the charge storage layer CSL to insulate the memory gate electrode MG from the control gate electrode CG. ing. Sidewalls 9 are formed on both side surfaces of the control gate electrode CG of the memory cell MC1, a part of the upper surface of the insulating film 7, and the surface of the side wall 8 on the source region Srm side so as to cover it. The sidewall 9 is made of, for example, silicon oxide and is a member mainly for forming the n ⁻ type semiconductor region 2a.

  In such a memory cell MC1, the charge contributing to data storage is injected from the substrate 1Sub to the trap level in the charge storage layer CSL at the position indicated by the arrow A by hot electron injection. Since this charge is stored discretely, even if a defect occurs in some part of the oxide film (insulating films 5b, 5t and sidewalls 8) surrounding the charge storage layer CSL and an abnormal leak occurs, the charge storage layer Since all charges in the CSL are not lost, the reliability of data retention can be improved. For this reason, since the insulating films 5b and 5t above and below the charge storage layer CSL can be made thin, it is possible to reduce the voltage of the write and erase operations. Further, when data is written, electrons are injected from the substrate 1Sub into the charge storage layer CSL by hot electron injection, so that electron injection efficiency is excellent and high-speed, low-current writing can be performed. On the other hand, when erasing data, a positive voltage is applied to the memory gate electrode MG side, and electrons in the charge storage layer CSL are extracted to the memory gate electrode MG side, so that control of writing and erasing operations is easy. The power supply circuit and the peripheral circuit can be simplified (downsized).

  FIG. 3 shows an example of voltage values applied to the respective parts during the data read operation Re, erase operation Er, and write operation Wr of the memory cell MC of FIGS.

  During the data read operation Re, the drain electrode D (drain region Drm) of the selected memory cell MC has a source electrode S (for example, about 1 V and the control gate electrode CG has a source electrode S (for example, about 1.5 V). For example, 0 (zero) V is applied to the source region Srm), the memory gate electrode MG, and the substrate 1Sub, and the nMISQnc for selecting the memory cell MC is turned on. At this time, the threshold voltage of the memory nMISQnm changes depending on the presence or absence of electrons in the charge storage layer CSL of the memory nMISQnm, and a current flows or does not flow between the drain region Drm and the source region Srm. Thus, the stored data is read out.

  In the data erasing operation Er, the drain electrode D (drain region Drm), the source electrode S (source region Srm) and the substrate 1Sub of the selected memory cell MC are applied to, for example, 0 (zero) V and the control gate electrode CG. For example, about 14V is applied to the memory gate electrode MG, for example, about 1.5V. As a result, electrons in the charge storage layer CSL escape to the memory gate electrode MG side by tunnel emission, and data is erased.

  Further, the data write employs the source side hot electron injection method. In the data write operation Wr, the drain electrode D (drain region Drm) and the substrate 1Sub of the selected memory cell MC are set to, for example, 0 (zero) V, the control gate electrode CG is set to, for example, about 1.5 V, and the memory gate electrode MG. For example, about 12V is applied to the source electrode S (source region Srm) of the selected memory cell MC, for example, about 6V. As a result, hot electrons generated in the channel of the memory cell MC are injected into the charge storage layer CSL to write data.

  4 and 5 show another example of a basic device section of the split gate electrode type memory cell MC. FIG. 4 illustrates a memory cell MC2 (MC) riding on the memory gate electrode, and FIG. 5 illustrates a memory cell MC3 (MC) riding on the side wall type memory gate electrode. In the memory cells MC2 and MC3 of FIGS. 4 and 5, a part of the memory gate electrode MG runs on the control gate electrode CG side, but the insulating film 10 made of silicon oxide or the like on the upper surface of the control gate electrode CG, the insulating film The control gate electrode CG and the memory gate electrode MG are insulated by 5b, 5t, the charge storage layer CSL, and the like. Also in this case, data reading, writing and erasing operations are the same as described above, and thus description thereof is omitted. The semiconductor device having the memory cell as described above is used in, for example, an IC (Integrated Circuit) card (memory card).

  Next, the problems found by the present inventors for the split gate type memory cells MC (MC1 to MC3) shown in FIGS. 2, 4, and 5 will be described with reference to FIGS.

  FIG. 6 shows a cross-sectional view of the main part of the substrate 1Sub during the process of forming the split gate electrode type memory cell MC1 of FIG. An n-type semiconductor region 6 is formed on the main surface of the substrate 1Sub (at this stage, a planar substantially circular wafer). In addition, a laminated pattern of the insulating film 5b, the charge storage layer CSL, the insulating film 5t, the memory gate electrode MG, and the insulating film 7 is formed on the main surface of the substrate 1Sub. A sidewall 8 made of, for example, silicon oxide is formed on the side surface of the laminated pattern. In the process of forming the memory cell MC1, first, as shown in FIG. 7, the formation region of the selection nMISQnc (see FIG. 2) is exposed on the main surface of the substrate 1Sub, and other regions are formed. A photoresist pattern (hereinafter simply referred to as a resist pattern) PR1 that is covered is formed. Subsequently, boron (B), for example, is introduced into the main surface of the substrate 1Sub by ion implantation using the resist pattern PR1, the sidewall 8 and a part of the memory gate electrode MG as a mask. Thereby, the p-type semiconductor region 4 is formed on the main surface of the substrate 1Sub in a self-aligned manner with respect to the memory gate electrode MG. At this time, in order to keep the threshold voltage of the selection nMISQnc high from the viewpoint of the operation stability of the memory cell MC1, the amount of boron introduced is set to the above-mentioned n-type already formed on the main surface of the substrate 1Sub. It is necessary to introduce it to such an extent that the conductivity type of the semiconductor region 6 is canceled. For this reason, since a large amount of impurities are implanted into the p-type semiconductor region 4, the total impurity concentration in the semiconductor region 4 (the sum of the impurity concentrations for forming each of the semiconductor regions 4 and 6). Get higher.

  As described above, in the split gate electrode type memory cells MC1 to MC3 of FIGS. 2, 4 and 5, the nMISQnc for selection and the nMISQnm for memory are adjacent to each other, so that the channel for forming nMISQnc and Qnm is formed. In order to form a semiconductor region, an impurity for forming one of the channel forming semiconductor regions is implanted into the substrate 1Sub, and then an impurity of the opposite conductivity type is implanted so as to cancel the other channel. It is necessary to form a semiconductor region for formation. For this reason, since a large amount of impurities are implanted into one of the channel forming semiconductor regions, the total impurity concentration of the channel forming semiconductor region into which the large amount of impurities has been implanted increases. As a result, deterioration of electrical characteristics such as a decrease in carrier mobility (a decrease in drain current Ids) and an increase in variation in threshold voltage occur.

  Therefore, in the first embodiment, in the step of forming the split gate electrode type memory cell MC having two MISs, the MIS for selection and the MIS for memory, the semiconductor region for channel formation of one MIS is etched. Then, a semiconductor region for forming the other MIS channel is formed. Thereby, since the impurity concentration of the semiconductor region for channel formation of the other MIS can be lowered, the carrier mobility can be improved and the drain current Ids in the memory cell MC can be increased. Therefore, the operation speed (data read speed) of the split gate electrode type memory cell MC can be improved. In addition, since variation in threshold voltage can be reduced, operation reliability of the split gate type memory cell MC can be improved.

  A specific example will be described with reference to cross-sectional views of the main part of the substrate 1Sub in the memory region during the manufacturing process of the semiconductor device of FIGS. Here, an application example to the memory cell MC1 in FIG. 2 will be described.

First, as shown in FIG. 8, arsenic (As), for example, is ion-implanted on the main surface of a substrate 1Sub made of p-type silicon (Si) single crystal (at this stage, a substantially circular planar semiconductor wafer). Introduce. Thereby, an n-type semiconductor region 6 for forming a channel of nMISQnm for memory (see FIG. 2) is formed on the main surface of the substrate 1Sub. In the first embodiment, there is an etching process in which the main surface of the substrate 1Sub is slightly etched in a process described later to partially remove the n-type semiconductor region 6, but impurity ions for forming the semiconductor region 6 are included. Is implanted to a very deep position, it is necessary to increase the etching amount of the substrate 1Sub during the above etching process. Therefore, it is preferable to reduce the implantation depth of impurity ions for forming the semiconductor region 6. preferable. The impurity ion implantation energy at this time is, for example, about 20 to 40 keV. Further, the implantation depth of impurity ions at this time (depth at which the impurity concentration becomes highest) is, for example, about 10 to 20 nm. The dose amount of impurity ions is, for example, about 1 × 10 ¹⁴ / cm ² . Although phosphorus can be used as an impurity, arsenic is used because the semiconductor region 6 can be formed at a shallow position on the substrate 1Sub even with the same implantation energy by using arsenic having an atomic weight (mass) heavier than phosphorus. This is because it is easy to form the semiconductor region 6 at the position.

  Next, as shown in FIG. 9, on the main surface of the substrate 1Sub, for example, an insulating film 5b made of silicon oxide, a charge storage layer CSL made of silicon nitride, an insulating film 5t made of silicon oxide, and a memory made of low-resistance polysilicon. The conductive film 11 for gate formation and the insulating film 7 made of silicon oxide are oxidized or sequentially deposited from the lower layer by the CVD (Chemical Vapor Deposition) method, and then these laminated films are photolithography (hereinafter simply referred to as lithography) technology and etching. By patterning with the technique, as shown in FIG. 10, a laminated pattern of the insulating film 5b, the charge storage layer CSL, the insulating film 5t, the memory gate electrode MG (11), and the insulating film 7 is formed. Subsequently, after an insulating film made of, for example, silicon oxide is deposited on the main surface of the substrate 1Sub by a CVD method or the like, this is etched back by an anisotropic dry etching method, so that as shown in FIG. Sidewalls 8 are formed on the side surfaces of the laminated pattern.

  Next, as shown in FIG. 12, a resist pattern PR2 is formed on the main surface of the substrate 1Sub so that the nMIS formation region for selection is exposed, and then the resist pattern PR2, the sidewalls 8 and the insulating film 8 are etched. As a mask, the main surface of the substrate 1Sub in the nMIS formation region for selection is partially etched by performing an etching process under such a condition that the etching selectivity between the material of the sidewall 8 and the material of the substrate 1Sub is increased. The recess 13 is formed. Thus, the n-type semiconductor region 6 in the nMIS formation region for selection is removed. At this time, since the sidewall 8 is used as an etching mask, a part of the substrate 1Sub can be selectively removed with good alignment. As an etching method at this time, for example, a dry etching method is used. However, a wet etching method may be used. When wet etching is used, a part of the substrate 1Sub can be removed without causing much damage to the substrate 1Sub. Further, the damaged layer of the substrate 1Sub generated during the dry etching may be removed by performing a slight wet etching process after the dry etching process. Thereby, the electrical characteristics of the nMIS for selection can be improved. If the depth of the recess 13 is too deep, it causes a step, and is preferably, for example, 50 nm or less, and specifically, for example, about 20 nm.

Next, as shown in FIG. 13, for example, boron difluoride (BF ₂ ) is ion-implanted into the main surface of the substrate 1Sub using the resist pattern PR2, the sidewall 8, and the memory gate electrode MG as a mask. The p-type semiconductor region 4 for forming the nMIS channel is formed. The implantation energy of impurity ions at this time is, for example, about 40 to 100 keV. In the first embodiment, since the main surface of the substrate 1Sub in the nMIS formation region for selection is cut and the n-type semiconductor region 6 is removed, the selection nMIS is compared with the case where the depression 13 is not formed. The impurity concentration of the p-type semiconductor region 4 for forming the channel can be lowered. For this reason, the mobility of carriers (electrons) can be improved, and the drain current Ids in the memory cell MC1 can be increased. Therefore, the operation speed (data read speed) of the split gate electrode type memory cell MC1 can be improved. In addition, since the variation in the threshold voltage of the nMIS for selection can be reduced, the operation reliability of the split gate electrode type memory cell MC1 can be improved.

Next, as shown in FIG. 14, the substrate 1Sub is oxidized to form a gate insulating film 3 made of, for example, silicon oxide on the main surface of the substrate 1Sub, and then on the main surface of the substrate 1Sub. In addition, a control gate electrode CG is formed by depositing a conductor film made of, for example, low-resistance polysilicon by the CVD method and patterning the conductor film by a lithography technique and a dry etching technique. Subsequently, as shown in FIG. 15, by using the control gate electrode CG and the memory gate electrode MG as a mask, for example, arsenic or phosphorus is ion-implanted into the main surface of the substrate 1Sub, thereby forming an n ⁻ type in the main surface of the substrate 1Sub. The semiconductor region 2a is formed in a self-aligned manner with respect to the control gate electrode CG and the memory gate electrode MG.

Next, an insulating film made of, for example, silicon oxide is deposited on the main surface of the substrate 1Sub by a CVD method, and then etched back by an anisotropic dry etching method, thereby forming a control gate as shown in FIG. Sidewalls 9 are formed on both side surfaces of the electrode CG, on the insulating film 7 and on the surfaces of the sidewalls 8. Subsequently, as shown in FIG. 17, by using the sidewall 9 and the control gate electrode CG as a mask, for example, arsenic or phosphorus is ion-implanted into the main surface of the substrate 1Sub, whereby an n ⁺ -type semiconductor is formed on the main surface of the substrate 1Sub. Region 2b is formed in a self-aligned manner with respect to control gate electrode CG and memory gate electrode MG. Thus, the drain region Drm and the source region Srm of the memory cell MC1 are formed, and the selection nMISQnc and the memory nMISQnm are formed. Thereafter, as shown in FIG. 18, a silicide layer 14 such as cobalt silicide (CoSi _x ) is formed on the main surface portion of the substrate 1Sub and the upper surface portion of the control gate electrode CG by a salicide (Salicide: Self Align silicide) process. To do.

  Next, as shown in FIG. 19, an insulating film 15 made of, for example, silicon oxide is deposited on the main surface of the substrate 1Sub by a CVD method, and then contact holes CNT are formed in the insulating film 15. Subsequently, a plug PLG is formed in the contact hole CNT. The plug PLG is, for example, a thin barrier film made of a laminated film of titanium (Ti) and titanium nitride (TiN), and a relative structure made of tungsten (W) or aluminum (Al) formed so as to be surrounded by the barrier film. And a thick conductor film. Thereafter, a first layer wiring M1 made of, for example, tungsten or aluminum (Al) is formed on the insulating film 15. Thereafter, a semiconductor device having a nonvolatile memory is manufactured through a normal semiconductor device manufacturing process.

(Embodiment 2)
In the second embodiment, which is a modification of the first embodiment, an example in which the substrate 1Sub portions on both sides of the memory gate electrode are etched is illustrated as an example of the substrate 1Sub in the manufacturing process of the semiconductor device of FIGS. This will be described with reference to a cross-sectional view of the main part.

First, after passing through the steps described in FIGS. 8 to 11 of the first embodiment, the substrate 1Sub is subjected to the same etching process as described in FIG. A recess 13 is formed in the main surface portion of the substrate 1Sub exposed from the formation region of the electrode MG and the sidewall 8. Here, both sides of the formation region of the memory gate electrode MG and the side wall 8 are removed. That is, the semiconductor region 6 where the source region of the memory cell is formed is also removed. Subsequently, as shown in FIG. 21, after forming a resist pattern PR2 on the main surface of the substrate 1Sub so that the nMIS formation region side for selection is exposed, the resist pattern PR2, the sidewall 8 and the insulating film 8 are formed. As a mask, for example, boron difluoride (BF ₂ ) is ion-implanted into the main surface of the substrate 1Sub to form a p-type semiconductor region 4 for forming a channel for selecting nMIS. The conditions at this time are the same as described in FIG. Also in the second embodiment, the same effect as described in FIG. 13 can be obtained. Thereafter, in the same manner as in the first embodiment, a memory cell MC1 (MC) is formed as shown in FIG.

In the second embodiment, the depression 13 is formed even in the portion where the source region Srm of the memory cell MC1 is formed, and the layer of the n-type semiconductor region 6 is removed. Therefore, the n ⁻ type of the drain region Drm and the source region Srm. The impurity concentration of the semiconductor region 2a and the n ⁺ type semiconductor region 2b can be reduced. Further, compared to the case where the depression 13 is not formed, the concentration adjustment of the n ⁻ type semiconductor region 2a and the n ⁺ type semiconductor region 2b can be facilitated.

(Embodiment 3)
The methods of the first and second embodiments can be applied to the memory cells MC2 and MC3 shown in FIGS. Therefore, in the third embodiment, an example in which the method of the second embodiment is applied to the memory cell MC2 of FIG. 4 will be described with reference to cross-sectional views of main parts in the manufacturing process of the semiconductor device of FIGS.

First, as shown in FIG. 23, for example, boron difluoride (BF ₂ ) is introduced into the main surface of a substrate 1Sub (the semiconductor wafer) made of, for example, p-type silicon (Si) single crystal by an ion implantation method. Thereby, a p-type semiconductor region 4 for channel formation of nMISQnc for selection (see FIG. 4) is formed on the main surface of the substrate 1Sub. In the third embodiment, there is an etching process in which the main surface of the substrate 1Sub is slightly etched in a process described later to partially remove the p-type semiconductor region 4, so that the n-type semiconductor region of the first embodiment is used. For the same reason as in FIG. 6, it is preferable to reduce the implantation depth of impurity ions for forming the p-type semiconductor region 4. The implantation energy of impurity ions at this time is, for example, about 40 to 100 keV. Further, the implantation depth of impurity ions at this time (depth at which the impurity concentration becomes highest) is, for example, about 40 to 100 nm. The dose amount of impurity ions is, for example, about 1 × 10 ¹⁴ / cm ² . Boron can be used as an impurity, but boron difluoride is used because the boron difluoride having an atomic weight (mass) heavier than boron is used so that the semiconductor region is located at a shallow position on the substrate 1Sub even with the same implantation energy. This is because it is easy to form the semiconductor region 4 at a shallow position.

  Subsequently, as shown in FIG. 24, the substrate 1Sub is oxidized to form a gate insulating film 3 made of, for example, silicon oxide on the main surface of the substrate 1Sub, and then the main surface of the substrate 1Sub. A conductive film 17 made of, for example, low-resistance polysilicon is deposited on the conductive film 17 by CVD, and an insulating film 10 made of, for example, silicon oxide is deposited on the conductive film 17 by CVD. Thereafter, the gate insulating film 3, the conductor film 17 and the insulating film 10 are patterned by lithography and dry etching techniques to form the control gate electrode CG as shown in FIG.

  Next, by performing the same etching process as described in FIG. 12 on the substrate 1Sub, as shown in FIG. 26, the depression 13 is formed in the main surface portion of the substrate 1Sub exposed from the formation region of the control gate electrode CG. To do. Here, the surface layer of the substrate 1Sub on both sides of the formation region of the control gate electrode CG is removed. That is, the p-type semiconductor region 4 in the nMIS formation region, source region, and drain region formation portion of the memory cell is also removed. Subsequently, as shown in FIG. 27, for example, arsenic or phosphorus is ion-implanted into the main surface of the substrate 1Sub using the control gate electrode CG as a mask, thereby forming an n-type semiconductor region for forming an nMIS channel for memory. 6 is formed. The impurity ion implantation energy at this time is, for example, about 20 to 40 keV. In the third embodiment, since the main surface of the substrate 1Sub in the nMIS formation region for the memory is cut and the p-type semiconductor region 4 is removed, the n-type semiconductor region 6 for forming the nMIS channel for the memory is removed. The impurity concentration of can be reduced. For this reason, the mobility of carriers (electrons) can be improved, and the drain current Ids in the memory cell MC2 can be increased. Therefore, the operation speed (data reading speed) of the split gate electrode type memory cell MC2 can be improved. In addition, since the variation in the threshold voltage of the memory nMIS can be reduced as compared with the case where the depression 13 is not formed, the operation reliability of the split gate type memory cell MC2 can be improved.

Next, as shown in FIG. 28, the insulating film 5b, the charge storage layer CSL insulating film 5t, and the memory gate forming conductor film 11 are sequentially deposited on the main surface of the substrate 1Sub from the lower layer by the CVD method. By patterning the laminated film by lithography and etching techniques, a laminated pattern of the insulating film 5b, the charge storage layer CSL, the insulating film 5t, and the memory gate electrode MG (11) is formed as shown in FIG. Subsequently, using the control gate electrode CG and the memory gate electrode MG as a mask, for example, arsenic or phosphorus is ion-implanted into the main surface of the substrate 1Sub, thereby forming the n ⁻ type semiconductor region 2a in the main surface of the substrate 1Sub. It is formed in a self-aligned manner with respect to the CG and the memory gate electrode MG. Thereafter, an insulating film made of, for example, silicon oxide is deposited on the main surface of the substrate 1Sub by a CVD method, and then etched back by an anisotropic dry etching method, so that a control gate is formed as shown in FIG. Sidewalls 9 are formed on one side of the electrode CG, on the insulating film 10 and on both sides of the memory gate electrode MG. Subsequently, using the sidewall 9 and the memory gate electrode MG as a mask, for example, arsenic or phosphorus is ion-implanted into the main surface of the substrate 1Sub, thereby forming the n ⁺ -type semiconductor region 2b in the main surface of the substrate 1Sub. And in a self-aligned manner with respect to the memory gate electrode MG. In this way, the drain region Drm and the source region Srm of the memory cell MC2 are formed, and the selection nMISQnc and the memory nMISQnm are formed. Thereafter, as shown in FIG. 31, the silicide layer 14 is formed on the main surface portion of the substrate 1Sub and the upper surface portion of the control gate electrode CG through the salicide process, as in the first embodiment, and the insulating film 15 is deposited. The contact hole CNT is formed, the plug PLG is formed, and the first layer wiring M1 is formed.

In the third embodiment, the depression 13 is formed also in the portion where the drain region Drm and the source region Srm of the memory cell MC2 are formed, and the layer of the p-type semiconductor region 4 is removed, so the drain region Drm and the source region Srm The impurity concentration of the n ⁻ type semiconductor region 2a and the n ⁺ type semiconductor region 2b can also be reduced. Further, compared with the case where the depression 13 is not formed, the concentration adjustment of the n ⁻ type semiconductor region 2a and the n ⁺ type semiconductor region 2b can be facilitated.

(Embodiment 4)
In the fourth embodiment, an example for solving another problem found for the first time by the inventor of the split gate electrode type memory cell will be described. The problem is that the semiconductor region for the source / drain of the MIS for selecting the memory cell is offset with respect to the control gate electrode.

  First, the problem will be described with reference to FIGS. FIG. 32 is a cross-sectional view of a main part of the memory cell MC1, and FIG. In the case of the memory cell MC1, the p-type semiconductor region 4 for channel formation of the nMISQnc for selection is formed after the formation of the n-type semiconductor region 6 for channel formation of the nMISQnm for memory. Therefore, the nMISQnm for selection In the boundary region with the memory nMISQnm, a part of the p-type semiconductor region 4 diffuses toward the n-type semiconductor region 6 side. In the region C in which the p-type semiconductor region 4 is diffused to the n-type semiconductor region 6 side, the originally n-type region may be changed to a p-type or a conductivity type close to p-type. As a result, when viewed from the control gate electrode CG of the selection nMISQnc, the n-type semiconductor region 6 serving as the source / drain region of the selection nMISQnc is separated by the length L1. That is, a portion (region C) where the control gate electrode CG that regulates the drain current Ids does not overlap in plan via the gate insulating film 3 is generated. When such an offset region is generated, the drain current Ids is reduced, and the operation speed of the memory cell MC1 is reduced.

  Therefore, in the fourth embodiment, a spacer is provided so that the end of the semiconductor region for forming the channel of the selection MIS is separated from the side surface of the sidewall formed on the side surface of the memory gate electrode of the memory MIS. After forming the side surface of the side wall of the memory gate electrode, an impurity for forming a semiconductor region for forming a channel of the MIS for selection is introduced into the substrate 1Sub. As a result, the end of the semiconductor region (semiconductor region for forming the channel of the MIS for the memory) that is to be the source / drain region of the MIS for selection can be surely overlapped under the control gate electrode. The drain current Ids of the gate type memory cell can be improved. Therefore, the operation speed (data reading speed) of the split gate type memory cell can be improved. A specific example of this will be described with reference to a cross-sectional view of the main part of the substrate 1Sub in the memory region during the manufacturing process of the semiconductor device shown in FIGS.

First, after the steps described with reference to FIGS. 8 to 11 of the first embodiment, the spacer insulating film 19 is deposited on the main surface of the substrate 1Sub (semiconductor wafer) as shown in FIG. . The insulating film 19 is made of, for example, silicon oxide and has a thickness of, for example, about 30 nm. In the fourth embodiment, the insulating film 19 is deposited by, for example, a CVD method using a mixed gas of TEOS (Tetraethoxysilane) and oxygen (O ₂ ). The lower insulating film 7 and the silicon oxide film for forming the sidewall 8 are deposited by, for example, a CVD method using a mixed gas of monosilane (SiH ₄ ) and oxygen (O ₂ ), and then the silicon oxide film It is formed by annealing treatment for densification. Thus, when the insulating film 19 is removed by wet etching as will be described later, the etching rate of the insulating film 19 is about twice that of the insulating film 7 and the sidewall 8 made of the same silicon oxide. Therefore, the insulating film 19 can be selectively removed. However, silicon nitride may be used as the material of the insulating film 19. Also in this case, when the insulating film 19 is removed by etching, the etching selectivity with silicon oxide can be made high.

Subsequently, as shown in FIG. 35, after forming the resist pattern PR2 on the main surface of the substrate 1Sub in the same manner as in the first embodiment, as described in FIG. 13 in the first embodiment, For example, ion implantation of boron difluoride (BF ₂ ) or boron (B) is performed to form the p-type semiconductor region 4 for forming the nMIS channel for selection. At this time, in the fourth embodiment, the insulating film 19 deposited on the side surface of the sidewall 8 on the nMIS formation region side for selection serves as a mask, and the end portion on the memory gate electrode CG side of the p-type semiconductor region 4 However, it is positioned away from the side surface of the sidewall 8 on the nMIS formation region side for selection by the thickness of the insulating film 19.

Thereafter, after removing the resist pattern PR2, the insulating film 19 is selectively removed by wet etching as shown in FIG. Subsequently, as shown in FIG. 37, after forming the gate insulating film 3 and the control gate electrode CG, as in the first embodiment, for example, arsenic or phosphorus is ion-implanted into the substrate 1Sub to form an n ⁻ type semiconductor. Region 2a is formed. Thereafter, as shown in FIGS. 38 and 39, similarly to the first embodiment, the selection nMISQnc and the memory nMISQnm are formed, and the split gate electrode type memory cell MC1 is formed.

  In the fourth embodiment, as shown in FIG. 39, the end of the n-type semiconductor region 6 extends from the side surface of the sidewall 8 so as to protrude below the control gate electrode CG by a length L2. That is, the control gate electrode CG that regulates the current of the memory cell MC1 is surely overlapped with the end portion of the n-type semiconductor region 6 (the source / drain region of the nMISQnc for selection) via the gate insulating film 3. be able to. Therefore, since the drain current Ids of the split gate electrode type memory cell MC1 can be improved, the operation speed (data read speed) of the split gate electrode type memory cell MC1 can be improved. The length L2 can be controlled by the thickness of the insulating film 19. Therefore, it is possible to relatively easily set the overlapping amount in which the control gate electrode CG overlaps the end portion of the n-type semiconductor region 6 through the gate insulating film 3. The length L3 indicates the thickness of the insulating film 19. A length L4 obtained by subtracting the length L2 from the length L3 indicates a length in which the p-type semiconductor region 4 is diffused to the n-type semiconductor region 6 side.

  However, in the fourth embodiment, the n-type semiconductor region 6 protrudes under the control gate electrode CG and the control gate electrode CG overlaps the end of the n-type semiconductor region 6 with the gate insulating film 3 interposed therebetween. As described above, in the fourth embodiment, it is sufficient that the end portion of the n-type semiconductor region 6 which is the source / drain region of nMISQnc for selection is not separated from the end portion of the control gate electrode CG, as shown in FIG. As described above, the control gate electrode CG does not overlap the end of the n-type semiconductor region 6 with the gate insulating film 3 interposed therebetween, but the end of the n-type semiconductor region 6 substantially coincides with the side surface position of the sidewall 8. However, this does not exclude a structure that is not separated from the end of the control gate electrode CG. Note that the length L5 in FIG. 40 is the thickness of the insulating film 19, and is also the length by which the p-type semiconductor region 4 is diffused to the n-type semiconductor region 6 side.

  Further, the fourth embodiment and the first and second embodiments may be combined. That is, after forming the recess 13 in the main surface of the substrate 1Sub, the insulating film 19 is deposited, and then an impurity for forming the p-type semiconductor region 4 is introduced.

  The insulating film 19 may also be used as an insulating film used in another process. In this case, since the process can be shared, the manufacturing time and cost of the semiconductor device can be shortened.

(Embodiment 5)
In the fifth embodiment, an example for solving the first problem found by the present inventor for the split gate electrode type memory cell will be described. The problem is that in the process of patterning the gate electrode on the side of the board, a part of the gate electrode on the side of the board on the side of the gate electrode that is not boarded is left, and when removing it, the main surface of the substrate This is a problem of etching.

  First, the problem will be described with reference to FIGS. 41 to 43 are cross-sectional views of the main part of the substrate 1Sub during the patterning process of the control gate electrode. As shown in FIG. 41, after forming the gate insulating film 3 on the main surface of the substrate 1Sub, the conductor film 21 made of low-resistance polysilicon or the like for forming the control gate electrode is formed on the main surface of the substrate 1Sub by the CVD method or the like. And a resist pattern PR3 for forming a control gate electrode is formed thereon. Subsequently, by performing an anisotropic dry etching process using the resist pattern PR3 as an etching mask, a control gate electrode CG is formed as shown in FIG. At this time, the conductor film 21a may be left on the side surface of the sidewall 8 on the side surface of the memory gate electrode MG. Therefore, in order to remove the remaining conductor film 21a, as shown in FIG. 43, a resist pattern PR4 that exposes the conductor film 21a is formed, and the conductor film 21a is removed by an etching process. However, since the conductor film 21a is made of polysilicon at this time, the main surface of the substrate 1Sub made of lower silicon is also partially etched, and the main surface of the substrate 1Sub is roughened. As a result, the drain current Ids of the memory cell is reduced. May decrease.

  Therefore, the fifth embodiment includes a step of forming a protective film on the substrate on the side where the gate electrode does not run. Thereby, when removing the etching residue generated during the processing of the gate electrode on the ride-on side, the main surface of the substrate 1Sub is protected by the protective film, so that the main surface of the substrate 1Sub can be prevented from being etched. A specific example will be described with reference to a cross-sectional view of the main part of the substrate 1Sub in the memory region during the manufacturing process of the semiconductor device of FIGS.

First, after the steps described with reference to FIGS. 8 to 11 of the first embodiment, as shown in FIG. 44, a resist pattern PR2 is formed on the main surface of the substrate 1Sub (semiconductor wafer). A p-type semiconductor region 4 is formed by ion implantation of boron bromide (BF ₂ ) or boron (B). Subsequently, after removing the resist pattern PR2, as shown in FIG. 45, an insulating film 23 made of, for example, silicon oxide or silicon nitride and having a thickness of about 20 nm is formed on the main surface of the substrate 1Sub. Thereafter, the same resist pattern PR5 as the resist pattern 2 used when forming the p-type semiconductor region 4 is formed on the insulating film 23, and then the insulating film 23 is etched using the resist pattern PR5 as an etching mask. As shown, a pattern of the insulating film 23 is formed. The pattern of the insulating film 23 is formed so as to cover a part of the upper surface of the insulating film 7, the surface of the side wall 8 on one side, and the upper surface of the substrate 1Sub on the source region side of the memory cell.

  Next, as shown in FIG. 47, after the gate insulating film 3 is formed on the main surface of the substrate 1Sub, a conductor film 21 made of, for example, low-resistance polysilicon is deposited thereon by a CVD method or the like. Subsequently, after forming a resist pattern PR3 for forming a control gate electrode on the conductor film 21, the conductor film 21 is patterned using this as an etching mask, so that the control gate electrode CG is formed as shown in FIG. Form. At this time, the conductor film 21a is left on the lower side surface of the side wall 8 on one side surface of the memory gate electrode MG. In the fifth embodiment, the conductor film 21a is left on the insulating film 23. Therefore, a resist pattern PR4 for removing the conductor film 21a is formed on the main surface of the substrate 1Sub by a lithography technique, and an etching process is performed using the resist pattern PR4 as an etching mask, whereby the conductor film 21a is shown in FIG. To remove. At this time, in the fifth embodiment, the etching selection ratio between the polysilicon and the insulating film 23 is increased, and the etching process is performed under the condition that the polysilicon is more easily etched than the insulating film 23. At this time, since the insulating film 23 thicker than the gate insulating film 3 is provided in the lower layer (between the conductive film 21a and the substrate 1Sub) of the remaining conductive film 21a made of polysilicon, the main surface of the substrate 1Sub. Is protected, and the main surface of the substrate 1Sub can be prevented from being etched. That is, it is possible to prevent the substrate 1Sub from being roughened due to the removal of the unetched conductor film 2a. Accordingly, since the drain current Ids of the split gate electrode type memory cell can be prevented from decreasing, the high speed operation (high speed data reading) of the split gate electrode type memory cell can be maintained.

Next, after removing the resist pattern PR4, as shown in FIG. 50, an impurity such as phosphorus or arsenic is ion-implanted into the substrate 1Sub to form the n ⁻ type semiconductor region 2a in the control gate electrode CG and the memory gate. It is formed in a self-aligned manner with respect to the electrode MG. Subsequently, as shown in FIG. 51, after the sidewall 9 is formed as in the first embodiment, the n ⁺ type semiconductor region 2b is formed on the main surface of the substrate 1Sub on the control gate electrode CG and the memory gate electrode MG. In a self-aligned manner. Thus, the drain region Drm and the source region Srm of the memory cell MC1 are formed, and the selection nMISQnc and the memory nMISQnm are formed. After that, as shown in FIG. 52, after the silicide layer 14 is formed by the salicide process, the insulating film 15 deposition process, the contact hole CNT formation process, the plug PLG formation process, A semiconductor device is manufactured through the formation process of the one-layer wiring M1.

  The fifth embodiment and the first and second embodiments may be combined. That is, after forming the recess 13 in the main surface of the substrate 1Sub, the p-type semiconductor region 4 is formed. Subsequently, after depositing the insulating film 23, the control gate electrode CG is formed.

  Further, the fifth embodiment and the fourth embodiment may be combined. In this case, the insulating films 19 and 23 may be deposited separately to achieve each purpose. However, when forming the protective insulating film 23 in the p-type semiconductor region 4 of the fourth embodiment, Alternatively, the spacer insulating film 19 may be used. That is, after the steps of FIGS. 8 to 11, the insulating film 23 is deposited on the main surface of the substrate 1 Sub, and impurities are introduced as in the fourth embodiment to form the p-type semiconductor region 4. The subsequent steps are the same as those described in FIG. In this case, the manufacturing process of the semiconductor device can be reduced as compared with the case where the protective insulating film 23 and the spacer insulating film 19 are separately deposited, and the manufacturing time and cost of the semiconductor device can be reduced. .

(Embodiment 6)
The sixth embodiment is a modification of the fifth embodiment, and an example in which the protective insulating film 23 is formed of another MIS gate insulating film on the same substrate will be described. According to the sixth embodiment, since the process can be shared, the manufacturing time and cost of the semiconductor device can be shortened.

  53 to 61 are cross-sectional views of main parts of the substrate 1Sub during the manufacturing process of the semiconductor device according to the sixth embodiment. The left side of each figure shows the main part cross-sectional view of the memory region, and the right side is a peripheral circuit. 2 is a cross-sectional view of a main part of a high breakdown voltage nMIS formation region for use.

  FIG. 53 is a cross-sectional view of the principal part of the substrate 1Sub (semiconductor wafer) after the steps described in FIGS. 8 to 11 and after the steps described in FIG. Further, on the right side of FIG. 53, a groove-type separation portion 25 called, for example, SGI (Shallow Groove Isolation) is formed on the main surface of the substrate 1Sub. The separation part 25 is formed by, for example, a silicon oxide film embedded in a groove dug in the substrate 1Sub. The isolation region 25 defines an active region.

  First, by subjecting such a substrate 1Sub to oxidation treatment, as shown in FIG. 54, a high-breakdown-voltage nMIS gate insulating film 23a is formed on the main surface of the substrate 1Sub. The gate insulating film 23a is made of, for example, silicon oxide having a thickness of about 2 to 8 nm, and is formed not only on the high breakdown voltage nMIS formation region but also on the main surface of the substrate 1Sub in the memory region. Subsequently, as shown in FIG. 55, an insulating film 23b made of, for example, silicon oxide having a thickness of about 13 nm is deposited on the main surface of the substrate 1Sub by a CVD method or the like, and then annealed to improve the film quality of the insulating film 23b. Apply processing. The insulating films 23a and 23b are high breakdown voltage nMIS gate insulating films. By depositing the insulating film 23b, the gate breakdown voltage of the high breakdown voltage nMIS can be improved. Thereafter, a resist pattern PR6 is formed on the main surface of the substrate 1Sub so that the nMIS formation region for selecting the memory region is exposed and covers the high breakdown voltage nMIS formation region by lithography. As shown in FIG. 56, the insulating films 23a and 23b are patterned. In the nMIS formation region for selecting the memory region, the insulating film 23a is removed, and the main surface of the substrate 1Sub is exposed. After that, as in the first embodiment, the main surface of the substrate 1Sub in the nMIS formation region for selection may be partially etched to form the recess 13 as a matter of course.

  Next, as shown in FIG. 57, the gate insulating film 3 made of, for example, silicon oxide is formed on the main surface of the substrate 1Sub by oxidizing the substrate 1Sub. This gate insulating film 3 becomes a gate insulating film of a low withstand voltage MIS such as a selection nMIS. The thickness of the gate insulating film 3 is thinner than the gate insulating films 23a and 23b of the high breakdown voltage nMIS, for example, about 2 to 5 nm. Subsequently, after a conductor film 21 made of, for example, low-resistance polysilicon is deposited on the main surface of the substrate 1Sub by a CVD method or the like, a resist pattern PR7 is formed thereon by a lithography technique. The resist pattern PR7 is a pattern that covers the control gate electrode formation region in the memory region and the gate electrode formation region in the high breakdown voltage nMIS formation region, and the others are exposed. Thereafter, by patterning the conductor film 21 by etching using the resist pattern PR7 as an etching mask, as shown in FIG. 58, the nMIS control gate electrode CG for selecting a memory cell and the high breakdown voltage nMIS gate of the peripheral circuit are formed. An electrode HG is formed. At this time, the conductor film 21a is left on the lower side surface of the side wall 8 on one side surface of the memory gate electrode MG. In the fifth embodiment, the conductor film 21a is left on the insulating films 23a and 23b. Therefore, in order to remove the conductor film 21a, a resist pattern PR8 is formed on the main surface of the substrate 1Sub so that the conductor film 21a is exposed and other areas are covered, and this is used as an etching mask. The conductive film 21a is removed as shown in FIG. 59 by performing the same etching process as that for removing the conductive film 21a of the fifth embodiment. As a result, since the laminated film of the insulating films 23a and 23b thicker than the gate insulating film 3 is provided under the conductive film 21a that remains after etching, made of polysilicon (between the conductive film 21a and the substrate 1Sub), the substrate It is possible to protect the main surface of the 1Sub and prevent the main surface of the substrate 1Sub from being etched. Accordingly, since the drain current Ids of the split gate electrode type memory cell can be prevented from decreasing, the high speed operation (high speed data reading) of the split gate electrode type memory cell can be maintained.

Next, after removing the resist pattern PR8, as shown in FIG. 60, for example, arsenic is introduced into the substrate 1Sub by an ion implantation method or the like while leaving the insulating films 23a and 23b. An n ⁻ type semiconductor region 2a is simultaneously formed. Subsequently, after an insulating film made of, for example, silicon oxide is deposited on the main surface of the substrate 1Sub by a CVD method or the like, this is etched back by an anisotropic dry etching method, so that as shown in FIG. Sidewalls 9 are simultaneously formed on the side surfaces of the control gate electrode CG and the memory gate electrode MG and the side surfaces of the gate electrode HG of the high breakdown voltage nMIS. At this time, the insulating films 23a and 23b exposed from the sidewall 9 are also removed. Thereafter, for example, phosphorus is introduced into the substrate 1Sub by an ion implantation method or the like, thereby forming the n ⁺ type semiconductor region 2b for the memory cell and the high breakdown voltage nMIS at the same time. In this manner, the drain region Drm and source region Srm of the memory cell MC1 and the drain region Drh and source region Srh of the high breakdown voltage nMISQnh are simultaneously formed.

  As described above, in the sixth embodiment, since the protective insulating films 23a and 23b are formed of the gate insulating films 23a and 23b of the high breakdown voltage nMISQnh, the process can be shared. Shortening and cost reduction are possible.

  In the sixth embodiment, the protective insulating film for removing the remaining conductor film 21a is formed of a laminated film of the insulating film 23a formed by the oxidation method and the insulating film 23b formed by the CVD method. However, the protective insulating film may be formed by either one of the insulating films 23a and 23b.

(Embodiment 7)
When the protective insulating film of the fifth and sixth embodiments is formed by the CVD method, the gate electrode formed at the beginning of the memory cell is covered with the protective insulating film, so that the silicide is formed on the gate electrode. A layer cannot be formed.

  Therefore, in the seventh embodiment, a silicide layer is formed in a contact hole portion to which a wiring such as a first layer wiring contacts and is connected among the gate electrodes formed first. Thereby, the contact resistance between the gate electrode formed first and the wiring such as the first layer wiring can be reduced.

  FIG. 62 shows a plan view of relevant parts of a memory area as a specific example of the seventh embodiment. 62 corresponds to the cross-sectional view taken along line X1-X1 in FIG. 44 to FIG. 52 of the fifth embodiment or the left cross-sectional view of FIG. 53 to FIG. 61 of the sixth embodiment. On the upper surface of the control gate electrode CG and the memory gate electrode MG, hatched hatching is given to the portion where the silicide layer 14 is formed by the salicide process. A silicide layer 14 is formed on the entire upper surface of the control gate electrode CG formed later. On the other hand, on the upper surface of the memory gate electrode MG formed first, the silicide layer 14 is formed only in the region where the contact hole CNT is disposed. Thereby, the contact resistance between the plug PLG in the contact hole CNT and the memory gate electrode MG can be reduced. The contact hole CNT is electrically connected to the first layer wiring.

  In order to form such a configuration, the pattern shape of the resist pattern PR5 in FIG. 45 is changed to the region of the memory gate electrode MG in which the contact hole CNT is disposed in the steps described with reference to FIGS. The pattern shape is such that the protective insulating film 23 portion is also exposed, and the protective insulating film 23 portion and the insulating film 7 portion in the region where the contact hole CNT of the memory gate electrode MG is disposed are also etched. Remove it. Then, after the process described with reference to FIG. 51, the salicide process is performed in a state where the upper surface of the control gate electrode CG and the region where the contact hole CNT of the memory gate electrode MG is disposed are exposed, thereby the control gate electrode The silicide layer 14 is formed in the region where the contact hole CNT is disposed on the entire upper surface of the CG and the upper surface of the memory gate electrode MG. Such a method is the same when the seventh embodiment is applied to the sixth embodiment.

  As a method for forming the silicide layer 14, for example, a refractory metal film such as a cobalt (Co) film, a titanium (Ti) film, or nickel (Ni) is formed by sputtering on the entire main surface of the substrate 1Sub. Thereafter, heat treatment is performed to react the polycrystalline silicon film of the control gate electrode CG and the memory gate electrode MG with the refractory metal film. Thereafter, the unreacted refractory metal film is removed, and a silicide layer 14 which is a metal / semiconductor reaction layer is formed.

  The seventh embodiment and the fifth and sixth embodiments can also be applied to the split gate type memory cells of FIGS.

(Embodiment 8)
In the eighth embodiment, an example for solving another problem that the present inventors have found for the first time with respect to the memory cell of the split gate electrode type will be described. The problem is that the erase operation becomes difficult as a result of the width (short dimension) of the charge storage layer CSL becoming larger than the width (short dimension) of the memory gate electrode.

  First, the problem will be described with reference to FIGS. FIG. 63 is a fragmentary cross-sectional view of the substrate 1Sub after the step of forming the memory gate electrode MG. In this step, usually, the memory gate electrode MG made of low-resistance polysilicon or the like is formed by a dry etching method, and then the charge storage layer CSL is selectively etched by wet etching. Here, the etching is performed in a state in which the end in the width direction (direction along the main surface of the substrate 1Sub, the short direction) of the memory gate electrode MG and the end in the width direction of the charge storage layer CSL coincide with each other. It has been broken. However, if a subsequent oxidation step is performed in such a state, both ends in the width direction of the memory gate electrode MG are oxidized as shown in FIGS. 64 and 65 (see the arrows in FIG. 65), while charge accumulation is performed. Since the layer CSL is made of silicon nitride and remains unoxidized, the substantial width (short dimension) of the memory gate electrode MG becomes shorter than the width of the charge storage layer CSL, and both end portions in the width direction of the charge storage layer CSL. Is protruded to the outside in the width direction of the memory gate electrode MG. Such protruding portions E at both ends of the charge storage layer CSL are less affected by the electric field from the memory gate electrode MG because the physical distance from the memory gate electrode MG is longer. For this reason, during the erasing operation of the memory cell MC, there arises a problem that it is difficult to extract the charge accumulated in the protruding portion E at both ends of the charge storage layer CSL. This problem does not depend on the direction in which the charge of the charge storage layer CSL is extracted during the erase operation of the memory cell MC. That is, in this embodiment, the erase operation is performed by extracting electrons to the memory gate electrode MG. However, the same problem occurs in a memory cell that extracts electrons to the substrate 1Sub. This problem also occurs in the memory cells shown in FIGS.

  Therefore, the eighth embodiment includes a step of overetching the charge storage layer to make it smaller than the upper memory gate electrode. Thereby, it is possible to finally prevent a part of the charge storage layer from protruding outside the memory gate electrode. That is, it is possible to prevent a portion that makes it difficult to erase data (withdrawing charges) from being formed in the charge storage layer. Therefore, the operation speed (data erasing speed) of the split gate electrode type memory cell can be improved. In addition, since the rate of occurrence of defective erasing operations can be reduced, the yield of semiconductor devices can be improved. A specific example will be described with reference to a cross-sectional view of the main part of the substrate 1Sub in the memory region during the manufacturing process of the semiconductor device of FIGS.

  First, in the process of forming the memory gate electrode of the memory cell, it is difficult to etch the charge storage layer (silicon nitride) and the memory gate electrode (polysilicon) at the same time, so the etching process is performed in two stages. That is, after the steps described in the first embodiment with reference to FIGS. 8 and 9, the memory gate electrode MG and the insulating film 5t are formed by dry etching using the resist pattern PR9 as an etching mask as shown in FIG. Pattern. At this time, the etching selectivity between polysilicon and silicon nitride is increased, and the etching process is performed under the condition that polysilicon is more easily etched.

  Subsequently, after removing the resist pattern PR9, the substrate 1Sub is subjected to a wet etching process using hot phosphoric acid or the like to selectively etch the charge storage layer CSL as shown in FIG. At this time, in the eighth embodiment, the overetching process is performed so that the exposed side surfaces at both ends in the width direction of the charge storage layer CSL are located inside the exposed side surfaces at both ends in the width direction of the memory gate electrode MG. As a result, the side surfaces at both ends in the width direction of the memory gate electrode MG are oxidized in the subsequent oxidation treatment step, so that the side surfaces at both ends in the width direction of the charge storage layer CSL are the side surfaces at both ends in the width direction of the memory gate electrode. It is possible to prevent such a problem that it is located outside of. The side etching amount (undercut amount or overetch amount) of the charge storage layer CSL can be changed by adjusting the wet etching time with hot phosphoric acid.

  Thereafter, through the same process as described in the first embodiment, as shown in FIGS. 68 and 69, a split gate type memory cell MC1 is formed. FIG. 69 shows an enlarged cross-sectional view of the main part of FIG. Also in the case of the eighth embodiment, as shown in FIGS. 68 and 69, the side surface portion of the memory gate electrode MG is oxidized by the oxidation process after the patterning of the charge storage layer CSL, and the width direction dimension of the memory gate electrode MG is reduced. It is slightly smaller than the design value. As described above, in the eighth embodiment, the width direction dimension of the charge storage layer CSL is reduced in anticipation of the memory gate electrode MG becoming smaller due to oxidation. The side surfaces at both end portions in the width direction substantially coincide with the side surfaces at both end portions in the width direction of the memory gate electrode MG, and do not protrude from the outside of the memory gate electrode MG. Therefore, at the time of erasing operation of the split gate type memory cell MC1, since there is no portion in the charge storage layer CSL where it is difficult to extract charges, the data erasing speed can be improved. In addition, since there is no portion in the charge storage layer CSL where it is difficult to extract charges, the occurrence rate of the erase operation failure can be reduced and the yield of the semiconductor device can be improved.

  In the above description, the case where the both ends in the width direction of the charge storage layer CSL substantially coincide with the both ends in the width direction of the memory gate electrode MG has been described. As long as the CSL does not protrude outside the memory gate electrode MG, the width of the charge storage layer CSL is finally smaller than the width of the memory gate electrode MG as shown in the enlarged cross-sectional view of the main part of the memory cell in FIG. The structure in which the both ends in the width direction of the charge storage layer CSL are located inside the both ends in the width direction of the memory gate electrode MG is not excluded. In the case of this structure, the entire planar area of the charge storage layer CSL is included in the entire planar area of the memory gate electrode MG. For this reason, it is possible to improve the certainty of the charge extraction at the time of erasing.

(Embodiment 9)
The ninth embodiment is a modification of the eighth embodiment, and describes an example in which the method of the eighth embodiment is applied to the method of forming the split gate electrode type memory cell MC2 of FIG. To do.

  First, after the steps described with reference to FIGS. 23 to 25, as shown in FIG. 71, an n-type semiconductor region 6 is formed on the substrate 1Sub in a self-aligning manner using the control electrode CG as a mask. Subsequently, the insulating film 5b, the charge storage layer CSL, the insulating film 5t, and the conductor film 11 are sequentially deposited on the substrate 1Sub from the lower layer by CVD or the like, and then a resist pattern PR10 for forming a memory gate electrode is formed thereon by lithography. Form by technology. Thereafter, using the resist pattern PR10 as an etching mask, the conductor film 11 and the insulating film 5t exposed therefrom are etched in the same manner as in the eighth embodiment, and then the resist pattern PR10 is removed. As shown in FIG. The gate electrode MG and the insulating film 5t are patterned.

  Next, as in the eighth embodiment, the charge storage layer CSL is selectively etched by wet etching, thereby patterning the charge storage layer CSL as shown in FIG. At this time, also in the ninth embodiment, the over-etching process is performed so that both exposed side surfaces of the charge storage layer CSL are located inside the exposed side surfaces of the memory gate electrode MG. As a result, the problem that the side surfaces of both ends in the width direction of the charge storage layer CSL are located outside the side surfaces of both ends in the width direction of the memory gate electrode MG can be prevented as in the eighth embodiment. .

  Thereafter, through the same process as described in the third embodiment, a split gate electrode type memory cell MC2 is formed as shown in FIG. Also in the ninth embodiment, as shown in FIG. 74, the side surface portion of the memory gate electrode MG is oxidized by the oxidation process after the patterning of the charge storage layer CSL, and the width direction dimension of the memory gate electrode MG is larger than the design value. Although slightly smaller, in the ninth embodiment as well, the width direction dimension of the charge storage layer CSL is reduced in anticipation of this, so that both end portions in the width direction of the charge storage layer CSL are finally obtained. The side surfaces of the memory gate electrode MG substantially coincide with the side surfaces of both ends in the width direction of the memory gate electrode MG, and do not protrude outside the memory gate electrode MG. As a result, in the erasing operation of the split gate type memory cell MC2, the data erasing speed can be improved as in the eighth embodiment. In addition, the rate of occurrence of defective erase operation can be reduced, and the yield of semiconductor devices can be improved. Of course, also in the ninth embodiment, as described with reference to FIG. 70, both end portions of the charge storage layer CSL may finally be positioned inside both end portions of the memory gate electrode MG.

(Embodiment 10)
The tenth embodiment is a modification of the eighth embodiment, and describes an example in which the method of the eighth embodiment is applied to the method of forming the split gate electrode type memory cell MC3 of FIG. To do.

  First, as in the ninth embodiment, after the process of depositing the conductor film 11 of FIG. 71 is performed, the conductor film 11 is etched back by anisotropic dry etching, as shown in FIG. The sidewall 11a of the conductor film 11 is formed on the side surface side of the control gate electrode CG. Subsequently, as shown in FIG. 76, a resist pattern PR11 for forming the memory gate electrode MG is formed on the main surface of the substrate 1Sub by lithography. After that, using the resist pattern PR11 as an etching mask, the conductor film 11a and the insulating film 5t exposed therefrom are etched in the same manner as in the eighth and ninth embodiments to pattern the memory gate electrode MG (11a) and the insulating film 5t. .

  Next, after removing the resist pattern PR11, as in the eighth and ninth embodiments, the charge storage layer CSL is selectively etched by wet etching, thereby forming the charge storage layer CSL as shown in FIG. Pattern. At this time, in the tenth embodiment, an over-etching process is performed so that both exposed side surfaces of the charge storage layer CSL enter and are recessed below the insulating film 5t. This prevents the problem that the side surface of one end in the width direction of the charge storage layer CSL is located outside the side surface of one end in the width direction of the memory gate electrode MG, as in the eighth and ninth embodiments. Can do.

  Thereafter, through the same process as described in the third embodiment, as shown in FIG. 78, a split gate type memory cell MC3 is formed. Also in the case of the tenth embodiment, as shown in FIG. 78, the surface portion of the memory gate electrode MG is oxidized by the oxidation process after the patterning of the charge storage layer CSL, and the width direction dimension of the memory gate electrode MG is larger than the design value. Although it is slightly smaller, in the tenth embodiment as described above, the width direction dimension of the charge storage layer CSL is reduced in anticipation of this, so that one end portion in the width direction of the charge storage layer CSL is finally obtained. Of the memory gate electrode MG substantially coincides with the side surface of one end in the width direction of the memory gate electrode MG, and does not protrude outside the memory gate electrode MG. As a result, in the erasing operation of the split gate type memory cell MC3, the data erasing speed can be improved as in the eighth and ninth embodiments. In addition, the rate of occurrence of defective erase operation can be reduced, and the yield of semiconductor devices can be improved. Of course, also in the tenth embodiment, as described with reference to FIG. 70, the side surface of one end portion of the charge storage layer CSL may finally be positioned inside the side surface of one end portion of the memory gate electrode MG.

(Embodiment 11)
In the eleventh embodiment, an example for solving the first problem found by the present inventor for the split gate type memory cell will be described. The problem is that when the impurity concentration of the memory gate electrode is high, the data erasing time is delayed.

  FIG. 79 shows how the erase characteristic of the split gate electrode type memory cell depends on the impurity concentration of the memory gate electrode. For example, n-type low-resistance polysilicon is used as the memory gate electrode material. The electric charge contributing to data storage is assumed to be electrons. As shown by the arrows in FIG. 79, it can be seen that the erase speed decreases as the impurity ion implantation concentration into the memory gate electrode is increased. This is because, in the case of an erasing method in which electrons contributing to data storage are extracted to the memory gate electrode side, both the effect of extracting electrons and the effect of recombination with holes injected from the memory gate electrode side are erase operations. It is assumed that it has an influence on FIG. 80 illustrates an energy band diagram. White circles indicate holes and black circles indicate electrons. The erase speed changes because holes injected from the memory gate electrode MG recombine with electrons extracted from the charge storage layer CSL to generate electron-hole pairs, or electrons in the charge storage layer CSL This is considered to be caused by flowing (tunneling) to the memory gate electrode MG side through a defect level in the insulating film 5t in the vicinity of the memory gate electrode MG. Therefore, it is possible to control the erase speed by depleting the interface between the memory gate electrode MG and the insulating film 5t.

  As a specific example, the memory gate electrode MG of the memory cell MC of the first to tenth embodiments is formed of n-type polysilicon, and the concentration of the n-type impurity in the memory gate electrode MG is lowered. Alternatively, the memory gate electrode MG of the memory cell MC of the first to tenth embodiments is formed of n-type polysilicon, and the impurity concentration in the interface region between the memory gate electrode MG and the insulating film 5t is set to the same memory gate electrode MG. The impurity concentration is made lower than that of other regions. That is, an impurity concentration difference is intentionally formed in the same memory gate electrode MG, and the impurity concentration in one region (first region) on the charge storage layer CSL side of the memory gate electrode MG is the other region (first region). 2 region) is relatively low. By adopting such a configuration, both effects of extracting electrons from the charge storage layer CSL and recombining with electrons contributing to data storage due to the injection of holes from the memory gate electrode MG side are generated. Since the data erasure proceeds by this, the erasing speed can be increased. Also in this case, the rate of occurrence of defective erasing operations can be reduced, so that the yield of the semiconductor device can be improved.

According to the study of the present inventors, the concentration of the n-type impurity in the entire memory gate electrode MG or the interface region with the insulating film 5t is, for example, about 1 × 10 ¹⁸ / cm ³ to 2 × 10 ²⁰ / cm ³ , Preferably, it is about 8 × 10 ¹⁹ / cm ^{3 to} 1.5 × 10 ²⁰ / cm ³ . This is because if it is lower than the lower limit value, depletion occurs or the resistance value of the memory gate electrode MG increases, causing other operational problems including erasure. This is because the problem arises.

Further, in a general split gate electrode type MONOS structure memory cell investigated by the present inventor, the memory gate electrode is formed in the same process as the other MIS gate electrodes constituting the peripheral circuit and the like. The concentration of the n-type impurity contained in the memory gate electrode is, for example, 2.5 × 10 ²⁰ / cm ³ or higher, which is higher than that of the eleventh embodiment. On the other hand, in the eleventh embodiment, the memory gate electrode MG and the gate electrode of the other MIS are formed separately. Then, the impurity concentration of the memory gate electrode MG is lowered as described above. Specifically, a polycrystalline silicon film into which a relatively low-concentration n-type impurity is introduced that becomes the memory gate electrode MG, and a relatively high-concentration n-type impurity that becomes the gate electrode of the other MIS. The introduced polycrystalline silicon film is formed separately by CVD, and then patterned. Of course, the memory gate electrode MG and the other MIS gate electrode on the same substrate are simultaneously patterned, and an n-type impurity is introduced into the other MIS gate electrode side by an ion implantation method or the like in a later step. The n-type impurity concentration in the MIS gate electrode may be relatively high. The formation method by this ion implantation method has an advantage that it is easier to control the impurity concentration of the gate electrode of the other MIS than the separate CVD method described above. In any case, in the eleventh embodiment, the concentration of the n-type impurity of the memory gate electrode MG of the memory cell MC and the concentration of the n-type impurity of the gate electrode of the other MIS formed on the same substrate are determined. The concentration of the n-type impurity in the memory gate electrode MG is intentionally lower than the concentration of the n-type impurity in the gate electrode of the other MIS.

From the same idea, by making the memory gate electrode MG of the memory cell MC of the first to tenth embodiments p-type, it is possible to increase the erase speed by the same operation as described above. According to the study by the present inventors, when the memory gate electrode MG is p-type, the concentration of the p-type impurity in the memory gate electrode MG is preferably about 1 × 10 ¹⁸ / cm ³ or more, for example. Therefore, also in this case, when there is an MIS having a p-type gate electrode in another MIS formed on the same substrate, the concentration of the p-type impurity in the memory gate electrode MG of the memory cell MC is different from the above-mentioned other In some cases, the concentration is intentionally higher than the concentration of the p-type impurity in the MIS p-type gate electrode.

(Embodiment 12)
In the twelfth embodiment, the impurity concentration difference is intentionally formed in the same memory gate electrode MG described in the eleventh embodiment, and the region of the memory gate electrode MG on the charge storage layer CSL side is the same. A case where the impurity concentration is relatively low will be described.

  FIG. 81 shows an example of a fragmentary sectional view of the memory cell MC1 of the twelfth embodiment. Here, the memory gate electrode MG has a laminated structure of the conductor layers 11a and 11b. The conductor layers 11a and 11b are both made of n-type low-resistance polysilicon, but the impurity concentration is intentionally different, and the conductor layer (first region) 11a on the lower layer side, that is, the side in contact with the insulating film 5t. The concentration of the n-type impurity is lower than the concentration of the n-type impurity in the conductor layer (second region) 11b thereon. Thereby, the erasing speed can be increased as in the eleventh embodiment. In addition, the yield of the semiconductor device can be improved. Furthermore, according to the twelfth embodiment, the case where the entire memory gate electrode MG is formed of a conductive film made of a low-concentration n-type impurity is exemplified in the eleventh embodiment. Since the conductor layer 11b having a high concentration is provided, the resistance of the memory gate electrode MG and the contact resistance with the wiring can be reduced.

  In order to form such conductor layers 11a and 11b, for example, the following first and second methods can be cited. The first method is a method in which the conductor films 11a and 11b are separately deposited by the CVD method. That is, the conductor films 11a and 11b are sequentially deposited from the lower layer during the conductor film 11 deposition step described with reference to FIG. At this time, the concentration of each n-type impurity in the conductor films 11a and 11b is adjusted. Thereafter, after depositing the insulating film 7 on the conductor film 11b, the laminated film is patterned in the same manner as described above, thereby forming the memory gate electrode MG. The second method is a method of forming an impurity concentration distribution by an ion implantation method. That is, when the n-type impurity is introduced into the conductor film 11 by depositing the conductor film 11 as described with reference to FIG. 9 and the like, the conditions such as the ion implantation energy and the dose amount are changed. By introducing, conductor layers 11 a and 11 b having different n-type impurity concentrations are formed in the conductor film 11. Then, after depositing the insulating film 7 on the conductor film 11, the laminated film is patterned in the same manner as described above, thereby forming the memory gate electrode MG. When this ion implantation method is employed, the impurity concentration and formation position of each conductor layer 11a, 11b can be easily controlled, so that the conductor layers 11a, 11b can be formed in a state closer to the aim.

  In the twelfth embodiment, the conductor layers 11a and 11b may be formed of p-type low resistance polysilicon. In this case, the concentration of the p-type impurity in the lower conductor film 11a in contact with the insulating film 5t is higher than the concentration of the p-type impurity in the upper conductor film 11b.

(Embodiment 13)
In the thirteenth embodiment, a modification of the twelfth embodiment will be described.

  FIG. 82 shows an example of a fragmentary sectional view of the memory cell MC1 of the thirteenth embodiment. Here, the memory gate electrode MG has a three-layer structure of the conductor layers 11a, 11b, and 11c. The conductor layers 11a, 11b, and 11c are all made of n-type low-resistance polysilicon, but the impurity concentration is intentionally different, and the conductor layer (first side in contact with the insulating film 5t) (first side) The concentration of the n-type impurity in the (region) 11a and the uppermost conductor layer 11c is lower than the concentration of the n-type impurity in the intermediate conductor layer (second region) 11b. The concentration of the n-type impurity in the conductor films 11a and 11c may be the same or different. Thereby, the erasing speed can be increased as in the eleventh and twelfth embodiments. In addition, the yield of the semiconductor device can be improved.

  The reason why the concentration of the uppermost conductor layer 11c is reduced is that when a high-concentration polysilicon layer exists at the top, a watermark remains on the surface of the high-concentration polysilicon layer after the cleaning process, or the high-concentration polysilicon layer This is to avoid abnormal oxidation, which may occur. That is, according to the thirteenth embodiment, since the impurity concentration of the uppermost conductor layer 11c is lowered, problems during the manufacturing process of the semiconductor device can be avoided, so that the reliability and yield of the semiconductor device can be improved. it can.

  The method for forming the conductor layers 11a, 11b, and 11c is the same as that in the twelfth embodiment, and a description thereof will be omitted.

  In the thirteenth embodiment, the conductor layers 11a, 11b, and 11c may be formed of p-type low resistance polysilicon. In this case, the concentration of the p-type impurity in the lower conductive film 11a in contact with the insulating film 5t is higher than the concentration of the p-type impurity in the upper conductive films 11b and 11c.

(Embodiment 14)
In the fourteenth embodiment, a modification of the twelfth embodiment will be described.

  FIG. 83 shows an example of a fragmentary sectional view of the memory cell MC1 of the fourteenth embodiment. Here, the memory gate electrode MG has a stacked structure of the conductor layer 11 and the silicide layer 14. The conductor film 11 is made of, for example, n-type low-resistance polysilicon, and the concentration of the n-type impurity is lowered as described in the eleventh and twelfth embodiments. The silicide layer 14 is made of, for example, cobalt silicide or tungsten silicide. As described above, even if the concentration of the n-type impurity in the conductor film 11 is lowered as described in the eleventh and twelfth embodiments, the memory gate is provided by providing the low resistance silicide layer 14 on the conductor film 11. The total resistance of the electrode MG can be kept low. However, also in Embodiment 14, the conductor film 11 may be formed of p-type polysilicon and the silicide layer 14 may be provided thereon.

(Embodiment 15)
In the fifteenth embodiment, another modification of the twelfth embodiment will be described.

  FIG. 84 shows an example of a fragmentary sectional view of the memory cell MC1 of the fifteenth embodiment. Here, the memory gate electrode MG has a three-layer structure including the conductor layer 11, the barrier metal layer 28, and the metal layer 29. The conductor film 11 is made of, for example, n-type low-resistance polysilicon, and the concentration of the n-type impurity is lowered as described in the eleventh and twelfth embodiments. The barrier metal layer 28 is made of, for example, tungsten nitride (WN). The metal layer 29 is made of, for example, tungsten. As described above, even if the concentration of the n-type impurity in the conductor film 11 is lowered as described in the eleventh and twelfth embodiments, the memory gate is provided by providing the low-resistance metal layer 29 on the conductor film 11. The total resistance of the electrode MG can be greatly reduced. However, in the fifteenth embodiment, the conductor film 11 may be formed of p-type polysilicon and the metal layer 29 may be provided on the conductor film 11 with the barrier metal layer 28 interposed therebetween.

(Embodiment 16)
In the sixteenth embodiment, an example for solving still another problem found by the present inventor for the split gate electrode type memory cell will be described. The problem is that the drain current of the memory cell tends to deteriorate.

  In a memory cell having a MONOS structure that uses silicon oxynitride (SiON) as a material for the charge storage layer CSL or the insulating film 5t, when a rewrite operation in which an interface state is formed by an erase operation is performed, the drain current Ids of the memory cell is increased. May decrease. Therefore, it is necessary to design the memory cell in consideration of current degradation when the memory cell is turned on.

  That is, in the sixteenth embodiment, in the split gate electrode type memory cell in which data is erased by the operation of extracting the charge stored in the charge storage layer to the memory gate electrode side, the state where the write level is the smallest is the memory cell. The initial threshold voltage Vth is set to be higher. As a result, the drain current can be allowed to deteriorate due to data writing and erasing. Therefore, the number of times of rewriting compensation of data in the memory cell can be improved, and the life of the memory cell can be improved.

  The structure (including read, write, and erase operations) of the memory cell of the sixteenth embodiment is the same as that described in the first to fifteenth embodiments. FIG. 85 shows the write and erase states of the memory cell of the sixteenth embodiment. The initial state In is a state in which the threshold voltage Vth is low, but the erase state after writing once is not set to the same state as the initial state In, but is set to a higher state. In the case of multilevel storage, the state in which the write level is the lowest is a state in which the state is higher than the initial state In. In the case of this method, even when the drain current Ids is reduced due to deterioration such as rewriting, there is a margin to erase deeply, so that the deterioration in the operation of the semiconductor device cannot be seen.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  For example, in the first to sixteenth embodiments, the case where the present invention is applied to a memory cell having a MONOS structure has been described. The present invention can also be applied to a memory cell having an MNOS (Metal Nitride Oxide Semiconductor) structure in which a polysilicon film (or a polycide or polymetal structure film similar to the above) is stacked in order from the lower layer.

  In the first to tenth embodiments, the case where the memory cell structure in which the charge contributing to the data storage is extracted to the memory gate electrode in the erasing operation is described. However, the present invention is not limited to this. The present invention can also be applied to a memory cell structure in which charges that contribute to storage are extracted to the substrate side.

  In the first to sixteenth embodiments, the case where the charge storage layer is formed of a nitride film has been described. However, the present invention is not limited to this. For example, a plurality of nano-particles formed in a dot shape in a silicon oxide film are used. It is good also as a structure which has a crystal | crystallization. The nanocrystal is a portion where charges contributing to data are captured, like a floating gate electrode of a general nonvolatile memory. Each nanocrystal is formed by, for example, a CVD method. Since each nanocrystal is formed in a physically separated state (that is, has a discrete trap level), only a part of electric charge is lost even if there is a leak path during data retention, and data retention characteristics Is excellent. In addition, even when erasing and writing data, the characteristics are averaged among many nanocrystals, so it is difficult to be affected by structural variations such as nanocrystal diameters or insulating films, or stochastic behavior. The yield of semiconductor devices having cells can be improved.

  Of course, each of the first to sixteenth embodiments may be combined with one or more of the other embodiments.

  In the above description, the case where the invention made mainly by the present inventor is applied to an IC card which is a field of use as the background has been described. However, the present invention is not limited to this, and a mobile object such as a mobile phone is used. The present invention can also be applied to a memory built in an information processing apparatus such as a communication device or a personal computer.

  The present invention is effective when applied to a semiconductor device manufacturing method and semiconductor device technology.

1Sub Semiconductor substrate 2a Semiconductor region 2b Semiconductor region 3 Gate insulating film 4 Semiconductor region 5b, 5t Insulating film 6 Semiconductor region 7 Insulating film 8 Side wall 9 Side wall 10 Insulating film 11 Conductive film 11a Conductive layer 11b Conductive layer 13 Recess 14 Silicide layer DESCRIPTION OF SYMBOLS 15 Insulating film 17 Conductive film 19 Insulating film 21 Conductive film 21a Conductive film 23 Insulating film 23a, 23b Insulating film 25 Separation part 28 Barrier metal layer 29 Metal layer MC, MC1-MC3 Memory cell D Drain electrode Drm Drain area S Source electrode Srm Source region CG Control gate electrode MG Memory gate electrode CSL Charge storage layer Qnc n-channel type MIS • FET
Qnm n-channel MIS • FET
PR1 to PR11 Photoresist pattern CNT Contact hole PLG Plug M1 First layer wiring HG Gate electrode

Claims (9)

In a semiconductor device having a nonvolatile memory cell having a first field effect transistor for memory adjacent to each other and a second field effect transistor for memory cell selection,
A gate electrode of the first field effect transistor, the first gate electrode formed on the semiconductor substrate;
A first gate insulating film of the first field effect transistor formed between the semiconductor substrate and the first gate electrode and having a charge storage layer for storing charges contributing to data storage;
A gate electrode of the second field effect transistor, the second gate electrode formed on the semiconductor substrate;
A second gate insulating film of the second field effect transistor formed between the semiconductor substrate and the second gate electrode;
A first semiconductor region formed in the semiconductor substrate adjacent to the first gate electrode;
A second semiconductor region formed in the semiconductor substrate adjacent to the second gate electrode;
Have
The first gate electrode, the second gate electrode, the first semiconductor region, and the second semiconductor region have a first conductivity type impurity,
The concentration of the first conductivity type impurity in the first gate electrode is higher on the upper surface side of the first gate electrode than in the vicinity of the interface between the first gate electrode and the first gate insulating film. A semiconductor device. In a semiconductor device having a nonvolatile memory cell having a first field effect transistor for memory adjacent to each other and a second field effect transistor for memory cell selection,
A gate electrode of the first field effect transistor, the first gate electrode formed on the semiconductor substrate;
A first gate insulating film of the first field effect transistor formed between the semiconductor substrate and the first gate electrode and having a charge storage layer for storing charges contributing to data storage;
A gate electrode of the second field effect transistor, the second gate electrode formed on the semiconductor substrate;
A second gate insulating film of the second field effect transistor formed between the semiconductor substrate and the second gate electrode;
A first semiconductor region formed in the semiconductor substrate adjacent to the first gate electrode;
A second semiconductor region formed in the semiconductor substrate adjacent to the second gate electrode;
Have
The second gate electrode, the first semiconductor region, and the second semiconductor region have impurities of a first conductivity type,
The first gate electrode has a second conductivity type impurity which is a conductivity type opposite to the first conductivity type,
The impurity of the second conductivity type of the first gate electrode is lower in concentration on the upper surface side of the first gate electrode than in the vicinity of the interface between the first gate electrode and the first gate insulating film. A semiconductor device characterized by the above. The semiconductor device according to claim 1,
The first gate electrode comprises a first polysilicon and a second polysilicon formed on the first polysilicon,
The semiconductor device according to claim 1, wherein a concentration of the first conductivity type impurity of the second polysilicon is higher than a concentration of the first conductivity type impurity of the first polysilicon. The semiconductor device according to claim 3.
The semiconductor device according to claim 1, wherein a film thickness of the first polysilicon is thinner than a film thickness of the second polysilicon. The semiconductor device according to claim 2,
The first gate electrode comprises a first polysilicon and a second polysilicon formed on the first polysilicon,
The semiconductor device according to claim 1, wherein a concentration of the second conductivity type impurity of the second polysilicon is lower than a concentration of the second conductivity type impurity of the first polysilicon. The semiconductor device according to claim 5.
The semiconductor device according to claim 1, wherein a film thickness of the first polysilicon is thinner than a film thickness of the second polysilicon. The semiconductor device according to claim 1 or 2,
The semiconductor device according to claim 1, wherein the charge storage layer includes discrete trap levels. The semiconductor device according to claim 1 or 2,
A semiconductor device, wherein the charge storage layer is made of silicon nitride. The semiconductor device according to claim 1 or 2,
A semiconductor device, wherein data is erased by drawing out charges in the charge storage layer to the first gate electrode side.

Images