Patent Application
20050002236
This non-provisional application claims priority under 35 U.S.C. 6119(a) on Patent Application No. 2003-142492 filed in Japan on May 20, 2003, the entire contents of which are hereby incorporated by reference.
1. Field of the Invention
The present invention relates to a semiconductor storage device comprising a plurality of memory elements, and a mobile electronic apparatus comprising the semiconductor storage device. More particularly, the present invention relates to a semiconductor storage device, in which multiple bits (2 bits or more) of data can be stored in at least one of two memory function sections contained in a memory element, and a mobile electronic apparatus comprising the semiconductor storage device.
2. Description of the Related Art
One representative type of conventional non-volatile memory is the flash memory.
In the flash memory, as shown in
The memory cell holds data depending on the more or less of the amount of charge in the floating gate 902. A plurality of memory cells are arranged into a memory cell array. In the memory array, data is rewritten into or read from a desired memory cell by applying a prescribed voltage thereto via a particular word line and bit line selected.
Such a flash memory has drain current (Id) vs. gate voltage (Vg) characteristics as shown in
However, in this flash memory, the insulating film 907 is functionally required to separate the floating gate 902 from the word line 903. In addition, it is difficult to obtain a thin gate insulating film, because the thinner film is likely to cause charge leakage from the floating gate 902. Therefore, it is difficult to practically obtain a thin insulating film 907 and a thin gate insulating film, which is an obstacle to the miniaturization of memory ealls.
According to an aspect of the present invention, a semiconductor storage device is provided, which comprises: a memory array comprising a plurality of memory elements, each of the plurality of memory elements comprising a gate electrode provided via a gate insulating film on a semiconductor layer, a channel region provided under the gate electrode, a first diffusion region and a second diffusion region provided on opposite sides of the channel region and having a conductivity type opposite to that of the channel region, and a first memory function section and a second memory function section provided on opposite sides of the gate electrode and having a function of retaining charges, the gate electrode of each memory element being connected to one of a plurality of word lines, the first diffusion region and the second diffusion region of each memory element each being connected to one of a plurality of bit line is a row decoder for selecting one of the plurality of word lines in accordance with a row address; and a write control circuit for applying a write pulse to a bit line, which is connected to one of the first diffusion region and the second diffusion region of the memory element connected to the selected word line, in accordance with a column address. The write control circuit controls the application of the write pulse so that a quantity of charges retained in one of the first memory function section and the second memory function section of the memory element corresponds to a value of multibit data.
In one embodiment of this invention, at least a portion of the first memory function section or the second memory function section of the memory element overlaps a portion of the first diffusion region or the second diffusion region.
In one embodiment of this invention, at least one of the first memory function section and the second memory function section of the memory element comprises a charge retaining film having a surface substantially parallel to a surface of the gate insulating film and having a function of retaining charges.
In one embodiment of this invention, at least one of the first memory function section and the second memory function section of the memory element comprises a charge retaining film having a function of retaining charges, the charge retaining film extending substantially parallel to a side of the gate electrode.
In one embodiment of this invention, a semiconductor layer of the memory element has a region in a vicinity of the diffusion region, the region having a concentration higher than that of a vicinity of a surface of the semiconductor layer under the gate electrode.
In one embodiment of this invention, the memory element comprises: a charge retaining film having a surface substantially parallel to a surface of the gate insulating film and having a function of retaining charges; and an insulating film separating the charge retaining film from a channel region or a semiconductor layer, the thickness of the insulating film being smaller than that of the gate insulating film and being 0.8 nm or more.
In one embodiment of this invention, at least a portion of the first memory function section and the second memory function section of the memory element overlaps a portion of the first diffusion region or the second diffusion region, at least one of the first memory function section and the second memory function section of the memory element comprises a charge retaining film having a function of retaining charges, and the charge retaining film is an insulating film.
According to another aspect of the present invention, a semiconductor storage device is provided, which comprises: a memory array comprising a plurality of memory elements, each of the plurality of memory elements comprising a gate electrode provided via a gate insulating film on a semiconductor layer, a channel region provided under the gate electrode, a first diffusion region and a second diffusion region provided on opposite sides of the channel region and having a conductivity type opposite to that of the channel region, and a first memory function section and a second memory function section provided on opposite sides of the gate electrode and having a function of retaining charges, the gate electrode of each memory element being connected to one of a plurality of word lines, the first diffusion region and the second diffusion region of each memory element each being connected to one of a plurality of bit lines; a row decoder for selecting one of the plurality of word lines in accordance with a row address; a column decoder for selecting at least one of the plurality of bit lines in accordance with a column address; and a sense amplifier for determining a value of multibit data stored in one of the first memory function section and the second memory function section with respect to the selected bit line of the plurality of bit lines. The sense amplifier comprises: a comparison section for comparing a voltage of the selected bit line with each of a plurality of different reference voltages, and outputting a plurality of comparison results; and a determination section for determining the value of the multibit data based on the plurality of comparison results.
In one embodiment of this invention, at least a portion of the first memory function section or the second memory function section of the memory element overlaps a portion of the first diffusion region or the second diffusion region.
In one embodiment of this invention, at least one of the first diffusion region and the second diffusion region of the memory element is offset from the gate electrode.
According to another aspect of the present invention, a mobile electronic apparatus is provided, which comprises any one of the above-described semiconductor storage devices.
According to the semiconductor storage device of the present invention, multibit data (2 bits or more) can be stored in at least one of the first memory function section and the second memory function section of a memory element. As a result, a memory capacity per memory element can be dramatically increased. That is, a large-capacity memory array can be realized without increasing the area of a memory array and the density of memory elements.
The mobile electronic apparatus of the present invention comprises the semiconductor storage device of the present invention, and therefore, has at least the same effect as that of the semiconductor storage device of the present invention.
Thus, the invention described herein makes possible the advantages of providing a semiconductor storage device, in which multiple bits (2 bits or more) can be stored in at least one of two memory function sections contained in a memory element which can be easily miniaturized, and a mobile electronic apparatus comprising the semiconductor storage device.
These and other advantages of the present invention will become apparent to those skilled in the art upon reading and understanding the following detailed description with reference to the accompanying figures.
Hereinafter, the present invention will be described by way of illustrative examples with reference to the accompanying drawings.
The semiconductor storage device of the present invention comprises at least one memory element.
The memory element mainly comprises a semiconductor layer, a gate insulating film, a gate electrode, a channel region, a diffusion region, and a memory function section. As used herein, the term “channel region” refers to a region which typically has the same conductivity type as that of the semiconductor layer and is located directly under the gate electrode. As used herein, the term “diffusion region” refers to a region which has a conductivity type opposite to that of the channel region.
Specifically, the memory element of the present invention may comprise a first conductivity type region as a diffusion region, a second conductivity type region as a channel region, a memory function section extending across a border between the first and second conductivity type regions, and an electrode provided via a gate insulating film. Preferably, the memory element of the present invention may comprise a gate electrode provided on a gate insulating film, two memory function sections provided on opposite aides of the gate electrode, two diffusion regions each provided on a side of the memory function section opposite to the gate electrode, and a channel region provided under the gate electrode.
The semiconductor apparatus of the present invention is provided as semiconductor layers on a semiconductor substrate, and preferably, on a first conductivity type well region provided within the semiconductor substrate.
The semiconductor substrate is not particularly limited as long as it can be used in semiconductor apparatuses. Examples of the semiconductor substrate include bulk substrates made of, for example, an elemental semiconductor, such as silicon, germanium, or the like, or a compound semiconductor, such as silicon germanium, Gabs, InGaAs, ZnSe, GaN, or the like. Substrates having a semiconductor layer on a surface thereof may be used, including SOI (Silicon on Insulator) substrates, SOS substrates, multilayer SOX substrates, glass or plastic substrates having a semiconductor layer, and the like. Particularly, a silicon substrate or an SOI substrate having a silicon layer on a surface thereof are preferable. The semiconductor substrate or the semiconductor layer may be monocrystalline (e.g., using epitaxy), polycrystalline, or amorphous, depending on which the amount of current passing therewithin varies.
Preferably, an isolation region is provided on the semiconductor layer. In addition, an element, such as a transistor, a capacitor, a resistor, or the like, may be provided on the semiconductor layer. These elements maybe combined into a circuit or a semiconductor apparatus. The circuit or semiconductor apparatus may have a single layer structure, or a multilayer structure with interlayer insulating film(s). The isolation region may be made of an isolation film, such as a LOCOS film, a trench oxide film, an STI film, or the like. The semiconductor layer may have P-type or N-type conductivity. Preferably, at least one first conductivity type (P-type or N-type) well region is provided on the semiconductor layer. The impurity concentrations of the semiconductor layer and the well region may be within a range known in the art. Note that when at SOI substrate is used as the semiconductor layer, a well region may be provided in a surface semiconductor layer, while a body region may be provided under a channel region.
The gate insulating film is not particularly limited as long as it is commonly used for semiconductor apparatuses. For example, the gate insulating film may be a single-layer or multilayer film of an insulating film, such as a silicon oxide film, a silicon nitride film, or the like; and a highly dielectric film, such as an aluminum oxide film, a titanium oxide film, a tantalum oxide film, a hafnium oxide film, or the like. Particularly, a silicon oxide film is preferable. The gate insulating film preferably has a film thickness of about 1 to 20 nm, and more preferably about 1 to 6 nm. The gate insulating film may be provided only directly under a gate electrode, or may be larger (wider) than the gate electrode.
The gate electrode, which is provided on the gate insulating film, is in a shape which is typically used in semiconductor apparatuses or in a shape which has a recess portion at a lower end portion thereof. Note that the gate electrode is preferably in an integrated form without being divided by a single-layer or multilayer conductive film. However, the gate electrode may be divided by a single-layer or multilayer conductive film. The gate electrode may have a side wall insulating film on a side wall thereof. The gate electrode is not particularly limited as long as it is commonly used in semiconductor apparatuses. The gate electrode may be, for example, a single-layer or multilayer conductive film made of, for example, polysilicon, a metal (e.g. copper, aluminum, etc.), a high melting point metal (e.g., tungsten, titanium, tantalum, etc.), a silicide of a high melting point metal, or the like. The gate electrode preferably has a film thickness of, for example, about 50 to 400 nm. Note that the channel region is provided under the gate electrode.
Note that, preferably, the gate electrode is provided only on a side wall of the memory function section described below, or may not overlap an upper portion of the memory function section. With this arrangement, a contact plug can be positioned closer to the gate electrode, thereby facilitating the miniaturization of the memory element. In addition, the above-described simple arrangement facilitates the manufacture of the memory element, resulting in an improvement in yield.
The memory function section has at least a function of retaining charge (hereinafter referred to as “charge retaining function”). In other words, the memory function section has a function of storing or retaining charge, trapping charge, or retaining a charge polarization state. This function is achieved by a particular film or region having a charge retaining function, which to contained in the memory function section. Such a film or region comprises silicon nitride, silicon, silicate glues containing an impurity (e.g., phosphor, boron, etc.), silicon carbide, alumina, a high dielectric (e.g., hafnium oxide, zirconium oxide, tantalum oxide, etc.), zinc oxide, a ferroelectric, a metal, or the like. Therefore, the memory function section may comprise a single-layer or multilayer structure of, for example, an insulating film comprising a silicon nitride film; an insulating film comprising a conductive film or a semiconductor layer therewithin an insulating film comprising at least one conductor or semiconductor dot; an insulating film comprising a ferroelectric film in which inner charge is polarized due to an electric field and the polarization is maintained; or the like. Particularly, silicon nitride film is preferable. The is because silicon nitride film has the following benefits: 1) silicon nitride film has a high trap level of charges, and therefore, can have large hysteresis characteristics; 2) silicon nitride film also has a long charge retaining time, so that there is substantially no charge leakage due to generation of a leak path, and therefore, the charge retaining function is satisfactory; and 3) silicon nitride film is a standard material for the manufacture of LSIs.
When an insulating film comprising therewithin a film having a charge retaining function, such as a silicon nitride film or the like, is used as a memory function section, the reliability of retaining memories can be enhanced. Since silicon nitride film is an insulating material, even if charge leakage occurs in a portion of the film, all charges cannot be immediately eliminated throughout the silicon nitride film. When a plurality of memory elements are arranged, a distance between memory elements may be small, and thereby, some adjacent memory function sections may contact each other. In this case, if the memory function sections are made of a conductor, information stored in the memory function sections can be lost. However, the insulating film memory function section can avoid such a problem. In addition, a contact plug can be positioned closer to the memory function section. In some cases, the contact plug can overlap the memory function section. Therefore, the miniaturization of the memory element can be facilitated.
In order to enhance the reliability of retaining memories, a film-like shape is not necessarily required for a film having a charge retaining function. It is preferable that a film having a charge retaining function is discretely provided in an insulating film. Specifically, a film having a charge retaining function is preferably distributed in the form of dots in a material which resists retaining a charge, such as silicon oxide or the like.
When a conductive film or a semiconductor layer is used as a charge retaining film, the charge retaining film is preferably provided via an insulating film so that the charge retaining film does not directly contact a semiconductor layer (a semiconductor substrate, a well region, a body region, a source/drain region, or a diffusion region) or a gate electrode. For example, conductive films and insulating films are alternately stacked (multilayer structure); a conductive film is distributed in the form of dots in an insulating film; a conductive film is provided in a portion of a side wall insulating film provided on a side wall of a gates and the like.
Preferably, by employing as a memory function section an insulating film comprising a conductive film or a semiconductor layer therewithin, the amount of charges to be injected into the conductor or semiconductor film can be controlled substantially freely, thereby obtaining a multivalued memory function section.
Preferably, by employing as a memory function section an insulating film comprising at least one conductor or semiconductor dot, write/erase operation due to direct tunneling of charge can be facilitated, resulting in lower power consumption.
A ferroelectric film, such as PZT, PLZT, or the like, in which a polarization direction varies depending on electric field, maybe used as a memory function section. In this case, charges are substantially generated on a surface of a ferroelectric film due to polarization, and such a state is maintained. Therefore, the ferroelectric film memory function section, can have hysteresis characteristics similar to that of a film having a memory function, such that charges are supplied from the outside of the film and are then trapped by the film. In addition, the ferroelectric film can retain charges without injection of charges from the outside of the film. In the ferroelectric film, hysteresis characteristics can be obtained only due to the polarization of charges within the film. Therefore, the ferroelectric film can preferably provide high-speed write/erase operation.
Note that the insulating film contained in the memory function section is preferably a film which has a region or function of obstructing escaping charges. An exemplary film having a function of obstructing escaping charges is a silicon oxide film or the like.
The charge retaining film contained in the memory function section is provided on opposite sides of the gate electrode directly or via an insulating film, or to provided on a semiconductor layer (a semiconductor substrate, a well region, a body region, or a source/drain region or a diffusion region) directly or via a gate insulating film. The charge retaining film on the opposite sides of the gate electrode preferably overlap the entire or a part of the side wall of the gate electrode directly or via an insulating film. In an exemplary application in which the gate electrode has a recess portion in a lower end portion thereof, the charge retaining film may be formed directly or via an insulating film to bury the recess portion completely or partially.
The diffusion region can function as a source/drain region, and has a conductivity type opposite to that of the semiconductor layer or the well region. The junction between the diffusion region and the semiconductor layer or the well region preferably has a steep gradient of impurity concentration. This is because hot electrons or hot holes are efficiently generated at a lower voltage, i.e., a high-speed operation can be achieved at a low voltage. The junction depth of the diffusion region is not particularly limited, and can be adjusted as appropriate, depending on the desired function of a semiconductor storage device or the like. Note that when an SOI substrate is used as a semiconductor substrate, the diffusion region may have a junction depth smaller than a film thickness of the surface semiconductor layer. Preferably, the junction depth is approximately equal to the film thickness of the surface semiconductor layer.
The diffusion region may be disposed to overlap an end of the gate electrode, to be contiguous to an end of the gate electrode, or to be offset from an end of the gate electrode. Particularly when the diffusion region is offset from an end of the gate electrode, the inversion of an offset region under the charge retaining film, which is caused by applying a voltage to the gate electrode, depends largely on the amount of charges accumulated in the memory function section. Therefore, preferably, the memory effect is increased and the short channel effect is reduced. However, when the off set is large, a driving current between the diffusion regions (source/drain) is significantly reduced. Therefore, the offset, i.e., a distance from one end of the gate electrode to the nearer diffusion region in the longitudinal direction of the gate electrode, is preferably shorter than a thickness of the charge retaining film in a direction parallel to the gate longitudinal direction. It is particularly important that at least a portion of a film or region in the memory function section, which has a charge retaining function, overlaps a portion of the diffusion region. The essence of the memory element contained in the semiconductor storage device of the present invention is that a memory is rewritten by an electric field across the memory function section generated by a voltage difference between the gate electrode present only on the side wall portion of the memory function section, and the diffusion region.
A portion of the diffusion region may be provided above a surface of the channel region, i.e., a lower surface of the gate insulating film. In this case, it is appropriate that a conductive film is formed on the diffusion region provided in the semiconductor substrate in a manner that integrates the conductive film with the diffusion region. Examples of such a conductive film include a semiconductor (e.g., polysilicon, amorphous silicon, etc.), silicide, the above-described metals, high melting point metals, and the like. Particularly, polysilicon is preferable. Since polysilicon has a very high impurity diffusion rate as compared with that of the semiconductor layer, it is easy to obtain a shallow junction depth of the diffusion region in the semiconductor layer, i.e., it is easy to suppress the short channel effect. Note that, in this case, at least a portion of the memory function section as well as the gate electrode are preferably located between the diffusion regions.
The memory element of the present invention can be formed in a manner similar to a technique, in which a typical semiconductor process is used to form a side-wall spacer having a single-layer or multilayer structure on, for example, a aide wall of the gate electrode. Specifically, after the gate electrode is formed, a single-layer film or multilayer film comprising a film having a charge retaining function (hereinafter referred to as “charge retaining film∞), a charge retaining film/an insulating film, an insulating film/a charge retaining film, an insulating film/a charge retaining film/an insulating film, or the like, are formed. The resultant film is subjected to etch back under appropriate conditions so as to leave a residual film in the form of a side-wall spacer. Alternatively, an insulating film or a charge retaining film is formed. Etch back is performed under appropriate conditions so that a residual film is left in the form of a side-wall spacer. Another charge retaining film or an insulating film is formed. Etch back is similarly performed so that a residual film is left in the form of a side-wall spacer. Alternatively, an insulating film material, in which particulate charge retaining materials are distributed, is applied or deposited on a semiconductor layer comprising a gate electrode. Etch back to performed under appropriate conditions so that a residual insulating film material is left in the form of a side-wall spacer. Alternatively, after a gate electrode is formed, the above-described single-layer film or multilayer film 18 formed. Patterning is performed using a mask. Alternatively, a charge retaining film, a charge retaining film/an insulating film, an insulating film/a charge retaining film, an insulating film/a charge retaining film/an insulating film, or the like to formed before forming a gate electrode. An opening is formed in a region which will be a channel region in these films. A gate electrode material film is formed on the entire surface of the resultant structure. The gate electrode material film is subjected to patterning in such a manner that a portion thereof, which buries the opening and has a size broader than that of the opening, is left. The present invention is not limited to these methods.
When a plurality of memory elements of the present invention are arranged into an array (memory cell array), the best mode of the memory element is obtained if all of the following requirements are satisfied: for example, (1) the gate electrodes of a plurality of memory elements are integrated together to function as a word line; (2) a memory function section is provided on opposite sides of the word line; (3) an insulator, particularly a silicon nitride film, retains charges in the memory function section; (4) the memory function section is made of ONE (Oxide Nitride Oxide) film, and the silicon nitride film has a surface substantially parallel to a surface of the gate insulating film; (5) the silicon nitride film in the memory function section is separated from the word line and the channel region by the silicon oxide film; (6) the insulating film (silicon nitride film) in the memory function section and the diffusion region overlap each other; (7) a thickness of the insulating film separating the silicon nitride film having a surface substantially parallel to a surface of the gate insulating film from the channel region or the semiconductor layer, is different from a thickness of the gate insulating films (8) writing and erasure of a single memory element are performed via a single word line; (9) no electrode (word line) having a function of aiding writing and erasure is possessed on the memory function section; and (10) a region having high impurity concentration and a conductivity type opposite to that of the diffusion region in provided directly under the memory function section and adjacent to the diffusion region. Note that a memory element which satisfies at least one of the requirements may be used heroin.
A particularly preferable combination of the above-described requirements include that (3) an insulator, particularly a silicon nitride film, retains charges in the memory function section; (6) the insulating film (silicon nitride film) in the memory function section and the diffusion region overlap each other; and (9) no electrode (word line) having a function of aiding write and erase operations is possessed on the memory function section.
A memory element, which satisfies the requirements (3) and (9), is considerably useful as described below. A bit line contact can be disposed closer to a memory function section on a sidewall of a word line. In addition, even when memory elements are close to each other, a plurality of memory function sections do not interfere with each other, thereby retaining memory information. Therefore, the miniaturization of memory elements is facilitated. Note that when the charge retaining region in the memory function section is made of a conductor, charge retaining regions interfere with each other due to capacity coupling as memory elements are disposed closer to each other, whereby memory information cannot be retained.
When the charge retaining region in the memory function section is made of an insulator (e.g., a silicon nitride film, etc.), it is not necessary that a separate memory function section is provided in each memory cell. For example, memory function sections, which are provided on opposite sides of a word line shared by a plurality of memory cells, need not be separated from each other in each memory cell. The memory function sections provided on the opposite sides of the word line can be shared by the plurality of memory cells sharing the word line. Therefore, a photolithography step and an etching step for separation of memory function sections are not required, thereby simplifying manufacturing processes. In addition, a positioning margin in a photolithography stop and a margin for film reduction by etching are not required, thereby making it possible to reduce a margin between memory cells. Therefore, as compared with when the charge retaining region in the memory function section is made of a conductor (e.g., polycrystalline silicon film), the footprint of a memory cell can be reduced even if the same level of micromachining is used. Note that when the charge retaining region in the memory function section to made of a conductor, a photolithography step and an etching step for separating memory function sections for each memory cell are required. In this case, a photolithography positioning margin and an etching film reduction margin are required.
In addition, no electrode having a function of aiding write and erase operations is provided on the memory function section, and therefore, the device structure is simple, whereby the number of steps can be reduced and the yield can be improved. Therefore, it is possible to integrate the memory element of the present invention with transistors constituting a logic circuit or an analog circuit. Moreover, inexpensive semiconductor storage devices can be obtained.
It is more preferable that the requirement (6) as well as the requirements (3) and (9) are satisfied. Specifically, when a charge retaining region in a memory function section and a diffusion region overlap each other, write and erase operations can be performed at a considerably lower voltage. Specifically, for example, write and erase operations can be performed at such low a voltage as 5 V or less. This is a considerably large effect in terms of circuit design. A high voltage, which is required for flash memories, is not required in a chip. Therefore, a charge pumping circuit, which requires a large footprint, can be omitted or its size can be reduced. Particularly, when a small-capacity memory is incorporated for adjustment into a logic LSI, the footprint of peripheral circuitry for driving the memory cell is dominant in the memory portion as compared with the footprint of the memory cell. Therefore, by omitting a voltage boostering circuit for a memory cell or reducing the size thereof, the size of a chip can be most effectively reduced.
When the requirement (3) is not satisfied, i.e., a conductor is used to retain charges in the memory function section, a write operation can be performed even if the requirement (6) is not satisfied, i.e., the conductor in the memory function section and a diffusion region do not overlap each other. This is because the conductor in the memory function section aids writing due to capacity coupling with the gate electrode.
When the requirement (9) is not satisfied, i.e., at electrode having a function of aiding write and erase operations is provided on the memory function section, writing can be performed even if the requirement (6) is not satisfied, i.e., an insulator in the memory function section and a diffusion region do not overlap each other.
In the semiconductor storage device of the present invention, one or both ends of the memory element may be connected to a transistor in series, or alternatively, the memory element and a logic transistor may be integrated together on the same chip. In this case, the semiconductor device of the present invention particularly the memory element, can be formed by a process having high affinity to a standard process for typical transistors (e.g., transistors, logic transistors, etc.). Therefore, the memory element and transistors can be simultaneously formed. Therefore, a process for integrating the memory element with a transistor or a logic transistor is simple, is thereby, making it possible to obtain an inexpensive integrated device.
In the semiconductor storage device of the present invention, one memory function section can store two or more-level information, whereby the memory element can store 4 or more-level information. Note that the memory element may store only two-level information. In addition, when the variable resistor effect of the memory function section is utilized, the memory element can function as a memory cell having functions of both a selection translator and a memory transistor.
The semiconductor storage device of the present invention may be combined with a logic element, a logic circuit, or the like to provide a wide variety of application of electronic apparatuses, such as data processing system. (e.g., personal computers, notebook computers, laptop computers, personal assistant/signal emitters, minicomputers, workstations, mainframes, multiprocessor computers or all other types of computer systems); electronic parts contained in data processing systems (e.g., CPUs, memories, data storage devices, etc.); communication apparatuses (e.g., telephones, PHS, modems, rooters, eta.); image display apparatuses (e.g., display panels, projectors, etc.), business machines (e.g., printers, scanners, copying machines, etc.); video apparatuses (e.g. a video cameras, digital cameras, etc.); entertainment apparatuses (e.g., game machines, music players, etc.); information apparatuses (e.g., personal digital assistants, clocks and watches, electronic dictionaries, etc.); apparatuses carried on cars (e.g., car navigation systems, car audio apparatuses, etc.); audio/video apparatuses for recording/reproducing information, such as moving images, static images, music, and the like; electrical appliances (e.g., washing machines, microwave ovens, refrigerators, rice cookers, dishwashers, vacuum cleaners, air conditioners, etc.); health care apparatuses (e.g., massaging machines, scales, sphygmomanometers, etc.); mobile storage devices (e.g., IC cards, memory cards, etc.), or the like. Particularly, the semi conductor storage device of the present invention is useful for mobile electronic apparatuses, such as mobile telephones, mobile information terminals, IC cards, memory cards, mobile computers, mobile game machines, digital cameras, portable video players, portable music players, electronic dictionaries, clocks and watches, and the like. Note that the semiconductor storage device of the present invention may be incorporated as at least a portion of a control circuit or a data storage circuit of an electronic apparatus, or may be optionally incorporated in a detachable manner.
A method for producing the above-described memory element will be described below with reference to
Thereafter, the silicon oxide film/silicon nitride film/silicon oxide film is etched back by anisotropic etching. As a result, memory function sections 261 and 262, which are optimal to storage, are formed in the form of a side-wall spacer on side walls of the gate electrode 217.
Thereafter, the gate electrode 217 and the side-wall spacer memory function sections 261 and 262 are used as a mask to perform ion implantation to form diffusion layer regions (source/drain region) 212 and 213. Thereafter, silicide steps or upper wiring steps may be performed using known procedures.
Hereinafter, embodiments of a semiconductor storage device, a mobile electronic apparatus, a method for controlling erasure and programming of data stored therein will be described with reference to accompanying drawings.
A semiconductor storage device according to Embodiment 1 of the present invention comprises a memory element 1 as shown in
The memory element 1 comprises a semiconductor substrate 101, a P-type well region 102, a gate insulating film 103, and a gate electrode 104. The P-type well region 102 is provided on the semiconductor substrate 101. The gate electrode 104 to provided via the gate insulating film 103 on the P-type well region 102. A silicon nitride film 109 having a trap level for retaining charges (charge retaining film) is provided on an upper surface and side surfaces of the gate electrode 104. Portions of the silicon nitride film 109 on aide wall portions of the gate electrode 104 are memory function portions 105a and 105b, respectively, which actually retain charges. The term “memory function portion” refers to a memory function section or a portion of the charge retaining film in which charges are actually accumulated due to a rewrite operation. N-type diffusion regions 107a and 107b, which function as a source region or a drain region, respectively, are provided on opposite sides of the gate electrode 104 and within the P-type well region 102. The diffusion regions 107a and 107b have an offset structure. Specifically, the diffusion regions 107a and 107b do not reach a region 121 under the gate electrode 104. An offset region 120 under the charge retaining film constitutes a portion of a channel region.
Note that the memory function portions 105a and 105b, which substantially retain charges, are side wall portions of the gate electrode 104. Therefore, the silicon nitride film 109 may be provided only on a region corresponding to the above-described portions ((a) of
The principle of writing information into the memory element will be described with reference to
To implant electrons into the second memory function section 131b (writing), as shown in
To implant electrons into the flat memory function section 131a (writing), as shown in
Next, the principle of erasure of the memory element will be described with reference to
In a first method for erasing information stored in the first memory function section 131a, as shown in
To erase information stored in the second memory function section 131b, the potential is exchanged between the first diffusion region and the second diffusion region.
in a second method for erasing information stored in the first memory function section 131a, as shown in
In the second method, even when a voltage applied to the PN junction between the P-type well region and the first diffusion region 107a is insufficient for creating a tunnel between bands so that hot holes are not generated in the PN junction, the electrons implanted from the second diffusion region 107b trigger generation of electron-hole pairs in the PN junction, so that hot holes can be generated. Therefore, a voltage required for erasure can be lowered. Particularly, when the offset region 120 (
Note that in the first erasure method, information stored in the first memory function section 131a is erased by applying a voltage of +5 V to the first diffusion region 107a. In the second erasure method, only +4 V is sufficient. Thus, the second method can reduce a voltage for erasure, i.e., can reduce power consumption, thereby making it possible to suppress deterioration of the memory element due to hot carriers.
In either the first or second method, the memory element resists over-erasure. The term “over-erasure” refers to a phenomenon that as the amount of holes accumulated in the memory function section is increased, the threshold is decreased without saturation. Over-erasure raises a significant problem in EEPROM, such as representatively flash memory and the like. Particularly, when the threshold is negative, it becomes impossible to select memory cells, which is a critical operation error. On the other hand, in the memory element of the semiconductor storage device of the present invention, even when a large amount of holes are accumulated in the memory function section, electrons are induced only under the memory function section, leaving the potential of the channel region under the gate insulating film substantially unaffected. Since the erasure threshold is determined by the potential under the gate insulating film, over-erasure it unlikely to occur.
A principle of a read operation of the memory element will be described with reference to
To read information stored in the first memory function section 131a, the first diffusion region 107a is caused to function as a source electrode of the transistor, while the second diffusion region 107b is caused to function as a drain electrode of the transistor. For example, a voltage of 0 V is applied to the first diffusion region 107a and the P-type well region 102, a Voltage of +1.8 V is applied to the second diffusion region 107b, and a voltage of +2 V is applied to the gate electrode 104. In this case, when electrons are not accumulated in the first memory function section 131a, a drain current is likely to flow. On the other hand, when electrons are accumulated in the first memory function section 131a, an inversion layer in unlikely to be formed in the vicinity of the first memory function section 131a, so that a drain current is unlikely to flow. Therefore, by detecting the drain current, it is possible to read information stored in the first memory function section 131a. Particularly when a voltage which causes a pinch-off operation is applied for a read operation, the state of charge accumulation in the first memory function section 131a can be precisely determined irrespective of the presence or absence of charge accumulation in the second memory function section 131b.
To read information stored in the second memory function section 131b, the second diffusion region 107b is caused to function as a source electrode of the transistor, while the first diffusion region 107a is caused to function as a drain electrode of the transistor. For example, a voltage of 0 V is applied to the second diffusion region 107b and the P-type well region 102, a voltage of +1.8 V is applied to the first diffusion region 107a, and a voltage of +2 V is applied to the gate electrode 104. Thus, as compared to when information stored in the first memory function section 131a is read out, by exchanging source/drain regions, information stored in the second memory function section 131bcan be read out.
Note that when a portion of the channel region remains not covered with the gate electrode 104 (such a portion is the offset region 120), an inversion layer is extinguished or formed in the portion of the channel region not covered with the gate electrode 104, depending on the presence or absence of excessive charges in the memory function sections 131a and 131b. As a result, a large hysteresis (threshold change) is obtained. Note that when the width of the offset region 120 is considerably large, a drain current is significantly reduced, leading to a significant reduction in the read rate. Therefore, the width of the off set region 120 is preferably determined so as to achieve a sufficient hysteresis and read rate.
Even when the diffusion regions 107a and 107b each reach the end of the gate electrode 104, i.e., the diffusion regions 107a and 107b each overlap the gate electrode 104, the threshold of the transistor is not substantially changed by a write operation. However, a parasitic resistance at a source/drain end is significantly changed, so that a drain current is significantly reduced (by one order of magnitude or more). Therefore, it to possible to perform a read operation by detecting the drain current, thereby obtaining a function as a memory. Note that when a greater memory hysteresis effect is required, it is more preferable that the diffusion regions 107a and 107b do not overlap the gate electrode 104 (i.e., the offset region 120 exists).
With the above-described operation method, it is possible to write and erase two bits of data per transistor. In addition, memory elements may be arranged into a memory array, where a word line WL is connected to the gate electrode 104, a first bit line BL1 is connected to the first diffusion region 107a, and a second bit line BL2 is connected to the second diffusion region 107b.
In the above-described operation method, by exchanging between the source electrode and the drain electrode, writing and erasure of two bits of data per transistor is achieved. Alternatively, the source electrode and the drain electrode may be fixed so that the transistor functions as a one-bit memory. In this case, one of the source and drain regions can be caused to have a common fixed voltage, so that the number of bit lines connected to the source/drain region can be reduced by half.
As is clearly understood from the above description in the memory element of the semiconductor storage device of the present invention, the memory function section is provided separately from the gate insulating film and is positioned on the opposite sides of the gate electrode, so that a two-bit operation can be achieved. The memory function sections are separated from each other by the gate electrode. Therefore, the interference between the memory function sections can be effectively suppressed when rewriting is performed. In addition, since the gate insulating film to separated from the memory function section, the gate insulating film can be a thin film, thereby making it possible to suppress a short channel effect. Therefore, it is easy to miniaturize a memory element, and further, a semiconductor storage device.
A memory element of a semiconductor storage device according to Embodiment 2 of the present invention has substantially the same structure as that of the memory element 1 of
The memory function sections 261 and 262 preferably each comprise a charge retaining film having a function of retaining charges and an insulating film for the purpose of improving the ability to retain memories. In Embodiment 2, a silicon nitride film 242 having a trap level of charges is used as a charge retaining film, while silicon oxide films 241 and 243 having a function of preventing dissipation of charges accumulated in the charge retaining film are used as insulating films. Since the memory function section comprises a charge retaining film and an insulating film, the dissipation of charges can be prevented, thereby improving the retaining ability. In addition, as compared with when a memory function section consists only of a charge retaining film, the volume of a charge retaining film can be reduced as appropriate. Thereby, the movement of charges within the charge retaining film is limited, so that the occurrence of changes in characteristics due to the movement of charges can be suppressed during the retaining of a memory. In addition, a silicon nitride film 242 is interposed between silicon oxide films 241 and 243, whereby the efficiency of charge implantation is increased during a rewrite operation so that a higher-speed operation can be obtained. Note that in this memory element, the silicon nitride film 242 may be replaced with one that is made of a ferroelectric material.
Regions (the silicon nitride film 242) of the memory function sections 261 and 262, which retain charges, overlap diffusion regions 212 and 213, respectively. In this case, the term “overlap” means that at least a portion of the region (silicon nitride film 242) for retaining charges is provided on at least a portion of the diffusion region 212 and 213. Note that reference numeral 211 indicates a semiconductor substrate, reference numeral 214 indicates a gate insulating film, reference numeral 217 indicates a gate electrode, a reference numeral 271 indicates an offset region between the gate electrode 217 and diffusion regions 212 and 213. Though not shown, a channel region is provided under the gate insulating film 214, i.e., on an uppermost surface of the semiconductor substrate 211.
An effect of overlapping of the silicon nitride film 242, which is a region for retaining charges in the memory function sections 261 and 262, and the diffusion regions 212 and 213, will be described below.
As shown in
In
Note that as shown in
As can be clearly seen from
Taking into account the results of the above-described device simulation, a memory cell array was produced where W2 was fixed to 100 nm and W1 was designed to be 60 nm or 100 nm. When W1 was 60 nm, the silicon nitride film 242 and each of the diffusion regions 212 and 213 were designed to overlap each other by 40 nm. When W1 was 100 nm, the silicon nitride film 242 and each of the diffusion regions 212 and 213 were designed not to overlap each other. Read times of these memory cell arrays were measured and the worst cases were compared with each other, taking variations thereof into account. As a result, a read access time was 100 times higher when W1 was designed to be 60 nm than when W1 was designed to be 100 nm. In practical applications, a preferable read access time is 100 nanoseconds or loss per bit. In the case of W1=W2, this condition cannot be achieved. When variations due to a manufacturing process is taken into account, (W2-W1)>10 nm to preferably established.
To read information from the memory function section 261 (region 281), preferably, the diffusion region 212 is caused to function as a source electrode, the diffusion region 213 is caused to function as a drain region, and a pinch-off point to provided in a portion of the channel region closer to the drain region, as in Embodiment 1. In other words, when information is read from one of the two memory function sections, a pinch-off point is preferably provided in a region of the channel region closer to the other memory function section. Thereby, it is possible to detect information stored in the memory function section 261 with high sensitivity irrespective of the storing state of the memory function section 262, significantly contributing to achievement of two-bit operations.
On the other hand, when information is stored only in one of the two memory function sections or two memory function sections are used in the same memory state, a pinch-off point is not necessarily provided during a read operation.
Note that, as not shown in
In addition, the memory function section preferably comprises a charge retaining film provided substantially parallel to a surface of the gate insulating film. In other words, preferably, an upper surface of the charge retaining film contained in the memory function section is equally spaced from an upper surface of the gate insulating film. Specifically, as shown in
Thus, the memory function section 262 comprises the silicon nitride film 242a substantially parallel to the surface of the gate insulating film 214. Due to this structure, the easiness of formation of an inversion layer in the offset region 271 can be effectively controlled by the more or less of the amount of charges accumulated in the silicon nitride film 242a, thereby increasing the memory effect. In addition, when the silicon nitride film 242a is provided parallel to the surface of the gate insulating film 214, a change in the memory effect can be reduced to a relatively small value even if the offset amount (W1) varies, thereby suppressing a variation in the memory effect. Moreover, the movement of charges toward an upper portion of the silicon nitride film 242a is suppressed, thereby making it possible to suppress a change in characteristics due to the movement of charges during the retaining of a memory.
In addition, preferably, the memory function section 262 comprises an insulating film (e.g. , a portion on the offset region 271 of a silicon oxide film 224), which separates the silicon nitride film 242a substantially parallel to the surface of the gate insulating film 214 from the channel region (or a well region). This insulating film suppresses the dissipation of charges accumulated in the charge retaining film, resulting in a memory element having a higher level of ability to retain a memory.
Note that by controlling the film thickness of the silicon nitride film 242a and controlling the film thickness of an insulating film (a portion on the off set region 271 of the silicon oxide film 244) under the silicon nitride film 242a to be constant, a distance from the surface of the semiconductor substrate to charges accumulated in the charge retaining film can be maintained substantially constant. In other words, the distance from the surface of the semiconductor substrate to the charges accumulated in the charge retaining film, can be controlled to be between the minimum film thickness of the insulating film, under the silicon nitride film 242a, and the sum of the maximum film thickness of the insulating film, under the silicon nitride film 242a, and the maximum film thickness of the silicon nitride film 242a. Thereby, the density of electric lines of force generated by charges accumulated in the silicon nitride film 242a can be substantially controlled, thereby considerably reducing a variation in the memory effect of the memory element.
A memory function section 262 in a semiconductor storage device according to Embodiment 3 of the present invention comprises a silicon nitride film 242 (charge retaining film) which has a substantially uniform film thickness as shown in
When a positive voltage is applied to the gate electrode 217, an electric line of force 283 in the memory function section 262 passes through the silicon nitride film 242 twice (the regions 282 and 281) as shown with an arrow. Note that when a negative voltage is applied to the gate electrode 217, the direction of an electric line of force is reverse. In this case, the relative dielectric constant of the silicon nitride film 242 to about 6, while the relative dielectric constant of the silicon oxide films 241 and 243 is about 4. Therefore, as compared with when only the region 281 (charge retaining film) is provided, the effective relative dielectric constant of the memory function section 262 in the direction of the electric line of force 283 is increased, thereby making it possible to reduce a difference in potential between opposite ends of the electric line of force. In other words, most of a voltage applied to the gate electrode 217 is used to enhance an electronic field in the offset region 271.
The reason charges are implanted into the silicon nitride film 242 during a rewrite operation is that the generated charges are attracted via the electronic field in the offset region 271. Therefore, by providing the charge retaining film indicated by arrow 282, the amount of charges implanted into the memory function section 262 during a rewrite operation is increased, resulting in an increase in a rewriting rate.
Note that when a silicon nitride film is provided instead of the silicon oxide film 243, i.e., the charge retaining film is not uniform with respect to a height corresponding to the surface of the gate insulating film 214, the movement of charges toward an upper portion of the silicon nitride film becomes significant, resulting in a deterioration in the ability to retain a memory.
The charge retaining film may be preferably made of a high dielectric material, such as hafnium oxide or the like, which has a considerably large relative dielectric constant, instead of the silicon nitride film.
In addition, preferably, the memory function section further comprise an insulating film (a portion on the offset region 271 of the silicon oxide film 241) which separates the charge retaining film substantially parallel to the surface of the gate insulating film from the channel region (or the well region). This insulating film suppresses the dissipation of charges accumulated in the charge retaining film, thereby further improving the ability to retain a memory.
In addition, preferably, the memory function section further comprise an insulating film (a portion adjacent to the gate electrode 217 of the silicon oxide film 241) which separates the gate electrode from the charge retaining film extending in a direction substantially parallel to a side of the gate electrode. This insulating film prevents charges from being implanted from the gate electrode to the charge retaining film, so that electric characteristics are not changed, thereby making it possible to improve the reliability of the memory element.
Moreover, as in Embodiment 2, the film thickness of the insulating film (a portion on the offset region 271 of the silicon oxide film 241) under the silicon nitride film 242 is preferably controlled to be constant. Also, as in Embodiment 2, the film thickness of the insulating film (a portion adjacent to the gate electrode 217 of the silicon oxide film 241) on a side of the gate electrode is preferably controlled to be constant. Thereby, the density of electric lines of force generated by charges accumulated in the silicon nitride film 242 can be substantially controlled, and the leakage of charges can be prevented.
In Embodiment 4, optimization of distances between a gate electrode, a memory function section, and source/drain regions in a memory element of a semiconductor storage device will be described below.
As shown in
In the above-described memory element, B<C is preferable. When such a relationship is satisfied, an offset region 271 exists between a portion under a gate electrode 217 of a channel region and diffusion regions 212 and 213. Thereby, the easiness of inversion effectively varied throughout the offset region 271 due to charges accumulated in the memory function sections 261 and 262 (silicon nitride film 242). Therefore, the memory effect is increased, particularly resulting in a high-rate read operation.
When the diffusion regions 212 and 213 are offset from the gate electrode 217, i.e., A<B is satisfied, the easiness of inversion of the offset region when a voltage is applied to the gate electrode 217 is significantly changed, depending on the amount of charges accumulated in the memory function section. The memory effect can be enhanced while the short channel effect can be reduced.
Note that the offset region 271 may not be necessarily required as long as the memory effect is obtained. Even when the offset region 271 is not provided, the memory effect may be obtained in the memory function sections 261 and 262 (the silicon nitride film 242) if the impurity concentration of the diffusion regions 212 and 213 are sufficiently low.
Accordingly, A<B<C is most preferable.
A memory element of a semiconductor storage device according to Embodiment 5 of the present invention has substantially the come structure as that of the memory element of Embodiment 2, except that the semiconductor substrate is an SOI substrate as shown in
In this memory element, a buried oxide film 28818 provided on a semiconductor substrate 286. An SOI layer is provided on the buried oxide film 288. Diffusion regions 212 and 213 are provided in the SOI layer, and the other region of the SOI layer is a body region 287.
This memory element has an action and effect similar to those of the memory element of Embodiment 2. In addition, the junction capacity between the diffusion regions 212 and 213 and the body region 287 can be significantly reduced, thereby making it possible to increase the speed of the element and reduce the power consumption thereof.
A memory element of a semiconductor storage device according to Embodiment 6 of the present invention has substantially the same structure as that of the memory element of Embodiment 2, except that a P-type high concentration region 291 is added to a channel side of the N-type diffusion regions 212 and 213 as shown in
Specifically, the P-type high concentration region 291 has a higher impurity (e.g., boron, etc.) concentration, which provides a P-type, than the impurity concentration of a region 292, which provides a P-type. An appropriate P-type impurity concentration of the P-type high concentration region 291 is, for example, about 5×1017 cm−3 to about 1×1019 cm−3. The P-type impurity concentration of the region 292 is, for example, 5×1016 to 1×1018 a cm−3.
Thus, by providing the P-type high concentration region 291, the junction between the diffusion regions 212 and 213 and the semiconductor substrate 211 is steep 20 directly under the memory function sections 261 and 262. Therefore, hot carriers are easily generated during write and erase operations, so that a voltage required for write and erase operations can be reduced, or the speed of write and erase operations can be increased. Moreover, since the impurity concentration of the region 292 is relatively low, the threshold to low when the memory element is in the erasure state, so that the drain current is increased. Therefore, the road rate is improved. As a result, it is possible to obtain a memory element having a low rewriting voltage or a high rewriting rate, and a high read rate.
In
A memory element of a semiconductor storage device according to Embodiment 7 of the present invention has substantially the same structure as that of the memory element of Embodiment 2, except that the thickness (T1) of an insulating film separating a charge retaining film (a silicon nitride film 242) from a channel region or a well region, is smaller than the thickness (T2) of a gate insulating film as shown in
The thickness (T2) of a gate insulating film 214 has a lower limit because of the requirement for a breakdown voltage during a memory rewrite operation. However, the thickness (T1) of the insulating film can be smaller than T2 irrespective of the requirement for a breakdown voltage.
In this memory element, the reason the degree of freedom of designing T1 is high as described above, is as follows.
In this memory element, the insulating film separating the charge retaining film from the channel region or the well region, is not interposed between the gate electrode and the channel region or the well region. Therefore, a high electronic field applying between the gate electrode and the channel region or the well region does not act directly on the insulating film separating the charge retaining film from the channel region or the well region. Instead, there is a relatively weak electronic field spreading horizontally from the gate electrode. Therefore, T1 can be smaller than T2 irrespective of the requirement for a breakdown voltage to the gate insulating film.
By decreasing T1, it can be made easier to implant charges into a memory function section, so that a voltage required for writing and erase operations can be lowered, or the speed of writing and erase operations can be increased. In addition, when charges are accumulated in a silicon nitride film 242, the amount of charges induced by the channel region or the well region is increased, thereby making it possible to enhance the memory effect.
Among the electric lines of force in the memory function section, there is a short electric line of force, which does not pace through the silicon nitride film 242, as shown with arrow 284 in
In contrast to this, for example, EEPROM, such as, representatively, a flash memory or the like, comprises an insulating film separating a floating gate from a channel region or a well region, which is interposed between a gate electrode (control gate) and the channel region or the well region. A high electronic field directly acts from the gate electrode onto the insulating film. Therefore, in EEPROM, the thickness of the insulating film separating the floating gate from the channel region or the well region is limited, so that the optimization of the function of a memory element is inhibited.
As is clearly understood from the above description, by setting T1<T2, the voltage required for write and erase operations can be reduced, or the speed of writing and erase operations can be increased, and the memory effect can be enhanced, without reducing the ability of a memory element to withstand with voltage. Note that the thickness T1 of the insulating film is preferably equal to or greater than 0.8 nm, in which range the thickness T1 can maintain a prescribed level of uniformity or film quality during manufacturing processes, and the ability to retain a memory is not considerably deteriorated.
Specifically, in the case of a liquid crystal driver LSI which has a high withstand voltage in a design rule, a maximum voltage of 15 to 18 V is required to drive a liquid crystal panel TFT. Therefore, generally, a thin gate oxide film cannot be obtained. When a non-volatile memory of the present invention for adjusting Images is incorporated into a liquid crystal driver LSI, the thickness of the insulating film separating the charge retaining film (silicon nitride film 242) from the channel region or the well region can be designed to be optimized, independently of the thickness of the gate insulating film, in the memory element of the present invention. For example, in a memory cell having a gate electrode length (word line width) of 250 nm, T1 and T2 are separately determined to be 20 nm and 10 nm, respectively. Thereby, a memory cell having a satisfactory level of writing efficiency can be obtained. (The reason a short channel effect does not occur even when T1 is greater than the thickness of typical logic transistors, is that a source/drain region is offset from the gate electrode).
A memory element of a semiconductor storage device according to Embodiment 8 of the present invention has substantially the same structure as that of Embodiment 2, except that a thickness (T1) of an insulating film separating a charge retaining film (silicon nitride film 242) from a channel region or a well region is greater than a thickness (T2) of a gate insulating film as shown in
The thickness (T2) of the gate insulating film 214 has an upper limit value due to the requirement of prevention of a short channel effect of the memory element. However, the thickness (T1) of the insulating film can be made greater than T2 irrespective of the requirement of prevention of a short channel effect. Thus, the thickness of the insulating film separating the charge retaining film (silicon nitride film 242) from the channel region or the well region can be optimized, independently of the thickness of the gate insulating film. Therefore, when the degree of scaling of miniaturization is increased (the gate insulating film becomes thinner), the memory function section does not obstruct the scaling.
In this memory element, the reason that the degree of freedom of design of T1 is high, is that the insulating film separating the charge retaining film from the channel region or the well region to not interposed between the gate electrode and the channel region or the well region, as described above. Therefore, it to possible to make T1 greater than T2 irrespective of the requirement of prevention of a short channel effect for the gate insulating film.
By Increasing T1, the dissipation of charges accumulated in the memory function section can be prevented, thereby improving the ability to retain a memory.
Therefore, by setting T1>T2, the ability to retain a memory can be improved without deteriorating the short channel effect of the memory element.
Note that the thickness (T1) of the insulating film is preferably 20 nm or less, taking a reduction in the rewriting rate into consideration.
Specifically, in a non-volatile memory, such as, representatively, a flash memory or the like, a selection gate electrode is a writing/erasure gate electrode and a gate insulating film (including a floating gate) corresponding to the writing/erasure gate electrode also functions as a charge accumulation film. Therefore, the requirement for miniaturization (a thin film is essential for suppression of a short channel effect) is contrary to the requirement for reliability (to suppress leakage of retained charges, the insulating film separating the floating gate from the channel region or the well region cannot be equal to or less than about 7 nm). Therefore, miniaturization is difficult. In fact, ITRS (international Technology Roadmap for Semiconductors) shows that there is no prospect that the physical length of a gate is miniaturized to about 0.2 μm or less. In this memory element, T1 and T2 can be separately designed as described above, thereby making it possible to realize miniaturization.
For example, for a memory cell having a gate electrode length (word line width) of 45 nm, T2 and T1 are separately set to be 4 nm and 7 nm, respectively, thereby making it possible to realize a memory element free from a short channel effect. The reason a short channel effect does not occur even when T2 is greater than the thickness of typical logic transistor, is that a source drain region is offset from the gate electrode.
Since a source/drain region is offset from the gate electrode in the memory element, miniaturization is easier as compared with typical logic transistors.
Thus, there is no electrode for aiding write and erase operations on an upper portion of the memory function section. Therefore, a high electronic field applied between an electrode for aiding write and erase operations and the channel region or the well region does not act directly on the insulating film separating the charge retaining film from the channel region or the well region only a relatively weak electronic field, which spreads horizontally from the gate electrode acts on the insulating film. Therefore, it is possible to realize a memory element having a gate length smaller than that of logic transistors for the same process generation.
Embodiment 9 relates to a change in electric characteristics of a memory element of a semiconductor storage device of the present invention when rewriting to performed.
An can be seen from
The emergence of such characteristics is a peculiar phenomenon which occurs because the gate electrode and the diffusion region are offset from each other and a gate electronic field is unlikely to reach the offset region. When the memory element is in a writing state, an inversion layer is considerably unlikely to occur in the offset region under the memory function section even if a positive voltage is applied to the gate electrode. This is responsible for the small gradient of the Id-Vg curve at the subthreshold region in the writing state.
When the memory element is in an erasure state, electrons are induced in the offset region with high density. Further, when a voltage of 0 V is applied to the gate electrode (i.e., OFF state), electrons are not induced in a channel under the gate electrode (therefore, an off current is small). For this reason, the gradient of the Id-Vg curve in the subthreshold region is large in the erasure state, and the increasing rate of a current (conductance) to high even in a region equal to or higher than the threshold.
As can be seen from the above description, the memory element contained in the semiconductor storage apparatus of the present invention can have a particularly large drain current ratio between write and erase operations.
Embodiment 10 relates to a semiconductor storage device comprising a plurality of memory elements of Embodiments 1 to 8 and circuits for rewrite and read operations, and a method for operating the device.
The semiconductor storage device has a portion 321, in which memory elements 301aA to 301aD, . . . , and 301nA to 301nD of Embodiments 1 to 8. Each memory element has two memory function sections. For example, the two memory function sections in the memory element 301aA are indicated with arrows A and B.
A selection transistor is connected to each of the opposite sides of each memory element, i.e., two selection transistors are connected to each memory element. For example, selection transistors 302aA and 303aA are connected to the memory element 301aA in series.
The gate electrodes of the memory elements 301aA to 301aD are connected via a word line 308a. The gate electrodes of other memory elements are similarly connected via wordlines. The selection transistors 302aA to 302aD are connected via a selection transistor word line 309a. The selection transistors 303aA to 303aD are connected via a selection translator word line 310a. The gate electrodes of the other selection transistors are similarly connected via selection transistor word lines.
The selection transistor 302aA to 302nA are each connected to a first bit line 316A1. The selection transistor 303aA to 303nA are each connected to a second bit line 316A2. The other selection transistors are similarly connected to a first or second bit line.
Each bit line is connected to a pair of operation selecting transistors 304 and 305. The operation selecting translators 304 and 305 are connected to first voltage input terminals 317A1 to 317D1 and second voltage input terminals 317A2 to 317D2, respectively. The gate electrodes of the operation selecting transistors 304 and 305 are connected to operation selection line 312 and 313, respectively. With such wiring, when the operation selection line 312 is selected, for example, the first bit line 316A1 is connected to the first voltage input terminal 317A1 and the second bit line 316A2 is connected to the second voltage input terminal 317A2. When the operation selection line 313 is selected, for example, the first bit line 316A1 is connected to the second voltage input terminal 317A2 and the second bit line 316A2 is connected to the first voltage input terminal 317A1. Thus, a pair of the bit lines 316A1 and 316A2 are connected to different voltage input terminals. By selecting a different operation selection line, a voltage input terminal to be connected can be exchanged.
Two pairs of bit lines (e.g., a pair of bit lines 316A1 and 316A2 and a pair of bit lines 316B1 and 316B2) are connected via a switching transistor to one and the other input of a sense amplifier having two inputs, respectively. The connection will be described below more specifically.
The first bit lines 316A1 to 316D1 are each connected to a first switching transistor 306. The second bit line 316A2 to 316D2 are each connected to a second switching transistor 307. The gate electrodes of the switching transistors 306 and 307 are connected to switching transistor selection lines 314 and 315, respectively. With such wiring, when the switching transistor selection line 314 is selected, for example, one input of a sense amplifier 318AB is connected to the first bit line 316A1, while the other input of the sense amplifier 318AB is connected to the first bit line 316B1. When the switching transistor selection line 315 is selected, for example, one input of the sense amplifier 318AB is connected to the second bit line 316A2, while the other input of the sense amplifier 318AB is connected to the first bit line 316B2.
Note that any amplifier capable of detecting a current output from a memory element can be used as a sense amplifier. For example, a differential amplifier can be used.
Although four pairs of bit lines are provided in
In another example, in which a memory element is applied to such a semiconductor storage device, one memory element may be connected to one input of a sense amplifier, while an external reference cell may be connected to the other input of the sense amplifier.
A method for operating the above-described semiconductor storage device will be described below. The semiconductor storage apparatus has a rewrite operation and a read operation. The rewrite operation includes a write operation and a read operation.
Firstly, a method for a write operation will be described. As an exemplary write operation, writing information into the memory element 301aA will be described.
The operation selection line 312 is selected to switch the operation selecting transistor 304 into the ON state. Thereby, for example, the first bit line 316A1 is connected to the first voltage input terminal 317A1, while the second bit line 316A2 is connected to the second voltage input terminal 317A2. The same is true of the other bit lines.
In addition, the selection transistor word lines 309a and 310a are selected. As a result a one (memory function section A) of the diffusion regions (source/drain) of the memory element 301aA is connected to the first voltage input terminal 317A1, while the other (memory function section B) is connected to the second voltage input terminal 317A2. The same is true of the memory element 301aB to 301aD connected to the word line 308a.
Prescribed voltages for writing are applied to the word line 308a and the first and second voltage input terminals 317A1 and 317A2. A voltage of, for example, +5 V is applied to the word line 308a. A voltage of +5 V is applied to the first voltage input terminal 317A1, and a voltage of 0 V is applied to the second voltage input terminal 317A2. As a result, writing is selectively performed in the memory function section A of the memory element 301aA.
In this case, if prescribed voltages are applied to other voltage input terminals, writing can also be performed for the memory elements 301aB to 301aD. For memory elements in which writing is not to be performed, a voltage of 0 V is input to the voltage input terminal or the voltage input terminal is set to be in an open state.
To perform writing for the memory function section B of the memory element 301aA, the selection line 313 is selected instead of the operation selection line 312, where the other selection operations and applied voltage conditions are similarly determined. A voltage applied to the first voltage input terminal 317A1 may be exchanged with a voltage applied to the second voltage input terminal 317A2.
Next, a method for an erase operation will be described. As an exemplary erase operation, erasure of the memory element 301aA will be described below.
As in the write operation, the operation selection line 312 to selected, the operation selecting transistor 304 is switched into the ON state, and the selection translator word lines 309a and 310a are selected.
Prescribed voltages for erasure are applied to the word line 308a and the first and second voltage input terminals 317A1 and 317A2. A voltage of −5 V is applied to the word line 308a. A voltage of +5 V is applied to the first voltage input terminal 317A1. A voltage of 0 V is applied to the second voltage Input terminal 317A2. As a result, the memory function section A of the memory element 301aA is selectively erased.
In this case, if prescribed voltages are applied to the other voltage input terminals, the memory element 301aB to 301aD can be erased. For memory elements not to be erased, a voltage of 0 V is applied to the voltage input terminals thereof, or the voltage input terminals thereof are set to be in the open state.
To perform erasure for the memory function section B of the memory element 301aA, the selection line 313 is selected instead of the operation selection line 312, where the other selection operations and applied voltage conditions are similarly determined. A voltage applied to the first voltage input terminal 317A1 may be exchanged with a voltage applied to the second voltage input terminal 317A2.
Next, a method for a read operation will be described. As an exemplary read operation, reading of information stored in the memory element 301aA will be described.
To read stored information from the memory function section A of the memory element 301aA, the operation selection line 313 is selected, the operation selecting transistor 305 is switched into the ON state, and the switching transistor selection line 314 is selected to switch the first switching transistor 306 into the ON state. Further, a voltage appropriate for a read operation (e.g., +2 V) is applied to the word line 308a. A voltage of, for example, +1.8 V is applied to the first voltage input terminal 317A1. The second voltage input terminal 317A2 is in an open state.
According to the above-described selecting operation and applied voltage conditions, a current frows from the first voltage input terminal 317A1 and then via the memory element 301aA to one input of the sense amplifier 318AB. By detecting a value of the current, information stored in the memory element 301aA can be determined. In this case, the memory function section A of the memory element 301aA serves as a source. Therefore, the value of the current passing through the memory element 301aA is affected mainly by the more or less of the amount of charges accumulated in the memory function section A. Therefore, only information stored in the memory function section A can be selectively read out.
As described in Embodiment 9, in the semiconductor storage device of the present invention, a memory element can have a particularly large drain current ratio between write and erase operations, whereby a writing state and an erasure state can be easily distinguished from each other.
To read information stored in the memory function section B of the, memory element 301aA, the operation selection line 312 is selected instead of the operation selection line 313, the switching transistor selection line 315 is selected instead of the switching transistor selection line 314, and the other selecting operations and applied voltage conditions are similarly determined.
During the read operation, by additionally applying a voltage of, for example, +1.8 V for a read operation to the first voltage input terminal 317B1, a current corresponding to information stored in the memory element 318aB is supplied to the other input of the sense amplifier 318AB. Therefore, in Embodiment 10, the sense amplifier 318AB can detect a difference between currents passing through the two memory elements 301aA and 301aB. In this case, two memory elements are used to store one bit or two bits of information. As described above concerning a read operation, when the direction of a current passing through a memory element during a read operation is reversed, and information stored in the memory function section A and information stored in the memory function section B are separately read out, the two memory elements can be used to perform a two-bit operation. Alternatively, by limiting the direction of a current passing through a memory element during a read operation only to one direction, two memory elements can be used to perform a one-bit operation.
(a) to (d) of
When a single memory element is connected to one input of a sense amplifier while an external reference device is connected to the other input of the sense amplifier, the single memory element can store two bits of information.
However, the outputs of two memory elements (memory element pair) are preferably input to the same sense amplifier, as in Embodiment 10. With such a structure, by detecting a difference between currents passing through two memory elements having a similar device structure, information stored in the memory element can be read out. On the other hand, for example, when a transistor having a typical structure is used as an external reference, a difference in temperature characteristics between the memory element and the transistor having a typical structure reduces the reliability of a read operation. Therefore, by supplying the outputs of a memory element pair to the same sense amplifier, the reliability of a read operation can be improved.
In the semiconductor storage device of this embodiment, a selection transistor is connected to each of the opposite sides of each memory element, i.e., two selection transistors is connected to each memory element. Therefore, during a rewrite operation, a rewriting voltage applied on a bit line is applied to the only selected memory element, and the voltage is not applied to the other memory elements connected to the same bit line pair. Therefore, it is possible to prevent erroneous rewriting of non-selected cells during a rewrite operation.
In typical memory cells, off currents of non-selected cells are added to a read current of a selected cell during a read operation, so that the reliability of a read operation is reduced. Such a problem is reduced by the present invention, particularly when the number of word lines is large and the number of cells connected to the same bit line pair is large.
Note that even when a single selection transistor is provided only on one side of each memory element, it is possible to reduce off current of non-selected cells during a read operation.
The word lines 308ato 308n, for example, can be formed by using upper metal conductor lines, which are connected to the gate electrode of each memory element. Preferably, the gate electrodes of memory elements (at least a pair of memory elements) collectively function as a word line, and the memory function sections of a memory element pair are collectively shared on the opposite aides of the gate electrode. For example, a linear polysilicon electrode extends over active regions of a plurality of semiconductor layers. When the polysilicon electrode is separated from the active region of a semiconductor layer via a gate insulating film, the polysilicon functions as a gate electrode on each active region, and the linear polysilicon electrode itself functions as a word line. In this case, a contact which connects the gate electrode with the upper metal conductor line can be significantly reduced, thereby improving the scale of integration of a semiconductor storage device. In addition, since the gate electrodes share a memory function section, it is not necessary to separate memory function sections for each memory element. Therefore, the manufacturing process can be simplified and an inexpensive and reliable semiconductor storage device can be obtained.
In the semiconductor storage device of Embodiment 10, logic transistors are preferably integrated together.
A procedure for forming the memory element of Embodiment 10 has a high level of affinity to a typical process for forming a standard transistor. Therefore, a process for producing a semiconductor storage device, i.e., a process for integrating memory elements and logic translators together, is considerably simplified. Referring to
More specifically, for example, the side-wall spacer 714 may be modified to have a structure similar to the memory function sections 261 and 262 of
The LDD region may be formed in a transistor constituting a standard logic portion by performing impurity implantation for formation of the LDD region after formation of the gate electrode and before formation of the memory function section (side-wall spacer). Therefore, the memory element and the standard transistor can be simultaneously formed only by masking the memory element with a photoresist during performing impurity implantation for formation of the LDD region. Thus, the memory elements and the standard transistors can be easily integrated together.
Note that the process for forming a conventional flash memory is significantly different-from the process for forming a standard logic circuit. Therefore, as compared with when a flash memory to used as a non-volatile memory and the flash memory and logic circuits or analog circuits are integrated together, the semiconductor storage device of the present invention can dramatically reduce the number of masks and the number of processing steps. Therefore, the yield of a chip having logic circuits or analog circuits along with a non-volatile memory element can be improved, thereby reducing manufacturing cost. As a result, an inexpensive, highly reliable semiconductor storage device can be obtained.
In the semiconductor storage device of Embodiment 10, opposing stored information is preferably stored in two memory elements connected to a single sense amplifier. During a read operation, the sense amplifier is preferably operated to detect a difference between the values of currents passing through the two memory elements. When opposing stored information is stored in two memory elements connected to a single sense amplifier, the reliability of a read operation is higher than when a single memory element and an external reference cell are connected to a single sense amplifier. Moreover, a read rate can be improved while a read current value can be reduced. The reason the reliability of a read operation is improved when opposing stored information is stored in two memory elements connected to a single sense amplifier, is described below. The two memory elements can be caused to have the same number of rewrite operations. Therefore, any change in characteristics of the two memory cell due to the deterioration of the memory element is substantially the same between the two memory cells. As a result, a difference between the values of currents passing through the two memory elements is unlikely to be changed. Similarly, the difference between the values of currents passing through the two memory elements is easily maintained to be high, thereby making it possible to improve the read rate. Also, similarly, the value of a read current can be easily reduced by increasing the sensitivity of the sense amplifier. If the value of a read current can be reduced, the density of a memory cell array can be increased by reducing the gate width of each memory element.
In the above-described operation method, opposing stored information is stored in two memory elements connected to a single sense amplifier, and during a read operation, the sense amplifier is operated to detect a difference between the values of currents passing through the two memory elements. This operation method is particularly preferable for the semiconductor storage device of the present invention.
As described in Embodiment 9, the memory element of the present invention can have a particularly high drain current ratio between writing and erase operations. Therefore, it is possible to realize a high-speed read operation by increasing the difference between the values of currents passing through the two memory elements. Even when the gate width of the memory element is reduced, a necessary difference between the current values can be obtained. Therefore, particularly, the density of a memory cell array can be easily improved by reducing the gate width of each memory element.
Note that in two memory elements connected to a single sense amplifier, a rewrite operation may be performed separately in one memory function section (A) and the other memory function section (B) (the operation method described in
Alternatively, in two memory elements connected to a single sense amplifier, one memory function section (A) and the other memory function section (B) may have the same memory state (the operation method described in
During a read operation, a drain current sensitively responds to the charge amount of a memory function section closer to the source (source-side memory function section), and is not very sensitive to the charge amount of a memory function section closer to the drain (drain-side memory function section). However, a drain current of a memory element is not completely free from the influence of the charge amount of the drain-side memory function section. The influence results in an interference effect, varying the value of a current to a sense amplifier and increasing the margin of a read current. Therefore, in two memory elements connected to a single sense amplifier, by causing one memory function section (A) and the other memory function section (S) to have the same memory state, the variation of the value of a current input to the sense amplifier can be reduced. As a result, the reliability of a read operation can be improved.
In Embodiment 10, the memory element of Embodiment 7 is preferably used. Specifically, the thickness (T1) of an insulating film separating the charge retaining film (silicon nitride film 242) from the channel region or the well region is preferably smaller than the thickness (T2) of the gate insulating film and is preferably 0.8 nm or more. When such a memory element is used in the semiconductor storage device of Embodiment 10, a voltage required for writing and erase operations can be reduced, or high-speed writing and erase operations can be achieved. In addition, since the memory effect of the memory element is increased, the read rate of the semiconductor storage device can be increased.
In Embodiment 10, the memory element of Embodiment 8 vis preferably used. Specifically, the thickness (T1) of an insulating film separating the charge retaining film (silicon nitride film 242) from the channel region or the well region is preferably greater than the thickness (T2) of the gate insulating film, and is preferably 20 nm or less. If such a memory element is used in the semiconductor storage device of Embodiment 10, the ability of retaining a memory can be improved without deteriorating the short channel effect of the memory element. Therefore, even when the scale of integration of a semiconductor storage device is increased, a sufficient ability to retain a memory can be obtained.
As described in Embodiment 2, in a memory element used in Embodiment 10, regions (silicon nitride film 242) for retaining charges in the memory function sections 261 and 262 preferably overlap the diffusion regions 212 and 213, respectively. When such a memory element is used in the semiconductor storage device of Embodiment 10, the read rate of the semiconductor storage device can be sufficiently increased.
As described in Embodiment 2, in a memory element used in Embodiment 10, the memory function section preferably comprises a charge retaining film substantially parallel to a surface of the gate insulating film. When such a memory element is used in the semiconductor storage device of Embodiment 10, the variation of the memory effect of the memory element can be reduced. Therefore, in this case, the variation of a read current in the semiconductor storage device can be suppressed. In addition, a change in characteristics of the memory element retaining a memory can be reduced, thereby improving the memory retaining ability of the semiconductor storage device.
As described in Embodiment 3, in a memory element used in Embodiment 10, the memory function section preferably comprises a charge retaining film substantially parallel to a surface of the gate insulating film, and a portion extending substantially parallel to a side of the gate electrode. When such a memory element is used in the semiconductor storage device of Embodiment 10, the rewriting speed of the memory element is increased. Therefore, the speed of the rewrite operation of the semiconductor storage device can be increased.
The memory element having the best mode as described above is most preferably used in Embodiment 10. Thereby, the best performance of a semiconductor storage device can be obtained.
A semiconductor storage device according to Embodiment 11 of the present invention is the same as the semiconductor storage device of Embodiment 10, except that a selection transistor is not used.
Memory elements 401aA to 401aD, . . . , 401nA to 401nD are connected directly to first bit lines 416A1 to 416D1 and second bit lines 416G2 to 416D2, respectively.
A method for operating the semiconductor storage device is similar to that of the semiconductor storage device of Embodiment 10 (FIG 20). Note that since no selection transistor is used, a rewriting voltage is applied to the source/drain of all memory elements connected to the same bit line pair during a rewrite operation. Therefore, it is necessary to design the memory elements so that any memory element, which is not connected to a selected word line, is not rewritten.
According to the semiconductor storage device of Embodiment 11, the scale of integration can be significantly improved since no selection transistor is used. Therefore, the manufacturing cost can be significantly reduced, resulting in an inexpensive semiconductor storage device.
The memory element of the present invention is free from over-pressure, which is a problem with EEPROM. Therefore, the memory element of the present invention is particularly suited for use in a semiconductor storage device which has a plurality of word line and no selection transistor, as in Embodiment 11. This is because when the threshold of a memory element (e.g., the memory element 401aA) is negative due to over-erasure, the first bit line 416A1 is constantly conductive to the second bit line 416A2, and therefore, memory elements connected between these bit lines cannot be selected.
In a semiconductor storage device according to Embodiment 12 of the present invention, the density of a memory element array region 521 is further increased.
In
In the semiconductor storage device, a read operation is performed as follows. Currents passing through two respective memory elements belonging to different blocks (e.g., memory elements 501aA1 and 501bB1) are supplied to one and the other inputs of a sense amplifier, respectively to detect a difference between the current. In this case, for example, one input of the sense amplifier is connected to a bit line A1, while the other input is connected to a bit line B1. In addition, a voltage (e.g., +1.8 V) appropriate for a read operation 19 applied to the bit lines A2 and 82. Paths of currents flowing in this situation are indicated by dashed lines with an arrow In
According to the semiconductor storage device of Embodiment 12, a bit line is shared by memory elements arranged in two neighboring lines, thereby making it possible to significantly improve the scale of integration. Therefore, the manufacturing cost is significantly reduced, resulting in an inexpensive semiconductor storage device.
Embodiment 13 relates to a semiconductor device comprising a plurality of memory elements, and more particularly, a semiconductor storage device, in which multiple bits (2 bits or more) of data can be stored in at least one of two memory function sections contained in a memory element.
The memory array 2100 comprises a plurality of memory elements. Each memory element maybe, for example, one described in Embodiments 1 to 9. Specifically, any memory element can be used, which comprises a gate electrode provided via a gate insulating film on a semiconductor layer, a channel region provided under the gate electrode, a first diffusion region and a second diffusion region provided on opposite sides of the channel region and having a conductivity type opposite to that of the channel region, and a first memory function section and a second memory function section provided on opposite sides of the gate electrode and having a function of retaining charges.
In
The gate electrode of a memory element is connected to one of a plurality of word lines. The first diffusion region and the second diffusion region of a memory element are each connected to one of the bit lines.
The semiconductor storage device 2000 is preferably provided on a single chip.
The semiconductor storage device 2000 further comprises a state machine 2114.
An address is supplied from a processor 2999 via an address bus 2210 to the state machine 2114. Date to supplied via a date bus 2212 to the state machine 2114. Various commands, such as a program command, an erase command, a read command, and the like, are supplied from the processor 2999 via the data bus 2212 to the state machine 2114. The state machine 2114 controls execution of a program operation with respect to the memory array 2100 in response to a program command. The state machine 2114 controls execution of an erase operation with respect to the memory array 2100 in response to an erase command. The state machine 2114 controls execution of a read operation with respect to the memory array 2100 in response to a read command. Data read from the memory array 2100 is output from the state machine 2114 via the data bus 2212 to the processor 2999.
The semiconductor storage device 2000 further comprises a row decoder 2110 and a write control circuit 2106.
To control execution of a program operation with respect to the memory array 2100, the state machine 2114 outputs a row address to the row decoder 2110, and outputs a column address to a write control circuit 2106. With a row address and a column address, the position (a row position and a column position) of a memory element to be subjected to a program operation and the position (a first memory function section A or a second memory function section B) of a memory function section to be subjected to the program operation are specified.
The row decoder 2110 selects one of a plurality of word lines in accordance with a row address. Such a selection is performed by applying a prescribed voltage to one of the word lines. The selected word line is the word line to which the prescribed voltage is applied by the row decoder 2110. For example, the prescribed voltage is generated by the row decoder 2110 based on a power supply voltage supplied from a voltage supply circuit 2112.
The write control circuit 2106 applies a write pulse to a bit line, which is connected to one of the first diffusion region and the second diffusion region of a memory element connected to the selected word line, in accordance with a column address. A voltage required for applying a write pulse to a bit line is, for example, generated an the write control circuit 2106 based on a power supply voltage supplied from the voltage supply circuit 2112.
In order to write data to the first memory function section A of the memory element 2102apositioned at the upper left corner of the memory array 2100 in
The write control circuit 2106 is capable of controlling the application of a write pulse in accordance with the value of multibit data to be written, in order to write multibit data (2 bits or more) into the first memory function section A of the memory element 2102a.
For example, to write two-bit data, the number of applied write pulses or the duration of application of write pulses may be controlled depending on the value of two-bit data to be written. For example, the number of applied write pulses or the duration of application of write pulses may be controlled as follows. When the value of data to be written “00” a write pulse is not applied. When the value of data to be written is “01”, two write pulses (2×T μs are applied. When the value of data to be written is “10”, four write pulses (4×T μs) are applied. When the value of data to be written is “11”, six write pulses (6×T μs) are applied. Here, T represents a prescribed time. For example, T is 20. Alternatively, the voltage level of a write pulse may be controlled depending on the value of two-bit data to be written. Alternatively, the application of a write pulse may be controlled in accordance with a combination of at least two of the number of applied write pulses, the application duration, and the voltage level.
By controlling the application of a write pulse depending on the value of multibit data to be written, the quantity of charges retained in the first memory function section A of the memory element 2102acan be controlled to correspond to the value of multibit data. For example, application of a write pulse is controlled as follows. When the value “00” of two-bit data is written, the quantity of charges retained in the first memory function section A of the memory element 2102ais approximately 0. When the value “01” of two-bit data is written, the quantity of charges retained in the first memory function section A of the memory element 2102ais approximately E1 (>0). When the value “10” of two-bit data is written, the quantity of charges retained in the first memory function section A of the memory element 2102ais approximately E2 (>E1). When the value “11” of two-bit data is written, the quantity of charges retained in the first memory function section A of the memory element 2102ais approximately E3 (>E2) Here, the values of E1, E2, and E3 are appropriately designed so as to be distinguished from one another.
in order to write data into the memory function section B of the memory element 2102a, for example, a voltage of +5 V is applied to the word line 2214a, the bit line 2216a is grounded, and a write pulse (+5 V) is applied to the bit line 2216b. A method for writing multibit date (2 bits or more) into the second memory function section B of the memory element 2102ais substantially the same as that for the first memory function section A of the memory element 2102a.
In this manner, multibit data can be stored in at least one of the memory function section A and the memory function section B of the memory element 2102a. Multibit data can be stored in at least one of the memory function section A and the memory function section B of the other memory elements contained in the memory array 2100 as well as the memory element 2102a.
In order to confirm the normal completion of writing of multibit data, for example, after a write operation is performed, data is read from a memory element into which the data has been written, and the read data is compared with reference data. Such an operation is referred to as a verify operation. A method for determining the value of multibit data read from a memory element will be described below.
Note that the state machine 2114 controls the execution of an erase operation with respect to the memory array 2100 in a manner similar to the method for controlling the execution of a program operation with respect to the memory array 2100. In order to erase data stored in the first memory function section A of the memory element 2102a, for example, a voltage of −5 V is applied to the word line 2214a, an erase pulse (+5 V) is, applied to the bit line 2216a, and the bit line 2216b is grounded. In order to erase data stored in the second memory function section B of the memory element 2102a, for example, a voltage of −5 V is applied to the word line 2214a, the bit line 2216a is grounded, and an erase pulse (+5 V) is applied to the bit line 2216b.
The semiconductor storage device 2000 further comprises a column decoder 2108 and a sense amplifier 2116.
In order to control execution of a read operation with respect to the memory array 2100, the state machine 2114 outputs a row address to the row decoder 2110, and outputs a column address to the column decoder 2108. With a row address and a column address, the position (a row position and a column position) of a memory element to be subjected to a read operation and a position (a first memory function section A or a second memory function section B) of a memory function section to be subjected to the read operation are specified.
The row decoder 2110 selects one of a plurality of word lines in accordance with a row address. Such a selection is performed by applying a proscribed voltage to one of the word lines. The selected word line is the word line to which the prescribed voltage is applied by the row decoder 2110. For example, the prescribed voltage is generated by the row decoder 2110 based on a power supply voltage supplied from a voltage supply circuit 2112.
The column decoder 2108 selects at least one of a plurality of bit lines in accordance with a column address. Such a selection is made by selectively connecting the same number of bit lines as that of data values to be simultaneously output among the bit lines, to the sense amplifier 2116. Thus, the selected bit line is one that is connected to the sense amplifier 2116.
In order to read data from the first memory function section A of the memory element 2102a, for example, a voltage of +2 V is applied to the word line 2214a, the bit line 2216a is grounded, and a read pulse (+1.8 V) is applied to the bit line 2216b. Such a control is achieved by the state machine 2114 controlling the row decoder 2110 and the column decoder 2108 (e.g., the sense amplifier 2116 may apply a read pulse (biased voltage) to a bit line connected thereto). With such a control, a voltage of +2 V is applied to the gate electrode of the memory element 2102a, a voltage of 0 V is applied to the first diffusion region of the memory element 2102a, and a read voltage of +1.8 V is applied to the second diffusion region of the memory element 2102a. As a result, data is read from the first memory function section A of the memory element 2102a.
In order to read data from the memory function section B of the memory element 2102a, for example, a voltage of +2 V is applied to the word line 2214a, a read pulse (1.8 V) to applied to the bit line 2216a, and the bit line 2216b is grounded.
In order to determine the value of multibit data (2 bits or more of data) read from one of the first memory function section A and the second memory function section B of the memory element 2102a, the sense amplifies 2116 comprises a comparison section for comparing the voltage of one selected bit line with the voltages of a plurality of different reference voltages and outputting a plurality of comparison results, and a determination section for determining the value of multibit data based on the comparison results.
When the voltage of a selected bit line is lower than a reference voltage Ref1 (e.g., 0.7 V), the value of two-bit data is determined to be “00”. When the voltage of a selected bit line is equal to or higher than the reference voltage Ref1, and is lower than a reference voltage Ref2 (e.g., 1.2 V), the value of two-bit data is determined to be “01”. When the voltage of a selected bit line is equal to or higher than the reference voltage Ref2, and is lower than a reference voltage Ref3 (e.g., 1.7 V), the value of two-bit data is determined to be “10∞. When the value of a selected bit line is equal to or higher than the reference voltage Ref3, the value of two-bit data is determined to be “11”. Here, the values of the reference voltages Ref1, Ref2, and Ref3 are previously designed, as appropriate, depending on the quantity of charges retained in the first memory function section A and the second function section B. The reference voltages Ref1, Ref2, and Ref3 are generated for example, by a reference voltage supply circuit 2118 based on the power supply voltage supplied from the voltage supply circuit 2112, and are supplied to the sense amplifier 2116.
The sense amplifier 2116 comprises a sense amplifier 2311, a sense amplifier 2312, a sense amplifier 2313, and an output decode circuit 2320. The sense amplifier 2311 compares the voltage of a selected bit line with the reference voltage Ref1, and outputs the result of the comparison C1 (“0” or “1”). The sense amplifier 2312 compares the voltage of a selected bit line with the reference voltage Ref2, and outputs the result of the comparison C2 (“0” or “1”). The sense amplifier 2313 compares the voltage of a selected bit line with the reference voltage Ref3, and outputs the result of the comparison C3 (“0” or “1”). The output decode circuit 2320 decodes the three comparison results C1 to C3 output from the sense amplifiers 2311 to 2313, respectively, and outputs two-bit data (“00”, “01”, “10”, or “11”). The two-bit data output from the output decode circuit 2320 is output to and stored in the data latch/buffer 2120.
When the quantity of charges stored in a memory function section subjected to a read operation corresponds to “00”, the outputs (C1, C2, C3) of the sense amplifiers 2311 to 2313 are (“0”, “0”, “0”). When the quantity of charges stored in a memory function section subjected to a read operation corresponds to “01”, the outputs (C1, C2, C3) of the sense amplifiers 2311 to 2313 are (“1”, “0”, “0”). When the quantity of charges stored in a memory function section subjected to a read operation corresponds to “10”, the outputs (C1, C2, C3) of the sense amplifiers 2311 to 2313 are (“1”, “1”, “0”) . When the quantity of charges stored in a memory function section subjected to a read operation corresponds to “11”, the outputs (C1, C2, C3) of the sense amplifiers 2311 to 2313 are (“1”, “1”, “1”). Therefore, the output decode circuit 2320 provides an output (“00”) for an input (“0”, “0”, “0”), an output (“01”) for an input (“1”, “0”, “0”), an output (“10”) for an input (“1”, “1”, “0”), and an output (“11”) for an input (“1”, “1”, “1”). Such a structure of the output decode circuit 2320 can be easily realized by combining, for example, logic gates.
Thus, in the example shown in
Note that
In the circuit configuration of
The reference voltage switch circuit 2340 outputs the reference voltage Ref1 as a reference voltage Ref. The comparison result C1 (“0” or “1”) of the sense amplifier 2311 is output to and retained in a one-bit latch 2331.
Thereafter, the reference voltage iswitch circuit 2340 switches the reference voltages Ref, and outputs the reference voltage Ref2 as a reference voltage Ref. The comparison result C2 (“0” or “1”) of the sense amplifier 2311 is output to and retained in a one-bit latch 2332.
Thereafter, the reference voltage switch circuit 2340 switches the reference voltages Ref, and outputs the reference voltage Ref3 as a reference voltage Ref. The a comparison result C3 (“0” or “1”) of the sense amplifier 2311 is output to and retained in a one-bit latch 2333.
The output decode circuit 2320 decode* the three comparison results C1 to C3 retained in the one-bit latches 2331 to 2333, and outputs two-bit data (“00”, “01”, “10” or “11”). The configuration and operation of the output decode circuit 2320 are similar to those described above with reference to
Thus, in the example shown in
Note that
Alternatively, the output decode circuit 2320 may be constructed to output one-bit data twice instead of outputting two-bit data at once. For example, a one-bit address (other than row addresses and column addresses) maybe input to the output decode circuit 2320. In response to an address value “0”, one-bit data (e.g., the most significant bit of two-bit data output from the output decode circuit 2320 of
Such an effect is more significant as the number of bits of multibit data to increased. In order to store M-bit (M to an integer of 2 or more), the output decode circuit 2320 may be preferably constructed to output one-bit data M times instead of outputting M-bit data at once.
In Embodiment 13, the semiconductor storage device 2000 is constructed to store two-bit data in at least one of the first memory function section and the second memory function section of a memory element. The semiconductor storage device 2000 can be similarly constructed to store 3 bits or more of data in at least one of the first memory function section and the second memory function section of a memory element.
Memory arrays of Embodiments 10 to 12 may be used instead of the memory array 2100 of
According to the semiconductor storage device 2000 of Embodiment 13, multibit data (2 bits or more) can be stored in at least one of the first memory function section and the second memory function section of a memory element. As a result, a memory capacity per memory element can be dramatically increased. That is, a large-capacity memory array can be realized without increasing the area of a memory array and the density of memory elements.
The above-described semiconductor storage device may be applied to, for example, a rewritable non-volatile memory for adjusting an image on a liquid crystal panel as shown in
A liquid crystal panel 1001 is driven by a liquid crystal driver 1002. The liquid crystal driver 1002 comprises a non-volatile memory portion 1003, an SRAM portion 1004, and a liquid crystal driver circuit 1005. The non-volatile memory portion comprises the memory element of the present invention, more preferably a semiconductor storage device according to any one of Embodiments 10 to 12. The non-volatile memory portion 1003 can be externally rewritten.
Information stored in the non-volatile memory portion 1003 is transcribed into the SRAM portion 1004 when the power of the apparatus is turned ON. The liquid crystal driver circuit 1005 can read information stored in the SRAM portion 1004 if required. By providing the SRAM portion 1004, the stored information can be read with high speed.
The liquid crystal driver 1002 may be provided external to the liquid crystal panel 1001 as shown in
Liquid crystal panels display gray scale images, which are varied by applying multiple levels of voltages to each pixel. Generally, there are variations among products in the correspondence between applied voltages and displayed gray levels. To address this problem, information for compensating for variations in individual products is stored after completion of the product, and based on the information, compensation is performed. Thereby, image quality can be made uniform among products. Therefore, a rewritable non-volatile memory for storing correction information is preferably carried on the product. As the non-volatile memory, the memory element of the present invention is preferably used, particularly a semiconductor storage device according to any one of Embodiments 10 to 12, with which the memory elements of the present invention are integrated.
When the memory element of the present invention is used as a non-volatile memory for adjusting an image on a liquid crystal panel, a process for integrating it with a circuit, such as a liquid crystal driver or the like, can be easily achieved, leading to a reduction in manufacturing cost. The semiconductor storage devices of Embodiments 10 to 12 are particularly suitable when a memory size is relatively small and the reliability or stability is valued. This is because the outputs of two memory elements (memory element pair) are input to the same sense amplifier, so that a difference between currents passing through the two memory elements having similar device structures can be detected, thereby performing a read operation reliably and stably. Therefore, an area per bit is large. However, when a memory size is small, the large area per bit of the semiconductor storage device can be accepted as compared to the areas of other circuits. A typical non-volatile memory for adjusting an image on a liquid crystal panel has a size of, for example, several kilobytes, i.e., a relatively small memory size. Therefore, the semiconductor storage devices of Embodiments 10 to 12 are particularly preferably used as non-volatile memories for adjusting an image on a liquid crystal panel.
A mobile telephone, which is a mobile electronic apparatus comprising the above-described semiconductor storage device, is shown in
The mobile telephone comprises a control circuit 811, a battery 812, an RF (radio frequency) circuit 813, a display portion 814, an antenna 815, a signal line 816, a power supply line 817, and the like. The control circuit 811 comprises the above-described semiconductor storage device of the present invention. Note that the control circuit 811 may be preferably an integrated circuit in which elements having the same structure as described in Embodiment 10 are used both as memory circuit elements and as logic circuit elements. As a result, the manufacture of integrated circuits can be facilitated, leading to a reduction in the manufacturing cost of mobile electronic apparatuses.
As described above, by using a semiconductor storage device in a mobile electronic apparatus, in which a memory portion and a logic circuit portion can be easily integrated together and a high-speed read operation can be obtained, the operation speed of the mobile electronic apparatus can be increased and the manufacturing cost can be reduced. Therefore, it to possible to obtain an Inexpensive, highly reliable, high-performance mobile electronic apparatus.
According to the semiconductor storage device of the present invention, multibit data (2 bits or more) can is be stored in at least one of the first memory function section and the second memory function section of a memory element. As a result, a memory capacity per memory element can be dramatically increased. That is, a large-capacity memory array can be realized without increasing the area of a memory array and the density of memory elements.
Each memory element in a memory array contained in the semiconductor storage device according to the present invention comprises a gate electrode provided via a gate insulating film on a semiconductor layer, a channel region provided under the gate electrode, a first diffusion region and a second diffusion region provided on opposite sides of the channel region and having a conductivity type opposite to that of the channel region, and a first memory function section and a second memory function section provided on opposite sides of the gate electrode and having a function of retaining charges. Therefore, the memory element can be formed by a process having high affinity to a standard process for typical transistors. Therefore, as compared with when a conventional flash memory is used as a non-volatile memory element and the flash memory and a peripheral circuit comprising a typical transistor are integrated together, the semiconductor storage device of the present invention can dramatically reduce the number of masks and the number of processing steps. Therefore, the yield of a chip can be improved, thereby reducing manufacturing cost.
Preferably, at least a portion of the first memory function section or the second memory function section of a memory element overlaps a portion of the first diffusion region or the second diffusion region. Thereby, a drain current (read current) can be increased. As a result, a read operation can be performed with high speed. In addition, a write operation can be performed at a low voltage without an auxiliary gate. As a result, the power consumption during a program operation can be reduced.
At least one of the first diffusion region and the second diffusion region of a memory element is preferably off set from the gate electrode. Thereby, a read current ratio between a write operation and an erasure operation can be increased. As a result, a read operation can be performed with high speed and the power consumption can be reduced.
At least one of the first memory function section and the second function section of a memory element preferably has a charge retaining film, which has a surface substantially parallel to a surface of a gate insulating film and has a function of retaining charges. Thereby, variations in a memory effect can be suppressed, whereby the probability of succeeding a rewrite operation can be increased. Therefore, the maximum pulse value can be set to be a value smaller than that of conventional techniques, thereby making it possible to complete a rewrite operation quickly in addition, the probability of failure in a rewrite operation can be reduced.
A semiconductor layer of the memory element preferably has a region in a vicinity of a diffusion region, which has a concentration higher than that of a vicinity of a surface of the semiconductor layer under the gate electrode. Thereby, a memory effect (i.e., a difference in threshold between a write operation and an erase operation) is enhanced. The memory element preferably comprises a charge retaining film, which has a surface substantially parallel to a surface of the gate insulating film and has a function of retaining charges, and an insulating film separating the charge retaining film from a channel region or a semiconductor layer. The thickness of the insulating film is smaller than that of the gate insulating film, and is preferably 0.8 nm or more. As a result, the amount of charges induced in the channel region or the well region is increased, thereby enhancing the memory effect.
Thus, the memory element of the present invention has a memory effect greater than that of conventional flash memories. Therefore, It is possible to obtain a greater design margin for determining whether the memory element is in the program (write) state or the erase state, thereby increasing the probability of a successful rewrite operation. As a result, the value of a maximum pulse can be set to be smaller than that of conventional techniques. Therefore, a rewrite operation can be completed quickly. Moreover, the possibility that an error occurs in a rewrite operation is reduced.
At least one of the first memory function section and the second memory function section of a memory element preferably comprises a charge retaining film having a function of retaining charges. The charge retaining film preferably extends substantially parallel to a side of the gate electrode. As a result, the rewriting rate to increased. A semiconductor layer of the memory element preferably is a region in a vicinity of a diffusion region, which has a concentration higher than that of a vicinity of a surface of the semiconductor layer under the gate electrode. Therefore, the rewriting rate is increased. The memory element comprises a charge retaining film for retaining charges, which has a surface substantially parallel to a surface of the gate insulating film, and an insulating film separating the charge retaining film from a channel region or a semiconductor layer. The thickness of the insulating film is smaller than that of the gate insulating film, and is preferably 0.8 nm or more. Thereby, it is made easier to implant charges into the memory function section. As a result, the rewriting rate is increased.
Thus, the memory element of the present invention has a rewriting rate higher than that of conventional flash memories. Therefore, a rewrite operation can be completed quickly as compared to when the rewriting procedure described in the embodiment above is applied to conventional flash memories.
At least a portion of the first or second memory function section of the memory element preferably overlays a portion of the first or second diffusion region. In addition, at least one of the first and second memory function sections of the memory element comprises a charge retaining film having a function of retaining charges. The charge retaining film is preferably an insulating film. Thereby, problems (i.e., a write operation cannot be performed with an effective voltage and time period) caused due to lack of an auxiliary gate can be overcome while maintaining ease of miniaturization, ease of process, and high yield. Moreover, a write voltage can be reduced.
According to a mobile electronic apparatus according to the present invention, the above-described semiconductor storage device is provided therein. Therefore, it is easy to integrate a memory portion and a logic circuit portion together. Thereby, the operating speed of the mobile electronic apparatus can be improved, resulting in a reduction in manufacturing cost. As a result, an inexpensive, highly reliable, mobile electronic apparatus can be obtained.
Various other modifications will be apparent to and can be readily made by those skilled in the art without departing from the scope and spirit of this invention. Accordingly, it is not intended that the scope of the claims appended hereto be limited to the description as set forth herein, but rather that the claim be broadly construed.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2003-142492 | May 2003 | JP | national |
