[go: up one dir, main page]

TW201104843A - A NAND based NMOS NOR flash memory cell, a NAND based NMOS NOR flash memory array, and a method of forming a NAND based NMOS NOR flash memory array - Google Patents

A NAND based NMOS NOR flash memory cell, a NAND based NMOS NOR flash memory array, and a method of forming a NAND based NMOS NOR flash memory array Download PDF

Info

Publication number
TW201104843A
TW201104843A TW99113777A TW99113777A TW201104843A TW 201104843 A TW201104843 A TW 201104843A TW 99113777 A TW99113777 A TW 99113777A TW 99113777 A TW99113777 A TW 99113777A TW 201104843 A TW201104843 A TW 201104843A
Authority
TW
Taiwan
Prior art keywords
voltage
transistor
flash memory
charge
line
Prior art date
Application number
TW99113777A
Other languages
Chinese (zh)
Inventor
Peter Wung Lee
Fu-Chang Hsu
Hsing-Ya Tsao
Original Assignee
Aplus Flash Technology Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US12/387,771 external-priority patent/US8072811B2/en
Application filed by Aplus Flash Technology Inc filed Critical Aplus Flash Technology Inc
Publication of TW201104843A publication Critical patent/TW201104843A/en

Links

Landscapes

  • Semiconductor Memories (AREA)
  • Read Only Memory (AREA)
  • Non-Volatile Memory (AREA)

Abstract

A NOR flash nonvolatile memory device provides the memory cell size and a low current program process of a NAND flash nonvolatile memory device and the fast, asynchronous random access of a NOR flash nonvolatile memory device. The NOR flash nonvolatile memory device has an array of NOR flash nonvolatile memory circuits. Each NOR flash nonvolatile memory circuit includes a plurality of charge retaining transistors serially connected in a NAND string. A drain of a topmost charge retaining transistor is connected to a bit line associated with the serially connected charge retaining transistors and a source of a bottommost charge retaining transistor is connected to a source line associated with the charge retaining transistors. Each control gate of the charge retaining transistors on each row is commonly connected to a word line. The charge retaining transistors are programmed and erased with a Fowler-Nordheim tunneling process.

Description

201104843 l 六、發明'說明: 【發明所屬之技術領域】 [0001] 本發明是關於一種非揮發性記憶體陣列結構和操作,特 別是一種以NAND為基礎的N0R快閃記憶體的結構和操作 【先前技術】 [0002]非揮發性記憶體是本技術領域的習知技術。非揮發性記 憶體的類型包括唯讀記憶體(R0M)、電子可編程唯讀記憶 體(EPROM)、電子可抹除可編程唯讀記憶(EEpR〇M)體、 NOR快閃記憶體和NAND快閃記憶體。目前,在諸如個人數 字助手、手機、小筆電和可攜式電腦、錄音機以及全球 的搜尋系統等等的應用中,快閃記憶體已成為更普遍的 非揮發性記憶體之一。快閃記憶體具有高密度、矽面積 小、低成本的優點,並且能在使用單一的低電壓的情況 下被重覆地編程和抹除。 []驾知的快閃記憶鱧結構使用譬如電荷儲存和電荷擷取的 電何保存結構。在非揮發性浮動閘記憶體裡電荷儲存 、、Ό構中代表數位數據的電荷被儲存在元件的浮動閘上。 被儲存的電荷會改變浮動閘記憶體的臨界電壓以使得數 位數據被儲存起來。然’在矽氧氮氧硅(S0N0S)或者金氧 氮氧硅(M0N0S)型單元裡的電荷擷取結構中,電荷是在雙 絕緣層之間的電荷擷取層被擷取。在S0N0S與MONOS元件 中’棟取層的電荷有同硅氮化物(SiNx)—樣相對高 的絕 緣係數(k)。 [0004] 099113777201104843 l VI. INSTRUCTIONS DESCRIPTION: TECHNICAL FIELD OF THE INVENTION [0001] The present invention relates to a non-volatile memory array structure and operation, and more particularly to a NAND-based NOR flash memory structure and operation [Prior Art] [0002] Non-volatile memory is a well-known technique in the art. Non-volatile memory types include read-only memory (R0M), electronically programmable read-only memory (EPROM), electronically erasable programmable read-only memory (EEpR〇M), NOR flash memory, and NAND Flash memory. Flash memory has become one of the more popular non-volatile memories in applications such as personal digital assistants, cell phones, small laptops and portable computers, recorders, and global search systems. Flash memory has the advantages of high density, small footprint, low cost, and can be repeatedly programmed and erased with a single low voltage. [] The flash memory structure of the driver knows how to store the structure, such as charge storage and charge extraction. In the non-volatile floating gate memory, the charge storage, the charge representing the digital data in the structure is stored on the floating gate of the component. The stored charge changes the threshold voltage of the floating gate memory so that the digital data is stored. However, in the charge extraction structure in the silicon oxynitride (S0N0S) or gold oxynitride (M0N0S) type unit, the charge is taken up by the charge extraction layer between the double insulating layers. In the S0N0S and MONOS components, the charge of the layer is relatively high (KN) compared to silicon nitride (SiNx). [0004] 099113777

非揮發性記憶體分為兩大類產品:快隨機存取非同步NOR 表單編號AOloi s 第4頁/共133頁 〇9! 201104843Non-volatile memory is divided into two categories: fast random access asynchronous NOR form number Aloi s Page 4 of 133 〇9! 201104843

[0005] 〇 快閃記憶體和數慢的串列存取同步NAND快閃記憶體。目 前所設計的NOR快閃記憶體有複數外接地址和數據引腳以 及適當的控制訊號。它的一個缺點是當記憶密度加倍時 ,由於增加一外部地址引腳將會引起所需的外接針腳數 目的增加。相反’ NAND快閃記憶體的優點是比NOR的引腳 數少而且無地址輸入引腳。當密度增加時,NAND快閃記 憶體引腳數量始終保持不變。作為當今生產中的兩個主 流’ NAND和NOR快閃記憶體的單元結構均使用一次充電 保留(電荷存儲或電荷擷取)電晶體記憶單元,用於把 數據的一位元當作零荷儲存,由此其亦通常被稱作單階 電位編程單元(SLC)。 NAND和NOR快閃記憶體的優點是可在系統内碥程和抹除 並且具有至少十萬次的忍耐週期。此外,因為單元尺寸 可南度擴展’單晶片NAND和NOR快閃記憶體都能提供 giga兆位元組密度。例如’目前一位元/單電晶體nand 皁元尺寸大約保持在4:,2 τ( λ是一半導體製程中最小的 特性尺寸)’而NOR單元尺寸大約是10λ2。又,除用兩個 臨界電壓(VtO和Vtl)的單階電位編程單元儲存數據之 外,單電晶體NAND和NOR快閃記憶體能夠在一實體的單 元中有四個多電位臨界電壓(VtO、Vtl、Vt2和Vt3)時, 至少可在每一單元内儲存兩位元或者每一電晶體儲存兩 位元。 目前,一單晶片雙層多晶矽閘NAND快閃記憶體的最高容 量是64Gb。相比之下,一雙層多晶矽閘N〇R快閃記憶體有 2Gb的容量。NAND快閃記憶體和NOR快閃記憶體之間容 099113777 表單煸號A0101 第5頁/共133頁 0992024278-0 [0006] 201104843 量’的大差距是由於N A N D快閃記憶體單元的擴展性比N 0 R快 閃記憶體單元的擴展性優越。一NOR快閃記憶體單元的從 汲極到源極之間的電壓(Vds)必須要5. 0V方能保持高電 流通道熱電子(Channel-Hot-Electron,CHE)的編程操 作。相反地,一 NAND快閃記憶體單元進行低電流福勒-諾德海姆(Fowler-Nordheim)穿随效應編程操作時只須 要其汲極到源極之間的電壓為0. 0V。上述的結果使得一 位元/單電晶體NAND單元的尺寸僅是一位元/單電晶體 N0R單元的一半,這使得NAND快閃記憶體適用於需要巨大 數據儲存的應用上,而NOR快閃記憶體則廣泛地被用作需 要少量數據儲存但要求快速和非同步的隨機讀取的編程 代碼儲存記憶體上。 [0007] 雙電晶體NOR快閃記憶體單元由兩個NM0S電晶體形成, 它的構造相當於一單階編程單元。該電晶體NOR單元中上 部的電晶體是一浮動閘電晶體而底部的電晶體是一常用 的NM0S選擇電晶體。僅僅上部的電晶體有能力儲存數據 。雙電晶體NOR快閃記憶體單元僅有一個電晶體可保存數 據,在該NOR快閃記憶體單元中,每一NAND單元對應一 選擇電晶體。 [0008]美國專利7, 263, 003 號(Edahiro, et al.)描述一雙 電晶體快閃記憶體使用一複製單元陣列來控制主要的單 元陣列的預先充電/放電和感應放大器電路。 [0009]美國專利5, 596, 523號(Endoh, et al.)提供了一NOR型 EEPR0M記憶體單元陣列中的一段。每兩個相鄰的NOR單 元連接到一相對應的位元線,其中一記憶體單元電晶體 099113777 表單編號A0101 第6頁/共133頁 0992024278-0 201104843 [0010][0005] 快 Flash memory and slow serial access synchronous NAND flash memory. The currently designed NOR flash memory has multiple external address and data pins and appropriate control signals. One disadvantage of this is that when the memory density is doubled, the addition of an external address pin will cause an increase in the number of external pins required. Conversely, the advantage of NAND flash memory is that it has fewer pins than NOR and has no address input pins. As the density increases, the number of NAND flash memory pins remains the same. As the two mainstream 'NAND and NOR flash memory cell structures in today's production, a one-time charge retention (charge storage or charge extraction) transistor memory cell is used to store one bit of data as a zero load. Thus, it is also commonly referred to as a single-order potential programming unit (SLC). The advantages of NAND and NOR flash memory are that they can be programmed and erased within the system and have a tolerance period of at least 100,000 times. In addition, because the cell size can be expanded south, 'single-chip NAND and NOR flash memory can provide giga megabyte density. For example, the current one-dimensional/single-crystal nand soap cell size is approximately maintained at 4:, 2 τ (λ is the smallest characteristic size in a semiconductor process) and the NOR cell size is approximately 10 λ2. In addition, single-crystal NAND and NOR flash memory can have four multi-potential threshold voltages (VtO) in a single unit, except for the single-stage potential programming unit with two threshold voltages (VtO and Vtl). For Vtl, Vt2, and Vt3), at least two bits may be stored in each cell or two cells may be stored in each transistor. At present, the maximum capacity of a single-chip double-layer polysilicon gate NAND flash memory is 64Gb. In contrast, a two-layer polysilicon gate N〇R flash memory has a 2Gb capacity. NAND flash memory and NOR flash memory between 099113777 Form nickname A0101 Page 5 / 133 pages 0992024278-0 [0006] 201104843 The large gap of the amount is due to the scalability ratio of NAND flash memory cells The N 0 R flash memory unit has excellent scalability. The voltage (Vds) from the drain to the source of a NOR flash memory cell must be 5.0 V to maintain the programming operation of the channel-Hot-Electron (CHE). Conversely, a low-current Fowler-Nordheim wear-through effect programming operation requires only a voltage between its drain and source of 0. 0V. The above results make the size of a single-element/single-crystal NAND cell only half that of a single-element/single-transistor NMOS cell, which makes NAND flash memory suitable for applications that require huge data storage, while NOR flashes. Memory is widely used as a programming code storage memory that requires a small amount of data storage but requires fast and asynchronous random reads. [0007] A dual transistor NOR flash memory cell is formed from two NMOS transistors, the construction of which is equivalent to a single order programming unit. The upper transistor in the transistor NOR cell is a floating gate transistor and the bottom transistor is a commonly used NMOS selective transistor. Only the upper transistor has the ability to store data. The dual transistor NOR flash memory cell has only one transistor to hold data, and in the NOR flash memory cell, each NAND cell corresponds to a select transistor. [0008] U.S. Patent No. 7,263,003 (Edahiro, et al.) describes a dual-cell flash memory using a replica cell array to control the pre-charge/discharge and sense amplifier circuits of the main cell array. [0009] U.S. Patent No. 5,596,523 (Endoh, et al.) provides a section of a NOR-type EEPROM memory cell array. Each two adjacent NOR cells are connected to a corresponding bit line, one of which is a memory cell transistor 099113777 Form No. A0101 Page 6 of 133 Page 0992024278-0 201104843 [0010]

[0011] Ο [0012] 099113777 的没極和另-單元電晶體的源極共同接連至_位.元線。 其他單元電晶體的源極和汲極共同連接到—個源極線。 該源極線由一選擇電晶體來提供。 美國專利6, 765’825號(Scott)描述了 —包括雙浮動閘 電晶體的微差NOR記憶體單元。每—電晶體的沒極終端連 接到-相應的微差位it線。該雙電晶體的源極終端連接 到一共用的電流源極或下沉極。每一控制閘終端連接到 一相應的字元線,該字元線可以與其它控制終端所連接 的相應的字元線相同或者不同《該浮動閘電晶體可以是 五終端元件’其包括一增加的井終端,在該情況下,用 來編程EEPRGM記憶體單元的組位元線與用來讀取EEpR〇M 記憶體單元的組位元線不同。當汲極終端連接到微差讀 " - * λ ‘ 15 取位元線時’每一井終端連接到一相對應的微差編程位 元線。 美國專利申請2006/0181 925號(Specht,et al.)是一 種非揮發性記憶體單元的排列°其中記憶電晶體被排成 橫列和直行。第一直行的記憶電晶體的源極與沒極終端 柄和到與第二直行的記憶電晶體的第一源極與沒極終端 不同的金屬面的導線。按這種方法才可能使記憶體裡相 鄰直行的記憶電晶體互相靠近。 【發明内容】 本發明的目的是提供一種NOR快閃記憶體,其既有NAND 的小尺寸和低電流編程操作的特點’又有NOR的快速和非 同步隨機讀取操作的特點。 為了達到前述目的,NOR快閃記憶體的一實施例包括把複、 表單煸號 A01Q1 % 7 I/* 133 I 0992024278-0 [0013] 201104843 數'電荷保存電晶體串連成一NAND串。‘ 電晶體的極連接到與料串連電> S的電何儲存 逆冤何保存電晶體有關的 :線,同時最底層的電荷儲存電晶體的源極連接到 與該等串連電荷保存電晶體有關的_源極線。在每一橫 列上的複數電荷保存電晶體的每—控_連接到同一字 ^線。該料連電荷保存電晶體在—第—類導電型井之 =⑽型二重井)。該第—類導電型井在—第二類導電 型井(則型井)之内形成。該第二類導電型深井又在-第 一類導電型的基板(P型基板)中形成。 [0014] [0015] 該等電荷保存電晶彻編程操作和抹除操作是#Fowler_ N〇rdheim«效應操作嘴將該請純存電晶體中一 被選擇的電制存電晶體料—單階如單元來進行編 程細作…大約+ 15._大約+2()肩的電壓以逐漸增大 的方式施加於«擇的電荷料電晶__閘和電荷 儲存電晶體的基極之間。該等未_擇的電荷保存電晶 體被-施加於該未被選擇的電趣存電萄的控制問和 基極之間的-少祕0.0V的中間„所阻止。該歷快 閃减懸電路佈局的尺寸大約是製造_快閃記憶體電路 製程技術的最小的特性尺寸的四倍。 為了抹除被選擇的電荷儲存電晶體,_大約出則大 約+ 20. 0V的高電壓被施加於被選擇的電荷儲存電晶體的 基極和控制閘之間。藉由對未被選擇的電荷保存電晶體 上施加偏壓使得該未被選擇㈣射轉電晶體的控制閉 和基極之間的㈣大财UV,從而可以禁止該等未被 選擇的電荷保存電晶體。 099113777 表單編號A0101 第8頁/共133頁 0992024278-0 201104843 [0016]當從該等以單^編程單元進行編程操作的電為保·存電晶 禮t讀取-被選擇的電荷錯存電晶趙時,源極線連子接= -電磨追隨感應電路。該被選擇的電荷餚存電晶鍾的閘 極和汲極的電壓被設置到—電壓源(VDD)的電位,大約 1,8V或者大約3· 0V ^在該等電荷保存電晶體之内所有未 被選擇的電荷保存電晶體的閘極的電壓均被設置於一大 於6. 0V的第一高讀取電壓,若N〇R快閃記憶體電路未被選 擇執行讀取操作,該等電荷保存電晶體裡未被選擇的電 荷保存電晶體的控制閘的電壓被設置到接地參考電壓以 〇 關閉該電荷保存電晶體。電壓追隨感應電路是一比較電 路,其參考端連接到參考電壓源。該參考電壓源被設置 在大約2. 0V以區別第一邏輯水平(〇)的臨界電壓和第二 邏輯水平(1)的臨界電壓。 [0017] 〇 當從該等Μ階編程單元進行編簡作的電荷保存電曰 體中讀取—被選擇的電荷,存電顧時,源極線連二 一電壓追隨感應電路。該被選擇的電荷儲存電晶體的閘 Γ沒極的電屋被設置到大約4肩的高電I在電荷保 存電晶體内所有未被選擇荷H 壓均設置於一大於7.0V的第_Τ4=的閑極的電 感應電路包括複數比較電路:讀取:壓广厘追隨 =體之内代表數據的臨界電壓的數目減―。每—比 參考電^參考端連接到—組參考電壓源的其中之…令 參考電壓源被設置到在每_ 該 區別存錯於電荷儲存電 壓之間的-電壓’以 。 體巾㈣界電壓所代表的數據 099113777 表單蝙號ΑΟίοι 第9 I共133頁 201104843 _ .[0011] [0012] The source of the 099113777 and the source of the other-cell transistor are connected in common to the _ bit. The source and drain of the other unit transistors are connected in common to one source line. The source line is provided by a selection transistor. U.S. Patent No. 6,765,825 (Scott) describes a differential NOR memory cell comprising a dual floating gate transistor. The terminal of each transistor is connected to the corresponding differential bit line. The source terminal of the dual transistor is connected to a common current source or sinker. Each control gate terminal is connected to a corresponding word line, which may be the same as or different from the corresponding word line to which the other control terminals are connected. "The floating gate transistor may be a five terminal element" which includes an increase The well terminal, in this case, the group bit line used to program the EEPRGM memory cell is different from the group bit line used to read the EEpR〇M memory cell. When the bungee terminal is connected to the differential read " - * λ ‘ 15 bit line, each well terminal is connected to a corresponding differential programming bit line. U.S. Patent Application No. 2006/0181 925 (Specht, et al.) is an arrangement of non-volatile memory cells in which memory cells are arranged in a row and straight. The source of the memory cell of the first row is the same as the terminal of the non-polar terminal and the wire of the metal surface different from the first source and the terminal of the second straight-line memory transistor. In this way, it is possible to bring the memory transistors in the memory adjacent to each other close to each other. SUMMARY OF THE INVENTION It is an object of the present invention to provide a NOR flash memory that combines the features of small size and low current programming operations of NAND with the features of NOR fast and asynchronous random read operations. In order to achieve the foregoing objectives, an embodiment of a NOR flash memory includes concatenating a form, a nickname A01Q1 % 7 I/* 133 I 0992024278-0 [0013] 201104843 number 'charge holding transistors into a NAND string. 'The pole of the transistor is connected to the string.> The electricity of the S is related to the storage of the transistor: the source of the lowest charge storage transistor is connected to the series of charge storage. The _ source line associated with the transistor. Each of the plurality of charge-preserving transistors in each row is connected to the same word line. The charge-contained transistor is in the - (10) type double well of the -type conductivity well. The first type of conductivity type well is formed within the second type of conductivity type well (the type well). The second type of conductivity type deep well is formed in a -first conductivity type substrate (P type substrate). [0015] The charge-preserving electro-cutter programming operation and the erasing operation are #Fowler_N〇rdheim« effect operating nozzles, which are selected in a purely stored transistor, a selected electrical storage crystal material - single order As the unit performs programming, the voltage of about +15_about +2() shoulder is applied in a gradually increasing manner between the selected charge cell __ gate and the base of the charge storage transistor. The unselected charge-storing transistors are blocked by the application of the unselected electrical charge control and the base between the base and the 0.0V. The size of the circuit layout is approximately four times the smallest feature size of the manufacturing-flash memory circuit process technology. To erase the selected charge storage transistor, a high voltage of approximately + 20. 0V is applied to approximately Between the base of the selected charge storage transistor and the control gate, by applying a bias voltage to the unselected charge holding transistor such that the unselected (four) transistor is controlled between the closed and the base (4) The big money UV, so that the unselected charge storage transistors can be prohibited. 099113777 Form No. A0101 Page 8 of 133 pages 0992024278-0 201104843 [0016] When programming from these single programming units The electric power is saved and stored in the crystal ceremony t-reading - the selected charge is stored in the memory crystal Zhao, the source line is connected to the sub-connection = - electric grinder follows the induction circuit. The gate of the selected electric charge storage electric crystal clock And the voltage of the drain is set to - the voltage source (VDD) The voltage of the gate of all the unselected charge-storing transistors in the charge-storing transistor is set to a first high-read voltage greater than 6.0 V. If the N〇R flash memory circuit is not selected to perform a read operation, the voltage of the control gate of the unselected charge holding transistor in the charge holding transistor is set to the ground reference voltage to turn off the charge storage. The voltage tracking follower circuit is a comparator circuit whose reference terminal is connected to a reference voltage source. The reference voltage source is set at about 2.0 V to distinguish the threshold voltage of the first logic level (〇) from the second logic level ( 1) The threshold voltage. [0017] 读取 When reading from the charge-preserving electrode body that is programmed from the first-order programming unit, the selected charge is stored, and the source line is connected with the voltage of the second one. Inductive circuit. The selected gate of the selected charge storage transistor is set to a high electrical potential of about 4 shoulders. All unselected H voltages in the charge holding transistor are set at a voltage greater than 7.0V. The idleth of the first _Τ4= The electric induction circuit includes a plurality of comparison circuits: reading: the number of threshold voltages representing the data within the body is reduced by -. Each of the reference voltage reference terminals is connected to the group of reference voltage sources. The reference voltage source is set to -voltage 'between the difference between the charge storage voltage and each of the differences. The data represented by the body towel (four) boundary voltage 099113777 form bat number ΑΟίοι 9 I 133 pages 201104843 _ .

[晒在*另一實施例中,一賴快閃記憶體包括一職快閃記憶 體電路的陣列,其中臓快閃記憶體電路的電荷保存電晶 體排列成直行和橫列。每-_快閃記憶體電路包括一在 一直行上被連㈣連成-議D串的複數電荷保存電晶體 以使得該等電荷保存電晶體中的至少一個作為一選擇 問極電晶體,用於避免當該等電荷保存電晶體沒有被選 擇進行讀取操作時流經該等電荷保存電晶體之間的電流 洩漏。每一N0R快閃記憶體電路中最上端的電荷儲存電晶 體的沒極連接到與N0R快閃記憶體電路所在的直行相對應 的—本地的位元線,該本地的位元線與該直行相平行。 每一 _快閃記㈣電路中的最下賴,難存電晶㈣ 源極連接到與N0R快閃記憶谶意路相對應的一本地的源極 線,該本地的源極線與對應的位元線相平行^每一橫列 上的電荷保存電晶體的每一控制閘共同連接到一字元線 [0019] N0R快閃記憶體包括一直每電壓控制電g。該直行 控制電路連接到與每—電荷保存電晶體的直行相對應的 本地的位元線和源極線,並且提供控制訊號到與每一電 荷保存電晶體的直行有關的本地位元線(l〇cal bit 1 i ne)和源極線。每一本地位元線透過一位元線選擇電晶 體連接到一組全域位元線(Global bit line)之一, 同時每一本地的源極線透過一源極線選擇電晶體連接到 一組全域的源極線之一。該等全域的位元線和全域的源 極線連接到直行電壓控制電路以傳輸控制訊號到被選擇 的本地的位元線和被選擇本地的源極線,以對N〇R快閃記 099113777 表單編號A0101 第10頁/共133頁 0992024278-0 201104843 =:Γ荷保存_執行讀取操作、編 [0020] Ο NOR快閃記憶體包括—橫列電壓控制電路。該橫列電壓 控制電路連接到與每—電荷保存電晶_橫列相對應的 字儿線’並且提供控制訊號到與每—電荷保存電晶體的 橫列相關的字树,同時本地的位元線選擇電晶體和源 極線選擇電晶體的閘極連接到每一本地的位元線。該橫 列控制電路為了讀取、編程和抹除_快閃記憶體電路 中被選擇的電聽存電U轉輸控制訊朗字元線。 該橫列電壓控制電路亦傳輸控制錢频選擇的位元線 選擇電晶體_選擇的祕“晶體,以將位元線和源 極線控制魏從直行電壓㈣電路物職^擇的本地 的位元線和被選擇本地的源極線。 [0021] 電荷保存電晶體的編程和抹除依靠_FQffle卜 穿隨效應。將該等電荷保存電晶射被選擇的電荷保存 電晶體作為單階電位編程單元進行編程時,該橫列電壓 控制電路提供-施加於被選_:電賴存電晶體的控制 閘和基極之間的編程電壓到字元線,該編程電壓大約為 + 15. 0V到大約+ 20. 0V。該橫列電壓控制電路提供一小於 + 10. 0V的中間電壓,s亥中間電壓被施加於未被選擇的電 荷儲存電晶體的控制閘和基極之間以遮蔽未被選擇的電 荷保存電晶體。NOR快閃記憶體電路的佈局要求是每一 NOR快閃記憶體電路的尺寸大約是N〇R快閃記憶體電路製 程技術最小的特性尺寸的四倍。 099113777 將該等電荷保存電晶體中一被選擇的電荷保存電 表單編號A0101 第11頁/共133頁 晶體作 0992024278-0 [0022] 201104843 為—多階電位編程單元進行編程時,橫列電壓控制電路 在被選擇的電荷保存電晶體的控制閘和電荷保存電晶體 的基極之間以逐漸增大地方式施加—大約+ 15·⑽到大約 20. 0V的編程電廢到被選擇的電荷保存電晶體的字元線 :在每-切大該編程電壓時檢驗所讀取的被選擇的電 荷保存電晶體的數據,—直達到正麵臨界電壓。該等 電荷保存電晶體中未被選擇的電晶體被—施加於未被選 擇的電荷儲存電晶體的控制閘和基極之間的小於+ 10. 的中間高電壓所遮蔽。 [0023] 爲抹除被選擇的電荷保存電晶體,職列電壓控制電路 施加一大約+ 15.0V到大約+ 20.0V的正極的抹除電壓到被 選擇的電何儲存電晶體的基極和控制間之間1該橫列電 壓控制電路在未被選擇的電荷保存電晶澈上施加-偏壓 使得控制閘和基極之間有一大約〇. 〇v電壓,以遮蔽電荷 保存電晶體中未被選擇的電晶體? 1一:ν :;r: [0024] 將NOR快閃記憶體電路的該等電荷保存電晶體中—被選擇 的電荷保存電晶髏作秦單階電位編程單元進行讀取操作 時,源極線連接到直行電壓控制電路内的一電壓追隨感 應電路。該橫列電壓控制電路設置被選擇的電荷保存電 晶體的字元線,也就是控制閘的電壓到大約1. 8V或者大 約3. 0V的電壓源的電壓(VDD)。該橫列電壓控制啟動本 地的位元線選擇電晶體連接至與被選擇的電荷保存電晶 體相對應的全域的位元線和本地的位元線。之後,該直 行電壓控制電路設置全域的位元線的電壓,也是連接到 被選擇的電荷儲存電晶體的汲極的本地的位元線的電壓 099113777 表單編號A0101 第12頁/共133頁 0992024278-0 201104843 到電壓源(VDD)的電壓’該電壓源的電壓大約為ΐ8ν或 大約3· 。該橫列電脸制電路設置字元線和被選擇 = 記憶體電路之_等電荷保存電晶體中所有未 選擇的電相存電晶體的控制閘的電壓為—大於6. 〇 ν 雷=—讀取Μ。電壓追_應電路為在直行電壓控制 之内的一比較電路,JL — ^ 〃考端連接到參考電壓源 Ο [0025] Ο [0026] :參考電Μ源的電壓被設置到大⑽,以區別代表 臨界=水平⑻的臨界電壓和代表第二邏輯水平⑴的 4電屢。該橫列電壓㈣電路設置字认的電麼,也 ==擇的_快閃記憶體電路内電荷保存電晶體中未 '擇的電荷保存電晶艘的控制間的電墨為接地參考電 ’從而關閉該電荷保存電晶體。 》[In another embodiment, a flash memory includes an array of flash memory circuits in which the charge-storing transistors of the flash memory circuit are arranged in a straight line and a horizontal row. Each--flash memory circuit includes a plurality of charge-storing transistors connected to the four-string series on a straight line to make at least one of the charge-holding transistors serve as a selective transistor. To avoid current leakage between the charge-storing transistors when the charge-storing transistors are not selected for a read operation. The bottom of the uppermost charge storage transistor in each of the NOR flash memory circuits is connected to a local bit line corresponding to the straight line where the NOR flash memory circuit is located, the local bit line and the straight line parallel. The most bleak, hard-to-charge (4) source in each _ flash (4) circuit is connected to a local source line corresponding to the NOR flash memory path, the local source line and the corresponding bit The lines are parallel. Each of the control gates of the charge-storing transistor on each column is commonly connected to a word line [0019]. The N0R flash memory includes control of the voltage per voltage. The straight line control circuit is coupled to the local bit line and source line corresponding to the straight line of each charge holding transistor, and provides a control signal to the local bit line associated with the straight line of each charge holding transistor (l 〇cal bit 1 i ne) and source line. Each of the status lines is connected to one of a set of global bit lines through a single line selection transistor, and each local source line is connected to a group through a source line selection transistor. One of the source lines of the whole domain. The global bit lines and the global source lines are connected to the straight line voltage control circuit to transmit control signals to the selected local bit lines and the selected local source lines to flash the N99 R099113777 form. No. A0101 Page 10 of 133 pages 0992024278-0 201104843 =: Load save _ perform read operation, edit [0020] Ο NOR flash memory includes - course voltage control circuit. The row voltage control circuit is coupled to the word line corresponding to each of the charge holding transistors and provides a control signal to the word tree associated with each row of charge holding transistors, while the local bits The gates of the line select transistor and the source line select transistor are connected to each local bit line. The row control circuit transfers the control word line in order to read, program, and erase the selected electro-acoustic memory U in the flash memory circuit. The horizontal voltage control circuit also transmits a bit line selection transistor that controls the money frequency selection to select a crystal "selective crystal" to control the bit line and the source line to control the local bit of the direct line voltage (four) circuit object. The source line and the selected source line. [0021] The programming and erasing of the charge-storing transistor relies on the _FQffle pass-through effect. The charge-preserving electron-crystal-selected charge-preserving transistor acts as a single-order potential When the programming unit is programmed, the horizontal voltage control circuit provides a programming voltage applied to the word line between the control gate and the base of the selected NMOS transistor, the programming voltage is approximately + 15.0V. Up to about + 20. 0 V. The horizontal voltage control circuit provides an intermediate voltage less than + 10. 0 V, and an intermediate voltage is applied between the control gate and the base of the unselected charge storage transistor to shield The selected charge holds the transistor. The layout of the NOR flash memory circuit requires that the size of each NOR flash memory circuit be approximately four times the size of the smallest feature of the N〇R flash memory circuit technology. 099113777 Such A selected charge is stored in the storage transistor. Form No. A0101 Page 11 of 133 Crystals 0992024278-0 [0022] 201104843 When the multi-level potential programming unit is programmed, the horizontal voltage control circuit is selected. The control gate of the charge-storing transistor and the base of the charge-storing transistor are applied in an increasing manner—about + 15·(10) to about 0.02 V. The programmed electricity is discarded to the character of the selected charge-preserving transistor. Line: verifying the data of the selected charge-preserving transistor read every time the program voltage is cut-to-high, reaching the front threshold voltage. The unselected transistors in the charge-preserving transistor are applied. The intermediate high voltage between the control gate and the base of the unselected charge storage transistor is less than + 10. [0023] To erase the selected charge storage transistor, the rank voltage control circuit applies a An erase voltage of about +15.0V to about +20.0V of the positive electrode to the selected electrode between the base of the storage transistor and the control room. The row voltage control circuit saves the electrolysis in the unselected charge. Applying a bias voltage such that there is an approximately 〇. 〇v voltage between the control gate and the base to shield the charge from holding the unselected transistor in the transistor? 1: ν :; r: [0024] NOR flash memory In the charge-preserving transistor of the body circuit, when the selected charge-preserving transistor is used as the Qin single-order potential programming unit for reading operation, the source line is connected to a voltage following sensing circuit in the straight-line voltage control circuit. The horizontal voltage control circuit sets the word line of the selected charge holding transistor, that is, the voltage of the voltage source (VDD) of the control gate to a voltage of about 1.8 V or about 3.0 V. The course voltage control starts local. The bit line selection transistor is connected to the global bit line and the local bit line corresponding to the selected charge holding transistor. Thereafter, the straight-through voltage control circuit sets the voltage of the global bit line, which is also the voltage of the local bit line connected to the drain of the selected charge storage transistor. 099113777 Form No. A0101 Page 12 of 133 Page 0992024278- 0 201104843 Voltage to voltage source (VDD) 'The voltage of this voltage source is about ν8ν or about 3·. The horizontal electric circuit setting circuit sets the word line and the voltage of the control gate of all unselected electric phase storage transistors in the charge storage transistor that is selected to be equal to or greater than 6. 〇ν雷= Read Μ. The voltage chasing circuit is a comparison circuit within the straight-line voltage control, and the JL — ^ reference is connected to the reference voltage source Ο [0025] Ο [0026]: the voltage of the reference power source is set to a large (10) to The difference represents the critical voltage of the criticality = level (8) and the four electrical repetitions representing the second logic level (1). The horizontal voltage (four) circuit sets the word recognition power, and also == selected _ flash memory circuit charge storage transistor does not select the charge to save the electric ink between the control of the electric crystal ship as the ground reference power' Thereby the charge holding transistor is turned off. 》

F 讀取以多階電位編程單元進行編程操作的該等電荷保存 ^晶體中-被選擇的電荷儲存電晶體時,源極線連接到 -電壓追隨感應電路。閘極和被選擇的電荷儲存電晶體 的沒極的電隸設置到大約為4 ()V的—適中的高電壓。 在電荷保存電晶體中所有未被選擇的電荷保存電晶體的 閘極的電壓被設置到一大於7._第二讀取電壓。該電 壓追隨感應電路中有複數比較電路,其數目等於存儲於 電荷儲存電晶體之内代表數據的臨界電壓的數目減一。 每一比較電路的-參考端連接到—參考電壓源組中之一 。該參考電壓職設1到在每—臨界電壓之間的一電壓 ,以區別存在電荷儲存電晶義臨界電壓職表的數據 0 在另-實施例中’形成—瞧快閃記憶體的方法包括: 099113777 表單編號A0101 第13頁/共133頁 0992024278-0 201104843 k供一基板;在基板上設置一NOR快閃記憶體電路的陣 列’並使得N0R快閃記憶體電路的電荷保存電晶體被排 列成橫列和直行。每一 N 0 R快閃記憶體電路是把一直行上 的電荷保存電晶體串連成一 NAND串而形成的,以使得該 等電荷保存電晶體中的至少一個作為一選擇閘極電晶體 ’用於避免當該等電荷保存電晶體沒有被選擇進行讀取 操作時流經該等電荷保存電晶體之間的電流浪漏。每一 N〇R快閃記憶體電路中最上端的電荷儲存電晶體的汲極連 接到與N0R快閃記憶體電路所在的直行相對應的一本地的 位元線,該本地的位元線與該直行相平行。每一N〇R快閃 記憶體電路中最下端的電荷儲存電晶體的源極連接到與 N0R快閃記憶體電路相對應的一本地的硃極婊,該本地的 源極線與對應的位元線相平行。每一橫列上的複數電荷 保存電晶體的每一控制閘共同連接到一字元線。 [0027] 形成一 n 〇 R快閃記憶體的方法包括形成一直行電壓控制 電路。該直行電壓控制電路用於提供拉制訊號到與每一 電荷保存電晶體的直行相對應的本地的位元線和源極線 。每一本地的位元線透過一位元線選擇電晶體連接到全 域的位元線組中之一,且每一本地的源極線透過一源極 線選擇電晶體連接到全域的源極線組中之一。為了讀取 、編程和抹除N0R快閃記憶體電路之内被選擇的電荷保 存電晶體,全域的位元線和全域的源極線連接到直行電 壓控制電路,以傳輸控制訊號到被選擇的本地的位元線 和被選擇的本地的源極線。 [0028] 形成NOR快閃記憶體的方法包括形成一橫列電壓控制電 099113777 表單編號A0101 第14頁/共133頁 0992024278-0 201104843 〇 ^ °亥橫列電愿控制電路用於提供控制訊號到與每一電 :保存電晶_橫列相對應的字元線以及與每—本地的 曰疋線相連的本地的位元線選擇電晶體和源極線選擇電 體的閑極。為了讀取、編程和抹除NOR快閃記憶體電 路之内被選擇的電荷保存電晶體,該橫列控制電路傳輸 控制訊號到字70線。該橫列電壓控制電路亦將控制訊號 傳輸給被選擇的位元線轉電晶㈣被賴的源極線電 日曰體以把位疋線和源極線控制訊號從直行電壓控制電 路傳輪到被選擇的本地的位元線和被選擇本地的源極線 〇 [0029] 對及等電拍存電晶體進行編程操作和抹除操作依靠一 Fawler-Moi'dheim穿隧效應來完成β將電荷保存電晶體 群中被選擇的電荷保存電晶體作為單階電位編程單元進 行編程時,該橫列電壓控制電路提供-大約15. 0V到大約 2〇· 0V的尚的電壓到字元線,該電壓被施加於被選擇的電 荷儲存電明體的控制閘和基招之間。該橫列電壓控制電 〇 路提供-小於+ 10. 〇ν的中間電學’該中間電壓被施加於 未被選擇電4铸存電晶體的控制閘和基極區之間從而遮 蔽該未被選擇的電荷保存電晶體。NOR快閃記憶體電路佈 局的尺寸大約是製造N〇R快問記憶體電路製程技術的最小 的特性尺寸的四倍。 [0030] 將電荷保存電晶體中-被選擇的電荷保存電晶體作為一 多1¾電位編程單元進行編程時,該橫列電壓控制電路在 被選擇的電荷保存電晶體的控制閘和電荷保存電晶體的 基極之間以逐漸增大地方式施加_大約+ 15.剛大約 099113777 表單編號A0101 第15頁/共133頁 0992024278-0 201104843 徵〇V的編㈣壓到魏_電料存電㈣的字元線 。在每-次逐漸增大該編程電壓時檢驗所讀取的被選擇 的電荷保存電晶體的數據,—直到達到正確的臨界電壓 。該等電荷保存電晶體t未被選擇的電晶體被施加於未 被選擇的電荷儲存電晶體的控制閑和基極之間的一小於 + 10.0V的中間高電壓所遮蔽。 [0031] [0032] 爲抹除被選擇的電荷保存電晶體,該橫列電歷控制電路 施加大約+ 15. 0V到大約+ 20· 0V的正極抹除電壓到被選 擇的電荷儲存電鍾的基極和控制閘之間。該橫列電壓 控制電路還在未被選擇的電荷保存電晶體上施加一偏慶 使得控制閘和基極之間有—“G. 〇v的電壓,從而遮蔽 電%保存電晶體中未被選擇'的電晶:體。 讀取NOR快閃記憶體電路的該等電荷保存電晶體中一作為 單階電位編程單元的被選擇的電荷保存電晶體時,源極 線連接到直行電壓控制電路内的一電壓追隨感應電路。 該橫列電壓控制電路設置被選擇的電荷保存電晶體的字 元線即控制閘的電壓為大約1. 8V或者大約3, 0V的電壓源 (VDD)。該橫列電壓控制電路啟動本地的位元線選择電晶 體,以將與被選擇的電荷保存電晶體相對應的全域的位 元線和本地的位元線相連。之後,該直行電壓控制電路 設置全域的位元線的電壓即連接到被選擇的電荷儲存電 晶體的汲極的本地的位元線的電壓到電壓源(VDD),該電 壓源為大約1. 8V或者大約3· 0V。該橫列電壓控制電路還 設置字元線和被選擇N 0 R快閃記憶體電路之内該等電荷保 存電晶體中所有未被選擇的電荷保存電晶體的控制開的 099113777 表單編號A0101 第16頁/共133頁 0992024278-0 201104843 電壓到一大於6. 〇ν 路設置字元線的電°該橫列電壓控制電 該等電荷保存電晶被選擇的膽快閃記憶體電路的 制間的電壓到接地參2選擇的電荷保存電晶體的控 ⑽參考電壓,以關 :追隨感應電路為在直行電壓控制二*比 :’其具有-參考端連接到參考„源=二 壓源的電壓被設置到大約20ν,以 "考電 平(〇)的臨界電壓和 。< 表第-邏輯水 表第-邏輯水平⑴的臨界電壓。 [0033] Ο 讀取電荷保存電晶體中以多階電位編程單 電晶體巾所有未被選擇的奸彳w Μ何保存 被設置到-大於7期第1=電晶_極的電壓 路包括複數比較電路,其; 體之内代表數據的臨界__目 Ο 的一參考端連制-參考_為中之…參 被設置到在每-臨界電麼水平之間的一電壓,以區别= 儲於電荷儲存電晶體的臨界電髮所代表的數據。子 【實施方式】 [0034] 圖1A是一 _S N姆快閃浮動間電晶體】0的俯視圖圖 1B是該_Ν爆快問浮動閑電晶體Π)的剖視圖。圖lc 是該_S NAND快問浮動間電晶體_示意圖。在由 _s NA_間浮動閘電晶體10形成的NAND單元串的技 通結構中,_S N綱快閃浮動閘電晶體Π)的沒極15^ 099113777 表單編號A0101 第17頁/共133頁 0992024278-0 201104843 源極20不需要觸點。在習知的NAND單元串中,最上層的 電晶體連接到一頂端的選擇電晶體,而最低層的電晶體 連接到一底端的選擇電晶體。頂端的選擇電晶體的没極 和底端的選擇電晶體的源極的觸點連接到位元線 (Bitline,BL)和源極線。這種習知的NAND單元串的結 構使得N Μ 0 S N A N D快閃浮動閘電晶體1 〇的尺寸在快間記 憶體結構中是最小的。 [0035] 該NMOS NAND快閃浮動閘電晶體10在p型基板(psuB)40 的最上層形成。一N型材斜擴散在p型暴板4〇的表層形成 一深N井(deep N-well,DNW)35。然後一p型材料擴散 在深N井35的表層形成一P井^0(一般稱,之為三重p井, triple P-well TPW)。然破一Ν型材料擴散到ρ井3〇的 表層中形成没極(D) 15和源極(S) 20。一第一多晶石夕層 在P井30上、汲極15和源極20之間的基極上方形成浮動閘 4 5。一第二多晶石夕層在浮動閘4 5的上方形成,以形成 NMOS NAND快閃浮動閘電晶體10的控制閘(G)25。該 NMOS NAND快閃浮動閘電晶體1 〇的閘長度是汲極丨5和源 極20之間的P井30的基極的通道。該NMOS NAND快閃浮動 閘電晶體10的通道寬度由汲極15和源極20的n擴散的寬 度決定。該NMOS NAND快閃浮動閘電晶體1〇一般的單元 尺寸大約是4;12,其中X軸長2λ以及Y軸長2λ。尺寸 Lambda (又)是製程内所能達到幾何特性的最小尺寸。 [0036] 該浮動閘45内儲存電子電荷,以改變NMOS NAND快閃浮 動閘電晶體10的臨界電壓。P型基板40在操作中被連接到 接地參考電壓源(GND)。深N井35連接到電壓源(yj)D)。 099113777 表單編號A0101 第18頁/共133頁 0992024278-0 201104843 在目前NMOS NAND快閃浮動閘電晶體10的設計中,電源 電壓是1.3V或者3.0V。三重P井30在正常讀取操作中連 接到接地參考電壓源。 [0037] 該NMOS NAND快閃浮動閘電晶體1〇在陣列中排列成橫列 和直行。第二多晶矽層即NMOS NAND快閃浮動閘電晶體 10的控制閘25延伸以形成一字元線,該字元線連接陣列 裡橫列上的每一NMOS NAND快閃浮動閘電晶體1〇。 [0038] Ο 一隧道氧化物50在汲極15和源極20之間的通道區32的上 方和浮動閘45之間被形痛-該隧道氧化物50—般的厚度 是 100A。電子在fowier-Nprdheiip填道編程和?〇*161·-^^(11^111通道抹除期間穿1^壤^^^擊_'(^161*_ 5 · -¾馨参馨 .IffF reads the charge stored in the multi-level potential programming unit. When the crystal is selected - the charge is stored in the transistor, the source line is connected to the - voltage follower sensing circuit. The gate and the selected pole of the charge storage transistor are set to a moderately high voltage of approximately 4 ()V. The voltage of the gate of all unselected charge-storing transistors in the charge-storing transistor is set to a second read voltage greater than 7.. The voltage follower sensing circuit has a plurality of comparison circuits equal to the number of threshold voltages representing data stored in the charge storage transistor minus one. The reference terminal of each comparison circuit is connected to one of the reference voltage source groups. The reference voltage is set to a voltage between each of the threshold voltages to distinguish the data in the presence of the charge storage electromorphic threshold voltage. In another embodiment, the method of forming a flash memory includes: 099113777 Form No. A0101 Page 13 of 133 Page 0992024278-0 201104843 k for a substrate; an array of NOR flash memory circuits is placed on the substrate 'and the charge storage transistors of the NOR flash memory circuit are arranged Align and go straight. Each of the N 0 R flash memory circuits is formed by serially connecting charge-holding transistors on a line to a NAND string such that at least one of the charge-holding transistors acts as a select gate transistor To avoid current leakage between the charge-holding transistors when the charge-storing transistors are not selected for a read operation. The drain of the uppermost charge storage transistor in each N〇R flash memory circuit is connected to a local bit line corresponding to the straight line where the NOR flash memory circuit is located, the local bit line and the Straight line parallel. The source of the lowermost charge storage transistor in each N〇R flash memory circuit is connected to a local Zhurong 相对 corresponding to the NOR flash memory circuit, the local source line and the corresponding bit The lines are parallel. Each control gate of the complex charge-storing transistor on each row is connected in common to a word line. [0027] A method of forming an n 〇 R flash memory includes forming a continuous line voltage control circuit. The straight line voltage control circuit is operative to provide a pull signal to a local bit line and source line corresponding to a straight line of each charge holding transistor. Each local bit line is connected to one of the global bit line groups through a one-element selection transistor, and each local source line is connected to the global source line through a source line selection transistor. One of the groups. To read, program, and erase the selected charge-storing transistor within the NOR flash memory circuit, the global bit line and the global source line are connected to the straight-line voltage control circuit to transmit the control signal to the selected The local bit line and the selected local source line. [0028] A method of forming a NOR flash memory includes forming a horizontal voltage control circuit 099113777 Form No. A0101 Page 14 / 133 pages 0992024278-0 201104843 〇 ^ °Haiyang electric control circuit is used to provide control signals to And each of the electricity: the word line corresponding to the save transistor _ course and the local bit line connected to each local 曰疋 line select the idle pole of the transistor and the source line selection. To read, program, and erase the selected charge holding transistor within the NOR flash memory circuit, the row control circuit transmits a control signal to word line 70. The row voltage control circuit also transmits a control signal to the selected bit line to the transistor (4) to the source line of the source line to transfer the bit line and source line control signals from the straight line voltage control circuit. To the selected local bit line and the selected local source line 〇 [0029] and the isoelectric memory transistor programming operation and erase operation rely on a Fawler-Moi'dheim tunneling effect to complete the β When the selected charge-storing transistor in the charge-storing transistor group is programmed as a single-order potential programming unit, the course voltage control circuit provides a voltage of about 15.0 V to about 2 〇·0 V to the word line. This voltage is applied between the control gate and the base of the selected charge storage electrical body. The series voltage control circuit provides - intermediate power less than + 10. 〇 ν 'the intermediate voltage is applied between the control gate and the base region of the unselected electric 4 cast transistor to shield the unselected The charge holds the transistor. The size of the NOR flash memory circuit layout is approximately four times the smallest feature size of the N〇R fast memory system process technology. [0030] When the charge-preserving transistor is programmed as a multi-level potential programming unit, the row voltage control circuit controls the gate and charge-storing transistor of the selected charge-holding transistor The bases are applied in a gradually increasing manner _about + 15. just about 099113777 Form No. A0101 Page 15 / 133 pages 0992024278-0 201104843 The compilation of the levy V (four) pressure to the word of Wei _ electricity storage (four) Yuan line. The data of the selected selected charge holding transistor is verified every time the programming voltage is gradually increased, until the correct threshold voltage is reached. The non-selected transistors of the charge holding transistors t are masked by an intermediate high voltage applied between the control idle and base of the unselected charge storage transistors of less than +10.0V. [0032] To erase the selected charge holding transistor, the course electrical control circuit applies a positive erase voltage of approximately + 15.0V to approximately +20·0V to the selected charge storage clock. Between the base and the control gate. The row voltage control circuit also applies a bias on the unselected charge-storing transistor so that there is a voltage of "G. 〇v" between the control gate and the base, so that the shielded electricity saves the transistor from being selected. 'Electrocrystal: body. When reading the charge-holding transistor of the NOR flash memory circuit as a single-stage potential programming unit in the charge-holding transistor, the source line is connected to the straight-line voltage control circuit. A voltage follower sensing circuit. The horizontal voltage control circuit sets the word line of the selected charge holding transistor, that is, the voltage of the control gate is about 1.8 V or a voltage source (VDD) of about 3, 0 V. The voltage control circuit activates a local bit line selection transistor to connect the global bit line corresponding to the selected charge holding transistor to the local bit line. Thereafter, the straight line voltage control circuit sets the global The voltage of the bit line is connected to the voltage of the local bit line of the drain of the selected charge storage transistor to the voltage source (VDD), which is about 1.8 V or about 3.0 V. Voltage The circuit also sets the word line and the selected N 0 R flash memory circuit within the charge storage transistor. All of the unselected charge holding transistors in the control transistor are turned on. 099113777 Form No. A0101 Page 16 of 133 Page 0992024278-0 201104843 Voltage to a greater than 6. 〇ν Road sets the power of the word line. The horizontal voltage control powers the charge to save the crystal. The voltage between the intersystems of the selected bile flash memory circuit is grounded. 2 Select the charge to hold the transistor's control (10) reference voltage to off: follow the induction circuit for the straight-line voltage control two * ratio: 'it has - the reference terminal is connected to the reference „ source=two voltage source voltage is set to about 20ν , to " test level (〇) threshold voltage and. < Table - Logical water table - the threshold voltage of the first logic level (1). [0033] 读取 Reading the charge-storing transistor to program a single-crystal lens with multiple-order potentials. All unselected traits w are saved to - more than 7 stages. The voltage path of the first = electro-crystal _ pole includes plural Comparing the circuit, which represents a critical __ directory of the data, a reference end connection - reference _ is in the middle ... the parameter is set to a voltage between each - critical level to distinguish = The data represented by the critical electric hair stored in the charge storage transistor. [Embodiment] [0034] FIG. 1A is a top view of a _S N flash flash transistor ○ 0 is a cross-sectional view of the 浮动 快 快 浮动 。 。 。 。 。 。 。 。 。 。 。). Figure lc is a schematic diagram of the _S NAND fast inter-floating transistor. In the technical structure of the NAND cell string formed by the _s NA_ floating gate transistor 10, the _S N class flash floating gate transistor 的) of the pole 15^ 099113777 Form No. A0101 Page 17 of 133 0992024278-0 201104843 Source 20 does not require contacts. In a conventional NAND cell string, the uppermost transistor is connected to a top select transistor, and the lowest layer transistor is connected to a bottom select transistor. The tip of the top select transistor and the source of the bottom select transistor are connected to the bit line (Bitline, BL) and the source line. The structure of this conventional NAND cell string is such that the size of the N Μ 0 S N A N D flash floating gate transistor 1 是 is the smallest in the fast memory structure. [0035] The NMOS NAND flash floating gate transistor 10 is formed on the uppermost layer of the p-type substrate (psuB) 40. An N-profile oblique diffusion forms a deep N-well (DNW) 35 on the surface of the p-type baffle. Then a p-type material diffuses in the surface of the deep N well 35 to form a P well ^0 (generally referred to as triple p-well TPW). However, a broken type of material diffuses into the surface layer of the well 3 to form a dipole (D) 15 and a source (S) 20. A first polycrystalline layer forms a floating gate 45 above the base between the P well 30 and the drain 15 and the source 20. A second polycrystalline layer is formed over the floating gate 45 to form a control gate (G) 25 of the NMOS NAND flash floating gate transistor 10. The gate length of the NMOS NAND flash floating gate transistor 1 汲 is the channel of the base of the P well 30 between the drain 丨 5 and the source 20. The channel width of the NMOS NAND flash floating gate transistor 10 is determined by the width of the n-diffusion of the drain 15 and source 20. The NMOS NAND flash floating gate transistor has a typical cell size of about 4; 12, where the X axis is 2λ long and the Y axis is 2λ. Dimensions Lambda (again) is the smallest dimension that can achieve geometrical properties within the process. [0036] The floating gate 45 stores an electron charge to change the threshold voltage of the NMOS NAND flash floating gate transistor 10. The P-type substrate 40 is connected to a ground reference voltage source (GND) in operation. The deep N well 35 is connected to a voltage source (yj) D). 099113777 Form No. A0101 Page 18 of 133 0992024278-0 201104843 In the current design of NMOS NAND flash floating gate transistor 10, the supply voltage is 1.3V or 3.0V. The triple P well 30 is connected to a ground reference voltage source during normal read operations. [0037] The NMOS NAND flash floating gate transistors 1 are arranged in a row and a straight line in the array. The second polysilicon layer, ie, the control gate 25 of the NMOS NAND flash floating gate transistor 10, extends to form a word line that connects each NMOS NAND flash floating gate transistor 1 in the array of the array. Hey. [0038] A tunnel oxide 50 is painfully formed between the upper portion of the channel region 32 between the drain 15 and the source 20 and the floating gate 45 - the thickness of the tunnel oxide 50 is 100A. Electronic programming in fowier-Nprdheiip and ? 〇*161·-^^(11^111 channel erasing period 1^ soil ^^^ _'(^161*_ 5 · -3⁄4 馨参馨.Iff

Nordheim通道抹&在習知的HAND'森作t (儲存的電子 從浮動閘45被射出並穿過隧道氧化物5〇到單元通道區32 ,最後進入三重P井30中。 [0039] Ο 圖1D是一單電晶體浮動閘龍0 S N A肋:决閃單元的編程電 位和抹除電位的雙臨界電壓分配表浮動閘45在抹除操 作之後,電子電荷減少使得NMOS NAND快閃浮動閘電晶 體10的臨界電壓降低。正常情況下,NMOS NAND快閃浮 動閘電晶體10在被抹除操作後,其臨界電壓大約是-2. 0V。相比之下,在Fowler-Nordheim通道編程過程中 ’電子被吸到浮動閘45處,使得NMOS NAND快閃浮動閘 電晶體10的臨界電壓增加到大約+2. 0V。習慣上,抹除操 作後大約-2. 0V的臨界電壓(VtO)被指定為邏輯數據值“ Γ ,編程後+ 2. 0V的臨界電壓(Vtl)被指定為的邏輯數 據值“0”。 099113777 表單编號A0101 第19頁/共133頁 0992024278-0 201104843 [0040] 電子的移除較難控制’因此在一陣列中,F〇wler_ Nordheim通道抹除過程中從浮動閘中移除電子電荷一般 是以一頁(512B)或者一扇區(64〇)的單位集體執行並 且抹除臨界電壓(VtO)有—較寬的分佈。相反地,編程操 作時把電子按照一可控制的方法注射到浮動閘中並且在 一位元一位元的基礎上執行(透過連接到汲極15的位元線 一次一NMOS NAND快閃浮動閘電晶體1〇執行),如此編程 臨界電壓(vti)的分佈是小於抹除臨界電壓(Vt〇)的分佈 並且被控制在0.5V之内。由於每一|^帅單元所儲存的抹 除狀態寬分佈的臨界電壓(Vt〇)和編程狀態窄分佈的臨界 電壓(vti)是兩個明顯不同的臨界電壓,關〇s Nand快閃 汙動閘電晶體10若僅儲存一二免位數拿--的位元,則被稱 為一單階電位編程或者SLC (Single-Level-Cell); NMOS NAND快閃浮動閘電晶體1〇儲存一錐元數據,則被 稱為單位元單電晶體(single_bit_〇ne_transi^〇r lblT)NM0S NAND快閃浮動閘:單元 [0041] 圖1E是一單電晶體浮動閘NM〇s NAND快閃單元的具有一 抹除電位和二個編程電位的四個臨界電壓分配表。 " I知 技術中,透過改變編程條件,根據NM〇s “肋快閃浮動 閘電晶體1G中浮動祕的電荷的數量可形成超過兩個臨 界電壓,一般指的是NMOS NAND快閃浮動閘單元的多階 電位編程或者MLC (multi-level cell)。在該實施例 中’有四個臨界電壓能被編程在NM〇s NAND快閃浮動門 電晶體10。最負極的臨界電壓Vt0是抹除電壓,其有一中 間值-2·〇ν,用以儲存一邏輯數據值“u,,。由於最負 099113777 表單編號A0101 第20頁/共133頁 0992024278-0 201104843 〇 極的臨界電壓vto是唯一的抹除狀態,即移去電子電荷, 因此談最負極的臨界電壓vto在臨界電壓(VtO、Vtl、 Vt2和Vt3)中有最寬闊的分佈。由於其它三個臨界電壓 (Vtl、Vt2和Vt3)從抹除狀態以可控制的方式增加電子 到浮動閘上’它們在編程狀態有比較狹窄的分佈。三個 正極、狹窄的編程臨界電壓被充分地分開而被檢測。在 該實施例中,三個臨界電壓的第一個Vtl具有大約+ 1.0V 的一標準值(normal value),用以儲存一邏輯數據值“ 10” 。三個臨界電壓的第二個Vt2具有大約+ 2. 0V的一標 準值,用以儲存一邏輯數據值“0Γ 。三個臨界電壓的 第三個Vt3具有大約+3. 0V的一標準值,用以儲存一邏輯 數據值“00” 。因為每一NMOS NAND快閃浮動閘電晶體 10均儲存四個明顯的臨界電壓狀態,每一NMOS NAND快 閃浮動閘電晶體10還均儲存兩個二進位數據位元並且被 稱為兩位元單電晶體NMOS NAND快閃單元(2b/lT)。 [0042] NMOS NAND快閃浮動閘電晶體10的臨界電壓(vtO、Vtl 〇 、Vt2和Vt3)的標準值可在不同的設計中有超過i. 〇v的 變化。在不同的NMOS NAND快閃浮動閘單元設計之間亦 可變化二位元的數據值與四臨界電壓狀態之間的分佈。 例如,一些NMOS NAND快閃浮動閘單元設計將邏輯數據 值“01”分配給第一正極臨界電壓Vtl、邏輯數據值“ 10 ”分配給第二正極的臨界電壓Vt2 ,或者負值抹除臨界電 壓vto可分配到邏輯數據值“00” 、第三正極的臨界電壓 Vt3可分配到邏輯數據值“11” 。 [0043] 圖2A是一 NMOS NOR快閃浮動閘電晶體11〇的俯視圖。圖 099113777 表單編號A0101 第21頁/共133頁 0992024278-0 201104843 ^是該關05 1^01?快閃浮動閘電晶體11〇的剖視圖。圖20 是該NMOS NOR快閃浮動閘電晶體110的示意圖。該NMOS NOR快閃浮動閘電晶體11 〇在三重p型基板14〇的最上的表 層中形成,從而形成於一三重P井之内。一N型材料擴散 到P型基板140的表層中以形成一深N井135。然後,一P 型材料擴散到深N井135的表層中以形成P井130(—般被 稱為三重P井)。然後,N型材料被擴散到P井130的表層 中以形成汲極(D) 115和自我對準源極(S) 120。第一多 晶矽層在P井130上、汲極115和源極120之間的基極上方 形成浮動閘145。在浮動閘145上方形成一第二多晶矽層 :::::: __ ::. ' : ’以形成NMOS NOR快閃浮動閘電晶體11 〇的控制閘(G) 125。自我對準源極120在兩個NMOS NOR快閃浮動閘電晶 體110的兩控制閘125的相鄰的兩柄第$多盖矽層之間自 我對準形成。自我對準源極120 —般用於減少NMOS NOR 快閃浮動閘電晶體110源極線的間距》 [0044] NMOS NOR快閃浮動閘電晶赛11〇的閘長度等於p井13〇上 、汲極115和源極120之間的基極的通道區132。NMOS f、 .. ..... ί β NOR快閃浮動閘電晶體11 0的通道寬度由汲極π 5和源極 120的Ν擴散的寬度決定。NMOS NOR快閃浮動閘電晶體 110 —般的單元尺寸大約是10又2,其中X軸長2. 5λ以 及Υ轴長4 λ。 [0045] 浮動閘145儲存電子電荷以改變NMOS NOR快閃浮動閘電 晶體110的臨界電壓。P型基板140在操作中連接到接地參 考電壓源(GND)。深N井135在讀取和編程操作中連接到 電壓源(VDD) ’然’其在Fowler-Nordheim通道抹除操 099113777 表單編號A0101 第22頁/共133頁 0992024278-0 201104843 作中的電壓為+ 10V左右。在目前NMOS NOR快閃浮動問電 晶體110的設計中,電源電壓是1. 3V或者3. 0V。三重p 井130在正常讀取和編程操作中連接到接地參考電壓,然 ,其在抹除操作中的電壓大約為+ 10V。換句話說,在 Fowler-Nordheim通道抹除操作期間,深n井135和=重 P井130有相同的偏電壓’其大約為+ ,以避免深n井 135和三重P井130之間P/N節點的前向漏電流。 [0046] 〇 NMOS NOR快閃浮動閘電晶體11〇在陣列中排列成橫列和 直打。第二多晶碎層即_s N0R快閃浮動間電晶體 的控制閘125延伸形成-字元線,該字元線連接到陣列裡 橫列上的每-_S臓快閃浮 [0047]The Nordheim channel wipe & is in the conventional HAND's work (the stored electrons are ejected from the floating gate 45 and pass through the tunnel oxide 5〇 to the cell channel region 32 and finally into the triple P well 30. [0039] Figure 1D is a single transistor floating sluice 0 SNA rib: double-voltage voltage distribution table of the programming potential and erase potential of the firing cell. Floating gate 45 after the erase operation, the electron charge is reduced so that the NMOS NAND flash floating gate The threshold voltage of the crystal 10 is lowered. Under normal conditions, the threshold voltage of the NMOS NAND flash floating gate transistor 10 after being erased is about -2.0 V. In contrast, during the Fowler-Nordheim channel programming process. The electrons are attracted to the floating gate 45, so that the threshold voltage of the NMOS NAND flash floating gate transistor 10 is increased to about +2.0 V. Conventionally, the threshold voltage (VtO) of about -2.0 V is erased after the erase operation. The logical data value "0", which is designated as the logical data value " Γ , the threshold voltage (Vtl) after programming + 2. 0V is designated as 0. 099113777 Form No. A0101 Page 19 / 133 pages 0992024278-0 201104843 [0040] The removal of electrons is more difficult to control, so it’s a while In the F〇wler_ Nordheim channel erase process, the removal of the electron charge from the floating gate is generally performed collectively in units of one page (512B) or one sector (64〇) and the threshold voltage (VtO) is erased. In contrast, the programming operation injects electrons into the floating gate in a controlled manner and performs on a one-bit one-bit basis (through the bit line connected to the drain 15 once, one NMOS NAND) The flash floating gate transistor is executed), so that the distribution of the threshold voltage (vti) is less than the distribution of the erase threshold voltage (Vt〇) and is controlled within 0.5 V. Since each unit is stored The widened threshold voltage (Vt〇) of the erased state and the narrowly distributed threshold voltage (vti) of the programmed state are two distinct threshold voltages, and only one or two of the s Nand flash slamming gate transistors 10 are stored. The bit-free bit is called a single-order potential programming or SLC (Single-Level-Cell); NMOS NAND flash floating gate transistor 1〇 stores a cone of data, which is called a unit. Single-single transistor (single_bit_〇ne_transi^〇r lblT NM0S NAND Flash Floating Gate: Unit [0041] Figure 1E is a four-voltage voltage distribution table with a erase potential and two programming potentials for a single transistor floating gate NM〇s NAND flash cell. In the technology, by changing the programming conditions, according to NM〇s “the number of floating secret charges in the rib flash floating gate transistor 1G can form more than two threshold voltages, generally refers to the multi-step of the NMOS NAND flash floating gate unit. Potential programming or MLC (multi-level cell). In this embodiment, there are four threshold voltages that can be programmed in the NM〇s NAND flash floating gate transistor 10. The most negative threshold voltage Vt0 is the erase voltage, which has an intermediate value of -2·〇ν for storing a logical data value "u,,. Since the most negative 099113777 form number A0101 page 20 / 133 pages 0992024278-0 201104843 The threshold voltage vto of the bungee is the only erased state, that is, the electron charge is removed, so the threshold voltage vto of the most negative electrode has the widest distribution among the threshold voltages (VtO, Vtl, Vt2 and Vt3). The threshold voltages (Vtl, Vt2, and Vt3) add electrons to the floating gate in a controlled manner from the erased state. They have a relatively narrow distribution in the programmed state. The three positive, narrow programming threshold voltages are sufficiently separated. Detected. In this embodiment, the first Vtl of the three threshold voltages has a normal value of approximately +1.0V for storing a logical data value of "10". The second of the three threshold voltages Each Vt2 has a standard value of approximately +2.00V for storing a logical data value of "0". The third Vt3 of the three threshold voltages has a standard value of approximately +3.00 V for storing a logical data value "00". Because each NMOS NAND flash floating gate transistor 10 stores four distinct threshold voltage states, each NMOS NAND flash floating gate transistor 10 also stores two binary data bits and is referred to as a two-bit binary. Single transistor NMOS NAND flash unit (2b/lT). [0042] The standard values of the threshold voltages (vtO, Vtl 〇 , Vt2, and Vt3) of the NMOS NAND flash floating gate transistor 10 may vary by more than i. 〇v in different designs. The distribution between the data values of the two bits and the state of the four threshold voltages can also vary between different NMOS NAND flash floating gate cell designs. For example, some NMOS NAND flash floating gate cell designs assign a logical data value of "01" to a first positive threshold voltage Vtl, a logical data value of "10" to a second positive threshold voltage Vt2, or a negative value to erase a threshold voltage. Vto can be assigned to the logical data value "00", and the third positive threshold voltage Vt3 can be assigned to the logical data value "11". 2A is a top plan view of an NMOS NOR flash floating gate transistor 11A. Figure 099113777 Form No. A0101 Page 21 of 133 0992024278-0 201104843 ^ This is a cross-sectional view of the closed 05 1^01? flash floating gate transistor 11〇. 20 is a schematic diagram of the NMOS NOR flash floating gate transistor 110. The NMOS NOR flash floating gate transistor 11 is formed in the uppermost surface layer of the triple p-type substrate 14A to be formed in a triple P well. An N-type material diffuses into the surface layer of the P-type substrate 140 to form a deep N well 135. A P-type material is then diffused into the surface of the deep N-well 135 to form a P-well 130 (generally referred to as a triple-P well). The N-type material is then diffused into the surface of the P-well 130 to form a drain (D) 115 and a self-aligned source (S) 120. The first polysilicon layer forms a floating gate 145 over the P well 130, above the base between the drain 115 and the source 120. A second polysilicon layer :::::: __ ::. ' : ' is formed over the floating gate 145 to form a control gate (G) 125 of the NMOS NOR flash floating gate transistor 11 . The self-aligned source 120 is self-aligned between the adjacent two handles of the two control gates 125 of the two NMOS NOR flash floating gate transistors 110. The self-aligned source 120 is generally used to reduce the pitch of the source line of the NMOS NOR flash floating gate transistor 110. [0044] The gate length of the NMOS NOR flash floating gate crystal game is equal to that of the p well. The channel region 132 of the base between the drain 115 and the source 120. The channel width of the NMOS f, .. ..... ί β NOR flash floating gate transistor 11 0 is determined by the width of the drain diffusion of the drain π 5 and the source 120. The NMOS NOR flash floating gate transistor 110 has a typical cell size of about 10 and 2, wherein the X-axis length is 2. 5λ and the x-axis length is 4 λ. [0045] The floating gate 145 stores electronic charge to change the threshold voltage of the NMOS NOR flash floating gate transistor 110. The P-type substrate 140 is operatively connected to a ground reference voltage source (GND). Deep N well 135 is connected to the voltage source (VDD) during read and program operations. 'Right' its voltage in Fowler-Nordheim channel eraser 099113777 Form No. A0101 Page 22 of 133 page 0992024278-0 201104843 + 10V or so. In the current design of the NMOS NOR, the power supply voltage is 1. 3V or 3. 0V. The triple p well 130 is connected to the ground reference voltage during normal read and program operations, however, its voltage during the erase operation is approximately +10V. In other words, during the Fowler-Nordheim channel erase operation, the deep n well 135 and the = heavy P well 130 have the same bias voltage 'which is approximately +' to avoid P/ between the deep n well 135 and the triple P well 130. Forward leakage current of the N node. [0046] NMOS NMOS NOR flash floating gate transistors 11 are arranged in a matrix in a row and straight. The second polycrystalline layer, i.e., the control gate 125 of the _s N0R flash floating inter-cell transistor, extends to form a word line that is connected to each of the -_S臓 flashes on the column of the array [0047]

一随道氧化物15〇在祕115和源極12〇之間的通道區i32 的上方和浮動閘145之間形成。暖道氧化物15〇-般的厚 度是100 A。電子電荷在高錢通道_子編程過程和 低電_FQWlendheim_抹除過程中穿過隨道氧 化物⑽。在習知的臟操作中,F〇wWheim通道 抹除操作把雌的電子從軸_5射出財過隧道氧化 物150到單元通道區132,最後進入三重1>井130中。 [0048] 099113777 在抹除操作後,儲存在浮動閘145㈣子電荷減少導致 NM〇S_快閃浮動閘電晶體川的第—臨界傾vt0) 減少到大約小於2. 5V。相比之下^ π 下’在通道熱電子編程操 作中,電子被吸入到浮動閘14 和⑺’以致NM〇s Ν〇Ι^^閃浮 動閘電晶體11 〇的第二臨界電龎 介I壓(vtl)被設置到大約大 於4. 0V。抹除狀態下的第— ^ &界電壓(Vt0)的寬分佈和 編程狀態下的第二臨界電壓 Utl)的窄分佈皆被設置到 表單編號_1 第23頁/共133頁 0992024278-0 201104843 正極,以避免任何由於NMOS NOR快閃浮動閘電晶體11〇 具有負極的臨界電壓所引起的誤讀取操作。 [0049] 圖2D是具有單階編程電位的一單電晶體浮動閘NM〇s N〇R 快閃單元的雙臨界電壓分配表。浮動閘145在抹除操作之 後,電子電荷減少使得NMOS NOR快閃浮動閘電晶體11〇 的臨界電壓降低。正常情況下,N Μ 0 S N 0 R快閃浮動閘電 晶體110在被抹除後’其臨界電壓的最大值大約是+ 2. 5V 。相比之下’在通道熱電子編程中,電子被吸入到浮動 閘145 ’以致NMOS NOR快閃浮動閘電晶體ι10的臨界電壓 增加到至少大約+4. 0V。習慣上,抹除操作後的大約 + 2.5V的臨界電壓(VtO)被指定為邏輯數捸值“丨,,,編 程操作後的+ 4.0V的臨界電壓:< V U >被指定為邏輯數據值 “0” 。與NMOS NAND快閃浮動閘電晶截_同,儲存單位 元數據的NMOS NOR快閃浮動閘電晶體丨丨〇被稱為單位元 單電晶體NMOS NOR快閃浮動閘單元(iblT)。 [0050] 圖2 E疋具有一個抹除電位和三個.編..程電位的一單電晶體 浮動閘NMOS NOR快閃單元的四個臨界電壓分配表。在習 知技術中,透過變化編程條件,並根據NMOS NOR快閃浮 動閘電晶體110中浮動閘14 5上電荷的數量,可形成超過 兩個的臨界電壓’一般被稱為NMOS NOR快閃浮動閘單元 的多階電位編程或者多階電位編程單元。本實施例中, 有四個臨界電壓能被編程在NMOS NOR快閃浮動閑電晶體 110。最小的正極寬分佈臨界電壓Vt0是抹除電壓,其有 一最大值+ 2· 5V,用以儲存一邏輯數據值“u” 。其他三 個正極窄分佈編程臨界電壓被充分地分開而允許被正確 099113777 表單編號Α0101 第24頁/共133頁 0992024278-0 201104843 心剛 檢測。本實施例中,三個臨界電壓中的第一個Vtl有一 大約+ 3. 5V的標準值,用以儲存一邏輯數據值“10” 。三 個臨界電壓中的第二個Vt2有一大約+4.5V的標準值,用 以儲存一邏輯數據值“01” 。三個臨界電壓的第三個Vt3 有一大約+ 5. 5V的標準值,用以儲存一邏輯數據值“00” 。由於每一 NMOS NOR快閃浮動閘電晶體110儲存了四 個明顯不同的正極臨界電壓狀態,每一NMOS NOR快閃浮 動閘電晶體110儲存了兩個二進位數據位元,故被稱為雙 位元單電晶體NMOS NOR快閃單元(2b/lT)。 NMOS NOR快閃浮動間電晶體110的臨界電壓Vtl和Vt2的 標準值可在不同的設計中有超過i. 0W的:變化'。臨界電壓 VtO和Vt3的標準值可以有一較寬界電癦分佈。例如 ,第一臨界電壓Vto可在大約1. 0V到大約2. 5V之間變化 。第四臨界電壓Vt3可有較寬的分佈,但它必須大於大約 4. 5 V以保證NMOS NOR快間浮動閛電晶體1 i 0在一非傳導 s I 狀態下。如前所述的NMOS NAND快閃浮動il單元,對應 〇 四個臨界電壓狀態的二位元數镇:值的分佈亦可在不同的 NMOS NOR快閃浮動閘單元設計之間變化。 [0052] wIntel StrataFlash ( Memory Technology Overview” ,Atwood等發表,英特爾技術期刊,第工卷 第2期,Q4 1997,wamv. intei.com,2007年4月 23 日, Intel Strst^F 13>sh( Memory Technology De — velopment and Implementation”,Fazio等發表於 英特爾技術期刊,第1卷苐2期,Q4 1997, www.intel.com’2009年4月 21 日,“ETOX ( Flash 099113777 表單編號A0101 第25頁/共133頁 0992024278-0 201104843A channel oxide 15 is formed between the channel region i32 between the secret 115 and the source 12 and the floating gate 145. The warmth oxide 15 〇-like thickness is 100 A. The electron charge passes through the accompanying oxide (10) during the high-cost channel _ sub-programming process and the low-power _FQWlendheim_ erase process. In conventional dirty operation, the F〇wWheim channel erase operation ejects female electrons from the axis _5 through the tunnel oxide 150 to the cell channel region 132 and finally into the triple 1 > well 130. 0伏。 [0048] 099113777 after the erase operation, stored in the floating gate 145 (four) sub-charge reduction caused by the NM 〇 S_ flash floating gate transistor transistor's first critical tilt vt0) reduced to about less than 2. 5V. In contrast, ^ π under 'in the channel hot electron programming operation, electrons are drawn into the floating gate 14 and (7)' so that NM〇s Ν〇Ι ^ ^ flash floating gate transistor 11 〇 second critical 厐 I 0伏。 The pressure (vtl) is set to be greater than about 4. 0V. The narrow distribution of the -^ & boundary voltage (Vt0) and the second threshold voltage Utl) in the programmed state are set to the form number_1 page 23/133 pages 0992024278-0 201104843 positive pole to avoid any mis-reading operation caused by the NMOS NOR flash floating gate transistor 11〇 having the negative voltage of the negative pole. 2D is a dual threshold voltage distribution table of a single transistor floating gate NM〇s N〇R flash cell having a single-order programming potential. [0049] FIG. After the erase gate 145 is erased, the electron charge is reduced such that the threshold voltage of the NMOS NOR flash floating gate transistor 11 降低 is lowered. Under normal conditions, the maximum value of the threshold voltage of the N Μ 0 S N 0 R flash floating gate transistor 110 after being erased is approximately + 2. 5V. In contrast, in the channel hot electron programming, electrons are drawn into the floating gate 145' so that the threshold voltage of the NMOS NOR flash floating gate transistor ι10 is increased to at least about +4.00 V. Conventionally, the threshold voltage (VtO) of approximately +2.5V after the erase operation is specified as the logical value 丨 "丨,,, the threshold voltage of + 4.0V after the programming operation: < VU > is designated as logic The data value is “0.” The same as the NMOS NAND flash floating gate transistor, the NMOS NOR flash floating gate transistor that stores the unit metadata is called the single-element single-transistor NMOS NOR flash floating gate unit. (iblT) [0050] FIG. 2 shows four threshold voltage distribution tables of a single transistor floating gate NMOS NOR flash cell having an erase potential and three .program potentials. In the prior art, By varying the programming conditions and according to the amount of charge on the floating gate 14 5 of the NMOS NOR flash floating gate transistor 110, more than two threshold voltages can be formed, which is generally referred to as an NMOS NOR flash floating gate unit. Potential programming or multi-step potential programming unit. In this embodiment, there are four threshold voltages that can be programmed in the NMOS NOR flash floating idler 110. The minimum positive width distribution threshold voltage Vt0 is the erase voltage, which has a maximum value. + 2· 5V for storing a logic According to the value "u". The other three positive polarity narrow programming threshold voltages are sufficiently separated to allow the correct 099113777 form number Α0101 page 24 / total 133 page 0992024278-0 201104843 heart detection. In this embodiment, three critical The first Vtl of the voltage has a standard value of approximately +3.5 V to store a logical data value of "10". The second of the three threshold voltages Vt2 has a standard value of approximately +4.5V for storage. A logical data value of "01". The third Vt3 of the three threshold voltages has a standard value of approximately +5.5 V to store a logical data value of "00". Since each NMOS NOR flashes the floating gate transistor 110 Four distinct positive threshold voltage states are stored. Each NMOS NOR flash floating gate transistor 110 stores two binary data bits, so it is called a dual-bit single-crystal NMOS NOR flash unit (2b). /lT) The standard values of the threshold voltages Vtl and Vt2 of the NMOS NOR flash floating inter-cell transistor 110 may exceed i. 0W in different designs: variation '. The standard values of the threshold voltages VtO and Vt3 may have a wider value. The distribution of electric power. For example, The voltage of the fourth threshold voltage Vt3 may have a wider distribution, but it must be greater than about 4.5 V to ensure that the NMOS NOR floats rapidly. 1 i 0 in a non-conducting s I state. As mentioned above, the NMOS NAND flash floating il unit corresponds to the two-bit number of the four threshold voltage states: the distribution of values can also be fast in different NMOS NORs Flash floating gate unit design changes. [0052] wIntel StrataFlash (Memory Technology Overview), published by Atwood, Intel Technical Journal, Volume 2, Q4 1997, wamv. intei.com, April 23, 2007, Intel Strst^F 13>sh( Memory Technology De — velopment and Implementation”, Fazio et al., Intel Technical Journal, Vol. 1, No. 2, Q4 1997, www.intel.com 'April 21, 2009, ETOX (Flash 099113777 Form No. A0101, 25th Page / Total 133 pages 0992024278-0 201104843

Memory Technology: Scaling and Integration Challenges” ,Fazio等發表於英特爾技術期刊,第6 卷第2期,2002年5月 ’ www. intel. com,20094月 21 日 ,討論了一浮動閘ET0X (快閃記憶電晶體,其結構可形 成如圖3A-3E所示的NMOS NOR快閃單元。圖3A是一雙電 晶.體浮動閘NMOS NOR快閃單元的俯視圖。圖3B是該雙電 晶體浮動閘NMOS NOR快閃單元的剖視圖。圖2C是該雙電 晶體浮動閘NMOS NOR快閃單元的示意圖。該雙電晶體浮 動閘NMOS NOR快閃單元210在p型基板240最上面的表層 中形成。一N型材料擴散到p型基板24〇的表層中,以形成 雙浮動閘電晶體205a、205t>钓汲極(D)方15a、215b和 自我對準源極(S,SAS)22Q、自我樹準源極(s)220被雙 浮動閘電晶體205a和2〇5b所共有。一第一多晶矽層在汲 極215a和215b與自我對準源極22〇之間的基極23(^和 230b上方形成浮動閘245a和245b。一第二多晶矽層在浮 動閉245a和245b上方形成雙_睛案晶體2〇5&和2〇51) 的控制間(G) 225a和225b。自我對卓源極220在雙浮動 f甲1電晶體205a和205b的兩個控制閘225a和225b中相鄰 的兩個第二多晶石夕層之間自我對準形成◊自我對準源極 220被共用於雙電晶體浮動閘2〇53及2〇51),以減少源極 線的間距。 [0053] [0054] 每一沒極215a和215b分別有一金屬觸點250a和250b。 該兩金屬觸點250a和25〇b共同連接到一金屬位元線255 〇 圖3D是該雙電晶體浮動閘NM〇s N〇R快閃單元2丨〇具有一 099113777 表單編號A0101 第26頁/共133頁 0992024278-0 單階編程電位的雙臨界電壓分配表。浮動閘245在抹除操 作之後電子電荷減少,使得雙浮動閘電晶體205a和205b 的臨界電壓降低。相反,在通道熱電子編程過程中,電 子被吸入到浮動閘245a和245b,以致雙浮動閘電晶體 205a和205b的臨界電壓增加。習慣上,抹除後的臨界電 壓(VtO)被指定為邏輯數據值“1” ,編程後的臨界電壓 (Vtl)被指定為邏輯數據值^該雙浮動閘電晶體 205a和205b儲存兩位元數據,被稱為雙位元雙電晶體 NMOS NOR快閃浮動閘單元(2b2T)。 圖3E是具有一個抹除電位和三個編程電位的雙電晶體浮 動閘NMOS NOR快閃單元210的四個臨界電壓分配表。習 知技術中,透過變化編程條件,並根據雙電晶體浮動閘 NMOS NOR快閃單元210中浮動閘245上電荷的數量,可形 成超過兩個的臨界電壓,一般被指為雙電晶體浮動閘 NMOS NOR快閃單元210的多階電位編程或者多階電位編 程單元。在本實施例中,有四個臨界電壓能在雙浮動閘 電晶體205a和205b被編程》最小的正極臨界電壓VtO是 抹除電壓,用以儲存一邏輯數據值“1Γ 。其他三個正 極編程臨界電壓為了允許被正確檢測而被充分地分開。 本實施例中,三個臨界電壓中的第一個電壓Vtl儲存一 邏輯數據值“10” 。三個臨界電壓的第二個電壓Vt2儲存 一邏輯數據值“01” 。三個臨界電壓的第三個電壓Vt3儲 存一邏輯數據值“〇〇” ^由於每一雙電晶體浮動閘NMOS NOR快閃單元210儲存有四個明顯不同的臨界電壓狀態, 每一雙電晶體浮動閘NMOS NOR快閃單元210儲存有二進 表單編號A0101 第27頁/共133頁 201104843 位數據雙位元,故被稱為雙位元單電晶體NM0S N0R快閃 單元(2b/lT)。 [0056] 雙電晶體浮動閘_〇5 Ν〇_閃單元21〇的臨界電壓vu和 Vt2的標準值在不同的設計中亦可變化。臨界電壓VtO和 Vt3的標準值能有一較寬的臨界電壓分佈。如前所述的 NMOS NAND快閃浮動閘單元’對應四個臨界電壓狀態的 二位元數據值的分佈亦可在不同的NMOS NOR快閃浮動閘 單元設計之間變化。 [0057] 圖4A是本發明一NM0S n〇R快閃記憶髏單元4〇〇的示意圖 。圖4B-1和4C-1是NMOS NOR快閃記憶體單元4〇〇的俯視 圖。圖4B-2和4C-2是NMOS NOR快閃記愧體單元4〇〇的剖 視圖。浮動閘型NMOS NOR快閃記憶體單元4〇〇在p型基板 440的最上面的表層上形成。一n型材料無散到p型基板 440的表層中以形成深n井435。然後,一p型材料擴散到 深N井435的表層中以形成p井430 ( —般被稱為三重p井) 。然後’ N型材料擴散到P井430的表層中,以形成NM〇s NAND快閃浮動閘電晶趙405a的汲極(D) 415、NMOS NAND快閃浮動閘電晶體405b的源極和自我對準源極/没 極(S/D) 420。源極/汲極420為NMOS NAND快閃浮動閘 電晶體405a的源極和NMOS NAND快閃浮動閘電晶體4〇5b 的汲極。一第一多晶矽層在P井430的NMOS NAND快閃浮 動閘電晶體405a的汲極415和源極420以及NMOS NAND快 閃浮動閘電晶體4 0 5 b的沒極4 2 0和源極4 2 2之間的基極上 方形成浮動閘445a和445b。第二多晶矽層在浮動閘445a 和445b上方形成NMOS NAND快閃浮動閘電晶體405a和 099113777 表單編號A0101 第28頁/共133頁 0992024278-0 201104843 405b的控制閘(G)425a和425b »自我對準源極/汲極420 在NMOS NAND快閃浮動閘電晶體405a和405b的兩控制 閘425a和425b所相鄰的第二多晶矽層之間自我對準形成 。自我對準源極420共同用於NMOS NAND快閃浮動閘電晶 體405a和405b,以減少源極線的間距。 [0058] Ο ο NMOS NAND快閃浮動閘電晶體405a和405b的閘極長度為 P井430裡NMOS NAND快閃浮動閘電晶體405a的汲極415 和源極4 2 0以及N Μ 0 S N A N D快閃浮動閘電晶體4 0 5 b的汲 極420和源極422之間基極的通道。NMOS NAND快閃浮動 閘電晶體405a及405b的通道宽度由汲極415、源極422和 源極/汲極420 N擴散的寬度決定。雙電晶:體:NM0S NOR快 閃記憶體單元4 0 0的一般單元尺寸在大韵丄2,1 2到大約14 λ 2之間。故’一位元的NOR單元的有效尺寸是大約6又2 。該一位元的NOR單元的有效尺寸(6 λ2)比習知技術的一 NAND單元尺寸稍大。然,該一位元的nor單元的有效尺寸 比習知用大於50nm半導體製程技術的n〇R單元尺寸(1〇入 2)小很多。由於小於5〇nm半導體製程的擴展因素,從而 使得前述的NOR單元結構尺寸預期增加到15 λ 2 ^ NMOS NOR快閃記憶體單元4〇〇的有效單位元/單電晶體尺寸仍然 保持大約6又2的有效的單元尺寸不變。不變的單元尺寸 是由於其擴展性與習知技術的NMOS NAND快閃記憶體單 元的擴展相同。 浮動閘445a和445b分別儲存電子電荷以改變NM〇s NANI) 快閃浮動閘電晶體405a和405b的臨界電壓。在所有諸如 讀取操作、編程操作和抹除操作中,p型基板44〇永遠連 099113777 表單編號A0101 第29頁/共133頁 0992024278-0 [0059] 201104843 接到接地參考電壓(GND)。深N井邮在讀取操作和編程 操作中連接到電源電壓(_),然,其在F〇wier_ N〇rdheim通道抹除操作過財連接的電壓大約為+ 20V。 在目前_S NOR快閃記憶體單元4〇〇的設計中電墨源 是1.8V或3.0V。與深N井的偏壓條件相同,三重p井43〇 在正常的讀轉作和編轉作巾連接到減參考電壓源 ,但疋在Fowler-Nordheim通道抹除操作中連接的電壓 大約為+ 20V。 [0060] [0061] [0062] 099113777 NMOS NAND快閃浮動閉電晶體4〇50〇4〇5b在其陣列中 排列成直行和橫列-第二多晶矽層即關⑽NAND快閃浮 動閘電晶體405a及405b的控制閉425a及425b延伸形成 子元線’該子元線連接到陣列裡的一..列上每一nm〇S NAND快閃浮動閘電晶體4〇5a及405b。 一隧道氧化物在NMOS N AND快閃浮動閘電晶體4〇 5a的汲 極415和源極4 2 0以及N M:0 S N A N D快蹲浮動閘電晶體 405b的没極420和源極42:¾^婿俄'通遽^區432a和432b的 上方和浮動閘445a和445b下方之間形成。隧道氧化物一 般的厚度是100 A。在Fowler-Nordheim通道編程操作 和抹除操作期間,電子電荷流經隧道氧化物。在習知的 NOR操作中,Fowler-Nordheim通道抹除操作把儲存的 電子從浮動閘445a和445b射出並穿過随道氧化物到單元 通道區432a和432b,最後進入三重P井430中。 在抹除操作後,儲存在浮動閘445a和445b的電子電荷減^ 少導致NMOS NAND快閃浮動閘電晶體405a和405b的第— 臨界電壓(VtO)降低。相比之下’在一Fowler- 表單編號A0101 第30頁/共133頁 0992024278 201104843Memory Technology: Scaling and Integration Challenges", Fazio et al., Intel Technical Journal, Vol. 6, No. 2, May 2002 'www.intel.com, April 21, 2009, discusses a floating gate ET0X (flash memory) The transistor has a structure to form an NMOS NOR flash cell as shown in Figures 3A-3E. Figure 3A is a top view of a double crystal body floating gate NMOS NOR flash cell. Figure 3B is the double transistor floating gate NMOS. A cross-sectional view of the NOR flash cell. Figure 2C is a schematic diagram of the dual transistor floating gate NMOS NOR flash cell. The dual transistor floating gate NMOS NOR flash cell 210 is formed in the uppermost surface layer of the p-type substrate 240. The type of material diffuses into the surface layer of the p-type substrate 24A to form double floating gate transistors 205a, 205t> fishing rod (D) sides 15a, 215b and self-aligned source (S, SAS) 22Q, self-aligned The source (s) 220 is shared by the double floating gate transistors 205a and 2〇5b. A first polysilicon layer is at the base 23 between the drains 215a and 215b and the self-aligned source 22〇 (^ and Floating gates 245a and 245b are formed over 230b. A second polysilicon layer is in floating closures 245a and 245b The square forms a double _ eye crystal 2〇5& and 2〇51) control rooms (G) 225a and 225b. The self-aligned source 220 in the double floating f-1 transistors 205a and 205b two control gates 225a and Self-aligned between two adjacent second polycrystalline layers in 225b, self-aligned source 220 is commonly used for dual transistor floating gates 2〇53 and 2〇51) to reduce the source line [0054] Each of the poles 215a and 215b has a metal contact 250a and 250b, respectively. The two metal contacts 250a and 25B are commonly connected to a metal bit line 255. FIG. 3D is the double power. Crystal floating gate NM〇s N〇R flash unit 2丨〇 has a 099113777 Form No. A0101 Page 26 of 133 Page 0992024278-0 Double-threshold voltage distribution table for single-order programming potential. Floating gate 245 after erase operation The electron charge is reduced, causing the threshold voltages of the dual floating gate transistors 205a and 205b to decrease. Conversely, during channel hot electron programming, electrons are drawn into the floating gates 245a and 245b such that the threshold voltages of the dual floating gate transistors 205a and 205b Increased. Conventionally, the threshold voltage (VtO) after erasing is specified as logic According to the value "1", the programmed threshold voltage (Vtl) is designated as a logical data value. The dual floating gate transistors 205a and 205b store two bits of data, which is called a two-bit dual-crystal NMOS NOR flash floating. Gate unit (2b2T). Figure 3E is a four threshold voltage distribution table for a dual transistor floating gate NMOS NOR flash cell 210 having an erase potential and three programming potentials. In the prior art, by changing the programming conditions and according to the amount of charge on the floating gate 245 in the double transistor floating gate NMOS NOR flash unit 210, more than two threshold voltages can be formed, generally referred to as a double crystal floating gate. Multi-level potential programming or multi-order potential programming unit of NMOS NOR flash unit 210. In this embodiment, there are four threshold voltages that can be programmed in the dual floating gate transistors 205a and 205b. The minimum positive threshold voltage VtO is the erase voltage for storing a logic data value of "1". The other three positive programming The threshold voltage is sufficiently separated in order to allow correct detection. In this embodiment, the first voltage Vtl of the three threshold voltages stores a logical data value "10". The second voltage Vt2 of the three threshold voltages stores one. The logical data value "01". The third voltage Vt3 of the three threshold voltages stores a logical data value "〇〇" ^ Since each dual transistor floating gate NMOS NOR flash unit 210 stores four distinct threshold voltages State, each double crystal floating gate NMOS NOR flash unit 210 stores a binary form number A0101 page 27 / 133 pages 201104843 bit data double bit, so it is called double bit single crystal NM0S N0R flash Unit (2b/lT) [0056] The standard values of the threshold voltages vu and Vt2 of the double transistor floating gate _〇5 Ν〇_flash unit 21〇 may also vary in different designs. The criteria for the threshold voltages VtO and Vt3 Value can have Wide threshold voltage distribution. The distribution of the two-bit data values corresponding to the four threshold voltage states of the NMOS NAND flash floating gate unit as described above can also vary between different NMOS NOR flash floating gate unit designs. 4A is a schematic diagram of an NM0S n〇R flash memory unit 4〇〇 of the present invention. FIGS. 4B-1 and 4C-1 are top views of an NMOS NOR flash memory unit 4〇〇. FIG. 2 and 4C-2 are cross-sectional views of the NMOS NOR flash memory cell unit 4A. The floating gate type NMOS NOR flash memory cell unit 4 is formed on the uppermost surface layer of the p-type substrate 440. An n-type material is absent. Dispersed into the surface layer of the p-type substrate 440 to form a deep n-well 435. Then, a p-type material diffuses into the surface layer of the deep N-well 435 to form a p-well 430 (generally referred to as a triple-p well). Then 'N The type of material diffuses into the surface layer of the P well 430 to form the drain (D) 415 of the NM〇s NAND flash floating gate transistor 405a, the source and self-aligned source of the NMOS NAND flash floating gate transistor 405b. Pole/dipole (S/D) 420. Source/drain 420 is the source of NMOS NAND flash floating gate transistor 405a and NMOS NAND flash floating The drain of the gate transistor 4〇5b. The first polysilicon layer is at the drain 415 and the source 420 of the NMOS NAND flash floating gate transistor 405a of the P well 430 and the NMOS NAND flash floating gate transistor 40 Floating gates 445a and 445b are formed above the base between the 5b 0 and the source 4 2 2 of 5 b. The second polysilicon layer forms NMOS NAND flash floating gate transistors 405a and 099113777 over floating gates 445a and 445b. Form No. A0101 Page 28 of 133 pages 0992024278-0 201104843 405b Control Gates (G) 425a and 425b » The self-aligned source/drain 420 is self-aligned between the second polysilicon layers adjacent to the two control gates 425a and 425b of the NMOS NAND flash floating gate transistors 405a and 405b. Self-aligned source 420 is commonly used for NMOS NAND flash floating gate transistors 405a and 405b to reduce the pitch of the source lines. [0058] The gate length of the NMOS NAND flash floating gate transistors 405a and 405b is the drain 415 and the source 4 2 0 and the N Μ 0 SNAND of the NMOS NAND flash floating gate transistor 405a in the P well 430. The channel of the base between the drain 420 and the source 422 of the floating floating gate transistor 4 0 5 b. The channel width of the NMOS NAND flash floating gate transistors 405a and 405b is determined by the width of the drain 415, the source 422, and the source/drain 420 N diffusion. Double crystal: Body: NM0S NOR Fast flash memory unit 400 The general unit size is between 2, 12 and approximately 14 λ 2 . Therefore, the effective size of a one-dimensional NOR unit is about 6 and 2. The effective size (6 λ2) of the one-bit NOR cell is slightly larger than the size of a NAND cell of the prior art. However, the effective size of the one-dimensional nor cell is much smaller than the conventional n〇R cell size (1 into 2) of semiconductor process technology greater than 50 nm. Due to the expansion factor of the semiconductor process less than 5〇nm, the structure size of the aforementioned NOR cell is expected to increase to 15 λ 2 ^ The effective unit cell/single transistor size of the NMOS NOR flash memory cell unit 4 remains approximately 6 The effective unit size of 2 is unchanged. The constant cell size is due to its scalability being the same as that of the conventional NMOS NAND flash memory cell. Floating gates 445a and 445b store electron charges to change the threshold voltage of NM〇s NANI) flash floating gate transistors 405a and 405b, respectively. In all such operations as read, program, and erase, the p-type substrate 44 is always connected. 099113777 Form No. A0101 Page 29 of 133 0992024278-0 [0059] 201104843 Connected to ground reference (GND). Deep N is connected to the supply voltage (_) during read operations and programming operations. However, its voltage at the F〇wier_ N〇rdheim channel erase operation is approximately +20V. In the current design of the _S NOR flash memory cell unit 4, the ink source is 1.8V or 3.0V. The same as the bias condition of the deep N well, the triple p well 43 is connected to the reduced reference voltage source during the normal read transition and the braided wipe, but the voltage connected in the Fowler-Nordheim channel erase operation is approximately + 20V. [0062] 099113777 NMOS NAND flash floating closed crystals 4〇50〇4〇5b are arranged in a straight row and a row in their array - a second polysilicon layer is off (10) NAND flash floating gate The control closures 425a and 425b of the crystals 405a and 405b extend to form a sub-element line that is connected to each of the nm 〇S NAND flash floating gate transistors 4〇5a and 405b on a column in the array. A tunnel oxide is in the NMOS N AND flash floating gate transistor 4〇5a, the drain 415 and the source 4 2 0 and the NM:0 SNAND fast floating gate transistor 405b, the pole 420 and the source 42: 3⁄4^ It is formed between the upper side of the 'Using' areas 432a and 432b and the lower side of the floating gates 445a and 445b. The tunnel oxide is typically 100 A thick. During the Fowler-Nordheim channel programming operation and erase operation, electron charges flow through the tunnel oxide. In conventional NOR operation, the Fowler-Nordheim channel erase operation ejects stored electrons from floating gates 445a and 445b and through the associated oxide to cell channel regions 432a and 432b, and finally into triple P well 430. After the erase operation, the reduced electron charge stored in the floating gates 445a and 445b causes the first threshold voltage (VtO) of the NMOS NAND flash floating gate transistors 405a and 405b to decrease. In contrast, in a Fowler-form number A0101, page 30 / 133 pages 0992024278 201104843

Nordheim通道編程操作中,電子被吸入到浮動閘.445a和 445b,以致NMOS NAND快閃浮動閘電晶體405a和405b第 二臨界電壓水平(Vtl)被設置到相對高的電壓。 [0063] ΟDuring the Nordheim channel programming operation, electrons are drawn into the floating gates .445a and 445b such that the second threshold voltage level (Vtl) of the NMOS NAND flash floating gate transistors 405a and 405b is set to a relatively high voltage. [0063] Ο

圖5A-5E為本發明雙電晶體浮動閘NMOS NOR快閃單元串 接所形成的陣列的其中一部分的線路連接的俯視圖。該 部份包含四橫行的雙電晶體NMOS NOR快閃記憶體單元 400和十二直列的雙電晶體NMOS NOR快閃記憶體單元400 ,或者八橫行的NMOS NAND快閃浮動閘電晶體405a和 405b。每一NMOS NOR快閃記億體單元400有如圖4A、 4B-1、4B-2、4C-1和4C-2所示的N+擴散没極415、源極 /汲極420和源極422。控制閉425a和4251)連接在字元線 WL0 450a和WL1 450b。位元線455a和455b以及源極線 460a和460b被形成做為圖4B-2和4C-2的第一層金屬 (455a和460b)或者第二層金屬(455b和460a)。該位元 線455a及455b分別透過過孔457a和457b與NMOS NAND ·" " I iji _ .- .,.· -· ψ-1 ^ ·" d ....... -.5 y :... 快閃浮動閘電晶體405a的汲極415相連。該源極線460a ϋ; : 1 ·;ί : Ψ 々· ·:: ο [0064] 及460b分別透過過孔462a及462b與NMOS NAND快閃浮動 閘電晶體405b的源極422相連。 在圖5B中,本地的Metal 1位元線到本地的Metal 2位 元線的連接和本地的Metal 1源極線到本地的Metal 2源 極線的連接是透過過孔(Vial)實現的。圖5C中,下一層 的本地的Metal 2位元線到本地的Metal 3位元線之間的 連接和本地的Metal 2源極線到本地的Metal 3源極線的 連接是透過過孔(VIA2)實現的。圖5D中,再下一層的本 地的Metal 3位元線到本地的Metal 4位元線的連接和本 099113777 表單編號A0101 第31頁/共133頁 0992024278-0 201104843 地的Metal 3源極線連接到本地的Metal 4源極線的連接 是透過過過孔(VIA3)實現的;圖5E中,又下一層的本地 的Metal 4位元線到本地的Metal 5位元線的連接和本地 的Metal 4源極線到本地的“1^1 5源極線的連接是透過 過孔(VIA4)實現的。十二條本地的位元線455a、455b和 十二條本地的源極線460a、46〇i)的NMOS NOR快閃記 憶體單元400的矩陣僅僅使用五層金屬與大約6 λ 2的一有 效的單元尺寸即可成功地連接在一起。每一全域的位元 線和每一全域的源極線分別地被兩條本地的位元線455a 和455b和本地的源極線460a和460b所共用。 [0065] 在圖5A-5E中所描述的結構中,有五層金靥線產生一單元 構造以使得一單位元電晶體》〇R單元的有效尺寸大約是6 λ2。金屬線之間的間距可在水平或者X軸方向較大,或 者NAND串可包括三個或更多的浮動閘電晶體,以使金屬 層減少至五層以下。金屬層數與NAND串數和在水平或X軸 方向的金屬線間距之間有1折衷的方索。NAND串數愈多 並且在X轴方向愈鬆散則金:屬層..愈少 [0066] 圖6A-6D是本發明雙電晶體浮動閘NMOS NAND快閃單元的 單電晶體的各實施例的臨界電壓圖◊圖6A是圖4A、 4B-1、4B-2、4C-1 和4C-2 中的NMOS NAND快閃浮動閘 電晶體405a和40 5b執行編程操作和抹除操作的一實施例 的臨界電壓水平示意圖。在該實施例中,一正極編程臨 界電壓(Vtl)具有一狹窄的分佈,其代表邏輯數據“〇,, ,一負極編程臨界電壓(VtO)亦具有一狹窄的分佈,其代 表邏輯數據1 。VtO和Vtl在編程狀態有較優越的窄 099113777 表單編號A0101 第32頁/共133頁 0992024278-0 201104843 分佈臨界電壓。在NMOS NAND快閃浮動閘電晶體405a和 405b的抹除操作中,•一+ 2〇v的電壓施加於NMOS NAND快 閃浮動閘電晶體4〇5a和405b所在的三重P井430,並且接 地參考電壓(()V)被施加於被選擇的NMOS NAND快閃浮動 閘電晶體的被選擇的控制閘425a和425b上,以在被選擇 的NMOS NAND快閃·浮動閘電晶體405a和405b的被選擇控 制閘425a和425b和通道區432a和432b之間形成20 V的電 壓差’以產生負極的Fowler-Nordheim通道穿隧效應。 由於NOR快閃記憶體陣列抹除操作習慣上是在n〇r快閃記 Ο [0067] 憶體陣列區塊裡以癸·Β轉單元執行.,故負極的臨界電壓 (V10 ) —般被認為是集體抹〗_翁. 々 - . ·ί _ 響; .-1¾¾ ^雪-:’广-' 在習知技術中,NAND快閃記憶體陣列的臨壓(vtO) :- >-i-= - <·ϊ 具有一寬電壓分佈。習慣上’負極臨界電壓(Vto)有大約 ❹ [0068] 2.0V的一個電壓範圍,即在-2.0V到大約0.0V之間變化 。臨界電壓(Vtl)的編程電壓大約是+ 2.5V,其在+ 2,0 V到+ 3. 0V之間變化。正極鸥界電塵(vtl )在電路操作中 不需要狹窄的0. 5V分#’只要在頁編程操作期間小於被 選擇的NAND快閃記憶體陣列區塊中的未被選擇的字元線 的6. 0V的通過電壓即可。 不同於一頁512位元的NAND快閃記憶體陣列的同步緩慢的 20us的線性讀取速度規範’ NOR快閃記憶體元件的迅速 、隨機以及非同步的讀取速度少於l〇〇ns。鑒於上述對 NMOS NOR快閃記憶體單元400的雙位元/雙電晶體的速度 須求,將NMOS NAND快閃浮動閘電晶體4〇5a和405b串接 時,負極臨界電壓(VtO)和正極臨界電壓(vti)的最理想 099113777 表單編號A0101 第33頁/共133頁 0992024278-0 201104843 的臨界t壓分佈在大約0.5V之内。負極的臨界電壓 (Vt〇)有大約-0.5V的標準電壓,正極的臨界電壓(Vtl) #大約+3. 〇V的標準電壓。為了使負極的臨界電壓(VtO) # '麵的臨界電壓(Vtl)有一狹窄的臨界電壓分佈,負極 的臨界電壓(VtO)和正極的臨界電壓(Vtl)可透過使用一 位元—位元的正極Fowler-Nordheim通道編程操作來達 到。NMOS NAND快閃浮動閘電晶體405a和405b的負極的 臨界電壓(VtO)狀態可透過兩個步驟來達到。第一個步驟 是在一頁或一區塊中用一敉寬的負極臨界電壓(VtO)分佈 執行負極的Fowler-Nordheim通道集體抹除,第二個步 .: 驟是用正極的一位元一位元Fowler~Nor(iheim通道編程 而獲得一狹窄的負極臨界電塵(Vt0)"根據積體電路製程 ,被選擇的NMOS NAND快問浮動閘電晶體405a和405b的 -正極的臨界電壓(Vtl)可透過單一步驟而變窄,即從大約 + 15. 0V到大約+20V逐漸增大被選擇的控制閘425a和 425b的編程電壓。對於NMOS NAND快閃浮知閘電晶體 405a和405b,負極的臨界電壓(VtO)和正極的臨界電壓 (Vtl)都有一分佈在大約〇. 5V的狹窄的編程狀態。5A-5E are top plan views of line connections of a portion of an array formed by a series connection of a dual transistor floating gate NMOS NOR flash cell. The portion includes four horizontal double crystal NMOS NOR flash memory cells 400 and twelve inline dual transistor NMOS NOR flash memory cells 400, or eight horizontal NMOS NAND flash floating gate transistors 405a and 405b . Each of the NMOS NOR flash memory cells 400 has an N+ diffusion dipole 415, a source/drain 420, and a source 422 as shown in FIGS. 4A, 4B-1, 4B-2, 4C-1, and 4C-2. Control blocks 425a and 4251) are connected to word lines WL0 450a and WL1 450b. The bit lines 455a and 455b and the source lines 460a and 460b are formed as the first layer metal (455a and 460b) or the second layer metal (455b and 460a) of Figs. 4B-2 and 4C-2. The bit lines 455a and 455b pass through the via holes 457a and 457b and the NMOS NAND ·"" I iji _ .- ., . . . . . . . . . . . . . . . . . . 5 y :... The drain 415 of the flash floating gate transistor 405a is connected. The source line 460a ϋ; : 1 · ; ί : Ψ · · 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 In Figure 5B, the connection of the local Metal 1 bit line to the local Metal 2 bit line and the connection of the local Metal 1 source line to the local Metal 2 source line are via vias (Vial). In Figure 5C, the connection between the local Metal 2 bit line of the next layer to the local Metal 3 bit line and the connection of the local Metal 2 source line to the local Metal 3 source line is through the via (VIA2). ) achieved. In Figure 5D, the connection of the local Metal 3 bit line of the next layer to the local Metal 4 bit line and the Metal 3 source line connection of this 099113777 Form No. A0101 Page 31 / 133 page 0992024278-0 201104843 The connection to the local Metal 4 source line is via the via (VIA3); in Figure 5E, the next layer of the local Metal 4 bit line is connected to the local Metal 5 bit line and the local Metal 4 source line to the local "1 ^ 1 5 source line connection through the via (VIA4). Twelve local bit lines 455a, 455b and twelve local source lines 460a, 46 The matrix of the NMOS NOR flash memory cell 400 of 〇i) can be successfully connected using only five layers of metal with an effective cell size of approximately 6 λ 2. Each global bit line and each global domain The source lines are shared by the two local bit lines 455a and 455b and the local source lines 460a and 460b, respectively. [0065] In the structure depicted in Figures 5A-5E, there are five layers of gold ridge lines generated. A unit is constructed such that the effective size of a unit cell transistor 〇R unit is approximately 6 λ2. The spacing between the lines may be larger in the horizontal or X-axis direction, or the NAND string may include three or more floating gate transistors to reduce the metal layer to less than five layers. The number of metal layers and the number of NAND strings There is a trade-off between the metal line spacing in the horizontal or X-axis direction. The more NAND string numbers and the looser in the X-axis direction, the more the gold: genus layer.. less [0066] Figures 6A-6D are double pairs of the present invention The threshold voltage diagram of various embodiments of a single transistor of a transistor floating gate NMOS NAND flash cell is shown in FIG. 6A as an NMOS NAND flash floating in FIGS. 4A, 4B-1, 4B-2, 4C-1, and 4C-2. The gate transistors 405a and 40b perform a threshold voltage level diagram of an embodiment of a program operation and an erase operation. In this embodiment, a positive programming threshold voltage (Vtl) has a narrow distribution representing logical data "〇 , , , A negative programming threshold voltage (VtO) also has a narrow distribution, which represents logical data 1 . VtO and Vtl are superior in programming state. 099113777 Form No. A0101 Page 32 of 133 Page 0992024278-0 201104843 Distribution threshold voltage. In the erase operation of the NMOS NAND flash floating gate transistors 405a and 405b, a voltage of +2〇v is applied to the triple P well 430 where the NMOS NAND flash floating gate transistors 4〇5a and 405b are located, and grounded. A reference voltage (()V) is applied to selected control gates 425a and 425b of the selected NMOS NAND flash floating gate transistor for selection of NMOS NAND flash/floating gate transistors 405a and 405b A voltage difference of 20 V is formed between the control gates 425a and 425b and the channel regions 432a and 432b to produce a Fowler-Nordheim channel tunneling effect of the negative electrode. Since the NOR flash memory array erase operation is customarily performed in the n〇r flash memory [0067] in the memory array block, the threshold voltage (V10) of the negative electrode is generally considered to be considered. It is a collective wipe〗 _ Weng. 々- . · ί _ 响; . -13⁄43⁄4 ^ Snow -: '广-' In the conventional technology, the NAND flash memory array of the pressure (vtO) :- >-i -= - <·ϊ has a wide voltage distribution. Conventionally, the negative voltage threshold (Vto) has a voltage range of approximately V [0068] of 2.0V, i.e., varies from -2.0V to about 0.0V. The programming voltage for the threshold voltage (Vtl) is approximately +2.5V, which varies between +2,0 V and +3.00V. The positive gull dust (vtl) does not require a narrow 0. 5V minutes #' as long as it is less than the unselected word line in the selected NAND flash memory array block during the page programming operation. 6. 0V pass voltage can be. A slower 20us linear read speed specification than a one-page 512-bit NAND flash memory array. The fast, random, and asynchronous read speeds of NOR flash memory components are less than 10 ns. In view of the above-mentioned speed requirement of the double/dual transistor of the NMOS NOR flash memory cell 400, when the NMOS NAND flash floating gate transistors 4〇5a and 405b are connected in series, the negative threshold voltage (VtO) and the positive electrode are connected. The ideal threshold voltage (vti) 099113777 Form No. A0101 Page 33 / 133 pages 0992024278-0 201104843 The critical t-pressure distribution is within about 0.5V. The threshold voltage (Vt〇) of the negative electrode has a standard voltage of about -0.5 V, and the threshold voltage (Vtl) of the positive electrode is about +3. The standard voltage of 〇V. In order to make the threshold voltage (VtO) of the negative electrode have a narrow threshold voltage distribution, the threshold voltage (VtO) of the negative electrode and the threshold voltage (Vtl) of the positive electrode can be transmitted through one-bit-bit. The positive Fowler-Nordheim channel programming operation is achieved. The threshold voltage (VtO) state of the negative terminals of the NMOS NAND flash floating gate transistors 405a and 405b can be achieved in two steps. The first step is to perform a collective erase of the Fowler-Nordheim channel of the negative electrode with a wide negative voltage threshold (VtO) distribution in one page or one block. The second step is to use one bit of the positive electrode. One element Fowler~Nor (iheim channel programming to obtain a narrow negative critical dust (Vt0)" according to the integrated circuit process, the selected NMOS NAND asks the threshold voltage of the positive gate of the floating gate transistors 405a and 405b (Vtl) can be narrowed by a single step of gradually increasing the programming voltages of the selected control gates 425a and 425b from about + 15.0V to about +20V. For NMOS NAND flash floating gate transistors 405a and 405b The threshold voltage (VtO) of the negative electrode and the threshold voltage (Vtl) of the positive electrode all have a narrow programming state distributed at about V5V.

[0069]圖 6B是圖4A、4B-1、4B-2、4C-1 和4C-2 中NMOS NAND 快閃浮動閘電晶體405a和405b的編程操作和抹除操作的 第二種實施例的臨界電壓圖。在該單電位單元(SLC)實 施例中,第一臨界電壓(vto)和第二臨界電壓(Vtl)中全 都是正極的’且具有大約0.5V的臨界電壓分佈。正極的 第一臨界電壓(VtO)亦透過兩步驟完成:第一步驟是執行 負極的Fowler-Nordheim通道集體抹除操作,第二步驟 099113777 表單編號A0101 第34頁/共133頁 0992024278-0 201104843 是執行如圖6a所述的正極的Fowler_Nordheim通道一位 元一位元編程操作。第一臨界電壓(VtO)和第二臨界電壓 (V11)皆是編程狀態而不是一抹除狀態和一編程狀態。 [〇〇7〇]第一臨界電壓(VtO)被設置為正極,其具有一〇.5V的標 準值和狹窄的0.5V分佈,即從大約+ 0.75V到大約 • +1.25V,用以儲存一邏輯數據“1” 。第二臨界電壓 (Vtl)為正極,其具有一3· 0V的標準值和從大約+ 2. 75V 到大約+ 3. 25V的狹窄的分佈,用以儲存一邏輯數據‘‘〇,, _ 。在一些實施例中,NOR快閃記憶體根據有些應用中的 〇 一 速度考量,需要具有一從+ 2· 5V到+3. 5V較寬闊的臨界電 壓分佈。6B is a second embodiment of the programming and erase operations of the NMOS NAND flash floating gate transistors 405a and 405b of FIGS. 4A, 4B-1, 4B-2, 4C-1, and 4C-2. Critical voltage map. In the single potential cell (SLC) embodiment, the first threshold voltage (vto) and the second threshold voltage (Vtl) are all positive and have a critical voltage distribution of about 0.5V. The first threshold voltage (VtO) of the positive electrode is also completed in two steps: the first step is to perform a collective erase operation of the Fowler-Nordheim channel of the negative electrode, the second step 099113777 Form No. A0101 Page 34 / 133 pages 0992024278-0 201104843 Yes A one-dimensional one-bit programming operation of the Fowler_Nordheim channel of the positive electrode as described in Figure 6a is performed. The first threshold voltage (VtO) and the second threshold voltage (V11) are both programmed states rather than an erased state and a programmed state. [〇〇7〇] The first threshold voltage (VtO) is set to the positive electrode, which has a standard value of 〇5V and a narrow 0.5V distribution, that is, from about +0.75V to about +1.25V for storage. A logical data "1". The second threshold voltage (Vtl) is a positive electrode having a standard value of 3.0 V and a narrow distribution from about + 2.75 V to about + 3.25 V for storing a logical data '', _, _. In some embodiments, the NOR flash memory needs to have a wider critical voltage distribution from +2.5V to +3.5V depending on the speed considerations in some applications.

[0071]圖 6C是圖 4A、4B-1、4B-2、4C-1 和4C-2 中 NMOS NAND 快閃浮動閘電晶體405a和405b的編程操作和抹除操作的 另一種實施方案的臨界電壓圖。該實施攝是關於一多電 位單元(MLC),其中所有的四個臨界電壓參平 r , , ' ·, ...... .... 、Vt2和Vt3)不論是正極或者負值皆有大約〇 5V的狹窄 〇 分佈。在該實施轉中,第一臨:界電壓(ντο)是負極並且亦6C is a critical diagram of another embodiment of the programming and erase operations of the NMOS NAND flash floating gate transistors 405a and 405b of FIGS. 4A, 4B-1, 4B-2, 4C-1, and 4C-2. Voltage map. This implementation is about a multi-potential cell (MLC) in which all four threshold voltages are r, r, ', , ....., Vt2, and Vt3), whether positive or negative. There is a narrow 〇 distribution of about 〇5V. In this implementation, the first front-end voltage (ντο) is the negative electrode and

是透過如前所述使用兩個步驟的寫的方法進入編程狀態 ,意味著第一臨界電壓水平(ντο)有大約0.5V的標準值 和從大約-0. 25V到大約-〇. 75V之間變化的分佈,用於 儲存一邏輯數據“1Γ 。第二臨界電壓(VT1)是儲存在 NMOS NAND快閃浮動閘電晶體405a和405b的第二種數據 狀態,其具有大約+ 1. Ον的標準值。第二臨界電壓(ντί) 的分佈在大約+ 0. 75V到大約+ 1. 25V之間變化,並用於儲 存一邏輯數據“10” 。第三臨界電壓(Vt2)是NM0S 099113777 表單編號A0101 第35頁/共133頁 0992024278-0 201104843 NAND快閃浮動閘電晶體405a和405b的第三種數據狀態, 具有大約+ 2.0V的標準值。第三臨界電壓(Vt2)的分佈 從大約+ 1. 75V到大約+ 2. 25V之間變化,並用於儲存一邏 輯數據“01” 。第四臨界電壓(Vt3)是NMOS NAND快閃 浮動閘電晶體405a和405b的第四種數據狀態,並且有大 約+ 3. 0V的標準值。第四臨界電壓水平(vt3)的分佈在大 約+2. 75V到大约+ 3. 25V之間變化,並用於儲存邏輯數據 “00、 [0072] [0073] 圖6D是圖4A、圖4B-1、圖4B-2、圖4C-1和圖4C-2中 NMOS快閃浮動閘電晶體4〇5a和405b執行編程操作和抹除 操作的另一個實施例的臨界電壓圖。第一臨养電壓^·^) 、第二臨界電壓(VT1)、第三臨界電壓(Vt2)和第四臨界 電壓(Vt3)都是正極的並且臨界電壓分佈都相對比較狹窄 。在該實施例中’第一臨界電壓(VT0>有大約+ 1.0V的標 準值,並用於儲存邏輯數據“u” ^第一臨界電壓(ντ〇) 的電壓分佈在+ 〇. 7 5 V到+1. 2 5 V之間變化。第二臨界電壓 . . ...... ....... ... (VT1)有大約+2. 0V的楳準值,並用於餘存一邏輯數據“ ίο” 。第二臨界電壓(VT1)在大約+1.75V到大約+2.25V 之間變化。第三臨界電壓(Vt2)有大約+ 3. 〇v的標準值, 並用於儲存—邏輯數據“01” 。第三臨界電壓(Vt2)的分 佈在大約+ 2. 75V到大約+ 3. 25V之間變化·。第四臨界電壓 (Vt3)有大約3. 〇v的標準值,並用於儲存一邏輯數據“ 〇〇” 。第四臨界電壓(vt3)的分佈在大約+3 75V到大約 + 4. 25V之間變化。 圖7A-7D是本發明雙電晶體浮動閘NM〇s N0R快閃單元的 099113777 表單編號A0101 第36頁/共133頁 0992024278-0 201104843 其他實施例的臨界電壓圖。圖6A_6D摇述了隨、圖 .1、圖4B-2、圖4(>1和圖4(:_2中_ nand快閃浮 動開電晶體405a和405b執行編程操作和抹除操作的常規 的臨界電壓圖。圖7A-7D中描述與圖6A_6D中相反的抹除 和編程臨界電壓圖。在圖7A中,第一臨界電壓(ντ〇) 第二臨界·(νΤ1)分職表賴數據“『和邏輯數據 1 ,並分別具有大約—〇. 5V和大約+ 3· 0V的標準值。 同樣地,在圖财’第-臨界電壓⑽)代表邏輯數據 Ο 第二臨界電壓(VT1)代表邏輯數據“丨”,並分 別具有大約H.0V和大約+ 3庸的標準值。在圖7c中刀 第—臨界電壓(ντο)有大約—標準值,用於儲存 =輯數據“ 00” ;第二臨界電贿T1)具有大約+ 1.二 標準值’並用於儲存邏輯數據“ 10” ;第三臨界電壓 Ο (八2)有大約+ 2. 0V的標準值,並用於儲存邏輯數據“〇1 ,第四臨界電壓(Vt3)有大約+3. Of的標準值,並用於 儲存-邏輯數據“GO”。在咖中,第—臨界電厘; (VT0)有大約+ 1.0V的標準值,並用於儲存邏輯數據“ °° ;第二臨界電壓(VT1)有大約+2.0V的標準值,並 於儲存邏輯數據“ίο” ;第三臨界電壓(vt2)有大約 + 3.0V的標準值,並用於儲存邏輯數值“〇1” ;第四 電壓⑽)有大約+4. 〇V的標準值,並用於儲存邏輯數據 _彡階電位單元的最高的臨界電壓即第四臨界電壓⑺ 狀態或者單階電位單元的第二臨界電壓(V⑴被指定作為 抹除操作的狀態。單階電位單元的第一臨界電壓(vt〇)和 099113777 表單编號A0101 第37頁/共133頁 0992024278-0 201104843 多階電位單元的第一臨界電壓(vto)、第二臨界電壓 (Vtl)、第三臨界電壓(vt2)是編程狀態。抹除臨界電壓 (多階電位單元的臨界電壓Vt3或者單階電位單元的臨界 電壓Vtl)是透過N0R快閃記憶體元件上的一頁的正極 Fowler-Nordheim通道穿隧效應而獲得,該nor快閃記 憶體元件施加大約+ 20.0V的電壓於圖4A、圖4B-1、圖 4B-2、圖4C-1和圖4C-2中的被選擇的NMOS NAND快閃浮 動閘電晶體405a和405b的被選擇的控制閘425a和425b ,以及施加一接地參考電壓(〇. 0V)在被選擇的基極。應 該注意的是,圖7C、?D#的多階電隹單元的第四臨界電 壓(Vt3)和圖7A、7B的單階r電位單元的第二臨界電壓 (VT1)的抹除狀態設置到可實'現p0wpeΓ-Nordheim通道 集體效應的一電壓值。由於只要檢驗抹除狀態臨界電壓 是否通過最小可接受的抹除狀態臨界電壓,而最大抹除 狀態電壓不須注意’不需要檢驗,故,臨界電壓的分佈 變化更大。 [0075] 在抹除操作之後’透過一位元一位元F〇wler_N〇rdheim 邊界編程過程來編程那些待編程的單元到其它邏輯數據 狀態’在該過程中,施加一大約-1〇.〇V的負極的電壓於 NOR快閃記憶體元件中一頁的被選擇字元線和施加大約 + 5 V到大約+1 〇 v的電壓於被選擇的nm〇S NAND快閃浮動 閘電晶體405a和405b的汲極。然後斷開被選擇的NM0S NAND快閃浮動閘電晶體405a和405b的源極到浮動狀態。 如前所述,NMOS NAND快閃浮動閘電晶體4〇5a和405b的 編程操作具有兩個步驟,其中第一個步驟是用正極的 099113777 表單編號A0101 第38頁/共133頁 0992024278-0 201104843The programming state is entered by using a two-step writing method as described above, which means that the first threshold voltage level (ντο) has a standard value of about 0.5 V and is from about -0.25 V to about - 〇. 75 V. The varying distribution is used to store a logical data "1". The second threshold voltage (VT1) is the second data state stored in the NMOS NAND flash floating gate transistors 405a and 405b, which has a standard of approximately + 1. Ον The value of the second threshold voltage (ντί) varies from approximately +0.75V to approximately +1.50V and is used to store a logical data "10". The third threshold voltage (Vt2) is NM0S 099113777 Form No. A0101 Page 35 of 133 pages 0992024278-0 201104843 The third data state of the NAND flash floating gate transistors 405a and 405b has a standard value of approximately +2.0 V. The distribution of the third threshold voltage (Vt2) is approximately + 1 Between 75V and about + 2.25V, and for storing a logical data "01". The fourth threshold voltage (Vt3) is the fourth data state of the NMOS NAND flash floating gate transistors 405a and 405b, and A standard value of approximately + 3. 0V. Fourth Pro The distribution of the boundary voltage level (vt3) varies between approximately +2.75V to approximately + 3.25V and is used to store logical data "00, [0072] Figure 6D is Figure 4A, Figure 4B-1, Figure 4B-2, FIG. 4C-1 and FIG. 4C-2 NMOS flash floating gate transistors 4〇5a and 405b perform a threshold voltage diagram of another embodiment of a program operation and an erase operation. The first temporary voltage ^·^), the second threshold voltage (VT1), the third threshold voltage (Vt2), and the fourth threshold voltage (Vt3) are both positive and the threshold voltage distribution is relatively narrow. In this embodiment, the first threshold voltage (VT0> has a standard value of about +1.0 V and is used to store the logical data "u" ^the first threshold voltage (ντ〇) is distributed at + 〇. 7 5 V to Between +1 and 2 5 V. The second threshold voltage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A logical data "ίο". The second threshold voltage (VT1) varies between approximately +1.75V and approximately +2.25V. The third threshold voltage (Vt2) has a standard value of approximately + 3. 〇v and is used for storage - Logic data "01". The distribution of the third threshold voltage (Vt2) varies from approximately + 2.75V to approximately + 3.25V. The fourth threshold voltage (Vt3) has a standard value of approximately 3. 〇v, and is used The logic data "〇〇" is stored. The distribution of the fourth threshold voltage (vt3) varies from about +3 75 V to about + 4.25 V. Figures 7A-7D show the dual transistor floating gate NM〇s N0R of the present invention. 099113777 of flash unit No. A0101 Page 36 of 133 page 0992024278-0 201104843 Threshold voltage diagram of other embodiments. Fig. 6A_6D is abbreviated with Fig. 1, Fig. 4B-2, Fig. 4 (> Figure 4 (:_2 _ nand flash floating floating crystals 405a and 405b perform a conventional threshold voltage diagram of a program operation and an erase operation. The erase and program threshold voltage diagrams opposite to those in FIGS. 6A-6D are described in FIGS. 7A-7D. In Fig. 7A, the first threshold voltage (ντ〇) second critical value (νΤ1) is divided into data "" and logical data 1 and has a standard value of approximately - 5 V and approximately + 3 · 0 V, respectively. In the graph, the 'threshold voltage (10) represents the logical data Ο the second threshold voltage (VT1) represents the logical data "丨" and has a standard value of about H.0V and about +3, respectively. In Fig. 7c The knife-threshold voltage (ντο) has an approximate - standard value for storing = series data "00"; the second critical electric bribe T1) has approximately + 1. 2 standard values and is used to store logical data "10"; The three threshold voltage Ο (eight 2) has a standard value of approximately + 2. 0V and is used to store the logical data "〇1, the fourth threshold voltage (Vt3) has a standard value of approximately +3. Of, and is used for storage - logical data "GO". In the coffee, the first critical value; (VT0) has a standard value of about + 1.0V, For storing logical data "°°; the second threshold voltage (VT1) has a standard value of about +2.0V, and stores the logical data "ίο"; the third threshold voltage (vt2) has a standard value of about +3.0V, And used to store the logical value "〇1"; the fourth voltage (10)) has a standard value of about +4. 〇V, and is used to store the logical data _ the highest threshold voltage of the 电位 step potential unit, that is, the fourth threshold voltage (7) state or single The second threshold voltage of the step potential unit (V(1) is designated as the state of the erase operation. First threshold voltage (vt〇) and 099113777 of single-stage potential unit Form No. A0101 Page 37 of 133 Page 0992024278-0 201104843 First threshold voltage (vto) and second threshold voltage (Vtl) of multi-level potential unit The third threshold voltage (vt2) is a programmed state. Erasing the threshold voltage (the threshold voltage Vt3 of the multi-step potential unit or the threshold voltage Vtl of the single-order potential unit) is obtained by tunneling the positive Fowler-Nordheim channel of a page on the NOR flash memory device, which is fast The flash memory device applies a voltage of approximately +20.0V to the selected NMOS NAND flash floating gate transistors 405a and 405b of FIGS. 4A, 4B-1, 4B-2, 4C-1, and 4C-2. The selected control gates 425a and 425b are applied as well as a ground reference voltage (〇. 0V) at the selected base. It should be noted that Figure 7C,? The fourth threshold voltage (Vt3) of the multi-level power unit of D# and the second threshold voltage (VT1) of the single-stage r-potential unit of FIGS. 7A and 7B are set to the real-time p0wpeΓ-Nordheim channel collective effect a voltage value. Since the threshold voltage of the erase state is checked by the minimum acceptable erase state threshold voltage, and the maximum erase state voltage does not need to be noted, the distribution of the threshold voltage is more varied. [0075] After the erase operation, 'the one-to-one bit F〇wler_N〇rdheim boundary programming process is used to program the cells to be programmed to other logical data states'. In the process, an approx. -1 〇 is applied. The voltage of the negative terminal of V is at a selected word line of one page in the NOR flash memory device and a voltage of about + 5 V to about +1 〇v is applied to the selected nm 〇 S NAND flash floating gate transistor 405a. And the bungee of 405b. The sources of the selected NM0S NAND flash floating gate transistors 405a and 405b are then turned off to a floating state. As mentioned earlier, the programming operation of the NMOS NAND flash floating gate transistors 4〇5a and 405b has two steps, the first of which is to use the positive 099113777 Form No. A0101 Page 38 of 133 Page 0992024278-0 201104843

Fowler-Nordheim通道操作來抹除被選擇的NOR快閃記 憶體元件的區塊,第二個步驟是用一位元一位元Fowler-Nordheim邊界隧道編程操作把最大臨界電壓修整成為期 望的電壓。 [0076] Ο Ο [0077] 圖8是一包含本發明雙電晶體浮動閘NM0S n〇R快閃單元 510的各實施例的NOR快閃記憶體元件500的示意圖。NOR 快閃記憶體元件500包括一排列成直行和橫列的雙電晶 體浮動閘NMOS NOR快閃單元510的陣列505。每一雙電晶 體浮動閘NMOS NOR快閃單元510包括兩個NMOS NAND快 閃浮動閘電晶體515a和515b。兩個NMOS NAND快閃浮動 閘電晶體5153和5151)的構^»操作錢_%4儿、圖46-1、 圖4C-2、圖4C-1和圖4C-2啊的關閃浮動閘電 晶體405a和405b。NMOS NAND快閃浮動閘電晶體515a的 汲極連接到本地的位元線520a、520b、…、520n-l和 520η之一。NMOS NAND浮動閘電晶體515b的源極被連接 到源極線530a、53Q|、…、530n-l 和530η之一。NMOS N A N D快閃浮動閘電晶體515 a的源極連接到N Μ 0 S Ν 0 R快 * j ψ * t (> 閃浮動閘電晶體515b的汲極。 與鄰近直行的雙電晶體浮動閘NMOS NOR快閃單元510有 關的本地的位元線520a、520b、…、520n-l和520η透 過位元線選擇電晶體560a..... 560η連接到全域的位元The Fowler-Nordheim channel operates to erase the block of the selected NOR flash memory element. The second step is to trim the maximum threshold voltage to the desired voltage using a one-bit one-dimensional Fowler-Nordheim boundary tunnel programming operation. 8 is a schematic diagram of a NOR flash memory component 500 including various embodiments of a dual transistor floating gate NM0S n〇R flash cell 510 of the present invention. [0077] FIG. The NOR flash memory component 500 includes an array 505 of dual-electric crystal floating gate NMOS NOR flash cells 510 arranged in straight and horizontal rows. Each dual transistor floating gate NMOS NOR flash cell 510 includes two NMOS NAND flash floating gate transistors 515a and 515b. Two NMOS NAND flash floating gate transistors 5153 and 5151) are configured to operate the _%4, Figure 46-1, Figure 4C-2, Figure 4C-1, and Figure 4C-2. Transistors 405a and 405b. The drain of the NMOS NAND flash floating gate transistor 515a is connected to one of the local bit lines 520a, 520b, ..., 520n-1, and 520n. The source of the NMOS NAND floating gate transistor 515b is connected to one of the source lines 530a, 53Q|, ..., 530n-1, and 530n. The source of the NMOS NAND flash floating gate transistor 515 a is connected to N Μ 0 S Ν 0 R fast * j ψ * t (> The flip-flop of the floating floating gate transistor 515b. The double transistor floating gate adjacent to the straight line The local bit lines 520a, 520b, ..., 520n-1, and 520n associated with the NMOS NOR flash cell 510 are connected to the global bit by the bit line selection transistors 560a.....

線525a..... 525η。與鄰近直行雙電晶體浮動閘NMOS NOR快閃單元510有關的本地的源極線530a、530b、…、 530n-l和530η透過源極線選擇電晶體565a..... 565η 被連接到全域的源極線540a、…、540η。全域的位元線 099113777 表單編號Α0101 第39頁/共133頁 0992024278-0 201104843 525a、…、525η以及全域的源極線540a、…、540η連 接到直行電壓控制電路(column voltage control circuit, COLUMN VOLTAGE CTL)555。直行電壓控制 電路555產生適當的電壓以選擇性地讀取、編程和抹除雙 電晶體浮動閘NMOS N0R快閃單元510。 [007S] 陣列505中每一橫列上的雙電晶體浮動閘丽〇S N0R快閃 單元510的NMOS NAND快閃浮動閘電晶體515a和515b的 每一控制閘連接到字元線545a、545b、…、545m中之一 。在橫列電壓控制電路(1"〇¥飞:〇1技轻6.00111:1'〇1(^1·- cuit, ROW V CTL CKT)550 中字元線545a、545b、… 、545m連接到字元線電壓控制子電路(ίττοΜ line voltage control circuit, WOKD LINE VOLTAGE CTL)552。 [0079]位元線選擇電晶體560a..... 560n的每一閘極連接到橫 列電壓控制電路550内的位元線選擇控制子電路(bit line select control sub-circuit, BL SEL CTL)551 ’以提供選擇訊號啟動位元線選擇電晶體56〇a .....56〇n,從而將一被選擇的本地的位元線520a、 520b.....520n-l和520η連接到與其相對應的全域的 位元線525a、…、525η。源極線選擇電晶體565a..... 565η的每一閘極連接到橫列電壓控制電路ho之内的源極 線選擇控制子電路(source line voltage control sub-circuit, SOURCE LINE VOLTAGE CTL) 553 > 以將本地的源極線53〇a、53〇b、…、53〇n—i和53〇n連 接到與其相對應的全域源極線54〇a、…、54〇n。 099113777 表單編號A0101 第40頁/共133頁 0992024278-0 201104843 [0080]源極線選擇電晶體565a..... 565η的每一閘極連接到橫 列電壓控制電路5 5 0之内的源極線選擇控制子電路5 $ 3, 以提供選擇訊號啟動源極線選擇電晶體565a、...、 ’從而將一被選擇的本地的源極線530a、530b..... 530n-l和530η連接到與其相對應的全域的源極線54〇& .....540n。源極線選擇電晶體565a..... 565η的每一 閘極連接到橫列電壓控制電路55〇之内的源極線選擇控制 子電路553,以將本地的源極線53〇a、53〇b、...、Line 525a..... 525η. Local source lines 530a, 530b, ..., 530n-1, and 530n associated with adjacent straight-line dual transistor floating gate NMOS NOR flash cells 510 are connected to the global domain through source line selection transistors 565a.....565n Source lines 540a, ..., 540n. Global bit line 099113777 Form number Α0101 Page 39/133 page 0992024278-0 201104843 525a,...,525η and global source lines 540a,...,540η are connected to the column voltage control circuit (COLUMN VOLTAGE) CTL) 555. The straight line voltage control circuit 555 generates an appropriate voltage to selectively read, program, and erase the dual crystal floating gate NMOS NOR flash unit 510. [007S] Each control gate of the NMOS NAND flash floating gate transistors 515a and 515b of the dual transistor floating gate SNOF flash unit 510 on each of the arrays 505 is coupled to word lines 545a, 545b. One of ..., 545m. In the horizontal voltage control circuit (1"〇¥飞:〇1技术轻6.00111:1'〇1(^1·- cuit, ROW V CTL CKT) 550, the word lines 545a, 545b, ..., 545m are connected to the character A line voltage control circuit (WOKD LINE VOLTAGE CTL) 552. [0079] Each gate of the bit line selection transistors 560a..... 560n is coupled to the course voltage control circuit 550. A bit line select control sub-circuit (BL SEL CTL) 551 ′ is provided to select a signal to activate the bit line selection transistor 56〇a .....56〇n, thereby selecting one The local bit lines 520a, 520b.....520n-1 and 520n are connected to their corresponding global bit lines 525a, ..., 525n. The source line selection transistors 565a.....565n Each gate is connected to a source line voltage control sub-circuit (SOURCE LINE VOLTAGE CTL) 553 > to the local source line 53〇a, 53〇b, ..., 53〇n-i and 53〇n are connected to their corresponding global source lines 54〇a, ..., 54〇n. 099113 777 Form No. A0101 Page 40 of 133 Page 0992024278-0 201104843 [0080] Each gate of the source line selection transistor 565a.....565n is connected to the source within the row voltage control circuit 550 The pole select control sub-circuit 5 $ 3 is operative to provide a select signal to activate the source line select transistor 565a, ..., ' so that a selected local source line 530a, 530b..... 530n-l And 530n are connected to their corresponding global source lines 54A & .... 540n. Each gate of the source line selection transistors 565a.....565n is connected to the course voltage control circuit 55. The source line selection control sub-circuit 553 within the 〇 to local source lines 53〇a, 53〇b, ...,

530Π-1和530n連接到與其相對應的全域的源極線“Μ、 …、540η。 闕圖9為橫列電壓控制電路55{)的示意圖。橫列電壓控制電 路550包括一控制解碼器(ccmuol decoder·, DCDR)6G5,用於接收編程時序和控制訊號61()、抹除時 序和控制訊號615以及讀取時序和控制訊號62〇。該控制 解碼益6〇5解碼編程時序和控制訊號6]〇、抹除時序和控 制讯號615以及讀取時序和控制訊g62〇以建立_快問 記憶體元件500的操作。該横列電麵制電路550包括一 位址解碼器 Uddress dec〇der,ADDRDCDR)625,用 於接收和解碼一位址訊號63〇,以提供待被編程操作、抹 除雜或者讀取操作的被選擇雙電晶趙浮動閘_s麵 快閃單元51〇的位置。 099113777 [0082] 該位元線選擇控制子電路551從控制解瑪器⑽接收已被 p的編簡作、抹_作和讀取操作的時序和控制訊 :還從位址解瑪H 625接收已被解碼的位址。位元線選 制子電路551選擇位⑽選擇訊號570a、...、570η 表單鵠號Α010Ι 第41頁/共133頁 0992024278-0 201104843 中的一個以啟動位元線選擇電晶體56〇a..... 560η,以 將已連接至NOR快閃記憶體元件5〇〇的本地的位元線 520a ' 520b.....52〇n-l和520η連接到相對應的全域 的位元線525a、…、525η。 [0083] 該源極線選擇控制子電路553從控制解碼器605接收已被 解碼的編程操作、抹除操作和讀取操作的時序和控制訊 號,還從位址解碼器625接收已被解碼的位址。該源極線 選擇控制子電路5 53選擇源極線選擇訊號575a..... 575η中的一個以啟動源極線選擇電晶艎565a..... 565n530Π-1 and 530n are connected to their corresponding global source lines “Μ, ..., 540η. Figure 9 is a schematic diagram of the horizontal voltage control circuit 55{). The course voltage control circuit 550 includes a control decoder ( Ccmuol decoder·, DCDR) 6G5, used to receive programming timing and control signals 61(), erase timing and control signals 615, and read timing and control signals 62. This control decodes the decoding of the programming timing and control signals. 6] 〇, erase timing and control signals 615 and read timing and control signals to establish the operation of the memory component 500. The horizontal electrical circuit 550 includes a bit address decoder Uddress dec〇der , ADDRDCDR) 625, for receiving and decoding the address signal 63 〇 to provide the position of the selected dual-electro-optic floating gate _s surface flash unit 51 待 to be programmed, erased or read. 099113777 [0082] The bit line selection control sub-circuit 551 receives timing and control information from the control numerator (10) that has been edited, erased, and read by p: also from the address address H 625 Receive the address that has been decoded. Bit line selection The path 551 selects the bit (10) to select the signal 570a, ..., 570n, the form 鹄 Α Ι Ι 41 41 41 41 41 41 41 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 2011 The local bit lines 520a' 520b..... 52 〇 nl and 520 η connected to the NOR flash memory device 5 连接 are connected to the corresponding global bit lines 525a, ..., 525 n. The source line selection control sub-circuit 553 receives the timing and control signals of the programmed operation, the erase operation, and the read operation that have been decoded from the control decoder 605, and also receives the decoded bits from the address decoder 625. The source line selection control sub-circuit 535 selects one of the source line selection signals 575a.....575n to activate the source line selection transistor 565a..... 565n

,以將已連接至N0U快閃記憶體元伴5〇〇的本地的源極線 530a、530b、…、連接到相對應的全域 的源極線540a、…、540η。The local source lines 530a, 530b, ... connected to the N0U flash memory cell are connected to the corresponding global source lines 540a, ..., 540n.

[0084] 該字元線電壓控制子電路552包括一編程電壓產生器635 、一抹除電壓產生器640 ' —讀取電壓產生器645和一橫 列選擇開關650。該編程電壓產生器包括一脈衝增大 電壓產生器636,以提供一從大約15 (^逐漸增大到大約 + 20. 0V的脈衝電纖,從而可以更精確穩定地設置圖8中 NMOS NAND浮動閘電晶體515a和515b的臨界電壓值。第 一實施例中,一正極編程電壓產生器637用於提供一大約 + 5. 0V的電壓;第二實施例中,該正極編程電壓產生器 637則用於提供一大約+ 2.5V的電壓,以防止圖8中未被 選擇的NMOS NAND快閃浮動閘電晶體5153和5151)的編程 操作被限制。在第二實施例中,抹除和編程條件如圖 7a-7d所描述的被反轉。根據圖7&_7(1中的電壓分配關係 ,負極編程電壓產生器638提供大約_1〇. 〇v的負電壓以 099113777 表單編號A0101 第42頁/共133頁 0992024278-0 201104843 對圖8中未被選擇的NMOS NAND快閃浮動閘電晶體515a和 515b進行編程操作。—接地參考電壓源639用於使得所有 位於一NOR快閃記憶體元件500之内的雙nm〇S NAND快閃 浮動閘電晶體515a和515b相互絕緣,以防止圖8中該等 NMOS NAND快閃浮動閘電晶體515a和515b中已建立的編 程被損壞。 [0085] Ο 該抹除電壓產生器640包括一正極抹除電壓產生器642, 用於提供必要的正極電壓以抹除N〇R快閃記憶體元件5〇〇 在第一實施例中未被選擇的字元線,從而防止圖8中未被 選擇的NMOS NAND快閃浮動閘電晶體5丨5&和5丨5b中的編 程被損壞。第二實施例中,灣辑辣除每產j生器642用於 提供所需要的電壓以對圖8 的閃浮動閘電 ί.ϋΊ: .:乂‘寒.1¾ 晶體515a和515b進行抹除操作。在第一實施例中,該抹 除電壓產生器640包括一負極抹除電壓產生器643,用於 對圖8中NMOS NAND快閃浮動閘電晶體5丨53和5丨5b進行 [0086] 抹除操作。在第二實施例中,該未被選擇的字元線的電 壓則被設置到接地參考電壓源644。 爲讀取單階單元數據,該讀取電壓產生器645包括一第一 高讀取電壓產生器646,該第—高讀取電壓產生器646用 於提供必要的讀取電壓VH給圖8中NMOS NAND快閃浮動閘 電晶體515a和515b的被選擇字元線控制閘。爲讀取多階 單元數據,該讀取電壓產生器645還包括一第二和第三高 讀取電壓產生器647和648,該第二和第三高讀取電壓產 生器647和648分別用於提供必要的讀取電壓VH1和VH2給 圖8中的NMOS NAND快閃浮動閘電晶體5〗5&和5丨肋的被 099113777 表單編號A0101 第43,頁/共133頁 0992024278-0 201104843 [0087] [0088] [0089] 099113777 選擇控制閘。該讀取電壓產生器645還提供〜電壓源產生 器649到圖8中的NMOS NAND快閃浮動閘電晶 515b的控制閘,以讀取單階單元數據。 該橫列電壓控制電路包括一橫列選擇開關,以傳輸編程 電壓產生器635、抹除電壓產生器640和讀取電壓產生器 645的編程電壓、抹除電壓和讀取電壓到被選擇的字元線 545a,545b,…,545m。 請參閱圖10 ’其描述直行電壓控制電路555。該直行電壓 控制電路5 5 5包括一控制魏碼器7 〇 5,該控制解碼器7 〇 5 用於接收編程時序和控制訊號710、抹除時序和控制訊號 715、讀取時序和控制訊號720 該控制解瑪器7〇5還用 於對編程時序和控制訊號710、抹除時序和控制訊號715 、讀取時序和控制訊號720進行解碼,议對NOR快閃記憶 體元件500進行操作。該直行電壓控制電路555還包括一 位址解碼器725,該位址解攀琴*725用於接收和解碼一位 址说號730 ’以提供選擇的雙電晶趙浮動閘NMOS NAND快 閃單凡51〇的位址,從而對其進行編程、抹除或讀取操作 〇 該直行電壓控制電路555還包括一編程電壓產生器735、 一抹除電壓產生器74〇、一讀取電壓產生器745及一直行 選擇開關750 °該編程電壓產生器735包括一編程電壓源 736 ’第一實施例中,該編程電壓源736用於提供一大約 + 10. 0V的編程抑制電壓給圖8中未被選擇的NM〇s NANd 快閃浮動閘電晶體515a和515b的汲極和源極,以抑制對 該未被選擇的NMOS NAND快閃浮動閘電晶體515a和515b 表單編號A0101 第44頁/共133頁 201104843 的編程操作。第二實施例中,在編程操作期間,該編程 〇 ^ [0090] 電壓源736用於提供一大約+ 5. 0V的電壓給圖8中被選擇 的NMOS NAND快閃浮動閘電晶體515a和515b的汲極。第 一實施例中’在編程操作期間,一接地參考電壓源737還 被提供給圖8中被選擇的NMOS NAND快閃浮動閘電晶體 515a和515b的汲極和源極。對一些圖8中未被選擇NMOS NAND快閃浮動閘電晶體515a和515b,該接地參考電壓源 737還被提供給未被選擇NMOS NAND快閃浮動閘電晶體 515a和515b,以抑制對其的編程摇作。 該抹除電壓產生器740包括一抹除電壓源742,該抹除電 壓源742用於提供必要的正極電壓,從而實現第一實施例 中對NOR快閃記憶體元件500的抹降栋作。圖8中未被選擇 的NMOS NAND快閃浮動閘電晶體515a和515b的汲極和源 極的電壓則被設置到接地參考電壓源743。 [0091] 爲讀取多階電位單元數據,該讀取電聲產生器745包括一 適中的高讀取電壓源747,該適中的高讀取電壓源747用 於提供必要的讀取電壓VHD給,8中被選擇的NMOS NAND 快閃浮動閘電晶體515a和515b的汲極。爲讀取單階單元 數據,該讀取電壓產生器745還包括一電壓源產生器,該 電壓源產生器用於提供電壓給圖8中的NMOS NAND快閃浮 動閘電晶體515a和515b的汲極。 [0092] 該直行電壓控制電路555包括一直行選擇開關75〇,該直 行選擇開關750用於將編程電壓產生器735、抹除電壓產 生器740和讀取電壓產生器745的編程電壓、抹除電壓和 讀取電壓傳送至被選擇的位元線525a、525b..... 525n 099113777 表單編號A0101 第45頁/共133頁 0992024278-0 201104843 以及源極線540a、540b、…、540η。 [0093] 圖11Α是圖4Α中NMOS NOR快閃記憶體單元400的各種實 施例中單階電位編程電壓追隨感應電路的示意圖。該示 意圖描述在一直行NMOS NAND快閃浮動閘電晶體裡的兩 個NMOS NAND快閃浮動閘電晶體405a和405b。該NAND快 閃浮動閘電晶體4 0 5 a和4 0 5 b中最上端的電晶體的汲極 415連接到本地的位元線805之後透過位元線選擇電晶體 810被連接到全域位元線815。該全域位元線815連接到 圖8中的直行電壓控制電路555。該位元線選擇電晶體810 的閘極連接到圖8的位元線選擇控制子電路5 51,以接收 啟動訊號啟動位元線選擇電〜晶:體810,從而使得最上端的 快閃浮動閘電晶體40 5a的及極415遂輳到電壓源VDD。 [0094] 最下端的NMOS NAND快閃浮動閘電晶體4〇5b的源極422 連接到本地的源極線825。該本地的源極線825透過源極 線選擇電晶體830連接到全域的位元線835。該全域的位 元線835連接到圖10中的直行電壓控制電路555中的感應 放大器755。該感應放大器755包括一比較電路850,該 比較電路850的一端連接到全域的源極線835,另外一端 連接到參考電壓源855。該參考電壓源855的電壓被設置 在代表邏輯“Γ和邏輯“0”的臨界電壓之間。該源極 線選擇電晶體8 3 0的閘極連接到圖8中直行電壓控制電路 555中的源極線電壓控制子電路553。該源極線電壓控制 子電路553用於提供啟動該源極線選擇電晶體83〇所必要 的電壓,從而將本地的源極線825連接到全域位元線835 ,也就是NMOS NOR快閃記憶體單元4〇〇的源極422。當該 099113777 表單編號A0101 第46頁/共133頁 0992024278-0 201104843 NMOS NAND快閃浮動閘電晶體4〇5a和405b被啟動時,其 操作行為與電壓追隨器相似。源極線電容845的電壓等於 電壓源減去NMOS NAND快閃浮動閘電晶體4〇5a或者405b 的編程臨界電壓(Vs = VDD-VtMSEL)。未被選擇的NMOS NAND快問浮動閘電晶體405a或者405b被驅動,以使得它 有一個最小的電壓下跌。取決於該選擇的NMOS ΝΑΟ快 閃浮動閘電晶體405a或者405b的編程臨界電壓,該比較 電路850的輸出電壓將代表邏輯“Γ或者邏輯“〇” 。 [0095] Ο 請參閱圖11B,為了讀取NMOS NANIX快閃浮動閘電晶體 405a和405b中最上端電晶體的單階電位編程單元儲存值 ,第一字元線fLO 450a的電澤被設置到,:¾源VDD的電 壓。該電壓源VDD的電壓為冬約+ 1,,.8V或者大約+3. 0V。 · .....; .· :.....·· . ο 第二字元線WL1 450b的電壓則被設置到一大於+6. ον的 較高讀取電壓,以開啟NMOS NAND快閃浮動閘電晶體 405b。最上端的NMOS NAND快閃浮動閘電磊體405a的汲 極的電壓透過本地的位元線8 0 5和全域的位元線815被設 置到電壓源VDD的電壓。若該NMOS NAND快閃浮動閘電晶 體405a的電壓被設定在第一臨界電壓Vt0(從大約-0· 75V 到大約-0. 25V),最下端的NMOS NAND快閃浮動閘電晶 體405b的源極422,也就是比較電路850的第一輸入端的 電壓值VS0大約等於電壓源VDD的電壓值。若浮動閘電晶 體405a的電壓被設定在第二臨界電壓Vtl (大於+ 3. 0V), 該下端的NMOS NAND快閃浮動閘電晶體405b的源極422 ,也就是比較電路850的第一輸入端的電壓值VS1大約等 於接地參考電壓(0. 0V)的電壓值。如此,該比較電路 099113777 表單編號A0101 第47頁/共133頁. 0992024278-0 201104843 850的輸出端的邏輯值由最上端的NMOS NAND快閃浮動閘 電晶體405a所被編程的臨界電壓值指定。 [0096] 為了讀取該NMOS NAND快閃浮動閘電晶體405a和405b中 最下端的電晶體的SLC儲存值,該第二字元線WL1 450b 的電壓值被設置為電壓源VDD的電壓值。該第一字元線 WLO 450a的電壓值被設置為一大於+ 6.0V的較高讀取電 壓,以開啟該NMOS NAND快閃浮動閘電晶體405a。最下 端的NMOS NAND快閃浮動閘電晶體405b的汲極的電壓透 過最上端的NMOS NAND快閃浮動閘電晶體405a、全域的 位元線815和本地的位元線805被設置為電壓源VDD。若 最下端的NMOS NAND快閃浮動阱電晶體405¾的電壓被設 定在第一臨界電壓VtO(從大約-Q.75Y到大約-0. 25V), 最下端的NMOS NAND快閃浮動閘電晶體4_b的源極422 ,也就是比較電路850的第一輸入端的電壓值VS0大約等 於電壓源VDD的電壓值。由於該NMOS NAND快閃浮動閘電 晶體405b的閘極電壓Vi)D小於Vtl,若NMOS NAND快閃浮 動閘電晶醴405b的電麇被設定在第二臨界電壓Vtl (大 於+ 3. 0V),最下端的NMOS NAND快閃浮動閘電晶體的 405b的源極422,也就是比較電路850的第一輸入端的電 壓值VS1大約等於接地參考電壓(0. 0V)的電壓值。如此 ,最下端的NMOS NAND快閃浮動閘電晶體405b則處於一 非傳導狀態,本地的位元線805即無電壓被傳遞到本地的 源極線選擇電晶體830,故VS1 = 0V。如此,該比較電路 850的輸出邏輯值則由最下端的NMOS NAND快閃浮動閘電 晶體405b所被編程的臨界電壓值指定。 099113777 表單編號A0101 第48頁/共133頁 0992024278-0 201104843 [0097] 在NMOS NOR快閃記憶體單元400的一陣列中,若一nm〇s NOR快閃記憶體單元4〇〇未被選擇讀取而另一個關⑽N〇R 快閃記憶單元被選擇讀取時,該未被選擇的⑽⑽Ν〇Ι^^ 閃記憶體單元400中的非被選擇的關〇5 NAND快閃浮動閘 電晶體405a和405b的控制閘極的電壓被設置到接地參考 電壓,以關閉該電荷保存電晶體》 [0098] Ο 圖11C是圖4A中NMOS NOR快閃記憶體單元400的多階電 位編程的電壓追隨感應電路的具體實施方式的示意圖。 如在圖11A中所描述的一直行NMOS NAND快閃浮動閘電晶 體’該示意圖說明除全域的位元線外,該兩NMOS NAND 快閃浮動閘電晶體405a和405b的電壓均被纖置到一第一 較高的讀取電壓源VHD。 [0099] Ο 在該具體實施例中’全域的位元線835連接到圖10中的直 行電壓控制電路555中的感應放大器755。在該實施例中 ’該感應放大器755包括三個比較電路860、870和880。 該三個比較電路860、870和8各0的每一個電路的第一輸 入端均連接到全域的位元線835,第二輸入端連接到參考 電壓源,其中第一比較電路860的第二輸入端連接到第一 參考電壓源865 REFV0 ;第二比較電路870的第二輸入 端連接到第二參考電壓源875 REFV1 ;第三比較電路 880的第二輸入端連接到第三參考電壓源885 REFV2。 該三個參考電壓源865、875和885的電壓值設置在代表 數據的邏輯值(“00” ,“01” ,“10” ,“11”)的臨 界電壓值之間。該源極線選擇電晶體830的閘極連接到圖 8中橫列電壓控制電路中的源極線電壓控制子電路553。 099113777 表單編號Α0101 第49頁/共133頁 0992024278-0 201104843 該源極線電壓控制子電路553用於提供必要的電壓,以使 得該源極線選擇電晶體830連接到本地的源極線825,也 就是NMOS NOR快閃記憶體單元400的源極422連接到全域 的位元線835。當該NMOS NAND快閃浮動閘電晶體405a 和405b啟動時,其類似一電壓追隨器。在源極線電容845 上的電壓等於電壓源減掉被選擇的NMOS NAND快閃浮動 閘電晶體405a或405b的編程臨界電壓(Vs = VDD-VtMSEL)。該未被選擇的NMOS NAND快閃浮動閘電晶體 405a或405b被驅動,以使其具有最小的電壓降。根據選 擇的NMOS NAND快閃浮動閘電晶體405a或405b的編程臨 界電壓水平,該比較電路850的輸出電壓將以被編程臨界 電壓代表數據邏輯值(“00” ,“01” ,“10” ,“11 ”)。應該注意的是,本實施例所描述的结構適用於一個 兩位元多階單元。可以理解,不脫於本發明精神,任何 數目的數據邏輯值均可以被NMOS NAND快閃浮動閘電晶 體405a和405b保存。The word line voltage control sub-circuit 552 includes a program voltage generator 635, an erase voltage generator 640'-read voltage generator 645, and a column select switch 650. The programming voltage generator includes a pulse increasing voltage generator 636 to provide a pulsed fiber from about 15 (^ gradually increasing to about + 20. 0V) so that the NMOS NAND floating of FIG. 8 can be more accurately and stably set. The threshold voltage values of the gate transistors 515a and 515b. In the first embodiment, a positive programming voltage generator 637 is used to provide a voltage of about +0.5 V; in the second embodiment, the positive programming voltage generator 637 is A programming operation to provide a voltage of approximately +2.5V to prevent unselected NMOS NAND flash floating gate transistors 5153 and 5151 in FIG. 8 is limited. In the second embodiment, the erase and program conditions are reversed as described in Figures 7a-7d. According to the voltage distribution relationship in FIG. 7 &_7 (1), the negative programming voltage generator 638 provides a negative voltage of about _1 〇 〇 以 以 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 表单 表单 表单 表单 表单 表单 表单 对 对 对 对 对 对 对 对 对 对 对 对 对 对 对 对The unselected NMOS NAND flash floating gate transistors 515a and 515b are programmed to operate. - The ground reference voltage source 639 is used to cause all of the dual nm 〇S NAND flash floating gates within a NOR flash memory component 500. The transistors 515a and 515b are insulated from each other to prevent the established programming in the NMOS NAND flash floating gate transistors 515a and 515b of Fig. 8 from being damaged. [0085] The erase voltage generator 640 includes a positive erase The voltage generator 642 is configured to provide a necessary positive voltage to erase the N 〇 R flash memory element 5 未被 unselected word line in the first embodiment, thereby preventing the unselected NMOS in FIG. 8 . The programming in the NAND flash floating gate transistors 5丨5& and 5丨5b is damaged. In the second embodiment, the Bay Series is used to provide the required voltage to flash the Figure 8 Floating brake power ϋΊ.ϋΊ: .:乂'寒.13⁄4 crystals 515a and 5 15b performs an erase operation. In the first embodiment, the erase voltage generator 640 includes a negative erase voltage generator 643 for the NMOS NAND flash floating gate transistors 5 丨 53 and 5 图 of FIG. 5b performs a [0086] erase operation. In the second embodiment, the voltage of the unselected word line is set to the ground reference voltage source 644. To read the single-order unit data, the read voltage generator 645 includes a first high read voltage generator 646 for providing the necessary read voltage VH to the selected ones of the NMOS NAND flash floating gate transistors 515a and 515b of FIG. The word line control gate. To read the multi-level cell data, the read voltage generator 645 further includes a second and third high read voltage generators 647 and 648 for generating the second and third high read voltages. 647 and 648 are respectively used to supply the necessary read voltages VH1 and VH2 to the NMOS NAND flash floating gate transistor 5 of FIG. 8 and the 5 ribs of the 099113777 form number A0101, page 43, total 133 Page 0992024278-0 201104843 [0087] [0089] 099113777 Select control gate. The voltage generator 645 also provides a voltage gate generator 649 to the control gate of the NMOS NAND flash floating gate transistor 515b of Figure 8 to read the single-order cell data. The row voltage control circuit includes a row selection A switch is applied to transfer the programming voltage generator 635, the erase voltage generator 640, and the read voltage generator 645 programming voltage, erase voltage, and read voltage to the selected word line 545a, 545b, ..., 545m. Referring to Figure 10', a straight-line voltage control circuit 555 is described. The straight line voltage control circuit 55 5 includes a control code coder 5 〇 5 for receiving a programming timing and control signal 710, an erase timing and control signal 715, a read timing, and a control signal 720. The control numerator 7〇5 is also used to decode the programming timing and control signals 710, erase timing and control signals 715, read timing and control signals 720, and to operate the NOR flash memory component 500. The straight line voltage control circuit 555 also includes a bit address decoder 725 for receiving and decoding the address number 730' to provide a selected dual crystal thyristor floating gate NMOS NAND flash The 51 〇 address is programmed, erased or read. The straight line voltage control circuit 555 further includes a program voltage generator 735, an erase voltage generator 74, and a read voltage generator 745. And the row select switch 750 °. The program voltage generator 735 includes a program voltage source 736. In the first embodiment, the program voltage source 736 is used to provide a program inhibit voltage of about + 10.0V to the The selected NM〇s NANd flash drain gates 515a and 515b are drained and sourced to suppress the unselected NMOS NAND flash floating gate transistors 515a and 515b. Form No. A0101 Page 44 of 133 Programming operation of page 201104843. In a second embodiment, during a programming operation, the programming source 736 is used to provide a voltage of approximately +5.0 V to the selected NMOS NAND flash floating gate transistors 515a and 515b of FIG. Bungee jumping. In the first embodiment, a ground reference voltage source 737 is also provided to the drain and source of the selected NMOS NAND flash floating gate transistors 515a and 515b of FIG. 8 during a programming operation. For some of the unselected NMOS NAND flash floating gate transistors 515a and 515b of FIG. 8, the ground reference voltage source 737 is also provided to the unselected NMOS NAND flash floating gate transistors 515a and 515b to suppress Programming shakes. The erase voltage generator 740 includes a wipe voltage source 742 for providing the necessary positive voltage to effect the erase of the NOR flash memory device 500 in the first embodiment. The drain and source voltages of the unselected NMOS NAND flash floating gate transistors 515a and 515b of Figure 8 are then set to ground reference voltage source 743. [0091] To read multi-level potential cell data, the read electroacoustic generator 745 includes a moderately high read voltage source 747 for providing the necessary read voltage VHD to The drains of the selected NMOS NAND flash floating gate transistors 515a and 515b. To read the single-order cell data, the read voltage generator 745 further includes a voltage source generator for supplying a voltage to the drain of the NMOS NAND flash floating gate transistors 515a and 515b of FIG. . [0092] The straight line voltage control circuit 555 includes a row select switch 75A for erasing the program voltages of the program voltage generator 735, the erase voltage generator 740, and the read voltage generator 745. The voltage and read voltage are transferred to selected bit lines 525a, 525b..... 525n 099113777 Form No. A0101 page 45/133 pages 0992024278-0 201104843 and source lines 540a, 540b, ..., 540n. 11 is a schematic diagram of a single-stage potential programming voltage following sensing circuit in various embodiments of the NMOS NOR flash memory cell 400 of FIG. The illustration is directed to two NMOS NAND flash floating gate transistors 405a and 405b in an NMOS NAND flash floating gate transistor. The drain 415 of the uppermost transistor of the NAND flash floating gate transistor 4 0 5 a and 4 0 5 b is connected to the local bit line 805 and then connected to the global bit line through the bit line selection transistor 810. 815. The global bit line 815 is coupled to the straight line voltage control circuit 555 of FIG. The gate of the bit line selection transistor 810 is connected to the bit line selection control sub-circuit 51 of FIG. 8 to receive the enable signal enable bit line select transistor 810, so that the uppermost flash floating gate The sum terminal 415 of the transistor 40 5a is connected to the voltage source VDD. [0094] The source 422 of the lowermost NMOS NAND flash floating gate transistor 4〇5b is connected to the local source line 825. The local source line 825 is coupled to the global bit line 835 through the source line select transistor 830. The global bit line 835 is coupled to the sense amplifier 755 in the straight line voltage control circuit 555 of FIG. The sense amplifier 755 includes a comparison circuit 850 having one end coupled to a global source line 835 and the other end coupled to a reference voltage source 855. The voltage of the reference voltage source 855 is set between a threshold voltage representing a logic "Γ and a logic "0". The gate of the source line selection transistor 830 is connected to the straight line voltage control circuit 555 of FIG. a source line voltage control sub-circuit 553. The source line voltage control sub-circuit 553 is configured to provide a voltage necessary to activate the source line select transistor 83, thereby connecting the local source line 825 to the global bit line 835, that is, the source 422 of the NMOS NOR flash memory cell 4 。. When the 099113777 form number A0101 page 46 / 133 pages 0992024278-0 201104843 NMOS NAND flash floating gate transistors 4〇5a and 405b are At startup, its operational behavior is similar to that of a voltage follower. The voltage of the source line capacitor 845 is equal to the voltage source minus the programming threshold voltage of the NMOS NAND flash gate transistor 4〇5a or 405b (Vs = VDD-VtMSEL). The selected NMOS NAND asks the floating gate transistor 405a or 405b to be driven such that it has a minimum voltage drop. Depending on the programmed threshold voltage of the selected NMOS ΝΑΟ flash floating gate transistor 405a or 405b, Than the output voltage of the circuit 850 represents a logical "Gamma] or a logical" square. " [0095] Referring to FIG. 11B, in order to read the single-order potential programming unit storage value of the uppermost transistor in the NMOS NANIX flash floating gate transistors 405a and 405b, the electric potential of the first word line fLO 450a is set to , : 3⁄4 source VDD voltage. The voltage of the voltage source VDD is about +1, .8V or about +3.00V. · .....; .. :.....·· . ο The voltage of the second word line WL1 450b is set to a higher read voltage greater than +6. ον to turn on NMOS NAND fast Flash floating gate transistor 405b. The voltage of the drain of the uppermost NMOS NAND flash floating gate body 405a is set to the voltage of the voltage source VDD through the local bit line 805 and the global bit line 815. If the voltage of the NMOS NAND flash floating gate transistor 405a is set at the first threshold voltage Vt0 (from about -0.75V to about -0.25V), the source of the lowermost NMOS NAND flash floating gate transistor 405b The pole 422, that is, the voltage value VS0 of the first input of the comparison circuit 850 is approximately equal to the voltage value of the voltage source VDD. If the voltage of the floating gate transistor 405a is set at the second threshold voltage Vtl (greater than +3.00V), the source 422 of the lower NMOS NAND flash floating gate transistor 405b, that is, the first input of the comparison circuit 850 The voltage value VS1 of the terminal is approximately equal to the voltage value of the ground reference voltage (0. 0V). Thus, the comparison circuit 099113777 Form No. A0101 Page 47 of 133. 0992024278-0 201104843 The logic value of the output of the 850 is specified by the threshold voltage value programmed by the uppermost NMOS NAND flash floating gate transistor 405a. [0096] In order to read the SLC storage value of the lowermost transistor in the NMOS NAND flash floating gate transistors 405a and 405b, the voltage value of the second word line WL1 450b is set to the voltage value of the voltage source VDD. The voltage value of the first word line WLO 450a is set to a higher read voltage greater than +6.0V to turn on the NMOS NAND flash floating gate transistor 405a. The voltage of the drain of the lowermost NMOS NAND flash floating gate transistor 405b is set to the voltage source VDD through the uppermost NMOS NAND flash floating gate transistor 405a, the global bit line 815, and the local bit line 805. If the voltage of the lowermost NMOS NAND flash floating well transistor 4053⁄4 is set at the first threshold voltage VtO (from about -Q.75Y to about -0.25 V), the lowermost NMOS NAND flash floating gate transistor 4_b The source 422, that is, the voltage value VS0 of the first input of the comparison circuit 850 is approximately equal to the voltage value of the voltage source VDD. Since the gate voltage Vi)D of the NMOS NAND flash floating gate transistor 405b is less than Vtl, if the power of the NMOS NAND flash floating gate transistor 405b is set to the second threshold voltage Vtl (greater than +3.00V) The source 422 of the 405b of the lowermost NMOS NAND flash floating gate transistor, that is, the voltage value VS1 of the first input terminal of the comparison circuit 850 is approximately equal to the voltage value of the ground reference voltage (0. 0V). Thus, the lowermost NMOS NAND flash floating gate transistor 405b is in a non-conducting state, and the local bit line 805, i.e., no voltage, is transferred to the local source line select transistor 830, so VS1 = 0V. Thus, the output logic value of the comparison circuit 850 is specified by the threshold voltage value programmed by the lowermost NMOS NAND flash floating gate transistor 405b. 099113777 Form No. A0101 Page 48 / Total 133 Page 0992024278-0 201104843 [0097] In an array of NMOS NOR flash memory cells 400, if a nm 〇s NOR flash memory cell unit 4 is not selected for reading When the other (10) N 〇 R flash memory cell is selected for reading, the unselected (10) (10) Ν〇Ι ^ ^ flash memory cell 400 in the non-selected 〇 5 NAND flash floating gate transistor 405a And the voltage of the control gate of 405b is set to the ground reference voltage to turn off the charge holding transistor. [0098] FIG. 11C is a multi-step potential programming voltage following induction of the NMOS NOR flash memory cell 400 of FIG. 4A. A schematic diagram of a specific embodiment of the circuit. The NMOS NAND flash floating gate transistor as described in FIG. 11A illustrates that the voltages of the two NMOS NAND flash floating gate transistors 405a and 405b are both fiberized except for the global bit lines. A first higher read voltage source VHD. [0099] In the particular embodiment, the global bit line 835 is coupled to the sense amplifier 755 in the straight voltage control circuit 555 of FIG. In this embodiment, the sense amplifier 755 includes three comparison circuits 860, 870, and 880. A first input of each of the three comparison circuits 860, 870, and 8 is coupled to a global bit line 835, the second input being coupled to a reference voltage source, wherein the second comparison circuit 860 is second The input is coupled to a first reference voltage source 865 REFV0; the second input of the second comparison circuit 870 is coupled to a second reference voltage source 875 REFV1 ; the second input of the third comparison circuit 880 is coupled to a third reference voltage source 885 REFV2. The voltage values of the three reference voltage sources 865, 875, and 885 are set between the threshold voltage values representing the logical values of the data ("00", "01", "10", "11"). The gate of the source line selection transistor 830 is connected to the source line voltage control sub-circuit 553 in the horizontal voltage control circuit of FIG. 099113777 Form Number Α0101 Page 49 of 133 Page 0992024278-0 201104843 The source line voltage control sub-circuit 553 is used to provide the necessary voltage to connect the source line select transistor 830 to the local source line 825, That is, the source 422 of the NMOS NOR flash memory cell 400 is connected to the global bit line 835. When the NMOS NAND flash floating gate transistors 405a and 405b are activated, they are similar to a voltage follower. The voltage on source line capacitance 845 is equal to the voltage source minus the programmed threshold voltage (Vs = VDD - VtMSEL) of the selected NMOS NAND flash floating gate transistor 405a or 405b. The unselected NMOS NAND flash floating gate transistor 405a or 405b is driven to have a minimum voltage drop. Depending on the programmed threshold voltage level of the selected NMOS NAND flash floating gate transistor 405a or 405b, the output voltage of the comparison circuit 850 will represent the data logic value ("00", "01", "10", with the programmed threshold voltage. "11"). It should be noted that the structure described in this embodiment is applicable to one two-dimensional multi-order unit. It will be appreciated that any number of data logic values may be preserved by NMOS NAND flash floating gate transistors 405a and 405b without departing from the spirit of the present invention.

[0100] 圖11D討論讀取NM Of N OR擎閃記憶體單元400多階電位編 程的偏壓。該第一字元線WL0 450a的電壓被設定為第一 較高讀取電壓VH0,以讀取該NMOS NAND快閃浮動閘電晶 體405a和405b最上端的電晶體。該第一較高讀取電壓 VH0大約為4. 0V。該第二字元線WL1 450b的電壓被設定 為一大於+ 7.0V的第二較高讀取電壓VH1,以開啟NMOS NAND快閃浮動閘電晶體405b。該最上端的NMOS NAND快 閃浮動閘電晶體405a的汲極的電壓透過本地的位元線805 和全域位元線815被設置到一第三較高電壓源VHD 099113777 表單編號A0101 第50頁/共133頁 0992024278-0 201104843 04. 0V) 〇 [0101] ο ο 若NMOS NAND快閃浮動閘電晶體405a的電壓被設定為第 —臨界電壓VtO(從大約-0.75V到大約-0.25V),則最下 端的NMOS NAND快閃浮動閘電晶體405b的源極422,也 就是比較電路850的第一輸入端的電壓VS0大約為第三高 讀取電壓VHD。若NMOS NAND快閃浮動閘電晶體405a的 電壓被設定為第二臨界電壓Vtl(大約+ 1.0V),最下端的 NMOS NAND快閃浮動閘電晶體405b的源極422的電壓VS1 ’也就是比較電路850的第一輸入端的電壓為大約3. 〇v。 若NMOS NAND快閃浮動閘電晶體4〇5a的電壓被設定為第 三臨界電壓Vt2(大約2. 0V),最下端的NMOS NAND快閃 浮動閘電晶體405b的源極422的電壓VS2,也就是比較 電路850的第一輸入端的電壓大約為2. OV »若NMOS NAND快閃浮動閘電晶體405a的電壓被設定為第二臨界電 壓Vt3(大約+ 3. 0V),最下端的MOS NAND快閃浮動閘電 晶體405b的源極422的電壓VS3,也就是比較電路850的 第一輸入端的電壓大約為接地參考電壓(1. 0V)。之後該 y ·. --%*-· ·ί! 比較電路850的輸出端所設定的邏輯狀態則由最上端的 NMOS NAND快閃浮動閘電晶體40 5a被編程的臨界電壓決 定。 [0102] 為了讀取該NMOS NAND快閃浮動閘電晶體405a和405b中 最下端的電晶體的多階電位編程,該第二字元線WL1 450b的電壓被設置到VHD的電壓。該第一字元線WL0 450a的電壓被設置到一大於+ 6. 0V的較高讀取電壓,以 開啟該NMOS NAND快閃浮動閘電晶體405a。被SLG [η] 099113777 表單编號Α0101 第51頁/共133頁 0992024278-0 201104843 [0103] 閘極控制的最下端的選擇電晶體的全域源極線的電壓G S L 是透過最下端的NMOS NAND快閃浮動閘電晶體405b、最 上端的NMOS NAND快閃浮動閘電晶體405a、本地的位元 線805、被BLG [ η ]閘門控制的最上端選擇電晶體 Msel、全域位元線815而設置地。頂端和下端的選擇電晶 體的閘極電壓必須要耦合至高讀取電壓與臨界電壓之和 (VHD + Vt)的水平,才能完全把充足的VHD電壓從GBL傳遞 到 GSL。 若NMOS NAND快閃浮動閘電晶艘4 〇 ib的電壓被設定在第 一臨界電壓VtO (從_約-0.75 V到大約-0.25V),最 下端的NMOS NAND快閃浮動勝電#a體4051)的源極422的 電壓VS0 ’也就是比較電路850的第一輪入端的電壓大約 等於一第三高讀取電壓源VHD。若該NMOS NAND快閃浮動 閘電晶體405b的電壓被設定在第二臨界電壓Vtl (大約 + 1· 0V) ’而且VHD大約等於4. 0V,則最下端的NMOS NAND快閃浮動閘電晶體40让以澈柽422的電壓VS1,也就 是比較電路850的第一輸入端的電壓大約等於3. 0V。若 NMOS NAND快閃浮動閘電晶體405b的電壓被設定在第三 臨界電壓Vt2 (大約2.0V),則最下端的NMOS NAND快閃 浮動閘電晶體405b的源極422的電壓VS2,也就是比較電 路850的第一輸入端的電壓大約為2. 0V。若該NMOS N A N D快閃浮動閘電晶體4 0 5 b的電壓被設定為第s臨界電 壓Vt3(大約+3. 0V),最下端的NMOS NAND快閃浮動閘電 晶體405b的源極422的電壓VS3,也是該比較電路850的 第一輸入端的電壓大約為1. 0V。如此,比較電路850的輸 099113777 表單編號A0101 第52頁/共133頁 099 201104843 出端顯示最下端的NMOS NAND快閃浮動閘電晶體4〇讥所 被編程的臨界電壓代表的邏輯值。 [0104]在圖11A和圖11C中’圖4B-2和4C-2中的三重p井430連 接到接地參考電壓(0.0V),深n井435則連接到電壓源 VDD 〇 [0105] Ο [0106] 在NMOS NOR快閃記憶體早元4 〇 〇的一陣列中,若一nm〇S NOR快閃記憶體單元400未被選擇進行讀取操作,而另一 個NM0S被選擇進行讀取操作時,該未被選擇的NM〇s N〇R 快閃記憶體單元400中未被選擇的NM0S NAND快閃浮動閘 電晶體4 0 5 a和4 0 5 b的控制閉的電麥則被;定為接地參考 電壓,以關閉電荷保存電晶纏。響、 圖 12A-12E是圖4A、4B-1 中雙電 晶體浮動閘NMOS NOR快閃單元的抹除偏電壓表。請參閱 圖12B-12E,該四個表格的抹除偏電壓用於提供一抹除條 件以使得圖4A、4B-1、4B-2 ' 4C-1和4CM2中汲極415、 420和源極420、422之間的基極通道运432a和432b與控 制閘4258或4251)之間的電壓降在|?0|161^01"(^丨111通道 抹除期間被設定為一大約+ 20· 0V的電壓。圖12A中,被 選擇的字元線450a或450b ’也就是控制閘425a或425b 的電壓被設定為一大約-10· 0V的負極抹除電壓。該汲極 415和420、源極420和422、三重ρ井43〇和深Ν井435的 電壓被設定為一大約+ 10. 0V的正極抹除電壓。該未被選 擇的字元線450a或450b,也就是未被選擇的控制閘425a 或425b的電壓則被設定為一大約+ 10. 〇v的遮蔽抹除電壓 099113777 表單編號Α0101 第53頁/共133頁 0992024278-0 201104843 [0107] 在圖12B中’該負極抹除電壓大約為-15 〇V,正極抹除 電壓大約為+ 5. 0V,正極遮蔽電壓大約為+ 5. 0V。在圖 12c中’該負極抹除電壓大約為_2〇. 〇v,正極抹除電愿 大約為0· 0V ’正極遮蔽電壓大約為〇· 〇v。在圖12])中, 電壓水平被反轉,該負極抹除電壓大約為〇. 〇v,正極抹 除電壓大約為+ 20.0V。圖12A-12D中每一電壓產生— Fowler-Nordheim通道穿隧效應,以減少被選擇的_〇s NAND快閃浮動閘電晶體4〇5a或405b的臨界電壓。 [0108] 圖4A、4B-1、4B-2、4C-1和4C-2中未被選擇的雙電晶 體浮動閘NMOS NAND快問單元不共用同一個三重p井43〇 和深N井435。該未被選擇的字元線450a或450b,也就是 控制閘425a或425b、汲極415和420、源極420和422和 三重P井430的電壓被設定為大約等於接地參考電壓。該 深N井435的電壓則被設定為電壓源極VDD的電壓。 [0109] 對於浮動閘NMOS NAND快閃單元的一個陣列中的子陣列( 經常為512Kb或者4Kb.的區塊),該等:未被抹除的子陣列 的深N井的電壓被設定為+ 20V,其字元線、汲極、源極和 三重P擴散井的電壓則被設定為接地參考電壓。該等在不 同深N井中未被選擇的子陣列的字元線、波極、源極、= 重P井和深N擴散井的電壓則被設定為接地參考電壓。 [〇11〇]圖12E中討論另一抹除和編程臨界電壓被反轉時的抹除操 作。在這種情況下,該被選擇的字元線45〇a或450b,也 就是控制閘425a或425b的電壓被設定為一大約+ 20· (^的 正極編程電壓。該控制閘4 2 5 a或4 2 5 b、没極415和4 2 〇、 源極420和422、三重P井430的電壓被設定為接地參考電 099113777 表單編號A0101 第54頁/共133頁 201104843 壓(0. OV)。該深N井435的電壓被設定為電壓源的電壓。 該等設置抹除的臨界電壓到正極電壓的條件和設置編程 的臨界電壓到負極電壓的條件如圖7A-7D所示。 [0111] 圖 13Α和 13Β是對圖4Α、4Β-1、4Β-2、4C-1 和4C-2 中雙 電晶體浮動閘NMOS NOR快閃單元進行編程操作時的編程 偏電麼表。在對圖4A、4B-1、4B_2、4C-1和4C-2中雙 電晶體浮動閘NMOS NAND快閃單元中被選擇的_〇s NAND快閃浮動閘電晶體405a或405b進行編程操作之前, 該單元必須被抹除。在如第8圖所示的一雙電晶體浮動閘 NMOS NAND快閃單元的一陣列中,抹除操作是針對一頁 或一區塊的單元進行的。 [0112] 對圖 4A、4B-1、4B-2、4C-l·和4G-2 中/¾隹擇的NM0S NAND快閃浮動閘電晶體405a或405b進行編程操作時,該 被選擇的字元線450a或450b,也就是控制閘425a或 425b的電壓被設定為一大約+ 15> 0.V至+ 20: 0V的正極編 程電壓。沒極415和420、_極420和422、通道區432a 和432b的電壓透過三重P井430被鲜定為接地參考電壓 (0. 0V)。未被選擇的NMOS NAND快閃浮動閘電晶體4〇5a 或405b的字元線450a或450b與其控制閘425a或425b相 連,以將其電壓設定為大約+ 5. 0V的遮蔽編程電壓。如圖 8中所示的一陣列内,位於被選擇的字元線450a或450b 上的未被選擇的浮動閘NMOS NAND快閃單元的汲極和源 極的電壓被設定為一從大約+ 7. 0V到大約+ 10. 〇v的正極 編程遮蔽電壓。如圖8中所示的一陣列内,具有正極遮蔽 電壓的共用位元線455a、455b和源極線460a、460b的 099113777 表單編號A0101 第55頁/共133頁 0992024278-0 201104843 未被選擇的浮動閘NMOS NAND快閃單元的字元線450a和 450b的電壓被設定為+ 5. 0V的正極遮蔽編程電壓。該等 字元線450a、450b或位元線455a、455b或源極線460a 、46 Ob中沒有與正極編程電壓或正極編程遮蔽電壓相連 的未被選擇浮動閘NMOS NOR快閃單元的電壓則被設定為 接地參考電壓(O.OV)。習知的是,當施加於控制閘425a 或425b的正極編程電壓愈高,編程操作之後的臨界電壓 Vt亦愈高。在編程操作期間,為了保持能精確控制NM0S NAND快閃單元的臨界電壓,閘極被施加一從大約+ 15. 0V 到大約+ 16. 0V的初始正極編赛電♦後在每一次編程 〇 操作時反覆地小量遞增該正極編程電壓。上述的編程電 壓適用於對單階電位單元或多階電位單元進行編程操作 ’其臨界電壓如圖6A-6D所示。 [0113] [0114] 隨著被選擇的區塊中被選擇的編程單元的沒極電壓和浮 動源極的電壓逐漸小量的增加負閘極電壓,這是反復的 編程操作和編程檢驗步驟。例如,汲極(采地的BL)電壓 被耦和到固定的+ 5V且本地的SL處於浮動狀態。圖8F為對 ◎ 被選擇的單元M0進行編程操作的較佳偏壓條件。-10V的 閘極電壓被施加於被選擇的單元M0的WL0上。該-10V的 閘極電壓可以從-5V開始然後逐漸下降至-10V。換句話 說’單元的電壓值Vt能精確地被控制至期望的電壓值。 請參閱圖13B,其描述如圖7A-7B所示的反轉編程和抹除 條件的編程電壓。本實施例中,被選擇的NMOS NAND快 閃浮動閘電晶體405a或405b的被選擇字元線450a或 450b的電壓被設置到大約-10. ov的負編程電壓◊汲極 099113777 表單編獍A0101 第56頁/共133頁 0992024278-0 201104843 Ο [0115][0100] FIG. 11D discusses reading the bias voltage of the multi-stage potential programming of the NM Of N OR flash memory cell unit 400. The voltage of the first word line WL0 450a is set to a first higher read voltage VH0 to read the uppermost transistor of the NMOS NAND flash floating gate transistors 405a and 405b. 0伏。 The first higher read voltage VH0 is about 4. 0V. The voltage of the second word line WL1 450b is set to a second higher read voltage VH1 greater than +7.0V to turn on the NMOS NAND flash floating gate transistor 405b. The voltage of the drain of the uppermost NMOS NAND flash floating gate transistor 405a is set to a third higher voltage source VHD 099113777 through the local bit line 805 and the global bit line 815. Form No. A0101 Page 50 / Total 133 pages 0992024278-0 201104843 04. 0V) 〇[0101] ο ο If the voltage of the NMOS NAND flash floating gate transistor 405a is set to the first threshold voltage VtO (from about -0.75V to about -0.25V), then The source 422 of the lowermost NMOS NAND flash floating gate transistor 405b, that is, the voltage VS0 of the first input of the comparison circuit 850 is approximately the third highest read voltage VHD. If the voltage of the NMOS NAND flash floating gate transistor 405a is set to the second threshold voltage Vtl (about + 1.0 V), the voltage VS1 ' of the source 422 of the lowermost NMOS NAND flash floating gate transistor 405b is also compared. The voltage at the first input of the circuit 850 is approximately 3. 〇v. If the voltage of the NMOS NAND flash floating gate transistor 4〇5a is set to the third threshold voltage Vt2 (about 2.0 V), the voltage VS2 of the source 422 of the lowermost NMOS NAND flash floating gate transistor 405b is also That is, the voltage at the first input terminal of the comparison circuit 850 is about 2. OV » If the voltage of the NMOS NAND flash floating gate transistor 405a is set to the second threshold voltage Vt3 (about + 3. 0V), the lowermost MOS NAND is fast. The voltage VS3 of the source 422 of the flash floating gate transistor 405b, that is, the voltage at the first input of the comparison circuit 850 is approximately the ground reference voltage (1.0 V). Thereafter, the logic state set by the output of the comparison circuit 850 is determined by the threshold voltage at which the uppermost NMOS NAND flash floating gate transistor 40 5a is programmed. [0102] To read the multi-level potential programming of the lowermost transistor in the NMOS NAND flash floating gate transistors 405a and 405b, the voltage of the second word line WL1 450b is set to the voltage of the VHD. The voltage of the first word line WL0 450a is set to a higher read voltage greater than +6.00V to turn on the NMOS NAND flash floating gate transistor 405a. By SLG [η] 099113777 Form No. 1010101 Page 51 / Total 133 Page 0992024278-0 201104843 [0103] The lowermost terminal of the gate control selects the voltage of the global source line of the transistor. The GSL is transmitted through the lowermost NMOS NAND. The flash floating gate transistor 405b, the uppermost NMOS NAND flash floating gate transistor 405a, the local bit line 805, the uppermost selection transistor Msel controlled by the BLG [[η] gate, and the global bit line 815 are provided. The gate voltage of the top and bottom select transistors must be coupled to the high sum of the read voltage and the threshold voltage (VHD + Vt) to fully transfer sufficient VHD voltage from GBL to GSL. If the voltage of the NMOS NAND flash floating gate transistor 4 〇 ib is set at the first threshold voltage VtO (from _about -0.75 V to about -0.25 V), the lowermost NMOS NAND flash floating flash power #a body The voltage VS0' of the source 422 of 4051), that is, the voltage at the first wheel input of the comparison circuit 850 is approximately equal to a third high read voltage source VHD. If the voltage of the NMOS NAND flash floating gate transistor 405b is set to the second threshold voltage Vtl (about + 1·0V)' and the VHD is approximately equal to 4.0V, the lowermost NMOS NAND flash floating gate transistor 40 0伏。 The voltage VS1 of the voltage 422, that is, the voltage of the first input of the comparison circuit 850 is approximately equal to 3. 0V. If the voltage of the NMOS NAND flash floating gate transistor 405b is set to the third threshold voltage Vt2 (about 2.0 V), the voltage VS2 of the source 422 of the lowermost NMOS NAND flash floating gate transistor 405b is also compared. The voltage at the first input of the circuit 850 is approximately 2.0 V. If the voltage of the NMOS NAND flash floating gate transistor 4 0 5 b is set to the sth threshold voltage Vt3 (about +3.0 V), the voltage of the source 422 of the lowermost NMOS NAND flash floating gate transistor 405b.伏之间。 The voltage of the first input terminal of the comparator circuit 850 is about 1. 0V. Thus, the input of the comparison circuit 850 099113777 Form No. A0101 Page 52 of 133 099 201104843 The output shows the logic value represented by the threshold voltage of the lowest NMOS NAND flash floating gate transistor. [0104] In FIGS. 11A and 11C, the triple p well 430 in FIGS. 4B-2 and 4C-2 is connected to the ground reference voltage (0.0 V), and the deep n well 435 is connected to the voltage source VDD 〇 [0105]. [0106] In an array of NMOS NOR flash memory early 4 ,, if one nm 〇 S NOR flash memory cell 400 is not selected for read operation, another NMOS is selected for read operation When the unselected NM〇s N〇R flash memory cell 400 is not selected, the NM0S NAND flash floating gate transistors 4 0 5 a and 4 0 5 b are controlled to be closed; The ground reference voltage is set to turn off the charge to save the electric crystal. Figure 12A-12E is the erase bias voltage meter of the double crystal floating gate NMOS NOR flash unit in Figures 4A and 4B-1. Referring to Figures 12B-12E, the erase bias voltages of the four tables are used to provide an erase condition such that the drains 415, 420 and source 420 of Figures 4A, 4B-1, 4B-2 '4C-1 and 4CM2 The voltage drop between the base channel 432a and 432b and the control gate 4258 or 4251) between 422 is set to one approximately + 20·0V during the channel erasing period of ^?0|161^01" In Fig. 12A, the selected word line 450a or 450b', that is, the voltage of the control gate 425a or 425b, is set to a negative erase voltage of about -10·0 V. The drains 415 and 420, the source The voltages of 420 and 422, triple ρ well 43 〇 and deep WELL 435 are set to a positive erase voltage of approximately + 10. 0 V. The unselected word line 450a or 450b, that is, unselected control The voltage of gate 425a or 425b is set to a masking erase voltage of approximately + 10. 〇v. 099113777 Form No. 1010101 Page 53 of 133 Page 0992024278-0 201104843 [0107] In Figure 12B, the negative erase voltage Approximately -15 〇V, the positive erase voltage is approximately + 5. 0V, and the positive shield voltage is approximately + 5. 0V. In Figure 12c, the negative eraser is Approx. _2〇. 〇v, positive erase is about 0. 0V 'The positive blocking voltage is about 〇· 〇v. In Figure 12]), the voltage level is reversed, and the negative erase voltage is about 〇 〇v, the positive erase voltage is approximately + 20.0V. Each of the voltages in Figures 12A-12D produces a Fowler-Nordheim channel tunneling effect to reduce the threshold voltage of the selected _〇s NAND flash floating gate transistor 4〇5a or 405b. [0108] The unselected double transistor floating gate NMOS NAND fast interrogation units of FIGS. 4A, 4B-1, 4B-2, 4C-1, and 4C-2 do not share the same triple p well 43 〇 and deep N well 435 . The voltage of the unselected word line 450a or 450b, i.e., the control gate 425a or 425b, the drains 415 and 420, the sources 420 and 422, and the triple P well 430 are set to be approximately equal to the ground reference voltage. The voltage of the deep N well 435 is set to the voltage of the voltage source VDD. [0109] For a sub-array in an array of floating gate NMOS NAND flash cells (often 512Kb or 4Kb.), the voltage of the deep N well of the un-erased sub-array is set to + At 20V, the voltage of the word line, drain, source, and triple P diffusion well is set to the ground reference voltage. The voltages of the word lines, the wave poles, the source poles, the = heavy P wells, and the deep N diffusion wells of the sub-arrays that are not selected in the different deep N wells are set to the ground reference voltage. [〇11〇] The erasing operation when another erase and programming threshold voltage is inverted is discussed in Fig. 12E. In this case, the selected word line 45A or 450b, that is, the voltage of the control gate 425a or 425b is set to a positive programming voltage of about +20. (^). The control gate 4 2 5 a Or 4 2 5 b, no pole 415 and 4 2 〇, source 420 and 422, triple P well 430 voltage is set to ground reference 099113777 Form No. A0101 Page 54 / 133 pages 201104843 Pressure (0. OV) The voltage of the deep N well 435 is set to the voltage of the voltage source. The conditions for setting the erased threshold voltage to the positive voltage and setting the programmed threshold voltage to the negative voltage are as shown in FIGS. 7A-7D. Figure 13A and Figure 13B are the programming biases for programming the dual transistor floating gate NMOS NOR flash cell in Figure 4Α, 4Β-1, 4Β-2, 4C-1, and 4C-2. The selected _〇s NAND flash floating gate transistor 405a or 405b in the dual transistor floating gate NMOS NAND flash cell in 4A, 4B-1, 4B_2, 4C-1, and 4C-2 is programmed before the cell is programmed. Must be erased. In an array of a double transistor floating gate NMOS NAND flash cell as shown in Figure 8, the erase operation is Performed on a page or a block of cells. [0112] The NM0S NAND flash floating gate transistor 405a selected from FIGS. 4A, 4B-1, 4B-2, 4C-1, and 4G-2 When the programming operation is performed by 405b, the selected word line 450a or 450b, that is, the voltage of the control gate 425a or 425b is set to a positive programming voltage of about + 15 > 0.V to + 20: 0V. The voltages of 415 and 420, _ poles 420 and 422, and channel regions 432a and 432b are freshly set to ground reference voltage (0. 0 V) through triple P well 430. Unselected NMOS NAND flash floating gate transistor 4 〇 5a Or word line 450a or 450b of 405b is coupled to its control gate 425a or 425b to set its voltage to a masking programming voltage of approximately +0.5 V. Within an array as shown in Figure 8, located in the selected character The drain and source voltages of the unselected floating gate NMOS NAND flash cells on line 450a or 450b are set to a positive programming masking voltage from about +7.00V to about +10 〇v. In the array shown in 8, the common bit line 455a, 455b having the positive blocking voltage and the source line 460a, 460b are 099113777 Form No. A 0101 Page 55 of 133 0992024278-0 201104843 The voltage of word lines 450a and 450b of the unselected floating gate NMOS NAND flash cell is set to a positive blocking programming voltage of + 5. 0V. The voltages of the unselected floating gate NMOS NOR flash cells that are not connected to the positive programming voltage or the positive programming masking voltage are not included in the word lines 450a, 450b or the bit lines 455a, 455b or the source lines 460a, 46 Ob. Set to ground reference voltage (O.OV). It is conventionally known that the higher the positive programming voltage applied to the control gate 425a or 425b, the higher the threshold voltage Vt after the programming operation. During the programming operation, in order to maintain the precise control of the threshold voltage of the NM0S NAND flash cell, the gate is applied with an initial positive programming from approximately + 15.0V to approximately + 16.0V. The positive programming voltage is incremented by a small amount over time. The above programming voltage is suitable for programming a single-order potential unit or a multi-level potential unit. The threshold voltage is shown in Figures 6A-6D. [0114] As the gate voltage of the selected programming cell in the selected block and the voltage of the floating source gradually increase by a small amount of negative gate voltage, this is an iterative programming operation and a program verifying step. For example, the drain (BL of ground) voltage is coupled to a fixed +5V and the local SL is floating. Fig. 8F is a preferred bias condition for programming a selected cell M0. A gate voltage of -10 V is applied to WL0 of the selected cell M0. The -10V gate voltage can start from -5V and then gradually drop to -10V. In other words, the voltage value Vt of the unit can be accurately controlled to a desired voltage value. Referring to Figure 13B, the programming voltage for the reverse programming and erase conditions as shown in Figures 7A-7B is depicted. In this embodiment, the voltage of the selected word line 450a or 450b of the selected NMOS NAND flash floating gate transistor 405a or 405b is set to about -10. ov of the negative programming voltage bungee 099113777 Form Compilation A0101 Page 56 of 133 pages 0992024278-0 201104843 Ο [0115]

[0116] G[0116] G

[0117] 415和420的電壓逐漸下降到一大約+ 5. ον的正汲極電壓 。源極420的電壓則停止浮動。該被選擇的NM〇s N〇R快 閃單元反復地被編程操作和被檢驗,以使得在編程操作 之後可精確地達到單元臨界電壓。本實施例中,編程條 件是基於FoWler-Nordheim邊界隧道編程操作。普遍的 FN邊界編程操作用於減少在編程之後所被選擇單元的電 壓Vt。然,被選擇的編程單元在㈣邊界編程之後的最後 電壓Vt—定要為正值,以避免由於bl透過被選擇的區塊 中未被選擇單元的滲漏而被誤讀。FN邊界發生在本發明 被選擇區塊中被選擇的NAND單元的汲極點和閘極點之間 的邊界處。 另,負極編程電壓可以從大約-7.〇V逐漸增加到大約- 10· Ον »中間的正極汲極電壓被固定在大約+ 5· 〇v »本實 施例中,每一反復步驟以大約〇. 3V逐漸碎負極編程電壓 〇 將未被選擇的字元線i5〇a或450b的電屬設置到一大約 + 2.5V的正極遮薇電義,以阻止該未被選擇的NMOS NAND 快閃浮動閘電晶體405a或405b不被編程操作。該未被選 擇的NMOS NAND快閃浮動閘電晶體405a或405b的汲極 415和三重P井430的電壓被設置到接地參考電壓(o.ov) ,該深N井435的電壓則被設置到電壓源VDD。 該被選擇的浮動閘NMOS NOR快閃單元的浮動間中的電子 從浮動閘445a或445b被驅逐。因此被選擇的浮動閘NM0S NOR快閃單元的臨界電壓能在單階單元和多階單元裡十分 精確地被控制。 099113777 表單編號A0101 第57頁/共133頁 0992024278-0 201104843 [0118] 第14圖是應用本發明形成一NOR快閃記憶體元件的流程圖 。浮動閘電晶體的一陣列在一基板上被形成(如方框905 所示)。該浮動閘電晶體被安排在一由行和列構成的矩陣 中。在一直行中串連連接至少兩個相鄰的浮動閘電晶體( 方框910),以形成一NOR記憶體單元θΝΑΝΙ)串。在以 NAND為基礎的NOR快閃記憶體單元的每一直行中的最上端 的浮動閘電晶體的汲極被連接到一相應的位元線(方框 915)。以NAND為基礎的NOR快閃記憶體單元的每一直行 中的最下端的浮動閘電晶體的源極被連接到一相應的源 極線(方框920) 〇 [0119] 本地的位元線透過一最上端的位元線選擇電晶體被連接 到一相應的全域位元線(方柩% 5 )»讓最上端的位元線選 擇電晶體的源極連接到本地的位元線,該最上端的位元 線選擇電晶體的沒極連接到全域位元線^該本地的源極 線透過一最下端的源極線選擇電晶體被連接到一相應的 全域源極線(方框930)。該最下端的滅&線選擇電晶體的 ·... 源極連接到本地的源極線'該最下端的源極線選擇電晶 體的汲極連接到全域源極線。 [0120] 一位元線閘極選擇控制線被連接到最上端的位元線選擇 電晶體的閘極(方棍935)。-源極線閘極選擇控制線被連 接到下端的源極線選擇電晶體的閘極(方框94〇)。在以 NAND為基礎的NOR快閃記憶體陣列的每一橫列中,每一浮 動閘電晶體的控制閘極連接到一相應的字元線(方框945) 。每一浮動閘電晶體橫列中的每—字元線被連接到一字 兀線電壓控制器(方框950) ’以提供對以謂㈣基礎的 099113777 表單編號A0101 第58頁/共133頁 201104843 峨快閃e’lt體陣列進行編轉作、抹除操作和讀取操作 所需要的偏壓。每-位元線選擇控制線被連接到_位元 線選擇控制器(方框955),以使得位元線選擇電晶體可以 有選擇地連接-被選擇的本地的位元㈣—全域位元線 同樣地,每-源極線選擇控制線被連接到一源極線選 擇控制gs (方框960),以使得源極線選擇電晶體可以有選 擇地連接-被選擇的本地的源極線到_全域源極線。[0117] The voltages of 415 and 420 gradually drop to a positive drain voltage of approximately + 5. ον. The voltage of source 420 stops floating. The selected NM〇s N〇R flash cell is repeatedly programmed and tested so that the cell threshold voltage can be accurately reached after the programming operation. In this embodiment, the programming conditions are based on the FoWler-Nordheim boundary tunnel programming operation. A common FN boundary programming operation is used to reduce the voltage Vt of the selected cell after programming. However, the final voltage Vt of the selected programming unit after (4) boundary programming is to be positive to avoid being misread due to leakage of bl through the unselected cells in the selected block. The FN boundary occurs at the boundary between the drain point and the gate point of the selected NAND cell in the selected block of the present invention. In addition, the negative programming voltage can be gradually increased from about -7. 〇V to about -10· Ο » » The middle positive drain voltage is fixed at about + 5 · 〇v » In this embodiment, each iterative step is about 〇 3V gradually breaks the negative programming voltage 〇 sets the power of the unselected word line i5〇a or 450b to a positive illuminating power of about +2.5V to prevent the unselected NMOS NAND from flashing. Gate transistor 405a or 405b is not programmed to operate. The voltages of the drain 415 and the triple P well 430 of the unselected NMOS NAND flash floating gate transistor 405a or 405b are set to the ground reference voltage (o.ov), and the voltage of the deep N well 435 is set to Voltage source VDD. The electrons in the floating compartment of the selected floating gate NMOS NOR flash cell are evicted from the floating gate 445a or 445b. Therefore, the threshold voltage of the selected floating gate NM0S NOR flash cell can be controlled very accurately in single-order cells and multi-level cells. 099113777 Form No. A0101 Page 57 of 133 0992024278-0 201104843 [0118] Figure 14 is a flow chart for forming a NOR flash memory component using the present invention. An array of floating gate transistors is formed on a substrate (as indicated by block 905). The floating gate transistor is arranged in a matrix of rows and columns. At least two adjacent floating gate transistors are connected in series (block 910) in a row to form a NOR memory cell θ ΝΑΝΙ string. The drain of the uppermost floating gate transistor in each of the NAND-based NOR flash memory cells is connected to a corresponding bit line (block 915). The source of the lowermost floating gate transistor in each row of the NAND-based NOR flash memory cell is connected to a corresponding source line (block 920) 〇[0119] local bit line The transistor is connected to a corresponding global bit line (square 柩% 5 ) through an uppermost bit line selection layer » the source of the uppermost bit line selection transistor is connected to the local bit line, the uppermost The bit line select transistor has a gate connected to the global bit line. The local source line is coupled to a corresponding global source line through a lowermost source line select transistor (block 930). The lowermost & line selects the transistor's ... source connected to the local source line'. The lowermost source line selects the drain of the transistor to be connected to the global source line. [0120] A one-bit gate select control line is connected to the uppermost bit line select gate of the transistor (square stick 935). The source line gate selection control line is connected to the gate of the lower source line selection transistor (block 94A). In each row of a NAND-based NOR flash memory array, the control gate of each floating gate transistor is coupled to a corresponding word line (block 945). Each word line in each floating gate transistor row is connected to a word line voltage controller (block 950)' to provide a base for the base (4). Form number A0101 Page 58 of 133 201104843 峨 flashing e'lt body array for the bias required for the splicing, erasing and reading operations. The per-bit line select control line is coupled to the _ bit line select controller (block 955) such that the bit line select transistor can be selectively connected - the selected local bit (4) - global bit Similarly, the per-source line select control line is coupled to a source line select control gs (block 960) such that the source line select transistor can be selectively coupled to the selected local source line. To the _ global source line.

[0121] G 每-全域位it線和直行位元線被連接到—直行電壓控制 器(方框965)。如上所述,該字元線電壓控制器和直行電 壓控制器用於提供適當的電壓給以NAND為基礎的N〇R快 閃記憶體單元,α對該N 〇 R快閃記憶體單纟進行編程操作 、抹除操作和讀取操作》 , [0122] Ο 圖15是一以NAND為基礎的多電晶體浮動閘NM〇s n〇r快閃 記憶體陣列的一具體實施方式的示意圖。在以NAND為基 礎的NMOS NOR快閃記憶體陣列的第8圖中,每一浮動閘 NMOS NOR快閃單元包括兩個浮動閘電晶體。在第15圖中 ’每一浮動閘NMOS NOR快閃單元1;,〇〇5中浮動閘電晶體 1010a、l〇l〇b、…、l〇i〇n中的至少兩個相互串連連接 ’就如第8圖所描述的雙電晶體串連的實施例。最上端的 浮動閘電晶體1010a的汲極連接到本地的位元線1015, 最下端的浮動閘電晶體1 〇 1 〇n的源極連接到本地的源極線 1 020。在以NAND為基礎的NMOS NOR快閃記憶體陣列裡 一相關的橫列上,每一字元線l〇25a、1 025b..... 1025η連接到浮動閘電晶體ιοί〇a、101 〇b.....101 On 的控制閘。單階單元在浮動閘NMOS NOR快閃單元裡所儲 099113777 表單編號A0101 第59頁/共133頁 0992024278-0 201104843 存的位元數目是每一電晶體具有一位元,如此浮動閘 NMOS N0R快閃單元就可被指定為〇位元/n電晶體單元。 在多階單元裡,位元數目取決於儲存在每一浮動閘電晶 體1010a、1010b、…、ι〇ι〇η中臨界電壓的數目。 [0123] 目則對NOR快閃把憶體元件技術的需求是讀取時間為大約 20uS到大約100ns之間。電晶體的數目確定了以NAND為 基礎的NOR快閃c憶體單元的性能。例如圖4A、、[0121] G-to-global bit it and straight bit lines are connected to a straight-through voltage controller (block 965). As described above, the word line voltage controller and the straight line voltage controller are used to supply an appropriate voltage to the NAND-based N〇R flash memory unit, and the N 〇R flash memory unit is programmed by α. Operation, Erase Operation, and Read Operation, [0122] Figure 15 is a schematic diagram of a NAND-based multi-crystal floating gate NM〇sn〇r flash memory array. In Figure 8 of a NAND-based NMOS NOR flash memory array, each floating gate NMOS NOR flash cell includes two floating gate transistors. In Fig. 15, 'each floating gate NMOS NOR flash unit 1; at least two of the floating gate transistors 1010a, l〇l〇b, ..., l〇i〇n in 〇〇5 are connected in series with each other 'An embodiment of the double transistor series as described in Fig. 8. The drain of the uppermost floating gate transistor 1010a is connected to the local bit line 1015, and the source of the lowermost floating gate transistor 1 〇 1 〇n is connected to the local source line 1 020. On a related NAND-based NMOS NOR flash memory array, each word line l〇25a, 1 025b.....1025η is connected to the floating gate transistor ιοί〇a, 101 〇 b.....101 On the control gate. The single-order unit is stored in the floating gate NMOS NOR flash unit. 099113777 Form No. A0101 Page 59 / 133 pages 0992024278-0 201104843 The number of bits stored is one bit per transistor, so the floating gate NMOS N0R is fast. The flash cell can be designated as a clamp/n transistor unit. In a multi-level cell, the number of bits depends on the number of threshold voltages stored in each of the floating gate transistors 1010a, 1010b, ..., ι〇ι〇η. [0123] The need for NOR flash to use the memory element technology is that the read time is between about 20 uS and about 100 ns. The number of transistors determines the performance of the NAND-based NOR flash memory unit. For example, Figure 4A,

4B-2、4C-1和4C-2中雙電晶體浮動閘NM〇s⑽以夬閃單 元的實;5&例中針對以從1 Gb到4 Gb容量的NAND為基礎的 NM〇S _快閃記憶想陣列的讀取時間大約為l〇〇nS。另 ’以1Mb到4Mb容量的NAND為基礎的龍⑽臓快閃記憶 體陣列的讀取時間為2〇ns,在--陣列中,隨機讀 二操作是以一位元組(8位元)、-字元(1雜元)或者一雙 子疋(32位το)為單位進行讀取操作;編程單位是以一頁 512位疋組或半頁256位元組為單位進行編程操作·抹除 操作是以區段為單位執行(-小區段4K位元組或_大區段 64K位元組)。口0 [0124] 099113777 在其它實施例中,以膽㈣基礎的_s N⑽快閃記憶 單疋有16個或32個電晶體相串連。對-個容量從1Gb到 32Gb的較長的排列的陣列的讀取時間減少到大約Mm 在該實施例巾’讀取操作是以半頁(256位元組)或一頁 (512位讀)的單位連續的讀取。同樣地,編程操作的 如全頁的512位元組或半頁的256位元組為單位進行《 程操作。抹除操作是以—區段51 te 組)或519“ ㈣位 ytes X 32 (16K位元組)為單位進行抹除^The double-crystal floating gates NM〇s(10) in 4B-2, 4C-1 and 4C-2 are in the form of strobe units; 5&s examples are based on NAND based on NAND from 1 Gb to 4 Gb capacity. Flash memory wants the array to read for approximately l〇〇nS. In addition, the read time of the Dragon (10) 臓 flash memory array based on NAND of 1Mb to 4Mb capacity is 2〇ns. In the array, the random read operation is a one-tuple (8-bit). , - character (1 hex) or a pair of 疋 (32 το) for reading operations; programming unit is a page of 512-bit 或 group or half-page 256-bit group for programming operations The divide operation is performed in units of sectors (-small section 4K bytes or _large section 64K bytes). Port 0 [0124] 099113777 In other embodiments, the _s N(10) flash memory unit based on the biliary (four) has 16 or 32 transistors in series. The read time for a longer array of capacities from 1Gb to 32Gb is reduced to approximately Mm. In this embodiment, the read operation is half page (256 bytes) or one page (512 bits read). The unit is continuously read. Similarly, a programming operation such as a full page of 512 bytes or a half page of 256 bytes is performed in units. The erase operation is erased in units of - segment 51 te group or 519 "four" bit ytes X 32 (16K bytes) ^

表單蝙鱿Α0101 第6〇頁/共133頁 0992024278-0 201104843 [0125] [0126] Ο [0127] Ο [0128] 作。 在各實施例中,以NAND為基礎的浮動閘NMOS NOR快閃記 憶體單元如前所述可以包含任何數目的電晶體。然,爲 保證符合目前對浮動閛NMOS NOR快閃記憶體單元性能的 要求,一般在一以NAND為基礎的浮動閘NMOS NOR快閃記 憶體單元中使用多達15個電晶體。 如上所述,以NAND為基礎的NMOS NOR快閃記憶體單元包 含浮動閘電晶體,該浮動閘電晶體用於儲存電荷《在N0R 快閃記憶體單元的每一NAND串中以浮動閘NMOS NOR快閃 記憶體單元為基礎的NAND均包括有S0N0S電荷操取NAND 電晶體。 一包括以N A N D為基礎的快閃記‘衝會單元螓列的積體電路 可以包括一NAND快閃記憶體陣列和使用本發明概念的以 NAND為基礎的NMOS NOR快閃記憶體單元陣列。以NAND 為基礎的NMOS NOR快閃記憶艟單元_列寸丨以更進一步與 易揮發性記憶髏結合以形成在單―積體:電路上的記憶功 忐。更進一步,以NAND為基礎的NMOS NOR快閃記憶體單 元可包括外圍的電路系統,以使以NAND為基礎的⑽⑽ NOR快閃記憶體單元適用於諸如pLj)或fpga。 综上所述,本發明符合發明專利要件,爰依法提出專利 申明。惟,以上所述者僅為本發明之較佳實施例,舉凡 熟悉本案技藝之人士 ’在爰依本發明精神所作之等效修 飾或變化,皆應涵蓋於以下之巾請專利範圍内。 【圖式簡單說明】 099113777 表單編號A0101 第61頁/共133 頁 0992024278-0 201104843 [0129] 圖1A是一單電晶體浮動閘匪OS NAND快閃單元的俯視圖 〇 [0130] 圖1B是一單電晶體浮動閘NMOS NAND快閃單元的剖視圖 〇 [0131] 圖1C是一單電晶體浮動閘NMOS NAND快閃單元的示意圖 〇 [0132] 圖1D是一單電晶體浮動閘NMOS NAND快閃單元兩個臨界 電壓的分配圖。 [0133] 圖1E是一單電晶體浮動閘NMOS NAND快閃單元四個臨界 電壓的分配圖。 [0134] 圖2A是一單電晶體浮動閘NMOS NOR快閃單元的俯視圖。 [0135] 圖2B是一單電晶體浮動閘NMOS NOR快閃單元的剖視圖。 [0136] 圖2C是一單電晶體浮動閘NMOS NOR快閃單元的示意圖。 [0137] 圖2D是一單電晶體浮動閘NMOS NOR快閃單元兩個臨界電 壓的分配圖。 [0138] 圖2E是一單電晶體浮動閘NMOS NOR快閃單元四個臨界 電壓的分配圖。 [0139] 圖3A是先前技術中一雙電晶體浮動閘NMOS NOR快閃單 元的俯視圖。 [0140] 圖3B是圖3A中的雙電晶體浮動閘NMOS NOR快閃單元的剖 視圖。 [0141] 圖3C是先前技術圖3A中的雙電晶體浮動閘NMOS NOR快閃 099113777 表單編號A0101 第62頁/共133頁 0992024278-0 201104843 單元的示意圖。 [0142] 圖3D是先前技術的一雙電晶體浮動閘NMOS NOR快閃單元 兩個臨界電壓的分配圖。 [0143] 圖3E是先前技術的一雙電晶體浮動閘NMOS NOR快閃單 元四個臨界電壓的分配圖。 [0144] 圖4A是本發明雙電晶體浮動閘NMOS N0R快閃單元的一實 施例的示意圖。Form 鱿Α0101 Page 6 of 133 pages 0992024278-0 201104843 [0126] 0 [0127] Ο [0128]. In various embodiments, the NAND-based floating gate NMOS NOR flash memory cell can include any number of transistors as previously described. However, in order to ensure compliance with current performance requirements for floating NMOS NMOS NOR flash memory cells, up to 15 transistors are typically used in a NAND-based floating gate NMOS NOR flash memory cell. As described above, the NAND-based NMOS NOR flash memory cell includes a floating gate transistor for storing charge "Floating gate NMOS NOR in each NAND string of the NOR flash memory cell. Flash memory cell-based NANDs all include a S0N0S charge operation NAND transistor. An integrated circuit comprising a flash memory unit based on N A N D may comprise a NAND flash memory array and a NAND-based NMOS NOR flash memory cell array using the inventive concept. The NAND-based NMOS NOR flash memory unit is further integrated with volatile memory to form a memory function on a single-integrated circuit. Furthermore, the NAND-based NMOS NOR flash memory cell can include peripheral circuitry to enable NAND-based (10) (10) NOR flash memory cells for applications such as pLj or fpga. In summary, the present invention complies with the requirements of the invention patent, and proposes a patent declaration according to law. However, the above description is only the preferred embodiment of the present invention, and equivalent modifications or variations made by those skilled in the art of the present invention in the spirit of the present invention should be covered by the following claims. [Simple description of the drawing] 099113777 Form No. A0101 Page 61 of 133 Page 0992024278-0 201104843 [0129] FIG. 1A is a top view of a single transistor floating gate OS NAND flash unit [0130] FIG. 1B is a single Sectional view of a transistor floating gate NMOS NAND flash cell 013 [0131] FIG. 1C is a schematic diagram of a single transistor floating gate NMOS NAND flash cell [0132] FIG. 1D is a single transistor floating gate NMOS NAND flash cell A distribution map of threshold voltages. [0133] FIG. 1E is a distribution diagram of four threshold voltages of a single transistor floating gate NMOS NAND flash cell. 2A is a top plan view of a single transistor floating gate NMOS NOR flash cell. 2B is a cross-sectional view of a single transistor floating gate NMOS NOR flash cell. 2C is a schematic diagram of a single transistor floating gate NMOS NOR flash unit. 2D is a distribution diagram of two threshold voltages of a single transistor floating gate NMOS NOR flash cell. 2E is a distribution diagram of four threshold voltages of a single transistor floating gate NMOS NOR flash cell. [0139] FIG. 3A is a top plan view of a dual transistor floating gate NMOS NOR flash unit of the prior art. [0140] FIG. 3B is a cross-sectional view of the dual transistor floating gate NMOS NOR flash cell of FIG. 3A. 3C is a schematic diagram of a unit of the dual transistor floating gate NMOS NOR flash in the prior art FIG. 3A. 099113777 Form No. A0101 Page 62 of 133 0992024278-0 201104843 Unit. 3D is a distribution diagram of two threshold voltages of a dual transistor floating gate NMOS NOR flash cell of the prior art. 3E is a distribution diagram of four threshold voltages of a prior art dual transistor floating gate NMOS NOR flash unit. 4A is a schematic diagram of an embodiment of a dual transistor floating gate NMOS NOR flash cell of the present invention.

[0145] 圖4B-1、4B-2、4C-1和4C-2是本發明雙電晶體浮動閘 NMOS NOR快閃單元的一實施例的俯視圖和剖視圖。 [0146] 圖5 A-5E是本發明雙電晶體浮動閘NMOS NOR快閃單元陣 列中的一段的連接示意圖。 [0147] 圖6 A-6D是本發明單電晶體浮動閘NMOS NOR快閃單元各 實施例的臨界電壓圖。 [0148] 圖7A-7D是本發明雙電晶體浮動閘NMOS NOR快閃單元其 它各實施例的臨界電壓圖。 [0149] 圖8是一包含本發明各實施例中雙電晶體浮動閘NMOS NOR快閃單元的NOR快閃記憶體元件的示意圖。 [0150] 圖9是圖8中N 0 R快閃記憶體的橫列電壓控制電路的電路 圖。 [0151 ] 圖1 0是圖8中N 0 R快閃記憶體的直行電壓控制電路的電路 圖。 [0152] 圖11A是本發明雙電晶體浮動閘NMOS NOR快閃單元中的 099113777 表單編號A0101 第63頁/共133頁 0992024278-0 201104843 一單階電位編程電壓追隨感應電路的示意圖。 [0153] 圖11B是本發明雙電晶體浮動閘NMOS N0R快閃單元中單 階電位編程讀取偏壓表。 [0154] 圖11C是本發明雙電晶體浮動閘NMOS N0R快閃單元中一 多階電位編程電壓追隨感應電路的示意圖。 [0155] 圖11D是本發明雙電晶體浮動閘NMOS NOR快閃單元中多 階電位編程讀取偏表。 [0156] 圖12A-12E是本發明雙電晶體浮動間NMOS N0R快閃單元 的抹除偏壓表。 [0157] 圖1 3A-1 3B是本發明雙電晶體浮動閘NMOS NOR快閃單元 的編程偏壓表》 [0158] 第14A-14B圖是形成NOR快閃記憶體的流程圖。 [0159] 第15圖是本發明多電晶體浮動閘NMOS NOR快閃單元的一 實施例的示意圖。 【主要元件符號說明】 [0160] NMOS NAND快閃浮動閘電晶體:1 0 [0161] NMOS NOR快閃浮動閘電晶體:110 [0162] 汲極:15、115、215a、215b、415 [0163] 源極:20、120、220、422 [0164] 控制閘:25、125、225a、225b、425a、425b [0165] P井:30、130、430 099113777 表單編號A0101 第64頁/共133頁 0992024278-0 201104843 [0166]通道區:32、132、432a、432b [0167] 深N井:35、135、435 [0168] P型基板:40、140、240、440 [0169] 浮動閘:45、145、245a、245b、445a、445b [0170] 隧道氧化物:50、150 [0171] 雙電晶體浮動閘NMOS NOR快閃單元:210 [0172] 浮動閘電晶體:205a、205b 〇 [0173] 基極· 2 3 0 〇、2 3 0.b ^ ' [0174] 金屬觸點:250a、250b [0175] 金屬位元線:255 [0176] NMOS NOR快閃記憶體單元:400 [0177] NMOS NAND快閃浮動閘電晶體,.:40.5a、40:5 b、515a、 i . i .·'"· : \, i s' .: g J !| 515b rfVVi:. 〇 [0178] 源極/汲極: 420 i [0179] 字元線:450a、450b、1 025a、1025b...l 025n、545 [0180] 位元線:455a、455b、520a、520b…520n、525a、 525b…525n、805、815、1015 [0181] 源極線:460a、460b、540a、540b…540n、530a、 530b…530n、825、1020 [0182] NOR快閃記憶體元件:500 099113777 表單編號A0101 第65頁/共133頁 0992024278-0 201104843 [0183] 雙電晶體浮動閘NMOS NOR快閃單元:510 [0184] 陣列:505 [0185] 橫列電壓控制電路:5 5 0 [0186] 位元線選擇控制子電路:551 [0187] 字元線電壓控制子電路:552 [0188] 源極線選擇控制子電路:553 [0189] 直行電壓控制電路:555 [0190] 位元線選擇電晶體:560a、560b…560η、810 [0191] 源極線選擇電晶體:565a、565b…565η、830 [0192] 位元線選擇訊號:57Ga..... 570η [0193] 源極線選擇訊號:575a、575b [0194] 控制解碼器:605、705 [0195] 編程時序和控制訊號:610、710 .. ......4B-1, 4B-2, 4C-1, and 4C-2 are top and cross-sectional views of an embodiment of a dual transistor floating gate NMOS NOR flash cell of the present invention. 5A-5E are connection diagrams of a segment of the dual transistor floating gate NMOS NOR flash cell array of the present invention. 6A-6D are threshold voltage diagrams of various embodiments of a single transistor floating gate NMOS NOR flash cell of the present invention. 7A-7D are threshold voltage diagrams of other embodiments of a dual transistor floating gate NMOS NOR flash cell of the present invention. 8 is a schematic diagram of a NOR flash memory component including a dual transistor floating gate NMOS NOR flash cell in accordance with various embodiments of the present invention. 9 is a circuit diagram of a course voltage control circuit of the N 0 R flash memory of FIG. 8. 10 is a circuit diagram of a straight-line voltage control circuit of the N 0 R flash memory of FIG. 8. 11A is a schematic diagram of a single-stage potential programming voltage following induction circuit in the double transistor floating gate NMOS NOR flash unit of the present invention. 099113777 Form No. A0101 Page 63 of 133 0992024278-0 201104843 11B is a single-order potential programming read bias table in the dual transistor floating gate NMOS NOR flash cell of the present invention. 11C is a schematic diagram of a multi-level potential programming voltage following sensing circuit in the dual transistor floating gate NMOS NOR flash unit of the present invention. 11D is a multi-level potential programming read offset table in a dual transistor floating gate NMOS NOR flash cell of the present invention. 12A-12E are eraser bias tables of the dual transistor floating-to-space NMOS NOR flash cell of the present invention. 1A-1 3B are programming bias tables of the dual transistor floating gate NMOS NOR flash cell of the present invention. [0158] FIGS. 14A-14B are flowcharts showing the formation of a NOR flash memory. Figure 15 is a schematic illustration of an embodiment of a polymorphic floating gate NMOS NOR flash cell of the present invention. [Main component symbol description] [0160] NMOS NAND flash floating gate transistor: 1 0 [0161] NMOS NOR flash floating gate transistor: 110 [0162] Bungee: 15, 115, 215a, 215b, 415 [0163] Source: 20, 120, 220, 422 [0164] Control gate: 25, 125, 225a, 225b, 425a, 425b [0165] P well: 30, 130, 430 099113777 Form number A0101 Page 64 of 133 0992024278-0 201104843 [0166] Channel area: 32, 132, 432a, 432b [0167] Deep N well: 35, 135, 435 [0168] P-type substrate: 40, 140, 240, 440 [0169] Floating gate: 45 , 145, 245a, 245b, 445a, 445b [0170] Tunnel oxide: 50, 150 [0171] Double transistor floating gate NMOS NOR flash unit: 210 [0172] Floating gate transistor: 205a, 205b 〇 [0173] Base · 2 3 0 〇, 2 3 0.b ^ ' [0174] Metal contact: 250a, 250b [0175] Metal bit line: 255 [0176] NMOS NOR flash memory unit: 400 [0177] NMOS NAND flash floating gate transistor, .: 40.5a, 40:5 b, 515a, i . i .·'"· : \, is' .: g J !| 515b rfVVi:. 〇[0178] Source /Bungee: 420 i [0179] Word line: 4 50a, 450b, 1 025a, 1025b...l 025n, 545 [0180] Bit lines: 455a, 455b, 520a, 520b...520n, 525a, 525b...525n, 805, 815, 1015 [0181] Source line: 460a, 460b, 540a, 540b...540n, 530a, 530b...530n, 825, 1020 [0182] NOR flash memory component: 500 099113777 Form number A0101 Page 65 of 133 page 0992024278-0 201104843 [0183] Crystal Floating Gate NMOS NOR Flash Unit: 510 [0184] Array: 505 [0185] Horizontal Voltage Control Circuit: 5 5 0 [0186] Bit Line Select Control Subcircuit: 551 [0187] Word Line Voltage Control Subcircuit :552 [0188] Source line selection control subcircuit: 553 [0189] Straight line voltage control circuit: 555 [0190] Bit line selection transistor: 560a, 560b...560η, 810 [0191] Source line selection transistor: 565a, 565b...565η, 830 [0192] Bit line selection signal: 57Ga..... 570η [0193] Source line selection signal: 575a, 575b [0194] Control decoder: 605, 705 [0195] Programming timing And control signals: 610, 710 .. ......

[0196] 抹除時序和控制訊號:615、715 [0197] 讀取時序和控制訊號:620、720 [0198] 位址解碼器:625、725 [0199] 位址訊號:630、730 [0200] 編程電壓產生器:635、735 [0201] 脈衝增大電壓產生器:636 0992024278-0 099113777 表單編號A0101 第66頁/共133頁 201104843 [0202] 正極編程電壓產生器:637 [0203] 負極編程電壓產生器:638 [0204] 接地參考電壓源:639、644、737、743 [0205] 抹除電壓產生器:640、740 [0206] 正極抹除電壓產生器:642 [0207] 負極抹除電壓產生器:643 [0208] 讀取電壓產生器:645、745 〇 ^ [0209] 第一高讀取電壓產生器:646 [0210] 第二高讀取電壓產生器:647 [0211] 第三高讀取電壓產生器:648 [0212] 電廢源產生器:649 [0213] 橫列選擇開關:650 [0214] 直行選擇開關:750 ❹ [0215] 編程電壓源:736 [0216] 抹除電壓源:742 [0217] 適中的高讀電壓源:747 [0218] 感應放大器:755 [0219] 源極線電容:845 [0220] 比較電路:850、860、870、880 099113777 表單編號A0101 第67頁/共133頁 0992024278-0 201104843 [0221] 參考電壓源:865、875、885、855 [0222] 浮動閘NMOS NOR快閃單元:1005 [0223] 浮動閘電晶體:1010a、1010b…1010η 099113777 表單編號A0101 第68頁/共133頁 C) 0992024278-0[0196] Erase Timing and Control Signals: 615, 715 [0197] Read Timing and Control Signals: 620, 720 [0198] Address Decoder: 625, 725 [0199] Address Signals: 630, 730 [0200] Programming voltage generator: 635, 735 [0201] Pulse increasing voltage generator: 636 0992024278-0 099113777 Form number A0101 Page 66 / 133 pages 201104843 [0202] Positive programming voltage generator: 637 [0203] Negative programming voltage Generator: 638 [0204] Ground reference voltage source: 639, 644, 737, 743 [0205] Erase voltage generator: 640, 740 [0206] Positive erase voltage generator: 642 [0207] Negative erase voltage generation Device: 643 [0208] Read Voltage Generator: 645, 745 〇 ^ [0209] First High Read Voltage Generator: 646 [0210] Second High Read Voltage Generator: 647 [0211] Third High Read Take voltage generator: 648 [0212] Electric waste source generator: 649 [0213] Horizontal selection switch: 650 [0214] Straight line selection switch: 750 ❹ [0215] Programming voltage source: 736 [0216] Erase voltage source: 742 [0217] Moderate high read voltage source: 747 [0218] sense amplifier: 755 [0219] Pole capacitance: 845 [0220] Comparison circuit: 850, 860, 870, 880 099113777 Form number A0101 Page 67 / 133 page 0992024278-0 201104843 [0221] Reference voltage source: 865, 875, 885, 855 [0222] Floating Gate NMOS NOR Flash Unit: 1005 [0223] Floating Gate Crystal: 1010a, 1010b...1010η 099113777 Form No. A0101 Page 68 of 133 C) 0992024278-0

Claims (1)

201104843 Ο G 七、申請專利範圍: •—種N0R快閃記憶體電路,包括: =數電荷保存電晶體連續地連接成-NAND串,其中: -最上端的電荷保存電晶體,其汲極連制與該等 接的電荷保存電晶體有關的一位元線; —最下端的電荷保存電晶體,其祕連制錢等連 接的電荷保存電晶體有關的一源極線;以及 ^ 該等電荷保存電晶體的每一控制閘均連接到一字元線。 如申清專利範圍第1項所瑪之臟快閃記憶體電路, 該等電荷保存電晶體形成於_第—類導電型的井内。、 .如申4專利範圍第2項所述之碰、快__電路 i第類導電型井形成於一率二類導奪型_井内 .如申清專利範圍第3項所述之N0R快閃贫靈體電路, 該第二類導電型深井形成於一第一類導電型的基板中。 .如申清專利範圍第1項所述之蠢快閃記解電路,其中 S等電荷保存電晶體採用穿隧效應進 行編程操作和抹除操作。 .如申请專利範圍第1項所述之N〇R快閃記憶體電路,其中 一+ 15· 0V到+ 20. 0V的編程電壓被施加於一被選擇的電荷 保存電的體的控制間和基極之間,以對該等電荷保存電晶 體中的-個破選擇的電荷保存電晶體進行編程操作,其中 該編程電廢以逐漸增加的方式被施加於該被選擇的電荷保 存電晶體的控制閘和基極之間。 •如申清專利範圍第6項所述之N〇R快閃記憶趙電路,其中 一小於10· 0V的電壓被施加於該等電荷保存電晶體中其他 其中 其中 099113777 表單編號A0101 第69頁/共133頁 0992024278-0 201104843 未被選擇的電荷保存電晶體的控制閘和基極之間,以遮蔽 該等未被選擇的電荷保存電晶體。 8 .如申請專利範圍第1項所述之NOR快閃記憶體電路,其中 該N0R快閃記憶體電路的佈局要求該N0R快閃記憶體電 路的大小是製造N0R快閃記憶體電路製程技術的最小的 特性尺寸的四倍到六倍。 9 .如申請專利範圍第1項所述之N0R快閃記憶體電路,其中 施加一+ 15. Ον到+ 20. 0V的正極抹除電壓於被選擇的電荷 保存電晶體的基極和控制閘之間,以對該被選擇的電荷保 存電晶體進行抹除操作。 C1 10 .如申請專利範圍第6項所述之NOR快閃記憶體電路,其中 一0. 0V的偏壓被施加於該等電荷保存電晶體中其他未被 選擇的電荷保存電晶體的控制閘和基極之間,以遮蔽該等 未被選擇的電荷保存電晶體。 11 .如申請專利範圍第1項所述之NOR快閃記憶體電路,其中 以一單階編程單元進行編程操作的該等電荷保存電晶體中 被選擇的電荷保存電晶體執行讀取操作時,包括: 連接源極線到一電壓追隨感應電路; U 設置該N 0 R快閃記憶體電路内被選擇的電荷保存電晶體 的汲極和閘極的電壓為電壓源的電壓; 該等電荷保存電晶體中未被選擇的電荷保存電晶體的閘極 的電壓被設定為一第一高讀取電壓;以及 比較該電壓追隨感應電路中在源極線產生的電壓和參考電 壓源,該參考電壓源被設置到2. 0V,以區分一在第一邏 輯值的臨界電壓和一在第二邏輯值的臨界電壓。 12 .如申請專利範圍第11項所述之NOR快閃記憶體電路,其 099113777 表單編號 A0101 第 70 頁/共 133 頁 0992024278-0 201104843 中該第一高讀取電壓大於6.〇v。 13 .如申請專利範圍第丨丨項所述之N〇R快閃記憶體電路,其 中該參考電壓源為2, 〇V。 ’、 14 .如申請專利範圍第丨丨項所述之N〇R快閃記憶體電路,其 中若N0R快閃記憶體電路未被選擇讀取操作,而當另— N0R快閃記憶體電路被選擇讀取操作時,未被選擇的該 N0R快閃記憶體電路中的該等電荷保存電晶體中未被選 擇的該電荷保存電晶體的控制閘的電壓被設置為接地參考 電壓,以關閉該電荷保存電晶體。 〇 15 .如申請專利範圍第1項所述之N0R快閃記憶體電路,其中 以一多階編程單元進行編程操作的該等電荷保存電晶趙中 被選擇的電荷保存電晶體執行讀取操作時,包括·· 連接源極線到一電廢追隨感應電路; 設置該被選擇的電荷保存電晶體的汲極和閘極的電壓為一 中間的高電壓; 將該等電荷保存電晶體之内·所有未被選蘀的電荷保存電晶 體的閘極的電壓設置為一第二高讀取電壓;以及 〇 比較該電壓追隨感應電路中在源極線產生的電壓和複數參 考電壓源,以決定一臨界電壓值用於代表儲存於電荷保存 電晶體内的數據。 16 .如申請專利範圍第15項所述之N〇R快閃記憶體電路’其 中該中間的高電壓是+ 4.〇V ° 17 .如申請專利範圍第15項所述之麵快閃記慎艘電路’其 中該第二高讀取電壓大於7.0V 。 18 .如申請專利範圍第15項所述之N0R快閃記憶體電路,其 中該參考電壓源被設置在每一臨界電壓之間以區别儲存於 0992024278-0 099113777 表單編號A0101 第71頁/共133頁 201104843 電荷保存電晶體内的數據的臨界電壓。 19 .如申請專利範圍第15項所述之N0R快閃記憶體電路,其 中若N0R快閃記憶體電路未被選擇讀取,而當另一N0R 快閃記憶體電路被選擇讀取時,該未被選擇的NOR快閃 記憶體電路中的該等電荷保存電晶體中未被選擇的電荷保 存電晶體的控制閘的電壓被設置到接地參考電壓,以關閉 該電何保存電晶體。 20 . —種N0R快閃記憶體,包括: 一包括複數N0R快閃記憶體電路的陣列,該陣列被安排 成橫列和直行,其中每一快閃記憶體電路包括: €1 每一直行内的複數電荷保存電晶體被串聯連接成一 NAND 串; 每一 N 0 R快閃記憶體電路中最上端的電荷保存電晶體的 汲極連接到與每一 N 0 R快閃記憶體電路所在的直行相對 應的一本地的位元線; 每一NOR快閃記憶體電路中最下端的電荷保存電晶體的 源極連接到與每一 N 0 R快閃記憶體電路所在的直行相對 應的一本地的源極線;以及_ D 每一橫列上的電荷保存電晶體的每一控制閘共同地連接到 一字元線。 21 .如申請專利範圍第20項所述之N0R快閃記憶體,其中該 位元線和源極線與該NOR快閃記憶體電路所在的直行相對 應並且是平行的。 22 .如申請專利範圍第20項所述之N0R快閃記憶體,還包含 一直行電壓控制電路,用於提供控制訊號到與每一直行電 荷保存電晶體相對應的本地的位元線和源極線。 099113777 表單編號A0101 第72頁/共133頁 0992024278-0 201104843 23 .如申請專利範圍第22項所述之卯只快閃記憶體,其t每 ~本地的位元線透過一位元線選擇電晶體連接到複數全域 位 元線中之 24 25 〇 26 27 〇 28 29 099113777 如申請專利範圍第23項所述之N0R快閃記憶體,其中每 一本地的源極線透過一源極線選擇電晶體連接到複數全域 源極線中之一, 如申請專利範圍第24項所述之Ν0Κ快閃記憶體,其t該 全域位元線和全域源極線連接到直行電壓控制電路,以傳 輪控制訊號到被選擇的本地的位元緣和被選擇的本地的源 極線’用於對NOR快閃記憶體電路中被選擇的電荷保存電 曰^曰體進行讀取操作、編程操作和抹除操作。 %中請專利範圍第20項所述之N0R快閃記憶體,還包含 —横列電壓控制電路,用於提供控制訊號到與每一橫列電 荷保存電晶體相對應的字元線。 如申請專利範圍第26項所述之N0R快閃記憶體,其中該 撗列控制電路傳輸控制訊號到字年線,以對該NOR快閃記 憶體電路巾㈣擇的電荷柄存電 晶體進行讀取操作、編程 操作和抹除操作。 申明專利fe圍第2〇項所述之快閃記憶體,還包含 一位70線選擇控制電路,連接所有本地的位元線選擇電晶 體的閘極;以及 複數源極線選擇電晶體,料祕㈣擇電㈣連接到每 一本地的位元線。 如申請專利範圍第20項所述之N〇R快閃記憶體,其中該 橫列控制電路用於傳輸字喊控制訊號到字元線,一 表單編號歷1 ^ 73 133 1 以對該 09920242, 201104843 30 _快閃記憶體電路中被選擇的 細作、編程操作和抹除操作,::電曰曰體進仃· 傳給你-⑶ 冊3㈣控制電路勒於分別 選擇錢及祕線選魏_賊擇的位Μ 選擇電晶體以及被選擇的源極線電晶體,以將位元線^ :線:制訊號從直行電壓控制電路傳輸到被選擇的本地: 几線和被選擇的本地的源極線。 〜 31 32 33 申料利範圍第20項所述之_快閃記憶體,其中該 —電何保存電晶體採用一 F〇wler_Lim穿随效應進 行編程操作和抹除操作。 .如申請專利範圍第20項所述之_快閃記憶體,其中一 15. 0Vfj + 2G,GV的編程電麼被施加於—被選擇的電荷保 存電曰曰體的控制閘和基極之間.,以對該被選擇的電荷保^ 電晶體進行編程操作。 子 .如申明專利範圍第31項所述之N〇R快閃記憶體,其中— 小於10. 0V的電壓被施加於該等電荷保存電晶體中其他未 被選擇的電荷保存電晶體的控制閘和基極之間,以遮蔽該 等未被選擇的電荷保存電晶體。 " 如申請專利範圍第20項所述之N0R快閃記憶體,其中該 N 0 R快閃記憶體電路佈局要求該N 〇 R快閃記憶體電路的大 小是製造NOR快閃記憶體電路製程技術的最小的特性尺寸 的四倍到六倍。 34 . 35 . 099113777 如申請專利範圍第20項所述之NOR快閃記憶體,其中施 加一+ 15. 0V到+ 2〇. 0V的正極抹除電壓於被選擇的電荷保 存電晶體的基極和控制閘之間,以對該被選擇的電荷保存 電晶體進行抹除操作。 如申請專利範圍第20項所述之NOR快閃記憶體,其中_ 表單編.號A0101 第74頁/共133頁 0992024278-0 201104843 36 . Ο 37 . 38 . ❹ 39 . 40 . 099113777 〇. 〇V偏壓被施加於該電荷保存電晶體群中其他未被選擇 的電荷保存電晶體雜制閘和基極之間,以遮蔽該等未被 選擇的電荷保存電晶體。 如申請專利範圍第20項所述之陶快閃記憶體,其中以 -單階編料元進行肺操作的該等電荷保存電晶體中被 選擇的電荷保存電晶體執行讀取操作時,包括. 連接源極線到_電壓追隨感應電路; 設置該_ 記憶體電路内被選擇的電荷保存電晶體 的汲極和閘極的電壓為電壓源的電壓; 將該等電荷保存電晶體中未被選擇的電荷保存電晶體的閉 極的電壓設置到第一高讀取電壓;以及 比較該電壓魏感應電路中在源極線產生的電壓和參考電 麼源’其中該參考電Μ被設置到2. 〇v,以區分一在第 -邏輯值的臨界電壓和-在第二邏輯值的臨界電刀壓。 如申請專利範圍第36項所述之_快閃記憶體,其中該 第一尚讀取電壓大於6 〇v。·: 如申請專利範詩36項所述之_怏閃 參考電壓源為2. 〇Γ。 " ’八中該 «專利扼圍第36項所述之NGR快閃記憶體, 元線及未被選擇的爾快閃記憶體電路的複仅罕 晶體中未被選擇的電荷保存電晶體的複 5r、子電 設置到接地參考《,以關電荷保存電電壓被 如申請專利__韻述之_ _ = -多階編程單元進行編程操作的該等電荷保^ ’、中从 選擇的電荷保存電晶體執行讀取操作時,包^晶體中被 連接源極線到一電壓追隨感應電路; . 表單編號第75頁/共133頁 0992024278-0 201104843 設置該被選擇的電荷保存電晶體的汲極和閘極的電壓為一 中間的高電壓; 將該等電4保存電晶體之内所有未被選擇的電何保存電晶 體的閘極的電壓設置為一第二高讀取電壓;以及 比較該電壓追隨感應電路中在源極線產生的電壓和複數參 考電壓源,以決定一臨界電壓值用於代表儲存於電荷保存 電晶體内的數據。 41 .如申請專利範圍第40項所述之N0R快閃記憶體,其中該 中間的高電壓是+4. 0V 。 42 .如申請專利範圍第40項所述之NOR快閃記憶體,其中該 第二高讀取電壓大於7.0V » 43 .如申請專利範圍第40項所述之NOR快閃記憶體,其中該 參考電壓源被設置在每一臨界電壓之間以區別儲存於電荷 保存電晶體内的數據的臨界電壓。 44 .如申請專利範圍第40項所述之NOR快閃記憶體,其中字 元線及未被選擇的NOR快閃記憶體電路的讓等電荷保存電 晶體中未被選擇的電荷保存電晶體的複數控制閘的電壓被 設置到接地參考電壓,以關閉電荷保存電晶體。 45 . —種形成NOR快閃記憶體的方法,包括以下步驟: 提供一基板;以及 將複數N 0 R快閃記憶體電路形成一包括橫列和直行的陣列 ,其中NOR快閃記憶體電路的形成包括如下步驟: 把複數電荷保存電晶體放置成直行和橫列; 線性連接直行中的複數電荷保存電晶體以形成一 N A N D串 連接每一NOR快閃記憶體電路中最上端的電荷保存電晶體 099113777 表單編號A0101 第76頁/共133頁 0992024278-0 201104843 46 . Ο 47 . 48 . ❹ 49 , 50 51 的汲極到與每一N〇L^閃記憶體電路所在的直行相對應的 一本地的位元線; 連接每一NOR快閃記憶體電路中最下端的電荷保存電晶體 的源極到與每一NOR快閃記憶體電路所在的直行相對應的 一本地的源極線;以及 連接每一橫列上的複數電荷保存電晶體的每一控制閘到一 字元線。 如申請專利範圍第45項所述之方法,還包括: 每一直行中NOR快閃記憶體電路有與其相對應的位元線和 源極線;以及 將位元線和源極線平行設置.。 如申請專利範圍第45項所述之方法,其中該NDR快閃記憶 體電路佈局要求該NOR快閃記憶體電路的λ小是製造NOR 快閃記憶體電路製程技街的最小的特性尺寸的四倍到六倍 〇 如申請專利範圍第45項所述乏方法,還包括: 形成一直行電壓控制電路;以及 • , Λ 連接該直行電壓控制電路,以提供控制訊號到與每一直行 電荷保存電晶體相對應的本地的位元線和源極線。 如申請專利範圍第45項所述之方法,還包括: 將每一本地的位元線透過一位元線選擇電晶體連接到複數 全域位元線中之一。 如申請專利範圍第45項所述之方法,還包括: 將每一本地的源極線透過一源極線選擇電晶鱧連接到複數 全域源極線中之一。 如申請專利範圍第46項所述之方法,還包括: 099113777 表單編號Α0101 第77頁/共133頁 0992024278-0 201104843 52 . 53 . 54 . 55 . 56 · 57 . 58 ' 099113777 將王域位7L線和全域源減連㈣直行轉控制 傳輸控制訊號到被選擇的本地的位元線和被選^以 源極線’用於對_快閃記憶體電路中被選擇的卜地的 電晶體進行讀取操作、編程操作以及抹除操作。何保存 如申請專職圍第46賴述之方法還包括: 形成一橫列電壓控制電路。 如申請專利範_52項賴之方法,還包括: 料該橫列電壓控制電路’以提供至少—控制訊號到 一橫列電荷保存電晶體相對應的字元線。 、 如申請專利範圍第52賴述之方法,還包括:201104843 Ο G VII. Patent application scope: • A N0R flash memory circuit, including: = number of charge-holding transistors are continuously connected into a -NAND string, where: - the uppermost charge holds the transistor, and its anode is connected a bit line associated with the charge-holding transistor; the lowermost charge-preserving transistor, the secret charge-connecting charge-holding transistor-related source line; and ^ the charge-preserving Each control gate of the transistor is connected to a word line. Such as the dirty flash memory circuit of the first paragraph of the patent scope, the charge storage transistor is formed in the well of the -type conductivity type. , such as the collision and fast __ circuit i of the scope of the patent scope of the application of the fourth type of conductivity type well formed in the first class of the type of guide type _ well. As stated in the third paragraph of the Shenqing patent scope N0R fast The flash-poor body circuit, the second type of conductive deep well is formed in a first conductivity type substrate. For example, the stupid flash memory circuit described in claim 1 of the patent scope, wherein the charge storage transistor of S is subjected to a tunneling effect for a programming operation and an erase operation. An N〇R flash memory circuit as described in claim 1, wherein a programming voltage of +15·0V to +2.00V is applied to a control of a selected charge-storing body and Between the bases, a programming operation is performed with a selected charge holding transistor in the charge holding transistors, wherein the programmed electrical waste is applied to the selected charge holding transistor in a gradually increasing manner. Between the control gate and the base. • The N〇R flash memory Zhao circuit as described in claim 6 of the patent scope, wherein a voltage less than 10·0 V is applied to the other of the charge-storage transistors. Among them, 099113777 Form No. A0101, page 69/ A total of 133 pages 0992024278-0 201104843 The unselected charge holds between the control gate and the base of the transistor to shield the unselected charge from holding the transistor. 8. The NOR flash memory circuit according to claim 1, wherein the layout of the NOR flash memory circuit requires the size of the NOR flash memory circuit to be a manufacturing process of the NOR flash memory circuit. Four to six times the smallest feature size. 9. The NOR flash memory circuit of claim 1, wherein a positive erase voltage of + 15. Ον to + 20. 0V is applied to the base of the selected charge storage transistor and the control gate In between, an erase operation is performed on the selected charge storage transistor. C1 10. The NOR flash memory circuit of claim 6, wherein a bias voltage of 0 V is applied to a control gate of the other unselected charge storage transistor in the charge storage transistor. Between the base and the base, the transistor is held to shield the unselected charges. 11. The NOR flash memory circuit of claim 1, wherein the selected charge holding transistor of the charge holding transistors that are programmed by a single order programming unit performs a read operation. The method includes: connecting a source line to a voltage following sensing circuit; U setting the voltage of the drain and gate of the selected charge holding transistor in the N 0 R flash memory circuit as a voltage source; The voltage of the gate of the unselected charge-storing transistor in the transistor is set to a first high read voltage; and the voltage is compared to the voltage generated at the source line and the reference voltage source in the sensing circuit, the reference voltage The source is set to 2.0 V to distinguish between a threshold voltage at a first logic value and a threshold voltage at a second logic value. 12. The NOR flash memory circuit of claim 11 is 099113777 Form No. A0101, and the first high read voltage is greater than 6.〇v. 13. The N〇R flash memory circuit of claim 2, wherein the reference voltage source is 2, 〇V. ', 14. The N〇R flash memory circuit as described in the scope of the patent application, wherein if the NOR flash memory circuit is not selected for the read operation, and the other N0R flash memory circuit is When the read operation is selected, the voltage of the control gate of the charge-holding transistor that is not selected in the charge-holding transistors in the unselected NOR flash memory circuit is set to the ground reference voltage to turn off the The charge holds the transistor. N15. The NOR flash memory circuit of claim 1, wherein the charge-preserving transistor in the charge-storing transistor in a program operation by a multi-level programming unit performs a read operation And including: connecting the source line to an electric waste following the sensing circuit; setting the voltage of the drain and the gate of the selected charge holding transistor to an intermediate high voltage; storing the electric charge in the transistor · setting the voltage of the gate of all unselected charge-storing transistors to a second high read voltage; and comparing the voltage to the voltage generated at the source line and the complex reference voltage source in the sensing circuit to determine A threshold voltage value is used to represent the data stored in the charge holding transistor. 16. The N〇R flash memory circuit as described in claim 15 wherein the intermediate high voltage is + 4. 〇V ° 17. The flash memory is as described in claim 15 The circuit 'where the second high read voltage is greater than 7.0V. 18. The NOR flash memory circuit of claim 15, wherein the reference voltage source is set between each threshold voltage to be stored separately at 0992024278-0 099113777 Form No. A0101 Page 71 / Total Page 133, 201104843 The charge holds the threshold voltage of the data in the transistor. 19. The NOR flash memory circuit of claim 15, wherein if the NOR flash memory circuit is not selected for reading, and when another NOR flash memory circuit is selected for reading, The voltages of the control gates of the unselected charge-storing transistors in the unselected NOR flash memory circuits in the unselected NOR flash memory circuit are set to a ground reference voltage to turn off the power and save the transistor. 20. An N0R flash memory, comprising: an array comprising a plurality of NOR flash memory circuits arranged in a row and a straight line, wherein each flash memory circuit comprises: €1 every line The plurality of charge-storing transistors are connected in series to form a NAND string; the drain of the uppermost charge-storing transistor in each of the N 0 R flash memory circuits is connected to a straight line corresponding to each N 0 R flash memory circuit a local bit line; the source of the lowest charge holding transistor in each NOR flash memory circuit is connected to a local source corresponding to the straight line where each N 0 R flash memory circuit is located Each of the control gates of the charge holding transistor on each row of _D is commonly connected to a word line. 21. The NOR flash memory of claim 20, wherein the bit line and the source line correspond to a straight line in which the NOR flash memory circuit is located and are parallel. 22. The NOR flash memory of claim 20, further comprising a continuous voltage control circuit for providing a control signal to a local bit line and source corresponding to each row of charge holding transistors Polar line. 099113777 Form No. A0101 Page 72 of 133 Page 0992024278-0 201104843 23 . As claimed in the 22nd paragraph of the patent application, only the flash memory is selected, and each bit of the local bit is selected by a bit line. The crystal is connected to a plurality of global bit lines of 24 25 〇 26 27 〇 28 29 099113777. The NOR flash memory of claim 23, wherein each local source line is selected through a source line. The crystal is connected to one of the plurality of global source lines, such as the Κ0Κ flash memory described in claim 24, wherein the global bit line and the global source line are connected to the straight line voltage control circuit for transmitting The control signal to the selected local bit edge and the selected local source line 'for reading, programming, and wiping the selected charge holding device in the NOR flash memory circuit In addition to the operation. The NOR flash memory described in claim 20 of the patent scope further includes a course voltage control circuit for providing a control signal to a word line corresponding to each row of charge storage transistors. The NOR flash memory according to claim 26, wherein the queue control circuit transmits a control signal to the word line to read the charge handle storage transistor of the NOR flash memory circuit (4) Take operations, program operations, and erase operations. The flash memory described in the second paragraph of the patent, including a 70-line selection control circuit, connecting all local bit line selection gates of the transistor; and a plurality of source line selection transistors, Secret (4) Select electricity (4) Connect to each local bit line. The N〇R flash memory according to claim 20, wherein the row control circuit is configured to transmit a word control signal to the word line, and a form number is 1^73 133 1 to the 09920242, 201104843 30 _Selected fine-grained, programmed operation and erase operation in the flash memory circuit, ::Electric 仃 仃 · · · · · (3) Book 3 (four) control circuit to choose the money and secret line to choose Wei _ The location of the thief selects the transistor and the selected source line transistor to transfer the bit line ^: line: signal from the straight-line voltage control circuit to the selected local: several lines and the selected local source Polar line. ~ 31 32 33 The _ flash memory described in item 20 of the scope of application, in which the electric storage transistor uses a F〇wler_Lim wear-through effect for programming operation and erasing operation. _ flash memory as described in claim 20, wherein a 15.0Vfj + 2G, the programming power of the GV is applied to the control gate and the base of the selected charge-storing electrode In between, to program the selected charge protection transistor. The N〇R flash memory as described in claim 31, wherein - a voltage less than 10.0 V is applied to the control gates of the other unselected charge-preserving transistors in the charge-holding transistors Between the base and the base, the transistor is held to shield the unselected charges. " N0R flash memory as described in claim 20, wherein the N 0 R flash memory circuit layout requires the size of the N 〇 R flash memory circuit to be a NOR flash memory circuit process The technology's smallest feature size is four to six times the size. 34 . 35 . 099113777 NOR flash memory according to claim 20, wherein a positive erase voltage of +1.5 0V to + 2 〇. 0V is applied to the base of the selected charge storage transistor Between the control gate and the control gate, an erase operation is performed on the selected charge storage transistor. For example, the NOR flash memory described in claim 20, wherein _ form number. A0101 page 74 / 133 pages 0992024278-0 201104843 36 . Ο 37 . 38 . ❹ 39 . 40 . 099113777 〇. 〇 A V bias is applied between the other unselected charge holding transistor miscellaneous gates and bases in the charge holding transistor group to mask the unselected charge holding transistors. The ceramic flash memory according to claim 20, wherein the selected charge storage transistor in the charge storage transistor in which the lung operation is performed by the single-stage coding element performs a reading operation, including: Connecting the source line to the _ voltage tracking sensing circuit; setting the voltage of the drain and gate of the selected charge holding transistor in the memory circuit to be the voltage of the voltage source; the charge is not selected in the holding transistor The charge-holding transistor's closed-pole voltage is set to a first high-read voltage; and the voltage is generated at the source line and the reference source is compared to the voltage in the sense circuit, where the reference voltage is set to 2. 〇v to distinguish between a threshold voltage at the first-logic value and a critical electrosurgical pressure at the second logic value. A flash memory as described in claim 36, wherein the first read voltage is greater than 6 〇v. ·: As described in the application for patent verse 36, the reference voltage source is 2. 〇Γ. " 'Eight in the 'Zhezhong' patents NGR flash memory, the unlined and unselected flash memory circuits of the unselected charge-preserving transistor Complex 5r, the sub-electricity is set to the ground reference ", to save the electric voltage with the off charge, as the patent application __ rhyme _ _ = - multi-level programming unit for the operation of the charge protection ^, the selected charge When the save transistor performs a read operation, the source line is connected to a voltage follow-up induction circuit in the crystal; Form number page 75/133 pages 0992024278-0 201104843 Set the selected charge to hold the transistor 汲The voltage of the pole and the gate is an intermediate high voltage; the voltage of the gate of all the unselected electricity and the holding transistor stored in the transistor is set to a second high reading voltage; The voltage follows the voltage developed at the source line and the complex reference voltage source in the sensing circuit to determine a threshold voltage value representative of the data stored in the charge holding transistor. 41. The N0R flash memory of claim 40, wherein the middle high voltage is +4. 0V. 42. The NOR flash memory of claim 40, wherein the second high read voltage is greater than 7.0V » 43. The NOR flash memory of claim 40, wherein A reference voltage source is provided between each threshold voltage to distinguish a threshold voltage of data stored in the charge holding transistor. 44. The NOR flash memory of claim 40, wherein the word line and the unselected NOR flash memory circuit save the unselected charge in the transistor to preserve the transistor The voltage of the complex control gate is set to the ground reference voltage to turn off the charge holding transistor. 45. A method of forming a NOR flash memory, comprising the steps of: providing a substrate; and forming a plurality of N 0 R flash memory circuits into an array comprising a horizontal and a straight line, wherein the NOR flash memory circuit Forming includes the steps of: placing a plurality of charge-preserving transistors in a straight row and a row; linearly connecting the plurality of charge-holding transistors in a straight row to form a NAND string to connect the topmost charge-storing transistor in each NOR flash memory circuit 099113777 Form No. A0101 Page 76 / 133 pages 0992024278-0 201104843 46 . Ο 47 . 48 . ❹ 49 , 50 51 The bungee to a local line corresponding to the straight line where each N 〇 L ^ flash memory circuit is located Bit line; connecting the source of the lowermost charge holding transistor in each NOR flash memory circuit to a local source line corresponding to the straight line where each NOR flash memory circuit is located; and connecting each The complex charge on one column holds each control gate of the transistor to a word line. The method of claim 45, further comprising: the NOR flash memory circuit has a corresponding bit line and source line in each line; and the bit line and the source line are arranged in parallel. . The method of claim 45, wherein the NDR flash memory circuit layout requires that the λ small of the NOR flash memory circuit is the minimum characteristic size of the NOR flash memory circuit. Up to six times, as in the method of claim 45, the method further includes: forming a line voltage control circuit; and, Λ connecting the straight line voltage control circuit to provide a control signal to save power with each line charge The local bit line and source line corresponding to the crystal. The method of claim 45, further comprising: connecting each local bit line to one of the plurality of global bit lines through a one-bit selection transistor. The method of claim 45, further comprising: connecting each local source line to one of the plurality of global source lines through a source line select transistor. For example, the method described in claim 46 includes: 099113777 Form No. Α0101 Page 77/133 Page 0992024278-0 201104843 52 . 53 . 54 . 55 . 56 · 57 . 58 ' 099113777 Will King 7L Line and global source de-connection (4) straight-forward control transmission control signal to selected local bit line and selected source line 'for transistor selected in the _ flash memory circuit Read operations, program operations, and erase operations. How to save The method of applying for the full-time section 46 includes: Forming a horizontal voltage control circuit. The method of applying the patent _52 further includes: ???said the horizontal voltage control circuit ‘to provide at least a control signal to a row of charge storage transistors corresponding to the word line. For example, the method of applying for the scope of patent application 52 also includes: 連接本地驗桃所腳的電8¾體的Μ姉本地的源極 所選擇的與每—本地的位元線相連的電晶體.的閘極。 如申請專利範圍第54項所述之方法,其中還包括: 從該橫列電壓控制電路傳輸訊號到字元線,以對_快閃 記憶體電路中被選擇的電荷保存電晶體進行讀取操作、編 程操作和抹除操作。The local source of the 83⁄4 body connected to the foot of the local peach tester. The gate of the transistor connected to each local bit line. The method of claim 54, further comprising: transmitting a signal from the horizontal voltage control circuit to the word line to perform a read operation on the selected charge holding transistor in the flash memory circuit , programming operations and erase operations. 如申請專利範圍第55項所述之方法,其中:從該橫列控制 電路傳輸選擇控制訊號到被選擇的位元線選擇電晶體和被 選擇的源極線選擇電晶體,用於把位元線和源極線控制訊 號從直行電壓控制電路傳輸到被選擇的本地的位元線和被 選擇的本地的源極線。 如申請專利範圍第45項所述之方法,其中該等電荷保存電 晶體應用一F〇wier_Nordheim穿隧效應進行編程操作和 抹除操作。 如申請專利範圍第45項所述之方法,其中一+ 15. 0V到 + 20. 0V的編程電壓被施加於該等電荷保存電晶體中—被 0992024278-0 表單編號A0101 第78頁/共133頁 201104843 59 60❹ 61 62G 099113777 選擇的電荷保存電晶體的控制閘和基極之間’以對該被選 擇的電荷保存電晶體進行編程操作’其中該編程電壓以逐 漸增加的步驟被施加於該被選擇的電荷保存電晶體的控制 閘和基極之間。 如申請專利範圍第45項所述之方法’其中一小於10. 〇v的 編程遮蔽電壓被施加於該等電荷保存電晶體中其他未被選 擇的電荷保存電晶體的控制閘和基極之間,以遮蔽該等未 被選擇的電荷保存電晶體。 如申請專利範圍第45項所述之方法,其中施加一+ 15. 〇v 到+ 20. 〇ν的負極抹除電麈於被選擇的電荷保存電晶體的 基極和控制閘之間,以對该被選擇的電荷保存電晶體進行 抹除操作。 如申請專利範圍第45項所述之方法,其中二0.0V的偏壓 被施加於該等電荷保存電晶體中其他未被選擇的電荷保存 電晶體控制閘和基極之間,以遮蔽該等未被選擇的電荷保 存電晶體。 如申請專利範圍第45項所述之方法―’其中被編程為—單階 編程單元的複數電荷保存電晶體中被選擇的電荷保存電晶 體通過以下步驟被讀取: 阳 連接源極線到一電壓追隨感應電路; 設置該_快閃記憶體電路内被選擇的電荷保存電晶體的 汲極和閘極的電壓為電壓源的電壓; 该等電荷保存電晶體中未被選擇的電荷保存電晶體的閉極 的電壓被設定為一第一高讀取電壓;以及 比較電堡追隨感應電路令在源極線產生的電壓和一參考電 壓源’該參考f Μ雜設置狀GV,簡分—在第 表單編號_1 第79頁/共⑶頁 0992024278-0 201104843 輯值的臨界電壓和一在第二邏輯值的臨界電壓。 63 .如申請專利範圍第62項所述之,其中該第一高讀取電壓大 於6· 0V。 64 .如申請專利範圍第62項所述之方法,其中該參考電壓源為 2. 0V。 65 .如申請專利範圍第62項所述之方法,其中字元線亦即未被 選擇的N0R快閃記憶體電路的電荷保存電晶體中未被選擇 的電荷保存電晶體的複數控制閘的電壓被設置到接地參考 電壓,以關閉電荷保存電晶體。 66.如申請專利範圍第45項所述之方法,其中被編程為一多階 (% 編程单元的複數電荷保存電晶體中被選擇的電何保存電晶 體通過以下步驟被讀取: 連接源極線到一電壓追隨感應電路; 設置該被選擇的電荷保存電晶體的汲極和閘極的電壓為一 中間的高電壓; 在電4保存電晶體之内所有未被選擇的電何保存電晶體的 閘極電壓設置到第二高讀取電壓;以及 比較電壓追隨感應電路中在源極線產生的電壓和複數參考 U 電壓源,以決定一臨界電壓值用於代表儲存於電荷保存電 晶體内的數據。 67 .如申請專利範圍第66項所述之方法,其中該標準的高電壓 是 +4.0V 。 68 .如申請專利範圍第66項所述之方法,其中該第二高讀取電 壓大於7. 0V ^ 69 .如申請專利範圍第66項所述之方法,其中該參考電壓源被 設置在每一臨界電壓之間,以區別儲存於電荷保存電晶體 099113777 表單編號 A0101 第 80 頁/共 133 頁 0992024278-0 201104843 70 . 71 · Ο ❹ 72 . 内的數據的臨界電壓。 如申請專利範圍第66項所述之方法,其中字元線,亦即未 被選擇的NOR快閃記憶體電路的電荷保存電晶體中未被選 擇的電荷保存電晶體的複數控制閘的電壓被設置到接地參 考電壓*以關閉該電何保存電晶體。 一種積體電路裝置,包括: 一包括複數NAND快閃記憶體電路的陣列,每一NAND快閃 記憶體電路包括: 複數電荷保存電晶體,排列成橫列和直行,其中每一直行 上的電荷保存電晶體以線性連接成一 NAND串,每一 NAND 串有一上端選擇電晶體和一下端選擇電晶體; 一包括複數N0R快閃記憶體電路的陣列,其中每一N0R 快閃記憶體電路包括: 複數電荷保存電晶體,排列成橫列和直行,其中每一直行 上的電荷保存電晶體群以線性連接成一 NAND串,其中舞 一N0R快閃記憶體電路中最上端的電荷保存電晶體的汲極 連接到每一N0R快閃記憶體電路所在的直行中相對應的一 本地的位元線,其中每一 N0R快閃記憶體電路中最下端的 電荷保存電晶體的源極連接到每一 N 0 R快閃記憶體電路所 在的直行中相對應的一本地的源極線,其中每一橫列上的 複數電荷保存電晶體中的每一控制閘連接到一字元線。 如申請專利範圍第71項所述之積體電路裝置,其中在每一 N0R快閃記憶體電路中,其中位元線和源極線與NOR快閃 記憶體電路所在的直行相對應並且是平行的。 如申請專利範圍第71項所述之積體電路裝置,其中每一 N0R快閃記憶體電路包括一直行電壓控制電路,用於提供 099113777 表單編號A0101 第81頁/共133頁 0992024278-0 73 . 201104843 74 . 75 . 76 · 77 . 78 · 79 80 控制訊號到與每一直行中電荷保存電晶體相對應的本地的 位元線和源極線。 如申請專利範圍第71項所述之積體電路裝置,其中在每一 N0R快閃記憶體電路中,每一本地的位元線透過一位元線 選擇電晶體連接到複數全域位元線中之一。 如申請專利範圍第74項所述之積體電路裝置,其中在每一 N0R快閃記憶體電路中,每一本地的源極線透過一源極線 選擇電晶體連接到複數全域源極線中之一。 如申請專利範圍第75項所述之積體電路裝置,其中在每一 N0R快閃記憶體電路中,全域位元線和全域源極線連接到 Γ〗 直行電壓控制電路,以傳輸控制訊號到被選擇的本地的位 元線和被選擇的本地的源極線,用於對N0R快閃記憶體電 路中被選擇的電荷保存電晶體執行讀取操作、編程操作和 抹除操作。 如申請專利範圍第74項所述之積體電路裝置,其中每一 N0R快閃記憶體電路包括一橫列電壓控制電路,用於提供 控制訊號到與每一橫列電荷保存電晶體相對應的字元線。 如申請專利範圍第77項所述之積體電路裝置,其中在每一 CJ N0R快閃記憶體電路中,該橫列控制電路傳輸訊號到字元 線,以對N0R快閃記憶體電路中被選擇的電荷保存電晶體 執行讀取操作、編程操作和抹除操作。 如申請專利範圍第78項所述之積體電路裝置,其中每一 N0R快閃記憶體電路還包括一位元線選擇控制電路,該位 元線選擇控制電路連接所有本地的位元線選擇電晶體的閘 極,複數源極線選擇電晶體則連接到每一本地的位元線。 如申請專利範圍第71項所述之積體電路裝置,其中在每一 099113777 表單編號A0101 第82頁/共133頁 0992024278-0 201104843 81 ·G 82 . 83 .Ο 84 · 099113777 NOR快閃記憶體電路中,橫列控制電路傳輪字元線控制訊 號到字it線’快㈣憶體電財鶴擇的電荷保 存電晶體執行讀取操作、編程操作和抹除操作,該橫列控 制電路還分別傳輪位元線選擇訊號和源極線選擇訊號到被 選擇的位it線選擇電晶體和被選擇的源極線電晶體,以將 位元線和雜線㈣减從直行電壓㈣電料輪到被選 擇的本地的位元線和被選擇本地的源極線。 如申請專利難第71項所述之積體電路裝置,其中在每一 隱快閃記憶體電路中,該等電荷保存電晶體採用一 Powler^ordheim穿隧效應進行編程操作和抹除操作。 如申請專利範圍第71項所述喝積體電路裝屋,,其中在每一 NOR快閃記憶體電路中,一+15 .〇v到舰喊編程電壓 被施加於複數電荷贿電晶备-被選義電荷保存電晶 體的控制閘和基極之間,以將該電荷紳電晶體編程為單 階編程單元’其中該編程電壓以逐漸增加的步驟被施加於 該被選擇的電荷鱗電晶體的㈣_祕之間。 如申凊專利朗第82項所述之祕電路裝置,其中在每一 R决閃》己It體電路中,—小於-1Q.GV的編程遮蔽電壓被 施加於複數f荷料電晶㈣巾其他未被選擇的電荷保存 電明體的控制’基極之間,以遮蔽該等未鶴擇的電荷 保存電晶體。 如申凊專利範圍第71項所述之積艘電路裝置,其中在每一 峨快閃記憶體電路中,該NOR快閃記憶體電路佈局要求 糊R快閃記憶體電路的大小是製造N 〇 R快閃記愧體電路 製程技術的最小的特性尺寸的四倍到六倍。 如申請專利範圍第71項所述之積體電路裝置,其中在每一 表單編號A0101 第83頁/請頁 0992024278-0 85 . 201104843 NOR快閃記憶體電路中,一+ 15. 0V到+ 20. 0V的負極抹除 電壓被施加於被選擇的電荷保存電晶體的基極和控制閘之 間,以對該被選擇的電荷保存電晶體進行抹除操作。 86 .如申請專利範圍第71項所述之積體電路裝置,其中在每一 N0R快閃記憶體電路中,一0.0V的偏壓被施加於該等電荷 保存電晶體中其他未被選擇的電荷保存電晶體的控制閘和 基極之間,以遮蔽該等未被選擇的電荷保存電晶體。 87 .如申請專利範圍第71項所述之積體電路裝置,其中在每一 N0R快閃記憶體電路中,複數電荷保存電晶體中被編程為 一單階編程單元的被選擇的電荷保存電晶體通過以下步驟 〇 被讀取: 連接源極線到一電壓追隨感應電路; 設置該N 0 R快閃記憶體電路内被選擇的電荷保存電晶體的 汲極和閘極的電壓為電壓源的電壓; 該電4保存電晶體群中未被選擇的電何保存電晶體的閘極 的電壓被設置為一第一高讀取電壓;以及 比較該電壓追隨感應電路中在源極線產生的電壓和參考電 壓源,該參考電壓源被設置到2. 0V,以區分一在第一邏 U 輯值的臨界電壓和一在第二邏輯值的臨界電壓。 88 .如申請專利範圍第87項所述之積體電路裝置,其中在每一 N0R快閃記憶體電路中,該第一高讀取電壓大於6. 0V。 89 .如申請專利範圍第87項所述之積體電路裝置,其中在每一 NOR快閃記憶體電路中,該參考電壓源的電壓為2. 0V。 90 .如申請專利範圍第87項所述之積體電路裝置,其中字元線 ,亦即未被選擇的N 0 R快閃記憶體電路的該等電荷保存電 晶體中未被選擇的電荷保存電晶體的複數控制閘的電壓被 099113777 表單編號A0101 第84頁/共133頁 0992024278-0 201104843 91 ❹ 92 93 Ο 94 .如 95 屋 099113777 ,以關閉電荷保存電晶 表單編號Α0101 設置到接地參考電壓,以關閉電荷保存電晶體。 如申請專利範圍第71項所述之積體電路裝置,其中在每一 麵快閃記憶體電路中,其中被編程為一多階編程單元的 複數電荷保存電晶體中被選擇的電荷保存電晶體通過以下 步驟被讀取: 連接源極線到一電壓追隨感應電路; 設置該被選擇的電荷保存電晶體的汲極和閘極的電壓為一 中間的高電壓; 將電荷保存電晶體群之⑽有未被選_電荷保存電晶體 的閑極的電壓設置為一第二高讀取電麼;以及 比較該驗轉感應電財.在祕線產_電壓和複數參 考電壓源’以蚊—臨界電㈣用於代表儲存於電荷保存 電晶體内的數據。 η 〜 如申請專利範圍第91項所述之_電路裝置,其 NOR快閃記憶體電路中麻 、 中間阿電壓為+4.0V» 。專利_第91項所述之積 画快閃記憶體電路中,令第’2置’其中在每一 如申靖專利電壓大於7.〇v。 麵快閃記/體電項所述之積體電路裝置,其中在每一 ㈣之間㈣顯該參考電壓源被設置在每一臨界 麼。 丨叫於電娜存電⑽_數據的臨界電 如申請專利範圍第 ,亦即未被選握从 述之積體電路裝置,其中字1 以未被選擇的峋快閃記憶 #中予几線 晶體中未被選擇的 "路的該等電荷保存電 設置到接地參/了保存電晶體的複數控制_電屢被 體 第85頁/共133頁 0992024278-0The method of claim 55, wherein: the selection control signal is transmitted from the row control circuit to the selected bit line selection transistor and the selected source line selection transistor for using the bit element The line and source line control signals are transmitted from the straight line voltage control circuit to the selected local bit line and the selected local source line. The method of claim 45, wherein the charge storage transistors employ a F〇wier_Nordheim tunneling effect for programming and erasing operations. A method as claimed in claim 45, wherein a programming voltage of + 15.0 V to + 20. 0 V is applied to the charge holding transistors - by 0992024278-0 Form No. A0101 Page 78 of 133 Page 201104843 59 60❹ 61 62G 099113777 The selected charge-holding transistor is between the control gate and the base 'to program the selected charge-storing transistor', wherein the programming voltage is applied to the The selected charge holds between the control gate and the base of the transistor. The method of claim 45, wherein one of the methods is less than 10. The programming mask voltage of 〇v is applied between the control gate and the base of the other unselected charge-storing transistors in the charge-storing transistors. To conserve the unselected charge to hold the transistor. The method of claim 45, wherein applying a negative electrode of + 15. 〇v to + 20. 〇ν erases electricity between the base of the selected charge-preserving transistor and the control gate, The selected charge holding transistor performs an erase operation. The method of claim 45, wherein a bias voltage of two 0.0V is applied between the other unselected charge storage transistor control gates and the bases of the charge storage transistors to shield the same The unselected charge holds the transistor. The method of claim 45, wherein the selected charge-storing transistor in the complex charge-storing transistor in which the single-stage programming unit is programmed is read by the following steps: cation connection source line to one a voltage follow-up sensing circuit; setting a voltage of the drain and gate of the selected charge-holding transistor in the flash memory circuit to a voltage source; the charge-preserving transistor has an unselected charge-preserving transistor The closed-pole voltage is set to a first high read voltage; and the comparison of the electric gate follows the induction circuit to cause the voltage generated at the source line and a reference voltage source 'the reference f noisy setting GV, simplification| Form No. _1 Page 79 / Total (3) Page 0992024278-0 201104843 The threshold voltage of the value and a threshold voltage at the second logic value. 63. The method of claim 62, wherein the first high read voltage is greater than 6.0V. 64. The method of claim 62, wherein the reference voltage source is 2. 0V. 65. The method of claim 62, wherein the word line, that is, the voltage of the unselected NOR flash memory circuit, of the unselected charge storage transistor, the voltage of the plurality of control gates of the transistor. It is set to the ground reference voltage to turn off the charge holding transistor. 66. The method of claim 45, wherein the selected one of the plurality of stages (the % of the plurality of charge-storing transistors selected to save the transistor is read by the following steps: connecting the source a line-to-voltage follow-up sensing circuit; setting the voltage of the drain and gate of the selected charge-storing transistor to an intermediate high voltage; all the unselected electricity within the cell 4 holding the transistor The gate voltage is set to a second high read voltage; and the comparison voltage follows the voltage generated at the source line and the complex reference U voltage source in the sense circuit to determine a threshold voltage value for use in the charge storage transistor The method of claim 66, wherein the high voltage of the standard is +4.0 V. 68. The method of claim 66, wherein the second high read voltage More than 7. 0V ^ 69. The method of claim 66, wherein the reference voltage source is set between each threshold voltage to be separately stored in the charge storage transistor 0991 13777 Form No. A0101 Page 80 of 133 Page 0992024278-0 201104843 70 . 71 · Ο ❹ 72. The threshold voltage of the data. For example, the method described in claim 66, wherein the word line, that is, The voltage of the complex control gate of the unselected charge-storing transistor in the charge-storing transistor of the selected NOR flash memory circuit is set to the ground reference voltage* to turn off the current and save the transistor. The method includes: an array comprising a plurality of NAND flash memory circuits, each NAND flash memory circuit comprising: a plurality of charge-storing transistors arranged in a row and a straight line, wherein the charge on each of the lines preserves the transistor to be linear Connected into a NAND string, each NAND string has an upper selection transistor and a lower selection transistor; an array comprising a plurality of NOR flash memory circuits, wherein each NOR flash memory circuit comprises: a plurality of charge storage transistors, Arranged into a horizontal row and a straight line, wherein each row of charge-holding transistor groups is linearly connected into a NAND string, wherein the dance-N0R flashes The drain of the uppermost charge-preserving transistor in the memory circuit is connected to a corresponding local bit line in the straight line where each of the NOR flash memory circuits is located, wherein the lowest end of each of the NOR flash memory circuits The source of the charge-storing transistor is connected to a corresponding local source line in a straight row in which each N 0 R flash memory circuit is located, wherein each of the plurality of charge-preserving transistors in each row holds the control The gate is connected to a word line. The integrated circuit device according to claim 71, wherein in each of the NOR flash memory circuits, the bit line and the source line and the NOR flash memory circuit are The straight line in which it is located corresponds and is parallel. The integrated circuit device of claim 71, wherein each of the NOR flash memory circuits comprises a line voltage control circuit for providing 099113777 form number A0101 page 81 / page 133 0992024278-0 73 . 201104843 74 . 75 . 76 . 77 . 78 · 79 80 Control signals to the local bit line and source line corresponding to the charge holding transistor in each row. The integrated circuit device of claim 71, wherein in each of the NOR flash memory circuits, each local bit line is connected to the plurality of global bit lines through a one-element selection transistor. one. The integrated circuit device of claim 74, wherein in each of the NOR flash memory circuits, each local source line is connected to the plurality of global source lines through a source line selection transistor. one. The integrated circuit device of claim 75, wherein in each of the NOR flash memory circuits, the global bit line and the global source line are connected to the 直 straight-line voltage control circuit for transmitting the control signal to The selected local bit line and the selected local source line are used to perform read operations, program operations, and erase operations on the selected charge holding transistors in the NOR flash memory circuit. The integrated circuit device of claim 74, wherein each of the NOR flash memory circuits includes a horizontal voltage control circuit for providing a control signal to correspond to each of the row charge storage transistors. Word line. The integrated circuit device of claim 77, wherein in each CJ NOR flash memory circuit, the row control circuit transmits a signal to a word line to be used in the NOR flash memory circuit. The selected charge holding transistor performs a read operation, a program operation, and an erase operation. The integrated circuit device of claim 78, wherein each of the NOR flash memory circuits further comprises a bit line selection control circuit, and the bit line selection control circuit is connected to all local bit lines to select electricity. The gate of the crystal, the complex source line select transistor is connected to each local bit line. The integrated circuit device as described in claim 71, wherein each 099113777 form number A0101 page 82 / 133 page 0992024278-0 201104843 81 · G 82 . 83 .Ο 84 · 099113777 NOR flash memory In the circuit, the row control circuit transmits the word line control signal to the word line “fast” (four), and the charge storage transistor performs the read operation, the programming operation and the erase operation, and the row control circuit further The transfer bit line selection signal and the source line selection signal are respectively selected to the selected bit it line selection transistor and the selected source line transistor to reduce the bit line and the impurity line (4) from the straight line voltage (four) electric material It is the turn of the selected local bit line and the selected local source line. The integrated circuit device according to claim 71, wherein in each of the hidden flash memory circuits, the charge storage transistors are subjected to a Powler^ordheim tunneling effect for programming operation and erasing operation. For example, in the application of the patent scope, item 71, in which the integrated circuit is installed, in each NOR flash memory circuit, a +15 .〇v to the ship shouting programming voltage is applied to the plurality of charge bribes - The selected charge holds between the control gate and the base of the transistor to program the charge transistor into a single-order programming unit' wherein the programming voltage is applied to the selected charge scale transistor in a stepwise increasing step (four) _ secret between. For example, in the circuit device of claim 82, wherein in each R-flashing circuit, the programming masking voltage of less than -1Q.GV is applied to the plurality of f-charged electric crystal (four) towels. Other unselected charges preserve the control of the electro-lumps between the bases to shield the unsupported charges from preserving the transistors. The apparatus circuit of claim 71, wherein in each of the flash memory circuits, the NOR flash memory circuit layout requires the size of the paste R flash memory circuit to be N 〇 The R flash flash memory circuit technology is four to six times the smallest feature size. The integrated circuit device as described in claim 71, wherein in each form number A0101, page 83 / page 0992024278-0 85 . 201104843 NOR flash memory circuit, one + 15. 0V to + 20 A negative erase voltage of 0 V is applied between the base of the selected charge holding transistor and the control gate to perform an erase operation on the selected charge holding transistor. 86. The integrated circuit device of claim 71, wherein in each of the NOR flash memory circuits, a bias of 0.0 V is applied to the other unselected ones of the charge holding transistors. A charge is held between the control gate and the base of the transistor to shield the unselected charge from holding the transistor. 87. The integrated circuit device of claim 71, wherein in each of the NOR flash memory circuits, a selected charge-storing device programmed into a single-order programming unit in the plurality of charge-storing transistors The crystal is read by the following steps: connecting the source line to a voltage follow-up sensing circuit; setting the voltage of the drain and gate of the selected charge-holding transistor in the N 0 R flash memory circuit to a voltage source a voltage; the voltage of the gate of the unselected electrical and holding transistor is set to a first high read voltage; and comparing the voltage to the voltage generated at the source line in the sensing circuit And a reference voltage source, the reference voltage source is set to 2.0V to distinguish a threshold voltage at the first logic value and a threshold voltage at the second logic value. The first high-reading voltage is greater than 6.0V in each of the NOx flash memory circuits, as described in claim 87. The voltage of the reference voltage source is 2.0 V in each of the NOR flash memory circuits. 90. The integrated circuit device of claim 87, wherein the word line, that is, the unselected N 0 R flash memory circuit, of the uncharged charge in the charge holding transistor The voltage of the complex control gate of the transistor is 099113777 Form No. A0101 Page 84 / Total 133 Page 0992024278-0 201104843 91 ❹ 92 93 Ο 94 . Such as 95 House 099113777, to close the charge to save the crystal form number Α 0101 set to the ground reference voltage To save the transistor by turning off the charge. The integrated circuit device of claim 71, wherein in each of the flash memory circuits, the selected charge-preserving transistor in the plurality of charge-preserving transistors programmed into a multi-level programming unit It is read by the following steps: connecting the source line to a voltage following sensing circuit; setting the voltage of the drain and gate of the selected charge holding transistor to an intermediate high voltage; storing the charge in the transistor group (10) The voltage of the idle pole of the unselected _charge-storing transistor is set to a second high-reading power; and the comparison of the inductive-inductive power is made. In the secret line, the voltage and the complex reference voltage source 'mosquito-critical Electricity (4) is used to represent the data stored in the charge-storing transistor. η~ As in the circuit device described in item 91 of the patent application, the voltage in the NOR flash memory circuit is +4.0V». In the flash memory circuit of the patent described in Item 91, the first '2' is placed in each of the applications such as Shenjing's patent voltage is greater than 7.〇v. The integrated circuit device described in the flash flash/physical item, wherein between each (four) (four), the reference voltage source is set at each threshold. The screaming of the electric power of the electric storage (10) _ data, such as the scope of the patent application, that is, the selection of the circuit device is not selected, in which the word 1 is in the unselected 峋 flash memory # The charge of the unselected "luster in the crystal is saved to the grounding parameter/the complex control of the holding transistor_Electric-repeat body page 85/133 pages 0992024278-0
TW99113777A 2009-05-07 2010-04-30 A NAND based NMOS NOR flash memory cell, a NAND based NMOS NOR flash memory array, and a method of forming a NAND based NMOS NOR flash memory array TW201104843A (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US12/387,771 US8072811B2 (en) 2008-05-07 2009-05-07 NAND based NMOS NOR flash memory cell, a NAND based NMOS NOR flash memory array, and a method of forming a NAND based NMOS NOR flash memory array

Publications (1)

Publication Number Publication Date
TW201104843A true TW201104843A (en) 2011-02-01

Family

ID=44816966

Family Applications (1)

Application Number Title Priority Date Filing Date
TW99113777A TW201104843A (en) 2009-05-07 2010-04-30 A NAND based NMOS NOR flash memory cell, a NAND based NMOS NOR flash memory array, and a method of forming a NAND based NMOS NOR flash memory array

Country Status (1)

Country Link
TW (1) TW201104843A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI868782B (en) * 2023-07-06 2025-01-01 旺宏電子股份有限公司 In memory searching device
US12400706B2 (en) 2023-07-06 2025-08-26 Macronix International Co., Ltd. In memory searching device

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI868782B (en) * 2023-07-06 2025-01-01 旺宏電子股份有限公司 In memory searching device
US12400706B2 (en) 2023-07-06 2025-08-26 Macronix International Co., Ltd. In memory searching device

Similar Documents

Publication Publication Date Title
JP3854348B2 (en) NAND flash memory device and driving method thereof
TW559815B (en) Semiconductor memory device having memory cell arrays capable of accomplishing random access
US6771536B2 (en) Operating techniques for reducing program and read disturbs of a non-volatile memory
US8345481B2 (en) NAND based NMOS NOR flash memory cell, a NAND based NMOS NOR flash memory array, and a method of forming a NAND based NMOS NOR flash memory array
TWI631563B (en) Method to reduce program disturbs in non-volatile memory cells
CN204991153U (en) Storage location and nonvolatile memory
TW200425149A (en) Semiconductor memory device and data write method
US6909639B2 (en) Nonvolatile memory having bit line discharge, and method of operation thereof
KR20060002759A (en) System and method for programming a cell in a nonvolatile memory device
JP2007520029A (en) Variable current sinking for non-volatile memory miscellaneous / fine programming
CN100447899C (en) Flash memory device with stable source line regardless of bit line coupling and loading effect
JP2007520028A (en) Charge packet measurement for non-volatile memory miscellaneous / fine programming
CN102460402A (en) Memory system having nand-based nor and nand flashes and sram integrated in one chip for hybrid data, code and cache storage
CN113450857B (en) Peak and average current reduction for sub-block memory operations
TW201106365A (en) Memory erase methods and devices
CN105229745B (en) Share in memory and support circuit
KR101747565B1 (en) Complementary decoding for non-volatile memory
US12198745B2 (en) Semiconductor storage device and writing method thereof
TWI352994B (en)
US8416628B2 (en) Local sensing in a memory device
CN102067235A (en) NADN based NMOS NOR flash memory cell, a NADN based NMOS NOR flash memory array, and a method of forming a nand based nmos nor flash memory array
US8437198B2 (en) Method for discharging a voltage from a capacitance in a memory device
TW201104843A (en) A NAND based NMOS NOR flash memory cell, a NAND based NMOS NOR flash memory array, and a method of forming a NAND based NMOS NOR flash memory array
TWI693602B (en) Operation method of low-current electronic erasable rewritable read-only memory array
JP2009301691A (en) Nonvolatile semiconductor memory device