TW201030947A - Single-polycrystalline silicon electrically erasable and programmable nonvolatile memory device - Google Patents
Single-polycrystalline silicon electrically erasable and programmable nonvolatile memory device Download PDFInfo
- Publication number
- TW201030947A TW201030947A TW99103898A TW99103898A TW201030947A TW 201030947 A TW201030947 A TW 201030947A TW 99103898 A TW99103898 A TW 99103898A TW 99103898 A TW99103898 A TW 99103898A TW 201030947 A TW201030947 A TW 201030947A
- Authority
- TW
- Taiwan
- Prior art keywords
- volatile memory
- potential
- source
- floating gate
- well
- Prior art date
Links
Landscapes
- Non-Volatile Memory (AREA)
Abstract
Description
201030947 六、發明說明: 【發明所屬之技術領域】 _]树明涉及-種非揮發性_電路記憶器件,特別涉及 -種單層多晶碎浮動閘電子可抹除可編程唯讀記憶體 (EEPR0M)和快fA]電子可抹除可編程唯讀記憶體(FLASH MEMORY) 〇 【先前技術] [0002] [0003] 〇 在半導體工業,一般來說,有兩大類重要的“⑽記憶體 :"揮發性"和"非揮發性"。"揮發性"記憶體指的是當低 電壓電源磋移去或被關閉時,所儲存的章料不會被保存 ’揮發性記憶艘包括靜態随¥接取(^^.動態隨機存 取(DRAM)纪憶鑪· ‘>、\ '% I:- “非發揮性”記憶艟(NVM)所儲存的資訊不會被破壞,通 常在電源電壓斷電以後亦能夠保存二十奉。如今有許多 不同種類的NVM記德鱗適合約應沒。螂如,最普遍的 sf ifelleCTurii ❹ 画快閃記料,它有寸大約是0.5T ,即4又2“2是-特定的補I#許的最小面積) ,NAND快閃記憶艎用於儲存大量的影音資料。2008年, 利用45nm的製程技術,NAND記憶體的有效密度最高可達 到16 Gb。 [0004] NVM第二大·類是NOR快閃記憶體,它的記憶單元為一面積 大約是10 λ 2的單電晶體’ NOR快閃記憶體用來儲存程式 代碼。如今市場上的NOR快閃記憶體的最大有效記憶密度 是2 Gb,其是2007年利用70nm的製程技術做出。第三類 NVM是雙電晶體浮動閘隧道氧化(FLOTOX) EEPR0M,它 099103898 表單編號A0101 第4頁/共106頁 0992007313-0 201030947 [0005]201030947 Sixth, invention description: [Technical field of invention] _] Shuming involves a kind of non-volatile _ circuit memory device, especially related to - single-layer polycrystalline floating gate electronic erasable programmable read-only memory ( EEPR0M) and fast fA] electronically erasable programmable read-only memory (FLASH MEMORY) 〇 [prior art] [0002] [0003] In the semiconductor industry, in general, there are two important categories of "(10) memory: "Volatile" and "non-volatile"."Volatile" memory means that the stored chapters are not preserved when the low-voltage power source is removed or turned off. The memory bank includes static information stored in ¥ (^^. Dynamic Random Access (DRAM) Memory Remembrance · '>, \ '% I:- "Non-Playing" Memory (NVM) will not be stored Destruction, usually after the power supply voltage is cut off, can also save twenty Feng. Now there are many different kinds of NVM gram scales suitable for the corresponding. For example, the most common sf ifelleCTurii ❹ draw flash notes, it has about inch 0.5T, that is, 4 and 2 "2 is the minimum area of the specific complement I#, NAND Flash memory is used to store a large amount of audio and video data. In 2008, with 45nm process technology, NAND memory has an effective density of up to 16 Gb. [0004] The second largest class of NVM is a NOR flash memory whose memory cell is a single transistor 'NOR flash memory' having an area of about 10 λ 2 for storing program code. The maximum effective memory density of NOR flash memory on the market today is 2 Gb, which was made in 2007 using 70 nm process technology. The third type of NVM is a double crystal floating gate tunnel oxidation (FLOTOX) EEPR0M, which is 099103898 Form No. A0101 Page 4 of 106 0992007313-0 201030947 [0005]
[0006][0006]
的記憶單元尺寸是8〇λ2。目前,EE_M的密度僅是嘟 左右’-般顧在位元組更改級的應用上,細纖或麵 快閃記憶體只允許大塊數據的改變不同,EEp_能夠達 到最大的編程/抹除(P/E)週期。目前,針對以位元組為 早位的少量數據的更改,EEPR0M可達到一百萬次〇 Million)P/E 週期。 議有幾個缺點:為了完成基本的抹除和編程操作它需 要將高電壓器件、電荷幫浦電路和複雜的雙層多晶石夕單 元結構都内建在同一個晶片上。目前,上述議單元器件 都是利用複雜的雙層多晶碲凑.電壓槊程。 前面提到的雙層多晶画單|^徽’。_ 4執行編程和 抹除操作時所需要㈣壓“所製造的 元器件而言太高。例奸’目前每個NAND單元結構的〇 5T 電晶體就需要2 〇 V來做福勒-諾德漢 (Fowler-N〇rdheim,FN)^|t^|^0^^ 0 單電晶體NOR*閃單元來外感釋街▼的編程作業需要 大約lov。然’ Fowle卜Ν(θ_娜道抹除操作卻需要 + 10V和-10V。目前,一個雙電晶體EEPR〇M記憶單元結構 需要+ 15V執行fowler-Nordheim編程和抹除操作。纟士果 ,若把以上三種NVM單元合起來内建在同一個晶片上進行 編程和抹除操作,就需要一個電荷幫浦電路提供大約1〇v 到20V的高電壓。故,NVM陣列週邊器件為了該等操作需 要一個尚擊穿(Breakdown)電壓。然,高擊穿電壓並不 相容於目前週邊的單層多晶低電壓邏輯器件製程技術, 若要達到該高擊穿電壓則必須修改製程,那就會增加生 099103898 表單編號A0101. 第5頁/共106頁 0992007313-0 201030947 產成本。 [0007] 已發表的名為"嵌入應用下低電壓低功率操作的新單層多 晶快閃記憶體單元"(A New Single-Poly Flash Memory Cell with Low-Voltage and Low-Power Operations for Embedded Applications, CHI, ETAL., THE 5TH ANNUAL IEEE DEVICE RESEARCH CONFERENCE DIGEST, JUNE 1997, PP:126 -127, POSTED IEEEXLORE. IEEE.ORG : 2002 -08 -06 [0008] 21:21:45.0)的文章討論到在三重井(triple wells) CMOS技術下單層多晶快閃記憶體單元結構以及工作電壓 不超過±Vcc下的新編程和抹除方蛛。_❸单择多晶 ... EPROM雖然完全相容於標準C»0Sft;程S卻县有須高電壓 操作、慢編程以及無法敵電抹除的缺點❶具有該新編程/ 抹除方案的快閃單元使系統單晶片CMOS混合訊號電路允 許低電壓、低功率以及非揮發記華體的應用。. 广 rif.'!;ecr ' 美國專利5, 929, 478(?81^!3以6^:€^ )描述了 一單層 非揮發記憶體。它包括一個ΐ鲁·,__ΡΕΤ和一個在p基板The memory cell size is 8 〇 λ2. At present, the density of EE_M is only the same as the application of the byte modification level. The fine fiber or the surface flash memory only allows the change of the large data to be different, and the EEp_ can achieve the maximum programming/erasing. (P/E) cycle. Currently, EEPR0M can achieve a million Million) P/E cycles for changes to a small amount of data in the early position of the byte. There are several disadvantages to this: in order to perform basic erase and program operations, it is necessary to build high voltage devices, charge pump circuits, and complex double-layer polycrystalline silicon unit structures on the same wafer. At present, the above-mentioned unit cells are all utilizing complex two-layer polysilicon. The double-layer polycrystalline single sheet|^ emblem' mentioned above. _ 4 required to perform programming and erasing operations (4) pressure "the components manufactured are too high. The smuggling of the current NAND5T transistor of each NAND cell structure requires 2 〇V to do Fowler-Nord Han (Fowler-N〇rdheim, FN)^|t^|^0^^ 0 Single crystal NOR* flash unit to the external sense of the street ▼ programming operations need about lov. Then 'Fowle divination (θ_ Na Dao wipe In addition to operation, it requires +10V and -10V. Currently, a dual-transistor EEPR〇M memory cell structure requires +15V to perform fowler-Nordheim programming and erase operations. Gentleman, if the above three NVM units are built together Programming and erasing operations on the same die requires a charge pump circuit to provide a high voltage of approximately 1 〇v to 20 V. Therefore, the NVM array peripheral device requires a Breakdown voltage for such operations. The high breakdown voltage is not compatible with the current single-layer polycrystalline low-voltage logic device process technology. To achieve this high breakdown voltage, the process must be modified, which will increase the number 099103898 Form No. A0101. Page 5 / Total 106 pages 0992007313-0 201030947 Production costs. [0007] A new Single-Poly Flash Memory Cell with Low-Voltage and Low-Power Operations for Embedded Applications ("A New Single-Poly Flash Memory Cell with Low-Voltage and Low-Power Operations for Embedded Applications" CHI, ETAL., THE 5TH ANNUAL IEEE DEVICE RESEARCH CONFERENCE DIGEST, JUNE 1997, PP:126 -127, POSTED IEEEXLORE. IEEE.ORG : 2002 -08 -06 [0008] 21:21:45.0) The article is discussed in the triple Triple wells Single-layer polycrystalline flash memory cell structure under CMOS technology and new programming and erasing spiders with operating voltage not exceeding ±Vcc. _❸Single polycrystal... EPROM is fully compatible with standard C»0Sft; Cheng S. County has the disadvantages of high voltage operation, slow programming and inability to erase the power. The flash unit with this new programming/erasing scheme allows the system single-chip CMOS mixed-signal circuit to allow low voltage, low power. And the application of non-volatile Chinese body.. Guangrif.'!;ecr' US Patent 5, 929, 478 (?81^!3 to 6^:€^) describes a single layer of non-volatile memory. It consists of a ΐLu·, __ΡΕΤ and a substrate on the p
❹ (substrate )上Ν磊晶(N-epitaxial)層的兩個ρ井中製 造出來的電容器。P f型重滲雜(sinker)和N型埋層在兩 個P井中提供隔離。NVM器件的編程和抹除操作是靠fet偏 壓和電容器的電荷在一個導電層的進出來完成》該導電 層相當於該FET的浮動閘,從NVM上讀取資料是靠在電容 器施加一個讀取電壓時感應FET電流而完成。 [0009] 099103898 美國專利6, 992, 927和7, 164, 606(Poplevine,et al.)提供了一個NVM陣列,其包括一個四電晶趙pm〇s非 表單編號A0101 第6頁/共106頁 t 201030947 f發記憶體_單元,其具有共同連接的浮動閘。四個電 晶體的任何-個用於執行不同的控制、抹除、寫和讀取 的操作。故允許每-個器件被個別選擇而優化它不同的 操作。 闕美國專利 7,263,001(Wang,etal.)_T___ 錢體的單元和陣列。轉列把_單元用橫列和直行來 排列。每一個NVM單元包括一浮動閘、—編程元件和一邏 輯存儲元件。在編程和抹除模式下,每—個單元的浮動 閘被充電至一個預先決定的電壓值。開始讀取時,所有 存儲元件被預先充電到一個高電值β在預先充電後,讀 取被選擇的_以確定其元浮動 據對應 的位元值,原來預先充電的電位會被拉下β 【發明内容】 _本發明的,的是提供]龆脸__記憶體 元件,它與習知之單層多程相比,只是 多加了 一層或二層的製程和g /的電源電壓。 [0012] 本發明的另外一個目的是提供一種單層多晶浮動閘NVM記 憶體元件,其編程和抹除電壓均小於6. 0V。 [0013] 進一步地,本發明的另夕| —個目的是提供一種單層多晶 浮動閘NVM記憶艘元件’其編程和抹除操作操作可以利用 極低電流的Fowler-Nordheim邊緣隧道效應來完成。 [0014] 本發明的另一個目的是提供一種單層多晶浮動閘NVM記憶 體元件,其與習知的NVM記憶體元件相比體積較小。 099103898 表單編號 A0101 第 7 頁/共 106 頁 0992007313-0 201030947 [0015] 本發明的另一個目的是提供一種單層多晶浮動閘NVM記憶 體元件,其具有多次編程和抹除的能力。 [0016] 為了實現上述目的中的至少一個,一種單層多晶矽浮動 閘非揮發性記憶體元件必須包括一MOS電容器和一MOS電 晶體。該M0S電晶體的製造尺寸利用目前低電壓邏輯積體 電路製程來製造。該MOS電容器的第一個電極板和M0S電 晶體的閘極連接,以使得該M0S電晶體的閘極可以浮動以 形成浮動閘。該MOS電容器的第二個電極板通常為汲極擴 散、源極擴散和MOS電晶體的基極(bulk)。該MOS電晶體 的汲極連接到一個位元線電,蜃源,源極逮接到一個源極电容器 Substrate A capacitor fabricated in two p-wells of the N-epitaxial layer. The Pf type sinker and N-type buried layers provide isolation in the two P wells. The programming and erasing operations of the NVM device are performed by the fet bias and the charge of the capacitor in and out of a conductive layer. The conductive layer is equivalent to the floating gate of the FET. Reading data from the NVM is based on the capacitor applying a read. The FET current is sensed when the voltage is taken. [0009] 099103898 U.S. Patent Nos. 6,992, 927 and 7, 164, 606 (Poplevine, et al.) provide an NVM array comprising a four-electron crystal pm 〇s non-form number A0101 Page 6 of 106 Page t 201030947 f memory_unit, which has a floating gate that is connected in common. Any one of the four transistors is used to perform different control, erase, write, and read operations. Therefore, each device is allowed to be individually selected to optimize its different operations.阙 US Patent 7,263,001 (Wang, etal.)_T___ Units and arrays of money. The column arranges the _ cells in horizontal and straight rows. Each NVM unit includes a floating gate, a programming component, and a logic storage component. In programming and erase modes, the floating gate of each cell is charged to a predetermined voltage value. At the beginning of reading, all the storage elements are pre-charged to a high electric value β. After pre-charging, the selected _ is read to determine the bit value corresponding to its meta-floating data, and the original pre-charged potential is pulled down. SUMMARY OF THE INVENTION The present invention provides a face-to-face memory component that is more than one or two layers of process and g/supply voltage compared to conventional single-layer multi-pass. 0伏。 [0012] Another object of the present invention is to provide a single-layer polycrystalline floating gate NVM memory element, the programming and erase voltage are less than 6. 0V. [0013] Further, another object of the present invention is to provide a single-layer polycrystalline floating gate NVM memory vessel element whose programming and erasing operation operations can be completed by Fowler-Nordheim edge tunneling of extremely low current. . Another object of the present invention is to provide a single layer polycrystalline floating gate NVM memory element that is relatively small compared to conventional NVM memory elements. 099103898 Form Number A0101 Page 7 of 106 0992007313-0 201030947 Another object of the present invention is to provide a single layer polycrystalline floating gate NVM memory element with multiple programming and erasing capabilities. [0016] In order to achieve at least one of the above objects, a single-layer polysilicon floating gate non-volatile memory device must include a MOS capacitor and a MOS transistor. The fabrication dimensions of the MOS transistor are fabricated using current low voltage logic integrated circuit processes. The first electrode plate of the MOS capacitor is connected to the gate of the MOS transistor such that the gate of the MOS transistor can float to form a floating gate. The second electrode plate of the MOS capacitor is typically the drain of the drain, the source diffusion, and the bulk of the MOS transistor. The drain of the MOS transistor is connected to a bit line, the source, and the source is connected to a source.
或更大。 理尺寸比 線。該M0S電容器的 較是相對地大。本發 [0017] MOS電容器和M0S電晶體物理尺寸之間的大比率提供了一 個80%以上的大耦合比率。當一個電壓被施加於M〇s電容 器的第二電極板時,該,舖於腦電容器 的第二電極板的電壓的一外部翁 到身動閘極。電壓 施加於MOS電晶艘的源極或在M〇s電晶體的閘極 氧化區產生一個電場氧化區,從而啟動F〇wler — Nordheim的邊緣隧道效應。當在第二電極板的電壓是負 值、M0S電晶艘的沒極的電壓是正值時 ,浮動閘上的電荷 被吸取’以完成對該浮動閘非揮發性記憶體元件的編程 操作°另’右第二電極板的電壓是正值,而M0S電晶體汲 或源極或基;電壓是負值時,電子將會注射進入浮 動閘與正電何結合而完成對該浮動閘非揮發性記憶體元 件的抹除操作。 099103898 表單編珑A0101 第8頁/共1〇6頁 201030947 [0018] —種非揮發性記憶體元件的實施方式是在基板的表面先 形成一第一類導電型(N-型)的第一深擴散井。該第一深 擴散井連接到一第一深擴散井偏壓源、之後在第一深擴 散井裡形成一第二類導電型(P型)的第一淺擴散井和一第 二淺擴散井。第一淺擴散井連接到第一井偏壓源。第二 淺擴散井連接到第二井偏磨源。 [0019] [0020] 參 該非揮發性記憶體元件包括編程電躲合金屬氧化半 導體(MOS)電晶體,該編程電荷耦合金屬氧化半導體 (MOS)電晶體連接成為-個電容器。該編程電荷輛合麗 電晶體包括m多晶鄉成的編糾合浮動問。 該編程耦合浮動閘的尺寸用輪縱.浮_¥^聲_戈擴散 井之間提供一個大嶽率:舊y|M〇s電晶體 包括一第一類導電型雜質的源極和没桎源極和沒極 均擴散在第一淺擴散井》 . 該非揮發f±記隱艘TL件包辞^。該存儲 荷存儲浮動閉 。該電荷存儲浮動閉連接到:臟糖4閘以使得被 柄合到粞合浮動_電荷會錄_電荷_浮動閑。 該== 體包括—第一類導電型雜質的源極和波極 極均擴散在第二淺擴散井,極連接到 源極線偏魏,該賴連接_树偏H 電晶體的臨界電壓被設定到抹㈣界 ==::行抹除操作,存 ==:Γ電壓,對_發-憶 099103898 表單編號Α0101 第9頁/共106頁 0992007313-0 201030947 [0021] 對該非揮發性記憶體元件的編程操作是利用Fowler-Nordheira電荷隧道效應在電荷存儲浮動閘和存儲M0S電 晶體汲極之間的閘極氧化而完成。該非揮發性記憶體元 件的編程操作是由設定第一井偏壓源的電位從大约-7. 0V 到大約-5. 0V,且位元線偏壓源的電位從大约+ 5. 0V到大 約+ 7. 0V,以實現設定存儲M0S電晶體的臨界電壓到編程 臨界電位,其中該編程臨界電位是從大約0. 0V到大約 1. 0V。 [0022] 對該非揮發性記憶體元件的抹除操作是經由設定存儲M0S 電晶體的臨界電位到抹除臨界電位來完成。該抹狳臨界 電位是從大約3. 0V到大約4. 0V » ' 画 Λ [0023] 讀取該非揮發性記憶體元件時會檢測存棟Μ 0 S電晶體的臨 界電壓為編程臨界電位或是否為抹除臨界電位。當讀取 該非揮發性記憶體元件時,第二淺擴散井的電位被設定 為大於編程臨界電位而又小於缽除臨界電位,且位元線 偏壓源的電位被設定在大約+ 1. 0V。 [0024] 另一個實施方式中,該非揮·發性記憶體元件還包括一選 擇閘電晶體。該選擇閘電晶體包括一連接到一選擇閘控 制端的選擇閘。一第一類導電型雜質的源極擴散在第二 淺擴散井,同時還連接到源極線偏壓源。一第一類導電 型雜質的汲極為存儲M0S電晶體的源極。 [0025] 對該非揮發性記憶體元件的編程操作是經由設定第一井 偏壓源的電位從大約-7. 0V到大約-5. 0V,以及位元線偏 壓源的電位從大約+ 5. 0V到九約+ 7. 0V。另,該選擇閘控 099103898 表單編號Α0101 第10頁/共106頁 201030947 制端被設置在一個非激活電位以關閉該選擇閘電晶體, 從而使得該存儲MOS電晶體的臨界電壓被設定在編程臨界 電位 [0026] [0027] [0028] 099103898 當讀取該非揮發性記憶體元件時,會檢測存儲M〇s電晶體 的臨界電壓是否為編程臨界電位或為抹除臨界電位。在 讀取該非揮發性滅體元料’該第二歸散井的電位 被設置為大於編程臨界電位且小於抹除臨界電位。該位 元線偏壓源的偏歷電位被設定在大約+ UV。源極線偏壓 源被狀在接地參考電位,該選_控制端被設定在激 活電位以開啟該選擇卿髋,從叫取該非揮發性 記憶體元件的編程狀態 < 办 一 在另外,實施方式中,養帽件還包括 -在基板表層形成的第-類導電型雜料第二深擴散井 。該第二深擴散井連接到第二深井偏壓源一第二類導 電型雜質的第三淺擴散井存tffeto讎形成,且連 接到第一井偏電晶體在第三淺 擴散井裡形成。該絲電私括—由一單 層多晶料成的抹除耗合浮動閘》該抹_合浮動閘連 接到編_合浮動閘和電荷存儲浮動閘,且該抹除叙合 浮動閉的尺寸小於編_合浮動閘。該抹除電荷耗合膽 電晶體包括-第-類導電型雜質的源極㈣極該源極 和汲極均擴散在第三淺擴散井。 該抹除電糾合MGS電晶㈣料連接在抹除線偏壓源上 。該非揮發性記憶體元件的編程操作是由經過一個存在 於:?二獅閉和存儲職電晶體沒極之間娜 第11頁/共1.06頁Or bigger. The size ratio line. The MOS capacitor is relatively large. The large ratio between the physical dimensions of the MOS capacitor and the MOS transistor provides a large coupling ratio of 80% or more. When a voltage is applied to the second electrode plate of the M〇s capacitor, an external voltage applied to the second electrode plate of the brain capacitor reaches the body gate. The voltage is applied to the source of the MOS cell or to generate an electric field oxidation zone in the gate oxide region of the M〇s transistor, thereby initiating F边缘wler — Nordheim's edge tunneling. When the voltage at the second electrode plate is a negative value and the voltage of the MOSFET of the MOS plate is positive, the charge on the floating gate is drawn to complete the programming operation of the floating gate non-volatile memory device. In addition, the voltage of the right second electrode plate is positive, and the MOS transistor is 汲 or source or base; when the voltage is negative, the electron will be injected into the floating gate and positively combined to complete the non-volatile operation of the floating gate. Erase operation of the memory component. 099103898 Form Compilation A0101 Page 8/Total 1 Page 6 201030947 [0018] An embodiment of a non-volatile memory element is to first form a first conductivity type (N-type) first on the surface of the substrate. Deep diffusion well. The first deep diffusion well is connected to a first deep diffusion well bias source, and then a second conductivity type (P type) first shallow diffusion well and a second shallow diffusion well are formed in the first deep diffusion well . The first shallow diffusion well is connected to the first well bias source. The second shallow diffusion well is connected to the second well eccentricity source. [0020] The non-volatile memory component includes a programmed electrically-occluded metal oxide semiconductor (MOS) transistor that is connected as a capacitor. The programming charge of the Heli crystal includes the m-poly crystals. The size of the programming coupling floating gate provides a large ratio between the wheel vertical and the floating _¥^ sound_go diffusion well: the old y|M〇s transistor includes a source of the first type of conductivity type impurity and no source Both the pole and the finite pole are diffused in the first shallow diffusion well. The non-volatile f±remembered TL pieces are included. The storage load is floating closed. The charge storage is floatingly closed to: the dirty sugar 4 gate so that the handle is closed to the coupled float _ charge will be recorded _ charge _ floating idle. The == body includes—the source and the wave pole of the first type of conductivity type impurity are diffused in the second shallow diffusion well, and the pole is connected to the source line, and the threshold voltage of the connection _tree bias H transistor is set To wipe (four) bound ==:: row erase operation, save ==: Γ voltage, _ hair-recall 099103898 form number Α 0101 page 9 / total 106 page 0992007313-0 201030947 [0021] The non-volatile memory component The programming operation is accomplished by Fowler-Nordheira charge tunneling in the gate oxidation between the charge storage floating gate and the storage M0S transistor drain. The programming operation of the non-volatile memory device is set by the potential of the first well bias source from about -7. 0V to about -5.00V, and the potential of the bit line bias source is from about + 5. 0V to about The volts are about 0. 0V to about 1. 0V. The voltage is set to a threshold voltage of about 0. 0V to about 1. 0V. [0022] The erase operation of the non-volatile memory element is accomplished by setting a threshold potential for storing the MOS transistor to erase the critical potential. The eraser critical potential is from about 3.0 V to about 4. 0 V » ' 画 Λ [0023] When reading the non-volatile memory component, the threshold voltage is detected. The threshold voltage of the S transistor is the programmed critical potential or whether To erase the critical potential. When the non-volatile memory element is read, the potential of the second shallow diffusion well is set to be greater than the programming threshold potential and less than the critical potential, and the potential of the bit line bias source is set at approximately + 1. 0V. . [0024] In another embodiment, the non-volatile memory component further includes a select gate transistor. The select gate transistor includes a select gate coupled to a select gate control terminal. A source of a first type of conductivity type impurity diffuses in the second shallow diffusion well while also being connected to the source line bias source. A ruthenium of a first type of conductivity type impurity stores the source of the MOS transistor. [0025] The programming operation of the non-volatile memory device is performed by setting a potential of the first well bias source from about -7. 0V to about -5.00V, and the potential of the bit line bias source is from about + 5 . 0V to nine about + 7. 0V. In addition, the selection gate control 099103898 form number Α0101 page 10 / total 106 page 201030947 system terminal is set at an inactive potential to turn off the selection gate transistor, so that the threshold voltage of the storage MOS transistor is set at the programming threshold Potential [0026] [0028] 099103898 When reading the non-volatile memory element, it is detected whether the threshold voltage of the memory M?s transistor is the programming threshold potential or the erase critical potential. The potential of the second relegation well is read to be greater than the programming threshold potential and less than the erase critical potential. The bias potential of the bit line bias source is set at approximately + UV. The source line bias source is shaped at a ground reference potential, and the select terminal is set at an activation potential to turn on the selected hip, from calling the programming state of the non-volatile memory element < In the method, the cap member further includes a second deep diffusion well of the first type conductivity type impurity formed on the surface layer of the substrate. The second deep diffusion well is connected to a second shallow well of a second deep well bias source, a second type of conductive impurity, and is formed in the third shallow diffusion well. The wire is privately-used by a single layer of polycrystalline material, and the wiper is connected to the coded floating gate and the charge storage floating gate, and the eraser is closed and closed. The size is smaller than the _ _ floating gate. The erase charge-consuming bile transistor includes a source (four) of a -type conductivity type impurity, and the source and the drain are both diffused in the third shallow diffusion well. The erased electric matching MGS electro-crystal (four) material is connected to the erase line bias source. The programming operation of the non-volatile memory component is performed by one: between the two lions and the storage of the occupational crystals between the poles of the nath page 11 / 1.06
表單编號A0101 0992007313-0 201030947 區的Fowler-Nordheim電荷隧道效應而完成。該非揮發 性記憶體元件的編程操作是透過設定第一井偏壓源的電 位從大約-7. 0V到大約-5. 0V、位元線偏壓源設定電位從 大約+ 5. 0V到大約+ 7. 0V,以設定存儲M0S電晶體的臨界 電壓為編程臨界電位。該編程臨界電位是從大約〇 〇v到 大約+ 1.0V。 [0029] [0030] 099103898 對該非揮發性記憶體元件的抹除操作是透過在抹除耦合 序動閘和抹除電荷麵合M0S電晶體的通道(channe 1)區之 間的閘極氧化區發生Fowler-Nordheiffl電荷隧道效應, 且同時設置存儲M0S電蟲艘的臨界電麈為袜除臨界電位來The form number A0101 0992007313-0 201030947 is completed by the Fowler-Nordheim charge tunneling effect. The programming operation of the non-volatile memory device is set by the potential of the first well bias source from about -7. 0V to about -5.00V, the bit line bias source sets the potential from about + 5. 0V to about + 7. 0V, set the threshold voltage for storing the M0S transistor as the programmed critical potential. The programming threshold potential is from about 〇 〇v to about + 1.0V. [0030] 099103898 The erase operation of the non-volatile memory element is through a gate oxidized region between the channel (channe 1) region of the erased coupled gate and the erased charge-faced MOS transistor. Fowler-Nordheiffl charge tunneling occurs, and at the same time, the critical electric enthalpy for storing the M0S worm is set to the critical potential of the socks.
約+4. 0V。 除線偏壓 完成的。該抹除臨界電位是約 該非揮锋牲記憶髏元件的 源的電位設定為從大約-7. 0V到大約-5. 0^、將第一井偏 壓源的電位設定為從大約+ 5. 0V到+ 7. 0V、將第三井偏壓 源的電位設定為從大約-7.,0V到办約-5· 0Y、將第一深井 4n! 丨乂奸太刼+7. OV,從而使 ΙΓϊΙρ*! 偏壓源的電位設定為從大約汁5. 0' 产 Γ-、, i j·-® 得存儲M0S電晶體的轉界電彳壓,|>|琴抹除臨界電位 在讀取該非揮發性記憶體元件時’該存儲M0S電晶體的臨 界電壓被檢測以決定該非揮發性記憶體元件處於編程臨 界電位或抹除臨界電位《在讀取該非揮發性記憶體元件 時,該第二淺擴散井的電位被設定為大於編程臨界電位 而又小於抹除臨界電位。另,該位元線偏壓源的電位被 設定在大約+ 1. 0V。 另外一個實施方式中,該非揮發性記憶髏元件還包括一 第一類導電型(N型)雜質的第二深擴散井,該第二深擴散 表單編號A0101 第12頁/共106頁 [0031] 201030947 [0032] ❹ 井在基板的表層形成’該第二深擴散井還連接到第二深 擴散井偏壓源。該抹除電荷_合祕電晶體在第二深擴散 井的表層上形成。該抹除電荷_合MOS電晶體包括—單層 多晶石夕抹除耗合浮動閉’該單層多晶梦抹除_合浮動問 連接到編_合浮㈣和電荷存儲浮動閘。與編程輪合 浮動閘相比,雜除電料合浮動_尺寸較小。該抹 除電荷搞合MGS電晶體的源極和祕由第—類導電型雜質 形成’且擴散在第二深擴散井裡^該抹除電荷搞合M〇s電 晶體的源極和汲極連接到抹除線偏壓源。 對該非揮發性記憶撤元件的採除操作是透過發生在抹除 耦合浮動閘和抹除電荷耦督㈣矫通蟓區之間的閘 極氧化區發生Fowler-Norc^l^lii^^mineD效應 ,且將齊儲M0S電晶鷀鈞臨界電壓設置到抹除臨界電位來 完成的。該抹除臨界電位從大約+3 〇v到大約+4 〇v。需 要抹除該非揮發性記憶體;牛時該抹除夢偏壓源的電 位被設定在從大約-7. G V , 井偏壓源的電 位被設定在從大約+5,至,該第三井偏壓源 的電位被設定在從大約-7. 0V到大約-5· 〇v、第一深井偏 壓源的電位被設定在從大約+ 5. 〇 v到大約+ 7. 〇 v,以設定 存儲M0S電晶體的臨界電壓為抹除臨界電位。 [0033]又,另一實施方式中,該非揮發性記憶體元件包括一第 一類導電型(N型)雜質的共同深擴散井,該共同深擴散井 在一基板上形成’且連接到深井偏壓源》一第二類導電 型(P型)雜質的第一淺擴散井和一第二淺擴散井在共同深 擴散井裡形成。一電荷耦合M0S電容器在第一淺擴散井裡 099103898 表單編號A0101 第13頁/共106頁 0992007313-0 201030947 形成,其包括由-單層多晶料成 人〜 合浮動閘的尺寸使得相合比率非常大。二閉’該耗 雜質的源極和祕擴散在第—淺擴 二類導電型 均被連接到第-井偏㈣。續第 。4極和沒極 二井偏壓源。 堯擴散井被連接到第 [0034] [0035] 099103898 該非揮發性記憶趙元件包括一存儲_電晶體,該存儲 MOS電晶體包括一由一單層多晶矽形成的電荷存儲浮動閘 ,該電荷存儲浮動閘連接到耦合浮動閘,以使得電荷被 搞合到輕合浮動閘時會被收集在電荷存儲浮動閘上。—About +4. 0V. The line bias is removed. The erase potential is set to a potential of about -7. 0V to about - 5. 0. The potential of the first well bias source is set to be from about + 5. 0V to + 7. 0V, set the potential of the third well bias source to be from about -7., 0V to the office -5·0Y, the first deep well 4n! 丨乂太刼+7. OV, thus Let the potential of the ΙΓϊΙρ*! bias source be set from about 5. 0' Γ-,, ij·-® to store the switching voltage of the M0S transistor, |>|The erased critical potential is read When the non-volatile memory component is taken, the threshold voltage of the memory MOS transistor is detected to determine that the non-volatile memory component is at a programming threshold potential or an erase critical potential. When reading the non-volatile memory component, the first The potential of the two shallow diffusion wells is set to be greater than the programming threshold potential and less than the erase critical potential. In addition, the potential of the bit line bias source is set at approximately + 1. 0V. In another embodiment, the non-volatile memory device further includes a second deep diffusion well of a first type of conductivity type (N type) impurity, the second deep diffusion form number A0101 page 12 / 106 [0031] 201030947 [0032] The well is formed on the surface layer of the substrate. The second deep diffusion well is also connected to a second deep diffusion well bias source. The erased charge-closed crystal is formed on the surface of the second deep diffusion well. The erased charge _ MOS transistor includes - a single layer of polycrystalline smear wipes out the floating closed ‘the single layer polycrystalline dream erase _ _ float ask to connect to the _ _ float (four) and the charge storage floating gate. Compared with the floating wheel of the programming wheel, the hybrid material is floating and the size is small. The erased charge is combined with the source and secret of the MGS transistor and formed by the first conductivity type impurity and diffused in the second deep diffusion well. The erased charge combines the source and the drain of the M〇s transistor. Connect to the erase line bias source. The non-volatile memory removal component is operated by the Fowler-Norc^l^lii^^mineD effect occurring in the gate oxidation region between the erased coupling floating gate and the erased charge coupling (4). And setting the threshold voltage of the M0S transistor to the erase critical potential is completed. The erase critical potential is from about +3 〇v to about +4 〇v. The non-volatile memory needs to be erased; the potential of the erase bias source is set at about -7. GV, and the potential of the well bias source is set at about +5 to the third well. The potential of the bias source is set from about -7. 0V to about -5 · 〇v, and the potential of the first deep well bias source is set from about + 5. 〇v to about + 7. 〇v to set The threshold voltage for storing the MOS transistor is the erase critical potential. [0033] In still another embodiment, the non-volatile memory element includes a common deep diffusion well of a first type of conductivity type (N type) impurity, and the common deep diffusion well forms 'on a substrate and is connected to the deep well A first shallow diffusion well of a second type of conductivity type (P type) impurity and a second shallow diffusion well are formed in a common deep diffusion well. A charge coupled MOS capacitor is formed in the first shallow diffusion well 099103898 Form No. A0101 Page 13 / 106 pages 0992007313-0 201030947, which consists of a single-layer polycrystalline adult ~ floating gate size such that the coincidence ratio is very large . The second closed 'source of the impurity and the secret diffusion in the first - shallow expansion of the second type of conductivity are connected to the first well (4). Continued. 4-pole and immersed two-well bias source. The 尧 diffusion well is connected to the [0035] [0035] 099103898 The non-volatile memory element includes a memory _ transistor including a charge storage floating gate formed by a single layer polysilicon, the charge storage floating The gate is connected to the coupled floating gate so that the charge is collected on the charge storage floating gate when it is engaged to the light-closed floating gate. -
第一類導電型雜質的琢輟和级接擴散在第二淺擴散井。The enthalpy and cascade of the first type of conductive impurities diffuse in the second shallow diffusion well.
位元線偏 抹除臨界電 該源極連接到源極線偏壓源kil^· 壓源。該存儲MOS電晶體的唪疥靄赛 位以對該非揮發性記億體元件進行抹除操作;該存儲M〇s 電晶體的臨界電壓被設置到編程臨界電位,以對該非揮 發性記憶體元件進行編程操作。『 對該非揮發性記憶體元件過發生在電荷存 儲浮動閘和存儲M0S電晶的閘極氧化區發生Bit line offset Erasing critical power The source is connected to the source line bias source kil^·. The memory of the memory MOS transistor is erased by the non-volatile memory element; the threshold voltage of the memory M?s transistor is set to a programming threshold potential to the non-volatile memory element Perform programming operations. 『This non-volatile memory component occurs over the gate storage oxidation gate of the charge storage floating gate and the memory M0S transistor.
Fowler-Nordheim電荷隧道效應而完成。要完成對該非 揮發性記憶體元件的編程操作,則必須設定第一井偏壓 源的電位從大約-7. 0V到大約-5. 0V,同時還須設定位元 線偏壓源的電位從大約+5. 0V到大約+7. 0V ’以設定存儲 M0S電晶體的臨界電位為編程臨界電位。該編程臨界電位 是從大約0. 0V到大約+ 1. 0V。 對該非揮發性記憶體元件的抹除操作是將該存儲M0S電晶 體的臨界電壓設定為抹除臨界電位。該抹除臨界電位從 表單編號A0101 第14頁/共106頁 0992007313-0 [0036] 201030947 大约+ 3. 0V到大約+ 4. 0V。 [0037] 在讀取該非揮發性記憶體元件時,該存儲M0S電晶體的臨 界電壓被檢測以決定該非揮發性記憶體元件是位於編程 臨界電位或是位於抹除臨界電位。在讀取該非揮發性記 憶體元件時,該第二淺擴散井的電位被設定為大於編程 臨界電位而又小於抹除臨界電位。又,該位元線偏壓源 的電位被設定在大約+ 1. 0V。 [0038] 另一個實施方式中 參 [0039] 099103898 成,該基板的表面包括一第一類導電型(N型)雜質的深擴 散井,該深擴散井連接到一渾f笮壓繆?.一第二類導電 型(P型)雜質的淺擴散井在 < 形成。該非 揮發性記憶體元件包括一電f私)半導體 (M0S)電晶體,該電荷耦合金屬氧化物半導體電晶體連接 作為一電容器。該電荷耦合M〇s電晶艘:包括一層由單層多 ::二:合浮動閘織,:酿寸必須使得 夺 擴散井的&科我齡f。-第-類導 電型雜質料極和没極 共同地連接到★擴散井和第-井偏壓源。 該非揮發性記 腸電晶想包括Γ 存儲嘱晶艘,該存健 該電荷存儲叫=1層多晶料成㈣財料動問, 合到轉合赫p _合浮動閘,以使得電荷被搞 ° ° 3會被收集到電荷存儲浮動閘上。一第一 類導電㈣質…她和—祕在基板裡 接到源極線偏壓湃^ ^ 心原極連 儲廳電曰體Γ該没極連接到位元線偏壓源。該存 表單編號二界電龍設置到—個抹除臨界電位,以 第15頁/共106頁 0992007313-0 201030947 對該非揮發性記憶想元件進行抹除操作;該存儲M〇s電晶 體的臨界電壓被設置到一編程臨界電位,以對該非揮發 性記憶體元件進行編程操作。 [0040] 對該非揮發性記憶體元件的編程操作是透過在電荷存儲 浮動閘和存儲M0S電晶體汲極之間的閘極氡化區發生Fowler-Nordheim charge tunneling is done. To complete the programming operation of the non-volatile memory component, the potential of the first well bias source must be set from about -7. 0V to about -5.00V, and the potential of the bit line bias source must also be set from Approximately +5.0 V to about +7. 0 V ' is set to store the critical potential of the M0S transistor as the programmed critical potential. The programming threshold potential is from about 0. 0V to about + 1. 0V. The erase operation of the non-volatile memory element is to set the threshold voltage of the memory MOS transistor to the erase critical potential. The erased critical potential is from the form number A0101 page 14/106 page 0992007313-0 [0036] 201030947 approximately + 3. 0V to approximately + 4. 0V. [0037] Upon reading the non-volatile memory component, the critical voltage of the memory MOS transistor is sensed to determine whether the non-volatile memory component is at a programmed critical potential or at an erase critical potential. When the non-volatile memory element is read, the potential of the second shallow diffusion well is set to be greater than the programming threshold potential and less than the erase critical potential. Also, the potential of the bit line bias source is set at approximately + 1. 0V. [0038] In another embodiment, the surface of the substrate includes a deep diffusion well of a first type of conductivity type (N type) impurity, and the deep diffusion well is connected to a 笮f笮 pressure 缪? A shallow diffusion well of a second type of conductivity (P type) impurity is formed in <. The non-volatile memory component includes an electrical (poly) semiconductor (MOS) transistor connected as a capacitor. The charge-coupled M〇s electro-crystalline ship: consists of a single layer consisting of a single layer: ::two: a floating damper: the brewing inch must make the diffusion well & - The first type of conductive impurity source and the pole are commonly connected to the ★ diffusion well and the first well bias source. The non-volatile enteroelectric crystal crystals are intended to include a Γ storage 嘱 crystal boat, and the storage of the charge is called =1 layer polycrystalline material into (four) financial entanglement, and is coupled to the turn-in hp _ floating gate to make the charge Engage ° ° 3 will be collected on the charge storage floating gate. A first type of conductive (four) quality... She and - secret in the substrate are connected to the source line bias 湃 ^ ^ Xinyuan pole connected to the storage hall, the pole is connected to the bit line bias source. The memory form number is set to - erase the critical potential, and the non-volatile memory element is erased by page 15 / page 106 0992007313-0 201030947; the criticality of the memory M〇s transistor The voltage is set to a programmed threshold potential to program the non-volatile memory element. [0040] programming operation of the non-volatile memory element occurs through a gate deuteration region between the charge storage floating gate and the storage MOSFET dipole
Fowler-Nordheim電荷隧道效應而完成。該非揮發性記 憶體元件的編程操作需要設置第一井偏壓源的電位從大 約-5. 0V到大約-7. 0V '設置位元線偏壓源的電位從大約Fowler-Nordheim charge tunneling is done. The programming operation of the non-volatile memory element requires setting the potential of the first well bias source from about -5.0 V to about -7. 0 V 'setting the potential of the bit line bias source from about
+ 5. 0V到大約+ 7. 0 V ’以設定存儲JI0S電晶體的臨界電壓 到編程臨界電位。...該.編程眩界電位是從大約〇. 〇 V到大約 [0041] 1.0V。 該非揮發牲記憶體元件的 ^ ; 抹#'德1頁#i 存儲Μ 0 S電晶 體的臨界電壓到抹除臨界電位從大約+3. 〇V到大約+4. 〇v 。讀取該非揮發性第憶體元件時’先要檢測存儲M〇s電晶 體的臨界電壓是位於編程懦_讎抹除臨界電 位。在讀取該非揮發性記二淺擴散井的 電位被設定在大铃編程臨θΙβ®#小於抹除臨界電位 。另’該位元線偏壓源的電位設定在大約+ 1〇V。 [0042]+ 5. 0V to approximately + 7. 0 V ' to set the threshold voltage for storing the JI0S transistor to the programmed critical potential. ...the programming glare potential is from about 〇. 〇 V to about [0041] 1.0V. ^ ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' S S S S S S S S S S S S S S S S S S S S S S S S S S S S S S S S 临界When reading the non-volatile memory element, the threshold voltage for detecting the memory of the M〇s is first detected by programming the 临界_雠 erase threshold. The potential of the non-volatile recording shallow diffusion well is set in the large bell programming pro θ Ι β ® # is less than the erase critical potential. The potential of the bit line bias source is set at approximately + 1 〇V. [0042]
另一實施方式t,一非揮發性記憶體元件在基板上形成 ,該基板的表面包括一第一類導電型(^型)雜質的深擴散 井,該深擴散井連接到一深井偏壓源。一第二類導電型 (P型)雜質的淺擴散井在共同深擴散井之内形成。該非揮 發性記憶體7G件包括一電荷耦合金屬氧化物半導體(M〇s) 電晶體,該電荷耦合金屬氧化物半導體電晶體連接作為 電容器。該電荷耦合M0S電晶體包括一由單層多晶矽形成 099103898 表單編號A0101 第16頁/共106頁 0992007313-0 201030947 =::::一其:寸使得從浮動— 在 、 料電型雜冑㈣'歸汲極擴散 [0043] [0044] 散井。該源極和汲極共同地連接到淺擴散井。 性記憶體元件還包括—電料合_電容器。該 =何tMGS電容器包括—由單層多晶料成_合浮動 ^其大小必驗得從浮動__散井_合比率是 播嵌* A帛—料€型雜f的_和祕擴散在淺 :散井。贿極源極和淺擴散井共同地連接到第一井偏 麼源。該存麵5“體_界轉被設"抹除臨界電 位以抹除非揮發性記憶|元件、若被設置到編程政界電 ^ ^ # # ^ ^ ^ ^ ^ 對該非,牲記憶體元件的^電荷存儲 浮動閘和存健MOS電晶體汲極之間的閘極氧化區發生 Fowler-Nordheim電荷隧道效應來完咸。F〇wUr_In another embodiment, a non-volatile memory component is formed on a substrate, the surface of the substrate includes a deep diffusion well of a first type of conductivity type, and the deep diffusion well is connected to a deep well bias source. . A shallow diffusion well of a second type of conductivity type (P type) impurity is formed within a common deep diffusion well. The non-volatile memory 7G device includes a charge coupled metal oxide semiconductor (M〇s) transistor that is connected as a capacitor. The charge coupled MOS transistor comprises a single layer polycrystalline germanium formed 099103898 Form No. A0101 Page 16 / 106 pages 0992007313-0 201030947 =:::: One of its: inch makes the floating - in, electricity type (4)汲 扩散 diffusion [0043] [0044] scattered wells. The source and drain are commonly connected to a shallow diffusion well. The memory device component also includes an electrical material capacitor. The =he tMGS capacitor includes - from a single layer of polycrystalline material into a _heavy float ^ its size must be verified from the floating __ 散 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Shallow: scattered wells. The bribe source and the shallow diffusion well are connected together to the first well source. The storage surface 5 "body_bound is set to " erase the critical potential to erase unless the volatile memory|component, if set to the programming political power ^ ^ # # ^ ^ ^ ^ ^ to the non, the memory component ^Fowler-Nordheim charge tunneling occurs in the gate oxidation region between the charge storage floating gate and the drain of the MOS transistor. F〇wUr_
Nordherm電_道效應被贈細發性記憶體元 件進行編程操作,此時需壓源的電位從 大約-7. 0V到大約-5· 〇'、⑩義^壓源的電位從大約 + 5.0V到大約+7.0V,以設置存儲M〇s電晶體的臨界電壓 到編程臨界電位。該編程臨界電位是從大約〇. 0V到大約 1.0V。 [0045] 對該非揮發性記憶體元件的抹除操作是透過設定存儲M〇s 電晶體的臨界電壓為抹除臨界電位。該抹除臨界電位是 從大約+3. 0V到大約+4. 0V。 [0046] 在讀取該非揮發性記憶體元件時,該存儲m〇S電晶體的臨 099103898 表單煸號A0101 第17頁/共106頁 0992007313-0 201030947 I電壓被檢測以决定該非揮發性記憶體元件是位於編程 臨界電位或是位㈣㈣界電位。在讀取該非揮發性記 隐體兀件時’第—淺擴散井的電位被設S大於編程臨界 電位而又小於抹除臨界電位。另,該位元線偏壓源的電 位u在大約+1. GV,並且源極線偏I源被設置到接地 參考電位。 [0047] [0048] 099103898 ⑬ 刀”㈣式中,在基板上形成一非揮發性記憶體元 件其在基板的表面包括—第一類導電型⑽)雜質的第 -深擴散井和第二深擴散井…第二類導電型㈣)雜質 的第-淺擴散井和第=淺擴散井分別形成於第—深擴散 井和第二轉散井之件包括- 電魏合金屬氧化物半導體轉I,,矮電荷耗合金 屬氧化物___連接作為―電‘.‘:該電晶體包 括一由單層多㈣義_合軸閘,料合㈣非常 的大。-第雜h〒和餘均擴散在第一 淺擴散井。該沒極、源極^1^^井共同地連接到 气井錢源。該第'二淺第二井偏壓源。 該第一深擴散井連接到第—深井ϋ源。”二深擴散 井連接到第二深井偏壓源。 、 該非揮發性記憶禮元件包括一存儲_電晶體,其包括一 由單層多晶料成㈣荷存料動閘,該錢 閉連接到編㈣合浮動閘,以使得電荷_合_= 動閘會被收集在電荷存财動閘上。_第—類導電 質的源極錄在第二_散井而且連接親極線偏壓源 。-第-類導電型雜質的汲極擴散在第二淺擴散井而且 表單編號A0101 第頁/共106頁 0992007313-0 連接到一位元線偏壓源。該非揮發性記憶體元件還包括 -選擇閘極MOS電晶體’其包括一連接到選擇閘極控制終 端的選擇閘極。一第一類導電型雜質的源極擴散在第二 淺擴散井而且連接到源極線偏壓源。該選擇閘極M〇s電晶 體包括一第一類導電型雜質的汲極,該汲極同時為存儲 MOS電晶體的源極。該存儲M〇s電晶體的臨界電壓被設置 為一抹除臨界電位,以抹除非揮發性記憶體元件;或被 設置到一編程臨界電位,以對該非揮發性記憶體元件進 行編程操作。 對該非揮發性記憶體元件_鏞程操作袭透過在電荷存儲 浮動閘和存儲Μ 0 S電晶艎汲"* Fowler-Nordheim^:^ ι區發生 owler- N o r d h e i m電荷隧道效應被激活從而可以對該非揮發性記 憶體元件進行編程操作,此時需要設置第一井偏壓源的 電位從大約-7.GV到大約源的電位從 大約+5,到大約+7.膊’槪g㈣!'電晶趙的臨界 電麼為編程臨界_^該^_|;彳|從大約Q廣到大 約 1. 0V。 、 、一 對該非揮發性記憶體元件的抹除操作是透過在抹除耦合 浮動閘和抹除電荷耦合M0S電晶體通道區之間的閘極氧化 區發生Fowler-Nordheim電荷隧道效應而完成的,同時 還需要設置存儲M0S電晶體的臨界電壓為抹除臨界電位。 該抹除臨界電位從大約+ 3. 0V到大約+4, ^該非揮發性 記憶體元件的抹除需要設置第一井偏壓源的電位從大約 + 5. 0V的大約+7. 〇V、第二井偏壓源的電位從對大約— 表單編號A0101 第19.頁/共106頁 201030947 7. 0V到大約-5. 〇V、第一深井偏壓源的電位從大約+5 〇v 的大約+ 7. 0V、並且源極線偏壓源的電位從大約_7 到 大約-5. 0V,以設置存儲M0S電晶體的臨界電壓為編程臨 界電位。 [0051] [0052]The Nordherm electric_channel effect is programmed by a fine-grained memory element. The potential of the voltage source is from about -7. 0V to about -5. The potential of the voltage source is about + 5.0V. To approximately +7.0V, to set the threshold voltage of the M〇s transistor to the programmed critical potential. The programming threshold potential is from about 〇.0V to about 1.0V. [0045] The erasing operation of the non-volatile memory element is to erase the critical potential by setting a threshold voltage for storing the M〇s transistor. 0伏。 The erased threshold potential is from about +3.00V to about +4. 0V. [0046] When reading the non-volatile memory element, the memory of the m〇S transistor is 099103898, the form number A0101, page 17 / 106 pages 0992007313-0 201030947 I voltage is detected to determine the non-volatile memory The component is located at the programmed critical potential or at the bit (four) (four) boundary potential. When reading the non-volatile cryptographic element, the potential of the first-diffusion well is set to S greater than the programmed critical potential and less than the erase critical potential. In addition, the potential u of the bit line bias source is about +1.0 GV, and the source line bias I source is set to the ground reference potential. [0048] 099103898 13 In the formula (4), a non-volatile memory element is formed on the substrate, and the surface of the substrate includes a first-type conductivity type (10) impurity-first deep diffusion well and a second deep Diffusion wells... The second type of conductivity type (4)) The first shallow shallow diffusion well and the first shallow shallow diffusion well are formed in the first deep diffusion well and the second reverse diffusion well, respectively - the electrical Weihe metal oxide semiconductor is transferred to I ,, the short charge-consuming metal oxide ___ is connected as "electric'.": The transistor consists of a single-layer multi- (four) sense-closed gate, and the material (four) is very large. - the first h and the remaining Both are diffused in the first shallow diffusion well. The immersed, source ^1^^ well is commonly connected to the gas well money source. The first 'two shallow second well bias source. The first deep diffusion well is connected to the first Deep well source.” The second deep diffusion well is connected to the second deep well bias source. The non-volatile memory device includes a memory_transistor including a single layer of polycrystalline material (4), a charge storage gate, and the money is connected to the (four) floating gate to make the charge_合_= The gate will be collected on the charge storage gate. The source of the first-type conductivity is recorded in the second_well and connected to the polar line bias source. - The drain of the first-type conductivity type impurity diffuses in the second shallow diffusion well and the form number A0101 is connected to a one-line bias source. The non-volatile memory component further includes - a select gate MOS transistor 'which includes a select gate connected to the select gate control terminal. A source of a first type of conductivity type impurity diffuses in the second shallow diffusion well and is connected to the source line bias source. The select gate M〇s transistor includes a drain of a first type of conductivity type impurity, which is also the source of the memory MOS transistor. The threshold voltage of the memory M?s transistor is set to erase the critical potential to erase the volatile memory element; or is set to a programmed threshold potential to program the non-volatile memory element. The non-volatile memory component is activated by the owler-N ordheim charge tunneling effect in the charge storage floating gate and the storage Μ 0 S electro-crystal 艎汲 "* Fowler-Nordheim^:^ ι area The non-volatile memory element is programmed to operate, and the potential of the first well bias source needs to be set from about -7. GV to about the source potential from about +5 to about +7. 槪'槪g (four)! The criticality of the electric crystal Zhao is the programming threshold _^ the ^_|; 彳| from about Q wide to about 1. 0V. And a pair of non-volatile memory elements are erased by Fowler-Nordheim charge tunneling in the gate oxide region between the erased floating gate and the erased charge coupled MOS transistor channel region. At the same time, it is also necessary to set the threshold voltage for storing the M0S transistor to erase the critical potential. The erasing threshold potential is from about +3.00V to about +4, ^ the erasing of the non-volatile memory element requires setting the potential of the first well bias source from about +1.0V to about +7. 〇V, The potential of the second well bias source is from approximately - Form No. A0101, page 19. / page 106, 201030947, 7. 0V to approximately -5. 〇V, the potential of the first deep well bias source is approximately +5 〇v Approximately +7.00V, and the potential of the source line bias source is from about _7 to about -5.00V, to set the threshold voltage for storing the MOS transistor to the programmed critical potential. [0052] [0052]
在讀取該非揮發性記憶體元件時,該存儲M0S電晶體的臨 界電壓被檢測以決疋該非揮發性記憶體元件是位於編程 臨界電位或是位於抹除臨界電位。在讀取該非揮發性記 憶體元件時,該第一淺擴散井的電位被設定為大於編程 臨界電位而又小於抹除臨界·電位》另,該位元線偏壓源 的電位被設定在大約+m:’:並1濂極線偏壓源被設置為 接地參考電位、選擇閘極控,^^被活電位以 啟動選擇閘極M0S電晶體,^讓非性記憶體元 件的編程狀態。 【實施方式】 ❹ 請參閱囷1 ’王(Wang)等人埤卷了一單層辛晶浮動閘單電 晶體快閃記憶想的结構。gp y包括一第一編程電 晶體M106、一第二編程電儲電晶體M102。 在讀取操作開始時,預先充電電路108用一電源電麗源 VDD對存儲電晶體M102進行預充電。第一編程電晶體 M106和第二編程電晶體M104合起來形成NVM單元C100的 編程單元。第一編程電晶體M106和第二編程電晶體Mi〇4 的閘極耦合在一起並且與存儲電晶體Ml02共用浮動結11〇 。第一編程電晶體M106的源極、汲極和N井終端一起被耗 合到第一編蘀電壓源VEE。第二編程電晶體Ml 〇4的源極、 汲極和N井終踹一起被耦合到第二編程電壓源vpp » 099103898 表單编號A0101 第20頁/共106頁 0992007313-0 201030947 [0053] [0054] Ο ❹[0055] 浮動結Π 〇储存被注入的電子Upon reading the non-volatile memory component, the critical voltage of the memory MOS transistor is sensed to determine whether the non-volatile memory component is at a programmed critical potential or at an erase critical potential. When reading the non-volatile memory element, the potential of the first shallow diffusion well is set to be greater than the programming threshold potential and less than the erase critical potential. In addition, the potential of the bit line bias source is set to approximately +m:': The 1 濂 line bias source is set to the ground reference potential, select the gate control, ^^ is activated to activate the select gate MOS transistor, and let the programming state of the non-volatile memory component. [Embodiment] ❹ Please refer to 囷1 ’Wang et al. for a single-layer symplectic floating gate single crystal flash memory. The gp y includes a first programming transistor M106 and a second programming capacitor M102. At the beginning of the read operation, the precharge circuit 108 precharges the storage transistor M102 with a power source VDD. The first programming transistor M106 and the second programming transistor M104 are combined to form a programming unit of the NVM cell C100. The gates of the first programming transistor M106 and the second programming transistor Mi〇4 are coupled together and share a floating junction 11〇 with the storage transistor M102. The source, drain and N well terminals of the first programming transistor M106 are consumed together with the first compiled voltage source VEE. The source, drain and N well of the second programming transistor M1 〇4 are coupled together to a second programming voltage source vpp » 099103898 Form No. A0101 Page 20 of 106 Page 0992007313-0 201030947 [0053] 0054] Ο ❹[0055] Floating Π 〇 Storage of injected electrons
何,目的是用於調整NVM n的臨界·⑽。 ^量會提升存儲電議1Q2的臨界電壓1反減少 儲存在洋動結110的電子数量 界電膨$姐 重會降低存儲電晶體_2的臨 電壓。第-編程電晶舰06的尺寸必 電晶體剛衫議軸2⑽的財規模編程 =程操作時,Fowler_Nc)rdheim通道随道效應會發生 第二編程電晶舰〇4通道和閉極之間的氧化區。當第 =程電細㈣峨蛛1询:編程電絲 冲經過第二編程電晶雄_4故注入到浮動結⑴,以辦 加存儲電議m臨界電略^作時,一曰 反向Fowler-Nordheim隧^應戈發糸在第二編程 體Μι_评和浮動結u°之間電子從; 動結110離開經過厚度小於1G喝第二褊程電晶簡〇4 的閘極氧化區而到達第二竹電鲁㈣在抹除 後’ _s存儲電晶體' ts編程操作時注射電子 進入到浮動結110是透過將第一編程電麼源VEE設置—高 電壓+ 1G. GV到-個把較大的第—編程電晶舰G6的沒極 、源極和在-起的終端點,並且透過將第二編程 電壓源VPP設置接地參考電壓(00v)到一個把較小的第 二編程電晶體M104的汲極、源極和N井都黏在一起的終端 點來達成的。該接地參考電壓(〇. 〇v)被施加於輸出位元 線Vo(BL)和連接到NMOS存儲電晶體jn〇2源極的字元線電 壓源WL。由於第一編程電晶體Mi〇6的尺寸大於第二編程 099103898 表單編號A0101 第21頁/共1〇6頁 0992007313-0 201030947 電晶體M104和NM0S存儲電晶體Ml〇2合起來的尺寸的四到 五倍’浮動結110的電位大約為9 QV。電子從第二編程電 晶體M104透過Fowler-Nordheinrit道隧道被吸引到浮動 結no。在編程操作以後’ NM0S存儲電晶體M1〇2的臨界 電壓會增加。 [0056] 〇 在抹除操作時,+10. ον的高電位從第二編程電壓源vpp 到較小的第二編程電晶體Mi〇4的源極、汲極和N井終端。 接地參考電壓(0. 0V)被施加於較大的第一編程電晶體 M106的源極 '汲極井終端、字元線電壓源札、NM〇s 存儲電晶魏M102的源極、输也位元線恥(社)以及NM〇s存 儲電晶體Ml 〇2‘的汲極。基於气七:偏‘壓碟壓浮動結 110的%處於+1V附近,等皆編、祥草晶體以〇4的第 二編程電麝源VPP和浮動結lio之間的壓降大約為+ 9V。 由於第二編程電晶體U1.04的閘極和n井之間的f〇w 1 er_ Nordheim通道隧嗔效應,芦子麵結^的電子從浮動結 110經過第二編程電晶艘Mlj4要"吸#今丨涂二編程電壓源 vpp。在袜除操作以後’該電晶體M102的臨界 'I! ί* 電壓會下降。 [0057] Wang等所描述的抹除操作和編程操作是根據F〇wler_What is the purpose of adjusting the criticality of NVM n (10). ^ The amount will increase the threshold voltage of the storage electricity 1Q2 and reduce the amount of electrons stored in the oceanic junction 110. The expansion will reduce the voltage of the storage transistor_2. The size of the first-programmed electro-crystalboard 06 must be the crystal size of the crystal shirt 2 (10). When the operation is performed, the Fowler_Nc)rdheim channel will have a second programming electro-optical ship between the 4 channel and the closed pole. Oxidation zone. When the first step is fine (four) 峨 spider 1 inquiry: the programming wire is rushed through the second programming electric crystal _4, so injected into the floating knot (1), in order to add storage electricity m critical power slightly, a reverse Fowler -Nordheim tunneling electrons from the second programming body Μι_ evaluation and floating junction u°; moving junction 110 leaves the gate oxidation region of thickness 2 less than 1G and drinking the second process electron crystal After reaching the second bamboo electric Lu (four) after the erasing '_s storage transistor' ts programming operation, the injection of electrons into the floating junction 110 is set by the first programming power source VEE - high voltage + 1G. GV to - The larger first-programmed crystallizer G6 has the pole, source and at-end termination points, and sets the ground reference voltage (00v) to a second programming voltage by placing the second programming voltage source VPP The end point of the crystal M104 whose drain, source and N well are stuck together is achieved. The ground reference voltage (〇. 〇v) is applied to the output bit line Vo(BL) and the word line voltage source WL connected to the source of the NMOS storage transistor jn〇2. Since the size of the first programming transistor Mi〇6 is larger than the second programming 099103898 Form No. A0101 Page 21/Total 1 Page 6 0992007313-0 201030947 The crystal M104 and the NM0S storage transistor Ml〇2 are combined in size four The potential of the five-fold 'floating junction 110' is approximately 9 QV. Electrons are attracted from the second programming transistor M104 through the Fowler-Nordheinrit tunnel to the floating junction no. The threshold voltage of the NM0S memory transistor M1〇2 increases after the programming operation. [0056] 〇 During the erase operation, the high potential of +10. ον is from the second programming voltage source vpp to the source, drain and N well terminals of the smaller second programming transistor Mi〇4. The ground reference voltage (0. 0V) is applied to the source of the larger first programming transistor M106, the terminal of the terminal well, the voltage source of the word line, the source of the NM〇s storage transistor M102, and the source The bit line shame (she) and the NM〇s store the drain of the transistor Ml 〇2'. Based on the gas seven: the partial pressure of the pressure plate 110 is near +1V, and the voltage drop between the second programming power source VPP and the floating junction lio of the 祥4 crystal is about +9V. . Due to the tunneling effect of the f〇w 1 er_ Nordheim channel between the gate of the second programming transistor U1.04 and the n-well, the electrons of the azimuth junction are passed from the floating junction 110 through the second programmed electro-crystallizer Mlj4. ; suction #今丨涂二programming voltage source vpp. After the socking operation, the critical 'I! ί* voltage of the transistor M102 drops. [0057] The erase operation and the programming operation described by Wang et al. are based on F〇wler_
Nordheim通道隧道效應來完成的。該單層多晶快閃Ννΐί 單元C100有低電流編程和抹除操作的好處。然,它具有 以下缺點:要求+ 10Vi^j壓才能有效地執行Fowler-Nordheim通道抹除和Fow ler-Nordheim通道編程。 + 10V的高壓要求太高,其無法從只能提供+ 6.0V擊穿電 壓(breakdown voltage)設備的大多数單層多晶邏輯 099103898 表單編號A0101 第22頁/共106頁 0992007313-0 201030947 製程得到》 [0058] 玟,有必要提供一種與單 電壓兼容的單Ma 製程和低電源操作 電壓兼—衫晶浮動閘Nv時儲單元 浮動闡NVM在蚀s二.... 千·Nordheim channel tunneling is done. This single-layer poly-crystal flash Ννΐί unit C100 has the benefits of low current programming and erase operations. However, it has the following disadvantages: a +10Vi^j pressure is required to effectively perform Fowler-Nordheim channel erasing and Fowler-Nordheim channel programming. + 10V high voltage requirements are too high, it can not be obtained from most single-layer polycrystalline logic 099103898 of the device that can only provide +6.0V breakdown voltage. Form No. A0101 Page 22 / 106 Page 0992007313-0 201030947 》 [0058] 玟, it is necessary to provide a single-Ma process compatible with a single voltage and a low power supply operating voltage and a crystal floating gate Nv when the storage unit floats NVM in Eclipse s....
[0059] -----",《丨甘两早兀^該單居 洋動間圆存儲單元的編程和抹_作的錢層在 + 6.糊例肩之間。該單層多晶浮細儲 要用到非常低的電流,以透過 道來完成編程和抹除操作。進-步需要的是,該單層多 晶浮動閘_存儲單元的難氡化區厚度小於刚α = 尺寸比習知_存儲單元的物理尺寸要小。值得注意的是 ’卓層多晶發浮動閉的特點是它被認為是一個金屬氧化 物半導體陶電晶體。單卿驗不是金屬,然 ,其仍然可Μ傳導電^具;場效应電 TMOS 電晶體的這一特點。 f _ 2^胃_ «性記憶體 2〇〇在Ρ型基板205上形成。的上表面形成 - N型雜質的第-深擴散井__瓣一深擴散井21〇透過 一N型連接點連接到一第一深井偏壓源265 ◊該?型基板是 經過連接擴散點270連接到接地參考電壓源(〇. 〇v)。該 第一深擴散井210是在當前單層多晶矽邏輯積體電路製程 上增加一層製程的微小變化而得到。 晶 [0060] 099103898 在第一深擴散井210之内形成一 p型雜質的一第一淺擴散 井215和一第二淺擴散井220 ^該第一和第二淺擴散井 215和220是在當前單層多晶秒邏輯積醴電路的製程基礎 上另外增加的一層製程而得到。 表單編號A0I01 第23頁/共106頁 0992007313-0 201030947 [0061] 一編程電荷耦合金屬氧化半導體(MOS)電晶體225被連接 作為一電容器。該編程電荷耦合M0S電晶體225形成在第 一淺擴散井215的表面並且包括都擴散在第一淺擴散井 215内的N型雜質的源極226和汲極227。該源極226、汲 極227和第一淺擴散井215共同連接到第一淺井偏壓源 VTPW1 230。該P型連接擴散點216連接第一淺擴散井 215到第一淺井偏壓源VTPW1 230。該第一淺擴散井215 透過連接擴散點216連接到第一淺井偏壓源VTPW1 230 。該閘極氧化區229上面具有由單層多晶矽形成的編程耦 合浮動閘228。該閘極氧化區229的材料和厚度與當前 CMOS領域常用的技術是一[0059] -----", "丨甘两早兀^ The single-occupation of the oceanic and horizontal storage unit programming and wiping _ the money layer in the + 6. paste shoulder. This single-layer polycrystalline floating reservoir uses very low current to pass through the channel for programming and erase operations. What is required for the further step is that the thickness of the hardened region of the single-layer polysilicon floating gate_storage unit is smaller than just α = the size is smaller than the physical size of the conventional storage unit. It is worth noting that the feature of the floating layer of the polycrystalline silicon is that it is considered to be a metal oxide semiconductor ceramic transistor. Single Qing is not a metal, however, it can still be used to conduct electricity; this feature of field effect electric TMOS transistors. f _ 2^ Stomach _ «Sexual memory 2 〇〇 is formed on the Ρ-type substrate 205. The upper surface is formed - a deep diffusion well of N-type impurity __ a deep diffusion well 21 〇 through an N-type connection point connected to a first deep well bias source 265 ◊? The type substrate is connected to a ground reference voltage source (〇. 〇v) via a connection diffusion point 270. The first deep diffusion well 210 is obtained by adding a small change in the process of the current single-layer polysilicon logic integrated circuit process. Crystal [0060] 099103898 A first shallow diffusion well 215 and a second shallow diffusion well 220 forming a p-type impurity within the first deep diffusion well 210. The first and second shallow diffusion wells 215 and 220 are The current single-layer polymorphic logic accumulation circuit is based on an additional layer of process. Form No. A0I01 Page 23 of 106 0992007313-0 201030947 [0061] A programmed charge coupled metal oxide semiconductor (MOS) transistor 225 is connected as a capacitor. The programmed charge coupled MOS transistor 225 is formed on the surface of the first shallow diffusion well 215 and includes a source 226 and a drain 227 of N-type impurities both diffused within the first shallow diffusion well 215. The source 226, the drain 227 and the first shallow diffusion well 215 are commonly connected to a first shallow well bias source VTPW1 230. The P-type connection diffusion point 216 connects the first shallow diffusion well 215 to the first shallow well bias source VTPW1 230. The first shallow diffusion well 215 is coupled to the first shallow well bias source VTPW1 230 through a connection diffusion point 216. The gate oxide region 229 has a programmed coupling floating gate 228 formed of a single layer of polysilicon. The material and thickness of the gate oxide region 229 is one of the techniques commonly used in the current CMOS field.
[0062] IT,:, 在該第二淺擴散井220内形贫擎臀择多辱鼴235。該 存儲M0S電晶趙235"具有一N型汲極236和一N型源極237, 該N型汲極236和N型源極237均擴散在該第二淺擴散井 220。該汲極236連接到一位元4電壓源叫240,該源極 | ^ I f J | f1 - f 237連接到源極線偏壓源2知:極贰化1區239的上面 具有由單層多晶矽形成的,4|^^>&閘238。該閘極氧 III 化區239的材料和厚度與當前(^⑽領域製程所常用的技術 是一致的。該第二淺擴散井220透過連接擴散點250連接 到第二淺井偏壓源vTPW2 255。[0062] IT,: In the second shallow diffusion well 220, the shape is poor and the hips are more insulting. The memory MOS transistor 235" has an N-type drain 236 and an N-type source 237, and the N-type drain 236 and the N-type source 237 are both diffused in the second shallow diffusion well 220. The drain 236 is connected to a bit 4 voltage source called 240, and the source | ^ I f J | f1 - f 237 is connected to the source line bias source 2: the top of the region 1 239 has a single Layered polycrystalline germanium formed, 4|^^>& gate 238. The material and thickness of the gate oxy-III region 239 is consistent with the current techniques commonly used in the process of the art. The second shallow diffusion well 220 is coupled to the second shallow well bias source vTPW2 255 through the connection diffusion site 250.
[0063] 該存儲M0S電晶體235的編程和抹除操作是根據肋…冗― Nordheim的隧道邊緣編程(FN邊緣編程)效應而實現。邊 緣編程意味着Fowler-Nordheim随道效應發生在單元的 邊緣靠近汲極236或源極237與浮動結之間,而不是在存 儲M0S電晶體235的通道區域280。在邊緣編程操作的過 099103898 表單編號A0101 第24頁/共1〇6頁 0992007313-0 201030947 [0064][0063] The programming and erasing operations of the memory MOS transistor 235 are implemented in accordance with the rib ... redundancy - Nordheim tunnel edge programming (FN edge programming) effect. Edge programming means that the Fowler-Nordheim effect occurs at the edge of the cell near the drain 236 or between the source 237 and the floating junction, rather than in the channel region 280 where the MOS transistor 235 is stored. Over the edge programming operation 099103898 Form number A0101 Page 24 / Total 1 page 6 0992007313-0 201030947 [0064]
[0065] [0066] 程中,電子從電荷存儲浮動閘238被吸入到存儲M0S電晶 體235的沒極236。在Fowler-Nordheim隧道邊緣編程 時’每個單層多晶矽浮動閘非揮發性記憶體2〇〇在編程操 作時僅消耗1.0 πΑ以下的電流。 該編程電荷耦合M0S電晶體225在編程耦合浮動閘228、 没極、源極以及通道區域275之間形成的編程電容器和存 儲M0S電晶體235在電荷存儲浮動閘238、汲極236、源極 237和通道區域280之間形成的存儲電容器之間具有一個 耦合比率,該耦合比率必須非常大,大於80¾^要達到該 耦合比率’編程電荷耦合MGS-電晶鱧225的實際物理尺寸 相對於存儲M0S電晶體常的大。 該編程電荷耦合M0S電晶體_ 體2 3 5之間 的尺寸比率要超過10倍。該耦合比率要求存儲M0S電晶體 235的編程電位如表1所顯示:[0066] During the process, electrons are drawn from the charge storage floating gate 238 to the gate 236 that stores the MOS transistor 235. When programming at the edge of the Fowler-Nordheim tunnel, each single-layer polysilicon floating gate non-volatile memory 2 consumes less than 1.0 π 〇〇 of current during programming operations. The programming charge coupled MOS transistor 225 is formed between a programming coupled floating gate 228, a gate, a source and a channel region 275, and a memory MOS transistor 235 at the charge storage floating gate 238, drain 236, source 237 There is a coupling ratio between the storage capacitor formed between the channel region 280 and the coupling ratio, which must be very large, greater than 803⁄4^ to achieve the coupling ratio 'the actual physical size of the programmed charge coupled MGS-the transistor 225 relative to the memory MOS The transistor is often large. The size ratio between the programmed charge coupled MOS transistors _ body 2 3 5 is more than 10 times. The coupling ratio requires that the programming potential of the memory MOS transistor 235 be as shown in Table 1:
InfeSiectual 電晶體 235 痛:動遍… 285 " 汲極Pr( ”236 7_fy VTPW2 220 NVRAM cell 200 Vt FN邊緣 編程 -5V 浮動/0V —-—^— + 5V ον Vtl 註:Vt2>Vtl [0067] 在對該浮動閘非揮發性記憶體200進行編程操作時,電子 從浮動閘結2 8 5被吸取到汲·極2 3 6。該浮動閘結2 8 5連接 至編程搞合浮動閘2 2 8和電荷存儲浮動·閘2 3 8。對於被選 099103898 表單編號Α0101 第25頁/共106頁 0992007313-0 201030947 擇的浮動閘非揮發性記憶體200來說,要將電子從浮動閘 結285吸取到汲極236,該浮動閘結285的電位必須在大 約-5. 0V,並且必須將汲極236的電位設置在大約+ 5. 0V 。而對於沒被選擇的浮動閘非揮發性記憶體200來說,要 將汲極236的電位設置為接地參考電壓源(0. 0V)。第二 淺擴散井220和第一深擴散井210被設置為接地參考電壓 源(0. 0V)。該源極237被設置為浮動狀態或接地參考電 壓源(0. 0V)。 [0068]InfeSiectual transistor 235 pain: move all over... 285 " bungee Pr( ”236 7_fy VTPW2 220 NVRAM cell 200 Vt FN edge programming -5V floating /0V —-—^— + 5V ον Vtl Note: Vt2>Vtl [0067] When the floating gate non-volatile memory 200 is programmed, electrons are drawn from the floating gate junction 285 to the 汲 pole 2 36. The floating gate junction 285 is connected to the programmed floating gate 2 2 8 and the charge storage floating gate 2 3 8. For the selected 099103898 form number Α 0101 page 25 / 106 page 0992007313-0 201030947 selected floating gate non-volatile memory 200, to send electrons from the floating gate 285 The drain 236 is drawn, the potential of the floating gate 285 must be about -5.0 V, and the potential of the drain 236 must be set at about + 5. 0 V. For the floating gate non-volatile memory that is not selected. 200, the potential of the drain 236 is set to a ground reference voltage source (0. 0V). The second shallow diffusion well 220 and the first deep diffusion well 210 are set as a ground reference voltage source (0. 0V). The source 237 is set to a floating state or a ground reference voltage source (0. 0V).
請參閱圖2b,為要達到如表1所描述的條件而發生f〇wi — er,Nordheim隧道邊緣埠敦效應所需婆施加於浮動閘非Referring to Figure 2b, in order to achieve the conditions described in Table 1, f〇wi-er, the edge of the Nordheim tunnel is required to be applied to the floating gate.
閘非揮發性 30的電位大 揮發性記憶體200上的電位^對 記憶體2〇〇來說,該第一淺孝偏i源! . ... 約設置在-6. 0V到-7. 0V之間,該位元穠電壓源bl 240 的電位大約設置在+ |.:=(|V到+ 7. 0V之間.。對於沒被選擇的Gate non-volatile 30 potential large volatile memory 200 potential ^ For memory 2 〇〇, the first shallow filial i source! . ... is set at about -6. 0V to -7. Between 0V, the potential of the voltage source bl 240 is set approximately between + |.:=(|V to + 7. 0V. For those not selected
浮動閘非揮發性記慊體200來說〖丨該位元轉電壓源bl 24〇被設置為接地參考電壓案一深井偏a 源265被設置為接地參考g,並且該第二淺 -\ J 1 I f ^ 井偏壓源VTPW2 255和源極線偏壓源245可被選擇性地言史 置為浮動狀態或接地參考電壓源(0.0V)。 [0069]如上所述,該編程電荷耦合M0S電晶體225的物理器件尺 寸大於s亥存儲M0S電晶體235,而使得該編程電容器和存 儲電容器之間的耦合比率超過8〇%。為了使該浮動閘結 285具有大約-5.0V的編程電壓,該第一淺井偏壓源 VTPW1 230的耦合電壓必須是在大約_6 (^到_7 〇v之間 。如此則在汲極236的邊緣產生低電流的F〇wler- 099103898 表單编號A0101 第26頁/共1〇6頁 0992007313-0 201030947 Φ [0070]For the floating gate non-volatile memory body 200, 丨 the bit turn voltage source bl 24〇 is set to the ground reference voltage case. A deep well bias a source 265 is set to the ground reference g, and the second shallow-\ J The 1 I f ^ well bias source VTPW2 255 and the source line bias source 245 can be selectively set to a floating state or a ground reference voltage source (0.0V). As described above, the physical size of the programmed charge coupled MOS transistor 225 is greater than the sigma storage MOS transistor 235 such that the coupling ratio between the programming capacitor and the storage capacitor exceeds 8〇%. In order for the floating gate 285 to have a programming voltage of approximately -5.0 V, the coupling voltage of the first shallow well bias source VTPW1 230 must be between approximately _6 (^ to _7 〇v. Thus at the drain 236 The edge produces a low current F〇wler- 099103898 Form No. A0101 Page 26 / Total 1 Page 6 0992007313-0 201030947 Φ [0070]
Nordheim随道效應感應。.被儲存的電子從該浮動閘結 285被吸入進汲極236 ’以到達該位元線電壓源bl 240 。在Fowler-Nordheim邊緣編程操作以後,在一預先設 定的編程時間之後’該浮動閘非揮發性記憶體2〇〇的編程 臨界電壓Vt2從它的抹除臨界電壓vtl開始下降。在抹除 操作以後,抹除臨界電壓Vtl從大約+ 3. 〇V上升到大約 + 4. 0V,而編程臨界電壓Vt2大約是+ l.〇v ^抹除臨界電 壓vti不應該為負,以避免過度抹除而進入存儲M〇s電晶 體235耗盡狀態。根據閘極氧化區239的厚度和位元線電 壓源BL 240以及浮動間结找5的電位.,Fowler-Nordheim編程時間可在幾嘴^规衝楽j衫缚^完成。 在一次操嬋中,該第碎㊣23〇被設 置為驗證讀取電位(VriiD),該驗證讀取ΐ位(V )被Nordheim senses with the road effect. The stored electrons are drawn from the floating gate 285 into the drain 236' to reach the bit line voltage source bl 240. After the Fowler-Nordheim edge programming operation, the programming threshold voltage Vt2 of the floating gate non-volatile memory 2A begins to decrease from its erase threshold voltage vtl after a predetermined programming time. After the erase operation, the erase threshold voltage Vtl rises from approximately + 3. 〇V to approximately + 4. 0V, and the programming threshold voltage Vt2 is approximately + l. 〇v ^ erase the threshold voltage vti should not be negative, Avoid excessive erasure and enter the memory M〇s transistor 235 depleted state. According to the thickness of the gate oxide region 239 and the potential of the bit line voltage source BL 240 and the floating junction, the Fowler-Nordheim programming time can be completed in a few mouths. In one operation, the first fragment is set to verify the read potential (VriiD), and the verification read clamp (V) is
READ 參 [0071] 耦合到浮動閘結2 85 ^碎驗證讀取電位(vread)被設置在 抹除臨界電位Vtl和編程臨界第繁八2之間位元線電 壓祕240的電位大約被41^|^第二淺井偏壓 源咖2 255和源極線偏置為 源(〇. ον)。該第—深井偏壓源26^被設置為電源電壓源 (VDD)的電位。 若該浮動閘非揮發性記憶體2〇〇被編程操作,那麼在該浮 動閘結285的電位大約為第二臨界電位,並且有導電電流 從位元線電壓源BL 240流經該浮動閘非揮發性記憶體 200而到達源極線偏壓源245。當該浮動閘非揮發性記憶 體200的臨界電位是第二臨界電位yt2時,所讀取的數據 為二進制的「0」^若該浮動閘非揮發性記憶體2〇〇沒有 099103898 表單编號A0101 第27頁/共106頁 0992007313-0 201030947 被選擇作編程操作,那麼它的臨界電壓依然是在它的抹 除臨界電位Vtl,故該浮動閘結285的電位小於抹除臨界 電位Vtl ^當該浮動閘非揮發性記憶體2〇〇沒有導電電流 時’並且該浮動閘非揮發性記憶體2〇〇的臨界電壓是抹除 臨界電位’則所讀取的二進制數據是「1」。 [0072] 該浮動閘非揮發性記憶體200的好處是:全部的抹除、編 程和讀取操作最大的電位要求大約是+ 5. 0V,其與當前單 層多晶矽邏輯積體電路製程兼容。READ Ref. [0071] Coupling to the floating gate junction 2 85 ^Verification The verify read potential (vread) is set between the erase critical potential Vtl and the programming critical first octave 2 potential of the bit line voltage secret 240 is approximately 41^ |^ The second shallow well bias source 2 255 and the source line are biased as the source (〇. ον). The first deep well bias source 26 is set to the potential of the power supply voltage source (VDD). If the floating gate non-volatile memory 2 is programmed to operate, then the potential of the floating gate 285 is approximately the second critical potential, and a conductive current flows from the bit line voltage source BL 240 through the floating gate. The volatile memory 200 reaches the source line bias source 245. When the critical potential of the floating gate non-volatile memory 200 is the second critical potential yt2, the read data is a binary "0". If the floating gate non-volatile memory 2 has no 099103898 form number A0101 page 27 / page 106 0992007313-0 201030947 is selected as the programming operation, then its threshold voltage is still at its erase critical potential Vtl, so the potential of the floating gate 285 is less than the erase critical potential Vtl ^ When the floating gate non-volatile memory 2 〇〇 has no conduction current 'and the threshold voltage of the floating gate non-volatile memory 2 是 is the erase critical potential', the binary data read is "1". [0072] The benefit of the floating gate non-volatile memory 200 is that the maximum potential requirement for all erase, program, and read operations is approximately +5.00V, which is compatible with current single-layer polysilicon logic integrated circuit processes.
[0073] 請參閱圖3 ’為本發明單層多晶矽浮動閘非揮發性記憶體 Q 的另一具體實施方式的結構。該具體實施方式的結構與Please refer to FIG. 3' for the structure of another embodiment of the single-layer polysilicon floating gate non-volatile memory Q of the present invention. The structure and structure of this embodiment
圖2a的結構的不同之處在於$理;2 ,編秦電荷耦合Μ 0 S 電晶體225被電荷耦合在該第一 '.Cl, 淺擴散井215上形成 < 具有二個Ρ型擴散區的電荷耦合MOS 電容器325,該二個ρ型擴散區連接到第一井偏壓源 VTPW1 23〇。該第一淺擴^μ丨5汽摹二淺擴散井mo的 深度和濃度與圖2a中相對;如此可簡化製 程處理的複雜度,從而可看禮晶製程的費用並且 〇 與當前邏輯CMOS製程兼容。該編程耦合浮動閘328被耦合 到電荷存儲浮動閘238以形成浮動閘結385。 [0074] 該電荷耦合M0S電容器325和存儲M0S電晶體235與圖2a中 的編程電荷耦合金屬氧化物半導體(M0S)電晶體225和存 儲M0S電晶體235有相同的操作偏壓條件。與圖2a中的浮 動閘非揮發性記憶體200相同,該浮動閘非揮發性記憶體 300的電荷耦合M0S電容器325與存儲M0S電晶體235的尺 寸比率必須足够大,以使得該耦合比率超過8〇%。 099103898 表單編號A0101 第28頁/共106頁 0992007313-0 201030947 [0075] 當該第一淺井偏壓源VTPW1 230被設置為編程電位(大 約從-7. 0¥到_5. 0V)時,在該浮動閘結385上的電壓在 編程操作時高於-5. 0V。吸取電子的結果與前面圖2a所描 述的和圖2b所顯示的偏廢電位相同。 [0076] 請參閲圖4a,為本發明單層多晶矽浮動閘非揮發性記憶 體的再一具體實施方式的結構。該單層多晶矽浮動閘非 揮發性記憶體400是在P型基板405上形成的。該p型基板 405的表面形成一 N型雜質的第一深擴散井41〇。該第一深 擴散井410只是在當前單層多晶矽遏輯積體電路製程增加 一層製程而得到。該深擴散井填過^一N型連接點連接 到一深井偏壓珠465。該 P 辑連释擴散點 470 連接到接挺夸着電壓源(〇. /: [0077] 在該深擴散井410之狗形成一p型雜質铪淺擴散井415。該 淺擴散井415是在當翁舉層多晶梦邏輯積體電路製程另外 增加層而付到° WelleckjolThe structure of Fig. 2a differs in that; 2, the Qin charge-coupled Μ 0 S transistor 225 is charge-coupled on the first '.Cl, shallow diffusion well 215. < has two Ρ-type diffusion regions The charge coupled MOS capacitor 325 is connected to the first well bias source VTPW1 23〇. The depth and concentration of the first shallow diffusion ^μ丨5 摹2 shallow diffusion well mo are opposite to those in FIG. 2a; this simplifies the complexity of the process processing, thereby observing the cost of the ritual process and the current logic CMOS process compatible. The program coupled floating gate 328 is coupled to a charge storage floating gate 238 to form a floating gate 385. The charge coupled MOS capacitor 325 and the memory MOS transistor 235 have the same operational bias conditions as the programmed charge coupled metal oxide semiconductor (MOS) transistor 225 and the memory MOS transistor 235 of FIG. 2a. As with the floating gate non-volatile memory 200 of FIG. 2a, the size ratio of the charge coupled MOS capacitor 325 of the floating gate non-volatile memory 300 to the memory MOS transistor 235 must be sufficiently large that the coupling ratio exceeds 8 〇%. 099103898 Form No. A0101 Page 28 / Total 106 Pages 0992007313-0 201030947 [0075] When the first shallow well bias source VTPW1 230 is set to the programming potential (approximately from -7.00 to _5. 0V), The voltage on the floating gate 385 is higher than -5.0 V when programmed. The result of the electron absorption is the same as that shown in Fig. 2a and the offset potential shown in Fig. 2b. Referring to FIG. 4a, a structure of still another embodiment of a single-layer polysilicon floating gate non-volatile memory of the present invention is shown. The single-layer polysilicon floating gate non-volatile memory 400 is formed on a P-type substrate 405. The surface of the p-type substrate 405 forms a first deep diffusion well 41 of an N-type impurity. The first deep diffusion well 410 is obtained by adding a further process to the current single-layer polysilicon suppression integrated circuit process. The deep diffusion well is connected to a deep well biased bead 465 via an N-type connection point. The P-series diffusion-diffusion point 470 is connected to a tapping voltage source (〇. /: [0077] The dog in the deep diffusion well 410 forms a p-type impurity shallow diffusion well 415. The shallow diffusion well 415 is When the Weng layer polycrystalline dream logic integrated circuit circuit process adds another layer and pays ° Welleckjol
剛在該淺擴散井4 i 5的表面形|^醜%今麵合金屬氧化物 半導體(M0S)電晶趙425,竣,編_胃荷耦合金屬氧化物半 導體電晶體425連接作為一電容器。該編程電荷耦合M〇s 電晶體425包括一 N型雜質的源極426和汲極427,該源極 426和汲極427擴散在該淺擴散井415。該源極426、汲極 427和淺擴散井415共同連接到一淺井偏壓源^{^143〇 。該淺擴散井415透過連接擴散點416連接到淺井偏祕 VTPW1 430。-單層多晶發在閘極氧化區似上形成一編 程搞合浮動閘428。該閘極氧化區429的材料和厚度與當 前CMOS領域所用的製程技術一致。 099103898 表單編號A0101 第29頁/共1〇6頁 0992( 201030947 [0079] [0080] [0081] 在該P型基板405上形成一存儲M〇s電晶體435。該存儲 MOS電晶體435有一 N型汲極型源極437擴散在p型基板 405裡。該汲極436連接到一位元線電壓源乩44〇,該源 極437連接到源極線偏壓源445。在該閘極氧化區439上 有-單層多晶石夕形成-電荷存儲浮動閘權。該閘極氧化 區439的材料和厚度與當前CM〇s領域所用的製程技術一致 〇 該浮動閘非揮發性記憶體4〇〇與圖2a的浮動閘非揮發性記 憶體200不同之處在於存儲M〇s電晶嬤435是直接地在基 @ 板405的表面形成的。在爵加中,讓滓動閛結285是編程 t M m H物半導體和存儲M〇s 電晶體2_連接點。由於電晶體435 直接在基板405的表面形成,該存儲^^^電晶體435和編 程電荷耦合金屬氧化物半導體⑽s)電晶體425所在的深 擴散井410之間的間距必須符合當前Cjj〇s,域製程的佈局 設計規則。如同圖2a,在該气成的編程電荷 耦合金屬氧化物半導艘(Μ(^華IX25在編程操作時被 肇 用來作為電壓麵合電容器^ ^該i‘4Q5的表面被形成❼ 存儲M0S電晶體435是一單層多晶存儲浮動閘極器件。 對該存儲M0S電晶體435的編程操作是一從該浮動閘非揮 發性記憶體400吸取電子的Fowler_N〇rdheim邊緣編程 操作。Fowler-Nordheim邊緣編程進行讀取和編程操作 所需要的電壓與表1類似,透過注射電子到該浮動閘結 485來完成抹除操作,並且該浮動閘非揮發性記憶魏4〇〇 的臨界電壓為抹除臨界電壓Vtl (大約+3 〇v到+4. 〇v)。 099103898 表單編號A0101 第30頁/共1〇6頁 0992007313-0 201030947 [0082] 请參閱圖4b,為本實施方式中讀取和編程操作的偏壓電 位的表格,在該淺井偏壓源VTpw 43〇的耦合電壓必須從 大約-6. 0V到-7. 0V,則可在該汲極436的邊緣產生低電 流Fowler-Nordheim隧道效應。被儲存的電子從該浮動 閘結485被吸取進入汲極436,進而到達該位元線電壓源 BL 440。在執行Fowler_N〇rdheim邊緣編程操作以後的 一段預設的編程時間之後,該浮動閘非揮發性記憶體 的臨界電壓Vt2從抹除臨界電壓Vtl減小。在抹除操作以Just in the surface of the shallow diffusion well 4 i 5 | ^ ug% of the surface of the metal oxide semiconductor (M0S) electro-crystal Zhao 425, 竣, _ gastric-coupled metal oxide semiconductor transistor 425 is connected as a capacitor. The programmed charge coupled M?s transistor 425 includes a source 426 and a drain 427 of an N-type impurity that is diffused in the shallow diffusion well 415. The source 426, the drain 427 and the shallow diffusion well 415 are commonly connected to a shallow well bias source ^^^. The shallow diffusion well 415 is connected to the shallow well VTPW1 430 through a connection diffusion point 416. - A single layer polycrystalline wafer forms a programmed floating gate 428 on the gate oxide region. The material and thickness of the gate oxide region 429 is consistent with the process technology used in the current CMOS field. 099103898 Form No. A0101 Page 29/Total 1 Page 6 0992 (201030947 [0079] [0081] A memory M s transistor 435 is formed on the P-type substrate 405. The memory MOS transistor 435 has a N The type drain source 437 is diffused in the p-type substrate 405. The drain 436 is connected to a one-bit line voltage source 乩44〇, which is connected to the source line bias source 445. The gate is oxidized The region 439 has a single-layer polycrystalline stone formation-charge storage floating gate. The material and thickness of the gate oxide region 439 are consistent with the process technology used in the current CM〇s field. The floating gate non-volatile memory 4 〇〇 is different from the floating gate non-volatile memory 200 of FIG. 2a in that the memory M 〇 电 嬷 435 is formed directly on the surface of the base @ 405. In the 加 加, let the 閛 閛 285 It is a programming M M H material semiconductor and a memory M 〇 s transistor 2_ connection point. Since the transistor 435 is formed directly on the surface of the substrate 405, the memory transistor 435 and the programmed charge coupled metal oxide semiconductor (10) s) The spacing between the deep diffusion wells 410 where the transistors 425 are located must conform to the current Cjj〇s, the fabric of the domain process. Design rules. As shown in Fig. 2a, the gas-charged programmed charge-coupled metal oxide semi-conducting vessel (Μ(^华IX25 is used as a voltage-surface-capacitor during programming operation) ^The surface of the i'4Q5 is formed. ❼ Storage M0S The transistor 435 is a single layer polycrystalline memory floating gate device. The programming operation of the memory MOS transistor 435 is a Fowler_N〇rdheim edge programming operation that extracts electrons from the floating gate non-volatile memory 400. Fowler-Nordheim The voltage required for edge programming for reading and programming operations is similar to that of Table 1. The erase operation is performed by injecting electrons into the floating gate 485, and the threshold voltage of the floating gate non-volatile memory is erased. The threshold voltage Vtl (approximately +3 〇v to +4. 〇v). 099103898 Form No. A0101 Page 30 / Total 1 〇 6 Page 0992007313-0 201030947 [0082] Please refer to FIG. 4b, read and For a table of bias potentials for programming operation, the coupling voltage at the shallow well bias source VTpw 43〇 must be from about -6. 0V to -7. 0V, and a low current Fowler-Nordheim can be generated at the edge of the drain 436. Tunneling effect The floating gate 485 is drawn into the drain 436 to reach the bit line voltage source BL 440. After a predetermined programming time after performing the Fowler_N〇rdheim edge programming operation, the floating gate non-volatile memory The threshold voltage Vt2 is decreased from the erase threshold voltage Vtl.
後,抹除臨界電壓Vtl大約在+ 3· 0V到+4· 0V之間,編程 臨界電壓Vt2大約為〇v。該抹除臨界電壓vtl不應該 為負,以避免該存儲M0S電孕濟435在等博_以後進入 耗盡狀態。根據該閘極氧仆^9#波七位#線電壓源虬 2 4 °以及:’動閘結4 8 5的電異纖e i m編程時 間可我從幾微秒到幾毫秒之間完成。 [0083]Thereafter, the erase threshold voltage Vtl is approximately between +3·0V and +4·0V, and the program threshold voltage Vt2 is approximately 〇v. The erase threshold voltage vtl should not be negative to prevent the stored MOS power supply 435 from entering the depleted state after the hop. According to the gate oxygen servant ^9 # wave seven bit # line voltage source 虬 2 4 ° and: 'moving junction 485 5 electric fiber e e m programming time can be completed from a few microseconds to a few milliseconds. [0083]
在讀取操作時,該淺井偏气源VT^W 430 取電拉(vREAD),該驗證 閘結485。該驗證讀取電位臨界電位ηι和 編程臨界電位Vt2之間。該位元^電壓源bl 440的電位 大約被設置為+ 1. 0V。該P型基板405和源極線偏壓源445 被設置為接地參考電壓源(0. 0V) »該深井偏壓源465的 被j設置為驗證讀QJ yi耦合到浮動 電位被設置為電源電壓源(VDD)的電位。 [0084]若該浮動閘非揮發性記憶體4〇〇被編程操作,那麼該浮動 閘結485的電位大約是第二臨界電位,並且有導電電流從 該位元線電壓源BL 440經過該浮動閘非揮發性記憶體 400到源極線偏壓源445。當該浮動閘非揮發性記憶體 099103898 表單編號A0101 第31頁/共106頁 0992007313-0 201030947 [0085] [0086] 099103898 40〇的臨界電位是第二臨界電位vt2時,所讀取的數據為 —進制的「〇」。若該浮動閘非揮發性記憶體4〇〇沒有被 、編程操作’則它的臨界電壓依然停留在抹除臨界電位Vtl 。故該浮動閘結485的電位比抹除臨界電位Vtl小。當該 浮動開非揮發性記憶體400沒有導電電流,而且該浮動閘 非揮發性記憶體400的臨界電位是抹除臨界電位時,所讀 取的數據為二進制的「1」。 請參閱圖5 ’為本發明單層多晶矽浮動閘非揮發性記憶體 的另一個具體實施方式的結構。該實施方式的結構與圖 4a本質上是相同的,除了圈4§1中褊衮電荷耦合M〇s電晶體 425用電荷耦合M〇s電容器之f 荷耦合M〇s 電谷器525舆圖3中的相應喝圈3的舍荷搞合M〇s 谷器325具有在第一淺擴散井215中形成的二個ρ型擴散 區,而且這二個Ρ型擴被區連接到第一井偏壓#VTpw 43〇。該淺擴散井41S的深芒亡秀婷與圖3录相同的,如此 可簡化製程的複雜度及減少:早巧多晶▲程的費用並且與 备月,』邏輯咖s製程兼容。該令餐轉合浮動閘528輛合到電 荷存儲浮動閘438而形成浮i閘\#5。 該電荷麵合MOS電容器525和存儲廳電晶體435與圖4&的 編程電荷麵合金屬氧化物半導體⑽s)電晶艘425和存儲 MOS電晶想435有相同的操作偏愿條件。如同_的浮動 閘非揮發性記憶趙400,該浮動閘非揮發性記憶艎5〇〇的 電荷麵合祕電容器525相對於存麵s電晶體邮的物理 尺寸比率要足够大,以使得_合比率超過8〇%。 當該淺井偏魏VTPW 43〇被設置為編程電位(大約從_ 表單編號A0101 第32頁/共106頁During the read operation, the shallow well bias source VT^W 430 is pulled (vREAD), which verifies the junction 485. This verification reads between the potential critical potential ηι and the programming threshold potential Vt2. The potential of the voltage source bl 440 is set to approximately + 1. 0V. The P-type substrate 405 and the source line bias source 445 are set to a ground reference voltage source (0 V). The deep well bias source 465 is set to verify that the read QJ yi is coupled to the floating potential and is set to the supply voltage. The potential of the source (VDD). [0084] If the floating gate non-volatile memory 4 is programmed to operate, the potential of the floating gate 485 is approximately the second critical potential, and a conductive current passes from the bit line voltage source BL 440 through the floating Gate non-volatile memory 400 to source line bias source 445. When the floating gate non-volatile memory 099103898 Form No. A0101 Page 31 / Total 106 Page 0992007313-0 201030947 [0085] [0086] 099103898 When the critical potential of 40〇 is the second critical potential vt2, the data read is - "进制" in hexadecimal. If the floating gate non-volatile memory 4 is not programmed, its threshold voltage remains at the erase critical potential Vtl. Therefore, the potential of the floating gate 485 is smaller than the erase critical potential Vtl. When the floating open non-volatile memory 400 has no conduction current, and the critical potential of the floating gate non-volatile memory 400 is the erase critical potential, the read data is a binary "1". Please refer to FIG. 5' for the structure of another embodiment of the single-layer polysilicon floating gate non-volatile memory of the present invention. The structure of this embodiment is essentially the same as that of FIG. 4a except that the 褊衮 charge coupled M 〇 电 transistor 425 is coupled with the charge coupled M 〇 s capacitor of the 〇 电 电 电 舆 舆 舆 舆 舆The corresponding 喝 搞 〇 〇 谷 325 325 325 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有 具有Bias #VTpw 43〇. The shallow-diffusing well 41S is the same as that recorded in Figure 3, which simplifies the complexity and reduction of the process: the cost of the pre-crystallized ▲ process is compatible with the standby month, the logic coffee s process. The order is converted into a floating gate 528 and connected to the charge storage floating gate 438 to form a floating gate\#5. The charge-to-surface MOS capacitor 525 and the memory cell 435 have the same operational bias conditions as the programmed charge-planar metal oxide semiconductor (10) s) cell 425 and the memory MOS die 435 of Figures 4 & Like the floating gate non-volatile memory Zhao 400 of _, the floating gate non-volatile memory 艎 5 〇〇 charge surface sealing capacitor 525 is larger than the physical size ratio of the storage surface s transistor post, so that _ The ratio is over 8%. When the shallow well Wei VTPW 43〇 is set to the programming potential (approximately from _ form number A0101 page 32 / total 106 pages
0992007313-0 [0087] 201030947 7. 0V到-5. 0V)時’該浮動閘結585的電壓在編程操作時 大於-5. 0V。所產生的電子吸取現象與前面圖切所描述的 和圖4b所顯示的偏壓電位是相同的。 [0088] 請參閱圖6a ’對於熟習本領域技術的人來說,圖5的細節 與圖6a是相同的。該浮動閘非揮發性記憶體2〇〇包括一編 程電荷耦合M0S電晶體225和一存儲M0S電晶體235,該編 程電荷耦合M0S電晶體225連接作為一M0S電容器,該存 儲M0S電晶體235連接作為一電荷儲存器。編程耦合浮動 閘2 2 8和電荷存儲浮動閘238相連接而形成一浮動閘結 285。該編程電荷耦合|f〇s奮晶艟225的源極、汲極和基 極(第一淺擴散井)連接到 味—Wl 230。 該存儲M0S電晶體23 5的汲^[钿源BL 24〇 °該存儲M0S電晶想235的源極連接到源極線偏壓源245 ·.; mi rr 。該第二淺擴散井’ %存儲M0S電晶體235的基極(bulk) 連接到第二淺井偏壓源VTPW2 » #注举的芩,圖心中存儲 M0S電晶兹4 3 5是在基板40Iplf ί故本實施方 [0089] .rf ΟΟΘΓι V 式可不提供該第二淺井偏辱竭fl_W2 'Urnce 請參閱圖6b,對於熟習本領域技藝的人來說,圖5的細節 與圖6b相同。該實施方式中的浮動閘非揮發性記憶體3〇〇 的結構與囷2a的結構的不同之處在於:圖2a中編程電 荷麵合M0S電晶體225被電荷麵合JJ0S電容器325替換。該 電荷耦合M0S電容器325的編程耦合浮動閘328和存儲M0S 電晶趙23 5的電荷存儲浮動閘238連接成為浮動閘結385 。兩P型擴散區連接到該第一淺井偏壓源VTPW1 230。在 圖2a中’該存儲M0S電晶體235的沒極連接到位元線電愿 099103898 表單編號A0I01 第33頁/共106頁 0992007313-0 201030947 源BL 240。該存儲M0S電晶體235的源極連接到源極線偏 壓源245 »該第二淺擴散井’即存儲M〇s電晶體235的基 極(bulk)連接到第二淺井偏壓#VTPW2 ^要注意的是, 在圖5中該存儲jjos電晶體435是在基板405的表面形成的 。故本實施方式可不提供該第二淺井偏壓源VTpff2。 [0090] 請參閱圖6c,為一非揮發性記憶器件6〇〇,它包括一由圖 6 a中‘洋動閘非揮發性記憶體2 〇 〇所形成的一個陣列6 〇 5 ^ 對於熟習本領域技藝的人來說,只需一些適當的修改,0伏。 0992007313-0 [0087] 201030947 7. 0V to -5. 0V) when the voltage of the floating gate 585 is greater than -5.00 in the programming operation. The resulting electron pickup phenomenon is the same as that described in the previous figure and the bias potential shown in Fig. 4b. [0088] Please refer to FIG. 6a'. The details of FIG. 5 are the same as those of FIG. 6a for those skilled in the art. The floating gate non-volatile memory 2A includes a programmed charge coupled MOS transistor 225 and a memory MOS transistor 235 connected as a MOS capacitor, the memory MOS transistor 235 being connected as A charge reservoir. The programming coupled floating gate 2 2 8 is coupled to the charge storage floating gate 238 to form a floating gate 285. The source, drain and base (first shallow diffusion well) of the programmed charge coupled |f〇s 艟 艟 225 are connected to the odor-Wl 230. The memory of the MOS transistor 23 5 is stored. The source of the memory MOS transistor 235 is connected to the source line bias source 245 ·.; mi rr . The second shallow diffusion well '% stores the base of the MOS transistor 235 to be connected to the second shallow well bias source VTPW2 » #Note 芩, the memory in the figure stores the M0S electric crystal 4 4 5 is on the substrate 40Iplf ί Therefore, the present embodiment [0089] .rf ΟΟΘΓι V may not provide the second shallow well partial insult fl_W2 'Urnce Please refer to FIG. 6b. For those skilled in the art, the details of FIG. 5 are the same as FIG. 6b. The structure of the floating gate non-volatile memory 3A in this embodiment is different from the structure of the 囷2a in that the programmed charge-surface MOS transistor 225 in Fig. 2a is replaced by a charge-junction JJ0S capacitor 325. The program coupled floating gate 328 of the charge coupled MOS capacitor 325 and the charge storage floating gate 238 of the memory MOS transistor 23 are connected to form a floating gate 385. Two P-type diffusion regions are coupled to the first shallow well bias source VTPW1 230. In Fig. 2a, the gate of the memory MOS transistor 235 is connected to the bit line. 099103898 Form No. A0I01 Page 33 of 106 0992007313-0 201030947 Source BL 240. The source of the storage MOS transistor 235 is connected to the source line bias source 245 » the second shallow diffusion well 'ie, the base of the storage M 〇 transistor 235 is connected to the second shallow well bias #VTPW2 ^ It is to be noted that the memory jjos transistor 435 is formed on the surface of the substrate 405 in FIG. Therefore, the second shallow well bias source VTpff2 may not be provided in this embodiment. [0090] Please refer to FIG. 6c, which is a non-volatile memory device 6A, which includes an array formed by the non-volatile memory 2 洋 of FIG. 6 a. Those skilled in the art will only need some appropriate modifications,
由該浮動閘非揮發性記憶體2 〇 〇所形成的陣列6 〇 5便可由 圖6b浮動閘非揮發性記憶體3〇〇替換,正如圖“所描述, 每一浮動閘非揮發性記憶體#一 -耦合M〇s電 晶體22b和存儲MOS電晶體2藏,於列尽Q5中每一橫列 内的編程電荷耦合MOS電晶體225,其汲極、源極和基極 (bulk)都連接到第一歲井偏壓源VTpwl[〇] 645a,...,The array 6 〇5 formed by the floating gate non-volatile memory 2 便可 can be replaced by the floating gate non-volatile memory 3 图 of Figure 6b, as described in the figure, each floating gate non-volatile memory #一-Coupled M〇s transistor 22b and memory MOS transistor 2, in the programmed charge-coupled MOS transistor 225 in each column of Q5, the drain, the source and the base are both Connected to the first year old well bias source VTpwl[〇] 645a,...,
VTPW1[1] 645b。每條位元線叫[〇] 640a,BL[1] 64〇b ’…’ BL[n] 64〇n連1黃到該'歹,j陣B™〇5中與其相關聯 的每一浮動閘非揮發牲記中的/每一存儲仙^電晶 體235的汲極。每一橫列内的存電晶體235的基極 (bulk)連接到與每一橫列相關聯的第二淺井偏壓源 VTPW2 [0] 650a’...’VTPW2 [1] 650b。每一橫列的 存儲M0S電晶體235的源極連接到相關聯的源極線偏壓源 655 ° 該位元線BL[0] 640a,BL[l] 640b,…,BL[n] 640η 連接到位元線感應放大器615,該位元線感應放大器615 包括位元線電壓源,以為該浮動閘非揮發性記憶體2〇〇提 099103898 表單編號Α0101 第34頁/共1〇6頁 0992007313-0 [0091] 201030947 供編程和讀取操作時必要的偏壓。另,該位元線感應放 大益615還包括一感應放大器電路,用以感應在讀取操作 時所被選擇的浮動閘非揮發性記憶體2〇〇的數據狀態並且 在數據输出口 635提供感應到的數據。 [酬—仃解碼控制器610包括第一、淺井偏壓源VTpwl⑷VTPW1[1] 645b. Each bit line is called [〇] 640a, BL[1] 64〇b '...' BL[n] 64〇n even 1 yellow to the '歹, j array BTM〇5 is associated with each float The gate of the non-volatile memory is stored in each of the drains of the transistor 235. The base of the storage transistor 235 in each row is connected to a second shallow well bias source VTPW2 [0] 650a'...'VTPW2 [1] 650b associated with each row. The source of each row of storage MOS transistors 235 is connected to an associated source line bias source 655 °. The bit lines BL[0] 640a, BL[l] 640b, ..., BL[n] 640η are connected. In place of the bit line sense amplifier 615, the bit line sense amplifier 615 includes a bit line voltage source to provide the floating gate non-volatile memory 2 to 099103898. Form number Α 0101 page 34 / total 1 page 6 page 0992007313-0 [0091] 201030947 A bias voltage necessary for programming and reading operations. In addition, the bit line sense amplifier 615 further includes a sense amplifier circuit for sensing the data state of the floating gate non-volatile memory 2〇〇 selected during the read operation and providing the sense at the data output port 635. Data to. [Reward-仃 decoding controller 610 includes the first, shallow well bias source VTpwl (4)
…,VTPWltl] 645b、第二淺井偏壓源 VTPW2[G] 65Ga,...,VTPW2[1] 65Qb和源極線偏屢源 655。一地址訊號62〇、數據訊號625和控制訊號63〇在被 解碼之後會連接到該行解碼控制器61〇和位元線感應放大 器615,以實現編程和讀取陣巧&认5中所被選擇的浮動閘 非揮發性記憶體200時所必綠蜂 [咖請參_7a,為本發明單权&4:^^發性記憶 體700的ϋ實施城的結構。該結構與嶋的不同之 處在於:增加-選擇雜電晶體790。在該實施方式中, =程1浮動閘228和電舖猶呢在一起 形成一㈣閉結785。 Property ......, VTPWltl] 645b, second shallow well bias source VTPW2[G] 65Ga,...,VTPW2[1] 65Qb and source line offset source 655. An address signal 62, a data signal 625, and a control signal 63 are coupled to the row decoding controller 61 and the bit line sense amplifier 615 after being decoded to implement programming and reading of the array & The selected floating gate non-volatile memory 200 must be a green bee [Career _7a, which is the structure of the implementation of the 单 && 4: ^ 发 记忆 记忆 memory 700. The structure differs from germanium in that it adds-selects a hybrid transistor 790. In this embodiment, the =1 floating gate 228 and the electric gate are formed together to form a (four) closure 785. Property ...
[酬該存儲M0S電晶體735的形中在第二淺擴散井 220裡形成存儲廳電晶艘235一樣。該存儲職電晶體 735有-N魏極736 — N型源極m擴散在第二淺擴散 井220。該汲極736連接到—位元線電壓源240。由一單 層多㈣在閘極氧化區739的上面形成電荷存儲浮動閉 738。該閘極氧化區739的材料和厚度與當前_s領域製 程常用的技術是一致的。 刪在該第二淺擴散井220還同時形成選擇 0992007313-0 099103898 表單編號A0101 第35頁/共1〇6頁 201030947 一P型源極793擴散在該第二淺擴散井22〇。該選擇閘極電 晶體790的汲極同時為該存儲M0S電晶體735的源極,即 擴散在該第二淺擴散井220的P型雜質。該選擇閘極電晶 體790的源極連接到該源線電壓源SL 245。由一單層多 晶矽在閘極氧化區794上面形成選擇閘極792 ,該選擇閘 極792連接到一選擇閘極控制訊號795。在編程和讀取操 作時,該選擇閘極電晶體790用於控制該存儲M〇s電晶體 735與源極線245的連接》 [0096] [0097] ❹ 如同圖2a和2 b所描述,該存儲MOS電晶體735的編程和抹 除操作是依照Fowler-Nordheim隧道邊緣編程理論。該 編程電荷耦合MO S電晶體2 2 一齡存儲M〇 s 電晶體735的存儲電容器的於8〇%。要 達到讓耦合比率’該編程電荷耦合M〇s電晶體22 5的物理 尺寸相對於該存儲M0S電晶體735的物理尺寸亦要非常大 。該編程電荷耦合M0S電晶體2 2奸目對於存儲M0S電晶體 ;·::: "5 ' 735的尺寸比率要大過十倍^ ^f .11:“:· * ί •j^;;ϋ 》The form of the storage MOS transistor 735 is formed in the second shallow diffusion well 220 to form the memory cell 235. The memory cell 735 has a -N Wei 736 - N source m diffused in the second shallow diffusion well 220. The drain 736 is coupled to a bit line voltage source 240. A charge storage floating closure 738 is formed over the gate oxide region 739 by a single layer (4). The material and thickness of the gate oxide region 739 is consistent with the techniques commonly used in current field processes. Deleted in the second shallow diffusion well 220 also formed a selection at the same time 0992007313-0 099103898 Form No. A0101 Page 35 / Total 1 page 6 201030947 A P-type source 793 diffuses in the second shallow diffusion well 22〇. The drain of the select gate transistor 790 is simultaneously the source of the memory MOS transistor 735, i.e., the P-type impurity diffused in the second shallow diffusion well 220. The source of the select gate transistor 790 is coupled to the source line voltage source SL 245. A select gate 792 is formed over the gate oxide region 794 by a single layer of polysilicon, and the select gate 792 is coupled to a select gate control signal 795. The select gate transistor 790 is used to control the connection of the memory M s transistor 735 to the source line 245 during programming and read operations. [0096] As described in Figures 2a and 2b, The programming and erasing operations of the memory MOS transistor 735 are in accordance with the Fowler-Nordheim tunnel edge programming theory. The programming charge coupled MO S transistor 2 2 is one year old storage M 〇 s transistor 735 of the storage capacitor of 8 〇 %. To achieve the coupling ratio 'the physical size of the programmed charge coupled M?s transistor 22 5 is also very large relative to the physical size of the memory MOS transistor 735. The programmed charge-coupled MOS transistor 2 2 is for storing M0S transistors; the size ratio of ··:: "5 ' 735 is greater than ten times ^ ^f .11: ":· * ί •j^;; ϋ 》
請參閱圖7b ’其為達到表1*翻蜂^錢件以產生f〇w 1 er-Please refer to Figure 7b' for reaching Table 1* to turn the bee to generate f〇w 1 er-
Nordheim隧道邊緣編程效應發生時所需要運用在該浮動 閘.非揮發性記憶體上的電位。對於被選擇的浮動閘非 揮發性記憶體700,該第一淺井偏壓源ντρ\Π 230的電位 被設置在大約-6. 0V和-7. 0V之間、位元線電壓源BL 240的電位被叙置在大約+ 5. 0V和+ 7. 0V之間。對於沒被 選擇的浮動閘非揮發性記憶體700,該位元線電壓源BL 240的電位被設置為接地參考電壓源(0. 0V)。該第一深 井偏壓源265的電位被設置為接地參考電壓源(ο.ον), 099103898 表單編號Α0101 第36頁/共1〇6頁 0992007313-0 201030947 並且該第二淺井偏壓源VTPW2 255和源極線偏壓源245被 設置為准許浮動或選擇性地被設置為接地參考電壓源 (0.0V) ’該選擇閘極控制訊號795的電位被設置為接地 參考電壓源(0. 0V)以關閉選擇閘極電晶體790,如此則 允許該源極737在編程操作中浮動,並且電荷可以從該編 程耦合浮動閘228吸取到該存儲MOS電晶體235的汲極236 [0098] ❹ 如上所述,該編程電荷耦合MOS電晶體225的物理器件尺 寸大於該存儲Μ 0 S電晶體73 5,而使得該編程電容器和存 儲電容器之間的耦合比率超過80¾ ·务:了在該浮動閘結 m有大約-5.0V的編程電,___壓源 VTPW1 23〇的耦合電壓必須掛气今0V之間 如此蹲在該汲極2 3 6的邊緣發生低電流的Fowler-The potential of the Nordheim tunnel edge programming effect is applied to the floating gate. Non-volatile memory. For the selected floating gate non-volatile memory 700, the potential of the first shallow well bias source ντρ\Π 230 is set between about -6. 0V and -7. 0V, and the bit line voltage source BL 240 The potential is set between approximately + 5. 0V and + 7. 0V. For the floating gate non-volatile memory 700 that is not selected, the potential of the bit line voltage source BL 240 is set to the ground reference voltage source (0. 0V). The potential of the first deep well bias source 265 is set to a ground reference voltage source (ο.ον), 099103898 Form No. Α0101, page 36/1, page 6, 0992007313-0 201030947 and the second shallow well bias source VTPW2 255 And the source line bias source 245 is configured to permit floating or selectively set to a ground reference voltage source (0.0V) 'The potential of the selected gate control signal 795 is set to a ground reference voltage source (0. 0V) The gate 790 is selected to be turned off, thus allowing the source 737 to float during the programming operation, and charge can be drawn from the programming coupled floating gate 228 to the drain 236 of the memory MOS transistor 235 [0098] The physical device size of the programmed charge coupled MOS transistor 225 is larger than the memory 73 0 S transistor 73 5 such that the coupling ratio between the programming capacitor and the storage capacitor exceeds 803⁄4. There is a programming power of about -5.0V, the coupling voltage of the ___voltage source VTPW1 23〇 must be hang between the current 0V so that Fowler- low current occurs at the edge of the bungee 2 3 6
Nordheim隧道效應。破儲存的電子從該浮動閘結785被 ❹ 吸取進入汲極236而到達气元線寫壓源BL ^4〇。在〜“一 er-Nordheim邊緣編程操编程時間之後 ,該浮動閘非揮發性記德<界/電壓v 12從抹除臨 界電壓vtl減少。在抹除ϋ後%該抹除臨界電壓vtl 大約在+ 3. 0V到大約+4. 〇V之間,而編程臨界電壓Vt2大 約是+1. ον。該抹除臨界電壓Vtl不應該為負,以避免過 度抹除而使該存儲_電晶想235進人耗盡狀態。根據該 閘極氧化739的厚度和位元線電壓源队 240以及浮動閘 結785的電位,Fowler—N〇rdheim編程時間可在幾微秒 到幾毫秒之間完成。 [0099] 在讀取操作中’該第—淺井偏Μ源VTPW1 23G的電位被 099103898 表單编號Α0101 第37頁/共106頁 0992007313-0 201030947 設置到驗證讀取電位(Vread),該驗證讀取電位(VREAD) 被相合到該浮動閘結785。該驗證讀取電位(V )被設 READy 置在抹除臨界電位Vtl和編程臨界電位Vt2之間。該位元 線電髮源BL 240的電位大約被設置在h.ov。該第二淺 井偏壓源VTPW2 255和源極線偏壓源245被設置在接地參 考電壓源(0. 0V)。該第一深井偏壓源265的電位被設置 在電源電壓源(VDD)的電位。 [0100] [0101]Nordheim tunneling effect. The stored electrons are sucked from the floating gate 785 into the drain 236 to reach the gas source write source BL^4. After ~ "one er-Nordheim edge programming operation programming time, the floating gate non-volatile memory < boundary / voltage v 12 is reduced from the erase threshold voltage vtl. After erasing ϋ%, the erase threshold voltage vtl is approximately Between + 3. 0V and about +4. 〇V, and the programming threshold voltage Vt2 is about +1. ον. The erase threshold voltage Vtl should not be negative to avoid over-wiping to make the memory_electric crystal I want the 235 to be depleted. According to the thickness of the gate oxide 739 and the potential of the bit line voltage source team 240 and the floating gate 785, the Fowler-N〇rdheim programming time can be completed in a few microseconds to a few milliseconds. [0099] In the read operation, the potential of the first shallow well bias source VTPW1 23G is set to the verify read potential (Vread) by 099103898 Form No. 1010101 Page 37/106 page 0992007313-0 201030947 The read potential (VREAD) is matched to the floating gate 785. The verify read potential (V) is set to READy between the erase critical potential Vtl and the programmed critical potential Vt2. The bit line is electrically generated by the BL 240 The potential is approximately set at h.ov. The second shallow well bias source VTPW2 255 Source line bias source 245 is disposed on a ground reference voltage source (0. 0V). The first deep well potential bias potential source 265 is provided at the power supply voltage source (VDD) of. [0100] [0101]
若該浮動閘非揮發性記憶體7〇〇被編程操作,那麼在該浮 動閘結785的電位大約是第二臨界電位,並且有導電電流 從該位元線電廢爆《^24♦讓雜繪幾動閛非揮發性記憶體 7〇〇而到達源極線偏壓源2 择發性記憶 體700的臨界電位是第二臨$電^取的數據 是二進制的「0」j若該浮動閘非揮發性記憶體7〇〇沒有 被選擇編程操作,那麼它的臨界電壓依..然是抹除臨界電 位vtl,故該浮動閘結785的n小於抹除抨界電位vtl 。當該浮動閘非揮發性記電流時 ,並且If the floating gate non-volatile memory 7 is programmed to operate, then the potential of the floating gate 785 is approximately the second critical potential, and a conductive current is electrically discharged from the bit line. Draw a few moving non-volatile memory 7〇〇 and reach the source line bias source 2 The critical potential of the selective memory 700 is the second data. The data is binary "0" if the floating If the gate non-volatile memory 7 is not selected for programming operation, then its threshold voltage depends on the erase potential vtl, so the n of the floating gate 785 is smaller than the erase boundary potential vtl. When the floating gate is non-volatile, the current is recorded, and
該浮動閘非揮發性記德想7/^44電1是抹除臨界電位 \ ^ s I ί~-Λ'6 ,所讀取的二進制數據為「i 广一 圖2a的單電晶體浮動閘非揮發性記憶艘綱的操作要點在 於在Fowler-Noniheim邊緣編程操作時該存儲M〇s電晶 體235的過麟除貞臨界電壓。若在—個㈣裡的任何存 儲M0S電晶體235有負臨界電壓,在正常讀取操作時它也 許會導致錯誤的感應數據。增加選擇閘極電晶㈣〇以形 成-「雙電晶體」的浮動閘非揮發性記憶體7⑽可消除在 編程操作時的過度抹I在讀取操作時該選擇閘極電晶 099103898 表單鎢號A0101 第38頁/共1〇6頁 0992007313-0 201030947 體790防止未被選擇的浮動閘非揮發性記憶體7〇〇的漏電 流。 [0102] ❹ [0103] ❹ 圖7c是圖7a所描述的浮動閘非揮發性記憶體7〇〇的示意圖 。該耦合電容器是前述的編程電荷耦合M〇s電晶體225。 該存儲MOS電晶體735和選擇閘極電晶體79〇的基極 (bulk)共同連接到第二淺井偏壓源VTpW2 255 〇與圖4a 相似,該存儲MOS電晶體735和選擇閘極電晶體79〇可直 接在基板205形成,則電荷耦合恥5電晶體225可以不需 要,使得該存儲MOS電晶趙735和選擇閘極電晶體790的 基極(bulk)透過基板205直接連接到到接地參考電壓源 (0.0V)。 說%_:·.. 該浮動閘非揮發性記憶想7.Q|勤草2 a的浮動閘 非揮發極記憶體200相诜是增加了單元的尺寸大小。然, 由於本發明浮動閘非揮發性記憶體的主要決定因素是在 該編程電荷搞合職電晶1晶體7 3 5之間 的大間距》在該浮動閘非@雜^^_700的存儲M0S電 晶體735裡增加的選擇閘所多加的面積是可 以忽略不計的。 [0104] 圖8a所示為一非揮發性記憶艘800,其包括由圖7c的浮動 閘非揮發性記憶體700所組成的陣列805 ^正如圖7c所描 述,每一浮動閘非揮發性記憶體70〇中包括一編程電荷耦 合MOS電晶體225、一存儲MOS電晶艎735和一選擇閘極電 晶體790。該陣列805中每一橫列的編程電荷耦合M〇s電 晶體225的汲極、源極和基極(buik)連接到字元線WL[0] 845a,…’ WL[1] 845b ’ 該等字元線WL[0] 845a,… 099103898 表單編號A0101 第39頁/共106頁 0992007313-0 201030947 ,WL[1] 845b都連接到該第一淺井偏壓源VTPW1。每 條位元線BL[0] 840a ’ BL[1] 840b,…,BL[n] 840η 連接到與陣列8 0 5相關聯的直行的浮動閘非揮發性記憶體 700的存儲MOS電晶體735的沒極。每一橫.列的浮動閘非 揮發性記憶體700的存儲MOS電晶體735和選擇閘極電晶 體790的基極(bulk)連接到相關聯的第二淺井偏壓源 VTPW2[0] 850a,...,VTPW2[1] 850b。每一橫列的 選擇閘極電晶體790的源極線連接到相關聯的源極線偏邀 源855。每一橫列的存儲MOS電晶體735的選擇閘極電晶 體790的選擇閘極79¾連接到相關聯的選擇閘極控制訊號 SG[0] 860a,…,SG[1 ] 8-60b » .,::.卜:':.::>'. 疆1 [0105] 元線感應放 ❹ 該位元線連接到位元線感應放九器815,泫 大器815包括位元線電壓源,該位元線電壓源可在編程和 讀取浮動閘非揮發性記憶體700時提供必要的偏壓。另, 該位元線感應放大器815還;包括τ:感應放六·器電路,該感 1 0| [r i; 1 ;! 應放大器電路可在讀取操作择的浮動閘非 h «.if ^ ^ |,- |j i ζ. I jj -·The floating gate non-volatile memory is 7/^44 electric 1 is the erase critical potential \ ^ s I ί~-Λ'6, the binary data read is "i Guangyi 1a single crystal floating gate The operational point of the non-volatile memory vessel is to store the threshold voltage of the M〇s transistor 235 during the Fowler-Noniheim edge programming operation. If any of the memory MOS transistors 235 in the (4) has a negative threshold Voltage, which may cause erroneous sensing data during normal read operation. Increase the selection of gated transistor (4) 〇 to form - "double transistor" floating gate non-volatile memory 7 (10) to eliminate excessive programming operation Wipe I select gate galvanic 099103898 during read operation Form Tungsten A0101 Page 38 / Total 1 〇 6 Page 0992007313-0 201030947 Body 790 Prevents leakage of unselected floating gate non-volatile memory 7〇〇 Current. [0102] FIG. 7c is a schematic diagram of the floating gate non-volatile memory 7〇〇 depicted in FIG. 7a. The coupling capacitor is the aforementioned programmed charge coupled M〇s transistor 225. The storage MOS transistor 735 and the base of the select gate transistor 79 are commonly connected to the second shallow well bias source VTpW2 255 相似 similar to FIG. 4a, the memory MOS transistor 735 and the select gate transistor 79. The germanium may be formed directly on the substrate 205, and the charge coupled dummy 5 transistor 225 may not be required, such that the bulk of the memory MOS transistor 735 and the select gate transistor 790 are directly connected to the ground reference through the substrate 205. Voltage source (0.0V). Say %_:·.. The floating gate non-volatile memory thinks 7.Q|Qincao 2 a floating gate The non-volatile pole memory 200 phase 诜 is the size of the unit. However, due to the floating thyristor non-volatile memory of the present invention, the main determinant is that the programming charge is engaged in the large spacing between the crystals of the crystal 1 crystal 5" in the floating gate non-hybrid ^ ^ _ 700 storage M0S The area added to the selection gate added to the transistor 735 is negligible. [0104] FIG. 8a shows a non-volatile memory boat 800 comprising an array 805 of floating gate non-volatile memory 700 of FIG. 7c. As described in FIG. 7c, each floating gate non-volatile memory The body 70 includes a programmed charge coupled MOS transistor 225, a memory MOS transistor 735, and a select gate transistor 790. The drain, source and base of the programmed charge coupled M〇s transistor 225 of each row in the array 805 are connected to word lines WL[0] 845a,...' WL[1] 845b ' The word line WL[0] 845a,... 099103898 Form No. A0101 Page 39/106 page 0992007313-0 201030947, WL[1] 845b are all connected to the first shallow well bias source VTPW1. Each bit line BL[0] 840a 'BL[1] 840b,...,BL[n] 840η is connected to the storage MOS transistor 735 of the floating floating gate non-volatile memory 700 associated with the array 850 Nothing. The storage MOS transistor 735 of each horizontal column of the floating gate non-volatile memory 700 and the base of the selection gate transistor 790 are connected to the associated second shallow well bias source VTPW2[0] 850a, ..., VTPW2[1] 850b. The source line of each row of select gate transistors 790 is coupled to an associated source line bias source 855. The select gates 829a of the select gate transistors 790 of each row of memory MOS transistors 735 are coupled to associated select gate control signals SG[0] 860a,..., SG[1] 8-60b » ., ::. Bu: ':.::>'. Xinjiang 1 [0105] The main line is connected to the bit line sensing device 815, and the 815 device includes a bit line voltage source. The bit line voltage source provides the necessary bias voltage when programming and reading the floating gate non-volatile memory 700. In addition, the bit line sense amplifier 815 also includes: τ: inductive discharge hex circuit, the sense 1 0| [ri; 1 ;! should be the amplifier circuit can be selected in the floating operation of the floating gate non-h «.if ^ ^ |,- |ji ζ. I jj -·
揮發性記憶體7 0 0的:數據狀辛數據输出口 835提供 :J T | I / f 所讀出的數據。 [0106] 該行解碼控制器810包括連接到字元線WL[0] 845a,... ,WL[1] 845b的第一淺井偏壓源、第二淺井偏壓源 VTPW2[〇] 850a,...,VTPW2[1] 850b、選擇閘極控制 訊號SG[0] 860a,…,SG[1] 860b和源極線偏壓源855 。另,一地址訊號820、數據訊號825和控制訊號830在 被解碼之後會連接到該行解碼控制器810和位元線感應放 大器815,以實現當編程操作和讀取該陣列805時對於所 099103898 表單編號A0101 第40頁/共106頁 0992007313-0 201030947 被選擇的浮動閘非揮發性記憶體700必要的控制和偏廢β [〇1〇7]請參閱圖8b,該非揮發性記憶體800包括由圖7c的浮動閘 非揮發性記憶體7〇〇所組成的陣列805。本實施方式中, 該浮動閘非揮發性記憶體70〇的結構本質上被反轉過來, 在該陣列805的直行上的每一浮動閘非揮發性記憶體7〇〇 的選擇閘極電晶體790的汲極對應連接到一位元線BL[〇] 840a ’BL[1] 840b,…,BL[n] 840η 和存儲 MOS 電晶 體225的源極^Volatile memory 700: Data sin data output port 835 provides: J T | I / f read data. [0106] The row decoding controller 810 includes a first shallow well bias source connected to the word lines WL[0] 845a, . . . , WL[1] 845b, and a second shallow well bias source VTPW2[〇] 850a, ..., VTPW2[1] 850b, select gate control signals SG[0] 860a,..., SG[1] 860b and source line bias source 855. In addition, an address signal 820, a data signal 825, and a control signal 830 are coupled to the row decoding controller 810 and the bit line sense amplifier 815 after being decoded to implement the programming operation and reading the array 805 for the 099103898. Form No. A0101 Page 40 / Total 106 Pages 0992007313-0 201030947 Selected floating gate non-volatile memory 700 necessary control and partial waste β [〇1〇7] Please refer to FIG. 8b, the non-volatile memory 800 includes An array 805 of floating gate non-volatile memory 7A of Figure 7c. In this embodiment, the structure of the floating gate non-volatile memory 70〇 is essentially reversed, and the gate transistor of each floating gate non-volatile memory 7〇〇 on the straight line of the array 805 is selected. The drain of 790 is correspondingly connected to one bit line BL[〇] 840a 'BL[1] 840b,...,BL[n] 840η and the source of the storage MOS transistor 225^
[0108] [0109][0109] [0109]
在該陣列805之内的所有編程電荷耦合M0S電晶體225的 汲極、源極和基極連接到共 元線CWL 845又連接到第 式中’該浮動閘非揮發性 中的陣列的好處是可以節省陣列8 〇 5的物理區域的尺寸。 每條位元線BL[〇] 840a,BL[1] 840b,...,BL[n] 840n對應連接到該陣列浮動閘非揮發 性記憶體700的存儲M0S電。每一橫列的 浮動閘非揮發性記憶體700〇f^§s電晶體735的基極 連接到相對應的第二淺井偏壓源VTPW2[〇] 85〇a,...,The benefits of connecting the drain, source and base of all of the programmed charge coupled MOS transistors 225 within the array 805 to the common line CWL 845 and to the array in the floating gate non-volatile are The size of the physical area of the array 8 〇 5 can be saved. Each of the bit lines BL[〇] 840a, BL[1] 840b, ..., BL[n] 840n corresponds to the stored MOS power connected to the array floating gate non-volatile memory 700. The floating gate of each row of non-volatile memory 700〇f^§s the base of the transistor 735 is connected to the corresponding second shallow well bias source VTPW2[〇] 85〇a,...,
,該共同字 。本實施方 5相比圖8a VTPW2[ 1 ] 850b。每一橫列的存儲M0S電晶體735的源極 連接到相對應的源極線偏壓源855。每一列的存儲M0S電 晶體735的每一選擇閘極電晶體790的選擇閘極792連接 到相對應的選擇閘極控制訊號SG[0] 860a,…,SG[1] 860b。 [0110] 099103898 該位元線連接到位元線感應放大器815,該位元線感應放 表單編號A0101 第41頁/共106頁 0992007313-0 201030947 大器815包括位元線電壓源,該位元線電壓源可在編程和 讀取浮動閘非揮發性記憶體700時提供必要的偏壓。另, 該位兀線感應放大器815遠包括一感應放大器電路,該感 應放大器電路可在讀取操作時感應被選擇的浮動閘非揮 發性s己憶體700的數據狀態並且在數據输出口 835提供所 讀出的數據。 [0111] [0112] 該行解碼控制器810包括連接到字元線WL[〇] 845a,… ’ WL[1] 845b的第一淺井偏壓源、第二淺井偏壓源 VTPW2[0] 850a , 訊號SG[〇] 860a 。另, VTPW2[1] 850b、選擇閘極控制 ’SG[1] 860b和源:極線偏壓源855 一地址訊號82〇、數制訊號830在 被解碼之後會連接到行解碼寿卞㉟8:#表·备元線感應放大 器815 ’以實現當今程操作和讀取談陣列gg 5時對於所被 選擇的浮動閘非揮發性記憶體7 〇 〇必要的控制和偏壓。 ❹ 在上述所有實施方式中’編__瓣wle卜 Nordhe i m的邊緣随道效應议旨減^例電晶體的臨界 電麼來完成的。在一般一:£_可:_的單層多晶石夕浮動 閘非揮發性記憶體裡,抹除操作是透過紫外光的照射而 抹除的。透過紫外光照射來抹除對於本發明的單層多晶 梦浮動閘非揮發性記憶體而言具有一個缺點,若為阻棚 紫外光而將包含浮動閘非揮發性記憶體的模塊封裝那 麼就不可能實現重複編程和抹除。在沒有專門的抹除和 編程設備的情況下,不可能再次修改封裝之内的資料數 據。故,所需要的是合併編程和抹除能力在同—浮動閘 非揮發性記憶體的設備。不同於一次性可編程的單層多 099103898 表單编號A0101 第42頁/共1〇6頁 0992007313-0 201030947 晶矽浮動閘非揮發性記憶體,本發明的編程或抹除耐力 週期超過一千次。 [0113] 請參閱圖9a,本發明單層多晶矽浮動閘非揮發性記憶體 900的又一個實施方式以電子編程和抹除的能力來達到避 免紫外光抹除的要求。編程電荷存儲單元7〇〇包括編程電 荷耗合MOS電晶艎225、存儲MOS電晶體735以及選擇間極 電晶體790。它們的構造和作用都與圖7a*7b中所插述的 相同。一P型材料的抹除元件905在第三淺擴散井915裡形 成。該第三淺擴散井915在由N型材料形成的第二深擴散 井910裡形成’而該第二深擴散井91〇又形成在該基板 2〇5的表面。該第一深擴散^ihg袴钵平择多晶梦邏 輯積體電路製程裡增加一層。當—_ 源極921和一 Ν型汲極9 2 2擴散在該^三^散井9丨5時, 將會形成一抹除電荷柄合MOS電晶體920 β —Ν型源極擴散 921和Ν型汲極擴散如2共呷地丨連薺到I抹 線路弘 95〇 〇 ίΠιβ1Ιβ€?υ〇ι ❹ [0114] -. .-,P-ooeriy 該抹除耦合浮動間923位於®#_區924上方,且由一 單層多晶矽而形成。該閘極氧化區924的材料和厚度與當 前CMOS領域的製程所常用的技術一致。該抹除耦合浮動 閘923、編程耦合浮動閘228和電荷存儲浮動閘738共同 地連接以形成一浮動閘結985。 [0115] 該第二深擴散井910經過連接擴散點925連接到第二深井 偏壓源930。該p型擴^連接點935連接第三淺擴散井915 到第三淺井偏壓源VTPW3 940 »該P型基板透過連接擴散 點270連接到接地參考電壓源(〇.〇v)。 099103898 表單編號A0101 第43頁/共106頁 0992007313-0 201030947 [〇116]該浮動問非揮發性記憶體900的抹除操 作是靠Fowlei·., the common word. This embodiment 5 is compared to Fig. 8a VTPW2[1] 850b. The source of each row of memory MOS transistors 735 is coupled to a corresponding source line bias source 855. The select gate 792 of each select gate transistor 790 of each column of memory MOS transistors 735 is coupled to a corresponding select gate control signal SG[0] 860a, ..., SG[1] 860b. [0110] 099103898 The bit line is connected to a bit line sense amplifier 815, which is placed on the form number A0101. Page 41/106 page 0992007313-0 201030947 The amplifier 815 includes a bit line voltage source, the bit line The voltage source provides the necessary bias voltage when programming and reading the floating gate non-volatile memory 700. In addition, the bit line sense amplifier 815 includes a sense amplifier circuit that senses the data state of the selected floating gate non-volatile memory 700 during the read operation and is provided at the data output port 835. The data read. [0112] The row decoding controller 810 includes a first shallow well bias source connected to word lines WL[〇] 845a,...' WL[1] 845b, and a second shallow well bias source VTPW2[0] 850a , signal SG [〇] 860a. In addition, VTPW2[1] 850b, select gate control 'SG[1] 860b and source: polar line bias source 855, an address signal 82, and a digital signal 830 are connected to the line decoding Shou 358 after being decoded: #表·Secondary line sense amplifier 815 'to achieve the necessary control and bias for the selected floating gate non-volatile memory 7 when operating the array and reading the array gg 5 . ❹ In all of the above embodiments, the edge of the Nordhe i m is determined by the effect of reducing the criticality of the transistor. In the general one: £_可: _ single-layer polycrystalline rock floating gate non-volatile memory, the erasing operation is erased by ultraviolet light. Erasing by ultraviolet light irradiation has a disadvantage for the single-layer polycrystalline dream floating gate non-volatile memory of the present invention, and if the module containing the floating gate non-volatile memory is packaged for blocking ultraviolet light, then It is impossible to achieve repetitive programming and erasing. In the absence of specialized erase and programming equipment, it is not possible to modify the data within the package again. Therefore, what is needed is a device that combines programming and erasing capabilities in the same-floating gate non-volatile memory. Different from one-time programmable single layer multi 099103898 Form No. A0101 Page 42 / Total 1〇6 Page 0992007313-0 201030947 Crystal floating gate non-volatile memory, the programming or erasing endurance period of the invention exceeds one thousand Times. Referring to FIG. 9a, yet another embodiment of the single layer polysilicon floating gate non-volatile memory 900 of the present invention achieves the requirement of avoiding ultraviolet light erasing by electronic programming and erasing capabilities. The programmed charge storage unit 7A includes a program charge MOS transistor 225, a memory MOS transistor 735, and a select transistor 790. Their construction and function are the same as those set forth in Figures 7a*7b. A wiper element 905 of a P-type material is formed in the third shallow diffusion well 915. The third shallow diffusion well 915 is formed in a second deep diffusion well 910 formed of an N-type material and the second deep diffusion well 91 is formed on the surface of the substrate 2〇5. The first deep diffusion ^ihg 袴钵 择 多 多 梦 梦 梦 梦 梦 梦 。 。 。 增加 增加 增加 增加 增加 增加 增加 增加When the - _ source 921 and the 汲-type 9 9 2 2 are diffused in the ^ 3 ^ well 9 丨 5, a wipe-off charge MOS transistor 920 β - 源 source diffusion 921 and Ν will be formed. Type bungee diffusion such as 2 呷 呷 荠 I I I I I 弘 弘 弘 弘 弘 弘 弘 弘 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 [0114] -. . Above 924, and formed by a single layer of polysilicon. The material and thickness of the gate oxide region 924 are consistent with techniques commonly used in current CMOS field processes. The erase coupling floating gate 923, the program coupled floating gate 228, and the charge storage floating gate 738 are commonly connected to form a floating gate 985. [0115] The second deep diffusion well 910 is connected to the second deep well bias source 930 via a connection diffusion point 925. The p-type expansion junction 935 connects the third shallow diffusion well 915 to the third shallow well bias source VTPW3 940 » the P-type substrate is connected to the ground reference voltage source (〇.〇v) through the connection diffusion point 270. 099103898 Form No. A0101 Page 43 of 106 0992007313-0 201030947 [〇116] The floating operation of the non-volatile memory 900 is based on Fowlei·.
Nordheim通道隧道效應來完成的 道随道效應比上述實施方式中㈣述的Fowl ―⑽賴道效應需要—個更㈣電位才能 執行編程操作和抹_作。表2是對該浮動閘非揮發性= 憶體_伽Urn叫糾道效應和邊道己 效應的電壓要求的比較。 崎The channel effect of the Nordheim channel tunneling effect is required to perform a programming operation and a wiper operation than the Fowl-(10) ray effect described in (4) of the above embodiment. Table 2 is a comparison of the voltage requirements for the floating gate non-volatile = memory-gamma Urn called the correction effect and the side effect. Saki
Fowler-Nordh er- ei [0117]表 2 [0118]Fowler-Nordh er- ei [0117] Table 2 [0118]
圖9b提供本實施方式*浮__輯記㈣_編程操 作時所需要的電位e SFowler-Nordheim通道隧道抹除 操作時,該抹除電荷耦合M0S電晶體920從它的第三淺擴 散井915的通道注射電子到該抹除耦合浮動閘923,並且 使知·電子經過該通道閘極氧化區924而到違該浮動閘結 985。該抹除耦合浮動閘923用作耦合一負浮動閘極電壓 到該浮動閘結985,從而感應導致執行Fowler_Figure 9b provides the potential e SFowler-Nordheim channel tunnel erase operation required for the present embodiment of the present invention. The erased charge coupled MOS transistor 920 is from its third shallow diffusion well 915. The channel injects electrons into the erase coupling floating gate 923 and causes the electrons to pass through the gate gate oxide region 924 to the floating gate 985. The erase coupled floating gate 923 acts to couple a negative floating gate voltage to the floating gate 985, thereby inducing the execution of Fowler_
Nordheim邊緣隧道編程操作。在該浮動閘非揮發性記憶 體單疋900的通道抹除操作或邊緣編程操作的過程中,該 099103898 表單編號A0101 第44頁/共106頁 0992007313-0 201030947 抹除電荷搞合MOS電晶體92G、編程電荷耗合廳電晶體 225和存儲_電晶體735的尺寸必須能使得在7. 〇v之下 可以完成編程操作和抹除操作。 [0119] 在Fowler-Nordheim通道隧道抹除操作過程中,當該第 -井偏壓源VTPW1 230的電位被設置為大約+ 7 〇v並且被 施加於第一淺擴散井215時,該浮閘結985及存儲浮動閘 228的電位必須為大約+ 6 3V。如此即會使得一個大約 + 6. 3V的電位被耦合到該浮動閘結985。該編程電荷耦合 M0S電晶體225的面積與抹除電荷耦合M〇s電晶體92〇和存 儲M0S電晶體735兩者合鵡來面澈麟尺寸相比要大過十倍 才能期望達到耦合比率超偏壓源 VTPW3 940的電位被設置對得第三淺擴 散井9丨各偏置。另’該抹除控制線路950的電位被設置到 大約-7. 0V,以額外保證第三淺擴散井915的通道區975 籲 的電位被設置到大約-7. 〇V# 為可以浮動地或被設置到Nordheim edge tunnel programming operations. During the channel erase operation or edge programming operation of the floating gate non-volatile memory unit 900, the 099103898 form number A0101 page 44/106 page 0992007313-0 201030947 erase the charge to engage the MOS transistor 92G The programming charge chamber hall 225 and the storage_transistor 735 must be sized such that the programming operation and the erase operation can be performed under 7. 〇v. [0119] During the Fowler-Nordheim channel tunnel erasing operation, when the potential of the first well bias source VTPW1 230 is set to approximately +7 〇v and is applied to the first shallow diffusion well 215, the floating gate The junction 985 and the storage floating gate 228 must have a potential of approximately + 6 3V. This will cause a potential of approximately + 6. 3V to be coupled to the floating gate 985. The area of the programmed charge coupled MOS transistor 225 is equal to the erase charge coupled M 〇s transistor 92 〇 and the stored MOS transistor 735. The potential of the bias source VTPW3 940 is set to be offset by the third shallow diffusion well 9丨. Further, the potential of the erase control line 950 is set to about -7.0 V to additionally ensure that the potential of the channel region 975 of the third shallow diffusion well 915 is set to about -7. 〇V# is floatable or Set to
該I二淺擴錄井220的電位 I二淺· (D. 0V)。該位元 ., ,· C- ' ί^:' 線電壓源BL 240、源極線與《霉||、2+45和選擇閘極控制訊. V ?F^iC© 號795均被設置到接地參考'^位(〇 〇v) ^該第一深井偏 壓源265和第一淺井偏壓源VTPWi 230的電位被設置到大 約+ 7. 0V。該第二淺井偏壓源VTPW2 255和第二深井偏壓 源930的電位為可以浮動地或被設置到接地參考電位 (0.0V)。 當該第三淺井偏壓源VTPW3 940的電位被設置到大約-7. 0V以使得第三淺擴散井915偏置時,該抹除電荷耦合 M0S電晶體920的閘極氧化區924將有大約13. 3V的電壓在 099103898 表單編號Α0101 第45頁/共1〇6頁 0992007313-0 [0120] 201030947 洋動閘結985和第三淺擴散井915之間。該抹除電荷轉合 M0S電晶體92G的氧化區厚度是足夠地厚,從而可導致產 生Fowler-Nordheim通道隧道效應。在該第三淺擴散井 915的電子經過抹除電荷耦合M0S電晶體920的通道閘極 氡化區924被注射進入浮動閘結985。由於+13. 3V的電場 是橫跨抹除電荷耦合M〇s電晶體92〇的整體通道975,故 該抹除操作被定義為Fowler_N〇rdheim通道抹除操作。 [0121] 由於該編程電荷耦合M0S電晶體225的尺寸較大,閘極氧The potential of the I-two shallow expansion well 220 is two shallow (D. 0V). The bit ., , · C- ' ί^: ' Line voltage source BL 240, source line and "Mold ||, 2+45 and select gate control. V ?F^iC© No. 795 are set To the ground reference '^ bit (〇〇v) ^ The potential of the first deep well bias source 265 and the first shallow well bias source VTPWi 230 is set to approximately +7.00V. The potential of the second shallow well bias source VTPW2 255 and the second deep well bias source 930 can be floating or set to a ground reference potential (0.0V). When the potential of the third shallow well bias source VTPW3 940 is set to about -7. 0V to bias the third shallow diffusion well 915, the gate oxidized region 924 of the erased charge coupled MOS transistor 920 will have approximately 13. The voltage of 3V is 099103898 Form No. 1010101 Page 45 / Total 1 〇 6 Page 0992007313-0 [0120] 201030947 Between the kinetic gate 985 and the third shallow diffusion well 915. The erased charge-transferred MOS transistor 92G has an oxidized region thickness that is sufficiently thick to cause Fowler-Nordheim channel tunneling. Electrons in the third shallow diffusion well 915 are injected into the floating gate junction 985 via the channel gate deuteration region 924 of the erase charge coupled MOS transistor 920. Since the electric field of +13. 3V is across the entire channel 975 of the erased charge coupled M〇s transistor 92, the erase operation is defined as the Fowler_N〇rdheim channel erase operation. [0121] Due to the large size of the programmed charge coupled MOS transistor 225, the gate oxygen
化區229在抹除操作時的電場僅有+〇. 7V。如此即可防止 編程電荷耦合M0S電晶嫌瑟2,5,的所有隧道效應。同樣地, 當浮動閘結9 8 5的電位為大的·+ 6) 3 V時努淺擴散井 220偏置於接地參考電位(〇^闌極氧化區 739的電壓位僅為+ 6. 3V ’其不夠大到可產生F〇wler_The electric field of the 229 is only +〇. 7V during the erase operation. This prevents all tunneling effects of programming charge coupled M0S electro-crystals 2,5. Similarly, when the potential of the floating gate junction 985 is large + 6) 3 V, the shallow diffusion well 220 is biased to the ground reference potential (the voltage level of the 氧化 阑 氧化 oxidation zone 739 is only + 6. 3V). 'It is not big enough to produce F〇wler_
Nordheim随道效應。如此,在抹除操作時,只有抹除電 荷搞合M0S電晶艘9 20具有r個分高的電揚可在抹除操 作時引起Fowler-Nordhei,i^;it效應。在:抹除操作完成 ^ r 『Τ\/Nordheim's accompanying effect. Thus, in the erasing operation, only the erased charge is combined with the M0S electric crystal boat 9 20 having r heights of the electric lift to cause the Fowler-Nordhei, i^;it effect during the erasing operation. In: erase operation is completed ^ r 『Τ\/
以後,由於電子在浮】了,從而使得存儲 M0S電晶體735的臨界電壓增加。本實施方式中,對於_ 個+ 3. 0V的電源電廢而言,該浮動閘非揮發性記憶體9〇〇 的抹除臨界電壓目標值被設計在超過+ 4. 〇V的電位。 [0122] 在Fowler-Nordheim邊緣隧道編程操作時,第一井偏壓 源VTPffl 230被設置在大約-7. 0V並且被施加於第一淺擴 散井215。位元線電壓源BL 240被設置在電位大約+ 5. 〇v ,以被施加於被選擇的浮動閘非揮發性記憶體900,或被 設置在接地參考電位(0.0V),以被施加於沒被選擇的浮 099103898 表單编號A0101 第46頁/共106頁 0992007313-0 201030947 動閘非揮發性記憶體900。抹除控制線路95〇被設置在接 地參考電位(0.0V)。第二淺擴散井220可以浮動或被設 置在接地參考電位(ο. ον)。源極線偏壓源245被允許可 以浮動或設置到接地參考電位(〇. 0V)。第三淺井偏壓源 VTPW3 940、抹除控制線路950、選擇閘極的控制訊號 795、第一深井偏壓源265和第二深井偏壓源93〇全部被 設置為接地參考電位(0.0V)» [0123] ❹ 當上述偏壓被應用在浮動閘非揮發性記憶鱧9〇〇時由於 90%耦合比率是從第一淺擴散井215的_ 7 〇v電位通過存 儲MOS電晶體735而來的’故浮動閉結985的電壓大約是_Later, since the electrons are floating, the threshold voltage of the memory MOS transistor 735 is increased. In the present embodiment, for the power supply waste of _ + 3.0 V, the target value of the erase threshold voltage of the floating gate non-volatile memory 9 被 is designed to exceed the potential of + 4. 〇V. [0122] In the Fowler-Nordheim edge tunnel programming operation, the first well bias source VTPff1 230 is set at approximately -7. 0V and applied to the first shallow diffusion well 215. The bit line voltage source BL 240 is set at a potential of approximately + 5. 〇v to be applied to the selected floating gate non-volatile memory 900, or is set at a ground reference potential (0.0V) to be applied to Float 099103898 Form No. A0101 Page 46 / Total 106 Pages 0992007313-0 201030947 Non-volatile memory 900. The erase control line 95 is set at the ground reference potential (0.0V). The second shallow diffusion well 220 can be floated or set at a ground reference potential (ο. ον). The source line bias source 245 is allowed to float or set to a ground reference potential (〇. 0V). The third shallow well bias source VTPW3 940, the erase control line 950, the select gate control signal 795, the first deep well bias source 265, and the second deep well bias source 93 are all set to the ground reference potential (0.0V). » [0123] ❹ When the above bias voltage is applied to the floating gate non-volatile memory 鳢9〇〇, since the 90% coupling ratio is from the _ 7 〇v potential of the first shallow diffusion well 215 through the storage MOS transistor 735 'The voltage of the floating closed junction 985 is about _
6.3V。由於位元線電壓源ΒΙ/2·4〇被 選擇閘極控制訊號795被控餘鮮緣〇ν) 故存儲MOS電晶體735的源極737是浮動的,且在存儲m〇S μ 0V,並且 電晶體735的汲極736處會感應Fowler—Nordheim邊緣隧 道編程效應◊電子從浮動1結9續透過汲轉736進入第二 淺擴散井22q。該存卿界電壓在經過 參 [0124] «无决疋的一段編程時間文今味磬/約+3 〇 v到大約 4. 0V的抹除臨界電位減小到大約_丨〇v到大約丨.的編 程臨界電壓期望值。 本實施方式中,由於存儲M0S電晶體735的雙電晶體單元 ^構和選擇閉極電晶想790的存在,該存儲M0S電晶體 735的負極臨界電壓是被允許的。由於該選擇閘極電晶體 790是一改進的NM0S器件,透過將選擇閘極控制訊號795 α置為接地參考電位(0.0V),可使得所有存儲M0S電晶 胃735的負極臨界電壓所漏出的電流都被選擇閘極電晶體 099103898 表單編號Α0101 第47頁/共1〇6頁 0992007313-0 201030947 [0125] 79〇所阻攔。 讀取操作請參閱圖7b,該第三淺井偏壓源VTPW3 940和 抹除控制線路950的電位被設置為接地參考電位(〇 〇Ό 。邊第二深井偏壓源910可以浮動或被設置為接地參考電 位(0.0V)。即使抹除編程電位比較大,其數據狀態的檢 測與上面仍然相同。 [0126] 請參閱圖8a,一組浮動閘非揮發性記憶體9〇〇被安排成行 和列與在圏8a顯示的陣列相似。必要的偏壓和控制訊號 被加到行解碼控制器81〇和位元線感應放大器815,以完 成對本實施例中的浮動閘非揮,發样記,體异〇 〇的抹除和編6.3V. Since the bit line voltage source ΒΙ/2·4 〇 is selected by the gate control signal 795 to control the residual margin 〇 ν), the source 737 of the memory MOS transistor 735 is floating and is stored at m 〇 S μ 0V, And the Fowler-Nordheim edge tunnel programming effect is induced at the drain 736 of the transistor 735. The electrons continue to pass from the floating 1 junction 9 through the twist 736 into the second shallow diffusion well 22q. The voltage of the sacred boundary is reduced by the reference [0124] «a period of programming without a decision 文 磬 约 / about +3 〇 v to about 4. 0V erase the critical potential is reduced to about _ 丨〇 v to about 丨The programming threshold voltage expectation. In the present embodiment, the negative threshold voltage of the memory MOS transistor 735 is allowed due to the presence of the dual transistor unit structure of the MOS transistor 735 and the selection of the closed-electrode crystal 790. Since the select gate transistor 790 is a modified NMOS device, by setting the select gate control signal 795α to the ground reference potential (0.0V), all the negative voltages of the memory of the MOS transistor 735 can be leaked. The current is selected by the gate transistor 099103898 Form No. 1010101 Page 47 / Total 1 〇 6 Page 0992007313-0 201030947 [0125] 79 〇 blocked. Read operation Referring to Figure 7b, the potential of the third shallow well bias source VTPW3 940 and the erase control line 950 is set to the ground reference potential (〇〇Ό. The second deep well bias source 910 can be floated or set to Ground reference potential (0.0V). Even if the erase programming potential is large, the detection of the data state is the same as above. [0126] Referring to Figure 8a, a set of floating gate non-volatile memory 9 is arranged in rows and The columns are similar to the array shown in Figure 8a. The necessary bias and control signals are applied to the row decode controller 81A and the bit line sense amplifier 815 to complete the floating gate non-swinging, sample recording, Wiping and editing
程操作 [0127] 1和地4、#: 請參閱厲靡a,為本發明單層多蠤V浮刼輪非揮發性記憶 體1 000的實施方式的橫切面示意圖。本實施方式中,該 單層多晶矽浮動閘非揮發性記憶體1〇〇〇在P型基板1005 上形成。一n型雜質的第一第二深擴散井 1014在基板1005的表面形鉍秦一策二深擴散井1〇12 ......1¾¾體電路之製程中增 w ψOperation [0127] 1 and ground 4, #: Please refer to the schematic diagram of a cross-sectional view of an embodiment of the single-layer multi-turn V floating wheel non-volatile memory 1000 of the present invention. In the present embodiment, the single-layer polysilicon floating gate non-volatile memory 1 is formed on the P-type substrate 1005. The first second deep diffusion well 1014 of an n-type impurity is increased in the process of the surface of the substrate 1005 in the process of the 1st 12th ...
加一處理層即可得到。第一深擴散井1012透過N型連接點 1062連接到第一深井偏壓源1030。第二深擴散井1014透 過N型連接點1060連接到第二深井偏壓源1 065。P型基板 1005透過連接擴散點1070連接到接地參考電壓源(0. 0V) P型雜質的第一淺擴散井1015形成於第一深擴散井1012 之内,P型雜質的第二淺擴散井1020則形成於第二深擴散 099103898 表單編獍A0101 第48頁/共106頁 0992007313-0 [0128] 201030947 井1014之内。電荷耦合M0S電晶體1025連接作為一電容 器。該電荷耦合M0S電晶艎1025在第一淺擴散井1015的 表面形成,並且該電荷耦合M0S電晶體1025具有擴散至第 一淺擴散井101 5内的N型雜質的源極1 026和汲極1027。 該源極1026、汲極1〇27和第一淺擴散井1〇15共同地連接 到第一淺井偏壓源VTPW1 1050。第一淺擴散井1015透過 連接擴散點1016連接到第一淺井偏壓1〇5〇。轉 合浮動閘1028在閘極氧化區1〇29由一單層多晶矽形成。 其中’該閘極氧化區1029的材料和厚度與當前CMOS領域 的製程中常用的技術一致βAdd a treatment layer to get. The first deep diffusion well 1012 is coupled to the first deep well bias source 1030 through an N-type connection point 1062. The second deep diffusion well 1014 is coupled to the second deep well bias source 1 065 through an N-type connection point 1060. The P-type substrate 1005 is connected to the ground reference voltage source (0. 0V) through the connection diffusion point 1070. The first shallow diffusion well 1015 of the P-type impurity is formed in the first deep diffusion well 1012, and the second shallow diffusion well of the P-type impurity 1020 is formed in the second deep diffusion 099103898 Form Compilation A0101 Page 48 / Total 106 Page 0992007313-0 [0128] 201030947 Well 1014. The charge coupled MOS transistor 1025 is connected as a capacitor. The charge coupled MOS transistor 1025 is formed on the surface of the first shallow diffusion well 1015, and the charge coupled MOS transistor 1025 has a source 1 026 and a drain that diffuse into the N-type impurity in the first shallow diffusion well 1015. 1027. The source 1026, the drain 1〇27 and the first shallow diffusion well 1〇15 are commonly connected to the first shallow well bias source VTPW1 1050. The first shallow diffusion well 1015 is connected to the first shallow well bias 1〇5〇 through the connection diffusion point 1016. The transfer floating gate 1028 is formed by a single layer of polysilicon in the gate oxide region 1〇29. Wherein the material and thickness of the gate oxide region 1029 are consistent with the techniques commonly used in current CMOS field processes.
. il|·^· ' IPjS.il|·^· ' IPjS
[0129] 0 之後,第二淺擴散井1〇2〇^|·歌成電晶體 1035。該舞儲M0S電晶體1035努一至弟二淺擴散井 .‘·., 士! ::ΐ 肖1038在j閘極氧化區 咖11〇39的材料 1020的Ν盤汲極1036和Ρ型源極1037。該汲極1036連接 到一位元線電壓源BL 1040,該源極1037則連接到一源 極線偏壓源1〇45。電荷存_澤鶊| 10 3 9内由一單層多晶矽形ί ▲ 广 PropeiT·; 和厚度與當前CMOS領域的所常用的技術一致。該 第二淺擴散井1020透過連it點1021連接到第二淺井 偏壓源VTPW2 1 055。 [0130] 同時,在第二淺擴散井1020内形成一選擇閘極電晶體 1090。該選擇閘極電晶體1090的N型源極1093擴散在第 二淺擴散井1020内。該選擇閘極電晶體1090的汲極1〇37 ,也就是存儲M0S電晶體1035的源極1037,是由一N型雜 質擴散在第二淺擴散井1020裡形成的。該選擇閘極電晶 體1 090的源極1 093連接到源極線電壓源SL 1045。該選 099103898 表單編號A0101 第49頁/共1〇6頁 0992007313-0 201030947 擇閘極電晶體1090的選擇閘極1〇92在閘極氧化區1〇94内 由一單層多晶矽形成。該選擇閘極1〇92連接到_選擇閘 極控制訊號1095。在編程和讀取操作時,該選擇閘極電 晶體1090用於控制連接該存儲M〇s電晶體1〇35到源極線 1045。 [0131] 與圖9a中單層多晶矽浮動閘非揮發性記憶鱧9〇〇類似編 程操作是使用Fowler-Nordheim邊緣隧道效應,而抹除 操作則是使用Fowler-Nordheim通道隧道效應。本實施 方式中,該電荷耦合M〇s電晶艟1〇25被要求執行内建在同 一晶片中的Fowler-Nordhei*通遒抹除,操作和Fowler-[0129] After 0, the second shallow diffusion well 1〇2〇^| The dance megaphone 1035 Nuo to the younger shallow diffusion well. ‘·., 士! ::ΐ Xiao 1038 in the gate oxidized zone of the j. The material of the coffee 11 〇 39 1020 Ν 汲 10 1036 and Ρ source 1037. The drain 1036 is coupled to a one-bit line voltage source BL 1040, which is coupled to a source line bias source 1〇45. Charge storage _ Ze 鶊 | 10 3 9 by a single layer polycrystalline ί ▲ Wide PropeiT ·; and thickness is consistent with the current commonly used technology in the field of CMOS. The second shallow diffusion well 1020 is connected to the second shallow well bias source VTPW2 1 055 through a connection point 1021. [0130] Meanwhile, a select gate transistor 1090 is formed in the second shallow diffusion well 1020. The N-type source 1093 of the select gate transistor 1090 diffuses within the second shallow diffusion well 1020. The drain 1 〇 37 of the selected gate transistor 1090, that is, the source 1037 of the MOS transistor 1035, is formed by diffusion of an N-type impurity in the second shallow diffusion well 1020. The source 1 093 of the selected gate transistor 1 090 is connected to the source line voltage source SL 1045. The selection 099103898 Form No. A0101 Page 49 / Total 1 Page 6 0992007313-0 201030947 The selection gate 1〇92 of the gate transistor 1090 is formed by a single layer of polysilicon in the gate oxide region 1〇94. The select gate 1〇92 is connected to the _select gate control signal 1095. The select gate transistor 1090 is used to control the connection of the memory M〇s transistor 1〇35 to the source line 1045 during programming and read operations. [0131] A similar operation to the single-layer polysilicon floating gate non-volatile memory 〇〇9〇〇 in Figure 9a is the use of Fowler-Nordheim edge tunneling, while the erase operation uses the Fowler-Nordheim channel tunneling effect. In this embodiment, the charge coupled M〇s transistor 〇1〇25 is required to perform Fowler-Nordhei* pass-through erase, operation and Fowler- built in the same wafer.
Nordheim 邊緣編 程操作 圖9a中,該存儲M0S電晶Nordheim Edge Programming Operation In Figure 9a, the memory MOS transistor
[0132][0132]
Nordheim通道抹除操作之後會被增加β另,該存儲M〇s 電晶競1035的臨界電麼在Fowler-Nordheim邊緣編程操 作以後會被減少。電子透礴_ §零肩體丨〇 3 5的閘極 氧化區1039被注射到浮動增加該存儲M〇s • .:.':: #The Nordheim channel erase operation will be increased by β. The memory of the M〇s Cryptography 1035 will be reduced after the Fowler-Nordheim edge programming operation. Electron 礴 § § zero shoulder body 丨〇 3 5 gate oxidized zone 1039 is injected into the float to increase the storage M〇s • .:.':: #
電晶體10 3 5的臨界電壓。f":5 [0133]請參閱圖10b ’在抹除操作時,第一淺井偏壓源VTPW1 1 050的電位被設置到大約+ 7. 0V,以被施加於該電荷耦 合M0S電晶體1025的第一淺擴散井1015。第二淺井偏壓 源VTPW2 1055的電位則被設置到大約-7.0V,以被施 加於第二淺擴散井1020。源極線電壓源1 045的電位大約 被設置到-7. 0V。位元線電壓源1040與存儲M0S電晶體 1035的汲極1036相分離而使得沒極1036浮動。第一深井 偏壓源1 030及第一深擴散井1〇12的電位被設置到大約 099103898 表單編號Α0101 第50頁/共1〇6頁 0992007313-0 201030947The threshold voltage of the transistor 10 3 5 . f":5 [0133] Please refer to FIG. 10b' During the erase operation, the potential of the first shallow well bias source VTPW1 1 050 is set to approximately +7.0 V to be applied to the charge coupled MOS transistor 1025. The first shallow diffusion well 1015. The potential of the second shallow well bias source VTPW2 1055 is set to approximately -7.0 V to be applied to the second shallow diffusion well 1020. The potential of the source line voltage source 1 045 is set to about -7. 0V. The bit line voltage source 1040 is separated from the drain 1036 storing the MOS transistor 1035 such that the pole 1036 floats. The potential of the first deep well bias source 1 030 and the first deep diffusion well 1〇12 is set to approximately 099103898 Form No. Α0101 Page 50/Total 1〇6 Page 0992007313-0 201030947
[0134] ❷ [0135] + 7. OV ’以導致在Fowler-Nordheim通道抹除操作的同 時防止不需要的向前電流在第一深擴散井1〇12、第二深 擴散井1014 '第一淺擴散井1〇15和第二淺擴散井1〇2〇裡 流動。選擇閘極控制訊鐃1 095和第二深井偏麗源1〇65的 電位被設置到接地參考電.位(0.0V)。與之前的單元结構 相同’電何耗合M0S電晶體1025的尺寸大於存健m〇s電晶 體1035。電荷耦合M0S電晶體1025和存儲M0S電晶體 1035之間理想的尺寸比率是10倍或更大以達到超過9〇% 的耦合比率’從而可以減少在低電壓的邏輯製程裡所必 需的高電壓。 七七J 攀:ί' ' II繼f 如上所述,由於從第一淺擴麩丨^比率 (9090再加上透過電荷耦合容器的電位 大約是+7. 0V,偏壓會強迫在浮動閘結1〇{^5的電位大約 是+ 6‘ 3V。當第二淺擴散井1〇2〇的電位大約為_7. “時, 一個大電場將建立在浮動竹,5和,二P擴散井1020 之間的通道閉極氧化属10母,T释潔g|〇wler_[0134] 7. [0135] + 7. OV 'to prevent unwanted forward currents in the Fowler-Nordheim channel erase operation while the first deep diffusion well 1〇12, the second deep diffusion well 1014' first The shallow diffusion well 1〇15 and the second shallow diffusion well flow in 1〇2〇. The potential of the gate control signal 1 095 and the second deep well 1 〇 65 is set to the ground reference voltage (0.0V). The same as the previous cell structure, the size of the MOS transistor 1025 is larger than that of the memory cell 1035. The ideal size ratio between the charge coupled MOS transistor 1025 and the memory MOS transistor 1035 is 10 times or more to achieve a coupling ratio of more than 〇%, thereby reducing the high voltage necessary in a low voltage logic process.七七J Climb: ί' 'II Following f As mentioned above, due to the ratio of the first shallow gluten ( ^ (9090 plus the potential through the charge coupled container is about +7.0 V, the bias will force the floating gate The potential of the junction 1〇{^5 is approximately +6' 3V. When the potential of the second shallow diffusion well 1〇2〇 is approximately _7. “When a large electric field will be established in floating bamboo, 5 and 2 P diffusion The closed-end oxidation of the channel between the wells 1020 is 10 mothers, and the T release is g|〇wler_
Nordheim通道隧瑱鵁應的第二淺擴散井 1020經過存儲M0S電晶體1035的i道閘極氧化區1〇39被 注射進入電荷存儲浮動閘1038到浮動閘結1〇85,故在通 道抹除操作以後,浮動閘結1 085的電子数量將被增加, 如此即可使得存儲M0S電晶體1035的臨界電壓增加到大約 4. 0V的期望值,適合該浮動閘非揮發性記憶體1〇〇〇的列 陣所使用的接近3. 0V的電源電壓源(vdd)。 在Fowler-Nordheim邊緣隧道編程操作時,第一井偏壓 源VDNW1 1030的電位被設置到大約_7〇v,以被施加於 099103898 表單編號A0101 第51頁/共106頁 0992007313-0 201030947 第一淺擴散井1015 ^該位元線電壓源BL 1〇4〇的電位被 設置到大約+ 5. 0V,以被施加於被選擇的浮動閘非揮發性 記憶體1000,或者被設置到接地參考電位(〇 〇v),以被 施加於沒被選擇的浮動閘非揮發性記憶體1〇〇〇。第二淺 井偏壓源VTPW2 1055被隔離或連接到接地參考電位 (0.0V),以使得第二淺擴散井1020可以被允許浮動或被 設置到接地參考電位(〇. ον)。源極線偏壓源1〇45可以允 許浮動或設置到接地參考電位(〇. 0V)。選擇閘極控制訊 號1095、第一深井偏壓源1030和第二深井偏壓源1〇65全 部被設置到接地參考電银(1观)。 ·. - · · ^ [0136]The second shallow diffusion well 1020 of the Nordheim channel tunnel is injected into the charge storage floating gate 1038 to the floating gate junction 1〇85 through the i-channel gate oxidation region 1〇39 of the storage MOS transistor 1035, so that the channel is erased. After operation, the number of electrons of the floating gate junction 1 085 will be increased, so that the threshold voltage of the memory MOS transistor 1035 is increased to an expected value of about 4.0 V, which is suitable for the floating gate non-volatile memory. A supply voltage source (vdd) of approximately 3.0 V used by the array. During the Fowler-Nordheim edge tunnel programming operation, the potential of the first well bias source VDNW1 1030 is set to approximately _7 〇 v to be applied to 099103898 Form No. A0101 Page 51 / Total 106 Pages 0992007313-0 201030947 First Shallow diffusion well 1015 ^ The potential of the bit line voltage source BL 1〇4〇 is set to approximately + 5. 0V to be applied to the selected floating gate non-volatile memory 1000, or set to the ground reference potential (〇〇v) to be applied to the non-volatile memory of the floating gate that is not selected. The second shallow well bias source VTPW2 1055 is isolated or connected to a ground reference potential (0.0V) such that the second shallow diffusion well 1020 can be allowed to float or be set to a ground reference potential (〇. ον). The source line bias source 1〇45 can be allowed to float or set to the ground reference potential (〇.0V). The select gate control signal 1095, the first deep well bias source 1030, and the second deep well bias source 1〇65 are all set to ground reference silver (1). ·. - · · ^ [0136]
當上述偏壓被施加於浮動閘^揮^•性,·钱嘴;〖〇〇〇時,由 於90%耦合從第一淺擴散井位透過電荷轉 合M0S電晶體1025而來,故在浮動閘結^85的電屋大約 為-6. 3V。由於位元線電壓源BL 1040的電位被設置在 + 5. 0V.,並且選擇閘極控,訊。號;||〇95被控言】在接地參考 電位CO. 0V:) ’故存儲nos推|1|歸^8‘1〇37是浮動 propo 卞莉When the above-mentioned bias voltage is applied to the floating gate, the money mouth; when 〇〇〇, since 90% of the coupling is transmitted from the first shallow diffusion well through the charge-transferred MOS transistor 1025, it is floating The electricity house of the gate junction ^85 is approximately -6. 3V. Since the potential of the bit line voltage source BL 1040 is set at + 5. 0V., and the gate is controlled, the signal is selected. No.||〇95 is being controlled] in the ground reference potential CO. 0V:) ‘so store nos push|1|return ^8‘1〇37 is floating propo 卞莉
的’存健H〇S電晶激1035的成牌纟Ϊρβ6$會發生F〇wier一 _ 麵嚷琴+5适爭.::w \ β ψ j I ή: -SiuieS' - -»· ~The '’健H〇S electric crystal 1035's card 纟Ϊρβ6$ will occur F〇wier one _ face 嚷琴+5 contend.::w \ β ψ j I ή: -SiuieS' - -»· ~
Nordheim邊緣隧道編程效應。電子從浮動閘結1〇85透過 汲極1 036進入第二淺擴散井1020。存儲M0S電晶體1035 的臨界電壓在經過預先決定的一個編程時間以後,從大 約+ 3. 0V到+4· 0V的抹除臨界電壓被減小到大約+ 1 的 編程臨界電壓期望值。 本實施方式的單層多晶發浮動閘非揮發性記憶體1 〇 0 0比 圏9a中的非揮發性記憶體900的好處是只需要一個nm〇S 電容器(電荷耦合M0S電晶體1025)來實現内建在同一晶 099103898 表單編珑A0101 第52頁/共1〇6頁 0992007313-0 [0137] 201030947 [0138] [0139] ❹ [0140] .❹ [0141] 片上的F〇wler-Nordheiin 抹除操作和 F〇wlerN〇rdheim 編程操作。且浮動閘非揮發性記憶體1〇〇〇的物理尺寸較 小,因而可減少基板1025的費用。 一組浮動閘非揮發性記憶體10〇〇可以被安排成行和列與 在圖8a顯示的列陣相似。必要的偏壓和控制訊號被加到 行解碼控制器810和位元線感應放大器,以實現浮動 閘非揮發性記憶體1〇〇〇的抹除和編程操作。 清參閲圖11a ’本發明單層多晶料動_揮發性記憶體 11 〇〇的另一具體實施方式的結構舆圖9线本質上是相同的 ,除了第二深擴散汫W0请^枣有多第二深擴散 井1110替換之外。第二深擴.散井壤二深 壓源113G。抹除單元替換。 圖lla中,編程和電荷存儲器件7〇〇包括编程電荷耦合 MOS電晶體225、存儲MOS電晶體735及:選擇閘極電晶體 790,該選擇閘極電晶艘79丨_^(^#酿^述的結構和 作用相同。抹除電荷耦合%、包括兩個形成在 第二深擴散井1110内的p型和H22,且該抹 除電荷耦合MOS電容器1020與抹除線控制訊號114〇相連 。抹除耦合浮動閘1123被耦合到電荷存儲浮動閘738和編 程搞合浮動閘228以形成浮動閘結丨185。 本實施方式中,抹除操作從圖9&中單層多晶矽浮動閘非 揮發性記憶體900的F〇wler_N〇rdheim通道隧道抹除操 作改變到Fowler-Nordheh邊緣隧道抹除操作,從而進 -步減少了内建在同-晶片上的執行寫(編程)操作時的 099103898 表單編號A0101 第53頁/共1〇6頁 0992007313-0 201030947 [0142] 工作電壓。 請參閱圖lib,其將對Fowler_NGrdheim邊緣隧道抹除 操作進行描述。第-淺井偏麗源VTpwi 23〇和第〆淺擴 散井215的電位被設置到大約+ 7〇v。第二淺井偏壓源 VTPW2 255被隔離或被設置到接地參考電位(〇 〇v),以 使得第二淺擴散井22G可以浮動或將第—淺擴散井215設 置到接地參考電位(uv)»第二深井偏磨源113()被隔離 或被設置到接地參考電位(〇〇V) 以使得第二深擴散井 1110可以浮動或被設置到接地參考電位(〇. 〇v)。連接到 PMOS抹除電荷耦合電容器112〇的擴散區1121和1122的Nordheim edge tunnel programming effect. Electrons enter the second shallow diffusion well 1020 from the floating gate 1 〇 85 through the drain 1 036. The threshold voltage for storing the MOS transistor 1035 is reduced from a threshold voltage of about + 3. 0V to +4·0V to a programmed threshold voltage of about + 1 after a predetermined programming time has elapsed. The advantage of the non-volatile memory 900 in the single-layer polycrystalline floating gate non-volatile memory 1 〇0 0 圏9a of the present embodiment is that only one nm 〇S capacitor (charge-coupled MOS transistor 1025) is required. Implementation built in the same crystal 099103898 Form compilation A0101 Page 52 / Total 1 page 6 page 0992007313-0 [0137] 201030947 [0138] [0139] ❹ [0140] .❹ [0141] On-chip F〇wler-Nordheiin wipe In addition to operation and F〇wlerN〇rdheim programming operations. Moreover, the physical size of the floating gate non-volatile memory 1 较 is small, so that the cost of the substrate 1025 can be reduced. A set of floating gate non-volatile memory 10's can be arranged in rows and columns similar to the array shown in Figure 8a. The necessary bias and control signals are applied to the row decode controller 810 and the bit line sense amplifier to effect the erase and program operation of the floating gate non-volatile memory. Referring to Figure 11a, the structure of another embodiment of the single-layer polycrystalline material _ volatile memory 11 本 of the present invention is substantially the same in line except for the second deep diffusion 汫W0 There are many second deep diffusion wells 1110 replaced. The second deep expansion. The deep well source of the deep well is 113G. Erase the unit replacement. In FIG. 11a, the programming and charge storage device 7A includes a programmed charge coupled MOS transistor 225, a memory MOS transistor 735, and a select gate transistor 790. The select gate is electrically crystallized 79 丨 _ ^ (^# The structure and effect are the same. The charge coupling % is erased, including two p-types and H22 formed in the second deep diffusion well 1110, and the erased charge-coupled MOS capacitor 1020 is connected to the erase line control signal 114 The erase coupling floating gate 1123 is coupled to the charge storage floating gate 738 and programmed to engage the floating gate 228 to form a floating gate 丨 185. In this embodiment, the erase operation is non-volatile from the single layer polysilicon floating gate of Figure 9 & The F〇wler_N〇rdheim tunnel tunnel erase operation of the memory 900 is changed to the Fowler-Nordheh edge tunnel erase operation, thereby further reducing the 099103898 form when performing write (program) operations built on the same-wafer No. A0101 Page 53 / Total 1 〇 6 Page 0992007313-0 201030947 [0142] Operating voltage. Please refer to Figure lib, which will describe the Fowler_NGrdheim edge tunnel erasing operation. The first - shallow well VLpwi 23〇 and Dijon Shallow expansion The potential of the well 215 is set to approximately + 7 〇 v. The second shallow well bias source VTPW2 255 is isolated or set to the ground reference potential (〇〇v) such that the second shallow diffusion well 22G can float or The shallow diffusion well 215 is set to the ground reference potential (uv)»the second deep well eccentric source 113() is isolated or set to the ground reference potential (〇〇V) such that the second deep diffusion well 1110 can float or be set to Ground reference potential (〇. 〇v) connected to the diffusion regions 1121 and 1122 of the PMOS erase charge coupled capacitor 112A
抹除線控制訊號U40的電位在灰务5·〇ν。第一 深井偏壓源265及第-深擴置到大約 + 7. 0V+位兀線電壓源BL 240、源择線偏纒源245和選擇 閘極控制訊號795都被設置到接地參考電位(〇 〇ν)。 [0143] 如上所描述的情況' ’ 1麵费鮮立在酬8抹 除電荷輕合電容器1120的珠声和舞極擴散區1121和1122 之間。PMOS抹除電魏合傷抹雜合浮動閘The potential of the line control signal U40 is erased in the ash 5·〇ν. The first deep well bias source 265 and the first deep-expanded to approximately +7.0V+bit line voltage source BL240, source select bias source 245, and select gate control signal 795 are all set to ground reference potential (〇 〇ν). [0143] The situation described above is a significant difference between the bead sound and the dance pole diffusion regions 1121 and 1122 of the charge light coupling capacitor 1120. PMOS erased electric Wei He wounded mixed floating gate
1123則被搞合到電位大約-6· 3V ^這是由於PM〇s抹除電 荷麵合電容器1120與編程電荷耦合M〇s電晶體225和存儲 M0S電晶體735兩者組合的耦合比率非常大(大約9〇%), 從而使得從第一淺擴散井215的+ 7.〇¥電壓的耦合會強迫 浮動閘結1185及抹除耦合浮動閘1123的電位在+ 6 3V。 當抹除線路控制訊號的電位是-5‘(^時,該pM〇s抹除電 荷耦合電容器1120會停留在非導電狀態。如此則會導致 Fowler-Nordheira邊緣隧道抹除作用在源極和汲極擴散 099103898 表單編號A0101 第54頁/共106頁 0992007313-0 201030947 區1121和1122的邊緣◊如同先前所描述的,這樣的電位 會防止在第一深擴散井210、第二深擴散井1110、第一淺 擴散井215和第二淺擴散井220有不受期待的前向電流。 [0144] 在Fowler-Nordheim邊緣隧道編程操作時,本實施方式 中的浮動閘非揮發性記憶體1100的電位本質上與前面所 述的Fowler-Nordheim邊緣隧道抹除操作所用的電位極 性是相反的。該第一淺井偏壓源VTPW1 230及第一淺擴 散井215的電位因而被設置到大約-7. 0V。該位元線電壓 φ 源BL 240的電位被設置到大約亨5.0V,以被施加於被選1123 is then brought to the potential of about -6·3V ^ because the coupling ratio of the PM〇s erased charge capacitor 1120 to the programmed charge coupled M〇s transistor 225 and the stored MOS transistor 735 is very large. (approximately 9〇%), such that the coupling of the + 7.〇¥ voltage from the first shallow diffusion well 215 forces the floating gate junction 1185 and the erased floating gate 1123 to have a potential of + 6 3V. When the potential of the erase line control signal is -5' (^, the pM〇s erased charge coupled capacitor 1120 will remain in a non-conducting state. This will cause the Fowler-Nordheira edge tunnel erase to act at the source and the 汲Pole diffusion 099103898 Form No. A0101 Page 54 / Total 106 pages 0992007313-0 201030947 The edges of the regions 1121 and 1122 are as previously described, such potentials are prevented in the first deep diffusion well 210, the second deep diffusion well 1110, The first shallow diffusion well 215 and the second shallow diffusion well 220 have undesired forward currents. [0144] The potential of the floating gate non-volatile memory 1100 in the present embodiment during the Fowler-Nordheim edge tunnel programming operation The potential of the potential of the first shallow well bias source VTPW1 230 and the first shallow diffusion well 215 is thus set to about -7. 0V. The potential of the bit line voltage φ source BL 240 is set to approximately +/- 5.0V to be applied to the selected
擇的浮動閘非揮發性記憶艟ιιαο,或被設置到接地參考 電位(O.OV) 憶體1100。 井215則被設置到接地參考電位(〇. 〇ν),第二淺井偏壓 源VTPf2 255被隔離或被設置到接地參考電位(〇.ον), 以使得第二淺擴散井2 2 0浮;動或賊設耳到铸地參考電位 (0. 0V:)。抹除線控制訊號W抹除電荷輕合 -r i〇fj^rh 參 電容器1020的擴散區1〇21产二從’而被設置到接地參 考電位(0· 0V)。該第一深井偏壓源265、第一深擴散井 210和第二深擴散井111〇被設置到接地參考電位(〇 〇y) 。該選擇閘極控制訊號795和偏壓源245被設置到接地參 考電位(0. 0V) » [0145]在編程操作時,該選擇閘極控制訊號795和抹除線路控制 訊號1140的電位控制在接地參考電位(〇 〇v),以關閉選 擇閘極電晶體790,從而使得存餘隱電晶體735的源極 可以浮動,以幫助Fowler-Nordheim邊緣隧道編程操作 0992007313-0 099103898 表單編號A0101 第55頁/共106頁 201030947 °當圖lib中的電壓被施加於浮動閘非揮發性記憶體11 00 時’一個低電流Fowler-Nordheim邊緣隧道現象將會發 生在存儲M0S電晶體735的汲極。如此,電子將會從浮動 閘結1185被吸收到被選擇的浮動閘非揮發性記憶體1100 的汲極736。在Fowler-Nordheim邊緣編程操作以後’ 臨界電壓將被減少到從-1. 〇V到1. 〇v的範圍。 [0146] 在讀取操作時’第一淺井偏壓源VTPW1 230被設置到一 驗證讀取偏壓(V )的電位,之後該驗證讀取偏廢Select the floating gate non-volatile memory 艟ιιαο, or set to ground reference potential (O.OV) memory 1100. The well 215 is set to the ground reference potential (〇. 〇ν), and the second shallow well bias source VTPf2 255 is isolated or set to the ground reference potential (〇.ον), so that the second shallow diffusion well 2 2 0 floats ; or the thief set the ear to the ground reference potential (0. 0V:). The erase line control signal W erases the charge light-r i〇fj^rh. The diffusion region 1 〇 21 of the capacitor 1020 is set to the ground reference potential (0·0 V). The first deep well bias source 265, the first deep diffusion well 210, and the second deep diffusion well 111A are set to a ground reference potential (〇 〇 y). The select gate control signal 795 and the bias source 245 are set to the ground reference potential (0. 0V). [0145] During the programming operation, the potentials of the select gate control signal 795 and the erase line control signal 1140 are controlled. Ground reference potential (〇〇v) to turn off select gate transistor 790 so that the source of residual crypto transistor 735 can float to aid Fowler-Nordheim edge tunnel programming operation 0992007313-0 099103898 Form No. A0101 55 Page / Total 106 pages 201030947 ° When the voltage in the diagram lib is applied to the floating gate non-volatile memory 11 00 'a low current Fowler-Nordheim edge tunneling phenomenon will occur in the drain of the storage MOS transistor 735. As such, electrons will be absorbed from the floating gate 1185 to the drain 736 of the selected floating gate non-volatile memory 1100. After the Fowler-Nordheim edge programming operation, the threshold voltage will be reduced to a range from -1. 〇V to 1. 〇v. [0146] At the time of the read operation, the first shallow well bias source VTPW1 230 is set to a potential for verifying the read bias voltage (V), after which the verification reads the waste
(V )的電位被耦合到浮動閘结1185。該驗證讀取The potential of (V) is coupled to floating gate 1185. The verification read
電位(VPotential (V
)被設置在 編程臨界 電位Vt2之間。位元線棄壓源BL 240的電位被設置到大 約1. 0V «•第二淺井偏壓源VTPW2 255、抹除控制訊號 114 0和源極線偏壓源2 4 5碑辱》置斷農弯秦考電壓源 (0. 0V)。第一深井偏壓源^第二典井偏壓源255被設 置到電源電壓源(VDD)的寶擇的浮動閘非揮發性 記憶體1100的選擇閘極控制訊號795被設置到電源電壓源 (VDD)。沒被選擇的浮動閘非揮發性記憶體11〇〇的選擇 閘極控制訊號795被設置到接地參考電位(〇 〇ν)。 [0147]) is set between the programming threshold potential Vt2. The potential of the bit line reject source BL 240 is set to approximately 1. 0V «• The second shallow well bias source VTPW2 255, the erase control signal 114 0 and the source line bias source 2 4 5 Insults Bend Qin test voltage source (0. 0V). The first deep well bias source ^the second well bias source 255 is set to the supply voltage source (VDD). The selected gate control signal 795 of the floating gate non-volatile memory 1100 is set to the power supply voltage source ( VDD). The selection of the floating gate non-volatile memory 11〇〇 that is not selected The gate control signal 795 is set to the ground reference potential (〇 〇ν). [0147]
若浮動閘非揮發性記憶體1100被編程操作,則在浮動閘 結1185的電位大約是在第二臨界電壓,並且從該位元線 電壓源BL 240到源極線偏壓源245有導電電流流經浮動 閘非揮發性記憶體。當該浮動閘非揮發性記憶體 1100的臨界電位是第二臨界電位vt2時,讀取到的數據則 099103898 表單編號A0101 第56頁/共1〇6頁 0992007313-0 201030947 [0148]If the floating gate non-volatile memory 1100 is programmed to operate, the potential at the floating gate junction 1185 is approximately at a second threshold voltage, and there is a conduction current from the bit line voltage source BL 240 to the source line bias source 245. Flow through the floating gate non-volatile memory. When the critical potential of the floating gate non-volatile memory 1100 is the second critical potential vt2, the read data is 099103898 Form No. A0101 Page 56 / Total 1 Page 6 0992007313-0 201030947 [0148]
[0149][0149]
被指定作為二進制的「〇」。若該浮動閘非揮發性記憶體 1100沒有被選擇作編程操作,則它的臨界電壓依然是在 它的抹除臨界電位Vtl。故該浮動閘結1185的電位比抹除 臨界電位V11小❶當該浮動閘非揮發性記憶體1丨〇 〇沒有導 電電流,並且浮動閘非揮發性記:It體1100的臨界電壓是 抹除臨界電壓時,讀取到的數據則被指定作為二進制的 「1」。 與其他實施方式相似,該編程電荷耦合MOS電晶體225的 物理尺寸大於存儲M〇s電晶體735和編程電荷耦合m〇s電 晶體225兩者組合在一起的物理尺寸。該編,程電荷輕合 MOS電晶體225的物理尺寸g 储M0s電晶 艘735和編程電荷耦的總面積 以達到超過90%的耦合比率,從而可以減少必需的電位 使得一般低電壓的邏輯積體電路製程亦可用來製造包括 浮動閘非揮發性記憶艘11 〇〇的一_产揮發存_器件。 一組浮動閘非揮發性記憶_安排成行和列與 在圖8a顯示的陣列相似。和控制訊號加到行 解碼控制器810和位元線感應放大器815,以對本實施方 式所描述的浮動閘非揮發性記憶體1〇〇〇進行抹除和編程 操作。 [0150] —單層多晶矽浮動閘非揮發性記憶體包括一MOS電容器和 一MOS電晶體,它的製造尺寸允許使用當前低電壓邏輯積 體電路製程。該MOS電容器的第一電極板連接到MOS電晶 體的閘極,如此MOS電晶體的閘極可以浮動並且形成浮動 閘非揮發性記憶體的一浮動閘結。該MOS電容器的第二電 099103898 表單編號A0101 第57頁/共106頁 0992007313-0 201030947 極板通常是MOS電晶體的汲_極擴散、源極擴散和大塊基極 (bulk)。該M0S電晶體的汲極連接到一位元線電壓源, 並且該M0S電晶體的源極連接到源極線。該m〇S電容器的 物理尺寸與M0S電晶體的物理尺寸比較是相對地大。本實 施方式中’該M0S電容器的物理尺寸比M0S電晶體的物理 尺寸大過10倍或更大。 [0151] [0152]It is designated as a binary "〇". If the floating gate non-volatile memory 1100 is not selected for programming operation, its threshold voltage is still at its erase critical potential Vtl. Therefore, the potential of the floating gate 1185 is smaller than the erase critical potential V11. When the floating gate non-volatile memory 1丨〇〇 has no conduction current, and the floating gate is non-volatile, the threshold voltage of the It body 1100 is erased. At the threshold voltage, the read data is designated as a binary "1". Similar to other embodiments, the physical size of the programmed charge coupled MOS transistor 225 is greater than the physical size of the combination of the memory M?s transistor 735 and the programmed charge coupled m?s transistor 225. The program, the physical size of the process charge MOS transistor 225, stores the total area of the M0s cell 735 and the programmed charge coupler to achieve a coupling ratio of more than 90%, thereby reducing the necessary potential so that the logical product of the general low voltage The bulk circuit process can also be used to fabricate a volatile device that includes a floating gate non-volatile memory bank. A set of floating gate non-volatile memories _ arranged in rows and columns is similar to the array shown in Figure 8a. And control signals are applied to row decode controller 810 and bit line sense amplifier 815 to erase and program the floating gate non-volatile memory 1 described in this embodiment. [0150] A single layer polysilicon floating gate non-volatile memory includes a MOS capacitor and a MOS transistor, the fabrication dimensions of which allow the use of current low voltage logic integrated circuit processes. The first electrode plate of the MOS capacitor is connected to the gate of the MOS transistor such that the gate of the MOS transistor can float and form a floating gate of the floating gate non-volatile memory. The second electrode of the MOS capacitor 099103898 Form No. A0101 Page 57 of 106 0992007313-0 201030947 The plate is usually the 汲_pole diffusion, source diffusion and bulk of the MOS transistor. The drain of the MOS transistor is connected to a one-bit line voltage source, and the source of the MOS transistor is connected to the source line. The physical size of the m〇S capacitor is relatively large compared to the physical size of the MOS transistor. In the present embodiment, the physical size of the MOS capacitor is more than 10 times or more larger than the physical size of the MOS transistor. [0152]
該M0S電容器和M0S電晶體之間的物理尺寸比率提供了 > 個大於80%的大耦合比率。當電壓施加於M0S電容器的第 二電極板時,大耦合比率能使大部分在M0S電容器第二電 極板的電壓耦合到浮動閛結*電麇被施加於M〇s電晶體的The physical size ratio between the MOS capacitor and the MOS transistor provides > a large coupling ratio of greater than 80%. When a voltage is applied to the second electrode plate of the MOS capacitor, the large coupling ratio enables most of the voltage at the second electrode plate of the MOS capacitor to be coupled to the floating 閛 junction * 麇 is applied to the M 〇 s transistor
汲極或源極,,以在M0S電晶内建立一個 電場’如此F〇wler-N〇rdh^^|i^^能被啟動 。當在第二電極板的電壓是負極,而被施加於汲極(或源 極)的M0S電晶體的電壓是正極時,在浮動閘的電荷則被 吸取去編程浮動閘非揮發馋記憶夢。另,苹第二電極板 的電歷是正極’而被施加於大塊基極 (bulk)的M0S電晶髏.電壓^浮動閘的電荷則被 注射去抹除浮動閘非揮發性記憶“乞 綜上所述,本發明符合發明專利要件,爰依法提出專利 申請。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,皆應涵蓋於以下之申請專利範圍内。 【圖式簡單說明】 圖1是習知的單層多晶梦浮動閘非揮發性記憶體的示意圈 099103898 表單編號A0101 第58頁/共106頁 0992007313-0 [0153] 201030947 [_ ®2a是本發明單層多晶料動閘非揮發性記憶體的第一 較佳實施方式的橫切面圖。 剛圖2b是圖2a中單層多晶料動_揮發性記龍的電麼 源的電位的示意圖。 _]則是本發明單衫晶料動_揮發性記㈣的第二較 佳實施方式的橫切面圖》 [_ ®4a是本剌單層多晶料動閘非揮發纽憶體的第 參 較佳實施方式的橫切面圖。 [0158] [0159] 圖4b是圖4a中單層多晶料動閉非都發性記憶體的電壓 源的電位的示意圖。 .二::,響卜':_.,Λ.P t =^v',s, 圖5是本發明單層多晶矽浮_ 佳實施方式的橫切面The drain or source, to establish an electric field in the M0S transistor, so that F〇wler-N〇rdh^^|i^^ can be activated. When the voltage at the second electrode plate is the negative electrode and the voltage of the MOS transistor applied to the drain (or source) is the positive electrode, the charge at the floating gate is absorbed to program the floating gate non-volatile memory dream. In addition, the electrical history of the second electrode plate is the positive electrode's and is applied to the bulk of the MOS transistor. The charge of the voltage ^ floating gate is injected to erase the floating gate non-volatile memory. In summary, the present invention meets the requirements of the invention patent, and the patent application is filed according to law. However, the above is only a preferred embodiment of the present invention, and those who are familiar with the art of the present invention are in accordance with the spirit of the present invention. Modifications or changes should be covered in the following patent application. [Simplified Schematic] Figure 1 is a schematic representation of a conventional single-layer polycrystalline dream floating gate non-volatile memory 099103898 Form No. A0101 Page 58 / Total 106 pages 0992007313-0 [0153] 201030947 [_2a is a cross-sectional view of the first preferred embodiment of the single-layer polycrystalline material gate non-volatile memory of the present invention. Figure 2b is a single Figure 2a Schematic diagram of the potential of the layered polycrystalline material _ volatile kelong's electric source. _] is the cross-sectional view of the second preferred embodiment of the single-crystal kinetic volatility (four) of the present invention. [_® 4a is the first reference of the non-volatile neon memory of the single-layer polycrystalline material A cross-sectional view of an embodiment. [0159] FIG. 4b is a schematic diagram of the potential of a voltage source of a single-layer polycrystalline material in a non-uniform memory in FIG. 4a. 2::, 响卜':_ , Λ.P t =^v', s, Figure 5 is a cross-section of a single-layer polycrystalline raft of the present invention
發植:纪號體的第四較 [0160] ❹ [0161] 圖6a是圖2a中單’多晶發,動閉非揮發性記憶想的示意'0 °· ,.Ι»Γ.Ι::Ϊ *ri%!iecluai _是圏3.中單記憶體的示意圖 _],是圖2a中單層多晶料動閘非揮發性記憶體的陣列 不意圖。 剛圖7a是本發明單層Η料動閘轉發性記憶趙的第五 較佳實施方式的橫切面圖。 _圖71)是圖7a中單層多晶矽浮動閉非揮發性記憶體的電壓 源的電位的示意囷。 099103898 表單編號A0101 第59頁/共1〇6頁 0992007313-0 201030947 [0165] [0166] [0167] [0168] [0169] [0170] [0171] [0172] [0173] [0174] [0175] [0176] 099103898 圖7c是圖7a中單層多晶矽浮動閘非揮發性記憶體的示意 圖。 圖8a是圖7a中單層多晶矽浮動閘非揮發性記憶體陣列示 意圖。 圖8b是圖7a中單層多晶矽浮動閘非揮發性記憶體的第六 較佳實施方式的示意圖。 圖9a是本發明單層多晶矽浮動閘非揮發性記憶體的第七 較佳實施方式的橫切面圖。 圖9b是圖9a中單層多晶矽浮動閘非揮發性記憶體的電壓 源的電位的不意圖。 圖10 a是本發明單層多晶矽浮動閘非揮發性記憶體的第八 較佳實施方式的橫切面圖。 圖10b是圖10a中單層多晶矽浮動閘非揮發性記憶體的電 壓源的電位的示意圖。 圖11a是本發明單層多晶矽浮動閘非揮發性記憶體的第九 @ 較佳實施方式的橫切面圖。 圖lib是圖11a中單層多晶矽浮動閘非揮發性記憶體的電 壓源的電位的示意圖。 【主要元件符號說明】 NVM單元:C100 第一編程電晶體:M106 第二編程電晶體.Μ10 4 表單編號Α0101 第60頁/共106頁 0992007313-0 201030947 [οΐπ] 存儲電晶體:Μ102 [0178] 預先充電電路:108 [0179] 電源電壓源:VDD [0180] 浮動結:110 [0181] 第一編程電壓源:VEE [0182] 第二編程電壓源:VPP [0183] 參 [0184] 字元線電壓源:WL 輸出位元線:Vo(BL) [0185] 浮動閘非揮發性記憶體:200、300、400、500、600、 700 、 800 、 900 、 1000 、 1100 [0186] 基板:205、405、1005 [0187] 第一深擴散井:210、410、101· f [0188] 第一深井偏壓源:265、465、10加 .. # 議 1 [0189] 連接擴散點:270、470、411^416、250、1070、1021 [0190] 第一淺擴散井:215、1015 [0191] 第二淺擴散井:220、1020 [0192] 編程電荷耦合M0S電晶體:225、425 [0193] 電荷耦合M0S電容器:325 [0194] 源極:226、237、426、437、737、1026、1037、 1093 099103898 表單編號Α0101 第61頁/共106頁 0992007313-0 201030947 [0195] 汲極:227、236、427、436、736、1027、1036 [0196] 第一淺井偏壓源:VTPW1 230、VTPW1[0] 645a,…, VTPW1[1] 645b 、 VTPW1 1050 [0197] 閘極氧匕區:229、239、429、439、739、794、924、 1029 [0198] 存儲MOS電晶體:235、435、735、1035 [0199] 源極線偏壓源:245、445、655、855 [0200] 電荷存儲浮動閘:238、438、738、1038 © [0201] 第二淺井偏壓源:VTPff2 255、VTPW2 [fl] 650a,…, VTPW2 [1] 650b ' VTPW^t J: 850a » - * VTPW2[ 1 ] 850b、VTPW2 1055 ’ [0202] 通道區域:280、275、975 [0203] 編程耦合浮動閘:228、3得、08*、爭28 : 11 11J : 1 ^ * i [0204] 浮動閘結:285、385、48^『龟,5、ϊ姑、985、1085、 1185 Office ® [0205] 位元線電壓源:BL 240 、BL 440 [0206] 淺擴散井:415 [0207] 淺井偏壓源:VTPW1 430 [0208] 電荷耦合M0S電容器:525 [0209] 陣列:605、805 [0210] 位元線:BL[0] 640a,BL[1] 640b,…,BL[n] 640η 099103898 表單編號A0101 第62頁/共106頁 0992007313-0 201030947 ’ BL[η] 840η [0211] [0212] [0213] [0214] [0215] [0216] 參 [0217] [0218] [0219] [0220] [0221] w [0222] [0223] [0224] [0225] [0226] [0227] [0228] 、BL[0] 840a,BL[1] 840b,… 位元線感應放大器:615、815 數據输出口 : 635、835 行解碼控制器:610、810 地址訊號:620、820 控制訊号:630、830 選擇閘極電晶體:790 選擇閘極:792、1092 選擇閘極控制訊號:795、SG[0] 8tt0a,… 860b、1095 字元線: WL[0] 845a,845b 數據訊號:825 y .會參I s _ ' i § p s _ 養 s ss Is g s 共同字元線CWL : 845 ;· 抹除元件:905 . 第三淺擴散井:915 第二深擴散井:910、1014、1110 抹除電荷耦合MOS電晶體:920 N型源極擴散:921 N型汲極擴散:922 抹除控制線路:EL 950 ? SG[1] 099103898 表單編號A0101 第63頁/共106頁 0992007313-0 201030947 [0229] 抹除耦合浮動閘:923、1123 [0230] 第三淺井偏壓源:VTPW3 940 [0231] N型連接點:1060、 1062 [0232] 電荷耦合M0S電晶體 :1025 [0233] 耦合浮動閘:1028 [0234] 位元線電壓源:BL 1040 [0235] 閘極電晶體:1090 [0236] 源極線電壓源:SL 1045 [0237] 抹除單元:1105 [0238] P型擴散區:1121、 1122 [0239] 抹除線控制訊號:1140 [0240] PM0S抹除電荷耦合電容器:1120: 099103898 表單編號A0101 第64頁/共106頁 ❿ 0992007313-0Hair transplant: the fourth comparison of the corpus body [0160] ❹ [0161] Fig. 6a is a schematic diagram of the single 'polycrystalline hair in Fig. 2a, moving non-volatile memory. '0 °·,.Ι»Γ.Ι: :Ϊ *ri%!iecluai _ is a schematic diagram of 中3. Medium memory _], which is an array of single-layer polycrystalline gate non-volatile memory in Figure 2a. Figure 7a is a cross-sectional view of a fifth preferred embodiment of the single layer smashing transfer memory of the present invention. _ Figure 71) is a schematic representation of the potential of the voltage source of the single-layer polysilicon floating closed non-volatile memory of Figure 7a. 099103898 Form No. A0101 Page 59 / Total 1 Page 6 0992007313-0 201030947 [0166] [0168] [0170] [0170] [0175] [0175] [0176] 099103898 FIG. 7c is a schematic diagram of the single-layer polysilicon floating gate non-volatile memory of FIG. 7a. Figure 8a is a schematic illustration of the single layer polysilicon floating gate non-volatile memory array of Figure 7a. Figure 8b is a schematic illustration of a sixth preferred embodiment of the single layer polysilicon floating gate non-volatile memory of Figure 7a. Figure 9a is a cross-sectional view showing a seventh preferred embodiment of the single-layer polysilicon floating gate non-volatile memory of the present invention. Figure 9b is a schematic illustration of the potential of the voltage source of the single-layer polysilicon floating gate non-volatile memory of Figure 9a. Figure 10a is a cross-sectional view of an eighth preferred embodiment of the single layer polysilicon floating gate non-volatile memory of the present invention. Figure 10b is a schematic illustration of the potential of a voltage source for a single layer polysilicon floating gate non-volatile memory of Figure 10a. Figure 11a is a cross-sectional view of a ninth preferred embodiment of a single layer polysilicon floating gate non-volatile memory of the present invention. Figure lib is a schematic illustration of the potential of a voltage source for a single layer polysilicon floating gate non-volatile memory in Figure 11a. [Main component symbol description] NVM unit: C100 First programming transistor: M106 Second programming transistor. Μ10 4 Form number Α0101 Page 60/Total 106 page 0992007313-0 201030947 [οΐπ] Storage transistor: Μ102 [0178] Precharge circuit: 108 [0179] Power supply voltage source: VDD [0180] Floating junction: 110 [0181] First programming voltage source: VEE [0182] Second programming voltage source: VPP [0183] Reference [0184] Word line Voltage source: WL output bit line: Vo(BL) [0185] Floating gate non-volatile memory: 200, 300, 400, 500, 600, 700, 800, 900, 1000, 1100 [0186] Substrate: 205, 405, 1005 [0187] First deep diffusion well: 210, 410, 101 · f [0188] First deep well bias source: 265, 465, 10 plus.. # 1 [0189] Connection diffusion point: 270, 470 411^416, 250, 1070, 1021 [0190] First shallow diffusion well: 215, 1015 [0191] Second shallow diffusion well: 220, 1020 [0192] Programming charge coupled MOS transistor: 225, 425 [0193] Charge-Coupled MOS Capacitor: 325 [0194] Source: 226, 237, 426, 437, 737, 1026, 1037, 1093 099103898 Form No. 0101 Page 61 / 106 pages 0992007313-0 201030947 [0195] Bungee: 227, 236, 427, 436, 736, 1027, 1036 [0196] First shallow well bias source: VTPW1 230, VTPW1[0] 645a ,..., VTPW1[1] 645b , VTPW1 1050 [0197] Gate oxime zone: 229, 239, 429, 439, 739, 794, 924, 1029 [0198] Storage MOS transistor: 235, 435, 735, 1035 [0199] Source Line Bias Source: 245, 445, 655, 855 [0200] Charge Storage Floating Gate: 238, 438, 738, 1038 © [0201] Second shallow well bias source: VTPff2 255, VTPW2 [fl] 650a,..., VTPW2 [1] 650b ' VTPW^t J: 850a » - * VTPW2[ 1 ] 850b, VTPW2 1055 ' [0202] Channel area: 280, 275, 975 [0203] Programmable coupling floating gate: 228, 3得得, 08*, 争28 : 11 11J : 1 ^ * i [0204] Floating gate: 285, 385, 48^ "Turtle, 5, Aunt, 985, 1085, 1185 Office ® [0205] Bit line voltage Source: BL 240, BL 440 [0206] Shallow diffusion well: 415 [0207] Shallow well bias source: VTPW1 430 [0208] Charge coupled M0S capacitor: 525 [0209] Array: 605, 805 [0210] Bit line: BL [0] 640a, BL [1] 640b,...,BL[n] 640η 099103898 Form No. A0101 Page 62/Total 106 Page 0992007313-0 201030947 'BL[η] 840η [0211] [0212] [0214] [0215] [0216 ] [0217] [0218] [0220] [0222] [0223] [0228] [0228] [BL], BL[0] 840a, BL[1] 840b,... Bit line sense amplifier: 615, 815 data output port: 635, 835 line decoding controller: 610, 810 address signal: 620, 820 control signal: 630, 830 select gate transistor: 790 select gate :792,1092 Select gate control signal: 795, SG[0] 8tt0a,... 860b, 1095 character line: WL[0] 845a, 845b data signal: 825 y. I will refer to I s _ ' i § ps _ s ss Is gs Common word line CWL : 845 ;· Wipe component: 905 . Third shallow diffusion well: 915 Second deep diffusion well: 910, 1014, 1110 Erasing charge coupled MOS transistor: 920 N-type source Diffusion: 921 N-type bungee diffusion: 922 erase control line: EL 950 ? SG[1] 099103898 Form No. A0101 Page 63 / Total 106 Page 0992007313-0 201030947 [0229] Erasing the coupled floating gate: 923, 1123 [ 0230] section Shallow well bias source: VTPW3 940 [0231] N-type connection point: 1060, 1062 [0232] Charge-coupled M0S transistor: 1025 [0233] Coupling floating gate: 1028 [0234] Bit line voltage source: BL 1040 [0235] Gate Transistor: 1090 [0236] Source Line Voltage Source: SL 1045 [0237] Erase Unit: 1105 [0238] P Type Diffusion Area: 1121, 1122 [0239] Erase Line Control Signal: 1140 [0240] PM0S Wipe the charge-coupling capacitor: 1120: 099103898 Form number A0101 Page 64 of 106 ❿ 0992007313-0
Claims (1)
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US12/378,036 US8472251B2 (en) | 2008-02-11 | 2009-02-10 | Single-polycrystalline silicon electrically erasable and programmable nonvolatile memory device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| TW201030947A true TW201030947A (en) | 2010-08-16 |
Family
ID=44860604
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW99103898A TW201030947A (en) | 2009-02-10 | 2010-02-09 | Single-polycrystalline silicon electrically erasable and programmable nonvolatile memory device |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TW201030947A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI645543B (en) * | 2014-04-18 | 2018-12-21 | 日商芙洛提亞股份有限公司 | Non-volatile semiconductor memory device |
| CN109473430A (en) * | 2018-11-30 | 2019-03-15 | 江苏集萃微纳自动化系统与装备技术研究所有限公司 | Single-layer polycrystalline EEPROM based on standard CMOS process |
| TWI810528B (en) * | 2020-02-26 | 2023-08-01 | 台灣積體電路製造股份有限公司 | Semiconductor device and fabrication method thereof |
| TWI842474B (en) * | 2023-04-12 | 2024-05-11 | 國立清華大學 | HETEROGENEOUS INTEGRATION CAPACITOR AND MoM CAPACITOR |
-
2010
- 2010-02-09 TW TW99103898A patent/TW201030947A/en unknown
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI645543B (en) * | 2014-04-18 | 2018-12-21 | 日商芙洛提亞股份有限公司 | Non-volatile semiconductor memory device |
| CN109473430A (en) * | 2018-11-30 | 2019-03-15 | 江苏集萃微纳自动化系统与装备技术研究所有限公司 | Single-layer polycrystalline EEPROM based on standard CMOS process |
| TWI810528B (en) * | 2020-02-26 | 2023-08-01 | 台灣積體電路製造股份有限公司 | Semiconductor device and fabrication method thereof |
| US11715781B2 (en) | 2020-02-26 | 2023-08-01 | Taiwan Semiconductor Manufacturing Co., Ltd. | Semiconductor devices with improved capacitors |
| US12148811B2 (en) | 2020-02-26 | 2024-11-19 | Taiwan Semiconductor Manufacturing Co., Ltd. | Method of fabricating a semiconductor device having capacitor material |
| TWI842474B (en) * | 2023-04-12 | 2024-05-11 | 國立清華大學 | HETEROGENEOUS INTEGRATION CAPACITOR AND MoM CAPACITOR |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8472251B2 (en) | Single-polycrystalline silicon electrically erasable and programmable nonvolatile memory device | |
| TWI394266B (en) | Complementary MOS semiconductor erasable stylized read-only memory and electronic erasable stylized read-only memory device and programmable complementary CMOS inverter | |
| TW475267B (en) | Semiconductor memory | |
| TWI342615B (en) | A multiple time programmable (mtp) memory cell and a method for operating the same | |
| TWI736763B (en) | Memory circuits, methods of operating a memory circuit and memory arrays | |
| US20030235082A1 (en) | Single-poly EEPROM | |
| USRE50512E1 (en) | Non-volatile semiconductor storage device | |
| TW200810095A (en) | Nonvolatile semiconductor memory and activation method thereof | |
| CN105390154B (en) | Page or Word Erasable Composite Nonvolatile Memory | |
| JP2009540545A (en) | Nonvolatile memory embedded in a conventional logic process and method of operating such a nonvolatile memory | |
| CN1866544B (en) | non-volatile semiconductor storage device | |
| JP2005501403A (en) | Nonvolatile semiconductor memory and method of operating the same | |
| TW200926185A (en) | Nonvolatile memory devices and methods of operating same to inhibit parasitic charge accumulation therein | |
| CN103515393B (en) | Nonvolatile memory device with single polysilicon layer memory cell | |
| JPH1032269A (en) | Semiconductor device | |
| CN108028061A (en) | Asymmetric Transfer Field-Effect Transistor for Nonvolatile Memory | |
| JP2008234821A (en) | Nonvolatile semiconductor memory device | |
| JP2009267185A (en) | Non-volatile semiconductor memory device | |
| US7612397B2 (en) | Memory cell having first and second capacitors with electrodes acting as control gates for nonvolatile memory transistors | |
| TW201138071A (en) | A universal dual charge-retaining transistor flash NOR cell, a dual charge-retaining transistor flash NOR cell array, and method for operating same | |
| CN107093456A (en) | Single-layer polysilicon non-volatile memory cell | |
| WO2002037502A9 (en) | Common source eeprom and flash memory | |
| TW201030947A (en) | Single-polycrystalline silicon electrically erasable and programmable nonvolatile memory device | |
| JP2001067885A (en) | Method for erasing non-volatile memory cell of field- programmable gate array | |
| US7164606B1 (en) | Reverse fowler-nordheim tunneling programming for non-volatile memory cell |