TW200903499A - Non-volatile storage with boosting using channel isolation switching and method thereof - Google Patents

Non-volatile storage with boosting using channel isolation switching and method thereof Download PDF

Info

Publication number
TW200903499A
TW200903499A TW097116862A TW97116862A TW200903499A TW 200903499 A TW200903499 A TW 200903499A TW 097116862 A TW097116862 A TW 097116862A TW 97116862 A TW97116862 A TW 97116862A TW 200903499 A TW200903499 A TW 200903499A
Authority
TW
Taiwan
Prior art keywords
word line
voltage
volatile storage
line
storage element
Prior art date
Application number
TW097116862A
Other languages
Chinese (zh)
Other versions
TWI386944B (en
Inventor
Ying-Da Dong
Jeffrey W Lutze
Shih-Chung Lee
Gerrit Jan Hemink
Ken Oowada
Original Assignee
Sandisk Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US11/745,082 external-priority patent/US7460404B1/en
Priority claimed from US11/745,092 external-priority patent/US7463522B2/en
Application filed by Sandisk Corp filed Critical Sandisk Corp
Publication of TW200903499A publication Critical patent/TW200903499A/en
Application granted granted Critical
Publication of TWI386944B publication Critical patent/TWI386944B/en

Links

Classifications

    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/34Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
    • G11C16/3418Disturbance prevention or evaluation; Refreshing of disturbed memory data
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/34Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/08Address circuits; Decoders; Word-line control circuits
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/30Power supply circuits

Landscapes

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

Abstract

Program disturb is reduced in non-volatile storage by preventing source side boosting in selected NAND strings. A self-boosting mode which includes an isolation word line is used. A channel area of an inhibited NAND string is boosted on a source side of the isolation word line before the channel is boosted on a drain side of the isolation word line. Further, storage elements near the isolation word line are kept in a conducting state during the source side boosting so that the source side channel is connected to the drain side channel. In this way, in selected NAND strings, source side boosting can not occur and thus program disturb due to source side boosting can be prevented. After the source side boosting, the source side channel is isolated from the drain side channel, and drain side boosting is performed.

Description

200903499 九、發明說明: 【發明所屬之技術領域】 本發明係關於非揮發性記憶體。 本申請案係關於以標題”Ncm-Volatile Storage With200903499 IX. Description of the Invention: TECHNICAL FIELD OF THE INVENTION The present invention relates to non-volatile memory. This application is related to the title "Ncm-Volatile Storage With

Boosting Using Channel Isolation Switching”提出申請的同 在申凊中之共同讓渡的美國專利申請案第__號(檔案號 SAND-1229US1),其以引用之方式併入本文中。 【先前技術】Boosting Using Channel Isolation Switching, U.S. Patent Application Serial No. __ (file No. SAND-1229US1), which is incorporated herein by reference.

半導體記憶體已愈加風行地用於各種電子裝置中。舉例 而言,非揮發性半導體記憶體用於蜂巢式電話、數位相 非行動計算裝置及其 機 '個人數位助理、行動計算裝置 他裝置中。在最為風行的非揮發性半導體記憶體當中有電 子可抹除可程式化唯讀記憶體(EEPR〇M)及快閃記憶體。 與傳統之全特徵化EEPR0M對比,在快閃記憶體(亦為 EEPROM類型)的情況τ,可單步地抹除整個記憶體陣列之 内容或記憶體之一部分的内容。 傳統EEPROM及快閃記憶體均利用浮動閉極,浮動閑極 定位於半導體基板中之通道區上方且與通道區絕緣。:動 閘極定位於源極區與汲極區之間。控制閘極提供於浮 極上方且與浮動開極絕緣。經如此形成之電晶體的臨限^ 壓(VTH)由保留於浮動閘極上之電荷量控制。亦即 啟電晶體以准許其源極與汲極之間傳導之前必須施加至= 制閘極之最小電壓量由浮動閘極上之電荷㈣_。工 —些EEPR〇M及快閃記憶體裝置具有用以儲存兩個電荷 131225.doc 200903499 子動閉極’且因此’可在兩個狀態(例如,經抹除 狀態與經程式化狀態)之間程式化/抹除記憶體元件。此快 閃記憶體裝置有時被稱為二進位快閃記憶體裝置 一、 一記憶體元件可儲存一資料位元。 ’、’、母 多狀態(亦被稱作多位準)快閃記憶體裝置係藉 個相異容許/有效經程式化臨限電壓範圍來實施。^一少 異臨限電壓範圍對應於記憶體裝置中所編碼之=目 合的預定值。舉例而言,當每一記憶體元件可置放在:應 ;四個相異臨限電壓範圍之四個離散電荷帶中—由、 時’元件可儲存兩個資料位元。 、 通吊將在程式操作期間施力口至控亲,】閑極之程式電厣 vPGM施加為量值隨時間而增加之—連串脈衝。在―;能二 法中’將脈衝之量值隨每一連續脈衝而增加預定步長月,匕例 ’ 0·2-〇·4 V。VpGM可施加至快閃記憶體元件之控制閘 圣°在程式脈衝之間的時期中,進行驗證操作。亦即 連續程式化脈衝之㈣取經並行地程式化之元件群之每一 元件的程式化位準,以判定其是等於還是大於元件經程式 化至之驗證位準。對於多狀態快閃記憶體S件陣列而 可針對元件之每-狀態來執行驗證步驟,以判定元件是否 =達到其資料關聯驗證位準。舉例而言,能夠在四個狀離 =存資料之多狀態記憶體元件可能需要針對三個比較點 來執行驗證操作。 此外’當程式化EEPROM或快閃記憶體裝置(諸如, N伽串中之财肋快閃記憶體裝置)時,通常將v_施加 131225.doc 200903499 至控制閘極且使位7L線接地,從而使得將來自單元或記憶 體元件(例如,儲存元件)之通道的電子注入至浮動閘極 中。當電子累積於浮動閘極中時,浮動閉極變得帶負電, 且記憶體元件之臨限電壓升高,使得認為記憶體元件處於 經程式化狀態中。可在標題為"source Side Self BoosiingSemiconductor memory has become increasingly popular in a variety of electronic devices. For example, non-volatile semiconductor memory is used in cellular telephones, digital phase non-mobile computing devices, and their personal personal assistants, mobile computing devices, and other devices. Among the most popular non-volatile semiconductor memories are electronically erasable programmable read-only memory (EEPR〇M) and flash memory. In contrast to the conventional fully characterized EEPR0M, in the case of flash memory (also of the EEPROM type), the content of the entire memory array or the contents of one part of the memory can be erased in a single step. Both conventional EEPROM and flash memory utilize a floating closed-pole that is positioned above the channel region in the semiconductor substrate and insulated from the channel region. The movable gate is positioned between the source region and the drain region. The control gate is provided above the float and insulated from the floating open. The threshold voltage (VTH) of the thus formed transistor is controlled by the amount of charge remaining on the floating gate. That is, the minimum amount of voltage that must be applied to the gate before the transistor is allowed to conduct between its source and drain is the charge on the floating gate (4)_. Some EEPR〇M and flash memory devices have two states for storing two charges 131225.doc 200903499, and thus can be in two states (eg, erased state and stylized state) Stylize/erase memory components. This flash memory device is sometimes referred to as a binary flash memory device. A memory device can store a data bit. The ',' and parent multi-state (also referred to as multi-level) flash memory devices are implemented by a different allowable/effective programmed threshold voltage range. ^ A small threshold voltage range corresponds to a predetermined value of the coded = in the memory device. For example, when each memory component can be placed in: four discrete charge bands of four distinct threshold voltage ranges, the two-bit elements can store two data bits. The crane will apply force to the control during the operation of the program.] The program of the idle pole vPGM is applied as a series of pulses that increase with time. In the "energy method", the magnitude of the pulse is increased by a predetermined step and length with each successive pulse, for example, '0·2-〇·4 V. VpGM can be applied to the control gate of the flash memory device. The verification operation is performed during the period between the program pulses. That is, the continuous stylized pulse (4) takes the stylized level of each component of the parallel-programmed component group to determine whether it is equal to or greater than the verify level to which the component is programmed. For a multi-state flash memory S-piece array, a verification step can be performed for each state of the component to determine if the component = reaches its data association verification level. For example, a multi-state memory component capable of four-segment data storage may need to perform verification operations for three comparison points. In addition, when stylized EEPROM or flash memory devices (such as the gambling flash memory device in N gamma series), v_ is usually applied to 131225.doc 200903499 to the control gate and the bit 7L line is grounded. Thereby electrons from channels of cells or memory elements (eg, storage elements) are injected into the floating gate. When electrons accumulate in the floating gate, the floating closed-pole becomes negatively charged, and the threshold voltage of the memory element rises, making the memory element considered to be in a stylized state. Available under the heading "source Side Self Boosiing

Technique For Non-Volatile Memory"之美國專利 6,859,397 及005年2月3日么布的標題為”⑽⑼以叩ο· pr〇grammedTechnique For Non-Volatile Memory" US Patent 6,859,397 and February 3, 005, titled "(10)(9) by 叩ο· pr〇grammed

Memory之美國專利申請案公開案中找到關 於此程式化之更多資訊;該等案則丨用之方式全文併入本 文中。 個問題為程式干擾。可在其他 然而,繼續成問題的 NAND串之耘式化期間在受抑制nand串處發生程式干 擾’且有時在經程式化贴動串自身處發生程式干擾。當 未選非揮發性儲存元件之臨限電壓歸因於其他非揮發性儲 存元件之程式化而移位時,發生程式干擾。可在先前經程 式化儲存7C件以及尚未程式化之經抹除儲存元件上 式干擾。 柱 【發明内容】 本發明藉由提供—種用於減少非揮發性儲存器中 干擾的方法來解決上述及其他問題。 工 實包Ή中,一種用於操作非揮發性儲存器之 == 線之Μ側上提升至少-_串之前在第: 一〜’I ^上執仃至少一 NAND串之第—提升,其中第 子線係在第—字線之沒極側上。包括第—字線及第二字 131225.doc 200903499 多字線與至少-函㈣相關聯,且至少一_d 二=多非揮發性儲存元件。方法進一步包括:在第一 中盘第一將電壓施加至第一字線以用於提供在傳導狀態 ::、二予線相關聯之第一非揮發性儲存元件;及將電壓 :加^第二字線以用於提供在傳導狀態中㈣二字線相關 聯之第—非揮發性儲存元件。 之# Λ楚^ 運步包括在第一提升 第二子線之没極側上執行至少-nand串之第二提 二之:時:電壓施加至第一字線以用於提供在非傳導狀態 中非揮發性儲存元件,且同時將程式電壓施加至第 一子線。因此,在施加程式脈衝之前發生源極侧提升。 實施财,—種用於操作非揮發性儲存器之方法 广至少_NAND串中之第一非揮發性儲 Π:,娜串之第-提升,其在程式化序列中係在 揮發性儲存元件之前。方法進一步包括在第-提升 二=:少一 之第一非揮發性儲存元件及在 純儲存元件之側上的第二非揮發性儲存元件, 式:序列中係在傳導嶋之第-非揮發性儲存元 I#法進一步包括在第一提升之後在第二非揮發性 储存凡件之側[•拙s , 式化序列中係在第:非串之第二提升’其在程 _發性存7L件之後,同時提供在 傳導狀恕中之第—儲存元件。 在另一實施例尹, 包括(a)在第—時段中 字線之源極側上的第 一種用於操作非揮發性儲存器之方法 ·’(i)將電壓施加至字線集合令之特定 一字線集合以用於提升至少—nand I3I225.doc •10· 200903499 串之第一通道區;(ii)將電壓施加至包括特定字線之第 字線集合以提供至少-NAND串中與第:字線集合相關聯 之在傳導狀態中的非揮發性儲存元件,第二字線集合係在 第一字線集合之汲極側上;及(iii)將電壓施加至第二字線 集合之汲極側上的第三字線集合以避免提升至少一 nand 串之第二通道區。方法進一步包括(b)在跟著第—時段之第 二時段中:⑴將電壓施加至第三字線集合以用於提升至少 一 NAND串之第二通道區;(ii)將程式電壓施加至第二字線More information on this stylization can be found in the US Patent Application Publications of Memory; the manner in which these cases are applied is incorporated in this article in its entirety. The problem is program disturb. However, program disturb can occur at the suppressed nand string during the NAND string that continues to be problematic, and sometimes program disturb occurs at the programmed post itself. Program disturb occurs when the threshold voltage of the unselected non-volatile storage element is shifted due to the stylization of other non-volatile storage elements. The 7C pieces can be stored in the previous process and the erased storage elements that have not yet been programmed can be interfered with. BACKGROUND OF THE INVENTION The present invention addresses these and other problems by providing a method for reducing interference in a non-volatile reservoir. In the practical package, a == line on the side of the line for operating the non-volatile memory is raised at least -_ string before the first: '~ ^ ^ at least one NAND string is lifted - wherein The first sub-line is on the non-polar side of the first word line. Including the first word line and the second word 131225.doc 200903499 The multi word line is associated with at least a letter (four), and at least one _d two = multiple non-volatile storage elements. The method further includes first applying a voltage to the first word line in the first mid-disc for providing a first non-volatile storage element associated with the conduction state::, two-pre-line; and applying a voltage: The two-word line is used to provide a first-non-volatile storage element associated with the (four) two-word line in the conductive state. # ^ ^ 运 运 ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ The medium and non-volatile storage elements, and simultaneously apply the program voltage to the first sub-line. Therefore, the source side rise occurs before the program pulse is applied. Implementing a method for operating a non-volatile memory is at least the first non-volatile storage in the NAND string: the first-elevation of the string, which is in the volatile storage element in the stylized sequence. prior to. The method further comprises a first non-volatile storage element on the first-lift two = one less and a second non-volatile storage element on the side of the pure storage element, wherein the sequence is in the first - non-volatile The storage element I# method further includes a side of the second non-volatile storage after the first promotion [•拙s, in the sequence of the second: non-string second promotion' After storing 7L pieces, the first storage element is provided in the conduction. In another embodiment, Yin includes (a) a first method for operating a non-volatile memory on a source side of a word line in a first period - '(i) applying a voltage to a word line set a particular set of word lines for boosting at least the first channel region of the -nand I3I225.doc •10.200903499 string; (ii) applying a voltage to the set of word lines including the particular word line to provide at least a -NAND string a: a set of word lines associated with a non-volatile storage element in a conductive state, a second set of word lines being on a drain side of the first set of word lines; and (iii) applying a voltage to the second set of word lines The third set of word lines on the drain side avoids lifting the second channel region of at least one nand string. The method further includes (b) in a second time period following the first time period: (1) applying a voltage to the third set of word lines for boosting the second channel region of the at least one NAND string; (ii) applying a program voltage to the Two word line

集合中之字線;及(iii)將電壓施加至特定字線以使第一通 道區與第二通道區隔離。 【實施方式】 本發明提供一種用於減少非揮發性儲存器中之程式干擾 的方法。 適用於實施本發明之記憶體系統之一實例利用nani^^ 閃記憶體結構,其包括在兩個選擇閘極之間串聯地配置多 個電晶體。串聯電晶體及選擇閘極被稱作NAND串。圖i為 展示一NAND串之俯視圖。圖2為其等效電路。圖i及圖二所 描繪之NAND串包括串聯且夾於第一選擇閘極12〇與第二選 擇閘極122之間的四個電晶體1〇〇、1〇2、1〇4及1〇6。選擇 閘極120閘控至位元線126之NAND串連接。選擇閘極122閘 控至源極線128之NAND串連接。藉由將適當電壓施加至控 制閘極120CG來控制選擇閘極12〇。藉由將適當電壓施加 至控制閘極122CG來控制選擇閘極122。電晶體1〇〇、 102、104及106中之每一者具有控制閘極及浮動閘極。電 131225.doc 200903499 晶體100具有控制閘極100CG及浮動閘極100FG。電晶體 102包括控制閘極102CG及浮動閘極102FG。電晶體104包 括控制閘極104CG及浮動閘極104FG。電晶體106包括控制 閘極106CG及浮動閘極106FG。控制閘極100CG連接至(或 為)字線WL3,控制閘極102CG連接至字線WL2,控制閘極 104CG連接至字線WL1,且控制閘極106CG連接至字線 WL0。在一實施例中,電晶體100、102、104及106各自為 儲存元件,亦被稱作記憶體單元。在其他實施例中,儲存 元件可包括多個電晶體,或可不同於圖1及圖2所描繪之儲 存元件。選擇閘極120連接至選擇線SGD。選擇閘極122連 接至選擇線SGS。 圖3為描繪三個NAND串之電路圖。利用NAND結構之快 閃記憶體系統的典型架構將包括若干NAND串。舉例而 言,在具有許多其他NAND串之記憶體陣列中展示三個 NAND串320、340及3 60。NAND串中之每一者包括兩個選 擇閘極及四個儲存元件。雖然為了簡單起見而說明四個儲 存元件,但現代NAND串可具有高達(例如)三十二個或六 十四個儲存元件。 舉例而言,NAND串320包括選擇閘極322及327以及儲存 元件323-326,NAND串340包括選擇閘極342及347以及儲 存元件343-346,NAND串360包括選擇閘極362及367以及 儲存元件363-366。每一 NAND串藉由其選擇閘極(例如, 選擇閘極327、347或367)而連接至源極線。選擇線SGS用 以控制源極側選擇閘極。各種NAND串320、340及360藉由 I31225.doc -12- 200903499 選擇閘極322、342、362等#中之選擇電晶體而連接至各 別位7L線321、341及361。此等選擇電晶體由汲極選擇線 SGD控制。在其他實施例中,選擇線未必需要在nand串 虽中為共同的;亦即,可為不同NAND串提供不同選擇 線。子線WL3連接至儲存元件323、343及363之控制閘 極。字線WL2連接至儲存元件324、344及364之控制閘 極。字線WL1連接至儲存元件325、345及365之控制閘 極子線WL0連接至儲存元件326、346及366之控制閘 極。可看出,每一位元線及各別NAND串包含儲存元件陣 列或集合之行。字線(WL3、WL2、wl^wl〇)包含陣列 或集合之列。每-字線連接列中之每—儲存元件的控制間 極。或’控制閘極可由字線自身提供。舉例而言,字線 WL2為儲存元件324、344及3M提供控制閘極。實務上, 在一子線上可存在數千個儲存元件。 每-儲存元件可儲存資料。舉例而言,當儲存一數位資 料位元時’將儲存元件之可能臨限電壓(VTH)範圍分成兩 個範圍’其經指派邏輯資料"i,,及,’〇"。在NAND型快閃記 憶體之一實例中’ Vth在抹除儲存元件之後為負,且經界 定為邏輯"1"。在程式操作 έ 。 玩栎作之後的vTH為正,且經界定為邏 輯”〇”。當vTH為負且試圖讀取時’儲存元件將開 邏輯被儲存。當vTH為正且試圖讀取操料,儲存元件 將不開啟,此指示邏輯” "袖健 铒被儲存。儲存元件亦可儲存多 個,訊㈣,例如’多個數位資料位元。在此狀況下,將 、值軏圍分成資料位準之數目。舉例而[若儲存四個 131225.doc -13- 200903499 貝Λ位準’則將存在經指派給資料值”11"、”1〇"、"01”及 〇〇之四個vTH範圍。在NAND型記憶體之一實例中,在抹 除刼作之後的VTH為負,且經界定為"u"。正Vth值用於狀 L 10 、"01”及"〇〇"。經程式化至儲存元件中之資料與元 件^臨限電壓範圍之間的特定關係視針對儲存it件所採用 之Η料編碼方案而定。舉例而言’美國專利第6,222,762號 及美國專利申請案公開案2_/G255G9G描述用於多狀態快And a word line in the set; and (iii) applying a voltage to the particular word line to isolate the first channel region from the second channel region. [Embodiment] The present invention provides a method for reducing program disturb in a non-volatile memory. One example of a memory system suitable for use in practicing the present invention utilizes a nani^ flash memory structure that includes a plurality of transistors arranged in series between two select gates. The series transistor and the select gate are referred to as NAND strings. Figure i is a top plan view showing a NAND string. Figure 2 is its equivalent circuit. The NAND strings depicted in FIG. 1 and FIG. 2 include four transistors 1〇〇, 1〇2, 1〇4, and 1〇 connected in series and sandwiched between the first selection gate 12〇 and the second selection gate 122. 6. The NAND string connection of the gate 120 gate to the bit line 126 is selected. The gate 122 is gated to the NAND string connection of the source line 128. The selection gate 12 is controlled by applying an appropriate voltage to the control gate 120CG. The selection gate 122 is controlled by applying an appropriate voltage to the control gate 122CG. Each of the transistors 1, 102, 104, and 106 has a control gate and a floating gate. Electric 131225.doc 200903499 The crystal 100 has a control gate 100CG and a floating gate 100FG. The transistor 102 includes a control gate 102CG and a floating gate 102FG. The transistor 104 includes a control gate 104CG and a floating gate 104FG. The transistor 106 includes a control gate 106CG and a floating gate 106FG. Control gate 100CG is coupled to (or) word line WL3, control gate 102CG is coupled to word line WL2, control gate 104CG is coupled to word line WL1, and control gate 106CG is coupled to word line WL0. In one embodiment, transistors 100, 102, 104, and 106 are each a storage element, also referred to as a memory unit. In other embodiments, the storage element may comprise a plurality of transistors, or may be different from the storage elements depicted in Figures 1 and 2. The selection gate 120 is connected to the selection line SGD. The selection gate 122 is connected to the selection line SGS. Figure 3 is a circuit diagram depicting three NAND strings. A typical architecture for a flash memory system utilizing a NAND structure would include several NAND strings. For example, three NAND strings 320, 340, and 3 60 are shown in a memory array having many other NAND strings. Each of the NAND strings includes two select gates and four storage elements. Although four storage elements are illustrated for simplicity, modern NAND strings can have up to, for example, thirty-two or sixty-four storage elements. For example, NAND string 320 includes select gates 322 and 327 and storage elements 323-326, NAND string 340 includes select gates 342 and 347 and storage elements 343-346, NAND string 360 includes select gates 362 and 367 and storage Elements 363-366. Each NAND string is connected to the source line by its select gate (e.g., select gate 327, 347, or 367). The selection line SGS is used to control the source side selection gate. The various NAND strings 320, 340, and 360 are connected to respective bit 7L lines 321, 341, and 361 by selecting a selection transistor in gates 322, 342, 362, etc., by I31225.doc -12-200903499. These selective transistors are controlled by the drain select line SGD. In other embodiments, the select lines do not necessarily need to be common in the nand string; that is, different select lines can be provided for different NAND strings. The sub-line WL3 is connected to the control gates of the storage elements 323, 343 and 363. Word line WL2 is coupled to the control gates of storage elements 324, 344 and 364. Control gate WL0, to which word line WL1 is coupled to storage elements 325, 345, and 365, is coupled to the control gates of storage elements 326, 346, and 366. It can be seen that each bit line and each NAND string contains a row of storage element arrays or sets. The word lines (WL3, WL2, wl^wl〇) contain arrays or arrays. Each word line connects each of the columns to the control element of the storage element. Or the 'control gate' can be provided by the word line itself. For example, word line WL2 provides control gates for storage elements 324, 344, and 3M. In practice, there can be thousands of storage elements on a sub-line. Each storage element can store data. For example, when storing a bit of data bits, the range of possible threshold voltages (VTH) of the storage elements is divided into two ranges', which are assigned logical data "i,,,,'". In an example of a NAND type flash memory, 'Vth is negative after erasing the storage element and is defined as a logical "1". In the program operation έ . The vTH after the game is positive and is defined as a logical "〇". When the vTH is negative and an attempt is made to read, the storage element will be logically stored. When the vTH is positive and an attempt is made to read the material, the storage component will not be turned on, and the indication logic is stored. The storage component can also store multiple messages (4), such as 'multiple digit data bits. In this case, divide the value into the number of data levels. For example, if you store four 131225.doc -13-200903499, you will be assigned the data value "11"," 1〇 ", "01" and the four vTH ranges of 〇〇. In one example of a NAND type memory, the VTH after the erase operation is negative and is defined as "u". The positive Vth value is used for the shape L 10 , "01" and "〇〇". The specific relationship between the data programmed into the storage element and the component's threshold voltage range is used for the storage of the piece. Depending on the data encoding scheme, for example, 'US Patent No. 6,222,762 and US Patent Application Publication No. 2_/G255G9G are described for multi-state fast

閃儲存7L件之各種資料編碼方案,該等案以引用之方式全 文併入本文中。 NAND型快閃記憶體及其操作之相關實例提供於美國專 利第 5,386,422 號、第 5,522,58〇號、第 5,57〇,315 號、第 5,774’397 號、第 6,〇46,935 &、帛 M56,528 號及第 M22,號中’該等專利中之每—者以引用之方式併入本 文中。 *程式化快閃健存元件時,將程式電塵施加至儲存元件 之控制閘極,且使與儲存元件相_之位元線接地。將來 =道之電子注人至浮動閘極中。當電子累積於浮動閑極 中知,洋動閘極變得帶負電’且儲存元件之、升 了將知·式電壓施加至妹兹-V' /μ々·' ,“王式化之儲存元件的控制閘極,將 彼程式電Μ施加於適當字線上。如上文所論述,^ 中之每-者中的一儲存元件共用同—字線。舉例而古,合 程式化圖3之儲存元件324時,亦將程式電摩施加至二 件344及364之控制閘極。 然而,可在其他Ν娜串之程式化期間在受抑制Nand 13I225.doc -14- 200903499 串處發生程式干擾,且有時在經程式化nand串自身處發 生私式干擾。當未選非揮發性儲存元件之臨限電壓歸因於 其他非揮發性儲存元件之程式化而移㈣,發生程式干 擾。可在先前經程式化儲存元件以及尚未程式化之經抹除 儲存元件上發生程式干擾。各種程式干擾機制可限制用於 =揮發性儲存裝置(諸如,NAND快閃記憶體)之可用操作 舉例而言’若NAND串320受抑制(例如,其為不含有告 前經程式化之儲存元件的未選糊〇串)且副㈣州: 程式化(例如,其為含有當前、《式化之儲存元件的所選 NAND串),則可在NAN”32G處發生程式干擾。舉例而 言’若通過電壓(pass VGltage)VpAss為低,則受抑制Μ. 串之通道未經良好地提升,且未㈣伽串之所選字線可 經無意地程式化。在另—可能情境中’經提升電壓可藉由 閘極引發沒極、;__或其他_制來降低,此導致 相同問題。其他效應(諸如,電荷儲存元件歸因於與稍後 經程式化之其他鄰近儲存元件之電容耗合的VTH移位)亦可 有助於程式干擾。 圖4描繪展示程式干擾機制之Nand串的橫截面圖。此 處利用(諸如)圖5c所描繪之經修訂抹除區域自我提升 (SB)模式6亥視圖為簡化的且未按比例。NAND串400 包括形成於基板490上之源極側選擇閘極4〇6、汲極側選擇 問極似,及八個儲存元件彻、410、412、414、416、 418 、 420及422 p井區自身形成 組件可形成於p井區上 131225.doc 15 200903499 於基板之η井區_。n井又可形成於p基板中。除了具有電 位VBL之位元線426以外,還提供具有電位乂咖似之源極 供應線404。在程式化期間,將VpGM提供於與所選儲存元 件418相關聯之所選字線(在此狀況下,為阳)上。另外, 回想到,可將儲存元件之控制閘極提供為字線之一部分。 舉例而言 ’ WLO、WL1、WL2、WL3、WL4'wu、wm 及WL7可分別經由儲存元件4〇8、4l〇、4i2、am、ye、 418、420及422之控制閘極而延伸。 在-實例提升模式中,當儲存元件418為所選儲存元件 時,將相對較低電壓VL0W(例如,2_6 v)施加至鄰近源極側 字線(WL3),而將隔離電壓Vis〇(例如,〇_4 v)施加至被稱 作隔離字線之另-源極側字線(WL2),且將v⑽施加至盥 NAND串400相關聯之剩餘字線(意即,WL0、WL1、 机4、WL6及WL7)。雖然^〇及%⑽之絕對值可在相對較 大且部分地重疊之範圍内變化,但在一可能實施例中, vlso之值始終低於Vlow。Vsgs施加至選擇閘極4〇6,且 VSGD施加至選擇閉極424。字線或非揮發性儲存元件之源 極側指代面向NAND串之源極端的側(例如,在源極供庫線 _處),而字線或非揮發性儲存元件之汲極側指代面向 NAND串之汲極端的側(例如,在位元線426處)。 圖5a至圖5h描繪自我提升模式之不同實例。注意,所描 緣之電壓指示在源極側提升之後發生之汲極側提升期間= 利用的電壓。亦見圖6至圖9。亦可利用各種其他方法。通 常,已開發各種類型之提升模式來對抗程式干擾。在所選 131225.doc -16 - 200903499 字線上之儲存元件的程式化期間, 至與當前未經程式化之储存元件、隹…電壓集^加 施提井槿……, 丁通信的未選字線來實 並"、&程式化之儲存元件與所選NAND串相關 V,』儲存元件與未選NAND串相關聯。 在所提供之實例中’字線為㈣至机 WLn,源極側選擇閘極控 圯予踝马 帝J線為SGS,且汲極側選擇閘極 控制線為SGD。亦描緣施加至控制線之電壓集合。程式化 可自NAND串之源極側至沒極側以—次―個字線^式化 序列來進行。然而,亦可利用其他程式化序列。舉例而 言,在兩步驟程式化技術中,可在第一進程中部分地程式 化NAND串之儲存元件’在第一進程中係自nand串之源 極側至沒極側-次一個字線地進行。接著在第二進程中完 成程式化,在第二進程中亦係自NAND串之源極側至汲極 側一次一個字線地進行。在另一選擇中,在兩上一下過程 中程式化儲存元件,例如,按以下序列:WL〇(部分程式 化)、WL1(部分程式化)、WL〇(完成程式化)、WL2(部分程 式化)、WL1 (完成程式化)、WL3(部分程式化),等等。 在圖5a所示之實例中,所施加之電壓包括:vSGs,其施 加至源極側選擇閘極控制線SGS ;通過電壓VPASS,其施加 至未選字線WL0至WLn-2及WLn+Ι至WLi中之每一者;程 式電壓VpGM, 其施加至所選字線WLn ;隔離電壓VIS〇,其 施加至相鄰於源極側上之所選字線的字線WLn-Ι ;及 VSGD ’其經由汲極側選擇閘極控制線SGD而施加^通常, VSGS為0 V ’使得源極側選擇閘極關閉,可施加在〇 5-1 $ 131225.doc -17- 200903499 v之範圍内的額外源極偏壓Vs〇urce以進一步改良源極側選 擇閘極之切斷行為。VSGD為約15_3 V,使得所選NANE)串 之汲極側選擇閘極歸因於相應低位元線電壓Vbl(諸如,οι V)之施加 而開啟 。未選 /受抑 制 NAND 串之 汲極側 選擇閘 極歸因於相應較兩VBL(諸如,1.5-3 V)之施加而關閉。在 圖5a之實例中,將在〇_4 v之典型範圍内的低隔離電壓ναό 施加至相鄰於源極側上之所選字線的字線。 另外,VPASS可為約7_1〇 v,且VpGM可自約12_25 v變 化。在一程式化方案中,將程式電壓之脈衝串施加至所選 字線。見圖20。脈衝串中之每一連續程式脈衝的振幅以階 梯方式來增加,通常為每脈衝增加約〇 3_〇 5 v。另外,可 將驗證脈衝施加於程式脈衝之間以驗證所選儲存元件是否 已達到目標程式化條件。亦注意,每一個別程式脈衝可具 有固疋振幅,或可具有變化振幅。舉例而言,一些程式化 方案施加具有如斜坡或階梯般變化之振幅的脈衝。可利用 任何類型之程式脈衝。 在WLn為經程式化之字線且程式化係自每一 nand串之 源極側至祕側進行的情況下,自從最後抹㈣作以來, 與WLOS WLn_i相關聯之儲存元件將已經至少部分地經程 式化,且當程式化WLn上之儲存元件時,與WLn+i至;U =關聯之儲存元件將經抹除或至少尚未完全程式化。未選 字線上之通過電壓耦合至與未選NAND串相關聯之通道,、 從而使在未選NAND串之通道中存在電壓,其傾向於藉由 降低跨越儲存兀件之随道氧化物的電塵來減少程式干擾。 13I225.doc •18- 200903499 圖5b描繪經修訂抹除區域自我提升模式。在此狀況下, 隔離電魔VIS0施加至WLn-2,且介於、〇與乂⑽之間的低 電壓VL0W施加至WLn-丨。Vl〇w亦可被視為隔離電壓,然 而,在一可能實施例中,乂_始終高於Vis〇且低於ν_。 在此方法中,vL0W充當中間電壓,使得在所選字線(wLn) 與相鄰源極側字線(WLn_丨及WLn_2)之間的通道中存在較 少突然電壓改變。舉例而言,Vl〇w可為(例如)2_6 v,且 VIS0可為(例如)〇-4 V。通道電壓之較少突然改變導致通道 區中之較低電場及較低通道電位,尤其係在與U字線相 關聯之儲存元件處。與Vis〇字線(如在圖中)相關聯之儲 存7L件之汲極側或源極側處的高通道電壓可使電荷載流子 (電子及電洞)藉由閘極引發汲極洩漏(GIDL)而產生。由 GIDL所產生之電子隨後可在所選字線與Vis。字線之間的區 域中之強電場中加速’且隨後可注入(經由熱電子注入)於 與所選字線相關聯之儲存元件中的—些中且因此引起程式 干擾。此程式干擾機制可藉由減少電場來避免或減少,諸 如,藉由添加以介於所選字線之電壓與Vis〇之間的中間電 壓所偏壓之一(或多個)字線。 剩餘未選字線接收VpASs。具體言之,VPASS施加至與 WL〇至U相關聯之第-儲存元件群,丨中第-群相鄰 於源極側:4擇閘極且在隔離字線WLn_2之源極側上。又, VPAS0&加至與WLn+1至WU相關聯之第二儲存元件群,其 中第二群相鄰於汲極側選擇閑極且在所選字線WLn之没極 側上。 131225.doc 19 200903499 圖5c描繪另一經修訂抹除區域自我提升模式。在此狀況 下’相鄰於所選字線(WLn)之源極側字線(wLn-Ι)接收 Vpass,下一字線(WLn-2)接收VL0W,且在彼之後的下一字 線(WLn-3)接收V1S0。剩餘未選字線接收VpAss。亦結合圖4 來論述此提升模式。具體言之,VPASS施加至與WL0至 WLn-4相關聯之第一儲存元件群,其中第一群相鄰於源極 側選擇閘極且在隔離字線WLn-3之源極側上。又,VpAs^fc 加至與WLn+1至WLi相關聯之第二儲存元件群,其中第二 群相鄰於汲極側選擇閘極且在所選字線WLn之汲極側上。 此方法之優勢在於:所選字線(其由於施加至彼字線之高 程式電壓VPGM而對程式干擾最敏感)更遠離να。及字 線。與所選字線相關聯之儲存元件較不可能受熱電子注入 干擾因為負貝產生熱載流子之電場位於更遠離所選字線 處0 圖ScUfe繪另一經修訂抹除區域自我提升模式。在此狀況 下,相鄰於所選字線(WLn)之源極側字線(U接收 VPASS ’下一字線(WLn_2)接收Vl〇w,下—字線(wLn_3)接 收vIS0’且下一字線接收Vl〇w。乘m未選字線接收v簡。 具體言之’ vPASS施加至與机0至心_5相關聯之第一儲存 元件群,*中第-群相鄰於源極側選擇閘極且在隔離字線 WLn-3之源極側上。又,VpAss施加至與饥川至和相關 如之第一儲存兀件群’ I中第二群相鄰於汲極側選擇閘極 且在所選字線WLn之汲極側上。在隔離字線之兩側處提供 vLOW可減少歸因於高度經提升源極側(例如’在與wl〇至 13I225.doc •20- 200903499 WL5相關聯之通道的一都八南、τ π < n j 4分處)而在隔離字線處發生gidl 的機率。 圖㈣繪另—經修訂抹除區域自我提升模式。在此狀況 下,相鄰於所選字線(WLn)之源極側字線(乳卜”接收 VPASS.HIGH,下—字線(WLn_2)接收v隨圓_,下一字線 (WLn-3)接收 VPASS_L0W,下—字線(WLn_4)接收 Vl〇w,下一 字線(WLn-5)接收VISO,且下一字線(WLn_6)接收Vl〇w。剩 餘未選字線接收VpASS。具體言之,VpAss施加至與wL〇至 WLn-7相關聯之第一儲存元件群,其中第一群相鄰於源極 側選擇閘極且在隔離字線WLn_5之源極側上。又,¥關施 加至與WLn+Ι至WLi相關聯之第二儲存元件群,其中第二 群相鄰於汲極側選擇閘極且在所選字線WLn之汲極側上。 因此,可同時利用多個vPASS電壓。舉例而言,可為 NAND串之汲極側及源極側利用不同VpAss值。另外,可在 汲極側及源極側處均利用多個VpAss電壓。例如,如所描 繪,可緊接於所選字線而利用較高VpAss(VpAss_)以用於 程式化。對於所選字線與隔離字線之間的字線而言,可具 有經偏壓至*同VpASS值(例如,VpAss__、〜似麵顆及 VPASS-HIGH)之多個字線。在一實施例中,VpGM>VpAss. hIGH>VPASS-MEDIUM>Vpass_l〇w>Vl〇w>Vis〇。注意,&㈣及 vIS〇之多個值亦為可能的。通常,所有Vis〇電壓均小於所 有VL0W電壓,所有vL〇w電壓又均小於所有VPASS電壓。藉 由增加所選字線與VIS0字線之間的字線之數目,且藉由逐 漸地減少彼等字線上之偏壓,可減少所選字線與v〗s〇字線 131225.doc 21 200903499 之間的電場且因此可減少裎式干擾。 圖5f描繪另一經修訂抹除區域自我提升模式。在此狀況 下,相鄰於所選字線(WLn)之源極側字線(WLn—丨)接收 VPASS-HIGH ’ 下一字線(WLn-2)接收 VPASS-MEDIUM,下一字線 (WLn-3)接收 VPASS_L0W,下一字線(WLn_4)接收 Vl〇w,下一 字線(WLn-5)接收VIS0、下一字線(WLn_6)接收Vl〇w,且下 一字線(WLn-7)接收VPASS.L〇w。剩餘未選字線接收。 具體言之,vPASS施加至與WL0至 WLn_8相關聯之第一儲存 元件群,其中第一群相鄰於源極側選擇閘極且在隔離字線 WLn-5之源極側上。又,vPASS施加至與WLn+Ι至WLi相關 聯之第二儲存元件群,其中第二群相鄰於汲極側選擇閘極 且在所選字線WLn之没極側上。 圖5g描繪另一經修訂抹除區域自我提升模式。此狀況與 圖5f之狀況的不同之處在於:相鄰於所選字線(WLn)之汲 極側字線(WLn+Ι)接收VPASS.H1GH而非VpASS 0 圖5h描繪另一經修訂抹除區域自我提升模式。在此狀況 下,額外隔離字線提供於經程式化字線之汲極側上。舉例 而言,與圖5 c之提升模式相比,在一可能實施例中, WLn+1 接收 VPASS-HiGH,且 WLn+3 接收 V丨s。。WLn + 2 接收 VPASS,其中VPASS-HIGH>vPASS。由於施加提升電壓及兩個 隔離電壓’在NAND串中提供三個經提升通道區域。舉例 而言,第一經提升通道區域係在WL0至WLn-4之區中,第 一經提升通道區域係在WLn-1至WLn+2之區中,且第三經 提升通道區域係在WLn+4至WLi之區中。(諸如)當可以下 J31225.doc •22- 200903499 部頁資料來部分地程式化WLn+1 (見(例如)圖1此之Βι狀態) 時,Vpass-high之利用移除對WLn+Ι之資料依賴性。可類似 地修改圖5d至圖5g之提升模式。 各種其他實把例為可能的。舉例而言,可將不同經提升 通道區域提升至不同位準。又,所選字線與額外汲極側隔 離字線之間的字線之數目可變化’施加至不同經提升通道 區域中之未選字線的電壓亦可變化。亦可提供具有兩個以 上隔離電壓及三個經提升通道區域之實施例。對於其他細 節,參考2006年9月2?日提出申請之標題為”ReducingVarious data encoding schemes for flash storage of 7L pieces are incorporated herein by reference. Examples of NAND type flash memory and its operation are provided in U.S. Patent Nos. 5,386,422, 5,522,58, 5,57,315, 5,774'397, 6, 6, 46,935 &帛M56, 528 and M22, each of which is incorporated herein by reference. * When the flashing component is programmed, the program dust is applied to the control gate of the storage component, and the bit line of the storage component is grounded. In the future = the electronic injection of the road into the floating gate. When the electrons accumulate in the floating idle pole, the oceanic gate becomes negatively charged, and the storage element rises and the known voltage is applied to the sister-V'/μ々·', "the storage element of the king." The gate is controlled and the program is applied to the appropriate word line. As discussed above, one of the storage elements of the ^ share the same word line. For example, the storage element 324 of Figure 3 is programmed. At the same time, the program is also applied to the control gates of the two pieces 344 and 364. However, during the stylization of the other strings, program disturb occurs at the suppressed Nand 13I225.doc -14- 200903499 string, and there is Private interference occurs at the programmed nand string itself. When the threshold voltage of the unselected non-volatile storage element is shifted due to the stylization of other non-volatile storage elements (4), program disturb occurs. Program disturb occurs on stylized storage elements and erased storage elements that have not yet been programmed. Various program interference mechanisms may limit the available operations for = volatile storage devices (such as NAND flash memory). String 320 is suppressed (for example, it is an unselected paste string that does not contain pre-programmed storage elements) and the secondary (four) state: stylized (for example, it is a selected NAND string containing the current, "typed storage element", Program disturb can occur at NAN"32G. For example, if the pass VGltage VpAss is low, it is suppressed. The channel of the string is not well boosted, and the selected word line of the (four) gamma string can be unintentionally programmed. In another-possible scenario, the boosted voltage can be reduced by the gate-induced immersion, __ or other _ system, which causes the same problem. Other effects, such as the VTH shift of the charge storage element due to the capacitance of other adjacent storage elements that are later programmed, may also contribute to program disturb. Figure 4 depicts a cross-sectional view of a Nand string showing the program interference mechanism. The revised erase area self-boosting (SB) mode 6-her view, such as depicted in Figure 5c, is simplified and not to scale. The NAND string 400 includes a source side selection gate 4〇6 formed on the substrate 490, a drain side selection pole, and eight storage elements, 410, 412, 414, 416, 418, 420, and 422 p wells. The zone self-forming component can be formed on the p-well region 131225.doc 15 200903499 in the n-well region of the substrate. The n well can be formed in the p substrate. In addition to the bit line 426 having the potential VBL, a source supply line 404 having a potential is provided. During stylization, VpGM is provided on the selected word line (in this case, yang) associated with the selected storage element 418. Additionally, it is recalled that the control gate of the storage element can be provided as part of the word line. For example, 'WLO, WL1, WL2, WL3, WL4'wu, wm, and WL7 may extend through the control gates of storage elements 4〇8, 4l〇, 4i2, am, ye, 418, 420, and 422, respectively. In the - instance boost mode, when the storage element 418 is the selected storage element, a relatively lower voltage VL0W (eg, 2_6 v) is applied to the adjacent source side word line (WL3), and the isolation voltage Vis is (eg, , 〇_4 v) is applied to the other-source side word line (WL2), which is referred to as an isolated word line, and v(10) is applied to the remaining word lines associated with the 盥 NAND string 400 (ie, WL0, WL1, machine 4. WL6 and WL7). Although the absolute values of ^ 〇 and % (10) may vary over a relatively large and partially overlapping range, in a possible embodiment, the value of vlso is always lower than Vlow. Vsgs is applied to select gate 4〇6 and VSGD is applied to select closed pole 424. The source side of the word line or non-volatile storage element refers to the side facing the source terminal of the NAND string (eg, at the source supply line _), while the drain side of the word line or non-volatile storage element refers to The side facing the extreme of the NAND string (eg, at bit line 426). Figures 5a through 5h depict different examples of self-elevation modes. Note that the voltage of the indicated voltage indicates the voltage used during the drain side lift occurring after the source side is boosted. See also Figures 6 to 9. Various other methods are also available. In general, various types of boost modes have been developed to combat program disturb. During the stylization of the storage elements on the selected 131225.doc -16 - 200903499 word line, to the currently unprogrammed storage elements, 隹...voltage set ^ 提提井槿..., unselected words of Ding Communication The line is ", & stylized storage elements associated with the selected NAND string V, the storage element is associated with the unselected NAND string. In the example provided, the word line is (4) to WLn, the source side selects the gate control 圯 to the 踝马 帝 J line is SGS, and the drain side select gate control line is SGD. The set of voltages applied to the control line is also depicted. The stylization can be performed from the source side to the immersed side of the NAND string in a sequence of - word lines. However, other stylized sequences can also be utilized. For example, in the two-step programming technique, the storage element of the NAND string can be partially programmed in the first process from the source side of the nand string to the bottom side of the nand string in the first process. Conducted. Then, the stylization is completed in the second process, and in the second process, the word line is also performed one line at a time from the source side to the drain side of the NAND string. In another option, the storage elements are stylized in two previous processes, for example, in the following sequence: WL〇 (partially stylized), WL1 (partially stylized), WL〇 (completed stylized), WL2 (partial program) , WL1 (completed stylization), WL3 (partially stylized), and so on. In the example shown in Figure 5a, the applied voltage includes: vSGs applied to the source side select gate control line SGS; pass voltage VPASS applied to the unselected word lines WL0 to WLn-2 and WLn+Ι To each of WLi; a program voltage VpGM applied to the selected word line WLn; an isolation voltage VIS〇 applied to the word line WLn-Ι adjacent to the selected word line on the source side; and VSGD 'It is applied via the drain side selection gate control line SGD. Normally, VSGS is 0 V ', so that the source side selection gate is closed and can be applied within the range of 〇5-1 $131225.doc -17- 200903499 v The additional source bias voltage Vs〇urce further improves the switching behavior of the source side select gate. The VSGD is about 15_3 V such that the drain side select gate of the selected NANE) string is turned on due to the application of the corresponding low bit line voltage Vbl (such as οι V). The unselected/suppressed NAND string's drain side select gate is turned off due to the application of two more VBLs (such as 1.5-3 V). In the example of Figure 5a, a low isolation voltage να 典型 within a typical range of 〇_4 v is applied to the word line adjacent to the selected word line on the source side. In addition, VPASS can be about 7_1 〇 v, and VpGM can vary from about 12 _25 v. In a stylized scheme, a pulse of program voltage is applied to the selected word line. See Figure 20. The amplitude of each successive program pulse in the burst is increased in a stepwise manner, typically by about 〇 3_〇 5 v per pulse. Alternatively, a verify pulse can be applied between the program pulses to verify that the selected storage element has reached the target stylized condition. Also note that each individual program pulse may have a solid amplitude or may have a varying amplitude. For example, some stylization schemes apply pulses with amplitudes that vary like ramps or steps. Any type of program pulse can be utilized. In the case where WLn is a stylized word line and the stylization is performed from the source side to the secret side of each nand string, the storage element associated with WLOS WLn_i has been at least partially since the last wipe (4) Stylized, and when the storage elements on WLn are programmed, the storage elements associated with WLn+i to; U= will be erased or at least not fully programmed. The pass voltage on the unselected word line is coupled to the channel associated with the unselected NAND string, thereby causing a voltage in the channel of the unselected NAND string, which tends to reduce the electrical charge across the oxide of the memory element Dust to reduce program interference. 13I225.doc •18- 200903499 Figure 5b depicts the revised erased area self-elevation mode. In this case, the isolated electric magic VIS0 is applied to WLn-2, and the low voltage VL0W between 〇 and 乂 (10) is applied to WLn-丨. Vl〇w can also be considered as an isolation voltage, however, in a possible embodiment, 乂_ is always higher than Vis 〇 and lower than ν_. In this method, vL0W acts as an intermediate voltage such that there is less abrupt voltage change in the channel between the selected word line (wLn) and the adjacent source side word lines (WLn_丨 and WLn_2). For example, Vl〇w can be, for example, 2_6 v, and VIS0 can be, for example, 〇-4 V. Less abrupt changes in channel voltage result in lower electric fields and lower channel potentials in the channel region, especially at the storage elements associated with the U-line. The high channel voltage at the drain side or source side of the stored 7L piece associated with the Vis 〇 word line (as shown in the figure) allows charge carriers (electrons and holes) to cause a drain leak through the gate Produced by (GIDL). The electrons generated by GIDL can then be in the selected word line with Vis. Acceleration in a strong electric field in the region between the word lines' and subsequent injection (via hot electron injection) into the memory elements associated with the selected word line and thus causing program disturb. This program disturbing mechanism can be avoided or reduced by reducing the electric field, such as by adding one (or more) word lines biased by an intermediate voltage between the voltage of the selected word line and Vis. The remaining unselected word lines receive VpASs. Specifically, VPASS is applied to the first-storage element group associated with WL 〇 to U, where the first-group is adjacent to the source side: 4 is the gate and is on the source side of the isolated word line WLn_2. Further, VPAS0& is applied to the second storage element group associated with WLn+1 to WU, wherein the second group is adjacent to the drain side select idle and on the bottom side of the selected word line WLn. 131225.doc 19 200903499 Figure 5c depicts another revised erased area self-elevation mode. In this case, the source side word line (wLn-Ι) adjacent to the selected word line (WLn) receives Vpass, and the next word line (WLn-2) receives VL0W, and the next word line after the other (WLn-3) receives V1S0. The remaining unselected word lines receive VpAss. This lifting mode is also discussed in conjunction with FIG. In particular, VPASS is applied to the first set of storage elements associated with WL0 through WLn-4, wherein the first group is adjacent to the source side select gate and on the source side of the isolated word line WLn-3. Further, VpAs^fc is applied to the second storage element group associated with WLn+1 to WLi, wherein the second group is adjacent to the drain side selection gate and on the drain side of the selected word line WLn. The advantage of this method is that the selected word line (which is most sensitive to program disturb due to the high program voltage VPGM applied to the word line) is further away from να. And word line. The storage element associated with the selected word line is less likely to be disturbed by hot electron injection because the electric field of the hot carrier generated by the negative shell is located further away from the selected word line. Figure 4Cufe depicts another modified erase area self-boost mode. In this case, the source side word line adjacent to the selected word line (WLn) (U receives VPASS 'the next word line (WLn_2) receives Vl〇w, and the lower-word line (wLn_3) receives vIS0' and A word line receives Vl〇w. Multiply the m unselected word line to receive v. Specifically, 'vPASS is applied to the first storage element group associated with machine 0 to heart_5, * the first group is adjacent to the source The pole side selects the gate and is on the source side of the isolation word line WLn-3. Further, VpAss is applied to the second group adjacent to the first group of the first storage element group 'I. Select the gate and on the drain side of the selected word line WLn. Providing vLOW at both sides of the isolated word line can be reduced due to the height of the boosted source side (eg 'in w1〇 to 13I225.doc •20 - 200903499 WL5 associated channel of all the eight nan, τ π < nj 4 points) and the probability of gidl occurring in the isolated word line. Figure (4) painted another - revised erase area self-improvement mode. In this situation Next, the source side word line adjacent to the selected word line (WLn) receives VPASS.HIGH, and the lower-word line (WLn_2) receives v with circle_, and the next word line (WLn-3) receives VPAS S_L0W, the lower-word line (WLn_4) receives Vl〇w, the next word line (WLn-5) receives VISO, and the next word line (WLn_6) receives Vl〇w. The remaining unselected word lines receive VpASS. VpAss is applied to the first storage element group associated with wL〇 to WLn-7, wherein the first group is adjacent to the source side selection gate and on the source side of the isolation word line WLn_5. a second storage element group associated with WLn+Ι to WLi, wherein the second group is adjacent to the drain side selection gate and on the drain side of the selected word line WLn. Therefore, multiple vPASSs can be utilized simultaneously For example, different VpAss values can be used for the drain side and the source side of the NAND string. In addition, multiple VpAss voltages can be utilized at both the drain side and the source side. For example, as depicted, it can be tight Higher VpAss (VpAss_) are used for programming when connected to the selected word line. For word lines between the selected word line and the isolated word line, there may be biased to *with VpASS values (eg, Multiple word lines of VpAss__, ~face-like and VPASS-HIGH. In one embodiment, VpGM>VpAss.hIGH>VPASS-MEDIUM>Vpass_l〇w>Vl 〇w>Vis〇. Note that multiple values of & (iv) and vIS〇 are also possible. Typically, all Vis〇 voltages are less than all VL0W voltages, and all vL〇w voltages are less than all VPASS voltages. The number of word lines between the selected word line and the VIS0 word line, and by gradually reducing the bias voltage on the word lines, the selected word line can be reduced from the v s 〇 word line 131225.doc 21 200903499 The electric field and therefore the ripple interference can be reduced. Figure 5f depicts another revised erased area self-elevation mode. In this case, the source side word line (WLn_丨) adjacent to the selected word line (WLn) receives VPASS-HIGH'. The next word line (WLn-2) receives VPASS-MEDIUM, the next word line ( WLn-3) receives VPASS_L0W, the next word line (WLn_4) receives Vl〇w, the next word line (WLn-5) receives VIS0, the next word line (WLn_6) receives Vl〇w, and the next word line (WLn) -7) Receive VPASS.L〇w. The remaining unselected word lines are received. Specifically, vPASS is applied to the first group of storage elements associated with WL0 through WLn_8, wherein the first group is adjacent to the source side select gate and on the source side of the isolated word line WLn-5. Further, vPASS is applied to the second storage element group associated with WLn + Ι to WLi, wherein the second group is adjacent to the drain side selection gate and on the bottom side of the selected word line WLn. Figure 5g depicts another revised erased area self-elevation mode. This condition differs from the condition of Figure 5f in that the drain side word line (WLn+Ι) adjacent to the selected word line (WLn) receives VPASS.H1GH instead of VpASS 0. Figure 5h depicts another modified erase Regional self-upgrading mode. In this case, additional isolation word lines are provided on the drain side of the programmed word line. For example, in a possible embodiment, WLn+1 receives VPASS-HiGH and WLn+3 receives V丨s, as compared to the boost mode of Figure 5c. . WLn + 2 receives VPASS, where VPASS-HIGH>vPASS. Three elevated channel regions are provided in the NAND string due to the applied boost voltage and two isolation voltages'. For example, the first elevated channel region is in the region of WL0 to WLn-4, the first elevated channel region is in the region of WLn-1 to WLn+2, and the third elevated channel region is in WLn. +4 to the area of WLi. (such as) When the WDn+1 can be partially programmed by J31225.doc •22-200903499 (see, for example, Figure 1 for the Βι state), the use of Vpass-high removes the WLn+ Data dependency. The lifting mode of Figures 5d to 5g can be similarly modified. Various other practical examples are possible. For example, different elevated channel areas can be raised to different levels. Also, the number of word lines between the selected word line and the extra drain side isolation word lines can vary. The voltage applied to the unselected word lines in the different elevated channel regions can also vary. Embodiments having more than two isolation voltages and three elevated channel regions are also provided. For other details, refer to the title of “Reducing” on September 2, 2006.

Program Disturb In Non-Volatile Storage"的樓案號為 SAND-1120/SDK-0868之美國專利申請案第11/535,628號, 其以引用之方式併入本文中。 關於不同通道區之提升的時序,各種實施例為可能的。 考慮WL0與WLn-4之間的第一通道區、WLn-Ι與WLn+2之 間的第二通道區及WLn+4與WLi之間的第三通道區。在一 方法中’共同提升第一通道區及第三通道區,此後提升第 二通道區。在一方法中,提升第一通道區,此後共同提升 弟一通道區及第三通道區。在一方法中,提升第一通道 區’此後提升第三通道區,此後提升第二通道區。通常, 應較佳不在第三通道區之前提升第二通道區,因為來自第 二通道區之電子將被吸引至經提升第二通道區,因此降低 第二通道區中之經提升通道電位,同時稍微提升第三通道 區。此為不當效應’因為所降低之提升可引起程式干擾。 注意’所有上述實例僅充當說明,因為其他偏壓條件及 131225.doc •23· 200903499 偏壓條件之不同組合為可能的 再次參看圖4,假定沿著Nand串400之儲存元件的程式 化係以自儲存元件408至儲存元件422之程式化序列進行, 則將已經至少部分地程式化儲存元件408-4 1 ό,且將尚未 完全程式化儲存元件42〇及422。因此,儲存元件408·416 中之全部或一些將具有經程式化至且儲存於其各別浮動閘 電子且可抹除或部分地程式化儲存元件420及 422,此視程式化模式而定。舉例而言,當先前已在兩步 驟程式化技術之第一步驟中程式化儲存元件42〇及422時, 可部分地程式化儲存元件420及422。 在EASB或REASB提升模式之情況下,在起始提升之後 的某寺將VIso^加至所選字線之一或多個源極側鄰近者, 且Vis〇足夠低以隔離基板中之經程式化通道區域與經抹除 通道區域。亦即,基板490在隔離字線412之源極側上的通 道區域與基板在隔離字線412之汲極側上的通道區域隔 離。源極側亦可被視為經程式化側,因為已程式化關聯儲 存π件中之大多數或全部,而汲極側亦可被視為未經程式 化側’因為尚未程式化關聯儲存元件。另夕卜,源極側上之 通,區域為基板49〇之藉由對WLG及wu施加v⑽而提升 的第一經提升區,而汲極側上之通道區域為基板49〇之主 要藉由對WL5施加VpGM且對WL4、WL6及机7施加%似而 提升的第二經提升區。 -般而言’經程式化區域經提升得較少,因為在經程式 化儲存元件下之通道電位僅〇pAss達到足夠高位準以開 131225.doc -24- 200903499 啟經程式化儲存元件之後才可開始增加(例如,經提升)。 另一方面,在經抹除條件下儲存元件之通道電位將在施加 VPASS之後(幾乎)立即開始增加,因為經抹除儲存元件中之 大多數(若非全部)將處於開啟狀態,甚至在施加至其相應 字線之vPASS電壓仍非常低時(在Vpass電壓之斜升(rampingU.S. Patent Application Serial No. 11/535,628, the disclosure of which is incorporated herein by reference. Various embodiments are possible with respect to timing of boosting of different channel regions. Consider the first channel region between WL0 and WLn-4, the second channel region between WLn-Ι and WLn+2, and the third channel region between WLn+4 and WLi. In a method, the first channel zone and the third channel zone are jointly raised, and thereafter the second channel zone is raised. In one method, the first channel area is raised, and thereafter the channel area and the third channel area are jointly raised. In one method, the first channel area is raised, and thereafter the third channel area is raised, and thereafter the second channel area is raised. In general, it is preferred that the second channel region be not raised prior to the third channel region because electrons from the second channel region will be attracted to the elevated second channel region, thereby reducing the elevated channel potential in the second channel region while Increase the third channel area slightly. This is an improper effect' because the reduced increase can cause program interference. Note that 'all of the above examples are for illustrative purposes only, as other bias conditions and different combinations of 131225.doc •23·200903499 bias conditions are possible. Referring again to Figure 4, it is assumed that the stylization of the storage elements along the Nand string 400 is From the staging sequence of storage element 408 to storage element 422, storage element 408-4 1 将 will have been at least partially programmed, and storage elements 42 and 422 will not yet be fully programmed. Thus, all or some of the storage elements 408·416 will have programmed and stored in their respective floating gate electronics and can erase or partially program the storage elements 420 and 422, depending on the stylized mode. For example, storage elements 420 and 422 may be partially programmed when the storage elements 42 and 422 have been previously programmed in the first step of the two-step stylization technique. In the case of EASB or REASB boost mode, VIso^ is added to one or more source side neighbors of the selected word line at a temple after the initial boost, and Vis is low enough to isolate the program in the substrate. The channel area and the erased channel area. That is, the substrate region of the substrate 490 on the source side of the isolation word line 412 is isolated from the channel region of the substrate on the drain side of the isolation word line 412. The source side can also be considered as a stylized side because the stylized association stores most or all of the π pieces, and the drain side can also be considered unstylized side 'because the associated storage elements have not been programmed yet . In addition, the pass on the source side is the first lifted region of the substrate 49 which is raised by applying v(10) to WLG and wu, and the channel region on the drain side is the substrate 49. A VpGM is applied to WL5 and a second lifted area that is lifted as a percentage is applied to WL4, WL6, and machine 7. In general, the stylized area has been improved less because the channel potential under the stylized storage element is only 〇pAss at a high enough level to open the 131225.doc -24-200903499 scripted storage element. Can start to increase (for example, improved). On the other hand, the channel potential of the storage element under erased conditions will begin to increase (almost) immediately after VPASS is applied, since most, if not all, of the erased storage elements will be on, even when applied When the vPASS voltage of the corresponding word line is still very low (the ramping of the Vpass voltage (ramping)

Up)期間)。因此,當隔離字線之沒極側上的通道區域與隔 離字線之源極側處的通道區域彼此隔離時,隔離字線之汲 極側上的通道區域與隔離字線之源極側處的通道區域相比 將經提升至較高電位。在一些實施例中,在充分地提升兩 個通道之後,將施加被施加至所選字線之程式化電壓 VpGM。 雖然上述實施例可減少特定程式干擾機制,但確實存在 /、他程式干擾機制。當vPASS相對較高時,一種其他程式 干擾失敗模式傾向於發生在較高字線上。此失敗模式發生 在’’至程式化之NAND串(例如,所選NAND串)上,且由來自 所選NAND串通道中之汲極側的熱載流子注入引起。當 VPASS達到特定位準時,此熱載流子注入由源極側通道中 之同提升電位引發。詳言之,在EASB及REASB之情況 下如所論述,藉由將隔離電壓vis〇施加於所選字線下 方之字線上而將NAND串分成源極側及汲極側。在所選 NAND串中,例如,在提升期間,沒極側通道電位將保持 於0 1 V。但,在源極側上,因為接收ναό之儲存元件被切 斷⑴如在非傳導狀態中提供,假定Viso<VTH,其中VTH 為儲存tl件之臨限電壓),所以仍提升通道。當源極側提 131225.doc -25- 200903499 升電位變高且汲極側通道電 向雷塥甘 4位保持於0v時,產生大橫 且引…Γ發至源極側上之儲存元件的熱载流子注入 越。此描繪於圖”,#中箭頭描綠跨 Γ之隔離^存元件412下之通道而移動且移動至儲存元件 之子動閉極中從而使儲存元件之臨限電壓升 一卜 0 為了防止所選NAND串中之此魅則认1』 之此頰別的程式干擾,更好的 疋在&升期間不使源極側通道與沒極側通道隔離。缺而, =無隔離之情況下’在受抑制NAND串通道中,没極側提 :將因源極側經程式化儲存元件而顯著地降低。詳言之, 當程式化高字線且源極側與汲極側通道電容比變大時"及 極側提升效率之降低可變得嚴重。為了克服此兩難問題 ⑽emma),基於源極側早期提升方案而提議通道隔離切換 方法。在此方法之情況下,隔離字線保持於相對較高電麗 VC一諸如,4 V),其足以開啟隔離儲存元件(即使隔離儲 存疋件係在最高經程式化狀態中),藉此在源極側提升期 間連接源極側通道與汲極側通道。為了進一步保證在所選 NAND串中源極側通道與汲極側通道之連接,亦可將 VC0ND施加至隔離儲存元件之汲極側上的字線,直至所選 字線打開關聯儲存元件為止,例如,因此其處於傳導狀蜞 或開啟。另外,若利用可至少部分地程式化所選儲存元件 之汲極側上之儲存元件的程式化技術,則亦可將VC⑽〇施 加至此等儲存元件以使其在源極側提升期間保持開啟。 由於源極側通道與汲極側通道經連接,所以在所選 131225.doc •26· 200903499 NAND串巾,料電位將㈣㈣·i v且料提升源極側。 結果,將消除或減少熱電子自通道之没極側至源極側的轉 移及汲極駭人干㈣型。為了保證在源極顺升時將源 極側通道與汲極側通道連接,應不遲於Vpass來施加 vC0ND。為了提供安全裕度,可在VPASS開始在源極側上斜 升之前不久來施加vCC)ND。 在源極側提升完成之後,應在汲極側提升開始之前將隔 離字線電壓降低至VIS。。以此方式,受抑制通道之沒極側 提升(在未選NAND串中)保持與源極側隔離。另外,改良 受抑制通道之提升效率,因為在源極側提升期間,汲極^ 通道中之許多電子將流動至源極側,從而在將Vmm施加 至汲極側字線之前有效地引起汲極側通道之一些提升。另 一方面,在所選NAND串中’源極側及汲極側上之通道電 位仍保持於0·1 V,且再次防止或減少汲極側注入干擾類 型。 、 圖6描繪基於圖“之自我提升模式的字線及其他電壓之 時間線。所示之時段描繪利用單一程式化脈衝之單一提升 及程式化循環。此循環通常繼之以驗證脈衝序列以判定儲 存7L件是否已達到所要程式化狀態。接著利用通常在加強 振^下之另一程式化脈衝來重複提升及程式化循環。見圖 2〇°亦注意,可選預充電時期可先於所示之時段,在預充 電蚪期中,可對汲極側通道部分地充電(預充電)(例如)丨5_ 3 V之位元線電壓,其藉由打開(在傳導狀態中提供)汲極選 擇閘極而轉移至通道。通常,在預充電期間將〇 V施加至 131225.doc -27- 200903499 Γ舉::’所選NAND串之位元線電壓不始終必須為。 /例言,所選NAND串之v-可為(例如)(M V。對於 =抑制NAND串而言,在對通道預充電之狀況下,甚至在 提升開始之前,V -Γ CH-DRA1N可南於0 V,但未必等於i 5_3 V,因為預充電之量視儲存㈣之經抹除VTH而;t。若非常 冰地抹除儲存兀件,則預充電實際上可達到…v位準。 典型預充電位準係在1 _2 v之範圍内。Up)). Therefore, when the channel region on the gate side of the isolation word line and the channel region at the source side of the isolation word line are isolated from each other, the channel region on the drain side of the isolation word line and the source side of the isolation word line The channel area will be boosted to a higher potential. In some embodiments, the programmed voltage VpGM applied to the selected word line will be applied after the two channels are sufficiently boosted. Although the above embodiment can reduce the specific program interference mechanism, there is indeed a / program interference mechanism. When vPASS is relatively high, one other program interference failure mode tends to occur on higher word lines. This failure mode occurs on the '' to the stylized NAND string (e.g., the selected NAND string) and is caused by hot carrier injection from the drain side of the selected NAND string channel. When VPASS reaches a certain level, this hot carrier injection is caused by the same boosting potential in the source side channel. In particular, in the case of EASB and REASB, the NAND string is divided into a source side and a drain side by applying an isolation voltage vis 于 to the word line below the selected word line as discussed. In the selected NAND string, for example, during boost, the gate side potential will remain at 0 1 V. However, on the source side, since the storage element receiving the ναό is cut (1) as provided in the non-conducting state, assuming that Viso<VTH, where VTH is the threshold voltage for storing tl pieces, the channel is still raised. When the source side raises 131225.doc -25-200903499, the potential of the rise becomes high and the drain side channel power is maintained at 0v, the large-horizontal and the storage element on the source side is generated. The hotter carrier injection is more. This is depicted in the figure "," in which the arrow moves over the channel under the isolation element 412 and moves to the sub-closed pole of the storage element to raise the threshold voltage of the storage element by 0. This enchantment in the NAND string recognizes that the other program of the cheek interferes with the better, and the better 疋 does not isolate the source side channel from the immersed side channel during the & liter. In the absence of isolation, there is no isolation. In the suppressed NAND string channel, the immersed side will be significantly reduced by the source side of the stylized storage element. In detail, when the high word line is programmed and the source side and drain side channel capacitance ratio changes In the case of this method, the isolation word line is kept in the case of the method of the method. In order to overcome this dilemma (10) emma), the channel isolation switching method is proposed based on the source side early promotion scheme. A relatively high current VC such as 4 V) is sufficient to open the isolated storage element (even if the isolated storage element is in the highest programmed state), thereby connecting the source side channel and the 在 during source side lift Extreme side channel. For further assurance The connection of the source side channel and the drain side channel in the selected NAND string may also apply VC0ND to the word line on the drain side of the isolation storage element until the selected word line opens the associated storage element, for example, It is in a conductive state or on. In addition, if a stylization technique that at least partially programs the storage elements on the drain side of the selected storage element is utilized, VC(10)〇 can also be applied to the storage elements to The source side channel is kept open during the boosting. Since the source side channel and the drain side channel are connected, in the selected 131225.doc •26·200903499 NAND string, the material potential will be (4)(4)·iv and the material is raised to the source side. It will eliminate or reduce the transfer of hot electrons from the non-polar side to the source side of the channel and the drain type (4). In order to ensure that the source side channel is connected to the drain side channel when the source is stepped up, vC0ND is applied later than Vpass. In order to provide a safety margin, vCC)ND may be applied shortly before VPASS begins to ramp up on the source side. After the source side boost is completed, it shall be isolated before the drain side lift begins. Word line The voltage is reduced to VIS. In this way, the dead-side side of the suppressed channel (in the unselected NAND string) remains isolated from the source side. In addition, the improved efficiency of the suppressed channel is improved because during source-side boosting Many of the electrons in the drain ^ channel will flow to the source side, effectively causing some boosting of the drain side channel before applying Vmm to the drain side word line. On the other hand, in the selected NAND string The channel potential on the source side and the drain side remains at 0·1 V, and the type of implant interference on the drain side is prevented or reduced again. Figure 6 depicts the time of the word line and other voltages based on the self-boost mode of the figure. line. The period shown depicts a single boost and stylized loop with a single stylized pulse. This loop is typically followed by a verification pulse sequence to determine if the stored 7L piece has reached the desired stylized state. The lifting and stylizing cycles are then repeated using another stylized pulse, usually under enhanced vibration. See also Figure 2 〇 ° Also note that the optional pre-charge period may precede the period shown, during the pre-charge period, the drain side channel may be partially charged (precharged) (for example) 丨 5_ 3 V bits The line voltage is transferred to the channel by opening (provided in the conduction state) the drain select gate. Typically, 〇 V is applied to 131225.doc -27- 200903499 during pre-charging:: The bit line voltage of the selected NAND string does not always have to be. / In other words, the v- of the selected NAND string can be (for example) (MV. For the = suppression NAND string, in the case of pre-charging the channel, even before the start of the upgrade, V - Γ CH-DRA1N can be south At 0 V, but not necessarily equal to i 5_3 V, because the amount of pre-charge depends on the storage (4) of the erased VTH; t. If the storage element is wiped off very cold, the pre-charge can actually reach the ...v level. Typical precharge levels are in the range of 1 _2 v.

波形800以簡化表示來描繪針對受抑制(未選)nand串之 位元線電壓VBL、對㈣娜串集合而言為共同之汲極選 擇問極電壓V⑽,及對㈣娜串集合而言為共同之源極 電左VS0URCE。實務上’ Vs〇urc—需等於V⑽及,且在 此等波形之間亦可存在時序差。波形805描繪針對所選 NAND串之位元線電壓Vbl及對於取難串集合而言為共同 之源極選擇閘極電壓Vsgs。在一替代例中,所選位元線之 VBL可具有一個以上位準。舉例而言,在快速通過寫入實 施例中,通常利用兩個位準,諸如,〇 ¥及通常為V 之較高位準。首先利用0 V來容許較快程式化,而接下來 利用較咼位準來提供幾乎已達到目標臨限電壓之經程式化 之儲存元件之臨限電壓的更精密控制。 波形810描繪施加至所選字線之汲極側上之字線的電 壓。WLi表示第丨個或最高字線,且WLn+1表示相鄰於汲極 侧上之所選字線(WLn)的字線。波形8 1 5描繪施加至所選字 線(WLn)之電壓。波形82〇描繪施加至相 鄰於源極側上之所 選字線之隔離字線(WLn—丨)的電壓。波形825描繪施加至隔 131225.doc -28- 200903499 離字線WLn-l之源極側上之字線(貿七〇至WLn_2)的電壓。 波形830及835分別描繪針對受抑制NAND串及所選NAND 串的存在於隔離字線之源極側上之基板之通道中的通道電 位(Vch-source)。波形840及845分別描繪針對受抑制NAND 串及所選NAND串的存在於隔離字線之汲極侧上之基板之 通道中的通道電位(vCH_DRA1N) α注意Vchdrain(波形84〇)如 何追蹤、/及極側k升電壓(波形8 1 〇)及程式電壓(波形8 1 5)。 程式電壓有助於汲極側提升之程度視汲極側處儲存元件之 數目而定。在汲極側處具有較少儲存元件之情況下,程式 電壓對汲極側提升之影響較大。 另外,注意,在源極側提升期間,Vchdrain(波形84〇)在 tl處稍微增加’因為汲極側通道中之電子流動至源極側, 從而在將VPASS施加至及極側字線之前有效地引起汲極侧 通道之一些提升,如先前所論述。 沿著時間線之底部的為時間點t〇_t9。詳言之,在⑺處, 如由波形綱所指示,受抑制(未選)nand串之VBL及VsGD 自0 v增加至(例如)1.5.3 V。又,VS0URCE自(例如)0·5 乂增 加至1.5 V在VSGS4〇 ν(波形8〇5)之情況下,此確保所有 NAND串之源極選擇閙摇比& βΒ ]極s保持關閉。對於所選nand串 而言,Vbl=〇(或對於快读桶讲仓λ Γ、迷通過寫入實施例而言稍高點), 使得在VSGD=1 ·5-3 V之愔π 丁 、η上Waveform 800 depicts, in a simplified representation, a bit line voltage VBL for a suppressed (unselected) nand string, a drain select voltage V(10) for a (four) nanostring set, and a (four) set of The common source is the electric left VS0URCE. In practice, 'Vs〇urc—needs equal to V(10) and there may be a timing difference between these waveforms. Waveform 805 depicts the bit line voltage Vbl for the selected NAND string and the source select gate voltage Vsgs that are common to the set of strings. In an alternative, the VBL of the selected bit line may have more than one level. For example, in a fast pass write embodiment, two levels are typically utilized, such as 〇 ¥ and a higher level, typically V. First, 0 V is used to allow for faster stylization, and then a more accurate level is used to provide more precise control of the threshold voltage of the programmed storage element that has reached the target threshold voltage. Waveform 810 depicts the voltage applied to the word lines on the drain side of the selected word line. WLi represents the second or highest word line, and WLn+1 represents the word line adjacent to the selected word line (WLn) on the drain side. Waveform 8 1 5 depicts the voltage applied to the selected word line (WLn). Waveform 82 〇 depicts the voltage applied to the isolated word line (WLn - 丨) of the selected word line on the source side. Waveform 825 depicts the voltage applied to the word line (trade 〇 to WLn_2) on the source side of word line WLn-1 from 131225.doc -28-200903499. Waveforms 830 and 835 depict channel potentials (Vch-source) in the channels of the suppressed NAND string and the selected NAND string present on the source side of the isolated word line, respectively. Waveforms 840 and 845 depict channel potentials (vCH_DRA1N) for the suppressed NAND string and the selected NAND string in the channel of the substrate on the drain side of the isolated word line, respectively. Note how Vchdrain (waveform 84〇) is tracked, / And the extreme side k-liter voltage (waveform 8 1 〇) and the program voltage (waveform 8 1 5). The extent to which the program voltage contributes to the buckle side lift depends on the number of storage elements at the drain side. In the case of fewer storage elements at the drain side, the program voltage has a greater effect on the trip side lift. In addition, note that during source side boost, Vchdrain (waveform 84〇) increases slightly at t1 because the electrons in the drain side channel flow to the source side, effectively before VPASS is applied to the pole side word line. This causes some elevation of the drain side channel, as discussed previously. Along the bottom of the timeline is the time point t〇_t9. In detail, at (7), as indicated by the waveform scheme, the VBL and VsGD of the suppressed (unselected) nand string are increased from 0 v to, for example, 1.5.3 V. Also, VS0URCE is increased from, for example, 0·5 1.5 to 1.5 V in the case of VSGS4〇 ν (waveform 8〇5), which ensures that the source selection of all NAND strings is kept off and & For the selected nand string, Vbl=〇 (or slightly higher for the fast-read bucket λ Γ, 迷 by the writing embodiment), so that VSGD=1 ·5-3 V 愔 丁, η

It况下’>及極選擇閘極打開以容許 發生程式化。雖然所提佴 圾仏之實例對應於圖5a之提升模式, 但可利用利用所選字、繞 踝之/原極側上之一或多個隔離字線的 基本上任何類型之提弁 方案。舉例而言,該實例可與區域 131225.doc -29- 200903499 自我提升(LSB)及/或經修訂LSB(RLSB)提升模式組合被利 用。在類LSB模式中,在没極側上亦可存在一或多個隔離 字線’使得鄰近所選字線之字線為Q v或其他隔離電壓, 且向剩餘未選字線供應VpAss或本文所描述之其他電壓。 R L S B類似於R E A S B。向隔離字線之緊㈣極及源極側字 線供應中間電壓VLQW,而向剩餘未選字線供應VpAssg 文所描述之其他電壓。 在ti處,將vC0ND施加至饥歧机“,使得關聯储存元 件開啟(例如,在傳導狀態中提供)。此容許在NAND串中 在隔離字線(WLn_ 1)之源極側與所選字線(WLn)线極側之 間的電荷轉移。 在t2處,藉由將VpASS施加至WL〇iWLn_2來起始源極側 通道之提升(波形825)。如所描繪,可相對於Vc〇nd而延遲 VPASS以保證在源極側提升時將源極側通道與汲極側通道 連接。通過電壓提升隔離字線之源極側上的nand串之通 道。注意vCH-S0URCE之相應增加(波形83〇)。在與在所選字 線之汲極側上在程式化序列中係在所選字線之後的冒。+1 至WLi相關聯之通道區中,歸因於諸如所施加之〇¥的電壓 而避免提升。但,可已歸因於自汲極側流動至經提升源極 側之電子而發生一些提升。在12與〇之間,發生源極側通 道之k升。在t3之後,施加vIS0以關閉隔離字線(WLn_i)之 關聯儲存元件,藉此阻止在NAND串中在隔離字線(冒1^_ 1)之源極側與所選字線(WLn)之汲極側之間的電荷轉移。 在為確js WLn- 1已達到VIS0位準所需要之延遲之後,且 131225.doc -30- 200903499 在Η處開始,藉由施加來起始沒極側通道之提升(波 形8⑼。注意VCH-DRAiN之相應增加(波形請)。源極及汲極 側通道之提升繼續,直至 且至t8為止。另外,在t5處,將VpGM1 施加至WLn,且在16處,將νρ_施加至·。因此,最初 可在第-位準且隨後在較高第二位準施加程式電壓。此方 法避免vCH_DRAIN之突然改變’其可由ν_之突然改變引 起。然而,可或者利用單一階梯式%⑽脈衝。此外注 意,在-些實施例中,VpGMi可等於,且在一此狀況 下,Η與t5之間的時間可等於零,使得υν_基本 上同%斜升。在t7處,移除程式電壓,在以處移除提升 電壓’且在t9處,提升及程式化循環結束。因此,在tl與 以之間發生源極側提升,且在“與以之間發生汲極側提 升0 歸因於源極側提升及在⑽^之間用於打開與I及 WLn-Ι相關聯之儲存元件之電壓的施加,在此時段期間在 源極側通道與汲極側通道之間可發生電荷轉移。舉例而 言,汲極側通道中之許多電子將流動至源極側,從而在將 VPASS施加至汲極側字線之前有效地引起汲極側通道之一 些提升。另外,在汲極側提升開始之前在^處乂⑶仙之移 除用來將受抑制通道之後續汲極侧提升與源極侧隔離。 圖7描繪基於圖5b之自我提升模式的字線及其他電壓之 時間線。圖7之時間線與圖ό之時間線的不同之處在於:在 所選字線WLn之汲極側上且相鄰於所選字線的字線WLn+i 在11與13之間接收\^〇1>113而非〇\^(波形812)。可(例如)在可 131225.doc -31 - 200903499 部分地程式化與WLn+1相關聯之非揮發性儲存元件時利用 此方法。另夕卜,在所選字線WLn與隔離字線WLn-2之間的 字線1在Μ與t8之間接收V_,其中V_>Vls。(波形 8i7)。此提供在-或多彳时时線上自ν_2至〜。之逐漸 轉變。接著毅形810施加至WLn+uWu,將波形82〇施 加至WLn-2,且將波形825施加至WL〇至WLn_3。In the case of '> and the pole selection gate is opened to allow for stylization. Although the example of the tampering corresponds to the boost mode of Figure 5a, substantially any type of tampering scheme utilizing one or more of the selected word, wrap/origin side or one of the isolated word lines can be utilized. For example, the example can be utilized in combination with the region 131225.doc -29-200903499 Self Lifting (LSB) and/or Revised LSB (RLSB) boost mode. In the LSB-like mode, one or more isolated word lines may also be present on the non-polar side such that the word lines adjacent to the selected word line are Qv or other isolation voltage, and VpAss or the supply is supplied to the remaining unselected word lines. Other voltages as described. R L S B is similar to R E A S B. The intermediate voltage VLQW is supplied to the tight (four) and source side word lines of the isolated word line, and the other voltages described in the VpAssg text are supplied to the remaining unselected word lines. At ti, vC0ND is applied to the hunger machine "so that the associated storage element is turned on (eg, provided in a conductive state). This allows the source side of the isolated word line (WLn_1) and the selected word in the NAND string) Charge transfer between the line side of the line (WLn). At t2, the boost of the source side channel (waveform 825) is initiated by applying VpASS to WL〇iWLn_2. As depicted, relative to Vc〇nd The VPASS is delayed to ensure that the source side channel is connected to the drain side channel when the source side is lifted. The nand string channel on the source side of the isolation word line is isolated by voltage boosting. Note that the corresponding increase of vCH-S0URCE (waveform 83) 〇). In the channel region associated with +1 to WLi in the stylized sequence on the side of the selected word line after the selected word line, due to, for example, the applied 〇¥ The voltage is prevented from increasing. However, some improvement may have occurred due to the flow from the drain side to the electrons on the boosted source side. Between 12 and 〇, the source side channel is k liters. After t3 Applying vIS0 to close the associated storage element of the isolated word line (WLn_i), thereby blocking the NAN The charge transfer between the source side of the isolated word line (1^_1) and the drain side of the selected word line (WLn) in the D string. In order to ensure that js WLn-1 has reached the VIS0 level After the delay, and 131225.doc -30- 200903499 starts at the Η, the lifting of the immersed side channel is initiated by the application (waveform 8(9). Note the corresponding increase of VCH-DRAiN (waveform please). Source and drain The lifting of the side channel continues until and until t8. In addition, at t5, VpGM1 is applied to WLn, and at 16, νρ_ is applied to ·. Therefore, initially at the first level and then at a higher level The second level applies a program voltage. This method avoids a sudden change in vCH_DRAIN 'which can be caused by a sudden change in ν_. However, a single stepped %(10) pulse can be utilized. Also note that in some embodiments, VpGMi can be equal to And in this case, the time between Η and t5 may be equal to zero, such that υν_ is substantially ramped up by %. At t7, the program voltage is removed, the boost voltage is removed at 'where, and at t9, The promotion and stylization loop ends. Therefore, the source side rise occurs between tl and And the application of the voltage of the storage element associated with the I and WLn-Ι between the (10)^ is caused by the application of the voltage between the source side and the (10)^. Charge transfer can occur between the source side channel and the drain side channel. For example, many of the electrons in the drain side channel will flow to the source side, effectively causing VPASS to be applied to the drain side word line. Some lifting of the bungee side channel. In addition, the removal of the 乂(3) 仙 before the start of the bungee side lift is used to isolate the subsequent drain side lift of the suppressed channel from the source side. Figure 7 depicts a timeline of word lines and other voltages based on the self-boost mode of Figure 5b. The time line of FIG. 7 differs from the time line of FIG. 7 in that the word line WLn+i on the drain side of the selected word line WLn and adjacent to the selected word line is received between 11 and 13\ ^〇1>113 instead of 〇\^ (waveform 812). This method can be utilized, for example, in the partial programming of non-volatile storage elements associated with WLn+1 at 131225.doc -31 - 200903499. In addition, word line 1 between the selected word line WLn and the isolated word line WLn-2 receives V_ between Μ and t8, where V_>Vls. (Waveform 8i7). This is provided on the - or multi-time line from ν_2 to ~. Gradually change. The shape 810 is then applied to WLn+uWu, the waveform 82〇 is applied to WLn-2, and the waveform 825 is applied to WL〇 to WLn_3.

亦有可能使VC0ND之位準對於其所施加至之不同字線而 變化。舉例而言,可基於相應非揮發性儲存元件之程式化 狀態來設定vCOND。vCOND在關聯非揮發性儲存元件具有較 高經程式化狀態時可較高,且在關聯非揮發性儲存元件具 有較低經程式化狀態時可較低。VC0ND僅需要足夠高以在 源極侧通道區域與汲極侧通道區域之間產生傳導路徑。提 供不同Vc〇ND位準谷許解決資料型樣依賴性之靈活性。視 後部型樣(例如,資料型樣)而定,作為一實例,WLn+1可 處於中下狀態B’(圖1 8a),而WLn及WLn下方之字線可處於 狀悲C(圖1 8 c) ’其為最高經程式化狀態。在此狀況下,可 將VC0ND_L0W&加至WLn+Ι ’且可將乂⑶仙-出即施加至WLn_ 2 至 WLn ’ 其中 Vc〇ND-HIGH〉Vc〇lSiD-LOW。 圖8描繪基於圖5C之自我提升模式的字線及其他電壓之 時間線。圖8之時間線與圖7之時間線的不同之處在於:在 所選字線WLn之源極側上且相鄰於所選字線的字線wLn-1 在t4與t8之間接收VpAss而非Vlow(波形8 1 6)。接著將波形 8 17施加至WLn-2,將波形820施加至WLn-3,且將波形825 施加至WL0至WLn-4。此提供在一或多個中間字線上自 131225.doc -32- 200903499 VPGM2至vIS0之甚至更高的逐漸轉變。 作為可利用之另一替代例’例如,當不程式化與一 相關聯之非揮發性儲存元件時,可在t_3之間將w而非 Vcond施加至 WLn+l。 圖9描綠作為圖8之時間線之替代例的字線及其他電壓之 時間線。圖9之時間線與圖8之時間線的不同之處在於:自 乂⑶㈣至後續電壓進行電壓逐漸轉變,例如,在why上 自%_至VpASS(波形9 i 2)及在机卜i上自%_至(波 形916)、在WLn上自Vc〇ND至VpG⑷(波形915),及/或在 WLn上自Vcond至VL0W(波形917)。在t3與Μ之間的時段 中’電壓因此可在源極與沒極側提升轉變之間自v_d直 接斜升或下降至VPASS或vLOW。 此方法之優勢在於:可防止或減少VIS0及/或VL0W字線 處之GIDL。在圖7及圖8之上述實例中,在施加電壓、⑽ 之前將VL0W字線下拉至〇 v。尤其與提升模型中之一些組 合,此可引起GIDL之增加。施加Vl〇w之目的為在提升期 間減少電場。然而,當Vl〇w字線上之電壓自Vc〇nd降低至〇 V時,彼子線之鄰近者中的電場歸因於經提升源極側而增 加,且可發生GIDL。可藉由使Vl〇w字線上之信號自Vc〇nd 直接斜升至VLOW來防止電場之此增加。 另外,若VL〇w>VCOND,貝彳(例如)在圖“之提升方案的情 况下’可有利的疋將vL0W而非vC0ND施加至字線,在此方 案中’將VL0W施加於WLn-4及WLn-2上,且將VIS0施加於 WLn-3上。在此狀況下,為了減少在WLn_3上發生 131225.doc -33- 200903499 GIDL(虽字線電壓自Vc〇ND轉變至V丨s〇時)或在wLn_4上發生 GIDL(知因於Vc〇nd)之機率,可較佳的是自開始就使wLn_ 4保持偏壓至。 可利用本文所論述之類似時間線來類似地實施圖&至% 之剩餘提升模式以及其他提升模式。舉例而言,在圖%之 提升模式的情況下,如所論述,可提升三個或三個以上不 同通道區。對於共同提升第一通道區及第三通道區,此後 提升第二通道區的狀況而言,可在圖6至圖9中被稱作源極 側提升之時段中提升第一通道區及第三通道區,而在被稱 作汲極側提升之時段中提升第二通道區。對於提升第一通 道區,此後共同提升第二通道區及第三通道區的狀況而 δ ’可在被稱作源極側提升之時段中提升第一通道區,而 可在被稱作汲極側提升之時段中提升第二通道區及第三通 道區。對於提升第一通道區,此後提升第三通道區此後 提升第二通道區的狀況而言,可在被稱作源極侧提升之時 段中提升第一通道區,可在被稱作源極側提升之時段之後 且在被稱作汲極側提升之時段之前的時段中提升第三通道 區,且可在被稱作汲極側提升之時段中提升第二通道區。 圖10描繪NAND串之源極側在NAND串之汲極側之前經 提升的程式化過程。結合圖8之提升方案來說明該過程, 但許多變化為可能的。程式化在步驟丨〇〇〇處開始,且在步 驟1005處選擇用於程式化之字線。源極側提升在步驟丨〇 j 〇 處開始。在步驟1〇丨5處’將VCOND設定於隔離字線(wLn_3) 至隔離字線之汲極側上之已用於程式化的最遠字線 131225.doc -34- 200903499 (Ln 1)上。在步驟ίο?。處,將vpAss設定於隔離字線之源 極側上的字線上。在步驟1〇25處,將〇 v設定於剩餘汲極 侧字線上,例如,WLn+2至WLi,且在步驟1030處,源極 側提升結束。亦即,通常,維持但不進一步提升經提升源 極側位準。在步驟1035處,汲極侧提升連同程式化一起開 始。如先前所說明,可在程式化之前起始汲極側提升。在 步驟1040處’根據所選提升模式而將電壓施加至未選字 線。在步驟1045處,將程式化脈衝施加至所選字線。汲極 侧提升及程式化脈衝在步驟1〇5〇處結束。 在步驟1055處執行驗證操作以判定所選儲存元件是否已 經程式化至所要目標臨限電壓位準,例如,Vva、Vvb或It is also possible to change the level of the VC0ND for the different word lines to which it is applied. For example, vCOND can be set based on the stylized state of the corresponding non-volatile storage element. vCOND can be higher when the associated non-volatile storage element has a higher programmed state and can be lower when the associated non-volatile storage element has a lower programmed state. The VC0ND only needs to be high enough to create a conduction path between the source side channel region and the drain side channel region. Provides flexibility for different Vc〇ND levels to resolve data type dependencies. Depending on the type of the back (eg, data type), as an example, WLn+1 can be in the lower-middle state B' (Fig. 18a), while the word lines below WLn and WLn can be in the shape of C (Figure 1). 8 c) 'It is the highest stylized state. In this case, VC0ND_L0W& can be added to WLn+Ι' and 乂(3)-- can be applied to WLn_ 2 to WLn ' where Vc 〇 ND - HIGH > Vc 〇 lSiD - LOW. Figure 8 depicts a timeline of word lines and other voltages based on the self-boost mode of Figure 5C. The time line of FIG. 8 differs from the time line of FIG. 7 in that the word line wLn-1 on the source side of the selected word line WLn and adjacent to the selected word line receives VpAss between t4 and t8. Instead of Vlow (waveform 8 1 6). Waveform 8 17 is then applied to WLn-2, waveform 820 is applied to WLn-3, and waveform 825 is applied to WL0 through WLn-4. This provides an even higher gradual transition from 131225.doc -32 - 200903499 VPGM2 to vIS0 on one or more intermediate word lines. As an alternative to being available'', for example, when not associated with a non-volatile storage element, w can be applied to WLn+1 between t_3. Figure 9 depicts green as a timeline for word lines and other voltages as an alternative to the timeline of Figure 8. The timeline of Figure 9 differs from the timeline of Figure 8 in that the voltage is gradually changed from 乂(3)(4) to the subsequent voltage, for example, on the why from %_ to VpASS (waveform 9 i 2) and on the machine i From %_ to (waveform 916), from Vc〇ND to VpG(4) (waveform 915) on WLn, and/or from Vcond to VL0W (waveform 917) on WLn. During the period between t3 and ’, the voltage can therefore ramp up or down from v_d to VPASS or vLOW between the source and the immersion side boost transition. The advantage of this method is that it prevents or reduces the GIDL at the VIS0 and/or VL0W word lines. In the above examples of Figures 7 and 8, the VL0W word line is pulled down to 〇 v before the voltage is applied, (10). This can lead to an increase in GIDL, especially in combination with some of the lifting models. The purpose of applying Vl〇w is to reduce the electric field during the boost. However, when the voltage on the Vl〇w word line is reduced from Vc〇nd to 〇 V, the electric field in the neighbor of the sub-line is increased due to the boosted source side, and GIDL can occur. This increase in electric field can be prevented by ramping the signal on the Vl〇w word line directly from Vc〇nd to VLOW. In addition, if VL 〇 w > VCOND, Bessie (for example) may advantageously apply vL0W instead of vC0ND to the word line in the case of the "improvement scheme" of the figure, in which VL0W is applied to WLn-4 And WLn-2, and VIS0 is applied to WLn-3. In this case, in order to reduce the occurrence of 131225.doc -33-200903499 GIDL on WLn_3 (although the word line voltage is changed from Vc〇ND to V丨s〇) Time) or the probability of GIDL (known as Vc〇nd) occurring on wLn_4, it may be preferable to keep wLn_4 biased from the beginning. The similar timeline discussed herein can be used to similarly implement the diagram & The remaining boost mode to % and other boost modes. For example, in the case of the graph % boost mode, as discussed, three or more different channel zones can be upgraded. The third channel region, and thereafter the condition of the second channel region is raised, the first channel region and the third channel region may be raised in the period of the source side lifting in FIGS. 6 to 9, and is referred to as Raising the second passage zone during the bungee side lift period. For lifting the first passage zone, Thereafter, the conditions of the second channel region and the third channel region are jointly raised, and δ' can raise the first channel region in a period called the source side lifting, and can be upgraded in a period called the bungee side lifting. The second channel zone and the third channel zone. For the condition that the first channel zone is raised, and thereafter the third channel zone is raised, and then the second channel zone is raised, the first channel zone can be raised in a period called the source side lifting. The third channel region may be raised in a period after the period called the source side boost and before the period called the drain side lift, and may be raised in the period called the drain side lift Channel area. Figure 10 depicts the stylized process of the source side of the NAND string before the drain side of the NAND string. This process is illustrated in conjunction with the lifting scheme of Figure 8, but many variations are possible. Stylized in steps丨The beginning begins and the word line for stylization is selected at step 1005. The source side boost begins at step 丨〇j 。. At step 1〇丨5, 'VCOND is set to the isolated word line (wLn_3) ) to the drain side of the isolated word line For the stylized farthest word line 131225.doc -34- 200903499 (Ln 1). At step ίο?, set vpAss on the word line on the source side of the isolated word line. In step 1〇25 At this point, 〇v is set on the remaining drain side word lines, for example, WLn+2 to WLi, and at step 1030, the source side boost ends. That is, generally, the raised source side position is maintained but not further boosted. The dipole side lift begins with stylization at step 1035. As explained previously, the drain side lift can be initiated prior to stylization. At step 1040, the voltage is applied to the selected boost mode. Select the word line. At step 1045, a stylized pulse is applied to the selected word line. The bungee side lift and stylized pulses end at step 1〇5〇. A verify operation is performed at step 1055 to determine if the selected storage element has been programmed to a desired target threshold voltage level, for example, Vva, Vvb, or

Vvc(圖16)。在決策區塊1060處,若針對當前字線之程式 化未完成,則在步驟1010處開始,重複額外源極側提升繼 之以汲極側提升及程式化循環。若針對當前字線之程式化 完成’但針對所有字線之程式化未完成,則在決策步驟 1065處’在步驟1〇75處選擇用於程式化之下一字線。若針 對當前字線及所有字線之程式化完成,則程式化在步驟 1070處結束。 注意’在替代實施例中,可利用字線依賴性,其中不利 用源極側提升繼之以汲極側提升之提升方案用於較低字 線,諸如’ 32字線NAND串中之WL0-WL22。確實利用源 極側提升繼之以汲極側提升之提升方案接著可用於較高字 線,諸如,WL23-WL31,其中所解決之程式干擾類型更 成問題。 I31225.doc 35- 200903499 圖11»兒明ΝΑΝΟ儲存元件(諸如,圖}及圖2所示之儲存元 件)陣列11GG之-實例。沿著每—行,位元線11()6耗合至 NAND串11 50之汲極選擇閘極的汲極端子i%。沿著 NAND串之每一列,源極線丨丨〇4可連接蕭串之源極選 擇間極的所有源極端子⑽。在美國專利第5,57G,315號、 第5’774,397號及第6,046,935號中找到作為記憶體系統之一 部分的NAND架構陣列及其操作之一實例。Vvc (Figure 16). At decision block 1060, if the programming for the current word line is not complete, then at step 1010, the additional source side boost is repeated followed by the bungee side boost and stylized loop. If the stylization for the current word line is completed 'but the stylization for all word lines is not completed, then at decision step 1065', a word line for stylization is selected at step 1 〇 75. If the stylization of the current word line and all of the word lines is completed, the stylization ends at step 1070. Note that in alternative embodiments, word line dependencies may be utilized where a source side boost followed by a drain side boosting boost scheme is used for lower word lines, such as WL0 in a '32 word line NAND string. WL22. A boost scheme that utilizes source side boost followed by a drain side boost can then be used for higher word lines, such as WL23-WL31, where the type of program disturb solved is more problematic. I31225.doc 35- 200903499 Figure 11 - Example of an array 11GG of storage elements (such as Figure} and storage elements shown in Figure 2). Along the line-by-row, bit line 11()6 is summed to the drain terminal i% of the drain select gate of NAND string 11 50. Along each column of the NAND string, the source line 丨丨〇4 can be connected to all of the source terminals (10) of the sinister source selection. An example of a NAND architecture array and its operation as part of a memory system is found in U.S. Patent Nos. 5,57,350, 315, 5,774,397, and 6,046,935.

儲存元件陣列分成大量儲存元件區塊。如對於快閃 EEPROM系統而言為常見的,區塊為抹除單元。亦即,每 一區塊包含經共同抹除的最小數目之儲存元件。每一區塊 L韦刀成π午多頁。頁為程式化單元。在一實施例中,個別 頁可分成若干片段,且片段可含有作為基本程式化操作而 經一次寫入的最少數目之儲存元件。一或多個資料頁通常 儲存於一儲存元件列中。一頁可儲存一或多個區段。區段 包括使用者資料及附加項資料。附加項資料通常包括已自 區"k之使用者資料所計算的錯誤校正碼(Ecc)。控制器(下 文所描述)之一部分在資料經程式化至陣列中時計算ECC, 且在自陣列讀取資料時亦檢查ECC。或者,將ECC及/或其 他附加項資料儲存於其所從屬之與使用者資料不同的頁乃 至不同的區塊中。 使用者資料區段通常為5 12個位元組,此對應於磁碟 驅動盗中之區段的大小。附加項資料通常為額外丨6至2〇個 位元組。大量頁形成一區塊,自8個頁(例如)直至32、64、 12 8或更多頁間的任何數量。在一些實施例中一 n 串 131225.doc -36- 200903499 列包含一區塊。 在-實關中,藉由如下方式來抹除記憶體儲存元件: =升/至抹除„(例如,14_22v)持__ ^所選一 區塊之字線接地,而源極線及位元線為浮動的。歸 於電谷耦合,未選字線、位元 , 进擇線及共同源極〇 亦:高至抹除電壓之有效部分。因此,當浮動間極 之電子通吊藉由F—穿 :時’將強電場施加至所選儲存元件之隨道氧化= 抹除所選儲存元件之資料。 電子自 '子動閘極轉移至Ρ井 區時,所選儲存元件之臨限電屋 降低。可對整個記憶體陣 、獨立區塊或另一儲存元件單元執行抹除。 圖12為利用單列/行解碼器及讀取/寫入電路之非揮發性 把憶體系統的方塊圖。該圖說明根據本發明之一實施例之 记憶體裝置1296,記憶體裝置129 G具有用於並行地讀取及 儲存元件頁之讀取/寫入電路。記憶體裝置⑽可 …或多個記憶體晶粒1298。記憶體晶粒1298包括二維 健存凡件陣列謂、控制電路121〇及讀取/寫入電路 =。在—些實施例中’儲存元件陣列可為三維的。記憶 ^陣列U00可由字線經由列解碼器咖而定址及可由位元 線經由行解碼器㈣而定址。讀取/寫入電路1265包括多 個感測區塊12〇()且容許並行地讀取或程式化儲存元件頁。 通吊控制益1250包括於與—或多個記憶體晶粒相同 之記憶體裝置1296(例如,抽取式儲存卡)中。命令及資料 係經由線路1220而在主機與控制器⑽之間轉移且經由線 131225.doc -37- 200903499 路1218而在控制器與一或多個記憶體晶粒a%之間轉移。 控制電路丨21〇與讀取/寫入電路1265協作以對記憶體陣 列1100執行記憶體操作。控制電路121〇包括狀態機1212、 晶片上位址解碼器1214、提升控制1215及功率控制模組 1216。狀態機1212提供記憶體操作之晶片級控制。晶片上 位址解碼器12 14在由主機或記憶體控制器所利用之位址至 由解碼器1230及1260所利用之硬體位址之間提供位址介 面。提升控制可用於設定提升模式,包括判定用於起 減極側及汲極職升之時序,如本文所論述。功率控制 模組1216控制在記憶體操作期間供應至字線及位元線之功 率及電壓。 在一些實施例中,可組合圖12之組件令的—些。在各種 設計η中,可將組件中除了錯存元件陣列咖以外之一或多 者(早獨或組合地)視為管理電路。舉例而言,一或多個管 理電路可包括控制電路1210、狀態機咖、解碼^ 1214/126G、功率控制模組⑵6、感測區塊丨、讀取/寫 入電路1265、控制器125〇等等中之任一者或組合。 圖為利用雙列/行解碼器及讀取/寫入電路之非揮發性 記憶體系統的方塊圖。此處,提供圖12所示之記憶體裝置 1296的另一配置。在陣列之相反侧上以對稱方式來實施由 各種周邊電路對記憶體陣和⑽之存取,使得每一側上之 存取線路及電路的密声姑+ I度減少一丰。因此,列解碼器分為列 解碼W23 0A及⑵⑽,且行解碼器分為行解碼器】鳩及 126〇Β。類似地,讀取/寫入電路分為自陣列U00之底部連 13l225.doc -38- 200903499 接至位元線的讀取/寫入電路1265A及自陣列11〇〇之頂部連 接至位元線的讀取/寫入電路丨265B。以此方式,讀取/寫 入模組之在、度基本上減少一半。如上文針對圖1 2之裝置所 描述’圖13之裝置亦可包括控制器。 圖14為描繪感測區塊之一實施例的方塊圖。個別感測區 塊1 200分成被稱作感測模組128〇之核心部分及共同部分 1290。在一實施例中,針對每一位元線將存在一獨立感測 模組1280,且針對多個感測模組128〇之集合將存在一共同 部分1290。在一實例中,一感測區塊將包括一共同部分 1290及八個感測模組128〇。一群中之感測模組中的每一者 將經由資料匯流排丨272而與關聯共同部分通信。對於其他 細節,參考2〇〇6年6月Μ日公布之標題為”N〇n_v〇latile Memory and Method with Shared Processing for an Aggregate of Sense Amplifiers"且以引用之方式全文併入本 文中的美國專利申請案公開案第2〇〇6/〇14〇〇〇7號。 感測模組1280包含感測電路127〇,感測電路127〇判定經 連接位元線中之傳導電流是高於還是低於預定臨限位準'。 感測模組12 8 0亦包括位元線鎖存器丨2 8 2,位元線鎖存器 1282用以設定經連接位元線上之電壓條件。舉例而言,鎖 存於位it線鎖存器1282中之預定狀態將導致經連接:元線 被拉至指定程式抑制之狀態(例如,1 5_3 V)。 共同部分i 290包含處理器1292、資料鎖存器集合1294, 及耦合於資料鎖存器集合1294與資料匯流排a:。之間的 I/O介面1296。處理器1292執行計算。舉例而言,其功能 131225.doc -39- 200903499 中之者為判定儲存於經残測找γ 判丨, 兑測儲存元件中之資料,且將 d疋貝科儲存於資料鎖存 用以扃π a 仔益集合中。資料鎖存器集合1294 用以在項取刼作期間儲 地理器1292所判定之資料位 几。其亦用以在程式操作期 間儲存自資料匯流排1220所導 入之貝枓位兀。所導入之資 體中之m 貝枓位兀表示思欲程式化至記憶 从 貝料。1/〇介面1296提供資料鎖存器1294與資 料匯流排1220之間的介面。 、 在讀取或感測期間,牵雄4 之彳呆作處於狀態機1 2 1 2之控制 下’狀態機1 2 12控制不同#击,丨鬥么 U衩制閘極電壓至經定址儲存元件 之供應。因為感測模組12 8 0步進诵Μ # $ < 7退通過對應於由記憶體所支 援之各種記憶體狀態的各種預 ^ w疋衩制閘極電壓,所以感測 模組1280可在此等電壓中之一去 Τ ^ 者下跳脫且將經由匯流排 1 士272而將輸出自感測模組128〇提供至處理器㈣。在彼 日守’處理器1292藉由考慮感測模組之跳脫事件及關於經由 輸入線1293而自狀態機所施加之控制閘極電壓的資訊來判 定所得記憶體狀態。其接著計算針對記憶體狀態之二進位 編碼且將所得資料位元儲存至資料鎖存器1294中。在核心 部分之另一實施例中,位元線鎖存器1282提供雙重用途, 既作為用於鎖存感測模組1 280之輸出的鎖存器且亦作為如 上文所描述之位元線鎖存器。 據預期,一些實施例將包括多個處理器1292。在一實施 例中,每一處理器1292將包括輸出線(未描繪),使得輸出 線中之每一者經共同線或(wired-OR)。在—些實施例中, 輪出線在連接至線或線之前反相。此組態使得能夠在程式 131225.doc -40· 200903499 驗證過程期間快速地判定程式化過程何時已完成,因為接 收線或之狀態機可判定經程式化之所有位元何時已達到所 要位準。舉例而言,當每一位元已達到其所要位準時,用 於彼位元之邏輯〇將發送至線或線(或資料丨反相)。當所有 位元均輸出資料0(經反相之資料lm,則狀態機知道終止 程式化過程。因為每—處理器與八個感測模組通信,所以 狀態機需要讀取線或線人次,或將邏輯添加至處理器咖 2累積關聯位元線之結果’使得狀態機僅f要讀取線或線 -次。類似地,藉由正確地選擇邏輯位準,全域狀態機可 積測第-位it何時改變其狀態且相應地改變演算法。 在程式或驗證期間,將待程式化之資料自資料匯流排 1220儲存於資料鎖存器集合1294中。在狀態機之控制下, 程式操作包含施加至經定址儲存元件之控制閉極的一連串 ,式化電屋脈衝。每一程式化脈衝繼之以讀回(驗證)以判 定儲存元件是否已經程式化至所要記憶體狀態。處理器 =292相對於所要記憶體狀態而監控所讀回之記憶體狀態。 田兩者’處理器】292設定位元線鎖存器,以便 使位元線被拉至指^程式抑制之狀態。此㈣心至 線之儲存元件進—牛叙斗„ 式程式化脈衝出現在其控 制極上。在其他實施例中,在驗證過程期間,處理器最 初載入位元線鎖存哭彳9 s 9 α ., 子1282,且感測電路將其設定為抑 值。 貧料鎖存器堆疊1294含有對應於感_組之資料鎖存号 堆疊。在-實施例中’每感測模組128時在三個資料鎖存 131225.doc •41 - 200903499 器。在—些實施例令(但並非所需 移位暫“,使得將料於Μ之::句存11實施為 料匯流排1220之串列次极,、 仃資料轉換為用於資 中,可將斜廡、 貝广,且反之亦然。在較佳實施例 資科财 具存元件之讀取/寫人區塊的所有 貝枓鎖存器聯接在一起 形成區塊移位暫存器,使得可藉 彳轉移來輸入或輸出資料區塊。詳言之,具有r個讀 取/寫入模組之組經調適成使得其資料鎖存器集合中之每 -:將資料依次移入或移出資料匯流排,> 同其為用於整 個項取/寫入區塊之移位暫存器的一部分。 可在以下各項中找到關於非揮發性儲存裝置之各種實施 例之結構及/或操作的額外資訊:〇) 2〇〇7年3月27曰發布 的標題為"Non-Volatile Memory And Method With Reduced Source Line Bias Errors"之美國專利 7,196 931 ; (2) 2〇〇6年 4月 4 日發布的標題為,,N〇n_Volatile Memory And Method with Improved Sensing”之美國專利7,〇23,736 ; (3) 2006 年 5 月 16 曰發布的標題為"improved Memory Sensing Circuit And Method For Low Voltage Operation"之美國專利 7,046,568 ; (4) 2006年 8月 5 曰公布的標題為"Compensating for Coupling During Read Operations of Non-Volatile Memory"之美國專利申請案公開案第2006/0221692號;及 (5) 2006年7月20曰公開的標題為1’Reference SenseThe array of storage elements is divided into a plurality of storage element blocks. As is common for flash EEPROM systems, the block is an erase unit. That is, each block contains a minimum number of storage elements that are collectively erased. Each block L Wei knife into π no more than one page. The page is a stylized unit. In one embodiment, individual pages may be divided into segments, and the segments may contain a minimum number of storage elements that are written once as a basic stylized operation. One or more data pages are typically stored in a storage element column. One page can store one or more sections. The section includes user data and additional items. The additional item data usually includes the error correction code (Ecc) calculated from the user data of the area. One of the controllers (described below) calculates the ECC when the data is programmed into the array, and also checks the ECC when reading data from the array. Alternatively, the ECC and/or other additional data may be stored in a different page from the user's profile than the user's profile. The user profile section is typically 5 12 bytes, which corresponds to the size of the zone in which the disk drive is stolen. The additional item data is usually an extra 6 to 2 bytes. A large number of pages form a block from any number of 8 pages (for example) up to 32, 64, 12 8 or more pages. In some embodiments an n string 131225.doc -36- 200903499 column contains a block. In the real-off, the memory storage element is erased by: = liter / to erase „ (for example, 14_22v) holding __ ^ the word line of the selected block is grounded, and the source line and the bit line The line is floating. Due to the coupling of the electric valley, the unselected word lines, bits, the selection line and the common source are also as high as the effective part of the erase voltage. Therefore, when the floating pole is electronically suspended by F - Wear: When the 'strong electric field is applied to the selected storage element, the oxidation of the selected storage element. Erasing the data of the selected storage element. When the electron is transferred from the sub-gate to the Sakai area, the selected storage element is limited. The house is lowered. The entire memory array, independent block or another storage element unit can be erased. Figure 12 is a block diagram of a non-volatile memory system using a single column/row decoder and a read/write circuit. The figure illustrates a memory device 1296 having a read/write circuit for reading and storing component pages in parallel, in accordance with an embodiment of the present invention. The memory device (10) can be ... or more Memory die 1298. Memory die 1298 includes two-dimensional memory array array, control The circuit 121 and the read/write circuit =. In some embodiments, the 'storage element array can be three-dimensional. The memory array U00 can be addressed by the word line via the column decoder and can be passed by the bit line via the row decoder. (4) Addressing. The read/write circuit 1265 includes a plurality of sensing blocks 12 () and allows reading or programming of the storage element pages in parallel. The pass control 1250 is included in - or a plurality of memory crystals In the same memory device 1296 (eg, removable memory card), the command and data are transferred between the host and the controller (10) via line 1220 and via the line 131225.doc-37-200903499 way 1218 in the controller Transferring between one or more memory grains a%. The control circuit 丨21〇 cooperates with the read/write circuit 1265 to perform a memory operation on the memory array 1100. The control circuit 121 includes a state machine 1212, a wafer The upper address decoder 1214, the boost control 1215, and the power control module 1216. The state machine 1212 provides wafer level control of the memory operation. The on-chip address decoder 12 14 is located at the address utilized by the host or the memory controller. Decoder 1 An address interface is provided between the hardware addresses utilized by 230 and 1260. The boost control can be used to set the boost mode, including determining the timing for the deceleration side and the bungee jump, as discussed herein. Power Control Module 1216 Controlling the power and voltage supplied to the word lines and bit lines during memory operation. In some embodiments, the components of Figure 12 can be combined. In various designs η, components can be excluded from the components. One or more of the array coffees (either alone or in combination) are considered as management circuits. For example, one or more management circuits may include control circuit 1210, state machine, decoding ^ 1214/126G, power control module (2) 6. Sensing block 丨, read/write circuit 1265, controller 125 〇, etc., or a combination thereof. The figure shows a block diagram of a non-volatile memory system using a dual column/row decoder and a read/write circuit. Here, another configuration of the memory device 1296 shown in Fig. 12 is provided. Access to the memory array and (10) by various peripheral circuits is performed in a symmetrical manner on the opposite side of the array such that the dense lines of the access lines and circuits on each side are reduced by a factor of one. Therefore, the column decoder is divided into column decodings W23 0A and (2) (10), and the row decoder is divided into row decoders 鸠 and 126 〇Β. Similarly, the read/write circuit is divided into a read/write circuit 1265A connected to the bit line from the bottom of the array U00, and a read/write circuit 1265A connected to the bit line from the top of the array U00. The read/write circuit 丨 265B. In this way, the degree of reading/writing of the module is substantially reduced by half. The apparatus of Fig. 13 as described above with respect to the apparatus of Fig. 12 may also include a controller. 14 is a block diagram depicting one embodiment of a sensing block. The individual sensing blocks 1 200 are divided into a core portion called a sensing module 128 and a common portion 1290. In one embodiment, an independent sensing module 1280 will be present for each bit line, and a common portion 1290 will exist for the collection of multiple sensing modules 128A. In one example, a sensing block will include a common portion 1290 and eight sensing modules 128A. Each of the sensing modules in the group will communicate with the associated common portion via the data bus 272. For other details, refer to the US Patent entitled "N〇n_v〇latile Memory and Method with Shared Processing for an Aggregate of Sense Amplifiers" published on the following day of June, 6th, and is incorporated by reference in its entirety. The application publication No. 2〇〇6/〇14〇〇〇7. The sensing module 1280 includes a sensing circuit 127〇, and the sensing circuit 127 determines whether the conduction current in the connected bit line is higher or lower. The sensing module 1280 also includes a bit line latch 丨282, and the bit line latch 1282 is used to set the voltage condition on the connected bit line. For example, The predetermined state latched in the bit it line latch 1282 will cause a connection: the element line is pulled to a specified program suppressed state (eg, 1 5_3 V). The common part i 290 includes the processor 1292, the data latch The set of devices 1294, and the I/O interface 1296 coupled between the data latch set 1294 and the data bus a: The processor 1292 performs the calculation. For example, its function 131225.doc -39- 200903499 In order to determine the γ judgment in the residual measurement, the test is stored. The data in the component, and the d 疋 疋 储存 储存 储存 储存 储存 储存 储存 储存 储存 储存 储存 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 It is also used to store the shells imported from the data bus 1220 during the program operation. The m-bed field in the imported entity indicates that the desire is stylized to memory from the bedding. 1/〇 interface 1296 provides an interface between the data latch 1294 and the data bus 1220. During the reading or sensing period, the dozer 4 is under the control of the state machine 1 2 1 2 'state machine 1 2 12 control Different #打,丨斗么U衩 gate voltage to the supply of the addressed storage element. Because the sensing module 12 8 0 step 诵Μ # $ < 7 retreat corresponds to the various memories supported by the memory The various states of the body state pre-modulate the gate voltage, so the sensing module 1280 can trip under one of the voltages and will output the self-sensing mode via the busbar 272. Group 128〇 is provided to the processor (4). In the case of the processor 1292, by considering the sensing module The off event and the information about the control gate voltage applied from the state machine via input line 1293 determine the state of the memory. It then calculates the binary encoding for the memory state and stores the resulting data bit to the data latch. In another embodiment of the core portion, bit line latch 1282 provides dual use as both a latch for latching the output of sensing module 1 280 and also as described above. The bit line latch. It is contemplated that some embodiments will include multiple processors 1292. In one embodiment, each processor 1292 will include an output line (not depicted) such that each of the output lines is wired-OR. In some embodiments, the rounds are inverted before being connected to a line or line. This configuration enables a quick determination of when the stylization process has been completed during the verification process of the program 131225.doc -40. 200903499, since the receive line or state machine can determine when all of the stylized bits have reached the desired level. For example, when each bit has reached its desired level, the logic used for that bit will be sent to the line or line (or the data is inverted). When all bits output data 0 (inverted data lm, the state machine knows to terminate the stylization process. Because each processor communicates with eight sensing modules, the state machine needs to read lines or line people, Or adding the logic to the processor 2 to accumulate the result of the associated bit line 'so that the state machine only f reads the line or line-time. Similarly, by correctly selecting the logic level, the global state machine can accumulate the first - When bit it changes its state and changes the algorithm accordingly. During program or verification, the data to be programmed is stored in data latch set 1294 from data bus 1220. Under the control of the state machine, the program operates A series of patterned electric house pulses applied to the control closed end of the addressed storage element. Each stylized pulse is then read back (verified) to determine if the storage element has been programmed to the desired memory state. 292 monitors the read back memory state relative to the desired memory state. The two 'processors' 292 set the bit line latches so that the bit lines are pulled to the state of the program suppression. (4) The heart-to-line storage component enters the syllabic pulse on its control electrode. In other embodiments, during the verification process, the processor initially loads the bit line to latch the crying 9 s 9 α Sub-1282, and the sensing circuit sets it to a value. The lean latch stack 1294 contains a data latch number stack corresponding to the sense group. In the embodiment, each sensor module 128 is Three data latches 131225.doc • 41 - 200903499. In some embodiments, (but not the required shift temporary), so that it will be implemented in the following:: sentence storage 11 is implemented as a series of material bus 1220 The secondary pole, the 仃 data is converted into the capital, and the slanting ridge, the scallop, and vice versa. In the preferred embodiment, all the shackles of the reading/writing block of the asset storage component are The registers are coupled together to form a block shift register so that the data block can be input or output by means of transfer. In particular, the group having r read/write modules is adapted such that its data lock Each in the set of registers - the data is moved into or out of the data bus in turn, > the same for the whole A portion of the shift register of the fetch/write block. Additional information regarding the structure and/or operation of various embodiments of the non-volatile storage device can be found in: 〇) 2〇〇7年3 The title of "Non-Volatile Memory And Method With Reduced Source Line Bias Errors" was published on the 27th of June, and was published on April 4, 2002. The title is N〇 U.S. Patent No. 7, 〇23, 736; (3) May 16, 2006 issued under the heading "improved Memory Sensing Circuit And Method For Low Voltage Operation" US Patent 7,046,568; (4) Published on August 5, 2006, titled "Compensating for Coupling During Read Operations of Non-Volatile Memory", US Patent Application Publication No. 2006/0221692; and (5) Title published on July 20, 2006 For 1'Reference Sense

Amplifier For Non-Volatile Memory”之美國專利申請案公 開案第20060158947號。所有五個上文剛剛所列出之專利 文件均以引用之方式全文併入本文中。 131225.doc -42· 200903499 圖1 5 e兑明針對全位元線記 m m '系構或針對奇偶記憶體架 構之圮憶體陣列至區塊之組織 ηΛΛ+ _ _ 螂妁貫例。描述記憶體陣列 之不把性結構。作為一實例,描述被分成⑽個區 塊之NAND快閃EEpR〇M。可同時抹除儲存於每一區塊中 =枓:在-實施例中,區塊為經同時抹除之最小健存元U.S. Patent Application Publication No. 20060158947 to Amplifier For Non-Volatile Memory. All of the five patent documents listed above are hereby incorporated by reference in their entirety herein. 5 e ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ ̄ An example is described as a NAND flash EEpR 〇M divided into (10) blocks. It can be erased and stored in each block at the same time = 枓: In the embodiment, the block is the smallest memory element that is erased at the same time.

仵單7G °在此實例中方各 P 夏财纟·% £塊中存在對應於位元線 :、BL1........BL8511之8,512個行。在被稱作全位元 、、卿L)架構(㈣151G)之—實施财,可在讀取及程式 操錢間同時選擇區塊之所有位元線。可同時程式化沿著 共同子線且連接至任一位元線之儲存元件。 在所提供之實例中,四個儲存元件經串聯地連接以形成 NAND串。㈣四個儲存元件經展*為包括於每—n勘 串中’但可利用多於四個或少於四個儲存元件(例如,16 個、32個、64個或另一數目)1娜串之一端子經由沒極 選擇閘極(連接至選㈣極汲極線SGD)而連接至相應位元 線’且另-端子經由源極選擇閘極(連接至選擇閘極源極 線SGS)而連接至共同源極。仵Single 7G ° In this example, there are 8 512 rows corresponding to the bit line in the block of the Xixia 纟·% £ :, BL1........BL8511. In the implementation of the all-in-one, qing L) architecture ((4) 151G), all the bit lines of the block can be selected simultaneously between reading and program money. Storage elements along a common sub-line and connected to any bit line can be programmed simultaneously. In the example provided, four storage elements are connected in series to form a NAND string. (d) Four storage elements are shown to be included in each -n survey but may utilize more than four or fewer than four storage elements (eg, 16, 32, 64 or another number) One of the series terminals is connected to the corresponding bit line ' via the gateless selection gate (connected to the selected (four) pole drain line SGD) and the other terminal is connected to the selected gate source line SGS via the source (selected gate source line SGS) And connected to a common source.

在被稱作奇偶架構(架構15〇〇)之另一實施例中,位元線 分成偶數位元線(BLe)及奇數位元線(BL〇)。在奇數/偶數位 元線架構中’沿著共同字線且連接至奇數位元線之儲存元 件在一時間經程式化,而沿著共同字線且連接至偶數位元 線之儲存元件在另一時間經程式化。在此實例中,在每一 區塊中存在被分成偶數行及奇數行之8,512個行。在此實 例中,四個儲存元件經展示為串聯地連接以形成NAND 131225.doc -43- 200903499 串。儘管四個儲存元件經展示為包括於每I·串中, 但可利用多於四個或少於四個儲存元件。 在讀取及程式化操作之一組態中’同時選擇4,256個儲 存兀件。所選擇之儲存元件具有同一字線及同一類別之位 凡線(例如,偶數或奇數)。因此,可同時讀取或程式化形 成邏輯頁之532個資料位元組’且記憶體之一區塊 至少八個邏輯頁(四個字缘,每一者 Ί14子琢 ^ 有具有奇數頁及偶數 頁)。料多狀態儲存元件而言,當每—儲存元件儲存兩 個資料位元(其中此等兩個位元中之每一者儲存於不同頁 中)時,一區塊儲存十六個邏輯頁。亦可利用其他大小之 區塊及頁。 對於ABL或奇偶架構而言,可藉由使口井升高至抹除電 壓(例如,20 V)且使所選區塊之字線接地來抹除儲存元 件。源極線及位元線為浮動的。可對整個記憶體陣列、獨 立區塊或為記憶體裝置之一部分的儲存元件之另_單元執 行抹除。電子自儲存元件之浮動閘極轉移至p井區,使得 儲存元件之VTH變為負的。 在讀取及驗證操作期間,選擇閘極(SGD及SGS)連接至 在2.5至4.5 V之範圍内的電壓,且未選字線(例如,在 為所選字線時之WLO、WL1及WL3)升高至讀取通過電壓 VrEAD(通常為在4.5至6 V之範圍内的電壓),以使電晶體作 為通過閘極而操作。所選字線WL2連接至電壓,電壓之位 準係針對每一讀取及驗證操作而被指定,以便判定相關儲 存元件之VTH是高於還是低於此位準。舉例而言,在針對 131225.doc -44 - 200903499 =準儲存元狀讀取操作巾,可使所選字線机2接地, /于偵測VTH是否高於0 V。在針對兩位準儲存元件之驗證 ,作中,所選字線WL2連接至(例如)G8 v,n驗證^ =達到至少0.8V。源極及,井處於0V。將所選位元線 (假足為偶數位元線(BLe))預充電至(例如)為〇 7 V之位準 若Vt„高於字線上之讀取或驗證位準,則與所關注儲存元 件相關聯之位元線(BLe)的電位位準由於非傳㈣存元件 而維持面位準。另一方而,\r in: ΤΗ低於麵取或驗證位準, 則相關位元線(BLe)之電位位準減小至低料,例如,小 於0.5 V,因為傳導儲存元件使位元線放電。可藉此由連 接至位元線之電Μ比較器感測放大器#測儲#元件之狀 態0 根據此項技術中已知之技術來執行上文所描述之抹除、 。貝取及驗也操作。因此,熟習此項技術者可改變所解釋之In another embodiment, referred to as an odd-even architecture (Architecture 15A), the bit lines are divided into even bit lines (BLe) and odd bit lines (BL〇). In an odd/even bit line architecture, storage elements along a common word line and connected to odd bit lines are programmed at one time, while storage elements along a common word line and connected to even bit lines are in another Stylized for a while. In this example, there are 8,512 rows divided into even rows and odd rows in each block. In this example, four storage elements are shown connected in series to form a NAND 131225.doc -43- 200903499 string. Although four storage elements are shown as being included in each I-string, more than four or fewer than four storage elements may be utilized. In the configuration of one of the read and program operations, 4,256 storage elements are simultaneously selected. The selected storage elements have the same word line and the same type of line (e.g., even or odd). Therefore, 532 data bytes of the logical page can be simultaneously read or programmed, and at least eight logical pages of one block of memory (four words, each of which has 14 pages) have odd pages and Even page). In the case of a multi-state storage element, when each storage element stores two data bits (where each of the two bits is stored in a different page), one block stores sixteen logical pages. Blocks and pages of other sizes can also be used. For an ABL or parity architecture, the storage element can be erased by raising the well to an erase voltage (e.g., 20 V) and grounding the word line of the selected block. The source line and the bit line are floating. Erasing can be performed on the entire memory array, the independent block, or another unit of the storage element that is part of the memory device. The floating gate of the electronic self-storage component is transferred to the p-well region such that the VTH of the storage component becomes negative. During read and verify operations, the select gates (SGD and SGS) are connected to voltages in the range of 2.5 to 4.5 V, and unselected word lines (eg, WLO, WL1, and WL3 when the selected word line is selected) Raising to the read pass voltage VrEAD (typically a voltage in the range of 4.5 to 6 V) to operate the transistor as a pass gate. The selected word line WL2 is coupled to a voltage that is specified for each read and verify operation to determine if the VTH of the associated memory element is above or below this level. For example, in the case of 131225.doc -44 - 200903499 = quasi-storage reading of the operating towel, the selected word line machine 2 can be grounded, / whether VTH is detected to be higher than 0 V. In the verification for the two quasi-storage elements, the selected word line WL2 is connected to, for example, G8 v, n verifying that ^ = at least 0.8V. Source and well, at 0V. Pre-charging the selected bit line (the foot line is an even bit line (BLe)) to, for example, the level of 〇7 V. If Vt„ is higher than the reading or verifying level on the word line, then The potential level of the bit line (BLe) associated with the storage element maintains the surface level due to the non-transmission (four) storage element. On the other hand, \r in: ΤΗ is lower than the face extraction or verification level, then the relevant bit line The potential level of (BLe) is reduced to a low level, for example, less than 0.5 V, because the conduction storage element discharges the bit line. This can be used by the Μ comparator sense amplifier connected to the bit line. State of the component 0 is performed in accordance with techniques known in the art to perform the erasing, decimation, and inspection operations described above. Thus, those skilled in the art can change the interpretation.

細節中的許多細節。亦可利用此項技術中已知之其他抹 除、讀取及驗證技術。 圖16描繪臨限電壓分布&單進程程式化之實例集合。針 對每-儲存it件儲存兩個資料位元之狀況來提供儲存元件 陣列之實例VTH分布。為經抹除儲存元件提供第一臨限電 壓刀布E亦七田繪經程式化儲存元件之三個臨限電壓分布 A B及C在一實施例中,£分布中之臨限電壓為負,且 A、B及C分布中之臨限電壓為正。 每相。臨限電壓範圍對應於資料4立元集合之預定值。 厶程式化至儲存元件中之資料與儲存元件之臨限電壓位準 131225.doc -45- 200903499 之間的特定關係才見針對健存元件戶斤採用t資料編碼方案而 疋牛例而s ,美國專利第6,222,762號及2004年12月16日 公開之吳國專利申請案公開案第2〇〇4/〇255〇9〇號描述用於 多狀態快閃儲存元件之各種資料編碼方案,該等案均以引 用之方式全文併入本文中。在一實施例中,利用格雷 (Gray)碼指派而將資料值指派給臨限電壓範圍,使得若浮 動:極之臨限電壓錯誤地移位至其鄰近實體狀態,則將僅 〜a個位70。一實例將"11,,指派給臨限電壓範圍E (狀態 E)、將” 10”指派給臨限電壓範圍A(狀態a)、將"〇〇"指派給 L限電壓竓圍B(狀態B),且將"〇1 "指派給臨限電壓範圍 C(狀態C)。然巾’在其他實施例中不湘格雷碼。雖然 展不四個狀態,但本發明亦可與包括彼等包括多於四個或 J於四個狀㉟之多狀態結構的其他多狀態結構—起被利 用。 一亦提供三個讀取參考電壓Vra、Vrb及Vrc以用於自儲存 -件瀆取貝料。藉由測試給定儲存元件之臨限電壓是高於 還是低於Vra'统可判定儲存元件所處之狀 態(例如,程式化條件)。 另外,提供三個驗證參考電壓Vva、Vvb&vve。當將儲 存疋件程式化至狀態A時,系統將測試彼等儲存元件是否 具有大於或等於Vva之臨限電麼。#將儲存元件程式化至 狀態_,系統將測試儲存元件是否具有大於或等於Vvb 之臨限電壓。當將儲存㈣程式化至狀態㈣,系統將判 定赌存元件是否具有其大於或等於Vve之臨限電壓。 I3l225.doc -46 - 200903499 在被稱為全序列程式化之一實施例中,可將儲存元件自 抹除狀態E直接程式化至經程式化狀態A、B*c中之任一 者。舉例而言,可首先抹除待程式化之儲存元件群體,使 得群體中之所有儲存元件皆處於經抹除狀態』。接著將利 用如由圖20之控制閘極電壓序列所描繪的一連串程式化脈 衝以將儲存元件直接程式化至狀態A、3或〇。在一些儲存 元件自狀態E經程式化至狀態辑,其他儲存元件自狀態』 經程式化至狀態B及/或自狀態E經程式化至狀態c。告在 心上自狀態』程式化至狀態c時,至ww下之相^動 開極之寄生搞合的量達到最大值,因為在乳打下之浮動問 極上的電荷量改變與在自狀態E程式化至狀態A或自狀w 程式化至狀態B時的電荷改變相比為最大的。當自狀態味 式化至狀態B時,至相鄰浮動閑極之辆合的量較少。當自 狀態E程式化至狀態a時,耗合之量甚至進—步減少。 圖⑺兒明程式化多狀態儲存元件之雙進程技術的一實 例,多狀態儲存元件儲存用於兩個不同頁之資料:下部頁 及上部頁。描繪四個狀態:狀態E⑴)、狀態A⑽、狀態 B(〇〇)及狀態C(01)。對於狀tE而言,兩個頁均儲存”1' 對於狀態A而言,下部頁錯存”〇”且上部頁儲存"〗、對於 狀態B而言’兩個頁均儲存,,〇”。對於狀態c而言 頁 儲存”1”且上部頁儲存丨丨〇丨丨。、七立 ^ ^ 只砵仔〇 。注意,儘管已將特定位元型樣 心派給狀態之每一者’但亦可指派不同位元型樣。 在第一程式化進程中,辟六— 儲存兀件之臨限電壓位準係根據 待程式化至下部邏輯頁中 4 伹兀采°又疋。若彼位元為邏輯 131225.doc -47- 200903499 ”ι” ’則臨限電壓不改變,因為其由於早先已經抹除而處 於適當狀‘%。然而,若待程式化之位元為邏輯,,〇",則儲 存元件之臨限位準增加以處於狀態A,如由箭頭17〇〇所 示。其終結第一程式化進程。 在第二程式化進程中,儲存元件之臨限電壓位準係根據 經程式化至上部邏輯頁中之位元來設^。^上部邏輯頁位 元將儲存邏輯”1”,則不發生程式化,因為儲存元件處於 狀態E或A中之-者’此視下部頁位元之程式化而定,兩 u @ ± 。若上部f位元將為邏輯 ”〇",則臨限電壓移位。若第一進程導致儲存元件保持於 經抹除狀態E,則在第二階段中,程式化儲存元件,使得 臨限電壓增加以在狀態,如由箭頭172〇所描繪。若儲 存元件已由於第一程式化進程而經程式化至狀態A,則在 第一進程中進一步程式化儲存元件,使得臨限電壓增加以 在狀態B中’如由箭頭1710所描繪。第二進程之結果為將 儲存元件程式化至經指定以針對上部頁來儲存邏輯"〇"之 狀態’而未改變下部頁之資料。在圖16及圖"中,至相鄰 字線上之浮動閘極之耦合的量視最終狀態而定。 在一實施射,若寫人足夠資料以填滿整頁,則可設置 系統以執行全序列寫入。若對於全頁而言未寫入足:資 料’則程式化過程可㈣所接收之資料來程式化下邻頁程 式化。當接收到後續資㈣’系統將接著程式化上部頁 在又一實施例中’系統可在程式化下部頁之模式中開始寫 入’且在隨後接收到足夠資料以填滿整個字線(或字線之 131225.doc -48- 200903499 大多數)之儲存元件時轉換至全序列程式化模式。2〇〇6年6 月 15 日公布的標題為"Pipelined pr〇gramming 〇f Non-Volatile Memories Using EaHy DaU"之美國 專利申 請案公 開案第2〇〇6/〇126390號中揭示此實施例之更多細節,該案 以引用之方式全文併入本文中。 圖18a至圖18c揭示用於程式化非揮發性記憶體之另 程’其藉由以下方式來減少浮動閘極至浮動閘極耦合效 應:對於任一特定儲存元件,在寫入至針對先前頁之相鄰 儲存元件之後,相對於特定頁而寫入至彼特定儲存元件。 在實例實施例中’非揮發性儲存元件利用四個資料狀態 而每儲存元件儲存兩個資料位元。舉例而言,假定狀態e 為經抹除狀態且狀態A、BAC為經程式化狀態。狀態 存資料η。狀態a儲存資料01。狀態B儲存資料1〇。狀態c 儲存資料00。此為非格雷編碼之實例,因為兩個位元均在 相鄰狀心A與B之間改變。亦可利用資料至實體資料狀態 之其他編碼。每-儲存元件儲存兩個資料頁。出於參考目 的,將此等資料頁稱為上部頁及下部頁;然而其可被給 予其他標記。關於狀態A,上部頁儲存位元〇且下部頁儲存 關於狀態B,上部頁儲存位元】且下部頁儲存位元 。關於狀態C ’兩個頁均儲存位元資料〇。 程^匕過程為兩步驟過程。在第一步驟中,程式化下部 、。右下部頁將保持資料ι,則儲存元件狀 二:貧料待程式化至。,則儲存元件之電厂堅臨限值升高, 儲存元件經程式化至狀態B,。圖制此展示儲存元 131225.doc •49· 200903499 件自狀態E至狀·% B之程式化。狀態B,為臨時狀態b ;因 此,將驗證點描繪為Vvb1,其低於\rvb。 在一實施例中,在儲存元件自狀態E經程式化至狀態B, 之後,NAND串中之其鄰近儲存元件(WLn+1)接著將相對 於其下部頁而經程式化。舉例而言,返回參看圖2,在程 式化儲存元件106之下部頁之後,將程式化儲存元件ι〇4之 下部頁。在程式化儲存元件104之後,若儲存元件ι〇4具有 自狀態E升高至狀態B’之臨限電壓,則浮動閘極至浮動閘 極耦合效應將使儲存元件1〇6之表觀臨限電壓升高。此將 具有使狀態B,之臨限電壓分布加寬至經描綠為圖⑽之臨 限電壓分布185〇之臨限電壓分布的效應。臨限電壓分布之 此表觀加寬將在程式化上部頁時得以矯正。 圖…描繪程式化上部頁之過程。若儲存元件處於經抹 除狀態E且上部頁將保持於!,則儲存元件將保持於狀態 E。若儲存元件處於狀態E且其上部頁資料待程式化至〇, 則儲存元件之臨限電壓將升高,使得儲存元件處於狀能 元件係在中間臨限電麗分布185°中且上部頁資 ^:持於1,則儲存元件將經程式化至最終狀態B。若健 存件係在中間臨限電壓分布185〇中且上 資料0 ’則儲存元件 貝厂月,餐:成 於狀能C Γ 電壓將升高,使得儲存元件處 …。由圖18a至圖18c所描繪之過程 —合效應,因為僅鄰近儲存元件之上部 將對給定傲古-^ 頁程式化 碼之-實例J 觀臨限電屋具有影響。替代狀態編 實例為在上部頁資料為1時自分布⑽移至狀態c, I31225.doc •50· 200903499 且在上部頁資料為0時移至狀態B。 儘官圖1 8a至圖1 8C提供關於四個資料狀態及兩個資料頁 之實例’但所教示之概念可應用於具有多於四個或少於四 個狀態及多於兩個或少於兩個頁之其他實施例。 圖19為描述用於程式化非揮發性記憶體之方法之一實施 例的流程圖。在一實施例中,在程式化之前抹除(以區塊 或其他單元)儲存元件。在步驟丨9〇〇中,由控制器發出"資 料載入叩令且由控制電路121〇接收輸入。在步驟 中,自控制器或主機將指定頁位址之位址資料輸入至解碼 器1214。在步驟1910中,將經定址頁之程式資料頁輸入至 負料緩衝器以用於程式化。將彼資料鎖存於適當鎖存器集 合中。在步驟1 9 1 5中,由控制器將,,程式"命令發出至狀態 機 1 2 1 2。 由’’程式”命令觸發’將利用圖20之施加至適當所選字線 之脈衝串2000的階梯式程式脈衝而將步驟丨9丨〇中所鎖存之 資料程式化至由狀態機1 2 12所控制之所選儲存元件中。在 步驟1920中,將程式電壓VpGM初始化至開始脈衝(例如, 12 V或其他值),且將由狀態機121 2所維持之程式計數器 (PC)初始化為0。在步驟1925中,施加源極提升,如先前 所論述。在步驟1930處,將第一VpGM脈衝施加至所選字線 以開始程式化與所選字線相關聯之儲存元件,且發生及極 側提升,如先前所論述。若邏輯”〇”儲存於特定資料鎖存 器中指示應程式化相應儲存元件’則使相應位元線接地。 另一方面,若邏輯"1"儲存於特定鎖存器中指示相應儲存 131225.doc -51 - 200903499 牛應保持其虽前資料狀態,則將相應位元 v以抑制程式化。 伐 在/驟1935令,驗證所選儲存元件之狀態 選:存元件之目標臨限電歷已達到適當位準,則將儲:斤 存器中之資料改變至邏輯T。若偵測到臨: ,〇 、到適备位準,則不改變儲存於相應資料鎖存3| 中之資料。以此t斗' „ ^ f益 存有邏輯,,ρ之位7 在相應資料鎖存器中儲 、铒1之位几線。當所有資料鎖存器均儲存邏輯”Ρ 時’狀態機(經由上文所描述之線或型機制)知曉所有所選 丄存凡件均已經程式化。在步驟1940中’進行關於所有資 料鎖存器是否均儲存邏輯”1”之檢查。若所有資料鎖存哭 邏輯"卜則程式化過程完成且成功,因為所有所 存兀件均已經程式化及驗證。在步驟1945中報告,,通 過狀態。在一些實施例中,即使在程式化時並未驗證所 有所選儲存元件’仍認為程式化過程完成且成功。在此狀 、。下可知因於不充分經程式化儲存元件而發生後續讀取 刼作期間之錯誤。然而,此等錯誤可由ECC校正。 ",右在步驟1940中判定並非所有資料鎖存器均儲存邏輯 ”1^’則程式化過程繼續。在—些實施例中,即使並非所 有資料鎖存器均儲存邏輯,Μ",程式過程仍停止。在步驟 mo中,對照程式極限值Pcmax來檢查程式計數器PC。程 式極限值之一實例為二十;然而,亦可利用其他數字。若 程式計數HPC不小於PCmax,則程式過程已失敗且在步驟 1955中報告”失敗"狀態。若程式計數器PC小於PCmax,則 131225.doc -52· 200903499 在步驟I960中使VpGM增加步長且使程式計數器pc遞增。 該過程接著循環回至步驟1930以施加下一 VpGM脈衝。 圖2 0描繪在程式化期間施加至非揮發性儲存元件之控制 閘極的實例脈衝串2〇〇〇,及在脈衝串期間發生之提升模式 切換。脈衝串2000包括施加至經選擇以用於程式化之字線 的一連串程式脈衝 2005、2010、2015、2020、2025、 2〇30、2035、2040、2045、2050 .......。在一實施例中, 程式化脈衝具有電壓VpGM,其開始於12 v且針對每一連續 程式化脈衝而增加增量(例如,0·5 v),直至達到為(例 如)20-25 V之最大值為止。在程式脈衝之間的為驗證脈 衝舉例而&,驗s登脈衝集合2006包括三個驗證脈衝。在 -些實施例中,針對資料所程式化至之每一狀態(例如, 狀態A、B及C)可存在一驗證脈衝。在其他實施例中,可 存在更多或更少驗證脈衝。例如,每—集合中之驗證脈衝 可具有為Vva、Vvb及Vvc(圖17)或Vvb,(圖18a)之振幅。 如所提及,當發生程式化時,例如’在程式脈衝之前或 期間,施加被施加至字線以實施提升模式之電壓。另一方 面,在(例如)發生於程式脈衝之間的驗證過程期間,不施 加提升電壓。實情為,將通常小於提升電壓之讀取電壓施 加至未選字線。讀取電壓具有—振幅,該振幅在當前經程 式化儲存元件之臨限電壓比作驗證位準時足以打開ΝΑ· 串中之先前經程式化儲存元件。 出 述0 於說明及描述之目的,已呈現本發明之 其不意欲為詳盡的或將本發明限於所揭 月述詳細描 示之精確形 131225.doc -53- 200903499 式。按照上述教示,許多修改及變化為可能的 述實施例以便最好地解釋本發明之原理及1際所每 藉此使得其他熟習此項技術者能夠最好地將本=用於以 種實施例中且在適合於所涵蓋之特定用 '各 況下最好地利用本發明。意欲藉由 > 改的情 圍來界定本發明之範嘴。 此處所附之申請專利範 【圖式簡單說明】 圖1為NAND串之俯視圖。 圖2為圖1之NAND串的等效電路圖。 圖3為NAND快閃儲存元件陣列之方塊圖。 圖4描繪展示程式干擾機制之顺仙串的橫截面圖。 圖5a至圖5h描繪自我提升模式之不同實例。 圖6描繪基於圖5a之自我提升模式的字線及其他電壓之 時間線。 圖7描繪基於圖5b之自我提升模式的字線及其他電壓之 時間線。 圖8描繪基於圖5c之自我提升模式的字線及其他電壓之 時間線。 圖9描繪作為圖8之時間線之替代例的字線及其他電壓之 時間線。 圖1 0描繪NAND串之源極側在NAND串之汲極側之前經 提升的程式化過程。 圖11為NAND快閃儲存元件陣列之方塊圖。 圖1 2為利用單列/行解碼器及讀取/寫入電路之非揮發性 131225.doc -54- 200903499 記憶體系統的方塊圖。 入電路之非揮發性 圖1 3為利用雙列/行解碼器及讀取/寫 記憶體系統的方塊圖。 圖14為描繪感測區塊之一實施例的方塊圖。 記憶體架 圖1 5說明針對全位元線記憶體架構或針對奇偶 構之s己憶體陣列至區塊之組織的實例。 圖1 6描繪臨限電壓分布及單進程程式化之實例集合。Many details in the details. Other erase, read, and verify techniques known in the art can also be utilized. Figure 16 depicts an example set of threshold voltage distribution & single process stylization. An instance VTH distribution of the array of storage elements is provided for the condition that two data bits are stored for each storage element. Providing a first threshold voltage knife E for the erased storage element and three threshold voltage distributions AB and C for the stylized storage element of the seven fields, in one embodiment, the threshold voltage in the £ distribution is negative, and The threshold voltage in the A, B, and C distributions is positive. Every phase. The threshold voltage range corresponds to a predetermined value of the data 4 epoch set. The specific relationship between the data stored in the storage component and the threshold voltage level of the storage component 131225.doc -45- 200903499 can be seen in the case of the storage component. U.S. Patent No. 6,222,762 and U.S. Patent Application Publication No. 2/4,255, filed on Dec. 16, 2004, the disclosure of which is incorporated herein by reference. The cases are hereby incorporated by reference in their entirety. In an embodiment, the data value is assigned to the threshold voltage range using a Gray code assignment such that if the floating: extreme threshold voltage is erroneously shifted to its neighboring entity state, then only ~a bits will be 70. An instance assigns "11, to the threshold voltage range E (state E), assigns "10" to the threshold voltage range A (state a), and assigns "〇〇" to the L-limit voltage range B (state B), and assign "〇1 " to the threshold voltage range C (state C). The wipes are not in the other embodiments. Although the four states are not exhibited, the present invention can also be utilized with other multi-state structures including those having more than four or six state structures of four. Three read reference voltages Vra, Vrb, and Vrc are also provided for self-storage - picking of the material. The state in which the storage element is located (e.g., stylized conditions) can be determined by testing whether the threshold voltage of a given storage element is above or below Vra'. In addition, three verification reference voltages Vva, Vvb & vve are provided. When staging storage elements to state A, the system will test if their storage elements have a power limit greater than or equal to Vva. # Program the storage component to state_, the system will test whether the storage component has a threshold voltage greater than or equal to Vvb. When the memory (4) is programmed to state (4), the system will determine if the component has its threshold voltage greater than or equal to Vve. I3l225.doc -46 - 200903499 In one embodiment referred to as full sequence stylization, the storage element can be directly programmed from the erase state E to any of the stylized states A, B*c. For example, the population of storage elements to be programmed may be erased first so that all storage elements in the population are erased. A series of stylized pulses as depicted by the control gate voltage sequence of Figure 20 will then be used to program the storage elements directly to state A, 3 or 〇. Some of the storage elements are programmed from state E to state set, and other storage elements are programmed from state to state B and/or from state E to state c. When the state is programmed from the state to the state c, the amount of parasitic fit to the bottom of the ww reaches the maximum, because the amount of charge on the floating pole of the milk is changed and the program is in the state E. The change in charge when the state A or the self-pattern w is programmed to the state B is the largest. When the state is flavored to state B, the amount of the approach to the adjacent floating idle is less. When staging from state E to state a, the amount of consumption is even reduced. Figure (7) shows an example of a two-process technique for stylized multi-state storage elements that store data for two different pages: the lower page and the upper page. Four states are depicted: state E(1)), state A(10), state B(〇〇), and state C(01). For the case tE, both pages store "1". For state A, the lower page is "〇" and the upper page is stored ", for state B, 'two pages are stored, 〇" . For status c, the page stores "1" and the upper page stores 丨丨〇丨丨. , 七立 ^ ^ Only 砵仔〇. Note that although a particular bit pattern has been assigned to each of the states', different bit patterns can also be assigned. In the first stylization process, the threshold voltage level of the six-storage component is based on the program to be converted to the lower logical page. If the bit is logic 131225.doc -47- 200903499 "ι" ’, the threshold voltage does not change because it is in the proper shape ‘% because it has been erased earlier. However, if the bit to be programmed is logical, 〇", then the threshold level of the storage element is increased to be in state A, as indicated by arrow 17〇〇. It ends the first stylization process. In the second stylization process, the threshold voltage level of the storage element is set according to the bits that are programmed into the upper logical page. ^The upper logical page bit will store the logic "1", then no stylization will occur, because the storage element is in state E or A - this depends on the stylization of the lower page bit, two u @ ± . If the upper f-bit will be a logical "〇", then the threshold voltage shifts. If the first process causes the storage element to remain in the erased state E, then in the second phase, the storage element is programmed to make the threshold The voltage is increased in the state as depicted by arrow 172. If the storage element has been programmed to state A due to the first stylization process, the storage element is further stylized in the first process, causing the threshold voltage to increase In state B, 'as depicted by arrow 1710. The result of the second process is to program the storage element to the state specified to store the logical "〇" for the upper page without changing the information of the lower page. In Figure 16 and Figure, the amount of coupling to the floating gates on adjacent word lines depends on the final state. In one implementation, if enough data is written to fill the entire page, the system can be set to execute the full Sequence write. If the full page is not written to the full: data 'the stylization process can (4) the received data to program the next page stylized. When receiving the follow-up capital (four) 'the system will then program the upper page In yet another In the example, the system can start writing in the mode of the programmed lower page and then receive enough data to fill the storage elements of the entire word line (or 131225.doc -48-200903499 most of the word line). Switch to full-serial stylized mode. The US Patent Application Publication No. 2〇〇6/, published on June 15, 2002, is titled "Pipelined pr〇gramming 〇f Non-Volatile Memories Using EaHy DaU" Further details of this embodiment are disclosed in pp. 126,390, which is hereby incorporated by reference in its entirety in its entirety. Reducing the floating gate to floating gate coupling effect: for any particular storage element, after writing to an adjacent storage element for a previous page, writing to a particular storage element relative to a particular page. In an example embodiment A non-volatile storage element utilizes four data states and stores two data bits per storage element. For example, assume that state e is erased and states A and BAC are stylized. Store data η. State a stores data 01. State B stores data 1 状态 state c stores data 00. This is an example of non-Gray coding because both bits change between adjacent centers A and B. Other codes of data can be used to the status of the physical data. Each storage element stores two data pages. For reference purposes, these data pages are referred to as upper and lower pages; however, they can be given other marks. The upper page stores the bit 〇 and the lower page stores the state B, the upper page stores the bit 】 and the lower page stores the bit. For the state C ' both pages store the bit 〇. The process is a two-step process . In the first step, stylize the lower part. The lower right page will keep the data ι, then store the component shape. 2: The poor material is to be programmed. , the power plant threshold of the storage component is increased, and the storage component is programmed to state B. Figure shows the storage unit 131225.doc •49· 200903499 Stylized from state E to shape·% B. State B, which is the temporary state b; therefore, the verification point is depicted as Vvb1, which is lower than \rvb. In one embodiment, after the storage element is programmed from state E to state B, its neighboring storage elements (WLn+1) in the NAND string will then be programmed relative to its lower page. For example, referring back to Figure 2, after the lower page of the storage element 106 is programmed, the lower page of the storage element ι4 will be programmed. After staging the storage element 104, if the storage element ι4 has a threshold voltage that rises from state E to state B', the floating gate-to-floating gate coupling effect will cause the storage element 1〇6 to look The voltage limit is increased. This will have the effect of widening the threshold voltage distribution of state B to the threshold voltage distribution of the threshold voltage distribution 185 of the graph (10). This apparent widening of the threshold voltage distribution will be corrected when the upper page is programmed. Figure... depicts the process of stylizing the upper page. If the storage element is in erased state E and the upper page will remain at! , the storage element will remain in state E. If the storage element is in state E and its upper page data is to be programmed to 〇, then the threshold voltage of the storage element will rise so that the storage element is in the middle of the energy distribution 185° and the upper page is ^: Hold at 1, the storage element will be programmed to final state B. If the health component is in the middle threshold voltage distribution 185〇 and the data 0 ’ is stored, the component is stored in the factory. The meal is in the form of C Γ The voltage will rise so that the storage component is located. The process-synthesis effect depicted by Figures 18a through 18c, since only the upper portion of the storage element will have an effect on the given instance of the code. An alternative state is exemplified by moving from distribution (10) to state c, I31225.doc •50·200903499 when the upper page data is 1, and moving to state B when the upper page data is zero. Figure 1 8a to Figure 1 8C provide examples of four data states and two data pages 'but the concepts taught can be applied to have more than four or less than four states and more than two or less Other embodiments of two pages. Figure 19 is a flow chart depicting one embodiment of a method for staging non-volatile memory. In one embodiment, the components are erased (in blocks or other units) prior to programming. In step 〇〇9〇〇, the controller issues a "data loading command and the input is received by the control circuit 121A. In the step, the address data of the specified page address is input to the decoder 1214 from the controller or the host. In step 1910, the program data page of the addressed page is entered into the negative buffer for stylization. The data is latched into the appropriate set of latches. In step 1 915, the controller sends the program " command to the state machine 1 2 1 2 . Triggered by the ''Program' command'' will use the stepped program pulse of pulse train 2000 applied to the appropriate selected word line of FIG. 20 to program the data latched in step 丨〇9丨〇 to state machine 1 2 In the selected storage element controlled by 12. In step 1920, the program voltage VpGM is initialized to a start pulse (e.g., 12 V or other value), and the program counter (PC) maintained by the state machine 121 2 is initialized to 0. In step 1925, source boosting is applied, as previously discussed. At step 1930, a first VpGM pulse is applied to the selected word line to begin programming the storage elements associated with the selected word line, and The pole side boost, as previously discussed. If the logic "〇" is stored in a particular data latch indicating that the corresponding storage element should be programmed, then the corresponding bit line is grounded. On the other hand, if the logic "1" is stored in In the specific latch, indicate the corresponding storage 131225.doc -51 - 200903499 The cow should maintain its pre-data status, then the corresponding bit v will be suppressed to stylize. In the 1935 order, verify the status of the selected storage element. Save If the target's target power limit has reached the appropriate level, the data in the memory: memory will be changed to logic T. If the detection of Pro:, 〇, to the appropriate level, then the data will not be stored in the corresponding data. The data in the latch 3|. With this t bucket ' „ ^ f beneficial logic, ρ bit 7 is stored in the corresponding data latch, 铒 1 bit line. When all data latches store a logical "Ρ" state machine (via the line or type mechanism described above), it is known that all selected memory items have been programmed. In step 1940, 'on all data locks' Whether the memory stores a check of logic "1". If all data is latched, the logic is "successful" and the program is successfully completed, because all the stored conditions have been programmed and verified. In step 1945, the report is passed. State. In some embodiments, even if all selected storage elements are not verified during stylization, the stylization process is considered to be complete and successful. In this case, it is known that subsequent operations are caused by insufficiently programmed storage elements. Errors during reading are read. However, such errors can be corrected by ECC. " Right In step 1940, it is determined that not all data latches store logic "1^" and the stylization process continues. In some embodiments, the program process stops even if not all of the data latches store logic. In step mo, the program counter PC is checked against the program limit value Pcmax. An example of a process limit value is twenty; however, other numbers may also be utilized. If the program count HPC is not less than PCmax, the program process has failed and the "failure" status is reported in step 1955. If the program counter PC is less than PCmax, 131225.doc -52· 200903499 increments VpGM in step I960 and The program counter pc is incremented. The process then loops back to step 1930 to apply the next VpGM pulse. Figure 20 depicts an example pulse train 2〇〇〇 applied to the control gate of the non-volatile storage element during stylization, and The boost mode switching occurs during the burst. The burst 2000 includes a series of program pulses 2005, 2010, 2015, 2020, 2025, 2〇30, 2035, 2040, 2045 applied to the word line selected for programming. 2050 . . . In one embodiment, the stylized pulse has a voltage VpGM that begins at 12 v and increments (eg, 0.5 volts) for each successive stylized pulse until it reaches For example, the maximum value of 20-25 V. Between the program pulses is an example of a verification pulse & the s pulse collection 2006 includes three verification pulses. In some embodiments, the data is programmed. Chemical There may be one verify pulse for each of the states (eg, states A, B, and C.) In other embodiments, there may be more or fewer verify pulses. For example, the verify pulse in each set may have Vva, The amplitude of Vvb and Vvc (Fig. 17) or Vvb, (Fig. 18a). As mentioned, when stylization occurs, for example, 'before or during the program pulse, a voltage applied to the word line to implement the boost mode is applied. On the other hand, during the verification process, for example, occurring between program pulses, no boost voltage is applied. In other words, a read voltage, typically less than the boost voltage, is applied to the unselected word line. The read voltage has an amplitude. The amplitude is sufficient to open the previously stylized storage element in the string when the threshold voltage of the current programmed storage element is compared to the verification level. Describing 0 For the purposes of illustration and description, the present invention is not intended In order to be exhaustive or to limit the invention to the precise description of the detailed description of the disclosure, the descriptions of the above-described teachings, many modifications and variations are possible in the preferred embodiments. The principles of the present invention and the use of the present invention will enable others skilled in the art to best utilize the present invention in the embodiments and in the particular application. The present invention is intended to define the scope of the present invention by the scope of the present invention. The application of the invention is hereby incorporated by reference. FIG. 1 is a top view of a NAND string. FIG. 2 is a NAND string of FIG. Equivalent Circuit Diagram Figure 3 is a block diagram of an array of NAND flash memory elements. Figure 4 depicts a cross-sectional view of the sinister string showing the program interference mechanism. Figures 5a through 5h depict different examples of self-elevation modes. Figure 6 depicts a timeline of word lines and other voltages based on the self-boost mode of Figure 5a. Figure 7 depicts a timeline of word lines and other voltages based on the self-boost mode of Figure 5b. Figure 8 depicts a timeline of word lines and other voltages based on the self-boost mode of Figure 5c. Figure 9 depicts a timeline of word lines and other voltages as an alternative to the timeline of Figure 8. Figure 10 depicts the stylized progression of the source side of the NAND string before the drain side of the NAND string. Figure 11 is a block diagram of an array of NAND flash memory elements. Figure 12 is a block diagram of a non-volatile 131225.doc-54-200903499 memory system utilizing a single column/row decoder and a read/write circuit. Non-volatile into the circuit Figure 13 is a block diagram of a dual column/row decoder and a read/write memory system. 14 is a block diagram depicting one embodiment of a sensing block. Memory Rack Figure 15 illustrates an example of the organization of a full bit line memory architecture or a block of suffix arrays to blocks of parity. Figure 16 depicts a set of examples of threshold voltage distribution and single-process stylization.

圖1 7描繪臨限電壓分布及雙進程程式化之實例集合。 圖18a至圖l8c展示各種臨限電壓分布,且描述用於 化非揮發性記憶體之過程。 工 圖19為描述用於程式化非揮發性記憶體之過程 AA_ '^施 圖20描繪在程式化期間施加 閘極的實例脈衝串。 至非揮發性儲存元件之控制 【主要元件符號說明】 100 電晶體 100CG 控制閘極 100FG 浮動閘極 102 電晶體 102CG 控制閘極 102FG 浮動閘極 104 電晶體 104CG 控制閘極 104FG 浮動閘極 131225.doc 55- 200903499 106 電晶體 106CG 控制閘極 106FG 浮動閘極 120 選擇閘極 120CG 控制閘極 122 選擇閘極 122CG 控制閘極 126 位元線 128 源極線 320 NAND 串 321 位元線 322 選擇閘極 323 ' 324 、 325、 326 儲存元件 327 選擇閘極 340 NAND 串 341 位元線 342 選擇閘極 343 ' 344 、 .345、 346 儲存元件 347 選擇閘極 360 NAND 串 361 位元線 362 選擇閘極 363 、 364 ' * 365 、366 儲存元件 367 選擇閘極 -56- 131225.doc 200903499 400 404 406 408 、 410 、 412 、 414 、 416 ' 418 、 420 、 422 424 426 800 810 815 820 825 830 835 840 845 1100 1104 1106 1126 1128 1150 1200 1210 NAND 串 源極供應線 選擇閘極 儲存元件 選擇閘極 位元線 波形 波形 波形 波形 波形 波形 波形 波形 波形 NAND儲存元件陣列 源極線 位元線 汲極端子 源極端子 NAND 串 感測區塊 控制電路 131225.doc -57- 200903499 1212 狀態機 1214 晶片上位址解碼 1215 提升控制 1216 功率控制模組 1218 線路 1220 貧料匯流排 1250 控制器 1260 解碼器 1260A 行解碼器 1260B 行解碼器 1265 讀取/寫入電路 1265A 讀取/寫入電路 1265B 讀取/寫入電路 1270 感測電路 1272 匯流排 1280 感測模組 1282 位元線鎖存器 1290 共同部分 1292 處理器 1293 輸入線 1294 資料鎖存器 1296 記憶體裝置 1298 記憶體晶粒 1500 奇偶架構 131225.doc -58 - 200903499Figure 17 depicts an example set of threshold voltage distribution and two-process stylization. Figures 18a through 18c show various threshold voltage distributions and describe the process for non-volatile memory. Figure 19 is a diagram depicting a process for staging non-volatile memory. AA_' Figure 20 depicts an example pulse train that applies a gate during stylization. Control to non-volatile storage components [Main component symbol description] 100 transistor 100CG control gate 100FG floating gate 102 transistor 102CG control gate 102FG floating gate 104 transistor 104CG control gate 104FG floating gate 131225.doc 55- 200903499 106 Transistor 106CG Control Gate 106FG Floating Gate 120 Select Gate 120CG Control Gate 122 Select Gate 122CG Control Gate 126 Bit Line 128 Source Line 320 NAND String 321 Bit Line 322 Select Gate 323 '324, 325, 326 storage element 327 select gate 340 NAND string 341 bit line 342 select gate 343 ' 344 , .345, 346 storage element 347 select gate 360 NAND string 361 bit line 362 select gate 363 364 ' * 365 , 366 storage element 367 selection gate - 56 - 131225.doc 200903499 400 404 406 408 , 410 , 412 , 414 , 416 ' 418 , 420 , 422 424 426 800 810 815 820 825 830 835 840 845 1100 1104 1106 1126 1128 1150 1200 1210 NAND string source supply line selection gate storage Piece selection gate bit line waveform waveform waveform waveform waveform waveform waveform NAND storage element array source line bit line 汲 terminal source terminal NAND string sensing block control circuit 131225.doc -57- 200903499 1212 state machine 1214 on-chip address decoding 1215 boost control 1216 power control module 1218 line 1220 lean bus 1250 controller 1260 decoder 1260A row decoder 1260B row decoder 1265 read / write circuit 1265A read / write circuit 1265B read Fetch/write circuit 1270 sense circuit 1272 bus bar 1280 sense module 1282 bit line latch 1290 common part 1292 processor 1293 input line 1294 data latch 1296 memory device 1298 memory die 1500 parity structure 131225.doc -58 - 200903499

1700 1710 1720 1850 2000 2005 ' 2010 2020 ' 2025 2035 、 2040 2050 2006 tO、tl、t2、 t5、t6、t7、 A B B' BLe BLo BLO、BL1、 BL3、BL4、 BL8511 C E SGD SGS 箭頭 箭頭 箭頭 臨限電壓分布 脈衝串 、2015、 程式脈衝 ' 2030 、 、2045 、 驗證脈衝集合 t3、t4、 時間點 t8 ' t9 狀態 狀態 狀態 偶數位元線 奇數位元線 BL2 ' 位元線 BL5、 狀態 狀態 選擇問極汲極線 選擇閘極源極線 131225.doc -59- 200903499 V bl 位元線電壓 V CH-DRA1N 通道電位 V CH-SOURCE 通道電位 V COND 電壓 V ISO 隔離電壓 Vlow 中間電壓 VpASS 通過電塵 V PASS-HIGH 高通過電壓 V PASS-LOW 低通過電壓 V PASS-MEDIUM 中間通過電壓 VpGM 程式電壓 VpGM 1 程式電塵 VpGM2 程式電壓 Vra 讀取參考電壓 Vrb 讀取參考電壓 Vrc 讀取參考電壓 VsGD 汲極選擇閘極電壓 VsGS 源極選擇閘極電壓 V SOURCE 源極偏壓 V th 臨限電壓 Vva 驗證參考電壓 Vvb 驗證參考電壓 Vvb' 驗證點 Vvc 驗證參考電壓 131225.doc -60- 2009034991700 1710 1720 1850 2000 2005 ' 2010 2020 ' 2025 2035 , 2040 2050 2006 tO, tl, t2, t5, t6, t7, ABB' BLe BLo BLO, BL1, BL3, BL4, BL8511 CE SGD SGS arrow arrow arrow threshold voltage Distributed pulse train, 2015, program pulse '2030, , 2045, verification pulse set t3, t4, time point t8 't9 state state state even bit line odd bit line BL2 'bit line BL5, state state selection Pole line selection gate source line 131225.doc -59- 200903499 V bl bit line voltage V CH-DRA1N channel potential V CH-SOURCE channel potential V COND voltage V ISO isolation voltage Vlow intermediate voltage VpASS through electric dust V PASS- HIGH high pass voltage V PASS-LOW low pass voltage V PASS-MEDIUM intermediate pass voltage VpGM program voltage VpGM 1 program electric dust VpGM2 program voltage Vra read reference voltage Vrb read reference voltage Vrc read reference voltage VsGD drain select gate Voltage VsGS Source Select Gate Voltage V SOURCE Source Bias V th Threshold Voltage Vva Verify Reference Voltage Vvb Verify Reference Voltage Vvb' Test point Vvc verification reference voltage 131225.doc -60- 200903499

Wli 字線 WLO、WL1、 WL2、 字線 WL3、WL4、 WL5、 WL6、WL7、 WL63 WLn 所選字線 WLn-1 字線 WLn-2 字線 WLn-3 字線 WLn-4 字線 WLn-5 字線 WLn-6 字線 WLn-7 字線 WLn-8 字線 WLn+1 字線 WLn+2 字線 WLn+3 字線 WLn+4 字線 131225.doc -61 -Wli Word Lines WLO, WL1, WL2, Word Lines WL3, WL4, WL5, WL6, WL7, WL63 WLn Selected Word Line WLn-1 Word Line WLn-2 Word Line WLn-3 Word Line WLn-4 Word Line WLn-5 Word Line WLn-6 Word Line WLn-7 Word Line WLn-8 Word Line WLn+1 Word Line WLn+2 Word Line WLn+3 Word Line WLn+4 Word Line 131225.doc -61 -

Claims (1)

200903499 十、申請專利範園: 1. 一種用於操作非揮發性儲存器之方法,包含: 在一第二字線之一汲極側上提升至少一 NAND串之前 在一第一字線之一源極側上執行該至少一 nand串之第 -提升’該第二字線係在該第—字線之—汲極側上,包 括省第一字線及該第二字線之複數個字線與該至少— nand串相關聯,且該至少—NAND串具有複數個非揮發 性儲存元件; 在該第一提升期間,將一電壓施加至該第一字線以用 於提供該複數個非揮發性儲存元件中與該第一字線相關 聯之在-傳導狀態中的一第一非揮發性儲存元件,且將 一電壓施加至該第二字線以用於提供該複數個非揮發性 儲存7L件中與該第二字線相關聯之在一傳導狀態中的一 苐一非揮發性儲存元件;及 在該第一提升之後在該第二字線之該汲極側上執行該 至v NAND串之第二提升,同時將_電麼施加至該第 一字線以用於提供在—非傳導狀態中之該第—非揮發性 儲存元件,1同時將一冑式電壓施加至該第二字線。 2.如請求項1之方法,其中: 在該第二提升期間’將在一第一位準之一電壓施加至 該複數個字線中在該第:字線之㈣極側上的字線,施 加至該第-字線之該電壓係在一小於該第一位準之第二 位準’且將在-大於該第二位準之位準的㈣施加至在 該第一字線與該第二字線之間的至少—中間字線。 131225.doc 200903499 3.如q求項〗之方法,進一步包含: 在提升期:,將一電遷施加至該複數個字線中 子線與该第二字線之間的至少-字線以用於提 =焚數個非揮發性儲存元件中在該第—字線與該第二 子線之間的在一傳導狀態中 非輝發性儲存元 件心^該至少_财輯中在該第—非揮發性儲存元 二非揮發性儲存元件之間的每—非揮發性儲存 几件係在一傳導狀態中被提供。 4.如請求項丨之方法,進一步包含: 5. =第一提升期間’將一電壓施加至該複數個字線中 一二第-子線之該汲極側上之相鄰於該第二字線的至少 子線,以用於提供該複數個非揮發性儲存元件中在該 第二字線之該汲極側上之相鄰於該第二導 狀態中之至少一非揮發性儲存元件。 傳導 如請求項1之方法,進_步包含: =弟-提升期間’將—㈣施加至該複數個字線中 =第:字線之該汲極侧上的一字線集合以用於避免在 子線之㈣極側上提升該至少-nand串。 6.如請求項丨之方法,進—步包含: 在=第—提升期間,將-電屋施加至該複數個字線中 在該第二字線之該汲極側上的至 複數個非揮發性儲存線於提供該 在一傳導狀態令之至少—第沒極上的 電^加至該複數個字線中在該至少一第三非揮發性 131225.doc 200903499 :存元件之-沒極側上的—字線集合,以用於避 X:非揮發性儲存元件之-及極側上提升該二 7·如請求項1之方法’進一步包含: 之二提升期間,將—電壓施加至該複數個字線中 線中在該第—字線與該第二字線之間的至少一 者以用於在該第一字後盥 财助串。 —線之間提升該至少一 8·如:求項1之方法,進-步包含,在該第二提升期間: 中t壓施加至—額外字線以用於提供在—非 態::一額外非揮發性儲存元件,該額外字線係在該第 = 沒極側上’該第二提升係對該至少一取ND 串在該第三字線與該額外字線之間的 額外字線一側上執行該至少=二之 弟二升。 9. 一種非揮發性儲存系統,包含·· 至夕NAND串,其具有複數個非揮發性儲存元件; 複數個字線,其與該至少—NAND串進行通信;及’ 一或多個控制電路,其與該複數個字線進行通信1 =多個控制電路:⑷在-第二字線之1極側上提: :=;、—NAND串之前在-第-字線之-源極側上執行 串之第一提升,該第二字線係在該第一 :至及極侧上;⑻在該第一提升期間,將一㈣施 加…-字線以提供該複數個非揮發性儲存元件中與 131225.doc 200903499 該第一字線相關聯之在—傳導狀態中@ 一第一非揮發性 儲存元件,且將—電壓施加至該第二字線以提供該複數 個非揮發性儲存元件中與該第二字線相關聯之在一傳導 狀態中的-第二非揮發性儲存元件;及⑷在該第一提升 之後在該第二字線之該沒極側上執行該至少一财仙串 之第二提升,同時將一電壓施加至該第一字線以提供在 一非傳導狀態中之該第—非揮發性儲存元件,且同時將 —程式電壓施加至該第二字線。 10. 如請求項9之非揮發性儲存系統’其中: ▲在該第二提升期間,將在一第一位準之一電壓施加至 /複數個子線中在该第二字線之該沒極側上的字線,施 加至該第一字線之該電壓係在一小於該第一位準之第二 2準a將在-大於該第二位準之位準的電壓施加至在 X第予線與β亥第二字線之間的至少一中間字線。 11. 如明求項9之非揮發性儲存系統,其中: 在該第一提升期間,該一或多個控制電路將—電壓施 加至該複數個字線中在該第二字線之該沒極側上的一字 線集合以用於避Φ为包r贫_ + 在〜第一子線之該汲極側上提 少一 NAND串。 ,-Η 非琿發性儲存…—,、, 靶加至該第二字線以提供在一傳導狀態中之該第二非 揮發性健存元件的該電麼係在—第_位準;且 一該一或多個控制電路藉由將在-大於該第一位準之第 -位準的一電壓施加至該複數個字線中在該第—字線之 13I225.doc 200903499 ”亥源極側上的一字線集合來執行該第一提升。 13.如請求項9之非揮發性儲存系統,其中: 知加至„亥第一字線以提供在一傳導狀態中之該第 揮發性儲存元件的該電壓係在—第一位準;且 該-或多個控制電路藉由將在一大於該第一位準之第 -位準的電壓施加該至少一NAND串中在該第—非揮發 :儲存70件之該源極側上的-非揮發性儲存元件集合來 執行該第一提升。 14·如請求項13之非揮純儲存系統,其中: 在執仃4第-提升時,該一或多個控制電路將在一小 於該第一位準之笛:r >,、住, 弟二位準的電壓施加至該複數個非揮發 欧儲存7L件中在該第二非揮發性儲存元件之—汲極側上 的一非揮發性儲存元件集合。 1 5.如》月求項I4之非揮發性儲存系統,其中: _在執仃該第—提升時,該一或多個控制電路將在該第 二位準之電壓施加至該至少一_串中在該第二非揮 發性儲存元件之該沒極側上的該非揮發性儲存元件集 合。 16· 一種非揮發性儲存系統,包含: 用於執行第—升壓之構件’其在―第二字線之一沒極 則上提升至少-N A N D串之前在—第_字線之—源極側 上執行該至少—NAND串之第—提升,該第二字線係在 :第-字線之一汲極側上,包括該第—字線及該第二字 、之複數個字線與該至少串相關聯,且該至少 131225.doc 200903499 NAND串具有複數個非揮發性錯存元件; 用於提供傳導狀態中非揮發件 該第-提升期間,將一電愿施加至該第一字:件,其在 供該複數個非揮發性儲存 1用於提 在一傳導狀能由M ^ 干甲/、該第一子線相關聯 導以中的-第-非揮發性儲存元件,且將一電 壓轭加至該第二字線以用 電 凡件中與該第二字線相關胪+ + ^ 储存 ^予深相關聯之在一傳導狀態中的—筮一 非揮發性儲存元件;及 弟一 用於執行第二升壓之構件,i 筮_〜 ,、在°亥第—獒升之後在該 弟-予線之該汲極側上執行該至少—财仙串之第二提 升,同時將一電壓施加 主該第—予線以用於提供在一非 傳導狀態中之該第一非揮發性儲存元件,且同時將一程 式電壓施加至該第二字線。 131225.doc200903499 X. Patent Application: 1. A method for operating a non-volatile memory, comprising: one of a first word line before lifting at least one NAND string on one of the second word lines Performing the first-upward of the at least one nand string on the source side, the second word line is on the drain side of the first word line, and includes a plurality of words of the first word line and the second word line a line associated with the at least nand string, and the at least - NAND string having a plurality of non-volatile storage elements; during the first boost, applying a voltage to the first word line for providing the plurality of non- a first non-volatile storage element in a conductive state associated with the first word line in the volatile storage element, and applying a voltage to the second word line for providing the plurality of non-volatile Storing a non-volatile storage element in a conductive state associated with the second word line in a 7L piece; and performing the up to the drain side of the second word line after the first boost v The second boost of the NAND string, while applying _ A word line is provided for providing the first non-volatile storage element in a non-conducting state, 1 simultaneously applying a voltage of the 字 to the second word line. 2. The method of claim 1, wherein: during the second boosting period, a voltage at a first level is applied to a word line on a (four)-pole side of the first word line in the plurality of word lines The voltage applied to the first word line is applied to a second level less than the first level and (4) at a level greater than the second level is applied to the first word line At least the middle word line between the second word lines. 131225.doc 200903499 3. The method of claim 9, further comprising: during the promotion period: applying an electromigration to at least the word line between the sub-line and the second word line of the plurality of word lines For extracting a plurality of non-volatile storage elements in a conductive state between the first word line and the second sub-line in a conductive state, at least in the - Each of the non-volatile storage elements between the non-volatile storage elements and the non-volatile storage elements is provided in a conductive state. 4. The method of claim 1, further comprising: 5. = a first boost period 'applying a voltage to the second side of the plurality of word lines on the drain side adjacent to the second At least a sub-line of the word line for providing at least one non-volatile storage element adjacent to the second conductive state on the drain side of the second word line of the plurality of non-volatile storage elements . Conducted as in the method of claim 1, the step _ includes: = 弟 - boost period 'will - (d) apply to the plurality of word lines = the first word line set on the drain side of the : word line for avoiding The at least-nand string is raised on the (four) pole side of the sub-line. 6. The method of claim 1, wherein the method further comprises: applying a power house to the plurality of non-numbers of the plurality of word lines on the drain side of the second word line during the first-elevation The volatile storage line is provided to the at least one of the first plurality of word lines in the at least one of the first plurality of word lines. The upper-word line set is used to avoid X: the non-volatile storage element - and the pole side is raised by the second method. The method of claim 1 further includes: during the second lifting period, applying a voltage to the At least one of the first word line and the second word line of the plurality of word line lines for use in the first word after the first word. - Raising the at least one between the lines. For example, the method of claim 1, the step-by-step includes, during the second boosting period: the medium voltage is applied to the - extra word line for providing the -nonstate:: An additional non-volatile storage element, the additional word line being on the first = no-pole side, the second boosting line is an additional word line between the third word line and the additional word line for the at least one ND string On the one side, execute at least = two brothers two liters. 9. A non-volatile storage system comprising: a NAND string having a plurality of non-volatile storage elements; a plurality of word lines communicating with the at least NAND string; and 'one or more control circuits And communicating with the plurality of word lines 1 = a plurality of control circuits: (4) on the 1st side of the - second word line: :=;, - before the NAND string is on the -first side of the -th word line Performing a first boost of the upper string, the second word line is on the first: to the extreme side; (8) during the first boosting period, applying one (four) to the word line to provide the plurality of non-volatile stores The first non-volatile storage element in the - conduction state associated with the first word line of 131225.doc 200903499, and a voltage applied to the second word line to provide the plurality of non-volatile storage a second non-volatile storage element in a conductive state associated with the second word line; and (4) performing the at least one on the non-polar side of the second word line after the first boost The second increase in the string, while applying a voltage to the first word line to provide A non-conducting state of the second - non-volatile storage element, and simultaneously - the program voltage is applied to the second word line. 10. The non-volatile storage system of claim 9 wherein: ▲ during the second boost period, applying a voltage at a first level to/in the plurality of sub-lines at the second word line a word line on the side, the voltage applied to the first word line is applied to the second level a less than the first level, and the voltage at a level greater than the second level is applied to the Xth At least one intermediate word line between the line and the second word line of the β. 11. The non-volatile storage system of claim 9, wherein: during the first boosting period, the one or more control circuits apply a voltage to the plurality of word lines at the second word line A set of word lines on the pole side is used to avoid Φ as a packet r _ + NAND string is reduced on the drain side of the ~ first sub-line. , - Η non-burst storage ... -,,, the target is applied to the second word line to provide the electrical conductivity of the second non-volatile storage component in a conductive state at the - _ level; And the one or more control circuits are applied to the plurality of word lines by a voltage at a level greater than the first level, at the first word line, 13I225.doc 200903499 A set of word lines on the pole side to perform the first boost. 13. The non-volatile storage system of claim 9, wherein: the first word line is added to the first word line to provide the first volatilization in a conductive state The voltage of the storage element is at a first level; and the one or more control circuits are applied to the at least one NAND string by applying a voltage greater than a first level of the first level - Non-volatile: A collection of 70 non-volatile storage elements on the source side of the 70 pieces is stored to perform the first boost. 14. The non-volatile storage system of claim 13, wherein: the one or more control circuits are at a scale less than the first level: r >, live, A two-level voltage is applied to a set of non-volatile storage elements on the x-pole side of the second non-volatile storage element in the plurality of non-volatile euro storage 7L pieces. 1 5. The non-volatile storage system of claim 1, wherein: _ the first or more control circuits apply a voltage at the second level to the at least one _ when the first-up is performed A collection of the non-volatile storage elements on the non-polar side of the second non-volatile storage element in the string. 16· A non-volatile storage system, comprising: a component for performing a first boosting step, wherein the one of the second word line is boosted by at least one of the NAND strings before the -th word line-source Performing at least the first-up of the NAND string on the side, the second word line being on a drain side of the first word line, including the first word line and the second word, the plurality of word lines and The at least string is associated, and the at least 131225.doc 200903499 NAND string has a plurality of non-volatile fault elements; for providing a non-volatile member in the conductive state, the first boost period, applying a wish to the first word a member for providing the plurality of non-volatile storages 1 for extracting a -first non-volatile storage element in a conductive state capable of being associated with the first sub-line, and Adding a voltage yoke to the second word line to be associated with the second word line 胪 + + ^ storage ^ deep associated with a non-volatile storage element in a conductive state; And the younger one is used to perform the second boosting component, i 筮 _~ , after the ° Haidi - soaring Performing a second boost of the at least one of the dice strings on the dipole side of the line, and applying a voltage to the first to prewire for providing the first non-volatile in a non-conducting state The component is stored and a program voltage is applied to the second word line at the same time. 131225.doc
TW097116862A 2007-05-07 2008-05-07 Non-volatile storage with boosting using channel isolation switching and method thereof TWI386944B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US11/745,082 US7460404B1 (en) 2007-05-07 2007-05-07 Boosting for non-volatile storage using channel isolation switching
US11/745,092 US7463522B2 (en) 2007-05-07 2007-05-07 Non-volatile storage with boosting using channel isolation switching

Publications (2)

Publication Number Publication Date
TW200903499A true TW200903499A (en) 2009-01-16
TWI386944B TWI386944B (en) 2013-02-21

Family

ID=39943949

Family Applications (1)

Application Number Title Priority Date Filing Date
TW097116862A TWI386944B (en) 2007-05-07 2008-05-07 Non-volatile storage with boosting using channel isolation switching and method thereof

Country Status (4)

Country Link
KR (1) KR101431195B1 (en)
CN (1) CN101715596B (en)
TW (1) TWI386944B (en)
WO (1) WO2008137687A1 (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10026487B2 (en) * 2016-06-03 2018-07-17 Sandisk Technologies Llc Non-volatile memory with customized control of injection type of disturb during program verify for improved program performance
US10283202B1 (en) * 2017-11-16 2019-05-07 Sandisk Technologies Llc Reducing disturbs with delayed ramp up of selected word line voltage after pre-charge during programming
US10741262B2 (en) * 2018-10-12 2020-08-11 Macronix International Co., Ltd. NAND flash operating techniques mitigating program disturbance
KR20210119084A (en) 2020-03-24 2021-10-05 에스케이하이닉스 주식회사 Semiconductor device and operating method of semiconductor device

Family Cites Families (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH10223866A (en) * 1997-02-03 1998-08-21 Toshiba Corp Semiconductor storage device
KR100297602B1 (en) * 1997-12-31 2001-08-07 윤종용 Method for programming a non-volatile memory device
JP3810985B2 (en) * 2000-05-22 2006-08-16 株式会社東芝 Nonvolatile semiconductor memory
KR100502412B1 (en) * 2002-10-23 2005-07-19 삼성전자주식회사 Non-volatile semiconductor memory device and program method thereof
US6859397B2 (en) * 2003-03-05 2005-02-22 Sandisk Corporation Source side self boosting technique for non-volatile memory
TWI220526B (en) * 2003-03-26 2004-08-21 Macronix Int Co Ltd An operation method of nonvolatile memory array
KR100562506B1 (en) * 2003-12-01 2006-03-21 삼성전자주식회사 Flash memory device and programming method thereof
US7023739B2 (en) * 2003-12-05 2006-04-04 Matrix Semiconductor, Inc. NAND memory array incorporating multiple write pulse programming of individual memory cells and method for operation of same
US7161833B2 (en) * 2004-02-06 2007-01-09 Sandisk Corporation Self-boosting system for flash memory cells
US7023733B2 (en) 2004-05-05 2006-04-04 Sandisk Corporation Boosting to control programming of non-volatile memory
US7307884B2 (en) * 2004-06-15 2007-12-11 Sandisk Corporation Concurrent programming of non-volatile memory
TWI297154B (en) * 2005-01-03 2008-05-21 Macronix Int Co Ltd Non-volatile memory cells, memory arrays including the same and methods of operating cells and arrays
KR100621634B1 (en) 2005-05-06 2006-09-07 삼성전자주식회사 Flash memory device and program method thereof

Also Published As

Publication number Publication date
KR101431195B1 (en) 2014-08-18
TWI386944B (en) 2013-02-21
WO2008137687A1 (en) 2008-11-13
CN101715596A (en) 2010-05-26
CN101715596B (en) 2013-08-21
KR20100029194A (en) 2010-03-16

Similar Documents

Publication Publication Date Title
KR101519081B1 (en) Enhanced bit-line pre-charge scheme for increasing channel boosting in non-volatile storage
KR101655352B1 (en) Forecasting program disturb in memory by detecting natural threshold voltage distribution
TWI424436B (en) Compensating for coupling during read operations in non-volatile storage
EP2831884B1 (en) Nonvolatile memory programming with string select transistor gate voltage dependent on location of selected word line
EP2929537B1 (en) Non-volatile storage nand string select gate voltage lowered during programming
EP2446441B1 (en) Reduced programming pulse width for enhanced channel boosting in non-volatile storage
EP2831883B1 (en) Nonvolatile memory programming with string select transistor diffusion region voltage dependent on location of selected word line
TW201017672A (en) Multi-pass programming for memory with reduced data storage requirement
EP2140457B1 (en) Reducing power consumption during read operations in non-volatile storage
EP2078303B1 (en) Reading of a nonvolatile memory cell by taking account of the stored state of a neighboring memory cell
WO2013147938A1 (en) Selected word line dependent programming voltage
TW200949844A (en) Sensing in non-volatile storage using pull down to regulated source voltage to remove system noise
EP2078302B1 (en) Reducing program disturb in non-volatile storage using early source-side boosting
TW200818195A (en) Method and system for reducing the impact of program disturb during read
KR100948200B1 (en) Reducing read disturb for non-volatile storage
TW200903499A (en) Non-volatile storage with boosting using channel isolation switching and method thereof
TW200841341A (en) Programming non-volatile memory with reduced program disturb by using different pre-charge enable voltages and non-volatile memory system
TW200849259A (en) NAND flash memory cell array and method with adaptive memory state partitioning
TWI352404B (en) Providing local boosting control implant for non-v
TW200822344A (en) Method and system for reducing program disturb in non-volatile storage
TWI357603B (en) Reducing program disturb in non-volatile memory us
TW200901201A (en) Systems and methods for programming non-volatile memory with reduced program disturb by removing pre-charge dependency on word line data

Legal Events

Date Code Title Description
MM4A Annulment or lapse of patent due to non-payment of fees