TW200830315A - Method of operating non-volatile storage and non-volatile memory system - Google Patents

Method of operating non-volatile storage and non-volatile memory system Download PDF

Info

Publication number
TW200830315A
TW200830315A TW96127570A TW96127570A TW200830315A TW 200830315 A TW200830315 A TW 200830315A TW 96127570 A TW96127570 A TW 96127570A TW 96127570 A TW96127570 A TW 96127570A TW 200830315 A TW200830315 A TW 200830315A
Authority
TW
Taiwan
Prior art keywords
level
storage elements
resistance
reverse bias
reset
Prior art date
Application number
TW96127570A
Other languages
Chinese (zh)
Other versions
TWI356415B (en
Inventor
Roy E Scheuerlien
Tanmay Kumar
Original Assignee
Sandisk 3D Llc
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 US11/461,424 external-priority patent/US7495947B2/en
Priority claimed from US11/461,431 external-priority patent/US7492630B2/en
Application filed by Sandisk 3D Llc filed Critical Sandisk 3D Llc
Publication of TW200830315A publication Critical patent/TW200830315A/en
Application granted granted Critical
Publication of TWI356415B publication Critical patent/TWI356415B/en

Links

Classifications

    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C13/00Digital stores characterised by the use of storage elements not covered by groups G11C11/00, G11C23/00, or G11C25/00
    • G11C13/0002Digital stores characterised by the use of storage elements not covered by groups G11C11/00, G11C23/00, or G11C25/00 using resistive RAM [RRAM] elements
    • G11C13/0007Digital stores characterised by the use of storage elements not covered by groups G11C11/00, G11C23/00, or G11C25/00 using resistive RAM [RRAM] elements comprising metal oxide memory material, e.g. perovskites
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C11/00Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C11/56Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using storage elements with more than two stable states represented by steps, e.g. of voltage, current, phase, frequency
    • G11C11/5685Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using storage elements with more than two stable states represented by steps, e.g. of voltage, current, phase, frequency using storage elements comprising metal oxide memory material, e.g. perovskites
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C13/00Digital stores characterised by the use of storage elements not covered by groups G11C11/00, G11C23/00, or G11C25/00
    • G11C13/0002Digital stores characterised by the use of storage elements not covered by groups G11C11/00, G11C23/00, or G11C25/00 using resistive RAM [RRAM] elements
    • G11C13/0021Auxiliary circuits
    • G11C13/0069Writing or programming circuits or methods
    • 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
    • G11C16/3404Convergence or correction of memory cell threshold voltages; Repair or recovery of overerased or overprogrammed cells
    • G11C16/3413Circuits or methods to recover overprogrammed nonvolatile memory cells detected during program verification, usually by means of a "soft" erasing step
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C5/00Details of stores covered by group G11C11/00
    • G11C5/02Disposition of storage elements, e.g. in the form of a matrix array
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C17/00Read-only memories programmable only once; Semi-permanent stores, e.g. manually-replaceable information cards
    • G11C17/14Read-only memories programmable only once; Semi-permanent stores, e.g. manually-replaceable information cards in which contents are determined by selectively establishing, breaking or modifying connecting links by permanently altering the state of coupling elements, e.g. PROM
    • G11C17/16Read-only memories programmable only once; Semi-permanent stores, e.g. manually-replaceable information cards in which contents are determined by selectively establishing, breaking or modifying connecting links by permanently altering the state of coupling elements, e.g. PROM using electrically-fusible links
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C13/00Digital stores characterised by the use of storage elements not covered by groups G11C11/00, G11C23/00, or G11C25/00
    • G11C13/0002Digital stores characterised by the use of storage elements not covered by groups G11C11/00, G11C23/00, or G11C25/00 using resistive RAM [RRAM] elements
    • G11C13/0021Auxiliary circuits
    • G11C13/0069Writing or programming circuits or methods
    • G11C2013/009Write using potential difference applied between cell electrodes
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C2211/00Indexing scheme relating to digital stores characterized by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C2211/56Indexing scheme relating to G11C11/56 and sub-groups for features not covered by these groups
    • G11C2211/564Miscellaneous aspects
    • G11C2211/5641Multilevel memory having cells with different number of storage levels
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C2213/00Indexing scheme relating to G11C13/00 for features not covered by this group
    • G11C2213/30Resistive cell, memory material aspects
    • G11C2213/32Material having simple binary metal oxide structure
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C2213/00Indexing scheme relating to G11C13/00 for features not covered by this group
    • G11C2213/30Resistive cell, memory material aspects
    • G11C2213/33Material including silicon
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C2213/00Indexing scheme relating to G11C13/00 for features not covered by this group
    • G11C2213/30Resistive cell, memory material aspects
    • G11C2213/34Material includes an oxide or a nitride
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C2213/00Indexing scheme relating to G11C13/00 for features not covered by this group
    • G11C2213/70Resistive array aspects
    • G11C2213/71Three dimensional array
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C2213/00Indexing scheme relating to G11C13/00 for features not covered by this group
    • G11C2213/70Resistive array aspects
    • G11C2213/72Array wherein the access device being a diode

Landscapes

  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Materials Engineering (AREA)
  • Computer Hardware Design (AREA)
  • Semiconductor Memories (AREA)
  • Static Random-Access Memory (AREA)

Abstract

A reverse bias trim operation for the reset state of a non-volatile memory system is disclosed. Non-volatile memory cells including a resistance change element undergo a reverse bias reset operation to change their resistance from a set state at a first level of resistance to a reset state at a second level of resistance. Certain memory cells in a set of cells that was reset may be deeply reset to a level of resistance beyond a target level for the reset state. A second reverse bias is applied to the set of memory cells to move the resistance of each cell that was deeply reset toward the target level of the reset state. A smaller reverse bias than used for the reset operation can shift the resistance of the cells back toward the set level and out of their deeply reset condition. The operation is self-limiting in that cells stop their resistance shifts upon reaching the target level. Cells that were not deeply reset are not affected.

Description

200830315 九、發明說明: 【發明所屬之技術領域】 依據本揭示内容之具體實施例係關於包含非揮發性記憶 體單元陣列而特定言之係該些併入被動元件記憶體單元的 、 陣列之積體電路。 Μ 【先前技術】 使用具有一可偵測位準的狀態變化(例如一電阻或相變) 馨之材料來形成各類基於非揮發性半導體之記憶體裝置。例 如,簡單的反熔絲係用於一次場可程式化(〇Τρ)記憶體陣 列中的二進制資料儲存,此係藉由將一記憶體單元之一較 氐電阻初始只體狀態指派給一第一邏輯狀態(例如邏輯"〇,,) 而將該單元之一較高電阻實體狀態指派給一第二邏輯狀態 (例如邏輯”1”)。某些材料可以在其初始電阻之方向上往^ 切換其電阻。此等材料可用於形成可重寫的記憶體單元。 材料中多個位準的可偵測電阻可進一步用於形成可能可以 I 或可能不可以重寫之一多狀態裝置。 ,’、有5己诚體效果(例如一可偵測的電阻位準)之材料常 常係與-引導元件串聯放置以形成_記憶體裝置H • #線⑨傳‘電流之二極體或其他裝置-般係用作該引導元 ^ 2所在許夕貝施方案中,一組字元線與位元線係配置為一 :貝垂直的組悲,❿_記憶體單元處於每一字元線與位 :: 交又點。可以將兩終端之記憶體單元構造於該等交 M H終端(例如’該單元之終端部分或該單元之 分離層)與形成個別字元線的導體接觸而另一終端與形成 123008.doc 200830315 個別位元線之導體接觸。 在此類况下’在將具有包含可切換電阻材料或相變材 料的被動元件兄憶體單元之非揮發性記憶體陣列實施為該 狀二、改變元件時’讀取及寫入操作期間的偏壓條件係一重 要考蓋口素。在1試生產包含可以可靠地製造、程式化及 ^ 讀取的-或多個被動元件記憶體單元陣列之—記憶體裝置 夺回/兔漏電流、程式干擾、讀取干擾等可能造成困難。 • 例如,此等因素常常藉由減少可以同時定址的單元數目而 限制系統性能,以便讓洩漏電流保持於可接受的位準。在 I忒針對尚頻寬讀取與程式操作而個別或同時定址多個單 元時個別"己彳思體單元之間微小的差異可能亦造成困難。 特定單兀可能具有可能導致電阻超出與此等操作後之一對 應貧料狀態之相關聯的範圍的特性。例如,與經受相同偏 壓條件之其他單元相比,特定的單元可以經歷不同數量的 電阻偏移。 • 【發明内容】 本發明揭示一種針對非揮發性記憶體系統的重新設定狀 態之反向偏壓調整操作。包括一電阻改變元件之非揮發性 ‘ €憶'體單元經歷—反向偏壓重新設定操作以將其電阻從處 v 力一第一電阻位準之一設定狀態改變為處於一第二電阻位 準之-重新設定狀態。可以將一已重新設定的單元华合中 之特定記憶體單元深度重新設定為超出一針對該重新設定 狀態的目標位準之-電阻位準。向該記憶體單元集合施加 -第二反向偏壓以將已深度重新設定的每一單元之電阻朝 123008.doc 200830315 =重新設疋狀態之目標位準移動。與用於該重 的反向偏屢相比之-更小的反向偏壓可以將該等翠元= 阻朝該設定位準往回偏移而脫離其深度重新設定 操作係自我限制性,因為單元在到達該目標位準:二 其電阻偏移。未深度重新設定之單元不受影變。不τ 在一具體實施例中,提供一種操作非料性儲存之方 法’其包括向複數個非揮發性儲存元件施加一第一200830315 IX. Description of the Invention: [Technical Field] The embodiments according to the present disclosure relate to an array comprising non-volatile memory cell arrays, in particular, those incorporating passive component memory cells. Body circuit. Μ [Prior Art] A variety of non-volatile semiconductor-based memory devices are formed using materials with a detectable level of state change (such as a resistance or phase change). For example, a simple antifuse is used to store binary data in a field programmable (〇Τρ) memory array by assigning one of the memory cells to the first body state of the resistor. A logic state (e.g., logic "〇,,) assigns one of the higher resistance entity states of the cell to a second logic state (e.g., logic "1"). Some materials can switch their resistance to ^ in the direction of their initial resistance. These materials can be used to form rewritable memory cells. Multiple levels of detectable resistance in the material can be further used to form a multi-state device that may or may not be rewritten. , 'The material with 5 honest effects (such as a detectable resistance level) is often placed in series with the - guiding element to form a _ memory device H • #线9传' current diode or other The device-like system is used as the guiding element ^2 in the Xu Xibei Shi scheme, a group of character lines and bit line systems are configured as one: the vertical group of sadness, the memory cell is in each word line With the bit:: Hand in and out. The memory units of the two terminals may be constructed at the MH terminals (eg, the terminal portion of the unit or the separate layer of the unit) and the conductors forming the individual word lines are in contact with each other and the other terminal is formed with 123008.doc 200830315 The conductor of the bit line is in contact. In such cases, 'a non-volatile memory array having a passive component of a passive component containing a switchable resistive material or a phase change material is implemented as a second, changing element during a read and write operation. Bias conditions are an important test. In a trial production, it is possible to reliably manufacture, program, and read - or a plurality of passive component memory cell arrays - memory devices recapture / rabbit leakage current, program interference, read interference, etc. may cause difficulties. • For example, these factors often limit system performance by reducing the number of cells that can be addressed simultaneously to keep the leakage current at an acceptable level. Small differences between individual "study units may also be difficult when I'm addressing multiple units individually or simultaneously for frequent bandwidth reading and program operations. A particular unit may have characteristics that may cause the resistance to exceed a range associated with one of the post-operational lean states. For example, a particular cell may experience a different amount of resistance offset than other cells that are subjected to the same bias conditions. • SUMMARY OF THE INVENTION The present invention discloses a reverse bias adjustment operation for a reset state of a non-volatile memory system. A non-volatile 'return' body unit including a resistance changing element undergoes a reverse bias reset operation to change its resistance from a position of one of the first resistance levels to a second resistance level Pre-set - reset the status. The specific memory cell depth in a reset cell can be reset to a resistance level that exceeds a target level for the reset state. A second reverse bias is applied to the set of memory cells to move the resistance of each cell that has been depth reset toward the target level of the 123008.doc 200830315 = reset state. The smaller reverse bias voltage compared to the reverse bias used for the weight can be offset from the set level by the offset, and the depth reset operation is self-limiting. Because the unit is reaching the target level: its resistance is offset. Units that are not re-set in depth are not affected by the change. In a specific embodiment, a method of operating a non-material storage is provided, which includes applying a first to a plurality of non-volatile storage elements.

ί向偏壓以在—第—方向上將每-儲存元件之-電阻從 弟一電阻狀態移動至-第二電阻狀態。該方法接著包 該複數個鱗發性料元件施加―第二位準的反向偏壓, 以將具有超出一針對該第二電阻狀態的目標電阻位準之一 電阻的該複數個儲存元件之_子集之—電阻在_第二 上朝該目標電阻位準移動。 一 U 在-具體實施例中,提供一種操作非揮發性儲存之方 法’其包括藉由向複數個非揮發性儲存元件施加_第—位 準的反向偏壓而將該等儲存元件從—較低電阻狀態切換為 一較高電阻狀態。在切換後接著向該等儲存元件施加一第 :位準的反向偏壓以降低具有超出一對應於該第二電阻狀 悲的目標位準之一電阻的該等儲存元件之一子集之一 阻。 术一電 在-具體實施例中提供一種非揮發性記憶體系統,直包 括:有至少-電阻改變元件以及與該複數個非揮發性健存 心牛通信的控制電路之複數個非揮發性儲存元件。該控制 電路執行-重新設定操作,其包括:藉由向該複數個_ 123008.doc 200830315 發性儲存元件施加至少一反向偏壓重新設定電壓來將該等 儲存元件從一較低電阻設定狀態重新設定為_較高電p且重 新設定狀態;以及向該複數個非揮發性儲存元件施加至少 一反向偏壓調整電壓以降低具有超出一針對該第二較高電 阻重新設定狀態的目標值之一電阻的該複數個非揮發性儲 存元件之一子集之一電阻。 【實施方式】 圖1說明可依據本揭示内容之具體實施例使用之一非揮 發性記憶體單元之一範例性結構。如圖1所示之一兩終端 記憶體單元100包括連接至一第一導體110之一第一終端部 分與連接至一第二導體112之一第二終端部分。該記憶體 單兀包括與一狀態改變元件1〇4及一反熔絲1〇6串聯之一引 導兀件102以提供非揮發性資料儲存。該引導元件可以採 取呈現一非線性傳導電流特徵之任何合適的裝置(例如一 簡單的二極體)為形式。該狀態改變元件將隨具體實施例 而變化並且可以包括諸多類型的材料來經由代表性的實體 狀態儲存資料。狀態改變元件1〇4可以包括電阻改變材 料、相變電P且材料等。例如,在一具體實施例中使用具有 兩個位準的可偵測電阻變化(例如,從低至高與從高 至低)之半^體或其他材料來形成一被動儲存元件丨〇〇。 f由將邏輯資料值指派給可以設定並從電阻改變元件104 :取之各種位準的電阻,記憶體單元剛可以提供可靠的 貝料項取/寫人能力。反料而可以進一步提供可以運用 於非揮發性資料儲存之電阻狀態改變能力。反㈣係製造 123008.doc 200830315 ^ 一高電阻狀態而且可以係跳脫或熔合成一較低電阻狀 您。反熔絲在處於其初始狀態時一般係非導電,而在處於 其跳脫或熔合狀態的低電阻條件下呈現高導電率。由於一 . 冑散裝置或元件可以具有—電阻及不同的電阻狀態,因此 • 術語電阻率及電阻率狀態係用來表示材料本身之特性。因 - & ’儘管一冑阻改變元件或裝置可以具有t阻狀態,但一 電阻率改變材料可以具有電阻率狀態。 p 反熔絲1〇6可以向記憶體單元100提供超出其狀態變化能 力之優‘點。例如,反熔絲可用於相對於與該單元相關聯的 讀取寫入電路而將該記憶體單元之開啟電阻設定於一適當 位準。此等電路一般係用於跳脫該反熔絲且具有一相關聯 的電阻m等電路驅動電壓及電流位準以跳脫該反熔 絲’因此在隨後的操作期間’該反溶絲傾向於將該記憶體 單元設定於針對此等相同電路之一適當的開啟電阻狀態。 應'日月白’可以在具體實施例中使用其他類㉟的兩終端非 > 揮發性記憶體單元。例如’一具體實施例不具有一反熔絲 106而僅包括狀態改變元件1〇4與引導元件1〇2。其他具體 實施例可以包括額外的狀態改變元件作為該反熔絲之替代 * 元件或額外元件。在名稱為"垂直堆疊的塲可程式化非揮 - 發性記憶體及製造方法"之美國專利案第6,034,882號中說 明各類合適的記憶體單元。可以使用各種其他類型的單 元,包括以下專利案中說明的該些單元:美國專利案第 6,420,215號及美國專利申請案序列號_97,7()5,其名稱 為"併入串聯鏈式二極體堆疊之三維記憶體陣列",申請於 123008.doc -11- 200830315 20〇1年6月29日;以及美國專利中請案序列號〇9/56〇 626, 其名稱為"三維記憶體陣列及製造方法”,其係申請於2〇〇〇 14月28日’ i述各案之全部内容皆係以引用的方式併入 於此。 依據本揭示内容之具體實施例,各種材料呈現適用於實 施狀態改變元件104之電阻率改變特性。適用於電阻狀態 改變元件104之材料之範例包括但不限於摻雜的半導體(例 如’多晶體的石夕、更-般而言係多晶石夕)、過渡金屬氧化 物、複合金屬氧化物、可程式化金屬化連接、相變電阻元 件、有機材料可變電阻器、碳聚合物膜 '摻雜的硫屬化合 物玻璃及包含改變電阻的行動原子之肖特基(Sch〇ttky)阻 障二極體。在某些情況下,此等材料之電阻率可以係僅設 定於一第一方向(例如,從高到低),而在其他情況下,該 電阻率可以係從一第一位準(例如,較高電阻)設定為一第 二位準(例如,較低位準)而接著重新設定回到該第一電阻 率位準。 可以將一電阻值範圍指派給一實體資料狀態以適應裝置 之間的差異以及在設定及重新設定循環後裝置内的變化。 術語設定與重新設定一般係分別用於表示將一元件從一高 電阻實體狀態改變為一低電阻實體狀態(設定)與將一元件 從一低電阻實體狀態改變為一高電阻實體狀態(重新設定) 之程序。依據本揭示内容之具體實施例可用於將記憶體單 元設定為一較低電阻狀態或將記憶體單元設定為一較高電 阻狀態。儘管可相對於設定或重新設定操作提供特定範 123008.doc -12- 200830315 例,但應明白&等範例僅係範例而本揭示内纟不受此限 制。 導體110及112-般係彼此正交而形成用以存取一記憶體 單元陣列1〇〇之陣列終端線。處於一層的陣列終端線(亦稱 為陣列線)可稱為字元線或X線。處於一垂直相鄰層之陣列 線可稱為位元線或γ線。一記憶體單元可以係形成於每一 子7G線與每一位元線之凸出的交叉點,並連接於個別交叉 子7G線與位元線之間(如圖中針對記憶體單元1〇〇之形成所 示)。具有至少兩個記憶體單元層級(即,兩個記憶體平面) 之一三維記憶體陣列可以利用一層以上之字元線及/或一 層以上之位元線。一單石三維記憶體陣列係一其中在一單 基板(例如,一晶圓)上形成多個記憶體層級之陣列,其 中無插入的基板。 圖2A及2B係對可用於各項具體實施例之範例性記憶體 單兀之更詳細的說明。圖2A中,在第一及第二金屬導電層 110與112之間形成記憶體單元丨2〇。該記憶體單元包括具 有一重度摻雜的η型區域122、本質區域124及一重度捧雜 的Ρ型區域126之一p_i-n型二極體。在其他具體實施例中, 區域122可以係p型,而區域126係11型。區域124係本質 的,或並非故意摻雜,但在某些具體實施例中其可為輕度 摻雜。未摻雜區域可能並非係極佳的電性中性,而產生缺 陷、污染物等使其性能如同輕度n摻雜或p摻雜一般。吾等 仍將此一二極體視為具有一本質中間層之一 型二極 體。還可以使用其他類型的二極體,例如^^接面二極 123008.doc -13 - 200830315 介於摻雜的p型區域126與導體110之間的係一反溶絲 128。反熔絲128在其初始狀態中呈現實質上非導電的特性 而在其$又疋狀態中呈現實質上導電的特性。可以依據具體 只&例而使用各類型反溶絲。在一般製造之反溶絲中,橫 , 跨該反熔絲而施加之一較大偏壓將熔合形成材料而使得該 反溶絲變成實質上導電。此操作一般係稱為跳脫該反熔 φ 絲。 記憶體單元120進一步包括由該二極體之一或多層形成 之一狀態改變元件。頃發現,用於在某些記憶體單元中形 成二極體之材料本身呈現電阻改變能力。例如,在一具體 實施例中,該二極體之本質區域係由多晶矽形成,該多晶 石夕經證實具有從一較高電阻率狀態設定為一較低電阻率狀 態而接著從該較低電阻率狀態設定回到一較高電阻率狀態 之能力。因此’該二極體本身或其一部分還可以形成如圖 隹 1所示之狀態改變元件104。在其他具體實施例中,可以將 一或多個額外層包括於記憶體單元120中以形成如圖1所示 之一狀態改變元件。例如,可以將如上所述的多晶矽、過 - 渡金屬氧化物等之一額外層包括於該單元中以提供一狀態 « 改變記憶體效果。可以將此額外層包括於該二極體與導體 112之間、該二極體與該反熔絲128之間或該反熔絲與導體 110之間。 圖2B解說一其中不存在一反熔絲128之簡單的記憶體單 元組態。記憶體單元140僅包括重度摻雜的n型區域142、 123008.doc -14- 200830315 本質區域144及重度摻雜的p型區域146。由此等區域形成 的二極體之一或多層用作如上所述針對該單元之記憶體效 果。在一具體實施例中,記憶體單元14〇還可以包括其他 層以形成用於該單元之一額外的狀態改變元件。 圖3A至3B說明可用於一具體實施例之一範例性單石三 維纪憶體陣列之一部分。但是,可以依據各項具體實施例 而使用其他記憶體結構,包括製造於一半導體基板上面、 上方或内部之二維記憶體結構。但是,該等字元線與位元 線層係在圖3A之透視圖所說明結構中的記憶體單元之間共 旱。此組悲常常係稱為一完全鏡射結構。複數個實質上平 行及共面的導體在一第一記憶體層級L〇形成一第一位元線 集合162。在此等位元線與相鄰的字元線之間形成處於層 級L0之記憶體單元152。在圖3A至3B之配置中,字元線 164係在記憶體層L0與L1之間共享,而因此進一步連接至 處於記憶體層級L1之記憶體單元170。一第三導體集合形 成針對此等處於層級L1的單元之位元線174。此等位元線 174進而係在記憶體層級L1與記憶體層級£2之間共享,如 圖3B之斷面圖所說明。記憶體單元178係連接至位元線 及字元線176以形成第三記憶體層級L2,記憶體單元182係 連接至子元線176及位元線180以形成第四記憶體層級, 而記憶體單元186係連接至位元線U0及字元線184以形成 弟五έ己憶體層級L5。該等二極體的極性之配置及該等字元 線與位元線之個別配置可隨具體實施例而變化。此外,可 以使用多於或少於五個記憶體層級。 123008.doc -15- 200830315 若在圖3A之具體實施例中使用p_i-n二極體作為該等記 憶體單元之引導元件,則記憶體單元17〇之二極體可以係 相對於該第一層級的記憶體單元152ip_i_n二極體而顛甸 形成。例如,若單元152包括一 n型底部重度摻雜區域與一 P型頂部重度摻雜區域,則在第二層級的單元i 7〇中,該底 邛重度摻雜區域可以係p型而該頂部重度摻雜區域係η型。 在一替代性具體實施例中,可以在相鄰的記憶體層級之 間形成一層間介電質。在記憶體層級之間不共享任何導 體。用於二維單石儲存記憶體之此類結構常常係稱為一非 鏡射結構。在某些具體實施例中,共享導體之相鄰記憶體 層級與不共享導體之相鄰記憶體層級可以係堆疊於同一單 石二維記憶體陣列中。在其他具體實施例中,某些導體係 共享而其他導體並非共享。例如,在某些組態中可以共享 僅該等字疋線或僅該等位元線。一第一記憶體層級可以 包括介於一位元線層級BL0與字元線層級WL〇之間的記憶 體單元。處於層級WL0之字元線可以係共享以在一記憶體 層級L1形成連接至一第二位元線BL1之單元。該等位元線 層並非共享,因此接下來的一層可以包括一層間介電質以 將位元線BL1與接下來之一層級的導體分離。此類組態常 常稱為半鏡射。記憶體層級不必皆形成為具有同類記㈣ 單元。若需要,使用電阻改變材料之記憶體層級可^使 用其他類型記憶體單元之記憶體層級等交替。 的記憶體陣列線之電晶體佈局組態")所說明之_二; 在如美國專利案第7,〇54,219號(其名稱為”針對緊密間距 施 123008.doc -16 - 200830315 例中,使用置放於該陣列的 +幻的* g子兀線層上之字元線片段 來形成字元m藉由―垂直連接將料諸連接成形 成-個別字元線。各字元線皆駐留於一分離層上且實質上 垂直對齊(儘管在某此層上在尤# — 你示一嘈上存在較小橫向偏移)之一群組的The bias voltage shifts the resistance of each of the storage elements from the first resistance state to the second resistance state in the -first direction. The method then applies a second level of reverse bias to the plurality of scale elements to: the plurality of storage elements having a resistance that exceeds a target resistance level for the second resistance state _ subset - the resistor moves on the _second toward the target resistance level. In a specific embodiment, a method of operating a non-volatile storage is provided that includes including the storage elements by applying a reverse bias to a plurality of non-volatile storage elements. The lower resistance state switches to a higher resistance state. Applying a first: level reverse bias to the storage elements after switching to reduce a subset of the storage elements having a resistance that exceeds a target level corresponding to the second resistive sorrow One resistance. In a specific embodiment, a non-volatile memory system is provided, comprising: a plurality of non-volatile storage elements having at least a resistance change element and a control circuit in communication with the plurality of non-volatile health care cattle . The control circuit performs a reset operation including: resetting the storage elements from a lower resistance state by applying at least one reverse bias reset voltage to the plurality of _123008.doc 200830315 Resetting to a higher power p and resetting the state; and applying at least one reverse bias voltage adjustment voltage to the plurality of non-volatile storage elements to reduce a target value exceeding a reset state for the second higher resistance One of a plurality of subsets of the plurality of non-volatile storage elements of the resistor. [Embodiment] FIG. 1 illustrates an exemplary structure in which one of the non-volatile memory cells can be used in accordance with a specific embodiment of the present disclosure. A two-terminal memory unit 100 as shown in FIG. 1 includes a first terminal portion connected to one of the first conductors 110 and a second terminal portion connected to one of the second conductors 112. The memory unit includes a boot device 102 in series with a state change element 1〇4 and an antifuse 1〇6 to provide non-volatile data storage. The guiding element can take the form of any suitable device (e.g., a simple diode) that exhibits a non-linear conduction current characteristic. The state change element will vary from embodiment to embodiment and may include many types of materials to store data via a representative physical state. The state changing element 1〇4 may include a resistance change material, a phase change electric P, a material, and the like. For example, a passive storage element 形成 is formed in a particular embodiment using a half-body or other material having two levels of detectable resistance change (e.g., low to high and high to low). By assigning the logical data value to a resistor that can be set and changed from the resistance change element 104: various levels, the memory unit can just provide reliable feed/write capability. Conversely, it is possible to further provide a resistance state change capability that can be applied to non-volatile data storage. Anti-(four) system manufacturing 123008.doc 200830315 ^ A high resistance state and can be tripped or melted into a lower resistance. The antifuse is generally non-conductive when in its initial state and exhibits high conductivity under low resistance conditions in its tripped or fused state. Since a dissipative device or component can have a resistance and a different resistance state, the term resistivity and resistivity states are used to indicate the properties of the material itself. A resistivity changing material may have a resistivity state, although a resistive changing element or device may have a t-resistive state. The p-fuse 1〇6 can provide the memory cell 100 with an excellent point beyond its state change capability. For example, an anti-fuse can be used to set the on-resistance of the memory cell to an appropriate level relative to a read write circuit associated with the cell. These circuits are typically used to trip the anti-fuse and have an associated resistance voltage m and other circuit drive voltage and current levels to trip the anti-fuse 'so during the subsequent operation' the anti-solving filament tends The memory cell is set to an appropriate turn-on resistance state for one of the same circuits. Other terminal 35 non- > volatile memory cells may be used in the specific embodiment. For example, a particular embodiment does not have an antifuse 106 and only includes state changing element 1〇4 and guiding element 1〇2. Other embodiments may include additional state changing elements as an alternative to the antifuse * element or additional element. Various suitable memory cells are described in U.S. Patent No. 6,034,882, the disclosure of which is incorporated herein by reference. Various other types of units may be used, including those described in the following patents: U.S. Patent No. 6,420,215 and U.S. Patent Application Serial No. _97,7() 5, entitled "Incorporating Series Chains A three-dimensional memory array of a diode stack is applied at 123008.doc -11- 200830315 on June 29, 20; and in US patents, serial number 〇9/56〇626, whose name is " "Three-dimensional memory arrays and methods of manufacture", which are hereby incorporated by reference in its entirety in its entirety in the entire entire entire entire entire entire entire entire entire disclosure The material exhibits a resistivity change characteristic suitable for implementing the state change element 104. Examples of materials suitable for the resistance state change element 104 include, but are not limited to, doped semiconductors (eg, 'polycrystalline, but more generally Spar), transition metal oxides, composite metal oxides, programmable metallization connections, phase change resistors, organic material variable resistors, carbon polymer film 'doped chalcogenide glass' and include modification a Schottky barrier diode of a variable electrical resistance atom. In some cases, the resistivity of such materials can be set only in a first direction (eg, from high to low). In other cases, the resistivity can be set from a first level (eg, higher resistance) to a second level (eg, a lower level) and then reset back to the first resistivity. A range of resistance values can be assigned to an entity data state to accommodate differences between devices and changes within the device after a set and reset cycle. Term setting and resetting are generally used to indicate that a component is being used. A high resistance physical state changes to a low resistance physical state (set) and a procedure for changing a component from a low resistance physical state to a high resistance physical state (reset). Embodiments in accordance with the present disclosure may be used The memory unit is set to a lower resistance state or the memory unit is set to a higher resistance state. Although a specific specification can be provided with respect to the setting or reset operation 123008.doc - 12-200830315 Example, but it should be understood that the examples such as & are merely examples and the disclosure is not limited thereto. The conductors 110 and 112 are generally orthogonal to each other to form a memory cell array. Array terminal lines. Array terminal lines (also referred to as array lines) in one layer may be referred to as word lines or X lines. Array lines in a vertically adjacent layer may be referred to as bit lines or gamma lines. It can be formed at the intersection of each sub 7G line and each bit line, and is connected between the individual cross sub 7G line and the bit line (as shown in the figure for the formation of the memory unit 1) One of the three-dimensional memory arrays having at least two memory cell levels (ie, two memory planes) may utilize more than one word line and/or more than one bit line. A monolithic three-dimensional memory array is one in which an array of a plurality of memory levels is formed on a single substrate (e.g., a wafer) without an interposed substrate. Figures 2A and 2B are a more detailed illustration of exemplary memory cartridges that may be used in various embodiments. In Fig. 2A, a memory cell 丨 2 形成 is formed between the first and second metal conductive layers 110 and 112. The memory cell includes a p_i-n type dipole having a heavily doped n-type region 122, an intrinsic region 124, and a heavily doped germanium region 126. In other embodiments, region 122 can be p-type and region 126 can be type 11. Region 124 is essential or not intentionally doped, but in some embodiments it may be lightly doped. Undoped regions may not be extremely electrically neutral, but defects, contaminants, etc., which cause their properties to be as mild as n-doping or p-doping. We still consider this diode to be a type of diode with an intrinsic intermediate layer. Other types of diodes can also be used, such as a junction dipole 123008.doc -13 - 200830315 between the doped p-type region 126 and the conductor 110. The anti-fuse 128 exhibits substantially non-conductive properties in its initial state and exhibits substantially conductive characteristics in its $-deuterium state. Each type of anti-solvent can be used depending on the specific & example. In a generally manufactured reverse dissolution filament, a large bias applied across the antifuse will fuse to form a material such that the reversely soluble filament becomes substantially electrically conductive. This operation is generally referred to as tripping the anti-melting φ wire. The memory unit 120 further includes a state changing element formed of one or more of the diodes. It has been found that the material used to form the diode in some memory cells exhibits a resistance change capability. For example, in one embodiment, the essential region of the diode is formed of polysilicon, which is shown to have a lower resistivity state from a higher resistivity state and then from the lower The ability of the resistivity state to return to a higher resistivity state. Therefore, the diode itself or a part thereof can also form the state changing element 104 as shown in Fig. 1. In other embodiments, one or more additional layers may be included in memory unit 120 to form a state change element as shown in FIG. For example, an additional layer of polycrystalline germanium, over-metal oxide, etc. as described above may be included in the cell to provide a state «change memory effect. This additional layer may be included between the diode and conductor 112, between the diode and the antifuse 128 or between the antifuse and conductor 110. Figure 2B illustrates a simple memory cell configuration in which an antifuse 128 is absent. Memory unit 140 includes only heavily doped n-type regions 142, 123008.doc -14 - 200830315 essential regions 144 and heavily doped p-type regions 146. One or more of the diodes formed by such regions are used as the memory effect for the unit as described above. In one embodiment, memory unit 14A may also include other layers to form additional state changing elements for one of the units. Figures 3A through 3B illustrate one portion of an exemplary single-rock three-dimensional memory array that can be used in an embodiment. However, other memory structures can be used in accordance with various embodiments, including two-dimensional memory structures fabricated on, above, or within a semiconductor substrate. However, the word lines and bit line layers are coherent between the memory cells in the structure illustrated in the perspective view of Figure 3A. This group of griefs is often referred to as a completely mirrored structure. A plurality of substantially parallel and coplanar conductors form a first set of bit lines 162 at a first memory level L. A memory cell 152 at level L0 is formed between the bit lines and adjacent word lines. In the configuration of Figures 3A through 3B, word line 164 is shared between memory layers L0 and L1 and thus further coupled to memory unit 170 at memory level L1. A third set of conductors forms a bit line 174 for the cells of level L1. These bit lines 174 are in turn shared between the memory level L1 and the memory level £2, as illustrated in the cross-sectional view of Figure 3B. The memory unit 178 is connected to the bit line and the word line 176 to form a third memory level L2, and the memory unit 182 is connected to the sub-element 176 and the bit line 180 to form a fourth memory level, and the memory The body unit 186 is connected to the bit line U0 and the word line 184 to form the syllabary level L5. The configuration of the polarities of the diodes and the individual configuration of the word lines and bit lines can vary from embodiment to embodiment. In addition, more or less than five memory levels can be used. 123008.doc -15- 200830315 If a p_i-n diode is used as the guiding element of the memory cells in the embodiment of FIG. 3A, the diode of the memory cell 17 can be relative to the first The hierarchical memory unit 152ip_i_n is formed by the diode. For example, if the cell 152 includes an n-type bottom heavily doped region and a P-type top heavily doped region, then in the second-level cell i 7〇, the bottom heavily doped region may be p-type and the top The heavily doped regions are n-type. In an alternate embodiment, an interlevel dielectric can be formed between adjacent memory levels. No conductors are shared between memory levels. Such structures for two-dimensional single stone storage memories are often referred to as a non-mirror structure. In some embodiments, adjacent memory levels of shared conductors and adjacent memory levels of shared conductors may be stacked in the same single-crystal two-dimensional memory array. In other embodiments, some of the conductive systems are shared while other conductors are not shared. For example, in some configurations only these word lines or only those bit lines can be shared. A first memory level can include a memory cell between a bit line level BL0 and a word line level WL. The word lines at level WL0 may be shared to form a unit connected to a second bit line BL1 at a memory level L1. The bit line layers are not shared, so the next layer can include an interlevel dielectric to separate the bit line BL1 from the next level of conductor. This type of configuration is often referred to as semi-mirror. The memory levels do not have to be formed to have similar (four) units. If necessary, the memory level of the material used to change the resistance can be alternated with the memory level of other types of memory cells. The memory array layout of the memory array line is configured as described in the US Patent No. 7, pp. 54, 219 (the name is "for the close spacing application 123008.doc -16 - 200830315, The word line segments are placed using the word line segments placed on the +phan* g sub-layer layer of the array to form the word-mesh by "vertical connection". Each word line resides. On a separate layer and substantially vertically aligned (although on a certain layer, there is a small lateral offset on the display)

字元線可統稱為-列。在_列内的字元線較佳的係共享該 列位址之至4 -部分。同樣,各位元線皆駐留於—分離層 上且實質上垂直對齊(同樣,儘管在某些層上存在較小橫 向偏移)之一群組的位元線可統稱為一行。在一行内的位 元線較佳的係共享該行位址之至少一部分。 圖4係包括一記憶體陣列2〇2之一積體電路之一方塊圖。 記憶體陣列202之陣列終端線包括組織為列的各層字元線 與組織為行的各層位元線。該積體電路2〇〇包括列控制電 路220,該列控制電路22〇之輸出2〇8係連接至該記憶體陣 列202之個別字元線。該列控制電路接收一群組M個列位 址信號與一或多個各種控制信號,而且一般可以包括諸如 列解碼器222、陣列終端驅動器224及區塊選擇電路226之 類兼用於璜取及寫入(即,程式化)操作之電路。該積體電 路200還包括行控制電路21〇,該行控制電路21〇之輸入/輸 出206係連接至該記憶體陣列202之個別位元線。該行控制 電路206接收一群組N個行位址信號與一或多個各種控制信 號,而且一般可以包括諸如行解碼器212、陣列終端接收 器或驅動器214、區塊選擇電路216以及讀取/寫入電路及 I/O多工器之類電路。諸如列控制電路22〇及行控制電路 210之類電路可統稱為控制電路,或因其係連接至該記憶 123008.doc -17- 200830315 體陣列2G2之各個陣列終端而稱為陣列終端電路。 併入一記憶體陣列之積體電路一般將該陣列細分成有時 數目車乂大的子陣列或區塊。可以將區塊進一步一起分組成 包3 (例如)16、32個或一不同數目的區塊之機架。就常用 f月況而S,子陣列係一連續的記憶體單元群組,其具有一 般不會因解碼器、驅動器、感測放大器及輸入/輸出電路 而斷開之連績的字元及位元線。此係基於各種原因中的任 何原因而實行。例如,因字元線及位元線之電阻及電容而 產生的沿該等線橫越之信號延遲(即,RC延遲)在一大陣列 中可能相當明顯。可藉由將一較大陣列細分成一群組較小 的子陣列以使得每一字元線及/或每一位元線之長度減 小,從而減小此等RC延遲。作為另一範例,與存取一群 組圯憶體單元相關聯之功率可以指示在一給定的記憶體循 環期間可以同時存取的記憶體單元數目之一上限。因此, 一較大的記憶體陣列常常係細分成較小的子陣列以減少同 時存取的記憶體單元數目。然而,為便於說明,還可以與 子陣列同義地使用陣列來表示具有一般不因解碼器、驅動 器、感測放大器及輸入/輸出電路而斷開的連續字元及位 元線之一連續的群組記憶體單元。一積體電路可以包括一 或一個以上記憶體陣列。 圖5係說明依據一具體實施例針對在一非揮發性記憶體 系統中之一記憶體單元集合之狀態之電阻分佈之一曲線 圖。圖5所描繪的範例性記憶體系統利用四個電阻狀態, 但可以結合利用不同數目及/或組合的電阻狀態之系統來 123008.doc •18· 200830315 使用依據本揭示内容之具體實施例。以線250來說明該記 憶體單元集合之原始(或初始)狀態。針對此等在製造後處 於其初始狀態的單元之電阻分佈係顯示為基於在一選定電 壓偏壓(例如,2 V)下該單元的傳導電流之一機率函數。該 等單元在製造後之原始狀態係一較高電阻狀態,其在該選 定電壓下具有約10’A至1〇-9Α之一傳導電流。The word lines can be collectively referred to as - columns. Preferably, the word lines in the _ column share the 4-bit portion of the column address. Similarly, the bit lines that reside on the separate layer and are substantially vertically aligned (again, although there is a small lateral offset on some layers) can be collectively referred to as a row. Preferably, the bit lines within a row share at least a portion of the row address. 4 is a block diagram of an integrated circuit including a memory array 2〇2. The array terminal lines of the memory array 202 include layers of word lines organized as columns and bit lines of each layer organized into rows. The integrated circuit 2A includes a column control circuit 220, and the output 2〇8 of the column control circuit 22 is connected to individual word lines of the memory array 202. The column control circuit receives a group of M column address signals and one or more various control signals, and may generally include both a column decoder 222, an array terminal driver 224, and a block selection circuit 226 for both capture and A circuit that writes (ie, stylizes) operations. The integrated circuit 200 further includes a row control circuit 21, and the input/output 206 of the row control circuit 21 is connected to individual bit lines of the memory array 202. The row control circuit 206 receives a set of N row address signals and one or more various control signals, and may generally include, for example, a row decoder 212, an array terminal receiver or driver 214, a block selection circuit 216, and a read. / Write circuits and circuits such as I/O multiplexers. Circuitry such as column control circuit 22 and row control circuit 210 may be collectively referred to as a control circuit or as an array termination circuit because it is connected to each of the array terminals of memory 128008.doc -17-200830315 body array 2G2. Integral circuits incorporated into a memory array typically subdivide the array into sub-arrays or blocks that are sometimes large in number. The blocks can be further grouped together into a rack of packages 3, for example, 16, 32 or a different number of blocks. In the case of S, the sub-array is a continuous group of memory cells with characters and bits that are generally not broken by decoders, drivers, sense amplifiers, and input/output circuits. Yuan line. This is based on any of a variety of reasons. For example, signal delays (i.e., RC delays) that traverse along the lines due to the resistance and capacitance of the word lines and bit lines can be quite significant in a large array. These RC delays can be reduced by subdividing a larger array into a smaller group of sub-arrays such that the length of each word line and/or each bit line is reduced. As another example, the power associated with accessing a group of memory cells can indicate an upper limit on the number of memory cells that can be simultaneously accessed during a given memory cycle. Therefore, a larger memory array is often subdivided into smaller sub-arrays to reduce the number of memory cells accessed simultaneously. However, for ease of illustration, the array may also be used synonymously with the sub-array to represent a continuous group of consecutive characters and bit lines that are generally not broken by the decoder, driver, sense amplifier, and input/output circuitry. Group memory unit. An integrated circuit can include one or more memory arrays. Figure 5 is a graph illustrating a resistance distribution for a state of a set of memory cells in a non-volatile memory system, in accordance with an embodiment. The exemplary memory system depicted in Figure 5 utilizes four resistive states, but may be combined with a system utilizing different numbers and/or combinations of resistive states. 123008.doc • 18· 200830315 uses specific embodiments in accordance with the present disclosure. Line 250 is used to illustrate the original (or initial) state of the set of memory cells. The resistance distribution for the cells in their initial state after fabrication is shown as a function of the probability of the conduction current of the cell at a selected voltage bias (e.g., 2 V). The original state of the cells after fabrication is a relatively high resistance state having a conduction current of about 10'A to 1〇-9Α at the selected voltage.

該裝置之一跳脫狀態係說明於線252。狀態252對應於該 裝置之一最低電阻狀態。處於狀態252之裝置在圖5所說明 的2 V所施加電壓位準下呈現約ι〇·5Α之一傳導電流。在一 具體實施例中,可以藉由跳脫一反熔絲將記憶體單元係從 其最高電阻初始狀態設定為最低電阻跳脫狀態。在其他具 體貝%例中’可以將諸如多晶石夕或一金屬氧化物之類的一 電阻改變材料之一電阻率切換成將單元設定成此較低電阻 狀態。在一具體實施例中,跳脫一反熔絲以將裝置設定成 如線252所說明之一跳脫狀態包括向該等單元施加一較大 的正向偏壓,例如約8 V。還可以針對此等操作使用其他 技術、偏壓條件及/或電壓位準。 線254說明針對該記憶體單元集合在從線252所說明的較 低電阻狀態重新設定為一較高電阻重新設定狀態後之電阻 分佈。處於此重新言史定狀態之記憶體單元在㈣加的Η 電壓位準下呈現約10、至1〇-、之一傳導電流。該重新設 定狀態與該較高電阻初始狀態相比處於—較低電阻,但在 其他具體實施例中可處於—較高電阻。在__具體實施例 中,如下所述之-反向偏壓重新設定操作可用於將該等記 123008.doc -19· 200830315 憶體單元之電阻從狀態252重新設定為狀態254。例如,在 一具體實施例中,可以藉由讓該等記憶體單元經受 約-10 V至-12 V等級之一反向偏壓來增加每一單元中之一 電阻率改變材料之電阻率。 線256說明該等記憶體單元之一設定狀態。可以將記憶 體單元從其較高電阻重新設定狀態254設定為一較低電阻 設定狀態256。處於設定狀態256之記憶體單元在所施加的 2V電壓位準下具有約1〇·6Α之一傳導電流。處於設定狀態 256之單疋電阻比該等單元處於跳脫狀態252時之電阻更 南’但比該等單元處於重新設定狀態254時的電阻更低。 在一具體實施例中可以使用約+8 V之一正向偏壓將一記憶 體單元之電阻從重新設定狀態254切換為設定狀態256。在 其他具體實施例中,可以使用其他偏壓條件及/或電壓位 準來設定該等記憶體單元。 圖5所說明之四個電阻狀態可用於形成各種類型的記憶 體系統。在一具體實施例中,該重新設定狀態轉換係用於 在 次可程式化記憶體陣列中進行一場程式化操作。併 入一電阻改變元件之一記憶體單元係從初始狀態25〇工廠 設定為較低電阻狀態252。接著將包括該記憶體單元之記 憶體陣列提供給一終端使用者。藉由在製造期間將該單元 從其較焉電阻初始狀態進行設定而獲得之較低電阻狀態對 應於該單之一格式化或未程式化狀態。將該記憶體陣列 提供給電路以依據從與該記憶體單元通信之一終端使用者 或主機裝置接收的資料將選定記憶體單元重新設定為較高 123008.doc •20- 200830315 電阻狀態254。 在另一具體實施例中’該等四個電阻狀態係用於形成一 多狀態記憶體系統。可以依據使用者資料將記憶體單元從 初始狀態250程式化為狀態252、254或256中的任一狀態 (或保留於狀態250)。在一此類具體實施例中,每一單元可 以儲存2位元的資料。在另一具體實施例中可以形成一可 重寫記憶體系統。可以將單元設定為狀態256而接著多次 重新設定回到狀態254以形成一位元的可重寫陣列。還可 以依據具體實施例使用其他類型的記憶體系統,以非限制 性範例而言包括:美國專利申請案第_ % (MD-294Y,律師檔案號碼1〇519-141),其名稱為”多用途 記憶體單元及記憶體陣列";美國專利申請案 第--號(MD-296Y律師檔案號碼1〇519-142),其 名稱為”混合使用記憶體陣列";美國專利申請案 第----號(MD-310Y律師檔案號碼10519_149),其 名稱為具有不同資料狀態之混合使用記憶體陣列";以及 美國專利申請案第--一_號(律師擋案號碼]\^- 163 1 ),其名稱為"使用包含具有可調式電阻的可切換半導 體記憶體元件之一記憶體單元的方法"。 -針對讀取、設定或重新設定狀態而偏壓兩終端記憶體單 陣列,可以產生程式干擾、讀取干擾以及可能影響功 ^肖耗以及該等讀取及程式化操作的可靠性之高浪漏電 =例如’當選擇在__陣列内的特定記憶體單元用於特定 呆時,該等偏壓條件可以引起無意中經由未選定記憶體 123008.doc -21 - 200830315 早疋之A漏電流。儘管在該記憶體陣列内使用引導元件, 但:能存在此等、茂漏電流。未選定記憶體單元之二極體在 經文較小的正或負偏壓條件時可能傳導少量的電流。 例如’將在某些兩终端記憶體陣列中實施之—正向偏壓 重新设定操作考量為—抹除操作。在該等!丨導it件從位元 線至字元隸態之條件下,藉由向一選定的位元線施加一 較大電壓而向一潠佘1 、疋的子兀線施加一低電壓或接地條件來 產生#x A㈤正偏壓。未選定的位元線可處於—較小正偏 壓而未選定的字it線處於—較大正偏壓。在以此方式偏壓 該::體陣列之條件下,在某些情況下可能存在經由沿該 ^定字元線或位元線之半選定的單元以及經由沿—未選定 的子元線及位it線之未選定單元的不可接受位準之電流茂 、、同樣在正向偏壓設定操作(可用於程式化一記憶 體早兀陣列)期間可能發生不可接受位準之洩漏電流。經 由未選定單I之較小$漏電流之累積效應限制在—時間可 以操作的選定記憶體單元之數目。 員發現,可以向具有電阻改變元件之記憶體單元施加一 反向偏壓以改變該單元之一可偵測的電阻。例如,可以藉 由讓諸如上述金屬氧化物、多晶矽之類材料經受一產生橫 跨該材料之一反向偏壓的電壓脈衝,來將此類材料從一較 低電阻率狀態重新設定為一較高電阻率狀態。在一具體實 中在一重新設定操作期間施加一反向偏壓以使得經 由該°己丨思體陣列之洩漏電流最小化。在某些實施方案中, 可以將一 士讲 、 質上為零之偏壓提供給特定的未選定記憶體單 123008.doc -22- 200830315 疋。由於該等洩漏電流係最小化,因此可以選擇更大數目 的記憶體單元用於重新設定操作。此藉由減小程式化及/ 或抹除時間而提供操作說明書之一改良。此外,該等低洩 漏電流可以藉由使得裝置性能正規化於預期位準内來促進 更可靠的操作。名稱為”併入可逆極性的字元線及位元線 解碼器之被動元件記憶體陣列”之美國專利申請案 第-—----號(MD-273律師檔案號碼023-0048)揭示可 用於使得經由未選定及半選定的記憶體單元之洩漏電流最 小化之一反向偏壓操作。 圖6係依據一具體實施例在一反向偏壓操作期間一記憶 體陣列之一部分之一電路圖。該等反向偏壓條件可用於將 記憶體單元設定為一低電阻狀態或將記憶體單元重新設定 為一高電阻狀態。下面可為方便起見而對一重新設定操作 作一特定參考,但此並不表示對應用所揭示的偏壓及技術 之一限制。一或多個選定字元線處於一正偏壓而一或多個 選疋位元線處於一負偏壓。例如,該等選定字元線可以接 收+1/2 VRR之一重新設定電壓信號Vwr,而透過重新設定 電壓信號VBR以約-1/2 VRR之一負偏壓驅動該等選定位元 線。VRR係重新設定該記憶體所需要的反向偏壓(或負電 壓)數里並且可以隨具體實施例而變化。在一範例性實施 方案中,VRR約為12 V而使得該等選定字元線接收+6V而 該等選定位元線接收-6 V來產生12V反向偏壓位準。未選 定的字元線及位元線皆係接地。用於該等選定記憶體單元 (表示為S)之引導元件受到反向偏壓,而讓一反向電流穿 123008.doc -23 - 200830315 過用於該等選定單元之電阻改變材料。在此反向偏壓條件 下,該電阻改變材料從一第一電阻狀態切換到一第二電阻 狀恶。圖6所說明之偏壓條件有利地提供針對該等未選定 單疋(表示為U)之一零偏壓條件。因此,可以獲得在程式 操作期間經由未選定與半選定的記憶體單元之低洩漏電 流。F表示沿一選定位元線之半選定的記憶體單元,而η表 示/〇選疋予元線之半選定的記憶體單元。此外,針對該 7選定陣列線之+Λ 1/2Vrr的選定位準在該驅動器電路上 提供幸又小的負載便可產生用於該反向偏壓重新設定操作的 電壓位準之需要。藉由橫跨該等陣列線使用正與負電壓位 準來分割該等偏壓,該驅動器電路僅需要產生在某些實施 方案中所需要的總電壓位準之一半。 遇可以使用其他偏壓條件來反向偏壓該等選定記憶體單 元以進行-重新設定操作。例如’在一具體實施例中,可 以向接地的選定字元線及選定位元線施加一正電壓偏壓 (例如’ VRR)。未選定的字元及位元線各可以接收 + 1/2VRR。此偏壓情形還將向選定的記憶體單元提供一反 向偏屢’該反向偏壓可心在—設定操作後將該等單元重 新設定回到較高電阻狀態。關於反向偏壓操作之更多資 訊’請見美國專利申請案第 彔乐—-一__號(MD-273律 師檔案號碼023顧8),其錢為"併人可逆極㈣字元線 及位π線解碼器之被動元件記憶體陣列,,。 /可能在某些記憶體實施方案中’針對-處於-重新設 疋狀怨的記憶體早疋陣列之—電阻分佈可能過寬或者包括 123008.doc -24 - 200830315 與所需範圍相比之一更大範圍的電阻。例如,針對處於圖 5之線252所示重新設定狀態中的記憶體單元之電阻分佈包 括一相對較大範圍的電阻。某些記憶體單元在所施加的電 壓位準下呈現約1〇-1之一傳導電流,而其他處於相同實 體狀恶之記憶體單元呈現約10·7α之一較大傳導電流。此 等傳導電流證明在皆希望處於相同重新設定實體狀態的單 元之間存在一較大的電阻差異。重要的係,深度重新設定 為一極高電阻的記憶體單元(較接近一 1〇·8Α之傳導電流的 該些圮憶體單元)與處於原始或初始狀態25〇的記憶體單元 並無較寬之分離。在某些實施方案中,此等兩個實體狀態 之間缺少限度可能會在讀取及寫入操作期間證明有問題。 較大範圍的電阻可能導致對儲存於該等記憶體單元中的資 料之錯誤讀數。例如,深度重新設定為一極高電阻之單元 ^施加-讀取參考電壓之情況下彳能不會充分傳導,從而 指示其處於重新設定實體狀態。可能會誤認為此等單元處 於初始或原始狀態250。在諸如程式化之類的其他操作期 F:可此在一驗證步驟期間不正確地讀取此等單元,從而 導致錯誤施加可能不需要的額外程式化電壓。 —依據一具體實施例’―調整操作制於或併人—重新設 疋操作’以針對處於_重新設定狀態之記憶體單元提供一 較小的電阻分佈。可以在記憶體單元重新設定後向其施加 調整偏壓來朝與該等重新設定單元相關聯之—所 標位準偏移θ_ 〆 —設定的^。在重新設㈣具有比 的電阻之記憶體單$可以使其電阻降低以 123008.doc •25- 200830315 與處於該重新设定狀態之其他單元更相近地匹配。 在-具體實施例中使用—自我限制反向偏壓調整操作。 頃發現’向記憶體單元施加一小反向偏壓可以增加其電 阻,而不像在針對該重新設定操作施加一較大反向偏壓之 ★ 情況下一樣減小其電阻。 , 圖5說明將一反向偏麼調整操作用於-反向偏壓重新設 定操作之效果°線258表示處於該重新設定狀態2M的記憶 • 體單元之一目標電阻(或所需平均電阻)。不同的電阻材料 可以依據其個別特性提供不同的目標電阻位準。針對該重 新設定狀態之目標電阻可由傾向於指#該材料之一在最自 然情況下的重新設定位準之此等特性決定。特定數目之單 元使其電阻低於圖5中的目標位準,而其他單元使其電阻 高於該目標位準。施加比該反向偏壓重新設定位準Vrr更 低之-數量的反向偏壓,從而導、致具有高於所需位準258 ^-電阻的記憶體單元偏移至一較低電阻。頃發現,相對 • 冑小數量的反向偏壓會以一自我限制的方式減小電阻改變 材料之電阻。具有一處於或高於一特定位準的電阻之單元 在經受-較小反向偏壓時會令其電阻增加。但是,低於該 • 肖定電阻位準之單元不受該反向偏壓之影響。在一實财 . 帛中’已顯示針對該反向偏壓之-範ϋ内的位準對單元電 阻產生類似的影響°例如,在從狀態252至狀態254之一重 =設定操作期間’約1〇ν至12V之一反向偏壓可以將一選 定記憶體單元之電阻增加約如圖5所示之數量。等於在該 重新設定操作期間所施加的反向偏壓約50%至60%(例如, 123008.doc -26- 200830315 如,6 V至7 V)之一反向偏壓可用於增加該等記憶體單元 中電阻高於一特定位準之特定記憶體單元之電阻。 在該調整操作中的電阻增加具有自我限制性,因為在達 到特足的電阻數篁時該等單元停止減小電阻。此外,僅具 有高於臨界位準之一電阻的該些單元受該調整操作之影 響。已經處於適當電阻範圍之單元不會經歷一電阻偏移, 即使其經受該反向偏壓調整電壓亦如此。因此,如圖5所One of the trip states of the device is illustrated on line 252. State 252 corresponds to one of the lowest resistance states of the device. The device in state 252 exhibits a conduction current of about 1 〇 5 在 at the voltage level of 2 V as illustrated in FIG. In one embodiment, the memory cell can be set from its highest resistance initial state to the lowest resistance trip state by tripping an antifuse. In other specific examples, the resistivity of a resistance changing material such as polycrystalline or a metal oxide can be switched to set the cell to this lower resistance state. In one embodiment, tripping an antifuse to set the device to one of the trip states as illustrated by line 252 includes applying a greater forward bias to the cells, such as about 8 volts. Other techniques, bias conditions, and/or voltage levels can also be used for such operations. Line 254 illustrates the resistance distribution for the set of memory cells after resetting to a higher resistance reset state from the lower resistance state illustrated by line 252. The memory cell in this re-stated state exhibits a conduction current of about 10 to 1 〇- at the (iv) plus 电压 voltage level. The reset state is at - lower resistance than the higher resistance initial state, but may be at - higher resistance in other embodiments. In a particular embodiment, a reverse bias reset operation as described below can be used to reset the resistance of the memory cell from state 252 to state 254. For example, in one embodiment, the resistivity of one of the resistivity changing materials in each cell can be increased by subjecting the memory cells to a reverse bias of one of about -10 V to -12 V. Line 256 illustrates one of the set states of the memory cells. The memory cell can be set from its higher resistance reset state 254 to a lower resistance set state 256. The memory cell in the set state 256 has a conduction current of about 1 〇 6 在 at the applied 2 V voltage level. The single turn resistance in the set state 256 is less than the resistance of the cells in the trip state 252 but lower than the resistance of the cells in the reset state 254. In one embodiment, the resistance of a memory cell can be switched from reset state 254 to set state 256 using a forward bias of about +8 V. In other embodiments, other bias conditions and/or voltage levels can be used to set the memory cells. The four resistance states illustrated in Figure 5 can be used to form various types of memory systems. In one embodiment, the reset state transition is used to perform a stylized operation in the secondary programmable memory array. One of the memory cells of a resistance change element is set from the initial state 25 〇 factory to the lower resistance state 252. The memory array including the memory unit is then provided to an end user. The lower resistance state obtained by setting the unit from its initial state of the resistor during manufacture corresponds to the single formatted or unprogrammed state. The memory array is provided to circuitry to reset the selected memory unit to a higher resistance state 254 in accordance with data received from an end user or host device in communication with the memory unit. In another embodiment, the four resistance states are used to form a multi-state memory system. The memory unit can be programmed from the initial state 250 to any of the states 252, 254, or 256 (or retained in state 250) based on the user profile. In one such embodiment, each unit can store 2 bits of data. In another embodiment, a rewritable memory system can be formed. The cell can be set to state 256 and then reset multiple times back to state 254 to form a one-bit rewritable array. Other types of memory systems may also be used in accordance with specific embodiments, including, by way of non-limiting example: US Patent Application No. _% (MD-294Y, attorney docket number 1〇519-141), entitled "Multiple Use memory unit and memory array "; US Patent Application No. - (MD-296Y Lawyer File Number 1〇519-142), the name is "mixed memory array"; US patent application ---- (MD-310Y lawyer file number 10519_149), the name is a mixed memory array with different data status "; and the US patent application No. - 1 (lawyer file number) \ ^ - 163 1), the name is "Method of using a memory cell containing one of the switchable semiconductor memory elements with adjustable resistance". - Biasing a single-array memory array for reading, setting, or resetting states, can cause program disturb, read interference, and high-wave leakage that may affect the power consumption and reliability of such read and program operations = For example, when a particular memory cell within the __ array is selected for a particular stay, the bias conditions can cause an A leakage current that is inadvertently passed through unselected memory 123008.doc -21 - 200830315. Although a guiding element is used within the memory array, there can be such a leakage current. A diode of a memory cell that is not selected may conduct a small amount of current when subjected to a small positive or negative bias condition. For example, the forward bias reset operation will be implemented in some two-terminal memory arrays as an erase operation. In these! Applying a low voltage or grounding condition to a sub-turned line of a 疋1, 疋 by applying a large voltage to a selected bit line under the condition of a bit line to a character state. To generate a #x A (five) positive bias. The unselected bit line can be at - a smaller positive bias and the unselected word it is at - a larger positive bias. In the case of biasing the :: body array in this manner, in some cases there may be via selected cells along the half of the word line or bit line and via the unselected sub-line and An unacceptable level of current in the unselected cells of the bit line, and an unacceptable level of leakage current during the forward bias setting operation (which can be used to program a memory array). The cumulative effect of the smaller $leak current by the unselected single I is limited to the number of selected memory cells that can be operated at time. It has been found that a reverse bias can be applied to a memory cell having a resistance changing element to change the detectable resistance of one of the cells. For example, such materials can be reset from a lower resistivity state to a lower voltage by subjecting a material such as the above-described metal oxide, polysilicon, to a voltage pulse that produces a reverse bias across one of the materials. High resistivity state. In a specific implementation, a reverse bias is applied during a reset operation to minimize leakage current through the array. In some embodiments, a one-way, zero-weight bias can be provided to a particular unselected memory list 123008.doc -22- 200830315 疋. Since these leakage currents are minimized, a larger number of memory cells can be selected for resetting operations. This provides an improvement in the operating instructions by reducing the stylization and/or erasing time. Moreover, such low leakage currents can facilitate more reliable operation by normalizing device performance to an expected level. U.S. Patent Application No.----- (MD-273 Lawyer File Number 023-0048) entitled "Positive Element Memory Array Incorporating Reversible Polar Character Lines and Bit Line Decoders" discloses that Reverse bias operation is minimized by minimizing leakage current through unselected and semi-selected memory cells. Figure 6 is a circuit diagram of a portion of a memory array during a reverse bias operation in accordance with an embodiment. These reverse bias conditions can be used to set the memory cell to a low resistance state or to reset the memory cell to a high resistance state. A specific reference to a reset operation may be made below for convenience, but this does not represent one of the limitations of the bias and technique disclosed in the application. One or more selected word lines are at a positive bias and one or more of the selected bit lines are at a negative bias. For example, the selected word lines can receive one of +1/2 VRR to reset the voltage signal Vwr, and the reset voltage signal VBR drives the selected positioning line with a negative bias of about -1/2 VRR. The VRR resets the number of reverse bias (or negative voltage) required for the memory and can vary from embodiment to embodiment. In an exemplary embodiment, the VRR is about 12 volts such that the selected word line receives +6 volts and the selected locating element lines receive -6 volts to produce a 12 volt reverse bias level. Unselected word lines and bit lines are grounded. The guiding elements for the selected memory cells (denoted as S) are reverse biased, and a reverse current is passed through the resistance changing material for the selected cells. Under this reverse bias condition, the resistance changes material from a first resistance state to a second resistance. The bias conditions illustrated in Figure 6 advantageously provide a zero bias condition for one of the unselected single turns (denoted U). Thus, low leakage current through unselected and semi-selected memory cells during program operation can be obtained. F denotes a selected memory cell along a half of the selected bit line, and η denotes/selects a half selected memory cell of the element line. In addition, the selection of + Λ 1/2 Vrr for the selected array line of the 7th requires a small enough load on the driver circuit to generate the voltage level for the reverse bias reset operation. By dividing the bias voltages across the array lines using positive and negative voltage levels, the driver circuit only needs to produce one-half of the total voltage level required in certain embodiments. Other selected bias conditions can be used to reverse bias the selected memory cells for the -reset operation. For example, in a particular embodiment, a positive voltage bias (e.g., 'VRR) can be applied to selected ground and selected bit lines of ground. Unselected characters and bit lines can each receive + 1/2VRR. This biasing condition will also provide a reverse bias to the selected memory cells. The reverse bias can be reset to return to the higher resistance state after the set operation. For more information on reverse bias operation, please see the US patent application No. _ __ (MD-273 lawyer file number 023 Gu 8), the money is "and the reversible pole (four) characters Passive component memory array for line and bit π line decoders,. / May be in some memory implementations 'targeted-in-re-arranged memory early-array arrays—the resistance distribution may be too wide or include 123008.doc -24 - 200830315 compared to the desired range A wider range of resistors. For example, the resistance distribution for a memory cell in the reset state shown in line 252 of Figure 5 includes a relatively large range of resistance. Some memory cells exhibit a conduction current of about 1 〇-1 at the applied voltage level, while other memory cells in the same solid state exhibit a large conduction current of about 10·7α. These conduction currents demonstrate a large difference in resistance between cells that wish to be in the same reset physical state. Importantly, the depth is reset to a very high-resistance memory cell (the memory cells that are closer to a conduction current of 1〇·8Α) than the memory cells in the original or initial state of 25〇. The separation of the width. In some embodiments, the lack of limits between these two entity states may prove problematic during read and write operations. A larger range of resistance may result in erroneous readings of the data stored in the memory cells. For example, if the depth is reset to a very high resistance unit, the 彳 can not be sufficiently conducted in the case of applying-reading the reference voltage, thereby indicating that it is in the reset physical state. It may be mistaken that these units are in the initial or original state 250. During other operational periods such as stylization F: This may be incorrectly read during a verification step, resulting in an erroneous application of an additional stylized voltage that may not be needed. - According to a specific embodiment, the "adjustment operation or the coincidence-reset operation" provides a smaller resistance distribution for the memory unit in the _reset state. After the memory unit is reset, an adjustment bias can be applied thereto to the set target offset θ_ 〆 - associated with the reset unit. Resetting the memory of the (4) resistor with a ratio of $ can cause its resistance to decrease to 123008.doc •25- 200830315 to match more closely with other units in the reset state. Used in a particular embodiment - self-limiting reverse bias adjustment operation. It has been found that 'applying a small reverse bias voltage to the memory cell increases its resistance without reducing its resistance as in the case of applying a large reverse bias for the reset operation. Figure 5 illustrates the effect of a reverse bias adjustment operation for the -reverse bias reset operation. Line 258 indicates the target resistance (or desired average resistance) of the memory unit in the reset state 2M. . Different resistive materials can provide different target resistance levels depending on their individual characteristics. The target resistance for this reset state can be determined by such characteristics that tend to refer to the resetting level of one of the materials in the most natural case. A certain number of cells have their resistance lower than the target level in Figure 5, while other cells have their resistance above the target level. A reverse bias of a number lower than the reverse bias reset level Vrr is applied, thereby causing the memory cell having a resistance higher than the desired level 258^-resist to shift to a lower resistance. It has been found that a relatively small amount of reverse bias reduces the resistance of the material in a self-limiting manner. A cell having a resistance at or above a particular level will increase its resistance when subjected to a small reverse bias. However, cells below the • Shore fixed resistance level are not affected by this reverse bias. In a real money. 已 'has been shown to have a similar effect on the cell resistance for the level of the reverse bias voltage. For example, during the operation from state 252 to state 254 = set operation period 'about 1 A reverse bias of 〇ν to 12V can increase the resistance of a selected memory cell by an amount as shown in FIG. A reverse bias voltage equal to about 50% to 60% of the reverse bias applied during the reset operation (eg, 123008.doc -26-200830315, eg, 6 V to 7 V) can be used to increase the memory The resistance of a particular memory cell in the bulk cell that is above a certain level of resistance. The increase in resistance in this adjustment operation is self-limiting because the cells stop reducing the resistance when a particularly large number of resistors is reached. In addition, the cells having only one resistance above the critical level are affected by the adjustment operation. A cell that is already in the proper resistance range does not experience a resistance offset, even if it is subjected to the reverse bias regulation voltage. Therefore, as shown in Figure 5

不具有一低於該目標位準258的電阻之該些單元不受針對 該調整操作的反向偏壓之影響。 圖7係說明依據一具體實施例針對一非揮發性記憶體系 統的重新設定狀態之一反向偏壓調整操作之一電路圖。以 一正電壓位準+vTT將一調整電壓信號Vwt提供給一或多個 選定的字元線以施加一用於該調整操作之反向偏壓。在一 具體實施例中,VTT係在調整期間向該等選定單元施加的 反向偏壓數量。在圖7所示偏壓情形中,選定的位元線係 接地,因此以+VTT將完全數量的調整偏壓施加於選定的字 元線,以ά生與此數量相等之一橫跨每一選定單元的反向 偏壓。在-具體實施例,,調整反向偏壓之數量〜等於 VRR的總反向重新設定偏壓位準之約6〇%。繼續結合以上 範例,若該反向偏壓重新設定電位等於約12v,則該反向 偏壓調整位準νττ可約為6¥或7¥。因此,在一具=實施 例中向該等選定字元線施加+6 V或+7 V。 八 、 依據本揭示内容之-或多項具體實施例中可以針對一調 整操作使用其他偏壓條件。例如,在—具體實施例中,向 123008.doc -27- 200830315 該等選定的字元線施M+1/2VTT之一電壓而向該等選定的 位元線施加-1/2 VTT之一電壓《橫跨每一單元所產生的反 向偏壓與先前所述之(VTT)相同,但是,已橫跨不同類型的 陣列線而分佈個別偏壓。 由於僅該些需要電阻偏移之單元受該反向偏壓調整之影 響,因此該操作肖資料無關而可以係針對高頻£應用來實 施。在一具體實施例中使用一高頻寬操作來同時調整一較 一次性針對該調整操The cells that do not have a resistance below the target level 258 are unaffected by the reverse bias for the adjustment operation. Figure 7 is a circuit diagram showing one of the reverse bias adjustment operations for a non-volatile memory system reset state in accordance with an embodiment. An adjusted voltage signal Vwt is supplied to one or more selected word lines at a positive voltage level + vTT to apply a reverse bias for the adjustment operation. In a specific embodiment, the VTT is the amount of reverse bias applied to the selected cells during the adjustment. In the biasing case shown in Figure 7, the selected bit line is grounded, so a full amount of adjustment bias is applied to the selected word line at +VTT to produce one of the same number as this one. Reverse bias of the selected unit. In a particular embodiment, the number of reverse biases is adjusted to be equal to about 6% of the total reverse reset bias level of the VRR. Continuing with the above example, if the reverse bias reset potential is equal to about 12v, the reverse bias adjustment level νττ can be about 6¥ or 7¥. Thus, in a = embodiment, +6 V or +7 V is applied to the selected word lines. Eight, other biasing conditions may be used for an adjustment operation in accordance with or in various embodiments of the present disclosure. For example, in a specific embodiment, one of M+1/2VTT is applied to the selected word lines of 123008.doc -27-200830315 and one of the -1/2 VTTs is applied to the selected bit lines. The voltage "reverse bias generated across each cell is the same as previously described (VTT), however, individual biases have been distributed across different types of array lines. Since only those cells requiring resistance offset are affected by the reverse bias adjustment, the operation is independent of the data and can be implemented for high frequency £ applications. In a specific embodiment, a high frequency wide operation is used to simultaneously adjust a one-time adjustment for the adjustment operation.

大群組的記憶體單元。在一範例中 選擇來自該區塊 作選擇一記憶體單元區塊之各個位元線。 之一字7G線並針對每一字元線重複該操作。此外,可以在 該調整操作期間同_選擇在一 P車列内❹㈣塊,但在一 具,實施例中選擇-單-區塊。利用此技術,—次性調整 大量單元以不對該重新設定操作之頻寬產生不合理的影 響。在其他具體實施例中,可使用其他群組。在_具體實 鉍例中,可以選擇來自橫跨該陣列的多個區塊之一或多個 位元線及一或多個字元線。 圖8係依據一具體實施例用於重新設定一記憶體單元之 方法之一流程圖,其併入一反向偏壓調整操作。在步驟 300中用於一記憶體單元陣列之列與行控制電路接收指定 用於重新設定的選定單元之位址及控制資訊。例如,在一 可重寫陣财可以接㈣對該等選定單元之_抹除請求, 而在2多狀態陣列中可以接收一寫入請求。滿足該寫入或 抹除明求可包括如圖所示重新設定該等選定單元。 /驟302中,向重新設定的單元施加一或多個重新設 123008.doc -28- 200830315 定電壓脈衝信號。可以藉由如上所述在該等選定字元及位 凡線上使用-電壓组合來橫跨該等選定單元施加該反向重 新設定偏壓。例如’在-具體實施例中該字元線電壓信號 、可以包括一正電壓脈衝(例如,+1/2Vrr)而該位元線電 ’ 壓信號VBR包括一負電壓脈衝(例如,-1/2Vrr)以反向偏壓 - 該等選定单元。 在步驟304中,若仍有欲重新設定之記憶體單元,則該 • 方法返回施加一(或多個)額外脈衝,或者,若所有欲重新 設定之單元皆已接收到一重新設定電壓脈衝則繼續到步驟 3〇6。在一具體實施例中,可以讓該陣列之較小部分經受 個別的重新設定脈衝以使得經由半選定或未選定單元之洩 漏電流最小化。例如,在一實施方案中,步驟3〇2與3〇4之 每一迭代可以向來自該記憶體内的許多機架(例如,16至 2024個機架或更多)之每一機架之一(或者在其他情況下的 一個以上)區塊施加一脈衝直至每一選定位元線皆已接收 馨到一重新設定電壓脈衝。在其他具體實施例中,可以在步 驟302中選擇其他數目之位元線及/或字元線。名稱為"併 入用於記憶體陣列區塊選擇的兩個資料匯流排之記憶體陣 • 列”之美國專利申請案第__號(MD-3〇3律師檔 • 案號碼023_0052)及名稱為"用於區塊可選擇的記憶體陣列 之階層式位元線偏壓匯流排”之美國專利申請案 第--—號(MD-307律師檔案號碼023-0053)說明用 於對一記憶體陣列(例如陣列302)進行增加的平行存取之技 術。 123008.doc -29- 200830315 在向母選定記憶體單元施加一反向偏壓電壓脈衝後, 藉由在步驟306中讀回該等裝置之電阻狀態來執行一驗證 操作。步驟306可以包括決定一記憶體單元之電阻是否已 〜加到處於或高於一最小臨界電阻。在步驟306中可以使 用包括感測一選定單元在一組參考偏壓條件下的電流或電 ^ 之各種技術來決定是否充分地重新設定一記憶體單 兀。名稱為"用以讀取一多層級被動元件記憶體單元陣列 φ 之設備,,之美國專利申請案第__號(MD-274律 師檔案號碼023_0049)說明可用於讓一讀回操作驗證該重新 。又定狀恶之合適的讀取技術。在步驟3〇8中,該重新設定 操作針對具有一未充分重新設定的記憶體單元之該些位元 線而分支進行。在選用步驟31Θ中向利用該等字元及/或位 兀線電壓信號VwR及Vbr的該些記憶體單元施加一重試脈 衝在具體實施例中向具有一未充分重新設定的記憶體 單元之各個位疋線同時施加該脈衝。可以使用該等位元線 籲 《各種群組而施加個別脈衝。在_具體實施例中,不施加 任何重試脈衝。若使用一重試脈衝,則在步驟312中針對 該等單70執仃一驗證操作。若在步驟314中決定未充分重 • 新"又定、、二歷該重试操作之單元,則在步驟3 16中使用錯誤 • &正碼來管理該等單元或以冗餘的記憶體單元來加以替換 等等。 由於在步驟302及3 14中施加的重新設定電壓,因此可能 已如上所述將特疋單疋深度重新設定為一高電阻狀態。針 對該重新設定狀態之一更緊密的電阻分佈將在狀態之間提 123008.doc -30- 200830315 供一更大的限度,而因此提供一更可靠的裝置。因此,在 步驟308及314中成功驗證單元後或在步驟316中處置未重 新設定的任何單元後,在步驟3 18中針對欲重新設定的單 元執行一調整操作。 在一具體實施例中針對經歷該重新設定操作的各個單元 同時執行步驟3 18。由於該操作係自我限制性,因此還可 以讓未完全或深度重新設定之單元經受該調整偏壓而不會 有負面影響。此等單元不會經歷進一步的電阻偏移。此 外,調整至一較低電阻的該些單元在其到達與該調整偏壓 相關聯之一位準時將停止其電阻變化。 如上所述,在一具體實施例中向該等單元施加一反向偏 壓調整電壓VTT。在一具體實施例中施加一較小數量的反 向偏壓以減小在一調整操作期間的電阻而非增加在一重新 設定操作期間的電阻。在向每一單元施加該反向偏壓調整 電壓VTT後,在步驟320中完成該重新設定操作。可以依據 具體實施例對圖8所示方法作諸多變化。例如,在針對所 有重新設定的單元完成步驟3 02及3 04之後而在步驟306中 進行驗證之前,可以併入步驟3 18中的調整操作。 圖9A說明亦可用於施加圖6之反向偏壓重新設定條件之 列控制電路220之一部分之一具體實施例。列解碼器422在 該重新設定脈衝期間對應於一選定字元線並向 NMOS/PMOS字元線驅動器電路(例如,圖4中的224)輸出 接地。針對該驅動器電路之接地輸入開啟上部PMOS裝置 402及404。該接地輸入使得該驅動器電路分別將反向源極 123008.doc •31- 200830315 選擇匯流排信號vWR及_傳遞至該選定字元線及與解碼 器似相關聯之每半選定字元線。對應於一未選定字元線 的每一列解碼益423向其個別驅動器電路輸出v皿,如圖 9B所示。VwR之正偏壓開啟該等未選定字元線之驅動器電 路之NMOS裝置416及418。據此,選擇源極選擇匯流排位 ' 準(兩者皆為GND)並在每一對應的未選定字元線上加以驅 動。在一具體實施例中,如前所述,該字元線反向重新設 φ 定電壓VwR等於約+1/2 VRR。VWR還可以提供其他電壓位 準。例如,可以針對該重新設定操作提供如下所述具有一 傾斜脈衝(例如,開始k+1/2Vrr而然後增加)之一或多個 反向重新設定電壓脈衝。 圖10A及10B係可用於施加針對該反向重新設定操作的 偏壓條件之行控制電路210之一部分之電路圖。行解碼器 5 12控制一選定的位元線驅動器來提供選定的位元線電壓 脈衝VBR。在一具體實施例中,vBR提供_1/2 Vrr之一電壓 籲 脈衝。行解碼器5 12可以係橫跨多個位元線驅動器(例如, 24個)而共旱而且還在即將施加該重新設定脈衝之前將該 等半選定的位元線連接至接地偏壓。在施加該脈衝期間, • 該等半選定的位元線浮動於接地附近。半選定的位元線上 • 大量未選定單元提供使得半選定的位元線保持為接近接地 之一洩漏電流。在一具體實施例中,在一重新設定操作期 間與該選定位元線共享一行解碼器之記憶體單元可以係半 選疋的δ己憶體早元。例如’在該重新設定操作期間該等單 元可以連接至該選定字元線。該選定行解碼器512將(}1^) 123008.doc -32- 200830315 輸出至用於該行解碼器的驅動器電路之輸入。在該驅動器 電路之NMOS/PMOS對處的GND輸入將開啟下部NMOS裝 置506。將該反向源極選擇匯流排位準VBR傳遞至該選定位 元線。未選定的列解碼器5 13向其個別驅動器電路之閘極 提供VBr,從而選擇在每一驅動器對的頂部之PMOS裝置。 將源極選擇匯流排信號位準(皆處於GND)提供給對應於解 碼器5 13的每一未選定字元線。 圖11A說明可用於施加圖7之反向偏壓調整條件之列控制 電路220之一部分之一具體實施例。選定的列解碼器422向 該NMOS/PMOS字元線驅動器電路輸出一字元線調整電壓 脈衝Vwt。Vwt係一正電壓並開啟下部NMOS裝置406及 408。該驅動器電路將該源極選擇匯流排信號VWT傳遞至該 等選定的字元線。每一未選定的列解碼器423向其個別的 驅動器電路輸出GND,如圖11B所示。開啟上部PMOS裝置 412及414,並將來自該反向源極選擇匯流排之GND信號傳 遞給每一未選定的字元線。 圖10 A係可用於施加針對該反向偏壓調整操作的偏壓條 件之行控制電路210之一部分之一電路圖。選定的行解碼 器512控制一選定的位元線驅動器,開啟該等上部PMOS裝 置並將GND傳遞給選定用於該重新設定操作之每一位元 線。該調整操作與資料無關,而在該操作之該自我限制性 質之條件下可以一次性選擇大量單元。因此,在一選定區 塊中的每一位元線接收該GND位準信號以施加該反向偏壓 調整電壓位準。 123008.doc -33- 200830315Large group of memory cells. In an example, the individual bit lines from the block are selected for selecting a memory cell block. One word 7G line and repeat this operation for each word line. In addition, a (four) block can be selected in a P train during the adjustment operation, but in one embodiment, a single-block is selected. With this technique, a large number of units are sub-adjusted so as not to unreasonably affect the bandwidth of the reset operation. In other embodiments, other groups may be used. In a particular embodiment, one or more bit lines and one or more word lines from a plurality of blocks across the array may be selected. Figure 8 is a flow diagram of a method for resetting a memory cell incorporating a reverse bias adjustment operation in accordance with an embodiment. The step and row control circuitry for a memory cell array in step 300 receives the address and control information for the selected cell designated for resetting. For example, a rewritable array can receive (d) an erroneous request for the selected cells, and a write request can be received in the 2 multi-state array. Satisfying the write or erase request may include resetting the selected cells as shown. / In step 302, one or more reset voltage signals of 123008.doc -28-200830315 are applied to the reset unit. The reverse reset bias can be applied across the selected cells by using a voltage combination on the selected characters and the bit lines as described above. For example, in a particular embodiment, the word line voltage signal can include a positive voltage pulse (e.g., +1/2Vrr) and the bit line voltage 'voltage signal VBR includes a negative voltage pulse (e.g., -1/). 2Vrr) with reverse bias - these selected units. In step 304, if there is still a memory unit to be reset, the method returns one or more additional pulses, or if all the units to be reset have received a reset voltage pulse. Continue to step 3〇6. In a specific embodiment, a smaller portion of the array can be subjected to individual reset pulses to minimize leakage current through the semi-selected or unselected cells. For example, in one embodiment, each iteration of steps 3〇2 and 3〇4 can be directed to each of a plurality of racks (eg, 16 to 2024 racks or more) from the memory. A pulse (or in more than one of the other cases) applies a pulse until each selected bit line has received a reset voltage pulse. In other embodiments, other numbers of bit lines and/or word lines may be selected in step 302. U.S. Patent Application No. __ (MD-3〇3 Lawyer File No. 023_0052), which is incorporated in the Memory Array of Columns for Memory Array Block Selection. US Patent Application No.--- (MD-307 Lawyer File Number 023-0053) for the name "hierarchical bit line bias bus for block selectable memory arrays A memory array (e.g., array 302) performs techniques for increased parallel access. 123008.doc -29- 200830315 After applying a reverse bias voltage pulse to the selected memory cell, a verify operation is performed by reading back the resistance states of the devices in step 306. Step 306 can include determining whether the resistance of a memory cell has been added to be at or above a minimum critical resistance. Various techniques, including sensing the current or current of a selected cell under a set of reference bias conditions, can be used in step 306 to determine whether to fully reset a memory bank. The device has the name "" for reading a multi-level passive component memory cell array φ, and the US Patent Application No. __ (MD-274 Lawyer File Number 023_0049) description can be used to verify the readback operation. Re-. It is also suitable for reading techniques. In step 3-8, the reset operation is branched for the bit lines having a memory unit that is not sufficiently reset. Applying a retry pulse to the memory cells utilizing the characters and/or bit line voltage signals VwR and Vbr in an optional step 31A, in a particular embodiment, to each of the memory cells having an insufficient reset The pulse is applied simultaneously to the bit line. This bit line can be used to apply individual pulses to various groups. In the particular embodiment, no retry pulses are applied. If a retry pulse is used, then a verification operation is performed for the orders 70 in step 312. If it is determined in step 314 that the unit of the retry operation is not fully re-defined, and the error is corrected in step 3 16 , the unit is managed or the redundant memory is used. Body units are replaced and so on. Due to the reset voltage applied in steps 302 and 314, it is possible to reset the characteristic single-turn depth to a high resistance state as described above. A tighter resistance distribution of the needle to one of the reset states will provide a greater limit between the states 123008.doc -30-200830315, thus providing a more reliable device. Therefore, after successfully verifying the unit in steps 308 and 314 or disposing of any unit that has not been reset in step 316, an adjustment operation is performed in step 3 18 for the unit to be reset. In a specific embodiment, step 3 18 is performed simultaneously for each unit undergoing the reset operation. Since the operation is self-limiting, it is also possible to subject the unit that is not fully or deeply reset to the adjustment bias without negative effects. These units will not experience further resistance shifts. In addition, the cells adjusted to a lower resistance will stop their resistance change when they reach a level associated with the adjustment bias. As described above, in a specific embodiment, a reverse bias voltage VTT is applied to the cells. A small amount of reverse bias is applied in a particular embodiment to reduce the resistance during an adjustment operation rather than increasing the resistance during a reset operation. After the reverse bias adjustment voltage VTT is applied to each cell, the reset operation is completed in step 320. Many variations can be made to the method of Figure 8 in accordance with a particular embodiment. For example, the adjustment operations in step 3 18 may be incorporated after steps 312 and 307 are completed for all of the reconfigured units and before verification is performed in step 306. Figure 9A illustrates one embodiment of a portion of a control circuit 220 that can also be used to apply the reverse bias reset conditions of Figure 6. Column decoder 422 corresponds to a selected word line during the reset pulse and outputs a ground to the NMOS/PMOS word line driver circuit (e.g., 224 in Figure 4). Upper PMOS devices 402 and 404 are turned on for the ground input of the driver circuit. The ground input causes the driver circuit to pass the reverse source 123008.doc • 31- 200830315 select bus signal vWR and _ to the selected word line and each semi-selected word line associated with the decoder, respectively. Each column of decoding 423 corresponding to an unselected word line outputs a v-disc to its individual driver circuit, as shown in Figure 9B. The positive bias of VwR turns on the NMOS devices 416 and 418 of the driver circuits of the unselected word lines. Accordingly, the source selection bus is selected as 'quad (both GND) and driven on each corresponding unselected word line. In one embodiment, as previously described, the word line is reversely re-set φ with a fixed voltage VwR equal to about +1/2 VRR. VWR can also provide other voltage levels. For example, one or more reverse reset voltage pulses having a tilt pulse (e.g., start k + 1/2 Vrr and then increase) can be provided for the reset operation as described below. 10A and 10B are circuit diagrams of portions of a row control circuit 210 that can be used to apply bias conditions for the reverse reset operation. Row decoder 5 12 controls a selected bit line driver to provide the selected bit line voltage pulse VBR. In one embodiment, the vBR provides a voltage pulse of _1/2 Vrr. Row decoder 5 12 may be spread across a plurality of bit line drivers (e.g., 24) and will also connect the semi-selected bit lines to a ground bias just prior to application of the reset pulse. During the application of this pulse, • the semi-selected bit lines float near ground. Semi-selected bit lines • A large number of unselected cells provide leakage current that keeps the semi-selected bit lines close to ground. In a specific embodiment, a memory unit that shares a row of decoders with the selected positioning element during a reset operation may be a semi-selected δ hexon precursor. For example, the cells can be connected to the selected word line during the reset operation. The selected row decoder 512 outputs (}1^) 123008.doc -32 - 200830315 to the input of the driver circuit for the row decoder. The GND input at the NMOS/PMOS pair of the driver circuit will turn on the lower NMOS device 506. The reverse source select bus level VBR is passed to the selected bit line. The unselected column decoders 5 13 provide VBr to the gates of their individual driver circuits, thereby selecting the PMOS devices at the top of each driver pair. The source select bus signal levels (both at GND) are provided to each unselected word line corresponding to decoder 51. Figure 11A illustrates one embodiment of a portion of a column control circuit 220 that can be used to apply the reverse bias adjustment conditions of Figure 7. The selected column decoder 422 outputs a word line adjustment voltage pulse Vwt to the NMOS/PMOS word line driver circuit. Vwt is a positive voltage and turns on the lower NMOS devices 406 and 408. The driver circuit passes the source select bus signal VWT to the selected word line. Each unselected column decoder 423 outputs GND to its individual driver circuit as shown in Figure 11B. The upper PMOS devices 412 and 414 are turned on and the GND signal from the reverse source select bus is passed to each of the unselected word lines. Figure 10A is a circuit diagram of one portion of a row control circuit 210 that can be used to apply bias conditions for the reverse bias adjustment operation. The selected row decoder 512 controls a selected bit line driver, turns on the upper PMOS devices and passes GND to each bit line selected for the reset operation. This adjustment operation is independent of the data, and a large number of units can be selected at one time under the self-limiting condition of the operation. Thus, each bit line in a selected block receives the GND level signal to apply the reverse bias to adjust the voltage level. 123008.doc -33- 200830315

與圖9A至10B所示列與行解碼器相關聯之驅動器電路可 以匕括形成用於額外字元線及位元線的驅動器選擇電路之 額外的NMOS/PMOS裳置對。例如,用於該列控制電路之 每一驅動器集合可以包括16,NM〇s/pM〇s對,該等“個 NMOS/PMQS對係連接至該陣列之16個不同的字元線且係 與一單一列解碼器相關聯。用於該行控制電路之每一驅動 器集合可以包括12個NMOS/PMOS對,該等12個 NMOS/PMOS對係連m車狀12個不同的字元線且係 與單的行解碼器相關聯。此組態係範例性的,而可以 ,據具體實施例使用其他組態。但是,如上所述之此一組 恶可以有利地在每一記憶體層級提供該等陣列線之一合理 的扇出。其還促進以與該驅動器電路相關聯的陣列線相同 之間距來放置該驅動H電路。除適應大量陣列線外,此配 置可以避免各種驅動器電壓位準向該陣列之長距傳輸而因 此提高功率效能。關於用以控制一記憶體陣列的驅動器及 控制電路之更多細節(在—具體實施例中包括適用於實施 與貧料相關的對選定與未選定字元及/或位元線之選擇之 E< Scheuerlein^Luca -號(MD-295律The driver circuit associated with the column and row decoders illustrated in Figures 9A through 10B can include additional NMOS/PMOS skirt pairs that form driver select circuits for additional word lines and bit lines. For example, each driver set for the column control circuit can include 16, NM 〇 s / pM 〇 s pairs connected to the 16 different word lines of the array and tied A single column decoder is associated. Each driver set for the row control circuit can include 12 NMOS/PMOS pairs, which are connected to 12 different word lines and are Associated with a single row decoder. This configuration is exemplary, and other configurations may be used in accordance with a particular embodiment. However, such a set of evils as described above may advantageously provide this at each memory level. One of the array lines has a reasonable fan-out. It also facilitates placing the drive H circuit at the same distance from the array lines associated with the driver circuit. This configuration avoids various driver voltage levels except for a large number of array lines. Long-distance transmission of the array and thus improved power efficiency. More details regarding the driver and control circuitry used to control a memory array (in the specific embodiment include the selection and unselection of pairs suitable for implementing lean materials) word And / or the selection of the bit line of the E < Scheuerlein ^ Luca - Number (MD-295 law

Fasoli的美國專利申請案第 師檔案號碼G23-GG51),其名稱為"用於將讀取/寫入電路輕 合至一記憶體陣列之與資料相關的雙重匯流排"。 裝置特徵之差異可以影響在適才說明的反向重新設定操 作期間記憶體陣列202内的個別記憶體單元之特性。記憶 體單元可以具有因該製程而產生之不同尺寸。裝置之間在 123008.doc -34- 200830315 材料成分(例如多晶石夕4 夕日日矽材枓)方面可能缺少特定的均勻性。 ::,與該陣列中之-平均單元之標稱位準相比, ^::可以在一較低電壓編其他單元可以在-較高 —j為依據本揭示内容之—具體實施例充分地重新 μ e fe體早兀陣列(包括個別記憶體單元之間的變 ,)’在一重新設定操作期間向該陣列的選定記憶體單元 、、之至7 t壓脈衝具有—有變化斜率之振幅,從而逐Fasoli's U.S. Patent Application No. 1 file number G23-GG51) is entitled "Data-Related Dual Bus" for Lightning the Read/Write Circuit to a Memory Array. Differences in device characteristics can affect the characteristics of individual memory cells within memory array 202 during a reverse reset operation as described. The memory cells can have different sizes resulting from the process. There may be a lack of specific uniformity between the devices in the composition of the material (eg, polycrystalline stone, day, day, day, and day). ::, compared to the nominal level of the averaging unit in the array, ^:: can be programmed at a lower voltage, other units can be -higher -j based on the present disclosure - the embodiment is adequately Re- fe fe 兀 ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( Thus

漸★曰加向該等選定記憶體單元施加之反向㈣。需要一較 新又疋電壓位準之單元將在該電壓脈衝之振幅已改變 後重新設定於較高的反向偏壓,而僅需要-較低重新設定 電壓位準之單元將重新設定於一較小位準的反向偏壓。此 技術適應裝置之間的變化而同時還提供不會損害重新設定 的裝置之一尚效率程序。由於可以施加一單一的重新設定 電壓脈衝以產生一反向偏壓重新設定條件範圍,因此使得 耗時的驗證操作得到避免或最小化。可以橫跨每一單元而 施加該單一脈衝而該振幅變化以增加該反向偏壓。在該重 新設定電壓脈衝之較低值重新設定之單元在其重新設定為 該較高電阻狀態時將自動關閉。重新設定後的較高電阻將 減小或停止經由此等裝置之電流流動,從而確保其不因較 高值的重新設定電壓而受到損害。 圖13A至13B說明依據一具體實施例可在如圖6所示之一 重新設定操作期間分別施加於選定字元線及位元線之重新 设疋電壓彳§號。圖13 A說明一字元線重新設定電壓信號 Vwr,其在所說明操作之部分之持續時間期間升高至約 123008.doc -35- 200830315 + 1/2 VRR之一最大值(例如,+6 V)。圖13B中說明針對每一 重新設定電壓脈衝具有一-1/2 VRR的起始值之一位元線重 新设定電壓信號VBR。該位元線上的重新設定信號具有一 依據一實質上不變的斜率而改變之振幅。圖13B中,該位 ^ 元線重新設定電壓信號從約-1/2 VRR之一初始值增加至 • 約-(1/2 VRR+2 V)之一終止值。針對每一負位元線脈衝的 振幅幅度增加約2 V(例如,達到-8 V)以使得橫跨該陣列的 φ 選定部分而施加之反向偏壓逐漸增加。藉由如圖1 〇B所示 之電何幫浦電路之輸出將該Vbr脈衝之振幅限制於圖13 B 中虛線所示之又抓偏壓位準。該Vbr偏壓位準藉由控制圖 14B中的计數器712而在施加該等vBR脈衝之間返回其初始 值。在每一記憶體單元之二極體如圖6所示從位元線至字 兀線對齊之條件下,該字元線重新設定電壓之不變值與該 位元線重新設定電壓信號 < 增加的負電壓使得向沿該選定 位兀線及選定字元線的每一記憶體單元施加之反向偏壓增 _ 加。針對该位元線重新設定電壓信號,顯示多個脈衝,其 可用於個別地重新設定該陣列之較小部分。例如,可以向 在數個選定區塊(子陣列)的每一區塊内之一位元線施加一 • 第一重新設定電壓脈衝,而向在數個選定區塊的每一區塊 • 内之一第二位元線施加一第二脈衝。向更多位元線施加更 多重新設定脈衝直至將該使用者提供的所有資料編碼。此 技術可能需要與為儲存一頁面的使用者資料而使用的區塊 數目成反比例關係之16至64個或更多的重新設定電壓脈 衝。 123008.doc -36- 200830315 針對vBR之起始及終止值可以隨實施方案而改變。在一 具體實施例中,使用統計資料或實驗來針對每一脈衝選擇 最佳的起始與終止值。例如,可以將該脈衝之初始值選擇 成產生一反向偏壓,該反向偏壓係決定為任何單元在從該 較低電阻狀態重新設定為該較高電阻狀態之前將會需要的 最小值。每一脈衝之終止值可以係選擇成產生為重新設定 該陣列的任何單元而一般需要的最大反向偏壓。藉由逐漸 施加增加的反向偏壓,重新設定於一較低重新設定反向 偏壓位準之記憶體單元可以避免在增加的反向偏壓位準受 到損害。當一記憶體單元重新設定為該較高電阻重新設定 狀態時,其將傳導較少的電流並以一自我限制的方式表 現。在其已成功地重新設定時,其會自行關閉或在一足夠 的程度上停止傳導。此自我限制截止點將避免在該等反向 偏壓條件下受損。應注意,將一重新設定脈衝之振幅從一 起始值逐漸增加到一較大的終止值以由此增加針對選定記 憶體單元之反向偏壓不會具有與施加具有一較大起始值的 不變脈衝相同之電性效應。具有一較大起始值之一脈衝可 旎損吾形成該電阻改變元件的材料或引起電阻之一永久偏 移。因此,所揭示技術之一具體實施例利用一傾斜反向重 新設定脈衝來成功而安全地抹除具有不同裝置特徵之記憶 體單元。 圖12A及12B分別說明在一具體實施例中可以提供重新 設定電壓信號的列控制電路與行控制電路之一部分。圖 14A中之一電荷幫浦706透過包括一反向源極選擇匯流排脈 123008,doc -37- 200830315 衝產生器之列控制電路將反向重新設定Vwr偏壓位準提供 給該反向源極選擇匯流排(例如,圖9八至12中的匯流排 430)並直接提供給列解碼器電路(例如,圖4中的解碼器 322)。參考電壓產生器702接收一供應電壓Vcc並將一參考 電壓Vref提供給電荷幫浦控制器7〇4。使用來自電荷幫浦 706的輸出之一回授信號,該控制器可以按需要提供約 1/2 VRR之一起始vWR偏壓位準。 圖12B所說明的行控制電路利用一計數器712及數位至類 比轉換器714來產生具有一負傾斜脈衝輸出(負位準及斜率) 之位元線重新設定電壓乂⑽偏壓位準。計數器712接收一脈 衝開始時間並使用一時脈信號,向DAC 714提供一脈衝輸 入來產生一類比傾斜脈衝輸出。DAC 714接收該數位輸入 並向該電荷幫浦控制器提供電壓位準。電荷幫浦718產生 依據一由該計數器產生的實質上不變且係負斜率而增加之 一負位元線重新設定電壓VBR。該負電壓VBR偏壓位準之振 幅依據所定義的斜率而增加以逐漸增加橫跨該記憶體陣列 而施加的反向偏壓。 圖15A及15B說明用以施加圖6之反向偏壓之一替代的電 壓信號集合。一正電壓脈衝VWR係施加於該(等)選定的字 元線並依據一正斜率而增加。在該(等)選定的位元線上施 加一負位元線電壓脈衝VBR。每一字元線電壓脈衝開始於 約+5 V之一起始值而增加2 V至約+7 V。該VWR脈衝之幅度 受限於來自該電荷幫浦電路的輸出之VWR偏壓位率且係顯 示為圖15A中的虛線。該等字元線與位元線重新設定脈衝 123008.doc -38- 200830315 之組合將橫跨每一選定記憶體單元提供一增加的反向偏 壓。額外的位元線重新設定電壓脈衝係說明為可用於設定 或重新設定額外的位元線群組。如圖9A至9B所示,在某 些具體實施例中圖11A至11B之脈衝可用於產生一正向偏 壓。在另一具體實施例中,該等脈衝並非傾斜的。例如, 可以向一第一陣列線施加具有一負極性之一第一電壓脈 衝’而向一第二陣列線施加具有一正極性之一第二電壓脈 衝以產生一反向偏壓。此配置還可以切換該等記憶體單元 之電阻’但不包括在該等脈衝上之一斜率或在所施加的偏 壓中之一所產生的偏移。 圖15A及15B之具體實施例包括一重試技術,該技術針 對在施加該初始電壓脈衝時不重新設定的記憶體單元使用 由該VWR偏壓位準決定之一略微更高的重新設定脈衝位 準。例如,可以在施加最後的重新設定電壓脈衝8〇4及814 後驗證該陣列之一選定部分之重新設定之結果。一驗證操 作可以包括讀回該記憶體單元之電阻狀態並將其與針對該 重新設定狀態的預定義位準相比較。可以讓不重新設定的 任何行或位元線經受一更高位準之一重試脈衝。該字元線 電壓脈衝806之起始值係增加至7 V並增加至一 9 V之位 準。任何重試脈衝之值可隨具體實施例而變化且可以係如 先前所述依據統計資料及/或測試來選擇。在圖15A及15B 中’向未通過針對一重新設定狀態的驗證之陣列之每一位 元線施加該重試脈衝。在其他具體實施例中,可以在個別 施加該等初始重新設定電壓脈衝後施加一重試脈衝(或多 123008.doc •39- 200830315 個脈衝)。若一行或其他群时- 、他群組的早疋在一重試脈衝(或多個 重試脈衝)後未通過針對該目標電阻狀態之驗證,列可以 ,用錯誤校正㈣技術來對其加以處置或以冗餘的記憶體 早元來加以替換。Gradually add the inverse (4) applied to the selected memory cells. A unit requiring a newer voltage level will be reset to a higher reverse bias after the amplitude of the voltage pulse has changed, and only the unit requiring a lower reset voltage level will be reset to one. A smaller level of reverse bias. This technique accommodates changes between devices while also providing an efficiency program that does not compromise resetting. Since a single reset voltage pulse can be applied to generate a reverse bias reset condition range, time consuming verify operations are avoided or minimized. The single pulse can be applied across each cell and the amplitude changes to increase the reverse bias. The cell reset at the lower value of the reset voltage pulse will automatically turn off when it is reset to the higher resistance state. The higher resistance after reset will reduce or stop the flow of current through these devices, thereby ensuring that it is not damaged by the higher value of the reset voltage. Figures 13A through 13B illustrate resetting voltages applied to selected word lines and bit lines, respectively, during a reset operation as shown in Figure 6, in accordance with an embodiment. Figure 13A illustrates a word line reset voltage signal Vwr that rises to a maximum of about 123008.doc -35 - 200830315 + 1/2 VRR during the duration of the portion of the illustrated operation (eg, +6) V). A bit line re-set voltage signal VBR having a start value of one -1/2 VRR for each reset voltage pulse is illustrated in Fig. 13B. The reset signal on the bit line has an amplitude that varies according to a substantially constant slope. In Fig. 13B, the bit line reset voltage signal is increased from an initial value of about -1/2 VRR to an end value of about - (1/2 VRR + 2 V). The amplitude amplitude of each negative bit line pulse is increased by about 2 V (e.g., to -8 V) such that the reverse bias applied across the selected portion of the array of φ is gradually increased. The amplitude of the Vbr pulse is limited to the biased bias level shown by the dashed line in Fig. 13B by the output of the electric Ho circuit as shown in Fig. 1B. The Vbr bias level returns to its initial value between the application of the vBR pulses by controlling the counter 712 of Figure 14B. Under the condition that the diode of each memory cell is aligned from the bit line to the word line as shown in FIG. 6, the word line resets the constant value of the voltage and the bit line resets the voltage signal < The increased negative voltage increases the reverse bias applied to each memory cell along the selected positioning line and the selected word line. A voltage signal is reset for the bit line, and a plurality of pulses are displayed that can be used to individually reset a smaller portion of the array. For example, a first reset voltage pulse can be applied to one of the bit lines in each of a plurality of selected blocks (sub-arrays), and to each block in a plurality of selected blocks. One of the second bit lines applies a second pulse. More reset pulses are applied to more bit lines until all data provided by the user is encoded. This technique may require 16 to 64 or more reset voltage pulses in inverse proportion to the number of blocks used to store user data for one page. 123008.doc -36- 200830315 The starting and ending values for vBR can vary from implementation to implementation. In a specific embodiment, statistics or experiments are used to select the optimal start and end values for each pulse. For example, the initial value of the pulse can be selected to produce a reverse bias that is determined to be the minimum value that any cell would need before resetting from the lower resistance state to the higher resistance state. . The end value of each pulse can be selected to produce the maximum reverse bias typically required to reset any cells of the array. By gradually applying an increased reverse bias, resetting the memory unit at a lower reset reverse bias level can avoid damage to the increased reverse bias level. When a memory cell is reset to the higher resistance reset state, it will conduct less current and behave in a self-limiting manner. When it has been successfully reset, it shuts itself down or stops conduction to a sufficient extent. This self-limiting cutoff point will avoid damage under these reverse bias conditions. It should be noted that the amplitude of a reset pulse is gradually increased from a starting value to a larger terminating value to thereby increase the reverse bias for the selected memory cell without having a larger starting value with the application. The same electrical effect of the constant pulse. A pulse having a larger starting value can damage the material forming the resistance changing element or cause a permanent deflection of one of the resistors. Accordingly, one embodiment of the disclosed technology utilizes a tilt reverse reset pulse to successfully and safely erase memory cells having different device characteristics. Figures 12A and 12B illustrate portions of a column control circuit and a row control circuit that may provide a reset voltage signal, respectively, in a particular embodiment. A charge pump 706 of FIG. 14A includes a reverse source select bus line 123008, and a doc-37-200830315 burst generator control circuit provides a reverse reset Vwr bias level to the reverse source. The poles are selected (e.g., bus bars 430 in Figures 9-8) and provided directly to the column decoder circuitry (e.g., decoder 322 in Figure 4). The reference voltage generator 702 receives a supply voltage Vcc and supplies a reference voltage Vref to the charge pump controller 7〇4. Using one of the outputs from charge pump 706 to feed back the signal, the controller can provide one of the starting vWR bias levels of about 1/2 VRR as needed. The row control circuit illustrated in Figure 12B utilizes a counter 712 and a digital to analog converter 714 to generate a bit line reset voltage 乂 (10) bias level having a negative ramp output (negative level and slope). Counter 712 receives a pulse start time and uses a clock signal to provide a pulse input to DAC 714 to produce an analog skew output. The DAC 714 receives the digital input and provides a voltage level to the charge pump controller. The charge pump 718 generates a negative bit line reset voltage VBR that is increased by a substantially constant and negative slope generated by the counter. The amplitude of the negative voltage VBR bias level is increased in accordance with the defined slope to gradually increase the reverse bias applied across the memory array. Figures 15A and 15B illustrate a set of voltage signals instead of one of the reverse biases of Figure 6. A positive voltage pulse VWR is applied to the selected word line and increases in accordance with a positive slope. A negative bit line voltage pulse VBR is applied to the (equal) selected bit line. Each word line voltage pulse begins at about one of the +5 V starting values and increases by 2 V to about +7 V. The magnitude of the VWR pulse is limited to the VWR bias bit rate from the output of the charge pump circuit and is shown as the dashed line in Figure 15A. The combination of the word line and bit line reset pulses 123008.doc -38 - 200830315 will provide an increased reverse bias across each selected memory cell. Additional bit line reset voltage pulses are described as available for setting or resetting additional bit line groups. As shown in Figures 9A through 9B, the pulses of Figures 11A through 11B can be used to generate a forward bias in some embodiments. In another embodiment, the pulses are not oblique. For example, a first voltage pulse having a negative polarity can be applied to a first array line and a second voltage pulse having a positive polarity can be applied to a second array line to generate a reverse bias. This configuration can also switch the resistance of the memory cells' but does not include the slope of one of the pulses or the offset produced by one of the applied biases. The embodiment of Figures 15A and 15B includes a retry technique for using a reset pulse level that is slightly higher by one of the VWR bias levels for a memory cell that is not reset when the initial voltage pulse is applied. . For example, the result of the resetting of a selected portion of the array can be verified after the last reset voltage pulses 8〇4 and 814 are applied. A verify operation can include reading back the resistance state of the memory cell and comparing it to a predefined level for the reset state. Any row or bit line that is not reset can be subjected to a higher level one retry pulse. The initial value of the word line voltage pulse 806 is increased to 7 V and increased to a level of 9 V. The value of any retry pulse can vary from embodiment to embodiment and can be selected based on statistics and/or testing as previously described. This retry pulse is applied to each bit line of the array that has not passed the verification for a reset state in Figs. 15A and 15B. In other embodiments, a retry pulse (or more 123008.doc • 39 - 200830315 pulses) may be applied after each of the initial reset voltage pulses is applied. If a row or other group - the early group of a group fails to pass the verification of the target resistance state after a retry pulse (or multiple retry pulses), the column can be disposed of by the error correction (4) technique. Or replace it with redundant memory.

圖6A及16B# a月依據_具體實施例可用於提供圖η a及 別之脈衝的列及行控制電路之部分。該等選定的字元線 θ供在此具體只轭例中具有依據一正斜率而增加之一振 te的JL重㈣定信號。在驅動電荷幫浦控制器9⑽時利用 一計數器904及數位至類比轉換器9〇6。控㈣9〇8使用 DAC 906之類比輸出並經由電荷幫浦91〇產生一正傾斜、 偏壓位準。將電荷幫浦910之輸出直接施加於該等字元線 解碼器並㈣反向源極選擇匯流排脈衝產生電路施加於該 反向源極選擇匯流排線。圖⑽說㈣以提供負%偏壓位 準之行控制電路210之一部分。一參考電壓產生器914將一 參考電壓vref輸送至電荷幫浦控制器916。該控制器利用來 自電荷幫浦918的輸出之一回授迴路來保持針對該位元線 重新設定電壓信號的vBR偏壓之一穩定值。 前述關於本發明的詳細說明係基於圖解及說明之目的而 提出。其並不希望飽攬無遺或將本發明限於所揭示的精確 形式。根據以上原理,可進行許多修改及變更。選擇所述 具體實施例係為了最佳地說明本發明之原理及其實際應 用,從而使其他熟習此項技術者能將本發明最佳地應用於 各種具體實施例中並作出適合特定預期用途的各種修改。 希望本發明之範疇由本文隨附申請專利範圍加以定義。 123008.doc -40 - 200830315 【圖式簡單說明】 圖m明依據-具體實施例之—範例性非揮發性記憶體 H2 一 早70。 圖2A及2B說明依據一具體實施例之範例性非揮發性記 憶體單元。 圖3A及3B係依據—具體實施例之—三維記憶體陣列之 個別透視圖及斷面圖。 圖4係依據-具體實施例之—非揮發性記憶體系統之一 方塊圖。‘ 圖5係5兄明依據-具體實施例針對—非揮發性記憶體系 統的各種狀態之電阻分佈之一曲線圖。 圖6係解說依據—具體實調針對—反向偏壓重新設定 操作的偏壓條件之一記憶體陣列之一簡化電路圖。 圖7係解⑤依據—具體實施例針對一反向偏壓調整操作 的偏壓條件之一記憶體陣列之一簡化電路圖。 圖8係依據具體實施例用於重新設定一記憶體單元陣 列之一方法之一流程圖,其併入一調整操作。 圖9A及9B係可以依據一具體實施例用於提供圖6之反向 偏壓重新設定條件之列控制電路之-部分之電路圖。 圖10A及10B係可以依據一具體實施例用於提供圖6之反 向偏壓重新叹定條件之行控制電路之一部分之電路圖。 圖11A及11B係可以依據一具體實施例用於提供圖7之反 向偏壓調整條件之列控制電路之-部分之電路圖。 圖12係可以依據一具體實施例用於提供圖7之反向偏壓 123008.doc 200830315 調整條件之行控制電路之一部分之一電路圖。 圖13 A及13B說明依據一具體實施例用以在重新設定操 作期間產生一增加的反向偏壓之範例性位元線及字元線重 新設定電壓信號。 圖14A及14B係可用於產生如圖ha及11B所示的傾斜脈 衝重新設定電壓信號之控制電路之一部分之電路圖。 圖15A及15B說明依據一具體實施例用以在重新設定操 作期間產生一增加的反向偏壓之其他範例性位元線及字元 線重新設定電壓。 圖16A及16B係可用於產生如圖9A及9B所示的傾斜重新 設定脈衝信號之控制電路之一部分之電路層級圖。 【主要元件符號說明】 100 記憶體單元/被動儲存元件 102 引導元件 104 狀態改變元件 106 反熔絲 110 第一導體/第一金屬導電層 112 第二導體/第二金屬導電層 120 記憶體單元 122 重度摻雜的η型區域 124 本質區域 126 重度摻雜的ρ型區域 128 反熔絲 140 記憶體單元 123008.doc •42- 200830315 142 重度摻雜的η型區域 144 本質區域 146 重度摻雜的ρ型區域 152 記憶體單元 162 第一位元線集合 164 字元線 170 記憶體單元 174 位元線 176 字元線 178 記憶體單元 180 位元線 182 記憶體單元 184 字元線 186 記憶體單元 200 積體電路 202 記憶體陣列 206 行控制電路210之輸入/輸出 208 列控制電路220之輸出 210 行控制電路 212 行解碼器 214 陣列終端接收器或驅動器 216 區塊選擇電路 220 列控制電路 222 列解碼器 123008.doc -43- 200830315 224 陣列終端驅動器 226 區塊選擇電路 250 線/狀態 252 線/狀態 254 線/狀態 256 線/狀態 258 線/目標位準 402 上部PMOS裝置 404 上部PMOS裝置 406 下部NMOS裝置 408 下部NMOS裝置 412 上部PMOS裝置 414 上部PMOS裝置 416 NMOS裝置 418 NMOS裝置 422 列解碼器 423 列解碼器 430 匯流排 506 下部NMOS裝置 512 行解碼器 513 未選定的列解碼器 702 參考電壓產生器 704 電荷幫浦控制器 706 電荷幫浦 123008.doc -44- 200830315Figures 6A and 16B# a month may be used to provide portions of the column and row control circuits of Figure η a and other pulses. The selected word line θ is provided with a JL-heavy (four) fixed signal having one of the vibrations added in accordance with a positive slope in this particular yoke example. A counter 904 and a digital to analog converter 9 〇 6 are utilized when driving the charge pump controller 9 (10). Control (4) 9〇8 uses the analog output of DAC 906 and generates a positive tilt and bias level via charge pump 91〇. The output of the charge pump 910 is applied directly to the word line decoder and (iv) the reverse source select bus pulse generation circuit is applied to the reverse source selection bus line. Figure (10) illustrates (d) a portion of the row control circuit 210 that provides a negative % bias level. A reference voltage generator 914 delivers a reference voltage vref to the charge pump controller 916. The controller utilizes one of the outputs from the charge pump 918 to return the loop to maintain a stable value of the vBR bias for resetting the voltage signal for the bit line. The foregoing detailed description of the invention has been presented for purposes of illustration It is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations are possible in light of the above principles. The embodiments were chosen to best explain the principles of the invention and its application, and thus, Various modifications. It is intended that the scope of the invention be defined by the scope of the appended claims. 123008.doc -40 - 200830315 [Simplified Schematic] FIG. 4 is an exemplary non-volatile memory H2 as early as 70. 2A and 2B illustrate exemplary non-volatile memory cells in accordance with an embodiment. 3A and 3B are individual perspective and cross-sectional views of a three-dimensional memory array in accordance with a particular embodiment. Figure 4 is a block diagram of one of the non-volatile memory systems in accordance with a particular embodiment. ‘ Figure 5 is a graph of the resistance distribution of various states of the non-volatile memory system according to the specific embodiment. Figure 6 is a simplified circuit diagram of one of the memory arrays in accordance with one of the bias conditions for the reverse bias reset operation. Figure 7 is a simplified circuit diagram of one of the memory arrays for one of the bias conditions for a reverse bias adjustment operation in accordance with an embodiment. Figure 8 is a flow diagram of one of the methods for resetting a memory cell array in accordance with a particular embodiment, incorporating an adjustment operation. Figures 9A and 9B are circuit diagrams of portions of a control circuit for providing a reverse bias reset condition of Figure 6 in accordance with an embodiment. 10A and 10B are circuit diagrams of a portion of a line control circuit for providing a reverse bias re-attack condition of FIG. 6 in accordance with an embodiment. 11A and 11B are circuit diagrams of portions of a control circuit for providing a reverse bias adjustment condition of FIG. 7 in accordance with an embodiment. Figure 12 is a circuit diagram of one portion of a row control circuit for providing a reverse bias voltage of 123008.doc 200830315 of Figure 7 in accordance with an embodiment. 13A and 13B illustrate exemplary bit line and word line reset voltage signals for generating an increased reverse bias during a reset operation in accordance with an embodiment. 14A and 14B are circuit diagrams of portions of a control circuit that can be used to generate a ramp pulse reset voltage signal as shown in Figures ha and 11B. 15A and 15B illustrate other exemplary bit line and word line reset voltages used to generate an increased reverse bias during a reset operation in accordance with an embodiment. Figures 16A and 16B are circuit level diagrams of portions of a control circuit that can be used to generate the ramp reset pulse signals as shown in Figures 9A and 9B. [Main Element Symbol Description] 100 Memory Unit/Passive Storage Element 102 Guide Element 104 State Change Element 106 Anti-Fuse 110 First Conductor/First Metal Conductive Layer 112 Second Conductor/Second Metal Conductive Layer 120 Memory Unit 122 Heavily doped n-type region 124 essential region 126 heavily doped p-type region 128 anti-fuse 140 memory cell 123008.doc • 42- 200830315 142 heavily doped n-type region 144 essential region 146 heavily doped ρ Type area 152 Memory unit 162 First bit line set 164 Word line 170 Memory unit 174 Bit line 176 Word line 178 Memory unit 180 Bit line 182 Memory unit 184 Word line 186 Memory unit 200 Integrated circuit 202 memory array 206 row control circuit 210 input/output 208 column control circuit 220 output 210 row control circuit 212 row decoder 214 array terminal receiver or driver 216 block selection circuit 220 column control circuit 222 column decoding Array 123008.doc -43- 200830315 224 Array Terminal Driver 226 Block Selection Circuit 250 Line / State 252 Line / State 254 Line / State 256 Line / State 258 Line / Target Level 402 Upper PMOS Device 404 Upper PMOS Device 406 Lower NMOS Device 408 Lower NMOS Device 412 Upper PMOS Device 414 Upper PMOS Device 416 NMOS Device 418 NMOS Device 422 Column Decoder 423 Column Decoder 430 Bus 506 Lower NMOS Device 512 Row Decoder 513 Unselected Column Decoder 702 Reference Voltage Generator 704 Charge Pump Controller 706 Charge Pump 123008.doc -44- 200830315

712 714 716 718 804 806 814 904 906 908 910 914 916 918 計數器 數位至類比轉換器/DAC 電荷幫浦控制電路 電荷幫浦 重新設定電壓脈衝 字元線電壓脈衝 重新設定電壓脈衝 計數器 數位至類比轉換器/DAC 電荷幫浦控制器 電荷幫浦 參考電壓產生器 電荷幫浦控制器 電荷幫浦712 714 716 718 804 806 814 904 906 908 910 914 916 918 Counter Digit to Analog Converter / DAC Charge Pump Control Circuit Charge Pump Reset Voltage Pulse Word Line Voltage Pulse Reset Voltage Pulse Counter Digit to Analog Converter / DAC charge pump controller charge pump reference voltage generator charge pump controller charge pump

123008.doc -45-123008.doc -45-

Claims (1)

200830315 申請專利範園: 一種操作非揮發性儲存之方法,其包含: 藉由複數個非揮發性儲存元件施加_第_位準 錢,將該等儲存元件從—較低電阻狀態切: 電阻狀態;以及 在切換該等儲存元件後向該等儲存元件施加一第二位 以降低具有超出一對應於該第二電阻狀態 的払位準之一電阻的該等儲存元件之一子集之一 阻。 /、 電 2·如請求項1之方法,其中·· 該第一位準的反向偏壓高於該第 3·如請求項2之方法,其中: 該第二位準的反向偏壓約為該第 該位準之60%。 二位準的反向偏壓 一位準的反向偏壓 〇 之 4·如請求項1之方法,其中:200830315 Patent Application: A method of operating non-volatile storage, comprising: applying _ _ bit quasi-money by a plurality of non-volatile storage elements, cutting the storage elements from a lower resistance state: resistance state And applying a second bit to the storage elements after switching the storage elements to reduce a subset of the subset of the storage elements having a resistance exceeding a level corresponding to the second resistance state . The method of claim 1, wherein the first level of reverse bias is higher than the third method of claim 2, wherein: the second level of reverse bias It is about 60% of the first level. Two-level reverse bias A one-order reverse bias 〇 4. According to the method of claim 1, wherein: ^加-第-位準的反向偏壓包含向與該等儲存元件通 仏之-第-Ρ車列線施加_正電壓脈衝而向與該等儲存元 件通信之一第二陣列線施加一負電壓脈衝;以及 施加-第二位準的反向偏壓包含向該第—陣列線施加 一正電壓脈衝而向該第二陣列線施加一固定偏壓。 5·如請求項1之方法,其中: 該複數個非揮發性儲存元件之每一元件皆包括一電阻 率改變材料; 於該複數個儲存 施加該第一位準的反向偏壓而增加用 123008.doc 200830315 兀=之每—儲存元件的該電阻率改變材料之一電阻率; 加該第二位準的反向偏壓而減小用隹 I 儲存元件七 、該子集之母一 干的該電阻率改變材料之該電阻率。 6.如請求項5之方法,其中: 該電阻率改變材料係多晶矽。 7·如請求項5之方法,其中: 該電阻率改變材料係一金屬氧化物。 8·如請求項5之方法,其中: =複數個非揮發性儲衫件之每—元件包括與該電阻 率改變材料串聯之一 ^ H1 ^ ^ ^ 導兀件該電阻率改變材料形成 該引導几件之至少一部分。 9·如請求項5之方法,其中: 阻發性健存元件之每一元件皆包括與該電 阻羊改k材料串聯之一反熔絲。 10·如請求項1之方法,其令: 馨 之::數個非揮發性儲存元件係-三維單石記憶體陣列 11.如請求項10之方法,其中: 包括第—複數個陣列線與實質上 複數個陣列線之第二複數個陣列線;以1 者數個陣列線與該第二複數個陣列線之至少-二==,層級之間共享的個別線。 該較低電阻妝能i 於針對該等健存元件之一設定狀 123008.doc • 2 - 200830315 態;以及 該較高電阻狀態對應 定狀態。 、士該等儲存元件之一重新設 13·如請求項12之方法,其中: 該較低電阻狀態係針 化狀 態 。 、該荨儲存元件之一程式 14·如請求項13之方法,其中:The plus-first-level reverse bias includes applying a positive-voltage pulse to the -first bus line that is in communication with the storage elements and applying a second array line to one of the communication elements A negative voltage pulse; and applying a second level of reverse bias includes applying a positive voltage pulse to the first array line and applying a fixed bias voltage to the second array line. 5. The method of claim 1, wherein: each of the plurality of non-volatile storage elements comprises a resistivity change material; and the plurality of stores apply the reverse bias of the first level to increase 123008.doc 200830315 兀 = each of the resistivity of the storage element changes the resistivity of the material; adding the second level of reverse bias to reduce the use of 隹I storage element VII, the mother of the subset This resistivity changes the resistivity of the material. 6. The method of claim 5, wherein: the resistivity change material is polycrystalline germanium. 7. The method of claim 5, wherein: the resistivity change material is a metal oxide. 8. The method of claim 5, wherein: = each of the plurality of non-volatile storage members comprises: one of the elements in series with the resistivity changing material ^ H1 ^ ^ ^ guiding member, the resistivity changing material forming the guiding At least a part of a few pieces. 9. The method of claim 5, wherein: each of the components of the stimuli-storing component comprises an anti-fuse in series with the resistive material. 10. The method of claim 1, wherein: ???:: a plurality of non-volatile storage element systems - a three-dimensional single-rock memory array. 11. The method of claim 10, wherein: the first plurality of array lines and The second plurality of array lines of the plurality of array lines; the individual lines shared by at least two of the array lines and the second plurality of array lines. The lower resistance makeup can be set to one of the storage elements 123008.doc • 2 - 200830315; and the higher resistance state corresponds to the state. And one of the storage elements is reset. 13. The method of claim 12, wherein: the lower resistance state is a pinned state. One of the storage elements of the program. 14. The method of claim 13, wherein: 該複數個非揮發性 a如請求扣之方法,其^件包括兩個以上的狀態。 非揮發性儲存元件係多狀態儲存元件。 16·如明求項12之方法,其中: 該複數個非揮發性儲存元件係形成—次場可程式化 記憶體之一陣列之部分; 態 該較低電阻狀㈣針對該等儲存元件之-格式化狀 17·如請求項1之方法,其中: 施加該第二位準的反向偏壓不實質上改變具有在該目 標位準内之一電阻的該複數個儲存元件之一電阻。 18. —種非揮發性記憶體系統,其包含: 複數個非揮發性儲存元件,其包括一電阻改變元件; 控制電路,其與該複數個非揮發性儲存元件通信,該 控制電路藉由施加一第一位準的反向偏壓將該等儲存元 件從一較低電阻狀態切換為一較高電阻狀態來重新設定 該複數個非揮發性儲存元件,在切換該等儲存元件後該 123008.doc 200830315 控制電路向該等儲存元件施加一 々 昂—仪準的反向偏壓, 以降低具有超出一對應於該第二 电且狀恶的目標位準之 -電阻的該等儲存元件之一子集之一電阻。 19·如請求項18之非揮發性記憶體系統,其中: 該第一位準的反向偏塵高於該第二位準的反向偏麼。 20.如請求項19之非揮發性記憶體系統,其中: 該弟一位準的反向偏壓约兔兮楚 . 徇縻匀為5亥弟一位準的反向偏壓之The plurality of non-volatile a, such as the method of requesting deduction, includes two or more states. The non-volatile storage element is a multi-state storage element. The method of claim 12, wherein: the plurality of non-volatile storage elements form part of an array of subfield programmable memory; the lower resistance (4) for the storage elements - The method of claim 1, wherein: applying the reverse bias of the second level does not substantially change a resistance of the plurality of storage elements having a resistance within the target level. 18. A non-volatile memory system comprising: a plurality of non-volatile storage elements including a resistance change element; a control circuit in communication with the plurality of non-volatile storage elements, the control circuit being applied A first level of reverse biasing switches the storage elements from a lower resistance state to a higher resistance state to reset the plurality of non-volatile storage elements, the 123008 after switching the storage elements. Doc 200830315 The control circuit applies a reverse-biased reverse bias to the storage elements to reduce one of the storage elements having a resistance beyond a target level corresponding to the second electrical and disgusting Set one of the resistors. 19. The non-volatile memory system of claim 18, wherein: the reverse level of the first level is higher than the reverse of the second level. 20. The non-volatile memory system of claim 19, wherein: the younger one of the opposite reverse biases is about the same as the rabbit. 該位準之60%。 21.如請求項18之非揮發性記憶體系統,其中 該控制電路藉由向與該等儲存元件通信之-第一陣列 線施加一正電壓脈衝而向與該等儲#元件通信之一第二 陣列線施加一負電壓脈衝來施加一第一位準的反向偏 壓;以及 該控制電路施加-第二位準的反向偏壓包含向該第— 陣列線施加一正電壓脈衝而向該第二陣列線施加一固定 φ 偏壓。 22.如請求項18之非揮發性記憶體系統,其中: 該複數個非揮發性儲存元件之每一元件皆包括一電阻 * 率改變材料; * 施加該第一位準的反向偏壓而增加用於該複數個儲存 元件之每一儲存元件的該電阻率改變材料之一電阻率; 施加該第二位準的反向偏壓而減小用於該子集之每一 儲存元件的該電阻率改變材料之一電阻率。 23·如請求項22之非揮發性記憶體系統,其中·· 123008.doc 200830315 該電阻率改變材料係多晶矽。 24·如請求項22之非揮發性記憶體系統,其中 該電阻率改變材料係一金屬氧化物。 25·如請求項22之非揮發性記憶體系統,其中 元件皆包括與該電 電阻率改變材料形 該複數個非揮發性儲存元件之每一 阻率改變材料串聯之一引導元件,該 成該引導元件之至少一部分。 26.60% of this level. 21. The non-volatile memory system of claim 18, wherein the control circuit communicates with the memory element by applying a positive voltage pulse to the first array line in communication with the storage elements. The second array line applies a negative voltage pulse to apply a first level of reverse bias; and the control circuit applies a second level of reverse bias comprising applying a positive voltage pulse to the first array line The second array line applies a fixed φ bias. 22. The non-volatile memory system of claim 18, wherein: each of the plurality of non-volatile storage elements comprises a resistance* rate changing material; * applying the first level of reverse bias Increasing a resistivity of the resistivity changing material for each of the plurality of storage elements; applying the second level of reverse bias to reduce the each of the storage elements for the subset The resistivity changes the resistivity of one of the materials. 23. The non-volatile memory system of claim 22, wherein: 123008.doc 200830315 The resistivity change material is polycrystalline germanium. 24. The non-volatile memory system of claim 22, wherein the resistivity changing material is a metal oxide. 25. The non-volatile memory system of claim 22, wherein the component comprises a guiding element in series with each resistivity changing material of the plurality of non-volatile storage elements. At least a portion of the guiding element. 26. 如請求項22之非揮發性記憶體系統,其中: 該複數個非揮發性儲存元件之每一元件皆包括與該電 阻率改變材料串聯之一反熔絲。 27·如請求項is之非揮發性記憶體系統,其進一步包含: 二維單石讀、體P車列,其包括該複數個非揮發性儲 存元件。 28·如請求項27之非揮發性記憶體系統,其進一步包含·· 第一複數個陣列線;以及 第二複數個陣列線,其實質上垂直於該第一複數個陣 列線; 其中該第-複數個陣列線與該第二複數個陣列線之至 少-者包括在該三料列的記憶體層級之間共享的個別 線。 29·如請求項18之非揮發性記憶體系統,其中: 該較低電阻狀態對應於針對該等健存元件之一設定狀 態;以及 該較高電阻狀態對應於針對該㈣存元件之—重新設 123008.doc 200830315 定狀態。 3〇·如晴求項29之非揮發性記憶體系統,其中·· 該較低電阻狀態係針對該等儲存元件之一程式化狀 態。 31_如請求項30之非揮發性記憶體系統,其中: 該複數個非揮發性儲存元件係可重寫的儲存元件。 32·如睛求項30之非揮發性記憶體系統,其中·· 該複數個非揮發性儲存元件包括兩個以上的狀態。 33_如請求項29之非揮發性記憶體系統,其中: 該複數個非揮發性儲存元件係形成一一次場可程式化 記憶體之一陣列之部分; 該較低電阻狀態係針對該等儲存元件之一格式化狀 態。 34.如請求項18之非揮發性記憶體系統,其中·· 施加該第二位準的反向偏壓不實質上改變具有在該目 標位準内之一電阻的該複數個儲存元件之一電阻。 3 5 ·如請求項18之非揮發性記憶體系統,其中·· 該控制電路包括列控制電路與行控制電路之至少一 者0 123008.docThe non-volatile memory system of claim 22, wherein: each of the plurality of non-volatile storage elements comprises an anti-fuse in series with the resistivity change material. 27. The non-volatile memory system of claim 1, further comprising: a two-dimensional single stone read, body P train including the plurality of non-volatile storage elements. 28. The non-volatile memory system of claim 27, further comprising: a first plurality of array lines; and a second plurality of array lines substantially perpendicular to the first plurality of array lines; wherein the - at least one of the plurality of array lines and the second plurality of array lines - comprising individual lines shared between the memory levels of the three streams. 29. The non-volatile memory system of claim 18, wherein: the lower resistance state corresponds to setting a state for one of the health storage elements; and the higher resistance state corresponds to a weight for the (four) memory component Newly set 123008.doc 200830315 to determine the status. 3. The non-volatile memory system of claim 29, wherein the lower resistance state is a stylized state for one of the storage elements. 31. The non-volatile memory system of claim 30, wherein: the plurality of non-volatile storage elements are rewritable storage elements. 32. The non-volatile memory system of claim 30, wherein the plurality of non-volatile storage elements comprise more than two states. 33. The non-volatile memory system of claim 29, wherein: the plurality of non-volatile storage elements form part of an array of one field programmable memory; the lower resistance state is for the One of the storage elements is formatted. 34. The non-volatile memory system of claim 18, wherein the applying the second level of reverse bias does not substantially alter one of the plurality of storage elements having a resistance within the target level resistance. 3: The non-volatile memory system of claim 18, wherein the control circuit comprises at least one of a column control circuit and a row control circuit.
TW96127570A 2006-07-31 2007-07-27 Method of operating non-volatile storage and non-v TWI356415B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US11/461,424 US7495947B2 (en) 2006-07-31 2006-07-31 Reverse bias trim operations in non-volatile memory
US11/461,431 US7492630B2 (en) 2006-07-31 2006-07-31 Systems for reverse bias trim operations in non-volatile memory

Publications (2)

Publication Number Publication Date
TW200830315A true TW200830315A (en) 2008-07-16
TWI356415B TWI356415B (en) 2012-01-11

Family

ID=38863123

Family Applications (1)

Application Number Title Priority Date Filing Date
TW96127570A TWI356415B (en) 2006-07-31 2007-07-27 Method of operating non-volatile storage and non-v

Country Status (2)

Country Link
TW (1) TWI356415B (en)
WO (1) WO2008016844A2 (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7522448B2 (en) 2006-07-31 2009-04-21 Sandisk 3D Llc Controlled pulse operations in non-volatile memory
US7719874B2 (en) 2006-07-31 2010-05-18 Sandisk 3D Llc Systems for controlled pulse operations in non-volatile memory
WO2008016833A2 (en) * 2006-07-31 2008-02-07 Sandisk 3D Llc Increasing write voltage pulse operations in non-volatile memory
US7978507B2 (en) 2008-06-27 2011-07-12 Sandisk 3D, Llc Pulse reset for non-volatile storage
TWI480980B (en) * 2012-09-26 2015-04-11 Lin Chrong Jung Memory array and non-volatile memory device of the same

Family Cites Families (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5818749A (en) * 1993-08-20 1998-10-06 Micron Technology, Inc. Integrated circuit memory device
US6483734B1 (en) * 2001-11-26 2002-11-19 Hewlett Packard Company Memory device having memory cells capable of four states
JP4660095B2 (en) * 2002-04-04 2011-03-30 株式会社東芝 Phase change memory device
JP4129170B2 (en) * 2002-12-05 2008-08-06 シャープ株式会社 Semiconductor memory device and memory data correction method for memory cell
US6813177B2 (en) * 2002-12-13 2004-11-02 Ovoynx, Inc. Method and system to store information
US7800932B2 (en) * 2005-09-28 2010-09-21 Sandisk 3D Llc Memory cell comprising switchable semiconductor memory element with trimmable resistance
US20050226067A1 (en) * 2002-12-19 2005-10-13 Matrix Semiconductor, Inc. Nonvolatile memory cell operating by increasing order in polycrystalline semiconductor material
US6930909B2 (en) * 2003-06-25 2005-08-16 Micron Technology, Inc. Memory device and methods of controlling resistance variation and resistance profile drift
KR100682895B1 (en) * 2004-11-06 2007-02-15 삼성전자주식회사 Nonvolitile Memory Device Using Resistance Material Having Mulistate Resistance and Function Method for the Same
US7812404B2 (en) * 2005-05-09 2010-10-12 Sandisk 3D Llc Nonvolatile memory cell comprising a diode and a resistance-switching material

Also Published As

Publication number Publication date
WO2008016844A3 (en) 2008-03-27
WO2008016844A2 (en) 2008-02-07
TWI356415B (en) 2012-01-11

Similar Documents

Publication Publication Date Title
US7492630B2 (en) Systems for reverse bias trim operations in non-volatile memory
US7495947B2 (en) Reverse bias trim operations in non-volatile memory
KR101573506B1 (en) Reverse set with current limit for non-volatile storage
KR101568327B1 (en) Simultaneous write and verify in a non-volatile storage
US7522448B2 (en) Controlled pulse operations in non-volatile memory
US7719874B2 (en) Systems for controlled pulse operations in non-volatile memory
KR101573507B1 (en) Pulse reset for non-volatile storage
US7499304B2 (en) Systems for high bandwidth one time field-programmable memory
US7499355B2 (en) High bandwidth one time field-programmable memory
TWI575525B (en) Non-volatile semiconductor memory device
JP5377633B2 (en) Capacitive discharge method for writing non-volatile memory
US8848430B2 (en) Step soft program for reversible resistivity-switching elements
TWI397924B (en) Method for controlled pulse operations in non-volatile memory and non-volatile memory
JP2012531004A (en) Write reversible resistance switching element
EP1929525A1 (en) Method for using a memory cell comprising switchable semiconductor memory element with trimmable resistance
JP2013520761A (en) Step software program for reversible resistive switching devices
TW200830315A (en) Method of operating non-volatile storage and non-volatile memory system
WO2008016835A1 (en) High bandwidth one time field-programmable memory
WO2008016420A2 (en) Multi-use memory cell and memory array and method for use therewith
TWI455130B (en) Mixed-use memory array and method for use therewith
WO2008016421A2 (en) Mixed-use memory array with different data states and method for use therewith

Legal Events

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