TW201212319A - Composition of memory cell with resistance-switching layers - Google Patents
Composition of memory cell with resistance-switching layers Download PDFInfo
- Publication number
- TW201212319A TW201212319A TW100121291A TW100121291A TW201212319A TW 201212319 A TW201212319 A TW 201212319A TW 100121291 A TW100121291 A TW 100121291A TW 100121291 A TW100121291 A TW 100121291A TW 201212319 A TW201212319 A TW 201212319A
- Authority
- TW
- Taiwan
- Prior art keywords
- layer
- electrode
- conductive
- resistance
- switching
- Prior art date
Links
Landscapes
- Semiconductor Memories (AREA)
Abstract
Description
201212319 六、發明說明: 【發明所屬之技術領域】 本發明技術係關於資料儲存。 此申請案主張2010年6月18曰申請之美國臨時專利申請 案第 61/356,327 號(檔案號:SAND-01478US0)及2011 年3月 25曰申請之美國臨時專利申請案第61/467,936號(檔案號: SAND-01478US1)之權益,該兩個臨時專利申請案皆以引 用方式併入本文中。 【先前技術】 各種材料展示出可逆電阻改變或電阻切換行為’其中材 料之電阻依據穿過該材料之電流及/或跨越該材料之電壓 之歷史而變。此等材料包含硫屬化物、碳聚合物、鈣鈦礦 以及某些金屬氧化物(MeOx)及金屬氮化物(MeN)。具體而 言,存在僅包含一種金屬且展現出可靠電阻切換行為之金 屬氧化物及氮化物。舉例而言,此群組包含氧化鎳 (NiO)、氧化鈮(Nb205)、二氧化鈦(Ti02)、氧化铪 (Hf02) '氧化鋁(A1203)、氧化鎂(MgOx)、二氧化鉻 (Cr02)、氧化釩(VO)、氮化硼(BN)及氮化鋁(A1N),如 Pagnia及 Sotnick在「Bistable Switching in Electroformed Metal-Insulator-Metal Device」(Phys. Stat. Sol. (A) 108,11 至65 (1988))中所闡述。可將此等材料中之一者之一電阻 切換層(RSL)形成為處於一初始狀態,舉例而言,一相對 低電阻狀態。在施加充分電壓時,該材料切換至一穩定高 電阻狀態,甚至在移除該電壓之後仍維持該穩定高電阻狀 156917.doc 201212319 態。此電阻切換係可逆的,以使得隨後施加一適當電流或 電壓可用來使該RSL返回至一穩定低電阻狀態,甚至在移 除该電壓或電流之後仍維持該穩定低電阻狀態。此轉化可 重複多次。對於某些材料,該初始狀態係高電阻而非低電 阻。一設定程序可係指將該材料自高電阻切換至低電阻, 而一重設程序可係指將該材料自低電阻切換至高電阻。一 電阻切換記憶體元件(RSME)可包含定位在第一電極與第 二電極之間的一RSL。 將此等可逆電阻改變材料用於非揮發性記憶體陣列中頗 受關注。舉例而言,一個電阻狀態可對應於一資料「〇」, 而另一電阻狀態對應於一資料「1」。某些此等材料可具有 兩個以上穩定電阻狀態。此外,在一記憶體單元中, RSME可與例如二極體之一操縱元件串聯,該操縱元件選 擇性地限制跨越該RSME之電壓及/或穿過該RSME之電 流。舉例而言,二極體可允許電流僅沿該RSME2 一個方 向流動’而基本上阻礙相反方向上之一電流。此—操縱元 件自身通常並非一電阻改變材料。而是,該操縱元件允許 在不影響一陣列中其他記憶體單元之狀態之情形下寫入至 一記憶體單元及/或自該記憶體單元讀取。 已知具有由電阻改變材料形成之儲存元件或單元之非揮 發性記憶體。舉例而言,以引用方式併入本文之標題為 「Rewriteable Memory Cell Comprising A Di〇de And a Resistance-Switching Material」之美國專利申請公開案第 2006/0250836號闡述一種包含與一電阻改變材料(例如” 156917.doc 201212319201212319 VI. Description of the invention: [Technical field to which the invention pertains] The technology of the present invention relates to data storage. This application claims US Provisional Patent Application No. 61/356,327, filed on Jun. 18, 2010 ( file number: SAND-01478US0) and US Provisional Patent Application No. 61/467,936, filed on March 25, 2011 ( Archives: SAND-01478US1), the entire contents of each of which are incorporated herein by reference. [Prior Art] Various materials exhibit reversible resistance change or resistance switching behavior. The resistance of the material varies depending on the history of the current passing through the material and/or the voltage across the material. These materials include chalcogenides, carbon polymers, perovskites, and certain metal oxides (MeOx) and metal nitrides (MeN). Specifically, there are metal oxides and nitrides which contain only one metal and exhibit reliable resistance switching behavior. For example, the group includes nickel oxide (NiO), niobium oxide (Nb205), titanium dioxide (Ti02), hafnium oxide (Hf02) 'alumina (A1203), magnesium oxide (MgOx), chromium dioxide (Cr02), Vanadium oxide (VO), boron nitride (BN) and aluminum nitride (A1N), such as Pagnia and Sotnick in "Bistable Switching in Electroformed Metal-Insulator-Metal Device" (Phys. Stat. Sol. (A) 108, 11 As explained in 65 (1988)). A resistance switching layer (RSL), one of the materials, may be formed to be in an initial state, for example, a relatively low resistance state. When a sufficient voltage is applied, the material switches to a stable high resistance state and maintains the stable high resistance even after the voltage is removed. 156917.doc 201212319 state. This resistance switching is reversible so that subsequent application of an appropriate current or voltage can be used to return the RSL to a stable low resistance state, even after the voltage or current is removed. This conversion can be repeated multiple times. For some materials, this initial state is high resistance rather than low resistance. A setting procedure may refer to switching the material from a high resistance to a low resistance, and a reset procedure may refer to switching the material from a low resistance to a high resistance. A resistance switching memory element (RSME) can include an RSL positioned between the first electrode and the second electrode. The use of such reversible resistance-changing materials in non-volatile memory arrays has received considerable attention. For example, one resistance state may correspond to one data "〇" and another resistance state corresponds to a data "1". Some of these materials may have more than two stable resistance states. Moreover, in a memory cell, the RSME can be in series with a steering element, such as one of the diodes, that selectively limits the voltage across the RSME and/or the current through the RSME. For example, the diode can allow current to flow only in one direction along the RSME2 and substantially block one of the currents in the opposite direction. This—the steering element itself is usually not a resistance change material. Rather, the manipulation element allows writing to and/or reading from a memory unit without affecting the state of other memory units in an array. Non-volatile memory having a storage element or unit formed of a resistance change material is known. For example, U.S. Patent Application Publication No. 2006/0250836, the disclosure of which is incorporated herein by reference in its entirety in its entirety in the entire entire entire entire entire entire entire entire entire disclosure 156917.doc 201212319
MeOx或MeN)串聯耦合之二極體之可再寫非揮發性記憶體 單元。 然而,繼續需要允許記憶體單元之大小按比例縮小之技 術。 【實施方式】 本發明提供一種δ己憶體系統,其包含具有兩個或兩個以 上電阻切換層(RSL)之可逆電阻率切換記憶體元件 (RSME)。在一實例性實施方案中,該RSME包含串聯之一 第一電極(E1)、一第一電阻切換層(RSL1)、被視為係一散 射層或耦合電極之一中間層(IL)、一第二rsl(RSL2)及一 第二電極(E2)。在一種方法中,該RSME具有一鏡像組 態’其中該RSME組態在該IL之任一側上對稱。然而,並 不要求此一鏡像組態。 通常,隨著基於RSME之記憶體裝置之大小按比例縮 小,一缺點係在該RSME之一設定或重設程序期間之一衝 擊電流可損壞與該RSME串聯之一相關聯操縱元件或甚至 阻礙極度減小之尺寸下之記憶體單元之操作。亦通常,極 多的基於RSL之記憶體裝置需要在彼期間使該RSL之初始 絕緣性質崩潰之一形成步驟。此形成步驟通常與一極短且 極高放電電流峰值相關聯,該放電電流峰值可設定該rsl 之導通電阻位準以用於隨後切換事件。若該導通電阻位準 係極低(例如’ 1 〇〇 kQ至30 kQ),則相關聯切換電流亦係 極高且因此該記憶體單元將不可在極小技術節點下操作。 一設定或重設程序係針對一 RSL及RSME之一種類型之電 I56917.doc -6- 201212319 阻切換操作。為解決此問題,提供在一導電比之任—側上 包含單獨RSL之一 RSME。 特定而言,包含本文中所提供之一 RSME之一記憶體單 元可藉由主動減小操作電流來限制衝擊電流過沖.。例如 環之-薄ILT防止電流過沖且可限制電流流動,藉此較 容易跨越個別RSL產生-大電場。由於電流減小,因此損 壞該單元之一操縱元件之可能性減小且可使用一較薄操 縱元件,從而促進記憶體裝置之按比例縮小且可能減小電 力消耗。由於仍然允許一離子電流,因此該單元之切換能 力得以維持。 該RSME係基於一個別rSl之一定性模型,該定性模型 闡述數個發現,包含:基於電子/電洞及離子傳導之一切 換電流、離子電流之指數型E場相依性及所量測電流係一 衝擊電流而不用於該切換機制。具體而言,該定性模型闡 述.(1) 一突崩型设定電流增加;(ϋ)為何難以將設定狀態 限於一高導通電阻狀態;(iii)循環良率對設定程序之敏感 度;(iv)為何重設電壓可高於設定電壓;(v)為何需要較高 重設電壓用於較深重設;及(vi)為何針對較深重設重設電 流較高。衝擊電流之模型亦可適用於任何其他「薄」儲存 材料/離子記憶體,例如TiSi、CBRAM(導電橋接RAM) » 對於MeOx之一 RSL,該等發現亦指示電子/電洞電流不引 起切換效應’而係在MeOx中衝擊地行進,從而僅將熱遞 送至觸點,且此不同於較厚碳或相變材料,在該等材料中 若記憶體單元足夠長則此電流在該單元中產生熱。 156917.doc 201212319 圖1係包含與一操縱元件104串聯耦合於一第一導體ι〇6 與一第二導體108之間的一 RS ME 10 2之一電阻切換記憶體 單元(RSMC)IOO之一項實施例之一簡化透視圖。 RSME 1〇2包含在一導電中間層(iL)133之任一側上之 RSL 130及135。如所提及,一 RSL具有可在兩個或兩個以 上狀態之間可逆地切換之一電阻率。舉例而言,一 RSL在 製造時可係處於一初始高電阻率(高電阻)狀態,在施加一 第一電壓及/或電流時該RSL可切換至一低電阻率狀態。施 加一第二電壓及/或電流可使該RSL返回至高電阻率狀態。 另一選擇為’該RSL在製造時可係處於一初始低電阻狀 態’在施加(若干)適當電壓及/或電流時該RSL可逆地切換 至一高電阻狀態。當用於一記憶體單元中時,每一 RSL之MeOx or MeN) a rewritable non-volatile memory cell of a diode coupled in series. However, there continues to be a need for techniques that allow the size of the memory cells to be scaled down. [Embodiment] The present invention provides a delta-resonance system comprising a reversible resistivity switching memory element (RSME) having two or more resistance switching layers (RSL). In an exemplary embodiment, the RSME includes one of the first electrodes (E1) in series, a first resistance switching layer (RSL1), an intermediate layer (IL) that is considered to be a scattering layer or a coupling electrode, and a The second rsl (RSL2) and a second electrode (E2). In one method, the RSME has a mirrored configuration 'where the RSME configuration is symmetric on either side of the IL. However, this mirror configuration is not required. Generally, as the size of the RSME-based memory device is scaled down, one drawback is that one of the inrush currents during one of the RSME setting or resetting procedures can damage the operating element associated with one of the RSME series or even hinder extremes. The operation of the memory unit at a reduced size. Also, in general, many RSL-based memory devices require one of the steps to cause the initial insulation properties of the RSL to collapse during that step. This formation step is typically associated with a very short and very high discharge current peak that sets the on-resistance level of the rsl for subsequent switching events. If the on-resistance level is extremely low (e.g., '1 〇〇 kQ to 30 kQ), then the associated switching current is also extremely high and therefore the memory cell will not operate at the very small technology node. A setting or resetting procedure is for a type of RSL and RSME type I56917.doc -6- 201212319 blocking switching operation. To solve this problem, one of the RSELs comprising a separate RSL is provided on either side of the conductivity ratio. In particular, one of the memory cells, including one of the RSMEs provided herein, can limit the inrush current overshoot by actively reducing the operating current. For example, the ring-thin ILT prevents current overshoot and limits current flow, thereby making it easier to generate a large electric field across individual RSLs. As the current is reduced, the likelihood of damaging one of the unit's steering elements is reduced and a thinner steering element can be used, thereby facilitating scaling of the memory device and possibly reducing power consumption. Since an ion current is still allowed, the switching capability of the unit is maintained. The RSME is based on a certain rSl model, which describes several findings, including: an index-based E-field dependence and a measured current system based on one of electron/hole and ion conduction switching current, ion current. An inrush current is not used for the switching mechanism. Specifically, the qualitative model illustrates. (1) a sudden collapse type set current increase; (ϋ) why it is difficult to limit the set state to a high on-resistance state; (iii) the sensitivity of the cycle yield to the setting procedure; Iv) why the reset voltage can be higher than the set voltage; (v) why a higher reset voltage is required for a deeper reset; and (vi) why the reset current is higher for a deeper reset. The inrush current model can also be applied to any other "thin" storage material/ion memory, such as TiSi, CBRAM (conductive bridged RAM). » For one of the MeOx RSLs, these findings also indicate that the electron/hole current does not cause a switching effect. 'And the impact travels in MeOx to deliver heat only to the contacts, and this is different from thicker carbon or phase change materials in which the current is generated in the cell if the memory cell is sufficiently long heat. 156917.doc 201212319 FIG. 1 is a resistor-switching memory unit (RSMC) 100 comprising an RS ME 10 2 coupled between a first conductor ι 6 and a second conductor 108 in series with an operating element 104. One of the embodiments simplifies the perspective. RSME 1〇2 includes RSLs 130 and 135 on either side of a conductive intermediate layer (iL) 133. As mentioned, an RSL has a resistivity that can be reversibly switched between two or more states. For example, an RSL can be fabricated in an initial high resistivity (high resistance) state that can be switched to a low resistivity state when a first voltage and/or current is applied. Applying a second voltage and/or current returns the RSL to a high resistivity state. Alternatively, the RSL can be in an initial low resistance state at the time of manufacture. The RSL is reversibly switched to a high resistance state when a suitable voltage and/or current is applied. When used in a memory unit, each RSL
一個電阻狀態(及RSME之一對應電,阻狀態)可表示該RSME 之二進制「0」’而每一RSL之另一電阻狀態(及'該RSME之 一對應電阻狀態)可表示該RSME之二進制「丨」。然而,可 使用兩個以上資料/電阻狀態。舉例而言,在上文所提及 之美國專利申請公開案第2006/0250836號中闡述眾多可逆 電阻改變材料及採用可逆電阻改變材料之記憶體單元之操 作。 在一項實施例中,將RSME自高電阻率狀態(例如,表示 二進制資料「0」)切換至低電阻率狀態(例如,表示二進制 資料「1」)之程序稱為設定或形成,且將該RSME自低電 阻率狀態切換至高電阻率狀態之程序稱為重設。在其他實 施例中,可反轉設定與重設及/或資料編碼。可針對一記 156917.doc 201212319 憶體單元執行該設定或重設程序以將其程式化至用以表示 二進制資料之一所期望狀態。 在某些實施例中,RSL 130及135可由金屬氧化物(MeOx) 形成,該金屬氧化物之一個實例係Hf02。 關於使用可逆電阻改變材料製造一記憶體單元之更多資 訊可在於2009年1月1日公佈之標題為「Memory Cell That Employs a Selectively Deposited Reversible Resistance Switching Element and Methods of Forming The Same」之 US 2009/0001343中找到,且該案以引用之方式併入本文 中〇 RSME 102包含電極132及134。電極132定位於RSL 130 與一導體108(例如一位元線或字線(控制線))之間。在一項 實施例中,電極132係由鈦(Ti)或氮化鈦(TiN)製成。電極 :134定位於RSL 133與一操縱元件104之間。在一項實施例 中,電極134係由氮化鈦(TiN)製成且充當一黏合與障壁 層。 操縱元件104可係二極體或藉由選擇性地限制跨越RSME 102之電壓及/或穿過RSME 102之電流而展現非歐姆傳導之 其他適合操縱元件。在一種方法中,該操縱元件允許電流 沿僅一個方向流過RSME,例如自位元線至字線。在另一 方法中,一操縱元件(例如,一穿通二極體)允許電流沿任 一方向流過該RSME。 該操縱元件充當一單向閥,從而使電流沿一個方向比另 一方向更容易傳導。低於沿正向方向之一臨界「導通」電 156917.doc 201212319 壓’該二極體傳導極小電流或不傳導電流。藉由使用適當 加偏壓方案,當選擇一個別rSME用於程式化時,只要當 沿正向方向施加時跨越相鄰RSME之電壓未超過該二極體 冬f通電壓或當沿相反方向施加時未超過反向崩潰電壓, 相鄰RSME之二極體即可用以電隔離該等相鄰RSME且因此 防止無意中之電阻切換。 具體而言’在一大型交叉點式RSME陣列中,當需要相 對大電壓或電流時,存在與欲定址之rSME共用頂部或底 部導體(例如’字線或位元線)之RSME將曝露於充分電壓 或電流從而導致不期望電阻切換的一危險。取決於所使用 之加偏壓方案,跨越未選定單元之過多洩漏電流亦可成為 一問題。使用二極體或其他操縱元件可克服此危險。 以此方式,記憶體單元1〇〇可用作二維或三維記憶體陣 列之部分且可在不影響該陣列中其他記憶體單元之狀態之 障形下將資料寫入至記憶體單元100及/或自記憶體單元 100讀取資料。操縱元件104可包含任一適合二極體,例如 垂直多晶p-n或p-i-n二極體(或是該二極體之一 η區位於 —Ρ區上面之上指或是該二極體之一 位於一 η區上面之 下指)。或者,甚至可使用一穿通二極體或一齊納二極 體,該等二極體可沿兩個方向操作。該操縱元件與該 RSME—起可呈一垂直柱之形狀。在其他方法中,rsmE2 若干部分在彼此側向配置,如下文進一步論述。 在某些實施例中,操·縱元件104可由一多晶半導體材料 (例如多晶矽、一多晶矽-鍺合金、多晶鍺或任一其他適合 1569l7.doc • 10· 201212319 材料)形成。舉例而言,操縱元件104可包含一重摻雜n+多 晶矽區142、位於n+多晶矽區142上面之一輕摻雜或一本質 (非故意摻雜)多晶矽區144及位於本質區144上面之一重摻 雜P+多晶矽區146。在某些實施例中,可在n+多晶矽區142 上形成一薄(例如,數百埃或更少)鍺及/或矽-鍺合金層(未 展不)(當使用矽-鍺合金層時具有約1〇%或更多之鍺),以防 止及/或減少摻雜劑自n+多晶矽區142遷移至本質區144 中(舉例而5)如於標題為「Deposited Semiconductor Structure To Minimize N-Type Dopant Diffusion And Method Of Making」之美國專利申請公開案第 2006/0087005號中所闡述,該申請公開案以引用之方式併 入本文中。應理解,可反轉^及P+區之位置。 當操縱元件104係由沈積矽(例如,非晶或多晶)製造而 成時’可在一極體上形成妙化物層以使該沈積石夕在製造時 處於一低電阻率狀態。此一低電阻率狀態允許更容易程式 化記憶體單元’乃因將該沈積石夕切換至一低電阻率狀態不 需要一大電壓。 如以引用之方式併入本文中之美國專利第7,176,064號 「Memory Cell Comprising a Semiconductor Junction Diode Crystallized Adjacent to a Silicide」中所闡述,石夕化物形 成材料(例如鈦及/或钻)在退火期間與沈積石夕發生反應以形 成矽化物層。石夕化鈥及石夕化鉛之晶格間距接近於矽之晶格 間距’且此等矽化物層看似可在沈積矽結晶時充當毗鄰沈 積碎之「結晶模板」或「晶種」(例如,石夕化物層在退火 156917.doc 201212319 期間增強矽二極體之結晶結構)。藉此提供較低電阻率之 矽。對於矽-鍺合金及/或鍺二極體而言,可達成類似結 果。 導體106及108包含任一適合導電材料,例如鎢、任一適 當金屬、重摻雜半導體材料、一導電矽化物、—導電妙化 物-鍺化物、一導電鍺化物或諸如此類。在圖丨之實施例 中,導體106及108係軌道形狀且沿不同方向(例如,實質 上彼此垂直)延伸。可使用其他導體形狀及/或組態。在某 些實施例中,可將障壁層、黏合層、抗反射塗層及/或諸 如此類(未展示)與導體106及1〇8—起使用以改良裝置效能 及/或幫助裝置製造。導體1〇6可係一字線而導體1〇8係一 位元線,或反之亦然。 儘管在圖1中將RSME 102展示為定位於操縱元件1〇4上 面’但將理解在替代實施例中,rSME 102亦可定位於操 縱7L件104下面。各種其他組態亦係可能的。—RSL可展 現出單極或雙極電阻切換特性β關於一單極電阻切換特 性,用於設定及重設程序兩者之電壓係相同極性,亦即, 皆為正的或皆為負的。相比之下,關於一雙極電阻切換特 性,將相反極性之電壓用於設定及重設程序。具體而言, 用於設定程序之電壓可係正的,而用於重設程序之電壓係 負的,或用於設定程序之電壓可係負的,而用於重設程序 之電壓係正的。 圖2Α係由複數個圖1之記憶體單元1〇〇形成之一第一記憶 體層級114之一部分之一簡化透視圖。為簡單起見,未單 156917.doc 201212319 獨展示RSME 102、操縱元件l〇4及障壁層113。記憶體陣 列114係一「交又點」式陣列,其包含多個記憶體^元耦 合至其之複數個位元線(第二導體108)及字線(第—導體 106)(如圖所示)。可使用其他記憶體陣列組態,如可使用 多個記憶體層級。 圖2Β係一單體式三維陣列116之一部分之一簡化透視 圖,該單體式三維陣列包含定位於一第二記憶體層級12〇 下面之一第一記憶體層級118。在圖3之實施例中,每一記 憶體層級118及120包含呈一交叉點式陣列之複數個記憶體 單元100。將理解,第一記憶體層級118與第二記憶體層級 120之間可存在額外層(例如,一層間電介質),但為簡單起 見在圖2Β中未展示。可使用其他記憶體陣列組態,如可使 用額外記憶體層級。在圖2Β之實施例中,所有二極體可 「指向」相同方向(例如相依於採用是在該二極體底部還 是頂部上具有一ρ摻雜區之p_i_n二極體而向上或向下),從 而簡化二極體製造。 在某些實施例中,可如美國專利第6,952,03〇號「High-Density Three-Dimensional Memory Cell 」 中所闡述來形成 該等記憶體層級,該專利以引用之方式併入本文中。舉例 而言,一第一記憶體層級之上部導體可用作定位於該第一 記憶體層級上面之一第二記憶體層級之下部導體,如圖2C 中所展示。在此等實施例中,毗鄰記憶體層級上之二極體 較佳地指向相反方向’如標題為「Large Array Of Upward Pointing P-I-N Diodes Having Large And Uniform 156917.doc 13· 201212319 CUrrentj之美國專利7,586,773中所闡述,該專利以引用之 方式併入本文中。舉例而言,第一記憶體層級ιΐ8之二極 體可係如由箭頭A1所指示之上指二極體(例如,其中p區位 於二極體底部)’而第二記憶體層級12〇之二極體可係如由 箭頭A2所指示之下指二極體(例如,其中n區位於二極體底 部),或反之亦然。 一單體式三維記憶體陣列係其中在一單個基板(例如, 一圓晶)上面形成多個記憶體層級而無介入基板之一個記 憶體陣列。形成一個記憶體層級之若干層直接沈積或生長 於一或若干現有層級之層上方。相比之下,如在Leedy之 美國專利第 5,915,167 號「Three Dimensional StructureA resistance state (and one of the RSMEs corresponding to the electrical, resistive state) may represent the binary "0" of the RSME and another resistance state of each RSL (and 'one of the RSMEs corresponding to the resistance state" may represent the binary of the RSME "丨". However, more than two data/resistance states can be used. For example, the operation of a plurality of reversible resistance changing materials and memory cells employing reversible resistance changing materials is set forth in the above-referenced U.S. Patent Application Publication No. 2006/0250836. In one embodiment, the process of switching the RSME from a high resistivity state (eg, representing binary data "0") to a low resistivity state (eg, representing binary data "1") is referred to as setting or forming, and will The procedure for switching the RSME from a low resistivity state to a high resistivity state is called resetting. In other embodiments, the settings and resets and/or data encoding can be reversed. This setting or reset procedure can be performed for a 156917.doc 201212319 memory unit to be programmed to represent the desired state of one of the binary data. In certain embodiments, RSLs 130 and 135 may be formed of a metal oxide (MeOx), an example of which is Hf02. More information on the use of reversible resistance-changing materials to create a memory cell can be found in US 2009/, entitled "Memory Cell That Employs a Selectively Deposited Reversible Resistance Switching Element and Methods of Forming The Same", published on January 1, 2009. Found in 0001343, and the disclosure of which is hereby incorporated by reference in its entirety herein in its entirety, the <RTIgt; Electrode 132 is positioned between RSL 130 and a conductor 108 (e.g., a bit line or word line (control line)). In one embodiment, electrode 132 is made of titanium (Ti) or titanium nitride (TiN). The electrode : 134 is positioned between the RSL 133 and an operating element 104. In one embodiment, electrode 134 is made of titanium nitride (TiN) and acts as a bond and barrier layer. The steering element 104 can be a diode or other suitable steering element that exhibits non-ohmic conduction by selectively limiting the voltage across the RSME 102 and/or the current through the RSME 102. In one method, the steering element allows current to flow through the RSME in only one direction, such as from a bit line to a word line. In another method, a steering element (e.g., a feedthrough diode) allows current to flow through the RSME in either direction. The steering element acts as a one-way valve to allow current to conduct more easily in one direction than the other. A critical "on" current below one of the forward directions. 156917.doc 201212319 Pressure 'The diode conducts very little or no current. By using an appropriate biasing scheme, when a different rSME is selected for stylization, the voltage across the adjacent RSME does not exceed the diode's winter f-pass voltage when applied in the forward direction or when applied in the opposite direction. When the reverse breakdown voltage is not exceeded, the adjacent RSME diodes can be used to electrically isolate the adjacent RSMEs and thus prevent inadvertent resistance switching. Specifically, in a large cross-point RSME array, when a relatively large voltage or current is required, there is an RSME that shares the top or bottom conductor (eg, 'word line or bit line') with the rSME to be addressed. The voltage or current thus causes a risk of undesired resistance switching. Excessive leakage current across unselected cells can also be a problem depending on the biasing scheme used. This danger can be overcome by using diodes or other operating elements. In this way, the memory unit 1 can be used as part of a two-dimensional or three-dimensional memory array and can write data to the memory unit 100 under a barrier that does not affect the state of other memory cells in the array. / or read data from the memory unit 100. The steering element 104 can comprise any suitable diode, such as a vertical polycrystalline pn or a pin diode (or one of the diodes is located above the Ρ region or one of the diodes is located) The upper part of the η area refers to the lower part. Alternatively, even a through diode or a Zener diode can be used, which can operate in both directions. The steering element can be in the shape of a vertical column together with the RSME. In other methods, portions of rsmE2 are laterally configured with each other, as discussed further below. In some embodiments, the motor element 104 can be formed from a polycrystalline semiconductor material (e.g., polysilicon, a polysilicon-germanium alloy, polysilicon or any other suitable material). For example, the steering element 104 can include a heavily doped n+ polysilicon region 142, a lightly doped or an intrinsic (unintentionally doped) polysilicon region 144 over the n+ polysilicon region 142, and a heavily doped one over the intrinsic region 144. P+ polysilicon region 146. In some embodiments, a thin (eg, hundreds of angstroms or less) tantalum and/or tantalum-niobium alloy layer (not shown) may be formed on the n+ polysilicon region 142 (when a tantalum-niobium alloy layer is used) Having about 1% or more of ruthenium) to prevent and/or reduce migration of dopants from the n+ polysilicon region 142 to the intrinsic region 144 (for example, 5) as described under the heading "Deposited Semiconductor Structure To Minimize N-Type" The disclosure of U.S. Patent Application Publication No. 2006/0087005, the entire disclosure of which is incorporated herein by reference. It should be understood that the position of the ^ and P+ regions can be reversed. When the manipulating element 104 is fabricated from deposited germanium (e.g., amorphous or polycrystalline), a layer of wonderful layers can be formed on a pole body such that the deposited stone is in a low resistivity state at the time of manufacture. This low resistivity state allows for easier programming of the memory cell' because a large voltage is not required to switch the deposition to a low resistivity state. As described in "Memory Cell Comprising a Semiconductor Junction Diode Crystallized Adjacent to a Silicide", U.S. Patent No. 7,176,064, the disclosure of which is incorporated herein by reference. During the reaction with sedimentation, a telluride layer is formed. The lattice spacing of Shi Xihua and Shi Xihua's lead is close to the lattice spacing of 矽 and these bismuth layers appear to act as "crystal templates" or "seeds" adjacent to the sedimentary crystallization during deposition crystallization ( For example, the lithium layer enhances the crystalline structure of the ruthenium dioxide during annealing 156917.doc 201212319). This provides a lower resistivity. Similar results can be achieved for bismuth-tellurium alloys and/or ruthenium diodes. Conductors 106 and 108 comprise any suitable electrically conductive material, such as tungsten, any suitable metal, heavily doped semiconductor material, a conductive germanide, a conductive dopant-telluride, a conductive germanide or the like. In the illustrated embodiment, conductors 106 and 108 are rail shaped and extend in different directions (e.g., substantially perpendicular to each other). Other conductor shapes and/or configurations can be used. In some embodiments, barrier layers, adhesive layers, anti-reflective coatings, and/or the like (not shown) can be used with conductors 106 and 1 to improve device performance and/or aid device fabrication. Conductor 1〇6 can be a word line and conductor 1〇8 can be a bit line, or vice versa. Although the RSME 102 is shown in Figure 1 as being positioned above the steering element 1〇4, it will be understood that in alternative embodiments, the rSME 102 can also be positioned below the operating 7L member 104. Various other configurations are also possible. —RSL exhibits a single pole or bipolar resistance switching characteristic β with respect to a single pole resistance switching characteristic, and the voltages used to set and reset the program are of the same polarity, that is, both positive or negative. In contrast, with regard to a bipolar resistance switching characteristic, voltages of opposite polarity are used for setting and resetting procedures. Specifically, the voltage used to set the program can be positive, and the voltage used to reset the program is negative, or the voltage used to set the program can be negative, and the voltage used to reset the program is positive. . Figure 2 is a simplified perspective view of one of the portions of the first memory level 114 formed by a plurality of memory cells 1 of Figure 1. For the sake of simplicity, the RSME 102, the operating element 104 and the barrier layer 113 are shown separately from 156917.doc 201212319. The memory array 114 is an "interleaved" array comprising a plurality of bit lines (second conductors 108) and word lines (first conductors 106) to which a plurality of memory cells are coupled (as shown in the figure). Show). Other memory array configurations can be used, such as multiple memory levels. 2 is a simplified perspective view of one of a portion of a monolithic three-dimensional array 116 comprising a first memory level 118 positioned below a second memory level 12〇. In the embodiment of FIG. 3, each of the memory levels 126 and 120 includes a plurality of memory cells 100 in a cross-point array. It will be appreciated that additional layers (e.g., an inter-layer dielectric) may be present between the first memory level 118 and the second memory level 120, but are not shown in Figure 2 for simplicity. Other memory array configurations can be used, such as the use of additional memory levels. In the embodiment of FIG. 2, all of the diodes can be "pointed" in the same direction (eg, depending on whether a p_i_n diode having a p-doped region at the bottom or top of the diode is used, up or down) , thus simplifying the manufacture of the diode. In some embodiments, the memory levels can be formed as set forth in U.S. Patent No. 6,952,03, entitled "High-Density Three-Dimensional Memory Cell", which is incorporated herein by reference. For example, a first memory level upper conductor can be used to position one of the second memory level lower conductors above the first memory level, as shown in Figure 2C. In such embodiments, the dipoles adjacent to the memory level are preferably pointed in the opposite direction as in U.S. Patent 7,586,773, entitled "Large Array Of Upward Pointing PIN Diodes Having Large And Uniform 156917.doc 13 · 201212319 CUrrentj. As stated, the patent is incorporated herein by reference. For example, the diode of the first memory level ι 8 may be as indicated by the arrow A1 (eg, where the p region is located at The second body of the second memory layer 12' can be referred to as the lower diode as indicated by the arrow A2 (eg, where the n region is at the bottom of the diode), or vice versa. A monolithic three-dimensional memory array is one in which a plurality of memory levels are formed on a single substrate (eg, a wafer) without an intervening substrate. Several layers forming a memory level are directly deposited or grown in one Or three layers of existing tiers. In contrast, U.S. Patent No. 5,915,167 to Leedy, "Three Dimensional Structure"
Memory」中’已藉由在單獨基板上形成若干記憶體層級 並將該等記憶體層級彼此上下地黏合來構造堆疊式記憶 體’該專利以引用之方式併入本文中。可在接合之前將該 等基板薄化或自該等記憶體層級移除,但由於該等記憶體 層級最初形成於單獨基板上方,因此此等記憶體並非真正 的單體式三維記憶體陣列。 根據所揭示之配置’以上實例展示呈一圓柱形或柱形形 狀之記憶體單元及呈軌道形狀之導體。然而,本文中所闡 述之技術並不限於一記憶體單元之任何一個具體結構。亦 可使用其他結構來形成包含RSME之記憶體單元。舉例而 言’美國專利 6,952,043 、6,951,780、6,034,882、 6,420,215、6,525,953及7,081,377(其每一者皆以引用之方 式併入本文中)提供可經調適以使用RSME之記憶體單元之 156917.doc 14 201212319 結構之實例。另外,其他類型之記憶體單元亦可與本文中 所閣述之技術一起使用。 圖3係繪示可實施本文中所闡述之技術之一記憶體系統 3 00之一項實例之一方塊圖。記憶體系統3〇〇包含可係如上 文所闡述之二維或三維記憶體單元陣列之一記憶體陣列 . 302。在一項實施例中,記憶體陣列302係一單體式三維記 憶體陣列。記憶體陣列302之陣列端子線包含組織為若干 列之各種字線層及組織為若干行之各種位元線層^然而, 亦可實施其他定向。 3己憶體系統300包含列控制電路320,其輸出308連接至 δ己憶體陣列302之各別字線。列控制電路320自系統控制邏 輯電路330接收含Μ個列位址信號之一群組及一或多個各 種控制信號’且通常可包含例如列解碼器322、陣列端子 驅動器324及區塊選擇電路326之電路以用於讀取及程式化 (例如’設定及重設)操作兩者。記憶體系統3〇〇亦包含行控 制電路310,其輸入/輸出306連接至記憶體陣列3〇2之各別 位元線。行控制電路306自系統控制邏輯330接收含Ν個行 位址k號之一群組及一或多個各種控制信號,且通常可包 例如行解碼器3 12、陣列端子接收器或驅動器3 14、區塊 選擇電路316之電路以及包含感測放大器318及1/(;)多工器 〜凟取/寫入電路。系統控制邏輯33〇自一主機接收資料及 命7並向主機提供輸出資料。在其他實施例中,系統控制 邏輯330自一單獨控制器電路接收資料及命令,並向彼控 制器電路提供輸出資料,其中該控制器電路與該主機通 1569l7.doc •15- 201212319 “。系統控制邏輯330可包含一或多個狀態機、暫存器及 用於控制記憶體系統300之操作之其他控制邏輯。舉例而 言’可提供下文進一步論述之寫入電路46〇、讀取電路461 及箝位控制電路464。 在一項實施例中,繪示於圖3中之所有組件係配置於一 單個積體電路上》舉例而言,系統控制邏輯33〇、行控制 電路310及列控制電路320可形成於一基板之表面上,且呈 一單體式二維記憶體陣列之記憶體陣列3〇2係形成於該基 板上面(且因此,在系統控制邏輯33〇、行控制電路31〇及 列控制電路320上面)。在某些情況下,控制電路之一部分 可形成於與該記憶體陣列之某些層相同之層上。 併入有一記憶體陣列之積體電路通常將陣列細分成數個 子陣列或區塊。可進一步將區塊編組在一起成為含有(舉 例而言)16個、32個或不同數目個區塊之格室,如通常使 用,一子陣列係一相連記憶體單元群組,其具有通常未被 解碼器、驅動器、感測放大器及輸入/輸出電路阻斷之相 連字線及位元線。出於種種原因而如此做。舉例而言,因 字線及位元線之電阻及電容而引起的向下遍曆此等線之信 號延遲(亦即,RC延遲)在一大陣列中可係非常顯著。可藉 由將一較大陣列細分成一較小子陣列群組以便減小每一字 線及/或每一位元線之長度來減小此等RC延遲。作為另一 實例,與存取一記憶體單元群組相關聯之電力可控管在一 既定記憶體循環期間可同時存取之記憶體單元數目之一上 限。因此,通常將一大記憶體陣列細分成較小子陣列以降 156917.doc -16 - 201212319 低同時存取之記憶體單元之數目。然而,為易於闡述,亦 可使用與子陣列同義之一陣列來指代具有通常未被解碼 器、驅動器、感測放大器及輸入/輸出電路阻斷之相連字 線及位元線之一相連記憶體單元群組。一積體電路可包含 一個或一個以上記憶體陣列。 如上文所述,可藉由可逆地切換RSME 102之RSL中之每 一者來在兩個或兩個以上狀態之間可逆地切換RSME 102。舉例而言,該RSME在製造時可係處於一初始高電阻 率狀態,在施加一第一電壓及/或電流時該RSME可切換至 一低電阻率狀態。施加一第二電壓及/或電流可使該RSME 返回至一高電阻率狀態。記憶體系統300可與本文中所述 之任·一 RSME一起使用。 圖4A係一單極RSL之一實例性實施例之電壓對電流之一 圖表。X軸繪示一電壓絕對值,y軸繪示電流且該等線經調 整以在該圖表之原點處匯合。在設定程序中,線404表示 當處於高電阻率重設狀態時RSL之I-V特性,且線406表示 在Vset下至低電阻率設定狀態之一轉變。在重設程序中, 線400表示當處於低電阻率設定狀態中時RSL之I-V特性且 線402表示在Vreset下至高電阻率重設狀態之一轉變。該實 例展示一單極操作模式,其中對於設定及重設切換兩者而 言電壓之極性係相同的。 為判定RSL之狀態,可跨越該RSL施加一電壓且量測所 得電流。一較高或較低所量測電流分別指示該RSL處於低 或高電阻率狀態。在某些情況下,高電阻率狀態實質上高 156917.doc -17- 201212319 於低電阻率狀態,例如係低電阻率狀態之兩個或三個數量 級(100至1,000)倍高。注意,具有不同I-V特性之一RSL之 其他變化形式亦可與本文中之技術一起使用。 當處於重設狀態時,該RSME回應於介於0與Vset之間的 一所施加電壓而展現出由線404所示之電阻特性。然而, 當處於設定狀態時,該RSME回應於介於0與Vreset之間的 一所施加電壓而展現出由線400所示之電阻特性,其中 Vreset<Vset。因此,取決於該RSME之電阻狀態,該RSME 因此回應於相同電壓範圍(例如,介於0與Vreset之間)中之 相同電壓而展現出不同電阻特性。在一讀取操作中,可回 應於所感測電流是設定狀態中之la還是重設狀態中之lb而 施加一固定電壓Vread<Vreset。可因此藉由識別一 RSL或 RSME之I-V特性之至少一個點來感測該RSL或RSME之狀 態。 在一種方法中,該RSME可包含每一者皆展現出一實質 上類似單極切換特性之多個RSL。"Memory" has been constructed by forming a plurality of memory levels on a separate substrate and bonding the memory levels to each other up and down. The patent is incorporated herein by reference. The substrates may be thinned or removed from the memory levels prior to bonding, but since the memory levels are initially formed over a separate substrate, such memory is not a true monolithic three dimensional memory array. According to the disclosed configuration, the above examples show a memory cell in the shape of a cylinder or a cylinder and a conductor in the shape of a track. However, the techniques described herein are not limited to any one particular structure of a memory unit. Other structures can also be used to form the memory unit containing the RSME. For example, 'US Patent Nos. 6,952,043, 6,951,780, 6,034,882, 6, 420, 215, 6, 525, 953, and 7, 081, 377, each of each of each of each of each of each of each of each of each of An example of the 201212319 structure. In addition, other types of memory cells can be used with the techniques described herein. 3 is a block diagram showing one example of a memory system 300 that can implement the techniques set forth herein. The memory system 3 includes an array of memory that can be one of two or three dimensional memory cell arrays as set forth above. In one embodiment, memory array 302 is a monolithic three dimensional memory array. The array terminal lines of the memory array 302 include various word line layers organized into columns and various bit line layers organized into rows. However, other orientations may be implemented. The memory system 300 includes a column control circuit 320 having an output 308 coupled to respective word lines of the delta memory array 302. Column control circuit 320 receives a group of one of a plurality of column address signals and one or more various control signals from system control logic circuit 330 and may typically include, for example, column decoder 322, array terminal driver 324, and block selection circuit. Circuitry 326 is used for both reading and stylizing (eg, 'set and reset') operations. The memory system 3A also includes a row control circuit 310 having input/output 306 coupled to respective bit lines of the memory array 3〇2. The row control circuit 306 receives from the system control logic 330 a group of one of the row address k numbers and one or more various control signals, and typically may include, for example, a row decoder 3 12, an array terminal receiver or a driver 3 14 The circuit of the block selection circuit 316 includes a sense amplifier 318 and a 1/(;) multiplexer ~ capture/write circuit. System control logic 33 receives data and data from a host and provides output data to the host. In other embodiments, system control logic 330 receives data and commands from a separate controller circuit and provides output data to the controller circuit, wherein the controller circuit communicates with the host 1569l7.doc •15-201212319. Control logic 330 may include one or more state machines, registers, and other control logic for controlling the operation of memory system 300. For example, write circuit 46A, read circuit 461, which is discussed further below, may be provided. And clamp control circuit 464. In one embodiment, all of the components illustrated in Figure 3 are configured on a single integrated circuit. For example, system control logic 33, row control circuit 310, and column control The circuit 320 can be formed on a surface of a substrate, and a memory array 3〇2 of a single-dimensional two-dimensional memory array is formed on the substrate (and thus, in the system control logic 33, the row control circuit 31) The column control circuit 320 is above. In some cases, a portion of the control circuit can be formed on the same layer as some of the layers of the memory array. The integrated circuit typically subdivides the array into a number of sub-arrays or blocks. The blocks can be further grouped together into a cell containing, for example, 16, 32 or a different number of blocks, as is commonly used, a sub- An array is a group of connected memory cells having connected word lines and bit lines that are typically not blocked by decoders, drivers, sense amplifiers, and input/output circuits. This is done for a variety of reasons. The signal delay (ie, RC delay) that traverses the lines down due to the resistance and capacitance of the word lines and bit lines can be significant in a large array. By using a larger array Subdividing into a smaller sub-array group to reduce the length of each word line and/or each bit line to reduce such RC delays. As another example, associated with accessing a memory cell group The upper limit of the number of memory cells that a power controllable can simultaneously access during a given memory cycle. Therefore, a large memory array is usually subdivided into smaller sub-arrays to drop 156917.doc -16 - 201212319 low simultaneous access Remember The number of body units. However, for ease of explanation, an array that is synonymous with a sub-array can also be used to refer to connected word lines and bits that are typically not blocked by decoders, drivers, sense amplifiers, and input/output circuits. One of the lines is connected to a group of memory cells. An integrated circuit can include one or more memory arrays. As described above, two or two can be switched by reversibly switching each of the RSLs of the RSME 102. The RSME 102 is reversibly switched between more than one state. For example, the RSME can be in an initial high resistivity state during manufacture, and the RSME can be switched to a low resistivity when a first voltage and/or current is applied. State. Applying a second voltage and/or current returns the RSME to a high resistivity state. Memory system 300 can be used with any of the RSMEs described herein. Figure 4A is a graph of voltage versus current for an exemplary embodiment of a monopole RSL. The X-axis shows an absolute value of the voltage, the y-axis shows the current and the lines are adjusted to meet at the origin of the graph. In the setup procedure, line 404 represents the I-V characteristic of RSL when in the high resistivity reset state, and line 406 represents one transition from the Vset down to the low resistivity set state. In the reset procedure, line 400 represents the I-V characteristic of RSL when in the low resistivity setting state and line 402 represents one transition from the Vreset to the high resistivity reset state. This example shows a unipolar mode of operation in which the polarity of the voltage is the same for both the set and reset switches. To determine the state of the RSL, a voltage can be applied across the RSL and the resulting current can be measured. A higher or lower measured current indicates that the RSL is in a low or high resistivity state, respectively. In some cases, the high resistivity state is substantially high in the low resistivity state, such as two or three orders of magnitude (100 to 1,000) times higher than the low resistivity state. Note that other variations of RSL with one of the different I-V characteristics can also be used with the techniques herein. When in the reset state, the RSME exhibits a resistive characteristic as indicated by line 404 in response to an applied voltage between 0 and Vset. However, when in the set state, the RSME exhibits a resistive characteristic as indicated by line 400 in response to an applied voltage between 0 and Vreset, where Vreset < Vset. Thus, depending on the resistance state of the RSME, the RSME thus exhibits different resistance characteristics in response to the same voltage in the same voltage range (e.g., between 0 and Vreset). In a read operation, a fixed voltage Vread < Vreset is applied in response to whether the sensed current is lb in the set state or the reset state. The state of the RSL or RSME can thus be sensed by identifying at least one point of the I-V characteristic of an RSL or RSME. In one approach, the RSME can include a plurality of RSLs each exhibiting substantially similar unipolar switching characteristics.
圖4B係繪示兩個實例性單極RSL之不同I-V特性之一圖 表。對於兩個或兩個以上單極RSL,I-V(電流-電壓)特性 可係實質上相同,以使得(舉例而言)1隨著V以一共同速率 增加,且設定及/或重設位準可係實質上相同。或者,RSL 之I-V特性可不同,以使得(例如)針對該等RSL中之一者I 隨著V更快地增加或設定及/或重設位準可係不同。在此實 例中,「A」表示一第一類型之RSL且「B」表示一第二類 型之RSL,其中該等RSL具有不同的單極電阻切換特性。X 156917.doc •18- 201212319 轴繪示電壓(V)且y轴繪示電流(I)。對於類型「A」RSL ’ 線400、402、404及406與圖4A中相同。亦對於類型「A」 RSL,VsetA係設定電壓,VresetA係重設電壓’ IresetA係 重設電流且Iset_limitA係電流設定限制。對於類型「B」 RSL,線 420、422、424 及 426 分別對應於線 400、402、 404及406。亦對於類型「B」RSL,VsetB係設定電壓, VresetB係重設電壓,IresetB係重設電流且Iset_limitB係電 流設定限制。在此處所示之方法中’ VsetA>VsetB、 VresetA>VresetB、IresetA>IresetB 且 Iset_limitA>Iset_limitB, 但此僅係一實例且可應用其他替代性關係。 當兩個或兩個以上RSL處於同一 RSME中時,該RSME之 切換特性將係該等RSL中之每一者之切換特性之一函數。 在一設定程序期間,舉例而言,若跨越每一 RSL相等地分 割電壓,則隨著V增加,類型「B」RSL可在類型「A」 RSL之前切換。類似地,在一重設程序期間,舉例而言, 假設在每一 RSL中施加相同電壓,則隨著V增加,類型 「B」RSL可在類型「A」RSL之前切換。 另一選擇為,類型「A」及「B」RSL可具有相反極性之 不同I-V特性。舉例而言’可使VsetA>0 V且VresetA>0 V, 而VsetB <0 V且VresetB<0 V。作為一實例,類型「A」 RSL之特性可係如圖4A中所述,而類型「B」RSL之特性 可係如下文圖4C中所述。理論上,一 RSME中之一個RSL 亦可具有一單極特性而該RSME中之另一RSL具有一雙極 特性。然而,在一 RSME中之所有RSL當中僅使用一種切 156917.doc -19- 201212319 換特性(單極或雙極)可允許一簡化之控制方案。 在某些情況下,一 RSME之一讀出切換該等RSL中之一 者之資料狀態。舉例而言,在一第一 RSL處於低電阻狀態 且一第二RSL處於高電阻狀態之情形下,假設高電阻狀態 比低電阻狀態高若干數量級,則一讀取操作基本上將偵測 不到電流。亦即,等於每一 RSL之電阻之總和之該RSME 之電阻將極高,因此電流將係極低或基本上為零。一讀取 操作可將第二RSL切換至低電阻狀態,以使得該RSME之 電阻係低,且穿過其之電流係相對高且可偵測"在將第二 RSL切換回至高電阻狀態之後,接下來可執行一回寫操 作。 當跨越一 RSME之電極施加一電壓時,將根據每一 RSL 之電阻跨越每一 RSL按比例劃分該電壓。當該第一 RSL處 於低電阻狀態且第二RSL處於高電阻狀態時,該第一 RSL 會將電極處之電位轉移至IL,以使得實質上跨越第二RSL 而施加所有電壓。若此電壓係適當量值及極性,則其將切 換第二RSL。 此外,一 RSL可使用可作為一單極或雙極裝置操作之一 材料,例如在Sun等人之「Coexistence of the bipolar and unipolar resistive switching behaviours in Au/SrTi03/Pt cells」(2011 年 3 月 10 曰,J· Phys. D: Appl. Phys. 44, 125404)中所述,該文章以引用之方式併入本文中。 圖4C係繪示另一實例性單極RSL之I-V特性之一圖表。 與圖4A之特性相比,在設定及重設程序期間使用負電壓代 156917.doc -20- 201212319 替正電壓。在設定程序中,線434表示當處於高電阻率重 設狀態時RSL之I-V特性,且線436表示在Vset下至低電阻 率設定狀態之一轉變。在重設程序中,線430表示當處於 低電阻率設定狀態時RSL之I-V特性,且線432表示在 Vreset下至高電阻率重設狀態之一轉變。Vread、Vreset、 Vset及Vf皆為負電壓。在一讀取操作中,可回應於所量測 電流係設定狀態中之la還是重設狀態中之lb而施加一固定 電壓 Vread>Vreset。 圖4D係繪示一實例性雙極RSL之I-V特性之一圖表。此 處,將相反極性電壓用於設定及重設程序。此外,將正電 壓用於設定程序且將負電壓用於重設程序。在此雙極RSL 中,當施加一正電壓時發生設定程序且當施加一負電壓時 發生重設程序。在設定程序中,線444表示當處於高電阻 率重設狀態時RSL之I-V特性,且線446表示在Vset下至低 電阻率設定狀態之一轉變。在重設程序中,線440表示當 處於低電阻率設定狀態時RSL之I-V特性,且線442表示在 Vreset下至高電阻率重設狀態之一轉變。Vset及Vf係正電 壓且Vreset係一負電壓。 圖4E係繪示另一實例性雙極RSL之I-V特性之一圖表。 在此雙極RSL中,當施加一正電壓時發生重設程序且當施 加一負電壓時發生設定程序。在設定程序中,線454表示 當處於高電阻率重設狀態時RSL之I-V特性,且線456表示 在Vset下至低電阻率設定狀態之一轉變。在重設程序中, 線450表示當處於低電阻率設定狀態時RSL之Ι-V特性,且 156917.doc •21 - 201212319 線452表示在Vreset下至高電阻率重設狀態之一轉變^ Vset 及Vf係正電壓且Vreset係一負電麼。 儘管圖4D及圖4C中之lreset位準高於iset位準,但應強 調其可反過來。此意指對於相反極性圖4D及圖4C中之Iset 位準可高於lreset位準。 圖5繪示用於讀取一記憶體單元之狀態之一電路之一實 施例。一記憶體陣列之一部分包含記憶體單元55〇、552、 5 54及556。圖中繪示諸多位元線中之兩者及諸多字線中之 兩者。位元線559耦合至單元550及554,且位元線557耦合 至單元552及556。舉例而言,位元線559係選定位元線且 可係處於2 V。舉例而言,位元線557係一未選定位元線且 可係處於接地。舉例而言,字線547係選定字線且可處於〇 V。舉例而言,字線549係—未選定字線且可處於2v〇 將用於該等位元線中之一者559之一讀取電路繪示成經 由電晶體558連接至該位元線,該電晶體受由行解碣器η〗 供應之一閘極電壓控制以選擇或取消選擇對應位元線。電 晶體558將該位元線連接至一資料匯流排563。寫入電路 560(其係系統控制邏輯330之部分)連接至該資料匯流排。 電晶體562連接至該資料匯流排且作為受箝位控制電路 564(其係系統控制邏輯33〇之部分)控制之一籍位裝置操 作。電晶體562亦連接至包含一資料鎖存器568之—感測放 大器566。感測放大器566之輸出連接至一資料輪出端子 (至系統控制邏輯330、一控制器及/或一主機)。寫入電路 560亦連接至感測放大器566及資料鎖存器568。 156917.doc •22· 201212319 當嘗試讀取RSME之狀態時,首先以Vread(例如,大約2 v) 給所有字線加偏壓且使所有位元線處於接地。然後將選定 字線拉至接地。舉例而言,此論述將假設選擇記憶體單元 550進行讀取i透過資料匯流排(藉由導通電晶體558)及箱 位裝置(電晶體562’其接收〜2 V+Vth(電晶體562之臨限電 壓))將一或多個選定位元線559拉至Vread。籍位裝置之閘 極係高於Vread但經控制以保持該位元線接近Vread。在一 種方法中,由選定記憶體單元550將電流自感測放大器中 之一感測節點拉動穿過電晶體562 ^該感測節點可接收介 於一高電阻率狀態電流與一低電阻率狀態電流之間的一參 考電流。該感測節點對應於該單元電流與該參考電流之間 的電流差而移動。感測放大器566藉由比較該所感測電壓 與一參考讀取電壓來產生一資料輸出信號。若該記憶體單 元電流大於該參考電流,則該記憶體單元係處於低電阻率 狀態且該感測節點處之電壓將低於該參考電壓。若該記悔 體單元電流小於該參考電流,則該記憶體單元處於高電阻 率狀態且該感測節點處之電壓將高於該參考電壓。將來自 感測放大器566之輸出資料信號鎖存於資料鎖存器568中。 再次參考圖4A,舉例而言,當處於高電阻率狀態時,若 施加電壓Vset及充分電流,則RSL將被設定為低電阻率狀 態。線404展示當施加Vset時之行為。電壓將保持相當恆 定且電流將朝向Iset 一 limit增加。在某一點處,將設定rsl 且該裝置行為將係基於線406。注意,第一次設定rsl 時,需謂形成電壓)來設定該裝置。之後,二以設 156917.doc •23· 201212319 定所使用之裝置。形成電壓Vf在絕對量值上可大於Vset » 當處於低電阻率狀態(線400)時,若施加vreset及充分電 流(Ireset) ’則rSl將被重設為高電阻率狀態。線4〇〇展示 當施加Vreset時之行為。在某一點處,將重設RSL且該裝 置行為將係基於線402。 在一項實施例中’ Vset係大約7 V,Vreset係大約9 V, Iset_limit係大約1〇 μΑ且Ireset可低達1〇〇 μΑ。此等電壓及 電流適用於其中使一 RSME與二極體_聯之圖5之電路。 圖6Α至圖6Μ可係(舉例而言)沿一 RSMEi 一垂直或水平 平面之一剖視圖。 圖6A繪示具有一 RSME及位於該RSME下面之一操縱元 件(SE)之一實例性記憶體單元。該記憶體單元可具有各種 組態。一個組態係一堆疊式組態,其中以一層形式提供每 一類型之材料,且將每一層定位於其上面之層之下方且通 常具有類似剖面面積。在另一可能組態中,可將一或多 個層與一或多個其他層端對端地配置(參見圖6F至圖6J)。 注意’在該等圖中,描纷為彼此此鄰之任何兩個層或材 料可彼此接觸。然而,除非另有說明,否則並不需要此, 且描繪為彼此毗鄰之任何兩個層或材料可由未描繪之一或 多個其他材料層分離。另外,在某些情況下,一材料可作 為一製造副產物而形成,例如形成於一 Si層上之一 Si〇x 層在圖中沒有必要繪示出此等副產物。此外,所闡述之 實施方案之若干變化形式係可能的。舉例而言,可反轉每 一實施方案中各層之次序以使得(例如)字線位於頂部上且 156917.doc -24- 201212319 位元線位於底部上。可在所繪示之該等層中之每一者之間 提供一或多個中間層。此外,可改變操縱元件之位置,以 使得其位於包含RSL之其他層之上面或下面。可將該等層 之定向自垂直修改為水平或任一其他定向。將可形成一共 同導電路徑之多個層或部分認為係串聯連接。 記憶體單元包含例如W或NiSi之一位元線觸點(BLC)材 料’該位元線觸點材料連接至一記憶體裝置之一位元線。 該位元線係一種類型之控制線,以使得該BLC亦係至一第 •-控制線之一觸點。在一串聯路徑中之BLC之後係例如 TiN之一第一黏合層(AL1),該黏合層有助於blc黏合至 RSME以及充當一障壁。可藉由任一習用方法(例如濺鍍) 沈積一 TiN層。在該串聯路徑中之rSN1e之後係一操縱元件 (SE) ’例如二極體。該操縱元件允許經由字線或位元線將 一信號(例如一電壓或電流)選擇性地施加至一或多個記憶 體單元以個別地控制該等單元以(例如)藉由切換其RSMe 來改變其各別資料狀態❶RSME之電阻切換行為不相依於 SE。SE自身可具有一電阻切換行為;然而,此行為將不 相依於RSME之電阻切換行為。 在該串聯路徑中之SE之後係一第二黏合層(AL2),例如 TiN。在該串聯路徑中之AL2之後係例如W或NiSi之一字線 觸點(WLC)材料,該字線觸點材料連接至一記憶體裝置之 一字線。字線係一種類型之控制線,以使得該WLC亦係至 一第二控制線之一觸點。記憶體單元之所繪示部分係如此 串聯地配置。 1569I7.doc • 25- 201212319 圖6B繪示具有一 rsME之一記憶體單元之一替代性組 態’其中操縱元件(SE)位於該RSME上面。亦可反轉其他 層自頂部至底部之次序,即自底部至頂部。 圖6C繪示作為呈一垂直堆疊之一鏡像電阻式開關(MRS) 之圖6A之RSME之一實例性實施方案。該RSME包含一第 一電極(E1)(其在某些組態中係一頂部電極)、一第一電阻 切換層(RSL1)及一導電中間層(江)(其充當一散射層、耦合 電極或麵合層)。該RSME亦包含一第二rsl(RSL2)及一第 二電極(EL2) ’該第二電極在某些組態中係一底部電極。 舉例而言’該等RSL可係可逆RSL。一可逆RSL可自一個 狀態切換至另一狀態且切換回至該一個狀態。IL係以電方 式位於E1與E2之間且與E1及E2電串聯。rsli係以電方式 位於E1與IL之間且與£1及比電串聯。RSL2係以電方式位 於E2與IL之間且與E2及IL電串聯。「以電方式位於 之 間」或類似詞可意指在一導電路徑中。舉例而言,IL可以 電方式位於E1與E2之間而實體上位於或不位於£1與£2之 間。 舉例而s ’可藉由將兩個雙極憶阻器(記憶體_電阻器)元 件反串聯地連接成一個鏡像電阻式開關(MRS)而形成一 RSME。一憶阻器係一被動兩端子電路元件,其中電阻依 據穿過該裝置之電流及跨越該裝置之電壓之歷史而變。此 一MRS可由一第一憶阻器元件製成,該第一憶阻器元件包 含例如η型矽之一 El、RSL1(其可係一過渡金屬氧化物,例 如氧化铪(Hf02)或氮氧矽铪(HfSiON))及一 IL(其可係能夠 156917.doc •26· 201212319 與氧進行一化學反應之一可氧化電極(例如TiN))。 RSME包含由相同(或不同)材料製成但以一相反次序之 一第二憶阻器元件,該第二憶阻器元件共用IL之可氧化電 • 極。此外,在一種方法中,第一憶阻器元件及第二憶阻器 70件可皆具有雙極或單極I-V(電流-電壓)特性。在另一方 _ 法中,該等憶阻器元件中之一者具有一單極特性且另—憶 阻15 70件具有一雙極特性。藉由將兩個憶阻器元件合併成 .一個RSME,該RSME具有係組成憶阻器元件之Ι-ν特性之 一疊加之一 I-V特性但具有該RSME*在比個別憶阻器元件 低得多之電流下操作之額外益處。 更一般而言’ RSME將具有係組成RSL之特性之—疊 加之一 I-V特性但能夠實現低電流下之操作。 IL藉由散射自RSL進入乩之電子藉此減慢無益於切換機 制之一電子流動以避免損壞一操縱元件而充當一散射層。 此外’ IL充當電容性耦合至藉由設定E1&E2之電位而施加 至該RSME之一電壓之一耦合電極或層。 透過此散射’ IL提供減小一設定或重設程序期間之一峰 值電流同時達成一低電流操作之一電阻。據信該電流限制 操作源自該IL層之兩個態樣。首先,熱電子藉由一電子_ 電子互動在IL層中被極佳地散射。第二,該等rsl中之一 者一開始崩潰且將過量電荷q遞送至比上,於該RSL上所 施加之電壓就有效地減小V=q/C,其中c係IL層朝向電極 E’,1及E2之電谷。同時’ 一較高電壓現在正處於第二尺队 處’從而誘發第二RSL之一崩潰。由於可用電荷q之量係 156917.doc -27· 201212319 有限的’因此此處可流動之電流亦係極有限的。以此方 式’此RSME實現記憶體單元在低電流下之操作。據信該 電阻係基於IL散射電子且產生對所施加偏壓電壓之極有效 的負回饋以使得形成允許在低電流下發生切換之小導電細 絲之能力。在無IL之情形下,當施加一電壓時將形成具有 極低電阻之一細絲,從而導致該記憶體單元中之一高電流 峰值(由於關係I=V/R)且所需切換電流亦將係極高。 RSME具有相對於IL之一鏡像組態,乃因一 rsl及一電 極之序列在該IL之任一側上延伸。一鏡像組態亦可使用相 同材料用於RSL及電極。El、RSL1及IL之組合形成一第一 憶阻器(記憶體-電阻器)元件,且E2、RSL2及IL之組合形 成一第二憶阻器元件。該兩個憶阻器元件可係反串聯地或 串聯地連接成一個鏡像電阻開關(MRS)之雙極憶阻器元 件。 在使用中’當跨越E1及E2施加一電壓時,產生一電場 (E) ’該電場係由e 1與E2之間的距離劃分之電壓e il可浮 動’此意指其並非由一電壓/電流信號直接驅動而是可電 谷性麵合至由一電壓/電流信號直接驅動之一或多個其他 電極(例如,Ε1及/或Ε2)。由於電容性耗合,Ε1與Ε2之間 的電壓之一個部分將係自Ε1強加至輕合層且跨越RSL1, 而El與E2之間的電壓之另一部分將係自耦合層強加至^ 且跨越RSL2 ^根據每一 RSL之電阻跨越每一 RSL按比例劃 分該電壓。 此外,第一憶阻器可具有一第一I-V特性,而第二憶阻 1569l7.doc • 28· 201212319 器具有一第二Ι-V特性,以使得該記憶體單元之總特性 係第一憶阻器與第二憶阻器之〗_v特性之一疊加,但具有 該記憶體#元將以比個別,隐阻器元件低得多之電流操作之 額外益處。在一種方法中,第一憶阻器及第二憶阻器之“ V特性不同,但具有相同極性。纟另__方法中,第一憶阻 器及第二憶阻器之I-V特性具有相反極性。先前所論述之 圖4A至圖4E提供一 RSL之實例性;[·ν特性。 可以下文進一步詳述之諸多可能組態提供RSME之元 件。用於E1之實例性材料包含n+ Si(多晶矽)、p+ Si(多晶 石夕)、TiN、TiSix、TiAIN、TiAl、W、WN、WSix、Co、 CoSi、p+ Si、Ni及NiSi。用於rsli及RSL2之實例性材料 包含例如Me〇x及MeN之金屬材料。然而,亦可使用非金 屬材料,如在本文中某些實施例中所論述。尺乩丨及尺乩之 吓係相同類型或不同類型。一 RSL亦可係一相變單元、基 於碳、基於碳奈米管、奈米離子記憶體、導電橋或改變其 相、自旋、磁分量等之一單元。RSL可具有在]^〇範圍(例 如1 ΜΩ至10 ΜΩ或更多)之一導通電阻(導電狀態電阻)。 此與可程式化金屬化單元(PMC)(例如,一導電橋接或 CBRAM)相反,可程式化金屬化單元形成量子點接觸且具 有約25 ΚΩ或更低之一低得多之電阻。該較高電阻提供一 低電流操作及更佳可縮放性。 用於E2之實例性材料包含n+ si、n+ SiC、p+ SiC及p+Figure 4B is a graph showing one of the different I-V characteristics of two exemplary unipolar RSLs. For two or more unipolar RSLs, the IV (current-voltage) characteristics may be substantially the same such that, for example, 1 increases with V at a common rate, and sets and/or resets the level Can be substantially the same. Alternatively, the I-V characteristics of the RSL may be different such that, for example, for one of the RSLs, the V may be increased or set faster and/or the level may be reset differently. In this example, "A" represents a first type of RSL and "B" represents a second type of RSL, wherein the RSLs have different unipolar resistance switching characteristics. X 156917.doc •18- 201212319 The axis shows the voltage (V) and the y-axis shows the current (I). The type "A" RSL' lines 400, 402, 404, and 406 are the same as in Fig. 4A. Also for type "A" RSL, VsetA sets the voltage, VresetA resets the voltage 'IresetA' resets the current and Iset_limitA sets the current limit. For type "B" RSL, lines 420, 422, 424, and 426 correspond to lines 400, 402, 404, and 406, respectively. Also for type "B" RSL, VsetB sets the voltage, VresetB resets the voltage, IresetB resets the current, and Iset_limitB sets the current limit. In the method shown here, 'VsetA>VsetB, VresetA>VresetB, IresetA>IresetB and Iset_limitA>Iset_limitB, but this is only an example and other alternative relationships can be applied. When two or more RSLs are in the same RSME, the switching characteristics of the RSME will be a function of the switching characteristics of each of the RSLs. During a setup procedure, for example, if the voltage is equally divided across each RSL, the type "B" RSL can be switched before the type "A" RSL as V increases. Similarly, during a reset procedure, for example, assuming that the same voltage is applied in each RSL, the type "B" RSL can be switched before the type "A" RSL as V increases. Alternatively, the type "A" and "B" RSLs may have different I-V characteristics of opposite polarities. For example, 'set VsetA> 0 V and VresetA> 0 V, and VsetB < 0 V and VresetB < 0 V. As an example, the characteristics of type "A" RSL may be as described in Figure 4A, and the characteristics of type "B" RSL may be as described in Figure 4C below. In theory, one RSL in an RSME can also have a unipolar characteristic and the other RSL in the RSME has a bipolar characteristic. However, using only one of all RSLs in an RSME 156917.doc -19- 201212319 switching characteristics (unipolar or bipolar) allows for a simplified control scheme. In some cases, one of the RSMEs reads the data state of one of the RSLs. For example, in the case where the first RSL is in a low resistance state and a second RSL is in a high resistance state, assuming that the high resistance state is several orders of magnitude higher than the low resistance state, a read operation will be substantially undetectable. Current. That is, the resistance of the RSME equal to the sum of the resistances of each RSL will be extremely high, so the current will be extremely low or substantially zero. A read operation can switch the second RSL to a low resistance state such that the resistance of the RSME is low and the current through it is relatively high and detectable " after switching the second RSL back to the high resistance state Then, you can perform a writeback operation. When a voltage is applied across the electrodes of an RSME, the voltage is divided proportionally across each RSL based on the resistance of each RSL. When the first RSL is in a low resistance state and the second RSL is in a high resistance state, the first RSL transfers the potential at the electrode to IL such that substantially all voltages are applied across the second RSL. If this voltage is of appropriate magnitude and polarity, it will switch the second RSL. In addition, an RSL can be used as a material for operation as a single or bipolar device, for example, in "Coexistence of the bipolar and unipolar resistive switching behaviours in Au/SrTi03/Pt cells" by Sun et al. (March 10, 2011)曰, J. Phys. D: Appl. Phys. 44, 125404), which is incorporated herein by reference. 4C is a graph showing one of the I-V characteristics of another exemplary monopolar RSL. Compared to the characteristics of Figure 4A, a negative voltage is used during the setup and reset procedure to replace the positive voltage with 156917.doc -20-201212319. In the setup procedure, line 434 represents the I-V characteristic of RSL when in the high resistivity reset state, and line 436 represents one transition from the Vset down to the low resistivity set state. In the reset procedure, line 430 represents the I-V characteristic of RSL when in the low resistivity setting state, and line 432 represents one transition from the Vreset to the high resistivity reset state. Vread, Vreset, Vset, and Vf are all negative voltages. In a read operation, a fixed voltage Vread>Vreset can be applied in response to the lb of the measured current system setting state or the reset state. 4D is a graph showing one of the I-V characteristics of an exemplary bipolar RSL. Here, the opposite polarity voltage is used to set and reset the program. In addition, a positive voltage is used to set the program and a negative voltage is used to reset the program. In this bipolar RSL, a setting procedure occurs when a positive voltage is applied and a reset procedure occurs when a negative voltage is applied. In the setup procedure, line 444 represents the I-V characteristic of RSL when in the high resistivity reset state, and line 446 represents one transition from the Vset down to the low resistivity set state. In the reset procedure, line 440 represents the I-V characteristic of RSL when in the low resistivity setting state, and line 442 represents one transition from the Vreset to the high resistivity reset state. Vset and Vf are positive voltages and Vreset is a negative voltage. 4E is a graph showing one of the I-V characteristics of another exemplary bipolar RSL. In this bipolar RSL, a reset procedure occurs when a positive voltage is applied and a setting procedure occurs when a negative voltage is applied. In the setup procedure, line 454 represents the I-V characteristic of RSL when in the high resistivity reset state, and line 456 represents one transition from the Vset down to the low resistivity set state. In the reset procedure, line 450 represents the Ι-V characteristic of RSL when in the low resistivity setting state, and 156917.doc •21 - 201212319 line 452 represents one transition from Vreset to high resistivity reset state ^Vset and Vf is a positive voltage and Vreset is a negative power. Although the lreset level in Figures 4D and 4C is higher than the iset level, it should be emphasized that it can be reversed. This means that for the opposite polarity, the Iset level in Figure 4D and Figure 4C can be higher than the lreset level. Figure 5 illustrates an embodiment of a circuit for reading a state of a memory cell. A portion of a memory array includes memory cells 55A, 552, 554, and 556. The figure shows two of the many bit lines and two of the many word lines. Bit line 559 is coupled to cells 550 and 554, and bit line 557 is coupled to cells 552 and 556. For example, bit line 559 selects a location line and can be at 2 V. For example, bit line 557 is an unselected location line and can be grounded. For example, word line 547 is the selected word line and can be at 〇 V. For example, word line 549 is the unselected word line and can be at 2v. One of the read circuits for one of the bit lines 559 is shown as being connected to the bit line via transistor 558, The transistor is controlled by a gate voltage supplied by the row de-energizer to select or deselect the corresponding bit line. The transistor 558 connects the bit line to a data bus 563. Write circuit 560, which is part of system control logic 330, is coupled to the data bus. Transistor 562 is coupled to the data bus and operates as a home device controlled by clamp control circuit 564, which is part of system control logic 33. The transistor 562 is also coupled to a sense amplifier 566 that includes a data latch 568. The output of sense amplifier 566 is coupled to a data wheel output terminal (to system control logic 330, a controller, and/or a host). Write circuit 560 is also coupled to sense amplifier 566 and data latch 568. 156917.doc •22· 201212319 When attempting to read the state of RSME, first bias all word lines with Vread (for example, approximately 2 v) and place all bit lines at ground. Then pull the selected word line to ground. For example, this discussion will assume that the selection memory unit 550 performs a read i-transmission data bus (by conducting a conductive crystal 558) and a bin device (the transistor 562' receives a ~2 V+Vth (transistor 562). Threshold voltage)) Pull one or more selected positioning elements 559 to Vread. The gate of the home device is higher than Vread but is controlled to keep the bit line close to Vread. In one method, a current is pulled from a sense node of the sense amplifier through the selected memory cell 550 through the transistor 562. The sense node can receive a high resistivity state current and a low resistivity state. A reference current between the currents. The sense node moves in response to a current difference between the cell current and the reference current. The sense amplifier 566 generates a data output signal by comparing the sensed voltage with a reference read voltage. If the memory cell current is greater than the reference current, the memory cell is in a low resistivity state and the voltage at the sense node will be lower than the reference voltage. If the reciprocal cell current is less than the reference current, the memory cell is in a high resistivity state and the voltage at the sense node will be higher than the reference voltage. The output data signal from sense amplifier 566 is latched into data latch 568. Referring again to Figure 4A, for example, when in a high resistivity state, if a voltage Vset and a sufficient current are applied, the RSL will be set to a low resistivity state. Line 404 shows the behavior when Vset is applied. The voltage will remain fairly constant and the current will increase towards Iset by a limit. At some point, rsl will be set and the device behavior will be based on line 406. Note that the first time you set rsl, you need to form a voltage) to set the device. After that, the second device is set to 156917.doc •23·201212319. The formation voltage Vf can be greater than Vset in absolute magnitude. » When in the low resistivity state (line 400), rSl will be reset to the high resistivity state if vsset and sufficient current (Ireset) are applied. Line 4〇〇 shows the behavior when Vreset is applied. At some point, the RSL will be reset and the device behavior will be based on line 402. In one embodiment, 'Vset is about 7 V, Vreset is about 9 V, Iset_limit is about 1 〇 μΑ and Ireset can be as low as 1 〇〇 μΑ. These voltages and currents are suitable for the circuit of Figure 5 in which an RSME is connected to a diode. 6A through 6 can be, for example, a cross-sectional view of a vertical or horizontal plane along an RSMEi. Figure 6A illustrates an exemplary memory unit having an RSME and one of the steering elements (SE) located below the RSME. This memory unit can have various configurations. A configuration is a stacked configuration in which each type of material is provided in a layer and each layer is positioned below the layer above it and typically has a similar cross-sectional area. In another possible configuration, one or more layers may be configured end-to-end with one or more other layers (see Figures 6F-6J). Note that in these figures, any two layers or materials that are adjacent to each other may be in contact with each other. However, this is not required unless otherwise stated, and any two layers or materials depicted as being adjacent to each other may be separated by one or a plurality of other layers of material that are not depicted. Further, in some cases, a material may be formed as a by-product of fabrication, for example, one of the Si?x layers formed on a Si layer, which is not necessarily shown in the drawing. In addition, several variations of the described embodiments are possible. For example, the order of the layers in each embodiment can be reversed such that, for example, the word line is on top and the 156917.doc -24 - 201212319 bit line is on the bottom. One or more intermediate layers may be provided between each of the illustrated layers. In addition, the position of the steering element can be varied such that it lies above or below other layers that comprise the RSL. The orientation of the layers can be modified from vertical to horizontal or any other orientation. Multiple layers or portions that may form a common conductive path are considered to be connected in series. The memory cell comprises a bit line contact (BLC) material such as W or NiSi. The bit line contact material is connected to a bit line of a memory device. The bit line is a type of control line such that the BLC is also tied to one of the first - control lines. After the BLC in a series path is, for example, one of the first adhesion layers (AL1) of TiN, which helps the bLC to adhere to the RSME and act as a barrier. A TiN layer can be deposited by any conventional method such as sputtering. An operating element (SE)', such as a diode, is placed after rSN1e in the series path. The steering element allows a signal (eg, a voltage or current) to be selectively applied to one or more memory cells via word lines or bit lines to individually control the cells to, for example, by switching their RSMe Change the status of its individual data. The resistance switching behavior of RSME is not dependent on SE. The SE itself can have a resistance switching behavior; however, this behavior will not be dependent on the resistance switching behavior of the RSME. A second adhesive layer (AL2), such as TiN, is placed after the SE in the series path. After the AL2 in the series path is, for example, a W or NiSi word line contact (WLC) material that is connected to a word line of a memory device. The word line is a type of control line such that the WLC is also tied to one of the second control lines. The depicted portions of the memory cells are arranged in series as such. 1569I7.doc • 25-201212319 Figure 6B illustrates an alternative configuration with one of the rsME memory cells 'where the steering element (SE) is located above the RSME. It is also possible to reverse the order of the other layers from top to bottom, from bottom to top. 6C illustrates an exemplary embodiment of the RSME of FIG. 6A as a mirrored resistive switch (MRS) in a vertical stack. The RSME includes a first electrode (E1) (which is a top electrode in some configurations), a first resistance switching layer (RSL1), and a conductive intermediate layer (Jiang) that acts as a scattering layer, coupled electrode Or face layer). The RSME also includes a second rsl (RSL2) and a second electrode (EL2). The second electrode is a bottom electrode in some configurations. For example, such RSLs may be reversible RSLs. A reversible RSL can switch from one state to another and switch back to that state. The IL is electrically connected between E1 and E2 and electrically connected in series with E1 and E2. The rsli is electrically connected between E1 and IL and in series with £1 and specific electricity. RSL2 is electrically placed between E2 and IL and electrically connected in series with E2 and IL. "Electrically located" or similar words may mean in a conductive path. For example, the IL can be electrically located between E1 and E2 and physically located or not between £1 and £2. For example, s ' can form an RSME by connecting two bipolar memristor (memory_resistor) elements in anti-series into a mirrored resistance switch (MRS). A memristor is a passive two-terminal circuit component in which the resistance varies according to the current through the device and the history of the voltage across the device. The MRS can be made of a first memristor element comprising, for example, an n-type ElEl, RSL1 (which can be a transition metal oxide such as hafnium oxide (Hf02) or nitrogen oxides HfSiON) and an IL (which can be 156917.doc •26·201212319 one of the chemical reactions with oxygen can oxidize the electrode (eg TiN)). The RSME comprises a second memristor element made of the same (or different) material but in reverse order, the second memristor element sharing the oxidizable electrode of the IL. Moreover, in one method, both the first memristor element and the second memristor 70 can have bipolar or monopolar I-V (current-voltage) characteristics. In the other method, one of the memristor elements has a unipolar characteristic and the other has a bipolar characteristic. By combining two memristor elements into one RSME, the RSME has one of the Ι-ν characteristics of the memristor element superimposed on one of the IV characteristics but with the RSME* being lower than the individual memristor elements Additional benefits of operation at multiple currents. More generally, 'RSME will have the characteristics of RSL - superimposing one of the I-V characteristics but enabling operation at low currents. The IL acts as a scattering layer by scattering electrons from the RSL into the germanium thereby slowing down the flow of electrons in one of the switching mechanisms to avoid damaging an operating element. Further, the 'IL acts as a coupling electrode or layer that is capacitively coupled to one of the voltages of the RSME by setting the potential of E1 & E2. Through this scatter 'IL' provides a resistor that reduces one of the peak currents during a set or reset procedure while achieving a low current operation. It is believed that this current limiting operation is derived from two aspects of the IL layer. First, the hot electrons are excellently scattered in the IL layer by an electron-electron interaction. Second, one of the rsl initially collapses and delivers an excess charge q to the ratio, and the voltage applied across the RSL effectively reduces V=q/C, where the c-line IL layer faces the electrode E. ', 1 and E2's electricity valley. At the same time 'a higher voltage is now at the second team' to induce a collapse of the second RSL. Since the amount of charge q available is limited to 156917.doc -27· 201212319, the current that can flow here is also extremely limited. In this way, the RSME achieves operation of the memory cell at low currents. It is believed that this resistance is based on IL scattering electrons and produces an extremely effective negative feedback to the applied bias voltage to form the ability to allow small conductive filaments to switch at low currents. In the absence of IL, when a voltage is applied, one of the filaments with a very low resistance will be formed, resulting in a high current peak in the memory cell (due to the relationship I=V/R) and the required switching current is also Will be extremely high. The RSME has a mirrored configuration relative to one of the ILs, since a sequence of rsl and one of the electrodes extends on either side of the IL. A mirrored configuration can also use the same material for the RSL and electrodes. The combination of El, RSL1 and IL forms a first memristor (memory-resistor) component, and the combination of E2, RSL2 and IL forms a second memristor component. The two memristor elements can be connected in series or in series to form a mirrored resistance switch (MRS) bipolar memristor element. In use, when a voltage is applied across E1 and E2, an electric field (E) is generated. 'The electric field is a voltage e il that is divided by the distance between e 1 and E 2 . 'This means that it is not caused by a voltage / The current signal is directly driven but can be electrically gated to directly drive one or more of the other electrodes (eg, Ε1 and/or Ε2) from a voltage/current signal. Due to capacitive fit, a portion of the voltage between Ε1 and Ε2 will be imposed from Ε1 to the light-to-layer layer and across RSL1, while another portion of the voltage between El and E2 will be imposed on the self-coupling layer and cross RSL2^ divides the voltage proportionally across each RSL based on the resistance of each RSL. In addition, the first memristor may have a first IV characteristic, and the second memristor 1569l7.doc • 28·201212319 has a second Ι-V characteristic such that the total characteristic of the memory cell is the first memristor The superimposing of one of the _v characteristics of the second memristor, but with the memory ## will have the added benefit of operating at a much lower current than the individual, hidden resistor elements. In one method, the first memristor and the second memristor have different "V characteristics, but have the same polarity. In the other method, the first memristor and the second memristor have opposite IV characteristics. Polarity. Figures 4A through 4E previously discussed provide an example of an RSL; [·v characteristics. Components of the RSME can be provided in a number of possible configurations as described in further detail below. Example materials for E1 include n+ Si (polysilicon) ), p+ Si (polycrystalline), TiN, TiSix, TiAIN, TiAl, W, WN, WSix, Co, CoSi, p+ Si, Ni, and NiSi. Exemplary materials for rsli and RSL2 include, for example, Me〇x And MeN metal materials. However, non-metallic materials can also be used, as discussed in some embodiments herein. The scales and scales are of the same type or different types. An RSL can also be a phase change. Unit, carbon-based, carbon nanotube-based, nano-ion memory, conductive bridge or a unit that changes its phase, spin, magnetic component, etc. RSL can have a range of ( Ω to 10 Μ Ω or More) One of the on-resistance (conductive state resistance). This and the programmable metal In contrast to a cell (PMC) (eg, a conductive bridge or CBRAM), the programmable metallization cell forms a quantum dot contact and has a much lower resistance of about 25 ΚΩ or less. This higher resistance provides a low current operation. And better scalability. Example materials for E2 include n+ si, n+ SiC, p+ SiC, and p+
Si(多晶矽)、TiN、TiAIN、TiAl、w、WN Co、CoSi、p+Si (polysilicon), TiN, TiAIN, TiAl, w, WN Co, CoSi, p+
Si、Ni及NiSi。呈不同層形式之特定材料組合可係有利 156917.doc -29- 201212319 的。下文進一步詳細論述各種組態β C·、. ' 用於1L之實例性材料包含TiN、TiN、Α卜Zr、La、Υ、 Τι、ΤιΑΙΝ、TixNy、TiAl合金及 p+ SiC。因此,il可由一 可氧化材料(例如,TiN、A卜Zr、La、γ ' Ti)或一不可氧 化材料(例如,TiA1N、TixNy、TiA1合金及碳(包含例如石 墨烯、非晶碳、碳奈米管、具有不同晶體結構之碳及p+ SiC))製成。通常,可將£1及£2之相同材料用於江層。在 某些情況下,有意或無意地將一或多個氧化物層形成為沈 積及形成步驟之一副產物。舉例而言,可藉由在Si頂部上 沈積MeOx來氧化Sie甚至TiN或其他所提出之金屬可在一 個側上藉由Me〇x沈積被氧化且可在介面處藉由1^(^與 TiN之一介面反應被氧化。 如所提及,El、E2及IL由一導電材料製成。一導電材料 之特性可在於其導電率σ=1/ρ或其倒數,其倒數係電阻率 P=E/J。以西門子每米(s/m)量測導電率,且以歐姆米(ω_ m)或Ω-cm量測電阻率。Ε係以v/m計之電場之量值且;係以 A/m2計之電流密度之量值。對於一絕緣體,p>1〇8 或 σ<10-8 S/cm。對於一半導體,1〇·3 icm <ρ<1〇8 Ω. 或 103 S/cm >σ>1〇-8 S/cm。對於一導體,1〇·3 Q_cm >p或 103 S/cm <σ。一半導體可區別於一導體之處在於一半導 體通常藉由將一絕緣體摻雜成ρ型或η型半導體而形成,而 一導體不依靠摻雜。一半導體區別於一導體之處亦在於一 半導體允許電流基於一所施加電壓之極性而流動,以使得 電流可沿一個方向而不沿相反方向強烈地流動。一半導體 156917.doc •30· 201212319 允許一正向電流沿其流動之方向取決於其係一 p型還是η型 半導體。相比而言,一導體允許電流沿任一方向同樣良好 地流動。一導電材料意指包含一半導體(一半導電材料)及 一導體。一導體亦可稱為一導電材料。一導體具有高於一 半導體之一導電率。 注意,RSME不依靠一高帶隙三層式堆疊(一相對高帶隙 材料位於一相對低帶隙材料之若干層之間),乃因比係可 接收一耦合電壓之一導電材料。 圖6D繪示在RSL之間使用多個不同類型之IL之圖6 A之 RSME之一實例性實施方案。使用多個毗鄰中間層,包含 類型「1」之一第一 IL(ILl)及類型「2」之一第二 ]:L(IL2)。此實施例之一優點係該等IL可係具有不同散射性 質及功函數之不同類型以提供調節該rSME之效能之一額 外能力。此外,使用多個相同或不同類型之IL可增加路徑 中之散射/電阻,藉此降低電流,此乃因I=v/R。多個毗鄰 IL可增加散射,可如一較厚單個IL那樣。然而,一較厚比 提出一按比例縮放之挑戰,即:若堆疊高度增加,則柱_ 蝕刻之縱橫比增加。因此,例如蝕刻、清潔及間隙填充之 製造製程變為極具挑戰性。較佳地,可具有兩個(或更多 個)她鄰(或非毗鄰)較薄IL(或類似或相異性質/材料)代替一 個較厚IL。舉例而言,兩個厚度為5 nm之比可提供與(例 如)20 nm之一單個較厚il相當之散射。 舉例而言,IL1及IL2可係具有不同電阻率及晶體結構之 不同材料。IL 1及IL2亦可係相同材料’但可具有將以不同 156917.doc •31 · 201212319 方式散射電荷載流子之不同晶體結構或定向或不同晶粒大 小。作為另一實例,一個IL可由一細晶粒材料或奈米粒子 (其可與另一IL相同或不同)組成。 若RSL1及RSL2係不同材料,且IL1及IL2係不同材料及/ 或不同材料類型,則該等IL相對於該等RSL之最佳放置將 係材料相依的。 一個可能實施方案使用一 pn-接面,其中IL1為n+ Si且 IL2為p+ Si。舉例而言,IL1及IL2可各自具有至少20 nm之 一厚度。另一可能組合使用例如TiN之一金屬用於該等IL 中之一者且使用n+或p+ Si用於該等IL中之另一者。舉例而 言,參見圖10C。 圖6E繪示使用一重複RSL/IL型樣之圖6A之RSME之一實 例性實施方案。一 RSL與一 IL之型樣或組合重複至少兩 次。舉例而言,除RSL2及一第二IL(IL2)之外,還提供 RSL1及一第一 IL(ILl)。第三RSL(RSL3)係毗鄰於E2。該 等RSL可係相同或不同類型,且該等IL可係相同或不同類 型。此實施例之一優點係多個散射層可增加該RSME之路 徑中之散射/電阻量。另外,使用不同類型之IL及RSL之能 力提供調節RSME之效能之一額外能力。 三個RSL可具有大量特性(全部相同、兩個相同且一個不 同、全部不同等)。與相異RSL—起使用一個以上IL將改變 該RSME之特性且提供調諧其效能之額外功能性。 當跨越RSME施加一電壓時,根據每一 RSL之電阻跨越 每一 RSL按比例劃分該電壓。在一個可能實施方案中,該 156917.doc -32- 201212319 等RSL中之兩者具有相同Ι-V特性,且另一 RSL具有一不同 I-V特性,以使得舉例而言當該另一 RSL處於一高電阻狀態 時該兩個RSL皆處於一低電阻狀態,或當該另一 RSL處於 一低電阻狀態時該兩個RSL皆處於一高電阻狀態。其他變 化形式亦係可能的。 圖6F繪示其中RSME之每一層水平延伸且該等層中之一 或多者端對端地配置之圖6A之RSME之一實例性實施方 案。代替一完全堆疊(垂直)組態,該RSME之若干部分在 該RSME之其他部分側向(側面)配置,或與該等其他部分 端對端地配置。舉例而言,El、RSL1及IL在一個堆疊 中,而RSL2及E2在另一堆疊中,且RSL2與該IL並排配 置。參考圖6A,BLC及AL1可提供於E1上面,且SE、AL2 及WLC可提供於E2下面。在一個可能方法中,一非導電 (NC)層可提供於IL下方,且與E2並排配置。該RSME之該 等部分/層仍串聯配置。在另一可能實施方案中,E2位於 RSL2之側面上而非位於其下方,以使得三個部分(IL、 RSL2及E2)端對端地配置。其他變化形式亦係可能的。使 該RSME之若干部分端對端或以其他方式彼此側向地延伸 提供調節該RSME之佈局之一額外能力。舉例而言,可減 小該RSME之高度。在一種方法中,BLC及AL1可提供於 E1上面,且SE、AL2及WLC可提供於E2下面。 圖6G繪示其中RSME之每一層水平延伸且該等層中之一 或多者端對端地配置之圖6A之RSME之另一實例性實施方 案。該RSME之若干部分在該RSME之其他部分側向配置或 156917.doc -33· 201212319 與該等其他部分端對端地配置。RSLl、IL及RSL2在一個 堆疊中,而E2、一非導電層(NC)及E2在另一毗鄰堆疊中。 E1端對端地配置在RSL1之侧面處且E2端對端地配置在 RSL2之側面處。仍可將該等部分認為係串聯配置,例如 呈El、RSLl、IL、RSL2、E2之一串聯路徑。在另一選項 中,舉例而言,E1在RSL側向且在RSL上面延伸且E2在 RSL側向且在RSL下面延伸。在一種方法中,BLC及AL1可 提供於E1上面,且SE、AL2及WLC可提供於E2下面。 一般而言,可認為至少一個El、E2、IL、RSL1及RSL2 可至少部分地在El、E2、IL、RSL1及RSL2中之至少一個 另一者側向配置。 在圖6F及圖6G中,側向配置係端對端的。舉例而言, RSL1係與E1端對端地側向配置及/或RSL2係與E2端對端地 側向配置。此外,IL與RSL1及RSL2中之至少一者端對端 地侧向配置。 圖6H繪示其中RSME之每一層垂直延伸之圖6A之RSME 之另一實例性實施方案。該RSME之若干部分在該RSME之 其他部分側向或與該等其他部分面對面地配置。舉例而 言,該BLC可位於E1上面、下面或側面,而WLC位於E2上 面、下面或側面。BLC及WLC與該RSME在一串聯路徑 中。製造可涉及一層沈積及一層間隔件蝕刻之η個重複循 環與一最後CMP步驟。舉例而言,可將Ε1層沈積為一水平 延伸層,接著對其進行蝕刻以形成所示之垂直延伸部分。 接著,可將RSL1層沈積為一水平延伸層,接著對其進行 156917.doc • 34- 201212319 蝕刻以形成所示之垂直延伸部分。針對IL、RSL2及E2部 分中之每一者重複此操作。在一種方法中,AL1及BLC(圖 6A)自E1垂直向上延伸且SE、AL2及WLC自E2垂直向下延 伸0 該等層中之兩者或兩者以上可彼此面對面地側向配置。 舉例而言,RSL1、IL及RSL2可各自彼此面對面地側向配 置。此外,El、RSL1、IL、RSL2及E2可各自彼此面對面 地側向配置。 與圖61之L形剖面及圖6J之U形剖面相比,例如,圖6D 至圖6H之RSME部分具有一矩形剖面。 圖61繪示包含RSL1、IL、RSL2及E2之L形部分之圖6A之 RSME之另一實例性實施方案。舉例而言,假設剖視圖係 沿具有垂直軸X及y之一垂直或水平平面。沿X方向,E1具 有一厚度tlx,RSL1具有一厚度t2x,IL具有一厚度t3x, RSL2具有一厚度t4x且E2具有一厚度t5x。沿y方向,El具 有一厚度tly,RSL1具有一厚度t2y,IL具有一厚度t3y, RSL2具有一厚度t4y且E2具有一厚度t5y。針對每一部分, X方向厚度可與對應y方向厚度相同或不同。可反轉該等層 之次序,以使得該等層以E2、RSL2、IL、RSL1、E1代替 E:1、RSL1、IL、RSL2、E2之次序延伸。舉例而言,BLC 可位於E1上面、下面或側面,而WLC位於E2之上面、下 面或側面。BLC及WLC與該RSME在一串聯路徑中。藉由 提供L形部分,導電細絲可形成於該RSME之一設定程序 中,其中該等細絲沿X及y兩個方向延伸。由於存在細絲在 156917.doc -35- 201212319 其上方延伸之一相對大面積’因此潛在地促進其形成。所 描繪之該等實施方案亦可旋轉9〇°或18〇°。 在此方法中’類似於圖6F至圖6H之概念’該等層之若 干部分彼此側向地配置’但該等層呈嵌套L形’具有彼此 成一直角延伸之兩個部分。舉例而言’ L形RSL2嵌套在L 形E2内,L形IL嵌套在L形RSL2内’且L形RSL1嵌套在L形 IL内。E1嵌套在L形RSL1内,但在此實例中其自身並非L 形。每一部分在一或多個尺寸方面可係相同或不同。 此處,可認為至少一個El、E2、IL、RSL1及RSL2至少 部分地在El、E2、IL、RSL1及RSL2中之至少一個另一者 侧向配置。 圖6J繪示包含RSL1、IL、RSL2及E2之U形部分之圖6A 之RSME之另一實例性實施方案。舉例而言,假設剖視圖 係沿具有垂直轴X及y之一垂直或水平平面。沿X方向,E1 具有一厚度tlx,RSL1具有厚度t2xa及t2xb,IL具有厚度 t3xa及t3xb,RSL2具有厚度t4xa及t4xb且E2具有厚度t5xa 及t5xb。沿y方向,El具有一厚度tly,RSL1具有一厚度 t2y,IL具有一厚度t3y ’ RSL2具有一厚度t4y,且E2具有 一厚度t5y » xa厚度可與對應Xb厚度相同或不同。此外, xy厚度可與對應xa及/或Xb厚度相同或不同。可反轉該等 層之次序,以使得該等層以E2、RSL2、IL、RSL1、E1代 替El、RSL1、IL、RSL2、E2之次序延伸。舉例而言, BLC可位於E1上面、下面或側面,而Wlc位於E2之上面、 下面或側面。BLC及WLC與該RSME在一串聯路徑中。藉 156917.doc •36· 201212319 由提供U形部分,導電細絲可形成於該rsme之一設定程 序中’其中該等細絲沿X方向在E1之任一側上且沿y方向延 伸。所描繪之實施方案亦可旋轉9〇。或180。。 在此方法中’類似於圖6F至圖6H之概念,該等層之若 干部分彼此側向地配置,但該等層呈嵌套U形,其具有與 一基底部分成一直角延伸之兩個平行部分。舉例而言,u 形RSL2係嵌套在u形E2内,U形IL係嵌套在U形RSL2内, 且U形RSL1係嵌套在U形IL内。E1係嵌套在U形RSL1内, 但在此實例中其自身並非U形。每一部分在一或多個尺寸 方面可係相同或不同。 一般而言’可將垂直堆疊實施例中之任一者調適為一L 形或U形實施例。 此處,可認為至少一個El、E2、IL、RSL1及RSL2係至 少部分地在El、E2、IL、RSL1及RSL2中之至少一個另一 者側向配置。 圖6K1繪示使用一個RSL及位於該RSL下面之一個崩溃層 之圖6A之RSME之一實例性實施方案。如先前所論述使用 RSL1,但在IL與E2之間使用一崩潰層代替一 RSL2。該崩 潰層係不具有一電阻切換行為而可在匕與以之間提供—阻 擋層之一材料。具有一電阻切換行為之一材料通常可在開 始與結束電阻狀態之間反覆切換。相比而言,一崩潰材料 係已藉由施加一相對高電壓及/或電流自具有一相關聯^乂 特性之一初始狀態崩潰為具有另一相關聯Ι·ν特性之—崩 潰狀態且通常可自該初始狀態至該崩潰狀態僅轉變一次之 156917.doc -37- 201212319 一材料。可認為一電阻切換材料係一可多次程式化之材 料’而可認為一崩潰材料係一可單次程式化之材料。此 處’可程式化可包含具有改變一電阻狀態之能力。儘管一 電阻切換材料可與一炫絲或反溶絲配對以形成一可單次程 式化材料,但該電阻切換材料自身仍係可多次程式化的。 例如,一可單次程式化之材料在設定一晶片之一唯一識別 符或設定例如一時脈或電壓參數之操作參數時有用。 用於崩潰層之實例性材料(及在初始狀態中崩潰之前某 些實例之一相關聯電阻率P範圍)包含:siN(25 c下針對Si, Ni and NiSi. Specific material combinations in different layers may be advantageous 156917.doc -29- 201212319. The various configurations of β C·, . ' for 1L are discussed in further detail below, including TiN, TiN, Zr, La, Υ, Τι, ΤιΑΙΝ, TixNy, TiAl alloy, and p+ SiC. Thus, il may be an oxidizable material (eg, TiN, Ab Zr, La, γ 'Ti) or a non-oxidizable material (eg, TiA1N, TixNy, TiAl alloy, and carbon (including, for example, graphene, amorphous carbon, carbon) Made of nanotubes, carbon with different crystal structures and p+ SiC). Typically, the same material of £1 and £2 can be used for the river layer. In some cases, one or more oxide layers are intentionally or unintentionally formed as a by-product of the deposition and formation steps. For example, Sie or TiN can be oxidized by depositing MeOx on top of Si or other proposed metal can be oxidized on one side by Me〇x deposition and can be passed through the interface by 1^(^ and TiN One of the interface reactions is oxidized. As mentioned, El, E2, and IL are made of a conductive material. A conductive material may be characterized by its conductivity σ = 1 / ρ or its reciprocal, and its reciprocal resistivity P = E/J. Measure the conductivity per metre (s/m) of Siemens and measure the resistivity in ohm-meter (ω_ m) or Ω-cm. The magnitude of the electric field in volts/v is the system The magnitude of the current density in A/m2. For an insulator, p > 1 〇 8 or σ < 10-8 S/cm. For a semiconductor, 1 〇 · 3 icm < ρ < 1 〇 8 Ω. 103 S/cm >σ>1〇-8 S/cm. For a conductor, 1〇·3 Q_cm > p or 103 S/cm < σ. A semiconductor can be distinguished from a conductor in that a semiconductor is usually It is formed by doping an insulator into a p-type or n-type semiconductor, and a conductor does not rely on doping. A semiconductor is distinguished from a conductor in that a semiconductor allows current to be based on the polarity of an applied voltage. Flow so that current can flow strongly in one direction and not in the opposite direction. A semiconductor 156917.doc •30· 201212319 allows a forward current to flow along its direction depending on whether it is a p-type or an n-type semiconductor. In contrast, a conductor allows current to flow equally well in either direction. A conductive material is meant to comprise a semiconductor (semiconducting material) and a conductor. A conductor may also be referred to as a conductive material. A conductor has a higher than one One conductivity of a semiconductor. Note that RSME does not rely on a high bandgap three-layer stack (a relatively high bandgap material is located between several layers of a relatively low bandgap material), which is capable of receiving a coupling voltage. An electrically conductive material. Figure 6D illustrates an exemplary embodiment of the RSME of Figure 6A using a plurality of different types of IL between RSLs. Using a plurality of adjacent intermediate layers, including a first IL of type "1" ( IL1) and one of the types "2" second:: L(IL2). One advantage of this embodiment is that the ILs can have different scattering properties and different types of work functions to provide an additional measure of the performance of the rSME. can In addition, the use of multiple identical or different types of IL increases the scattering/resistance in the path, thereby reducing the current due to I=v/R. Multiple adjacent ILs can increase the scattering, as can be a thicker single IL. However, a thicker ratio presents the challenge of scaling up, that is, if the stack height is increased, the column-etch aspect ratio is increased. Therefore, manufacturing processes such as etching, cleaning, and gap filling become extremely challenging. Preferably, there may be two (or more) adjacent (or non-adjacent) thinner ILs (or similar or dissimilar properties/materials) in place of a thicker IL. For example, two ratios of 5 nm in thickness provide a scatter comparable to, for example, a single thicker il of 20 nm. For example, IL1 and IL2 can be of different materials having different resistivities and crystal structures. IL 1 and IL 2 may also be of the same material 'but may have different crystal structures or orientations or different grain sizes that will scatter charge carriers in different ways 156917.doc • 31 · 201212319. As another example, one IL may be composed of a fine grain material or nano particles (which may be the same or different from another IL). If RSL1 and RSL2 are different materials, and IL1 and IL2 are different materials and/or different material types, the optimal placement of these ILs relative to the RSLs will be material dependent. One possible embodiment uses a pn-junction where IL1 is n+Si and IL2 is p+Si. For example, IL1 and IL2 can each have a thickness of at least 20 nm. Another possible combination is to use one of the metals such as TiN for one of the ILs and n+ or p+ Si for the other of the ILs. For example, see Figure 10C. Figure 6E illustrates an exemplary embodiment of the RSME of Figure 6A using a repeating RSL/IL pattern. An RSL and an IL type or combination are repeated at least twice. For example, in addition to RSL2 and a second IL (IL2), RSL1 and a first IL (IL1) are provided. The third RSL (RSL3) is adjacent to E2. The RSLs may be the same or different types, and the ILs may be the same or different types. An advantage of this embodiment is that multiple scattering layers increase the amount of scattering/resistance in the path of the RSME. In addition, the ability to use different types of IL and RSL provides an additional capability to adjust the performance of RSME. The three RSLs can have a large number of characteristics (all identical, two identical and one different, all different, etc.). Using more than one IL with a different RSL will change the characteristics of the RSME and provide additional functionality to tune its performance. When a voltage is applied across the RSME, the voltage is divided proportionally across each RSL based on the resistance of each RSL. In one possible implementation, both of the RSLs of 156917.doc -32 - 201212319 have the same Ι-V characteristic, and the other RSL has a different IV characteristic such that, for example, when the other RSL is in a Both RSLs are in a low resistance state in the high resistance state, or both RSLs are in a high resistance state when the other RSL is in a low resistance state. Other variations are also possible. 6F illustrates an exemplary implementation of the RSME of FIG. 6A in which each layer of the RSME extends horizontally and one or more of the layers are configured end-to-end. Instead of a fully stacked (vertical) configuration, portions of the RSME are configured laterally (side) at other portions of the RSME, or are configured end-to-end with the other portions. For example, El, RSL1, and IL are in one stack, while RSL2 and E2 are in another stack, and RSL2 is configured side by side with the IL. Referring to FIG. 6A, BLC and AL1 may be provided on E1, and SE, AL2, and WLC may be provided under E2. In one possible approach, a non-conductive (NC) layer can be provided below the IL and configured side by side with E2. The parts/layers of the RSME are still arranged in series. In another possible embodiment, E2 is located on the side of RSL2 rather than below it, such that the three sections (IL, RSL2 and E2) are configured end to end. Other variations are also possible. Extending portions of the RSME end-to-end or otherwise laterally to each other provides an additional capability to adjust one of the RSME's layouts. For example, the height of the RSME can be reduced. In one method, BLC and AL1 can be provided on E1, and SE, AL2 and WLC can be provided below E2. 6G illustrates another exemplary implementation of the RSME of FIG. 6A in which each layer of the RSME is horizontally extended and one or more of the layers are configured end-to-end. Portions of the RSME are configured side-by-side with other portions of the RSME or 156917.doc -33.201212319 are configured end-to-end with the other portions. RSL1, IL and RSL2 are in one stack, while E2, a non-conductive layer (NC) and E2 are in another adjacent stack. E1 is disposed end-to-end at the side of RSL1 and E2 is disposed end-to-end at the side of RSL2. The portions may still be considered to be in series configuration, such as in series with one of El, RSL1, IL, RSL2, E2. In another option, for example, E1 is lateral to the RSL and extends above the RSL and E2 is lateral to the RSL and extends below the RSL. In one method, BLC and AL1 can be provided on E1, and SE, AL2 and WLC can be provided below E2. In general, at least one of El, E2, IL, RSL1, and RSL2 can be considered to be at least partially laterally disposed on at least one of El, E2, IL, RSL1, and RSL2. In Figures 6F and 6G, the lateral configuration is end-to-end. For example, RSL1 is laterally configured with E1 end-to-end and/or RSL2 and E2 end-to-side. Further, at least one of the IL and RSL1 and RSL2 is laterally disposed end to end. Figure 6H illustrates another exemplary embodiment of the RSME of Figure 6A in which each layer of the RSME extends vertically. Portions of the RSME are placed laterally or face to face with other portions of the RSME. For example, the BLC can be located above, below or to the side of E1, while the WLC is located above, below or to the side of E2. The BLC and WLC are in a series path with the RSME. Manufacturing may involve a repeating cycle of a layer of deposition and a spacer etch and a final CMP step. For example, the Ε1 layer can be deposited as a horizontally stretched layer, which is then etched to form the vertically extending portions shown. Next, the RSL1 layer can be deposited as a horizontally stretched layer, which is then etched 156917.doc • 34-201212319 to form the vertically extending portions shown. Repeat this for each of the IL, RSL2, and E2 sections. In one method, AL1 and BLC (Fig. 6A) extend vertically upward from E1 and SE, AL2, and WLC extend vertically downward from E2. 0 or more of the two layers may be laterally disposed face to face with each other. For example, RSL1, IL, and RSL2 can each be laterally configured face to face with each other. Further, El, RSL1, IL, RSL2, and E2 may each be laterally disposed face to face with each other. Compared to the L-shaped cross section of Fig. 61 and the U-shaped cross section of Fig. 6J, for example, the RSME portion of Figs. 6D to 6H has a rectangular cross section. 61 illustrates another exemplary embodiment of the RSME of FIG. 6A including the L-shaped portions of RSL1, IL, RSL2, and E2. For example, assume that the cross-sectional view is along a vertical or horizontal plane with one of the vertical axes X and y. In the X direction, E1 has a thickness tlx, RSL1 has a thickness t2x, IL has a thickness t3x, RSL2 has a thickness t4x and E2 has a thickness t5x. In the y direction, El has a thickness tly, RSL1 has a thickness t2y, IL has a thickness t3y, RSL2 has a thickness t4y and E2 has a thickness t5y. For each portion, the thickness in the X direction may be the same or different from the thickness in the corresponding y direction. The order of the layers may be reversed such that the layers extend in the order of E2, RSL2, IL, RSL1, E1 instead of E:1, RSL1, IL, RSL2, E2. For example, the BLC can be located above, below or to the side of E1, while the WLC is located above, below or to the side of E2. The BLC and WLC are in a series path with the RSME. By providing an L-shaped portion, conductive filaments can be formed in one of the RSME setting procedures, wherein the filaments extend in both X and y directions. Due to the relatively large area of filaments extending above 156917.doc -35 - 201212319, the formation of the filament is potentially promoted. The embodiments depicted may also be rotated by 9 〇 or 18 〇 °. In this method, 'similar to the concept of Figs. 6F to 6H', the same portions of the layers are disposed sideways to each other 'but the layers are nested L-shaped' having two portions extending at right angles to each other. For example, 'L-shaped RSL2 is nested within L-shaped E2, L-shaped IL is nested within L-shaped RSL2' and L-shaped RSL1 is nested within L-shaped IL. E1 is nested within L-shaped RSL1, but in this example it is not itself L-shaped. Each portion may be the same or different in one or more dimensions. Here, at least one of El, E2, IL, RSL1, and RSL2 may be considered to be at least partially laterally disposed at least one of El, E2, IL, RSL1, and RSL2. 6J illustrates another exemplary embodiment of the RSME of FIG. 6A including U-shaped portions of RSL1, IL, RSL2, and E2. For example, assume that the cross-sectional view is along a vertical or horizontal plane with one of the vertical axes X and y. In the X direction, E1 has a thickness tx, RSL1 has thicknesses t2xa and t2xb, IL has thicknesses t3xa and t3xb, RSL2 has thicknesses t4xa and t4xb, and E2 has thicknesses t5xa and t5xb. In the y direction, El has a thickness tly, RSL1 has a thickness t2y, IL has a thickness t3y ' RSL2 has a thickness t4y, and E2 has a thickness t5y » xa thickness which may be the same or different from the corresponding Xb thickness. Furthermore, the xy thickness may be the same or different than the corresponding xa and/or Xb thickness. The order of the layers may be reversed such that the layers extend in the order of E2, RSL2, IL, RSL1, E1 instead of El, RSL1, IL, RSL2, E2. For example, the BLC can be located above, below or to the side of E1, while Wlc is located above, below or to the side of E2. The BLC and WLC are in a series path with the RSME. By 156917.doc • 36· 201212319, by providing a U-shaped portion, conductive filaments can be formed in one of the rsme setting procedures wherein the filaments extend on either side of E1 in the X direction and in the y direction. The depicted embodiment can also be rotated by 9 turns. Or 180. . In this method, 'similar to the concept of Figures 6F to 6H, portions of the layers are arranged laterally to each other, but the layers are in a nested U shape having two parallel extending at right angles to a base portion section. For example, the u-shaped RSL2 is nested within the u-shaped E2, the U-shaped IL is nested within the U-shaped RSL2, and the U-shaped RSL1 is nested within the U-shaped IL. The E1 is nested within the U-shaped RSL1, but in this example it is not U-shaped. Each part may be the same or different in one or more dimensions. In general, any of the vertically stacked embodiments can be adapted to an L-shaped or U-shaped embodiment. Here, at least one of El, E2, IL, RSL1, and RSL2 may be considered to be at least partially laterally disposed at least one of El, E2, IL, RSL1, and RSL2. Figure 6K1 illustrates an exemplary embodiment of the RSME of Figure 6A using an RSL and a collapse layer located below the RSL. RSL1 is used as discussed previously, but a crash layer is used instead of an RSL2 between IL and E2. The collapse layer does not have a resistance switching behavior and can provide a material of one of the barrier layers between the turns. A material that has a resistance switching behavior typically toggles between the start and end resistance states. In contrast, a crash material has collapsed from an initial state with one of the associated characteristics by applying a relatively high voltage and/or current to a crash state with another associated Ι·ν characteristic and usually A material that can be changed only once from the initial state to the collapsed state. 156917.doc -37- 201212319 A material. A resistor-switching material can be considered to be a material that can be programmed multiple times, and a crash material can be considered to be a single-programmed material. This can be programmed to include the ability to change a resistive state. Although a resistive switching material can be paired with a dazzling or anti-solving filament to form a single-staged material, the resistive switching material itself can be programmed multiple times. For example, a single stylized material is useful when setting a unique identifier for a wafer or setting operational parameters such as a clock or voltage parameter. The example material used for the crash layer (and the associated resistivity P range for one of the instances before collapse in the initial state) contains: siN (for 25 c
Si3N4,p=i〇i4 Q-cm)、Si02(25 C下,p=l〇14至1〇16〇- cm)、SiC(p=l〇2至 106 Ω-cm)、SiCN、SiON或可崩潰之任 一層,例如自一較高電阻(通常係非導電狀態)改變為一較 低電阻(導電狀態),但其自身通常不稱為一可逆電阻切換 材料。該崩潰層可係維持至少約1河口至丨〇 ΜΩ之一電阻同 時在朋潰狀態中導電之一材料。初始狀態中之電阻通常比 在朋潰狀態中高一或多個數量級。若該層之電阻係太低, 則其作為一保護層不那麼有效。崩潰層材料之電阻係 pl/A其中1係材料之長度且A係剖面面積。該長度係崩 潰層厚度。知曉p及R之後,可使用八及丨選擇材料之大小。 該崩潰層可係一可單次程式化之崩潰層。可認為此一崩 潰層係一不可切換崩潰層或一可單次切換之崩潰層乃因 該崩潰程序係不可逆的。亦即’一旦該崩潰層自開始非導 電狀態崩潰,該崩潰層便保持在該崩潰狀態中且不可返回 至開始狀態《相比而言’在某些情況下,一單極或雙極單 156917.doc *38· 201212319 元可以可單次程式化之模式操作,但通常實際上未崩潰 同時在導電時維持至少約i ]^〇至1〇]^〇之一電阻。 例如,可藉由將一相對高電壓或電流施加至RSL來將一 或多個RSL組態成一崩潰狀態。舉例而言,一所施加電壓 可顯著高於該材料之臨限電壓。該崩潰程序可係部分地由 於熱效應。關於進一步細節參見圖6K2及圖6K3。 圖6K2係展示一崩溃層自一初始狀態至一崩潰狀態之一 轉變之一圖表。該轉變可藉由跨越該崩潰層施加一電流或 電壓達一時間週期而達成,該時間週期可延長(例如)數分 鐘。在時間tb ’穿過該崩潰層之一電流逐步增加(由於該電 阻逐步降低),此時發生一崩潰事件。在某些情況下,可 發生多個崩潰事件。對於一所施加電壓,根據該崩潰層及 RSL1之各別電阻跨越該崩潰層及RSL1按比例劃分施加於 該RSME兩端之電壓《可將rSL1組態成一低電阻狀態,以 使得基本上跨越該崩潰層施加所有電壓。 圖6K3係展示一崩潰層處於一初始狀態(實線)及一崩潰 狀態(虚線)之一 I-V特性之一圖表。對於一給定電壓,電流 在崩潰狀態中係較高(且電阻係較低)。其中一崩潰層處於 初始狀態之一 RSME可區別於其中該崩潰層處於崩潰狀態 之一 RSME,以使得可根據該崩潰層之狀態儲存一資料位 元。可進一步在兩個狀態之間調變RSL以儲存一資料位 元。藉由施加適當讀取電壓,可判定該崩潰層及RSL之狀 態。 圖6L繪示使用一個可逆RSL(RSLl)及位於RSL1上面之一 156917.doc -39- 201212319 個崩潰RSL之圖6A之RSME之一實例性實施方案。此係圖 6K1之組態之一替代組態。 圖6M繪示其中電阻切換層(rsl)係不同類型之圖6A之 RSME之一實例性實施方案。rSL1及RSL2可由具有不同切 換特性之不同類型之材料製成(例如)以允許該RSme儲存 一個以上資料位元。用於RSL 1及RSL2之實例性材料包 含:Ti02、NiOx、HfSiON、HfOx、Zr02及 ZrSiON。 圖7A繪示作為一 Si二極體之圖6A之記憶體單元之操縱 元件(SE)之一實例性實施方案。該冗係具有一 η型區、一 本質(i)區及一 ρ型區之一 Si二極體。如所提及,一 SE選擇 性地限制跨越該RSME之電壓及/或穿過該rsmE之電流。 該SE允許在不影響一陣列中之其他記憶體單元之狀態之情 形下寫入至一記憶體單元及/或自一記憶體單元讀取。 圖7B繪示作為一穿通二極體之圖6A之記憶體單元之操 縱元件(SE)之一實例性實施方案。該穿通二極體包含一 n+ 區、P-區及n+區《—穿通二極體可沿兩個方向操作。特定 而言’一穿通二極體允許一交叉點式記憶體陣列之雙極操 作’且可具有一對稱非線性電流/電壓關係。該穿通二極 體針對選定單元在高偏壓下具有一高電流且針對未選定單 元在低偏壓下具有一低茂漏電流。因此,其可與具有電阻 切換元件之交又點式記憶體陣列中之雙極切換相容。該穿 通二極體可係一 n+/p-/n+裝置或一 p+/n-/p+裝置》 儘管提供涉及具有二極體作為一操縱元件之一記憶體單 元之實例性實施方案,但本文中提供之技術通常可應用於 156917.doc -40· 201212319 其他裝置及操縱元件,包含一電晶體、一穿通電晶體、一 穿通二極體、一 PN二極體、NP二極體、一 PIN二極體、齊 納二極體、一 NPN二極體、PNP二極體、一肖特基二極 體、一 MIN二極體、碳聚矽氧二極體、一電晶體佈局等 等。 在另一方法中,操縱元件可係一電晶體,例如一雙極或 CMOS電晶體。 此外,在某些組態中,不需要使用一操縱元件。 圖8繪示連接於一位元線與一字線之間的圖6A之記憶體 單元之一實例性實施方案。位元線觸點(BLC)係W或 NiSi,第一黏合劑層(AL1)係TiN,第一電極(E1)係n+ Si, RSL1 係 MeOx(例如 Hf02),IL 係 TiN,RSL2 係 MeOx(例如 Hf02)、為係操縱元件(SE)之Si二極體提供一額外黏合劑 層(AL),第二黏合劑層(AL2)係TiN且字線觸點(WLC)係W 或NiSi。另外,可使用選自由以下構成之群組之一材料來 提供一或多個頂蓋層:TiOx、A1203、ZrOx、LaOx及 YOx。通常,該頂蓋層可係一金屬氧化物。在此實例中, 該等頂蓋層係毗鄰於IL及RSL。具體而言,一個頂蓋層 (〇&?1)位於1181^1與11^之間又毗鄰於118[1及11^中之每一 者,且另一頂蓋層(Cap2)位於IL與RSL2之間又毗鄰於IL及 RSL2。自MeOx之角度來看,一頂蓋層可充當一氧源或吸 氧劑,此促進一 RSL中之切換。當充當一吸氧劑時,舉例 而言,該頂蓋層可幫助將氧自一MeOx RSL提供至一IL/電 極。當充當一氧源時,舉例而言,該頂蓋層可幫助將氧自 156917.doc -41 · 201212319 一 IL/電極提供至一 MeOx RSL。一吸氣劑係其中例如氧之 一材料移動至之一位置。吸氣係其中例如氧之材料移動至 該吸氣劑位置之一程序。該吸氣劑位置係其中由於氧處於 一較低能量狀態而將優先駐存於其中之一替代位置。 RSME係由自E1延伸至E2之層組成。在一實例性實施方 案中,舉例而言,E1及E2各自具有約1至3 nm或約1至1〇 nm之一厚度或高度,且舉例而言,江可具有約1至5 nm或 約1至10 nm之一厚度或高度。因此,該rSME之總厚度可 係極小。 圖9A繪示圖6C之RSME之一實施例,其中E1係由Co、 CoSi、n+ Si、p+ Si或p+ SiC製成且E2係由n+ Si製成。自 頂部至底部之層次序係:El、RSL1、Capl、IL、Cap2、 RSL2、E2。該RSME亦包含例如MeOx之一 RSL 1、例如TiN 之一 IL、例如MeOx之一 RSL2及例如n+ Si之一第二電極 (E2)。另外,在RSL^IL(Capl)之間且在 之間提供例如TiOx之頂蓋層。當£1及£2係由不同材料製成 時,此實施例可提供一不對稱結構。舉例而言,由鈷(c〇) 製成之El係期望的,乃因其具有接近犯之功函數之約_5 eV之一相對高功函數且可導致更佳切換。此係由於可係具 有一咼功函數之一益處之一較高障壁高度。在另一方法 中由石夕化姑(c〇Si)製成之E1亦係期望的,乃因其亦具有 相對向功函數。在另一方法中,E1係由n+si(多晶矽)製 成’其提供一高功函數(約4丨至4 15 eV)以及抗氧化之益 處。其他適合材料包含p+ Si(多晶矽),其具有約5」至5 2 156917.doc -42· 201212319 eV之功函數,及p+碳化碎(训),自於一高能隙其具有 約 6.6至 6.9 eV^· — ·*·, ^ 極咼功函數。參見圖9C。舉例而言,對 於(4H多型體)’能隙為約3_23 eV且對於6H α多型體,能隙 為約3·05 eV。此等能隙顯著高於(舉例而言)si之能隙,對 於Si ’能隙為約ij eV。 在一項實施例中’可沈積P+ SiC且接著(例如)藉由離子 植入用一摻雜劑(例如,B '以、Be或Ga)摻雜至每立方公 分約10E19至10E20個原子之一濃度。此係原位摻雜之一實 例。SiC化學上係惰性的且因此抗氧化。由於27〇〇 ^之一 昇華溫度,SiC實際上不熔化,且具有3 6至4 9 w/(cm*K) 之一高導熱率(與Si之1.49 W(cm*K)相比),該高導熱率可 由於高電流密度而有益於記憶體單元操作。 圖9B繪示圖6C之RSME之一實施例,其中E1及IL係由p+ SiC製成’且E2係由n+ Si、n+ SiC或p+ SiC製成。從頂部 至底部之層次序係:El(例如,p+ SiC)、RSL1、IL(例 如’ p+ SiC)、RSL2、E2。El及IL之高功函數可引起單元 電流減小,其中IL充當一散射層。此外,藉由調變il之摻 雜’可調變層電阻以增加散射及減小電流❶隨著摻雜增 加,IL電阻性減小,以使得在空乏層上存在較小空乏寬度 及較小電壓降。 此外,E2可由n+ Si、n+ SiC或p+ SiC製成。當E2係由n+ SiC製成時’存在於製造期間形成於E2與RSL2之間的一較 薄Si02層。由於避免跨越一 Si02層之一電壓降,因此操作 電壓減小。相比而言,在一 n+ Si底部電極之情況下,一較 156917.doc •43· 201212319 厚’2層可形成於E2與RSL2之間。作為n+ Sia替代實 例’ E2可由p+ siC製成。舉例而言,rsli及RSL2可係 MeOx。 在一種方法中’ IL可(例如)藉由將IL提供為一奈米結晶Si3N4, p=i〇i4 Q-cm), SiO 2 (at 25 C, p=l〇14 to 1〇16〇-cm), SiC (p=l〇2 to 106 Ω-cm), SiCN, SiON or Any layer that can collapse, for example, changes from a higher resistance (typically a non-conductive state) to a lower resistance (conductive state), but is not itself referred to as a reversible resistance-switching material. The collapse layer can maintain a resistance of at least about one estuary to one of the 丨〇 Ω Ω while conducting one of the materials in the puddle state. The resistance in the initial state is typically one or more orders of magnitude higher than in the peer state. If the resistance of the layer is too low, it is less effective as a protective layer. The resistance of the material of the collapse layer is pl/A, the length of the material of the 1 series and the cross-sectional area of the A system. This length is the thickness of the collapse layer. After knowing p and R, you can use eight and 丨 to select the size of the material. The crash layer can be a single-programmed crash layer. It can be considered that this collapse layer is a non-switchable crash layer or a crash layer that can be switched once because the crash procedure is irreversible. That is, once the crash layer has collapsed from the beginning of the non-conducting state, the crash layer remains in the crash state and cannot be returned to the start state. In contrast, in some cases, a single pole or bipolar single 156917 .doc *38· 201212319 The unit can be operated in a single stylized mode, but usually does not actually crash and maintains at least about one of the resistances of i)^1 to 1〇. For example, one or more RSLs can be configured into a collapsed state by applying a relatively high voltage or current to the RSL. For example, an applied voltage can be significantly higher than the threshold voltage of the material. This crash procedure can be partly due to thermal effects. See Figure 6K2 and Figure 6K3 for further details. Figure 6K2 shows a graph of one of the collapse levels from an initial state to a collapsed state. This transition can be achieved by applying a current or voltage across the breakdown layer for a period of time that can be extended, for example, by a few minutes. At time tb', the current through one of the collapsed layers is gradually increased (due to the gradual decrease in the resistance), and a collapse event occurs at this time. In some cases, multiple crash events can occur. For an applied voltage, the voltage applied across the RSME is proportionally divided across the breakdown layer and RSL1 according to the breakdown layer and the respective resistors of RSL1. "rSL1 can be configured to a low resistance state so as to substantially cross the The breakdown layer applies all voltages. Figure 6K3 shows a graph of one of the I-V characteristics of a collapsed layer in an initial state (solid line) and a collapsed state (dashed line). For a given voltage, the current is higher in the collapsed state (and the resistance is lower). One of the crash layers is in an initial state. The RSME can be distinguished from one of the RSMEs in which the crash layer is in a crash state so that a data bit can be stored according to the state of the crash layer. The RSL can be further modulated between the two states to store a data bit. The state of the collapse layer and the RSL can be determined by applying an appropriate read voltage. Figure 6L illustrates an exemplary embodiment of the RSME of Figure 6A using a reversible RSL (RSL1) and one of the RSL1 156917.doc -39 - 201212319 crash RSLs. One of the configurations of this figure 6K1 replaces the configuration. Figure 6M illustrates an exemplary embodiment of the RSME of Figure 6A in which the resistance switching layer (rsl) is of a different type. rSL1 and RSL2 can be made of different types of materials having different switching characteristics, for example, to allow the RSme to store more than one data bit. Exemplary materials for RSL 1 and RSL 2 include: Ti02, NiOx, HfSiON, HfOx, Zr02, and ZrSiON. Figure 7A illustrates an exemplary embodiment of a steering element (SE) of the memory cell of Figure 6A as a Si diode. The redundancy has an n-type region, an intrinsic (i) region, and a p-type one Si dipole. As mentioned, a SE selectively limits the voltage across the RSME and/or the current through the rsmE. The SE allows writing to and/or reading from a memory unit without affecting the state of other memory cells in an array. Figure 7B illustrates an exemplary embodiment of an operational element (SE) of the memory cell of Figure 6A as a feedthrough diode. The punch-through diode includes an n+ region, a P-region, and an n+ region. The punch-through diode can operate in two directions. In particular, a pass-through diode allows for a bipolar operation of a cross-point memory array and can have a symmetric nonlinear current/voltage relationship. The feedthrough diode has a high current at high bias for selected cells and a low leakage current at low bias for unselected cells. Therefore, it is compatible with bipolar switching in a point-and-point memory array having a resistance switching element. The punch-through diode can be an n+/p-/n+ device or a p+/n-/p+ device. Although an example implementation involving a memory cell having a diode as a steering element is provided, in this context The technology provided is generally applicable to other devices and operating elements of 156917.doc -40· 201212319, including a transistor, a through-current crystal, a punch-through diode, a PN diode, an NP diode, and a PIN two. A polar body, a Zener diode, an NPN diode, a PNP diode, a Schottky diode, a MIN diode, a carbon polyoxynitride, a transistor layout, and the like. In another method, the steering element can be a transistor, such as a bipolar or CMOS transistor. In addition, in some configurations, it is not necessary to use a steering element. Figure 8 illustrates an exemplary embodiment of the memory cell of Figure 6A coupled between a bit line and a word line. The bit line contact (BLC) is W or NiSi, the first adhesive layer (AL1) is TiN, the first electrode (E1) is n+Si, RSL1 is MeOx (for example, Hf02), IL is TiN, and RSL2 is MeOx ( For example, Hf02) provides an additional adhesive layer (AL) for the Si diode of the system control element (SE), the second adhesive layer (AL2) is TiN and the word line contact (WLC) is W or NiSi. Alternatively, one or more cap layers may be provided using one selected from the group consisting of TiOx, A1203, ZrOx, LaOx, and YOx. Typically, the cap layer can be a metal oxide. In this example, the cap layers are adjacent to the IL and RSL. Specifically, a cap layer (〇 & 1) is located between 1181^1 and 11^ and adjacent to each of 118 [1 and 11^, and another cap layer (Cap2) is located at IL. It is adjacent to IL and RSL2 between RSL2. From the perspective of MeOx, a cap layer can act as an oxygen source or oxygen scavenger, which facilitates switching in an RSL. When used as an oxygen absorbing agent, for example, the cap layer can help provide oxygen from a MeOx RSL to an IL/electrode. When acting as an oxygen source, for example, the cap layer can help provide oxygen from a 156917.doc -41 · 201212319 - IL / electrode to a MeOx RSL. A getter is one in which a material such as oxygen is moved to a position. The getter is a program in which, for example, the material of oxygen moves to the position of the getter. The getter position is where one of the alternate locations will preferentially reside due to the oxygen being in a lower energy state. RSME consists of layers extending from E1 to E2. In an exemplary embodiment, for example, E1 and E2 each have a thickness or height of about 1 to 3 nm or about 1 to 1 〇 nm, and for example, jiang can have about 1 to 5 nm or about One thickness or height from 1 to 10 nm. Therefore, the total thickness of the rSME can be extremely small. 9A illustrates an embodiment of the RSME of FIG. 6C in which E1 is made of Co, CoSi, n+Si, p+Si, or p+ SiC and E2 is made of n+Si. The order from top to bottom is: El, RSL1, Capl, IL, Cap2, RSL2, E2. The RSME also comprises, for example, one of MeOx RSL 1, such as one of TiN, one of, e.g., one of MeOx, RSL2, and one of, for example, n+Si, a second electrode (E2). Further, a cap layer such as TiOx is provided between and between RSL^IL(Capl). This embodiment provides an asymmetric structure when £1 and £2 are made of different materials. For example, an El made of cobalt (c〇) is desirable because it has a relatively high work function of approximately _5 eV close to the work function of the guilt and can result in better switching. This is due to the fact that one of the benefits of one of the work functions can be tied to a higher barrier height. In another method, E1 made of Shi Xihua Gu (c〇Si) is also desirable because it also has a relative work function. In another method, E1 is made of n+si (polycrystalline germanium) which provides a high work function (about 4 丨 to 4 15 eV) and the benefits of oxidation. Other suitable materials include p+Si (polycrystalline germanium) having a work function of about 5" to 5 2 156917.doc -42 · 201212319 eV, and p + carbonized powder, which has a 6.6 to 6.9 eV from a high energy gap. ^· — ·*·, ^ Extreme work function. See Figure 9C. For example, for (4H polytype) the energy gap is about 3-23 eV and for the 6H alpha polytype, the energy gap is about 3·05 eV. These energy gaps are significantly higher than, for example, the energy gap of si, which is about ij eV for the Si' energy gap. In one embodiment, 'p+ SiC can be deposited and then doped with a dopant (eg, B', Be or Ga) by ion implantation to about 10E19 to 10E20 atoms per cubic centimeter, for example. A concentration. This is an example of in-situ doping. SiC is chemically inert and therefore resistant to oxidation. SiC does not actually melt due to a sublimation temperature of 27〇〇^, and has a high thermal conductivity of 3 6 to 49 w/(cm*K) (compared to 1.49 W (cm*K) of Si), This high thermal conductivity can benefit memory cell operation due to high current density. Figure 9B illustrates an embodiment of the RSME of Figure 6C in which E1 and IL are made of p+ SiC' and E2 is made of n+ Si, n+ SiC or p+ SiC. The order from the top to the bottom is: El (e.g., p+ SiC), RSL1, IL (e.g., 'p+ SiC), RSL2, E2. The high work function of El and IL can cause a decrease in cell current, where IL acts as a scattering layer. In addition, by adjusting the doping of the il's adjustable layer resistance to increase the scattering and reduce the current ❶ as the doping increases, the IL resistivity decreases, so that there is a small depletion width and a small amount on the depletion layer. Voltage drop. Further, E2 may be made of n+Si, n+ SiC or p+ SiC. When E2 is made of n+ SiC, there is a thin SiO 2 layer formed between E2 and RSL2 during fabrication. Since the voltage drop across one of the Si02 layers is avoided, the operating voltage is reduced. In contrast, in the case of an n+ Si bottom electrode, a thicker layer of 156917.doc •43·201212319 can be formed between E2 and RSL2. As an alternative to n+ Sia, 'E2 can be made of p+ siC. For example, rsli and RSL2 can be MeOx. In one method, 'IL can be provided, for example, by providing IL to one nanocrystal.
SiC膜而由奈米粒子製成。例如,參見下文論述之w, 等 人0 圖9C係繪示p+ SiC相對於其他材料之費米能階之一圖 示。上文曾提及p+ Sic由於一高能隙而具有約6.6至6.9 ev 之一極高功函數。為說明此事實,針對4H-SiC提供一能量 圖’該圖繪示一真空中之能階(Evacuum)、導帶之能階 (Ec)、本質能階(Ei)及價帶之能階(Εν)。該圖示係來自以引 用方式併入本文中之T. Ayalew之論文「SiC SemiconductorThe SiC film is made of nano particles. For example, see w, discussed below, and Figure 0C shows one of the Fermi levels of p+ SiC versus other materials. It has been mentioned above that p+Sic has a very high work function of about 6.6 to 6.9 ev due to a high energy gap. To illustrate this fact, an energy diagram is provided for 4H-SiC. This figure shows the energy level (Evacuum) in the vacuum, the energy level (Ec) of the conduction band, the essential energy level (Ei), and the energy level of the valence band ( Εν). This illustration is from T. Ayalew's paper "SiC Semiconductor", incorporated herein by reference.
Devices Technology, Modeling And Simulation」(2004年 1 月 ’ Institute for Microelectronics,Vienna,Austria)。亦繪 示其他實例性材料及其費米能階:Al(4.28 eV)、Ti、Devices Technology, Modeling And Simulation" (January 2004 'Institution for Microelectronics, Vienna, Austria). Other example materials and their Fermi level are also shown: Al (4.28 eV), Ti,
Zn(4.33 eV)、W(4.55 eV)、M〇(4.60 eV) ' Cu(4.65 eV)、 Ni(5.10 eV)、Au(5.15 eV)及 Pt(5.65 eV)。如所提及,p+Zn (4.33 eV), W (4.55 eV), M〇 (4.60 eV) 'Cu (4.65 eV), Ni (5.10 eV), Au (5.15 eV), and Pt (5.65 eV). As mentioned, p+
SiC具有一相對高功函數。特定而言,費米能階將接近價 帶能階。 實務中,未經摻雜SiC具有約4.5至4.8 eV或若被氧覆蓋 約4.9 eV之一功函數。然而,對於p+ SiC,費米能階將更 接近價帶’以使得功函數係較高。相依於P+摻雜之位準及 SiC多型體(對於4H-SiC ’能帶間隙Eg=3.23-3.26eV或對於 611-8丨(:,£8=3.056¥)’功函數£1(|)]^可係如所示約6.6至6.9 156917.doc • 44 - 201212319 eV。SiC has a relatively high work function. In particular, the Fermi level will be close to the valence band. In practice, undoped SiC has a work function of about 4.5 to 4.8 eV or about 4.9 eV if covered by oxygen. However, for p+ SiC, the Fermi level will be closer to the valence band' to make the work function higher. Dependent on P+ doping level and SiC polytype (for 4H-SiC 'band gap Eg=3.23-3.26eV or for 611-8丨(:, £8=3.056¥)' work function £1(| )]^ can be as shown in the figure 6.6 to 6.9 156917.doc • 44 - 201212319 eV.
可藉由在不過高之一適當溫度下進行沈積來施加siC。 針對相對低溢度沈積可存在各種技術。舉例而言,在 750 C下之沈積已闡述於以引用之方式併入本文中之LThe siC can be applied by depositing at a temperature that is not too high. There are various techniques for depositing relatively low spills. For example, deposition at 750 C is set forth in the context of which it is incorporated herein by reference.
Golecki 專人之「Single-crystalline, epitaxial cubic SiC films grown on (100) Si at 750 °C by chemical vapor deposition」(1992年4 月,Applied Physics Letter,第 60 卷’第14期’ 1703至1705頁)》在此方法中,SiC膜係使用 曱基矽烷(SiCH3H3)(—種具有1:1之si:c比率之一單個前驅 物)及H2藉由低壓化學氣相沈積生長而成。 在另一實例性方法中,已在低溫下使用分子束磊晶來沈 積SiC ’如在以引用之方式併入本文中之a Fissel等人之 rLow-temperaturegrowthofSiCthinfilmsonSiand6H-SiC by solid-source molecular beam epitaxy」(1995年 6月, Applied Physics Letter,第 66卷、第 23 期、3182 頁至 3184 頁)中所述。此方法涉及在約800至1000°C之低溫下使用受 一基於四極質譜儀之通量計控制之固源分子束磊晶於 Si(lll)及2°至5°偏斜定向之6H-SiC(0001)基板上磊晶生長 化學計量SiC。在SiC(0001)之情況下,在展示(3x3)及(2x2) 超結構之Si穩定之表面上獲得膜。在T>900°C下於6H-SiC(0001)上之生長期間之反射高能衍射(rheED)圖案及阻 尼RHEED振盪指示階地上之二維成核係主要生長程序。 用於沈積SiC之另一實例性低溫方法係闡述於以引用之 :¾式併入本文中之W. Yu等人之「Low temperature 156917.doc •45- 201212319 deposition of hydrogenated nanocrystalline SiC films by helicon wave plasma enhanced chemical vapor deposition」 (2010年 9 月 3 曰之 J. Vac. Sci. Technol. A 28(5),American Vacuum Society,1234頁至1239頁)中。此處,藉由使用螺 旋波電漿增強型化學氣相沈積技術在低基板溫度下沈積氫 化奈米結晶碳化矽(nc-SiC:H)膜。已調查了射頻(rf)功率及 基板溫度對所沈積nc-SiC:H膜之性質之影響。已發現氫化 非晶SiC膜係在一低rf功率下製造而成,而具有嵌入於非晶 對等部分中之SiC奈米晶體之一微結構之nc-SiC:H膜可係 當rf功率為400 W或以上時沈積而成。自電容性主導放電 至具有高電漿強度之螺旋波放電之電漿轉變影響膜微結構 及表面形態。對在各種基板溫度下沈積之膜之分析揭示 SiC結晶之開始發生在低達150°C之基板溫度下。 圖10A繪示闡述替代性IL材料之圖6C之RSME之一實施 例。自頂部至底部之層次序係:El(例如,TiN)、E1(例 如,n+ Si)、RSL1 (例如,MeOx)、cap 1 (例如,TiOx)、 IL(例如,TiN)、cap2(例如,TiOx)、RSL2(例如, MeOx)、E2(例如,n+ Si)。在一項實施方案中,El包含一 n+ Si層上面之一 TiN層之一組合。另外,在RSL1與IL之間 且在IL與RSL2之間提供頂蓋層,例如TiOx。一進一步Ti觸 點(圖中未展示)可位於E1上面。作為一替代方案,IL可選 自由以下構成之群組:Al、Zr、La、Y、Ti、TiAIN、 TixNy及TiAl合金。此等材料係可實現較低V及I單元操作 之較佳耦合層。此實施例提供關於IL·之一鏡像結構,乃因 156917.doc -46- 201212319 一頂蓋層、一 RSL及一電極以相同次序自該IL之兩側延伸 且視情況係相同材料(例如,IL上面及下面之相同頂蓋層 材料,例如TiOx,後跟IL上面及下面之相同RSL材料,例 如MeOx,後跟IL上面及下面之相同電極材料,例如n+ Si)。 圖1 0B繪示呈一倒置鏡像堆疊組態之圖6C之RSME之一 實施例。自頂部至底部之層次序係:El(例如,TiN)、 capl(例如,TiOx)、RSL1(例如,MeOx)、IL(例如 ’ n+ Si)、RSL2(例如,MeOx)、cap2(例如,TiOx)、E2(例如, TiN)。在一種方法中,El係由TiN製成,IL係由n+ Si製成 且E2係由TiN製成。舉例而言,IL層可係具有一 10至100 nm厚度之n+Si。此實施例係一倒置鏡像組態,相對於圖 10A之實施例,該組態提供一倒置堆疊,乃因n+ Si層現在 係IL而非El或E2層,且頂蓋層係位於RSL與電極層之間 (capl在RSL1與E1之間;cap2在RSL2與E2之間)代替在RSL 與IL之間。具體而言,一 RSL、一頂蓋層及一電極以相同 次序自IL之兩側延伸且視情況係相同材料(例如,在IL上 面及下面之相同RSL材料,例如MeOx,後跟在IL上面及下 面之相同頂蓋層材料,例如TiOx,後跟在IL上面及下面之 相同電極材料,例如TiN)。 圖10C繪示呈一不對稱、直立堆疊組態之圖6C之RSME 之一實施例。自頂部至底部之層次序係:E1 (例如, TiN)、capl(例如,TiOx)、RSL1(例如,MeOx)、IL(例 如,n+ Si)、IL(例如,TiN)、cap2(例如,TiOx)、RSL2(例 156917.doc -47- 201212319 如,MeOx)、E2(例如,n+ Si)。在一種方法中,IL係一 TiN層上面之一 n+ Si層(例如,10至100 nm厚度)之一組 合。在MeOx層上面且此鄰於MeOx層提供例如TiOx之頂蓋 層。舉例而言,capl位於RSL1上面且®比鄰於RSL1,且 cap2位於RSL2上面且毗鄰於RSL2。該組態係不對稱的且 呈一直立堆疊,其中所有層係垂直配置。不使用一鏡像組 態。該組態係不對稱的,乃因在IL(n+ Si)上面延伸之層包 含RSL1後跟capl,而在IL(TiN)下面延伸之層包含cap2後 跟RSL2。該組態係直立的,乃因capl位於RSL1上面且 cap2位於RSL2上面。 圖10D繪示呈一不對稱、倒置堆疊組態之圖6A之RSME 之一實施例。自頂部至底部之層次序係:El(例如, TiN)、E1(例如,n+ Si)、RSL1(例如,MeOx)、capl(例 如,TiOx)、IL(例如,TiN)、IL(例如,n+ Si)、RSL2(#J 如,MeOx)、cap2(例如,TiOx)、E2(例如,TiN)。不使用 一鏡像組態。該組態係不對稱的,乃因在IL上面,一頂蓋 後跟一 RSL,而在IL下面,一 RSL後跟一頂蓋。該組態相 對於圖10C之實施例倒置,乃因n+ Si層現在為E1層而非E2 層,且TiN層現在為E2層而非下部E1層。舉例而言,以與 圖10C之方式相反之一方式,IL層可係具有一 10至100 nm 厚度之n+Si與TiN之一組合。 IL之其他實施例使用一或多種金屬,例如選自由以下構 成之群組之一金屬:TiAIN、WN、W、NiSi、CoSi及C。 圖11A繪示展示當E2為n+ Si時SiOx之生長之圖6C之 156917.doc -48- 201212319 RSME之一實施例。自頂部至底部之層次序係:El(例如, n+ Si)、RSL1(例如,MeOx)、cap 1 (例如,TiOx)、IL(例 如,TiN)、cap2(例如,TiOx)、RSL2(例如,MeOx)、 SiOx、E2(例如,n+ Si)。當E2係由Si製成且RSL2包括一 金屬氧化物時,由於形成於RSL2與E2之間的一 SiOx層之 厚度變化,因此RSL中之形成電壓存在一大變化。舉例而 言,當RSL2係金屬氧化物且係直接沈積於含有n+ Si之一 E2上且與該E2接觸時,n+ Si層之一頂部部分被氧化,從 而產生一 SiOx層。在一實例性實施方案中,一 1至2 nm之 SiOx層可形成於RSL2與E2之間,其中該等RSL各自係由2 到4 nm之一 MeOx(例如,Hf02)製成且E2係由n+ Si製成。 另一選擇為,El及/或E2可係由p+ Si、氮化鎢(例如, WN、WN2、N2W3)、TiN或 SiGe製成。 圖11B繪示展示當E2為TiN時例如TiOx之一低帶隙材料 之生長之圖6C之RSME之一實施例。自頂部至底部之層次 序係:El(例如,n+ Si)、RSL1(例如,MeOx)、capl(例 ^ ,TiOx)、1[({列如,TiN)、cap2(#J 如,TiOx)、RSL2(你J 如,MeOx)、Ti/TiOx、E2(例如,TiN)。為防止 SiOx形 成,可用沈積於一 TiN電極上之例如Ti之一材料替換E2之 n+Si層。可認為Ti層係該電極之部分。具體而言,在於Ti 層上面沈積例如HfOx之一 MeOx層(RSL2)期間,該Ti層之 一頂部部分(約1至5 nm)被氧化且轉換成一 TiOx層。該 TiOx層之厚度取決於MeOx沈積之温度。在此情況下,第 二電極(E2)包括位於一 TiN層上之一 Ti層,該第二電阻切 156917.doc -49- 201212319 換層(RSL2)包括MeOx,且一 TiOx層係形成於該Ti層上並 與該第二電阻切換層接觸。"Single-crystalline, epitaxial cubic SiC films grown on (100) Si at 750 °C by chemical vapor deposition" by Golecki (April 1992, Applied Physics Letter, Vol. 60 'No. 14 '1703 to 1705) In this method, the SiC film is formed by using decyl decane (SiCH3H3) (a single precursor having a ratio of 1:1 si:c) and H2 by low pressure chemical vapor deposition. In another exemplary method, molecular beam epitaxy has been used to deposit SiC at low temperatures, as described in reference herein by a Fissel et al., rLow-temperaturegrowthofSiCthinfilmson Siand6H-SiC by solid-source molecular beam epitaxy. (Applied Physics Letter, Vol. 66, No. 23, pp. 3182 to 3184, June 1995). The method involves the use of solid-state molecular beam epitaxy controlled by a quadrupole mass spectrometer-based flux meter at Si (lll) and 6° to 5° skew oriented 6H-SiC at a low temperature of about 800 to 1000 °C. (0001) Epitaxial growth of stoichiometric SiC on the substrate. In the case of SiC (0001), a film was obtained on the surface of the Si-stabilized surface exhibiting (3x3) and (2x2) superstructures. The reflection high energy diffraction (rheED) pattern and the resistive RHEED oscillation during growth at 6H-SiC (0001) at T > 900 °C indicate the two-dimensional nucleation system main growth procedure on the terrace. Another exemplary cryogenic process for depositing SiC is described in the reference to W. Yu et al., "Low temperature 156917.doc •45-201212319 deposition of hydrogenated nanocrystalline SiC films by helicon wave Plasma enhanced chemical vapor deposition" (September 3, 2010, J. Vac. Sci. Technol. A 28(5), American Vacuum Society, pages 1234 to 1239). Here, a hydrogenated nanocrystalline lanthanum carbide (nc-SiC:H) film is deposited at a low substrate temperature by using a spiral wave plasma enhanced chemical vapor deposition technique. The effects of radio frequency (rf) power and substrate temperature on the properties of the deposited nc-SiC:H film have been investigated. It has been found that a hydrogenated amorphous SiC film is fabricated at a low rf power, while an nc-SiC:H film having a microstructure of one of the SiC nanocrystals embedded in the amorphous equivalent portion can be used as the rf power. Deposited at 400 W or more. The plasma transition from self-capacitative dominant discharge to spiral wave discharge with high plasma strength affects the microstructure and surface morphology of the film. Analysis of the films deposited at various substrate temperatures revealed that the onset of SiC crystallization occurred at substrate temperatures as low as 150 °C. Figure 10A illustrates one embodiment of the RSME of Figure 6C illustrating an alternative IL material. The order of layers from top to bottom is: El (eg, TiN), E1 (eg, n+Si), RSL1 (eg, MeOx), cap 1 (eg, TiOx), IL (eg, TiN), cap2 (eg, TiOx), RSL2 (eg, MeOx), E2 (eg, n+Si). In one embodiment, El comprises a combination of one of the TiN layers on an n+ Si layer. Additionally, a cap layer, such as TiOx, is provided between RSL1 and IL and between IL and RSL2. A further Ti contact (not shown) may be located above E1. As an alternative, IL may be selected from the group consisting of Al, Zr, La, Y, Ti, TiAIN, TixNy and TiAl alloys. These materials are preferred coupling layers that enable lower V and I unit operation. This embodiment provides a mirror image structure for IL. 156917.doc -46-201212319 A cap layer, an RSL, and an electrode extend from both sides of the IL in the same order and are optionally the same material (eg, The same capping material above and below the IL, such as TiOx, followed by the same RSL material above and below the IL, such as MeOx, followed by the same electrode material above and below the IL, such as n+Si). Figure 10B illustrates an embodiment of the RSME of Figure 6C in an inverted mirror stack configuration. The order of layers from top to bottom is: El (eg, TiN), capl (eg, TiOx), RSL1 (eg, MeOx), IL (eg, 'n+Si), RSL2 (eg, MeOx), cap2 (eg, TiOx) ), E2 (for example, TiN). In one method, El is made of TiN, IL is made of n+Si, and E2 is made of TiN. For example, the IL layer can have n+Si having a thickness of 10 to 100 nm. This embodiment is an inverted mirror configuration that provides an inverted stack with respect to the embodiment of Figure 10A, since the n+ Si layer is now IL rather than the El or E2 layer, and the cap layer is located at the RSL and electrodes. Between the layers (capl is between RSL1 and E1; cap2 is between RSL2 and E2) instead of between RSL and IL. Specifically, an RSL, a cap layer, and an electrode extend from both sides of the IL in the same order and are optionally the same material (eg, the same RSL material above and below the IL, such as MeOx, followed by IL) And the same top cover layer material below, such as TiOx, followed by the same electrode material above and below the IL, such as TiN). Figure 10C illustrates one embodiment of the RSME of Figure 6C in an asymmetric, upright stack configuration. The order from top to bottom is: E1 (eg, TiN), capl (eg, TiOx), RSL1 (eg, MeOx), IL (eg, n+Si), IL (eg, TiN), cap2 (eg, TiOx) ), RSL2 (eg, 156917.doc -47 - 201212319 eg MeOx), E2 (eg, n+ Si). In one method, one of the n+ Si layers (e.g., 10 to 100 nm thickness) of the IL-based TiN layer is combined. A top cover layer such as TiOx is provided on top of the MeOx layer and adjacent to the MeOx layer. For example, capl is above RSL1 and ® is adjacent to RSL1, and cap2 is above RSL2 and adjacent to RSL2. The configuration is asymmetrical and is an upright stack with all layers vertically configured. Do not use a mirrored configuration. This configuration is asymmetrical because the layer extending over IL(n+Si) contains RSL1 followed by capl, while the layer extending below IL(TiN) contains cap2 followed by RSL2. The configuration is upright because capl is above RSL1 and cap2 is above RSL2. Figure 10D illustrates an embodiment of the RSME of Figure 6A in an asymmetric, inverted stacked configuration. The order of layers from top to bottom is: El (eg, TiN), E1 (eg, n+Si), RSL1 (eg, MeOx), capl (eg, TiOx), IL (eg, TiN), IL (eg, n+ Si), RSL2 (#J, eg, MeOx), cap2 (eg, TiOx), E2 (eg, TiN). Do not use a mirrored configuration. The configuration is asymmetrical because of the top of the IL, a top cover followed by an RSL, and below the IL, an RSL followed by a top cover. This configuration is inverted relative to the embodiment of Figure 10C because the n+ Si layer is now the E1 layer instead of the E2 layer, and the TiN layer is now the E2 layer rather than the lower E1 layer. For example, in one of the ways opposite to that of Figure 10C, the IL layer can have a combination of n+Si and TiN having a thickness of 10 to 100 nm. Other embodiments of IL use one or more metals, such as one selected from the group consisting of TiAIN, WN, W, NiSi, CoSi, and C. Figure 11A depicts an embodiment of 156917.doc-48-201212319 RSME of Figure 6C showing the growth of SiOx when E2 is n+Si. The order of layers from top to bottom is: El (eg, n+Si), RSL1 (eg, MeOx), cap 1 (eg, TiOx), IL (eg, TiN), cap2 (eg, TiOx), RSL2 (eg, MeOx), SiOx, E2 (for example, n+ Si). When E2 is made of Si and RSL2 includes a metal oxide, there is a large change in the formation voltage in the RSL due to the thickness variation of an SiOx layer formed between RSL2 and E2. For example, when RSL2 is a metal oxide and is deposited directly on one of n+Si and is in contact with E2, one of the top portions of the n+ Si layer is oxidized, thereby producing an SiOx layer. In an exemplary embodiment, a 1 to 2 nm SiOx layer may be formed between RSL2 and E2, wherein the RSLs are each made of one of 2 to 4 nm MeOx (eg, Hf02) and the E2 system is composed of Made of n+ Si. Alternatively, El and/or E2 may be made of p+Si, tungsten nitride (e.g., WN, WN2, N2W3), TiN or SiGe. Figure 11B depicts an embodiment of the RSME of Figure 6C showing the growth of a low band gap material such as TiOx when E2 is TiN. The order from the top to the bottom is: El (for example, n+ Si), RSL1 (for example, MeOx), capl (example ^, TiOx), 1 [({column, TiN), cap2 (#J, TiOx) , RSL2 (you J, eg MeOx), Ti/TiOx, E2 (for example, TiN). To prevent SiOx formation, the n+Si layer of E2 may be replaced by a material such as Ti deposited on a TiN electrode. The Ti layer can be considered to be part of the electrode. Specifically, during deposition of, for example, one of HfOx MeOx layers (RSL2) over the Ti layer, a top portion (about 1 to 5 nm) of the Ti layer is oxidized and converted into a TiOx layer. The thickness of the TiOx layer depends on the temperature at which the MeOx is deposited. In this case, the second electrode (E2) includes a Ti layer on a TiN layer, the second resistor cut 156917.doc -49 - 201212319 The change layer (RSL2) includes MeOx, and a TiOx layer is formed thereon. The Ti layer is in contact with the second resistance switching layer.
Ti/TiOx之帶隙比SiOx之帶隙低得多,因此可避免形成 電壓之大變化。E1可係n+ Si或一高功函數材料,例如Ni 或NiSi。在一實例性實施方案中,該等RSL各自由2至4 nm 之一 MeOx(例如,Hf02)製成。 此外,可將一高功函數材料用於E1以減小操作電流。舉 例而言,可使用具有5.1 eV之一功函數之Ni。NiSi係另一 替代材料。相比較而言,一 TiN之功函數為約4.2至4.7 eV,且n+ Si之功函數為約4.1至4.3 eV。 圖11C繪示其中該等RSL係由一經摻雜金屬氧化物製成 以減小操作電壓之圖6C之RSME之一實施例。自頂部至底 部之層次序係:E1(例如,n+ Si)、RSL1(例如,經摻雜 MeOx)、capl(例如,TiOx)、IL(例如,TiN)、cap2(例如, TiOx)、RSL2(例如,經摻雜 MeOx)、SiOx、E2(例如,n+ Si)。舉例而言,可使用一經重摻雜MeOx層,例如HfOx或 HfSiON。MeOx之摻雜可係藉由以約0.01%至5%之一濃度 將例如Ti、A1或Zr之一摻雜劑植入或擴散至該MeOx層中 達成。測試結果指示此等摻雜劑提供良好性質。舉例而 言,可使用離子植入或原位原子層沈積(ALD)。在一實例 性實施方案中,該等RSL各自由2至4 nm之一MeOx(例如, Hf02)製成,且一 1至2 nm之SiOx層係形成於E2上,該E2 係 n+ Si 0 圖1 ID繪示其中E2係TiN代替n+ Si之圖11C之RSME之一 156917.doc -50- 201212319 實施例。自頂部至底部之層次序係:El(例如,n+ Si)、 RSL1(例如,經摻雜 MeOx)、capl(例如,TiOx)、IL(例 如,TiN)、cap2(例如,TiOx)、RSL2(例如,經摻雜 MeOx)、Ti/TiOx、E2(例如,TiN)。在一實例性實施方案 中,該等RSL各自由2至4 nm之一 MeOx(例如,Hf02)製成 且一 Ti/TiOx層係形成於E2上。 圖11E繪示呈一不對稱鏡像單元組態之圖6C之RSME之 一實施例,其中該等RSL由不同材料製成。自頂部至底部 之層次序係:El(例如,n+ Si)、RSL1(例如,類型A MeOx)、capl(例如,TiOx)、IL(例如,TiN)、cap2(例如, TiOx)、RSL2(例如,類型 B之MeOx)、SiOx、E2(例如,n+ Si)。使RSME沿正方向及負方向兩者切換可係成問題的, 因此,以某一極性切換可係較佳的。一個可行性解決方案 係針對RSL1及RSL2使用不同材料。舉例而言,RSL1可係 類型「A」,而RSL2係類型「B」。舉例而言,可使用兩種 不同類型之MeOx來控制切換極性,以使得RSL1係類型 「A」之MeOx,而RSL2係類型「B」之MeOx。MeOx之實 例包含 AlOx、TiOx、NiOx、ZrOx、CuOx、WOx,以使得 RSL 1可使用此等材料中之一者且RSL2可使用此等材料中 之另一者。該等RSL材料可經選擇以獲得一所期望切換效 能,其中切換發生在例如所規定I-V條件之期望條件下。 舉例而言,E1及E2可由n+ Si或TiN製成。 圖11F繪示呈不含SiOx之一不對稱鏡像單元組態之圖6C 之RSME之一實施例《自頂部至底部之層次序係:e 1 (例 156917.doc •51· 201212319 如 ’ n+ Si)、RSL1(例如,類型 a之 MeOx)、capl(例如, TiOx)、IL(例如,TiN)、cap2(例如,TiOx)、RSL2(例如, 類型B之MeOx)、Ti/TiOx、E2(例如’ TiN)。在此情況下, 第二電極(E2)係例如TiN之一材料代替n+ Si,以使得在製 造期間不形成Si02層。如結合圖所論述,Ti係沈積於 一 TiN電極上,且在於Ti上面沈積例如HfOX之一 Me〇x層期 間’該Ti層之一頂部部分被氧化,從而產生一τί〇χ層。 圖12缯·示圖6C之RSME之一能量圖。水平_繪示沿 RSME自E1至E2之一距離’而垂直軸表示一能階。Ec係導 帶’其位於自E1與RSL1之間的接面處之一高位準ec2至E2 與RSL2之間的接面處之一低值ec 1之範圍。ee 1係E1之能 階,EIL係IL之能量且EE2係E2之能階^ Εν係價帶。導帶 中之一凹口表示在IL處實現之一較低能階,如下文所述。 一 MRS依靠離子導電作為一切換機制。在離子導體中, 電流係藉由離子四處移動以及藉由電子及電洞之移動來運 送。舉例而s ’經由離子或離子及電子/電洞之電流運送 存在於稱為電解液之導電液及亦稱為固態電解液之離子導 電固體中。此外,離子導電率對於諸多產品(例如j型及π 型電池(亦即’常規型及充電型),燃料電池、電致變色窗 及顯示器、固態感測器),尤其對於反應氣體、導電橋接 切換及本文令所述之雙極MeOx切換係重要的。。 與單純電子電流運送相比,存在與電流相聯繫之一化學 反應(例如,該系統隨時間改變),該化學反應發生在將離 子電流轉化成一電子電流之任何地方,亦即觸點或電極 156917.doc -52· 201212319 處。與藉助電子(或電洞)之電流相比,此係一顯著差別, 在藉助電子(或電洞)之電流中,針對跨越觸點之電流不需 要化學反應。雙極MeOx切換嘗試移動Me〇x中之氧空位以 形成一金屬細絲,藉此將氧儲存於介面處。可藉由包含以 下各項之機制來提供電子導電:傅勒_諾德翰、蕭特基、 空間電荷限制電流(SCLC)、SCLC與普爾.夫倫克爾(pF)一 起、PF及希爾定律。離子導電包含導電率、擴散及場類 型。 典型的離子導電率值係相對低且取決於一電極之自空氣 之氧供應、溫度及電場(以指數方式)。 圖13繪示在一RSL之一設定程序中一高電場之施加。此 掃描電子顯微鏡影像繪示包含Si02之一生長層之n+ si之 一左手邊電極(EL)、Hf02之一 RSL及TiN之一右手邊電極 (ER)。可施加一尚電場以將氧移動至例如Hf〇2之MeOx之 一 RSL中。此處,在一實例性實施方案中,一高電場存在 於3至5 nm寬之一Hf02區中。在使用5 nm之值時,該電場 因此係 5 V / 5 nm = 10 MV/cm。 圖14A至圖14D繪示在一 RSL之一設定程序中形成一導電 細絲中之不同階段。圖中繪示一單個MeOX膜之通常發生 之崩潰。舉例而言,將左手邊電極(EL)設定在〇 V處作為 一接地電極’中間區表示例如Hf02之一 RSL,且右手邊區 表示5 V下之一受驅動右手邊電極(ER)。5 V係不存在電流 限制器(電阻器)之情況之一近似值。此等圖指示具有此等 RSL中之兩者或兩者以上之一 RSME之預期行為。應記 156917.doc •53- 201212319 付’在一 RSME中’右手邊電極將接收一耗合電壓且不直 接被驅動。 在一設定或形成程序中,該RSL最初係不導電。一開放 或白圓圈表示一氧離子,且一封閉或黑圓圈表示金屬。高 電場搞合至帶負電之氧離子,從而自励2抽取氧離子且將 其吸引至ER。在圖14A之狀況之後,存在圖14B之狀況, 其中已抽取某些氧離子並儲存於ER處(如ER處之開放圓圈 所表示)’且自其抽取氧iHf〇2區變為具金屬性(如由封閉 圓圈所表示)。此程序繼續,以使得在圖14B之狀況之後, 達到圖14C之狀況,其中已抽取額外氧離子並將其儲存於 中間電極處,且自其抽取氧之額外]^^^區變為具金屬性。 最後,在圖14C之狀況之後,達到圖14D之狀況,其中已 抽取額外氧離子並將其儲存於該£11處,且自其抽取氧之 Hf02之一充分部分變為具金屬性,從而形成穿過該rsl之 一導電細絲或路徑作為該等電極之間的一短路。 因此,存在自一關斷狀態(其中該RSL處於一相對高電阻 狀態,類似於一開(不導電)路)至一導通狀態(其中該RSL處 於一相對低電阻狀態,類似於一短路(導電)或閉路)之一轉 變。 圖14E、圖14F及圖14G係分別闡述圖14A、圖14B及圖 14D之设定程序階段之能量圖。y軸繪示能量且乂軸繪示在 RSME中之距離。峰值表示由Hf〇2中之氧所強加之對電子 運送之障壁。該等峰值遵循導帶Ec,其介於自以丨至jgC2之 範圍。在圖14E至圖14G中,該導帶維持此固定範圍。EEl 156917.doc •54- 201212319 係EL之能量’且EER係Er之能量。此外,圖中繪示線性帶 彎曲之一理想情況。 在該程序之開始時’假設跨越EL及ER施加5 V且EL與 ER分離5 nm,則電場(E)處於1〇 MV/cm(5 V/5 nm)之一開 始位準。相對少量電流流動,如細的虛箭頭所表示(圖 14E)。隨著該程序繼續,自RSL抽取氧且用係一生長細絲 之部分之一金屬性區替換氧。該金屬性區基本上變為“電 極之一延伸部’以使得EL與ER之間的有效距離降低,例 如自5 nm至4 nm,且E場對應地增加至12 MV/cm(5 V/4 nm)。由於該較高場,因此較大量之電流流動,如由較粗 虛箭頭所表示(圖14F)。隨後,自Hf02抽取額外氧,以使 得該細絲生長且EL與ER之間的有效距離降低,例如自4 rim至1 nm,且由於該場與距離之間的指數關係,e場增加 至50 MV/cm(5 V/1 nm)。由於甚至更高之場,一甚至更大 量之電流流動成為一衝擊電流,如由甚至更粗之虛箭頭所 表示(圖14G)。 注意,在圖14E至圖14G中,第一能量峰值與最後一個 能量峰值之高度大約係相同,但存在較少峰值指示對電子 運送之一較低障壁。所提出之RSME可因此藉由IL層之電 流限制效應有利地避免一形成及設定程序令之一衝擊電 流。 圖15A至圖15C繪示在一 RSL之一重設程序中移除一導電 細絲之不同階段。 圖15D、圖15E及圖15F係分別闡述圖15A、圖15B及圖 156917.doc •55- 201212319 15C之重設程序階段之能量圖。左手邊區表示一接地電極 (EL),中間區表示例如Hf02之一 RSL且右手邊區表示一受 驅動電極(ER)。所繪示之電壓及電子係在不存在IL之電流 限制效應之情況之一近似值。此等圖指示一 RSL之預期行 為。應記得,一 RSME由至少兩個串聯RSL構成,且在一 RSME中,右手邊電極將接收一耦合電壓而不被直接驅動 且因此有效減小電流。 該重設程序實質上與圖14A至圖14D之設定程序相反。 在該重設程序之開始時(圖15A及圖15D),E場係50 MV/cm,且相對小數目個氧離子返回至Hf02之靠近ER之 一部分,從而阻斷由細絲形成之短路。舉例而言,使用與 設定程序相比係相反之一極性跨越ER及EL而施加-5 V之一 電壓。因此,在重設期間,舉例而言,可以-5 V開始。舉 例而言,跨越其施加E場之一有效距離係1 nm,從而產生 5 0 MV/cm之一 E場。隨後,跨越1.3 nm之一距離施加-7 V 之一電壓,從而產生53 MV/cm之一 E場(圖15B及圖15E)。 隨後,跨越1.6 nm之一距離施加-9 V之一電壓,從而產生 56 MV/cm之一 E場(圖15C及圖15F)。此程序在一 RSME中 係完全不同,可因此有利地避免一重設程序中之衝擊電 流。 在一雙極MeOx切換之情形下,提供一離子移動,其中 自RSL移除離子以使得RSL變為更具金屬性。此係一自放 大效應,乃因一移除一個離子,則由於場增加其他離子之 移除就加速,且移動對場之相依性係成指數的。因此,若 156917.doc •56- 201212319 已移除一個離子,則場已增加且離子移動之遷移率以指數 方式增加。因此,裝置具有一較快突崩效應。此闡釋設定 及形成相依性。 除離子移動外,電子可藉由象徵性地跳過能量峰值而同 時在S亥RSL中移動。最初,僅少量電子流動。但電場一增 加,更多電子就可溢出能量峰值且其流動起來更容易。最 後,大量電子朝向IL衝擊性地流動。然而,此電子流動係 不期望的’乃因該等電子無益於依靠個別離子之移動之切 換機制。為使該等離子移動,需要建立一充分電場。相關 聯電子流動係不期望的,乃因若具有與RSL串聯之一操縱 元件(例如二極體),則該二極體需要能夠不僅使來自小離 子電流之電流而且使來自較大電子電流之電流持續。 此外’在重設期間,氧移回至電阻切換元件,且因此, 1L與E1或E2之間的有效距離再次增加。產生允許大量電子 流動之一電場。 該RSME結構允許足以使離子移動些許而不使電子流動 太多之一電場得以建立。該RSME基本上提供不傳導極多 電子之一差勁導體。此外,IL提供使電子停止且反射電子 之一障壁。與電容性耦合效應一起,可因此使離子移動而 不使太多電子電流流動。 該RSME可係大體對稱的(在RSLi與RSL2之間具有一 IL) ’因此可在IL處(在該等RSL之間)關注該切換機制。該 IL允許在該裝置之中心建立電場,以使得離子將在 移動但不會跨越中間區中之IL。該IL係一導體,且能夠儲 156917.doc •57· 201212319 存氧離子。該IL可係具金屬性的,但其亦可不具金屬性β 該IL可係極薄’且應能夠反射及/或保持電子以使得該等 電子落在IL處。該IL之電容可藉由使其厚度變化來調整。 此對於按比例縮小之裝置而言可係尤其重要。 一目標係提供具有例如圖12中所繪示之一能量圖且包含 在彼處電子被反射但仍存在一所建立之電場之一電位臺階 之一 RSME »可使用一對稱構造,其中尺此丨及尺儿之具有相 同厚度’或RSL1及RSL2亦可具有不同厚度。一個RSL可 稍厚於另一 RSL,以使得可建立一場而不誘發切換。此將 導致基於RSL1及RSL2之厚度移位如圖η中所示之帶隙 圖。若該等RSL之厚度係相同,則其場將表現相同,且將 藉助相同電場切換。另一方面,藉由引入一不對稱性,可 僅對一個RSL調變,在此情況下,另一 RSL變為一阻擋層 而不進行切換。 關於衝擊電流,發生此現象係由於IL與E1或E2之間的距 離係如此短以至於沒有機會與體積相互作用。在一電導體 中 電子在電场中加速且在一平均自由路徑中行進直 至其被-電子與電子、電子與光子、電子與雜質或電子與 介面機制散射為止。對於例如矽或銅之典型導體,一典型 散射平均自由路徑係約4〇 nm。在―按比例縮放之記憶體 裝置中’電流係衝擊性的,乃因典型之尺寸係小得多,以 使得電子過冲且在電極内部深處被散射,並且不將能量遞 送至切換區。 圖16A繪示圖6八之尺咖之一設定程序。在步驟16〇〇 156917.doc •58· 201212319 處,針對-記憶體單元開始該設定程序。實務中,可針對 -記憶體裝置中之多個記憶體單元藉由將適當電壓施加至 適當位元線及字線來同時勃并__ 不丨』吁執仃一设定或重設程序。在步驟 1602處,跨越第一電極及第二電極施加一設定電壓。經由 與電阻切換記憶體單元串聯之—操縱元件跨越該電阻切換 記憶體單it之第-電極及第二電極而施加該電壓。 設定電壓可具有一所需波形,舉例而言,例如一或若干 固定振幅之脈衝、斜升式或樓梯式脈衝。因此,該電壓可 係一隨時間變化之電壓信號,例如量值隨時間增加。·對 於一固定振幅之脈衝,舉例而言,該振幅可係在例如 Vset(圖4A)之一位準處或高於該位準。對於一斜升式或樓 梯式脈衝’該設定電壓可在低於Vset之一位準處開始且增 加至Vset或更尚。在一種方法中,在一規定時間週期中, 盲目地施加設定電壓,而不判定實際上是否達成設定狀 態。在此情況下,該設定電壓具有基於該記憶體裝置之一 先前統計分析足以使接近100%的所有記憶體單元達成設 定狀態之一持續時間及/或量值。 在另一方法中’當施加設定電壓時監視該記憶體單元之 狀態’且當該監視指示已達到設定狀態時移除該設定電 壓。移除一電壓可意指允許第一電極及第二電極浮動。此 方法進一步闡述於(例如)2010年4月8日公佈、標題為「SetThe band gap of Ti/TiOx is much lower than that of SiOx, so that large variations in voltage formation can be avoided. E1 may be n+ Si or a high work function material such as Ni or NiSi. In an exemplary embodiment, the RSLs are each made of one of 2 to 4 nm MeOx (eg, HfO 2 ). In addition, a high work function material can be used for E1 to reduce the operating current. For example, Ni having a work function of 5.1 eV can be used. NiSi is another alternative material. In comparison, a TiN has a work function of about 4.2 to 4.7 eV, and a work function of n+ Si is about 4.1 to 4.3 eV. Figure 11C illustrates one embodiment of the RSME of Figure 6C in which the RSLs are made of a doped metal oxide to reduce the operating voltage. The order of layers from top to bottom is: E1 (eg, n+ Si), RSL1 (eg, doped MeOx), capl (eg, TiOx), IL (eg, TiN), cap2 (eg, TiOx), RSL2 ( For example, it is doped with MeOx), SiOx, E2 (for example, n+Si). For example, a heavily doped MeOx layer such as HfOx or HfSiON can be used. Doping of MeOx can be achieved by implanting or diffusing a dopant such as Ti, Al or Zr into the MeOx layer at a concentration of about 0.01% to 5%. The test results indicate that these dopants provide good properties. For example, ion implantation or in situ atomic layer deposition (ALD) can be used. In an exemplary embodiment, the RSLs are each made of one of 2 to 4 nm MeOx (eg, HfO 2 ), and a 1 to 2 nm SiOx layer is formed on E2, which is an n+ Si 0 map. 1 ID depicts one of the RSMEs of Figure 11C in which E2 is a TiN instead of n+Si. 156917.doc -50 - 201212319 Embodiment. The order of layers from top to bottom is: El (eg, n+Si), RSL1 (eg, doped MeOx), capl (eg, TiOx), IL (eg, TiN), cap2 (eg, TiOx), RSL2 ( For example, doped with MeOx), Ti/TiOx, E2 (eg, TiN). In an exemplary embodiment, the RSLs are each made of one of 2 to 4 nm MeOx (e.g., HfO 2 ) and a Ti/TiOx layer is formed on E2. Figure 11E illustrates an embodiment of the RSME of Figure 6C in an asymmetric mirrored unit configuration, wherein the RSLs are made of different materials. The order of layers from top to bottom is: El (eg, n+Si), RSL1 (eg, type A MeOx), capl (eg, TiOx), IL (eg, TiN), cap2 (eg, TiOx), RSL2 (eg, , Type B of MeOx), SiOx, E2 (eg, n+ Si). Switching the RSME in both the positive and negative directions can be problematic, so switching at a certain polarity may be preferred. A viable solution uses different materials for RSL1 and RSL2. For example, RSL1 can be of type "A" and RSL2 of type "B". For example, two different types of MeOx can be used to control the switching polarity such that RSL1 is a MeOx of type "A" and RSL2 is a MeOx of type "B". Examples of MeOx include AlOx, TiOx, NiOx, ZrOx, CuOx, WOx such that RSL 1 can use one of these materials and RSL 2 can use the other of these materials. The RSL materials can be selected to achieve a desired switching effect, wherein the switching occurs under desired conditions such as the specified I-V conditions. For example, E1 and E2 can be made of n+ Si or TiN. Figure 11F illustrates one embodiment of the RSME of Figure 6C in an asymmetric mirror unit configuration that is free of SiOx. The layer order from top to bottom: e 1 (eg 156917.doc • 51· 201212319 eg 'n+ Si ), RSL1 (eg, MeOx of type a), capl (eg, TiOx), IL (eg, TiN), cap2 (eg, TiOx), RSL2 (eg, MeOx of type B), Ti/TiOx, E2 (eg, 'TiN). In this case, the second electrode (E2) is a material such as TiN instead of n + Si so that the SiO 2 layer is not formed during the fabrication. As discussed in connection with the figures, Ti is deposited on a TiN electrode and a top portion of the Ti layer is oxidized by depositing, for example, one of the HfOX Me〇x layers on the Ti, thereby producing a τί〇χ layer. Figure 12 is an energy diagram of one of the RSMEs of Figure 6C. Horizontal_shows a distance ' along RSME from E1 to E2' and the vertical axis represents an energy level. The Ec-based conduction band 'is located at a low level ec 1 from one of the junctions between the high level ec2 to E2 and the RSL2 at the junction between E1 and RSL1. Ee 1 is the energy level of E1, EIL is the energy of IL and EE2 is the energy level of E2. One of the notches in the conduction band indicates that one of the lower energy levels is achieved at the IL, as described below. An MRS relies on ion conduction as a switching mechanism. In ionic conductors, current is transported by the movement of ions and by the movement of electrons and holes. For example, s 'current transport via ions or ions and electrons/holes is present in an electrically conductive liquid called an electrolyte and an ionically conductive solid, also known as a solid electrolyte. In addition, ionic conductivity is available for many products (eg, j-type and π-type batteries (ie, 'regular and rechargeable), fuel cells, electrochromic windows and displays, solid state sensors), especially for reactive gases, conductive bridging Switching and bipolar MeOx switching as described in this document are important. . Compared to pure electron current transport, there is a chemical reaction associated with the current (eg, the system changes over time) that occurs anywhere where the ion current is converted into an electron current, ie, the contact or electrode 156917 .doc -52· 201212319. This is a significant difference compared to the current by means of electrons (or holes), in which the current across the contacts does not require a chemical reaction in the current through the electrons (or holes). Bipolar MeOx switching attempts to move the oxygen vacancies in Me〇x to form a metal filament whereby oxygen is stored at the interface. Electron conduction can be provided by mechanisms including: Fuller_Nordham, Schottky, Space Charge Limit Current (SCLC), SCLC together with Poole Frenkel (pF), PF and Hill's law . Ion conduction includes conductivity, diffusion, and field type. Typical ionic conductivity values are relatively low and depend on the oxygen supply, temperature and electric field from the air (in an exponential manner). Figure 13 illustrates the application of a high electric field in one of the RSL setting procedures. The scanning electron microscope image shows a left-hand electrode (EL) of n+si comprising one growth layer of SiO2, one of Hf02 RSL and one right-hand electrode (ER) of TiN. An electric field can be applied to move the oxygen into an RSL such as MeOx of Hf 〇 2 . Here, in an exemplary embodiment, a high electric field is present in one of the 3 to 5 nm wide Hf02 regions. When using a value of 5 nm, the electric field is therefore 5 V / 5 nm = 10 MV/cm. Figures 14A through 14D illustrate different stages in forming a conductive filament in one of the RSL setting procedures. The figure shows the typical collapse of a single MeOX film. For example, the left-hand electrode (EL) is set at 〇 V as a ground electrode. The middle region represents, for example, one of the Hf02 RSLs, and the right-hand edge region represents one of the 5 V-driven right-hand electrodes (ER). The 5 V system has no approximation of the current limiter (resistor). These figures indicate the expected behavior of an RSME with one or more of these RSLs. It should be noted that 156917.doc •53- 201212319 pays 'in an RSME' the right-hand electrode will receive a current-consuming voltage and is not directly driven. The RSL is initially non-conductive during a setup or formation process. An open or white circle indicates an oxygen ion, and a closed or black circle indicates a metal. The high electric field is coupled to the negatively charged oxygen ions, so that the oxygen ions are extracted from the excitation 2 and attracted to the ER. After the condition of Figure 14A, there is the condition of Figure 14B, in which some oxygen ions have been extracted and stored at the ER (as indicated by the open circles at the ER)' and the oxygen iHf〇2 region is changed from metallic to metallic (as indicated by a closed circle). This procedure continues so that after the condition of Figure 14B, the condition of Figure 14C is reached, where additional oxygen ions have been extracted and stored at the intermediate electrode, and the additional ^^^^ region from which oxygen is extracted becomes metal Sex. Finally, after the condition of FIG. 14C, the condition of FIG. 14D is reached, in which extra oxygen ions have been extracted and stored at the £11, and one of the Hf02 from which the oxygen is extracted is sufficiently partially metallic to form A conductive filament or path through one of the rsl acts as a short between the electrodes. Therefore, there is a self-off state (where the RSL is in a relatively high resistance state, similar to an open (non-conducting) path) to a conducting state (wherein the RSL is in a relatively low resistance state, similar to a short circuit (conductive) ) or closed circuit) one of the changes. 14E, 14F, and 14G illustrate energy diagrams of the setup procedure stages of Figs. 14A, 14B, and 14D, respectively. The y-axis shows the energy and the x-axis shows the distance in the RSME. The peak indicates the barrier to electron transport imposed by the oxygen in Hf〇2. These peaks follow the conduction band Ec, which is in the range from 丨 to jgC2. In Figures 14E to 14G, the conduction band maintains this fixed range. EEl 156917.doc •54- 201212319 is the energy of EL' and EER is the energy of Er. In addition, the figure shows one ideal case of linear band bending. At the beginning of the procedure, assuming that 5 V is applied across the EL and ER and the EL is separated from the ER by 5 nm, the electric field (E) is at one of the starting levels of 1 〇 MV/cm (5 V/5 nm). A relatively small amount of current flows, as indicated by a thin dashed arrow (Fig. 14E). As the procedure continues, oxygen is drawn from the RSL and the oxygen is replaced with a metallic region that is part of a growing filament. The metallic region essentially becomes "one extension of the electrode" such that the effective distance between EL and ER is reduced, for example from 5 nm to 4 nm, and the E field is correspondingly increased to 12 MV/cm (5 V/ 4 nm). Due to this higher field, a larger amount of current flows, as indicated by the thicker dashed arrow (Fig. 14F). Subsequently, additional oxygen is extracted from Hf02 to allow the filament to grow and between EL and ER The effective distance is reduced, for example from 4 rim to 1 nm, and due to the exponential relationship between the field and the distance, the e-field is increased to 50 MV/cm (5 V/1 nm). Because of even higher fields, even A larger amount of current flows into an inrush current, as indicated by an even thicker virtual arrow (Fig. 14G). Note that in Figs. 14E to 14G, the first energy peak is approximately the same as the height of the last energy peak. However, there are fewer peaks indicating a lower barrier to one of the electron transports. The proposed RSME can thus advantageously avoid an inrush current in one of the forming and setting commands by the current limiting effect of the IL layer. Figure 15A to Figure 15C Removing different stages of a conductive filament in an RSL reset procedure 15D, 15E, and 15F are energy diagrams illustrating the reset procedure stages of FIGS. 15A, 15B, and 156917.doc • 55-201212319 15C, respectively. The left-hand side area represents a ground electrode (EL), and the middle area represents, for example, Hf02. One of the RSLs and the right-hand side region represents a driven electrode (ER). The voltage and electrons are plotted as an approximation of the current limiting effect of IL. These figures indicate the expected behavior of an RSL. It should be remembered that An RSME consists of at least two series RSLs, and in an RSME, the right hand electrode will receive a coupling voltage without being directly driven and thus effectively reducing the current. The reset procedure is substantially the same as the settings of Figures 14A-14D. The procedure is reversed. At the beginning of the reset procedure (Fig. 15A and Fig. 15D), the E field is 50 MV/cm, and a relatively small number of oxygen ions are returned to a portion of Hf02 near the ER, thereby blocking the formation of filaments. Short circuit. For example, one of the opposite polarity is applied across the ER and EL and a voltage of -5 V is applied compared to the setting procedure. Therefore, during the reset, for example, it can start at -5 V. For example, Say, apply E across it One of the effective distances of the field is 1 nm, resulting in an E field of 50 MV/cm. Subsequently, a voltage of -7 V is applied across one distance of 1.3 nm, resulting in an E field of 53 MV/cm (Fig. 15B). And Figure 15E). Subsequently, a voltage of -9 V is applied across one distance of 1.6 nm, resulting in an E field of 56 MV/cm (Figure 15C and Figure 15F). This procedure is completely different in an RSME. It is therefore advantageous to avoid inrush currents in a reset procedure. In the case of a bipolar MeOx switch, an ion shift is provided in which ions are removed from the RSL to make the RSL more metallic. This is a self-amplifying effect, because as soon as one ion is removed, the other ions are accelerated as the field is removed, and the dependence of the movement on the field is exponential. Therefore, if an ion has been removed from 156917.doc •56-201212319, the field has increased and the mobility of ion mobility has increased exponentially. Therefore, the device has a faster sag effect. This explains the settings and forms a dependency. In addition to ion movement, electrons can move in the S-HSL by symbolically skipping energy peaks. Initially, only a small amount of electrons flowed. But as the electric field increases, more electrons can overflow the energy peak and it flows more easily. Finally, a large amount of electrons flow impingingly toward the IL. However, this electronic flow is undesirable because the electrons are not beneficial to the switching mechanism that relies on the movement of individual ions. In order to move the plasma, a sufficient electric field needs to be established. The associated electron flow is undesirable because if one has a steering element (eg, a diode) in series with the RSL, the diode needs to be able to not only source current from a small ion current but also from a larger electron current. The current continues. Further, during resetting, oxygen moves back to the resistance switching element, and therefore, the effective distance between 1L and E1 or E2 increases again. An electric field is generated that allows a large amount of electron flow. The RSME structure allows an electric field to be established that is sufficient to move the ions a little while not allowing the electrons to flow too much. The RSME basically provides a poor conductor that does not conduct much electrons. In addition, the IL provides a barrier to stopping electrons and reflecting electrons. Together with the capacitive coupling effect, the ions can thus be moved without causing too much electron current to flow. The RSME can be generally symmetrical (having an IL between RSLi and RSL2) so the switching mechanism can be followed at the IL (between the RSLs). The IL allows an electric field to be established at the center of the device such that the ions will move but do not cross the IL in the intermediate region. The IL is a conductor and is capable of storing oxygen ions at 156917.doc •57·201212319. The IL may be metallic, but it may also be non-metallic. The IL may be extremely thin' and should be capable of reflecting and/or holding electrons such that the electrons fall at the IL. The capacitance of the IL can be adjusted by varying its thickness. This can be especially important for scaled down devices. A target system provides one of the potential steps, such as one of the energy diagrams depicted in FIG. 12, and includes one of the potential steps of the electric field that is reflected but still exists in one of the fields. RSME can be used, wherein a symmetrical configuration can be used. And the ruler has the same thickness' or RSL1 and RSL2 may also have different thicknesses. One RSL can be slightly thicker than the other RSL so that a field can be established without inducing a switch. This will result in a band gap map as shown in Figure η based on the thickness shift of RSL1 and RSL2. If the thicknesses of the RSLs are the same, their fields will behave the same and will switch with the same electric field. On the other hand, by introducing an asymmetry, only one RSL can be modulated, in which case the other RSL becomes a barrier without switching. With regard to the inrush current, this phenomenon occurs because the distance between IL and E1 or E2 is so short that there is no opportunity to interact with the volume. In an electrical conductor, electrons accelerate in an electric field and travel in an average free path until they are scattered by electrons and electrons, electrons and photons, electrons and impurities, or electrons and interface mechanisms. For a typical conductor such as germanium or copper, a typical scattering mean free path is about 4 〇 nm. In a "scaled memory device" the current is shock sensitive because the typical dimensions are much smaller so that the electrons are overshooted and scattered deep inside the electrode and no energy is delivered to the switching region. FIG. 16A is a diagram showing a setting procedure of the ruler of FIG. At step 16〇〇 156917.doc •58· 201212319, the setup procedure is started for the -memory unit. In practice, a plurality of memory cells in a memory device can be simultaneously set or reset by applying an appropriate voltage to the appropriate bit line and word line. At step 1602, a set voltage is applied across the first electrode and the second electrode. The voltage is applied across the resistance-switching memory cell in series with the first and second electrodes of the memory switching device. The set voltage can have a desired waveform, such as, for example, one or several pulses of fixed amplitude, ramp-up or stair-type pulses. Thus, the voltage can be a voltage signal that varies over time, e.g., the magnitude increases over time. • For a pulse of a fixed amplitude, for example, the amplitude can be at or above one of the levels of Vset (Fig. 4A). For a ramp or floor ladder pulse, the set voltage can start at one level below Vset and increase to Vset or more. In one method, the set voltage is blindly applied for a predetermined period of time without determining whether the set state is actually achieved. In this case, the set voltage has a duration and/or magnitude that is sufficient to cause nearly 100% of all memory cells to reach a set state based on one of the prioritized statistical analyses of the memory device. In another method, 'the state of the memory cell is monitored when a set voltage is applied' and the set voltage is removed when the monitor indicates that the set state has been reached. Removing a voltage may mean allowing the first electrode and the second electrode to float. This method is further described, for example, on April 8, 2010, entitled "Set
And Reset Detection Circuits For Reversible Resistance Switching Memory Material」之 US 2010/0085794 及 2008 年 6 月 24 曰發佈、標題為「Memory device for protecting 156917.doc -59- 201212319 memory cells during programming」之US 7,391,638 中,該 兩個專利皆以引用之方式併入本文中。 在步驟1604處,將電壓耦合至中間層(IL),且該IL散射 自RSL進入IL之電子。在步驟1606處,一或多個細絲形成 於RSL中。亦參見圖14A至圖14D。在不同RSL中,細絲之 形成可以不同速率進行且在不同時間完成。舉例而言,參 考圖4B,當設定電壓達到VsetB時,將首先針對類型 「B」之RSL達到設定狀態,且隨後當設定電壓達到VsetA 時針對類型「A」之RSL達到設定狀態。該設定電壓足以 在該等RSL中之每一者中形成一細絲以在該等RSL中提供 一導電路徑,藉此提供貫穿該RSME及記憶體單元之一導 電路徑。因此,在該等RSL中之每一者中且在RSME中達 成一低電阻狀態。可將RSME之低電阻狀態指派給一第一 個二進制資料狀態,例如,0或1。在步驟1608處,移除設 定電壓且該記憶體單元(包含該RSME)放電。注意,步驟 1602至1606至少部分地同時發生。 視情況,可僅該等RSL中之一者完成該設定程序,或少 於該RSME中之所有RSL之RSL完成該設定程序。 圖16B繪示針對圖6A之RSME之一重設程序。在步驟 1620處,針對一記憶體單元開始該重設程序。在步驟1622 處,跨越第一電極及第二電極施加一重設電壓(Vreset,參 見圖4A)。經由與電阻切換記憶體單元串聯之一操縱元件 跨越該電阻切換記憶體單元之第一電極及第二電極而施加 該電壓。該設定電壓可具有一所需波形,例如一固定振幅 156917.doc •60- 201212319 之脈衝或一斜升式脈衝◦因此,該電壓可係一隨時間變化 之電壓信號,例如量值隨時間增加。如前文,在一種方法 中,盲目地施加該設定電壓,而不判定實際上是否達成設 定狀態。在此情況下’重設電壓具有足以使接近!00。/〇的 所有記憶體單元達成重設狀態之一持續時間及/或量值。 在另一方法中,當施加重設電壓時監視記憶體單元之狀 態’且當該監視指示已達到重設狀態時移除該重設電壓。 此方法進一步闡述於上文所提及之US 2010/008 5794及US 7,391,638 中。 在步驟1624處,將電壓耦合至中間層,且IL散射自該等 RSL進入IL之電子。在步驟1626處,在該等RSL中移除或 毀壞一或多個細絲。亦參見圖15A至圖15C。在不同RSL 中’該等細絲之移除可以不同速率進行且在不同時間完 成。舉例而言’參考圖4B,當重設電壓達到VresetB時, 將首先針對類型「B」之RSL達到重設狀態,且隨後當重 設電壓達到VresetA時針對類型「A」之RSL達到重設狀 態。該重設電壓足以移除該等RSL中之每一者中之細絲以 移除該等RSL中之一導電路徑’藉此移除穿過該rSme及 該記憶體單元之一導電路徑。因此,在該等RSL中之每一 者中且在該RSME中達成一高電阻狀態。可將該rSme之高 電阻狀態指派給與低電阻資料狀態相反之一第二個二進制 資料狀態’例如1或0 ^在步驟1628處,移除重設電壓且該 記憶體單元(包含該RSME)放電。注意,步驟1622至1626 至少部分地同時發生。 156917.doc -61- 201212319 視情況’可僅該等RSL中之一者完成該重設程序,或少 於該RSME中之所有RSL之RSL完成該重設程序。 以上方法可包含:跨越電阻切換記憶體單元之第一電極 及第二電極施加一電壓以在該記憶體單元中設定一第一資 料狀態’其中該電壓係電容性耦合至一導電中間層,該導 電中間層以電方式位於該第一電極及該第二電極之間且與 該等電極串聯,且該電壓致使在以下各項中之至少—者中 切換一電阻狀態:(a)一第一電阻切換層,其以電方式位於 該第一電極與該導電中層之間且與該第一電極及該導電中 間層串聯,及(b)—第二電阻切換層,其以電方式位於該第 二電極與該導電中間層之間且與該第二電極及該導電中間 層串聯,及移除該電壓以允許該電阻切換記憶體單元放 電。該等電阻切換層可係可逆或不可逆的。 以上方法亦可包含藉由以下步驟改變一電阻切換記憶體 單兀中之一電阻狀態:(a)增加跨越該電阻切換記憶體單元 而施加之一隨時間變化之電壓之一量值直至在該電阻切換 記憶體單元之第一電阻切換層及第二電阻切換層中之一者 中切換一電阻狀態為止;及(b)隨後,進一步增加跨越該電 阻切換記憶體單元而施加之該隨時間變化之電壓之該量值 直至在該電阻切換記憶體單元之該第一電阻切換層及該第 一電阻切換層中之另一者中切換一電阻狀態為止。該電阻 狀態之該切換可係可逆的或不可逆的。 以上方法亦可包含:跨越第一控制線及第二控制線而施 加一電壓,其中該第一控制線係連接至一電阻切換記憶體 156917.doc •62· 201212319 單7C之一端,該第二控制線係連接至與該電阻切換記憶體 單元串聯之-操縱元件;及跨越該電阻切換記憶體單元之 第-電阻切換層及第二電阻切換層且跨越該第一電阻切換 層與該第二電阻切換層之間的一導電令間層而施加該電 堡;及移除該電壓以允許該電阻切換記憶體單元放電。該 專電阻切換層可係可逆的或不可逆的。 -因此’可看到,在一項實施例中,一電阻切換記憶體單 元包括··第-電極及第二電極;以電方式位於該第一電極 與該第二電極之間且與該等電極串聯之一導電中間層;以 電方式位於該第一電極與該導電中間層之間且與該第一電 極及該導電中間層㈣之-第—電阻切換層;及以電方式 位於該第二電極與該導電中間層之間且與該第二電極及: 導電中間層串聯之-第二電阻切換層,該第_電阻切換層 及該第二電阻切換層皆具有一雙極切換特性或皆具有一 極切換特性。 在另一實施例中,一電阻切換記憶體單元包括:一個一 極體操縱元件;及與該二極體操縱元件 甲唧之一電阻切換 記憶體元件,該電阻切換記憶體元件包 枯·第一電極及第 二電極;以電方式位於該第一電極與該第二電極之間且與 該等電極串聯之一導電或半導電中間層; 、 電方式位於該 負,一電極與該導電或半導電中間層之間且與誃 "^ 電 及 該導電或半導電中間層串聯之一第一電 , y 刀換層;及以電 力式位於該第二電極與該導電或半導電中 -第 極及該 电甲間層之間且與該 -第二電阻切換 156917.doc 63 · 201212319 層。 在另一實施例中,一記情體驻 6隐骽裝置包括:一記憶體陣列, ;包括複數個電阻切換記憶體單元,每-電阻切換記憶體 単元包括與-電阻切換記憶體元件串聯之一操縱元件,每 一電阻切換記憶體元件包括 :電阻切換層之間的-中間層;複數個字線及位元 母—電阻切換記憶體單元具有與該複數個 位元線中之 :各別位元線連通之-端,及與該複數個字線中之一各別 子線連通之另-端;及與該複數個字線及位元線連通之控 制電路,該控制電路經由其該各別位元線及該字線將一電 =加至該等電阻切換記憶體單元中之至少一者,以致使 =電:: 刀換記憶體單元中之該至少一者之該電阻切換記 隐體兀件自一個電阻狀態切換至另—電阻狀態。 在另f施例中’一電阻切換記憶體單元包括:第一電 極及第二電極;以電方式位於該第一電極與該第二電極之 間且與該等電極串聯之一導電或半導電中間層;以電方式 位於該第一電極與該導電或半導電中間層之間且與該第一 電極及該導電或半導電中間層串聯之一第一電阻切換層; 及以電方式位於該第二電極與該導電或半導電中間層且盘 該第二電極及該導電或半導電中間層串聯之一第二電阻:刀 換層」該第-電極、該第二電極、該導電或半導電令間 層》亥第t阻切換層及該第二電阻切換層中之至少一者 係至少部分地在該第一電極、該第二電極、該導電或半導 電中間層、該第一電阻切換層及該第二電阻切換層中之至 156917.doc -64 - 201212319 少一個其他者側向配置。 在另-實施财,一電阻切換記憶體單元包 極及第二電極;以電方式位於該第—電極電 ;且與該等電極串聯之-導電或半導電中間層:電;; 位於該第一電極與該導電 式 T守电甲間層之間 該導電或半導電中間層串聯之-第-電阻=層 且與該第二電極及該導電或半導電中間層串聯之 =換層,該導電或半導電中間層及該第—電阻切換層及 該第一電阻切換層係一個L形及U形中之至少一者。 甘在另一實施例一記憶體裝置包括::記憶體陣列, 二包括複數個電阻切換記憶體單元,每—電阻切換記憶體 早兀包括與-電阻切換記憶體元件争聯之一操縱元件,每 一電阻切換記憶體元件包括以電方式位於第一電阻切換層 與第二電阻切換層之間的一中間層,以及第一電極及第二 電極;對於每一電阻切換記憶體單元:該第一電極、該第 二電極、該導電或半導電中間層、該第一電阻切換層及該 第二電阻切換層十之至少一者至少部分地在該第一電極、 該第二電極、1¾導電或半導電中間層、該第一電阻切換層 及該第二電阻切換層中之至少一個其他者側向配置;複數 個字線及位元線·,每-電阻切換記憶體單元具有與該複數 個位元線中之一各別位元線連通之_端及與該複數個字線 中之-各別字線連通之及與該複數個字線及位元 線連通之控制電路,該控制電路經由其該各別位元線及該 156917.doc •65· 201212319 字線將一電壓施加至該等電阻切換記憶體單元中之至少一 者,以致使該等電阻切換記憶體單元中之該至少一者之該 電阻切換記憶體元件自一個電阻狀態切換至另一電阻狀 態。 在另一實施例中,一種用於改變一電阻切換記憶體單元 中之一電阻狀態之方法包括:跨越該電阻切換記憶體單元 之第一電極及第二電極施加一電壓以在該記憶體單元中設 疋一第一資料狀態’該電壓係電容性轉合至一導電或半導 電中間層,該導電或半導電中間層以電方式位於該第一電 極與該第二電極之間且與該等電極串聯,該電壓致使在以 下各項中之至少一者中切換一電阻狀態:(a)以電方式位於 該第一電極與該導電或半導電中間層之間且與該第一電極 及該導電或半導電中間層串聯之一第一電阻切換層;及(b) 以電方式位於該第二電極與該導電或半導電中間層之間且 與該第二電極及該導電或半導電中間層串聯之一第二電阻 切換層,及移除該電壓以允許該電阻切換記憶體單元放 電。 在另一實施例中,一種用於改變一電阻切換記憶體單元 中之電阻狀態之方法,其包括:增加跨越該電阻切換記 憶體單元而施加之一隨時間變化之電壓之一量值直至在該 電阻切換記憶體單元之第一電阻切換層及第二電阻切換層 中之一者中切換一電阻狀態為止;及隨後,進一步增加跨 越該電阻切換記憶體單元而施加之該隨時間變化之電壓之 該量值直至在該電阻切換記憶體單元之該第一電阻切換層 156917.doc •66- 201212319 及該第二電阻切換層中之另_者中切換一電阻狀態為止。 在另一實施例中,-種用於改變-電阻切換記憶體單元 中之-電阻狀態之方法包括:跨越第一控制線及第二控制 、線:施加-電壓’該第一控制線係連接至一電阻切換記憶 體單元之4 ’該第二控制線係連接至與該電阻切換記憶 體單兀串聯之-操縱元件,跨越該電阻切換記憶體單元之 第一電阻切換層及第二電阻切換層且跨越以電方式位於該 第-電阻切換層肖該第二電阻切換層之間的一㈣或半導 電中間層兩端施加該電壓;及移除該電壓以允許該電阻切 換記憶體單元放電。 在另-實施例甲,一電阻切換記憶體單元包括一操縱元 件及與該操縱元件串聯之一電阻切換記憶體元件,該電阻 切換記憶體元件包括:第一電極及第二電極;位於該第— 電極與該第二電極之間且與該等電極串聯之一導電或半導 電中間層;位於該第-電極與該導電或半導電中間層之間 且與該第一電極及該導電或半導電中間層串聯之一第—電 阻切換層;及位於該第二電極與該導電或半導電中間層之 . 間且與該第二電極及該導電或半導電中間層串聯之一第二 * 電阻切換層。 • 在另—實施例中,一電阻切換記憶體元件包括:第—電 極及第二電極;位於該第一電極與該第二電極之間且與該 等電極串聯之一導電或半導電中間層;位於該第一電極與 忒導電或半導電中間層之間且與該第一電極及該導電或半 導電中間層串聯之-第一電阻切換層,該第一電阻切換層 156917.doc 201212319 包括Me〇x ;位於該第二電極與該導電或半導電中間層之 間且與該第二電極及該導電或半導電中間層串聯之一第二 電阻切換層,該第一電阻切換層包括Me〇x,·及位於該導 電或半導電中間層與該第一電極之間的一頂蓋層,該頂蓋 層係選自由TiOx、AI203、Zr〇x、La0x、丫〇}{構成之群 組,自該第一電阻切換層之一角度看,該頂蓋層充當一氧 源或吸氧劑。 在另一實施例中,一記憶體裝置包括:一記憶體陣列, 其包括複數個記憶體單元,每一記憶體單元包括與一電阻 切換記憶體元件串聯之一操縱元件,每一電阻切換記憶體 元件包括位於第一電阻切換層與第二電阻切換層之間的一 中間層;複數個字線及位纟線;每一記憶體單元具有與該 複數個位元線中之一各別位元線連通之一端及與該複數個 子線t之-各別字線連通之另一端;及與該複數個字線及 位元:連通之控制電路,該控制電路經由其該各別位元線 及該子線將一電壓施加至該等記憶體單元中之至少一者, 以致使該等記憶體單元中之該至少一者之該電阻切換記憶 體元件自一個電阻狀態切換至另一電阻狀態。 在另一實施例中,一電阻切換記憶體單元包括:第一電 極及第一電極’以電方式位於該第-電極與該第二電極之 間且與該等電極串聯之一導電中間層;以電方式位於該第 電極與該導電中間層之間且與該第一電極及該導電中間 層串聯之一電阻切換層;及以電方式位於該第二電極與該 導電中門層之間且與該第三電極及該導電中間層串聯之一 156917.doc -68 - 201212319 崩潰層,該崩潰;S名& π _ _ 至Π)廳之-電阻處於—導電狀態時維持至少約ιμΩ =實施例中’―電阻切換記憶體單元包括 讀及與該操縱元件串聯之—電阻切換記憶體元件,= 阻切換記憶體元件包括:第一電極及第二電極 以電 位於該第一電極與該第二電極之間且與該等電極串^式 導電中間層;以電方式位於該第一電極與該導電中;— 間且與該第—電極及該導電令間層串聯之一電2之 及=方式位於該第二電極與該導電中間層矣二 一導電狀離時層,該崩潰層在處於 導電狀態時維持至少約1廳至10廳之一電阻。 在另-實施财…電阻㈣記龍單 ::與該操縱元件串聯之-電阻切換記憶體元件:: :::憶;元件包括:第-電極及第二…電: =該第—電極與該第二電極之間且與該等電㈣聯之 =導電令間層,·以電方式位於該第一電極與該導電或 層串聯導電或半導電令間 養電戈丰式位於該第二電極與該 4電或+導電中間層之間且與該第二電極及該導電 2間層串聯之-崩潰層,該崩潰層在處於—導電狀態時 ,准持至少約1 ΜΩ至10ΜΩ之一電阻。 實施例中’一記憶體裝置包括記憶體陣列,該記 -體陣列包括複數個記憶體單元,每一記憶體單元包括與 -電阻切換記憶體單元串聯之一操縱元件。每—電阻切換 156917.doc •69· 201212319 記憶體元件包括··第一電極及第二電極;以電 第一電極錢第三電極 ,於該 主道帝上 j ,、β亥等電極串聯之一導電或 中間層,’以電方式位於該第一 一 第—電極及該導電或半導電中間層争 :=阻切換層;及以電方式位於該第二電極與該導電 或+導電中間層之間且與該第二電極及該導電或半導電中 :層串聯之一崩潰層’該崩潰層在處於一導電狀態時維持 /約1 Μ Ω至10 Μ Ω之—電阻。該記憶體裝置亦包括:複 數個字線及位元線;每-記憶體單元具有與該㈣個位元 線中之-各別位元線連通之一端,及與該複數個字線中之 Γ各別字線連通之另一端;及與該複數個字線及位元線連 通之㈣電路’該控制電路經由其該各別位元線及該字線 將一電壓施加至該等記憶體單元中之至少一者以致使該 等圯憶體單元中之該至少一者之該電阻切換記憶體元件自 一個電阻狀態切換至另一電阻狀態。 上文已出於圖解說明及閱述之目的提供對本文中之技術 之詳細闡述。本文並非意欲窮盡或將本技術限制於所揭示 之精確形式。鑒於上文之教示可作出諸多修改及變化形 式。選擇所闡述之實施例旨在最佳地闡釋本技術之原理及 其實際應用,以藉此使得熟習此項技術者能夠在各種實施 例中並藉助適合於所涵蓋之特定使用之各種修改更佳地利 用本技術。本技術之範疇意欲由隨附申請專利範圍來界 定。 【圖式簡單說明】 156917.doc -70- 201212319 圖1係包含與一操縱元件串聯之一 RSME之一記憶體單元 之一項實施例之一簡化透視圖。 圖2A係由複數個圖1之記憶體單元形成之一第一記憶體 層級之一部分之一簡化透視圖。 圖2B係由複數個圖1之記憶體單元形成之三維記憶體陣 列之一部分之一簡化透視圖。 圖2C係由複數個圖1之記憶體單元形成之三維記憶體陣 列之一部分之一簡化透視圖。 圖3係一記憶體系統之一項實施例之一方塊圖。 圖4A係繪示一實例性單極RSL之I-V特性之一圖表。 圖4B係繪示兩個實例性單極RSL之不同I-V特性之一圖 表。 圖4C係繪示另一實例性單極RSL之I-V特性之一圖表。 圖4D係繪示一實例性雙極RSL之I-V特性之一圖表。 圖4E係繪示另一實例性雙極RSL之I-V特性之一圖表。 圖5繪示用於讀取一記憶體單元之狀態之一電路之一實 施例。 圖6A繪示具有一 RSME及位於該RSME下面之一操縱元 件(SE)之一實例性記憶體單元。 圖6B繪示具有一 RSME之一記憶體單元之一替代組態, 其中該操縱元件(SE)位於該RSME上面。 圖6C繪示作為呈一垂直堆叠之一鏡像電阻式開關(MRS) 之圖6A之RSME之一實例性實施方案。 圖6D繪示在RSL之間使用多個中間層(IL)之圖6A之 156917.doc -71 - 201212319 RSME之一實例性實施方案。 圖6E繪示使用一重複RSL/IL型樣之圖6A之RSME之一實 例性實施方案》 圖6F繪示其中RSME之每一層水平延伸且該等層中之一 或多者端對端地配置之圖6A之RSME之一實例性實施方 案。 圖6G繪示其中RSME之每一層水平延伸且該等層中之一 或多者端對端地配置之圖6A之RSME之另一實例性實施方 案。 圖6H繪示其中RSME之每一層垂直延伸之圖6A之RSME 之另一實例性實施方案。 圖61繪示包含RSL1、IL、RSL2及E2之L形部分之圖6A之 RSME之另一實例性實施方案。 圖6J繪示包含RSL1、IL、RSL2及E2之U形部分之圖6A 之RSME之另一實例性實施方案。 圖6K1繪示使用一個RSL及位於該RSL下面之一個崩潰層 之圖6A之RSME之一實例性實施方案。 圖6K2係展示一崩潰層自一初始狀態至一崩潰狀態之一 轉變之一圖表。 圖6K3係展示一崩潰層處於一初始狀態(實線)及處於一 崩潰狀態(虛線)之一 I-V特性之一圖表。 圖6L繪示使用一個RSL及位於該RSL上面之一個崩潰層 之圖6A之RSME之一實例性實施方案。 圖6M繪示其中該等RSL係不同類型之圖6A之RSME之一 156917.doc -72- 201212319 實例性實施方案。 圖7A繪示作為一 Si二極體之圖6A之記憶體單元之操縱 元件(SE)之一實例性實施方案。 圖7B繪示作為一穿通二極體之圖6A之記憶體單元之操 縱元件(SE)之一實例性實施方案。 圖8繪示連接於一位元線與一字線之間的圖6A之記憶體 單元之一實例性實施方案。 圖9A繪示圖6C之RSME之一實施例,其中E1係由Co、 CoSi、n+ Si、p+ Si 或 p+ SiC 製成且 E2 係由 n+ Si 製成。 圖9B繪示圖6C之RSME之一實施例,其中El及IL係由p+ SiC製成且E2係由n+ Si、n+ SiC或p+ SiC製成。 圖9C係繪示p+ SiC相對於其他材料之費米能階之一圖 示。 圖10A繪示闡述替代IL材料之圖6C之RSME之一實施 例。 圖10B繪示呈一倒置鏡像堆疊組態之圖6C之RSME之一 實施例。 圖10C繪示呈一不對稱、直立堆疊組態之圖6C之RSME 之一實施例。 圖10D繪示呈一不對稱、倒置堆疊組態之圖6A之RSME 之一實施例。 圖11A繪示展示當E2為n+ Si時SiOx之生長之圖6C之 RSME之一實施例。 圖11B繪示展示當E2為TiN時一低帶隙材料(諸如,TiOx) 156917.doc -73- 201212319 之生長之圖6C之RSME之一實施例。 圖11C繪示圖6C之RSME之一實施例,其中該等RSL係由 一經摻雜金屬氧化物製成以減小操作電壓。 圖11D繪示圖11C之RSME之一實施例,其中E2為TiN代 替 n+ Si。 圖11E繪示呈一不對稱鏡像單元組態之圖6C之RSME之 一實施例,其中該等RSL係由不同材料製成。 圖11F繪示呈一不含SiOx之不對稱鏡像單元組態之圖6C 之RSME之一實施例。 圖12繪示圖6C之RSME之一能量圖。 圖13繪示在一 RSL之一設定程序中一高電場之施加。 圖14A至圖14D繪示在一RSL之一設定程序中形成一導電 細絲中之不同階段。 圖14E、圖14F及圖14G係分別闡述圖14A、圖14B及圖 14D之設定程序階段之能量圖。 圖15A至圖15C繪示在一RSL之一重設程序中移除一導電 細絲中之不同階段。 圖15D、圖15E及圖15F係分別闡述圖15A、圖15B及圖 1 5C之重設程序階段之能量圖。 圖16A繪示圖6A之RSME之一設定程序。 圖16B繪示圖6A之RSME之一重設程序。 【主要元件符號說明】 100 電阻切換記憶體單元 102 電阻切換記憶體元件 156917.doc • 74· 201212319 104 操縱元件 106 第一導體 108 第二導體 113 障壁層 114 第一記憶體層級 116 單體式三維陣列 118 第一記憶體層級 120 第二記憶體層級 130 電阻切換層 132 電極 133 導電中間層 134 電極 135 電阻切換層 142 重播雜n.+多晶石夕: 144 輕摻雜或本質(非 146 重換雜P+多晶石夕 300 記憶體系統 302 記憶體陣列 306 輸入/輸出 308 輸出 310 行控制電路 312 行解碼器 314 陣列端子接收器 316 區塊選擇電路 156917.doc -75- 201212319 320 列控制電路 322 列解碼器 324 陣列端子驅動器 326 區塊選擇電路 330 系統控制邏輯 400 線 402 線 404 線 406 線 420 線 422 線 424 線 426 線 430 線 432 線 434 線 436 線 440 線 442 線 444 線 446 線 450 線 452 線 454 線 156917.doc •76· 201212319 456 線 547 字線 549 字線 550 記憶體單元 552 記憶體單元 554 記憶體單元 556 記憶體單元 557 位元線 558 電晶體 559 位元線 560 寫入電路 562 電晶體 563 資料匯流排 564 箝位控制電路 566 感測放大器 568 資料鎖存器 A1 箭頭 A2 箭頭 AL 黏合劑層 AL1 第一黏合層 AL2 第二黏合層 BLC 位線觸點 Cap 1 一個頂蓋層 Cap2 另一頂蓋層 156917.doc -77- 201212319And U.S. Patent No. 7,391,638, issued to U.S. Patent Application Serial No. No. No. No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No No. Both patents are incorporated herein by reference. At step 1604, a voltage is coupled to the intermediate layer (IL), and the IL scatters electrons from the RSL into the IL. At step 1606, one or more filaments are formed in the RSL. See also Figures 14A-14D. In different RSLs, filament formation can occur at different rates and at different times. For example, referring to Fig. 4B, when the set voltage reaches VsetB, the RSL of type "B" will first reach the set state, and then the RSL for type "A" reaches the set state when the set voltage reaches VsetA. The set voltage is sufficient to form a filament in each of the RSLs to provide a conductive path in the RSLs thereby providing a conductive path through the RSME and memory cells. Therefore, a low resistance state is reached in each of the RSLs and in the RSME. The low resistance state of RSME can be assigned to a first binary data state, for example, 0 or 1. At step 1608, the set voltage is removed and the memory cell (including the RSME) is discharged. Note that steps 1602 through 1606 occur at least partially simultaneously. Depending on the situation, only one of the RSLs may complete the setup procedure, or less than the RSL of all RSLs in the RSME to complete the setup procedure. Figure 16B illustrates a reset procedure for one of the RSMEs of Figure 6A. At step 1620, the reset procedure is initiated for a memory unit. At step 1622, a reset voltage (Vreset, see Figure 4A) is applied across the first and second electrodes. The voltage is applied across a first electrode and a second electrode of the memory cell by switching the memory cell in series with a resistor switching memory cell. The set voltage can have a desired waveform, such as a fixed amplitude 156917.doc • 60-201212319 pulse or a ramp-up pulse. Therefore, the voltage can be a time-varying voltage signal, for example, the magnitude increases with time. . As before, in one method, the set voltage is blindly applied without determining whether or not the set state is actually reached. In this case the 'reset voltage is enough to make it close! 00. All memory cells of /〇 reach a duration and/or magnitude of the reset state. In another method, the state of the memory cell is monitored when a reset voltage is applied and the reset voltage is removed when the monitor indicates that the reset state has been reached. This method is further described in the above-mentioned US 2010/008 5794 and US 7,391,638. At step 1624, a voltage is coupled to the intermediate layer, and IL scatters electrons from the RSL into the IL. At step 1626, one or more filaments are removed or destroyed in the RSLs. See also Figures 15A-15C. The removal of these filaments in different RSLs can be done at different rates and at different times. For example, referring to FIG. 4B, when the reset voltage reaches VresetB, the RSL of type "B" will first be reset, and then the RSL of type "A" reaches the reset state when the reset voltage reaches VresetA. . The reset voltage is sufficient to remove filaments in each of the RSLs to remove one of the conductive paths of the RSLs thereby removing a conductive path through the rSme and the memory cells. Therefore, a high resistance state is achieved in each of the RSLs and in the RSME. The high resistance state of the rSme can be assigned to a second binary data state opposite the low resistance data state, such as 1 or 0. At step 1628, the reset voltage is removed and the memory cell (including the RSME) is removed. Discharge. Note that steps 1622 through 1626 occur at least partially simultaneously. 156917.doc -61- 201212319 Depending on the situation, only one of the RSLs may complete the reset procedure, or less than the RSL of all RSLs in the RSME to complete the reset procedure. The method may include: applying a voltage across the first electrode and the second electrode of the resistance switching memory cell to set a first data state in the memory cell, wherein the voltage is capacitively coupled to a conductive intermediate layer, The conductive intermediate layer is electrically located between the first electrode and the second electrode and in series with the electrodes, and the voltage causes a resistance state to be switched in at least one of: (a) a first a resistance switching layer electrically disposed between the first electrode and the conductive intermediate layer and in series with the first electrode and the conductive intermediate layer, and (b) a second resistance switching layer electrically located at the first The second electrode is in series with the conductive intermediate layer and in series with the second electrode and the conductive intermediate layer, and the voltage is removed to allow the resistance switching memory cell to discharge. The resistive switching layers can be reversible or irreversible. The above method may further comprise: changing a resistance state of a resistor switching memory unit by: (a) increasing a voltage value of one of the voltage changes over time by switching the memory unit across the resistor until the Switching a resistance state to one of the first resistance switching layer and the second resistance switching layer of the resistance switching memory cell; and (b) subsequently increasing the time variation applied by switching the memory cell across the resistance The magnitude of the voltage is until a resistance state is switched in the other of the first resistance switching layer and the first resistance switching layer of the resistance switching memory cell. This switching of the resistance state can be reversible or irreversible. The method may further include: applying a voltage across the first control line and the second control line, wherein the first control line is connected to a resistance switching memory 156917.doc • 62·201212319 one end of the single 7C, the second The control line is connected to the control element connected in series with the resistance switching memory unit; and the first resistance switching layer and the second resistance switching layer of the memory unit are switched across the resistance and span the first resistance switching layer and the second A conductive inter-layer between the resistive switching layers applies the electric bunker; and the voltage is removed to allow the resistive switching memory cell to discharge. The dedicated resistance switching layer can be reversible or irreversible. - Thus, it can be seen that in one embodiment, a resistive switching memory cell comprises: a first electrode and a second electrode; electrically located between the first electrode and the second electrode and The electrode is connected in series with one of the conductive intermediate layers; electrically disposed between the first electrode and the conductive intermediate layer and with the first electrode and the conductive intermediate layer (four) - the first resistance switching layer; and electrically located at the first a second resistance switching layer between the second electrode and the conductive intermediate layer and in parallel with the second electrode and the conductive intermediate layer, the _ resistance switching layer and the second resistance switching layer both have a bipolar switching characteristic or All have a pole switching feature. In another embodiment, a resistance switching memory unit includes: a one-pole operating element; and a resistor switching memory element with the one of the diode operating elements, the resistor switching memory element An electrode and a second electrode; an electrically conductive or semiconductive intermediate layer electrically disposed between the first electrode and the second electrode and in series with the electrodes; electrically located at the negative, an electrode and the conductive or Between the semiconducting intermediate layers and one of the first and second layers of the conductive or semiconductive intermediate layer, and the y knife is layered; and electrically located in the second electrode and the conductive or semiconductive layer - The first pole and the armor layer are switched between the 156917.doc 63 · 201212319 layer and the second resistor. In another embodiment, a device 6 concealing device includes: a memory array, including a plurality of resistance switching memory cells, each of the resistance switching memory cells including the resistor-switching memory device in series An operating element, each of the resistance switching memory elements comprises: an intermediate layer between the resistance switching layers; a plurality of word lines and a bit mother-resistance switching memory unit having the same plurality of bit lines: each a terminal connected to the bit line, and another end connected to each of the plurality of word lines; and a control circuit connected to the plurality of word lines and the bit line, the control circuit via the control circuit Each of the bit lines and the word line adds an electric= to at least one of the resistance switching memory cells such that the resistance switching of the at least one of the = memory: memory cells The hidden component switches from one resistance state to another. In another embodiment, a resistor switching memory unit includes: a first electrode and a second electrode; electrically electrically located between the first electrode and the second electrode and electrically or semiconductively connected to the electrodes in series An intermediate layer; a first resistance switching layer electrically disposed between the first electrode and the conductive or semiconductive intermediate layer and in series with the first electrode and the conductive or semiconductive intermediate layer; and electrically located a second electrode and the conductive or semiconductive intermediate layer and the second electrode and the conductive or semiconductive intermediate layer are connected in series with a second resistor: a knife-changing layer, the first electrode, the second electrode, the conductive or half At least one of the conductive interlayer and the second resistive switching layer is at least partially at the first electrode, the second electrode, the conductive or semiconductive intermediate layer, and the first resistor In the switching layer and the second resistance switching layer, there is one other side configuration to 156917.doc -64 - 201212319. In another implementation, a resistor switches between the memory cell package pole and the second electrode; electrically located at the first electrode; and the conductive or semiconductive intermediate layer in series with the electrodes: electricity; Between the electrode and the conductive T-conserving inter-layer, the conductive or semi-conductive intermediate layer is connected in series with a -first resistance = layer and is connected in series with the second electrode and the conductive or semiconductive intermediate layer. The conductive or semiconductive intermediate layer and the first resistance switching layer and the first resistance switching layer are at least one of an L shape and a U shape. In another embodiment, a memory device includes: a memory array, and a plurality of resistor switching memory cells, each of which includes a control element for competing with the resistor-switching memory device. Each of the resistance switching memory elements includes an intermediate layer electrically disposed between the first resistance switching layer and the second resistance switching layer, and the first electrode and the second electrode; and the memory unit is switched for each resistor: the first Conducting at least one of an electrode, the second electrode, the conductive or semiconductive intermediate layer, the first resistance switching layer and the second resistance switching layer at least partially at the first electrode, the second electrode, and the first electrode Or at least one of the semiconductive intermediate layer, the first resistance switching layer, and the second resistance switching layer are laterally disposed; the plurality of word lines and the bit lines ·, each of the resistance switching memory cells having the complex number a control circuit for connecting the _ terminal of each of the bit lines and the respective word lines of the plurality of word lines and the plurality of word lines and the bit lines, the control The circuit applies a voltage to at least one of the resistance switching memory cells via the respective bit lines and the 156917.doc • 65·201212319 word lines such that the resistors switch to the memory cells At least one of the resistance switching memory elements switches from one resistance state to another. In another embodiment, a method for changing a resistance state of a resistance switching memory cell includes: applying a voltage across the first electrode and the second electrode of the resistance switching memory cell to the memory cell Providing a first data state, wherein the voltage is capacitively coupled to a conductive or semiconductive intermediate layer electrically disposed between the first electrode and the second electrode and The electrodes are connected in series, the voltage causing a resistance state to be switched in at least one of: (a) electrically located between the first electrode and the conductive or semiconductive intermediate layer and with the first electrode and The conductive or semiconductive intermediate layer is in series with one of the first resistance switching layers; and (b) is electrically located between the second electrode and the conductive or semiconductive intermediate layer and with the second electrode and the conductive or semiconductive The intermediate layer is in series with one of the second resistive switching layers, and the voltage is removed to allow the resistive switching memory cell to discharge. In another embodiment, a method for changing a resistance state in a resistance switching memory cell includes: increasing a value of a voltage that varies over time by switching a memory cell across the resistance until And switching one of the first resistance switching layer and the second resistance switching layer of the resistance switching memory unit to a resistance state; and subsequently, further increasing the voltage that is applied across the resistance switching memory unit The magnitude is until a resistance state is switched in the other of the first resistance switching layers 156917.doc • 66-201212319 and the second resistance switching layer of the resistance switching memory unit. In another embodiment, a method for changing a resistance state in a resistance-switching memory cell includes: crossing a first control line and a second control, a line: applying a voltage 'the first control line connection 4' to the resistance switching memory unit 4', the second control line is connected to the control element connected in series with the resistance switching memory unit, and the first resistance switching layer and the second resistance switching of the memory unit are switched across the resistance And applying a voltage across a layer of a (four) or semiconducting intermediate layer electrically between the second resistive switching layer; and removing the voltage to allow the resistive switching memory cell to discharge . In another embodiment, a resistive switching memory unit includes an operating element and a resistor switching memory element in series with the operating element, the resistive switching memory element comprising: a first electrode and a second electrode; a conductive or semiconductive intermediate layer between the electrode and the second electrode in series with the electrodes; between the first electrode and the conductive or semiconductive intermediate layer and with the first electrode and the conductive or semi a conductive first layer connected in series with the first resistance switch; and a second* resistor between the second electrode and the conductive or semiconductive intermediate layer and in series with the second electrode and the conductive or semiconductive intermediate layer Switch layers. In another embodiment, a resistive switching memory device includes: a first electrode and a second electrode; a conductive or semiconductive intermediate layer between the first electrode and the second electrode and in series with the electrodes a first resistance switching layer between the first electrode and the germanium conductive or semiconductive intermediate layer and in series with the first electrode and the conductive or semiconductive intermediate layer, the first resistance switching layer 156917.doc 201212319 includes Me〇x; a second resistance switching layer between the second electrode and the conductive or semiconductive intermediate layer and in series with the second electrode and the conductive or semiconductive intermediate layer, the first resistance switching layer comprising Me 〇x, · and a cap layer between the conductive or semiconductive intermediate layer and the first electrode, the cap layer is selected from the group consisting of TiOx, AI203, Zr〇x, La0x, 丫〇}{ The cap layer serves as an oxygen source or an oxygen absorbing agent from the perspective of the first resistance switching layer. In another embodiment, a memory device includes: a memory array including a plurality of memory cells, each memory cell including a steering element in series with a resistive switching memory component, each resistor switching memory The body element includes an intermediate layer between the first resistance switching layer and the second resistance switching layer; a plurality of word lines and a bit line; each memory cell has a different bit from the plurality of bit lines a terminal connected to the source line and another end connected to the respective word lines of the plurality of sub-lines t; and a control circuit connected to the plurality of word lines and the bit line, the control circuit via the respective bit line And the sub-line applies a voltage to at least one of the memory cells such that the at least one of the memory cells switches from the one resistance state to another resistance state . In another embodiment, a resistance switching memory cell includes: a first electrode and a first electrode ' electrically connected between the first electrode and the second electrode and a conductive intermediate layer in series with the electrodes; a resistive switching layer electrically disposed between the first electrode and the conductive intermediate layer and in series with the first electrode and the conductive intermediate layer; and electrically disposed between the second electrode and the conductive intermediate gate layer In conjunction with the third electrode and the conductive intermediate layer, one of the 156917.doc-68 - 201212319 collapse layers, the collapse; S name & π _ _ to Π) - the resistance is maintained at - conductive state at least about ιμΩ = In the embodiment, the “resistive switching memory unit includes a read-and-synchronized-switching memory element in series with the operating element, and the resistive switching memory element includes: the first electrode and the second electrode are electrically located at the first electrode and the Conducting an intermediate layer between the second electrodes and the electrodes; electrically located between the first electrode and the conductive; and electrically connected to the first electrode and the conductive interposed layer And = method is located in the first Electrode and the conductive intermediate layer twenty-one carry away from the shaped conductive layer, and the collapse of the layers in the conductive state is maintained at least about 1 Living Room resistor 10 to one. In the other - implementation of the ... resistance (four) record dragon single:: in series with the operating element - resistance switching memory components :::: recall; components include: the first electrode and the second ... electricity: = the first electrode and Between the second electrodes and the electric (four) = conductive inter-layer, electrically located in the first electrode and the conductive or layer series electrically or semi-conductive, between the second a collapsing layer between the electrode and the 4-electrode or +-conducting intermediate layer and in series with the second electrode and the conductive inter-layer 2, the collapsing layer being at least about 1 ΜΩ to 10ΜΩ when in a conducting state resistance. In the embodiment, a memory device includes a memory array including a plurality of memory cells, each memory cell including a steering element in series with a resistance switching memory cell. Each-resistance switch 156917.doc •69· 201212319 The memory component includes··the first electrode and the second electrode; the first electrode of the electric first electrode and the third electrode, and the electrodes of the main channel, j, and β Hai are connected in series a conductive or intermediate layer, electrically located in the first first electrode and the conductive or semiconductive intermediate layer: = resistive switching layer; and electrically located in the second electrode and the conductive or + conductive intermediate layer A breakdown layer between and in contact with the second electrode and the conductive or semiconductive layer: the breakdown layer maintains / resistance of about 1 Μ Ω to 10 Ω Ω when in a conductive state. The memory device also includes: a plurality of word lines and bit lines; each of the memory cells has one end connected to the respective bit lines of the (four) bit lines, and the plurality of word lines The other end of the respective word line connection; and the fourth circuit connected to the plurality of word lines and the bit line', the control circuit applies a voltage to the memory via the respective bit line and the word line At least one of the cells such that the at least one of the ones of the memory cells switch from one resistive state to another. The detailed description of the techniques herein is provided for the purposes of illustration and description. This document is not intended to be exhaustive or to limit the invention to the precise form disclosed. Many modifications and variations are possible in light of the above teachings. The embodiments described herein are chosen to best explain the principles of the present invention and its application, and thus, in the embodiments of the invention, Use this technology. The scope of the technology is intended to be defined by the scope of the accompanying claims. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a simplified perspective view of one embodiment of a memory unit comprising one of the RSMEs in series with a steering element. Figure 2A is a simplified perspective view of one of the first memory levels formed by a plurality of memory cells of Figure 1. Figure 2B is a simplified perspective view of one of the portions of a three dimensional memory array formed by a plurality of memory cells of Figure 1. Figure 2C is a simplified perspective view of one of the portions of a three-dimensional memory array formed by a plurality of memory cells of Figure 1. 3 is a block diagram of an embodiment of a memory system. 4A is a graph showing one of the I-V characteristics of an exemplary monopolar RSL. Figure 4B is a graph showing one of the different I-V characteristics of two exemplary unipolar RSLs. 4C is a graph showing one of the I-V characteristics of another exemplary monopolar RSL. 4D is a graph showing one of the I-V characteristics of an exemplary bipolar RSL. 4E is a graph showing one of the I-V characteristics of another exemplary bipolar RSL. Figure 5 illustrates an embodiment of a circuit for reading a state of a memory cell. Figure 6A illustrates an exemplary memory unit having an RSME and one of the steering elements (SE) located below the RSME. Figure 6B illustrates an alternate configuration of one of the memory cells having an RSME, wherein the steering element (SE) is located above the RSME. 6C illustrates an exemplary embodiment of the RSME of FIG. 6A as a mirrored resistive switch (MRS) in a vertical stack. Figure 6D illustrates an exemplary embodiment of 156917.doc-71 - 201212319 RSME of Figure 6A using multiple intermediate layers (IL) between RSLs. 6E illustrates an exemplary embodiment of the RSME of FIG. 6A using a repeating RSL/IL pattern. FIG. 6F illustrates that each layer of the RSME extends horizontally and one or more of the layers are configured end-to-end. An exemplary embodiment of the RSME of Figure 6A. 6G illustrates another exemplary implementation of the RSME of FIG. 6A in which each layer of the RSME is horizontally extended and one or more of the layers are configured end-to-end. Figure 6H illustrates another exemplary embodiment of the RSME of Figure 6A in which each layer of the RSME extends vertically. 61 illustrates another exemplary embodiment of the RSME of FIG. 6A including the L-shaped portions of RSL1, IL, RSL2, and E2. 6J illustrates another exemplary embodiment of the RSME of FIG. 6A including U-shaped portions of RSL1, IL, RSL2, and E2. Figure 6K1 illustrates an exemplary embodiment of the RSME of Figure 6A using an RSL and a collapse layer located below the RSL. Figure 6K2 shows a graph of one of the collapse levels from an initial state to a collapsed state. Fig. 6K3 is a graph showing one of the I-V characteristics of a collapsed layer in an initial state (solid line) and in a collapsed state (dashed line). Figure 6L illustrates an exemplary embodiment of the RSME of Figure 6A using an RSL and a collapse layer located above the RSL. Figure 6M illustrates an exemplary embodiment in which the RSLs are of a different type of RSME of Figure 6A 156917.doc-72-201212319. Figure 7A illustrates an exemplary embodiment of a steering element (SE) of the memory cell of Figure 6A as a Si diode. Figure 7B illustrates an exemplary embodiment of an operational element (SE) of the memory cell of Figure 6A as a feedthrough diode. Figure 8 illustrates an exemplary embodiment of the memory cell of Figure 6A coupled between a bit line and a word line. Figure 9A illustrates an embodiment of the RSME of Figure 6C in which E1 is made of Co, CoSi, n+ Si, p+ Si or p+ SiC and E2 is made of n+ Si. 9B illustrates an embodiment of the RSME of FIG. 6C in which El and IL are made of p+ SiC and E2 is made of n+Si, n+ SiC, or p+ SiC. Figure 9C is a graph showing one of the Fermi energy levels of p+ SiC relative to other materials. Figure 10A illustrates one embodiment of the RSME of Figure 6C illustrating alternative IL materials. Figure 10B illustrates an embodiment of the RSME of Figure 6C in an inverted mirror stack configuration. Figure 10C illustrates one embodiment of the RSME of Figure 6C in an asymmetric, upright stack configuration. Figure 10D illustrates an embodiment of the RSME of Figure 6A in an asymmetric, inverted stacked configuration. Figure 11A depicts an embodiment of the RSME of Figure 6C showing the growth of SiOx when E2 is n+Si. Figure 11B depicts an embodiment of the RSME of Figure 6C showing the growth of a low bandgap material (such as TiOx) 156917.doc -73 - 201212319 when E2 is TiN. Figure 11C illustrates an embodiment of the RSME of Figure 6C wherein the RSLs are made of a doped metal oxide to reduce the operating voltage. Figure 11D illustrates an embodiment of the RSME of Figure 11C, wherein E2 is TiN instead of n+Si. Figure 11E illustrates an embodiment of the RSME of Figure 6C in an asymmetric mirror unit configuration wherein the RSLs are made of different materials. Figure 11F illustrates an embodiment of the RSME of Figure 6C in an asymmetric mirror unit configuration without SiOx. Figure 12 is a diagram showing an energy diagram of the RSME of Figure 6C. Figure 13 illustrates the application of a high electric field in one of the RSL setting procedures. Figures 14A through 14D illustrate different stages in forming a conductive filament in one of the RSL setting procedures. 14E, 14F, and 14G illustrate energy diagrams of the setup procedure stages of Figs. 14A, 14B, and 14D, respectively. Figures 15A through 15C illustrate different stages in the removal of a conductive filament in one of the RSL reset procedures. 15D, 15E, and 15F are energy diagrams illustrating the reset procedure stages of Figs. 15A, 15B, and 15C, respectively. FIG. 16A illustrates one of the RSME setting procedures of FIG. 6A. Figure 16B illustrates one of the RSME reset procedures of Figure 6A. [Main component symbol description] 100 resistance switching memory unit 102 resistance switching memory element 156917.doc • 74· 201212319 104 steering element 106 first conductor 108 second conductor 113 barrier layer 114 first memory level 116 single three-dimensional Array 118 first memory level 120 second memory level 130 resistance switching layer 132 electrode 133 conductive intermediate layer 134 electrode 135 resistance switching layer 142 replay miscellaneous n. + polycrystalline stone: 144 lightly doped or essential (non-146 heavy Complementary P+ polycrystalline 300 memory system 302 memory array 306 input/output 308 output 310 row control circuit 312 row decoder 314 array terminal receiver 316 block selection circuit 156917.doc -75- 201212319 320 column control circuit 322 Column Decoder 324 Array Terminal Driver 326 Block Selection Circuit 330 System Control Logic 400 Line 402 Line 404 Line 406 Line 420 Line 422 Line 424 Line 426 Line 430 Line 432 Line 434 Line 436 Line 440 Line 442 Line 444 Line 446 Line 450 Line 452 line 454 line 156917.doc •76· 201212319 456 line 547 words 549 Word Line 550 Memory Unit 552 Memory Unit 554 Memory Unit 556 Memory Unit 557 Bit Line 558 Transistor 559 Bit Line 560 Write Circuit 562 Transistor 563 Data Bus 564 Clamp Control Circuit 566 Sense Amplifier 568 data latch A1 arrow A2 arrow AL adhesive layer AL1 first adhesive layer AL2 second adhesive layer BLC bit line contact Cap 1 one cap layer Cap2 another cap layer 156917.doc -77- 201212319
El 第一電極 E2 第二電極 Ec 導帶之能階 EE1 E1之能階 EE2 E2之能階 EEL EL之能量 EER ER之能量 Eg 能帶間隙 Ei 本質能階 EIL IL之能量 EL 左手邊電極 ER 右手邊電極 Ev 價帶之能階 EV 價帶 IL 中間層 IL1 第一中間層 IL2 第二中間層 IresetA 重設電流 IresetB 重設電流 Iset_limitA 重設電流限制 Iset_limitB 重設電流限制 NC 非導電層 ςφΜ 功函數 RSL1 第一電阻切換層 156917.doc -78- 201212319 RSL2 第二電阻切換層 RSL3 第三電阻切換層 RSME 電阻切換記憶體元件 SE 操縱元件 tlx 厚度 tly 厚度 t2x 厚度 t2xa 厚度 t2xb 厚度 t2y 厚度 t3x 厚度 t3xa 厚度 t3xb 厚度 t3y 厚度 t4x 厚度 t4xa 厚度 t4xb 厚度 t4y 厚度 t5x 厚度 t5xa 厚度 t5xb 厚度 t5y 厚度 Vf 形成電壓 Vread 固定電壓 -79- 156917.doc 201212319El first electrode E2 second electrode Ec conduction band energy level EE1 E1 energy level EE2 E2 energy level EEL EL energy EER ER energy Eg energy band gap Ei essential energy level EIL IL energy EL left hand side electrode ER right hand Side electrode Ev valence band energy level EV valence band IL intermediate layer IL1 first intermediate layer IL2 second intermediate layer IresetA reset current IresetB reset current Iset_limitA reset current limit Iset_limitB reset current limit NC non-conductive layer ςφΜ work function RSL1 First resistance switching layer 156917.doc -78- 201212319 RSL2 second resistance switching layer RSL3 third resistance switching layer RSME resistance switching memory element SE steering element tlx thickness tly thickness t2x thickness t2xa thickness t2xb thickness t2y thickness t3x thickness t3xa thickness t3xb Thickness t3y Thickness t4x Thickness t4xa Thickness t4xb Thickness t4y Thickness t5x Thickness t5xa Thickness t5xb Thickness t5y Thickness Vf Forming voltage Vread Fixed voltage -79- 156917.doc 201212319
VresetVreset
VresetAVresetA
VresetBVresetB
VsetVset
VsetAVsetA
VsetBVsetB
WLC 重設電壓 重設電壓 重設電壓 設定電壓 設定電壓 設定電壓 字線觸點 -80- 156917.docWLC Reset Voltage Reset Voltage Reset Voltage Set Voltage Set Voltage Set Voltage Word Line Contact -80- 156917.doc
Claims (1)
Applications Claiming Priority (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US35632710P | 2010-06-18 | 2010-06-18 | |
US46793611P | 2011-03-25 | 2011-03-25 | |
US13/157,204 US8520424B2 (en) | 2010-06-18 | 2011-06-09 | Composition of memory cell with resistance-switching layers |
Publications (1)
Publication Number | Publication Date |
---|---|
TW201212319A true TW201212319A (en) | 2012-03-16 |
Family
ID=46763779
Family Applications (4)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW100121291A TW201212319A (en) | 2010-06-18 | 2011-06-17 | Composition of memory cell with resistance-switching layers |
TW100121289A TW201212318A (en) | 2010-06-18 | 2011-06-17 | Memory cell with resistance-switching layers and lateral arrangement |
TW100121287A TW201212317A (en) | 2010-06-18 | 2011-06-17 | Memory cell with resistance-switching layers |
TW100121288A TW201209824A (en) | 2010-06-18 | 2011-06-17 | Memory cell with resistance-switching layers including breakdown layer |
Family Applications After (3)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW100121289A TW201212318A (en) | 2010-06-18 | 2011-06-17 | Memory cell with resistance-switching layers and lateral arrangement |
TW100121287A TW201212317A (en) | 2010-06-18 | 2011-06-17 | Memory cell with resistance-switching layers |
TW100121288A TW201209824A (en) | 2010-06-18 | 2011-06-17 | Memory cell with resistance-switching layers including breakdown layer |
Country Status (1)
Country | Link |
---|---|
TW (4) | TW201212319A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI484679B (en) * | 2012-12-17 | 2015-05-11 | Winbond Electronics Corp | Non-volatile memory |
US9159918B2 (en) | 2013-01-08 | 2015-10-13 | National Tsing Hua University | Resistive random access memory |
TWI779482B (en) * | 2020-02-18 | 2022-10-01 | 美商應用材料股份有限公司 | Soft reset for multi-level programming of memory cells in non-von neumann architectures |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9978810B2 (en) | 2015-11-04 | 2018-05-22 | Micron Technology, Inc. | Three-dimensional memory apparatuses and methods of use |
US10134470B2 (en) | 2015-11-04 | 2018-11-20 | Micron Technology, Inc. | Apparatuses and methods including memory and operation of same |
US10446226B2 (en) | 2016-08-08 | 2019-10-15 | Micron Technology, Inc. | Apparatuses including multi-level memory cells and methods of operation of same |
US10157670B2 (en) | 2016-10-28 | 2018-12-18 | Micron Technology, Inc. | Apparatuses including memory cells and methods of operation of same |
CN111342424B (en) * | 2020-02-11 | 2021-12-07 | 常熟理工学院 | Circuit automatic protection device based on memristor |
-
2011
- 2011-06-17 TW TW100121291A patent/TW201212319A/en unknown
- 2011-06-17 TW TW100121289A patent/TW201212318A/en unknown
- 2011-06-17 TW TW100121287A patent/TW201212317A/en unknown
- 2011-06-17 TW TW100121288A patent/TW201209824A/en unknown
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI484679B (en) * | 2012-12-17 | 2015-05-11 | Winbond Electronics Corp | Non-volatile memory |
US9159918B2 (en) | 2013-01-08 | 2015-10-13 | National Tsing Hua University | Resistive random access memory |
TWI513074B (en) * | 2013-01-08 | 2015-12-11 | Nat Univ Tsing Hua | Resistive random access memory |
TWI779482B (en) * | 2020-02-18 | 2022-10-01 | 美商應用材料股份有限公司 | Soft reset for multi-level programming of memory cells in non-von neumann architectures |
US11790989B2 (en) | 2020-02-18 | 2023-10-17 | Applied Materials, Inc. | Soft reset for multi-level programming of memory cells in non-von neumann architectures |
Also Published As
Publication number | Publication date |
---|---|
TW201212317A (en) | 2012-03-16 |
TW201212318A (en) | 2012-03-16 |
TW201209824A (en) | 2012-03-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP2583323B1 (en) | Composition of memory cell with resistance-switching layers | |
US8724369B2 (en) | Composition of memory cell with resistance-switching layers | |
US8520425B2 (en) | Resistive random access memory with low current operation | |
US8693233B2 (en) | Re-writable resistance-switching memory with balanced series stack | |
US9437658B2 (en) | Fully isolated selector for memory device | |
TW201212319A (en) | Composition of memory cell with resistance-switching layers | |
US9012879B2 (en) | Morphology control of ultra-thin MeOx layer |