I301M 16402twf.doc/006 九、發明說明: 【發明所屬之技術領域】 置,且特別是 本發明疋有關於-種非揮發性記憶體装 有關於一種電荷捕捉快閃記憶體。 【先前技術】 基於被稱為電可抹除可程式化唯讀記憶體 (EEPROM)和快閃記憶體的電荷儲存結構的電可程式化 =可抹除非揮發性儲存技術用於多種現代應用。快閃 體以可獨立程式化和讀取的記憶料元_崎計。快閃 =體^錢放大器驗確定儲存在非揮發性記憶體中 的-個或夕個㈣值。在典型的感應方案中,電流感應放 ^器將感應到的通過記憶體單福電流與參考電流相比 較0 荖二I""結構用於職〇M和快閃記憶體。隨 =貝=路的尺寸縮小,由於製造過程的可縮放性和簡單 =荷敵介電層的記憶體單元結 大。基於電荷捕捉介電層的記憶體單元結構包括 ^I301M 16402twf.doc/006 IX. Description of the Invention: [Technical Field of the Invention] In particular, the present invention relates to a non-volatile memory package relating to a charge trapping flash memory. [Prior Art] Electrically programmable based on a charge storage structure known as electrically erasable programmable read only memory (EEPROM) and flash memory = erasable unless volatile storage technology is used in many modern applications. The flash body is a memory element that can be independently programmed and read. Flash = Body = Money Amplifier to determine the value of - or eve (four) stored in non-volatile memory. In a typical sensing scheme, the current sense transducer will sense the current through the memory and the reference current compared to the reference current. The I" structure is used for the M and flash memory. The size of the channel is reduced with the scalability of the manufacturing process and the simplicity of the memory cell of the enemy dielectric layer. The memory cell structure based on the charge trapping dielectric layer includes ^
^ )Nitride Read-Only Memory > SONGS ^ PHINES 的結構。記M單元結構通過在電荷她介電層(例 ° , 4體早福賴輕增加。觸 移除來降低記賴單元响始·。做讀捕捉層 氮化物唯讀記憶體裝置使用相對較厚 例如大於3奈米且通常為約5到9奈米,來防止電^’。 1301¾ 以 5 16402twf.doc/006 代替直接穿随,帶對帶穿随引發的熱電洞注入(btbthh) =二:述單元。然而’熱電洞注入導致氧化物傷害, ,,致南起始單元中的電荷難和低起始單元中的 增扭。此外’歸因於電荷捕捉結構中難以抹除的電荷^ ^在程式化和抹除迴圈期間必須逐漸增加抹除時間)因貝 二子注入點彼此不重合’且在抹除脈衝後 二電子退打來’所以發生此電荷積累。另外 物唯讀記憶體快閃記憶體裝置的磁區 = 變化(例如溝道長度變化),每個單元的抹除速= 2。抹t速度中的此差異導致抹除狀態的較大vt分佈,i -些早讀得難輯除且—些單元被過度抹除。因此了 在許多程式化和抹除迴圈後,目標起始v t視窗_, 。隨著所述技術繼續按比例縮小,此現象將變得更 ^型的氮化物唯讀記憶體㈣記憶體單元結構將 物-氮化物-氧化物層安置於導電多晶石夕與晶體石夕半 底之間。所述襯底制是由基礎溝道區域分關的源極區 。可通過汲極感應或源極感應來執行快閃記 體早兀頃取。對於源極側感應來說,一個或一個以上 極線麵合耽憶體單元的源極區域,以讀 2 列中的特定記《單元的錢。 u體陣 傳統的浮動閘極裝置的每個單元只儲存丨位元,但 現了氮化物唯讀記憶體單元,其中每織化物唯讀記情 單元的每個快閃單元提供2位元,所述快閃單元將電相 16402twf.doc/006 ΐΓΐίΓΓ 結構中,氮化物層用作安置於頂部 二:動層之間的捕捉材料。⑽0層結構有效地 置中的閉極電介質。可在氮化物唯讀記憶 中的ϊ荷。貝^右側上捕捉具有氮化物層的〇n〇電介質 外s於程式域化物唯讀記龍陣财軌化物唯讀 早_技術為溝道熱電子注人法。在抹除操作期 a #於抹除,體單元的常見技術稱為帶對帶熱電洞注 氮化物唯㈤5仏體單元的與被抹除的側不同的側的 除能力上具有橫向電場效應。估計氮化物唯讀 體陣列的耐久性和保持性,抹除能力缺乏均勾性導致 ^ 口於圈和烘培的餘量損耗。氮化物唯讀記憶體單元的 ^側為浮動(或連接到地面),其可麵合到不確定電壓位 2例:1伏特或4伏特)’這導致陣列單元的抹除起始的 ^ 又導致抹除操作後的Vt分佈更寬。不確定電壓位 ^的變化可導致過度抹除。另一方面,如果另一側連接到 “面,那麼當位線偏壓超過穿通電壓時,穿通可導致幫浦 電路(pumpeifeuit)崩潰。因此,在區塊的抹除操作期間, 氮化物唯讀記憶體單元(其中一些節點為左浮動)可導致 ,用於抹除氮化物唯讀記憶體陣列中的氮化物唯讀 單元的電壓位準缺乏均勻性。 3快閃記憶體裝置中的技術進步,需要設計提供更高封 裝密度和更優裝置可縮放性的電荷捕捉快閃記憶體單=結 130職 I6402twf.doc/006 構 【發明内容】 本發明描述用於製造具有多個兩位元電荷捕捉單元 的電荷捕捉快閃S己憶體的方法,其中每個記憶體單元形成 有石夕翼片(Si-FIN)層’所述&侧層具有反轉的源極和 沒極區域。-行記憶體單元中的所選擇的多晶石夕問極的每 個鄰近多晶石夕閘極用於產生反轉區域,充當源極區域或沒 極區域’用於傳送所需的電壓,其在每個記憶體單元的源 極區域和汲極區域均未摻雜的情況下保存記憶體單元的密 度。所述快閃記憶體包括與多個Si_Fm層交叉的多個多曰^ 矽層:當在-行記憶體單元中選擇一記憶體單元以用於執 二釭式,作、抹除操作或讀取操作時 單元的左側上的第一組記憶體單 ^伴扪°己匕體 憶體單元傳遞第-電_傳遞_,; =右第二組記憶體單元充當用= 體早兀傳遞弟—電壓的傳遞閘極。 〜 穿隨(FN)抹除P極物富雷·諾特海姆式 式化和對所選擇的門=擇的閉極進行熱電洞(HH)程 中,所述方>去包括對其中每個辦一在弟—操作方法 域均未摻_電荷捕捉記憶體陣=以的源極和沒極區 選擇開極進行溝道熱電子(c 下,作:針對所 私式化、針對所選擇的 8 1301 概 16402twf.doc/006 閘極進行熱電洞抹除和 對於每個操作步驟來說,選摆斤;:擇,極進行讀取操作。 用於執行程式化、抹除或=二中的特定記憶體單元以 單元充當用=傳遞需要電壓的傳遞=仃中的其他記憶體 體單元,具有閘極電極、未摻 電何捕捉記憶 的第-側而安置,所述二單鄰記憶體單元 摻雜的源極區域和未摻雜的二1有,極、未 體單元,鄰近第-記憶體單元的第何捕捉, 記憶體單元具有閘極電極、未 所述第二 汲極區域;和Si_FIN層,越過第域和未摻雜的 第二電荷概記題單元和第捉記鐘單元、 延伸。 乐—窀何捕捉記憶體單元正交 古;4Γ =發明產生更緊凑的電荷捕捉記憶體,与 同度縮放’因為每個單元的源極 ::: ::且每個單元均不保留或佔用用於源極和== 下。 所附圖式,作詳細說明如 【實施方式】 現參看圖1A,其展示說明具有 麵0Μ快閃單元結構⑽的結構圖。所述結== 15 16402twf.doc/006 多晶矽(或多晶矽閘極)層no,其覆蓋氧化物-氮化物-氧化物(ΟΝΟ)薄膜120,源極區域130在左側且汲極區 域140在右側。所述ΟΝΟ薄膜120為第一位在左側且第 二位在右侧的兩位元操作提供電荷捕捉結構。Si-FIN層 150在垂直於多晶石夕層110的方向上延伸。源極區域130 和沒極區域140未換雜,即源極區域130和没極區域140 中無植入物。確切地,結構100使用鄰近的多晶矽閘極11〇 來產生反轉區域以充當用於傳遞所需電壓的源極/汲極區 ® 域。將所述源極/汲極區域限制到較小區域以進行反轉,這 導致較小單元尺寸和較高可縮放性。^ )Nitride Read-Only Memory > SONGS ^ PHINES structure. Record the M-cell structure by charging her dielectric layer (eg °, 4 body early blessing light increase. Touch removal to reduce the counting unit sing. · Read the capture layer nitride read-only memory device is relatively thick For example, greater than 3 nanometers and usually about 5 to 9 nanometers to prevent electricity. 13013⁄4 replaces direct wear with 5 16402twf.doc/006, with hot hole injection (btbthh) = two: Said unit. However, 'hot hole injection causes oxide damage, , and the charge in the south start unit is difficult and the increase in the low start unit. In addition, 'attributable to the charge trapped in the charge trapping structure ^ ^ The erasing time must be gradually increased during the stylization and erasing of the loop.) Since the two injection points do not coincide with each other and the two electrons are repelled after the erasing pulse, this charge accumulation occurs. In addition, the magnetic area of the memory-only flash memory device = change (for example, channel length variation), and the erase speed of each cell = 2. This difference in wipe speed results in a large vt distribution of the erased state, i - some early read hard to erase and some cells are over erased. So after many stylization and erase loops, the target starts v t window _, . As the technology continues to scale down, this phenomenon will become more of a nitride-only read-only memory. (IV) The memory cell structure places the material-nitride-oxide layer on the conductive polycrystalline stone and the crystal stone. Between the half bottom. The substrate is a source region that is separated by a base channel region. The flash recording can be performed by bungee sensing or source sensing. For source side sensing, one or more poles face the source area of the unit, to read the specific unit in the 2 columns. Each unit of the conventional floating gate device of the u-body array only stores the 丨 bit, but now a nitride-only memory unit, in which each flash unit of each woven material read-only unit provides 2 bits. The flash cell has an electrical phase 16402 twf.doc/006 ΐΓΐ ΓΓ ΓΓ structure in which the nitride layer is used as a capture material disposed between the top two: moving layers. (10) A closed-pole dielectric in which the 0-layer structure is effectively placed. Can be used in nitride read-only memory. On the right side of the shell, the 〇n〇 dielectric with a nitride layer is captured. The outer surface of the 程式n于 程式 程式 程式 程式 程式 唯 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ In the erasing operation period a #在抹除, the common technique of the body unit is called the paired-band thermoelectric hole. The nitride-only (5) 5-body unit has a lateral electric field effect on the side different from the erased side. It is estimated that the durability and retention of the nitride read-only array, the lack of wiping ability, leads to the loss of the ring and baking. The side of the nitride-reading memory unit is floating (or connected to the ground), which can be folded to an indeterminate voltage level of 2 cases: 1 volt or 4 volts) 'This leads to the erasing of the array unit ^ This results in a wider Vt distribution after the erase operation. Uncertain changes in the voltage level ^ can result in excessive erasure. On the other hand, if the other side is connected to the "face, then when the bit line bias exceeds the punch-through voltage, the punch-through can cause the pump circuit to collapse. Therefore, during the erase operation of the block, the nitride is read only. Memory cells (some of which are left floating) can result in a lack of uniformity in the voltage level used to erase the nitride-reading cells in the nitride read-only memory array. 3 Technological advances in flash memory devices Need to design a charge trapping flash memory that provides higher packing density and better device scalability. The invention is described for manufacturing a plurality of two-element charges. A method of capturing a charge of a snap cell of a flash memory, wherein each memory cell is formed with a Si-FIN layer, the & side layer having an inverted source and a non-polar region. - each adjacent polylithic gate of the selected polycrystalline silicon in the row memory cell is used to generate an inversion region, acting as a source region or a non-polar region 'for transmitting a required voltage, In each memory The density of the memory cell is preserved in the case where the source region and the drain region of the cell are undoped. The flash memory includes a plurality of multiple 曰 矽 layers crossing a plurality of Si_Fm layers: when in-line memory Selecting a memory unit in the body unit for performing the second type, the first group of memory on the left side of the unit during the erasing, erasing operation or reading operation, and the first unit of the unit Electricity_transfer_,; = right second group of memory cells acts as the transfer gate of the voltage-transfer with the body-body. ~ wear-follow (FN) erase P-pole Fule Noltheim style And in the thermoelectric hole (HH) process for the closed gate of the selected gate=selection, the side> includes including each of the ones in the operating method domain, the unloaded_charge trapping memory array= The source and the immersion region are selected to open the channel for the hot electrons (c: for the privateization, for the selected 8 1301, the 16402twf.doc/006 gate for the hot hole erase and for each In terms of the operation steps, select the pendulum; select, the pole performs the reading operation. It is used to perform the specific memory of stylization, erasure or = two The body unit acts as a unit for transmitting the required voltage = other memory unit in the ,, with the gate electrode, the first side of the uncharged and captured memory, the two single adjacent memory unit doping The source region and the undoped two have, the pole, the non-body unit, the adjacent capture of the first memory cell, the memory cell has a gate electrode, the second drain region is not; and the Si_FIN layer Crossing the first and undoped second charge enumeration unit and the first capture unit, extending. Le-his memory memory cell orthogonal; 4Γ = invention produces a more compact charge trapping memory, Scale with the same degree 'because the source of each cell is ::::: and each cell is not reserved or occupied for the source and == under. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Referring now to Figure 1A, there is shown a block diagram showing a structure (10) having a facet flash cell structure. The junction == 15 16402 twf.doc / 006 polycrystalline germanium (or polysilicon gate) layer no, which covers the oxide-nitride-oxide (germanium) film 120, with the source region 130 on the left side and the drain region 140 on the right side . The tantalum film 120 provides a charge trapping structure for a two bit operation in which the first bit is on the left side and the second bit is on the right side. The Si-FIN layer 150 extends in a direction perpendicular to the polycrystalline layer 110. The source region 130 and the gate region 140 are not interchanged, that is, there are no implants in the source region 130 and the gate region 140. Specifically, structure 100 uses adjacent polysilicon gates 11A to create an inversion region to serve as a source/drain region ® domain for delivering the desired voltage. The source/drain regions are limited to smaller regions for inversion, which results in smaller cell sizes and higher scalability.
如圖1B中所示,其是說明記憶體單元的陣列16〇的 佈局圖,其中多晶矽層與Si-FIN層交又安置。將多數行的 多晶矽層170、171、172和173安置成大體上相對於多數 列的 Si-FIN 層 180、181、182、183 和 184 正交。所述 si_piN 層中的母一個均越過多個多晶石夕層延伸。舉例來說, 180層越過多晶矽層170、多晶矽層171、多晶矽層172和 • 多晶石夕層173而延伸。圖1B中展示圖1A中的i極區域 140,其中多晶石夕層170的一部分與Si_FlN^ 18 分交叉。 電荷捕捉快閃單元結構100能夠執行兩位元操作,立 中說明性第-類型涉及FN抹除和熱電洞程式化',且說明 及Γ洞抹除和溝道熱電子程式化。可以兩 種:法,具有FN抹除和熱電洞程式化的所述類型的 插作。在弟-方法中,將電荷捕捉單元的兩側上的兩個位 I301U2 15 16402twf.doc/006 =職狀態(FN穿隨,電子在氮化物薄臈 Ϊ元電壓施力,所選擇的“捕捉 將約零伏特和約i 5伏==、具有高Vt的右側。當分別 將電荷捕捉單元的兩側中 :在弟-方法中’ 電洞注入來對氮化物薄膜(即,。使用熱 行程式化。當分別將約! ^=捉_)的右側進 所選擇的單元的最左端和最右端^:伏特的賴施加到 式化側。當分別將約零伏特=取具有低Vt的被程 選擇的單元的最左端和最右日士,诗大特的一電壓施加到所 式化側。 π σ貝取具有尚Vt的未被程 程式行;;熱=:除和溝道熱電子 元的兩侧上的兩個位抹㈣t;;方將電荷捕极單 和約零伏特的電壓施加到所選;一田^將約15伏特 時,讀取具有低Vt^右者八、早70的取左端和最右端 特的電壓施加到所選擇一田刀別將約零伏特和約1.5伏 ;有…左;::擇第r方r中最::和 子注入來對氮化物薄膜 ]如,如果以溝道熱電 捉單元的兩側中的一側程:進:程式:化,那麼將電荷捕 伏特和約零伏特的 ^到回狀恶。當分別將約1.5 右端時,讀取具有 & 口舍所選擇的單元的最左端和最 幻所_的電荷敝單元的最左 >015 16402twf.doc/006 端和最右端時,讀取具有低Vt的未被程式化侧。 在圖2A至圖2C中,展示分別說明電荷捕捉快閃記憶 體陣列160的抹除操作200、程式化操作240和讀取操作 260中的第一操作方法的電路圖。在此第一操作方法中, 以下面的方法來執行所述方法:使用富雷-諾特海姆式穿隧 (Fowler-Nordheim tunneling)來抹除記憶體單元;使用熱 黾洞’主入來對所述記憶體單元進行程式化;和讀取所述記 憶體單元。如圖2A中所示,如果要抹除閘極203,那麼剩 下的閘極單元,即閘極201、閘極202、閘極204、閘極205、 閘極206和閘極207均充當用於傳送電壓的傳遞閘極。本 說明書中的術語“閘極,,還稱為“閘極單元,,,或“電荷 捕捉圮憶體單元中的閘極”。將閘極單元,即閘極2〇1、 閘極202、閘極204、閘極2〇5、閘極2〇6和閘極2〇7充電 f第二電壓位準,例如10伏特,而將閘極2〇3單元充電到 咼達第二電壓水準,例如20伏特,以使用FN穿隧來抹除 閘極203單元。Si_FIN層21〇從閘極2〇1的左側一直延伸 到閘極207的右側;延伸穿過閘極2〇1、閘極搬、問極 203、閘極204、閘極205、閉極206和閘極2〇7。閘極2〇1、 閘極20j、間極203、間極2〇4、間極2〇5、閉極施和間 極207 5又置在氧化物層22〇 ±。石夕23〇鄰近閉極術的右 側深度摻雜,且設置在氧化物層22〇的上面。 在抹除操作後,順序進行到程式化操作240,使用孰 電财人料電荷雜記憶料元進躲式化,如圖2B 中m閘極2〇3為具有兩個位的氮化物唯讀記憶體單 12 1301319 P940015 16402twf.doc/006 元’其中第一位在左侧上且第二位在右側上。將使用熱電 洞注入對閘極203進行的程式化應用到閘極203的右侧 203-r。相對於閘極203的左側2034,其設定為約零伏特, 閘極203的右側203-r具有較高電壓電位,其設定為約5 伏特。右側203-r與左側203-1之間的5伏特的電壓差在右 側上產生熱電洞情況,這導致與閘極2〇3相關的氮化物層 中的儲存。As shown in Fig. 1B, it is a layout diagram illustrating an array 16 of memory cells in which a polysilicon layer is placed in contact with a Si-FIN layer. The majority of the rows of polysilicon layers 170, 171, 172, and 173 are disposed substantially orthogonal to the plurality of columns of Si-FIN layers 180, 181, 182, 183, and 184. One of the mothers in the si_piN layer extends over a plurality of polycrystalline layers. For example, the 180 layers extend over the polysilicon layer 170, the polysilicon layer 171, the polysilicon layer 172, and the polycrystalline layer 173. The i-pole region 140 of Figure 1A is shown in Figure 1B, wherein a portion of the polycrystalline layer 170 intersects the Si_FlN^18. The charge trapping flash cell structure 100 is capable of performing a two bit operation, and the illustrative first type relates to FN erasing and thermoelectric hole staging, and illustrates the cavity erasing and channel hot electron stylization. There are two types: the method, the type of interpolation with FN erasure and thermoelectric hole stylization. In the brother-method, two bits on both sides of the charge trapping unit are I301U2 15 16402twf.doc/006 = status (FN wears, electrons are applied at the nitride thin cell voltage, the selected "capture" Will be about zero volts and about i 5 volts ==, with the high side of the high Vt. When the two sides of the charge trapping unit are respectively: in the method - the hole is injected into the nitride film (ie, using the hot line) Stylized. When the right side of the ~^= catching_) is respectively entered into the leftmost and rightmost of the selected unit, the volt: is applied to the side of the formula. When respectively, about zero volts = the one with the low Vt The leftmost and rightmost Japanese of the unit selected by the process, a voltage of Shida special is applied to the side of the formula. π σ is taken from the unprocessed line with Vt; heat =: and the channel hot electron element The two bits on both sides of the wiper (four) t;; square charge charge single and about zero volts voltage applied to the selected; a field ^ will be about 15 volts, read with low Vt ^ right eight, early 70 The voltage applied to the left end and the far right end is applied to the selected field knife and will be about zero volts and about 1.5 volts; there is ... left;:: the most r: r and r: Into the nitride film] If, for example, in the channel thermoelectric capture unit on one side of the unit: into: program: then, the charge is trapped and about zero volts to the return of the evil. When respectively At about the right end of the 1.5, when reading the leftmost & most significant _ of the cell selected by the & port, the leftmost >015 16402twf.doc/006 end and the rightmost end, the read has a low Vt The unprogrammed side. In Figures 2A-2C, circuit diagrams illustrating the first method of operation of the erase operation 200, the stylization operation 240, and the read operation 260 of the charge trapping flash memory array 160, respectively, are shown. In this first method of operation, the method is performed in the following manner: using Fowler-Nordheim tunneling to erase memory cells; using a hot hole to enter The memory unit is programmed; and the memory unit is read. As shown in FIG. 2A, if the gate 203 is to be erased, the remaining gate unit, that is, the gate 201 and the gate 202 , gate 204, gate 205, gate 206 and gate 207 all serve to transmit electricity The transfer gate of the voltage. The term "gate" in this specification is also referred to as "gate cell,", or "gate in charge trapping unit". The gate unit, ie the gate 2〇 1. Gate 202, gate 204, gate 2〇5, gate 2〇6, and gate 2〇7 charge f the second voltage level, for example 10 volts, and charge the gate 2〇3 unit to 咼A second voltage level, such as 20 volts, is used to erase the gate 203 cell using FN tunneling. The Si_FIN layer 21 一直 extends from the left side of the gate 2〇1 to the right side of the gate 207; extends through the gate 2 〇 1, gate pole, pole 203, gate 204, gate 205, pole 206 and gate 2〇7. The gate 2〇1, the gate 20j, the interpole 203, the interpole 2〇4, the interpole 2〇5, the closed electrode and the interpole 207 5 are again placed on the oxide layer 22〇±. Shi Xi 23〇 is deeply doped adjacent to the right side of the closed-pole technique and is disposed above the oxide layer 22〇. After the erase operation, the sequence proceeds to the stylization operation 240, using the 孰 财 财 人 电荷 电荷 电荷 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , Memory single 12 1301319 P940015 16402twf.doc/006 yuan 'where the first position is on the left side and the second position is on the right side. The stylization of the gate 203 using thermowell injection is applied to the right side 203-r of the gate 203. With respect to the left side 2034 of the gate 203, it is set to about zero volts, and the right side 203-r of the gate 203 has a higher voltage potential, which is set to about 5 volts. A voltage difference of 5 volts between the right side 203-r and the left side 203-1 produces a hot hole condition on the right side, which results in storage in the nitride layer associated with the gate 2〇3.
或者,可將使用熱電洞注入對閘極2〇3進行的程式化 應用到閘極203的左側203-1。相對於閘極2〇3的右側 203-r ’其没疋為約零伏特,閘極203的左側203—1具有較 间電壓電位,其設定為約5伏特。左側203-1與右側2〇3-r 之間的5伏特的電壓差在左側上產生熱電洞,這導致與閘 極203單元相關的氮化物層中的電荷儲存。因此,可將使 用熱電洞注入對閘極203進行的程式化應用到閘極2〇3的 右側203-r或左側203-1。Alternatively, the stylization of the gate 2〇3 using the thermal hole injection can be applied to the left side 203-1 of the gate 203. The right side 203-r' of the gate 2〇3 is not about zero volts, and the left side 203-1 of the gate 203 has a relatively high voltage potential, which is set to about 5 volts. A voltage difference of 5 volts between the left side 203-1 and the right side 2〇3-r creates a thermoelectric hole on the left side, which results in charge storage in the nitride layer associated with the gate 203 cell. Therefore, the stylization of the gate 203 using the hot hole injection can be applied to the right side 203-r or the left side 203-1 of the gate 2〇3.
在程式化操作後,順序進行到讀取操作260以讀取閘 f 203單元。電荷翻在閘極203的右侧2〇3_r上,使得 讀取操作在閘極203的左側咖上進行。隨著時間的變 化’起始電壓Vt將趨向於飄移到更高。儘管上 :_二在問極加的右側咖上且讀取操作在閘極 早7^左側2〇3_1上進行,但是所屬領域的—般技術人 ’電荷射儲存在左側上,而讀取操作在右側 當選擇間極203以執行抹除操作、程式化操作或讀取 13 I3〇13p1215 16402twf.doc/006 操作時,剩下的相鄰單元,即間極2〇卜間極2〇2 旭、閘極2〇5、間極施和閘極2〇7充當傳遞閑極。样 ,的-個優勢在於其通過將相鄰單元用作傳遞雜,科 疋通過要求與每個,己憶體單元相關的附加閉極來作為傳遞 閉極而操作的常規解決方法,來減少記憶體陣列中的電路 =數量。舉例來說,如果將閘極2〇5選擇為要抹除,那麼 維20卜閘極202、閘極2〇3、閘極綱、閘極施和間 極207充當用於將㈣傳遞到閘極2〇5的傳遞間極。 圖3A至圖冗是制電荷捕捉賴記鐘陣列的程式 ΐ除和讀取中的第二操作方法的電路圖。在第二摔作 =二,是程式化記憶體單元、抹除所述記憶體單元 3〇Γ= 如圖3Α中所示,如果選擇閘極 3〇3以使用溝道熱電子(CHE)來進行程式化操作_ 麼閘極303的糊303_Γ約為5伏特,而問極3〇 3= 約為零伏特。在閉極3〇3的左側上,間極3q =After the stylization operation, the sequence proceeds to read operation 260 to read the gate f 203 unit. The charge is turned on the right side 2〇3_r of the gate 203, so that the reading operation is performed on the left side of the gate 203. As time changes, the starting voltage Vt will tend to drift higher. Although the above: _ 2 is on the right side of the question and the reading operation is performed on the left side of the gate 7^2_3_1, but the general-purpose person in the field 'charges the charge on the left side, and the read operation When the interpole 203 is selected on the right side to perform an erase operation, a program operation, or a read 13 I3〇13p1215 16402twf.doc/006 operation, the remaining adjacent units, that is, the interpole 2 inter-electrode 2〇2 Asahi The gate 2〇5, the interpole and the gate 2〇7 serve as the transfer idle. The advantage of this is that it reduces the memory by using adjacent cells as transfer memes, which is a conventional solution to the operation of transmitting the closed pole by requiring additional closed poles associated with each of the memory cells. Circuit in the body array = number. For example, if the gate 2〇5 is selected to be erased, the dimension 20 gate 202, the gate 2〇3, the gate, the gate, and the interpole 207 serve to pass (4) to the gate. The pole of the pole 2〇5. Fig. 3A to Fig. 3A are circuit diagrams of a second operation method in the process of erasing and reading the charge trapping array. In the second fall = two, is the stylized memory unit, erase the memory unit 3 〇Γ = as shown in Figure 3, if the gate 3 〇 3 is selected to use channel hot electrons (CHE) For the stylization operation, the paste 303_Γ of the gate 303 is about 5 volts, and the polarity is 3 〇 3 = about zero volts. On the left side of the closed-end 3〇3, the interpole 3q =
Lt 電峨到間極303的左側的傳遞間極。 =則將零伏特傳遞到閘極规,閘極302將零伏特傳 遞到閘極303的左側购。在閘極3〇3的右側上,問極 、閘極305、閘極3〇6和閘極3〇7充當用於 到間極303的右側的傳遞間極 伏 :極鄕’問極3。6將5伏特傳遞到間極二 將5伏特傳遞到閘極3〇4 305 3〇3的右側303_Γ 閑極304將5伏特傳遞到間極 在將零伏特傳遞到閘極期的左側303-i且將5伏特 i3〇im 16402twf.doc/006 傳遞到閘極303的右側303-r後,產生溝道熱電子。來自 左側303-1的零伏特使電子朝約為5伏特的右側3〇3_r加 速,且歸因於施加了 7伏特的多晶矽閘極3〇3的較高電位, 稱為熱電子的加速的電子被注入到氮化物層(〇Ν〇薄膜 120)中。接著通過溝道熱電子注入來對閘極3〇3的右側 303-r進行程式化。 圖3B是說明電荷捕捉快閃記憶體陣列16〇的通過熱 黾洞注入的抹除操作330的電路圖。如果選擇閘極來 抹除,那麼剩下的閘極單元,即閘極3〇卜閘極3〇2、閘極 304、閘極305、閘極306和閘極307充當用於傳送電壓的 傳遞閘極。將閘極單元,即閘極3〇卜閘極3〇2、閘極3〇4、 閘極305、和閘極307充電到第一電壓位準,即1〇伏特, 而以閘極303的右側303-r上的-5伏特通過熱電洞注入閘 極303。SkFIN 310層從閘極301的左側一直延伸到閘極 307的右側’包括延伸穿過閘極3〇1、閘極、閘極303、 間極304、閘極305、閘極306和閘極307。將閘極3〇1、 問極302、閘極303-、閘極304、閘極3〇5、閘極3〇6和閘 極307設置在氧化物層32〇上。矽翼片(smc〇n Fin) 33〇 深度摻雜在閘極307的右側和閘極301的左侧上。將矽翼 片330設置在氧化物層32〇的上面。 、 在抹除操作後,順序進行到讀取操作36〇以讀取閘極 3〇3。電荷儲存在閘極303的右側3〇3_r上,使得讀取操作 在閘極303的左側303-1上進行。隨著時間的變化,起始 私壓Vt將趨向於飄移到更高。儘管上文的說明展示電荷儲 15 13〇13故15 16402twf.doc/006 存在閘極303的右側303_r上,且讀取操作在閘極3〇3的 4 J 303 1上進行,但是所屬領域的一般技術人員應認識 到,,何,可儲存在左側上,而讀取操作在右側上進行。 第一插作方法200與第二操作方法3〇〇之間的一個差 異在於,低起始電壓與高起始電壓之間的相反關係。第一 • 操作綠200使用FN ?隨來抹除記憶體單元,這將記憶 • 料元抹_高起始。第二操作方法使用熱電洞 触來抹除記憶體單元,這將記憶體單元抹除到 壓。 一 圖4是說明執行抹除操作2〇〇、程式化操作24〇和讀 取操作260電荷捕捉快閃記憶體陣列16〇的第一操作方法 4〇〇的過程的流程圖。在步驟41〇處,第一操作方法4〇〇 在行中選擇閘極2〇3以通過使用冑雷_諾特海姆式穿隨來 抹除,而其他閘極,即閘極2〇卜閘極2〇2、閘極2〇4、閘 極205、閘極206和閘極207充當用於傳送電壓的傳遞閘 極。將未被選擇的閘極,即閘極2〇卜閘極2〇2、閘極2〇4、 _ 閘極205、閘極206和閘極207充電到第一電壓位準,例 如10伏特,而將閘極4 204充電到高達第二電壓位準,例 如20伏特,以通過使用fn穿隧來抹除閘極203。 在步驟420處,第一操作方法4〇〇在電荷捕捉快閃記 憶體陣列160中使用熱電洞注入來對閘極2〇3進行程式 化。對閘極203進行的程式化使用熱電洞注入,其被應用 到閘極203的右側203-r。相對於閘極203的左側203-1, 其設定為約零伏特,閘極203的右側203-r具有較高電麈 16 13〇13狄 16402twf.doc/006 粒’其設定域5伏特。如果選擇所相極,那 的-5伏特施加到間極203。右側2〇3_r與頂側閑極^之 =的10伏特的電壓差在右側2,上產生熱電洞注入,這 導致與閘極203相關的氮化物層中的儲存。 在步驟430中’第一操作方法_包括在電荷捕捉快 ^己憶體陣列160中執行對閘極2〇3進行的讀取摔作。電 閘極2〇3的右側加·,上,使得讀取操作在問極 2〇的左側购上進行。當閘極2〇3被抹除時,隨著時間 的殳化,起始電壓Vt將趨向於飄移到更高。儘管上文的說 明展示電荷儲存在閘極203的右側203-r上,且讀取操作 在閘,203的左側2〇3_丨上進行,但是所屬領域=一般技 術人員應認識到,電荷可儲存在左侧上,而讀取操作在右 j上進^ 圖5是說明進行程式化操作3〇〇、抹除操作和讀 取操作360電荷捕捉快閃記憶體陣列16〇的操作的第二操 作方法500的過程的流程圖。在步驟51〇處,第二操作方 法500包括使用溝道熱電子技術來對所選擇的間=二3進 行程式化。閘極303的右側3〇3-r約為5伏特,而閘極3〇3 的左侧303-1約為零伏特。在閘極3〇3的左側上,閘極3〇i 和閘極302充當用於將電壓傳遞到閘極3〇3的左侧的傳遞 閘極。閘極301將零伏特傳遞到閘極3〇2,閘極3〇2將零 伏特傳遞到閘極303的左側3〇3-卜在閘極303的右侧上, 閘極304、閘極305、閘極306和閘極307充當用於將程式 化電壓傳遞到閘極303的右側的傳遞閘極。閘極307將5 17 I3013A1 16402twf.doc/006 伏特傳遞到閘極306、閘極306將5伏特傳遞到閘極3〇5, 閘極305將5伏特傳遞到閘極304,閘極3〇4將5伏特傳 遞到閘極303 #右側303_r。在將零伏特傳遞到問極3〇3 的左側303-1且將5伏特傳遞到閘極3〇3的右側3〇3_γ後, 產生溝道熱電子情況。來自左侧期-1的零伏特使電子朝 • 、約為5伏特的右側303_r加速,且由於施加了 7伏特的多 .曰曰曰石夕閘極303的較高電位,將稱為熱電子的加速的電子注 • 彡到氮化物層(ΟΝΟ薄膜12G)中。接著通過溝道熱電子 注入來對閘極3〇3的右側303-r進行程式化。 在步驟520處,第二操作方法5〇〇包括執行電荷捕捉 ^閃記憶體陣列3GG的通過熱電洞注人的抹除操作。如果 選擇閘極303來抹除,那麼剩下的閘極單元,即閘極3〇1、 閘極302、閘極304、閘極3〇5、閘極3〇6和閘極3们充當 用於將電壓傳送到閘極3〇3的傳遞閘極。將閘極單元,^ 閘極=1、間極302、閘極304、閘極305、閘極7 3〇7充 朗S電壓位準,例如1〇伏特,而通過使用注 • S.PIN ^ 31〇 303 :上的5伏特來抹除閘極3〇3,且將_5伏特施加到閘極 的53Q中’第二操作方法5()()包括執行閑極303 ^取㈣。電荷儲存在閘極期的右侧地_ ΓΓ在閘極303的左側綱上進行。隨著時間= 已始電壓Vt將趨向於飄移到更高。 雖然本發明已以實施例揭露如上,然其並非用以限定 18 13013 αΐ5 16402twf.doc/006 本發明,任何熟習此技蓺者, 圍内,當可作些許之更;與潤飾,因本=之精神和範 當視後附之申請專·_界定者鱗本U之保護範圍 【圖式簡單說明】 圖1AS說明根據本發明的—個實施例的 的電荷捕捉快閃單元的結構圖。 ” 1 圖1B是說明根據本發明的一個實施例 的電荷捕捉糊單元的佈局圖。 八有 圖2A-2C是說明根據本發明的一個實施例的 =和讀取電荷捕捉快閃記憶體陣列的第一操作方法的電 枯說明根據本發明的一個實施例的程式化、 未除和項取電何捕捉快閃記憶斜列的第二操作方法的電 路圖。 圖4是根據本發明的一個實施例的執行第一操作方法 的過程的流程圖。 / 圖5疋根據本發明的一個實施例的執行第二操作方法 的過程的流程圖。 μ / 【主要元件符號說明】 100 ·•結構 110、170、171、172、173 :多晶矽層 120:氧化物-氮化物-氧化物薄膜 130 :源極區域 140 :汲極區域 19 B〇i3ja5 16402twf.doc/006 150、180、18卜 182、183、184、210 : Si-:FIN 層 160 :陣列 200、330 :抹除操作 2(Π、202、203、204、205、206、207、301、302、 303、304、305、306、307 :閘極 203-1、303-1 :左側 203-r、303-r :右侧 220 :氧化物層 230 :矽 240、300 :程式化操作 260、360 :讀取操作 400 :第一操作方法 410〜430 :步驟 500 :第二操作方法 510〜530 ··步驟The Lt is electrically connected to the transfer pole on the left side of the interpole 303. = then zero volts is passed to the gate gauge, and gate 302 passes zero volts to the left side of gate 303. On the right side of the gate 3〇3, the gate, the gate 305, the gate 3〇6, and the gate 3〇7 serve as the transfer volts for the right side of the interpole 303: the pole 问 'question pole 3. 6 Pass 5 volts to the interpole 2 Pass 5 volts to the gate 3〇4 305 3〇3 to the right 303_Γ The idle pole 304 passes 5 volts to the interpole and passes zero volts to the left side of the gate 303-i And after transferring 5 volts i3〇im 16402twf.doc/006 to the right side 303-r of the gate 303, channel hot electrons are generated. The zero volt from the left side 303-1 accelerates the electron toward the right side 3〇3_r of about 5 volts, and is attributed to the higher potential of the polycrystalline germanium gate 3〇3 to which 7 volts is applied, called accelerated electrons of hot electrons. It is injected into the nitride layer (tantalum film 120). The right side 303-r of the gate 3〇3 is then programmed by channel hot electron injection. Figure 3B is a circuit diagram illustrating an erase operation 330 of a charge trapping flash memory array 16 through thermal cavity injection. If the gate is selected for erasing, the remaining gate cells, namely the gate 3, the gate 3, the gate 304, the gate 305, the gate 306 and the gate 307 serve to transfer the voltage. Gate. The gate unit, that is, the gate 3 闸 gate 3 〇 2, the gate 3 〇 4, the gate 305, and the gate 307 are charged to a first voltage level, that is, 1 volt, and the gate 303 -5 volts on the right side 303-r is injected into the gate 303 through a thermoelectric hole. The SkFIN 310 layer extends from the left side of the gate 301 to the right side of the gate 307' including extending through the gate 3〇1, the gate, the gate 303, the interpole 304, the gate 305, the gate 306, and the gate 307. . A gate 3A1, a transistor 302, a gate 303-, a gate 304, a gate 3〇5, a gate 3〇6, and a gate 307 are disposed on the oxide layer 32A. A flap (smc〇n Fin) 33〇 is deeply doped on the right side of the gate 307 and the left side of the gate 301. The flaps 330 are placed on top of the oxide layer 32A. After the erase operation, the sequence is read to the read operation 36 to read the gate 3〇3. The charge is stored on the right side 3〇3_r of the gate 303, so that the reading operation is performed on the left side 303-1 of the gate 303. As time changes, the initial private pressure Vt will tend to drift higher. Although the above description shows that the charge reservoir 15 13 tw 13 15 15 402 twf.doc / 006 exists on the right side 303_r of the gate 303, and the reading operation is performed on the 4 J 303 1 of the gate 3 〇 3, but the field One of ordinary skill will recognize that, where, it can be stored on the left side and the read operation is performed on the right side. One difference between the first interpolation method 200 and the second operation method 3A is the inverse relationship between the low starting voltage and the high starting voltage. First • Operation Green 200 uses FN to erase the memory unit, which will memorize • Element _ High Start. The second method of operation uses a thermocouple to erase the memory cell, which erases the memory cell to the voltage. Figure 4 is a flow diagram illustrating the process of performing the erase operation 2, the stylization operation 24, and the read operation 260 of the first method of operation of the charge trapping flash memory array 16A. At step 41, the first method of operation 4 选择 selects the gate 2〇3 in the row to be erased by using the 胄雷_诺特海式穿穿, and the other gates, ie the gate 2 The gate 2 〇 2, the gate 2 〇 4, the gate 205, the gate 206, and the gate 207 serve as transfer gates for transmitting voltage. Charging the unselected gates, ie, the gate 2, the gate 2, the gate 2, the gate 205, the gate 206, and the gate 207, to a first voltage level, such as 10 volts, The gate 4 204 is charged up to a second voltage level, such as 20 volts, to erase the gate 203 by tunneling using fn. At step 420, the first method of operation 〇〇 uses the hot hole injection in the charge trapping flash memory array 160 to program the gate 2〇3. The stylization of the gate 203 uses thermowell injection, which is applied to the right side 203-r of the gate 203. With respect to the left side 203-1 of the gate 203, it is set to about zero volts, and the right side 203-r of the gate 203 has a higher power 16 13 〇 13 di 16402 twf.doc / 006 granules whose setting domain is 5 volts. If the phase is selected, then -5 volts is applied to the interpole 203. A voltage difference of 10 volts on the right side 2 〇 3 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ In step 430, the 'first operational method' includes performing a read miss on the gate 2〇3 in the charge trapping fast complement array 160. On the right side of the electric gate 2〇3, the reading operation is performed on the left side of the pole 2〇. When the gate 2〇3 is erased, the initial voltage Vt will tend to drift higher as time passes. Although the above description shows that charge is stored on the right side 203-r of the gate 203, and the read operation is performed on the left side 2〇3_丨 of the gate 203, one of ordinary skill in the art should recognize that the charge can be Stored on the left side, and the read operation is performed on the right j. FIG. 5 is a second illustration of the operation of performing the program operation 3〇〇, the erase operation, and the read operation 360 of the charge trapping flash memory array 16〇. A flowchart of the process of operating method 500. At step 51, the second method of operation 500 includes the use of channel thermo-electronic techniques to program the selected two-to-two-stroke. The right side 3 〇 3-r of the gate 303 is about 5 volts, and the left side 303-1 of the gate 3 〇 3 is about zero volts. On the left side of the gate 3〇3, the gate 3〇i and the gate 302 serve as transfer gates for transmitting a voltage to the left side of the gate 3〇3. The gate 301 transmits zero volts to the gate 3〇2, and the gate 3〇2 transfers zero volts to the left side of the gate 303. 3〇3- is on the right side of the gate 303, the gate 304 and the gate 305 Gate 306 and gate 307 serve as pass gates for delivering a stylized voltage to the right side of gate 303. Gate 307 transfers 5 17 I3013A1 16402twf.doc/006 volts to gate 306, gate 306 transfers 5 volts to gate 3〇5, gate 305 transfers 5 volts to gate 304, gate 3〇4 Transfer 5 volts to the gate 303 # right 303_r. After passing zero volts to the left side 303-1 of the sense electrode 3〇3 and 5 volts to the right side 3〇3_γ of the gate 3〇3, a channel hot electron condition is generated. Zero volts from the left phase-1 causes the electrons to accelerate toward the right side 303_r of approximately 5 volts, and due to the higher potential of the 7 volts of the ruthenium gate 303, it will be called hot electrons. Accelerated Electron Note • 彡 into the nitride layer (ΟΝΟ film 12G). The right side 303-r of the gate 3〇3 is then programmed by channel hot electron injection. At step 520, the second method of operation 5 includes performing a charge trapping operation of the flash memory array 3GG through the hot hole injection. If the gate 303 is selected for erasing, the remaining gate cells, namely the gate 3〇1, the gate 302, the gate 304, the gate 3〇5, the gate 3〇6, and the gate 3 serve The voltage is transmitted to the transfer gate of the gate 3〇3. The gate unit, ^ gate=1, the interpole 302, the gate 304, the gate 305, and the gate 7 3〇7 are charged to the S voltage level, for example, 1 volt, and by using the note • S. PIN ^ 31 〇 303: 5 volts on the upper to erase the gate 3〇3, and _5 volts applied to the gate 53Q'. The second operation method 5()() includes performing the idle pole 303 ^ f (four). The charge is stored on the right side of the gate period _ 进行 on the left side of the gate 303. With time = the initial voltage Vt will tend to drift higher. Although the present invention has been disclosed above by way of example, it is not intended to limit the invention of 18 13013 αΐ5 16402 twf.doc/006. Anyone skilled in the art, within the circumference, may make some more; with retouching, due to this = The spirit and the scope of the application are defined as the scope of protection of the scales U. [FIG. 1AS] FIG. 1AS illustrates a structure diagram of a charge trapping flash unit according to an embodiment of the present invention. 1B is a layout diagram illustrating a charge trapping paste unit in accordance with an embodiment of the present invention. FIGS. 2A-2C are diagrams illustrating = and reading a charge trapping flash memory array in accordance with an embodiment of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS A circuit diagram of a second method of operation for staging, undividing, and term taking power to capture a flash memory ramp in accordance with one embodiment of the present invention. FIG. 4 is an embodiment of the present invention. A flowchart of a process of performing a first method of operation. / Figure 5 is a flowchart of a process of performing a second method of operation according to an embodiment of the present invention. μ / [Major component symbol description] 100 ·• Structure 110, 170 171, 172, 173: polycrystalline germanium layer 120: oxide-nitride-oxide thin film 130: source region 140: drain region 19 B〇i3ja5 16402twf.doc/006 150, 180, 18b 182, 183, 184 210: Si-: FIN layer 160: arrays 200, 330: erase operation 2 (Π, 202, 203, 204, 205, 206, 207, 301, 302, 303, 304, 305, 306, 307: gate 203-1, 303-1: 203-r, 303-r on the left side: 220 on the right side: oxide layer 230: 矽 240, 300: stylized operation 260, 360: read operation 400: first operation method 410 to 430: step 500: second operation method 510 to 530 · · steps
2020