TW468277B - Non-volatile flash memory cell with non-symmetric starting voltage - Google Patents
Non-volatile flash memory cell with non-symmetric starting voltage Download PDFInfo
- Publication number
- TW468277B TW468277B TW89116493A TW89116493A TW468277B TW 468277 B TW468277 B TW 468277B TW 89116493 A TW89116493 A TW 89116493A TW 89116493 A TW89116493 A TW 89116493A TW 468277 B TW468277 B TW 468277B
- Authority
- TW
- Taiwan
- Prior art keywords
- flash memory
- memory cell
- channel region
- patent application
- substrate
- Prior art date
Links
Landscapes
- Non-Volatile Memory (AREA)
- Semiconductor Memories (AREA)
Abstract
Description
ο 8 27 7 五、發明說明(1) 1發明領域 本發明係有關於具有不對稱起始電壓之非揮發性快閃 記憶體晶胞,特別是藉由提高部份通道區域之起始電壓來 防治抹除不當之非揮發性快閃記憶體晶胞。 5-2發明背景: 由於快閃記憶體(flash memory)具有可電除( electrically erase)且可程式(programmable)的特徵, 並且可以同時對整個記憶體陣列(a r r a y)中的各快閃記憶 體晶胞(ce 1 1 )進行電除與程式,因此已廣泛地被應用來作 為各種既須要儲存之資料不會因電源中斷而消失而要須要 可以重覆讀寫資料的記憶體,例如數位相機的底片或主機 板之基本輸入輸出系統。因此,如何提昇快閃記憶體的性 能與降低快閃記憶體的成本,便成為一個重要的課題β 如第一 Α圖所示,常見之快閃記憶體晶胞的結構為堆 疊式結構’基本上至少包含源極11、汲極1 2、浮動閘( floating gate) 13與控制閘(C0ntr0l gate)14,其中源極 11、汲極1 2與控制閘1 4係分別連接到不同的電源以控制快 閃記憶體的寫入(program)、讀取(read)與抹除(erase), 並且浮動閘1 3與控制閘1 4係被位於底材丨〇上的介電質層丄5ο 8 27 7 V. Description of the invention (1) 1 Field of the invention The present invention relates to a non-volatile flash memory cell with an asymmetric starting voltage, in particular by increasing the starting voltage of some channel regions. Prevents improper erasure of non-volatile flash memory cells. 5-2 Background of the Invention: Because flash memory has the characteristics of being electrically erasable and programmable, and can simultaneously access each flash memory in the entire memory array. The unit cell (ce 1 1) performs the division and program, so it has been widely used as a memory that can repeatedly read and write data, such as digital cameras, which need to store data that will not disappear due to power interruption. The basic input or output system of the film or motherboard. Therefore, how to improve the performance of flash memory and reduce the cost of flash memory has become an important issue. As shown in Figure 1A, the common flash memory cell structure is a stacked structure. It includes at least source 11, drain 1, 2, floating gate 13 and control gate (C0ntr0l gate) 14, where source 11, drain 12, and control gate 14 are respectively connected to different power sources to The flash memory is programmed, read, and erased, and the floating gates 13 and 14 are controlled by a dielectric layer on the substrate. 5
第5頁 d 6 8 27 7 五'發明說明(2) 所圍繞β 就Ν型快閃記憶體之晶胞而言(底材1 0是Ν槊底材)’當 資料要寫入時,係將源極11接地並分別對控制閘1 4與沒極 1 2都施加正電壓,此時由於並沒使用輕摻雜汲極,因此部 份的電子經散射進入浮動閘13,並且因為週遭介電質層I5 的位能障(potential barrier)而被陷(trap)在浮動閘13 。由於位於浮動閘1 3的電子會影響源極1 1與汲極1 2間通道 區域(channel region)的啟始電壓(threshold voltage) ,進而控制通道區域的導通與否,因此位於浮動閘13中的 電子便代表了資料,而可以由通道區域是否導通來讀取。 而在抹除快閃記憶體已有資料時,係將源極11接地,並對 控制閘1 4施加較汲極1 2低的正電壓,利用卩〇|1〇1·-Nordheim tunnel ing使得浮動閘13的電子經汲極1 2而消失 ’藉以確保通道區域的導通不再受先前輸入到浮動閘丨3之 電子的影響。 顯然地’若抹除時未能將浮動閘1 3回復到電子被注入 前的狀況,不論是電子殘留在浮動閘中的抹除不足(under erase)或者是不只注入之電子被移走連原屬浮動間丨3之電 子都被移走而使得浮動閘1 3帶正電荷1 6的過度抹除(〇ver :),t第一 B圖示’都會影響快問記憶體晶胞的性能 。例如,w某快閃記憶體晶胞發生過度抹除時, 記憶體晶胞又再;g枯宜 :*也 目丨丨# λ 全I ' ^入丹度被寫入貝料’則進入沣動閘丨3的電子會Page 5d 6 8 27 7 Fifth 'Explanation of the invention (2) Surrounded β As for the unit cell of the N-type flash memory (substrate 10 is Ν 槊 substrate)' When the data is to be written, the system Ground the source 11 and apply positive voltages to the control gates 14 and 12 respectively. At this time, since the lightly doped drain is not used, part of the electrons are scattered into the floating gate 13 due to the surrounding media. The potential barrier of the electrical layer I5 is trapped at the floating gate 13. Since the electrons located in the floating gate 13 will affect the threshold voltage of the channel region between the source 11 and the drain 12 and thus control the conduction of the channel region, they are located in the floating gate 13 The electrons represent the data and can be read by whether the channel area is conducting. When erasing the existing data in the flash memory, the source electrode 11 is grounded, and the control gate 14 is applied with a positive voltage lower than the drain electrode 12 by using 卩 〇 | 1〇1 · -Nordheim tunnel ing to make The electrons of the floating gate 13 disappear through the drain 12 to ensure that the conduction of the channel region is no longer affected by the electrons previously input to the floating gate 3. Obviously, if the floating gate 13 cannot be restored to the state before the electrons are injected during erasing, whether it is the under erase of the electrons remaining in the floating gate or not only the injected electrons are removed, even the original All the electrons belonging to the floating space 3 are removed, so that the floating gate 13 is over-erased with a positive charge 16 (0ver :), and the first B graph 'will affect the performance of the memory cell. For example, when a certain flash memory cell is over-erased, the memory cell is again; g Kuyi: * 也 目 丨 丨 # λ All I '^ Into the degree of dendrite is written into the shell material' then enter 沣The electronic meeting of the brake 丨 3
,1 6 8 27 7 _ _ ________****- —-- 五、發明說明(3) 被浮動閘1 3中的正電荷1 6所中和,若中和的數量大到使得 剩餘之電子不能改變通道區域的導通’亦即浮動閘1 3中的 電子無法在讀取資料時被偵測到’使得這個快閃記憶體晶 胞無法發揮儲存資料的作用。再者,若過度抹除的程度強 烈到使得浮動閘1 3上的正電荷1 6多到使得通道區域自動被 導通,此時這個快閃記憶體晶胞將無法被用來儲存任何資 料。除此之外,由於實際應用中快閃記憶體陣列通常同時 包含許多個快閃記憶體晶胞,如低密度高反應速率”非或( NOR) ”結構中的位元線’因此單—個快閃記憶體晶胞的不 正常運作’往往會導致整個快閃記憶體陣列的失效。 在個記快包顯 需一閃成晶明 必每快形體會 於對個來憶又 由針整用記程 ,須得以閃過 題必使可快除 問且會中各抹 的而只粒試入 當,不晶測寫 不路,少 一新 除電路減逐重 抹拭電會,的。 理測試又者時本 處入測且再當成 來加加而。不試 式地添,置除測 方外都化位抹與 的額胞雜的生間 路中晶複列發時 電1)體構陣及試 以el憶結體以測 若(C記的憶程加 粒閃體記過增 晶决It閃的地 另一種處理抹除不當問題的作法是使用所謂的分裂 極(spiU gateh如第一 C圓所示,此時通道區域可^ 為二部份,一部份係僅有控制開η於其上,另一部份 時有控制17與浮動開18與其上。顯然地,如第一 ,即使此時因抹除不當使得浮動閘i 8中出現正電 6 8 27 7 •、發明說明(4) ---— --—— 有 γ立 、一 但 制 並 敎 式 不位於,動閘18下方之通道區域的導通與否無法控制’ 閑17控:動=8:方:通道=域的導通與否仍可以由控 且大多 ' 抹除不當的問題可以被有效防治, 載子係使用效率是及極熱載子注入法約1〇〇倍的源極 與運作〇法將電子注入浮動閘1 8。當然,分裂閘即的型 86, m與 Τ 有4許丄的變化’如 u. s_4,639·89 3、u.s. 5,4 LS· 4’868, 629等所揭露。 裂蘭ΐ論如何,比較第—A圖與第一 C圖可以看出,使用八 4快閃記憶體晶胞的結構 ^ =的結構複雜,特別是控制問】 』的= 顯者不同,也因此製程會苍町也狀 ,在浮動f,1 18與浮動間13_者 並且成本也較高。再者 制問Π的長度必須較控能相同的前提τ,由於控 極之快閃記憶體晶胞的面^也較=度長’因此使用分裂閉 由前面的討論可以看出, 晶胞的結構,不是不能诂=知技藝中各種快閃記憶體 不當但又會面臨製程複雜;2不當,〶是雖能防治抹除 =新的快閃記憶趙晶胞'以本因此’必須 胞之可快速寫入與祙除等優:,充伤發揮快閃記憶體晶 5'3發明目的及概述:1 6 8 27 7 _ _ ________ ****---- 5. Explanation of the invention (3) Neutralized by the positive charge 16 in the floating gate 13 if the number of neutralization is so large that the remaining electrons The continuity of the channel region cannot be changed, that is, the electrons in the floating gate 13 cannot be detected when reading the data, so that this flash memory cell cannot play the role of storing data. Furthermore, if the degree of over-erase is so strong that the positive charge 16 on the floating gate 13 is so large that the channel area is automatically turned on, the flash memory cell cannot be used to store any data at this time. In addition, since the flash memory array usually contains many flash memory cells at the same time in practical applications, such as bit lines in a low-density, high-reaction-rate "NOR" structure, so a single The abnormal operation of the flash memory cell often leads to the failure of the entire flash memory array. In a note pack, it is necessary to flash into a crystal. Every time you realize it, you must remember it again and use the needle to record the process. You must be able to pass the question. In the right place, the crystal can't measure and write. There is a new one to remove the circuit. The physical test is also tested locally and added as a result. Add it without trial, and set the frontal cell mixed with the frontal cell except for the test side to regenerate the interphase. 1) The body array and try to remember the structure to test if (C recorded Yicheng added granules and flashes. It ’s been added to the crystal. It ’s another way to deal with the problem of erasing improperly is to use the so-called split pole (spiU gateh is shown in the first circle C. At this time, the channel area can be divided into two parts. One part only has the control opening η on it, and the other part has the control 17 and the floating opening 18 on it. Obviously, like the first, even if the floating gate i 8 appears in the floating gate due to improper erasure at this time. Positive electricity 6 8 27 7 • Description of the invention (4) ----- --—— There is γ stand, once the system is not located, the conduction in the channel area under the moving gate 18 cannot be controlled. Control: dynamic = 8: square: channel = domain conduction or not can still be controlled and most of the problem of improper erasure can be effectively prevented, the carrier system use efficiency is about 100 times that of the extremely hot carrier injection method The source and operation of the method inject electrons into the floating gate 18. Of course, the split gates, ie, the type 86, m and Τ have 4 variations, such as u. S_4,639 · 89 3. us 5,4 LS · 4'868, 629, etc. The split-lane theory, comparing Figure -A and Figure C can be seen, using the structure of the 8 4 flash memory cell structure ^ = structure It ’s complicated, especially the control question.] == The difference is obvious, so the process will also be Cangcho, and the cost is higher between the floating f, 1 18 and the floating 13_. Moreover, the length of the system must be longer than The premise of the same control τ, because the surface of the flash memory cell of the control pole is also longer than = degree, so the use of split closure can be seen from the previous discussion, the structure of the cell is not impossible. All kinds of flash memory are improper but will face complex process; 2 Improper, 〶 is able to prevent erasure = new flash memory Zhao Jing cell 'for this reason' must be able to be quickly written and erased, etc .: The purpose and summary of the invention of flash memory crystal 5'3:
8 27? -----— 五、發明說明(5) ,本發明背景中’傳統快閃記憶趙晶胞之缺點 發性快閃記憶體。 〜町m抹除不當之非揮 晶胞ΐ!:::一目的是在不明顯變動堆叠式快閃記憶體 提下…可以避免快閃記憶趙的功能因 徠除不當而失效之快閃記憶體晶胞結構。 本發明的目的還句括防·止_ 因過度抹除而失效的問ι ⑷閃記憶體在_結構中 針對前述之各目的’本發明提 壓之非揮發性快閃記憶趙晶胞,上:具不對稱啟始電 雜區、浮動閘與控制閘^其中,^,3 ·通道區域、摻 面内,並位於源極與汲極之間’在區域係位於底材的表 材的表面内;摻雜區係位於通道區=極與汲極皆位於底 且摻雜到摻雜區之多數個第一雜靠近源極的一端,並 多數個第二雜質的電性相同;控龆^性與底材所摻雜之 與通道區域絕緣…動閉係位===區域上並 ,並同時與控制閘與通道區域絕緣。道Q域與控制閘之間 再者,本具對稱啟始電壓之非揮發 尚可以擴展到不-定使用摻雜區,只要通道::憶體晶胞 遇迢&域可分為較8 27? ------V. Description of the invention (5), in the background of the present invention, the disadvantages of the traditional flash memory Zhao Jing cell: hair flash memory. ~ Machine erases improper non-volatile cell ΐ! ::: one purpose is to raise the stack flash memory without obvious changes ... to prevent flash memory from failing due to improper removal Unit cell structure. The purpose of the present invention also includes the problem of preventing and stopping the flash memory from being invalidated due to excessive erasure. The flash memory is in the structure for the aforementioned purposes. : With asymmetric starting electric miscellaneous area, floating gate and control gate ^ wherein, ^ 3, channel area, mixed surface, and located between the source and drain, in the region is located on the surface of the surface of the substrate Inside; the doped region is located in the channel region = both the pole and the drain are at the bottom and most of the first impurities doped to the doped region are close to the source, and the majority of the second impurities have the same electrical properties; Insulation and substrate doped insulation with the channel area ... The dynamic closing system is parallel to the area, and at the same time insulated from the control gate and the channel area. Between the Q-domain of the channel and the control gate, the non-volatile with a symmetrical starting voltage can still be extended to indefinite use of the doped region, as long as the channel :: memory cell
:6 8 277 五、發明說明(6) 靠近源極的第一通道區域以及較靠近汲極的第二通道區域 ,並且第一通道區域的啟始電壓較第二通道區域的啟始電 壓大即可。換言之,本發明的關鍵是不避免抹除不當的發 生,但以一啟始電壓較大的通道區域來防止非揮發性快閃 記憶體晶胞的不正常運作。 5-4發明詳細說明: 本發明的發明人指出一個關鍵:針對抹除不當時浮動 閘會使得位於浮動閘下方之通道區域不當導通的缺失,一 個有效的解決機制是將通道區域分成在會被抹除不當之浮 動閘所導通以及不會被抹除不當之浮動閘所導通的二部份 ,然後藉由調整不會被抹除不當之浮動閘所導通之通道區 域的的導通與否’便可以使得整個通道區域的導通與否不 會受到浮動閘下方通道區域之不當導通的影響β 顯然地,習知之分裂閘極僅是將通道區域分成在會被 抹除不當之浮動閉所導通以及不會被抹除不當之浮動閘所 導通的一種作法。針對習知使用分裂間極 胞的主要缺失:製程複雜與所佔面積較大,由於控;通道 區域導通與否的關鍵是在於通道區域的啟始電壓而不是閘 極與底材(通道區域)間的距離’本發明的發明人提出另一 種!3道區域分段的_作法:不改變堆叠式快閃記憶體晶胞: 6 8 277 V. Description of the invention (6) The first channel region near the source and the second channel region closer to the drain, and the starting voltage of the first channel region is greater than the starting voltage of the second channel region. can. In other words, the key of the present invention is to avoid the occurrence of improper erasure, but to prevent the abnormal operation of the non-volatile flash memory cell with a channel region with a large initial voltage. 5-4 Detailed description of the invention: The inventor of the present invention pointed out a key point: To eliminate the lack of improper conduction of the channel area below the floating gate when the floating gate is erased, an effective solution mechanism is to divide the channel area into Erase the two parts that are conducted by the improper floating gate and the two parts that will not be erased by the improper floating gate, and then adjust the conduction of the channel area that is not erased by the improper floating gate. Can make the entire channel area conductive or not not affected by the improper conduction of the channel area below the floating gate. Obviously, the conventional split gate only divides the channel area into the floating gate which will be improperly erased and will not be conductive. A method of conducting by improperly removed floating gates. According to the main lack of conventional use of interstitial cells: the process is complex and the area occupied is large, due to control; the key to the conduction of the channel area is the starting voltage of the channel area, not the gate and substrate (channel area). The distance between the 'inventors of the present invention proposed another! 3 methods of segmenting the area: Do not change the stacked flash memory cell
第10頁 -;8 27 7 五、發明說明(7) ---------- 之控制閘與浮動開的配置’但改變通道區域中啟始 分佈’以不對稱之啟始電壓分佈將通道區域分段。後的 根據上述之討論’本發明的發明人提出一種非揮發 快閃記憶體晶胞’如第二A圖所示,至少包含♦坌一1 ^ ΰ 通道區 域2卜第二通道區域22、源極23、汲極24、浮動閘25與控 制閘27。 ” 通道區域係位於底材20 (如P型底材)的表面内,並位於 底材2 0之表面内的源極2 3與ί及極2 4之間,在此通道區域可 分為較靠近源極2 3的第一通道區域2 1以及較靠近沒極2 4的 第二通道區域22,並且第一通道區域21的啟始電壓較第二 通道區域2 2的啟始電壓大。 控制閘2 6係位於通道區域上並與通道區域絕緣。浮動 閘25則係位於通道區域與控制閛25之間,並且同時與控制 閘26與通道區域二者絕緣。當然浮動閘23與控制閘24係位 於介電質層27中,藉以使得底材20、浮動閘23與控制閘24 三者彼此間都被絕緣。 明顯地,由於第一通道區域21的啟始電壓比第二通道 區域22分的臨界電壓大,因此在寫入資料到此快閃記憶體 晶胞時’沒極24附近的反轉區域(depletion region)會向 源極2 3延伸,亦即係以源極熱載子注入法將電子(P型底材Page 10-; 8 27 7 V. Description of the invention (7) ---------- The configuration of the control gate and floating opening 'but change the initial distribution in the channel area' with an asymmetric starting voltage The distribution segments the channel area. According to the above discussion, 'the inventor of the present invention proposes a non-volatile flash memory cell', as shown in the second diagram A, it contains at least a channel region 2 and a second channel region 22 and a source. Electrode 23, drain 24, floating gate 25 and control gate 27. The channel area is located in the surface of the substrate 20 (such as a P-type substrate) and between the source 2 3 and ί and the electrode 2 4 in the surface of the substrate 20. In this channel area, it can be divided into more The first channel region 21 near the source electrode 23 and the second channel region 22 near the electrode 24 and the start voltage of the first channel region 21 are larger than the start voltage of the second channel region 22. Control Gates 2 and 6 are located on the channel area and insulated from the channel area. Floating gates 25 are located between the channel area and control 閛 25, and are insulated from both the control gate 26 and the channel area. Of course, floating gate 23 and control gate 24 Is located in the dielectric layer 27, so that the substrate 20, the floating gate 23 and the control gate 24 are insulated from each other. Obviously, because the starting voltage of the first channel region 21 is 22 minutes lower than that of the second channel region The critical voltage is large, so when writing data to this flash memory cell, the depletion region near the end 24 will extend to the source 23, that is, the source hot carrier is injected. Method will electronics (P-type substrate
第11頁 λ ο 8 27 7 五'發明說明(8) 的少數載子)輸入到浮動閘23’但隨著輪入電子數目之増 加電子的輸入點會逐漸向沒極2 4靠近。 均 由於此時第一通道區域2 1的啟始電壓較大,亦即控制 閘26施加電壓使第二通道區域22恰好導通時,第一通道區 域21上未導通’必須對第一通道區域21施加較大的電壓: 能使第一通道區域21導通。因此如第二b圖所示般,即使 抹除不當(如出現正電荷28在浮動閘25中)使得浮動閘25下 之第二通道區域2 2發生不正常導通,只要第一通道區域21 的啟始電壓高到不會因正電荷28的出現而被導通,整個快 ^記憶體晶胞仍不會失效β換句話說,藉由調整第一通道 區域2 1的啟始電壓,便可使得此快閃記憶體晶胞的功能不 會被使用過程中之抹除不當所影響。當然,可以由測試過 程之資料來判斷第一通道區域2 1之啟始電壓必需多大才能 防治抹除不當的影響β 夕在此,調整啟使電壓最常用的方法是以離子植入法將 個雜質293(如硼離子)打入到通道區域靠近源極23 的了端’藉以形成摻雜區296(摻雜區29 6所在之通道區域 可視為第一通道區域21,通到區域的其它部份可以視為第 一通道區域22),如第二D圖所示(在此介電質層2 7僅畫出 部份)。甘 異中摻雜到摻雜區2 9 6之雜質的電性係與底材2 0所 #雜之雜質的電性相同,並且由於啟始電壓與底材2〇表面 的狀態密切,因此摻雜區29 6的厚度通常係較通道區域的Page 11 λ ο 8 27 7 Five 'invention description (8) Minority carriers) are input to the floating gate 23', but as the number of electrons in turn increases, the input point of electrons will gradually approach the electrode 24. Because at this time the starting voltage of the first channel region 21 is large, that is, when the control gate 26 applies a voltage to make the second channel region 22 just turn on, the first channel region 21 is not turned on. Applying a larger voltage: The first channel region 21 can be turned on. Therefore, as shown in the second figure b, even if the erasure is improper (such as the occurrence of a positive charge 28 in the floating gate 25), the second channel region 22 under the floating gate 25 is abnormally connected, as long as the The starting voltage is so high that it will not be turned on by the appearance of the positive charge 28, and the entire fast memory cell will not fail. In other words, by adjusting the starting voltage of the first channel region 21, you can make The function of the flash memory cell is not affected by improper erasure during use. Of course, the data of the test process can be used to determine how large the starting voltage of the first channel area 21 must be to prevent the improper erasure. Here, the most common method for adjusting the starting voltage is to use ion implantation. Impurities 293 (such as boron ions) are driven into the channel region near the end of the source electrode 23 to form a doped region 296 (the channel region where the doped region 29 6 is located can be regarded as the first channel region 21 and pass to other parts of the region The portion can be regarded as the first channel region 22), as shown in the second D diagram (only a part of the dielectric layer 27 is drawn here). The electrical properties of the impurities doped in the doped region 296 are different from the electrical properties of the impurities in the substrate 20, and because the starting voltage is close to the state of the substrate 20 surface, the doping The thickness of the miscellaneous region 29 6 is usually greater than that of the channel region.
第12頁 468 277Page 12 468 277
厚度來得小(比較第二D圊與第二A 以大角度佈植技術被打入底材2〇時,、。並且當雜質29 3係 材20的方向與底材20之表面的角度這些雜質293打入底 汉大約為20度。 再著,由於啟使電壓與浮 的電容成反比,第一通道區域 容可以調整到比該第二通道區 電容來得小。亦即也可以不用 過形成不同電容之介電質層在 25和底材20間之介電層 與該浮動閘間介電質層的電 域與該浮動閘間介電質層的 ,質來調整啟始電壓,而透 洋動閘之下來調整啟始電壓 最後,當須要抹除此晶胞 壓,並對控制閘2 6施加負電壓 Nordheim tunneling將浮動間 時’只要將汲極2 3接上正電 ’即可利用Fowlor-2 5的電子拉到汲極2 4。 顯然地,由於本發明可以有效地防治抹除不當所造成 成的快閃記憶體晶胞失效,因此本發明可以防止Ν型快閃 記憶體在NOR結構中因過度抹除而失效等快閃記憶比應用 上常見的問題。並且,由於本發明係直接經修改快閃記憶 想BS胞的結構來防治抹除不當’因此不需要再以測試電路 來俄測形成好的快閃記憶體陣列,既可以節省晶粒面積, 也可以節省測試時間與降低成本。 藉由比較第二A圖、第一人圓與第一 c圖,可以看出本The thickness is small (comparing the second D 圊 and the second A when they are driven into the substrate 20 with a large angle implantation technique.) And when the direction of the impurities 29 3 series 20 and the angle of the surface of the substrate 20 these impurities The penetration of 293 is about 20 degrees. Furthermore, because the starting voltage is inversely proportional to the floating capacitance, the area capacity of the first channel can be adjusted to be smaller than the capacity of the second channel area. That is to say, it is not necessary to form a difference. The dielectric layer of the capacitor is between the dielectric layer 25 and the substrate 20 and the electrical domain of the floating inter-gate dielectric layer and the floating inter-gate dielectric layer are used to adjust the starting voltage and pass through the ocean. Adjust the starting voltage by moving the brake. Finally, when it is necessary to erase this cell voltage and apply a negative voltage to the control gate 2 6 Nordheim tunneling will float between the 'as long as the drain 2 3 is connected to a positive power' can use Fowlor The electron of -2 5 is drawn to the drain electrode 24. Obviously, since the present invention can effectively prevent the failure of the flash memory cell caused by improper erasure, the present invention can prevent the N-type flash memory from being damaged in the NOR. Flash memory than structure is invalid due to over-erase Also, since the present invention directly modifies the structure of the BS cell to prevent improper erasure by directly modifying the flash memory, it is not necessary to use a test circuit to test the formation of a good flash memory array, which can save crystals. The grain area can also save test time and reduce costs. By comparing the second A picture, the first person circle and the first c picture, we can see that
第13頁 46 8 27 7 五、發明說明(ίο) 發明基本上是一個堆養式快閃記憶體晶胞,控制開2 6與浮 動閘25二者大致平行’控制閘26之底部較浮動閘25之頂部 遠離底材20’而且二者的形狀也很簡單,都可以用一個沉 積程序與一個微影蝕刻程序形成,而且在控制閘26的形成 也不需要任何的程序來形成如第一C圖所示之彎曲的控制 閘26,亦即本發明之製程較習知使用分裂閘極之快閃記憶 體的製程簡單。 再者,雖然在第二A圖中浮動閘23與底材20平行,但 本發明並不限於此。本發明的重點係在於通到區域中啟始 電壓的不對稱’利用靠近源極2 3—端的啟始電壓來防治抹 除不當時的不正常導通β至於浮動閘25與控制閘26二者的 配置是可以調整。 最後’為增加介電係數與增加電子在浮動閘25内的错 存時間,如第二C圖所示般,控制閘2 6與浮動閘2 5之間係 以複合介電質層2 9所分隔。在此複合介電質層2 9係由重叠 的三層介電質層所形成’中間一層的材料係為下列之—: 氮化碎或氣·氧化$夕’而表面的二層則為氧化層。 以上所述僅為本發明之較佳實施例而已,並非用以限 定本發明之申請專利範圍;凡其它未脫離本發明所揭示之 精神下所完成之等效改變或修飾,均應包含在下述之申請 專利範圍内。Page 13 46 8 27 7 V. Description of the invention (ίο) The invention is basically a stockpile flash memory cell, which is controlled to be parallel to the floating gate 25 and the floating gate 25. The bottom of the control gate 26 is more than the floating gate. The top of 25 is far away from the substrate 20 'and the shapes of the two are also very simple. Both can be formed by a deposition process and a lithography etching process, and the formation of the control gate 26 does not require any process to form the first C The curved control gate 26 shown in the figure, that is, the manufacturing process of the present invention is simpler than the conventional flash memory using split gates. Furthermore, although the floating gate 23 is parallel to the substrate 20 in the second A diagram, the present invention is not limited to this. The focus of the present invention is on the asymmetry of the starting voltage in the area. The use of the starting voltage close to the source 23-terminal to prevent erasure of improper abnormal conduction β, as for both the floating gate 25 and the control gate 26. The configuration is adjustable. Finally, to increase the dielectric constant and increase the time of electrons in the floating gate 25, as shown in Figure 2C, the control gate 26 and the floating gate 25 are connected by a composite dielectric layer 29. Separated. Here the composite dielectric layer 29 is formed by overlapping three dielectric layers. The material of the middle layer is one of the following:-Nitrided or gas-oxidized and the two layers on the surface are oxidized Floor. The above are merely preferred embodiments of the present invention, and are not intended to limit the scope of patent application for the present invention; all other equivalent changes or modifications made without departing from the spirit disclosed by the present invention shall be included in the following Within the scope of patent application.
第14頁 8 277 圖式簡單說明 第一 A圖為常見之堆疊式快閃記憶體晶胞的結構示意 圖; 第一 B圖為常見之堆疊式快閃記憶體晶胞遇到過度抹 除時之缺失的示意圖; 第一 C圖為常見之使用分裂閘之快閃記憶體晶胞的結 構不意圖, 第一 D圖為常見之使用分裂閘之快閃記憶體晶胞遇到 過度抹除時之缺失的示意圖; 第二A圖為本發明所提出之快閃記憶體晶胞之一種結 構的不意圖, 第二B圖為本發明所提出之快閃記憶體晶胞防治不當 抹除之機制的示意圖; 第二C圖為本發明所提出之快閃記憶體晶胞之另一種 結構的不意圖,以及 第二D圖為本發明所提出之快閃記憶體晶胞之另一種 結構的不意圖。Page 14 8 277 Brief description of the diagram The first diagram A is a schematic diagram of the structure of a common stacked flash memory cell. The first diagram B is a diagram of a common stacked flash memory cell when it encounters excessive erasure. Missing schematic diagram; Figure C shows the structure of a common flash memory cell using a split gate, and Figure D shows a common flash memory cell using a split gate when it is over-erased. Missing schematic diagram; Figure A is the intention of a structure of the flash memory cell proposed by the present invention, and Figure B is the mechanism of the improper erasure prevention of the flash memory cell proposed by the present invention Schematic diagram; the second diagram C is the intention of another structure of the flash memory cell proposed by the present invention, and the second diagram D is the intention of another structure of the flash memory cell proposed by the present invention .
第15頁 3 277 圖式簡單說明 主要部分之代表符號: 10 底 材 11 源 極 12 汲 極 13 浮 動 閘 14 控 制 閘 15 介 電 質 層 16 正 電 荷 17 控 制 閘 18 浮 動 閘 20 底 材 21 第 一 通 道 區 域 22 第 二 通 道 區 域 23 源 極 24 汲 極 25 浮 動 閘 26 控 制 閘 27 介 電 質 層 28 正 電 何 293 雜 質 296 摻 雜 區Page 15 3 277 Schematic representation of the main symbols: 10 substrate 11 source 12 drain 13 floating gate 14 control gate 15 dielectric layer 16 positive charge 17 control gate 18 floating gate 20 substrate 21 first Channel region 22 Second channel region 23 Source 24 Drain 25 Floating gate 26 Control gate 27 Dielectric layer 28 Positive charge 293 Impurity 296 Doped region
第16頁Page 16
Claims (1)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW89116493A TW468277B (en) | 2000-08-16 | 2000-08-16 | Non-volatile flash memory cell with non-symmetric starting voltage |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW89116493A TW468277B (en) | 2000-08-16 | 2000-08-16 | Non-volatile flash memory cell with non-symmetric starting voltage |
Publications (1)
Publication Number | Publication Date |
---|---|
TW468277B true TW468277B (en) | 2001-12-11 |
Family
ID=21660792
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW89116493A TW468277B (en) | 2000-08-16 | 2000-08-16 | Non-volatile flash memory cell with non-symmetric starting voltage |
Country Status (1)
Country | Link |
---|---|
TW (1) | TW468277B (en) |
-
2000
- 2000-08-16 TW TW89116493A patent/TW468277B/en not_active IP Right Cessation
Similar Documents
Publication | Publication Date | Title |
---|---|---|
TWI328882B (en) | Methods to resolve hard-to-erase condition in charge trapping non-volatile memory | |
CN100590853C (en) | Semiconductor memory and method for forming same | |
TW469601B (en) | Dual bit trench type gate non-volatile flash memory cell structure and the operating method thereof | |
JPH0368542B2 (en) | ||
TW512350B (en) | Method for erasing memory cells in a nonvolatile memory | |
TW476144B (en) | Non-volatile memory | |
US20020003252A1 (en) | Flash memory circuit with with resistance to disturb effect | |
JPH0770626B2 (en) | Non-volatile memory cell | |
TW535163B (en) | Erasing method in non-volatile memory device | |
JP2005184029A (en) | Nonvolatile storage element and semiconductor integrated circuit device | |
KR20070082241A (en) | Non-volatile memory device | |
TW468277B (en) | Non-volatile flash memory cell with non-symmetric starting voltage | |
US10797063B2 (en) | Single-poly nonvolatile memory unit | |
JP2926545B2 (en) | Method for manufacturing flash memory device | |
CN108831888B (en) | SONOS memory and manufacturing method thereof | |
JP2000223598A (en) | Nonvolatile memory element | |
JP3464955B2 (en) | Semiconductor storage device and storage method | |
CN103872059A (en) | P-type channel flash memory and manufacturing method thereof | |
JPS5836506B2 (en) | semiconductor storage device | |
TW565932B (en) | Flash memory cell with buried floating gate and method for operating such a flash memory cell | |
JPH0354869A (en) | Semiconductor non-volatile memory | |
JP2005184028A (en) | Nonvolatile storage element | |
TWI695489B (en) | Low-voltage fast erasing method of electronic writing erasing type rewritable read-only memory | |
JP2719641B2 (en) | Semiconductor nonvolatile memory | |
TW200534361A (en) | A split-gate p-channel flash memory cell with programming by band-to-band hot electron method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
GD4A | Issue of patent certificate for granted invention patent | ||
MK4A | Expiration of patent term of an invention patent |