42086 2^ 拷 五、發明說明(1) 本發明係有關於一種半導體記憶裝置,特別係有關於 一種靜態隨機存取記憶體及其製作方法,其中,每一個記 憶胞使用四個掩埋接觸窗及雙閘極。 靜態隨機存取記憶體是所有半導體記憶體中,處理速 度最快的一種記憶體,其主要是以記憶胞内電晶體的導電 狀態來儲存資料。在習知技藝中,靜態隨機存取記憶體 (以下簡稱SRAM)有幾種不同的設計方式,例如可使用6個 NM0S電晶體構成SRAM的電路;或可用多晶矽作為負載並由 4個電晶體構成SRAM電路;以及以CMOS設計構成SRAM電路 等。42086 2 ^ Copy 5. Description of the invention (1) The present invention relates to a semiconductor memory device, and more particularly to a static random access memory and a manufacturing method thereof, wherein each memory cell uses four buried contact windows and Double gate. Static random access memory is the fastest processing memory of all semiconductor memories. It mainly stores data in the conductive state of the memory cells. In the conventional art, there are several different designs of static random access memory (hereinafter referred to as SRAM). For example, 6 NM0S transistors can be used to form a SRAM circuit; or polycrystalline silicon can be used as a load and composed of 4 transistors. SRAM circuit; and SRAM circuit with CMOS design.
請參閱第1圖’習知以CMOS所設計的SRAM記憶單元係 包括兩個PM0S及四個NM0S,其中兩個pm〇S 1 0、1 2為SRAM 的上拉電晶體’NM0S 18、20為SRAM的傳輸閘(pass gate) 電晶體,NMOS 14、16為SRAM的下拉電晶體。其中,pm〇s 10、12與NM0S 14、16的閘極分別共用,10、12的源 極共同接至電源Vcc ’而NM0S 14、16的没極則共同接地。 雖然’上述幾種不同方式所構成的SR·電路,其操作 方式並無不同’但是在SRAM電路處於待機(standby)模式 時’由CMOS電晶體所構成的SRAM電路所消耗的功率,會較 上述另兩種方式構成的別…電路低。因此,就實際的^用 而言,自然是以由CMOS電晶體構成SRAM電路的方式較符八 需求。不過,利用此種方式設計構成SRAM電路,卻會有二 憶胞的尺寸較大的缺點,對於提升電路積集度有不良的影Please refer to Figure 1 'The conventional SRAM memory cell system designed with CMOS includes two PM0S and four NM0S, of which two pm0S 1 0 and 12 are SRAM pull-up transistors' NM0S 18 and 20 are Pass gate transistors for SRAM. NMOS 14, 16 are pull-down transistors for SRAM. Among them, the gates of pm0s 10, 12 and NM0S 14, 16 are shared respectively, the sources of 10, 12 are commonly connected to the power source Vcc 'and the non-poles of NM0S 14, 16 are commonly grounded. Although 'the SR and circuits constructed by the above several different methods have no difference in operation', when the SRAM circuit is in standby mode, the power consumed by the SRAM circuit composed of CMOS transistors will be higher than the above The other two ways are different ... the circuit is low. Therefore, as far as practical applications are concerned, it is natural that the SRAM circuit is composed of CMOS transistors to meet the eight requirements. However, using this method to design and construct the SRAM circuit has the disadvantage of a large memory cell size, which has a negative impact on improving the circuit accumulation degree.
第5頁 4 2 0862 ^ 五、發明說明(2) 有鑑於此’為了克服習知技術的缺點,本發明之主要 目的即在於提出一種靜態隨機存取記憶體及其製作方法, 其可減小由CMOS電晶體構成的SRAM記憶胞的大小。 為了達成上述目的,本發明提出新的SRAM佈局,並且 使用四個掩埋接觸窗及雙閘極的製程’來製作SRAM記憶 胞。 接著配合附圖說明本發明之實施例,以更清楚地說明 本發明之架構、步驟及優點,其中: 第1圖係繪示習知技藝中*CM0S電晶體構成的SRAM之 電路圖。. 第2a圖至第2f圖係繪示用以說明本發明之靜態隨機存 取記憶體的製作方法之佈局圖。 參考標號之說明 上拉電晶體10、12 ’下拉電晶體14、16,傳輸閘電 晶體1 8、20,主動區1〇〇,掩埋接觸窗丨丨仏〜丨〗〇d,閘極 材料120 ’閘極122a〜122f,接觸孔130a、130b、 132a〜132d 、 134a 、 134b 、 136a 、 136b 、 150a 、 150b ,電 源線1 4 0 ’連接線1 2 4、1 4 2 ’下拉電晶體的汲極之接地線 144 ’字元線連接線146,插塞148,位元線160。 實施例之說明 凊參閱第2a圖至第2f圖’本發明之靜態隨機存取記憶 體的製作方法係包括下列步驟:(i )首先,定義主動區“ 1 00,形成P井區及η井區’同時利用植入的步驟調整NM〇s 及PM0S的臨界電壓值’並且成長厚約4〇〜15〇埃的閘極氧化Page 5 4 2 0862 ^ V. Description of the invention (2) In view of this, in order to overcome the shortcomings of the conventional technology, the main purpose of the present invention is to propose a static random access memory and a manufacturing method thereof, which can reduce The size of the SRAM memory cell composed of CMOS transistors. In order to achieve the above-mentioned object, the present invention proposes a new SRAM layout, and uses a process of four buried contact windows and a double-gate process to make SRAM memory cells. Next, the embodiments of the present invention will be described with reference to the drawings to more clearly explain the architecture, steps, and advantages of the present invention. Among them: FIG. 1 is a circuit diagram showing an SRAM composed of a * CM0S transistor in the conventional art. Figures 2a to 2f are layout diagrams illustrating the method for manufacturing the static random access memory of the present invention. Description of reference numerals Pull-up transistors 10, 12 'Pull-down transistors 14, 16, Transmission gate transistors 18, 20, Active area 100, Buried contact window 丨 丨 仏 ~ 丨〗 0d, Gate material 120 'Gate 122a ~ 122f, contact holes 130a, 130b, 132a ~ 132d, 134a, 134b, 136a, 136b, 150a, 150b, power line 1 4 0' connection line 1 2 4, 1 4 2 ' The pole ground line 144 'word line connection line 146, plug 148, bit line 160. Description of the embodiment 凊 Refer to FIG. 2a to FIG. 2f. The method for manufacturing the static random access memory of the present invention includes the following steps: (i) First, define an active area "100, and form a P-well area and an η-well. The region 'simultaneously uses the implantation step to adjust the threshold voltages of NMOs and PMOS' and grows a gate oxide with a thickness of about 40 to 15 angstroms.
層,沉積厚約500-1500埃的未摻雜多晶矽,並定義出用以 使上拉電晶體與下拉電晶體連接的掩埋接觸窗丨丨仏〜丨^/ 的位置,如第2a圖所示;(ii)請參閱第2b圖,接著沉 約5 0 0 ~ 1 5 0 0埃的未摻雜多晶石夕或非晶石夕以作為閘極材料 120,並且定義出閘極I22a~122i的位置,同時定義出上拉 電晶體的汲極與下拉電晶體的源極的連接線〗24 ; (i丨i ) a 參閱第2c圖’然後進行n型淡汲極摻雜及?型淡没極雜明 接著形成厚約600〜2500埃的間隔層’利用光罩微影技術與 植入技術’分別進行n +摻雜及p+摻雜,以分別形成·與 PMOS的源極/汲極與閘極的區域及掩埋接觸窗的植入,然 後形成鈦金屬石夕化物作為閘極與源極/;;及極,接著沉積一 層氧化物作為隔離層’再形成接觸孔13〇a、i30b、 132a〜132d、134a、134b、136a、136b ;(iv)請參閱第2d 圖,形成一金屬層’並且加以定義,以得到電源線丨4〇、 作為上拉電晶體與下拉電晶體閘極的多晶碎與上拉電晶體 的沒極及下拉電晶體的源極的連接線142、下拉電晶體的 >及極之接地線144、字元線的連接線(strapping)146及插 塞148,(v)請參閱第2e圖’接著再次形成一廣氧化物作為 隔離層,並形成接觸孔150a、150b ;(vi)請參閱第2f圖, 然後再形成一金屬層,並定義出位元線160。 上述N型淡汲極摻雜係以能量5 Kev〜60KeV,濃度 1E13〜3E14/cm2,植入磷或砷。上述P型淡汲極摻雜則係 以能量2 KeV〜60KeV,濃度1E13~ 3E14/cm2,植入硼或說 化硼。Layer, depositing about 500-1500 angstroms of undoped polycrystalline silicon, and defining the position of the buried contact window 丨 丨 仏 ~ 丨 ^ / to connect the pull-up transistor to the pull-down transistor, as shown in Figure 2a ; (Ii) Please refer to FIG. 2b, and then sink approximately 500 to 150 angstroms of undoped polycrystalline or amorphous stone as the gate material 120, and define gates I22a to 122i (I 丨 i) a Refer to Figure 2c 'and then do n-type light-drain doping and? Then, a spacer layer with a thickness of about 600 to 2500 angstroms is formed, and then a mask layer using a photolithography technique and an implantation technique is used to perform n + doping and p + doping, respectively, so as to form a source and a PMOS / The area of the drain and gate electrodes and the buried contact window are implanted, and then a titanium metal oxide compound is formed as the gate and source electrodes; and the electrode, and then an oxide layer is deposited as an isolation layer to form a contact hole 13〇a , I30b, 132a ~ 132d, 134a, 134b, 136a, 136b; (iv) Please refer to Figure 2d to form a metal layer 'and define it to obtain the power line. 40. As a pull-up transistor and a pull-down transistor The polycrystalline chip of the gate and the connection line 142 of the pull-up transistor and the source of the pull-down transistor, the > of the pull-down transistor and the ground line 144 of the electrode, and the strapping 146 of the character line Plug 148, (v) please refer to FIG. 2e ', then form a wide oxide again as an isolation layer, and form contact holes 150a, 150b; (vi) refer to FIG. 2f, and then form a metal layer and define Out bit line 160. The N-type light-drain doping is implanted with phosphorus or arsenic at an energy of 5 Kev ~ 60KeV and a concentration of 1E13 ~ 3E14 / cm2. The P-type light-drain doping is implanted with boron or boron at an energy of 2 KeV ~ 60KeV and a concentration of 1E13 ~ 3E14 / cm2.
420862420862
上述Ν+摻雜係以能量10 KeV〜6〇Key,濃度2Ει5〜 6EI5/cm2 ,植入砷。上述ρ+摻雜則係以能, 濃度2Ε15〜6E15/cm2,植入硼或氟化硼。 上述形成閘極與源極/汲極的方式係利用兩個快速昇 溫退火(RTA)步驟形成鈦或鈷。換言之,先利用第一個快 速昇溫退火步驟,以溫度6〇〇〜700 °C處理形成的金屬,然 後選擇性地移除未反應的金屬,再利用第二個快速昇溫退 火步驟,以溫度800〜900 °C處理形成的金屬。 上述形成介電層作為隔離層的步驟係沉積硼磷四氧乙 基矽(BPTEOS),其包括卜2K埃的電漿促進四氧乙基矽 (PETEOS)及3〜12K埃的硼磷四氧乙基矽,再以750〜9 0 0 °C填 入BPTEOS 。The N + doping is implanted with arsenic at an energy of 10 KeV ~ 60Key and a concentration of 2E5 ~ 6EI5 / cm2. The above ρ + doping is implanted with boron or boron fluoride at a concentration of 2E15 ~ 6E15 / cm2. The gate and source / drain formation methods described above utilize two rapid temperature annealing (RTA) steps to form titanium or cobalt. In other words, the first rapid heating annealing step is used first to process the formed metal at a temperature of 600 to 700 ° C, and then the unreacted metal is selectively removed, and then the second rapid heating annealing step is used at a temperature of 800 ~ 900 ° C processed metal. The above step of forming a dielectric layer as an isolation layer is the deposition of borophosphotetraoxyethyl silicon (BPTEOS), which includes a 2K angstrom plasma-promoted tetraoxyethyl silicon (PETEOS) and a 3 ~ 12K angstrom borophosphotetraoxysilane. Ethyl silicon, then fill BPTEOS at 750 ~ 900 ° C.
第8頁Page 8