TW536562B - Low defect density, vacancy dominated silicon - Google Patents

Low defect density, vacancy dominated silicon Download PDF

Info

Publication number
TW536562B
TW536562B TW87105351A TW87105351A TW536562B TW 536562 B TW536562 B TW 536562B TW 87105351 A TW87105351 A TW 87105351A TW 87105351 A TW87105351 A TW 87105351A TW 536562 B TW536562 B TW 536562B
Authority
TW
Taiwan
Prior art keywords
crystal
axially symmetric
ingot
symmetric region
patent application
Prior art date
Application number
TW87105351A
Other languages
Chinese (zh)
Inventor
Robert Falster
Steve A Markgraf
Seamus A Mcquaid
Joseph C Holzer
Paolo Mutti
Original Assignee
Memc Electronic Materials
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Memc Electronic Materials filed Critical Memc Electronic Materials
Priority to TW87105351A priority Critical patent/TW536562B/en
Application granted granted Critical
Publication of TW536562B publication Critical patent/TW536562B/en

Links

Landscapes

  • Crystals, And After-Treatments Of Crystals (AREA)

Abstract

The present invention relates to single crystal silicon, in ingot or wafer form, which contains an axially symmetric region in which vacancies are the predominant intrinsic point defect and which is substantially free of agglomerated vacancy intrinsic point defects wherein the first axially symmetric region comprises the central axis or has a width of at least about 15 mm and a process for the preparation thereof.

Description

536562 經濟部中央標準局貝工消贽合作社印製 A7 五、發明説明( tgg背景 本發明大致係料半導體級單晶々之製備,用於電子组 =製造中:特別是本發明關於單晶矽塊與晶圓,其具有 θ格工位爲王材料(軸向對稱區,而無凝結内在點瑕疵 ,以及關於其製備方法。 〃一 ' 早晶石夕係大部分半導體電子組件製造過程上之啓始材料 j通常利用俗稱C咖hralski(ez)法製備,在此方法中,聚 晶體石夕㈣)用於料且加以熔化,_晶财接觸於溶態 曰單印缸以缓忮抽拉而生長。在一頸部之生成完成後 ’印體直徑即藉由減慢拉單晶速率及/或熔化溫度以擴大, 直到取知所需或目帛纟徑爲止。具有大致固《直徑之晶體 筒形主體隨後藉由控制拉單晶率與溶化溫度而生長,並且 補仏:減少〈熔化程度。接近於生長過程末尾但是在坩堝 7熔悲矽用完之前,晶體直徑需逐步減小以形成一端錐, =錐通常係因增加拉單晶率與供至坩堝之熱而形成,當直 徑變得夠小時,晶體即脱離熔態。 ,近年來已知單晶矽中有許多瑕疵會在晶體固化冷却後形 成於日日肢生長室中,此瑕疵有一部分係因過量(即濃度在溶 解限度以上)之内在點瑕疵存在而引起,此即習知之晶格空 位與行間充。自一熔態物生長之矽晶體一般係以過量之一 種或其他種内在點瑕疵生長、即晶體格子狀之晶體空位(v) 或石夕行間充(I),經建議在矽中之諸點瑕疵類型與初期濃度 係在固化時間做判斷,若濃度達到系統中之一臨界超飽和 度且點瑕疵之移動性相當高,則一反應或一凝結現象將易 (請先閱讀背面之注意事項再填寫本頁}536562 Printed by A7 of the Central Laboratories of the Ministry of Economic Affairs, Cooperating Consumers Co., Ltd. V. Description of the invention (tgg background The present invention is roughly based on the preparation of semiconductor-grade single-crystal rhenium, which is used in the electronics group = manufacturing: especially the present invention is about single-crystal silicon Blocks and wafers, which have θ lattice stations as king materials (axially symmetric regions, without condensed intrinsic point defects, and methods for their preparation. 〃 一 'Precious spar is one of the most important semiconductor electronic component manufacturing processes. The starting material j is usually prepared by the commonly known Chralski (ez) method. In this method, polycrystalline stone is used to melt and melt the material. And the growth. After the generation of a neck is completed, the diameter of the printed body is enlarged by slowing the rate of pulling the single crystal and / or the melting temperature until the required or mesh diameter is known. The cylindrical body of the crystal is then grown by controlling the rate of single crystal growth and the melting temperature, and supplemented by: reducing the degree of melting. Close to the end of the growth process, but before the crucible 7 is used up, the crystal diameter needs to be gradually reduced to form End cone, = cone is usually formed by increasing the pulling rate of single crystal and the heat supplied to the crucible. When the diameter becomes small enough, the crystal will leave the molten state. In recent years, many defects in single crystal silicon are known to The crystals are formed in the day-limb growth chamber after solidification and cooling. Part of this defect is caused by the existence of internal point defects in excess (that is, the concentration is above the limit of dissolution). This is the conventional lattice vacancy and row filling. Silicon crystals grown in a state of matter are generally grown with an excess of one or other intrinsic point defects, that is, crystal lattice-shaped crystal vacancies (v) or interstitial filling (I). It is recommended that the types of point defects in silicon and The initial concentration is judged at the curing time. If the concentration reaches a critical supersaturation in the system and the mobility of point defects is quite high, a reaction or a condensation phenomenon will be easy (please read the precautions on the back before filling this page) }

、1T -4- 536562 經濟部中央標準局货工消贽合作社印製 A7 B7 五、發明説明(2 ) 於發生,秒中之凝結内在點瑕疵在複雜且高積體電路製造 中嚴重衝擊到材料之生產量。、 1T -4- 536562 A7 B7 printed by the Goods and Consumers Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 5. Description of the invention (2) The condensed internal point defect in the second occurred in the manufacturing of complex and high-integrity circuits. Production volume.

晶格空位型瑕戚公認爲此可見式晶體瑕疵之原始,如D 形瑕疵、流線形瑕疵(FPDs)、閘氧化體(G0I)瑕疵、晶體原 始粒子(cop)瑕疵、晶體原始光點瑕疵(LPIH)'以乂可由 紅外線散射技術觀察到之特定級大型瑕疵,如利用掃描紅 外線顯微鏡與雷射掃描攝影。過量晶格空位之區域中亦存 在核狀瑕戚’成爲環氧化謗發堆疊錯誤(〇ISF)。經推測, 此特殊之瑕戚係因過量晶格空位存在而催化之高溫核化氧 凝結。 相關於行間充之瑕疵較少研究,其通常視爲行間充錯位 環或網路之密度,此瑕疵並無關於閘氧化體失效,即一項 重要之晶圓性能準據,但是其廣爲認定係其他裝置失效之 原因,一般係關於漏電問題。 此存在於CzochralSk;L矽中之晶格空位與行間充凝結瑕疵 之欲度一般範圍大約爲l*l〇3/cm3至iMOVcm3 ,諸値雖然較 低’但是凝結之内在點瑕疵在裝置製造時係快速增加,事 實上現在已成爲裝置製程中之產量限制因素。 目前主要有三種方式可處理凝結内在點瑕疵之問題,第 一万式包括針對拉單晶技藝之方法,用於減低錠塊中之凝 結内在點瑕疵之密度。此方备可進一步細分爲若干方法, 即具有拉單晶條件之方法,其造成晶格空位爲主材料之生 成,以及具有拉單晶條件之方法,其造成行間充爲主材料 之生成。例如,經建議則凝結瑕疵之密度可由以下方法減 -5- 本纸張尺度適用中國國家標準(CNS ) Λ4規格(210X 297公釐) (請先閲讀背面之注意事項再填寫本頁) t Γ 536562 經濟部中央標準局買工消贽合作社印¾ A7 B7 五、發明説明(3 ) — 低:(i)控制V/G。以生長出一晶體,其格子狀之晶格空位主 要爲内在點瑕疵,及(ii)藉由改變(通常爲減缓)拉單晶過程 期間矽塊自大約11001至1050^:之冷却速率,以利影變凝 結瑕疵之核成型速率。本方式雖可減少凝結瑕疵之密^, 但是無法避免其生成,由於裝置製造者之要求-越來。越嚴格 ,諸瑕疵之存在將繼續成爲一大問題。 其他方法亦建議在晶體主體生長期間減緩拉單晶率至大 約0.4mm/min以下,惟,此項建議亦不合適,因爲缓慢之拉 單晶率會導致各種拉晶機減產,較重要的是,此拉單晶率 會導致具有一向濃度行間充之單晶矽生成,再因此高濃度 造成凝結行間充瑕疵之生成與相關於此瑕疵之所有街生問 題。 第二種方式處理凝結内在點瑕疵之問題時係包括針對凝 結内在點瑕戚生成後之溶解或消滅方法,大體上其係利用 晶圓狀秒之南溫熱處理而達成,例如Fusegawa等人提出之 歐洲專利申請案50381/A1號所示,矽塊以超過〇 8 mm/min 之生長率生長,並且以U5(TC至1280Ό溫度範圍將錠塊裁 切下來之晶圓做熱處理,以減低接近晶圓表面薄區中之瑕 疵密度。所需之特定處理係依據晶圓中之凝結内在點瑕疫 濃度與位置而定,自一晶體切下之不同晶圓並無均勻之# 向瑕疵濃度,因此需要不同乏後生長處理條件。此外,此 晶圓熱處理較爲昴貴,可能使金屬污物進入矽晶圓,且對 所有之晶體瑕戚並無適用效果。 第三種方式在處理凝結内在點瑕疵之問題時係做單晶碎 -6- 本纸張尺度適用中國國家標準(CNS ) Λ4規格(2I0X29?公釐) (請先閲讀背面之注意事項再填寫本頁)Lattice vacancy type defects are generally recognized as the origin of visible crystal defects, such as D-shaped defects, streamlined defects (FPDs), gate oxide (G0I) defects, crystal original particle (cop) defects, and crystal original light spot defects ( LPIH) 'is a large-scale defect of a specific grade that can be observed by infrared scattering technology, such as using a scanning infrared microscope and laser scanning photography. Nucleiform defects are also present in the region of excessive lattice vacancies, becoming epoxidation stacking errors (〇ISF). It is speculated that this particular defect is the coagulation of high-temperature nucleated oxygen catalyzed by the presence of excessive lattice vacancies. The defects related to inter-row charging are rarely studied. It is usually regarded as the density of inter-row charging dislocation rings or networks. This defect is not related to gate oxide failure, which is an important criterion for wafer performance, but it is widely recognized It is the cause of the failure of other devices, and it is generally about the leakage problem. This desire exists in CzochralSk; the lattice vacancies in the L silicon and the desire for filling condensation defects between the rows generally range from about 1 * l03 / cm3 to iMOVcm3. This is a rapid increase, and in fact has now become a production limiting factor in the device manufacturing process. At present, there are mainly three ways to deal with the problem of condensed intrinsic point defects. The tenth formula includes a method for pulling single crystal technology to reduce the density of condensed intrinsic point defects in ingots. This method can be further subdivided into several methods, namely the method with single crystal pulling condition, which causes the generation of lattice vacancies as the main material, and the method with single crystal pulling condition, which results in the generation of interstellar filling as the main material. For example, the density of coagulation defects can be reduced by the following methods after suggestion. -5- This paper size applies Chinese National Standard (CNS) Λ4 specification (210X 297 mm) (Please read the precautions on the back before filling this page) t Γ 536562 Printed by the Central Bureau of Standards, Ministry of Economic Affairs, Consumers' Cooperatives ¾ A7 B7 V. Description of Invention (3) — Low: (i) Control V / G. In order to grow a crystal, the lattice-shaped lattice vacancies are mainly intrinsic point defects, and (ii) by changing (usually slowing down) the cooling rate of the silicon block from about 11001 to 1050 ^ during the single crystal pulling process, In order to change the nucleation rate of coagulation defects. Although this method can reduce the density of condensation defects, it cannot be avoided due to the requirements of device manufacturers. The more stringent, the existence of flaws will continue to be a major problem. Other methods also suggest slowing down the pulling rate of single crystal to below 0.4mm / min during the growth of the crystal body. However, this suggestion is not suitable, because the slow pulling of single crystal rate will lead to the reduction of various crystal pulling machines. The more important thing is that This pulling single crystal rate will result in the formation of single crystal silicon with a constant concentration of interstitial interstitials, and therefore the high concentration will cause the generation of condensation interstitial interstitial defects and all the problems associated with this defect. The second way to deal with the problem of condensed intrinsic point defects includes the method of dissolving or eliminating the condensed intrinsic point defects after generation. Generally speaking, it is achieved by heat treatment at the south temperature of the wafer-like seconds. For example, Fusegawa et al. As shown in European Patent Application No. 50381 / A1, silicon blocks are grown at a growth rate exceeding 0 8 mm / min, and wafers cut from the ingots at a temperature range of U5 (TC to 1280 ° C) are heat-treated to reduce access The density of defects in thin areas on the wafer surface. The specific treatment required is based on the concentration and location of the condensed spots in the wafer. Different wafers cut from a crystal have no uniform #directional defect concentration, Therefore, different post-growth processing conditions are required. In addition, the wafer heat treatment is relatively expensive, which may allow metal contaminants to enter the silicon wafer, and has no applicable effect on all crystal defects. The third method is inherent in the treatment of condensation The problem of point defects is made of single crystals. -6- This paper size applies the Chinese National Standard (CNS) Λ4 specification (2I0X29? Mm) (Please read the precautions on the back before filling this page)

經濟部中央標準局負工消费合作社印製 536562 A7 ---:_— ___ B7 五、發明説明~~~~ '~' — 晶圓表面上之薄石夕晶層外延式沉積,此方法提供之單晶石夕 係具有-大致無凝結内在點瑕戚之表面。惟,外延式沉積 會增加晶圓之成本。 由諸發展以觀,有必要使用一種單晶矽之製備方法,用 於防止凝結内在點瑕疵之生成,其藉由抑制造時之凝結 反應而成,其並非簡單地限制此瑕疵之形成率或在其形成 後才試以消除一些瑕疵。抑制凝結反應之方法可產生一矽 基材其大致上典凝結内在點瑕戚,此方法亦提供單晶石夕 晶圓具有增加之產量,其利用每晶圓需要之數個積體電路 .即可,而無外延製程之相關高成本。 發明概述 本發明之目的中,緣是以提供一種錠塊狀或晶圓狀之單 晶矽’其具有一大致徑向寬度之軸向對稱區,大致上不具 有晶格空位或矽行間充凝結生成之瑕疵;及提供一種製備 單晶硬塊之方法,其中晶格空位與行間充之濃度係經控制 ’以防止當鍵塊自固华溫度冷却時内在點瑕疵在錠塊固定 直徑段之一軸向對稱段内凝結。 因此,簡言之,本發明係指一種單晶矽晶圓,具有_中 心舶線、一前側與一後側且皆大致垂直於中心軸線、一環 側緣、及一自晶圓中心軸線延伸至環測緣之半徑。晶圓包 含一第一軸向對稱區,其中乏晶格空位主妻為内在點瑕 症且大致上操凝結晶格S位内在點瑕疵•,其中第一轴向對 稱區包含中心軸線或具有一至少大約丨5腿之寬度。 本發明另指一種單晶石夕塊,具有一中心軸線、一晶種錐 -7- 本紙張尺度適用中國國家ϋ ( CNS ) Λ4規格(210X 297公釐) "^ (請先閱讀背而之注意事項再填寫本頁)Printed by the Central Standards Bureau of the Ministry of Economic Affairs and Consumer Cooperatives 536562 A7 ---: __ ___ B7 V. Description of the Invention ~~~~ '~' — Epitaxial deposition of thin stone crystalline layers on the wafer surface. This method provides The monocrystalline stone has a surface that is substantially free of condensed internal spots. However, epitaxial deposition increases the cost of the wafer. From the perspective of developments, it is necessary to use a method for preparing single crystal silicon to prevent the formation of intrinsic point defects in condensation, which is formed by inhibiting the condensation reaction at the time of production. It does not simply limit the formation rate of this defect or Only try to eliminate some blemishes after they are formed. The method of inhibiting the coagulation reaction can produce a silicon substrate which is approximately the typical intrinsic point of coagulation. This method also provides single crystal wafers with increased yield, which uses several integrated circuits required per wafer. Yes, without the high costs associated with epitaxial processes. SUMMARY OF THE INVENTION The object of the present invention is to provide an ingot-shaped or wafer-shaped single crystal silicon which has an axially symmetrical region with a substantially radial width, and does not substantially have lattice vacancies or silicon interstitial condensation. Generated flaws; and a method for preparing single crystal hard blocks, in which the concentration of lattice vacancies and interstitial filling is controlled to prevent intrinsic point flaws on one axis of the fixed diameter section of the ingot when the key block cools from the Guhua temperature Condensing into a symmetrical segment. Therefore, in short, the present invention refers to a single crystal silicon wafer having a center line, a front side and a back side, all of which are substantially perpendicular to the central axis, a ring side edge, and an extending from the central axis of the wafer to Measure the radius of the edge. The wafer contains a first axially symmetric region, in which the main wife of the lack of lattice vacancies is an intrinsic point defect and the crystal lattice S site is substantially defective. The first axially symmetric region includes a central axis or has a At least about 5 legs wide. The present invention also refers to a monocrystalline stone block, which has a central axis and a seed cone. 7- This paper size is applicable to the Chinese National Cricket (CNS) Λ4 specification (210X 297 mm) " ^ (Please read the back first (Notes for filling in this page)

536562 A7 B7 ------- ------------------------------ 五、發明説明(5 ) 、一端錐、及一位於晶種錐與端錐之間之固定直徑段,固 定直徑段具有一環側緣及一自中心軸線延伸至環側緣之半 徑,單晶矽塊之特徵在於,當錠塊生長且自固化溫度冷却 後,固定直徑段即含有一第一軸向對稱區,其中之晶格空 位主要爲内在點瑕戚且大致上無凝結晶格空:内在力點瑕戚 ,其中第一軸向對稱區包含中心軸線或具有一至少大約15 咖之寬度,具有一長度,係沿錠塊固定直徑段之至少大約 20%長度之中心轴線而測量。 本發明又指一種生長一單晶碎塊之方法,其中鍵塊包含 一中心軸線、一晶種錐、一端錐、及一位於晶種錐與端錐 之間之固定直徑段,固定直徑段具有一環側緣及一自中心 線延伸至環側緣之半徑,錠塊係自熔態矽生長且依 Czochral方法而自固化溫度冷却,該方法包含在晶體固定 直馒般生長期間自固化至不低於大約1325°C之溫度範圍上 ’控制一生長速度v及一平均軸向溫度梯度G〇,以造成一 第一軸向對稱段生成Z其中當錠塊自固化溫度冷却時,晶 格空位即爲主要之内在點瑕疵且大致上無凝結内在點瑕疵 其中第一軸向對稱區延伸具有一寬度至少大約爲15匪或 含有中心軸線。 本發明之其他目的與特性可在文後明瞭與指明。 圖式簡單説明 · 圖1係一圖表,揭示行間充與晶格空位之初期濃度⑴、[V] 如何隨著比率V/G0之値增加而變化之範例,V係生長率及 G〇係平均軸向溫度梯度。 (請先閲讀背面之注意事項再填寫本頁)536562 A7 B7 ------- ------------------------------ 5. Description of the invention (5), one end cone And a fixed diameter section between the seed cone and the end cone. The fixed diameter section has a ring side edge and a radius extending from the central axis to the ring side edge. The single crystal silicon block is characterized in that when the ingot grows and After cooling at the solidification temperature, the fixed-diameter section contains a first axially symmetric region, in which the lattice vacancies are mainly intrinsic point defects and are substantially free of condensed crystal lattice holes: intrinsic force point defects, where the first axial direction The symmetry zone includes the central axis or has a width of at least about 15 cm and has a length measured along the central axis of at least about 20% of the length of the fixed diameter section of the ingot. The invention also refers to a method for growing a single crystal fragment, wherein the key block includes a central axis, a seed cone, an end cone, and a fixed diameter section between the seed cone and the end cone. The fixed diameter section has A ring side edge and a radius extending from the centerline to the ring side edge. The ingot is grown from molten silicon and cooled at a self-curing temperature according to the Czochral method, which includes self-curing to not low during the growth of a fixed crystal. Control a growth rate v and an average axial temperature gradient G0 over a temperature range of about 1325 ° C to cause a first axially symmetric segment to generate Z. Where the ingot cools at the solidification temperature, the lattice vacancies are Is a major intrinsic point defect and is substantially free of condensation intrinsic point defects where the first axially symmetric region extends with a width of at least about 15 mils or contains a central axis. Other objects and features of the present invention will be made clear and specified later. Brief description of diagrams. Figure 1 is a chart showing examples of how the initial concentration ⑴ and [V] of row filling and lattice vacancies change as the ratio V / G0 increases. V is the growth rate and G is the average. Axial temperature gradient. (Please read the notes on the back before filling this page)

、1T 鲤濟部中央榡準局貝工消贽合作社印製 - 8-, 1T Printed by the Beijin Consumers Cooperative of the Central Bureau of Standards of the Ministry of Lijing-8-

536562 經濟部中央標準局^:工消费合作社印製 A7 B7 五、發明説明(6 ) ^ 圖2係一圖表,揭示凝結行間充瑕疵生成所需自由能量中 之變化△巧如何隨著行間充即定初期濃度⑴之溫度τ而變化 之範例 圖3係一圖表,揭示行間充與晶格空位之初期濃度⑴、 如何在v/G〇値因G〇値增加而下降時沿著一錠-塊-或‘圓半徑 變化。應注意在V/I邊界處,一過渡現象發生自晶格空位爲 主材料至行間充爲主材料。 圖4係一單晶矽塊或晶圓之頂視圖,分別揭示晶格空位區 V與行間充區I爲主材料,以及存在於其間之V/I邊界。 圖5係一單晶矽塊之縱向橫截面圖,詳示錠塊之一固定直 徑段之一軸向對稱區。 … 圖6係在系列乳 >几殿熱處理後掃描一鍵塊轴向切片少數 載體#命所生之影像,詳示一大致筒形之晶格空位爲主材 料區,一大致環形之行間充爲主材料之軸向對稱區、存在 於其間之V/I邊界、及一凝結行間充瑕疵區。 圖7係以扭單晶率(即晶種上昇)爲晶體長度函數之圖表, 揭示拉單晶率如何在晶體之一段長度上呈直線下降。 圖8係在一系列氧沉澱熱處理後掃描一錠塊軸向切片少數 載體壽命所生之影像,如範例1所述。 圖9係以拉單晶率爲四牧單晶矽塊各別晶體長度函數之圖 表’矽塊分別以1 -4標示,用於產生一曲線v * (z),如範例工 所述。 圖10係在熔態/固態'介面處之平均軸向溫度梯度(}。做爲巧 向位置之函數之圖表,用於範例2所述之二個不同例子。 -9- 本纸張尺度適用中國國家標準(CNS ) Α4規格(210X297公釐) (請先閱讀背面之注意事項再填寫本頁} • —-#裳.536562 Central Bureau of Standards, Ministry of Economic Affairs ^: Printed by the Industrial and Consumer Cooperatives A7 B7 V. Description of the Invention (6) ^ Figure 2 is a chart that reveals the change in the free energy required to generate the defects between the lines of charging. Example of changing the initial concentration τ temperature τ Figure 3 is a chart that reveals the initial concentration 行 of interstitial charge and lattice vacancies, how to go along an ingot-block when v / G〇 値 decreases due to increase in G〇 値-Or 'Circle radius changes. It should be noted that at the V / I boundary, a transition phenomenon occurs from the lattice vacancy as the main material to the interspace charge as the main material. FIG. 4 is a top view of a single crystal silicon block or wafer, revealing that the lattice vacancy region V and the row interstitial region I are the main materials and the V / I boundary existing therebetween. Fig. 5 is a longitudinal cross-sectional view of a single crystal silicon block, showing in detail an axially symmetric region of a fixed diameter section of an ingot. … Figure 6 is a series of images of a small number of carriers #scanned by scanning a key block axially after heat treatment of a series of milks. It shows a roughly cylindrical lattice space as the main material area, and a generally circular row of interstitials. An axially symmetric region of the main material, a V / I boundary existing therebetween, and a condensed line defect region. Figure 7 is a graph of twisted single crystal rate (ie, seed crystal rise) as a function of crystal length, revealing how the pulled single crystal rate decreases linearly over a length of crystal. Fig. 8 is an image generated by scanning the axial life of a few ingots of a small number of carriers after a series of oxygen precipitation heat treatments, as described in Example 1. Figure 9 is a graph of the individual crystal length functions of a Simu single crystal silicon block with a pull single crystal rate. Table ′ silicon blocks are labeled 1-4 respectively to generate a curve v * (z), as described in the example. Figure 10 is the average axial temperature gradient (} at the melt / solid state 'interface. As a function of coincidence position, it is used for two different examples described in Example 2. -9- This paper scale applies China National Standard (CNS) Α4 specification (210X297 mm) (Please read the precautions on the back before filling this page} • —- ##.

、1T 經濟部中央標準局負工消费合作社印1i 536562 A7 __________ 五、發明説明(7 ) 圖11係晶格空位或行間充初期濃度[V]、[I]做爲徑向位置 之函數之圖表,用於範例2所述之二個不同例子。 圖12係以溫度爲軸向位置函數之圖表,揭示用於範例3户斤 述二個不同例子之錠塊中軸向溫度輪廓。 圖13係圖12所示二種冷却條件所生成且範例了詳述Α之行間 充濃度圖表。 圖14係在一系列氧沉澱熱處理後掃描整個錠塊軸向切片 少數載體壽命所生之影像,如範例4所述。 圖15係以圖表説明V/I邊界位置做爲單晶矽塊長度之函數 ,如範例5所述。 圖16 a係在’ 系列氧》>几殿熱處理後掃描· 一距離鍵塊肩部大 約100至250腿範圍之一小段錠塊軸向切片少數載體壽命所 生之影像,如範例6所述。 圖16b係在一系列氧沉澱熱處理後掃描一距離錠塊肩部大 約250至400腿範圍之一小段錠塊軸向切片少數載醴壽命所 生之影像,如範例6所述。 圖17係軸向溫度梯度G。在一錠塊不同軸向位置處之圖表 ,如例例7所述。 圖18係平均軸向溫度梯度内之徑向變化在一錠塊不同 位置處之圖表,如例範7所述。 圖19係以圖表説明軸向對耦區寬度與冷却率之間關係, 如範例7所述。 圖20係銅飾與瑕疵描示蝕刻後,距離錠塊肩部大約23 5至 350腿範圍之一小段錠塊軸向切片照片,如範例7所述。 -10- 本纸張尺度適用中國國家標準(CNS ) A4規格(2丨0χ 297公釐) (請先閱讀背面之注意事項再填寫本頁j、 1T Printed by the Central Standards Bureau of the Ministry of Economic Affairs and Consumer Cooperatives 1i 536562 A7 __________ V. Description of the Invention (7) Figure 11 is a graph of the initial concentration [V], [I] of the lattice vacancies or interrow filling as a function of the radial position For two different examples described in Example 2. Figure 12 is a graph using temperature as a function of axial position, revealing the axial temperature profile in an ingot used in Example 3 households and two different examples. FIG. 13 is a graph showing the intercharge density of A in detail, which is generated by the two cooling conditions shown in FIG. Fig. 14 is an image generated by scanning the whole ingot axial slice after a series of oxygen precipitation heat treatments, and the life of a few carriers, as described in Example 4. Figure 15 illustrates the position of the V / I boundary as a function of the length of the monocrystalline silicon block in a chart, as described in Example 5. Figure 16a: Scanning after a series of heat treatments in a series of "Oxygen" > A section of an ingot with a small distance from the shoulder of the key block about 100 to 250 legs is axially sliced from the life of a few carriers, as described in Example 6. . Fig. 16b is an image generated by scanning a small section of an ingot axially sliced from a small section of the ingot about 250 to 400 legs away from the ingot shoulder after a series of oxygen precipitation heat treatments, as shown in Example 6. Fig. 17 is an axial temperature gradient G. The graphs at different axial positions of an ingot are as described in Example 7. Fig. 18 is a graph of radial changes in the average axial temperature gradient at different positions of an ingot, as described in Example 7. FIG. 19 is a diagram illustrating the relationship between the width of the axial coupling region and the cooling rate, as described in Example 7. Fig. 20 is a series of copper ornaments and flaws. After the etching, a small section of the ingot is sliced axially from the shoulder of the ingot about 23 5 to 350 legs, as shown in Example 7. -10- This paper size applies to the Chinese National Standard (CNS) A4 specification (2 丨 0χ 297 mm) (Please read the precautions on the back before filling in this page j

536562 經濟部中夾標攀局兵工消费合作社印裂 A7 B7 五、發明説明(8) 圖21係銅飾與瑕疵描示蝕刻後,距離錠塊肩部大约3〇5至 460腿範圍之一小段錠塊軸向切片照片,如範例7所述。 圖22係銅飾與瑕疵描示蝕刻後,距離錠塊肩部大约14〇至 275腿範圍之一小段錠塊軸向切片照片,如範例7所述。 圖23係銅飾與瑕疵描示蝕刻後,距離錠塊肩一部大韌6〇〇至 730腿範圍之一小段錠塊軸向切片照片,如範例7所述。 圖24係以圖表説明平均軸向溫度梯度中之徑向變化G〇⑴ ,其可發生於不同結構之熱區中。 圖25係以圖表説明一錠塊在四個不同熱區結構中之軸向 溫度輪廓。 較佳實例詳細説明 —— 根據實驗證明得知,内在點瑕疵之種類與初期濃度係先 行決足,而錠塊自固化溫度(即大約141〇。〇)冷却至13〇〇。〇 以上之溫度(即至少大約1325。(:,至少大約135〇。〇或甚至至 少大約1375°C ),亦即諸瑕疵之種類與初期濃度係由v/G。比 率所控制,v係成長速·度而G。係在此溫度範圍内之平均軸 向溫度梯度。 參閲圖1,爲了增加v/Gg値,一由減少行間充爲主生長 至增加晶格空位爲主生長之過渡發生在接近於一臨界値 WG0處,根據已知資料,其大致爲2·1χ价^^汉,g在 軸向溫度梯度恒定於上述溫度範圍内之情況下決定,在此 臨界値時,諸内在點瑕疵之濃度即呈平衡。 當v/GQ値超過臨界値時,晶格空位之濃度即增加,同樣 ,當V /G。値落下至臨界値以下時,行間充之濃度即增加。 -11- 本紙張尺度適用中國國家標準(CNS ) Λ4規格(210X297公釐) (請先閱讀背面之注意事項再填寫本頁)536562 In the Ministry of Economic Affairs, the Bureau of Industry and Trade Cooperatives of the Panpan Bureau printed and cracked A7 B7. V. Description of the invention (8) Figure 21 series of copper ornaments and flaws after etching, one of the range from 305 to 460 legs from the shoulder of the ingot A photograph of axial sectioning of small ingots, as described in Example 7. Fig. 22 is a series of copper ornaments and flaws. After etching, a small section of the ingot is sliced axially from about 14 to 275 legs away from the ingot shoulder, as shown in Example 7. Fig. 23 is a series of copper ornaments and flaws. After etching, a small section of the ingot is axially sliced from 600 to 730 legs away from the shoulder of the ingot, as described in Example 7. FIG. 24 is a graph illustrating the radial change in the average axial temperature gradient, G0, which can occur in hot zones of different structures. Figure 25 graphically illustrates the axial temperature profile of an ingot in four different hot zone structures. Detailed description of the best example —— According to the experimental evidence, it is known that the type and initial concentration of intrinsic point defects are determined in advance, and the ingot self-curing temperature (ie, about 141.0%) is cooled to 1300. Above the temperature (ie at least about 1325. (:, at least about 135.0. 0 or even at least about 1375 ° C)), that is, the type and initial concentration of the defects are controlled by the v / G. Ratio, v is growing Speed · degree and G. It is the average axial temperature gradient in this temperature range. Refer to Figure 1. In order to increase v / Gg 一, a transition occurs from reducing row interstitial growth to increasing lattice vacancy mainly growth. Close to a critical 値 WG0, according to known data, it is approximately 2.1 valence ^^ Han, g is determined under the condition that the axial temperature gradient is constant within the above temperature range. At this critical ,, the internal The concentration of dot defects is in equilibrium. When v / GQ 値 exceeds the critical value, the concentration of lattice vacancies increases. Similarly, when V / G. 値 falls below the critical value, the concentration of row filling increases. -11 -This paper size applies to Chinese National Standard (CNS) Λ4 specification (210X297 mm) (Please read the precautions on the back before filling this page)

536562 A7 五、發明説明(9 ) 右諸/辰度到達系統中(一臨界超飽和程度,且若點瑕癌之 移動性相當高時,一反應或凝結現象應即容易發生,而石夕 中之凝結内在點瑕疵會嚴重衝擊到複雜與高積體電路製造 中之材料降伏性。 依本發明所示,經發現碎基質内之晶格空(反應:產生凝 結晶格空位瑕疵之反應以及矽基質内之行間充反應產生凝 結行間充瑕病之反應皆可抑制,在不限於任意之特定原理 下,相信晶格空位與行間充之濃度可在本發明製程中之晶 塊生長與冷却期間加以控制,使得系統之自由能量中之變 化不致超出一臨界値,臨界値即凝結反應自然發生而產生 凝結晶格空位或行間充瑕疵時之値。 大體上,可驅動反應使凝結晶格空位瑕疵由晶格空位點 瑕疵形成或凝結行間充瑕疵由單晶矽内行間充原子形成之 系統自由能量中之變化係由程式⑴導出·· AGv/I = kT ln(JV^L ⑴ [V/I〆536562 A7 V. Description of the invention (9) Right / Chen reached the system (a critical degree of supersaturation, and if the mobility of spot cancer is quite high, a reaction or coagulation phenomenon should occur easily, and Shi Xizhong The condensed intrinsic point defect will seriously impact the material drop in the manufacture of complex and high-integration circuits. According to the present invention, it is found that the lattice voids in the broken matrix (reaction: reactions that produce defects in the condensed crystal lattice vacancies and silicon The reaction of row-to-row interstitial reaction in the matrix, which can cause coagulation and interstitial defect, can be suppressed. It is believed that the concentration of lattice vacancies and row-to-row interstitials can be applied during the growth and cooling of the crystal block in the process of the present invention, without being limited to any specific principle. Control so that the change in the free energy of the system does not exceed a critical threshold, which is the time when condensation reactions occur naturally and condensate lattice vacancies or interstitial defects are generated. In general, the reaction can be driven to cause condensate lattice vacancy defects Lattice vacancy point defects are formed or condensed interstitial defects are changed in the free energy of the system formed by interstitial atoms in single crystal silicon. AGv / I = kT ln (JV ^ L ⑴ [V / I〆

XI 其中 △ Gv/I係用於形成凝結晶格空位瑕戚之反應或形成行間 充瑕痴之反應所需自由能量中之變化, k係波茲曼常數, T係溫度,單位爲K, 一 ' [V/I]係可用之晶格空位或行間充在單晶矽中之空間與時 間中一點處之濃度,及 ' [V/I]eq係可用之晶格空位或行間充在發生[V/I]且溫度T狀 12- (請先閱讀背面之注意事項再填寫本頁) • Ji 1 - J— · \ 經滴部中央標準局π貝工消费合作社印?木 本紙張尺度適用中國國家標隼(<:]^)八4規格(2】0'/297公釐) 經濟部中央標準局工消费合作社印¾ 536562 A7 ________ B7 五、發明説明(ι〇Γ " 態下,在空間與時間中相同點處之平衡濃度。 依此程式,以一定之晶格空位濃度[V]而言,溫度降低量 Τ即因[V;]eq隨溫度銳降而大致造成△ Gv增加,同樣地,以 一疋之行間充濃度[I]而言,溫度降低量T即因[I]eq隨溫度銳 降而大致造成△ 增加。 —一 ' 圖2係簡示用於一自固化溫度冷却之錠塊之石夕行間充之△ h與濃度變化,而不使用某些裝置以抑制石夕行間充之濃度 。當錠塊冷却時,△ q即因[I]增加超飽和而依程式(1)增大 ’且用於凝結行間充瑕病生成之能量阻隔物即接近達成, 而當冷却持續時,此能量阻隔物實際上已超量,在此點處 即發生一反應,此反應造成凝結行間充瑕疵之生成以及伴 隨之△ Gz減低,而超飽和系統係在釋放狀態,即[I]濃度減 低時。 同樣地,當一錠塊自固化溫度冷却且同時不使用某些裝 置以抑制晶格空位濃度時,△ Gv即因[V]增加之超飽和而 依程式(1)增加,且用於凝結晶格空位瑕疵生成之能量阻隔 物接近於達成,而當冷却持續時,此能量阻隔物實際上已 超量,在此點處即發生一反應,此反應造成凝結晶格空位 瑕疵之生成及△Gv之伴隨減低,而超飽和系統係釋放狀態。 晶格2位與行間充之凝結可分別在晶格空位與行間充爲 主之材料區域中避免,而此择鍵塊係自固化溫度冷却,且 其藉由保持晶格2位系統與行間充系統之自由能量爲小於 凝結反應發生時之値而達成。易言之,系統可經控制以利 於不致在晶格空位或行間充中王臨界式超飽和。此可藉由 -13 - 本紙浪尺度適用中國國家標準(CNS ) A4規格(2丨0X29*7公釐) (請先閱讀背面之注意事項再填寫本頁)XI where △ Gv / I is the change in the free energy required to form a crystal lattice vacancy defect reaction or to form an interstitial defect response, k is the Boltzmann constant, T is the temperature, and the unit is K. '[V / I] is the concentration of available lattice vacancies or interstitial spaces at a point in space and time in single crystal silicon, and' [V / I] eq is the available lattice vacancies or interstitial spaces occurring [ V / I] and temperature T-like 12- (Please read the precautions on the back before filling this page) • Ji 1-J— · \ Printed by the Central Standards Bureau of the Ministry of Industry π Beigong Consumer Cooperatives? The standard of woody paper is applicable to Chinese national standard (<:] ^) 8 4 specifications (2) 0 '/ 297 mm) Printed by the Industrial and Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 536 562 A7 ________ B7 V. Description of the invention (ι〇 Γ " Equilibrium concentration at the same point in space and time in the state. According to this formula, in terms of a certain lattice vacancy concentration [V], the temperature reduction amount T is due to the sharp drop of [V;] eq with temperature. Δ Gv is generally increased. Similarly, in the case of a row of interstitial concentration [I], the temperature reduction amount T is roughly caused by [I] eq as the temperature drops sharply. — 一 'Figure 2 is a brief illustration △ h and concentration change of Shiyuki line filling for an ingot cooled at a solidification temperature without using some devices to suppress the concentration of Shiyake line filling. When the ingot is cooled, △ q is due to [I] Increasing supersaturation and increasing according to the formula (1), and the energy barrier used to condense the formation of interstitial defect is close to reaching, and when the cooling continues, this energy barrier is actually excessive, at this point it is A reaction occurs, which results in the formation of filling defects between coagulation lines and the accompanying decrease in △ Gz The supersaturation system is in the release state, that is, when the [I] concentration is reduced. Similarly, when an ingot is cooled at the self-solidification temperature and some devices are not used to suppress the lattice vacancy concentration, △ Gv is due to [V] The increased supersaturation increases according to formula (1), and the energy barrier used for the formation of crystal lattice vacancy defects is close to being achieved, and when the cooling continues, this energy barrier is actually excessive, and at this point it is A reaction occurs, which results in the formation of condensed crystal lattice vacancy defects and the accompanying reduction of △ Gv, and the supersaturated system is in a released state. The condensation of the lattice 2 and the interspace filling can be mainly based on the lattice vacancy and the interspace filling, respectively. It is avoided in the material area, and this selective bond block is cooled by the self-curing temperature, and it is achieved by keeping the free energy of the lattice 2 position system and the interstitial system to be less than that at the time of the condensation reaction. In other words, the system can Controlled so as not to cause critical supersaturation in lattice vacancies or inter-row charging. This can be achieved by -13-this paper wave scale applies Chinese National Standard (CNS) A4 specification (2 丨 0X29 * 7 mm) (please first Read the back (Please fill in this page again)

536562 A7 B7 怒濟部中央標準局負工消费合作社印製 五、發明説明(U) ~^ 建iL晶格空位與行間充之初期濃度而達成(利用文後之v (Γ)控制)’使其相當地低而不致達到臨界式之超飽和。惟, 實施時此濃度係難以在跨越整個晶體半徑時達成,因此, 大體上危險之超飽和現象可因抑制晶體固化後之初期晶格 空位;辰度與初期行間充濃度而避免,亦即在,建立由v⑴ 決定之初期濃度後。 令人訝異的是,經發現由於大約1(Hcm 2/sec之較大行間 充移動率以及較小之晶格空位移動率,其可在較大距離上 達成晶格空位與行間充之抑制,·即大約5至10〇11以上之距離 ,利用仃間充在晶體表面之徑向擴散下沉或在晶體内之晶 格空位爲主區域,徑向擴散可有效地用於抑制行間充與晶 格工位 < 歌度,但需假定有足夠時間可容許内在點瑕疵之 m度做瓜向擴散。大體上’擴散時間係取決於行間充 與晶格空位之初期濃度中之徑向變化,較少之徑向變化則 需較短之擴散時間。通¥平均軸向溫度埽度G。係以單晶矽所增加半徑之函數 而增加,其依據Cz〇chralski法而成長,此意指v/g〇値在跨 越一錠塊之半徑時並非鼙一, 、 早 此交化 < 結果爲内在點之型 〜G、Γ::度並不恒定。若圖3,4所示WI邊界2所指之 界値沿著錠塊半徑4到達某些點時,材料即由晶格 全位爲主而轉變爲行間充爲^。 此外,錠塊將含有行間充 馬王足材料6足一車由向盤森r^/计山 % 5 (其中矽行間充原子之初期濃 度係以增大半徑之函游 材料f?、 ^ 而增加^即圍繞於晶格空位爲主之 材科—大致筒形區 ; Ί.------ (請先閱讀背面之注意事項再填寫本頁) 訂 J0 ,I ! -I · U中日曰也工位足初期濃度係以 -14-536562 A7 B7 Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Nuxi V. Description of the invention (U) ~ ^ The initial concentration of the crystalline lattice vacancies and the filling between rows was achieved (using v (Γ) control later in the article) to make It is quite low without reaching critical supersaturation. However, this concentration is difficult to achieve across the entire crystal radius during implementation. Therefore, the generally dangerous supersaturation phenomenon can be avoided by suppressing the initial lattice vacancies after the crystal is solidified; After establishing the initial concentration determined by v⑴. Surprisingly, it has been found that due to the large inter-row charge mobility and small lattice vacancy move rate of about 1 (Hcm 2 / sec), it can achieve the suppression of lattice vacancies and inter-row charge over a large distance. , I.e., a distance of about 5 to 1011 or more, the radial diffusion can be used to suppress the interstitial charge and the vacancy in the crystal as the main area. Lattice station < song degree, but it is necessary to assume that there is sufficient time to allow the m point of the internal point defect to diffuse to the ground. Generally, the 'diffusion time' depends on the radial change in the initial concentration of row filling and lattice vacancies. A smaller radial change requires a shorter diffusion time. The average axial temperature 埽 degree G is increased as a function of the radius of the single crystal silicon, which grows according to the Czochralski method, which means v / g〇 値 is not unitary when it crosses the radius of an ingot. The result is that the result is an intrinsic point ~ G, Γ :: degrees are not constant. If the WI boundary shown in Figures 3 and 4 When the boundary indicated by 2 reaches certain points along the ingot radius 4, the material is in full position by the crystal lattice. Mainly changed to interstitial charge ^. In addition, the ingot will contain interstitial charge Mawangzu material 6 feet and one car from Xiangpansen r / Jishan% 5 (The initial concentration of silicon interstitial atoms is to increase the radius Correspondence materials f ?, ^ and increase ^ means that the material department is mainly around the lattice vacancies-roughly cylindrical area; Ί .------ (Please read the precautions on the back before filling this page) Order The initial concentration of J0, I! -I · U in Japan and Japan is -14-

536562 經濟部中央榡準局員工消費合作衽印製 A7 B7 五、發明説明(12) 半徑之函數而增加)。 當含有一 V/I邊界之錠塊自固化溫度冷却時,行間充原子 與晶格空位之徑向擴散即因行間充與晶格空位之重組而造 成V/I邊界中之控向朝内變動。此外,行間充至晶體表面之 徑向擴散將隨著晶體冷却而發生,晶體之表衛-可在、晶體冷 却時保持近平衡點之瑕疵濃度,點瑕疵之徑向擴散易減低 V/I邊界外側之行間充濃度與V/I邊界内側之晶格空位濃度 。因此’若有足夠時間可供擴散,則任意處之晶格空位與 行間充之濃度可爲△ Gv與△ GI小於晶格空位凝結反應與行 間充凝結反應發生時之臨界値。— 請即參閱圖5,一單晶之矽塊10係依本發明方法第一實例 中之Czochralski法而生長,石夕塊包含一中心軸線12、一晶 種錐14、一端錐16及一位於晶種錐與端錐之間之固定直徑 #又18,固足直徑段具有一環側緣2〇及一自中心軸線η延伸 至環側緣20之半徑4。 包括生長速度v,平·均軸向溫度梯度G()、及冷却率在内 之曰日體生長條件最好可經控制,以利生成一行間充爲主材 料之軸向對稱區6,及一含有一凝結内在點無瑕疵材料軸向 對稱區9之晶格空位爲主材料大致筒形區8。軸向對稱區9具 有一寬度,係沿自V/I邊界2延伸至軸線12之半徑而測得, 在本發明之一實例中,其至少爲大約15_寬度,且寬度至 少爲錠塊固定直徑段半徑之7 5%,較佳爲至少大約15%, 又較理想爲至少大約25%,及最好爲至少大約50%。在一特 疋之較佳實例中,軸向對稱區9包括錠塊軸線12,即軸向對 -15- (請先閱讀背面之注意事項再填寫本頁}536562 Consumption Cooperation by Employees of the Central Bureau of Standards, Ministry of Economic Affairs, printed A7 B7 V. Description of Invention (12) Increased as a function of radius). When the ingot containing a V / I boundary cools down at the solidification temperature, the radial diffusion of interstitial atoms and lattice vacancies is caused by the reorganization of row interstitials and lattice vacancies, which causes the control direction in the V / I boundary to change inward. . In addition, the radial diffusion between the rows and the surface of the crystal will occur as the crystal cools. The watch guard of the crystal can maintain the defect concentration near the equilibrium point when the crystal cools. The radial diffusion of point defects easily reduces the V / I boundary. The interstitial concentration on the outer row and the lattice vacancy concentration on the inner side of the V / I boundary. Therefore, if there is sufficient time for diffusion, the concentration of lattice vacancies and interstitial spaces anywhere can be Δ Gv and Δ GI less than the critical thresholds at which the lattice vacancy condensation reaction and interstitial condensation reaction occur. — Please refer to FIG. 5. A single crystal silicon block 10 is grown according to the Czochralski method in the first example of the method of the present invention. The stone block includes a central axis 12, a seed cone 14, an end cone 16, and a The fixed diameter # 18 between the seed cone and the end cone. The diameter of the fixed foot has a ring side edge 20 and a radius 4 extending from the central axis n to the ring side edge 20. Solar growth conditions including growth rate v, horizontal and average axial temperature gradient G (), and cooling rate are best controlled to facilitate the formation of a line of interstitial axially symmetrical regions 6 and A lattice vacancy containing an axially symmetrical region 9 of a condensed intrinsic point flawless material is a substantially cylindrical region 8 of the main material. The axially symmetric region 9 has a width measured along a radius extending from the V / I boundary 2 to the axis 12. In one example of the present invention, it is at least about 15 mm in width, and the width is at least ingot fixed. The diameter section has a radius of 75%, preferably at least about 15%, more preferably at least about 25%, and most preferably at least about 50%. In a preferred example, the axially symmetric region 9 includes the ingot axis 12, which is the axial pair -15- (Please read the precautions on the back before filling this page}

本紙 Λ4 規格 經濟部中央標準局員工消t合作社印誓 536562 A7 _ _B7____ _ 五、發明説明(13) 稱區9與大致筒形區8係一致,易言之,錠塊10包括一晶格 空位爲主材料之大致筒形區8,其至少一部分係無凝結瑕疵 。此外,軸向對稱區9延伸過錠塊固定直徑段之至少大約 20%長度,較佳爲至少大約40%,較理想爲至少大約60%, 而最理想爲至少大約80%。 <一 Λ 軸向對稱區6(如果存在時)大致具有一寬度,係自環側緣 20以徑向向内趨向中心軸線12而測得,其至少大約爲錠塊 固定直徑段半徑之30%,在一些實例中至少大約40%、至少 大約60%、或甚至至少大約80%·,此外,軸向對稱區大致延 伸過錠塊固定直徑段之至少大約20%長度,較佳爲至少大 約40%,較理想爲至少大約60%,最理想爲至少大約8〇0/〇。 軸向對稱區6、9之寬度可沿中心軸線12之長度而具有某 些變化,因此,以一既定長度之軸向對稱區而言,軸向對 稱區6之寬度係由測量錠塊10環側緣2〇徑向趨向最遠於中心 軸線之一點處之距離而決定,易言之,測得之寬度可使軸 向對稱區6既定長度内、之最小距離得以決定。同樣地,軸向 對稱區9之寬度係藉測量由ν/Ι邊界2徑向趨向最遠於中心軸 線之一點處之距離而決定,易言之,所測得之寬度可使軸 向對稱區9既定長度内之最小距離得以決定。 生長速度ν及平均軸向溫度梯度(如前所定義)經由控 制而使V/Gg比値範圍在大約-0.5至2 5倍之v/G。臨界値(即依 v/G。臨界値資料所取得之大約丨χ 1〇_5cm2/sK^5x 1〇-5cm2/sK) ,v/GQ比値最好爲大約〇.6至15倍之V/G。臨界値(即依 臨界値資料所取得之大約1·3χ 10-5cm2/sKS3x 1〇-5cm2/sK), -16 - 本紙張尺度適用中國國家標準(CNS)八4^^210>< 297公釐 -----~~·. ;---r----- (請先閱讀背面之注意事項再填寫本頁)Paper Λ4 Specifications Staff of the Central Bureau of Standards, Ministry of Economic Affairs, Consumer Cooperatives, sworn 536562 A7 _ _B7____ _ V. Description of the invention (13) The area 9 is consistent with the roughly cylindrical area 8 series. In other words, the ingot 10 includes a lattice space The substantially cylindrical region 8 of the main material, at least a part of which is free of condensation defects. In addition, the axially symmetric region 9 extends over at least about 20% of the length of the fixed diameter section of the ingot, preferably at least about 40%, more preferably at least about 60%, and most preferably at least about 80%. < A Λ axial symmetry zone 6 (if present) has approximately a width, measured from the ring side edge 20 toward the center axis 12 radially inward, and is at least about 30 of the radius of the fixed diameter section of the ingot %, In some examples at least about 40%, at least about 60%, or even at least about 80%. In addition, the axially symmetric region extends approximately at least about 20% of the length of the fixed diameter section of the ingot, preferably at least about 40%, more preferably at least about 60%, and most preferably at least about 80/0. The width of the axially symmetric regions 6, 9 may have some variations along the length of the central axis 12. Therefore, for an axially symmetric region of a given length, the width of the axially symmetric region 6 is determined by measuring the ingot 10 rings. The side edge 20 is determined to be the distance at a point farthest from the center axis. In other words, the measured width can determine the minimum distance within the predetermined length of the axially symmetric region 6. Similarly, the width of the axially symmetric region 9 is determined by measuring the distance at a point where the ν / Ι boundary 2 radially moves farthest from the center axis. In other words, the measured width can make the axially symmetric region 9 The minimum distance within a given length is determined. The growth rate ν and the average axial temperature gradient (as previously defined) are controlled such that the V / Gg ratio 値 ranges from approximately -0.5 to 25 times v / G. Critical 値 (that is, according to v / G. The critical 値 data obtained is about 丨 χ 1〇_5cm2 / sK ^ 5x 1〇-5cm2 / sK), and the v / GQ ratio 値 is preferably about 0.6 to 15 times V / G. Critical volume (that is, about 1.3 × 10-5cm2 / sKS3x 1〇-5cm2 / sK obtained from the critical volume data), -16-This paper size applies Chinese National Standard (CNS) 8 4 ^^ 210 > < 297 Mm ----- ~~ ·.; --- r ----- (Please read the notes on the back before filling this page)

、1T 536562 經濟部中央標準局員工消费合作社印製 A7 B7 五、發明説明(14) 最理想的I v/Gg比値範圍大約爲〇 75至丨25倍之V/G。臨界 値(即依v/Gg臨界値資料所取得之大约16>< 1〇_5cmVsK至2.1 χ 1〇·5αη 2/sK)。在一特定較佳實例中,大致筒形區8内之 v/G。値係在v/Gq臨界値與1.1倍V/G。臨界値之間。 欲令軸向對稱區9之寬度爲最大値時,最好-疑-塊固化溫 度冷却至超過大約1050X:達一段時間,即⑴至少5小時,較 佳爲至少大約10小時,及較理想爲至少大約丨5小時,用於 150mm額定直徑之矽晶體,(丨丨)至少大約5小時,較佳爲至少 大約10小時,較理想爲至少大約2〇小時,又較理想爲至少 大約25小時,最理想爲至少大約3〇小時,用於2〇〇謹額定直 徑之碎晶體,及(iii)至少大約20小時,較隹爲至少大約40 小時’較理想爲至少大約60小時,及最理想爲至少大約75 小時’用於200腿以上額定直徑之矽晶體。冷却率之控制可 利用多種習知技術而達成,以減少熱傳導,包括使用絕熱 器、加熱器、放熱罩、及磁場。 平均軸向溫度梯度之控制可透過晶體拉單晶機之「熱 區」設計而達成,即製成加熱器、絕熱器、熱與放熱罩、 其他物件等之石墨(或其他材料),雖然設計係特別依晶體 拉單晶機之製作與模型而變化,大體上G。仍可利用習知技 術中用以控制熔態/固態介面間熱傳導之任意裝置而控制, 包括反射器、放熱罩、洗滌管、光管、及加熱器。大體而 言’ 中之徑向變化可利用定位此一裝置於熔態/固態介面 上方大約一個晶體直徑内而大幅減小,G0另可利用調整裝 置相對於溶態物與晶體之位置而控制,此係因調整裝置在 -17- 本纸張尺度適用中國國家標準(CNS ) M規格(2歐Μ?公楚) (請先閱讀背面之注意事項再填寫本頁)1T 536562 Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the invention (14) The optimal I v / Gg ratio 値 ranges from about 75 to 25 times V / G. Critical 値 (i.e., about 16 > < 10-5 cmVsK to 2.1 x 10 · 5αη 2 / sK obtained from v / Gg critical 値 data). In a particularly preferred embodiment, v / G within the generally cylindrical region 8. It is at the v / Gq threshold and 1.1 times V / G. Critical 値. When the width of the axially symmetric region 9 is to be maximized, it is best to cool the solidification temperature of the block to more than about 1050X: for a period of time, that is, at least 5 hours, preferably at least about 10 hours, and more preferably At least about 5 hours, for silicon crystals with a nominal diameter of 150mm, (丨 丨) at least about 5 hours, preferably at least about 10 hours, more preferably at least about 20 hours, and more preferably at least about 25 hours, Most preferably at least about 30 hours, for use in crushed crystals with a nominal diameter of 200 mm, and (iii) at least about 20 hours, more preferably at least about 40 hours, more preferably at least about 60 hours, and most preferably At least about 75 hours' for silicon crystals with a nominal diameter of more than 200 legs. Cooling rate control can be achieved using a variety of conventional techniques to reduce heat transfer, including the use of insulators, heaters, heat shields, and magnetic fields. The control of the average axial temperature gradient can be achieved through the "hot zone" design of the crystal pulling single crystal machine, which is made of graphite (or other materials) such as heaters, heat insulators, heat and radiation hoods, and other objects. It depends on the production and model of the crystal pulling single crystal machine, and it is generally G. It can still be controlled by any device known in the art for controlling heat transfer between the molten / solid interface, including reflectors, heat sinks, wash tubes, light tubes, and heaters. Generally speaking, the radial change in 'can be greatly reduced by positioning this device within about a crystal diameter above the molten / solid interface, and G0 can also be controlled by adjusting the position of the device relative to the dissolved substance and the crystal. This is because the adjustment device is -17- This paper size is applicable to the Chinese National Standard (CNS) M specification (2 Euro M? Public Chu) (Please read the precautions on the back before filling this page)

536562 經濟部中央標準局負工消t合作社印製 A7 五、發明説明(15) 〜 熱區中之位置或調整熔態表面在熱區中之位置而達成。此 外,當使用一加熱器時,G〇另可利用調整供給至加熱器之 月匕里而控制,諸方法之任一或所有者皆可在一批次 CzochralsKi製程期間使用,其中熔態物體積在製程期間係 完全排出。 一一 Λ 對本發明之某些實例而言,平均軸向溫度梯度最好較 爲恒定,且爲錠塊直徑之函數,惟,應注意的是由於熱區 設計中之改良可供G〇變動最少,因此相關於一固定生:率 <機械性項目即漸增爲一重要因素,此係因爲生長過程逐 漸極敏感於拉單晶率中之任意變化,再由此直接影響生長 率V。藉由製程控制,其即指有利於使、値不同於錠塊之 半位淮〇〇値中之顯著差異會造成大致朝向晶圓邊緣增 加之行間充較大濃度,藉以增加難度於避免凝結内在點瑕 疵之生成上。 由前又以觀,之控制牵涉到減少G〇中之徑向變化與取 得有利製程控制條件冬間平衡,因此,通常在大約一直徑 之晶體長度後,拉單晶率範園大約爲〇 2至〇 8111111/111111,拉 單晶率較佳爲〇·25至〇.6 mm/min,理想爲大約〇 3至〇 5 mm/mm。應注意的是,拉單晶率係取決於晶體直徑與晶體 拉單日曰機汉计二者,前述之範圍一般係用於2〇〇 直徑晶 體。一般而言,拉單晶率將故著晶體直徑增加而減小,惟 ,晶體拉單晶機可設計成容許拉單晶率超過前述者,因此 ,晶體扛單晶機最好設計成使拉單晶率越快越好,而仍可 容許一軸向對稱區依本發明而生成。 -18- 本紙張尺度關-— ^---Γ---II (請先閲讀背面之注意事項再填寫本頁) 、11 經濟部中央標準局資工消费合作社印裝 536562 A7 ____ B7 五、發明説明(16) 行間充之擴散量係在錠塊自固化溫度(大约141〇r )冷却 至碎行間充典法移動時之溫度下’精由控制冷却率而控制 之,矽行間充在接近於矽之固化溫度時較具移動性,即大 約1410Ό,惟,此移動性會隨著單晶矽塊之溫度下降而下 降。大體而言,行間充之擴散缓慢至一相當-程-度,使其在 700°C以下之一般實施時間無法移動,且甚至在8〇(Γ(:、9〇〇 °C、i〇〇〇°C 或 i〇5〇°C 亦然。 亦應注意的是,雖然發生行間充凝結反應時之溫度在理 論上可於廣泛溫度範圍内變化,但是實際上此範圍對於一 般czochralski式生長之矽而言較窄,此係因依Cz〇ch^ski 万法生長之秒中所取得之初期行間充濃度範園較窄所致。 因此,大體上一行間充凝結反應可發生於大約i 10(rc至800 c之溫度範圍内,一般大約爲1050〇c。 在行間充似呈移動性之溫度範圍内,且依熱區中之溫度 而足,則冷却率通常爲大約〇 lX: /min至3。〇 /min,冷却率較 佳爲大約0.1 c /min至/min,較理想爲大約〇n /min至 1 C/min ’ 及最理想爲大約O.rC/min 至 0.5QC/min。 藉由控制使行間充似呈移動性之溫度範圍内之錠塊冷却 率,則行間充可得到較多時間以擴散至晶體表面處之槽, 或擴散至使其滅絕之晶格空位爲主區域,因此,此行間充 I濃度可加以抑制,用於防-止凝結發生。藉由控制冷却率 而使用行間充之擴散率可放鬆嚴格之v/G〇規定,而此規定 爲取彳于抶凝結瑕疵之軸向對稱區所必要者。易言之,事實 上冷却率可加以控制,以供行間充有較多時間擴散,相對 -19- 張尺度適用中(CNS ]7^^〇-χ 297公禮)-----— (請先閱讀背面之注意事項再填寫本頁)536562 Printed by the Central Bureau of Standards, Ministry of Economic Affairs, Co-operative Society. A7 V. Description of Invention (15) ~ Position in the hot zone or adjust the position of the molten surface in the hot zone. In addition, when a heater is used, G0 can also be controlled by adjusting the supply of the moon to the heater. Any or all of the methods can be used during a batch of CzochralsKi process, where the volume of molten material It is completely discharged during the process. For some examples of the present invention, the average axial temperature gradient is preferably relatively constant and is a function of the ingot diameter. However, it should be noted that due to the improvements in the design of the hot zone, G0 can be changed the least. Therefore, it is related to a fixed birth rate: The mechanical item is gradually increasing as an important factor. This is because the growth process is becoming extremely sensitive to any change in the pull single crystal rate, and then directly affects the growth rate V. By process control, it means that the significant difference in the difference between the half position of the ingot and the ingot will result in a larger concentration between the rows that generally increase toward the edge of the wafer, thereby increasing the difficulty to avoid condensation. The generation of dot defects. From the previous perspective, the control involves reducing the radial change in G0 and achieving favorable process control conditions in the winter equilibrium. Therefore, after a crystal length of about one diameter, the pull single crystal rate range is about 0 2 To 〇8111111 / 111111, the pulling single crystal rate is preferably from 0.25 to 0.6 mm / min, and is desirably about 0.3 to 0.5 mm / mm. It should be noted that the pulling rate depends on both the crystal diameter and the crystal pulling rate. The foregoing range is generally used for 200-diameter crystals. Generally speaking, the pulling rate of single crystal will decrease due to the increase of crystal diameter. However, the single crystal pulling machine can be designed to allow the pulling single crystal rate to exceed the previous one. Therefore, it is best to design the crystal carrying single crystal machine so that The faster the single crystal rate is, the better, while still allowing an axially symmetric region to be generated according to the present invention. -18- The paper size is off --- ^ --- Γ --- II (Please read the notes on the back before filling this page), 11 Printed by the Central Standards Bureau of the Ministry of Economic Affairs, Industrial and Commercial Consumer Cooperatives 536562 A7 ____ B7 V. Description of the invention (16) The amount of inter-row interstitial diffusion is the temperature at which the ingot is cooled from the solidification temperature (approximately 1410r) to the temperature at which the inter-row interstitial method is moved. The precision is controlled by controlling the cooling rate. Silicon has more mobility at the curing temperature, which is about 1410 ° F. However, this mobility will decrease as the temperature of the single crystal silicon block decreases. Generally speaking, the diffusion between banks is slow to a considerable range-degree, making it impossible to move the general implementation time below 700 ° C, and even at 80 (Γ (:, 900 ° C, i〇〇). This is also true for 〇 ° C or i050 ° C. It should also be noted that, although the temperature at which the inter-row charge condensation reaction occurs can theoretically vary over a wide range of temperatures, in practice this range is for general czochralski-type growth. Silicon is narrower, which is due to the narrower initial row interstitial concentration range obtained in the second growth time of Czochchski method. Therefore, the intersegment condensation reaction can generally occur at about i 10 (In the temperature range from rc to 800 c, it is generally about 1050 ° C. In the temperature range where the row charge appears to be mobile, and depends on the temperature in the hot zone, the cooling rate is usually about 0lX: / min To 3.0 / min, the cooling rate is preferably about 0.1 c / min to / min, more preferably about 0 / min to 1 C / min 'and most preferably about O.rC / min to 0.5QC / min By controlling the cooling rate of the ingots within the temperature range that makes the inter-bank charge appear to be mobile, the inter-bank charge can get more It mainly diffuses to the grooves on the surface of the crystal, or to the lattice vacancies that make it extinct. Therefore, the concentration of interstitial I in this row can be suppressed to prevent the occurrence of condensation. It is used by controlling the cooling rate. The diffusion rate of line charging can relax the strict v / G0 requirement, which is necessary to take the axially symmetric region of the condensate defect. In other words, the cooling rate can be controlled in fact for the line charging There is a lot of time diffusion, relatively -19- Zhang scale is applicable (CNS) 7 ^^ 〇-χ 297 礼 礼 -----— (Please read the precautions on the back before filling this page)

經濟部中央標準扃資工消f合作社印製 536562 A7 ---一______B7 五、發明説明(17) . - 於臨界値之大範圍v/G〇値即可接受而用於取得—無凝結瑕 症之軸向對稱區。 欲在晶體固定直徑段之可觀長度上取得此冷却率,則亦 需考量錠塊端錐之生長製程,以及錠塊在端錐生長完成時 &lt;處理。通常一旦錠塊固定直徑段之生長完成—時,拉單晶 率將增加而開始進行形成端錐所需之錐化,惟,此一拉單 晶率之增加將造成固定直徑段之底段在行間充可充分移動 之溫度範圍内較快冷却。結果,諸行間充即無足夠時間可 擴散至欲滅絕之槽處,亦即此底段中之濃度無法抑制至— 相當之程度’且行間充瑕魏之凝結可能生成。 爲了防止此瑕疵生成於錠塊之底段中,因此最好錠塊之 固定直徑段具有一依據Czochralski方法之均句熱歷史,_ 均勻熱歷史可利用在一較固定速率下自熔態矽拉出單晶塊 而取得,此不僅是在固定直徑段之生長期間如此,同時在 晶體之端錐生長期間以及可能後續之端錐生長期間亦如此 。較固足足速率可取得,例如藉由⑴減低端錐生長期間之 坩堝與晶體相對於晶體固定直徑段生長期間之坩堝與晶體 之轉動率,及/或(ii)增加端錐生長期間加熱溶態碎之加熱 器供給能量相對於端錐生長期間之一般供給能量。製程變 數之其他調整係可個別或併合地發生。 茴‘錐之生長開始時,端銥之拉單晶率即建立,使得溫 度仍超過大约1 050X:之錠塊固定直徑段之任意小段可經歷 過包含鹆凝結内在點瑕痴之一軸向對稱區且已冷却至丨〇5〇 °(:以下之錠塊固定直徑段之其他小段之相同熱歷史。 -20- 本纸張尺度適用中國國家標準(CNS ) Λ4規格(210X29^7公楚)&quot; --- (請先閱讀背面之注意事項再填寫本頁} 、11 536562 經濟部中央標準局賀工消f合作社印製 A7 五、發明説明(18) : —~~-- 如前所示’晶格空位爲主區域之—最小半徑係存在以利 達成凝結行間充瑕斑之抑制,最小半徑値取決於⑽r)及 冷却率,由於晶體拉單晶機與熱區之設計會改變,因此上 迷用於WG。⑴之範圍、拉單晶率、及冷却率亦將改變,同 樣地,諸條件也可能沿著_生長晶體之長度_而一改變S如上 所示,無凝結性行間充瑕戚之行間充爲主區域之寬度最好 爲最大値,因此必須保持此區域之完度爲—盡量接近且不 超過晶體半徑與-既有晶體拉單晶機内沿生長晶體長度之 TO格空位爲主區域最小半徑之間差異之値。 。軸向對稱區6、理想寬度及用於已知晶體拉單晶機熱 區設計之所需理想晶體拉單晶率輪廓可依實驗決定,大體 而。,此實驗方式係關於先在軸向溫度輪廓上取得徑向可 用資料,以用於在一特定晶體拉單晶機内生長之錠塊,以 及取得在平均軸向溫度梯度中之徑向變化,以用於在相同 拉單晶機内生長之錠塊。總言之,此資料係用於拉單晶— 或多牧單晶矽塊,然像分析凝結性行間充瑕疵之存在,依 此,一理想之拉單晶率輪廓即可確定。 圖6係在一系列氧沉澱熱處理以發現瑕疵分佈圖形後,掃 描一 200 mm直徑錠塊一段之軸向切片少數載體壽命所產生 之影像,其説明一範圍,其中一近乎理想之拉單晶率輪廓 係用於一已知之晶體拉單晶機熱區設計。在此範例中,_ 過渡現象發生於自一行間充爲主區域最大寬度超出時之 v/G〇 (r)値(造成凝結行間充瑕疵區域28之生成)至一轴向對 稱區具有最大寬度之理想v/G。⑴値。 -21- 本紙張尺度適用中國國家標準(CNS ) A4規格(210X 297公釐) (請先閱讀背面之注意事項再填寫本頁) P*衣· --口 536562 A7 五、發明説明(I9) 除了增大G〇於錠塊半徑上所生成&quot;士之徑向變化外, v/G0亦可因v變化而在軸向變化,或因製程所 生之gg自然變化。以一標準之Czochralski製程而言,v係 隨著拉單晶率在生長循環中調整而改變,以利保持鍵塊爲 一固定直徑,拉單晶率中之諸此調整或變化_再由此使〜仏 在錠塊固定直徑段之長度上變化。依本發明之製程所示, 拉單晶率因而可做控制,以利加大錠塊之固定直徑段寬度 ,惟,其結果是發生錠塊半徑之變化。爲了確使生成之錠 塊具有一固定直徑,因此錠塊最好生長至大於所需之長徑 ’錠塊隨後進行標準製程,以自表面去除多餘材料,藉此 使鍵塊具有一固定直徑段。 大體上,當軸向溫度梯度Gg⑴之徑向變化降至最小時, 其較易於使晶格空位爲主材料無凝結瑕疵;參閲圖乃,其 説明用於四個分別熱區結構之軸向溫度輪廓;圖24呈現2 晶體中心至晶體半徑一半處之軸向溫度梯度0。⑴中之變化 經濟部中央標準局S〈工消费合作社印製 ,其係將自固化溫度至X軸線上所示溫度之梯度加以平均 而得。當晶體在熱區Ver丨與Ver 4中拉單晶時,此二熱區 係在G〇⑴中具有較大之徑向變&amp;,則其無法取得晶體自任 意軸向長度中心至邊緣處皆無凝結瑕斑之晶格空位爲主材 料。惟,當晶體在熱區Ver 2與Ver 3中拉單晶時,此二埶 區:在G“r)中具有較少徑向-變化,則其可取得晶體且具有 在晶體某些轴向長度上自中心至邊緣處無凝結瑕戒之晶格 空位爲主材料。 對於一依本發明製程製備且具有-V/I邊界之錠塊而言,Printed by the Central Standards of the Ministry of Economic Affairs, Consumers, and Cooperatives, printed by 536562 A7 --- ______B7 V. Description of Invention (17).-In the critical range of v / G〇, it can be accepted and used for acquisition-no condensation Axial symmetry area of the defect. In order to obtain this cooling rate at a considerable length of the fixed diameter section of the crystal, it is also necessary to consider the growth process of the end cone of the ingot, and the &lt; treatment when the ingot is grown. Generally, once the growth of the fixed diameter section of the ingot is completed, the single crystal pulling rate will increase and the taper required to form the end cone will begin. However, this increase in single crystal pulling rate will cause the bottom section of the fixed diameter section to Faster cooling within the temperature range where the intercharging can be fully moved. As a result, there is not enough time for the inter-row charging to spread to the trough to be extinct, that is, the concentration in this bottom section cannot be suppressed to a considerable degree 'and the condensation of inter-row filling defects may occur. In order to prevent this defect from being generated in the bottom section of the ingot, it is best that the fixed diameter section of the ingot has an average sentence thermal history according to the Czochralski method. _ The uniform thermal history can be used at a relatively constant rate. Obtained from a single crystal block, not only during the growth of the fixed diameter section, but also during the end cone growth of the crystal and possibly the subsequent end cone growth. A more solid enough rate can be obtained, for example, by reducing the crucible and crystal rotation rate during the growth of the end cone with respect to the crucible and crystal growth period during the growth of the fixed diameter section of the crystal, and / or (ii) increasing the heating during the end cone growth The energy supplied by the disintegrated heater is relative to the general energy supplied during the growth of the end cone. Other adjustments to process variables can occur individually or in combination. At the beginning of the growth of the anise 'cone, the pull single crystal rate of the terminal iridium is established, so that the temperature still exceeds about 1 050X: any small section of the fixed diameter section of the ingot can experience axial symmetry including one of the intrinsic point defects And the same thermal history of other small sections of the fixed diameter section of the ingot below. -20- This paper size applies to the Chinese National Standard (CNS) Λ4 specification (210X29 ^ 7gongchu) &quot; --- (Please read the precautions on the back before filling out this page}, 11 536562 He Gongxiao, Central Standards Bureau of the Ministry of Economic Affairs, printed A7 Cooperative Fifth, Invention Description (18): — ~~-As previously mentioned It shows that the lattice vacancies are the main area-the minimum radius is to suppress the condensed spots between condensation lines. The minimum radius 値 depends on ⑽r) and the cooling rate. Because the design of the crystal pulling single crystal machine and the hot zone will change, Therefore, the range for the WG. ⑴, the single crystal pulling rate, and the cooling rate will also change. Similarly, the conditions may also change along the length of the _growth crystal, and a change S is shown above, without condensing interline. The width of the main area should be as large as Large, so it is necessary to keep the completeness of this area as close as possible and not exceed the crystal radius and the difference between the minimum radius of the main area of the TO lattice vacancies along the length of the growing crystal in the existing crystal pulling single crystal machine. Directional symmetry zone 6, ideal width and required ideal crystal pulling single crystal rate profile used for known crystal pulling single crystal machine hot zone design can be determined by experiments, generally. This experimental method is about the axial temperature profile first. Obtain radial available data for ingots grown in a specific crystal pulling single crystal machine, and obtain radial changes in average axial temperature gradient for ingots grown in the same pulling single crystal machine In short, this data is used to pull single crystals-or polymorph single crystal silicon blocks, but it is like analyzing the existence of condensing interstitial defects. Based on this, an ideal pull single crystal rate profile can be determined. Figure 6 After a series of oxygen precipitation heat treatments to find the defect distribution pattern, scanning the axial life of a 200 mm diameter ingot with a section of a small number of carrier lifetimes, it illustrates a range, one of which is a nearly ideal pull single crystal rate The profile is used for the design of a known hot zone of a crystal pulling single crystal machine. In this example, the _ transition phenomenon occurs when v / G〇 (r) 値 (which causes condensation between rows) Generation of defect-filled area 28) to an ideal v / G with the largest width in an axially symmetric area. ⑴ 値. -21- This paper size applies to China National Standard (CNS) A4 (210X 297 mm) (Please read first Note on the back, please fill in this page again.) P * 衣 ·-口 536562 A7 V. Description of the invention (I9) In addition to increasing the radial change of G0 in the radius of the ingot, v / G0 also It can change in the axial direction due to the change of v, or the natural change of gg due to the process. In a standard Czochralski process, v is changed as the single crystal pulling rate is adjusted in the growth cycle to maintain the bond block. For a fixed diameter, various adjustments or changes in the pull single crystal rate are made _ and then ~~ is changed over the length of the fixed diameter section of the ingot. According to the process of the present invention, the single crystal pulling rate can be controlled to increase the width of the fixed diameter section of the ingot. However, the result is a change in the radius of the ingot. In order to ensure that the resulting ingot has a fixed diameter, it is best to grow the ingot to a length larger than the required long diameter. The ingot is then subjected to a standard process to remove excess material from the surface, thereby giving the key block a fixed diameter segment . In general, when the radial change of the axial temperature gradient Gg⑴ is minimized, it is easier to make the lattice vacancies free of condensation defects in the main material; see the figure for the description of the axial direction for four separate hot zone structures Temperature profile; Figure 24 presents the axial temperature gradient of 2 from the center of the crystal to half the radius of the crystal. Changes in the middle Printed by the Central Standards Bureau of the Ministry of Economic Affairs S <Industrial and Consumer Cooperatives, which is obtained by averaging the gradient from the curing temperature to the temperature shown on the X axis. When the crystal is pulled into a single crystal in the hot regions Ver 丨 and Ver 4, these two hot regions have a large radial change in G0⑴, and it cannot obtain the crystal from the center of any axial length to the edge. Lattice vacancies without condensation spots are the main material. However, when the crystal is pulled into a single crystal in the hot regions Ver 2 and Ver 3, the two fluorene regions: with less radial-change in G "r), it can obtain the crystal and has some axial directions in the crystal Lattice vacancies with no condensation flaws or rings from the center to the edge in length are the main material. For an ingot prepared according to the process of the present invention and having a -V / I boundary,

536562 A7 B7 五、發明説明(2G ) 即一含有晶格空位爲主材料之錠塊,依經驗顯示以低氧含 量之材料爲佳,即低於大約13 PPMA(依ASTM標準F-121-83 所指每百萬原子之份量),較佳的是單晶矽含有12 PPMA以 下之氧,較理想是11 PPMA以下,最理想是10 ppma以下, 此係因爲在中至高氧含量之晶圓中,即14H pp克A者, V/I邊界内側之氧謗發堆疊錯誤生成與增進氧群集帶漸爲重 要’其在已知之積體電路製程中各爲一潛在之問題源。 加強式氧群集之效應可進一步利用多種方法降低之,單 獨或併合使用皆可,例如氧沉澱核心一般會形成退火溫度 大約350 C至750°C之矽,因此,對某些應用上而言,最好 晶體爲一「短」晶體,亦即一晶體已在Cz〇chralsKi製程中 生長’直到晶種端已自矽之熔點(大約141〇。〇)冷却至大約 750 C,其後錠塊即迅速冷却。依此方式,花費在對核心生 成上十分重要之溫度範圍之時間係保持在最小値,且氧沉 殿核心具有不適切之時間形成於晶體拉單晶機内。 绖濟部中央標準局員工消費合作社印製 (請先閱讀背面之注意事項再填寫本頁) 惟’在單晶生長期闓所形成之氧沉澱核心最好利用退火 單晶碎而溶解,假設其尚未進行一穩定之熱處理,則氧沉 殿核心可藉由快速加熱矽至至少大約875χ:而將矽退火,且 最好持續昇溫至至少i〇〇(rc、至少;n〇(rC、或更高。在矽 到達i〇〇〇r以前,大致上所有(例如99%以上)之諸此瑕疵已 退火,重要的是晶圓係快速、熱至諸溫度,即溫度昇高率 至少每分鐘大約1〇。(:且最好每分鐘至少大約5(rc,否則部 刀或所有之氧沉澱核心會利用熱處理而呈穩定。平衡似乎 在較短之時間周期内到達達,即大約60秒以下,據此,單 - 23- 本^用中 標率(CNS )八4祕(210 Χ29ϋ^ &quot; -- 經濟部中央標準局負工消费合作社印裂 536562 A7 ____ __ B7 五、發明説明(21 ) 晶石夕中之氧沉澱核心可藉由退火至至少大約875。〇而溶解, 較佳爲至少大約950°C,且理想爲至少大約1100°c,其周期 至少大約5秒,且理想爲至少大約1〇分鐘。 溶解可在一般烤爐内或一快速熱退火(RTA)系統中實施, 矽之快速熱退火可在多種一般快速熱退火(RrA)爐之\壬一者 中進行’其中之晶圓係利用成列之高功率燈做個別加熱。 RTA烤爐可快速加熱一矽晶圓,例如其可在數秒内加熱一 晶圓自室到達1200 C,其中一種RTA爐爲取自AG協會(加州 觀景山)之610型烤爐。此外,溶解可在矽塊或矽晶圓上進 行,而較佳爲晶圓。 在本發明製程之一實例中,矽行間充原子之初期濃度係 控制在錠塊10之軸向對稱行間充爲主區6内,復參閲圖1, 大體上矽行間充原子之初期濃度係藉控制晶體生長速度v 及平均軸向溫度梯度gq而達成,使v/Gg比値較接近於此値 之臨界値,即ν/ι邊界發生處。此外,平均軸向溫度梯度 可建jl,而使Gg之變、化G。⑴(以及V/G。⑴)成爲G。之一變 異(即v/G。)’其成爲錠塊半徑之函數且亦受到控制。 本發明之另一實例中,v/Go亦經控制而使得錠塊之至少 一部分長度上沿著半徑並無V/I邊界存在,在此長度中,矽 係自中心至環側邊緣呈晶格空位爲主,而凝結晶格空位瑕 疵則王要利用控制v/G〇以避克於自錠塊環側緣呈徑向向内 延伸之一軸向對稱區内,亦即生長條件爲控制方式以令 WGg具有一値介於v/Gg臨界値與丨丨倍乂犯。臨界値之間。 應注意的是,依本發明製備之晶圓係適用做爲 其 ---;-------- (請先閱讀背面之注意事項再填寫本頁)536562 A7 B7 V. Description of the invention (2G) is an ingot with lattice vacancies as the main material. According to experience, materials with low oxygen content are better, that is, less than about 13 PPMA (according to ASTM standard F-121-83). The weight per million atoms), it is preferred that the single crystal silicon contains oxygen below 12 PPMA, more preferably below 11 PPMA, and most preferably below 10 ppma. This is because in wafers with medium to high oxygen content In the case of 14H pp g A, the generation of stacking errors and the enhancement of oxygen clustering bands inside the V / I boundary become increasingly important. They are each a potential source of problems in known integrated circuit manufacturing processes. The effect of the enhanced oxygen cluster can be further reduced by various methods, which can be used alone or in combination. For example, the oxygen precipitation core generally forms silicon with an annealing temperature of about 350 C to 750 ° C. Therefore, for some applications, Preferably, the crystal is a "short" crystal, that is, a crystal has been grown in the CzochralsKi process until the seed end has cooled from the melting point of silicon (about 141.0%) to about 750 ° C, after which the ingot is Cools quickly. In this way, the time spent in the temperature range that is important for core generation is kept to a minimum, and the core of the oxygen sink has unsuitable time to form in the crystal pulling single crystal machine. Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Health (please read the notes on the back before filling this page). However, the oxygen precipitation core formed during the single crystal growth period is best dissolved by annealing single crystal crushing, assuming it The core of the oxygen sink can be annealed by rapidly heating the silicon to a temperature of at least about 875x without performing a stable heat treatment, and the temperature is preferably continuously increased to at least 100 (rc, at least; no (rC, or more) High. Before silicon reaches i00〇r, almost all (for example, more than 99%) of these defects have been annealed, it is important that the wafer system is fast and heated to the temperature, that is, the temperature rise rate is at least about every minute 10: (: and preferably at least about 5 (rc) per minute, otherwise the knife or all oxygen precipitation cores will be stabilized by heat treatment. The equilibrium seems to reach within a short period of time, that is, about 60 seconds or less, Based on this, Shan-23- the official bid winning rate (CNS) eighty four secrets (210 χ29ϋ ^ &quot;-the work of the Central Standards Bureau of the Ministry of Economic Affairs Consumer Cooperatives printed 536562 A7 ____ __ B7 V. Description of the invention (21) The core of oxygen precipitation in Shi Xi can be annealed At least about 875 ° C and dissolving, preferably at least about 950 ° C, and ideally at least about 1100 ° C, with a period of at least about 5 seconds, and ideally at least about 10 minutes. Dissolution can be in a general oven or Implemented in a rapid thermal annealing (RTA) system, rapid thermal annealing of silicon can be performed in a variety of general rapid thermal annealing (RrA) furnaces, where the wafers are individually heated using a row of high-power lamps The RTA oven can quickly heat a silicon wafer. For example, it can heat a wafer from the chamber to 1200 C in a few seconds. One of the RTA ovens is a Model 610 oven from the AG Association (Scenic Mountain, California). In addition, The dissolution can be performed on a silicon block or a silicon wafer, and is preferably a wafer. In one example of the process of the present invention, the initial concentration of silicon row interstitial atoms is controlled in the axially symmetrical row interstitial of the ingot 10 as the main region. Within 6, refer to Figure 1. In general, the initial concentration of silicon interstitial atoms is achieved by controlling the crystal growth rate v and the average axial temperature gradient gq, so that the v / Gg ratio 値 is closer to the critical threshold 値, Where the ν / ι boundary occurs. In addition, the average axial The degree gradient can build jl, so that the change of Gg and G.⑴ (and V / G.⑴) become a variation of G. (that is, v / G.) 'Which becomes a function of the ingot radius and is also controlled. In another example of the present invention, v / Go is also controlled so that at least a part of the length of the ingot does not have a V / I boundary along the radius. In this length, the silicon system has a lattice from the center to the edge of the ring side. Vacancy is the main one, and the defect of condensed crystal lattice is to control v / G0 to avoid grams in an axially symmetrical area extending radially inward from the side edge of the ingot ring, that is, the growth conditions are controlled. In order to make WGg have a threshold between v / Gg and 乂 乂 times. Critical 値. It should be noted that the wafers prepared in accordance with the present invention are suitable as its ---; -------- (Please read the precautions on the back before filling this page)

、1T —24- 經濟部中央標準局負工消费合作社印繁 536562 A7 r-— ---- B7 五、發明説明(22) —'~~&quot; '一 - 上可供積置-上表層,表層積置可利用習知技術進行。、 1T —24- Working and Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs, Yinfan 536562 A7 r ----- B7 V. Description of Invention (22) — '~~ &quot;' 一-Available for accumulation on the upper surface The surface layer deposition can be performed using conventional techniques.

〃此外,亦應注意的是,依本發明製備之晶圓適合結W 氫或氬退火處理,如第503816 A1號歐洲專利申請案所示之 處理方式。 凝結瑕施之目視彳貞測 一 一 2 凝結之瑕疵可利用多種不同之技藝偵測,例如流線式瑕 疵或D形瑕疵通常利用在— Secc〇蚀刻液中蝕刻單晶矽樣品 大約30分鐘,及随後令樣品做顯微鏡檢視而偵測之(例如參 閲 H. Yamagishi 等人在 1992年.之 Semic〇nd % τ_η〇ι 7 A135作品所示p雖然對凝結晶格瑕疵之偵測係屬標準, 此方法亦可用於偵測凝結之行間充瑕疵,使用此技藝時, 此瑕戚若有存在則會在樣品表面上呈現爲較大之孔洞。 凝結瑕疵亦可利用雷射散射技藝而偵測,例如雷射散射 X光斷層攝影,其比較於其他蝕刻技藝而具有一低瑕魏密 度之偵測限制。 此外,凝結内在點塚疵可藉由將諸瑕疵飾以一金屬而做 目視偵測,該金屬可在加熱時擴散入單晶矽基質内。特別 是單晶矽樣品,例如晶圓、晶體片或緩動環,若有瑕疵存 在則可先以含有金屬之組合物塗於樣品表面而做目視檢視 ’该金屬可飾出諸瑕戚’而該組合物例如可爲^ __濃縮之石肖 酸銅溶液;塗覆之樣品隨後ia熱至大約900。〇與1 〇〇(TC之間 長達大約5至1 5分鐘,以利擴散金屬至樣品内;熱處理後之 樣品再冷却至室溫’使金屬逐漸到達超飽和且沉殿在含有 瑕戚之樣品基質内。 _2 5_ 度適用中國國家標準(CNS ) A4規格(210X297公釐) '' : ~ (請先閲讀背面之注意事項再填寫本頁) 、11 經濟部中央標準局眞工消f合作社印製 536562 A7 __________B7 五、發明説明(23) 冷却後’樣品先進行一無瑕疵描繪蝕刻,以利去除表面 歹欠留物與沉殿物,其係以_發亮钱刻液處理樣品大約8至12 刀鐘,典型之發壳蝕刻液包含大约55%硝酸(7〇%溶液重量 比)、大約20%氫氟酸(49%溶液重量比)、及大約25%氫氯酸( 濃縮液)。 , 樣品隨即以去離子水清洗及利用將樣品浸入或處理以一 SeCC0或Wright蝕刻溶液大約35至55分鐘,而進行第二次蝕 刻步驟,通常樣品係以Secc〇蚀刻溶進行蝕刻,其包含 比率之0.15莫耳重鉻酸鉀與氫氟酸(49%溶液重量比),此蝕 刻步驟用於發現或描繪出可能存在之凝結瑕疵。 大體上,供凝結瑕疵之行間充與晶格空位爲主之材料區 可相互區別,以及利用上述之銅飾技術而區別於含有凝結 瑕疵之材料,無瑕疵行間充爲主材料區並未含有蝕刻所現 之塗飾特性,而無瑕疵晶格空位爲主之材料區(在高溫氧核 溶解處理之前)則因氧核之銅飾而含有小蚀孔。 定義 . 本文所用之以下措辭或語句係具有既定之意旨:「凝結 内在點瑕疵」係指瑕疵來自於⑴晶格空位凝結而產生〇形 瑕疵、流線形瑕疵、閘氧化積體瑕疵、晶體原始粒子瑕疫 、日Ei m原:icr光點瑕痴、及其他此類晶格空位相關瑕症之反 應,或(ii)行間充凝結而產生錯位環與網絡、及其他此續^一 間充相關瑕疵之反應;「凝結行間充瑕疵」係指凝結之内 在點瑕痴來自於石夕行間充原子凝結之反應;「凝、纟士 P格卢 歸」係指凝結晶格空位娜由晶格空位凝結之反: fm tmtmmt tamml fm im 11&gt;1« νϋϋ ϋ—l mi «m . (請先閱讀背面之注意事項存填寫本頁) 訂 -26- 536562 經濟部中夾標準局貞工消費合作社印製 A7 B7 五、發明説明(24) 〜-- 產生,「半杈」係指自一晶圓或錠塊之中心軸線至一環側 緣所測%^距離;「大致無凝结&amp;在點瑕痛」係指—小於 諸瑕疵偵測限制之凝結瑕疵濃度,大約爲1〇3瑕疵/公分3 ; 「V/I邊界」係指沿著一錠塊或晶圓半徑之位 料即由晶格空位爲主」變爲行間充爲主;及卞晶= 王」與「行間充爲主」係指内在點瑕疵主要分別爲晶格空 位或行間充。 ' 範例 如以下I範例所示,本發明係提供一種製備一單矽晶塊 之方法,當錠塊依d〇vhtsldki方法而自固化溫度冷却時, 内在點瑕疵之凝結可避免發生於供裁切成晶圓之錠塊固定 直徑段之軸向對稱區内。 以下範例即載明一組可用於達成所需結果之條件,另有 若干方式可用於決定一既有晶體拉單晶機之理想拉單晶率 輪廓,例如不同於以多種拉單晶率生長一系列錠塊的是, 單to體可利用沿著晶體長度而增減之拉單晶率生長。在 此方式中’&amp;疋結之行間充瑕戚在單晶體生,長期間會出現與 消失多次,然後即可在多個不同之晶體位置上決定最佳之 拉單晶率’據此’以下範例並非自我設限之型式。 範例1 用於一具有預設爇區設計之拉單晶機 之最佳程序 一第一 200 mm單晶砍塊係在以下條件下生長,即拉單晶 率在晶體長度上自大約0.75 mm/min呈斜線至〇 35 mm/min, -27- 本紙張尺度適用中國國家標準(CNS ) A4規格(210X 297公釐) (請先閱讀背面之注意事項再填寫本頁) 訂 536562 A7 B7 五、發明説明(25 ) 圖7揭示拉單晶率爲晶體長度之函數,請考量在拉單晶機中 一生長200 mm錠塊之預先建立軸向溫度輪廓及平均軸向溫 度梯度G。中之預先建立徑向變化,即在容態/固態介面處之 軸向溫度梯度,諸拉單晶率·經選定以確使錠塊一端自其中 心至邊緣處呈晶格空位爲主之材料,而在錠 '塊/另端a則自其 中心至邊緣處呈行間充爲主之材料。生成之錠塊做縱向裁 切,且經分析以判斷凝結行間充瑕疵開始生成之處。 圖8係在一系列氧沉澱熱處理以發現瑕疵分佈圖形後,掃 描一錠塊自肩部起大約635 mm至760 mm—段之軸向切片少 數載體壽命所產生之影像;在大約680 mm之一晶體位置處 ,可看見一帶狀凝結行間充瑕疵28,此位置相當於V*(680 mm)=0.3 3 mm/min之臨界拉單晶率。在此點處,軸向對稱區 6之寬度(此區係行間充爲主材料區,但是無凝結之行間充 瑕疵)在其最大値,而晶格空位爲主區8之寬度Rv*(680)係大 約35 mm,且軸向對稱區之寬度1^*(680)係大約65 mm。 經漪部中央標準局負工消费合作社印製 (請先閱讀背面之注意事項再填寫本頁) 一列四牧之單晶矽塊隨後以穩定之拉單晶率生長,其係 略大於及略小於取得第一 200 mm鍵塊軸向對稱區最大寬度 時之拉單晶率。圖9揭示拉單晶率爲標號1-4之各牧晶體之 晶體長度函數^此四牧晶體隨後分析以判斷凝結式行間充 瑕疵最先出現或消失之軸向位置(及相對應之拉單晶率), 此四牧實驗判斷而得之點(€上*記號)示於圖9,在諸點之 間插値及自諸點外插即產生一曲線,標示爲圖9中之v*(Z) ,此曲線代表一第一近似値,即做爲拉單晶機中長度函數 之200 mm晶體之拉單晶率,此時之軸向對稱區在其最大寬 -28- 本纸張尺度適用中國國家標準(CNS ) A4規格(210X 297公釐) 度 經濟部中央標準局具工消费合作社印製 536562 五、發明说明(2δ ) 不同拉單晶率之其他晶體之生長以及這些晶體之進一步 分析可進一步推敲出ν *(Ζ)之實驗定義。 範例 G0(r)徑向變化之減低 ---' 圖10、11説明品質之改良,係藉由熔態/固態介面軸向溫 度梯度中之徑向變化G〇 (r)減少而達成,晶格空位與行間充 之初期濃度(大約距離熔態/固態介面i cm處)係依二個不同 例子而有不同之G〇⑺:⑴G0 (r)=2.65+5 X l〇-V/(K/mm)及⑵ G0 (〇=2·65 + 5 X 10-5rV(K/mm),每一例子之拉單晶率可經調 整以令冨含晶格空位之矽與富含行間充之矽之間邊界爲3 cm半徑。用於例子卜2之拉單晶率分別爲〇4、〇35mm/min 。圖11中可清楚看出晶體富含行間充段内之行間充初期濃 度係隨著初期軸向溫度梯度中之徑向變化減少而呈戲劇性 地減Λ即導致材料品質改善,因爲其較易於避免行間 充瑕疵群因行間充之超飽和而生成。 範例3 增加行間充之擴散時間 圖13 Λ明質(改良’其可藉由增加行間充擴散時 達成仃間充辰度依二個不同例子而有不同之晶體 向溫度輪廓dT/dz,溶態:/固態介面處之軸向溫度梯度 能11子自相同,.因此,行間充之初期濃度(大約距離熔 :固態介面w處)對二例子皆相同。在此範例中,拉單 ^可調整以使整個晶體富含行間充,二例之拉單晶率同 (請先閱讀背面之注意事項再填寫本頁)〃 In addition, it should also be noted that the wafer prepared according to the present invention is suitable for the junction hydrogen or argon annealing treatment, as shown in the European Patent Application No. 503816 A1. Visual inspection of condensing defects 彳 彳 2 Condensing defects can be detected using a variety of different techniques, such as streamline defects or D-shaped defects are usually used to etch single crystal silicon samples in Secc0 etching solution for about 30 minutes, And then let the sample be inspected by microscopy (see, for example, H. Yamagishi et al., 1992. Semic 0nd% τ_η〇ι 7 A135 works), although the detection of crystal lattice defects is a Standard, this method can also be used to detect condensate filling defects. When using this technique, if there is such a defect, it will appear as a large hole on the surface of the sample. Condensation defects can also be detected using laser scattering technology. Compared with other etching techniques, laser scattering X-ray tomography has a detection limit of low defect density. In addition, the condensed intrinsic point defects can be visually detected by decorating the defects with a metal. It is measured that the metal can diffuse into the single crystal silicon matrix when heated. Especially for single crystal silicon samples, such as wafers, crystal wafers or slow-moving rings, if there is a flaw, the sample can be coated with a metal-containing composition first. table While doing a visual inspection, 'the metal can decorate various flaws' and the composition can be, for example, a concentrated copper schistate solution; the coated sample is then heated to approximately 900 ° C and 100 ° C (TC It lasts about 5 to 15 minutes to facilitate the diffusion of metal into the sample; the heat-treated sample is cooled to room temperature again to allow the metal to gradually reach supersaturation and sink in the sample matrix containing the flaw. _2 5_ degree Applicable to China National Standard (CNS) A4 specification (210X297 mm) '': ~ (Please read the notes on the back before filling this page), 11 Printed by the Central Standards Bureau of the Ministry of Economic Affairs, Fong Cooperative Cooperative 536562 A7 __________B7 V. Description of the invention (23) After cooling, the sample is first subjected to a flawless drawing and etching to remove the surface leftovers and sunken objects. It is about 8 to 12 knife clocks for processing the sample with _Fengliang money carving solution, typically Hair shell etchant contains approximately 55% nitric acid (70% solution weight ratio), approximately 20% hydrofluoric acid (49% solution weight ratio), and approximately 25% hydrochloric acid (concentrated solution). The sample is then deionized Water washing and utilization. Immerse or treat the sample with a SeCC0 or Wrigh The etching solution is about 35 to 55 minutes, and the second etching step is performed. Usually, the sample is etched with Secc0 etching solution, which contains a ratio of 0.15 mole of potassium dichromate to hydrofluoric acid (49% solution weight ratio). This etching step is used to find or depict possible condensation defects. In general, the row filling for condensation defects and the material area dominated by lattice vacancies can be distinguished from each other, and the above-mentioned copper decoration technology is used to distinguish them from condensation. Defective materials, non-defective interstitial filling is the main material area and does not contain the coating characteristics present in the etching, while the non-defective lattice vacancy-dominated material area (prior to the high-temperature oxygen nucleation dissolution treatment) is due to the copper decoration of the oxygen nucleus. Contains small pits. Definition. The following wording or phrase used in this article has the intended meaning: "condensation intrinsic point defect" means that the defect comes from the condensing of the ⑴ lattice vacancies, resulting in 0-shaped defects, streamline defects, gate oxide accumulation defects, crystal original particles Defects, Ei m original: ICR light spot defects, and other such lattice vacancy-related defects, or (ii) interstitial condensation caused by dislocation rings and networks, and other continuations Defect response; "Condensation between rows of defects" refers to the reaction of the intrinsic point defect of condensation from Shi Xixing's interstitial atomic condensation; "Condensation and warrior P Gurugui" refers to the vacancy of the condensed crystal lattice. Condensing the opposite: fm tmtmmt tamml fm im 11 &gt; 1 «νϋϋ ϋ—l mi« m. (Please read the notes on the back and fill in this page first) Order-26- 536562 Printed by Zhengong Consumer Cooperative, Standards Bureau, Ministry of Economic Affairs System A7 B7 V. Description of the invention (24) ~-Produced, "half branch" refers to the% ^ distance measured from the center axis of a wafer or ingot to the side edge of a ring; "roughly no condensation &amp; "Pain" means-less than the limits of flaw detection The concentration of condensation defects is about 103 defects / cm 3; "V / I boundary" means that the material along the radius of an ingot or wafer changes from lattice vacancies to interstitial filling; And "Jingjing = King" and "mainly interstitial" mean that the internal point defects are mainly lattice vacancies or interstitials. '' For example, as shown in the following I example, the present invention provides a method for preparing a single silicon crystal block. When the ingot is cooled at the self-curing temperature according to the dovhtsldki method, the condensation of intrinsic point defects can be avoided for cutting. The axially symmetrical region of the fixed diameter section of the ingots forming the wafer. The following example sets out a set of conditions that can be used to achieve the desired result. There are several other methods that can be used to determine the ideal single crystal rate profile of an existing crystal pulling single crystal machine. In the series of ingots, a single to body can be grown using a pull single crystal rate that increases or decreases along the length of the crystal. In this way, '&amp; knots are filled with defects in the single crystal, which will appear and disappear multiple times over a long period of time, and then the best pull single crystal rate can be determined at a number of different crystal positions' based on this' The following examples are not self-limiting. Example 1 Optimal procedure for a single crystal pulling machine with a preset chirped area design-The first 200 mm single crystal chopping block is grown under the following conditions, that is, the single crystal pulling rate is from about 0.75 mm / Min is oblique to 〇35 mm / min, -27- This paper size applies to Chinese National Standard (CNS) A4 (210X 297 mm) (Please read the precautions on the back before filling this page) Order 536562 A7 B7 V. Description of the invention (25) FIG. 7 reveals that the pulling single crystal is a function of crystal length. Please consider the pre-established axial temperature profile and average axial temperature gradient G of a 200 mm ingot grown in a pulling single crystal machine. The radial change in the medium, that is, the axial temperature gradient at the capacitive / solid interface, and the pull single crystal rate are selected to ensure that one end of the ingot is mainly dominated by lattice vacancies from its center to the edge. On the ingot's block / another end a, the material is mainly filled from the center to the edge. The generated ingots are cut longitudinally and analyzed to determine where the condensed line defects begin to form. Figure 8 is an image produced by scanning the life of a small number of carriers in the axial section of a segment approximately 635 mm to 760 mm from the shoulder after a series of oxygen precipitation heat treatments to find the defect distribution pattern; one of approximately 680 mm At the crystal position, a band-shaped condensed interstitial filling defect 28 can be seen. This position is equivalent to a critical pulling single crystal rate of V * (680 mm) = 0.3 3 mm / min. At this point, the width of the axially symmetric region 6 (this region is filled with the main material region, but without condensation between the row filled defects) is at its maximum, and the lattice vacancy is the width of the main region 8 Rv * (680 ) Is about 35 mm, and the width of the axially symmetric region 1 ^ * (680) is about 65 mm. Printed by the Central Standards Bureau of the Ministry of Economic Affairs and Consumer Cooperatives (please read the precautions on the back before filling out this page). A row of Simu's monocrystalline silicon blocks was subsequently grown at a stable pull single crystal rate. Single crystal pulling rate at the maximum width of the axially symmetric region of the first 200 mm key block. Figure 9 reveals the crystal length function of each crystal with a single crystal rate of 1-4. The four crystals are then analyzed to determine the axial position where the condensing interstitial defects appear or disappear first (and the corresponding single crystal). Crystal rate), the points obtained by the four animal husbandry experiments (indicated by the * mark) are shown in FIG. 9, a curve is generated by inserting between the points and extrapolating from the points, which is marked as v * in FIG. 9 (Z), this curve represents a first approximation, that is, the pull single crystal rate of a 200 mm crystal as a function of the length in a single crystal machine. At this time, the axial symmetry region is at its maximum width -28- The scale is applicable to the Chinese National Standard (CNS) A4 specification (210X 297 mm). It is printed by the Central Standards Bureau of the Ministry of Economic Affairs and printed by the Consumer Cooperative. 536562 5. Description of the invention (2δ) The growth of other crystals with different single crystal rates and the growth of these crystals Further analysis can further delineate the experimental definition of ν * (Z). Example G0 (r) Reduction of Radial Change --- 'Figures 10 and 11 illustrate the improvement of quality, which is achieved by reducing the radial change G0 (r) in the axial temperature gradient of the molten / solid interface. The initial concentration of the lattice vacancies and the interspace charge (approximately at the distance of 1 cm from the molten / solid interface) are different according to two different examples. G⑺: ⑴G0 (r) = 2.65 + 5 X l0-V / (K / mm) and ⑵ G0 (〇 = 2.65 + 5 X 10-5rV (K / mm), the pull single crystal rate of each example can be adjusted to make the silicon with lattice vacancies and The boundary between silicon is a radius of 3 cm. The pull single crystal rates used in Example 2 are 〇04 and 〇35mm / min. It can be clearly seen in Figure 11 that the initial concentration of row filling in the crystal-rich row filling section varies with As the radial change in the initial axial temperature gradient decreases, the dramatic decrease in Λ leads to an improvement in material quality, because it is easier to avoid the formation of interstitial defect groups due to the supersaturation of interstitial charging. Example 3 Increasing the diffusion time of interstitial charging Figure 13 Λ Brightness (improved 'It can be achieved by increasing the interstitial charge and diffusion when the interstitial degree is different. There are two different examples of the crystal to the temperature wheel. dT / dz, the axial temperature gradient energy at the dissolved state: / solid interface can be the same. Therefore, the initial concentration of row charging (approximate distance from the melting point: solid interface w) is the same for both examples. In this example The pull single ^ can be adjusted so that the entire crystal is rich in row charge. The pull single crystal rate is the same in both cases (please read the precautions on the back before filling this page)

----____ _29- 536562 A7 B7 五、發明説明(27 ) 樣爲0.32 mm/min,例2中之行間充向外擴散時間較長則會 造成行間充濃度之整體下降,此導致材料品質之改善,因 爲其較易於避免行間充瑕疵群因行間充之超飽和而生成。 範例4 一枚700 mm長且150 mm直徑之晶體係以一:變動之拉單晶 率生長,拉單晶率呈近乎直線狀自肩部以1.2 mm/min變化 成距離肩部430 mm處之0.4 mm/min,然後再以呈近乎直線 狀回到距離肩部700 mm處之0.65 mm/min,在此條件下之該 特定拉單晶機中,整個半徑即在晶體長度上依富含行間充 條件生長於距離晶體肩部大約320至525 mm處。參閱圖14, 在大約525 mm之一軸向位置處且大約0.47 mm/min之拉單晶 率,晶體於跨越整個直徑上並無凝結之内在點瑕疵群,易 言之,有某一小段之晶體中,軸向對稱區之寬度係等於錠 塊之半徑,即大致無凝結瑕疵瑕疵區域。 範例5 經濟部中央標準局負工消费合作社印裝 (請先閱讀背面之注意事項再填寫本頁) 如範例1所示,一列岸晶秒塊係以變動之拉單晶率生長且 隨即分析以判斷凝結行間充瑕疵最先出現或消失處之軸向 位置(及對應之拉單晶率),在諸點之間插値及自諸點外插 而繪示於拉單晶率v、軸向位置之圖表上即產生一曲線, 其代表第一近似値,即做爲拉單晶機中長度函數之200 mm 晶體之拉單晶率,此時之軸 &gt; 對稱區在其最大寬度。其他 晶體以不同之拉單率生長,且諸晶體之進一步分析可用於 推敲此實驗性判斷之最佳拉單晶率輪廓。 使用此資料及依循此最佳拉單晶率輪廓,則一大約1000 - 30- 本纸張尺度適用中國國家標準(CNS ) Λ4規格(210X 297公釐) 536562 mm 經濟部中央標準局負工消资合作社印製 A7 B7 五、發明説明(28 ) 長度且200 mm直徑之晶體即可生長,自不同軸向位置取 得之生長晶體切片隨即利用此技藝中之氧沉澱法標準加以 分析,以利⑴判斷是否有凝結行間充瑕疵形成,(ii)判斷 V/I邊界之位置,其爲切片半徑之函數。依此,可判斷出一 軸向對稱區之存在,以及做爲晶體長度或位覆〜函數+之此區 域寬度。 自錠塊肩部大約200 mm至950 mm範圍之軸向位置所取得 之結果係如圖15圖表所示,諸結果顯示一拉單晶率輪廓可 經決定以用於一單晶矽塊之生長,使得錠塊之固定直徑段 可包含一具有一寬度之軸向對稱區,寬度係自環側緣以徑 向朝向錠塊中心軸線測得,至少大約爲固定直徑段半徑長 度之40%。此外,諸結果顯示此軸向對稱區可具有一長度 ,係沿錠塊之中心軸線測得,其大致爲錠塊固定直線段長 度之75%。 範例6 一具有大約1100 miQ長度且15〇 mm直徑之單晶矽塊係以 減小之拉單0曰率生長,在錠塊固定直徑段肩部之拉單晶 率大約1 mm/min,拉單晶率呈指數性減小至大約〇 4 mm/nun,此相當於距離肩部大約2〇〇111111處之軸向位置,拉 單晶率隨即呈直線性減小,直到大約〇 3咖⑽之速率到 達接近於錠塊固定直徑段末邊。----____ _29- 536562 A7 B7 V. Description of the invention (27) The sample is 0.32 mm / min. The longer outward diffusion time of the row charge in Example 2 will cause the overall decrease in the charge amount of the row charge, which leads to the quality of the material. The improvement is because it is easier to avoid the generation of inter-row filling defect groups due to super-saturation of inter-row filling. Example 4 A crystal system with a length of 700 mm and a diameter of 150 mm grows at a rate of varying single crystal growth rate. The single crystal growth rate is almost linear from the shoulder at 1.2 mm / min to a distance of 430 mm from the shoulder. 0.4 mm / min, and then returned in a nearly straight line to 0.65 mm / min from the shoulder 700 mm. Under this condition, in this particular single crystal pulling machine, the entire radius is rich in the length of the crystal. The condition is about 320 to 525 mm from the shoulder of the crystal. Referring to FIG. 14, at an axial position of about 525 mm and a pull single crystal rate of about 0.47 mm / min, the crystal has no condensed intrinsic point defect group across the entire diameter. In other words, there is a small section of In the crystal, the width of the axially symmetric region is equal to the radius of the ingot, that is, a region free of condensation defects. Example 5: Printed by the Central Standards Bureau of the Ministry of Economic Affairs and Consumer Cooperatives (please read the precautions on the back before filling out this page). As shown in Example 1, a row of shore crystal second blocks are grown at a variable pull single crystal rate and then analyzed to Determine the axial position (and the corresponding pull single crystal rate) at which the filling line defects appear first or disappear, and insert and extrapolate from the points to plot the single crystal rate v, axial direction A curve is generated on the position chart, which represents the first approximation, that is, the pulling single crystal rate of a 200 mm crystal as a function of the length in the pulling single crystal machine. At this time, the axis &gt; symmetry region is at its maximum width. Other crystals are grown at different pull rates, and further analysis of the crystals can be used to determine the optimal pull single crystal rate profile for this experimental determination. Using this information and following this best single crystal rate profile, a paper size of approximately 1000-30- applies to the Chinese National Standard (CNS) Λ4 specification (210X 297 mm) 536562 mm Cooperatives print A7 B7 V. Description of the invention (28) Crystals with a length of 200 mm in diameter can grow, and the growing crystal slices obtained from different axial positions are then analyzed using the oxygen precipitation method standard in this technology to facilitate It is judged whether there is a condensed line defect, (ii) the position of the V / I boundary is determined as a function of the slice radius. Based on this, the existence of an axially symmetric region and the width of this region as a function of the crystal length or potential coverage ~ function + can be determined. The results obtained from the axial position of the ingot shoulder in the range of approximately 200 mm to 950 mm are shown in the graph of Fig. 15. The results show that a single crystal rate profile can be determined for the growth of a single crystal silicon block. , So that the fixed diameter section of the ingot can include an axially symmetric region with a width measured from the side edge of the ring radially toward the central axis of the ingot, at least about 40% of the radius of the fixed diameter section. In addition, the results show that this axially symmetric region may have a length, measured along the central axis of the ingot, which is approximately 75% of the length of the fixed straight section of the ingot. Example 6 A single crystal silicon block having a length of about 1100 miQ and a diameter of 15 mm is grown at a reduced pull rate of about 0 mm. The pull single crystal rate at the shoulder of a fixed diameter section of an ingot is about 1 mm / min. The single crystal rate decreases exponentially to about 0.4 mm / nun, which is equivalent to the axial position at about 200111111 from the shoulder. The single crystal rate then decreases linearly until about 0.33 The rate reaches near the end of the fixed diameter section of the ingot.

在這些製程條件下之此特定散F 此付Ά £結構中,生成之錠塊含 有一區域,其中軸向對稱區具有 ^ 大約寺於錠塊半徑之寬 度。請即參閲圖10a、Mb,其係._ 八你在連_氧沉澱熱處理後 - 31- ^氏張尺度I用中國i?i^CNS ) (請先閲讀背面之注意事項再填寫本頁}Under these process conditions, the specific structure and the structure of the ingot, the resulting ingot contains a region, where the axially symmetric region has a width approximately equal to the radius of the ingot. Please refer to Fig. 10a and Mb, which are the following. _ BA You after the _ oxygen precipitation heat treatment-31- ^ Zhang Zhang scale I for China i? I ^ CNS) (Please read the precautions on the back before filling this page }

536562 A7 B7 五、發明説明(29 ) 掃描一段錠塊之軸向切片之少數載體壽命所生之影像,鍵 塊之連續分段即呈現,其範圍在大約1〇〇 mm至250 mm處及 250 mm至400 mm處之軸向位置中,由圖中可看出一區域存 在於鍵塊内,範圍在距離肩部大約170至290 mm之軸向位 置中,其在整個直徑上並無凝結之内在點瑕疵、易^言之, 一區域存在於錠塊内,其中軸向對稱區亦即大致上無凝結 行間充瑕疵之區域之寬度係大約等於錠塊之半徑。 此外,在一自大約125 mm至170 mm及大約自290 mm至 400 mm以上之軸向位置範圍内之區域中,有若干無凝結内 在點瑕疵之行間充爲主材料軸向對稱區圍繞於亦無凝結内 在點瑕痴之晶格空位爲主材料之一大致筒形核心。 最後,在大約100 mm至125 mm之一轴向位置範圍内之區 域中,有一無凝結瑕疵之行間充爲主材料之軸向對稱區圍 繞於晶格空位爲主材料之一大致筒形核心。在晶格空位爲 主之材料内,有一無凝結瑕戚之軸向對稱區圍繞於一含有 凝結晶格空位瑕疵之核心。 範例7 V/I邊界之冷却率及位置 經濟部中央標準局負工消费合作社印製 (請先閲讀背面之注意事項再填寫本頁) 一列單晶秒塊(150及200 mm額定直徑)係依Czochralski 方法生長,使一般技藝設計之不同熱區結構,以影響超過 大約1050°C溫度時之矽殘餘蛑間,各錠塊之拉單晶率輪廓 係沿著錠塊長度變化,試以產生一自凝結晶格空位點瑕疵 區至凝結行間充點瑕疵區之過渡。 一旦生長時,錠塊即沿著平行於生長方向之中心轴線做 -32- 本紙張尺度適用中國國家標準(CNS ) A4規格(210X 297公釐) 536562 A7 ^---________ B7 五、發明説明(30) : &quot; ^ 縱向裁切,且進—步分割成各爲大约2 mm厚度之數段,使 用則逑t銅飾技藝時,一組之此縱向段進行加敎且故意以 銅冷飾之,加熱條件則適合於一銅行間充之高濃度溶解。 在此項熱處理後,樣品即迅速冷却,其間銅斑係向外擴散 或沉澱在出現氧化物群或凝結行間充瑕疵之處—。在a一標準 又瑕疵描繪蝕刻後,樣品以目視檢查沉澱污斑之存在,那 些無沉澱污斑之區域即相當於無凝結行間充瑕疵之區域。 雕=一組縱向段則進行一系列之氧沉澱熱處理,以利於載 月且可命以圖表不之前先造成新氧化物群之形成核心與生長 。壽命圖表中之對比帶係用於判斷及測量各錠塊中不同軸 向位置處之瞬間熔態/固態介面形狀,然後使用熔態/固態 介面形狀上之資料,如下所述,以評估平均轴向溫度梯度 &lt;絕對値及徑向變化,此資料亦配合拉單晶率使用,以 評估v/Gq中之徑向變化。 經濟部中央標準局員工消費合作社印製 爲了進一步檢查生長條件在一單晶矽塊生成品質上之作 用,依據現有之實驗證明,據信有多種假設可用於判斷。 首先,爲了藉由冷却至發生行間充瑕疵凝結溫度所花之時 間以簡化熱歷史之處理,假設大約1〇5(rc爲發生矽行間充 凝結之溫度之合理近似値,此溫度似乎符合於使用不同冷 却率之實驗期間所觀察到之凝結行間充瑕疵密度變化。如 上所述,雖然凝結發生與否-亦爲行間充濃度之一項因素, 但疋相仏凝結將不致發生於l〇5〇°C以上溫度,因爲以 Czochralski式生長過程之行間充濃度而言,其可合理假設 系統在此溫度以上時不致變成行間充臨界超飽和。易言之 - 33- 本纸張尺度適用中國國家標準(CNS ) Λ4規格(210X297公釐) 536562 A7 B7 五、發明説明(31 ) ,以Czochralski式生長過程之典型行間充濃度而言,可合 理假^又系統不致,交成臨界超飽和,且因此在丨以上不 致發生凝結狀況。 第二項假設爲將單晶矽品質上之生長條件作用做成變數 ,而依附於矽行間充擴散率之溫度則省略不計、易。,言之, 其假設行間充係在大約14〇(rc與1〇5〇Ό之間之所有溫度下 ,以相同速率擴散,應瞭解到大約1050X:可視爲凝結溫度 &lt; 一合理近似値,此項假設之基本點在於自熔點起之冷却 曲’泉、、’田部並不重要,擴散距離僅取決於花費在自熔點冷却 至大約1050°c之總時間。 使用軸向溫度輪廓資料於各熱區設計及眞實拉單晶率輪 廓於一特足錠塊時,自大约l4〇〇°C至i〇5〇°C之總冷却時間 即可計算出來,應注意的是,各熱區之溫度變化率相當均 勻 匕員勻度思扎用於凝結行間充瑕痖之核成型溫度選擇 中 &lt; 任意誤差,例如大約1050Ό,將僅造成計算冷却時間 中之成比例誤差。 爲了決足鍵塊之晶格空位爲主區域之徑向範圍(R晶格空位) 經濟部中央標準局賀工消费合作社印製 ---*--------__ (请先閱讀背面之注意事項存填寫本頁) 或者軸向對稱區之寬度,其進一步假設由壽命圖表所決定 之叩格艾位爲主核心半徑係相等於v /G〇 = v/G〇臨界固化點 易口之轴向對稱區之寬度大致假設爲依據冷却至室溫 後IV/I邊界位置。如前所述'此係因爲當錠塊冷却時,晶 格文位與砂行間充之重組即可發生,當重組確實發生時, V/I邊界I眞實位置即向内移向錠塊之中心軸線,此即本文 所指之最終位置。536562 A7 B7 V. Description of the invention (29) Scanning the image of the life of a small number of carriers of the axial section of an ingot, the continuous segmentation of the key block is presented, and its range is about 100 mm to 250 mm and 250 In the axial position from mm to 400 mm, it can be seen from the figure that an area exists in the key block, and the range is in the axial position about 170 to 290 mm from the shoulder, which does not condense over the entire diameter. Intrinsic point defects, in other words, a region exists in the ingot, and the width of the axially symmetric region, that is, a region substantially free of condensation and interstitial defects, is approximately equal to the radius of the ingot. In addition, in a range of axial positions ranging from approximately 125 mm to 170 mm and approximately from 290 mm to 400 mm or more, there are a number of non-condensing intrinsic point defects between the rows filled with the axial symmetry zone of the main material. The lattice vacancies of the non-condensing intrinsic dot defect are roughly cylindrical cores as one of the main materials. Finally, in an area within an axial position range of approximately 100 mm to 125 mm, there is an axially symmetric region of the main material filled with non-condensing defects that surrounds the substantially cylindrical core of the main material. In the material where the lattice vacancies are dominant, an axially symmetric region without condensation defects surrounds a core containing the crystal lattice vacancy defects. Example 7 Cooling rate and location of the V / I boundary Printed by the Consumers Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs (please read the precautions on the back before filling this page) A series of single crystal seconds blocks (150 and 200 mm rated diameters) are based on The growth of the Czochralski method makes different hot zone structures designed by ordinary techniques to affect the residual silicon content at temperatures exceeding about 1050 ° C. The pull single crystal rate profile of each ingot changes along the length of the ingot. The transition from the defect area of the coagulation crystal lattice vacancy point to the filling defect area between the condensation rows. Once grown, the ingot is made along the central axis parallel to the growth direction. -32- This paper size applies the Chinese National Standard (CNS) A4 specification (210X 297 mm) 536562 A7 ^ ---________ B7 V. Invention Note (30): ^ Longitudinal cutting, and further divided into several sections with a thickness of about 2 mm each. When using the copper decoration technique, a set of this longitudinal section is added and intentionally copper For cold decoration, the heating conditions are suitable for high concentration dissolution of a copper row. After this heat treatment, the sample was rapidly cooled, during which the copper spots spread outward or precipitated where there were defects in oxide groups or condensation between rows— After the a-standard and defect-delineation etching, the sample was visually inspected for the presence of precipitated stains, and those areas without precipitated stains corresponded to the areas without condensation and filled with defects. Carved = a series of longitudinal sections are subjected to a series of oxygen precipitation heat treatment to facilitate the moon load and can be caused to form the core and growth of new oxide groups before the chart. The comparison band in the life chart is used to judge and measure the instant melt / solid interface shape at different axial positions in each ingot, and then use the data on the melt / solid interface shape as described below to evaluate the average axis The temperature gradient &lt; absolute chirp and radial change, this data is also used in conjunction with the pull single crystal rate to evaluate the radial change in v / Gq. Printed by the Consumers' Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs. In order to further examine the role of growth conditions in the quality of a monocrystalline silicon block, according to existing experiments, it is believed that there are a variety of assumptions that can be used for judgment. First of all, in order to simplify the processing of thermal history by cooling the time it takes to reach the condensation temperature of inter-row filling defects, it is assumed that approximately 105 (rc is a reasonable approximation of the temperature at which silicon inter-row filling condensation occurs). This temperature seems to be consistent with the use Changes in density of interstitial filling defects observed during the experiment with different cooling rates. As mentioned above, although condensation occurs or not-it is also a factor of interstitial concentration, but the phase condensate will not occur at 105. The temperature above ° C, because of the row intercalation concentration in the Czochralski-type growth process, it can reasonably be assumed that the system does not become critical supersaturation of the intercharging when it is above this temperature. In other words-33- This paper scale applies Chinese national standards (CNS) Λ4 specification (210X297 mm) 536562 A7 B7 V. Description of the invention (31) In terms of the typical row interstitial concentration of the Czochralski-type growth process, it can be reasonably false and the system is not inconsistent, and it becomes a critical supersaturation, and therefore Above 丨 no condensation will occur. The second assumption is to change the growth conditions on the quality of single crystal silicon into a variable and rely on the interstitial diffusion of silicon. The temperature is omitted and easy. In other words, it assumes that the interstellar charge is diffused at the same rate at all temperatures between about 140 ° C and 1050 ° F. It should be understood that about 1050X: can be considered as condensation Temperature &lt; A reasonable approximation 値, the basic point of this assumption is that the cooling tunes 'spring,' and 'field part' from the melting point are not important, and the diffusion distance depends only on the total time spent cooling from the melting point to about 1050 ° C. When using the axial temperature profile data to design in each hot zone and the profile of single solid crystal rate in a special foot ingot, the total cooling time from about 140 ° C to 105 ° C can be calculated. Note that the temperature change rate of each hot zone is quite uniform. The uniformity of the dagger uniformity is used to select the core molding temperature for condensing defects. Arbitrary error, such as about 1050 °, will only result in the success of calculating the cooling time. Proportional error. In order to determine the radial range of the main area of the lattice vacancies of the key block (R lattice vacancies), printed by He Gong Consumer Cooperative, Central Standards Bureau of the Ministry of Economic Affairs ---------------__ ( (Please read the precautions on the back and fill in this page) The width of the axially symmetric region is further assumed to be determined by the life graph. The main grid radius is equal to v / G0 = v / G〇. The width of the axially symmetric region is approximately equal to the critical solidification point. It is assumed to be based on the IV / I boundary position after cooling to room temperature. As mentioned earlier, this is because when the ingot is cooled, the reorganization of the lattice level and the interstellar filling can occur. When the reorganization does occur, V / The I boundary is the inward position, which is the inward direction of the ingot, which is the final position referred to herein.

(匸奶)/\4規格(210\ 297公釐) 536562 經濟部中央標準局貝工消费合作社印製 A7 B7 五、發明説明(32) 欲簡化固化時間中之晶體平均軸向溫度梯度G〇,熔態/固 態介面形狀係假設爲熔點等溫線,晶體表面溫度係利用有 限凡素模塑(FEA)技藝與熱區設計細節計算,晶體内之整個 /皿度範圍以及G係以正確之邊界條件解出Lapiace程式而推 淪出,即沿著熔態/固態介面之熔點及沿著晶-體/軸線a表面溫 度I FEA結果,自其中一製備且評估過之錠塊不同軸向位 置處所取得之結果即如圖丨7所示。 欲估算G。中徑向變化在初期行間充濃度上之作用時,一 徑向位置R,,亦即V/I邊界與晶體表面之間中途之位置,其 可假設爲一矽行間充距離錠塊中一槽溝之最遠點,而槽溝 在w格工位爲主區内或在晶體表面上。藉由使用生長率與 G0値於上述錠塊,在位置及,之計算所得v/G。與在ν/ι邊界之 v/Gg之間差異(即臨界v /G〇値)提供初期行間充濃度中之徑 向變化指示,以及其用於行間充上之能力,以到達晶體表 面上或晶格空位爲主區域内之一槽溝。 對於此特足之資料钽而言,似乎在v /G。内之徑向變化上 並無晶體品質之系統式依存性,如圖18所見,錠塊中之軸 向依存性在此樣品中爲最小,在此實驗系列中所涉及之生 長條件代表G〇徑向變化中之一相當窄之範圍。結果,此資 料組太窄而無法解出Gq徑向變化上之可辨識依存性(即一 凝結内在點瑕疵帶之有無)-。· 如上所述,所製備之各錠塊樣品係在凝結行間充瑕疵存 在或不存在時之不同軸向位置求値,對於各檢查過之軸向 t置而Q 了在樣品品質與軸向對稱區寬度之間進行其相 (請先閱讀背面之注意事項再填寫本頁)(匸 奶) / \ 4 specifications (210 \ 297 mm) 536562 Printed by the Baker Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the invention (32) To simplify the average axial temperature gradient G of the crystal in the curing time. The melting / solid interface shape is assumed to be the melting point isotherm. The surface temperature of the crystal is calculated using finite vane molding (FEA) technology and hot zone design details. The entire range / degree range in the crystal and G are calculated correctly. The boundary conditions are solved by the Lapiace formula, that is, the melting point along the melting / solid interface and the surface temperature I FEA along the crystal-body / axis a. The different axial positions of the ingots prepared and evaluated from one of them The results obtained on the premises are shown in Figure 丨 7. To estimate G. When the effect of the mid-radial change on the initial row charge density is a radial position R, that is, the position halfway between the V / I boundary and the crystal surface, it can be assumed that a silicon row charge distance is a slot in the ingot. The furthest point of the groove, and the groove is in the main zone of the w-lattice station or on the crystal surface. Calculate v / G by using the growth rate and G0 in the above ingots at the position and. The difference from v / Gg at the ν / ι boundary (ie, critical v / G〇 値) provides an indication of the radial change in the initial row interstitial concentration and its ability to be used to interstitially reach the crystal surface or The lattice vacancy is a groove in the main region. For this particular material, tantalum seems to be at v / G. There is no systematic dependence on the crystal quality in the radial change within. As seen in Figure 18, the axial dependence in the ingot is the smallest in this sample. The growth conditions involved in this experimental series represent the G0 diameter. One of the changes is quite narrow. As a result, this data set is too narrow to resolve the discernable dependencies on radial changes in Gq (ie, the presence or absence of a condensed intrinsic point defect band)-. · As mentioned above, the prepared ingot samples are obtained at different axial positions when the condensing line is full of flaws or not. Q is set for each checked axial t and the sample quality is symmetrical with the axial direction. Compare the width of the zones (please read the precautions on the back before filling this page)

536562 A7 ----------B7 五、發明説明(33) 關性。請參閲圖19,可製成一圖表 “、λ α ν既有樣品之品質比 Γ μ軸向位置時自固化溫度冷却至刪。^時 請 先 閱 讀 背 φ 注 意 事 項 再 填 寫 本 頁 R ΠΓ:望者,此圖表顯示出轴向對稱區之寬度(即 。晶雜謝山有-強烈之依存性於此特定溫度範園内之樣 品冷却歷史上。爲使軸向對稱區寬度増大,其一趨勢:建議吾 人增長擴散時間或減緩冷却率係必要的。 根據此圖表中之資料,可計笪屮 σ T f异出—^配之線,代表碎 印貝中自「艮」(即無瑕病)至「劣」(即含瑕旬之變化過渡 ,其A此特疋溫度範園内—既定鍵塊直徑所容許之冷却時 間之函數,軸向對稱區與冷却率之間之大致關: 程式表示: … 卜 (R 晶體-R過渡)2 =Deff*t105(rc 其中 R晶雜係錠塊之半徑, R過渡係在樣品内一軸向位置處,行間充爲主材料内發生 由無瑕疵至有瑕疵之過渡之軸向對稱區半徑,或反之亦然 j 經濟部中央標準局負工消费合作社印製536562 A7 ---------- B7 V. Description of the invention (33) Relevance. Please refer to Figure 19, a graph can be made, “λ α ν Existing sample quality ratio Γ μ At the axial position, the self-curing temperature is cooled to delete. When ^, please read the back φ precautions before filling in this page R ΠΓ : Wanter, this chart shows the width of the axially symmetric region (ie. Crystal Xieshan has a strong dependence on the cooling history of the samples in this particular temperature range. To make the axially symmetric region wide, one of the Trend: It is necessary for me to increase the diffusion time or slow down the cooling rate. According to the information in this chart, we can calculate 笪 屮 σ T f out of the ^ distribution line, which represents the "gen" (ie, flawless disease) ) To "bad" (i.e. a flawed change transition, which is a function of the cooling temperature allowed in a given temperature range in a specific temperature range, the approximate relationship between the axial symmetry region and the cooling rate: :… (R crystal-R transition) 2 = Deff * t105 (rc where the radius of the R crystal heterogeneous ingot, the R transition is at an axial position in the sample, and the line filling occurs in the main material from flawless to Radius of the axially symmetric region of the defective transition, or vice versa j Printed by the Consumers' Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs

Deff係-常數,大約爲9.3,-“],代表行間充擴散 率之平均時間與溫度,及 ^1〇5〇。(:係樣品既定軸向位置由固化冷却至大約1〇5〇乇所需 之時間。 -一 復參閲圖19,其可看出對一既定之鍵塊直徑而言,一冷 却時間可做估算以利取得一所需直徑之軸向對稱區。例I 以一大約150 mm直徑之鍵塊而言,—具有寬度大約等於錠 — 36-Deff system-constant, about 9.3,-"], represents the average time and temperature of interstitial charge diffusivity, and ^ 1050. (: The sample's predetermined axial position is cooled from solidification to about 1050 ° The time required.-Refer to Figure 19 again, it can be seen that for a given key block diameter, a cooling time can be estimated to obtain an axially symmetric region of the required diameter. Example I For a 150 mm diameter key block, — with a width approximately equal to the ingot — 36-

536562 A7 &quot;~~'~'~~ --—_ Β7 五、發明説明(34 ) &quot;~: 一' ---— 塊半徑之軸向對稱區可取得時,則在大约⑷代與⑽代 T溫度範圍間疑塊之此㈣部分應可冷却大約聰15小時 。同樣,以—大约200 mm直徑之鍵塊而言,若在此溫度範 圍間鍵塊之此特定部分可冷却大約25至35小時,則可取得 具有寬度大約等讀塊半徑之軸向對稱區.「若^線進— 步做外插,則可能需要大約65至75小時之冷却時間,以利 取得4向對稱區’而其寬度大約等於_具有大約鳩職 直徑之錠塊半徑,在此應注意的是,當錠塊直徑增加時, 由於距離增加致使行間充需擴散以到達鍵塊表面或晶格空 位之槽溝,因此需要額外之冷却時間。 請即參閱圖20、21、22、23 /可看出各不同鍵塊增加冷 却時間之效應,各圖説明一具有2〇〇mm額定直徑之錠塊一 部分,其冷却時間係自固化溫度至1〇5〇τ,依序自圖2〇增 加至圖23。 經濟部中央標準局賀工消费合作社印製 參閱圖20,其揭示距離肩部大約235至35〇 mm軸向位置範 圍内之錠塊邵分’在太約255 mm之軸向位置處,無凝結行 間充瑕疵之軸向對稱區之寬度在最大値,其大約爲錠塊半 收之4 5 %,在此位置以外,一過渡發生於自一無此瑕拖之 區域至一存在此瑕戚之區域。 請即參閱圖21 ’其揭示距離肩部大約305至460 mm抽向位 置範圍内之鍵塊邵分’在大:約360 mm之軸向位置處,無凝 結行間充瑕疵之軸向對稱區之寬度在最大値,其大約爲鍵 塊半徑之65%,在此位置以外,取疵開始生成。 請即參閱圖22,其揭示距離肩部大約140至275 mm軸向位 -37- 本^^尺度適用中國國家標準(〇奶)六4規格(210'乂29^釐) ' 一 = 536562 A7 B7 五、發明説明(35 ) 置範圍内之鍵塊邵分,在大约21 〇 mm之軸向位置處,軸向 對稱區之寬度大約等於錠塊半徑,亦即此範圍内之一小部 分錠塊係無凝結性内在點瑕疵。 请即參閱圖23 ’其揭示距離肩部大约6〇〇 mm至730 mm軸 向位置範圍内之錠塊部分,在大約64〇 mm至沉5 m^,範圍之 軸向位置上,軸向對稱區之寬度大約等於錠塊半徑。此外 ,在軸向對稱區寬度大约等於錠塊半徑之錠塊小段之長度 係大於圖22錠塊之觀察所得。 因此,當併合來看時,圖2〇 ·、21、22、23説明在無瑕疵 軸向對%區t寬度與長度上冷却至1〇5〇。。之效果,大體上 ,含有凝結行間充瑕疵之區域係晶體拉單晶率持續減低造 成初期行間充濃度之結果,其太大而無法減少該晶體部分 ,冷却時間。軸向對稱區之—較大長度意指可取得—較大 範圍之拉單晶率(即初期行間充濃度)用於此無瑕戚材料之 生長’所增加之冷却時間容許初期較高之行間充濃度,因 爲用於徑向擴散之充分時間可取得以抑制濃度在行間充瑕 疵凝:所需之臨界濃度以下。易言之,對於較長之冷却時 經濟部中央標準局負工消费合作社印製 。略低〈拉單晶率(以及較高之初期行間充濃度)仍 將導致最大之軸向對稱區6。因此,較長之冷却時間導致容 许拉早晶率變化中之增加,大致爲最大轴向對稱區直徑所 需心條件,且令製程控制上之限制變得寬鬆,結果,在鍵 塊大長度上形成一軸向對稱區之製程變得較容易。 :參:圖23,在距離晶體肩部大約665咖至73〇_上 巳U向位置上,無凝結瑕蚊_晶格空位爲主 -38- 536562 A7 B7 五、發明説明(36 ) 係存在且區域之寬度等於錠塊之半徑。 由以上資料可看出,藉由控制冷却率“亍間充 藉容許較多時間供行間充擴散至其可消失之處㈣^度^ 果,凝結行間充瑕戚之生&lt;可防止在單晶石夕塊之主要: 内發生。 ^ 刀 由上以觀,可看出本發明之多項目的皆可達成。 由於多種變化仍可在上述組合物與製程中達成且未脱離 本發明之範疇,因此前文中之所有事項應視爲用於説明, 而非具有設限之意味。 - 經濟部中夾標準局負工消費合作社印裝 本纸張尺度適用中國國家標準(CNS ) A4規格(210X 297公釐)536562 A7 &quot; ~~ '~' ~~ ---_ Β7 V. Description of the Invention (34) &quot; ~: 1 '----- When the axially symmetric area of the block radius is available, This part of the block between the T temperature ranges should be able to cool for about 15 hours. Similarly, for a key block of about 200 mm diameter, if this particular part of the key block can be cooled for about 25 to 35 hours in this temperature range, an axially symmetric region having a width of approximately equal reading block radius can be obtained. "If the line advances-extrapolation, it may take about 65 to 75 hours to cool down in order to obtain a 4-way symmetry zone 'and its width is approximately equal to the radius of the ingot with a diameter of approximately 鸠Note that when the diameter of the ingot is increased, the increase in distance causes the inter-row charge to diffuse to reach the surface of the key block or the slot of the lattice vacancy, so additional cooling time is required. Please refer to Figure 20, 21, 22, 23 / The effect of increasing the cooling time of different key blocks can be seen. Each figure illustrates a part of an ingot with a nominal diameter of 200 mm. The cooling time is from the solidification temperature to 1050τ, in order from Figure 2. Increase to Figure 23. Printed by He Gong Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs, see Figure 20, which reveals that the ingots in the axial position range of about 235 to 35 mm from the shoulder are in the axial direction of about 255 mm. At the location, there is no condensation between lines The width of the axially symmetric region is at the maximum, which is about 45% of the half of the ingot. Outside of this position, a transition occurs from an area without this defect to an area with this defect. Please That is, refer to FIG. 21 'It reveals the key block shaofen within a range of about 305 to 460 mm away from the shoulder' at a large: about 360 mm in the axial position, the width of the axially symmetric region without condensing defects between lines. At the maximum, it is about 65% of the radius of the key block. Beyond this position, defects are generated. Refer to Figure 22, which reveals the axial position about 140 to 275 mm from the shoulder. Applicable to Chinese National Standard (〇 奶) 6.4 specifications (210 '乂 29 ^ cent)' One = 536562 A7 B7 V. Description of the invention (35) The key block within the range is in the axial position of about 2 0mm Where the width of the axially symmetric region is approximately equal to the ingot radius, that is, a small part of the ingot within this range has no condensed intrinsic point defects. Please refer to FIG. 23 ', which reveals that the distance from the shoulder is about 600mm to The ingot portion within the 730 mm axial position range is from about 64 mm to 5 m ^. In the axial position, the width of the axially symmetric region is approximately equal to the ingot radius. In addition, the length of the ingot segments with the width of the axially symmetric region approximately equal to the ingot radius is greater than the observation of the ingot in Fig. 22. Therefore, when merged When viewed, Figures 20, 21, 22, and 23 illustrate the cooling of the% area t width and length in the axial direction of the defect-free area to 1050. Generally, the effect is that the area containing condensed lines is filled with defects. The result of the initial decrease of the single crystal rate caused by the continuous reduction of the crystal single crystal rate is too large to reduce the crystal portion and the cooling time. The larger length of the axially symmetric region means that a larger range of single crystal rate can be obtained (That is, the initial row filling density) The increased cooling time used for the growth of this flawless material allows higher initial row filling density because sufficient time for radial diffusion can be obtained to suppress the concentration of flaw filling between the rows: Below the required critical concentration. In other words, for longer cooling times, printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs. Slightly lower (pulling single crystal rate (and higher initial row interstitial concentration) will still result in the largest axially symmetric region 6. Therefore, a longer cooling time results in an increase in the allowable early crystallization rate change, which is roughly the core condition required for the diameter of the largest axially symmetric region, and loosens process control restrictions. As a result, the key block has a large length. The process of forming an axially symmetric region becomes easier. : Refer to: Figure 23, at about 665 to 73 ° from the shoulder of the crystal, the U-direction position of non-condensing mosquitoes _ lattice vacancies dominate -38- 536562 A7 B7 V. Description of the invention (36) The width of the area is equal to the radius of the ingot. From the above data, it can be seen that, by controlling the cooling rate, "interstitial charging allows more time for interstitial charging to diffuse to where it can disappear. ^^^ As a result, condensed interstitial charging defects can be prevented in monocrystalline stones. The main part of the evening block: it happens from the inside. ^ From the top view, it can be seen that many items of the present invention can be achieved. Because a variety of changes can still be achieved in the above composition and process without departing from the scope of the present invention, Therefore, all the matters in the foregoing should be regarded as illustrations, and not meant to have restrictions.-The printed paper size of the Zhongjia Standard Bureau of the Ministry of Economic Affairs and Consumer Cooperatives is printed in accordance with the Chinese National Standard (CNS) A4 (210X 297) Mm)

Claims (1)

六、申請專利範圍Scope of patent application 536562 第87105351號專利申請案 中文申請專利範圍_ 1 · 一種早晶矽晶圓,具有一中心軸線、一前側與一後側且皆 大致垂直於中心軸線、一環側緣、及一自晶圓中心軸線延 伸至環側緣之半徑,晶圓包含 一第一軸向對稱區,其中之晶格空位主要為内在點瑕疵 且大致上無凝結晶格空位内在點瑕疵,其中第一軸向對稱 區包含中心軸線或具有一至少15 mm之寬度。 2·如申請專利範圍第1項之晶圓,其中晶圓包含一第二軸向 對稱區’其中矽行間充原子係主要為内在點瑕疵且大致上 無凝結矽行間充内在點瑕疵。 3·如申請專利範圍第1項之晶圓,其中第一軸向對稱區之寬 度至少為半徑之15%。 — 4.如申請專利範圍第3項之晶圓,其中晶圓包含一第二軸向 對稱區’其中矽行間充原子係主要為内在點瑕疵且大致上 無凝結矽行間充内在點瑕疵。 5·如申請專利範圍第1項之晶圓,其中第一軸向對稱區之寬 度至少為半徑之25%。 6.如申請專利範圍第5項之晶圓,其中晶圓包含一第二軸向 對稱區’其中矽行間充原子係主要為内在點瑕疵且大致上 無凝結矽行間充内在點瑕疵。 7·如申請專利範圍第1項之晶圓,其中第一軸向對稱區之寬 度至少為半徑之50%。 8·如申請專利範圍第7項之晶圓,其中晶圓包含一第二軸向 對稱區’其中矽行間充原子係主要為内在點瑕疵且大致上 無凝結矽行間充内在點瑕疵。 (請先閲讀背面之注意事項再填寫本頁) 、1Τ 經濟部中央標準局員工消費合作社印製536562 Patent No. 87105351 Chinese patent application scope_ 1 · An early-crystal silicon wafer with a central axis, a front side and a back side, all of which are approximately perpendicular to the central axis, a ring side edge, and a self-wafer center The axis extends to the radius of the side edge of the ring. The wafer includes a first axially symmetric region, in which the lattice vacancies are mainly intrinsic point defects and there are substantially no condensed crystal lattice vacancies. The first axially symmetric region includes The central axis may have a width of at least 15 mm. 2. As for the wafer in the first item of the patent application scope, wherein the wafer includes a second axially symmetric region, wherein the silicon row interstitial atomic system is mainly an intrinsic point defect and is substantially free of condensed silicon row interstitial intrinsic point defect. 3. If the wafer of the first patent application scope, the width of the first axially symmetric region is at least 15% of the radius. — 4. The wafer as claimed in item 3 of the patent application scope, wherein the wafer includes a second axially symmetric region, wherein the silicon row interstitial atomic system is mainly an intrinsic point defect and is substantially free of condensed silicon row interstitial intrinsic point defect. 5. If the wafer of the first patent application scope, the width of the first axially symmetric region is at least 25% of the radius. 6. The wafer according to item 5 of the patent application scope, wherein the wafer includes a second axially symmetric region, wherein the silicon row interstitial atomic system is mainly an intrinsic point defect and is substantially free of condensed silicon row interstitial intrinsic point defect. 7. If the wafer of the first patent application scope, the width of the first axially symmetric region is at least 50% of the radius. 8. The wafer according to item 7 of the patent application scope, wherein the wafer includes a second axially symmetric region, wherein the silicon row interstitial atomic system is mainly an intrinsic point defect and is substantially free of condensed silicon row interstitial intrinsic point defect. (Please read the precautions on the back before filling out this page), 1T Printed by the Consumer Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 536562 A8 B8 C8 D8 申請專利範圍 經濟部中央標準局員工消費合作社印製 9·如申请專利範圍第1項之晶圓,其中第一軸向對稱區包含 中心軸線。 1〇·如申請專利範圍第9項之晶圓,其中晶圓包含一第二軸向 對%區,其中矽行間充原子係主要為内在點瑕疵且大致上 無凝結矽行間充内在點瑕疵。 11·如申請專利範圍第1項之晶圓 PPMA之氧含量。 12·如申請專利範圍第1項之晶圓 PPMA之氧含量。 13·如申請專利範圍第1項之晶圓 核心。 — 14·一種單晶矽塊,具有一中心軸線、一晶種錐、一端錐、及 一位於晶種錐與端錐之間之固定直徑段,固定直徑段具有 一環側緣及一自中心軸線延伸至環側緣之半徑,單晶矽塊 之特徵在於,當錠塊生長且自固化溫度冷却後,固定直徑 段即含有一第一軸向對稱區,其中之晶格空位主要為内在 點瑕疵且大致上無凝結晶格空位内在點瑕疵,其中第一轴 向對稱區包含中心軸線或具有一至少15 mm之寬度,且具 有一長度’係沿錠塊固定直徑段之至少20%長度之中心轴 線而測量。 _ 15.如申請專利範圍第14項之單晶矽塊,其中錠塊包含一第二 軸向對稱區,係與該第一軸向對稱區同中心,第二軸向對 稱區含有行間充原子做為主要之内在點瑕戚,且大致上無 凝結矽行間充内在點瑕疵。 其中晶圓具有少於大約13 其中晶圓具有少於大約11 其中晶圓具有一無氧沉澱 -2- 本紙張尺度適用中國國家標準(CNS ) A4現格(210X297公釐) (請先閱讀背面之注意事項再填寫本頁)536562 A8 B8 C8 D8 Patent application scope Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs 9 · If the patent application scope of the first wafer, the first axial symmetry area contains the central axis. 10. The wafer of item 9 in the scope of patent application, wherein the wafer includes a second axially opposite% region, in which the silicon row interstitial atomic system is mainly an intrinsic point defect and is substantially free of condensed silicon row interstitial intrinsic point defect. 11. Oxygen content of wafer PPMA such as the first item in the scope of patent application. 12. Oxygen content of wafer PPMA such as the first item in the scope of patent application. 13. The core of the wafer as in the first patent application. — 14 · A monocrystalline silicon block having a central axis, a seed cone, an end cone, and a fixed diameter section between the seed cone and the end cone, the fixed diameter section has a ring side edge and a self-center axis Extending to the radius of the ring side edge, the single crystal silicon block is characterized in that when the ingot is grown and the self-solidification temperature is cooled, the fixed diameter section contains a first axially symmetric region, wherein the lattice vacancies are mainly intrinsic point defects. And there are substantially no point defects in the crystal lattice vacancies, in which the first axially symmetric region includes the center axis or has a width of at least 15 mm and has a length of 'at least 20% of the length along the fixed diameter section of the ingot Axis. _ 15. The monocrystalline silicon block according to item 14 of the application, wherein the ingot contains a second axially symmetric region that is concentric with the first axially symmetric region, and the second axially symmetric region contains interstitial atoms. As the main intrinsic point defect, and there is almost no intrinsic point defect of the condensed silicon row. Wherein the wafer has less than about 13 where the wafer has less than about 11 where the wafer has an anaerobic precipitation-2- This paper size is applicable to Chinese National Standard (CNS) A4 (210X297 mm) (please read the back first) (Notes for filling in this page) 536562 A8 B8 C8 . ___ D8 六、申請專利範圍 16·如申請專利範圍第14項之單晶矽塊,其中軸向對稱區之長 度至少為錠塊固定直徑段長度之4〇0/〇。 17·如申請專利範圍第16項之單晶矽塊,其中錠塊包含一第二 軸向對稱區,係與該第一軸向對稱區同中心,第二軸向對 稱區含有行間充原子做為主要之内在點瑕疵,且大致上無 凝結矽行間充内在點瑕疵。 18·如申請專利範圍第16項之單晶矽塊,其中第一軸向對稱區 之寬度係至少為半徑之丨5%。 19·如申請專利範圍第16項之單晶矽塊,其中第一軸向對稱區 之寬度係至少為半徑之25%。 20·如申請專利範圍第16項之單晶矽塊,其中第一軸向對稱區 之長寬至少為錠塊固定直徑段長度之6〇0/〇。 21·—種生長一單晶矽塊之方法,其中錠塊包含一中心軸線、 一晶種錐、一端錐、及一位於晶種錐與端錐之間之固定直 徑段’固足直徑段具有一環側緣及一自中心軸線延伸至環 側緣之半徑’錠塊係自溶態碎生長且依Cz〇chrai方法而自 固化溫度冷却,該方法包含 經濟部中央標準局員工消費合作社印製 (請先閲讀背面之注意事項再填寫本頁} 在晶體固定直徑段生長期間自固化至不低於大約丨325艺 之溫度範圍上,控制一生長速度v及一平均軸向溫度梯度 G0 ’以造成一第一軸向對稱段生成,其中當錠塊自固化 溫度冷却時,晶格空位即為主要之内在點瑕疵且大致上無 凝結内在點瑕疵,其中第一軸向對稱區延伸具有一寬度至 少為15 mm或含有中心軸線。 22.如申請專利範圍第21項之方法,其中第一軸向對稱區之長 -3- 本紙張尺度適用巾關家標準(CNS ) ( 21()&gt;&lt;297公慶) &quot; 536562 申請專利範 圍 A8 B8 C8 D8 度係至少為錠塊固定直徑段長度之4〇%。 23如由、 • 請專利範圍第12項之方法,其中第一軸向對稱區之長 度係至少為錠塊固定直徑段長度之6〇0/〇。 •17申請專利範圍第11項之方法,其中第一軸向對稱區具有 寬度’係至少為錠塊固定直徑段半徑長度之60〇/〇。 25如由、 •甲請專利範圍第22項之方法,其中該第一軸向對稱區域 /、有一寬度’該寬度至少是該錠塊之固定直徑部分之半徑的 25%。 &amp; ' 26如申請專利範圍第22項之方法,其中該錠塊包括一第二轴 向對稱區域,係與該第一軸向對稱區同中心,第二轴向對 稱區含有行間充原子做為主要之内在點瑕疵,且實質上無 ’焚結矽行間充内在點瑕疵。 7’如申睛專利範圍第23項之方法,其中該第一軸向對稱區域 具有一寬度,該寬度至少是該錠塊之固定直徑部分之半_ 的 15%。 二 28·如申請專利範圍第23項之方法,其中該第一轴向對稱區域 具有一寬度,該寬度至少是該錠塊之固定直徑部分之半巧 的 25%。 &amp; 經濟部中央榡隼局員工消費合作社印製 (請先閲讀背面之注意事項再填寫本頁} 、11 29·^申凊專利範圍第23項之方法,其中該錠塊包括_第_ # 向對稱區域,係與該第一軸向-對稱區同中心,第二轴向對 稱區含有行間充原子做為主要之内在點瑕疵,且實質上無 凝結矽行間充内在點瑕疵。 3〇·如申請專利範圍第21項之方法,其中該第一軸向對稱區域 包括一中心軸。 -4- 本紙張尺度適用中國國家標準(CNS ) A4規格(210X297公釐) A8 Βδ C8 '^D8 、申請專利範圍 (請先閲讀背面之注意事項再填寫本頁) 31·&amp;申請專利範圍第21項之方法,其中該錠塊包括一第二軸 向對稱區域,係與該第一軸向對稱區同中心,第二軸向對 稱區含有行間充原子做為主要之内在點瑕疵,且實質上無 _結矽行間充内在點瑕疵。 • u申請專利範圍第21項之方法,其中控制成長速率V與瞬 間轴向溫度梯度gg,以便使比值v/Gq為臨界比值v/c^之 0·6 〜1.5 倍。 33·^申請專利範圍第21項之方法,其中控制成長速率v與瞬 間轴向溫度梯度Gq,以便使比值v/Gq為臨界比值v/G〇之 0·75 〜1.25 倍。 34·如申請專利範圍第21項之方法,其中控制成長速率V與瞬 間軸向溫度梯度Gg,以便使比值V/Gq為臨界比值V/G〇之 1〜1.1倍。 35·如申請專利範圍第21項之方法,其中該結晶具有15〇 mm之 直彼’其中該方法更包括控制該結晶之冷卻速率,以便該 結晶在至少1〇小時期間内,從固化溫度冷卻至1〇5〇〇c。 經濟部中央標準局員工消費合作社印製 36·如申請專利範圍第21項之方法,其中該結晶具有15〇 mm之 直徑’其中該方法更包括控制該結晶之冷卻速率,以便該 結晶在至少15小時期間内,從固化溫度冷卻至1〇5〇。〇。 37·如申請專利範圍第21項之方法_,其中該結晶具有200 mm之 直徑,其中該方法更包括控制該結晶之冷卻速率,以便該 結晶在至少10小時期間内從固化溫度冷卻至1〇5(rc。 38·如申請專利範圍第21項之方法,其中該結晶具有200 m m 足直徑,其中該方法更包括控制該結晶之冷卻速率,以便 __ -5- 本紙張ϋ適用中國國家標準(CNS ) A4規格(210^^^----* 536562 A8 B8 C8 D8 六、申請專利範圍 々、.、口日g在至少2G小時期間内,從固化溫度冷卻至1()5〇。〇。 39.如申請專利範圍第21项之方法,其中該結晶具有200 mm之 直拴其中忒方法更包括控制該結晶之冷卻速率,以便該 結晶在至少40小時期間内,從固化溫度冷卻至1〇5(rc。 4〇·如申請專利範圍第2丨,之方法,其中該結晶具有200 mm之 直徑,其中孩方法更包括控制該結晶之冷卻速率,以便該 結晶在至少60小時期間内從固化溫度冷卻至1〇5〇t。 (請先閲讀背面之注意事項再填寫本頁) 經濟部中央標隼局員工消費合作社印製 本紙張尺度適用中國國家標準(CNS ) A4規格(210X 297公釐)536562 A8 B8 C8. ___ D8 6. Scope of patent application 16. If the monocrystalline silicon block of item 14 of the patent application scope, the length of the axial symmetry area is at least 400 / 〇 of the length of the fixed diameter section of the ingot. 17. The monocrystalline silicon block according to item 16 of the patent application, wherein the ingot contains a second axially symmetric region that is concentric with the first axially symmetric region, and the second axially symmetric region contains interstitial atoms. It is the main intrinsic point defect, and there is almost no intrinsic point defect of the condensed silicon row. 18. The single crystal silicon block according to item 16 of the application, wherein the width of the first axially symmetric region is at least 5% of the radius. 19. The single crystal silicon block according to item 16 of the application, wherein the width of the first axially symmetric region is at least 25% of the radius. 20. The single crystal silicon block according to item 16 of the patent application scope, wherein the length and width of the first axially symmetric region is at least 60/0 of the length of the fixed diameter section of the ingot. 21 · —A method for growing a single crystal silicon block, wherein the ingot includes a central axis, a seed cone, an end cone, and a fixed diameter section between the seed cone and the end cone. A ring side edge and a radius' ingot that extends from the central axis to the ring side edge are ingots grown in a molten state and cooled at a solidification temperature according to the Czochrai method, which includes printing by the Consumer Cooperative of the Central Standards Bureau of the Ministry of Economy ( Please read the precautions on the back before filling in this page} During the growth of the fixed diameter section of the crystal, self-curing to a temperature range not lower than about 325 ° C, control a growth rate v and an average axial temperature gradient G0 'to cause A first axially symmetric segment is generated, wherein when the ingot is cooled at the solidification temperature, the lattice vacancies are the main intrinsic point defects and there are substantially no condensation intrinsic point defects, wherein the first axially symmetric region extends with a width of at least 15 mm or including the central axis. 22. The method according to item 21 of the patent application range, wherein the length of the first axial symmetry zone is -3- This paper size applies to the family standard (CNS) (21 () &gt; & lt (297 public celebrations) &quot; 536562 The scope of patent application A8 B8 C8 D8 is at least 40% of the length of the fixed diameter section of the ingot. 23 If by, • The method of item 12 of the patent scope, where the first axis is symmetrical The length of the zone is at least 60/0 of the length of the fixed diameter section of the ingot. • 17 The method of applying for item 11 of the patent scope, wherein the first axially symmetric zone has a width 'is at least the radius of the fixed diameter section of the ingot. 60 // 25. If you use the method described in item 22 of the patent, wherein the first axially symmetric region /, has a width 'the width is at least 25% of the radius of the fixed diameter portion of the ingot &Amp; '26 The method according to item 22 of the scope of patent application, wherein the ingot includes a second axially symmetric region that is concentric with the first axially symmetric region, and the second axially symmetric region contains interstitial atoms As the main intrinsic point defect, and there is substantially no 'incorporated point defect' of the incinerated silicon row. 7 'The method of item 23 in the patent scope, wherein the first axially symmetric region has a width, and the width is at least Is the solid of the ingot 15% of the half of the fixed diameter portion. 28. The method according to item 23 of the scope of patent application, wherein the first axially symmetric region has a width that is at least half the size of the fixed diameter portion of the ingot. 25%. &Amp; Printed by the Consumers' Cooperative of the Central Government Bureau of the Ministry of Economic Affairs (please read the precautions on the back before filling out this page), 11 29 · ^ The method of claim 23 in the patent scope, where the ingot includes _ The _ # direction symmetry region is concentric with the first axial-symmetric region, and the second axial symmetry region contains interstitial atoms as the main intrinsic point defect, and there are substantially no condensed silicon interstitial region interstitial defects. 30. The method of claim 21, wherein the first axially symmetric region includes a central axis. -4- This paper size applies to Chinese National Standard (CNS) A4 specification (210X297 mm) A8 Βδ C8 '^ D8, scope of patent application (please read the notes on the back before filling this page) 31 · &amp; scope of patent application The method of item 21, wherein the ingot includes a second axially symmetric region that is concentric with the first axially symmetric region, and the second axially symmetric region contains interstitial atoms as the main intrinsic point defect, and There are virtually no pitfalls in the interstitial charge. • The method of item 21 of the u patent application, in which the growth rate V and the instantaneous axial temperature gradient gg are controlled so that the ratio v / Gq is 0.6 to 1.5 times the critical ratio v / c ^. 33. The method of claim 21, wherein the growth rate v and the instantaneous axial temperature gradient Gq are controlled so that the ratio v / Gq is 0.775 to 1.25 times the critical ratio v / G0. 34. The method of claim 21, wherein the growth rate V and the instantaneous axial temperature gradient Gg are controlled so that the ratio V / Gq is 1 to 1.1 times the critical ratio V / G0. 35. The method of claim 21, wherein the crystal has a diameter of 15 mm, wherein the method further includes controlling the cooling rate of the crystal so that the crystal is cooled from the solidification temperature for at least 10 hours. To 1050c. Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 36. If the method of applying for the scope of the patent No. 21, wherein the crystal has a diameter of 150mm ', where the method further includes controlling the cooling rate of the crystal so that the crystal is at least 15 During the hour, the solidification temperature was cooled to 1050. 〇. 37. The method of claim 21, wherein the crystal has a diameter of 200 mm, and the method further includes controlling the cooling rate of the crystal so that the crystal is cooled from the solidification temperature to 1 within a period of at least 10 hours. 5 (rc. 38. The method according to item 21 of the patent application, wherein the crystal has a diameter of 200 mm, and the method further includes controlling the cooling rate of the crystal so that this paper ϋ applies Chinese national standards (CNS) A4 specifications (210 ^^^ ---- * 536562 A8 B8 C8 D8 VI. Patent application scope 々, .., g, within a period of at least 2G hours, cooled from the curing temperature to 1 () 50. 39. The method according to item 21 of the application, wherein the crystal has a diameter of 200 mm. The method further includes controlling the cooling rate of the crystal so that the crystal is cooled from the solidification temperature to at least 40 hours. 105 (rc. 40). The method according to the patent application No. 2 丨, wherein the crystal has a diameter of 200 mm, and the method further includes controlling the cooling rate of the crystal so that the crystal is in a period of at least 60 hours. Conggu The temperature is cooled down to 1050t. (Please read the precautions on the back before filling this page.) Printed on the paper by the Consumers' Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs. The paper size applies to the Chinese National Standard (CNS) A4 (210X 297 mm). )
TW87105351A 1998-04-09 1998-04-09 Low defect density, vacancy dominated silicon TW536562B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
TW87105351A TW536562B (en) 1998-04-09 1998-04-09 Low defect density, vacancy dominated silicon

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
TW87105351A TW536562B (en) 1998-04-09 1998-04-09 Low defect density, vacancy dominated silicon

Publications (1)

Publication Number Publication Date
TW536562B true TW536562B (en) 2003-06-11

Family

ID=29268235

Family Applications (1)

Application Number Title Priority Date Filing Date
TW87105351A TW536562B (en) 1998-04-09 1998-04-09 Low defect density, vacancy dominated silicon

Country Status (1)

Country Link
TW (1) TW536562B (en)

Similar Documents

Publication Publication Date Title
TW577939B (en) A process for growing a single crystal silicon ingot
TW550681B (en) Process for producing silicon on insulator structure having intrinsic gettering by ion implantation
US6743289B2 (en) Thermal annealing process for producing low defect density single crystal silicon
US20050238905A1 (en) Vacancy-dominated, defect-free silicon
TW200426255A (en) Process for preparing single crystal silicon using crucible rotation to control temperature gradient
US20010045184A1 (en) Crystal puller for growing low defect density, self-interstitial dominated silicon
TW536562B (en) Low defect density, vacancy dominated silicon

Legal Events

Date Code Title Description
GD4A Issue of patent certificate for granted invention patent
MM4A Annulment or lapse of patent due to non-payment of fees