1271450 玖、發明說明 (發明說明應敘明：發明所屬之技術領域、先前技術、内容、實施方式及圖式簡單說明） 【發明所屬之技術領域3 發明背景 本發明係關於攙雜鑽石，且更特別的是關於藉由化學 5 氣相沉積（Chemical Vapour Deposition)來製造之鑽石（以下 簡稱為CVD鑽石）。 t ]| 對一粒可被廣泛應用之鑽石而言，攙雜鑽石層體所具 有之特殊尺寸、均一攙雜濃度及與電子及/或光學有關之 1〇性質皆為有利。視其應用之内容，這種材料需要實質地排 除有害之具有電子或光學活性之陷阱（traps)或瑕疵。目前 尚無此型材料。 ，4 叩π六匁/子/义γ 15 ΙΟΟΟμη!，且側面積居於i x lmm2 — 5〇 χ 5〇mm2之範圍内之大 型表膜獨立鑽石。對於一個競爭市場上具可行性之生產而言 ’供用於這些結構中鑽石較有利是以―種大型材料長成並予 以加工成最終之成品。此外，大片可進行晶片切割加工，並 ，一步降低成品製造成本。對於諸如祕及吸收能量之量測 器之光學應用而言，大尺寸且厚之原料是—種對器材本質之 需求。因此，合成厚的層體是具有多種利益的。 爛是僅知之具有良好之相當淺接之攙雜表現特性之鑽 石攙雜物。文獻上報導其他尚在研發中之有潛力之淺接捷 雜物包含：硫⑻、卿)、氧⑼，，但其等還不是 值^賴之大量（bulk)攙雜物。有許多需要攙雜鑽石之電 20 1271450 玖、發明說明 子應用，其等通常是於—大面積上，且需要具有均一之特 性。然而，於合成中攙雜爛是該特定晶槽所具有之一種非 *鼓敏之特性。聚晶鑽石係包含—種隨機選擇之長晶槽， 然而雖然該硼之平均濃度於_個大於顆粒之尺度上是均一 5的’在同該顆粒尺寸之尺度下，該局部之硼濃度在點與點 之間是具有實質差異的。 攙雜物亦可藉由長晶之後處理來予以攙入。目前唯一 可信賴之適用於鑽石之長晶後處理是離子植入法，且其係 為一種可用以製造不具攙雜均一性之層疊鑽石結構之方法 10 。例如，可藉由使用一種適用於將硼植入一粒高品質之天 然Ila型鑽石之劑量及能量來製造一種VI，化型本質）結構 。不幸地，殘量損害（空隙及裂縫）常於離子植入法之狀態 下發生。雖然黏合（annealing)處理能夠減少此種損害，但 此種損害無法完全被移除。該損害會由於硼受體之散射及 15 補償不良而導致帶電載體性質衰降。 藉由化學氣相沉積（CVD)於一基材上沉積或生長例如 鑽石之方法，疋目刖已建立且被廣泛描述於專利案及其他 文獻中。由於鑽石是藉由CVD來沉積於一種基材上，因此 该等方法通常涉及提供一種氣體混合物，其於解離時能夠 20提供呈原子形式之氫或一種鹵素（例如：氟、氣）、及碳或 含碳自由基及其他具反應性物質（例如·· CHX、CFX，其中X 可以是1至4)。此外，含氧來源物可以存在，以及提供氮及 提供侧之來源物。於多種方法中惰性氣體諸如氦、氖或氬 亦可存在。因此，一種典型的來源氣體（s〇urce gas)混合物 1271450 玖、發明說明 會包含：碳氫化合物（CxHy，其中X及y可各自是1至1〇)或鹵 化石反（CxHyHalz ’其中X及z可各自是1至1〇，且y可以是〇至 10)以及任擇之下列一者或數者：C〇x(其中χ可以是〇至2)、 〇2、h2、N2、NH3、B2H6及一種惰性氣體。每一種氣體可 5以呈其天然同位素比例存在，或者可以呈人工控制相對之 同位素比例；例如：氫可以氘或氚存在，且碳可以或1271450 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明 发明It is about diamonds manufactured by Chemical Vapour Deposition (hereinafter referred to as CVD diamonds). t ]| For a diamond that can be widely used, it is advantageous for the noisy diamond layer to have a special size, a uniform doping concentration, and an electron and/or optical related property. Depending on the application, this material needs to substantially eliminate harmful electronic or optically active traps or flaws. There is currently no such material. , 4 叩π 六匁/子/义 γ 15 ΙΟΟΟμη!, and the side area is in the range of i x lmm2 — 5〇 χ 5〇mm2 of large-size independent diamond. For a viable production in a competitive market, it is advantageous for diamonds used in these structures to be grown from a large material and processed into the final product. In addition, large pieces can be processed by wafer cutting, and the manufacturing cost of finished products can be reduced in one step. For optical applications such as thermal and energy absorbing gauges, large and thick materials are essential to the nature of the equipment. Therefore, the synthesis of thick layers is of various benefits. Rotten is only known as a drill stone with a good and relatively shallow performance. The literature reports that other shallow joints that are still in development have the potential to include: sulfur (8), Qing), oxygen (9), but they are not the bulk of the bulk. There are many electricity that require noisy diamonds. 20 1271450 玖, invention description Sub-applications, etc. are usually on a large area and need to have uniform characteristics. However, the smashing in the synthesis is a non-drum-sensitive property of the particular crystal cell. The polycrystalline diamond contains a randomly selected crystal growth tank, although the average concentration of boron is uniformly 5 on the scale larger than the particle size, and the local boron concentration is at the same level as the particle size. There is a substantial difference between the points and the points. The impurities can also be intruded by treatment after the crystal growth. Currently the only reliable long-crystal post-treatment for diamonds is ion implantation, and it is a method that can be used to make laminated diamond structures without noisy uniformity 10 . For example, a VI, chemically active structure can be fabricated by using a dose and energy suitable for implanting boron into a high quality natural Ila diamond. Unfortunately, residual damage (voids and cracks) often occurs in the state of ion implantation. Although the annealing treatment can reduce such damage, such damage cannot be completely removed. This damage can lead to a decline in the nature of the charged carrier due to scattering of the boron acceptor and poor compensation of 15 . Methods for depositing or growing, for example, diamonds on a substrate by chemical vapor deposition (CVD) have been established and are widely described in patents and other literature. Since diamonds are deposited on a substrate by CVD, such methods generally involve providing a gas mixture capable of providing hydrogen or a halogen (eg, fluorine, gas), and carbon in atomic form upon dissociation. Or carbon-containing free radicals and other reactive substances (such as CHX, CFX, where X can be 1 to 4). In addition, an oxygen-containing source may be present, as well as providing nitrogen and providing a source of the side. Inert gases such as helium, neon or argon may also be present in a variety of ways. Therefore, a typical source gas (s〇urce gas) mixture 1271450 玖, the description of the invention will include: hydrocarbons (CxHy, where X and y can each be 1 to 1 〇) or halogen fossils (CxHyHalz 'where X and z may each be 1 to 1 〇, and y may be 〇 to 10) and optionally one or more of the following: C〇x (where χ may be 〇 to 2), 〇2, h2, N2, NH3, B2H6 and an inert gas. Each gas may exist in a ratio of its natural isotopes, or may be artificially controlled to a relative isotope ratio; for example, hydrogen may be present in helium or neon, and carbon may or
13 I c存在。來源氣體（source gas)混合物可藉由一種能量來源 ，例如：微波、RF(高週波）能量、一種火燄、一種熱鎢絲 或以喷射為主之技術來產生解離，且可容許該等藉此產生 10 之具有反應性氣體被沉積於一個基材上來形成鑽石。 CVD鑽石可以被產生在各種不同的基材上。視該基材 之本質及製法之化學性質可製成多晶或單晶CVD鑽石。 對3午多其他的攙雜物而言，於沉積之過程中使硼攙雜 入固體中是較不困難的。該於固體中硼之攙雜比例是攙雜 15之硼（B)對比碳（C)之濃度（[B]/[C]:固體）較諸於沉積氣體 之[B]/[C] ·氣體而吕’通常（於長晶槽内）是大約1， 然而其可因多種因素而異。有多種可以於製程中以硼攙雜 CVD鑽石之方法。以微波電漿、熱鎢絲及電弧噴射技術， 可添加乙棚烷（Β#6)或某些其他適合之氣體至氣體流内、 20可將該導入氣體打氣通入包含氧化硼（β203)之甲醇或丙酮 内可置放侧酸粉末於晶槽内、或***硼電極至電聚中。 對藉由燃燒火談方法來長晶而言，一種包含蝴酸之曱醇煙 硌可以一種噴霧器來予以注射入氣流内。鑽石薄膜亦可以 在非刻意之情形下被攙雜，例如：電漿已分解一種以硝酸 1271450 玖、發明說明 硼製成之基材。 氮氣亦可呈多種形式來予以導入合成之電漿内，其等 之典型是氮氣（NO、氨（NH3)、空氣及乙硼烯（b2H4)。 雖然高品質之單晶（SCV)CVD鑽石於高能電子上扮演 5 一個重要的角色，但有潛力之應用數量必須在一粒具有均 一且優良電子性能之CVD攙雜鑽石是可得之情況下才會實 質地增加。此外，硼攙雜鑽石還能夠被應用在其他以硼 (B)攙雜可有利地產生均一之顏色、亮度或其他相關特性 之應用上。 10 【明内溶1】 發明概要 依據本發明之第一個方面，本發明係提供一種以氣相 沉積（CVD)製造之㈣雜錯石，其中蝴攙雜之整體濃度是 均一的，於大部分體積内之差異小於5〇%，且較佳是小於 15 2〇°/❾，於每一個量測點以一種側向解析來予以量測是小於 50μηι，較佳是於每一個量測點以一種側向解析來予以量 測是小於30μιη，且具有特性（iHiii)之至少一種： ⑴層體係形成自一個較佳是一個由{1〇〇}、{ii3卜 {111}及{110}所構成之單一養晶槽，以及更佳是該等 2〇 {1〇〇}晶槽。 (11)層體厚度超過100^m，且更佳是超過500μιη，以及 (in)層體之體積超過lmm3，更佳是超過1〇mm3，甚至 更佳是超過30mm3。 該術語“大部分層體，，被使用於本說明書及申請專利範 1271450 玖、發明說明 圍中係意指鑽石層體總體積之至少70%，較佳是大於8〇% ’且更佳是大於95%。 本發明之CVD單晶硼攙雜鑽石層體亦可包含以氮作為 種攙雜物。該鑽石層體一般會包含一種不超過1/5硼濃 5度之氮濃度，且較佳是低於硼濃度之1/50。 該鑽石層體較佳是具有“優良之結晶性質，，。此“優良 之結晶性質，，容許硼原子及氮原子以攙雜物以及與點瑕疵 有關之例如諸等包含：空隙、氫及類似物之存在。 單晶硼攙雜鑽石層體亦可於該鑽石之大部分體積内具 10有一種或多種下列特性，該大部分體積係如上文所定義： 0)該層體包含一種是高於i χ 1〇14原子/cm3且低於 1 X l〇2G原子/cm3之非補償性硼濃度，較佳是一種是高於i X 1〇15原子/cm3且低於2 x 10i9原子/cm3之未補償硼濃度， 且更佳是一種是高於5 X 1〇15原子/cm3且低於2 χ 1〇i8原子 15 /cm3之未補償硼濃度， (b) —種於300K下量測之電洞移動率（Mh)係超過 Ph=G χ 2.1 χ 1〇10 /(Nh0·52) 當Nh小於或等於8 χ 1〇15原子/cm3時 (函數1) gh=G χ 1 χ 1〇18 /Nh 當Nh大於8 χ 1015原子/cm3時 (函數2) 其中Nh是電洞濃度（或當量，離子化蝴接受子之濃度） ’ Ph與Nh之函數關係是依據現行模式且該〇之數值代表最 10 1271450 玖、發明說明 佳之現有讀值。G是一個大於1.1之數值，且較佳是一個大 於1.4之數值，且更佳是一個大於1.7之數值，甚至更佳是 大於2.0。 (c)具有在與氮-空隙（N-V)色蕊相關之575nm及637nm 5 下為微弱或無冷光特性。特別地，當於77K下以514nm氬 離子雷射激發來量測時，該氮空隙於575nm對比637nm峰 值之零聲子譜線之積分強度是小於該落在1332(:1^1之鑽石 拉曼譜線（Raman line)之積分強度的1/50，較佳是小於 1/100，且更佳是小於1/300。 10 (d) —條於300K下以514nm氬離子雷射激發下量測之 拉曼譜線的寬度半高寬值（FWHM)是小於4CHT1，較佳是小 於3cm_1，且更佳是小於2.5cm-1。 (e) 當以傅立葉轉換紅外線光譜儀（FTIR)使用下述方 法來量測時，其未補償硼濃度具有高度之非均一性。特別 15 是，該未補償硼濃度之分佈頻率是於一取樣自該層體之代 表性取樣以傅立葉轉換紅外線光譜儀（FTIR)來予以量測， 於90%之量測中的差異，以平均值之百分比表示，是小於 50%，更佳是小於30%。 (f) 一種於紫外線激發之下，使用下述之方法量測 20 77K下238nm之結合激子散射（BE)，該結合激子散射（BE) 與未補償性取代硼原子之固體濃度一致。特別地，該藉由 此種方法量測之B E頻率分佈，於任何一個該層體之代表性 表面或該層體之取樣下，其90%之量測所具有之差異，以 平均值之百分比表示，是小於50%，更佳是小於30%。 1271450 玖、發明說明13 I c exists. The source gas mixture can be dissociated by an energy source such as microwave, RF (high frequency) energy, a flame, a hot tungsten wire or a spray-based technique, and can be tolerated A reactive gas that produces 10 is deposited on a substrate to form a diamond. CVD diamonds can be produced on a variety of different substrates. The polycrystalline or single crystal CVD diamond can be made depending on the nature of the substrate and the chemical nature of the process. For other noodles in the 3rd afternoon, it is less difficult to mix boron borax into the solid during the deposition process. The doping ratio of boron in the solid is boron (B) versus carbon (C) concentration ([B] / [C]: solid) compared to the [B] / [C] gas of the deposition gas. Lu's (usually in the long crystal bath) is about 1, however it can vary depending on a variety of factors. There are a variety of methods that can be used to make boron-doped CVD diamonds in the process. With microwave plasma, hot tungsten wire and arc spray technology, ethyl sane (Β#6) or some other suitable gas can be added to the gas stream, and 20 can be pumped into the containing boron oxide (β203). In the methanol or acetone, the side acid powder may be placed in the crystal bath or the boron electrode may be inserted into the electropolymer. For the growth of the fire by the combustion method, a decyl alcohol containing cinnamic acid can be injected into the gas stream by a spray. Diamond films can also be noisy in unintentional situations. For example, the plasma has been decomposed into a substrate made of boron with 1271450 硝酸, the invention. Nitrogen can also be introduced into the synthesized plasma in a variety of forms, such as nitrogen (NO, ammonia (NH3), air, and diborane (b2H4). Although high quality single crystal (SCV) CVD diamonds are used in High-energy electrons play an important role in 5, but the number of potential applications must be substantially increased in the case where a CVD-doped diamond with uniform and excellent electronic properties is available. In addition, boron-doped diamonds can be It is used in other applications where boron (B) doping can advantageously produce uniform color, brightness or other related properties. 10 [Inventive Dissolution 1] Summary of the Invention According to a first aspect of the present invention, the present invention provides (4) Miscellaneous stone produced by vapor deposition (CVD), wherein the overall concentration of the butterfly doping is uniform, the difference in most of the volume is less than 5%, and preferably less than 15 2 〇 ° / ❾, in each The measurement points are measured by a lateral analysis to be less than 50 μm, preferably measured at each measurement point by a lateral resolution of less than 30 μm, and having at least one of the characteristics (iHiii): (1) body It is formed from a single crystal growth tank composed of {1〇〇}, {ii3b{111} and {110}, and more preferably these 2〇{1〇〇} crystal grooves. The thickness of the layer body exceeds 100 μm, and more preferably exceeds 500 μm, and the volume of the (in) layer exceeds 1 mm 3 , more preferably exceeds 1 mm 3 , and even more preferably exceeds 30 mm 3 . The term “most layers, It is used in the present specification and the patent application No. 1271450. The invention description means at least 70%, preferably more than 8%, and more preferably more than 95% of the total volume of the diamond layer. The single crystal boron doped diamond layer may also comprise nitrogen as a dopant. The diamond layer generally comprises a nitrogen concentration of no more than 1/5 boron concentration of 5 degrees, and preferably less than 1/50 of the boron concentration. Preferably, the diamond layer has "excellent crystalline properties," which is an excellent crystalline property, and allows boron atoms and nitrogen atoms to be contained in dopants and related to dots, such as voids, hydrogen, and the like. The presence of matter. The single crystal boron doped diamond layer can also have 10 in most of the volume of the diamond. A plurality of the following characteristics, the majority of which are as defined above: 0) the layer comprises a non-compensating boron concentration which is higher than i χ 1 〇 14 atoms/cm 3 and lower than 1 X l 〇 2 G atoms/cm 3 , Preferably, one is an uncompensated boron concentration higher than i X 1 〇 15 atoms/cm 3 and lower than 2 x 10 i 9 atoms/cm 3 , and more preferably one is higher than 5 X 1 〇 15 atoms/cm 3 and lower than 2 χ 1〇i8 atom 15 /cm3 uncompensated boron concentration, (b) - The hole mobility (Mh) measured at 300K exceeds Ph=G χ 2.1 χ 1〇10 /(Nh0·52) When Nh is less than or equal to 8 χ 1〇15 atoms/cm3 (function 1) gh=G χ 1 χ 1〇18 /Nh When Nh is greater than 8 χ 1015 atoms/cm3 (function 2) where Nh is the hole concentration (or Equivalent, ionized butterfly acceptor concentration) 'The relationship between Ph and Nh is based on the current model and the value of the 代表 represents the most existing reading of 10 1271450 发明, invention description. G is a value greater than 1.1, and is preferably a value greater than 1.4, and more preferably a value greater than 1.7, and even more preferably greater than 2.0. (c) has weak or no luminescent properties at 575 nm and 637 nm 5 associated with nitrogen-void (N-V) color. In particular, when measured by 514 nm argon ion laser excitation at 77 K, the integrated intensity of the nitrogen void at 575 nm versus the 637 nm peak zero phonon line is less than the diamond falling at 1332 (: 1^1) The integral intensity of the Raman line is preferably 1/50, preferably less than 1/100, and more preferably less than 1/300. 10 (d) - The strip is excited by a 514 nm argon ion laser at 300K. The width half width (FWHM) of the measured Raman line is less than 4 CHT1, preferably less than 3 cm_1, and more preferably less than 2.5 cm-1. (e) When using Fourier Transform Infrared Spectrometer (FTIR), the following uses In the method of measurement, the uncompensated boron concentration has a high degree of heterogeneity. In particular, the distribution frequency of the uncompensated boron concentration is a representative sample taken from the layer by a Fourier transform infrared spectrometer (FTIR). To measure, the difference in the 90% measurement is expressed as a percentage of the average value, which is less than 50%, more preferably less than 30%. (f) Under ultraviolet excitation, the following method is used. 238 nm combined exciton scattering (BE) at 20 K, the combined exciton scattering (BE) and unfilled The concentration of the solid substituted boron atom is uniform. In particular, the BE frequency distribution measured by this method is 90% of the measured surface of any one of the layers or the sampling of the layer. The difference, expressed as a percentage of the average, is less than 50%, more preferably less than 30%. 1271450 玖, invention description
(g)於紫外線激發之下，使用下述方法量測其於77K 之自由激子散射（FE)強度是具有高度均一性。特別地，該 藉由此種方法量測之FE的頻率分佈，於任何一個該層體之 代表性表面或該層體之取樣下，其9〇%之量測所具有之差 5異’以平均值之百分比表示，是小於50%，更佳是小於 30%。 發現本發明CVD鑽石中具有高移動率是令人驚訝的。 現行之供用於該等載體濃度大於8 χ 1〇!5原子/cm3之載體（ 或離子化接受子）濃度所產生之移動率差異模式是依據接 10受子硼原子是主要的散射機制，且其分佈是之本質是依據 其存在。因此，這個模式暗示不可能達到高於這個數值。 因此相反地，於本說明書中所述之本發明結果顯示該模式 是錯誤的，此係由於先前文獻中所報導之攙雜鑽石中所具 有之移動率是受限於其他可被移除之因子。 15 本發明之CVD單晶硼攙雜鑽石層體可以是表膜獨立或 構成一粒較大之鑽石個體或層體之一個層體或區域。該較 大之鑽石層體或個體可以是以CVD或其他合成方法製造之 早晶或聚晶鑽石。該較大之鑽石層體或個體可被模雜以删 、氮或或其他元素。 ：〇 本發明之鑽石層體或個體可以是一種寶石之形式。 依據本發明之另一個方面，本發明係提供一種用以製 仏一種硼攙雜單晶CVD鑽石層體。此方法包含之步驟是： 提供-種具有_個實f不具有結晶㈣表面之鑽石基材， 提供一種來源氣體（source gas)(此種來源氣體包含一種硼 12 1271450 玖、發明說明 之來源），分解該來源氣體，然後容許均相磊生鑽石於該 實質不具有結晶瑕疵之表面上生長，藉此來產生一種單晶 硼攙雜鑽石層體，較佳是具有上述型式者。該方法之基本 要件是該鑽石必須是於一個實質不具有結晶瑕疵之表面上 5 生長。 於本發明之方法中可額外地包含使用控制性添加氮至 來源氣體中。於來源氣體中之氮可提供對生長之單晶所產 生之形體具有額外的控制，且氮的攙雜比例會實質地低於 蝴。因此氮氣之添加，以氮氣分子來計算，是居於該高於 10 〇·5且低於lOOOOppm之範圍内，且較佳的是居於該高於丄且 低於lOOOppm之範圍内，更佳的是居於該高於3且低於 200Ppm之範圍内，其等由於該攙雜材料之硼是以一種散射 色蕊存在，因此對硼攙雜層體之電子特性不具有顯著負面 影響，且會顯著地增加{1〇〇}晶體生長槽之尺寸，並縮小 15競爭性晶體生長槽（例如：{111})之尺寸。此意指，對生長 於一個{100}板塊上而言，該添加之氮氣能使其生長而維 持實質之{ 100}晶體生長槽。那些熟習此項技藝人士可瞭 解的是使用氮氣來改變形體之階段可與之分離或接續該均 一删攙雜層體長晶之階段。 2〇 因此本發明之均一硼攙雜鑽石可被應用於一寬廣範圍 之領域中，其例如：電子、價測儀、高能電子，等等。此 外，硼攙雜鑽石還能夠被應用在其他以硼（B)攙雜可有利 地產生均-之顏色、亮度或其他相關特性之應用上。例如 ，於某些諸如一種切割刀之應用中，硼可被使用來使該鑽 13 1271450 玖、發明說明 石具有顏色，藉此改善視覺控制，且其顏色之均一度可當 做種心不品質之因子。可選擇性該鑽石可被使用於諸如 抛光寶石之㈣性應用上，其一般係相同地可以顏色之均 度μ做一種指示品質之因子。 對於上述各種不同的應用而言，該鑽石這體或層體可 藉由例如·切割來產生兩個或數個以及通常是大量之小顆 粒或單70，俾以供用於上述之一種或多種應用中。該顆粒 或單元之尺寸可視其應用而定。 【實摊^方式】 10較佳實施例之詳細說明 除了上述特徵之外，本發明單晶硼攙雜CVD鑽石層體 可於其大部分體積中具有一種或數種下列之特徵，該大部 分特徵係如上文所定義： 1·由任何一種單一雜質：矽（si)、磷（P)、硫（s)、鎳 15 (Nl)、姑（Co)、鋁（A1)、錳（Μη)、鐵自由激子散射（FE)所構 成之個位准是居於不超過1 ppm，以及一個由這些雜質所 構成之總量是不超過5ppm。上述之“雜質，，不包含氫及其同 位素。 2· —個落在575nm寬帶之陰極冷光（CL)發散訊號很低 20 或不存在，且一個於77K下以514nm氬離子雷射激發（稱為 300mW入射光束）來量測之光學冷光（PL)相關譜線係具有 一個較該落在1332〇11_1之拉曼譜線（Raman line)波峰積分面 積小1/100之波峰積分面積，較佳是小於丨/50，更佳是小於 1/300 〇 14 1271450 玖、發明說明 3. 於電子磁共振光譜（EPR)中，一個帶電中性之單一氮 色蕊[N-C]Q之濃度是低於40ppb，更典型地是低於lOppb。 4. 於電子磁共振譜（EPR)中，一個於g=2.0028之轉動 密度是小於1 X l〇17cm·3，更典型是小於5 X 1016cnT3。於單 5 晶鑽石中，此落在g=2.0028之譜線係與晶格瑕疵濃度有關 ，其較大值典型地存在於：天然之Ila型鑽石中、於CVD鑽 石中（經由凹陷來予以可塑性地破壞）、及於品質不良之均 相蠢生（homoepitaxial)鑽石中。 5. 具有優良之光學性質，其紫外線（UV)/可見光及紅 10 外線（IR)穿透性近似lib型鑽石之理論極大值，且更特別地 ，單一取代氮於紫外線（UV)之270nm的吸收值很低或無， 於紅外線光譜之2500至3100CHT1之範圍内的C-H吸收寬帶 很小或無。半導體硼攙雜鑽石之吸收光譜特徵是一個起始 自近紅外線光譜區域之落在大約370meV向大約2.2eV延伸 15 之連續吸收峰。此吸收峰係負責產生藍色（對濃度〜5 X 1015cm_3而言是淡藍色，對濃度〜5 X 1019cm_3而言是深藍色 至黑色）。可於能量之連續寬帶閥值以下觀測到三個居於 304、348及363meV之顯著寬帶，其等若以高解析度之低 溫觀測時，會展現一個呈相當數量之結構。 20 6· X-射線形貌學（X-ray topography)顯示與長晶有關之 特性中，原始基材之<100>稜線向外生長形成<110>稜線。 由於具有潛力之補償氮氣實質低於硼，因此於未補償 硼之分佈均一性通常可代表所有的硼濃度之均一性。此外 ，電子特性主要是視未補償之硼濃度而非所有的硼濃度而 15 1271450 玖、發明說明 定。因此未補償硼濃度之均一性是一項重要的參數。 含有未補償删之鑽石顯示一種具有一極大值落在 1282cm·1 (159meV)之單一聲子吸收特徵。據發現該未補償 硼濃度與此寬帶對1282cm·1之吸光係數效應間是呈一種線 5性關係。當於室溫下進行量測時，該以ppm計之硼濃度是 1·2 X (1282cm_1之吸光係數）。 包含未補償硼之鑽石亦顯示一種可藉由扣除本質性兩 聲子吸收光譜所顯現之一個落在2457cm-i之吸收特徵。當 其1282cm 1之特徵太弱而無法使用時，其未補償硼濃度可 10 使用關係式： 未補償蝴（B)濃度（ppm)=0 00142 X 之積分吸光係數來推算其落在2457cnfi之積分吸光係數。 於一粒兩側邊平行之鑽石樣品中，一巨觀量測該未補 償硼濃度之均一度可以下列方式來使用傅立葉轉換紅外線 光瑨儀（FTIR)吸光儀。一個整體樣品之代表性紅外光圖譜 特徵了藉由收集该等於室溫下以一個〇 · 5 1解析度及一個 〇.5mm之視角尺寸量測之傅立葉轉換紅外線光譜儀”奵幻 光瑨來予以製成，該圖譜包含一個數據點最小值為20點。 其後以其平均量測為依據，挑選上述關係之一，並使用其 2〇來推算出每一個位置之未補償硼濃度。其後由其濃度量測 之頻率圖來判斷其均一度，以其遠離群組標準差極限值之 平均值來評估量測之百分比。 阿品質蝴攙雜鑽石之紫外線冷光光譜（於77K下記錄） 顯不一個落在5.22eV(237.5nm)有很強的結合激發散射，且 16 1271450 玖、發明說明 於5.27eV(235.2nm)有自由激發散射。對具有硼濃度超過大 約lppm之高品質鑽石而言，這兩個於77K下散射之積分強 度與未補償硼之濃度呈近似正比例。其關係式為： [以ppm計之未補償硼濃度]= 1.86 X 1(結合激發強度）/ 5 1(自由激發強度）。 於整個樣品内不同位置之此比例於整個量測之硼濃度 所涵概範圍内可以被用來判斷該鑽石於接近表面區域之鑽 石均一性。該樣品被塗覆以一薄層（5nm)、均一層金，俾 以避免電荷效應，配置於溫度77K下一部掃瞄電子顯微鏡 10 ，並使用一部MonoCL系統以一 15kV之加速電壓、一0.2毫 安培之電流以及一尺寸小於10 μπι X 10 μηι之點來收集。 該樣品之UV CL特性可藉由收集該等落在以兩組具有 間隔500μπι或1mm(視其涵概之面積）之垂直線所交錯而成 之格線内的交叉點來予以定義之譜線作圖，收取最少30點 15 之數據。其後由量測之濃度所製成之頻率圖來判斷，評估 90%量測之全部寬度分佈，並以其平均值之百分比來表示 。這個方法被施用來量測結合激發及自由激發散射之強度 ，並計算兩種強度之比例。 由於在抑制結合激發散射之捕捉瑕蔽上有顯著差異， 20 而使此等於其後增加該等被觀測到之結合激發散射，除非 每一處之結合激發散射都被完全抑制。 一個強烈的自由激發之存在係意指其實質沒有諸如差 排（dislocation)及雜質之瑕疵。對於聚晶CVD鑽石合成中 之個別結晶而言，低瑕疵及雜質密度與高自由激發自由激 17 1271450 玖、發明說明 子散射（FE)散射之間的關連性已於先前被報導。於較高之 棚位准（典型是於固體中高於2〇-25ppm)時，其自由激發散 射最終會被高的硼點缺陷密度抑制，而不是因為諸如差排 之結晶缺陷。自由激發散射之均一度是一種優良之不存有 5 局部南缺陷密度量測。 典型之一次離子質禮儀（SIMS)分析法係以一 i〇kv之主 電壓、一典型為1μΑ之電子束電流以及一小於5〇^m之空間 解析度來施加於主要之02+電子束。其典型地是循該等於 個於層體表面上標定〇.5mm或lmm之分析點，典型是由 10每一面得到最少20點，更佳是最少40點。藉由比較植入之 標準點來做校正。得自二次離子質譜儀（SIMS)之數據是藉 由求取群組平均值來予以分析，其後再以求取數據所涵概 之範圍，來表示數據群組中不同％之部分，對於一層體之 兩個相反的主要表面設定大約是相等之加重計分，俾以界 15疋個體積。於一偵測極限為大約2-5 X 1〇14原子/cm3之下 ，典型之二次離子質譜儀（SIMS)再現性可視其情形落在夂 5% 〇 為了界疋一材料之體積，兩個相反的表面是以SIMS 及be/fe圖4來予以界定，並藉由紅外線（ir)吸光來界定 2〇穿透樣品之厚度。量測技術（供用於BE、FE及未補償硼濃 度之SEM分析法，以及供用於所有硼濃度之SIMS分析法） 之解析度與諸等能夠於鑽石中觀測到之不同型態之硼濃度 有關例如·於具有一 ΙΟΟμπχ之典型顆粒尺寸之聚晶鑽石 中，以一個lmm分析點掃瞄樣品可能會平均掉而於個別之 18 1271450 玖、發明說明 晶粒或長晶槽之間觀察不到實質之硼（B)濃度差異。以一 5〇μπι或更低之解析度來取樣2〇個或更多之數據點時，可 能會展現此種微小等級差異是不存在的。 對製造本發明之均一硼攙雜CVD鑽石單晶層體而言， 5於一粒實質不具有結晶瑕疵之鑽石表面上長晶是很重要的 。於此考量下，瑕疵主要是意指差排（disl〇cati〇n)及微細 裂縫’其亦包含：雙晶分界（twin boundaries)、本質與攙 雜氮然關之點瑕疵、低頻散射分界及任何一種其他之晶格 延伸性損壞。較佳地，該基材是一種低雙折射1&型電中性 10 、Ib或na高壓/高溫合成鑽石或一種CVD合成單晶鑽石。 瑕疵會呈兩種方式來損害材料：負面影響電子性質（例如 :電洞之移動性）及影響局部的硼導入。由於差排長晶 (dislocation multiplication)是在厚層體長晶期間發生，因 此於基材及長晶早期控制差排是特別重要的。 15 瑕疵密度可以最容易地使用一種最適於顯現該瑕疵之 電漿或化學蝕刻（被稱為一種顯示電漿蝕刻）來做光學檢視 ，其可使用例如：一種如下述型式之簡易電漿蝕刻。可顯 現兩種型式之瑕疵： 1) 諸等基材材料品質之本質。於挑選之天然鑽石中 20 ’這些瑕疯密度可低至50/mm2，較典型之數值是1〇/mm2 ，或時而高至106/mm2或更高。 2) 諸等由拋光所產生者，包含沿拋光線之差排結構 及被細裂縫所形成之細碎裂痕。此等之密度可於一樣品中 具有差異，於不良抛光區域或樣品中，其典型之數值是落 19 1271450 玖、發明說明 在大約l〇2/mm2至超過l〇4/mm2之範圍内。 以如上述之與瑕疵有關之表面蝕刻特徵密度而言，瑕 庇密度較佳是少於5 X l〇3/mm2，且更佳是少於1〇2/mm2。 因此，於CVD長晶之基材表面或該表面之下的瑕疵位 5准可藉由小心製備基材來使具有及低於瑕疵位准減至最低 。由於當完成一個基材時，每一個階段皆可能影響落在最 終會形成基材表面之材料内部之瑕疵密度，因此於本案之 製造中包含任何一種施用於該等得自礦場採石（於天然鑽 石之情形）或合成（於合成材料之情形）材料之方法。特別的 H)加工步驟可包含習用之鑽石製法，例如：機械磨削、磨光 (lapping)及拋光（於此申請案中特別適用於低瑕疵位准）， 以及較少使用之技術，例如：雷射製造或離子植入及剝離 技術、化學/機械拋光、以及液體及電聚兩種化合 物製造技術。此外，該表面Rq(以點針剖面儀（styius 15 Profilometer)，較佳是以大於〇 〇〇8mm之長度來量測之平 坦表面剖面）應降至最低，於電漿蝕刻前之典型數值是不 超過數奈米（即低於10奈米）。 一種特別之減少基材表面損壞之方法係包含一種於均 質蟲生鑽石長晶之表面原位（in situ)電裝姓刻。理論上這 20種钱刻並不需要原位，亦不需要於其後立刻進行長晶，但 原位可達到最高效益，因為其可避免任何一種進一步物理 性損壞或化學性汙染之風險。當長晶亦是以電紧為主之方 法時’-種原位蚀刻也通常是最便利的 類似沉積或鑽石長晶之方法，但不使用任何一種含碳來源 20 1271450 坎、發明說明 氣體且通常是於一略低之溫度下，俾以產生較佳之蝕刻速 率控制。例如，其可由下列之一或數者所構成： (i) 一種氧氣蝕刻法，其主要是使用氫氣，可任擇地 具有一少量之氬氣，並需要-少量之〇2。典型之氧氣㈣ 條件是50-450 X 102 Pa，一種包含氧氣含量K4%之姓刻氣 體，-種0-30%之氬氣含量，以及平衡之氣氣，所有的百 分比是以體積計，具有一個600_110(rc之基材溫度（更典型 疋800。〇以及一個典型為歷時3-6〇分鐘之期間。 (ii) 一種類似於⑴之氫氣蝕刻法，但其中不含氧氣。 ⑽選擇性之供用於㈣之方法，但不單獨以氬氣為 主’氫氣及氧氣亦可被使帛，例如：諸等使用鹵素、其他 惰性氣體或氮氣。 ^ 15 蝕刻法係典型地由一種氧氣蝕刻，繼之以一種氫氣蝕 刻，然後直接移入該藉由導入碳來源氣體之合成中。可選 擇蚀刻時間/溫度，俾使由製造留存在表面之損壞能夠被 移除仁不致形成一種兩度粗糖之表面且沒有廣泛沿延 伸性瑕蚊钮刻（例如：貫穿表面之差排（diSl〇cation))，及 藉此產生深凹洞。由於蚀刻是侵錄，因此這些階段特別 重要的是其組份之合成槽設計及材料選擇，俾使沒有材料 被電衆轉移入氣相或移入基材表面。該氯氣蚀刻繼之以氧 就蚀刻之方法對該等由氧氣蚀刻所造成之角度磨損之結晶 瑕純不具專一性，其等會侵害性攻擊此種瑕疲並提供一曰 種較平緩、較適合接續長晶之平面。 CVD鑽石於其上長晶之鑽石基材面或表面較佳是 21 1271450 玖、發明說明 {100}、{110}、{113}或{111}表面。受限於製法，實際樣 品表面方位可與這些理論之方位相差高達5度，某些個例 可高達10度，雖然這會不利地影響再現性而較不為所欲。 於本發明方法中亦很重要的是適當控制Cvd長晶之環 5境中的雜質含量。更特別地，該鑽石長晶必須發生在一實 質不含污染物之空氣中，且該刻意添加之硼（以及設若使 用氮）被適當地控制。所需控制之硼及氮攙雜濃度係視實 施例而定，但典型是需要安定地大於2〇% ,更典型是需要 安定地大於10%，甚至更典型是需要安定地大於3%。此種 10控制必須小心控制來源氣體之氮雜質，因為氮是一種常見 之污染物。為了達到這種控制程度，於小心添加氮之前， 該來源氣體中之氮的位准通常被維持於氣相中低於500 ppb(以氣體總體積之分子分量計），較佳是低於3〇〇ppb，更 佳是低於lOOppb。量測氣相中低至…帅沖之絕對及相對氮 15 (或硼）濃度需要精密之監測儀器，例如：氣體層析法來予 以達成。此種方法之一個實例被描述如下： 標準氣體層析法（GC)包含：使用一個被最佳化來供最 大流速及最小無效空間（dead v〇iume)之窄孔樣品線，並自 感興趣之點抽取之一個氣體樣品流，然後在被通入廢棄物 20之丽，通過GC樣品線圈。該Gc樣品線圈是一段以一固定 且已知體積來環繞成之管線（對標準大氣壓注射而言，典 5L疋1 cm )’其可自其居於樣品之位置被轉移入該注入氣 體層析管柱中之載體氣體（高純度之氦氣）線。此可將一個 已知體積之氣體樣品置入流入管柱之氣體流中；於該項技 22 1271450 玖、發明說明 藝中，這個步驟被稱為樣品注射。 樣品注射是藉由載體氣體通過第1個GC管柱（填充一種 被取佳化來供分離簡單無機氣體之分子濾器），然後被部 分地分離，然而高濃度之原始氣體（例如··氦、氬）會致使 5該官柱產生完全分離之飽和（例如：氮氣滯流（nitr〇gen difficult))。其後該自第i個管柱流出之部分被注入第2個管 柱’藉此來避免大部分之其他氣體被通入第2個管柱，避 免官柱飽和並使該標的氣體（氮氣N2)完全分離。這個步驟 被稱為“心臟切割，，(“heart-cutting”）。 1〇 第2個管柱之輸出流通過一部排氣離子化偵測儀（DID) ，其可偵測該等由於樣品存在所造成之經由載體氣體漏放 流增加。化學結構可藉由該等以標準氣體混合物來予以校 正之氣體滯留時間來予以鑑定。該DID之反應於5倍之級數 内疋線性的且其是使用特殊校正之氣體混合物來予以校正 15 ，其等典型是居於10_l〇〇PPm之範圍内，藉由測定重量之 刀析法，其後由供應廠商做驗證。該DID之線性關係可以 謹慎之稀釋試驗來予以驗證。 此種已知技藝之氣體層析法已被更進一步修改，並研 么如下來供用於此申請案：於本案分析之方法典型是於 2〇 5〇·5〇〇 X 1〇2 Pa下操作。正常GC操作係使用超過來源氣體 之氣壓之過量壓力來驅動氣體通過樣品線。在本案中，樣 品是藉由在管線之廢棄物端連接一部真空幫浦來予以驅動 ，且該樣品是在低於空氣壓力下被抽取。然而，由於氣體 在管線中流動阻抗會造成管線壓力顯著降低，影響校正及 23 1271450 玖、發明說明 莖敏度。因此，於該樣品線圈及真空幫浦之間置放一個閥 ’其於樣品注射前之一段歷時很短的期間内被關閉，俾使 樣品線圈之壓力能夠穩定並以一壓力計來予以量測。為確 保注入一足量之樣品氣體，該樣品線圈之體積被增加至大 5 約5 cm3。視該樣品管線之設計而定，此技藝可有效地在 低至大約70 X 102Pa之壓力下***作。該gC之校正可視注 入之樣品體積而定，且可藉由同在分析下能夠供應與樣品 使用相同之壓力校正GC來得到最高之準確度。必須觀測 非常高標準之真空及氣體操作，來確保量測是準確的。 10 取樣點可以落在··合成槽之上游（俾以定性輸入氣體） 、居於槽内（俾以定性槽内環境）、或槽之下游。 典型地以乙硼烷（B2H6)來被添加至製程中之硼（B)是使 用一種校正且稱為lOOppm之配於&之乙硼烷（b2h6)來簡化 控制，相同地以氮氣（NO來被添加至製程中之氮是使用一 15種校正且稱為10〇PPm之配於札之氮氣（n2)來簡化控制。添 加之硼（B)及氮（N)皆以ppm來表示，其係以[B2H6]/[所有的 氣體]來計算硼（B)，其中[B2H6]係代表乙硼烷（b2h6)之莫耳 數，而[所有的氣體]則代表所有存在氣體之莫耳數），相同 地其係以[N2]/[所有的氣體]來計算氮氣（n2)。 2〇 遠來源氣體可以是任何一種該技藝中所已知且包含含 碳材料，其可解離產生自由基或其他具反應性之物種。該 氣體混合物亦通常會包含適合提供呈原子形式之氫氣或一 種鹵素。 該來源氣體之解離較佳是於一部使用微波能量之反應 24 1271450 玖、發明說明 器（其等之範例係為該技藝中所已知）内進行。然而，自該 反應器轉移任何-種雜質應減至最少。可使用—種微波系 統來確保電襞被置放在除了該鑽石長晶之基材表面及其基 座（基材載體）之外，遠離所有的表面處。一個較佳之基座 5材料的實例是：翻、鶴、石夕及碳化石夕。—個較佳之反應槽 材料的實例是··不銹鋼、鋁、銅、金及翻。 應使用可產生高微波功率（典型是3_6〇千瓦（kw)來供 用於直徑25-300mm之基材載體）之一種高電漿密度，以及 高氣體壓力（50-500 x l〇2Pa，較佳是 1〇〇 45〇χ ι〇2 ^)。 使用上述條件下，可製造出具有厚的高品質獨捷雜單 晶CVD鑽石層體，其具有異常高之流動性之電荷載體並具 有一種形體最適合用以製造該等供用於商業產品之均一大 體積。 以下描述本發明之實施例。 15 實施例1 適用於合成本發明單晶CVD鑽石之基材可以製備如下： ⑴最佳之原料材料（Ia型天然寶石及比型高壓高溫 (HPHT)寶石）挑選是以顯微鏡審視及雙折射影像來鑑定基 材是不含斑點及雜質。 20 (11)雷射切割、磨光（laPPin§)及拋光來使表面瑕疵減 至最少，使用一種顯示電漿姓刻法來測定該藉由加工來被 導入之瑕/疵位准。 (iii)於最佳化之後，可依慣例來製造基材，其所具有 之可由-種顯示電聚蚀刻㈣予以量測之瑕疲密度係主要 25 1271450 玖、發明說明 是視材質而定，且少於5 X 103/mm2，且通常是少於102 /mm2。藉由此方法來製備基材，並於其後使用於接續之合 成法中。 一種高溫/高壓合成之la型鑽石係於一高壓壓製下長晶 5 ，並使用上述方法來使基材瑕疵減至最少，俾以形成一個 尺寸為7.65mm X 8.25 mm2且厚度為0.54mm之每一面皆為 {100}之拋光平板。於此階段之表面粗糙度RQ是小於lnm 。該基材是使用一高溫之銅鋅合金焊接來予以設置於一鎢 基材上。其被導入一反應器内，並起始一個如上述之電漿 10 及長晶週期· 1) 該2.45GHz反應器被預先以配備之純淨機來調降 居於進料氣體流内之不為所欲之污染物，俾使其低於 80ppb 〇 2) 於270 X 102 Pa及一個753°C之基材溫度下，使用 15 15/75/600sccm(標準每秒立方公分）之氧氣/氬氣/氫氣 (02/Ar/H2)來進行一種原位（in situ)氧氣電漿蝕刻。 3) 此可在不干擾之下，於一 758°C之溫度（歷時10分 鐘内）自氣體流中移除氧氣並移入一種氫氣蝕刻。 4) 此係藉由添加碳源（於此實施例中為CH4)及攙雜氣 20 體來移入長晶之製程。於此實施例中，甲烷（CH4)於氣相中 之流動是30sccm。乙硼烷（B2H6)被使用作為硼攙雜物之來 源。該乙硼烷（B2H6)氣體濃度是1.4ppm。溫度為780°C。 5) 於長晶期完成時，自反應器移除該基材，然後自 該基材移除該生長於上述之一低瑕疵密度之表面之CVD鑽 26 1271450 玖、發明說明 石層體。 6) 其後拋平此層體來產生一粒具有<100>稜線且側 面尺寸大約為5 X 5mm2之均一硼攙雜、厚度735μιη層體。 7) 此被鑑定為CD-1之層體被清洗並以喷氧氣（02)來 5 清除其表面，然後使用沃氏技術（Hall technique)來測試其 移動性。測得其於300K為360cm2/Vs，於440K為185 em2/VS 〇此數據與一個以聲學聲子散射之模式所預測之Τ·2/3 倚賴性是一致的。 8) 使用二次離子質譜儀（SIMS)來分析此層體，可測 10 得其具有均一性之硼（B)總濃度為6.2 X 1018原子/cm3。 9) 使用沃氏技術（Hall technique)測試其載體濃度於 200K為 4·5 X 1013，於300K為 4 X 1015，於 500K為 1·6 X 1017 。該於300K為4 X 1015之載體濃度依照函數（1)所推算之移 動性上限值為163cm2/Vs，其測量值則為360cm2/Vs。因此 15 所顯示之G因子（定義如上述之函數1)較習知技藝之材料高 2.2。 實施例2 以下列改變之條件來重覆實施例1所述之方法： 1) 拋光之高壓高溫（HPHT)基材為厚度500μηι之所有 20 表面皆為{100}之5 X 5mm正方形。 2) 於333 X 102 Pa及一個800°C之基材溫度下，使用 15/75/600sccm(標準每秒立方公分）之氧氣/氬氣/氫氣 (02/Ar/H2)來進行一種原位（in situ)氧氣電漿蝕刻。 3) 繼之以一個歷時30分鐘之氫氣蝕刻，其中氧氣 27 1271450 玖、發明說明 (〇2)自氣體流中被移除，且維持溫度810°C。 4)長晶是藉由添加36sccm之甲烷（CH4)氣流，以及分 別可產生氣體濃度0.05及7ppm之乙硼烷（B2H6)及氮氣（N2) 氣流。溫度為812°C。 5 5)於長晶期完成時，自反應器移除該基材，然後自 該基材移除該CVD鑽石層體。 6)其後拋平此被鑑定為CD-2之層體來產生一粒具有 <100>棱線且側面尺寸大約為7 X 7mm2之均一侧攙雜、厚 度410μιη層體。 10 7)使用SIMS來分析該層體，一系列量測顯示該層體 具有一均一棚濃度為6·1 X 1〇18原子/cm3。該侧（b)濃度之 SIMS圖譜顯示：在該圖譜所具有之解析度下沒有濃度差 異，該圖譜具有一小於30mm之側面空間解析度及一高於 10%之量測靈敏度位准。量測到之氮濃度皆低於5 χ 1〇15原 15 子/cm3 〇 8)此被鑑定為CD_2之層體被清洗並以噴氧氣（〇2)來 清除其表面，然後測試其移動性及載體濃度。測得其載體 濃度超過4.5 χ 1013，且量測其流動性係超過2·5 χ ι〇3 cm2/Vs，產生G值為大約1.5。 2〇 9) CD_2可以下述數據來做進一步定性： (i) 其CL光譜顯示自由及結合激發，且無其他特 徵。 (ii) 電子磁共振譜（EPR)顯示無電中性取代氮且口 具有一 g=2.0028之微弱譜線。 28 1271450 玖、發明說明 (iii) 光譜顯示其具有除了與一6.5 X 1016原子/cm3之 未補償硼濃度有關之特徵性吸收之外之近似理論穿透度。 (iv) X射線振動曲線（X-ray rocking curve)圖譜顯示 該樣品之角度分佈是小於10 arc sec。 5 (v)拉曼光譜顯示一譜線寬度之半高寬值（FWHM) 為大約201^1。 實施例3 以下列改變之條件來重覆實施例1所述之方法： 氬氣（Ar) 75sccm，氫氣（H2) 600sccm，甲烷（CH4) 30 10 seem，330 x 102 Pa，795°C，4.4kW，乙硼烷（B2H6)及氮氣 (N2)氣體濃度分別為15及0.5ppm。 其後適當地加工及分析該厚度為300μιη之CVD層體的 兩面。 其頂面之一個SIMS圖譜顯示一個硼濃度為1.75 X 1018 15 原子/cm3，其底面之一個平均SIMS濃度為1.98 X 1018原子 /cm3 〇 實施例4 以下列改變之條件來重覆實施例1所述之方法： 氬氣（Ar) 50sccm，氫氣（H2) 600sccm，甲烷（CH4) 40 20 seem，330 x 102 Pa，795°C，4.4kW，乙硼烷（B2H6)及氮氣 (N2)氣體濃度分別為0.05及0.7ppm。其後適當地加工及分 析該厚度為113μπι之CVD層體的兩面。 其頂面之一個SIMS圖譜係量測自一個面積2mm X 4.5 mm内之0.5mm凹槽深度及一個面積5mm X 6mm内之1mm凹 29 1271450 玖、發明說明 槽深度。底面之數據則係量測自一個lmm之凹槽深度。因 此，分析之體積為3.4mm3。 量測到之頂面平均硼濃度為0.56，而底面為0.52ppm 。因此測定出該材料具有一平均值居於特定濃度範圍内的 5 體積百分比被顯示於表1 : 表1-於lmm凹槽上分析之SIMS濃度及分佈 SIMS 標準 單 位 詳細 說明 表面 體積 硼濃度平均值 ppm 頂面 底面 0.54 1.0mm 凹槽 0.56 0.52 數值範圍 % 100% -24% 至 +23% -14% 至 + 16% -21 %至+27% (範圍48%) 95% -17% 至 +20% -14% 至 + 11% -17% 至+ 18% (範圍35%) 85% -11% 至 + 14% -11% 至 + 11% -15% 至+13% (範圍28%) 70% -9%至 +9% -7%至 +9% -9%至+ 10% (範圍28%) 因此，由表1顯示100%之硼量測皆落在一個佔樣品頂 面47%之總範圍内，及一個佔樣品底面30%之總範圍内， 10 及於合併分析時佔兩主要表面結合體積之30%。類似地， 這些量測之70%皆落在佔兩表面合併19%之範圍内。 於該層體内量測到之氮濃度是低於〇.〇6ppm，此上限 被設為所使用量測條件之靈敏度。 該樣品底面被進一步使用MonoCL系統來分析自由激 15 發發射自由激子散射（FE)及結合激發發射（BE)強度，求取 一6 X 6矩陣（36個數據點）之lmm凹槽上之數據，其結果顯 示於表2。 30 1271450 玖、發明說明 表2-FE及BE量測之分佈 量測包含之％ 總範圍之數值（平均值之％) 頂面 底面 BE FE BE/FE BE FE BE/FE 100% 41 34 31 95% 39 29 28 90% 25 18 25 85% 20 15 24 70% 14 12 17 因此，該樣品頂面90%之硼（B)量測的自由激發皆落在 平均值25%之範圍内，自由激發是落在25%，結合激發是 5 落在18%，BE/FE比例是落在25%。 實施例5 以實施例4所述之方法長晶成一個層體。其後適當地 加工及分析該厚度為233 μπι之CVD層體的兩面。分析之體 積為 7.0mm3。 10 測定其頂面之彌濃度為0.34ppm，底面為0.29ppm，平 均為0.32ppm。因此測定出該材料具有一平均值居於特定 濃度範圍内的體積百分比被顯示於表3 : 表3-SIMS之硼（B)量測之分佈 層體體積之％ 量測到結合及硼（B)濃度之分佈 下限 上限 範圍 100% -22% + 24% 46% 95% -21% + 19% 40% 85% -13% + 13% 26% 70% -10% + 9% 19% 15 於該層體内量測到之氮濃度是低於0.03ppm，此上限 31 1271450 玖、發明說明 被設為所使用量測條件之靈敏度。 該樣品之頂面及底面被進一步使用MonoCL系統來分 析自由激發發射自由激子散射（FE)及結合激發發射（BE)強 度，求取一 6 X 6矩陣（36個數據點）之1mm凹槽上之數據， 5 其結果顯示於表4。 表4-FE及BE量測之分佈 量測包含之％ 總範圍之數值（平均值之％) 頂面 底面 BE FE BE/FE BE FE BE/FE 100% 20 14 26 19 29 32 95% 16 12 22 17 24 21 90% 13 11 18 14 21 17 85% 11 9 17 13 14 14 70% 10 8 14 11 9 12 使用頂面及底面當做兩個主要表面，這顯示90%自由 激發、結合激發及BE/FE比例皆實質地落在大約平均值 10 30%之分佈範圍内。 實施例6 以實施例4所述之方法長晶成一個層體。其後適當地 加工及分析該厚度為538μιη之CVD層體的兩面。分析之體 積為 16.1mm3。 15 測定其頂面之棚濃度為0.52ppm，底面為0.34ppm，平 均為0.43ppm。因此測定出此層體70%體積是落在平均值 之-23.3至+23.4的範圍内，其為全範圍之46.7%。 其後於該長晶面上，以低於30μιη之解析度來重覆硼 (Β)之SIMS圖譜，進一步顯現局部硼攙入均一度，其結果 32 1271450 玖、發明說明 顯示於下列表5。碳以外之元素分析顯示無雜質超過偵測 限值之0.5ppm。 於該層體内量測到之氮濃度是低於0.03ppm，此上限 被設為所使用量測條件之靈敏度。 5 該樣品之頂面及底面被進一步於SEM使用MonoCL系 統來分析自由激發發射自由激子散射（FE)及結合激發發射 (BE)強度，求取一6 X 6矩陣（36個數據點）之1mm凹槽上之 數據，其結果顯示於表5。 表5-硼（B)濃度及FE及BE量測之分佈 量測包 含之％ 總範圍之數值（平均值之％) 頂面 底面 硼濃 度*1 硼濃 度*2 BE FE BE/FE 硼濃 度*2 BE FE BE/FE 100% 25 30 41 33 30 29 20 13 25 95% 24 24 39 28 27 25 16 12 22 90% 25 18 24 13 10 18 85% 15 22 19 15 23 16 11 8 17 70% 12 14 14 12 17 9 9 8 13 10 mSIMS，解析度 <30μηι + 2SIMS，解析度 <50μιη 此層體亦使用紅外線（IR)吸收光譜來量測一面積5 χ 5mm(36個數據點）上1mm凹槽之未補償硼之差異。90%之 15 量測皆落在一個居於平均值大約34%之範圍内。 使用514nm氬離子雷射冷光於77K下量測其拉曼/光學 冷光光譜。其光譜之特徵為居於大約之鑽石拉曼 譜線，該譜線寬度之半高寬值（FWHM)為1.6CHT1。偵測其 居於575及637nm之零聲子譜線，顯示一波峰強度對比拉漫 20 波峰強度之比例極大值為1:1000。 33 1271450 玖、發明說明 實施例7 以實施例4所述之方法長晶成一個層體。其後適當地加 工為厚度818μιη之層體。使用紅外線（IR)吸收光譜來量測一 面積5 X 5mm(36個數據點）上1mm凹槽之未補償蝴之差異。 5 90%之量測皆落在一個居於平均值大約13%之範圍内。 【圖式簡單說明】 (無） 【圖式之主要元件代表符號表】 (無） 34(g) Under the ultraviolet excitation, the free exciton scattering (FE) intensity at 77 K was measured by the following method to have high uniformity. In particular, the frequency distribution of the FE measured by such a method has a difference of 9 〇% of the measurement of the nominal surface of any one of the layers or the sampling of the layer. The percentage of the average value is less than 50%, more preferably less than 30%. It has been found that the high mobility in the CVD diamonds of the present invention is surprising. The current mode of difference in mobility caused by the concentration of the carrier (or ionizing acceptor) at a concentration of more than 8 χ 1 〇 5 atoms/cm 3 is based on the fact that the boron atom is the main scattering mechanism, and Its distribution is based on its existence. Therefore, this mode implies that it is impossible to reach above this value. In contrast, the results of the present invention described in this specification show that the mode is erroneous because the mobility in the noisy diamonds reported in the prior literature is limited by other factors that can be removed. The CVD single crystal boron doped diamond layer of the present invention may be a layer or region in which the film is independent or constitutes a larger individual or layer of diamond. The larger diamond layer or individual may be an early or polycrystalline diamond produced by CVD or other synthetic methods. The larger diamond layer or individual can be molded to remove nitrogen, or other elements. :〇 The diamond layer or individual of the present invention may be in the form of a gemstone. According to another aspect of the present invention, the present invention provides a layer of a boron-doped single crystal CVD diamond layer. The method comprises the steps of: providing a diamond substrate having a surface having no crystalline (iv) surface, providing a source gas (the source gas comprising a boron 12 1271450 玖, source of the invention) Decomposing the source gas and then allowing the homogeneous epitaxial diamond to grow on the surface substantially free of crystalline germanium, thereby producing a single crystal boron doped diamond layer, preferably having the above type. The basic element of the method is that the diamond must be grown on a surface that does not substantially have a crystalline enthalpy. The use of controlled addition of nitrogen to the source gas may additionally be included in the process of the invention. The nitrogen in the source gas provides additional control over the shape produced by the growing single crystal, and the nitrogen doping ratio is substantially lower than that of the butterfly. Therefore, the addition of nitrogen, calculated as nitrogen molecules, is in the range of more than 10 〇·5 and less than 1000 ppm, and preferably in the range of higher than 丄 and less than 1000 ppm, more preferably Residing in the range of above 3 and below 200 Ppm, the boron of the doped material is present as a scattering color core, and thus has no significant negative influence on the electronic properties of the boron germanium layer, and is significantly increased { 1〇〇} The size of the crystal growth cell and the size of 15 competitive crystal growth grooves (for example: {111}). This means that for growth on a {100} plate, the added nitrogen can grow to maintain a substantial {100} crystal growth cell. Those skilled in the art will understand that the stage of changing the shape using nitrogen can be separated from or continued with the phase of the uniform removal of the heterogeneous layer. 2 Therefore, the uniform boron-doped diamond of the present invention can be applied to a wide range of fields such as an electron, a price measuring instrument, a high-energy electron, and the like. In addition, boron-doped diamonds can be used in other applications where boron (B) doping can advantageously produce uniform color, brightness or other related properties. For example, in some applications such as a cutting blade, boron can be used to make the diamond 13 1271450 玖, the invention stone has a color, thereby improving visual control, and the uniformity of color can be regarded as a kind of heart quality. factor. Alternatively, the diamond can be used in applications such as polishing gemstones, which are generally the same as the color uniformity μ as a factor indicative of quality. For various applications as described above, the body or layer of diamond may be produced by, for example, cutting to produce two or several and usually a large number of small particles or sheets 70 for use in one or more of the above applications. in. The size of the granule or unit may depend on its application. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS In addition to the above features, the single crystal boron doped CVD diamond layer of the present invention may have one or more of the following features in most of its volume, most of which are characterized. As defined above: 1. From any single impurity: bismuth (si), phosphorus (P), sulfur (s), nickel 15 (Nl), agar (Co), aluminum (A1), manganese (Μη), The level of iron free exciton scattering (FE) is not more than 1 ppm, and the total amount of these impurities is not more than 5 ppm. The above-mentioned "impurities, do not contain hydrogen and its isotopes. 2 - a cathode cold light (CL) divergence signal falling at a wide wavelength of 575 nm is very low or absent, and one is excited by a 514 nm argon ion laser at 77K. The optical luminescence (PL) correlation line measured for a 300 mW incident beam has a peak integration area smaller than the Raman line peak integration area of 1332 〇 11_1, preferably 1/100. Is less than 丨/50, more preferably less than 1/300 〇14 1271450 玖, invention description 3. In electronic magnetic resonance spectroscopy (EPR), the concentration of a charged neutral single nitrogen color core [NC] Q is lower than 40 ppb, more typically less than 10 ppb. 4. In electronic magnetic resonance spectroscopy (EPR), a rotational density at g = 2.0028 is less than 1 X l 〇 17 cm · 3, more typically less than 5 X 1016cnT3. Among the 5 crystal diamonds, the spectral line falling at g=2.0028 is related to the concentration of lattice enthalpy, and the larger value is typically present in the natural Ila diamond, which is plastically destroyed in the CVD diamond (via the depression). ), and in the homologetaxial diamonds of poor quality. 5. Has excellent optical properties, its ultraviolet (UV) / visible light and red 10 outer (IR) penetration is similar to the theoretical maximum of lib type diamond, and more specifically, single substituted nitrogen in the ultraviolet (UV) of 270nm The absorption value is very low or absent, and the CH absorption band in the range of 2500 to 3100 CHT1 of the infrared spectrum is small or absent. The absorption spectrum characteristic of the semiconductor boron doped diamond is a starting from the near-infrared spectral region falling at about 370 meV to about The continuous absorption peak of 2.2eV extension 15. This absorption peak is responsible for producing blue (light blue for concentration ~5 X 1015cm_3, dark blue to black for concentration ~5 X 1019cm_3). Three significant broadbandes at 304, 348, and 363 meV were observed below the continuous broadband threshold, which would exhibit a significant amount of structure when viewed at high resolutions. 20 6· X-ray morphology ( X-ray topography) shows the properties of the original substrate in relation to the crystal growth <100> ridgeline grows outward <110> ridgeline. Since the potential compensated nitrogen is substantially lower than boron, the uniformity of distribution in uncompensated boron generally represents the uniformity of all boron concentrations. In addition, the electronic characteristics are mainly based on uncompensated boron concentration rather than all boron concentrations. 15 1271450 玖, invention description. Therefore, the uniformity of uncompensated boron concentration is an important parameter. Diamonds containing uncompensated diamonds exhibit a single phonon absorption characteristic with a maximum value of 1282 cm·1 (159 meV). It has been found that the uncompensated boron concentration has a linear relationship with the effect of the broadband on the absorption coefficient of 1282 cm·1. When measured at room temperature, the boron concentration in ppm is 1·2 X (absorption coefficient of 1282 cm_1). A diamond containing uncompensated boron also exhibits an absorption characteristic that appears to be 2457 cm-i by subtracting the essential two-phonon absorption spectrum. When the characteristic of 1282cm 1 is too weak to be used, its uncompensated boron concentration can be used. 10 Relational formula: Uncompensated butterfly (B) concentration (ppm) = 00142 X The integral absorption coefficient is used to calculate the integral of 2457cnfi. Absorbance coefficient. In a diamond sample parallel to both sides, a uniform measurement of the uniformity of the uncompensated boron concentration can be performed using a Fourier Transform Infrared Photometer (FTIR) absorptimeter in the following manner. A representative infrared spectrum characteristic of a whole sample is characterized by collecting a Fourier-converted infrared spectrometer "奵 瑨 等于 等于 等于 等于 等于 等于 等于 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温 室温The map contains a data point with a minimum of 20 points. Thereafter, based on its average measurement, one of the above relationships is selected and 2〇 is used to derive the uncompensated boron concentration for each position. The concentration map of the concentration is used to determine the uniformity, and the percentage of the measurement is estimated by the average value of the limit value away from the group standard deviation. The ultraviolet luminescence spectrum of the quality butterfly (recorded at 77K) is not one. Falling at 5.22 eV (237.5 nm) has strong binding excitation scattering, and 16 1271450 玖, the invention shows free excitation scattering at 5.27 eV (235.2 nm). For high quality diamonds with boron concentrations above about 1 ppm, this The integrated intensity of the two scattering at 77K is approximately proportional to the concentration of uncompensated boron. The relationship is: [Uncompensated boron concentration in ppm] = 1.86 X 1 (combined excitation intensity) / 5 1 (free The ratio of the different positions in the entire sample to the range of the boron concentration measured throughout the sample can be used to determine the diamond uniformity of the diamond near the surface area. The sample is coated with a thin layer. (5nm), one layer of gold, to avoid the charge effect, is placed in the next scanning electron microscope 10 at a temperature of 77K, and uses a MonoCL system with an acceleration voltage of 15kV, a current of 0.2mA and a size smaller than The spot of 10 μπι X 10 μηι is collected. The UV CL characteristics of the sample can be obtained by collecting the grid lines which are interlaced by two vertical lines with intervals of 500 μm or 1 mm (depending on the area of the culvert). The intersection point within the curve is defined by the spectral line, and a minimum of 30 points and 15 data is collected. Thereafter, the frequency map made by the measured concentration is used to judge the full width distribution of the 90% measurement, and The percentage is expressed as a percentage. This method is applied to measure the intensity of the combined excitation and free excitation scattering and to calculate the ratio of the two intensities. Identical, 20, such that this is equal to the subsequent increase in the observed excitation scattering, unless the combined excitation scattering is completely suppressed at each point. The existence of a strong free excitation means that there is no such thing as a difference ( Dislocation) and impurities. For individual crystals in polycrystalline CVD diamond synthesis, the correlation between low enthalpy and impurity density and high free excitation free radical 17 1271450 玖, invention scatter scattering (FE) scattering has been It has been previously reported that at higher shelf levels (typically above 2〇-25ppm in solids), the free excitation scattering is eventually suppressed by high boron dot defect densities, not because of crystal defects such as poor alignment. The uniformity of freely excited scattering is an excellent one that does not exist. 5 Local South Defect Density Measurement. A typical primary ion spectroscopy (SIMS) analysis method is applied to the primary 02+ electron beam with a main voltage of i〇kv, a beam current of typically 1 μΑ, and a spatial resolution of less than 5 μm. It is typically followed by an analysis point equal to 标.5 mm or 1 mm on the surface of the layer, typically a minimum of 20 points, more preferably a minimum of 40 points, from each side of 10. Correction is made by comparing the implanted standard points. The data obtained from the secondary ion mass spectrometer (SIMS) is analyzed by finding the group average, and then the range of the data is used to represent the different % of the data group. The two opposite major surface settings of a layer are approximately equal weighted scores, with a margin of 15 。. At a detection limit of about 2-5 X 1 〇 14 atoms/cm3, the typical secondary ion mass spectrometer (SIMS) reproducibility can be seen at 5% 〇 for the volume of a material, two The opposite surface is defined by SIMS and be/fe Figure 4, and the thickness of the 2 〇 penetration sample is defined by infrared (ir) absorption. The resolution of the measurement technique (SEM for BE, FE, and uncompensated boron concentrations, and SIMS analysis for all boron concentrations) is related to the boron concentration that can be observed in different types of diamonds. For example, in a polycrystalline diamond with a typical particle size of ΙΟΟμπχ, scanning the sample with a lmm analysis point may average off and not observed in the individual 18 1271450 玖, the invention shows that the grain or the crystal growth slot The difference in boron (B) concentration. When sampling 2 or more data points with a resolution of 5 〇 μπι or lower, it may not be possible to exhibit such a small level difference. For the production of the uniform boron doped CVD diamond single crystal layer of the present invention, it is important to grow crystals on a diamond surface having no crystal enthalpy. In this consideration, 瑕疵 mainly means poor ( ( dis 及 及 及 及 及 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其 其One other type of lattice extensibility damage. Preferably, the substrate is a low birefringence 1& type electrically neutral 10, Ib or na high pressure/high temperature synthetic diamond or a CVD synthetic single crystal diamond. There are two ways to damage the material: negatively affecting the electronic properties (eg, mobility of the hole) and affecting local boron introduction. Since dislocation multiplication occurs during thick layer growth, it is particularly important to control the difference between the substrate and the crystal growth. 15 瑕疵 Density can be most easily accomplished using a plasma or chemical etch (referred to as a display plasma etch) that is most suitable for visualizing the ruthenium, which can be used, for example, by a simple plasma etch as described below. Two types of flaws can be found: 1) The nature of the quality of the various substrate materials. Among the selected natural diamonds, these can be as low as 50/mm2, which is 1〇/mm2, or sometimes as high as 106/mm2 or higher. 2) Those produced by polishing include a row structure along the polishing line and a fine crack formed by fine cracks. Such densities may vary in a sample, and in a poorly polished region or sample, the typical value is 19 1271450 玖, and the invention is in the range of about l〇2/mm2 to over l〇4/mm2. In terms of the surface etched feature density associated with ruthenium as described above, the occlusion density is preferably less than 5 X l 〇 3 / mm 2 , and more preferably less than 1 〇 2 / mm 2 . Therefore, the surface of the substrate of the CVD crystal growth or the surface of the surface can be minimized by having carefully prepared the substrate. Since each stage may affect the density of the crucible falling inside the material that will eventually form the surface of the substrate when finished, it is included in the manufacture of the case to apply to the quarry from the mine (in natural The case of diamonds) or the method of synthesizing (in the case of synthetic materials) materials. Special H) processing steps may include customary diamond processes such as mechanical grinding, lapping and polishing (especially suitable for low level in this application), and less frequently used techniques such as: Laser fabrication or ion implantation and stripping techniques, chemical/mechanical polishing, and liquid and electropolymer fabrication techniques. In addition, the surface Rq (with a styius 15 Profilometer, preferably a flat surface profile measured over a length of 〇〇〇8 mm) should be minimized, and the typical value before plasma etching is Not more than a few nanometers (ie less than 10 nanometers). A particular method of reducing surface damage to a substrate comprises in situ electroforming the surface of the homogenous worm diamond. In theory, these 20 kinds of money do not need to be in situ, and there is no need to grow crystals immediately thereafter, but the maximum benefit can be achieved in situ because it avoids any risk of further physical damage or chemical contamination. When the crystal growth is also based on the electric tight method, the in-situ etching is also usually the most convenient method of depositing or diamond-like crystal growth, but does not use any carbon-containing source 20 1271450 kel. It is usually at a slightly lower temperature to produce a better etch rate control. For example, it may be composed of one or more of the following: (i) An oxygen etching method which mainly uses hydrogen gas, optionally with a small amount of argon gas, and requires - a small amount of ruthenium 2. Typical oxygen (iv) conditions are 50-450 X 102 Pa, a gas containing a gas content of K4%, a argon content of 0-30%, and a balanced gas, all percentages by volume, with A 600_110 (rc substrate temperature (more typically 疋800. 〇 and a typical period of 3-6 〇 minutes. (ii) A hydrogen etching method similar to (1), but without oxygen. (10) Selective For the method of (4), but not argon alone. Hydrogen and oxygen can also be used. For example, halogens, other inert gases or nitrogen are used. ^ 15 The etching method is typically etched by an oxygen gas. Etching with a hydrogen gas and then directly moving into the synthesis by introducing a carbon source gas. The etching time/temperature can be selected so that the damage left by the surface of the manufacturing can be removed without forming a surface of a two-degree crude sugar and There is no extensive engraving of the mosquitoes (for example, diSl〇cation), and deep dimples are created by this. Since etching is an invasion, it is particularly important at these stages that the components are combined. The groove design and material selection are such that no material is transferred into the gas phase by the electricity collector or moved into the surface of the substrate. The chlorine etching is followed by oxygen etching to crystallize the angle caused by oxygen etching. Specificity, which will infringe on this kind of fatigue and provide a flatter, more suitable plane for continuous growth of crystals. The surface or surface of the diamond substrate on which the CVD diamond is grown is preferably 21 1271450 玖, Description of the invention {100}, {110}, {113} or {111} surface. Due to the manufacturing method, the actual sample surface orientation can differ from the theoretical orientation by up to 5 degrees, and some cases can be as high as 10 degrees, although this will It is not preferable to adversely affect the reproducibility. It is also important in the method of the present invention to appropriately control the content of impurities in the ring of the Cvd crystal growth. More specifically, the diamond crystal must occur in a substantial In the air containing contaminants, and the intentionally added boron (and if nitrogen is used) is appropriately controlled. The boron and nitrogen doping concentrations required for control are determined according to the examples, but typically need to be more than 2% by weight. More typical is needed It is more than 10% stable, and even more typically needs to be more than 3%. This kind of 10 control must carefully control the nitrogen impurities of the source gas, because nitrogen is a common pollutant. To achieve this degree of control, carefully add nitrogen. Previously, the level of nitrogen in the source gas was typically maintained in the gas phase below 500 ppb (based on the molecular component of the total gas volume), preferably below 3 ppb, and more preferably below 100 ppb. The measurement of the absolute and relative nitrogen 15 (or boron) concentrations in the gas phase requires a sophisticated monitoring instrument, such as gas chromatography. An example of such a method is described below: Standard gas layer The method of analysis (GC) consists of using a narrow-hole sample line optimized for maximum flow rate and minimum dead space (dead v〇iume), and extracting a gas sample stream from the point of interest, and then passing Into the waste 20, pass the GC sample coil. The Gc sample coil is a line that is surrounded by a fixed and known volume (for standard atmospheric injection, 5 L 疋 1 cm), which can be transferred into the injected gas chromatography tube from its position at the sample. A carrier gas (high purity helium) line in the column. This allows a sample of a known volume of gas to be placed in the gas stream flowing into the column; this step is referred to as sample injection in the art. Sample injection is carried out by a carrier gas through a first GC column (filling a molecular filter that is used to separate a simple inorganic gas) and then partially separated, whereas a high concentration of the original gas (eg, ·, Argon) will cause the column to produce a complete separation of saturation (eg, nitr〇gen difficult). Thereafter, the portion from the i-th column is injected into the second column 'to prevent most of the other gas from being passed into the second column, avoiding the saturation of the column and the target gas (nitrogen N2) ) completely separated. This step is called “heart-cutting.” 1〇 The output of the second column passes through a exhaust ionization detector (DID), which detects the sample due to the sample. There is an increase in leakage flow through the carrier gas. The chemical structure can be identified by the gas residence time corrected by the standard gas mixture. The reaction of the DID is linear in the order of 5 times and it is Use a specially calibrated gas mixture to calibrate 15 , which is typically in the range of 10 〇〇 〇〇 PPm, by the method of determining the weight of the knife, and then verified by the supplier. The linear relationship of the DID can be carefully diluted The test is to be verified. This known technique of gas chromatography has been further modified and studied as follows for the application: the method of analysis in this case is typically 2〇5〇·5〇〇X 1 Operating at Pa2 Pa. Normal GC operation uses excess pressure above the pressure of the source gas to drive the gas through the sample line. In this case, the sample is connected to a vacuum pump at the waste end of the line. Driven, and the sample is drawn below air pressure. However, due to the flow resistance of the gas in the pipeline will cause a significant decrease in line pressure, affecting the correction and 23 1271450 玖, the invention states the stem sensitivity. Therefore, in this sample A valve is placed between the coil and the vacuum pump. It is closed for a short period of time before the sample is injected, so that the pressure of the sample coil can be stabilized and measured by a pressure gauge. With a sufficient amount of sample gas, the volume of the sample coil is increased to a height of about 5 cm3. Depending on the design of the sample line, this technique can be effectively operated at pressures as low as about 70 X 102 Pa. Correcting the volume of the sample to be visually injected, and by providing the same pressure-corrected GC as the sample can be used to obtain the highest accuracy. It is necessary to observe very high standards of vacuum and gas operation to ensure that the measurement is Accurate. 10 The sampling point can fall on the upstream of the synthesis tank (for qualitative input gas), in the tank (in the environment of the qualitative tank), or in the tank Boron (B), which is typically added to the process with diborane (B2H6), is simplified using a calibrated and called 100 ppm of diborane (b2h6) to reduce control, similarly to nitrogen. (NO added to the process nitrogen is simplified using a 15 calibration and called 10 〇 PPm nitrogen gas (n2). The added boron (B) and nitrogen (N) are expressed in ppm. , which calculates [B2H6]/[all gases] to calculate boron (B), where [B2H6] represents the number of moles of diborane (b2h6), and [all gases] represents all gases present. The number of ears) is the same as [N2] / [all gases] to calculate nitrogen (n2). 2 〇 far source gas can be any of the known in the art and contains carbonaceous materials, which can be dissociated Free radicals or other reactive species. The gas mixture will also typically comprise hydrogen or a halogen suitable to provide the atomic form. Dissociation of the source gas is preferably carried out in a reaction using microwave energy 24 1271450 玖, an inventor of the invention (the examples of which are known in the art). However, any impurities transferred from the reactor should be minimized. A microwave system can be used to ensure that the electrode is placed away from all surfaces except the surface of the diamond crystal and its base (substrate carrier). An example of a preferred susceptor 5 material is: turn, crane, stone eve and carbonized stone eve. An example of a preferred reaction tank material is stainless steel, aluminum, copper, gold and turn. A high plasma density that produces high microwave power (typically 3-6 kW (kw) for substrate carriers 25-300 mm in diameter) and high gas pressure (50-500 x l 〇 2 Pa, preferably 1〇〇45〇χ ι〇2 ^). Under the above conditions, a thick high-quality single-crystal single-crystal CVD diamond layer can be produced, which has an abnormally high fluidity charge carrier and has a shape which is most suitable for manufacturing such a product for commercial use. volume. Embodiments of the invention are described below. 15 Example 1 The substrate suitable for the synthesis of the single crystal CVD diamond of the present invention can be prepared as follows: (1) The best raw material (type Ia natural gemstone and specific high pressure high temperature (HPHT) gemstone) is selected by microscopic examination and birefringence image. To identify the substrate is free of spots and impurities. 20 (11) Laser cutting, polishing (laPPin§) and polishing to minimize surface enthalpy, using a display plasma surname method to determine the 瑕/疵 level that is introduced by processing. (iii) After optimization, the substrate can be manufactured by convention, and the enthalpy density measured by the electro-concentration etching (4) is mainly 25 1271450 玖, and the invention description depends on the material. And less than 5 X 103/mm2, and usually less than 102 / mm2. The substrate is prepared by this method and used thereafter in the subsequent synthesis. A high temperature/high pressure synthetic la type diamond is subjected to a high pressure pressing of the long crystal 5, and the above method is used to minimize the substrate crucible to form a size of 7.65 mm X 8.25 mm 2 and a thickness of 0.54 mm. Polished plate with {100} on one side. The surface roughness RQ at this stage is less than 1 nm. The substrate is placed on a tungsten substrate using a high temperature copper-zinc alloy weld. It is introduced into a reactor and starts a plasma 10 and a crystal growth cycle as described above. 1) The 2.45 GHz reactor is pre-equipped with a pure machine to reduce the amount of residence in the feed gas stream. Desirable, make it less than 80ppb 〇 2) At 270 X 102 Pa and a substrate temperature of 753 ° C, use 15 15/75/600sccm (standard cubic centimeters per cubic centimeter) of oxygen / argon / Hydrogen (02/Ar/H2) was used to perform an in situ oxygen plasma etching. 3) This allows oxygen to be removed from the gas stream and transferred to a hydrogen etch at a temperature of 758 ° C (within 10 minutes) without interference. 4) This is done by adding a carbon source (CH4 in this embodiment) and a dopant gas 20 to move into the crystal growth process. In this embodiment, the flow of methane (CH4) in the gas phase was 30 sccm. Diborane (B2H6) is used as a source of boron dopants. The diborane (B2H6) gas concentration was 1.4 ppm. The temperature is 780 °C. 5) Upon completion of the long crystal phase, the substrate is removed from the reactor, and then the CVD drill 26 1271450, which is grown on the surface of one of the above low density, is removed from the substrate, and the stone body is invented. 6) After the flat layer is flattened to produce a <100> The ridge line and the side surface size is approximately 5 X 5 mm 2 of a uniform boron doped, thickness 735 μm layer. 7) The layer identified as CD-1 was cleaned and its surface was removed by oxygen (02) 5 and then tested for mobility using the Hall technique. It is measured to be 360 cm2/Vs at 300K and 185 em2/VS at 440K. This data is consistent with a Τ·2/3 dependence predicted by the pattern of acoustic phonon scattering. 8) Using a secondary ion mass spectrometer (SIMS) to analyze this layer, the total concentration of boron (B) with uniformity is 6.2 X 1018 atoms/cm3. 9) The carrier concentration was tested using a Hall technique to be 4·5 X 1013 at 200K, 4 X 1015 at 300K, and 1.6×1017 at 500K. The carrier concentration of 4 X 1015 at 300 K is estimated to be 163 cm 2 /Vs according to the function (1), and the measured value is 360 cm 2 /Vs. Thus the G factor shown in 15 (defined as function 1 above) is 2.2 higher than the material of the prior art. Example 2 The method described in Example 1 was repeated under the following modified conditions: 1) The polished high-pressure high-temperature (HPHT) substrate was a thickness of 500 μm, and all 20 surfaces were {100} 5 X 5 mm square. 2) Perform an in-situ operation at 333 X 102 Pa and a substrate temperature of 800 °C using 15/75/600 sccm (standard cubic centimeters per cubic centimeter) of oxygen/argon/hydrogen (02/Ar/H2) (in situ) oxygen plasma etching. 3) This was followed by a 30 minute hydrogen etch in which the oxygen 27 1271450 玖, the invention description (〇2) was removed from the gas stream and maintained at a temperature of 810 °C. 4) The crystal growth is obtained by adding a 36 sccm methane (CH4) gas stream and a diborane (B2H6) and nitrogen (N2) gas stream having gas concentrations of 0.05 and 7 ppm, respectively. The temperature is 812 °C. 5 5) Upon completion of the long crystal phase, the substrate is removed from the reactor and the CVD diamond layer is then removed from the substrate. 6) thereafter flattening the layer identified as CD-2 to produce a <100> The ridge line has a side surface size of about 7 x 7 mm2, and the layer side is doped and has a thickness of 410 μm. 10 7) Using SIMS to analyze the layer, a series of measurements showed that the layer had a uniform shed concentration of 6.1 x 1 〇 18 atoms/cm3. The SIMS spectrum of the concentration of the side (b) shows that there is no difference in concentration at the resolution of the map, and the map has a side space resolution of less than 30 mm and a measurement sensitivity level of more than 10%. The measured nitrogen concentrations were all below 5 χ 1〇15 original 15/cm3 〇8) The layer identified as CD_2 was cleaned and the surface was removed by oxygen (〇2), and then tested for mobility. And carrier concentration. The carrier concentration was measured to be more than 4.5 χ 1013, and the fluidity was measured to exceed 2. 5 χ ι 〇 3 cm 2 /Vs, resulting in a G value of about 1.5. 2〇 9) CD_2 can be further characterized by the following data: (i) Its CL spectrum shows free and combined excitation, and has no other characteristics. (ii) Electron magnetic resonance spectroscopy (EPR) showed no neutral neutral substitution of nitrogen and the mouth had a weak line of g = 2.0028. 28 1271450 发明, DESCRIPTION OF THE INVENTION (iii) The spectrum shows an approximate theoretical penetration in addition to the characteristic absorption associated with an uncompensated boron concentration of 6.5 X 1016 atoms/cm3. (iv) The X-ray rocking curve map shows that the angular distribution of the sample is less than 10 arc sec. 5 (v) Raman spectroscopy shows that the full width at half maximum (FWHM) of a line width is approximately 201^1. Example 3 The method described in Example 1 was repeated under the following modified conditions: Argon (Ar) 75 sccm, Hydrogen (H2) 600 sccm, Methane (CH4) 30 10 seem, 330 x 102 Pa, 795 ° C, 4.4 The kW, diborane (B2H6) and nitrogen (N2) gas concentrations were 15 and 0.5 ppm, respectively. Thereafter, both sides of the CVD layer body having a thickness of 300 μm were appropriately processed and analyzed. A SIMS spectrum of the top surface showed a boron concentration of 1.75 X 1018 15 atoms/cm3 and an average SIMS concentration of the bottom surface of 1.98 X 1018 atoms/cm3. Example 4 was repeated with the following changes. Method: argon (Ar) 50sccm, hydrogen (H2) 600sccm, methane (CH4) 40 20 seem, 330 x 102 Pa, 795 ° C, 4.4 kW, diborane (B2H6) and nitrogen (N2) gas concentration They are 0.05 and 0.7 ppm, respectively. Thereafter, both sides of the CVD layer body having a thickness of 113 μm were appropriately processed and analyzed. A SIMS map of the top surface is measured from a groove depth of 0.5 mm in an area of 2 mm X 4.5 mm and a 1 mm recess in an area of 5 mm X 6 mm 29 1271450 玖, invention description groove depth. The data of the bottom surface is measured from a groove depth of 1 mm. Therefore, the volume of the analysis is 3.4 mm3. The measured top boron concentration was 0.56 and the bottom surface was 0.52 ppm. Therefore, it is determined that the material has a mean value of 5 volume percent in a specific concentration range. Table 1 shows: Table 1 - SIMS concentration and distribution analyzed on the lmm groove SIMS standard unit details surface volume boron concentration average ppm Top surface 0.54 1.0mm Groove 0.56 0.52 Value range % 100% -24% to +23% -14% to + 16% -21 % to +27% (range 48%) 95% -17% to +20% -14% to + 11% -17% to + 18% (range 35%) 85% -11% to + 14% -11% to + 11% -15% to +13% (range 28%) 70% - 9% to +9% -7% to +9% -9% to + 10% (range 28%) Therefore, Table 1 shows that 100% of the boron measurement falls on a total range of 47% of the top surface of the sample. Within, and a total range of 30% of the bottom of the sample, 10 and 30% of the combined volume of the two major surfaces in the combined analysis. Similarly, 70% of these measurements fall within a range of 19% of the combined surface. The nitrogen concentration measured in the layer is below 〇.〇6 ppm, which is set to the sensitivity of the measurement conditions used. The bottom surface of the sample was further analyzed by the MonoCL system to analyze the free exciton emission (FE) and the combined excitation (BE) intensity of the 15 excitations, and a 6×6 matrix (36 data points) on the lmm groove was obtained. The data is shown in Table 2. 30 1271450 玖, inventive description Table 2-FE and BE measurement distribution measurement included % Total range value (% of average value) Top surface bottom BE FE BE/FE BE FE BE/FE 100% 41 34 31 95 % 39 29 28 90% 25 18 25 85% 20 15 24 70% 14 12 17 Therefore, the free excitation of the 90% boron (B) measurement on the top surface of the sample falls within the range of 25% of the average value, freely excited It is falling at 25%, the combined excitation is 5 at 18%, and the BE/FE ratio is at 25%. Example 5 The crystal was crystallized into a layer by the method described in Example 4. Thereafter, both sides of the CVD layer body having a thickness of 233 μm were appropriately processed and analyzed. The volume of the analysis is 7.0 mm3. 10 The concentration of the top surface was determined to be 0.34 ppm, the bottom surface was 0.29 ppm, and the average was 0.32 ppm. Therefore, it is determined that the volume percentage of the material having a mean value within a specific concentration range is shown in Table 3: Table 3 - % of the distribution layer volume measured by boron (B) of SIMS, and binding and boron (B) were measured. The upper limit of the concentration lower limit range is 100% -22% + 24% 46% 95% -21% + 19% 40% 85% -13% + 13% 26% 70% -10% + 9% 19% 15 at this layer The measured nitrogen concentration in the body is less than 0.03 ppm, and the upper limit is 31 1271450 玖, and the description of the invention is set to the sensitivity of the measurement conditions used. The top and bottom surfaces of the sample were further analyzed using the MonoCL system to analyze the free exciton emission free exciton scattering (FE) and the combined excitation emission (BE) intensity, and a 1 mm groove of 6 x 6 matrix (36 data points) was obtained. The above data, 5 the results are shown in Table 4. Table 4 - Distribution of FE and BE measurements. % of total range (% of average) Top surface BE FE BE/FE BE FE BE/FE 100% 20 14 26 19 29 32 95% 16 12 22 17 24 21 90% 13 11 18 14 21 17 85% 11 9 17 13 14 14 70% 10 8 14 11 9 12 Use the top and bottom surfaces as the two main surfaces, which shows 90% free excitation, combined excitation and BE The /FE ratios all fall substantially within a distribution of approximately 10 30% of the average. Example 6 Crystallization into a layer was carried out in the same manner as in Example 4. Thereafter, both sides of the CVD layer having a thickness of 538 μm were appropriately processed and analyzed. The volume of the analysis is 16.1 mm3. 15 The ceiling concentration of the top surface was measured to be 0.52 ppm, the bottom surface was 0.34 ppm, and the average was 0.43 ppm. Therefore, it was determined that 70% by volume of the layer body falls within the range of -23.3 to +23.4 of the average value, which is 46.7% of the entire range. Thereafter, the SIMS spectrum of boron (Β) was overlaid on the crystal face at a resolution of less than 30 μm to further reveal local boron enthalpy uniformity. The result is 32 1271450 发明, and the invention description is shown in Table 5 below. Elemental analysis other than carbon showed no impurities exceeding 0.5 ppm of the detection limit. The nitrogen concentration measured in the layer is less than 0.03 ppm, which is set to the sensitivity of the measurement conditions used. 5 The top and bottom surfaces of the sample were further analyzed by SEM using the MonoCL system to analyze the free exciton emission free exciton scattering (FE) and the combined excitation emission (BE) intensity, and a 6×6 matrix (36 data points) was obtained. The data on the 1 mm groove is shown in Table 5. Table 5 - Distribution of boron (B) concentration and FE and BE measurements. % of total range (% of average) Top surface boron concentration *1 Boron concentration *2 BE FE BE/FE Boron concentration* 2 BE FE BE/FE 100% 25 30 41 33 30 29 20 13 25 95% 24 24 39 28 27 25 16 12 22 90% 25 18 24 13 10 18 85% 15 22 19 15 23 16 11 8 17 70% 12 14 14 12 17 9 9 8 13 10 mSIMS, resolution <30μηι + 2SIMS, resolution <50μιη This layer also uses infrared (IR) absorption spectroscopy to measure the difference in uncompensated boron in a 1 mm groove on an area of 5 χ 5 mm (36 data points). 90% of the 15 measurements are in a range of approximately 34% of the average. The Raman/optical luminescence spectrum was measured at 77 K using 514 nm argon ion laser luminescence. Its spectrum is characterized by a diamond Raman line, which has a full width at half maximum (FWHM) of 1.6 CHT1. The zero phonon line at 575 and 637 nm is detected, showing a peak intensity of 1 peak with a maximum intensity of 1:10. 33 1271450 发明, DESCRIPTION OF THE INVENTION Example 7 The crystal was grown into a layer by the method described in Example 4. Thereafter, it is suitably processed into a layer body having a thickness of 818 μm. An infrared (IR) absorption spectrum was used to measure the difference in the uncompensated butterfly of a 1 mm groove on an area of 5 x 5 mm (36 data points). 5 90% of the measurements fall within a range of approximately 13% of the average. [Simple description of the diagram] (none) [The main components of the diagram represent the symbol table] (none) 34