九、發明說明： C發明所屬之技術領域3 發明領域 本發明係有關於半導體發光裝置，特別是包含穿燧接 合件與散射結構之III族氮化物半導體發光裝置。 I：先前技術3 發明背景 半導體發光裝置’包括發光二極體(LEDs)、共鳴腔發 光二極體(RCLEDs)、垂直腔雷射二極體(VCSELs)，以及邊 緣發光雷射’為目前可資利用之最具效率的光源。用以製 造得以在可見光譜範圍内操作之高亮度發光裝置的材料系 統包括群組III-V族半導體，特別是鎵、紹、銦與氮之二元、 二元及四元合金，亦稱ΠΙ族氮化物材料。基本上，hi族氮 化物發光裝置係透過金屬有機化學氣相沉積(MOCVD)、分 子束取向附生(MBE)，或其他取向附生技術，在一藍寶石、 碳化矽、III族氮化物，或其他適當基材上，以取向附生方 式生長一具不同成分與摻雜物濃度之半導體層的堆疊完成 的。該堆疊通常包括一或多層形成於基材上之摻雜比方說 矽的η型層、一形成於該nS層上之發光或主動區域以及 一或多層形成於該主動區域上之摻雜比方說鎂的p型層。形 成於導電基材上之m族氮化物裝置可將其P#n型接觸件形 成於該裝置之相對側邊上。通常，m族氮化物裝置是在絕 緣基材如藍寶石基材上製作，並將兩接觸件形成於該裝置 之同-側邊上。此《置使綠可賤過該接觸件（亦稱往 1356504 上取向附生裝置）或透過該裝置之與該接觸件對立的表面 (亦稱正反器裝置)被抽取。 【發明内容】 發明概要 5 根據本發明之實施例，一III族氮化物裝置包括一第一η 型層、一第一ρ型層，以及一隔離該第一ρ型層與該第一η型 層之主動區域。在某些實施例中，該裝置包括一第二η型層 及一隔離該第一及第二η型層之穿熥接合件。第一及第二接 觸件被電性連接至該第一及第二η型層。該第一及第二接觸 10 件係以相同材料做成，即對該主動區域所發射之光線具有 至少75%反射比的材料。在某些實施例中，該裝置包括一 結構層。在包括結構層及穿縫接合件之裝置中，該結構層 可以設置於該第二η型層與該第二接觸件之間。在不具備穿 燧接合件之裝置中，該裝置可包括一基材，且該結構層可 15 以形成於該基材之與裝置層對立的表面上。 圖式簡單說明 第1圖例示一 III族氮化物正反器發光裝置； 第2及3圖例示包括穿燧接合件之裝置； 第4圖標繪兩個在η型氮化鎵上移位之鋁接觸件的串聯 20 電阻與障礙電壓對溫度之關係； 第5圖例示一多層接觸件； 第6圖係所計算出之铭與銀的反射比； 第7Α、7Β及8圖例示包括散射結構之裝置； 第9及10圖為一小型接合件發光裝置之平面圖與橫斷 6 1356504 面圖； 第11及12圖為一大型接合件發光裝置之平面圖與橫斷 面圖； 第13及14圖為一頂部發光裝置之平面圖與橫斷面圖； . 5 帛15_示~*經封裝之發光裝置；以及 第16圖闡述外部量子效率作為一電流的函數，該電流 用於兩個根據第13及14圖之裝置，其中一者具有結構層而 另一者則不具有結構層。 【實施冷式】 · 10較佳實施例之詳細說明 第1圖例示一職氮化物正反器發光裝置之範例，其包 括-藍寶石基材卜一n型區域2、一主動區域3，以及一p型 區域4。該p型區域與該主動區域之一部分被餘刻去除以使 該η型區域2之-部分暴露出來。一n型接觸件⑺被形成於該 15 η型區域2之該暴露部分之上…p型接觸件9被形成於該p型 區域4之其餘部分之上。 若干因素會限制第！圖之裝置可製造且有利抽取的& « 線量。 首先’銀製P型接觸件之使用會限制第1圖之裝置可操 ' 2〇作的最大接合溫度。該p型接觸件之接觸區域—般大於該n · 型接觸件之接顧域’ 便使該裝置之發光區域最大化， 因為該11鲤接觸件之形成需要將該主動區域之一部分蝕刻 去除。該接觸件9及10被選擇以符合低接觸電阻率，俾減少 必須施加至該裝置之電壓，且符合高反射比，俾將入射於 7 該接觸件上之光線反射回該裝置以透過第1圖之該正反器 的該基材1被抽取《由於該P型接觸件一般大於該η型接觸 件’該ρ型接觸件特別需要具高反射比。高反射比與低接觸 電阻率之結合對III族氮化物裝置，如第1圖所示之裝置的Ρ 型接觸件來說始终難以達成。舉例來說，鋁具有合理的反 射比但無法與Ρ型III族氮化物材料產生良好的歐姆接觸。銀 常被使用，因為其可產生良好的Ρ型歐姆接觸且極具反射 性，但銀對III族氮化物層的附著力不佳且易受電子徙動影 響’而這可能導致惨烈的裝置失敗。為了避免銀製接觸件 中之電子徙動問題’該接觸件可以一或多層金屬層保護。 為了增加裝置之光線輸出，該裝置中之電流必須增加。隨 者電流之增加’裝_置的細作溫度亦會上升。在高於250之 溫度下，該銀製ρ型接觸件上方之保護層與該銀製ρ型接觸 件本身之間之熱膨脹係數的差異可能致使該Ρ型接觸件脫 離該裝置之半導體層，進而導致令人無法接受之高前進電 壓與不均勻之光線輸出。這會限制最大電流密度並終而限 制該裝置之光線輸出。 其次，III族氮化物層（η〜2.4)之高折射指數會產生數個 折射指數具有極大對比之介面；舉例來說，該藍寶石基材 (η〜1.8)與該III族氮化物層之間的介面。折射指數具有極大 對比之介面容易將光線限制在該裝置之中。 根據本發明之貫施例，提供可以增加該裝置之最大操 作溫度’並干擾將光線限制在該裝置中之介面的結構，進 而提高該裝置中所產生且可有利抽取的光線量。以下所述 之範例為III族氮化物發光裝置》III族氮化物裝置之半導體 層具有此一公式AlxINyGazN，其中〇£χ$ΐ、〇$y$i、〇$z$l ' x+y+z=l«III族氮化物裝置層可進一步包括群組m元素如硼 及鉈，且可將氮化物之若干元素置換成磷、砷、銻、或鉍》 雖然下列範例所例示的是III族氮化物裝置，本發明之實施 例亦可以其他in族v族材料系統做成，包括m族磷化物及 III族砷化物、ιι-ν族材料系統，以及任何其他適合製作發 光裝置之材料糸統。 第2及3圖例示本發明之第一實施例。在第2圖之裝置 中，η型區域2、主動區域3，以及p型區域4被形成於一適當 的基材1上之後，一穿燧接合件1〇〇被形成，然後是另一 層7。第3圖例示一含有一穿縫接合件之裝置的替代實作範 例。第3圖之該穿縫接合件1〇〇位於該主動區域之下方而 非如第2圖之實作範例般將其設置於該主動區域之上方。第 3圖之该穿燧接合件1〇〇介於該n型層2與該p型層4之間。因 此，第3圖之該裝置的極性與第2圖之該裝置的極性相反。 該穿燧接合件1〇〇允許生長於該穿燧接合件上方之材料相 對於下方之材料地做導電性變更。 該穿燧接合件100包括一重度摻雜ρ型層5，亦稱ρ++ 層以及重度摻雜η型層6’亦稱η++層。舉例來說，該ρ++ 層5在藍光裝置中可以是氮化銦鎵或氮化鎵，在紫外光裝置 中可以是氮化鋁銦鎵或氮化鋁鎵’並以大約丨〇丨8cm·3至大約 5X10 cm3之濃度摻雜一受體如鎂或鋅。在某些實施例中， 該P㈠層5以大約2xl〇2W至大約4xl〇2W之濃度摻雜雜 質。舉例來說，該n++層6在藍光裝置中可以是氮化銦鎵或 氮化鎵’在紫外光裝置中可以是八丨氮化銦鎵或氮化鋁鎵， 並以大約1018cm·3至大約5xl〇2Gcm·3之濃度摻雜一贈體如 石夕、鍺、硒或碲。在某些實施例中，該n++層6以大約 7xl019cm·3至大約9xl019cm-3之濃度摻雜雜質。該穿燧接合 件100通常極薄，比方說，該穿燧接合件1〇〇可具有大約2奈 米至100奈米之總厚度，且該p+十層5及該n++層6可各自具有 大約1奈米至50奈米之總厚度。在某些實施例中，該？++層5 及該n++層6可各自具有大約25奈米至35奈米之總厚度。該 P++層5及該n++層6不一定要具有相同的厚度。在—實施例 中，該P++層5為15奈米厚之摻雜鎂的氮化銦鎵，而該n++ 層6則為30奈米厚之摻雜矽的氮化鎵。該叶+層5及該於+層6 可具有分等級之摻雜物濃度。舉例來說，該p++層5之靠近 下方P型層4的部分可具有一摻雜物濃度，該摻雜物濃度從 '亥下方p型層之摻雜物濃度開始分級至該p+十層$中之期望 摻雜物濃度。類似地’該奸+層6可具有一摻雜物濃度，該 摻雜物濃度從靠近該？++層5之最大值分級至靠近該η型層7 之最小值。該穿燧接合件1〇〇被做成足夠薄且充分摻雜，以 使。亥穿縫接合件丨⑻在反向偏斜時接近歐姆，亦即，該穿縫 接合件_在反向偏斜模式中導引電流時展現低串聯電壓 下降及低電阻。在某些實關巾，該親接合件1G0在反向 偏斜時的電壓下降在2〇〇編2之電流密度下約為〇 a到 1 V 〇 该穿縫接合件10 〇被做成’當一電壓被施加至該接觸件 1356504 9及10以使遠主動區域3及該p型層4之間的p-n接合處向前 偏斜時，該穿燧接合件100會快速斷裂且以極小的電壓下降 在反向偏斜方向上導電。該穿縫接合件丨〇〇中之各該層不需 要具有相同的成分、厚度或摻雜物成分。該穿燧接合件100 5 亦可在該P++層5及該n++層6之間包括一同時含有p及η型摻 雜物的附加層。 包含一穿燧接合件之發光裝置允許使用兩個η型接觸 件而非不同的η&ρ型接觸件，因為該雙接觸件皆被形成於η 型層上，即該層2及7。兩個η型接觸件之使用可省卻上述銀 10製Ρ型接觸件且導致最大操作溫度之限制。任何對該主動區 域所發射之光線具有大於75%反射比的η型接觸件皆可使 用於正反器裝置中。適當的!!型接觸件範例為鋁。鋁對蝕刻 與未蝕刻之η型III族氮化物皆可產生低電阻接觸。第6圖例 示鋁與銀在介於250至550奈米之波長下所計算出的反射 15 比。第6圖顯示鋁在所例示之範圍内具有高反射比，且在紫 外線波長中比銀更具反射性。由於兩接觸件可以相同材料 作成，某些用以使不同接觸材料沉積於該裝置之該ρ&η型 區域上的沉積及蝕刻步驟可以省略。 該穿燧接合件100亦做為一擴孔層以分配該ρ型層4中 20 之正電荷載體。η型III族氮化物材料中之載體比ρ型hi族氮 化物材料中之載體具有更長的擴散長度，因此電流在η型層 中可以比其在ρ型層中更易於擴散。由於該ρ·η接合部之ρ侧 上的電流擴散產生於該η型層7中’第2及3圖中所示之該裝 置比不具備穿燧接合件之裝置具有更佳的ρ側電流擴散。 11 1356504 第4圖例示—具有鋪觸件之測試裝置的效能。電流與 電磨量測在兩個沉積於同一 n型層上之接觸件之間進行且 紀錄其電阻與障礙電壓(通過非0電流所需之最小電壓）。如 第4圖所示，當溫度上升至6_時，電阻與障礙電壓沒冑 5什麼明顯變化，顯示接觸安定。第2及3圖中所示之該接觸 · 件可以是單層或多層接觸件。單層接觸件可具有大約介於 ; 0.5到5微米之厚度。多層接觸件之範例顯示於第5圖。第5 圖中所示之該接觸件9具有兩層，一厚度大約介於750A與 5000A之間之提供高品質反射器的鋁層9A，以及一厚度大 · 10約介於0·5到5微米之間的鋁合金層9B。該合金層9B避免鋁 在高電流密度下在該紹層9A中產生電子徙動。該合金層9B 中之鋁除外的元素可以少量出現，只要大至足以填滿該鋁 之顆粒邊界即可，比方說少於5%。適當的合金範例為鋁_ 矽、鋁_石夕-鈦、鋁_銅、以及鋁_銅_嫣。該層9A及9B之成分 15可選擇以具有類似的熱膨脹係數俾在升高溫度時避免應力 相關的脫層。 第7A及7B圖例示一包含結構層之裝置的實施例以改 · 善該裝置之光子抽取。結構層12被形成於該第二η型層7 濬 上。由於該結構層基本上與和其最接近之下方層具有相同 2〇的導電類型，在第7Λ及7Β圖所示之實施例中，該結構層12 ' 為一η型層，雖然在其他實施例中，可將一ρ型層結構化。 該結構層12可以任何III族Ν半導體組成，雖然對該主動區域 所發射之光線來說呈透明狀的通常是氮化鎵或Α1氮化銦鎵 之組合物。該結構層12干擾該III族氮化物層之光滑表面並 12 1356504 將光線散射至該裝置外面。該結構層12可以許多此項技藝 所習知之技術形成。舉例來說，該結構層可以在該結構層 生長之前藉由於該裝置上沉積一SiNx“奈米光罩”，亦即一 具不同範圍之8丨队薄層形成。矽在該裝置上的出現可使後 5續生長之氮化鎵的生長模式從二維變成三維，進而造成一 結構表面。該結構層之特性可以透過改變該奈米光罩之厚 度以及在該奈米光罩之頂部沉積氮化鎵所使用的生長條件 來加以調整，如習知技藝所為。 在第7A圖所示之實施例中，該結構層12包括以袋狀物 10 16分離之半導體材料的三角錐或柱，該袋狀物16可以填充 空氣或另一種和III族氮化物材料比較之下具有低折射係數 的材料。舉例來說，低折射係數材料可具有大約少於2的折 射係數。該層12可具有大約200A到1〇,〇00人之厚度，通常大 約介於500A到4000A之間。該袋狀物與該材料之比例可於 15該層12做為袋狀物之容量的大約10%到該層12做為袋狀物 之容量的大約90%之範圍内做改變，其中該層12做為袋狀 物之容量一般大約介於50%到90%之間。 在第7A及7B圖所示之實施例中，一接觸件形成於該結 構層12之上。該接觸件9可以比方說蒸鍍或濺鍍法沉積於該 2〇 結構層12之上以在該結構層12上形成一保角層，如第7B圖 所示。在第7A圖所示之實施例中’可將一具低折射係數之 材料沉積於該袋狀物16中該結構層12之上以做為一厚層， 然後使其圖案化以在該低折射係數材料中開孔至該結構層 12。之後’接觸件13可以比方說蒸鍍或濺鍍法沉積。選擇 13 1356504 性地，第7A圖中之該接觸件π可為—焊接至該結構層12之 平滑的金屬鏡，以將空氣限制於該袋狀物16中。該鏡13可 以藉由在一宿主基材上沉積一反射金屬薄膜做成，其熱性 質類似該裝置，諸如比方說氮化鎵、砷化鎵、氧化鋁、鋼、 5鉬、或矽。然後該鏡/宿主基材之結合體在上升溫度（比方 ' 說，Λ約介於200°C到1，000。(：之間）以及壓力（比方說，大約 - 介於50 psi到500 psi之間）被焊接至該LED晶圓之洗淨表 面，以使該金屬鏡面對該LED晶圓之結構表面。此一金屬 層或透明材料如氧化銦錫層可以在焊接前沉積於該結構表 馨 10面上。此外，該結構層12中之該空氣袋狀物可以在該鏡被 焊接則填充一低折射係數介電質如氟化鎮。該鏡之材料與 焊接方式被選擇以使該裝置之前進電壓不會受到該鏡13之 實質影響。 一使該主動區域所發射之光子極性的可選用極性選擇 15層14，如線柵極化器，可以形成於該基材之與該裝置層相 對的侧邊。線柵極化器在美國專利第6,122,103及6,288,840 號中有詳細說明，其内容在此以參照方式併入本說明書。 ® 線柵極化器反射一與該線平行之極的光子，並傳輸一與該 線垂直之極的光子。如果一光子從該主動區域被發射出來 20並具有一使其從該線柵極化器被反射之極，則它會朝該結 構表面增殖。從該結構表面反射後，該光子之極向會改變， 可能允許該光子穿過該極化器。之後，發射至該裝置外面 的光線會被線性極性。該線柵極化器與該反射結構表面之 結合使該光子再循環直到它們達到某一極性程度為止。該 14 1356504 極性選擇層14可以在加工的任何階段中形成，且通常在燒 結晶圓上之晶粒之如形成最終加工步驟。該線柵極化器可 以下列方式形成：一金屬層被沉積於該晶圓上，然後一光 阻層被沉積於該金屬層上。該光阻通常藉由使其曝光至輻 5射之方式被圖案化，舉例來說，使短波長光線照射一上方 .， 業已形成s玄線拇極化器圖案之光罩、使用來自兩雷射束之 ， 干擾圖案將一具不同強度之光線線條的陣列投射至該光 阻、或者以一電子束將該線柵極化器圖案描繪在該光阻 上。一旦該光阻被曝光，它會被顯影且漂淨，使光阻線條 鲁 10遺留於該金屬層上。該金屬層以化學品（溼式蝕刻）、反廡 離子束(RIE)、電漿強化反應離子束、誘導耗合電漿(icp)、 或其他此項技藝所習知的適當技術被蝕刻。然後殘留之光 阻以化學方式從該晶圓被剝離，進而在該晶圓上留下一金 屬線條圖案。§亥線拇極化器中之線的週期性可以依據該茫 15置之發光波長做最佳化，進而促成相當高的反射效率。 包含散射層12、焊接金屬層13，及極性栅14中任—者 之穿燧接合件裝置亦可形成於一將第7八及7]3圖所示之裝 鲁 置極性反轉的裝置中，如第3圖所示。 在一具有穿燧接合件之裝置上生長一結構層可帶來許 2〇多優點。第7八及7]8圖所示之該裝置中的該穿燧接合件允許 . 該結構層12在一η型層上生長。使psm族氮化物層結構化 具有若干缺點。首先，蝕刻至p型氮化物層之散射層一般無 法提供適合電性接觸之表面。形成於此等散射層上之接觸 件通常明顯增加該裝置之前進電壓且展現不佳的可靠度。 15 1356504 此外’以一siNm罩將一p型結構層形成於—p型層上 會有問題，因為該奈米光罩中之贈體矽的存在可能導致一 P-η接合部被形成，其會增加該LED之前進偏斜電壓。再者， p型結構層中之袋狀物會不利地降低可供電流擴散之p型材 5料的里。結構層在該n型層7上之形成可消除形成於p型m 族氮化物層上之結構層的上述電性及可靠度問題。 第7 A及7B圖所示之該穿燧接合件亦使該結構層可以 位於》玄裝置之該主動區域上方，進而允許該主動區域之生 長可以在该結構層之生長前發生。由於結構化m族氮化物 10層之位移密度通常大於平滑之爪族氮化物層中的位移密 度，後難在一結構表面上生長一高品質主動區域。該穿燧 接合件之使用避免使P型區域及在該主動區域前生長之區 域結構化。 將一鏡13焊接至一結構層12也可改善該裝置中之光線 15抽取。將一扁平鏡焊接至該結構層12會在該鏡與該散射層 之間產生空氣袋狀物16。這些空氣袋狀物亦做為散射中 如果4接觸件係以傳統技術如濺鍵、蒸鑛、或電錢， 而非以焊接方式沉積，則此等袋狀物可能不會形成。 使用該結構層12，及該極性選擇層14，如果極性被認 2〇為有必要的話’可解決某些與透過吸收不正確極性之光線 運作的傳統極化器有關之不足。該結構層12做祕性隨機 益° *具不被期待之極性的光子從該極性選擇層14被反射 時’匕們可能再次從改變該光子之極向的該結構層12被反 射。在經過該極性選擇層14及該結構層12之間的一或多次 16 反射後，該光子可獲得正確的極性以穿過該極化器。因此， 從該主動區域發射之具不正確極性的光子在最終可以獲得 正確的極性。在使用一外部吸收極化器時，最初具有不正 確極性的光子被吸收然後遺失。在沒有該結構層的情況 下’被反射之不正確極性光線的極向鮮少具備隨機性。因 此，此一光線會在該LED中前後反射，直到其被吸收然後 遺失。 第8圖例示一含有一結構化構造之裝置的替代實施 例，以增進該裝置之光子抽取。該結構化構造12被形成於 s亥基材1之背部，與該裝置層相對立。在此一實施例中，該 基材必須具有一實質高於環境媒介之折射係數，以使來自 該主動區域之大部分光線可以與該結構表面產生反應。該 基材之折射係數應該大於1.8。因此’該基材丨通常為碳化 矽(η〜2.5)。第8圖所示之該裝置不需要穿燧接合件。該結構 化構造12可以是比方說一粗糙的η型氮化錄層型接觸 件皆被形成於該基材之與職構對立的㈣。該結構層可 以在该LED裝置層在該基材之對立側邊上生長之前，以取 向附生生長方式沉積。該結構之特徵與參照第7^7b圖所 作之說明相同。 第9圖為一小型接合裝置（亦即小於！平方公羞之面積) 的平面圖。第顧為第9圖中所示裝置之沿轴^顯示的产 斷面圖。第9及剛例示-可使用於第2、3、m、7b^、 中所示之任—取向附生結構2G的接觸件配置。第9㈣圖所 示之裝置具有一單一導孔2卜該單一導孔21向下蝕刻至該 1356504 主動區域下方之取向附生結構20的η型層。-η型接觸件1〇 被沉積於該導孔21中。該η型導孔21位於該裝置之中央 供電流與發光之均句性。一Ρ型接觸件9提供與該取向附生 結構20之該主祕域的㈣側邊進行之電性接觸。在 縫接合件之實施例中，該㈣接觸件9可以形成於__ 且可與該㈣接觸件1G具有相同的結構及材料。在其财施 例中，都型接觸件9可以形成於—p型層上且可做為 層13’如該圖所示。在其他更多的實施财，該P型接觸 10 件9包括一覆蓋—和型接觸件之可選用防護金屬層（圖中 ’以及―設置於該防護金屬層上方之厚P型金屬層。 _接觸件1G以—或多層介電層22與該ρ型接觸件9分 離。-P型基麵接部24，Μ說可連接至焊料之可濕型金 属’連接至該ρ型接觸件9,而一η型基座連接部_連接至 該η型接觸件1〇。 15 如第9圖所示，該裝置以三個基座連接部，即兩個ρ型 基座連接部24及—個η型基座連接部23，被連接至一基座。 如型基座連接部23可以位於該η型接觸區域1〇(由絕緣層 22%、’堯）中任何位置且不需要位於該導孔^之正上方。類似 •基座連接部24可以位於該ρ型接觸區域9中任何位 置如此》玄裝置與該基座之連接並不受限於該ρ型接觸件 9與遠n型接觸件1()之形狀錢置位置。 第Q為大型接合裝置（亦即大於或等於1平方公釐 之面積）的平面圖。第12圖為第U圖中所示裝置之沿轴DD 顯示的橫斷面圓。第U及12圖同樣例示-可使用於第2小 18 1356504 7A、7B及8圖中所示之任一取向附生結構20的接觸件配 置。該取向附生結構20之該主動區域以三個其中具有該η型 接觸件10之溝渠被分割成四個區域。每一區域皆以四個形 成於該ρ型接觸件9上之ρ型基座連接部24被連接至一基 5 座。如上所述，在具有穿燧接合件之裝置中，該ρ型接觸件 9可以形成於一η型層上且可與該η型接觸件10具有相同的 結構及材料。在其他實施例中，該ρ型接觸件9可以形成於 一 ρ型層上且可具有與該η型接觸件10不同的結構或材料， 又該ρ型接觸件9可做為一焊接層13，如第7Α圖所示。該η 10 型接觸件10包圍該四個主動區域。該η型接觸件10以六個η 型基座連接部23被連接至一基座。該η及ρ型接觸件可以一 絕緣層22被電性隔離。 示於第9至12圖之裝置普遍以正反器組態安裝，以便大 部分退出該裝置之光線皆經由該生長基材1退出。第13及14 15 圖例示一頂部發光裝置，其中大部分退出該裝置之光線皆 經由取向附生層之與該接觸件被形成之表面相同的頂部表 面退出。第13圖為該頂部發光裝置之平面圖。第14圖為第 13圖之一部分沿軸Ε顯示的橫斷面圖。雖然第14圖顯示一結 構化頂部取向附生層，該取向附生層20可為第2、3、7Α、 20 7Β及8圖中所示之任何取向附生結構。該ρ型接觸件9之指狀 物插入該η型接觸件10之指狀物。如果該接觸件9及10係以 一對該裝置之該主動區域所發射之光線具吸收性的材料做 成，則其所覆蓋之面積可以縮小。該裝置可以金屬線接合 至一封裝體之引線。 19 1356504 第16圖例示該相對外部量子效率（a.u.)作為電流& $ 數，該電流用於兩個如第13及14圖所示之裝置，其中’素 具有一形成於一穿燧接合件上之結構層而另一者則包含該 穿燧接合件但不具有該結構層。第16圖中之虛線代表具有 5 一結構層之裝置，而實線則代表不具該結構層之裝置°如 第16圖所示，具有一結構層之該裝置比不具該結構層之該 裝置具有更高的外部量子效率，這表示該結構層玎助長你1 該裝置抽取之光線的量。 10 15 20 第15圖為一經封裝之發光裝置的分解圖。一散熱槽100 被置入一插入模鑄式導線架1〇6中。該插入模鑄式導線架 1〇6為比方說一沿一提供電力路徑之金屬框架模造的填充 塑膠材料。該槽1〇〇可包括一可選用之反射杯102。玎為任 何上述裝置之發光裝置的晶粒104被直接或透過一熱傳導 基座103間接安裝至該槽100。可增設一光學透鏡108。 本發明經以上詳細說明後，熟悉此項技藝之人士將理 解基於本發明之揭露，可以在沒有違背此處所述之發明 的精神下對本發明進行修飾。因此，本發明之範圍不 且限制於上文所例示且說明的實施例。 【圖式簡單說明】 第1圖例示~ΙΠ族氮化物正反器發光裝置； 第及3圖例示包括穿縫接合件之裝置；IX. INSTRUCTIONS OF THE INVENTION: TECHNICAL FIELD OF THE INVENTION C FIELD OF THE INVENTION The present invention relates to semiconductor light-emitting devices, and more particularly to group III nitride semiconductor light-emitting devices comprising a pinch-through junction and a scattering structure. I: Prior Art 3 BACKGROUND OF THE INVENTION Semiconductor light-emitting devices 'including light-emitting diodes (LEDs), resonant cavity light-emitting diodes (RCLEDs), vertical cavity laser diodes (VCSELs), and edge-emitting lasers are currently available The most efficient source of light. Material systems for fabricating high-intensity illumination devices that operate in the visible spectral range include group III-V semiconductors, particularly binary, binary, and quaternary alloys of gallium, germanium, indium, and nitrogen, also known as germanium. Group nitride material. Basically, the hi-nitride illuminating device is permeable to metal organic chemical vapor deposition (MOCVD), molecular beam orientation epitaxy (MBE), or other epitaxial epitaxy techniques, in a sapphire, tantalum carbide, group III nitride, or On other suitable substrates, a stack of semiconductor layers of different compositions and dopant concentrations is grown in an epitaxial manner. The stack generally includes one or more layers of doped n-type layer formed on the substrate, a light-emitting or active region formed on the nS layer, and one or more layers of doping ratio formed on the active region. a p-type layer of magnesium. The group m nitride device formed on the conductive substrate can have its P#n type contacts formed on opposite sides of the device. Typically, the m-nitride device is fabricated on an insulating substrate such as a sapphire substrate and the two contacts are formed on the same side of the device. The "green" can be pulled through the contact (also referred to as the epitaxial device on the 1356504) or through the surface of the device opposite the contact (also known as the flip-flop device). SUMMARY OF THE INVENTION According to an embodiment of the present invention, a III-nitride device includes a first n-type layer, a first p-type layer, and an isolation of the first p-type layer and the first n-type The active area of the layer. In some embodiments, the apparatus includes a second n-type layer and a pinch joint that isolates the first and second n-type layers. The first and second contacts are electrically connected to the first and second n-type layers. The first and second contacts 10 are made of the same material, i.e., a material having a reflectance of at least 75% for the light emitted by the active region. In some embodiments, the device includes a structural layer. In a device comprising a structural layer and a seamed joint, the structural layer may be disposed between the second n-type layer and the second contact. In a device that does not have a piercing joint, the device can include a substrate and the structural layer can be formed on a surface of the substrate opposite the device layer. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 illustrates a III-nitride flip-flop illuminating device; Figures 2 and 3 illustrate a device including a splicing joint; and Figure 4 depicts two aluminum displaced over n-type GaN Series 20 resistance of the contact and the relationship between the barrier voltage and temperature; Figure 5 illustrates a multilayer contact; Figure 6 shows the calculated reflectance of silver and silver; and Figures 7, 7 and 8 illustrate the scattering structure. Figure 9 and Figure 10 are a plan view of a small joint light-emitting device and a cross-sectional view of a cross-section 6 1356504; Figures 11 and 12 are a plan view and a cross-sectional view of a large joint light-emitting device; Figures 13 and 14 A plan view and a cross-sectional view of a top light-emitting device; . 5 示 15_ shows a *-encapsulated illuminating device; and Figure 16 illustrates external quantum efficiency as a function of a current used for two according to the 13th And the device of Figure 14, one of which has a structural layer and the other does not have a structural layer. [Implemented Cold Mode] 10 Detailed Description of Preferred Embodiments FIG. 1 illustrates an example of a nitride flip-flop light-emitting device, which includes a sapphire substrate, an n-type region 2, an active region 3, and a P-type region 4. The p-type region and a portion of the active region are removed in detail to expose a portion of the n-type region 2. An n-type contact (7) is formed over the exposed portion of the 15 n-type region 2... a p-type contact 9 is formed over the remaining portion of the p-type region 4. Several factors will limit the number! The device of the figure can be manufactured and advantageously extracted & First, the use of 'silver P-type contacts' limits the maximum junction temperature at which the device of Figure 1 can be operated. The contact area of the p-type contact is generally greater than the contact area of the n-type contact to maximize the illumination area of the device because the formation of the 11-inch contact requires partial etching of the active area. The contacts 9 and 10 are selected to comply with a low contact resistivity, reduce the voltage that must be applied to the device, and conform to a high reflectance, and reflect the light incident on the contact member back to the device to transmit the first The substrate 1 of the flip-flop is drawn "Because the P-type contact is generally larger than the n-type contact", the p-type contact is particularly required to have a high reflectance. The combination of high reflectance and low contact resistivity is always difficult to achieve for a Group III nitride device, such as the Ρ-type contact of the device shown in Figure 1. For example, aluminum has a reasonable reflectance but does not produce good ohmic contact with the bismuth-type III nitride material. Silver is often used because it produces good erbium-type ohmic contact and is highly reflective, but silver has poor adhesion to the III-nitride layer and is susceptible to electron migration' and this can lead to tragic device failure. . In order to avoid electron migration problems in silver contacts, the contacts may be protected by one or more layers of metal. In order to increase the light output of the device, the current in the device must be increased. As the current increases, the temperature of the installation will also rise. At a temperature above 250, a difference in thermal expansion coefficient between the protective layer over the silver p-type contact and the silver p-contact itself may cause the 接触-shaped contact to detach from the semiconductor layer of the device, thereby causing Unacceptable high forward voltage and uneven light output. This limits the maximum current density and ultimately limits the light output of the device. Secondly, the high refractive index of the III-nitride layer (η~2.4) produces a plurality of interfaces with greatly different refractive indices; for example, between the sapphire substrate (η~1.8) and the III-nitride layer Interface. The refractive index has a very contrasting interface that tends to limit light to the device. In accordance with a consistent embodiment of the present invention, a structure is provided which can increase the maximum operating temperature of the device and interfere with the interface that limits light to the device, thereby increasing the amount of light that is generated in the device and that can be advantageously extracted. The semiconductor layer of the group III nitride device of the group III nitride light-emitting device described below has the formula AlxINyGazN, where χ£χ$ΐ, 〇$y$i, 〇$z$l 'x+y+ The z=l«III nitride device layer may further include group m elements such as boron and germanium, and may replace some elements of the nitride with phosphorus, arsenic, antimony, or antimony. Although the following examples illustrate the III group Nitride device, embodiments of the invention may also be made of other in-group v-material systems, including m-group phosphide and III-group arsenide, ιι-ν material system, and any other material system suitable for fabricating illuminating devices . Figures 2 and 3 illustrate a first embodiment of the present invention. In the apparatus of Fig. 2, after the n-type region 2, the active region 3, and the p-type region 4 are formed on a suitable substrate 1, a piercing joint member 1 is formed, and then another layer 7 is formed. . Figure 3 illustrates an alternative embodiment of a device incorporating a seamed joint. The seam joint 1 第 of Fig. 3 is located below the active area and is disposed above the active area as in the example of Fig. 2. The threaded joint 1 〇〇 of Fig. 3 is interposed between the n-type layer 2 and the p-type layer 4. Therefore, the polarity of the device of Fig. 3 is opposite to the polarity of the device of Fig. 2. The piercing joint 1 〇〇 allows the material grown above the piercing joint to be electrically altered relative to the material underneath. The threaded joint 100 comprises a heavily doped p-type layer 5, also referred to as a ρ++ layer and a heavily doped n-type layer 6', also known as a η++ layer. For example, the ρ++ layer 5 may be indium gallium nitride or gallium nitride in the blue light device, and may be aluminum indium gallium nitride or aluminum gallium nitride in the ultraviolet light device and may be approximately cm8 cm. • Doping a receptor such as magnesium or zinc at a concentration of from 3 to about 5 x 10 cm3. In some embodiments, the P(i) layer 5 is doped with impurities at a concentration of from about 2 x 1 〇 2 W to about 4 x 1 〇 2 W. For example, the n++ layer 6 may be indium gallium nitride or gallium nitride in the blue light device. In the ultraviolet light device, it may be tantalum indium gallium nitride or aluminum gallium nitride, and may be about 1018 cm·3 to about. The concentration of 5xl 〇 2Gcm·3 is doped with a gift such as Shi Xi, 锗, selenium or bismuth. In certain embodiments, the n++ layer 6 is doped with impurities at a concentration of from about 7 x 1 019 cm.3 to about 9 x 1019 cm-3. The piercing joint member 100 is generally extremely thin. For example, the piercing joint member 1 can have a total thickness of about 2 nm to 100 nm, and the p+ ten layer 5 and the n++ layer 6 can each have about The total thickness of 1 nm to 50 nm. In some embodiments, what? The ++ layer 5 and the n++ layer 6 may each have a total thickness of between about 25 nanometers and 35 nanometers. The P++ layer 5 and the n++ layer 6 do not have to have the same thickness. In the embodiment, the P++ layer 5 is a 15 nm thick magnesium-doped indium gallium nitride, and the n++ layer 6 is a 30 nm thick germanium-doped gallium nitride. The leaf + layer 5 and the + layer 6 may have a graded dopant concentration. For example, the portion of the p++ layer 5 near the lower P-type layer 4 may have a dopant concentration that is graded from the dopant concentration of the p-type layer below the sub-hierarchy to the p+ ten layer. The desired dopant concentration. Similarly, the + layer 6 can have a dopant concentration from which the dopant concentration is close to? The maximum value of layer ++ is graded to a minimum near the n-type layer 7. The piercing joint member 1 is made thin enough and sufficiently doped. The seam joint 丨 (8) approaches ohms in the reverse deflection, i.e., the seam joint _ exhibits a low series voltage drop and low resistance when conducting current in the reverse skew mode. In some actual closures, the voltage drop of the self-joining member 1G0 during reverse deflection is about 〇a to 1 V at a current density of 2〇〇2, and the seaming joint 10 is made into ' When a voltage is applied to the contacts 1356504 and 10 to deflect the pn junction between the far active region 3 and the p-type layer 4 forward, the piercing joint 100 will break rapidly and be extremely small. The voltage drop conducts in the reverse skew direction. Each of the layers in the seaming member does not need to have the same composition, thickness or dopant composition. The threaded joint 100 5 may also include an additional layer containing both p and n dopants between the P++ layer 5 and the n++ layer 6. A illuminating device comprising a splicing joint allows the use of two n-type contacts instead of different η&p-type contacts, since both contacts are formed on the n-type layer, i.e., layers 2 and 7. The use of two n-type contacts eliminates the above-described silver 10 tantalum contacts and results in a maximum operating temperature limit. Any n-type contact having a reflectance greater than 75% of the light emitted by the active region can be used in the flip-flop device. An example of a suitable!! type contact is aluminum. Aluminum can produce low resistance contacts for both etched and unetched n-type Group III nitrides. Figure 6 shows the reflectance ratio of aluminum and silver calculated at wavelengths between 250 and 550 nm. Figure 6 shows that aluminum has a high reflectance in the range exemplified and is more reflective than silver in the ultraviolet wavelength. Since the two contacts can be made of the same material, some deposition and etching steps for depositing different contact materials on the ρ&n-type region of the device can be omitted. The feedthrough joint 100 also acts as a perforated layer to distribute the positive charge carriers in the p-type layer 4 of 20. The carrier in the n-type Group III nitride material has a longer diffusion length than the carrier in the p-type hi-group nitride material, so that the current can be more easily diffused in the n-type layer than in the p-type layer. Since the current spreading on the ρ side of the ρ·η junction is generated in the n-type layer 7, the device shown in FIGS. 2 and 3 has a better ρ-side current than the device not having the 燧-jointed joint. diffusion. 11 1356504 Figure 4 illustrates the performance of a test device with a contact. Current and electrogrind measurements were made between two contacts deposited on the same n-type layer and recorded their resistance and barrier voltage (the minimum voltage required to pass a non-zero current). As shown in Figure 4, when the temperature rises to 6_, the resistance and the barrier voltage are not significantly changed, indicating contact stability. The contacts shown in Figures 2 and 3 can be single or multi-layer contacts. The single layer contact can have a thickness of between about 0.5 and 5 microns. An example of a multilayer contact is shown in Figure 5. The contact member 9 shown in Fig. 5 has two layers, an aluminum layer 9A providing a high quality reflector between about 750 A and 5000 A, and a thickness of about 10 to about 0.5 to 5 Aluminum alloy layer 9B between microns. The alloy layer 9B prevents aluminum from generating electron migration in the layer 9A at a high current density. The element other than aluminum in the alloy layer 9B may be present in a small amount as long as it is large enough to fill the particle boundary of the aluminum, for example, less than 5%. Examples of suitable alloys are aluminum _ 矽, aluminum _ Shi Xi - titanium, aluminum _ copper, and aluminum _ copper _ 嫣. The composition 15 of the layers 9A and 9B can be selected to have a similar coefficient of thermal expansion and to avoid stress-related delamination at elevated temperatures. Figures 7A and 7B illustrate an embodiment of a device comprising a structural layer to improve photon extraction of the device. A structural layer 12 is formed on the second n-type layer 7 。. Since the structural layer has substantially the same conductivity type as the lower layer closest thereto, in the embodiment shown in Figures 7 and 7, the structural layer 12' is an n-type layer, although in other implementations. In an example, a p-type layer can be structured. The structural layer 12 can be composed of any Group III germanium semiconductor, although a transparent form of the light emitted by the active region is typically a combination of gallium nitride or germanium indium gallium nitride. The structural layer 12 interferes with the smooth surface of the Ill-nitride layer and 12 1356504 scatters light outside of the device. The structural layer 12 can be formed by a number of techniques known in the art. For example, the structural layer can be formed by depositing a SiNx "nanomask" on the device prior to growth of the structural layer, i.e., a thin layer of 8 不同 different layers. The presence of ruthenium on the device can change the growth mode of gallium nitride grown from the second to the third dimension, resulting in a structured surface. The characteristics of the structural layer can be adjusted by varying the thickness of the nanomask and the growth conditions used to deposit gallium nitride on top of the nanomask, as is known in the art. In the embodiment illustrated in Figure 7A, the structural layer 12 includes a triangular pyramid or post of semiconductor material separated by a pocket 10 16 that can be filled with air or another material that is compared to a Group III nitride material. A material with a low refractive index. For example, a low index of refraction material can have a refractive index of less than about two. The layer 12 can have a thickness of about 200A to 1 〇, 〇00 people, and typically between about 500A and 4000A. The ratio of the pouch to the material can vary from 15 to 12% of the capacity of the pouch to about 90% of the capacity of the layer 12 as a pouch, wherein the layer The capacity of 12 as a bag is generally between 50% and 90%. In the embodiment shown in Figures 7A and 7B, a contact is formed over the structural layer 12. The contact member 9 can be deposited, for example, by evaporation or sputtering, onto the 2 〇 structural layer 12 to form a conformal layer on the structural layer 12, as shown in Figure 7B. In the embodiment shown in Fig. 7A, a material having a low refractive index can be deposited on the structural layer 12 of the pouch 16 as a thick layer and then patterned to be low. A hole is made in the refractive index material to the structural layer 12. Thereafter, the contact member 13 can be deposited, for example, by evaporation or sputtering. Alternatively, the contact π in Figure 7A can be a smooth metal mirror welded to the structural layer 12 to confine air to the pocket 16. The mirror 13 can be formed by depositing a reflective metal film on a host substrate, the thermal properties of which are similar to the device, such as, for example, gallium nitride, gallium arsenide, aluminum oxide, steel, molybdenum, or hafnium. The mirror/host substrate combination is then at an elevated temperature (say, say, between about 200 ° C and 1,000 ° (between) and pressure (say, about - between 50 psi and 500 psi) Between the solder surface of the LED wafer being soldered to the surface of the LED wafer. The metal layer or a transparent material such as an indium tin oxide layer may be deposited on the structure prior to soldering. In addition, the air pocket in the structural layer 12 can be filled with a low refractive index dielectric such as fluorinated town when the mirror is soldered. The material of the mirror and the soldering method are selected so that The forward voltage of the device is not affected by the physical influence of the mirror 13. A selectable polarity of the polarity of the photon emitted by the active region is selected as 15 layers 14, such as a wire gater, which may be formed on the substrate The opposite side of the device layer. The wire gate is described in detail in U.S. Patent Nos. 6,122,103 and 6,288,840, the disclosure of which is incorporated herein by reference. a pole of photons and transmits a line perpendicular to the line Photon. If a photon is emitted 20 from the active region and has a pole that is reflected from the line concentrator, it will proliferate toward the surface of the structure. After reflection from the surface of the structure, the photon The polar orientation may change, possibly allowing the photon to pass through the polarizer. Thereafter, the light emitted to the outside of the device will be linearly polarized. The combination of the wire gater and the surface of the reflective structure causes the photons to recirculate until they Up to a certain degree of polarity, the 14 1356504 polarity selective layer 14 can be formed in any stage of processing, and typically the grain on the sintered wafer forms a final processing step. The line gater can be formed in the following manner a metal layer is deposited on the wafer, and then a photoresist layer is deposited on the metal layer. The photoresist is typically patterned by exposing it to a radiation pattern, for example, to make it short The wavelength light illuminates above. The smear of the sinusoidal polarizer pattern is formed, and the interference pattern is used to project an array of light lines of different intensities to the photoresist. Or drawing the line gater pattern on the photoresist with an electron beam. Once the photoresist is exposed, it is developed and rinsed, leaving the photoresist line 10 on the metal layer. The metal layer is etched with a chemical (wet etch), a reverse iridium ion beam (RIE), a plasma enhanced reactive ion beam, an induced plasma (icp), or other suitable technique as is known in the art. The residual photoresist is chemically stripped from the wafer, leaving a metal line pattern on the wafer. The periodicity of the line in the thumb polarizer can be based on the wavelength of the light emitted by the 茫15 Optimization, which in turn contributes to a relatively high reflection efficiency. The device comprising the scattering layer 12, the solder metal layer 13, and the polarity gate 14 can also be formed in a 7th and 7th 3th The device shown in Figure 3 is shown in Figure 3. The growth of a structural layer on a device having a piercing joint provides many advantages. The piercing joints in the apparatus shown in Figures 7 and 7] allow the structural layer 12 to grow on an n-type layer. Structuralizing the psm nitride layer has several disadvantages. First, the scattering layer etched into the p-type nitride layer generally does not provide a surface suitable for electrical contact. Contacts formed on such scattering layers typically significantly increase the forward voltage of the device and exhibit poor reliability. 15 1356504 Furthermore, it is problematic to form a p-type structural layer on a p-type layer with a siNm mask, because the presence of a donor raft in the nano-mask may result in a P-n junction being formed, which would Increase the forward bias voltage of the LED. Furthermore, the pockets in the p-type structural layer can disadvantageously reduce the amount of p-profile material available for current spreading. The formation of the structural layer on the n-type layer 7 can eliminate the above-mentioned electrical and reliability problems of the structural layer formed on the p-type m-type nitride layer. The piercing joint shown in Figures 7A and 7B also allows the structural layer to be positioned above the active region of the "destructive" device, thereby allowing the growth of the active region to occur prior to growth of the structural layer. Since the displacement density of the structured m-nitride 10 layer is generally greater than the displacement density in the smoothed-claw nitride layer, it is difficult to grow a high-quality active region on a structured surface. The use of the piercing joint avoids structuring the P-type region and the region grown prior to the active region. Welding a mirror 13 to a structural layer 12 also improves the extraction of light 15 in the device. Welding a flat mirror to the structural layer 12 creates an air pocket 16 between the mirror and the scattering layer. These air pockets are also used as scattering. If the 4 contacts are deposited by conventional techniques such as splashing, steaming, or electricity, rather than by soldering, such pockets may not form. The use of the structural layer 12, and the polarity selective layer 14, if the polarity is recognized as necessary, can address some of the deficiencies associated with conventional polarizers that operate by absorbing light of incorrect polarity. The structural layer 12 is made to be secretly random. When a photon having an undesired polarity is reflected from the polarity selective layer 14, it is possible that the structural layer 12, which changes the direction of the photon, is reflected again. After one or more reflections between the polarity selective layer 14 and the structural layer 12, the photons can obtain the correct polarity to pass through the polarizer. Therefore, photons with incorrect polarity emitted from the active region can eventually obtain the correct polarity. When an external absorption polarizer is used, photons that initially have an incorrect polarity are absorbed and then lost. In the absence of this structural layer, the polarity of the incorrectly polarized light that is reflected is less random. Therefore, this light will be reflected back and forth in the LED until it is absorbed and then lost. Figure 8 illustrates an alternate embodiment of a device having a structured construction to enhance photon extraction of the device. The structured structure 12 is formed on the back of the substrate 1 opposite the device layer. In this embodiment, the substrate must have a refractive index substantially above ambient medium such that a substantial portion of the light from the active region can react with the surface of the structure. The substrate should have a refractive index greater than 1.8. Therefore, the substrate 丨 is usually ruthenium carbide (η to 2.5). The device shown in Figure 8 does not require a piercing joint. The structured structure 12 may be, for example, a rough n-type nitride layer type contact member formed on the substrate opposite to the structure (4). The structural layer can be deposited in an epitaxial growth mode prior to growth of the LED device layer on opposite sides of the substrate. The features of this structure are the same as those described with reference to Figure 7^7b. Figure 9 is a plan view of a small joint device (i.e., smaller than the area of the square shame). The above is a cross-sectional view of the device shown in Fig. 9 along the axis ^. Ninth and just-exemplified - the contacts for the any-oriented epitaxial structure 2G shown in the second, third, m, and 7b can be arranged. The device shown in Figure 9(4) has a single via 2 which is etched down to the n-type layer of the epitaxial structure 20 below the active region of the 1356504. An n-type contact member 1 is deposited in the via hole 21. The n-type via 21 is located at the center of the device for the uniformity of current and illumination. A 接触-type contact member 9 provides electrical contact with the (four) side of the main domain of the oriented epitaxial structure 20. In the embodiment of the seam joint, the (four) contact member 9 may be formed in __ and may have the same structure and material as the (four) contact member 1G. In its example, the contact member 9 may be formed on the -p type layer and may be used as the layer 13' as shown in the figure. In other implementations, the P-contact 10 piece 9 includes an optional protective metal layer for the cover-and-type contact (in the figure 'and a thick P-type metal layer disposed over the protective metal layer. _ The contact 1G is separated from the p-type contact 9 by a multi-layer dielectric layer 22. A P-type base contact portion 24, to which the wettable metal 'which is connectable to the solder is connected, is connected to the p-type contact member 9, And an n-type base connection portion _ is connected to the n-type contact member 1 15 15 as shown in Fig. 9, the device has three base connection portions, that is, two p-type base connection portions 24 and one The n-type base connecting portion 23 is connected to a pedestal. The sized base connecting portion 23 may be located at any position of the n-type contact region 1 〇 (by the insulating layer 22%, '尧) and need not be located at the guide The hole ^ is directly above. Similarly, the base connecting portion 24 can be located at any position in the p-type contact region 9 such that the connection between the sinuous device and the pedestal is not limited to the contact of the p-type contact member 9 with the far n-type. The shape of the piece 1() is set to the position. The Qth is a plan view of the large joint device (that is, an area larger than or equal to 1 square mm). 2 is a cross-sectional circle along the axis DD of the device shown in Figure U. Figures U and 12 are also exemplified - can be used for any orientation shown in Figure 2 1356504 7A, 7B and 8 The contact structure of the epitaxial structure 20. The active region of the epitaxial structure 20 is divided into four regions by three trenches having the n-type contact 10. Each of the regions is formed by four in the p The p-type base connecting portion 24 on the type contact member 9 is connected to a base 5 seat. As described above, in the device having the piercing joint member, the p-type contact member 9 may be formed on an n-type layer and The n-type contact 10 can have the same structure and material. In other embodiments, the p-type contact 9 can be formed on a p-type layer and can have a different structure or material than the n-type contact 10. Further, the p-type contact member 9 can be used as a solder layer 13, as shown in Fig. 7. The η 10 type contact member 10 surrounds the four active regions. The n-type contact member 10 has six n-type pedestals. The connecting portion 23 is connected to a pedestal. The η and p-type contacts can be electrically isolated by an insulating layer 22. As shown in Figures 9-12 The arrangement is generally installed in a flip-flop configuration such that most of the light exiting the device exits via the growth substrate 1. Figures 13 and 14 15 illustrate a top illumination device in which most of the light exiting the device is oriented. The top surface of the epiphytic layer that is identical to the surface on which the contact is formed exits. Figure 13 is a plan view of the top illumination device. Figure 14 is a cross-sectional view of the portion of Figure 13 along the axis. Figure 14 shows a structured top-oriented epitaxial layer which may be any of the oriented epitaxial structures shown in Figures 2, 3, 7Α, 20 7Β and 8. The fingers of the p-type contact 9 The fingers of the n-type contact 10 are inserted. If the contacts 9 and 10 are made of a material that absorbs light from the active area of the device, the area covered can be reduced. The device can be wire bonded to the leads of a package. 19 1356504 Figure 16 illustrates the relative external quantum efficiency (au) as a current & $ number for two devices as shown in Figures 13 and 14, wherein the element has a bond forming member The upper structural layer and the other comprise the piercing joint but does not have the structural layer. The dotted line in Fig. 16 represents a device having a structural layer of 5, and the solid line represents a device having no structural layer. As shown in Fig. 16, the device having a structural layer has a device having no structural layer. Higher external quantum efficiency, which means that the structural layer contributes to the amount of light that you extract from the device. 10 15 20 Figure 15 is an exploded view of a packaged light-emitting device. A heat sink 100 is placed in an insert molded lead frame 1〇6. The insert molded lead frame 1〇6 is, for example, a filled plastic material molded along a metal frame providing a power path. The slot 1 can include an optional reflector cup 102. The die 104 of the illumination device of any of the above devices is indirectly mounted to the slot 100 either directly or through a thermal conduction pedestal 103. An optical lens 108 can be added. Having described the invention in detail, it will be understood by those skilled in the art that the present invention may be modified without departing from the scope of the invention. Therefore, the scope of the invention is not limited to the embodiments illustrated and described above. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 illustrates a ~ ΙΠ nitride nitride flip-flop illuminating device; and Fig. 3 exemplifies a device including a seaming joint;
圖彳讀兩個在_氮化鎵上移位之祕觸件的串聯 且與障礙電壓對溫度之關係； 第5圖例示-多層接觸件; 20 1356504 第6圖係所計算出之鋁與銀的反射比； 第7A、7B及8圖例示包括散射結構之裝置； 第9及10圖為一小型接合件發光裝置之平面圖與橫斷 面圖； 5 第11及12圖為一大型接合件發光裝置之平面圖與橫斷 面圖； 第13及14圖為一頂部發光裝置之平面圖與橫斷面圖； 第15圖例示一經封裝之發光裝置；以及 第16_闡述外部量子效率作為一電流的函數，該電流 10 用於兩個根據第13及14圖之裝置，其中一者具有結構層而_ 另一者則不具有結構層。 【主要元件符號說明】 1.··藍寶石基板基材 16…袋狀物 2…η型區域 20…取向附生結構 3···主動區域 21…導孔 4…ρ型區域 22…介電層 5···重度摻雜ρ型層 24···ρ型基座連接部 6···重度捧雜η型層 23…η型基座連接部 7…η型層 100…穿燧接面接合件 9···ρ型接觸件接觸面 100…散熱槽 9Α…紹層 106···插入模鑄式導線架 10…η型接觸面接觸件 102…反射杯 12…結構層 104···晶粒 13…接觸件接觸面 103…熱傳導基座 14…極性選擇層 108…光學透鏡 21The figure reads the series connection of two secret contacts displaced on _ gallium nitride and the relationship between the barrier voltage and temperature; Figure 5 illustrates the multilayer contact; 20 1356504 Figure 6 shows the calculated aluminum and silver Reflectance ratio; Figures 7A, 7B and 8 illustrate devices including a scattering structure; Figures 9 and 10 are plan and cross-sectional views of a small junction light-emitting device; 5 Figures 11 and 12 show a large joint illumination Plan view and cross-sectional view of the device; Figures 13 and 14 are plan and cross-sectional views of a top light-emitting device; Figure 15 illustrates a packaged light-emitting device; and Figure 16 illustrates external quantum efficiency as a function of current This current 10 is used for two devices according to Figures 13 and 14, one of which has a structural layer and the other has no structural layer. [Description of main component symbols] 1.·Sapphire substrate substrate 16...Bag 2...n-type region 20...Oriented epitaxial structure 3··active region 21...via 4...p-type region 22...dielectric layer 5··· heavily doped p-type layer 24···p-type base connection part 6···heavy n-type layer 23...n-type pedestal connection part 7...n-type layer 100...penetration joint Piece 9···p-type contact contact surface 100...heat-dissipating groove 9Α...sole layer 106···inserted die-type lead frame 10...n-type contact surface contact piece 102...reflective cup 12...structural layer 104···crystal Particle 13... Contact contact surface 103... Thermal conduction pedestal 14... Polarity selection layer 108... Optical lens 21