201034790 六、發明說明： 【發明所屬之技術領域】 本發明係關於一種拋光機（polishing machine)、於研機 (lapping machine)的具備四主軸之全自動研磨裳置，用於自 地對半^體晶圓尚效地貫施超1¾精度的研磨加工。 【先前技術】 ◎ 本發_此種半導體晶圓，係使用於電料電子相關機哭、 所謂的0A機器⑽ice Aut〇mati〇n，辦公自動化機器）等_ 體電路，亚且其開發正日益進步，所⑽要求有伴隨機器本身小 型化的極薄化、考慮成品率的更加超高精度的加工精度、以及考 慮生產性的更進一層的擴徑化。 -般而言’料體晶圓是先將圓柱形的石夕結晶體削切為一定 的厚度’為了進-步獲得所期望的厚度，再對經過_的半導體 〇 晶圓進行研削加工。 通⑦研削加工分為粗研削、最終拋光研削等，是通過安 裝在研削盤的主軸下端的杯形砂輪鑽石研磨石來進行的，但是 近來=要切半導體晶圓，要有超高精度的平坦精度和鏡面加 二^用如習知的杯形砂輪鑽石研磨石·。故現有技術會 在半導體晶圓上留下因研細 祕研則而導致的損傷，而不可能實施現在 所：度的平坦精度、鏡面加工，因此有必要在研削 加工後進订進一步的研磨加工。 201034790 另外’現有技術的全自動研磨裝置，是從生產性的觀點來 看在為比較大的直徑的下方平臺與大致相同直徑的上部平臺之 間，借助載體等夾壓多片半導體晶圓，同時實施精研加工 (lapping)或者是拋光加工的研磨加工，但是近來的極薄化、 擴徑化的半導體晶圓則有如下實際情況卻令人擔憂：在加工前 後安裝平臺時或者加工中屢有破損發生，另外，加工後的加工 精度不均勻，從而不能謀求品質的均勻化等。 【發明内容】 本發明爲了解決上述之問題，努力研究所完成者。亦即，本 發明包括：基座；輸出側移送單元，其輸出移送半導體晶圓； 研磨前臨時放置台，其承載利用輪出移送單元輸出移送的兩片 半=體晶圓；第-研磨單元和第二研磨單元，其由具備貼附研 f前臨時放置台所承載的兩片半導體晶圓的卡盤、並列設置的 1主軸所組成；大致正四面检形的旋轉柱，其將第一研磨單 讀第二研磨單元支承在―側面與平行㈣—，而且配設 ^可旋轉；第-研磨台與第二研磨台，其配設為位於第一研磨 早叫及第二研磨單it的正下方；研磨後臨時放置台，其承載 用第二研磨台進行研磨加卫的兩片半導體晶圓；收_移送單 二，其從所述研錢臨時放置㈣納移送半導體㈣；並且研 磨前臨時放置台與第二研磨台是配設在旋轉柱的—面側鮮行 201034790 的另外-面側，第-研料與研贿科放置4配設在旋轉 柱的一面側與平行的另外—面側。 本發明通過上述的構成，可以同時用四主轴來研磨加工兩 片半導體晶圓，並且藉由分開為第—次研磨加1第二次研磨 加工而進行研磨加卫，來謀求研磨加μ及可加工性的效率 化’又’由於各個半導體晶圓貼附在卡盤的下面，所以即使被 極薄化、雜化，也可以在不產生破壞的情況下進行研磨加工， 此外，可以全自動地將多片收_輸出側晶圓盒的半導體晶圓 進行輸出、清洗、第-次研磨加玉、第二次研磨加工、清洗、 收納等步驟’而且本發明可將研磨加工面轉超高精度、為割 時代的發明。 — 因此’本發明的目的在於謀求拋光機、精研機的全自動化， 同時連續地對兩片半導體晶圓進行研磨加卫，從㈣化過程、 提高效率，且本發明的目的還在於創造並提供_種全自動研磨 裝置’可充分對應近麵要求的超高精度、極薄化、擴徑化的 半導體晶圓。 本發明的半導體晶圓的全自動研磨裝置，是在大致正四邊 柱形的旋轉柱的-侧面與平行的另外—側面上，支承分別由一 對主軸組成的第-研磨單元與第二研磨單元，還在旋轉柱的一 面側與平行的另外-面側上’分郝設研麵臨時放置台與第 二研磨台、以及第一研磨台與研磨後臨時放置台，因此可以同 201034790 T用第-研磨單㈣附研磨前臨時放置台所承载的兩片半導體 曰曰圓，亚且以第一研磨台對第一研磨進行加工，接著，使旋轉 柱旋轉四分之-轉，以第二研磨台對第二研磨進行加卫，此時， 在第-研磨臺上可以將接下來的兩片_導體晶圓分別卡在一對 主軸上，通過同時對兩片半導體晶圓實施研磨加工，以及分為 第-研磨加工與第二研磨加工進行研磨加工，來謀求研磨加工 以及可加工性的效率化，另外，由於半導體晶圓分別貼附在卡 盤的下面，所以即使極薄化、擴徑化，也可以在不產生破壞的 情況下進行研磨加工。 【實施方式】 以下’根據實施例的圖式，說明達成上述目的之本發明之具 備四主軸之半導體晶圓的全自動研磨裝置。 圖1是用於說明本發明之具備四主軸之半導體晶圓的全自動 研磨裝置’其實施狀婦平面目，目2是驗綱本發明之具 備四主軸之半導體晶_全自動研縣置，其實施狀概要侧面 圖。 如圖卜2所示，本發明是—種自動地對半導體晶圓^高效地 實施超高精度的研磨加X的、拋賴、精研機的賤四主轴之全 自動研磨裝置’其是用於在半導體晶圓上自動地實施研磨加工的 «裝置，包括：基座1 ;輸_移送單元2，其是將在所述基座 201034790 上具有的半導體晶圓單品地進行輸出移送；研磨前臨時放置台 3,其具有同時承載利用所述輸出移送單元輸出移送的兩片半導體 晶圓w、w的空間；第-研磨單元4和第二研磨單元5，其由在下 端具有分獅賴述研磨前臨時放置台3承_以半導體晶圓 w的卡盤4a、4a、5a、5a ’且並列設置的各自—對的主軸4b、4b 和5b、5b所組成；大致正四面柱形的旋轉柱6，其是以將所述第 一研磨單元4與第二研磨單元5支承在—側面與平行的另一側 〇面，而且配設為可旋轉；第一研磨台7與第二研磨台8，其位於所 述第-研磨單元4以及第二研磨單元5的各自的正下方而配設， 並且分顧於第-次研磨加功第二次研磨加工；研磨後 置台9,其具有同時承載用所述第二研磨台8進行研磨加工的加工 後的兩片半導體晶圓W的空間；以及收納移送單元1〇，其將半導 體晶圓W單品地從所述研磨後臨時放置台9進行收納移送；並且 所述研磨前臨時放置台3與所述第二研磨台8分別配設在旋轉柱6 ©的-綱與平行的另外—麻’所述第—研磨台？與所述研磨後 臨時放置台9分別配設在旋轉柱6的一側面與平行的另外一側面。 [實施例] 亦即，本發明是關於一種可充分對應近來所要求的半導體晶 圓W的極薄化、擴徑化、超高精度的平坦精度、鏡面加工，並且 使用四主軸4b、5a、5a、5b的全自動的拋光機、精研機的入自動 201034790 研磨裝置。 本發明的基座] 整體上呈長方形狀， 當設定。 是用於配設所述研磨裝置的後述的各機構， 且其高度可在考射加m的基礎上進行適 為驅動源；^ 元2包&發軸，魏設在側面而作 ，棒螺釘，、帶等傳導機構自：=: =:::所述輸出側升降機_-框形的輪： 多^半導體ί=所述輸出側晶圓盒載置台2a上載置的研磨前的 +wBaHw，以及棒形的輪出側移送臂此等，盆 附墊、以氣體或油等使半導體晶„單品地魏的圓筒則 ^體晶siw是單品地收納於由所述輸⑽晶圓盒心 置的夕個棒體所形成的支架上。 秘1且’在所述輪出側晶圓盒25中所收納的半導體晶UW是通 讀出側升降機構的升降，移送臂2c的進退、旋轉、驅動，以及 =的真_來_，並且每次兩_在研磨前臨時放 :著，研磨前臨時放置台3具有同時承载以輪出側移送單元2 不机出的程度不停地以純水等液體的水流進、 水流分別將半導體晶圓W、W抬起，同時清洗研磨力^^過液體的 201034790 而且，第一研磨單元4以及第二研磨單元5分別由-對主轴 4b、处、5b、5b組成，在設置各自的主轴处、处、％ ' %通過液 體^者f空將半導體晶圓W _在下端的卡盤4a、4a、5a、5a的 同時’第-研磨單元4的主軸4b、4b以及第二研磨單元5的主轴 b刀别與自由紅轉且自由升降的旋轉柱6平行並被支承。 此外’所_走轉柱6在相對位置上支承第一研磨單元4與第 W磨單元5 ’所述旋轉柱6重複進行旋轉四分之—轉、停止的動 〇 作。 接著，第-研磨台7與第二研磨台8分別在頂面設置拋光用 疋盤、精磨用定盤等研磨用定盤，所述帛—研磨纟7與第二研磨 台8是位於通過所述旋轉柱6的旋轉而旋轉的第一研磨單元4以 及第一研磨單元5的正下方砸馈，所述第—研磨台7用於進行 粗研磨加工等第一次研磨加工，所述第二研磨台8是用於進行最 終拋光研磨加工等第二次研磨加工，通過4b、4b、5b、5b的下降 〇 而將在各自的主軸4b、4b、5b、5b的下端所設置的卡盤4a、4a、 5a、5a上所貼附的各自的半導體晶圓w、w、W、W夾在第一研磨台 7以及第二研磨台8之間，並且實施拋光或者精研等研磨加工。 所述第一研磨台7以及第二研磨台8是通過驅動源的發動機 可旋轉地立設在基座1上，各自的研磨台7、8的軸心與主轴4b、 4b、5b、5b稱微有些偏移’各自向相反方向旋轉而實施第一次研 磨加工以及第二次研磨加工。 201034790 且研磨後臨時放置台9具有承載通過第二研磨單元5進行第 二次研磨加工，並且，同時進行最終拋光研磨加工的兩片半導體 晶圓W、W的空間，通常’通過液體的水流分別將半導體晶圓w、w 抬起，同時清洗研磨加工面。 從所述研磨後臨時放置台9將半導體晶圓w單品地進行收納 移运的收納移送單元10，是可以從研磨後臨時放置台9形成純水 等的液體的流路l〇a’並且通過水流將半導體晶圓w收納在收納侧 晶圓盒10b中，也可以反向配設所述輸出側移送單元2，並且通過 收納側移送臂（未圖示）的進退與吸附墊的吸附而收納在收納側 晶圓盒10b中。 而且，收納側晶圓盒1 〇b載置在收納側晶圓盒載置台i 〇c上， 通過頂面具有所述收納側晶圓盒載置台此的收納側升降機構(未 圖示）的升降’而在收納側晶圓盒1〇b巾的支架中單品地收納半 導體晶圓W。 本發明的半導體晶圓W的賤四主敵全自動研雜置中， 研磨前臨時放置台3與第二研磨台δ分職設於大致正四邊柱形 的旋轉柱6的-面侧與平行的另外—面側，並且第—应 研磨後臨時放置台9分別配設在旋轉柱6的—面側與平行的糾 一面側。 接著，說明本發明的半導體晶圓w㈣磨加卫的行程， 將多片半導體晶圓饮收納於輸 徇出側曰曰012b中’所述輪出側晶圓 201034790 二 =::: 而且’第-片的半導體晶圓w通過在輸 出=送臂2。上所具有的吸― Ο ^ 2c的進退、升降和旋轉而移送到研磨前臨時放置台3的頂 面’通過解除吸附墊的真空_喊餘研磨前臨時放置台3的 ^面’接著’囉地將第二片半導體晶_置於研磨前臨時放置 口 3的頂面。 、所述研磨前臨時放置台3從底面不斷有純轉的液體喷流， 半導體晶圓w在被水流抬起的同時被清洗。 接著，由旋轉柱6所支承的-對主軸4b、4b所組成的第一研 磨手段4下降’在下端的各自的卡盤4a、4a上貼賴研磨前臨時 放置台3抬起的兩片半導體晶圓W、w，帛—研磨單元*上升，旋 轉桂6在水準方向㈣9〇度後停止，通過上述方式半導體晶圓被 移送到第一研磨台7的上方。 且通過使第-研磨單元4下降’兩片半導體晶圓w、w被爽在 ~對主軸4b、4b各自的卡盤4a、4a下面與第一研磨台7的頂面 之間而實施第一次研磨加工。 接著’通過第-研磨單元4的上升與旋轉柱6再次旋轉9〇度 和铋止，第一次研磨加工後的兩片半導體晶圓w、w被移送到第二 201034790 研磨台8的上方，並且通過使第一研磨單元4下降，在第二研磨 台8的頂面實施第二研磨加工。 另外，通過第-研磨單元4的上升與旋轉柱6的再次旋轉9〇 度和停止，半導體晶圓被移送到研磨後臨時放置台9的上方，並 且通過使第-研磨單元4下降，利用喷射氣體等方法開發卡= 4a、4a ’來將半導體晶圓w、w載置在研磨後臨時放置台9上。風 而且’在研磨後臨時放置台9中純水等液體不停地喷流，兩 片半導體晶圓W、W利用收納側移送單㈣，通過收納側升降機構 的升降，依次被收納於内設在收納側晶圓盒⑽中的多個棒體之 間。 上述是根據最初兩片半導體晶圓w、w進行的說明，但此是連 續並自動實施的’故在大致正四邊柱形的旋轉柱6的—側面與平 行的另外-侧面上分別支承第—研磨單元4與第二研磨單元5，在 旋轉柱6的-面側與平行的另外—_上，分別配設研磨前臨時 放置台3與第二研磨台8以及第-研磨台7與研磨後臨時放置台 9，因此，在將研磨前臨時放置台3的兩片半導體晶圓w、w貝^ 在第-研磨單元4時，第二研磨單元5位於第二研磨台8的上方， 在其他的兩片半導體晶圓W、W上實施第二次研磨加工。 而旋轉柱6旋轉90度，第-研磨單元4在第一研磨台？上實 施兩片半導體晶圓w、w的第-次研磨加卫時，第二研磨單元5位 於研磨後臨時放置台9的上方，在_後臨時放置台9上開放卡 12 201034790 盤5a、5a而載置研磨後的其他兩片半導體晶圓⑻、饮。 此外’旋轉柱6旋轉9〇度，第—研磨單元4在第二研磨 上實施兩片半導體晶圓W、W的第二次研磨加工時，第二研磨單元 0w、w’此外’如圖i的假想線所圖示，旋轉柱6旋轉舒，第 Ο -研磨單元4位於研磨後臨時放置台9的上方_置兩片半導體 晶圓W ' W時，第二研磨單元5在第—研磨 片半導體晶圓實施第一次加工。 者'/、他兩 本發明可以全自_且有規則地域上述麵，並且可以對 收納在-個輸出側晶圓盒25中的全部轉體晶圓B次實施研磨 加工，輸出結束的輸出側晶圓盒此是可以和未加工的半導體晶圓 W被收納的接著的輸出側晶圓盒此進行替換的，收納側晶圓盒他 如果滿了’也會和㈣收納側晶圓盒1Gb進行替換。 〇 【圖式簡單說明】 圖1係用於說明本發明之具備四主軸之半導 體晶圓的全自動研磨 裝置’其實施例之概要平面圖。 圖2係用於說明本發明之具備 两王釉之+導體晶圓的全自動研 裝置，其實施例之概要側面圖。 磨 13 201034790 【主要元件符號說明】 W 半導體晶圓 1 基座 2 輸出侧移送單元 2a 輸出側晶圓盒載置台 2b 輸出側晶圓盒 2c 輸出側移送臂 3 研磨前臨時放置台 4 第一研磨單元 4a 卡盤 4b 主軸 5 第二研磨單元 5a 卡盤 5b 主轴 6 旋轉柱 7 第一研磨台 8 第二研磨台 9 研磨後臨時放置台 10 收納側移送單元 10a 水路 10b 收納側晶圓盒 10c 收納側晶圓盒載置台201034790 VI. Description of the Invention: [Technical Field] The present invention relates to a polishing machine and a lapping machine having a four-spindle automatic polishing skirt for use in a half-to-ground manner The wafer is effectively processed with a precision of 13⁄4 precision. [Prior Art] ◎ This is a semiconductor wafer that is used in electric electronic related machines, so-called 0A machines (10) ice Aut〇mati〇n, office automation machines, etc. (10) It is required to have an extremely thinner size with the miniaturization of the machine itself, a more ultra-high-precision machining accuracy in consideration of the yield, and an increase in diameter in consideration of productivity. In general, the "material wafer is first cut into a certain thickness of the cylindrical stone crystal". In order to obtain the desired thickness in advance, the semiconductor wafer is subjected to grinding. Through 7 grinding processing is divided into rough grinding, final polishing and grinding, etc., which is carried out by a cup-shaped grinding wheel diamond grinding stone installed at the lower end of the spindle of the grinding disc, but recently, it is necessary to cut the semiconductor wafer and have ultra-high precision flatness. Accuracy and mirror plus two ^ use the conventional cup-shaped grinding wheel diamond grinding stone. Therefore, the prior art will leave damage on the semiconductor wafer due to the research and development, and it is impossible to carry out the current flatness and mirror processing. Therefore, it is necessary to order further polishing after the grinding process. 201034790 In addition, the prior art fully automatic polishing apparatus sandwiches a plurality of semiconductor wafers by means of a carrier or the like between a lower platform having a relatively large diameter and an upper platform having substantially the same diameter from the viewpoint of productivity. Performing lapping or polishing processing, but recent thinning and expanding semiconductor wafers have the following realities: worrying when installing platforms before and after processing or during processing Damage occurs, and processing accuracy after processing is not uniform, and quality uniformity or the like cannot be achieved. SUMMARY OF THE INVENTION The present invention has been made in an effort to solve the above problems. That is, the present invention includes: a susceptor; an output side transfer unit whose output is transferred to the semiconductor wafer; a temporary placement stage before the grinding, which carries the two-and-a-half-body wafer transferred by the output of the wheel-out transfer unit; the first-grinding unit And a second polishing unit comprising: a chuck provided with two semiconductor wafers carried by the temporary placement stage before the deposition, and a main shaft arranged in parallel; a substantially four-sided rotating column, which will be the first grinding The single-reading second grinding unit is supported on the "side and parallel (four)", and is equipped with a rotatable; the first-polishing table and the second polishing table are arranged to be located in the first grinding early call and the second grinding single it Bottom; temporarily placed after grinding, carrying two semiconductor wafers polished and reinforced by a second polishing table; receiving _ transfer single two, which temporarily transfers (four) from the money to transfer semiconductor (four); and temporarily before grinding The placing table and the second grinding table are disposed on the other side of the fresh-lined 201034790 side of the rotating column, and the first and the slabs are placed on one side of the rotating column and parallel to each other. side. According to the above configuration, the two semiconductor wafers can be simultaneously polished by the four spindles, and the polishing is performed by separately performing the first polishing and the second polishing processing to obtain the grinding and the grinding. The efficiency of the processability is 'and' because each semiconductor wafer is attached to the underside of the chuck, so that even if it is extremely thinned and hybridized, it can be polished without damage, and it can be fully automatic. The semiconductor wafer of the plurality of receiving and outputting wafer cassettes is subjected to steps of outputting, cleaning, first grinding, jade, second polishing, cleaning, and storage, and the present invention can rotate the polished surface to a high precision. For the invention of the era. - Therefore, the object of the present invention is to achieve full automation of the polishing machine and the lapping machine, while continuously grinding and cultivating two semiconductor wafers, from (iv) the process, improving efficiency, and the object of the present invention is to create and Providing a fully automatic polishing device that can fully meet the requirements of ultra-high precision, extremely thinner, and expanded semiconductor wafers. The fully automatic polishing apparatus for a semiconductor wafer of the present invention supports a first-grinding unit and a second grinding unit each composed of a pair of main shafts on a side surface of a substantially square columnar rotating column and a parallel other side surface Also, on one side of the rotating column and on the other side of the parallel side, the placement table and the second polishing table, and the first polishing table and the temporary placement table after grinding are placed, so that the same can be used with the 201034790 T Single (4) with two semiconductor rounds carried by the temporary placement stage before grinding, and the first grinding is processed by the first polishing table, and then the rotating column is rotated by four-turn to the second polishing table. The second polishing is carried out. At this time, the next two _conductor wafers can be respectively clamped on a pair of main shafts on the first polishing table, and the two semiconductor wafers are simultaneously ground and divided into - The polishing process and the second polishing process are performed to improve the polishing process and the workability, and the semiconductor wafers are attached to the lower surface of the chuck, so that they are extremely thin. It can also be ground without any damage. [Embodiment] Hereinafter, a fully automatic polishing apparatus for a semiconductor wafer having four spindles of the present invention which achieves the above object will be described based on the drawings of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a view showing a fully automatic polishing apparatus for a semiconductor wafer having a four-spindle according to the present invention, and a second embodiment of the present invention. A schematic side view of the implementation. As shown in FIG. 2, the present invention is a fully automatic polishing apparatus for automatically performing an ultra-high-precision polishing and X-removing, four-spindle spindle of a semiconductor wafer. The apparatus for automatically performing polishing processing on a semiconductor wafer includes: a susceptor 1; a transfer-transfer unit 2 for output-transferring a semiconductor wafer which is provided on the susceptor 201034790; a front temporary placement stage 3 having a space for simultaneously carrying two pieces of semiconductor wafers w, w transferred by the output transfer unit; a first-grinding unit 4 and a second grinding unit 5, which have a branch at the lower end The pre-grinding temporary placement table 3 is composed of the chucks 4a, 4a, 5a, 5a' of the semiconductor wafer w and the respective pairs of main shafts 4b, 4b and 5b, 5b arranged in parallel; substantially square-sided cylindrical a rotating column 6 which supports the first polishing unit 4 and the second polishing unit 5 on the other side of the side and parallel, and is configured to be rotatable; the first polishing table 7 and the second polishing Stage 8, which is located in the first grinding unit 4 and the second polishing sheet 5 is disposed directly under the respective ones, and is divided into a second grinding process of the first grinding and the second grinding process; and a polishing post 9 having the same processing for carrying out the grinding process by the second polishing table 8 a space for two semiconductor wafers W; and a storage transfer unit 1 that stores and transports the semiconductor wafer W from the post-polishing temporary placement table 9; and the pre-polishing temporary placement table 3 and the The second polishing table 8 is respectively disposed on the first stage of the spin column 6 - the parallel and the other - the first grinding table? The post-polishing temporary placement table 9 is disposed on one side of the spin column 6 and the other side surface in parallel. [Embodiment] In other words, the present invention relates to an extremely thinning, a diameter-enhancing, ultra-high-precision flatness precision, mirror processing, and a four-spindle 4b, 5a, which can sufficiently meet the requirements of the recent semiconductor wafer W. 5a, 5b automatic polishing machine, lapping machine into the automatic 201034790 grinding device. The susceptor of the present invention has a rectangular shape as a whole and is set. It is a mechanism to be described later for arranging the polishing apparatus, and its height can be appropriately driven on the basis of the test plus m; ^ yuan 2 pack & hair shaft, Wei set on the side, rod Screw, belt and other conduction mechanism from:=: =::: The output side elevator _-frame-shaped wheel: Multi-semiconductor ί=The pre-grinding +wBaHw placed on the output side wafer cassette mounting table 2a And a rod-shaped wheel-side shifting arm, etc., a basin-attached pad, a gas crystal or a gas, etc., a semiconductor crystal, a cylinder of a single product, a body crystal siw, is housed in a single product, and is contained in the (10) crystal. On the support formed by the arbor of the round box, the semiconductor crystal UW accommodated in the wheel-side wafer cassette 25 is lifted and lowered by the read-side lifting mechanism, and the transfer arm 2c is moved. Advance and retreat, rotation, drive, and true_to_, and each time two_ temporarily placed before grinding: the temporary placement table 3 before grinding has the same degree of carrying the wheel-side transfer unit 2 does not stop The ground flows with water such as pure water, and the water flow lifts the semiconductor wafers W and W, respectively, while cleaning the grinding force and the liquid. 201034790 Moreover, the first polishing unit 4 and the second polishing unit 5 are respectively composed of a pair of main shafts 4b, 5b, 5b, and at the respective main shafts, at % '%, the semiconductor wafer W is emptied by the liquid _At the lower end of the chucks 4a, 4a, 5a, 5a, the main shafts 4b, 4b of the first grinding unit 4 and the main shaft b of the second grinding unit 5 are parallel to the revolving column 6 which is freely red-turned and freely movable. Further, the 'stationary column 6 supports the first polishing unit 4 and the second grinding unit 5' at opposite positions to repeat the rotation of the rotating column 6 by four revolutions. Each of the first polishing table 7 and the second polishing table 8 is provided with a polishing plate for polishing, such as a polishing disk, a polishing plate, and the like, and the polishing pad 7 and the second polishing table 8 are located The first polishing unit 4 that rotates by the rotation of the rotating column 6 and the first polishing unit 5 are directly fed downward, and the first polishing table 7 is used to perform a first polishing process such as rough polishing processing, the second polishing The stage 8 is used for the second polishing process such as final polishing and polishing, and passes 4b, 4b, 5b, The falling of the 5b is sandwiched between the respective semiconductor wafers w, w, W, W attached to the chucks 4a, 4a, 5a, 5a provided at the lower ends of the respective main shafts 4b, 4b, 5b, 5b. A polishing process such as polishing or lapping is performed between the first polishing table 7 and the second polishing table 8. The first polishing table 7 and the second polishing table 8 are rotatably mounted on the base by an engine of a driving source. In the seat 1, the axes of the respective polishing tables 7, 8 and the main shafts 4b, 4b, 5b, and 5b are slightly offset, and each of them rotates in the opposite direction to perform the first polishing process and the second polishing process. The post-grinding temporary placement table 9 has a space for carrying a second polishing process by the second polishing unit 5, and simultaneously performing two final semiconductor wafers W and W, usually by a liquid flow through the liquid. The wafers w and w are lifted and the polished surface is cleaned. The storage transfer unit 10 that stores and transports the semiconductor wafer w in a single product from the post-polishing temporary placement table 9 is a flow path l〇a' that can form a liquid such as pure water from the temporary placement stage 9 after polishing. The semiconductor wafer w is stored in the storage-side wafer cassette 10b by a water flow, and the output-side transfer unit 2 may be disposed in the reverse direction, and the storage-side transfer arm (not shown) may be moved forward and backward and adsorbed by the adsorption pad. It is accommodated in the storage side wafer cassette 10b. Further, the storage-side wafer cassette 1 〇b is placed on the storage-side wafer cassette mounting table i 〇c, and the storage-side lifting mechanism (not shown) of the storage-side wafer cassette mounting table is provided on the top surface. The semiconductor wafer W is housed in a single product in a holder for accommodating the side wafer cassette. In the fourth embodiment of the semiconductor wafer W of the present invention, the temporary placement stage 3 and the second polishing stage δ are disposed on the side of the rotating column 6 of the substantially square column and parallel. On the other side, and the first to be ground, the temporary placement table 9 is disposed on the side of the surface of the spin column 6 and the side of the parallel correction side. Next, the stroke of the semiconductor wafer w (four) of the present invention will be described, and a plurality of semiconductor wafers will be stored in the output side 曰曰 012b. 'The round-side wafer 201034790 2=::: and ' - The semiconductor wafer w of the sheet passes through the output = arm 2 . The suction ― 2 ^ 2c has the suction, lift and rotation and is transferred to the top surface of the temporary placement table 3 before the grinding. ' By releasing the vacuum of the adsorption pad _ shouting the surface of the temporary placement table 3 before grinding, then '啰The second semiconductor crystal is placed on the top surface of the temporary opening 3 before grinding. The pre-grinding temporary placement table 3 continuously has a purely liquid jet from the bottom surface, and the semiconductor wafer w is cleaned while being lifted by the water flow. Next, the first polishing means 4 composed of the pair of main shafts 4b, 4b supported by the rotating column 6 descends 'on the respective chucks 4a, 4a at the lower end, and the two semiconductors lifted up by the temporary placement stage 3 before grinding are attached. The wafer W, w, the crucible-grinding unit* is raised, and the rotating casket 6 is stopped at a level of (four) 9 degrees, and the semiconductor wafer is transferred to the upper side of the first polishing table 7 by the above-described method. And by first lowering the first polishing unit 4, the two semiconductor wafers w and w are cooled between the lower surfaces of the chucks 4a and 4a of the main spindles 4b and 4b and the top surface of the first polishing table 7, and the first Secondary grinding process. Then, 'the second polishing wafer w and w after the first polishing process are transferred to the second 201034790 polishing table 8 by the rise of the first polishing unit 4 and the rotation of the rotating column 6 again by 9 degrees. Further, by lowering the first polishing unit 4, a second polishing process is performed on the top surface of the second polishing table 8. Further, by the rise of the first-grinding unit 4 and the re-rotation of the spin column 6 by 9 degrees and stops, the semiconductor wafer is transferred to the upper portion of the post-polishing temporary placement table 9, and by lowering the first-grinding unit 4, the ejection is utilized. A method such as a gas development card = 4a, 4a' is used to mount the semiconductor wafers w and w on the temporary placement stage 9 after polishing. In the wind, the liquid such as pure water in the temporary placement stage 9 is continuously sprayed, and the two semiconductor wafers W and W are transported by the storage side (four), and are sequentially stored in the interior by the lifting and lowering of the storage side lifting mechanism. Between the plurality of rods in the storage side wafer cassette (10). The above is based on the description of the first two semiconductor wafers w and w, but this is a continuous and automatic implementation, so that the first side of the rotating column 6 of the substantially square column shape and the other side of the parallel side are respectively supported. The polishing unit 4 and the second polishing unit 5 are disposed on the surface side of the spin column 6 and on the parallel other side, respectively, and the pre-polishing temporary placement table 3 and the second polishing table 8 and the first polishing table 7 and after polishing. The table 9 is temporarily placed. Therefore, when the two semiconductor wafers w and w of the stage 3 are temporarily placed before the polishing, the second polishing unit 5 is located above the second polishing table 8, and the other The second polishing process is performed on the two semiconductor wafers W and W. While the spin column 6 is rotated 90 degrees, the first grinding unit 4 is at the first grinding table? When the first-time polishing of the two semiconductor wafers w and w is performed, the second polishing unit 5 is located above the temporary placement table 9 after the polishing, and the card 12 is opened on the temporary placement table 9 after the _10,347,740 disks 5a, 5a The other two semiconductor wafers (8) after polishing are placed and drinked. In addition, when the rotating column 6 is rotated by 9 degrees, the first polishing unit 4 performs the second polishing process of the two semiconductor wafers W and W on the second polishing, and the second polishing unit 0w, w' is further illustrated in FIG. The imaginary line shows that the rotating column 6 rotates, the third polishing unit 4 is located above the temporary placement table 9 after grinding. When two semiconductor wafers W'W are placed, the second polishing unit 5 is in the first polishing sheet. The semiconductor wafer is processed for the first time. The invention can be fully self-contained and has the above-mentioned surface in a regular area, and can perform grinding processing on all the rotating wafers accommodated in one output side wafer cassette 25, and the output side of the output end The wafer cassette is replaced by the subsequent output side wafer cassette that can be stored with the unprocessed semiconductor wafer W. If the storage side wafer cassette is full, it will be carried out with the (4) storage side wafer cassette 1Gb. replace. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic plan view showing an embodiment of a fully automatic polishing apparatus having a four-spindle semiconductor wafer of the present invention. Fig. 2 is a schematic side view showing an embodiment of a fully automatic polishing apparatus for a + conductor wafer having two king glazes according to the present invention. Grinding 13 201034790 [Description of main component symbols] W Semiconductor wafer 1 pedestal 2 Output side transfer unit 2a Output side wafer cassette mounting table 2b Output side wafer cassette 2c Output side transfer arm 3 Temporary placement table 4 before grinding First grinding Unit 4a Chuck 4b Spindle 5 Second polishing unit 5a Chuck 5b Spindle 6 Rotating column 7 First polishing table 8 Second polishing table 9 Temporary placement table after polishing Storage side transfer unit 10a Water path 10b Storage side wafer cassette 10c Storage Side wafer cassette mounting table