201017341 六、發明說明： 【發明所屬之技術領域】 本發明是關於光照射裝置。尤其是，關於使用於半導 體元件或液晶顯示基板的製造用曝光裝置的光照射裝置。 【先前技術】 傳統上，在被使用於半導體晶圓或液晶基板的曝光裝 ❿ 置的光源，使用著數kW至數10kW的超高壓水銀燈。該 燈是具高信賴性之故，因而以往就被使用，惟最近，隨著 工件的大面積化，例如在液晶曝光用燈等也使用著25kW 的大型燈。但是，燈的大型化是與構成燈的燈泡（發泡管） 或電極材料的大型化有直接關聯，而會大幅度地增加製造 成本與製作工數有關聯，因而逐漸接近極限。 一方面，作爲燈大型化的對策，例如在日本特開 20 04-3 61 746號（專利文獻1)，提案一種不是以1個大型燈 ® 構成光源，而是排列複數小型燈的構成。 提示於上述專利文獻1的光照射裝置，是例如排列放 電燈與反射鏡所構成的35個單元以構成1個光源部者， 而使用積分器將從該光源部所放射的光照射在工件。 又介紹著將積分器的入射面的光的入射領域，設計成 比入射面本身的面積還要小，就可將來自光源的放射光以 高利用率照射在工件的技術。 然而，一般放電燈是隨著經過點燈時間使得電極前端 會磨耗，而電極間距離會變長。因此，當初，即使可將來 -5- 201017341 自光源的放射光以高利用率入射在積分器，隨著經過點燈 時間，無法入射於該積分器的光比率也變多，結果，會降 低光的利用率，而在工件的照度也會降低。 尤其是，電極間距離爲數mm的放電燈，或是在封入 多量水銀的放電燈，點燈中的電極溫度成爲極高溫之故， 因而電極損耗是比通常的放電燈還要顯著地發生。又，點 燈電力愈大的放電燈，電極損耗愈容易發生之故，因而容 易發生光利用率的降低或照度降低。 @ 又，即使工件作成大型化，對於該工件的全表面，也 必須給予一定量的照射能量。所以，配合工件的大型化也 必須加大光源的電力（燈電力）。 揭示於上述與專利文獻1的光照射裝置，是爲了加大 光源電力，可考量增加單元數，惟若增加單元數，則因從 光對於積分器的入射角之關係，必須將光源的位置設定在 距積分器有段距離的位置。亦即，會導致光照射裝置的大 型化。 ❹ 綜合以上背景： U)作爲使用於半導體晶圓或液晶基板的曝光裝置的光 源，在使用1個大型燈上有極限。 (b) 如揭示於專利文獻1地，也提案一種排列複數小型 燈而形成1個光源部的方法，惟在此所使用的小型燈，是 因電極損耗厲害，因此隨著經過點燈時間，會降低光的利 用率與降低工件的照度。 (c) 揭示於專利文獻1的光照射裝置的情形，對應於工 -6 - 201017341 件的大型化，若加大光源部的電力（燈電力），則光源部與 積分器的距離會變大。 專利文獻1 :日本特開2004-361746號公報 專利文獻2 :日本特開平1 1 -297268號公報 專利文獻3 :日本特開2000-8232 1號公報 【發明內容】 ® 欲解決該發明的課題，是在利用複數小型燈以構成光 源部的光照射裝置中，提供一種都可解決 (a) 不受點燈時間之影響，可維持高利用率與高照度， (b) 即使增大光源部的電力，也不必加大光源部與積分 器之距離， 的光照射裝置。 本發明的一種光照射裝置，屬於至少具有：排列被封 入有水銀與鹵素的放電燈與反射鏡所構成的複數個光源單 ® 元的光源部，及對於各放電燈供應電力的饋電裝置，及入 射有從光源部所放射的光的積分器。又，上述饋電裝置是 對於上述放電燈供應交流電流，而且從上述光源部所放射 的光對於上述積分器的入射率爲90 %以下，爲其特徵者。 又從上述光源部所放射的光對於上述積分器的入射率 爲50%以上，爲其特徵者。 又，上述放電燈是額定2 00W以上，而電極間距離爲 1.0mm以上，爲其特徵者。 又上述放電燈是封入有〇.〇8〜0.25mg/cc的範圍的水 201017341 銀，爲其特徵者。 又，上述放電燈是封入有5xl(T5〜7χ10·3μιηο1/ιηηι3的 範圍的鹵素，爲其特徵者。 (Α)本發明的光照射裝置是將封入水銀與鹵素的放電燈予 以交流點燈，在該放電燈的電極前端形成突起，而該突起 在燈點燈中不會消耗，利用可維持大到同一距離的性質， 以解決降低照度維持率的問題。 (Β)又，將從光源部所放射的光中，入射於積分器的光的 ❹ 比率，亦即將光的利用率作成90%以下，即使增加電極間 距離，也不受其影響而可達成高照度維持率。 【實施方式】 第1圖是表示本發明的第1實施例的光照射裝置的槪 略構成。 放射光的光源部10是由複數放電燈Ν所構成，各放 電燈Ν是個別地被支撐於共通的支撐體3。各放電燈Ν是 @ 內設燈1與反射鏡2。支撐體3是沿著大約抛物面或是大 約橢圓面的平緩曲面形狀，從各光源單元Ν所放射的光， 重疊在光照射領域的積分器20的入射面的方式，隨著朝 向支撐體的周邊部，徐徐地傾斜光源單元加以支撐。積分 器20是將照度分布作成均勻的光學元件。在第1圖中， 記載著接近光源部10與積分器20，惟實際上，光源部1〇 與積分器20之距離是比圖示還要長，又，形成於支撐體3 的曲面形狀是更平緩。 -8 - 201017341 在積分器20重疊入射有來自複數的光源單元N的光 。從積分器20所出射的光，是利用準直光管21成爲平行 光，被照射在經由被保持在屏蔽平台22的屏蔽23而被保 持在工件平台24上的塗佈有光阻等感光劑的液晶基板或 半導體元件的所謂工件W。在屏蔽23形成有圖案，而該 圖案爲曝光形成於工件W上的感光劑。 在各光源單元N，獨立地連接有將電力供應於各該放 Ο 電燈1的饋電裝置30。又，各饋電裝置30的控制電路是 被連接於未圖示的光照射裝置的裝置控制部，而對於放電 燈的點燈，熄燈，或點燈時對於放電燈的電力供應，是光 源單元N別地被控制。針對於饋電裝置30的構成及動作 及裝置控制部如後述。 第2圖是表示光源單元N的擴大構造。1個的光源單 元N是由放電燈1，反射鏡2，及圍繞此些的收納盒4所 構成。又，放電燈1是利用饋電裝置30被交流點燈。物 ® 理性機構是如後述，惟若交流點燈水銀與鹵素的放電燈， 則在電極前端可形成突起。反射鏡2是圍繞放電燈1的凹 面反射鏡，配置成使得放電燈1的電極軸與反射鏡2的光 軸成爲一致。反射鏡2是使用例如橢圓鏡或拋物鏡。收納 盒4是內設放電燈1與反射鏡2的木屐箱狀著，而在後方 壁或是側壁設有冷卻風用開口。 第3圖是表示燈1的擴大圖。燈1是所謂放電燈，具 有藉由石英玻璃所成的放電容器所形成的槪略球形的發光 部10。在該發光部中形成有發光空間S，而在空間內 -9- 201017341 以1 mm〜2 mm的間隔相對配置有同一電極2。在發光部ίο 的兩端部形成有側管部1 1，而在該側壁部1 1例如利用收 縮密封氣密地埋設有鉬所成的導電用金屬箔3。在金屬箔 3的一端接合有電極2的軸部21，又，在金屬箔3的另一 端接合有外部引線4而從外部的饋電裝置進行饋定。在發 光部10封入有水銀與稀有氣體及鹵素氣體。水銀是爲了 得到所必需的紫外光波長，例如得到波長300〜3 60nm的 放射光者，封入〇.〇8〜0.25 mg/mm3。該封入量是藉由溫 _ 度條件也不相同，惟點燈時成爲80氣壓以上的高蒸氣壓 力。 稀有氣體是例如氬氣體封入大約13 kPa。其功能是在 於改善點燈始動性。鹵素是碘、溴、氯等與水銀或其他金 屬以化合物的形態被封入。鹵素的封入量是從5x1 (Γ5〜7x 1(Γ3μιη〇1/πιιη3的範圍被選擇。鹵素的功能是利用所謂鹵素 循環的長壽命化，惟如本發明的放電燈地極小型又極高點 燈蒸氣壓者，也有防止放電容器的透明消失的作用。表示 © 燈的數値例，例如發光部1〇的最大外徑9.5mm，電極間 距離1.5mm，發光管內容積7 5mm3，額定電壓70V，額定 電力200W，而以350Hz被交流點燈。 電極2的前端（相對於另一方的電極的端部）是隨著燈 的點燈，形成有突起。形成有突起的現象並不一定明顯， 惟如下地被推測。亦即，在燈點燈中從電極前端附近的高 溫部所蒸發的鎢（電極的構成材料），是與存在於發光管內 的鹵素成殘存氧氣結合，例如鹵素爲Br，則存在作爲 -10- 201017341 WBr、WBr2、WO、W02、W02Br、W02Br2 等的鎢化合物 。此些化合物是在電極前端附近的氣相中的高溫部被分解 而成爲鎢原子或陽離子。溫度擴散（氣相中的高溫部=自電 弧中，而低溫部=朝電極前端近旁的鎢原子的擴散），及在 電弧中鎢原子被電離而成爲陽離子，當陰極動作時藉由電 場朝著陰極方向引拉（漂移），藉此，電極前端附近的氣相 中的鎢蒸氣密度變高，而析出在電極前端，會形成突起。 # 第4圖是表示電極前端及突起的模式圖。電極2是由 球部2a與軸部21所構成，而在球部2a的前端形成有突 起2b。該突起2b是即使在開始燈的點燈時未存在的情形 ，利用其後的點燈，可以說也自然發生地被形成。在此， 突起2b是任何放電燈並不定也發生。在電極間距離爲 1mm〜2mm，將0.08 mg/mm3以上的水銀，及稀有氣體， 及5χ1(Γ5〜7χ1(Γ3μιηο1/ηιπι3的範圍的鹵素封入於發光部的 短弧型放電燈，隨著燈點燈，形成有突起2b，而在突起 ® 2b彼此間之間形成的電弧。 如此地，本案發明是在封入水銀與鹵素的燈中，利用 進行交流點燈，突起可形成在電極前端的技術，藉此相當 解決電極間距離的增大，與隨著該增大的照度維持率的降 低的問題。又，將點燈頻率以周期性地低頻進行點燈，則 更確實地可維持電極間距離。例如在350Hz的點燈中，周 期性地以40Hz進行點燈。 以下，放電燈的點燈形態與電極間距離，針對於照度 維持率所受的影響進行實驗。 -11 - 201017341 實驗是使用交流點燈型燈的電極間距離1.6mm(燈1) ，交流點燈型燈的電極間距離1.4mm(燈2)，交流點燈型 發的電極間距離1.2mm(燈3)，交流點燈型燈的電極間距 離1.0mm(燈4)，直流點燈型燈的電極間距離l.〇mm(燈5) ，直流點燈型燈的電極間距離〇.7mm(燈6)的6種類的'燈 。6種類的燈的電極間距離都表示點燈前的大小’而點燈 形態與電極間距離以外的條件，是基本上作成相同。實驗 是製作如第2圖所示的形態的6種類的具200W反射器的 參 燈，而對表示於第1圖的光源部總數53燈的裝置調查照 度維持率。交流點燈燈都以350Hz進行點燈。照度是使用 日本牛尾電機製UIT250照度計與S365受光器，求出在工 件面的照度維持率。照度維持率是隨著點燈時間的經過進 行測定照度，表示作爲對於點燈初期的照度的相對値。尤 其是，將從開始點燈經過75 0小時後的照度維持率抓住作 爲業者的指標，而從照度維持率的觀點上以表示75%以上 的樣品作爲合格。 ® 第5圖是表示實驗結果。縱軸是表示照度維持率（％) ，而橫軸是表示點燈時間（時間）。由圖可說明如下。 (1) 進行交流點燈的放電燈（燈1〜燈4)，是與進行直流點 燈的放電燈（燈5與燈6)相比較，照度維持率格外地優異 。該原因可能爲上述的突起成長在交流點燈的放電燈中有 良好的功能者。 (2) 即使進行交流點燈的放電燈，電極間距離愈大的燈照度 維持率也優異’具體上，燈4(積分器l.〇mm)是在點燈750 -12- 201017341 小時爲照度維持率75%，而在點燈1 500小時降低至60% ’對此，燈1(積分器1.6mm)是在750小時爲照度維持率 90%以上，而在點燈1 500小時也有接近90%的照度維持率 。該原因可能是電極間距離愈小，電極損耗愈厲害。 該結果，200W的交流點燈型放電燈的情形，則可知 若電極間距離爲1 .0mm以上，就可發揮業者可認定的照度 維持率。 Ο 以下，本發明人等，以200W以外的燈電力進行與上 述同樣的實驗。具體上，以200W的交流點燈型放電燈， 300W的交流點燈型放電燈，420W的交流點型放電燈作爲 對象。 針對於各燈，求出具有與上述同樣的照度維持率（在 點燈750小時維持75%的照度）的燈的電極間距離，則在 燈電力爲250W時是電極間距離1.1mm以上，在燈電力爲 300W時是電極間距離1.2mm以上，而在燈電力爲420W Φ 時是電極間距離1.4mm以上。 該結果 (3)照度維持率爲滿足業者水準所用的電極間距離，是藉由 燈的額定電力（燈電力）有所不同。具體上200W時是電極 間距離1.0mm以上，250W時是電極間距離l.lmm以上， 300W時是電極間距離1.2mm以上，而420W時是電極間 距離1 _ 4mm以上的燈，燈電力愈大，則滿足要求的電極間 距離的愈大。 又，電極間距離是與光的利用率（從光源部所放射的 -13- 201017341 光中被入射於積分器的比率）。因爲，電極間距離愈小的 燈，可將放電弧實質視作爲點之故，因而能將放電弧的光 1 0 0%地入射於積分器。一方面，電極間距離大的燈是將放 電弧視作爲有張大小之故，因而，除非使用具有需要以上 的入射面的積分器，就不能將放電弧的光100%地入射於 積分器，會發生無法射入於積分器的光，亦即會發生浪費 的光。 如此，對於表示於第1圖的光照射裝置，依序組裝點 @ 燈電力與電極間距離不同的複數燈，嘗試測定入射於積分 器的光。 具體上，針對於使用200W電極間距離1.0mm的燈的 裝置（裝置A)，使用250W電極間距離1.1 mm的燈的裝置（ 裝置B)，使用300W電極間距離1.2mm的燈的裝置（裝置 C)，及使用420W電極間距離1.4mm的燈的裝置（裝置D) ，進行測定光的利用率。 又，各裝置是調整燈（單元）數成爲曝光面的照度會相 @ 等。具體上，裝置A是單元數53個，總電力10.7KW，裝 置B是單元數42個，總電力10.6KW，裝置C是單元數 36個，總電力10.8KW，而裝置D是單元數25個，總電 力10.5KW，曝光面照度是被統一成45mW/cm2。 又，測定該實驗的各裝置的光利用率，則裝置A是成 爲89.9%，裝置B是成爲88.7%，裝置C是成爲88.0%， 而裝置D是成爲89.3%。 該結果，可知以下事項。 -14- 201017341 (4)可知照度維持率充分的光照射裝置是不管燈電力，光利 用率在90%以下。換言之，只要光利用率爲90%以下，則 不管燈電力具充分的照度維持率。在此，具充分的照度維 持率是指滿足業者基準値者。 在此，說明光利用率的測定方法。 第6圖是說明光利用率的測定方法的圖式，（a)是表示 燈與積分器的配置關係的圖式，（b)是表示積分器的入射面 Φ 的圖式，（c)是表示積分器的入射面的直徑方向的照度分布 〇 在（a)中，隔著所定距離（例如2600mm)設置燈與積分 器。該距離是在組裝表示於第1圖的裝置時所設定的數値 ，實際上最適値是藉由燈的裝載支數有所變化。被照射在 積分器的入射面的光是如（b)所示地，也有照射積分器的入 射面的成分。惟也存在未被照射在入射面的成分。在積分 器的入射面，以比入射面還要應的領域作爲對象而朝積分 ® 器的直徑方向移動受光器的位置。具體上，配置機器人（ 朝XY方向運轉的機構），例如以20mm間隔離行移動照度 計的受光器，而在X，Y的各點測定照度（mW/cm2)。又， 受光器的移動是手動也可以。相乘所測定的照度値與其位 置的面積（環狀部分）的面積（cm2)，則求得光束（W)。如此 地，在積分器的入射面全體全面地進行照度測定，就可求 出燈的全光束。又，入射於積分器的領域是入射面的光束 (積分器入射光束）之故，因而若計算積分入射光束/全光束 ，就可求出在積分器有幾％的光束被入射。具體上，以 -15- 201017341 2 0 mm間隔在左右上下各17點合計33點（中心是有掉1次 )進行測定Φ 720的圓內，作爲全光束。 又，光利用率是在90%以下就可達成超過業者基準的 照度維持率，而利用率愈小，照度維持率愈優異。光利用 率愈小，就是電極間距離會變大，而不容易發生電極損耗 。順便地，在表示於第5圖的200W的燈中，將交流點燈 型燈（燈1〜燈4)作爲光源入射於表示於第1圖的光照射裝 置，而針對於各該裝置，進行測定光利用率。又，與上述 同樣地，各裝置是調整燈（單元）數成爲曝光面的照度幾乎 相等。 其結果，光利用率是燈1爲58%，燈2爲67%，燈3 爲77%，燈4爲90%，燈5爲90%，燈6爲100%。 亦即，由該實驗，照度維持率充分的光照射裝置是不 管燈電力，光的利用率爲90%以下。但是若光利用率低於 5 0%，則未利用的光變多，而在與接通電力的關係不理想 。因此光利用率是50%以上90%以下較佳。 如此地，本案發明是一面重複各種試製或實驗一面找 出以下事實。 (1) 交流點燈放電燈與直流點燈的情形相比較，可發揮 各外地優異的照度維持率。 (2) 電極間距離與照度維持率是具有正的相關係，電極 間距離愈大的放電燈，照度維持率愈優異。 (3) 滿足作爲業者水準的照度維持率（所定的照度維持 率）的最小電極間距離，是藉由燈電力有所不同。燈電力 -16- 201017341 愈大，則滿足所定的照度維持率的電極間距離是愈大。 (4)所定的照度維持率是不管燈電力，與光利用率有關 聯。若光利用率爲90%以下，則不管燈電力，照度維持率 是可滿足業者水準。 然而，工件的液晶基板或半導體晶圓是近年來大型化 ，在液晶基板來說，也有在畫面對角超適40英吋者。曝 光此種大型的液晶基板時，即使曝光面積變大，每一單位 〇 面積的照射能量是成爲與從前需要同樣量。被塗佈於液晶 基板上的光阻感光所用的能量爲與曝光面積無關地一定。 亦即，若曝光面積變大，則僅其分量必須提高從光源部所 發生的照射能量。例如，爲了在曝光面欲得到與使用1燈 5kW的高壓水銀燈的光照射裝置同等的特性，則成爲以電 力換算若爲100W的放電燈，則需要50支（50台光源單元 )而光源部是成爲更大型化。 一方面，在積分器（積分透鏡），存在著光入射於積分 ® 器內部所用的入射角，而從該角度以外的角度所入射的光 ，是在積分器表面會被反射，而不會進入內部。亦即，光 源單元爲成爲50台會導致光源部大型化，則會發生無法 入射於積分器的光。 所以使用複數燈來構成光源部時，一面調整燈數（單 元數）與積分器之距離，一面可將來自光源部的放射光良 好地入射於積分器，且在曝光面必須設定僅可提供充分的 照射能量的構造。201017341 6. Description of the Invention: TECHNICAL FIELD The present invention relates to a light irradiation device. In particular, the present invention relates to a light irradiation device for use in an exposure apparatus for manufacturing a semiconductor device or a liquid crystal display substrate. [Prior Art] Conventionally, in a light source used for an exposure apparatus of a semiconductor wafer or a liquid crystal substrate, an ultrahigh pressure mercury lamp of several kW to several tens of kW is used. This lamp has been used for a long time and has been used in the past. However, recently, with the large area of the workpiece, for example, a large lamp of 25 kW is used for a liquid crystal exposure lamp. However, the enlargement of the lamp is directly related to the enlargement of the bulb (foaming tube) or the electrode material constituting the lamp, and the manufacturing cost is greatly increased in association with the number of fabrications, and thus gradually approaches the limit. On the other hand, as a countermeasure for the enlargement of the lamp, for example, Japanese Laid-Open Patent Publication No. 20 04-3 61 746 (Patent Document 1) proposes a configuration in which a plurality of small lamps are arranged instead of one large lamp. In the light irradiation device of Patent Document 1, for example, 35 units including a discharge lamp and a mirror are arranged to constitute one light source unit, and light emitted from the light source unit is irradiated onto the workpiece by using an integrator. Further, a technique in which the incident field of light on the incident surface of the integrator is designed to be smaller than the area of the incident surface itself, and the emitted light from the light source is irradiated to the workpiece at a high utilization rate. However, in general, the discharge lamp is worn at the front end of the electrode as the lighting time elapses, and the distance between the electrodes becomes longer. Therefore, at the beginning, even if the radiation from the light source can be incident on the integrator at a high utilization rate in the future -5 - 201017341, the ratio of light that cannot be incident on the integrator increases as the lighting time passes, and as a result, the light is lowered. The utilization rate of the workpiece will also decrease. In particular, a discharge lamp having a distance between electrodes of several mm or a discharge lamp in which a large amount of mercury is sealed, the temperature of the electrode in the lighting is extremely high, and thus the electrode loss is more remarkable than that of a normal discharge lamp. Further, the discharge lamp having a higher lighting power has an electrode loss which is more likely to occur, so that the light utilization rate is lowered or the illuminance is lowered. @ Also, even if the workpiece is made larger, a certain amount of irradiation energy must be given to the entire surface of the workpiece. Therefore, in order to increase the size of the workpiece, it is necessary to increase the power of the light source (lamp power). The light irradiation device disclosed in Patent Document 1 described above is intended to increase the number of cells in order to increase the power of the light source. However, if the number of cells is increased, the position of the light source must be set due to the relationship between the incident angle of the light and the integrator. At a distance from the integrator. That is, it causes a large size of the light irradiation device. ❹ Combining the above background: U) As a light source for an exposure device used for a semiconductor wafer or a liquid crystal substrate, there is a limit in using one large lamp. (b) As disclosed in Patent Document 1, a method of forming a single light source unit by arranging a plurality of small lamps is also proposed. However, since the small lamp used here is extremely worn out due to the lighting time, It will reduce the utilization of light and reduce the illumination of the workpiece. (c) In the case of the light irradiation device disclosed in Patent Document 1, the distance between the light source unit and the integrator becomes larger as the electric power of the light source unit (lamp power) is increased in accordance with the enlargement of the -6 - 201017341 piece. . Patent Document 1: Japanese Laid-Open Patent Publication No. Hei. No. Hei. No. Hei. No. Hei. No. Hei. No. Hei. In the light irradiation device that uses a plurality of small lamps to constitute the light source portion, it is possible to provide (a) unaffected by the lighting time, maintaining high utilization and high illumination, and (b) even increasing the light source portion. The electric light does not need to increase the distance between the light source unit and the integrator, and the light irradiation device. A light irradiation device according to the present invention is directed to a light source unit having at least a plurality of light source cells constituting a discharge lamp and a mirror in which mercury and a halogen are sealed, and a power feeding device that supplies electric power to each of the discharge lamps. And an integrator that is incident on the light emitted from the light source unit. Further, the power feeding device is characterized in that an alternating current is supplied to the discharge lamp, and an incident rate of light emitted from the light source unit with respect to the integrator is 90% or less. Further, the light emitted from the light source unit is characterized in that the incident rate of the integrator is 50% or more. Further, the discharge lamp is characterized in that it is rated at 200 W or more and the distance between electrodes is 1.0 mm or more. Further, the above-mentioned discharge lamp is characterized by being filled with water of 201017341 in the range of 〜.〇8 to 0.25 mg/cc. Further, the discharge lamp is characterized in that a halogen of a range of 5x1 (T5 to 7χ10·3μιηο/ιηηι3) is enclosed. (Α) The light irradiation device of the present invention illuminates a discharge lamp in which mercury and a halogen are sealed, A protrusion is formed at the tip end of the electrode of the discharge lamp, and the protrusion is not consumed in the lamp lighting, and the problem of reducing the illuminance maintenance rate is solved by utilizing the property of maintaining the same distance. (Β) Further, the light source unit is removed. Among the emitted light, the ratio of the light incident on the integrator is 90% or less, that is, even if the distance between the electrodes is increased, the high illuminance maintenance ratio can be achieved without affecting the distance between the electrodes. Fig. 1 is a schematic view showing a configuration of a light irradiation device according to a first embodiment of the present invention. The light source unit 10 for emitting light is composed of a plurality of discharge lamps, and each of the discharge lamps is individually supported by a common support. 3. Each discharge lamp @ is a built-in lamp 1 and a mirror 2. The support body 3 is a gentle curved shape along an approximately parabolic surface or an approximately elliptical surface, and the light radiated from each light source unit is superposed on the light irradiation field. of The form of the incident surface of the integrator 20 is gradually tilted toward the peripheral portion of the support body to support the light source unit. The integrator 20 is an optical element that makes the illuminance distribution uniform. In the first figure, the approaching light source unit is described. 10 and the integrator 20, but actually, the distance between the light source part 1A and the integrator 20 is longer than the illustration, and the curved shape formed on the support body 3 is more gentle. -8 - 201017341 In the integrator 20 The light from the plurality of light source units N is incident on the overlapping light. The light emitted from the integrator 20 is parallel light by the collimator light pipe 21, and is irradiated on the workpiece via the shield 23 held by the shield platform 22. A so-called workpiece W of a liquid crystal substrate or a semiconductor element coated with a photosensitive agent such as a photoresist on the stage 24. A pattern is formed on the shield 23, and the pattern is a photosensitive agent exposed on the workpiece W. In each of the light source units N, The power feeding device 30 that supplies electric power to each of the discharge lamps 1 is independently connected. Further, the control circuit of each power feeding device 30 is connected to a device control unit of a light irradiation device (not shown), and for the discharge lamp The power supply to the discharge lamp at the time of lighting, turning off, or lighting is controlled by the light source unit N. The configuration and operation of the power feeding device 30 and the device control unit will be described later. Fig. 2 is a view showing the light source unit N. The enlarged light source unit N is composed of a discharge lamp 1, a mirror 2, and a storage case 4 surrounding the same. Further, the discharge lamp 1 is electrically lit by the power feeding device 30. The mechanism is as described later, but if a discharge lamp that lights mercury and halogen is exchanged, a protrusion can be formed at the tip end of the electrode. The mirror 2 is a concave mirror surrounding the discharge lamp 1, and is arranged such that the electrode axis and the mirror of the discharge lamp 1 are arranged. The optical axis of 2 is identical. The mirror 2 is, for example, an elliptical mirror or a parabolic mirror. The storage case 4 is in the form of a wooden box in which the discharge lamp 1 and the mirror 2 are housed, and is provided with cooling air for the rear wall or the side wall. Opening. Fig. 3 is an enlarged view showing the lamp 1. The lamp 1 is a so-called discharge lamp having a substantially spherical light-emitting portion 10 formed by a discharge vessel made of quartz glass. The light-emitting space S is formed in the light-emitting portion, and the same electrode 2 is disposed to face each other at intervals of 1 mm to 2 mm in the space -9-201017341. The side tube portion 1 is formed at both end portions of the light-emitting portion ίο, and the conductive metal foil 3 made of molybdenum is hermetically embedded in the side wall portion 1 1 by, for example, a shrink seal. The shaft portion 21 of the electrode 2 is joined to one end of the metal foil 3, and the outer lead 4 is joined to the other end of the metal foil 3 to be fed from an external power feeding device. Mercury, a rare gas, and a halogen gas are sealed in the light emitting unit 10. Mercury is used to obtain the necessary wavelength of ultraviolet light, for example, to obtain a radiation having a wavelength of 300 to 3 60 nm, and is enclosed in 〇.〇8 to 0.25 mg/mm3. The amount of the seal is different depending on the temperature, but it is a high vapor pressure of 80 or more at the time of lighting. The rare gas is, for example, argon gas sealed at about 13 kPa. Its function is to improve the start-up of lighting. Halogen is iodine, bromine, chlorine, etc., and is encapsulated in the form of a compound with mercury or other metals. The halogen encapsulation amount is selected from the range of 5x1 (Γ5 to 7x 1 (Γ3μιη〇1/πιιη3. The function of the halogen is to extend the life of the so-called halogen cycle, but the discharge lamp of the present invention is extremely small and extremely high. The lamp vapor pressure also has the effect of preventing the transparency of the discharge vessel from disappearing. It means that the number of lamps is as follows, for example, the maximum outer diameter of the light-emitting portion 1〇 is 9.5 mm, the distance between the electrodes is 1.5 mm, and the inner volume of the light-emitting tube is 7 5 mm 3 , the rated voltage 70V, rated power 200W, and AC lighting at 350Hz. The front end of the electrode 2 (relative to the end of the other electrode) is formed with the lamp lit up. The phenomenon of forming the protrusion is not necessarily obvious. However, it is estimated that tungsten (the constituent material of the electrode) evaporated from the high temperature portion near the tip end of the electrode in the lamp lighting is combined with the residual oxygen present in the halogen in the arc tube, for example, halogen is Br, there are tungsten compounds such as -10-201017341 WBr, WBr2, WO, W02, W02Br, W02Br2, etc. These compounds are decomposed into tungsten atoms or cations at a high temperature portion in the gas phase near the front end of the electrode. Temperature diffusion (high temperature in the gas phase = from the arc, and low temperature = diffusion of tungsten atoms near the front end of the electrode), and tungsten atoms in the arc are ionized to become cations, when the cathode is operated by the electric field When the film is pulled (drifted) toward the cathode, the density of tungsten vapor in the gas phase near the tip end of the electrode is increased, and the protrusion is formed at the tip of the electrode. #图图图图图图图图图The electrode 2 is composed of the ball portion 2a and the shaft portion 21, and a projection 2b is formed at the tip end of the ball portion 2a. The protrusion 2b is a case where the light is not present even when the lighting of the lamp is started, and the lighting is utilized thereafter. It can be said that it is naturally formed. Here, the protrusion 2b is not fixed in any of the discharge lamps. The distance between the electrodes is 1 mm to 2 mm, mercury of 0.08 mg/mm3 or more, and rare gas, and 5χ1 (Γ5) A short arc type discharge lamp in which a halogen in the range of χ3μιηο1/ηιπι3 is enclosed in the light-emitting portion, and an arc formed between the protrusions 2b and the protrusions 2b is formed as the lamps are lit. Thus, the present invention Is sealing In the mercury and halogen lamps, the technique of performing the alternating current lighting and the protrusions can be formed at the front end of the electrode, thereby considerably solving the problem of an increase in the distance between the electrodes and a decrease in the illuminance maintenance rate with the increase. When the lighting frequency is periodically turned on at a low frequency, the distance between the electrodes can be more reliably maintained. For example, in a 350 Hz lighting, the lighting is periodically performed at 40 Hz. Hereinafter, the lighting pattern and the electrode of the discharge lamp The distance is measured for the influence of the illuminance maintenance rate. -11 - 201017341 The experiment is to use the distance between the electrodes of the AC lighting type lamp of 1.6 mm (lamp 1), and the distance between the electrodes of the AC lighting type lamp is 1.4 mm ( Lamp 2), the distance between the electrodes of the AC lighting type is 1.2 mm (light 3), the distance between the electrodes of the alternating current lighting type lamp is 1.0 mm (light 4), and the distance between the electrodes of the direct current lighting type lamp is 1. 〇mm ( Lamp 5), 6 types of 'lights' between the electrodes of the DC lighting type lamp 〇.7mm (light 6). The distance between the electrodes of the six types of lamps indicates the size before lighting, and the conditions other than the lighting pattern and the distance between the electrodes are basically the same. The experiment was to produce six types of lamps having a 200 W reflector as shown in Fig. 2, and to investigate the illuminance maintenance rate for the device having the total number of lamps of the light source unit shown in Fig. 1 . The AC lighting lights are illuminated at 350 Hz. The illuminance was obtained by using the Japanese oxtail mechanism UIT250 illuminometer and the S365 receiver, and the illuminance maintenance rate on the workpiece surface was obtained. The illuminance maintenance rate is an illuminance measured as the lighting time elapses, and indicates a relative illuminance as an illuminance at the initial stage of lighting. In particular, the illuminance maintenance rate after 75 hours from the start of lighting is taken as an indicator of the manufacturer, and from the viewpoint of the illuminance maintenance rate, a sample indicating 75% or more is qualified. ® Figure 5 shows the experimental results. The vertical axis represents the illuminance maintenance rate (%), and the horizontal axis represents the lighting time (time). The figure can be explained as follows. (1) The discharge lamp (lamp 1 to lamp 4) for AC lighting is particularly excellent in the illuminance maintenance rate compared with the discharge lamp (lamp 5 and lamp 6) that performs DC lighting. The reason may be that the above-mentioned protrusion grows well in a discharge lamp of an AC lighting. (2) Even if a discharge lamp for AC lighting is used, the lamp illumination maintenance rate is higher as the distance between the electrodes is larger. Specifically, the lamp 4 (integrator l.〇mm) is illuminated at 750 -12 to 201017341 hours. The maintenance rate is 75%, and it is reduced to 60% when lighting for 1,500 hours. For this, the lamp 1 (integrator 1.6mm) has an illuminance maintenance rate of 90% or more at 750 hours, and is close to 90 at 1,500 hours of lighting. % illuminance maintenance rate. The reason may be that the smaller the distance between the electrodes, the more the electrode loss is. As a result, in the case of the 200 W AC lighting type discharge lamp, it is understood that the illuminance maintenance rate which can be recognized by the operator can be exhibited when the distance between the electrodes is 1.0 mm or more. In the following, the inventors of the present invention conducted the same experiment as described above with lamp power other than 200 W. Specifically, a 200W AC lighting type discharge lamp, a 300W AC lighting type discharge lamp, and a 420W AC point type discharge lamp are targeted. For each of the lamps, the distance between the electrodes having the illuminance maintenance ratio (the illuminance at which 75% is maintained for 750 hours) is obtained. When the lamp power is 250 W, the distance between the electrodes is 1.1 mm or more. When the lamp power is 300 W, the distance between the electrodes is 1.2 mm or more, and when the lamp power is 420 W Φ, the distance between the electrodes is 1.4 mm or more. The result (3) The illuminance maintenance rate is the distance between the electrodes used to satisfy the level of the operator, and is determined by the rated power (lamp power) of the lamp. Specifically, the distance between electrodes is 1.0 mm or more at 200 W, the distance between electrodes is more than 1.1 mm at 250 W, the distance between electrodes is 1.2 mm or more at 300 W, and the distance between electrodes is 1 _ 4 mm or more at 420 W. Larger, the greater the distance between the electrodes that meets the requirements. Further, the distance between the electrodes is the ratio of the light source (the ratio of the light incident on the integrator from -13 to 201017341 emitted from the light source unit). Since the lamp having a smaller distance between the electrodes can be regarded as a point of the discharge arc, the arc discharge light can be incident on the integrator at 100%. On the one hand, a lamp having a large distance between electrodes has a discharge arc as a sheet size. Therefore, unless an integrator having an incident surface or more is used, the arc-dissipating light cannot be incident on the integrator 100%. Light that cannot be incident on the integrator occurs, that is, wasted light. As described above, in the light irradiation device shown in Fig. 1, a plurality of lamps having different distances between the lamp power and the electrodes are sequentially assembled, and it is attempted to measure the light incident on the integrator. Specifically, for a device (device A) using a lamp having a distance of 1.0 W between electrodes of 200 W, a device (device B) using a lamp having a distance of 1.1 mm between electrodes of 250 W, and a device using a lamp having a distance of 1.2 mm between electrodes of 300 W (device) C), and a device (device D) using a lamp having a distance of 420 W between the electrodes of 1.4 mm, the utilization of the measurement light was performed. Further, in each device, the number of adjustment lamps (units) becomes the illuminance of the exposure surface, and the like. Specifically, the device A is 53 units, the total power is 10.7 KW, the device B is 42 units, the total power is 10.6 KW, the device C is 36 units, the total power is 10.8 KW, and the device D is the number of units 25 The total power is 10.5 KW, and the exposure surface illumination is unified to 45 mW/cm2. Further, when the light utilization efficiency of each device in the experiment was measured, the device A was 89.9%, the device B was 88.7%, the device C was 88.0%, and the device D was 89.3%. The result is as follows. -14- 201017341 (4) It is known that the light irradiation device having a sufficient illuminance maintenance rate is 90% or less regardless of the lamp power. In other words, as long as the light utilization rate is 90% or less, the lamp power has a sufficient illuminance maintenance rate. Here, sufficient illuminance maintenance rate refers to meeting the industry benchmark. Here, a method of measuring the light utilization rate will be described. Fig. 6 is a view for explaining a method of measuring the light utilization rate, wherein (a) is a diagram showing an arrangement relationship between a lamp and an integrator, (b) is a diagram showing an incident surface Φ of the integrator, and (c) is a The illuminance distribution in the radial direction of the incident surface of the integrator is shown in (a), and the lamp and the integrator are placed at a predetermined distance (for example, 2600 mm). This distance is the number set when the device shown in Fig. 1 is assembled. Actually, the optimum number is changed by the number of loadings of the lamp. The light that is incident on the incident surface of the integrator is also a component that illuminates the incident surface of the integrator as shown in (b). However, there are also components that are not irradiated on the incident surface. On the incident surface of the integrator, the position of the photoreceptor is moved toward the diameter of the integrator in the field corresponding to the incident surface. Specifically, the robot (mechanism that operates in the XY direction) is placed, for example, a light receiver that moves the illuminance meter between 20 mm, and the illuminance (mW/cm2) is measured at each of X and Y. Also, the movement of the light receiver is manual. The beam (W) is obtained by multiplying the measured illuminance 値 and the area (cm 2 ) of the area (annular portion) of the position. In this way, the entire illuminance is measured on the entire entrance surface of the integrator, and the full beam of the lamp can be obtained. Further, the field of incidence on the integrator is the beam incident on the incident surface (integrator incident beam). Therefore, if the integrated incident beam/full beam is calculated, it can be determined that a few percent of the beam is incident on the integrator. Specifically, in the circle of Φ 720, the total light beam is measured by a total of 33 points (one at the center) at a distance of -15-201017341 2 0 mm. Further, when the light utilization rate is 90% or less, the illuminance maintenance rate exceeding the operator's standard can be achieved, and the smaller the utilization rate, the more excellent the illuminance maintenance rate. The smaller the light utilization rate, the larger the distance between the electrodes is, and the electrode loss is less likely to occur. By the way, in the 200W lamp shown in FIG. 5, the AC lighting type lamp (lamp 1 to lamp 4) is incident on the light irradiation device shown in Fig. 1 as a light source, and is performed for each of the devices. The light utilization rate was measured. Further, in the same manner as described above, the illuminance of each device is such that the number of adjustment lamps (cells) becomes the exposure surface is almost equal. As a result, the light utilization rate was 58% for the lamp 1, 67% for the lamp 2, 77% for the lamp 3, 90% for the lamp 4, 90% for the lamp 5, and 100% for the lamp 6. In other words, in this experiment, the light irradiation device having a sufficient illuminance maintenance rate does not require lamp power, and the light utilization rate is 90% or less. However, if the light utilization rate is less than 50%, the amount of unused light is increased, and the relationship with the power-on is not satisfactory. Therefore, the light utilization rate is preferably 50% or more and 90% or less. Thus, the invention of the present invention finds the following facts while repeating various trial productions or experiments. (1) Compared with the case of DC lighting, the AC lighting discharge lamp can provide excellent illuminance maintenance ratio in each field. (2) The distance between the electrodes and the illuminance maintenance rate are positively correlated. The discharge lamp with the larger distance between the electrodes has an excellent illuminance maintenance rate. (3) The minimum inter-electrode distance that satisfies the illuminance maintenance rate (determined illuminance maintenance rate) at the operator level is different depending on the lamp power. The larger the lamp power -16- 201017341, the larger the distance between the electrodes that satisfies the specified illuminance maintenance rate. (4) The specified illuminance maintenance rate is related to the light utilization rate regardless of the lamp power. If the light utilization rate is 90% or less, the illuminance maintenance rate can satisfy the industry level regardless of the lamp power. However, the liquid crystal substrate or the semiconductor wafer of the workpiece has been enlarged in recent years, and in the liquid crystal substrate, there is also a case where the screen diagonally exceeds 40 inches. When such a large liquid crystal substrate is exposed, even if the exposure area is increased, the irradiation energy per unit area is the same as that required in the past. The energy used for photoresist resistion applied to the liquid crystal substrate is constant regardless of the exposure area. In other words, if the exposure area is increased, only the component must increase the amount of irradiation energy generated from the light source unit. For example, in order to obtain a discharge lamp having a performance equivalent to a light irradiation device using a high-pressure mercury lamp of 1 kW and 5 kW, a discharge lamp having a power of 100 W is required, and 50 (50 light source units) are required, and the light source unit is Become bigger. On the one hand, in the integrator (integral lens), there is an incident angle at which light is incident on the inside of the integrator, and light incident from an angle other than the angle is reflected on the surface of the integrator without entering internal. In other words, when the number of the light source units is 50, the size of the light source unit is increased, and light that cannot be incident on the integrator occurs. Therefore, when the light source unit is configured by using a plurality of lamps, the distance between the number of lamps (the number of cells) and the integrator can be adjusted, and the emitted light from the light source unit can be incident on the integrator satisfactorily, and the exposure surface must be set only to provide sufficient The construction of the illuminating energy.
具體而言，在傳統的一般性的光照射裝置（1燈5k W -17- 201017341 的高壓水銀燈’照射領域500mmx600mm’視角 n)中， 曝光面所必須的照射能量是45mW/cm2 °將此置換成以如 本案發明的複數放電燈構成光源部的光照射裝置，則以 100W的放電燈就使用61個，對於同一照射領域，以同一 視覺可構成同一照射能量。這時候’ 6 1個的放電燈是在總 電力成爲6.1kW。 —方面，以200 W的放電燈則使用18個，對於同一 照射領域，以同一視覺可構成同一照射能量。這時候18 φ 個的放電燈是在總電力成爲3.6kW，而與使用l〇〇W的燈 的情形相比較，提昇約59 %的電力效率，亦即可知使用 200W的放電燈構成光源部者，比使用100W的放電燈構 成光源部者，燈數變少，而電力效率變高。 同樣地，針對於傳統的一般性的光照射裝置（1燈 10kW的高壓水銀燈，照射領域750mmx650mm，視角2.0° )’若置換成本案發明的光照射裝置，則以1 00W的放電燈 就用94個（總電力9.4k W)，對此，若200W的放電燈就用 參 29個（總電力5.8kW)，與上述同樣地，可知使用200W的 放電燈以構成光源部者，電力效率變高。 在此’本發明的放電燈是封入有〇.〇8〜0.25 mg/mm3 範圍的水銀。 第7圖是表示〇·ΐ5 mg/mm3的水銀，及封入鹵素的放 電燈的分光分布。如圖示地，可知多放射波長300〜 3 50nm的紫外線。若水銀是比〇·〇8 mg/mm3還要少，則除 了波長300〜35〇nm的發光以外，也增加依3 00nm以下水 -18- 201017341 銀所致的發光而在曝光上有不好影響之故，因而較不理想 。又，350〜45 Onm附近的連續光譜也降低之故，因而較 不理想。若水銀比 0.25 mg/mm3還要多，則波長 3 00〜 3 5 Onm的發光變少較不理想。 第8圖是表示點燈放電燈的饋電裝置。 饋電裝置3是由：供應有直流電壓的降壓斬波電路31 ，及連接於降壓斬波電路31的輸出側，將直流電壓變更 ❿ 成交流電壓而供應於放電燈1的全電橋型反相電路32(以 下，也稱爲「全電橋電路」），及串聯連接於放電燈的線 圈11，電容器C1，起動電路33，及控制電路34所構成 又，藉由降壓斬波電路31，全電橋電路32，起動電 路33，控制電路34構成饋電裝置，包括放電燈1稱爲點 燈裝置。 降壓斬波電路31是由：連接於直流電源VDC，開關 ® 元件Qx，及二極體Dx，及線圈Lx，及平滑電容器Cx， 及開關元件Qx的驅動電路Gx所構成。開關元件Qx是藉 由開/關驅動。藉由該驅動，被調整開關元件Qx的作用比 ，而控制被供應於放電燈10的電流或電力。 全電橋電路32是由：電橋狀地連接的電晶體或FET 的開關元件Q1〜Q4，及開關元件Q1〜Q4的驅動電路G1 〜G4所構成。又，在開關元件Q1〜Q4，分別並聯地二極 體爲也逆並聯地連接的情形，惟該實施例中省略二極體。 控制電路34是由：電力轉換器340，比較器341，脈 -19- 201017341 寬調變電路342，控制部343，全電橋電路驅動電路344 所構成。電力轉換器340是將在電阻Rl、R2、R3所檢測 的電壓訊號或電流訊號轉換成電力訊號。電力訊號是在比 較器341與基準電力値相比較且經由脈寬調變電路342， 進行反饋控制開關元件Qx。藉此，將燈的點燈電力作成 —定値。實施所謂定電力控制。又，開關元件Q 1〜Q4是 經由控制部3 43，利用全電橋電路驅動電路3 44被驅動。 全電橋電路32的動作是交互地重複開，閉開關元件 參 Q1、Q4，及開關元件Q1、Q4，及開關元件Q2、Q3。當 開關元件Ql、Q4導通時，則電流流在降壓斬波電路1 — 開關元件Q1—線圈L1—放電燈1—開關元件Q4—降壓斬 波電路1。一方面，當開關元件Q2、Q3導通時，則以降 壓斬波電路1—開關元件Q3—放電燈1—線圈L1—開關元 件Q2-降壓斬波電路1的路徑，將交流矩形波電流供應 於放電燈1。 驅動上述開關元件Q1〜Q4之際，爲了防止開關元件 馨 Q1〜Q4的同時導通，切換交流矩形波的極性時，設有將 元件開關元件Q1〜Q4都作成斷開之期間（空檔時間Td)。 又，供應於放電燈1的交流矩形波輸出的頻率，是由 60〜1 000Hz(穩定頻率）的範圍所選擇者，例如350Hz。又 ，上述空檔時間期間是由〇·5 μβ〜10μ3的範圍所選擇。 在此，本發明的放電燈點燈裝置是藉由表示於第3圖 的饋電裝置，以穩定頻率（6〇〜1000Hz)點燈表示於第1圖 的放電燈，而在其中定期地入低頻。該低頻是比穩定比率 -20- 201017341 還要低的頻率，由5〜200Hz的範圍被選擇，又，被插入 的波數以半周期作爲1單位而從1單位至10單位的範圍 被選擇，又’被插入在穩定頻率的間隔爲由0.01秒〜120 秒的範圍被選擇。 本發明的光照射裝置是以具有光源單元，及饋電裝置 ，及積分器者作爲要件者，其以外的構成要素，例如包含 折回鏡’濾波器，照度監測器等也可以。 〇 以上如所述地，本發明的光照射裝置，是交流點燈封 入水銀與鹵素的放電燈，在該放電燈的電極前端形成突起 ’該突起在燈點燈中不會消耗，利用維持大約同一大小的 性質’以解決電極間距離的增大與隨著其增大與隨著的照 度維持率降低的問題。又，從光源部所放射的光中，被入 射於積分器的光的比率，亦即將劣的利用率作成90%以下 ’則即使增加電極間距離，也不會受到其影響而可達成高 照度維持率。 ❹ 【圖式簡單說明】 第1圖是表示本發明的光照射裝置的槪略構成。 第2圖是表示本發明的光照射裝置的光源單元。 第3圖是表示本發明的光照射裝置的放電燈。 第4圖是表示說明本發明的放電燈的原理的模式圖。 第5圖是表示本發明的光照射裝置的實驗結果。 第6(a)圖至第6(c)圖是表示測定本發明的光照射裝置 的入射率的實驗的說明圖。 -21 - 201017341 第7圖是表示本發明的放電燈的放射波長。 第8圖是表示本發明的饋電裝置的電路構成。 【主要元件符號說明】 1 :放電燈 2 :反射鏡 3 :支撐體 4 :收納盒 _ 1 〇 :光源部 1 1 :發光部 1 2 :密封部 1 3 :金屬箔 14 :外部引線 20 :積分器 21 :準直平管 22 :屏蔽平台 · 23 :屏蔽 30 :饋電裝置 N :單元 W :工件 -22-Specifically, in the conventional general light irradiation device (500 mm x 600 mm viewing angle n of the high-pressure mercury lamp 'illumination field of 1 lamp 5k W -17- 201017341'), the irradiation energy necessary for the exposure surface is 45 mW/cm 2 °. In the light irradiation device in which the plurality of discharge lamps of the present invention constitute the light source portion, 61 lamps are used for the discharge lamp of 100 W, and the same illumination energy can be configured for the same illumination region. At this time, the '6 1 discharge lamp was 6.1 kW at the total power. On the other hand, 18 lamps are used for 200 W discharge lamps, and the same illumination energy can be used for the same illumination field. At this time, the 18 φ discharge lamp has a total power of 3.6 kW, and compared with the case of using a lamp of 10 〇〇W, the power efficiency is increased by about 59%, and it is also known that a 200 W discharge lamp is used to constitute the light source unit. When the light source unit is constituted by using a discharge lamp of 100 W, the number of lamps is reduced, and the power efficiency is increased. Similarly, for a conventional general light irradiation device (1 lamp of 10 kW high-pressure mercury lamp, irradiation field of 750 mm x 650 mm, viewing angle of 2.0°), if the light irradiation device of the invention is replaced by the discharge lamp of the invention, 94 is used for the discharge lamp of 100 watts. In this case, in the case of a discharge lamp of 200 W, 29 (total electric power: 5.8 kW) is used. As in the above, it is known that a 200 W discharge lamp is used to constitute a light source unit, and power efficiency is high. . Here, the discharge lamp of the present invention is sealed with mercury in the range of 8 to 0.25 mg/mm3. Fig. 7 is a view showing the spectral distribution of mercury of 〇·ΐ 5 mg/mm3 and a halogen-encapsulated discharge lamp. As shown in the figure, it is understood that a plurality of ultraviolet rays having a wavelength of 300 to 3 50 nm are emitted. If the mercury is less than 〇·〇8 mg/mm3, in addition to the luminescence at a wavelength of 300 to 35 〇nm, the luminescence caused by the water of -18-201017341 or less is also increased. The reason is therefore less than ideal. Moreover, the continuous spectrum near 350 to 45 Onm is also lowered, which is less desirable. If the mercury is more than 0.25 mg/mm3, the light emission at a wavelength of 300 to 3 5 Onm is less desirable. Figure 8 is a diagram showing a power feeding device for a discharge lamp. The power feeding device 3 is provided by a step-down chopper circuit 31 to which a DC voltage is supplied, and an output bridge connected to the step-down chopper circuit 31, and the DC voltage is changed to an AC voltage to supply the full bridge of the discharge lamp 1. The inverter circuit 32 (hereinafter also referred to as "full bridge circuit"), and the coil 11 connected in series to the discharge lamp, the capacitor C1, the start circuit 33, and the control circuit 34 are constructed by step-down chopper The circuit 31, the full bridge circuit 32, the start circuit 33, and the control circuit 34 constitute a power feeding device, and the discharge lamp 1 is referred to as a lighting device. The step-down chopper circuit 31 is composed of a drive circuit Gx connected to a DC power source VDC, a switch ® element Qx, and a diode Dx, and a coil Lx, and a smoothing capacitor Cx, and a switching element Qx. The switching element Qx is driven by on/off. By this driving, the ratio of the switching element Qx is adjusted to control the current or electric power supplied to the discharge lamp 10. The full bridge circuit 32 is composed of a bridge-connected transistor or FET switching elements Q1 to Q4 and switching elements Q1 to Q4. Further, in the case where the switching elements Q1 to Q4 are connected in parallel, the diodes are also connected in anti-parallel, but the diode is omitted in this embodiment. The control circuit 34 is composed of a power converter 340, a comparator 341, a pulse -19-201017341 wide modulation circuit 342, a control unit 343, and a full bridge circuit drive circuit 344. The power converter 340 converts the voltage signal or current signal detected by the resistors R1, R2, and R3 into a power signal. The power signal is compared with the reference power 在 at the comparator 341 and via the pulse width modulation circuit 342, the feedback control switching element Qx is performed. Thereby, the lighting power of the lamp is made to be fixed. Implementing so-called constant power control. Further, the switching elements Q 1 to Q4 are driven by the full bridge circuit driving circuit 34 via the control unit 3 43 . The operation of the full bridge circuit 32 is alternately repeated, with the switching elements Q1, Q4, and the switching elements Q1, Q4, and the switching elements Q2, Q3 being closed. When the switching elements Q1, Q4 are turned on, current flows in the step-down chopper circuit 1 - switching element Q1 - coil L1 - discharge lamp 1 - switching element Q4 - step-down chopper circuit 1. On the one hand, when the switching elements Q2, Q3 are turned on, the AC rectangular wave current is supplied in the path of the step-down chopper circuit 1 - the switching element Q3 - the discharge lamp 1 - the coil L1 - the switching element Q2 - the step-down chopper circuit 1. For the discharge lamp 1. When the switching elements Q1 to Q4 are driven, in order to prevent the switching elements Xin Q1 to Q4 from being turned on at the same time, when the polarity of the AC rectangular wave is switched, a period in which the element switching elements Q1 to Q4 are turned off is set (the neutral time Td). ). Further, the frequency of the AC rectangular wave output supplied to the discharge lamp 1 is selected from the range of 60 to 1 000 Hz (stable frequency), for example, 350 Hz. Further, the above neutral time period is selected from the range of 〇·5 μβ to 10 μ3. Here, the discharge lamp lighting device of the present invention is shown in the discharge device shown in FIG. 3 at a stable frequency (6 〇 to 1000 Hz), and is shown in the discharge lamp of Fig. 1, and periodically enters therein. Low frequency. The low frequency is a frequency lower than the stable ratio -20-201017341, and is selected from the range of 5 to 200 Hz. Further, the number of inserted waves is selected from one unit to 10 units in a half cycle as one unit, Also, the interval of being inserted at the stable frequency is selected from the range of 0.01 second to 120 seconds. The light-emitting device of the present invention is a component having a light source unit, a power feeding device, and an integrator, and other components may include, for example, a folding mirror's filter, an illuminance monitor, and the like. As described above, the light irradiation device of the present invention is a discharge lamp in which an alternating current lamp is sealed with mercury and a halogen, and a projection is formed at the tip end of the electrode of the discharge lamp. The projection is not consumed in the lamp lighting, and the use is maintained. The property of the same size 'to solve the problem of an increase in the distance between the electrodes and a decrease in the illuminance maintenance rate as it increases. Further, among the light emitted from the light source unit, the ratio of the light incident on the integrator is about 90% or less. Therefore, even if the distance between the electrodes is increased, the high illumination can be achieved without being affected. Maintenance rate. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic view showing a schematic configuration of a light irradiation device of the present invention. Fig. 2 is a view showing a light source unit of the light irradiation device of the present invention. Fig. 3 is a view showing a discharge lamp of the light irradiation device of the present invention. Fig. 4 is a schematic view showing the principle of the discharge lamp of the present invention. Fig. 5 is a view showing experimental results of the light irradiation device of the present invention. 6(a) to 6(c) are explanatory views showing an experiment for measuring the incident rate of the light irradiation device of the present invention. -21 - 201017341 Fig. 7 is a view showing the emission wavelength of the discharge lamp of the present invention. Fig. 8 is a view showing the circuit configuration of the power feeding device of the present invention. [Description of main component symbols] 1 : Discharge lamp 2 : Mirror 3 : Support 4 : Storage case _ 1 〇: Light source part 1 1 : Light-emitting part 1 2 : Seal part 1 3 : Metal foil 14 : External lead 20 : Integration 21: Collimation Flat Tube 22: Shielding Platform · 23: Shield 30: Feeder N: Unit W: Workpiece-22-