JP4258035B2 - Exposure apparatus and device manufacturing method - Google Patents

Description

但し、Ａ：投影光学系３０の露光領域の直径
ＮＡ_UL：投影光学系３０のウエハＷ側の開口数
ＢＬ：ベースライン長
で求められる。上記最大値ｈ₀は、投影光学系３０を構成する光学素子の厚みや、該光学素子を保持する保持部材の厚みを考慮していない値であり、ミラー１８とウエハＷとの距離ｈは、
ｈ≦ｈ₀ （Ｂ）
である。
【００２４】
ここで、ベースライン長ＢＬは、短ければ短いほど良いが（理由は後述）、このベースライン長ＢＬが短すぎると、上記（Ａ）、（Ｂ）式より距離ｈが短くなりすぎ、第１対物レンズ２を配置することが困難となる。本発明では、第１対物レンズ２の物体側開口数が条件式（１）を満足する構成であるため、ミラー１８とウエハＷとの距離ｈは、
３５ｍｍ≦ｈ≦７０ｍｍ
となる。ここで、ミラー１８とウエハＷとの間には、第１対物レンズ２を配置する必要があるため、この第１対物レンズ２の全長を考慮し、上記条件式（２）を得る。
条件式（２）を満足しない場合には、第１対物レンズ２及びミラー１８を、投影光学系３０とウエハＷとの間のスペースに配置することができなくなり、ベースライン長ＢＬを短く維持することが難しくなる。
【００２５】
また本発明では、ベースライン長ＢＬが、
５０ｍｍ≦ＢＬ≦２００ｍｍ （３）
を満足することが好ましい。図３において、ベースライン長ＢＬが、条件式（３）の上限を超える場合には、ウエハＷを保持するためのウエハステージ５０の拡大化や、スループットの低下、更にはアライメント精度の長期安定性の低下を招くことになる。
まず、ウエハステージ５０の拡大化について説明する。オフアクシスアライメント方式においてウエハＷの全面を位置検出することを可能にするためには、ベースライン長ＢＬが長ければ、その分ウエハステージ５０を大きくしなければならない。ウエハステージ５０が拡大化すると、装置の床面積（フットプリント）も大きくなる。また、ウエハステージ５０が拡大化すると、その重量も増すため、ウエハステージ５０移動時に、装置が振動するといった問題も生じる。
【００２６】
次に、スループット低下について説明する。オフアクシスアライメント方式においては、例えば特開平６−２７５４９６号公報に提案されているような位置合わせ方式が用いられる。すなわち、ウエハＷ上の複数のアライメントマークの位置を計測し、その結果より統計的手法を用いてショット領域の座標を予測し、その後、ウエハＷ上のショット領域が投影光学系３０の露光領域となるようにウエハステージ５０を移動させる。この移動時間はベースライン長ＢＬの長さに比例して大きくなるため、上記条件式（３）の上限を上回るときには、スループットの低下は著しくなる。
次に、アライメント精度の長期安定性に関して説明する。オフアクシスアライメント方式では、投影光学系３０の投影中心軸Ｚ１からベースライン長ＢＬだけ離れたアライメント位置で、アライメントを行っている。このとき、ベースライン長ＢＬを管理するために、ウエハステージ５０上には基準マークが設置されるが、上記条件式（３）の上限を超えてベースライン長ＢＬが長くなるときには、長期的な熱変動により基準マークの位置ずれ、回転等の影響が大きくなる。
【００２７】
また、実際に露光動作を行う際には、干渉計により計測したアライメント位置でのウエハステージ座標から、ベースライン長ＢＬをオフセットとして除去した座標を、露光位置として取り扱うが、上記熱変動により、ウエハステージ５０上に設けられる干渉計の移動鏡等の回転による影響が生じる。条件式（３）の上限を超える場合には、この回転による計測位置誤差が大きくなる。
また、上記条件式（３）の下限を超えてベースライン長ＢＬが短くなる場合には、条件式（１）を満足する第１対物レンズ２を配置することが困難になるか、あるいは投影光学系３０のウエハＷ側の開口数の確保や投影領域の確保が困難になる。
【００２８】
以下、図３における具体的数値例を示す。ウエハＷとミラー１８との距離ｈをｈ＝５０ｍｍとし、べースライン長ＢＬをＢＬ＝１００ｍｍとし、第１対物レンズ２の物体側開口数はＮＡ＝０．６とする。更に、第１対物レンズ２の各面の研磨面精度を平均でＲＭＳ＝０．０１λ（λ＝６３３ｎｍ）とすることで、低次の収差成分は非常に小さくなる。このとき、開口数がＮＡ＝０．２程度の従来の第１対物レンズ２のコマ収差と、本数値例による開口数がＮＡ＝０．６の第１対物レンズ２のコマ収差とは、ほぼ等しくなる。プロセスオフセットとコマ収差との関係は、
プロセスオフセット＝コマ収差／ＮＡ
となるので、本数値例によるプロセスオフセットは、従来のプロセスオフセットの約１／３の大きさになる。
【００２９】
【実施例】
本発明の実施例を図面によって説明する。図４にて、本発明による露光装置の第１実施例を示す。同図（Ｂ）に示すように、光ファイバー８より発した光束Ｔは、開口絞り７によって適当な照明光束Ｔとなって、コンデンサーレンズ６、視野絞り５、照明リレーレンズ４の順に透過した後、ハーフプリズム３に入射する。
同図（Ａ）に示すように、ハーフプリズム３で反射した光束Ｔは、各ミラー１８ａ、１８ｂ、１８ｃで反射した後に、第１対物レンズ２を透過して、ウエハＷ上のアライメントマークを照明する。
【００３０】
ここでウエハＷは、ウエハステージ５０上に設置されている。ウエハステージ５０は、露光装置の投影光学系３０の投影中心軸Ｚ１、及びアライメントマークに入射される光束Ｔの計測中心軸Ｚ２に垂直なｘｙ平面に設置されており、ウエハステージ駆動部（不図示）によりｘｙ方向及びｚ方向に移動可能となっている。
また第１対物レンズ２は、その光軸が、投影光学系３０の投影中心軸Ｚ１、及びアライメントマークに入射される光束Ｔの計測中心軸Ｚ２に対して平行となるように配置されている。
また、第１対物レンズ２の物体側開口数は、前述した条件式（１）を満足する。また、第１対物レンズ２の各レンズ面の研磨面精度のＲＭＳの平均値は、０．０１λ以下（λは、光束の波長）となっている。また、ミラー１８ｃと第１対物レンズ２のレンズ上面との距離は、前述した条件式（２）を満足する。また、ベースライン長ＢＬは、前述した条件式（３）を満足する。
【００３１】
ウエハＷ上のアライメントマークで反射した光束Ｔは、第１対物レンズ２を透過して、各ミラー１８ｃ、１８ｂ、１８ａで反射した後に、ハーフプリズム３に入射する。ハーフプリズム３を透過した光束Ｔは、第２対物レンズ９を透過して、ミラー３５で反射された後に、指標板１０上にアライメントマーク像を形成する。このアライメントマーク像を形成した光束Ｔは、リレーレンズ１１、１２を透過して、ｘｙ分岐ハーフプリズム１３に入射する。ｘｙ分岐ハーフプリズム１３に入射した光束Ｔのうち、ｘｙ分岐ハーフプリズム１３の光束分割面を透過した光束は、Ｘ座標用ＣＣＤ１４の撮像面に再結像される。
一方、ｘｙ分岐ハーフプリズム１３の光束分割面で反射した光束は、Ｙ座標用ＣＣＤ１５の撮像面に再結像される。これらＸ座標用ＣＣＤ１４及びＹ座標用ＣＣＤ１５から出力される撮像信号は、信号処理部４０で処理された後、アライメントマークの位置情報として主制御系（不図示）に転送される。この位置情報を基に、露光装置に設置されたマスク（不図示）とウエハＷの位置合わせを行った後に、マスクパターンをウエハＷ上の狙いの位置に投影光学系３０を介して転写露光する。
【００３２】
なお、本第１実施例では、ハーフプリズム３に対して、光ファイバー８から発した光束Ｔをハーフプリズム３の反射光路側で用い、ウエハＷで反射した光束Ｔをハーフプリズム３の透過光路側で用いているが、本発明は、逆に、光ファイバー８から発した光束Ｔをハーフプリズム３の透過光路側で用い、ウエハＷで反射した光束Ｔをハーフプリズム３の反射光路側で用いても良い。
また、本第１実施例では、ハーフプリズム３と第１対物レンズ２との間の光路内に、ミラー１８ａ、１８ｂ、１８ｃを３枚配置しているが、必要に応じてミラーの枚数を増減しても良い。
以上のように本第１実施例では、ベースラインＢＬが短いので、スループットを高くすることができる。また、第１対物レンズ２の焦点距離（作動距離）が短いので、全体の光路長が短くなり、アライメント装置としての安定性を高くすることができ、更に第１対物レンズ２の物体側開口数を比較的大きく設定できるので、プロセスオフセットを小さくすることができる。
【００３３】
次に図５にて、本発明による露光装置の第２実施例を示す。本第２実施例と前記第１実施例との構成の違いは、図４における第１対物レンズ２を分割して、第１対物レンズ前群２ａと第１対物レンズ後群２ｂとを用いた点である。そして、図４の第１対物レンズ２の位置に第１対物レンズ前群２ａを設置して、ミラー１８とハーフプリズム３の間の光路に第１対物レンズ後群２ｂを配置している。図４のような構成では、第１対物レンズ２が大き過ぎて、装置に挿入できない場合に、図５の構成は有効となる。すなわち、装置内に装着するには大き過ぎる第１対物レンズ２を、第１対物レンズ前群２ａと第１対物レンズ後群２ｂとに分割することで装置内に装着可能とする。これにより、本第２実施例においても、第１実施例と同等の効果を得ることができる。
なお、本第２実施例では、第１対物レンズ群の分割数を、第１対物レンズ前群２ａと第１対物レンズ後群２ｂとの２つとしたが、必要に応じて、それ以上の群数に分割することも可能である。
また、図４のようにミラー１８ａ、１８ｂ、１８ｃを３枚配置した場合は、第１対物レンズ後群２ｂを両ミラー１８ｂ、１８ｃの間等、ハーフプリズム３と第１対物レンズ前群２ａとの間の光路内のどこへでも配置することができる。
【００３４】
次に図６にて、本発明による露光装置の第３実施例を示す。本第３実施例と前記第１実施例との構成の違いは、アライメント装置内に、ウエハＷの合焦度を検出するオートフォーカス機構を併設した点である。すなわち、第１実施例のアライメント装置がウエハＷのｘｙ方向の位置合わせのみ行うのに対して、本第３実施例では更にウエハＷのｚ方向の位置合わせも行う。
同図（Ｂ）に示すように、光ファイバー８より発した光束Ｔは、開口絞り７によって適当な照明光束Ｔとなって、コンデンサーレンズ６を透過して、視野絞り板１６に入射する。ここで、同図（Ｃ）に示すように、視野絞り板１６は、オートフォーカス用スリット２７と、ｘｙ方向の位置合わせ用の視野絞り５とで構成されている。視野絞り板１６を射出した光束Ｔは、照明リレーレンズ４を透過して、ハーフプリズム３に入射する。
【００３５】
同図（Ａ）に示すように、ハーフプリズム３で反射した光束Ｔは、各ミラー１８ａ、１８ｂ、１８ｃで反射した後に、第１対物レンズ２を透過して、ウエハＷ上に視野絞り５とオートフォーカス用スリット２７の投影像を形成する。
ウエハＷ上で反射した光束Ｔは、第１対物レンズ２を透過して、各ミラー１８ｃ、１８ｂ、１８ａで反射した後に、ハーフプリズム３に入射する。ハーフプリズム３を透過した光束Ｔは、第２対物レンズ９を透過して、ビームスプリッター２０に入射する。ビームスプリッター２０で反射した光束Ｔは、前記第１実施例と同様にＸ座標用ＣＣＤ１４及びＹ座標用ＣＣＤ１５にアライメントマーク像を形成する。他方、ビームスプリッター２０を透過した光束Ｔは、オ−トフォーカス用スリット２７の投影像を選択する選択板２１を透過して、リレーレンズ２２を介して屋根型ミラー２３に入射する。ここで屋根型ミラー２３は、開口絞り７とほぼ共役となる位置に配置されて、入射した光束Ｔを２光束に分割して反射する。
【００３６】
屋根型ミラー２３で反射した２光束は、リレーレンズ２４を透過して、１次元ＣＣＤ等の焦点検出用の撮像素子２５の撮像面上に、オートフォーカス用スリット２７の２つの投影像Ａ、Ｂを形成する。ここで、屋根型ミラー２３の影響により、ウエハＷから撮像素子２５に至る結像光学系のテレセントリック性は崩れている。そのため、ウエハＷが合焦点から±ｚ方向にずれた場合、撮像素子２５上でのオートフォーカス用スリット２７の２つの投影像Ａ、Ｂ（同図Ｄ）の間隔は、計測方向に変化することになる。
この撮像素子２５から出力される撮像信号は、撮像信号処理部４５で処理された後、ウエハＷの焦点位置情報として主制御系（不図示）に転送される。この焦点位置情報を基に、その合焦度が最大となるように、露光装置に設置されたウエハＷの焦点位置合わせを行う。その後、ウエハＷ上のアライメントマークのｘｙ方向の位置合わせを行う。
【００３７】
以上のように本第３実施例では、前記第１実施例と同等の効果に加え、ウエハＷの合焦度の検出を行う上での効果を得ることができる。すなわち、第３実施例では、第１対物レンズ２の物体側開口数を大きく設定できるため、測定に十分な測定光束の光量を確保することができる。これにより、結像光学系としてそれ程倍率を上げることなく、高分解能なオートフォーカスを得ることができる。
なお、オートフォーカス系の開口数を、アライメント系の開口数より大きくしたい場合は、アライメント系に結像開口絞りを設けてやれば良い。
【００３８】
次に図７にて、本発明による露光装置の第４実施例を示す。本第４実施例の構成は、ｘｙ方向の位置検出用の光ファイバー８と、焦点検出用光源２９との２つの独立した照明系をもつ点で、前記第３実施例の構成とは異なる。
同図（Ｂ）に示すように、焦点検出用光源２９より発した焦点検出用光束は、ＡＦ用コンデンサーレンズ２８を透過して、オートフォーカス用スリット２７を照明する。オートフォーカス用スリット２７を射出した焦点検出用光束は、ダイクロイックミラー２６に入射する。ダイクロイックミラー２６を透過した焦点検出用光束は、照明リレーレンズ４を透過して、ハーフプリズム３に入射する。ハーフプリズム３で反射した後の焦点検出用光束は、第３実施例と同様の光路を通過して、ダイクロイックミラー３３に入射する。ダイクロイックミラー３３を透過した焦点検出用光束は、第３実施例と同様の光路を通過して、最後に撮像素子２５に入射する。
【００３９】
一方、同図（Ｂ）に示すように、光ファイバー８より発した位置検出用光束は、コンデンサーレンズ６を透過して、視野絞り５を照明する。視野絞り５を射出した位置検出用光束は、ダイクロイックミラー２６に入射する。ダイクロイックミラー２６で反射した位置検出用光束は、照明リレーレンズ４を透過して、ハーフプリズム３に入射する。ハーフプリズム３で反射した後の位置検出用光束は、第３実施例と同様の光路を通過して、ダイクロイックミラー３３に入射する。ダイクロイックミラー３３で反射した位置検出用光束は、第３実施例と同様の光路を通過して、最後にＸ座標用ＣＣＤ１４及びＹ座標用ＣＣＤ１５に入射する。
【００４０】
なお、本第４実施例では、ダイクロイックミラー２６、３３に対して、焦点検出用光源２９から発した焦点検出用光束をダイクロイックミラー２６、３３の透過光路側で用い、光ファイバー８から発した位置検出用光束をダイクロイックミラー２６、３３の反射光路側で用いているが、本発明は、逆に、焦点検出用光源２９から発した焦点検出用光束をダイクロイックミラー２６、３３の反射光路側で用い、光ファイバー８から発した位置検出用光束をダイクロイックミラー２６、３３の透過光路側で用いても良い。
以上により、本第４実施例においても、前記第３実施例と同等の効果を得ることができる。
【００４１】
【発明の効果】
以上のように本発明は、アライメントマークの位置を高精度に検出して、微細な線幅に対応できる露光装置を提供することができる。更に本発明は、ベースライン長が短くスループットが高く、全体の光路長が短く結像光学系としての安定性が高く、加えて高分解能のオートフォーカスを装備可能な露光装置を提供することができる。
【図面の簡単な説明】
【図１】結像光学系に（Ａ）コマ収差がないときの結像状態を示す概略図と、（Ｂ）コマ収差があるときの結像状態を示す概略図である。
【図２】物体側開口数によるコマ収差とプロセスオフセットとの関係を示す図である。
【図３】露光装置における第１対物レンズとミラーの配置を示す概略図である。
【図４】本発明の第１実施例による露光装置を示す（Ａ）概略図と、（Ｂ）Ａ図中Ｘ−Ｘ矢視図である。
【図５】本発明の第２実施例による露光装置を示す（Ａ）概略図と、（Ｂ）Ａ図中Ｘ−Ｘ矢視図である。
【図６】本発明の第３実施例による（Ａ）露光装置を示す概略図と、（Ｂ）Ａ図中Ｘ−Ｘ矢視図と、（Ｃ）Ｂ図中Ｙ−Ｙ矢視図と、（Ｄ）撮像素子上の投影像を示す概略図である。
【図７】本発明の第４実施例による露光装置を示す（Ａ）概略図と、（Ｂ）Ａ図中Ｘ−Ｘ矢視図である。
【図８】第１対物レンズを投影光学系領域外に配置した露光装置を示す図である。
【図９】従来技術による露光装置を示す図である。
【符号の説明】
１…落射プリズム ２…第１対物レンズ
２ａ…第１対物レンズ前群 ２ｂ…第１対物レンズ後群
３…ハーフプリズム ４…照明リレーレンズ
５…視野絞り ６…コンデンサーレンズ
７…開口絞り ８…光ファイバー
９…第２対物レンズ １０…指標板
１１、１２、２２、２４、２８…リレーレンズ
１３…ｘｙ分岐ハーフプリズム
１４…Ｘ座標用ＣＣＤ １５…Ｙ座標用ＣＣＤ
１６…視野絞り板
１８、１８ａ、１８ｂ、１８ｃ、３５…ミラー
２０…ビームスプリッター ２１…選択板
２３…屋根型ミラー ２５…撮像素子
２６、３３…ダイクロイックミラー
２７…オートフォーカス用スリット
２９…焦点検出用光源 ３０…投影光学系
４０…信号処理部 ４５…撮像信号処理部
５０…ウエハステージ
Ｗ…ウエハ Ｚ１…投影中心軸
Ｚ２…計測中心軸 Ｔ…光束[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an exposure apparatus used for transferring a projection original (a mask, a reticle, etc.) onto a substrate in a lithography process for manufacturing a device such as a semiconductor element, an imaging element, a liquid crystal display element, or a thin film magnetic head. .
[0002]
[Prior art]
Since devices such as semiconductor elements are generally formed by stacking a plurality of layers of circuit patterns on a wafer (or a substrate such as a glass plate) coated with a photosensitive material, an exposure apparatus for exposing circuit patterns onto the wafer. Is provided with an alignment device for aligning a reticle pattern to be exposed from now and each shot area of a wafer on which a circuit pattern has already been formed.
Conventionally, as this type of alignment apparatus, as disclosed in JP-A-4-65603, JP-A-4-273246, etc., light having a wide wavelength band emitted from a light source such as a halogen lamp is used. An off-axis system that illuminates an alignment mark (wafer mark) on the wafer, forms an enlarged image of the alignment mark on the image sensor, performs image processing on the obtained image signal, and detects the position of the alignment mark. A type of alignment apparatus is known. The detection system of the imaging type alignment apparatus is also called an FIA (Field Image Alignment) system.
[0003]
A conventional example of an exposure apparatus provided with an FIA alignment apparatus will be described with reference to FIG. A wide-band wavelength light beam T emitted from the optical fiber 8 becomes an appropriate illumination light beam T by the aperture stop 7, passes through the condenser lens 6, the field stop 5, and the illumination relay lens 4 in this order, and then enters the half prism 3. . The light beam T transmitted through the half prism 3 is transmitted through the first objective lens 2 and is incident on the incident light prism 1. The light beam T reflected by the light beam reflecting surface of the epi-illumination prism 1 illuminates the alignment mark on the wafer W.
Here, the wafer W is placed on the wafer stage 50. The wafer stage 50 is installed on an xy plane perpendicular to the projection center axis Z1 of the projection optical system 30 of the exposure apparatus and the measurement center axis Z2 of the light beam T incident on the alignment mark, and a wafer stage drive unit (not shown). ) Is movable in the xyz direction.
Further, a distance BL between the projection center axis Z1 of the projection optical system 30 and the measurement center axis Z2 of the light beam T incident on the alignment mark. ₁ In order to reduce the (baseline length), the first objective lens 2 is installed so that its optical axis is parallel to the xy plane, and there is an epi-prism on the optical path between the first objective lens 2 and the wafer W. 1 is installed.
[0004]
The light beam T reflected by the alignment mark on the wafer W is reflected by the light beam reflecting surface of the incident prism 1, passes through the first objective lens 2, and enters the half prism 3. The light beam T reflected by the half prism 3 is transmitted through the second objective lens 9 to form an alignment mark image on the indicator plate 10. The light beam T that forms this alignment mark image passes through the relay lenses 11 and 12 and enters the xy branch half prism 13. Of the light beam T incident on the xy branch half prism 13, the light beam transmitted through the light beam splitting surface of the xy branch half prism 13 is re-imaged on the imaging surface of the CCD 14 for X coordinate.
[0005]
On the other hand, the light beam reflected by the light beam splitting surface of the xy branch half prism 13 is re-imaged on the imaging surface of the Y-coordinate CCD 15. The imaging signals output from the X coordinate CCD 14 and the Y coordinate CCD 15 are processed by the signal processing unit 40 and then transferred to a main control system (not shown) as position information of the alignment mark. Based on this positional information, a mask (not shown) installed in the exposure apparatus is aligned with the wafer W, and then the mask pattern is transferred and exposed to a target position on the wafer W via the projection optical system 30. .
[0006]
[Problems to be solved by the invention]
The conventional alignment apparatus has an advantage that the influence of the thin film interference on the photoresist layer on the wafer is reduced because of the broadband illumination.
However, in the imaging optical system of the conventional imaging type alignment apparatus, a slight amount of aberration remains in the manufacturing process such as processing, assembly, and adjustment. If the aberration remains in this way, the contrast of the wafer mark image on the imaging surface is lowered, or the wafer mark image is distorted, resulting in a mark position detection error. In recent years, as the line width of a circuit pattern becomes finer, higher precision alignment is required, and a decrease in alignment accuracy due to the slight aberration as described above has become a problem.
[0007]
Of these remaining aberrations, the effect of coma on the detection of wafer mark images is particularly large. Coma aberration that is axially symmetric about the optical axis and uniform decentering coma aberration over the entire field of view, etc. When an asymmetric lateral aberration occurs in the optical system, the alignment mark image projected on the imaging surface is measured with a positional shift compared to the case of ideal imaging. Also, when the alignment mark shape such as pitch, duty ratio, step, etc. changes, or when the alignment mark is defocused, if the alignment mark image has coma aberration, the degree of asymmetry of the image varies. It changes, and the amount of deviation of the measurement position changes variously.
Since the shape of the alignment mark is different for each semiconductor manufacturing process, alignment of the wafer with an optical system with residual coma aberration may cause alignment errors due to process offsets and poor reproducibility of overlay accuracy. descend.
Accordingly, an object of the present invention is to provide an exposure apparatus that can detect the position of an alignment mark, which is a test object, with high accuracy and can cope with a fine line width.
[0008]
[Means for Solving the Problems]
The present invention has been made to solve the above-described problems. That is, an exposure apparatus according to the present invention includes an exposure optical system that projects a pattern provided on a projection original plate onto a photosensitive substrate. An illumination optical system for supplying alignment light having a predetermined wavelength to a position detection pattern provided on the photosensitive substrate, and alignment light from the position detection pattern illuminated by the alignment light A condensing optical system having an objective lens system provided at a position capable of receiving light, at least one light beam deflecting unit, and a photoelectric converting unit for photoelectrically converting alignment light via the condensing optical system. The numerical aperture on the photosensitive substrate side of the objective lens system is NA _W And when
0.35 ≦ NA _W ≦ 0.85
Is satisfied. At that time, the objective lens system in the condensing optical system is composed of optical elements whose average RMS value of polished surface accuracy of each surface is 0.01λ or less. The Where λ is the center wavelength used.
[0009]
According to a preferred aspect of the present invention, at least one of the at least one light beam deflecting means is disposed in the optical path between the light beam deflecting means disposed closest to the photosensitive substrate and the photosensitive substrate. An optical member having positive refractive power is disposed.
According to a preferred aspect of the present invention, the objective lens system has a front group and a rear group that are located closest to the photosensitive substrate, and the luminous flux is between the front group and the rear group. The deflecting means is located.
[0010]
Further, according to a preferred aspect of the present invention, the distance D between the light beam deflecting means arranged on the most photosensitive substrate side and the surface farthest from the photosensitive substrate of the objective lens system is:
D ≦ 30mm
It is arrange | positioned so that it may become.
According to a preferred aspect of the present invention, the objective lens system of the condensing optical system has a distance BL between a measurement center position of the objective lens system and a projection center position of the projection optical system,
50mm ≦ BL ≦ 200mm
It is arranged so as to satisfy.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described with reference to the drawings. In the present invention, the objective lens system in the alignment system is
0.35 ≦ NA _W ≦ 0.85 (1)
However, NA _W : Numerical aperture on the photosensitive substrate (wafer) side of the objective lens system
Configured to satisfy. Before describing the conditional expression (1), the influence of coma aberration of the optical system on image formation will be described with reference to FIG. Here, FIG. 1A shows an imaging state when no coma aberration remains in the optical system, and FIG. 1B shows an imaging state when coma aberration remains in the optical system. . On the object plane O, a line and space pattern with a pitch of p (hereinafter referred to as a pattern with a pitch p) and a line and space pattern with a pitch of 2p (hereinafter referred to as a pattern with a pitch of 2p). )).
[0012]
When these patterns are illuminated by illumination light (not shown), zero-order light and ± first-order light P (+1) and P (−1) are generated from the pattern of pitch p, and zero-order light is generated from the pattern of pitch 2p. And ± first-order light 2P (+1) and 2P (−1) are generated. In addition, from the patterns of the pitches p and 2p, higher-order diffracted light of ± 2nd order or higher is also generated, but here, the higher-order diffracted light is ignored.
In FIG. 1A, ± first-order diffracted lights P (+1) and P (−1) from a pattern with a pitch p, and ± first-order diffracted lights 2P (+1) and 2P (−1) from a pattern with a pitch 2p are Then, the light passes through the optical system including the lens groups L1 and L2 and the aperture stop AS, and reaches a predetermined point on the image plane I of the optical system. Here, for comparison with the optical system of FIG. 5B, the wavefront Wi on the pupil plane (position of the aperture stop AS) in the optical system of FIG.
[0013]
As shown in FIG. 1B, when coma aberration remains in the optical system including the lens groups L1 and L2 and the aperture stop AS, the ± first-order diffracted light P from the patterns of pitch p and 2p. The wavefront aberration Wc corresponding to the coma aberration is given to (+1), P (−1), 2P (+1), and 2P (−1). At this time, the wavefront aberration amount ΔWc (p (−1)) is given to the −1st order diffracted light P (−1) from the pattern of the pitch p, and the wavefront aberration amount ΔWc is given to the + 1st order diffracted light P (+1). (P (+1)) is given. The wavefront aberration amount ΔWc (2p (−1)) is given to the −1st order diffracted light 2P (−1) from the pattern of the pitch 2p, and the wavefront aberration amount ΔWc ( 2p (+1)). As described above, when there is a wavefront aberration corresponding to the coma aberration of the optical system, the wavefront aberration amounts ΔWc (p (−1)), ΔWc (p (+1)) and the wavefront aberration amount ΔWc (2p (−1)). , ΔWc (2p (+1)). Accordingly, there are positions where ± 1st order diffracted lights P (−1) and P (+1) reach image plane I and positions where ± 1st order diffracted lights 2P (−1) and 2P (+1) reach image plane I. Different. That is, when coma is present in the optical system, there is a problem that the image forming position of the pattern varies depending on the size of the pattern pitch.
[0014]
Further, not only the difference in pattern pitch as described above but also the change of the duty ratio, step, etc., the imaging position changes.
When aligning wafers using an optical system in which coma has remained during alignment in the semiconductor manufacturing process, the shape of the alignment mark (pitch, duty ratio, step, etc.) differs for each semiconductor manufacturing process. This causes a so-called process offset in which the detection position varies.
[0015]
Next, the conditional expression (1) will be described with reference to FIG. FIG. 2 is a graph showing a simulation result of the relationship between the coma aberration remaining in the alignment system and the process offset. The vertical axis represents the process offset and the horizontal axis represents the coma aberration. In this simulation, a process offset (an image shift amount from an ideal position) when a predetermined amount of residual coma aberration occurs is calculated for each object-side numerical aperture (NA). Thus, if the numerical aperture is the same, the smaller the remaining coma aberration, the smaller the process offset, and the same coma aberration, the larger the numerical aperture of the optical system. It can be seen that the process offset can be reduced.
In order to reduce the remaining coma, it is necessary not only to correct aberrations in optical design, but also to minimize errors in the manufacturing process, but to reduce the amount of coma caused by manufacturing errors. There are limits. As shown in FIG. 2, when the numerical aperture of the optical system is 0.35 or less, even if the manufacturing error is pushed to the limit, the process offset is sufficient to satisfy the strict alignment accuracy required with circuit miniaturization. It is difficult to make small.
[0016]
On the other hand, as the numerical aperture of the optical system is increased, the process offset can be reduced. However, when the numerical aperture exceeds 0.85, the effect of reducing the process offset by increasing the numerical aperture is effective. Less. That is, the change in the slope of the graph by increasing the numerical aperture is reduced.
In this case, the burden on the optical design also increases, and the occurrence of coma aberration on the higher-order optical design that cannot be corrected in the optical system following the objective lens system increases. In other words, coma aberration occurs, resulting in an increase in process offset.
In this case, it is also very difficult to ensure a so-called working distance between the wafer and the lens surface on the most wafer side of the objective lens system.
Therefore, the range defined by the conditional expression (1) is a range that is effective in reducing the process offset.
[0017]
Further, for example, as proposed in Japanese Patent Application No. 9-33283 by the applicant of the present application, in order to improve the autofocus resolution when attempting to implement the TTL autofocus through the alignment objective lens. For example, increase the magnification between the object to be detected and the autofocus detector while keeping the numerical aperture of the objective lens system small, or increase the numerical aperture of the objective lens system to increase the magnification not much. A street is conceivable. Here, in the former case, that is, when the value falls below the lower limit of the conditional expression (1), the amount of light reaching the autofocus detector decreases, and the autofocus measurement accuracy and reproducibility deteriorate, resulting in alignment. There is a risk of deteriorating accuracy. In the latter case, the amount of light reaching the autofocus detector is sufficiently maintained, so that the autofocus measurement accuracy and reproducibility are kept good.
[0018]
In the present invention, it is preferable that the average RMS value of the polishing accuracy of each surface of the optical element constituting the objective lens system is 0.01λ or less based on the conditional expression (1). Where λ is the center wavelength used.
As described above, the size of the process offset is proportional to the coma aberration and inversely proportional to the numerical aperture. Therefore, when the coma is constant, the process offset decreases when the numerical aperture is large.
Here, when the average value of the RMS of the polishing accuracy of each surface of the optical element that constitutes the objective lens system is 0.01λ or less, the inventors of the present application are low-order among the aberrations remaining in the objective lens system. It has been found that the aberration of the above is sufficiently small and the higher order aberration becomes dominant.
[0019]
At this time, in the pupil plane of the objective lens system, aberrations are randomly distributed in the pupil plane, and even if the numerical aperture is reduced, the coma aberration does not decrease much, and the process offset by increasing the numerical aperture The reduction of becomes large. On the other hand, when low-order aberrations are dominant, the coma aberration is reduced by reducing the numerical aperture, but the average value of the RMS of the polishing accuracy of each surface of the optical element described above. If the numerical aperture is 0.01λ or less, the coma value itself becomes larger than when the numerical aperture is increased, and as a result, the process offset cannot be improved.
In the present invention, the polishing accuracy is expressed by RMS, but the polishing accuracy can be paraphrased by a difference between a maximum value and a minimum value (so-called peak-to-valley or peak-to-peak) or an aberration coefficient. .
[0020]
In the present invention, as shown in FIG. 3, the first objective lens 2 (objective lens system) is disposed in the optical path between the mirror 18 (light beam deflecting member) arranged closest to the wafer W and the wafer W. It is preferable to arrange.
In the projection optical system 30 for projecting an image of a mask (projection original) on a wafer W (photosensitive substrate), the numerical aperture on the wafer W side has been increasing in recent years, and accordingly the wafer of the projection optical system 30 has been increased. It is very difficult to increase the working distance on the W side. Further, the outer diameter of the first objective lens 2 having the numerical aperture defined by the conditional expression (1) inevitably increases. Therefore, when the first objective lens 2 has the conventional arrangement shown in FIG. 9, the distance between the wafer W and the projection optical system 30 becomes very narrow, making it difficult to arrange the first objective lens 2. Furthermore, since it is difficult to secure a long working distance (focal length) with the first objective lens 2 having the numerical aperture defined by the conditional expression (1), the epi-prism 1 is used as the first objective as shown in FIG. A space for interposing the lens 2 on the object side cannot be secured.
[0021]
At this time, as shown in FIG. 8, a configuration in which the incident prism 1 is not interposed on the object side of the first objective lens 2 is conceivable. In this case, in order to secure an optical path for guiding the light flux T from the first objective lens 2 to the CCD 14 for X coordinates and the CCD 15 for Y coordinates, outside the region where the barrel of the projection optical system 30 is projected onto the xy plane, The first objective lens 2 needs to be provided. As described above, since the numerical aperture of the projection optical system 30 itself is also increased, the diameter of the lens barrel is large. Therefore, at this time, the baseline length BL ₁ Will be very long.
Therefore, by adopting the configuration as shown in FIG. 3, the numerical aperture of the first objective lens 2 can be increased in order to make the baseline length BL the same as the conventional length and to reduce the process offset.
[0022]
At that time, the first objective lens 2 shown in FIG. 3 is divided into a first objective lens front group and a first objective lens rear group, and the first objective lens front group and the first objective lens rear group, A mirror 18 for deflecting the light beam from the front group of the first objective lens can be disposed in the optical path between the first objective lens and the first objective lens. At this time, since the light beam having a large numerical aperture is converted into a light beam having a small numerical aperture by the convergence action of the front group of the first objective lens at the position of the mirror 18, the size of the mirror 18 needs to be so large. Absent. Furthermore, since the total length of the front group of the first objective lens (the distance from the lens surface closest to the object side to the lens surface closest to the image side) is shorter than the total length of the first objective lens 2, the working distance of the projection optical system 30 is Even in a short case, the first objective lens front group, the first objective lens rear group, and the mirror 18 can be installed in the installation space to secure an optical path for guiding the light beam to the detector.
[0023]
In the present invention, the distance D between the light beam deflecting member arranged closest to the photosensitive substrate and the surface farthest from the photosensitive substrate of the objective lens system is:
D ≦ 30mm (2)
It is preferable to arrange so that. In FIG. 3, the maximum value h of the distance between the mirror 18 (light beam deflecting member) and the wafer W (photosensitive substrate). ₀ Is

Where A: the diameter of the exposure area of the projection optical system 30
NA _UL : Numerical aperture on the wafer W side of the projection optical system 30
BL: Baseline length
Is required. Maximum value h ₀ Is a value that does not take into account the thickness of the optical element constituting the projection optical system 30 and the thickness of the holding member that holds the optical element, and the distance h between the mirror 18 and the wafer W is:
h ≦ h ₀ (B)
It is.
[0024]
Here, the shorter the baseline length BL, the better (the reason will be described later). However, if the baseline length BL is too short, the distance h becomes too short from the above formulas (A) and (B). It becomes difficult to arrange the objective lens 2. In the present invention, since the object-side numerical aperture of the first objective lens 2 satisfies the conditional expression (1), the distance h between the mirror 18 and the wafer W is
35mm ≦ h ≦ 70mm
It becomes. Here, since it is necessary to dispose the first objective lens 2 between the mirror 18 and the wafer W, the conditional expression (2) is obtained in consideration of the total length of the first objective lens 2.
If the conditional expression (2) is not satisfied, the first objective lens 2 and the mirror 18 cannot be disposed in the space between the projection optical system 30 and the wafer W, and the baseline length BL is kept short. It becomes difficult.
[0025]
In the present invention, the baseline length BL is
50mm ≦ BL ≦ 200mm (3)
Is preferably satisfied. In FIG. 3, when the baseline length BL exceeds the upper limit of the conditional expression (3), the wafer stage 50 for holding the wafer W is enlarged, the throughput is lowered, and the alignment accuracy is long-term stable. Will be reduced.
First, enlargement of the wafer stage 50 will be described. In order to make it possible to detect the position of the entire surface of the wafer W in the off-axis alignment method, if the base line length BL is long, the wafer stage 50 must be enlarged accordingly. When the wafer stage 50 is enlarged, the floor area (footprint) of the apparatus also increases. Further, when the wafer stage 50 is enlarged, the weight of the wafer stage 50 is increased, so that the apparatus vibrates when the wafer stage 50 is moved.
[0026]
Next, throughput reduction will be described. In the off-axis alignment method, for example, an alignment method as proposed in Japanese Patent Laid-Open No. 6-27596 is used. That is, the positions of a plurality of alignment marks on the wafer W are measured, and the coordinates of the shot area are predicted using a statistical method based on the results. Thereafter, the shot area on the wafer W is determined as the exposure area of the projection optical system 30. The wafer stage 50 is moved so that Since the moving time increases in proportion to the length of the baseline length BL, when the upper limit of the conditional expression (3) is exceeded, the throughput is significantly reduced.
Next, the long-term stability of alignment accuracy will be described. In the off-axis alignment method, alignment is performed at an alignment position separated from the projection center axis Z1 of the projection optical system 30 by the baseline length BL. At this time, in order to manage the baseline length BL, a reference mark is placed on the wafer stage 50. However, when the baseline length BL becomes longer than the upper limit of the conditional expression (3), a long-term operation is performed. Due to the thermal fluctuation, the influence of the positional deviation, rotation, etc. of the reference mark becomes large.
[0027]
Further, when actually performing the exposure operation, the coordinates obtained by removing the baseline length BL as an offset from the wafer stage coordinates at the alignment position measured by the interferometer are handled as the exposure position. The influence of rotation of a moving mirror of an interferometer provided on the stage 50 occurs. When the upper limit of conditional expression (3) is exceeded, the measurement position error due to this rotation becomes large.
Further, when the baseline length BL becomes shorter than the lower limit of the conditional expression (3), it is difficult to dispose the first objective lens 2 that satisfies the conditional expression (1), or the projection optics. It becomes difficult to secure the numerical aperture on the wafer W side of the system 30 and the projection area.
[0028]
Hereinafter, specific numerical examples in FIG. 3 will be shown. The distance h between the wafer W and the mirror 18 is h = 50 mm, the base line length BL is BL = 100 mm, and the object-side numerical aperture of the first objective lens 2 is NA = 0.6. Further, by setting the average polished surface accuracy of each surface of the first objective lens 2 to RMS = 0.01λ (λ = 633 nm), the low-order aberration component becomes very small. At this time, the coma aberration of the conventional first objective lens 2 having a numerical aperture of about NA = 0.2 and the coma aberration of the first objective lens 2 having a numerical aperture of NA = 0.6 according to this numerical example are almost equal. Will be equal. The relationship between process offset and coma is
Process offset = coma / NA
Therefore, the process offset according to this numerical example is about 1/3 of the conventional process offset.
[0029]
【Example】
Embodiments of the present invention will be described with reference to the drawings. FIG. 4 shows a first embodiment of an exposure apparatus according to the present invention. As shown in FIG. 5B, the light beam T emitted from the optical fiber 8 becomes an appropriate illumination light beam T by the aperture stop 7 and passes through the condenser lens 6, the field stop 5, and the illumination relay lens 4 in this order. The light enters the half prism 3.
As shown in FIG. 2A, the light beam T reflected by the half prism 3 is reflected by the mirrors 18a, 18b, 18c, and then passes through the first objective lens 2 to illuminate the alignment mark on the wafer W. To do.
[0030]
Here, the wafer W is placed on the wafer stage 50. The wafer stage 50 is installed on an xy plane perpendicular to the projection center axis Z1 of the projection optical system 30 of the exposure apparatus and the measurement center axis Z2 of the light beam T incident on the alignment mark, and a wafer stage drive unit (not shown). ) Is movable in the xy and z directions.
The first objective lens 2 is arranged such that its optical axis is parallel to the projection center axis Z1 of the projection optical system 30 and the measurement center axis Z2 of the light beam T incident on the alignment mark.
Further, the object-side numerical aperture of the first objective lens 2 satisfies the above-described conditional expression (1). Also, the average RMS value of the polished surface accuracy of each lens surface of the first objective lens 2 is 0.01λ or less (λ is the wavelength of the light beam). The distance between the mirror 18c and the upper surface of the first objective lens 2 satisfies the conditional expression (2) described above. Further, the baseline length BL satisfies the conditional expression (3) described above.
[0031]
The light beam T reflected by the alignment mark on the wafer W passes through the first objective lens 2 and is reflected by the mirrors 18c, 18b, and 18a, and then enters the half prism 3. The light beam T transmitted through the half prism 3 is transmitted through the second objective lens 9 and reflected by the mirror 35, and then forms an alignment mark image on the index plate 10. The light beam T that forms this alignment mark image passes through the relay lenses 11 and 12 and enters the xy branch half prism 13. Of the light beam T incident on the xy branch half prism 13, the light beam transmitted through the light beam splitting surface of the xy branch half prism 13 is re-imaged on the imaging surface of the CCD 14 for X coordinate.
On the other hand, the light beam reflected by the light beam splitting surface of the xy branch half prism 13 is re-imaged on the imaging surface of the Y-coordinate CCD 15. The imaging signals output from the X coordinate CCD 14 and the Y coordinate CCD 15 are processed by the signal processing unit 40 and then transferred to a main control system (not shown) as position information of the alignment mark. Based on this positional information, a mask (not shown) installed in the exposure apparatus is aligned with the wafer W, and then the mask pattern is transferred and exposed to a target position on the wafer W via the projection optical system 30. .
[0032]
In the first embodiment, with respect to the half prism 3, the light beam T emitted from the optical fiber 8 is used on the reflection light path side of the half prism 3, and the light beam T reflected by the wafer W is used on the transmission light path side of the half prism 3. However, in the present invention, conversely, the light beam T emitted from the optical fiber 8 may be used on the transmitted light path side of the half prism 3, and the light beam T reflected by the wafer W may be used on the reflected light path side of the half prism 3. .
In the first embodiment, three mirrors 18a, 18b and 18c are arranged in the optical path between the half prism 3 and the first objective lens 2, but the number of mirrors can be increased or decreased as necessary. You may do it.
As described above, in the first embodiment, since the baseline BL is short, the throughput can be increased. Further, since the focal length (working distance) of the first objective lens 2 is short, the entire optical path length is shortened, the stability as the alignment apparatus can be increased, and the numerical aperture on the object side of the first objective lens 2 is further increased. Can be set relatively large, the process offset can be reduced.
[0033]
Next, FIG. 5 shows a second embodiment of the exposure apparatus according to the present invention. The difference in configuration between the second embodiment and the first embodiment is that the first objective lens 2 in FIG. 4 is divided and the first objective lens front group 2a and the first objective lens rear group 2b are used. Is a point. Then, the first objective lens front group 2 a is installed at the position of the first objective lens 2 in FIG. 4, and the first objective lens rear group 2 b is arranged in the optical path between the mirror 18 and the half prism 3. In the configuration shown in FIG. 4, the configuration shown in FIG. 5 is effective when the first objective lens 2 is too large to be inserted into the apparatus. That is, the first objective lens 2 that is too large to be mounted in the apparatus can be mounted in the apparatus by dividing the first objective lens front group 2a and the first objective lens rear group 2b. Thereby, also in this 2nd Example, the effect equivalent to 1st Example can be acquired.
In the second embodiment, the number of divisions of the first objective lens group is two, that is, the first objective lens front group 2a and the first objective lens rear group 2b. It can also be divided into numbers.
Further, when three mirrors 18a, 18b, and 18c are arranged as shown in FIG. 4, the first objective lens rear group 2b is placed between the mirrors 18b and 18c, and so on, and the half prism 3 and the first objective lens front group 2a. Can be placed anywhere in the light path between.
[0034]
Next, FIG. 6 shows a third embodiment of the exposure apparatus according to the present invention. The difference in configuration between the third embodiment and the first embodiment is that an autofocus mechanism for detecting the degree of focus of the wafer W is provided in the alignment apparatus. That is, while the alignment apparatus of the first embodiment only performs alignment of the wafer W in the xy direction, the alignment of the wafer W in the z direction is further performed in the third embodiment.
As shown in FIG. 5B, the light beam T emitted from the optical fiber 8 becomes an appropriate illumination light beam T by the aperture stop 7, passes through the condenser lens 6, and enters the field stop plate 16. Here, as shown in FIG. 2C, the field stop plate 16 is composed of an autofocus slit 27 and a field stop 5 for alignment in the xy direction. The light beam T emitted from the field stop plate 16 passes through the illumination relay lens 4 and enters the half prism 3.
[0035]
As shown in FIG. 6A, the light beam T reflected by the half prism 3 is reflected by the mirrors 18a, 18b, and 18c, then passes through the first objective lens 2 and on the wafer W with the field stop 5 and A projection image of the autofocus slit 27 is formed.
The light beam T reflected on the wafer W is transmitted through the first objective lens 2 and reflected by the mirrors 18c, 18b, 18a, and then enters the half prism 3. The light beam T transmitted through the half prism 3 passes through the second objective lens 9 and enters the beam splitter 20. The light beam T reflected by the beam splitter 20 forms an alignment mark image on the X coordinate CCD 14 and the Y coordinate CCD 15 as in the first embodiment. On the other hand, the light beam T that has passed through the beam splitter 20 passes through the selection plate 21 that selects the projection image of the autofocus slit 27 and enters the roof mirror 23 via the relay lens 22. Here, the roof-type mirror 23 is disposed at a position substantially conjugate with the aperture stop 7 and reflects the incident light beam T by dividing it into two light beams.
[0036]
The two light fluxes reflected by the roof-type mirror 23 are transmitted through the relay lens 24, and are projected on the image pickup surface 25 of the focus detection image pickup device 25 such as a one-dimensional CCD. Form. Here, the telecentricity of the imaging optical system from the wafer W to the image sensor 25 is broken due to the influence of the roof-type mirror 23. Therefore, when the wafer W is deviated from the focal point in the ± z direction, the interval between the two projected images A and B (D in the figure) of the autofocus slit 27 on the image sensor 25 changes in the measurement direction. become.
The imaging signal output from the imaging element 25 is processed by the imaging signal processing unit 45 and then transferred to the main control system (not shown) as the focal position information of the wafer W. Based on the focal position information, the focal position of the wafer W installed in the exposure apparatus is adjusted so that the degree of focusing becomes maximum. Thereafter, the alignment marks on the wafer W are aligned in the xy direction.
[0037]
As described above, in the third embodiment, in addition to the same effects as in the first embodiment, an effect in detecting the degree of focus of the wafer W can be obtained. That is, in the third embodiment, since the object-side numerical aperture of the first objective lens 2 can be set large, it is possible to secure a sufficient amount of measurement light flux for measurement. Thereby, high-resolution autofocus can be obtained without increasing the magnification as much as the imaging optical system.
If it is desired to make the numerical aperture of the autofocus system larger than the numerical aperture of the alignment system, an imaging aperture stop may be provided in the alignment system.
[0038]
Next, FIG. 7 shows a fourth embodiment of the exposure apparatus according to the present invention. The configuration of the fourth embodiment is different from the configuration of the third embodiment in that it has two independent illumination systems of an optical fiber 8 for position detection in the xy direction and a light source 29 for focus detection.
As shown in FIG. 5B, the focus detection light beam emitted from the focus detection light source 29 passes through the AF condenser lens 28 and illuminates the autofocus slit 27. The focus detection light beam emitted from the autofocus slit 27 enters the dichroic mirror 26. The focus detection light beam that has passed through the dichroic mirror 26 passes through the illumination relay lens 4 and enters the half prism 3. The focus detection light beam reflected by the half prism 3 passes through the same optical path as in the third embodiment and enters the dichroic mirror 33. The focus detection light beam transmitted through the dichroic mirror 33 passes through the same optical path as in the third embodiment, and finally enters the image sensor 25.
[0039]
On the other hand, as shown in FIG. 5B, the position detection light beam emitted from the optical fiber 8 passes through the condenser lens 6 and illuminates the field stop 5. The position detection light beam emitted from the field stop 5 enters the dichroic mirror 26. The position detection light beam reflected by the dichroic mirror 26 passes through the illumination relay lens 4 and enters the half prism 3. The position detecting light beam reflected by the half prism 3 passes through the same optical path as in the third embodiment and enters the dichroic mirror 33. The position detection light beam reflected by the dichroic mirror 33 passes through the same optical path as in the third embodiment, and finally enters the X coordinate CCD 14 and the Y coordinate CCD 15.
[0040]
In the fourth embodiment, the position detection light emitted from the optical fiber 8 is used with respect to the dichroic mirrors 26 and 33 using the focus detection light beam emitted from the focus detection light source 29 on the transmission light path side of the dichroic mirrors 26 and 33. However, the present invention conversely uses the focus detection light beam emitted from the focus detection light source 29 on the reflection light path side of the dichroic mirrors 26, 33. The position detecting light beam emitted from the optical fiber 8 may be used on the transmitted light path side of the dichroic mirrors 26 and 33.
As described above, also in the fourth embodiment, an effect equivalent to that of the third embodiment can be obtained.
[0041]
【The invention's effect】
As described above, the present invention can provide an exposure apparatus that can detect the position of the alignment mark with high accuracy and cope with a fine line width. Furthermore, the present invention can provide an exposure apparatus that has a short baseline length, a high throughput, a short overall optical path length, a high stability as an imaging optical system, and can be equipped with a high-resolution autofocus. .
[Brief description of the drawings]
FIGS. 1A and 1B are a schematic diagram showing an imaging state when there is no coma aberration in the imaging optical system, and FIG. 1B is a schematic diagram showing an imaging state when there is coma aberration.
FIG. 2 is a diagram showing a relationship between coma aberration and process offset due to numerical aperture on the object side.
FIG. 3 is a schematic view showing an arrangement of a first objective lens and a mirror in the exposure apparatus.
4A is a schematic view showing an exposure apparatus according to the first embodiment of the present invention, and FIG. 4B is a view taken along line XX in FIG.
5A is a schematic view showing an exposure apparatus according to a second embodiment of the present invention, and FIG. 5B is a view taken along the line XX in FIG.
6A is a schematic view showing an exposure apparatus according to a third embodiment of the present invention, FIG. 6B is a schematic view showing an XX arrow in FIG. A, and FIG. (D) It is the schematic which shows the projection image on an image pick-up element.
7A is a schematic view showing an exposure apparatus according to a fourth embodiment of the present invention, and FIG. 7B is a view taken along line XX in FIG.
FIG. 8 is a view showing an exposure apparatus in which a first objective lens is disposed outside a projection optical system region.
FIG. 9 is a view showing an exposure apparatus according to the prior art.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Incident prism 2 ... 1st objective lens
2a ... first objective lens front group 2b ... first objective lens rear group
3 ... Half prism 4 ... Lighting relay lens
5 ... Field stop 6 ... Condenser lens
7 ... Aperture stop 8 ... Optical fiber
9 ... Second objective lens 10 ... Indicator plate
11, 12, 22, 24, 28 ... relay lens
13 ... xy branching half prism
14 ... CCD for X coordinate 15 ... CCD for Y coordinate
16 ... Field stop plate
18, 18a, 18b, 18c, 35 ... Mirror
20 ... Beam splitter 21 ... Selection plate
23 ... Roof-type mirror 25 ... Image sensor
26, 33 ... Dichroic mirror
27 ... Slit for autofocus
29 ... Light source for focus detection 30 ... Projection optical system
40: Signal processing unit 45 ... Imaging signal processing unit
50. Wafer stage
W ... wafer Z1 ... projection center axis
Z2 ... Measurement center axis T ... Light flux

Claims (4)

In an exposure apparatus provided with a projection optical system that projects a pattern provided on a projection original plate onto a photosensitive substrate,
An illumination optical system for supplying alignment light of a predetermined wavelength to a position detection pattern provided on the photosensitive substrate;
A condensing optical system having an objective lens system provided at a position capable of receiving alignment light from the position detection pattern illuminated by the alignment light, and at least one light beam deflecting unit;
Photoelectric conversion means for photoelectrically converting alignment light via the condensing optical system;
Means for guiding the alignment light to the objective lens system, and guiding the alignment light from the position detection pattern via the objective lens system to the photoelectric conversion means;
With
When the numerical aperture on the photosensitive substrate side of the objective lens system is NA _W ,
0.35 ≦ NA _W ≦ 0.85
Satisfied,
The objective lens system includes a front group including an optical member that is located closest to the photosensitive substrate and has at least one positive refractive power, and a rear group.
The light beam deflecting member arranged closest to the photosensitive substrate among the at least one light beam deflecting means is located between the front group and the rear group ,
Wherein in said condensing optical system objective lens system, the exposure apparatus the mean value of the RMS of each surface of the polishing surface accuracy is characterized Rukoto consists of the following optical elements 0.01Ramuda. The distance D between the light beam deflecting means arranged on the most photosensitive substrate side and the surface of the objective lens system farthest from the photosensitive substrate is:
D ≦ 30mm
The exposure apparatus according to claim 1 , wherein the exposure apparatus is arranged so that The objective lens system of the condensing optical system has a distance BL between the measurement center position of the objective lens system and the projection center position of the projection optical system,
50mm ≦ BL ≦ 200mm
The exposure apparatus according to claim 1 or claim 2 wherein you being arranged so as to satisfy. 4. A device manufacturing method comprising: exposing a circuit pattern on a wafer using the exposure apparatus according to claim 1 ; and stacking a plurality of layers of circuit patterns.

Images