JP3618907B2 - Pattern formation state detection apparatus and projection exposure apparatus using the same - Google Patents

Description

ここでｎ１、ｎ２はそれぞれ格子のＬ（ライン部分）とＳ（スペース相当部分）の屈折率である。ライン部分Ｌとスペース部分Ｓは、レジストの潜像の場合、Ｌがレジスト、Ｓが露光されたレジストであり、レジストを現像した場合、ＬがレジストでＳが空気等の気体である。
【００５８】
また上記屈折率の式は周期構造の周期が波長に対して十分小さい場合の近似式であり、厳密に電界を計算することで周期が波長同等の場合や、周期構造以外の屈折率も計算できる。
【００５９】
偏光解析のモデルはウエハ基板上の所定の厚みの複屈折媒質の複屈折率を測定することに相当する。偏光解析法とは、このウエハ基板上に、Ｐ／Ｓ（Ｐ偏光／Ｓ偏光）の位相差０、振幅比１の直線偏光光を所定の角度θで入射させ、その反射光の位相差（Δ）と振幅比（ψ）を測定し、ｎ‖、ｎ⊥を求めることができるものである。そして、予め測定したライン部分Ｌとスペース部分Ｓの屈折率ｎ１、ｎ２及びレジスト厚ｄと、この偏光解析法により求まったｎ‖、ｎ⊥の値から、上記式や厳密に解いた式を用いてデューティｔを求めている。またこの偏光解析法は公知であるので詳細は省略する。
【００６０】
入射光束の条件（入射角、パターン方向に対する方位角、波長）についてはレジストの屈折率や、膜厚、パターンの形状などで最適な条件が異なる。現像レジストパターンの測定と、潜像レジストパターンの測定とでも条件は異なる。更に潜像の場合、被露光部と未露光部との屈折率差が小さいため、露光条件の微小な変化でパターンの形状に発生する微小な変化を光束の微小な違いで分解検出するためには露光条件によって光束の変化が大きい入射条件で測定している。
【００６１】
そのために本発明の一実施形態は複数の入射条件で入射可能な装置であり、上記のように偏光解析法を利用してレジスト内の潜像パターン、もしくは現像後の凹凸パターンのデューティを測定している。
【００６２】
本発明の一実施形態ではある入射条件で測定した値から最適な入射条件を計算し、この最適な入射条件で測定することで分解能を高めて測定している。
【００６３】
偏光解析法では入射条件によっては異なるレジスト膜厚やレジストパターン形状においても同じ偏光状態が再現されることがある。そのため、本発明の一実施形態では複数の入射条件で反射光の変化を検出することで、ある入射条件では重なっていたレジスト形状を別の入射条件で分解して測定値を求めている。
【００６４】
本発明の一実施形態はＬ＆Ｓのパターンを異なる露光条件で感光基板上に転写して複数の感光パターンを形成する工程と、前記複数の感光パターンに対し、順次、前記光束を照射し、その反射光の偏光状態を検出し、前記偏光状態から感光パターンのデューティを算出する工程と、前記デューティが所望になる露光条件を決定する工程と、その露光条件でその後、複数のウエハに焼き付ける工程を含み、前記光束入射手段は入射光束の入射条件を可変であることを特徴とする。
【００６５】
（実施例１）
次に本発明の実施形態を図を用いて説明する。図１は本発明の実施例１の要部概略図、図２〜図５は図１の一部分を抽出したときの説明図である。
【００６６】
本実施例は感光パターンを介した入射光束の変化として偏光状態の変化を利用した場合である。
【００６７】
図１において１０１は縮小投影レンズであり、後述する露光用の光源１０７からの露光光で照明したレチクル１０２面上の回路パターン１０２ａをウエハ１０３上に投影している。１０４はウエハチャックであり、ウエハ１０３を吸着している。１０５は粗微動ステージであり、ウエハチャック１０４をＺ方向の粗微動させている。１０６はＸＹステージであり、ＸＹ方向にウエハチャック１０４を移動させている。１０７は露光用の光源である。１０８は構造体であり、光源１０７、レチクル１０２、鏡筒１０１及びウエハステージ１０６、アーチ３０６を支えている。
【００６８】
次に本実施形態で用いているウエハ１０３の投影レンズ１０１の光軸方向の位置、即ちフォーカス位置を検出する為のフォーカス位置制御装置とレチクル１０２を露光光で照明する際の露光量制御装置とを有する制御手段について図４を参照して説明する。
【００６９】
図３において２０１は半導体レーザ等の高輝度の光源である。光源２０１から出たレーザ光は折曲げミラー２０８により方向を変えられた後、ウエハ１０３の表面に入射する。ウエハ１０３の測定点Ｐ１で反射した後、光束は折曲げミラー２０９で方向を変えられた後、入射光束の２次元位置を検出する検出素子２０２に入射する。検出素子２０２は、例えばＣＣＤ等から成り、光束の入射位置を検知している。ウエハ１０３の表面のＺ方向（投影レンズ１０１の光軸方向）の位置変化が検出素子２０２上の入射位置の位置ずれとなるように検出している。この検出素子２０２からの信号に基づいてフォーカス制御装置２０３はウエハ１０３のＺ方向の位置、即ちフォーカス位置を粗微動ステージ１０５で制御している。
【００７０】
２０６はシャッター開閉機構、２０７はハーフミラー、２０５は照度センサーであり、ハーフミラー２０７で反射した光源１０７からの光束の露光量を検出している。積算露光制御装置２０４は照度センサー２０５からの信号に基づいてシャッターの開閉機構２０６を制御して光源１０７からの光束の積算光量を制御している。これによりレチクル１０２を照射する露光量が予め設定した値となるように調整している。
【００７１】
本実施例では、このような露光量制御装置とフォーカス位置制御装置より成る制御手段を利用してレチクル１０２面上のパターンをウエハ１０３面上に投影する際の露光条件を制御している。
【００７２】
次に光源１０７からの露光光で照明したレチクル１０２面上のパターンをウエハ１０３面上に投影する場合について説明する。
【００７３】
図４はレチクル１０２面上の基準パターン（以下「パターン」という）１０２ｐの説明図である。パターン１０２ｐはライン（Ｌ）とスペース（Ｓ）より成るＬ＆Ｓのパターン１０２１、１０２２を互いに直交させて構成している。
【００７４】
本実施形態ではＬ＆Ｓのパターン１０２１、１０２２より成るパターン１０２ｐを描画したレチクル１０２をレチクルステージ１０２ａにセットし、レジストを塗布したウエハ１０３をウエハチャック１０４にセットし、レチクル１０２のパターン１０２ｐをステップ＆リピート方式でウエハ１０３上に順次露光していく。このとき前述したフォーカス制御装置２０３及び積算露光制御装置２０４を用いて図６に示すように１ショットである領域１０３１を順次焼き付け、パターン１０２１、１０２２の潜像１０３１２および１０３１１を焼き付ける。
【００７５】
そしてＸ方向のショット位置に応じて、フォーカスオフセットを予想最適位置（予想ベストフォーカス位置）を中心に一定量ずつ変えながらステップ移動し、Ｙ方向のショットに対しては同様に、最適露光量（シャッター時間）を中心に露光量を変えながら露光していく。図５の例では説明の都合上、３×３のマトリックスであるが、このショット数は多い方が条件を出し易くなる。
【００７６】
図６は、このように順次露光したときのウエハ１０３のレジストの断面の説明図である。露光後のウエハ１０３のレジストは図６に示すようにレジスト内に潜像を形成する。潜像は露光光によってレジストが化学変化等で性質が変化して構成されたもので、この斜線で示した部分が露光した部分であり、一般的に屈折率が変化している。図６のマトリックス番地は図５のそれぞれに対応した位置での断面であり、（１）、（２）、（３）はそれぞれのチップのパターン像１０３１２の断面を表している。
【００７７】
次に、露光後のウエハ１０３のレジストは、図６に示す様に潜像を、ウエハチャック１０４から外すことなく、光源部３０１、受光部３０２、ドライバー３０３、そして偏光処理装置３０４で構成される偏光解析装置３００により入射光に対する反射光の振幅比ψと位相差Δを測定する。
【００７８】
図２は、この時の図１の偏光解析装置３００の主要部分を抽出した概略図である。図１０は図２の光路を展開した説明図である。図２、図１０において光源部（光束入射手段）３０１は光源３０１１（ウエハ１０３上のＬ＆Ｓピッチ以上の波長でＨｅＮｅレーザや半導体レーザばかりでなく分光器の単色光でも良い）と偏光素子３０１２（グラントムソン等）からなる。偏光素子３０１２はウエハ１０３に対し、Ｐ偏光成分（紙面に平行）とＳ偏光成分（紙面に垂直）が等量になるように紙面と偏光面が４５度になるように設置している。従ってこの光束Ｐ，Ｓの位相差Δは０で、振幅比ψは１である。
【００７９】
受光部（受光手段）３０２は、異方性の軸が光束３０５に直交したλ／４波長板３０２４、アナライザーであるところの偏光素子３０２２（グラントムソン等）、光電変換素子３０２１とを有している。更にλ／４板３０２４は光束３０５方向を回転軸とする回転機構３０２３内に保持され、ドライバー３０３の指令により一定速度で回転している。
【００８０】
今、光源部３０１からの射出光３０５がウエハ１０３上のレジスト表面とウエハ表面で反射し、その合波光束がウエハ１０３上のレジストの複屈折率ｎ１、ｎ２等に応じてＰ、Ｓ偏光の位相差Δと振幅比ψが変化する。
【００８１】
その光束を回転するλ／４板３０２４とアナライザー３０２２を通してディテクタ３０２１で検出し、位相差Δと振幅比ψに応じた正弦波の電気信号を得て、その振幅と直流成分の大きさ正弦波の位相情報から、上記Δ、ψを求めている。
【００８２】
そして、偏光解析法で測定されたΔ、ψより求められた ｎ‖、ｎ⊥と、予め測定されているレジストの未露光部分の屈折率ｎ１、被露光部分の屈折率ｎ２、レジストの厚みｄ、および基板の複素屈折率ｎｓより、デューティｔ⊥（ ｎ⊥を用いたときのデューティ）を以下の式で求めている。
【００８３】
ｔ⊥＝ｎ１＊＊２・（ｎ２＊＊２−ｎ⊥＊＊２）／（ｎ⊥＊＊２・（ｎ２＊＊２−ｎ１＊＊２））
また、デューティｔ‖（ ｎ‖を用いたときのデューティ）は、
ｔ‖＝（ｎ‖＊＊２−ｎ２＊＊２）／（ｎ１＊＊２−ｎ２＊＊２）
となる。この２値を次のように平均化してデューティｔの精度を高めている。
【００８４】
ｔ＝（ｔ⊥＋ｔ‖）／２
【００８５】
図２で光源部３０２と受光部３０１はアーチ３０６に固定されており、ドライバー３０３１が駆動部３０７を稼動してこのアーチ３０６の上を円弧状に動き、入射角を変えることができる。例えば測定対象が現像処理前の潜像である場合、デューティを精度高く測定できる最適な入射角はレジストによって差があり、未露光部分と被露光部分の屈折率差や、膜厚などによって最適な角度を選んでいる。また、測定対象が現像処理前の潜像である場合と、現像処理後の現像パターンである場合とでも最適入射角は大きく異なる為、コンピューター３０４からの司令により測定対象に応じた入射角に変更している。
【００８６】
一方、図７はサンプルのＬ＆Ｓを偏光解析によって得られたデューティとそのサンプルを現像し、ＳＥＭで測定した値を比較することにより最適露光条件を見出す方法の説明図である。これにより偏光解析法で発生する下地の構造等で発生するデューティのオフセットを計り、以降の測定ではこのオフセット値を差し引いた値を補正された正しい測定値とする。このＳＥＭとの比較はプロセス等の条件がかわる最初に一度だけすれば良く、以後はＳＥＭは必要としない。
【００８７】
次に図８、図９に示す露光条件の最適化のフローチャートを説明する。図８は１枚のウエハに８×６個のチップ露光をし、その偏光解析結果をΔ−ψマップ上に表した例である。例えば、ライン４１０上の測定点は露光量が一定でフォーカスが変化しているものであり、一方ライン４０１上の測定点は逆にフォーカスが一定で露光量が変化している。
【００８８】
このΔ−ψのマップ上の四角枠ＡＸ内に示されたショットは、前述したようなＳＥＭ測定との対応により最適なデューティになっている範囲を示している。従って露光ショットを偏光解析法により測定し、Δ、ψがこの枠ＡＸに入れば、最適デューティをしめしていることになる。
【００８９】
以上をフロー化したフローチャートが図９である。まず、ウエハにレジストをコートする。レジストの厚みがわかっていなければ、このとき測定する。次に前述したようにステージをステップ移動しながら、フォーカスと露光量（シャッター時間）を変えて試し焼きをする。入射角θを基板の複素屈折率ｎｓ、レジストの未露光部、被露光部の屈折率ｎ１、ｎ２、厚さｄ、波長λなどによって測定感度が高くなる最適な入射角に固定する。
【００９０】
次にウエハをウエハチャックから取り外すことなく、同様にステージを移動しながら偏光解析法で次々にショットを変えてΔ、ψを測定する。偏光解析結果が図８で示された所定の範囲のΔ、ψであれば、そのショットを焼き付けた露光条件を量産焼きの最適露光条件とする。
【００９１】
この偏光解析方法によるデューティチェックををウエハの量産焼付け工程内の途中で随時行い、歩留を上げている。本実施形態では以上のようにして最適な露光条件を設定した後にウエハを所定の現像処理工程を介して、これにより高集積度のデバイスを製造している。
【００９２】
ここまでは現像を行う前の潜像を測定対象とした場合を中心に説明したが、本実施例では入射角可変である為、レジストを現像した現像パターンの偏光解析を行うこともできる。図１３は本実施例に関わるレジスト現像後の感光パターンの断面形状の説明図である。図１３から明らかのように、潜像パターンの測定に比べて屈折率ｎ２が略１．０の気体である為、Ｌ＆Ｓの屈折率差が大きく取れるので、デューティの測定精度が有利であり、条件が厳しい場合、現像パターンを測定することが望ましい。
【００９３】
図１４は本実施例で現像パターンを測定する場合のフローチャートである。図１５は本実施例で現像パターンで測定する場合の説明図である。ここではレジストの現像用のデベロッパーとＳＥＭを更に同居させ、潜像ではなく現像で測定を行うとともに、ＳＥＭを偏光解析法によるパターン形状測定の校正用として現像したレジスト像を確認する必要が生じた時に使用している。図１４のフローチャートで測定を行う前に入射角θを基板の複素屈折率ｎｓ、レジストの屈折率ｎ１、厚さｄ、波長λなどによって測定感度が高くなる最適な入射角に固定してから測定を行っている。
【００９４】
（実施例２）
本発明の実施例２は図２において入射方向可変とする実施例である。入射方向とは、感光パターン面上のある方向に対して、入射光束光軸と感光パターンの法線によって定義される入射面がなす入射方位角である。図２においてウエハチャック１０４の下のＸＹステージの下に不図示の回転ステージを有し、ＸＹステージが測定対象の感光パターンを入射スポットに移動した後、レジストの下地、レジストの未露光部、被露光部の屈折率ｎ１、ｎ２、厚みｄ、露光したパターンの周期、角度、本数などに応じた最適な入射方向に回転ステージが感光パターンを光源部、受光部に対して回転する。こうして測定を感度良く行える最適な入射方向での偏光解析の測定を行っている。その他の構成は実施例１と同様である。
【００９５】
同様の効果はウエハのＸＹステージの下の回転ステージの替わりに図２のアーチ３０６を入射点を通るウエハの法線を軸に回転する駆動系を用いても得られる。
【００９６】
（実施例３）
本発明の実施例３は図２において入射波長可変とする実施例である。例えば光源部３０１の光源をハロゲンランプと分光器による構成にしたものである。潜像レジストパターンを測定対象とする場合、レジストの感光波長に近い短波長で測定を行うと未露光部が感光し、パターンが変化する可能性があるため、レジストが感度を持たない十分長い波長で測定している。一方現像レジストパターンを測定対象とする場合、現像処理後のパターンであるため、精度の高い測定を行うときには感度の高い短波長で測定を行っている。
【００９７】
また、レジストの下地、レジストの未露光部、被露光部の屈折率ｎ１、ｎ２、厚みｄ、露光したパターンの周期、角度、本数などに応じて最適な入射波長は異なるため、測定対象に適した最適な入射波長で測定することが望ましい。
【００９８】
その他の構成は実施例１と同様である。
【００９９】
同様の入射波長可変の効果はハロゲンランプからの光を分光器に通し、分光器の射出口から光ファイバーで光源部に光を導くことでも得られる。ハロゲンランプと分光器以外にもチューナブルレーザーや複数の半導体レーザーなどを導入することでも得られる。
【０１００】
（実施例４）
図１１は本発明の実施例４の説明図である。本実施例では実施例１に比べて露光装置上に物理的制約等がある場合に、偏光解析装置を露光装置のウエハステージ上とは別に設けている点が異なっているだけであり、その他の構成は同じである。図１２は、本実施例の変形例の説明図である。実施例４ではウエハ面上のレジストを現像しない潜像のままで偏光解析していたが、この変形例ではレジストを現像してから偏光解析を行っている。
【０１０１】
（実施例５）
本発明の実施例５は１つの感光パターンについて複数入射角で測定を行うものである。実施例１では予め測定してある未露光部と被露光部の屈折率ｎ１，ｎ２，レジスト厚みｄ、基板の複素屈折率ｎｓを用いて一つの最適入射角を合わせたのに対して、本実施例では複数の入射角で入射して光束の変化を検出し、複数の入射角での検出値を使ってより高精度に測定を行うことを特徴とする。
【０１０２】
本実施例によると予め測定してある未露光部と被露光部の屈折率ｎ１，ｎ２，レジスト厚みｄ、基板の複素屈折率ｎｓを用いて最適な入射角の近傍の複数の入射角を選び、これらの入射角で測定を行う。感光パターンのある形状を仮定すると所望の入射角で光束を入射したときの光束の変化を計算できる。測定した入射角のそれぞれに対して、あるデューティｔを仮定した計算による光束の変化と、実際に測定で得られた光束の変化を比較し、例えば３つの入射角θ１、θ２、θ３で測定している場合、測定で得られたＰ偏光、Ｓ偏光の比、ψ１、ψ２、ψ３、とＰ偏光、Ｓ偏光の位相差、Δ１、Δ２、Δ３、と計算で得られたＰ偏光、Ｓ偏光の比、ψ１’（ｔ）、ψ２’（ｔ）、ψ３’（ｔ）、Ｐ偏光、Ｓ偏光の位相差、Δ１’（ｔ）、Δ２’（ｔ）、Δ３’（ｔ）を用いた関数、
Ｄ（ｔ）＝（ψ１−ψ１’（ｔ））＋（ψ２−ψ２’（ｔ））＋（ψ３−ψ３’（ｔ））＋（Δ１−Δ１’（ｔ））＋（Δ２−Δ２’（ｔ））＋（Δ３−Δ３’（ｔ））
が最小に成るようにｔを決定することでより高精度にデューティｔを求めている。
【０１０３】
また、レジストは露光や現像によって膜厚が変化することが知られている為、予め測定した厚みｄで入射角の最適化をしても必ずしも最適な入射角とは限らない。膜厚によって偏光状態も変化する為、デューティの決定の場合は更に計算のパラメターとして厚みｄを加え、関数
Ｄ（ｔ）＝（ψ１−ψ１’（ｔ，ｄ））＋（ψ２−ψ２’（ｔ，ｄ））＋（ψ３−ψ３’（ｔ，ｄ））＋（Δ１−Δ１’（ｔ，ｄ））＋（Δ２−Δ２’（ｔ，ｄ））＋（Δ３−Δ３’（ｔ，ｄ））
が最小になるようにデューティｔと厚みｄを決定することで感光パターンの形状がより厳密に測定される。このように実際のレジストの形状を決定する場合、パラメターが多いため、複数の入射角による検出値で測定することが望ましい。その他の構成は実施例１と同様である。
【０１０４】
（実施例６）
本発明の実施例６は１つの感光パターンについて複数入射方向で測定を行うものである。入射方向とは、感光パターン面上のある方向に対して、入射光束光軸と感光パターンの法線によって定義される入射面がなす入射方位角である。本実施例は実施例２と同様に図２においてウエハチャック１０４の下のＸＹステージの下に不図示の回転ステージを有し、ＸＹステージが測定対象の感光パターンを入射スポットに移動した後、偏光解析の測定を行っている。
【０１０５】
パターンに対してＸの方向に入射して測定を行い、その後回転ステージを駆動して光の入射点を中心に感光パターンを９０°ＸＹ平面内で回転することで前の測定とは直交した入射方向による測定を行っている。
【０１０６】
本実施例は例えばレジストむらや下地の構造等による測定値の変動がある場合に対応したもので、ある方向からの偏光解析測定に対し、それに直交した入射方向により前記測定した測定点と略同一点を測定することにより以下のようにして精度向上を図っている。
【０１０７】
この場合も実施例１でｔを求めた時と同様に
ｔ（９０°）＝（ｔ⊥＋ｔ‖）／２
となり、実施例１で求めたｔ（０°）をつかって
ｔ＝（ｔ（０°）＋ｔ（９０°））／２
と平均することにより、デューティｔの測定精度を高める事もできる。
【０１０８】
又、パターンの細部に渡る形状を測定する場合は感光パターンの溝に平行な入射面とそれに直交する入射面の２方向のみでなく、その中間の入射方向で測定した値を解析することが要求される。
【０１０９】
同様の効果はウエハのＸＹステージの下の回転ステージの替わりに図２のアーチ３０６を入射点を通るウエハの法線を軸に回転する駆動系を用いても得られる。その他の構成は実施例１と同様である。
【０１１０】
（実施例７）
本発明の実施例７は１つの感光パターンについて複数波長で測定を行うものである。本実施例は実施例３と同様に図２において例えば光源部３０１の光源をハロゲンランプと分光器による構成にしたものである。
【０１１１】
実施例３では予め測定してあるレジストの下地、レジストの未露光部、被露光部の屈折率ｎ１、ｎ２、厚みｄ、露光したパターンの周期、角度、本数などの測定対象の値に応じた一つの最適な入射波長で測定を行ったのに対して、本実施例では複数の波長で入射して光束の変化を検出し、複数の波長での検出値を使ってより高精度に測定を行うことを特徴とする。
【０１１２】
本実施例によると予め測定してある、レジストの下地、レジストの未露光部、被露光部の屈折率ｎ１、ｎ２、厚みｄ、露光したパターンの周期、角度、本数などの測定対象の値を用いて最適な波長の近傍の複数の波長を選び、これらの波長で測定を行う。
【０１１３】
実施例５と同様に感光パターンのある形状を仮定すると所望の波長での光束を入射したときの光束の変化を計算できる。入射した波長のそれぞれに対して、あるデューティｔを仮定した計算による光束の変化と、実際に測定で得られた光束の変化を比較し、例えば３つの波長λ１、λ２、λ３で偏光状態の変化を測定している場合、測定で得られたＰ偏光、Ｓ偏光の比、ψ１、ψ２、ψ３、とＰ偏光、Ｓ偏光の位相差、Δ１、Δ２、Δ３、と計算で得られたＰ偏光、Ｓ偏光の比、ψ１’（ｔ）、ψ２’（ｔ）、ψ３’（ｔ）、Ｐ偏光、Ｓ偏光の位相差、Δ１’（ｔ）、Δ２’（ｔ）、Δ３’（ｔ）を用いた関数、
Ｄ（ｔ）＝（ψ１−ψ１’（ｔ））＋（ψ２−ψ２’（ｔ））＋（ψ３−ψ３’（ｔ））＋（Δ１−Δ１’（ｔ））＋（Δ２−Δ２’（ｔ））＋（Δ３−Δ３’（ｔ））
が最小に成るようにｔを決定することでより高精度にデューティｔを求めている。
【０１１４】
レジストは露光や現像によって膜厚が変化することが知られている為、パターンがない部位で予め測定したレジスト厚みｄでは計算が適当ではない場合がある。そのため本実施形態によれば、複数の波長によって測定することでデューティやレジスト形状の他にパターン位置でのレジスト膜厚も同時に測定するものである。
【０１１５】
また、レジストは露光や現像によって膜厚が変化することが知られている為、パターンがない部位で予め測定した厚みｄでの計算が適当ではない場合がある。そのため本実施形態によれば、複数の波長によって測定することでデューティやレジスト形状の他にパターン位置でのレジスト膜厚も同時に測定するものである。
【０１１６】
膜厚によって偏光状態が変化する為、上記のようなデューティの決定の場合は、更に計算のパラメターとして厚みｄを加え、関数
Ｄ（ｔ）＝（ψ１−ψ１’（ｔ，ｄ））＋（ψ２−ψ２’（ｔ，ｄ））＋（ψ３−ψ３’（ｔ，ｄ））＋（Δ１−Δ１’（ｔ，ｄ））＋（Δ２−Δ２’（ｔ，ｄ））＋（Δ３−Δ３’（ｔ，ｄ））
が最小になるようにデューティｔと厚みｄを決定することで感光パターンの形状がより厳密に測定される。このように実際のレジストの形状を決定する場合、パラメターが多いため、複数の波長による検出値で測定することが望ましい。
【０１１７】
同様の複数入射波長による測定はハロゲンランプからの光を分光器に通し、分光器の射出口から光ファイバーで光源部に光を導く構成でも得られる。また、チューナブルレーザーなどの複数波長光源あるいは異なる波長の光源を複数備えることでも同じ効果は得られる。その他の構成は実施例１と同様である。
【０１１８】
（実施例８）
本発明の実施例８は１つの感光パターンについて任意の入射条件で光源が入射光束を照射し、その光束の変化の検出値から、処理手段によって最適な入射条件を算出し、この最適な入射条件で光源が入射光束を照射し、感光パターンの形成状態を求めるものである。
【０１１９】
前述の通り、レジストは露光や現像処理により膜厚などが変動し、測定の誤差要因となる。そこで予め測定してある、レジストの下地の屈折率ｎｓ、レジストの未露光部、被露光部の屈折率ｎ１、ｎ２、厚みｄ、露光したパターンの周期、角度、本数などの測定対象の値を用いて計算した最適な入射条件で一度測定を行い、その測定結果と計算値の比較によって、大きく予備値と異なる数値（例えば膜厚）を補正して最適な入射条件を計算し直す。その新たな最適入射条件に光源を合わせ、光束を入射して変化を受光系で検出してレジストパターンを測定している。
【０１２０】
この入射条件とは入射角であり、入射方向であり、入射波長である。
【０１２１】
次に上記説明した露光方法を利用したデバイスの生産方法の実施例を説明する。
【０１２２】
図１６は微小デバイス（ＩＣやＬＳＩ等の半導体チップ、液晶パネル、ＣＣＤ、薄膜磁気ヘッド、マイクロマシン等）の製造のフローを示す。ステップ１（回路設計）では半導体デバイスの回路設計を行なう。ステップ２（マスク製作）では設計した回路パターンを形成したマスクを製作する。一方、ステップ３（ウエハ製造）ではシリコン等の材料を用いてウエハを製造する。ステップ４（ウエハプロセス）は前工程と呼ばれ、上記用意したマスクとウエハを用いて、リソグラフィ技術によってウエハ上に実際の回路を形成する。次のステップ５（組み立て）は後工程と呼ばれ、ステップ４によって作製されたウエハを用いて半導体チップ化する工程であり、アッセンブリ工程（ダイシング、ボンディング）、パッケージング工程（チップ封入）等の工程を含む。ステップ６（検査）ではステップ５で作製された半導体デバイスの動作確認テスト、耐久性テスト等の検査を行なう。こうした工程を経て半導体デバイスが完成し、これが出荷（ステップ７）される。
【０１２３】
図１７は上記ウエハプロセスの詳細なフローを示す。ステップ１１（酸化）ではウエハの表面を酸化させる。ステップ１２（ＣＶＤ）ではウエハ表面に絶縁膜を形成する。ステップ１３（電極形成）ではウエハ上に電極を蒸着によって形成する。ステップ１４（イオン打込み）ではウエハにイオンを打ち込む。ステップ１５（レジスト処理）ではウエハに感光剤を塗布する。ステップ１６（露光）では上記説明した露光装置によってマスクの回路パターンをウエハに焼付露光する。ステップ１７（現像）では露光したウエハを現像する。ステップ１８（エッチング）では現像したレジスト像以外の部分を削り取る。ステップ１９（レジスト剥離）ではエッチングが済んで不要となったレジストを取り除く。これらのステップを繰り返し行なうことによって、ウエハ上に多重に回路パターンが形成される。
【０１２４】
本実施例の製造方法を用いれば、従来は製造が難しかった高集積度の半導体デバイスを低コストに製造することができる。
【０１２５】
【発明の効果】
本発明によれば以上のように、露光によるレジストの感光状態（潜像）あるいは現像後のＬ＆Ｓ等の感光パターンの形成状態を入射光束の変化、例えば反射光の強度の変化や偏光状態の変化を利用して測定し、その測定値から最適な露光条件を決定し、その最適露光条件でウエハを量産露光していくことにより短時間で最適な露光条件を設定することができ、高集積度の投影パターンが容易に得られるパターン形成状態検出装置及びそれを用いた投影露光装置を達成することができる。
【０１２６】
特に従来のＳＥＭを用いる方法に比べて、周期性を持つパターン、例えばＬ＆Ｓパターンをレチクルに構成した露光条件測定用のレチクルを用いて、このパターンのレジスト像（現像する場合、もしくは現像しない潜像の場合）のデューティを反射光の情報を用いることにより最適露光条件を、高精度でしかも短時間で得られるという効果がある。
【図面の簡単な説明】
【図１】本発明の実施例１の要部断面図
【図２】図１の偏光解析装置一部分の説明図
【図３】露光装置における、フォーカス検出と露光量制御の一部分の概略図
【図４】露光条件測定用レチクルパターンの説明図
【図５】露光条件を振って焼いたウェハ上のレジスト潜像の説明図
【図６】図５のウエハのレジスト潜像の断面図
【図７】偏光解析装置による測定値のＳＥＭによる校正の説明図
【図８】Δ−ψマップを示す図
【図９】偏光解析法を露光条件設定に適用した場合のフローチャートの説明図
【図１０】偏光解析法の説明図
【図１１】偏光解析装置を別置きした実施例４の説明図
【図１２】偏光解析装置を別置きした実施例４の変形例の説明図
【図１３】レジスト現像後の断面形状の例を示す図
【図１４】現像工程をいれたフローチャートの説明図
【図１５】現像工程をいれる場合の実施例１の説明図
【図１６】デバイスの製造フローを現の説明図
【図１７】ウエハプロセスの説明図
【符号の説明】
１０１ 縮小型の投影レンズ
１０２ レチクル
１０３ ウエハ
１０４ ウエハチャック
１０５ 粗微動チルトＺステージ
１０７ 光源
１０８ フレーム
１０９ 中央処理装置
２０１ フォーカス検知用光源
２０２ フォーカス検知用光電変換器
２０３１ フォーカス制御装置
２０４１ 積算露光量検出回路
２０４２ 積算露光量制御回路
３０１ 偏光解析光源部
３０２ 偏光解析受光部
３０３ 回転ステージドライバー
３０４ 偏光解析処理装置[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a pattern formation state detection apparatus and a projection exposure apparatus using the same, for example, a process for manufacturing a semiconductor device such as an IC or LSI, an imaging device such as a CCD, a display device such as a liquid crystal panel, or a device such as a magnetic head. Among these, the exposure state of a projection exposure apparatus used in the lithography process is measured, and the optimum exposure conditions are determined in real time or quickly, which is suitable for exposure.
[0002]
[Prior art]
In recent years, higher integration of semiconductor devices such as ICs and LSIs has been accelerated, and the progress of microfabrication technology of semiconductor wafers accompanying this has been remarkable. As this fine processing technology, various reduction projection exposure apparatuses (steppers) that form circuit pattern images of a mask (reticle) on a photosensitive substrate by a projection optical system (projection lens) and expose the photosensitive substrate by a step-and-repeat method are proposed. Has been.
[0003]
In this stepper, the circuit pattern on the reticle is reduced and projected to a predetermined position on the wafer surface via a projection optical system having a predetermined reduction magnification, and after one projection transfer is completed, the wafer is The entire surface of the wafer is exposed by repeating the step of moving the stage placed by a predetermined amount and transferring again.
[0004]
In general, in order to transfer a fine circuit pattern using a stepper having a projection optical system, the exposure conditions such as the exposure amount on the wafer surface and the focus position of the wafer (position in the optical axis direction of the projection optical system) are appropriately set. Setting is important.
[0005]
For this reason, in a conventional stepper, printing is performed on a photosensitive substrate while changing at least one of an exposure condition, that is, a focus position and an exposure amount (shutter time) for each shot in a trial baking process (send head) before entering a mass production process. Thereafter, the photosensitive substrate is developed, and the line width of the pattern on the straight line is measured with an optical microscope or a line width measuring device to determine the optimum exposure conditions.
[0006]
For example, in the horizontal direction of the shot region array on the wafer, exposure is performed by changing the exposure amount (shutter time) by a constant amount while keeping the focus value constant, and in the vertical direction of the shot array, the exposure amount is fixed and focused. Exposure is performed by changing the value by a certain amount.
[0007]
Then, the line width of the resist pattern (L & S pattern) of the line (L) & space (S) in each shot formed after development is measured and detected by a scanning electron microscope (SEM). The optimum focus position and the optimum exposure amount are calculated.
[0008]
[Problems to be solved by the invention]
In order to set the optimum exposure conditions (exposure amount and focus position) in a conventional stepper, the line width of the resist pattern formed on the wafer is measured with an SEM or the like, so that there is a problem that processing time is required.
[0009]
In the present invention, the exposure state of the resist by exposure (latent image) or the formation state of the photosensitive pattern such as L & S after development changes in the incident light beam incident under a plurality of incident conditions, for example, the change in the intensity of reflected light and the polarization state. Measurement is performed using changes, the optimum exposure conditions are determined from the measured values, and the wafers are mass-produced and exposed under the optimum exposure conditions. It is an object of the present invention to provide a pattern formation state detection apparatus and a projection exposure apparatus using the pattern formation state detection apparatus that can easily obtain a projection pattern of a predetermined degree.
[0010]
[Means for Solving the Problems]
Pattern formation state detection apparatus of the present invention Is an apparatus for detecting the formation state of a photosensitive pattern formed by exposing a photosensitive member applied to an object to a periodic pattern via an optical system, and an irradiating means for irradiating the photosensitive pattern with a light beam, A light receiving means for receiving a light beam from the photosensitive pattern; a processing means for detecting a change in the polarization state of the light beam received by the light receiving means; and determining the formation state of the photosensitive pattern based on the detection result; and the illumination means. Incident condition changing means for changing at least one of the incident angle or wavelength or incident azimuth angle of the light beam irradiated to the photosensitive pattern; It is characterized by having.
[0047]
DETAILED DESCRIPTION OF THE INVENTION
First, the characteristics of the optimum exposure condition setting method when the pattern on the first object (reticle) surface is projected and exposed to the second object (wafer) in the present invention will be described.
[0048]
In the present invention, when a pattern is transferred to a resist,
(B) Focus position
(B) Exposure amount
This is based on the principle that the average value or effective value of the refractive index of the resist pattern changes with the change of.
[0049]
Therefore, in the present invention, a mask in which an L & S exposure condition measurement pattern having a period in one direction, for example, is formed on the reticle (R), and an image of the reference pattern is applied to the wafer (W) at least of the exposure amount and focus. Exposure is sequentially performed on the wafer while changing one of the conditions.
[0050]
Then, a plurality of latent images in the resist on the wafer formed by this exposure (images composed of portions where the refractive index has changed due to a chemical change or the like caused by exposure) or a concavo-convex pattern after development from the light beam incident means. The incident light beam is irradiated under the incident condition, and the signal light beam from the photosensitive pattern is received by the light receiving means. Detection of changes in incident light flux (change in polarization state or intensity of incident light flux, etc.) using a signal from the light receiving means, and detection of changes in multiple light fluxes corresponding to multiple incident conditions. Is determined by the processing means. Based on the signal from the processing means, the control means controls the exposure conditions on the wafer surface (such as the exposure amount and the position of the wafer in the optical axis direction).
[0051]
Next, a case where the change in the polarization state of the incident light beam is used as the change in the light beam will be described as an example.
[0052]
A light beam having a predetermined wavelength and a predetermined polarization state is incident on the resist at a predetermined incident angle with respect to the photosensitive pattern. Measures the polarization state of the light beam that is transmitted through the resist, reflected on the wafer substrate, and then combined with the light beam that has passed through the resist again and emitted directly from the resist surface. .
[0053]
It is known that the concavo-convex pattern phase type diffraction grating does not generate diffracted light at a wavelength longer than the pitch and has birefringence characteristics.
[0054]
In the embodiment of the present invention, the case where reflected light is detected is mainly shown. However, when the pitch of the resist pattern is larger than the wavelength, diffracted light is generated, and the same measurement can be performed with this diffracted light.
[0055]
Next, the ellipsometric method used in the present invention will be described.
[0056]
Now, it is assumed that the grating thickness is d, the duty ratio (ratio of the remaining resist portion to the period) is t, and laser light having a wavelength longer than the period is perpendicularly incident on the birefringent element. At this time, it is known that the refractive indexes n‖ and n⊥ at the periodic structure portion of the birefringent element are given by the following equations depending on whether the polarization state of incident light is parallel or perpendicular to the grooves of the grating. (Principle of optics III: by Born Wolff). Assuming that the effective refractive index for light parallel to the grooves of the grating is n‖ and the effective refractive index for perpendicular light is n⊥,
[0057]
[Outside 1]

Here, n1 and n2 are the refractive indexes of L (line portion) and S (space equivalent portion) of the lattice, respectively. In the case of a latent image of a resist, the line portion L and the space portion S are a resist and L are exposed resist. When the resist is developed, L is a resist and S is a gas such as air.
[0058]
The above refractive index formula is an approximate expression when the period of the periodic structure is sufficiently small with respect to the wavelength, and by calculating the electric field strictly, the refractive index other than the periodic structure can be calculated when the period is equal to the wavelength. .
[0059]
The ellipsometric model corresponds to measuring the birefringence of a birefringent medium of a predetermined thickness on the wafer substrate. In the ellipsometry, linearly polarized light having a P / S (P-polarized / S-polarized) phase difference of 0 and an amplitude ratio of 1 is incident on the wafer substrate at a predetermined angle θ, and the phase difference of the reflected light ( Δ) and the amplitude ratio (ψ) can be measured to obtain n‖ and n⊥. Then, using the above equation or a precisely solved equation based on the refractive indexes n1 and n2 and the resist thickness d of the line portion L and space portion S measured in advance and the values of n‖ and n⊥ obtained by the ellipsometry. Thus, the duty t is obtained. Further, since this ellipsometry is known, the details are omitted.
[0060]
Regarding the conditions of the incident light flux (incident angle, azimuth angle with respect to the pattern direction, wavelength), the optimum conditions differ depending on the refractive index of the resist, the film thickness, the pattern shape, and the like. Conditions differ between measurement of the developed resist pattern and measurement of the latent image resist pattern. Furthermore, in the case of a latent image, the difference in refractive index between the exposed and unexposed areas is small, so that minute changes that occur in the pattern shape due to minute changes in exposure conditions can be resolved and detected with minute differences in light flux. Is measured under an incident condition in which a change in the luminous flux is large depending on the exposure condition.
[0061]
Therefore, one embodiment of the present invention is an apparatus that can be incident under a plurality of incident conditions. As described above, the duty of the latent image pattern in the resist or the concavo-convex pattern after development is measured using the ellipsometry. ing.
[0062]
In one embodiment of the present invention, an optimum incident condition is calculated from a value measured under a certain incident condition, and the measurement is performed with the optimum incident condition to increase the resolution.
[0063]
In the ellipsometry, the same polarization state may be reproduced even with different resist film thickness and resist pattern shape depending on the incident conditions. For this reason, in one embodiment of the present invention, a change in reflected light is detected under a plurality of incident conditions, and a resist shape that has been overlapped under a certain incident condition is decomposed under another incident condition to obtain a measured value.
[0064]
In one embodiment of the present invention, an L & S pattern is transferred onto a photosensitive substrate under different exposure conditions to form a plurality of photosensitive patterns, and the plurality of photosensitive patterns are sequentially irradiated with the light flux and reflected. Detecting a polarization state of light, calculating a duty of a photosensitive pattern from the polarization state, determining an exposure condition where the duty is desired, and thereafter baking the wafer on a plurality of wafers under the exposure condition The light beam incident means is characterized in that the incident condition of the incident light beam is variable.
[0065]
(Example 1)
Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic diagram of a main part of Embodiment 1 of the present invention, and FIGS. 2 to 5 are explanatory views when a part of FIG. 1 is extracted.
[0066]
In this embodiment, a change in polarization state is used as a change in incident light flux through the photosensitive pattern.
[0067]
In FIG. 1, reference numeral 101 denotes a reduction projection lens, which projects a circuit pattern 102 a on the surface of a reticle 102 illuminated with exposure light from an exposure light source 107 described later onto a wafer 103. A wafer chuck 104 adsorbs the wafer 103. A coarse / fine movement stage 105 moves the wafer chuck 104 in the Z direction. Reference numeral 106 denotes an XY stage, which moves the wafer chuck 104 in the XY directions. Reference numeral 107 denotes a light source for exposure. Reference numeral 108 denotes a structure that supports the light source 107, reticle 102, lens barrel 101, wafer stage 106, and arch 306.
[0068]
Next, a focus position control device for detecting the position in the optical axis direction of the projection lens 101 of the wafer 103 used in this embodiment, that is, a focus position, and an exposure amount control device for illuminating the reticle 102 with exposure light, The control means having the above will be described with reference to FIG.
[0069]
In FIG. 3, reference numeral 201 denotes a high-intensity light source such as a semiconductor laser. The laser light emitted from the light source 201 is changed in direction by the bending mirror 208 and then incident on the surface of the wafer 103. After being reflected at the measurement point P <b> 1 of the wafer 103, the light beam is changed in direction by the bending mirror 209 and then incident on the detection element 202 that detects the two-dimensional position of the incident light beam. The detection element 202 is composed of, for example, a CCD and detects the incident position of the light beam. A change in position in the Z direction (in the optical axis direction of the projection lens 101) on the surface of the wafer 103 is detected so as to be a displacement of the incident position on the detection element 202. Based on the signal from the detection element 202, the focus control device 203 controls the position of the wafer 103 in the Z direction, that is, the focus position by the coarse / fine movement stage 105.
[0070]
Reference numeral 206 denotes a shutter opening / closing mechanism, 207 a half mirror, and 205 an illuminance sensor, which detect the exposure amount of the light beam from the light source 107 reflected by the half mirror 207. The integrated exposure control device 204 controls the integrated light quantity of the light beam from the light source 107 by controlling the shutter opening / closing mechanism 206 based on the signal from the illuminance sensor 205. As a result, the exposure amount for irradiating the reticle 102 is adjusted to a preset value.
[0071]
In the present embodiment, the exposure condition at the time of projecting the pattern on the reticle 102 surface onto the wafer 103 surface is controlled using the control means comprising such an exposure amount control device and a focus position control device.
[0072]
Next, a case where a pattern on the surface of the reticle 102 illuminated with exposure light from the light source 107 is projected onto the surface of the wafer 103 will be described.
[0073]
FIG. 4 is an explanatory diagram of a reference pattern (hereinafter referred to as “pattern”) 102 p on the surface of the reticle 102. The pattern 102p is configured by making L & S patterns 1021 and 1022 including lines (L) and spaces (S) orthogonal to each other.
[0074]
In this embodiment, the reticle 102 on which the pattern 102p composed of the L & S patterns 1021 and 1022 is drawn is set on the reticle stage 102a, the resist-coated wafer 103 is set on the wafer chuck 104, and the reticle 102 pattern 102p is stepped and repeated. In this manner, the wafer 103 is sequentially exposed. At this time, using the focus control device 203 and the integrated exposure control device 204 described above, an area 1031 that is one shot is sequentially printed as shown in FIG. 6, and the latent images 10312 and 10311 of the patterns 1021 and 1022 are printed.
[0075]
Then, according to the shot position in the X direction, the focus offset is changed step by step by a fixed amount around the predicted optimal position (predicted best focus position). The exposure is performed while changing the exposure amount around the time). In the example of FIG. 5, for convenience of explanation, a 3 × 3 matrix is used. However, the larger the number of shots, the easier the condition is set.
[0076]
FIG. 6 is an explanatory view of the resist cross section of the wafer 103 when sequentially exposed in this way. The resist on the wafer 103 after exposure forms a latent image in the resist as shown in FIG. The latent image is formed by changing the properties of the resist by chemical change or the like by exposure light. The hatched portion is the exposed portion, and the refractive index generally changes. The matrix addresses in FIG. 6 are cross sections at positions corresponding to those in FIG. 5, and (1), (2), and (3) represent cross sections of the pattern image 10312 of each chip.
[0077]
Next, the resist on the wafer 103 after exposure is composed of a light source unit 301, a light receiving unit 302, a driver 303, and a polarization processing unit 304 without removing the latent image from the wafer chuck 104 as shown in FIG. The ellipsometer 300 measures the amplitude ratio ψ and the phase difference Δ of the reflected light with respect to the incident light.
[0078]
FIG. 2 is a schematic diagram in which main parts of the ellipsometer 300 in FIG. 1 at this time are extracted. FIG. 10 is an explanatory diagram in which the optical path of FIG. 2 is developed. 2 and 10, a light source unit (light beam incident means) 301 includes a light source 3011 (which may be a monochromatic light of a spectrometer as well as a HeNe laser or a semiconductor laser at a wavelength equal to or greater than the L & S pitch on the wafer 103) and a polarizing element 3012. Thomson etc.). The polarizing element 3012 is installed with respect to the wafer 103 so that the P-polarization component (parallel to the paper surface) and the S-polarization component (perpendicular to the paper surface) are equal to each other with the paper surface and the polarization surface being 45 degrees. Therefore, the phase difference Δ between the light beams P and S is 0, and the amplitude ratio ψ is 1.
[0079]
The light receiving unit (light receiving unit) 302 includes a λ / 4 wavelength plate 3024 whose anisotropy axis is orthogonal to the light beam 305, a polarizing element 3022 (Gran Thompson, etc.) as an analyzer, and a photoelectric conversion element 3021. Yes. Further, the λ / 4 plate 3024 is held in a rotation mechanism 3023 having the light beam 305 direction as a rotation axis, and is rotated at a constant speed according to a command from the driver 303.
[0080]
Now, the emitted light 305 from the light source unit 301 is reflected by the resist surface on the wafer 103 and the wafer surface, and the combined light flux is P or S-polarized light depending on the birefringence indices n1, n2, etc. of the resist on the wafer 103. The phase difference Δ and the amplitude ratio ψ change.
[0081]
The light beam is detected by a detector 3021 through a rotating λ / 4 plate 3024 and an analyzer 3022, and a sine wave electric signal corresponding to the phase difference Δ and the amplitude ratio ψ is obtained. The above Δ and ψ are obtained from the phase information.
[0082]
Then, n‖ and n⊥ obtained from Δ and ψ measured by ellipsometry, the refractive index n1 of the unexposed portion of the resist, the refractive index n2 of the exposed portion, and the resist thickness d measured in advance. , And the complex refractive index ns of the substrate, the duty t⊥ (duty when n⊥ is used) is obtained by the following equation.
[0083]
t⊥ = n1 ** 2 · (n2 ** 2-n⊥ ** 2) / (n⊥ ** 2 · (n2 ** 2-n1 ** 2))
Also, the duty t‖ (duty when n‖ is used) is
t‖ = (n‖ ** 2-n2 ** 2) / (n1 ** 2-n2 ** 2)
It becomes. The accuracy of the duty t is increased by averaging these two values as follows.
[0084]
t = (t⊥ + t‖) / 2
[0085]
In FIG. 2, the light source unit 302 and the light receiving unit 301 are fixed to the arch 306, and the driver 3031 operates the driving unit 307 to move in an arc shape on the arch 306, thereby changing the incident angle. For example, when the object to be measured is a latent image before development processing, the optimum incident angle at which the duty can be measured with high accuracy varies depending on the resist, and the optimum angle depends on the difference in refractive index between the unexposed part and the exposed part, the film thickness, etc. Choose an angle. In addition, since the optimum incident angle differs greatly between the case where the measurement target is a latent image before development processing and the case where the development pattern is after development processing, the incident angle is changed according to the measurement target by a command from the computer 304. doing.
[0086]
On the other hand, FIG. 7 is an explanatory diagram of a method for finding the optimum exposure condition by comparing the L & S of the sample obtained by ellipsometry with the duty obtained by developing the sample and comparing the value measured by the SEM. As a result, the offset of the duty generated by the structure of the base generated by the ellipsometry is measured, and the value obtained by subtracting the offset value is corrected in the subsequent measurement. The comparison with the SEM needs to be performed only once at the beginning of the change of process conditions, and no SEM is required thereafter.
[0087]
Next, flowcharts for optimizing the exposure conditions shown in FIGS. 8 and 9 will be described. FIG. 8 shows an example in which 8 × 6 chip exposure is performed on one wafer and the polarization analysis result is displayed on the Δ-ψ map. For example, the measurement point on the line 410 has a constant exposure and the focus changes, while the measurement point on the line 401 has a constant focus and the exposure changes.
[0088]
The shots shown in the square frame AX on the map of Δ−ψ show the range where the optimum duty is obtained by the correspondence with the SEM measurement as described above. Therefore, when the exposure shot is measured by the ellipsometry, and Δ and ψ enter this frame AX, the optimum duty is shown.
[0089]
FIG. 9 is a flowchart showing the above flow. First, a resist is coated on the wafer. If the thickness of the resist is not known, it is measured at this time. Next, as described above, trial baking is performed while changing the focus and exposure amount (shutter time) while stepping the stage. The incident angle θ is fixed to an optimum incident angle at which the measurement sensitivity is increased depending on the complex refractive index ns of the substrate, the unexposed portions of the resist, the refractive indexes n1 and n2 of the exposed portion, the thickness d, the wavelength λ, and the like.
[0090]
Next, without removing the wafer from the wafer chuck, Δ and ψ are measured by changing shots one after another by ellipsometry while moving the stage in the same manner. If the result of the polarization analysis is Δ or ψ in the predetermined range shown in FIG. 8, the exposure condition for printing the shot is set as the optimum exposure condition for mass production printing.
[0091]
The duty check by this ellipsometry is performed at any time during the wafer mass production and baking process to increase the yield. In this embodiment, after setting the optimum exposure conditions as described above, the wafer is manufactured through a predetermined development processing step, thereby manufacturing a highly integrated device.
[0092]
Up to this point, the description has focused on the case where the latent image before development is used as the measurement object. However, in this embodiment, since the incident angle is variable, it is possible to perform polarization analysis of the development pattern in which the resist is developed. FIG. 13 is an explanatory diagram of the cross-sectional shape of the photosensitive pattern after resist development according to this embodiment. As apparent from FIG. 13, since the refractive index n2 is a gas having a refractive index of approximately 1.0 as compared with the measurement of the latent image pattern, a large difference in the refractive index of L & S can be obtained. Is difficult, it is desirable to measure the development pattern.
[0093]
FIG. 14 is a flowchart for measuring the development pattern in this embodiment. FIG. 15 is an explanatory diagram in the case of measuring with a development pattern in this embodiment. Here, the developer for developing the resist and the SEM were allowed to coexist, and it was necessary to check the developed resist image for calibration of pattern shape measurement by ellipsometry, while performing measurement by development instead of latent image. Sometimes used. Before the measurement in the flowchart of FIG. 14, the incident angle θ is measured after fixing the incident angle θ to an optimum incident angle that increases the measurement sensitivity depending on the complex refractive index ns of the substrate, the refractive index n1 of the resist, the thickness d, the wavelength λ, and the like. It is carried out.
[0094]
(Example 2)
The second embodiment of the present invention is an embodiment in which the incident direction is variable in FIG. The incident direction is an incident azimuth angle formed by the incident surface defined by the normal axis of the incident light beam optical axis and the photosensitive pattern with respect to a certain direction on the photosensitive pattern surface. In FIG. 2, a rotary stage (not shown) is provided under the XY stage below the wafer chuck 104, and after the XY stage moves the photosensitive pattern to be measured to the incident spot, the resist underlayer, the resist unexposed portion, The rotating stage rotates the photosensitive pattern with respect to the light source unit and the light receiving unit in an optimal incident direction according to the refractive indexes n1, n2, thickness d of the exposed part, the period, angle, number of the exposed patterns, and the like. In this way, the polarization analysis is measured in the optimum incident direction so that the measurement can be performed with high sensitivity. Other configurations are the same as those of the first embodiment.
[0095]
A similar effect can be obtained by using a drive system that rotates the arch 306 of FIG. 2 about the normal line of the wafer passing through the incident point instead of the rotary stage below the XY stage of the wafer.
[0096]
(Example 3)
The third embodiment of the present invention is an embodiment in which the incident wavelength is variable in FIG. For example, the light source of the light source unit 301 is configured by a halogen lamp and a spectroscope. When measuring a latent image resist pattern, measuring at a short wavelength close to the photosensitive wavelength of the resist may expose the unexposed areas and change the pattern. Therefore, the resist is not sensitive enough. It is measured by. On the other hand, when the development resist pattern is a measurement target, the pattern is a pattern after development processing. Therefore, when measuring with high accuracy, measurement is performed at a short wavelength with high sensitivity.
[0097]
In addition, the optimal incident wavelength differs depending on the resist base, the unexposed portion of the resist, the refractive index n1, n2, and the thickness d of the exposed portion, the period, angle, and number of the exposed patterns, so it is suitable for the measurement target. It is desirable to measure at the optimum incident wavelength.
[0098]
Other configurations are the same as those of the first embodiment.
[0099]
A similar effect of changing the incident wavelength can also be obtained by passing light from a halogen lamp through a spectroscope and guiding the light from the emission port of the spectroscope to the light source unit using an optical fiber. In addition to a halogen lamp and a spectroscope, it can also be obtained by introducing a tunable laser or a plurality of semiconductor lasers.
[0100]
(Example 4)
FIG. 11 is an explanatory diagram of Embodiment 4 of the present invention. This embodiment is different from the first embodiment in that there is a physical restriction on the exposure apparatus, except that the polarization analyzer is provided separately from the wafer stage of the exposure apparatus. The configuration is the same. FIG. 12 is an explanatory diagram of a modification of this embodiment. In the fourth embodiment, the polarization analysis is performed while the resist on the wafer surface is not developed, but the polarization analysis is performed after the resist is developed.
[0101]
(Example 5)
In Example 5 of the present invention, one photosensitive pattern is measured at a plurality of incident angles. In the first embodiment, one optimum incident angle is adjusted by using the refractive indexes n1, n2, resist thickness d of the unexposed portion and exposed portion measured in advance, and the complex refractive index ns of the substrate. The embodiment is characterized in that a change in a light beam is detected by entering at a plurality of incident angles, and measurement is performed with higher accuracy using detected values at a plurality of incident angles.
[0102]
According to the present embodiment, a plurality of incident angles in the vicinity of the optimum incident angle are selected using the refractive indexes n1, n2, resist thickness d of the unexposed portion and exposed portion that have been measured in advance, and the complex refractive index ns of the substrate. Measure at these incident angles. Assuming a certain shape of the photosensitive pattern, it is possible to calculate a change in the light beam when the light beam is incident at a desired incident angle. For each of the measured incident angles, the change in the light flux calculated by assuming a certain duty t is compared with the change in the light flux actually obtained by the measurement, for example, measured at three incident angles θ1, θ2, and θ3. P polarization, S polarization ratio obtained by measurement, ψ1, ψ2, ψ3 and P polarization, S polarization phase difference, Δ1, Δ2, Δ3, and P polarization obtained by calculation, S polarization Ratio, ψ1 ′ (t), ψ2 ′ (t), ψ3 ′ (t), phase difference between P-polarized light and S-polarized light, Δ1 ′ (t), Δ2 ′ (t), Δ3 ′ (t) were used. function,
D (t) = (ψ1-ψ1 ′ (t)) + (ψ2-ψ2 ′ (t)) + (ψ3-ψ3 ′ (t)) + (Δ1-Δ1 ′ (t)) + (Δ2-Δ2 ′) (T)) + (Δ3−Δ3 ′ (t))
By determining t so that is minimized, the duty t is obtained with higher accuracy.
[0103]
Further, since it is known that the film thickness of the resist changes due to exposure and development, even if the incident angle is optimized with the thickness d measured in advance, the optimum incident angle is not always obtained. Since the polarization state also changes depending on the film thickness, the thickness d is added as a calculation parameter when determining the duty, and the function
D (t) = (ψ1-ψ1 ′ (t, d)) + (ψ2-ψ2 ′ (t, d)) + (ψ3-ψ3 ′ (t, d)) + (Δ1-Δ1 ′ (t, d) )) + (Δ2−Δ2 ′ (t, d)) + (Δ3−Δ3 ′ (t, d))
By determining the duty t and the thickness d so as to be minimized, the shape of the photosensitive pattern can be measured more strictly. Thus, when determining the actual resist shape, since there are many parameters, it is desirable to measure with the detection value by several incident angles. Other configurations are the same as those of the first embodiment.
[0104]
(Example 6)
In Example 6 of the present invention, one photosensitive pattern is measured in a plurality of incident directions. The incident direction is an incident azimuth angle formed by the incident surface defined by the normal axis of the incident light beam optical axis and the photosensitive pattern with respect to a certain direction on the photosensitive pattern surface. As in the second embodiment, this embodiment has a rotating stage (not shown) under the XY stage below the wafer chuck 104 in FIG. 2, and after the XY stage moves the photosensitive pattern to be measured to the incident spot, the polarized light is polarized. Analytical measurement is performed.
[0105]
Incident in the X direction with respect to the pattern, and then the rotation stage is driven to rotate the photosensitive pattern in the 90 ° XY plane around the incident point of the light, so that the incident is orthogonal to the previous measurement. Measuring by direction.
[0106]
This example corresponds to the case where there are fluctuations in measured values due to, for example, resist unevenness or the structure of the base, and for the ellipsometric measurement from a certain direction, it is substantially the same as the measurement point measured in the incident direction orthogonal thereto. By measuring one point, the accuracy is improved as follows.
[0107]
In this case as well, when t is obtained in the first embodiment.
t (90 °) = (t⊥ + t‖) / 2
And using t (0 °) obtained in Example 1
t = (t (0 °) + t (90 °)) / 2
And the measurement accuracy of the duty t can be increased.
[0108]
In addition, when measuring the shape over the details of the pattern, it is required to analyze the measured values not only in the two directions of the incident surface parallel to the groove of the photosensitive pattern and the incident surface perpendicular to it, but also in the intermediate incident direction. Is done.
[0109]
A similar effect can be obtained by using a drive system that rotates the arch 306 of FIG. 2 about the normal line of the wafer passing through the incident point instead of the rotary stage below the XY stage of the wafer. Other configurations are the same as those of the first embodiment.
[0110]
(Example 7)
In the seventh embodiment of the present invention, one photosensitive pattern is measured at a plurality of wavelengths. In this embodiment, as in the third embodiment, in FIG. 2, for example, the light source of the light source unit 301 is configured by a halogen lamp and a spectroscope.
[0111]
In Example 3, the resist base, the unexposed portion of the resist, the refractive index n1, n2, and the thickness d of the exposed portion, the period of the exposed pattern, the angle, and the number of objects to be measured were measured in advance. In contrast to measuring at one optimal incident wavelength, this example detects changes in the light flux incident at multiple wavelengths, and uses the detection values at multiple wavelengths to measure with higher accuracy. It is characterized by performing.
[0112]
According to the present embodiment, the values of the measurement target such as the resist base, the unexposed portion of the resist, the refractive indexes n1 and n2 of the exposed portion, the thickness d, the period of the exposed pattern, the angle, and the number, which are measured in advance, are measured. A plurality of wavelengths in the vicinity of the optimum wavelength are selected and measured at these wavelengths.
[0113]
As in the fifth embodiment, assuming a shape having a photosensitive pattern, a change in the luminous flux when a luminous flux having a desired wavelength is incident can be calculated. For each incident wavelength, the change in the light flux calculated by assuming a certain duty t is compared with the change in the light flux actually obtained by measurement. For example, the change in the polarization state at three wavelengths λ1, λ2, and λ3 , The ratio of P-polarized light and S-polarized light obtained by measurement, ψ1, ψ2, ψ3, and P-polarized light, the phase difference of S-polarized light, Δ1, Δ2, Δ3, and P-polarized light obtained by calculation , S polarization ratio, ψ1 ′ (t), ψ2 ′ (t), ψ3 ′ (t), P polarization, S polarization phase difference, Δ1 ′ (t), Δ2 ′ (t), Δ3 ′ (t) A function using
D (t) = (ψ1-ψ1 ′ (t)) + (ψ2-ψ2 ′ (t)) + (ψ3-ψ3 ′ (t)) + (Δ1-Δ1 ′ (t)) + (Δ2-Δ2 ′) (T)) + (Δ3−Δ3 ′ (t))
By determining t so that is minimized, the duty t is obtained with higher accuracy.
[0114]
Since it is known that the thickness of a resist changes due to exposure or development, the calculation may not be appropriate for the resist thickness d measured in advance in a portion where there is no pattern. Therefore, according to this embodiment, the resist film thickness at the pattern position is simultaneously measured in addition to the duty and the resist shape by measuring with a plurality of wavelengths.
[0115]
In addition, since it is known that the film thickness of a resist is changed by exposure or development, there is a case where calculation with a thickness d measured in advance in a portion where there is no pattern is not appropriate. Therefore, according to this embodiment, the resist film thickness at the pattern position is simultaneously measured in addition to the duty and the resist shape by measuring with a plurality of wavelengths.
[0116]
Since the polarization state changes depending on the film thickness, when determining the duty as described above, the thickness d is further added as a calculation parameter, and the function
D (t) = (ψ1-ψ1 ′ (t, d)) + (ψ2-ψ2 ′ (t, d)) + (ψ3-ψ3 ′ (t, d)) + (Δ1-Δ1 ′ (t, d) )) + (Δ2−Δ2 ′ (t, d)) + (Δ3−Δ3 ′ (t, d))
By determining the duty t and the thickness d so as to be minimized, the shape of the photosensitive pattern can be measured more strictly. Thus, when determining the actual resist shape, since there are many parameters, it is desirable to measure with detection values by a plurality of wavelengths.
[0117]
The same measurement with a plurality of incident wavelengths can be obtained by a configuration in which light from a halogen lamp is passed through a spectroscope and light is guided from an emission port of the spectroscope to a light source unit through an optical fiber. The same effect can be obtained by providing a plurality of light sources having different wavelengths such as a tunable laser or a plurality of light sources having different wavelengths. Other configurations are the same as those of the first embodiment.
[0118]
(Example 8)
In the eighth embodiment of the present invention, a light source irradiates an incident light beam under an arbitrary incident condition with respect to one photosensitive pattern, and the optimum incident condition is calculated by the processing means from the detected value of the change in the light beam. The light source irradiates the incident light flux to determine the formation state of the photosensitive pattern.
[0119]
As described above, the film thickness of the resist varies due to exposure and development processing, which causes measurement errors. Therefore, the values of the measurement target such as the refractive index ns of the resist base, the unexposed portion of the resist, the refractive indexes n1 and n2 of the exposed portion, the thickness d, the period of the exposed pattern, the angle, and the number, which have been measured in advance, are measured. The measurement is performed once under the optimal incident conditions calculated using the above, and by comparing the measurement result with the calculated value, a numerical value (for example, film thickness) largely different from the preliminary value is corrected to recalculate the optimal incident condition. The resist pattern is measured by aligning the light source with the new optimum incidence condition, entering a light beam, detecting the change with a light receiving system.
[0120]
This incident condition is an incident angle, an incident direction, and an incident wavelength.
[0121]
Next, an embodiment of a device production method using the above-described exposure method will be described.
[0122]
FIG. 16 shows a flow of manufacturing a microdevice (a semiconductor chip such as an IC or LSI, a liquid crystal panel, a CCD, a thin film magnetic head, a micromachine, etc.). In step 1 (circuit design), a semiconductor device circuit is designed. In step 2 (mask production), a mask on which the designed circuit pattern is formed is produced. On the other hand, in step 3 (wafer manufacture), a wafer is manufactured using a material such as silicon. Step 4 (wafer process) is called a pre-process, and an actual circuit is formed on the wafer by lithography using the prepared mask and wafer. The next step 5 (assembly) is referred to as a post-process, and is a process for forming a semiconductor chip using the wafer produced in step 4, such as an assembly process (dicing, bonding), a packaging process (chip encapsulation), and the like. including. In step 6 (inspection), inspections such as an operation confirmation test and a durability test of the semiconductor device manufactured in step 5 are performed. Through these steps, the semiconductor device is completed and shipped (step 7).
[0123]
FIG. 17 shows a detailed flow of the wafer process. In step 11 (oxidation), the wafer surface is oxidized. In step 12 (CVD), an insulating film is formed on the wafer surface. In step 13 (electrode formation), an electrode is formed on the wafer by vapor deposition. In step 14 (ion implantation), ions are implanted into the wafer. In step 15 (resist process), a photosensitive agent is applied to the wafer. In step 16 (exposure), the circuit pattern of the mask is printed on the wafer by exposure using the exposure apparatus described above. In step 17 (development), the exposed wafer is developed. In step 18 (etching), portions other than the developed resist image are removed. In step 19 (resist stripping), unnecessary resist after etching is removed. By repeating these steps, multiple circuit patterns are formed on the wafer.
[0124]
By using the manufacturing method of this embodiment, it is possible to manufacture a highly integrated semiconductor device that has been difficult to manufacture at low cost.
[0125]
【The invention's effect】
According to the present invention, as described above, the resist exposure state (latent image) by exposure or the formation state of a photosensitive pattern such as L & S after development changes the incident light flux, for example, the change in reflected light intensity or the change in polarization state. The optimum exposure conditions can be set in a short time by determining the optimum exposure conditions from the measured values and performing mass-production exposure of the wafers under the optimum exposure conditions. A pattern formation state detection apparatus and a projection exposure apparatus using the same can be achieved.
[0126]
In particular, compared with a method using a conventional SEM, a resist image (a latent image that is developed or not developed) of a pattern having a periodicity, for example, a reticle for measuring exposure conditions in which an L & S pattern is configured on a reticle. In this case, the optimum exposure condition can be obtained with high accuracy and in a short time by using the reflected light information.
[Brief description of the drawings]
FIG. 1 is a sectional view of an essential part of Embodiment 1 of the present invention.
FIG. 2 is an explanatory diagram of a part of the ellipsometer of FIG.
FIG. 3 is a schematic view of a part of focus detection and exposure amount control in the exposure apparatus.
FIG. 4 is an explanatory diagram of a reticle pattern for measuring exposure conditions.
FIG. 5 is an explanatory diagram of a latent resist image on a wafer baked under different exposure conditions.
6 is a cross-sectional view of a latent resist image on the wafer of FIG. 5;
FIG. 7 is an explanatory diagram of calibration by SEM of measured values by an ellipsometer.
FIG. 8 is a diagram showing a Δ-ψ map;
FIG. 9 is an explanatory diagram of a flowchart when the ellipsometry is applied to the exposure condition setting.
FIG. 10 is an explanatory diagram of ellipsometry
FIG. 11 is an explanatory diagram of Example 4 in which an ellipsometer is installed separately.
FIG. 12 is an explanatory diagram of a modification of the fourth embodiment in which the ellipsometer is placed separately.
FIG. 13 shows an example of a cross-sectional shape after resist development.
FIG. 14 is an explanatory diagram of a flowchart including a developing process.
FIG. 15 is an explanatory diagram of Example 1 when a development process is performed
FIG. 16 is a diagram for explaining the device manufacturing flow.
FIG. 17 is an explanatory diagram of a wafer process.
[Explanation of symbols]
101 Reduction type projection lens
102 reticle
103 wafers
104 Wafer chuck
105 Coarse / Fine Tilt Z Stage
107 Light source
108 frames
109 Central processing unit
201 Light source for focus detection
202 Photoelectric converter for focus detection
2031 Focus control device
2041 Integrated exposure amount detection circuit
2042 Integrated exposure amount control circuit
301 Ellipsometric light source
302 Ellipsometric light receiving unit
303 Rotating stage driver
304 ellipsometry processor

Claims (10)

In a device for detecting the formation state of a photosensitive pattern formed by exposing a periodic pattern to a photoreceptor applied to an object via an optical system,
Irradiating means for irradiating the photosensitive pattern with a light beam ;
And a light-receiving means for receiving the light beam from the photosensitive pattern,
Processing means for detecting a change in the polarization state of the light beam received by the light receiving means, and determining the formation state of the photosensitive pattern based on the detection result ;
Incident condition changing means for changing at least one of an incident angle or a wavelength or an incident azimuth angle of the light beam applied to the photosensitive pattern by the illumination means;
The pattern formation state detection apparatus characterized by having. The pattern according to claim 1 , wherein the processing unit has a function of determining an incident condition based on a detection result of the change in the polarization state obtained by the irradiation unit irradiating a light beam under an arbitrary incident condition. Formation state detection device. 3. The pattern formation state detection apparatus according to claim 1, wherein the formation state of the photosensitive pattern is a duty of the photosensitive pattern. 7. The pattern formation state detection apparatus according to claim 1, wherein the formation state of the photosensitive pattern is a cross-sectional shape of the photosensitive pattern. 5. The pattern formation state detection apparatus according to claim 1 , wherein the change in the polarization state is a change in the polarization state of the light beam received by the light receiving unit with respect to the light beam applied to the pattern. The photosensitive pattern patterned state detecting apparatus of claims 1 to 5, wherein it is a latent image pattern before the development processing. The photosensitive pattern patterned state detecting apparatus of claims 1 to 5, wherein it is a developing pattern after development. 8. A projection exposure apparatus for projecting and exposing a pattern on a first object surface illuminated with exposure light onto a second object surface coated with a photosensitive member by a projection optical system to form a photosensitive pattern. A projection exposure apparatus comprising the pattern formation state detection apparatus. 9. An exposure method comprising: projecting and exposing a pattern on a first object surface onto a second object surface coated with a photoreceptor by a projection optical system using the projection exposure apparatus according to claim 8. A device manufacturing method, wherein a device is manufactured using the exposure method according to claim 9.

Images