JP4497569B2 - Evaluation method of coma aberration of projection optical system - Google Patents

Description

ここで
PL＝P1＋P2＋P3
PR＝P4＋P5＋P6
とする。強度差I def は、上記 強度IL, IRの差分なので
I def ＝IL −IR
と定義できる。
【００１９】
次に強度差I def の実部のみの数値積分を求めると位相PL、PRの実部Re(PL)、 Re(PR) は以下の様になる。なお、虚部の数値積分の結果も同じとなるので、以下説明は実部のみで行う。
【００２０】
Re(PL)＝cos(ωt) +cos(ωt ＋H) +cos(ωt ＋H ＋CM )
Re(PR)＝cos(ωt ＋H)+ cos(ωt ＋H ＋CM ) + cos( ωt ＋CM )
以上の式を用いて数値積分を行い、強度差I def とクロムの高さH との関係を解析する。
【００２１】
図３は縦軸に強度差I def 、横軸にクロムの高さH を取り、投影光学系３のコマ収差CMの量をλ/10 とλ/20 の二条件とした時の計算結果である。計算ではSL1 〜SL6 の散乱光の強度は全て等しいとした。また、横軸のクロムの高さH は中心露光波長λで正規化して表現してある。
【００２２】
図３のグラフよりコマ収差CMの量をλ/10 とλ/20 とした両方の曲線から、
１）I def はサイン関数的な変化を行う。
２）Idefの変化の周期はλである。
３）I def の振幅はコマ収差CMの発生量に比例する。等のことがわかる。
【００２３】
また同じく図３のグラフより、強度差I def が最大となるクロムの高さH はコマ収差CMの発生量によらず
H ＝ (2m−1)・λ/4 （m は正の整数） …(A)
の関係を満足する、λ/4、3 λ/4、、、、であることが分かる。
【００２４】
しかしながら(A) 式の条件は、照明σが大きい場合についてであり、照明σが小さくなった場合には更に限定が必要となる。
【００２５】
コヒーレンシーを表わすσの大小の目安は、図4 、図5 に示す様な光学像の光量分布のサイドローブの出方によって異なるが、大雑把には0.5 近傍の値が境界となる。より細かく定義すれば光学像の光量分布に細かいサイドローブが出ればσが小さい条件、サイドローブが観察できなければσが大きいとしてよい。
【００２６】
照明系のコヒーレンシーによる結像状態の違いを示したのが図4 及び図5 で、図４はσが大きい場合、図5 はσが小さい場合の光学像強度を示す。図４の左右の出力差は、前述の様にCr厚が波長の半分の整数倍の時、対称性を持つ様になる。図５の差分a −b についてもCr厚が波長の半分の整数倍の時、対称性を持つ様になるのは同様である。
【００２７】
一方、照明のσが小さくなると図5 に示す様にエッジ部の外側にもc 、d で示したサイドローブによるサブピークが発生する。コマ収差による非対称性に関連した差分c −d の挙動 は前述の差分a −b の挙動と異なる事が実験、シミュレーション双方から判明した。
【００２８】
Cr厚H に対する像の非対称性パラメータである差分a −b 、c −d の関係を示したのが図6 である。図6 では縦軸に任意ユニットで光の強度分布の非対称性SOI を、横軸にCr厚H を取ってある。 差分a −b は周期がλとなるサイン波であるが、差分c −d は周期λは持っているもののsin ( πH/λ) の自乗の関数であることが判る。
【００２９】
また内側に発生するサブピークの差分e −f も非対称性を表わすパラメータとなるが、差分c −d と正負が反対で、周期λではあるが、 sin (πH/λ) の自乗の関数となる。
【００３０】
図6 より差分a −b で示される像の非対称性はCr厚H が波長の4 分1 の奇数倍の時に最悪となるが、 差分c −d で表わされるの像の非対称性は、H が波長の半分の奇数倍でしか最悪とならない。従って、 照明σが小さい場合にコマ収差を評価するには、Cr厚H を波長の半分の奇数倍とする必要があることが判る。
【００３１】
次いで本発明に係るレチクルを使用して投影光学系の収差の評価を行う手順について説明する。高解像力を達成するためには、投影光学系の収差を正確に評価することが必須要件である。従って、投影光学系の評価では条件に応じてレチクルを使い分ける必要がある。
【００３２】
例えば現行で照明σを小さくして使用する条件としては、ハーフトーンレチクルを使用してコンタクトホールを形成する露光するケースが挙げられる。ハーフトーンの材質の屈折率をNd 、厚さをT とすると、ハーフトーンレチクルでは膜透過部と非透過部の位相を180 度ずらすため、T は
(Nd−1)T ＝n λ/2 ： nは整数
という条件を満足する必要がある。
【００３３】
現在ハーフトーンレチクルに使用されている多くの材質の屈折率はほぼ2 に近いものがほとんどなので、一番厚みが薄い条件を考えると、上記式の整数n ＝１である。従って
(2−1)T ＝１＊λ/2
より
T ＝λ/2
という厚さに対する条件を求めることができる。しかしながらT ＝λ/2という条件は波長の半分の整数倍なので、 差分c −d という像の非対称性パラメータが最大となる条件になる。言い換えれば現行のハーフトーンレチクルの条件は、照明σを小さくした時に像の対称性が最悪になる条件であり、コマ収差の評価には最も好ましい条件と言える。
【００３４】
なお、ここまではコマ収差の評価について述べてきたが、投影光学系のコマ収差は設計値と製造誤差の両者から発生する。本提案は両者を精度よく評価できるが、現状のリソグラフィーに使用される投影光学系においては後者の製造誤差分が殆どで、レンズ単体の誤差、偏心誤差等でコマ収差を発生させている。従って、上記レチクルを用いて評価を行った後に、コマ収差が大きく更に調整が必要であると判断された場合には、投影光学系は調整工程に入る。
【００３５】
差分a −b で表わされる像の非対称性は、照明系の非対称性によっても同様にサイン関数的に変化する。照明系の対称性とは、例えばtelecentricityと呼ばれてるフォーカス変化に伴う倍率変化等を意味する。従って、投影光学系のコマ収差を正確に評価するには照明系の対称性が必要である。
【００３６】
照明系の対称性が保証されていないと、差分a −b で表わされる非対称性は投影光学系のコマ収差と照明系の非対称性のミックスした評価量となる。例えば光量分布に発生する非対称性がコマ収差と照明系で同じ大きさであるが逆向きの貢献度を持つ場合は、差分a −b で表わされる非対称性はキャンセルして観測できない。
【００３７】
一方、差分c −d で表わされる像の非対称性はコマ収差のみで発生し、照明系の状態にはよらないことが分かっている。
【００３８】
なお、前述の様にコヒーレンシーを表わすσの大小の決定は光学像の光量分布に図5 に示したようなサイドローブが出るか否かによって判断される。σの値としては、大雑把にいって0.5 近傍の値が一つの指標である。光学像の光量分布に細かいサイドローブが出て差分c −d で表わされる像の非対称性を定義できる場合はσが小さい条件、図4 の様にサイドローブが観察できず差分c −d で表わされる像の非対称性が定義できなければσが大と考える。
【００３９】
図7，図8本発明に係るバイナリレチクルの適用形態を示したものである。図7は本発明を二層クロムレチクルに適用した実施形態、図8は本発明を三層クロムレチクルに適用した実施形態である。図9は本発明の参考例のレチクルであり、ハーフトーンレチクルに適用した例である。図7〜図9において、1が石英ガラス基板、2がクロム、10、11が酸化クロム、12がハーフトーン遮光部材、Hがクロムまたは遮光部材の高さを示している。
【００４０】
次に、本発明の別の実施形態として露光光にEUV を用いた投影光学系と、該投影光学系に使用する反射型レチクルについて説明する。EUV を用いた投影光学系としては例えばTakeo Watanabe,Kiyoto Mashima,Masahito Niibe and Hiroo Kinoshita：Microprocesses and Nanotechnology ‘97：Jpn. J. Appl. Phys. Vol. 36 (1997) pp. 7597-7600をあげることができる。該文献に示されるように、EUV 露光系では反射型レチクル面上のパターンが投影光学系によりウエハー上に投影露光される。
【００４１】
図10は吸収体の高さH の反射型レチクル（反射型のバイナリレチクル）MSを使用した本発明の露光装置の模式図で、M0 、M1、M2、M3は反射型投影光学系を構成するミラー、WFはウエハーである。また、図11は石英ガラス基板１を用いた露光光が反射の有無の２値のみから成るバイナリレチクルMSの拡大図で、反射層6を介して高さH の吸収体13を構成したバイナリレチクルMS上のパターンが投影光学系３により、ウエハーＷＦの上に露光される。
反射型レチクルMSにおける吸収体の高さＨは、反射効果で光路長差が透過型より2 倍大きくなるため、中心露光波長に対する吸収体の高さＨの基本量は透過型の場合の1/２倍となる。よって、中心露光波長λに対する吸収体の高さＨは、
反射型のレチクルを用い、σが大きい場合
H ＝ (2N−1)・λ/8 （N は正の整数）
反射型のレチクルを用い、σが小さい場合
H ＝ (2N−1)・λ/4 （N は正の整数）
となる。反射型レチクルの場合は右辺の基本量が透過型レチクルの1/2 、即ち透過型のλ/4に対応する量がλ/8、透過型のλ/2に対応する量がλ/4となっていることが特徴である。
【００４２】
EUV の露光波長を13nmとすると、σが大きい場合のＨは1.625nm の奇数倍、また、σが小さい場合はＨを3.25nmの奇数倍とすれば、投影光学系のコマ収差による非対称性が最も大きく現われ、光学系の収差状態を正確に把握することができる。
【００４３】
なお、反射系においてもコヒーレンシーを表わすσの大小の決定が像の光量分布に図5 に示したようなサイドローブが出るか否かによって判断されることは同様である。σの値は大雑把にいうと0.5 近傍の値が一つの指標である。像の光量分布に細かいサイドローブが出て差分c −d で表わされる像の非対称性を定義できる場合はσは小さい条件であり、図4 の様にサイドローブが観察できず差分c −d で表わされる像の非対称性が定義できなければσは大として考える。
【００４４】
また、表１は以上の結果を、各種露光波長に対して表にしたものである。
【００４５】
【表１】

【００４６】
【発明の効果】
以上説明したように、本発明の投影光学系のコマ収差の評価方法においては、中心露光波長λに従って決定されるレチクルパターンの高さＨを投影光学系に残存するコマ収差の影響が最大となる条件とすることで、投影光学系に残存する微小なコマ収差を評価することを可能とした。該評価結果に基づいて投影光学系を調整すれば、投影光学系の完成度を上げることができ、グレードの高い高解像力光学系を実現できる。
【００４７】
また特にコマ収差が最大となる条件で評価すれば、実際のレチクルパターンの高さ（厚さ）Ｈが変化しても、コマ収差の影響がそれ以上に大きくなることが無く、像の最悪の状態を評価してあるため所望の解像性能を得ることが可能である。
【図面の簡単な説明】
【図１】 本発明のレチクルのクロム部での散乱光の説明図、
【図２】 半導体露光装置の構成図、
【図３】 信号の強度差とクロム高さの関係を数値積分で求めたグラフ
【図４】 照明σが大きい場合の像のグラフ、
【図５】 照明σが小さい場合の像のグラフ、
【図６】 信号の強度差とクロム高さの関係を、数値積分で求めたグラフ
【図７】 本発明を二層クロムレチクルに適用した実施形態、
【図８】 本発明を三層クロムレチクルに適用した実施形態、
【図９】 本発明の参考例のハーフトーンレチクルに適用した例、
【図１０】 本発明の反射型レチクルを、露光光にＥＵＶを用いた半導体露光装置に配置してウエハー上にICパターンを転写する実施例を説明する図
【図１１】 レチクルＭＳの拡大図
【符号の説明】
1 石英ガラス基板
2 クロム
3 投影光学系
4 ウエハー、
6 反射層、
10 酸化クロム、
11 酸化クロム、
12 ハーフトーン遮光部材、
13 吸収体、
21 ウエハーチャック、
22 Z 駆動ステージ、
23 X 駆動ステージ、
24 Y 駆動テージ、
SL1 〜SL6 レチクルのクロムパターン部からの散乱光、
H クロムまたは遮光部材の高さ[0001]
BACKGROUND OF THE INVENTION
BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for evaluating coma aberration of a projection optical system . In particular, in an exposure apparatus for manufacturing a semiconductor that optically transfers a pattern, the projection optical system is formed on an original plate surface such as a reticle or mask that is a first object. IC, LSI, a fine electronic circuit pattern such as VLSI, is suitable to be that eXPOSURE aPPARATUS used to transfer on a wafer as the second object.
[0002]
[Prior art]
2. Description of the Related Art A projection exposure apparatus for manufacturing a semiconductor is required to project and expose a circuit pattern on a reticle or mask surface on a wafer surface with a high resolving power as the integrated circuit has a higher density. As a method for improving the projection resolving power of the circuit pattern, for example, a method of increasing the NA of the projection optical system by fixing the wavelength of the exposure light, or a method of shortening the exposure light to a shorter wavelength, for example, from g-line to i-line, i-line Therefore, a method using an excimer laser oscillation wavelength, and a method using EUV shorter than the excimer laser oscillation wavelength is being pursued.
[0003]
In the following description, the reticle and the mask are synonymous and will be described as being unified with the reticle.
[0004]
[Problems to be solved by the invention]
Recently, when projection exposure is performed on a wafer surface with high resolving power, it has been required to reduce the amount of aberration of a projection optical system that projects a reticle pattern onto the wafer. One of the particularly problematic aberrations is coma, and in increasing NA, it is necessary to reduce the remaining amount of coma from both the design and manufacturing aspects.
[0005]
However, in the process of achieving the above requirements, in the projection optical system and the illumination system, 1) the complexity and size increase due to the increase in the number of optical system components, and 2) the lengthening of manufacturing to reduce manufacturing errors. 3) There is a problem of high cost due to the above 1 and 2.
[0006]
Further, at the time of manufacturing an exposure apparatus, a method for reliably evaluating even a minute coma is desired. The present invention has been made in view of the above problems, and an object thereof is to provide a method for evaluating a coma aberration remaining in a projection optical system in an exposure apparatus.
[0007]
[Means for Solving the Problems]
The evaluation method for coma aberration of the projection optical system according to the invention of claim 1 is:
In the evaluation method of coma aberration of the projection optical system,
Preparing a transparent binary reticle;
Illuminating the binary reticle under illumination conditions where σ is less than 0.5 using a luminous flux with a center wavelength λ;
Projecting an image of the pattern of the binary reticle with the projection optical system;
Evaluating the coma aberration of the projection optical system based on the asymmetry of the sub-peak due to the side lobe outside the edge portion of the light amount distribution of the image, and
The height of the pattern is
(2N-1) · λ / 2 (N is a positive integer)
It is characterized by being.
The evaluation method of coma aberration of the projection optical system according to the invention of claim 2 is:
In the evaluation method of coma aberration of the projection optical system,
Preparing a reflective binary reticle;
Illuminating the binary reticle under illumination conditions where σ is less than 0.5 using a luminous flux with a center wavelength λ;
Projecting an image of the pattern of the binary reticle with the projection optical system;
Evaluating the coma aberration of the projection optical system based on the asymmetry of the sub-peak due to the side lobe outside the edge portion of the light amount distribution of the image, and
The height of the pattern is
(2N-1) · λ / 4 (N is a positive integer)
It is characterized by being.
[0008]
DETAILED DESCRIPTION OF THE INVENTION
The method for evaluating the coma aberration of the projection optical system according to the present invention is made possible by using a reticle configured with a pattern having a predetermined height (thickness) H with respect to the exposure center wavelength λ.
[0009]
The amount H of the predetermined height is
When σ, which is the coherency of the illumination system, is large with a transmissive reticle
H = (2N – 1) · λ / 4 (N is a positive integer)
When σ is small with a transmissive reticle
H = (2N – 1) · λ / 2 (N is a positive integer)
Reflective reticle with large σ
H = (2N – 1) · λ / 8 (N is a positive integer)
Reflective reticle with small σ
H = (2N – 1) · λ / 4 (N is a positive integer)
This condition satisfies or is in the vicinity of the condition.
[0010]
The approximate value of σ representing coherency varies depending on how the side lobes of the light quantity distribution of the image described later appear, but roughly speaking, the value near 0.5 is the boundary. If more finely defined, a condition where σ is small if a fine side lobe appears in the light amount distribution of the image, and a condition where σ is large if the side lobe cannot be observed is acceptable.
[0011]
FIG. 1 illustrates the structure of the reticle according to the first embodiment of the present invention and the scattered light at the chrome portion of the reticle in order to explain the principle of the present invention. The reticle shown in FIG. 1 is a so-called binary reticle consisting of only binary values indicating whether or not the exposure light is transmitted.
[0012]
In the figure, 1 is a quartz glass substrate, 2 is chromium, and H is the height (thickness) of chromium. The arrow indicates the scattered light generated by the reticle. SL1 is the scattered light at the upper left side of the chrome, SL2 is the scattered light to the left at the lower left side of the chrome, SL3 is the scattered light to the right at the lower left side of the chrome, SL4 Indicates the scattered light in the left direction at the lower right side of the chrome, SL5 indicates the scattered light in the right direction at the lower right side of the chrome, and SL6 indicates the scattered light at the upper right side of the chrome.
[0013]
FIG. 2 is a schematic view of an embodiment of the present invention in which a transmission reticle R 1 having a chromium height H 2 is placed in a semiconductor exposure apparatus to transfer an IC pattern onto a wafer. In the figure, R is a reticle, 3 is a projection optical system, 4 is a wafer, 21 is a wafer chuck, 22 is a Z drive stage, 23 is an X drive stage, 24 is a Y drive stage, and is formed below the reticle R. The pattern is exposed on the wafer 4 by the projection optical system 3.
[0014]
Here, the light intensity when the scattered light SL1, SL2, SL3, SL4, SL5, SL6 at the edge of the chrome 2 in FIG. 1 passes through the projection optical system 3 and forms an image on the wafer 4 is obtained.
[0015]
If the phase in imaging is SL1 reference, the phase of the scattered light (SL2, SL3, SL4, SL5) below the chrome 2 is delayed by the height H of the chrome 2. If the coma remaining in the projection optical system 3 is CM and the scattered light (SL3, SL5, SL6) directed in the right direction in FIG. 1 is delayed in phase by coma CM, the scattered light SL1, SL2, SL3, SL4, The phases of SL5 and SL6 can be expressed as follows.
[0016]
SL1 phase P1 EXP (j ωt)
SL2 phase P2 EXP {j (ωt + H)}
SL3 phase P3 EXP (j (ωt + H + CM))
SL4 phase P4 EXP {j (ωt + H)}
SL5 phase P5 EXP {j (ωt + H + CM)}
SL6 phase P6 EXP {j (ωt + CM)}
Next, how the coma aberration of the projection optical system 3 affects the imaging performance will be considered using the above phases P1 to P6.
[0017]
The influence of coma aberration on the wafer 4 of the projection optical system 3 is as follows. The scattered light SL1, SL2 and SL3 scattered on the left side with the chromium 2 in FIG. 1, and the scattered light SL4, SL5 and SL6 scattered on the right side. It is expressed using the difference I def from the intensity IR at. The intensities IL and IR can be expressed by an expression obtained by integrating the squares of the phases PL and PR with time.
[0018]
[Expression 1]

here
PL = P1 + P2 + P3
PR = P4 + P5 + P6
And The intensity difference I def is the difference between the above intensity IL and IR.
I def = IL −IR
Can be defined.
[0019]
Next, when the numerical integration of only the real part of the intensity difference I def is obtained, the real parts Re (PL) and Re (PR) of the phases PL and PR are as follows. In addition, since the result of the numerical integration of the imaginary part is the same, the following description will be made only with the real part.
[0020]
Re (PL) = cos (ωt) + cos (ωt + H) + cos (ωt + H + CM)
Re (PR) = cos (ωt + H) + cos (ωt + H + CM) + cos (ωt + CM)
Numerical integration is performed using the above equation, and the relationship between the intensity difference I def and the chromium height H is analyzed.
[0021]
Fig. 3 shows the calculation results when the vertical axis represents the intensity difference I def, the horizontal axis represents the chromium height H, and the amount of coma aberration CM of the projection optical system 3 is two conditions of λ / 10 and λ / 20. is there. In the calculation, the intensities of scattered light of SL1 to SL6 are all equal. In addition, the height H of chromium on the horizontal axis is expressed by normalizing with the center exposure wavelength λ.
[0022]
From the graph in Fig. 3, from both curves with coma aberration CM amount of λ / 10 and λ / 20,
1) I def changes like a sine function.
2) The period of change of Idef is λ.
3) The amplitude of I def is proportional to the amount of coma aberration CM generated. I understand that.
[0023]
Similarly, from the graph of FIG. 3, the chromium height H at which the intensity difference I def is maximum is independent of the amount of coma aberration CM generated.
H = (2m−1) · λ / 4 (m is a positive integer)… (A)
It can be seen that λ / 4, 3 λ / 4,.
[0024]
However, the condition of the formula (A) is for the case where the illumination σ is large, and further limitation is required when the illumination σ becomes small.
[0025]
The approximate value of σ representing coherency varies depending on how the side lobes of the light quantity distribution of the optical image as shown in FIGS. 4 and 5 appear, but roughly a value near 0.5 is the boundary. If it is defined more finely, σ may be small if a small side lobe appears in the light amount distribution of the optical image, and σ may be large if the side lobe cannot be observed.
[0026]
FIGS. 4 and 5 show the difference in the imaging state due to the coherency of the illumination system. FIG. 4 shows the optical image intensity when σ is large, and FIG. 5 shows the optical image intensity when σ is small. The output difference between right and left in FIG. 4 has symmetry when the Cr thickness is an integral multiple of half the wavelength as described above. Similarly, the difference a−b in FIG. 5 is symmetrical when the Cr thickness is an integral multiple of half the wavelength.
[0027]
On the other hand, when the illumination σ decreases, sub-peaks due to side lobes indicated by c 1 and d 2 also occur outside the edge portion as shown in FIG. It was found from both experiments and simulations that the behavior of the difference c −d related to the asymmetry due to coma aberration is different from the behavior of the difference a −b described above.
[0028]
FIG. 6 shows the relationship between the differences a −b and c −d, which are image asymmetry parameters with respect to the Cr thickness H. In FIG. 6, the vertical axis shows the asymmetric SOI of the light intensity distribution in arbitrary units, and the horizontal axis shows the Cr thickness H. The difference a −b is a sine wave with a period of λ, but the difference c −d has a period λ but is a function of the square of sin (πH / λ).
[0029]
The difference e −f between the sub-peaks generated inside is also a parameter representing asymmetry, but is opposite to the difference c −d and is a function of the square of sin (πH / λ) although it has a period λ.
[0030]
Figure 6 shows that the asymmetry of the image indicated by the difference a −b is worst when the Cr thickness H is an odd multiple of a quarter of the wavelength, but the asymmetry of the image expressed by the difference c −d It is worst only at odd multiples of half the wavelength. Therefore, it can be seen that in order to evaluate the coma aberration when the illumination σ is small, the Cr thickness H 2 must be an odd multiple of half the wavelength.
[0031]
Next will be described a procedure for evaluating the aberration of the projection optical system by using the reticle according to the present invention. In order to achieve high resolution, it is essential to accurately evaluate the aberration of the projection optical system. Accordingly, in the evaluation of the projection optical system, it is necessary to use different reticles according to conditions.
[0032]
For example, as a condition for using the illumination σ at a small size at present, there is a case where exposure is performed in which a contact hole is formed using a halftone reticle. If the refractive index of the halftone material is N d and the thickness is T, the halftone reticle shifts the phase of the membrane transmission and non-transmission areas by 180 degrees.
(N d −1) T = n λ / 2: n must satisfy the condition of an integer.
[0033]
Since the refractive index of many materials currently used for halftone reticles is almost close to 2, considering the condition of the thinnest thickness, the integer n = 1 in the above equation. Therefore
(2-1) T = 1 * λ / 2
Than
T = λ / 2
The condition for the thickness can be obtained. However, since the condition T = λ / 2 is an integral multiple of half the wavelength, the image asymmetric parameter of the difference c −d becomes the maximum condition. In other words, the condition of the current halftone reticle is a condition in which the symmetry of the image becomes the worst when the illumination σ is reduced, and can be said to be the most preferable condition for evaluating the coma aberration.
[0034]
Although the evaluation of the coma aberration has been described so far, the coma aberration of the projection optical system is generated from both the design value and the manufacturing error. Although this proposal can evaluate both with high accuracy, most of the latter production errors are in the projection optical system used in the current lithography, and coma aberration is generated due to errors of the lens alone, decentration errors, and the like. Therefore, after the evaluation using the reticle, if it is determined that the coma is large and further adjustment is necessary, the projection optical system enters the adjustment process.
[0035]
The asymmetry of the image represented by the difference a −b similarly changes in a sine function depending on the asymmetry of the illumination system. The symmetry of the illumination system means, for example, a magnification change associated with a focus change called telecentricity. Therefore, the symmetry of the illumination system is necessary to accurately evaluate the coma aberration of the projection optical system.
[0036]
If the symmetry of the illumination system is not guaranteed, the asymmetry represented by the difference a −b is an evaluation amount that is a mixture of the coma aberration of the projection optical system and the asymmetry of the illumination system. For example, when the asymmetry generated in the light quantity distribution has the same magnitude in the coma aberration and the illumination system but has a contribution in the opposite direction, the asymmetry represented by the difference a−b cannot be canceled and observed.
[0037]
On the other hand, it is known that the asymmetry of the image represented by the difference c −d occurs only with coma aberration and does not depend on the state of the illumination system.
[0038]
As described above, the magnitude of σ representing coherency is determined by whether or not a side lobe as shown in FIG. 5 appears in the light amount distribution of the optical image. As a value of σ, roughly, a value near 0.5 is one index. When a small side lobe appears in the light intensity distribution of the optical image and the image asymmetry represented by the difference c −d can be defined, the side lobe cannot be observed as shown in FIG. If the image asymmetry cannot be defined, σ is considered large.
[0039]
FIG. 7 and FIG. 8 show an application form of the binary reticle according to the present invention. FIG. 7 shows an embodiment in which the present invention is applied to a two-layer chrome reticle, and FIG. 8 shows an embodiment in which the present invention is applied to a three-layer chrome reticle . FIG. 9 shows a reticle of a reference example of the present invention , which is an example applied to a halftone reticle . 7 to 9 , 1 is a quartz glass substrate, 2 is chromium, 10, 11 is chromium oxide, 12 is a halftone light shielding member, and H is the height of chromium or the light shielding member.
[0040]
Next, as another embodiment of the present invention, a projection optical system using EUV as exposure light and a reflective reticle used in the projection optical system will be described. Examples of projection optical systems using EUV include Takeo Watanabe, Kiyoto Mashima, Masahito Niibe and Hiroo Kinoshita: Microprocesses and Nanotechnology '97: Jpn. J. Appl. Phys. Vol. 36 (1997) pp. 7597-7600. Can do. As shown in this document, in an EUV exposure system, a pattern on a reflective reticle surface is projected and exposed on a wafer by a projection optical system.
[0041]
FIG. 10 is a schematic diagram of an exposure apparatus according to the present invention using a reflective reticle (reflective binary reticle) MS having an absorber height H. M0, M1, M2, and M3 constitute a reflective projection optical system. The mirror and WF are wafers. Further, FIG. 11 is a binary reticle exposure light having a quartz glass substrate 1 is constituted by an enlarged view of a binary reticle MS consisting of only two values of the presence or absence of the reflection, the absorber 13 of the height H over the reflective layer 6 The pattern on the MS is exposed on the wafer WF by the projection optical system 3.
The height H of the absorber in the reflective reticle MS is twice as large as that of the transmissive type due to the reflection effect, so the basic amount of the height H of the absorber with respect to the center exposure wavelength is Doubled. Therefore, the height H of the absorber with respect to the center exposure wavelength λ is
When a reflective reticle is used and σ is large
H = (2N−1) · λ / 8 (N is a positive integer)
When a reflective reticle is used and σ is small
H = (2N−1) · λ / 4 (N is a positive integer)
It becomes. In the case of a reflective reticle, the basic amount on the right side is 1/2 of that of the transmissive reticle, that is, the amount corresponding to λ / 4 of the transmissive type is λ / 8, and the amount corresponding to λ / 2 of the transmissive type is λ / 4. It is a feature.
[0042]
If the exposure wavelength of EUV is 13 nm, H when σ is large will be an odd multiple of 1.625 nm, and if σ is small, if H is an odd multiple of 3.25 nm, the asymmetry due to coma in the projection optical system will be reduced. It appears most greatly, and the aberration state of the optical system can be accurately grasped.
[0043]
In the reflection system, the determination of the magnitude of σ representing coherency is similarly determined by whether or not the side lobe as shown in FIG. 5 appears in the light quantity distribution of the image. Roughly speaking, the value of σ is a value near 0.5. When a small side lobe appears in the light intensity distribution of the image and the image asymmetry represented by the difference c −d can be defined, σ is a small condition, and the side lobe cannot be observed as shown in FIG. If the asymmetry of the represented image cannot be defined, σ is considered large.
[0044]
Table 1 shows the above results for various exposure wavelengths.
[0045]
[Table 1]

[0046]
【The invention's effect】
As described above, in the method for evaluating coma aberration of the projection optical system according to the present invention, the influence of the coma aberration remaining in the projection optical system is maximized on the reticle pattern height H determined according to the center exposure wavelength λ. By setting the conditions, it was possible to evaluate minute coma aberration remaining in the projection optical system. If the projection optical system is adjusted based on the evaluation result, the completeness of the projection optical system can be increased, and a high-resolution high-resolution optical system can be realized.
[0047]
In particular, if the evaluation is performed under the condition that the coma aberration is maximized, even if the height (thickness) H of the actual reticle pattern changes, the influence of the coma aberration does not increase any more, and the worst of the image. Since the state is evaluated, desired resolution performance can be obtained.
[Brief description of the drawings]
FIG. 1 is an explanatory diagram of scattered light at a chrome part of a reticle of the present invention,
FIG. 2 is a block diagram of a semiconductor exposure apparatus,
FIG. 3 is a graph obtained by numerical integration of the relationship between the signal intensity difference and the chrome height. FIG. 4 is a graph of an image when the illumination σ is large.
FIG. 5 is a graph of an image when the illumination σ is small;
6 is a graph obtained by numerical integration of the relationship between the signal intensity difference and the chromium height. FIG. 7 is an embodiment in which the present invention is applied to a double-layer chromium reticle.
FIG. 8 shows an embodiment in which the present invention is applied to a three-layer chrome reticle;
FIG. 9 shows an example applied to a halftone reticle of a reference example of the present invention;
FIG. 10 is a diagram for explaining an embodiment in which the reflective reticle of the present invention is placed in a semiconductor exposure apparatus using EUV as exposure light and an IC pattern is transferred onto a wafer. FIG. 11 is an enlarged view of a reticle MS. Explanation of symbols]
1 Quartz glass substrate
2 Chrome
3 Projection optics
4 wafers,
6 reflective layer,
10 chromium oxide,
11 Chromium oxide,
12 Halftone shading material,
13 Absorber,
21 Wafer chuck,
22 Z drive stage,
23 X drive stage,
24 Y drive tee,
SL1-SL6 Light scattered from the chrome pattern of the reticle,
H Height of chrome or shading member

Claims (2)

In the evaluation method of coma aberration of the projection optical system,
Preparing a transparent binary reticle;
Illuminating the binary reticle under illumination conditions where σ is less than 0.5 using a luminous flux with a center wavelength λ;
Projecting an image of the pattern of the binary reticle with the projection optical system;
Evaluating the coma aberration of the projection optical system based on the asymmetry of the sub-peak due to the side lobe outside the edge portion of the light amount distribution of the image, and
The height of the pattern is
(2N-1) · λ / 2 (N is a positive integer)
A coma aberration evaluation method for a projection optical system, characterized in that: In the evaluation method of coma aberration of the projection optical system,
Preparing a reflective binary reticle;
Illuminating the binary reticle under illumination conditions where σ is less than 0.5 using a luminous flux having a center wavelength λ;
Projecting an image of the pattern of the binary reticle with the projection optical system;
Evaluating the coma aberration of the projection optical system based on the asymmetry of the sub-peak due to the side lobe outside the edge portion of the light amount distribution of the image, and
The height of the pattern is
(2N-1) · λ / 4 (N is a positive integer)
A coma aberration evaluation method for a projection optical system, characterized in that:

Images