JP3718584B2 - Film thickness measuring method and apparatus - Google Patents

Description

【００３８】
なお、同表および後で説明する表における「画素番号」の欄の「Ｎ1」は、最も短波長λ1のスペクトルを受光する画素の番号を示しており、長波長（λ2＜λ3＜…）側に行くに従って「Ｎ2」、「Ｎ3」、…としている。
【００３９】
こうして、絶対波長較正が完了すると、低圧水銀ランプを取外し、ハロゲンランプ５１２を点灯し（ステップＳ２）、しばらくしてハロゲンランプ５１２からの発光量が安定化すると、テーブルコントローラ６３９からの指令にしたがって六軸テーブル４１を移動させてチップホルダ４１３を照明位置ＩＬに位置決めする（ステップＳ３）。これによって、照明光学ユニット５１から照射される照明光が保護ガラス７３を介してホルミウムガラス７２を通過してリファレンスチップ７１で反射された後、再度ホルミウムガラス７２および保護ガラス７３を通過して結像光学ユニット５２に導光される。さらに、反射光は結像光学ユニット５２を介して分光ユニット５４に入射して分光される。
【００４０】
そこで、次のステップＳ４では、ラインセンサ５４２から出力される電気信号に基づき各画素で受光されるスペクトルの輝度を求めて分光特性を求める。こうして測定された分光特性は、例えば図１０に示すように、ホルミウムガラス７２の吸収特性の影響を直接的に受けて複数の特定波長でスペクトルの輝度レベルが大きく低下している。
【００４１】
次のステップＳ５では、可視域全体を３つの波長帯Ｒa、Ｒb、Ｒcに分け、各波長帯Ｒa、Ｒb、Ｒcにおいて最も顕著に輝度レベルの変化が生じている部分（同図の斜線部分）の重心波長λa、λb、λcを求めている（ステップＳ５）。ここで、具体的には、各波長帯Ｒa、Ｒb、Ｒcに応じたしきい値として最大輝度の４０％、８０％および８２％にそれぞれ設定して輝度レベルが大きく変化している部分を求める。そして、各部分に対応する重心画素の番号Ｎa、Ｎb、Ｎcをそれぞれ求めた後、表１に示す対応関係に基づき各画素番号Ｎa、Ｎb、Ｎcに対応する波長λa、λb、λcを求め、これらを重心波長λa、λb、λcとする。
【００４２】
こうして求めた重心波長λa、λb、λcは上記したようにホルミウムガラス７２の吸収特性を直接的に反映したものであり、この実施形態では、これらを特定波長λa、λb、λcとして制御ユニット６に記憶する（ステップＳ６）。これらの一連の処理（ステップＳ１〜Ｓ６）によって、被測定試料であるウエハの分光特性を測定するための準備処理が完了する。
【００４３】
なお、上記実施形態では、重心波長を求め、それらを特定波長としているが、重心波長を求める（ステップＳ５）代わりに、各波長帯Ｒa、Ｒb、Ｒcにおいて最も輝度レベルが低下している波長を特定波長とするようにしてもよい。また、この実施形態では、膜厚測定装置を用いて特定波長λa、λb、λcを実際のウエハの分光特性の測定に先立って測定しているが、予め別の分光測定装置によってホルミウムガラス７２の吸収特性に基づく特定波長を測定し、制御ユニット６に記憶させるようにしてもよい。
【００４４】
上記のようにして、分光測定の準備処理が完了すると、図１１に示すフローチャートにしたがって、被測定試料たるウエハの分光特性、さらに該分光特性に基づく膜厚測定を行う。
【００４５】
まず、ステップＳ１１で、被測定試料たるウエハをセットする。すなわち、この膜厚測定装置では、インデクサロボット２１によりウエハカセット１１又は１２からウエハを取り出し、中心調整装置２２によりセンタリングを行った後、供給側ステージ４８１に載置する。そして、該供給側ステージ４８１から六軸テーブル４１の保持プレート４１１にウエハを受渡し、プリアライメント用センサ４７による検出（プリアライメント）が行われた後、チップホルダ４１３を測定ヘッド５直下の所定の測定位置（照明位置ＩＬ）に配置する。
【００４６】
そして、準備処理におけるステップＳ２〜Ｓ４と同様にして、チップホルダ４１３に保持されているホルミウムガラス７２の分光特性を測定する。すなわち、ハロゲンランプ５１２を点灯し（ステップＳ１２）、しばらくしてハロゲンランプ５１２からの発光量が安定化すると、テーブルコントローラ６３９からの指令にしたがって六軸テーブル４１を移動させてチップホルダ４１３を照明位置ＩＬに位置決めして（ステップＳ１３）、可視域の照明光を上記構造体に照射する。そして、該構造体からの反射光を結像光学ユニット５２を介して分光ユニット５４に入射して分光させ、さらにラインセンサ５４２から出力される電気信号に基づき各画素で受光されるスペクトルの輝度を求めて分光特性を求める（ステップＳ１４）。
【００４７】
そして、各波長帯Ｒa、Ｒb、Ｒcにおいて最も顕著に輝度レベルの変化が生じている部分の重心画素の番号Ｎa、Ｎb、Ｎcをそれぞれ求めた（ステップＳ１５）後、制御ユニット６から準備処理で求めた特定波長λa、λb、λcを読み出し、次式に基づき互いに隣り合った特定波長間、つまり特定波長λa、λbの間、および特定波長λb、λcの間での補間係数α1、α2を求める（ステップＳ１６）。すなわち、特定波長λa、λbの間について、
α1＝（Ｎb−Ｎa）／（λb−λa）
に基づき、また特定波長λb、λcの間について、
α2＝（Ｎc−Ｎb）／（λc−λb）
に基づき、補間係数α1、α2を求め、制御ユニット６に記憶する。
【００４８】
このように補間係数α1、α2を求めておくことで、仮に経時変化や温度変化などの影響によって結像光学ユニット５２や分光ユニット５４などの光学特性が変化し、その結果、ラインセンサ５４２の各画素と分光によるスペクトルの波長との対応付けが変化したとしても、次の波長較正式によって各画素と波長との対応関係を較正することができる。すなわち、画素番号Ｎkを有する画素によって受光されるスペクトルの波長λk′は、
Ｎk≦Ｎbのとき、
λk′＝λa＋（Ｎk−Ｎa）／α1
Ｎb＜Ｎkのとき、
λk′＝λb＋（Ｎk−Ｎb）／α2
によって求まる。
【００４９】
上記のようにして補間係数α1、α2が求まると、テーブルコントローラ６３９からの指令にしたがって六軸テーブル４１を移動させて被測定試料たるウエハを照明位置ＩＬに位置決めする（ステップＳ１７）。これによって、照明光学ユニット５１から照射される照明光がウエハで反射され、この反射光が結像光学ユニット５２を介して分光ユニット５４に入射して分光される。そして、分光ユニット５４のラインセンサ５４２から出力される電気信号に基づき各画素で受光されるスペクトルの輝度を求めて分光特性を求める（ステップＳ１８）。なお、こうして求められた分光特性を示すデータは、表２に示すように、画素番号とスペクトル輝度との組み合わせデータである。
【００５０】
【表２】

【００５１】
次に、ステップＳ１６で求められた補間係数を用い、上記波長較正式にしたがって各画素番号に対応する波長を演算し、波長較正を行う（ステップＳ１９）。すなわち、画素番号Ｎ1の画素に対応する波長λ1については、
λ1′＝λa＋（Ｎ1−Ｎa）／α1
によって求められた波長λ1′に較正され、後の画素番号についても同様にして較正され、表３に示す分光特性が得られる。
【００５２】
【表３】

【００５３】
このように、この実施形態では、ホルミウムガラス７２の吸収特性を利用して被測定試料の測定ごとに波長較正を行っているため、経時変化や温度変化などの影響によって結像光学ユニット５２や分光ユニット５４などの光学特性が変化し、その結果、ラインセンサ５４２の各画素と分光によるスペクトルの波長との対応関係が変化したとしても、測定ごとにその対応関係が較正され、再現性のよい分光特性の測定を行うことができる。しかも、該ホルミウムガラス７２を六軸テーブル４１の移動に応じて移動自在に構成しているため、波長較正を行う際には、単にホルミウムガラス７２を照明位置ＩＬに位置決めするのみで済むので、波長較正の度に低圧水銀ランプを照明位置ＩＬに配置する必要があった従来例に比べて、波長較正を簡単に、効率よく行うことができる。
【００５４】
また、この膜厚測定装置では、上記のようにして分光特性の測定が完了する（ステップＳ１９）と、その波長較正された分光特性に基づきウエハの表面に形成された薄膜の膜厚測定を行う（ステップＳ２０）。このように分光特性を再現性よく測定することから、その分光特性に基づく膜厚測定を行っているため、必然的に膜厚測定も再現性よく行うことができる。なお、ステップＳ２１で膜厚測定を継続すると判断される間、上記一連の処理（ステップＳ１１〜Ｓ２０）を繰り返して行う。
【００５５】
Ｃ．変形例
なお、上記実施形態では、ホルミウムガラス７２の吸収特性を利用して特定波長を設定しているが、ホルミウムガラス７２の代わりに、透過率が特定波長で極大的に変化する物質（例えばジジムフィルタ）や多層膜などを用いてもよい。また、ホルミウムガラス７２の代わりに、特定波長で反射率が変化する反射膜を用いてもよく、この場合、リファレンスチップ７１が不要となる。
【００５６】
【発明の効果】
以上のように、この発明によれば、照明光の波長域に存在し、しかも互いに異なる少なくとも３つの特定波長λ a 、λ b 、λ c （ただし、λ a ＜λ b ＜λ c ）において、透過率あるいは反射率が極大的に変化する参照光学部材を準備し、被測定試料の分光特性を測定するのに先立って、前記参照光学部材を前記照明位置に位置決めすることで、前記参照光学部材の分光特性を測定し、前記複数の画素のうち前記特定波長λ a 、λ b 、λ cのスペクトルを受光する画素Ｎ a 、Ｎ b 、Ｎ cを特定し、前記少なくとも３つの特定波長λ a 、λ b 、λ c のうち、互いに隣り合う特定波長λ a 、λ b の間、および特定波長λ b 、λ c の間における所定の補間係数α 1 、α 2 を求めると共に、該補間係数α 1 、α 2 のいずれかと前記画素Ｎ a 、Ｎ b 、Ｎ c のいずれかとを利用して、ラインセンサの画素の各々で受光されるスペクトルの波長を演算した後、被測定試料を照明位置に位置決めし、当該被測定試料から反射される光を分光し、分光特性を実測するとともに、当該分光特性を上記演算結果に基づき波長較正するようにしているので、再現性の良い分光特性の測定を、効率良く行うことができる。
【図面の簡単な説明】
【図１】 本発明にかかる分光測定方法を適用可能な膜厚測定装置を示す斜視図である。
【図２】 チップホルダの斜視図である。
【図３】 図２のＢ−Ｂ線矢視図である。
【図４】 ホルミウムガラスの透過率特性を示すグラフである。
【図５】 測定ヘッド５を示す図である。
【図６】 図５のＡ−Ａ線矢視図である。
【図７】 図５の測定ヘッドの光学的および電気的構成を示す模式図である。
【図８】 シャッター機構を示す斜視図である。
【図９】 膜厚測定装置による分光測定における準備処理を示すフローチャートである。
【図１０】 図１の膜厚測定装置によって実測されるホルミウムガラスの分光特性を示すグラフである。
【図１１】 膜厚測定装置による分光測定方法および膜厚測定方法を示すフローチャートである。
【符号の説明】
６…制御ユニット
４１…六軸テーブル
５１…照明光学ユニット
５２…結像光学ユニット
５４…分光ユニット
７２…ホルミウムガラス
５１２…ハロゲンランプ
５１３…重水素ランプ
５４２…ラインセンサ
６１１…ＣＰＵ
６１２…ＲＯＭ
６１３…ＲＡＭ
６３９…テーブルコントローラ
ＩＬ…照明位置
Ｗ…ウエハ（被測定試料）
α1、α2…補間係数
λa〜λc…特定波長[0001]
BACKGROUND OF THE INVENTION
In the present invention, when a sample to be measured is positioned at an illumination position irradiated with illumination light in a predetermined wavelength range, the light emitted from the sample to be measured is dispersed, and each spectrum is received by a line sensor including a plurality of pixels. The present invention relates to a film thickness measurement method and apparatus for measuring the film thickness of a thin film formed on the surface of a sample to be measured by measuring spectral characteristics.
[0002]
[Prior art]
In a spectroscopic measurement apparatus for film thickness measurement, a sample to be measured such as a semiconductor wafer having a transparent thin film formed on a substrate surface is held on a stage, and a predetermined wavelength region (for example, visible wavelength) is directed toward the sample to be measured. And the light reflected by the sample is guided to the spectroscopic unit through the imaging optical unit. This spectroscopic unit is composed of a spectroscope that splits light from the imaging optical unit and a line sensor that receives a spectrum that is split by the spectroscope, and each spectrum corresponds to the spectroscope. The light is incident on the pixel, and an electrical signal corresponding to the intensity of the spectrum is output from each pixel to measure the spectral characteristics of the sample.
[0003]
In the spectroscopic measurement apparatus configured as described above, the absolute wavelength calibration is performed after the spectroscopic measurement apparatus is assembled and the optical axis is adjusted in order to obtain in advance the wavelength of the spectrum received by each pixel of the line sensor. It is carried out. In this absolute wavelength calibration, a low-pressure mercury lamp having a known emission line spectrum is placed in the illumination position in advance, and light from the low-pressure mercury lamp is guided to the spectroscopic unit through the imaging optical unit to perform absolute wavelength calibration. Each pixel of the line sensor is associated with a spectrum wavelength by spectroscopy. After the absolute wavelength calibration is completed, the low-pressure mercury lamp is removed, and the spectral characteristics are measured by irradiating the sample to be measured with illumination light.
[0004]
[Problems to be solved by the invention]
By the way, even after performing absolute wavelength calibration as described above, the optical characteristics of the imaging optical unit, spectroscopic unit, etc. change due to the effects of changes over time, temperature changes, etc. The correspondence between the pixel and the spectrum wavelength by spectroscopy may change. If there is such a change, measurement with good reproducibility cannot be performed.
[0005]
On the other hand, such problems can be resolved by performing absolute wavelength calibration periodically. However, in order to perform absolute wavelength calibration, it is necessary to remove the cover of the spectrometer and place a low-pressure mercury lamp at the illumination position. There is a great deal of effort for wavelength calibration. In particular, such a spectroscopic measurement apparatus is often used for continuous film thickness measurement of a thin film formed on the surface of a semiconductor wafer or the like, and a low-pressure mercury lamp is disposed and removed every time wavelength calibration is performed. The measurement efficiency is low and the throughput (the number of sheets that can be processed per unit time) is reduced.
[0006]
The present invention has been made in view of the above problems, and an object of the present invention is to provide a film thickness measuring method and apparatus capable of efficiently measuring spectral characteristics with good reproducibility.
[0007]
[Means for Solving the Problems]
The invention of claim 1 is a film thickness measuring method for measuring the film thickness of a thin film formed on the surface of a sample to be measured, and the sample to be measured is positioned at an illumination position irradiated with illumination light in a predetermined wavelength range. In the film thickness measuring method , the light reflected from the surface of the sample to be measured is dispersed, each spectrum is received by a line sensor composed of a plurality of pixels, and the spectral characteristics are measured to measure the film thickness of the thin film . (A) At least three specific wavelengths λ a , λ b , λ c (provided that λ a <λ b <λ c ) that are present in the wavelength range and are different from each other have a maximum transmittance or reflectance. Preparing a reference optical member that changes to (b), (b) storing at least three specific wavelengths λ a , λ b , and λ c of the reference optical member, and (c) measuring the spectral characteristics of the sample to be measured Prior to, the reference optical member is positioned at the illumination position. By Mesuru to measure the spectral characteristics of the reference optical member, wherein the specific wavelength lambda a of the plurality of pixels, lambda b, lambda pixel N a for receiving a spectrum of c, N b, identifies the N c process and, (d) said step (b) at least three specific wavelengths lambda a stored in, lambda b, lambda of c, the particular wavelength lambda a mutually adjacent, during the lambda b, and the specific wavelength lambda b, λ predetermined interpolation coefficients alpha 1 between c, together with obtaining the alpha 2, interpolation coefficients alpha 1, pixel N a identified in the process as one of α 2 (c), N b , with any of the N c by using a step of calculating the wavelength of the spectrum of light received by each pixel of the line sensor, after (e) the step (d), positioning the DUT on the illumination position, the object to be and dispersing the light reflected from the sample, after measuring the spectral characteristics, based on the spectral characteristics in the step (d) of the operation result It comprises the step of wavelength calibration can, a.
[0008]
The invention according to claim 2 is a film thickness measuring device for measuring the film thickness of a thin film formed on the surface of the sample to be measured, and positions the sample to be measured at an illumination position irradiated with illumination light in a predetermined wavelength range. In a film thickness measuring apparatus for measuring the film thickness of the thin film by dispersing light reflected from the surface of the sample to be measured and receiving each spectrum with a line sensor composed of a plurality of pixels and measuring spectral characteristics An illumination optical unit that irradiates an illumination position with illumination light of a predetermined wavelength range, and at least three specific wavelengths λ a , λ b , λ c (provided that λ a <λ b <λ c ) , a reference optical member whose transmittance or reflectance changes maximally, a moving means for selectively moving the sample to be measured and the reference optical member to the illumination position, and positioning at the illumination position The measured An imaging optical unit that collects light emitted from a sample or the reference optical member at a predetermined position, a spectroscope that splits the light from the imaging optical unit, and a plurality of pixels. A spectral unit having a line sensor for receiving the measured spectrum, and obtaining spectral characteristics of the sample to be measured and the reference optical member based on an electric signal output from each pixel of the line sensor while controlling each part of the apparatus. A film thickness measuring apparatus including a control unit, wherein the control unit measures at least three specific wavelengths λ a , λ b , λ c of the reference optical member, and measures spectral characteristics of the sample to be measured Prior to, the spectral characteristics of the reference optical member are measured by positioning the reference optical member at the illumination position, and the identification among the plurality of pixels is performed. Pixel identifying means for identifying the pixels N a , N b , N c that receive the spectra of the wavelengths λ a , λ b , λ c , and at least three specific wavelengths λ a , λ b , λ c stored in the memory which of the specific wavelength lambda a mutually adjacent, during the lambda b, and the specific wavelength lambda b, predetermined interpolation coefficients alpha 1 between lambda c, together with obtaining the alpha 2, interpolation coefficients alpha 1, the alpha 2 of or from the specific pixel N a, N b, by utilizing and either N c, a calculating means for calculating a wavelength of the spectrum of light received by each pixel of the line sensor, the illumination position measured sample Wavelength calibration means for positioning and calibrating the actually measured spectral characteristics, and the moving means includes a table for holding the sample to be measured, and a holder fixed to the table while holding the reference optical member And driving the table to measure the sample to be measured. Fine said reference optical member and a table drive means for moving integrally.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a perspective view showing a film thickness measuring apparatus according to the present invention. Hereinafter, after describing the outline of the film thickness measuring apparatus, a spectroscopic measurement method performed by the film thickness measuring apparatus will be described in detail.
[0010]
A. Film Thickness Measuring Device The film thickness measuring device shown in this figure is a device that measures the spectral characteristics of a semiconductor wafer and measures the film thickness of a thin film formed on the semiconductor wafer based on the spectral characteristics, and mounts a wafer cassette. A mounting stage 1, an indexer unit 2 for taking a semiconductor wafer (hereinafter abbreviated as “wafer”) into and out of a wafer cassette, and the spectral characteristics of the semiconductor wafer are measured, and further formed on the surface of the semiconductor wafer based on the spectral characteristics. And a measuring unit 3 for measuring the thickness of the thin film. The film thickness measurement apparatus according to this embodiment has a layout configuration in which the above-described placement stage 1 and measurement unit 3 are arranged around the indexer unit 2.
[0011]
A-1. Placement stage 1
In the illustrated example, two placement stages 1 are provided. In each stage 1, wafer cassettes 11 and 12 that store a plurality of wafers in multiple stages have their storage openings facing the indexer unit 2. It is placed in the state. In the apparatus of this embodiment, two types of wafers having a diameter of 200 mm and 300 mm are measurement targets.
[0012]
A-2. Indexer unit The indexer unit 2 includes an indexer robot 21 serving as a wafer transfer unit, and a center adjusting device 22 for correcting the center position of the wafer taken out from the wafer cassettes 11 and 12 by the indexer robot 21. Is provided.
[0013]
The indexer robot 21 is a horizontal articulated robot that can move up and down, and a thin plate-like hand 211 for holding a wafer is attached to the tip of the arm. A suction hole 212 is opened on the surface of the hand 211, and a negative pressure is supplied through the suction hole 212 so that the wafer is sucked and held by the hand 211.
[0014]
The center adjustment device 22 corrects the wafer position with respect to the hand 211, and includes a pair of chucks 221 that can be contacted and separated, and the wafers taken out from the wafer cassettes 11 and 12 are radially moved by the chuck 221. The wafer position is mechanically corrected by holding it.
[0015]
A-3. Measurement stage 4
The measurement unit 3 includes a measurement stage 4 serving as a holding portion of a wafer to be measured, and a measurement head 5 for measuring a film thickness disposed above the measurement stage 4.
[0016]
The measurement stage 4 is provided with a six-axis table 41 for holding the wafer movably, a pre-alignment sensor 47, and an auxiliary stage 48 for waiting the wafer.
[0017]
The six-axis table 41 includes a disk-shaped holding plate 411 for sucking and holding a wafer, and a six-axis direction, that is, an X-axis, a Y-axis, a Z-axis, and about each of these axes (θY direction, θX direction, The support mechanism 412 is supported so as to be displaceable in the θ direction). The wafer is transferred to and from the auxiliary stage 48 while moving the holding plate 411 in each of the above directions, and the measurement head 5 and the pre-alignment. The wafer is arranged at a predetermined measurement position of the sensor 47 for use.
[0018]
Further, a chip holder 413 (FIG. 8 described later) is attached to a side surface portion of the holding plate 411. As shown in FIGS. 2 and 3, a concave portion 413a is formed in the upper surface portion of the chip holder 413. Inside the concave portion 413a, a reference chip 71 such as silicon and holmium glass as a reference optical member are formed. 72 and a protective glass 73 whose surface is coated with an antireflection film are laminated in this order from the bottom. Therefore, when the holding plate 411 is moved in each of the above directions, the holmium glass 72 can be moved to an arbitrary position together with the holding plate 411. Thus, in this embodiment, the moving means for selectively moving the wafer and the holmium glass 72 to the measurement position by the six-axis table 41, the chip holder 413, and the drive mechanism (not shown) of the six-axis table 41. Is configured.
[0019]
Note that the holmium glass 72 has transmittance characteristics as shown in FIG. 4, for example, and as is clear from the figure, the transmittance is maximally reduced at a specific wavelength. Therefore, after the six-axis table 41 is moved and the holmium glass 72 is positioned at the illumination position IL, the illumination light from the measurement head 5 passes through the holmium glass 72 via the protective glass 73 and is reflected by the reference chip 71. The light passes through the holmium glass 72 and the protective glass 73 again and is guided to the measuring head 5 side. At this time, among the wavelength components of the reflected light, the specific wavelength component is absorbed by the holmium glass 72 and a reduction in luminance level is recognized. In this embodiment, wavelength calibration is performed using this absorption characteristic. This will be described in detail later.
[0020]
Returning to FIG. 1, the description of the configuration of the measurement stage 4 will be continued. The pre-alignment sensor 47 includes a light irradiation unit 471 and a light receiving unit 472 arranged vertically with a predetermined inspection space interposed therebetween. The pre-alignment sensor 47 rotates the wafer with a wafer edge portion interposed in the inspection space. Thus, prealignment, that is, the eccentricity of the wafer center with respect to the center of the holding plate 411 is detected.
[0021]
The auxiliary stage 48 is disposed between the indexer unit 2 and the six-axis table 41. In the illustrated example, the auxiliary stage 48 waits for the wafer before measurement and the supply-side stage 481 for waiting the wafer before measurement. A discharge side stage 482 is provided side by side. Each of the stages 481 and 482 is formed in a substantially U shape having a notch portion for interposing the holding plate 411. For example, the transfer of the wafer from the supply side stage 481 to the six-axis table 41 is a supply The operation is performed by moving the supply side stage 481 and the holding plate 411 relatively up and down in a state where the holding plate 411 is disposed below the notch lower portion of the side stage 481. On the other hand, the delivery of the wafer from the six-axis table 41 to the discharge side stage 482 is performed based on the reverse operation.
[0022]
A-5. Measuring head 5
5 is a diagram showing the measurement head 5, FIG. 6 is a view taken along the line AA in FIG. 5, and FIG. 7 is a schematic diagram showing an optical and electrical configuration of the measurement head 5. As shown in FIG. As described above, the measuring head 5 includes the illumination optical unit 51 as the illumination optical apparatus according to the present invention. In the illumination optical unit 51, a halogen lamp 512 and a deuterium lamp 513 are arranged in a predetermined light source arrangement area (lower right corner area in FIG. 6) 511. Lights L 1 and L 2 emitted from the lamps 512 and 513 are incident on the imaging optical unit 52 through the illumination optical unit 51. In this embodiment, the light source arrangement region 511 is provided with the protective cover 515 so as to cover the halogen lamp 512 and the deuterium lamp 513, and the protective cover 515 is illustrated via the exhaust duct 516. The exhaust fan to be omitted is connected.
[0023]
The illumination optical system 514 to which the lights L1 and L2 emitted from both lamps 512 and 513 are incident includes a filter turret 514a in which a plurality of filters are arranged concentrically, and an arbitrary filter by rotating the filter turret 514a. A filter switching motor 514b for positioning on the optical axis OA, a lens 514c, a total reflection mirror 514d, a field stop 514e, and a half mirror 514f are provided, and the white light L1 emitted from the halogen lamp 512 is light. Light in the visible region (hereinafter referred to as “visible light”) corresponding to the filter passing through the filter positioned on the axis OA is imaged through the lens 514c, the total reflection mirror 514d, the field stop 514e, and the half mirror 514f. The light enters the optical unit 52.
[0024]
The illumination optical system 514 further includes a total reflection mirror 514g and a lens 514h. After the ultraviolet light L2 emitted from the deuterium lamp 513 is reflected by the total reflection mirror 514g, the lens 514h and the half mirror 514f are provided. Then, the light enters the imaging optical unit 52.
[0025]
The high-energy ultraviolet light L2 emitted from the deuterium lamp 513 is always incident on the total reflection mirror 514g, and when the surface of the total reflection mirror 514g is exposed to ultraviolet light for a long time, the surface becomes cloudy. Arise. Therefore, in this embodiment, only when the shutter mechanism 517 is provided between the deuterium lamp 513 and the total reflection mirror 514g and the sample to be measured is irradiated with the ultraviolet light L2, the shutter mechanism 517 is opened and the deuterium lamp 513 is opened. The ultraviolet light L2 from is guided to the total reflection mirror 514g.
[0026]
FIG. 8 is a perspective view showing the shutter mechanism. FIG. 8A shows a state where the shutter is opened, while FIG. 8B shows a state where the shutter is closed. As shown in the figure, the shutter mechanism 517 has a support plate 517b attached to the front end surface of a base 517a having a substantially L-shaped cross section. A through hole 517c is provided in the lower left portion of the support plate 517b, and ultraviolet light L2 can be incident on the total reflection mirror 514g fixed to the front end surface of the base 517a through the through hole 517c. Yes.
[0027]
In addition, a shutter 517e is pivotally supported on the support plate 517b so as to be rotatable around a rotation shaft 517d provided at the upper center portion. Further, a shutter drive motor 517f is connected to the rotation shaft 517d, and the shutter 517e rotates about the rotation shaft 517d as a rotation center in response to the operation of the motor 517f, thereby opening the through hole 517c. It is configured to be occluded.
[0028]
Accordingly, when the sample to be measured is irradiated with the ultraviolet light L2, as shown in FIG. 5A, the shutter 517e is retracted from the through hole 517c and the ultraviolet light L2 is reflected by the total reflection mirror 514g to be measured. When the sample is guided to the sample and the sample to be measured does not need to be irradiated with the ultraviolet light L2, the shutter 517e is positioned in the through hole 517c and the ultraviolet light L2 is totally reflected as shown in FIG. Incident to the mirror 514g can be prohibited. As a result, it is possible to prevent the surface of the total reflection mirror 514g from being irradiated with the ultraviolet light L2 for a long time, and to effectively prevent fogging of the surface of the total reflection mirror 514g.
[0029]
Next, returning to FIGS. 5 and 7, the imaging optical unit 52 that receives the visible light and the ultraviolet light emitted from the illumination optical unit 51 configured as described above, and the light reflected by the wafer are dispersed. The description of the configuration of the spectroscopic unit 54 will be continued.
[0030]
The imaging optical unit 52 includes a lens turret 522 having a plurality of objective lenses 521a and 521b concentrically arranged and rotatable around a rotation axis 522a, and an arbitrary drive by rotating the lens turret 522 around the rotation axis 522a. An objective lens switching motor 523 for positioning the objective lens on the optical axis OA, a beam splitter 524, and a tube lens 525 are provided. The illumination light (visible light or ultraviolet light) from the illumination optical unit 51 is transmitted by the beam splitter 524. The light is reflected and irradiated to the predetermined illumination position IL through the objective lens positioned on the optical axis OA by the objective lens switching motor 523.
[0031]
The six-axis table 41 (FIG. 1) is configured to be movable over a relatively wide range including the illumination position IL, and a wafer W as a sample to be measured is placed on the holding plate 411 of the six-axis table 41. The six-axis table 41 is configured so that an arbitrary minute area on the surface of the wafer W can be positioned at the illumination position IL. Although not shown in the drawings, the six-axis table 41 detects its position (X, Y coordinates, etc.), and gives the position information to the control unit 6 that controls the entire apparatus. It has become.
[0032]
The light reflected by the minute region of the wafer W located at the illumination position IL is condensed at a predetermined position on the optical axis OA via the objective lens 521, the beam splitter 524, and the tube lens 525. A plate 532 having a pinhole 531 at the center is disposed in the vicinity of the light collecting position. A shutter 533 is disposed in the vicinity of the plate 532, and the shutter 533 is driven to open and close based on a signal from the control unit 6. Whether to control or not.
[0033]
The spectroscopic unit 54 includes a concave diffraction grating 541 that splits reflected light and a line sensor 542 that detects the spectral light intensity of the diffracted light diffracted by the concave diffraction grating 541. The line sensor 542 is configured by, for example, a photodiode array or a CCD, and is arranged in a conjugate relationship with the pinhole 531. Therefore, the light taken into the spectroscopic unit 54 is split into the concave diffraction grating 541, and a signal corresponding to the spectral light intensity of the light is given from the line sensor 542 to the control unit 6.
[0034]
Note that a prism 551 is disposed on the optical axis OA between the tube lens 525 and the plate 532 so that a part of the light from the wafer W is taken out. Further, the extracted light is condensed at a predetermined position via the lens 552. An image pickup element 553 is disposed at this condensing position, and an image signal corresponding to an image of a partial region (film thickness measurement region) on the surface of the wafer W is given to the control unit 6. As described above, in this embodiment, an image of the measurement region can be captured by the imaging unit 55, but the arrangement position of the imaging unit 55 is not limited between the tube lens 525 and the plate 532. In other words, it is optional as long as it is on the optical axis OA between the objective lens 521 and the spectroscopic unit 54.
[0035]
As shown in FIG. 7, the control unit 6 includes a well-known CPU 611 that executes logical operations, a ROM 612 that stores various programs for controlling the CPU 611 in advance, and temporarily stores various data during operation of the apparatus. The control part 61 provided with RAM613 to perform is provided. The control unit 61 includes the following components via the I / O unit 62:
・ Operation unit 31,
-Monitor 32,
Data processing unit 631: Performs predetermined processing on the signal from the line sensor 542,
Shutter controller 632: Controls opening / closing of the shutter 533,
Image processing unit 633: performs predetermined processing on the image signal from the image sensor 553,
Filter switching controller 634: Rotation control of the filter switching motor 514b is performed to switch the filter.
Lighting circuit 635: Controls lighting / extinguishing of the halogen lamp 512,
Lighting circuit 636: Controls lighting / extinguishing of the deuterium lamp 513,
Shutter controller 637: Controls opening / closing of the shutter 517e,
Lens turret controller 638: The objective lens switching motor 523 is rotationally controlled to switch the objective lens.
Table controller 639: Drives and controls the six-axis table 41.
And are electrically connected.
[0036]
B. Spectroscopic Measurement Method Next, a spectroscopic measurement method in the film thickness measuring apparatus configured as described above will be described with reference to FIGS. In this film thickness measuring apparatus, as in the conventional case, after assembling the apparatus and adjusting the optical axis, absolute wavelength calibration is performed (step S1). In this absolute wavelength calibration, a low-pressure mercury lamp having a known emission line spectrum is arranged in advance at the illumination position IL, and light from the low-pressure mercury lamp is guided to the spectroscopic unit 54 via the imaging optical unit 52 to obtain an absolute wavelength. Calibration is performed to associate each pixel of the line sensor 542 with a spectrum wavelength by spectroscopy. As a result, the correspondence shown in Table 1 is obtained and stored in the control unit 6.
[0037]
[Table 1]

[0038]
Note that “N1” in the column of “pixel number” in the table and the table described later indicates the number of the pixel that receives the spectrum of the shortest wavelength λ1, and is on the long wavelength (λ2 <λ3 <...) Side. As you go to “N2”, “N3”, and so on.
[0039]
Thus, when the absolute wavelength calibration is completed, the low-pressure mercury lamp is removed, the halogen lamp 512 is turned on (step S2), and after a while, the light emission amount from the halogen lamp 512 is stabilized, the six steps are performed according to a command from the table controller 639. The axis table 41 is moved to position the chip holder 413 at the illumination position IL (step S3). Thereby, the illumination light irradiated from the illumination optical unit 51 passes through the holmium glass 72 through the protective glass 73 and is reflected by the reference chip 71, and then passes through the holmium glass 72 and the protective glass 73 again to form an image. The light is guided to the optical unit 52. Further, the reflected light enters the spectroscopic unit 54 via the imaging optical unit 52 and is split.
[0040]
Therefore, in the next step S4, the spectral characteristic is obtained by obtaining the luminance of the spectrum received by each pixel based on the electrical signal output from the line sensor 542. For example, as shown in FIG. 10, the spectral characteristics measured in this way are directly affected by the absorption characteristics of the holmium glass 72, and the luminance level of the spectrum is greatly reduced at a plurality of specific wavelengths.
[0041]
In the next step S5, the entire visible range is divided into three wavelength bands Ra, Rb, and Rc, and the most prominent change in luminance level occurs in each wavelength band Ra, Rb, and Rc (shaded area in the figure). Centroid wavelengths λa, λb, and λc are obtained (step S5). Specifically, the threshold level corresponding to each wavelength band Ra, Rb, and Rc is set to 40%, 80%, and 82% of the maximum luminance, respectively, and the portion where the luminance level is greatly changed is obtained. . Then, after obtaining the center-of-gravity pixel numbers Na, Nb, and Nc corresponding to the respective portions, the wavelengths λa, λb, and λc corresponding to the pixel numbers Na, Nb, and Nc are obtained based on the correspondence shown in Table 1. These are defined as centroid wavelengths λa, λb, and λc.
[0042]
The center-of-gravity wavelengths λa, λb, and λc thus obtained directly reflect the absorption characteristics of the holmium glass 72 as described above. In this embodiment, these are given to the control unit 6 as specific wavelengths λa, λb, and λc. Store (step S6). By these series of processes (steps S1 to S6), the preparation process for measuring the spectral characteristics of the wafer as the sample to be measured is completed.
[0043]
In the above embodiment, the centroid wavelengths are obtained and set as specific wavelengths. Instead of obtaining the centroid wavelengths (step S5), the wavelength with the lowest luminance level in each wavelength band Ra, Rb, Rc is used. You may make it be a specific wavelength. In this embodiment, the specific wavelengths λa, λb, and λc are measured using the film thickness measurement device prior to the measurement of the actual spectral characteristics of the wafer, but the holmium glass 72 is previously measured by another spectral measurement device. A specific wavelength based on the absorption characteristic may be measured and stored in the control unit 6.
[0044]
When the preparation process for spectroscopic measurement is completed as described above, according to the flowchart shown in FIG. 11, the spectral characteristics of the wafer to be measured and the film thickness measurement based on the spectral characteristics are performed.
[0045]
First, in step S11, a wafer to be measured is set. That is, in this film thickness measurement apparatus, the wafer is taken out from the wafer cassette 11 or 12 by the indexer robot 21, centered by the center adjustment apparatus 22, and then placed on the supply side stage 481. Then, after the wafer is delivered from the supply side stage 481 to the holding plate 411 of the six-axis table 41 and detected by the pre-alignment sensor 47 (pre-alignment), the chip holder 413 is subjected to a predetermined measurement directly below the measuring head 5. It arrange | positions in a position (illumination position IL).
[0046]
And the spectral characteristic of the holmium glass 72 currently hold | maintained at the chip | tip holder 413 is measured similarly to step S2-S4 in a preparatory process. That is, the halogen lamp 512 is turned on (step S12), and after a while, the light emission amount from the halogen lamp 512 is stabilized, the six-axis table 41 is moved in accordance with a command from the table controller 639, and the chip holder 413 is moved to the illumination position. After positioning to IL (step S13), the structure body is irradiated with illumination light in the visible range. Then, the reflected light from the structure is incident on the spectroscopic unit 54 via the imaging optical unit 52 to be spectrally separated, and the luminance of the spectrum received by each pixel based on the electric signal output from the line sensor 542 is obtained. The spectral characteristics are obtained (step S14).
[0047]
Then, the numbers Na, Nb, and Nc of the center-of-gravity pixels of the portions where the luminance level changes most significantly in each wavelength band Ra, Rb, and Rc are obtained (step S15), and then the control unit 6 performs a preparation process. The obtained specific wavelengths λa, λb, and λc are read out, and interpolation coefficients α1 and α2 between specific wavelengths adjacent to each other, that is, between the specific wavelengths λa and λb and between the specific wavelengths λb and λc, are obtained based on the following equation: (Step S16). That is, between the specific wavelengths λa and λb,
α1 = (Nb−Na) / (λb−λa)
And between the specific wavelengths λb and λc
α2 = (Nc−Nb) / (λc−λb)
Based on the above, interpolation coefficients α 1 and α 2 are obtained and stored in the control unit 6.
[0048]
By obtaining the interpolation coefficients α1 and α2 in this way, the optical characteristics of the imaging optical unit 52 and the spectroscopic unit 54 change due to the influence of changes over time and temperature, and as a result, each of the line sensors 542 is changed. Even if the correspondence between the pixel and the spectral wavelength is changed, the correspondence between each pixel and the wavelength can be calibrated by the following wavelength calibration formula. That is, the wavelength λk ′ of the spectrum received by the pixel having the pixel number Nk is
When Nk ≦ Nb,
λk ′ = λa + (Nk−Na) / α1
When Nb <Nk
λk ′ = λb + (Nk−Nb) / α2
It is obtained by.
[0049]
When the interpolation coefficients α1 and α2 are obtained as described above, the six-axis table 41 is moved in accordance with a command from the table controller 639 to position the wafer to be measured at the illumination position IL (step S17). As a result, the illumination light emitted from the illumination optical unit 51 is reflected by the wafer, and the reflected light enters the spectroscopic unit 54 via the imaging optical unit 52 and is split. Then, based on the electric signal output from the line sensor 542 of the spectroscopic unit 54, the luminance of the spectrum received by each pixel is obtained to obtain the spectral characteristics (step S18). The data indicating the spectral characteristics thus obtained is a combination data of the pixel number and the spectral luminance as shown in Table 2.
[0050]
[Table 2]

[0051]
Next, the wavelength corresponding to each pixel number is calculated according to the wavelength calibration equation using the interpolation coefficient obtained in step S16, and wavelength calibration is performed (step S19). That is, for the wavelength λ1 corresponding to the pixel with the pixel number N1,
λ1 '= λa + (N1-Na) / α1
Is calibrated to the wavelength λ1 ′ obtained by the above, and the subsequent pixel numbers are similarly calibrated, and the spectral characteristics shown in Table 3 are obtained.
[0052]
[Table 3]

[0053]
As described above, in this embodiment, since the wavelength calibration is performed for each measurement of the sample to be measured using the absorption characteristic of the holmium glass 72, the imaging optical unit 52 and the spectral component are affected by the influence of the change over time and the temperature change. Even if the optical characteristics of the unit 54 and the like change, and as a result, the correspondence between each pixel of the line sensor 542 and the spectrum wavelength by spectrum changes, the correspondence is calibrated for each measurement, and the spectral characteristics with good reproducibility are obtained. Measurement of characteristics can be performed. Moreover, since the holmium glass 72 is configured to be movable in accordance with the movement of the six-axis table 41, when performing wavelength calibration, it is only necessary to position the holmium glass 72 at the illumination position IL. Compared to the conventional example in which a low-pressure mercury lamp needs to be arranged at the illumination position IL each time calibration is performed, wavelength calibration can be performed easily and efficiently.
[0054]
Further, in this film thickness measurement apparatus, when the measurement of the spectral characteristics is completed as described above (step S19), the film thickness of the thin film formed on the surface of the wafer is measured based on the spectral characteristics whose wavelength has been calibrated. (Step S20). Since the spectral characteristics are measured with good reproducibility in this way, the film thickness is measured based on the spectral characteristics, so that the film thickness can be measured with high reproducibility. While it is determined in step S21 that the film thickness measurement is to be continued, the above series of processing (steps S11 to S20) is repeated.
[0055]
C. In the above embodiment, the specific wavelength is set by using the absorption characteristics of the holmium glass 72. Instead of the holmium glass 72, a substance (for example, didymium) whose transmittance changes maximally at the specific wavelength. A filter) or a multilayer film may be used. Further, instead of the holmium glass 72, a reflective film whose reflectance changes at a specific wavelength may be used. In this case, the reference chip 71 is not necessary.
[0056]
【The invention's effect】
As described above, according to the present invention, at least three specific wavelengths λ a , λ b , λ c (provided that λ a <λ b <λ c ) exist in the wavelength range of the illumination light and are different from each other. Preparing a reference optical member having a maximum change in transmittance or reflectance, and positioning the reference optical member at the illumination position prior to measuring the spectral characteristics of the sample to be measured. the spectral characteristics of the member were measured, said specific wavelength lambda a of the plurality of pixels, lambda b, pixel N a which receives the spectrum of lambda c, N b, identifies the N c, wherein said at least three specific wavelengths lambda Among the a , λ b , λ c , predetermined interpolation coefficients α 1 , α 2 between specific wavelengths λ a , λ b adjacent to each other and between the specific wavelengths λ b , λ c are obtained , and the interpolation coefficients alpha 1, wherein either of the alpha 2 pixels N a, N b, use and either N c Te After calculating the wavelength of the spectrum of light received by each pixel of the line sensor, with positioning the DUT to the lighting position, and dispersing the light reflected from the DUT to the measured spectral characteristics, Since the wavelength characteristics of the spectral characteristics are calibrated based on the calculation result, it is possible to efficiently measure the spectral characteristics with good reproducibility.
[Brief description of the drawings]
FIG. 1 is a perspective view showing a film thickness measuring apparatus to which a spectroscopic measuring method according to the present invention can be applied.
FIG. 2 is a perspective view of a chip holder.
FIG. 3 is a view taken along the line BB in FIG. 2;
FIG. 4 is a graph showing transmittance characteristics of holmium glass.
FIG. 5 is a diagram showing a measurement head 5;
6 is a view taken along line AA in FIG.
7 is a schematic diagram showing an optical and electrical configuration of the measurement head of FIG. 5. FIG.
FIG. 8 is a perspective view showing a shutter mechanism.
FIG. 9 is a flowchart showing a preparation process in spectroscopic measurement by a film thickness measuring apparatus.
10 is a graph showing spectral characteristics of holmium glass actually measured by the film thickness measuring apparatus shown in FIG.
FIG. 11 is a flowchart showing a spectroscopic measurement method and a film thickness measurement method by a film thickness measurement apparatus.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 6 ... Control unit 41 ... Six axis table 51 ... Illumination optical unit 52 ... Imaging optical unit 54 ... Spectral unit 72 ... Holmium glass 512 ... Halogen lamp 513 ... Deuterium lamp 542 ... Line sensor 611 ... CPU
612 ... ROM
613 ... RAM
639 ... Table controller IL ... Illumination position W ... Wafer (sample to be measured)
α1, α2: Interpolation coefficients λa to λc: Specific wavelength

Claims (2)

A film thickness measuring method for measuring a film thickness of a thin film formed on a surface of a sample to be measured, wherein the surface of the sample to be measured is positioned when the sample to be measured is positioned at an illumination position irradiated with illumination light in a predetermined wavelength range. in the film thickness measuring method of light spectrally, measures the thickness of the thin film by measuring the spectral characteristics is received by the line sensor comprising a respective spectra from the plurality of pixels reflected from,
(a) In at least three specific wavelengths λ a , λ b , λ c (provided that λ a <λ b <λ c ) that exist in the wavelength range and are different from each other, the transmittance or reflectance is maximized. Providing a changing reference optical member;
(b) storing at least three specific wavelengths λ a , λ b , λ c of the reference optical member;
(c) measuring the spectral characteristics of the reference optical member by positioning the reference optical member at the illumination position prior to measuring the spectral characteristics of the sample to be measured; Identifying pixels N a , N b , N c that receive the spectra of wavelengths λ a , λ b , λ c ;
; (d) step (b) at least three specific wavelengths lambda a stored in, lambda b, lambda of c, the particular wavelength lambda a mutually adjacent, during the lambda b, and the specific wavelength lambda b, the lambda c predetermined interpolation coefficients alpha 1 between, the seek alpha 2, utilizing and either interpolation coefficients alpha 1, alpha pixel N a identified in 2 either said step (c), N b, N c Calculating a wavelength of a spectrum received by each of the pixels of the line sensor ;
(e) After performing the step (d), the sample to be measured is positioned at the illumination position, the light reflected from the sample to be measured is dispersed, the spectral characteristics are measured, and then the spectral characteristics are measured in the step a step of wavelength calibration based on the calculation result of (d),
A film thickness measuring method comprising: A film thickness measuring device for measuring a film thickness of a thin film formed on a surface of a sample to be measured, wherein the surface of the sample to be measured is positioned when the sample to be measured is positioned at an illumination position irradiated with illumination light in a predetermined wavelength range In a film thickness measuring apparatus for measuring the film thickness of the thin film by dispersing light reflected from the light, receiving each spectrum with a line sensor composed of a plurality of pixels, and measuring spectral characteristics,
An illumination optical unit that illuminates an illumination position with illumination light in a predetermined wavelength range;
Reference in which the transmittance or reflectance changes maximally in at least three specific wavelengths λ a , λ b , λ c (where λ a <λ b <λ c ) that exist in the wavelength range and are different from each other An optical member;
Moving means for selectively moving the sample to be measured and the reference optical member to the illumination position;
An imaging optical unit that collects light emitted from the sample to be measured or the reference optical member positioned at the illumination position at a predetermined position;
A spectroscopic unit that splits light from the imaging optical unit, and a line sensor that includes a plurality of pixels and that receives a spectrum separated by the spectroscope;
A film thickness measuring apparatus comprising a control unit that obtains spectral characteristics of the sample to be measured and the reference optical member based on an electric signal output from each pixel of the line sensor while controlling each part of the apparatus,
The control unit includes a memory for storing at least three specific wavelengths λ a , λ b , λ c of the reference optical member;
Prior to measuring the spectral characteristic of the sample to be measured, the spectral characteristic of the reference optical member is measured by positioning the reference optical member at the illumination position, and the specific wavelength λ a of the plurality of pixels is determined . pixel specifying means for specifying the pixels N a , N b , and N c that receive the spectra of λ b and λ c ;
Among the at least three specific wavelengths λ a , λ b , λ c stored in the memory, predetermined interpolation coefficients between the specific wavelengths λ a , λ b adjacent to each other and between the specific wavelengths λ b , λ c α 1 and α 2 are obtained, and light is received by each of the pixels of the line sensor using one of the interpolation coefficients α 1 and α 2 and one of the specified pixels N a , N b , and N c. Computing means for computing the wavelength of the spectrum;
A wavelength calibration means for positioning the sample to be measured at the illumination position and wavelength-calibrating the actually measured spectral characteristics ;
The moving means integrally holds the sample to be measured and the reference optical member by driving the table while holding the reference optical member, a table that holds the sample to be measured, and the table fixed to the table. A film thickness measuring apparatus comprising: a table driving means for moving the table to the position .

Images