DE102017220408A1 - Optical system for microlithography, as well as methods for position determination - Google Patents

Abstract

The invention relates to an optical system for microlithography, and a method for position determination. An optical system according to the invention comprises at least one component and at least one interferometer associated with said position determining unit having a light source unit, said component being an optical element, said light source unit having a light source for generating at least a first partial beam having a first frequency ( f ₁ ) and a tuning device, via which an adjustment of the first frequency (f ₁ ) to different values in a predetermined frequency sequence is feasible having.

Description

BACKGROUND OF THE INVENTION

Field of the invention

The invention relates to an optical system for microlithography, and a method for position determination.

State of the art

Microlithography is used to fabricate microstructured devices such as integrated circuits or LCDs. The microlithography process is carried out in a so-called projection exposure apparatus which has an illumination device and a projection objective. The image of a mask (= reticle) illuminated by means of the illumination device is hereby projected onto a substrate (eg a silicon wafer) coated with a photosensitive layer (photoresist) and arranged in the image plane of the projection objective in order to apply the mask structure to the photosensitive coating of the Transfer substrate.

In EUV-designed projection exposure equipment, i. at wavelengths below 15 nm (e.g., about 13 nm or about 7 nm), mirrors are used as optical components for the imaging process due to the lack of availability of suitable transparent refractive materials.

In operation of such projection lenses designed for EUV, in which mask and wafer are usually moved relative to one another in a scanning process, the positions of the mirrors, which are partially movable in all six degrees of freedom, must be set and maintained both with each other and with the mask or wafer with high accuracy to avoid, or at least reduce, aberrations and concomitant impairments to the imaging result. In this position determination, e.g. be demanded over a path length of 1 meter accuracies of the length measurement in the picometer (pm) range.

Various approaches are known in the prior art for interferometrically measuring the position of the wafer or the wafer stage, the reticle plane and the individual objective mirrors. This measurement can be carried out in particular using a heterodyne interferometer, in which the partial beams which run in two separate partial beam paths have mutually different frequencies and mutually orthogonal polarization states.

11 shows only a schematic representation of a possible conventional structure of a heterodyne interferometer. According to 11 creates a light source 1101 two light beams or light beam components with mutually different frequencies f ₁ and f ₂ , These components have mutually orthogonal polarization states, so that via a polarization beam splitter 1110 a splitting takes place on different beam paths. The sub-beams which propagate in mutually different directions depending on the polarization state become, after passing through a first retardation plate (lambda / 4-plate) 1125 and reflection on a first mirror 1120 or passing through a second delay plate (lambda / 4-plate) 1135 and reflection on a second mirror 1130 over the polarization beam splitter 1110 recombined. The recombined rays are rotated by polarizers 1102 . 1106 (45 ° position between the two orthogonally polarized beams) brought to interference, wherein the frequency of the interference corresponds to the beat frequency. A detector 1140 serves to determine the phase difference between the signal resulting from the corresponding superposition S ₂ and a reference signal S ₁ ,

It is e.g. It is known to generate the above-mentioned frequency-shifted partial beams using a Zeeman-stabilized helium-neon laser, wherein the frequency splitting is effected via an externally applied magnetic field. In other approaches, acousto-optic modulators are used for frequency splitting.

A problem that arises here in practice, among other things, is that an additional (zero-point) sensor system is required as a result of the unambiguity range, which in principle is limited to dividers of the respective wavelength for incremental interferometers, for determining the respective zero point determination constructive effort and required in the area of the interferometer or the component to be measured space can be increased.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide an optical system for microlithography and a method for position determination, which allow a highly accurate determination of the absolute position of a component while satisfying the requirements of microlithography and avoiding the problems described above.

This object is achieved according to the features of the independent claims.

• - eine Lichtquelle zur Erzeugung wenigstens eines ersten Teilstrahls mit einer ersten Frequenz (f₁ ); und
• - eine Durchstimm-Einrichtung, über welche eine Einstellung der ersten Frequenz (f₁ ) auf unterschiedliche Werte in einer vorgegebenen Frequenzabfolge durchführbar ist.

An optical system according to the invention for microlithography has at least one component and at least one interferometer associated with this component for position determination with a light source unit, wherein the component is an optical element and wherein the light source unit comprises:

• a light source for generating at least a first partial beam having a first frequency ( f ₁ ); and
• a tuning device, via which a setting of the first frequency ( f ₁ ) to different values in a given frequency sequence is feasible.

The invention is based in particular on the concept of realizing an absolute determination of the position of a component in an optical system for microlithography without the need for additional sensors for establishing a reference position by providing a sufficiently accurate and sufficiently fast frequency tuning of the respective (first) frequency of the first partial beam is realized.

In embodiments, the interferometer is a heterodyne interferometer, wherein the light source is further for generating a second sub-beam having a second frequency ( f ₂ ), the second frequency having a predetermined frequency spacing ( .delta.f ) to the first frequency. However, the invention is not limited to the realization with a heterodyne interferometer, but can also be implemented advantageously with a homodyne interferometer. In the homodyne limiting case (f ₂ → f ₁ ), a sensor device based on the quadrature principle is typically used. Both cases have in common that two derived signals are used in the person skilled in the art for phase acquisition.

The invention is based on the consideration that in the above-mentioned frequency tuning and the corresponding repeated determination or evaluation of the phase of the generated interference signal, an absolute position determination as explained in more detail below can take place from the ratio of the phase change to the frequency change, with the result that that, when tuned with a sufficiently high number of frequency steps, it is possible to deduce the absolute length of an interferometer arm of the interferometer. As a result, according to the invention, a highly accurate distance or position determination can be realized with a measuring range of a few meters (m) down to the picometer range (10 ^-12 m).

The concept according to the invention has the particular advantage that the components of the interferometer present in addition to the light source unit according to the invention, as well as the electronic detection of the interferometer signals, can in principle be kept unchanged, so that an already existing interferometer can be replaced by simply exchanging the light source unit according to the invention can be retrofitted.

In embodiments, when implemented with a heterodyne interferometer, generation of the second frequency f ₂ by polarization filtering from the first (in its frequency f ₁ tunable) partial beam and subsequent frequency shift with an acousto-optic modulator (AOM). In further embodiments, the generation of the first frequency f ₁ and the second frequency f ₂ Also be done via two separate light sources or laser modules, the second light source can be controlled to a fixed frequency distance to the first light source.

According to one embodiment, the setting of the first frequency to different values each with a frequency fluctuation of at most 1000Hz, in particular of at most 100Hz, feasible. The setting of the first frequency to different values can each take place in a response time of typically 10 ms.

According to one embodiment, this setting of the first frequency to different values over a frequency range of at least (10-50) GHz, in particular at least (10-100) GHz, more particularly at least (10-200) GHz, feasible. In principle, by enlarging this (tuning) range, the accuracy of the length measurement can be increased.

According to one embodiment, the adjustment of the first frequency ( f ₁ ) to discrete frequency values in defined frequency steps. These frequency steps may in particular be in the range of (10-100) MHz.

According to one embodiment, in the tuning, a continuous variation of the first frequency takes place within the predetermined frequency range.

According to one embodiment, the tuning results in a variation of the first frequency with a non-linear time dependence. Such an embodiment of the scanning process has, as explained below, the advantage that disturbing linear drift processes can be separated from the actual position information by the frequency sequence set during the frequency tuning being distinguishable from drift-like position changes. In this case, a variation of the first frequency, which is not monotonous in its time dependency and has a sign change, can in particular also take place within the predetermined frequency range.

According to an embodiment, the light source unit further comprises an input interface for inputting the freely programmable frequency sequence to be maintained in the tuning.

According to an embodiment, the light source unit further comprises a reference cavity having a reference atmosphere. As a result, atmospheric influences can be detected.

According to one embodiment, the light source unit has an acousto-optic modulator for setting the desired frequency spacing between the first frequency and the second frequency.

According to one embodiment, the light source has a first laser module for generating the first partial beam with the first frequency and a separate second laser module for generating the second partial beam with the second frequency.

According to one embodiment, the light source has a control for setting the desired frequency spacing between the first frequency and the second frequency. According to one embodiment, the component is assigned six interferometers, each of which enables position determination in one degree of freedom.

In one embodiment, the component is a mirror.

According to one embodiment, the optical system is a microlithographic projection exposure apparatus.

The invention further relates to a method for determining the position of a component in an optical system for microlithography, wherein the component is an optical element and wherein the position determination is performed with at least one interferometer in one degree of freedom, wherein the interferometer is a light source unit with a light source and having a tuning means, wherein on the light source at least a first partial beam having a first frequency (f ₁₎ is generated, wherein the first frequency (f ₁₎ is set to different values in a predetermined frequency sequence.

According to one embodiment, the interferometer is a heterodyne interferometer, wherein a second partial beam with a second frequency (2) is further transmitted via the light source. f ₂ ) is generated, wherein the second frequency has a predetermined frequency spacing (.DELTA.f) to the first frequency.

According to one embodiment, for the different values of the first frequency ( f ₁ ) in each case a determination of the phase of the generated interference signal.

According to one embodiment, the ratio of the phase change to the frequency change for the determination of an absolute length of an interferometer arm of the interferometer is determined in each case for the different values of the first frequency.

According to one embodiment, via a further light source, a third partial beam with a third frequency ( f ₃ ) and a fourth sub-beam having a fourth frequency ( f ₄ ) generated.

According to one embodiment, perturbations occurring in the system using the third sub-beam and the fourth sub-beam are detected.

Further embodiments of the invention are described in the description and the dependent claims.

The invention will be explained in more detail with reference to embodiments shown in the accompanying drawings.

list of figures

• 1 eine schematische Darstellung zur Erläuterung des Aufbaus einer beispielhaften Ausführungsform einer in einem erfindungsgemäßen optischen System vorhandenen Lichtquellen-Einheit;
• 2-3 schematische Darstellung weiterer möglicher Ausführungsformen einer in einem erfindungsgemäßen optischen System vorhandenen Lichtquellen-Einheit;
• 4 eine schematische Darstellung zur Erläuterung einer möglichen Realisierung der Rekonstruktion einer während des erfindungsgemäßen Verfahrens eingestellten Frequenzrampe;
• 5 eine schematische Darstellung zur Erläuterung einer bei dem erfindungsgemäßen Verfahren durchgeführten Auswertung;
• 6 ein Diagramm einer beispielhaften, während des erfindungsgemäßen Verfahrens einstellbaren Frequenzrampe;
• 7-10 schematische Darstellungen zur Erläuterung weiterer Ausführungsformen der Erfindung;
• 11 eine schematische Darstellung zur Erläuterung des prinzipiellen Aufbaus eines Heterodyninterferometers; und
• 12 eine schematische Darstellung zur Erläuterung des möglichen Aufbaus einer für den Betrieb im EUV ausgelegten mikrolithographischen Projektionsbelichtungsanlage.

Show it:

• 1 a schematic representation for explaining the structure of an exemplary embodiment of a light source unit present in an optical system according to the invention;
• 2-3 schematic representation of further possible embodiments of a present in an optical system according to the invention light source unit;
• 4 a schematic representation for explaining a possible realization of the reconstruction of a frequency ramp set during the process according to the invention;
• 5 a schematic representation for explaining an evaluation carried out in the inventive method;
• 6 a diagram of an exemplary, adjustable during the process according to the invention frequency ramp;
• 7-10 schematic representations for explaining further embodiments of the invention;
• 11 a schematic representation for explaining the basic structure of a heterodyne interferometer; and
• 12 a schematic representation for explaining the possible structure of a designed for operation in EUV microlithographic projection exposure apparatus.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

1 shows a schematic representation for explaining the possible structure of a light source unit present in an optical system according to the invention in one embodiment.

Although reference is made to a heterodyne interferometer in the embodiments described below, the invention is not limited to implementation with a heterodyne interferometer, but can also be implemented advantageously with a homodyne interferometer.

The heterodyne interferometer itself as well as the electronic detection of the interferometer signals is in 1 not shown and may basically have any conventional structure.

According to 1 is first about a laser module 10 a first partial beam having a first frequency f ₁ the two required for the heterodyne laser frequencies f ₁ . f ₂ generated. The laser module 10 In the embodiment of FIG 1 an active laser medium 102 , a Auskoppelspiegel 103 and a positionable mirror 101 , via which the resonator length can be adjusted by an electrical signal (for example by a piezoelectric crystal). By this variation of the resonator length, the laser line can be shifted in frequency to a variation or tuning of the frequency f ₁ to reach. This variation or tuning of the frequency f ₁ is also referred to as "scanning" hereinafter. A Fabry-Perot etalon 104 in the laser resonator of the laser module 10 allows the selection of the longitudinal mode and can be adjusted synchronously with the resonator length to prevent mode jumps during the scanning process. The synchronous control of positionable mirror 101 and Fabry-Perot etalon 104 is according to 1 by a common control electronics 110 reached.

In the embodiment of 1 Thus, a tuning means for tuning the first frequency f ₁ over a given frequency range by the control electronics 110 for positioning the mirror 101 for the purpose of setting or varying the resonator length of the laser module 10 educated.

To generate the second frequency f ₂ for the heterodyne method is according to 1 a portion of the first sub-beam of the first frequency f ₁ over two beam splitters 204 . 205 decoupled and with the help of an acousto-optic modulator (AOM) 206 in its frequency by the amount of an externally supplied reference frequency 207 pushed. The setting of orthogonal polarization states of the two partial beams with frequency required for the heterodyne interferometer f ₁ or f ₂ = f ₁ + Δf is through a lambda / 2 plate 210 causes, which according to 1 in the beam path of the first partial beam of the first frequency f ₁ located.

In embodiments, in addition, an absorption cell with photodetector 209 be used as a frequency reference. Here, a partial beam is branched off for frequency analysis, then through a cavity filled with a gas 208a and finally brought to an intensity-measuring detector (eg photodiode). The referencing to the known in frequency absorption lines (eg iodine cell, methane cell, ...) of the in the cavity 208a filled gas allows the highly accurate determination of the present frequency.

According to 2 can also be a Fabry-Perot interferometer 208b high quality with its frequency sufficiently sharp and equidistant transmission lines instead of the absorption cell for frequency referencing. Alternatively, the use of a frequency comb for frequency referencing, which may include feedback to (atomic clock) frequency standards, is also possible.

Further possible embodiments for realizing the tuning of the frequency according to the invention f ₁ (ie, the scanning process) will be described with reference to FIG. 3ff. described.

By the above-described variation or tuning of the frequency f ₁ can be realized as explained below, an absolute measurement in the heterodyne interferometer:

If we denote a control parameter indicative of the tuning with q, the values of the frequencies set during this tuning apply f ₁ . f ₂ : $$\begin{array}{l}{f}_1\left(q\right)=f_{1.0}+\Delta f_1\left(q\right).\\ f_2\left(q\right)=f_1\left(q\right)+f_b\end{array}$$

The tuning starts at the zero frequencies f _1.0 and f _1.0 + f _b with the common displacement Δf ₁ (q), where f _b denotes the heterodyne and constant beat frequency (= "beat frequency").

For the dependence of the phase change of the frequency change is neglecting disturbances, the following relationship, which in principle on the ratio of the phase change to the frequency change, a determination of the absolute length L, as in 5 shown for exemplary values, allows: $$\Delta \phi \left(q\right)=L\frac{2\pi }{c}\Delta f_1\left(q\right)$$

Therein C stands for the speed of light and Δφ for the phase change associated with the frequency change Δf ₁ (q), which is measured and recorded with the detector and the downstream phase evaluation unit.

The frequency tuning requires in reality a finite time interval T _S. Over this time interval, length changes due to unavoidable vibration-like and drift-like variations in the measurement configuration become disturbances. Including these perturbations, Equation (2) is modified to: $$\Delta \phi \left(q\right)=L\frac{2\pi }{c}\Delta f_1\left(q\right)+\frac{2\pi f_1\left(q\right)}{c}\delta L,$$

wobei die driftartigen Variationen durch ein Polynom in der Zeit bis zur Ordnung K mit entsprechenden Koeffizienten d₁,...,d_K beschrieben werden und die innerhalb der Durchstimmzeit als solche wahrnehmbaren Vibrationen als Superposition einer endlichen Anzahl J von harmonischen Oszillationen mit den entsprechenden Amplituden a_j und b_j sowie den Kreisfrequenzen Ω_j. The perturbations concerned can be represented mathematically in the time domain as follows: $$\delta L\left(t\right)=d_{1}t+d_2{t}^2+...+d_K{t}^K+{\displaystyle \underset{j=1}{\overset{J}{\Sigma }}\left(a_j\text{sin}\left({\mathrm{\Omega }}_{j}t\right)+b_j\text{cos}\left({\mathrm{\Omega }}_{j}t\right)\right).}$$

the drift-like variations being described by a polynomial in time to order K with corresponding coefficients d ₁ , ..., d _K and the vibrations perceptible as such within the tuning time as superposition of a finite number J of harmonic oscillations with the corresponding amplitudes a _j and b _j and the angular frequencies Ω _j .

The combination of equations (3) and (4) in the time domain finally forms the basis of the method for absolute measurement of the distance by means of frequency tuning: $$\begin{array}{c}\Delta \phi \left(q\left(t\right)\right)=L\frac{2\pi }{c}\Delta f_1\left(q\left(t\right)\right)+\frac{2\pi f_1\left(q\left(t\right)\right)}{c}\left(d_{1}t+d_2{t}^2+...+d_K{t}^K\right)+...\\ +\frac{2\pi f_1\left(q\left(t\right)\right)}{c}{\displaystyle \underset{j=1}{\overset{J}{\Sigma }}\left(a_j\text{sin}\left({\mathrm{\Omega }}_{j}t\right)+b_j\text{cos}\left({\mathrm{\Omega }}_{j}t\right)\right)},\end{array}$$

mit einer von f_b verschiedenen Schwebungsfrequenz $f_b^{\text{'}}$

zur Beobachtung und Feststellung der Störungen zwecks anschließender Korrektur. Beide Verfahren werden im Folgenden kurz beschrieben.To master the disturbances δL (q (t)) occurring during the tuning, there are in principle two possible solutions. The first approach is based on a sufficiently complete modeling of the disturbances and their control by filtering. The second approach uses another heterodyne measuring beam with fixed frequencies $f_3\text{and}{f}_4=f_3+f_b^{\text{'}}$

with a beating frequency different from f _b $f_b^{\text{'}}$

to observe and determine the disturbances for subsequent correction. Both methods are briefly described below.

mit $$\begin{array}{cccc}\tau =\frac{t-t_0}{T_S},& P=\frac{\Delta \phi }{2\pi },& \alpha =\frac{L}{c},& \begin{array}{cc}Q\left(\tau \right)=\frac{f_1\left(\tau \right)}{c},& drif{t}_k\left(\tau \right)={\tau }^k,\end{array}\end{array}$$

$$mod{e}_{j,s}\left(\tau \right)=\{\begin{array}{l}Q\left(\tau \right)\text{sin}\left({\mathrm{\Omega }}_j\tau \right),s=1\\ Q\left(\tau \right)\tau \text{sin}\left({\mathrm{\Omega }}_j\tau \right),s=2\\ Q\left(\tau \right)\text{cos}\left({\mathrm{\Omega }}_j\tau \right),s=3\\ Q\left(\tau \right)\tau \text{cos}\left({\mathrm{\Omega }}_j\tau \right),s=4\end{array}$$

und den entsprechend an die Notation angepassten Amplituden d_k, m_j,s und Frequenzen Ω_j der Störbeiträge. Vorteilhafterweise werden noch, um Unsicherheiten in den Frequenzen der Störoszillationen zu begegnen, die Ableitungen der Sinus- bzw. Kosinus-Terme nach der Frequenz mit im Modell der Moden aufgenommen (siehe s = 2, s = 4).In the elimination of the disturbances by modeling and filtering a sufficient knowledge of the occurring disturbances in the form of a model according to equation (5) is required. This knowledge can, for example, be obtained in advance by recording the measured phases for a non-excited and nominally stationary system over a sufficiently long observation period and analyzing the frequency content of the position changes derived from the measured phase changes (Fourier analysis). This pre-characterization is advantageously provided as a separate measurement mode (vibrometer functionality) and executed in advance of the frequency tuning during the start procedure. Knowing the model of the disturbances, the free parameters in the form of the amplitudes and angular frequencies then remain to be determined in order to eliminate the disturbances in the reconstruction of the measuring length from the recorded measured phases. In constructing the model, the terms in equation (5) are simplified by suitable substitutions, and one arrives at the compact representation $$P\left(\tau \right)=\alpha scan\left(\tau \right)+{\displaystyle \underset{\text{k}=1}{\overset{\text{K}}{\Sigma }}drif{t}_k\left(\tau \right)d_k+}{\displaystyle \underset{j=1}{\overset{J}{\Sigma }}{\displaystyle \underset{s}{\Sigma }mOd{e}_{j.s}\left(\tau \right)m_{j.s}}}$$

With $$\begin{array}{cccc}\tau =\frac{t-t_0}{T_S}.& P=\frac{\Delta \phi }{2\pi }.& \alpha =\frac{L}{c}.& \begin{array}{cc}Q\left(\tau \right)=\frac{f_1\left(\tau \right)}{c}.& drif{t}_k\left(\tau \right)={\tau }^k.\end{array}\end{array}$$

$$mOd{e}_{j.s}\left(\tau \right)=\{\begin{array}{l}Q\left(\tau \right)\text{sin}\left({\mathrm{\Omega }}_j\tau \right).s=1\\ Q\left(\tau \right)\tau \text{sin}\left({\mathrm{\Omega }}_j\tau \right).s=2\\ Q\left(\tau \right)\text{cos}\left({\mathrm{\Omega }}_j\tau \right).s=3\\ Q\left(\tau \right)\tau \text{cos}\left({\mathrm{\Omega }}_j\tau \right).s=4\end{array}$$

and the correspondingly adapted to the notation amplitudes d _k , m _{j, s} and frequencies Ω _{j of} the interference contributions. Advantageously, in order to counteract uncertainties in the frequencies of the spurious oscillations, the derivatives of the sine or cosine terms are also included according to the frequency in the model of the modes (see s = 2, s = 4).

derart gestaltbar (programmierbar), dass sie idealerweise orthogonal zu den Störbeiträgen ist oder der Überlapp mit diesen zumindest minimal ist. Hierzu kann beispielsweise - anstelle eines Scanvorgangs mit linearer Zeitabhängigkeit - bei der Durchstimmung eine nicht-monotone bzw. einen Vorzeichen-Wechsel aufweisende Variation der ersten Frequenz f₁ innerhalb des vorgegebenen Frequenzbereichs gemäß 6 realisiert werden. 6 zeigt ein Bespiel für die zeitliche Form der Frequenz-Durchstimmung in Form einer Dreiecksrampe mit insgesamt 2N sowohl in der Zeit als auch in der Frequenz äquidistanten diskreten Schritten $$f_1\left(t_0+n{\Delta }_t\right)=\{\begin{array}{c}{f}_{\mathrm{1,0}}+{\Delta }_{f}n,n=\mathrm{0,1,2,}\dots ,N\\ f_{\mathrm{1,0}}+{\Delta }_f\left(N-n\right),n=N+\mathrm{1,}N+\mathrm{2,}\dots \mathrm{,2}N\end{array},$$

wobei Δ_f und Δ_t das Frequenz- bzw. Zeitinkrement bei der Durchstimmung bezeichnen. Damit wird die Durchstimmung insbesondere im Gegensatz zu einer in der Zeit über das Messintervall linearen Durchstimmung unterscheidbar von einer zeitlich linearen Drift, welche erfahrungsgemäß die Störungen in den meisten Fällen dominiert. Erfindungsgemäß kann die Störanfälligkeit bzw. Anfälligkeit auf Drift-Vorgänge und Vibrationen durch die Wahl einer geeigneten Frequenzabfolge reduziert werden. In Ausführungsformen kann eine Eingabeeinheit (Schnittstelle) zur Eingabe einer bei der Durchstimmung einzuhaltenden Frequenzabfolge vorgesehen sein.Preferably, the temporal shape of the frequency tuning scan (τ) = Δf ₁ (q (τ)) is chosen such that the actually interesting portion of the phase change due to the frequency change as well as possible from the disturbances is separable. According to the invention, the frequency tuning is in the sense of the scalar product between functions $$\begin{array}{l}<scan.drif{t}_k>={\displaystyle {\int }_0^{1}d\tau scan\left(\tau \right)}drif{t}_k\left(\tau \right)\to \text{min}.\forall k.\\ <scan.mOd{e}_{j.s}>={\displaystyle {\int }_0^{1}d\tau scan\left(\tau \right)}mOd{e}_{j.s}\left(\tau \right)\to \text{min}.\forall j.s\end{array}$$

be designed (programmable) so that it is ideally orthogonal to the interference contributions or the overlap with these is at least minimal. For this purpose, for example - instead of a scan with linear time dependence - in the tuning a non-monotonous or a sign change having variation of the first frequency f ₁ within the predetermined frequency range according to 6 will be realized. 6 shows an example of the temporal form of frequency tuning in the form of a triangular ramp with a total of 2N in both the time and in the frequency equidistant discrete steps $$f_1\left({\mathrm{<figure-callout\; id="0"\; label="t"\; filenames="DE102017220408A1\_0027.png,DE102017220408A1\_0028.png"\; state="\{\{state\}\}">t</figure-callout>}}_0+n{\Delta }_t\right)=\{\begin{array}{c}{f}_{1.0}+{\Delta }_{f}n.n=\mathrm{0,1,2,}....N\\ f_{1.0}+{\Delta }_f\left(N-n\right).n=N+\mathrm{1,}N+\mathrm{2,}...2N\end{array}.$$

where Δ _f and Δ _t denote the frequency or time increment in the tuning. In particular, in contrast to a linear tuning that is linear over the measurement interval, the tuning becomes distinguishable from a temporally linear drift, which experience has shown that the interference predominates in most cases. According to the invention, the susceptibility or susceptibility to drift processes and vibrations can be reduced by the choice of a suitable frequency sequence. In embodiments, an input unit (interface) may be provided for inputting a frequency sequence to be maintained during the tuning.

The filtering of the disturbances succeeds, for example, by using the method of least error squares. For this purpose, a frequency tuning is performed, and the set frequency changes Δf ₁ = scan , the measured phase changes P and the normalized measurement times τ are each stored as column vectors. With the recorded measurement times, the interference patterns identified in advance with the unknown amplitude to be determined are calculated over the tuning interval. Thus, equation (6) in its discretized form yields: $$\underset{\_}{P}=\underset{\_}{scan}\alpha +{\displaystyle \underset{\text{k}=1}{\overset{\text{K}}{\Sigma }}{\underset{\_}{drift}}_k{d}_k+{\displaystyle \underset{j=1}{\overset{J}{\Sigma }}{\displaystyle \underset{s}{\Sigma }{\underset{\_}{mOde}}_{j.s}{m}_{j.s},}}}$$

und nach Stapelung sämtlicher Parameter zu einem Spaltenvektor u gemäß $$\underset{\_}{P}=\underset{\underset{\_}{\underset{\_}{sDM}}}{\underbrace{\left(\underset{\_}{scan}\underset{\_}{\underset{\_}{Drifts}}\underset{\_}{\underset{\_}{Modes}}\right)}}\underset{\underset{\_}{u}}{\underbrace{\left(\begin{array}{c}\alpha \\ \underset{\_}{d}\\ \underset{\_}{m}\end{array}\right)}}$$

(die Notation erschließt sich dem Fachmann unmittelbar aus Gleichung (9)) erhält man die im Rahmen der Stichprobe beste Schätzung der Parameter durch die Standardlösung überbestimmter linearer Gleichungssysteme zu $$\left(\begin{array}{c}\alpha \\ \underset{\_}{d}\\ \underset{\_}{m}\end{array}\right)=\text{inv}\left({\underset{\_}{\underset{\_}{sDM}}}^T\underset{\_}{\underset{\_}{sDM}}\right){\underset{\_}{\underset{\_}{sDM}}}^T\underset{\_}{P}.$$

This sets the measured phases in a linear relationship with the known calculated patterns of the interference contributions and their unknown and to be determined coefficients. After merging all pattern vectors into a matrix $\underset{\_}{\underset{\_}{sDM}}$

and after stacking all the parameters into a column vector u according to $$\underset{\_}{P}=\underset{\underset{\_}{\underset{\_}{sDM}}}{\underset{\}}{\left(\underset{\_}{scan}\underset{\_}{\underset{\_}{Drifts}}\underset{\_}{\underset{\_}{MOdes}}\right)}}\underset{\underset{\_}{u}}{\underset{\}}{\left(\begin{array}{c}\alpha \\ \underset{\_}{d}\\ \underset{\_}{m}\end{array}\right)}}$$

(the notation is directly understood by the skilled person from equation (9)) one obtains the best estimate of the parameters in the sample by the standard solution of overdetermined linear systems of equations $$\left(\begin{array}{c}\alpha \\ \underset{\_}{d}\\ \underset{\_}{m}\end{array}\right)=\text{inv}\left({\underset{\_}{\underset{\_}{sDM}}}^T\underset{\_}{\underset{\_}{sDM}}\right){\underset{\_}{\underset{\_}{sDM}}}^T\underset{\_}{P},$$

legt die zu bestimmende absolute Länge ohne Störungen frei (die Störbeiträge werden durch die Freigabe ihrer Koeffizienten bei der Rekonstruktion gewissermaßen „kurzgeschlossen“).The ultimately interesting parameter $\alpha =\frac{L}{c}$

sets the absolute length to be determined without disturbances (the interference contributions are to a certain extent "short-circuited" by the release of their coefficients during the reconstruction).

The execution of the model of the disturbances in linear form is only an example. Non-linear dependencies of the perturbations on the parameters to be determined can also be established in a non-linear optimization (eg, iterative Newton method), and then the parameters can be determined by non-linear parameter adaptation. This can be useful and necessary, for example, if the frequencies of the oscillations are variable in time so that they can be determined sufficiently accurately for the respective time interval of tuning only by appropriate consideration in the model of the disturbances.

möglichst nahe kommen. Beispielsweise sollte bei einer erforderlichen Frequenz-Stabilität von 100Hz eine Einstellzeit für einen Frequenzschritt von nicht wesentlich länger als etwa 10ms realisiert werden.In order to keep the number of oscillations to be considered as low as possible and in order to be able to identify high-frequency oscillations as such, the frequency tuning is preferably carried out within as short a time as possible. Ideally, in the special case of a discrete stepwise tuning, the response time τ should be for a frequency increment Δf of the limit based on the uncertainty relation of the Fourier transformation τ ≥ (1 / Δf) $\mathrm{\tau }\ge \frac{1}{\mathrm{\Delta }\text{f}}$

get as close as possible. For example, with a required frequency stability of 100 Hz, a setup time for a frequency step of not much longer than about 10 ms should be realized.

By accepting the higher expenditure on equipment a mastery of the disturbances is also expediently possible by applying these disturbances to a further, fixed frequency heterodyne measuring beam ( f ₃ . f ₄ ), which ideally with the tunable heterodyne measuring beam ( f ₁ . f ₂ ) is coupled into the interferometer detected. This is in 7 schematically illustrated, wherein the further heterodyne measuring beam ( f ₃ . f ₄ ) the frequencies f ₃ and f ₄ = f ₃ + Δf ₃ . In 7 is with " 60 "The (in addition to the analogue 1 trained and shown here simplified tunable light source 10 provided) further light source 60 , With " 606 "An acousto-optic modulator, with" 607 "An externally supplied reference frequency and with" 610 "A lambda / 2 plate called. For the simultaneously measured phase values Δφ _1,2 and Δφ _3,4 for the two heterodyne measuring beams ( f ₁ . f ₂ ) respectively. ( f ₃ . f ₄ ) apply the two relationships $$\begin{array}{l}\Delta {\phi }_{1.2}\left(q\right)=L\frac{2\pi }{c}\Delta f_1\left(q\right)+\frac{2\pi f_1\left(q\right)}{c}\delta L\left(t\left(q\right)\right).\\ \Delta {\phi }_{3.4}\left(q\right)=\frac{2\pi f_3}{c}\delta L\left(t\left(q\right)\right),\end{array}$$

zur Messung der Absolut-Länge mittels Frequenz-Durchstimmung bei expliziter Störungserfassung, wobei $$\Delta {\phi }_c\left(q\right)=\Delta {\phi }_{\mathrm{1,2}}\left(q\right)-\frac{f_1\left(q\right)}{f_3}\Delta {\phi }_{\mathrm{3,4}}\left(q\right)$$

die um die Störungen korrigierte gemessene Phasenverschiebung bezeichnet.Elementary transformation finally leads to the determining equation $$\Delta {\phi }_c\left(q\right)=L\frac{2\pi }{c}\Delta f_1\left(q\right)$$

for measuring the absolute length by means of frequency tuning for explicit interference detection, wherein $$\Delta {\phi }_c\left(q\right)=\Delta {\phi }_{1.2}\left(q\right)-\frac{f_1\left(q\right)}{f_3}\Delta {\phi }_{3.4}\left(q\right)$$

denotes the measured phase shift corrected for the perturbations.

From the plot of the measured and corrected phase shift Δφ _c (q) against the frequency shift Δf ₁ (q), the absolute length L to be determined in the form of the slope, as in FIG 5 for exemplary values.

In an advantageous manner, the method of frequency tuning with an additional heterodyne observation beam, in contrast to the method of interference detection by modeling and filtering, in principle, no limiting requirements for duration and shape of the frequency tuning.

The invention is not further limited with respect to the selected wavelength range, with the wavelength being only in the range of (400-1600) nm by way of example. Furthermore, the invention with respect to the specific embodiment of the (heterodyne) interferometer is not further limited, which can be configured in a conventional manner with plane mirrors or retroreflectors.

In the following, further, as an alternative to 1 and 2 possible embodiments for the realization of tuning the frequency according to the invention f ₁ (ie, the "scan") with reference to FIG. 3ff. described.

According to 3 unlike the embodiments of 1 and 2 two separate laser modules 10 and 11 used to the frequencies required for heterodyne operation f ₁ . f ₂ to provide. The structure and function of the two laser modules correspond to those of the laser module 10 out 1 ,

The frequency spacing between the frequencies f ₁ and f ₂ is according to 3 by regulation of the second laser module 11 to a constant frequency spacing Δf to the first laser module 10 achieved, leaving the first laser module 10 a partial beam with the frequency f ₁ and the second laser module 11 generates a partial beam with the frequency f ₂ = f ₁ + .DELTA.f. For this purpose, portions of both partial beams or frequencies via beam splitters 204 and 215 on a photodetector 211 brought to interference. The beat signal of both frequencies f ₁ and f ₂ gets at a mixer 213 with an external reference frequency 212 compared. The comparison signal is used to control the second frequency f ₂ a laser control 120 fed. In the embodiment according to 3 This control of the second frequency takes place f ₂ faster than the shift of the (first) frequency f ₁ through the scanning process. The setting of orthogonal polarization states of the two partial beams with frequency required for the heterodyne interferometer f ₁ respectively. f ₂ is according to 3 also by a lambda / 2 plate 210 set.

In order to obtain information about the exact shape of the frequency ramp during the scanning process, it is explained in 4 parallel to the actual heterodyne interferometer 400 (with components 401 to 409 ) Part of the light intensity of both frequencies f ₁ . f ₂ via a beam splitter 301 and a deflecting mirror 302 separated and a second heterodyne interferometer 500 (its components with " 501 " to " 510 "Are designated) supplied.

The first heterodyne interferometer 400 has polarizers 402 . 408 (each in 45 ° position), a photodetector 403 for the reference signal, a photodetector 409 for the interferometer signal, a polarization beam splitter 404 , a retro reflector 405 in the reference arm, a retroreflector 406 in the measuring arm and beam splitter 401 . 407 on. The second heterodyne interferometer 500 has polarizers 502 . 508 (each in 45 ° position), a photodetector 503 for the reference signal, a photodetector 509 for the interferometer signal, a polarization beam splitter 504 , a retro reflector 505 in the reference arm, a retroreflector 506 in the measuring arm and beam splitter 501 . 507 on. Unlike the first heterodyne interferometer 400 is the distance of the retroreflector 506 in the second heterodyne interferometer 500 over a mechanical reference length 510 (as a length standard) fixed. The evaluation of the interferometer signals of this second heterodyne interferometer 500 allows a reconstruction of the frequency ramp.

aus der an ihm gemessenen Phasenänderung Δφ₀ bestimmt.In the embodiment of 4 becomes the frequency change by means of the second heterodyne interferometer 500 , which has an invariable drift and vibration-free cavity of the (eg by calibration) known length L ₀ , according to $\Delta f_1=\frac{c}{2\pi }\frac{\Delta {\phi }_0}{{\mathrm{<figure-callout\; id="0"\; label="L"\; filenames="DE102017220408A1\_0027.png,DE102017220408A1\_0028.png"\; state="\{\{state\}\}">L</figure-callout>}}_0}$

determined from the phase change Δφ ₀ measured on it.

8th shows a schematic representation of a possible arrangement of inventive heterodyne interferometer for determining the position of a mirror 801 , At the mirror 801 In the exemplary embodiment, this is a mirror of a projection objective in a microlithographic Projection exposure system, wherein - without the invention being limited thereto - for example, from US 6,864,988 B2 is based on known structure, in which both a load-bearing support structure 803 ("Force frame") as well as an independently provided for this measurement structure 804 ("Sensor frame") are present. According to 8th are both support structure 803 as well as measurement structure 804 independently of one another via mechanical connections (eg springs) acting as dynamic decoupling 805 respectively. 806 to a base plate or base 830 mechanically connected to the optical system. The mirror 801 in turn is about a mirror mount 802 on the supporting structure 803 attached. In 8th schematically drawn are two heterodyne interferometers according to the invention 810 and 820 whose measuring path 811 respectively. 821 from the measuring structure 804 to the mirror 801 runs.

To measure the position of a component or a mirror in all six degrees of freedom, six heterodyne interferometers are required, one possible configuration being shown schematically in FIG 9a is shown. Marked are six measuring sections 905 with one each on the measuring frame 906 located starting point 904 and one at the mirror 901 located endpoint 903 , Like also in 9a indicated is located at the starting point 904 the measuring section 905 a polarization-dependent beam splitter 912 and retroreflector 913 constructed interferometer, and at the end point 903 the measuring section 905 one at the mirror 901 arranged retroreflector 911 , 9b is merely illustrative of the simplified representation of the interferometer structure 9a ,

10 shows a schematic representation of a possible arrangement of inventive heterodyne interferometer for determining the position of a component 1050 in the form of a mask table or a wafer table in a microlithographic projection exposure apparatus, wherein the component has mirrored side surfaces 1051 . 1052 for interferometric "measurement" has. According to 10 are to this interferometric measurement or determination of the measuring sections L _x and L _y each from polarization-dependent beam splitter 1001 respectively. 1011 and retroreflectors 1002 . 1003 respectively. 1012 . 1013 constructed interferometer provided. Further, similar to those described above with reference to 1-9 described embodiments with " 1008 " respectively. " 1018 "Beam splitter, with" 1007 " respectively. " 1017 "Photodetectors for the interferometer signal and with" 1010 " respectively. " 1020 "Photodetectors for the reference signal called. With " 1006 "," 1009 " respectively. " 1016 "," 1019 Are each analogous to the embodiments described above polarizers (each in 45 ° position) referred to. With " 1005 " respectively. " 1015 "Is in each case a deflection mirror, and with" 1004 "or" 1014 "Each denotes a lambda / 2 plate.

12 shows a schematic representation of an exemplary designed for operation in EUV microlithographic projection exposure apparatus. The (heterodyne) interferometer (s) according to the invention can be used in this projection exposure apparatus for measuring the distance of the wafer or the wafer stage, the reticle plane and the individual mirrors in the projection objective or in the illumination device. However, the invention is not limited to the application in systems designed for operation in the EUV, but also in the measurement of optical systems for other operating wavelengths (eg in the VUV range or at wavelengths less than 250nm) feasible.

According to 12 has a lighting device in a designed for EUV projection exposure system 1200 a field facet mirror 1203 and a pupil facet mirror 1204 on. On the field facet mirror 1203 becomes the light of a light source unit, which is a plasma light source 1201 and a collector mirror 1202 includes, steered. In the light path after the pupil facet mirror 1204 are a first telescope mirror 1205 and a second telescope mirror 1206 arranged. In the light path below is a deflection mirror 1207 arranged, which reflects the radiation impinging on an object field in the object plane of a six mirror 1251 - 1256 comprehensive projection lens steers. At the location of the object field is a reflective structure-bearing mask 1221 on a mask table 1220 arranged, which is imaged by means of the projection lens in an image plane in which a substrate coated with a photosensitive layer (photoresist) 1261 on a wafer table 1260 located.

The invention can be used for the absolute measurement of the position, for example, of one of the mirrors 1251 - 1256 or the mask table 1220 or the wafer table 1260 in the projection exposure machine 1200 be used.

In other applications, the invention can also be realized in a mask inspection system or a wafer inspection system.

Although the invention has also been described with reference to specific embodiments, numerous variations and alternative embodiments will be apparent to those skilled in the art, eg, by combining and / or replacing features of individual embodiments. Accordingly, it is understood for the Those skilled in the art will appreciate that such variations and alternative embodiments are included in the present invention, and the scope of the invention is limited only in terms of the appended claims and their equivalents.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• US 6864988 B2 [0074]

Claims (23)

Optical system for microlithography, comprising at least one component and at least one of these components for position determination interferometer with a light source unit, wherein the component is an optical element and wherein the light source unit comprises: • a light source for generating at least a first partial beam with a first frequency (f1); and • a tuning means, via which a setting of the first frequency (f ₁₎ is carried out to different values in a predetermined frequency sequence. Optical system after Claim 1 , characterized in that this setting of the first frequency (f ₁ ) to different values each with a frequency fluctuation of at most 1000Hz, in particular of at most 100Hz, is feasible. Optical system after Claim 1 or 2 , characterized in that this setting of the first frequency (f ₁ ) to different values over a frequency range of at least (10-50) GHz, in particular of at least (10-100) GHz, more particularly of at least (10-200) GHz, is feasible. Optical system according to one of Claims 1 to 3 , characterized in that in this tuning the adjustment of the first frequency (f ₁ ) to discrete frequency values in defined frequency steps. Optical system after Claim 4 , characterized in that said frequency steps are in the range of (10-100) MHz. Optical system according to one of Claims 1 to 3 , characterized in that in this tuning a continuous variation of the first frequency (f ₁ ) takes place within the predetermined frequency range. Optical system according to one of the preceding claims, characterized in that in this tuning, the variation of the first frequency (f ₁ ) takes place with a non-linear time dependence. Optical system according to one of the preceding claims, characterized in that the light source unit further comprises an input interface for inputting the freely programmable frequency sequence to be maintained during the tuning. Optical system according to one of the preceding claims, characterized in that the light source unit further comprises a reference cavity having a reference atmosphere. Optical system according to one of the preceding claims, characterized in that the interferometer is a heterodyne interferometer, wherein the light source is further configured to generate a second partial beam having a second frequency (f ₂ ), wherein the second frequency a predetermined frequency spacing (Δf) to the first Owns frequency. Optical system after Claim 10 , characterized in that the light source unit comprises an acousto-optic modulator (206) for setting the desired frequency spacing (Δf) between the first frequency (f ₁ ) and the second frequency (f ₂ ). Optical system after Claim 10 or 11 , characterized in that the light source comprises a first laser module (10) for generating the first partial beam at the first frequency (f ₁ ) and a separate second laser module (11) for generating the second partial beam at the second frequency (f ₂ ). Optical system after Claim 12 , characterized in that the light source unit has a control for setting the desired frequency spacing (Δf) between the first frequency (f ₁ ) and the second frequency (f ₂ ). Optical system according to one of the preceding claims, characterized in that the component is assigned to six interferometers, each of which enables a position determination in each case one degree of freedom. Optical system according to one of Claims 1 to 14 , characterized in that the component is a mirror. Optical system according to one of the preceding claims, characterized in that this is a microlithographic projection exposure apparatus. A method for determining the position of a component in an optical system for microlithography, wherein the component is an optical element and wherein the position determination is performed with at least one interferometer in one degree of freedom, the interferometer a light source unit with a light source and a tuning device wherein the first frequency (f ₁₎ is set to different values in a predetermined frequency sequence, wherein the light source via at least a first partial beam having a first frequency (f ₁₎ is generated. Method according to Claim 17 , characterized in that the interferometer is a heterodyne interferometer, wherein on the light source further a second partial beam having a second frequency (f ₂ ) is generated, wherein the second frequency has a predetermined frequency spacing (Δf) to the first frequency. Method according to Claim 17 or 18 , characterized in that for each of these different values of the first frequency (f ₁ ) there is a determination of the phase of the generated interference signal. Method according to Claim 19 , characterized in that for each of these different values of the first frequency (f ₁ ), the ratio of the phase change to the frequency change is determined to determine an absolute length of an interferometer arm of the interferometer. Method according to one of Claims 17 to 20 , characterized in that a third partial beam having a third frequency (f ₃ ) and a fourth partial beam having a fourth frequency (f ₄ ) are generated via a further light source. Method according to Claim 21 , characterized in that are detected using the third sub-beam and the fourth sub-beam interference occurring in the system. Method according to one of Claims 17 to 22 , characterized in that the optical system is an optical system according to any one of Claims 1 to 16 is.

Images