Present Status of the ASET At-Wavelength Phase-Shifting Point Diffraction Interferometer

Size: px

Start display at page:

Download "Present Status of the ASET At-Wavelength Phase-Shifting Point Diffraction Interferometer"

Lisa Riley
5 years ago
Views:

$Present Status of the ASET At-Wavelength Phase-Shifting Point Diffraction Interferometer Katsumi Sugisaki Yucong Zhu a Yoshio Gomei amasahito Niibe b Takeo Watanabe b Hiroo Kinoshita b a Association$

1 Present Status of the ASET At-Wavelength Phase-Shifting Point Diffraction Interferometer Katsumi Sugisaki Yucong Zhu a Yoshio Gomei amasahito Niibe b Takeo Watanabe b Hiroo Kinoshita b a Association of Super-Advanced Electronics Technology, Atsugi-shi, Kanagawa , Japan b Himeji Institute of Technology, Ako-gun, Hyogo , Japan ABSTRACT The precise alignment of Extreme Ultra-Violet Lithography (EUVL) imaging system is necessary in order to achieve diffraction-limited performance. Interferometric testing at the exposure wavelength is needed to ensure proper alignment and to achieve an acceptable fmal wavefront. We have built a prototype at-wavelength interferometer at the NewSUBARU facility. This interferometer is a phase-shifting point diffraction interferometer (PS/PDI) testing specially constructed Schwarzschild optics. Preliminary experiments using visible light were performed in order to learn the PS/PD!. The Schwarzschild optics were aligned using visible wavefront measurements with the interferometer. The precision of the visible measurements was evaluated. Experiments using EUV radiation have been started. Keywords: Optical testing, interferometry, phase-shifting point diffraction interferometer, EUV lithography 1. INTRODUCTION In order to achieve the required image quality, an EUVL optical system must be very precisely aligned. The allowable wavefront error of the optical system must be less than mm RMS for diffraction-limited imaging. The most practical means of ensuring proper alignment is through interferometric analysis of the wavefront emerging from the system. Interferometric data taken at many different points within the field can then be analyzed to determine the adjustments needed to bring the system into proper alignment. Since EUVL systems consist only of reflective elements, the interferometric testing can be done using either visible radiation (such as that produced by a He-Ne laser) or "at-wavelength" using 13.5nm radiation. A visible light interferometer suitable for testing the EUVL projection system has been developed at Lawrence Livermore National ry' Nevertheless, the use of visible radiation for interferometric testing has several drawbacks. First, the accuracy of the system (measured in a part of waves) must be much higher than when using 13.5nm radiation due to the difference in wavelength. For instance, to achieve an accuracy of O.25nm rms a visible light interferometer must have a resolution of /2500. For the same accuracy, an at-wavelength interferometer must achieve a resolution of/50. Second, visible radiation is reflected off of the top surface of the reflective multi-layer coating, leaving the structure of the multi-layer coating un-probed. Since errors in the multilayer can lead directly to errors in the wavefront, it is important that the system be tested at the exposure wavelength. Testing at-wavelength, however, poses significant problems since there can be no refractive material used. To solve this problem, a variation of the common Point Diffraction Interferometer (PD!) is used. This variation, the Phase Shifting Point Diffraction Interferometer (PS/PDI),25 has several distinguishing features. The system is relatively simple and there is no refractive material needed. As the name implies, the PS/PD! provides standard phase shifting capability, simplifying the analysis of the wavefront. In addition, the phase-shifting motion is determined by the geometry of the system and not the wavelength. Thus the motion is relatively large (in this case 3.75 imistep) and easy to make accurately. * Correspondence: sugisakieuv.aset-unet.ocn.ne.jp Soft X-Ray and EUV Imaging Systems, Winfried M. Kaiser, Richard H. Stulen, Editors, Proceedings of SPIE Vol (2000) 2000 SPIE X/00/$

2 In order to conduct a proof of concept, and to gain experience in the operation and manufacture of the at-wavelength interferometry, we have started to develop an interferometer called as the interferometric test stand (ITS).6 The ITS has been constructed at the undulator beam line at the NewSUBARU synchrotron radiation source. In this paper, we report on the present status of our system. Preliminary testing using visible radiation has been performed with the ITS. The undulator has been operating. We have started the at-wavelength experiments. 2. INTERFEROMETER TEST STAND (ITS) The ITS is designed as PS/PDI using radiations emitted from the undulator. The ITS consists of an initial pinhole, a diffraction grating, the optical element under test, a pinhole mask and a CCD camera. Figure 1 shows the concept of the PS/PDI. Initial Binary Test Pinhole Pinhole Grating Optic Mask Test Wavefront Pinhole Mask Camera Fig. I Concept ofthe PSIPDI. The initial pinhole is an optically small pinhole that diffracts light from the source into an essentially perfect diverging beam. It is placed at the object plane (mask side) of the imaging system. The grating is a coarse, freestanding binary diffraction grating that is placed before the optic under test. The pinhole mask is a freestanding mask, located at the image plane of the test optic (wafer-side), with an optically small pinhole next to a relatively large window. The light from the initial pinhole is diffracted by the grating into multiple diffraction orders. The system is aligned such that the un-diffracted 0th order falls on the pinhole of the pinhole mask, and the 1st order falls in the center of the window. All other orders are blocked by the mask. The 1st order passes through the window relatively unchanged while the 0th order is diffracted into a relatively perfect expanding beam. The two beams interfere and are imaged onto the CCD. By translating the grating perpendicularly to the grating lines, a relative phase-shift between the O and 1st orders can be achieved. The ITS is designed to be able to test by both visible and EUV radiations in order to learn about PS/PDI concept and compare the results using visible light and EUV radiations. The system is illustrated in Fig. 2. We prepared two sets of optical components. Each specifications are listed in table I. The window size of the pinhole mask is determined by the target spatial frequency region of a measured wavefront. The pitch ofthe grating is determined by the distance between the secondary pinhole and the window on the mask. In this case, the distance between the pinhole and the window on the mask is the same as the window size. Consequently, the target spatial frequency region determines the number of fringes and the pitch of the grating. To compare the results using visible and EUV radiation in the same spatial frequency region, we set the pitch of the grating for visible radiation the same as for EUV radiation. 48 Proc. SPIE Vol. 4146

3 Schwarzschild Optics NewSUBARU Monoch rometer Pinhole Grating Secondary Primary Pinhole Mirror Mirror Mask CCD Undulator BL9 Computer for Mechanical control Computer for Fringe analysis Fig. 2 Schematic diagram ofthe ITS. Table I. Optical components for visible and EUV radiation. Visth1eliht EUV Source He-Ne laser (632.8nm) Undulator (13.lnm) Initial Pinhole 3Oprn O.Sjim SUS sheet SUS foil Grating 1 5im pitch (66.7 lines/mm) 1 5.tm pitch (66.7 lines/mm) Crpauemongss FreestandingTapattem Schwarzschild Al coated Mo/Si multilayered Pinhole Mask 3.tm pinhole 65nm pinhole I 26.6j.tm Window 2.7jtm Window - c paftem :X Tapattem: Two sets of the Schwarzschild optics are specially manufactured as the test optics for visible and EUV radiation. These mirrors for visible radiation are coated with aluminum. The mirrors for EUV radiation are coated with and Mo/Si multilayers with peak reflectivity at 13.1 nm. The demagnification of the optics is 10:1 and it has a full aperture NA=O.1, which is obstructed by the primary mirror within the 34% of NA in diameter. The system is designed to be used on-axis using the full numerical aperture. The Schwarzschild optic and the pinhole mask mount are constructed as a single unit with invar bars to ensure temperature stability. The secondary mirror (the larger mirror) of the optics is mounted in a fixed mount while the primary mirror has adjusting capability to allow for aligning of the Schwarzschild system. The pinhole mask mount has translation capability along both transverse axes to allow for fine alignment of the mask pattern to the focal spots, and to allow for the changing of the mask patterns without the need to break the vacuum. The ITS has been constructed at the end of the undulator beam line at the NewSUBARU. Since the undulator can generate bright, quasi-monochromatic EUV radiation, it is suitable for the interferometry. Furthermore, the EUV radiation coming from the undulator is refined by passing through a monochrometer, which was designed to have resolution of more than Proc. SPIE Vol

4 1000. The EUV radiation focuses on the initial pinhole with the beam size ofabout 50 tim. 3. VISIBLE LIGHT TESTING Before performing the "at-wavelength" testing, we roughly evaluated the Schwarzschild optic using visible radiation. All of the optical components for visible light testing were then attached. In order to evaluate in the same region of the spatial frequency for both visible and EUV radiation, the pitch of the grating for visible light testing is the same as for EUV radiation. Since the visible beam largely diffracted by the grating compared to the BUy, we tilt the zero order beam against the optical axis to enter both 0th and 1st order beams into the pupil of the Schwarzschild optic as shown in Fig. 3. It is noted that the virtual object point of the 1st order beam, testing the optic, is on the optical axis. In this case, the pinhole acts as an ellipsoidal one because of the inclination of the pinhole against the incident beam, resulting in inducing a small astingmatism. Grating I 1st Order Beam Window Initial Pinhole 0th Order Beam Pinhole Fig. 3 Schematic diagram of the configuration of the visible testing. Figure 4 shows a typical image of interference fringes obtained using visible radiations. Phase-shifted fringe images were obtained by moving the grating. These images were analyzed by a computer software (Diffraction International Ltd.). The optical path difference (OPD) map of the wave front of the Schwarzschild optic is derived from the analysis. Fig. 4 Typical fringe image taken by the ITS using visible radiation. 50 Proc. SPIE Vol. 4146

Next, we aligned the Schwarzschild optic based on the measured wavefront error. Before alignment, the Schwarzschild optic has large coma and focus aberrations.

5 Next, we aligned the Schwarzschild optic based on the measured wavefront error. Before alignment, the Schwarzschild optic has large coma and focus aberrations. These aberrations are produced by the decenter of primary mirror and defocus of the optics. The amounts of such aberrations generated by displacements of the primary mirror and camera focus were calculated (Table 2). This simple two-mirror camera does not have complicated relation between degrees of freedom of each mirrors and the aberrations induced by the misalignments. Therefore, the adjusting axes and amounts can be easily derived from this table. Table 2. Aberrations produced by the displacement of the optics. These values are Zernike coeffient magnitudes generated by moving each axes with 0.01mm. Zernikes X-decenter Y-decenter Primary Mirror move Camera focus zi Z Z Z z Z Z Z Z Unit: wave632.8nm Table 3 shows the result of the alignment using this procedure. Focus and coma Y were successfully removed. Coma X is remained because the decenter in x-axis of the primary mirror cannot be quantitatively controlled. Figure 5 shows the derived optical path difference (OPD) map after the alignment. The measured wavefront error is listed in Table 4. The tilt corresponds to 40 fringes. The obtained wavefront has large astingmatism. Astingmatism may be caused by the elliptic shape ofthe initial pinhole. Table 3. Result ofthe alignment. Aberration Before align. Axes Calc. Actual After align. Zernike #5 (Focus) wave Camera Z 3038 urn 2823 urn 0.00 wave Zemike #8 (Coma X) wave PM X 469 urn --- urn wave Zernike #9 (Coma Y) wave PM Y 1 50 urn 204 urn 0.00 wave After the alignment, we evaluated repeatability of the measurements that were measured with the sarne optical configuration. Table 4 also shows the dispersion ofthe rneasured aberrations. These figures indicate that our interferornetry using visible radiation is insufficient to align the EUVL optical system. Fig. 5 OPD map of the wave-front of the Schwarzschild optic after the alignment. Coma aberration is seen in this figure. The wave like firnire seen in this image due to the damaue in the mirror.. Proc. SPIE Vol

6 Table 4. Results of the measurements of the aberration and the dispersions. Average Standard deviation PV(wave) RMS (wave) Seidel Tilt (wave) Focus (wave) Astigmatism (wave) Coma(wave) Spherical (wave) Wave=632.8nm 4. PRELIMINARY EUV EXPERIMENT In order to learn the EUV system, we investigated the spectral characteristics of the Schwarzschild optic and ensure the alignment for the small pinhole using EUV radiation. 4.1.Alignment for the EUV system The multilayered mirrors of Schwarzschild optic were replaced to the mirrors for visible radiation. The mirrors were aligned by the same manner as for visible testing. The wavefront error of nm RMS was achieved. After the alignment, the pinhole, the grating and the pinhole mask for visible radiation were replaced with those for EUV. To evaluate the spectral characteristics of the Schwarzschild optic, we imaged the incident beam to the ITS using the Schwarzschild optic without the initial pinhole, the grating and the pinhole mask. The intensity of the incident beam changed with the wavelength. We assured that the maximum reflecting wavelength was 1 3.mmwhich was designed Alignment for the pinhole The alignment of the incident EUV beam to the pinhole was a difficult problem. First, we tried the alignment of the pinhole using relatively large pinhole of 30 j.tm. The beam was aligned using He-Ne laser. We then aligned the EUV beam to go to the same path of He-Ne laser. It was neccesary to slightly adjust the EUV beam to concentrate on the pinhole. The EUV beam passing through the pinhole was successftilly imaged by the CCD. Furthermore, we verified the alignment procedure using 0.5 pm pinhole, which will be used for evaluating in the full aperture of the Schwarzschild optic. To use the same procedure, the EUV beam aligned to concentrate on the pinhole. The EUV radiation passing through the 0.Sjim pinhole was detected by a MCP as a monitor. 5. CONCLUSIONS The ASET ITS has been constructed at the end of the undulator beam line at the NewSUBARU facility. We have started the experiments and. visible testing was performed before EUV testing. The wavefront errors of the Schwarzschild optic were measured by analyzing the interferograms which were obtained by the same PS/PDI system as for EUV. The precision of the PS/PDI measurement using visible radiation was evaluated. Preliminary experiment using EUV radiation emitted from the undulator was performed. We will continue the experiments and evaluate the effectiveness of the at-wavelength testing. ACKNOWLEDGEMENTS The authors would like to thank K. A. Goldberg, P. Naulleau, J. Boker and D. Attwood of Center for X-ray Optics, Lawrence Berkeley National Laboratory for valuable discussion. This research was supported by NEDO as a Ministry of 52 Proc. SPIE Vol. 4146

7 International Trade and Industry project. REFERENCES 1. G. E. Sommargren, "Phase shifting diffraction interferometry for measuring extreme ultraviolet optics," in OSA Trends in Optics and Photonics, Vol. 4. Extreme Ultraviolet Lithography, G. D. Kubiak and D. R. Kania, Eds., Optical Society ofamerica, Washington, D. C., , K. A. Goldberg, R. Beguiristain, J. Bokor, H. Medecki, D. T. Attwood, K. Jackson, E. Tejnil, G. E. Sommargren, "Progress toward A /20 extreme ultraviolet interferometry," J. Vac. Sci. Technol. B13(6), , H. Medecki, E. Tejnil, K. A. Goldberg, J. Bokor, "Phase-shifting point diffraction interferometer," Opt. Left. 21 (19), , E. Tejnil, K. A. Goldberg, S. Lee, H. Medecki, P. J. Batson, P. E. Denham, A. A. MacDowel!, J. Bokor, D. Attwood, "At-wavelength interferometry for extreme ultraviolet lithography," J. Vac. Sci. Technol., B 15(6), , P. P. Naulleau, K. A. Goldberg, S. H. Lee, C. Chang, D. Attwood, J. Bokor, "Extreme-ultraviolet phase-shifting point-diffraction interferometer: a wave-front metrology tool with subangstrom reference-wave accuracy," Appi. Opt., 38 (35), , B. Jacobsen, Y. Gomei, H. Kinoshita, T. Watanabe, S. Kakunai, "Planned EUVL at-wavelength interferometer at the New Subaru facility," Proceedings on1999 International Workshop on X-ray and Extreme Ultraviolet Lithography, 2-2-2, Yokohama, Proc. SPIE Vol

At-wavelength characterization of the EUV Engineering Test Stand Set-2 optic

P6-4 At-wavelength characterization of the EUV Engineering Test Stand Set-2 optic Patrick Naulleau 1, Kenneth A. Goldberg 1, Erik H. Anderson 1, Phillip Batson 1, Paul E. Denham 1, Keith H. Jackson 1,