Soft x-ray microscopy with coherent x rays...

Imaging

Soft x-ray microscopy with coherent x rays (invited) J. Kit-z, H. Ade, C. Jacobsen, C.-H. Ko, S. Lindaas, I. McNulty, D. Sayre, S. Williams, and X. Zhang Physics Department, State University of New York at Stony Brook, Stony Brook, New York II 794

M. Howells ALS, Lawrence Berkeley Laboratory, Berkeley, California 94720

(Presented on 16 July 1991)

The Soft X-ray Undulator beamline at the NSLS supports a soft x-ray imaging program including scanning transmission microscopy, scanning photoemission microscopy, Gabor and Fourier transform holography, and large angle diffraction. Zone plates from an LBL Center for X-ray Optics/IBM collaboration are used as optical elements. The current instrumentation of the beamline and the experimental stations is discussed, along with plans for the future at the NSLS and prospects for imaging programs at third generation sources.

1. INTRODUCTION II. BEAMLINE

Much of the current effort in developing state of the art source of synchrotron radiation is aimed at high brightness undulator sources. X-ray microscopy is a major beneficiary of this trend, and we are fortunate to have as our source the Xl Soft X-ray Undulator at the National Synchrotron Light Source. The undulator and the associated beamline has been described in detail.’ In this report, we provide an overview of the experimental program at the beamline, and future plans for improvements at Xl and for the next generation instruments, such as at the Advanced Light Source in Berkeley.

While bending magnets and wigglers provide a fan of radiation that is easy to subdivide into smaller angular segments to serve several experiments, undulators produce a narrow, highly collimated beam, the useful central cone of which is more difficult to share. At the X 1 undulator, we have nevertheless made an effort to maximize utilization of the beam.

The work at the Xl A beamline depends on the large coherent flux in the beam. This makes it possible to use zone plates to form microprobes and to record holograms. The zone plates we have been using are made under a collaborative agreement between the LBL Center for X-ray Optics and the Nanofabrication Group at the IBM Re- search Center.2

Because the undulator is located in a straight section of the storage ring where the horizontal divergence of the electron beam is relatively high, the radiation divergence is dominated by that of the electron beam, with a horizontal divergence of 0.25 mrad (a,,,,,). As a consequence, the spectral linewidth of the undulator is broadened from the ideal (zero emittance) case. In practical terms, the foot- print of the central cone of the beam on the first mirror, located 13 m from the source, is about 15 cm wide at

Coherent illumination is not universally advantageous: Imaging microscopes operating in absorption contrast mode, such as the Gijttingen imaging microscope at BESSY,3 benefit from spatially incoherent illumination of the specimen. A source that concentrates all the radiation into a very small phase space volume is poorly matched to such applications. To redistribute the phase space, x-ray equivalents of rotating ground glass screens are necessary. In fact, such a device has already been used in the projec- tion lithography experiments at the U13 undulator by the AT&T group.4

1:; .----.-- Calculated (12% coupling) k

The development and application of x-ray microscopes was the subject of an international symposium a year ago, and the recently published proceedings’ is a rich source of information in the field. More recent reports on photoelectron microscopy,6 other work on soft x-ray microscopy,7 hard x-ray microscopy,8 and microtomography’ are available in this volume.

h' a 6.0

74 5: 4.5

2 -& 3.0 2 fg 1.5 2 EI

0.0 1.5 2 2.5 3 3.5 4 4.5 5

Wavelength (nm)

FIG. 1. The NSLS soft x-ray undulator spectral intensity for magnetic deflection parameter K = 1.5. The lines are calculated for 10% and 12% coupling between horizontal and vertical emittance in the storage ring, while the boxes are for measured spectral intensities. The structure near the carbon edge (4.36 nm) is an artifact (from Ref. I).

557 Rev. Sci. Instrum. 63 (l), January 1992 QQ34-6746/92/010557-Q7~Q2.00 0 1992 American Institute of Physics 557 Downloaded 23 Aug 2006 to 129.49.56.64. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp

Woiw Cooled Coppa ond Lend tlockrlop

vhn(ilh Branch ) l foaminp Mirror*

(Walrr Coolodl

I 1 1 I 1 0 I I

i4 2’2 i0 if3 is 14

Meters to Center of Undulator

12 10

4O-mrad grazing angle. Furthermore, as illustrated in Fig. 1, the spectral peaks are widened toward longer wave- lengths, and even harmonics as well as odd ones peak on axis. As a result, it is possible to share even this narrow fan by inserting the first mirror only part-way into the beam, with another beamline (XlB) intercepting the remaining portion.” Because we are only using the coherent portion of the beam, and because there are a large number of horizontal modes present, this arrangement (to be implemented in the near future) involves no compromise for XlA. Both beamlines must agree, however, on the undulator gap which determines the resulting spectrum.

The XlA beamline (Fig. 2) is further subdivided, with the long wavelength branch using the undulator fundamental, and the short wavelength branch using the second harmonic. These two branches are separated in angle at the output of the spherical grating bichromator, and further separated past their respective exit slits with small mirrors. The scanning transmission microscope is the dominant user of the long wavelength branch, while scanning photoemission microscopy, Gabor holography, and soft x-ray diffraction are major users of the short wavelength branch. Again, the experimenters at the two branches must agree on the spectral characteristics, and in this case on the ad- ditional common element, the bichromator entrance slit, which in part determines the bandwidth/flux trade-off that is available. In practice, there is simultaneous operation about 75% of the time.

This bandwidth/flux trade-off, as well as the size of the phase space element that we admit, has a profound effect on count rates and throughput. An aspect of this can be seen from Fig. 3. As we relax the strictest criteria of coherence and monochromaticity, the resolution worsens, while the flux increases, as one would expect. We usually operate with slit settings which provide a large gain in flux for a modest worsening of the resolution.

558 Rev. Sci. Instrum., Vol. 63, No. 1, January 1992

FIG. 2. Schematic diagram of the XIA beamline showing the elements of the short and long wavelength branches.

In many of our experiments, zone plates form microprobes by demagnifying a small source. In the present beamline there is no single small source, however, in the horizontal plane, the spherical grating focuses the beam onto the exit slits (typically 3 m from the zone plate), while in the vertical plane the source is at the undulator (23 m away) in the case where there is no focusing mirror downstream of the exit slit. The separation of the vertical and horizontal sources causes astigmatism in the beam, which is insignificant without vertical refocusing, but can become severe with it.” Until now, we ran mostly without refocusing or, in one case where the effect is severe, cor- rected for it using a specially fabricated elliptical zone plate.12 For the longer term, the first mirror in the beamline will be reconfigured so that it provides vertical focusing with greatly reduced astigmatism. This will push the light gathering power of the beamline closer to the maxi- mum possible, and should lead to a fivefold gain in flux for most experiments.

We use 250~pm square silicon nitride windows to separate the ultrahigh vacuum section of the beamline from rough vacuum on the short wavelength branch when needed, and from atmospheric pressure on’the long wavelength branch. These windows are quite reliable and we observe onIy mild contamination on them after a year of use.

Spatial and temporal stability of the microprobe is maintained partly by a feedback systemI based on position monitors located 9 m from the undulator, and partly by overfilling the slits and apertures (including the zone plates) +

Imaging beamlines at the third generation sources will operate with sources between one and two orders of magnitude higher in brightness and coherent power than Xl. These sources will be even more collimated, and the spectral features will not be as angle-averaged as they are at Xl.

Synchrotron radiation 558

Downloaded 23 Aug 2006 to 129.49.56.64. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp

Raction of Spatially Coherent Flux in the SXU fmtdamental

half-Airy-disk 4

i! lo-’ crllerlon

_____.---- _______--------

3 10” _e--- ____.----- B

doe. 2 & * 3 I iv 4 I * 6 1 * 0 I * 7 1 ’ 3 0 ’ s 1 7 10

Wavelength (nm)

FIG. 3. Fraction of the flux in the Xl undulator fundamental accepted for imaging. While the single Gaussian mode criterion (solid line) leads to effectively ideal coherence, the half-Airy-disk criterion (dashed line) results in a 20% increase in the focal spot size.

As a result, the methods we are currently using to share the beam cannot be implemented. However, the increased brightness makes time sharing more attractive, for while the time to collect an image is expected to drop to the second or subsecond range, the time required for a user to make an intelligent decision about the next image to be acquired will presumably not change! We envision the use of a system of computer controlled small mirrors which would deflect the beam to the next experimenter in line, while the one whose image has just been completed evalu- ates the results and decides on the next task. In addition to redirecting the beam, switching to the next user may also involve a change in the setting of the monochromator, and eventually of the undulator gap as well.

Stability of the beam will again be an important con- sideration at the new sources. With a much larger fraction of the beam having good enough coherence properties for the type of experiments we are interested in, apertures cannot be overfilled by a large factor without losing useful photons.

Ill. SCANNING TRANSMISSION X-RAY MICROSCOPE (STXM)

For high-resolution imaging of biological material, two considerations are of importance: The specimen should be exposed to a minimum amount of radiation dose for the given information-gathering task, and the specimen should remain as close to its natural state as possible (typically wet and at atmospheric pressure). The scanning transmission microscope is well suited to these specifications. The zone plate defines the resolution, and forms the microprobe in air (or in a helium enclosure to reduce losses). X-rays transmitted by the specimen are recorded with a high-efficiency, high-rate proportional counter. In addition to STXM at the NSLS,14 similar instruments are operating at Daresbury” and at BESSY.‘6


FIG. 4. STXM image of a gold test pattern after deconvolution of the point spread function. Test pattern was fabricated by Erik Anderson (from Ref. 19).

Recently, the STXM has been fitted with a high-resolution, high-efficiency nickel phase zone plate with 45-nm outer zones,17 and a new flexure stage with Queensgate positioners.18 The system has shown a performance that is very close to theoretical expectations. Based on the detailed study of the modulation transfer function (MTF), we were able to perform a deconvolution of the image.tg The result in the case of a resolution test pattern is shown in Fig. 4, where detail down to 36-nm lines and spaces is clearly reproduced. Typical operating energy is 350 eV.

Biological applications of STXM involve a growing number of users, and include the study of secretion granules,20 chromosomes2* (Fig. 5), cell cultures,22 and the mapping of calcium in sections of tendon and cartilage23 (Fig. 6).

In routine operations a 400 x 400 pixel image takes 3-6 min to record today, and 30-50 images are recorded on a typical day. With expected improvements to the XlA beamline, the data rate should increase by another order of magnitude within the next year, and with the increased brightness at the third-generation sources there is a need for a radical rethinking of STXM design. In order to make images at the rate allowed by the x-ray flux in the probe, the detector will have to handle rates in the gigahertz range, and the scanning stage will need to achieve 106-pixel/s scan rates. The data must be stored, displayed, and analyzed at rates that match these capabilities. To accomplish these tasks, we are exploring the use of inte- grating detectors, and continuous “on the fly” scanning of the zone plate in one dimension, while advancing the specimen only along the slower axis.

IV. THE SCANNING PHOTOEMISSION MICROSCOPE (SPEM)

Much of what we know about surface chemical com- position and electronic structure has been determined us-

Synchrotron radiation 559 Downloaded 23 Aug 2006 to 129.49.56.64. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp

3 2500.0

g 2000.0 u

1500.0

1000.0

500.0

FIG. 5. STXM image of a fixed hydrated chromosome of the bean Vicia faba.

ing photoemission spectroscopy. Until recently, the spatial resolution of this technique was limited, and for studies on heterogeneous surfaces Auger electron microscopes were generally employed. Due to charging and damage considerations,24 these latter instruments can only be used with rugged, conductive samples.

More recently, several types of synchrotron radiation based photoemission microscopes have been built, some of which form the image using electron optics,25p26 while others,27Y28 including the X1 SPEM,29 use a scanned microprobe to form the image (and if desired, to obtain the photoelectron spectrum) pixel by pixel.

In the current Xl SPEM a specially designed zone plate corrects for astigmatism,12 and forms a 150-nm mi-

FIG. 6. STXM images of a fixed, unstained, 150~nm-thick section from a patient with tendonitis. Top: large scale survey scan, covering an area nearly 1 mm on the side (scale bar: 0.1 mm). Bottom: the indicated portion of the same section on a finer scale (scale bar: 5 pm). The images at the left were collected at the calcium-absorptive wavelength of 3.560 nm, and the images at the right were at the calcium-transmissive wavelength of 3.595 nm. Calcium distribution can be determined from a pixel- by-pixel comparison of absorptivity. Courtesy of Christopher Buckley, King’s College, London, and Yusuf Ah, Royal Orthopedic Hospital, Stanmore, England.


0.0 ’ I I I 8 t I 300 350 400 4% 500 550 600

Electron energy (eV)

FIG. 7. Photoelectron spectrum from a 0.2~pm area of an uncleaned aluminum oxide surface. Spectral peaks corresponding to oxygen and carbon, in addit ion to aluminum are prominent.

croprobe on the surface to be examined. Photoelectrons are analyzed using a cylindrical mirror analyzer. Total yield is measured at the same time. Typical photon energy is 680 eV. A typical spectrum is shown in Fig. 7, while Fig. 8 is the image of a microelectronic device taken with the SPEM.

At present, it takes close to an hour to collect an image, and a reasonably high statistics spectrum from a single spot requires about 45 min. Our plans are to increase the data rate by about an order of magnitude partly by the beamline upgrade described earlier (also incIuding a blazed grating to provide higher efficiency for the short wavelength branch), and partly by the use of a more effi- cient and higher resolution photoelectron spectrometer. The expected electron energy resolution after these modi- fications is below 1 eV.

Scanning surface microprobes have the potential of recording and analyzing not only photo- and Auger electrons, but also the fluorescence signal, and the atomic and molecular species that leave the surface by photon stimu- lated desorption. Although there are plans to develop the instrumentation for PSD at the Xl SPEM, such devices will come into their full potential at the third generation machines.

V. HOLOGRAPHY AND DIFFRACTION

While scanning microscopes examine the specimen pixel by pixel, holographic imaging methods collect the full field all at once. Both Gabor and Fourier transform holog raphy have been implemented at the XlA beamline.

In the Gabor geometry a plane reference wave inter- feres with the waves scattered by the object. The object field should be sparse to transmit much of the reference beam unaltered, but can be as large as the beam itself, The transverse resolution in the reconstructed image is limited to the resolution of the detector. To reach into the submicron resolution range, the x-ray resist PMMA was intro- duced as the recording material.30 W ith electron micro-


(b)

(a) FIG. 8. SPEM images of a microelectronic device. The image on the left is recorded with total electron yield as the signal, while the one on the right is with oxygen KVV Auger electrons. The image field is 35 pm wide (from Ref. 29).

scope readout of the resist, and subsequent numerical reconstruction of the hologram, 60-nm transverse resolution has been demonstrated.31 It takes about 3 min to record a hologram in vacuum, or about 20 min if the specimen is wet and in an atmospheric environment. Recent work has been aimed at reducing the twin-image back- ground in the reconstruction using phase retrieval techniques,32 and replacing the electron microscope by atomic force microscopy for the readout of the resist. This approach promises to increase sensitivity and dynamic range, and simplify the interpretation of the results. An- other approach being taken is that of Joyeux ef ~1.~~ who are building a UV optical reconstruction system for use with holograms to be recorded at LURE, for which they will soon have a dedicated beamline on SuperACO.

In the Fourier transform geometry a spherical reference beam is made to interfere with the object wave. In this case, the resolution is determined by the precision of the reference beam (determined by the zone plate focus), and the resolution of the recording medium determines the size of the object field, rather than the image resolution. In our recent work,34 a Fresnel zone plate was used as the beam splitter, to illuminate the specimen, and also to form the spherical reference beam. Holograms of test patterns were recorded with a cooled CCD camera, and reconstructed numerically (Fig. 9). Over the limited object field diameter of 14 pm, structure with 50-nm lines and spaces is revealed in the reconstruction. At the Photon Factory, Aoki and Kikuta recorded Fourier transform holograms using pho- tographic film.35

Both in Gabor and Fourier transform holography, our reconstructions are essentially two dimensional, in that our resolution in depth is comparable to the thickness of typical specimens. To reach full three-dimensional reconstruction, the simplest road seems to be to make use of multiple views.36*37

To reach the highest resolution in all three dimensions (eventually half the wavelength), the scattered wave must be recorded over a large solid angle, and for many illumination directions. While it is difficult to generate a reference wave for large-angle recording, reconstruction from scattered wave patterns alone is the domain of crystallog


(b)

FIG. 9. Test specimen (a), Fourier transform hologram (b), and the reconstructed image (c). The lines and spaces of the test specimen de- crease from 125 nm at the periphery to 50 nm at the inner extreme. Specimen fabricated by Erik Anderson.

raphy. To extend the techniques of crystallography to the study on noncrystalline specimens requires that these specimens be particularly radiation resistant, as the large angle scattering pattern is on average quite weak. It is, nevertheless, this approach that promises the highest resolution in


soft x-ray imaging, and efforts to detect high-angle scattering are making significant progress.38

VI. SUMMARY

The brightness of the soft x-ray undulator and the ex- cellent zone plates fabricated for us by Erik Anderson have made it possible to develop several coherence-dependent forms of soft x-ray imaging at the NSLS. Each of these will continue to be improved and developed both at the NSLS and at the new sources that will come on line within the next few years. Among the lessons we have learned is that it is necessary to approach the design of imaging systems as an integrated whole, from the undulator, through the beamline, the microscope, its optics, the detector, and the data handling system.

As the brightness of harder x-ray sources improves, there will be new opportunities to extend submicron resolution imaging to the harder x-ray region. There are excit- ing possibilities for phase contrast microscopy of relatively thick biological specimens.39,40 In addition, elemental imaging by scanning fluorescence microscopy4t will be ex- tended to the submicron resolution regime using zone plates that are somewhat thicker than the ones we have been using. Several groups are working toward this goa1.42-44

ACKNOWLEDGMENTS

The imaging program at the NSLS is the product of the efforts of a large number of people. The NSLS staff and management (past and present) provided the undulator source that is ideal for our needs. The beamline and the STXM instrument itself was built with critical help from Harvey Rarback, Chris Buckley, Mark Rivers, and Dem- ing Shu. The optics fabricated by Erik Anderson with major support from Dieter Kern and David Attwood have been indispensible. Much of the biological work on STXM has been performed by Stephen Rothman, Kaarin Goncz, Bill Loo, Jerry Pine, John Gilbert, and Chris Buckley. The SPEM was built in collaboration with Erik Johnson, Steven Hulbert, Erik Anderson, and Dieter Kern. We are grateful to all these individuals. We thank the National Science Foundation for support under Grant No. DIR- 9005893, and the Department of Energy for support under Grant No. DE-FG02-89ER60858. Much of this work was carried out at the NSLS which is supported by the Depart- ment of Energy, Office of Basic Energy Sciences.

‘H. Rarback, C. Buckley, H. Ade, F. Camiiio, R. DiGennaro, S. Heii- man, M. Howeiis, N. Iskander, C. Jacobsen, J. Kirz, S. Krinsky, S. Lindaas, I. McNuity, M. Oversiuizen, S. Rothman, D. Sayre, M. Shar- noff, and D. Shu, J. X-ray Sci. Technoi. 2, 274 (1990).

’ E. Anderson, “Fabrication technology and applications of zone plates,” in X-ray/EUV Optics for Astronomy and Micmscopy, edited by R. B. Hoover (SPIE Proc., 1990), Vol. 1160, p. 2.

‘D. Rudolph, B. Niemann, G. Schmahi, and 0. Christ, “ The Gottingen x-ray microscope and x-ray microscopy experiments at the BESSY Stor- age Ring,” in X-Ray Microscopy, edited by G. Schmahi and D. Rudolph (Springer, Berlin, 1984), p. 192.

4A. A. McDowell, J. E. Bjorkhoim, J. Bokor, L. Eichner, R. R. Free- man, T. E. Jeweii, W. M. Mansfield, J. Pastaian, L. H. Szeto, D. M. Tennant, W. K. Waskiewicz, D. L. White, D. L. Windt, 0. R. Wood II,


and W. T. Siifvast, J. Vat. Sci. Technoi. (to be published). ‘A. Michette, G. Morrison, and C. Buckley, eds., X-ray Microscopy III

(Springer, Berlin, 1991). 6B. P. Tonner, these proceedings; J. Voss, ibid.; W. Ng, J. P. Wallace, A.

K. Ray-Chaudhuri, C. Capasso, and F. Cerrina, ibid. ‘C. J. Buckley, R. E. Burge, G. F. Foster, S. Y. Aii, C. A. Scotchford, J.

Kim, and M. Rivers, these proceedings; G. R. Morrison and M. T. Browne, ibid.; G. F. Foster, P. M. Bennet, C. J. Buckley, and R. E. Burge, ibid.; Y. Kagoshima, S. Aoki, and M. Ando, ibid.

sY. Suzuki, these proceedings; W. Yun, ibid.; V. V. Ariatov, Yu. A. Basov, A. A. Snigirev, and V. A. Yunkin, ibid.; U. Bonse, C. Riekei, and A. A. Snigirev, ibid.

9K. L. D’Amico, these proceedings: I. P. Doibnya, N. G. Gavriiov, S. G. Kuryio, N. A. Mezentsev, and V. F. Pindyurin, ibid.; Y. Nagata, H. Yamaji, K. Hayashi, K. Kawashima, K. Hyodo, H. Kawata, and M. Ando ibid.; I. P. Doibnya, ibid.

“W. Eberhardt, K. J. Randall, J. Feidhaus, A. M. Bradshaw, R. F, Garrett, and M. L. Knotek, Phys. Ser. 41, 745 (1990).

“A stigmatic microprobe is formed if both the horizontal and vertical sources imaged by the zone plate are effectively at infinity. Quantita- tively the source distance is large enough if its image is within the depth of focus from the focal point. This distance is d&/4.88& or about 0,5 m for STXM, and 2.2 m for SPEM.

“H. Ade, C.-H. Ko, and E. Anderson (to be published), “R. Nawrocky, J. W. Bittner, Li Ma, H. M. Rarback, D. P. Siddons, and

L. H. Yu, Nuci. Instrum. Methods A26,96 ( 1988). 14H. Rarback, C. Buckley, K. Goncz, H. Ade, E. Anderson, D. Attwood,

P. Batson, S. Heiiman, C. Jacobsen, D. Kern, J. Kirz, S. Lindaas, I. McNuity, M. Oversiuizen, M. Rivers, S. Rothman, D. Shu, and E. Tang, Nuci. Instrum. Methods A 291, 54 ( 1990).

“D. Morris, C. J. Buckley, G. R. Morrison, A. G. M.ichette, P. A. F. Anastasi, M. T. Browne, R. E. Burge, P. S. Charaiambous, G. F. Fos- ter, J. R. Palmer, and P. J. Duke, Scanning 13, 7 (1991).

16B. Niemann, “The Gottingen scanning x-ray microscope,” in Soft X-ray Optics and Technology, edited by E. E. Koch and G. Schmahi (SPIE Proc., 1987), Vol. 733, p. 423.

“E. H. Anderson and D. Kern, in X-ray Microscopy III, edited by A. Michette, G. Morrison, and C. Buckley (Springer, Berlin, 1991).

“M. Browne . in X-ray Microscopy 121, edited by A. Michette, G. Morti- son, and C. Buckley (Springer, Berlin, 1991).

“C. Jacobsen, S. Williams, E. Anderson, M. T. Browne, C. J. Buckley, D, Kern, J. Kirz, M. Rivers, and X. Zhang, Opt. Corm-nun. (to be pub- iished) .

‘OS. S. Rothman, N. Iskander, D. Attwood, Y. Vladimirsky, K. Mc- Quaid, J. Grendeii, J. Kirz, H. Ade, I. McNuity, D. Kern, T. H, P. Chang, and H. Rarback, Biochim. Biophys. Acta 991,484 ( 1989); S. S. Rothman, K. Goncz, and B. Loo Jr., in X-Ray Microscopy ii& edited by A. Michette, G. Morrison, and C. Buckley (Springer, Berlin, 1991).

“S. Williams , C. Jacobsen, J. Kim, S. Lamm, and J. van? Hof, in X-ray Microscopy III, edited by A. Michette, G. Morrison, and C, Buckley (Springer, Berlin, 199 1) .

22J. R. Gilbert, J, Pine, J. Kim, C. Jacobsen, S. Williams, C. 3. Buckley, and H. Rarback, in X-ray Microscopy 121, edited by A. Michette, G. Morrison, and C. Buckley (Springer, Berlin, 1991).

“C. J. Buckley, G. F. Foster, R. E. Burge, S. Y. Ali, and C. A. Scotch- ford, in X-ray Microscopy III edited by A. Michette, G. Morrison, and C. Buckley (Springer, Berlin, 1991).

*“J. Cazaux, Appi. Surf. Sci. 20, 457 ( 1985). 25B. Tonner and G. R. Harp, Rev. Sci. Instrum. 59,853 (1988); see, also,

Ref. 6. 26P. Pianetta, I. Lindau, P. L. King, M. Keenlyside, 6. Knapp, and R.

Browning, Rev. Sci. Instrum. 60, 1686 (1989). 27A. Moewes, H. Dadra, C. Kunz, H. Sievers, I. Storjohan, and J. Voss,

in X-ray Microscopy III, edited by A. Michette, G. Morrison, and G. Buckley (Springer, Berlin, 1991); see, also, Ref. 6.

“F. Cerrina, G. Margaritondo, J. H. Underwood, M. Hettrick, M. A. Green, L. J. Briiison, A. Franciosi, H. Hochst, P. M. Deiuca, Jr., and N. Gould, Nuci. Instrum. Methods A 266, 303 (1988).

29H. Ade J Kirz, S. Huibert, E. Johnson, E. Anderson, and D. Kern, Appi. Whys. Lett. 56, 1841 (1990); H. Ade, 3. Kirz, S. Huibert, E. Johnson, E. Anderson, and D. Kern, J. Vat. Sci. Technoi. A 9, 1902 (1991).

“M. Howeiis, C. Jacobsen, J. Kim, R. Feder, K. McQuaid, and S. Roth- man, Science 238, 514 (1987).

Synchrotron radiation 562

Downloaded 23 Aug 2006 to 129.49.56.64. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp

“C. Jacobsen, M. Howells, J. Kirz, and S. Rothman, J. Opt. Sot. Am. A 7, 1847 (1990).

“C. Jacobsen and S. Lindaas, in X-ray Microscopy ZZZ, edited by A. Michette, G. Morrison, and C. Buckley (Springer, Berlin, 1991).

“D. Joyeux and F. Polack, in Short Wavelength Coherent Radiation: Generation and Applications, edited by R. W. Falcone and J. Kirz (Op- tical Society of America, Washington, DC, 1988), Vol. 2, p. 295.

“1. McNulty, J. Kirz, C. Jacobsen, M. R. Howells, and E. H. Anderson, in X-ray Microscopy ZIZ, edited by A. Michette, G. Morrison, and C. Buckley (Springer, Berlin, 1991).

‘% Aoki and S. Kikuta, in Short Wavelength Coherent Radiation: Gen- eration and Application, edited by D. T. Attwood and J. Bokor (Amer- ican Institute of Physics, New York), p. 49.

36M R Howells, C. Jacobsen, J. Kirz, K. McQuaid, and S. S. Rothman, . . in Modern Microscopies, edited by P. J. Duke and A. G. Michette (Plenum, New York, 1990), p. 119.

“5. Trebes, C. Annese, D. Birdsall, J. Brase, J. Gray, S. Lane, R. London, D. Matthews, D. Peters, D. Pinkel, G. Stone, D. Rapp, M. Rosen, U. Weir, and T. Yorkey, in X-ray Microscopy ZZZ, edited by A. Michette, G.


Morrison, and C. Buckley (Springer, Berlin, 1991). ‘* D. Savre, in Proceedings of the 1990 Erice Workshop on Direct Methods,

edited by H. Schenk (Plenum, New York, 1991) (in press). 39D. Rudolph, G. Schmahl, and B. Niemann, in Modern Microscopies,

edited by P. J. Duke and A. G. Michette (Plenum, New York, 1990). “C!. Jacobsen, in X-ray Microscopy IZZ, edited by A. Michette, G. Mor-

rison, and C. Buckley (Springer, Berlin, 1991). “For a recent review and list of references see M. Rivers, S. R. Sutton,

and K. W. Jones, Synch. Rad. News 4, 23 ( 1991). “D. Rudolph, B. Niemann, and G. Schmahl, in High Resolution of Soft

X-ray Optics, edited by E. Spiller (SPIE Proc., 1981), Vol. 316, p. 103; see two separate articles by R. Hilkenbach and J. Maser in X-ray Mi- croscopy III, edited by A. Michette, G. Morrison, and C. Buckley (Springer, Berlin, 199 1) .

43R. M. Bionta, E. Ables, 0. Clamp, 0. D. Edwards, P. C. Grabriele, K. Miller, L. L. Ott, K. M. Skulina, R. Tilley, and T. Viada, Opt. Eng. 29, 576 (1990); see, also, W. Yun, Ref. 8.

“K. Saitoh, K. Inagawa, K. Kohra, C. Hayashi, A. Iida, and N. Kato, Rev. Sci. Instrum. 60, 1519 (1989).


Soft x-ray microscopy with coherent x rays...

Documents

Transcript of Soft x-ray microscopy with coherent x rays...