Investigations on Laser-Generated Plasma Sources
T. Wilhein
Georg-August-Universität Göttingen, Forschungseinrichtung Röntgenphysik,
Geiststraße 11, D-37073 Göttingen, Germany
Abstract. In order to measure the spectral brilliance of laser plasma X-ray sources,
a spectrograph has been developed which allows simultaneous recording of the
wavelength depending source diameter and the source spectrum. The optical
system is a new single element X-ray optic [7], which produces a series of lateral
displaced enlarged images of the X-ray source at different wavelength on the
detector, a cooled slow scan CCD camera with a thinned, back illuminated CCD.
First brilliance measurements of different laser plasma sources in the wavelength
range λ=1.5...5nm have been performed, showing that laser plasma sources can
serve as bright laboratory X-ray sources for applications in the soft X-ray region.
1 Introduction
During the last years, the fields of research using soft X-rays have experienced a remarkable growth. Applications like, e.g., X-ray microscopy, X-ray photoelectron
spectroscopy or X-ray lithography have prooved to be powerful scientific tools [1].
Experiments utilizing these technics are mostly performed at synchrotron radiation
facilities, which provide quasi-continuous highly collimated X-radiation over a wide
spectral range [2]. To extend the applications of these methods, intense laboratory Xray sources have to be developed. Laser generated plasmas promise to become bright
X-ray sources, supplying pulsed radiation at wavelength down to around 1nm using
commercially available laser systems and even much shorter, if terrawatt laser pulses
are employed [3, 4]. In order to decide wether a laser plasma is well suited to serve
as an X-ray source for a particular application, its emission properties have to be
quantified. The main figure of merit used to characterize X-ray sources is the
spectral brilliance, which is defined as the number of photons emitted per unit time,
source area, solid angle and relative bandwidth [2]. However, for pulsed sources it is
more suitable to define the pulse brilliance bpulse as
b pulse =
number of photons
pulse, source area, solid angle, rel. bandwidth
Units for the pulse brilliance are photons/(pulse×µm2×sr×0.1% BW). The exposure
time required to carry out a certain experiment can be estimated by integrating the
pulse brilliance over the corresponding experimental properties, e.g. the angle of
acceptance, and taking into account the repition rate of the source. To be able to
measure the pulse brilliance of laser generated plasma sources in the soft X-ray
range, a spectrograph has been developed which allows simultaneous measurement
of the spectral distribution and the source size by means of an off-axis reflection zone
plate (ORZ) [7] in combination with CCD image detection.
V - 26
T. Wilhein
2 Off-Axis Reflection Zone Plates for Spectral Imaging
A universal idea of point-to-point image formation is shown in Fig. 1 (cp.[5]). A
spherical wave emitted by a point source S will converge into image point I if the
surface of equal phase describing the wave deflection is given by an ellipsoid with
focal points at S and I. Because phase differences being integer multiples of 2⋅π will
result in constructive interference, a set of ellipsoidal shells corresponding to optical
path differences 2⋅π⋅λ for every two adjacent shells characterize the imaging process.
Any single element non-refractive optic can be described using this scheme. For thin,
plane diffractive optical elements the pattern providing the appropriate wave deflection is given by the figure which results from cutting the system of ellipsoidal shells
with the substrate plane. The general shape of this figures are non-concentric
ellipses, which may in special cases lead to, e.g., concentric circles, describing onaxis
zone
substrate
substrate plane
s
ϕ
wave front
S
d
optical axis
I
ellipsoidal shells
cn
Fig. 1. Origin of the off-axis reflection zone plate pattern (see text)
plates. If the straight line between S and I cuts the substrate plane, the optic is operating in transmission, whereas the opposite case leads to a reflective element. For the
off-axis reflection zone plate, the pattern is given (see Fig. 1 for descriptions) by

2
 p + w ⋅ 1 + m2 v 


=1
+
2
2
2
2
2
2
an − s + w v 1+ m
v ⋅ an − s + w2 v
q2
(
) (
)
p, q = coords in substr. plane, m = tan ( ϕ ), c n = n ⋅ λ 2 + const, n = 0,1, 2... = zone number
a 2
1
d
⋅
= normalized excentricity
a n 2 = c n 2 − d 2 , v: = m 2 + n 2 , w: = m ⋅ s, ε =
c
cn
1 + m2
n
Investigations on Laser-Generated Plasma Sources
V - 27
Off-axis reflection zone plates (ORZs) [7] are manufactured by e-beam writing of
the desired part of this pattern into PMMA and subsequent microstructuring steps.
The ORZs used in the below described experiments consists of about 7000 line pairs,
written with a LION LV1 e-beam writer (LEICA, Jena, Germany) and structured in
≈20 nm Ge on a glass substrate, in some cases coated with a thin Ni layer for enhanced reflectivity.
Table 1. Design parameters of the first ORZ
size: 2 stripes, each 1×8mm2, 1mm spacing,
≈7000 line pairs, grating constant along optic
axis g =1131-1138nm, +1. diffr. order
λ0
wavelength
2.4nm
a0
object distance
1500mm
b0
3000mm
image distance
f0
Focal length
1000mm
M0
Magnification
2
α0
Input angle (gr.inc.)
1.5°
Output angle (gr.inc.) β0
4.0°
dβ/dλ 13mrad/nm
Angular disp. at λ0
Linear disp. at λ0
dx/dλ
38mm/nm
Ω0
Collected solid angle
2×10-7sr
δy
Spatial diffr. limit
3µm
λ/∆λ
Spectral diffr. limit
> 1000
(dep. on source size)
Fig. 2. Scheme of the first ORZ
spectrum on CCD
focal spot
at λ0
CCD
reflecting
plane
X-ray
source
optical
axis
β
substrate
α
zone plate figure
Fig. 3. Image formation with an off-axis reflection zone plate [7]
V - 28
T. Wilhein
Detailed reports on the e-beam lithography and microfabrication processes are
presented in ref. [6–8]. A scheme of the first built ORZ is given in Fig. 2, its
paramters listed beside it. In a first order approximation, the optic can be treated as a
diffraction grating with slightly variing grating constant g on the zone plate axis and
curved lines. Figure 3 explains the appearance of spectrally displacedimages: for its
design paramters, that is geometry and wavelength, the ORZ forms a diffaction
limited image of a point source at the location of the image point. Because the focal
length of zone plates vary as f∝λ-1, longer / shorter wavelength will have their focus
in front of / behind the detector plane, leading to defocused and, due to the off-axis
character of the ORZ, laterally displaced images on the detector. Thus, the image
consists of a spectral and a spatial direction, the latter allowing to measure the source
size at the focused wavelength. The diffraction limits given in the table beside Fig. 2
for both spatial resolution and spectral resolving power takes into account the
effective aperture, the spectral resolving power in addition being influenced by the
average angular dispersion of the optic. The off-axis nature of the ORZ in combination with the grazing incidence properties give rise to remarkable aberrations
when used in geometries different from the design geometry. A ray-tracing program
has been developed to calculate the spectral and spatial behaviour of ORZs under
arbitrary conditions, e.g., extended or displaced sources, changes in the input angle
etc. For example, it has been found that the aberrations arising from setting the focus
of the ORZ to a wavelength λ different from the design wavelength λ0 can be
minimized by keeping input angle and source distant constant, and placing the
detector in image distance corresponding to the focal length fλ=f0⋅λ0/λ, where f0, λ0
are design parame-ters, and with respect to the change in the diffraction angle.
3 The ORZ-Based Imaging Spectrograph
The spectrograph consists of three vacuum chambers, connected with vacuum tubes
(Fig. 4). The first chamber contains the target system, the second the ORZ and a
metal filter (Al or Ti) to block visible, IR and UV light, the third chamber carries the
detector, a cooled clow scan CCD camera (Photometrics AT 200L), equipped with a
thinned, back illuminated CCD (TK1024AB). The camera provides cooling down to
–40°C, 16 bit/40 kHz readout and a readout noise of ≈10e-. The high dynamic range
allows to make precise measurements of line/background intensities in the spectral
images. In the present setup the achievable spatial resolution is limited by the pixel
size (24×24 µm2) of the CCD and approx. 2× magnification to ≈20µm and the
spectral resolution to ≈0.002 nm. To record a wider spectral range than that fitting to
the CCD size, the input angle α can be choosen by turning the ORZ, keeping the
total deflection angle γ constant. The spectra shown below have been obtained by
taking images at 3...6 different input angles. In order to adjust the optical axis of the
ORZ in a way that it hits the source, which is necessary to minimize aberrations, the
ORZ can be rotated precisely perpendicular to the substrate plane using a stepper
motor with 80000 steps/360°.
Investigations on Laser-Generated Plasma Sources
V - 29
laser
target
focusing
optic
ORZ
filter
X-rays
vacuum chambers
motor
thinned, back
illuminated CCD
Fig. 4. Scheme of the imaging spectrograph [15]
Measuring absolute brilliances means determination of absolute photon numbers.
To be able to do so, all components of the spectrograph have to be absolutely
calibrated. The filter transmittance and the diffraction efficiency of the ORZ have
been measured using the X-ray test chamber of the FE Röntgenphysik at BESSY for
wavelength λ=1...5nm. The quantum efficiency of the thinned, back illuminated
CCD was determined at the radiometry beamline of the Physikalisch Technische
Bundesanstalt (PTB) at BESSY [9] in the wavelength region λ=1.5...20nm. ORZ
diffraction efficiencies up to 6% have been obtained (at λ=2.2nm) [10], and the
quantum efficiency of the CCD was found to be ≥ 0.6.
4 First Experiments with the Imaging Spectrograph
Three experiments with the new imaging spectrograph are presented below, carried
out in collaboration with three different groups using different laser and target
systems. In each case the spectrograph was adapted to the respectively existing target
chambers. For applications like X-ray microscopy or X-ray photoelectron
spectroscopy strong emission in well separated lines is preferable, thus target
elements are choosen to have lines from H-like and He-like ions in the desired
spectral range. In the water window, λ≈2.3...4.3nm, which is of particular interest
for investigations on biological samples, nitrogen and carbon fullfill this condition
and therefore targets containing these elements were choosen for the experiments.
Due to the calibration procedure of the ORZ and the CCD the error in the given
photon numbers can be expected to be less then 50%.
V - 30
T. Wilhein
4.1 Experiment with a Droplet-Target Laser Plasma Source
at the Lund Institute of Technology
The first experiment with the ORZ spectrograph has been performed in collaboration
with H. Hertz, L. Rymell, M. Berglund at the Dep. of Physics, Lund Institute of
Technology, applying the droplet-target laser-plasma X-ray source [10]. This source
uses ≈10 µm droplets as target, resulting in a practically debris-free plasma source.
The frequency doubled Nd:YAG laser supplies 70 mJ laser pulses with 100–120 ps
pulse width at λ=532 nm and 10Hz repition rate, giving an intensity of ≈5⋅1014
W/cm2 in a 12 µm focal spot. A detailed description of the source can be found in
ref. [11, 12]. In this experiment, 30% NH3 dissolved in H2O has been used as target
liquid, and a 3 mJ UV prepulse has been employed for enhanced X-ray emission
[13].
Fig. 5. Spectral image of the Lund laser plasma source taken with the imaging spectrograph.
Focus set to 2.88 nm. Target: droplets, 30% NH3 dissolved in H2O, laser pulse energy 70 mJ,
pulse width ≈120 ps, intensity on target ≈5⋅1014W/cm2 [12], no prepulse
Figure 5 shows a spectral image of the Lund laser plasma source. The focus of the spectrograph was set to the nitrogen He-α line (N VI 1s2-1s2p) at λ=2.879 nm, because this
line is of particular interest for X-ray microscopy. Due to the target composition (only
oxygen and nitrogen contribute to the X-ray emission) the spectrum below 5 nm ends with
the focused line, thus the spectral image shows only defocused lines in the short
wavelength direction. The picture is composed of three different exposures with variing
input angle α. A part of the spectrum taken with the imaging spectrograph is represented
in Fig. 6. The expected nitrogen and oxygen lines can easily be identified. The pulse
brilliances under the given experimental conditions was found to be ≈9×107photons/
(pulse×µm2 ×sr×0.1%BW) in the N Ly-α line (λ=2.478nm), and ≈5×107photons/
(pulse×µm2 ×sr×0.1%BW) in the N He-α line (λ=2.879nm), the relative linewidth were
measured to λ/∆λ≥350 and 450 (FWHM), respectively.
1.0
V - 31
N
Ly-α
O
He-α
0.8
0.6
N
He-α
0.4
0.2
N
He-β
N
Ly-β
N
Ly-γ
satellites
N
He-γ
8
10 Photons / (pulse × µm2 × sr × 0.1% BW)
Investigations on Laser-Generated Plasma Sources
0.0
2.0
2.2
2.4
2.6
2.8
3.0
λ [nm]
Fig. 6. Part of the spectrum of the Lund laser plasma source, 30% NH3 dissolved in H2O,
laser pulse: E=70mJ, τ≈120ps, λ=532nm, 3mJ UV prepulse [10]
The measured linewidth is probablylimited by the effect of the extended source on
the spectrograph performance. The source size of ≈30µm is in good agreement with
other measurements and mainly influenced by the expansion of the plasma during
the time delay between pre- and main pulse [13]. Integrating over the entire
linewidth and the source size gives for the Ly-α line 4.4×1011photons/(pulse×sr) and
3.2×1011photons/(pulse×sr) for the He-α line [10].
4.2 Experiment with 8 ns IR Laser Pulses
at the Fraunhofer Institut für Lasertechnik in Aachen
An experiment using ≈8ns IR laser pulses from a Nd:YAG laser has been carried out
in collaboration with G. Schriever, S. Mager, K. Gäbel and R. Lebert at the Fraunhofer Institut für Lasertechnik in Aachen [16]. The ≈1000 mJ laser pulses at
λ=1064 nm supplied a target intensity of ≈1013W/cm2. Spectra have been taken from
cryogenic (frozen nitrogen) and solid (boron nitride) targets, the target refresh
limiting the repition rate to 0.03–1 Hz in the given experiment [17]. Figures 7 and 8
represent the measured pulse brilliances. The relatively large measured source size of
≈80µm corresponds to the comparatively long laser pulses and may influence the
appearance of the spectral line width, as discussed above. With the frozen nitrogen
target, relative linewidth in the Ly-α line at λ=2.478 nm and the He-α line at
λ=2.879 nm have been found to be λ/∆λ≥330 and 200, integration over line width
and source size give 1.0×1012 photons/(pulse×sr) and 2.4×1012 photons/(pulse×sr),
respectively.
1.6
N
He-β
0.8
0.4
N
He-α
N
Ly-α
1.2
8
10 Photons / (pulse × µm2 × sr × 0.1% BW)
T. Wilhein
N
N Ly-β
Ly-γ
N
He-δ
N
He-γ
satellites
0.0
2.0
2.2
2.4
2.6
2.8
3.0
λ [nm]
Fig. 7. Spectral brilliance of the Aachen laser plasma source, target: cryogenic,
frozen nitrogen laser pulse: E≈1000mJ, τ≈8ns, λ=1064nm [16]
1.6
N
Ly-α
B
Ly-α
N
He-α
1.2
O
He-α
N
He-β
0.8
0.4
N
N
Ly-β He-γ
N
Ly-γ
8
10 Photons / (pulse × µm2 × sr × 0.1% BW)
V - 32
B
Ly-β
B
C
Ly-α Ly-δ
B
Ly-γ
0.0
2.0
2.5
3.0
3.5
4.0
4.5
5.0
λ [nm]
Fig. 8. Spectrum of the Aachen laser plasma source, target: solid, boron nitride
laser pulse: E≈1000mJ, τ≈8ns, λ=1064nm [16]
Investigations on Laser-Generated Plasma Sources
V - 33
4.3 Experiment with High Intensity UV-Laser Pulses at the University of Jena
In collaboration with the U. Teubner, D. Altenbernd, and E. Förster, Max-PlanckArbeitsgruppe "Röntgenoptik", W. Theobald, R. Haeßner, and R. Sauerbrey, Institut
für Optik und Quantenelektronik, University of Jena, investigations on a laser
plasma created with a KrF∗ -Laser [14] have been performed [15]. The ≈20mJ UV
laser pulses, λ=248nm, pulse width ≈700fs, produced an intensity of ≈1016W/cm2
onto the solid target (used targets: boron nitride and carbon) with a repition rate of
0.5-1Hz. Employing a carbon rod as target, strong emission in the C Ly-α
(λ=3.373nm) and the C He-α (λ=4.027nm) line could be observed, the photon
numbers found in these lines being 1.3×1011photons/(pulse×sr) and 9×1010photons/
(pulse×sr) [15], the relative line width λ/∆λ≥100 and 320, respectively. The
relatively broad appearance of the Ly-α line may indicate that the high laser
intensity lead to line broadening due to the Stark effect. The source size was found to
be ≈25µm, thus, close to the limit given by spectrograph (s. above).
4.4 Discussion of the Experimental Results
The data from the described experiments show that it is possible to achieve pulse
brilliances in the range of 108photons/(pulse×µm2×sr×0.1%BW) in the water
window with the laser plasma sources under investigation, the Lund droplet-target
laser plasma source due to the small source size being the most effective in
converting laser pulse energy into X-ray pulse brilliance. The absolute photon
numbers calculated by integrating over line width and source size divided by the
laser pulse energy gives approximately the same values for the Aachen frozen
nitrogen and the Lund NH3/H2O droplet target, where in addition to the nitrogen
emission strong oxygen lines are observed. The Lund laser plasma source already
operating very stable at a repition rate of 10Hz seems to be well suited to act as a
bright laboratory X-ray source for applications like, e.g., X-ray microscopy, because
it provides high average photon numbers in narrow bandwidth lines. Also the
Aachen laser plasma source promise to become an intense X-ray source, if the
repition rate can be increased. Laser pulses with very high intensities create
comparatively broad line emission in the soft X-ray region, thus a monochromator
may be needed if the plasma should work as a monochromatic X-ray source.
5 Summary
A spectrograph using off-axis reflection zone plates has been developed and tested
with laser produced plasma sources in the spectral range λ=1.5...5nm. The spatial
resolution has been found to be better than 25µm, the spectral resolving power
λ∆λ≥450 at 2.88nm. A ray tracing program allows to control the design of ORZs for
a particular application. With the calibrated spectrograph absolute measurements of
the pulse brilliance of laser generated plasma sources have been performed, the results prooving that pulse brilliances bpulse >108photons/(pulse×µm2×sr×0.1%BW) and
V - 34
T. Wilhein
integrated photon numbers N >1012 photons/(pulse×sr) in a single are achievable in
narrow bandwidth line emission.
Acknowledgements
The author gratefully acknowledges the excellent collaborations with all the researchers in the labs in Lund, Jena and Aachen. Thanks to B. Niemann, T. Schliebe,
G. Schneider, P. Guttmann, J. Herbst, P. Nieschalk from the FE Röntgenphysik,
Göttingen, R. Plontke and his team from LEICA, Jena, and G. Schmahl and D. Rudolph for their continous support of this work.
This project has been funded by the German Federal Minister for Education and
Research (BMBF) under contract number 13N6491.
References
1
2
3
4
5
6
7
8
9
10
11
12
13
See, e.g., G. Schmahl, D. Rudolph, B. Niemann, P. Guttmann, J. Thieme,
G. Schneider, Naturwissenschaften 83 (1996), 61.
F. J. Wuilleumier, Y. Petroff, and I. Nenner, eds., Vacuum Ultraviolet
Radiation Physics, Proc. of the 10th VUV conference (World Scientifc,
Singapore, 1993).
E.-E. Koch, D. E. Eastman, and Y. Farge, Handbook on Synchrotron Radiation
(North Holland, Amsterdam, 1983).
M. C. Richardson, G. A. Kyrala, eds. Applications of Laser Plasma Radiation
II, Proc. SPIE Vol. 2523 (1995).
C. Tillmann, A. Persson, C.-G. Wahlström, S. Svanberg, K. Herrlin, Appl.
Phys. B 61, 333 (1995).
A. G. Michette and C. J. Buckley, eds. X-Ray Science and Technology,
(Institute of Physics, Bristol, 1993).
T. Schliebe, Diffraktive Kondensoroptiken für die Röntgenmikroskopie,
Elektronenstrahllithographie und Nanostrukturierung, PhD-thesis, (Göttingen,
in preparation).
B. Niemann, T.Wilhein, T. Schliebe, R. Plontke, O. Fortagne, I. Stolberg,
M. Zierbock, Microelectr. Eng. 30, 49 (1996).
B. Niemann, New X-ray optical elements generated by the electron beam
lithography system LION LV1, this volume.
T. Wilhein, D. Rothweiler, A. Tusche, F. Scholze, and W. Meyer-Ilse, in X-ray
Microscopy IV, V.V. Aristov and A.I. Erko, eds. (Bogorodskii Pechatnik
Publishers, Chernogolovka, Moscow region, 1994) p. 470.
T. Wilhein, D. Hambach, B. Niemann, M. Berglund, L. Rymell,
H. M. Hertz, Appl. Phys. Lett. 71 (28), (1997).
L. Rymell and H. M. Hertz, Opt. Commun. 103, 105 (1993).
L. Rymell, M. Berglund and H. M. Hertz, Appl. Phys. Lett. 66, 2625 (1995).
M. Berglund, L. Rymell, and H. M. Hertz, Appl. Phys. Lett. 69, 1683 (1996).
Investigations on Laser-Generated Plasma Sources
14
15
16
17
V - 35
U. Teubner, C. Wülker, W. Theobald, E. Förster, Phys. Plasmas 2, 972 (1995);
U. Teubner, W. Theobald, C. Wülker, J. Phys. B 29 (1996) 4333.
S. Szatmari, F. P. Schäfer, Opt. Comm. 68, 196 (1988).
T. Wilhein, R. Häßner, D. Altenbernd, U. Teubner, W. Theobald, E. Förster,
R. Sauerbrey, X-ray brilliance measurements of a subpicosecond laser-plasma
using an elliptical off-axis reflection zone plate, subm. to Opt. Lett.
T. Wilhein, G. Schriever, S. Mager, K. Gäbel, R. Lebert, to be published.
R. Lebert, G. Schriever, S. Mager, A. Naweed, O. Treichel, K. Bergmann,
W. Neff, Laser Produced and Pinch Plasmas: Narrowband X-Ray sources for
Applications. X-Tech 96 Workshop, Berlin, Sept. 29–Oct. 2, 1996.