The term excimer laser covers a family of laser systems that uses a
mixture excited by electrical discharges as active medium. Excimer
any diatomic-molecule such as xenon chloride (XeCl) that can exist only
excited: in fact, excimer molecules are bound only in the excited state
unbound in the ground state. Such excimers make good lasing materials
the population inversion is guaranteed as long as excited molecules are
The high wall-plug efficiency and the improved reliability of
excimer lasers promoted their widespread use in many applications
intense ultraviolet light, e.g., photolithography, micro-electronics,
material processing, micromachining, remote sensing,
corneal-corrective-surgery. However, commercial excimer lasers are not
suitable for important applications, such as large area material
processing and laser beam propagation down optical
fibres. This is due, respectively, to the limited (< 5 cm²)
size (which in turn limits the output energy) and to the short (< 30
laser pulsewidth typical of commercial excimer laser systems. These
were addressed at the ENEA Frascati back in the late seventies, when
started on designing and constructing of XeCl lasers with large active
high output energy and long pulsewidth.
During 1999, the work at Frascati was mainly focused on using the XeCl laser facility Hercules to anneal amorphous silicon panels and to drive a plasma X-ray source. These two activities were done in the frame, respectively, of the European Project FOTO (aimed at achieving high-mobility thin-film-transistors) and of an INTAS Project for the development of an high-charge, heavy-ions source driven by laser-plasmas. In addition, work was done on the optimisation of the X-ray laser-plasma source for microlithography applications, on the optimisation of an hybrid laser system emitting ultrashort UV laser pulses and on pictures restoration by selective laser paint removal.
The project FOTO up
The European project FOTO aims at the realisation of a clean room
laser annealing of amorphous silicon (a-Si) in order to achieve
high-mobility thin-film-transistors (TFTs) and high-efficiency
photovoltaic cells. This project involves many partners: among them, we
had a close collaboration with
the ERG FORI Division (ENEA Portici), CNR IESS (Rome) and the Company
S.p.A. (Calenzano). In this frame, the main tasks of our Laboratory
Hercules L excimer laser
The new XeCl excimer laser, named “Hercules L”, is an improved version of the laser facility Hercules described in the 1998 Progress Report of the ENEA Applied Physics Division. Table 1 compares the performance of Hercules and Hercules L. The main improvements can be summarised as follows:
Table 1. Main features of the laser systems Hercules (Frascati) and Hercules L (Portici)
|HERCULES (Frascati)||HERCULES L (Portici)|
|Active medium pressure||3,5 bar||5 bar|
|Wavelength||308 nm||308 nm|
|Preionisation||X-rays (70 keV)||X-rays (80 keV)|
|Discharge voltage||40 kV||50 kV|
|Laser pulse width||120-160 ns||100-150 ns|
|Max output energy||8 J||> 10 J|
|Peak power||50 – 70 MW||> 100 MW|
|Maximum repetition rate||10 Hz||10 Hz|
|Average power||35 – 40 W||> 100 W|
|Laser beam size||5 cm x 10 cm||5 cm x 10 cm|
Laser beam homogeniser
The second task was the design and test of a laser beam homogeniser for improving the intrinsic spatial fluctuations of the laser beam energy. A good spatial uniformity of the laser beam incident onto the a-Si panel is a preliminary condition to ensure a good uniformity of the poly-Si grain size and of the electric properties of the TFTs. During 1998, a first prototype of beam homogeniser was designed, mounted and tested on Hercules, as detailed in the 1998 Progress Report of the ENEA Applied Physics Division. We used this homogeneiser to irradiate a-Si films, obtaining the results detailed in the following paragraph. Based on these results, a new beam homogeniser was designed to irradiate the a-Si at high (0.8 J/cm2) and low (0.4 J/cm2) energy density level. The new homogeniser was designed thanks to dedicated software we have developed for (see Fig. 2.1). The basic principle of the beam homogeniser is using cylindrical lenses to divide the incoming beam in a 2-D matrix of secondary beamlets that overlap in the focal plane of the condenser lens. In this way, local spatial fluctuations of the beam energy are averaged due to the overlap of different portions of the input beam. Figure 1 shows the calculated ray-tracing of the homogeniser acting along one direction, with four secondary beamlets.
As a further step toward a versatile and general-purpose beam homogeniser, we have studied a novel optical scheme that allows to continuously change the spot size of the homogenised beam, still maintaining the same degree of homogeneity and, within proper limits, the same length of the homogeniser.
Fig. 2 – 2-D schematic of the working
the laser beam homogeniser as drawn by the ENEA software
Optimum working point
The third task within the project FOTO was seeking the optimum irradiation conditions to achieve poly-Si grains uniformly distributed and having a size as large as possible. Irradiation experiments were carried out at Frascati by using Hercules and the first prototype homogeniser. Colleagues of the ERG-FORI Division at ENEA Portici Centre prepared the a-Si samples. Preliminarily, we obtained a curve grain-size vs. laser-energy-density. This measurement was done in single-shot on 50-nm-thick a-Si films deposited on glass by using the Hercules beam without the homogeniser. By a careful measurement of both intensity profile and grain size (the latter performed in collaboration with the CNR-IESS laboratories), we obtained the plot shown in Fig. 2.
Fig. 2 - Plot of the average grain size of
poly-Si vs. the laser energy density, as obtained by irradiating a-Si
films on glass
with Hercules. The experimental uncertainty of the laser fluence values
less than 10%.
In the figure, we can discern three regions: the first and the third
are characterised by a grain size of about 100-200 nm and a behaviour
independent of the laser energy density. In the second one, small
of the energy density have dramatic consequences on the size of the
grains: we found a gradient (grain size / laser fluence) as large as
This phenomenon is known as Super-Lateral-Growth (SLG) and it is
characterised by a total melting of the a-Si layer. From Fig. 2 it is
clear that this working
point is hardly manageable, so that we decided to work away the SLG
Other measurements carried at Frascati and at the CNR-IESS laboratories have shown for the first time the advantage of using a long-duration laser pulse in order to increase the grain size on the poly-Si film. Fig. 3 shows the maximum size of the grains obtained in the SLG regime by using a short pulse (25 ns, emitted by a commercial laser system) and a long-pulse (150 ns, emitted by Hercules) XeCl laser.
Fig. 3 – Maximum lateral growth of the grains
films having different thickness,
respectively irradiated by Hercules (squares) and by a Lambda Physik XeCl laser (triangles).
An unexpected phenomenon we observed during the a-Si irradiation was a sort of fringe texture, occasionally visible on the surface of the poly-Si panels, as shown in Fig. 4. The results of accurate measurements showed that this pattern is due to the interference of the photons diffracted by the borders of the cylindrical lens arrays of the homogeniser. This interference appears only when approaching the highly non-linear SLG regime, because very small energy density variations just above and just below the SLG threshold are magnified by the non linear grain-size behaviour, see Fig. 2. An experimental method to eliminate this effect is in progress.
Fig. 4 - Photo of an irradiated poly-Si
texture-like pattern due to the interference
of the laser beam diffracted by the cylindrical lens array of the homogeneizer.
The dark regions are characterized by very large grains (size > 1 mm, due to the SLG),
while, in the clear zones, the grain size is about 0.2 mm.
Selective paint removal for old pictures restoration up
In collaboration with the Istituto Centrale del Restauro (Rome) we
have performed some experiments devoted to find the optimum laser
to selectively remove thin layers from ancient oil paintings.
were done irradiating a number of pictures made by overlapped layers of
painting, covered by a layer of garanza-lacquer. The raw-wood substrate
previously prepared according to the old recipe based on chalk of
linseed oil and rabbit glue. The proper Hercules energy density to
the external layer of oil paint without damaging the underlying paint
1 J/cm2. This energy density value is substantially larger
the 0.4 J/cm2 reported in literature by using excimer lasers
a pulsewidth six times shorter than the Hercules one. These results
a question about the relative importance among energy density and power
in paint removal processes. We also observed that multiple shots remove
thicker paint layer than a single shot, for the same total energy
This phenomenon is probably related to the screening effect of the
The laser ion source up
Our laser-plasma X-ray source is driven by focusing the laser beam
on a tape target, as described in the 1998 Progress Report of the ENEA
Physics Division. It emits ions at a kinetic energy up to more than 10
and with a peak current density up to more than 10 mA/cm2 (measured at
meter from the source). Potentially, laser ion sources are promising
single-turn injection into a synchrotron. During 1998-1999 we were
in an European-funded Project (INTAS open 97-2090) for the development
highly charged heavy ions source for the CERN hadrons collider. This
involves 6 participants (3 from Europe and 3 from the New Independent
of the former Soviet Union) and it has the purpose of developing and
different laser ion sources. In particular, the ENEA group is the
of the project and has the responsibility of a laser ion source. The
analysis of the ions emitted by the laser plasma driven by Hercules
a poor ionisation state (see the 1998 Progress Report). During 1999, a
investigation of the ionisation-state as a function of the flying
of ions has been carried on by a spectral analysis of the soft X-rays
by the ions. We discovered that close to the source (up to few mm of
the ionisation state is very high (+19 for a copper target and much
for heavier targets like tantalum) but during the fly the ionisation
reduces due to the strong recombination. A special optical system
a coaxial irradiation scheme is going to be developed, in order to
ions in the same direction of the laser beam. In this way, the
should be much less effective due to the higher temperature of the
during the early stage of expansion.
Microlithography applications of the laser-plasma source up
Potentially, the Extreme Ultra-Violet projection Lithography (EUVL)
a sub-100 nm spatial resolution for microchip fabrication. This is a
goal for the future microelectronics and nano-mechanics. Our
source is very efficient in the EUV (up to 20% in the 40-70 eV spectral
so that a moderate repetition rate of 100 Hz would be sufficient to
a good production yield of 60 wafer layers (having a 300-mm-diameter)
A main problem of solid-target plasma-sources is the emission of debris (small drops of melted target), which limits the lifetime of the first mirror that collects the EUV radiation from the source and reflects it to the following optics. During 1999 we continued to investigate the possibility to reduce the flux of the debris emitted by our laser-plasma source. We successfully damped the large-size debris (diameter F > 1mm) by a proper choice of the target material and by reducing the laser intensity on the target. In this way, in fact, the plasma temperature falls down and the spectral emission shifts to longer (EUV) wavelengths. The debris flux was reduced by more than two orders of magnitude, as shown in Fig. 5.
Fig. 5 – Images of debris flux in different experimental conditions. FL= focused laser beam diameter; IL= laser intensity.
450 shots in 1 bar He
600 shots in 1 bar He
500 shots in vacuum
Even the speed of the debris considerably decreased as a result of the lower plasma pressure, thus allowing the use of mechanical stoppers. Finally, we could reach a debris flux of only 6 debris/shot/sr/J, corresponding to a lifetime of the first mirror of about 107 shots. In the condition of low laser intensity, we also got a very good shot-to-shot stability (see Fig. 6), which is an important parameter for EUVL applications. We are investigating the interaction of small-size debris (F < 1mm) and of ions with different gases (see Fig. 7). The final goal is the development of a clean EUVL prototype demonstrating a sub-100 nm resolution. This project will involve different institutions (ENEA, Universities, private companies, etc.) and will be based both on the solid target laser-plasma source developed at ENEA and on the gas plasma source (pumped by a capillary electric discharge) developed in collaboration with the INFN laboratories of L'Aquila.
Fig. 6 - Burst of 28 X-ray pulses @ hn=40-70 eV, obtained from a Ta target at a laser
of 1010 W/cm² with a repetition rate of 0.5 Hz. The EUV
energy fluctuation is 1.1% rms.
To keep the figure compact, the sampling is stopped after each shot (laser beam: 120 ns - 4 J/shot).
Fig. 7 - The soft X-ray laser plasma source.
Note the blue-light emitted by the plume, due to the interaction of the
emitted ions with air at 0.7 mbar (a) or with a helium gas at 20 mbar
(b). The laser
beam, focused by the lens on the right hand of the photos is not
On the copper tape target is visible the chain of holes due to copper
by the former laser shots.
The ultrashort UV
laser source up
In the frame of a project ENEA-MURST, we installed and tested a commercial laser system emitting ultrashort (< 10-12 s) pulses at 308 nm. This FAMP laser system uses an excimer laser to pump several dye cells in cascade, emitting 100 ps pulses within a broad spectrum in the visible. These pulses pump a distributed-feedback dye laser (DFDL) unit, which generates ultrashort pulses (< 1 ps) tunable to the desired wavelength, in our case 616 nm. The DFDL pulses are then amplified and frequency-doubled to 308 nm by a second harmonic generation crystal. The final goal is to use it as a seed pulse to be injected in a large XeCl laser systems for picosecond pulse amplification. Figure 8 shows a photo of this hybrid excimer-dye-DFDL laser source.
Fig. 8 - Photo of the picosecond laser
operative in the excimer laboratory.
The plastic cables are coloured by the flowing dyes used as active media.
The measured energy and the time duration of the radiation emitted
laser are, respectively, 10 mJ and 0.5 ps,
figure 9. According to the simulations (performed in collaboration with
Pecs University), a proper amplification of this ultrashort pulses with
Hercules or Ianus system should deliver a pulsed beam with up to 0.5 J
1 ps at 10 Hz, corresponding to a peak power of 0.5 Terawatt.
Fig. 9 – Curve achieved by a Michelson-like,
The FWHM laser pulse duration estimated by this autocorrelation curve is 0.5x10-12 s.
T. Letardi, P. Di Lazzaro, G. Giordano, C.E. Zheng: "A note for
design” RT/INN/99 (1999).
S. Bollanti, P. Di Lazzaro, F. Flora, G. Giordano, T. Letardi, D. Murra, G. Schina, C. E. Zheng: “Pulsed X-ray diode with a long-lifetime plasma cathode” Appl. Phys. B 68, 683 (1999)
S. Bollanti, G. Clementi, P. Di Lazzaro, F. Flora, G. Giordano, T. Letardi, F. Muzzi, G. Schina, C.E. Zheng: "Excimer lamp pumped by a triggered discharge" IEEE Trans. Plasma Sci. 27, 211 (1999)
F. Rosmej, A. Faenov, T. Pikuz, I. Skobolev, F. Flora, S. Bollanti, P. Di Lazzaro, T. Letardi, K. Vigli-Papadaki , A. Reale, L. Palladino, G. Tomassetti, A. Scafati, L. Reale, T. Auguste, P. Oliveira, S. Hulin, P. Monot, A. Zigler, M Frankel: “Radiation from autoionising levels correlated with single excited states of highly charged ions in dense cold plasmas” Physica Scripta 80, 547 (1999)
S. Bollanti, F. Bonfigli, P. Di Lazzaro, F. Flora, T. Letardi and D. Murra: “Repetition-rate influence on the beam quality of a XeCl excimer laser” Opt. Commun. 167, 291 (1999)
M. Belli, A. Nottola, F. Flora, T. Jin: “Irradiation of cultured mammalian cells with ultrasoft X-rays: experimental set-up and dose calculation for non-monochromatic beams”, Radiation Physics and Chemistry, 54, 393 (1999)
S.V. Kukhlevsky, Cs Vér, J. Kaiser, L. Kozma, L. Palladino, A Reale, G. Tomassetti, F. Flora and G. Giordano “Generation of high density, pure metal vapour plasma by capillary discharge”, Appl. Phys. Lett., 74, 1 (1999)
S.V. Kukhlevsky, J. Kaiser, L. Palladino, A. Reale, G. Tommasetti, F. Flora, G. Giordano, L. Kozma, M. Lišca and O. Samek: "Physical processes in high-density ablation-controlled capillary plasmas", Physics Letters A 258, 335 (1999)
D. Della Sala, C. Privato, P. Di Lazzaro, G. Fortunato:
“Microelettronica gigante” Energia, Ambiente Innovazione 4, 52 (1999)
F. Flora, S. Bollanti, P. Di Lazzaro, G. Giordano, T. Letardi, A. Nottola, A. Marinai, K. Papadaki, G. Schina, P. Albertano, L. Palladino, A. Reale, A. Scafati, L. Reale, M. Belli, F. Ianzini, A. Tabocchini, A. Grilli, A. Faenov, T. Pikuz: “Applications of a soft X-ray plasma source pumped by a long pulse excimer laser", Proc. Int. Conf. on Lasers '98, V.J. Corcoran, T. Goldman Eds. (STS Press Mc Lean, VA 1999) pp. 454 – 461.
T.A. Pikuz, A. Ya. Faenov, M. Fraenkel, A. Zigler, F. Flora, S. Bollanti, P. Di Lazzaro, T: Letardi, A. Grilli, L. Palladino, G. Tomassetti, A. Reale, L. Reale, A. Scafati, T. Limongi: “Large-field high resolution X-ray monochromatic microscope, based on spherical crystal and high-repetition-rate laser-produced plasmas”, EUV, X-Ray, and Neutron Optics and Sources, C. McDonald, K. Goldberg, J. Maldonado, H. Chen-Mayer, S. Vernon Eds. SPIE vol. 3767, (1999) pp. 67 - 78.
S. Bollanti, P. Di Lazzaro, F. Flora, T. Letardi, A. Marinai, A.
Nottola, K. Vigli-Papadaki, A. Vitali, F. Bonfigli, N. Lisi, L.
Palladino, A. Reale, C.E. Zheng: “Toward a high average power and
debris free soft X-ray source for microlithography, pumped by a long
pulse excimer laser” EUV, X-Ray, and
Neutron Optics and Sources, C. McDonald, K. Goldberg, J. Maldonado, H.
S. Vernon Eds. SPIE vol. 3767, (1999) pp. 33 - 44.
S. Bollanti, P. Di Lazzaro, D. Murra, X. Cheng: “Omogeneizzazione spaziale di fascio laser: disegno delle ottiche ed esperimenti” SIF 99, Congresso della Società Italiana di Fisica (Pavia, Settembre 1999)
P. Di Lazzaro, S. Bollanti, F. Bonfigli, F. Flora, T. Letardi, D. Murra: “Beam quality of an XeCl laser vs. the repetition rate” Int. Forum on AHPLA ’99 Advanced High-Power Lasers (Osaka, Giappone, Novembre 1999)
T. Letardi, A. Baldesi, S. Bollanti, F. Bonfigli, P. Di Lazzaro, F. Flora, G. Giordano, D. Murra, G. Schina, C.E. Zheng “ Industrial large aperture XeCl laser for surface processing” Int. Forum on AHPLA ’99 Advanced High-Power Lasers (Osaka, Giappone, Novembre 1999)
S. Bollanti, F. Bonfigli, D. Della Sala, P. Di Lazzaro, D. Murra “Experimental results on silicon annealing by a long pulse XeCl laser” Int. Conf. ALT ’99, Advanced Laser Technologies (Potenza-Lecce, Settembre 1999)
S. Bollanti, F. Bonfigli, E. Burattini, P. Di Lazzaro, F. Flora, A.
T. Letardi, N. Lisi, A. Marinai, A. Nottola, L. Palladino, A. Reale, K.
C.E. Zheng “High efficiency XUV plasma source at 10-30 nm for
microlithography, pumped by a long pulse excimer laser” MNE ’99, Micro
Nano Engeneering (Roma, Settembre 1999)
P. Di Lazzaro, D. Murra: “Omogeneizzazione di fascio laser e
annealing di silicio amorfo: risultati sperimentali nell’ambito del