Laser and Particle Beams, page 1 of 7, 2014.

© Cambridge University Press, 2014 0263-0346/14 $20.00

doi:10.1017/S0263034614000524

Continuous production of a thin ribbon of solid hydrogen

S. GARCIA, D. CHATAIN, AND J. P. PERIN

University of Grenoble Alpes, CEA INAC-SBT, Grenoble, France
(RECEIVED 18 June 2014; ACCEPTED 8 August 2014)

Abstract
The development of very high power laser opens up new horizons in a various field, such as protontherapy in medicine or
laser-matter interaction in physics. The Target Normal Sheath Acceleration phenomenon is used in the first one. After the
laser-matter impact, a plasma is generated, and free electrons move forward. It creates an electrostatic field, which can
accelerate protons at the rear side of the target. The generated beam can be able to contain energetic protons with a
large spectrum(1–200 MeV). This energy distribution depends on the laser power and the nature of the target. This
technique has been validated by accelerating protons coming from hydrogenated contaminant (mainly water) at the rear
of metallic target up to 58 MeV at Lawrence Livermore National Laboratory using the Nova Petawatt laser. However,
several laser teams would like to study this interaction with pure targets. In this context, the low temperature
laboratory, at CEA-Grenoble has developed a cryostat able to continuously produce a thin hydrogen ribbon (50 and
100 μm thick). A new extrusion concept, without any moving part has been carried out, using only the thermodynamic
properties of the fluid. First results and perspectives are presented in this paper.
Keywords: Cryogenic target; hydrogen ribbon; proton beam

1. INTRODUCTION (Cowan, 2004). Typically, conversion efficiencies (ratio be-

tween the laser energy and the total ion beam energy) up to
Recently, the dramatic rise in attainable laser intensity by
4–15 percents are measured (Brenner, 2014; Green, 2014).
means of high power femtosecond lasers has generated a
Various physical mechanisms of laser-driven ion acceler-
fast evolution of a new research field known as laser-driven
ation have been investigated to date. The mechanism most in-
acceleration. The production and acceleration of protons up
vestigated experimentally is the Target Normal Sheath
to 60-MeV level on few micrometers are clearly visible re-
Acceleration (TNSA) when ions are accelerated at the rear
sults of this evolution (Snavely, 2000). The European project
side of a thin target in a quasi-electrostatic sheath formed
ELI will be able to produce a beam of some 1023 W.cm−2.
by fast electrons propagating from the target front side
The spectacular increase in brightness and decrease in pulse
(Hatchett, 2000; Maksimchuk, 2000; see Fig. 1). Usually,
duration of laser-driven particle beams will revolutionize the
protons from surface impurities on metal foils (hydrocar-
way of investigating matter (Malka, 2008). Fundamental
bons) are preferentially accelerated as their charge-to-mass
events in biology, femtochemistry and solid-state physics
ratio is the highest among all ion species, thus no screening
can be investigated with unprecedented space and time resolu-
effect is present. Nevertheless, the concentration of such im-
tion of 10−15–10−10s (Malka, 2008). Moreover, great attention
purities strongly depends on the environmental conditions
is devoted to medical applications of laser-accelerated light
thus causes large fluctuations in the proton current which
ions, especially in the hadron-therapy field (Malka, 2004).
make laser-accelerated beams not suitable for applications.
Using present-day common laser facilities in the 100-TW
In this paper, we propose to investigate experimentally new
power range, ions are typically accelerated from thin (∼ μm)
target geometries able to produce reliable and more stable
foils up to few tens of MeV energy per mass unit. Such
laser driven ion sources for applications. Our scheme is
beams show a very low-emittance (below 10−2 mm mrad for
based on the use of an advanced cryogenic system producing
protons above 10 MeV, which is 100 times better than typical
pure solid hydrogen target which allows minimizing the fluc-
RF accelerators) and extremely high ion current (kA range)
tuations in the produced proton current. For this purpose, a
cryostat has been developed with a specific design of an ap-
Address correspondence and reprint requests to: Stephane Garcia, CEA/
DSM/INAC/SBT/LCF, 17, rue des Martyrs, Bâtiment D1 – Bureau 318, paratus capable to deliver the required targets to a high
38 054 Grenoble Cedex 09, France. E-mail stephane.garcia@cea.fr vacuum chamber.
1
2 S. Garcia et al.

• Figure 2A: The Valve V1 is opened; hydrogen is intro-

duced in the cell and immediately condenses and blocks
the extrusion nozzle.
• Figure 2B: the cell is completely filled with liquid hy-
drogen (the temperature T2 is regulated above the
triple point).
• Figure 2C: The temperature T2 is decreased slowly
below the triple point to have a cell completely filled
with solid.
• Figure 2D: The valve V1 is closed.
• Figure 2E: The temperature T2 is increased, the pressure
in the cell rises, the temperature T1 is regulated close to
the triple point and some solid is extruded through the
nozzle.

Fig. 1. (Color online) Sketch showing the TNSA principle. An hot electron By precisely controlling the temperature of T1 and T2 one
cloud is generated at the rear of the target, creating an electrostatic field can control the solid extrusion velocity at a value close to
which accelerates the protons. 10 mm/s. The solid film then evaporates under the effect
of heat radiation and is pumped with a vacuum pump. This
system was the object of a patent deposited by our laboratory.
The construction of new high power laser facilities
(e.g., high repetition rate petawatt-class lasers as ELI- 3. EXPERIMENTAL SETUP
Beamlines) will clearly enable numerous prospective appli-
As shown in the Figure 3, the cryostat contains a cell sur-
cations based on secondary sources of energetic particles.
rounded by a thermal shroud at 60 K to hit radiation load
In particular, the use of the proposed solid hydrogen
heat. A burst disk prevents the cell if the pressure rises
cryogenic target along with these emerging laser technolo-
over than 20 MPa. A 200 L.s−1 turbo molecular pump
gies will allow demonstrating future medical applications
allows maintaining the cryostat under vacuum (10−6 mbar).
such as hadron therapy (Ledingham, 2007; Margarone,
The three heat exchangers of the cell and the thermal valve
2013).
are cooled with helium coming from a 100 or 250 liquid
The cryostat developed by the low temperature laboratory
liter tank. Four flow control valves are used to adjust the
of the CEA-Grenoble, in France, enables to produce a contin-
helium flow in each heat exchanger. Some heaters are used
uous film of solid H2 of some tens of microns in thickness
to adjust the required temperature of each part. The cryostat
and one millimeter in width. A new extrusion technique is
is equipped with several viewports to be able to observe
used, without any mobile part. Thermodynamic properties
the solid Hydrogen film at the outlet of the nozzle (Fig. 4).
of the fluid are used to make the pressure rise in a cell and
The observation can be done according to two perpendicular
push the solid H2 through a calibrated nozzle.
directions. For the moment, the solid film hydrogen, is ex-
truded in a box which is pumped by a 35 m3/h scroll
pump which keeps the pressure around 10−1 mbar (depend-
2. PRINCIPLE OF SOLID FILM EXTRUSION
ing of the Hydrogen film velocity). This box was mandatory
As discussed above, the solid hydrogen film target produc- because, as it is discussed later, we did not have a pumping
tion is realized by a new extrusion process, developed at apparatus allowing to maintain a sufficient low pressure in
CEA Grenoble. In this one, no moving part is used to the cryostat (this pressure must be lower than 10−3 mbar to
apply the required pressure for the extrusion. The main prin- avoid heat transfer by convection/conduction between the
ciple is based on the use of the thermodynamic properties of vacuum vessel and the cell) during solid hydrogen extrusion.
the fluid. As presented in Figure 2, the experimental cell is When experiences will be performed in lasers vacuum
equipped with two heat exchangers, one situated at its top vessel, this box and all the windows will be suppressed be-
and the other one situated at its bottom, enabling to achieve cause turbo molecular pumps of 2000 L.s−1s will allow ob-
the required temperatures at theses points. taining the required pressure in the vacuum vessel.
The working principle of the cell is the following:
3.1. Required Pumping Means during Extrusion
• First, when the cell is under vacuum, the temperature of
the heat exchanger E1 is regulated below the triple point As the solid hydrogen film sublimates, to insure the laser ex-
of the gas (i.e. 13 K for hydrogen) and the temperature periment, the vacuum level of the experimental chamber has
of the heat exchanger E2 is regulated above (typically to be maintained at a pressure lower than 10−4 mbar. To
20 K for hydrogen). achieve such requirement, the sublimated solid has to be
Thin ribbon of solid hydrogen 3

Fig. 2. (Color online) Sketch showing the solid film extrusion principle.

pumped. The pumping speed of the vacuum system is given following equation:
by the following equation:
2(e + L) ∗ H ∗ σ
S ∗ V ∗ ρsol ∗ 22.4 Pe = . (2)
Q= . (1) e∗L
M ∗P

Where Q is in l/s at standard conditions for temperature and Where σ is the limit shearing stress. This value, depending on
pressure, S is the area of the solid hydrogen film (in m2), V the temperature, is given by Leachman (2010) in Figure 6 for
is the film velocity (in m/s), ρsol is the density of the solid hydrogen, deuterium and neon.
film (in kg/m3) M is the molar mass (in kg/mol) and P is For example, in case of hydrogen at 12 K, σ ∼ 50 kPa, if
the required pressure in the vacuum vessel (in bar). Example: H = 2 mm and e = 10 μm we find an extrusion pressure of
for a H2 solid film of 1 mm × 100 μm having a velocity of 20 MPa. This pressure is easily achievable because, as
10 mm/s, if the required pressure in the vacuum vessel is shown in Figure 7 (doted line), we can see that if we fill
5 × 10−5 mbar, one needs a 5000 L.s−1 pump. the cell with solid at a pressure of 1 MPa at 12 K, the pressure
rises to 20 MPa if the cell is heated at 20 K after having
closed the filling valve of the cell.
3.2. Required Pressure for Extrusion
To extrude through a nozzle, a solid of thickness e, width L
and height H (Fig. 5), the required pressure Pe is given by the

Fig. 3. (Color online) Scheme of the cryostat. Fig. 4. (Color online) Cryostat Sophie II at CEA Grenoble.
4 S. Garcia et al.

imposed on the top of the cell by an heat exchanger. Increas-

ing the filling time enables to rise up the solidification line,
and thus to put more hydrogen in the cell.

3.4. Characterisation of the Target

The characterization of the ribbon is currently realized by two
cameras placed at the front and the side view of the target
(see Figs. 4 and 9). The H2 ribbon is transparent. Preliminary
measurements using an ombroscopy method show that for a
given ribbon velocity, the thickness is constant all along the
observed field (4 × 9 mm2 for the front view and 2 × 4 mm2
for the profile view). The velocity is measured by an image
cross-correlation method. Other techniques are in study to
get a more precise measurement of the thickness and the
roughness of the target.

Fig. 5. (Color online) Description of the applied force on the sheared solid 4. EXPERIMENTAL RESULTS
hydrogen.
First ribbon 1 mm × 100 μm and 1 mm × 50 μm sized have
been obtained. Figure 9 shows such a ribbon. The extrusion
3.3. Required Time to Solidify all the Volume of was realized continuously during 7 hours.
Hydrogen
4.1. Relation between the Temperature of the Cell and
Before closing the filling valve of the cell (V1 in Fig. 2), we
the Pressure
must be sure that all the volume of the cell is well solidified
otherwise the pressure in the cell would not rise as we want. The pressure of the gas (drawn in orange in Fig. 2) has to be
For this reason, we have modeled the cell and calculated the maintained constant. To reach this requirement during
solidification time. extrusion, the gas above the solid part has to be heated up,
Figure 8 shows isotherm lines in the cell after a filling time following the perfect gases law: since the gas volume
of 25 minutes. Assuming a steady state, it enables to get the decreases, the temperature has to be increased to keep a cons-
solidification front. The cell is considered filled when the top tant pressure. The initial mass (m) of hydrogen contained in
right edge’s temperature is below the solidification point: no the cell is of importance for the extrusion behavior. Indeed,
more hydrogen can enter in it. as shown in the Figure 10, obtained using hydrogen proper-
The dark blue part corresponds to temperature below the ties and equation of states, as the initial filling of the cell in-
solidification point (14.6 K at 10 bar) whereas the light creases, the needed temperature for the top of the cell
blue corresponds to temperature above (up to 15.1 K). The decreases to achieve a given pressure. Moreover, the less
red zone represents edges of the steel cell. The filling and the initial filling, the higher the temperature gradient between
solidification time depend on the temperature evolution two given pressures.

Fig. 6. (Color online) Curves of limit strength of hydrogen, deuterium, and neon.
Thin ribbon of solid hydrogen 5

4.2. Influence of the Nozzle Temperature on the

Hydrogen Film Velocity
Experiments have been carried out to study the influence of
the temperature at the extrusion nozzle Tnozzle, the tempera-
ture at the bottom of the cell (T ), and the applied pressure
inside the cell for the 100 μm nozzle as shown in Figure 11.
As the temperature on the bottom of the cell decreases, the
needed pressure to achieve a given velocity increases. It is in
total agreement with Figure 6, which shows that the lower the
temperature, the higher the shearing stress, and thus the
needed pressure to extrude.
The temperature difference between the bottom of the cell
Fig. 7. (Color online) Phase diagram of hydrogen—blue and pink line are (T ) and the extrusion nozzle Tnozzle seems to have a larger in-
respectively solid/liquid and liquid/gas limits—the red dotted curve is an fluence for low temperatures of the nozzle (11.5 K) than for
isochore line. dτ0
higher one (12.5 K), due to the shearing stress gradient ,
dT
shown in Figure 6. Here also, the higher the temperature of
the bottom, the lower the needed pressure for a given velocity

4.3. Influence of the Pressure on the Hydrogen film

Velocity
Figure 11 also shows that for a given themperature on the
bottom of the cell, the higher the pressure, the higher the

Fig. 8. (Color online) 2D-axisymetric representation of the solidification line Fig. 9. (Color online) Snapshot of the front view of hydrogen ribbon during
for a filling time of 25 minutes. Red zone represents edges of the steel cell. extrusing through an extrusion nozzle 100 μm large.
6 S. Garcia et al.

First laser shots on these solid H2 ribbons are expected in

January 2015 at IoP Prague. The image cross-correlation al-
gorithm will also be improved. It will enable to increase the
precision of the velocity measurements by decreasing the
error bar down to pixels resolution.

6. CONCLUSION
A cryostat allowing the production of a continuous thin solid
hydrogen ribbon was studied and tested at CEA-Grenoble
(France). Ribbons of 100 μm and 50 μm in thickness and
1 mm in width were obtained continuously during more
than 7 hours at a velocity in the range of 1–14 mm/s. This
project enabled to validate a new extrusion concept in
which no moving part is used. First results are promising
and open several perspectives for the use of solid H2 targets
in miscellaneous field such as protontherapy in medicine or
laser-matter interaction in physics. Indeed, even if the
Fig. 10. (Color online) Pressure-temperature relation into the cell for differ- 100 μm targets are too thick for TNSA experiments, it en-
ent initial fillings (m).
abled to solve technological limitation and leaded to the cur-
rent thickness of 50 μm, corresponding to the upper limit of
velocity, which is the expected behavior. It also shows that the typical thickness for TNSA target (1–50 μm). This tech-
there is a minimum threshold, which depends on the temper- nique is currently the only one enabling to continuously pro-
ature, to achieve before the solid starts to be extruded. duce suitable target for laser experiment. Discussion with IoP
team enabled to take into account all the geometric con-
straints of typical laser setups. First laser shots are expected
at the beginning of 2015 at PALS.
5. NEXT STEP
The development has been done with a 100 μm nozzle, but
the extrusion nozzle has already been replaced by a 50 μm 7. ACKNOWLEDGMENTS
one and will be replaced by smaller ones, to obtain thinner Authors are grateful to the whole support team working on this
hydrogen ribbons (25 μm and then 10 μm thick) and to estab- project at CEA-SBT Grenoble, and the IoP team (especially Dr.
lish a precise rheological behavior of the solid during D.Margarone and J.Prokupek) for its future collaboration with the
extrusion. implantation of the cryostat in PALS vacuum chamber. This work
Finite elements simulation will also be carried out to is supported by LANEF Grenoble through a LANEF PhD thesis.
compare this model with the bibliography (Vinyar,2000;
Leachman, 2010).
REFERENCES
BRENNER, C. (2014). High energy conversion efficiency in laser-
proton acceleration by controlling laser energy deposition onto
thin foil targets. App. Phys. Lett. 104.
COWAN, T. (2004). Ultralow emittance, multi-mev proton beams
from a laser virtual-cathode plasma accelerator. Phys. Rev.
Lett. 92, 204801–1.
GREEN, J. (2014). High efficiency proton beam generation through
target thickness control in femtosecond laser-plasma interac-
tions. App. Phys. Lett. 104.
HATCHETT, S. (2000). Electron, photon, and ion beams from the rel-
ativistic interaction of petawatt laser pulses with solid targets.
Phys. Plasmas 7, 2076.
LEACHMAN, J. (2010). Thermophysical properties ans modeling of a
hydrogenic pellet production system. Mechanical engineering,
University of Wisconsin-Madison.
LEDINGHAM, K. (2007). Laser-driven proton oncology – a unique
Fig. 11. (Color online) Velocity of the hydrogen ribbon as function of new cancer therapy? British J. Radiol. 80, 855–858.
applied pressure for different temperature of the extrusion nozzle 0.1 × Maksimchuk (2000). Forward ion acceleration in thin films driven
1 mm2. by a high-intensity laser. Phys. Rev. Lett. 84, 4108–4111.
Thin ribbon of solid hydrogen 7

MALKA, V. (2004). Practicability of protontherapy using compact Applications at ELI Workshop and Panel, volume 1546, p. 1,
laser systems. Medical Phys. 31, 1587–1592. Catania, Itay. ELIMED 2012.
MALKA, V. (2008). Principles and applications of compact laser- SNAVELY, R. (2000). Intense high-energy proton beams from
plasma accelerators. Nat. Phys. 4, 447–453. petawatt-laser irradiation of solids. Phys. Rev. Lett 85, 2945–2948.
MARGARONE, D. (2013). Preface: 2nd elimed workshop and panel. In VINYAR, L. (2000). Screw extruder for solid hydrogen. Techn. Phys.
A. C. Proceedings (ed.), 2nd MEDical and Multidisciplinary 45, 106–111.