Multi-MeV proton source investigations in ultraintense laser- foil
interactions
Borghesi, M., Mackinnon, A. J., Campbell, D. H., Hicks, D. G., Kar, S., Patel, P. K., Price, D., Romagnani, L.,
Schiavi, A., & Willi, O. (2004). Multi-MeV proton source investigations in ultraintense laser- foil interactions.
Physical Review Letters
,
9
(5), [055003]. https://doi.org/10.1103/PhysRevLett.92.055003
Published in:
Physical Review Letters
Document Version:
Publisher's PDF, also known as Version of record
Queen's University Belfast - Research Portal:
Link to publication record in Queen's University Belfast Research Portal
Publisher rights
© 2004 The American Physical Society
General rights
Copyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or other
copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated
with these rights.
Take down policy
The Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made to
ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in the
Research Portal that you believe breaches copyright or violates any law, please contact openaccess@qub.ac.uk.
Download date:09. Aug. 2022
Multi-MeV Proton Source Investigations in Ultraintense Laser-Foil Interactions
M. Borghesi,
1
A. J. Mackinnon,
2
D. H. Campbell,
3
D. G. Hicks,
2
S. Ka r,
1
P. K . P a t e l ,
2
D. Price,
2
L. Romagnani,
1
A. Schiavi,
4
and O. Willi
5
1
Department of Pure and Applied Physics, The Queen’s University of Belfast, Belfast BT7 1NN, United Kingdom
2
Lawrence Livermore National Laboratory, Livermore, California 94551, USA
3
The Blackett Laboratory, Imperial College, London SW7 2BZ, United Kingdom
4
Dipartimento di Energetica, Universita
´
La Sapienza, 00161 Rome, Italy
5
Heinriche-Heine-Universitat, 40225 Dusseldorf, Germany
(Received 17 February 2003; published 5 February 2004)
A study of the properties of multi-MeV proton emission from thin foils following ultraintense laser
irradiation has been carried out. It has been shown that the protons are emitted, in a quasilaminar
fashion, from a region of transverse size of the order of 100–200 m. The imaging properties of the
proton source are equivalent to those of a much smaller source located several hundred m in front of
the foil. This finding has been obtained by analyzing proton radiographs of periodically structured test
objects, and is corroborated by observations of proton emission from laser-heated thick targets.
DOI: 10.1103/PhysRevLett.92.055003 PACS numbers: 52.38.Kd, 52.59.–f
The production of protons during ultraintense laser
interaction with thin foils is a subject which has recently
received a great deal of attention, from both experimental
[1–3] and theoretical [4 –6] points of view. The protons
are accelerated via the space-charge force setup by the
displacement of energetic electrons directly accelerated
by the laser pulse. The location of the dominant accel-
eration mechanism is still an object of debate [1–6]. The
properties of the proton beam are of interest for a number
of important applications, which include the ignition of
compressed fusion capsules [7] and probing of laser-
plasma experiments [8]. One of the attractions presented
by this latter application is the high spatial resolution
achievable [8,9] when back-illuminating an object with
the proton beam. In a point-projection imaging scheme,
this indicates emission from a source with a radius of a
few microns. In this Letter we present the results of a
series of investigations of the source properties, all con-
sistently showing that protons are emitted in a laminar
fashion from an area of the target much larger than
suggested by the resolution tests. The imaging properties
of this source are equivalent to those of a virtual source
located several hundred microns in front of the target.
This study is the first to demonstrate directly this very
important feature of proton emission, only hinted at in
previous work [6]. These results are of primary impor-
tance for the correct understanding both of the emission
properties of the proton beams and of their envisaged and
present applications.
The experiments were carried out using the JanUSP
and VULCAN laser facilities, respectively, located at the
Lawrence Livermore National Laboratory (LLNL) and at
the Rutherford Appleton Laboratory (RAL, U.K.). The
chirped pulse amplified (CPA) JanUSP pulse at LLNL has
wavelength 0:8 m and 100 fs duration. It was focused on
target by an f=2 off-axis parabola (OAP), p polarized, at
an angle of incidence of 22
, with a focal spot of 3–5 m,
full width at half maximum (FWHM). This spot con-
tained 30%– 40% of the energy, giving a peak intensity
in excess of 10
20
W=cm
2
. The VULCAN laser, operating
in the CPA mode, provides 1 :054 m, 1 ps pulses with
energy up to 100 J. When focused on target by an F/3.5
OAP (usually p polarized, at a 15
incidence with the
target normal), the focal spot varied between 8 and
10 m in diameter FWHM, containing 30%– 40% of
the energy (up to 10 J), and giving intensities up to
5–710
19
W=cm
2
. The pulses were focused onto the
surface of Al foils of various thicknesses, and the protons
emitted at the back of the target were detected with
stacks of radiochromic film (RCF) and nuclear track
detector (CR39) layers. It should be noted here that pro-
tons can be accelerated even from a metallic target due to
a hydrocarbon (or water vapor) contaminant layer always
present at the target surface in standard experimental
conditions [10]. The energy spectrum and divergence of
the proton beam were consistent with previous observa-
tions carried out using the same systems [2,3].
In the JanUSP experiment the proton source (from
3 m Al foils) was used to backlight static objects.
Strong modulations can be imprinted in the proton
beam even with thin objects, where the collisional stop-
ping of protons is negligible. The modulations arise from
multiple small-angle scattering in the object which in-
creases the local divergence of the beam [11]. In the case
of narrow objects, this effect can produce a high-contrast,
magnified shadow of the object.
In Fig. 1, shadows of electroformed Cu mesh grids are
shown. The meshes are formed by 10 m lines with
31 m spacing and 5 m thickness. The mesh was placed
parallel to the proton-producing foil at a distance d from
the foil of 0:6 0:05 mm in the case of Fig. 2(a) and
1:0 0:05 mm [Fig. 2(b)]. In both cases the RCF layer
shown was located at a distance L 23:8 0:5mmfrom
the foil. The position of the layer in the multilayered
PHYSICAL REVIEW LETTERS
week ending
6 FEBRUARY 2004
VOLUME 92, NUMBER 5
055003-1 0031-9007=04=92(5)=055003(4)$22.50 2004 The American Physical Society 055003-1
detector was such as to record protons with energies
around 15 MeV. A line out in optical density (OD) across
the shadow of the mesh of Fig. 1(b) is shown in 1(c).
Important information was provided by magnification
tests carried out using the periodic structures. The mag-
nification expected for a point-projection imaging
scheme is simply the ratio M
G
L=d, where L is the
source-to-detector distance and d is the source-to-object
distance. The experimental magnification was measured
in separate shots by placing the mesh at different dis-
tances from the target, while the detector was kept at the
fixed distance of 23.8 mm. The averaged period of the
shadow in the detector plane was divided by the mesh
period, yielding the experimental magnification M
exp
.
The measured values of M
exp
for various mesh positions
are plotted in Fig. 2 against the value of M
G
expected,
and are consistently smaller than the values expected
from purely geometrical considerations. The effect be-
comes more and more important as the magnification is
increased, i.e., as the mesh is placed closer to the source.
This discrepancy can be explained by supposing that
the point source is not located at the target plane, but at a
distance x in front of the target. In this case the magni-
fication is M
exp
L x=d x <M
G
L=d. If one
makes this assumption, the experimental magnification
M
exp
can be expressed as a function of the geometrical
magnification M
G
expected from a point source located at
the target plane, i.e., M
exp
M
G
L x=L M
G
x.The
data of Fig. 2 have been fitted using such a function, where
L 23:8mmand x is a free parameter. As can be seen,
the fit reproduces remarkably well the trend of the data.
The best fit, plotted in Fig. 3, gives x 400 m.The
standard deviation due to shot-to-shot fluctuations is ap-
proximately x 150 m . The average 1=e radius of the
proton beam at the detector plane is R 0:5cm(with a
shot-to-shot random error of 0.1 cm), giving an average
ρ
θ
∆θ
x
Virtual
source
Target
0
5
10
15
20
25
30
0 10203040
M
G
(a)
(b)
a
M
exp
FIG. 2. (a) Plot of experimental magnification versus pre-
dicted geometrical magnification for meshes backlit with
15 MeV laser-produced protons; the dashed line is the best fit
using the function defined in the text, while the solid line
corresponds to M
exp
M
G
; the error bars for M
exp
and M
G
arise from experimental uncertainties in the measurement of
the distances L and d, and of the periods of the mesh grid and
its shadow in the detector. (b) Sketch illustrating the virtual
source description of quasilaminar emission from the target.
FIG. 1. (a),(b) Shadow of grid meshes (period: 31 m, line-
width: 10 m, line thickness: 5 m) impressed in the proton
beam profile, with the grid placed, respectively, at 0.6 and 1 mm
from the Al foil producing the protons; the images are col-
lected with RCF placed at 23.8 mm from the Al foil. (c) Profile
of optical density (OD, solid line) across mesh shadow in (b).
The calculated OD profile for a 10 m diameter virtual source
size is also plotted (dashed line). (d) Comparison between the
same experimental OD (solid black line) and calculated profiles
for various source sizes (Gaussian sources, 1=e beam diame-
ter): 1 m (gray solid line), 20 m (dashed line), and 50 m
(dotted line).
FIG. 3. (a) Proton beam profile on the fourth layer of RCF
(corresponding to a beam energy E 10 MeV) from an undis-
turbed Al 250 m foil. (b) Proton beam profile obtained when
a 100 m radius plasma was preformed on the back of the
target. (c) Interferogram showing the preformed plasma at the
time of CPA irradiation. (d) Schematic showing the relative
position of proton emitting region and preformed plasma on the
back surface of the target.
PHYSICAL REVIEW LETTERS
week ending
6 FEBRUARY 2004
VOLUME 92, NUMBER 5
055003-2 055003-2
emission half-angle of 11
. The interception of the emis-
sion cone at the target plane provides an estimate of the
area emitting protons, having a radius of 80 30 m.
Similar considerations can be inferred by close obser-
vation of one of the mesh images. For example, by consid-
ering the shadow in Fig. 1(a) one obtains from M
exp
25
that the virtual source for the corresponding event was
located at x 380 70 m in front of the target. The
divergence given by the cross section at the detector plane
(R 4:1 0:1mm) indicates that the beam, if emitted
from the virtual source, would encompass an area of the
mesh with radius given by r Rd x=L x.Inour
conditions r 170 30 m, corresponding to N
11 1 mesh periods across the proton beam diameter.
As a matter of fact, across the experimental cross section
one can count 12 and 9 periods, respectively, along a
vertical and an horizontal line out, giving an average of
10.5, consistent with the virtual source model. If the point
source were located at the target surface, the beam would
encompass only 6:5: 0:5 mesh periods.
All these considerations are based on the assumption
that the protons forming the image are emitted in a
quasilaminar fashion, and that, in the absence of colli-
sions, they propagate on a straight line after their emis-
sion. This assumption is supported by the fact that the
shadow of the mesh is sharp and that the periodicity of the
mesh structures does not change across the image. As a
matter of fact, if the proton emission were far from
laminar, it would not be possible to project an image of
the mesh at all. The resolution of a 150 m nonlaminar
source would be equivalent to 5–6 periods of the mesh.
An estimate of the size of the virtual source is provided
by a computational code, which has been developed to
simulate the proton beam propagation in the conditions of
the experiment. The program uses the Monte Carlo cal-
culations provided by the code
SRIM [12] to simulate the
collisional effects undergone by the protons when travers-
ing matter. The spatially resolved 2D dose distribution
and, consequently, the optical density distribution in the
RCF layers of the detector are calculated from the
SRIM
outputs. The experimental parameters for magnification
and mesh characteristics are used. The source size can be
adjusted until the calculated optical density modulation
reproduces the experimental one (details of these calcu-
lations will be provided elsewhere). By assuming a
Gaussian source, the best match was obtained for a 1=e
diameter of a 10 m, which is reproduced in Fig. 1(c).
Line outs expected for other source sizes are reproduced
in Fig. 1(d), and show the strong dependence of the OD
profile from the source size. In Fig. 2(b), ‘‘a’’ is the
extension of the virtual source located at a distance x in
front of the target, from which the degree of laminarity
of the source can be estimated; in other words, a circular
section of the proton-emitting region of the target, lo-
cated at a distance from its center, emits particles
within the angle , where =x and a=2x
10 mrad [see Fig. 2(b)]. For such a source, one can roughly
estimate the transverse normalized emittance [13] as
"
0
, where
0
is the radial extent of the proton-
emitting area on the back surface of the target and
v=c is the ratio of the proton velocity to the speed of light.
By taking the experimental values 10 mrad and
0
80 m, one obtains, for 15 MeV protons, "
0:1 mm mrad, which compares well with independently
reported measurements [14]. Additional effects (e.g.,
electrostatic charging of the grid) not included in the
simple collisional model used could, in principle, lead
to even lower estimates for a and ".
A further study of the proton source characteristics
was carried out at the Rutherford Appleton Laboratory
using the VULCAN laser. In a previous experiment, it
was seen that it was possible to eradicate the proton
beam detected at the rear of the target by preforming a
plasma at the back of target ahead of the short pulse
interaction with the foil [2]. In [2], the heating pulse
and the proton-producing pulse focal spots were directly
facing each other on opposite sides of the target. Here
we present similar results, but obtained with the heat-
ing pulse slightly displaced off-axis. The targets used
for this test were 250 m Al foils. A thick foil was chosen
in order to decouple front and back surfaces. The CPA
Vulcan laser pulse was focused onto the front surface of
the foil at an intensity of about 8 10
19
W=cm
2
.The
heating pulse (5 J, 600 ps, 0:527 m) was focused at
an irradiance of 2 10
13
W=cm
2
onto the rear of the
foil, in a focal spot of radius 100 m. The RCF detec-
tor pack was placed at 22 mm from the back of the Al
foil. A transverse 4! probe with ps resolution was em-
ployed to obtain interferograms of the preformed plasma.
For details of the setup please refer to Fig. 1 of Ref. [2]. In
Fig. 3 the proton signal detected onto the fourth RCF
layer (corresponding to an energy of 10 MeV) is shown
for two different conditions: (a) no heating pulse;
(b) heating pulse on, beginning 100 ps before the CPA
interaction. The heating pulse was in this case displaced
from the axis of the interaction pulse. The amount of
vertical displacement can be determined from the corre-
sponding interferogram, taken at the time of CPA irra-
diation, and shown in Fig. 3(c), by measuring the distance
b between the center of symmetry of the preformed
plasma and the CPA interaction axis (in this case
around 65 m). The latter can be identified by the self-
emission and/or the center of symmetry of the plasma
created by CPA on the front of the target. The dimensions
of the plasma at the time of the proton production can also
be obtained from the interferogram; the density contour
corresponding to n
e
10
18
cm
3
has a radius of approxi-
mately 100 m at the target surface.
While the beam displayed in 3(a) shows the character-
istic round shape always observed in proton beams pro-
duced from thick targets, the shape of the beam cross
section is changed in 3(b), as a part of the proton beam
appears to be destroyed by the presence of the plasma.
The right top corner of the beam has disappeared, and the
PHYSICAL REVIEW LETTERS
week ending
6 FEBRUARY 2004
VOLUME 92, NUMBER 5
055003-3 055003-3
cut is with good approximation circular. As shown in
Fig. 3(d) a circle with radius 100 m and with its center
at a distance of 65 m from the CPA interaction axis
matches reasonably well the circular cut. The phenome-
non was repeatable and the position of the cut changed
consistently when the plasma was formed at a different
distance from the CPA axis. When the delay was in-
creased by 100 ps, the whole proton beam disappeared,
down to the energy cutoff determined by our diagnostic
setup (E 2 MeV), as observed in [2]. Interferograms
indicate that at this time the plasma had a radius (ne
10
18
cm
3
)of300 m.
These observations confirm that the protons and/or the
main acceleration mechanism are located at the back
surface of the target, and provide a further indication
that the protons are emitted from a rather large area. If
one takes the distance b as a scale in the target plane, the
radius of the area emitting the protons is given as
130 m. While a plasma with transverse extension of
100 m is able to inhibit plasma acceleration for only a
portion of the beam, a plasma of radius 300 m is able
to cover the whole emitting area and inhibit completely
the acceleration process. Two effects can be responsible
for this: (1) heating and vaporization of the hydrocarbon
layer at the back surface ahead of the arrival of hot
electrons from the front, and (2) dramatic decrease of
the efficiency of back-surface acceleration due to the
increased ion scale length [2]. Also at RAL the virtual
source was much smaller than the emitting area. Knife-
edge measurements of the proton emittance for similar
energies carried out on the same facility indicated a
virtual source size of no more than 10 m.
All these measurements are consistent with quasi-
laminar emission from an extended area of the target.
Similar conclusions have been obtained, with a different
technique, in independent experiments by a separate
group [14]. Quasilaminarity of proton emission has
been predicted in computational work [6] as a conse-
quence of the shape of the Debye sheath accelerating
the electrons at the back of the target. Particle-in-cell
simulations [5,6] show that, in the presence of a spatially
varying transverse electron distribution, the ion front in
the accelerating Debye sheath becomes curved during the
acceleration phase. This determines a 3D accelerating
field with a strong radial component, which becomes
progressively more important away from the axis, ulti-
mately leading to quasilaminar flow [6].
The experimental results show that the emitting area is
considerably larger than the laser focal spot size. In thin
targets, electrons can acquire transverse momentum and
spread transversely across the target, leading to an ex-
tended Debye sheath, due to the ‘‘fountain effect’’ driven
by magnetic fields near the back surface [5]. In addition
to this effect, while refluxing through the target [3],
electrons can acquire transverse momentum due to reflec-
tions from curved sheaths. In thick targets, intrinsic di-
vergence of the electron beam produced at the front of the
target can also be responsible for the large emission
region observed. As a matter of fact, the 20
–30
elec-
tron beam divergence reported in various experiments
[15] is broadly consistent with the emission area inferred
at RAL.
In conclusion, experiments investigating multi-MeV
proton emission from thin foils irradiated with ultra-
intense laser pulses have provided new insight on the
characteristics of the proton source. The protons in the
energy range considered are emitted from an extended
area of radius 100 m, in a quasilaminar fashion, with
very low emittance. For projection purposes, the emission
properties are equivalent to those of a much smaller
virtual source located several hundred microns in front
of the target. This has been proven by careful magnifica-
tion checks employing proton backlighting of periodic,
cold structures. This evidence is reinforced by measure-
ments of proton emission employing targets preheated by
moderate-intensity irradiation, also confirming emission
from a similar-sized area. The finding is of great impor-
tance for applications employing the protons as a high
resolution backlighting or heating source.
This work was funded by LDRD and EPSRC grants,
and by Direct Access to the VULCAN facility. M. B.
acknowledges support from an IRTU-Europe Network-
ing grant. A. J. M. acknowledges partial support from
QUB IRCEP. The authors acknowledge discussions with
Dr. M. Zepf (QUB), Dr. H. Ruhl (General Atomics), and
Professor S. Bulanov (RAS, Moscow).
[1] E. L. Clark et al., Phys. Rev. Lett. 84, 670 (2000);
A. Maksimchuk et al., Phys. Rev. Lett. 84, 4108 (2000);
R. D. Snavely et al., Phys. Rev. Lett. 85, 2945 (2000);
M. Zepf et al., Phys. Rev. Lett. 90, 064801 (2003).
[2] A. J. Mackinnon et al., Phys. Rev. Lett. 86, 1769 (2001).
[3] A. J. Mackinnon et al., Phys. Rev. Lett. 88, 215006
(2002).
[4] S. Hatchett et al., Phys. Plasmas 7, 2076 (2000); H. Ruhl
et al., Plasma Phys. Rep. 27, 363 (2001); T. Zh. Esirkepov
et al., Phys. Rev. Lett. 89, 175003 (2002).
[5] A. Pukhov, Phys. Rev. Lett. 86, 3562 (2001).
[6] S. Wilks et al., Phys. Plasmas 8, 542 (2001).
[7] M. Roth et al., Phys. Rev. Lett. 86, 436 (2001).
[8] M. Borghesi et al., Phys. Plasmas 9, 2214 (2002); M. Bor-
ghesi et al., Phys. Rev. Lett. 88, 135002 (2002).
[9] J. A. Cobble et al., J. Appl. Phys. 92, 1775 (2002).
[10] S. J. Gitomer et al., Phys. Fluids 29, 2679 (1986).
[11] D. West and A. C. Sherwood, Nature (London) 239, 157
(1972).
[12] J. F. P. Ziegler, J. P. Biersack, and U. Littmark, The Stop-
ping and Range of Ions in Solids (Pergamon, New York,
1996).
[13] S. Humpries, Jr., Charged Particle Beams (Wiley and
Sons, New York, 1990).
[14] M. Roth et al., Plasma Phys. Controlled Fusion 44,B99
(2002).
[15] K. B. Wharton et al., Phys. Rev. Lett. 81, 822 (1998);
J. Santos et al., Phys. Rev. Lett. 89, 025001 (2002).
PHYSICAL REVIEW LETTERS
week ending
6 FEBRUARY 2004
VOLUME 92, NUMBER 5
055003-4 055003-4
