A Multi-Wavelength Interferometry System for the Orion Laser
Facility
S.Patankar1(a),, E.T.Gumbrell2,T.S.Robinson1,H.F.Lowe1, S.Giltrap1, C.J.Price1, N.H.Stuart1,
P.Kemshall2,J.Fyrth2,J. Luis2, J. W.Skidmore2, R.A.Smith1
1Blackett
2AWE,
Laboratory, Imperial College, London, SW7 2AZ, UK
Aldermaston, Reading, Berkshire, RG7 4PR, UK
We report on the design and testing of a multi-wavelength interferometry system for the Orion laser
facility based upon the use of self-path matching Wollaston prisms. The use of UV corrected
achromatic optics allows for both easy alignment with an eye safe light source and small (~
mm)offsets to the focal lengths between different operational wavelengths. Inteferograms are
demonstrated at wavelengths corresponding to first, second and fourth harmonics of a 1054nm
Nd:glass probe beam. Example data confirms the broadband achromatic capability of the imaging
system with operation from the UV (263nm) to visible (527nm) and demonstrates that features as
small as 5µm can be resolved for object sizes of 15mm by 10mm. Results are also shown for an off
harmonic wavelength that will underpin a future capability. The primary optics package is
accommodated inside the footprint of a Ten Inch Manipulator (TIM) to allow the system to be
deployed from a multitude of viewing angles inside the 4m diameter Orion target chamber.
I. INTRODUCTION
The recently commissioned Orion1 laser facility at the UK Atomic Weapons Establishment (AWE),
uses ten ~kJ class long pulse lasers and two independent sub-picosecond petawatt class short pulse
lasers to underpin a wide range of plasma physics experiments. A comprehensive array of diagnostic
capabilities2,3 exist to cover the full spectrum of radiation and particles emitted from intense laser
matter interactions ranging from Hohlraums to shocks. The optical diagnostics suite consisting of a
multi-channel pyrometry system, a Velocity Interferometer System for Any Reflector (VISAR), Gated
Optical Imagers (GOI) and short pulse interferometry provides key experimental tools for studying
plasma properties. This paper describes the optical setup and testing of a robust and flexible
interferometry system devised for use at the Orion laser facility to probe two dimensionalplasma
electron density profiles.
Optical interferometry using short pulse lasers is a commonly applied technique for diagnosing
plasma electron densities in under dense conditions4. Alaser pulse passing through the plasma
acquires a phase shift which is proportional to the density-length product of the free electron density.
When compared to a reference fringe pattern for light which has not traversed the plasma, it is
possible to extract total imparted optical phase and consequently determine a 2D projection of the
electron density-length product. The use of short pulse lasers with sub-picosecond pulse duration also
offers excellent temporal resolution and extends this technique to the study of rapidly evolving
hydrodynamic motion. For a typical 500fs pulse from a Nd:glass laser, the motion blur from
hydrodynamics at 1000km/s would only be 0.5µm which is typically less the optical resolution of the
imaging system.Two optical probe beams on Orionare specified to deliver 400fs 263nm (4ω) pulses
with a nominal energy of 50mJ. Although the system we describe here is designed specifically for
probing at the fourth harmonic, its capability to operate effectively at different wavelengths and
configurations is demonstrated. Additional probing wavelengths are not currently delivered to target
but they are available in the target chamber for the 4th harmonic generation process and can be used
(a) Author to whom correspondences should be address. Electronic Mail - [email protected]
for active imaging in the future. Whilst shorter wavelengths are able to probe denser plasmas (the
critical electron density Ncr= 1.6 × 1022 cm-3 at 263nm), longer wavelengths offer greater sensitivity in
terms of fringe displacement. Interactions of ultrashort laser pulses with solid targets can produce
strong harmonic self-emission from the generated plasma. Experimental evidence from AWE's
HELEN5 laser system for example has indicated that the generated emission results in a weaker selfemission background when probing at the fourth harmonic. Figure 1 shows a 4ω backlit optical image
of a fused silica plate irradiated at ~1020 Wcm-2 from the HELEN shortpulse system and probed
transversely 6.6ps later. It is clear that only the region immediately in the vicinity of the laser focus is
saturated with self-emission and important early time plasma dynamics such as front surface
expansion is clearly visible. This is in stark contrast to the second harmonic image which was
completely saturated by self-emission.
Focal region saturated
Fused
silica
slide
Front surface
expansion
Laser
150 µm
FIG 1:Fourth harmonic transverse shadowgraphy of a short pulse laser irradiation of a fused silica plate at 1020 W cm-2 .
Whilst self-emission in the fourth harmonic band saturates the imaging systemclose to the laser focus, early time plasma
dynamics such as the expansion of the front surface of the target is clearly visible.A reduction in background light from
plasma self-emission is a key advantage of probing with the fourth harmonic.
A range of optical schemes have been realised for plasma interferometry including Mach-Zehnder,
Michelson and Normarski configurations4. All methodologies rely on interference between light
pulses which have traversed the plasma and those which have not. This is achieved either by splitting
the probe pulse and recombining it with a reference copy, or by overlapping spatial regions in the
probe beam which have travelled the plasma with those which have not.
The contrast of the interferogram (the ratio of the intensities of the light and dark zones of the fringe
pattern) is dependent upon the relative intensity of the two beams being interfered and the spatial and
temporal coherence of the two pulses of light. The latter places rigid constraints on the alignment of
optics, particularly for use with short duration probe pulses. For example, the expected ~400fs probe
pulse on Orion would require path length matching to better than 20µm should the pulse be split and
then recombined.
Given the operational and engineering constraints on Orion (e.g. a low shot rate, potentially high
radiation Class 4 laser environment with carefully controlled access), a robust and flexible optical
system based on the use of Wollaston prisms was devised to accommodate a range of laser probing
scenarios. Although interferometry is the primary purpose of the system, shadowgraphy4 and
Schlieren imaging4 modes are also possible to implement with minor modifications to the optical
package. Polarised light interferometry has been demonstrated on laser experiments previously6 and
offers unique advantages from an operational perspective. The lack of a reference beam removes the
need for complex in-vacuo engineering along with the condition to have it path matched to
inteferometric accuracy with the probe. This allows the optical system to be stable and compact
enough to fit inside a TIM.
II. DESCRIPTION OF OPTICAL SETUP
The optical configuration for the interferometry package is based around the use of UV corrected
triplet lenses which greatly simplifies the alignment procedure for the fourth harmonic probing
system. UV pulses are only available at Orion on a single shot basis, and as a result, pre-shot
alignment is performed using an off wavelengthlow power CW 532nm Nd:YAG laser. The need for
additional correcting optics to account for the large focal length offsets between UV and visible is
avoided by using custom manufactured UVachromats which require very minor adjustments across a
broad wavelength range. Additionally, any wavelength from the UV to NIR can be imaged with the
appropriate minor correction to lens positionswhich removes the need to rebuild the system with
different optics when radically changing the mode of operation.
V
U
OP
L
L1 F1
F2
Z
WP
F3
CP
IP1
L2
L3
IP2
FIG2: A schematic of the optical configuration of the Wollaston prism based interferometry system. OP - Object
plane; L1 - Lens 1; WP - Wollaston Prism; CP - Cube Polariser; L2 - Lens 2; L3 - Lens 3; U - Object distance;
V - Image distance; L - Distance from the Wollaston to IP1; F2 - Focal length of lens 2 ;F3 - focal length of
lens 3; IP1 & IP2 - Image planes. Blue lines show image rays forming a double image at IP1 and IP2.
The layout of the optical system is detailed in figure 2. The object plane is imaged using a
180mmfocal length F/8UV triplet (Edmund optics 64-840) which is located at a distance of 216mm
(U in the figure 2) from the object plane. An image of the object plane is formed at a distance of
1080mm (V) behind L1, which gives a linear magnification of 5. For typical laser interaction
experiments, the target is relatively "small" (few mm scale) which demands the use of high
magnification optics to resolve key physical processes. It is also possible to change the magnification
by changing the focal lengths of the telescope between L2 and L3 which relay the image outside the
TIM vacuum chamber. The Wollaston prism is orientated at 45° to the incoming polarisation
(vertical) such that two orthogonally polarised beams of equal intensity are generated. The cube
polariser (CP) is orientated at 45° to the orthogonally polarised beams to ensure the transmitted beams
have the correct polarisation state for optimal fringe contrast. This optic is wavelength dependent and
will have to be changed depending on the wavelength configuration. The two orthogonally polarised
beams diverge at an angle ε, which is set by the geometry of the prism (and is a fixed quantity for a
specific prism set). At the first image plane (IP1), the centres of the two images are spatially separated
by an amount given by the relation εZ, where Z is the distance between the centre of the Wollaston
and the first image plane (IP1).The fringe spacing is given by the relation5 (λ/ε)/(Z/L). Fringes are
visible in the region where the two images overlap. It is crucial that the overlapped region contains
one image of the plasma and one "empty" region to allow unambiguous interpretation of the data.
This necessitates the use of Wollaston prisms with sufficient divergence to separate the images
adequately, and this will be influenced by the precise detail of a given experimental configuration, e.g.
target size and region of interest. Prisms with divergence angles of 0.5o and 2owere tested and shown
to suffice for even quite large laser targets up to scales of 15 mm. Fringe orientation can be chosen
arbitrarily if a half wave plate is used to adjust the probe beam polarisation before passing through the
target.Fringe spacing can be adjusted by changing the position of the Wollaston prism (effectively
adjusting the ratio of Z/L); however, this is limited to a range where the two images are still
sufficiently separated such that unambiguous data can be obtained. For a given image separation,
fringe spacing can also be changed by switching to a different probing wavelength, however, the
advantages of shorter wavelength probing (e.g. less harmonic selfemission) would potentially be lost.
Whilst the magnified image at position IP1 can be used to accommodate a detector, it lies only
1280mm from OrionTarget Chamber Centre (TCC) which means that any active camera would
require vacuum compatibility (and potentially EMP and radiation hardness) to operate inside the
chamber, which would add significantly to the complexity of the system.An additional pair of
achromatic relay lenses (L2 & L3) allows the detector along with UV and optical filters to be
placedon a small platform outside the TIM vacuum chamber and also allows the magnification to be
altered. The lens L2 (200mm F/4 B.Halle7 UV triplet) is positioned at its focal length from IP1 in
order to project the image rays to infinity. These rays are then collected and refocused using lens L3
(300mm F/5 B.Halle UV triplet) placed at its focal length from the front surface of a large format
CCD detector. The combination of L2 and L3 provides an additional magnification of 1.5 from IP1
which increases the total image magnification to 7.5. It is also possible to change magnification
externally to 2.5 by changing L3 to a 100mm F/2 UV quintuplet lens6. This offers additional
flexibility depending on target size and required field of view. It should also be noted that the UV
achromats do not have anti-reflection coatings at 263nm, however, due to the limited temporal width
of the probe pulses, multiple reflections were not found to cause any unwanted secondary fringes
during testing.Figure 3 shows a 3D representation of the TIM based optics package with the
interferometry optics placed in a line along the system axis.
CP
WP
L1
L2
TIM Payload frame
FIG.3: 3D layout of the TIM based optical probing payload frame along with the key optics which are
labelled as per figure 1. The first lens also hosts a removable debris shield and a ring light to aid focusing.
In addition to interferometry, the optical package can easily be modified for Schlieren imaging by
placing a stop at the probe ray focus of L1 and removing the Wollaston prism. Shadowgraphy is also
possible to implement by removing both the Wollaston prism and Schlieren stop. Modified
commercial digital single lens reflex (DSLR) cameras are used as time integrated detectors for the
probe beam light.High temporal resolution is provided by the limited duration of the probe pulse and
its greater intensity compared to self-emission. A Canon 350D DSLR with a 24mm by 16mm chip
size and 6.4µm pixel size was used for detecting the UV light. A large Nikon D810 detector with a
36mm by 24mm chip size and 4.8µm pixel size is also available for use in the visible and IR. If selfemission from the initiallaser-target interaction is relatively high, then a time gated imaging system
(GOI) can be used as the detector, albeit with significantly lower spatial resolution and dynamic
range.
The availability of two independent probe beams on Orion along with the flexibility of multiple
wavelengths and imaging configurations offers a compelling capability for a wide range of
experiments. Advanced probing scenarios such as two colour inteferometry8 and Faraday
rotation9imaging of magnetic fields are key future considerations included in the design of this
system.
III. Active System Testing
Optical testing and metrology of the complete interferometry and imaging system was carried out at
Imperial College London due to the limited availability of access toOrion. An OPCPA Nd:Glass
hybrid laser system10 with equivalent specifications to the Orion probe beam was used for
interferometry and a 1 TW (500mJ, 500fs, 1054nm) drive laser was used to create test laser plasmas
that acted as a short lived, rapidly evolving phase object. The probe beams were generated using two
cascaded KDP SHG crystals to provide co-propagating 1ω, 2ω and 4ω pulses which were separated
using appropriate filters and dichroic mirrors. Although the plasma conditions probed were not
comparable with those created by the very high energy drive lasers at Orion, the aim was to verify the
fidelity and flexibility of the probing system under a range of dynamic scenarios.
Test interferograms obtained using the system described above along with the 0.5oWollaston prism
are shown below in figure 3. The images have been cropped to the centre of the field of view.
Lineouts showing the fringe contrast are also attached for different test wavelengths. Table I shows
the maximum fringe contrast ratio obtained at each wavelength. The contrast obtained at 1ω and 2ω
are very good with very clear transitions from bright to dark zones, the 4ω fringes are a factor of two
lower in contrast as difficulties were encountered in driving the nonlinear conversion process used to
create the DUV probe light into saturation. This resulted in a modulated spatial profile which lead to
unequal intensities in the two arms of the interferometer system and consequently, reduced fringe
contrast. This issue was isolated solely to the quality of the available probe beam and is not a
fundamental limitation of the optical components. Fringe patterns are shown in three typically used
configurations although it should be noted that the orientation can be changed arbitrarily by using half
wave plates.
FIG 3. Typical interferograms obtained at centre of the image for 1ω (a), 2ω (b) and 4ω (c) probe pulses. Corresponding
lineouts are taken in the direction indicated by the dashed white line. Excellent fringe contrast is obtained at the first and
second harmonic owing to better beam quality. We noticed internal defects inside UV interference filters which resulted in a
localised reduction in fringe contrast. Difficulties were also encountered in obtaining a smooth UV beam profile which
resulted in non-equal intensities in the interferometer arms. These issues are not attributed as an optical fault with the
imaging system.
Table I: Measured fringe contrast at image centre on a typical interferogram for the first, second and fourth harmonic probe
beams.
Wavelength
1054nm
527nm
263nm
Peak Fringe Contrast
58.9%
71.8%
35%
A. Experimental Examples
Experimental validation was performed at both the second and fourth harmonic using two different
scale length targets with the low (0.5°) and high(2°) divergence Wollaston prisms. Detailed
interpretations of the experiments are notthe subject of this paper and will be addressed in a separate
publication. The probing direction was set orthogonal to the drive beam. For the UV probing test, a
laser interaction with a small scale solid density target was performed.A 50µm diameter silicon oil
droplet11 attached to 9µm diameter Carbon fibre was irradiated with 150GW of power using F/20
focusing. Figure 4a shows the pre-shot UV interferogram of the micro-droplet attached to a stalk.
Given the small scale length of the target, only a modest image separation is required to isolate the
two images while ensuring a uniform field of fringes. The imaging configuration was setup as per
figure 2 with the 0.5° Wollaston placed at a distance of 850mm from IP1 (L = 50mm). This gives an
image separation of ~ 7.4mm which even at a magnification of 5 gives enough isolation (IS ~ 1.3mm)
between the two images. The resultant fringe spacing is ~ 2µm which is well resolved by the detector
(Canon 350D). It should be noted that the relatively poor fringe contrast seen in the images was due to
inhomogenities in the interference filter and non-uniformities in the spatial profile over the field of
view. Figure 4b shows the deflection in the fringe pattern 3ns after the drive laser hitting the oil
droplet. It should be noted that the UV images are not exactly in focus as adjustments were not made
from the 532nm alignment laser. This highlights the achromatic capability of the imaging system as
useable data can be obtained without a UV alignment beam.
A
B
Drive
Laser
9 µm
FIG4: 263nm short pulse interferometry of an oil droplet attached to a 9µm diameter carbon wire. A)
Background interferogram showing a 50µm oil droplet attached to the wire. B) Fringe pattern deflection
observed 3ns after irradiation with a 100mJ 650fs laser pulse. The relatively low fringe contrast is due to
difficulties encountered in uniform UV illumination and not an optical defect. The bright spot in the
corner of the second image is scattered light from the drive laser.
Low density gaseous targets are transparent at optical wavelengths and therefore ideally suited for
testing a number of probing scenarios. Noble gas clusters from pulsed gas jets have been shown to be
very efficient at absorbing short pulse laser light12 and make interesting targets for experiments in
areas including laboratory astrophysics13 and x-ray generation14.To verify the imaging system
capabilities at the second harmonic wavelength, argon gas clusters from a large aperture (12mm) gas
nozzle were irradiated with 1 TW pulses using F/20 focusing. The measured average atomic density
of the cluster gas produced by this system is ~ 1019 cm-3and the measured mean cluster size is ~5nm.
The interaction was imaged perpendicular to the drive laser using simultaneous interferometry and
Schlieren imaging.Given the size of the gas jet, the imaging system was modified from the
configuration shown in figure 2 by changing L1 to 300mm F/5 UV triplet and the Wollaston
divergence to 2°. The magnification was lowered to 2 and a larger area detector (Nikon D810) was
used to increase the field of view for the interferometry. The additional image relay shown in the
design was also not used as it is specific for TIM compatibility and consequently images were taken at
the IP1 position. A 4" beamsplitter was used after L1 to split the probe beam for performing Schlieren
imaging independently. A 500µm wire was placed at the focus of L1 to block the probe rays and
perform dark field Schlieren. To generate sufficient image separation for the long plasma, the 2°
Wollaston was placed 55cm from the detector giving a fringe separation of ~66µm and image
separation greater than 1 cm. The finite size of the probe pulse limits the region with fringes to
approximately half the interferogram and accounting for optical apertures of the probe beam in Orion,
sets a limit of ~15mm on the region that can probed.
FIG 5: Schlieren image of Argon clusters (50 bar backing pressure density, >1019 cm-3, cluster size ~ 5nm) irradiated with
1TW 1054nm pulse and F/20 focusing. Transverse probing was performed using 2ω 500fs pulses 16ns after the laser-cluster
interaction. Here a well-defined thin -shelled blast wave is launched into the surrounding gas, and is preceded by a radiative
precursor
FIG 6 : Interferogram taken under identical conditions (but different shot) as the Schlieren image in the previous figure
along with an inset figure from the white box. Left hand side image shows the large field of view afforded by a 36mm by
24mm detector. This allows the entire 12mm bore gas jet to be imaged. Right Close-up of the boxed region shows high
resolution interferometricdata can be obtained at large field of view.
Figure 5 shows a dark field Schlieren image of a blastwave formed 16ns after the laser-cluster
interaction. The faint halo surrounding the interaction region is due to a radiative precursor, where
plasma self-emission at soft x-ray wavelengths (XUV) ionises the cluster medium, creating an
electron density gradient that is visualised by the Schlieren technique15. Figure 6 shows the
interferogram obtained under identical conditions, this can be unfolded using an Abel inversion
technique to extract the plasma electron density4. It is worth highlighting the large field of view
possible with a 36mm by 24mm detector which can encompass large targets such as a 12 mm bore gas
jet. The close up image inset demonstrates that fine details in the turbulent blast wave region can be
well resolved.
B. OFF HARMONIC CAPABILITY
In short pulse driven experiments (currently 1054nm and 527nm drive wavelengths are available at
Orion), harmonic emission and optical transition radiation from the laser-target interaction can be a
powerful source of background light in addition to plasma self-emission. Since the Orion probe beam
is also a harmonic of the drive beam, it is not possible to filter out this light for discrimination
purposes using bandpass or interference filters. This makes the currentscheme for the short
pulseimaging probing system using open shutter cameras potentially incompatible with short pulse
drive experiments that create very large self-emission backgrounds. Two solutions exist for this
problem, one of which would be to use a gated detector (such as GOIs) which would allow the initial
bright background from the interaction to be filtered in the time, though at the expense of spatial
resolution and dynamic range. The other option would be use an off-harmonic wavelength for
probing, in which case the current methodology will be applicable with the simple addition of filters
to remove harmonic wavelengths. Whilst Orion has access to GOIs, they are significantly lower
resolution devices than CCD camerasand have a smaller field of view. The use of Raman-shifted
probe16 beams at 630nm has been demonstrated as a method to observe small scale structure close to
irradiated targets and would be a useful probing capability for Orion. Imaging performance at 630nm
was assessedby selecting a narrow spectrum of a laser driven supercontinuum17source using
interference filters. Fringes obtained under an identical setup to that used in figure 3are shown in
figure 7.Whilst the fringe quality is limited by the spatial coherence of the generated supercontinuum
light, the preliminary results highlight the possibility of using this imaging system over a very wide
wavelength range.
FIG. 7: Fringes obtained using a supercontinuum source which is spectrally apertured at 630nm. Whilst fringe quality is
limited by the spatial coherence and phase aberrations of the broadband light, the possibility of using Raman shifted probe
wavelengths is demonstrated.
IV. Conclusion
We report on the development and testing of a multi-wavelength TIM compatible imaging system for
short pulse probing on the ORION laser facility. Multi-harmonic capability together with operation
over a wide wavelength range without the need for extensive system realignment is achieved using
custom manufactured highly corrected UV-triplets. Interferometry is achieved using a self-path
matched configuration based on Wollaston prisms. This allows for a compact linear optical
configuration which is easy to align and fits within a restricted engineering package suitable for use in
a demanding experimental environment. Data is presented which highlights promising performance at
second and fourth harmonic wavelengths of anNd:glass probe beam. Preliminary measurements also
highlights the possibility of using Raman shifted probe wavelengths to perform off-harmonic
interferometry.
ACKNOWLEDGEMENTS
This work was performed for and funded by the Atomic Weapons Establishment (AWE),
Aldermaston, UK. N.H.Stuart, T.S.Robinson, S.Giltrap and H.F.Lowe are funded by AWE/EPSRC
CASE Studentships. S. Patankar is funded by EPSRC KTS fellowship RSRO P43480.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Hopps, Nicholas, et al. Plasma Physics and Controlled Fusion 57.6 (2015): 064002.
Bentley, C. D., et al. Review of Scientific Instruments 83.10 (2012): 10D732.
Oades, Kevin, et al. Review of scientific instruments 75.10 (2004): 4222-4224.
Hutchinson, Ian H. Principles of plasma diagnostics. Cambridge university press, 2005.
Norman, Michael J., et al. .Applied optics 41.18 (2002): 3497-3505.
Benattar, R., C. Popovics, and R. Sigel.Review of Scientific Instruments 50.12 (1979): 15831586.
Custom UV optics manufacturer, Bernhard Halle Nachfl. GmbH, Hubertusstraße, 10
D - 12163 Berlin, Germany
Moosman, B., et al. " Review of scientific instruments 70.1 (1999): 672-676.
Stamper, J. A., et al,.Physical Review Letters, 40(18), 1177.
Bigourd, D., et al. Applied Physics B 113.4 (2013): 627-633.
Kurt. J. Lesker 704 vacuum pump oil was used for its low vapour pressure.
Ditmire, T., et al. Physical review letters 78.16 (1997): 3121.
Moore, A. S., Symes, D. R., & Smith, R. A. (2005). Physics of plasmas, 12(5), 52707-52707.
Ditmire, T., et al. " Applied physics letters 71.2 (1997): 166-168.
Ditmire, T., et al. The Astrophysical Journal Supplement Series 127.2 (2000): 299.
RS Adrain et al, Optics Communications, 15 (2), pages 290-292, (1975 )
Alfano, Robert R. "The supercontinuum laser source." (1989): 458.