Monomode filtering and High Angular Resolution

Transcription

Monomode filtering and High Angular Resolution
Sylvestre Lacour – Observatoire de Paris
Between astronomy and instrumentation: career
ASTROPHYSICS
50%
•  Interstellar medium (FUSE)
INSTRUMENTATION 50%
•  FUSE Science Data Processing Team (Engineer at JHU)
•  Evolved stars (Imaging with IOTA)
•  Development of Fibered interferometers (PhD in Paris)
•  Faint companion detection around young stellar objects (Sydney and Paris)
•  Implementation of aperture masking on NaCo (Post doc at Sydney)
55 rank A papers
1765 citations (H-­‐‑25)
Project LITHIUM
2
Between astronomy and instrumentation: project
ASTROPHYSICS
WP1:
Toward a beXer understanding of planetary formation in transition disks
Project LITHIUM
50%
INSTRUMENTATION
WP2:
Increase dynamic range of classical aperture masking techniques
50%
WP3:
Bring aperture masking to new levels by combining it to nulling
3
Aperture Masking
PALOMAR
KECK
GEMINI SUBARU
VLT LBT JWST E-­‐‑ELT
Pupil wheel on SPHERE
Pupil images from NACO Project LITHIUM
4
Basic Principle
PSF
FT
CP
Phases
Disk model by Olofsson et al. 2013
Project LITHIUM
Companion model by Huelamo, Lacour et al. 2011
5
The capabilities of aperture masking
HD142527B
1,0E+00
1,0E-­‐‑01
T Cha (b?)
HD142527B
Flux Ratio
1,0E-­‐‑02
LkCa15 (b?)
Beta pic b
1,0E-­‐‑03
HR 8799 b
HR 8799 c
1,0E-­‐‑04
HR 8799 d
1,0E-­‐‑05
1,0E-­‐‑06
10
HR 8799 e
100
1000
Separation [mas]
Project LITHIUM
6
The capabilities of aperture masking
1,0E+00
1,0E-­‐‑01
Long
Baseline
Interferometry
Flux Ratio
1,0E-­‐‑02
1,0E-­‐‑03
1,0E-­‐‑04
1,0E-­‐‑05
1,0E-­‐‑06
10
Aperture Masking
(actual)
ADI
SDI
Coronagraphic
Techniques
(From SPHERE user manual)
100
1000
T Cha (b?)
HD142527B
LkCa15 (b?)
Beta pic b
HR 8799 b
HR 8799 c
HR 8799 d
HR 8799 e
Separation [mas]
Project LITHIUM
7
How are planets formed?
•  Core accretion ?
Or
•  Gravitational instabilities ?
Br detection limits obtained with RDI are alsoOr
indicated along with 2 Myr stellar
tracks using the BCAH98 model47 .
•  A mixture of both ?
gap-crossing
streams
Figure 1 ALMA observations of HD 142527, with a horseshoe dust continuum surrounding a cavity that still contains gas. We see diffuse CO gas in Keplerian rotation
(coded in doppler-shifted colours), and filamentary emission in HCO+ , with nonKeplerian flows near the star (comparison models illustrative of Keplerian rotation
are shown in SI). The near-IR emission abuts onto the inner rim of the horseshoe-
HD142527, ALMA data
Project LITHIUM
10
Simulation of accreting planets
from Casassus et al. Nature 2013
8
HD142527
The Astrophysical Journal Letters, 753:L38 (5pp), 2012 July 10
Biller et
0.570 Msun
2.5 0.500 Msun
1.5•105
0.300 Msun
3.5 0.250 Msun
⇒ 
4.0 Accreting planet ?
or
4.5
⇒ 
5.0 Stellar companion?
5.0•10
0.200 Msun
0.175 Msun
Mh (mag)
5
3.0 0.350 Msun
N
1.0•10
0.450 Msun
0.400 Msun
H
K
L
4
0.150 Msun
0.130 Msun
0.110 Msun
0.100 Msun
5.5
0
0.0
6.0
0.2
0.4
0.6
0.8
1.0
Mass (Solar Masses)
1.2
1.4
0.000
0.005
age (Gyr)
0.010
0.015
Figure 3. Left: mass estimate histograms for HD 142527B. We adopt Monte Carlo methods to account for the range of possible ages for this system. An ensem
of 106 possible ages are drawn from a Gaussian in log space, centered on log(age) = 6.7 and with σ (log(age)) = 0.4. We then interpolate with age and single-b
absolute magnitude to find the best mass for the companion from the models of Baraffe et al. (1998). We plot here the resulting mass distributions from single-b
absolute magnitudes in H, K, and L! . It is instantly apparent that the mass estimate for the companion is not well constrained at these ages but is most likely to li
the range from 0.1 to 0.4 M" . Right: age vs. absolute magnitude in the H band. We plot isomass contours (from the models of Baraffe et al. 1998) as a function
age and absolute magnitude, as well as the assumed age and absolute magnitude of the companion. The models themselves are highly dependent on age; the reg
constrained by the observations (yellow rectangle) is consistent with companion masses from 0.1 to 0.4 M" .
(A color version of this figure is available in the online journal.)
Discovery of a companion with aperture masking by Biller, Lacour, et al. 2012
Age: 2-­‐‑
5Myr
estimate for each band of 0.28 ± 0.15 M" , 0.34 ± 0.19 M" , a
0.60 ± 0.29 M" in H, K, and L! respectively. However,
Temp: 3000K
mass distributions from our Monte Carlo simulations are hig
Radius: 1.2 Rsun
non-Gaussian, with significant probability to find considera
higher companion masses. All mass estimates are within 2σ
Accretion: 1.7Msun/yr
each other, but the L! -band mass estimate is particularly high a
we note that the companion appears anomalously bright in
-­‐‑-­‐‑-­‐‑-­‐‑-­‐‑-­‐‑
While the H − K colors are similar to what would be expec
for
a young red companion, K − L! ∼ 0.9 mag, considera
in agreement with a divergent from the expected value of 0.4 mag. We thus do n
attempt to estimate spectral type using H − K and K −L! colo
M=0.13Msun from The models of Siess et al. (2000) yield a similar mass range
evolutionary models the companion of 0.1–0.4 M" for ages of 2–12 Myr. T
models themselves are highly dependent on age; to illustr
(Baraffe et al. 1998)
this, we plot isomass contours (from the models of Bara
Project LITHIUM
from the 2MASS survey. Within a 1 deg radius of the primary,
2MASS detects 1918 objects with H of 10.7 mag or brighter
and 1505 objects with Ks of 10.0 mag or brighter. Thus, adopting the approach of Brandner et al. (2000), in particular their
Equation (1), we estimate the probability of finding an unrelated
source at least as bright as the observed companion within 0.!! 088
of the primary to be ∼1.1 × 10−6 in the H band and ∼8.3 × 10−7
in Ks. We also considered simulated stellar populations along
this line of sight (Galactic latitude and longitude of 335.◦ 6549,
+08.◦ 4804) using the Besançon Galactic population synthesis
models (Robin et al. 2003). This line of sight is directly into the
Galactic bulge, so the models yield 882 background sources per
square degree that are brighter than K = 10.5 mag. However,
the chances of finding one of these within 0.!! 088 of the primary
are still vanishingly small—∼1.6×10−6 —and these objects are
predominantly M giant stars, with considerably bluer expected
colors (for an M5III star, H − K = 0.29 mag and K − L! =
0.22 mag; Tokunaga 2000) than measured for the detected companion. It is therefore extraordinarily unlikely that the companion is unrelated to the primary, although proper motion confirmation in a year will be necessary to finally determine this.
et al. 1998) as a function of absolute H magnitude and age
Figure 3.
Orbital motion and SED (Lacour, Biller, et al., in preparation)
4.3. Mass Estimate
9 on the Orbit
4.4. Constraints
We estimate the semimajor axis of HD 142527B’s or
from its observed separation. Assuming a uniform eccentric
distribution between 0 < e < 1 and random viewing angl
Dupuy et al. (2010) compute a median correction factor betwe
Use and increase the capabilities of aperture masking
1,0E+00
1,0E-­‐‑01
Long
Baseline
Interferometry
Flux Ratio
1,0E-­‐‑02
Aperture Masking
WP1:
(actual)
Understanding the physics of planet formation
Imaging
(From SPHERE user manual)
1,0E-­‐‑03
1,0E-­‐‑04
1,0E-­‐‑05
1,0E-­‐‑06
10
T Cha (b?)
100
HD142527B
LkCa15 (b?)
Beta pic b
HR 8799 b
HR 8799 c
HR 8799 d
HR 8799 e
1000
Separation [mas]
Project LITHIUM
10
Dynamic range limitation: closure phase accuracy
The Astrophysical Journal Letters, 753:L38 (5pp), 2012 July 10
Biller et al.
400
1.5
(a)
(b)
(c)
(d)
(e)
(f)
1.0
0.5
0.0
200
0
−1.0
Closure Phase (deg)
dec [mas]
v (meters)
5
−0.5
0
∆mag=5.2
PA=133deg
Sep=88 mas
1.0
0.5
0.0
−0.5
−1.0
−1.5
1.5
1.0
−200
−5
−1.5
1.5
0.5
0.0
−0.5
−1 deg
−5
−0.5 deg
0.5 deg
0
5
1 deg
−1.0
−400
400
−1.5
200
u (meters)
0
RA [mas]
−200
−400
−45
−40
−35
−30
−25
−20
−45
−40
−35
−30
−25
−20
Parallactic angle (deg)
Figure 1. Left panel: L! -band UV coverage on HD 142527. The size and colors of the markers are relative to the phase measured. The larger the size, the higher
the value of the phase. The colors denote the sign of the phase (red are negatives values). The plot shows diagonal stripes orthogonal to the direction of the binary
companion (indicated by the arrow). Middle panel: L! -band χ 2 surface as a function of R.A. and decl. obtained from the best-fit binary model to the closure phases.
A clear minimum indicates the position of the stellar companion, coherent with the orientation of the stripes in the Fourier domain. The red contours correspond to
3σ and 5σ error bars in the detection. Right panel: L! -band closure phase as a function of parallactic angle for the six largest three-hole triangles. Calibrator data are
represented as colored squares (red HD 142695, green HD 144350, and blue HD 142384), while HD 142527 data are plotted as black squares. The solid line is the
closure phase predicted by the best-fitting binary system model.
(A color version of this figure is available in the online journal.)
HD142527 Dataset (Biller et al. 2012)
10−7Project LITHIUM
Primary + Disk
Companion
10−8
0.1–0.3 mag, which may suggest variability for this system.
11 No
errors are provided for the Malfait et al. (1998) photometry; we
assume error bars are similar to the 2MASS photometry.
Raw photometry for this system is comprised of light from
Dynamic range limitation: closure phase accuracy
1) Systematic errors:
Bias on the phase limits the closure phase accuracy to several tenths of a degree (typical 0.2 degree)
Project LITHIUM
12
Use and increase the capabilities of aperture masking
1,0E+00
1,0E-­‐‑01
Long
Baseline
Interferometry
Flux Ratio
1,0E-­‐‑02
1,0E-­‐‑03
1,0E-­‐‑04
1,0E-­‐‑05
1,0E-­‐‑06
10
Aperture Masking
WP1:
(actual)
T Cha (b?)
Understanding the physics of planet formation
WP2:
Understanding the Imaging
(From SPHERE user manual)
systematics
100
HD142527B
LkCa15 (b?)
Beta pic b
HR 8799 b
HR 8799 c
HR 8799 d
HR 8799 e
1000
Separation [mas]
Project LITHIUM
13
Dynamic range limitation: closure phase accuracy
1) Systematic errors:
Bias on the phase limits the closure phase accuracy to several tenths of a degree (typical 0.2 degree)
2) Photon noise:
The flux of the companion is mixed with the flux of the star: the photon noise limits the detectability
Project LITHIUM
14
Use and increase the capabilities of aperture masking
1,0E+00
1,0E-­‐‑01
Long
Baseline
Interferometry
Flux Ratio
1,0E-­‐‑02
1,0E-­‐‑03
1,0E-­‐‑04
1,0E-­‐‑05
1,0E-­‐‑06
10
Aperture Masking
WP1:
(actual)
T Cha (b?)
Understanding the physics of planet formation
WP2:
Understanding the Imaging
(From SPHERE user manual)
systematics
WP3:
Overcoming photon noise
100
HD142527B
LkCa15 (b?)
Beta pic b
HR 8799 b
HR 8799 c
HR 8799 d
HR 8799 e
1000
Separation [mas]
Project LITHIUM
15
LiNbO3
2.3.
EFFET ÉLECTRO-OPTIQUE : VERS DES INTERFÉROMÈTRES INTÉGRÉS C
Lithium Niobate: Summary of Physical Properties and Crystal Structure
mirror
plane
T
1,9
+
1.9
J
O Niobium
O Lithium
Fig. 3. Three mirror plane symmetry elements associated with
+ a 1s
lithium
niobate (3m
class) On voit les trois plans de
Figure 2.7 – Structure cristalline
du Niobate
decrystal
Lithium.
Crystalline structure of LiNbO3, from Weiss et al. 1985
axis
(classe 3m). Source : Weiss et al. 1985 [59]
Fig. 4
Project LITHIUM
L’effet électro-optique
16 of the Li
qualitatively by considering the movement
and Nb ions relative to the oxygen octahedra. The
undisturbed positions of the Li and Nb ions relative to
the oxygen octahedra at room temperature are de-
hexa
The a
both
indic
guide, mais cet aspect de mon travail sera détaillé plus loin.
Notes sur les pertes
LiNbO3
Dans un guide d’onde diélectrique, les pertes sont théoriquement essentiellement dues aux pertes
du matériau. Dans le cas du Niobate de Lithium, la transmission est de l’ordre de 75% après 0.6 cm
de propagation autour de 3 µm [50]. Nous verrons que les pertes mesurées pour les guides diffusion
de titane en bande L sont bien supérieures aux pertes du matériau. En effet, en diffusant le titane
à haute température, on altère la structure cristalline du matériau. En outre, des observations au
98
CHAPITRE - 3. RÉALISATION ET CARACTÉRISATION
DE GUIDES
L présence de grains dans les guides. Ces grains diffusant peuvent être à
microscope
nousEN
onBANDE
révélé la
l’origine des pertes élevées que présentent nos guides. La diffusion de titane a permis de réaliser des
présentant des pertes de l’ordre de 0.3 dB/cm @ 632.8 nm [51]. Avec l’échange protonique,
3.3 Réalisation technologique de guides d’onde guides
RALIS
des guides APE (Annealed Proton Exchange) avec des pertes comprises entre 0.2 et 0.4 dB/cm @
La fabrication de guides d’onde, d’un point de vue technologique, est
uneµm
science
à part
entière.
1.55
ont été
obtenus
[52]. Les technologies diffusion de titane et échange protonique peuvent
Ce n’était pas initialement vu comme le cœur du sujet de ma thèse. donc
J’ai néanmoins
été
amené
permettre un guidageà avec une bonne transmission. Il s’agit de procéder aux ajustements
réaliser les guides en salle blanche et donc à me familiariser avec lestechnologiques
techniques lithographiques
pour y parvenir en bande L.
de réalisation de guides. C’est pourquoi je détaillerai dans cette partie les étapes technologiques
qui nous ont permis de réaliser nos composants. Sur la Figure 3.23 est schématisé le résultat
final des étapes de réalisation : sur un substrat de Niobate de Lithium est diffusé du titane par
chauffage, constituant le guide d’onde, et de part et d’autre du guide sont déposées les électrodes de
contrôle de phase. Pour augmenter l’efficacité électro-optique d’un facteur deux, les électrodes ont
été placées en configuration push-pull entre les entrées du composant comme le montre la Figure
3.24 représentant une jonction Y.
•  Transparent from 400nm
to 5μm
•  Electro-optic (variable
refractive index as a
function of electric field)
LiNbO3
Ti:LiNbO3
X
Y
+
-
Z
Figure 2.4 – Transmission du Niobate
de Lithium après 0.6 cm de propagation en fonction de la longueur
Transmission LiNbO3, from Nassau et al. 1966 d’onde. Source : Nassau et al. [50]
Figure 3.23 – Configuration d’un guide T i : LiN bO3 X-Cut. En couleur,
les électrodes
de contrôle de
de filtrage
Couplage
et longueur
phase (sans leurs connections).
Project LITHIUM
LiNbO
17 lumineuse vers les
Lorsqu’on injecte de la lumière dans un guide d’onde, on couple de la puissance
modes guidés. La quantité de lumière couplée à chaque mode dépend de l’intégrale de recouvrement
entre le spot d’injection et le profil du mode guidé concerné. Le profil de chaque mode guidé dépend
Ti:LiNbO
Double Mach-­‐‑Zehnder for nulling
•  And it routinely work:
o  Telecom applications
o  Spatial applications
> 40 dB!
! MXPE-LN-10 high extinction ratio modulator
Project LITHIUM
18
Closure phase and nulling
Phase and Amplitude control
Nuller stage
Closure phase stage
Project LITHIUM
19
WP3: Overcoming photon noise
Figure II-36 : déphaseur simple. Le déphasage décroît avec la longueur d’onde.
Il est possible de « compenser » cette dépendance en longueur d’onde en concaténant
plusieurs segments de guides, chacun de largeur différente :
Improve transmission
Test Y Junction
Improve Y junction
2016
Guide 1
Guide 2
First realisation (curves and straight lines)
Test Mach-­‐‑
Zehnder
L1
1
1
neff
2
2
2
neff1
neff2
WP 3.1
Figure II-37 : déphaseur compensé en longueur d’onde
First double Mach Zehnder
82
Achromatized Pierre Labeye : « Composants
First design 3 optiques intégrés pour l’interférométrie astronomique », 2008.
double Mach-­‐‑
sub-­‐‑apertures
Zehnder
WP 3.2
1
6
En notant ∆ni la différence d’indice
effectif sur le segment « i » due à la différence de
Improve Mach-­‐‑
Zehnder
First Achromatized Mach-­‐‑Zendher
largeur des guides, on obtient
un
ϕ:
Nullerdéphasage
stage
G
G
1
6
1/2
G2
2
3
First design 4 sub-­‐‑aperture
Prototype design VLTI planet finder?
G3
3/4
5
G4
4
Achromatized 3 sub-­‐‑apertures
First design 6 sub-­‐‑apertures
Prototype design planet finder CHARA?
Prototype instrument already on SCExAO
C(3
34)
Contact with ESO to build a Project LITHIUM
planet finder for the VLTI
Improve double Mach-­‐‑
Zehnder
G5
4
2
Prototype Mach Zehnder λ = 3.8µμm
Improve Achromatized Mach-­‐‑Zehnder
5/6
C(12 56)
Closure
phase stage
C(1
Prototype Mach-­‐‑Zehnder λ = 600nm
2017
L2
1
neff
56)
2015
Prototype Nuller Nanosat?
First design 18 sub-­‐‑aperture
Prototype SUBARU?
Technology ideal for 20
space observations
Secret level
Project LITHIUM
21
Use and increase the capabilities of aperture masking
1,0E+00
1,0E-­‐‑01
Long
Baseline
Interferometry
Flux Ratio
1,0E-­‐‑02
1,0E-­‐‑03
1,0E-­‐‑04
1,0E-­‐‑05
1,0E-­‐‑06
10
Project LITHIUM
Aperture Masking
WP1:
(actual)
T Cha (b?)
Understanding the physics of planet formation
WP2:
Understanding the Imaging
(From SPHERE user manual)
systematics
WP3:
Overcoming photon noise
100
Separation [mas]
WP4:
Going to space
HD142527B
LkCa15 (b?)
Beta pic b
HR 8799 b
HR 8799 c
HR 8799 d
HR 8799 e
1000
22
WP4: Going to space
2015
First realisation (curves and straight lines)
Improve transmission
Test Y Junction
Improve Y junction
Improve Mach-­‐‑
Zehnder
First Achromatized Mach-­‐‑Zendher
Prototype Mach-­‐‑Zehnder λ = 600nm
2017
Test Mach-­‐‑
Zehnder
Prototype Mach Zehnder λ = 3.8µμm
Improve Achromatized Mach-­‐‑Zehnder
First double Mach Zehnder
Improve double Mach-­‐‑
Zehnder
Achromatized double Mach-­‐‑
Zehnder
First design 4 sub-­‐‑aperture
Prototype design VLTI planet finder?
Achromatized 3 sub-­‐‑apertures
First design 6 sub-­‐‑apertures
Prototype design planet finder CHARA?
First design 3 sub-­‐‑apertures
Prototype Nuller Nanosat?
First design 18 sub-­‐‑aperture
Prototype SUBARU?
2019
Project LITHIUM
Space interferometer
prototype
23
WP4: Going to space
2015
First realisation (curves and straight lines)
Improve transmission
Test Y Junction
Improve Y junction
Improve Mach-­‐‑
Zehnder
First Achromatized Mach-­‐‑Zendher
Prototype Mach-­‐‑Zehnder λ = 600nm
2017
Test Mach-­‐‑
Zehnder
Prototype Mach Zehnder λ = 3.8µμm
Improve Achromatized Mach-­‐‑Zehnder
First double Mach Zehnder
Improve double Mach-­‐‑
Zehnder
Achromatized double Mach-­‐‑
Zehnder
First design 4 sub-­‐‑aperture
Prototype design VLTI planet finder?
Achromatized 3 sub-­‐‑apertures
First design 6 sub-­‐‑apertures
Prototype design planet finder CHARA?
First design 3 sub-­‐‑apertures
Prototype Nuller Nanosat?
First design 18 sub-­‐‑aperture
Beta Pictoris b transit
Prototype SUBARU?
2019
Project LITHIUM
Space interferometer
prototype
24
WP4: Going to space
- 
- 
- 
- 
- 
- 
Beta Pictoris b radius = 1.5Rjup
Beta Pictoris radius 1.8Rsun
Expected contrast ratio: 1/100
Semi-major axis: 8AU
Transit duration: a few hours
Observed November 1981:
•  Why Beta Pictoris b:
•  Photometry is a first step toward interferometry
•  The planet is unique in the context of planetary formation (Nature paper 100% guaranteed)
Space interferometer
Project LITHIUM
prototype
•  Why not from the ground?
•  We do not know when the transit will happen (between April 2017 and Feb 2018)
•  So we want a dedicated observatory, able to observe 24/24h, 7/7d for 6 month.
25
I. Mission profile
Visibilité station sol : station Paris-Meudon:
Simulation des temps de passage sur 1 jour
(VTS):
• 
• 
Tableau des coordonnées de la station
Latitude
48°48’
Longitude
Altitude
2°14’
162m
Image représentant les deux orbites en 2D
Pour l’orbite héliosynchrone:
o  Temps de visibilité maximal par jour: 13 min 83’’
o  Temps de visibilité minimal par jour: 2 min 33’’
o  Nombre de passages par jour: 7
• 
Pour l’orbite terrestre basse:
o  Temps de visibilité maximal par jour: 9 min
o  Temps de visibilité minimal par jour: 4 min
o  Nombre de passages par jour: 5
Conclusion: Temps de visibilité plus long pour
l’orbite SSO => quantité de données reçues
importantes : 1 Mo pour les visibilités maximales
pour un transfert de 9600 bits/s.
Travail effectué par S. Arroub
26
II. Scientific payload
•  Nécessité d’avoir une précision de pointage de 0,01°.
•  Systèmes de contrôle d’aXitude et d’orbite existant pointent avec une précision de 0,1°.
Solutions apportées:
•  Fixer les fibres optiques sur des systèmes piézos à trois axes.
•  Utiliser un capteur de position permeXant la localisation précise de l’étoile.
Pièzo
Photons
provenant
de l’étoile
pointée
Capteur de position Carte de contrôle
Microprocesseur
27
III. Platform (from ISIS)
Logiciel utilisé:
-  IDM CIC, extension d’Excel développée par le CNES
-  Sketch up, logiciel de modélisation 3D
Plusieurs vues du nanosatellite agencé:
Tableau de masse totale du nanosatellite:
Masse Masse avec [kg] marge [kg] Plateforme 2,6535
3,1842
1,04
1,248
3,6935
4,4322
Charge utile Masse nanosatellite Extrait du tableau rempli sur IDM CIC:
Composant
X
Structure
Externe
0
ADACS
0
Batterie
0
Panneau
Face 1
0
Y
Z
Rx Ry 0
0
Rz
Rx 80 100 Ry Rz
Rx Ry 0
60
Rz
Rx 8,75 -2,15 Ry Rz
R1
R2
R3
Masse
[Kg]
Masse
avec
marge
0
0
0
0,6
0,72
90
360
0
0,907
1,0884
0
0
0
0,1995
0,2394
0
0
0
0,059
0,0708
Vue de dessous
Vue de dessus
Conclusion:
Masse dépasse légérement 4 kg =>
-  Changer quelques composants
-  Augmenter volume CubeSat
28
Vue de face
IV. Launch (we need the CNES)
C’mon CNES… be usefull
Project LITHIUM
29