MARX

Miroirs Actifs Rayons X

pour micro et nanofocalisation

X-ray Active Optics for micro and

nanofocalisation

Goal of the MARX Project :

– To develop an X-ray active mirrors for micro or nano

focalisation

1st part of the talk1st part of the talk

– To develop at wavelength an in situ alignment system

2nd part part of the talk2nd part part of the talk

– To validate the potential of X-ray active mirrors (for micro

and nano focalisation)

Still in progressStill in progress

Outline

MARX PROJECT Team and fundingMARX PROJECT Team and funding

• Partners

• SOLEIL - project management and experimentation

• ISP SYSTEM - development of the active mirror

• IMAGINE OPTIC – development of wave front sensor

• Finances

• 50% from the ANR (Agence Nationale pour la

Recherche)

• 50% partners financing

MARX main GOALSMARX main GOALS

– Elliptic shape error less than 0.5 !rad RMS

– Focal distance 300 to 350 mm

– High range of elliptic shapes

– Possibility to correct residual polishing

defaults of the mirror

– Possibility to correct aberrations from the

optical system

Active optics = Magic Mirror25 years old 45 years old

Adaptive mirrors is the solution

Active optics = Magic Mirror

true example

We can win a lot from the Astrophysics research and expertise in

adaptive optics and wavefront analysis

Active optics = Magic Mirror

MARX

1st Part

Active mirror

PRINCIPLE OF ACTIVE MIRRORPRINCIPLE OF ACTIVE MIRROR

– The design of MARX permits to obtain naturally an elliptic shape withonly 1 bender

– The AME actuators apply correction strengths to the initial form inorder to obtain best focalisation and reduce the optical aberrations;

– Mirror characteristics:• Dimensions : 350x50x8mm

• Material : Silicium

– The ISP SYSTEM original concept has been recently patented

AME 2 AME 3 AME 4

320mm

AME 5AME 1 AME 9AME 8AME 7AME 6 AME 10

– 10 actuators generating micro forces along themirror

– 1 bender

!"##$#%

&' &( &) &* &+ &, &- &. &/ &'0

ISP SYSTEM DESIGN

• Active Kirckpatrick-Baez

mirrors system :

2 mirrors activated by

– 2 kinds of actuators :

• 1 bender

• and 10 micro strength

actuators (AME)

MARX Mirror DESCRIPTION

• The mirror is supported :

– At the first extremity by athin plate with strictionswhich allows 1 rotationand 1 translation. The jointis glued to the mirror

– At the other side, by apivot joint. The mirror ishold by a tighteningcontrolled system (jaws).

• The AME are fastened to softpads glued on the back face ofthe mirror. Soft pads are usedto limit the “print effects” on theactive face of the mirror.

1234567895#:!"##$#%;"6:<9542

:="4;:>4#"34"$6>

?"##$#:<5@

A5=>

B26@2#?"3#$:C4#2674;:D34854$#>:ED?F)0G

B26@2#

H2I2#:5#!

A5=>

?"##$#

?"3#$:C4#2674;:D34854$#>

:ED?F)0G

MARX Mirror DESCRIPTION

• A dedicated Control

Rack has been specially

developed and realized

for MARX application.

• The 19’’ Rack includes :

– 10 AME and 1

bender actuators

controllers with

integrated

microcontroller et

power driver

– Interface

communication

from a PC via RS

232

– Power supply

MARX Mirror CONTROL RACK

• Every actuator controller includes an algorithm of calibration with a dedicatedmathematical grading

ISP SYSTEM’s Micro Strength Actuator (AME)

• AME have been especially developed by ISP SYSTEM to be used in activeoptic systems which require very precise shapes (focalisation or wave frontcorrection).

• The concept, based on a calibrated strength generation, has been patented since2002.

• The strength generation is obtained by coupling a screw-nut system energized by abipolar stepping motor, with a floating head including springs.

• The strength range is about +/-30N with a repeatability of 10mN (othersconfigurations are available for customer applications).

D?F)0:8>2@:J$#:?D1K D?F(0:8>2@:J$#:95>2#>L:<#"!2@:54:?MN1OPO1D:(00,E"642#654"$659:2Q;"R"4"$6:$J:!"3#$:!23;56"3:"6@8>4#SG

APPLICATION FOR LASERS Mégajoule

Laser example

• Laser wave front correction

system

• Mirror : BK7

• shape : square mirror

• size :400x400mm

• material : BK7 glass

• Fitted of 39 ISP System’s

AME20 actuators

Mirror : Flat rectangular Silicon mirror (350 mm ! 40 mm ! 8 mm)Slopes errors measured on LTP at 0.5 "rad rms over 340 mm pupil size

Working conditions : P = 35000 mm / Q = 300 à 350 mm / # = 0.35°! Working curvature from 100 to 115 m! Working sag about 100 "m

Target ellipse

MARX Mirror targets

MARX

2nd Part

Active mirror

Metrology

• 1D local slopes measurement.

• Active surface can be face-up, face-down or on the side.

• Maximum length : 1 m.

• Precision : Curvature 0.3%, slope errors 0.2 "rad r.m.s.

• Calibration : on a flat reference surface - precision : 0,1 %.

• Radii of curvature : from 3 m to infinite.

• Gratings groove density measurements.

• Shape reconstruction by Stitching algorithm available.

Long Trace Profiler (LTP) – Laboratoire de Métrologie Optique @ SOLEIL

Polarizationinterferometer

MARX MIRROR OFF-LINE METROLOGY

M. Thomasset, S. Brochet, F. Polack

LTP measurement of the standing alone mirrorMirror dimensions: 350 mm ! 40 mm ! 8 mm

3 traces spaced by 10 mmMeasurement over 340 mm by 1 mm steps

LD: R = 22.2 km $ = 0.5 "rad rmsLC: R = 19.1 km $ = 0.5 "rad rmsLG: R = 19.5 km $ = 0.6 "rad rms

Slopes ("rad)

Heights (nm)

Roughness1.7 Å rms8 nm PV

Twist correction - interferometer- LTP

Alignment of the mirror

LTP measurements

Measurement over

300 mm by 1 mm steps

-374.5 mm

350 mm

Activejaw

-24.5 mm

Flexor

LTP direction of measurementEnding point

-40 mmStarting point

-340 mm

300 mm

11 10 9 8 7 6 5 4 3 2

0

20

40

60

80

100

120

140

0 5000 10000 15000 20000 25000 30000

couple Actionneur 1 (mN)

erre

ur

de

form

e (n

m)

0

0.5

1

1.5

2

2.5

0 5000 10000 15000 20000 25000 30000

couple Actionneur 1 (mN)

erre

ur

de

form

e (!

rad

)Curvature actuator : classical x-ray benderCurvature actuator : classical x-ray bender

Residual shape errors ("rad rms)

Residual shape errors (nm)

PV

rms

The mirror is pre-curved (R = 140 m)Curvature dynamic range : 140 to 72 m

Heigth at the center : 80 to 160 "m

-80000

-60000

-40000

-20000

0

20000

40000

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

hau

teurs

(nm

)

0

1000

5000

10000

15000

20000

25000

29000

-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

fonct

ion d

'infl

uen

ce (

nm

/mN

)

5000

10000

14000

Pousser/Tirer autour de 15000

70

80

90

100

110

120

130

140

150

0 5000 10000 15000 20000 25000 30000

couple Actionneur 1 (mN)

R (

m)

Curvature (m) vs. AME1 strength

In principle able to go from Flat to 55 m

Shape correction actuators

! Case of actuator n°6

We realized successive shape measurements for different strength values of AME6.(the nominal shape of the mirror is subtracted from all measurements).! Deformation of the mirror surface is symmetric.

The influence function is the same on the whole range of the actuator.

!Demonstrate the linearity of the system

(true for all actuators)

INFLUENCE FUNCTIONS

-80000

-60000

-40000

-20000

0

20000

40000

60000

80000

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

hau

teurs

(nm

)

0

10000

20000

29000

-10000

-20000

-29000

-0.5

0

0.5

1

1.5

2

2.5

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

fonct

ion d

'infl

uen

ce (

nm

/mN

)

10000

20000

29000

The actuators at the edges of the mirror induce 4 times less deformation of the surface (~0.5nm/mN) than those in the middle (~2 nm/mN).

INFLUENCE FUNCTIONS

-0.5

0

0.5

1

1.5

2

2.5

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

fon

cti

on

d'in

flu

en

ce

(n

m/m

N) Act2

Act3

Act4

Act5

Act6

Act7

Act8

Act9

Act10

Act11

Shape correction actuators

Shape correction actuators EIGEN MODES

SLOPE

HEIGHT

-4

-2

0

2

4

6

8

10

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

erre

urs

de

pen

tes

(!ra

d)

$ = 1.42 "rad rmsh = 24 nm rms

H = 81.6 nm PV

y = 6.861E-06x3 - 2.266E-02x2 + 7.167E+00x + 4.351E+03

0

1000

2000

3000

4000

5000

6000

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

pen

tes

(!ra

d)

Residual to bestelliptical fit

Curvature actuator @ 0

R = 139.519 m

Nominal shape measurement (all actuators @ 0)

Shape correction using the 10 inside actuatorsTarget: Best elliptical fit from nominal shape measurement

! = 0.59 "rad rms

h = 3.921 nm rms

H = 16.317 nm PV

Residual to bestelliptical fit

R = 139.6 m

Residual to target

R = 139.519 m! = 1.65 "rad rms

h = 389.4 nm rms

H = 1291.4 nm PV

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

erre

urs

de

pen

tes

(!ra

d)

-10

-8

-6

-4

-2

0

2

4

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

erre

urs

de

pen

tes

(!ra

d)

! Correction in a single iteration! Small drift in final curvature (can be corrected using AME1)

(strengths between -7N and +6N)

Mirror is hold in correction and curved to 100 m

! = 0.84 "rad rms

h = 9.86 nm rms

H = 39.66 nm PV

y = 6.861E-06x3 - 2.266E-02x2 + 7.167E+00x + 4.351E+03

0

1000

2000

3000

4000

5000

6000

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

pen

tes

(!ra

d)

Residual to bestElliptical fit

-3

-2

-1

0

1

2

3

4

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

erre

urs

de

pen

tes

(!ra

d)

Small degradation of the surface shape errors by curving the mirror

Curvature actuator @ 12N

R = 100 m

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

erre

urs

de

pen

tes

(!ra

d)

-3

-2

-1

0

1

2

3

4

5

6

7

-200 -150 -100 -50 0 50 100 150 200

abscisses (mm)

erre

urs

de

pen

tes

(!ra

d)

! = 0.55 "rad rms

h = 3.06 nm rms

H = 15.9 nm PV

Residual to bestelliptical fit

R = 100.32 m

Residual to target

R = 100 m

! = 1.357 "rad rms

h = 204.4 nm rms

H = 630 nm PV

! Correction in 2 iterations using the same interaction matrix

! Small drift in curvature

Shape correction using the 10 inside actuatorsTarget: Best elliptical fit from previous shape measurement (strengths between -10N and +9N)

- Curvature range from 140 to 72 m (possible to go from flat to 55 m)

- The mirror system is linear.- 1 single interaction matrix can be used.- Slope errors can be corrected to 0.6 "rad rms, in 1 or 2 iterations.- Small drift on curvature respect to the target ellipse.

- Upgrades:Use of AME1 to correct the drift in curvature: 2 steps correction

1- Curvature adjustement (use of AME1)

2- Residual shape errors correction (use of the 10 inside actuators)

- Coupling with a hard X-ray Hartmann wavefront sensor and closed loop experiment on

synchrotron radiation beamlines (end of 2008).

CONCLUSION

3rd Part

Wavefront

sensor

Principle of Shack Hartmann

wavefront sensor

f

%x&&

Microlenses array

Wavefront

sensingCCD

M(x)

tan(")=#x / f

Principle of Hartmann

wavefront sensor

L

Hole array EUV CCD camera

! = "y/L

x

y

z

Ref

eren

ce f

lat

wa

vef

ron

t

Reference spot

centroid position"y!

Ab

erra

ted

wa

vef

ron

t

Aberrated spot

centroid position

Hartmann wavefront sensor

1st TEST

0.6 "m

Perfect diffracted wavefront

Hartmann wave-front measurement at 13.4 nm with # EUV 120

accuracy Optics Letters, Vol. 28 Issue 17 Page 1534 (September 2003)

P. Mercère, P. Zeitoun, Mourad Idir, S. Le Pape, D. Douillet, X. Levecq, G. Dovillaire, S. Bucourt, K.A. Goldberg, P. Naulleau, S. Rekawa

CXRO Beamline 12.0 ALS

0.6"m pinhole

Reproducibility Sensibility – Precision

0.6"m pinhole

1.86 "m in the X direction

#EUV/125 rms , #EUV/16 PV

0.1 nm rms, 0.8 nm PV

#EUV/100 rms , #EUV/19 PV

0.13 nm rms, 0.8 nm PV

• Wavefront precision ~ #/100 rms (0.1 nm rms)

• Tilt Precision ~ 0.02 !rad rms

• Focal distance precision ~ 1. 10-5 m –1 rms

EUV Calibration

ALS beamline 12.0 wavefront measurement without spatial filtering

010

2030

40

0

10

20

30

40-4

-2

0

2

4

illuminated subpupils

illuminated subpupils

wavefront

(x Lambda=13.4nm)

42 "

m

23 "m

Measured spot size

(YAG crystal+"scope objective)

Ray tracing calculated

with Hartmann sensor software

based on phase measurement

KB optimization

SLS LUCIA wavefront system

Generation II

Hartmann Hartmann WavefrontWavefront measurement andmeasurement and correction correction

Calibration of the sensor on a spatially filtered reference beam at 700 eV

1-"m pinhole diffracted waveseen by the sensor

Absolute wavefront measurement after calibration

Closed loop correction at 3.64 keV

-6090 -6080 -6070 -6060

-500

0

500

1000

1500

2000

FWHM = 2.40 "m

Knife Edge Scan of Vertical X-ray beam at LUCIA

Derivative of Knife Edge Data Gauss Fit

Derivative o

f F

luore

scence

Vertical Position (!m)

5025 50305030 5035 50405040 5045 50505050 5055 50605060 5065

-450

00

450

900900

1350

18001800

2250

27002700

Derivative of Knife Edge Data Gauss Fit

Derivative o

f F

luore

scence

Horizontal Position (!m)

Knife Edge Scan of Horizontal X-ray beam at LUCIA

FWHM = 2.55 "m

Before correction After correction

7.7 nm rms

30.9 nm PV

0.8 nm rms

4.6 nm PV

Accuracy of the sensor is limited by shot noise.

But signal to noise ratio can be improved by accumulation of several images :

répétabilité vs moyennage image à 900 eV

0

0.5

1

1.5

2

2.5

3

0 5 10 15 20 25 30 35 40 45 50

nombre d'images cumulées

rép

éta

bilit

é (

nm

)

0

20

40

60

80

100

120

140

160

180

rap

po

rt s

ign

al/b

ruit

rms PV rapport signal sur bruit

Signal to noise ratio

Repeatability nm PV

Repeatability nm rms

répétabilité vs moyennage image à 2100 eV

0

0.5

1

1.5

2

2.5

3

0 5 10 15 20 25 30 35 40 45 50

nombre d'images cumulées

rép

éta

bilit

é (

nm

)

0

20

40

60

80

100

120

rap

po

rt s

ign

al/b

ruit

rms PV rapport signal sur bruit

Signal to noise ratio

Repeatability nm PV

Repeatability nm rms

Use of a high readout rate (30 Hz)

Hamamatsu CCD camera" Possible

IMPROVEMENTS

Accumulation of 500 images : 15 s

Repeatability : 0.038 nm rms

0.28 nm PV

Accumulation of 100 images : 3 s

Repeatability : 0.06 nm rms

0.42 nm PV

Direct detection

Hartmann

GridCCD

Wavefront1000x1000 pixels of 8!m

Grid pitch of 57 !m

Distance grid / CCD of 100mm

Problems :

The shot noise limits the sensor sensitivity

The CCD is slowly « destroyed »

by hard Xrays

The direct detection sensor is

adapted to soft Xrays

Hartmann wavefront sensors for Xray beams

Indirect detection : adapted for hard Xrays

640x480 pixels of 7.4!m

Visible magnification x4.5

Grid pitch of 20 !m

Distance grid / YAG of 14mm

Sensor sensitivity : 0.023 nm rms

Sensor accuracy : 0.25 nm rms

limited by the calibration process

The X-ray Hartmann wave-front sensor

for in situ alignment

Calibration process :

For visible wavefront sensors

Source on translation stages

Single mode fiber : no aberration

Measured tilts and curvature must fit the real movement of the source

Sensor

For XRays wavefront sensors

1/ We use a visible source, the Talbot diffraction effect gives us a calculable

image to adjust the main parameters of the calibration

2/ « at wavelength » on the most possible « aberration free » beam, some

measurements are done to finalize the calibration

Soft Xrays : A small pinhole is set in the beam to diffract a pure diverging

beam.

Hard Xrays : Some small areas of the beam are used for different positions

of the sensor relatively to the beam. These measurements are averaged.

Hartmann wavefront sensors for Xray beams

Hard Xrays : measurement of the influence of a curvable cristal on CRISTAL

beam line at SOLEIL at 10.6 keV

We plotted the curvature measured by the sensor (in dioptries) in function of the

position of the motors. The fit is obtained by using the laser propagation theory.

influence of a curvable cristal on the measured curvature by a WFS

-0.01

0

0.01

0.02

0.03

0.04

0.05

0.06

0 100 200 300 400 500 600 700 800 900

C1-C2 motors postions in steps

measure

d

curv

ature

in

1

/m

measurements

Laser theory

• Xrays beams can be modelized by

the propagation of a gaussian beam

in vaccum.

• Even at this short wavelength, the

effect of diffraction must be taken into

account.

• The Xrays source can be considered

as a « Waist »

19 m 18 m

SourceHard Xrays WFS

Motors to change the cristal curvature

Hartmann wavefront sensors for Xray beams

Conclusion

• Hartman based Wavefront sensor are available

from VUV to hard X-ray (Imagine Optic)

• Small actuator design on specs are available (ISP

System)

• A full adaptive optics solution is available

(Wavefront sensor + mirror on specifications) are

avalabile (Imagine Optic)

More to do

Test on a beamline the full system

SOLEIL and DIAMOND end of November

1. Development of KB mirrors with mirror cooling fixture for more

powerful X-ray beams ( MARX2)

2. Smaller mirror possible

130 x 40mm (torpédo shape)

2 AME40 (+40N) for curvature

8 AME3 (+/-3N) for small correction

10 mm distance

- Minimum radius 23 m

Mini MARX

– SYNCHROTRON SOLEIL• Mourad IDIR

• Pascal MERCERE• Thierry MORENO

• Murielle THOMASSET + Sylvain Brochet

– IMAGINE OPTIC• Xavier LEVECQ

• Samuel BUCOURT

• Guillaume DOVILLAIRE

• Johan FLORIOT

– ISP SYSTEM• Paul SAUVAGEOT

• Lionnel ESCOLANO

• Nicolas NIVELET

• Benjamin SAUX

THANKS TO THE MARX PROJECT TEAM