Force controlfornanohandling insideScanningElectronMicroscopes - - PowerPoint PPT Presentation

▶
force controlfornanohandling
SMART_READER_LITE
LIVE PREVIEW

Force controlfornanohandling insideScanningElectronMicroscopes - - PowerPoint PPT Presentation

Dec 15, 2023 413 likes •708 views

TC-ANH Force controlfornanohandling insideScanningElectronMicroscopes Daniel Jasper Division Microrobotics and Control Engineering 1 Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow TC-ANH Contents

slide-1
SLIDE 1

TC-ANH

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

1

Daniel Jasper

Force controlfornanohandling insideScanningElectronMicroscopes

slide-2
SLIDE 2

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

2

Daniel Jasper, ICRA 2009 17.05.2009

Contents

►Introduction ►Limits of currentSEM-basedtracking ►Trackingwithbeamcontrol ►Isthe SEM a perfectsensor? ►Force sensinggripper ►Conclusions and Outlook

slide-3
SLIDE 3

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

3

Daniel Jasper, ICRA 2009 17.05.2009

Introduction

►Why do weneed a global sensorfornanohandling? ►Insufficient robot / sensoraccuracy

► Buy betterrobots! ► But: MorethanthreeDoF?

►Imprecisemounting / manufacturing

► Do calibration! ► But: Usingwhatreference? ► But: Feedingsystem?

►Drift

► Do drift compensation (AFM..)! ► But: Whatifmoretoolsarerequired?

Robot A Robot B Someconnection Tool Specimen

several cm nm accuracy needed

slide-4
SLIDE 4

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

4

Daniel Jasper, ICRA 2009 17.05.2009

Introduction

►Requirement: Nanoobjectsmustbedetectable ►Availalbe global sensors:

► Scanning Probe Microscopes

► Seriousspacerestrictions, tinyworkingspace ► Speed

► Transmission ElectronMicroscope

► See SPM

► ScanningElectronMicroscope

► Usable ► Key challenge: Efficientobjecttracking

► Otherideas?

slide-5
SLIDE 5

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

5

Daniel Jasper, ICRA 2009 17.05.2009

Limits of currentSEM-basedtracking

►Image acquisitionslow ►Distortedmovingobjects ►Image processingslow ►Limited update rate ►Significantdead time ►Idea: Do notacquireimages

slide-6
SLIDE 6

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

6

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol – Idea

►Interpolation in high-resolution linear encoders ►Required: Sine and cosinesignal

slide-7
SLIDE 7

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

7

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol

►Sine / cosinevaluefrom image data

slide-8
SLIDE 8

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

8

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol

►Whatifthesignalisnotexactlysine-shaped? ►Sine/Cosineinterpolationwiththelissajousplot

slide-9
SLIDE 9

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

9

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol

►Scanning a specific, grid-like target structure ►Standard SEM calibration chessboard

slide-10
SLIDE 10

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

10

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol

►Results of 10 measurements on each step

► Accurate measurements ► Sub-pixel resolution ► Low noise ► 16 nm step visible

slide-11
SLIDE 11

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

11

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol

►Advantages

► Computationally simple ► Robust against noise ► Robust against changing brightness / contrast ► Sub-pixel accurate

►Disadvantages

► Needs target pattern ► Needs acquisition of complete images ► Limited to a number of pixels

slide-12
SLIDE 12

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

12

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol

►Line-scan is sufficient for position detection ►Scanning a square on the target ►If target moves, adjust the scanned square

slide-13
SLIDE 13

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

13

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol – Movementdetection

►Target moves

► Signals change correspondingly

►New target center position can be calculated exactly with

sine/cosine interpolation

►Resolution/Noise <10 nm

slide-14
SLIDE 14

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

14

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol – Using a real target

►Target: SEM calibration grid

► Smallest square: 1 µm

►Works without exact sine waveform ►Beam-induced deposition visible

► Drift?

slide-15
SLIDE 15

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

15

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol – Results

►Defocused image ►Unaffected as long as pattern still visible

slide-16
SLIDE 16

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

16

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol – Results

►Brightness and

contrastchange

►Outliers

►Rotating

►No influence on x and y

slide-17
SLIDE 17

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

17

Daniel Jasper, ICRA 2009 17.05.2009

Tracking with beam control – Results

►Different specimen currents ►Reduced sharpness: Not important, no edge detection ►Better signal-to-noise ratio: Less noise

slide-18
SLIDE 18

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

18

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol – Results

►Magnification ►Almost identicalresults on 1000x and 10000x

magnification

slide-19
SLIDE 19

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

19

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol – Long rangescan

►1000x magnification ►Working range: ~200 µm ►Little to no impact on noise and

resolution

slide-20
SLIDE 20

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

20

Daniel Jasper, ICRA 2009 17.05.2009

Trackingwithbeamcontrol – Properties

►Can measure x, y and φ ►Update rates up to 1 kHz and more ►Noise <10 nm ►Immune to brightness/contrast/focus changes ►Long-range scanning (±100 µm)

► Four orders of magnitude

►Combination with imaging possible

► Multi-tasking / Scheduling ► E.g. Measure position after each line of image scan

slide-21
SLIDE 21

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

21

Daniel Jasper, ICRA 2009 17.05.2009

So, isthe SEM a perfectsensor?

►No! ►Inherentlyserialprocess, no batch-processing

► Similar to otherrequiredtechnologies ► Increasethroughput

►Hard to integrateintostandardmanufacturingline

► Vacuum Airlocksnecessary ► Feedingsystems to putobjectsontonanorobots ► ... ► But: Excellentforprototyping☺

By esec semiconductor

slide-22
SLIDE 22

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

22

Daniel Jasper, ICRA 2009 17.05.2009

So, isthe SEM a perfectsensor?

►Time-variantrelationbetween SEM and worldcoordinates

► But: 99% of operationsonlyrequired relative coordinates

►Complextargetpatternsneed to becreated

► Canbesimplified ► Work in progress: Automatically find objectmodel and edges

►Trackingworks in plane, butwhatabout 6 DoF

► Z/Height-informationrequired ► Work in progress: Focus information, Stereoscopic SEM imaging

►Position tracking good, but I need force detection

► Visual force detection

y y x x SEM world

slide-23
SLIDE 23

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

23

Daniel Jasper, ICRA 2009 17.05.2009

Simplifiedscanningpattern

►No rotation measured ►Onlyonepoint-shapedpatternrequired

slide-24
SLIDE 24

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

24

Daniel Jasper, ICRA 2009 17.05.2009

Force-sensinggripper

Gripperdesigned and fabricatedby DTU, Denmark

slide-25
SLIDE 25

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

25

Daniel Jasper, ICRA 2009 17.05.2009

Pushing a CNT againstthegripper

slide-26
SLIDE 26

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

26

Daniel Jasper, ICRA 2009 17.05.2009

Pushing a CNT againstthegripper

►10-fold averaging ►100 Hz update rate ►Stiffness: ~4 N/m ►Max.: 12 nm ≅ 48 nN ►Noise: 2 nm ≅ 8 nN ►Carefullstiffnessselection:

►Too soft: no gripping ►Toohard: lowresolution

slide-27
SLIDE 27

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

27

Daniel Jasper, ICRA 2009 17.05.2009

Conclusions

►High update rates ►nm resolutionforpositionmeasurements (x, y and φ) ►nN resolutionfor force measurements ►Force-feedbackgripperwith no extra wiring ►Lessbeam-inducedcontamination on importantparts

slide-28
SLIDE 28

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

TC-ANH

28

Daniel Jasper, ICRA 2009 17.05.2009

Outlook

►Position / Force measurement without special pattern

► Using model of objects ► Detecting edges

►Height measurements

► Focus / Sharpness ► Stereoscopic images

►High-resolution (field-emission) SEM

slide-29
SLIDE 29

TC-ANH

Division Microrobotics and Control Engineering Technology Cluster Automated Nanohandling Head: Prof. Dr.-Ing.habil. S.Fatikow

29

ThankYou