+ All Categories
Home > Documents > Microscopy and Microanalysis

Microscopy and Microanalysis

Date post: 23-Feb-2016
Category:
Upload: arlen
View: 27 times
Download: 1 times
Share this document with a friend
Description:
Latest Developments in Dynamic TEM: Revealing Material Processes at Nanometer and Nanosecond Scales. Microscopy and Microanalysis. August, 2012. - PowerPoint PPT Presentation
24
LLNL-PRES-569336 The work presented in this article was performed at LLNL under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 and supported fully by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering. Latest Developments in Dynamic TEM: Revealing Material Processes at Nanometer and Nanosecond Scales Microscopy and Microanalysis B. W. Reed , 1 T. LaGrange, 1 M. K. Santala, 1 J. T. McKeown, 1 W. J. DeHope, 1 G. Huete, 1 R. M. Shuttlesworth, 1 J. S. Kim, 1 T. Topuria, 2 S. Raoux, 3 S. Meister, 4 Y. Cui, 4 A. Kulovits, 5 J. M. K. Wiezorek, 5 L. Nikolova, 6 M. J. Stern, 7 J.-C. Kieffer, 6 B. J. Siwick, 7 F. Rosei, 6 and G. H. Campbell 1 August, 2012 1 Lawrence Livermore National Laboratory 2 IBM Research Division, Almaden Research Center 3 IBM T. J. Watson Research Center 4 Department of Materials Science and Engineering, Stanford University 5 Department of Mechanical Engineering and Materials Science, University of Pittsburgh 6 Institut National de la Recherche Scienti que 7 Departments of Physics and Chemistry, McGill University
Transcript

2011 LLNL Template

Latest Developments in Dynamic TEM: Revealing Material Processes at Nanometer and Nanosecond ScalesMicroscopy and MicroanalysisB. W. Reed,1 T. LaGrange,1 M. K. Santala,1 J. T. McKeown,1 W. J. DeHope,1 G. Huete,1 R. M. Shuttlesworth,1 J. S. Kim,1 T. Topuria,2 S. Raoux,3 S. Meister,4 Y. Cui,4 A. Kulovits,5 J. M. K. Wiezorek,5 L. Nikolova,6 M. J. Stern,7 J.-C. Kieffer,6 B. J. Siwick,7 F. Rosei,6 and G. H. Campbell1August, 20121Lawrence Livermore National Laboratory2IBM Research Division, Almaden Research Center3IBM T. J. Watson Research Center4Department of Materials Science and Engineering, Stanford University5Department of Mechanical Engineering and Materials Science, University of Pittsburgh6Institut National de la Recherche Scientique7Departments of Physics and Chemistry, McGill UniversityLLNL-PRES-569336The work presented in this article was performed at LLNL under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 and supported fully by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Materials Sciences and Engineering.Lawrence Livermore National LaboratoryLLNL-PRES-569336#1Single-Shot Dynamic TEM (DTEM) is aimed at solving problems in materials science . . .For materials science, in situ TEM is often concerned with:Microstructural evolutionPhase transformationsChemical reactions involving nanostructured materialsDamage

Laser-pulse driven photoelectron cathodeCathode drive laser Nd:YLF(5w) l = 211nm10 ns FWHM pulse widthSample drive laserNd:YAG= 1064nm or 355nm12 ns FWHM pulse widthe-sProbee-se-s

Sample LocationCCD CameraBeam ShifterIn most applications, these things never unfold exactly the same way twiceLawrence Livermore National LaboratoryLLNL-PRES-569336#. . . yet what you really want is to capture not a single shot but a movie.

Nucleation and growth of a single crystalline grain in laser-heated amorphous GeTe

This is not a montage of individual experiments

This is a set of nine TEM images captured in under 3 sLawrence Livermore National LaboratoryLLNL-PRES-569336#The Arbitrary Waveform Generation (AWG) Laser can produce temporally shaped laser pulses over an unprecedented temporal range with high spatial mode quality and energy.

Produces variable pulse widths from 5 ns to 10 s and pulse trains in 250 s temporal windowDesigned to deliver energies up to 1 J per pulse for high photoelectron yield after UV conversionFiber-based optical components, solid-state diode amplifiers and adaptive optics provide high stability and beam quality

Lawrence Livermore National LaboratoryLLNL-PRES-569336#

Movie Mode works by synchronizing an arbitrary series of laser pulses with a fast post-sample deflectorLawrence Livermore National LaboratoryLLNL-PRES-569336#

Movie Mode works by synchronizing an arbitrary series of laser pulses with a fast post-sample deflectorElectrostatic DeflectorAWG cathode laser generates a UV pulse trainSample drive laser1 pulse or drive event9 electron pulses with temporal resolution from 5ns to 10s and interframe times from 25ns to 250s 2k x 2k Single-electron Sensitive CCD camera High Voltage Electrostatic Deflector rasters pulse train onto the CCDPulse Train of ElectronsThe CCD is read out at the end of experiment and segmented into frames. Lawrence Livermore National LaboratoryLLNL-PRES-569336#66Method pioneered by Takaoka and Ura, and BostanjogloThe images are rastered in two dimensions

What the camera seesHow it's interpretedLawrence Livermore National LaboratoryLLNL-PRES-569336#The system is extremely programmable and can potentially operate in many different modes The laser itself can have essentially any temporal profile, subject to:~5 ns resolution~250 s total run time (with some limits)~1 J totalThe electrostatic deflectors can be made to operate in any sequence at all, switching on and off arbitrarilyThis includes switching during an electron pulse for streakingAnd of course all the usual TEM imaging and diffraction modes are available, along with in situ sample holdersNovel methods of driving the sample, beyond just a laser pulse, are of interestThe AWG laser may also be used as a sample drive laserLawrence Livermore National LaboratoryLLNL-PRES-569336#Example Application: Metastable phase transformations in Ge2Sb2Te5 (GST) Laser amorphization in Ge2Sb2Te5 takes ~10ns, laser crystallization ~100ns Amorphization requires very rapid quenching, ~1010 K/s Time-resolved crystallization experiments performed on amorphous films In situ laser switching achieved by use of a novel specimen geometry

crystallineamorphous

crystallineamorphous

Laser powerhigher power amorphizing pulselower power crystallizing pulse

time

Lawrence Livermore National LaboratoryLLNL-PRES-569336#9Phase change materials: Ge2Sb2Te5 crystallization022002113004111microcrystallineamorphouscurves offset for visibility

Intensity (arbitrary units)4 8 12k (nm-1)022002222133111024224Time-resolved crystallizationAs-deposited amorphous Ge2Sb2Te5 films were laser crystallizedVarious time-delays recorded

Rotationally averaged data used to map out the time to crystallization

50 ns delaySantala et al. J. Appl.Phys. 111 (2012)Lawrence Livermore National LaboratoryLLNL-PRES-569336#10Movie mode reveals previously invisible details of nucleation, growth, and impingement

How fast is the nucleation? When does it happen? How fast does the front move? When do stresses start to become important? How does the evolution change after impingement? Why is the microstructure nonuniform at the end?All these questions can now be answered directly.Lawrence Livermore National LaboratoryLLNL-PRES-569336#

a-Ge

5 mConventional TEM image of laser crystallized area of a-Ge filmZone I: NanocrystallineZone II: Elongated radial grainsZone III: Layers ofelongated grains oriented circumferentially and nanocrystalsExample application: Crystallization of amorphous germaniumyields three zones of differing crystalline morphology

Lawrence Livermore National LaboratoryLLNL-PRES-569336#

Example application: Development of microstructure in laser melt and resolidification of metallic thin films

NanocrystallineMicrocrystallineLiquidBeforeDuringAfter

Al -MetalunmeltedSixNy Membrane SiO2 Capping LayerSi SubstrateLiquidLaser IrradiationWe can observe morphological changes in liquid-solid interface of rapid lateral solidification (RLS) of molten metal films and quantitatively measure interface velocities (e.g., images show a planar front moving at ~3.5 m/s).Lawrence Livermore National LaboratoryLLNL-PRES-569336#Rapid Alloy Solidification: Al-7 at.% CuTime-resolved imaging reveals the evolution of the microstructure as the melted film re-solidifies

As-depositedAs-deposited films are nanocrystalline and~80 nm thick

As-deposited15 s20 s30 s25 sRe-solidified

Lawrence Livermore National LaboratoryLLNL-PRES-569336#14Regions 2 and 3: Large Columnar and Interior GrainsThere are 3 morphologies produced in the the film by laser irradiation:Small Al-rich grains with Cu segregated to the grain boundaries Region 1: Small Al-rich GrainsRapid Alloy Solidification: Al-7 at.% CuLarge columnar grains with [100] cube-direction growth axesColumnar grains produce two large interior grains and multiple wraparound grains

Dark-field STEM imageEDS Cu Map

[200]

[200]

Lawrence Livermore National LaboratoryLLNL-PRES-569336#15Rapid Alloy Solidification: Effects of Cu Content

60 s50 sRe-solidifiedRe-solidified40 s30 s

20 sRe-solidified25 s

7 at. % Cu12 at. % Cu18 at. % Cu(Eutectic)Effects of Increasing Cu ContentTime-resolved imaging shows, with increasing Cu content: Increased times for the alloy to solidifyA more jagged solid/liquid interfaceA shrinking columnar regionIncreasing number of interior grains Lawrence Livermore National LaboratoryLLNL-PRES-569336#16Example Application: Reactive multilayer foils (RMLF)

Reacted FoilUnreacted Foil~13 m/s (nm/ns) PropagationNiAlIntermetallicAtomic DiffusionThermal DiffusionReaction ZoneAtomic MixingLaserspot

1 mm

1 mmReacted foilNiAl B2 intermetallicUnreacted foilCross sectionPlan viewAdvantages of DTEM studies:Conventionally, direct metastable state observations at the nanoscale cannot be donePast studies have only used electron microscopy on quenched RMLFs

50 mm14.6 ms after drive initiation

LaserElectron PulseLawrence Livermore National LaboratoryLLNL-PRES-569336#1717Off-equiatomic compositions produce short-lived striations behind the reaction frontPeriodicityLength2 Al : 3 Ni670 nm40 mm3 Al : 2 Ni390 nm3.2 mmThe 3 Al:2 NiV foils have a shorter region of transient morphology and smaller periodicity than those grown at 2 Al:3 NiV

The observed Al loss in Al-rich samples from surface evaporation may result in the rapid quenching of the sample and explain the shorter length

8 mm2 Al : 3 NiV

3 Al : 2 NiV3 mmLawrence Livermore National LaboratoryLLNL-PRES-569336#

3 mm2 Al : 3 NiV

3 mm1 Al : 1 NiV

3 Al : 2 NiV3 mmA very likely interpretation is that the striations are liquidThe estimated temperature rise from an adiabatic calculation is ~100K below the congruent melting pointLawrence Livermore National LaboratoryLLNL-PRES-569336#19With the Movie-mode capability, we can directly follow rapid phase transitions and morphological changes at the reaction front

Lawrence Livermore National LaboratoryLLNL-PRES-569336#Similarly, in Ti-B multilayer foils, we have observed a thin transient zone behind the reaction front that could contain multiple phases

1 mThis transient zone may be two-phase as in the NiAl system, containing TiB2 in both solid and liquid phases.Dark-contrast regions between grains have been observed in a zone roughly 3-4m in size behind the propagating front.Liquid?Lawrence Livermore National LaboratoryLLNL-PRES-569336#We have observed the phase transitions with time-resolved electron diffraction in a 1.75 m region at the propagating reaction front

The Ti-B RMLF transforms from nc-Ti/a-B layered structure to nc-TiB2 within 625ns.This implies an extremely high cooling rate.Lawrence Livermore National LaboratoryLLNL-PRES-569336#Movie Mode Dynamic TEM has arrived.Where do we go from here?Tweak the system for better performanceRe-engineer the gun for better brightnessUnderstand the real resolution limits. How bad is stochastic blur? What about damage?Develop new operating modes based on this technologyIntegrate in situ sample holders and sample drivesIntegrate with advanced detectors

And of course, now that the capability exists,Do the best dynamic materials science that we can! Capture the crucial in between moments that have never been seen before.Lawrence Livermore National LaboratoryLLNL-PRES-569336#23To calibrate models on reactive powder compacts, we are studying reaction dynamics in Ti-B RMLFs

Three foil compositions: Ti-2B, 2Ti-3B and 3Ti-2B.Films are capped with Ti and have 6.5 bilayers with a 39nm bilayer thickness, giving a 250nm total thicknessReaction front propagates at 13.2 0.1m/sAdiabatic heat rise (~3500K) suggests melting of both layers near front (Z.A. Munir Am. Ceram. Soc. Bull. 67, 342 (1988))Observed Ti loss suggests that evaporative cooling may be importantWork in Collaboration with David Adams at SNL AlbuquerqueLawrence Livermore National LaboratoryLLNL-PRES-569336#


Recommended