1. Laser Plasma and Laser-Matter Interactions Laboratory Particulate Formation in Laser Plasma M. S. Tillack, S. S. Harilal, C. V. Bindhu and D. Blair Center for Energy Research and Mechanical and Aerospace Engineering Department Jacobs School of Engineering http://aries.ucsd.edu/LASERLAB 2003 Simposio en Fisica de Materiales Centro de Ciencias de Materia Condensada Universidad National Autónoma de México 24 January 2003

2. Laser plasmas have numerous applications in science and industry  Micromachining  Thin film deposition  Cluster production  Nanotube production  Surface modification  Surface cleaning  Elemental analysis  X-ray laser  Photolithography  Medicine  Inertial fusion energy

3. Problems in micromachining are caused by workpiece & equipment contamination Contaminated surfaceAfter cleaning Laser ink-jet printer head before and after cleaning (courtesy of HP) Laser entrance window on our ablation chamber

4. Use of lasers in thin film deposition (PLD) is also limited by a lack of control over particulate Problems to be addressed Particulates Quality of films depends on deposition conditions – Detailed study of plume dynamics is necessary Lack of adequate theory Advantages Almost any material Doesn’t require very low pressures Reactive deposition possible Multilayer epitaxial films

5. Control of nanotube and nanoparticle fabrication requires a better understanding of the production mechanisms  Understanding why laser ablation produces such high nanotube yields is a high priority  Species responsible for growth?  Spatial distribution and transport?  Growth times and rates? Schematic of nanotube synthesis (D. B. Geohegan, ORNL)

6.  Absorption, reflection  Heat transfer  Thermodynamics (phase change)  Plasma breakdown  Shock waves (gas)  Stress waves (solid)  Laser-plasma interactions  Gas dynamic expansion  Atomic & molecular processes  Others Understanding the mechanisms of particulate formation and methods to control them will enable greater use of lasers A wide range of physics is involved:

7. Subtopics 1.Experimental studies of the expansion dynamics of plumes interpenetrating into ambient gases 2.Modeling and experiments on homogeneous nucleation and growth of clusters (surface ejection is another topic of interest to us!)

8. 1.Experimental studies of the expansion dynamics of plumes interpenetrating into ambient gases 0.01Torr 1Torr 0.1Torr 10Torr 100Torr

9. Lasers used in UCSD Laser Plasma and Laser-Matter Interactions Laboratory Lambda Physik 420 mJ, 20 ns multi-gas excimer laser (248 nm with KrF) Spectra Physics 2-J, 8 ns Nd:YAG with harmonics 1064, 532, 355, 266 nm

10. Experimental setup for studies of ablation plume dynamics Target : Al Laser Intensity : 5 GW cm -2 Ambient : 10 -8 Torr – 100 Torr air

11. Approximate plasma parameters Electron Density: Measured using Stark broadening Initial ~ 10 19 cm -3 Falls very rapidly within 200 ns Follows ~1/t – Adiabatic Temperature: Measured from line intensity ratios Initial ~8 eV falls very rapidly (Experiment Parameters: 5 GW cm -2, 150 mTorr air)

12. Plume behavior at low pressure

13. Below 10 mTorr the plume expands freely Laser intensity = 5 GWcm -2, Intensification time = 2 ns Each image is obtained from a single laser pulse Plume edge maintains a constant velocity (~ 10 7 cm/s) P = 10 -6 TorrP = 10 -2 Torr

14. Plume behavior in weakly collisional transition regime

15. The plume splits and sharpens at 150 mTorr Strong interpenetration of the laser plasma and the ambient low density gas Observed plume splitting and sharpening. This pressure range falls in the region of transition from collisionless to collisional interaction of the plume species with the gas Enhanced emission from all species

16. Plume behavior in strongly collisional transition regime

17. Instabilities appear at 1.3 Torr  Plume decelerates  Instability appears  Intensity peaks in slower component

18. Plume behavior in high pressure regime

19. Above 10 Torr the plume remains confined P = 10 Torr airP = 100 Torr Air

20. Summary of plume dynamics vs. pressure Fitting Models:  Free expansion: R ~ t  Shock Model: R ~ (E o /  o ) 1/5 t 2/5  Drag Model: R = R o (1–e -bt )  Best fit at 150 mTorr R ~ t 0.445 (Harilal et. al, Journal of Physics D, 35, 2935, 2002)

21. Using both spectroscopy and imaging, a triple-fold plume structure was observed Al (396nm) at 18 mm in 150 mTorr air  First peak in the TOF is not seen in the imaging studies – low dynamic range of ICCD?  The delayed peak does not match well with a SMB fit  A convolution of two SMB fits matches well  Faster peak – a small group of high KE – suffer negligible background collisions  Slower peak – undergo numerous collisions with background and decelerate (Harilal et. al, J. Applied Physics, in press, 2003)

22. 2.Modeling and experiments on homogeneous nucleation and growth of clusters Target Contact Surface Shock Condensed Particulates

23. Classical theory of aerosol nucleation and growth Homogeneous Nucleation (Becker-Doring model) Condensation Growth Coagulation where the coagulation kernel is given by Convective Diffusion and Transport Particle Growth Rates

24. Dependence of homogeneous nucleation rate and critical radius on saturation ratio High saturation ratios result from rapid cooling from adiabatic plume expansion Extremely small critical radius results Reduction in S due to condensation shuts down HNR quickly; Competition between homogeneous and heterogeneous condensation determines final size and density distribution

25. Effect of ionization on cluster nucleation Ion jacketing results in an offset in free energy (toward larger r * ) Dielectric constant of vapor reduces free energy Cluster birthrate vs. saturation ratio (Si, 10 9 W/cm 2, 1% ionization)

26. A 1-D multi-physics model was developed  Laser absorption  Thermal response  Evaporation flux  Transient gasdynamics  Radiation transport  Condensation  Ionization/recombination absorption I o e – x, w/plasma shielding cond., convection, heat of condensation 2-fluid Navier-Stokes simple Stephan-Boltzmann model modified Becker-Doring model modified Saha, 3-body recombination Target : Si Laser Intensity : 10 7 –10 9 W cm -2 Ambient : 500 mTorr He

27. Model prediction of expansion dynamics High ambient pressure prevents interpenetration (in any case, the 2-fluid model lacks kinetic effects)

28. The plume front is accelerated to hypersonic velocities Spatial distribution of nucleation and growth rates at 500 ns ~62 eV Thermal energy is converted into kinetic energy; collisions also appear to transfer energy from the bulk of the plume to the plume front Surface temperature and laser irradiance vs. time ~2 eV

29. Model prediction of cluster birth and growth Spatial distribution of nucleation (*) and growth (o) rates at 500 ns Time-dependence of growth rate/birth rate Clusters are born at the contact surface and grow behind it Nucleation shuts down rapidly as the plume expands

30. Besides spectroscopy and Langmuir probes, witness plates served as our primary diagnostic Start with single crystal Si HF acid dip to strip native oxide Spin, rinse, dry Controlled thermal oxide growth at 1350 K to ~1  m, 4 Å roughness Ta/Au sputter coat for SEM Locate witness plate near plume stagnation point Witness plate prior to exposure, showing a single defect in the native crystal structure Witness plate preparation technique:

31. Measurement of final condensate size 500 mTorr He 5x10 8 W/cm 2 5x10 9 W/cm 2 5x10 8 W/cm 2 5x10 7 W/cm 2 Good correlation between laser intensity and cluster size is observed. Is it due to increasing saturation ratio or charge state?

32. Saturation ratio and charge state derived from experimental measurements Maximum ionization state derived from spectroscopy, assuming LTE Saturation ratio derived from spectroscopy, assuming LTE Saturation ratio is inversely related to laser intensity!

33. Cluster size distribution – comparison of theory & experiment The discrepancy at low irradiance is believed to be caused by anomolously high charge state induced by free electrons

34. Summary and future work We have obtained a better understanding of the mechanisms which form particulate in laser plasma –Clusters in the size range from 5-50 nm are routinely produced at moderate laser intensity –Model predictions appear to match experimental data In-situ particle measurements (scattering, spectroscopy) would be very useful to further validate the mechanisms Better control of size distribution and enhanced yield are desired Model improvements are needed: 2-D, kinetic treatment Applications of nanoclusters & quantum dots will be explored research supported by the US Department of Energy, Office of Fusion Energy Sciences and the Hewlett Packard Company, Printing and Imaging Group