Download to read offline
Course entry t hz and ultrafast
- 1. Beyond the visible:A tour to future of spectroscopy and imaging Barmak Heshmat Dr. Ramesh Raskar Dr. C. Barsi 1
- 2. The big picture •Beyond the visible/IR spectrum (THz spec.) – New hardware trends – New computational trends • Beyond the line of sight (multihop imaging) – Seeing around the corners – Seeing through the diffusers • Beyond the resolvable (subwavelength imaging) – New hardware trends(course p1) – New computational trends(course p2) 2
- 3. Spectroscopy • EM waves •Many types of spectroscopy 3
- 4. Wave of spectrometers •They were all there in the lab but now they are entering consumer market! – Optical absorption diagnostic – Raman food analysis – THz skin, cosmetics, pharm. Electronics starting to become portable Optics starting to become portable
- 5. Example Just like supercomputers we still need the accurate lab spectrometers but portable versions can be used in limited applications. • Raman spectrometer from lab to the key chain! Tellspec DeltaNu® ReporteR™ Smiths Detection RespondeR™ RCI Microphazir™ Horiba T64000 ?
- 6. Hyperspectral and multispectralimaging 6 http://www.markelowitz.com/Hyperspectral.html
- 7.
- 8.
- 9. Beating the diffractionlimit 9 Superlensing Enhanced near field probes Fluorescence imaging Super oscillatory lenses Diffraction limit has limited our resolution in imaging now we are learning ways to go beyond this limit.
- 10. Seeking light afterscattering • Going from imaging for human to imaging for computers (measurement in other mathematical spaces and reconstructing the image) • Going from single scattering imaging to multi- scattering imaging. 10 2nd Bounce 1st Bounce 3rd Bounce
- 11.
- 12. New hardware trends •Introductions • Applications • PC Switches – New Materials for THz – Optimizing Excitation of PC Switches – Nanoplasmonic Structures • Summary • Questions? 12
- 13. …THz 400 THz Frequency(Hz) 800 THz Unique spectroscopy capabilities Study ofTHz dynamics Faster communication Imaging and inspection 13
- 14. Why THz • Noninvasive •Water in biological systems, protein folding, disease state of tissue • Vibrational modes for organic molecules • Picosecond time scale dynamics 14
- 15. THz and tissues •Can measure absorption and refraction index together through pulsed imaging. 15
- 16. THz imaging • Securityapps, (mm wave <> THz) • More inspection and analysis apps 16 See a whole gallery here: http://thznetwork.net/index.php/thz-images Jefferson Lab Ken O, UT, Texas Startiger project D. Mittleman Rice U Q. Hu, MITBESSY, Germany- (100um res)
- 17. THz microscopy 17 R. Kersting,THz-ANSOM 150nmEpithelial tumor cell, A. Tredicuccii, ~15um Diffractionlimit Ordinary imaging Near field imaging Scanning probes D. Zimdars, Picometrix, Inc,
- 18. New trends inhardware 18
- 19.
- 20.
- 21.
- 22.
- 23.
- 24. H2O, Here is whatis detected Temporal profile Frequency composition Math 24
- 25. Our ultimate dream was! Last10 years in our lab This year in our lab Future, in our hand The miniaturization process 25
- 26.
- 27.
- 28. New Materials forTHz 28
- 29. Conventional Materials The philosophyof an optical switch defines the desired properties of the substrate material. highest level of fast photoconductivity modulations: • high optical density • high thermal breakdown limit • high mobility, and Vb and Vsat • short carrier lifetime (sub-picosecond) • low dark conductance • PC switching started by Austin on Si in 1975 (D.H. Auston, Appl. Phys. Lett., 26 (3) 101 (1975)) • C.H. Lee used GaAs in 1977(C.H. Lee, Appl. Phys. Lett., 30 (2) 84 (1977)) • M.Y. Frankel used LT-GaAs in 1990 (M.Y. Frankel, et al, IEEE Trans on Elec. Devices, 37, 2493, 1990). 29
- 30. LT-GaAs • LT-GaAs hasshort carrier lifetime (<1ps) • It has low mobility as well GaAsBi • Bi is a group V poor metal GaAsBi is shrinking bandgap material 30
- 31. GaAsBi Results • 500GHz bandwidth improvement • Interesting emissions! 31
- 32. Effect of GaAsBigrowth condition • THz emission with variation of different parameters 32
- 33. Carbon nanotubes Increasing the performance withcarbon nanotubes between the gold electrodes of the chip 33
- 34. So we madesamples. 34
- 35.
- 36. Nanoplasmonics • Engineering surface electrondensity waves in the metallic nanostructures to achieve an enhanced optical response. • A key property of nanoplasmonics is its capability to efficiently couple light into subwavelength structures. 36
- 37. Nanoplasmonics: An Example Tuningannular nano- apertures B. Heshmat, D. Li, T. E. Darcie, R. Gordon, " Tuning plasmonic resonances of an annular aperture in metal plate "Optics Express, Vol. 19, Iss. 7, pp. 5912–5923 (2011). 37
- 38. Nanoplasmoincs for THzPC Switches 38
- 39. Nanoplasmoincs in THzPC switches B. Heshmat, H. Pahlevaninezhad,Y. Pang, M. Masnadi, R. Lewis, T. Tiedje, R. Gordon and T. E. Darcie "Nanoplasmonic Terahertz Photoconductive Switch" Nano letter, accepted. 39
- 40. Results of UsingNanoplasmonic Structures Peak-to-peak response enhancements of 40×, 10×, and 2×, compared to GaAs, LT-GaAs and Commerical device. 40
- 41.
- 42. Challenges • THz waveshave long wavelength; biological structures, many important ones, are small… • Living things need water: THz radiation and water are not “best friends”… • Unless you work hard, no clear spectroscopic features at THz are visible for many samples. • Some solutions to above problems are coming out. 42
- 43. Summary of newtrends in hardware • 100 GHz to 10THz region of EM waves are called THz, have been unexplored, but we are finally closing the gap. • Main challenge is detection and generation. • Major sources and QCLs, schottky diodes, PC switches and nonlinear crystals. • There is room for enhancement through material, optics and nanoplasmonics. • Many exciting applications from early cancer detection to inspection of organic materials and faster telecommunication. 43
- 44.
- 48. • They alsoinvestigated the difference between a random mask and an optimized one.
- 49.
- 51. The optimal blocksize for the block-based CS is a function of the local image characteristics, and different block sizes can be assigned to different regions.
- 55. Summary of computationaltrends • Compressive measurements, where you measure the minimum amount of points to reconstruct an image with known priors. • Layer separation based on pulse features • Reference-free measurements in THz imaging • Here is a demo: 55
- 56. Beyond the lineof sight 56
- 57. Time-of-flight In Situ remotesensing Require direct path between objects sensor JPL Hyperspectral Imaging Spectroscopic Monterrey Bay Aquarium Research Institute http://www.mbari.org/coastal/ http://earthobservatory.nasa.gov/Features/Lidar/ http://aviris.jpl.nasa.gov/html/aviris.freedata.html Optical remote sensing
- 58. What if thereis no direct path? Receiver Source ?
- 59. Computation + optics J.Bertolotti, et al. Nature 491 (2012).S. M. Popoff, et al. Nat. Commun. 1 (2010) • Relies on coherence/correlation • Small field of view • Short standoff distance
- 60. 60Nature Photonics 6,549–553 (2012)
- 61. A. Velten, etal. Nat. Commun. 3 (2012). Time is a parameter for imaging
- 62. x t Hyperbola x Laser Streak camera Diffuser Object Time-resolved imageformation Source: Ti:Sapph (λ0 =795nm, but could use other wavelengths) Detector: Streak Camera (δt ≈2ps) Different ray paths register at different times hyperbolic impulse response (x – ct)
- 63. Il (x,t)= I0G(xl, x, ¢x )N(qin )N(qout )R( ¢x )d ct -(rl ( ¢x )+rc ( ¢x ))( )d ¢xò Time-resolved image formation Geometry Diffuser Object Reflectance Time constraint
- 64. Il R(x), N(qin,out){ } 1)(0for,ˆˆ1 minarg 1 2(.)),( xRII L L l num ll meas l NxR Inverse problem Given a set of streak images Find the unknown reflectance R(x)
- 65.
- 66.
- 67. Experimental setup • Needto know something about diffuser
- 68.
- 69. Unknown reflectance • Assumeobject geometry known (can get from previous work) • Wide field reconstruction • Works for incoherent light
- 70. Moving on tothe miniaturization
- 71. Time of flightcamera • Continuous wave instead of pulsed • Cheaper, safer, more compact, but less accurate. R. Raskar, et al., “Coded Time of Flight Cameras: Sparse Deconvolution to Address Multipath Interference and Recover Time Profiles”, SIGGRAPH Asia 2013.
- 72. 3d imaging through turbulence Solvingocclusion problems www.picassodreams.com/photos/nyc_skyscrapers/ http://www.nasa.gov/vision/earth/lookingatearth/h2005_katrina.html http://www.fjellandfjord.com/article.php?id=166 http://www.soest.hawaii.edu/GG/HCV/loihi.html Generalizations for remote imaging
- 73. Summary of timeof flight imaging • Moving from single scattering to multiscattering (multihop) imaging • Different reconstruction techniques that rely on previous optimization techniques can be used. • Moving from expensive ultrafast hardware to cheaper slow hardware that operates on modulated light • Now we can recover what is in the visible volume of these cameras N. Naik, C. Barsi, A. Velten, R. Raskar. “Estimating spatially varying reflectance through scattering layers using time-resolve inversion.” JOSA A.
- 74. Two picosecond timeresolution Streak camera details