BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Time and Frequency as New Frontiers in Microscopy - Archie Howie\,
  Department of Physics\, University of Cambridge
DTSTART:20061017T140000Z
DTEND:20061017T150000Z
UID:TALK5712@talks.cam.ac.uk
CONTACT:Edmund Ward
DESCRIPTION:Electron energy loss spectroscopy in the 1eV to 2keV range wit
 h\nspatial resolution on the near-atomic scale is now well established\nas
  a vital ancillary technique in high resolution electron\nmicroscopy. Taki
 ng a wider view of the whole energy loss or frequency\nspectrum however\, 
 it has been noted that the situation is rather less\nimpressive. Over many
  decades of energy loss below 1eV or frequencies\nbelow 1015 Hz\, scanned 
 probe microscopy\, near-field microscopy or\nscanning optical microscopy t
 echniques appear to be more effective.\nOn the evidence these provide\, th
 ere is no reason to reject the view\nthat these spectral regions are heavi
 ly populated with interesting\nphenomena. Using RF and other pulsing techn
 iques\, the frequency range\nbelow about 1MHz has been explored in a few i
 solated TEM or SEM\ninvestigations.\n\nWithin the limits of linear behavio
 ur\, a complete knowledge of the\ncomplex response as a function of freque
 ncy is sufficient to specify\nthe time response of a specimen to any appli
 ed signal including a\nsharp impulse. Indeed the standard EELS method\, be
 ing based on the\ndelivery of a series of such impulses from individual el
 ectrons in\nthe beam\, operates that principle in reverse.  When the respo
 nse to a\nperiodic input signal of frequency w is mapped out by phase-sens
 itive\ndetection methods\, as for example in scanning electron acoustic\nm
 icroscopy (SEAM)\, it is even possible to study non-linear features\nof be
 haviour by measuring the response at say 2w.\n\nAlthough it is not so far 
 possible to scan whole images by scanning\ntunneling microscopy in less th
 an about 1ms\, the basic non-\nlinearities of the tunneling process facili
 tate the probing of high\nfrequency dynamic response at selected positions
 .  Methods include\nmagnetostrictive oscillation of the tunneling distance
  and the use of\nlaser pulses in sum-difference frequency methods and in o
 verlapping\npulse techniques. With split laser pulses in the traditional p
 ump-\nprobe procedure\, the probe pulse can also be used to collect the\ns
 ignal at a controllable time delay by using it to trigger a photo-\nactiva
 ted gate in the tunneling circuit. Pump-probe optical pulse\ntechniques ca
 n also be employed in apertureless near-field microscopy\nwhere the tip fi
 eld enhancement effect provides the necessary\nlocalization of the pump pu
 lse. Scanned probe microscopy has led the\nfield in exploring these new po
 ssibilities\, but it is probably true\nto say that we still await really e
 xciting applications.\n\nTEM imaging in pump-probe operation has recently 
 been demonstrated by\nsplitting a train of laser pulses to pump the sample
  and to generate\na probe pulse of electrons by exciting a photocathode el
 ectron source\nwith a controllable delay time.  Images at x100\,000 were o
 btained in\nseconds using sub-100fs pulses with as few as one electron per
  pulse\nand a repetition rate of 80MHz. This must be the best approach for
 \ndynamic imaging of non-damaging phenomena which can be repeated many\nti
 mes under identical starting conditions.  More general\, ultra-fast\nimagi
 ng requires single shot operation with millions of electrons in\none pulse
  and for high spatial resolution raises truly formidable\nproblems of elec
 tron optics.
LOCATION:Gordon Seminar Room\, Austin Building\, Materials Science and Met
 allurgy\, Department of
END:VEVENT
END:VCALENDAR