Transcript Lecture 11.Super-resolution methods

Super-resolution Methods
I - PALM
Detecting A Single Fluorescent Molecule?
• Size: ~ 1nm
• Absorption Cross-section: ~ 10-16 cm2
• Quantum Yield: ~1
Absorbance of 1 molecule = ?
How many fluorescence photons per excitation photons?
Single Molecule “Blinks”
Myosin V -- a motor protein.
De-convolution Microscopy
Thompson, RE; Larson, DR; Webb, WW, Biophys. J. 2002,
Paul Selvin
Accuracy
 
( s  a / 12 ) / N
2
2
# of photons
Photo-activation
De-convolution
Photo-switchable Fluorescent Protein
Gurskaya NG et al. 2006 Nat. Biotechnol.
Photo-activation Localization
Microscopy (PALM)
stochastic optical reconstruction microscopy
STORM
Ground-State Depletion (GSDIM)
What Next?
•
•
•
•
Z-resolution
Better fluorescent proteins
Multiple-color labeling
Cryo-temperature imaging
II. NSOM
Super-Resolution: Beyond Diffraction Limit of λ/2:
Near-Field: Distance <<Optical Wavelength
Light not yet diffracted at sample
Resolution not diffraction
Limited, no diffraction,
Limited by aperture size
Aperture Diameter<<Wavelength: 50-100 nm
Aperture-Surface distance<<Wavelength: 20 nm
Probes made from pulled fiber-optics
Experimental Geometries with Fiber-based Probes
trans
epi
•Transmission mode most common (far-field collection)
•Epi-illumination good for two-photon excitation
•Far-field excitation, Near field Collection mode good for SHG
(not shown here)
Fabrication of Tapered Fiber tips:
cannot with standard pipette puller for electrophysiology
CO2 Laser
Pull-solenoid
Pull down to 30-100 nm diameter
Very fragile, fabrication not highly reproducible
EM of Uncoated Tip
Hallen lab, NC State
Uncoated tips do not confine light well
for one photon excitation
Good for NLO modes (intrinsic peak power confinement)
Much higher transmission than coated tips
Coating tips with
Evaporated aluminum
Coating confines light
Rotate at magic angle
For even coverage
Bell Jar
Hallen lab, NC State
Signal Strength vs Resolution
Resolution only depends on aperture, not wavelength
Theoretical: 1/r6 scaling
50 nm practical limit:
106 throughout loss of laser
Hallen lab, NC State
Scanning Probe Feedback Mechanism:
AFM and NSOM same implementation
Need constant tip-specimen distance for near-field
Measuring forces
Use second NIR laser and 2-4
Sectored position sensitive diode
Probe has mirror on top
Experimental Geometry with AFM type Feedback
Tapered fibers use same
Feedback as AFM
Control piezo for
Axial control
Nanonics Design
Sits on
Inverted
Microscope
Far-field
collection
Nonlinear excitation and NSOM with probe collection
Use uncoated probes:
•Higher efficiency
•Metals can interact with
Strong laser field,
perturb sample
(e.g. quench fluorescence)
Confinement from NLO
Don’t need coating
Far-field excitation,
NSOM collection
Saykally, J. Phys. Chem. B, (2002)
Shear force (topography), transmission NSOM, and
fluorescence NSOM images of a phase separated polymer
blend sample (NIST)
Limitations
• Shallow depth of view.
• Weak signal
• Very difficult to work on cells, or other soft
samples
• Complex contrast mechanism – image
interpretation not always straightforward
• Scanning speed unlikely to see much
improvement
Practical Concerns
↑
Hallen lab, NC State
- Coating can have small pinholes:
Loss of confinement
- Easily damaged in experiment
Aperture vs Apertureless NSOM
Principle of the Apertureless NSOM
Sharp tip of a electric conductor enhance (condense)
the local electric field.
Raman spectrum (SERRS) of Rh6G with and without AFM tip
Apertureless NSOM Probes
III. STED
Stimulated Emission Rate:
Absorption Rate:
-σ12FN1
Absorption
Cross-Section
Units → cm2
Number of atoms or
molecules in lower
energy level (Unit:
per cm3)
Photon Flux
Units → #/cm2sec
-σ21FN2
Stimulated emission
Cross-Section
Units → cm2
(typical value ~ 10-19
to 10-18 cm2)
σ12 = σ21
Number of atoms or
molecules in lower
energy level (Unit:
per cm3)
Photon Flux
Units → #/cm2sec
Stimulated Emission Depletion (STED)
Drive down to ground state with second “dump”pulse,
Before molecule can fluoresce
Quench fluorescence and Combine with spatial control
to make “donut”, achieve super-resolution in 3D (unlike NSOM)
Setup
STED Experimental Setup and PSF’s
100 nm
Axial and lateral
PSFs
Need two tunable lasers,
Overlapped spatially, temporally
Hell et al
And synchronized
Resolution increase with STED
microscopy applied to synaptic
vesicles
The real physical reason for the
breaking of the diffraction barrier is
not the fact that fluorescence is
inhibited, but the saturation (of the
fluorescence reduction).
Fluorescence reduction alone would
not help since the focused STEDpulse is also diffraction-limited.
RESOLFT: Extending the STED Idea
• Triplet – Singlet
• PAFP
• Photochromic Dye
4-pi Microscopy
4pi Microscopy: Improves Axial Resolution
Excite high NA top and bottom
Standing Wave interference makes sidelobes
Need deconvolution to remove sidelobes from image
The resolution is largely given by the extent of the effective 4Pispot, which is 3-5 times sharper than the spot of a regular
confocal microscope
~100 nm Axial Resolution
2-photon confocal
2-photon 4pi
2-photon 4pi
With sidelobes
gone
4-pi scope readily works for cell imaging
GFP-labeled mitochondrial compartment of live
Saccharomyces cerevisiae.
Combine STED with 4 pi for improved 3D resolution
Over STED or 4Pi alone
30 nm Resolution: 15 fold improvement over Diffraction Limit
Comparing to Confocal