EasySTED segmented waveplate

[George McNamara]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Confocal (and nanoscopes) Listserv,

I have been researching improvements in STED fluorescence nanoscopy. The
biggest improvements I came across were (order not meant to imply criticality
or price) are:

* 3D-EasySTED segmented waveplate(s) ... vs current Leica (2D) Vortex
Phase Filter.
* Pulsed depletion (i.e. 80 MHz Ti:Sa instead of continuous wave (CW) laser.
Note that this ideally by around 100 to 300 ps duration and start some number
of picoseconds after the excitation pulse has left (needed to give excited
state to reach the lowest S1 state).
* Time gated (g-STED, aka FLIM STED) instead of continuous detection (note
that this requires pulsed excitation, and fits well with pulsed depletion). See
Vicidomini et al 2011 Nat Methods for recent publication (note that Fig 1 is on
fluorescent nanodiamonds - not a typical fluorophore. Look at the figures for
those using Alexa Fluor 488, etc).
* New Leica HyD detector(s). Sadly the CW-STED demo system that just left
Miami had only one HyD.
* Ultra stable vibration isolation table (unlike the demo system).

EasySTED was published in 2010 by Reuss et al in Optics Express (open
access at http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-2-1049 ).

I found online Matthias Reuss's PhD dissertation (http://archiv.ub.uni-
heidelberg.de/volltextserver/volltexte/2011/11539/pdf/Reuss_Dissertation.pdf ),
 which has lots of cool stuff (though temperature wise does not come close to
A. Giske's cryoSTED dissertation, http://archiv.ub.uni-
heidelberg.de/volltextserver/volltexte/2008/7969/pdf/CryoSTED_microscopy_Ph
DThesis_agiske.pdf ).

Reuss's dissertation has stuff from the 2010 Optics Express article, a little bit
on time-gating STED but the item that caught my eye was 3D-EasySTED.

I have done more research on EasySTED and in exchanging emails with Goetz
Zinner at the manufacturer (www.b-halle.de), Goetz wrote that:

"An interesting result I found is that a 6.5 lambda plate at 710 nm would have
7.5 lambda at 622 nm, 8.5 lambda 554 nm and 9.5 lambda 501 nm. This would
enable to use only one mode converter for all four depletion wavelengths."

Transmitting any of the half lambda wavelengths through a single segmented
waveplate produces STED depletion doughnuts at each of these wavelengths.
From Reuss's dissertation it seems (to me) that a tolerance of a few percent
(ex. 2% would be 10 nm at 501 nm) would still produce an acceptable
doughnut.

But wait, there is more - the same segmented waveplate can also tranmit
whole lambda wavelengths, and these can be calculated using the data
available from a paper Goetz mentioned in his email:

G. Ghosh 1999 Dispersion-equation coefficients for the refractive index and
birefringence of calcite and quartz crystals.  Opt Commun 163: 95–102
(available online at www.sciencedirect.com).

Using Goetz'z equation (actually, the standard Lambda retardance equation -
see Reuss) and Ghosh's data, I made an Excel file that presents Goetz's
results. Besides confirming that one segmented waveplate can have multiple
half and whole lambda's, one of the coolest features is that (interpolating from
Ghosh's data), 467.5 and 890 um thick quartz segmented waveplates produce
whole and/or half Lambda retardance values for a variety of thicknesses that
are round number of micrometers in thickness (200, 300, 400 um, for
example).

For those interested in EasySTED, I am happy to send anyone the Excel file.
Best to contact me through my office email address,
[hidden email].

Sincerely,

George

[POUVREAU SANDRINE]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear confocal listserv.
I am taking advantage of the fact that the STED topic has been started to raise a concern that a lot of physiologists colleagues have. I am trying to do live cell organelles imaging with STED system (in neurons), the organelles being stained with fluorescent proteins (EYFP). The problem is that the level of staining required with our STED set up seems to be incompatible with live cell physiology. Indeed, expression levels that show very bright staining in confocal are not bright enough for our STED system, to the point we cannot do any picture.  Higher expression levels will just damage the organelle, or induce leak of the fluorophore into the cytosol, so this is not an option.  It also seems to me that, although the signal is very weak, the intensity of the excitation is very high, which may cause further damage to the organelles.  So far, my colleagues have been able to do nice images of the cells (neurons) with a cytosolic volume label (fluorescent protein), although they require high staining too, so it seems that our system is just fine but we cannot work with small volumes. Is this a common problem with STED systems? We are using filters in the detection path, no AOBS.
Thanks
Have a nice day
Sandrine


________________________________________
De : Confocal Microscopy List [[hidden email]] de la part de George McNamara [[hidden email]]
Date d'envoi : samedi 10 septembre 2011 23:59
À : [hidden email]
Objet : EasySTED segmented waveplate

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Confocal (and nanoscopes) Listserv,

I have been researching improvements in STED fluorescence nanoscopy. The
biggest improvements I came across were (order not meant to imply criticality
or price) are:

* 3D-EasySTED segmented waveplate(s) ... vs current Leica (2D) Vortex
Phase Filter.
* Pulsed depletion (i.e. 80 MHz Ti:Sa instead of continuous wave (CW) laser.
Note that this ideally by around 100 to 300 ps duration and start some number
of picoseconds after the excitation pulse has left (needed to give excited
state to reach the lowest S1 state).
* Time gated (g-STED, aka FLIM STED) instead of continuous detection (note
that this requires pulsed excitation, and fits well with pulsed depletion). See
Vicidomini et al 2011 Nat Methods for recent publication (note that Fig 1 is on
fluorescent nanodiamonds - not a typical fluorophore. Look at the figures for
those using Alexa Fluor 488, etc).
* New Leica HyD detector(s). Sadly the CW-STED demo system that just left
Miami had only one HyD.
* Ultra stable vibration isolation table (unlike the demo system).

EasySTED was published in 2010 by Reuss et al in Optics Express (open
access at http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-2-1049 ).

I found online Matthias Reuss's PhD dissertation (http://archiv.ub.uni-
heidelberg.de/volltextserver/volltexte/2011/11539/pdf/Reuss_Dissertation.pdf ),
 which has lots of cool stuff (though temperature wise does not come close to
A. Giske's cryoSTED dissertation, http://archiv.ub.uni-
heidelberg.de/volltextserver/volltexte/2008/7969/pdf/CryoSTED_microscopy_Ph
DThesis_agiske.pdf ).

Reuss's dissertation has stuff from the 2010 Optics Express article, a little bit
on time-gating STED but the item that caught my eye was 3D-EasySTED.

I have done more research on EasySTED and in exchanging emails with Goetz
Zinner at the manufacturer (www.b-halle.de), Goetz wrote that:

"An interesting result I found is that a 6.5 lambda plate at 710 nm would have
7.5 lambda at 622 nm, 8.5 lambda 554 nm and 9.5 lambda 501 nm. This would
enable to use only one mode converter for all four depletion wavelengths."

Transmitting any of the half lambda wavelengths through a single segmented
waveplate produces STED depletion doughnuts at each of these wavelengths.
From Reuss's dissertation it seems (to me) that a tolerance of a few percent
(ex. 2% would be 10 nm at 501 nm) would still produce an acceptable
doughnut.

But wait, there is more - the same segmented waveplate can also tranmit
whole lambda wavelengths, and these can be calculated using the data
available from a paper Goetz mentioned in his email:

G. Ghosh 1999 Dispersion-equation coefficients for the refractive index and
birefringence of calcite and quartz crystals.  Opt Commun 163: 95–102
(available online at www.sciencedirect.com).

Using Goetz'z equation (actually, the standard Lambda retardance equation -
see Reuss) and Ghosh's data, I made an Excel file that presents Goetz's
results. Besides confirming that one segmented waveplate can have multiple
half and whole lambda's, one of the coolest features is that (interpolating from
Ghosh's data), 467.5 and 890 um thick quartz segmented waveplates produce
whole and/or half Lambda retardance values for a variety of thicknesses that
are round number of micrometers in thickness (200, 300, 400 um, for
example).

For those interested in EasySTED, I am happy to send anyone the Excel file.
Best to contact me through my office email address,
[hidden email].

Sincerely,

George

[George McNamara]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Sandrine,

Having just hosted a three week CW-STED demo at UM, I have some
suggestions. I am assuming you are on a Leica CW-STED system.

Short term:

* No need for full power depletion: I suggest you start by comparing
confocal (STED power slider at 0% and red/black shutter button
unselected = black) to low power STED. Since LAS AF allows you to make
lots of scan tracks, I suggest sequential scan - frame - mode
(excitation laser same for all tracks): confocal, STED 10%, STED 25%,
STED 75%, STED 100%, confocal.

* There is a clear interaction between excitation and depletion: On a
new field of view, same as above, but with less excitation power (for
example, 25% of your usual power).

* Do STED first: Jonathan Boyd of Leica pointed out during our demo that
the point of being on a STED machine is to do STED. I had typically done
confocal frame scan then STED. Jonathan convinced me that doing STED
first makes sense.

* Go fast: 1400 Hz (min zoom 6, I am referring to the high resolution
scanner) may produce better STED and confocal data than slower speeds.
If you have not been evaluating different scan speeds, your Leica
training person did a poor job. Jonathan also pointed out that on the
SP5 high resolution scanner can operate at 600 Hz at 1x zoom, so
everyone should switch from 400 Hz default to 600 Hz (of course from a
core billing perspective, we could make more money by getting everyone
to use 100 Hz). With respect to 600 Hz, hopefully Leica will add this as
a drop down list item in near future LAS AF release. Jonathan also
showed me that resonant scan mode can produce better data. For example
10 frame averaging (this was on our SP5, no HyD, so no photon count
accumulation) at 8000 Hz produced better data than 1 scan at 800 Hz.

* Get the riboflavin out ... Bogdanov et al (see Evrogen DMEMgfp),
Matsuda ... Sedat (PLoS One), and more recently a visit from Essen
Bioscience reps (thanks Daniel), show that, respectively, riboflavin in
culture media causes EGFP and related FPs - ex. EYFP - to photobleach
~8x faster, riboflavin greatly facilitates photoswtiching of EGFP to red
fluorescence (they used this for PALM on their OMX), and riboflavin is
by far the major source of green autofluorescence in tissue culture
media. I was surprised that the last was not Phenol red, but I will
still recommend "no riboflavin (for 24 hours - see Bogdanov), no Phenol
red, if you want to use 25 mM Hepes make sure it plus light does not
mess up your cell's physiology.

* ROXS like live cell media ... see Kasper et al 2010 Small, then
references cited within and newer related papers (many form Tinnefeld
and Sauer) for live cell equivalents.

The bottom line of the above is that if you can STED to say 140 nm XY
resolution, a 1/3rd improvement over confocal (214 nm for pinhole 1.0,
lambda = 500 nm, NA = 1.4) use it and do biology.

* Long term: if you do not have HyD detector(s), get a demonstration. Be
sure to use accumulation mode, not averaging. HyD's have substantially
higher quantum efficiency than the standard Leica PMT's, and lower
electronic noise (or at least the use of photon counting mode hides the
noise up front - same thing from a data perspective). For me, operating
with the data in photon counts makes understanding the data much easier.
At times 20 photon maximum signal in a few pixel size object (single
molecule?) is enough to get useful results, since the noise floor is
typically ~0.5 photons in the same image.

* Long term: EYFP is pH and chloride sensitive. You should eventually
switch to mVenus, unless you are trying to study pH and/or chloride ion
changes in your organelles (in which case you probably need a second
color for ratio ... or FLIM). keep an eye on the literature in case an
even better yellow FP comes out.

* Very long term (start planning now to find money): FLIM HyD(s) for
g-STED (and pulsed laser[s]). Note that the current HyD (not sure if it
is the detector module, electronics or both - ask your Leica rep), are
photon counting but lack something for FLIM timing. Leica should have
these out in 2012. If you do not have the other FLIM hardware (the
Leica/PicoQuant SMD module), start looking for money for that too. If
leica (and PicoQuant) nail this combination, g-STED will not only be
able to gate out the "pedestal" photons (see FLIM section of Reuss
dissertation or Fig 1 of Vicidomini et al), but also any other sources
of "zero time" photons.

* Unknown term: The current Leica Vortex Phase Filters improve in XY
only. The STED point spread function is skinnier than confocal (I think
of STED as a pencil, confocal as an American football), so there is some
improvement in axial resolution (10%?), but either adding an annular
filter (2nd depletion path) or 3D-EasySTED points the way to where Leica
is going. Hopefully Leica will simply have the field service engineers
replace all customer 2D-VFPs with 3D-? as part of routine service.

***

Additional point to my previous post ... A major goal for our STED grant
proposal(s) is to get the money for an instrument to enable our users to
image and count single molecules in or on cells, with "the usual colors"
that our user base uses now. For example, Alexa Fluor 488, Alexa Fluor
568, and (occasionally) Alexa Fluor 647, or EGFP, mCherry, possbile 3rd
color (there are now plenty of hybrid small organic molecule - protein
targeting sequences, such as Halo-Tag, Snap-Tag, etc, that these could
be alternatives to standard FPs ... I was also pleased with one user's
Panomics breanched DNA detection RNA FISH result and saw a nice paper
recently in Biotechniques reminding me about proximity ligation assay -
rolling circle amplification "prxoimity immunofluorescence"). By single
molecule, I do not necessarily mean a single Alexa Fluor 488 dye
molecule. I want to provide to my users the ability to image single
antigens, typically one antigen, one monoclonal primary antibody, ~3
secondary antibodies, each with ~4 fluorophores. With optimized mounting
media (ex. ROXS - Kasper et al 2010, also shows a clear benefit in
confocal), optimized multicolor STED setup (see below and Reuss 2011
dissertation and g-STED, etc), I expect to get to get 50x50x200 nm
optical resolution from multiple channels not just green for CW-STED).
If we count to 25 photons, the precision localization is denominator
sqrt(25) = 5, which means 10x10x40 nm. Since the indirect
immunofluorescence layers are >10 nm diameter, no need to count too far
past 25 photons. Hopefully Leica (maybe with some help from PicoQuant)
will get P.L. into STED sooner than later (perhaps as simple as a port
from GSD). Of course the Ag-1stAb-2ndAbs-~10 fluorophores is exactly
what everyone has been using for immunofluorescence for a long time.
Improvement in mounting media is doable (perhaps Invitrogen will come
out with Prolong ROXS with out DAPI?) and will help users evaluate their
experiments in their own labs or on our core microscopes.



Best wishes,

George



On 9/11/2011 6:04 AM, POUVREAU SANDRINE wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Dear confocal listserv.
> I am taking advantage of the fact that the STED topic has been started to raise a concern that a lot of physiologists colleagues have. I am trying to do live cell organelles imaging with STED system (in neurons), the organelles being stained with fluorescent proteins (EYFP). The problem is that the level of staining required with our STED set up seems to be incompatible with live cell physiology. Indeed, expression levels that show very bright staining in confocal are not bright enough for our STED system, to the point we cannot do any picture.  Higher expression levels will just damage the organelle, or induce leak of the fluorophore into the cytosol, so this is not an option.  It also seems to me that, although the signal is very weak, the intensity of the excitation is very high, which may cause further damage to the organelles.  So far, my colleagues have been able to do nice images of the cells (neurons) with a cytosolic volume label (fluorescent protein), although they require high staining too, so it seems that our system is just fine but we cannot work with small volumes. Is this a common problem with STED systems? We are using filters in the detection path, no AOBS.
> Thanks
> Have a nice day
> Sandrine
>
>
> ________________________________________
> De : Confocal Microscopy List [[hidden email]] de la part de George McNamara [[hidden email]]
> Date d'envoi : samedi 10 septembre 2011 23:59
> À : [hidden email]
> Objet : EasySTED segmented waveplate
>
> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Dear Confocal (and nanoscopes) Listserv,
>
> I have been researching improvements in STED fluorescence nanoscopy. The
> biggest improvements I came across were (order not meant to imply criticality
> or price) are:
>
> * 3D-EasySTED segmented waveplate(s) ... vs current Leica (2D) Vortex
> Phase Filter.
> * Pulsed depletion (i.e. 80 MHz Ti:Sa instead of continuous wave (CW) laser.
> Note that this ideally by around 100 to 300 ps duration and start some number
> of picoseconds after the excitation pulse has left (needed to give excited
> state to reach the lowest S1 state).
> * Time gated (g-STED, aka FLIM STED) instead of continuous detection (note
> that this requires pulsed excitation, and fits well with pulsed depletion). See
> Vicidomini et al 2011 Nat Methods for recent publication (note that Fig 1 is on
> fluorescent nanodiamonds - not a typical fluorophore. Look at the figures for
> those using Alexa Fluor 488, etc).
> * New Leica HyD detector(s). Sadly the CW-STED demo system that just left
> Miami had only one HyD.
> * Ultra stable vibration isolation table (unlike the demo system).
>
> EasySTED was published in 2010 by Reuss et al in Optics Express (open
> access at http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-2-1049 ).
>
> I found online Matthias Reuss's PhD dissertation (http://archiv.ub.uni-
> heidelberg.de/volltextserver/volltexte/2011/11539/pdf/Reuss_Dissertation.pdf ),
>   which has lots of cool stuff (though temperature wise does not come close to
> A. Giske's cryoSTED dissertation, http://archiv.ub.uni-
> heidelberg.de/volltextserver/volltexte/2008/7969/pdf/CryoSTED_microscopy_Ph
> DThesis_agiske.pdf ).
>
> Reuss's dissertation has stuff from the 2010 Optics Express article, a little bit
> on time-gating STED but the item that caught my eye was 3D-EasySTED.
>
> I have done more research on EasySTED and in exchanging emails with Goetz
> Zinner at the manufacturer (www.b-halle.de), Goetz wrote that:
>
> "An interesting result I found is that a 6.5 lambda plate at 710 nm would have
> 7.5 lambda at 622 nm, 8.5 lambda 554 nm and 9.5 lambda 501 nm. This would
> enable to use only one mode converter for all four depletion wavelengths."
>
> Transmitting any of the half lambda wavelengths through a single segmented
> waveplate produces STED depletion doughnuts at each of these wavelengths.
> > From Reuss's dissertation it seems (to me) that a tolerance of a few percent
> (ex. 2% would be 10 nm at 501 nm) would still produce an acceptable
> doughnut.
>
> But wait, there is more - the same segmented waveplate can also tranmit
> whole lambda wavelengths, and these can be calculated using the data
> available from a paper Goetz mentioned in his email:
>
> G. Ghosh 1999 Dispersion-equation coefficients for the refractive index and
> birefringence of calcite and quartz crystals.  Opt Commun 163: 95–102
> (available online at www.sciencedirect.com).
>
> Using Goetz'z equation (actually, the standard Lambda retardance equation -
> see Reuss) and Ghosh's data, I made an Excel file that presents Goetz's
> results. Besides confirming that one segmented waveplate can have multiple
> half and whole lambda's, one of the coolest features is that (interpolating from
> Ghosh's data), 467.5 and 890 um thick quartz segmented waveplates produce
> whole and/or half Lambda retardance values for a variety of thicknesses that
> are round number of micrometers in thickness (200, 300, 400 um, for
> example).
>
> For those interested in EasySTED, I am happy to send anyone the Excel file.
> Best to contact me through my office email address,
> [hidden email].
>
> Sincerely,
>
> George
>    

--


George McNamara, PhD
Analytical Imaging Core Facility
University of Miami

[Andreas Bruckbauer]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****


 Hi Sandrine, George,

would structured illumination with EMCCD camera based detection not be more efficient for very waek samples? How does a STED with HyD detectors compare to the OMX in terms of collection efficiency? Translating the signal to number of molecules is very usefull but watch out that molecules can bleach even when they are not contributing to your signal.

best wishes

Andreas


 


 

 

-----Original Message-----
From: George McNamara <[hidden email]>
To: CONFOCALMICROSCOPY <[hidden email]>
Sent: Sun, 11 Sep 2011 18:29
Subject: RE : EasySTED segmented waveplate


*****

To join, leave or search the confocal microscopy listserv, go to:

http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy

*****



Hi Sandrine,



Having just hosted a three week CW-STED demo at UM, I have some

suggestions. I am assuming you are on a Leica CW-STED system.



Short term:



* No need for full power depletion: I suggest you start by comparing

confocal (STED power slider at 0% and red/black shutter button

unselected = black) to low power STED. Since LAS AF allows you to make

lots of scan tracks, I suggest sequential scan - frame - mode

(excitation laser same for all tracks): confocal, STED 10%, STED 25%,

STED 75%, STED 100%, confocal.



* There is a clear interaction between excitation and depletion: On a

new field of view, same as above, but with less excitation power (for

example, 25% of your usual power).



* Do STED first: Jonathan Boyd of Leica pointed out during our demo that

the point of being on a STED machine is to do STED. I had typically done

confocal frame scan then STED. Jonathan convinced me that doing STED

first makes sense.



* Go fast: 1400 Hz (min zoom 6, I am referring to the high resolution

scanner) may produce better STED and confocal data than slower speeds.

If you have not been evaluating different scan speeds, your Leica

training person did a poor job. Jonathan also pointed out that on the

SP5 high resolution scanner can operate at 600 Hz at 1x zoom, so

everyone should switch from 400 Hz default to 600 Hz (of course from a

core billing perspective, we could make more money by getting everyone

to use 100 Hz). With respect to 600 Hz, hopefully Leica will add this as

a drop down list item in near future LAS AF release. Jonathan also

showed me that resonant scan mode can produce better data. For example

10 frame averaging (this was on our SP5, no HyD, so no photon count

accumulation) at 8000 Hz produced better data than 1 scan at 800 Hz.



* Get the riboflavin out ... Bogdanov et al (see Evrogen DMEMgfp),

Matsuda ... Sedat (PLoS One), and more recently a visit from Essen

Bioscience reps (thanks Daniel), show that, respectively, riboflavin in

culture media causes EGFP and related FPs - ex. EYFP - to photobleach

~8x faster, riboflavin greatly facilitates photoswtiching of EGFP to red

fluorescence (they used this for PALM on their OMX), and riboflavin is

by far the major source of green autofluorescence in tissue culture

media. I was surprised that the last was not Phenol red, but I will

still recommend "no riboflavin (for 24 hours - see Bogdanov), no Phenol

red, if you want to use 25 mM Hepes make sure it plus light does not

mess up your cell's physiology.



* ROXS like live cell media ... see Kasper et al 2010 Small, then

references cited within and newer related papers (many form Tinnefeld

and Sauer) for live cell equivalents.



The bottom line of the above is that if you can STED to say 140 nm XY

resolution, a 1/3rd improvement over confocal (214 nm for pinhole 1.0,

lambda = 500 nm, NA = 1.4) use it and do biology.



* Long term: if you do not have HyD detector(s), get a demonstration. Be

sure to use accumulation mode, not averaging. HyD's have substantially

higher quantum efficiency than the standard Leica PMT's, and lower

electronic noise (or at least the use of photon counting mode hides the

noise up front - same thing from a data perspective). For me, operating

with the data in photon counts makes understanding the data much easier.

At times 20 photon maximum signal in a few pixel size object (single

molecule?) is enough to get useful results, since the noise floor is

typically ~0.5 photons in the same image.



* Long term: EYFP is pH and chloride sensitive. You should eventually

switch to mVenus, unless you are trying to study pH and/or chloride ion

changes in your organelles (in which case you probably need a second

color for ratio ... or FLIM). keep an eye on the literature in case an

even better yellow FP comes out.



* Very long term (start planning now to find money): FLIM HyD(s) for

g-STED (and pulsed laser[s]). Note that the current HyD (not sure if it

is the detector module, electronics or both - ask your Leica rep), are

photon counting but lack something for FLIM timing. Leica should have

these out in 2012. If you do not have the other FLIM hardware (the

Leica/PicoQuant SMD module), start looking for money for that too. If

leica (and PicoQuant) nail this combination, g-STED will not only be

able to gate out the "pedestal" photons (see FLIM section of Reuss

dissertation or Fig 1 of Vicidomini et al), but also any other sources

of "zero time" photons.



* Unknown term: The current Leica Vortex Phase Filters improve in XY

only. The STED point spread function is skinnier than confocal (I think

of STED as a pencil, confocal as an American football), so there is some

improvement in axial resolution (10%?), but either adding an annular

filter (2nd depletion path) or 3D-EasySTED points the way to where Leica

is going. Hopefully Leica will simply have the field service engineers

replace all customer 2D-VFPs with 3D-? as part of routine service.



***



Additional point to my previous post ... A major goal for our STED grant

proposal(s) is to get the money for an instrument to enable our users to

image and count single molecules in or on cells, with "the usual colors"

that our user base uses now. For example, Alexa Fluor 488, Alexa Fluor

568, and (occasionally) Alexa Fluor 647, or EGFP, mCherry, possbile 3rd

color (there are now plenty of hybrid small organic molecule - protein

targeting sequences, such as Halo-Tag, Snap-Tag, etc, that these could

be alternatives to standard FPs ... I was also pleased with one user's

Panomics breanched DNA detection RNA FISH result and saw a nice paper

recently in Biotechniques reminding me about proximity ligation assay -

rolling circle amplification "prxoimity immunofluorescence"). By single

molecule, I do not necessarily mean a single Alexa Fluor 488 dye

molecule. I want to provide to my users the ability to image single

antigens, typically one antigen, one monoclonal primary antibody, ~3

secondary antibodies, each with ~4 fluorophores. With optimized mounting

media (ex. ROXS - Kasper et al 2010, also shows a clear benefit in

confocal), optimized multicolor STED setup (see below and Reuss 2011

dissertation and g-STED, etc), I expect to get to get 50x50x200 nm

optical resolution from multiple channels not just green for CW-STED).

If we count to 25 photons, the precision localization is denominator

sqrt(25) = 5, which means 10x10x40 nm. Since the indirect

immunofluorescence layers are >10 nm diameter, no need to count too far

past 25 photons. Hopefully Leica (maybe with some help from PicoQuant)

will get P.L. into STED sooner than later (perhaps as simple as a port

from GSD). Of course the Ag-1stAb-2ndAbs-~10 fluorophores is exactly

what everyone has been using for immunofluorescence for a long time.

Improvement in mounting media is doable (perhaps Invitrogen will come

out with Prolong ROXS with out DAPI?) and will help users evaluate their

experiments in their own labs or on our core microscopes.







Best wishes,



George







On 9/11/2011 6:04 AM, POUVREAU SANDRINE wrote:

> *****

> To join, leave or search the confocal microscopy listserv, go to:

> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy

> *****

>

> Dear confocal listserv.

> I am taking advantage of the fact that the STED topic has been started to

raise a concern that a lot of physiologists colleagues have. I am trying to do

live cell organelles imaging with STED system (in neurons), the organelles being

stained with fluorescent proteins (EYFP). The problem is that the level of

staining required with our STED set up seems to be incompatible with live cell

physiology. Indeed, expression levels that show very bright staining in confocal

are not bright enough for our STED system, to the point we cannot do any

picture.  Higher expression levels will just damage the organelle, or induce

leak of the fluorophore into the cytosol, so this is not an option.  It also

seems to me that, although the signal is very weak, the intensity of the

excitation is very high, which may cause further damage to the organelles.  So

far, my colleagues have been able to do nice images of the cells (neurons) with

a cytosolic volume label (fluorescent protein), although they require high

staining too, so it seems that our system is just fine but we cannot work with

small volumes. Is this a common problem with STED systems? We are using filters

in the detection path, no AOBS.

> Thanks

> Have a nice day

> Sandrine

>

>

> ________________________________________

> De : Confocal Microscopy List [[hidden email]] de la part de

George McNamara [[hidden email]]

> Date d'envoi : samedi 10 septembre 2011 23:59

> À : [hidden email]

> Objet : EasySTED segmented waveplate

>

> *****

> To join, leave or search the confocal microscopy listserv, go to:

> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy

> *****

>

> Dear Confocal (and nanoscopes) Listserv,

>

> I have been researching improvements in STED fluorescence nanoscopy. The

> biggest improvements I came across were (order not meant to imply criticality

> or price) are:

>

> * 3D-EasySTED segmented waveplate(s) ... vs current Leica (2D) Vortex

> Phase Filter.

> * Pulsed depletion (i.e. 80 MHz Ti:Sa instead of continuous wave (CW) laser.

> Note that this ideally by around 100 to 300 ps duration and start some number

> of picoseconds after the excitation pulse has left (needed to give excited

> state to reach the lowest S1 state).

> * Time gated (g-STED, aka FLIM STED) instead of continuous detection (note

> that this requires pulsed excitation, and fits well with pulsed depletion).

See

> Vicidomini et al 2011 Nat Methods for recent publication (note that Fig 1 is

on

> fluorescent nanodiamonds - not a typical fluorophore. Look at the figures for

> those using Alexa Fluor 488, etc).

> * New Leica HyD detector(s). Sadly the CW-STED demo system that just left

> Miami had only one HyD.

> * Ultra stable vibration isolation table (unlike the demo system).

>

> EasySTED was published in 2010 by Reuss et al in Optics Express (open

> access at http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-2-1049 ).

>

> I found online Matthias Reuss's PhD dissertation (http://archiv.ub.uni-

> heidelberg.de/volltextserver/volltexte/2011/11539/pdf/Reuss_Dissertation.pdf

),

>   which has lots of cool stuff (though temperature wise does not come close to

> A. Giske's cryoSTED dissertation, http://archiv.ub.uni-

> heidelberg.de/volltextserver/volltexte/2008/7969/pdf/CryoSTED_microscopy_Ph

> DThesis_agiske.pdf ).

>

> Reuss's dissertation has stuff from the 2010 Optics Express article, a little

bit

> on time-gating STED but the item that caught my eye was 3D-EasySTED.

>

> I have done more research on EasySTED and in exchanging emails with Goetz

> Zinner at the manufacturer (www.b-halle.de), Goetz wrote that:

>

> "An interesting result I found is that a 6.5 lambda plate at 710 nm would have

> 7.5 lambda at 622 nm, 8.5 lambda 554 nm and 9.5 lambda 501 nm. This would

> enable to use only one mode converter for all four depletion wavelengths."

>

> Transmitting any of the half lambda wavelengths through a single segmented

> waveplate produces STED depletion doughnuts at each of these wavelengths.

> > From Reuss's dissertation it seems (to me) that a tolerance of a few percent

> (ex. 2% would be 10 nm at 501 nm) would still produce an acceptable

> doughnut.

>

> But wait, there is more - the same segmented waveplate can also tranmit

> whole lambda wavelengths, and these can be calculated using the data

> available from a paper Goetz mentioned in his email:

>

> G. Ghosh 1999 Dispersion-equation coefficients for the refractive index and

> birefringence of calcite and quartz crystals.  Opt Commun 163: 95–102

> (available online at www.sciencedirect.com).

>

> Using Goetz'z equation (actually, the standard Lambda retardance equation -

> see Reuss) and Ghosh's data, I made an Excel file that presents Goetz's

> results. Besides confirming that one segmented waveplate can have multiple

> half and whole lambda's, one of the coolest features is that (interpolating

from

> Ghosh's data), 467.5 and 890 um thick quartz segmented waveplates produce

> whole and/or half Lambda retardance values for a variety of thicknesses that

> are round number of micrometers in thickness (200, 300, 400 um, for

> example).

>

> For those interested in EasySTED, I am happy to send anyone the Excel file.

> Best to contact me through my office email address,

> [hidden email].

>

> Sincerely,

>

> George

>    





--





George McNamara, PhD

Analytical Imaging Core Facility

University of Miami


 

[George McNamara]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Andreas,

Operating a back illuminated EMCCD in EM mode cuts the effective quantum
efficiency in half (assuming the user knows to turn on / leave on the EM
gain register). At our OMX demo in March, API did not show me any data
in photon counting mode (the demo mostly used an EMCCD(s) V3, we wrote
the S10 for a 5 megapixel scientific CMOS V4 to reduce price and get a
bigger field of view).

Bleaching during acquisition - check out the ROXS paper (Kasper et al
2010, Small). This and similar reagents - including some for live cell
imaging - should improve STED, OMX (and the other 3D-SIM), and confocal,
and widefield imaging. Of course ultra-stable fluorophores would be
problematic for PALm/STORM/Monet (the user could always change solutions
for the latter modes).

One of the advantages of STED over confocal is the smaller volume being
imaged. At 50x50x200 nm (g-EastSTED, see Blom et al 2011 MRT, Vicidomini
et al 2011 Nat Meth, Reuss 2011 PhD dissertation) STED is operating with
an (approximately) 4x4x3 = 48 fold smaller volume than confocal. Smaller
volume means less autofluorescence from the specimen, as well as a much
sparser distribution of wanted fluorophores than confocal. Of course it
implies scanning either for longer, or smaller volume, or both, to get
done. Hell's lab has published on RESCUE-STED, something like Mander's
CLEM, to presumably get done sooner. I believe the RESCUE-STED paper
used a 400 usec pixel dwell time and 15x15 nm pixel size, and a 4x4 um
field of view (TEMesque field of view). At our Miami CW-STED demo we
often used 3072x3072 pixels, 25x25 nm pixel size, and 4 to 10 line or
frame accumulation, 75x75 um field of view, getting done in 20 seconds
per 2D image (single channel).

Enjoy,

George
p.s. please convince NIH to fund both our S10's. If we get both OMX V4
(sCMOS) and STED, you are welcome to visit Miami to do a head to head
comparison.

On 9/12/2011 5:41 AM, Andreas Bruckbauer wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
>
>   Hi Sandrine, George,
>
> would structured illumination with EMCCD camera based detection not be more efficient for very waek samples? How does a STED with HyD detectors compare to the OMX in terms of collection efficiency? Translating the signal to number of molecules is very usefull but watch out that molecules can bleach even when they are not contributing to your signal.
>
> best wishes
>
> Andreas
>
>
>
>
>
>
>
>
>
> -----Original Message-----
> From: George McNamara<[hidden email]>
> To: CONFOCALMICROSCOPY<[hidden email]>
> Sent: Sun, 11 Sep 2011 18:29
> Subject: RE : EasySTED segmented waveplate
>
>
> *****
>
> To join, leave or search the confocal microscopy listserv, go to:
>
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>
> *****
>
>
>
> Hi Sandrine,
>
>
>
> Having just hosted a three week CW-STED demo at UM, I have some
>
> suggestions. I am assuming you are on a Leica CW-STED system.
>
>
>
> Short term:
>
>
>
> * No need for full power depletion: I suggest you start by comparing
>
> confocal (STED power slider at 0% and red/black shutter button
>
> unselected = black) to low power STED. Since LAS AF allows you to make
>
> lots of scan tracks, I suggest sequential scan - frame - mode
>
> (excitation laser same for all tracks): confocal, STED 10%, STED 25%,
>
> STED 75%, STED 100%, confocal.
>
>
>
> * There is a clear interaction between excitation and depletion: On a
>
> new field of view, same as above, but with less excitation power (for
>
> example, 25% of your usual power).
>
>
>
> * Do STED first: Jonathan Boyd of Leica pointed out during our demo that
>
> the point of being on a STED machine is to do STED. I had typically done
>
> confocal frame scan then STED. Jonathan convinced me that doing STED
>
> first makes sense.
>
>
>
> * Go fast: 1400 Hz (min zoom 6, I am referring to the high resolution
>
> scanner) may produce better STED and confocal data than slower speeds.
>
> If you have not been evaluating different scan speeds, your Leica
>
> training person did a poor job. Jonathan also pointed out that on the
>
> SP5 high resolution scanner can operate at 600 Hz at 1x zoom, so
>
> everyone should switch from 400 Hz default to 600 Hz (of course from a
>
> core billing perspective, we could make more money by getting everyone
>
> to use 100 Hz). With respect to 600 Hz, hopefully Leica will add this as
>
> a drop down list item in near future LAS AF release. Jonathan also
>
> showed me that resonant scan mode can produce better data. For example
>
> 10 frame averaging (this was on our SP5, no HyD, so no photon count
>
> accumulation) at 8000 Hz produced better data than 1 scan at 800 Hz.
>
>
>
> * Get the riboflavin out ... Bogdanov et al (see Evrogen DMEMgfp),
>
> Matsuda ... Sedat (PLoS One), and more recently a visit from Essen
>
> Bioscience reps (thanks Daniel), show that, respectively, riboflavin in
>
> culture media causes EGFP and related FPs - ex. EYFP - to photobleach
>
> ~8x faster, riboflavin greatly facilitates photoswtiching of EGFP to red
>
> fluorescence (they used this for PALM on their OMX), and riboflavin is
>
> by far the major source of green autofluorescence in tissue culture
>
> media. I was surprised that the last was not Phenol red, but I will
>
> still recommend "no riboflavin (for 24 hours - see Bogdanov), no Phenol
>
> red, if you want to use 25 mM Hepes make sure it plus light does not
>
> mess up your cell's physiology.
>
>
>
> * ROXS like live cell media ... see Kasper et al 2010 Small, then
>
> references cited within and newer related papers (many form Tinnefeld
>
> and Sauer) for live cell equivalents.
>
>
>
> The bottom line of the above is that if you can STED to say 140 nm XY
>
> resolution, a 1/3rd improvement over confocal (214 nm for pinhole 1.0,
>
> lambda = 500 nm, NA = 1.4) use it and do biology.
>
>
>
> * Long term: if you do not have HyD detector(s), get a demonstration. Be
>
> sure to use accumulation mode, not averaging. HyD's have substantially
>
> higher quantum efficiency than the standard Leica PMT's, and lower
>
> electronic noise (or at least the use of photon counting mode hides the
>
> noise up front - same thing from a data perspective). For me, operating
>
> with the data in photon counts makes understanding the data much easier.
>
> At times 20 photon maximum signal in a few pixel size object (single
>
> molecule?) is enough to get useful results, since the noise floor is
>
> typically ~0.5 photons in the same image.
>
>
>
> * Long term: EYFP is pH and chloride sensitive. You should eventually
>
> switch to mVenus, unless you are trying to study pH and/or chloride ion
>
> changes in your organelles (in which case you probably need a second
>
> color for ratio ... or FLIM). keep an eye on the literature in case an
>
> even better yellow FP comes out.
>
>
>
> * Very long term (start planning now to find money): FLIM HyD(s) for
>
> g-STED (and pulsed laser[s]). Note that the current HyD (not sure if it
>
> is the detector module, electronics or both - ask your Leica rep), are
>
> photon counting but lack something for FLIM timing. Leica should have
>
> these out in 2012. If you do not have the other FLIM hardware (the
>
> Leica/PicoQuant SMD module), start looking for money for that too. If
>
> leica (and PicoQuant) nail this combination, g-STED will not only be
>
> able to gate out the "pedestal" photons (see FLIM section of Reuss
>
> dissertation or Fig 1 of Vicidomini et al), but also any other sources
>
> of "zero time" photons.
>
>
>
> * Unknown term: The current Leica Vortex Phase Filters improve in XY
>
> only. The STED point spread function is skinnier than confocal (I think
>
> of STED as a pencil, confocal as an American football), so there is some
>
> improvement in axial resolution (10%?), but either adding an annular
>
> filter (2nd depletion path) or 3D-EasySTED points the way to where Leica
>
> is going. Hopefully Leica will simply have the field service engineers
>
> replace all customer 2D-VFPs with 3D-? as part of routine service.
>
>
>
> ***
>
>
>
> Additional point to my previous post ... A major goal for our STED grant
>
> proposal(s) is to get the money for an instrument to enable our users to
>
> image and count single molecules in or on cells, with "the usual colors"
>
> that our user base uses now. For example, Alexa Fluor 488, Alexa Fluor
>
> 568, and (occasionally) Alexa Fluor 647, or EGFP, mCherry, possbile 3rd
>
> color (there are now plenty of hybrid small organic molecule - protein
>
> targeting sequences, such as Halo-Tag, Snap-Tag, etc, that these could
>
> be alternatives to standard FPs ... I was also pleased with one user's
>
> Panomics breanched DNA detection RNA FISH result and saw a nice paper
>
> recently in Biotechniques reminding me about proximity ligation assay -
>
> rolling circle amplification "prxoimity immunofluorescence"). By single
>
> molecule, I do not necessarily mean a single Alexa Fluor 488 dye
>
> molecule. I want to provide to my users the ability to image single
>
> antigens, typically one antigen, one monoclonal primary antibody, ~3
>
> secondary antibodies, each with ~4 fluorophores. With optimized mounting
>
> media (ex. ROXS - Kasper et al 2010, also shows a clear benefit in
>
> confocal), optimized multicolor STED setup (see below and Reuss 2011
>
> dissertation and g-STED, etc), I expect to get to get 50x50x200 nm
>
> optical resolution from multiple channels not just green for CW-STED).
>
> If we count to 25 photons, the precision localization is denominator
>
> sqrt(25) = 5, which means 10x10x40 nm. Since the indirect
>
> immunofluorescence layers are>10 nm diameter, no need to count too far
>
> past 25 photons. Hopefully Leica (maybe with some help from PicoQuant)
>
> will get P.L. into STED sooner than later (perhaps as simple as a port
>
> from GSD). Of course the Ag-1stAb-2ndAbs-~10 fluorophores is exactly
>
> what everyone has been using for immunofluorescence for a long time.
>
> Improvement in mounting media is doable (perhaps Invitrogen will come
>
> out with Prolong ROXS with out DAPI?) and will help users evaluate their
>
> experiments in their own labs or on our core microscopes.
>
>
>
>
>
>
>
> Best wishes,
>
>
>
> George
>
>
>
>
>
>
>
> On 9/11/2011 6:04 AM, POUVREAU SANDRINE wrote:
>
>    
>> *****
>>      
>    
>> To join, leave or search the confocal microscopy listserv, go to:
>>      
>    
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>      
>    
>> *****
>>      
>    
>>      
>    
>> Dear confocal listserv.
>>      
>    
>> I am taking advantage of the fact that the STED topic has been started to
>>      
> raise a concern that a lot of physiologists colleagues have. I am trying to do
>
> live cell organelles imaging with STED system (in neurons), the organelles being
>
> stained with fluorescent proteins (EYFP). The problem is that the level of
>
> staining required with our STED set up seems to be incompatible with live cell
>
> physiology. Indeed, expression levels that show very bright staining in confocal
>
> are not bright enough for our STED system, to the point we cannot do any
>
> picture.  Higher expression levels will just damage the organelle, or induce
>
> leak of the fluorophore into the cytosol, so this is not an option.  It also
>
> seems to me that, although the signal is very weak, the intensity of the
>
> excitation is very high, which may cause further damage to the organelles.  So
>
> far, my colleagues have been able to do nice images of the cells (neurons) with
>
> a cytosolic volume label (fluorescent protein), although they require high
>
> staining too, so it seems that our system is just fine but we cannot work with
>
> small volumes. Is this a common problem with STED systems? We are using filters
>
> in the detection path, no AOBS.
>
>    
>> Thanks
>>      
>    
>> Have a nice day
>>      
>    
>> Sandrine
>>      
>    
>>      
>    
>>      
>    
>> ________________________________________
>>      
>    
>> De : Confocal Microscopy List [[hidden email]] de la part de
>>      
> George McNamara [[hidden email]]
>
>    
>> Date d'envoi : samedi 10 septembre 2011 23:59
>>      
>    
>> À : [hidden email]
>>      
>    
>> Objet : EasySTED segmented waveplate
>>      
>    
>>      
>    
>> *****
>>      
>    
>> To join, leave or search the confocal microscopy listserv, go to:
>>      
>    
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>      
>    
>> *****
>>      
>    
>>      
>    
>> Dear Confocal (and nanoscopes) Listserv,
>>      
>    
>>      
>    
>> I have been researching improvements in STED fluorescence nanoscopy. The
>>      
>    
>> biggest improvements I came across were (order not meant to imply criticality
>>      
>    
>> or price) are:
>>      
>    
>>      
>    
>> * 3D-EasySTED segmented waveplate(s) ... vs current Leica (2D) Vortex
>>      
>    
>> Phase Filter.
>>      
>    
>> * Pulsed depletion (i.e. 80 MHz Ti:Sa instead of continuous wave (CW) laser.
>>      
>    
>> Note that this ideally by around 100 to 300 ps duration and start some number
>>      
>    
>> of picoseconds after the excitation pulse has left (needed to give excited
>>      
>    
>> state to reach the lowest S1 state).
>>      
>    
>> * Time gated (g-STED, aka FLIM STED) instead of continuous detection (note
>>      
>    
>> that this requires pulsed excitation, and fits well with pulsed depletion).
>>      
> See
>
>    
>> Vicidomini et al 2011 Nat Methods for recent publication (note that Fig 1 is
>>      
> on
>
>    
>> fluorescent nanodiamonds - not a typical fluorophore. Look at the figures for
>>      
>    
>> those using Alexa Fluor 488, etc).
>>      
>    
>> * New Leica HyD detector(s). Sadly the CW-STED demo system that just left
>>      
>    
>> Miami had only one HyD.
>>      
>    
>> * Ultra stable vibration isolation table (unlike the demo system).
>>      
>    
>>      
>    
>> EasySTED was published in 2010 by Reuss et al in Optics Express (open
>>      
>    
>> access at http://www.opticsinfobase.org/oe/abstract.cfm?uri=oe-18-2-1049 ).
>>      
>    
>>      
>    
>> I found online Matthias Reuss's PhD dissertation (http://archiv.ub.uni-
>>      
>    
>> heidelberg.de/volltextserver/volltexte/2011/11539/pdf/Reuss_Dissertation.pdf
>>      
> ),
>
>    
>>    which has lots of cool stuff (though temperature wise does not come close to
>>      
>    
>> A. Giske's cryoSTED dissertation, http://archiv.ub.uni-
>>      
>    
>> heidelberg.de/volltextserver/volltexte/2008/7969/pdf/CryoSTED_microscopy_Ph
>>      
>    
>> DThesis_agiske.pdf ).
>>      
>    
>>      
>    
>> Reuss's dissertation has stuff from the 2010 Optics Express article, a little
>>      
> bit
>
>    
>> on time-gating STED but the item that caught my eye was 3D-EasySTED.
>>      
>    
>>      
>    
>> I have done more research on EasySTED and in exchanging emails with Goetz
>>      
>    
>> Zinner at the manufacturer (www.b-halle.de), Goetz wrote that:
>>      
>    
>>      
>    
>> "An interesting result I found is that a 6.5 lambda plate at 710 nm would have
>>      
>    
>> 7.5 lambda at 622 nm, 8.5 lambda 554 nm and 9.5 lambda 501 nm. This would
>>      
>    
>> enable to use only one mode converter for all four depletion wavelengths."
>>      
>    
>>      
>    
>> Transmitting any of the half lambda wavelengths through a single segmented
>>      
>    
>> waveplate produces STED depletion doughnuts at each of these wavelengths.
>>      
>    
>>>  From Reuss's dissertation it seems (to me) that a tolerance of a few percent
>>>        
>    
>> (ex. 2% would be 10 nm at 501 nm) would still produce an acceptable
>>      
>    
>> doughnut.
>>      
>    
>>      
>    
>> But wait, there is more - the same segmented waveplate can also tranmit
>>      
>    
>> whole lambda wavelengths, and these can be calculated using the data
>>      
>    
>> available from a paper Goetz mentioned in his email:
>>      
>    
>>      
>    
>> G. Ghosh 1999 Dispersion-equation coefficients for the refractive index and
>>      
>    
>> birefringence of calcite and quartz crystals.  Opt Commun 163: 95–102
>>      
>    
>> (available online at www.sciencedirect.com).
>>      
>    
>>      
>    
>> Using Goetz'z equation (actually, the standard Lambda retardance equation -
>>      
>    
>> see Reuss) and Ghosh's data, I made an Excel file that presents Goetz's
>>      
>    
>> results. Besides confirming that one segmented waveplate can have multiple
>>      
>    
>> half and whole lambda's, one of the coolest features is that (interpolating
>>      
> from
>
>    
>> Ghosh's data), 467.5 and 890 um thick quartz segmented waveplates produce
>>      
>    
>> whole and/or half Lambda retardance values for a variety of thicknesses that
>>      
>    
>> are round number of micrometers in thickness (200, 300, 400 um, for
>>      
>    
>> example).
>>      
>    
>>      
>    
>> For those interested in EasySTED, I am happy to send anyone the Excel file.
>>      
>    
>> Best to contact me through my office email address,
>>      
>    
>> [hidden email].
>>      
>    
>>      
>    
>> Sincerely,
>>      
>    
>>      
>    
>> George
>>      
>    
>>
>>      
>
>
>
>
>    

--


George McNamara, PhD
Analytical Imaging Core Facility
University of Miami