In answer to the following letter by Susana Zanello:
>
> Dear All:
>
...

Hi, Susana,
I'm not quite sure whether the others are bothered by this discussion.
If somebody thinks so, please protest. Then we do this discussion
private. Thank you. Also, before delivering my sermon, I should like
you to keep in mind Mark Cannell's yesterday email; might be his
suggestion is good.

>
> 1) What precisely do you mean when you say that the beam properties
>     will be no longer as they began after a trip through the fiber optic?
>     I hate to be obtuse but I need a more exact explanation (again due
>     to the density of my hair)

A laser "beam" has very specific properties; this is what actually
allows us to call it a "beam" (although it is not a beam in the strict
mathematical meaning of the word). One of these properties is e.g. the
divergence of the beam, another one its mode structure, possibly its
state of polarization a.s.o. Some lasers, e.g. the HeCds, do perfer to
not emit light in the so called TEM00 state (don't bother about the name,
it just means that you have a beam "as you would expect it") "voluntarily";
their "beam" rather looks like a tube (TEM*01), and forcing them to
produce a real TEM00 beam causes a loss of power.  All these magnitudes
are well defined and normally must be specified by a producer (unless
the machine is to be used in a so called "rock concerto" or comparable
totally stupid and unnecessary events).
It does not make any sense to explain this in detail now.

I am going to give an example of practical importance:
If a laser beam e.g. has a specified divergence of 0.5mrad (typical value
for a UV beam from an Ar-Ion-Laser in all-lines-central-UV configuration)
and a beam diameter (we do not bother now about the proper definition of
"beam diameter") of 1.5mm at the output-coupler (i.e. the mirror from
which the beam is emitted on the laser) it is possible to calculate its
diameter (i.e. the spot size) at any distance.
A laser beam can be transported in a fiber at which this property is
conserved, i.e. the beam divergence will not be increased by orders of
magnitude when the beam leaves the fiber at the outcoupling end, IF THE
FIBER IS BUILT ACCORDINGLY AND THE INCOUPLING AND OUTCOUPLING DONE
PROPERLY. For "deeper" opto-technical reasons the core diameter of a
"proper" laser beam fiber (the fiber is not simply a line of plastic but
has an inner structure) is wavelength dependent. Proper UV fibers
have a core diameter of less than a micron (if I remember correctly I once
read the figure of 100nm). It is less a problem of production of such a
fiber which makes it commercially unavailable. Rather, the unsolved
problem of properly coupling in and out a 1.5mm beam into a 100nm
fiber core at a N.A. of 0.1 or 0.2 makes this fiber being not useful,
and therefore it is not produced and sold.
If you, however, use an "unsuitable" fiber (i.e. what is called an
"optical cable" in some catalogues), the light which will leave the
fiber at its outcoupling end, will not at all have any defined "beam
divergence" any longer! It will hardly be possible to call it a "beam",
hence. Therefore, it will not be possible to calculate some kind of
spotsize!

Assume you are intending to do an experiment on Ca2+ diffusion in the
dendrite of a neuron. Let us assume that you are interested exclusively in
the Ca2+ concentration in a certain spine or that you want to release
caged Ca2+ exclusively in a 1um^3 large volume at the membrane of the
nucleus. If your fiber indeed WOULD produce a 100nm light spot, then you
could, provided you have a good kind of micromanipulator and carefully
manoeuvre the fiber close to the spine or the nucleus, do an educated
guess and state: "I detect exclusively fluorescence from the spine."
If your fiber releases an undefined "beam", then you will definitely not
be able to call this guess "educated". Depending on the kind of referee
of a paper manuscript this could be a critical point. (I'm not (yet) a PhD,
I've no ambitions to be a paper referee, no danger from my side).
There are a multitude of examples where you might be dependent of indeed
knowing that you are exciting fluorescent light exclusively in a
volume of a um^3 or something like that.

I am not talking now about the fact that probably not all UV light of
the beam emitted from the (hypothetical) fiber would be stopped by
the spine. It's a 3D problem and scattered UV light as well as
light transmitted via the spine might cause fluorescence elsewhere!
A promising approach to solve this problem, to my mind, can only be
2photon excitation with a high NA water immersion lens, since then you,
indeed, know that only in a very small 3D volume close to the focal centrum
of the objective light intensity will be sufficient to cause fluorescence.



> 2) What are the characteristics of these rumored UV-capable (fused silica
>     fiber optics?  Johannes - you can just give me the references

I've never worked with them. I had considered this for some weeks and was
desillusionized by the experts from Schott, Newport, ... As far as I know,
again, there is no such fiber available. What is called "UV-cable" can
be purchased from Schott, Newport, a.s.o. (I'm going to send you the
adresses later, I've to look them up for myself).

> 3) Does the following sound practical/possible/adequate...
>
> I was thinking of a light source of wide range (Xenon, etc) followed (in
> the light path) by a set of interference and neutral density filters to select
> wavelength and adjust intensity of the light.  Then this would be collected
> with a suitable fiber optic (fused silica?) that would end in a light guide
> which is tapered at the end - similar to a patch-pipette.  This is then held
> by a micromanipulator and "focussed" onto the cell(s).  This is in use
> with some investigators doing another thing with visible range light.

I do not dare to answer this question. All I had to say about it I've already
told.  I know there are some persons who are trying to open this path and,
who knows, might be they are going to succeed. I wish them a good success,
anyway and do not like to demotivate. A bad-willing referee however,
see above, might have an argument to stop a publication...

>
> My Questions:
>
> 1) Can I use a conventional fiber optic in this system?

The losses of intensity in a conventional fiber will be strong if
UV light is used. This depends of course on the length of the path
in the fiber.

> 2) If not, what is required?

Fused silica cables.

> 3) Can I do something to avoid the use of the micromanipulator via
>      an actuated dual-shutter to rapidly switch from the stimulating "beam"
>      to the excitation beam of the CLSM?  (This would of course be through
>      the objective of the scope)

Hm, I do not know. As Aryeh pointed out there exist opto-mechanical shutters
by Vincent Associates or other factories. To do faster things, the use
of electro-optical modulators might be suitable (but then we are again
talking about several tens of k$).

>
> We'll have a Bar-B-Q after my Nobel Prize if you help me.

Great, I'm going to deliver a dinner-speech, but I won't wear any necktie.

>
> Susi
>

Johannes

--
*****************************************************************************
Paul Johannes Helm

Mailadress:     Department Physics 4
                The Royal Institute of Technology
                S-100 44 Stockholm
                Sweden

Visitingadress: Department Physics 4
                The Royal Institute of Technology
                Teknikringen 14/4tr.
                S-100 44 Stockholm
                Sweden

Voice:          +46 8 790 7219
Fax:            +46 8 205609
Telex:          11421 kth
WWW:            http://www.fysik4.kth.se/~johannes
email:          [log in to unmask]
*****************************************************************************