Re: TIRF objective for routine imaging

Posted by James Pawley on
URL: http://confocal-microscopy-list.275.s1.nabble.com/TIRF-objective-for-routine-imaging-tp4778151p4786139.html

Hello Charu,

If using this objective on a CLSM is your plan, then I would go for a lower magnification (the 60x/1.49 instead of the 100x/1.49 for example) due to transmission efficiencies. Objective with the same N.A. but with lower magnifications are almost always a better choice when working with point scanners. Just "zoom" in to meet your Nyquist!

Pete

Hi Charu and Pete,

"Efficiency" is a tricky thing. Everyone is in favor of it but it often means different things to different people.

In the days of widefield systems with Hg arcs and excitation optics that made a large image of the arc in the objective BFP, there was a rumor that the "brightness" of a fluorescent image would vary with the fourth power of the NA and inversely with the square of the magnification. So high NA, 40x (and now 20x!) objectives became popular. The problem was that this brightness was related to visibility by eye. Once you wanted to record the low-mag image on film, you had to use slower film with finer grain to preserve the details (Ok. Many people didn't care about the details and used fast film just to see if there was some stain, but this approach lost the details.)

When we got CCD cameras, we all became a bit more conscious of Nyquist sampling and realized that lower mag meant smaller camera pixels and, in the end, we collected the same number of photons/pixel no matter what objective mag you choose.  Assuming that you had settled on a camera built around one of the ubiquitous SONY chips designed for the Japanese HDTV standard, the pixels were 6.7µm and you would just have to choose the right coupling tube mag to allow you to meet Nyquist.

So that is the mag part: In terms of image intensity, mag isn't important with laser instruments (including laser TIRF), unless you need low lag to cover a larger field and are willing to scan slowly, with a huge raster and tolerate some off-axis image degradation.

What about the NA part? As a mercury arc is about the same dimensions as the BFP of the objective (say 3-5 mm. It varies with arc power: 50w, 100w, 200w), any efficient illumination system will focus this into the BFP at about 1:1 because 1:1 optics have the highest potential light throughput.

Therefore, the objective that accepts light from the largest fraction of this arc-image and conveys it to the specimen will excite the most fluorescence. As the diameter of the BFP varies directly with NA, its area varies with (NA)*2. As the same should follow for the efficiency with which the objective collects the excited fluorescence, it will also vary with (NA)*2 and the total overall brightness will vary with (NA)*4.

Some BFP background

All other things being equal, a 100x objective has a focal length that is 2.5x shorter than a 40x lens. If, for a moment, you can just imagine the simplest lens diagram with on-axis, parallel illumination striking the lens and then converging to form a focused spot, one focal length away from the optical center of the lens (the optical center is the plane about which it seems to act). No matter what the objective magnification, the boundary of the angle of convergence is set only by the NA. Therefore, as the 40x lens has a 2.5x longer focal length, its principle plane must be 2.5x farther from the focal plane compared to that of the 100x lens. And as the divergence is the same, from similar triangles, the diameter of the parallel ray bundled needed to fill the 40x lens must be 2.5x larger than that needed to fill the 100x lens.

Put simply, the diameter of the clear aperture that you see when you look at a 100x NA 1.3 is about only about 40% as large as that seen when looking at the back of a 40x NA 1.3, and if you focus a more-or-less uniform image of the arc onto this plane, and it is big enough to "fill" the BFP of the 40x, then you will find that about (2.5)*2 = 6.25x more light comes out of the 40x, compared to the 100x. (On the other hand, if your excitation optics make an image of the arc in the BFP that is much smaller than that of the 40x BFP, this effect will be proportionally less. Assuming Kohler illumination, you should be able to see the relative size of the arc and the BFP if you just let the illumination proceed from the objective and fall onto a piece of lens tissue placed a cm or two on the far side of the focus plane.)

However, if you have a laser-based illumination (such as is found on many laser confocals), things are a little different. First all all, on the illumination side, excitation intensity itself is not a usually a problem. In fact we have to adjust the laser power down to prevent singlet-state saturation and undue photodamage. So one of the two (NA)*2 intensity terms become irrelevant. (On the imaging side, of course, larger NA will collect more light and, in the absence of spherical aberration etc, will deliver more photons to the detector.)

On the other hand, in laser-based systems, the diameter of the diffraction-limited spot is inversely proportional not to the NA engraved on the objective barrel but to the fraction of this NA actually filled by the laser beam. A 1mm diam unexpanded laser beam filling a 60x 1.3NA lens, with a BFP about 8mm in diameter will form a focused spot that is 8x LARGER than it should be. (The confocal image may still look sort of OK as the NA on the imaging  side will still be 1.3, but the X,Y and Z resolution will still be lower.)

Now, confocal manufacturers know all this and they try to set things up so that the laser beam is large enough to "fill the BFP" of the most common lenses with more-or-less uniform laser intensity. However, you can see their problem, unless the illumination optics actually adjusts to match each lens, then we are back to the widefield system: if it is adjusted so that the laser "fills" the 40x BFP, then only 1/6th of that light will be used with the 100x. The rest will be bouncing around causing troublesome reflections. So they compromise, with the result that, in practice, one will often record slightly better resolution in  X, Y and Z with the higher mag objective because the illumination will better fill its aperture.

So wasn't that a long story. I hope you are not all asleep (but I will be checking!)

Sorry. Must be Spring! And sorry for the typos...

Jim P.

              **********************************************

Prof. James B. Pawley,                                      Ph.  608-263-3147 
Room 223, Zoology Research Building,                                  FAX  [phone]
1117 Johnson Ave., Madison, WI, 53706                                [hidden email]
3D Microscopy of Living Cells Course, June 12-24, 2010, UBC, Vancouver Canada

Info: http://www.3dcourse.ubc.ca/                Applications still being accepted

               "If it ain't diffraction, it must be statistics." Anon.

On Mar 23, 2010, at 04:55 AM, charu tanwar wrote:

Thank you all for your valuable inputs.

Charu

CHARU TANWAR
Imaging Specialist
Advanced Instrumentation Research Facility
Jawaharlal Nehru University
New Delhi 110067
India.

--- On Tue, 23/3/10, Tim Feinstein <[hidden email]> wrote:

From: Tim Feinstein <[hidden email]>
Subject: Re: TIRF objective for routine imaging
To: [hidden email]
Date: Tuesday, 23 March, 2010, 5:55 AM

We use the same objective as Neeraj on a A1/TiE for confocal, widefield and TIRF.  No complaints.

cheers,

Tim Feinstein

On Mar 22, 2010, at 7:39 PM, Neeraj Gohad wrote:

Hi Charu,

 

We have A Nikon TIRF module on our Nikon TiE so we have the Apo 60X 1.49 NA TIRF Objective, I have used this for regular confocal imaging with great success.

 

Best,

 

Neeraj.

 

Neeraj V. Gohad, Ph.D.

Postdoctoral Fellow

Okeanos Research Group

Department of Biological Sciences

132 Long Hall

Clemson University

Clemson,SC-29634

Phone: [phone]

Fax: [phone]

 

Website: http://www.clemson.edu/okeanos

 

Please note my new email address: http://in.mc83.mail.yahoo.com/mc/compose?to=neerajg@...

 

 

From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Marco Dal Maschio
Sent: Monday, March 22, 2010 3:56 PM
To: http://in.mc83.mail.yahoo.com/mc/compose?to=[e-mail]
Subject: Re: TIRF objective for routine imaging

 

Dear Charu,

I had the opportunity to test from the same company 60x 1.45tirf // 60x VC // 60x 1.49 tirf.

I remember that probably 60x 1.49 is not plan, related to the  flat field correction.

But this objective is currently used by colleagues  performing tracking and imaging in neuronal cultures

with optimal image quality. If you want I can send you images comparing 60x VC and 60x 1.49 tirf.

Sorry but no experience about 100x.

 

best

Marco 

 

On Mon, Mar 22, 2010 at 10:29 AM, Charu Tanwar <http://in.mc83.mail.yahoo.com/mc/compose?to=tanwar_charu@...> wrote:

Dear List

Anybody please let me know whether we can use TIRF 100XH (N.A.
1.49,Working Distance 0.12mm,Coverglass correction 0.13-.20mm - this is a
special TIRF objective) for routine confocal imaging of bacteria and RBC's with
low flourescence signal.
Can it be mounted on a confocal microscope from TIRF microscope if both the
systems are from same company.

Thanks in advance

Charu Tanwar
Imaging Specialist
Advanced Instrumentation Research Facility
Jawaharlal Nehru University
New Delhi
India.

        

---

Your Mail works best with the New Yahoo Optimized IE8. Get it NOW!.

--

              **********************************************
Prof. James B. Pawley,                                          Ph.  608-263-3147 
Room 223, Zoology Research Building,                                  FAX  [phone]
1117 Johnson Ave., Madison, WI, 53706                                [hidden email]
3D Microscopy of Living Cells Course, June 12-24, 2010, UBC, Vancouver Canada
Info: http://www.3dcourse.ubc.ca/            Applications due by March 15, 2010

               "If it ain't diffraction, it must be statistics." Anon.