Cover slip thickness

Everything relating to microscopy hardware: Objectives, eyepieces, lamps and more.
Post Reply
Message
Author
ZodiacPhoto
Posts: 59
Joined: Sat Sep 24, 2022 6:53 pm

Cover slip thickness

#1 Post by ZodiacPhoto » Mon Dec 12, 2022 12:04 am

Just a newbie question: why most objectives are made for 0.17mm cover slip thickness, and most cover slips I see for sale are #1 (0.13-0.16mm) rather than #1.5 (0.16-0.19mm)?
And is it even noticeable?
Thanks!

BramHuntingNematodes
Posts: 1538
Joined: Tue Jan 21, 2020 1:29 am
Location: Georgia, USA

Re: Cover slip thickness

#2 Post by BramHuntingNematodes » Mon Dec 12, 2022 12:24 am

I believe it is because routine lab work often ends up with a thin layer between the cover slip and sample, so the slip is made thinner to compensate. It's also not noticeable until NA gets a kve the .6 or .65 of your typical 40x lab objective. For these high dry objectives I have found that coverslip accuracy becomes very important, although often these objectives also have a collar for adjusting to different coverslip thicknesses and so can be used with thinner slips.

10x and 20x objectives typically show no ill effects at all from the thinner slip.
1942 Bausch and Lomb Series T Dynoptic, Custom Illumination

EYE C U
Posts: 288
Joined: Wed Jul 15, 2020 5:18 pm

Re: Cover slip thickness

#3 Post by EYE C U » Mon Dec 12, 2022 1:27 am

be careful as hell with them..they can trick you into thinking you are holding one only then the other falls on the floor.next thing you know your bleeding..

ZodiacPhoto
Posts: 59
Joined: Sat Sep 24, 2022 6:53 pm

Re: Cover slip thickness

#4 Post by ZodiacPhoto » Mon Dec 12, 2022 2:01 am

be careful as hell with them..they can trick you into thinking you are holding one only then the other falls on the floor.next thing you know your bleeding..
I've been lucky so far...
However, I noticed that even the cover slips that are sold as "pre-cleaned" are dirty right out of the box, and they do stick together:
slips.jpg
slips.jpg (145.06 KiB) Viewed 2887 times

Sure Squintsalot
Posts: 393
Joined: Mon May 16, 2022 3:44 pm

Re: Cover slip thickness

#5 Post by Sure Squintsalot » Mon Dec 12, 2022 2:08 am

If you're an impatient gorilla like me, filthy cover slips are a major PITA to clean. I'm sorry to hear that despite the added cost, even "clean" ones are dirty.

BramHuntingNematodes
Posts: 1538
Joined: Tue Jan 21, 2020 1:29 am
Location: Georgia, USA

Re: Cover slip thickness

#6 Post by BramHuntingNematodes » Mon Dec 12, 2022 2:32 am

I bought .17 if for nothing else they break less frequently
1942 Bausch and Lomb Series T Dynoptic, Custom Illumination

GerryR
Posts: 183
Joined: Sun May 22, 2022 11:44 am
Location: Virginia, USA

Re: Cover slip thickness

#7 Post by GerryR » Mon Dec 12, 2022 5:06 pm

Cover glass thickness and resolution.doc
(26 KiB) Downloaded 103 times
Copied this off the web. Not very detailed, but gives a good overview:

ZodiacPhoto
Posts: 59
Joined: Sat Sep 24, 2022 6:53 pm

Re: Cover slip thickness

#8 Post by ZodiacPhoto » Tue Dec 13, 2022 1:42 am

The specimen is embedded in mounting medium. The thickness of this medium must be added to the thickness of the cover glass. A specimen which is located deep in the medium will have a larger “effective” cover glass thickness than a specimen which is located right beneath the cover glass.
Ok, that explains why cover glass < 0.17mm is more common.
I mostly look at pond life and diatoms, and the thickness of the water above the specimen may affect the resolution...

GerryR
Posts: 183
Joined: Sun May 22, 2022 11:44 am
Location: Virginia, USA

Re: Cover slip thickness

#9 Post by GerryR » Tue Dec 13, 2022 1:59 pm

ZodiacPhoto wrote:
Tue Dec 13, 2022 1:42 am
The specimen is embedded in mounting medium. The thickness of this medium must be added to the thickness of the cover glass. A specimen which is located deep in the medium will have a larger “effective” cover glass thickness than a specimen which is located right beneath the cover glass.
Ok, that explains why cover glass < 0.17mm is more common.
I mostly look at pond life and diatoms, and the thickness of the water above the specimen may affect the resolution...
I may be wrong on this, but I believe that concave-cavity slides are used for pond water (n~=1.33) studies, so the cover glass basically rests on the slide with barely a film between slide and cover glass. Of course this is not for permanent mounting.

SuiGenerisBrewing
Posts: 78
Joined: Tue Mar 29, 2022 10:39 pm
Location: Canada
Contact:

Re: Cover slip thickness

#10 Post by SuiGenerisBrewing » Tue Dec 13, 2022 2:41 pm

The answer is "its complicated", as you've likely already realised from some of the above posts. Most microscope lenses intended to work with coverslips that are #1.5 in thickness. But this assumes that the sample is attached to the coverslip. If your sample is on the slide, or suspended in medium between the coverslip and slide, that "rule" doesn't hold out. In my lab we do some super-resolution microscopy, and careful matching of coverslip thickness, sample thickness, index matching, etc, is critical to getting good images. And dealing with coverslip thickness is a major PITA in that process - and in many cases ends up being a compromise between the mounting mediums optical density, the sample thickness, and the thickness of the cover slip.

It can be a real challenge to get the "perfect" coverslip - for example, if you prepare two slides with the same sample, mounted the same distance away from the coverslip, but in one sample use an aqueous mounting medium and in the other use a high optical density mounting medium, you'll likely need a different thickness coverslip for each sample to get the best resolution.

The good news is that with the kinds of microscopes used by hobbyists, the impact of using the wrong thickness coverslip is pretty minor in terms of your ultimate image quality and resolution. And, as others have mentioned above, a thinner coverslip may actually be superior, depending on how you've prepared your samples. For my home setup I generally use #1.5s, but I also mostly image yeast and bacteria, and prepare my mounts directly on the coverslip, or use coverslip-less preparations.

Bryan

Macro_Cosmos
Posts: 108
Joined: Mon Jul 19, 2021 2:20 am
Location: 192.0.0.1
Contact:

Re: Cover slip thickness

#11 Post by Macro_Cosmos » Thu Dec 15, 2022 12:42 am

SuiGenerisBrewing wrote:
Tue Dec 13, 2022 2:41 pm
The answer is "its complicated", as you've likely already realised from some of the above posts. Most microscope lenses intended to work with coverslips that are #1.5 in thickness. But this assumes that the sample is attached to the coverslip. If your sample is on the slide, or suspended in medium between the coverslip and slide, that "rule" doesn't hold out. In my lab we do some super-resolution microscopy, and careful matching of coverslip thickness, sample thickness, index matching, etc, is critical to getting good images. And dealing with coverslip thickness is a major PITA in that process - and in many cases ends up being a compromise between the mounting mediums optical density, the sample thickness, and the thickness of the cover slip.

It can be a real challenge to get the "perfect" coverslip - for example, if you prepare two slides with the same sample, mounted the same distance away from the coverslip, but in one sample use an aqueous mounting medium and in the other use a high optical density mounting medium, you'll likely need a different thickness coverslip for each sample to get the best resolution.

The good news is that with the kinds of microscopes used by hobbyists, the impact of using the wrong thickness coverslip is pretty minor in terms of your ultimate image quality and resolution. And, as others have mentioned above, a thinner coverslip may actually be superior, depending on how you've prepared your samples. For my home setup I generally use #1.5s, but I also mostly image yeast and bacteria, and prepare my mounts directly on the coverslip, or use coverslip-less preparations.

Bryan
Very insightful post, thank you for sharing your expertise.
Do you combine superresolution methods with deconvolution? I am interested in the optimum procedure to combine imaging with practical deconvolution, by utilising microspheres to obtain an experimental PSF. From what I have read, the optimal protocol involves making the microsphere sides to obtain the PSF in exactly the same matter as the sample -- same type of coverslip, same mountant, same protocol and preferably matched to the same fluorescence dyes. This gets very esoteric. I would love some input from scientists that deal with superresolution! Thanks. :D

SuiGenerisBrewing
Posts: 78
Joined: Tue Mar 29, 2022 10:39 pm
Location: Canada
Contact:

Re: Cover slip thickness

#12 Post by SuiGenerisBrewing » Thu Dec 15, 2022 1:39 pm

Macro_Cosmos wrote:
Thu Dec 15, 2022 12:42 am
Very insightful post, thank you for sharing your expertise.
Do you combine superresolution methods with deconvolution? I am interested in the optimum procedure to combine imaging with practical deconvolution, by utilising microspheres to obtain an experimental PSF. From what I have read, the optimal protocol involves making the microsphere sides to obtain the PSF in exactly the same matter as the sample -- same type of coverslip, same mountant, same protocol and preferably matched to the same fluorescence dyes. This gets very esoteric. I would love some input from scientists that deal with superresolution! Thanks. :D
Deconvolution and super-resolution are similar, but different.

There are good free tools to do deconvolution, and they're not too hard to implement with a bit of know-how. My personal favorite (which we use extensively at work) is DeconvoluitonLab2 (paper | website). It runs in ImageJ (a free scientific imaging package). You can use either a hypothetical PSF, or you can use a resolution-limited bead or other particle to generate a true PSF. The biggest limitation you'll find with deconvolution is that they are generally monochromatic in nature - e.g. good for fluorescence, but less so for widefield/phase contrast/etc images. There are brightfield-orientated approaches (example), but I've not used any of those and cannot comment on how well they work. Honestly, for brightfield images, standard photo-stacking/extended focal depth algorithms are about as good as it gets.

For deconvolutionlab2, the best resource to learn how to use it is Johanna M. Dela Cruz's youtube channel. She has videos on how to use it with both theoretical and real PSFs.

Super-res isn't really deconvolution, but it is all mathematical recreations of the "ground truth" image. We use two approaches:

#1 - Enhanced Super-Resolution Radial Fluctuation (eSRRF), which runs via ImageJ. All fluorophores blink to some extent, which is why (if you've ever looked through the eye pieces of a fluorescence scope) fluorescent samples often look like there is a heat mirage in front of it. That blinking encodes some spatial information, which can be used to generate an image with ~2 times the physical resolution of the microscope. The downside to the approach is that you need to capture 50 to 500 images of the sample to create one eSRRF image. But with a super-sensitive camera, we can do this with live cells, giving us super-resolution imaging of living samples. I've embedded an image I captured of this on Monday, its a macrophage stained for DNA (white), actin (yellow), plasma membrane (purple), and has phagocytosed (eaten) some bacteria (cyan). I've also included a single frame of the original capture that was used to generate the eSRRF image

eSRRF
RGB-1.png
RGB-1.png (173.9 KiB) Viewed 2570 times
Origonal
RGB2.png
RGB2.png (116.62 KiB) Viewed 2570 times
#2 - Ground State Depletion Microscopy (GSDM), which is run on a custom Leica microscope/software package. This is a type of RESLOFT microscopy where you use special tricks to force most of the fluorophores in a sample into a dark (non-fluorescent) state. You then image the active fluorophores, and because you can resolve the individual PSFs of each fluorophore, you're able to calculate the position of each fluorophore by fitting a Gaussian PSF to the imaged PSF. You then use another "trick" to turn on another portion of the fluorophore (and push into darkstate the ones you just imaged), and repeat. 20,000 or so images later you can produce an image that contains most of the fluorophores in the sample. We routinely get resolutions as good as 12 nm with this method (20-25 nm is typical), but it can only be performed on fixed cells and takes over an hour to generate a single image, so we only use it when we want to image inter-protein interactions directly.

Macro_Cosmos
Posts: 108
Joined: Mon Jul 19, 2021 2:20 am
Location: 192.0.0.1
Contact:

Re: Cover slip thickness

#13 Post by Macro_Cosmos » Thu Dec 15, 2022 11:19 pm

SuiGenerisBrewing wrote:
Thu Dec 15, 2022 1:39 pm
Deconvolution and super-resolution are similar, but different...
Thanks for all the information! I have used SVI's Huygens for widefield and DIC (brightfield methods) deconvolution, the results were very good. For just brightfield, on a screen level, sensible sharpening gives comparable results. Good sharpening techniques, such as setting the estimation radius to something tiny is similar to blind deconvolution anyway, and often yields very acceptable results for brightfield microscopy. Here is a devonvolved image of some fibroblast cells.
Image

I read an article regarding STED combined with deconvolution: https://svi.nl/Huygens-STED-Software
Perhaps not much, but combining deconvolution with some superresolution methods seems to bring an improvement.

Thanks for sharing the SRRF links, I thought it was no longer available after Andor's acquisition of the entire team, they have implemented SRRF+ for their new iteration of Zyla and Sona cameras. I will check out DeconvolutionLab, Huygens is great but the software is extremely expensive for an amateur.

SuiGenerisBrewing
Posts: 78
Joined: Tue Mar 29, 2022 10:39 pm
Location: Canada
Contact:

Re: Cover slip thickness

#14 Post by SuiGenerisBrewing » Fri Dec 16, 2022 1:33 pm

SRRF can be run with any camera capable of fast acquisitions. We use a 12-year old Photometrics Evolve 512 Delta with good results. Some of the high-QD CMOS's also work. Andors implementation is simply on the camera itself, whereas we do ours in post. Andor also didn't acquire the team - SRRF is developed as an open-source algorithm by Ricardo Henriques's lab at UCL.

Sure Squintsalot
Posts: 393
Joined: Mon May 16, 2022 3:44 pm

Re: Cover slip thickness

#15 Post by Sure Squintsalot » Fri Feb 03, 2023 8:28 pm

It's hard to photograph but these coverslip surfaces have a fine orange-peel texture to them, viewable with incident light. I'm pretty sure coverslip surfaces should be smooth as, well, glass. Right?

Not only are these filthy, glazed with a fine greasy substance, but they have this bizarre texture to them. Did I get a reject batch? I didn't exactly get them for cheap:
Screenshot 2023-02-03 132119.jpg
Screenshot 2023-02-03 132119.jpg (55.22 KiB) Viewed 2155 times

Javier
Posts: 806
Joined: Tue May 09, 2017 11:19 am
Location: Buenos Aires, Argentina.

Re: Cover slip thickness

#16 Post by Javier » Mon Feb 06, 2023 1:17 pm

ZodiacPhoto wrote:
Mon Dec 12, 2022 2:01 am
be careful as hell with them..they can trick you into thinking you are holding one only then the other falls on the floor.next thing you know your bleeding..
I've been lucky so far...
However, I noticed that even the cover slips that are sold as "pre-cleaned" are dirty right out of the box, and they do stick together:
slips.jpg
Just keep looking for the right ones. I found some superb rebranded coverslips that are sold here in Buenos Aires that come pristine out of the box. They are a bit more expensive, but they make a great difference.

charlie g
Posts: 1831
Joined: Mon Oct 20, 2014 7:54 pm

Re: Cover slip thickness

#17 Post by charlie g » Tue Feb 07, 2023 3:24 am

I just offer a 'shout out' for NJ/ USA firm: " Globe Scientific.com" ( Paramus/ NJ, USA location).

Their coverslips are low cost, these cover slips are clean..right out of the box. I've been useing their variety of slides, and coverslips..for decades.

Please notice..all you 'pond life/ live wetmount slide microscopists'.. the use of #1.0 large rectangular coverslips...offers a huge amount of area far from the 'danger of mucking your dear high mag./low working distance objectives'.

I find these #1.0 coverslips easy to handel..you verify if you have more than one slip in your hand..before transfer to your wetmount droplet...proper protocol..and no bloody injuries.

I do have the: 1.5 slips...but the 1.0 slips are not so delicate as to be a safety problem.
Attachments
DSCN1820.JPG
DSCN1820.JPG (102.52 KiB) Viewed 1990 times

Post Reply