Examining an Apochromatic Objective that Might be Decentred

Introduction

Welcome to my incoherent ramblings about this exotic piece of glass.

The “requirement” for a lower magnification water immersion objective led to the acquisition of an exotic piece of glass. To be fair, the word “requirement” is being used loosely, “rationalising” would be the better term, or just “thing, I want”.

This objective was purchased from a Danish feller who does not know anything about it. I assisted him with other transactions as well. The objective was by all means quite expensive, at a price tag of US$1300, it is one of my higher valued objectives.

The objective I purchased is an Olympus UApo N 340 40x/1.15W bfp1 FN22. I will decipher this mess in the appendix. Notably, the letter “U” here denotes “ultraviolet”, not “universal”.

Revelation

To try out this objective, I ran a differential interference contrast (DIC) Z-stack on the massive Arachnoidiscus deficiens diatom frustule. With a PE3.3, the total magnification was 132x. The results were quite unsatisfactory.

The full resolution photomicrograph can be accessed here (opens in a new tab): Flickr Link

The photomicrograph was sharpened and slightly retouched. While the centre resolution is simply insane, the corners are horrible. Notably, the bottom and top right corners are worse than the other two. Indeed, rotating the objective will displace the fuzzy corners. Is the objective decentred? Subsequently, I compared this Z-stack to a couple other stacks of the same subject using different objective lenses.

The individual, full-resolution files of the 40x W, 40x Dry and 60x W are here (opens in a new tab): 40xW Flickr Link 40x Dry Flickr Link 60xW Flickr Link

Here are the full-resolution crops of each z-stack.

Centre: Link

Corners: Link

Both the 40x dry objective and 60x water immersion objective were plan-corrected super-apochromatic objectives. The 60x z-stack looks different due to the different types of DIC used. I accidentally used a PE2.5x for the 60x z-stack, which pushed the magnification to 150x. From the 60x, fine details concerning the pores of the frustule are displayed. They are not smooth, there is a tiny symmetrical indent in each pore. The lacklustre numerical aperture of 0.95 and no immersion liquid translated to the loss of fine detail regarding the 40x dry objective. Immersion media enables better light transmission, a higher refractive index translates to a brighter image, which contributes to better detail and less interference. Interference manifested as false detail I do not even want to try and fix was clearly present in the 40x dry objective z-stack. Regardless of the horrible bottom right corner, the 40x water immersion objective takes the “W”. Despite the faults, the output was solid and completely on par with my favourite 60x objective.

Methodology

To further analyse this objective, I utilised tiny fluorescence beads. These humble balls are indispensable tools to characterise an objective’s performance for the calibration purposes of confocal microscopy and deconvolution. The beads I have access to are large enough to acquire a measured point spread function (PSF) of the objective, but the process is too tiresome. Nonetheless, it will provide insight into the objectives’ performance due to their nature of being a point-source.

One bead is nice and round in the centre portion, the ones away from the centre have some weird halos. This image already illustrates the non-flatness of this objective. It is not plan-corrected. With ImageJ and some Z-stacking, I can acquire the PSF of this objective across different points. The appropriate size of these beads will depend on your microscopy needs and the appropriate calculations based on modifications of the Rayleigh Criterion. The formula is n*λ/NA, where “n” depends on the microscopy method and specific needs. n=0.61 is used in the famous Rayleigh Criterion. I will be using n=0.51 here, which corresponds to the FWHM (full width half maximum) resolution limit, defined by two beads that are close to each other. 244nm fluorescence beads are appropriate in our situation, divide this by 2.3 allows us to satisfy the Nyquist sampling criterion, giving beads of 110nm. The beads in the image above are around 3um in diameter, excited with green light, using a FITC fluorescence cube. Ideally, refractive indices should match as well, to ensure the minimisation of spherical aberrations.

Defining resolution as the objective’s ability to discriminate between two closely placed beads with adequate contrast, I found a triplet of beads and imaged them at various locations in the frame.

As exemplified above, the right side is distinctively worse than the rest and exhibits far more variation along the z-axis. This, to me, indicates that the objective is quite badly decentered. A series of images at different focal planes were taken for some of the corners — it was difficult to nail the focus. In both the bottom right and top right, the objective almost fails to distinguish between the three beads. There is a lot of comatic aberration.

The problem with this conclusion is clear as well, is this performance within specifications? This objective is not plan-corrected and only has a field number of 22mm, I do not own another objective that is remotely similar to assess this disparity.

Compared to the output from my 60x plan-corrected objective, the differences are clear. I used a PE2x for this one as well, which means the objective is being used outside of its comfortable image circle. 35mm format cameras couple with the PE2.5x.

Here is a video of the beads in motion, the 40x is followed by the 60x, on the same focal plane.

Conclusion

Summing up, I would prefer to not go through the returning process and keep this objective. The performance I am seeing here is clearly problematic. Due to my inexperience with such objectives, I currently do not know if the aberrations are within specifications. An apochromatic objective that is neither plan-corrected nor large field is highly unusual. Perhaps the objective was designed to strictly work with confocal or multiphoton applications where only a small central portion matters. I cannot see this objective being adequate for any widefield applications, it should at least guarantee resolution within its field number.

To solve this conundrum, I have scheduled a meeting with local Olympus engineers. I would like to report their findings and decide if the objective should be returned to the seller, or kept.

References

Guide on the Rayleigh criterion: https://www.edinst.com/us/news/the-rayleigh-criterion-for-microscope-resolution/

Modifications to the formula for confocal and two-photon microscopy: https://www.olympus-lifescience.com/en/microscope-resource/primer/digitalimaging/deconvolution/deconresolution/

Resource on microspheres: https://svi.nl/Recording-Beads

Appendix

  • U: The letter “U” here, in Olympus’ infinite wisdom to cause duplication with another use case, means “Ultraviolet”. Ordinarily, “U” denotes “Universal”. A simple method to distinguish the two is to examine the text that follows, all “Universal” objectives are “plan” objectives. The text will always say “UPlan…”.
  • Apo: Apochromatic
  • N: “New”, meaning it is a UIS2 objective. This is a lazy use of the letter which will no doubt make some Internet dwellers chuckle nowadays, for all the wrong reasons.
  • UIS2: Universal Infinity System 2, Olympus’ lovely self-corrected infinity system. Take notes, Zeiss.
  • 340: The objective boasts adequate transmission at 340nm ultraviolet light.
  • 40x/1.15: 40x with a numerical aperture of 1.15.
  • W: Water immersion
  • bfp1: Back focal plane one. This simply means the BFP location is at -25mm, objectives without this label are either at -19.1mm, incompatible with DIC, or you know this already. The A-line XLPlan objectives have their BFP placed at -48.1mm, necessitating specially designed DIC prisms.
  • FN22: Field Number of 22mm. Most apochromatic objectives boast a field number of 26.5mm.
  • 0.13-0.25: This is the range of the correction collar which is measured in millimetres. The correction collar is an important feature among high numerical aperture dry and water immersion objectives, it enables the compensation for specimen thickness and non-standard coverslips. Most objectives only allow 0.13-0.23mm. Notable lenses that offer such a feature are the DC-Nikkor 105mm and 135mm, Canon’s recent reanimation of this feature in their newer RF-mount macro lens, and the compensation ring found on Schneider Macro Varon lenses. It is in essence a spherical aberration compensation ring.
  • The infinity symbol simply means it is infinity-corrected.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this:
search previous next tag category expand menu location phone mail time cart zoom edit close