Evaluation of a Prototype BF-DF-Oblique-Circular Oblique Lighting (BF-DF-Obl-COL) Condenser

Introduction

An earlier article on ModernMicroscopy.com, Machining a Darkfield Insert for the Olympus BH2 1.25 NA Condenser, noted that the darkfield inserts were made for Peter Cooke (MICA, Chicago, IL) to be used to teach high-resolution dispersion staining in his advanced microscopy classes using Olympus BH2 microscopes. Peter has asked whether there is a better condenser design that will reduce the time and effort needed to switch between brightfield and darkfield. This has not been an issue for my student microscopes with external fiber-optic illumination systems because the darkfield stop is inserted at the end of the light-guide, which serves as the light source. This is not feasible for microscopes such as the Olympus BH2 with built-in illumination systems. My second student microscope is a modified LOMO Biolam (Multiscope) with more capability in transmitted light than my modified Monolux microscope¹. This added capability includes the phase contrast upgrade for the Multiscope, an aplanatic 1.40 NA condenser, and a ball-bearing rotary stage. The LOMO phase contrast condenser design seemed to provide the basis for a prototype to demonstrate an answer to Peter’s need for rapid change between brightfield and darkfield when the darkfield stop must be near the aperture diaphragm of the condenser. The phase contrast condenser allows this rapid switch only for the 10X objective, which needs the phase annulus for the 100X objective to achieve darkfield. The open position in the annulus wheel provides brightfield with an aperture diaphragm. I was not satisfied with the 0.8 NA condenser optics when used for brightfield with the 40X 0.65 NA objective. I found that the LOMO 1.25 NA Abbe condenser is able to achieve acceptable Köehler illumination with the 40X objective. My idea was to replace the phase contrast annuli in the centerable wheel with a set of darkfield stops and use the 1.25 NA Abbe lens in place of the 0.8 NA lens. Chris Vander Tuuk of LOMO America was generous enough to donate a base of the LOMO phase contrast condenser for building a prototype condenser. Since the top lens can be unscrewed from the top of the 1.25 NA condenser, my plan was to use this configuration with a suitable stop in the wheel with the 4X objective.  McCrone Microscopes and Accessories donated a diatom test plate used to demonstrate the performance of the prototype condenser. Oblique illumination is obtained with this condenser by decentering the stop and setting the aperture diaphragm at almost the full NA of the objective. Annular illumination, recently named circular oblique lighting (COL) by Paul James, is achieved by using a smaller stop so that the rear focal plane of the objective shows a narrow ring of illumination at just below the maximum NA of the objective².

Prototype Condenser Mounted in the Modified Biolam Microscope

Figure 1 shows my modified Biolam microscope fitted with the prototype condenser shown in Figure 2. The one-piece LOMO bracket shown in Figure 3, with a slot for the annulus wheel and an upper threaded ring for the condenser lens, was replaced with a two-piece design shown in Figure 4. The aperture diaphragm in the lower part of the bracket is now centerable while viewing the objective rear focal plane before final tightening of the two socket-head cap screws. Note that a male thread needed to mount the LOMO 1.25 NA Abbe lens assembly has replaced the female thread in the one-piece bracket.

The wheel, before the phase annuli were replaced with darkfield stops, is shown in Figure 5. (Close-up examination of the center of the wheel will show a partially completed detent spring needed to replace the spring accidentally broken in a failed attempt to disassemble the center pivot assembly.) All of the annuli were removed and the openings for them bored out to contain the darkfield stops shown mounted in the wheel in Figure 6. Observation of the rear focal planes of the objectives was of critical importance for determining the stop size just sufficient to give a good dark ground as well as to achieve proper centering of the stop using the centering screws shown in Figure 4. These operations were done using a 25X Klein loupe slipped over the eyepiece as shown in Figure 7. The short eyepoint of this loupe, fabricated from a 30 mm stereo microscope 25X eyepiece, does not permit digital image recording with my Nikon CoolPix^® 995 camera. This image recording can be done with this microscope using its drawtube end-mounted 1X objective shown in my online article in Micscape³. That article also contains operating ray diagrams along with close-up views of the system components including the dovetailed attachment for the analyzer and polymer wave-plates, and the mating drawtube holder.

Imaging the Diatom Test Plate with the 4X Objective

Figure 8 shows the condenser in the lowered position with the top lens removed for use with the 4X Zeiss objective. (I now use a Zeiss 4X 160 mm tube length objective instead of the LOMO 4X objective because the Zeiss lens is corrected for use with a compensating Zeiss Kpl eyepiece found best for use with the higher power LOMO objectives.)  Figure 9 shows the diatoms imaged in brightfield with the 4X objective with the field diaphragm adjusted so its image falls just within the 18 mm diameter intermediate image field size of the stop in the 10X Zeiss Kpl high eyepoint eyepiece. Figure 10 shows the same field imaged in darkfield after selecting and centering the stop in the wheel and opening the aperture diaphragm in the condenser. This low magnification is important for surveying the field before switching to a higher power objective.

Imaging Pleurosigma angulatum with Darkfield Illumination

The diatom test slide has the diatom Pleurosigma angulatum which serves as a very good resolution test target for the 40X 0.65 NA objective because the stria spacing of about 0.52 micrometers matches the theoretical resolution of a 0.65 NA objective, when used with a matching illumination NA of just under 0.65. In his response to a letter by Robert B. McLaughlin,⁴ Dr. Walter McCrone republished a very high resolution optical photomicrograph of Pleurosigma angulatum taken in the early 20^th Century by Spitta. This image is shown in Figure 11. It was evidently taken with darkfield illumination, probably with blue light and perhaps with even shorter wavelength ultraviolet. The definition of “just resolved” means that the periodicity of the structure will be detectable but this fine structure will not be faithfully resolved. Figure 12 shows Pleurosigma angulatum recorded with the LOMO 40X objective and darkfield illumination from the prototype condenser. The condenser height had to be raised from the brightfield setting in order for the high NA rays to illuminate the specimen for darkfield. This same problem exists with the Olympus Abbe condenser used with my darkfield inserts and will be discussed in a subsequent paragraph. The CoolPix^® lens zoom control was set so the diagonals of the recorded field  the field seen with the 10X 18 mm FN eyepiece and the image was subsequently cropped in Adobe PhotoShop^®. The periodicity is recorded along with lines that initially were suspicious of being alias lines from the camera sensor resolution being close to the optical resolution. The CoolPix^® lens was zoomed to cover about half the field size of the eyepiece for the cropped portion of the field shown in Figure 13. The lines are still present and therefore not from digital camera aliasing because they are also evident on a close examination through the eyepiece. Tony Havics of pH2, LLC previously tested my modified Biolam. Tony found that darkfield, with the stop at the fiber-optic light guide end with the aplanatic condenser and the same 40X objective, was capable of resolving the first three sets of lines of the HSE/NPL Test Slide, as shown in Figure 14. Resolving the three sets of lines is a requirement for counting asbestos fibers using phase contrast microscopy.

Comparison of Results with Brightfield, Oblique and COL Illumination

My normal practice has been to align the components of the external fiber-optic illumination system using a 9X objective and establish good Köehler illumination using the aplanatic condenser with its aperture diaphragm left fully open. The aperture diaphragm at the light-guide end is used and the darkfield stops are also inserted at that location. I did the same with the prototype condenser and then opened the diaphragm fully at the light-guide end, and subsequently adjusted the diaphragm of the condenser for Köehler illumination. The 40X objective was then swung in on the turret and the field diaphragm imaged just outside the field of view of the eyepiece. I found that I had to raise the condenser from the height setting with the 10X objective in order to be able to fully fill the rear focal plane of the 40X objective with an image of the light source. The field diaphragm was then poorly imaged, as seen when the substage mirror was tilted slightly. I found that the stria pattern on Pleurosigma angulatum was not detectable until the illumination NA set with the aperture diaphragm almost matched that of the objective. This was the aperture setting also used for oblique and circular oblique lighting. The stria pattern in brightfield had very low contrast making detection and focusing difficult. The CoolPix^® was zoomed to record about half the field size and the resulting images were cropped to show the same field as the darkfield image in Figure 13. In order to attempt to match the eyepiece image quality, this image and the other images of the diatom have not had their contrast enhanced digitally. The cross hatch pattern from the stria is shown for brightfield in Figure 15.  The contrast is far inferior to the darkfield image in Figure 13. Oblique illumination from rotating the wheel to decenter the stop for the 40X objective gave much improved contrast as shown in Figure 16. The resolution is now directional, with only one set of parallel lines visible. There is now a camera lens artifact visible in this image as well as in the COL image. A concentric ring pattern is evident in these images and believed to result from residual tool marks in the mold subsequently replicated on the surface of one of the molded aspheric lens elements in the CoolPix^®. This artifact will not be present when using digital microscopy systems from the major manufacturers of microscopes. The contrast of the COL image in Figure 17 is far superior to the brightfield image in Figure 15. The stop for the COL images is the same stop used for the 4X objective, which is somewhat smaller than the stop for the 40X objective as seen in Figure 6.

Role of Condenser Spherical Aberration on Image Quality

The need to raise the condenser from the height setting set for Köehler illumination with the 9X objective for fully filling the rear focal plane of the 40X objective and an additional adjustment upward needed for darkfield illumination is clear evidence of spherical aberration. I decided to image the same field using the LOMO aplanatic condenser, both with the field diaphragm imaged just outside the field of view and fully open and the aperture diaphragm at the end of the light-guide source. I found that the image with the diaphragm fully open, see Figure 18, had better contrast than the brightfield image in Figure 15 taken with the prototype condenser. The image taken with the field diaphragm just outside the field of view had the highest contrast as expected, see Figure 19. These results stress the importance of the condenser being corrected for spherical aberration (aplanatic). The wave optical treatment of image formation and resolution assumes that the object be illuminated with spherical or plane wave fronts. This is not the case for illumination with spherical aberration. The wave front phase relationship for proper destructive and constructive interference to form the image is altered by the spherical aberration. 

The late Edward P. Herlihy, who was a fellow of the Royal Microscopical Society and Vice-President of the Quekett Microscopical Club, lamented the almost universal use of the Abbe condenser, which has an aplanatic aperture far below its claimed aperture value⁵. Herlihy notes that only the aplanatic aperture is useful for microscopy and that the achromatic condenser is far superior. Dr. Walter McCrone has stated in his requirements for a good polarized light microscope that the condenser be aplanatic⁶. Barry Ellam’s article in The Amateur Diatomist notes that the Abbe condenser can be satisfactorily used with darkfield stops⁷. Barry notes that there are problems caused by the lack of correction for spherical aberration in brightfield. He notes that these difficulties can be easily overcome by use of annular illumination (renamed COL by Paul James) long favored by diatomists for resolving the most difficult specimens, especially with a green filter.

Effects of Condenser Spherical Aberration at Rear Focal Plane of a High NA Objective

My first exposure to the problems caused by condenser spherical aberration were while attempting to demonstrate for John Delly the imaging of the interference figure from a  film using my modified Monolux microscope¹. He showed how to fill the rear focal plane of the 60X 0.85 NA objective with much more of the figure by significantly raising the condenser from the position giving Köehler illumination with the 10X objective and then fully opening the field diaphragm. The resulting images were included in my student microscope article and are reproduced in this article along with ray diagrams recently done for this article that indicate what is occurring.

I first set up proper Köehler illumination for the 10X objective using a tissue. I then swung in the 60X 0.85 NA objective and reduced the field diaphragm opening just outside the field of the 60X objective. I then replaced the tissue section with a thin film of polyester. John realized that the rear focal plane was not being fully filled with illumination even when the aperture diaphragm was fully open shown in Figure 20. He suggested raising the condenser, which brought a ring of illumination at the outer edge of the rear focal plane along with an inner dark zone and a center bright spot shown in Figure 21. He then suggested fully opening the field diaphragm along with inserting the analyzer.

This gave us an interference figure for the polyester film fully filling the rear focal plane or exit pupil of the objective as shown in Figure 22. The ray diagram in Figure 23 explains why opening the aperture diaphragm did not illuminate the outer portion of the rear focal plane of the 60X objective when the condenser height was previously established to give best K
öehler illumination with the 10X objective. The high NA rays converged below the object focal plane because of uncorrected spherical aberration and diverged without passing through the field of the 60X objective. Figure 24 illustrates what happened when the condenser was raised.  The intermediate NA rays now passed around the field of the objective and converged above.  The axial rays still go through the object field. Figure 25 illustrates what happened when the field diaphragm was opened fully so that the intermediate NA rays could pass through the object focal plane along with the high and low NA rays. Figure 26 illustrates the situation for an aplanatic condenser without spherical aberration. Lonert indicated the same behavior for an Abbe condenser but provided no examples or ray diagrams⁸.

Summary

The prototype condenser demonstrates that there is an easy design solution that provides rapid switching among the various illumination modes. The imaging tests with Pleurosigma angulatum of the Abbe condenser versus the aplanatic condenser confirm Dr. McCrone’s requirement that a good polarized light microscope (I believe he would also apply this requirement to the biological microscope) must have an aplanatic condenser. This requirement also avoids the need to adjust condenser height for high NA brightfield or darkfield after K
öehler illumination has been properly established for the 10X objective. I was tempted to use the LOMO aplanatic condenser lens for the prototype, but that would have prevented the study of the effects of spherical aberration with the LOMO Abbe condenser. LOMO has a separate single-element aspherical lens that interchanges in the same base as the high NA lens. This design should be as acceptable as having the top lens removable for use with a 4X objective, as I have done with the Abbe condenser.

Acknowledgement

I could not have done this study without the donation of the phase contrast condenser base by Chris Vander Tuuk of LOMO America. The historical background on Abbe condensers was provided by John Delly. Discussions with John and his reviews of the test results also contributed to this article.

References

1. Clarke, T. M. Building an Affordable Universal Student Microscope. The Microscope, 2000;48:19-39.

2. James, P. “Circular Oblique Lighting” Balsam Post, 2003;61:3-10.

3. Clarke, T. M. (2001). A versatile low power microscopy set-up with a modified Russian Biolam stand. Micscape, September 2001, http://www.microscopy-uk.org.uk/mag/artsep01/tcmacro.html

4. McCrone, W. C. The Microscope, 1985;33:87.

5. Herlihy, E. P. An Amateur’s Retrospect. The New York Microscopical Society Annual, 1968:14-21.

6. McCrone, W. C. Response to Letter-to-the-Editor from Oppenheimer Goldberg. The Microscope, 1985;33:71.

7. Ellam, B. Diatoms by Dark-ground Illumination-with Some Historical Musings-Part II. The Amateur Diatomist, 2003;1
(4):33-42.

8. Lonert, A. C. Turtox Microscopy Booklet, General Biological Supply House, 1962.