Forays into fluorescence

Simple transmitted blue light autofluorescence of mosses and algae imaged with a digital SLR

 by David Walker, UK

Professional fluorescence microscopy typically uses dedicated high intensity lamps with epi-equipment and more recently sophisticated laser techniques, but simple transmitted light fluorescence studies have been described in the literature for the enthusiast using a compound microscope and suitable quartz halogen lamp, filters and darkfield (ref. 1 - 3).

Another approach now possible for the enthusiast is the use of low cost, high power LEDs together with sensitive but affordable cameras. Ely Silk has provided an excellent comprehensive resource on his website www.viewsfromscience.com with detailed practical advice and striking examples of the results. Some microscope makers (e.g. Zeiss, ref. 8) also now offer LED based fluorescence models; third party kits e.g. by Fraen for transmitted light LED fluorescence are also available to retrofit to some modern microscopes.

I hadn't tried any sort of fluorescence studies before, so for my first foray I was interested in basic transmitted studies of selected plant autofluorescence using a microscope's 100W quartz halogen lamp for visible light excitation using darkfield with filters suggested in the literature; I was particularly keen to see whether my digital SLR was able to capture the weak images, given the high ISO and low noise performance of modern DSLRs. A longer term project is to see how the imagery from such a setup compares with those from a microscope with an epi-fluorescence head.

Basic plant chlorophyll autofluorescence can be studied using a deep blue excitation filter and an orange barrier filter (ref. 1). Leaves from mosses, liverworts and freshwater algae might be good practice subjects for such studies as they are easy to mount and give good imagery of the chlorophyll containing plant structures. A superb book which includes three chapters on the autofluorescence of plant, animal and other subjects is Volume II of Rost's 'Fluorescence Microscopy' (ref. 6).

Filter choice
Wratten gelatine filters have been recommended and can be cut to suit, stacked if needed and can be protected from handling by mounting between glass (ref. 3). Small modern glass equivalents are now available e.g. from Astronomica, often a comparable price and are more robust to heat and handling but there may be fitting limitations of glass filters above the objective depending on microscope. The author's Zeiss Photomicroscope accepts a 25 mm diameter filter in the main accessory slot above the objectives. Lee Filters made of acetate could also be tried, the filters' spectral characteristics are on the back of e.g their 'ColourMAGIC' starter packs.

The Schott glass filters are an alternative and are available from optical and scientific suppliers, typically ca. £17 each in UK for 25 mm diameter, 3 mm thick (Knight Optical, UK). Although a higher initial outlay, the barrier / exciter filters in dedicated epi-fluorescence heads (e.g by Zeiss and Leitz) were often of the Schott type so the maker's fluorescence literature is useful for choosing filters for an excitation light of interest (or those suitable for a given fluorochrome dye) (ref. 4, 5).

I opted for two filter approaches summarised in table below; one set using filters to hand from the spares box but not necessarily well matched, the second a Zeiss matched set*. Only blue light excitation was studied.

*These were salvaged from a Zeiss III RS epi fluorescence head where all filters and dichromatic beam splitters had suffered surface damage, possibly from heat, destroying the delicate beam splitters, but the filters after home re-polishing looked excellent.
(Each filter works on the epi head with its matched beam splitter which reflects the excitation light onto the subject and only transmits the emitted light, but using darkfield for transmitted studies helps prevent excitation light reaching the eyepiece.)

* Zeiss III RS filter data from ref. 4.
** A red suppression filter is often a typical filter in epi-fluorescence outfits because some excitation filters as well as transmitting in a chosen narrow band can also emit in the deep red / near IR, e.g. the BG3 (ref. 4).
*** Filter thickness and / or stacking same filter can be used to reduce residual unwanted light in a filter's spectrum, Zeiss use two BG12 filters in the III RS for blue excitation.
**** A heat blocking filter, typically a KG1 is also a vital filter for typical epi-fluorescence with mercury arc and xenon lamps, possibly less important for transmitted darkfield with a typical 100W quartz halogen microscope lamp as none was fitted as standard on own scope.

Initial setting up trials

Excitation with transmitted brightfield versus transmitted darkfield
Darkfield is recommended for dyed filters to minimise direct light reaching the eyepiece which may swamp weak fluorescence (ref. 2,3). But an example using blue excitation autofluorescence of algae is shown below in brightfield (left) and darkfield (middle, right) images. Brightfield shows the residual blue light passing the exciter and barrier filters but no observable fluorescence of the algae. Darkfield blocks direct light and effective at showing red emission from the chloroplasts and now a deep black background.

The third image shows the use of one BG12 filter in the Zeiss set rather than the stacked two for the middle image. The residual light passing the filters is significantly increased and passes a pale turquoise blue. The third image is thus a combination of darkfield in this light (of the cell walls) and the red fluorescence tonality modified by the blue. The combination lighting was useful for showing both non-fluorescing and fluorescing features for some subjects.

Subject: Fresh filamentous freshwater algae, aqueous mount.
Objective: Zeiss 6.3X, NA0.16 planachro'.
Filters: Zeiss blue, 100W halogen lamp, full intensity.
Condenser: Zeiss brightfield 'J' setting, 'D' setting on phase, water immersion
Camera exposure: ISO 800, 20 secs.
Comments: Images out of camera, apart from resize.

Just a darkfield image or subject autofluorescence?
With darkfield confirmed as the best lighting, a subject was tried that had little to no autofluorescence but active in darkfield. From trials, processed cellulose fibres from toilet tissues showed very little blue light fluorescence so used as a test subject. (Silica diatom frustules seemed an option but found all my prepared slides had mountant autofluorescence). Darkfield should be illuminating any refractive part of the subject in the filter set leakage colour, whereas fluorescence will only show for plant structures that fluoresce under the excitation light chosen, and in the colour expected e.g. typically red for plant chloroplasts.

Processed cellulose fibres as a weak to non-fluorescing control for blue excitation. Zeiss 16x/0.35 objective, 100W lamp full intensity except for lefthand image.
Left: normal darkfield with a pale blue filter. (1/20th sec, ISO 800)
Middle: darkfield with the Zeiss blue excitation filter set (2 x BG12, BG38, '50'), light leakage minimal. (20 secs, ISO 800)
Right: darkfield with the Zeiss filter set but just 1 x BG12.  (20 secs, ISO 800). Filter leakage significant to show a darkfield type image. But as remarked on below, this was found useful to delineate the non-fluorescing features of a subject with fluorescence included. (Mixed lighting is commonly used, e.g. epi-fluorescence with transmitted phase and DIC.)

Visible studies
A darkened room and dark adapted eyes are recommended for these sort of studies; I did see a weak image by eye with binocular head under these conditions for the plants studied and normal NA objectives, but personally found it impractical both from an inability to really see detail and of microscopy in an almost dark room. This is where a dedicated fluorescence microscope with high intensity mercury arc and high NA objectives would be better. Although I'm fortunate to have access to one, it does not have a standard blue filter cube so direct comparisons weren't possible.

Digital imaging
I used a Nikon D300 DLSR primarily, a current mid-priced model with CMOS APS-C sensor but see quick comparison with a Nikon D50 DSLR below. It has ISO 200 - 3200 as standard settings and shutter speeds to 30 secs plus bulb; also a Live View feature using the main sensor. It was used exclusively with Nikon's Control Pro 2 software to remotely control and instantly transfer and assess the results on a PC.
Focussing - despite weakly visible fluorescent images, there's an easy workaround to focus. Just prior to a fluorescence exposure, the filters were replaced with a single red filter on the field lens to give standard darkfield imaging, plenty bright enough to focus using Live view (or camera viewfinder or parallax eyepiece if preferred). As the colour was in a similar band as that emitted under fluorescence the focus plane should be close to if not exactly correct.
Exposure measurement was by trial and error, but as for normal photography it's surprising how quickly the correct exposure can be judged to within a stop or so. On the author's scope / camera combination, ISO 800 and exposures from 10 - 20 secs coped with most of the subjects tried with 12V / 100W lamp always set at full rated voltage and was a good compromise between greater noise on higher ISO and excessively long exposures. I didn't use the camera's built-in noise reduction software.

Moss leaf with epiphytic algae. Left: Nikon D50, Nikon's entry level model 2005, right Nikon D300, current mid-range model.
Out of camera, resized. 15 secs, ISO 800. Zeiss 16x/0.35 objective. Zeiss blue excitation set. Oiled 'D' darkfield.

Detail from above. Crops and resized to be comparable. Left Nikon D50 (6 Mpixel, CCD sensor), right Nikon D300 (12 Mpixel CMOS sensor).

An older budget DSLR like the D50 seems competent from a quick comparison, although the D300 seemed to pick up the weak green for this example. 'Pixel peeping' showed more colour noise if lightened the black at 100% images cf the D300, but unlikely to see in web images and typical print sizes.

Some Results

Lamp: Zeiss 12V, 100W transmitted microscope lamp at full intensity. All light sent to camera on Photomicroscope beam splitter. Optovar 1.25X.
Camera: Nikon D300 DSLR. White balance 'Auto'. Images all taken as jpegs (i.e. 3x8 bit RGB, the RAW image offers 14 bits a channel). 'Vivid' mode set to combat the camera's tendency to flatten colours on a microscope when jpeg set to 'neutral'.
Condenser: Darkfield 'D setting on phase condenser, water or oil immersed. There didn't seem much observable advantage in using oil on the Zeiss phase condenser, so used water in the main. A trial of the Zeiss high NA ultra darkfield oil immersion condenser gave about a stop better exposure but no marked difference otherwise, so phase condenser with darkfield setting was found more practical.

Filter sets:
Zeiss blue: Excitation - 2xBG12 (each 3mm thick) + BG38 red suppression. Barrier - 'filter 50' pale yellow.
Blue set 1: Excitation - BG3 (3 mm thick)  + CM500S red suppression. Barrier - OG550.

Comments to date
From trials to date the images looked quite usable from modest NA objectives, giving genuine fluorescence and providing greater insight into plant structures. The backgrounds were a good black with little to no post-capture image processing, suggesting the 'D' setting of the Zeiss condenser was working well with the filter pairs used. The in-optic darkfield setting of this type of condenser could well be superior to a simple patch stop in filter tray. Both a modern and earlier model DSLR tried seem very capable of imaging at the mid range ISO settings and long time exposures required.

Apparently plant tissue autofluorescence is typically much brighter than for animal tissues (ref. 6a), so the subjects chosen above may have been within the capabilities of the transmitted set-up with a DSLR.

Simple blue light fluorescence studies following procedures recommended in the literature has certainly opened up to the author a new way of studying suitable plant subjects with imagery totally different to that by brightfield and normal contrast enhancement. For readers with a DSLR and microscope with darkfield and halogen lamp, the outlay of suitable filters is worth considering.

Comments to the author David Walker are welcomed.

References and Resources
1) W. Nachtigall, 'Exploring with the microscope', Sterling Publishing, 1995, p. 40.

2) W. G. Hartley, 'Fluorescence in amateur microscopy', Quekett Journal of Microscopy, 1978, 33, 372 - 375. 

3) D. J. Cook, 'Fluorescence microscopy using gelatine filters and quartz-iodine illumination', Quekett Journal of Microscopy, 1980, 34, 113 - 118. This author used a Leitz 35 mm slide projector as light source.

4) Zeiss West, 'Epi-fluorescence condenser III RS. Operating instructions.', 12 - 13. Downloadable from Zeiss website and www.science-info.net.

5) Leitz Wetzlar, K. Koch, 'Fluorescence microscopy. Instrument, methods, applications', 1972, 10 - 16. Downloadable from www.science-info.net. Includes an excellent practical introduction to fluorescence microscopy.

6) F. W. D. Rost, 'Fluorescence microscopy', vols. I and II, CUP, 1992. Obtainable by interlibrary loan or a local university library may have it and can also be 'dipped into' on Google Books: 'Vol. II, Ch. 10, 'Autofluorescence in plants, fungi and bacteria'. Vol. I includes a chapter 'Commercial fluorescence microscopes' covering both older and more recent designs, some of which may be in the amateur's budget on the used market.

6a) Above ref., Vol. II, page 16.

7) B. Herman, 'Fluorescence microscopy', Bios Scientific Publishers, RMS Handbook No. 40, 2nd edition 1998, p. 19.

8) B. Hohman, Carl Zeiss, 'LED Light Sources: A Major Advance in Fluorescence Microscopy', Microscopy and Analysis, November 2008.

9) G. Nordhorn-Richter, 'Primary fluorescence of mosses', Leitz Scientific and Technical Information, 1984, 8, 167 -170. A useful ref. from Rost's book, ref. 6.

Microscopy UK Front Page
Micscape Magazine
Article Library

Please report any Web problems or offer general comments to the Micscape Editor .