UV-A-Induced VIS Fluorescence—Methods

Warning

Ultraviolet-A (UV-A) radiation at high irradiance can be damaging to eyes and skin, especially when there is little or no visible light together with it. Eye protection is recommended in all cases but is especially important when using Xenon flashes as light sources as they emit all the way down to UV-C. It is much safer to use only the so called “black light” corresponding to UV-A1 at wavelengths near 365 nm. However, even in this case when using flashlights or any other sources of UV-A that emits a concentrated beam, use of safety goggles is a must. When selecting goggles or safety spectacles pay attention to their UV-protection rating (e.g., UV400 CE, or increasing protection from ANSI U2 to U6 markings) as some goggles are certified to protect only from mechanical impacts (e.g., ANSI Z87+ with no U marking) (Figure 1). It is important that they provide protection also from the sides.

A: In white light.

B: In ultraviolet-A radiation.

Figure 1: Protection eye wear protects the eyes by blocking all UV-A radiation from reaching the eye. The ones in the photographs are simple ones, given away by an optician to me. They are anyway effective, looking black and casting a dark shadow on the white background. Photographs taken with a full-spectrum converted Olympus E-M1 camera, using either no filter on the lens and Sunwayfoto FL-96 fill light video lamp or a Jaxman Uc1 UV-A flashlight filtered, and a stack comprised of filters ZWB2 stacked with TSN575 on the lens. Lens: Sigma 30 mm 1:2.8 DN MFT. The background was a 5 mm-thick slab of PTFE. A black reference is included in the top-right corner to facilitate consistent editing.

Viewing photographs

Clicking on photographs opens a gallery view with them displayed at a larger size and at higher resolution. This view also allows navigation among multiple photographs from a given Figure.

Fluorescence

Fluorescence occurs when a molecule absorbs radiation and re-emits the absorbed energy also as radiation. The usual case is the capture of a single photon, dissipation of a (small) fraction of the energy as heat and the emission of the rest of the energy as a new photon. As the emitted photon carries less energy than the absorbed one, fluorescence takes place at a longer wavelength than the absorbed radiation.

Both UV and visible light can induce the fluorescence of chlorophyll in plants. Strong fluorescence is emitted at wavelengths of 680 nm and longer. We do not see the fluorescence in normal daylight as the high sensitivity of human vision with a peak in green prevent it. With a camera modified to increase sensitivity to infra-red radiation and a filter that blocks wavelengths shorter than 680–700 nm the fluorescence can be imaged. A flash of strong light is most effective and is used in the study of plant’s photosynthesis (Figure 2).

Figure 2: NIR chlorophyll fluorescence induced by a flash of white light in plants of fava beans. Camera: full-spectrum converted Olympus E-M1, Sigma 30 mm 1:2.8 DC DN MFT lens, Heliopan RG680 filter on lens. Strong white light illumination from a device used to study photosynthesis of plants.

The quantum yield of fluorescence depends on the molecules involved. In most cases the fraction of the incident radiation that is emitted as fluorescence is small, frequently not even detectable. When we illuminate a plant or object with ultraviolet-A (UV-A) radiation only some of its components, if any, fluoresce strongly, emitting visible light (VIS).

Thus, to induce fluorescence across the whole visible spectrum we have to use radiation of shorter wavelengths, such as UV-A radiation. Our eyes’ low sensitivity to UV-A radiation and easily available eyeglasses and goggles that block all UV radiation help in allowing us to see the fluorescence without it being overwhelmed by the radiation used as excitation.

Different lichens contain pigments that fluoresce brightly when exposed to UV-A radiation (Figure 3). Colours of the fluorescence vary between yellow, red, occasionally blue and rarely green.

A: In white light.

B: In ultraviolet-A radiation.

Figure 3: Lichens on the trunk of an aspen tree, photographed at night illuminated with either white light or ultraviolet-A radiation. Camera: Olympus E-M1 Mk II, Sigma 30 mm 1:1.4 DC DN MFT lens. A: White light illumination from a white LED light source, B: UV-A illumination (365 nm Convoy 2+ flashlight with VIS-blocking filter) and Tiffen Haze 2A + Firecrest UV400 filters on lens. Photographs white-balanced and edited in Capture One Pro version 16.5.

Can the flourescence be emitted at a shorter wavelength than the excitation?

It can in exceptional cases when the energy acquired by a molecule by absorbing more than one photon is emitted as a single photon. In such a case, even if there is thermal energy dissipation, and the emitted photon carries less energy than the sum of the absorbed photons, it will carry more energy than each absorbed photon individually. This happens only in exceptional cases because the re-emission is normally fast.

Photographing fluorescence

In many respects photographing fluorescence is similar to photographing in low-light. It usually requires long exposure-times and use of a tripod or other camera support. In addition we need to make sure that what we photograph is really fluorescence emitted by the object being photographed rather than contamination by excitation radiation, ambient light or even fluorescence by nearby objects.

Is it really fluorescence?

We can check if the light in the image is from the light source or not by including a “white” reflectance reference ([#fig-lichens-fluo-02]). I use a slab of white PTFE (Teflon) about 5 mm thick. It should have a matte surface and be clean, as dirt can easily fluoresce. (I use wet/dry sandpaper 600 grit under running water from time to time to clean it.) If the white slab appears darker, nearly black, in an otherwise brighter image, we can conclude that what has been recorded is fluorescence.

A: In white light.

B: In ultraviolet-A radiation.

Figure 4: Lichens on a rock. To the left of the rock the white PTFE slab can be seen. Camera: Olympus E-M1 Mk II, Sigma 30 mm 1:1.4 DC DN MFT lens. A: White light illumination from a white LED light source, B: UV-A illumination (365 nm Convoy 2+ flashlight with VIS-blocking filter) and Tiffen Haze 2A + Firecrest UV400 filters on lens. In B: a weak blue-violet shine on the reference slab was subtracted from the whole image. Photographs white-balanced and edited in Capture One Pro version 16.5.

Location and surroundings

As fluorescence is weak compared to other light sources it can only be observed and photographed in the absence of other sources of visible light. Outdoors, normally at night and far from street lamps and other light sources, or in a darkened room indoors. The main problem indoors is the presence of fluorescent objects other than those of interest. These include paper, light-coloured and white cloths and clothes. The reason why UV-induced blue fluorescence is pervasive indoors is that blue-fluorescing compounds are normally added to white paper and some white plastics so that they look whiter and brighter. In fact, these chemical compounds are frequently called whiteners. Many laundry powders and liquids contain whitening and/or brightening agents that remain on clothes after washing them. Less frequently, white and light-coloured paints can be formulated with fluorescent substances in addition to white pigments that reflect visible light.

Note

Post-it and similar yellow notes do not fluoresce much when exposed to UV-A radiation and can be used for labels.

UV-A sources

I find UV-A LEDs most useful as UV-A sources as they intrinsically emit little VIS radiation. Little is anyway in most cases too much when photographing fluorescence and they need to be combined with optical filters that effectively block visible radiation. Flash lights and other small sources of UV-A are readily available as they are used for checking authenticity of bank notes and cleanliness (urine fluoresces strongly). LEDs emitting at a wavelength of 365 nm are suitable (Figure 5), while those emitting at longer wavelengths are not as they emit too much violet and blue light. LEDs emitting at wavelengths shorter than 360 nm tend to have much weaker light emission, making then cumbersome for regular use.

autoplot(leds.mspct$LedEngin_LZ1_10UV00_365nm, geom = "spct")

Figure 5: Emission spectrum of an UV-A LED with nominal peak of emission at 365 nm. LED type 3W SMD LED type LZ1-10UV00 from LED Engin.

Black-light-blue and UV-A fluorescent tubes can also be used indoors, or where mains power is available. They usually emit over a broader range of wavelengths than individual UV LEDs, something that can sometimes be an advantage Figure 6. Many of them emit a significant fraction of the total radiation in the visible. This makes fluorescent tubes less suitable as blocking the unwanted VIS and NIR radiation is more difficult than with UV-A LEDs.

autoplot(lamps.mspct[c("QPanel.FT.UVB340.40W", "Eiko.F36T8.BLB")], 
         geom = "spct", idfactor = "Lamp", facets = 1, span = 101)

Figure 6: Emission spectra of black-light-blue (BLB) fluorescent tube and of a UV-A fluorescent tube.

In principle a flash modified to emit UV-A radiation could be used as a light source. Xenon lamps do emit over a very broad range of wavelengths, making them more difficult to filter Figure 7.

autoplot(lamps.mspct$Godox.XeF.AD200.H200.flash, geom = "spct", span = 51)

Figure 7: Emission spectra Xenon flash lamp with a glass envelope and no filter. Godox AD200 flash with a non-standard flash bulb in a bare-lamp head.

To be able to photograph fluorescence isolated from VIS and NIR radiation, without contamination from the UV-A sources, it is necessary to use UV-pass VIS-blocking filters on the light source. Convenient light sources for use in the field are UV-A flashlights Figure 9. When exposure time is relatively short or even illumination is desired, a flood flashlight works better. In the case of long exposure times and use of light painting the more concentrated beam from the Convoy 2+ can be useful. The Convoy 2+ is also suitable for illuminating small objects. Both flashlights have Nichia LEDs with a nominal peak of emission at 365 nm. The exact types were not specified when I bought the flashlights, but at least the Jaxman Uc1 most likely has a NVSU233B(T) LED from Nichia.

A: Convoy S2+ with a 3 W Nichia UV-A 365nm LED.

B: Jaxman Uc1 flashlight with a 6 W UV-A 365nm LED attached to an ad-hoc holder.

C: Jaxman Uc1 flashlight with a 6 W UV-A 365nm LED attached to an ad-hoc holder.

D: Jaxman Uc1 flashlight with a 6 W UV-A 365nm LED detail view of the VIS-blocking filter.

Figure 8: Photographs of two very similarly looking UV-A flashlights, that differ mainly on the LED power and optics. The Jaxman is a “flood” light with a broad and even beam and the Convoy is a “spot” light with a narrower beam that is less even. Both flashlights were bought from AliExpress sellers. The Jaxman was delivered with the VIS-blocking filter installed, and the Convoy was delivered with a clear glass that I replaced with a filter bought separately, also from an AliExpress seller.

autoplot(lamps.mspct$Convoy.2plus.LED.UVA.flashlight, geom = "spct")

Figure 9: Convoy 2+ flashlight with a 3W Nichia LED emitting at 365 nm and a ZWB? UV-pass VIS-blocking replacement filter installed. This flashlight emits a narrow beam of high intensity light with a very obvious hot-spot at the centre.

autoplot(lamps.mspct$Jaxman.U1c.LED.UVA.flood.flashlight, geom = "spct")

Figure 10: Jaxman U1c flashlight with a 6W LED emitting at 365 nm with a built-in filter of unknown brand and type. This is a “flood” flashlight with broader and more evenly illuminated beam.

Choosing an UV-A flashlight

There are many different UV-A or black-light flashlights available. Most suitable are those emitting at 365 nm, because those emitting at 385 nm and longer wavelengths emit much more visible light, that will either interfere with the recording of the visible fluorescence or be blocked by the necessary UV+blue-blocking filter. It is also important to consider the beam uniformity and breadth. Finally, the power of most flashlights is not given by sellers as energy flux of UV-A radiation but instead as electrical power consumed. In some cases the power values given are clearly inflated, possibly based on the maximum power the LEDs withstand rather than the power at which they work in the flashlight or are simply imaginary.

Convoy 2 and 2+ flashlights

The Convoy 2 and 2+ flashlight have been available from some years. They are modular, available in many different configurations and even as parts for self assembly. Flashlight using the same metal body and other parts are sold under other brands. For example, the Jaxman is very similar to my Convoy 2+ but it has, apparently, a different LED and or optics, and a fifferent LED driver.

The same body Convoy 2+ is sold with different LEDs installed, different wavelengths, different powers, and with different packages. There are two types of reflectors, clear glass, AR-coated glass, and UV-pass VIS-block windows. Drivers differ in the maximum current, that needs to be matched to the LED used, and in the user interface. The user interface can allow dimming or not, and even in some cases have a bliking option. Some drivers have a single user interface and others have multiple ones to choose from. One has to be careful with the driver and LED, as a driver that feeds a very high current to the LED will heat the flashlight. Several of the drivers have over-heating protection that decreases the power as needed to avoid overheating. This works when use is briefly, but for a longer photography session one want constant illumination.

All in all it is best to buy a ready assembled Convoy flashlight with Nichia UV LED with peak at 365 nm and an UV-pass filter. It is possible, however, to buy the parts separately, even from different suppliers and assemble a custom version with a specific LED of one’s choice. When I bought the Convoy 2+ it was not available with a pre-installed filter. Nowadays there are multiple UV-A versions, and those using LEDs from other brands than Nichia are cheaper.

Configurations ready available are with white light of different colour temperatures and powers, UV-C. UV-A, blue, green, read and infra-red. Convoy has an official store in AliExpress (“Convoy Flashlight Store”) that has this and several other flashlights.

The standard configuration uses a standard 18650 rechargeable Lithium battery. The body of the flashlight has a diameter of about 20 mm and is 80 mm long. To hold it I am using a holder that is sold to attach small cameras to the handlebar of a bicycle. It can be found in AliExpress under the name of “1/4 Screw Metal Cycling Bike Mount Motorcycle Handlebar Holder for GoPro 13 12 11 10 9 8 Insta360 X3 X4 DJI Osmo Action 4 5 Pro” by multiple sellers. This adds as attachment point a 1/4”-20 male thread, the standard size used for cameras. In the photograph, I have attached a male “flash cold shoe”.

Exciter filter

As mentioned above, if we want to photograph fluorescence across the whole visible range, the light source used for excitation should not emit any radiation in the visible range. In fact, more generally we need to use complementary filters in the light source and camera. Most frequently we aim at a cross-over of the transmittances so that neither of the two filters transmits at 400 nm. In practice, we need a “dead-zone” as the cut-off and cut-in wavelengths of filters have tolerances of a few nanometres and the there can be gradual change in transmittance within a range of wavelengths. We look first at the filters commonly used on light sources or “exciter filters”.

Digital cameras have built-in filters on their sensors, but these filters let through some UV-A radiation. Additional filters must be chosen considering the light source in use. For example, many UV-pass filters are not effective at blocking the near infra-red radiation (NIR) that some cameras can “see” Figure 11. Fluorescent lamps emit some NIR radiation while UV-A LEDs do not emit it ([compare @ig-Qpanel340-lamp to @fig-LedEngin365nm]).

autoplot(filters.mspct[c("Tangsinuo_ZWB1_2.1mm_52mm", "Tangsinuo_ZWB2_2.0mm_52mm")], 
          geom = "spct", idfactor = "Filter", span = 51)

Figure 11: Spectral transmittance of some short-pass UV filters. These are both glass filters not anti-reflexion coated.

Filter types ZWB1 and ZWB2 from various Chinese suppliers are cheaper than the German made Schott UG1 and UG5. Specifications are less tight for the Chinese filters, and can sometimes present optical defects. ZWB1 is a filter type for use with UV-A LEDs. These filters are usually fine if bought from reliable suppliers. Filters ready cut to size for common flash light types are readily available through AliExpress and eBay sellers.

The sensitivity of cameras to NIR radiation varies. In LED UV-A light sources it is best to use a VIS-blocking filter.

Visible light as excitation

If we are interested in red or NIR fluorescence we can use blue- or UV + blue radiation for excitation, with suitable filters and LEDs or lamps. As with UV short-pass filters, blue short-pass filters block NIR completely very poorly Figure 12. So, once again, easiest it is to use LEDs as light sources.

autoplot(filters.mspct[c("Tangsinuo_ZB1_2.0mm_52mm", 
                         "Tangsinuo_ZB2_2.0mm_52mm")],
         geom = "spct", idfactor = "Filter", 
         annotations = c("+", "wls"),
         span = 101)

Figure 12: Transmittance spectra of two types of blue glass filters.

Barrier filter

Barrier filters are used in front or behind the camera lens to block the excitation light completely. In this case there is an additional difficulty, yellow, orange and red glass filters themselves fluoresce when exposed to UV radiation. The solution is to use a UV-blocking filter that either reflects UV-radiation, i.e., interference filters or based on a different pigment than that used in glass filters, in front of the glass filters Figure 13. Kodak publications from the film era recommend using a Kodak Wratten 2A or 2B filter in front of other barrier filters. Tiffen still sells filters very similar to the Kodak ones. In these filter a coloured light-absorbing film, in most cases using gelatine as medium for the pigment, is encased between two layers of clear optical glass. These filters used to be also available as bare gelatine films. The Zeiss UV T* filter is based on interference and reflects radiation below 405 nm, it is AR coated and less affected by reflections at the transmitted wavelengths than the Tiffen. The Zeiss T* filter prevents the fluorescence of other glass filters as effectively or better than the Tiffen Haze 2A, but it is rather expensive.

autoplot(filters.mspct[c("Tiffen_Haze_2A_2.6mm_52mm", 
                         "Zeiss_UV_Tstar_2.0mm_52mm")],
         geom = "spct", idfactor = "Filter", 
         annotations = list(c("-", "peaks"), c("+", "wls")))

Figure 13: Two UV-blocking filters that do not fluoresce when exposed to UV-radiation. A Tiffen Haze 2A, equivalent to Kodak Wratten 2A and Zeiss UV T-* filter. Only the Zeiss filter has an anti-reflection coating.

In some cases we may want to photograph fluorescence only of some colours. If we are interested only in longer wavelength we can use a long-pass filter, behind a UV-blocking filter Figure 14.

autoplot(filters.mspct[c("Hoya_Y_(K2)_HMC_2.3mm_52mm",
                         "Heliopan_Yellow_5_SH_PMC_2.3mm_52mm", 
                         "Heliopan_Orange_22_SH_PMC_2.2mm_52mm",
                         "Heliopan_Red_25_SH_PMC_2.2mm_30.5mm",
                         "Heliopan_RG665_2.3mm_46mm")],
         geom = "line", idfactor = "Filter", 
         annotations = list(c("-", "peaks"), c("+", "wls")))

Figure 14: Spectral transmittance of some long-pass VIS filters. These are all glass filters with anti-reflexion multi-coating.

With a camera modified to “see” a broader spectrum by replacement of the sensor filter, it is possible to photograph NIR fluorescence by blocking visible fluorescence with long-pass barrier filters Figure 16. Compared to an unmodified camera leaks of NIR from the excitation radiation source become an even bigger problem. NIR long-pass filters require “protection” from UV to prevent their fluorescence.

Fluorescence of glass filters

Yellow, orange, red and NIR long-pass glass filters fluoresce when exposed to UV-A radiation, most strongly the yellow filters. Yellow filters do emit a very strong yellow fluorescence. The fluorescence of orange and red filters is not so intense at least to the naked eye. The fluorescence of NIR filters is in the IR and not visible to a normal camera or the naked eye. Using a full-spectrum converted camera and a long-pass filter on the lens, we can see that all four glass filters fluoresce in the IR. To prevent this an additional uV-reflecting or absorbing filter must be stacked in front of these coloured glass filters.

A: Filters in white light.

B: Filters in UV-A (365 nm) radiation, visible fluorescence (true colour).

C: Filters in UV-A (365 nm) radiation, near-infrared fluorescence (false colour).

D: Filters in UV-A (365 nm) radiation, near-infrared fluorescence (false colour).

Figure 15: Photographs of four long-pass glass filters under white visible and ultraviolet-A radiation. The background is in all photographs the same slab of white PTFE (Teflon), but in B–D the filters used on the camera lens blocked the UV-A radiation used to excite fluorescence. Clockwise from the top left corner: Hoya Red 25A (603 nm), Hoya R72 (716 nm),Heliopan Orange 22 (568 nm) and Heliopan Yellow 5 (466 nm), all AR coated. Light sources: Sunwayfoto FL96 at 5500K LED fill light and Jaxman Uc1 flashlight. Cameras A, B: Olympus OM-1; C, D: Olympus E-M1 full-spectrum converted. For visible fluorescence barrier filter: Firecrest UV400 + Tiffen 2A, and for NIR fluorescence Firecrest UV400 + C:Zomei IR680 or D:Haida IR720 (726 nm) + Tiffen 2A. Lens: M.Zuiko 25 mm 1:1.2 Pro. Photographs white-balanced and edited in Capture One Pro version 16.5.

autoplot(filters.mspct[c("Heliopan_RG695_2.2mm_52mm", 
                         "Heliopan_RG780_2.3mm_52mm",
                         "Hoya_R72_2.4mm_52mm",
                         "Zomei_IR850_2.1mm_52mm")],
         geom = "line", idfactor = "Filter", 
         annotations = list(c("-", "peaks"), c("+", "wls")))

Figure 16: Spectral transmittance of some long-pass NIR filters. These are all glass filters, of which the Hoya R72 is anti-reflexion coated.

In principle band-pass filters can also be used, although this is a specialized situation. Of band-pass filters, the only sold for photography ready mounted on rings are the UV-IR cut filters. The actual wavelength boundaries vary rather broadly among brands and types Figure 17. In AliExpress and eBay sometimes unusual ones are offered.

autoplot(filters.mspct[c("Firecrest_UVIR_Cut_0.96mm_52mm", 
                         "Heliopan_UVIR_CUT_Digital_2.2mm_52mm",
                         "Rocolax_UVIR_Cut_445nm_650nm_1.1mm_52mm",
                         "Rocolax_UVIR_Cut_PRO_HD_(W)_1.1mm_52mm",
                         "Fotga_UVIR_CUT_0.54mm_52mm")],
         geom = "line", idfactor = "Filter", facets = 1,
         annotations = list(c("-", "peaks", "summaries"), c("+", "wls")))

Figure 17: Transmittance spectra of an assortment of UVIR-cut filters sold for photography and video.

Filters can be stacked, and this is frequently necessary with UV-pass filters that also transmit some NIR radiation.

A huge selection of band-pass filters mounted on suitable rings are available from suppliers of equipment for industrial machine vision. For example, MIDOPT (distributed in Europe by Stemmer Imaging), is a well known supplier, but such filters tend to be more expensive than those sold for photography. Square filters in a 2” \(\times\) 2” or 50 mm \(\times\) 50 mm size are common in the catalogues of optics suppliers like Hebo glass, UQG Optics, and some local optics shops, e.g., Teknofokus in Finland.

Cameras and camera settings

Normal cameras can be used for photographing the visible fluorescence. They are preferable to full-spectrum converted cameras as the built-in sensor filter help block unwanted UV and IR radiation. To photograph NIR fluorescence, an IR or full-spectrum-converted camera is needed.

A key requirement is that the camera supports long exposure times, i.e., that it has at least a Bulb setting for shutter speed as exposures may extend to minutes. Cameras that can display the live image as is “builds” during a long exposure are extremely convenient. In Olympus/OM-System cameras this mode is called and is what I always use for UVIVF photography in the field at night. In the case of close-up and macro photography it is possible to use exposure times that are shorter and to illuminate the whole view evenly.

A lens with a large aperture helps, but is not a must as long as subjects do not move. Any good lens is suitable as the UV radiation is used to excite the fluorescence and we block it with a filter before it enters the lens. Thus, it does not require the especial lenses and modified cameras that are required for reflected UV photography.

Close-up photography of fluorescence

Combining the close-up or macro-photography with ultraviolet-induced visible fluorescence in the field can be difficult as exposure times are long and focus stacking is frequently needed for subjects with relief. A few examples from last night can be seen in Figure 18.

A: In visible light.

B: UV-A-induced visible fluorescence.

C: In visible light..

D: UV-A-induced visible fluorescence.

E: In visible light..

F: UV-A-induced visible fluorescence.

G: In visible light..

H: UV-A-induced visible fluorescence.

I: In visible light..

J: UV-A-induced visible fluorescence.

K: In visible light..

L: UV-A-induced visible fluorescence.

Figure 18: Photographs of four dead tree branches, covered with lichens and mosses. The background is in all photographs the same black velvet cloth. The edge of a slab of white PTFE (Teflon) is visible on the top-right corner of the photographs. In, B, D, F, H, and J the filters used on the camera lens blocked the UV-A radiation used to excite fluorescence. Light sources: Sunwayfoto FL96 at 5500K LED fill light and Jaxman Uc1 flashlight. Camera: Olympus OM-1. For visible fluorescence barrier filter: Firecrest UV400 + Tiffen 2A. Lens: A–J M.Zuiko 25 mm 1:1.2 Pro, K-L M.Zuiko 90 mm 1:3.5 Macro Pro. Photographs white-balanced and edited in Capture One Pro version 16.5. Focus-stacks of 20 or 80 photographs were merged in Helicon Focus 8.2.18 Pro to increase the depth of field.

With suitable filters and lenses we can even attempt macro-photography of UV-A-induced chlorophyll fluorescence with a modified mirrorless camera (Figure 19).

A: In visible light.

B: UV-A-induced near-infrared flousrescence.

Figure 19: Moss on a dead branch. Illuminated with A: white light or B: UV-A radiation. Only difference with images in Figure 18 in the approach was adding a long-pass filter with a cut-in at 650 nm (HB650). Using a filter with cut-in at 695 nm (Heliopan RG695) resulted in a similar image but required a longer exposure time. Camera: full-spectrum converted Olympus E-M1 camera. See legend to Figure 18 for other details.

Conclusion

Photographing UV-A-induced visible fluorescence is interesting and less demanding in equipment and technique than photography in reflected UV radiation. An unmodified digital camera can be used and long pass filters are in general cheaper and more common than UV short-pass filters that block visible and infrared radiation effectively enough. UV-A as excitation is convenient and easy to work with, blue light could be used instead if we are interested in fluorescence at longer wavelengths or for safety reasons in teaching situations. Using UV-B radiation to induce UV-A and violet fluorescence is in principle possible but more demanding in practice. In fact, I have yet to attempt using wavelengths shorter than UV-A1 for excitation. I have just ordered a Convoy 2+ flashlight like that in Figure 8 A: but with a blue LED emitting at 450 nm instead of the more common UV-A or white light versions. Photographing visible-light induced near infrared fluorescence is more demanding in equipment, but opens interesting possibilities. Could we with comparatively cheap equipment use photography to qualitatively assess changes epidermal UV transmittance under controlled conditions?

Related content

In the following Flickr albums you will find additional examples of my UVIVF photographs: Lichens (Finland): UV-A-induced fluorescence, Mouldy fruit / Fruta enmohecida, Lichens (French Alps): UVA-induced autofluorescence, and Plants under UV radiation.