Home » Short Wave UV Imaging » diy Shortwave UV Image Converters for Solar-Blind and RUVIS Imaging

diy Shortwave UV Image Converters for Solar-Blind and RUVIS Imaging

David Prutchi, Ph.D. SWUV SBUV converters

DIY shortwave-ultraviolet imaging converters. Top: Converter based on fluorescent coating. Bottom: Short-wave ultraviolet viewer based on a surplus RCA UV image converter tube.

Imaging in the shortwave ultraviolet spectrum (wavelengths below 300 nm) enables some very exciting applications. Light in these wavelengths is completely invisible (but potentially very harmful) to the unaided eye. I am most interested in a band known as “Solar Blind UV” or SBUV. As shown in the following figure, solar radiation in the 240 nm to 280 nm range is completely absorbed by the ozone in the atmosphere and cannot reach Earth’s surface, thus allowing ultraviolet-emitting phenomena (e.g. electrical discharges, hydrogen combustion, etc.) to be detectable in full daylight.

Solar Blind Spectrum David Prutchi PhD

Certain ultraviolet-emitting phenomena, for example electrical discharges and breakdown in high-voltage power lines are almost undetectable in the visible range (a), but emit strongly in the UV-C region (b) which daylight lacks, allowing visualization in full daylight by superimposing the UV-C and visible images (c). d) The spectral content of sunlight is heavily modified by the Earth’s atmosphere, and shows various dips that correspond to different gases absorbing energy at different points in the spectrum. Ozone in the atmosphere absorbs very heavily in the UV-C band (100 nm to 280 nm), so virtually no sunlight in that wavelength range reaches Earth’s surface. The darkest part of that range – 240 nm to 280 nm is known as the “solar-blind” band. Pictures a-c courtesy of Eran Frisch, Ofil, Ltd

SBUV images don’t show the context in which the SBUV-emitting phenomenon occurs, so SBUV images are often superimposed onto a visible picture. Panel (a) of the following figure shows a simplified block diagram for a camera made by Ofil, Ltd. (www.ofilsystems.com) – an Israeli company that is the worldwide leader in the development of solar-blind UV-C imaging systems. In Ofil’s camera, light from the scene is split by a two-way mirror (a component that in optics is known as a beamsplitter) and directed simultaneously towards a visible color camera and to a SBUV camera. The latter comprises a lens that is transparent to shortwave-UV, one of Ofil’s specialized SBUV filters and an image intensifier with a SBUV-sensitive photocathode. The intensifier’s output is imaged by a monochrome CCD camera. The video streams from the visible and SBUV cameras are finally combined into one superimposed video channel that shows SBUV-emitting phenomena in context with the visible landscape in real time.

SBUV Camera Ofil David Prutchi PhD

Ofil, Ltd.’s solar-blind imaging camera is able to combine images obtained from a solar-blind UV imager (operating in the 240 nm to 280 nm range) and a visible-light camera (a). b) It is next to impossible to see a rocket-propelled grenade (RPG) being launched at a distance of 500 m in broad daylight. However, the rocket plume is clearly visible in the solar-blind UV band, allowing for immediate detection and identification of a launch site. Images courtesy of Eran Frisch, Ofil, Ltd.

A dramatic demonstration of the usefulness of such a system is shown in the lower panels. (b) shows a scene in which a rocket-propelled grenade is being launched at a distance of 500m from the camera, yet nothing unusual is seen in this daylight picture. However, it’s impossible to ignore the rocket’s plume when the SBUV image is superimposed onto the visible image as shown in (c)!

Ofil has also developed a system for detecting shoulder-launched missiles to civilian aircraft from terrorist threats.   In their system, a SBUV sensor detects the UV radiation emitted by an approaching missile and hands over its coordinates to a fine-tracking and jamming system that can release chaff and flares to deviate the missile. SBUV imaging is much better for this type of application than any other remote sensing technology because it allows unequivocal detection of imminent threats with very low false-alarm rate.

SBUV imagers are widely used worldwide for more mundane tasks such as locating electrical discharges in bright daylight, hence saving power distribution and electric-train companies millions by allowing them to detect corona discharges before an insulation break turns into a catastrophic failure. Chemical industries also use SBUV cameras to detect and locate fires caused by fuels such as hydrogen that don’t produce bright visible flames.

Imaging in the shortwave UV also has significant uses in forensics – most importantly, “Reflected UV Imaging Systems” (RUVIS) are used by CSI units to discover latent fingerprints under shortwave UV illumination (254 nm) without the use of powders or chemicals. Illumination is commonly provided by a portable SW-UV lamp, and an imager sensitive to this wavelength is used to produce a picture visible to the operator. Fingerprint residues containing oils and/or amino acids reflect and scatter shortwave UV, making them evident against most smooth, non-porous backgrounds that reflect shortwave UV poorly. Untreated sweaty prints show as bright reflective ridges on a black background, while oily prints appear as strong, dark ridges on a shiny background. This allows an investigator equipped with a sortwave camera or image converter to search, view, and capture latent prints not visible to the unaided eye. RUVIS is also used to enhance the contrast of faint or invisible latent prints that have been exposed to cyanoacrylate fumes. The microscopic fibers occurring with high humidity cyanoacrylate development (known as the cyanoacrylate bloom) reflect short-wave UV very strongly, so cyano-developed latent prints show very brightly under RUVIS against a jet-black substrate.

Imaging in the Shortwave UV

Unlike infrared and visible light, ultraviolet light has very little penetrating power into otherwise transparent or semitransparent materials. Because of its short wavelength, it is easily scattered by surface scratches and imperfections that are not apparent at longer wavelengths.   These characteristics make ultraviolet imaging an ideal inspection tool in production lines.

Unfortunately, these same characteristics make traditional camera sensors very inefficient at ultraviolet wavelengths because UV light can’t penetrate through the silicon bulk to the photosensitive sites on a camera sensor. Cameras specifically designed for ultraviolet or for very low light level imaging commonly use a “back-illuminated” sensor. Thinned-down back-illuminated sensors improve the sensitivity of a camera to UV light quite dramatically, but silicon imaging technology nevertheless reaches its sensing limit at around 300 nm.

A simple, inexpensive and very effective trick used by camera companies to enhance a camera sensor’s UV response for shorter wavelengths is to coat the sensor with a substance that fluoresces under ultraviolet light. Think about the way in which fluorescent paints glow brightly under “black light” – which is simply long-wave UV. Actually, the ultraviolet light has much higher intensity than the paint’s glow, but our eyes are significantly more sensitive to the fluorescence wavelength (e.g. neon green, pink, or orange) than to the virtually invisible 365 nm “black light”.

In the same way, a very thin fluorescent layer or “phosphor” (not to be confused with the element phosphorus) deposited on the sensor can be used to convert UV outside of the sensor’s range to a longer wavelength that the sensor can easily detect. One fluorescent coating used in early UV astronomy experiments was sodium salicylate – a close relative to Aspirin – which glows around 400 nm when exposed to light with wavelengths below 350 nm.   However, as shown in the following figure, sodium salicylate’s glow is on the edge of an image sensor’s sensitivity.   Later, coronene, also known as the aromatic hydrocarbon superbenzene, was used for UV astronomy because its fluorescence glow is more easily detected by a camera’s sensor. Modern cameras use composite phosphors made specifically for this purpose – the most common being Metachrome and Lumogen to increase the sensitivity of cameras in the 120 nm to 430 nm range.

SW-UV conversion Phosphors David Prutchi PhD

Fluorescence of a number of coating materials is used to extend the wavelength range of image sensors into the ultraviolet. These curves show the wavelength distribution of the glow that these materials output when illuminated by UV at wavelengths below 350 nm. Lumogen is the best because it glows at wavelengths that more closely match the sensor’s sensitivity.

Lumogen glows in the 540 nm to 580 nm range when illuminated by violet/ultraviolet light with wavelengths shorter than 450 nm. At wavelengths longer than 460 nm, the very thin layer of Lumogen applied to the sensor becomes transparent, allowing it to work normally in the visible and infrared portions of the spectrum. A problem common to Lumogen coatings is a steady drop in sensitivity with accumulated ultraviolet exposure. However, new Lumogen-type coatings such as Photometrics’ Metachrome II (www.photometrics.com) are known to remain stable for long periods of time under constant UV exposure.

The dedicated phosphors are not easy to come by just for experimenting. That is why I decided to mention sodium salicylate and coronene, which can be easily purchased from chemical supply houses, allowing an enthusiast to experiment with DIY coating of sensors.   However, applying these coatings usually involves evaporation of a very thin (in the nm range) layer of the phosphor under a vacuum, which may restrict this type of project only to advanced hobbyists.

There are a number of video cameras optimized for ultraviolet imaging that use sensors that are specifically designed for this application. Many of these are based on Sony’s ICX-407BLA CCD chip, which is a Lumogen-coated sensor usable down to 200 nm. These cameras are very expensive (over $3,000) – at least from my viewpoint as a photography enthusiast.

Shortwave-UV Image Conversion

Shortwave-UV cameras with lumogen-coated sensors are simply too expensive for the small number of things that I could imagine doing with one. There are many other pieces of photographic kit that I’d rather buy for that kind of money. I thus decided to build a simple image converter to photograph shortwave-UV phenomena by projecting the ultraviolet image onto a translucent screen coated with a phosphor that fluoresces in the visible spectrum in response to shortwave ultraviolet light. The visible image can then be viewed by a standard camera using a macro lens.

Fluorescent UV converter David Prutchi PhD www.UVIRimaging.com

A DIY shortwave UV image converter can be made by projecting the short-wave UV scene onto a piece of glass coated with a thin fluorescent coating that responds to shortwave UV photons by emitting light in the visible range (a). b) My DIY SW-UV camera consists of a C-mount UV-transparent lens preceded by a narrowband SW-UV filter (Edmund Optics model 67-809, 254 nm CWL 40 nm bandwidth) projecting onto a glass blank coated with a very thin layer of LDP LLC UVSWG fluorescent powder. The visible image is then captured with a conventional CCTV camera fitted with a close-focusing lens.

As the phosphor, I used one of the shortwave-UV fluorescent powders sold by LDP LLC (www.maxmax.com). These come in three different types, each responding to a different peak wavelength by fluorescing in a different part of the visible spectrum.

Table 1 – Shortwave-UV fluorescent powders sold by LDP LLC. These powders glow brightly when exposed to shortwave ultraviolet light, but do not respond to near-UV light.

LDP LLC Product Code Shortwave Ultraviolet Absorption Peak (nm) Visible Fluorescence Emission Peak (nm)
UVSWG 254 525 (green)
UVSWR 270 610 (red)
UVSWB 293 480 (blue)

These fluorescent powders are inorganic phosphors with a stable crystalline structure and an average particle size of 4 µm. I made dilute suspensions of these three powders in acetone (hardware store, from the paint thinner aisle) and just a tiny bit of Beacon 527 multi-use glue (from crafts store). To make a converter screen, I applied one drop of suspension at a time to a spinning 1” diameter glass blank and let it evaporate very evenly. I used a filtered shortwave UV lamp to gauge the thickness and evenness of the coating after each drop.

Commercial fluorescent UV converters are produced in a similar manner, and are commonly available as accessories for laser beam profilers – CCD cameras in which a laser beam is attenuated before shining onto a CCD to measure its beam profile directly.   Lasers that emit in the shortwave UV (e.g. UV excimer lasers) can’t be profiled directly by a silicon CCD, so fluorescent plates are available as accessories that convert UV radiation into visible light that is then imaged onto the CCD of the beam profiling camera.

Electronic Image Conversion and Intensifiers

Fluorescent image conversion relies on the energy of the incoming ultraviolet light to produce light through fluorescence that can thereafter be seen by the camera. This passive form of wavelength conversion is relatively inefficient, so it can be used only to image relatively strong shortwave ultraviolet sources.

An alternative to the passive fluorescent method of image conversion is to use an active electronic system to intensify the image during the conversion process. In an image conversion tube, like the one shown in the schematic diagram of the following figure, ultraviolet photons strike a photocathode, causing the ejection of electrons that are accelerated by a high-voltage power supply towards a fluorescent screen. Each incident photon that strikes the photocathode surface causes the release of many photons from the fluorescent screen, thus compensating for inefficiencies in the conversion process.

Image Converter Intensifier Tube David Prutchi PhD

In an image converter tube, invisible ultraviolet photons release electrons from a transparent metallic electrode. The electrons are accelerated by a high voltage power supply and strike a fluorescent screen to produce a visible image. This way, each incident ultraviolet photon causes the release of multiple visible photons, thus compensating for inefficiencies in the wavelength conversion process.

The first image-converter tube was developed in 1934 by G. Holst and his colleagues at Philips in the Netherlands, allowing infrared images to be seen in real-time for the first time. The same concept was applied by General Electric engineers in 1953 to develop an image converter tube for the ultraviolet solar-blind band.

1P25 UVIR image converter tube  David Prutchi PhD

The 1P25 is an early image converter tube sensitive to infrared and ultraviolet. It is relatively easy to find in the surplus market.

I built a shortwave-UV image converter using an image converter tube that I found on eBay®. The tube, made by RCA has a cesium-telluride (CsTe) photocathode and a quartz window, making it sensitive below 280 nm. Brand-new tubes with 180 nm to 1,350 nm sensitivity are sold by FJW Optical Systems (www.findrscope.com) as replacements for their FIND-R-SCOPE UV-viewers. Similar conversion tubes designed for infrared viewing, and which are widely available in the surplus market (such as RCA’s 6914) are also very sensitive in the near-UV, down to 300 nm. [6032 Image Converter Tube Specs]

diy RUVIS SW UV image Viewer David Prutchi PhD

My home-made shortwave ultraviolet viewer is based on a surplus RCA UV image converter tube. a) A simple power supply is used to bias the tube at around 15,000 Volts DC from a 9V battery. b) The image converter tube is enclosed within Thorlabs SM-2 optical tube components, and the 9V-to-15kV converter is housed in a Radio Shack plastic project box. A simple UV-transparent lens, suitable UV-pass filter, and eyepiece lens complete the viewer.

The power supply for the tube needs to produce around 15,000 Volts at low current. My home-built power supply based on a design by Bob Iannini for an infrared image converter tube. The circuit consists of a single-transistor oscillator powered by a 9V battery that drives a model 28K077 miniature ferrite high frequency transformer purchased from Bob’s company “Information Unlimited” (www.amazing1.com). The transformer outputs around 2,500 Volts AC which are rectified and multiplied by a number of high-voltage diodes and capacitors until the required 15,000 Volts are reached. The high-voltage positive output connects to the image converter’s tube electrode at the eyepiece end, while the ground terminal connects to the photocathode.   The tube also requires a focusing voltage at approximately 1/6 of the total potential. This is derived from some high-voltage resistors mounted directly on the image converter tube. The image converter is housed in an enclosure made of PVC pipe, and a simple UV-transparent lens and eyepiece lens complete the viewer.

My lens consists of a double-convex element made of UV-grade fused silica with 90% transmittance from 2,000 nm down to 200 nm (25.4 mm diameter × 60 mm focal length, Edmund Optics www.edmundoptics.com model NT46-294) housed inside a Thorlabs (www.thorlabs.com) model SM1V05 Ø1″ SM1 lens tube with ½” long external threads, which provides sufficient travel for focusing when mounted on a ½”-long SM1 extension tube (Thorlabs SM1L05).

Even higher sensitivity to shortwave-UV can be achieved by using an image intensifier tube. The simplest kind (known as “Generation I”) is not much different than an image converter tube, just that it is designed for higher efficiency, leading to moderate gain (a few hundred times). In contrast, later-generation tubes, like the ones shown in the diagram of the following figure, employ electron multipliers to boost gain. That is, not only the energy but also the number of electrons between the photocathode and the screen is significantly increased.

Gen II and Gen III Image Intensifiers David Prutchi PhD

Generation II and III image-intensifier tubes employ MicroChannel plate (MCP) electron multipliers to boost gain (a). b) MCPs are very thin plates of conductive glass containing scores of μm-diameter holes, c) in each of which a cascade of secondary electron emission occurs, directing an enormous amount of electrons towards the fluorescent screen for each electron entering the MCP.

Multiplication is achieved by use of very thin plates of conductive glass containing scores of minute holes (each a few μm in diameter) inside which a cascade of secondary electron emission occurs. Gains of one to ten million times are common with so-called MCP-based intensifier tubes, making infrared- and visible-light-sensitive tubes ideal for passive night vision scopes.

However, UV-sensitive intensifiers are less common because they are used mostly for highly specialized applications in solar-blind imaging. They are produced using MgF2 UV-transparent input windows and UV-sensitive CsTe or RbCsTe photocathodes. The fluorescent screen is often directly coupled to a silicon imaging sensor through a fiber-optic bundle to build what is commonly known as an “Intensified CCD” or ICCD camera. Lastly, a new technology competing with the sensitivity of ICCDs is known as “Electron-Multiplying CCD” or EMCCD in which secondary electrons are produced in the silicon chip to produce gain.

ICCDs and EMCCDs are also out of budget for most enthusiasts, but I believe they are worth mentioning because your curiosity in the field may lead you to a scientific or commercial application where the performance of one of these cameras justifies its cost.

Filters for Shortwave Ultraviolet Imaging

As previously mentioned, in “solar blind” band between 240 nm and 280 nm, solar radiation is completely blocked by ozone in the stratosphere, leaving virtually no in-band background radiation reaching the Earth’s surface. Without interference from background light, even very weak levels of ultraviolet light are detectable, making it possible to image electrical discharges, rocket plumes, industrial fires and other sources of short-wave-UV even in bright daylight.

However, the Sun is such a powerful light source that nearly all photons outside the 240 nm to 280 nm range must be eliminated to make it possible to distinguish the ultraviolet emitters from background light outside the SBUV band. Conventional optical filters cannot provide sufficient attenuation of sunlight outside the solar-blind band, so special filters based on dye-doped polymer films are needed to achieve the critical rejection of all out-of-band light. A popular filter for these applications is made by Ofil.   For example, their SB-BDF filter transmits a peak of 15% of ultraviolet light at 264 nm, while allowing only 10-6% of near-UV light to pass through. These filters are meant to be mounted on specialized, ultra-sensitive ultraviolet cameras.

Solar Blind UV filters David Prutchi PhD

Filters made to observe allowing shortwave ultraviolet-emitting phenomena (e.g. electrical discharges, hydrogen combustion, etc.) in full daylight typically allow 15% of ultraviolet light at 264 nm to pass through, while transmitting barely 10-6 % of near-UV light (a). A typical use of a solar-blind imager is in the detection of industrial fires. For example, an alcohol fire is almost undetectable in the visible range when it happens in broad daylight (b), yet it is easy to see if an image of its UV-C emissions is superimposed on the visible image. Photographs b and c courtesy of Eran Frisch, Ofil Systems.

Ofil’s filters and cameras are well beyond the typical enthusiast’s budget. Nevertheless, experimenting with shortwave-UV imaging is possible using a home-built image converter and a more affordable filter, like one of Edmund Optics’ UV bandpass interference filters (e.g. model 67-809, 254 nm CWL, 40 nm bandwidth). This filter won’t allow a camera to detect fire while pointing in the direction of the sun, but is good enough for exploring the solar-blind band. Edmund’s other “traditional coated bandpass interference filters” are perfect for other applications involving narrow-band ultraviolet imaging in the 193 nm to 400 nm range.

For example, a shortwave UV bandpass filter (centered at 254 nm with maximum 25 nm bandwidth) can be used for RUVIS to provide sufficient contrast against the substrate’s reflections and fluorescence at other wavelengths. Commercial RUVIS devices, like the one shown in the following figure, are usually based on image intensifier tubes sensitive to short-wave UV (commonly a Generation II type) fitted with a quartz lens and a 254 nm filter. Common focal lengths used for searching and acquisition of latent prints are in the range of 50 mm to 80 mm with maximum apertures around f/3.5. In addition to image intensifiers, video cameras optimized for short-wave ultraviolet imaging can also be used for RUVIS fingerprint acquisition.

RUVIS www.UVIRimaging.com

The KrimesiteTM RUVIS imager made by Sirchie® incorporates a Gen II UV-sensitive intensifier tube, a 6-element quartz lens with 60 mm focal length and f/3.5 aperture, and a narrowband 254 nm filter into a compact viewer that is easy to use to find latent prints and other evidence not visible to the unaided eye. The small laser on the side helps investigators locate the area where evidence is found with the RUVIS imager. Image courtesy of Sirchie®.

A complete whitepaper on these diy SBUV and RUVIS shortwave UV imagers is available for download at: Prutchi – diy Shortwave UV Converters and cameras

Other interesting sources of information related to this project:

For more diy UV photography and imaging projects, check out my new book:

Exploring Ultraviolet Photography book by David Prutchi PhD.



3 Responses