- Email: [email protected]
Fluorescent identification of biological and other stains on skin by the use of alternative light sources
J O U R N A L
O F
CLINICAL FORENSIC MEDICINE Journal of Clinical Forensic Medicine 12 (2005) 296–301 www.elsevier.com/locate/jcfm
Original communication
Fluorescent identification of biological and other stains on skin by the use of alternative light sources Jacob Wawryk a, Morris Odell b
b,*
a University of Melbourne, Australia Victorian Institute of Forensic Medicine, 57-83 Kavanagh Street, Southbank, Vic. 3006, Australia
Available online 31 May 2005
Abstract The detection of body fluids in cases of sexual assault or abuse is important for purposes of evidence collection and DNA testing. In cases where semen is deposited on skin, a method for detection of semen may be a valuable aid to evidence collection [Gabby T, Winkleby M, Boyce T, Fisher D, Lancaster A, Sensabaugh G. Sexual abuse of children. The detection of semen on skin. AJCD 1992; 146:700–703]. Semen is known to fluoresce when exposed to light of certain wavelengths. For this study, we placed various body fluids, as well as lubricants and moisturizers on the forearms of volunteers. The areas were illuminated using Alternative Light Sources (ALS) with peak wavelengths of between 370 and 500 nm for examination soon after deposition, and again after 24 h. No fluorescence was visible from any of the substances in the majority of volunteers examined. In a few subjects, semen and urine were found to fluoresce faintly under the more powerful lights. In these cases, the quality of fluorescence provided by urine and semen was noticeably different. For comparison, semen was applied to cloth, and fluoresced well at the expected wavelengths. While ALS is useful for identification of stains on clothing, its use in detecting stains on skin is currently very limited. Ó 2005 Elsevier Ltd and AFP. All rights reserved. Keywords: Semen; Fluorescence; ALS; Ultraviolet; Rape
1. Introduction The detection of seminal stains on humans in cases of sexual assault or abuse is important for purposes of evidence collection1,2 and DNA testing. Nevertheless, there has been very little research done in the field. In the past, standard sexual assault investigation technique3,4 included fluorescent identification for semen stains using a WoodÕs Lamp (WL, UV light source of around 360 nm wavelength) although this seems to have been done without any experimental verification of its efficacy. The WL is known to cause fungi, bacteria and chemicals to fluoresce. Fluorescence is the ability of a material to absorb light of a particular wavelength (color), and then re-emit light at another, longer wave*
Corresponding author. Tel.: +613 9684 4480; fax: +613 9684 4481. E-mail address: [email protected] (M. Odell).
length. This re-emission of light occurs immediately – there is no delay to the onset of fluorescence, nor is there continued emission of light after the source of incident illumination is removed. This differs from phosphorescence, where the material continues to glow for a certain time after incident light is removed. Fluorescent substances may be excitable at one, or many different wavelengths, and may fluoresce at different wavelengths depending on the excitation wavelength used. A recent paper by Santucci et al.5 cast doubt on the efficacy of a WoodÕs Lamp for the detection of biological stains, with a finding that the WL does not cause semen to fluoresce appreciably. A 1991 paper by Stoilovic6 documented the fluorescent excitation spectra of dried semen, and showed that it is excitable from 300 to 480 nm, in a normal distribution skewed towards higher wavelengths, peaking at between 380 and 430 nm. The fluorescent emission
1353-1131/$ - see front matter Ó 2005 Elsevier Ltd and AFP. All rights reserved. doi:10.1016/j.jcfm.2005.03.005
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301
spectrum of semen when excited at 350 and 450 nm, has peak emissions at 460 and 520 nm, respectively7. Semen can be made to phosphoresce, but this only occurs when it has been cooled significantly. Calloway et al.8 were able to produce marked phosphorescence at 77 ° (the temperature of liquid nitrogen), though this is obviously not usable in the clinical setting. Semen is phosphorescent when illuminated by very low UV light (254 nm). Special ‘‘forensic’’ light sources such as the Polilight or Poliray (Rofin Australia) are available that are capable of exploiting semenÕs wide excitation band, and are tunable (with filters) to emit only wavelengths in a narrow band, so as to produce a specific emission spectrum for easy detection. They are also usable for detection of blood and gunshot residues. They are in common use in forensic laboratories where they are used to guide examinations of clothing and similar material9. Unfortunately, these forensic light sources are extremely expensive, costing thousands of dollars, and are also bulky which reduces their portability and thus their usefulness for investigators to take to, and use at a sexual assault examination in the field. Recently there have been great advances in alternative light source technologies for forensic purposes10. In particular the use of light emitting diode (LED) technologies promises a great reduction in the size, weigh, cost and power consumption of such sources. These are solid state devices that convert electrical energy directly into electromagnetic radiation. LEDs have previously been available in only a very limited range of wavelengths, and with quite poor light intensity. However, they are now available in almost every color of the visible spectrum, as well as in UV/infrared, and some (such as the Lumileds LuxeonTM ) are capable of high luminous intensities. Their luminous intensity compares favorably with a standard 75 W incandescent globe while using only a fraction of the energy and generating minimum heat. LEDs have many advantages over other light sources – they are small, light, cheap, have a very long lifetime and are relatively sturdy. A useful feature of LEDs with regard to possible use in forensic examinations is that they have a relatively narrow emission spectrum. A typical incandescent lamp produces much of its output towards the longer wavelength end of the spectrum (greens, yellows, and reds), and this is all wasted energy that must be filtered out. It is possible to pick a LED that will cause the substance being studied to fluoresce best, in much the same way that a much more expensive forensic light source must be tuned with filters. Fluorescence emitted by a substance is always of much lower intensity than the light used to stimulate it. Therefore, in order to view fluorescence it is necessary to filter out incident light (if that light is in the visible spectrum). The majority of the excitation spectrum of
297
semen is in the visible wavelength, and therefore it is necessary to use filters that will block the bright incident light, but allow any fluorescent light through. Semen fluoresces orange when illuminated with blue light, so the use of an orange filter allows proper visualization of fluorescence. Under UV illumination semen fluoresces green, and so yellow goggles are used. Most of the studies on fluorescence of biological fluids and stains have been conducted using inert substrates and there has been very little emphasis on the usefulness or otherwise of the technique on the living11. The aim of this study was to identify a technology (if any) suitable for a cheap compact ALS that could be carried in the field by forensic medical officers who perform sexual assault examinations.
2. Materials For this study we chose 11 LED light sources to investigate, with peak emission wavelengths ranging from 370 to 480 nm. Source 1 was a LedDynamics EverLEDTM replacement flashlight bulb, which we installed in a 3 D cell MagliteTM torch. Though marketed as ‘‘blue’’ this bulb has a decidedly turquoise color to it, and we found it to have a peak wavelength of 482.5 nm. At the time of the study (late 2003) EverLEDTM sold for about AU$55. Sources 2–6 were individual LEDs purchased from various electronics stores and from the Internet. They varied in price from AU$1 to $10 each. Source 2 is unusual for an LED light source in that it has a rather broad, flat peak, from around 415 to 430 nm. All of the other LEDs tested had much more peaked emission characteristics. Source 3 peaked at 402.0 nm, source 4 at 456.2 nm. Source 5 peaked at 463.6 nm wavelength and source 6 at 399.7 nm. These LEDs were all driven with a constant 20 mA, the maximum current recommended to prevent thermal runaway. Sources 7 and 8 were purchased as completed torches with a single LED light in ‘‘blue’’ and ‘‘UV’’. These are a typical keychain type light, with a thumb operated on– off switch and power provided by two 3v flat cell batteries. The ‘‘blue’’ torch produced a peak wavelength of 463.9 nm, and the ‘‘UV’’ torch produced a peak of 410.6 nm – more violet than true UV, though on the edge of the visible spectrum. These torches sold for about AU$20. Source 9 was a penlight LED torch purchased from www.maxmax.com. It produced a peak wavelength of 376.3 nm, well into the UV spectrum, and is powered by 3 AAAA cell batteries (1.5 V each). As the lower UV LEDs are still rather expensive, the torch had a price of $AU80. Sources 10 and 11 were LuxeonTM Star V LEDs. These are a newly developed LED technology by
298
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301
Lumileds (www.Lumileds.com) and produce around 40 times the luminous flux of standard hi-intensity LEDs, up to 220 lm. This compares with a standard 60 W household light globe that produces around 850 lm. Source 10 was marketed as ‘‘royal blue’’ and 11 as ‘‘blue’’, and they were found to have peak wavelengths of 452.9 and 466.9 nm, respectively. These LEDs were driven at 700 mA, their recommended maximum continuous current rating, and sold for around AU$40 each. Source 12 was a PolirayTM (Rofin Australia) forensic light source, used for comparison. It is not an LED light source, but a 75 W halogen lamp and 450 nm band pass filter. Poliray sells for over AU$1000, with many kits and options available. Wavelengths were measured at the RMIT Applied Optics laboratory on an Ocean Optics model PC2000 fibre-optic spectrometer,. This instrument has a wavelength resolution of around 2 nm, and the wavelength scale is accurate to within 1 nm (see Fig. 1, Table 1). In addition to looking for fluorescence in semen on skin, we also examined several possible confounding materials that might be present in a sexual assault examination. These included several types of lubricants (Wet ProbeTM personal lubricant – Claredale Distributors Victoria, Eros superconcentrated Bodyglide – pjur group: Luxemburg, Satin GelTM – Calvista Wholesale
Table 1 Summary of light sources Source
Peak wavelength (nm)
Cost (AU$)
1 2 3 4 5 6 7 8 9 10 11 12
482.5 415–430 402.0 456.2 463.6 399.7 463.9 410.6 376.3 452.9 466.9 450
$55 $1–10 $1–10 $1–10 $1–10 $1–10 $20 $20 $80 $40 $40 $1000+
Pty. Ltd.), condoms (LiaisonTM – Searle Australia Pty. Ltd., ChekmateTM – Ansell International, Lifestyles SportTM colored – Ansell International) and hand creams (Nivea BodyTM Nourishing Lotion – Smith & Nephew Pty. Ltd. Australia, PondÕsTM Hydro-Nourishing Moisturising cream – Unilever Australasia).
3. Method Subjects were recruited from the Victorian Institute of Forensic Medicine (VIFM), as well as the University
Fig. 1. Light source wavelength spectra.
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301
of Melbourne. They were offered AU$10 plus travel expenses for their participation in this trial. Approval for experimentation using human tissue (semen, blood, saliva, urine) was obtained from the VIFM Ethics committee. Subjects were asked to collect a semen sample on the day they came in for testing, and to keep it refrigerated in the hours after collection. Urine was collected just before application, and in the case of male volunteers it was checked that this was not their first urination since semen collection, to avoid contamination. Blood was drawn by fingerprick just before application, and saliva was also provided at this time. Subjects were not told to use any special techniques in washing or otherwise preparing the skin areas used for the study. This was judged to be as close as possible to the real life situation where there is normal random variation in the condition of victimsÕ skin. Substances were applied to delineated areas of the forearms of volunteers, and allowed to air dry for a few minutes, until they no longer appeared wet. The room was darkened, and the forearm to which substances had been applied illuminated with the various light sources. Care was taken not to touch the test areas with the light source or any of the other equipment used for the study. Such contact could potentially lead to contamination of both the test area and the equipment. These areas were then observed for fluorescence by researchers with the naked eye and while wearing orange or yellow filter goggles, and also photographed with a Nikon Coolpix 950 digital camera equipped with an orange filter. Volunteers were asked not to wash the area to which substances had been applied, and to return the following day so that we could observe changes with time. Normal contact with clothing was permitted. Examination on the second day followed the same method as the first, with careful observation followed by photography.
299
The everLEDTM MagliteTM replacement bulb produced some fluorescence in three of the eight subjects when viewed through orange goggles, though it was very faint. In all of these three cases it was much easier to determine the presence of some substance on the skin through reflection rather than fluorescence. Nothing was detected with this lamp on the second day of observation. Sources 10, 11 and 12 (the two LuxeonTM LEDs and the PolirayTM ) resulted in visible fluorescence in five of the eight subjects in the first round of testing. However, despite this apparent success, it was still easier to determine the presence of something on the skin through reflection than fluorescence. In three of the five cases where fluorescence was detected the first day, fluorescence was still visible on day two, though it was extremely faint. Photographs were taken of every subject, under illumination by every light source, with incident light filtered out. However due to the faintness of fluorescence detected (in those where fluorescence was detected at all) it was not possible to obtain photographs showing this phenomenon. None of the subjectsÕ saliva was fluorescent under any of the light sources. Urine was not seen to fluoresce with any of the lower powered light sources, or the everLEDTM MagliteTM replacement bulb. It did display some fluorescence under sources 10, 11 and 12 in four of the eight subjects tested on day one, of similar intensity to the semen. No fluorescence was detected on day two. Blood was easily visualized under all of the light sources on day one, though this was mostly due to its pigmented nature. Blood absorbs light around 430 nm wavelength (violet), appearing black under this illumination. No blood could be detected on day two. None of the lubricants or hand creams tested were visibly fluorescent under any of the light sources.
5. Discussion 4. Results Eight subjects were recruited for the trial. Ages ranged from 18 through mid 50s, with a mean in the low 20s. All of these subjects were of Caucasian descent. None were taking any medications. Two had undergone vasectomy some years before. Semen samples were used an average of 15 h after collection, after being kept refrigerated but not frozen. Blood, urine and saliva were provided immediately before observation. Despite efforts to recruit more females for this study only one female was tested using her partnerÕs semen. In every case the low powered LED lights (source 2– 9) proved unable to excite visible fluorescence in the semen.
In all cases where the fluorescence was visible on the skin of test subjects it was much easier to notice the presence of a substance on the skin under normal white light, without looking for fluorescence. As the semen dries on the skin it forms a crust (much like that formed when superglue dries on the skin, for comparison), and this creates a transparent shiny patch that can be observed as light reflects from it at an angle. Any fluorescence observed with the tested light sources through goggles was less obvious than simply looking at the surface normally (though this reflection is completely nonspecific). This crust was not visible on the second day, though some fluorescence remained in three of the eight subjects. It is not possible to say for how long the crust
300
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301
remains on the skin after drying of the semen. Obviously daily activities, contact with clothes, etc., would have a great bearing on this. In viewing semen fluorescence with these light sources it was necessary to position the light very close to the surface of the skin (<3 cm) so as to obtain sufficient intensity to create visible fluorescence. Fluorescence can then be faintly observed from a distance of 20 cm, with filter goggles. Observing the skin from this distance without filter goggles is dazzling, much like looking directly at an incandescent light bulb. We also applied small amounts of semen to cloth, and observed it in the same manner as we had looked for fluorescence on skin. Fluorescence was quite obvious when observed in this manner (with faint fluorescence visible even with some of the lower powered sources). It is possible to differentiate between semen and urine (which also fluoresces under these wavelengths of illumination) as the semen fluoresces a noticeably more orange hue compared to the urineÕs almost yellow. Also, urine does not fluoresce as brightly, and is more even in texture (the semen fluorescence has a patchy quality to it). However, unlike semen, urine does not leave a crust on the skin where it dries. This means that it is not visible via reflection, and ALS fluorescence is the only it may be visualized (however poorly). It was initially anticipated that we would be able to compare the fluorescence of azoospermic semen provided by the men who had been vasectomized, with the fluorescence of ‘‘normal’’ semen. However as all fluorescence observed was so faint, such a comparison is not possible, except to state that one of the five samples where some fluorescence was detected, was azoospermic. Azoospermic semen appears to fluoresce to some degree, indistinguishable from ‘‘normal’’ semen, on skin, under illumination of the intensity used in this study.
6. Limitations of the study In cases of sexual assault, seminal stains are unlikely to be found on the forearms of victims. They are much more likely to be located on skin around the genital region, in the pubic hair, the face, or over the breasts. It is unknown whether the skin of the forearm would behave in the same way as skin in these other regions. However, ethical considerations, as well as simple logistics prevented us from testing in these areas. This trial was not blinded in any way. We knew which substance was which, and which light source was which. Nothing can be done about knowing which light source was used as the light produced by each was quite individual. It is possible that some bias was introduced by our knowing which substance was which; however such circumstances would serve only to heighten
our ability to perceive what little fluorescence we did see in the semen.
7. Conclusion Identification of semen stains is important for the collection of evidence and DNA samples in cases of sexual assault. It is possible with new technologies to identify semen stains on fabrics and clothing. However at present, the same cannot be said for the identification of semen stains on skin. This research indicates that while such stains do fluoresce, they do so excessively faintly – too faintly to be of use in their ‘‘quick and easy identification’’ with lights of currently available intensity. Why it is that stains on skin do not fluoresce nearly as noticeably is not known, though it may be due to some substance excreted in sweat or sebum interfering with fluorescence. In contrast, it may be that something causes semen to be more fluorescent when deposited on fibres. Any stains present on fibres would also likely have been created by a larger volume of semen than that tested in this trial. This would lead to better fluorescence. A potential limitation of this technique might be an effect of the ALS on semen. It is known that DNA can be degraded by intense ultraviolet light. The light sources used in this study were mostly not ultraviolet and of low intensity compared to natural radiation such as sunlight. Sunlight is known to cause DNA changes in skin after prolonged exposure but there is no literature regarding its effect on semen. The use of ultraviolet light in automated DNA analysis machines would suggest that this will not be a practical problem should the technique ever be refined further.12 It may be possible to identify semen stains on skin quickly and easily in the future, with more powerful alternative light sources, however the possibility of damage to the skin from such intense light would then need to be considered. Until that time however, it seems sexual assault examiners will need to continue to rely on the victimÕs account, along with unguided swabs when collecting evidence.
References 1. Gabby T, Winkleby M, Boyce T, Fisher D, Lancaster A, Sensabaugh G. Sexual abuse of children. The detection of semen on skin. AJCD 1992;146:700–3. 2. McGregor M, Du Mont J, Myhr T. Sexual assault forensic medical examination: is evidence related to successful prosecution? Annal Emergency Med 2002;39:639–47. 3. Crowley, Sharon R. Sexual assault, The medico-legal examination. Stamford: Appleton & Lange; 1999.
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301 4. Girardin, Faugno B, Seneski D, Slaughter P. Color atlas of sexual assault. St. Louis: Mosby; 1997. 5. Santucci K, Nelson D, McQuillen K, Duffy S, Liniakis J. WoodÕs lamp utility in the detection of semen. Pediatrics 1999;104:1342–4. 6. Stoilovic M. Detection of semen and blood stains using polilight as a light source. Forensic Sci Int 1991;51:289–96. 7. Springer E, Almog J, Frank A, Ziv Z, Bergman P, Gui Qiang W. Detection of dry body fluids by inherent short wave UV luminescence: preliminary results. Forensic Sci Int 1994;66:89–94. 8. Calloway A, Jones P, Siegel S, Stupian G. Location of seminal stains by their phosphorescence and its use in determining the
9.
10. 11. 12.
301
order of deposition of overlapping seminal and bloodstains. J Forensic Sci Soc 1973;13:223–30. Kobus H, Silenieks E, Scharnberg J. Improving the effectiveness of fluorescence for the detection of semen stains on fabrics. J Forensic Sci 2002;47:819–23. Nelson D, Santucci K. An alternate light source to detect semen. Acad Emerg Med 2002;9:1045–8. Oepen I. Identification of characteristics in blood and semen stains – a review. Forensic Sci Int 1988;36:183–91. Ricci U et al.. Identification of forensic samples by using an infrared-based automatic DNA sequencer. Croat Med J 2003;44(3):299–305.
O F
CLINICAL FORENSIC MEDICINE Journal of Clinical Forensic Medicine 12 (2005) 296–301 www.elsevier.com/locate/jcfm
Original communication
Fluorescent identification of biological and other stains on skin by the use of alternative light sources Jacob Wawryk a, Morris Odell b
b,*
a University of Melbourne, Australia Victorian Institute of Forensic Medicine, 57-83 Kavanagh Street, Southbank, Vic. 3006, Australia
Available online 31 May 2005
Abstract The detection of body fluids in cases of sexual assault or abuse is important for purposes of evidence collection and DNA testing. In cases where semen is deposited on skin, a method for detection of semen may be a valuable aid to evidence collection [Gabby T, Winkleby M, Boyce T, Fisher D, Lancaster A, Sensabaugh G. Sexual abuse of children. The detection of semen on skin. AJCD 1992; 146:700–703]. Semen is known to fluoresce when exposed to light of certain wavelengths. For this study, we placed various body fluids, as well as lubricants and moisturizers on the forearms of volunteers. The areas were illuminated using Alternative Light Sources (ALS) with peak wavelengths of between 370 and 500 nm for examination soon after deposition, and again after 24 h. No fluorescence was visible from any of the substances in the majority of volunteers examined. In a few subjects, semen and urine were found to fluoresce faintly under the more powerful lights. In these cases, the quality of fluorescence provided by urine and semen was noticeably different. For comparison, semen was applied to cloth, and fluoresced well at the expected wavelengths. While ALS is useful for identification of stains on clothing, its use in detecting stains on skin is currently very limited. Ó 2005 Elsevier Ltd and AFP. All rights reserved. Keywords: Semen; Fluorescence; ALS; Ultraviolet; Rape
1. Introduction The detection of seminal stains on humans in cases of sexual assault or abuse is important for purposes of evidence collection1,2 and DNA testing. Nevertheless, there has been very little research done in the field. In the past, standard sexual assault investigation technique3,4 included fluorescent identification for semen stains using a WoodÕs Lamp (WL, UV light source of around 360 nm wavelength) although this seems to have been done without any experimental verification of its efficacy. The WL is known to cause fungi, bacteria and chemicals to fluoresce. Fluorescence is the ability of a material to absorb light of a particular wavelength (color), and then re-emit light at another, longer wave*
Corresponding author. Tel.: +613 9684 4480; fax: +613 9684 4481. E-mail address: [email protected] (M. Odell).
length. This re-emission of light occurs immediately – there is no delay to the onset of fluorescence, nor is there continued emission of light after the source of incident illumination is removed. This differs from phosphorescence, where the material continues to glow for a certain time after incident light is removed. Fluorescent substances may be excitable at one, or many different wavelengths, and may fluoresce at different wavelengths depending on the excitation wavelength used. A recent paper by Santucci et al.5 cast doubt on the efficacy of a WoodÕs Lamp for the detection of biological stains, with a finding that the WL does not cause semen to fluoresce appreciably. A 1991 paper by Stoilovic6 documented the fluorescent excitation spectra of dried semen, and showed that it is excitable from 300 to 480 nm, in a normal distribution skewed towards higher wavelengths, peaking at between 380 and 430 nm. The fluorescent emission
1353-1131/$ - see front matter Ó 2005 Elsevier Ltd and AFP. All rights reserved. doi:10.1016/j.jcfm.2005.03.005
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301
spectrum of semen when excited at 350 and 450 nm, has peak emissions at 460 and 520 nm, respectively7. Semen can be made to phosphoresce, but this only occurs when it has been cooled significantly. Calloway et al.8 were able to produce marked phosphorescence at 77 ° (the temperature of liquid nitrogen), though this is obviously not usable in the clinical setting. Semen is phosphorescent when illuminated by very low UV light (254 nm). Special ‘‘forensic’’ light sources such as the Polilight or Poliray (Rofin Australia) are available that are capable of exploiting semenÕs wide excitation band, and are tunable (with filters) to emit only wavelengths in a narrow band, so as to produce a specific emission spectrum for easy detection. They are also usable for detection of blood and gunshot residues. They are in common use in forensic laboratories where they are used to guide examinations of clothing and similar material9. Unfortunately, these forensic light sources are extremely expensive, costing thousands of dollars, and are also bulky which reduces their portability and thus their usefulness for investigators to take to, and use at a sexual assault examination in the field. Recently there have been great advances in alternative light source technologies for forensic purposes10. In particular the use of light emitting diode (LED) technologies promises a great reduction in the size, weigh, cost and power consumption of such sources. These are solid state devices that convert electrical energy directly into electromagnetic radiation. LEDs have previously been available in only a very limited range of wavelengths, and with quite poor light intensity. However, they are now available in almost every color of the visible spectrum, as well as in UV/infrared, and some (such as the Lumileds LuxeonTM ) are capable of high luminous intensities. Their luminous intensity compares favorably with a standard 75 W incandescent globe while using only a fraction of the energy and generating minimum heat. LEDs have many advantages over other light sources – they are small, light, cheap, have a very long lifetime and are relatively sturdy. A useful feature of LEDs with regard to possible use in forensic examinations is that they have a relatively narrow emission spectrum. A typical incandescent lamp produces much of its output towards the longer wavelength end of the spectrum (greens, yellows, and reds), and this is all wasted energy that must be filtered out. It is possible to pick a LED that will cause the substance being studied to fluoresce best, in much the same way that a much more expensive forensic light source must be tuned with filters. Fluorescence emitted by a substance is always of much lower intensity than the light used to stimulate it. Therefore, in order to view fluorescence it is necessary to filter out incident light (if that light is in the visible spectrum). The majority of the excitation spectrum of
297
semen is in the visible wavelength, and therefore it is necessary to use filters that will block the bright incident light, but allow any fluorescent light through. Semen fluoresces orange when illuminated with blue light, so the use of an orange filter allows proper visualization of fluorescence. Under UV illumination semen fluoresces green, and so yellow goggles are used. Most of the studies on fluorescence of biological fluids and stains have been conducted using inert substrates and there has been very little emphasis on the usefulness or otherwise of the technique on the living11. The aim of this study was to identify a technology (if any) suitable for a cheap compact ALS that could be carried in the field by forensic medical officers who perform sexual assault examinations.
2. Materials For this study we chose 11 LED light sources to investigate, with peak emission wavelengths ranging from 370 to 480 nm. Source 1 was a LedDynamics EverLEDTM replacement flashlight bulb, which we installed in a 3 D cell MagliteTM torch. Though marketed as ‘‘blue’’ this bulb has a decidedly turquoise color to it, and we found it to have a peak wavelength of 482.5 nm. At the time of the study (late 2003) EverLEDTM sold for about AU$55. Sources 2–6 were individual LEDs purchased from various electronics stores and from the Internet. They varied in price from AU$1 to $10 each. Source 2 is unusual for an LED light source in that it has a rather broad, flat peak, from around 415 to 430 nm. All of the other LEDs tested had much more peaked emission characteristics. Source 3 peaked at 402.0 nm, source 4 at 456.2 nm. Source 5 peaked at 463.6 nm wavelength and source 6 at 399.7 nm. These LEDs were all driven with a constant 20 mA, the maximum current recommended to prevent thermal runaway. Sources 7 and 8 were purchased as completed torches with a single LED light in ‘‘blue’’ and ‘‘UV’’. These are a typical keychain type light, with a thumb operated on– off switch and power provided by two 3v flat cell batteries. The ‘‘blue’’ torch produced a peak wavelength of 463.9 nm, and the ‘‘UV’’ torch produced a peak of 410.6 nm – more violet than true UV, though on the edge of the visible spectrum. These torches sold for about AU$20. Source 9 was a penlight LED torch purchased from www.maxmax.com. It produced a peak wavelength of 376.3 nm, well into the UV spectrum, and is powered by 3 AAAA cell batteries (1.5 V each). As the lower UV LEDs are still rather expensive, the torch had a price of $AU80. Sources 10 and 11 were LuxeonTM Star V LEDs. These are a newly developed LED technology by
298
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301
Lumileds (www.Lumileds.com) and produce around 40 times the luminous flux of standard hi-intensity LEDs, up to 220 lm. This compares with a standard 60 W household light globe that produces around 850 lm. Source 10 was marketed as ‘‘royal blue’’ and 11 as ‘‘blue’’, and they were found to have peak wavelengths of 452.9 and 466.9 nm, respectively. These LEDs were driven at 700 mA, their recommended maximum continuous current rating, and sold for around AU$40 each. Source 12 was a PolirayTM (Rofin Australia) forensic light source, used for comparison. It is not an LED light source, but a 75 W halogen lamp and 450 nm band pass filter. Poliray sells for over AU$1000, with many kits and options available. Wavelengths were measured at the RMIT Applied Optics laboratory on an Ocean Optics model PC2000 fibre-optic spectrometer,. This instrument has a wavelength resolution of around 2 nm, and the wavelength scale is accurate to within 1 nm (see Fig. 1, Table 1). In addition to looking for fluorescence in semen on skin, we also examined several possible confounding materials that might be present in a sexual assault examination. These included several types of lubricants (Wet ProbeTM personal lubricant – Claredale Distributors Victoria, Eros superconcentrated Bodyglide – pjur group: Luxemburg, Satin GelTM – Calvista Wholesale
Table 1 Summary of light sources Source
Peak wavelength (nm)
Cost (AU$)
1 2 3 4 5 6 7 8 9 10 11 12
482.5 415–430 402.0 456.2 463.6 399.7 463.9 410.6 376.3 452.9 466.9 450
$55 $1–10 $1–10 $1–10 $1–10 $1–10 $20 $20 $80 $40 $40 $1000+
Pty. Ltd.), condoms (LiaisonTM – Searle Australia Pty. Ltd., ChekmateTM – Ansell International, Lifestyles SportTM colored – Ansell International) and hand creams (Nivea BodyTM Nourishing Lotion – Smith & Nephew Pty. Ltd. Australia, PondÕsTM Hydro-Nourishing Moisturising cream – Unilever Australasia).
3. Method Subjects were recruited from the Victorian Institute of Forensic Medicine (VIFM), as well as the University
Fig. 1. Light source wavelength spectra.
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301
of Melbourne. They were offered AU$10 plus travel expenses for their participation in this trial. Approval for experimentation using human tissue (semen, blood, saliva, urine) was obtained from the VIFM Ethics committee. Subjects were asked to collect a semen sample on the day they came in for testing, and to keep it refrigerated in the hours after collection. Urine was collected just before application, and in the case of male volunteers it was checked that this was not their first urination since semen collection, to avoid contamination. Blood was drawn by fingerprick just before application, and saliva was also provided at this time. Subjects were not told to use any special techniques in washing or otherwise preparing the skin areas used for the study. This was judged to be as close as possible to the real life situation where there is normal random variation in the condition of victimsÕ skin. Substances were applied to delineated areas of the forearms of volunteers, and allowed to air dry for a few minutes, until they no longer appeared wet. The room was darkened, and the forearm to which substances had been applied illuminated with the various light sources. Care was taken not to touch the test areas with the light source or any of the other equipment used for the study. Such contact could potentially lead to contamination of both the test area and the equipment. These areas were then observed for fluorescence by researchers with the naked eye and while wearing orange or yellow filter goggles, and also photographed with a Nikon Coolpix 950 digital camera equipped with an orange filter. Volunteers were asked not to wash the area to which substances had been applied, and to return the following day so that we could observe changes with time. Normal contact with clothing was permitted. Examination on the second day followed the same method as the first, with careful observation followed by photography.
299
The everLEDTM MagliteTM replacement bulb produced some fluorescence in three of the eight subjects when viewed through orange goggles, though it was very faint. In all of these three cases it was much easier to determine the presence of some substance on the skin through reflection rather than fluorescence. Nothing was detected with this lamp on the second day of observation. Sources 10, 11 and 12 (the two LuxeonTM LEDs and the PolirayTM ) resulted in visible fluorescence in five of the eight subjects in the first round of testing. However, despite this apparent success, it was still easier to determine the presence of something on the skin through reflection than fluorescence. In three of the five cases where fluorescence was detected the first day, fluorescence was still visible on day two, though it was extremely faint. Photographs were taken of every subject, under illumination by every light source, with incident light filtered out. However due to the faintness of fluorescence detected (in those where fluorescence was detected at all) it was not possible to obtain photographs showing this phenomenon. None of the subjectsÕ saliva was fluorescent under any of the light sources. Urine was not seen to fluoresce with any of the lower powered light sources, or the everLEDTM MagliteTM replacement bulb. It did display some fluorescence under sources 10, 11 and 12 in four of the eight subjects tested on day one, of similar intensity to the semen. No fluorescence was detected on day two. Blood was easily visualized under all of the light sources on day one, though this was mostly due to its pigmented nature. Blood absorbs light around 430 nm wavelength (violet), appearing black under this illumination. No blood could be detected on day two. None of the lubricants or hand creams tested were visibly fluorescent under any of the light sources.
5. Discussion 4. Results Eight subjects were recruited for the trial. Ages ranged from 18 through mid 50s, with a mean in the low 20s. All of these subjects were of Caucasian descent. None were taking any medications. Two had undergone vasectomy some years before. Semen samples were used an average of 15 h after collection, after being kept refrigerated but not frozen. Blood, urine and saliva were provided immediately before observation. Despite efforts to recruit more females for this study only one female was tested using her partnerÕs semen. In every case the low powered LED lights (source 2– 9) proved unable to excite visible fluorescence in the semen.
In all cases where the fluorescence was visible on the skin of test subjects it was much easier to notice the presence of a substance on the skin under normal white light, without looking for fluorescence. As the semen dries on the skin it forms a crust (much like that formed when superglue dries on the skin, for comparison), and this creates a transparent shiny patch that can be observed as light reflects from it at an angle. Any fluorescence observed with the tested light sources through goggles was less obvious than simply looking at the surface normally (though this reflection is completely nonspecific). This crust was not visible on the second day, though some fluorescence remained in three of the eight subjects. It is not possible to say for how long the crust
300
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301
remains on the skin after drying of the semen. Obviously daily activities, contact with clothes, etc., would have a great bearing on this. In viewing semen fluorescence with these light sources it was necessary to position the light very close to the surface of the skin (<3 cm) so as to obtain sufficient intensity to create visible fluorescence. Fluorescence can then be faintly observed from a distance of 20 cm, with filter goggles. Observing the skin from this distance without filter goggles is dazzling, much like looking directly at an incandescent light bulb. We also applied small amounts of semen to cloth, and observed it in the same manner as we had looked for fluorescence on skin. Fluorescence was quite obvious when observed in this manner (with faint fluorescence visible even with some of the lower powered sources). It is possible to differentiate between semen and urine (which also fluoresces under these wavelengths of illumination) as the semen fluoresces a noticeably more orange hue compared to the urineÕs almost yellow. Also, urine does not fluoresce as brightly, and is more even in texture (the semen fluorescence has a patchy quality to it). However, unlike semen, urine does not leave a crust on the skin where it dries. This means that it is not visible via reflection, and ALS fluorescence is the only it may be visualized (however poorly). It was initially anticipated that we would be able to compare the fluorescence of azoospermic semen provided by the men who had been vasectomized, with the fluorescence of ‘‘normal’’ semen. However as all fluorescence observed was so faint, such a comparison is not possible, except to state that one of the five samples where some fluorescence was detected, was azoospermic. Azoospermic semen appears to fluoresce to some degree, indistinguishable from ‘‘normal’’ semen, on skin, under illumination of the intensity used in this study.
6. Limitations of the study In cases of sexual assault, seminal stains are unlikely to be found on the forearms of victims. They are much more likely to be located on skin around the genital region, in the pubic hair, the face, or over the breasts. It is unknown whether the skin of the forearm would behave in the same way as skin in these other regions. However, ethical considerations, as well as simple logistics prevented us from testing in these areas. This trial was not blinded in any way. We knew which substance was which, and which light source was which. Nothing can be done about knowing which light source was used as the light produced by each was quite individual. It is possible that some bias was introduced by our knowing which substance was which; however such circumstances would serve only to heighten
our ability to perceive what little fluorescence we did see in the semen.
7. Conclusion Identification of semen stains is important for the collection of evidence and DNA samples in cases of sexual assault. It is possible with new technologies to identify semen stains on fabrics and clothing. However at present, the same cannot be said for the identification of semen stains on skin. This research indicates that while such stains do fluoresce, they do so excessively faintly – too faintly to be of use in their ‘‘quick and easy identification’’ with lights of currently available intensity. Why it is that stains on skin do not fluoresce nearly as noticeably is not known, though it may be due to some substance excreted in sweat or sebum interfering with fluorescence. In contrast, it may be that something causes semen to be more fluorescent when deposited on fibres. Any stains present on fibres would also likely have been created by a larger volume of semen than that tested in this trial. This would lead to better fluorescence. A potential limitation of this technique might be an effect of the ALS on semen. It is known that DNA can be degraded by intense ultraviolet light. The light sources used in this study were mostly not ultraviolet and of low intensity compared to natural radiation such as sunlight. Sunlight is known to cause DNA changes in skin after prolonged exposure but there is no literature regarding its effect on semen. The use of ultraviolet light in automated DNA analysis machines would suggest that this will not be a practical problem should the technique ever be refined further.12 It may be possible to identify semen stains on skin quickly and easily in the future, with more powerful alternative light sources, however the possibility of damage to the skin from such intense light would then need to be considered. Until that time however, it seems sexual assault examiners will need to continue to rely on the victimÕs account, along with unguided swabs when collecting evidence.
References 1. Gabby T, Winkleby M, Boyce T, Fisher D, Lancaster A, Sensabaugh G. Sexual abuse of children. The detection of semen on skin. AJCD 1992;146:700–3. 2. McGregor M, Du Mont J, Myhr T. Sexual assault forensic medical examination: is evidence related to successful prosecution? Annal Emergency Med 2002;39:639–47. 3. Crowley, Sharon R. Sexual assault, The medico-legal examination. Stamford: Appleton & Lange; 1999.
J. Wawryk, M. Odell / Journal of Clinical Forensic Medicine 12 (2005) 296–301 4. Girardin, Faugno B, Seneski D, Slaughter P. Color atlas of sexual assault. St. Louis: Mosby; 1997. 5. Santucci K, Nelson D, McQuillen K, Duffy S, Liniakis J. WoodÕs lamp utility in the detection of semen. Pediatrics 1999;104:1342–4. 6. Stoilovic M. Detection of semen and blood stains using polilight as a light source. Forensic Sci Int 1991;51:289–96. 7. Springer E, Almog J, Frank A, Ziv Z, Bergman P, Gui Qiang W. Detection of dry body fluids by inherent short wave UV luminescence: preliminary results. Forensic Sci Int 1994;66:89–94. 8. Calloway A, Jones P, Siegel S, Stupian G. Location of seminal stains by their phosphorescence and its use in determining the
9.
10. 11. 12.
301
order of deposition of overlapping seminal and bloodstains. J Forensic Sci Soc 1973;13:223–30. Kobus H, Silenieks E, Scharnberg J. Improving the effectiveness of fluorescence for the detection of semen stains on fabrics. J Forensic Sci 2002;47:819–23. Nelson D, Santucci K. An alternate light source to detect semen. Acad Emerg Med 2002;9:1045–8. Oepen I. Identification of characteristics in blood and semen stains – a review. Forensic Sci Int 1988;36:183–91. Ricci U et al.. Identification of forensic samples by using an infrared-based automatic DNA sequencer. Croat Med J 2003;44(3):299–305.