A method for measuring dermal exposure to multifunctional acrylates Jouni Surakka,* Stina Johnsson, Gunnar Rose�n, Tomas Lindh and Torkel Fischer National Institute for Working Life, S±112 79 Stockholm, Sweden. E-mail: Jouni.Surakka@niwl.se Received 16th June 1999, Accepted 4th October 1999 UV-curable acrylates are used increasingly for coating wood surfaces in the furniture industry. One of the active components, tripropylene glycol diacrylate (TPGDA), is known to be both an allergen and irritant to the skin.Methods to measure dermal exposure to skin irritants and allergens, such as acrylates, are insufÆcient for exposure assessment and there is none for this compound. The aim of this investigation was to develop a skin and surface sampling method, based on tape stripping, and a gas chromatographic method for quantitative analysis for assessing occupational skin exposure to multifunctional acrylates.Twelve adhesives were tested for their efÆciency to remove TPGDA and UV-coating from a glass surface, the skin of guinea pigs and human volunteers employing the tape-stripping method in order to Ænd the best performing tape. Variables that affect removal efÆciency such as the applied dose and its retention time on the skin, tape adhesion time on the skin, and the number of strippings required to detect the contaminant from the skin were studied.Fixomull1 tape performed the best during sampling and analysis and had the most consistent removal efÆciencies for the studied substances. The average removal efÆciency with a single stripping at the 2 ml TPGDA exposed skin sites was 85% (RSD~14.1), and for UV-resin exposed sites 63% (RSD~20.2).The results indicated that this method can be used for measuring dermal exposure to multifunctional acrylates efÆciently, accurately, and economically. This method provides a sensitive and powerful tool for the assessment of dermal exposure to multifunctional acrylates both from the skin and from other contaminated surfaces in occupational Æeld settings.Introduction The use of ultraviolet radiation curable coatings (UV-coatings) has increased rapidly in the furniture and parquet industry. UV-technology provides several advantages over the use of traditional solvent-based coating. UV-technology is economical, fast, saves energy, produces a hard, Ænished surface resistant to chemicals and the solvent emissions in the process are negligible.There are both physical and chemical risks for hazardous exposure while working on the UV-curing line. The physical hazard of concern is dermal exposure to the high intensity UVradiation. 1 UV-lacquers are important chemical risk factors to the skin. In Sweden, there are about 50 facilities (1997) that employ the technique of UV-radiation curable coatings on wood productsand approximately 350 workers are potentially exposed to UV-radiation and UV-curable coatings.2 Unintentional skin contact with uncured UV-coatings e.g., from tools and surfaces contaminated with UV-coatings is common.UV-coatings are usually free from evaporating solvent and contain reactive and biologically active acrylates.Yet, the methods to measure dermal exposure to UV-curable paints and lacquers are lacking. In Europe, the wood surface coating industry is the dominant sector (50% of the market) in the use of UV-curable acrylate coatings. In the last 15 years, the use of these coatings has doubled every Æfth year with a 15% annual growth.3 In 1998, 37 780 tons of multifunctional acrylates (MuFAs) were used in Europe with a predicted market growth to 51 590 tons in 2003.This means an expected doubling of the use of MuFAs in the next 10 years.4 Commonly used UV-coatings are composed of three basic components; an acrylate prepolymer (e.g., urethane acrylate, polyester acrylate), a MuFA monomer [e.g., tripropylene glycol diacrylate (TPGDA), trimethylolpropane diacrylate], and a photoinitiator system (e.g., benzophenone, benzil dimethyl ketal).5 Acrylates are well known as skin contact irritants and sensitizers and may also induce respiratory hypersensitivity.6,7 Allergic reactions from acrylates may develop from a wide variety of products and in different occupations.8±13 The risk of inducing contact allergy depends both on the sensitization capacity of the chemical and its penetration into the skin, which depends on the physico-chemical properties of the substance, concentration, skin barrier function and time of exposure.14,15 The American Conference of Governmental Industrial Hygienists has established occupational threshold limit values (TLV1s) for four monofunctional acrylates: methyl acrylate, ethyl acrylate, n-butyl acrylate, and methyl methacrylate. 16 Only methyl acrylate has been assigned with a skin notation indicating that a signiÆcant contribution to the total exposure may occur by direct skin contact with the substance.The American Industrial Hygiene Association (AIHA) recommends an 8 h time-weighted average (TWA1 8h) workplace environmental exposure level (WEEL) of 1 mg m23 with a skin notation for MuFAs.17 Recently, TPGDA and two other methacrylate compounds (ethylene glycol dimethacrylate and 2-hexyl propylene methacrylate) have been classiÆed as skin sensitizing agents in Finland.18 Tripropylene glycol diacrylate (TPGDA) is the most commonly used multifunctional acrylate and it has a strong sensitizing capacity.TPGDA and other multifunctional acrylates have low vapour pressure and low water solubility. Thereby they remain for a relatively long period of time on the skin after deposition. With this in mind we wanted to develop a method for measuring skin exposure to commonly used UV-coatings in the furniture and parquet industry.Methods for measuring dermal exposure to chemicals In many occupations, dermal exposure can be the dominant entry of hazardous chemicals into the skin.19,20 Dermal penetration has been studied for many compounds through analysis of the exposed compound and its metabolites in urine, faeces and respired air.Assessment of dermal exposure from J. Environ. Monit., 1999, 1, 533±540 533 This journal is # The Royal Society of Chemistry 1999analysis of skin deposition has seldom been used and has primarily focused on exposures to pesticides21±23 and toxic metals.24 Few methods have been so far developed to measure dermal exposure to allergens (e.g., acrylates).Skin sampling techniques for estimating dermal exposure to chemicals have been divided into three categories. (1) Surrogate skin techniques, e.g., the use of patch or garment samples,24 (2) removal techniques, e.g., washing, wiping, scraping and brushing of the skin, and (3) direct reading instruments that employ Øuorescent tracer techniques, e.g., video imaging.24,25 Aim of investigation The tape-stripping technique has been widely accepted as a dermal sampling technique in dermatology.26±31 Adhesives have also been used in occupational hygiene but mainly for sampling surfaces contaminated by asbestos or glass Æbres, toxic metals, or fungi.Recently, this method has been used for developing a quantitative dermal exposure assessment method for hazardous metals.32 This technique is applicable for compounds which have low volatility and remain on the skin for a signiÆcant period of time like MuFAs. Methods to measure dermal exposure to UV-coatings in the occupational or environmental setting are lacking.The aim of this investigation was to develop a reliable and simple skin and surface sampling method that is based on tape stripping, and a gas chromatographic method for quantitative analysis of samples. The method we developed is consistent, sensitive and accurate for measuring skin and surface exposure to MuFAs.This method provides a new powerful tool for identifying workers at risk of dermal exposure to allergens and is applicable for several compounds and occupations. Materials and methods The selection of an adhesive for measuring the dermal deposition of TPGDA was performed in three stages. First, tape-stripping studies were conducted in order to evaluate the applicability of eleven adhesive tes and a tissue adhesive. Second, nine adhesive tapes and the tissue adhesive were further tested using the skin of recently euthanized guinea pigs.Third, two adhesive tapes were tested on the palms and the volar region of the lower arms of human volunteers. During the Ærst two study stages, only TPGDA was used. During the third study stage, both TPGDA and UV-resin were used.A priori determined amounts of TPGDA and/or UV-resin were used in all experiments. The selected adhesives were tested concerning their efÆciency and accuracy for removing the deposited material. Selection of the materials Adhesive tapes. Eleven commercially available adhesive tapes and one tissue adhesive (referred to as `tissue adhesive' from hereon) were chosen for the study (Table 1).Mylar1 Ælm was used as a backing of the `tissue adhesive' for the collection and analysis of the compound deposited on a glass surface. The adhesives were selected based on their potential technical properties for skin sampling purposes and previous experience in dermatology.30,33±35 Eight tapes were cut to size 25 mm640 mm (10 cm2) using a surgical knife and stored individually in containers at room temperature.Two of the adhesives were commercially available for dermal sampling per se (D-Squame1 and Sebutape1). Multifunctional acrylate and UV-lacquer. Tripropylene glycol diacrylate (TPGDA, CAS No. 42978-66-5, technical grade) was obtained from Akzo Nobel Industrial Coatings AB, Malmo» , Sweden. A commercially available UV-resin, which contained 40±50% TPGDA, was obtained from Becker Acroma AB, Ma» rsta, Sweden (Table 2).We determined the purity of the TPGDA, according to our method, to be 86% (w/ w). This is in accordance with the previously published studies36,37 that reported TPGDA to be 81±91% pure in studies where acrylates used in dermal patch testing were investigated. Tape-stripping study The following parameters were evaluated: I Sampling from glass slides; (1) the efÆciency with which the adhesive removed the deposited compound from the surface (removal efÆciency) and (2) the ease with which the adhesive could be handled when applying and removing it from the surface.II Sampling from guinea pig skin; (1) the ease with which the adhesive could be applied to and removed from the skin, (2) solubility of the adhesive material in the solvent, and (3) the indifference of the gas chromatographic analysis to dissolved adhesive material. III Sampling from human skin; the most promising adhesives were further tested on human skin to Ænd the best adhesive for skin sampling.Tape stripping of glass surfaces. TPGDA (5 ml) was carefully applied onto a factory cleaned microscopic slide (76626 mm, KEBOLab, Stockholm, Sweden) using a calibrated digital micropipette (Biohit Proline1 0.5±10 ml, Biohit Oy, Finland).Since TPGDA is a viscous liquid, residue outside the pipette tip was carefully wiped off by a lint-free lens tissue before application in order to deliver a constant amount of the material to the test surface. After 20 min, the adhesive was applied over the surface where TPGDA was deposited.The adhesive Ælm was inspected for any defects prior to application. After one minute, the adhesive was peeled off from the surface using a clean forceps and constant force. The adhesive was removed in a 30±45 degree angle in order to minimize the effect of the peeling force on removal efÆciency.27 The forceps were rinsed in acetone between each stripping to avoid contamination.The testing of each adhesive on a glass slide was performed in duplicate and each glass slide was stripped three successive times. Tape stripping of guinea pig skin. Hair was shaved around the torso, between the fore and hind limbs. TPGDA (5 ml) was applied neat to the skin using a micropipette and the exposed site was marked in 1.25 cm radius with a black marker.The TPGDA-exposed site was retained uncovered for 1, 15, 30, or 120 min. Adhesive was retained over the exposed skin site for 1 min and removed using the same standardized manner as with glass application. Three to four strippings per exposure site were performed. If the exposed surface area was longer than the sampling area of the adhesive, the second stripping was arranged to cover also the unsampled area.A surface thermometer (ALMEMO 2290-8, Ahlborn Mess- und Regelungsteknik, Holtzkirchen, Germany) equipped with a sensor (Pt 100, same manufacturer) was used to continuously monitor skin temperature during testing. The study was approved by the Stockholm's Northern Ethical Research Council for the Use of Laboratory Animals (nr N 204/97).The animals were delivered and prepared by the staff at the Department of Laboratory Animals, National Institute for Working Life, Solna, Sweden. Tape stripping of human volunteers. PredeÆned areas on subject's Ængertips, palms, and inner volar region of the lower arms were exposed to TPGDA or UV-resin (see Fig. 1). The compounds were applied in a similar manner as described previously.TPGDA was applied at dosages of 5, 2.5, 2 or 1.0 ml and UV-resin at a dose of 2.5 or 2.0 ml. The exposed site was marked within a 1.5 cm radius and retained uncovered for 30 min before stripping by the same standardized technique as 534 J. Environ. Monit., 1999, 1, 533±540in the previous studies. The skin was exposed to the study compounds without cleaning in order to imitate the natural effects of dermal excretions (sebum, sweat) and accumulated dirt, which is the case in the normal occupational exposure settings.First, we investigated the effect of the adhesion (contact) time of the adhesive on the skin (1 or 2 min) on the removal efÆciency. Eight predeÆned areas, four on the left and four on the right arm, were exposed to 5 ml of TPGDA.This provided four similarly exposed skin sites sampled by each adhesive. Each exposed site was sampled by three successive strippings (see Fig. 1, Test A). Additionally, one sample, serving as a reference, was taken from an unexposed site with each of the adhesives tested. Second, reproducibility was investigated by applying 2.5 ml of TPGDA or UV-resin onto predeÆned areas on the lower arm and palm (see Fig. 1, Test B). Two successive tape strippings were performed using a 2 min contact time for the adhesive. The temperature of the skin was continuously monitored by a surface thermometer (ALMEMO 2290-8) and a sensor (Pt 100) Table 1 The adhesives investigated in the study Adhesive tape Description and dimensions Purpose Supplier Bioclusive1 Opaque rubber-like Ælm, adhesive undeclared, cut to size 25 mm640 mm Medical, dermatology, wound Ælm Johnson & Johnson AB, Sweden Blenderm1 Opaque, elastic Ælm, adhesive undeclared, cut to size 25 mm640 mm Medical, dermatology, wound Ælm Johnson & Johnson AB, Sweden D-Squame1-disc Clear adhesive discs, diameter of 22 mm, strong but undeclared adhesive Skin surface sampling/ assessment, visualize patterns of dry skin CuDerm Europe, Denmark Fixomull1 Self-adhesive gauze, woven polyester backing, polyacrylate adhesive, cut to size 25 mm640 mm Medical, dermatology Beiersdorf AB, Sweden MeÆx1 Nonwoven polyester Æbres, polyacrylate adhesive, cut to size 25 mm640 mm Medical, self-adhesive gauze Mo» lnlycke AB, Sweden Polyester tape, 850 Opaque tape, adhesive not known, cut to size 25 mm 6 40 mm Dermatology, for fungus sampling 3M Svenska AB, Sweden Scanpor1 Nonwoven polyester Æbres, polyacrylate adhesive, cut to size 25 mm640 mm Dermatology, medical Norgeplaster A/S, Kristiansand, Norway Sebutape1 Adhesive patches with tabs, adhesive undeclared, size 32 mm619 mm Skin assessment, tape sensitive to follicle sebum production CuDerm Europe, Denmark Tegaderm1 1626 Cut to size 25 mm640 mm, adhesive undeclared Dermatology, wound Ælm 3M Svenska AB, Sweden Tesa 4287 Orange (13) polypropylene backing, natural rubber adhesive, width 19 mm Technical, packaging Beiersdorf AB, Sweden Tesa 4663 Soft aluminium foil tape, natural rubber adhesive Technical, strong adhesive for construction work Beiersdorf AB, Sweden Tissue adhesive Histoacryl 1 blue and Mylar1 -Ælm Ultrapure cyanoacrylate drop spread to polyester Ælm, cut to size 25 mm640 mm Surgical tissue adhesive and Ælm as a backing material for the cement B.Braun Melsunge AG, Woundhealing Division, Germany and Dupont, USA Fig. 1 Diagram of the different tape-stripping experiments conducted on human volunteers. Table 2 Chemical and physical properties of tripropylene glycol diacrylate (TPGDA) and the technical information of the UV-curable coating SpeciÆcation TPGDA UV-resin Molecular formula C15H24O6 NA CAS Numbers 42978-66-5 42978-66-5 55818-7-0 (Diacrylate of bisphenol A) Molecular weight 300.25 NAa Density at 25 �C (DINb 51757) 1.0355 g cm23 1.08 g cm23 Boiling point (DIN 51751) at 0.3 mbar 109 �C NA Flash point NA 105 �C Viscosity (DIN 51562) at 25 �C 10.5 mPa s NA at 45 �C 5.2 mPa s Vapour pressure at 20 �C v1025 mbar NA at 25 �C v1023 mbar Solubility in water at 25 �C 0.036 g/100 ml NA European Union ClassiÆcation Irritating Harmful aNA~Not available.bDIN~Deutsches Institut fu» r Normung, the German Institute for Standardization. J. Environ. Monit., 1999, 1, 533±540 535attached by an adhesive tape inside the arm 5 cm from the head of the ulna.Third, the effect of the dose of TPGDA on removal efÆciency was investigated by exposing the same skin sites as in the previous test. Six TPGDA doses were deposited onto one palm and arm (see Fig. 1, Test C); sites #1±4 received a dose of 2.5 ml each and sites #5 and #6 received a dose of 1 ml each. The adhesive was retained on the exposed site for 2 min.Fourth, the removal efÆciency of the best performing adhesive was investigated in detail on ten human volunteers (see Fig. 1, Test D). Three predeÆned areas on the left thumb, palm and the lower arm were exposed to TPGDA (2 ml on each site). Three corresponding areas on the right palm and the lower arm were exposed to UV-resin (2 ml on each site). In order to provide an even exposure the applied dose was smeared by a pipette tip to cover an area of about 1 cm2.The exposed site was retained uncovered for 30 min before sampling. One stripping using 2 min contact time was performed on each exposed site. Before stripping, the area over which TPGDA had spread during this time was registered when it was visually larger than the area of the adhesive.One sample was taken from an unexposed site on each arm. All tests of the human volunteers were conducted in the laboratories at the National Institute for Working Life in Solna. The study was approved by the Karolinska Institute Ethical Committee, Stockholm, Sweden (Dnr 2.2 AI 36/95). Sample handling and extraction The extraction efÆciency for the spiked TPGDA samples was studied with Æve different solvents: acetone, ethyl acetate, ethanol, toluene, and carbon disulÆde.Acetone produced the most consistent and repeatable results with high yield and was therefore evaluated as the most suitable for sample extraction. Adhesive samples were handled from the perimeters of the adhesive and then placed into the labelled 20 ml scintillation vials (Wheaton, Millville, NJ, USA).The vials were Ælled with 10 ml of acetone (p.a. quality) that was spiked with n-nonane (both from Merck, Darmstadt, Germany) as an internal standard, and closed with a cap with an aluminium insert. There was no signiÆcant difference between the extraction efÆciency for spiked TPGDA samples with 20 or 30 min extraction time or between the ultrasound or rotator shaked samples.Therefore, samples were shaken in a rotator shaker at 250 rpm for 30 min, and allowed to settle for 30 min, before aliquoting into autosampler vials for gas chromatographic analysis. Blank samples (one for every 10 samples) were prepared with the same procedure. Gas chromatographic analysis A modiÆcation of a previously published gas chromatographic (GC) method for analysis of airborne MuFAs38 was developed for the quantitative analysis of TPGDA from the samples.Analyses were performed on a Hewlett-Packard 5880A gas chromatograph equipped with a HP 7671A autosampler (Hewlett-Packard, Waldbronn, Germany) in splitless/split injection mode. Two different columns, HP-1 (30 m, 0.53 mm id, 2.65 mm Ælm thickness of 100% dimethylpolysiloxane; Hewlett-Packard, Waldbronn, Germany), and a DB-5.625 column (30 m, 0.32 mm id, 1.0 mm Ælm thickness of 95% dimethyl-, 5% diphenyl-polysiloxane, J&W ScientiÆc, Inc., California), were employed.The injector and the Øame ionization detector temperatures were 250 �C and 300 �C, respectively. The initial oven temperature was held at 50 �C for 1 min and increased at 20 �C min21 to 230 �C where it was held for 6 min.After analysis, the oven temperature was increased to 280 �C for 6 min to clean the column from compounds with higher retention times. The carrier gas (helium) Øow was 2.3 ml min21. The retention time for TPGDA peaks at (10 psi) column head pressure was 11 min. Statistical analysis Differences in means on the removal efÆciency were calculated employing Student's two-tailed t-test.The differences were classiÆed as signiÆcant if the p-value was v0.05. Relative standard deviation (RSD), expressed as accuracy, was used to compare the relative variation between different kinds of measurements. In the storage stability test, a linear regression model with time as an explanatory variable was used to analyse the data. Accuracy of the applied dose The accuracy of the applied dose was determined by applying 2 ml of UV-resin ten times directly onto the bottom of the 20 ml sample vial using a calibrated digital pipette (inaccuracy of 0.70% and imprecision of 1.67%) (Biohit Proline1 0.5±10 ml, Biohit Oy, Finland) equipped with a disposable polypropylene pipette tip.UV-resin is a viscous liquid and therefore residues outside the pipette tips were carefully wiped off by a tissue prior to application to deliver a constant amount of the material to the surface.Samples were processed and analysed as described previously. Storage stability test Samples were prepared by spiking the Fixomull1 tape pieces (25640 mm) with 2 ml of UV-resin using a calibrated digital pipette as described above.Adhesives were inspected for any surface defect prior to testing. The samples were placed into Table 3 Removal efÆciencies obtained with the different adhesives using three sequential strippings of TPGDA dose (5 ml, 20 min exposure) from the glass surface Removal efÆciency (%) Adhesive tape Strip 1 Strip 2 Strip 3 Total Bioclusive1 71.3 14.3 NDa 85.5 Blenderm1 67.8 19.5 0.5 87.8 D-Squame1 88.0 4.8 ND 92.8 Fixomull1 68.0 18.3 ND 86.3 MeÆx1 101.8 ND ND 101.8 Scanpor1 72.5 15.0 1.3 88.8 Sebutape1 66.8 15.8 ND 82.5 Tegaderm1 77.5 13.0 0.8 91.3 Tesa 4287 44.8 20.5 8.3 73.5 Tissue adhesive 29.0 7.8 32.0 68.8 aND~not detected.Fig. 2 The inØuence of the adhesion times of the Fixomull1 and DSquame 1 tapes on the removal efÆciencies when tested on the skin of guinea pigs.Each site was exposed to 5 ml of TPGDA doses for 30 min and stripped three consecutive times. 536 J. Environ. Monit., 1999, 1, 533±54020 ml scintillation vials, the cap was closed, and the samples were stored in the dark at room temperature for 0, 24, 48, 72, and 96 h. Five spiked samples and one blank were analysed for each time period. At the end of the storing period, the samples were extracted with 10 ml of acetone and processed as described previously.After extraction, the samples were stored in a freezer at 218 �C until analysis by GC. Limit of detection The analytical limit of detection (ALOD) for TPGDA (determined purity 86% w/w) and employing Fixomull1 adhesive as a sampling medium was 9 ng, which corresponds to an average surface concentration of 4.5 mg cm22 in the sampling area.For the UV-resin (contains 55% TPGDA) used in this study the respective ALOD corresponds to a surface concentration of 11.4 mg cm22. Removal efÆciency The removal efÆciency was calculated based on the predetert substances applied onto the test surface. All the results were calculated as applied TPGDA mass onto surface and converted to percentage removal efÆciency.Results Tape stripping from glass surface Two adhesives, 3M 850 Polyester tape and Tesa 4663 aluminium tape, were omitted from the study during the Ærst test on a glass surface due to difÆculties during sampling and processing of the samples and their poor removal efÆciencies. The calculated mean removal efÆciencies (recovery) for the remaining ten adhesives are presented in Table 3.A single tape stripping removed from 29% to 102% of the deposited compound with an average removal efÆciency of 68.7%. Although MeÆx1 had the highest removal efÆciency, it was rejected due to difÆculties in sampling and handling. Tape stripping the skin of guinea pigs Four adhesives were selected for tests on guinea pig skin (Table 4).The total removal efÆciency varied between 42±93% depending on skin exposure time and the type of the adhesive. On an average, the Ærst stripping removed 55% or more of the deposited TPGDA whereas the second stripping removed 6±29%. Fixomull1 had the most consistent removal efÆciency throughout the range of adhesion times and, in addition, the adhesive was easy to handle during sampling.Blenderm1 and the tissue adhesive were rejected from further studies due to difÆculties in sampling and handling. D-Squame1 did not cover totally the area of some TPGDA doses spread onto the skin, which decreased the removal efÆciency and increased the variation of the results. The effect of the adhesion time of the tape on the skin on the removal efÆciency was studied using D-Squame1 and Fixomull 1 adhesives.The skin of two guinea pigs was exposed to 5 ml of TPGDA on six sites for 30 min while a constant skin temperature was maintained. Three successive samples were obtained from each exposed site. Removal efÆciencies of the studied adhesives increased along with skin sampling time (Fig. 2). The removal efÆciency at 5 s sampling time was 62% for D-Squame1 and 65% for Fixomull1, while 60 s sampling time provided 95% and 96% removal efÆciency for D-Squame1 and Fixomull1, respectively.Tape stripping of human skin D-Squame and Fixomull adhesives were further tested on human skin. First investigation. The effect of skin sampling time (1 and 2 min) on the removal efÆciency of the deposited compound from the skin (Fig. 1, Test A). One-minute sampling time and with three successive strippings provided 46.4% yield of the deposited dose with D-Squame1, and 72.7% yield with Fixomull1. Doubling of the sampling time (2 min) increased the removal efÆciency to over 70% for both the adhesive types (Fig. 3). Observations prior to stripping indicated that the sites #1 and 2 on both arms had no visible residue left, but sites #3 and 4 had visible residues that had spread, following the furrows of the skin, to cover an area of about 1 cm2.Second investigation. The reproducibility of a single stripping (Fig. 1, Test B). D-Squame1 was used for stripping sites #1, 2, 5, 6, and 9, and Fixomull1 for sites #3, 4, 7 and 8. Each of ten volunteers received eight TPGDA doses (2.5 ml each) on sites #1±8 and one UV-resin dose (2.5 ml) on site #9.Visual observation of the exposed skin sites prior to stripping revealed that on sites #1±4 there were visible residues that covered an Table 4 Removal efÆciencies (%) of the adhesives obtained with sequential strippings of TPGDA (5 ml) from the skin of guinea pigs when four different exposure times and an adhesion time of 1 min were used Exposure time (min) Adhesive tape Strip No. 1 15 30 120 Blenderm1 1 57 47 58 47 2 21 16 21 9 3 7 4 3 3 4 NSa NS 3 1 Total 85 67 85 60 D-Squame1 1 70 26 79 43 2 15 13 6 13 3 4 3 1 5 4 NS NS 0.1 1 Total 89 42 86.1 62 Fixomull1 1 73 63 70 58 2 16 14 12 14 3 4 5 2 6 4 NS NS 1 3 Total 93 82 85 81 Tissue adhesive 1 NS 42 65 62 2 NS 29 16 8 3 NS 6 1 15 4 NS NS 0.3 3 Total N/Ab 77 82.3 88 aNS~sampling not performed.bN/A~not applicable. Fig. 3 Test `A', the inØuence of the adhesion times of the Fixomull1 and D-Squame1 tapes on the removal efÆciencies when tested on one volunteer. Four sites of the volar region of the lower arms were exposed to 2.5 ml of TPGDA for 30 min and stripped three consecutive times after 1 or 2 min adhesion time. J. Environ.Monit., 1999, 1, 533±540 537area of 3±5 mm in diameter. On sites #5 and #6, there were typically no residues visible to the eye. On sites #7 and #8, the TPGDA doses were visible and had spread over an area of about 1 cm2 following the furrows of the skin. UV-resin typically spread out symmetrically and covered an area of diameter of 3 to 4 mm; on some occasions TPGDA had spread out following the skin patterns on the palm.The highest removal efÆciency of 93.6% was obtained by DSquame 1 on the sites #1 and 2, whereas Fixomull1 at the lower palm sites #3 and 4 removed an average of 86.2% of the deposited TPGDA dose (Fig. 4). On the volar region of the wrist (Fig. 1, Test B, #5±6), TPGDA had spread over an area greater than D-Squame1 could cover. The average removal efÆciency at these sites for TPGDA was 63.3% and RSD 31.3%.A single tape stripping from exposed sites #7 and 8 gave an average removal efÆciency of 78.2% and RSD of 12.7%. UVresin exposed site #9 was stripped by D-Squame1 and the average removal efÆciency was 64.5% and RSD 11.4% (Fig. 4). The mean difference between the removal efÆciency for DSquame and Fixomull was not statistically signiÆcant at the 0.05 level.Third investigation. The removal efÆciency of Fixomull1 adhesive to TPGDA doses of 2.5 and 1 ml (Fig. 1, Test C). Prior to skin sampling there were clearly visible residues on sites #1± 3, whereas on sites #4±6 there were no visible residues. The removal efÆciency was dependent on the volume of the applied dose; for 1 ml dose it was 57±70% and for 2.5 ml dose it was 70± 96% (Fig. 5). The average skin temperature on volunteers during the 90 min long testing period was 28.4°1.9 �C. Fourth investigation. The efÆciency of Fixomull1 to remove TPGDA and UV-resin from the skin (Fig. 1, Test D). Test substances were applied to three predeÆned areas; on the thumb (site #1), on the palm (site #2) and on the volar region of the lower arm (site #3).TPGDA was applied to the left hand and UV-resin to the right hand on respective areas. For site #1 and #2 a visual inspection prior to stripping showed that both TPGDA and UV-resin could be seen on the skin, but on site #3 only UV-resin was still visible. Residues were analysed from Æve pipette tips; on average there was 0.15°0.08 mg (95% CI) TPGDA left after application. As shown in Table 5 a single tape stripping, for ten volunteers, removed on the average 85% of TPGDA and 62% of UV-resin.The difference in proportions was statistically signiÆcant at the 0.05 level. For TPGDA the result for each site was similar to the summary value. However, for UV-resin, site #1 (thumb) showed 3 out of ten samples with extremely low readings, which was probably due to errors.Excluding site #1, the following is a summary of results for UV-resin: Min~50.6%, Max~80.4%, Average~65.9%, s~7.8, n~30, RSD~11.8. The addition of a second tape stripping made the total removal efÆciency higher. For TPGDA the average removal efÆciency was 92.8% (RSD~13.5, n~30) and for UV-resin 77.6% (RSD~21.3, n~30). Excluding site #1 for UV-resin: 82.9% (RSD~13.2, n~20). Residues of the test substances were analysed from Æve pipette tips (2 for TPGDA and 4 for UV-resin).The average of the analysed TPGDA mass on pipette tips was 0.15 mg°0.08 (95% CI) (average for TPGDA tips was 0.11 mg and for UVresin tips 0.16 mg). Storage stability We investigated the stability of the TPGDA on UV-resin (2 ml) that was applied onto Fixomull1 adhesive and stored up to four days (96 h).We found that the regression coefÆcient for TPGDA break down inilligrams was (mg)~20.0000746 storing time (h)z1.10 (intercept). The test indicated that TPGDA concentration on the UV-lacquer could be sampled on Fixomull1 adhesive and stored at room temperature in darkness up to 4 days from the sampling without loss of TPGDA from the samples.Accuracy of the applied dose Based on TPGDA concentration in the UV-resin the average applied dose was 1.10°0.02 mg with an inaccuracy of 5.5% (coefÆcient of variation). According to our analysis UV-resin contained 55% (w/w) TPGDA. Discussion Good technical properties and the demand to decrease the use of solvents have increased the use of UV-curable coatings in the wood surface coating industry.3,39 Exposure to MuFAs poses a risk for skin irritation and sensitization and a potential for carcinogenic effects.2,40,41 Therefore, methods for studying dermal exposure to MuFAs are important to develop.Fig. 4 Test `B', the average removal efÆciencies and standard deviations (s) for Fixomull1 and D-Squame1 tapes when tested on ten volunteers. Each site was exposed to 2.5 ml of TPGDA or UV-resin for 30 min and stripped once after 1 min adhesion time.Fig. 5 Test `C', the average removal efÆciencies and standard deviations (s) for Fixomull1 tape when tested on Æve volunteers. Sites #1±4 (on palm and on the volar region of the lower arm) were exposed to 2.5 ml of TPGDA and sites #5 and 6 (on the volar region of the lower arm) were exposed to 1.0 ml of TPGDA for 30 min and stripped once after 2 min adhesion time.Table 5 Test `D': Removal efÆciency of a single tape stripping summarized for all 3 exposed skin sites employing Fixomull1 adhesive TPGDA UV-resin Statistic (%) (%) Average 85.0 62.0 Min 67.4 35.3 Max 106.2 80.4 s 12.0 12.5 n 30 30 RSD 14.1 20.2 538 J. Environ. Monit., 1999, 1, 533±540Fenske25 has indicated three categories of sampling techniques for estimation of dermal exposure.Each of them has advantages and limitations when considering the qualitative and quantitative value for measuring dermal exposure to a speciÆc chemical. We evaluated these sampling techniques in order to Ænd the optimal method that could be developed further for measuring dermal exposure to MuFAs.The Ærst category consisted of surrogate skin techniques (e.g., the use of patch or garment samples). This technique was not considered appropriate because it overestimates the actual dermal exposure and is not suitable for use on the hand.2,24,40 The second category included removal techniques (e.g., washing, wiping, scraping and brushing). These methods suffered from multiple sources of errors and they did not fulÆl the requirements for satisfactory quantitative sampling.The third category included direct reading instruments using Øuorescent tracer techniques (e.g., video imaging).24,25 An addition of a Øuorescent trace into the UV-curable coating used for commercial products is not possible, and therefore this technique had to be neglected. Tape stripping with an adhesive has been widely accepted as a dermal sampling technique.26±31 Tape stripping in fact is the only exposure assessment technique that had the potential to meet the requirements set for the ideal assessment of MuFAs.42 Adhesives have been used for sampling of asbestos Æbres, glass Æbres, toxic metals, and fungi from contaminated surfaces.However, Fenske did not consider this technique as a skin sampling method.25 This technique is applicable for compounds, which have a low volatility and remain on the skin for a signiÆcant period of time like MuFAs.The tape-stripping method has also been used as a quantitative dermal exposure assessment method for hazardous metals.32 There are a large number of adhesives with different properties available for dermatological, medical, and technical purposes.The adhesive to be used for measuring dermal exposure had to be validated. Two adhesives are commercially available for dermal sampling (D-Squame1 and Sebutape1) but none has been aimed for assessing a dermal exposure to chemicals. The key factors that need to be determined for an adhesive considered to be used for measuring dermal exposure are: removal efÆciency, adhesion time, extraction of the contaminant from the adhesive and non-interference of the sampling medium with acrylate analysis.42,43 Often other factors were more important and overcame the results from statistical tests when the tapes were ranked for further studies.The Ærst experiments on the glass surface provided necessary information about the suitability of the adhesives for sampling MuFA.We observed that the Ærst stripping of MuFA from the glass surface gave a wide variation in sampling efÆciency between the adhesives. However, several adhesives performed well indicating that many of them can be used for sampling skin exposed to MuFA. The skin of the guinea pig was used to study the effect of exposure time, adhesion time and the removal efÆciency of the applied TPGDA dose of sequential strippings.Four adhesives were selected for this study phase. Blenderm1 and Fixomull1 adhesives gave the more consistent removal efÆciency for TPGDA, than D-Squame1 and the tissue adhesive. The tissue adhesive had the strongest adherence to the skin. However, the hardened cement had a tendency to separate from the Øexible Mylar1-Ælm used as backing leaving a residue of the cement on the skin.D-Squame1 was observed to have a good adhesion to the skin and, therefore, it was included in further testing on human volunteers. The skin of the guinea pig has similar properties to human skin and therefore this animal was used for initial testing of the adhesives. However, the hairiness of the guinea pig skin, compared to the almost hairless human skin, is a matter of concern.Therefore, the selected adhesives were further tested on human skin. In the third study phase two adhesives qualiÆed for more detailed studies of the effects of the adhesive type, adhesion time, number of strippings and the type of contaminant on the removal efÆciency. The tests of adhesion time implied that both of the selected adhesives had lower performance on human than on guinea pig skin, but this could be overcome with two minutes adhesion time.Reproducibility tests conÆrmed that a single tape stripping, employed with either of the adhesives, could efÆciently remove TPGDA from the palm and lower arm. The differences between the test on guinea pig and on the Ærst human test fall within the error limits.The skin type and skin furrows were observed to affect the spreading of the contaminant on the skin. Spreading was more symmetric on the palm and the lower arm than on the wrist, where the skin is softer, thinner, and has more furrows. Both adhesive types had similar accuracy at two-minute adhesion time, but Fixomull1 had results that were more consistent over the studied adhesion times and for different skin areas.This is the basis for our choice of Fixomull1 adhesive. The Fixomull1 adhesive also has the advantage that it can be cut to the desired size. The third test (the effect of the deposited dose on the removal efÆciency) indicated that Fixomull1 adhesive provides stable performance at all exposed skin areas, and that as small doses as 1 ml are measurable after 30 min exposure time to the skin.The fourth test on human skin was to compare the removal efÆciency of pure TPGDA and UV-lacquer employing the Fixomull1 adhesive. Fixomull adhesive had high removal efÆciency for TPGDA and moderate for UV-resin. This lower removal efÆciency for UV-resin probably depends on an increased error with smaller applied dose.UV-resin has higher viscosity than TPGDA and was observed to stick to the pipette tips more than TPGDA. Analysis of some pipette tips indicated that small amounts of TPGDA remained on the tips, especially after the application of UV-resin. This may also explain the lower removal efÆciency of UV-resin compared to TPGDA. However, the RSD was moderately low indicating that the method is well Ætted for screening purposes.We performed a second stripping to ensure the removal efÆciency. Even though the UV-lacquer contained less than 60% TPGDA, which means a lower dose of TPGDA to the exposed skin site as compared to pure TPGDA, a single tape stripping with two minutes adhesion time provided good sensitivity and good accuracy with single stripping and the second stripping is not necessary.We were aware that our reference TPGDA compound was not pure, because this is not technically possible to produce. TPGDA that is used for dermal patch testing is reported to be 81±91% pure.36,37 TPGDA concentration analysis of the UVresin was in good agreement with the information obtained from the Material Safety Data Sheet in which the TPGDA content was stated to be 40±50%.The quantiÆcation of TPGDA from the adhesive samples was performed uniformly and therefore the results are comparable throughout the study. Despite the sources of error the total error for the Fixomull1 method was less than 20%. Our results indicate that Fixomull1 is a good adhesive for sampling TPGDA surface contamination at concentrations that may exist both on the exposed skin and on other surfaces at workplaces.In conclusion, we have developed a simple, consistent, and accurate method to measure dermal exposure to MuFAs. This method employs a single stripping with the Fixomull1 adhesive using a 2 min adhesion time on the surface of the skin or other surfaces to be investigated. In order to obtain a better understanding of dermal exposure to MuFAs experimentally derived information about the efÆciency of the tape-stripping method must be integrated J.Environ. Monit., 1999, 1, 533±540 539with occupational exposure measurements. Further studies are warranted in order to gain knowledge about the levels of dermal exposure that may occur in the wood surface coating industry.Acknowledgements We thank Dr Anders Boman (National Institute for Working Life, Solna, Sweden) for his assistance with laboratory animals, Dr. Eero Priha (Tampere Regional Institute of Occupational Health, Finland) and Mr. Bengt-Ove Lundmark (National Institute for Working Life, Solna, Sweden) for their comments in the development of the GC analysis. We are grateful for Associate Professor Leena A.Nylander-French at the University of North Carolina at Chapel Hill, NC, USA, for her comments during this work. We are grateful to all the volunteers; their valuable input made this investigation possible. We also thank Akzo Nobel Industrial Coatings AB, Malmo» and Becker Acroma AB, Ma» rsta, Sweden for providing the test chemicals. This research was supported by The Swedish Council for Working Life (Research Contract No. 95-0590). References 1 J. Surakka, T. Fischer, G. Rose�n and L. A. Nylander-French, Appl. Occup. Environ. Hyg., 1997, 12(4), 261. 2 L. A. Nylander-French, T. Fischer, M. Hultengren, M. Lewne� and G. Rose�n, Appl. Occup. Environ. Hyg., 1994, 9(12), 962. 3 J. Pernell, An overview of the radiation curing market in Europe for 1995, in 5th International UV/EB Processing Conference, Nashville, TN, 1996, RadTech International North America, 1996. 4 H. H. Bankowsky, E. Beck, W. Reich, P. Enenkel and M. Lokai, Radiation Curing in Europe. Internet: http://www.radcurenet.de/ marketframe.htm. Curt R. Vincentz Publishing, Hannover, Germany, 1999. 5 N. S. Allen, M. A. Johnson, P. K. T. Oldring and M. S. Salim, Reactive diluents for UV and EB curable formulations, in Prepolymers and Reactive Diluents for UV and EB Curable Formulations, ed.P. K. T. Oldring, SITA Technology, London, 1991, vol. 2, p. 237. 6 B. Bjo» rkner, Contact Derm., 1984, 11(4), 236. 7 P. Piirila» , L. Kanerva, H. Keskinen, T. Estlander, M. Hyto»nen, M. Tuppurainen and H. Nordman, Clin. Exp. Allergy, 1998, 28(11), 1404. 8 L. Kanerva, T. Estlander and R. Jolanki, Curr. Probl. Dermatol., 1996, 25, 86. 9 M. Kiec-Swierczynska, Medycyna Pracy, 1994, 45(4), 297. 10 R. E. Ranchoff and J. S. Taylor, J. Am. Acad. Dermatol., 1985, 13(6), 1015. 11 T. Rustemeyer and P. J. Frosch, Contact. Derm., 1996, 34(2), 15. 12 J. S. Taylor, Acrylic reactions–ten years' experience, in Current Topics in Contact Dermatitis, ed.P. J. Frosch, A. Dooms- Goossens, J.-M. Lachapelle, R. Rycroft and R. Scheper, Springer, Berlin, 1989, p. 346. 13 A. Tosti, L. Guerra, C. Vincenzi and A. M. Peluso, Toxicol. Indust. Health, 1993, 9(3), 493. 14 V. Fiserova-Bergerova, Ann. Occup. Hyg., 1993, 37(6), 673. 15 P. Grandjean, Skin Penetration: Hazardous Chemicals at Work, Taylor & Francis, London, 1990. 16 ACGIH, Threshold Limit Values for Chemical Substances and Physical Agents, American Conference of Governmental Industrial Hygienists, Cincinnati, OH, 1999. 17 AIHA, Hexanediol diacrylate, in Workplace Environmental Exposure Level Guide, American Industrial Hygiene Association, Washington, 1981, p. 51. 18 The Finnish ScientiÆc Committee on Health Effects of Chemicals, ClassiÆcation on Ethylene Glycol Dimethacrylate (EGDMA), 2- Hydroxypropyl Methacrylate (2-HPMA) and Tripropylene Glycol Diacrylate (TPGDA) for their Skin Sensitizing Properties, Helsinki, Finland, 1996. 19 J. W. Cherrie and A. Robertson, Ann. Occup. Hyg., 1995, 39(3), 387. 20 J. G. van Rooij, M. M. Bodelier-Bade and F. J. Jongeneelen, Br. J. Ind. Med., 1993, 50(7), 623. 21 J. C. Reinert, A. P. Nielsen, C. Lunchick, O. Hernandez and D. M. Mazzetta, Toxicol. Lett., 1986, 33(1±3), 183. 22 D. H. Brouwer, E. J. Brouwer and J. J. van Hemmen, Am. J. Ind. Med., 1994, 25(4), 573. 23 G. Chester, Ann. Occup. Hyg., 1993, 37(5), 509. 24 S. A. Ness, Surface and Dermal Monitoring for Toxic Exposures, van Nostrand Reinhold, New York, 1994. 25 R. A. Fenske, Ann. Occup. Hyg., 1993, 37(6), 687. 26 H. Pinkus, J. Invest. Dermatol., 1951, 19, 431. 27 J. K. Prall, Arch. Biochem. Cosmetol., 1966, 9, 87. 28 D. Porter and S. Schuster, J. Invest. Dermatol., 1967, 49, 251. 29 H. L. Jenkins and J. A. Tresise, J. Soc. Cosmetic Chemists, 1969, 20, 451. 30 J. Serup, A. Winther and C. Blichmann, Clin. Exp. Dermatol., 1989, 14, 277. 31 M. J. Gerritsen, P. E. van Erp, I. M. van Vlijmen-Willems, L. T. Lenders and P. C. van de Kerkhof, Arch. Dermatol. Res., 1994, 286(8), 455. 32 C. Cullander, D. Imbert and G. Bench, Fundam. Appl. Toxicol., 1997, 36(1, Part 2, Supplement), 192. 33 F. Dreher, A. Arens, J. J. Hostynek, S.Mudumba, J. Ademola and H. I. Maibach, Acta Derm. Venereol., 1998, 78(3), 186. 34 T. Fischer, A . Dahlen and B. Bjarnason, Contact Derm., 1999, 40(1), 32. 35 J. E. Hale, Postgrad. Med. J., 1970, 46, 447. 36 M.-L. Henricks-Eckerman and L. Kanerva, Am. J. Contact Derm., 1997, 8(1), 20. 37 L. Kanerva, M.-L. Henricks-Eckerman, R. Jolanki and T. Estlander, Clin. Dermatol., 1997, 15, 533. 38 L. A. Nylander-French, E. Priha, G.-B. Berglund and G. Rose�n, Appl. Occup. Environ. Hyg., 1994, 9(12), 977. 39 K. Lawson, Current status of UV/EB curing in North America, in RadTech Report Northbrook, IL, 1996, vol. 10, pp. 15±19. 40 T. Fischer, L. A. Nylander-French and G. Rose�n, Am. J. Contact Derm., 1994, 5(4), 201. 41 L. A. Nylander-French and J. E. French, Toxicol. Pathol., 1998, 26(4), 476. 42 J. J. van Hemmen and J. K. Brouwer, Sci. Tot. Environ., 1995, 168, 131. 43 D. Satas, Handbook of Pressure Sensitive Adhesive Technology, Van Nostrand Reinhold, New York, 1989, pp. 61±95. Paper 9/04816B 540 J. Environ. Monit., 1999, 1, 533&