Abstract Background Diisocyanates (isocyanates), including methylene diphenyl diisocyanate (MDI), are the primary reactive components of spray polyurethane foam (SPF) insulation. They are potent immune sensitizers and a leading cause of occupational asthma. Skin exposure to isocyanates may lead to both irritant and allergic contact dermatitis and possibly contribute to systemic sensitization. More than sufficient evidence exists to justify the use of protective garments to minimize skin contact with aerosolized and raw isocyanate containing materials during SPF applications. Studies evaluating the permeation of protective garments following exposure to SPF insulation do not currently exist. Objectives To conduct permeation testing under controlled conditions to assess the effectiveness of common protective gloves and coveralls during SPF applications using realistic SPF product formulations. Methods Five common disposable garment materials [disposable latex gloves (0.07 mm thickness), nitrile gloves (0.07 mm), vinyl gloves (0.07 mm), polypropylene coveralls (0.13 mm) and Tyvek coveralls (0.13 mm)] were selected for testing. These materials were cut into small pieces and assembled into a permeation test cell system and coated with a two-part slow-rise spray polyurethane foam insulation. Glass fiber filters (GFF) pretreated with 1-(9-anthracenylmethyl)piperazine) (MAP) were used underneath the garment to collect permeating isocyanates. GFF filters were collected at predetermined test intervals between 0.75 and 20.00 min and subsequently analyzed using liquid chromatography-tandem mass spectrometry. For each garment material, we assessed (i) the cumulative concentration of total isocyanate, including phenyl isocyanate and three MDI isomers, that effectively permeated the material over the test time; (ii) estimated breakthrough detection time, average permeation rate, and standardized breakthrough time; from which (iii) recommendations were developed for the use of similar protective garments following contamination by two-component spray polyurethane foam systems and the limitations of such protective garments were identified. Results Each type of protective garment material demonstrated an average permeation rate well below the ASTM method F-739 standardized breakthrough rate threshold of 100.0 ng/cm2 min−1. Disposable latex gloves displayed the greatest total isocyanate permeation rate (4.11 ng/cm2 min−1), followed by the vinyl and nitrile gloves, respectively. The Tyvek coverall demonstrated a greater average rate of isocyanate permeation than the polypropylene coveralls. Typical isocyanate loading was in the range of 900 to 15,000 ng MDI/cm2. Conclusion Permeation test data collected during this study indicated that each type of protective garment evaluated, provided a considerable level of protection (i.e. 10–110-fold reduction from the level of direct exposure) against the isocyanate component of the SPF insulation mixture. Nitrile gloves and polypropylene coveralls demonstrated the lowest rate of permeation and the lowest cumulative permeation of total isocyanate for each garment type. diisocyanate, permeation, protective garment, spray polyurethane foam Introduction Increased emphasis on energy loss and green building design has resulted in a higher demand for efficient insulating materials, including spray polyurethane foam (SPF) insulations. SPF insulations have a higher thermal efficiency than air-based insulating materials and offer increased resistance to moisture, microbial growth, and pests. SPF insulation may be blown into place in-situ. This allows for direct application to irregular surfaces, cracks, and seams to effectively eliminate air infiltration and provide increased structural reinforcement (Randall and Lee, 2002; American Chemistry Council, 2012). Professionally applied SPF insulations are most commonly two-component systems that are applied over large surface areas using high-pressure delivery systems. Low-pressure systems or self-contained kits may also be used for smaller applications. The ‘A-side’ component is typically a blend of monomeric and polymeric methylene diphenyl diisocyanate (MDI); however, it also includes small amounts of phenyl isocyanate that are produced due to secondary reactions during the manufacturing process. The ‘B-side’ of the mixture is comprised of polyols, catalysts, surfactants, blowing agents, and fire retardants. The B-side formulation has the most influence over the final physical and mechanical properties of the foam and has the most variability between specific product-lines and manufacturers. The precise B-side formulations are usually considered trade secret by manufacturers and not made publicly available to consumers or applicators (American Chemistry Council, 2012). Two-component high-pressure SPF insulations are combined at a 1:1 ratio and distributed onto the substrate at a thickness of one to three inches using a proportioner system and spray gun. The proportioner system pumps each chemical component from separate 55-gallon drums through heated hoses to decrease the viscosity of each component. A rapid polymerization reaction occurs once the components are mixed at the spray gun head, resulting in the rapid expansion and curing of the foam. Self-contained kits consist of two pressurized cylinders that deliver each component through nylon hoses and a spray nozzle. Two-component SPF insulations are commonly applied at high volumes and over large surface areas under the non-standard conditions of a dynamic construction environment. Personal exposures to isocyanates are likely highly dependent on the environmental conditions, application methods, and SPF formulations used during application. The manual application process requires the worker to stand near the point of application, creating significant potential for both airborne exposure and dermal contact with the aerosolized raw materials. During field work, we regularly observed contractors in significantly contaminated conditions, with wide-spread deposits of SPF insulation coating their personal protective equipment and clothing during and following application. The health effects of isocyanates are well documented and include acute irritation of the skin, mucous membranes, and respiratory and gastrointestinal tracts. Immune sensitization may also lead to severe respiratory diseases such as hypersensitive pneumonitis, reactive airway dysfunction syndrome, and isocyanate induced asthma (Rom and Markowitz, 2007). Although inhalation exposures to isocyanates have historically been the primary concern in the workplace, skin exposures are also known to cause serious health effects, including localized irritation that is mild or temporary in nature (Larsen et al., 2001; Daftarian et al., 2002) and allergic contact dermatitis following occupational skin exposures to polyurethane foams and adhesives containing MDI (Engfeldt et al., 2013; Kieć-Świerczyńska et al., 2014). It has also been suggested that dermal exposures may lead to systemic sensitization and an increased risk for occupational asthma. The relationship between skin exposure to isocyanates and systemic respiratory sensitization has been well documented with animal models, but there is presently limited epidemiological evidence to confirm this association in humans (Bello et al., 2007; Redlich, 2010). Although there is a need for further research, more than sufficient evidence exists to justify the use of protective garments to minimize skin contact and the potential for isocyanate sensitization (Heederik et al., 2012; Jones, 2016; Jones et al., 2017). Unfortunately, dermal exposures to isocyanates are not currently regulated and occupational exposure limits do not include a skin notation to indicate their potential for systemic toxicity following dermal contact. There are currently no published studies that characterize the potential for dermal exposure during SPF insulation application. Preliminary data from our ongoing dermal exposure assessment during SPF applications points to high potential for dermal exposure to isocyanates on hands and other body parts (Bello et al., 2017). The permeation testing of protective garments is typically conducted by manufacturers in accordance with one of the following three standard methods: International Organization for Standardization (ISO) Standard 6529 (ISO 6529, 2013), American Society for Testing and Materials (ASTM) method F-739 (ASTM F-739, 2012), or British and European Standard (BS EN) 16523-1 (BS EN 16523-1, 2015). There are significant limitations to the practical applications of data obtained using these methods, primarily because they only evaluate garment permeation using a single test chemical rather than a formulation. Exposure to a single chemical is often unrealistic under many real-world working conditions, particularly in the construction trades. Consideration must be given to differences in permeation between pure chemicals and chemical mixtures (Mickelsen et al., 1986; Ceballos et al., 2011). In addition, these test methods do not account for the toxicity of the specific test chemical when identifying an unacceptable level of breakthrough. Each of these methods has established a ‘standardized’ or ‘normalized’ break through time as the threshold of acceptability. This threshold is defined as the point at which the permeation rate has reached 0.1 μg/cm2 min−2 for all three methods, but because this threshold does not account for the toxicity of the specific test chemical, these methods cannot infer a safe level of exposure. Another factor that should be considered when dealing with potent sensitizing agents is the acceptable level of cumulative permeation. Mäkelä et al. (2014) evaluated the resistance of a variety of protective gloves to pure 4,4′-MDI vapors including thin materials such as 0.07–0.10 mm-vinyl and nitrile, as well as thicker materials such as 0.51–0.79 mil natural rubber and neoprene rubber. A cumulative permeation end-point of 1.0 μg/cm2 was recommended during this study, due to the sensitizing capacity of MDI. Cumulative permeation is not currently evaluated by ISO 6529, ASTM F-739 or BS EN 16523-1. Permeation studies have been conducted to evaluate the resistance of protective gloves to polymerizing isocyanate containing spray paints, including a mixture of hexamethylene diisocyanate (HDI) and isophorone diisocyanate (IPDI) monomers and pre-polymers. These studies have demonstrated the immediate breakthrough of both thin latex and nitrile materials (i.e. 0.10–0.13 mm thickness; Ceballos et al., 2011, 2013). There are currently no published studies that evaluate the permeation of protective garment materials following exposure to a mixture of reacting SPF insulation. Therefore, the objectives of this study were to (i) measure the cumulative concentrations of total isocyanate and its mixture components that effectively permeate a common selection of disposable protective garments; (ii) estimate the breakthrough detection time, average permeation rate, and standardized breakthrough time; and (iii) identify the limitations of such protective garments and develop recommendations for the use of similar protective garments following contamination by two-component SPF systems. Methods Protective garment selection and preparation Cost, availability, thickness, comfort, and color of protective gloves have been identified as influential factors in the selection and use of protective gloves during automotive spray painting, an industry with a similar risk for skin exposure to isocyanates (Ceballos et al., 2014). Ceballos et al. (2014) found that chemical compatibility was not a primary consideration during glove selection and that employees commonly relied on glove distributers and not manufacturers for chemical compatibility information. This study also found that spray painters maintained a preference for thin gloves, due to their increased dexterity and comfort. Anecdotal evidence obtained during conversations with SPF insulation contractors and the direct observation of SPF insulation application, supports similar findings with SPF contractors and indicates that the use disposable ‘off-the-shelf’ protective garments is a common practice within the industry. The intent of this study was to evaluate the most common types of protective garment materials used by SPF insulation contractors. We focused on low-cost disposable protective garments that were readily available from national ‘brick-and-mortar’ home improvement retailers to accomplish this objective. Our garment selection was further supported by field observations that were conducted as part of a larger exposure assessment study among spray foam applicators in the New England region (manuscript under preparation). Prior to conducting air and dermal sampling, contextual information was collected on tasks/activities, including their duration, number of workers, product information and personal protective equipment. These field observations were used to guide the current study design with regard to the types of garments, exposure levels, exposure durations, and work practices. Among the 53 study participants, 84% (40) wore 3-mil thick nitrile gloves, 12% (6) wore firm grip nitrile coated gloves, 4% (2) wore latex gloves, and 2% (1) did not wear any gloves. All but one sprayer were observed wearing Tyvek coveralls. Sprayers often wore their gloves for the entire duration of spraying tasks. It was typical for the task durations to run up to 2 h, after which the change-out of raw materials was necessary. Although sprayers would often change disposable protective gloves between each spraying interval, nitrile-coated cotton gloves were sometimes reused. An in-person and on-line review of local and national retailers identified that latex, nitrile, and vinyl protective glove materials with a nominal thickness of 0.07 mm were commonly available off-the-shelf, and they were therefore selected for testing. Thicker disposable gloves were often marketed as ‘Heavy Duty’ and sold at a higher price-point at some retailers. Tyvek and polypropylene disposable coveralls with an approximate thickness of 0.13 mm were also available off-the-shelf at the home improvement retailers visited and were therefore included in our testing. Due to the thin construction of these disposable protective garments, it is likely that this study not only presents the most likely scenario for PPE use, but also the worst-case scenario for permeation when compared with higher cost and less readily available heavy duty or non-disposable protective garments. Descriptions of each type of protective garment material evaluated are provided in Table 1. Each protective garment specimen consisted of a 7.62 cm diameter circle cut from the palm of a protective glove or a contiguous portion of coverall. Thickness was measured at the center of the specimen using a digital caliper. Individual specimens were inspected for defects prior to testing and discarded if inconsistencies were visually identified. Table 1. Total isocyanate permeation rate (ng NCO cm−2 min−1), breakthrough detection time, and protection factor by garment type. Garment type Material Manufacturer Average measured thickness (mm) Cumulative permeation, at 20 min (ng/cm2) Protection factor Linear regression breakthrough detection (min) Permeation rate R2 Glove Latex Blue Hawk 0.07 87.38 10 ≤0.75 4.11 0.88 Glove Nitrile Blue Hawk 0.05 8.05 110 ≤0.75 0.14 0.29 Glove Vinyl HDX 0.07 46.05 20 ≤0.75 2.29 0.90 Coverall Polypropylene 3M 0.12 14.76 60 ≤0.75 0.43 0.54 Coverall Tyvek Pro Line 0.13 31.66 30 ≤0.75 0.67 0.37 Garment type Material Manufacturer Average measured thickness (mm) Cumulative permeation, at 20 min (ng/cm2) Protection factor Linear regression breakthrough detection (min) Permeation rate R2 Glove Latex Blue Hawk 0.07 87.38 10 ≤0.75 4.11 0.88 Glove Nitrile Blue Hawk 0.05 8.05 110 ≤0.75 0.14 0.29 Glove Vinyl HDX 0.07 46.05 20 ≤0.75 2.29 0.90 Coverall Polypropylene 3M 0.12 14.76 60 ≤0.75 0.43 0.54 Coverall Tyvek Pro Line 0.13 31.66 30 ≤0.75 0.67 0.37 Time-interval data and the predicted trend of permeation data indicated that all glove materials permeated prior to or at the first test interval of 0.75 min. Nitrile gloves demonstrated the most resistance to total isocyanates with the lowest average rate of permeation (0.14 ng NCO/cm2 min-1) of all glove materials and the lowest cumulative permeation at 20 min (8.05 ng NCO/cm2). The average nitrile glove thickness was 0.05 mm. Glove permeation results and the 95% confidence interval for the slope of each regression line are presented in Fig. 3. View Large Table 1. Total isocyanate permeation rate (ng NCO cm−2 min−1), breakthrough detection time, and protection factor by garment type. Garment type Material Manufacturer Average measured thickness (mm) Cumulative permeation, at 20 min (ng/cm2) Protection factor Linear regression breakthrough detection (min) Permeation rate R2 Glove Latex Blue Hawk 0.07 87.38 10 ≤0.75 4.11 0.88 Glove Nitrile Blue Hawk 0.05 8.05 110 ≤0.75 0.14 0.29 Glove Vinyl HDX 0.07 46.05 20 ≤0.75 2.29 0.90 Coverall Polypropylene 3M 0.12 14.76 60 ≤0.75 0.43 0.54 Coverall Tyvek Pro Line 0.13 31.66 30 ≤0.75 0.67 0.37 Garment type Material Manufacturer Average measured thickness (mm) Cumulative permeation, at 20 min (ng/cm2) Protection factor Linear regression breakthrough detection (min) Permeation rate R2 Glove Latex Blue Hawk 0.07 87.38 10 ≤0.75 4.11 0.88 Glove Nitrile Blue Hawk 0.05 8.05 110 ≤0.75 0.14 0.29 Glove Vinyl HDX 0.07 46.05 20 ≤0.75 2.29 0.90 Coverall Polypropylene 3M 0.12 14.76 60 ≤0.75 0.43 0.54 Coverall Tyvek Pro Line 0.13 31.66 30 ≤0.75 0.67 0.37 Time-interval data and the predicted trend of permeation data indicated that all glove materials permeated prior to or at the first test interval of 0.75 min. Nitrile gloves demonstrated the most resistance to total isocyanates with the lowest average rate of permeation (0.14 ng NCO/cm2 min-1) of all glove materials and the lowest cumulative permeation at 20 min (8.05 ng NCO/cm2). The average nitrile glove thickness was 0.05 mm. Glove permeation results and the 95% confidence interval for the slope of each regression line are presented in Fig. 3. View Large SPF test-chemical selection A two-part slow-rise SPF insulation kit (Tiger Foam™ #TF-600SR) was chosen as the test chemical for this study (shown in Fig. 1). This low‐pressure formulation was similar in chemical composition to high‐pressure proportioner-applied mixtures, but was more suitable for use in the laboratory setting. Field testing using two-component high-pressure test mixtures was not practical due to the logistics of SPF application (truck, long tubing lines, etc.), rapid cure times of SPF mixtures, short sample collection intervals required for this study, and the experimental protocols necessary to prevent sample contamination during collection. The slow-rise formulation was found to be more suitable for this experiment due to its delayed foaming action, lower expansion ratio (1:1.25), and slower tack-free time than other formulations (i.e. 60–90 s). The properties exhibited by this formulation permitted the application of an expanding SPF mixture to the test cell under controlled laboratory conditions. Figure 1. View largeDownload slide Tiger Foam™ #TF-600SR two-part slow-rise SPF insulation test-chemical kit. A-side and B-side components are contained in separate pressurized cylinders. The chemical mixture is expelled through two-nylon hoses and ejected through a spray nozzle. Figure 1. View largeDownload slide Tiger Foam™ #TF-600SR two-part slow-rise SPF insulation test-chemical kit. A-side and B-side components are contained in separate pressurized cylinders. The chemical mixture is expelled through two-nylon hoses and ejected through a spray nozzle. The SPF insulation kit was comprised of two one-gallon pressurized cylinders. The A-side component consisted of a mixture of 30–60% pure MDI, 30–60% polymeric MDI, and 5–10% 1,1,1,2-Tetrafluoroethane. Commercial MDI is composed primarily of the 4,4′-MDI monomer (~98%) and impurities of the 2,2′-MDI and 2,4′-MDI isomers (Harari et al., 2016). The composition of polymeric MDI varies by manufacturer and generally contains 30–80% 4,4′-MDI and higher molecular weight MDI oligomers [trimer, tetramer, and pentamer each containing three, four, or five rings, respectively (MAK-Collection, 2012)]. The trimer is the most abundant of these oligomeric species as determined by liquid chromatography (LC). Additionally, the production processes for pure MDI commonly results in trace amounts of phenyl isocyanate that are normally not reported on the safety data sheet. Current production processes generally maintain the level of phenyl isocyanate below 0.005% (w/w) ratio (MAK-Collection, 2012). However, even at these trace levels, we have frequently quantified pure MDI in personal air and dermal samples collected during SPF application (unpublished data). The smaller size and higher vapor pressure of phenyl isocyanate increase its potential for airborne and dermal exposure (National Center for Biotechnology Information, 2018). Because phenyl isocyanate is a potent respiratory and skin sensitizer, it is important to document exposures to this chemical and the effectiveness of protective equipment. The B-side component consisted of a mixture of 30–60% proprietary polyol blend, 15–45% tris (1-chloro-2-propyl) phosphate (TCPP), 10–30% 1,1,1,2-Tetrafluoroethane, and 1–5% Pentamethyldiethylenetriamine. The kit functioned by expelling each the two components from their respective pressurized cylinder, through nylon hoses to the spray-gun nozzle where mixing occurred. Permeation test cell design A permeation test cell system designed by Ceballos et al. (2011) was used to evaluate the dermal protective clothing against polymerizing materials. The three-component system consisted of a bezel, cell body, and cell base used to house the test material and solid media (shown in Fig. 2). The cell bezel held the test material in place and created an airtight seal between the cell body and solid media to prevent inadvertent contamination of the sampling media. The cell base was inserted into the back of the test cell to hold the 25 mm diameter sampling media flush against the backside of test material, while allowing for its quick collection during the test process. The 2.54 cm diameter cell-bezel window allowed for exposure to a standardized area of garment material. Figure 2. View largeDownload slide Garment material was held tight between the cell bezel and cell body. The cell base was inserted into the body to hold the sampling media flush against the garment material. Cardboard containers were used to house the test cells and contain the expanding SPF. Figure 2. View largeDownload slide Garment material was held tight between the cell bezel and cell body. The cell base was inserted into the body to hold the sampling media flush against the garment material. Cardboard containers were used to house the test cells and contain the expanding SPF. Use of the original permeation panel system was not possible due to the rapid expansion of SPF insulation and the short collection intervals necessary for media recovery with this rapidly polymerizing material. The system utilized by Ceballos et al. (2011, 2013) was designed for use with slower reacting polymerizing materials such as HDI containing clear coats used in the automotive collision repair industry. Automotive clear coats are applied to a thickness of only a few millimeters and commonly have an air-cure-time of more than 16 h. Individual 20 × 20 × 10 cm cardboard containers were used to house each test cell and contain the expanding test chemical while allowing for the efficient collection of solid test media from the rear of the container. Test chemical application and interval determination SPF insulation application was conducted under laboratory conditions in a ventilated fume hood to minimize sample contamination during media collection and personal exposure. Test chemical cylinders were maintained at approximately 75°F prior to and during application to maintain the correct viscosity of both components as recommended by the manufacturer. A test spray was conducted prior to coating each batch of test cells to ensure proper mixing. A maximum of five test cells were arranged on a rack at a 45-degree angle and sprayed at 30-s intervals to provide sufficient time for media retrieval between each sample. Test cells were coated consistently to achieve an approximately 2.0 cm thick layer of uncured foam on each cell window. Within 15 s of the pre-established collection time, each test container was positioned at the hood opening with the cell base facing outward for media collection. Clean nitrile gloves and tweezers were used to collect solid media from each test cell. Solid media was placed directly into a desorbing solution vial of 0.5 mM 1-(9-anthracenylmethyl)piperazine (MAP) in acetonitrile and sealed to prevent contamination. Test intervals were established based upon the published 60–90 s tack-free time of the slow-rise SPF insulation kit. The test interval was defined as the time between SPF contact and sample collection. Initial test intervals were set at 0.75, 1.50, 2.25, 3.00, 5.00, 10.00, 15.00, and 20.00 min. Closer intervals were used immediately following application up to twice the tack-free time (i.e. 3.00 min) to best characterize the permeation profile during the period of peak unreacted MDI availability. Samples were then collected at 5-min intervals up to the final collection period at 20 min. Actual collection times for each sample were recorded to the nearest second. Each test container was air dried for a minimum of 24 h following the spray period to permit the full cure of each SPF sample. Test cell containers were disassembled and the thickness of the cured SPF insulation over each cell window was measured with a dial caliper. Test cells were then cleaned using N-Methyl-2-pyrrolidone (NMP) to remove any residual cured SPF insulation and sonicated for a minimum of 15 min in isopropyl alcohol. Each cell component was rinsed twice with ultrapure water following sonication and then mechanically washed with an aqueous cleaning agent. The effectiveness of the cell cleaning procedure was evaluated separately, and no contamination was introduced through the media retrieval process. Field blanks cells were prepared in an identical manner to test samples and placed in the laboratory hood with standard samples. Blanks were not directly sprayed with SPF insulation and were collected from their test cells 45 s after spraying adjacent samples to quantify unintended contamination by suspended aerosols during collection. Filter loading with the SPF test material was assessed by collecting three filters per test interval (24 samples in total) that were directly coated with the two-part SPF test chemical at each testing time interval. These ‘baseline’ samples were collected at their predetermined sample intervals and analyzed in a manner identical to the standard garment test cells Sample analysis and quality control Samples from 128 test cells and the 24 baseline samples (that assessed filter loading) were collected for analysis during this study. These 128 samples included an additional six latex glove samples that were obtained during the pilot phase of this study. One vinyl glove sample was punctured during collection and excluded from sample analysis, resulting in less than three data points at the 2.25-min test interval. Collection media consisted of 25-mm glass fiber filters impregnated with a MAP reagent. Samples were analyzed using a well-established ultra-high-performance liquid chromatography-UV absorbance-tandem mass spectrometry in the positive electrospray ionization mode (UHPLC-UV-ESI-MS/MS) method (Harari et al., 2016). This method provided for the quantitative analysis of phenyl isocyanate, the three isomers of MDI (i.e. 4,4′-MDI, 2,4′-MDI, 2,2′-MDI) and the MDI trimer. The limit of detection (LOD) for each species of isocyanate was as follows: 0.25 ng/ml for phenyl isocyanate, 0.20 ng/ml for 4,4′-MDI, 0.04 ng/ml for 2,4′-MDI and 2,2′-MDI, and 2.00 ng/ml for the MDI trimer for 10 µl injection and 1 ml final sample preparation volume. Normalized limit of detection data equated to the following values: 0.05 ng/cm2 for phenyl isocyanate, 0.04 ng/cm2 for 4,4′-MDI, 0.01 ng/cm2 for 2,4′-MDI and 2,2′-MDI, and 0.39 ng/cm2 for the MDI trimer. Eight field blanks were submitted for analysis to assess the level of contamination introduced through the test method. Laboratory blanks were analyzed with each sample preparation batch to ensure contamination was not introduced during the analytical process. Even with a highly sensitive analytical method, the blanks were very clean and no blank corrections to sample data were necessary. Data analysis Univariate analysis was conducted to summarize SPF densities and thicknesses data for comparison with the manufacturer’s specifications. The average thickness of foam over the cell window was 52.3 mm, with a coefficient of variation (CV) of 19.6% and a range of 26.4–81.6 mm. Simple linear regression analysis was conducted separately for each protective garment material to determine the average permeation rate, the standardized breakthrough time, and the 95% confidence intervals for the slope of each regression line. The independent variable [i.e. total isocyanate (NCO)] for each garment material was the sum of each quantifiable species (i.e. phenyl isocyanate, 4,4′-MDI, 2,4′-MDI, and 2,2′-MDI). MDI trimer was not detected in any of the permeation samples and was therefore excluded from the total isocyanate value and regression analysis. Non-detectable data points for each isocyanate species were substituted with the constant values of their respective LOD/√2. Simple substitution was determined to be appropriate due to the small sample size of each time interval (n ≤ 3) and the constant LODs for each species (Hewett and Ganser, 2007). One nitrile glove sample was determined to be statistically an outlier and was excluded from data analysis. Additional linear regression analysis was conducted to identify permeation characteristics by individual species for phenyl isocyanate, 4,4′-MDI, 2,4′-MDI, and 2,2′-MDI. Baseline filter loading data obtained by directly coating filter media with the SPF test chemical were highly variable (900 to 15,726 ng/cm2 total NCO). During the recovery process, the SPF adhered to the filter media, often leaving behind variable pieces of foam. This introduced considerable experimental error and was largely responsible for this variability. As such, these data have been presented as a worst-case range for potential direct exposure. Results Glove permeation data Permeation rates for each barrier material are presented in Table 1. The latex gloves displayed the greatest total isocyanate permeation rate (4.11 ng/cm2 min−1) and maximum cumulative permeation at 20 min (87.38 ng NCO cm−2) of all the gloves tested. The average latex glove thickness was 0.07 mm. Vinyl gloves displayed the second greatest permeation rate of (2.29 ng NCO/cm2 min−1) and maximum cumulative permeation at 20 min (46.05 ng NCO/cm2). The average vinyl glove thickness was 0.07 mm Glove permeation results and the 95% confidence intervals for the slope of each regression line are presented in Fig. 3. Figure 3. View largeDownload slide Total isocyanate permeation by glove material and simple linear regression results with 95% confidence intervals. Figure 3. View largeDownload slide Total isocyanate permeation by glove material and simple linear regression results with 95% confidence intervals. Coverall permeation data The Tyvek coverall displayed the greatest total isocyanate permeation rate (0.67 ng NCO/cm2 min−1) and maximum cumulative permeation at 20 min (31.66 ng NCO/cm2) of the two coveralls tested (Table 1). The average Tyvek coverall thickness was 0.13 mm. The polypropylene coverall displayed a lower permeation rate (0.43 ng NCO/cm2 min−1) and maximum cumulative permeation at 20 min (14.76 ng NCO/cm2). The average polypropylene coverall thickness was 0.12 mm. Time-interval data and the predicted trend of permeation indicated that both the Tyvek and polypropylene coveralls permeated at or prior to the 0.75-min time point. Coverall permeation results and the 95% confidence interval for the slope of each regression line are presented in Fig. 4. Figure 4. View largeDownload slide Total isocyanate permeation for a Tyvek and polypropylene coveralls and simple linear regression results with 95% confidence intervals. Figure 4. View largeDownload slide Total isocyanate permeation for a Tyvek and polypropylene coveralls and simple linear regression results with 95% confidence intervals. Isocyanate permeation by species Simple linear regression was performed for each detectable species of isocyanate using the latex glove permeation data (shown in Fig. 5), which demonstrated the greatest rate of permeation by garment material. Phenyl isocyanate displayed the greatest total isocyanate permeation rate (3.40 ng/cm2 min−1) and maximum cumulative permeation at 20 min (71.79 ng/cm2) of each species. 4,4′-MDI permeated at a rate of 0.55 ng/cm2 min−1 with a maximum cumulative permeation at 20 min of 15.23 ng/cm2. 2,4′-MDI permeated at a rate of 0.13 ng/cm2 min−1 with a maximum cumulative permeation at 20 min of 3.17 ng/cm2. 2,4′-MDI demonstrated the lowest permeation rate of 0.02 ng/cm2 min−1 and maximum cumulative permeation at 20 min of 0.53 ng/cm2. Figure 5. View largeDownload slide Isocyanate permeation by and simple linear regression results with 95% confidence intervals. Figure 5. View largeDownload slide Isocyanate permeation by and simple linear regression results with 95% confidence intervals. Discussion Limited guidance is currently provided by protective garment manufacturers about the effectiveness of their materials as a barrier to reacting SPF formulations. Specific data on the permeation rates and breakthrough times for the MDI and phenyl isocyanate components of these mixtures are not available. Additionally, SPF insulation product safety data sheets (SDS) commonly refer to protective garment types in general terms (e.g. disposable coveralls, protective gloves, etc.), as opposed to recommending specific barrier materials and thicknesses. The American Chemistry Council (2012) provided guidelines for appropriate personal protective equipment (PPE) during high-pressure applications that included chemical-resistant gloves, protective clothing, eye and face protection, and respiratory protection. Gloves made of nitrile, neoprene, butyl, or PVC were suggested to ‘generally’ provide ‘adequate’ protection against the isocyanate component of A-side materials. It was further suggested that applicators and helpers typically wear disposable coveralls to prevent contact of SPF with their skin and street clothing. Each of the protective garments evaluated in this study provided a considerable level of dermal protection against the MDI component of the SPF formulation, although the latex gloves demonstrated the least effective performance. The following conservative protection factors were calculated for each protective garment as the ratio of the minimum total isocyanate loading concentration (900 ng/cm2 from baseline filter loading data) to the maximum cumulative concentration of isocyanate permeation at 20 min: 0.07 mm latex gloves (10×), 0.05 mm nitrile gloves (110×), 0.07 mm vinyl glove (20×). 0.12 mm polypropylene coverall (60×), and 0.13 mm Tyvek coverall (30×). The 0.07 mm latex gloves provided the least effective level of protection for all protective garments, with a maximum cumulative isocyanate permeation at 20 min of 87.38 ng/cm2. Sample media directly exposed to the SPF insulation mixture resulted in a range of isocyanate exposure from 900 to 15,000 ng/cm2 total NCO, indicating that a significant level of dermal protection is still provided even when using latex barrier materials. Additionally, all garments demonstrated some level of detectable breakthrough. Data for all protective garments tested indicated that detectable breakthrough occurred prior to or at the initial time-point of 0.75 min, which is expected, because the raw materials are still in a liquid form and have the highest mobility/flux. As the foam polymerizes, it becomes increasingly more solid with active ingredients trapped inside the foam network. ASTM method F-739 method uses a standardized break through time to classify an unacceptable level of breakthrough. This standardized threshold does not account for the toxicity of a specific test chemical and therefore cannot infer a safe level of exposure. Study data indicated that following detectable isocyanate breakthrough, each type of protective garment tested prevented an average permeation rate at or above the standard breakthrough threshold of 0.1 μg/cm2 min−1 (100.0 ng/cm2 min−1). Latex gloves allowed for the greatest average rate of permeation at 4.11 ng/cm2 min−1, only 4.1% of the ASTM F-739 standardized breakthrough rate. However, the permeation rate did not directly correspond to the concentration of isocyanate species in the A-side component. Phenyl isocyanate, the species with the lowest molecular weight and a trace impurity in polymeric MDI, demonstrated the greatest rate of permeation despite the much higher percentage of 4,4′-MDI present in the A-side mixture. Makela et al. (2014) recommended a cumulative permeation end-point of 1.0 μg/cm2 (1000 ng/cm2) based upon the sensitizing capacity of MDI. The greatest level of cumulative permeation during this study occurred with latex gloves, resulting in a maximum cumulative permeation of 87.38 ng/cm2 total NCO. Based upon the average permeation rate of for this test material (4.11 ng/cm2 min−1), it is highly unlikely that this level of cumulative permeation would be reached prior to the completed reaction of the SPF mixture or the change-out of disposable gloves. Direct comparison of our results with other protective garment permeation studies was not possible due to the varying test methods, garment materials, and test chemical formulations. Mäkelä et al. (2014) heated pure MDI to a test temperature of 23°C to evaluate permeation through various protective garments. Compared to our 4,4′-MDI permeation data for vinyl glove permeation, Mäkelä et al. (2014) found a much higher rate of permeation. Although both garment materials consisted of 0.07 mm vinyl material and demonstrated breakthrough shortly after exposure, our average 4,4′-MDI permeation rate during the first 20 min of exposure was 0.68 ng/cm2/min with SPF, versus ~55 ng/cm2/min with heated pure MDI. Conversely, our latex glove data demonstrated a significantly higher rate of permeation for total isocyanates when compared to permeation data obtained by Ceballos et al. (2011). The automotive clear coats tested by Ceballos were based on slow curing aliphatic polyisocyanates systems. One explanation for these differences is the rapid cure rate of the aromatic isocyanates used in SPF formulations relative to the aliphatic isocyanates found in automotive coatings and the unreacted MDI formulation used by Mäkelä. The rapid foam curing process results in a high rate of isocyanate consumption and nearly immediate foam expansion and solidification that incorporates a cellular structure. These factors likely reduce the mobility/diffusivity of permeating chemicals through the foam. Our data demonstrated an average permeation rate of 4.1 ng/cm2/min for total isocyanates (phenyl isocyanate, 4,4′-MDI, 2,4′-MDI, and 2,2′-MDI) with an average glove thickness of 0.07mm and a total exposure time of 20 min. Automotive clearcoat data demonstrated an average permeation rate of 2.9 ng/cm2/min for total isocyanates (HDI and IPDI) with a glove thickness range of 0.10 mm to 0.13 mm mils and a total exposure time of 91 min. These differences further emphasize the critical importance of conducting permeation testing using relevant product formulations and realistic task stimulations. One significant advantage of our study design was the sensitivity of the UHPLC-UV-ESI-MS/MS method used for isocyanate analysis. This method provided for a significantly lower limit of isocyanate detection that was two orders of magnitude more sensitive than the HPLC-UV methods used in the other permeation studies (Ceballos et al., 2011; Mäkelä et al. 2014). Use of more sensitive methods for isocyanate quantification could provide a significant advantage when characterizing permeation behavior, particularly at time-points shortly after garment exposure. Despite the significant advantages of decreased detection limits, concerns must also be noted about the need for increased quality control to prevent unintentional contamination at these lower levels during sample preparation, recovery, and analysis. Conclusions and recommendations As previously mentioned, there is a considerable amount of variability in the B-side formulations between specific product-lines and manufacturers. This variability can affect foam density and cure rates, and ultimately isocyanate permeation. Therefore, these results and recommendations should be applied only to similar SPF insulation formulations. Permeation test data collected during this study indicated that each of the protective garments evaluated provided a considerable level of protection against the isocyanate component of the SPF insulation mixture (i.e. 10–110-fold reduction from the level of direct exposure). The 0.05 mm nitrile gloves were the most effective and are readily available from most commercial building suppliers. The nitrile gloves also provided the slowest rate of permeation and the lowest cumulative permeation of isocyanate. It should be noted that disposable gloves of this thickness are particularly susceptible to puncture or tearing in the construction work environment. Therefore, a more tear resistant disposable (≥0.13 mm) or non-disposable nitrile material is highly recommended for SPF applications. Tyvek and polypropylene protective coveralls were readily available from commercial building suppliers and provide better air-flow than non-permeable materials, allowing for greater comfort and reduced opportunity for thermal related stressors. Although both materials demonstrated almost immediate breakthrough, the average rate of permeation was low for both coverall materials (≤0.43 ng/cm2 min−1). Despite this low permeation rate, significant exposure potential (or cumulative dose) may still exist due to the large surface area of the coverall (i.e. whole body). Workers also commonly wear additional layers of clothing and undergarments beneath disposable coveralls, which will likely provide an additional dermal protection. Additionally, SPF application requires moderate to strenuous work activity, often in environments with limited airflow or environmental control. The use of respiratory protection is highly recommended during SPF insulation applications, but they may increase susceptibility to thermal stressors. As a result, breathable coverall materials may be a more acceptable choice than impermeable non-breathable selections. Our field observations and discussions with SPF applicators indicate that breathable coveralls are preferred in the field. Acknowledgements This research was supported by CPWR: The Center for Constructions Research and Training (CPWR) through NIOSH Cooperative Agreement Number U60-OH009762. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of CPWR or NIOSH. References American Chemistry Council . ( 2012 ) Spray Polyurethane Foam: Benefits for Passive Houses. Book Spray Polyurethane Foam . Available at https://polyurethane.americanchemistry.com/Spray-Foam-Coalition/CPI-PHIUS-Paper-Benefits-to-Foam.pdf ( accessed 25 May 2017 ). ASTM F-739 . ( 2012 ) Standard test method for permeation of liquids and gases through protective clothing materials under conditions of continuous contact . West Conshohocken, PA : ASTM International . Bello D , Herrick CA , Smith TJ et al. ( 2007 ) Skin exposure to isocyanates: reasons for concern . Environ Health Perspect ; 115 : 328 – 35 . Google Scholar CrossRef Search ADS PubMed Bello A , Xue Y , Woskie SR , Bello D . ( 2017 ) Assessment and control of exposures to reactive chemical resins in construction. I. SPF applications . Presented at the Plenary talk at the British Occupational Hygiene Society , Harrogate, UK , April 26 . BS EN 16523-1 . ( 2015 ) Determination of material resistance to permeation by chemicals. Permeation by liquid chemical under conditions of continuous contact . Brussels, Belgium : European committee for standardization, CEN . Ceballos D , Reeb-Whitaker C , Glazer P et al. ( 2014 ) Understanding factors that influence protective glove use among automotive spray painters . J Occup Environ Hyg ; 11 : 306 – 13 . Google Scholar CrossRef Search ADS PubMed Ceballos DM , Sasakura M , Reeb-Whitaker C et al. ( 2013 ) Testing of glove efficacy against sprayed isocyanate coatings utilizing a reciprocating permeation panel . Ann Occup Hyg ; 58 : 50 – 9 . Google Scholar PubMed Ceballos DM , Yost MG , Whittaker SG et al. ( 2011 ) Development of a permeation panel to test dermal protective clothing against sprayed coatings . Ann Occup Hyg ; 55 : 214 – 27 . Google Scholar PubMed Daftarian HS , Lushniak BD , Reh CM et al. ( 2002 ) Evaluation of self-reported skin problems among workers exposed to toluene diisocyanate (TDI) at a foam manufacturing facility . J Occup Environ Med ; 44 : 1197 – 202 . Google Scholar CrossRef Search ADS PubMed Engfeldt M , Isaksson M , Zimerson E et al. ( 2013 ) Several cases of work-related allergic contact dermatitis caused by isocyanates at a company manufacturing heat exchangers . Contact Dermatitis ; 68 : 175 – 80 . Google Scholar CrossRef Search ADS PubMed Harari H , Bello D , Woskie S et al. ( 2016 ) Development of an interception glove sampler for skin exposures to aromatic isocyanates . Ann Occup Hyg ; 60 : 1092 – 103 . Google Scholar CrossRef Search ADS PubMed Heederik D , Henneberger PK , Redlich CA ; ERS Task Force on the Management of Work-related Asthma . ( 2012 ) Primary prevention: exposure reduction, skin exposure and respiratory protection . Eur Respir Rev ; 21 : 112 – 24 . Google Scholar CrossRef Search ADS PubMed Hewett P , Ganser GH . ( 2007 ) A comparison of several methods for analyzing censored data . Ann Occup Hyg ; 51 : 611 – 32 . Google Scholar PubMed ISO 6529 . ( 2013 ) Protective clothing -- protection against chemicals -- determination of resistance of protective clothing materials to permeation by liquids and gases . Geneva, Switzerland : International Organization for Standardization . Jones K . ( 2016 ) Dermal uptake of diisocyanates – evidence from workplace exposures . Presented at the Occupational and Environmental Exposure of Skin to Chemicals Conference , Manchester, UK . Available at http://oeesc2016.org/files/2016/09/Session-9a-Jones-20_09_16.pdf ( accessed 25 May 2017 ). Jones K , Johnson PD , Baldwin PEJ et al. ( 2017 ) Exposure to diisocyanates and their corresponding diamines in seven different workplaces . Ann Work Expo Health ; 61 : 383 – 93 . Google Scholar CrossRef Search ADS PubMed Kieć-Świerczyńska M , Swierczyńska-Machura D , Chomiczewska-Skóra D et al. ( 2014 ) Occupational allergic and irritant contact dermatitis in workers exposed to polyurethane foam . Int J Occup Med Environ Health ; 27 : 196 – 205 . Google Scholar CrossRef Search ADS PubMed Larsen TH , Gregersen P , Jemec GB . ( 2001 ) Skin irritation and exposure to diisocyanates in orthopedic nurses working with soft casts . Am J Contact Dermat ; 12 : 211 – 4 . Google Scholar PubMed MAK-Collection . ( 2012 ) 4,4′-Methylene diphenyl isocyanate (MDI) and polymeric MDI (PMDI) [MAK Value Documentation, 1997]. the MAK-collection for Occupational Health and Safety : Wiley-VCH Verlag GmbH & Co. KGaA . Mäkelä EA , Henriks-Eckerman ML , Ylinen K et al. ( 2014 ) Permeation tests of glove and clothing materials against sensitizing chemicals using diphenylmethane diisocyanate as an example . Ann Occup Hyg ; 58 : 921 – 30 . Google Scholar PubMed Mickelsen RL , Roder MM , Berardinelli SP . ( 1986 ) Permeation of chemical protective clothing by three binary solvent mixtures . Am Ind Hyg Assoc J ; 47 : 236 – 40 . Google Scholar CrossRef Search ADS PubMed National Center for Biotechnology Information . ( 2018 ) PubChem Compound Database; CID=7672 . Available at https://pubchem.ncbi.nlm.nih.gov/compound/7672 ( accessed 30 January 2018 ) Randall D , Lee S . ( 2002 ) The polyurethanes book. [Everberg, Belgium] ; New York : [Huntsman Polyurethanes]; Distributed by John Wiley & Sons . ISBN 0470850418 9780470850411. Redlich CA . ( 2010 ) Skin exposure and asthma . Ann Am Thorac Soc ; 7 : 134 – 37 . Google Scholar CrossRef Search ADS Rom WN , Markowitz S . ( 2007 ) Environmental and occupational medicine . Philadelphia : Wolters Kluwer/Lippincott Williams & Wilkins . ISBN 0781762995 9780781762991. © The Author(s) 2018. Published by Oxford University Press on behalf of the British Occupational Hygiene Society. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
Annals of Work Exposures and Health (formerly Annals Of Occupational Hygiene) – Oxford University Press
Published: May 12, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera