TY - JOUR
AU1 - Ehnes,, Colin
AU2 - Genz,, Manfred
AU3 - Duwenhorst,, Jörn
AU4 - Krasnow,, Jurij
AU5 - Bleeke,, Jan
AU6 - Schwarz,, Katharina
AU7 - Koch,, Wolfgang
AU8 - Schuchardt,, Sven
AB - Abstract The aerosol release during the professional application of two different isocyanate based two component spray systems was identified and the physicochemical properties of the released airborne aerosols were characterized. For this purpose, aerosol release fractions were measured using a mass balance method described by Schwarz and Koch. Besides the release of total aerosol mass special emphasis was directed to the content of free monomeric MDI (4,4′- and 2,4′-diphenylmethane diisocyanate) in three particle size fractions relevant for inhalation uptake: inhalable, thoracic, and respirable size fraction. Two products were investigated: a two component PUR (polyurethane) spray foam (Elastopor) and a polyurea spray coating (Elastocoat). The mass fraction of the applied products released with the overspray as inhalable aerosol is 6.3 × 10−4 (Elastopor) and 4.0 × 10−4 (Elastocoat). Of the released total overspray aerosol 75 or 80% were in the thoracic size range, and 26 or 47% in the respirable regime for the PUR spray foam or the polyurea spray coating, respectively. At the time point of release the content of monomeric MDI in the aerosol corresponds to the composition of the bulk product. However, analysis of air samples indicates that <1% of the spray foam aerosol mass release fraction is attributed to free monomeric 4,4′- and 2,4′-MDI. For the Spray Coating the monomeric MDI fraction is <0.1%. Higher oligomers of MDI and prereacted oligomeric reaction products make up a few percent of the aerosol. This results in a total fraction of 0.0023% (spray foam) and 0.00015% (spray coating), respectively, of the sprayed monomeric MDI that is transferred into an inhalable aged aerosol. This data demonstrates, that during professional spraying only a small fraction of the total applied mass is released as airborne aerosol. The potential distribution of the theoretically inhalable aerosol in the respiratory tract and a low residual monomer content is described, significantly contributing to a refined safety assessment of the spray applications at the workplaces. aerosol curing, aerosol release fraction, diphenylmethane diisocyanate, inhalable, respirable, safety assessment, thoracic Introduction Di- and poly-isocyanates are organic compounds which contain two or more common functional isocyanate-groups (–N=C=O). In the production of technically versatile polyurethane plastics (PUR), these highly reactive NCO groups are for example reacted with a polyol, resulting in a polymeric network of urethane, urea and other groups derived of those. Some of these manifold polymer uses of isocyanates are spray applications. Important spray application technologies of the aromatic diisocyanate MDI (diphenylmethane diisocyanate) are for example rigid foam insulations of buildings (PUR spray foam), as well as polyurea coatings of surfaces to become permanently protected (polyurea spray coating). In such two component applications, the polyhydroxylated or polyaminated organic component A is sprayed together with the isocyanate component B in a defined stochiometric ratio. A typical spray apparatus consists of a mixing unit into which the two components are pumped under high pressure at controlled mass flux and an air assisted dispersion nozzle. The technical specifications of the spray apparatus are designed to bring a maximum of the mixture onto the surface or into the hollow space. Consistent with this technical setup the reactions of component A with component B begin with the mixing in the spray apparatus and the generated aerosol droplets represent reacting units. A small fraction of the sprayed product will not deposit on the surface and escapes as overspray into the air, demanding a particularly thorough safety assessment for the operator. Due to the very low vapor pressure of the reactants and the increasing viscosity due to the polymerization reaction, a significant vapor generation during the process does usually not occur. Exposure assessment as well as evaluation of the intrinsic inhalation toxicity of the spray application requires a detailed characterization of this overspray aerosol fraction in the breathing zone of the worker. Next to the total mass of airborne particles, information on the release of aerosols in health-related particle size ranges and the content of residual reactive monomer under the conditions of use of the product is desirable. Since the deposition of aerosol droplets in the respiratory tract depends on the aerodynamic diameter, conventions for health-related size ranges of airborne particles were defined in the European Standard CEN 481 (Comité Européen de Normalisation [CEN], 1992). Such conventions are defined for the inhalable, thoracic, and respirable fractions, and approximate the mass fraction which may penetrate to the respective regions of the respiratory tract of a healthy adult under average conditions. Relative to total airborne particles, the particle size having 50% penetration for the thoracic and respirable fractions are 10 and 4.0 μm, respectively. Particle size selective sampling according to these conventions yields in a better relationship between measured concentration and type and potency of a potential health hazard in the respiratory tract. The possible air concentration of the product as well as the concentration of free MDI depends on the overspray generation rate and the local conditions at the workplace such as room volume and ventilation conditions and, regarding the free MDI content, the curing within the aerosol droplets. There are numerous publications on MDI concentrations at workplaces in general and MDI originating from two component spray processes in particular. Lesage et al. (2007) report on MDI concentrations associated with spray foaming in residential areas. Concentrations as well as particle size distribution related to the overspray at various distances from the source are measured. Puscasue et al. (2015) compare the impinger technique with a combination of denuder and filtration for MDI analysis in aerosols of spray foam processes. They found that the filter method underestimates the MDI concentration significantly compared with the impinger method. From the perspective of a generalized safety assessment for a specific spray application, it is important to characterize the aerosol source strength (overspray generation rate) in the first place rather than absolute exposure concentrations under specific conditions of use, since this is directly connected to the technology and chemistry of the application process. Air concentrations can then be derived from these source data by using indoor or outdoor dispersion modeling, their details depending on the specific environmental conditions under which the technology is used in practice. This study describes the experimental characterization of the physico-chemical properties of the overspray of two exemplary spray applications of MDI, that is a spray foaming and a spray coating. A mass balance method described in Schwarz and Koch (2017) is applied for this purpose. The three health-related particle size regimes defined in the CEN standard 481 (CEN, 1992) are considered in this context, that is the process specific aerosol release into the respirable, the thoracic, and the inhalable size range. Data are presented for the total and the fractionated aerosol masses of the reacting two component systems, as well as for the unreacted free MDI in the aged aerosol. The effect of aerosol curing on the potential exposure concentration of bystanders to free MDI is discussed for an indoor and outdoor scenario. Materials and methods Investigated products and spray technology Two products were investigated: a two component polyurethane spray foam (Product A, Elastopor® 1622/201) and a polyurea spray coating (Product B, Elastocoat® C 6335/101). For the foaming application, component A consists of a mixture of polyols and additives (including blowing agents), whereas component B is the so called polymeric diphenylmethane diisocyanate (PMDI CAS# 9016-87-9). The content of monomeric MDI (mMDI) is 21.12% and the content of oligomeric MDI (oMDI) is 30.1% of the total applied product mass. The ratio of the 2,4′- and the 4,4′ MDI isomers is 1/10. The oligomeric MDI consists of three-ring and higher-ring oligomers. The heated reactive liquids are mixed under high pressure and released for application using a spray foam machine, type GAMA Evolution VR (GAMA, BARCELONA, Spain), equipped with Master II Gun. The system is operated at a mass release rate of 200 g s−1. For the spray coating, the two components are a preparation based on aminoterminated polyols, catalyst, and additives (component A) and the B component being a modified MDI mixture containing an equimolar ratio of monomeric 2,4′- and 4,4′-MDI resulting in a total content of monomeric MDI of 20.28% and a content of prereacted MDI of 6.5% of the total applied product mass. These components are mixed by high pressure and are sprayed (with a mass output rate of 150 g s−1) with a spraying machine type Graco Reactor EXP2 dosing system (Graco Distribution, Maasmechelen, Belgium). The mixing head of type Fusion, AW2828 is used in combination with a spray nozzle of 0.3 mm diameter. In the following, the term oligomeric MDI (oMDI) is applied to describe the higher ring oligomers of MDI in the original isocyanate product, the oligomers in the original product from prereaction of MDI with a polyol, as well as oligomeric products formed during the polymerization reaction of the components postspray. General procedure for determination of release fractions The professional application of sprayed isocyanate products is accompanied by the formation of a small quantity of airborne aerosols. Droplets not deposited on the target surface during the spray process will become airborne. Depending on the droplet size distribution, the aerosol may contain a certain mass fraction that can be inhaled and transported into various regions of the respiratory tract that is the respirable (r), the thoracic (t), and the inhalable (i) fraction of the airborne material. The potential of a spray process to generate health relevant airborne particles can be quantified by release fractions, Rr,t,i ⁠, defined as the mass, mr,t,i ⁠, of the product generated as aerosol in the three size classes normalized to the total mass, M ⁠, of the spray product released: Rr,t,i=mr,t,iM (1) If specific chemical compounds are of concern such as monomeric MDI, the release fractions can be specified also with respect to the release of these substances. A mass balance method described in detail in Schwarz and Koch (2017) is employed to determine the release fractions experimentally. In brief, this method uses a well-mixed spray booth, with volume V, where a short spray bolus of mass, M, of product is sprayed onto a designated surface. Under the well stirred conditions the overspray is spatially homogenized leading to concentrations cr,t,i of the overspray inside the control volume immediately after the spraying. The associated aerosol mass is calculated by multiplying the concentration with the control chamber volume: mr,t,i=cr,t,i⋅V. (2) Experimentally, the concentrations cr,t,i are derived from time average values, c¯r,t,i ⁠, by aerosol sampling over a sufficiently long measuring time, T, to collect enough material for chemical and gravimetric analysis. Wall losses of material and material removal from the spray booth by air exchange can be taken into account by measuring the mean residence times, τr,t,i ⁠, and using the formula valid for well stirred mixing conditions (Schwarz and Koch, 2017) relating the concentration, cr,t,i ⁠, right after the end of the spray process with the measured time average concentration, c¯r,t,i cr,t,i=c¯r,t,iT/τr,t,i1−e−T/τr,t,i (3) This relationship is obtained by time averaging of the exponential decrease of the aerosol concentration Cr,t,i(t)=cr,t,i⋅Exp(−t/τr,t,i) over the sampling time, T. Aerosol measurements The aerosol inside the spray booth is characterized using the aerosol monitoring system RESPICON® (Helmut Hund GmbH, Wetzlar). The RESPICON operates at a total nominal sampling flow rate of 3.1 l min−1. It consists of three stages and combines filter sampling for measurement of the average mass concentration (total mass) in the three size fractions, c¯r,t,i ⁠, and aerosol photometry (constant angle light scattering sensor) for time resolved aerosol recording for determination of the decay times, τr,t,i (Koch et al., 1999). The aerosol mass collected on the internal quartz fiber filters of the RESPICON is determined by gravimetric analysis. The mass fraction of free MDI, φr,t,i ⁠, on the filters is measured using HPLC-analysis by derivatization to its corresponding urea compounds. For this purpose, the aerosol atmosphere is sampled on impregnated quartz fiber filters placed in the RESPICONs. The derivatization agent used was 1-(2-methoxylphenyl)piperazine. The chemical analysis is carried out following the ISO Standard 16702. To back-up the findings obtained with the RESPICON aerosol samplers, additional sampling is carried out for the spray coating using a combination of wash bottle and impregnated filter, as described in the ISO Standard 16702. Samples are taken in parallel to the RESPICON samplers and are analyzed for monomeric MDI. Test chamber The release tests are carried out in spray chambers equipped as shown in Fig. 1. Four RESPICON sampling instruments are used in parallel: two equipped with unimpregnated (R1, R2) and two with impregnated quartz fiber filters (G1, G2). The instruments R1 and R2 are suited to also measure the time pattern of the concentrations. A mixing fan in a corner of the chamber is intended to implement well mixed conditions. The product is sprayed onto a chalk board located flat on the floor in the center of the chamber. The equipment for the ISO sampling method is placed in between the RESPICON sampling devices. The heights of aerosol inlets are 1.50 m in all cases. Figure 1. View largeDownload slide Principles of the sampling set-up for spray foam analysis. Positions of measuring instruments R1 and R2 (equipped with unimpregnated filters), G1 and G2 (equipped with impregnated filters) and ISO sampler as well as the ventilator. The indicated floor area was treated with the respective spray application. The sampling set-up for the spray coating analysis was equivalent, with room dimensions of 3.2 m (height), 4.1 m (width), and 6.5 m (length). Figure 1. View largeDownload slide Principles of the sampling set-up for spray foam analysis. Positions of measuring instruments R1 and R2 (equipped with unimpregnated filters), G1 and G2 (equipped with impregnated filters) and ISO sampler as well as the ventilator. The indicated floor area was treated with the respective spray application. The sampling set-up for the spray coating analysis was equivalent, with room dimensions of 3.2 m (height), 4.1 m (width), and 6.5 m (length). Conduct of the tests The two products were investigated in two different spray chambers. In order to determine the worst-case conditions for application specific aerosol release fractions, the LEV (Local Exhaustion Ventilation) of the chambers were shut down. However, chamber design-related differences resulted in differences in chamber volume (53 and 80 m3 for the spray foam and spray elastomer, respectively) and air exchange rates in the spray chambers (13 and 2 h−1 for the spray foam and spray elastomer, respectively). This air exchange was caused by an existing negative pressure difference between the chamber and the outside environment allowing for air still flowing into the chamber from the outside through leaks. The process parameters of the dosing machines were adjusted to the intended values and a functioning check on spray performance was carried out before the release test. The chamber was ventilated with room air with a ventilator and spray actions of duration between 10 and 60 s (depending on the sprayed product) were carried out to determine the release fractions. During spraying the operator stood in front of the treated area and the spray nozzle was held approximately perpendicular to the surface. The duration of aerosol sampling was between 20 and 30 min. Thus, the aerosol sampling time was large compared to the aerosol release time in all tests. At the end of the measurement the impregnated filters were removed from the instruments G1 and G2 and transferred into flasks containing the derivatization agent. The solution was chemically analyzed for isocyanate compounds in the laboratory. The pre-weighed quartz fiber filters from the instruments R1 and R2 were stored in sealed Petry dishes and post-weighed in the laboratory to determine the deposited aerosol mass. Two release experiments were carried out for each product. The release fractions were calculated from the measured average concentrations and the time constants using equations (1)–(3). Average values were obtained from the two RESPICONs and the two test runs. Error bars were calculated as the maximum difference between Test 1 and Test 2 for each product. Comparison with the ISO method was done for the spray coating system only. Results Representative examples of the temporal concentration profiles after short-term spraying of 30 s duration are shown in Fig. 2 for a PUR spray foam and a polyurea spray coating product. In each case, the spray was released ~10 min after the aerosol monitoring instruments were turned on. For both products the steep increase in concentration during spraying is followed by an exponential decrease after termination of the spray-process. This pattern confirms fast mixing by the applied fan and the applicability of equation (3) for data analysis as outlined above. The sampling time after spray release was 20 min. The variations in the decay times reflect the different air exchange rates established in the two spray chambers used for the testing. The aerosol concentration in the inhalable size range decreases faster than that of the respirable range, which can be explained by faster sedimentation of higher weight aerosol droplets. It has to be emphasized, that the absolute aerosol concentrations shown here do not simulate realistic workplace situations but worst case conditions, since the spray chambers were deliberately operated with very low air ventilation rate to collect a sufficient aerosol mass for chemical analysis of the ingredients. Figure 2. View largeDownload slide Aerosol mass concentration in health relevant size regimes inside the spray chamber after the release of a spray bolus during 30 s. Results of R1 (see Figure 1) with a total released mass of 4 kg foaming or 4.5 kg coating product, respectively. Top: Foaming product measured in chamber 1 with aerosol residence time τ = 4.6 min; Bottom: coating product measured in chamber 2 with τ = 30 min. In all cases, the concentration decrease with time is exponential. Figure 2. View largeDownload slide Aerosol mass concentration in health relevant size regimes inside the spray chamber after the release of a spray bolus during 30 s. Results of R1 (see Figure 1) with a total released mass of 4 kg foaming or 4.5 kg coating product, respectively. Top: Foaming product measured in chamber 1 with aerosol residence time τ = 4.6 min; Bottom: coating product measured in chamber 2 with τ = 30 min. In all cases, the concentration decrease with time is exponential. The release rates calculated from the chamber experiments are displayed in Table 1 and Fig. 3. Table 1. Release fractions of the total aerosol, and free monomeric and oligomeric MDI in the overspray at the application of PUR spray foam or polyurea spray coating. Numbers for the total aerosol, and free and oligomeric MDI are indicated as mean values of n measurements (± error). Total (n = 4) 2,4′-MDI (n = 2) 4,4′-MDI (n = 2) oMDI (n = 2) PUR spray foam Mean value r 1.7E-04 3.7E-07 1.3E-06 3.7E-06 t 4.7E-04 8.3E-07 3.1E-06 9.1E-06 i 6.3E-04 1.0E-06 3.8E-06 1.2E-05 Error r 4.1E-05 1.6E-07 5.4E-07 1.2E-06 t 1.0E-04 2.5E-07 6.2E-07 2.9E-06 i 1.3E-04 3.8E-07 1.0E-06 4.3E-06 Polyurea spray coating Mean value r 1.9E-04 3.2E-08 9.1E-09 1.0E-05 t 3.2E-04 1.5E-07 2.6E-08 1.4E-05 i 4.0E-04 2.7E-07 4.1E-08 1.5E-05 Error r 2.2E-05 1.5E-10 4.4E-11 1.8E-06 t 5.2E-05 8.8E-09 2.2E-09 1.5E-06 i 5.8E-05 1.3E-08 2.2E-09 1.4E-06 Total (n = 4) 2,4′-MDI (n = 2) 4,4′-MDI (n = 2) oMDI (n = 2) PUR spray foam Mean value r 1.7E-04 3.7E-07 1.3E-06 3.7E-06 t 4.7E-04 8.3E-07 3.1E-06 9.1E-06 i 6.3E-04 1.0E-06 3.8E-06 1.2E-05 Error r 4.1E-05 1.6E-07 5.4E-07 1.2E-06 t 1.0E-04 2.5E-07 6.2E-07 2.9E-06 i 1.3E-04 3.8E-07 1.0E-06 4.3E-06 Polyurea spray coating Mean value r 1.9E-04 3.2E-08 9.1E-09 1.0E-05 t 3.2E-04 1.5E-07 2.6E-08 1.4E-05 i 4.0E-04 2.7E-07 4.1E-08 1.5E-05 Error r 2.2E-05 1.5E-10 4.4E-11 1.8E-06 t 5.2E-05 8.8E-09 2.2E-09 1.5E-06 i 5.8E-05 1.3E-08 2.2E-09 1.4E-06 r, respirable; t, thoracic; i, inhalable. View Large Table 1. Release fractions of the total aerosol, and free monomeric and oligomeric MDI in the overspray at the application of PUR spray foam or polyurea spray coating. Numbers for the total aerosol, and free and oligomeric MDI are indicated as mean values of n measurements (± error). Total (n = 4) 2,4′-MDI (n = 2) 4,4′-MDI (n = 2) oMDI (n = 2) PUR spray foam Mean value r 1.7E-04 3.7E-07 1.3E-06 3.7E-06 t 4.7E-04 8.3E-07 3.1E-06 9.1E-06 i 6.3E-04 1.0E-06 3.8E-06 1.2E-05 Error r 4.1E-05 1.6E-07 5.4E-07 1.2E-06 t 1.0E-04 2.5E-07 6.2E-07 2.9E-06 i 1.3E-04 3.8E-07 1.0E-06 4.3E-06 Polyurea spray coating Mean value r 1.9E-04 3.2E-08 9.1E-09 1.0E-05 t 3.2E-04 1.5E-07 2.6E-08 1.4E-05 i 4.0E-04 2.7E-07 4.1E-08 1.5E-05 Error r 2.2E-05 1.5E-10 4.4E-11 1.8E-06 t 5.2E-05 8.8E-09 2.2E-09 1.5E-06 i 5.8E-05 1.3E-08 2.2E-09 1.4E-06 Total (n = 4) 2,4′-MDI (n = 2) 4,4′-MDI (n = 2) oMDI (n = 2) PUR spray foam Mean value r 1.7E-04 3.7E-07 1.3E-06 3.7E-06 t 4.7E-04 8.3E-07 3.1E-06 9.1E-06 i 6.3E-04 1.0E-06 3.8E-06 1.2E-05 Error r 4.1E-05 1.6E-07 5.4E-07 1.2E-06 t 1.0E-04 2.5E-07 6.2E-07 2.9E-06 i 1.3E-04 3.8E-07 1.0E-06 4.3E-06 Polyurea spray coating Mean value r 1.9E-04 3.2E-08 9.1E-09 1.0E-05 t 3.2E-04 1.5E-07 2.6E-08 1.4E-05 i 4.0E-04 2.7E-07 4.1E-08 1.5E-05 Error r 2.2E-05 1.5E-10 4.4E-11 1.8E-06 t 5.2E-05 8.8E-09 2.2E-09 1.5E-06 i 5.8E-05 1.3E-08 2.2E-09 1.4E-06 r, respirable; t, thoracic; i, inhalable. View Large Figure 3. View largeDownload slide Release fractions in the three health relevant size regimes for the total aerosol mass and the monomeric and oligomeric MDI components in the reaction product. Left: PUR spray foam, right: polyurea spray coating. Figure 3. View largeDownload slide Release fractions in the three health relevant size regimes for the total aerosol mass and the monomeric and oligomeric MDI components in the reaction product. Left: PUR spray foam, right: polyurea spray coating. For both spray applications the release fraction for the total inhalable aerosol was well below 0.1% of the total amount used in the initial reaction mixture. This total inhalable aerosol consists of 75 or 80% of the mass in the thoracic and 26 or 47% aerosol mass in the respirable size range for the spray foam and spray coating application, respectively. Accordingly, <1000 mg inhalable, <500 mg thoracic, and <200 mg respirable aerosol is transferred into the air from 1 kg sprayed material, the so-called overspray. For example, 0.047 and 0.032% of the sprayed material is transferred into the thoracic aerosol for the spray foam and spray coating application, respectively. The impregnated filters were analyzed for monomeric and oligomeric MDI. Representative results for the air concentrations of 4,4′-MDI averaged over the sampling time are shown in Fig. 4. The 4,4′-MDI concentration strongly depends on the particle size, with roughly 30% of the free 4,4′-MDI in the respirable size range. Since the vapor phase MDI should be collected in the same stage as the respirable aerosol, this means, that under worst case conditions a maximum of 30% of the measured MDI could be theoretically attributed to the vapor phase. In conclusion, the 4,4′-MDI concentration analyzed on the filters is not dominated by the vapor phase, although the measured concentrations are below the saturated vapor concentration of pure MDI which can be calculated with the ideal gas law and the vapor pressure of 4.4′-MDI. In addition, the 4,4′-MDI evaporation from the aerosol droplets is thermodynamically hindered in the reacting mixture. Figure 4. View largeDownload slide Time averaged airborne concentration of 4,4′-MDI as measured in the four tests carried out in the control chambers. Figure 4. View largeDownload slide Time averaged airborne concentration of 4,4′-MDI as measured in the four tests carried out in the control chambers. Regarding the mass fractions, φr,t,i ⁠, of free monomeric MDI and oligomeric MDI, the chemical composition of the free aerosol has drastically changed compared with the primary product mix (Table 2). Only a minor fraction (<1%) of free monomeric MDI remains in the aerosol released from spraying the PUR spray foam product which initially contains approx. φ0 = 21% of free monomeric MDI. For the polyurea spray coating, the free monomeric MDI content in the aerosol is even an order of magnitude lower, with a similar initial content of approximately φ0 = 20% in the starting material. Table 2. MDI fraction, φMDI in the total released aerosol and proportion of unreacted MDI of the PUR spray foam and polyurea spray coating starting material. Numbers for 2,4′-MDI, 4,4′-MDI, and the oligomeric MDI content are indicated as mean values of n = 2 measurements (± error). PUR spray foam Polyurea spray coating 2,4′-MDI 4,4′-MDI oMDI 2,4′-MDI 4,4′-MDI oMDI MDI fraction in matured aerosol r 2.1E-03 (±4.7E-04) 7.2E-03 (±1.5E-03) 2.2E-02 (±1.9E-03) 1.7E-04 (±2.0E-05) 5.0E-05 (±5.7E-06) 5.8E-02 (±1.7E-02) t 1.7E-03 (±1.4E-04) 6.6E-03 (±1.5E-04) 1.9E-02 (±2.0E-03) 4.8E-04 (±5.1E-05) 8.4E-05 (±7.1E-06) 4.5E-02 (±1.2E-02) i 1.6E-03 (±2.7E-04) 5.9E-03 (±4.0E-04) 1.8E-02 (±3.0E-03) 6.9E-04 (±6.8E-05) 1.0E-04 (±9.9E-06) 4.0E-02 (±9.4E-03) Unreacted fraction of initial MDI content r 1.1E-1 (±4.9E-2) 4.0E-2 (±1.7E-2) 7.2E-2 (±2.4E-2) 1.7E-3 (±0.0E-2) 5.0E-4 (±0.0E-2) 7.69E-2 (±1.3E-2) t 8.7E-2 (±2.6E-2) 3.5E-2 (±6.0E-3) 6.2E-2 (±2.0E-2) 4.7E-3 (±2.0E-4) 8.0E-4 (±1.0E-4) 5.83E-2 (±6.0E-3) i 8.0E-2 (±3.0E-2) 3.1E-2 (±9.0E-3) 6.1E-2 (±2.2E-2) 6.7E-3 (±3.0E-4) 1.0E-3 (±3.0E-4) 5.26E-2 (±5.0E-3) PUR spray foam Polyurea spray coating 2,4′-MDI 4,4′-MDI oMDI 2,4′-MDI 4,4′-MDI oMDI MDI fraction in matured aerosol r 2.1E-03 (±4.7E-04) 7.2E-03 (±1.5E-03) 2.2E-02 (±1.9E-03) 1.7E-04 (±2.0E-05) 5.0E-05 (±5.7E-06) 5.8E-02 (±1.7E-02) t 1.7E-03 (±1.4E-04) 6.6E-03 (±1.5E-04) 1.9E-02 (±2.0E-03) 4.8E-04 (±5.1E-05) 8.4E-05 (±7.1E-06) 4.5E-02 (±1.2E-02) i 1.6E-03 (±2.7E-04) 5.9E-03 (±4.0E-04) 1.8E-02 (±3.0E-03) 6.9E-04 (±6.8E-05) 1.0E-04 (±9.9E-06) 4.0E-02 (±9.4E-03) Unreacted fraction of initial MDI content r 1.1E-1 (±4.9E-2) 4.0E-2 (±1.7E-2) 7.2E-2 (±2.4E-2) 1.7E-3 (±0.0E-2) 5.0E-4 (±0.0E-2) 7.69E-2 (±1.3E-2) t 8.7E-2 (±2.6E-2) 3.5E-2 (±6.0E-3) 6.2E-2 (±2.0E-2) 4.7E-3 (±2.0E-4) 8.0E-4 (±1.0E-4) 5.83E-2 (±6.0E-3) i 8.0E-2 (±3.0E-2) 3.1E-2 (±9.0E-3) 6.1E-2 (±2.2E-2) 6.7E-3 (±3.0E-4) 1.0E-3 (±3.0E-4) 5.26E-2 (±5.0E-3) r, respirable; t, thoratic; i, inhalable. View Large Table 2. MDI fraction, φMDI in the total released aerosol and proportion of unreacted MDI of the PUR spray foam and polyurea spray coating starting material. Numbers for 2,4′-MDI, 4,4′-MDI, and the oligomeric MDI content are indicated as mean values of n = 2 measurements (± error). PUR spray foam Polyurea spray coating 2,4′-MDI 4,4′-MDI oMDI 2,4′-MDI 4,4′-MDI oMDI MDI fraction in matured aerosol r 2.1E-03 (±4.7E-04) 7.2E-03 (±1.5E-03) 2.2E-02 (±1.9E-03) 1.7E-04 (±2.0E-05) 5.0E-05 (±5.7E-06) 5.8E-02 (±1.7E-02) t 1.7E-03 (±1.4E-04) 6.6E-03 (±1.5E-04) 1.9E-02 (±2.0E-03) 4.8E-04 (±5.1E-05) 8.4E-05 (±7.1E-06) 4.5E-02 (±1.2E-02) i 1.6E-03 (±2.7E-04) 5.9E-03 (±4.0E-04) 1.8E-02 (±3.0E-03) 6.9E-04 (±6.8E-05) 1.0E-04 (±9.9E-06) 4.0E-02 (±9.4E-03) Unreacted fraction of initial MDI content r 1.1E-1 (±4.9E-2) 4.0E-2 (±1.7E-2) 7.2E-2 (±2.4E-2) 1.7E-3 (±0.0E-2) 5.0E-4 (±0.0E-2) 7.69E-2 (±1.3E-2) t 8.7E-2 (±2.6E-2) 3.5E-2 (±6.0E-3) 6.2E-2 (±2.0E-2) 4.7E-3 (±2.0E-4) 8.0E-4 (±1.0E-4) 5.83E-2 (±6.0E-3) i 8.0E-2 (±3.0E-2) 3.1E-2 (±9.0E-3) 6.1E-2 (±2.2E-2) 6.7E-3 (±3.0E-4) 1.0E-3 (±3.0E-4) 5.26E-2 (±5.0E-3) PUR spray foam Polyurea spray coating 2,4′-MDI 4,4′-MDI oMDI 2,4′-MDI 4,4′-MDI oMDI MDI fraction in matured aerosol r 2.1E-03 (±4.7E-04) 7.2E-03 (±1.5E-03) 2.2E-02 (±1.9E-03) 1.7E-04 (±2.0E-05) 5.0E-05 (±5.7E-06) 5.8E-02 (±1.7E-02) t 1.7E-03 (±1.4E-04) 6.6E-03 (±1.5E-04) 1.9E-02 (±2.0E-03) 4.8E-04 (±5.1E-05) 8.4E-05 (±7.1E-06) 4.5E-02 (±1.2E-02) i 1.6E-03 (±2.7E-04) 5.9E-03 (±4.0E-04) 1.8E-02 (±3.0E-03) 6.9E-04 (±6.8E-05) 1.0E-04 (±9.9E-06) 4.0E-02 (±9.4E-03) Unreacted fraction of initial MDI content r 1.1E-1 (±4.9E-2) 4.0E-2 (±1.7E-2) 7.2E-2 (±2.4E-2) 1.7E-3 (±0.0E-2) 5.0E-4 (±0.0E-2) 7.69E-2 (±1.3E-2) t 8.7E-2 (±2.6E-2) 3.5E-2 (±6.0E-3) 6.2E-2 (±2.0E-2) 4.7E-3 (±2.0E-4) 8.0E-4 (±1.0E-4) 5.83E-2 (±6.0E-3) i 8.0E-2 (±3.0E-2) 3.1E-2 (±9.0E-3) 6.1E-2 (±2.2E-2) 6.7E-3 (±3.0E-4) 1.0E-3 (±3.0E-4) 5.26E-2 (±5.0E-3) r, respirable; t, thoratic; i, inhalable. View Large In the same way, the content of free oligomeric MDI is reduced from approximately φ0 = 31% (higher ring-MDI-oligomers) in the starting material to 2% (higher ring MDI-oligomers or oligomeric reaction products of MDI with a polyol) in the aerosol particles of the aged PUR spray foam application. In the polyurea spray coating application however, the content of oligomeric MDI in the starting material (prereacted MDI) and in the released free aerosol (prereacted MDI and/or oligomers formed postspray) remains approximately constant at 5–6% of the total mass. Table 2 also indicates the proportion from the monomeric MDI initially released into the respective fraction which was not reacted. This illustrates a change in the isomer composition from 1:10 in the starting material to ~1:3.7 in the aerosol for the spray foam, and from 1:1 to 1:0.2–0.3 in the aerosol fraction for the spray coating, respectively. This shows, that in both applications the reaction of the 4,4′-MDI with a hydroxy- or amino-terminated polyol proceeds more efficiently than that of the 2,4′-MDI. Furthermore, there seems to be a particle size dependence of the free monomeric MDI content in the aerosol. This can be deduced from Fig. 5 where the content of the measured isocyanate compounds are given for the respirable (r), the trachea-bronchial (t.b.), and the extra-thoracic (e.t.) size regimes representing roughly the aerosol fraction with particle diameters in the range <5, 5–10, and >10 µm. The latter two are obtained by subtracting the respirable fraction (r) from the thoracic fraction (t), or by subtracting the thoracic fraction (t) from the inhalable fraction (i), respectively. The free monomeric MDI content seems to decrease with increasing particle size for the foam spray whereas it increases for the coating spray. For example, the 2,4′-MDI content of the coating spray aerosol is 0.017% in the finest size regime (respirable), while it is a factor of 10 larger for the coarse particles (0.15% for particles > 10 µm). In other words, for the spray coating application more unreacted free monomeric MDI remains in the large droplets of the inhalable fraction compared with the small droplets of the respirable fraction. A contrary trend is recognizable for the spray foam application. Figure 5. View largeDownload slide Amount of isocyanate compound in extra-thoracic (e.t.; >10 µm), trachea-bronchial (t.b.; 5–10 µm) and respirable (r.; <5 µm) size regime. Left: foam spray, right: coating spray. e.t., extra thoracal; t.b., tracheobronchial; r., respirable (see text for explanation). Figure 5. View largeDownload slide Amount of isocyanate compound in extra-thoracic (e.t.; >10 µm), trachea-bronchial (t.b.; 5–10 µm) and respirable (r.; <5 µm) size regime. Left: foam spray, right: coating spray. e.t., extra thoracal; t.b., tracheobronchial; r., respirable (see text for explanation). In experiments with the spray coating (Tests 3 and 4), samples were taken using the ISO method in parallel to the RESPICON samplers and were analyzed for 4,4′-MDI. Since the ISO method does not provide size resolution, the time averaged concentration obtained from the ISO sampling method was compared with the average concentration of the inhalable fraction measured with the RESPICON. The ISO samples result in 29% (Test 3) and 13% (Test 4), higher values of the 4,4′-MDI concentration than the RESPICON samples (Table 3). In the ISO method, most of the material (especially the large particle fraction) is collected in the wash bottles where the particles are thoroughly mixed with the derivatization agent allowing the free MDI to react immediately. This could mean that unreacted MDI still present in the aerosol at the time of sampling has derivatized faster compared to when collected on impregnated filters such as in the RESPICON. This leads to a higher MDI concentration under the respective conditions. Unreacted MDI will be present in the aerosol during a time period of the order of the reaction time for polymerization after the release process. By the immediate derivatization in the wash bottle, the polymerization reaction is quenched whereas in the aerosol it is ongoing until the reactants have been used up, which would result in a lower content of free MDI in the filter sample as compared with the wash bottle sample. Therefore, the result of the comparison of the two sampling methodologies for the spray coating application may not necessarily be considered as representative for the spray foam application. For example, if different states of curing of the aerosols were present at the time of sampling, resulting in different residual monomeric MDI concentrations. Due to potentially less efficient derivatization on impregnated filters, higher residual MDI concentrations at the time of sampling may result in a more pronounced underestimation of the actual concentration in comparison to wash bottles. For spray foam, Lesage et al. (2007) indicate that filter sampling approaches underestimate MDI levels in some situations, with the effectiveness of stabilization being a plausible explanation. Table 3. Time averaged concentration of 4,4′-MDI (in µg m−3) after sampling polyurea spray coating aerosols with the ISO 16702 and RESPICON method. To compare sampling efficiencies, the total concentration analyzed with the ISO method was compared with the inhalable concentration of the RESPICON. Test no. ISO sampling method Respicon Wash bottle Filter Total r t i 3 2.8 1.1 3.9 1.2 2.2 3.1 4 3.5 1.2 4.7 1.5 3.0 4.2 Test no. ISO sampling method Respicon Wash bottle Filter Total r t i 3 2.8 1.1 3.9 1.2 2.2 3.1 4 3.5 1.2 4.7 1.5 3.0 4.2 r, respirable; t, thoracic; i, inhalable. View Large Table 3. Time averaged concentration of 4,4′-MDI (in µg m−3) after sampling polyurea spray coating aerosols with the ISO 16702 and RESPICON method. To compare sampling efficiencies, the total concentration analyzed with the ISO method was compared with the inhalable concentration of the RESPICON. Test no. ISO sampling method Respicon Wash bottle Filter Total r t i 3 2.8 1.1 3.9 1.2 2.2 3.1 4 3.5 1.2 4.7 1.5 3.0 4.2 Test no. ISO sampling method Respicon Wash bottle Filter Total r t i 3 2.8 1.1 3.9 1.2 2.2 3.1 4 3.5 1.2 4.7 1.5 3.0 4.2 r, respirable; t, thoracic; i, inhalable. View Large Discussion This study pursued the goal to characterize the aerosol release associated with the application process of stoichiometric mixing and spraying of polyurethane spray foam and spray coating systems under realistic conditions. For that purpose, the mass releases into the three health-related particle size fractions relevant for inhalation uptake were quantified, and the free residual monomeric MDI content was determined in the related aerosol droplets. The result is a source term for the aerosol, Sr,t,i ⁠, and the MDI, Sr,t,imMDI ⁠, respectively, determined by the total product mass flow, M˙p ⁠, the measured release fractions, Rr,t,i ⁠, and the free MDI content, φr,t,i ⁠. The source term can be calculated as Sr,t,i=Rr,t,i⋅M˙p and Sr,t,imMDI=Rr,t,i⋅φr,t,i⋅M˙p ⁠. The quantitative data generated in this study help to describe the physicochemical processes in the airborne aerosols generated during the application of isocyanate spray applications. Therefore, this data enables an improved evaluation of the inhalation toxicity of realistic aerosols and exposure modeling at workplaces. In this context, an important finding of the study is that only a minor part of <0.1% of the applied mass is released as airborne aerosol during the application of the spray products. This inhalable fraction is defined as all particles that can enter the respiratory tract during normal breathing. The thoracic size fraction of particles, which can pass the upper respiratory tract and can reach the trachea and bronchi, was determined to be ~75–80% of the total airborne aerosol. The respirable fraction, which can reach the peripheral airways, that is bronchiole and the alveolar region was determined to be 26–50% of the total airborne aerosol. In conclusion, it can be demonstrated, that only a very minor share of the applied material for the spray applications is released as airborne aerosol. From the measured approximate size distribution of the released aerosol it is possible to estimate the deposition fractions in the different regions of the respiratory tract. This allows a comparison with highly respirable exposure atmospheres generated in animal toxicity assays performed in accordance to the OECD Guidance Document No. 39 (OECD, 2009). Therefore, a heterogeneous deposition of the realistic aerosol within the respiratory tract can be assumed. Pulmonary irritation is the underlying mechanism for the acute inhalation toxicity of MDI, and this effect is partly determined by the site of deposition of the inhaled droplets in the respiratory tract. In addition, the low vapor pressure of MDI excludes a significant evaporation in the respiratory tract. With these substance properties, the data on the particle size distribution of the aerosol generated in this study, can be applied to translate the acute toxic potency from experimental inhalation toxicity data to realistic aerosols encountered at the specific workplaces (Pauluhn, 2008). The data furthermore describes a significant change in the chemical composition from the primary product mix to the free aerosol. The strong decrease in monomeric MDI indicates that the aerosol droplets represent reactive units in which the reaction of the isocyanate component with the polyol component takes place, and that these droplets are in an advanced state of curing. The reaction time scale determines the curing state of the droplets until they may reach the breathing zone of the worker. The influence of reaction time scale on the aerosol composition is discussed below. Differences in residual MDI content can qualitatively be explained by the differences in reaction kinetics, that is a lower reactivity of a functional isocyanate group towards a hydroxylated polyol in the PUR spray foam application (<1% remaining monomeric MDI in aerosol) compared with an aminated polyol in the polyurea spray coating application (<0.1% remaining monomeric MDI in aerosol). Higher component temperatures due to the higher viscosity and higher reaction enthalpies in the spray coating compared with the spray foam application are likely to play an additional role for this observation. The apparently slower reaction kinetic of the spray foam application in comparison to the spray coating application results in a less advanced state of curing of the aerosols at the time of sampling, with a higher residual monomeric MDI concentration. If the effectiveness of MDI stabilization is compromised on the filters, a higher residual MDI concentration may result in a more pronounced underestimation of the actual MDI concentration. Accordingly, the findings of the MDI analyses in the release fractions of the airborne aerosols provide valuable mechanistic information, however, the quantitative information for the spray foam application should be interpreted with care as the filter method may significantly underestimate the actual MDI concentration, as for example described by Puscasue et al. (2015). In addition, the data illustrate a clear change in the monomeric MDI isomer ratios between the starting material and the aerosol droplets. A plausible explanation for this observation is the reduced reactivity of a NCO-group in 2-position of 2,4′-MDI compared with a NCO group in 4′-position. These differences in relative reactivities can be explained by steric hindrance of the NCO group in 2-position and were experimentally confirmed by application technological data for a reaction with a polyetherol (data not shown). Based on the existing data it is difficult to identify a distinct explanation for the observed particle size dependence of the free monomeric MDI content in the aerosol droplets. A correlation to the differences in reaction kinetics between an aminoterminated and a hydroxyterminated reaction system may be plausible, resulting in a lower process enthalpy in the spray foam compared with the spray coating. This may have an impact due to an improved heat dissipation over a larger surface of small particles. The time-dependent decrease in density and increase in particle size of spray foam aerosols due to the use of a blowing agent may have an additional impact. However, further kinetic studies are required to better describe this observation. The results of this study can also be applied to estimate the exposure of workers and bystanders in varying indoor and outdoor application scenarios. Realistic exposure estimates can, for example, be modeled, with the identified application specific source strengths. Approximately 500 mg inhalable aerosol is released with every kilogram product sprayed onto the surface; for the respirable particles it is about 200 mg kg−1. The process rate of the spray apparatus is about 200 g s−1 yielding a source strength of released aerosol mass of 40 mg s−1 for the respirable aerosol and 100 mg s−1 for the inhalable fraction. This source strength together with the exposure scenario such as, among others, room size and ventilation rate for indoor application and wind velocity and distance from the source for outdoor scenarios determines the airborne mass concentration of the aerosol. Regarding exposure to free MDI a conservative view would be to use the original MDI mass fraction of the product mixture around φr,t,i0≈0.2 in the source term Sr,t,imMDI ⁠. However, aerosol curing leads to a significant depletion of the free MDI due to the polymerization reaction taking place in the aerosol droplets released in the spray application process. For the indoor scenario, the aerosol dispersion is mainly by diffusion due to turbulence generated mechanically by the spray process and thermally by the temperature gradients above the foam layer. The air turbulence is parameterized by the turbulent diffusivity, K. Following diffusion theory, the maturation time for the aerosol to diffuse from the source to a bystander at distance, x, can be estimated as: τm=x2/(6⋅K) ⁠. Outdoors, aerosol is transported by convection with the wind and the elapsed time between source and receptor is simply given by τm=x/u ⁠, where u is the wind velocity. Assuming first-order kinetics for the polymerization process with the polymerization time, τp, the free MDI content, φr,t,i(t) as function of time is given as φr,t,i(t)=φr,t,i0⋅exp⁡(−t/τp) yielding φr,t,i(x)=φr,t,i0⋅exp⁡(−x2/(6Kτp)) and φr,t,i(x)=φr,t,i0⋅exp⁡(−x/(uτp)) for the reduction of free MDI as a function of source-receptor distance, x, in the indoor and the outdoor scenario, respectively. Figure 6 shows examples for realistic dispersion parameters under the conservative assumption of a polymerization reaction time of 30 s. For generic exposure scenarios, such simulations result in realistic exposure estimates for free residual monomeric MDI contained in partly cured aerosol particles, depending on the distance to the source. Further research should be aimed at getting more quantitative information on the polymerization reaction times scale in the airborne particles which may differ from the value estimated from the bulk reaction during foam formation. Figure 6. View largeDownload slide Estimated free MDI content (in %) in the aerosol released from a spray process as function of source-receptor distance (in m) for different indoor (left) and outdoor (right) scenarios. The initial content is 20%. Figure 6. View largeDownload slide Estimated free MDI content (in %) in the aerosol released from a spray process as function of source-receptor distance (in m) for different indoor (left) and outdoor (right) scenarios. The initial content is 20%. Declaration Funding of this project was provided by BASF Polyurethanes GmbH. At the time the project was funded, Colin Ehnes was an employee of BASF SE as a producer of MDI. Manfred Genz, Jörn Duwenhorst, Jurij Krasnow, and Jan Bleeke were employees of BASF Polyurethanes GmbH. The authors declare no other conflict of interest relating to the material presented in this article. References CEN . ( 1992 ) Workplace atmospheres: size fraction definitions for measurement of airborne particles in the workpklace . CEN Standard EN. Berlin: Beuth Verlag. 481. Google Preview WorldCat COPAC Koch W , Dunkhorst W , Lödding H . ( 1999 ) Design and performance of a new personal aerosol monitor . Aerosol Sci Technol ; 31 : 231 – 46 . Google Scholar Crossref Search ADS WorldCat Lesage J , Stanley J , Karoly WJ et al. ( 2007 ) Airborne methylene diphenyl diisocyanate (MDI) concentrations associated with the application of polyurethane spray foam in residential construction . J Occup Environ Hyg ; 4 : 145 – 55 . Google Scholar Crossref Search ADS PubMed WorldCat OECD Environment, Health and Safety Publications . ( 2009 ) Series on testing and assessment, no. 39, guidance document on acute inhalation toxicity testing, ENV/JM/MONO(2009)28 . Available at https://one.oecd.org/document/ENV/JM/MONO(2009)28/en/pdf. Google Preview WorldCat COPAC Pauluhn J . ( 2008 ) Inhalation toxicology: methodological and regulatory challenges . Exp Toxicol Pathol ; 60 : 111 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat Puscasu S , Aubin S , Cloutier Y et al. ( 2015 ) Comparison between the ASSET EZ4 NCO and impinger sampling devices for aerosol sampling of 4,4′-methylene diphenyl diisocyanate in spray foam application . Ann Occup Hyg ; 59 : 872 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat Schwarz K , Koch W . ( 2017 ) Thoracic and respirable aerosol fractions of spray products containing non-volatile compounds . J Occup Environ Hyg ; 14 : 831 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2019. 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)
TI - Characterization of Aerosol Release during Spraying of Isocyanate Products
JF - Annals of Work Exposures and Health (formerly Annals Of Occupational Hygiene)
DO - 10.1093/annweh/wxz045
DA - 2019-08-07
UR - https://www.deepdyve.com/lp/oxford-university-press/characterization-of-aerosol-release-during-spraying-of-isocyanate-jfPzaYwDYV
SP - 773
VL - 63
IS - 7
DP - DeepDyve
ER -