A sample of each product was digested and screened for 31 elements using inductively coupled plasma-optical emission spectrometry (ICP-OES) using a modified NIOSH Method 7303 by Bureau Veritas (Novi, MI) (NIOSH, 2003). If chromium (Cr) was quantified in a sample, the resin was further analyzed to speciate Cr(VI) using ion chromatography with ultraviolet–visible detection in accordance with a modified OSHA Method ID 215 (version 2) (OSHA, 2006). Details of the analytical method are provided in the Supplemental File. Table S2 lists the elements and their respective limits of detection (LOD) and quantification (LOQ).

Samples of each product were quantitatively analyzed for 10 target irritants and/or sensitizers and qualitatively analyzed to identify other organic compounds present using gas chromatography-mass spectrometry (GC-MS) by Bureau Veritas. The target substances (acetone, isopropyl alcohol, n-hexane, benzene, methyl methacrylate, toluene, ethylbenzene, styrene, m,p-xylene, o-xylene) were selected based on their presence in resins (unpublished data) used in a study of VP printer emissions (Stefaniak et al., 2019). Details of the analysis are provided in the Supplemental File. Target compounds were quantified by comparing the mass spectral patterns and retention times to authentic standards. Other organic substances in the samples were qualitatively identified from the remaining chromatographic peaks (tentatively identified compounds [TICs]) by the analyst based on the mass spectral matches with mass spectral library software (Wiley Registry 8th Edition/NIST, 2008 Mass Spectral Library, Wiley-Blackwell) and a manual review. The semi-quantitative mass concentration of each TIC was estimated from the total area response of the identified compound compared with the total area response of the closest internal standard assuming a 1:1 response ratio. The reporting limit for TICs was 2.5 ng/μL.

For all elements and organic substances present, currently available literature was searched to determine whether each was a skin irritant and/or sensitizer. Product SDSs were reviewed for whether they listed individual ingredients, their Chemical Abstract Services (CAS) number, and product classification as a sensitizer and/or irritant in accordance with the European Commission regulation No. 1272/2008 on Classification, Labelling and Packaging of Substances and Mixtures (CLP Regulation). This regulation provides a decision logic by which manufacturers use set criteria and a weight-of-evidence approach to categorize a product based on types of hazards (e.g., skin sensitizer). Under the CLP regulation, products that are skin sensitizers can be classified into one of three categories (Category 1 = skin sensitizer, Category 1A = high frequency of skin sensitization in humans and/or a high potency in animals, and Category 1B = low to moderate frequency of skin sensitization in humans and/or a low potency in animals). Products that cause skin irritation (reversible damage) are assigned to CLP regulation Category 2 (skin irritation) or Category 3 (mild skin irritation).

The exact compositions of resin products were not declared on some product SDSs so there was no a priori expectation that the numbers and concentrations of elements and organic substances followed a normal distribution or had equal variances. As such, statistical analyses were performed only using the products in which the elements and organic substances were present. Results from characterization of the 39 products (herein designated 1, 2, …39) were compared by manufacturer (A – G), one- or two-part system (VP resins only), AM process type (MJ or VP), and color (clear, white, black, grey, yellow, green, and purple). The footnote in Table 1 lists the number of samples by factor (e.g., AM process type) and level (e.g., MJ or VP). Supplemental Table S1 lists the inventory of resin samples by product number and their properties. For each factor, the median number of sensitizers or irritants and the median mass concentration of sensitizers or irritants were compared in JMP (v13.5.1, SAS Institute Inc., Cary, NC) using the non-parametric Steel-Dwass test with multiple comparisons, which preserves the overall error rate (α = 0.05). Data were plotted using SigmaPlot (version 13.0, Systat Software, Inc., San Jose, CA).

Approximately 40% (15/39) of SDSs listed ingredients by name and included their CAS number. Thirty-three percent (13/39) of SDSs listed at least one ingredient as a generic class (e.g., photoinitiators) without a CAS number; three (products 18, 19, 20) listed all ingredients using generic classes without CAS numbers. Approximately 20% (8/39) of SDSs listed at least one ingredient as proprietary (trade secret) and did not provide its CAS number; of these eight products, two (numbers 16, 24) listed all ingredients as proprietary. Finally, 7% (3/39) of SDSs listed at least one ingredient as a generic class without a CAS number and at least one ingredient as proprietary without a CAS number (numbers 13, 22, 32). In accordance with the CLP regulation, manufacturers categorized 5% (2/39) of products as an irritant only, 18% (7/39) as a sensitizer only, 72% (28/39) as both an irritant and sensitizer, and 5% (2/39) as neither an irritant nor a sensitizer.

Of the 23 elements and 54 organic substances that were categorized as irritants (Fig. 1), product SDSs listed only three of these elements and nine of these organic substances.

Of the 22 elements and 23 organic substances that were categorized as skin sensitizers, product SDSs only listed four of these elements and six of these sensitizing organic substances.

Products 3 and 10 were neither classified as a skin irritant nor as a sensitizer on their SDSs. Product 3 contained 3525.98 mg/kg of skin irritants and/or sensitizers. Specifically, it contained five elements that were irritants and sensitizers (cadmium, calcium, cobalt, phosphorous, and zinc) at a summed concentration of 251.42 mg/kg, two elements that were irritants only (barium and magnesium) at a sum of 1.33 mg/kg, and one element that was a sensitizer only (titanium) at 0.23 mg/kg. This product also contained four irritant organic substances (acetone, toluene, 2,4,6-trimethylbenzoic acid, and 2,4-bis(1,1-dimethylethyl) phenol) at a summed concentration of 3273 mg/kg. Product 10 contained 4308.32 mg/kg of skin irritants and sensitizers (irritant and sensitizing elements arsenic, phosphorous, selenium, and tin at a summed concentration of 8.27 mg/kg, irritant only element barium at 0.054 mg/kg, and the irritant and sensitizing organic substance triethanolamine at 4300 mg/kg).

ICP and GC analyses revealed that all SDS were incomplete with regard to ingredients that were skin irritants and/or sensitizers. Overall, the total number of skin irritant and/or sensitizer ingredients that were not listed on SDSs ranged from 5 (product 10) to 32 (product 32); corresponding mass concentrations were 8.32 mg/kg and 3521.41 mg/kg, respectively. For elements, the mass concentration of skin irritant and/or sensitizer ingredients that were not listed on SDSs ranged from 2.32 mg/kg (product 4) to 978.73 mg/kg (product 22). For organic substances, the mass concentration of skin irritant and/or sensitizer ingredients that were not listed on SDSs ranged from zero (product 10) to 356,000 mg/kg (product 22).

The SDSs for products 16 and 24 listed all ingredients as proprietary. Product 16 contained aluminum, arsenic, calcium, molybdenum, phosphorous, potassium, selenium, tin, and zinc (irritants and sensitizers) at a summed concentration of 534.78 mg/kg and barium and magnesium (irritants only) at a summed concentration of 3.43 mg/kg. Additionally, this product contained tetraethylene glycol diacrylate at 3800 mg/kg (irritant and sensitizer) and toluene, acetic acid, 1-methoxy-2-propanol, and propylene glycol methyl ether acetate (irritants only) at a summed concentration of 2305 mg/kg. Product 24 contained arsenic, calcium, chromium, nickel, phosphorous, potassium, selenium, tin, and zinc (irritants and sensitizers) with a summed concentration of 212.21 mg/kg and barium (irritant only) at 0.07 mg/kg. Isopropyl alcohol, methyl methacrylate, methyl acrylate, 2-hydroxyethyl methacrylate, and butylated hydroxytoluene (irritants and sensitizers) were present at 62,638 mg/kg and acetone and methacrylic acid (irritants only) were identified at a summed concentration of 335 mg/kg.

From Fig. 3, no single element was quantified at a concentration that exceeded 1% by weight (10,000 mg/kg). As shown in Fig. 5, there were 12 irritant and/or sensitizing organic substances present at concentrations above 1% by wt. in one or more products in which the substance was detected (in mg/kg): 2-hydroxyethyl acrylate in 5/14 products (13,000 to 29,000), 2-hydroxyethyl methacrylate in 11/12 products (39,000 to 350,000), 2-hydroxypropyl methacrylate in 5/5 products (84,000 to 170,000), 3,3′-sulfonyldianiline in 1/1 product (39,000), bisphenol A in 1/3 products (97,000 mg/kg), bisphenol A diglycidyl ether in 2/2 products (24,000 to 110,000), glycerin in 1/1 product (75,000), hexamethylene diacrylate in 2/2 products (13,000 to 83,000), methyl methacrylate in 3/14 products (37,000 to 190,000), N,N-dimethylacrylamide in 2/2 products (37,000 to 360,000), tetramethoxymethane in 2/2 products (18,000 in both), and triethylene glycol dimethacrylate in 1/1 product (91,000).

Under the OSHA Hazard Communication Standard, relevant ingredients (excluding carcinogens and reproductive toxins) in a mixture are defined as those present at greater than or equal to 10,000 mg/kg or 1% by wt. (OSHA, 2012). Disclosure of an individual chemical ingredient in a mixture at a concentration of <1% by wt. on a product SDS depends upon whether that ingredient affects the hazard classification (OSHA, 2012). Thus, if an ingredient is present at < 1% by wt., it does not have to be listed on a product SDS, unless the person interpreting the standard considers that the ingredient will affect the product hazard classification (LeBouf et al., 2019). Products 3 and 10 were neither classified as a skin irritant nor as a skin sensitizer under the CLP regulation, despite both containing ingredients that possess these properties. No individual ingredient identified in product 3 exceeded 1% by wt., so none were required to be reported (nor were they voluntarily listed) on the product SDS, despite their potential skin hazard. The concentration of each irritant and sensitizer identified in product 10 was below 1% so none were required to be reported on the product SDS, though triethanolamine (0.43%), both an irritant and sensitizer, was voluntarily listed for this product.

The OSHA Hazard Communication Standard does not require that a chemical constituent be listed on a product SDS regardless of its percent concentration by weight if a mixture is designated by the manufacturer as proprietary (OSHA, 2012). Though the proprietary designation protects intellectual property, it can obscure chemical hazards posed by products used in workplaces. The SDSs for products 16 and 24 listed all ingredients as proprietary, but they contained many known irritants and/or sensitizers. The incompleteness of ingredient listings on product SDSs was not limited to these products. In fact, all product SDSs, regardless of the amount of ingredient information given, did not list from 5 to 32 irritants and/or sensitizers. In accordance with the CLP regulation, product 16 was categorized as a skin sensitizer and irritant and product 24 was categorized as a skin sensitizer. Among the remaining resin products, all but two were categorized as an irritant only, as a sensitizer only, or as both an irritant and sensitizer. Though these designations under the CLP regulation provided some warning to users that precautions should be taken to protect skin when working with these products, under-reporting of ingredients could inadvertently infer a “false-safe” impression to users (i.e., the hazard is not as severe as communicated on the SDS). Additionally, it is difficult to properly select types of skin protection (e.g., gloves, aprons, arm covers) and specific product materials (e.g., nitrile or vinyl, disposable or reusable) without knowledge of specific product ingredients.

Twelve organic substances were present at concentrations that exceeded the 1% reporting threshold of the Hazard Communication Standard (see Fig. 5). Three of these 12 substances (3,3′-sulfonyldianiline, glycerin, and N,N-dimethylacrylamide) were listed on their respective product SDSs. Hexamethylene diacrylate was identified in two products but only listed on the SDS for one product. 2-Hydroxyethyl acrylate exceeded 1% by wt. in five products; however, this substance was only listed on the SDS for one product. None of the remaining substances (2-hydroxyethyl methacrylate, 2-hydroxypropyl methacrylate, bisphenol A, bisphenol A diglycidyl ether, methyl methacrylate, tetramethoxymethane, and triethylene glycol dimethacrylate) were listed on SDSs. The reason why these skin hazards were not reported on product SDSs is unknown but highlights that product SDSs alone might not be sufficient for understanding health risks.

Seven products were categorized as sensitizers though they also contained irritants; it is possible that sensitization was considered by the manufacturer to be a more severe endpoint because avoidance or substitution are the best ways to prevent further exposure and elicitation of an allergic reaction (Martin et al., 2018) whereas irritation is considered to be temporary and reversible (Novak-Bilíc et al., 2018; Bains et al., 2019). Products 12 and 18 were categorized as irritants, but they contained sensitizers. The rationale for categorizing these products as irritants but not sensitizers is unclear. Product 12 contained the irritants and sensitizers 2-hydroxyethyl methacrylate at 5.6% and triethylene glycol dimethacrylate at 9.1% and product 18 contained the irritant and sensitizer 2-hydroxyethyl acrylate at 1.7%.

It is important to have both transparency and completeness of ingredients on SDSs of resin products to permit accurate hazard assessments. If an employee develops dermatitis, a complete list of ingredients and concentrations on product SDSs can help facilitate the identification of the offending agent. For example, patch testing with standard “series” of common causative agents such as fragrances, metals, etc. is often used to inform the diagnosis of ACD in persons with skin disorders (Bourke et al., 2009); however, these series might not include a specific allergen present in a workplace. Rømyhr et al. followed industrial painters and conducted patch testing using standard allergen series and a special study series based on known or suspected sensitizers present in their workplaces. Several workers had positive patch tests to chemicals only in the special study series, which indicated that it was useful to test with known and suspected skin sensitizers specific to patients’ workplaces (Rømyhr et al., 2006). Creytens et al. reported two cases of ACD among workers exposed to light curing resins in an AM facility and noted that they could not fully evaluate potential allergens in these products because some SDSs did not specify all ingredients, did not include concentrations of ingredients, or listed only a generic substance class (e.g., diacrylate esters) without a CAS number (Creytens et al., 2017). Note that under the Hazard Communication Standard, a health care professional can request disclosure of specific chemicals and amounts of proprietary ingredients for exposure assessment and medical treatment purposes (OSHA, 2012).

Firstly, we characterized a convenience sample of 39 commercial products, though it is unknown how many more MJ and VP resins are currently on the market, which could limit the generalizability of the results. Secondly, analysis of organic substances was done by methanol extraction followed by GC screening; however, it is possible that extractions were incomplete (recoveries not determined) and/or more organic irritants and sensitizers not amenable to this analytical technique were present in the products, so values presented herein should be considered underestimates. Volatilization or aerosolization of resin constituents could present an inhalation hazard such as irritation, sensitization, or cancer (Väisänen et al., 2019, 2022). Though the current study focused on the dermal exposure pathway, risk assessments should not neglect potential adverse health effects for the inhalation pathway. Additionally, 79 of the organic substances were TICs, so their identities were based on manual matching with spectral libraries by the analyst and concentration values were semi-quantitative based on the total area response of the identified compound compared with the total area response of the closest internal standard assuming a 1:1 response ratio. None-the-less, presentation of their concentration provided valuable information for future studies to quantify these substances using authentic standards. Thirdly, we categorized elements and organic substances as irritants and/or sensitizers. This simplified approach could overestimate risk because it gives equal probability to all elements and substances, though some are more common irritants and sensitizers than others. On the other hand, this approach could underestimate risk because it assumes all elements and organic substances have equal potency. Additionally, it ignores the possibility of effects from co-exposure to irritants and sensitizers, such as a ‘danger’ model, whereby an antigenic signal in the presence of a non-specific danger signal (e.g., irritancy) is needed to activate the immune system (Nagtegaal et al., 2012) and the possibility of multiple sensitivity from cross-reactivity or multiple sensitization (Thyssen and Menne, 2010) or permissive effects from other non-irritant and non-allergen components.