Analytical survey of tattoo inks—A chemical and legal perspective with focus on sensitizing substances

Tattoo inks have been reported to elicit allergic contact dermatitis.


| INTRODUCTION
Tattooing is done by injecting colored inks under/into the dermis layer of the skin to leave a permanent design. The inks consist of pigments and auxiliary compounds, such as solvents, binders, and pH regulators. [1][2][3] Tattoo art has been an increasing fashion phenomenon globally, and already involves 12% of Europeans and up to 30% of United States' citizens, in particular in young generations. [3][4][5][6] In parallel, tattoo removal is becoming more frequent. Tattoo inks might contain sensitizing/hazardous substances that may cause adverse health effects linked to the application and removal of tattoos, and a certain proportion of the ink could be transported within the body via the blood. 5,7 These effects include acute allergy directly after tattooing or delayed hypersensitivity after long-term exposure to the chemicals in the inks. 4,[6][7][8][9][10] As an example, about 70% of 3411 tattooed individuals reported skin problems immediately or a few weeks after tattooing. 10 Skin cancer risks from tattooing have been neither proved nor excluded. 4,11 Sensitizing substances might induce allergic contact dermatitis (type IV hypersensitivity), an inflammatory skin reaction caused by direct contact with these substances. 12 A patch test is a clinical diagnostic standard method for type IV hypersensitivity, aiming to identify an allergen in an allergic patient by applying the diluted substance under occlusion on the skin under standardized conditions. 12 It can be used to detect specific allergies in a patient with an allergic reaction to a tattoo. A patch test study on 90 patients with a selection of tattoo ink stock products revealed only nine individuals with positive reactions, mainly associated with red inks. 13 This suggests that many culprit allergens in tattoo inks are neither not yet known nor included in baseline and specialized tattoo ink patch test series. 13 The pigments used in tattoo inks are produced mainly for largescale applications in construction or cosmetics industries, not specifically for use in injecting into the skin, and they generally show low purity (70%-90%). 3,4,14 Metals are often used in different substances as dyes or pigments, either in inorganic pigments, such as metal oxides, or in metal-organic complexes. Tattoo inks have been confirmed to contain harmful impurities that are known or suspected to cause adverse effects in humans, such as hexavalent chromium (Cr VI ) in Cr oxides; nickel (Ni), copper (Cu), and cobalt (Co) in iron (Fe) oxides; aromatic amines in azo-colorants; and polycyclic aromatic hydrocarbons in carbon black. 4,[14][15][16] Considering the increasing popularity of tattooing and the possible presence of harmful substances in the products used for tattoos, there is a need for rules to limit the risks posed by unsuitable tattoo  4 The Swedish Medical Product Agency has published a regulation on tattoo inks in 2012, covering product directory, labeling, product information, and importation and usage of tattoo inks. 18 A report of the Joint Research Center (JRC) of the European Commission (EC), compiled by experts from research and risk assessment, aimed to set a legislative framework to protect consumer safety. 4 Based on the evidence provided by the JRC of the presence of tattoo inks on the European market not complying with the limits set by the CoE, the European Chemicals Agency (ECHA) submitted in 2019 a restriction proposal on substances used in tattoo inks and PMU to the Committees for Risk Assessment (RAC) and Socio-economic Analysis (SEAC) for their evaluation. 14 Finally, a legal requirement for substances in tattoo inks or PMU at the EU-wide level was published on December 14, 2020, and will come into force on January 5, 2022 due to a transition period. 19 Several relatively recent studies have reported the occurrence and potential risks posed by hazardous chemicals in tattoo inks. Bocca et al. 15 found that Cr VI in tattoo inks could be a possible cause of dermal adverse reactions, and 90% of the investigated inks contained Cr VI above the maximum allowed level (0.2 μg/g), but no information appeared on the label. An investigation on a set of tattoo inks with various shades 16 showed that the concentrations of Cr, Cu, and lead (Pb) were above (5-to 500-fold), the maximum allowed levels regulated in ResAP(2008)1. In another published market study in Italy, several toxic elements, such as cadmium (Cd), antimony (Sb), Pb, vanadium (V), and manganese (Mn), exceeded 1 μg/g in some cases. 20 In the same study, the sensitizing metals Cr, Ni, and Co were above the safe limit in 62.5%, 16.1%, and 1.8% of the studied 56 tattoo inks, respectively. The presence of the prohibited pigments and the prevailing pigments behind chronic allergic reactions (Pigments Red 22, Red 210, and Red 170) were revealed in several studies on tattoo inks, by different analytical techniques such as Raman spectroscopy and mass spectrometry. 9,16,21 According to a previous report compiled by the Swedish Chemicals Agency in 2010, only 5 of 31 analyzed tattoo inks in various shades were free of hazardous substances, and the others contained aromatic amines (classified as carcinogenic, mutagenic, and allergenic) and different metals at levels above the recommended limits. In a Swiss study (2009), 41% of the samples had nonpermitted chemical contents. 5 This study aimed at assessing potential hazards with tattoo inks, and how those are related to concomitant content of substances/impurities, to labeling, to color, and to brand. This study increases knowledge about which substances are relevant to include in a patch test when testing a patient with an allergic reaction to a tattoo. In this study, a total of 73 tattoo inks known to be used in Sweden and many other countries, were either collected from a store and a tattoo studio in Sweden or ordered online. These samples were investigated on their contents of metals and pigments, and whether their labeling fulfilled legal requirements.
Matrix-assisted laser desorption/ionization time of flight tandem mass spectrometry (MALDI-ToF-MS n ) was used for identification of organic pigments and inductively coupled plasma mass spectroscopy (ICPMS) for the quantification of metal present in the tattoo inks.  Table S1, Appendix S1. These investigated tattoo inks were manufactured by a range of top brands, including World Famous Tattoo Ink (abbreviated as "WF"), Intenze Advanced Tattoo Ink ("In"), Radiant Colour ("RC"), Fusion Tattoo Ink ("Fu"), Eternal Ink ("Et"), Solid Ink ("So"), Dynamic ("Dy"), Tang Dragon Tattoo ("TD"), and Kuro Sumi Colours ("KS"). Samples 1-56 were bought between March 2019 and January 2020, and samples 57-73 were old or previously opened samples kindly provided by a tattoo studio. The latter samples were excluded from some evaluations, as their selling date might be prior to some legal requirements, and their previous opening could have caused evaporation, resulting in higher concentrations of substances. The label information on each tattoo ink bottle was inspected to investigate compliance with the requirements set in ResAP(2008)1 1 published by CoE. Correct label reading was confirmed by two persons.   To identify which pigments were present in each sample using MS, a corresponding peak for the pigment had to be found in the mass spectrum and the isotope pattern. For MS/MS identification, a probability score higher than 70% had to be obtained from the database search.

| Chemicals
The target plate was washed with deionized water and liquid detergent and wiped gently with Kimwipes (Kimberly-Clark, Irving, TX, USA) until all visible pigments were removed, and washed extensively with deionized water to remove the detergent. The target plate was wiped with isopropanol before sonicating the plate in ultra-pure water for 15 minutes. Thereafter, isopropanol was used to wipe the target plate twice, and the plate was sonicated in isopropanol for 15 minutes.
The target plate was then placed in an oven at 250 C for 3 hours.

| ICPMS
Quantitative analysis of both total (through microwave assisted digestion with concentrated HNO 3 ) and water-soluble (extracted in 0.9% NaCl, see below) trace metals in tattoo ink samples was conducted with ICPMS.
For total trace metals, the tattoo ink samples were digested using an Ultraclave IV microwave digestion system (MLS GmbH, Leutkirch, Germany). A total of 0.1 g of the tattoo ink was weighed into 10 mL quartz vessels, and 4.5 mL of sub-boiled concentrated HNO 3 was added into the vessel before closing it. The vessels were then placed in the autoclave with a pressure of 4 Â 10 6 Pa of argon (grade 5.0, Messer, Austria). More details on the autoclave settings can be found in Table S2 in Appendix S1. After the samples were cooled down, the solutions were transferred into 50 mL tubes, and HCl and ultrapure water were added into the tubes to obtain a final concentration of 9% HNO 3 and 1% HCl in the solutions. The blank samples containing PBS were also diluted with ultrapure water (9% HNO 3 and 1% HCl) at a ratio of 1 + 9. White precipitates were observed in many samples, suggesting nonsoluble (under these conditions) titanium dioxide.
For the extraction of the water-soluble metals from the tattoo inks, an aliquot of the tattoo inks (0.5 g) was mixed with 10 mL 0.9% NaCl and extracted in a shaking water bath at 37 C for 12 hours.
After extraction and cooling down, the samples were centrifuged at 30 000 g. Afterwards, the supernatant was diluted 10 times with a solution of 1% HNO 3 and 0.1% HCl. A precipitate was observed after the addition of acids, so the samples were centrifuged again. Some samples had to be additionally filtered because the precipitates could not be removed.

| Statistical analysis
where LQ is the lower quartile-the data value located halfway between the median and the smallest data value; IQD is the interquartile distance-the distance between the upper and lower quartiles (UQ -LQ); and UQ is the upper quartile-the data value located halfway between the median and the largest data value.

| Inspection on label information
According to the instructions and requirements for labeling tattoos regulated in ResAP (2008) Table S1, Appendix S1) were inspected, since the other samples might have been older than the regulation. The results are summarized in Table 1 We could prove incorrect ingredients' labeling for 15 of 20 "TD" samples. For 5 of 20 "TD" samples, we could not disprove the correctness of the ingredients list; however, it would be impossible to make different colors with only a white and a black pigment, which were the only labeled pigments. All manufacturers had marks with conditions of use and warnings, even though they had a large variety of descriptions (shown in Table S3, Appendix S1).
(See Discussion section.) The degree of violation of labeling requirements varied among the brands ( Table 1)

| Identification of the pigments in tattoo inks
The pigments used in tattoo ink samples were analyzed both with and without MALDI matrix, since the matrix could result in interfering peaks in the lower m/z region. Detection without matrix was possible because the pigments were able to absorb laser energy.  Figure S1 (Appendix S1).  were declared. Other pigments were detected only with either MS or MS 2 and not included in the statistical analysis (below).
Polyethylene glycol (or PEG) was detected in 27 (37.5%) tattoo ink samples (Table S4), which none of the tattoo ink samples had declared. PEG is a common contamination in MALDI-TOF, and to confirm the presence of PEG in the tattoo ink samples, freshly prepared samples were analyzed. If the mass spectra contained PEG in both sets (with and without matrix), the sample was considered containing PEG.

| Quantification of metals in tattoo inks
The inorganic pigment CI77891 (TiO 2 ) was declared in many labels. Ti was not analyzed in this study; however, its presence was confirmed by the white precipitates observed after the digestion of the samples.
Cu originates from the phthalocyanine pigment group (starting with CI74, Pigment Blue 15, and Pigments Green 7 and 36) and was found mainly in green, blue, purple, and gray inks. Copper ion is the central atom in the structures of these pigments and is firmly bonded to the base structure. 25  The total amount (μg/g) of selected detected metals (Al, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Sr, Mo, Pd, Cd, Sn, Sb, Ba, W, Hg, Pb, and Bi) in the tattoo inks by means of ICPMS is summarized in Figure 2. Samples 57, 58, and 66 were excluded because they were partially dried out.
Metals found in larger quantities (0.3 μg/g -270 mg/g) were Fe, Al, and Cu. Fe showed the highest concentrations (4.39 μg/g -270 mg/ g) in some inks but its use or concentration is not restricted. Fe oxides have been approved as coloring agents in cosmetics 24 and food. 27 More hazardous metals (such as Cd, Pb, and Mn) and strongly sensitizing elements (such as Ni and Cr) were present in relatively lower amounts (shown in Table S5, Appendix S1). Hg, Sb, and Co were only above the detection limits in a few cases. Figure 3 shows the total or regulated under EU, but total Ba (0.051-166 μg/g) was found above the CoE's limit (50 μg/g) in a few inks. Although the metals mentioned above are known as skin sensitizers and/or hazardous substances after shortor long-term human exposure, 28 very few samples exceeded the restricted amounts of these impurities. Cr (0.35-139 μg/g) was found in almost all samples. However, this study did not determine the Cr speciation. It is therefore not possible to judge whether the maximum allowed concentration level of 0.5 μg/g Cr VI as defined by the EU regulation was exceeded. The restriction defined for Ni is 5 μg/g, and three inks were found to exceed it. However, the resolution ResAP(2008)1 recommended "as low as technically achievable" for Ni. 1 All tattoo inks contained quantifiable levels of Ni (0.1-41 μg/g). Although certain pigments containing no Ni could be found on the market, this is not true for all pigments, for example, inorganic Fe oxides pigments. 25 Both Cr and Ni are considered sensitizing elements, and to minimize potential health risk for sensitive individuals, it is recommended that its levels should not exceed 1 μg/g. 16,29,30 According to the newly released EU regulation, the pres- The total metal content (μg/g) of Cu, Cr, Ni, Pb, and Ba is shown for the different brands investigated in this study in Figure 4. The highest median levels of Cu, Cr, and Ni were all observed in "Fu" inks (only statistically significant for Cu compared with "So" brand). Cr contents were statistically significantly greater in the "WF" brand as compared with "In" and "TD." The Ni contents in "In" were statistically significantly greater compared with the "TD" brand. Otherwise, there was no statistically significant difference in these metal contents among the brands. We also found a clear difference in Sr content, with higher levels in the "TD" (0.4 μg/g -8.0 mg/g) and "KS" (1.8-275 μg/g) brands as compared to all other brands (0.2-12.5 μg/g), although this metal impurity is not regulated.  between Cr, Mn, Co, and Ni showed in all cases a large r value close to 1 (r > 0.9) with a highly significant correlation (P < .001). This means that if a tattoo ink contains Cr, it most likely also contains Mn, Co, and Ni. Fe, which is of special interest due to its high content in the tattoo inks ( Figure 2) and common presence in pigments, had a positive, statistically significant correlation with Cr, Co, Ni, Mn, Zn, As, and Pb. Cu, the other common and pigment-included element, had only a positive, statistically significant, correlation with Mo. The impurities Mn, Co, Zn, As, and Pb were strongly correlated with several metals.
The CoE ResAP(2008)1, 1 recommends a maximum concentration of 25 μg/g soluble Cu in tattoo inks, but this concentration limit is increased to 250 μg/g by ECHA. 19 ECHA justifies its proposal in that soluble substances are not expected to accumulate in the organism but are excreted quickly (within a few weeks). 14 The new limit was exceeded by only one blue-colored sample from the brand "Et" (7760 μg/g) among samples 1-56 (Figure 3 The restriction limit of 50 μg/g Ba in tattoo inks refers to total Ba content in the CoE ResAP(2008)1. 1 However, a soluble Ba limit of 500 μg/g has been regulated by ECHA. 19 As can be seen in

| DISCUSSION
This study revealed some alarming trends. From a consumer and medical perspective, the mislabeling of ingredients might be most problematic. There was some indication that mislabeling occurred intentionally, since confirmed (detected by both MS and MS n ) present pigments, not labeled on the ingredients list, were more likely to be among restricted, suspected nonsuitable, or discussed to be banned, pigments (29) as compared to other pigments (9) in this study. However, it cannot be ruled out that this trend is due to analytical limitations or sample selection in this study.
All samples from green, blue, and gray tattoo inks and 75% of the samples from pink and purple inks evaluated in this study contained the pigments that were either identified as not allowed to be used in cosmetics other than rinse-off products by the cosmetics regulation 24 (Pigments Violet 19, Violet 23, and Red 122) or were discussed to be banned but delayed due to the lack of alternatives for tattooing (Pigments Blue 15 and Green 7). 19 Pigment Blue 15 is banned for use in hair dyes, and Pigment Green 7 is banned for use in hair dyes and eye products. Cu-phthalocyanine colorants such as blue and green pigments are very common in cosmetics. 25 High Cu contents in blue and green tattoo inks were also reported in previous studies. 16,25,26 Up to 4310 μg/g soluble Cu in tattoo inks was also reported in an European market survey by EC. 4 The proposed ban or restriction of many between Fe and Ni with P < .001. Battistini et al. 26 found that a mixture of different kinds of metals were often observed simultaneously in tattoo inks, and that the mixture may alter the original toxicity of one metal.
Although this study did not quantify the amount of Ti, it confirmed its presence. Ti originates from the very common white pigment TiO 2 (CI77891, Pigment White 6). A large presence of Ti in tattoo inks was also found by Manso et al. 16 This pigment might be of comparably low concern; however, it is not totally harmless. TiO 2 as nanoparticles (like in pigments) is suggested to cause cancer and other adverse health outcomes. 35 Allergic contact dermatitis to Ti exists in T A B L E 2 Significant correlations between different metals (of 27) in tattoo inks investigated by JASP, expressed as Pearson's correlation coefficient ("r") with its P value (*P < .05, **P < .01, ***P < .001) Note: nsample size (only combinations with both content values above the detection limit were investigated). Corresponding scatter plots in Figure S2 (Appendix S1) rare cases. 36,37 Al is another element found in relatively high concentrations, possibly related to aluminum oxides and silicon oxides (Si was not analyzed), with similar and relatively low, but not absent, toxicity and sensitization potential. 38,39 Ba in tattoo inks originates from BaSO 4 , which is used to brighten darker shades and as a stabilizer. 40 BaSO 4 is of low concern, but soluble impurities can cause a number of adverse health effects, 41  The tattoo needles themselves can be a source of many nano-or micrometer-sized particles (rich in Ni and Cr), especially for inks that contain TiO 2 , as described recently. 7,9 Hence, the mean concentrations of Ni and Cr in tattooed skin could be far higher than measured in the inks due to the tattoo needle wear. Both elements are common allergens, and their target levels in consumer products should be less than 1 μg/g. 16,29,30 This study found polyethylene glycol (or PEG) in several tattoo inks. It is very common to find other substances in tattoo inks, in addition to the pigments, like binders, solvents, and additives, and a plausible source for PEG could be the use of polymeric binders, or surfactants Tween and Triton, which both have PEG as a sidechain. 44 Ninety-three percent of the investigated tattoo inks violated at least one of the legal requirements for labeling by the CoE ResAP (2008)1. 1 In this study, the brands "In," "RC," "Fu," and "Dy" recommended an allergy or patch test before use, without instructions F I G U R E 6 Box plots of water-soluble Cu and Ba content (μg/g) as a function of the brand (A, C) and the color (B, D) in 70 samples. Note that three dried-out samples (samples 57, 58, and 66) were not included, and data are the mean value of triplicate measurements for each sample. Corresponding data in Table S5 (Appendix S1) on how or where to conduct the test. A self-made patch test could be wrongly conducted or read, and even result in sensitization or a wrong belief of absent allergy. In addition, a negative patch test is never a guarantee that allergy is not developed in future (due to long-term exposure to the tattoo ink). Several manufacturers also declared a disclaimer that they would not be responsible for any allergic reaction.
This study is limited by its sample selection, its analytical method limitations, and sample size. However, the studied tattoo inks are sold and used globally. The analytical limitations mean that Ti and Si were not measured and that many possibly hazardous organic compounds were not investigated. This results in an underestimation of possibly hazardous substances in the tattoo inks of this study. Future studies could widen the pigment mass spectrometry library and improve the pigment analysis in tattoo inks in terms of detection limits, interferences, and quantification so that further pigments would be able to be detected. The sample size was primarily of concern for statistical comparisons among brands and colors, since some of the color and brand groups contained only a few samples.
This analytical survey provides color-and brand-resolved information on common pigments in typical tattoo inks and can therefore be used to select patch test substances to find culprit allergens in patients with a contact allergy to certain tattoo inks.

| CONCLUSIONS
The following main conclusions were drawn: warnings, but had various descriptions, some with misleading or dangerous information on skin allergy and patch tests. Only three "WF" and one "So" inks were free of any violations, and the inks from "TD" did not fulfill most of the requirements on tattoo ink labeling. Half of the tattoo inks declared at least one ingredient incorrectly on the label, with a higher probability to not declare a pigment listed as nonsuitable, sensitizing, or discussed to be banned within the EU. Among the detected pigments, only 37.7% were declared on the labels.

CONFLICT OF INTERESTS
The authors declare they have no conflicts of interest.

DATA AVAILABILITY STATEMENT
The data that supports the findings of this study are available in the supplementary material of this article. The data that support the findings of this study are also available from the corresponding author upon reasonable request.