In Vitro Quantitative Chemical Analysis of Tattoo Pigments

Background  The composition of cosmetic tattoos might prove relevant to their treatment by high-powered lasers.

Objectives  To test the accuracy and completeness of information supplied by the tattoo ink manufacturers and to perform an elemental assay of tattoo pigments using scanning electron microscopy with energy-dispersive x-ray analysis.

Design  Samples of 30 tattoo inks were examined using "standardless" energy-dispersive spectrometry. This technique uses quantitative electron x-ray microanalysis. The technique reliably identifies all elements with the exception of those elements with atomic numbers less than 11.

Setting  A major national referral laboratory for microscopic examination and biochemical analysis of tissue. These results were compared with ink compositions compiled from manufacturer-supplied material safety data sheets.

Main Outcome Measures  (1) The percentage of any given element in whole tattoo pigments. (2) The presence or absence of elements and/or compounds as recorded in material safety data sheets supplied by the tattoo ink manufacturers.

Results  Of the 30 tattoo inks studied, the most commonly identified elements were aluminum (87% of the pigments), oxygen (73% of the pigments), titanium (67% of the pigments), and carbon (67% of the pigments). The relative contribution of elements to the tattoo ink compositions was highly variable between different compounds. Overall, the manufacturer-supplied data sheets were consistent with the elemental analysis, but there were important exceptions.

Conclusion  The composition of elements in tattoo inks varies greatly, even among like-colored pigments. Knowledge of the chemical composition of popular tattoo inks might aid the clinician in effective laser removal.

THE COMPOSITION of tattoo pigments is of interest not only because of photoallergic, granulomatous, and anaphylactic reactions,^1,2 but also because it has been suggested that the chemical composition might be a predictive factor in the response of tattoos to laser irradiation.^3,4 Pigment compositions of certain ink colors have changed over time as some metals produce adverse reactions and have been almost eliminated from tattoo pigments.⁵ Cinnabar (mercuric sulfide), for example, once used in red tattoos, has been removed from pigments produced in the United States.

Researchers have proposed that some tattoos may be resistant to treatment not only because of poor absorption of laser energy by tattoo pigment but also as a result of oxidative-reductive changes that occur in some metals on excitation from laser light.⁵ For example, tattoos with iron oxides and titanium dioxide (TiO₂) have been shown to undergo darkening with high-powered laser treatment.^3,4,6 It follows that identification of these compounds prior to treatment might prove helpful in predicting the response to laser treatment. Accordingly, using energy-dispersive x-ray analysis, we identified the elemental composition of commonly used tattoo pigments.

Twenty-nine tattoo pigments were obtained from one of the leading tattoo ink suppliers in the United States (Huck Spaulding Enterprises Inc, Voorheesville, NY). The pigments comprised a kit supplied by the company, and included a range of their most popular pigments. In addition, material safety data sheets (MSDSs) for each pigment in the kit were requested from the manufacturer. Also, a sample of india ink was obtained from another manufacturer (Eberhard Faber Inc, Lewisburg, Tenn) and analyzed. The stock numbers, colors, and pigment names are listed in Table 1.

The pigments were transferred from the stock bottles to 1.8-mL cryotube vials (Nalge Nunc International, Rochester, NY) and mailed to one of us (F.B.J.) for analysis. Samples of the tattoo inks were stirred, then spread on the smooth surface of 1-in- (2.45-cm-) diameter carbon disks. The specimens were dried and examined, uncoated, using a scanning electron microscope (DSM 960A; Carl Zeiss Inc, Thornwood, NY) and an integrated microanalyzer for images and x-rays (IMIX; Princeton Gamma-Tech, Inc, Princeton, NJ). The technique, quantitative electron x-ray microanalysis by energy-dispersive spectrometry (EDS), has become over the past 30 years the dominant spectrometric technique in instruments such as the scanning electron microscope.⁷ The technique reliably identifies all elements with the exception of those with atomic numbers less than 11. Elements with atomic numbers less than 11 include the following: hydrogen, helium, lithium, beryllium, boron, carbon, nitrogen, oxygen, fluorine, and neon. Hydrogen, helium, and lithium are not recognized by EDS. Depending on the spectrometer window transmission characteristics, beryllium, boron, carbon, nitrogen, oxygen, and fluorine may or may not be identified. The EDS technique used in this study used an ultrathin window that allowed for identification of many of these lighter elements. Despite the fact that newer EDS techniques have made it possible to identify elements with atomic numbers less than 11, quantifying these elements by EDS remains problematic. For elements with atomic numbers less than 11, mass spectrometry remains the standard criterion.

On bombardment with electrons, the samples emit x-rays that characterize the elemental composition of the samples and give an indication of the relative amounts of various elements present. Once a qualitative analysis has been made, the relative quantities can be assessed and expressed in mass or atomic fraction. Conventional qualitative analysis is performed with standards. In this study, we used a "standardless" technique⁷ that seeks to simplify the measurement process by requiring the analyst only to measure the EDS spectrum of the unknown. Rather than depend on direct measurement of standard intensities, the standardless technique relies on independent calculations and the analysis report is generated by a software system. The popularity of the standardless analysis is attributed to simplicity of operation, close relation to the conventional standards procedure, and the reassuring analytical total of unity. The total elemental composition should total 100%.

The results of the energy-dispersive x-ray diffraction are listed in Table 1. The elements are listed as a percentage of the total composition of each tattoo ink compound. Overall, oxygen (73% of the pigments), titanium dioxide (TiO₂) (67% of the pigments), and carbon (67% of the pigments) were the most frequently identified elements. The frequency with which a certain element was identified in like-colored inks is shown in Figure 1. For elements that were commonly identified, the relative contribution of that element to the overall ink composition varied widely between different inks, even in inks with the same gross color.

Review of the MSDSs is given in Table 2. The composition of tattoo inks was typically the sum of other inks. For example, in the case of blush (stock No. 9097), the ink contained orange 2915, yellow 972-3119, white (R-900), plus a solvent mix composed of distilled water and isopropyl alcohol. The designation of an "x" signifies the presence of an element (the relative contributions were not included in the MSDSs). Overall, there was good correlation between the MSDSs and the experimental results, with the following exceptions. Pigment stock No. 9093 (misty blue) was listed in the MSDSs as containing copper; however, chemical analysis failed to demonstrate this element. This finding is the most substantial discrepancy. Others are detailed as follows. Five pigments (stock Nos. 9017 [lemon yellow], 9025 [emerald green], 9094 [tulip yellow], 9095 [peony pink], and 9098 [wild violet]) were identified as having barium in the MSDSs; however, quantitative analysis did not reveal the presence of barium. Nitrogen was reported in the MSDSs of numerous pigments; yet, analysis failed to detect its presence. This finding can be attributed to the inherent difficulty in identifying light elements with EDS as was discussed in the "Materials and Methods" section.⁷ Indeed, the chemical formulas of the azo dyes provided by the tattoo manufacturer revealed only small amounts of nitrogen as a constituent of the pigments. Carbon and oxygen (also elements with atomic numbers <11) were detected by EDS, primarily owing to the spectrometer window used.

This study showed, not unexpectedly, that individual tattoo inks are complex compounds whose composition may include organic dyes, metals, and solvents. Adding to this complexity, the individual inks are often mixed prior to final manufacturing. Finally, on delivery of the tattoo inks to the artist, it is common for he or she to mix these already complex compounds with other complex pigments. Accordingly, the final color of an in vivo tattoo is often the product of the manufacturer's tendencies and the artist's license. We do not suggest that biopsy specimens be submitted as a guiding tool in laser treatment. However, the findings in this study allow the treating physician to make educated guesses regarding tattoo ink composition. For example, we discovered the likelihood of a blue ink containing TiO₂ is 100%, and that the percent composition of TiO₂ranged from 38% to 95% in those blue inks. Figure 1 is included in the article to assist the clinician in predicting what elements may be present in a particular color category. One would expect that these inks, although showing good absorption by spectroscopy,⁸ might be resistant to treatment based on laser-induced darkening. Titanium dioxide is added to paints and tattoo pigments as a brightening agent. It is also a common ingredient in sunscreens. The chemical properties of TiO₂are such that it absorbs UV light well in the range between 280 and 400 nm. At wavelengths greater than 400 nm, TiO₂permits light transmittance with little absorption.⁹ However, because of the high intensity of the nanosecond pulses usually applied in laser treatment, even the small relative absorption of other wavelengths is enough to observe the darkening reaction, which is based on reduction of TiO₂in the compound.⁶ It is unclear whether the amount of TiO₂ affects the degree of darkening. Presumably, tattoos such as 8007 white (98.55% TiO₂) would darken more than those inks with less TiO₂, but this can only be proven through experiment.

The other compound associated with darkening is iron oxide.³ Where it was found, the relative percentage was high, which may explain the high clinical frequency at which these tattoos darken during laser treatment. Given the data presented in this article, it may be wise to counsel patients with tattoo colors that are known to contain TiO₂ and iron oxide about the potential for treatment resistance or even tattoo darkening. And since some, but not all, laser-darkened tattoos have been shown to subsequently lighten with additional Q-switched laser treatment,¹⁰ test sites should be considered in these patients before treating the entire tattoo.

The lack of perfect correlation between the MSDSs' composition and the experimental results suggests that although these manufacturers' sheets are helpful, they cannot be relied on for a complete characterization of the tattoo ink composition. Nor are they useful for quantifying the amount of element present.

Previous investigators^1,5,11 have examined the elemental ink composition in biopsy specimens submitted from patients. Their results show that ink chemistries have changed considerably over the past 15 years, even in same-colored pigments. Particularly noteworthy is the absence of cadmium, cobalt, mercury, and lead from the inks in our study. These changes in part are because of the Food, Drug, and Cosmetic Act of 1976, which limited the concentration of lead and mercury in cosmetics to 10 ppm and 3 ppm, respectively.

In vivo absorption spectra of tattoos have been suggested as a guide in the choice of laser wavelength.⁸ Certainly, absorption of laser light is a prerequisite for any effective treatment, and so this kind of guidance is a first step in efficient laser tattoo lightening. However, photon absorption by the ink does not guarantee lightening, only that there presumably will be localized particle heating. As noted previously, TiO₂ and iron oxide darken with high-intensity laser irradiation.⁴ Moreover, identification of TiO₂ within a composite ink might not always be possible by standard absorption spectra using visible wavelengths.

It would be of value in the future to encourage tattoo artists to record in the client's record the manufacturer and stock number of the pigments used. In the future, physicians may have access to the absorption spectra, as well as the chemical constituents, of these pigments. We consider the identification of ink constituents as a first, but not final, step in the improvement of tattoo laser treatment. Once achieved, and assuming that manufacturers do not change the composition without notice, ex vivo experiments can be performed to correlate tattoo ink composition with absorption spectra and response to laser irradiation. We are performing these experiments. The final step will be to investigate the response of these inks in vivo and finally generate a list of inks that might prove "permanent" but "laser removable." This final step would require the cooperation of the tattoo industry. The result is a win-win situation for the consumer and the tattoo industry. After all, the concept of a removable tattoo might prove attractive to some clients and actually increase the absolute number of tattoo recipients. Also, working with the tattoo industry, cosmetically elegant inks could still be manufactured, but the composition would be known to physicians and scientists. This knowledge would enable the laser industry and laser physicians to optimize treatment.

Accepted for publication October 11, 2000.

Presented at the annual meeting of Lasers in Surgery and Medicine, Reno, Nev, April 8, 2000.

The views expressed in this article are those of the authors and do not reflect the official policy or position of the US Department of the Navy, US Department of Defense, or the US government.

Corresponding author: CDR E. Victor Ross, MC, USNR, Department of Dermatology, Naval Medical Center, 34520 Bob Wilson Dr, Suite 300, San Diego, CA 92134-2000.