This method for two-step pyrolysis online coupled to gas chromatography with mass spectrometric detection and data evaluation protocol can be used for multi-component analysis of tattoo inks and discrimination of counterfeit products.
Tattoo inks are complex mixtures of ingredients. Each of them possesses different chemical properties which have to be addressed upon chemical analysis. In this method for two-step pyrolysis online coupled to gas chromatography mass spectrometry (py-GC-MS) volatile compounds are analyzed during a first desorption run. In the second run, the same dried sample is pyrolyzed for analysis of non-volatile compounds such as pigments and polymers. These can be identified by their specific decomposition patterns. Additionally, this method can be used to differentiate original from counterfeit inks. Easy screening methods for data evaluation using the average mass spectra and self-made pyrolysis libraries are applied to speed up substance identification. Using specialized evaluation software for pyrolysis GS-MS data, a fast and reliable comparison of the full chromatogram can be achieved. Since GC-MS is used as separation technique, the method is limited to volatile substances upon desorption and after pyrolysis of the sample. The method can be applied for quick substance screening in market control surveys since it requires no sample preparation steps.
Tattoo inks are complex mixtures consisting of pigments, solvents, binders, surfactants, thickening agents, and, sometimes, preservatives1. The increased popularity of tattooing in the last decades has led to the establishment of legislation addressing tattoo ink safety across Europe. In most instances, color-giving pigments and their impurities are restricted and therefore should be monitored by state laboratory market surveys to control their compliance with law.
Using the approach of online pyrolysis-gas chromatography mass spectrometry (py-GC-MS) described here, multiple ingredients can be identified simultaneously. Since volatile, semi-volatile and non-volatile compounds can be separated and analyzed within the same process, the variety of target compounds is high compared to other methods used for tattoo ink analysis. Liquid chromatography methods are mostly carried out with pigments solubilized in organic solvents2. Raman spectroscopy as well as Fourier-transform infrared (FT-IR) spectroscopy have been described as suitable tools for the identification of pigments and polymers but are limited with multi-ingredient mixtures since no separation technique is used in standard laboratory applications3,4. Laser desorption/ionization time-of-flight mass spectrometry (LDI-ToF-MS) has also been used for pigment and polymer identification5,6. Altogether, most methods lack the analysis of volatile compounds. The lack of suitable commercial spectral libraries is a common disadvantage of all of these methods. The identification of inorganic pigments has often been carried out with either inductively coupled plasma mass spectrometry (ICP-MS)7,8 or energy dispersive X-ray spectroscopy (EDX)4,9. Also, FT-IR and Raman spectroscopy have been used for the analysis of inorganic pigments such as titanium dioxide or iron oxides in other research fields10,11,12,13.
The goal of this study was to establish a method applicable in standard analytical laboratories with moderate financial costs to upgrade existing and common devices. Py-GC-MS as described here is a non-quantitative approach for identification of organic ingredients from mixtures. Upon identification of suspicious substances in a py-GC-MS screening, target substances can be quantified with more specialized approaches. It is especially interesting for the analysis of non-volatile and non-soluble substances like pigments and polymers.
The described method can be adapted for inks and varnishes in other fields of application. The data evaluation methods described are applicable to all pyrolysis investigations. Also, counterfeit products, mostly from Asian markets, display a potential source of risk to the consumer and a financial burden to manufacturers (personal communication at the 3rd ECTP in Regensburg, Germany, 2017). The method described here can be used to compare the characteristics of putative counterfeit inks to an original bottle, similar to published forensic approaches for car varnish identification14.
1. Tattoo ink preparation and sample mounting
2. Analysis of ink samples by py-GC-MS
3. Data evaluation approaches
NOTE: Data evaluation should be adapted depending on the individual analytical questions, e.g., the search for volatiles, non-volatile compounds, hazardous cleavage products from azo pigments, or similar.
The method includes a two-step chromatographic approach for each sample (Figure 1). In the first run, the sample is dried inside the injector system at 90 °C before volatile compounds are transferred onto the column. Since the drying process is incomplete in most cases, residual solvents and volatile compounds are transferred and analyzed. In the second run, the previously dried sample is subsequently pyrolyzed to facilitate the analysis of non-volatile organic components.
Figure 1: Schematic diagram of the pyrolysis workflow. Please click here to view a larger version of this figure.
Well produced inks with highly pure ingredients and a limited number of components result in chromatograms easy to interpret with standard libraries, since most peaks can be identified (Figure 2). But even in the high-quality inks, non-declared ingredients have been found. For example, propylene glycol is often found in addition to the declared glycerol (Figure 2 and Figure 3).
Other substances such as formaldehyde might be added as a preservative. 1-Hydroxy-2-propanone may result as an impurity of pigment synthesis and is therefore an example of a non-intentionally added substance (NIAS).
Figure 2: Representative results of the py-GC-MS analysis of a tattoo ink with only a few, pure ingredients. (A) 1st run: desorption for identification of volatile compounds. (B) 2nd run: pyrolysis for identification of less and non-volatile compounds. Declared and identified ingredients are indicated below. Please click here to view a larger version of this figure.
Inks containing multiple ingredients and impurities will result in a pyrogram that is difficult to interpret (Figure 3). Most peaks occurring in the second run may not be baseline separated from each other, making identification difficult, even when utilizing data deconvolution. Some substances might also result in peaks below the threshold set during data evaluation (e.g., 0.2% of total peak area). A solution to this problem might be a step-wise approach using 400 °C, 600 °C, and 800 °C in consecutive pyrolysis steps for the very same sample (see Figure 4).
Some pigment decomposition products may descent from multiple pigments (Supplementary Table 1). For example in Figure 3 and Figure 4, acetyl cyanide can derive from multiple yellow or orange pigments. The degradation product 2-methoxyphenylisocyanate may also derive from the pigment Red 9, and o-anisidine from the pigments Red 170 and Red 9. However, due to the combination with the degradation product 4-methoxy-2-nitro-aniline and the yellow appearance of the ink, only pigments yellow 65 and 74 would be plausible as parent compounds. These two pigments are regional isomers and cannot be distinguished from another with this method. Pigment orange 13 — which was declared on the ink bottle — has not been identified. If the pigment was only present in low amounts, it might have been below the limit of detection. On the other hand, inks are often declared falsely27.
Figure 3: Representative results of the py-GC-MS analysis of the yellow tattoo ink “banana cream” with many, impure ingredients. (A) 1st run: desorption for identification of volatile compounds. (B) 2nd run: pyrolysis for identification of less and non-volatile compounds. Declared and identified ingredients are indicated below. Please click here to view a larger version of this figure.
Figure 4: Gradual pyrolysis of the yellow ink “banana cream” displayed in Figure 3. A-C) Consecutive pyrolysis runs at 400 °C, 600 °C, and 800 °C. Please click here to view a larger version of this figure.
A positive result for counterfeit product identification is displayed in the following example (Figure 5). Three “lemon yellow” inks have been purchased either from a licensed distributor of the US-based ink manufacturer, via an internet auction platform or via an Asian market place. All inks have been analyzed with the two-step py-GC-MS method. In this example, the differences in peak numbers and retention times are already visible by eye.
The chromatogram from the 1st desorption run and the pyrogram from the 2nd run of the original ink were compared against three independent acquisitions of the original ink and the two counterfeit products using pyrogram evaluation software. The software was found to be highly useful in distinguishing the different inks. The forward match factor was above 0.9 (with 1 being the perfect match) only towards pyrograms or desorption chromatograms of the same ink, respectively.
Also, forward matches above 0.9 were only achieved with the same ink when comparing the ink to library contained pyrograms of 84 inks of various colors and manufacturers.
Alternatively, a statistical comparison as proposed by Yang et al. for car varnishes may be applied14.
Figure 5: Identification of counterfeit products by py-GC-MS. Three “lemon yellow” inks from licensed distributor (A), an online auction platform (B), and an Asian market place (C) have been analyzed. Please click here to view a larger version of this figure.
Py-GC-MS is a useful screening method for a broad range of substances in tattoo inks that can also be used for the analysis of other products. Compared to other methods, py-GC-MS can be conducted with only minimal sample preparation. GC-MS devices can be found in most analytical laboratories compared to more specialized methods such as MALDI-ToF-MS and EDX.
The data evaluation of pyrograms may be challenging, since the list of possible ingredients is infinite in theory and library searches that also account for the combination of substances towards a parent compound in the library are necessary. The data evaluation methods described here allow for reliable fast screening of substances that have been added to standard pyrogram libraries. Conversely, testing for counterfeit products is a fast and straightforward approach that can be conducted without any pre-build libraries, since the identification of single substances is irrelevant.
To obtain the best possible results, the amount of ink added to pyrolysis should neither be too high nor too low. This will either result in a contamination of the pyrolysis unit, liner or column or a lack of significant peaks for proper pyrogram interpretation. Therefore, using a defined volume of ink as described in this method with adjusted split ratios is highly recommended. As shown in Figure 3, impurities or polymers can overload the pyrogram with peaks impairing the identification of single substances. Therefore, the detection limit for pigments is highly dependent on the corresponding mixtures. In such cases, pigments might first be separated from other ink ingredients by dilution and precipitation in alcohol-water solvents.
The limitation of the method is the analysis of organic pigments without specific cleavage sides such as quinacridones, perylenes and perinones16,17,18. Also, if a mixture of multiple pigments with the same cleavage group occurs (e.g., with azo pigments), the identification might be challenging (cf. ink displayed in Figure 3). In addition, the pyrolysis products must be able to enter the gas phase. Polymers like hydroxyethyl-cellulose consisting of sugar monomers that have to be chemically modified for GC-MS analysis cannot be detected by py-GC-MS. As in all other methods, only pigments with known pyrograms can be identified. However, main decomposition products can be concluded from the pigment structure, especially in the case of azo pigments. Therefore, checking for plausibility of a declared pigment can be carried out even if the pigment has never been analyzed in py-GC-MS before.
The method can be used to discriminate original inks from counterfeit products. However, a reliable sample of the original ink must be available. Since the composition of inks may change over time, inks produced in the same time range or at best from the same batch must be used for comparison. In future, py-GC-MS might be used to monitor tattoo ink ingredients and thereby reveal declaration fraud and the use of banned pigments and possible other ingredients. A further application of these methods might be the identification of counterfeit products14.
The authors have nothing to disclose.
This work was supported by the intramural research project (SFP #1323-103) at the German Federal Institute for Risk Assessment (BfR).
99.999% Helium carrier gas | Air Liquide, Düsseldorf, Germany | – | |
5975C inert XL MSD with Triple-Axis Detectors | Agilent Technologies, Waldbronn, Germany | – | |
7890A gas chromatograph | Agilent Technologies, Waldbronn, Germany | – | |
AMDIS software (Version 2.7) | The National Institute of Standards and Technology, Gaithersburg, MD, USA | – | can be used for GC/MS peak integration, e.g. for transfer to pyrogram evaluation software |
Cold Injection System (CIS) | Gerstel, Mühlheim, Germany | – | |
electron impact (EI) source | Agilent Technologies, Waldbronn, Germany | – | |
Enhanced ChemStation (E02.02.1431) | Agilent Technologies, Waldbronn, Germany | – | used to generate Average Mass Spektra (AMS), can be used for peak integration and standard GC/MS library search |
J&W HP-5MS GC Column, 30 m, 0.25 mm, 0.25 µm, 5975T Column Toroid Assembly | Agilent Technologies, Waldbronn, Germany | 29091S-433LTM | |
MassHunter Software | Agilent Technologies, Waldbronn, Germany | – | no Version specified, can be used for GC/MS peak integration and standard GC/MS library search |
Microcapillary tube Drummond Microcaps, volume 2 µL | Sigma-Aldrich, St. Louis, MO, USA | P1549-1PAK | |
MS ChromSearch (Version 4.0.0.11) | Axel Semrau GmbH & Co. KG, Sprockhövel, Germany | – | specialized pyrogram evaluation software |
NIST MS Search Program (MS Search version 2.0g) | The National Institute of Standards and Technology, Gaithersburg, MD, USA | – | used for MS and AMS library generation and corresponding substance search with selfmade and commercial libraries |
NIST/EPA/NIH Mass Spectral Library (EI) mainlib & replib (Data version: NIST v11) | The National Institute of Standards and Technology, Gaithersburg, MD, USA | – | used commercial mass spectral library |
Polystyrene (average Mw ~192,000) | Sigma-Aldrich, St. Louis, MO, USA | 430102-1KG | |
Pyrolysis tubes, tube type – quartz glass – lenght 25 mm; 100 Units | Gerstel, Mühlheim, Germany | 018131-100-00 | |
Pyrolyzer Module for TDU | Gerstel, Mühlheim, Germany | – | |
Quartz wool | Gerstel, Mühlheim, Germany | 009970-076-00 | |
Steel sticks | Gerstel, Mühlheim, Germany | – | |
Thermal Desorption Unit (TDU 2) | Gerstel, Mühlheim, Germany | – | |
Transport adapter | Gerstel, Mühlheim, Germany | 018276-010-00 | |
Tweezers for Pyrolysis tubes | Gerstel, Mühlheim, Germany | 009970-074-00 | |
Zebron Z-Guard Hi-Temp Guard Column, GC Cap. Column 10 m x 0.25 mm, Ea | Phenomenex Ltd. Deutschland, Aschaffenburg, Germany | 7CG-G000-00-GH0 |