Multimodal imaging of hallucinogens 25C- and 25I-NBOMe on blotter papers

Due to the much lower production costs but similar effects to lysergic acid diethylamide (LSD), phenethylamine derivatives are sold as a cheaper replacement or deceptively as LSD itself. These potent hallucinogenic substances can lead to severe intoxication, thus a more profound understanding of their use is required. This includes the elucidation of the manufacturing processes for the commonly used blotter papers and the assessment of the risk of overdosing because of a heterogeneous distribution on the blotter papers. Besides the rapid detection of the analytes, the manufacturing process was elucidated by three different imaging techniques and liquid chromatography-mass spectrometry (LC – MS). A blotter paper sample, containing the two hallucinogenic phenethylamine derivatives 25I-NBOMe and 25C-NBOMe, was analyzed by complementary techniques such as micro x-ray fluorescence ( μ XRF), laser ablation (LA)-inductively coupled plasma-optical emission spectroscopy (ICP-OES), matrix assisted laser desorption ionization (MALDI)-MS, and with LC – MS after extraction. Using the signal from chlorine and iodine within the compounds, μ XRF proved to be the fastest, cheapest and easiest method for identification, requiring no sample preparation at all. LA-ICP-OES provided three-dimensional information of the elements in the blotter paper. These results helped to confirm the assumption that manufacturers spray the compounds onto the paper. Whereas μ XRF and LA-ICP-OES detected signals for chlorine and iodine, MALDI-MS-imaging showed the molecular distribution of both analytes. LC – MS analyses as a complementary method support the imaging results. Quantitative results for different drug hotspots revealed a heterogeneous distribution of the drugs on the blotter paper implying an inherent risk of overdosing for consumers.


iodo-2,5-dimethoxyphenyl)-N-[(2-methoxyphenyl)methyl]ethylamine]
(25I-NBOMe), were developed for pharmacological studies on the 5-HT 2A receptor and for positron emission tomography (PET) imaging, respectively. [1][2][3] Their rising popularity since 2010 as designer drugs of abuse can be traced back to the much lower costs of synthesis but similar effects compared with the better known hallucinogenic compound lysergic acid diethylamide (LSD). [4][5][6] LSD and NBOMe derivatives as well, are highly potent ligands and partial agonists at the 5-HT 2A -receptor, which induces the typical and desired effects of these drugs. [7][8][9] For NBOMe compounds, effects such as stimulation, euphoria, changes in consciousness and hallucinations are described. [10][11][12][13] Hence, drug producers and dealers sell them as a replacement for LSD, or misleadingly as LSD itself. [14][15][16] Because of the similar effects and the low effective doses of LSD and NBOMes, unintentional effects of NBOMes by consumers may occur. Besides the reported clinical effects such as tachycardia, hypertension, agitation, and aggression, toxicological issues have to be considered as well. 17 Maurer et al. showed an extensive metabolism of 25I-NBOMe, nevertheless, they claim a lack of systematic studies on the metabolism of NBOMe derivatives. 18,19 As already described in many cases, NBOMe consumption may lead to severe health problems or even death, caused by an overdose. [20][21][22][23] To assess the risk of overdose incidents, blotter papers have to be investigated in more detail. NBOMes often are sold on perforated blotter sheets, from which small rectangles, the so-called "trips", can be torn off. One common manufacturing process is the immersion of "empty" blotter paper into a drug solution and subsequent dry hanging. The amount of NBOMes on a single trip ranges from approximately 500 to 1500 μg, while the single effective dose for an unexperienced user is around 100 μg. 24 Coelho et al. used direct ATR-FTIR spectroscopy for rapid detection of NBOMes on blotter papers, which required no sample preparation and was directly applied on blotter papers. The data indicated a heterogeneous distribution of the NBOMes. 25 Furthermore, the infrared spectra showed the presence of a plastic polymer on the front side of the blotter papers.
A blotter sheet, seized by the German police, containing approximately 20 "trips", was analyzed by three different imaging techniques. These 2D analyses will reveal a possible heterogeneous appearance of NBOMes in some areas of the blotter paper, which could lead to overdose incidents if these hotspots are torn off. First, rapid and non-destructive analysis of the elemental distribution was performed by means of micro-X-ray fluorescence (μXRF) spectroscopy and was subsequently supported by matrix assisted laser desorption ionization (MALDI)-mass spectrometry (MS) for molecular imaging. Additionally, imaging utilizing laser ablation (LA)-inductively coupled plasma (ICP)-optical emission spectroscopy (OES) was performed for 3D analysis since LA enables the acquisition of a depth profile. The obtained imaging results were then confirmed via high performance liquid chromatography (HPLC)-MS as a complementary technique.

| Bench-top μXRF
Initial imaging analyses were realized by a laboratory bench-top μXRF

| MALDI-MS
Matrix application was done using an airbrush (infinity, Harder und Steenbeck GmbH und Co. KG, Norderstedt, Germany). A total amount of 600 μL of DHB (0.1 M; 0.1% TFA) was sprayed onto the blotter paper (7 × 7 mm) by applying 100 μL stepwise from an approximate distance of 10 cm. The subsequent analysis of the molecular distribution experiments was carried out, using MALDI-MS imaging (iMScope TRIO, Shimadzu Corp., Kyoto, Japan) with a mass resolution of R = 10,000 and a mass accuracy better than 5 ppm (calculated with external calibration). With a spot size and pitch of 50 μm, accumulations were set to 1 time/pixel, the sample voltage was 3.50 kV and the detector voltage was set to 1.90 kV. The acquired m/z range was 200-500. The parameters of the 355 nm Nd:YAG laser such as the number of shots were set to 100 AU, the repetition rate was 1000 Hz and the laser intensity was set to 61.5 AU. Images were created using the instrument controlling software Imaging MS Solution, version 1.20.14 (Shimadzu Corp., Kyoto, Japan).

| LA-ICP-OES
Imaging data, as well as the depth profile data, were carried out using a laser ablation system with a 213 nm wavelength Nd:YAG-laser (LSX 213 G2+, CETAC Technologies, Omaha, NE, USA) hyphenated to an ICP-OES (Arcos, SPECTRO Analytical Instruments GmbH, Kleve, Germany). For the first image creation, a laser fluence of 5 J/cm 2 and 20 Hz laser shot frequency were chosen. For the following depth profile analyses, the laser energy was reduced by 50%, whereas the laser spot size (50 μm) and scan rate (100 μm/s) were maintained. Helium as carrier gas for the laser ablation cell was set to 0.

| HPLC-MS/MS
To extract both NBOMe compounds from certain regions with higher concentrations (hotspots) or regular areas, specific punch pliers were The oven temperature was set to 50 C. The HESI spray voltage in positive ionization mode was set to 3.5 kV. The cone temperature and heated probe temperature was 350 C, while the cone and probe gas flow were set to 20 AU. The nebulizer gas was 60 AU. Table 1 shows the HPLC-MS/MS separation parameters such as retention time, precursor ions, and corresponding product ions, the scan time and applied collision energy.
An independent two-tail t-test was applied to certify the significance of the statistical difference between the two areas.

| μXRF-imaging
The aim of the project, to obtain two-dimensional as well as threedimensional information required the use of complementary analytical techniques. First, analyses were carried out using the non-destructive μXRF technique, which allows two-dimensional mapping of elemental distributions and does not require any form of sample preparation. In this case, chlorine for 25C-NBOMe was detected using its K-α-line (2.6 keV) and iodine, for 25I-NBOMe, necessarily via its L-β-line (3.9 keV), due to interference of L-α-lines from calcium and iodine.
The L-β-line delivers a much less intense signal than the K-α-line of chlorine, which results in a less explicit image (see Figure 1B). However, iodine and chlorine ( Figure 1B and 1C) exhibit heterogeneous distributions on the blotter paper with hotspots of both elements, respectively. Outlines of the sun on the blotter paper can be slightly observed in the chlorine and iodine distributions in Figures 2B and   2C. This could be caused by a heterogeneous distribution, but more likely because of different X-ray interaction behavior of the blotter paper matrix. 26 Additionally, the elemental signals show that 25C-NBOMe and 25I-NBOMe have the same distribution since both patterns are identical. This leads to the hypothesis that the compounds were transferred to the blotter paper from a single solution. Furthermore, the presence of hotspots indicates an application from above, via a spray device or similar. Some of the hotspots fit the darker spots, which can already be recognized in the photograph of the blotter paper ( Figure 1A).
T A B L E 1 HPLC-MS/MS parameters including retention time (Rt), precursor ion (Prec. ion) and corresponding product ions (Prod. ion) and collision energy (CE) for MRM-mode

| MALDI-MS-imaging
In order to validate that the chlorine and iodine signals observed in  homogeneously onto the blank blotter paper. Nevertheless, regions with black ink showed higher signal intensities for 25C-NBOMe ( Figure S2).

This effect reduces the visibility of the hotspots in the MALDI-
MS images since the scale in these images is normalized and signal intensities are highest in regions with black ink.

| LA-ICP-OES imaging and depth profile
Although MALDI is considered as a semi-destructive method, enough material was left over, so that further analyses could be carried out Afterward, a small area (small red rectangle, Figure 4B) that contained a hotspot was ablated six times consecutively. The obtained data provide three-dimensional information of the compounds on blotter paper. Figure 4C shows

| HPLC-MS/MS
With the obtained LA-ICP-OES data, a depth profile could be created which indicates a decreasing analyte concentration in the third dimension. Nonetheless, it is possible that hotspots only seem to have a higher amount of drugs per cm 2 due to less diffusion in the third dimension. This hypothesis was further investigated via HPLC-MS/ MS after corresponding sample preparation. Therefore, five hotspot regions as well as five regular areas, which did not contain a hotspot, were cut out manually using punch pliers. Afterwards, the analytes were extracted and then analyzed via LC-MS. A recovery rate of 101% was determined using self-produced trips with known concentrations of 25C-NBOMe. Table 2 and Table 3 show the results of the HPLC-MS/MS experiments for 25C-NBOMe and 25I-NBOMe, respectively. The mean amount of 25C-NBOMe in hotspots is significantly higher than in regular areas, which was proved using the independent two-tail t-test  Figure 3C shows a depth profile of the chlorine signal from a hotspot (small rectangle in Figure 4B), which was obtained by ablating the same area six times via LA-ICP-OES. Experiments were performed using a spot size of 50 μm and a scan rate of 100 μm/s [Colour figure can be viewed at wileyonlinelibrary.com] the different techniques, the hypothesis of hotspots, which contain a significantly higher amount of drug compared with regular areas, was shown. Furthermore, a manufacturing process via a spray device could be confirmed.  Table 3.
Similar to the results of 25C-NBOMe, the mean amount of 25I-NBOMe in the hotspots is significantly higher than in the regular areas. A total number of five hotspots and five regular areas was used.

| CONCLUSIONS
In this proof-of-concept study, non-destructive μXRF was used as the The observed decreasing signal intensity in the third dimension elucidated the manufacturing process, where the compounds are sprayed onto the blotter sheets from one side, in contrast to the frequently applied manufacturing process of immersing the blotter paper in a solution of the active substance followed by air-drying. The occurrence of hotspots can be caused by a droplet formation during the spray process. Therefore, the complementary results of all three imaging techniques support the two major hypotheses, namely inhomogeneous distribution and manufacturing process.
As an approach that is independent of the imaging techniques, HPLC-MS/MS analyses and statistical evaluation were applied to confirm these results and to assess the potential risk of overdose because of heterogeneous distribution of the active substances on the investigated blotter paper. The HPLC-MS results did not show major differences referring to the consumption unit of single trips.
This can be attributed to the manufacturing process by one-sided spraying, which seems to be less prone to cause inhomogeneous distribution on the paper surface compared with immersing and drying. The main source of heterogeneity found for the studied blotter papers was the hotspots caused by the spray process. These hotspots could contribute to an overdose incident if an unusually high number of them were located on a single trip, but their overall contribution to the total amount of active substance in a trip is less than originally expected.
T A B L E 3 Mean amounts in cutouts of 1 cm 2 , standard deviations and relative standard deviations of the quantification via LC-MS of 25I-NBOMe after extraction of the compounds from hotspot regions and regular areas