Arsenic and arsenic species in MOD, POD, and disposable POD electronic cigarette aerosols: a pilot study

The growing popularity of electronic cigarettes (e-cig) has raised questions about the health effects of e-cig use, or vaping. Previous studies have reported on the potential of exposure to arsenic (As) and other metal(loid)s from vaping, but little is known about the speciation of As in the inhaled aerosols, an important determinant of toxicity. Inorganic As (iAs) species AsIII and AsV are generally more hazardous than organic As species. This study aimed to investigate total and speciated As in condensed aerosols of popular commercial e-cig products and to compare them with regulatory exposure limits. High-performance liquid chromatography and inductively-coupled plasma mass spectrometry were used for As measurements of e-cig aerosol condensates. The analysis included samples from three types of e-cig devices: MODs, PODs, and disposable pod (d-POD) devices. iAs species were identified in all 23 analyzed e-cig aerosol condensate samples, with the highest aerosol concentrations measured in MODs. The geometric mean (range) iAs concentration of 2.3 (1.2–5.1) μg/m3 observed in MOD devices in this study exceeded the recommended exposure limit of 2 μg/m3 for 15-min or shorter inhalation exposures set by the United States National Institute for Occupational Safety and Health. These preliminary results suggest that iAs species are present in inhalable aerosols of some MOD products at levels above regulatory limits for iAs inhalation.


Table of Contents Description
Page #

Table S1
Orbitrap LC-MS instrumental parameters for full scan and dd-MS 2 in positive and negative polarity modes.S2 Table S2 Overview of quality control mix compounds and analytical results.Mass error is calculated based on the detected adduct mass (n=8 replicates).S4 Table S3 Spikes of quality control mix into PG/VG base and Juul e-liquid.S5 Table S4 Compound Discoverer workflow parameters used.

S7 -S10
Table S5 Non-target workflow results for known compounds in the QC mixture and reference standard mixture.Sfit% and mzCloud best match scores were obtained from a single CD analysis of data from multiple runs using the nontarget instrumental method.For the standard mix, data from aerosols and eliquids over two runs was included.For the QC mix, data from all replicates (n=8) over 4 runs were included.

Figure S1
Example standard addition calibration curves for (a) tributylphosphine oxide in Juul aerosol and (b) vanillin in Mi-Salt e-liquid.Un-spiked calibrators (0 ng/mL) not shown due to log scale.

Figure S4
Kendrick Mass Defect plots for methylene homologous series in e-liquids and aerosols (vaping session 1) of all commercial products analyzed.Red symbols in Blu Aerosol-1 represent three series of homologous compounds detected only after vaping in Blu, all classified as lipid-like by MSCC and sharing similar retention times.

Figure S1 :
Figure S1: Example standard addition calibration curves for (a) tributylphosphine oxide in Juul aerosol and (b) vanillin in Mi-Salt e-liquid.Un-spiked calibrators (0 ng/mL) not shown due to log scale.

Figure S4 :
Figure S4: Kendrick Mass Defect plots for methylene homologous series in e-liquids and aerosols (vaping session 1) of all commercial products analyzed.Red symbols in Blu Aerosol-1 represent three series of homologous compounds detected only after vaping in Blu, all classified as lipidlike by MSCC and sharing similar retention times.

Figure S6D :
Figure S6D: MS 2 spectra of reference standard and e-cig sample for harmaline.

Table S1 :
Orbitrap LC-MS instrumental parameters for full scan and dd-MS 2 in positive and negative polarity modes.

Table S2 :
Overview of quality control mix compounds and analytical results.Mass error is calculated based on the detected adduct mass (n=8 replicates).

Table S3 :
Spikes of quality control mix into PG/VG base and Juul e-liquid.

Table S4A :
Compound Discoverer workflow parameters used.

Table S5 :
Non-target workflow results for known compounds in the QC mixture and reference standard mixture.Sfit% and mzCloud best match scores were obtained from a single CD analysis of data from multiple runs using the non-target instrumental method.For the standard mix, data from aerosols and e-liquids over two runs was included.For the QC mix, data from all replicates (n=8) over 4 runs were included.