Characterization of emissions from a desktop 3D printer and indoor air measurements in office settings

ABSTRACT

Emissions from a desktop 3D printer based on fused deposition modeling (FDM) technology were measured in a test chamber and indoor air was monitored in office settings. Ultrafine aerosol (UFA) emissions were higher while printing a standard object with polylactic acid (PLA) than with acrylonitrile butadiene styrene (ABS) polymer (2.1 × 10⁹ vs. 2.4 × 10⁸ particles/min). Prolonged use of the printer led to higher emission rates (factor 2 with PLA and 4 with ABS, measured after seven months of occasional use).

UFA consisted mainly of volatile droplets, and some small (100–300 nm diameter) iron containing and soot-like particles were found. Emissions of inhalable and respirable dust were below the limit of detection (LOD) when measured gravimetrically, and only slightly higher than background when measured with an aerosol spectrometer. Emissions of volatile organic compounds (VOC) were in the range of 10 µg/min. Styrene accounted for more than 50% of total VOC emitted when printing with ABS; for PLA, methyl methacrylate (MMA, 37% of TVOC) was detected as the predominant compound. Two polycyclic aromatic hydrocarbons (PAH), fluoranthene and pyrene, were observed in very low amounts. All other analyzed PAH, as well as inorganic gases and metal emissions except iron (Fe) and zinc (Zn), were below the LOD or did not differ from background without printing.

A single 3D print (165 min) in a large, well-ventilated office did not significantly increase the UFA and VOC concentrations, whereas these were readily detectable in a small, unventilated room, with UFA concentrations increasing by 2,000 particles/cm³ and MMA reaching a peak of 21 µg/m³ and still being detectable in the room even 20 hr after printing.

KEYWORDS:

• Fused deposition modeling
• indoor
• office
• ultrafine aerosols
• volatile organic compounds

Acknowledgments

Special thanks go to Konrad Schelbert for VOC analytics, to Christoph Bossard (Suva) and Sven Hoffmann (ETH Zurich) for their very helpful and benevolent supervision, and last, but not least, to the reviewers for their stimulating remarks.

Funding

This study was conducted in the framework of a master of advanced studies at the federal institute of technology (ETH) Zurich. The author wishes to thank his employer, Swiss national accident insurance fund (Suva), for its financial and logistic support.