Investigating environmentally persistent free radicals (EPFRs) emissions of 3D printing process

Abstract

In recent years, the emission of particles and gaseous pollutants from 3D printing has attracted much attention due to potential health risks. This study investigated the generation of environmentally persistent free radicals (EPFRs, organic free radicals stabilized on or inside particles) in total particulate matter (TPM) released during the 3D printing process. Commercially available 3D printer filaments, made of acrylonitrile–butadiene–styrene (ABS) in two different colors and metal content, ABS-blue (19.66 μg/g Cu) and ABS-black (3.69 μg/g Fe), were used for printing. We hypothesized that the metal content/composition of the filaments contributes not only to the type and number of EPFRs in TPM emissions, but also impacts the overall yield of TPM emissions. TPM emissions during printing with ABS-blue (11.28 μg/g of printed material) were higher than with ABS-black (7.29 μg/g). Electron paramagnetic resonance (EPR) spectroscopy, employed to measure EPFRs in TPM emissions of both filaments, revealed higher EPFR concentrations in ABS-blue TPM (6.23 × 10¹⁷ spins/g) than in ABS-black TPM (9.72 × 10¹⁶ spins/g). The presence of copper in the ABS-blue contributed to the formation of mostly oxygen-centered EPFR species with a g-factor of ~2.0041 and a lifetime of 98 days. The ABS-black EPFR signal had a lower g-factor of ~2.0011, reflecting the formation of superoxide radicals during the printing process, which were shown to have an “estimated tentative” lifetime of 26 days. Both radical species (EPFRs and superoxides) translate to a potential health risk through inhalation of emitted particles.

1. Introduction

3D printers utilizing fused filament fabrication (FFF) are widely used in educational, industrial, and household settings. The principle of FFF is the extrusion of melted thermoplastic filament from a heated extrusion head and deposition of the melted material in thin layers onto a moving platform. FFF filaments are available as different thermoplastics (e.g., acrylonitrile butadiene styrene (ABS), polycarbonate (PC), polylactic acid (PLA), etc.) often mixed with other materials such as wood particles, metals, and colorants. ABS is a commonly used filament in 3D printing. During heating, partial decomposition and emission of particulate matter (PM) and volatile gases occurs, which can deteriorate indoor air quality [1–9]. The thermoplastic extrusion method in 3D printing has been recognized to emit hazardous compounds such as acrylonitrile, polyaromatic hydrocarbons (PAHs), and volatile organic compounds (VOCs) as a function of the material used [2,6,10,11]. Exposure to emissions containing thermal decomposition products of thermoplastics can have toxic effects in animals [12–14]. Inhalation of VOCs or adsorption of VOCs onto the surface of particles that travel through the respiratory tract may lead to adverse health effects in humans, including allergies, asthma, obstructive pulmonary disease, and lung cancer [15–17]. Exposure to airborne ultrafine particles (UFPs) has been linked to adverse health effects [18–22] to the cardiovascular and respiratory systems, with many of those effects associated with generation of reactive oxygen species (ROS) and oxidative stress, which can cause DNA damage, denaturation of proteins and enzymes, and oxidation [23]. There are several reported cases of occupational diseases related to 3D printing including a report of work related asthma [24], cases of contact rhinitis [25], and a case of hypersensitivity pneumonitis associated with nylon powder used in 3D printing [24].

The filament brand, color, metal additives, extruder temperature, and shape of printing objects can affect the particle emissions during 3D printing [26]. It has been shown that particle emissions during 3D printing using ABS was higher than that of PLA [2,17]. Studies regarding the emission of particles from 3D printing primarily focus on particle size distribution or particle number concentrations [2,6,8,27,28]. In contrast, only a few studies are available on the effect of metal additives on these thermoplastics [7,29–31]. These studies show that metal additives in filaments can lead to more particle generation during printing, which may cause adverse health effects to the respiratory and cardiovascular systems. Stabile et al. [29] reported that the particle emission rate for copper-infused PLA was 1–2 orders of magnitude greater than that of neat PLA (i.e., PLA filaments without metal additives), despite using the same printing temperature. Similarly, according to Erik et al. [32], particle emissions from ABS and PLA filaments with tungsten and copper additives were higher than their neat counterparts, even when printing at the same temperature.

Particle emissions during 3D printing may also contain environmentally persistent free radicals (EPFRs), which can add an additional layer of complexity to the potential health and environmental risks. EPFRs (i.e., organic free radicals stabilized on or inside particles) are typically associated with particulate emissions from thermal sources. They are found in significant concentrations in ambient air PM [33–37]. While a gas phase free radical is typically highly reactive, with less than a -millisecond lifetime, surface-bound EPFRs are stabilized by their association with thermally-generated PM and can exist for much longer timeframes, ranging from minutes to months [33,38–40]. EPFRs have been detected in combustion systems [41], coal and wood burning sites [36], and residues produced by various plastics [23,34,42,53]. It is proposed that EPFRs are formed by chemisorption of organic compounds at a metal oxide surface site, leading to electron transfer and formation of surface-associated EPFRs and the reduced metal [55]. As a result, the use of filaments containing metal additives could promote the production of EPFRs during the 3D printing process. Dellinger and colleagues observed that EPFRs are present in fine and ultrafine PM [51–54]. PM containing EPFRs has been demonstrated to produce cardiovascular and pulmonary dysfunction in animal models and exhibit greater toxicity as compared to PM with no EPFRs [43–46]. EPFRs generate ROS [47,48], which induce oxidative stress and can damage DNA, proteins and enzymes, leading to multiple health dysfunctions [49,50].

In the present work, we tested two ABS-type printing filaments from the same manufacturer with distinctly different metal content. To our knowledge, no data is available regarding the formation and presence of EPFRs on 3D printer emission particles, which constitutes a significant gap in understanding the potential health and environmental risks associated with 3D printing. As the presence of metals is one of the critical conditions for EPFR formation, this study can shed light on EPFR formation during 3D printing. This work aims to investigate the presence of EPFRs in TPM and examine the influence of filament additives (i. e., color pigments and metal) on particle emissions. TPM was collected, quantified, and investigated for the presence of EPFRs using electron paramagnetic resonance (EPR) spectroscopy. The materials and methods used for qualitative and quantitative analysis of EPFRs in the collected TPM are provided, including details on TPM sampling, EPR analysis, and lifetime (aging) analysis. Detailed mechanisms for the formation and decay of EPFRs and superoxide species are also given, as well as discussion of the potential human health effects of exposure to these radicals and precautions to take to reduce exposure during 3D printing.

2. Material and methods

2.1. 3D printer set up and thermoplastic filament

The printer used in this study was a LulzBot TAZ 6 (Fargo Additive Manufacturing Equipment 3D, LLC, Fargo, ND). The printer filament feedstock was fed to a narrow-heated nozzle at a constant speed, where it was melted and extruded through the nozzle. The nozzle was moving in two dimensions while the z-axis movement was turned off.

The printing program was set to print a cylinder (17.7 mm × 17.7 mm × 102.4 mm). The printing parameters were as follows: layer height 0.25 mm, infill density 20 %, printing speed 60 mm/s, and temperature 230 °C. The printing time for one cylinder was 4 h. ABS-black and ABS-blue filaments (1.75 mm) were purchased from the same manufacturer. The elemental composition of these filaments is presented in Table 1.

Table 1.

Elemental composition of filaments.

Elements	ABS-black(μg/g)	ABS-blue(μg/g)
Aluminium	9.88 ± 1.04	18.62 ± 3.14
Antimony	0.36 ± 0.17	0.20 ± 0.04
Arsenic	0.04 ± 0.01	0.04 ± 0.01
Calcium	17.17 ± 2.57	19.25 ± 1.60
Chromium	0.30 ± 0.09	0.25 ± 0.06
Copper	0.27 ± 0.13	19.66 ± 3.25
Iron	3.69 ± 0.23	4.41 ± 1.36
Lead	0.15 ± 0.01	0.17 ± 0.05
Magnesium	10.14 ± 1.17	17.28 ± 2.63
Manganese	0.06 ± 0.02	0.23 ± 0.13
Nickel	0.14 ± 0.03	0.14 ± 0.03
Phosphorus	8.34 ± 1.70	10.22 ± 0.57
Silicon	9.97 ± 1.71	12.88 ± 1.17
Strontium	0.12 ± 0.01	0.14 ± 0.01
Tin	0.06 ± 0.03	0.03 ± 0.01
Vanadium	0.13 ± 0.002	0.12 ± 0.02
Zinc	9.61 ± 6.61	8.03 ± 3.86

2.2. TPM sampling train

TPM emitted during 3D printing was captured using a sampling train (Fig. 1). During sampling, the sampling chamber was attached to the printing head, and the head was raised to a height of 215 mm and kept at a hanging position during printing. The sampling chamber around the head was designed in such a way as to vacuum-pull all fumes and particles released during printing through the sampling train. The actual collection cartridge was a 4 mm ID × 40 mm Suprasil tube filled with 100 mg of collection matrix, located instream of the gas flow, with all gases passing through the cartridge. The cartridge was designed and constructed at Louisiana State University, with an easily recoverable, non-reactive, bio-inert collection matrix (Cab-o-Sil, with 200 nm grain size) capable of achieving 100 % efficiency in particulate collection. The cartridge, packed with 100 mg of the collection matrix, was positioned vertically in a holding vial (10 mm × 60 mm) with the help of a septum to ensure a proper air flow path. As shown in Fig. 1, the sampling train pulled the gases surrounding the printing head through the sampling chamber with a negative pressure pull (600 mmHg) created by a dry vacuum pump (20 % vacuum conditions). Particles were collected during 24 h of printing (printing of 6 cylinders). The cartridge was weighed before and after the collection process to evaluate the mass of collected material. Emission values were given in μg/g of a printed object (mass of TPM/mass of used filament). After TPM collection, the entire cartridge was removed from the holder and transferred to a 10-mm EPR tube for EPR analysis.

Fig. 1.

Sampling diagram of PM collection on collection matrix (silica).

2.3. EPR analysis

After TPM collection, the samples were weighed and analyzed by EPR spectroscopy for the presence of EPFRs. EPR spectroscopy is the only analytical technique capable of detection and measurement of species with unpaired electrons, such as short-lived (e.g., ${}^{·}\text{O}\text{H}$) and long-lived (e.g., semiquinone) free radicals. Basically, there are three parameters used in EPR spectroscopy that allow for quantification of EPFRs in PM: spin concentration, used to calculate the level of EPFRs in PM, and g-factor and peak-to-peak line width (Δ H_p-p), used to distinguish the type of radicals present in the sample. All EPR spectra were recorded at room temperature using a Bruker EMX 10/2.7 EPR spectrometer (Bruker Instruments, Billerica, MA) with dual cavities and X-band. The instrument parameters used for radical signal measurement were as follows: a microwave frequency of 9.7 GHz, a microwave power of 2.0 mV, a modulation frequency of 100 kHz, a modulation amplitude of 4.0 Gauss (G), a center field of 3470 G, a receiver gain of 3.54 × 10⁴, a sweep time of 167.77 s, a time constant of 40.96 ms, and 3 scans. Each analysis was repeated two times. The radical concentration (spins) was calculated by comparing the signal peak area calculated from peak-to-peak line widths (Δ H_p-p)² multiplied by the relative signal intensity to that of a 2,2-diphenyl-1-picrylhydrazyl (DPPH) standard. The results were normalized to the mass of collected TPM and given in spins/g which corresponds to the number of radicals per gram of collected TPM. The values of the g-factor were determined by Bruker’s WINEPR Sim-Fonia 2.3 software.¹

2.4. Lifetime analysis

The lifetime of EPFRs is usually longer than that of free radicals, such that EPFRs stabilized on particle surfaces can be persistent in an atmospheric environment. Due to their persistency, they accumulate in the environment and can potentially be transported long range. Atmospheric aging of the collected TPM was done in the open air environment of the cartridge tube, with both ends of the cartridge open to the air. Aging was performed using a static approach (i.e., no air flow was purged through the cartridge containing the collected particles) and air exposure was based on a natural air diffusion process. Based on the EPFR concentration changes during aging, decay kinetic studies were conducted to determine the stability and 1/e lifetime (t_1/e) of radicals in the air. EPR spectra were measured at specified time intervals to determine the change in radicals concentration as a function of time for samples exposed to ambient air for 90 days. The relative radical concentration was plotted versus time. The lifetime of EPFRs was evaluated using the kinetic expression for first-order decay (Eq. 1), where $t_{1/e}$ corresponds to the point at which the radical concentration $\left(R/R_0\right)$ is equal to $1/e$ .

$$\ln \left(R/R_0\right)=-\text{k}t\hspace{0.17em}\hspace{0.17em}\text{and}\hspace{0.17em}\hspace{0.17em}{t}_{1/\text{e}}=1/\text{k}.$$ (1)

The rate constant $\left(\text{k}\right)$ was found from the slope of the correlation between the logarithm of radical concentration $\left(R/R_0\right)$ and time, and the $1/e$ lifetime was calculated.

3. Results and discussion

Metals and other polymer additives, such as pigments, can impact the emission of chemicals during heating [56,57]. Very little is known about such effects during 3D printing, although a few studies have examined particle emissions from 3D printers affected by additives, such as pigments or metals [2,7,29–31,58]. Erik et al. [32] reported that particle emissions were higher for filaments with metal additives compared to neat filaments (i.e., filaments without metal additives). We focused this study on two ABS polymer filaments, ABS-blue and ABS-black, that not only differ by pigment but also by varying metal content (see Table 1).

An average TPM emission rate of 11.28 μg/g from ABS-Blue and 7.29 μg/g from ABS-black was calculated from the two replicate analyses (Table 2). These analyses agree with Erik et al. [32] observations and indicate that metal additives may induce higher particle emission; Stabile et al. [29] had similar results. Possible mechanisms behind this observation include a depolymerization process, oxidation of polymer fragments, or gas phase coagulation of particles around the metal.

Table 2.

Summary of EPR spectral parameters of collected TPM.

Filament Type	Printing Time (hr)	g-factor	Δ Hp-p (G)	Average TPM Emission Rate (μg/g)	EPFR Concentration (spins/g)
ABS-blue	24	2.00405 ± 0.0006	17.30 ± 1.39	11.28 ± 0.85	(3.70 ± 0.77) × 1017
ABS-black	24	2.00106 ± 0.0001	3.14 ± 0.17	7.77 ± 0.75	(9.72 ± 1.90) × 1016

g-factor: proportionality factor; ΔH_p-p: peak-to-peak width.

EPFRs are stable and relatively unreactive radicals in the ambient air [59], which distinguishes them from common short-lived radicals such as hydroxyl, phenyl, vinyl and methyl radicals. EPFR stability is associated with resonance stabilization and a metal center. The resonance stabilized radicals can be detected by EPR, and g-factor values commonly used to identify carbon-centered or oxygen-centered radicals. In environmental matrices, the g-factor of EPFRs changes with metals and temperature. According to researchers, there are three types of EPFRs: oxygen-centered radicals (g > 2.004), carbon-centered radicals (g < 2.003), and carbon-centered radicals with an adjacent oxygen atom (g = 2.003–2.004) [59]. The oxygen-centered radicals are more stable in the environment, whereas carbon-centered radicals are more susceptible to oxidation in air. Formation of EPFRs is inherently associated with the metal centers present on the particulates and, therefore, typically involves interaction between organic decomposition byproducts and transition metals in particles to form free radical-particle conjugates (Fig. 2) [33,50,60].

Fig. 2.

General mechanism of EPFR formation on a metal oxide (M(II)O) surface.

The presence and concentration of EPFRs in TPM generated from 3D printing was tested using an EPR spectrometer. EPR parameters including the g-factor and peak-to-peak line width (Δ H_p-p) were used to distinguish the type of radicals present in PM. Observed EPFR concentrations on TPM (expressed as spins/g) and the TPM emission levels (expressed as μg/g) from both filaments are shown in Table 2. TPM collected during 3D printing with the ABS-blue filament revealed a broad paramagnetic signal with a center field of 3465.73 G, a peak-to-peak width (Δ H_p-p) of 17.30 G, and a g-factor of 2.00405 (Fig. 3A, Table 2). Although the signal strength was relatively weak, the signal parameters indicate the presence of an organic radical and in particular, points to an oxygen-centered phenoxyl type radical [61].

Fig. 3.

EPR spectra of TPM from 3D printing using (A) ABS-blue and (B) ABS-black filaments. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

In comparison, TPM collected during 3D printing using the ABS-black filament showed a low g-factor of 2.00106 and a Δ H_p-p of 3.14 G (Fig. 3B). The origin of this signal is not clear; however, it is possible that it is related to a superoxide radical generated on the iron oxide nanocluster emitted during heating of the filament. Formation of such superoxide species adsorbed on iron oxide nanoparticles, with stability at room temperature, was described previously by Herring et al. [62]. Iron, while known to form EPFRs, is less reactive compared to copper, though still able to create stabilized superoxide species [39].

Due to the nature of observed EPFRs, it is very important to understand their time-dependent behavior. After collection, TPMs emitted by both polymers were exposed to ambient air for aging studies (Fig. 4).

Fig. 4.

Aging of TPM radicals, collected during 3D printing versus time of exposure to ambient air.

Exposure of ABS-blue TPM to air resulted in a characteristic radical decay pattern and associated changes in the spectral g-factor. Fresh TPM emitted by ABS-blue contained 7.94 × 10¹⁷ EPFRs/g of TPM (expressed as spins/g) with a g-factor of 2.00390 – 2.00425. This type of spectra based on the g-factor value can be attributed to oxygen-centered phenoxyl type radicals (g-factor ~2.004 – 2.0045). However, given the spectral width is relatively broad (Δ H_p-p = 20.74 G), this is indicative of a more complex composition of EPFRs, among which the phenoxyl type species dominate; a contribution of carbon-centered radicals is also likely (g-factor ~2.003–2.0035). During aging in air, the radical concentration decreased with time, reaching its minimum value of 2.94 × 10¹⁷ spins/g in 90 days, with a g-factor of 2.00364 ± 0.0003. The shift in g-factor value during aging further supports a multicomponent EPFR composition, with some species decaying faster compared to others. Considering the change in g-factor, we concluded that the radical decay was associated with gradual oxidation/reaction of oxygen-centered EPFRs (such as phenoxyl), while carbon-centered EPFRs remained relatively unreactive. Based on the kinetic fit (Fig. 5) and kinetic description, the lifetime of ABS-blue EPFRs was found to be 98 days.

Fig. 5.

Kinetic decay fits for two types of TPM collected: ABS-blue (upper) and ABS-black (lower). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

A different aging profile was observed when exposing ABS-black TPM to air. Both Figs. 4 and 5 indicate two different kinetic regimes during the aging process – an initial increase of the EPFR concentration (from 1 × 10¹⁷ to 4.15 × 10¹⁷ spins/g between day 1 and day 60), followed by a relatively accelerated decay (4.15 × 10¹⁷ to 1.3 × 10¹⁷ spins/g) – with the g-factor (~2.00106 ± 0.0001) remaining relatively constant and a narrow signal (Δ H_p-p = 3.10 G). Such changes during the TPM aging process are consistent with the generation of new radicals during air exposure; the lack of major change in the g-factor value indicates the same type of radicals were generated during aging [63]. In fact, the observed g-factor values of TPM emitted from ABS-black are unlikely to be associated with an organic fraction, but rather are characteristic of the formation of surface-bound and stabilized superoxide species. Such superoxide species have been observed by others and are considered a lattice anion vacancy in small crystallites occupied by superoxide [64,65]. Considering the change in radicals concentration during aging, a single kinetic description of the EPFRs decay in TPM from ABS-black cannot be definitively provided (Fig. 5). The fit to the kinetic data is very poor over the entire time period (R² = 0.04). However, aging data can be separated into two regimes: radical formation and radical decay (Fig. 5). The radical generation time is characterized by a doubling time t₂ of 65 days. At this point, we can only speculate on the origin of the formation of superoxide EPFRs in these samples. Since, as indicated by elemental analysis, ABS-black contains metals and other inorganic elements (including Fe, Zn, Mg, Si and Al), it is likely that during the printing process and partial depolymerization [9], the emitted TPM contained nanocrystallites of oxides which, due to their size and high surface area, actively adsorbed oxygen. Over time, superoxides may have become incorporated into the lattice, causing a loss in radical character, which is manifested in gradual radical decay beyond 60 days of exposure. This final decay, although based on only two data points, can be assigned an “estimated tentative” lifetime of 26 days.

It is very apparent that TPM emitted from ABS-blue and ABS-black is very different – the former producing EPFRs of mostly phenoxyl type, while the latter generates surface-stabilized superoxides. The general composition of both filaments should be the same as they are both ABS polymers and produced by the same company. However, the elemental analysis of both filaments revealed some differences in the content of inorganic elements. In particular, ABS-blue had high levels of copper as compared to ABS-black (see Table 2). The differences in TPM and EPFR concentration and speciation can be correlated with the elemental composition differences of both filaments and specifically, the metal additives. ABS-blue contains ~20 μg/g of Cu, while ABS-black has a much lower Cu content (0.72 μg/g) and a composition of other metals comparable to that of ABS-blue, including Fe (~4 μg/g), Zn (~8 μg/g), and Ni (0.14 μg/g) (see Table 1). In general, EPFR formation is closely related to metal content [51,55,66–68], and copper is one of the most prominent metals in the formation of EPFRs [33,39]. When the presence of copper is lacking, such as in the case of the ABS-black filament, other metals become significant in respect to oxygen adsorption and other ABS components (such as nanocarbon) can also play a role.

4. Conclusions

The aim of the current study was to investigate the presence of paramagnetic species, in particular EPFRs, in TPM emitted during the 3D printing process and elucidate the factors contributing to their presence. In fact, we did detect the presence of paramagnetic species, either EPFRs or superoxides, both associated with the TPM surface; the type of species most likely dependent on the composition of the original filament used in printing. The data provide evidence that TPM emissions and EPFRs in TPM during 3D printing were higher in ABS-blue, a filament characterized by high copper content. These EPFRs displayed a typical phenoxyl type characteristic, with a lifetime of 98 days. The presence of oxygen-centered EPFRs in TPM generated during 3D printing with ABS-blue filament raises a serious concern regarding the potential exposure and health risks of using such filaments in small rooms or in facilities containing multiple professional devices. In the case of ABS-black, the observance of superoxide species formation is even more concerning, as very little is known on the health impacts of surface-associated superoxides. Judging by the general properties of superoxides, these species are potentially very dangerous. Exposure to high concentrations of ultra fine particles and EPFRs are associated with cardiovascular and respiratory system effects; therefore, it is prudent to take the following precautions to reduce exposures in schools, private homes and in public locations and universities where 3D printing processes take place: (a) 3D printers should be used in well-ventilated areas; (b) to minimize inhalation of emitted particles, the user should stay away from the printer while it is running (ideally outside of the printing room); and (c) a low emitting filament and printer should be chosen, if possible.

Future research on the life cycle of TPM generated during 3D printing and the potential formation of reactive oxygen species during the printing process can provide additional insight into the potential health effects of exposure to 3D printer emissions.

Data availability

Data will be made available on request.