ArticlePDF Available

Emissions of Nanoparticles and Gaseous Material from 3D Printer Operation

Authors:
  • National Institute of Chemical Safety - Ministry of Environment

Abstract and Figures

This study evaluated the emissions characteristics of hazardous material during fused deposition modeling type 3D printing. Particulate and gaseous materials were measured before, during, and after 3D printing in an exposure chamber. One ABS and two PLA (PLA1 and PLA2) cartridges were tested three times. For online monitoring, a scanning mobility particle sizer, light scattering instrument, and total volatile organic compound (TVOC) monitor were employed and a polycarbonate filter and various adsorbent tubes were used for offline sampling. The particle concentration of 3D printing using ABS material was 33-38 times higher than when PLA materials were used. Most particles were nanosize (<100 nm) during ABS (96%) and PLA1 (98%) use, but only 12% were nanosize for PLA2. The emissions rates were 1.61  1010 ea/min and 1.67  1011 ea/g cartridge with the ABS cartridge and 4.27-4.89 108 ea/min and 3.77-3.91x109 ea/g cartridge with the PLA cartridge. TVOCs were also emitted when the ABS was used (GM; 155 ppb, GSD; 3.4), but not when the PLA cartridges were used. Our results suggest that more research and sophisticated control methods, including the use of less harmful materials, blocking emitted containments, and using filters or adsorbents, should be implemented.
Content may be subject to copyright.
Emissions of Nanoparticles and Gaseous Material from 3D Printer
Operation
Yuna Kim,
Chungsik Yoon,*
,,
Seunghon Ham,
,
Jihoon Park,
Songha Kim,
Ohhun Kwon,
and Perng-Jy Tsai
§
Department of Environmental Health, School of Public Health, Seoul National University, 1, Gwanak-ro, Gwanak-gu, Seoul 151-742,
Republic of Korea
Institute of Health and Environment, School of Public Health, Seoul National University, 1, Gwanak-ro, Gwanak-gu, Seoul 151-742,
Republic of Korea
§
Department of Environmental and Occupational Health, Medical College, National Cheng Kung University, 138, Sheng-Li Rd.,
Tainan 70428, Taiwan
*
SSupporting Information
ABSTRACT: This study evaluated the emissions characteristics of
hazardous material during fused deposition modeling type 3D printing.
Particulate and gaseous materials were measured before, during, and after
3D printing in an exposure chamber. One ABS and two PLA (PLA1 and
PLA2) cartridges were tested three times. For online monitoring, a
scanning mobility particle sizer, light scattering instrument, and total
volatile organic compound (TVOC) monitor were employed and a
polycarbonate lter and various adsorbent tubes were used for oine
sampling. The particle concentration of 3D printing using ABS material
was 3338 times higher than when PLA materials were used. Most
particles were nanosize (<100 nm) during ABS (96%) and PLA1 (98%)
use, but only 12% were nanosize for PLA2. The emissions rates were
1.61 ×1010 ea/min and 1.67 ×1011 ea/g cartridge with the ABS cartridge
and 4.274.89 ×108ea/min and 3.773.91 ×109ea/g cartridge with
the PLA cartridge. TVOCs were also emitted when the ABS was used (GM; 155 ppb, GSD; 3.4), but not when the PLA
cartridges were used. Our results suggest that more research and sophisticated control methods, including the use of less harmful
materials, blocking emitted containments, and using lters or adsorbents, should be implemented.
INTRODUCTION
Three-dimensional (3D) printers are used in various industrial
sectors, including electronics, automotive, aerospace, medical
science, and education.
1
Several 3D printing techniques are
available, including fused deposition modeling (FDM), selective
laser sintering (SLS), and stereo lithographic apparatus (SLA). In
SLS printing, a laser beam strikes the heated powder and uses
continuously laminated molding to make a product. Metal
powder, ceramic powder, and thermoplastic powder have been
used as cartridges. In SLA printing, an ultraviolet laser strikes the
liquid material cartridge containing the photopolymer resin to
make a product. Two cartridges are required; one for the product
and another for the support material.
2
The FDM printer is popular because it is inexpensive and easy
to use compared to the SLS and SLA printers. In FDM printing, a
thermoplastic material, supplied as a wire, is heated at the
extrusion nozzle head and hardened immediately to form
successive layers as designed. The FDM method is similar to
thermoplastic extrusion, which can emit various hazardous
compounds depending on the types of plastic used.
3
Most FDM
3D printers use acrylonitrile-butadiene-styrene (ABS) and
polylactic acid (PLA) as ller material.
4
The major ABS
breakdown products are acrylonitrile, 1,3-butadiene, and styrene
at 160180 °C.
5
Acrylonitrile is toxic by inhalation and is
classied as Group 2B (possibly carcinogenic to humans) by the
International Agency for Research on Cancer (IARC). 1,3-
Butadiene is classied as Group 1 (conrmed human
carcinogen) by the IARC and may cause heritable genetic
damage. Styrene is known to be irritating to the skin and eyes.
6
PLA cartridges contain lactic acid, which is known to be an
environmentally friendly material, but further information is
required regarding the health hazards following exposure by
inhalation.
All 3D printing systems (FDM, SLS, SLA, or others) are
expected to be used more extensively in the near future. When
considering the energy (heat, laser beam, etc.) and raw materials
(metal powder, thermoplastics, etc.) used, hazardous agents
Received: June 8, 2015
Revised: August 19, 2015
Accepted: September 24, 2015
Published: September 24, 2015
Article
pubs.acs.org/est
© 2015 American Chemical Society 12044 DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
could be emitted into the air.
7
To our knowledge, only one
technical report on nanoparticle emissions during 3D printing
has been published.
4
There are also no published data for gaseous
emissions during 3D printing. The objective of this study was to
evaluate the emission of particulate matter and gaseous materials
during FDM 3D printing.
MATERIALS AND METHODS
Study Design. Two FDM 3D printers were used. The
printers had a similar principle of operation but were produced
by dierent manufacturers. Dierent cartridges were recom-
mended for use in the models. The brand A 3D printer (Cube,
3D Systems) was tested using both an ABS cartridge and a PLA
cartridge (PLA1). The brand B 3D printer (3DISON Plus, Rokit,
Korea) was tested with its own PLA cartridge (PLA2). Both
printers were open types and the extrusion temperatures of the
ABS and two PLA cartridges were 250 °C and 210220 °C,
respectively.
Sampling and Analysis. Particulate matter and gaseous
materials emitted during 3D printing were measured in a test
chamber. The test chamber (1 ×1×1 m) was constructed from
acryl. The only air withdrawn by the sampling instruments (total,
9.35 L/min) was supplied by ve ventilation holes, which were
located in the lower side of each wall. Air was ltered into the
chamber through a high eciency particulate arrestor (HEPA)
lter and a charcoal absorbent bed. Total volatile organic
compound (TVOC) and particulate concentrations were
measured inside the test chamber and the particulate level was
demonstrated to be below that outdoors (<4000 particles/cm3in
the test chamber vs 600020 000 particles/cm3outdoors). The
TVOC concentration was close to zero.
The operating time varied according to the cartridge type (2 h
30 min, 1 h 55 min, and 2 h 50 min for the ABS, PLA1, and PLA2
cartridges, respectively), although the printed products were of
identical size. The 3D printers were operated under the
conditions recommended by the manufacturers to produce a
mock-up bobbin (5 ×5×3 cm) (Figure 1).
The sampling period was divided into three blocks: before (1
h), during (operating time), and after the printing period (equal
to the operating time). For example, when using the ABS
cartridge, the periods 1 h of before operation, 2 h 30 min during
operation, and 2 h 30 min after operation were monitored. Each
cartridge was tested three times. Cartridge consumption during
each operation was measured by subtracting the weight of the
cartridge after use from the weight of the cartridge before use; the
results were 14.52 ±0.04 g, 14.90 ±0.14 g, and 18.58 ±3.22 g for
the ABS, PLA1, and PLA2 cartridges, respectively.
The 3D printer was located at the center of the test chamber.
Sampling cassettes or inlets of the online (real-time) monitoring
instruments were located in the upper part of the test chamber.
Both online and oine (integrated) monitoring was used to
measure particulate and gaseous materials. Online samples were
collected at 1 min intervals during the sampling period.
To measure the particulate concentration in the air with the
online method, three instruments were used to collect data at 1
min intervals. A scanning mobility particle sizer (SMPS, model
3910, TSI Inc.) with a detectable size range of 10 to 420 nm was
used to investigate the particle size distribution and number
concentrations. A condensation particle counter (CPC, P-Trak
model 8525, TSI Inc.) was used to measure the total particle
number concentration, with a range of 20 nm to 1000 nm and an
upper limit of 500 000 particles/cm3. The CPC data are not
presented here because the measurements during ABS cartridge
use were often over the upper limit. A DustTrak (DRX aerosol
monitor model 8533, TSI Inc.) with a size range of 0.1 to 15 μm
wasusedtomeasurethemassconcentration.Themass
concentration was recalculated using the C-factor. The C-factor
was calculated using an integral polyvinyl chloride (PVC) lter
placed within the instrument.
8
All real-time instruments were calibrated by the manufacturers
and a zero calibration using a HEPA lter was performed before
and after sampling.
Particle mass concentration was measured by oine
monitoring using a polycarbonate (PC) lter (37 mm, 0.4 μm,
SKC Inc.) in an open-face three-piece cassette using a high
volume pump (Escort ELF, MSA) at a constant ow rate of 2.0
L/min. After sampling, the cassettes were sealed using silicon
tape and kept in a desiccator in a weighing room (Temp 20 °C±
1°C, RH 50% ±5%). A microbalance (XP6 microbalance,
Mettler-Toledo, Switzerland), with a sensitivity of 1 μg, was used
for gravimetric analysis.
Particle emissions rates during printer operation were
calculated as follows. Because the air in the chamber was
withdrawn only through a sampling lter or sampling tube placed
20 cm above the upper part of the extrusion head of the 3D
printer, we assumed all pollutants passed through the sampling
lter or sampling tube because the small gap around the sampling
tube was tightly sealed with a sealant.
Emissions rates were calculated based on time (particles per
min, ea/min) and cartridge use (particles per amount of cartridge
used, ea/g cartridge used), as demonstrated below. An example
using the ABS data in Table 1 is shown in parentheses:
Correction of before operationconcentration: to
measure only particles emitted during 3D printer
operation, the before operationconcentration was
subtracted from the during operationconcentration of
particles.
−=
(
1 731 578e/cc 5, 021ea/cc 1 726 557ea/cc)
(1)
Calculation of particles entering the SMPS instrument per
min: the sampling ow rate of the SMPS was 0.75 L/min.
Therefore, particles entering the SMPS instrument per
min was calculated by multiplying the concentration
corrected above (1) by the 0.75 L/min SMPS ow rate.
The result was multiplied by 1000 to adjust for volume
Figure 1. Image of the printed bobbins (left, using an ABS cartridge;
right, using a PLA cartridge).
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12045
××
=
(
1 726 557ea/cc 0.75L/min 1000cc/L
1 298 638 500ea/min in the SMPS) (2)
Calculation of emissions rates based on time: the total
sampling rate of all sampling instruments was 9.35 L/min.
Therefore, emissions rates were calculated by the
proportional equation of ow rate.
=
=
≑×
(
1 298 638 500ea/min in the SMPS
: 0.75L/min SMPS flow rate emissions rate
: 9.35L total flow rate, therefore, the emissions
rate of the ABS cartridge 16 190 254 300ea/min
1.62 10 ea/min)
10 (3)
Calculation of emissions rate based on the amount of
cartridge used: the emissions rate based on time was
multiplied by the operation time to obtain the total
emissions during operation (16 190 254 300 ea/min ×
150 min = 2.43 ×1012 ea/min), and then divided by the
cartridge mass consumed
×÷
(
2.43 10 ea/min 14.52g of cartridge use
d
1.67 10 ea/g cartridge)
12
11 (4)
To measure the TVOCs, a ppbRAE monitor (ppbRAE 3000,
RAE Systems Inc.) was used for online sampling. Before
sampling, the ppbRAE was calibrated using 10 ppm isobutene
standard gas (CALGAZ Inc.), and zeroed with an adsorbent tube
attachment.
Airborne aldehydes were measured using the National
Institute for Occupational Safety and Health (NIOSH) method
2016. Aldehydes were sampled using a 2,4-dinitropenylhydrazine
(2,4-DNPH) cartridge (Supelco), with an ozone scrubber
(WAT054420, Waters Sep-Pak) using a low-volume pump
(Gillian LFS-114), with a constant ow rate of 0.2 L/min.
9
After sampling, all samples were sealed using aluminum foil to
protect against light exposure and kept below 4 °C until
desorption.
To desorb aldehydes, solvent extraction equipment (Vacuum
Elution Rack, Supelco) was used. Aldehydes were slowly eluted
with 5 mL of acetonitrile (ACN, high pressure liquid
chromatography (HPLC) grade, J.T.Baker). All of the
experimental equipment used for analysis was cleaned using
ACN and baked at 6080 °C. Extracted aldehydes were analyzed
using HPLC (Ultimate 3000, Thermo Scientic) with an
ultraviolet (UV) detector (Ultimate 3000 Variable Wavelength,
Dionex, Germany) and an autosampler (Ultimate 3000, Dionex).
A C-18 column (Eclipse XDB, Agilent Technologies) was used
for analysis; the column was 4.6 m in length, 250 mm in inner
diameter, and 5 μminlm thickness; the wavelength of the
detector set to 360 nm. ACN (60%) and water (40%; HPLC
grade, J.T.Baker) was used as a carrier liquid, with the ow rate
set to 1.0 mL/min. The carrier liquid was ltered before analysis.
The injection volume was 20 μL and the column temperature
was 25 °C. An aldehyde standard solution (T011/IP-6A
Aldehyde/Ketone-DNPH mix, Supelco) was used in the analysis.
Field blank and spiked samples were analyzed simultaneously for
quality control. The limit of detection (LOD) was estimated
from seven replicates of the lowest standard solution. The LOD
of the aldehyde compounds is presented in Supporting
Information Table 1.
Benzene, toluene, ethylbenzene, and m-, p-xylene (BTEX)
were measured according to a standard NIOSH method using an
adsorbent charcoal tube (cat. no. 226-01, SKC Inc.) with a ow
rate of 0.2 L/min.
10
Before analysis, charcoal tubes were
extracted with 1 mL of carbon disulde (CS2). The extraction
sample (1 μL) was injected by autosampler (CombePAL, CTC
Analytics, Switzerland) into a gas chromatograph (HP 6890,
Hewlett-Packard Co.) - mass spectrometer (HP 5975C, Agilent
Technologies) (GC-MS). A capillary column (DB-5 ms, Agilent
Technologies) was used for the analysis (30 m ×0.25 mm ×0.25
μm). The GC-MS was programmed with a 1.60 min solvent
delay, and an initial temperature of 35 °C for 3 min, rising by 10
°C/min to 100 °C. The selected-ion monitoring (SIM) mode
was set for m/z78 (benzene and m-, p-xylene) and 91 (toluene).
Field blank and spiked samples were analyzed simultaneously for
quality control and the LOD was estimated from seven replicates
of the lowest standard solution.
Table 1. Particle Concentrations Determined by SMPS Before, During, And After 3D Printing, And Emissions Rates During 3D
Printing
SMPS (size range: 10420 nm) emissions rates during operation
cartridge before operation during operation after operation outdoors ea/min ea/g cartridge used
ABS
c)
GM (GSD)
a
5021 (1.46) 1 731 578 (1.47) 6373 (2.31) 20 363 (1.63) 1.61 ×1010 1.67 ×1011
range 244723 993 76613395,120 25151597,471 844746 616 7.16 ×1073.17 ×1010 7.40 ×1083.28 ×1011
CMD (nm)
b)
67.9 (1.18) 32.60 (1.11) 50.49 (1.07) 54.74 (1.26)
ratio
f)
0.655 0.957 0.887 0.768
PLA1
d)
GM (GSD) 1997 (1.28) 52 252 (1.98) 1374 (1.81) 6222 (1.17) 4.89 ×1083.77 ×109
range 13054612 2961212 154 53511 988 363412 306 2.77 ×1071.98 ×1092.14 ×1081.53 ×1010
CMD (nm) 44.46 (1.07) 27.94 (1.14) 43.37 (1.08) 40.92 (1.10)
ratio 0.794 0.975 0.844 0.812
PLA2
e)
GM (GSD) 2174 (1.20) 45 690 (2.50) 1582 (1.55) 5939 (1.24) 4.27 ×1083.91 ×109
range 3743497 2469474 000 1283-21 711 143111 138 2.31 ×1074.43 ×1092.11 ×1084.06 ×1010
CMD (nm) 84.44 (1.15) 188.18 (1.26) 82.90 (1.11) 59.64 (1.16)
ratio 0.543 0.119 0.553 0.708
a
Geometric mean (geometric standard deviation): unit (particles/cm3).
b)
Count median diameter, geometric mean (geometric standard deviation).
c)
Acrylonitrile-butadiene-styrene (ABS) (Cube, 3D Systems).
d)
Polylactic acid (PLA1) (Cube, 3D Systems).
e)
Polylactic acid (PLA2) (3DISON
plus, Rokit, Korea).
f)
Ratio of 100 nm to total number concentration.
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12046
Five phthalates were collected and analyzed according to an
Occupational Safety and Health Administration (OSHA)
standard method.
11
Five substances were selected as phthalate
components that can be emitted during 3D printer operation:
(1) dimethyl phthalate (DMP) (2) diethyl phthalate (DEP) (3)
dibutyl phthalate (DBP) (4) di-2-ethylhexyl phthalate (DEHP),
and (5) di-n-octyl phthalate (DnOP). Phthalates were collected
using XAD-2 tubes (cat. no. 226-30-06, SKC Inc.) with a ow
rate of 1.0 L/min. Sampled phthalates were extracted with 4 mL
of a hexane and acetone mixture (9:1). All of the equipment used
in the analysis was cleaned using ACN. Extracted samples were
injected by the autosampler into the GC-MS. A capillary column
(DB-5 ms, Agilent Technologies) was used for the analysis (30 m
×0.25 mm ×0.25 μm). The GC-MS was programmed with a 5
min solvent delay, and an initial temperature of 60 °C for 5 min,
increasing by 10 °C/min to 100 °C, 20 °C/min to 220 °C, and
then 5 °C/min to 300 °C for 3 min. The SIM mode was set for
m/z149 (DEP, DBP, DEHP, and DnOP), and 163 (DMP).
12
Field blank and spiked samples were analyzed simultaneously for
quality control and the LOD was estimated from seven replicates
of the lowest standard solution.
Statistical analyses were performed using the SAS 9.13
software (SAS Institute Inc.). For the statistical analysis of
variance we used a post hoc Tukey test to determine any
signicant dierences between the cartridges, for which we also
used Excel 2010 (Microsoft Corp, USA). Online instrument data
were presented as geometric means (GM) (geometric standard
deviation [GSD]), and oine data were collected from three
samples and presented as arithmetic means (AM) (standard
deviation [SD]). A graphical representation was generated using
SigmaPlot 10.0 (Systat).
RESULTS
Number Concentration of Particles. The number
concentration, range of concentration, count median diameter
(CMD) (unit: nm), fraction of nanosize particles (<100 nm),
and emissions rates are summarized in Table 1. During 3D
printer operation, the GM (GSD) number concentration for the
ABS cartridge, as measured by the SMPS was 1 731 578 (1.47)
particles/cm3, while the values for the PLA1 and PLA2 cartridges
were 52 252 (1.98) particles/cm3and 45 690 (2.50), respec-
tively.
For all cartridges, the number concentration of particles was
considerably higher during 3D printing than before operation,
after operation, and the outdoor concentration. When using the
ABS cartridge, the GM of the number concentration during 3D
printing (1 731 578 ea/cm3), as measured by the SMPS, was 345
times higher than before operation (5021 ea/cm3), 273 times
higher than after operation (6373 ea/cm3), and 85 times higher
than that in outdoor air (20 363 ea/cm3).
The same trend was found for the PLA1 and PLA2 cartridges.
The GMs during 3D printing were 26 times (PLA1; 52 252 ea/3)
and 21 times (PLA2; 45 690 ea/cm3) higher than before printing
(PLA1; 1997 ea/cm3, PLA2; 2174 ea/cm3).
The particle emissions rates based on both time and cartridge
use (g) for the ABS cartridge were 2 orders of magnitude higher
Figure 2. SMPS output for the various size categories before, during, and after 3D printer operation. (a) ABS cartridge (b) PLA1 cartridge (c) PLA2
cartridge.
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12047
than for the PLA cartridges. The emissions rates from ABS
cartridge use were 1.61 ×1010 ea/min based on time and 1.67 ×
1011 ea/g based on cartridge use.
The emissions rates based on time and cartridge use were of
the same order of magnitude for PLA1 and PLA2. The rates were
4.89 ×108ea/min and 3.77 ×109ea/g cartridge for PLA1, and
4.27 ×108ea/min and 3.91 ×109ea/g cartridge use for PLA 2.
The outdoor concentration measured by SMPS varied. During
ABS cartridge sampling, the GM (GSD) of the number
concentration was 20 363 (1.63) particles/cm3as measured by
the SMPS and 6222 (1.17) particles/cm35939 (1.24) particles/
cm3as measured by the CPC.
Figure 2 shows the SMPS output during sampling. In the
gure, the red line indicates the duration of printer operation.
During 3D printer operation, the number concentration rapidly
increased for all cartridges. The use of the ABS cartridge
produced the highest number concentration, whereas the PLA1
cartridge emitted the smallest particles. Particles of less than 100
nm were produced rapidly when the ABS and PLA1 cartridges
were used, whereas the use of the PLA2 cartridge produced
particles larger than 100 nm.
The data measured by the SMPS are presented as the GM
(GSD) of the CMD in Figure 3(a). For all cartridges, before and
after operation, the particle size was less than 100 nm. During 3D
printing, the CMD using the ABS was 33 (1.11) nm, whereas for
the PLA1 and PLA2 cartridges it was 28 (1.14) nm and 188
(1.26) nm, respectively. The 3D printing using the ABS and
PLA1 cartridges produced nanoparticle emissions.
Figure 3. (a) The number concentrations according to the count median diameter (CMD) measured by SMPS. Box plots indicate the median (line
within box), upper quartile and lower quartile (top and bottom part of the box, respectively), the maximum excluding outliers and the minimum
excluding outliers (upper and lower bars on whiskers, respectively), and outliers (black circles). (b) Mean dierence in the CMD between the cartridges.
The dierence between means (black circles), and the 95% condence intervals (upper and lower bars on whiskers) are indicated.
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12048
Figure 3(b) shows the results of an analysis of variance using a
post hoc Tukey test to determine any dierences between the
cartridges (ABS, PLA1, and PLA2). Depending on the cartridges,
printers, and testing conditions, statistically signicant dier-
ences in the CMD of particles were observed between cartridges
(p< 0.05). Before and after operation there was a similar trend,
while Figure 3(b) shows that during operation particles
produced using the PLA2 (C) cartridge had the highest CMD,
whereas particles produced using the PLA1 (B) cartridge had the
lowest CMD.
Figure 4 shows the number concentration measured by SMPS.
As soon as 3D printing began, the total number concentration of
particles, which ranged from 10 to 420 nm, increased rapidly with
all cartridges. The number concentration with the ABS cartridge
was consistently much higher than for the PLA1 or PLA2
cartridges. The total number concentration of particles when
using the PLA1 and PLA2 cartridges displayed similar lower
trends. After 3D printer operation nished, the particle
concentration decreased rapidly.
Mass concentration of Particles. Table 2 shows the total
mass concentrations of the particles emitted for each cartridge
using online monitoring by DustTrak. During 3D printer
operation, the GMs (GSD) of the particles emitted for each
cartridge were: ABS; 63.74 (1.10) μg/m3, PLA1; 31.89 (1.01)
μg/m3, and PLA2; 153.20 (1.69) μg/m3. The mass concentration
of particles using the PLA2 cartridge was the highest, and there
were statistically signicant dierences between the levels before
and during operation for all cartridges. Figure 5 shows a
comparison of the mass concentration of particles emitted with
each cartridge (ABS, PLA1, and PLA2). Statistical analysis of
variance was conducted using a post hoc Tukey test to identify
any signicant dierences between the cartridges. During printer
operation, the mass concentration of particles during use of the
PLA2 cartridge was higher than for the other cartridges (p<
0.05).
The data measured by the PC lter are presented in Figure 5.
The mass concentrations measured during use of the ABS and
PLA2 cartridges increased during printer operation, whereas for
PLA1 the concentrations were similar to those before operation.
During operation, the mass concentrations were approximately
1.9 times and 2.1 times higher than before operation when the
ABS and PLA2 cartridges were used, respectively. After
operation, the mass concentration decreased compared to the
levels before and during operation for all three cartridges.
Gaseous Materials. Table 3 shows the TVOC and BTEX
concentrations during sampling. When the PLA1 and PLA2
cartridges were used, TVOCs were not detected. When the ABS
cartridge was used, during 3D printing the average GM (GSD)
and maximum TVOC concentrations were 154.9 (3.4) ppb and
453.3 ppb, respectively. The TVOC concentration rapidly
Figure 4. Real-time monitoring by SMPS during 3D printing with the
use of ABS, PLA1, and PLA2 cartridges (particle size ranged from 10 to
420 nm).
Table 2. Total Mass Concentration As Measured by Dusttrak
(Real-Time Monitoring) Before, During, And After 3D
Printer Operation
concentration (unit: μg/m3)
cartridge GM (GSD)
a
range
ABS before operation 58.34 (1.06) 54.2367.88
during operation 63.74 (1.10) 58.0291.40
after operation 51.93 (1.06) 49.6164.51
PLA1 before operation 32.32 (1.01) 31.9034.19
during operation 31.89 (1.01) 31.6133.36
after operation 32.10 (1.03) 31.6438.57
PLA2 before operation 11.19 (1.01) 11.1211.60
during operation 153.20 (1.69) 11.43387.43
after operation 11.33 (1.20) 10.4347.34
a
Geometric mean (geometric standard deviation).
Figure 5. Particle mass concentrations measured using a PC lter. Error bars indicate the standard deviation of the mean (n= 3).
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12049
decreased after 3D printer operation, but some TVOCs were still
detected 15 min after printing nished (Figure 6).
Benzene concentrations were below the LOD for all cartridges.
The toluene, ethylbenzene, and m- and p-xylene concentrations
increased when the ABS and PLA2 cartridges were used. When
the PLA1 cartridge was used, the concentration during printer
operation was lower than that before operation. During printing,
the highest ethylbenzene concentration (16.4 times higher than
the outdoor concentration) was recorded when the ABS
cartridge was used. When the PLA2 cartridge was used, m- and
p-xylene were detected at higher concentrations (2.9 times
higher than those outdoors) than when the other cartridges were
used.
Table 4 shows the concentrations of aldehydes and phthalate
when each cartridge was used. During 3D printing operation, the
concentrations of formaldehyde, acetaldehyde, and isovaleralde-
hyde increased. Formaldehyde was the most abundant aldehyde
when the PLA2 cartridge was used, whereas acetaldehyde and
isovaleraldehyde were detected when the ABS cartridge was
used.
During operation, the indoor concentration of aldehydes was
higher than that outdoors, except for isovaleraldehyde when the
PLA1 cartridge was used. When the ABS cartridge was used,
isovaleraldehyde and acetaldehyde concentrations were 11.9
times, and 3.2 times higher, respectively, than the outdoor
concentration. When the PLA2 cartridge was used, formaldehyde
concentrations were 5.2 times higher than the outdoor
concentration.
The phthalate concentrations were almost at the LOD and all
outdoor samples were below the LOD, but DEP, DBP, and
DEHP were detected with the ABS cartridge; DEP and DEHP
were detected with the PLA1 cartridge; and DBP was detected
with the PLA2 cartridge. DnOP and DMP were below the LOD
for all cartridges. These results show that the concentrations of
the phthalates that were present at levels above the LOD
increased during printing. When the PLA1 cartridge was used,
the concentrations of DEP and DBP after printing were slightly
higher than during operation.
DISCUSSION
We determined the particulate and gaseous emissions from 3D
printing, and investigated the impact of various 3D printer
cartridges on emissions. Our results conrm that FDM 3D
printing can be hazardous due to the high concentrations of
emitted nanoparticles and aldehydes, including formaldehyde,
which is carcinogenic to humans (Group 1); acetaldehyde and
isovaleraldehyde; some phthalates such as DEP, DBP, and
DEHP; and VOCs such as toluene and ethylbenzene.
Various characteristics of nanoparticles, including the size and
morphology, are known to inuence their toxicity, with small
particles being more toxic than large particles.
1315
Nano-
particles can penetrate to the alveoli upon inhalation. Nano-
particles have large surface area and can transfer toxic agents to
blood vessels and tissue cells, leading to potential health
hazards.
16
Only one study was available for comparison of the emissions
rates of particles during 3D printing.
4
The emissions rate using
the ABS cartridge (1.61 ×1010 ea/min) in this study was 1 order
of magnitude lower than that of the previous study (1.9 ×1011
ea/min); that using the PLA cartridge was approximately 2
orders of magnitude lower than in the previous study (4.274.89
×108ea/min vs 2.0 ×1010 ea/min). These dierences may be
due to several reasons including the dierent brands of 3D
Table 3. TVOC and BTEX Concentrations Before, During, And After 3D Printer Operation
concentration (ppb)
a
ABS PLA1 PLA2
substance before
operation printer
operation after
operation outdoors before
operation printer
operation after
operation outdoors before
operation printer
operation after
operation outdoors
TVOCs ND ND ND ND ND ND ND ND ND
ND 154.9 (3.4) 57.9 (5.1) ND ND ND ND ND ND
ND ND ND ND ND ND ND ND ND
VOCs benzene < LOD
b
< LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD
toluene < LOD 3.7 (5.5) 2.5 (3.5) < LOD 29.5 (28.2) 16.2 (15.3) 20.5 (25.0) 4.9 (7.6) 1.5 (1.6) 2.7 (3.9) 10.6 (9.3) 1.3 (1.0)
ethylbenzene < LOD 11.5 (7.0) 2.0 (1.2) 0.7 (0.6) 1.5 (1.0) 0.8 (0.5) 1.0 (0.5) 0.6 (0.2) < LOD 1.2 (0.7) 1.6 (1.0) < LOD
m-, p-xylene < LOD < LOD 0.6 (0.2) < LOD 1.5 (1.0) 0.8 (0.4) 0.9 (0.4) < LOD < LOD 1.3 (0.8) 1.8 (1.1) < LOD
a
Arithmetic mean (standard deviation).
b
Limits of detection (ppb): 21.29 (benzene), 0.52 (toluene), 0.45 (ethylbenzene), 0.45 (m-, p-xylene). ND, not detected.
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12050
printers and cartridges used, the dierent environments in which
tests were conducted (a small 1 m3chamber in this study vs a
large 45 m3oce room in the previous study), and dierent
calculation methods (direct calculation in this study vs model
estimation in the previous study); the same SMPS model was
used in both studies.
We also calculated emissions rates based on the amount of
cartridge used (particle numbers/g cartridge used) as well as
emissions rates based on time (ea/min). Emissions rates based
the amount of cartridge used can provide useful information for
comparing dierent cartridge products once a good deal of data
are accumulated. Interestingly, in this study the emissions rate
varied widely with the same cartridge, as shown in Table 1. For
example, the emissions rate with the ABS cartridge use varied
from 7.16 ×107to 3.17 ×1010 during printing. This seems
reasonable when we examine Figure 4, which shows a high peak
at the beginning of printing and a gradual decrease thereafter. In
the previous study,
4
they assumed constant emissions during
printing and the range of their emissions rate was narrow (i.e., 1.8
×10112.0 ×1011 with ABS cartridge use).
There could be some uncertainty when calculating the
emissions rates of particles. We assumed that all particles emitted
from the 3D printer were withdrawn evenly regardless of the
sampling lter and sampling tube, and we assumed isokinetic
sampling, however, this was never veried. We adapted this
calculation method from a previous study on welding fumes.
17
However, all fumes generated during welding in a welding fume
chamber are collected with only one large lter, so no assumption
of even distribution or isokinetic sampling is necessary. If there
was leakage or unsampled particles inside the chamber, the
calculated emissions rate might have been underestimated.
Despite the sampling process being conducted in a test
chamber, the outdoor conditions and concentrations of the
target materials could possibly impact the test results, despite use
of a charcoal and HEPA lter to lter the air entering the
chamber. There were several possible leak points where sampling
tubes were connected. Prior to sampling, the background particle
concentration was detected by the real-time instruments.
Therefore, measurements of the 3D printer emissions only
began when the mean particle number concentration in the
chamber, as measured by the real-time instruments, was below
5000 particles/cm3, which was much lower than in the outdoor
air.
To determine the mass concentration of particulate matter, a
DustTrak online monitoring instrument and PC lters, tradi-
tional oine sampling media, were used. Online and oine
results exhibited the same trend, but did not match exactly
because the former was for real-time monitoring and the latter
was for integrated sampling.
A ppbRAE monitor was used to measure TVOCs, and at the
same time the SMPS and CPC were used to measure the number
concentration of particulate matter. The use of isopropyl alcohol
in the CPC could have aected the ppbRAE monitor when
measuring the TVOC concentration, although the CPC
instrument was located outside the test chamber and the
adsorbent was attached to the hole of the test chamber. The test
chamber was made with an acrylic box and located inside the
laboratory. The chamber was ventilated before and after
sampling. The laboratory conditions might have resulted in
natural airow into the chamber due to the suction of inside air
with simultaneous sampling instrument operation. In this study,
we performed the experiment after verifying that the TVOC
concentration in the test chamber was zero, but the TVOC
instrument does not measure all tested pollutants.
Some VOCs including BTEX have been reported to be
emitted during thermal processing of plastics.
57
During 3D
printing, high temperatures of 210250 °C are applied to melt
the plastic in the feed cartridges. The TVOC concentration
increased when the ABS cartridge was used, as shown in Figure 6.
In contrast, the BTEX concentration was very low or below the
LOD, as shown in Table 3, although slightly higher
concentrations of ethylbenzene and toluene were measured
with ABS cartridge use compared to before or after printer
operation. A more detailed study is required to analyze thermo
Figure 6. Real-time monitoring of TVOC concentrations during sampling.
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12051
decomposition products including acrylonitrile, 1,3-butadiene,
and styrene.
The XAD-2 used for phthalate analysis might have a loss of
sample during the desorption process. DEHP and DBP are also
commonly used laboratory tools during sampling and analytical
work.
18
Therefore, the phthalate concentration could have been
underestimated.
Outdoor sampling was performed near the ventilation opening
of a laboratory building. Therefore, the concentrations of VOCs,
phthalates, and aldehydes were expected to be higher than the
general outdoor concentration, and the outdoor concentration
was inuenced by ne dust in the air. The concentrations of all
measured pollutants before 3D printing operation in the test
chamber were lower than outdoor levels, except for form-
aldehyde (Table 4). The formaldehyde concentration before 3D
operation ranged from 26.4 to 111 ppb, while the outdoor level
ranged from 21.3 to 30.2 ppb. However, it was clear that
formaldehyde was emitted during 3D printer operation because
its level increased during operation, was higher than before
operation and in outdoor air, and then decreased after operation.
There was little dierence between the structures of each
cartridge and the cartridge nozzle temperature setting con-
ditions. The same product was produced during each experi-
ment, but the printing time diered depending on the cartridge
and manufacturer. The emission characteristics diered based on
the use of dierent cartridges. The use of the ABS cartridge
resulted in the highest number concentration of particles,
whereas when the PLA1 cartridge was used the lowest particle
size and a low number concentration were recorded. The use of
the PLA1 and PLA2 cartridges produced similar number
concentrations, but with the PLA2 cartridge the particle size
was larger.
In this study, some measurement data obtained before
operation had unexpectedly high values even though the
exposure chamber was cleaned between experiments and an
HEPA lter was used during operation. For example, the particle
mass concentrations measured by lters before the ABS and
PLA2 cartridge tests (Figure 5), toluene before the PLA1 test
(Table 3), and formaldehyde concentrations before all cartridge
tests (Table 4) were relatively high. It is not clear why these
materials were detected at high levels but one partial explanation
is that there may have been preexisting pollutants in the inside air
before the tests. Between each experiment, the 3D printer was
replaced and the operating conditions were veried by printing
for a very short time. After checking for normal operation, the
printer was turned oand the inside surface of the chamber was
cleaned with wet tissues. After cleaning, we plugged the air inlet
holes with an HEPA lter and an adsorbent charcoal tube.
Therefore, the preexisting air in the exposure chamber was not
cleaned out before sampling, whereas during the test only ltered
air entered the chamber. Also, large particles could have been
emitted from the surface of the 3D printers or some particles may
have been produced during the setup process. Despite these
limitations, it is clear that concentrations of small particles such as
SMPS, TVOCs, BTEX, aldehydes except formaldehyde, and all
phthalates were very low or below the LOD before operation,
while most of the concentrations increased during operation.
Although it was not low before operation, the formaldehyde
concentration increased during operation compared to before
operation.
This study has several limitations that should be addressed in
future research. First, only two types of 3D printer (both FDM
types) and two cartridge types (ABS and PLA) were tested. More
Table 4. Aldehyde and Phthalate Concentrations during Use of the 3D Printer Cartridges
concentration (ppb)
a
ABS PLA1 PLA2
substance before
operation printer
operation after
operation outdoors before
operation printer
operation after
operation outdoors before
operation printer
operation after
operation outdoors
aldehydes formaldehyde 39.3 (6.6) 68.0 (12.7) 43.8 (21.2) 21.3 (10.7) 26.4 (16.4) 54.0 (23.3) 40.4 (6.0) 23.9 (13.0) 111.0 (59.1) 155.9 (24.3) 114.7 (41.7) 30.2 (5.7)
acetaldehyde < LOD2) 31.9 (6.2) 11.6 (1.7) < LOD 11.2 (2.1) 30.4 (5.0) 10.8 (1.5) 10.9 (1.7) < LOD 18.4 (14.7) < LOD < LOD
isovaleraldehyde < LOD 90.8 (22.5) 37.6 (17.2) < LOD < LOD < LOD 14.6 (12.0) < LOD < LOD 27.2 (33.9) < LOD < LOD
phthalates DMP < LOD
b
< LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD
DEP < LOD 2.2 (0.8) 1.0 (0.6) < LOD < LOD 0.9 (0.4) 1.0 (0.4) < LOD < LOD < LOD < LOD < LOD
DBP < LOD 0.7 (0.2) < LOD < LOD < LOD < LOD 1.5 (1.2) 1.0 (0.4) < LOD 2.7 (3.2) < LOD < LOD
DEHP < LOD 1.4 (0.3) < LOD < LOD < LOD 1.4 (0.2) < LOD < LOD < LOD < LOD < LOD < LOD
DnOP < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD < LOD
a
Arithmetic mean (standard deviation).
b
Limits of detection (ppb): 0.40 (DMP; dimethyl phthalate), 0.67 (DEP; diethyl phthalate), 0.81 (DBP; dibutyl phthalate), 1.25 (DEHP; di-2-ethylhexyl
phthalate), 0.75 (DnOP; di-n-octyl phthalate).
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12052
information could be obtained from a wider range of 3D printers
and cartridges, although FDM printers are popular and are
usually used with ABS or PLA cartridges. Second, sampling was
conducted in a chamber, and no exposure monitoring was
undertaken.
This study did not examine worker or occupant exposure, but
rather the emitted concentrations measured at the top of the
extrusion nozzle head of a 3D printer in a test chamber. Because
of the high temperature (>200 °C) at the extrusion head, most
emissions were rising upward, directly to the inlet head of each
sampling instrument. Therefore, the analytes in this study had
little time to interact with the chamber wall to substantially
inuence the measurement results. We present our results as
emissions rates, which can be compared to other studies.
19
It is
possible that some emissions of measured contaminants were not
sampled because of the xed ow rate of the sampling
instruments. However, all experiments were performed within
the same chamber with the same ventilation rate, and our results
show that 3D printers should be used with caution.
In conclusion, we have conrmed that concentrations of
particulate and gaseous materials increase during FDM 3D
printing. The concentrations of particulate matter and gaseous
materials diered depending on the cartridge type and
manufacturer. Nanosize particles were emitted at high
concentrations regardless of cartridge type, and several aldehydes
including carcinogenic formaldehyde, some phthalates, and
VOCs such as toluene and ethylbenzene were emitted. Use of
the ABS cartridge resulted in a higher particle concentration and
emissions rate than use of the PLA cartridge.
Many expect 3D printing to be popular in the near future. Our
results suggest that more research and sophisticated preventive
control methods, including using less harmful materials, blocking
emissions, and using lters or adsorbents, should be
implemented to protect the health of users including susceptible
populations such as the elderly, pregnant women, and children.
ASSOCIATED CONTENT
*
SSupporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.est.5b02805.
Target aldehydes from standard solution and limit of
detection (LOD) (PDF)
AUTHOR INFORMATION
Corresponding Author
*Phone: +82-2-880-2734; fax: +82-2-745-9104; e-mail:
csyoon@snu.ac.kr.
Notes
The authors declare no competing nancial interest.
ACKNOWLEDGMENTS
This work was supported by the Basic Science Research Program
through the National Research Foundation of Korea (NRF)
funded by the Ministry of Education, Science and Technology
(No. 2011-0002926) and BK21 Plus project (No. 5280-
20140100) of National Research Foundation of Korea (NRF).
REFERENCES
(1) Kim, G. 3D Printing Industry: Technology, market and
opportunity. Kor. Manag. Consult. Review. 2014,14 (1), 265285.
(2) Kwak, K. H.; Park, S. W. Global 3D printer industry technology
trend. J. Kor. Soc. Mechan. Eng. 2013,53 (1), 5864.
(3) Forrest, M. J.; Jolly, A. M.; Holding, S. R.; Richards, S. J. Emissions
from processing thermoplastics. Ann. Occup. Hyg. 1995,39 (1), 3553.
(4) Stephens, B.; Parham, A.; Zeineb, E. O.; Tiffanie, R. Ultrafine
particle emissions from desktop 3D printers. Atmos. Environ. 2013,79,
334339.
(5) Unwin, J.; Coldwell, M. R.; Keen, C.; McAlinden, J. J. Airborne
emissions of carcinogens and respiratory sensitizers during thermal
processing of plastics. Ann. Occup. Hyg. 2013,57 (3), 399406.
(6) Health and Safety Executive. Investigation of potential exposure to
carcinogens and respiratory sensitizers during thermal processing of
plastics. 2010. http://www.hse.gov.uk/pubns/ (accessed 5 October
2014).
(7) Lithner, D. Environmental and health hazard ranking and
assessment of plastic polymers based on chemical composition. Sci.
Total Environ. 2011,409, 33093324.
(8) Park J. Y. Occupational exposure assessment for nanoparticles,
Retrieved from the University of Minnesota Digital Conservancy. 2009.
http://purl.umn.edu/55807 (accessed 10 November 2014).
(9) National Institute for Occupational Safety and Health. NIOSH
method No. 2016. 2003. http://www.cdc.gov/NIOSH/nmam/pdfs/
(accessed 10 July 2014).
(10) National Institute for Occupational Safety and Health. NIOSH
method No. 1501. 2003. http://www.cdc.gov/NIOSH/nmam/pdfs/
(accessed 30 July 2014).
(11) Occupational Safety and Health Administration. (1994) OSHA
method No. 104. https://www.osha.gov/dts/sltc/methods/organic/
org104/org104.html (accessed 18 June 2014).
(12) Kim, W.; Choi, I.; Hung, Y.; Lee, J.; Min, S.; Yoon, C. Phthalate
levels in nursery schools and related factors. Environ. Sci. Technol. 2013,
47, 1245912468.
(13) Peters, A. H.; Wichmann, H. E.; Tuch, T.; Heinrich, J.; Heyder, J.
Respiratory effects are associated with the number of ultrafine particles.
Am. J. Respir. Crit. Care Med. 1997,155, 13761383.
(14) Ramachandran, G.; Ostraat, M.; Evans, D. E.; Methner, M. M.;
OShaughnessy, P.; DArcy, J.; Geraci, C. L.; Stevenson, E.; Maynard, A.;
Rickabaugh, K. A strategy for assessing workplace exposures to
nanomaterials. J. Occup. Environ. Hyg. 2011,8, 673685.
(15) Ham, S. H.; Yoon, C. S.; Lee, E. S.; Lee, K. Y.; Park, D. U.; Chung,
E. K.; Kim, P. J.; Lee, B. C. Task-based exposure assessment of
nanoparticles in the workplace. J. Nanopart. Res. 2012,14, 1126.
(16) Oberdörster, G.; Oberdörster, E.; Oberdörster, J. Nano-
toxicology: an emerging discipline evolving from studies of ultrafine
particles. Environ. Health Perspect. 2005,113, 823839.
(17) Yoon, C. S.; Paik, N. W.; Kim, J. H. Fume generation and content
of total chromium and hexavalent chromium in flux cored arc welding.
Ann. Occup. Hyg. 2003,47 (8), 671680.
(18) Park, Y. H.; Hwang, Y. J.; Seo, Y. K.; Beak, S. O. Determination of
phthalates compounds in the ambient atmosphere (II): Evaluation of
experimental artifacts and sample clean-up procedures. J. Kor. Soc.
Environ. Eng. 2010,32 (5), 455460.
(19) Destaillats, H.; Maddalena, R. L.; Singer, B. C.; Hodgson, A. T.;
Mckone, T. E. Indoor pollutants emitted by office equipment: A review
of reported data and information needs. Atmos. Environ. 2008,42,
13711388.
Environmental Science & Technology Article
DOI: 10.1021/acs.est.5b02805
Environ. Sci. Technol. 2015, 49, 1204412053
12053
... The emission rate of particulates and VOCs, and the size distribution of particles are clearly impacted by the type of feedstock material. Various studies suggested that the feedstock materials requiring lower operating temperature (e.g., PLA) emit less particulates than the feedstock materials requiring higher operating temperature (such as ABS) (Stephens et al. 2013;Kim et al. 2015;Mendes et al. 2017;Zhang et al. 2017Zhang et al. , 2019. While the level of toxicity was not the major objective for these studies, the chemical composition of the particulates and VOCs would most likely depend on the feedstock material to some extent. ...
... While the level of toxicity was not the major objective for these studies, the chemical composition of the particulates and VOCs would most likely depend on the feedstock material to some extent. For instance, chemicals, such as aldehydes-known to be carcinogenic-were found in the filament composition (Kim et al. 2015; Albers, Defesche, and Mulling 2019). Zhang and his team perceived that certain additives in these feedstock materials, such as ABS, are potential factors in the characteristics of particulates generated, including size, concentration, and emission rate. ...
... 3.1.2. Printing head/extruder temperature Regardless of which feedstock materials are used for printing, the extruder temperature seems to have a positive relation to the emission rate (Stephens et al. 2013;Kim et al. 2015;Kwon et al. 2017;Mendes et al. 2017;Simon, Aguilera, and Zhao 2017). Particularly, when the extruder temperature is raised beyond the manufacturer's recommended level for a certain feedstock material, the emission rate significantly increases (Kwon et al. 2017;Mendes et al. 2017). ...
Article
Full-text available
As additive manufacturing (AM) has become an evolving discipline in many industries, including manufacturing, medical, and aerospace, it becomes important to identify the risk coming from human exposure to particulates and Volatile Organic Compounds (VOCs) in AM which can lead to serious and chronic health issues. To address this issue, this article first provides a summary of previously reported particulate and VOCs characterization studies during AM processes, including equipment, environmental setups, variables studied, and instrumentation reported in the literature. We then reported a synopsis of the nature of the exposure, characteristics of the emitted particulates and VOCs, and associated health risks for different AM settings in a systematic manner. The key factors contributing to the harmful emissions include the use of toxic material compounds, high operating temperature, manual handling of hazardous materials, and limitations of the underlying printing technology among others. For that matter, we have outlined potential pathways to control hazardous exposure. Our recommendations include adopting safer operational practices, developing regulatory frameworks for facilities and equipment manufacturers, and seeking better technologies that minimize harmful emissions. Our aim is to benefit early-stage researchers, regulators, and industry practitioners in understanding and advancing knowledge of health hazards, safer practices, and technologies in AM.
... The 3D printer is not intended to handle faults; hence, it normally keeps printing even if one occurs. Proceeding with the printing process could harm the 3D printer and even waste materials or the output itself [4]. One of the prominent printing error is nozzle clogging [5]. ...
Article
Full-text available
Additive manufacturing is one of the rising manufacturing technologies in the future; however, due to its operational mechanism, printing failures are still prominent, leading to waste of both time and resources. The development of a real-time process monitoring system with the ability to properly forecast anomalous behaviors within fused deposition modeling (FDM) additive manufacturing is proposed as a solution to the particular problem of nozzle clogging. A set of collaborative sensors is used to accumulate time-series data and its processing into the proposed machine learning algorithm. The multi-head encoder–decoder temporal convolutional network (MH-ED-TCN) extracts features from data, interprets its effect on the different processes which occur during an operational printing cycle, and classifies the normal manufacturing operation from the malfunctioning operation. The tests performed yielded a 97.2% accuracy in anticipating the future behavior of a 3D printer.
... As such, their indoor operation may be considered unsafe as particulate matter, volatile organic compounds (VOC), and other gaseous compounds potentially hazardous to human health can be emitted. Interestingly, such concern may be easily circumvented by redesigning the process (43,76); that is, (i) enclosing the printing system, (ii) using highly efficient air filters, (iii) introducing a ventilation system, and (iv) thoroughly considering factors like the filament color, build plate and extruder temperature, printer model, and use of bio-based polymers. In addition, the printing of parts closer to or exactly at the consumer locations can reduce or eliminate transportation and carbon footprint. ...
Article
Material extrusion 3D printing has been widely used in industrial, educational and residential environments, while its exposure health impacts have not been well understood. High levels of ultrafine particles are found being emitted from 3D printing and could pose a hazard when inhaled. However, metals that potentially transfer from filament additives to emitted particles could also add to the exposure hazard, which have not been well characterized for their emissions. This study analyzed metal (and metalloid) compositions of raw filaments and in the emitted particles during printing; studied filaments included pure polymer filaments with metal additives and composite filaments with and without metal powder. Our chamber study found that crustal metals tended to have higher partitioning factors from filaments to emitted particles; silicon was the most abundant element in emitted particles and had the highest yield per filament mass. However, bronze and stainless-steel powder added in composite filaments were less likely to transfer from filament to particle. For some cases, boron, arsenic, manganese, and lead were only detected in particles, which indicated external sources, such as the printers themselves. Heavy metals with health concerns were also detected in emitted particles, while their estimated exposure concentrations in indoor air were below air quality standards and occupational regulations. However, total particle exposure concentrations estimated for indoor environments could exceed ambient air fine particulate standards.
Article
Three-dimensional (3D) printing is an additive manufacturing process that increases in application and consumer popularity. Studies with 3D printers have shown that the printing process releases particles and volatile organic compounds (VOCs). This review looked at 50 studies that analyzed the most commonly used printing process in consumer 3D printers, the material extrusion or so-called fused filament fabrication (FFF) method and summarizes the most important results. Although the reviewed studies often used different methods, general assumptions can be drawn: Higher printing temperature resulted in higher emissions, styrene was the main VOC emitted during printing with ABS, the size of released particles was in the nano range and filaments with additives could pose a higher risk due to the possible release of e.g., carbon nanotubes (CNTs). In vivo and in vitro studies showed toxic effects. Thus, we recommend: printing in a separated and ventilated room, using the lowest possible print temperature and be cautious with filaments containing particulate additives.
Article
Nanoplastics are emerging contaminants of concern for living organisms and ecosystems, yet nanoplastics are difficult to extract and analyse. Once released into the environment, the fate and behavior of nanoplastics are controlled by physical, chemical, and biological factors. Here, we review nanoplastics weathering, aggregation, biofouling, and bioavailability. Nanoplastics adsorb and transport metals and organic contaminants. Ingestion of nanoplastics by aquatic organims such as microbes, algae, invertebrates, and fish, induces toxicological effects on organism growth, behavior, and reproduction.
Article
Background TVOC (total volatile organic compounds) has been used as a sum parameter in indoor air sciences for over 40 years. In the beginning, individual VOC concentrations determined by gas chromatography were simply added together. However, several methods for calculating TVOC have become established over time. Methods To understand the manifold definitions of TVOC, one must trace the history of indoor air sciences and analytical chemistry. Therefore, in this work, the original approaches of TVOC are searched and explained. A detailed description of the measurement methods is followed by a critical evaluation of the various TVOC values and their possible applications. The aim is to give the reader a deeper understanding of TVOC in order to use this parameter correctly and to be able to better assess published results. In addition, related sum values such as TSVOC and TVVOC are also addressed. Results A milestone was the analytical definition of VOCs and TVOC in 1997. A list of VOCs that should at least be considered when calculating TVOC was also provided. This list represented the status at that time, is no longer up-to-date and is being updated by a European working group as part of a harmonization process. However, there is still confusion about the exact definition and reasonable application of TVOC. The signals of other sum parameters, measured with photoacoustics, flame ionization, photoionization or electrochemical sensors, are also often given under the term TVOC. Conclusions It was recognized early that TVOC is not a toxicologically based parameter and is therefore only suitable for a limited number of screening purposes. Consequently, TVOC cannot be used in connection with health-related and odor-related issues. Nevertheless, such references are repeatedly made, which has led to controversial scientific discussions and even court decisions in Germany about the correct and improper use of TVOC.
Article
Full-text available
3D Printing has a wide range of technologies and applications used in healthcare. In this work, a thorough literature review, concerning the last decade, was conducted in major bibliographic databases. The aim is to report the main applications and materials used in healthcare. In addition, this work tries to identify possible health issues due to particles emissions made by the 3D printing materials.
Article
3D fused deposition modeling (FDM) printing has been extensively applied in various building environments. However, particles are unintentionally emitted during the printing and potentially caused adverse health effects to users. To comprehend these emitted particles, this review summarized the background information, formation mechanisms, measurement methods, characterization and emission control methods based on over 100 literatures published from 2015 to 2022. Although discrepancies of the data amongst the literatures were seen, in general, ABS filaments emitted the highest particle concentration with the largest particle size while PLA generated least particles with the smallest size. The median emission rates in the print of ABS, PLA and other filaments (average of 20 other filaments) were 2.2 × 10¹⁰, 6.0 × 10⁸ and 3.9 × 10⁹ particles/min, respectively. The first quartile percentile and median geometric mean diameter (GMD) were 28.0 and 35.0 nm for ABS, 23.8 and 29.2 nm for PLA and 24.9 and 31.8 nm for the average of other filaments. Filaments with metal additives emitted smaller particles than the non-additive ones. Extra attentions should be paid as the particles from 3D FDM printing are with high concentration and very small sizes, which have a high deposition rate in alveoli. Effective control methods include modifying chemical composition, avoiding printing filaments higher than the recommended temperature, choosing reasonable low infill density and height, applying local ventilation with an enclosure and filters (particles can be removed more than 90%) and adjusting environment humidity if possible.
Article
Full-text available
This study was performed to investigate the fume generation rates (FGRs) and the concentrations of total chromium and hexavalent chromium when stainless steel was welded using flux-cored arc welding (FCAW) with CO 2 gas. FGRs and concentrations of total chromium and hexavalent chromium were quantified using a method recommended by the American Welding Society, inductively coupled plasma-atomic emission spectroscopy (NIOSH Method 7300) and ion chromatography (modified NIOSH Method 7604), respectively. The amount of total fume generated was significantly related to the level of input power. The ranges of FGR were 189-344, 389-698 and 682-1157 mglmin at low, optimal and high input power, respectively. It was found that the FGRs increased with input power by an exponent of 1.19, and increased with current by an exponent of 1.75. The ranges of total chromium fume generation rate (FGR Cr ) were 3.83-8.27, 12.75-37.25 and 38.79-76.46 mg/min at low, optimal and high input power, respectively. The ranges of hexavalent chromium fume generation rate (FGR Cr6+ ) were 0.46-2.89, 0.76-6.28 and 1.70-11.21 mglmin at low, optimal and high input power, respectively. Thus, hexavalent chromium, which is known to be a carcinogen, generated 1.9 (1.0-2.7) times and 3.7 (2.4-5.0) times as the input power increased from low to optimal and low to high, respectively. As a function of input power, the concentration of total chromium in the fume increased from 1.57-2.65 to 5.45-8.13% while the concentration of hexavalent chromium ranged from 0.15 to 1.08%. The soluble fraction of hexavalent chromium produced by FCAW was ∼80-90% of total hexavalent chromium. The concentration of total chromium and the solubility of hexavalent chromium were similar to those reported from other studies of shielded metal arc welding fumes, and the concentration of hexavalent chromium was similar to that obtained for metal inert gas-welding fumes.
Article
Full-text available
The development of low-cost desktop versions of three-dimensional (3D) printers has made these devices widely accessible for rapid prototyping and small-scale manufacturing in home and office settings. Many desktop 3D printers rely on heated thermoplastic extrusion and deposition, which is a process that has been shown to have significant aerosol emissions in industrial environments. However, we are not aware of any data on particle emissions from commercially available desktop 3D printers. Therefore, we report on measurements of size-resolved and total ultrafine particle (UFP) concentrations resulting from the operation of two types of commercially available desktop 3D printers inside a commercial office space. We also estimate size-resolved (11.5 nm-116 nm) and total UFP (<100 nm) emission rates and compare them to emission rates from other desktop devices and indoor activities known to emit fine and ultrafine particles. Estimates of emission rates of total UFPs were large, ranging from ˜2.0 × 1010 # min-1 for a 3D printer utilizing a polylactic acid (PLA) feedstock to ˜1.9 × 1011 # min-1 for the same type of 3D printer utilizing a higher temperature acrylonitrile butadiene styrene (ABS) thermoplastic feedstock. Because most of these devices are currently sold as standalone devices without any exhaust ventilation or filtration accessories, results herein suggest caution should be used when operating in inadequately ventilated or unfiltered indoor environments. Additionally, these results suggest that more controlled experiments should be conducted to more fundamentally evaluate particle emissions from a wider arrange of desktop 3D printers.
Article
Full-text available
This article describes a highly tailorable exposure assessment strategy for nanomaterials that enables effective and efficient exposure management (i.e., a strategy that can identify jobs or tasks that have clearly unacceptable exposures), while simultaneously requiring only a modest level of resources to conduct. The strategy is based on strategy general framework from AIHA® that is adapted for nanomaterials and seeks to ensure that the risks to workers handling nanomaterials are being managed properly. The strategy relies on a general framework as the basic foundation while building and elaborating on elements essential to an effective and efficient strategy to arrive at decisions based on collecting and interpreting available information. This article provides useful guidance on conducting workplace characterization; understanding exposure potential to nanomaterials; accounting methods for background aerosols; constructing SEGs; and selecting appropriate instrumentation for monitoring, providing appropriate choice of exposure limits, and describing criteria by which exposure management decisions should be made. The article is intended to be a practical guide for industrial hygienists for managing engineered nanomaterial risks in their workplaces.
Article
This study was carried out to find out any inherent problems occurring in the sampling and analytical procedures, and to suggest the relevant solutions to the problems. In addition, an optimal condition of clean-up process was developed, which was based on a method using silica glass column. As a result of experiments to test any artificial contamination of blank samples such as glassware and collection media, artifacts of DBP and DEHP appeared to be detected in various kinds of laboratory tools and apparatuses used in the sampling and analytical works. Therefore, it is necessary to investigate a degree of contamination before laboratory works by conducting a prior check any possible contaminations in all experimental tools and apparatus. It is also necessary to devise a method to avoid a tool, if possible, or to use a substitute of phthalate free. If the use of any plastic tool to cause contamination is inevitable, it should be properly corrected with a blank level, as is equally treated as the sample. The clean-up process demonstrated in this study can give us a significant benefit in terms of the quantity and quality of a target compound by GC/MS analysis.
Article
Although task-based sampling is, theoretically, a plausible approach to the assessment of nanoparticle exposure, few studies using this type of sampling have been published. This study characterized and compared task-based nanoparticle exposure profiles for engineered nanoparticle manufacturing workplaces (ENMW) and workplaces that generated welding fumes containing incidental nanoparticles. Two ENMW and two welding workplaces were selected for exposure assessments. Real-time devices were utilized to characterize the concentration profiles and size distributions of airborne nanoparticles. Filter-based sampling was performed to measure time-weighted average (TWA) concentrations, and off-line analysis was performed using an electron microscope. Workplace tasks were recorded by researchers to determine the concentration profiles associated with particular tasks/events. This study demonstrated that exposure profiles differ greatly in terms of concentrations and size distributions according to the task performed. The size distributions recorded during tasks were different from both those recorded during periods with no activity and from the background. The airborne concentration profiles of the nanoparticles varied according to not only the type of workplace but also the concentration metrics. The concentrations measured by surface area and the number concentrations measured by condensation particle counter, particulate matter 1.0, and TWA mass concentrations all showed a similar pattern, whereas the number concentrations measured by scanning mobility particle sizer indicated that the welding fume concentrations at one of the welding workplaces were unexpectedly higher than were those at workplaces that were engineering nanoparticles. This study suggests that a task-based exposure assessment can provide useful information regarding the exposure profiles of nanoparticles and can therefore be used as an exposure assessment tool.
Article
Phthalate esters, which are known endocrine disruptors, are ubiquitously present throughout indoor environments. Leaching from building materials may be a major source of phthalate esters. In this study, we evaluated phthalate ester concentrations in dust samples from 64 classrooms located in 50 nursery schools and explored the critical factors affecting phthalate concentrations, especially with regard to building materials. Dust was sampled by a modified vacuuming method, and building materials were assessed using a portable X-ray fluorescence (XRF) analyzer to determine whether they contained polyvinyl chloride. Di-n-butyl phthalate (DBP), di(2-ethylhexyl) phthalate (DEHP), and di-isononyl phthalate (DINP) were the most frequently detected phthalates. Of these, DEHP was the most abundant phthalate, with a geometric mean of 3,170 µg/g dust, and concentrations were significantly correlated with the area of polyvinyl chloride (PVC)-verified flooring. DINP, which has not been well-reported in other studies, was the second-most abundant phthalate, with a geometric mean of 688 µg/g dust, and showed a critical relationship with the number of children in the institution and the agency operating the nursery school. This is the first study to verify the sources of phthalates with an XRF analyzer and to evaluate the relationship between phthalate concentrations and PVC-verified materials.
Article
A 2-year study has been carried out into the emissions produced during the processing of thermoplastic materials. One of the main reasons for the inception of the work was the perceived need by the plastics processing industry and material suppliers for data in order to comply with recent work-place legislation. Very few data obtained under ‘real life’ situations were available for consultation prior to the start of this study. The principal objective of the project therefore was to determine the effect that the processing of thermoplastics had on the workplace environment by the collection both of qualitative and of quantitative chemical data. During the study a wide range of bulk commercial thermoplastic materials were covered, including polyvinyl chloride (PVC), Nylon 6, acrylonitrile-butadiene-styrene (ABS), high impact polystyrene (HIPS), low density polyethylene (LDPE) and high density polyethylene (HDPE). In order to investigate the effect the type of process had on the emissions produced two principal fabrication methods were studied, namely injection moulding and extrusion-based processes. A wide range of species was detected in each process environment, it being possible to detect the relevant monomer(s) in some cases. However, none of the situations studied were found to generate a high level of process fume. The concentrations of the species detected were found to be in the range 0–2 mg m ⁻³ under standard processing conditions and up to ∼10mg m ⁻³ during purging operations. In none of the situations studied was any individual chemical species found at a concentration above the present occupational exposure limit. The data obtained shows that a higher level of fume is generated by extrusion-based processes than by those involving injection moulding. Emissions data were obtained both by personal exposure monitoring and from a number of static momtors positioned around the process equipment. This revealed the important effect that the monitoring position had on the data generated and the need to employ an effective sampling strategy if representative data was to be obtained. The results obtained also showed how the choice of sampling adsorbent could influence the data obtained. Tenax has been found to be a satisfactory general-purpose adsorbent material for this type of study.
Article
Objectives: Thermoplastics may contain a wide range of additives and free monomers, which themselves may be hazardous substances. Laboratory studies have shown that the thermal decomposition products of common plastics can include a number of carcinogens and respiratory sensitizers, but very little information exists on the airborne contaminants generated during actual industrial processing. The aim of this work was to identify airborne emissions during thermal processing of plastics in real-life, practical applications. Methods: Static air sampling was conducted at 10 industrial premises carrying out compounding or a range of processes such as extrusion, blown film manufacture, vacuum thermoforming, injection moulding, blow moulding, and hot wire cutting. Plastics being processed included polyvinyl chloride, polythene, polypropylene, polyethylene terephthalate, and acrylonitrile-butadiene-styrene. At each site, static sampling for a wide range of contaminants was carried out at locations immediately adjacent to the prominent fume-generating processes. Results: The monitoring data indicated the presence of few carcinogens at extremely low concentrations, all less than 1% of their respective WEL (Workplace Exposure Limit). No respiratory sensitizers were detected at any sites. Conclusions: The low levels of process-related fume detected show that the control strategies, which employed mainly forced mechanical general ventilation and good process temperature control, were adequate to control the risks associated with exposure to process-related fume. This substantiates the advice given in the Health and Safety Executive's information sheet No 13, 'Controlling Fume During Plastics Processing', and its broad applicability in plastics processing in general.
Article
Computers, printers, copier machines and other electronic equipment are a common part of the home and office environments. However, human exposure to potentially harmful pollutants emitted from office equipment has not been systematically evaluated, and is currently not well understood. In this review, we summarize available information on emission rates and/or indoor concentrations of various pollutants that are related to office equipment use, briefly describe experimental methods used to characterize emissions and identify critical research needs in this field. The office equipment evaluated in this review includes computers (desktops and notebooks), printers (laser, ink-jet and all-in-one machines) and photocopy machines. We summarize reported emission rates for the following pollutant groups: volatile organic chemicals (VOCs), ozone, particulate matter and several semivolatile organic chemicals (SVOCs). The latter includes phthalate esters, brominated flame retardants, organophosphate flame retardants, and polycyclic aromatic hydrocarbons (PAHs). We also review studies reporting airborne concentrations in indoor environments (offices, residences, schools, electronics recycling plants) where office equipment was present and deemed to be a significant contributor to the total pollutant burden. We find that for some pollutants, such as organophosphate flame retardants, the link between office-equipment emissions and indoor air concentrations is relatively well established. However, indoor VOCs, ozone, PAHs and phthalate esters can originate from a variety of sources, and their source apportionment is less straightforward. We then observe that personal exposures may be significantly larger than those estimated through average pollutant indoor concentrations, due to proximity of users to the sources over extended periods of time. Finally, we observe that the magnitude of emissions, the link from emissions to personal exposure, the toxicological significance of the chemicals emitted, and the costs and impacts of alternate materials should all be considered in order to evaluate potential importance of human exposures and health risks.
Article
Plastics constitute a large material group with a global annual production that has doubled in 15 years (245 million tonnes in 2008). Plastics are present everywhere in society and the environment, especially the marine environment, where large amounts of plastic waste accumulate. The knowledge of human and environmental hazards and risks from chemicals associated with the diversity of plastic products is very limited. Most chemicals used for producing plastic polymers are derived from non-renewable crude oil, and several are hazardous. These may be released during the production, use and disposal of the plastic product. In this study the environmental and health hazards of chemicals used in 55 thermoplastic and thermosetting polymers were identified and compiled. A hazard ranking model was developed for the hazard classes and categories in the EU classification and labelling (CLP) regulation which is based on the UN Globally Harmonized System. The polymers were ranked based on monomer hazard classifications, and initial assessments were made. The polymers that ranked as most hazardous are made of monomers classified as mutagenic and/or carcinogenic (category 1A or 1B). These belong to the polymer families of polyurethanes, polyacrylonitriles, polyvinyl chloride, epoxy resins, and styrenic copolymers. All have a large global annual production (1-37 million tonnes). A considerable number of polymers (31 out of 55) are made of monomers that belong to the two worst of the ranking model's five hazard levels, i.e. levels IV-V. The polymers that are made of level IV monomers and have a large global annual production (1-5 million tonnes) are phenol formaldehyde resins, unsaturated polyesters, polycarbonate, polymethyl methacrylate, and urea-formaldehyde resins. This study has identified hazardous substances used in polymer production for which the risks should be evaluated for decisions on the need for risk reduction measures, substitution, or even phase out.