Research on 3D printers has focused on laboratory-based testing, using exposure chambers, and simulated ‘real world’ settings. There seems to be a knowledge gap about the use of desktop 3D printers in home settings and the potential for user exposure [63]. In addition, each of the previous research studies has considered slightly different testing scenarios, both in terms of the environment and the sampling protocol, making comparisons between studies more difficult. A method of standardisation should be adopted for testing in different environments to allow comparison and validation of research. For example, a chamber of a consistent size and ventilation or a room environment with a consistent air exchange. The sensors and samplers used should also be arranged in a consistent placement to the source to allow for comparisons between studies.

During their research, Stefaniak et al. [51] stated that it is unknown whether the VOCs or the ultrafine PM, or a combination of the two, were mainly responsible for any negative health effects or biological changes occurring post-exposure. This shows that further toxicological studies should be carried out to assess the impacts of PM and the VOCs identified from 3D printing. These toxicological studies may include cell death in response to exposure, proteomic response to exposures or DNA/RNA changes, based on exposures to the mixtures of VOCs emitted from 3D printers in addition to previous studies looking at particles emitted. These would offer insight into the short-, and long-term responses of the body to the effects of these chemical exposures.

Whilst each of the identified VOCs was well below the occupational limit values, the combined effects of printer VOCs have not been considered i.e., the TVOC concentration. In addition, the mixture of VOCs may interact with each other or the environment forming secondary VOCs/particles not primarily emitted from the printing process. The presence of these secondary VOCs may further increase TVOC. Owing to the large variety of chemical structures identified from past research, the combination of VOCs may also have the potential to interact with or exacerbate health implications.

During the previous research conducted by Väisänen et al. [32], the composition of the identified VOCs differed depending on the resin used. The alcohol compounds identified during dental printing were not present during clear resin printing, and the methacrylate compounds identified from clear resin printing were not among the most abundant during dental resin printing. This difference in VOC emissions may relate to the different compositions of these two resins formulated to provide different properties for the printed component. The different molecules used in the resin liquid would interact causing the bond strength to differ as well as the flexibility of the overall structure. Smaller molecules which can bind together more strongly may create a more rigid structure, whilst longer molecules may introduce flexibility into the item e.g. Polyurethane [10]. In addition to the physical properties, the overall health impact would be dependent on the purpose of the item being created. Dental or medical resins which would have direct, long-term contact with the body would also need to be assessed for the potential for harm at a more rigorous level as observed by fewer TVOC emissions by Pham [64], or by specifically looking at biocompatibility and the ability of the object to survive sterilisation [65], which may impact the chemical composition further.

Research conducted by Pham et al. indicated that biological-based resins are adapted for their purpose and as such emit fewer VOCs than their non-biological counterparts [64]. Tough resin was found to emit ten-fold greater emissions than BioMed or Surgical resins after the post-processing curing process [64].

During research conducted by Väisänen et al., the post-processing stage of the resin print procedure was reported to have a twenty-fold increase in the amount of VOC emissions recorded than the active printing stage. This twenty-fold increase in emissions during the post-processing stage of resin printing may indicate that further emission mitigation techniques may be required for this stage when the operator is more likely to be handling the printed component [32]. This was supported by Yang et al. [49] who discussed the phenomenon of VOC emissions becoming trapped within printer hoods, only to be released in much higher concentrations when the operator opens the hood to remove the printed structure. The increase in VOC emission during the post-processing is likely due to the washing of the printed structure in alcohol before over-curing under UV light. The VOC increase can be attributed to the alcohol bath, in addition to the opening of the UV hood surrounding resin printers, releasing any VOCs trapped within the hood into the greater environment.

When the printing process was investigated by Bowers et al. [38] the post-printing tasks including IPA washing and curing were also found to increase VOC emissions. The authors also established that pre-printing processes, including pouring the resin into the resin bed, emit high concentrations of VOCs and are a further potential exposure for the operators, despite being the task with the shortest duration. Within the entire printing lifecycle, each process emitted a different mixture and quantity of these compounds. IPA was prominent during the cleaning stages, whilst 2-hydroxypropyl methacrylate was quantified up to 58.5 µg/m3 during the recovery stage after printing ended. Acetaldehyde, acetone, ethanol, and styrene were also quantified at their highest concentrations in the recovery stage. This leads to the potential operator exposures being greater after the printing ends, rather than peaking during the printing process.

The methods used for exposure and sample collection varied between studies. The majority of exposure measurements were taken inside exposure chambers, which varied in size from 0.18 m3 to 3 m3 [7, 22, 26, 34], and exposure sampling rooms, 81 m3 to 126 m3 [7, 24], whilst the samplers themselves were in different positions within their environments. The difference in these environments and sampling protocols complicates direct comparisons. In addition, the measurements were reported using different units. Therefore, the reported values have been converted into standard units to be more easily comparable (µg/m3 or µg/min).

Previous research discussed whether if the printer had been used previously that day, this would increase emissions [66]. The use of 3D printers after being in an inactive state would be more representative of hobbyists and home users, rather than industrial use where the 3D printer may be continuously used.

For each filament type (ABS and PLA), concentrations from VOC emissions are discussed in Table 5 and emission rates are tabulated in Table 6. The studies provided consistent evidence for the same types of emissions when the same filament was tested. The identity and concentrations of the VOCs are mostly consistent between the studies, strengthening the overall findings. However, for ABS filament there were ~200-fold differences in the concentrations of acetaldehyde and ethanol emissions. The differences in acetaldehyde and ethanol emissions found between studies could be due to the testing methodology being different as well as the sample collection and sample analysis methods. The differences in collection and analysis methods may account for minor changes in the VOC concentrations quantified, however, the greatest differences are expected to be the test methodology, including print time, distance from the printer and the room volume.

For the VOCs identified from previous literature, the printer emissions fell below the recommended time-weighted average exposure limits in their safety data sheets [42], indicating a limited risk for operators of a single ME printer over 8 h. The TVOC, any repeated exposures, and pre-existing risk factors should be considered for long-term implications to health.

Published studies have reported that VOC emissions throughout the printing lifecycle are affected by the different printing variables. When cured and uncured printed clear resins were used to build surgical components, curing the product reduced VOC emissions ten-fold [58]. The differences between clear and surgical resins only made a small contribution to these VOC emissions [58], indicating that the cured status of the resin may have a larger impact on VOC emission than the resin type.

The chronology of the resin printing also affected emissions. Peak emissions were identified by Yang [49] after the printing finished and the build plate rose out of the liquid resin vat, leaving a large surface area for volatilisation to occur. Also, an emission peak was observed by Yang [49] when the printed component was post-processed by washing it with alcohol to clean the surface. This is supported by work from Han et al., where TVOC emissions peaked when the build plate rose out of the resin bed [37]. Another study found that user exposure was twenty times greater during the post-processing steps than during the active print cycle [32].

In research led by Bowers, the separate stages of VP printing were investigated. They quantified concentrations of VOCs during pouring resin, printing, recovery, and the curing process [38]. The time spent during the tasks was not correlated with the VOC concentrations, as the pouring stage emitted some of the highest concentrations for all quantified compounds despite being the quickest stage. Additional research led by Zhang identified that a resin printer switched off and cold remained a source of VOC emissions, due to the volatilisation of resin compounds at room temperature [36]. Ventilation and adequate storage were recommended, and exposures should be controlled [36].

Research has shown that the filament type impacts the quantity and identity of VOCs emitted [29, 35, 67, 68]. Multiple authors found that the colour of the filament also impacts the VOC emissions [31, 35, 41, 49, 67, 69]. However, Zhang et al. [70] found that filament colour was not a contributing factor for ABS emissions, but filament brand and printer brand were both significant variables with p < 0.0001 Alternatively, for the PLA filament, Zhang et al. [70] found that none of the factors were significantly important for differences in emissions. The largest differences were seen for the different printer brands.

Increased VOC emissions have been found to occur at higher temperatures, particularly for ABS filament which requires a hotter extruder and printing bed temperature than used for PLA [67]. The number of printer head nozzles did not make a difference to VOC emissions when one nozzle was compared to two nozzles to build a small hair comb [67]. However, the nozzle temperature did make a difference in the VOC emissions [71, 72]. Testing at 200 °C, 230 °C and 300 °C resulted in increases in concentrations of VOC emissions as the temperature increased [71]. The relative humidity was also found to alter VOC emission, with greater emission related to higher humidity [72]. The temperature of the build plate and nozzle were not considered to impact VOC emissions, however, only a small sample size was used in this study and therefore can only indicate a trend [73].

The location of the 3D printer will also affect the VOC concentrations within the environment, depending on the size and ventilation of the room. The ventilation may be increased with open windows, fans, or air conditioning [74]; and may be reduced by closing doors or windows [75].

The total personal exposure is the combination of concentration and duration, which is different from the total emission from the printer. Ventilation rates affect exposure as they can alter the concentration of compounds within the environment, with higher ventilation leading to lower concentrations due to dilution into a greater volume and increased removal of air. The exposure that a person experiences may affect how the compounds impact health.

Many previous studies have focused on filament extrusion printers, due to the stability of the feed material, widescale adoption and affordability compared to other models. Resin bed printers are a more recent development and consequently, fewer studies about emissions from this type of printer have been published. Most of the previous research has focused on particle emissions from resin printers, but the presence of VOC has been examined in some studies [26, 32, 38, 55, 56, 75]. In previous research, only a few oxygenate compounds; methacrylates, acetone, benzaldehyde, butyraldehyde, isopropanol, formaldehyde, hexaldehyde and propionaldehyde [32, 56] were quantified from the printing process. There seems to be a gap in knowledge concerning VOC emissions from resin bed printers, and specifically for home users where ventilation may be poor compared to industrial settings. Personal exposure studies would benefit research into the safety of affordable 3D printing technology, which is becoming more mainstream in daily life for many work and school sectors.

The differences in identity and quantity of VOCs emitted from the two types of filaments used during ME 3D printers are summarised in Tables 5 and  6. The considerable number of VOCs identified from various filaments and past research for ME shows how varied the VOCs are. In comparison, Table 4 lists the VOCs quantified during resin bed printing as mainly carbonyl compounds or methacrylate compounds. This difference in composition is likely to be caused by the composition of the feedstock materials themselves. The VOCs emitted from the thermoplastics mainly derived from the compounds that the filaments were made from. While the resins emit the monomers used to make the polymer. The variety of VOCs emitted can lead to varied exposures and therefore varied implications for the human body.

An additional difference between VP and ME printing is the temperature of the processes. Whilst ME printing melts the thermoplastics at high temperatures, VP printers only increase to around 40 °C. The VOC emission is dependent on temperature for liquids, as increased temperatures allow more energy per molecule and a higher chance of the molecule partitioning into the gas phase and being emitted as VOC emissions.

Each of the VOCs quantified from previous research was emitted by 3D printers at concentrations well below the published 8-h GB WEL values, noting the caveat that the printer studies relate to emissions and not personal exposure assessments to VOCs. It is unlikely that the operators of the 3D printers would have continual exposure for eight hours at the same emission rate, due to moving within rooms, leaving rooms, increasing ventilation, and shorter print times amongst other reasons. Work carried out by Runstrom et al. identified the amount of time that VP and ME 3D industry printer operators spent on non-3D printing tasks was 97% and 96% respectively, limiting potential exposure periods [76]. Owing to the probability of a shorter or lower exposure, the dose that the user is exposed to is likely to be even lower than the GB WELs for these VOCs.

One of the more commonly identified VOCs from ME and VP printers is methyl methacrylate [22, 25]. Methyl methacrylate has been identified as a respiratory irritant [77] but its status as a respiratory allergen is uncertain. Methyl methacrylate caused lung inflammation in mice after they were exposed to 150 ppmv (6 × 105 µg/m3) for 120–200 min [78]. However, when Muttray et al. [79] exposed volunteers to 50 ppmV methyl methacrylate (2 × 105 µg/m3) over four hours in a test chamber they found no significant impact on the exposed volunteers other than reports of headaches, which diminished 35 min after exposure [79]. Overall, a review assessing methyl methacrylate determined there was insufficient evidence to classify it as a respiratory sensitiser [80]. The GB WELs are 50 and 100 ppmV for the 8-h time-weighted average and short-term exposure limits respectively [81]; and are several orders of magnitude higher than the concentrations of VOCs quantified from printer emissions [22, 23, 25, 32].

A similar compound linked to 3D printing is 2-hydroxypropyl methacrylate [32]. This compound has been linked to skin irritation, eye damage, skin sensitization, and single-exposure inhalation organ toxicity [82]. Though no GB WEL has been set for hydroxypropyl methacrylate, the 8-h time-weighted GB WEL exposure for methacrylic acid is listed as 70,000 µg/m3, which is four orders of magnitude greater than the reported printer emissions of 6–8 µg/m3 [32]. There are no GB WEL limits for 2-hydroxypropyl methacrylate, however, there is a GB WEL for the structurally and chemically similar molecule 2-hydroxypropyl acrylate which is 2.7 mg/m3 [42], which is again, several orders of magnitude greater than identified. 2-hydroxypropyl acrylate is very similar to 2-hydroxypropyl methacrylate in terms of chemical structure, reactivity, size and polarity, and has similar health impacts such as skin and respiratory irritation and sensitization, however, 2-hydroxypropyl acrylate is also a skin corrosive [83].

Benzene and toluene were also reported in 3D printer emissions and have separately been linked with risk for asthma and other respiratory diseases [84]. Benzene is associated with increasing the risk of leukaemia [84] and is considered a multi-organ carcinogen [85] where exposure can cause chromosomal aberrations [85]. From Table 5, benzene was quantified up to 11.5 µg/m3. The GB WEL for benzene is 3.25 mg/m3 [42], which is 2 orders of magnitude greater than identified.

Toluene has been shown to have neurotoxic effects post-exposure of up to 200 ppm (7.5 × 105 µg/m3) and has an 8-h GB WEL of 50 ppm (1.8 × 105 µg/m3) [63]. These are several orders of magnitude above the literature emission values of up to 61 µg/m3. The GB WEL for 8 h is 191 mg/m3 and the 15-min exposure with 384 mg/m3 [42].

Another major VOC linked with 3D printing is styrene [7, 22, 25, 41, 86]. Ambient styrene concentration was associated with blood styrene concentration [87], as well as being associated with a reduction in the vibrotactile sensitivity of the participants [87] and reduced stability whilst standing on one leg [87]. Styrene exposure has also been associated with poor colour vision [88]. The reported concentrations of styrene from desktop printer emissions are below the workplace exposure limits of 430 mg/m3 [42] by several orders of magnitude

Two of the compounds exceeded the UKHSA guidelines for exposures during ME printing, styrene during ABS printing, and formaldehyde during PLA printing. The guideline air quality value for styrene is averaged over a year, so the potential health risks for people exposed to 3D printer fumes are still likely to be low. However, the guideline for formaldehyde was averaged over 30 min, which may potentially place the operator at risk should they exceed a 30-min exposure at the same concentration. As only one out of two PLA filaments exceeded the UKHSA guidelines, the variation between filaments may be high, even when the same material is used. The variation in VOC emission may lead to differing exposures experienced by the operators, and so caution should be undertaken.

The setting of occupational exposure limits and control guidance values do not typically consider vulnerable non-occupational groups. For example, those with pre-existing respiratory conditions such as asthma or COPD [89], or the elderly or young children. Any of these vulnerable groups may experience a non-proportional response to the VOC exposure. Previous research has highlighted associations between exposure to VOCs and symptoms affecting the respiratory, cardiovascular, and neurological systems [89]. In addition to pre-existing risk factors, chronic exposure needs to be considered. The impact of repeated exposure may cause long-term symptoms, even at low doses.

Previous research led by Karwasz investigated the printing habits of ME printer users and modelled potential exposure scenarios when using a non-ventilated hood. 15% of participants reported headaches when using the printer, 70% used printers with an exposed print chamber and 57% did not use any filtration with the remaining participants being unsure [90]. When the modelling scenarios were analysed, the opening of the chamber door resulted in high levels of pollutants within 3 s, regardless of the ventilation within the room [90]. These scenarios identify the possibility of short-term high exposures to VOC emissions from 3D printers.