Graphene is a two-dimensional (2-D), carbon-based, nanomaterial that demonstrates exceptional physical and chemical properties such as electrical conductivity, mechanical strength, and thermal stability (Mbayachi et al. 2021). The past two decades have witnessed an extraordinary increase in graphene research and innovation since the material was isolated in 2004. Graphene’s unique properties have facilitated its use in a variety of manufacturing applications, including semiconductors, composites, sensors, energy storage, functional inks, polymer additives, tires, coatings, electronics, and filtration (Novoselov et al. 2012; Randviir et al. 2014).

Graphene is often referred to as the graphene family of nanomaterials (GFNs) since it is a class of materials based on their surface properties, number of carbon layers, and size. GFNs can include few-layered graphene (FLG, 2–5 layers), multi-layered graphene (MLG, 2–10 layers), graphene nanoplatelets (GNP, stacks of graphene sheets > 10 layers), as well as materials based on oxygen content, which include graphene oxide (GO) and reduced graphene oxide (rGO, <5% oxygen content) (Niang et al. 2025). These materials can be considered pristine, which is their unaltered form, or functionalized, where they undergo surface modifications to increase solubility and reactivity (Georgakilas et al. 2012; Chowdury et al. 2024). There is a growing consensus on GFN characterization and classification as recent standards have been adopted (ASTM Standard E3220 2020; ISO 2025).

The worldwide graphene market size was estimated at $0.76 billion USD in 2023 and is expected to expand at a rate of 40.4% over the next 5 years, reaching $4.13 billion USD by 2028 (Research and Markets 2024). Additionally, surveys conducted in 2016 and 2019 found that graphene was one of the most reported carbonaceous nanomaterials produced or used by U.S. facilities (Babik et al. 2018; Neu-Baker et al. 2022). With the growing production and application of GFNs, the likelihood of workplace exposures throughout their life cycle is expected to rise. Despite the fast pace of growth that promises further innovations and applications in various fields, the number of workers exposed to GFNs in the U.S. is largely unknown.

Currently, there is limited information regarding potential human health effects associated with occupational exposures to GFNs. To date, only one human clinical research study has been published. Andrews et al. (2024) conducted a controlled acute inhalation (2 hr) study that aerosolized 200 μg/m³ of GO to expose human participants and reported no changes in respiratory or cardiovascular health and a limited effect on blood coagulability. In the absence of other human data, information on the potential effects of graphene has largely been assessed by animal toxicity studies.

Reviews of acute and sub-chronic animal studies have shown the potential to cause non-cancerous adverse lung effects, including inflammation, fibrosis, or granuloma formation following inhalation, intratracheal instillation, or pharyngeal aspiration in rats and mice (Kong et al. 2024; Niang et al. 2025). Other in vivo studies, which used various routes and dosing methods, suggest that pristine and functionalized GFNs may exhibit potential pulmonary, systemic, reproductive, developmental, and genotoxic outcomes in laboratory mammals. (Ema et al. 2017; Rhazouani et al. 2021; Niang et al. 2025). However, the toxicological data remain relatively limited, especially when considering the many different types of GFNs and their potential surface modifications.

Several carbonaceous materials, such as synthetic and natural graphite, carbon black, coal dust, and carbon nanotubes, have occupational exposure limits (OELs) in the U.S. that range from 3.5 to 15 milligrams per cubic meter (mg/m³) at the total/inhalable aerosol size fractions and 0.001 to 5.0 mg/m³ at the respirable aerosol size fraction (See Supplemental Table S1; NIOSH 2013a; NIOSH 2019a, 2019b, 2019c, 2019d). Inhalable particles (D₅₀ = 100 μm) are those that may enter the mouth or nose, while respirable particles (D₅₀ = 4 μm) may penetrate the unciliated or gas-exchange regions of the respiratory system (Vincent et al. 1999).

There are currently no international OELs published by authoritative organizations to ensure the safe handling of GFNs in workplaces. However, several independent research groups have conducted risk assessments that derived preliminary OELs of 18 micrograms per cubic meter (μg/m³) and 212 μg/m³ from different types of GFNs (Lee et al. 2019; Spinazzè et al. 2019). Additionally, researchers have also used the NIOSH Occupational Exposure Banding process on GFNs. Results of the banding process indicated that graphene exposures may have potential health consequences at air concentrations <10 μg/m³ (NIOSH 2019e; Niang et al. 2025).

To date, there is limited occupational exposure data for GFNs captured outside of small-scale research and development settings, and no published exposure assessment research has been conducted within U.S. workplaces to determine exposures to GFNs. Therefore, the primary objective of this study was to conduct an industry-wide exposure study of graphene at representative primary manufacturing facilities and downstream secondary manufacturing facilities. The aim was to (1) demonstrate the feasibility of using facility background corrected elemental carbon (EC) measurements as a marker for assessing GFN exposures, (2) understand the range of occupational exposures within U.S. workplaces, and (3) inform risk assessors and toxicologists on realistic exposure information and material characteristics of GFNs.

All four secondary manufacturing facilities that participated in this study processed or handled GO (Table 1). Of the seven primary manufacturing facilities visited, three produced GNP, two produced FLG, and one each produced MLG and rGO. The range of employees directly handling graphene at each facility was between 2 and 15 workers. Ten of the 11 facilities visited had the capability to produce or handle kilogram (kg) quantities during the workday. The maximum quantity handled was at two primary manufacturers of GNP that handled over 100 kg per day. Facility personnel, except in one facility, were observed to use respiratory protection for at least some of the tasks sampled, and all but two facilities used some form of engineering control.

This study assessed occupational exposures to GFNs for 44 participants from 11 primary and secondary manufacturing facilities located across the U.S. Participating facilities produced or used different types of GFNs with varying amounts of carbon layers, and over half of the facilities produced or used kilogram quantities daily. Additionally, most of the represented facilities were small businesses with fewer than 30 employees.

Generally, many of the smaller facilities did not have the resources to have a dedicated occupational and environmental health and safety professional to anticipate hazards, develop worker training, perform exposure monitoring, and implement proper engineering controls. However, every facility had at least one employee who worked either full-time or part-time on these issues and expressed awareness of the importance of controlling exposures to GFN.

Most facilities used some combination of exposure control measures and PPE to reduce GFN exposures. However, the extent of their use and capacity to mitigate exposures varied greatly from facility-to-facility. Nearly every facility and 45% of participants were observed using either NIOSH-approved filtering facepiece or elastomeric half-face respirators for respiratory protection. However, participants at two facilities were observed using acid/gas respirator cartridges, which do not protect against GFN or other particulate exposures.

All but two of the facilities had some type of engineering control(s) in place to reduce workplace GFN exposures. Interestingly, workers who used a ventilated or unventilated enclosure were found to have the highest exposures, followed by workers who did not use any engineering controls (Table 3). This was likely attributable to many of the production processes being enclosed, except for short periods when large amounts of material were manually handled or transferred. It should be noted that engineering controls have been proven to be effective at controlling exposures for nanomaterial production and downstream processes (NIOSH 2013b).

To date, there is no consensus on exposure assessment or air sampling methods for GFNs. However, exposure assessment methods that utilize elemental carbon (EC) as a marker for exposure, usually collected alongside a confirmatory sample analyzed by electron microscopy, have been developed and validated for other carbonaceous nanomaterials and continue to gain consensus in the literature (NIOSH 2013a; Canu et al. 2020; McCormick et al. 2021). Recent international studies have also used similar methods, which have further demonstrated the feasibility of using EC to assess occupational exposures to GFNs (Lee et al. 2016; Vaquero et al. 2019; Lovén et al. 2021; Storsjö et al. 2024; Poikkimäki et al. 2025).

Lee et al. (2016) collected personal samples to assess EC levels at the respirable size fraction across two research laboratories producing and handling small quantities of GO and rGO and creating graphene thin films. No personal samples collected for EC were found to be above the LOD (0.3 μg/cm²). Vaquero et al. (2019) assessed GO and rGO exposures at a pilot production plant with the capacity to produce up to 1 kg per day of material, and the only personal respirable sample collected contained EC levels below the LOD (1 μg/cm²). Lovén et al. (2021) assessed EC levels at two facilities making ink formulations for surface coatings. EC levels captured in the personal breathing zone ranged from below the LOD (0.06 μg/cm²) to 5.6 μg/m³. Storsjö et al. (2024) collected total aerosol personal samples from two workers at a graphene production facility that produced less than 100 grams of graphene powder per day and reported EC exposures of 20 and 60 μg/m³ with an LOD of 0.2 μg/cm². Poikkimäki et al. (2025) assessed respirable EC levels from seven laboratory-scale exposure scenarios handling rGO, GO, and FLG and did not report any EC concentrations above the LOD (0.31 μg/cm²).

Other studies assessing GFN exposures have focused on exposure metrics such as gravimetric mass analysis and the use of direct reading instruments to assess mass, lung deposited surface area, and particle number concentration (Heitbrink et al. 2015; Spinazzè et al. 2016, 2018; Boccuni et al. 2018, 2020; Iavicoli et al. 2018; Bellagamba et al. 2020, 2023; Tombolini et al. 2021; López et al. 2023). Generally, these studies assessed exposures at facilities producing low volumes of GFNs (<kg). Additionally, previous studies on other carbonaceous nanomaterials have shown that the use of direct reading instruments may not be the most appropriate tools for collecting quantitative exposure assessment data because of the inability to assign gradient levels of exposure (Evans et al. 2010; Dahm et al. 2013).

In this study, workers had a wide range of exposures to GFNs at the respirable (<0.01 to 1825.23 μg/m³) and inhalable (0.01 to 6327.10 μg/m³) aerosol size fractions. The highest exposure at the inhalable size fraction in this study was several orders of magnitude higher than what has been previously reported in the literature for GFN exposures. Generally, the primary manufacturing sites had the highest exposures in this study, a finding that was supported in the regression model, as two of the three types of secondary manufacturers were found to have significantly decreased exposures compared to primary manufacturers. This difference in exposure was likely attributable to many of the primary manufacturers producing and handling larger quantities of dry powdered GFN for longer durations throughout the workday, while secondary manufacturers used GFN materials dispersed in liquids more frequently.

It remains a challenge to place the range of exposures demonstrated in this study within the appropriate context to adequately protect workplace health and safety. First, the potential risks associated with GFN exposures will likely differ by specific material characteristics (e.g., size, carbon layers, impurities), which makes interpreting the growing amount of animal toxicity data challenging. Future study designs could consider using the graphene classification parameters recommended in ISO/TS 9651:2025 Nanotechnologies—Classification Framework for Graphene-Related 2D Materials (ISO 2025). Additionally, there has only been one human clinical research study to assess the potential effects of acute inhalation exposures to graphene oxide that found limited health effects (Andrews et al. 2024). While this finding is encouraging, epidemiological information is important to understanding the effects of chronic exposures to GFNs, which are more likely to be experienced by workers (Pelin et al. 2018). Finally, there are no current international OELs published by authoritative organizations to compare exposure sampling results and ensure the safe handling of GFN in workplaces.

Several OELs exist from authoritative organizations for bulk synthetic and natural graphite in the form of OSHA PELs, ACGIH^VR TLVs^VR, and NIOSH RELs. These OELs range from 2 to 5 mg/m³ at the respirable size fraction and 10 to 15 mg/m³ at the inhalable size fraction (Supplemental Table S1). Graphite is a typical precursor material for a top-down production process of GFNs, which were commonly observed in this study. Multiple facilities were familiar with the OELs for graphite and acknowledged they were trying to maintain exposures below those levels. However, several samples collected from this study for EC were either over the action level (half of the OEL) or approached these OELs. Previous research has suggested that ultrafine (nano) particles may be more detrimental to human health when compared to similar bulk materials (Gupta and Xie 2018). This increase in potential toxicity can be attributed to deeper lung penetration and higher reactivity due to particle size and the increase in surface area (Bakand et al. 2012; Andrews et al. 2024).

Results from this study found that 28% and 10% of the TWA personal samples collected at the respirable size fraction exceeded the suggested OELs of 18 μg/m³ by Lee et al. (2019) and 212 μg/m³ by Spinazzè et al. (2019), respectively. Meanwhile, 38% of the respirable personal samples collected were above the suggested OEB of <10 μg/m³ established by Niang et al. (2025). Similarly, results of personal samples in the inhalable size fraction of 49% and 10% exceeded the suggested OELs of 18 and 212 μg/m³, respectively, while 53% of these personal samples exceeded the suggested OEB of <10 μg/m³.

The discrepancies between suggested exposure limits and exposures observed in this study further emphasize the importance of consensus exposure assessment methods and safety guidance for GFNs. Since there are no consensus exposure assessment methods for GFNs, Niang et al. (2025) suggested sampling for EC at the respirable size fraction due to adverse lung effects from animal toxicity data near the alveolar region. When feasible, collecting samples at both the inhalable and respirable size fractions can be informative, especially if little is known about the material and aerosol characteristics of the GFNs being sampled. A recent study by Evans et al. [forthcoming] examined the dustiness of several GFNs and graphite materials and found that the thoracic aerosol fraction constituted 56% of the inhalable fraction. Similarly, this study found the average respirable-to-inhalable ratio was 50% but varied by site, environmental conditions, and material. This information could be useful to retrospectively convert a value in case future OELs are developed, but the ratio may differ substantially by GFN type.

Additionally, collecting a sample for a confirmatory analysis by electron microscopy could also be useful if resources allow, but it may not always be necessary, especially if the occupational hygienist’s professional judgment indicates that exposures are occurring and there are limited organic and elemental carbon interferences in the workplace.

The marginal model for inhalable EC demonstrated higher within-facility and within-worker correlations than the respirable EC model, which still exhibited moderate correlations. These findings underscore the suitability of marginal regression for capturing within-facility variability in clustered exposure data and within-worker variability in longitudinal exposure data. Additionally, the model for inhalable EC also found more significant parameter estimates, most notably for the facility-level determinants. A few of the job tasks were significantly associated with decreasing or increasing exposure potentials. Moreover, a subjective assessment of overall engineering control adequacy was found to be the strongest parameter estimate that increased the exposure potential. An analogous engineering control rating was used in a model to predict carbon nanotube exposures with similar results (Dahm et al. 2019). The variable was categorical and based upon observations with ratings of adequate (appeared to effectively control exposures and used consistently and properly by employees), semi-adequate (many appeared to be effective however some were likely deficient and were not consistently or properly used by employees), and inadequate (lacked proper engineering controls and/or controls present likely did not effectively control exposures). This may suggest that professional judgment of engineering control adequacy could be useful in future models to assist in estimating occupational exposures. Especially since it has been shown that industrial hygienists who use professional judgment alone to make exposure judgments have a wide range of accuracy (Logan et al. 2009).

This study should be considered cross-sectional in nature and may not be wholly representative of exposures that U.S. workers handling GFN may encounter. Additionally, the overall summary statistics for all sites and for several individual facilities are likely an under-estimation of the true exposure, as some samples contained large amounts of EC that exhibited incomplete combustion during EC analysis. The standard deviations and geometric standard deviations indicated large variances in the exposures from facility-to-facility and for some job tasks. This was likely due to the wide range of job duties and exposure potentials at each facility. The job tasks in this study were assigned based on similar duties, but many facilities were performing similar tasks in different ways and with varying levels of controls. Finally, the summary statistics displayed in Tables 3 and 4 should be interpreted with caution since task-based samples were not collected. A participant likely performed multiple tasks throughout their workday that contributed to their overall exposure.

In summary, this is the first multi-site investigation conducted in the U.S. to describe the range of occupational exposures for various types of production and manufacturing facilities. There are currently no well-recognized OELs for GFNs, but samples collected from this study for EC were found to either be over the action level or approached OELs for bulk carbonaceous materials, such as graphite. This exposure assessment study, along with other recent studies, demonstrates the feasibility of using EC as a marker to assess GFN exposures and inform risk assessors and toxicologists on realistic exposure information and characteristics of GFNs. The findings underscore the importance of standardized exposure assessment methods and the establishment of formal occupational exposure limits for the growing family of graphene materials to safeguard worker health.

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15459624.2026.2633099. AIHA and ACGIH members may also access supplementary material at http://oeh.tandfonline.com.