Influences of coffee roasting stages and variables on the formation of diacetyl and 2,3-pentanedione: A field assessment in a café setting

Influences of coffee roasting stages and variables on the formation of diacetyl and 2,3-pentanedione: A field assessment in a café setting | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Influences of coffee roasting stages and variables on the formation of diacetyl and 2,3-pentanedione: A field assessment in a café setting Ta-Yuan Chang, Tzu-Chieh Su, Yeh-Chung Chien This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8615798/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This study examined airborne alpha-diketones (diacetyl and 2,3-pentanedione), known respiratory irritants, across different coffee processing variables. A factorial design assessed roast degree and processing stage, supported by separate tests on other factors. All experiments were performed in triplicate at a single site to reduce variability. Real-time measurements of total volatile organic compounds (TVOC) and PM 2.5 aerosols were also conducted when possible. Significant differences (p roasting > packaging), but not by roast degree. Grinding emitted much higher TVOC than roasting (1131.7 vs. 88.3 ppb), while roasting produced higher PM 2.5 (1079.6 vs. 90.1 µg/m³). TVOC levels strongly correlated with diacetyl (r = 0.827) and 2,3-pentanedione (r = 0.848). Beans from Brazil (sun-dried) released more alpha-diketones than Ethiopian (washed) beans. Grinding zone levels of diacetyl and 2,3-pentanedione were 8.4 and 4.9 times higher, respectively, than n the nearby bar area (~ 7 m away). Bean aging affected 2,3-pentanedione levels, peaking on the first day after roasting. Grind size and quantity showed no significant effects. These findings highlight key factors influencing alpha-diketone emissions, and reflect realistic exposures in café environment. Routine monitoring and source control are essential to protect worker and consumer health. Elevated PM 2.5 levels during roasting also merit further study on their chemical characteristics and toxicity. Coffee roasting diacetyl 2 3-butanedione 2 3-pentanedione TVOC PM2.5 Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Highlights 1. This study systematically examined diacetyl and 2,3-pentanedione levels across various coffee processing variables, highlighting key emission variables such as processing stage (grinding > roasting > packaging), bean origin (Brazil > Ethiopia), sampling distance, and bean aging. Roast degree, grind size and quantity had no significant effect. 2. Findings reflect real-world exposures in small-scale cafe environments, with implications for worker and consumer health. 3. Notably, roasting produced significantly higher fine aerosol (PM) levels than grinding, identifying a new area for research. 4. This work bridges food science and occupational health by addressing both aroma-related compounds and airborne hazards. 1. Introduction Coffee is one of the most widely consumed beverages globally, with its flavor and aroma primarily developed through the roasting process. During roasting, complex chemical reactions such as Maillard reactions, caramelization, and pyrolysis occur, producing a wide range of volatile organic compounds (VOCs) that contribute to coffee’s characteristic sensory properties [ 1 – 4 ]. Among these VOCs, 2,3-butanedione (diacetyl) and 2,3-pentanedione are particularly noteworthy due to their distinct aroma and potential health implications. While these alpha-diketones impart desirable butter-like flavors at trace levels, high concentrations have been associated with respiratory illnesses, notably bronchiolitis obliterans, in occupational settings such as commercial coffee processing factories. Similarly, workers in commercial popcorn factories have experienced obliterative bronchiolitis due to high exposures to diacetyl, an artificial buttery flavoring ingredient [ 5 – 6 ]. Furthermore, diacetyl and 2,3-pentanedione have been identified in flavored e-cigarettes, raising new concerns regarding consumer health [ 7 ]. The presence of alpha-diketones in various environments has prompted increased regulatory and research attention in recent years [ 8 – 10 ]. Additionally, "neo-formed contaminants" (NFCs) like acrylamide, which occur during coffee processing, have raised food safety concerns due to their potential long-term risks for consumer health [ 11 ]. Limited research has systematically compared how different roasting protocols affect the formation dynamics of various coffee-derived compounds. One study found that the concentration of flavoring organics in coffees, assessed by bean extractions, can vary significantly depending on roasting parameters such as time, temperature, airflow, and the type of roasting process (e.g., drum roasting, hot air roasting, fluid bed roasting) [ 12 ]. VOCs emitted during coffee roasting were stage-dependent. After an initial drying phase, concentrations of VOCs like acetic acid, acetaldehyde, and pyridine rapidly increased, reaching a maximum at a medium roast level. Sporadic bursts of some volatiles, including 2,3-pentanedione, coincided with bean popping [ 13 ]. High within-batch and between-batch variability were observed among the 50 identified volatile compounds [ 14 ]. Field measurements in coffee processing plants or café shops have identified the highest diacetyl and 2,3-pentanedione levels in grinding or packaging areas [ 9 , 15 – 21 ]. However, due to the nature of field assessments, the specific influencing factors for VOC release are challenging to evaluate. Understanding the impact of specific coffee roasting variables on the generation of various compounds in a café setting is crucial not only for optimizing flavor profiles but also for ensuring the safety of workers or baristas in the coffee industry, and consumers/patrons. Accordingly, this study aims to analyze and compare the levels of airborne diacetyl and 2,3-pentanedione produced under different coffee roasting settings, thereby contributing to both food science and occupational/environmental health. 2. Materials and Methods 2.1 Study Design This study investigated the levels of airborne diacetyl and 2,3-pentanedione during various coffee roasting stages and under different variables. A factorial experiment was conducted to test the effect of roast degree, roasting stage, and their interaction. This was supplemented by independent tests examining the effects of grinding fineness, bean origin, sampling distance, grind quantity, and bean aging. All tests were performed within a single coffee establishment to minimize variability. Real-time measurements of total volatile organic compounds (TVOC) and particulates with an aerodynamic diameter less than 2.5 micrometers (PM 2.5 ) were also performed where feasible. 2.2 Coffee Roasting Processes (Stages) To facilitate comparisons, this study standardized coffee roasting processes into three stages: Green Bean Roasting (R): Green (raw) beans were roasted to different degrees, determined by the final bean temperature: light (189–192°C), medium (198–201°C), and dark roast (207–210°C). The approximate roast times were 120 minutes, 130 minutes, and 150 minutes, respectively. Before roasting, green beans were stored in a room with a temperature between 20°C and 25°C and packaged in vacuum-sealed bags to maintain their initial moisture content between 10% and 12% for optimal flavor. During roasting, 3 kg per batch of beans were loaded into the roaster and followed the roast routine. A total of 6 batches were processed to complete the entire roasting procedure, taking 2 to 2.5 hours total. Packaging and Storing of Roasted Bean (P): Roasted coffee beans were then poured into an iron bucket for natural cooling (~ 1.5 hours) and subsequently stored in one-pound capacity bags with one-way degassing valves. The overall procedure took approximately 2 hours. Grinding and Packing (G): Roasted coffee beans were ground using a coffee grinder (Model CM-520, Flying Eagle Co., Taiwan). The resulting coffee powder was then divided into quarter-pound capacity bags with degassing valves. Two kg of beans were ground for most experiments, and the overall procedure took approximately 20 minutes. 2.3 Test Variables This study comprised a 3x3 factorial experiment (roast degree * roasting stages) and independent tests. The roast degrees included Light, Medium, and Dark roast, while the roasting stages included R, P, and G, as described above. Test variables in the independent experiments included: Grinding fineness: Coarse (particle size < 1.40 mm) or fine (particle size < 0.85 mm), achieved by grinder settings and assured by metal mesh. Bean origin: Arabica variety ( Coffea arabica ) beans from (A) Ethiopia (1900-2200m grown, washed), (B) Guatemala (1600-1800m grown, washed), and (C) Brazil (750-1200m grown, natural/sun-dried) were tested. Sampling distance: Distances away from the grinding zone (0 m), seating zone (~ 3 m), and bar zone (~ 7 m). Grind quantity: 2 kg or 0.5 kg per batch. Bean aging: Days post-roasting: 1, 5, or 7 days. 2.4 Test Facility The test facility was a privately-owned coffee shop located in Douliu City, central Taiwan, with dimensions of 10.5 m (length) * 8 m (width) * 4 m (height). It was divided into five major zones: roasting, grinding, packaging, seating, and bar area. The store's ventilation primarily relied on natural ventilation during sampling. The roasting zone was equipped with a commercial coffee roaster (Model GS-W6M, Giesen BV, Netherlands) capable of roasting 6 kg of beans per batch. Figure 1 illustrates the floor plan and sampling spots of the facility. 2.5 Sampling and Analysis of Diacetyl and 2,3-Pentanedione Sampling and analysis of airborne diacetyl and 2,3-pentanedione generally followed OSHA 1012 and OSHA 1016 methods with minor modifications [ 22 ]. Samples were collected using an active pump connected to a silica gel tube (#226 − 183, SKC Inc., US). Samples were collected at typical operator positions, 150 cm above the ground, to mimic normal breathing zones. Sampling time was typically 2 hours, with 20 minutes for short-term sampling. The sampling flow rate was 50–200 mL/min to achieve optimal analytical sensitivity. Ambient temperature and humidity were also recorded during sampling. Samples were stored in a freezer (4°C) and analyzed within 7 days. Sample analysis was based on separation via gas chromatography coupled with a flame ionization detector (GC-FID, Model Trace 1300, Thermo Scientific, US). Samples were first desorbed, with the aid of vibration, using 2 mL of ethanol/DI water (95:5) solution for both the front and rear tube sections. Finally, 1.0 µL of the aliquot was injected for analysis. 2.6 Measurement of TVOC and PM 2.5 Total volatile organic compounds (TVOC) were continuously monitored using a portable volatile gas monitor, which employed photoionization detection with a 10.6 eV UV-discharge lamp (Model ppbRAE 3000, RAE systems, US). Particulate (PM 2.5 ) levels were continuously measured using an aerosol monitor, which utilized light-scattering laser photometry to provide aerosol mass concentration (DustTrak 8530, TSI Inc., US). Sampling locations for these instruments were adjacent to those of the silica-gel samples for comparison. 2.7 Test Procedure Due to limited resources, only selected conditions were tested to examine the potential influences of variables on diacetyl and 2,3-pentanedione formation. The test schemes generally followed the procedure outlined in Section 2.2 , but were specific to each sub-test. Each experiment was run in triplicate. All tests were performed between October 2023 and June 2024. Factorial experiment (roast degree * process): Ethiopia bean (ground to coarse size) was tested. Sampling occurred at R, P, and G zones. TVOC and PM 2.5 levels were measured at R and G zones. Test for grind fineness: Ethiopia bean (medium roast), with coarse or fine particle size, was tested. Sampling occurred at the G zone. Test for bean origin: Three beans (Ethiopia, Guatemala, and Brazil; medium roast, and ground to coarse size) were tested. Sampling occurred at the G zone. Test for sampling distance: Ethiopia bean (light roast, and ground to coarse size) was tested. Sampling occurred at the G, seating, and bar zones. Test for bean aging: Ethiopia bean (dark roast, and ground to coarse size) was tested. Sampling occurred at the G zone. Test for grind quantity: Ethiopia bean (dark roast, and ground to coarse size) was tested. Sampling occurred at the G zone. 2.8 Data Analysis Data were expressed as mean ± standard deviation, where applicable. Statistical analysis was performed using SPSS 20.0. Non-parametric Scheirer-Ray-Hare test was used for the factorial experiment, while Kruskal-Wallis or Mann-Whitney U tests were used for independent experiments to assess the significance among variables. The correlation between diacetyl and 2,3-pentanedione concentrations and those from direct-reading instruments, measured at the same spot with corresponding time frame, was assessed using Spearman’s rank correlation coefficient. The significance level (α) was set at 0.05. 3. Results and Discussion Coffee roasting is a complex process, with chemical formation influenced by variables such as bean type, roasting temperature, and time. This study conducted limited tests to explore and compare the potential impact of various roast-related factors on the generation of airborne diacetyl and 2,3-pentanedione, which may have health implications for processing workers or consumers. 3.1 Effect of Roast Degree Across Stages Mean diacetyl and 2,3-pentanedione concentrations measured for different roast degrees across stages are shown in Figure 2. The highest airborne concentrations of both diacetyl (Figure 2, top) and 2,3-pentanedione (Figure 2, bottom) were observed during the grinding stage, followed by roasting and packaging. For roast degree, dark roasting generally yielded higher diacetyl and 2,3-pentanedione concentrations than light or medium roast. Statistical analysis revealed that the roasting stage had a significant effect on diacetyl and 2,3-pentanedione concentrations (p<0.001, Scheirer–Ray–Hare test), whereas roasting degree and the interaction between both factors did not reach statistical significance (Table 1). Post-hoc comparisons among the three processes indicated that grinding had significantly higher (p<0.05) diacetyl and 2,3-pentanedione concentrations than roasting. Similarly, roasting had significantly higher (p<0.05) diacetyl and 2,3-pentanedione concentrations than packaging. Current findings align with prior results that highlights grinding as a major emission source of alpha-diketones [9, 15-21]. Diacetyl and 2,3-pentanedione concentrations increase with increasing degrees of roast (Schenker et al., 2002), and dark (French) roast was associated with the highest mass emission factor of diacetyl [9]. Table 1 Effect of different roast degree and stage on diacetyl and 2,3-pentanedione concentrations. Roast Degree Roast Stage 1 Test Statistics 2 Roasting (R) Packaging (P) Grinding (G) Factors H value p value Diacetyl Light 39.47±3.09 3 18.29±3.11 182.49±12.98 Degree 1.580 0.454 Medium 45.80±2.17 20.68±3.38 170.83±9.94 Stage 23.143 <0.001 (P<R<G) Dark 57.00±7.80 25.47±4.12 301.67±29.79 Degree*Stage 0.314 0.988 2,3-pentanedione Light 25.06±1.78 14.66±1.98 100.61±11.86 Degree 1.342 0.511 Medium 28.95±3.20 16.64±1.36 109.68±5.32 Stage 23.150 <0.001 (P<R<G) Dark 35.54±3.58 15.75±2.13 169.62±15.93 Degree*Stage 0.307 0.989 1 R (Green bean roasting); P (Packaging and storing of roasted bean); G (Grinding and packing) 2 Scheirer–Ray–Hare test & Dunn's test; α=0.05 3 Mean±SD, in ppb; n=3 TVOC levels were continuously monitored at the roasting and grinding zones in the factorial experiment, and the results from the roasting zone are shown in Figure 3. Abrupt increases in TVOC levels occurred nearly every 10 minutes after loading, coinciding with the highest roasting temperature reached and bean popping in each of the six roasting loads. TVOC levels tended to increase as the load progressed or as the roast degree increased from light to dark. Statistical analysis revealed that roasting stage had a significant effect on TVOC concentrations (p<0.001, Scheirer–Ray–Hare test), whereas roasting degree and the interaction between both factors did not reach statistical significance (Table 2). Post-hoc comparisons among the three processes indicated that grinding had significantly higher (p<0.05) TVOC concentrations than roasting. Table 2 Effect of different roast degree and stage on mean TVOC levels (ppb) Roast Degree Roast Stage Test Statistics 1 Roasting (R) Grinding (G) Factors H value p value Light 56.09±22.65 2 884.37±186.38 Degree 2.889 0.236 Medium 71.56±15.46 678.62±13.75 Stage 12.790 <0.001 (R<G) Dark 137.32±11.05 3332.95±231.21 Degree*Stage 0.105 0.949 1 Scheirer–Ray–Hare test & Dunn's test; α=0.05 2 Mean±SD; n=3 Similarly, PM 2.5 levels were continuously monitored at the roasting and grinding zones in the factorial experiment, with results from the roasting zone shown in Figure 4. PM 2.5 levels at the roasting zone generally followed the TVOC trend, showing abrupt increases every 10 minutes after loading, though the increase was not as sharp as that for TVOC. At this point (~10 min post loading), the beans had undergone dehydration and browning. Statistical analysis revealed that the roasting stage had a significant effect on PM 2.5 concentrations (p<0.001, Scheirer–Ray–Hare test), whereas roasting degree and the interaction between both factors did not reach statistical significance (Table 3). Post-hoc comparisons among the three processes indicated that grinding had significantly higher (p<0.05) PM 2.5 concentrations than roasting. Table 3 Effect of different roast degree and stage on PM 2.5 levels (μg/m 3 ) Roast Degree Roast Stage Test Statistics 1 Roasting (R) Grinding (G) Factors H value p value Light 969.57±66.63 2 54.78±14.41 Degree 2.211 0.331 Medium 1105.27±165.81 55.30±7.32 Stage 12.789 <0.001 (G<R) Dark 1164.02±38.78 160.24±25.42 Degree*Stage 0.222 0.895 1 Scheirer–Ray–Hare test & Dunn's test; α=0.05 2 Mean±SD; n=3 Interestingly, grinding had significantly higher overall mean TVOC concentrations than roasting (1131.7 vs. 88.3ppb), while roasting had significantly higher overall mean PM 2.5 aerosols than grinding (1079.6 vs. 90.1μg/m 3 ). A reasonable explanation for this discrepancy is that during roasting, green beans undergo complex chemical reactions ( i.e ., pyrolysis, Maillard reaction, Strecker degradation, and caramelization), creating a complex array of volatile compounds such as acetic acid, aldehydes, 2,3-pentanedione, pyridine, furans, and pyrazines [13]. These roasting-oriented volatiles generally have molecular weights less than 110 and boiling points less than 115 °C, and thus relatively high vapor pressures. As the roasting temperature rapidly elevates to the target of ~200 °C, these volatile chemicals are abruptly released through bean cracking/popping (Figure 3). Meanwhile, as many of these volatiles are common components or precursors of aerosols, aerosols are subsequently formed due to condensation when these vapors meet cooler room air. Such pathway is compatible with common aerosol-forming mechanism, and consequently leading to formation of high levels of PM 2.5 aerosols [23]. Conversely, in grinding, fine particles resulting from the mechanical breaking down of roasted beans, rather than aerosols, prevail. The grinding process having higher TVOC levels than roasting can be explained by grinding releasing already created but retained volatiles by breaking down the bean's porous structure, drastically increasing surface areas, and allowing for rapid off-gassing [24-25]. 3.2 Effect of Grind Fineness The diacetyl and 2,3-pentanedione concentrations measured at the grinding zone for the two bean finenesses are shown in Figure 5. The fine-ground (diameter <0.85mm) group exhibited higher mean diacetyl and 2,3-pentanedione concentrations than their coarse-ground (diameter <1.40mm) counterparts. This finding is consistent with previous results indicating that finer particles tend to create more surface areas, facilitating chemical release [26]. Nonetheless, these differences did not reach statistical significance. 3.3 Effect of Grind Quantity The diacetyl and 2,3-pentanedione concentrations at the grinding zone from roasted beans with two different grind loadings are shown in Figure 6. The 2-kg group had higher mean diacetyl and 2,3-pentanedione concentrations than the 0.5-kg group. This aligns with earlier discussions that larger particle surface areas facilitate chemical release. Grinding a larger quantity of beans will create overall more surface areas and release more TVOC. This was verified by higher TVOC levels (data not shown) measured for the 2-kg grinding group. Although the concentration differences between the two groups appeared obvious, they did not reach statistical significance (p=0.1 for both chemicals), likely due to a limited sample size. 3.4 Effect of Bean Origin The diacetyl and 2,3-pentanedione concentrations measured at the grinding zone among the three test beans (A: Ethiopia, B: Guatemala, and C: Brazil) are shown in Figure 7. Bean C exhibited the highest mean diacetyl and 2,3-pentanedione concentrations compared to beans A and B. However, statistical significance (p<0.05) for diacetyl and 2,3-pentanedione concentrations only reached between beans C and A. Notably, the test beans represent specific commercial supplies from selected countries and may not fully reflect the broader variability of the global coffee supply. Although all beans tested were of the Arabica variety, they were processed differently after harvesting; bean C was treated by a natural/sun-dried procedure, while beans A and B were washed. Differences in chemical release among beans have also been observed previously. Diacetyl and 2,3-pentanedione concentrations were higher during grinding of soft (Brazil) than during grinding of hard (Honduras) beans, which are grown at higher altitudes and thus grow slower [21]. Total dust and endotoxin exposure were significantly lower in Arabica than in Robusta (using dry pre-processing method) coffee factories [27]. 3.5 Effect of Sampling Distance The diacetyl and 2,3-pentanedione concentrations measured at the grinding zone (0 m) and nearby sampling sites with different distances, i.e., seating zone (~3 m) and bar zone (~7 m), are shown in Figure 8. It is clear that the diacetyl and 2,3-pentanedione concentrations decreased as the distance from the pollution source (grinding zone) increased. Statistical analysis indicated that diacetyl and 2,3-pentanedione concentrations at the grinding zone were 8.4 times and 4.9 times higher (p<0.05) than those in the bar zone, respectively. The chemical distribution in a room primarily depends on ventilation and diffusion. An inverse relationship between diacetyl and 2,3-pentanedione concentrations and pollution distances has also been found previously [21]. As current coffee roasting is a batch-type process, the short-term concentration gradient among sampling distances is apparent. These findings suggest that closely monitoring diacetyl and 2,3-pentanedione concentrations in seating areas is important for customer health, especially in cafés where the full roasting process is performed. 3.6 Effect of Bean Aging The diacetyl and 2,3-pentanedione concentrations at the grinding zone for processing roasted beans with different aging (storage) periods are shown in Figure 9. Both diacetyl and 2,3-pentanedione concentrations showed a decreasing trend with increasing aging time, indicating that aging after roasting may decrease the release of both chemicals. However, only 2,3-pentanedione concentrations in the 1st day group were statistically higher (p<0.05) than that of the 7th day group. Previous study also found higher emission factors of diacetyl and 2,3-pentanedione on the first day of storage, followed by a decrease [26], further strengthening the effect of storage on releasing. It is known that volatiles generated during roasting are trapped inside beans and off-gassing gradually. As the time passes, less volatiles, including alpha-diketone are retained, resulting in lower concentration when grinding. However, roast degree seemed increasing in bean porosity, which in turn affects the emission characteristics of chemicals [26]. The beans used in this sub-test was from the same processing batch, thus facilitating the comparison. On the other hand, the amount of diacetyl generated during grinding is temperature-dependent, with warmer temperatures (~40°C) creates significantly higher amount [28]. However, this study ground cooled beans that had been stored for at least one day after roasting. 3.7 Correlation of Alpha-Diketones with TVOC and PM 2.5 Concentrations Correlation analysis based on concurrent measurements of diacetyl, 2,3-pentanedione, TVOC, and PM 2.5 at roasting and grinding zones is shown in Figures 10 and 11. A high and positive correlation was found between TVOC levels and diacetyl (r=0.827, p<0.001) or 2,3-pentanedione (r=0.848, p<0.001) (Figure 10). A high correlation (r=0.974) between diacetyl and TVOC measured from all data was also found previously (Echt et al., 2021). Therefore, TVOC levels can be considered a good surrogate of exposure index for diacetyl and 2,3-pentanedione in roasting and grinding zones. Nearly no correlation was found between PM 2.5 levels and diacetyl (r=－0.207) or 2,3-pentanedione (r=－0.205) when data from roasting and grinding zones were analyzed together (Figure 11, a and b). However, a high correlation was found between PM 2.5 levels and diacetyl (r=0.823, p<0.001) or 2,3-pentanedione (r=0.836, p<0.001) if data from the grinding zone was used (Figure 11, c and d). These findings provide further evidence that the generation of fine particles (PM 2.5 ) in roasting and grinding zones may follow different mechanisms. During grinding, fine particles mainly originate from the physical breakdown of roasted beans, a pathway that also releases VOCs, including alpha-diketones, thus resulting in a positive high correlation. Nonetheless, during roasting, fine particles primarily originate from the condensation of volatile chemicals, and such conversion may not follow a linear relationship, which in turn affects the correlations between alpha-diketone and PM 2.5 aerosol levels. 3.8 Human Exposures and Health Actual human exposures to airborne chemicals depend on field concentration and contact time. Numerous agencies have established occupational exposure limits for diacetyl and 2,3-pentanedione. For instance, US NIOSH [29] has set Recommended Exposure Limits (REL) of 5 ppb and 9.3 ppb, respectively, for typical 8-hour time-weighted average (TWA) daily exposure, and 25 ppb and 31 ppb, respectively, for short-term 15-minute exposure. Previous field studies for coffee processing workplaces have consistently found higher exposures in the grinding zone, with overall diacetyl concentrations ranging from a few tenths to nearly a few hundreds of ppb and mean 2,3-pentanedione concentrations from a few to high tenths of ppb [9, 15-21]. Many of these measurements exceeded occupational exposure limits, indicating that proper control measures should be adopted. Engineering controls, such as the adoption of enclosures while ventilating the grinder, have proven effective [30]. While the results are based on a single site and specific test conditions, the concentrations measured in various zones likely reflect realistic exposure scenarios for similar small-scale or artisanal coffee operations. Based on current measurements, mean 8-hour diacetyl and 2,3-pentanedione concentrations in roasting and packaging zones generally remained below 10 ppb (diacetyl: 3.9-16.2 ppb; 2,3-pentanedione: 3.3-9.9 ppb), which approaches or exceeds the REL. The highest diacetyl and 2,3-pentanedione concentrations also occurred in the grinding zone, corresponding to short-term 15-minute exposures of 172.9-331.1 ppb (diacetyl) and 87.6-182.1 ppb (2,3-pentanedione). These levels are well above the short-term exposure limits, indicating that cafés relying on manual, open-system operations may pose a significant occupational health risk. Consumers sitting a few meters away from the grinding zone may still receive excessive exposures (Section 3.5). Nonetheless, such exposure is likely deceptively pleasant due to the coffee aroma. Moreover, exposures to gaseous diacetyl and 2,3-pentanedione have recently been linked to occupational irreversible lung disease (bronchiolitis obliterans) in the coffee processing industry. However, coffee bean dust has long been considered an occupational asthmagen [31-32] (AIHW, 2008; HSE, 2001). Sakwari and coworkers [33] reported airborne dust concentrations of 1.23 mg/m 3 (geometric mean) for coffee workers in Kilimanjaro and identified a higher prevalence of respiratory symptoms such as cough with sputum and chest tightness than controls. Although not specified, coffee bean dust may comprise particles from coffee chaff, particles from grinding roasted coffee cherry, and aerosols formed from the condensation of volatile chemicals generated during roasting, as discussed earlier. Consequently, quantification of the chemical composition of coffee bean dust would further distinguish the contribution of each component to respiratory effects. 4. Conclusions Diacetyl and 2,3-pentanedione have garnered significant attention due to their distinct aroma characteristics and potential health implications. This study systematically and controllably examined the levels of these two hazardous alpha-diketones under various coffee processing stages and variables. The obtained results are valuable for understanding the relative importance of each tested variable. While site-specific, the actual field concentrations reflect realistic situations in a small-scale café environment. 4.1 Summary of Findings The stages of roasting process significantly influenced airborne concentrations of both diacetyl and 2,3-pentanedione (i.e. , grinding > roasting > packaging). This aligns well with previous field findings, thus highlighting a critical control point for exposure. Bean origin also affected emissions. Specifically, beans from Brazil (processed by natural drying) emitted significantly more alpha-diketones than the washed beans from the other two countries. Accordingly, the importance of treatment type on alpha-diketone release warrants further investigation. Grind-related variables such as grind fineness and quantity affected emissions. Finer particles or grinding larger amounts of beans tended to increase alpha-diketone emissions, supporting the hypothesis that surface area favors VOC emissions. Furthermore, aged (days after roasting) beans released fewer alpha-diketones during grinding. Therefore, understanding all potential influencing factors may optimize grinding procedures to avoid high levels of alpha-diketones in café environments. Roasting showed significantly higher mean PM 2.5 aerosols than grinding, likely resulting from the condensation of roasting-oriented volatiles. Future studies focusing on the forming mechanism and chemical composition of these aerosols, or broadly coffee bean dust, would help distinguish their impacts on respiratory effects. 4.2 Practical Implications TVOC levels were positively and significantly correlated with both alpha-diketones in roasting and grinding zones. Therefore, routine monitoring of TVOC should be considered as a low-cost, real-time surrogate for alpha-diketone exposure assessment, beneficial both for workers and consumers. To mitigate exposures, engineering controls on source such as isolating or relocating grinding zones, utilizing localized exhaust ventilation, or employing enclosed grinding equipment should be prioritized, particularly in small-scale facilities operating under open and manual processing conditions that can potentially generate hazardous levels of alpha-diketones. Declarations Funding This work has been supported by the internal funds from National Yunlin University of Science and Technology (#109H500515). Author contributions Interpretation of data, prepare and revise manuscript (Ta-Yuan Chang); Performing experiment, sample and data analysis (Tzu-Chieh Su); Design and supervise study, finalize the content to be published, address all responses (Yeh-Chung Chien). All authors read and approved the final manuscript. Data availability All data associated with this study are reported in the manuscript. Conflict of interest The authors declare no competing interests Human Ethics and Consent to Participate declarations : not applicable References da Costa DS, Albuquerque TG, Costa HS, Bragotto APA (2023) Thermal Contaminants in Coffee Induced by Roasting: A Review. Int J Environ Res Public Health 20:5586 McCoy MJ, Hoppe Parr KA, Anderson KE, et al. (2017) Diacetyl and 2,3-pentanedione in breathing zone and area air during large-scale commercial coffee roasting, blending and grinding processes. Toxicol Rep 4:113-122 Toledo P, Pezza L, Pezza HR, Toci AT (2016) Relationship Between the Different Aspects Related to Coffee Quality and Their Volatile Compounds. Compr Rev Food Sci Food Saf 15:705-719 Poisson L, Schieberle P (2008) Characterization of the most odor-active compounds in an Arabica coffee brew. J Agric Food Chem 56:5836–5842 Hubbs AF, Cummings KJ, McKernan LT, et al. (2008) Respiratory toxicologic pathology of inhaled diacetyl in rats. Toxicol Pathol 36:330–344 Kreiss K, Gomaa A, Kullman G, et al. (2002) Clinical bronchiolitis obliterans in workers at a microwave-popcorn plant. N Engl J Med 347:330-338 Allen JG, Flanigan SS, LeBlanc M, et al. (2016) Flavoring chemicals in e-cigarettes: diacetyl, 2, 3-pentanedione, and acetoin in a sample of 51 products, including fruit-, candy-, and cocktail-flavored e-cigarettes. Environ Health Perspect 124:733-739 Hubbs AF, Kreiss K, Cummings KJ, et al. (2019) Flavorings-Related Lung Disease: A Brief Review and New Mechanistic Data. Toxicol Pathol 47:1012-1026 Duling MG, LeBouf RF, Cox-Ganser JM, et al. (2016) Environmental characterization of a coffee processing workplace with obliterative bronchiolitis in former workers. J Occup Environ Hyg 13:770-781 Bailey RL, Cox-Ganser JM, Duling MG, et al. (2015) Respiratory morbidity in a coffee processing workplace with sentinel obliterative bronchiolitis cases. Am J Ind Med 58:1235-45 Barrios-Rodríguez YF, Gutiérrez-Guzmán N, Pedreschi F, Mariotti-Celis MS (2022) Rational design of technologies for the mitigation of neo-formed contaminants in roasted coffee. Trends Food Sci Technol 120:223-235 Moon JK, Shibamoto T (2009) Role of roasting conditions in the profile of volatile flavor chemicals formed from coffee beans. J Agric Food Chem 57:5823–5831 Yeretzian C, Jordan A, Badoud R, Lindinger W (2002) From the green bean to the cup of coffee: Investigating coffee roasting by on-line monitoring of volatiles. Eur Food Res Technol 214:92–104 Caporaso N, Whitworth MB, Cui C, Fisk ID (2018) Variability of single bean coffee volatile compounds of Arabica and robusta roasted coffees analysed by SPME-GC-MS. Food Res Int 108:628-640 Echt H, Dittmore M, Coker M, et al. (2021) Characterization of Naturally Occurring Alpha-Diketone Emissions and Exposures at a Coffee Roasting Facility and Associated Retail Café. Ann Work Expo Health 65:715-726 LeBouf RF, Blackley BH, Fortner AR, et al. (2020) Exposures and Emissions in Coffee Roasting Facilities and Cafés: Diacetyl, 2,3-Pentanedione, and Other Volatile Organic Compounds. Front Public Health 8:561740 McClelland TL, Boylstein RJ, Martin SB, et al. (2019). Health Hazard Evaluation Report: HHE-2016-0109-3343, March 2019. Evaluation of exposures and respiratory health at a coffee roasting and packaging facility and two off-site retail cafés. Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services. Pengelly I, O’Shea H, Smith G, Coggins MA (2019) Measurement of diacetyl and 2, 3-pentanedione in the coffee industry using thermal desorption tubes and gas chromatography–mass spectrometry. Ann Work Expo Health 63:415-425 Stanton M, Martin S, Nett R. (2018). Health Hazard Evaluation Report: HHE-2016-0012-3302, January 2018. Evaluation of Exposures and Respiratory Health at a Coffee Roasting and Packaging Facility. Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services. Hawley B, Martin S, Duling M, Bailey R (2017) Health Hazard Evaluation Report: HHE-2016-0013-3294, October 2017. Evaluation of Exposures and Respiratory Health at a Coffee Roasting and Packaging Facility. Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services Gaffney SH, Abelmann A, Pierce JS, et al. (2015) Naturally occurring diacetyl and 2,3-pentanedione concentrations associated with roasting and grinding unflavored coffee beans in a commercial setting. Toxicol Rep 2:1171-1181 OSHA. (2008) OSHA sampling and analytical methods. Salt Lake City (UT): U.S. Department of Labor, Occupational Safety and Health Administration (OSHA). Acetoin and Diacetyl (Method 1012); 2,3-Pentanedione (Method 1016). https://www.osha.gov/chemicaldata/sampling-analytical-methods Seinfeld JH, Pandis SN. (2016) Atmospheric Chemistry and Physics: From Air Pollution to Climate Change (3rd ed.). John Wiley & Sons. Fadai NT, Melrose J, Please CP, et al. (2017) A heat and mass transfer study of coffee bean roasting. Int J Heat Mass Transf 104:787–99 Akiyama M, Murakami K, Ohtani N, et al. (2003) Analysis of volatile compounds released during the grinding of roasted coffee beans using solid-phase microextraction. J Agric Food Chem 51:1961–9 LeBouf RF, Ranpara A, Fernandez E, et al. (2022) Model Predictions of Occupational Exposures to Diacetyl and 2,3-Pentanedione Emitted From Roasted Whole Bean and Ground Coffee: Influence of Roast Level and Physical Form on Specific Emission Rates. Front Public Health 10:786924 Sakwari G, Mamuya SHD, Bråtveit M, et al. (2013) Personal Exposure to Dust and Endotoxin in Robusta and Arabica Coffee Processing Factories in Tanzania. Ann Occup Hyg 57:173–183 HSE (2023) “Exposure to diacetyl vapour in food and drink manufacture” Safety and Health Bulletin #EPD01-2023. Health and Safety Executive (HSE)/Norwich, Norfolk, UK NIOSH. (2016) Criteria for a recommended standard: occupational exposure to diacetyl and 2,3-pentanedione. Cincinnati (OH): National Institute for Occupational Safety and Health (NIOSH)/Atlanta (GA): centers for Disease Control (CDC) (Publication No. 2016-111). Stanton ML, McClelland TL, Beaty M, et al. (2022) Case Study: Efficacy of Engineering Controls in Mitigating Diacetyl and 2,3-Pentanedione Emissions During Coffee Grinding. Front Public Health 10:750289 AIHW (Australian Institute of Health and Welfare) (2008) Occupational asthma in Australia. Bulletin no. 59. Cat no. AUS 101. AIHW, Canberra HSE (2001) “Asthmagen? Critical assessments of the evidence for agents implicated in occupational asthma. Health and Safety Executive (HSE)/Norwich, Norfolk, UK Sakwari G, Bråtveit M, Mamuya SH, Moen BE (2011) Dust exposure and chronic respiratory symptoms among coffee curing workers in Kilimanjaro: a cross sectional study. BMC Pulm Med 11:54 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8615798","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":577177002,"identity":"6cb57c09-53d3-4599-bda3-b4aa9bc6c550","order_by":0,"name":"Ta-Yuan Chang","email":"","orcid":"","institution":"China Medical University","correspondingAuthor":false,"prefix":"","firstName":"Ta-Yuan","middleName":"","lastName":"Chang","suffix":""},{"id":577177003,"identity":"a5915bcf-f957-4faa-8bd2-c94ad1bf5c7a","order_by":1,"name":"Tzu-Chieh Su","email":"","orcid":"","institution":"National Yunlin University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Tzu-Chieh","middleName":"","lastName":"Su","suffix":""},{"id":577177011,"identity":"64c4b728-2e49-43ce-8bbd-6f7b930647ca","order_by":2,"name":"Yeh-Chung Chien","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5UlEQVRIiWNgGAWjYHACNoYEBgY5KOcAECcQp8UYyGBsgGs5QEgLECQ2EK3F4PjhZw8eVNxJ33Aj9/iDDzV3GPjZcwyYP7bh0XImzdwg4cyz3A038hIbZxx7xiDZ88aA4SA+LTd42CQS2w4DteQYNvM2HAaK5AC1bCOk5d/hdAOQlr9ALfbEaWk4nADWwgiyRYKAFskzaWYSCccOG84888ZwZs+xwzwSZ54VHDj7D7cWPmCISf6oOSzPdzzH4AOQIcffnrzxQcUZ3FoUDqAxeEDEAWxKYUC+AZ0xCkbBKBgFowAdAACX1WATavtAUAAAAABJRU5ErkJggg==","orcid":"","institution":"National Yunlin University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Yeh-Chung","middleName":"","lastName":"Chien","suffix":""}],"badges":[],"createdAt":"2026-01-16 08:21:43","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8615798/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8615798/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":100786655,"identity":"aaa45052-f5fb-48a6-85fd-11cc50d5abd1","added_by":"auto","created_at":"2026-01-21 11:59:59","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":673458,"visible":true,"origin":"","legend":"","description":"","filename":"paper3.docx","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/29d6e30f765e7ad606c7868d.docx"},{"id":100787118,"identity":"385bcd22-26ae-4b63-8bd7-01ac3ec7f3a4","added_by":"auto","created_at":"2026-01-21 12:00:51","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5204,"visible":true,"origin":"","legend":"","description":"","filename":"7ad3c4e9f0384b40b01b7a887ccb8710.json","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/3832f07b581fd1eaffb45f36.json"},{"id":100786636,"identity":"2a7c5bbe-8ccb-4fd5-982c-c3e6f8a84cf9","added_by":"auto","created_at":"2026-01-21 11:59:56","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":109456,"visible":true,"origin":"","legend":"","description":"","filename":"7ad3c4e9f0384b40b01b7a887ccb87101enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/80da6ddf24160c8120ce79cc.xml"},{"id":100786568,"identity":"e6386fa2-1958-4991-90cc-3d9a86348f03","added_by":"auto","created_at":"2026-01-21 11:59:47","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":128951,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/0ec49f6c41e9d7f6a178009c.png"},{"id":100786784,"identity":"85d36455-c46a-49fe-b1b0-92ccce01249f","added_by":"auto","created_at":"2026-01-21 12:00:15","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":276917,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/9251ec03166a2458e22231a9.jpeg"},{"id":100786501,"identity":"d7c87292-28e7-4e7f-9e7d-147191cbeb0c","added_by":"auto","created_at":"2026-01-21 11:59:38","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":515492,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage11.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/39f2353a4d6af001cb4ab861.jpeg"},{"id":100787023,"identity":"b1dd0923-adbd-4a73-8a69-30d50fb9e644","added_by":"auto","created_at":"2026-01-21 12:00:40","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":290023,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/0fd98cae7d1b5d53be91df7e.jpeg"},{"id":100786596,"identity":"46f38d83-12a1-4d4e-8a1f-3dc782561325","added_by":"auto","created_at":"2026-01-21 11:59:50","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":674273,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/ff47521faecca2ca3b1f3176.jpeg"},{"id":100786429,"identity":"63a77323-2a3a-461c-94d6-649c0a94596d","added_by":"auto","created_at":"2026-01-21 11:59:27","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":745240,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/e6b5e11f2c55b023b662eb0f.jpeg"},{"id":100786759,"identity":"0b1eb4bf-4f74-4c4d-856b-a8ee39fec928","added_by":"auto","created_at":"2026-01-21 12:00:11","extension":"emf","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17812,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.emf","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/449949d9dbdcc4ea3d4d01b5.emf"},{"id":100786703,"identity":"3e9be1d5-36f5-49ee-bf90-becbf16a86df","added_by":"auto","created_at":"2026-01-21 12:00:07","extension":"emf","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17900,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.emf","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/7e8f35c6bc507aa69dfde7f3.emf"},{"id":100786571,"identity":"315491cd-0f8f-435f-a03e-0fb3d8839108","added_by":"auto","created_at":"2026-01-21 11:59:48","extension":"emf","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":24140,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.emf","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/9656fdcccd1375f0d73f7897.emf"},{"id":100787143,"identity":"186470c4-9a3d-488e-9ff1-c7a645b45bc4","added_by":"auto","created_at":"2026-01-21 12:00:54","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":68501,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/043c3c8598a6d6b4e6e0dbd9.png"},{"id":100786257,"identity":"41d28ad3-f58b-4340-936c-583c2c158074","added_by":"auto","created_at":"2026-01-21 11:59:15","extension":"emf","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25280,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.emf","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/c5001a7e2fc950645ff24436.emf"},{"id":100787357,"identity":"9ca060ba-f365-4d86-9aa1-a389f8f8beab","added_by":"auto","created_at":"2026-01-21 12:01:31","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":28578,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/98a2d754deca162975c11ba2.png"},{"id":100796772,"identity":"0d87c17f-db95-4e1c-8382-fd0759b23073","added_by":"auto","created_at":"2026-01-21 13:45:41","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":67415,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/e95937a570bcb065c94ed3bf.png"},{"id":100797358,"identity":"c0d4b55c-0cf2-42dd-b72d-799df7970b10","added_by":"auto","created_at":"2026-01-21 13:49:13","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":123850,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/fb246f4f844c94a0066a28df.png"},{"id":100786715,"identity":"3508dcd1-614b-4343-b216-90ea0bfa92f6","added_by":"auto","created_at":"2026-01-21 12:00:08","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":70903,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/ecc0916b4d7ea58bd99003a1.png"},{"id":100787137,"identity":"47686215-83d6-4c3f-85ea-7e338ecc47d2","added_by":"auto","created_at":"2026-01-21 12:00:53","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":112417,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/b16d90e8037906ebc670725b.png"},{"id":100786739,"identity":"73d0d19e-88c8-43f0-aab9-68e3e171120e","added_by":"auto","created_at":"2026-01-21 12:00:10","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145517,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/ddfc021ea68c4e34e4182a0b.png"},{"id":100786471,"identity":"460f515c-e94c-434d-bd18-89a2be0efd6a","added_by":"auto","created_at":"2026-01-21 11:59:37","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5783,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/448338bf6df9e2ad1182b4ca.png"},{"id":100787030,"identity":"fb9a2e61-6ff2-4405-b874-d209e8c57dfc","added_by":"auto","created_at":"2026-01-21 12:00:41","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5402,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/ca612bdab48647f3440dfb90.png"},{"id":100787234,"identity":"55e29b73-f603-41bf-9ea8-a5041c9f091a","added_by":"auto","created_at":"2026-01-21 12:01:11","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6122,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/3853144120d5db926954d5b8.png"},{"id":100786469,"identity":"e725469e-072a-4fbd-8e05-6174650735a7","added_by":"auto","created_at":"2026-01-21 11:59:37","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14744,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/fbfbc219c2ff8619105a240f.png"},{"id":100786538,"identity":"e8d61d8d-25b9-43c7-bdd9-9d59b5c78b2c","added_by":"auto","created_at":"2026-01-21 11:59:44","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6250,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/0d958c83cbc63e46fcefbc09.png"},{"id":100787033,"identity":"29a293f5-140e-449d-b3b5-7bf3de41774e","added_by":"auto","created_at":"2026-01-21 12:00:42","extension":"xml","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":106198,"visible":true,"origin":"","legend":"","description":"","filename":"7ad3c4e9f0384b40b01b7a887ccb87101structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/ff03f28b979c1c831fbee98c.xml"},{"id":100787035,"identity":"b037a184-ffe9-45f9-a455-9860d18138b8","added_by":"auto","created_at":"2026-01-21 12:00:42","extension":"html","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":120865,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/e6c2c1665e3f198cf03ee6b7.html"},{"id":100786540,"identity":"924b9481-ae65-4321-93ea-f0190710e03f","added_by":"auto","created_at":"2026-01-21 11:59:44","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":119231,"visible":true,"origin":"","legend":"\u003cp\u003eFloor plan of the test facility\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/89506305f5730b31ce8503f0.png"},{"id":100786799,"identity":"8913ce88-7096-44f2-885b-bfd18a6dbf76","added_by":"auto","created_at":"2026-01-21 12:00:19","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":451052,"visible":true,"origin":"","legend":"\u003cp\u003eEffect of different roast degree and stage on diacetyl (top) and 2,3-pentanedione (bottom) concentrations. Error bars represent standard deviation of triplicate.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/252a9024701e25ce820b9432.png"},{"id":100786910,"identity":"25c8d7f7-505c-4163-891e-d3cdfc19f923","added_by":"auto","created_at":"2026-01-21 12:00:26","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":959060,"visible":true,"origin":"","legend":"\u003cp\u003eTVOC trends for different roast degree during roasting\u003c/p\u003e\n\u003cp\u003e(each graph represents data of triplicate measurements)\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/7a4396952464d4c7574e5916.png"},{"id":100786535,"identity":"10f84a5e-5cbf-486a-825a-c803ccc274f5","added_by":"auto","created_at":"2026-01-21 11:59:43","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1013662,"visible":true,"origin":"","legend":"\u003cp\u003ePM\u003csub\u003e2.5\u003c/sub\u003e trends for different roast degree during roasting\u003c/p\u003e\n\u003cp\u003e(each graph represents data of triplicate measurements)\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/2f0df941e89fa4d0bd2f0ff2.png"},{"id":100796712,"identity":"e6ff40c8-9a5f-42bf-8621-abe1bb43e502","added_by":"auto","created_at":"2026-01-21 13:45:15","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":96646,"visible":true,"origin":"","legend":"\u003cp\u003eEffect of different grind fineness on diacetyl and 2,3-pentanedione concentrations. Error bars represent standard deviation of triplicate.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/2b31888eead94bc4b9752c01.png"},{"id":100798065,"identity":"099233b1-5f7e-4a6d-b6ca-1b02671f10a7","added_by":"auto","created_at":"2026-01-21 13:52:29","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":94536,"visible":true,"origin":"","legend":"\u003cp\u003eEffect of grind quantity on diacetyl and 2,3-pentanedione concentrations.\u003c/p\u003e\n\u003cp\u003eError bars represent standard deviation of triplicate.\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/889c7d48a474ee246651c19f.png"},{"id":100786466,"identity":"8b8aebca-ecea-4312-a3e2-452195b29c03","added_by":"auto","created_at":"2026-01-21 11:59:36","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":106088,"visible":true,"origin":"","legend":"\u003cp\u003eEffect of bean origin on diacetyl and 2,3-pentanedione concentrations.\u003c/p\u003e\n\u003cp\u003eError bars represent standard deviation of triplicate.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/361e26eb4acec94416a05bf5.png"},{"id":100804223,"identity":"d291b3a4-616d-444a-8ba3-88ca94fe31bf","added_by":"auto","created_at":"2026-01-21 14:39:39","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":68501,"visible":true,"origin":"","legend":"\u003cp\u003eEffect of sampling distance on diacetyl and 2,3-pentanedione concentrations. Error bars represent standard deviation of triplicate.\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/ff91df4dc5f02aebdf8427ed.png"},{"id":100786428,"identity":"5b2bcfe8-6479-4296-b683-e10baa4fc962","added_by":"auto","created_at":"2026-01-21 11:59:27","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":113193,"visible":true,"origin":"","legend":"\u003cp\u003eEffect of bean aging on diacetyl and 2,3-pentanedione concentrations.\u003c/p\u003e\n\u003cp\u003eError bars represent standard deviation of triplicate.\u003c/p\u003e","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/18ff77318fe016d65046e4dd.png"},{"id":100786704,"identity":"37a92e6e-e503-4a76-9aed-1afa15e440b2","added_by":"auto","created_at":"2026-01-21 12:00:07","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":399681,"visible":true,"origin":"","legend":"\u003cp\u003eSpearman correlation between TVOC and diacetyl (a) or 2,3-pentanedione (b) concentrations at roasting and grinding zones\u003c/p\u003e","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/430a56753119727d6e54108f.png"},{"id":100786431,"identity":"be5fb5a8-9c75-495b-8273-d52b5d5139c3","added_by":"auto","created_at":"2026-01-21 11:59:27","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":734690,"visible":true,"origin":"","legend":"\u003cp\u003eSpearman correlation between PM\u003csub\u003e2.5\u003c/sub\u003e and diacetyl or 2,3-pentanedione concentrations at roasting and grinding zones\u003c/p\u003e","description":"","filename":"floatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/8e68c275fd2972f487417090.png"},{"id":101206437,"identity":"20cc4c64-58e0-45cd-ad47-639129a04be3","added_by":"auto","created_at":"2026-01-27 09:56:13","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":4995260,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8615798/v1/794d3f6c-3e45-41ee-b547-9068c7110e06.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Influences of coffee roasting stages and variables on the formation of diacetyl and 2,3-pentanedione: A field assessment in a café setting","fulltext":[{"header":"Highlights","content":"\u003cp\u003e1. This study systematically examined diacetyl and 2,3-pentanedione levels across various coffee processing variables, highlighting key emission variables such as processing stage (grinding\u0026thinsp;\u0026gt;\u0026thinsp;roasting\u0026thinsp;\u0026gt;\u0026thinsp;packaging), bean origin (Brazil\u0026thinsp;\u0026gt;\u0026thinsp;Ethiopia), sampling distance, and bean aging. Roast degree, grind size and quantity had no significant effect.\u003c/p\u003e\u003cp\u003e2. Findings reflect real-world exposures in small-scale cafe environments, with implications for worker and consumer health.\u003c/p\u003e\u003cp\u003e3. Notably, roasting produced significantly higher fine aerosol (PM) levels than grinding, identifying a new area for research.\u003c/p\u003e\u003cp\u003e4. This work bridges food science and occupational health by addressing both aroma-related compounds and airborne hazards.\u003c/p\u003e"},{"header":"1. Introduction","content":"\u003cp\u003eCoffee is one of the most widely consumed beverages globally, with its flavor and aroma primarily developed through the roasting process. During roasting, complex chemical reactions such as Maillard reactions, caramelization, and pyrolysis occur, producing a wide range of volatile organic compounds (VOCs) that contribute to coffee\u0026rsquo;s characteristic sensory properties [\u003cspan additionalcitationids=\"CR2 CR3\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Among these VOCs, 2,3-butanedione (diacetyl) and 2,3-pentanedione are particularly noteworthy due to their distinct aroma and potential health implications. While these alpha-diketones impart desirable butter-like flavors at trace levels, high concentrations have been associated with respiratory illnesses, notably bronchiolitis obliterans, in occupational settings such as commercial coffee processing factories. Similarly, workers in commercial popcorn factories have experienced obliterative bronchiolitis due to high exposures to diacetyl, an artificial buttery flavoring ingredient [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Furthermore, diacetyl and 2,3-pentanedione have been identified in flavored e-cigarettes, raising new concerns regarding consumer health [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. The presence of alpha-diketones in various environments has prompted increased regulatory and research attention in recent years [\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Additionally, \"neo-formed contaminants\" (NFCs) like acrylamide, which occur during coffee processing, have raised food safety concerns due to their potential long-term risks for consumer health [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eLimited research has systematically compared how different roasting protocols affect the formation dynamics of various coffee-derived compounds. One study found that the concentration of flavoring organics in coffees, assessed by bean extractions, can vary significantly depending on roasting parameters such as time, temperature, airflow, and the type of roasting process (e.g., drum roasting, hot air roasting, fluid bed roasting) [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. VOCs emitted during coffee roasting were stage-dependent. After an initial drying phase, concentrations of VOCs like acetic acid, acetaldehyde, and pyridine rapidly increased, reaching a maximum at a medium roast level. Sporadic bursts of some volatiles, including 2,3-pentanedione, coincided with bean popping [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. High within-batch and between-batch variability were observed among the 50 identified volatile compounds [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Field measurements in coffee processing plants or caf\u0026eacute; shops have identified the highest diacetyl and 2,3-pentanedione levels in grinding or packaging areas [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan additionalcitationids=\"CR16 CR17 CR18 CR19 CR20\" citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. However, due to the nature of field assessments, the specific influencing factors for VOC release are challenging to evaluate.\u003c/p\u003e \u003cp\u003eUnderstanding the impact of specific coffee roasting variables on the generation of various compounds in a caf\u0026eacute; setting is crucial not only for optimizing flavor profiles but also for ensuring the safety of workers or baristas in the coffee industry, and consumers/patrons. Accordingly, this study aims to analyze and compare the levels of airborne diacetyl and 2,3-pentanedione produced under different coffee roasting settings, thereby contributing to both food science and occupational/environmental health.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study Design\u003c/h2\u003e \u003cp\u003eThis study investigated the levels of airborne diacetyl and 2,3-pentanedione during various coffee roasting stages and under different variables. A factorial experiment was conducted to test the effect of roast degree, roasting stage, and their interaction. This was supplemented by independent tests examining the effects of grinding fineness, bean origin, sampling distance, grind quantity, and bean aging. All tests were performed within a single coffee establishment to minimize variability. Real-time measurements of total volatile organic compounds (TVOC) and particulates with an aerodynamic diameter less than 2.5 micrometers (PM\u003csub\u003e2.5\u003c/sub\u003e) were also performed where feasible.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Coffee Roasting Processes (Stages)\u003c/h2\u003e \u003cp\u003eTo facilitate comparisons, this study standardized coffee roasting processes into three stages:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eGreen Bean Roasting (R): Green (raw) beans were roasted to different degrees, determined by the final bean temperature: light (189\u0026ndash;192\u0026deg;C), medium (198\u0026ndash;201\u0026deg;C), and dark roast (207\u0026ndash;210\u0026deg;C). The approximate roast times were 120 minutes, 130 minutes, and 150 minutes, respectively. Before roasting, green beans were stored in a room with a temperature between 20\u0026deg;C and 25\u0026deg;C and packaged in vacuum-sealed bags to maintain their initial moisture content between 10% and 12% for optimal flavor. During roasting, 3 kg per batch of beans were loaded into the roaster and followed the roast routine. A total of 6 batches were processed to complete the entire roasting procedure, taking 2 to 2.5 hours total.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003ePackaging and Storing of Roasted Bean (P): Roasted coffee beans were then poured into an iron bucket for natural cooling (~\u0026thinsp;1.5 hours) and subsequently stored in one-pound capacity bags with one-way degassing valves. The overall procedure took approximately 2 hours.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eGrinding and Packing (G): Roasted coffee beans were ground using a coffee grinder (Model CM-520, Flying Eagle Co., Taiwan). The resulting coffee powder was then divided into quarter-pound capacity bags with degassing valves. Two kg of beans were ground for most experiments, and the overall procedure took approximately 20 minutes.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Test Variables\u003c/h2\u003e \u003cp\u003eThis study comprised a 3x3 factorial experiment (roast degree * roasting stages) and independent tests. The roast degrees included Light, Medium, and Dark roast, while the roasting stages included R, P, and G, as described above. Test variables in the independent experiments included:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eGrinding fineness: Coarse (particle size\u0026thinsp;\u0026lt;\u0026thinsp;1.40 mm) or fine (particle size\u0026thinsp;\u0026lt;\u0026thinsp;0.85 mm), achieved by grinder settings and assured by metal mesh.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eBean origin: Arabica variety (\u003cem\u003eCoffea arabica\u003c/em\u003e) beans from (A) Ethiopia (1900-2200m grown, washed), (B) Guatemala (1600-1800m grown, washed), and (C) Brazil (750-1200m grown, natural/sun-dried) were tested.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eSampling distance: Distances away from the grinding zone (0 m), seating zone (~\u0026thinsp;3 m), and bar zone (~\u0026thinsp;7 m).\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eGrind quantity: 2 kg or 0.5 kg per batch.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eBean aging: Days post-roasting: 1, 5, or 7 days.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Test Facility\u003c/h2\u003e \u003cp\u003eThe test facility was a privately-owned coffee shop located in Douliu City, central Taiwan, with dimensions of 10.5 m (length) * 8 m (width) * 4 m (height). It was divided into five major zones: roasting, grinding, packaging, seating, and bar area. The store's ventilation primarily relied on natural ventilation during sampling. The roasting zone was equipped with a commercial coffee roaster (Model GS-W6M, Giesen BV, Netherlands) capable of roasting 6 kg of beans per batch. Figure\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e illustrates the floor plan and sampling spots of the facility.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Sampling and Analysis of Diacetyl and 2,3-Pentanedione\u003c/h2\u003e \u003cp\u003eSampling and analysis of airborne diacetyl and 2,3-pentanedione generally followed OSHA 1012 and OSHA 1016 methods with minor modifications [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Samples were collected using an active pump connected to a silica gel tube (#226\u0026thinsp;\u0026minus;\u0026thinsp;183, SKC Inc., US). Samples were collected at typical operator positions, 150 cm above the ground, to mimic normal breathing zones. Sampling time was typically 2 hours, with 20 minutes for short-term sampling. The sampling flow rate was 50\u0026ndash;200 mL/min to achieve optimal analytical sensitivity. Ambient temperature and humidity were also recorded during sampling. Samples were stored in a freezer (4\u0026deg;C) and analyzed within 7 days. Sample analysis was based on separation via gas chromatography coupled with a flame ionization detector (GC-FID, Model Trace 1300, Thermo Scientific, US). Samples were first desorbed, with the aid of vibration, using 2 mL of ethanol/DI water (95:5) solution for both the front and rear tube sections. Finally, 1.0 \u0026micro;L of the aliquot was injected for analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Measurement of TVOC and PM\u003csub\u003e2.5\u003c/sub\u003e\u003c/h2\u003e \u003cp\u003eTotal volatile organic compounds (TVOC) were continuously monitored using a portable volatile gas monitor, which employed photoionization detection with a 10.6 eV UV-discharge lamp (Model ppbRAE 3000, RAE systems, US). Particulate (PM\u003csub\u003e2.5\u003c/sub\u003e) levels were continuously measured using an aerosol monitor, which utilized light-scattering laser photometry to provide aerosol mass concentration (DustTrak 8530, TSI Inc., US). Sampling locations for these instruments were adjacent to those of the silica-gel samples for comparison.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Test Procedure\u003c/h2\u003e \u003cp\u003eDue to limited resources, only selected conditions were tested to examine the potential influences of variables on diacetyl and 2,3-pentanedione formation. The test schemes generally followed the procedure outlined in Section \u003cspan refid=\"Sec4\" class=\"InternalRef\"\u003e2.2\u003c/span\u003e, but were specific to each sub-test. Each experiment was run in triplicate. All tests were performed between October 2023 and June 2024.\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eFactorial experiment (roast degree * process): Ethiopia bean (ground to coarse size) was tested. Sampling occurred at R, P, and G zones. TVOC and PM\u003csub\u003e2.5\u003c/sub\u003e levels were measured at R and G zones.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eTest for grind fineness: Ethiopia bean (medium roast), with coarse or fine particle size, was tested. Sampling occurred at the G zone.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eTest for bean origin: Three beans (Ethiopia, Guatemala, and Brazil; medium roast, and ground to coarse size) were tested. Sampling occurred at the G zone.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eTest for sampling distance: Ethiopia bean (light roast, and ground to coarse size) was tested. Sampling occurred at the G, seating, and bar zones.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eTest for bean aging: Ethiopia bean (dark roast, and ground to coarse size) was tested. Sampling occurred at the G zone.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eTest for grind quantity: Ethiopia bean (dark roast, and ground to coarse size) was tested. Sampling occurred at the G zone.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.8 Data Analysis\u003c/h2\u003e \u003cp\u003eData were expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation, where applicable. Statistical analysis was performed using SPSS 20.0. Non-parametric Scheirer-Ray-Hare test was used for the factorial experiment, while Kruskal-Wallis or Mann-Whitney U tests were used for independent experiments to assess the significance among variables. The correlation between diacetyl and 2,3-pentanedione concentrations and those from direct-reading instruments, measured at the same spot with corresponding time frame, was assessed using Spearman\u0026rsquo;s rank correlation coefficient. The significance level (α) was set at 0.05.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results and Discussion","content":"\u003cp\u003eCoffee roasting is a complex process, with chemical formation influenced by variables such as bean type, roasting temperature, and time. This study conducted limited tests to explore and compare the potential impact of various roast-related factors on the generation of airborne diacetyl and 2,3-pentanedione, which may have health implications for processing workers or consumers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.1 Effect of Roast Degree Across Stages\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMean diacetyl and 2,3-pentanedione concentrations measured for different roast degrees across stages are shown in Figure 2. The highest airborne concentrations of both diacetyl (Figure 2, top) and 2,3-pentanedione (Figure 2, bottom) were observed during the grinding stage, followed by roasting and packaging. For roast degree, dark roasting generally yielded higher diacetyl and 2,3-pentanedione concentrations than light or medium roast. Statistical analysis revealed that the roasting stage had a significant effect on diacetyl and 2,3-pentanedione concentrations (p\u0026lt;0.001, Scheirer\u0026ndash;Ray\u0026ndash;Hare test), whereas roasting degree and the interaction between both factors did not reach statistical significance (Table 1). \u003cem\u003ePost-hoc\u003c/em\u003e comparisons among the three processes indicated that grinding had significantly higher (p\u0026lt;0.05) diacetyl and 2,3-pentanedione concentrations than roasting. Similarly, roasting had significantly higher (p\u0026lt;0.05) diacetyl and 2,3-pentanedione concentrations than packaging.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCurrent findings align with prior results that highlights grinding as a major emission source of alpha-diketones [9, 15-21]. Diacetyl and 2,3-pentanedione concentrations increase with increasing degrees of roast (Schenker et al., 2002), and dark (French) roast was associated with the highest mass emission factor of diacetyl [9].\u003c/p\u003e\n\u003cp\u003eTable 1\u0026nbsp;Effect of different roast degree and stage on diacetyl and 2,3-pentanedione concentrations.\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"978\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 104px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRoast\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eDegree\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 302px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRoast Stage\u003c/strong\u003e\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd colspan=\"7\" style=\"width: 321px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTest Statistics\u003csup\u003e2\u003c/sup\u003e\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eRoasting (R)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003ePackaging (P)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eGrinding (G)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eFactors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003eH value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDiacetyl\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eLight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e39.47\u0026plusmn;3.09\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e18.29\u0026plusmn;3.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e182.49\u0026plusmn;12.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eDegree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e1.580\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e0.454\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eMedium\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e45.80\u0026plusmn;2.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e20.68\u0026plusmn;3.38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e170.83\u0026plusmn;9.94\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eStage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e23.143\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u0026lt;0.001 (P\u0026lt;R\u0026lt;G)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eDark\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e57.00\u0026plusmn;7.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e25.47\u0026plusmn;4.12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e301.67\u0026plusmn;29.79\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eDegree*Stage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.314\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e0.988\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 208px;\"\u003e\n \u003cp\u003e\u003cstrong\u003e2,3-pentanedione\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eLight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e25.06\u0026plusmn;1.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e14.66\u0026plusmn;1.98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e100.61\u0026plusmn;11.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eDegree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e1.342\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e0.511\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eMedium\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e28.95\u0026plusmn;3.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e16.64\u0026plusmn;1.36\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e109.68\u0026plusmn;5.32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eStage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e23.150\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e\u0026lt;0.001 (P\u0026lt;R\u0026lt;G)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003eDark\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e35.54\u0026plusmn;3.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 95px;\"\u003e\n \u003cp\u003e15.75\u0026plusmn;2.13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 104px;\"\u003e\n \u003cp\u003e169.62\u0026plusmn;15.93\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 9px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eDegree*Stage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e0.307\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"3\" style=\"width: 151px;\"\u003e\n \u003cp\u003e0.989\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 241px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"15\" style=\"width: 99.7323%;\"\u003e\n \u003cp\u003e\u003csup\u003e1\u003c/sup\u003e R (Green bean roasting); P (Packaging and storing of roasted bean); G (Grinding and packing)\u003c/p\u003e\n \u003cp\u003e\u003csup\u003e2\u003c/sup\u003e Scheirer\u0026ndash;Ray\u0026ndash;Hare test \u0026amp; Dunn\u0026apos;s test; \u0026alpha;=0.05\u003c/p\u003e\n \u003cp\u003e\u003csup\u003e3\u0026nbsp;\u003c/sup\u003eMean\u0026plusmn;SD, in ppb; n=3\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTVOC levels were continuously monitored at the roasting and grinding zones in the factorial experiment, and the results from the roasting zone are shown in Figure 3. Abrupt increases in TVOC levels occurred nearly every 10 minutes after loading, coinciding with the highest roasting temperature reached and bean popping in each of the six roasting loads. TVOC levels tended to increase as the load progressed or as the roast degree increased from light to dark. Statistical analysis revealed that roasting stage had a significant effect on TVOC concentrations (p\u0026lt;0.001, Scheirer\u0026ndash;Ray\u0026ndash;Hare test), whereas roasting degree and the interaction between both factors did not reach statistical significance (Table 2). \u003cem\u003ePost-hoc\u003c/em\u003e comparisons among the three processes indicated that grinding had significantly higher (p\u0026lt;0.05) TVOC concentrations than roasting.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 2\u0026nbsp;Effect of different roast degree and stage on mean TVOC levels\u0026nbsp;(ppb)\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"601\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRoast\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eDegree\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 227px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRoast Stage\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd colspan=\"4\" style=\"width: 273px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTest Statistics\u003c/strong\u003e\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eRoasting (R)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eGrinding (G)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 107px;\"\u003e\n \u003cp\u003eFactors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003eH value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 83px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u0026nbsp;\u003c/em\u003evalue\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003eLight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e56.09\u0026plusmn;22.65\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e884.37\u0026plusmn;186.38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 107px;\"\u003e\n \u003cp\u003eDegree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e2.889\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 83px;\"\u003e\n \u003cp\u003e0.236\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003eMedium\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e71.56\u0026plusmn;15.46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e678.62\u0026plusmn;13.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 107px;\"\u003e\n \u003cp\u003eStage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e12.790\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 83px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003cp\u003e(R\u0026lt;G)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003eDark\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e137.32\u0026plusmn;11.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e3332.95\u0026plusmn;231.21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 107px;\"\u003e\n \u003cp\u003eDegree*Stage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.105\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 83px;\"\u003e\n \u003cp\u003e0.949\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"10\" style=\"width: 99.8336%;\"\u003e\n \u003cp\u003e\u003csup\u003e1\u003c/sup\u003e Scheirer\u0026ndash;Ray\u0026ndash;Hare test \u0026amp; Dunn\u0026apos;s test; \u0026alpha;=0.05\u003c/p\u003e\n \u003cp\u003e\u003csup\u003e2\u0026nbsp;\u003c/sup\u003eMean\u0026plusmn;SD; n=3\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eSimilarly, PM\u003csub\u003e2.5\u003c/sub\u003e levels were continuously monitored at the roasting and grinding zones in the factorial experiment, with results from the roasting zone shown in Figure 4. PM\u003csub\u003e2.5\u003c/sub\u003e levels at the roasting zone generally followed the TVOC trend, showing abrupt increases every 10 minutes after loading, though the increase was not as sharp as that for TVOC. At this point (~10 min post loading), the beans had undergone dehydration and browning. Statistical analysis revealed that the roasting stage had a significant effect on PM\u003csub\u003e2.5\u003c/sub\u003e concentrations (p\u0026lt;0.001, Scheirer\u0026ndash;Ray\u0026ndash;Hare test), whereas roasting degree and the interaction between both factors did not reach statistical significance (Table 3). \u003cem\u003ePost-hoc\u003c/em\u003e comparisons among the three processes indicated that grinding had significantly higher (p\u0026lt;0.05) PM\u003csub\u003e2.5\u003c/sub\u003e concentrations than roasting.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTable 3\u0026nbsp;Effect of\u0026nbsp;different roast degree and stage\u0026nbsp;on\u0026nbsp;PM\u003csub\u003e2.5\u003c/sub\u003e levels (\u0026mu;g/m\u003csup\u003e3\u003c/sup\u003e)\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"601\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\" style=\"width: 87px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRoast\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eDegree\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 227px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eRoast Stage\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd colspan=\"4\" style=\"width: 260px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eTest Statistics\u003c/strong\u003e\u003csup\u003e1\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eRoasting (R)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003eGrinding (G)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 107px;\"\u003e\n \u003cp\u003eFactors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003eH value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 70px;\"\u003e\n \u003cp\u003e\u003cem\u003ep\u0026nbsp;\u003c/em\u003evalue\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003eLight\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e969.57\u0026plusmn;66.63\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e54.78\u0026plusmn;14.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 107px;\"\u003e\n \u003cp\u003eDegree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e2.211\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 70px;\"\u003e\n \u003cp\u003e0.331\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003eMedium\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e1105.27\u0026plusmn;165.81\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e55.30\u0026plusmn;7.32\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 107px;\"\u003e\n \u003cp\u003eStage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e12.789\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 70px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003cp\u003e(G\u0026lt;R)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 87px;\"\u003e\n \u003cp\u003eDark\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e1164.02\u0026plusmn;38.78\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 113px;\"\u003e\n \u003cp\u003e160.24\u0026plusmn;25.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 6px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 107px;\"\u003e\n \u003cp\u003eDegree*Stage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.222\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\" style=\"width: 70px;\"\u003e\n \u003cp\u003e0.895\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20px;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"10\" style=\"width: 99.8336%;\"\u003e\n \u003cp\u003e\u003csup\u003e1\u003c/sup\u003e Scheirer\u0026ndash;Ray\u0026ndash;Hare test \u0026amp; Dunn\u0026apos;s test; \u0026alpha;=0.05\u003c/p\u003e\n \u003cp\u003e\u003csup\u003e2\u0026nbsp;\u003c/sup\u003eMean\u0026plusmn;SD; n=3\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eInterestingly, grinding had significantly higher overall mean TVOC concentrations than roasting (1131.7 vs. 88.3ppb), while roasting had significantly higher overall mean PM\u003csub\u003e2.5\u003c/sub\u003e aerosols than grinding (1079.6 vs. 90.1\u0026mu;g/m\u003csup\u003e3\u003c/sup\u003e). A reasonable explanation for this discrepancy is that during roasting, green beans undergo complex chemical reactions (\u003cem\u003ei.e\u003c/em\u003e., pyrolysis, Maillard reaction, Strecker degradation, and caramelization), creating a complex array of volatile compounds such as acetic acid, aldehydes, 2,3-pentanedione, pyridine, furans, and pyrazines [13]. These roasting-oriented volatiles generally have molecular weights less than 110 and boiling points less than 115 \u0026deg;C, and thus relatively high vapor pressures. As the roasting temperature rapidly elevates to the target of ~200 \u0026deg;C, these volatile chemicals are abruptly released through bean cracking/popping (Figure 3). Meanwhile, as many of these volatiles are common components or precursors of aerosols, aerosols are subsequently formed due to condensation when these vapors meet cooler room air. Such pathway is compatible with common aerosol-forming mechanism, and consequently leading to formation of high levels of PM\u003csub\u003e2.5\u003c/sub\u003e aerosols [23]. Conversely, in grinding, fine particles resulting from the mechanical breaking down of roasted beans, rather than aerosols, prevail. The grinding process having higher TVOC levels than roasting can be explained by grinding releasing already created but retained volatiles by breaking down the bean\u0026apos;s porous structure, drastically increasing surface areas, and allowing for rapid off-gassing [24-25].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.2 Effect of Grind Fineness\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe diacetyl and 2,3-pentanedione concentrations measured at the grinding zone for the two bean finenesses are shown in Figure 5. The fine-ground (diameter \u0026lt;0.85mm) group exhibited higher mean diacetyl and 2,3-pentanedione concentrations than their coarse-ground (diameter \u0026lt;1.40mm) counterparts. This finding is consistent with previous results indicating that finer particles tend to create more surface areas, facilitating chemical release [26]. Nonetheless, these differences did not reach statistical significance.\u0026nbsp;\u003c/p\u003e\n\u003cp id=\"_Toc188576984\"\u003e\u003cstrong\u003e3.3 Effect of Grind Quantity\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe diacetyl and 2,3-pentanedione concentrations at the grinding zone from roasted beans with two different grind loadings are shown in Figure 6. The 2-kg group had higher mean diacetyl and 2,3-pentanedione concentrations than the 0.5-kg group. This aligns with earlier discussions that larger particle surface areas facilitate chemical release. Grinding a larger quantity of beans will create overall more surface areas and release more TVOC. This was verified by higher TVOC levels (data not shown) measured for the 2-kg grinding group. Although the concentration differences between the two groups appeared obvious, they did not reach statistical significance (p=0.1 for both chemicals), likely due to a limited sample size.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.4 Effect of Bean Origin\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe diacetyl and 2,3-pentanedione concentrations measured at the grinding zone among the three test beans (A: Ethiopia, B: Guatemala, and C: Brazil) are shown in Figure 7. Bean C exhibited the highest mean diacetyl and 2,3-pentanedione concentrations compared to beans A and B. However, statistical significance (p\u0026lt;0.05) for diacetyl and 2,3-pentanedione concentrations only reached between beans C and A. Notably, the test beans represent specific commercial supplies from selected countries and may not fully reflect the broader variability of the global coffee supply.\u003c/p\u003e\n\u003cp\u003eAlthough all beans tested were of the Arabica variety, they were processed differently after harvesting; bean C was treated by a natural/sun-dried procedure, while beans A and B were washed. Differences in chemical release among beans have also been observed previously. Diacetyl and 2,3-pentanedione concentrations were higher during grinding of soft (Brazil) than during grinding of hard (Honduras) beans, which are grown at higher altitudes and thus grow slower [21]. Total dust and endotoxin exposure were significantly lower in Arabica than in Robusta (using dry pre-processing method) coffee factories [27].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.5 Effect of Sampling Distance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe diacetyl and 2,3-pentanedione concentrations measured at the grinding zone (0 m) and nearby sampling sites with different distances, i.e., seating zone (~3 m) and bar zone (~7 m), are shown in Figure 8. It is clear that the diacetyl and 2,3-pentanedione concentrations decreased as the distance from the pollution source (grinding zone) increased. Statistical analysis indicated that diacetyl and 2,3-pentanedione concentrations at the grinding zone were 8.4 times and 4.9 times higher (p\u0026lt;0.05) than those in the bar zone, respectively. The chemical distribution in a room primarily depends on ventilation and diffusion. An inverse relationship between diacetyl and 2,3-pentanedione concentrations and pollution distances has also been found previously [21]. As current coffee roasting is a batch-type process, the short-term concentration gradient among sampling distances is apparent. These findings suggest that closely monitoring diacetyl and 2,3-pentanedione concentrations in seating areas is important for customer health, especially in caf\u0026eacute;s where the full roasting process is performed.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.6 Effect of Bean Aging \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe diacetyl and 2,3-pentanedione concentrations at the grinding zone for processing roasted beans with different aging (storage) periods are shown in Figure 9. Both diacetyl and 2,3-pentanedione concentrations showed a decreasing trend with increasing aging time, indicating that aging after roasting may decrease the release of both chemicals. However, only 2,3-pentanedione concentrations in the 1st day group were statistically higher (p\u0026lt;0.05) than that of the 7th day group. Previous study also found higher emission factors of diacetyl and 2,3-pentanedione on the first day of storage, followed by a decrease [26], further strengthening the effect of storage on releasing.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIt is known that volatiles generated during roasting are trapped inside beans and off-gassing gradually. As the time passes, less volatiles, including alpha-diketone are retained, resulting in lower concentration when grinding. However, roast degree seemed increasing in bean porosity, which in turn affects the emission characteristics of chemicals [26]. The beans used in this sub-test was from the same processing batch, thus facilitating the comparison. On the other hand, the amount of diacetyl generated during grinding is temperature-dependent, with warmer temperatures (~40\u0026deg;C) creates significantly higher amount [28]. However, this study ground cooled beans that had been stored for at least one day after roasting.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.7 Correlation of Alpha-Diketones with TVOC and PM\u003csub\u003e2.5\u003c/sub\u003e Concentrations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCorrelation analysis based on concurrent measurements of diacetyl, 2,3-pentanedione, TVOC, and PM\u003csub\u003e2.5\u003c/sub\u003e at roasting and grinding zones is shown in Figures 10 and 11. A high and positive correlation was found between TVOC levels and diacetyl (r=0.827, p\u0026lt;0.001) or 2,3-pentanedione (r=0.848, p\u0026lt;0.001) (Figure 10). A high correlation (r=0.974) between diacetyl and TVOC measured from all data was also found previously (Echt et al., 2021). Therefore, TVOC levels can be considered a good surrogate of exposure index for diacetyl and 2,3-pentanedione in roasting and grinding zones.\u003c/p\u003e\n\u003cp\u003eNearly no correlation was found between PM\u003csub\u003e2.5\u0026nbsp;\u003c/sub\u003elevels and diacetyl (r=－0.207) or 2,3-pentanedione (r=－0.205) when data from roasting and grinding zones were analyzed together (Figure 11, a and b). However, a high correlation was found between PM\u003csub\u003e2.5\u003c/sub\u003e levels and diacetyl (r=0.823, p\u0026lt;0.001) or 2,3-pentanedione (r=0.836, p\u0026lt;0.001) if data from the grinding zone was used (Figure 11, c and d). These findings provide further evidence that the generation of fine particles (PM\u003csub\u003e2.5\u003c/sub\u003e) in roasting and grinding zones may follow different mechanisms. During grinding, fine particles mainly originate from the physical breakdown of roasted beans, a pathway that also releases VOCs, including alpha-diketones, thus resulting in a positive high correlation. Nonetheless, during roasting, fine particles primarily originate from the condensation of volatile chemicals, and such conversion may not follow a linear relationship, which in turn affects the correlations between alpha-diketone and PM\u003csub\u003e2.5\u0026nbsp;\u003c/sub\u003eaerosol levels.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.8 Human Exposures and Health\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eActual human exposures to airborne chemicals depend on field concentration and contact time. Numerous agencies have established occupational exposure limits for diacetyl and 2,3-pentanedione. For instance, US NIOSH [29] has set Recommended Exposure Limits (REL) of 5 ppb and 9.3 ppb, respectively, for typical 8-hour time-weighted average (TWA) daily exposure, and 25 ppb and 31 ppb, respectively, for short-term 15-minute exposure. Previous field studies for coffee processing workplaces have consistently found higher exposures in the grinding zone, with overall diacetyl concentrations ranging from a few tenths to nearly a few hundreds of ppb and mean 2,3-pentanedione concentrations from a few to high tenths of ppb [9, 15-21]. Many of these measurements exceeded occupational exposure limits, indicating that proper control measures should be adopted. Engineering controls, such as the adoption of enclosures while ventilating the grinder, have proven effective [30].\u003c/p\u003e\n\u003cp\u003eWhile the results are based on a single site and specific test conditions, the concentrations measured in various zones likely reflect realistic exposure scenarios for similar small-scale or artisanal coffee operations. Based on current measurements, mean 8-hour diacetyl and 2,3-pentanedione concentrations in roasting and packaging zones generally remained below 10 ppb (diacetyl: 3.9-16.2 ppb; 2,3-pentanedione: 3.3-9.9 ppb), which approaches or exceeds the REL. The highest diacetyl and 2,3-pentanedione concentrations also occurred in the grinding zone, corresponding to short-term 15-minute exposures of 172.9-331.1 ppb (diacetyl) and 87.6-182.1 ppb (2,3-pentanedione). These levels are well above the short-term exposure limits, indicating that caf\u0026eacute;s relying on manual, open-system operations may pose a significant occupational health risk. Consumers sitting a few meters away from the grinding zone may still receive excessive exposures (Section 3.5). Nonetheless, such exposure is likely deceptively pleasant due to the coffee aroma.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMoreover, exposures to gaseous diacetyl and 2,3-pentanedione have recently been linked to occupational irreversible lung disease (bronchiolitis obliterans) in the coffee processing industry. However, coffee bean dust has long been considered an occupational asthmagen [31-32] (AIHW, 2008; HSE, 2001). Sakwari and coworkers [33] reported airborne dust concentrations of 1.23 mg/m\u003csup\u003e3\u003c/sup\u003e (geometric mean) for coffee workers in Kilimanjaro and identified a higher prevalence of respiratory symptoms such as cough with sputum and chest tightness than controls. Although not specified, coffee bean dust may comprise particles from coffee chaff, particles from grinding roasted coffee cherry, and aerosols formed from the condensation of volatile chemicals generated during roasting, as discussed earlier. Consequently, quantification of the chemical composition of coffee bean dust would further distinguish the contribution of each component to respiratory effects.\u003c/p\u003e"},{"header":"4. Conclusions","content":"\u003cp\u003eDiacetyl and 2,3-pentanedione have garnered significant attention due to their distinct aroma characteristics and potential health implications. This study systematically and controllably examined the levels of these two hazardous alpha-diketones under various coffee processing stages and variables. The obtained results are valuable for understanding the relative importance of each tested variable. While site-specific, the actual field concentrations reflect realistic situations in a small-scale caf\u0026eacute; environment.\u003c/p\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Summary of Findings\u003c/h2\u003e \u003cp\u003eThe stages of roasting process significantly influenced airborne concentrations of both diacetyl and 2,3-pentanedione \u003cem\u003e(i.e.\u003c/em\u003e, grinding\u0026thinsp;\u0026gt;\u0026thinsp;roasting\u0026thinsp;\u0026gt;\u0026thinsp;packaging). This aligns well with previous field findings, thus highlighting a critical control point for exposure. Bean origin also affected emissions. Specifically, beans from Brazil (processed by natural drying) emitted significantly more alpha-diketones than the washed beans from the other two countries. Accordingly, the importance of treatment type on alpha-diketone release warrants further investigation.\u003c/p\u003e \u003cp\u003eGrind-related variables such as grind fineness and quantity affected emissions. Finer particles or grinding larger amounts of beans tended to increase alpha-diketone emissions, supporting the hypothesis that surface area favors VOC emissions. Furthermore, aged (days after roasting) beans released fewer alpha-diketones during grinding. Therefore, understanding all potential influencing factors may optimize grinding procedures to avoid high levels of alpha-diketones in caf\u0026eacute; environments. Roasting showed significantly higher mean PM\u003csub\u003e2.5\u003c/sub\u003e aerosols than grinding, likely resulting from the condensation of roasting-oriented volatiles. Future studies focusing on the forming mechanism and chemical composition of these aerosols, or broadly coffee bean dust, would help distinguish their impacts on respiratory effects.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Practical Implications\u003c/h2\u003e \u003cp\u003eTVOC levels were positively and significantly correlated with both alpha-diketones in roasting and grinding zones. Therefore, routine monitoring of TVOC should be considered as a low-cost, real-time surrogate for alpha-diketone exposure assessment, beneficial both for workers and consumers.\u003c/p\u003e \u003cp\u003eTo mitigate exposures, engineering controls on source such as isolating or relocating grinding zones, utilizing localized exhaust ventilation, or employing enclosed grinding equipment should be prioritized, particularly in small-scale facilities operating under open and manual processing conditions that can potentially generate hazardous levels of alpha-diketones.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e This work has been supported by the internal funds from National Yunlin University of Science and Technology (#109H500515).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e Interpretation of data, prepare and revise manuscript (Ta-Yuan Chang); Performing experiment, sample and data analysis (Tzu-Chieh Su); Design and supervise study, finalize the content to be published, address all responses (Yeh-Chung Chien). All authors read and approved the final manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e All data associated with this study are reported in the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of interest\u003c/strong\u003e The authors declare no competing interests\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHuman Ethics and Consent to Participate declarations\u003c/strong\u003e: not applicable\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eda Costa DS, Albuquerque TG, Costa HS, Bragotto APA (2023) Thermal Contaminants in Coffee Induced by Roasting: A Review. Int J Environ Res Public Health 20:5586\u003c/li\u003e\n \u003cli\u003eMcCoy MJ, Hoppe Parr KA, Anderson KE, et al. (2017) Diacetyl and 2,3-pentanedione in breathing zone and area air during large-scale commercial coffee roasting, blending and grinding processes. Toxicol Rep 4:113-122\u003c/li\u003e\n \u003cli\u003eToledo P, Pezza L, Pezza HR, Toci AT (2016) Relationship Between the Different Aspects Related to Coffee Quality and Their Volatile Compounds. Compr Rev Food Sci Food Saf 15:705-719\u003c/li\u003e\n \u003cli\u003ePoisson L, Schieberle P (2008) Characterization of the most odor-active compounds in an Arabica coffee brew. J Agric Food Chem 56:5836\u0026ndash;5842\u003c/li\u003e\n \u003cli\u003eHubbs AF, Cummings KJ, McKernan LT, et al. (2008) Respiratory toxicologic pathology of inhaled diacetyl in rats. Toxicol Pathol 36:330\u0026ndash;344\u003c/li\u003e\n \u003cli\u003eKreiss K, Gomaa A, Kullman G, et al. (2002) Clinical bronchiolitis obliterans in workers at a microwave-popcorn plant. N Engl J Med 347:330-338\u003c/li\u003e\n \u003cli\u003eAllen JG, Flanigan SS, LeBlanc M, et al. (2016) Flavoring chemicals in e-cigarettes: diacetyl, 2, 3-pentanedione, and acetoin in a sample of 51 products, including fruit-, candy-, and cocktail-flavored e-cigarettes. Environ Health Perspect 124:733-739\u003c/li\u003e\n \u003cli\u003eHubbs AF, Kreiss K, Cummings KJ, et al. (2019) Flavorings-Related Lung Disease: A Brief Review and New Mechanistic Data. Toxicol Pathol 47:1012-1026\u003c/li\u003e\n \u003cli\u003eDuling MG, LeBouf RF, Cox-Ganser JM, et al. (2016) Environmental characterization of a coffee processing workplace with obliterative bronchiolitis in former workers. J Occup Environ Hyg 13:770-781\u003c/li\u003e\n \u003cli\u003eBailey RL, Cox-Ganser JM, Duling MG, et al. (2015) Respiratory morbidity in a coffee processing workplace with sentinel obliterative bronchiolitis cases. Am J Ind Med 58:1235-45\u003c/li\u003e\n \u003cli\u003eBarrios-Rodr\u0026iacute;guez YF, Guti\u0026eacute;rrez-Guzm\u0026aacute;n N, Pedreschi F, Mariotti-Celis MS (2022) Rational design of technologies for the mitigation of neo-formed contaminants in roasted coffee. Trends Food Sci Technol 120:223-235\u003c/li\u003e\n \u003cli\u003eMoon JK, Shibamoto T (2009) Role of roasting conditions in the profile of volatile flavor chemicals formed from coffee beans. J Agric Food Chem 57:5823\u0026ndash;5831\u003c/li\u003e\n \u003cli\u003eYeretzian C, Jordan A, Badoud R, Lindinger W (2002) From the green bean to the cup of coffee: Investigating coffee roasting by on-line monitoring of volatiles. Eur Food Res Technol 214:92\u0026ndash;104\u003c/li\u003e\n \u003cli\u003eCaporaso N, Whitworth MB, Cui C, Fisk ID (2018) Variability of single bean coffee volatile compounds of Arabica and robusta roasted coffees analysed by SPME-GC-MS. Food Res Int 108:628-640\u003c/li\u003e\n \u003cli\u003eEcht H, Dittmore M, Coker M, et al. (2021) Characterization of Naturally Occurring Alpha-Diketone Emissions and Exposures at a Coffee Roasting Facility and Associated Retail Caf\u0026eacute;. Ann Work Expo Health 65:715-726\u003c/li\u003e\n \u003cli\u003eLeBouf RF, Blackley BH, Fortner AR, et al. (2020) Exposures and Emissions in Coffee Roasting Facilities and Caf\u0026eacute;s: Diacetyl, 2,3-Pentanedione, and Other Volatile Organic Compounds. Front Public Health 8:561740\u003c/li\u003e\n \u003cli\u003eMcClelland TL, Boylstein RJ, Martin SB,\u0026nbsp;et al.\u0026nbsp;(2019). Health Hazard Evaluation Report: HHE-2016-0109-3343, March 2019. Evaluation of exposures and respiratory health at a coffee roasting and packaging facility and two off-site retail caf\u0026eacute;s. Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services.\u003c/li\u003e\n \u003cli\u003ePengelly I, O\u0026rsquo;Shea H, Smith G, Coggins MA (2019) Measurement of diacetyl and 2, 3-pentanedione in the coffee industry using thermal desorption tubes and gas chromatography\u0026ndash;mass spectrometry. Ann Work Expo Health 63:415-425\u003c/li\u003e\n \u003cli\u003eStanton M, Martin S, Nett R. (2018). Health Hazard Evaluation Report: HHE-2016-0012-3302, January 2018. Evaluation of Exposures and Respiratory Health at a Coffee Roasting and Packaging Facility. Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services.\u003c/li\u003e\n \u003cli\u003eHawley B, Martin S, Duling M, Bailey R (2017) Health Hazard Evaluation Report: HHE-2016-0013-3294, October 2017. Evaluation of Exposures and Respiratory Health at a Coffee Roasting and Packaging Facility. Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, U.S. Department of Health and Human Services\u003c/li\u003e\n \u003cli\u003eGaffney SH, Abelmann A, Pierce JS, et al. (2015) Naturally occurring diacetyl and 2,3-pentanedione concentrations associated with roasting and grinding unflavored coffee beans in a commercial setting. Toxicol Rep 2:1171-1181\u003c/li\u003e\n \u003cli\u003eOSHA. (2008) OSHA sampling and analytical methods. Salt Lake City (UT): U.S. Department of Labor, Occupational Safety and Health Administration (OSHA). Acetoin and Diacetyl (Method 1012); 2,3-Pentanedione (Method 1016). https://www.osha.gov/chemicaldata/sampling-analytical-methods\u003c/li\u003e\n \u003cli\u003eSeinfeld JH, Pandis SN. (2016) \u003cem\u003eAtmospheric Chemistry and Physics: From Air Pollution to Climate Change\u003c/em\u003e (3rd ed.). John Wiley \u0026amp; Sons.\u003c/li\u003e\n \u003cli\u003eFadai NT, Melrose J, Please CP, et al. (2017) A heat and mass transfer study of coffee bean roasting. Int J Heat Mass Transf 104:787\u0026ndash;99\u003c/li\u003e\n \u003cli\u003eAkiyama M, Murakami K, Ohtani N, et al. (2003) Analysis of volatile compounds released during the grinding of roasted coffee beans using solid-phase microextraction. J Agric Food Chem 51:1961\u0026ndash;9\u003c/li\u003e\n \u003cli\u003eLeBouf RF, Ranpara A, Fernandez E, et al. (2022) Model Predictions of Occupational Exposures to Diacetyl and 2,3-Pentanedione Emitted From Roasted Whole Bean and Ground Coffee: Influence of Roast Level and Physical Form on Specific Emission Rates. Front Public Health 10:786924\u003c/li\u003e\n \u003cli\u003eSakwari G, Mamuya SHD, Br\u0026aring;tveit M, et al. (2013) Personal Exposure to Dust and Endotoxin in Robusta and Arabica Coffee Processing Factories in Tanzania. Ann Occup Hyg 57:173\u0026ndash;183\u003c/li\u003e\n \u003cli\u003eHSE (2023) \u0026ldquo;Exposure to diacetyl vapour in food and drink manufacture\u0026rdquo; Safety and Health Bulletin #EPD01-2023. Health and Safety Executive (HSE)/Norwich, Norfolk, UK\u003c/li\u003e\n \u003cli\u003eNIOSH. (2016) Criteria for a recommended standard: occupational exposure to diacetyl and 2,3-pentanedione. Cincinnati (OH): National Institute for Occupational Safety and Health (NIOSH)/Atlanta (GA): centers for Disease Control (CDC) (Publication No. 2016-111).\u003c/li\u003e\n \u003cli\u003eStanton ML, McClelland TL, Beaty M, et al. (2022) Case Study: Efficacy of Engineering Controls in Mitigating Diacetyl and 2,3-Pentanedione Emissions During Coffee Grinding. Front Public Health 10:750289\u003c/li\u003e\n \u003cli\u003eAIHW (Australian Institute of Health and Welfare) (2008) Occupational asthma in Australia. Bulletin no. 59. Cat no. AUS 101. AIHW, Canberra\u003c/li\u003e\n \u003cli\u003eHSE (2001) \u0026ldquo;Asthmagen? Critical assessments of the evidence for agents implicated in occupational asthma. Health and Safety Executive (HSE)/Norwich, Norfolk, UK\u003c/li\u003e\n \u003cli\u003eSakwari G, Br\u0026aring;tveit M, Mamuya SH, Moen BE (2011) Dust exposure and chronic respiratory symptoms among coffee curing workers in Kilimanjaro: a cross sectional study. BMC Pulm Med 11:54\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Coffee roasting, diacetyl, 2,3-butanedione, 2,3-pentanedione, TVOC, PM2.5","lastPublishedDoi":"10.21203/rs.3.rs-8615798/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8615798/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study examined airborne alpha-diketones (diacetyl and 2,3-pentanedione), known respiratory irritants, across different coffee processing variables. A factorial design assessed roast degree and processing stage, supported by separate tests on other factors. All experiments were performed in triplicate at a single site to reduce variability. Real-time measurements of total volatile organic compounds (TVOC) and PM\u003csub\u003e2.5\u003c/sub\u003e aerosols were also conducted when possible.\u003c/p\u003e \u003cp\u003eSignificant differences (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05) in alpha-diketone concentrations were found among processing stages (grinding\u0026thinsp;\u0026gt;\u0026thinsp;roasting\u0026thinsp;\u0026gt;\u0026thinsp;packaging), but not by roast degree. Grinding emitted much higher TVOC than roasting (1131.7 vs. 88.3 ppb), while roasting produced higher PM\u003csub\u003e2.5\u003c/sub\u003e (1079.6 vs. 90.1 \u0026micro;g/m\u0026sup3;). TVOC levels strongly correlated with diacetyl (r\u0026thinsp;=\u0026thinsp;0.827) and 2,3-pentanedione (r\u0026thinsp;=\u0026thinsp;0.848). Beans from Brazil (sun-dried) released more alpha-diketones than Ethiopian (washed) beans. Grinding zone levels of diacetyl and 2,3-pentanedione were 8.4 and 4.9 times higher, respectively, than n the nearby bar area (~\u0026thinsp;7 m away). Bean aging affected 2,3-pentanedione levels, peaking on the first day after roasting. Grind size and quantity showed no significant effects.\u003c/p\u003e \u003cp\u003eThese findings highlight key factors influencing alpha-diketone emissions, and reflect realistic exposures in caf\u0026eacute; environment. Routine monitoring and source control are essential to protect worker and consumer health. Elevated PM\u003csub\u003e2.5\u003c/sub\u003e levels during roasting also merit further study on their chemical characteristics and toxicity.\u003c/p\u003e","manuscriptTitle":"Influences of coffee roasting stages and variables on the formation of diacetyl and 2,3-pentanedione: A field assessment in a café setting","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-21 11:37:47","doi":"10.21203/rs.3.rs-8615798/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"21a59915-1b0c-4a08-8dc6-3d9e3ed4ff26","owner":[],"postedDate":"January 21st, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-25T22:23:51+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-21 11:37:47","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8615798","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8615798","identity":"rs-8615798","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}