Physiology or Psychology: What Drives Human Emissions of Carbon Dioxide and Ammonia?

Humans are the primary sources of CO2 and NH3 indoors. Their emission rates may be influenced by human physiological and psychological status. This study investigated the impact of physiological and psychological engagements on the human emissions of CO2 and NH3. In a climate chamber, we measured CO2 and NH3 emissions from participants performing physical activities (walking and running at metabolic rates of 2.5 and 5 met, respectively) and psychological stimuli (meditation and cognitive tasks). Participants’ physiological responses were recorded, including the skin temperature, electrodermal activity (EDA), and heart rate, and then analyzed for their relationship with CO2 and NH3 emissions. The results showed that physiological engagement considerably elevated per-person CO2 emission rates from 19.6 (seated) to 46.9 (2.5 met) and 115.4 L/h (5 met) and NH3 emission rates from 2.7 to 5.1 and 8.3 mg/h, respectively. CO2 emissions reduced when participants stopped running, whereas NH3 emissions continued to increase owing to their distinct emission mechanisms. Psychological engagement did not significantly alter participants’ emissions of CO2 and NH3. Regression analysis revealed that CO2 emissions were predominantly correlated with heart rate, whereas NH3 emissions were mainly associated with skin temperature and EDA. These findings contribute to a deeper understanding of human metabolic emissions of CO2 and NH3.


■ INTRODUCTION
Carbon dioxide (CO 2 ) and ammonia (NH 3 ) are two important inorganic gases found in both outdoor and indoor environments.Outdoors, CO 2 , as a well-known greenhouse gas, strongly contributes to global warming and serves as a driver of climate change. 1,2NH 3 is a major player in atmospheric chemistry, combining with acidic gases to form aerosol particles with direct implications for climate feedback and public health. 3,4−7 For instance, NH 3 has been found to be associated with nanocluster formation. 8The combination of CO 2 and NH 3 alters the pH of water films in indoor aerosols and on indoor surfaces and thus has meaningful impacts on the persistence of airborne viruses 9,10 and indoor surface chemistry. 11Although in typical offices and residences, the CO 2 and NH 3 concentrations are insufficient to trigger acute health effects, 12,13 elevated levels can still lead to discomfort and affect the performance of occupants. 14,15umans are the major emitters of CO 2 and NH 3 in indoor environments.In addition to emissions from human activities such as cooking and cleaning, 16 endogenous emissions from human metabolism contribute considerably to the buildup of indoor CO 2 and NH 3 levels.Humans emit CO 2 mainly through exhalation; dermal emissions could account for only up to 3.5%. 17The emission rate depends on the metabolic rate, sex, age, and environmental parameters. 18,19−34 However, dermal emission is the dominating pathway of NH 3 emission, an order of magnitude larger than emission from breath. 35−37 Physiological and psychological engagement are linked to human metabolic processes. 38,39−42 Physiological engagement, such as doing physical exercise, elevates the emission rate of CO 2 . 23,24Performing cognitive tasks, a form of psychological engagement, has also been linked to higher CO 2 emission rates, although the evidence is limited. 43Yet, the influence of physiological and psychological engagement and their relative importance for NH 3 emissions from humans have not been well-documented.In addition, quantitative relationships between CO 2 and NH 3 emissions and human physiological responses are largely unknown.
In summary, although the relationship between physiological activities and human CO 2 emissions has been well studied, there is a scarce body of research exploring the influence of psychological engagement on metabolic CO 2 emissions from humans.Furthermore, there is a lack of research on the influence of physiological or psychological engagement on human NH 3 emissions.The concurrent measurement of human CO 2 and NH 3 emissions also remains unexplored, which is important for elucidating their distinct emission behaviors and mechanisms.In addition, the correlation between the metabolic emissions of these two gases and human physiological responses is not well-understood.To contribute to existing knowledge, this study measured human metabolic emissions of CO 2 and NH 3 in a well-controlled climate chamber occupied by human subjects.We investigated the influence of physiological factors (walking and running) and psychological factors (meditation and cognitive tasks) on emission rates.We also recorded human physiological data, including skin temperature, EDA, and heart rate during the experiments and explored the relationships between CO 2 and NH 3 emissions and physiological responses.The results have the potential to deepen our understanding of emission behaviors and mechanisms of CO 2 and NH 3 from humans, enabling more accurate modeling of human emissions and their impacts on indoor air quality.
■ METHODS Climate Chamber.A climate chamber study has been demonstrated to be an effective approach for investigating human emissions of air pollutants. 8,35,44,45We performed a series of experiments in a 62 m 3 climate-controlled chamber at the E ́cole Polytechnique Fedeŕale de Lausanne (EPFL) (Figure S1).The chamber was ventilated with 100% outdoor air that was filtered by a combination of a newly installed HEPA filter and an activated carbon molecular filter.The filtered air was distributed through a supply diffuser and exhausted via an outlet, both located in the ceiling.A dedicated Heating, Ventilation, and Air-Conditioning (HVAC) system controlled the air temperature and relative humidity inside the chamber at 24 ± 0.5 °C and 50 ± 5%, respectively.Two pedestal fans, located in the chamber corners and facing the walls, ensured efficient air mixing (Figure S1).Inside the chamber, three tablets were provided for the participants.In the experiments involving physiological engagement, the chamber was furnished with three treadmills, one table, and three chairs (Figure S1).The air change rate was controlled at 2.87 ± 0.04 h −1 .In the psychological engagement experiments, there were three tables and four chairs inside the chamber (Figure S1).In these experiments, we adjusted the air change rate to 1.44 ± 0.01 h −1 to ensure adequate signals for the gas measurement instruments.The chamber and furniture surfaces were thoroughly cleaned prior to the experiments and at regular intervals during the campaign.
Experimental Design and Procedure.We recruited four groups of three young adults in each group (age range: 19−32, BMI range: 21.5−29.3,see Table S1), labeled as G2A, G2B, G3A, and G3B.G2A and G2B were involved in the physiological engagement experiments, while the other two groups underwent psychological engagement.Each group consisted of two males and one female, except for G2B, which had one male and two females.The participants were asked to take a shower in the evening prior to the experiments, using the provided perfume-and odorant-free soap and shampoo.They were asked not to apply any personal care products.During the experiment days, the students were also asked to avoid drinking alcohol or eat spicy food, garlic, onions, and asparagus.On the experimental day, 30 min prior to entering the chamber, participants changed into shortsleeved T-shirts and shorts provided by the researchers.These new clothes were washed with perfume-and odorant-free laundry detergent immediately after purchase, tumble-dried, and sealed in individual zip-lock bags.In addition, all participants were provided with the same food and drink (a light sandwich with tomato and cheese and a bottle of water), which they finished consuming 10 min before entering the chamber.Participants were not allowed to bring any personal items into the chamber.
We performed physiological engagement experiments by asking participants to walk and run on treadmills.To investigate the influence of activity level on human emissions of CO 2 and NH 3 , we asked participants to exercise at two metabolic rates: 2.5 and 5 met.Prior to the experiments, we gathered participants for a pretest session in order to adjust the treadmills to match the nominal metabolic rate of each participant (Table S1).Each experiment lasted for 2.5 h, including a 1 h preexercise sitting session, a 1 h exercise session, and a 30 min postexercise sitting session.
In the case of psychological experiments, participants engaged in two types of activities: online-guided meditation and cognitive tasks (d2 test 46,47 and Stroop and multitasking test 48,49 ).Each experiment consisted of two 45 min free-sitting periods with a 40 min engagement session in between.Details of the experimental design and procedure can be found in Section S1 and Table S2.
Instrumentation and Quality Control.CO 2 concentration inside the well-mixed chamber was measured by a highaccuracy gas analyzer at 0.5 s time intervals (LI-850, LI-COR Biosciences GmbH, Germany, range: 0−20,000 ppm, accuracy: ±1.5%).An air pump (AirCheck TOUCH, SKC Inc., U.K.) drew chamber air at 0.75 L/min into the gas analyzer.The NH 3 level was monitored by a high-precision NH 3 analyzer at 30 s time intervals (LSE NH 3 -1700 Analyzer, LSE Monitors, Netherlands, range: 0−15 ppm, noise: 1 ppb, precision: 2 ppb).The sampling flow rate was 140 mL/min.Both gas analyzers were placed outside the chamber, with sampling lines passing through a hole on the chamber wall at a height of 1 m.The NH 3 sampling line was as short as possible (10 cm) to minimize NH 3 loss in the sampling tube, whereas the length of the CO 2 sampling line was ∼0.5 m.Inside the chamber, there were two air temperature and humidity sensors (HOBO U12− 012, Onset Computer Co.) placed on the table.

Environmental Science & Technology
In addition to measuring gases and environmental conditions, we equipped participants with wearable wristbands (E4, Empatica Inc.) to measure their physiological data.The wristbands recorded skin temperature (recording frequency: 4 Hz), electrodermal activity (EDA, 4 Hz), and heart rate (1 Hz).All participants were asked to put on wristbands on their left hands before entering the chamber.
All of the instruments were calibrated before the experimental campaign.As shown in Table S2, each experiment had one replicate, and the differences observed were generally within 15%, indicating good reproducibility of the experimental results.
Data Analysis.We calculated per-person emission rates of CO 2 and NH 3 based on the material-balance equation, as described in Section S2.For statistical analysis and comparisons, the average emission rates were calculated for three subsessions of each experiment: before engagement, during engagement, and after engagement.The physiological data for each participant were averaged across the last 15 min of each subsession, as this period was found to approximate steady-state conditions of the activities.We performed paired t tests to examine the significance of the difference in emission rates and physiological data before, during, and after engagement.In addition, to investigate the relationship between gas emissions and human physiological data, we calculated 15 min averages of all parameters for the entire set of experiments.Subsequently, the 15 min average emission rates, physiological data, and chamber temperature and humidity during full experiments were analyzed by multilinear regression using MATLAB (2022b).

Characteristics of Gas Emissions and Physiological
Data. Figure 1 shows a time-series of NH 3 and CO 2 mixing ratios and human physiological data during the experiments of physiological engagement involving walking at 2.5 met.An instant increase of the CO 2 and NH 3 levels was observed after participants entered the chamber, indicating that humans are a potent source of these two gases.The physiological data gradually stabilized as CO 2 and NH 3 reached a steady state while participants remained seated.An immediate increase in CO 2 levels occurred after the walking session began.However, the NH 3 level started to climb up sharply only 30 min after participants began to walk.This difference demonstrates the distinct mechanisms underlying CO 2 (respiratory) and NH 3 (dermal) emissions.
−52 Two mechanisms govern skin temperature variations during physiological exercise: peripheral vasoconstriction and vasodilation.The former contributes to the initial decrease in skin temperature while the latter increases the skin temperature.Both occur in response to changes in blood supply related to thermoregulation and increased metabolic heat production. 53,54During 2.5 met walking, peripheral vasoconstriction likely dominated.EDA and heart rate level increased, which is also consistent with previous findings. 42After the participants stopped exercising, the levels of CO 2 inside the chamber decreased, whereas NH 3 continued climbing for another 15 min and then gradually decreased.The skin temperature progressively increased, and EDA declined to prewalking levels.The heart rate rapidly returned to the value before walking,

Environmental Science & Technology
When the participants were physiologically engaged at a higher metabolic rate (5 met), the trends in indoor CO 2 and NH 3 levels, as well as human physiological data, were generally similar to those at 2.5 met but with considerably higher levels (Figure 2).After the participants stopped running, the NH 3 level kept climbing and did not decrease until they exited the chamber.The skin temperature dropped rapidly at the beginning of running and then increased to higher level relative to prerunning.This indicates that the peripheral vasodilation tended to dominate during the 5 met running to cope with excess heat production. 47,48The skin temperature decreased slowly after the engagement stopped.The differences in the levels of CO 2 and NH 3 , as well as physiological indicators remained pronounced between the two groups (see also Figures S4−S5).
In meditation experiments, CO 2 and NH 3 concentrations did not vary as much as in the physiological engagement experiments, as shown in Figure 3 (see also Figures S6−S7).CO 2 concentration increased and reached a steady state almost at the end of the experiment.NH 3 exhibited a similar trend initially but slightly decreased after 1 h in the chamber, starting during the meditation period.The skin temperature generally followed a descending trend, especially for group G3A, in line with the fact that most participants felt slightly cool after settling down in the chamber (Figure S10), similar to previous findings. 56,57EDA levels reached a low value within 30 min after the participants entered the chamber and remained low during the rest of the experiment.The heart rate showed large variations, especially for group G3A, which was mainly caused by one participant, as made evident by the large standard deviation.
Cognitive tasks contributed to similar CO 2 and NH 3 concentrations as meditation, as illustrated in Figure 4 (see also Figures S8−S9).There were no meaningful changes in the CO 2 and NH 3 levels after the engagement started or ended.The skin temperature and heart rate also showed a descending trend and large variations, respectively.Notably, EDA levels exhibited a discernible difference: they slightly rose when the participants started cognitive tasks and fell back when the tasks ended.This echoes previous findings that cognitive stress induces changes in human EDA levels. 40mission Rate: Statistics and Comparisons.Figure 5 summarizes the average CO 2 and NH 3 emission rates and average steady-state human physiological data across all of the experiments with physiological engagement.The average CO 2 emission rate when the participants were seated before exercise was 19.6 L/h per person, which is slightly higher than the average per-person CO 2 generation rate in offices and conference rooms (17.3 L/h), 18 probably owing to two participants with large BMI (29.3 and 25.6 kg/m 2 , respectively) in G2A (Table S1).The participants generated ∼2× and 5× more CO 2 when walking at 2.5 met and running at 5 met, respectively.The increase was proportional to the designed metabolic rate.30 min after walking, the CO 2 emission rates returned to the prewalking level.Postrunning CO 2 emission rates were slightly higher than the prerunning rates but without statistical significance (average: 25.2 vs 20.0 L/h, p = 0.09).The average per-person NH 3 emission rate before running was 2.7 mg/h, within the range reported by Li et al. (0.57−5.2 mg/h for short-clothing scenarios). 35The NH 3 emission rate increased ∼2× and 3× when participants exercised at 2.5 and 5 met, respectively.Unlike CO 2 , participants continued to emit NH 3 at an elevated rate for 30 min after exercise.Regarding physiological data, the skin temperature significantly dropped during 2.5 met walking and returned to the previous level 30 min after the exercise session

Environmental Science & Technology
ended.In contrast, during 5 met running, the skin temperature significantly increased and remained high after running.Both EDA and heart rate sharply increased during exercise.30 min after exercise, they decreased but remained significantly higher than the preexercise values, except for the heart rate after 2.5 met walking.However, it is foreseen that the EDA and heart rate would return to preexercise levels if participants had rested longer, given the steep decay of these two physiological signals after exercise stopped (Figures 1 and 2).
Both CO 2 and NH 3 emission rates in the psychological engagement experiments were within the range reported in the literature.We did not observe a significant change in CO 2 generation during meditation or cognitive tasks relative to the preengagement session.However, when comparing between the two types of engagement, a slightly higher (p < 0.05) CO 2 generation was found during cognitive tasks relative to meditation (Figure 6).The finding is in accordance with the study of Gall et al. 43 that compared human CO 2 emissions between relaxed and stressed activities.Nonetheless, it should be noted that their relaxed and stressed sessions were performed sequentially in a consecutive experiment, whereas this study investigated the two types of psychological engagement in separate experimental runs.A significant increase in EDA signals was found during cognitive tasks compared to before those tasks, as well as during meditation.NH 3 emission rates decreased with time, following a similar trend with skin temperature reflecting the thermoregulation of the participants.Surprisingly, the heart rate did not exhibit a meaningful relationship with psychological engagement, except that after cognitive tasks, when the heart rate significantly decreased relative to during and before the engagement.To summarize, physiological engagement significantly elevated human emissions of CO 2 and NH 3 , while the influence of psychological engagement was not significant.
The whole-body CO 2 emission rates from sedentary humans have been widely reported in the literature.The average perperson CO 2 emission rate during the preengagement sitting period across all experiments in this study was 19.0 ± 1.3 L/h (BMI: 24.1 ± 2.8 kg/m 2 ).The value was somewhat higher than those from other chamber studies, such as 14.1 ± 3.3 L/h from Kuga et al., 19 12.3 ± 1.7 L/h from Qi et al., 22 16.8 ± 0.7 L/h from Gall et al., 43 and 16.1 ± 0.8 L/h from Sakamoto et al. 58 and Li et al. 17 In addition, Sakamoto et al. found higher CO 2 emissions in the afternoon relative to the morning due to the increased metabolism from diet-induced thermogenesis. 58e did not observe such a difference, probably because we had two distinct groups of participants in the morning and afternoon sessions, and because the controlled diet and time of food consumption right before the experiments were the same in both sessions.Nevertheless, the CO 2 emission rates in this study were within the range reported for sedentary adults (12.3−23L/h per person). 18,22,59Moreover, the proportional

Environmental Science & Technology
increment of CO 2 emission rate with increased metabolism (2.5 and 5 met exercise) echoes the linear relationship between the metabolic rate and the CO 2 generation rate in the literature. 18Data on whole-body NH 3 emission rate from humans are scarce at present.The average per-person NH 3 emission rate during the preengagement sitting period across all of the experiments was 2.0 ± 0.9 mg/h, within the range reported by Li et al. for sedentary participants. 35To our knowledge, this study was the first to report human NH 3 emissions during physiological and psychological engagements and to demonstrate that increased metabolism can elevate human emissions of NH 3 .
Correlation and Regression.Previous findings in this study have demonstrated that CO 2 and NH 3 emissions are both related to human metabolic processes, although they have distinct emission mechanisms.Figure 7 shows the correlations between CO 2 and NH 3 emissions across all of the experiments.Even though the overall correlation was moderate (Pearson's r = 0.48), when correlating the data separately during and after physiological and psychological engagement, we obtained strong and significant correlations between the emissions of the two gases.Such a correlation illustrates the "delayed" emission of NH 3 relative to CO 2 and reflects that the response to changes in human physiological status may take longer for dermal emissions relative to respiratory emissions.It may also be related to their difference in internal metabolism in human bodies.CO 2 is a product generated from cellular respiration, after which it is transported in the bloodstream to the lungs and then expelled from the body via exhalation.In comparison,

Environmental Science & Technology
NH 3 originates from protein metabolism and moves to the liver by the bloodstream, where it is converted to urea.NH 3 remaining in the blood can diffuse through the skin or be emitted in sweat. 60The conversion of NH 3 by the livers may take a longer time relative to the instant expulsion of CO 2 by the lungs, leading to the "delayed" emission of NH 3 .In addition, NH 3 excreted with sweating can be continuously emitted along with the evaporation of perspiration after exercise, which may also contribute to the "delayed" emissions.This may also be attributed to NH 3 's stickiness, resulting in a longer equilibration time inside the chamber.−11 Regression analysis conducted across all of the experiments further highlights the differences between CO 2 and NH 3 emissions from humans, as seen in Table 1: 86% of the variability in CO 2 emission rates could be explained by skin temperature, EDA, heart rate, air temperature, and relative humidity, with the heart rate being the most significant predictor, directly linked to the human metabolic rate.For NH 3 emission rates, 58% of the variability could be explained by the same set of variables, with skin temperature and EDA potentially playing an important role (Table 1).Both of these are physiological indicators related to skin properties.However, given the lower R 2 and the complexity of NH 3 metabolism and emissions, the results should be interpreted with caution and a more detailed investigation of the biological processes affecting NH 3 emissions is warranted.Although the selection of the data-averaging interval can influence the exact values of the coefficients (Table S3), the results remained similar when using average values for the whole periods of each subsession.
Limitations.Several limitations should be acknowledged when interpreting the results.NH 3 is known as a sticky gas that can be absorbed on chamber surfaces including walls, furniture, and human surfaces.This property may introduce bias to gasphase NH 3 measurements and emission rate calculation.The absorbed amount of NH 3 onto surfaces depends on the gasphase NH 3 concentration (Section S3 and Figure S13), surface-bound NH 3 (Section S3), surface properties, and air temperature and humidity. 61Given the case-specific NH 3 absorption properties, obtaining a simple correction factor to adjust the measured NH 3 or calculated emission rates is not feasible.We examined the uncertainty caused by the absorption/desorption processes of NH 3 , which was found to be within 13% (Section S3 and Table S4).In addition, NH 3 emission rates reported in this study can be considered as a "net" rate, including both NH 3 emission from and deposition onto humans.
Another potential source of bias to the emission rate calculation was the assumption of constant outdoor CO 2 and NH 3 levels during each experiment.Due to the limited number of instruments, we approximated the outdoor CO 2 and NH 3 levels using the 20 min average concentrations before participants entered the chamber and assumed them to be constant during the experiment.As we did not notice significant aperiodic variations of indoor CO 2 and NH 3 levels that were potentially caused by outdoor fluctuation, we expect that the assumption of a constant outdoor level during the experiments was justified.
Males and females have different chemical emission rates. 23his study, however, considered the average emissions in a mixed group and thus negated the potential influence of sex.In the experiments involving physiological engagement, we shortened the after-exercise subsession to 30 min considering the comfort of the participants after potential sweating.This period was insufficient for gaseous emissions and physiological data to return to preexercise levels, especially after 5 met running.In the experiments involving psychological engagement, the participants' activities on the tablets were not strictly regulated before or after the engagement, which could bring some uncertainties in the during−before and after−during comparisons.Finally, air temperature and relative humidity were controlled within a narrow range in this study, which may have caused their insignificant predictive power in the linear regression models for CO 2 and NH 3 emissions (Table 1).Previous studies have demonstrated the influence of air temperature on CO 2 and NH 3 emissions. 26,35Hence, regression results presented in this study should be applied with caution.
Implications and Future Outlook.Understanding the human emission rates of CO 2 and NH 3 is critical for indoor ventilation control and for ensuring acceptable indoor air quality for occupants.The findings of this study contribute to the knowledge of human emissions of air pollutants and the significance of human physiological and psychological factors, especially in the case of NH 3 , which has been relatively understudied.This knowledge lays the groundwork for constructing mathematical models for human-associated gas emissions that can be extrapolated to various indoor environments.These findings provide valuable insights into the dynamics of gas emissions and physiological responses during various engagement activities, shedding light on the complex interactions between human activities and indoor air quality.
−66 Future work exploring the effect of human gaseous emissions on personal exposure is warranted.
NH 3 is the dominant neutralizer of acidity in indoor environments for airborne particles, aqueous surface films, and water bulk.Human emissions of NH 3 are typically sufficient to neutralize the acidifying effects of exhaled CO 2 . 5However, given the strong influence of personal (e.g., clothing coverage) The regression used 15 min of average data from all experiments (152 data points in total, both before, during, and after physiological and psychological engagement).*p < 0.05, **p < 0.01, and ***p < 0.001.

Environmental Science & Technology
and environmental factors (e.g., air temperature) on human NH 3 emissions 35 and the inconsistent correlation between NH 3 and CO 2 emissions (Figure 7), future studies should consider broader emission scenarios to investigate the influence of human-emitted NH 3 and CO 2 on indoor acid− base balance and chemistry.
Human physiological data serve as indicators of a dynamic human−environment interaction.For instance, the skin temperature reflects the balance between human heat production and the thermal environment based on thermoregulation.The correlation between human gaseous emissions and physiological data reminds us of the importance of considering the interplay among the environment, human physiology, perception, and human emissions.Therefore, future studies focused on human emissions of air pollutants should not only consider environmental parameters and chemicals 16,67−70 but also physiological indicators.Moreover, while this study has established preliminary correlations between gas emissions and human physiological indicators, the regression results should be interpreted with caution, given the complexity of metabolism and emission of the two gases.More detailed investigations into the biological, physical, and chemical processes associated with CO 2 and NH 3 emissions are warranted.In addition, the role of human perception of indoor environments (e.g., thermal and acoustics) in human emissions and consequent indoor air quality merits closer attention.

* sı Supporting Information
The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.est.3c07659.Details of experimental procedure (Section S1); calculation of CO 2 and NH 3 emission rates (Section S2); discussion about NH 3 deposition onto chamber surfaces (Section S3); schematic layout of the climate chamber (Figure S1); time-series plot of measured parameters in the experiment of 2.5 met walking on August 19, 2022 (Figure S2); time-series plot of measured parameters in the experiment of 2.5 met walking on August 22, 2022 (replicate) (Figure S3); time-series plot of measured parameters in the experiment of 5 met running on August 23, 2022 (Figure S4); time-series plot of measured parameters in the experiment of 5 met running on August 24, 2022 (replicate) (Figure S5); time-series plot of measured parameters in the experiment of meditation on August 26, 2022 (Figure S6); time-series plot of measured parameters in the experiment of meditation on August 30, 2022 (replicate) (Figure S7); time-series plot of measured parameters in the experiment of cognitive tasks on August 29, 2022 (Figure S8); time-series plot of measured parameters in the experiment of cognitive tasks on August 31, 2022 (replicate) (Figure S9); thermal perception of participants collected 30 min after entering the chamber across all of the experiments (Figure S10); self-reported stress level from participants immediately before and after the 40 min psychological engagement session (Figure S11); d2 test sheet used in the psychological engagement experiments (Figure S12); difference between the NH 3 removal rate and air change rate in relation to the average NH 3 level during the occupied period in each experiment (Figure S13); physiological data of participants in each group and treadmill settings for each participant in the physiological engagement experiments (Table S1); summary of experimental conditions and associated average CO 2 and NH 3 emission rates, human physiological data, and chamber temperature and humidity (Table S2); multilinear regression coefficients for CO 2 and NH 3 emission rates with human physiological data and air temperature and relative humidity (Table S3); and NH 3 removal rate in each experiment after participants left the chamber and comparison with air change rate (Table S4

Figure 1 .
Figure 1.Time-series of NH 3 and CO 2 concentrations (top) and physiological data from participants: skin temperature and EDA (middle), and heart rate (bottom) in the experiments of physiological engagement by walking at 2.5 met.Morning and afternoon experiments were performed with participant groups G2A and G2B, respectively.The lines of the physiological data represent averages of all three participants in each group.Shaded areas represent standard deviation.Missing CO 2 data represent periods of direct breath sampling (Section S1).Data are presented based on a single experimental day (date: August 19, 2022).

Figure 2 .
Figure 2. Time-series of NH 3 and CO 2 concentrations (top) and physiological data from participants: skin temperature and EDA (middle) and heart rate (bottom) in the experiments of physiological engagement by running at 5 met.Morning and afternoon experiments were performed with participant groups G2A and G2B, respectively.The lines of the physiological data represent averages of all three participants in each group.Shaded areas represent the standard deviation.Missing CO 2 data represent periods of direct breath sampling (Section S1).Data are presented based on a single experimental day (date: August 23, 2022).

Figure 3 .
Figure 3. Time-series of NH 3 and CO 2 concentrations (top) and physiological data from participants: skin temperature and EDA (middle) and heart rate (bottom) in the experiments of psychological engagement by meditation.The morning and afternoon experiments were performed with participant groups G3A and G3B, respectively.The lines of the physiological data represent averages of all three participants in each group.Shaded areas represent standard deviation.Missing CO 2 data represent periods of direct breath sampling (Section S1).Data are presented based on a single experimental day (date: August 26, 2022).

Figure 4 .
Figure 4. Time-series of NH 3 and CO 2 concentrations (top) and physiological data from participants: skin temperature and EDA (middle) and heart rate (bottom) in the experiments of psychological engagement by cognitive tasks.The morning and afternoon experiments were performed with participant groups G3A and G3B, respectively.The lines of the physiological data represent averages of all three participants in each group.Shaded areas represent standard deviation.Missing CO 2 data represent periods of direct breath sampling (Section S1).Data are presented based on a single experimental day (date: August 31, 2022).

Figure 5 .
Figure 5. CO 2 and NH 3 emission rates (top) and human physiological data (bottom) before (1 h), during (1 h), and after (30 min) physiological engagement.The emission rates represent the average values across all respective subsessions (including replicates).The physiological data represent the average values of the last 15 min across all respective subsessions (including replicates), when they approximately reached steady state.The asterisks "*" indicate that the during−before or after−during difference was significant (p < 0.05).The dots "•" indicate that the after− before difference was significant (p < 0.05).

Figure 6 .
Figure 6.CO 2 and NH 3 emission rates (top) and human physiological data (bottom) before (45 min), during (40 min), and after (45 min) psychological engagement.The emission rates represent the average values across all respective subsessions (including replicates).The physiological data represent the average values of the last 15 min across all respective subsessions (including replicates) when they approximately reached steady state.The asterisks "*" indicate that the during−before or after−during difference was significant (p < 0.05).The dots "•" indicate that the after−before difference was significant (p < 0.05).

Figure 7 .
Figure 7. Pearson correlations between CO 2 and NH 3 emission rates before, during, and after physiological and psychological engagement.The data points include all emission rates obtained per subsession (48 in total).***p < 0.001.