Reducing CO2-Induced Driver Drowsiness With A Sodium Hydroxide Impregnated Air Filter
Written by Francisco Sebastiano with statistical analysis by Max Welter
Everyone produces CO2 (about 1 kg per day, in fact). It’s how we breathe — Oxygen comes in and Carbon Dioxide comes out. What everyone doesn’t do is consider where that CO2 goes. When metabolically produced CO2 accumulates in indoor spaces, increasing in-air CO2 concentrations, the parasympathetic nervous system is activated causing drowsiness and lethargy [1]. This is cause for concern in the automotive industry as car cabins have small-volumes and poor ventilation, diminishing driver reaction time and creating dangerous roads when cabin CO2 concentrations get too high [2,3].
Why Don’t You Just Crack A Window?
This is the most common question I get about my research. The fundamental assumption behind it, that introducing outdoor air into the car cabin will push out the CO2 concentrated air, is correct. Opening a window will bring CO2 concentrations inside a car from dangerous levels (2500 Parts-Per-Million) to ideal levels (400 PPM). As anyone who’s ever rolled their window down while on the highway knows, however, driving with windows down can be extremely unpleasant and fuel inefficient. Even on surface streets hot or cold climates make driving with rolled down windows a non-option for many people.
Why Don’t You Use Your Car’s External Air Ventilation?
This question is slightly less common as many people don’t know how their cars air ventilation system works — I definitely didn’t. Here’s the basic run-down:
Any air that enters your car cabin first goes through the cabin air filter — a pleated piece of fibrous paper which captures dust and other ultrafine particulate matter (UFP). When your car is in recirculation mode, meaning it’s recirculating the same air repeatedly through the filter, this works quite well at reducing dust and causes UFP concentrations to near 0 [4]. When your car is in external air circulation mode, however, air only passes through the filter once or twice before reaching your lungs. This isn’t just any old air, either. Highway and road air is polluted with harmful particulate matter (dust) and vehicle emissions which can enter the cabin when external air circulation is turned on [5,6].
So Where’s The Safe Air?
This question has been the subject of Indoor Air Quality (IAQ) research for over a decade and no true solution has emerged. The most promising method of IAQ management is fractional air circulation — the process of switching between recirculation and external circulation periodically [7]. Fractional recirculation still results in high CO2 and UFP exposures but does so in an alternating fashion. The method is highly flawed and has consequently seen zero instances of automotive industry application.
Carbon Capture Technology
With carbon dioxide induced climate change coming into both the scientific and public eye, recent literature has seen a massive increase in carbon capture research [8]. Direct Air Capture (DAC), the process of directly pulling CO2 from ambient air, is one of the processes at the forefront of this increase. DAC offers a method of removing CO2 from a space without introducing external air. A 2021 study published in Heliyon used a sodium hydroxide (NaOH) sorbent to remove CO2 from a sealed work space [9]. Concentrations in the 1000 ft³ room went from 3400 to less than 2000 ppm within an hour. Though not significant enough to eliminate a ventilation system, the observed changes in concentration were projected to reduce the needed amount of externally circulated air by 50%. The study’s findings suggest that car cabins, which have volumes far smaller than the volume of a room, could be effectively scrubbed with a similar sorbent, entirely reducing the need for external air ventilation. Unable to find any studies on car-cabin DAC, I aim in this article to apply the technology to automotive air filtration systems.
CO2 Scrubber Design
Remember that cabin air filter from earlier? It turns out to be a perfect medium for CO2 scrubbing. Borrowing from the aforementioned study on workspace DAC [9] an STP cabin air filter model CAF1774 was impregnated with 20 grams (0.5 moles) of Sodium Hydroxide (NaOH). The reaction the scrubber uses goes as follows:
Reaction 1: 2NaOH(s) + CO2(g) + H2O(l) → Na2CO3(aq) + 2H2O(l)
Reaction 2: Na2CO3(aq) + CO2(g) + H20(l)→ 2NaHCO3(aq)
Overall: 2NaOH(s) + 2CO2(g) + H2O(l) → 2 NaHCO3(aq)
Methodology
Based on the reaction equation at 100% molar efficiency a scrubber would capture 1 mole of CO2 for every mole of NaOH used. Since humans breath out 0.5 moles of CO2 every 30 minutes and since the average commute is about 30 minutes [5], the filter was impregnated with 0.5 moles of NaOH and was tested for 30-minute long in-car trials. Conducted were 5 control trials with no scrubber running, 5 intervention trials with the scrubber running, and 5 intervention trials with external air circulation turned on and no scrubber running.
Safety Measures and The Human-Metabolic-Analog
Though great at scrubbing CO2 and perfectly safe when fully reacted, unreacted NaOH is extremely caustic. For that reason, a human-metabolic-analog was developed which reacted acetic acid with aqueous sodium bi-carbonate to produce CO2 at the same rate as a breathing human. The reaction was mediated by a solenoid which throttled the amount of NaHCO3 allowed into the reaction chamber. The above figure describes this process in more depth. The human-metabolic-analog’s validity was confirmed by comparison between CO2 concentration data from the car with a human occupant and CO2 data from the car with the human analog. The CO2 concentrations between the two trial sets were nearly identical, as is evident in the below graph.
Measurement
Two MHZ19B NDIR CO2 sensors were used to measure car-cabin CO2 concentrations. One sensor was placed in the back seat of the car (2004 Toyota Highlander). The other sensor was placed in the drivers seat of the car. The purpose of the layout was to measure difference in CO2 between the front and back of the car. In addition to the CO2 sensors, two DHT 22 sensors were used to measure temperature and relative humidity. Relative humidity readings were converted into absolute humidity using the absolute humidity formula. One sensor was placed inside the car on the dashboard. The other sensor was placed outside of the car on the roof. The sensor layout is described by the figure below.
One of the largest contingencies associated with alakali-based carbon capture is caustic vapor emission. The NaOH + CO2 reaction produces water, meaning that a certain amount of NaOH saturated water vapor may be picked up by the air as it passes air through the impregnated filter. An atmospheric water condenser (a cup with ice in it) was used to collect samples of the air emitted by the scrubber. These samples were tested for changes in pH with a pH meter.
Results
Each of the below trial categories is the average result of five different tests. For example, the “CO2 Scrubber On” Trial was repeated 5 times for a total of 2.5 hours of experiment time. These 2.5 hours were averaged into a singular 30 minute/1800 second plot.
The CO2 scrubber intervention reduced in-car CO2 concentrations after 30 minutes from 2250 PPM to 880 PPM. This brings the CO2 in the car within the ASHRAE recommended limit of 1000 PPM [10]. Compared to external air circulation, however, the scrubber intervention’s average CO2 concentration of 738 PPM is outperformed by the external air intervention’s 450 PPM average. The scrubber nonetheless shows promise, especially when it’s considered that the scrubber only required 20 grams of NaOH per trial and that the scrubber set up could be modified to increase the amount of air that the NaOH contacts per filter pass.
The back seat CO2 data showed similar results to the front seat data. The only important difference is that the overall CO2 concentrations in the back of the car were lower than those from the front of the car. This is expected as the car’s air does not diffuse ideally, producing a localization effect in the front of the car where CO2 is produced by the driver. Interestingly, this localization effect was not seen in the intervention trials.
The absolute humidity in the car during the analog control trials was constant, not changing at all. The absolute humidity of the external air intervention decreased as air was ventilated in, presumable because the air outside the car was dryer than the air inside the car. Most importantly, the absolute humidity during the scrubber intervention trials increased at a rate of 60 mg per minute with a correlation value r of 0.95. There was a p-value of 0.027 between the control and intervention trials. There is therefore strong evidence that the scrubber is the cause of the observed increases in humidity. This happens because the NaOH + CO2 reaction produces water. That water can enter the air circulation system when air passes through the NaOH filter.
The temperature inside of the car increased by an average of 7.3 degrees Celsius when the scrubber was turned on for 30 minutes. In contrast, the average temperature increase during the human-metabolic analog control trials was less than 1 degree Celsius (see above figure). The correlational coefficient between the amount of CO2 removed by the scrubber and the temperature inside the car was .932 (see below figure). The temperature increase occurs because NaOH + CO2 is an exothermic reaction, meaning heat is released from the scrubber into the air as it passes through the filter. In this, the scrubber functions as both a heater and a carbon capture device. In cold climates this is good. In hot climates, however, the hot air created by the scrubber would require cooling via an air conditioning system.
Estimation of Reaction Enthalpy: Three Methods
The CO2 concentration varies across front and back of the car. A model which accounts for this variation is necessary in order to calculate the total number of moles of CO2 in the car at any given time. In order to construct this model a simplified cross section of the car was made with the help of a tape measure (see above image for more details). This model was used to determine the total number of moles of CO2 in the car left after 30 minutes of the analog control and scrubber intervention trials. The analog control and scrubber intervention trials ended with 0.286 moles and 0.159 total moles of CO2 in the car respectively.
Though the molar difference in CO2 between the control and scrubber intervention trials is known, the actual amount of CO2 scrubbed by the scrubber is not. Even when fully sealed a significant amount of air exits and leaves the car. With an AER of zero the concentration of CO2 in the car after 30 minutes of driving would be around 8,000 PPM. The actual concentration is about 2200 PPM. A significant amount of scrubbed air is therefore removed from the cabin via air flow. This removal needs to be accounted for in order to get an accurate estimate of how much NaOH reacted with CO2. There are many ways to do this calculation, all of which rely on slightly different assumptions and produce slightly different results.
Method 1: Ratio Approximation
This is the simplest and most presumptive method of molar calculation. If we assume that the ratio of moles CO2 produced by the scrubber to moles CO2 left in the car at t = 30 minutes is constant, we can estimate how much CO2 was scrubbed with a general k factor. The equation is described in more depth by the above figure. This method produces an estimation of 0.289 moles CO2 scrubbed.
Method 2: Derivation
Derivation is the most complicated method of determining how much CO2 was scrubbed. The major assumptions associated with this method are a constant air exchange rate, constant CO2 production rate, and even diffusion of CO2 across the car. It involves first solving for the air exchange rate (AER) during the analog control trials, then using that AER to solve for moles CO2 scrubbed during the scrubber intervention trials. The equation used for these calculations is derived above. The method produces an estimation of 0.285 moles CO2 scrubbed, which is nearly identical to the estimation produced by method number 1.
Method 3: Empirical Formula
A 2012 study of 73 different cars produced an empirical equation which describes the relationship between CO2 concentration and AER. The equation is pictured above. This method produces an estimation of 0.31 moles CO2 scrubbed. All of the methods produced similar estimations. For the purpose of simplicity, the method number 1 estimation of 0.289 moles CO2 will be used for bond enthalpy calculations.
Bond Enthalpy Calculations
An air density of 1.222 kilograms per cubic meter, specific heat of 1kJ/kg/C, and change in temperature of 7.3 Celsius during the scrubber intervention was used to estimate an overall change in heat of 33.9 kilojoules. This makes a molar reaction enthalpy of 121 kilojoules per mole. For every mole of CO2 reacted the temperature inside of the car will increase by 26 degrees Celsius.
Atmospheric pH Test
Results from the atmospheric condensation trials suggest that the scrubber increased the amount of vaporous NaOH in the car. However, the P-Value between the atmospheric control trials and the scrubber intervention trials was 0.15 (2 tailed Pair Test). The results from this specific experiment are, therefore, not very significant. Since the methods used to obtain the samples prevented pre-dilution weight measurements, it is not possible to discern how much NaOH was emitted into the air by the scrubber. At high enough concentrations vaporous NaOH can be harmful to human health. Further research into the topic of aerated NaOH is therefore needed before an NaOH scrubber can be safely implemented indoors. NaOH emission is not an unsolvable issue, however, as NaOH could be removed from scrubbed air via dehumidification or pH neutralization. It’s nonetheless a setback which decreases the overall feasibility of the system
Discussion and Summary
These results show overall promise for the future of automotive DAC for IAQ management. Looking forward, I intend to refine the atmospheric condensation testing method so that it is more robust and closer to an academia-publishable measurement. More experimentation is also needed in the space of filter/contactor designs as the filter impregnation method used in this article didn’t fully react 0.5 moles of NaOH with 0.5 moles CO2 in the 30 minute time period. Increasing the surface area of NaOH contacting the air would help in speeding up the reaction rate. As the reaction rate speeds up, however, so will the rate of temperature increase. Improvements in scrubber efficiency will therefore need to be paired with research into methods of scrubber heat dissipation.
Feel free to contact me with questions if you’re interested in IAQ, DAC, Arduino, or anything else mentioned in this article.
All Journal Articles Referenced are available on Sci-Hub
References:
[1] M. Nawaz, H. Suleman, and A. S. Maulud, “Carbon capture and utilization: A Bibliometric analysis from 2007–2021,” Energies, vol. 15, no. 18, p. 6611, 2022.
[2] N. Hudda and S. A. Fruin, “Carbon dioxide accumulation inside vehicles: The effect of ventilation and driving conditions,” Science of The Total Environment, vol. 610–611, pp. 1448–1456, 2018.
[3] J. G. Allen, P. MacNaughton, J. G. Cedeno-Laurent, X. Cao, S. Flanigan, J. Vallarino, F. Rueda, D. Donnelly-McLay, and J. D. Spengler, “Airplane pilot flight performance on 21 maneuvers in a flight simulator under varying carbon dioxide concentrations,” Journal of Exposure Science & Environmental Epidemiology, vol. 29, no. 4, pp. 457–468, 2018.
[4] Y. Zhu, A. Eiguren-Fernandez, W. C. Hinds, and A. H. Miguel, “In-cabin commuter exposure to ultrafine particles on Los Angeles freeways,” Environmental Science & Technology, vol. 41, no. 7, pp. 2138–2145, 2007.
[5] S. Fruin, D. Westerdahl, T. Sax, C. Sioutas, and P. M. Fine, “Measurements and predictors of on-road ultrafine particle concentrations and associated pollutants in Los Angeles,” Atmospheric Environment, vol. 42, no. 2, pp. 207–219, 2008.
[6] A. J. Cohen, M. Brauer, R. Burnett, H. R. Anderson, J. Frostad, K. Estep, K. Balakrishnan, B. Brunekreef, L. Dandona, R. Dandona, V. Feigin, G. Freedman, B. Hubbell, A. Jobling, H. Kan, L. Knibbs, Y. Liu, R. Martin, L. Morawska, C. A. Pope, H. Shin, K. Straif, G. Shaddick, M. Thomas, R. van Dingenen, A. van Donkelaar, T. Vos, C. J. Murray, and M. H. Forouzanfar, “Estimates and 25-year trends of the global burden of disease attributable to ambient air pollution: An analysis of data from the global burden of diseases study 2015,” The Lancet, vol. 389, no. 10082, pp. 1907–1918, 2017.
[7] H. S. Jung, M. L. Grady, T. Victoroff, and A. L. Miller, “Simultaneously reducing CO2 and particulate exposures via fractional recirculation of Vehicle Cabin Air,” Atmospheric Environment, vol. 160, pp. 77–88, 2017.
[8] T. Vehviläinen, H. Lindholm, H. Rintamäki, R. Pääkkönen, A. Hirvonen, O. Niemi, and J. Vinha, “High indoor CO2 concentrations in an office environment increases the transcutaneous CO2 level and sleepiness during cognitive work,” Journal of Occupational and Environmental Hygiene, vol. 13, no. 1, pp. 19–29, 2016.
[9] G. S. N. V. K. S. N. Swamy, “Development of an indoor air purification system to improve ventilation and air quality,” Heliyon, vol. 7, no. 10, 2021.
[10] North Carolina Energy Office, “Energy saving fact sheet ventilation and indoor air quality — NC,” North Carolina DEQ, Apr-2010. [Online]. Available: https://deq.nc.gov/media/1440/download. [Accessed: 30-Jan-2023].