Greenhouse gas capture by triboelectric charging

A B S T R A C


Introduction
The Greenhouse effect keeps temperatures on Earth in a range where life can flourish.However, the observed increase in the atmospheric concentration of greenhouse gases since the Industrial Revolution may lead to temperatures that with time will jeopardize current ecosystems.Effective means of controlling the atmospheric content of greenhouse gases are therefore in high demand.We have recently shown that experiments simulating wind-mediated triboelectric charging of grains in a Martian environment lead to excitation energies sufficient to dissociate molecules like methane [1].The dissociation products are highly reactive and provide a potential pathway for removal of methane on the planet.This finding has motivated us to investigate if triboelectric charging can be instrumental in removing greenhouse gases from the Earth's atmosphere and thus mitigate climate change.We consider the major greenhouse gases [2], CO 2 , CH 4 , and N 2 O.The global average concentration of these gases for 2019 are [3] 410.5 ± 0.2 ppm (CO 2 ), 1877 ± 2 ppb (CH 4 ), and 332 ± 0.1 ppb (N 2 O).The concentration increase of these gases since the Industrial Revolution is linked to global warming [4].Currently, much attention is given to limit the global temperature increase to a few degrees by reducing emissions of greenhouse gases.Such reductions face a huge challenge, making capture of greenhouse gases, already in the atmosphere, a useful addition to counteract temperature increase.

Experimental
Fig. 1 shows the apparatus used in the investigation.It consists of two units.One unit (A in Fig. 1) is a 20 cm long cylindrical quartz ampoule with a diameter of 2.5 cm.The ampoule contains 10 g of quartz grains (Merck, 1.07536) and 972 mbar of a gas mixture consisting of 20.8 % O 2 , 78.7 % N 2 and 0.5 % of one of the greenhouse gases, CO 2 , CH 4 , and N 2 O.We have chosen a common concentration of the greenhouse gases in our experiments for ease of comparison.The other unit in the apparatus (B in Fig. 1) is a quartz ampoule with two CaF 2 windows.Unit B is designed to fit into an IR spectrometer.Units A and B are connected with a quartz pipe with a quartz filter, which allows free flow of the enclosed gases but prevents the transfer of quartz grains between the two units.The quartz grains are activated by rotating (tumbling) the apparatus end over end at 30 RPM in a turning wheel shown in the Supplementary Material.In the rotation process, the quartz grains rub against each other and the walls of the ampoule.The tumbling erodes the grains and thereby exposes the ambient gas to pristine quartz surfaces.At the same time, the triboelectric effect [5][6][7][8][9][10] charges the grains.Previous experiments have shown that this activation process increases the total surface area of the quartz grains linearly with tumbling time [11].Exchange of charge between the grains can lead to excitation energies high enough to dissociate methane to CH 3 , CH 2 , and CH and ionize it to CH 4 + .CO 2 and N 2 O have ionization energies lower than that of methane and are expected to be susceptible to dissociation/ionization as well.The quartz grains have been prepared in an elaborate way (See Supplementary Material) to ensure they correspond to a well-defined and reproducible state.However, we have found that pristine surface sand behaves in much the same way as quartz.In a typical experiment, the apparatus is tumbled for a couple of weeks.The tumbling is interrupted for an hour on some days to record the IR spectrum of the gas in unit B

Results
Below we present the results of the experiments with the three greenhouse gases.According to Lambert-Beer's law, the absorbance of infrared radiation by the gases is proportional to their concentration.We have recorded the IR spectrum of the gas phase before the mechanical activation starts and during the activation in the following days.Fig. 2a  and b show the IR absorption from CO 2 in the ampoule, which contains 0.5% of CO 2 in a dry atmosphere initially.Fig. 2a shows the absorption spectrum associated with the asymmetric stretch mode of CO 2 around 2350 cm − 1 at selected days, while Fig. 2b shows the absorbance of CO 2 at 2360 cm − 1 as a function of time.Fig. 2b shows that the CO 2 content is reduced to 0.6 % of its initial value after 13 days of tumbling and further tumbling reduces the CO 2 concentration below the 0.1 % detection limit.Subsequently, the ampoule rests for five days.IR spectra recorded after this period show no increase in the CO 2 concentration.
We have also monitored the concentration of the dominant nongreenhouse gases, O 2 and N 2 , during this experiment using Raman spectroscopy and found that the intensity of the Raman lines are constant within the experimental error (Fig. S1).
Fig. 2c and d refer to the ampoule, which initially is loaded with 0.5 % of CH 4 in a dry atmosphere.Fig. 2c shows the IR absorption spectrum associated with the asymmetric stretch mode of CH 4 around 3000 cm − 1 at selected days, while Fig. 2d shows the absorbance of CH 4 at 3016 cm − 1 as a function of time.The CH 4 content of the ampoule remains nearly constant at its initial value during the first five days of tumbling after which it decreases with an increasing rate to less than 0.1 % of the initial value after 25 days.The IR spectra recorded on the ampoule with CH 4 reveal the production of CO and CO 2 as indicated in Fig. 2c by the absorption associated with the CO stretch mode around 2150 cm − 1 and the absorption pertaining to the CO 2 asymmetric stretch mode around 2350 cm − 1 .The removal of CH 4 is linked to a concurrent production of CO and CO 2 .Thus, the CO concentration increases to a maximum while the CH 4 is still close to its initial value and then drops to a level below the detection limit concurrent with the disappearance of CH 4 .The CO 2 concentration in turn increases for as long as CH 4 is present and then drops below the detection limit during the next couple of days after the disappearance CH 4 .After 36 days of tumbling, the ampoule rests for five days.IR spectra recorded after this period show no absorption from CO 2 , CH 4 or CO.We expect the reactions leading from CH 4 to CO and CO 2 involves triboelectric ignited combustion of CH 4 and note that triboelectric charging in related experiments [1] leads to energies capable of dissociating O 2 .
Fig. 2e and 2f refer to the ampoule, which initially is loaded with 0.5 % of N 2 O in a dry atmosphere.Fig. 2e shows the IR absorption spectrum associated with the asymmetric stretch mode of N 2 O around 2225 cm − 1 at selected days, while Fig. 2f shows the time dependence of the maximum absorbance of N 2 O at 2235 cm − 1 .After 15 days of tumbling, there is no absorption from N 2 O and the ampoule rests for five days.IR spectra recorded after these five days show no reappearance of the N 2 O absorption.

Summary and perspective
Our experiments have demonstrated that triboelectric charging of quartz can remove three major greenhouse gases from an air-like atmosphere quantitatively.Quartz serves as abrasive, which creates the triboelectric effect.The choice of quartz is due to its chemical simplicity and high purity, which facilitate theoretical investigations.We are in the process of exploring the fate of the gases removed from the atmosphere and expect CO 2 to be chemically bound in the solid phase, as it has not returned to the atmosphere -several days after its removal.
The concentrations of the greenhouse gases in the experiments are much higher than the current concentrations in ambient air.We have used dry air for convenience.Initial experiments in a moist atmosphere indicated that triboelectric charging removed water in a couple of days (See Fig. S2).Considering there are efficient techniques for water removal from air, it seemed prudent to use dry air.
The capture of greenhouse gases is not specific to the use of quartz of high purity.In a separate experiment, we have found that replacing the quartz grains with pristine sand captures CO 2 in much the same way as quartz (Fig. S3), suggesting that readily available materials may be used instead.
We consider the outcome of our experiments as a proof of principle of a hitherto unexplored method to sequester greenhouse gases from the atmosphere.In the presented study, we have activated quartz grains using a rotation speed of 30 RPM and found that the greenhouse gases are removed quantitatively in a couple of weeks.We expect a modified activation process can dramatically shorten this time.We encourage a feasibility study to address if capture of greenhouse gases from ambient air by triboelectric charging can effectively be up-scaled and ultimately can restore favourable atmospheric greenhouse gas concentrations for a sustainable future [12].

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.Apparatus used in the experiments.A: Quartz ampoule containing quartz grains (SiO 2 ) and the gas mixture.B: Quartz ampoule with two CaF 2 windows.B fits into the IR spectrometer.Two filters (pore size: 40-100 µ) protect B and the vacuum system from the quartz grains, but allow the exchange of gas between the two ampoules.

Fig. 2 .
Fig. 2. IR spectra and absorption dynamics of a gas mixture of 20.8 % O 2 , 78.7 % N 2 and 0.5 % of one of the greenhouse gases CO 2 , CH 4 and N 2 O in the tribo-electric mediated activation with quartz grains.a: IR spectra of CO 2 in the gas mixture recorded during 13 days of activation.b: The absorbance from CO 2 at 2360 cm − 1 as a function of activation time.c: IR spectra of CH 4 in the gas mixture recorded during 32 days of activation.CH 4 produces CO and CO 2 as indicated by the IR absorption from CO at 2150 cm − 1 , and the IR absorption from CO 2 around 2350 cm − 1 .d: The absorbance from CH 4 at 3016 cm − 1 , from CO at 2175 cm − 1 and from CO 2 at 2360 cm − 1 as a function of activation time.(Note the concentration of CO is increased by a factor of 100).e: IR spectra of N 2 O in the gas mixture recorded during 23 days of activation.f: The absorbance from N 2 O at 2230 cm − 1 as a function of activation time.