Oxygenates production in a microfluidic dielectric barrier discharge device sustained in Ar/CH₄/O₂

Reforming of methane (CH₄) is a chemical process used widely to produce syngas, a mixture of carbon monoxide (CO) and hydrogen (H₂). CH₄ reforming can be performed through two processes—steam reforming in the presence of H₂O and dry reforming in the presence of CO₂.¹ While steam reforming of CH₄ is commonly used in industrial processes, it is an endothermic process requiring elevated pressures and temperatures.¹ The syngas produced in steam reforming of CH₄ is often further converted to other value-added compounds. One of the most common value-added compounds produced is methanol (CH₃OH).² From CH₃OH, formaldehyde (CH₂O) is produced by oxidation.

Atmospheric pressure plasmas, when sustained in CH₄, convert CH₄ to other compounds.^3–7 Previous research has examined dry reforming of CH₄ in plasmas, where CH₄ and CO₂ are converted to syngas and other compounds.^8–15 Plasmas have the potential to be more energy efficient than other CH₄ reforming processes as plasmas can be sustained near ambient temperature and at atmospheric pressure. Hot (several eV) electrons dissociate and ionize the neutral feedstock gases, leading to complex plasma chemistry while retaining low gas temperatures. CH₄ conversion proceeds through direct electron-impact dissociation of CH₄ or reactions of other plasma-produced radicals and excited states with CH₄, while the low gas temperature enables selectivity.

Conversion of CH₄ in the presence of a noble gas has been previously investigated.^16–19 Janeco et al. examined electron kinetics in He/CH₄/CO₂ mixtures and showed that addition of He shifted the electron energy distribution to higher energies.²⁰ The higher energy electrons led to increased conversion of CH₄ and CO₂. Rahmani and Nikravech showed that addition of Ar increased the electron density and electron temperature in CH₄/CO₂ mixtures.²¹ Jo et al. showed that the identity of the noble gas (He, Ne, Ar, Kr, or Xe) affected the electron temperature and density.²² Ozkan et al. examined the difference in CH₄ and CO₂ conversion in the presence of either Ar or He.¹² They showed that the conversion of CH₄ was higher for He than for Ar, while the conversion of CO₂ was higher for Ar than He.

Studies have also focused on the production of oxygenates, including CH₃OH and CH₂O.²³ De Bie et al. modeled the production of oxygenates in CH₄/O₂ mixtures and CH₄/CO₂ mixtures in a nanosecond pulsed dielectric barrier discharge (DBD).²⁴ They predicted the formation of many oxygenated species, including CH₃OH, CH₃OOH, CH₂O, C₂H₅OOH, C₂H₅OH, and CH₃CHO. Biswas et al. examined the production of oxygenates from CO₂ and C₂H_6.²⁵ The dominant product was CO, and other hydrocarbons and oxygenated species were also formed. Kolb et al. investigated the production of oxygenates in a DBD with 97% He.²⁶ The addition of O₂ to He/CH₄/CO₂ was shown to increase CH₃OH and CH₂O yields. Li et al. examined oxygenate production from CH₄/CO₂/O₂ discharges.²⁷ The increased amounts of O, OH, and HO₂ were responsible for enhancing the oxygenate production with O₂ addition. Oxygenate production in plasma-catalytic systems has also been studied.^28–30 

As we advance our knowledge of oxygenate production from CH₄ plasmas, the manner of implementing those processes also becomes important. That implementation should consider the method of excitation (e.g., DBD, microwave, glow, pulsing) and throughput. For example, pulsed ns DBDs have some advantages over microwave excitation when maintaining low gas temperatures, which is an important consideration in the conversion process. Micro-plasmas have the advantage of rapidly rising voltage pulses, well-controlled power deposition, and temperature control. DBD plasmas sustained in microchannels (as found in microfluidic devices) enable such control as well as management of products through, for example, solvation into solution. Large throughput is achieved by parallel processing.

In this study, we used the 0-dimensional (0D) model GlobalKin to assess the production of CH₃OH and CH₂O in a pulsed nanosecond DBD sustained in Ar/CH₄/O₂ flowing through a microfluidic channel, ideal for producing uniform flows with high power deposition at atmospheric pressure. The goal is to examine mechanisms for CH₃OH and CH₂O production while minimizing the production of the greenhouse gas CO₂. The dominant production and consumption mechanisms of the plasma-produced radicals and long-lived species are identified. Common oxygenates formed include CH₃OH, CH₂O, and CO. The plasma-produced species that do not contain oxygen include H₂, C₂H₄, and C₂H₆. After examining the base case, the consequences of the gas mixture are examined. When increasing the Ar content while keeping CH₄/O₂= 1/1, CH₃OH and CH₂O densities decrease while the density of CO₂ increases to 90% Ar. A mixture of 50% Ar and 35%–40% CH₄ (10%–15% O₂) maximizes CH₃OH and CH₂O production while also minimizing the amount of CO₂ formed.

The model is described in Sec. II. The results of the base case are discussed in Sec. III, including delineation of the production and consumption mechanisms of abundant radicals and long-lived species. Consequences of the gas mixture are discussed in Sec. IV, and conversion and energy expenditures are discussed in Sec. V. Concluding remarks are presented in Sec. VI.

GlobalKin is a 0D plasma chemistry model that integrates continuity equations for neutral and charged species.³¹ Sources and losses for these species include chemical reactions, flow, and diffusion to bounding surfaces of the plasma. Electron temperature T_e and gas temperature T_gas are calculated using their respective energy equations. A stationary solution of Boltzmann's equation produced using a two-term spherical harmonic expansion is used to compute the electron energy distribution at different values of E/N (electric field/gas number density). The resulting values of electron temperature T_e are then the basis of creating a look-up table of electron-impact rate coefficients and transport coefficients as a function of T_e. The look-up table is updated at the beginning of each discharge pulse by recomputing the electron energy distributions to reflect the changes in the gas mole fractions.

The gas mixture examined is Ar/CH₄/O₂, and the base case conditions are Ar/CH₄/O₂ = 50/25/25. The reaction mechanism includes 122 species, listed in Table I, and 3265 reactions between those species. The Ar/O₂/H₂O mechanism is based on Van Gaens and Bogaerts,³² and the Ar/CH₄/H₂O mechanism is based on our previous work.³³ Additional reactions between hydrocarbon species and oxygen species were added to the mechanism and are listed in the supplementary material. Important reactions in the production and consumption of radicals and long-lived species are listed in Tables II and III, respectively. These reactions will be discussed in the following section.

The geometry in this work replicates that of our previous study of methyl radical production in a microfluidic channel.³³ The microfluidic channel serves as a dielectric barrier discharge (DBD). The channel dimensions are 500 × 500 μm². The power is pulsed on ns timescales, ramping up over 15 ns, staying constant for 30–45 ns, and decreasing to 0 W at 60 ns. The peak power deposition is 228 kW cm⁻³, and the energy deposited into the plasma is 10 mJ cm⁻³ per pulse. The pulse repetition rate is 10 kHz (0.1 ms period), and the evolution over 20 pulses (2 ms) was tracked. The flow rate is 0.1 SCCM to approximate plug flow, and the outflow was adjusted to maintain 1 atm at a timescale of 0.1 ms.

Since GlobalKin is a 0D model, it does not account for spatial dynamics such as ionization waves in the bulk plasmas or along surfaces. The computational expediency of global models is that they enable detailed analysis of the plasma chemistry and can be used to compute plasma chemical dynamics over multiple pulses and long time scales. We acknowledge that chemical conversion processes that depend on the specifics of these spatial dynamics will not be captured in a global model. However, we expect that the trends predicted by global models will hold for higher dimensional models and experiments.

In this section, the radical and long-lived species production and consumption for the base case of Ar/CH₄/O₂ = 50/25/25 are discussed. The plasma properties are discussed in Sec. III A. The production and consumption mechanisms of radicals and long-lived species are discussed in Secs. III B and III C, respectively.

The plasma properties are shown in Fig. 1 over the last of 20 pulses. The electrons are initially seeded with a density of 10⁹ cm⁻³ at the beginning of each discharge pulse. As power is applied with this initially small electron density, T_e increases to 4.9 eV, above the quasi-steady-state value, to sustain an electron avalanche. With the increase in electron density, T_e decreases. During the constant power portion of the pulse, the electron density and T_e reach their quasi-steady state values of 1.8 × 10¹³ cm⁻³ and 3.1 eV. As the power decreases at the end of the pulse, T_e rapidly decreases to lower the power dissipation per electron. Electrons rapidly recombine with positive ions, dominantly dissociative recombination, or attach to form negative ions including OH⁻, O₂⁻, and O⁻. The gas temperature increases from its initial value of 300 K up to 403 K at 2 ms.

The dominant radicals formed in the Ar/CH₄/O₂ system are shown in Fig. 2, and the production and consumption reactions of each radical are detailed in Table II for those processes that contribute more than 5% of the total production or consumption of the radical. The most important reactions will be discussed in this section.

HO₂ remains at an elevated density between the pulses, reaching 6.7 × 10¹⁴ cm⁻³ before the final pulse. HO₂ is consumed in reactions with O₂^(1Δ) (56%) and O (32%). Both O₂^(1Δ) and O are produced during the pulse and consumed in the afterglow, decreasing the amount of HO₂ consumed in the afterglow. OH relies on HO₂ to be produced, both through reactions with O₂^(1Δ) (45%) and formation of O₂ from HO₂ (26%). OH is consumed in H abstraction from CH₄ (25%) and CH₂O (20%).

Both of these reactions consume 33% of CH. The most abundant C₂H_x radical is C₂H₅, requiring other radicals to be formed before it can be produced. C₂H₅ is formed from CH₂ associating with CH₄ (reaction 5, 98%). C₂H₅ is primarily consumed in forming the oxygenated radical C₂H₅O₂ by association with O₂ (95%).

The oxygenated radicals are shown in Fig. 2(c). Contrary to the ROS and hydrocarbon radicals, all oxygenated radicals rely on other radicals for their formation. Therefore, their densities all reach their maxima after the discharge pulse, following the production of their precursors during the discharge pulse. The most abundant oxygenated radical is CH₃O₂, formed by the association of CH₃ and O₂ (99%). CH₃O₂ is primarily consumed in reactions with O (60%) and OH (34%), forming CH₃O and CH₃OH. Since CH₃O₂ is consumed in reactions with other radicals, CH₃O₂ persists between pulses with a density of 4.6 × 10¹⁴ cm⁻³ at the end of 20 pulses.

The long-lived species that are formed by repetitive pulsing are shown in Fig. 3, and the dominant production and consumption reactions of each long-lived species are listed in Table III. Consumption reactions appear in Table III if they represent more than 5% of the total integrated production rate of that long-lived species.

Reaction (18) contributes 58% to C₃H₆ consumption, while reaction (19) contributes 41%.

The long-lived oxygenated species formed in the plasma are shown in Fig. 3(c). The most abundant long-lived oxygenates are CO, CH₃OH, and CH₂O. CO is the most abundant after 20 pulses with a final density of 3.8 × 10¹⁶ cm⁻³. CO is primarily formed from H abstraction by O₂ from CHO (93%). As O₂ is a feedstock gas, this reaction is limited by the availability of CHO. The CO/H₂ ratio, relevant for syngas applications, reaches 4.3 at the end of 20 pulses. CH₃OH has a density of 2.1 × 10¹⁶ cm⁻³ after 20 pulses. CH₃OH is produced by H abstraction from CH₃O by CH₃O (53%) and O removal from CH₃O₂ (45%). The rates of these reactions are shown over the last pulse in Fig. 4(a). The peak rates occur at different times, attributed to when the densities of the reactants peak. The formation of CH₃O₂ occurs by CH₃ association with O₂, while CH₃O requires CH₃O₂ to form. Therefore, the formation of CH₃OH from CH₃O occurs at a later time than the formation from CH₃O₂. CH₃OH is consumed in reactions with OH, shown in Fig. 4(b), though these reactions occur at 25% of the rate of CH₃OH formation.

Reaction (20) contributes 13% to CH₂O formation and reaction (21) contributes 11%. As shown in Fig. 4(c), reaction (21) occurs during the discharge pulse, as both O and CH₃ are produced from the electron-impact dissociation of feedstock gases. The other three reactions that produce CH₂O rely on CH₃O, which reaches its maximum density 2 μs after the last pulse begins. As a result, the maximum rates of those reactions occur near 2 μs. While CH₂O is formed at an integrated rate of 5.6 × 10¹⁶ cm⁻³ over the 20 pulses, 4.0 × 10¹⁶ cm⁻³ of CH₂O is consumed over these 20 pulses. The main consumption mechanisms of CH₂O are CHO formation through H abstraction by OH (79%) and by O (18%), shown in Fig. 4(d). H abstraction by O reaches its maximum rate as the power begins to decrease, corresponding to the maximum in O density. OH, however, is formed later from the reactions of HO₂, and therefore, the rate peaks later at 0.7 μs.

A goal of this study is to determine the conditions that maximize the density of CH₃OH (and, to a lesser extent, CH₂O) while minimizing the amount of CO₂ formed. The gas mixture in which the plasma is sustained affects the fundamental plasma properties (electron density, T_e), as well as the long-lived species formed and their relative concentrations. In Sec. IV A, the fraction of Ar in the mixture is varied while holding the ratio CH₄/O₂ constant at 1/1. In Sec. IV B, the ratio of CH₄/O₂ is changed while keeping a constant 50% Ar mole fraction.

Previous studies have shown that increasing the noble gas content in CH₄/CO₂ plasmas increases the conversion of CH₄ and CO₂.^18,19,21 In this section, the effect of changing the Ar percentage while keeping the CH₄/O₂ ratio constant at 1/1 as in the base case is examined.

As the Ar percentage increases from 0% to 95% Ar, the electron density when the power begins to decrease over the last pulse (45 ns) increases from 9.4 × 10¹² to 7.9 × 10¹³ cm⁻³. With the increase in Ar, T_e increases from 2.98 to 3.27 eV. During the quasi-steady state, electron temperature is determined by the balance of ionization sources and charged particle losses. With the power profile being held constant, electron density is then determined by the power dissipation per electron at the self-sustaining electron temperature. The first inelastic electron-impact threshold with Ar is electronic excitation at an energy of 11.55 eV, while the ionization threshold is 16 eV. With Ar dilution, the mole-fraction weighted power dissipation per electron decreases as the rate of power dissipation by electrons with CH₄ or O₂ is larger due to their rotational–vibrational modes and lower threshold energies for electronic excitation and ionization. This condition is somewhat mediated at high power deposition when multistep ionization of argon begins to dominate, and at intermediate mole fractions of the molecular gas when Penning processes can be important.

The frequency of radical formation by electron-impact on a molecular species is $f r= n e ∑ i k i ( T e ),$ where the sum is over rate coefficients for radical formation. The total rate of radical formation by, for example, electron-impact on CH₄ is $f t= n e[ C H 4] ∑ i k i ( T e )$⁠. With the increases in electron temperature and density with Ar dilution, $f r$ monotonically increases while $[ C H 4]$ monotonically decreases. These trends result in $f t$ being maximum at an intermediate Ar dilution.

The maximum oxygenated species densities (CH₃OH, CH₂O, CO, and CO₂) over the last pulse are shown in Fig. 5(a) as a function of Ar percentage. CH₃OH and CH₂O both decrease as the Ar percentage increases. For CH₃OH, the decrease is slow below 40%, while above 40% Ar, the decrease becomes more rapid. While the production of CH₃OH due to CH₃O reacting with CH₃O and CH₃O₂ reacting with OH increases up to a 40% Ar mole fraction, the consumption of CH₃OH by OH also increases. The increase in the consumption of CH₃OH outweighs the increase in CH₃OH production, thereby decreasing its density.

Above a mole fraction of 40% Ar, the consumption of CH₃OH by OH continues to increase while the production of CH₃OH decreases, leading to a more rapid decrease in the CH₃OH density. Similarly, CH₂O production relies on CH₃O and CH₃, and increases through 40% Ar. However, the consumption of CH₂O due to OH also increases, thereby outweighing the increase in production to decrease the CH₂O density. Above 40% Ar, the density of CH₂O is not well correlated to the density of CH₃OH. While the production of CH₂O due to CH₃O reactions decreases above 40% Ar, its production due to CH₃ reacting with O increases. Therefore, the decrease in CH₂O production is less pronounced than in CH₃OH.

The production of both CH₃OH and CH₂O relies on the CH₃O radical. The production of CH₃O by CH₃O₂ reacting with O increases up to 40% Ar. CH₃O₂ production by association of CH₃ and O₂ also increases up to 40% Ar, following the production of CH₃ by H abstraction from CH₄. CH₃O₂ and CH₃ production by H abstraction both increase with increasing T_gas, with T_gas reaching a maximum of 414 K in the afterglow at 40% and 50% Ar.

While the densities of CH₃OH and CH₂O decrease with increasing Ar, the densities of CO and CO₂ increase up to Ar mole fractions of 70%–90%, as shown in Fig. 5(a). CO reaches a maximum density of 4.1 × 10¹⁶ cm⁻³ at 70% Ar, while CO₂ reaches 4.4 × 10¹⁵ cm⁻³ at 90% Ar. CO relies on its production by H abstraction from CHO by O₂, which increases to 70% Ar. CHO is formed by H abstraction from CH₂O by OH. Since CH₂O decreases as Ar increases, the increase in CHO production occurs because of the increase in T_gas as well as the increase in OH density. The density of OH increases up to 90% Ar as the production from HO₂ reacting with O increases. This trend prevails in spite of a decrease in OH production from HO₂ reacting with O₂(¹Δ). CO₂ production from CH₃CO reacting with O and OH reacting with CO increase to 80% Ar and 90% Ar, respectively. These reactions lead to a maximum CO₂ density at 90% Ar.

The density of C₂H₄ exhibits a small decrease from 0% Ar to 80% Ar before increasing to 97% Ar. The production of C₂H₄ from CH reacting with CH₄ decreases with the increasing Ar percentage, but the production from C₂H₅⁺ reacting with H₂O and from H abstraction from C₂H₅ increases. That increase offsets the decrease from CH reacting with CH₄. For mole fractions larger than 80% Ar, the production from C₂H₅⁺ reacting with H₂O and H abstraction from C₂H₅ further increases, driving an increase in C₂H₄ density. The density of C₂H₆ increases with increasing Ar mole fraction to 97%, though the increase becomes more rapid after 50% Ar. C₂H₆ is produced dominantly by the mutual association of CH₃, and the maximum CH₃ density over the last pulse increases nearly linearly with the Ar percentage from 1.4 × 10¹⁵ cm⁻³ at 0% Ar to 1.8 × 10¹³ cm⁻³ at 95% Ar. After 97%–98% Ar, H₂, C₂H₄, and C₂H₆ densities decrease as CH₄ makes up less than 3% of the discharge—the increase in plasma density and electron temperature with increasing Ar mole fraction cannot compensate for the low mole fraction of CH₄.

The fraction of the inlet O₂ remaining is between 70% and 80% for the change in Ar content shown in Fig. 5. On that basis alone, one might expect that the production of, for example, CH₃OH and CH₂O would decrease based on there being fewer O atoms available due to the depletion of the O₂ feedstock. However, the degree of dissociation of O₂ does not straightforwardly increase or decrease the production of CH₃OH and CH₂O. There should clearly be sufficient O atom production to enable the stoichiometry of the products. However, one of the eventual products of O₂ dissociation is OH which is involved in both the production and consumption reactions of CH₂O and CH₃OH. At some intermediate degree of O₂ dissociation, sufficient O atoms are produced while not also feeding the destruction pathways.

These results indicate that for the ratio of CH₄/O₂ = 1/1, operating with a small Ar mole fraction will maximize the CH₃OH and CH₂O densities while producing little CO₂. In fact, at 95% Ar, the density of CO₂ is larger than the density of either CH₃OH or CH₂O. These trends are, however, sensitive to the ratio of CH₄/O₂.

Ultimately, the production of oxygenates depends on the reactions between carbon-containing and oxygen-containing species, while higher order hydrocarbon production depends largely on the reactions between non-oxygen containing species. In this regard, the branching to different oxygenated and hydrocarbon products likely depends on the ratio of CH₄ to O₂ in the mixture. In this section, these trends are discussed while the Ar mole fraction is held constant at 50%, and the CH₄ and O₂ percentages are varied. At low CH₄ percentages (high O₂ percentages), oxygenate formation is expected to dominate over hydrocarbon formation as oxygen is abundant. At high CH₄ percentages (low O₂ percentages), hydrocarbon formation is expected to dominate over oxygenate formation as oxygen is less abundant.

The electron density as the last power pulse begins to ramp down (45 ns) increases from 1.3 × 10¹³ cm⁻³ at 5% CH₄ to 3.4 × 10¹³ cm⁻³ at 45% CH₄. This increase is due, in part, to decreased negative ion formation, with the maximum total negative ion density occurring at 20% CH₄. The negative ions are dominantly the negative water cluster ions O₂⁻(H₂O), O₂⁻(H₂O)₂, and O⁻(H₂O) that follow the formation of O₂⁻ and O⁻. While the maximum densities of O₂⁻ and O⁻ occur at 5% CH₄ (45% O₂), the maximum of the cluster ion densities occurs at 20% CH₄ as H₂O is formed in the discharge instead of being present as one of the feedstock gases. While the electron density increases, T_e decreases from 3.21 eV at 5% CH₄ to 2.85 eV at 45% CH₄.

The dominant long-lived species are shown in Fig. 6 with the oxygenates shown in Fig. 6(a) as a function of CH₄ and O₂ percentage. CH₃OH and CH₂O reach their maximum densities at 35% and 40% CH₄, respectively. This is surprising because the O₂ content in the discharge is low (10%–15%). The production of both CH₃OH and CH₂O by CH₃O reacting with CH₃O increases to 35% CH₄ which explains the maximum CH₃OH density at 35% CH₄. However, CH₂O reaches a maximum at 40% CH₄ due to its additional production from CH₃ reacting with O increasing to 40% CH₄. Both CH₃OH and CH₂O production rely on CH₃O. The maximum CH₃O density occurs at 30% CH₄, based largely on the production from CH₃O₂ which reaches a maximum at 25% CH₄. The CH₃O₂ density is maximum at 30%–35% CH₄. While the density of CH₃ reaches a maximum at 40% CH₄, the association reaction to form CH₃O₂ depends on T_gas, which reaches its maximum at 25% CH₄ (414 K).

The fraction of inlet O₂ remaining is between 70% and 80% for the changing CH₄/O₂ content shown in Fig. 6 while the production of CH₃OH and CH₂O is maximum for an O₂ inlet flow of 10–15%. The small amount of O₂ at this maximum in production may be counter-intuitive. The production of CH₃OH and CH₂O generally requires CH₃O as a precursor. CH₃O is produced from CH₃O₂, which is generated by the reactions between CH₃ + O₂. The CH₃ content is critical for CH₃OH and CH₂O production and can be the rate-limiting step, provided, there is a critical amount of O₂ remaining.

The densities of CO and CO₂ follow a different trend than those of CH₃OH and CH₂O. These species are formed preferentially at low CH₄ percentages, with densities reaching a maximum at 20% CH₄ (30% O₂). The maximum in CO density follows the maximum of its production from H abstraction from CHO by O₂. CHO is produced by H abstraction from CH₂O by either OH or O. The rates of these reactions increase to 20%–25% CH₄ due to the increase in both CH₂O density and T_gas. The production mechanisms of CO₂ (OH reacting with CO and CH₃CO reacting with O) reach their maximum at less than 25% CH₄, leading to a maximum CO₂ density at 20% CH₄.

The densities of long-lived hydrocarbons and H₂ are shown in Fig. 6(b). As expected, these species increase with increasing CH₄, as they do not require oxygen to form. In fact, their production is enhanced since the hydrocarbon radicals which would otherwise be consumed by forming oxygenates are more readily available for hydrocarbon formation as the fraction of O₂ decreases. H₂ and C₂H₄ both reach their maximum density at 40% CH₄. H₂ is produced primarily from the electron-impact dissociation of CH₄. While T_e decreases as the CH₄ percentage increases, the electron and CH₄ densities increase. These increases lead to an increase in the rate of electron-impact dissociation of CH₄ through 40% CH₄. The production of C₂H₄ by C₂H₅⁺ reacting with H₂O decreases above 15% CH₄, but the production by CH reacting with CH₄ increases. This increase occurs as the rate of electron-impact dissociation of CH₄ producing CH increases to 40% CH₄. The density of C₂H₆ increases through 45% CH₄ which is primarily formed through mutual reactions of CH₃. The integrated rate of mutual association of CH₃ increases from 8.4 × 10¹² cm⁻³ at 5% CH₄ to 1.1 × 10¹⁵ cm⁻³ at 45% CH₄.

Increasing the CH₄/O₂ ratio with a constant mole fraction of 50% Ar increases the production of CH₃OH and CH₂O and simultaneously decreases CO₂. The results presented in this section show that at 50% Ar, 35%–40% CH₄ (10%–15% O₂) optimize CH₃OH and CH₂O production. At these conditions, the production of CO₂ decreases relative to its concentration at CH₄/O₂ = 1/1.

Although not a direct focus of this investigation, plasma conditions can be optimized to generate specific end products. For example, CH₃CHO (acetaldehyde) is widely used as an intermediate to generate acetic acid, peracetic acid, pyridine bases, and other chemicals. The production of CH₃CHO from ethylene³⁵ by the Wacker process is one of the most important processes in the chemical industry,³⁶ while acetaldehyde is itself a hazardous liquid. Less capital intensive and safer, point-of-use production of CH₃CHO would be advantageous for on-demand organic syntheses. For our setup, plasma generation of CH₃CHO is maximum for conditions similar to those of the other C₂ compounds. For 50% argon mixtures, CH₃CHO production is maximum with 40% CH₄ (10% O₂) resulting in a density of 8 × 10¹⁴ cm⁻³ after 20 pulses. For a ratio of CH₄/O₂= 1/1, CH₃CHO production is maximum in the absence of argon, generating a density of 6 × 10¹⁴ cm⁻³ after 20 pulses.

The intent of this investigation was to address reaction mechanisms and not necessarily optimize conversion and energy efficiency for CH₄ oxygenation. Although global modeling is well suited to discussions of mechanisms, often, energy efficiencies are dependent on geometrical considerations that are difficult to include in global models—such as division in power deposition between sheaths and bulk, propagation of streamers or surface ionization waves, and gas flow (replenishment and pumping of gases).

That said, some assessments can be made. The conversion rates of CH₄ and O₂ were assessed for a sealed-off system to eliminate the uncertainties in accounting for gas flow and pressure stabilization while varying the CH₄/O₂ ratio with 50% Ar. The maximum CH₄ conversion was 2.2% at 5% CH₄, decreasing to 0.7% for 45% CH₄. The maximum O₂ conversion was 6.4% at 5% O₂, decreasing to 3.0% for 45% O₂. Xu and Tu measured CH₄ conversion in an AC DBD with pure CH₄ and reported a conversion percentage of 25.2% at the longest residence time (16 s).³⁷ Chen et al. reported a maximum conversion of 31.9% in a pure CH₄ nanosecond pulsed DBD.³⁸ Jo et al. reported a CH₄ conversion percentage of 7%–11% for Ar/CH₄/O₂ plasmas in an AC DBD for varying O₂ and CH₄ concentrations.²² As these rates of conversion are dependent on residence time and energy deposition-per-molecule, direct comparisons are better made normalized by energy deposition.

The energy required for conversion (expressed as eV/molecule) was evaluated in a sealed-off system while varying the CH₄/O₂ ratios at 50% Ar. For these conditions, the minimum energy cost for the conversion of CH₄ was 13.8 eV/molecule for Ar/CH₄/O₂ = 50/30/20. Xu and Tu measured a maximum energy efficiency of 0.26 mmol/kJ (40 eV/molecule) for CH₄ conversion in an AC DBD operating in pure CH₄.³⁷ While the calculated energy cost in this work was lower than that reported in Xu and Tu, the plasma in this work was formed in Ar/CH₄/O₂ instead of pure CH₄. The calculated energy cost for the production of CH₃OH was minimum at Ar/CH₄/O₂ = 50/35/15 at 46.6 eV/molecule. Wang et al. reported a maximum energy efficiency of CH₃OH production of 0.83 mol/kWh (45 eV/molecule) in an AC DBD operating in CH₄/CO₂,²³ comparable to the calculated energy cost in this work.

The formation of value-added chemicals in atmospheric pressure plasmas can provide an alternative to energy-intensive steam and dry reforming of CH₄. The production of oxygenates, including CH₃OH and CH₂O, using plasmas sustained in Ar/CH₄/O₂ was investigated using a 0D plasma chemistry model GlobalKin. The geometry was focused on narrow channels as might be used in microfluidic devices where a scale is achieved by massively parallel sets of microchannels. The goal of this study was to investigate reaction mechanisms resulting in the production of CH₃OH and CH₂O while minimizing the production of the greenhouse gas CO₂.

First, the plasma chemistry in Ar/CH₄/O₂ = 50/25/25 was examined. Of the many radical species that are formed in the plasma, CH₃ and O are formed directly by the electron-impact dissociation of CH₄ and O₂. Others, particularly the oxygenated radicals CH₃O₂, CH₃O, and CHO, depend on those initial radicals for their formation. Further reaction of these species branch to several products. The oxygenates, including CO, CH₃OH, and CH₂O, comprise most of the carbon-containing products while the greenhouse gas CO₂ is also formed. Pure hydrocarbon products (C₂H₄ and C₂H₆) are also formed, albeit at an order of magnitude of lower density. The formation of CH₃OH relies on CH₃O and CH₃O₂, while the formation of CH₂O relies on CH₃O and CH₃. CO is formed almost exclusively by H abstraction from CHO.

The mole fractions of the feed gas affect the selectivity of the process and, in particular, the production of CH₃OH, CH₂O, and CO₂. First, the consequences of Ar percentage were examined while keeping the ratio CH₄/O₂ = 1/1. CH₃OH and CH₂O have maximum densities at 0% Ar, while the density of CO is the highest at 70% Ar and CO₂ is the highest at 90% Ar. If the goal is, for example, CO production for syngas, moderate Ar additions are beneficial. However, if the goal is CH₃OH or CH₂O, Ar dilution is not beneficial. When varying the CH₄/O₂ ratio at a constant 50% Ar, the densities of CH₃OH and CH₂O are maximum at 35%–40% CH₄ (10%–15% O₂) while the densities of CO and CO₂ are the maximum at 20% CH₄. Therefore, a CH₄ rich discharge of 35%–40% favors CH₃OH and CH₂O production while minimizing CO₂ formation.

Conventional microfluidic labs-on-a-chip are intentionally designed to have long residence times (many to tens of seconds or more), a goal achieved with channel lengths of up to several meters. Although optimum for microfluidic processing, this is not necessarily the geometry that would be implemented for plasma conversion using microchannels. With these long channel lengths and residence times, the desired products would also be subject to lengthy electron-impact dissociation and ionization. An advantage of plasma conversion using microchannels is that the precise time of plasma exposure can be set by the length of the channel and gas flow. Although not the focus of this investigation, we anticipate that optimized plasma conversion using microchannels would consist of large arrays of parallel channels, with residence time in any given channel being tens to hundreds of ms. These parameters would be chosen to optimize the conversion of the feedstock gas while not detrimentally undergoing plasma dissociation of the desired products.

See the supplementary material for additions and revisions to the prior published reaction mechanism33 for plasmas sustained in Ar/CH4 mixtures to account for the addition of oxygen.

This material was based upon work supported by the National Science Foundation under Award Nos. ECO-CBET-2032604 and ECO-CBET-2032664. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. This work was also supported by the Department of Energy Office of Fusion Energy Sciences (No. DE-SC0020232) and the Army Research Office MURI program (No. W911NF-20-1-0105).

The authors have no conflicts of interest.

Mackenzie Meyer: Conceptualization (lead); Data curation (lead); Formal analysis (lead); Investigation (lead); Methodology (equal); Writing – original draft (lead); Writing – review & editing (equal). Ryan Hartman: Funding acquisition (equal); Project administration (equal); Supervision (equal); Writing – review & editing (equal). Mark J. Kushner: Formal analysis (supporting); Funding acquisition (equal); Investigation (supporting); Methodology (equal); Project administration (equal); Resources (equal); Supervision (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.