Novel investigation of perovskite membrane based electrochemical nitric oxide control phenomenon

The combustion of hydrocarbon fuels within the automotive industry results in harmful and reactive incomplete combustion byproducts. Specifically, nitric oxide emissions (NO) lead to increased smog, acid rain, climate change, and respiratory inflammation within the population [Nitrogen Dioxide | American Lung Association]. Current methods for treating combustion exhaust include the catalytic converter in conjunction with nitrogen oxide traps. However, there is no active, continuous reduction method that does not require restrictions on the combustion environment (Hirata in Catal Surv Asia 18:128–133, 2014). Here, a small voltage potential oscillation across a newly designed electro-chemical catalytic membrane significantly reduces NO emissions. A ceramic membrane consisting of two dissimilar metal electrodes, sandwiching a dielectric layer, is able to achieve an NO reduction in excess of 2X that of a platinum group metal (PGM) three way catalytic converter. An analysis of the exhaust effluent from the membranes indicates N2O as a precursor to N2 and O2 formation, without the introduction of ammonia (NH3), during the reaction of NO indicating a divergence from current literature. Our results demonstrate how an oscillatory electric potential on a catalytic surface may alter anticipated reaction chemistry and interaction between the catalytic surface and fluid flow.

normal operating conditions. The model exhausts were created by combining O 2 , CH 4 , CO, CO 2 , N 2 , and NO via benchtop flow meters.
The results shown below in Table 1 represent the time averaged results of each specimen. The specimen was introduced into the furnace 60 min prior to the collection of any effluent, for analysis to allow each sample to obtain steady state operation. The effluent of each sample specimen was then collected, analyzed, and recorded by the E-Instruments emission analyzer and gas chromatography-mass spectroscopy (GC-MS). The results were averaged over a constant 3 h test for each specimen.
The electrochemical-catalytic membrane is able to significantly outperform the PGM catalytic converter in the reduction of NO and CO, and closely matches the PGM catalytic converter in the reduction of hydrocarbons. Additionally, the electrochemical-catalytic membrane is able to maintain function in the presence of high concentrations of oxygen. The ability of the electrochemical-catalytic converter to maintain reactivity of NO at varying levels of NO has been previously reported 17 .
The novel electrochemical-catalytic membrane is able to consistently outperform traditional catalytic membranes for continuous NO reduction across multiple operation conditions. However, during testing, a minute oscillatory electrical voltage was seen to develop across the membrane.
Therefore, this work is designed to investigate the fundamental relation between electrical activity and reaction mechanism by which the novel electrochemical-catalytic membrane is able to react NO.
Catalytic perovskite materials are arranged such that two metal-based catalytic layers are separated by and in contact with a dielectric membrane for the study of NO reduction. The resulting electrochemical catalytic membrane is subjected to model exhaust flows within a controlled laboratory to serve as an initial investigation into the emission control potential.   [18][19][20][21][22][23] . Although these materials are commonly utilized in the production of high temperature solid oxide fuel cells, their purpose here is to act only as a catalytic membrane. As such, the nickel oxide was not reduced to nickel prior to testing. Pure silver wire electrodes were then added to both the support and outer layers via silver paste at the edge of the layer. A test was constructed in order to investigate the fundamental mechanism by which the cathode layer of the electrochemical-catalytic membrane was able to significantly breakdown NO. The new electrochemical-catalytic membrane, with 0.81 cm 2 of reactive surface area, is tested for NO reduction against: 1. VAG catalytic converter with 0.81 cm 2 of reactive surface area; 2. The same electrochemical-catalytic membrane with the anode and cathode wired together to resemble a short circuit condition (hereafter referred to as an external short circuit or short circuited); and 3. The same electrochemical-catalytic membrane with an external 3 V, 8000 μs (4% duty cycle) pulse width modulation (PWM) signal with 0.81 cm 2 of reactive surface area.
Each specimen was held at 600 °C and subject to 5 ml/min flow of a certified premixed gas cylinder of 10 vol% NO and 90 vol% N 2 . All testing equipment was cleaned, dried, and flushed with Ar for 1 h within the furnace at 600 °C prior to testing, to ensure no contamination of results. Only N 2 and NO were supplied to the testing apparatus. During each test, the internal furnace environment surrounding the testing equipment was flooded with Ar, in order to ensure an external inert environment and to detect any leakage. For the electro-chemical catalyst, the NO flow was sent to the cathode side of the membrane. The voltage signal for both the electrochemical catalytic membrane and the short-circuited membrane were recorded with a 100 MHz oscilloscope.
For both of the experimental setups, the effluent of each specimen, recorded with a Hiden Analytic QGA MS with a 100 ppb detection limit, verified via GC by a SRI 8610C and an E-instruments E-8500 emission analyzer, is compared to the baseline effluent. The experimental setup for the second, one sided test is depicted below in Fig. 2. The experimental setup for the initial study is the same, except before the effluent leaves the testing chamber it is recycled over the opposite side of the test specimen and then is exhausted.

Results and discussion
Experimental results of NO reduction phenomenon. The electro-chemical catalytic membrane, consisting of two metal electrode surfaces separated by a dielectric electrolyte, was able to achieve a reduction in NO concentration greater than 2X that of a traditional PGM catalyst as seen in Table 2 above. Additionally, wiring an external oscillating voltage signal between the two metal electrodes of the membrane increased reduction potential by an additional 10%. This behavior is in stark contrast to that of the electro-chemical catalytic membrane with the addition of an external short circuit, which was only able to convert an amount of NO similar to a traditional PGM catalyst. Figure 3 shows the comparison of the voltage signal generated by the electro-chemical catalytic membrane and the electro-chemical catalytic membrane with an external short circuit during the NO reduction. The externally short-circuited membrane is centered on a voltage potential of 0, whereas the membrane without a short circuit is centered around -6.0 mV, under the standard convention of the cathode being the electrically positive electrode.
Additionally, the oscillations of the membrane without a short circuit are 2X the magnitude of those seen in the short-circuited membrane. Amplification of voltage oscillations within the membrane, as with the addition of a PWM external source, increases the total amount of NO reduction. Therefore, the ability of the electro-chemical catalytic membrane to charge and discharge while interacting with an external flow is the key difference allowing for significantly increased reaction rates.
The charge, discharge behavior of the electro-chemical catalytic membrane significantly differs from the reaction pathway previously accepted in literature. www.nature.com/scientificreports/ The electro-chemical catalytic membrane shows significant deviations from all other specimens in the mass spectrum at 17, 18, 30, 32 and 44 amu, as shown in Fig. 4. The difference at 30 and 32 amu represent a reduction in NO and production of O 2, respectively. The large peak at 18 amu, followed by sub-peaks at 17 and 16 amu indicates lone and charged oxygen. The peak at 44 amu indicates a concentration of N 2 O. The decrease in magnitude at 18 amu and the substantial peak at 44 amu represents a significant deviation from the majority of literature for NO breakdown. NO reduction has traditionally been separated into two categories: 1. Chemical gradient and equilibrium with or without a catalyst present 24,28 ; 2. Combustion reaction [29][30][31][32] .
A high concentration of NO will drive an equilibrium reaction given sufficient energy and/or the presence of a catalyst [31][32][33] .
NO may also undergo a combustion reaction with ammonia (NH 3 ), air and/or hydrocarbons (C x H y ) in which NO will typically pass through an intermediary of NO 2 29,34-38 . In this study, there is clear formation of N 2 O without additional reactants being sourced to aid the reduction of NO. Previously, an intermediary of N 2 O during the breakdown of NO had only been recorded while supplying ammonia to a platinum catalyst at 850 °C 29,[31][32][33][34][35][36][37] . It is therefore believed that the electrical oscillation activity of the electro-chemical membrane is the reason for an altered reaction pathway and significant reduction of NO. This hypothesis is primarily investigated through a Gibbs free energy analysis. The following table lists potential reactions and respective changes in Gibbs free energy for a temperature range near the testing point of 600 °C in the breakdown of NO and eventual formation of N 2 and O 2 .  www.nature.com/scientificreports/ For simplicity of analysis, it is assumed that the first step in the breakdown of NO is the formation of N 2 O 2 , as has been reported previously in literature [17][18][19][20][21][22][23][24][25][26][27][31][32][33][34][35][36][37][38][39][40][41][42] . Therefore, the key equations to investigate in Table 3, given sufficient concentration of N 2 O 2 , are reaction equations numbers 8 and 10. It is clear that at 600 °C the formation of NO 2 is energetically favorable to the formation of NO 2 , reaction 10, if the Gibbs free energy minimization is the only driving factor. However, it also indicates that the formation of N 2 O, reaction 8, could also be spontaneous but is less probable to occur than reaction 10. Therefore, it is reasonable to assume that an electrochemical manipulation/influence could drive the likelihood of occurrence toward reaction 8 and the formation of N 2 O.  www.nature.com/scientificreports/ Additionally, the transfer efficiency from NO to N 2 is considered in conjunction with the Gibbs free energy analysis. Table 3 indicates in reactions 12 and 14 that N 2 O as an intermediary species should result in a higher yield of N 2 when compared to NO 2. Table 4 above calculates the transfer efficiency factor from NO reduction to N 2 production. The ideal increase in N 2 is calculated from the reduction of NO reported in Table 2, and represents the amount of N 2 that could be produced if the transfer efficiency were equal to 1. That is, if all NO reduction resulted in the formation of N 2 . It is clear to see that the electrochemical catalytic membrane with enhanced electrical activity has a significantly better transfer efficiency at 0.94. There is a direct correlation between N 2 O production and increased N 2 transfer efficiency. The negative factor for the PGM catalytic converter is not considered to be abnormal for catalytic converter operation. It has been reported that commercial PGM catalytic converters do have limited N 2, NO storage capability, which would account for the discrepancy in N 2 concentration in the exhaust 4-12 . Formation of N 2 O is seen in the mass spectrum analysis of the electrochemical catalytic membrane and more predominately in the electrochemical catalytic membrane with forced electrical oscillations. It is not as evident in the effluent of the traditional PGM catalyst, nor the electrochemical catalytic membrane with an electrical connection between the anode and cathode. Therefore, when the electrical activity on the membrane is not present, the formation of N 2 O is significantly diminished.
The deposition, removal, and rearrangement of electrons from/on the catalytic surface is believed to be the key in altering the probability of reaction pathways. Traditional catalytic reaction chemistry with a solid, nonconsumable catalyst without the presence or possibility of water formation or the presence of a proton acceptor/ donor, assumes that the catalytic surface, after absorbing the reactant, acts primarily as an electron transport mechanism around the reactant molecule 43,44 . For the above experimental results, the apparatus was free of any water vapor, as described in the methods section, and the test flow consisted of only nitrogen and oxygen, Table 3. Gibbs free energy of reaction analysis summary of potential chemical reaction equations.  www.nature.com/scientificreports/ eliminating the possibility of acid-base catalysis. Within any catalyst in which there is electron movement around the reactant, there must be an ohmic like resistance. However, if the charge on the catalytic surface is forced to vary, i.e. the surface is forced to charge and discharge, electron movement to and from the surface is established prior to the absorption of the reactant molecule. The addition and removal behavior of electrons that is established on the surface is believed to manipulate the natural reaction pathways increasing the overall reactivity of the system. In addition, this alternating voltage potential creates a fluctuating external electric field capable of doing work and exciting any incoming polar molecule. The external field may also manipulate the collisions of the incoming reacting species by forcing varying molecular alignment as the flow approaches the catalytic surface.

Conclusions
The unique configuration of a metal-based catalytic surface connected through a dielectric membrane to another metal-based electrode allows for the potential of surface charging and charge communication without significant internal current flow or internal charge balancing. This charging and discharging of the catalytic surface allows for the altered chemical reaction pathways demonstrated in this work. When the two metal electrodes of the electro-chemical catalytic membrane are electrically connected, the reactivity decreased by a factor of ~ 2. Any charge that began to develop on the surface was able to relax back into equilibrium via current flow. Therefore, the surface charge was no longer capable of reacting with the incoming flow, decreasing efficacy of NO conversion. Further investigation into this phenomenon has led to the consideration of CO and CO 2 conversion to particulate carbon and gaseous oxygen. Although this exploration is in its infancy, it does indicate potential for the electrochemical manipulation of carbon based emissions through electric potential oscillation. However, unlike the work described in the background of this work 17 , initial testing has shown significant carbon deposition and carbon coking of the electrode layer during prolonged operation in a purely CO and CO 2 environment, decreasing reactivity over time. Additional work is required to develop a technology capable of continually reacting CO and CO 2 into particulate carbon and gaseous oxygen without the addition of other reactants to prevent coking.