Sobhan Neyrizi

Sobhan Neyrizi Imidazolium co-catalysts for efficient electrochemical reduction of CO2 in acetonitrile


IMIDAZOLIUM CO-CATALYSTS FOR EFFICIENT ELECTROCHEMICAL REDUCTION OF CO2 IN ACETONITRILE DISSERTATION to obtain the degree of doctor at the University of Twente, on the authority of the rector magnificus, prof. dr. ir. A. Veldkamp, on account of the decision of the Doctorate Board to be publicly defended on Thursday 30 November 2023 at 14.45 hours by Sobhan Neyrizi born on the 22nd of May, 1989 in Ray, Iran

This dissertation has been approved by: Promotor G. Mul Co-promotor dr. M.A. Hempenius Cover design: Ella Maru studio Printed by: Ridderprint Lay-out: Sobhan Neyrizi ISBN (print): 978-90-365-5912-6 ISBN (digital): 978-90-365-5913-3 URL: © 2023 Sobhan Neyrizi, The Netherlands. All rights reserved. No parts of this thesis may be reproduced, stored in a retrieval system or transmitted in any form or by any means without permission of the author. Alle rechten voorbehouden. Niets uit deze uitgave mag worden vermenigvuldigd, in enige vorm of op enige wijze, zonder voorafgaande schriftelijke toestemming van de auteur.

Graduation Committee: Chair / secretary: prof.dr. J.L. Herek Promotor: G. Mul University of Twente, TNW, Photocatalytic Synthesis Co-promotor: dr. M.A. Hempenius University of Twente, TNW, Sustainable Polymer Chemistry Committee Members: prof. dr. J. Huskens University of Twente, TNW, Molecular Nanofabrication prof. dr. J. Lange University of Twente, TNW, Sustainable Process Technology prof. dr. M.T.M. Koper Leiden University, Leiden Institute of Chemistry prof. dr. P.J.A. Kenis University of Illinois Urbana-Champaign, Chemical and Biomolecular Engineering dr. S. Er DIFFER

This thesis is dedicated to the memory of my father, Qasem Neyrizi (), may his memory forever be a comfort and a blessing.

Summary The combustion of carbon-based fuels has led to the persistent accumulation of carbon dioxide (CO2) in the Earth's atmosphere, resulting in far-reaching climate changes. To address this pressing global issue, there is an urgent need for sustainable and efficient methods to harness CO2 as an energy source, thus mitigating its environmental impact. One of the most promising approaches for CO2 utilization is electrochemical reduction, often carried out in water due to its green and abundant nature. However, the intrinsic electrochemical reactivity of water presents challenges in this process This PhD thesis explores the potential for improving carbon dioxide (CO2) reduction efficiency by transitioning from water-based electrochemical processes to non-aqueous media, using acetonitrile as a solvent. To enhance efficiency, imidazolium cations are introduced as cocatalysts and essential components of the electrolyte. The central aim of this work is to gain a fundamental understanding of the pivotal role of imidazolium cations in promoting nonaqueous electrochemical CO2 reduction. Additionally, it investigates their potential in facilitating efficient electrochemical conversion across a variety of affordable transition metals. By analyzing the structure-activity relationship for imidazolium electrolyte cations, this research provides insights into the selection or synthesis of effective electrolyte cations to enhance non-aqueous electrochemical CO2 reduction.

Samenvatting De verbranding van koolstofhoudende brandstoffen heeft geleid tot de aanhoudende ophoping van koolstofdioxide (CO2) in de atmosfeer van de aarde, met als gevolg ingrijpende klimaatveranderingen. Om dit urgente wereldwijde probleem aan te pakken, is er een dringende behoefte aan duurzame en efficiënte methoden om CO2 als energiebron te benutten en daarmee de milieueffecten te verminderen. Een van de meest veelbelovende benaderingen voor CO2gebruik is elektrochemische reductie, vaak uitgevoerd in water vanwege de groene en overvloedige aard ervan. De intrinsieke elektrochemische reactiviteit van water brengt echter uitdagingen met zich mee in dit proces. Dit proefschrift verkent de mogelijkheden om de efficiëntie van de reductie van kooldioxide (CO2) te verbeteren door over te stappen van elektrochemische processen in water naar nietwaterige media, waarbij acetonitril als oplosmiddel wordt gebruikt. Om de efficiëntie te verbeteren, worden imidazoliumkationen geïntroduceerd als mede-katalysatoren en essentiële componenten van de elektrolyt. Het centrale doel van dit onderzoek is een fundamenteel begrip te verkrijgen van de cruciale rol die imidazoliumkationen spelen bij het bevorderen van elektrochemische CO2-reductie in niet-waterige omgevingen. Daarnaast onderzoekt het hun potentieel om efficiënte elektrochemische conversie te vergemakkelijken over een verscheidenheid aan betaalbare overgangsmetalen. Door de structuur-activiteitsrelatie voor imidazolium-elektrolytkationen te analyseren, biedt dit onderzoek inzicht in de selectie of synthese van effectieve elektrolytkationen om de niet-waterige elektrochemische CO2-reductie te verbeteren.

Table of Contents Chapter 1: CO2 Reduction in Non-Aqueous Media: Advantages, Mechanism, and Promotional Effects of Imidazolium Cation …………………………….……………….…………….…….……………………….....[9-15] Chapter 2: Advancements in Non-Aqueous Electrochemical CO2 Reduction: Design and Optimization of Experimental Protocols ……………………………………………......[17-46] Chapter 3: Unveiling the Role of Imidazolium Cations in Promoting Electrochemical CO2 Reduction……………………………………………………………………………......[47-98] Chapter 4: Investigating the Influence of C4, C5-Substituted Imidazolium Cations on Electrochemical CO2 Conversion……………………………………………………...[99-133] Chapter 5: Exploring the Structure-Activity Relationship of Late-Transition Metal Catalysts in Imidazolium-Assisted Electrochemical CO2 Reduction……………..….[135-164] Chapter 6: Exploring Electrochemical CO2 Reduction Catalyzed by 2-Methylated Imidazolium: Insights from ATR-FTIR Spectroscopy…………………………….…[165-188] Chapter 7: Exploring the Influence of Electrolyte Identity on Non-Aqueous CO2 Reduction: Impact of Alkyl Chain Length and Alkali Metal Cations.……………….[189-200] Chapter 8: Summary and Perspectives…………………………………………..[201-208] Appendix…………………………………………………………………………...[209-212] References………………………………………………………………………….[213-220] Acknowledgments…………………………………………………………………….[221] Publications.…………………………………………………………...……………….[222]

Chapter 1: CO2 Reduction in Non-Aqueous Media: Advantages, Mechanism, and Promotional Effects of Imidazolium Cation

 1.1. Why electrochemical reduction of CO2 and associated challenges The electrochemical reduction of CO2 (CO2RR), driven by electrical energy from renewable sources, holds significant promise in mitigating the escalating levels of atmospheric CO2 and the resulting impact on Earth's climate. Moreover, such process presents an alternative avenue to traditional fossil resources as a sustainable carbon source9. To achieve efficient electrochemical CO2 reduction, substantial advancements in catalyst properties 10-12, electrode design (such as gas diffusion electrodes)13-15, and electrolyte composition16-18 have been reported. However, several challenges still need to be addressed before the technology becomes commercially viable19-20 , particularly in the context of aqueous electrolytes. These challenges encompass the low solubility of CO2 in water 21-23 , the acidification of the electrolyte due to CO2 dissolution (resulting in bicarbonate/carbonate formation and salt precipitation) 24-26, the concurrent occurrence of the competing hydrogen evolution reaction27-28, and the instability of metal and metal oxide catalysts in acidic aqueous environments29-30. 1.2. CO2 reduction in non-aqueous solvents The electrochemical inertness of organic solvents, such as acetonitrile, presents potential solutions to alleviate some of the issues mentioned earlier with aqueous media21, 23, 31. Notably, CO2 solubility in acetonitrile is approximately eight times higher than in water 32, a critical parameter that promotes enhanced CO2 utilization within the context of reactor design. The absence of water ensures a high selectivity towards CO2 reduction products. This is further reinforced by the consideration of the broad electrochemical window (electrochemical stability) of organic media, as well as the fact that CO2 becomes the sole redox-active species on the electrode-catalyst surface. For instance, acetonitrile and propylene carbonate have been reported to exhibit electrochemical windows of 6.1 V and 6.6 V, respectively 33.

 This extended electrochemical window provides ample room for accommodating diverse redox chemistries, a considerable advantage compared to the constraints posed by water as a solvent, with its electrochemical window typically around 2.06 V (the exact value may depend on the solution's pH). The advantages discussed above have spurred investigations into CO2 reduction in non-aqueous media, although the number of such studies is not as extensive as those conducted in aqueous media. By utilizing a two-layered carbon-free lead (Pb) gas diffusion electrode (GDE), Konig et al.34 demonstrated a 53% Faradaic efficiency for oxalate production at a current density of approximately 80 mA/cm2 and a potential of approximately -2.5 V vs. Ag/Ag+. Their study employed a 0.1 M tetraethylammonium tetrafluoroborate solution in acetonitrile as the electrolyte. In a different study, Tomita et al.35 observed that for a Pt electrode, the main product was oxalic acid at a current density of 5 mA/cm2 using a 0.1 M tetraethylammonium perchlorate acetonitrile-water mixture. However, an increase in water concentration led to a decrease in oxalic acid formation and an increase in formic acid production. At higher water concentrations, hydrogen evolution became dominant. Another study by Figueiredo et al.21 reported CO as the primary product in wet acetonitrile using an ammonium-based electrolyte. The formation of CO, accompanied by formate and carbonate, was also reported by Christensen and Hamnett36 using a 0.1 M tetrabutylammonium tetrafluoroborate solution in acetonitrile over an Au electrode. As we can observe, the performance of CO2 reduction in non-aqueous media has exhibited significant variations dependent on factors such as the composition of the electrolyte, the nature of the electrode, and the availability of proton donors. For readers interested in an extensive review, a recent work by Reis et al.37 offers a comprehensive overview of CO2 electrochemical reduction studies in non-aqueous media.

 1.3. Electron transfer and reaction intermediates mediated by imidazolium cations From a mechanistic perspective, the initiation of CO2 reduction has been argued to involve a first electron transfer to the CO2 molecule 19, 38. In the case of a metal electrode catalyst, this step follows CO2 adsorption. The following central equation can be considered for CO2 activation:   [Eq 1.1] Due to the high reactivity and short lifespan of this intermediate, as identified by Bard et al.39 , the predominant reaction and subsequent product distribution considerably depend on the nature of the electrode material40, the surrounding microenvironment41-42, and the availability of protons43. In situations involving low-proton aprotic media, numerous studies have documented the formation of CO, oxalate, and carbonate as the primary products. Additionally, in the presence of residual water, other products like hydrogen, glyoxalate, glycolic acid, glyoxylic acid, and formic acid have been observed in the reaction mixture44-48. The formation of CO in non-aqueous media has predominantly been linked to a disproportionation reaction. In this process, upon the second electron transfer to an adsorbed *CO2 intermediate, carbonate and CO are simultaneously generated in equimolar proportions. Oxalate, on the other hand, has been proposed to emerge from a self-coupling reaction of adsorbed *CO2 intermediates, while the presence of water has been suggested to drive the formation of formate 31, 36.The first reaction stage mentioned in Eq. [1.1] requires a significantly negative standard potential, as reported to be -2.21 V vs. SCE on a mercury electrode in DMF solvent44. Recent work by Koper et al. 49 has even demonstrated the absence of CO2 conversion in the absence of solvation effects that stabilize the negatively charged CO2 intermediate. Without water as a potential

 moderator for energy requirements during the initial electron transfer, the utilization of electrolyte cations to engineer the kinetics of CO2 reduction appears to be a promising approach. Among the electrolyte cations, imidazolium cations have shown great promise for CO2 reduction in both aqueous and non-aqueous media. Rosen et al. reported a 96% faradaic efficiency for CO formation on an Ag electrode in an ethyl methyl imidazolium-water mixture (18 mol % EMIM BF4 in water) with a cell voltage of 1.5 Volts (+0.2 V overpotentials)50. Following this seminal work, various attempts have been made to utilize imidazolium-type molecules to enhance CO2 reduction in non-aqueous media. Lau et al. employed C2functionalized imidazolium cations to achieve an onset potential of around -2.0 V vs. Fc/Fc+ for CO2 reduction over an Ag electrode in acetonitrile 8, and Atifi et al. reported the use of butyl methyl imidazolium hexafluorophosphate for the conversion of CO2 to CO with 85% FE using a Bi electrode51. Sung et al. also demonstrated the improved efficiency of a molecular Lehntype catalyst through the incorporation of imidazolium species into the secondary coordination sphere52. Despite these efforts, the reported onset potentials and associated energy losses in imidazolium solutions remain relatively high. Moreover, the fundamental understanding of the function of imidazolium cations is limited, and several hypotheses regarding the promotion mechanism have been proposed. These hypotheses include: i) the coordination of the cation with the adsorbed CO2 intermediate, the nature of which still requires clarification 3, 50, ii) the suppression of the H2 evolution reaction 53-54, , iii) the formation of an intermediate imidazolium carboxylate providing a low-energy pathway for the conversion of CO2 to CO3-4, 55, or iv) the stabilization of the high-energy *CO2¯ intermediate (on Ag surfaces) through hydrogen bonding via the C4-H or C5-H functionality of C2-substituted imidazolium cations (Figure 1.1) 8. However, different reaction conditions have been previously applied, including variations in applied salts (different anions and C2-substituted imidazolium cations) and variable water

 content in solvent compositions, making it difficult to discriminate between these hypotheses and to properly assess the general function of the cations based solely on existing literature. Unraveling the underlying role of these cations in CO2 reduction performance holds the potential to open new avenues for the efficient design of non-aqueous electrochemical CO2 reduction systems. This thesis embarks on a comprehensive investigation into the influence of imidazolium cations on the performance of (predominantly) Au electrodes in the electrochemical reduction of CO2 in non-aqueous media. Rigorous measures have been taken to maintain anhydrous conditions, mitigating the potential interference from water (Chapter 2). Au electrodes were chosen as a benchmark catalyst facilitating a systematic exploration. The research strategy involves a combination of synthetic methodologies, electrochemical Figure 1.1. (a) Schematic representation of interactions involving imidazolium, as proposed by Kamet et al.3 This includes cation coordination with adsorbed intermediates, carboxylation, and a side reaction leading to oxalate formation as a byproduct on a Pd electrode. (b) Proposed mechanism by Lau et al.8 involving hydrogen bonding to adsorbed *CO2 intermediate as the main mechanism for C2-methylated imidazolium on Ag electrode. These studies were conducted in an acetonitrile environment. a b

 experiments, and Density Functional Theory (DFT) calculations aimed at providing deep insights into the promotional effect that imidazolium cations exert on CO2 reduction. In particular, the influence of substituents at both C1, C3- and at the N1, N3- positions of the imidazolium ring on its catalytic performance will be scrutinized (Chapters 3 and 4). This analysis will shed light on the structure-activity relationships governing the impact of cations on CO2 reduction. Expanding the investigation, the performance of anhydrous imidazoliumacetonitrile will be systematically examined on other electrode materials, including Ag, Zn, Cu, and Ni (Chapter 5). This broader exploration aims to understand the generality of the observed effects and the potential for cation-mediated enhancement across different catalysts. Furthermore, the performance of C2-methylated imidaozlium cation will be investigated using in-situ FTIR spectroscopy (Chapter 6). Finally, the impact of the alkyl chain length of 1-alkyl3-methyl imidazolium cations on their catalytic efficiency will be also discussed (Chapter 7).

Chapter 2: Advancements in Non-Aqueous Electrochemical CO2 Reduction: Design and Optimization of Experimental Protocols

 Summary Chapter 2 delves into an investigation of the applicability of common practices and experimental protocols for non-aqueous electrochemical CO2 reduction. The chapter focuses on exploring the influence of widely used Ag/Ag+ reference systems and metallic (such as Ag) counter electrodes, shedding light on the associated artifacts that arise when studying the CO2 reduction process in non-aqueous media. Building upon the limitations of existing methods, a novel experimental protocol is developed to establish a robust approach. This work introduces key modifications that address two critical aspects: a) ensuring reliable potential recording for the working electrode, which serves as a vital parameter for assessing catalytic performance, and b) providing detailed steps to prepare and conduct experiments under anhydrous conditions.

 Anhydrous Electrochemistry: Methods and Considerations The main objective of this study was to investigate the co-catalyzed reduction of CO2 by imidazolium cations over Au electrodes in the absence of water. To ensure anhydrous conditions throughout our experiments, significant emphasis was placed not only on maintaining anhydrous conditions during preparation, but also throughout the entire course of the experiments. To achieve this, a detailed protocol was established and elaborated on in this section, which ensured relatively consistent water content (<50 ppm). This was essential to ensure the integrity of the results and prevent any interference from variations in water content on the performance and assessment of the reaction mechanism. Throughout this chapter, a model system consisting of 1,3-dimethyl imidazolium NTf2 (MM NTf2) in anhydrous acetonitrile was employed to investigate the experimental approach. The performance of the MM cation in promoting CO2 reduction will be discussed in detail in Chapter 3. Electrodes and reactor. For each electrolysis measurement, the electrodes underwent an initial polishing process using sandpaper until a smooth and shiny surface was achieved. In the case of voltammetry electrodes, a polishing pad (Prosense, QVMF 1040) was moistened with ethanol and the electrode was gently polished for a duration of 4 minutes (no alumina was utilized on the polishing pad). Subsequently, the electrodes were thoroughly rinsed with MilliQ water (Milli-Q® Reference, 18.2 M, 5 ppb TCO, Merc) and subjected to a 10-minute sonication in 0.5 molar HNO3. Following this step, another round of sonication in ethanol was performed, concluding with a 10-minute sonication in HPLC grade acetonitrile (99.9%, Sigma Aldrich) to ensure the electrodes' cleanliness. It is important to note that the electrodes were always rinsed with Milli-Q water between each sonication step. Noble electrodes such as Au underwent additional electrochemical cleaning (Supporting Information Section III) as necessary. However, general electrochemical cleaning as proposed

 by some studies56 for electrodes such as Cu resulted in poor electrochemical results with unassigned peaks (Supporting Information Figures S2.7a-S2.7c) . Thus, it is recommended to avoid electrochemical cleaning with non-noble metals, and if necessary, rigorous sandpaper polishing (mechanical cleaning) should be applied. The glass reactor, reference compartment, graphite rod, gas inlet, and gas outlet tubes underwent a thorough rinsing process using Milli-Q water, ethanol, and HPLC grade acetonitrile. Subsequently, the working electrode and all other reactor components were assembled, and the entire setup was sonicated for 2 sets of 5 minutes using HPLC grade acetonitrile. Following this step, the reactor was subjected to a continuous purge of super dry helium (Helium A/Zero Grade N4.6, Linde) for a duration of 30 minutes prior to the introduction of the electrolyte. In experiments involving CO2, a dryer (ZPure DS H2O, ChromRes) was employed to ensure the complete removal of moisture from the CO2 inlet (Carbon Dioxide Food Grade, Linde) before it entered the reactor. This additional step ensured that the CO2 supplied to the reactor was free from any residual moisture. Figures S2.1 displays the results of SEM and EDX analysis from Ni and Au electrodes after the above-mentioned cleaning procedure. The surface appears to be sufficiently clean for being tested in electrochemical experiments. Counter electrode. A graphite rod was chosen as the counter electrode to mitigate additional artifacts resulting from metal oxidation and subsequent deposition onto the working electrode. The choice of counter electrode plays a crucial role in the electrochemical CO2 reduction process. Figure 2.1 presents a comparison of chronoamperometry results for CO2 reduction on an Au electrode using different counter electrodes: Au and Ag (Figure 2.1a) as well as graphite (Figure 2.1b). The results show that employing Au and Ag as counter electrodes leads to an increase in current, suggesting the occurrence of additional reactions. On the other hand, when graphite is used as the counter electrode, the performance of the CO2 reduction remains stable,

 with a consistent Faradaic efficiency for CO production. Figure S2.8 further supports the influence of the counter electrode choice. It displays images of an Au voltammetry electrode before and after conducting linear sweep voltammetry (LSV) experiments for CO2 reduction, with Ag employed as the counter electrode. The deposition of silver on the Au electrode during the experiment results in a noticeable transformation, resembling the appearance of silverish. However, when using graphite as the counter electrode, no significant changes in the electrode's appearance are observed after the voltammetry experiments. These observations emphasize the importance of selecting the appropriate counter electrode material in non-aqueous electrochemical CO2 reduction studies to minimize unwanted side reactions and artifacts. Glassware to prepare electrolyte solutions. All glassware intended for solution preparation underwent a thorough cleaning process. Firstly, the glassware was sonicated in Milli-Q water for a duration of 10 minutes. Following that, a subsequent round of sonication was conducted using ethanol for an additional 10 minutes. This ensured the removal of impurities or residues from the glass surfaces, ensuring a clean environment for solution preparation. After the cleaning process, the glassware was dried at a temperature of 200 degrees Celsius for a minimum of 2 hours, immediately transferred to the glove box antechamber while hot, where Figure 2.1. Chronoamperometry results for Au electrode under CO2 purging with 0.5 mol% of MM NTf2 as electrolyte in anhydrous acetonitrile. (a) with Au and Ag as counter electrode. and (b) with graphite as counter electrode. Both experiments were performed at -1.8 V vs. Ag/Ag+. The higher current observed for the experiment with graphite is due to the larger size of the working electrode (Au foil). a b

 it was subjected to five evacuation-N2 refilling cycles before it was taken into the glove box. This ensured the complete elimination of any remaining moisture. The dried glassware was utilized for the preparation of electrolyte solutions and reference solutions using anhydrous acetonitrile. Electrolyte solutions always contained 0.5 molar percent of the electrolyte salt target to study. The electrolyte solutions were formulated to contain a target concentration of 0.5 molar percent of the electrolyte salt under investigation. Reference electrode and solution preparation. In order to avoid the cross-contamination of the working solution with other organic cations, for each cation, a new reference solution was prepared for the electrochemical measurement. A reference solution always contained 0.1 mol% (0.02 molar) of Ag OTf with 0.4 mol% of the electrolyte subject to study (in total 0.5 molar percent of salt concentration in anhydrous acetonitrile)57. The reference solution was separated from the working solution by an ultrafine frit. Ag wire was used as the pseudoreference electrode. The potential recorded versus the Ag reference electrode immersed in an electrolyte containing 0.02 molar silver salt in acetonitrile can be converted into the SHE scale by the following equation58:  +542 mV vs. SHE (± 45 mV) Electrochemical measurements. Solutions (electrolyte and reference solutions) were transferred with caution into the reactor with gas-tight syringes. Before solution injection, the reactor was washed twice with anhydrous acetonitrile (transferred from the glove box). After injection, the solutions were kept under He/CO2 purging for 1hr to remove any remaining oxygen. Karl-Fischer titrations were performed to measure the water content from solutions inside the glove box and after the injection into the reactor. 48-55 ppm water was found for all solutions and it was confirmed that no water was introduced to the solutions upon transferring

 from the glove box into the reactor. Figure S2.9 illustrates the electrochemical set-up utilized in this study. For recording potentials versus Ag/Ag+, an Ag wire was used as the pseudo-reference electrode. 0.1 mol% Ag OTf in 0.1 mL anhydrous acetonitrile was always used as the reference solution and was separated from the working solution with an ASTM ultrafine frit. Working solutions always contained 0.5 mol% of the electrolyte and were all prepared inside the glove box. NTf2 was always used as the common electrolyte anion and anhydrous acetonitrile was the common solvent for all measurements. A graphite rod (99.99% Sigma Aldrich) was used as the counter electrode. Gas chromatography (Compact GC 4.0, Inter science) was used to analyze the gas products from the reactor. He was used as the carrier gas and the Pulsed Discharge Detector of the GC was calibrated for 1 to 100,000 ppm CO (Carbon Monoxide CP Grade N3.0, Linde). All electrochemical measurements were performed with a Biologic SP-300. RDE measurements were performed using a WaveVortex 10 Electrode Rotator (Pine research). Figure 2.2 illustrates the sensitivity of our non-aqueous setup to the presence of water. The figure showcases the LSV (Linear Sweep Voltammetry) results obtained from the Ni electrode under He purge, with varying levels of residual water concentration. Notably, the findings reveal a discernible increase in the reductive current upon the introduction of water into the system in the potential range of -1.0 to -1.5 V, lower than required for reduction of CO2. The reduction of H2O becomes mass transfer limited in the range of -1.5 V to -2.0 V. To ensure consistent hydrodynamics and minimize variations during linear sweep voltammetry (LSV) and electrolysis (EL) measurements, several parameters were kept constant. These parameters included maintaining a consistent reaction volume size, counter electrode size, and gas flow rate (5 ml/min). These measures were implemented to ensure reliable and meaningful comparisons of the apparent relative activities when studying different electrolytes or electrodes. Prior to any electrochemical measurement, it was ensured that a stable open-circuit

 voltage was achieved. Reproducibility of measurements was checked to ensure reliable data. Due to its inherent activity for water reduction, a Nickel electrode was used to probe the impact of the presence of water experimentally. Gas Analysis and Faradaic Efficiency Evaluation Method. To evaluate Faradaic efficiency for a reaction, gas and liquid products were analyzed. For gas analysis, the GC was first calibrated for the relevant ppm's of CO and other potential products for CO2 reduction. Figure S2.4 shows examples of the calibration data for CO and H2. The pressure drop over the lines of the GC was taken into account for calculating CO concentration. The flow in was always contrasted with the flow out to ensure a leak-less experiment. Before electrolysis, a few sequences from the reactor were recorded to obtain the background with no reaction. After obtaining the background, electrolysis started and CO was detected (Figure S2.5). After GC measurement, the pressure drop was re-measured to ensure that the experimental conditions were the same as at the beginning of the electrolysis. Based on calibration measurements, CO ppm was evaluated and then converted to mol fraction by the use of GC pressure drop, lab temperature, and total volume of the gas. After electrolysis, the solution was always analyzed by NMR to inspect the presence of other possible products. Figure 2.2. LSV results for a Ni electrode under He purge at varying levels of water content. Introducing 500 ppm (red) and 2000 ppm (blue) of water induces a noticeable increase in the reductive current within the same electrochemical window.

 Artifacts and Protocol for Potential Recording in Non-Aqueous Electrochemical CO2 Reduction In our preliminary approach to investigate non-aqueous CO2 reduction through electrochemical measurements, we employed a three-electrode cell, a widely adopted configuration in electrochemical research, as extensively reported by previous researchers59-61. This configuration, as detailed in the previous section, involved the utilization of an Ag wire as the pseudo-reference electrode. For recording potentials relative to Ag/Ag+, a reference solution was employed, consisting of 0.1 mol% (0.02 molar) of Ag OTf with 0.4 mol% of the electrolyte under study. To maintain consistency, the salinity of the reference solution was matched with that of the working solution, thereby preventing salt transport induced by concentration potential between the compartments. To ensure proper separation between the reference solution and the working solution, we incorporated an ASTM ultrafine frit. Moreover, we chose to use a graphite counter electrode to mitigate any potential issues arising from metal deposition from a metallic counter electrode, which could interfere with the reaction mechanism under investigation at the working electrode. Figure 2.3. (a) Linear sweep voltammetry (LSV) results for CO2 reduction on a Pt wire electrode in anhydrous acetonitrile. The electrolyte and co-catalyst used were 0.5 mol% of MM NTf2. Subsequent cycles are indicated by the corresponding numbers. (b) Electrolysis results for the same system demonstrating the consistent production of CO with 100% faradaic efficiency. a b

 Figure 2.3 (a) illustrates the linear sweep voltammetry (LSV) results for CO2 reduction at a Pt electrode using the three-electrode configuration described earlier. The electrolyte and cocatalyst utilized were 0.5 mol% of 1,3-dimethyl imidazolium bis(trifluoromethylsulfonyl)imide (MM NTf2). The detailed discussion on the role of MM cation and other imidazolium cations in CO2 reduction will be presented in future chapters. In this context, the system serves to demonstrate the establishment of the experimental protocol. Form Figure 2.3 (a) we observed a slight increase in the reductive current for the Pt electrode with subsequent cycles. While these incremental changes may not be readily apparent in just two consecutive cycles (as shown in Figure 2.4), they become more noticeable when multiple cycles are overlapped, revealing an overall increase in the current. Figure 2.5 presents the EDX-SEM results obtained from the Pt electrode after electrolysis, as detailed in Figure 2.3. The SEM image highlights two distinct regions. In one of these regions, marked by spectrum 30, a dendritic structure is clearly observed. The corresponding EDX analysis in the top-right corner confirms the presence of Ag (43%), C (29%), and N (18%) elements in this region. Conversely, in the second region, represented by spectrum 31, no dendritic structure is evident. The EDX analysis from this region further supports the hypothesis Figure 2.4. Overlapping the first two cycles of LSVs from Figure 2.3 for demonstration purposes.

 that the presence of Ag, C, and N in spectrum 30 is specifically associated with the dendritic structure. It is worth noting that NHCs (parent N-heterocyclic compounds) have been documented for their ability to coordinate with metallic nanoparticles and heterogeneous metallic surfaces (as shown in the bottom left panel in Figure 2.5)62-64. Metal-NHC complexes have found applications in the fields of catalysis and biochemistry65. The results obtained from Figures 2.3-2.5 allow us to propose a hypothesis that, upon applying a reductive potential, Ag particles from the reference solution and imidazolium compounds present in the working solution may form NHC-imidazole clusters on the surface of the Pt catalyst. However, a more detailed analytical and electrochemical study is required to determine the exact composition and coordination number of these clusters. It is important to highlight that the observed surface coverage with Ag-MM species, as mentioned earlier, is attributed to the use of Ag NTf2 reference solution, which reaches a maximum concentration of 20 ppm in the electrolyte solution if all Ag leaks from the reference solution compartment. Various reference capillaries, both commercially available and custommade, were tested to investigate the possibility of preventing Ag leakage from the reference solution. Unfortunately, regardless of the type of frit used in the capillaries, the deposition of Ag was consistently observed. However, it is important to note that the extent of Ag deposition varied depending on factors such as the magnitude of the reductive potential applied and the duration of electrolysis.

 To mitigate the challenge of reductive Ag deposition in electrolysis experiments, a novel protocol was developed and implemented in this study, which is illustrated in Figures 2.6 and 2.7. In this new approach, the initial step involved conducting all linear sweep voltammetry (LSV) measurements and electrolysis experiments using a graphite rod as both the counter and reference electrode (Figure 2.6a). A critical requirement of this protocol was that the size of the graphite counter electrode consistently exceeded that of the working electrode. Additionally, the applied current in the cell was deliberately restricted to the bulk concentration of the reactant, specifically CO2 in acetonitrile. By adhering to these conditions, the undesired impacts of the counter electrode on the working electrode and its associated reaction were effectively minimized. In the second step of the protocol, once a stable electrochemical response was obtained using the two-electrode configuration, a reference solution containing Figure 2.5. Scanning electron microscopy (SEM) results along with associated energy-dispersive X-ray spectroscopy (EDX) analysis of the Pt electrode investigated in Figure 2.3. The bottom left panel illustrates a schematic representation of the coordination of metal nanoparticles with NHC (N-heterocyclic compound) compounds, adapted from reference5.

 Ag NTf2 salt, identical in composition to the previously described solution, was introduced into the reference capillary. This was followed by the recording of a single linear sweep voltammogram (LSV) using a three-electrode configuration. The setup involved an Ag wire as the pseudo-reference electrode, graphite as the counter electrode, and the working electrode being the Au foil (Figure 2.6b). Importantly, a four-electrode connection was utilized to measure the potentials of both the working electrode and the counter electrode relative to Ag/Ag+. Subsequently, the Ag solution was withdrawn from the capillary to minimize the potential impact of Ag-imidazolium deposition on the working electrode. In the third step, the potential recorded versus Ag/Ag+ in the second step is contrasted with the potential recorded versus graphite in the first step (Figure 2.7a). This comparison allows for the alignment of the previously recorded potential versus graphite with respect to the Ag/Ag+ reference using the following equation:     This alignment step ensures the accurate calibration of the potential recorded versus graphite as a function of potential ( with respect to the Ag/Ag+ reference, enabling reliable comparison and analysis of the electrochemical data. Finally, in the last step, the cell potential recorded in the three-electrode configuration was compared to that of the two-electrode configuration to examine the overlap between the two profiles obtained in steps 1 and 2 (Figure 2.7b). This comparison aimed to ensure the absence of any artifacts or anomalous effects arising from the presence of the Ag salt reference solution. By assessing the consistency and agreement between the cell potentials obtained from both configurations, the reliability and validity of the experimental data were verified, further confirming the effectiveness of the developed protocol in mitigating potential issues associated with reductive Ag deposition during electrolysis experiments. Furthermore, Figure 2.7c presents the scanning electron microscopy (SEM) results along with energy-dispersive X-ray spectroscopy (EDX) analysis of the Au foil used as

 the working electrode in the aforementioned showcase experiment. Contrasting the results of SEM-EDX analysis from the Au electrode with those from the Pt working electrode in Figure 2.5 provides supporting evidence for the efficiency of the experimental protocol devised in this work. Specifically, no evidence of Ag deposition is observed, reinforcing the accuracy of the potentials recorded in steps 1 and 2 as described earlier. WE CE/Ref(Gr) Potentiostat    a     WE CE(Gr) Potentiostat b Figure 2.6 . Illustrating the protocol devised for acquiring electrochemical data in non-aqueous CO2 reduction. (a) Stage 1: Potential recorded using a two-electrode configuration with a graphite rod serving as both the reference and counter electrode. (b) Stage 2: Introduction of a reference solution containing 0.1 mol% of Ag NTf2 to the reference capillary. LSV measurement is conducted in a three-electrode configuration, with an Au foil as the working electrode. The potentials of both the working and counter electrodes are recorded relative to Ag/Ag+ using a four-electrode connection to the potentiostat. The electrolyte employed is 0.5 mol% of MM NTf2.

 Figures 2.8 and 2.9 investigate the sensitivity of our devised protocol to Ag impurities. Deviations in the LSV results recorded in stage 3 suggest interference from Ag leakage and potential Ag deposition (Figure 2.8). Subsequent SEM-EDX analysis (Figure 2.9) provides further evidence of Ag deposition. These findings demonstrate the capability of our protocol to detect and identify the interference from Ag (or other metals) deposition during electrochemical LSV experiments. Figure 2.7 . Illustrating the protocol devised in this study for acquiring electrochemical data in non-aqueous CO2 reduction. (a) Stage 3: Conversion of the potentials recorded versus graphite in step 1 to the corresponding potentials versus Ag/Ag+ by comparing the profiles obtained in steps 1 and 2. Note that each potential may require a specific calibration constant (. (b) Stage 4: Verification of the protocol's accuracy by assessing the overlap of cell potentials obtained in steps 1 and 2. Any deviation between the two profiles could indicate potential interference from Ag leakage- see Figures 2.8-2.9.(c) SEM-EDX analysis providing evidence of a Ag free experiment performed following the devised protocol in this study. a b c

 Figure 2.8. Sensitivity of the devised protocol for acquiring electrochemical data in non-aqueous CO2 reduction. Deviations in the current-potential profiles recorded in stage 3 indicates the possible Ag deposition during the calibration stage, see Figure 2.9 Figure 2.9. Sensitivity of the devised protocol for acquiring electrochemical data in non-aqueous CO2 reduction. SEM-EDX analysis providing evidence of Ag deposition arising from Ag leakage in stage 3.

 The significant implications of conducting electrochemical experiments accurately are further illustrated by comparing the electrolysis results obtained when potentials are recorded versus Ag/Ag+ reference electrode and when our devised protocol is applied (Figure 2.10). Figure 2.10 Electrolysis results for CO2 reduction at -1 mA/cm2 using Ag, Ni, and Zn electrodes in anhydrous acetonitrile. The electrolyte consisted of 0.5 mol% of MM NTf2. a, The reference solution contained 0.1 mol% (0.02 molar) of Ag OTf, and an Ag wire served as the pseudo-reference electrode. b, Potentials were corrected using the experimental protocol devised in this study, relative to the Ag/Ag+ reference. a b

 Figure 2.10a depicts the experimental results for three catalysts (Ni, Ag and Zn) where 0.1 mol% (0.02 molar) of Ag OTf was added to the reference capillary, with an Ag wire serving as the reference electrode. The corresponding electrodes are labeled as Ag 500, Ni 500, and Zn 500. In contrast, Figure 2.10b shows the results obtained when the electrolysis experiments were conducted according to the protocol devised in this study. Interestingly, there are no noticeable differences in the activity of Ni, Zn, and Ag catalysts for CO2 reduction when Ag OTf is consistently used as the reference solution (Ni 500, Ag 500, and Zn 500). However, when the reference solution is removed from the capillary and our protocol is applied to perform electrolysis, the differences between the three catalysts become evident. We also note that all three electrodes display more negative overpotentials for the same current density as compared to Ag 500, Ni 500 and Zn 500. After conducting EDX-SEM analysis on the electrode surfaces following electrolysis (Supporting Information Figure S2.2), it was observed that Ag, carbon, and oxygen were present. This finding is reminiscent of the situation observed for the Pt electrode, as shown in Figure 2.5. It is important to note that the concentration of Ag in the reactor was 20 ppm at maximum, indicating that the deposition of Ag alone cannot account for the improved overpotentials observed for Ni 500, Ag 500, and Zn 500 electrodes compared to the situation when Ag OTf was not used. Therefore, we can conclude that the similar catalytic performance observed for Ni 500, Ag 500, and Zn 500 can be attributed to a modified catalytic behavior associated with Ag-MM deposition during the electrolysis process. While EDX-SEM analysis for electrodes used in Figure 2.9b did not show any evidence of Ag deposition, carbon, nitrogen, and oxygen species were still present for all three catalysts (Supporting Information Figure S2.3). This observation implies the potential deposition of carbonyl and MM cation during electrolysis. Our experiment with the Zn electrode for 100 hours of electrolysis (Figure S2.3c) showed a very slight shift in the overpotential that could always be compensated by the gradual addition of MM cation. In line with the EDX-SEM

 analysis, this observation allows us to assume the gradual deposition of cations over the electrode surface during prolonged electrolysis experiments. In conclusion, the experiments described above show that anhydrous conditions can be assessed easily by noting a reduction potential which is significantly lower than required for the reduction of CO2, while the deposition of Ag can be prevented using the described protocol for measurement of the reference potential.


 Supporting Information Table of Contents I- Results of EDX-SEM analysis II- GC analysis and calibration III- Electrochemical cleaning IV- The effect of counter electrode V- Set-up

 I- Results of EDX-SEM analysis Figure S2.1. The results of scanning electron microscopy (SEM) in conjunction with the associated energy-dispersive X-ray spectroscopy (EDX) analysis from (a) Ni and (b) Au electrodes after undergoing the cleaning procedure described in the main text. a b

 Figure S2.2. The results of scanning electron microscopy (SEM) in conjunction with the associated energy-dispersive X-ray spectroscopy (EDX) analysis from (a) Ag 500 , (b) Ni 500, and (c) Zn 500 electrodes after 5 hrs of CO2 electrolysis. Ag OTf was used as reference solution. a b c

 Figure S2.3. The results of scanning electron microscopy (SEM) in conjunction with the associated energy-dispersive X-ray spectroscopy (EDX) analysis from (a) Ni, (b) Ag, and (c) Zn electrodes after 3 hrs (Ni), 5 hrs (Ag) and 100 hrs (Zn) of CO2 electrolysis. The protocol devised in this study was applied to perform the experiments. a b c

 II- GC analysis and calibration                               Figure S2.4. Calibration of the GC for H2 and CO.

        Figure S2.5. Subsequent sequences of the GC for detection of the products for CO2 reduction. The black line is the background obtained before electrolysis. In this example CO is the only product. The electrolyte was 0.5 mol% MM NTf2 in anhydrous MeCN.

 III- Electrochemical cleaning For noble electrodes such as Au, an electrochemical cleaning procedure was employed after the electrolysis experiments as follows: After sonication in 0.5 molar HNO3, the electrode was subjected to electrochemical washing in a 0.1 molar H2SO4 aqueous solution. The electrochemical cleaning process involved the recording of cyclic voltammograms (CVs) with 20 cycles at a scan rate of 1 V/sec, followed by 10 cycles at 100 mV/sec within the voltage range of 0.2 to 1.5 V versus Ag/AgCl. Figure S2.6 provides an illustration of the electrochemical cleaning procedure for the Au electrode. The cleaning process was continued until repeatable cycles were achieved. However, electrochemical cleaning for non-noble metals such as Cu resulted in poor electrochemical behavior. Figure S2.7a illustrates the cleaning results of a Cu electrode, which does not exhibit a reversible reduction and oxidation cycle as observed for the Au electrode. Comparatively, Figures S2.7b and S2.7c present the cyclic voltammetry results for Au and Cu Figure S2.6. Electrochemical cleaning of Au electrode in a 0.1 molar H2SO4 aqueous solution. (a)The first 20 cycles demonstrate the gradual attainment of repeatable cycles. The initial cycle displays a noticeable oxidation peak, indicating the removal of carbon species from the electrode surface, which was previously exposed to CO2 electrolysis experiments. (b) Subsequent cycles exhibit consistent and repeatable behavior. The electrochemical cleaning process was conducted immediately after CO2 electrolysis in a 0.5 mol% MM NTf2 electrolyte solution in anhydrous MeCN. a b

 electrodes, respectively, after the electrochemical cleaning process. In the case of the Cu electrode, a noticeable unassigned peak is observed during the reduction cycle, while the Au electrode exhibits the expected behavior of a clean electrode. Figure S2.7. (a) Electrochemical cleaning of Cu electrodes in a 0.1 molar H2SO4 aqueous solution. b and c, Cyclic voltammetry in CO2 saturated anhydrous acetonitrile for Cu (b) and Au (c) electrodes after electrochemical cleaning. 0.5 mol% MM NTf2 was added as supporting electrolyte. a b c

 IV- The effect of counter electrode Figure S2.8. Au voltammetry electrode before (a) and after (b) voltammetry experiments under CO2 purging with Ag as the counter electrode. The deposition of silver is visibly observed after the voltammetry experiments. The supporting electrolyte used was 0.5 mol% MM NTf2. a b

 V- Set-up Figure S2.9. Experimental set-up designed for conducting electrochemical CO2 reduction under anhydrous conditions. Solutions were consistently transferred from the glove box and injected into the reactors while maintaining a continuous purge of super dry He or CO2 gas inlet.

Chapter 3: Unveiling the Role of Imidazolium Cations in Promoting Electrochemical CO2 Reduction This chapter is partially based on:                     