Electric-Field Catalysis of Diels–Alder Reactions Authors: Bing-Ru Shao, Shen-Yi Guo, Miguel Paraja, Qing-Xia Zhang, Augustina Jozeliu̅naitė, Naomi Sakai, Stefan Matile Published: 2026 ABSTRACT The dream to catalyze reactions with externally applied electric fields is so obvious and so promising that it has attracted immense scientific attention. The Diels–Alder (DA) reaction has become the primary reaction to elaborate on this dream, but electric-field catalysis (EFC) of DA reactions under practical organic synthesis conditions remained impossible. The objective of this study was to challenge the emerging principles of scalable EFC in microfluidic capacitors with this holy grail in the field. Functional Helmholtz layers are assembled to translate applied electric fields into effective electric fields that are, according to chronoamperometry, strong enough for the EFC. Yields increase nonlinearly from 0% up to 90% with applied voltage. In chiral cationic Helmholtz layers assembled from polyarginines, benzylated quinines, or ammonium binaphthyls, enantioselectivity up to 31% ee is obtained. These results demonstrate that the emerging principles of EFC in microfluidic capacitors pass the ultimate test in the field with distinction: Besides realizing DA-EFC with externally applied electric fields under practical conditions, this study also introduces chiral Helmholtz layers as (i) the primary catalysts, with an impressive efficiency of EC50 = 0.2 mol %, and (ii) the unique source of chirality for asymmetric EFC. Introduction Every chemical reaction involves the movement of electrons from one place to another. The dream to enable, accelerate, and direct these charge displacements with externally applied electric fields (AEFs) has inspired chemists for decades. − It is supported by studies in elegant model systems, − by the local electric fields that enzymes have evolved to accomplish their performance, − and by numerous theoretical investigations. − ,− The elephant in the room, however, is its incompatibility with practical organic synthesis conditions. For electric-field catalysis (EFC), the Diels–Alder reaction has emerged as the primary reaction. − ,− ,− ,− In normal electron demand Diels–Alder (NEDDA) reactions, the charge transfer Q CT occurs from the HOMO of the electron-rich diene to the LUMO of the electron-poor dienophile (Figure A). In their pioneering computational analysis in 2010, Shaik and co-workers showed already that effective electric fields (EEFs) applied along this z-axis either accelerate (+z, + EEF z ) or inhibit the NEDDA (−z), depending on the direction. For inverse electron demand Diels–Alder (IEDDA) reactions, inverse Q CT from the HOMO of the electron-rich dienophile to the LUMO of the electron-poor diene is catalyzed and inhibited by EEFs of the opposite direction. Only recently, Ramanan and co-workers computed that strong EEFs in decelerating −z orientation can cause EFC because the reaction mechanisms change from NEDDA to IEDDA. This concept has been further expanded to Lewis acid catalyzed DA reactions. Polarization of the substrates in the z direction by +z EEFs, visualized best in molecular electrostatic potential surfaces, induces macrodipoles μ z that support the reaction, and can proceed to charge-transfer complexes between the substrates as reactive intermediates. The original computational study by Shaik and co-workers also predicted that EEFs applied along the y axis should influence endo/exo selectivity, and that EEFs could change from concerted to stepwise mechanisms with zwitterionic intermediates. Subsequent studies further predicted that EEFs in the x axis should influence the enantioselectivity of DA reactions. , (A) The Diels–Alder reaction as primary reaction in theoretical studies of EFC. Effective electric fields (EEFs) oriented along the z axis of the charge transfer Q CT during normal (NEDDA) and inverse electron demand Diels–Alder (IEDDA) cycloadditions catalyze or inhibit depending on their direction (+,−); μ defines the notation used also for macrodipoles. (B, C) Previous experimental studies of DA-EFC. Single-molecule break junction experiments with favorably (TS-1), unfavorably (TS-2), and flexibly oriented transition states (TS-3), and in situ MS detection experiments that disfavor operational EFC in water and MeCN microdroplets (TS-4). Experimental support for these computational predictions first emerged 10 years ago from single-molecule studies with scanning tunneling microscopy break junctions (STM-BJs, Figure B). Norbornyl-derived dienophile 1 was covalently attached to a flat gold surface, and diene 2 was attached to a golden STM tip. In this construct, voltage applied between the gold tip and plate produces an AEF that is oriented in the z axis of the NEDDA reaction with transition state TS-1. The occurrence of a reaction can be detected from the tunneling current across the molecular bridge formed. In STM break junctions, single-molecule current reporting NEDDA reactions is blinking because after the reaction, the S–Au bonds of junctions break rapidly. NEDDA reactions were thus characterized as the blinking frequency. This blinking frequency increased linearly with +z-oriented AEFs up to saturation around 500 mV, while inverse −z-oriented AEFs did not catalyze the reaction. Three years later, the complementary IEDDA with AEFs in the x orientation was reported using mechanically controllable break junctions (MCBJs). The pyridyl termini of the tetrazine 3 were placed to bind to the gold electrodes of the MCBJ and thus orient the diene in the x-direction of the AEF. The IEDDA with dihydrofuran 4 along TS-2 is followed by retro-DA to release N2 and ring-opening aromatization. All intermediates were characterized by their characteristic single-molecule currents, which allowed monitoring of different reaction kinetics by counts in the respective current histograms. Measured for 1 mM substrates in mesitylene/dichloromethane 4:1, AEFs catalyzed the final ring-opening aromatization with a linear dependence on applied voltage until saturation characterized by effective voltages around 20 mV and saturation around 200 mV. Consistent with theory, the IEDDA was not catalyzed by the AEFs in the x orientation. Two years later, in 2021, graphene–molecule–graphene single-molecule junctions (GMG-SMJs) were used to explore NEDDA-EFC with flexible substrates. The terminal amino groups of the rigid-rod scaffold of 5 were reacted with carboxylates of oxygen plasma etched graphene. This tethered the maleimide dienophile flexibly along the scaffold for NEDDA with furan 6 along TS-3. Single-molecule reactions with furan as vapor and in toluene/DMSO solution were characterized by stochastic switching between six conductive states in I-t curves. In the histograms, they were assigned to the single substrate 5, exo/endo reactant (charge-transfer) complexes with furan, and exo/endo products appearing at increasing temperature. Stochastic spikes observed only at high temperatures (>95 °C, furan vapor) and voltage were assigned to a zwitterionic Michael addition intermediate with a large macrodipole. Their reversible conversion into DA products, directly observed in single-molecule I-t curves, supported a mechanistic crossover that was predicted to occur at EEFs > 2.6 V/nm. Increasing DA products with decreasing polarity of toluene/DMSO mixtures at 300 mV was considered as support for operational EFC. Single-molecule I-t curves provided rates for reversible conversion of all intermediates and thus equilibria for the concerted DA mechanism. Lowered activation energies from single-molecule Arrhenius plots for forward and reverse cycloaddition with 300 mV compared to bulk solution supported the occurrence of DA-EFC with furan 6 in toluene at 25 °C, while equilibria depended on temperature and solvent polarity. Compared to single-molecule DA-EFC with oriented substrates 1 and 2, these results with flexible substrates 5 and 6 were particularly important in the context of the present study because they supported that self-orientation of freely tumbling substrates in an AEF is sufficient for EFC to occur. In other words, the results imply that fixed substrate orientation in oriented AEFs is not a prerequisite for EFC of Diels–Alder reactions. Originating from ultrafast reactions observed during electrospray mass spectrometry (ESI-MS), possible EFC in microdroplets is a topic that causes intense current controversy. As early as 2017, the Diels–Alder cyclization of 3,5-hexadienyl acrylate ester catalyzed by indium­(III) triflate was shown to work in bulk water but, according to MS detection, not in water microdroplets. In the same year, the same Zare group reported that, again according to in situ MS detection, the Sharpless [2σ + 2σ + 2π] cycloaddition between the strained quadricyclane and diethyl azodicarboxylate occurs in water microdroplets and attributed the observed rate enhancement to concentrating “on-water” effects originating from poor solubility. In 2024, the DA reaction between a furan derivative and ortho-diiodotetrafluorobenzene in water microdroplets, detected by MS in situ, has been explained with reactive oxygen species that produce ortho-diiodotetrafluorobenzene radical anions as “quasi-benzyne” dienophiles, that is by electron transfer chemistry rather than EFC. In the same year 2024, Diels–Alder chemistry in water and MeCN microdroplets has been explored with cyclopentadiene 7, produced in situ by retro-DA from dimer 8, and acrylonitrile 9 (Figure C). This is an attractive choice because competition with oxidation of substrate 7, substrate dimers 8, and product 10 into ketones 11–13 provides an internal control for the occurrence of radical redox chemistry instead of EFC. According to in situ MS detection, oxidation products were the main products under all of the conditions. They were proposed to originate from reactive oxygen species generated during vaporization, possibly involving EEFs. The ratio of DA against oxidation products decreased with increasing spray potential up to 5000 V, that is, increasing EEFs. Increasing DA product ratios with droplet flight time, nebulizing gas pressure, substrate concentrations, solvent apolarity, and apolar (but not polar) droplet size (spray solution flow rate) all supported that increasing effective concentrations by droplet evaporation and insolubility-related on-water confinement, rather than EEFs, catalyze DA reactions in microdroplets. These results were supported by computational simulations that highlighted concerns about ultrafast EEF fluctuations without orientation in the low picosecond time scale in water microdroplets. The EFC of Diels–Alder reactions under organic synthesis conditions has not been reported so far. This is not surprising because, as stated at the beginning, the incompatibility with practical organic synthesis is the grand central problem with catalysis with externally applied electric fields. Extrapolating from early results with other electrochemical constructs, − we and others have recently suggested that this challenge could be tackled with microfluidic capacitors. The key was the introduction of catalytic Helmholtz layers that translate intrinsically insufficient AEFs into effective electric fields (EEFs) that are strong enough to directly catalyze reactions. − In this study, Helmholtz layer catalysis in microfluidic capacitors is used to realize the seemingly impossible and catalyze Diels–Alder reactions with external electric fields under organic synthesis conditions. The functional chiral Helmholtz layers constructed for this purpose are the first example for engineered Helmholtz layers that function as (i) primary catalyst and (ii) only source of chirality in asymmetric EFC under practical conditions. The chemistry of anthrone 14 was selected to elaborate on DA-EFC under organic synthesis conditions not only because cycloaddition to electron-poor dienophiles 15–17 gives products 18–20 with interesting chirality (Figure ). Alternative Michael addition (MA) products 21–23 can serve as intrinsic chemoselectivity probes for mechanistic transitions from concerted DA toward the stepwise MA alternatives that were predicted to occur at high voltage in above single-molecule GMG-SMJ experiments with tethered dienophile 5. Moreover, the oxidation product 24 can serve as an intrinsic probe for transitions from EFC to the radical redox chemistry that dominates with 11–13 in water microdroplets (Figure vs C). , Current study on DA-EFC under organic synthesis conditions in microfluidic capacitors (charged wavy arrow) is realized with classical anthrone (14) chemistry because it offers products with interesting stereochemistry (18–20) and intrinsic probes for the occurrence of stepwise mechanisms (21–23) and radical redox chemistry (24). Introduced in 1989 by Koerner and Rickborn, anthrone DA chemistry has attracted much attention for asymmetric base catalysis, − first realized by Riant and Kagan. Michael addition products 21–23 have been observed along with DA product , using Lewis acid and hydrogen-bonding catalysts, , but apparently, not much Brønsted acid catalysis has been explored for anthrone DA chemistry. Results and Discussion Systems Design The microfluidic reactors introduced to achieve electric-field catalysis compatible under practical organic synthesis conditions are commercially available (Figure A,B; photos: Figures S1–S3). They have received much recent attention for a completely different purpose. − Namely, they are expected to make organic radical redox chemistry more controllable and safely scalable for use in industrial production. Intrigued by their properties, we thought that they could also be ideal to realize EFC in capacitor mode before the onset of electron transfer. (A) Side view of microfluidic capacitors and (B) top view of the FEP foil with the flow system inbetween the graphite plate electrodes. (C) Emerging principles for EFC with externally applied fields under organic synthesis conditions: Layer component generation (LCG, in situ or by addition) is followed by layer component separation (LCS) in response to AEFs (optionally supported by breakers B*), layer components translocation and Helmholtz layer assembly (HLA) to translate weak AEFs into strong EEFs. These EEFs are strong enough for Helmholtz layer catalysis (HLC) by stabilizing the transition states (TS) from substrate (S) to products (P), supported by structural information transfer, also from (*) optional ion-pair breakers (B*) and catalysts (C*). The selected microfluidic capacitors are composed of two 5 × 5 cm2 plate electrodes that are separated by a thin fluorinated ethylene propylene (FEP) foil with a flow channel. A thickness of d = 100–250 μm is commonly used; d < 100 μm is possible but increasingly problematic for routine work. This sandwich is then placed between metal plates and in the external housing that holds it all together. Graphite plates were the electrodes of choice because their surfaces show the least structural changes at high voltage. For maximal reproducibility, they were systematically polished on a dedicated machine before each experiment. The assembled capacitor was then connected to power supply or potentiostat to apply voltage and measure currents as a function of time. Finally, a syringe pump was connected to flow the reaction mixture through the capacitor at a controlled rate. Electric fields are the applied voltage divided by the distance between the electrodes. According to theoretical predictions, EFC requires electric fields >1 V/nm to significantly stabilize transition states. − ,− These predictions imply that microfluidic capacitors with plate electrodes spaced >100 μm apart will not be capable of delivering strong enough AEFs. This conclusion called for a shift of attention from the AEFs as such to their consequences. In capacitors, these consequences are electric double layers (EDLs). Ion pairs present in apolar solvents separate and move to the respective electrode where they assemble into compact yet dynamic Helmholtz or Stern layers of uniformly charged ions, followed by less ordered diffuse layers with some ions of the opposite charge present as well. − Polar solvents or substrates can also produce EDLs with Helmholtz layers composed of highly ordered, uniformly oriented neutral but polar molecules. With both ions or polar molecules, Helmholtz layers can be considered like supramolecular electrodes. The short distance to plate electrodes of opposite charge (1–5 nm) generates local effective electric fields (EEFs) that can exceed the AEFs by more than one million , and are with >1 V/nm strong enough to directly catalyze reactions. These considerations encouraged the rational design of Helmholtz layers components with functional motifs (Figure C). Their assembly into supramolecular 2D architectures would then position these functional motifs within the EEF they create to both enable EFC and contribute recognition motifs for information transfer to the product. Applying early lessons from cell penetration and sensing, − the concept of catalytic Helmholtz layers was explored first with bioinspired ion pairs composed of polyarginine and hydrophobic anions for EFC of proline-catalyzed aldol condensations. , Then it was validated and generalized with catalytic Helmholtz layers constructed from anionic Brønsted acids and ion-pair breakers. The results were compared to extract mechanistic principles for scalable EFC in catalytic Helmholtz layers (Figure C). Without going into details, the current understanding envisions a four-step mechanism composed of layer component generation (LCG), layer component separation (LCS), Helmholtz layer assembly (HLA), and Helmholtz layer catalysis (HLC). − LCG can include added ion pairs, , proton transfer, and even reactive intermediates (unpublished). LCS can be supported by optional ion-pair breakers, which then integrate into the Helmholtz layers, and, like optional organocatalysts, − contribute to structural information transfer to the product. HLA of the separated components produces the EEFs needed for EFC. HLC can thus be defined as EFC with EEFs and (optional) structural information transfer from the compact layers of (engineered) EDLs. Diels–Alder Reactions in Catalytic Helmholtz Layers Electric-field catalysis of Diels–Alder reactions in microfluidic capacitors was explored first with functional Helmholtz layers composed of the original, bioinspired polyarginine 25 (pR) ,− as cationic layer component LC and dodecyl sulfate (DS) 26 as anionic layer component LA at 1 mM concentrations in CHCl3/DMSO 3:1 (Figure A). They were loaded together with 50 mM diene 14, 75 mM dienophile 15, and mesitylene as an internal standard into a syringe pump and flowed through the capacitor with a constant rate at room temperature (Figures , A). Yields Y after one passage through the microfluidic capacitor were determined by 1H NMR spectroscopy or chiral HPLC and described as microfluidic yields Y m. Microfluidic yields Y m 0 report results without applied voltage. Under above conditions, Y m 0 = 0% was recorded for DA product 18; anthrone DA chemistry did not occur without electric fields (Figure B). (A) DA-EFC in catalytic Helmholtz layers: Flowing into the AEF of the microfluidic capacitor, layer component separation (LCS) prepares for Helmholtz layer assembly (HCA) of LC 25 and LA 26 at the plate electrode of opposite charge to translate weak AEFs into EEFs. These EEFs are strong enough to catalyze NEDDA reactions of anthrone 14 with dienophiles 15–17 by (a) supporting intermolecular charge transfer along the reaction axis, (b) polarizing substrate complexes to induce supportive dipoles, perhaps up to charge transfer complex formation and inversion to IEDDA mechanisms, and (c) increasing phenol acidity up to deprotonation as formal invisible base catalysts, as outlined in transition states TS-5 and TS-6. (B, C) YV curves (C) without and (B) with LC 25 and LA 26 (1 mM, 2 mol %): Dependence of the microfluidic yields Y m (black, left, 18: circles, 21: squares, 24: diamonds; *experimental replicates) and enantioselectivity (right, red, 18: triangles) on the applied voltage V for 14 (50 mM) and 15 (75 mM) in CHCl3/DMSO 3:1, flow rate 5 μL/min, 100 μm foil, graphite electrodes, room temperature, with curve fit to Hill equation and indication of Y m max, Y m 0, V c, V 50, n g, and ee max; X = NMe, NH, NPh. The application of voltage afforded DA product 18. Remarkably, Y m did not increase linearly. Reminiscent of voltage-gated ion channels, sigmoidal YV curves were obtained instead (Figure B). The maximum experimental yield, Y m max = 75%, was obtained at 1.5 V, and saturation could not be reached because DMSO as (co)­solvent is incompatible with high voltages. Saturation behavior in YV curves has been observed regularly and attributed to the assembly of the final catalytic Helmholtz layer whose activity cannot be further improved with increasing applied voltage. − Curve fit to the Hill equation gave an effective voltage V 50 = 0.9 V ((Y m max + Y m 0)/2), a critical voltage V c = 0.6 V (Y m 0 + 10%), and a gating (Hill) coefficient n g = 4. The removal of Helmholtz layer components LC 25 and LA 26 strongly reduced DA-EFC (Figure C). Most importantly, Y m max dropped from Y m max = 75% to Y m max = 32%. Interestingly, YV curves remained voltage-gated with n g = 4. The poor Y m max and a minor Y m 0 = 4% resulted in V 50 = V c = 0.7 V at Y m 50 = 14%. With and without LC 25 and LA 26, EFC in microfluidic capacitors gave the DA product 18 almost exclusively (Figure B,C, circles). The oxidation product 24 was negligible and independent of AEFs (Figure B,C, diamonds). Contrary to disfavored DA reactions in microdroplets (Figure C), this selectivity supported the idea that successful DA reactions in catalytic Helmholtz layers occur by EFC without involving radical redox processes. The validity of this conclusion was supported by chronoamperometry (see below). Significant amounts of MA product 21 were not obtained either (Figure B,C, squares). As described in the introduction, a transition from DA to MA products has been observed in single-molecule junctions at high temperature and high voltage (Figure B). Supported by high-level computational studies, this transition has been explained with the stabilization of the large macrodipole of the zwitterionic (rather than diradical) MA intermediate by strong electric fields. The absence of MA product 21 implied that in EEFs in catalytic Helmholtz layers, the formation of DA product 18 occurs by concerted cycloaddition and not by a stepwise mechanism. Mechanistic Considerations These results were consistent with significant DA-EFC in functional Helmholtz layers composed of LC 25 and LA 26 that translate weak AEFs into strong EEFs of >1 V/nm. For EFC, these strong EEFs could support intermolecular charge transfer along the reaction axis in transition states TS-5 and TS-6 as expected from theoretical studies (Figure Aa vs A). Gradually increasing EEFs − are expected to support diffusion of separate substrates or substrate complexes with strong macrodipoles across the diffuse into the compact catalytic Helmholtz layer. The dynamic, fluid 2D architectures also of this compact layer provide more than enough flexibility to adapt and integrate substrates spontaneously. Contributions from noncovalent interactions with Helmholtz layer components to support substrate integration and orientation are best documented by chirality transfer from chiral catalytic Helmholtz layers (Figure B, red triangles, see below for higher values). Self-orientation of freely diffusing substrate complexes in EEFs has been supported experimentally for TS-3 in single-molecule BJ studies and theoretically for TS-4 (Figures B,C). Computational predictions also support that EEFs could further stabilize TS-5 and TS-6 by polarizing the substrates, inducing supportive dipoles (Figure Ab, A). , Finally, EEFs could further catalyze the DA reaction by increasing phenol acidity up to deprotonation (Figure Ac). Here, EEFs would act as formal, invisible base catalysts, and base catalysis is well-established in anthrone DA chemistry. , Weak anthrone DA chemistry without functional Helmholtz layers implied primary contributions from EFC under organic synthesis conditions (Figure B vs C). The residual, equally voltage-gated DA-EFC without functional Helmholtz layers (Figure C) was likely to occur in Helmholtz layers formed by uniformly oriented polar components of the solvent mixtures, here DMSO. In principle, proton transfer between the two substrates in response to AEFs could further produce cationic and anionic Helmholtz layers at the respective plate electrodes with reactant complexes activated by protonation and deprotonation, respectively. Although conceivable, such an intersubstrate proton transfer seemed rather unlikely with this system. Similar V c values with and without dedicated Helmholtz components instead suggested that V c reports mostly on the AEF needed to direct neutral polar substrate complexes toward either cationic or anionic Helmholtz layers to access TS-5 and TS-6, respectively. The distinctly different Y m max values with (75%) and without (32%) functional Helmholtz components would then support that the catalytic activity of Helmholtz layers including LC 25 and LA 26 exceeds Helmholtz layers only from presumably neutral, uniformly oriented substrates 14 and 15 in oriented solvent decisively. This difference evinced the catalytic activity of the engineered Helmholtz layers as the origin of EFC of Diels–Alder reactions under organic synthesis conditions. Quantitative support for this interpretation was obtained by chronoamperometry (see the following). Quininium Cations As Helmholtz Components The dependence of DA-EFC on the nature of functional Helmholtz layers was explored first by replacing Helmholtz components LC 25 and LA 26 with N-benzylquininium chloride (QUIBEC, NBQ), , which offers the quininium cation 27 as LC and chloride as LA (Figure ). Dependence of DA-EFC on voltage (A, C–E) and concentration (B) of Helmholtz layer components (27), ion-pair breakers (28, A, D, gray), substrates (17, D, E), and solvents (CHCl3, C, E). YV curves for the reaction of diene 14 (50 mM) and dienophiles 15 (A–C) or 17 (D–E, 75 mM) in the presence of 27-Cl (A, C–E: 1 mM; B: 0–10 mM) and 28 (A, D, gray, 2.5 mM) in CHCl3/DMSO 3:1 (A, B, D) or CHCl3 (C, E), reporting Y m (left, 18/20: circles, 21/23: squares, 24: diamonds) and enantioselectivity (right, red, 18/20: triangles) as a function of the applied voltage V (A, C–E: varied; B: 0, black, or 1 V, dark red), flow rate 5 μL/min, 100 μm foil, graphite electrodes, room temperature, fit to Hill equation, affording V c, V 50, and n g, *excluded from curve fit. This Helmholtz layer substitution did not significantly disturb the DA-EFC between anthrone 14 and maleimide 15 (Figure A). The YV curve gave similar Y m max (80%, 75%), V c (0.4 V, 0.6 V), and n g (3, 4), suggesting that the catalytic activity of the final Helmholtz layers did not change significantly. An increase of Y m 0 from 0% to 13% implied weak contributions of the quininium cation to DA catalysis already without applied voltage. Preservation of the trace enantioselectivity supported, together with similar Y m max, that DA-EFC occurs preferably in the chiral cationic Helmholtz layer, the supramolecular anode, along TS-5 rather than TS-6 in the achiral, inorganic Helmholtz layers (Figure , see below). Catalyst Concentration Dependence The concentration dependence of catalytic Helmholtz layers was determined for DA-EFC with 50 mM diene 14 and 75 mM dienophile 15 at 1 V for the quininium 27 as LC and chloride as LA in CHCl3/DMSO 3:1 (Figure B). The Y m max of the reaction increased gradually to a maximum at around 1 mM, which was followed by a significant decrease at 10 mM. Hill analysis up to 3 mM afforded an EC50 ≈ 100 μM. Although dependent on conditions, particularly solubility in less polar solvent mixtures, Helmholtz components were used at the maximum at 1 mM. With the chiral Helmholtz layers serving as unique catalysts for DA-EFC, this calculated to 2 mol % and an EC50 = 0.2 mol % for formal catalyst loading, referring to monomeric LC catalysts 27 relative to the anthrone substrate 14. Saturation behavior in the concentration dependence of catalytic Helmholtz layers was related to saturation behavior in the YV curves. Both were expected to mark completion of the assembly of the final catalytic Helmholtz layer, whose activity cannot be further improved with increasing voltage − and even can decrease with too many Helmholtz components in the surrounding solution. Solvent Dependence Solvent polarity is important for EFC under organic synthesis conditions in engineered Helmholtz layers for three main reasons. First, the onset of electron-transfer chemistry with certain, mostly polar solvents can limit access to high voltage. This is the case for DMSO above 2 V, for instance, but not for CHCl3. , Second, in response to AEFs, polar solvents can orient uniformly, form EDLs, and generate EEFs that are strong enough for EFC. Intrinsic competition from polar solvents can complicate or even hinder the construction of catalytic Helmholtz layers with extrinsic components, which makes studies in apolar solvents preferable or at least less complicated. Third, the difficulty of ion-pair separation by AEFs for Helmholtz layer assembly increases with decreasing solvent polarity. With the current understanding of EFC, this should increase the critical voltage, that is, voltage gating, and the impact of ion-pair breakers. Contrary to the limitation of CHCl3/DMSO 3:1 to V < 2 V, YV curves in CHCl3 could thus be recorded up to V = 20 V without the emergence of oxidation products 24 (Figure C,E). The YV curves for substrates 14 and 15 with 2 mol % NBQ 27 in CHCl3 revealed preserved voltage gating (Figure C). With decreasing solvent polarity, product traces obtained without voltage decreased from Y m 0 = 13% to Y m 0 = 5%. This was as expected from the intrinsic reactivity in a stirred solution. Maximal microfluidic yields at high voltages increased slightly to a new record of Y m max = 87%. A significant increase of the critical voltage with decreasing solvent polarity from V c = 0.4 V to V c = 1.5 V confirmed , that V c in less polar solvents reports on increasingly demanding ion-pair separation by the AEFs. With decreasing solvent polarity, the enantioselectivity of DA-EFC of substrates 14 and 15 increased from ee max = 7% to ee max = 26% (Figure A,C, triangles).This result was important because it provided the first hints of asymmetric EFC with chiral information originating exclusively from the catalytic Helmholtz layers, rather than additional chiral organocatalysts such as prolines. This result thus supported the concept of asymmetric EFC by chirality transfer from chiral Helmholtz layers as realistic and worth further investigation. Substrate Dependence DA-EFC with anthrone 14 in Helmholtz layers assembled from quininium 27 was compatible with maleimide 16 in place of maleimide 15 (Figure ). Nearly identical YV curves were obtained in CHCl3/DMSO 3:1 (Y m max = 73%; Figures S49, A). Removal of the quininium Helmholtz layers decreased DA-EFC with 16 as significantly as removal of pR Helmholtz layers decreased DA-EFC with 15 (Y m max = 34%; Figures S51, C). The main difference was that the additional hydrogen-bond donor in 16 fully suppressed, rather than increased, the trace ee max = 7% with 15 (Figures S12, A). In clear contrast, DA-EFC in the same chiral quininium Helmholtz layers 27 n with anthrone 14 and phenyl maleimide 17 increased the stereoselectivity to ee max = 13% (Figure D). Compared with Y m max = 80% for 18, the yields of DA product 20 dropped to Y m max = 53% (Figure D). This Y m max = 53% was, however, impressive because yields without Helmholtz layers 27 n were Y m max = 13% with 17, even lower than Y m max = 32% with 15 (Figures S8 and S53). In the less polar CHCl3, DA-EFC with maleimides 16 and 17 in place of maleimide 15 gave products 19 and 20 with up to 31% ee (Figures E and S23). Redox probes 24 were not obtained, also at high voltages (Figure E, diamonds). The best enantioselectivity ee max = 31% for DA-EFC was observed with phenyl maleimide 17 in CHCl3 at 10 V (Figure E, triangles). The recorded enantioselectivity with a phenyl substituent in maleimide 17 was consistent with theoretical predictions from the literature. For an asymmetric DA-EFC with up to ee max = 31%, the chiral catalytic quininium Helmholtz layers are the only possible origin of enantioselectivity (Figure E). The dependence on flow rates was not remeasured because in previous studies with other systems, decreasing flow rates gave increasing yield, which was reasonable given the longer reaction times. − Since parameters other than reaction time can contribute to changes, flow rates should not be considered as a single point in reaction kinetics without due caution. Poor accessibility to the reaction kinetics remains a technical limitation of microfluidic capacitors. The impact of alternating current and electrode material on previous systems was weak − and thus not assessed in this study either. Chronoamperometry Potential-step chronoamperometry (PS-CA) is routinely used to quantify EEFs in EDLs of capacitors from current decay kinetics and to probe for the occurrence of radical redox processes with currents at steady state. ,− Applied to EFC under standard conditions, one of the two graphite electrodes was configured as working electrode, while the other served as both the counter and reference electrode. After the microfluidic reactor was filled with reaction solution containing quininium 27 as LC and chloride as LA together with substrates 14 and 15 in CHCl3/DMSO 3:1, PS-CA measurements were carried out while maintaining a continuous flow. A step potential was applied by imposing 1.0 V across the electrodes for 40 s to charge the EDL capacitor, followed by a step back to 0.0 V for 40 s to discharge (Figure Ab). The charging of the capacitor at 1.0 V produced a current that decreased rapidly, indicating the assembly of EDLs, including the catalytic Helmholtz layers (Figure Aa). The step change to 0.0 V caused a current to flow in the opposite direction to discharge the capacitor (applied first at the beginning in Figure A). The decay kinetics of this current report on EDL disassembly and should be similar to those of EDL assembly kinetics. This two-step procedure was repeated to probe the reproducibility. (A) Chronoamperogram with charge/discharge step potential changes of V = ±1.0 V in 40 s intervals (b) for 14 (50 mM), 15 (75 mM), and 27-Cl (both 1 mM) in CHCl3/DMSO 3:1, flow rate 5 μL/min (a). (B, C) Representative charge profiles at (B) early and (C) late time points after application of V = +1.0 V at t = 0 s, without (blue) and with 14 and 15 (red), 27-Cl (orange) and 14, 15 and 27-Cl (green) in CHCl3/DMSO 3:1. Very early data points (B) were fitted to monoexponential decay to calculate EEFs by applying the RC circuit model and Gauss law (SI, 3.6). Electrode area = 12 cm2. Errors represent standard errors of 3 charge kinetics analyses. The initial charging/discharging kinetics of the capacitor were then fitted to an equivalent resistor-capacitor (RC) circuit model to determine the capacitance (for charging: C = 0.72 ± 0.04 mF, Figure B, green; for detailed procedures, see SI, 3.6, Equations S2–S6, Figures S57–S60). − The EEF produced by the EDL was then estimated based on Gauss law, using mathematical average dielectric constants of the mixed solvent (Equation S7, Tables S27–S30). The EEF = 4.4 ± 0.1 V/nm obtained was consistent with the >1 V/nm needed to directly catalyze reactions. The absolute EEF values should however not be overinterpreted because of several fitting assumptions, such as unknown local dielectric constants (contributing to underestimates), − ignored distance dependence, multiexponential decay due to complex EDLs, and so on. Chronoamperograms were recorded for solutions containing anthrone 14 and maleimide 15 but not Helmholtz components 27-Cl (Figures B, red, and S58). Smaller initial current and slower current decay revealed that the assembly of neutral EDLs with substrates 14 and 15 occurs but is less favorable than the one formed by charged Helmholtz components 27-Cl (Figure A). This more reluctant assembly was consistent with neutral EDLs formed by oriented polar substrates 14 and 15 and substrate complexes. The EEF = 1.9 ± 0.2 V/nm from charge kinetics with substrate EDLs was much weaker than the EEF = 4.4 ± 0.1 V/nm obtained in the presence of the charged components 27-Cl. This trend coincided with better DA-EFC (Y m max = 80% vs Y m max = 32%, Figures A, C) and thus provided corroborative quantitative evidence for both validity and significance of the concept of Helmholtz-layer catalysis. The EEFs Helmholtz components 27-Cl produced without substrates 14 and 15 were calculated to EEF = 4.0 ± 0.1 V/nm. These charged EEFs exceeded the neutral EEF = 1.9 ± 0.2 V/nm, although the concentration of ionic components was more than 50 times smaller (Figure B, red vs orange). The shape of chronoamperograms recorded for 27-Cl did not depend strongly on the presence or the absence of substrates 14 and 15 (Figure B, orange vs green). These trends supported that substrates 14 and 15, used at higher concentrations, insert via the diffuse layers into the still fluid, dynamic Helmholtz layers assembled by LC 27 and proceed to cycloaddition as soon as their orientation aligns with the EEF produced by the catalytic Helmholtz layer. This alignment of reactant complexes along the EEF is supported intrinsically by their macrodipole and in addition by noncovalent interactions with catalytic Helmholtz layer, as evinced by chirality transfer. The EEF = 1.8 ± 0.1 V/nm measured for the CHCl3/DMSO 3:1 solvent should originate from 2D assembly of uniformly oriented DMSO solvent molecules (Figure B, blue). Then, the dielectric constant of DMSO would probably be more appropriate to use, yielding an even smaller EEF (0.59 ± 0.04 V/nm). After 50 s of EDL assembly at 1 V, very small microcurrents decreasing toward steady state were observed. After 40 min, which are usually allowed for equilibration before evaluating the reaction, decreasing presteady-state currents I ≈ 24 μA were obtained for architectures with substrates 14 and 15 with/without LC 27-Cl and I ≈ 7 μA for solvent with/without LC 27-Cl (Figure C). The slightly larger steady state currents I ≈ 24 μA with substrates 14 and 15 but without 27-Cl were consistent with the more than 50 times higher concentrations used as well as their integration into EDL architectures. These presteady-state microcurrents <25 μA thus reported final small adjustments of the EDL architectures. Their smallness excluded significant contributions from radical redox chemistry to the found DA-EFC (Equation S8). ,, Most importantly, presteady-state currents I ≈ 24 μA did not change in the presence of Helmholtz components 27-Cl (Figure C, green vs orange), while the microfluidic yield of DA products 18 almost tripled from 26% to 71% at 1 V (Figures A, C). Negligible contributions from faradaic processes were consistent with the absence of the oxidation product 24. This result was important because it provided support for the existence and significance of the DA-EFC under organic synthesis conditions. Ion-Pair Breakers Ion-pair breakers were introduced this year following the emergence of voltage-gated EFC. Ion-pair separation in response to AEFs in microfluidic capacitors was considered to account for the observed critical voltage needed to access EFC. For functional anions, neutral anion binders centered around Schreiner thiourea 28 − were thus considered to bind to these anions, separate them from their countercations, and integrate structural information into the anionic Helmholtz layers (Figure C). A significant decrease of critical voltages, an increase of yields, and changes in selectivity in the presence of ion-pair breakers provided experimental support for the concept. Bifunctional thiourea catalysis of anthrone DA chemistry was reported. However, the presence of the original ion-pair breaker 28 did not significantly change the YV curve for DA-EFC between anthrone 14 and maleimides 15 or 17 in Helmholtz layers made from quininium 27 and chloride in CHCl3/DMSO 3:1 (Figures A and D, black vs gray). Critical and effective voltages were affected but remained at comparably low values. This poor activity of ion-pair breaker 28 was meaningful because they were introduced for tighter ion pairs in less polar solvents characterized by V c = 5.0 V. The weaker ion pairs in CHCl3/DMSO 3:1 naturally gave smaller V c = 0.5 V. The poor activity of Schreiner thiourea 28 thus supported that for this DA-EFC in CHCl3/DMSO 3:1, ion-pair separation by AEFs is not the limiting process reported by the critical voltage V c. This finding provided corroborative support that in this system, the formation of oriented neutral reactant complexes in the catalytic Helmholtz layers determines the critical voltage V c. Helmholtz Component Screening Scope and limitations of functional Helmholtz layer architectures to catalyze DA reactions with externally applied electric fields were explored under standard conditions with anthrone 14 and maleimide 15 in CHCl3/DMSO 3:1 at a flow rate of 5 μL/min and graphite electrodes separated by 100 μm (Figure ). CHCl3/DMSO 3:1 was preferred over the more attractive, less polar CHCl3 to prevent that eventual solubility problems of the tested Helmholtz components would influence the results. Anionic Helmholtz layer components were varied first, while the original, bioinspired polyarginine 25 , was kept constant as cationic Helmholtz component (Figures B, A, Table ). Polyarginine with chloride counterions was combined with different anions as sodium salts, which meant that chloride and sodium ions were present in all measurements as additional ions at constant concentrations limited by their solubility. Helmholtz layer component screening for DA-EFC, with YV curve summary for (A) LA and (B) LC variations. YV curves for the reaction of diene 14 (50 mM) and dienophile 15 (75 mM) in the presence of components 25, 26, and 30–39 (1 mM, 2 mol %), reporting Y m (18; Y m max of 21 and 24 < 4%) as a function of the applied voltage V (CHCl3/DMSO 3:1, flow rate 5 μL/min, 100 μm foil, graphite electrodes, room temperature, fit to Hill equation), for (A) LC = 25 with LA = 26 (gray dotted line), 31 (blue dashed), 32 (green solid), 34 (magenta dash-dot), or 35 (purple dash-dot-dot), and (B) LA = 30 with LC = 25 (red solid), 36 (blue dashed), or 37 (orange dot) and LA = 39 with LC = 38 (cyan dash-dot-dot). 1Helmholtz Component Screening for DA-EFC entryLC LA Y m max (%) Y m 0 (%) V c (V) V 50 (V) n g ee max (%) 1 25 26 7500.60.94*72––3240.70.7403 27 Cl– 80130.40.73*74 25 29 72240.60.76*155 25 30 66260.70.93*156 25 31 68240.50.7377 25 32 76110.50.8408 25 33 6370.70.9709 25 34 57160.60.93010 25 35 6500.40.83011Na+ 29 61160.60.83012 36 30 22121.01.03013 37 30 2401.11.25014 38 39 4901.11.38*015 40 Br– 6190.60.93*11aResults of Hill analysis for DA-EFC of diene 14 (50 mM) and dienophile 15 (75 mM) to 18 in the presence of 25–27 and 29–40 (1 mM, 2 mol %) in CHCl3/DMSO 3:1, flow rate 5 μL/min, 100 μm foil, graphite electrodes, at room temperature.bCationic layer component.cAnionic layer component.dHighest observed Y m (%).e Y m (%) observed at 0 V.fThe critical voltage equals Y m = Y m 0 + 10%.gThe effective voltage is equal to Y m = (Y m max + Y m 0)/2.hGating coefficients, estimated by Hill fits; uncertainties ≤ 1 unless marked with *.i ee values, estimated by chiral HPLC analysis. Helmholtz Component Screening for DA-EFC Results of Hill analysis for DA-EFC of diene 14 (50 mM) and dienophile 15 (75 mM) to 18 in the presence of 25–27 and 29–40 (1 mM, 2 mol %) in CHCl3/DMSO 3:1, flow rate 5 μL/min, 100 μm foil, graphite electrodes, at room temperature. Cationic layer component. Anionic layer component. Highest observed Y m (%). Y m (%) observed at 0 V. The critical voltage equals Y m = Y m 0 + 10%. The effective voltage is equal to Y m = (Y m max + Y m 0)/2. Gating coefficients, estimated by Hill fits; uncertainties ≤ 1 unless marked with *. ee values, estimated by chiral HPLC analysis. Replacement of the original DS 26 by (S)-TRIP 29, a popular chiral binaphthylphosphodiester, − caused the appearance of weak activity Y m 0 = 24% without applied voltage (Table , entry 4, Figure S24). Present with Y m 0 = 16% also in the absence of pR 25, this observation was best explained with very weak base catalysis by (S)-TRIP 29 (entry 11). The YV curve with (S)-TRIP 29 and pR 25 showed a voltage-gated EFC with unchanged V c = 0.6 V. Compared with DS 26, lower V 50 = 0.7 V and higher n g = 6 suggested that (S)-TRIP 29 facilitated the assembly of catalytic Helmholtz layers. With Y m max = 72%, the catalytic activity of the final Helmholtz layer was only marginally weaker compared to that of DS 26 (Y m max = 75%, entry 1 vs 4). Replacement of an achiral DS-Helmholtz cathode 26 n with a chiral (S)-TRIP assembly 29 n increased the maximal enantioselectivity from ee max = 7% to ee max = 15%. These relatively small increases of Y m max and ee max with chiral Helmholtz cathodes supported that DA-EFC mainly occurs in the chiral pR-Helmholtz anodes 25 n along TS-5. A decrease to Y m max = 61% and ee max = 0% upon replacement of this chiral pR anode 25 n by inorganic sodium was consistent with this interpretation (entry 11). The trends observed with (S)-TRIP 29 were overall preserved with chiral phosphodiesters 30 and 31 as partners of pR 25, particularly with regard to the presence of weak enantioselectivity and weak base catalysis without voltage (Table , entries 5 and 6, Figures A, blue, S27, S29). This minor enantioselectivity fully disappeared, rather than increased, with more sophisticated chiral phosphodiesters 32–35 known or extrapolated (33, 34) from asymmetric catalysis − (entries 7–10, Figure A, green, magenta, purple, Schemes S5–S9). Moreover, yields for EFC at high voltage and base catalysis without voltage decreased down to Y m max = 63% and Y m 0 = 0%, respectively (entries 8, 10). This poor performance of more sophisticated chiral phosphodiesters confirmed that the emerging principles of scalable EFC at high voltage differ and can be almost contrary to established synthetic methodology. The dependence on their nature demonstrated that anionic Helmholtz components contribute to DA-EFC at constant cationic Helmholtz components. However, the most important contributions to the yield and enantioselectivity of DA-EFC seemed to originate from the constant cationic Helmholtz component 25. More sophisticated anionic Helmholtz components failed to improve much from DS 26 and at worst disturbed the catalytic Helmholtz layers. The conclusion that TS-5 in catalytic Helmholtz layers accounts for most of DA-EFC was already supported by preserved top performance upon substitution of pR 25 with the quininium 27 and the poor activity of ion-pair breakers 28, integrating into the anionic rather than the cationic Helmholtz layer (Figures A, A,D). This interpretation was supported by more important changes in response to substitution of the cationic rather than the anionic Helmholtz layer component. Arginine monomers 36 in place of pR 25 were incapable of significant EFC, although they were teamed up with phosphodiesters 30 that were impressive with 25 (Table , entries 5 vs 12, Figure B, blue). Y m max = 22% was below Y m max = 32% without charged catalytic Helmholtz layers. The presence of a carboxylic acid in the Helmholtz component 36 might contribute to this inactivity. However, also the carboxylate-free guanidinium variant 37, together with the same phosphodiester 30, performed poorly (Y m max = 24%; entry 13 Figure B, orange). The more hydrophobic arginine monomer 38 together with the TFA anion 39 showed an unusual YV curve with strong voltage gating (entry 14, Figure B, cyan). V c = 1.1 V and n g = 8 characterized a sudden turn-on of DA-EFC from Y m 0 = 0% to Y m max = 49% around V 50 = 1.3 V (Y m max = 75% at 2 V). This interesting cooperativity could originate from the assembly into cationic Helmholtz layers reminiscent of those of biomimetic lipid monolayers. The Maruoka phase-transfer catalyst 40 having an ammonium cation next to a chiral binaphthyl as in the phosphodiester series provided corroborative support that DA-EFC occurs preferably in the cationic Helmholtz layer along TS-5 (Figure A). High Y m max = 61% against nearly negligible Y m 0 = 9% coincided with the reemergence of stereoselectivity at ee max = 11% (Table , entry 11, Figure S47). This weakly asymmetric DA-EFC was in the range of the best chiral cations pR 25 (ee max = 15%, entries 4 and 5) and quininium 27 (ee max = 7%, entry 3). However, while asymmetric DA-EFC with quininium Helmholtz layers 27n maximized to ee max = 26–31% in less polar CHCl3 at high voltage (Figure E), a similar jump was not observed with Maruoka Helmholtz layers 40n . Enantioselectivity remained at ee max = 15% in CHCl3 (Y m max = 90%, Figure S55). The strong responsiveness to variation of cationic Helmholtz layer components, exceeding the functional impact of anionic components by far, supported that DA-EFC occurs preferably in cationic Helmholtz layers along TS-5 (Figure ). This interpretation was interesting because EEFs acting like invisible base catalysts to activate the DA reaction by phenol deprotonation should perform better in anionic rather than cationic Helmholtz layers, that is, TS-6, contributing another anionic layer component (Figure c). DA-EFC occurring in cationic Helmholtz layers along TS-5 thus supported that the EEFs catalyzed the cycloaddition directly by enabling the charge transfer along the reaction axis in reactant complexes, as described by theory. − ,− ,− ,− This interpretation implied that, also supported by theory, the oriented EEFs produced by the catalytic Helmholtz layers will orient integrated reactant complexes according to their dipole and thus install the directionality needed for EFC (Figures and A). Conclusions The Diels–Alder reaction has emerged as the primary reaction in electric field catalysis. − ,− ,− ,− Despite this central importance, electric-field catalysis (EFC) of Diels–Alder (DA) reactions has not been possible so far under practical organic synthesis conditions. In this study, this seemingly impossible is achieved. Within microfluidic capacitors, functional Helmholtz layers are installed to translate the applied electric fields (AEFs) into effective electric fields (EEFs) that are strong enough to directly catalyze the reactions. In these EEFs, the yield of DA products from maleimides and the classical anthrones increases nonlinearly with increasing applied voltage, i.e., shows voltage-gated EFC. Chronoamperometry measurements demonstrate the formation of EEFs in the catalytic Helmholtz layers. With up to 4.4 V/nm, they are strong enough to directly catalyze the reactions. Negligible steady-state currents exclude significant contributions from Faradaic processes under these conditions. These chronoamperometry data are valuable because they quantitatively demonstrate the occurrence and relevance of EFC under scalable bulk conditions in rationally designed functional Helmholtz layers. We define Helmholtz-layer catalysis (HLC) as EFC with EEFs and (optional) structural information transfer from the compact layers of (engineered) electric double layers (EDLs). Beyond realizing scalable DA-EFC, this study marks the first time that HLC is achieved without the presence of additional small-molecule catalysts that are first activated by EEFs to then accelerate the reaction of interest. In this study, the functional Helmholtz layer is the only catalyst. Enantioselectivity up to 31% ee, which originates exclusively from these chiral catalytic Helmholtz layers, is equally unprecedented. Together with the strong dependence on layer components, this unprecedented chirality transfer from chiral catalytic Helmholtz to the transition state of the reaction highlights how noncovalent interactions from engineered Helmholtz layers can contribute to catalysis beyond their EEFs. The understanding of the assembly of catalytic Helmholtz layers on the molecular level thus emerges as a primary topic for supramolecular chemistry experts to contribute to the development of a conceptual framework for an organic synthesis with externally applied electric fields under practical synthesis conditions. The promises, from basic organic chemistry to sustainable industrial production and the understanding of the origin of life, are tantalizing. Supplementary Material