Search

The storage of solar energy in chemical bonds will depend on pH-universal catalysts that are not only impervious to acid, but actually thrive in it. Whereas other homogeneous water oxidation catalysts are less active in acid, we report a catalyst that maintained high electrocatalytic turnover frequency at pH values as low as 1.1 and 0.43 (kcat =1501±608 s-1 and 831±254 s-1 , respectively). Moreover, current densities, related to catalytic reaction rates, ranged from 15 to 50 mA cm-2 mM-1 comparable to those reported for state-of-the-art heterogeneous catalysts and 30 to 100 times greater than those measured for two prominent literature homogeneous catalysts at pH 1.1 and 0.43. The catalyst also exhibited excellent durability when a chemical oxidant was used (CeIV , 7400 turnovers, TOF 0.88 s-1 ). Preliminary computational studies suggest that the unusual active-site sulfonate group acts a proton relay even in strong acid, as intended.

A new strategy to create highly redox-responsive H-bond dimers based on proton-coupled electron transfer is proposed that capitalizes on the importance of secondary H-bonds in determining overall binding strength in H-bond dimers. Electron transfer induced proton transfer across a H-bond can be used to significantly strengthen the overall binding by both creating strong ionic H-bonds and changing the secondary H-bonds from unfavorable to favorable. The viability and potency of this approach are demonstrated with an electroactive DAD (A = H-acceptor, D = H-donor) array, H(MQ+)H, paired with an electroinactive ADA array, O(NH)O. NMR titration of H(MQ+)H with O(NH)O in 0.1 M NBu4PF6/CD2Cl2 gives a Kassoc of 500 M-1, typical of DAD-ADA dimers. However, upon two-electron reduction in 0.1 M NBu4PF6/CH2Cl2, cyclic voltammetry studies indicate a 1.8 × 105 increase in binding strength, corresponding to a very large Kassoc of 9 × 107 M-1. The latter value is typical of DDD-AAA H-bond dimers, consistent with proton transfer across the central H-bond upon reduction.

H-bonds can exert a substantial impact on the course of organic electrode reactions due to their ability to stabilize charged intermediates and products formed during these reactions, as well as facilitate proton-coupled electron transfer (PCET) reactions. This has fundamental implications for the mechanism of organic electrode reactions, but also practical impact in supramolecular chemistry and potentially synthetic electrochemistry. My group's main focus has been on the supramolecular applications, using electron transfer to alter the strength of H-bonds to create highly redox-responsive H-bond dimers. Initially we sought to avoid proton transfer because we feared that would lead to irreversible electrochemistry. However, inevitably proton transfer did show up, but, to our surprise, did not lead to irreversible electrochemistry. To explain this, we developed a new mechanism, the "wedge scheme", that shows how H-bonding can facilitate reversible electron and proton transfer. This insight recently led us to a new PCET-based design strategy for the creation of our most highly redox-responsive H-bond dimers yet.

We report studies on the photocatalytic formation of C–S bonds to form benzothiazoles via an intramolecular cyclization and sulfenylated indoles via an intermolecular reaction. Cyclic voltammetry (CV) and density functional theory studies suggest that benzothiazole formation proceeds via a mechanism that involves an electrophilic sulfur radical, while the indole sulfenylation likely proceeds via a nucleophilic sulfur radical adding into a radical cationic indole. These conditions were successfully extended to several thiobenzamides and indole substrates.

Giardiasis and other protozoan infections are major worldwide causes of morbidity and mortality, yet development of new antimicrobial agents with improved efficacy and ability to override increasingly common drug resistance remains a major challenge. Antimicrobial drug development typically proceeds by broad functional screens of large chemical libraries or hypothesis-driven exploration of single microbial targets, but both strategies have challenges that have limited the introduction of new antimicrobials. Here, we describe an alternative drug development strategy that identifies a sufficient but manageable number of promising targets, while reducing the risk of pursuing targets of unproven value. The strategy is based on defining and exploiting the incompletely understood adduction targets of 5-nitroimidazoles, which are proven antimicrobials against a wide range of anaerobic protozoan and bacterial pathogens. Comprehensive adductome analysis by modified click chemistry and multi-dimensional proteomics were applied to the model pathogen Giardia lamblia to identify dozens of adducted protein targets common to both 5’-nitroimidazole-sensitive and -resistant cells. The list was highly enriched for known targets in G. lamblia, including arginine deiminase, α-tubulin, carbamate kinase, and heat shock protein 90, demonstrating the utility of the approach. Importantly, over twenty potential novel drug targets were identified. Inhibitors of two representative new targets, NADP-specific glutamate dehydrogenase and peroxiredoxin, were found to have significant antigiardial activity. Furthermore, all the identified targets remained available in resistant cells, since giardicidal activity of the respective inhibitors was not impacted by resistance to 5’-nitroimidazoles. These results demonstrate that the combined use of click chemistry and proteomics has the potential to reveal alternative drug targets for overcoming antimicrobial drug resistance in protozoan parasites., Author summary Giardia lamblia and other protozoan parasites annually infect hundreds of millions of individuals worldwide and are major causes of clinically important diseases. Vaccines against giardiasis are not available, making treatment with antimicrobial drugs critical for the management of the infection. Resistance to all major drugs against G. lamblia has been reported and threatens their future utility, yet development of new and improved drugs remains a major scientific and economic challenge. Here we report a strategy for identifying new drug targets by exploring the targets of a successful class of existing antimicrobial drugs with multiple targets. Comprehensive identification of these targets revealed possibilities for alternative drugs that can overcome existing forms of resistance. This strategy has the potential to accelerate identification of promising novel drug targets as key rate-limiting step in the development of alternative antimicrobial drugs.

The electrostatic nature of the hydrogen bond makes it straightforward, in principle, to electrochemically perturb the strength of hydrogen bonds by using oxidation to create stronger H-donors or reduction to create stronger H-acceptors. However, oxidation or reduction can also lead to proton transfer and subsequent electron transfer, which can nullify the effect of the initial electron transfer. Illustrative examples of these complications based on a series of investigations of electroactive, urea-containing 2 H-bond dimers are described. It is also shown that proton transfer accompanying electron transfer can be used to advantage with two examples of oxidative dissociation of 4 H-bond dimers that are of interest for the construction of supramolecular polymers and gels. In both cases proton transfer is key to achieving the desired outcome. Moving forward, it is argued that proton transfer should always be considered as a possible useful component in the design of electrochemically controllable H-bonding systems.

Recently we have reported a new strategy to create highly redox-responsive H-bond dimers based on proton-coupled electron transfer.1 The strategy capitalizes on the importance of secondary H-bonds in determining overall binding strength in H-bond dimers. Specifically, this study shows that electron-transfer induced proton transfer across a H-bond can be used to significantly strengthen the overall binding by both creating strong ionic H-bonds and by changing the secondary H-bonds from unfavorable to favorable. The power of this approach is demonstrated with the electroactive DAD (A = H-acceptor, D = H-donor) array 1 paired with the electroinactive ADA array 2. NMR titration in 0.1 M NBu4PF6/CD2Cl2 gives a Kassoc of 500 M− 1 in the starting oxidized state, a value typical of DAD-ADA dimers. However, upon two electron reduction in 0.1 M NBu4PF6/CH2Cl2, CV studies indicate a 1.8×105 increase in binding strength, corresponding to a very large Kassoc of 9×107 M− 1. The latter value is typical of DDD-AAA H-bond dimers, consistent with proton transfer across the central H-bond upon reduction. In this presentation, the results of a full investigation of this system will be reported, including UV-vis spectroelectrochemical studies that support proton transfer and NMR studies of possible 1-1 and 2-2 homodimerization. Furthermore, theoretical CV’s that include, if needed, homodimerization will be fit to the experimental CV’s to provide a more exact measure of the binding enhancement in this system. 1H. Choi, K. Baek, S. Toenjes, J. L. Gustafson, and D. K. Smith, J. Am. Chem. Soc. 2020, 142, 17271-17276. DOI: 10.1021/jacs.0c07841 Figure 1

Hantzsch’s ester or amide is commonly used as a stoichiometric hydride transfer reagent in organic synthesis. The goal of this research is to see if Hantzsch’s ester or amide can be converted into an electrochemically-reversible hydride transfer mediator. The first step in achieving this is to realize the reversible 2 electron, 1 proton reduction of the corresponding pyridinium, 1, as shown below. Normally reduction of pyridiniums such as 1 are chemically irreversible due to the dimerization of the initially formed uncharged radical. We believe the key to avoid this pathway is to add a H-donor that can H-bond to the carbonyl O’s in such a way that an acidic H is properly positioned to H-bond to the radical, facilitating a second electron-transfer, proton-transfer step to generate the hydride donor. This presentation will describe the results of our initial efforts to achieve reversibility in this system. Figure 1

Two novel ruthenium complexes, [Ru(2,2′-bipyridine-6,6′-diphosphonato)(pic)2] (2) (pic=4-picoline) and [Ru(6,6′-diisopropyl-2,2′-bipyridine-6,6′-diphosphonato)(pic)2] (3) bearing phosphonate groups have been synthesized and characterized by NMR spectroscopy, elemental analyses, X-ray crystallography, cyclic voltammetry, and UV/Vis spectroscopy. Both complexes show catalytic water oxidation activity by electrochemistry. At pH 7, the RuII/III redox couple of 2 is observed at a lower potential than that of 3, yet significantly, 3 oxidizes water at a lower onset potential. At pH 1 however, 2 and 3 have comparable catalytic reactivity using sacrificial oxidant CeIV. We propose that water oxidation activities of 2 and 3 are influenced by overall charges. For example, in oxidizing from RuII to RuIII at pH 7, 2 acquires a −1 overall charge whereas 3 acquires a +1 charge. Charge-dictated electrostatic effects may govern binding of a water molecule to the metal site.

Supramolecular polymers are constructed along weaker, non-covalent interactions such as hydrogen-bonding. Compared to conventional polymers where monomer units react to form covalent bonds, non-covalent interactions of monomer units grant these supramolecular polymers dynamic behavior. One example of these systems that utilizes non-covalent interactions to construct supramolecular polymers are the well-investigated ureidopyrimidinones, UPy. UPy’s dimerize along their 4 strong directional hydrogen bonds in non-polar organic solvents. UPy’s have been covalently linked at the ends of smaller monomer chains so as to function as chain-linkers in the construction of larger polymer structures under favorable dimerization conditions. One interest in the use of these systems in supramolecular applications is their inherent ability of self-healing. Applying heat or mechanical stress to these systems will prevent polymerization at the H-bonds. However, upon relief of these external stimuli, H-bonding resumes. In this study, the use of electron transfer as an external stimulus to perturb dimerization and affect UPy chain polymerization will be explored. In order to exemplify the aims of this study, previously studied electroactive UPy’s will be investigated as redox-responsive chain terminators with regards to linear polymer chains that are functionalized with UPy units at their terminal ends, allowing for long-chain polymerization indicated by an increase in solution viscosity. In this study, a decrease in solution viscosity should be observed once these long chain bi-functionalized UPy polymers are in the presence of redox-responsive mono-functional UPy. Upon electrochemical oxidation / reduction of the mono-functional electroactive UPy units, an increase in viscosity is expected. The weakened H-bonding of the mono-functional UPy units will prevent chain termination and long-chain polymerization between bi-functional long chain UPy’s is expected to resume.

Unlike conventional polymers, in which monomers are reacted to form strong covalent connections, supramolecular polymers are formed by linking together monomers via weaker, non-covalent interactions such as H-bonds, ion-dipole interactions and so on. One very well-studied class of supramolecular polymers are based on the ureidopyrimidones, UPy’s, which dimerize via 4 strong, linear H-bonds in non-competitive solvents. Monomers for supramolecular polymerization can be made by linking together 2 or more UPy units with short covalent chains. Under dimerization conditions, the UPy’s link up to form longer polymeric chains. One of the useful attributes of such polymers is their self-healing ability. Application of heat or mechanical stress will readily break the monomers apart at the H-bonds, but, upon cooling or relief of stress, the bonds reform. The goal of this research is to see if electron transfer can be used as an alternative, more selective stimulus to break apart the UPy chains. As a first step in this direction, the use of our previously studied electroactive UPy’s, 1a and 1b, are being investigated as redox-active chain terminators. For these studies the known UPy monomer, 2, was synthesized. Viscosity measurements of solutions of 2 in 0.1 M NBu4PF6/CH2Cl2 give a critical polymerization concentration of 20 mM. Based on this, 60 mM 2 was chosen as the working concentration for initial experiments giving a solution with a specific viscosity of 3.0 mPa∙s. Upon addition of 2 mM of the ferrocene containing UPy 1a, the viscosity dropped to 1.1 mPa∙s, indicating that 1a is an effective chain terminator, decreasing the lengths of the chains which results in a decrease in the viscosity of the solution The ferrocenes were then completely oxidized to the ferrocenium form in a two compartment electrochemical cell. This caused the viscosity of the solution to increase from 1.0 to 1.7 mPa∙s, suggesting decreased effectiveness of the ferrocenium form as a chain terminator. This is consistent with poorer binding of the ferrocenium form to the bis-UPy. Importantly, reduction back to the ferrocene led to a decrease in the viscosity back to close to where it was initially. In this presentation, the results of further studies on the ability of both 1a and 1b to act as redox-responsive chain terminators for UPy-based supramolecular polymers will be reported. Figure 1

Ureidopyrimidones (UPy’s) are well-known to dimerize via 4 strong H-bonds in noncompetitive solvents. Although UPy dimers have been widely studied, there are few examples of UPy’s containing redox-active groups, and even fewer in which oxidation/reduction has been shown to affect dimerization. In this study, a thorough electrochemical investigation of a UPy with a ferrocene attached to the 6-position of the pyrimidone ring, UPy(Fc), demonstrates that strong and reversible redox-control of dimerization is possible in the UPy system. 1H NMR shows that the reduced UPy(Fc) is fully dimerized under electrochemical conditions in CH2Cl2. Cyclic voltammetry (CV) shows that oxidation of the ferrocene to the ferrocenium converts the dimer to another species with a less positive E1/2, and that reduction of this species reforms the dimer in a chemically reversible fashion. Analysis of the scan rate and concentration dependence of the CV’s along with the relative diffusion coefficient measurements strongly suggest tha...

Cyclic voltammetric and spectroelectrochemical evidence is presented showing that the H-bonded dimer formed from a ureidopyrimidone derivative containing a phenylenediamine redox couple can be reversibly broken apart at mM concentrations in CH2Cl2 by an electrochemically induced proton-coupled electron transfer reaction.

It is well-established that very weak acids such as water and alcohols strongly H-bond to quinone dianions, Q2–, in aprotic solvents. This results in thermodynamic stabilization of Q2– and a shift of the formal potential of the Q–/2– couple to less negative values. This study shows that the strong H-bonding of water also results in a type of kinetic stabilization of Q2–. CVs of the naturally occurring naphthoquinone Vitamin K1 in very dry 0.1 M NBu4PF6/CH2Cl2 show a reversible Q0/– wave but a chemically irreversible Q–/2– wave. Similar behavior is seen with anthraquinone and duroquinone. Evidence suggests that this is due to nucleophilic attack of Q2– on CH2Cl2 to give ether products. Addition of water results not only in the expected positive shift in potential of the second wave but also in an increase in the chemical reversibility. This indicates that the H-bonding of water to Q2– blocks the irreversible reaction with CH2Cl2 by significantly decreasing the rate of that reaction. Further experiments sho...

Recently we introduced the “wedge scheme” as a convenient means to include H-bonded intermediates in an overall proton-coupled electron transfer (PCET) mechanism by showing that it can nicely explain the unusual solvent- and concentration-dependent voltammetry of a phenylenediamine-based urea, U(H)H. This compound undergoes an apparent 1 e– reversible oxidation in CH2Cl2 that actually corresponds to 2 e– oxidation of half of the ureas to the quinoidal cation accompanied by transfer of a H+ and deactivation of the other half of the ureas. The reversibility of the process is due to the second electron transfer and proton transfer proceeding through a H-bond complex between the oxidized urea and the dimethylamino group on another urea. In this study the effect of adding 1,8-naphthyridine, naph, which H-bonds to the U(H)H in its reduced state, is examined using cyclic voltammetry. Addition of naph causes both an increase in the current of the reversible U(H)H cyclic voltammetric wave as well as the appearance...

Unlike conventional polymers, in which monomers are reacted to form strong covalent connections, supramolecular polymers are formed by linking monomers via weaker, non-covalent interactions such as H-bonds, ion-dipole interactions and so on. These imbue supramolecular polymers with unique properties since the linkages are readily broken, but also easily reformed as environmental conditions change. One particularly well-studied class of supramolecular polymers are based on the ureidopyrimidones, UPy’s, which dimerize via 4 strong, linear H-bonds in non-competitive solvents. Monomers for supramolecular polymerization can be made by linking together 2 or more UPy units with short covalent chains. Under dimerization conditions, the UPy’s link up to form longer polymeric chains. One of the useful attributes of such polymers is their self-healing ability. Application of heat or mechanical stress will readily break the monomers apart at the H-bonds, but, upon cooling or relief of stress, the bonds reform. The goal of this research is to see if electron transfer can be used as an alternative, more selective stimulus to break apart the UPy chains. As a first step in this direction, we have synthesized a known UPy monomer, 1, which is expected to polymerize at concentrations above ~20 mM in CH2Cl2. This can be verified by measuring the change in specific viscosity as a function of the concentration of 1. By adding simple UPy’s as chain terminators the degree of polymerization should decrease, with a corresponding decrease in the viscosity of solution. For this work, we plan to add our previously studied electroactive UPy’s, 2a and 2b. We have previously shown oxidation breaks apart the 2a and 2b dimers at mM concentrations in CH2Cl2. Thus, upon electrochemical oxidation, 2a or 2b will no longer be able to act as chain terminators and the linked UPy 1 should reform polymer chains. This can be detected by an increase in the viscosity of solution upon oxidation. Figure 1

Proton-coupled electron transfer reactions (PCET) play a significant role in many chemical and biological mechanisms. However, although these reactions have been known and studied for over a century, it is only within the last few decades that the importance of H-bonding intermediates in these reactions has come to be appreciated. In previous studies, we examined how explicit consideration of electron transfer and proton transfer within a H-bond intermediate (the “wedge” scheme) could explain the voltammetry of oxidation reactions involving phenylenediamines and bases in acetonitrile. The focus of this new study is to see if the wedge scheme can also explain the chemically reversible PCET reaction due to addition of acidic alcohols to the N-methyl-4,4’-bipyridinium redox couple (commonly called “monoquat” or MQ) in acetonitrile. MQ, which starts out as a +1 cation, undergoes two reversible reductions in acetonitrile, first to an uncharged radical and then to a quinoidal anion. For this study, different types of acidic alcohols, such as trifluoromethanol, methanol, and dichloroethanol, were added at different concentrations. The cyclic voltammetry shows a positive shift in the E1/2 of the second reduction after adding the acidic alcohol as a guest. This is likely due to protonation of the monoquat anion by the alcohol. The overall reaction was expected to be MQ + ROH + e- = MQH + RO-, which would show a 60 mV per log[ROH] change in E1/2. However, analysis of the collected data contradicts this hypothesis by giving a 180 mV per log[ROH] E1/2 shift, indicating that three alcohols are involved in the overall reaction. Since it is highly unlikely that the MQ anion would be triply protonated, our current hypothesis is that the product conjugate base of the alcohol forms a strongly H-bonded complex with two other alcohols, making the overall reaction, MQ + 3ROH + e- = MQH + (RO-)(ROH)2. Results of on-going research to either disprove or provide further support for this hypothesis will be reported.

An important goal in supramolecular chemistry is the development of stimuli-responsive systems in which the strength of the intermolecular interactions can be altered by external signals such as changes in light, temperature, pH or voltage of an electrode. Our group is exploring the latter possibility, primarily by trying to develop systems in which electron transfer perturbs the strength of H-bonding interactions between molecules. The underlying principle is straightforward: an oxidation that decreases the negative charge on a H-acceptor (A) or a reduction that decreases the positive charge on a H-donor (D) will weaken a H-bond. Alternatively, reduction that increases the negative charge on a H-acceptor or oxidation that increases the positive charge on a H-donor will increase the strength of a H-bond. However, in the latter case, it is possible that oxidation or reduction could also lead to full proton transfer. If this occurs across the H-bond, the primary H-bonds will remain, but the secondary H-bonds will change. This can lead to an increase in unfavorable secondary interactions, which would counteract the effect of the initial proton transfer. However, with proper design, proton transfer could lead to an increase in favorable secondary interactions, which would enhance the effect of initial transfer. The goal of this project is to do the latter. For this work, the 3 H-bond DAD array, 1, that contains a N-methyl-4,4'-bipyridinium or “monoquat” redox couple (see figure) has been synthesized. Compound 1 forms a three H-bond dimer with the non-electroactive ADA array, 2, in CH2Cl2. Typically, DAD-ADA arrays such as this have relatively weak association constants of ~102 M−1 in non-competitive solvents such as CH2Cl2 due to the three, favorable primary H-bonds (shown as solid, double-headed, green arrows in the figure) being counterbalanced by four unfavorable secondary interactions (shown as dashed, double-headed, red arrows). Preliminary NMR studies suggest stronger binding in this case (>103 M−1), probably due to the greater acidity of the amide NH’s in 1 because of the strongly electron-withdrawing nature of the pyridinium. In the absence of 2, initial cyclic voltammetry studies of 1 display the expected two, sequential, 1 e− reduction waves of the monoquat redox couple in CH2Cl2, corresponding to reduction of 1 to the radical and then the quinoidal anion. Addition of 2 results in no change in the E1/2 of the first reduction, but the second reduction shifts ~0.24 V positive. Further addition of 2 causes no additional change in the E1/2, consistent with a 1:1 complex. The 0.24 V positive shifts indicate a ~104 increase in binding strength upon overall 2 e− reduction of 1. Combined with the initial association constant of >103 M−1, this indicates an association constant of >107 M−1 in the fully reduced state. We believe the most likely explanation of such strong H-bonding in a 3 H-bond array is that the second reduction induces proton transfer across the central H-bond in the complex, thus converting the DAD-ADA array to a DDD-AAA array. The latter is expected to have significantly stronger H-bonding because all of the secondary interactions, in addition to the primary interactions, are favorable. If confirmed by further studies, the association constant in the reduced form would represent, to the best of our knowledge, the largest achieved to date in a redox-dependent H-bonding system, thereby demonstrating the utility of incorporating proton transfer to amplify the effect of electron transfer in these systems. Figure 1

Ureidopyrimidones (UPy’s) are well-known to dimerize in weakly polar solvents such as CH2Cl2, via the formation of four strong, linear H-bonds. This, coupled with their relative ease of synthesis, has led to their use as a linker in supramolecular polymers and gels. The attraction of such materials is their inherent self-healing properties, whereby a defect can be repaired by using heat or mechanical stress to reversibly break the polymer at the H-bond sites. This increases the fluidity of the material, allowing the defect to fill in. Upon cooling or relief of stress, the H-bonds reform and the defect is repaired. However, while the use of heat or mechanical stress as stimuli are clearly useful, neither is inherently very selective. On-going research in our lab is focused on increasing the versatility of the UPy system by creating electroactive UPy’s in which dimerization strength can be controlled in a more selective manner through oxidation and reduction. Previously, our group has shown that dimers based on the ferrocene-UPy derivative, 1, break apart upon oxidation of the ferrocene to the ferrocenium form at mM concentrations in CH2Cl2. This could be due both to the creation of electrostatic repulsion and a decrease in H-bond strength due to the reduced H-donor ability of the O and N on the pyrimidone side. In this project, another Fc-UPy has been prepared, 2, in which the ferrocene is attached to the urea side of the molecule. In contrast to 1, oxidation of 2 shows a single reversible ferrocene CV wave in CH2Cl2 at mM concentrations. Under these conditions, 1H NMR indicates that 2 is fully dimerized, thus, it appears that, unlike with 1, oxidation has no significant effect on dimerization of 2. In both cases, oxidation would increase electrostatic repulsion. However, unlike in 1, oxidation of 2 should actually increase H-bond strength by making the urea NH a stronger H-donor. Therefore, it appears that the effect of oxidation on the H-donating or accepting ability is a more crucial factor for dimerization control than electrostatics in these systems. Figure 1

The formation of supramolecular polymers is possible through directional and favorable intermolecular interactions that, unlike conventional covalent bonds, are weaker and purely electrostatic. The electrostatic nature of these interactions grants their respective systems reversibility in binding interactions, allowing molecular systems to achieve dynamic behavior. An important goal in supramolecular chemistry is the development of stimuli-responsive systems where the strengths of these binding interactions can be perturbed. Perturbation of these intermolecular forces is achieved through the use of external signals such as light absorption, changes in temperature and/or pH, etc. However, in an almost unexplored area, the use of an applied voltage as an external signal is expected to perturb intermolecular binding via electrochemical redox reactions. One such system known for self-complementary hydrogen bonding is the Ureidopyrimidinone (UPy). Introduced by Meijer, UPy can dimerize with another UPy along its complementary 4 hydrogen bond frame. Due to their directionality and strength of association, UPy’s have been utilized as cross-linkers in oligomerization and as capping agents in oligomer chain termination. In these cases, association and/or dissociation of UPy dimers can be influenced by external signals that were previously described. However, the use of voltage as a means of dimerization control of the UPy system and other H-bonded systems is virtually unheard of. Preliminary studies on electroactive UPy derivatives – UPy’s with substituents rendering them responsive to electrochemical stimuli – have been explored. One such UPy derivative is UPyFc, where a ferrocene substituent acts as an electrochemical redox center of the UPy derivative. Cyclic voltammetry and bulk electrolysis / steady-state voltammetry studies reveal that dimerization of UPyFc can be electrochemically controlled, and the mechanism of the electrochemical reaction indicates a square scheme for the UPyFc / UPyFc+ redox couple in dichloromethane. The fully reduced UPyFc exists as a dimer. Upon oxidation to UPyFc+, the dimer breaks apart to form the monomer. Furthermore, this process appears to be completely reversible in dichloromethane. Based off of these preliminary studies, electro-inactive UPy oligomers have been developed with the goal of altering the physical properties of bulk non-polar organic solutions containing these UPy oligomer units and electro-active UPy derivatives. It is expected that, in bulk solution, the formation of UPy hetero-dimers between UPyFc and electro-inactive UPy oligomer units will be observed and the bulk solution will show low viscosity. Oxidizing UPyFc to UPyFc+ will break these hetero-dimers apart and UPy oligomerization will be observed, in which the bulk solution should show an increase in viscosity. Based off of preliminary studies, this should be a reversible process in which physical properties of bulk solutions can be altered by controlling H-bonding interactions via electrochemical redox reactions.

The formation of supramolecular polymers is possible through directional and favorable intermolecular interactions that, unlike conventional covalent bonds, are weaker and purely electrostatic. The electrostatic nature of these interactions grants their respective systems reversibility in binding interactions, allowing molecular systems to achieve dynamic behavior. An important goal in supramolecular chemistry is the development of stimuli-responsive systems where the strengths of these binding interactions can be perturbed. Perturbation of these intermolecular forces is achieved through the use of external signals such as light absorption, changes in temperature and/or pH, etc. However, in an almost unexplored area, the use of an applied voltage as an external signal is expected to perturb intermolecular binding via electrochemical redox reactions. One such system known for self-complementary hydrogen bonding is the Ureidopyrimidinone (UPy). Introduced by Meijer, UPy can dimerize with another UPy along its complementary 4 hydrogen bond frame. Due to their directionality and strength of association, UPy’s have been utilized as cross-linkers in oligomerization and as capping agents in oligomer chain termination. In these cases, association and/or dissociation of UPy dimers can be influenced by external signals that were previously described. However, the use of voltage as a means of dimerization control of the UPy system and other H-bonded systems is virtually unheard of. Preliminary studies on electroactive UPy derivatives – UPy’s with substituents rendering them responsive to electrochemical stimuli – have been explored. One such UPy derivative is UPyFc, where a ferrocene substituent acts as an electrochemical redox center of the UPy derivative. Cyclic voltammetry and bulk electrolysis / steady-state voltammetry studies reveal that dimerization of UPyFc can be electrochemically controlled, and the mechanism of the electrochemical reaction indicates a square scheme for the UPyFc / UPyFc+ redox couple in dichloromethane. The fully reduced UPyFc exists as a dimer. Upon oxidation to UPyFc+, the dimer breaks apart to form the monomer. Furthermore, this process appears to be completely reversible in dichloromethane. Based off of these preliminary studies, electro-inactive UPy oligomers have been developed with the goal of altering the physical properties of bulk non-polar organic solutions containing these UPy oligomer units and electro-active UPy derivatives. It is expected that, in bulk solution, the formation of UPy hetero-dimers between UPyFc and electro-inactive UPy oligomer units will be observed and the bulk solution will show low viscosity. Oxidizing UPyFc to UPyFc+ will break these hetero-dimers apart and UPy oligomerization will be observed, in which the bulk solution should show an increase in viscosity. Based off of preliminary studies, this should be a reversible process in which physical properties of bulk solutions can be altered by controlling H-bonding interactions via electrochemical redox reactions.

Oxidation and reduction of organic redox-couples typically leads to huge changes in their acidity/basicity, with the result that proton transfer (PT) often accompanies electron transfer (ET). These types of reactions are known as proton-coupled-electron-transfer or PCET reactions. While the prevalence of PCET reactions in reversible organic redox transformations is something that has been recognized for over 100 years, it is only much more recently that it has become apparent that H-bonding intermediates often play a significant role in the mechanism of these reactions. One way in which H-bonding can do so is through what we have termed a “wedge-scheme” mechanism in which ET and PT occur through the H-bond intermediate formed between the redox-active acid or base and the added base or acid.1 This can provide an easier path for ET-PT in either direction and thus increase the reversibility of the overall reaction. Another means through which H-bonding can affect PCET is by stabilizing the products. In the PCET reaction to be discussed in this presentation, reduction of N-methyl-4,4'-bipyridinium (commonly called “monoquat” or MQ) in acetonitrile in the presence of acidic alcohols, it appears that both H-bond effects are present. MQ undergoes two reversible one electron reductions in acetonitrile, the first corresponding to reduction of the starting cation, MQ+, to an uncharged radical, MQ. This is followed by a second reduction at significantly more negative potentials of the radical to the anion, MQ−, which formally contains a negatively-charged nitrogen and is thus quite basic. Not surprisingly, CV studies show that addition of trifluoroethanol or dichloroethanol, both fairly acidic alcohols, result in no change in E1/2 of the first reduction but a significant positive shift in the E1/2 of the second reduction, most likely due to protonation of the anion. The second reduction wave is slightly broad, but continues to shift positive and remain chemically reversible up to at least 400 equivalents of added alcohol, at which point it has begun to merge with the first reduction. What is surprising, in addition to the chemical reversibility, is that analysis of the observed E1/2 as a function of added alcohol concentration is consistent with a single 1e− transformation involving three alcohols throughout the entire range. Since it is highly unlikely that the MQ anion would be triply protonated, our current hypothesis is that the product conjugate base of the alcohol forms a strongly H-bonded complex with two other alcohols, making the overall reaction MQ + 3ROH + e− = MQH + (RO−)(ROH)2. The mechanism would involve initial 1 e−, 1H+ transfer occurring through the H-bond intermediate (wedge scheme), followed by rapid formation of the homoconjugate complex (RO−)(ROH)2. The results of current and future work on this system to further elucidate the mechanism, and either support or refute the wedge-scheme/homoconjugate hypothesis, will be presented. 1 L. A. Clare, A. T. Pham, F. Magdaleno, J. Acosta, J. E. Woods and D. K. Smith, Journal of the American Chemical Society 2013, 135, 18930-18941.

The kinetics of many proton-coupled electron transfer (PCET) reactions cannot be adequately described by step-wise proton and electron transfer. Concerted electron-proton transfer (CPET) is another possibility, but examples exist where step-wise mechanisms are not viable yet there is no compelling evidence for CPET. This study investigates such a reaction, the oxidation of an NH-containing phenylenediamine radical cation, H2PD+, in the presence of pyridines in acetonitrile. H2PD+ is formed by a net one electron oxidation of 2,3,5,6-tetramethylphenylenediamine in acetonitrile via a rather complicated mechanism that likely proceeds through a H-bonded dimer. Once formed, it can be further oxidized at more positive potentials to the quinoidal dication, H2PD2+, which will be considerably more acidic than the radical cation. Indeed, the E1/2 for the radical oxidation jumps to a considerably more negative potential upon addition of 1 equivalent of the weak base pyridine. The CV wave broadens but stays chemically reversible. Further addition of pyridine leads to smaller E1/2 shifts with continued reversibility. This behavior could possibly be explained by stabilization of H2PD2+ through H-bonding or proton transfer to pyridine, however, UV-vis spectroelectrochemical experiments provide definitive evidence for proton transfer. Furthermore, CV studies with 4-substituted pyridine derivatives of weaker basicity show that the observed E1/2 with 1 equivalent of the pyridine depends in a Nernstian fashion on the pKa of the conjugate acid of the pyridine, indicating that all the pyridines studied deprotonate H2PD2+ to give the quinoidal cation, HPD+. The continued E1/2 shift observed upon further addition of the pyridines can then be completely explained by application of the Nernst equation to the overall reaction H2PD+ + pyr → HPD+ + Hpyr+ + e−. However, while this explains the thermodynamics, the classic step-wise proton-electron transfer mechanism for this reaction cannot explain the observed reversibility at high base concentration. In contrast, the reversibility can be explained by a “wedge” scheme mechanism in which electron and proton transfer occur within the H-bond complex formed as an intermediate in proton transfer. In this case, the electron-proton transfer could be concerted, however, the kinetics for the second oxidation in the presence of pyridine show no significant isotope effect. Furthermore, a new reduction peak that appears at faster scan rates suggests the presence of an intermediate in the electron-proton transfer. Both results strongly suggest that the electron-proton transfer is step-wise, not concerted, within the H-bond complex. This result points to the important role H-bonding may play in PCET even without CPET.

Ureidopyrimidinone (UPy) derivatives, introduced by Meijer and co-workers, are capable of forming dimers linked by four strong H-bonds. They have been widely used as polymer crosslinkers in self-healing supramolecular polymers and gels. While their donor-donor-acceptor-acceptor (DDAA) H-bond motif contributes to polymer stretchiness, their relatively high association constant of 106 also contributes to their ability to autonomously self-repair. This presentation shows the result of substituting a UPy with a N,N- dimethyl-p-phenylenediamine and therefore adding a redox reactive dimension to self-repair. In particular, this work involves the electrochemical investigation of UPyH (shown below) with a series of different bases, specifically pyridine and substituted pyridines. The focus is on finding the appropriate electroinactive base that facilitates both the dimer break up upon oxidation and proton transfer and dimer reformation by reduction and back proton transfer. Cyclic voltammetry (CV) scans of UPyH only, conducted in CH2Cl2 at slower scan rates, show two reversible redox waves. The first wave is of one electron height, corresponding to a two-electron transfer per dimer. At a higher oxidation potential, the second oxidation wave is one half the height of the first and then on the return scan, a slightly more negative potential is needed for reduction. At a faster scan rate, the CV response for the second wave shows that it becomes irreversible. Also, at a faster scan rate, a new reduction wave appears at a more negative peak potential than the reduction wave for the first electron transfer process. This type of CV return wave behavior for UPyH at faster scan rates was previously seen in another dimethylphenylurea redox active compound with only two H-bonding sites, UHH. In the UHH investigation, CV and spectroelectrochemical results pointed to the dimethylamino moiety being basic enough to abstract an N-H proton in the radical cation oxidation state and therefore allowing for a two electron transfer wave to give the quinoidal cation. A similar reduction wave negative of the first redox wave was seen in acetonitrile at all scan rates, while in CH2Cl2 this wave was seen only at faster scan rates. At slower scans the first oxidation remained reversible. Based upon the UHH study, interpretation of the UPyH second oxidation wave is that it correspond with a second e- transfer per redox center. Each urea-phenylenediamine N-H site in the dimer then becomes acidic enough to transfer its proton to the carbonyl oxygen on the UPyH isocytosine. This would then change the favorable binding motif of DDAA to a weaker binding motif of ADAD. In addition, a 2+ charge forms on each monomer in the dimer contributing to strong repulsive forces. This results in dimer break-up, where upon the protonated isocytosine carbonyl on each monomer is exposed to the more basic dimethylamino site on a fully reduced UPyH dimer. Proton transfer from the isocytosine to the dimethyl amino renders the fully reduced dimer electroinactive, explaining the half wave height. Thus, the overall reaction, shown in the accompanying scheme, corresponds to break up of the half the starting dimers, with two electron oxidation and deprotonation of each monomer to give two UPy+. However, the other half of the UPy’s remain dimerized in the protonated form as (HUPyH+)2. Further investigation focuses on the addition of a series of substituted pyridines as a base and their effect on dimer break up and reformation as it competes with the dimethylamino moiety on the phenylenediamine. The goal is to find an external base that will allow full oxidation and break up of all the UPyH dimers, while retaining the chemical reversibility of the process. UPy has a unique self-healing property that can be observed electrochemically as a redox couple which can be useful for supramolecular polymers. With the use of UPyH’s H-bond array and a phenylenediamine redox couple attached to the urea, this would allow H-donating of the NH urea. The main focus of this study is UPyH+ base will be used as a redox-controlled chain terminator that can control the degree of polymerization. The viscosity of the supramolecular polymers will determine if UPyH base would act as a chain terminator, and if it would be electroactive to oxidize to deactivate the polymer chain to reform a chain with the base. When the polymers reduce, UPyH would go back at its starting state by reforming with itself.

Water-oxidation catalysts (WOCs) can potentially be improved by installing pendant electron-donor groups that may also be proton donors or acceptors. We have modified one of the most well-studied WOCs with alkoxy or hydroxy substituents on the bidentate bipyridine ligand (N,N), thereby forming [(terpy)RuII(N,N)X] (X = Cl, H2O; terpy = 2,2′;6′,2"-terpyridine). A combination of NMR spectroscopy (particularly 15N chemical-shift data), UV/Vis spectroscopy, X-ray diffraction, and oxygen evolution data point to interesting and beneficial effects of an oxygenated group proximal to X. A methoxy group on the 2,2′-bipyridyl (bipy) ring cis to X = Cl is shown to facilitate ionization of the chloride ligand in aqueous acetone, perhaps by acceptance of a hydrogen bond from the aquo ligand. Hydrogen-bond donation of a proximal hydroxy group to a bound aquo ligand is shown by X-ray diffraction. Distinct differences in pKa values for the 4,4′- and 6,6′-dihydroxy bipy complexes are seen. In water oxidation driven by ceric ammonium nitrate, the 6,6′-dimethoxy species is somewhat faster and longer-lived than the analogue that lacks the oxygenated groups [a turnover number (TON) of 215 instead of 138 in 10 h, and a turnover frequency (TOF) of 0.36 min–1 instead of 0.23 over the same time period]. Taken together, oxygenated groups near the WOC active site are promising electron or proton donors and/or hydrogen-bond acceptors, and are the subject of further scrutiny.

A cyclic voltammetric investigation of nitrobenzene in DMSO in the presence of various acids including boc- and methylester-protected cysteine is reported. It has been hypothesized that the reaction of thiols with reduced nitroaromatics is at the root of the biological activity for a variety of nitroaromatics, including a number of medicinally important nitroheterocycles used to treat infections by anaerobic microorganisms. This study shows that the only reaction occurring between cysteine and the 1e − reduction product, the nitro radical anion, is proton transfer. However, by comparing the voltammetry observed in the presence of cysteine to that observed in the presence of a range of different acids in DMSO we found that further reaction occurs between cysteine and the 2e − reduction products, either the nitrogen diol or the nitroso, greatly enhancing the overall rate of the reduction. This solution reaction proceeds through a rapidly formed, oxidizably-detectable intermediate, which, by analogy to known chemistry in aqueous solution, is likely the semi-mercaptal. Thus this study shows that cyclic voltammetry in aprotic solvents is a very useful technique for studying the reactivity of nitroaromatics with thiols and confirms that the substantive reaction occurs at the 2e − stage of the reduction.

Supramolecular polymers can be formed through directional intermolecular interactions between designated molecular subunits in order to achieve dynamic behavior. The strength of these intermolecular interactions can be altered through the use of external signals such as light absorption, changes in temperature and/or pH, or in an almost unexplored caveat, the application of voltage. Such is one of the important goals in supramolecular chemistry: the development of stimuli-responsive systems where the strength of intermolecular interactions can be perturbed through the use of external signals. Applications of supramolecular chemistry include self-healing polymers and gels, controlled release of entrapped molecules, and the development of smart materials. Ureidopyrimidinones (UPy) – introduced by Meijer and coworkers – can form two self-complementary 4 hydrogen bond arrays with binding motifs AADD or ADAD and have a relatively small dissociation constant of 10-6. Due to their directionality and strength of dimerization, UPys have been widely imployed as polymer cross-linkers in oligomerization and also as capping agents in chain termination. In these cases, association and/or dissociation of UPy hydrogen bonded complexes are influenced by external signals that were previously described. However, the application of voltage to affect binding of the UPy supramolecular complex is a relatively new approach towards dimerization control. In order for UPy to show electrochemical responsiveness, an electroactive substituent needs to be attached to the UPy framework. We have prepared and characterized such a system in which a ferrocene (Fc) is attached to the pyrimidone ring in the UPy. Concentration and scan-rate-dependent cyclic voltammetry and diffusion coefficient data from bulk electrolysis coupled with steady-state voltammetry indicate a square scheme mechanism for the UPyFc / UPyFc+ redox couple in dichloromethane. The fully reduced UPyFc exists as a dimer. Upon oxidation to UPyFc+, the dimer breaks apart to form the monomer. However, it is unknown how conversion between the AADD and ADAD binding motifs – due to tautomerization – factors into the square scheme mechanism. Nonetheless CV simulations with a simple square scheme mechanism (neglecting tautomerization) provide reasonable fits to the experimental data. Importantly, the overall process appears to be completely chemically reversible even over the relatively long time scale of bulk electrolysis. Based off of this preliminary work, syntheses of covalently-linked UPyFc oligomers are underway with the goal of creating a redox-responsive supramolecular polymer whose physical properties can be altered upon oxidation. As it has been confirmed that the fully reduced UPyFc exists as a dimer, the physical properties of the material should reflect this dimerization by showing rigidity or in some cases, a highly viscous material. Upon oxidation, dimerization should cease, causing the ferrocene-UPy oligomer to break apart and show physical properties that reflect a loss in rigidness and greater fluid, thereby achieving a more selective process of perturbing electrostatic interactions in supramolecular assemblies.

Oxidation and reduction of organic redox-couples typically leads to huge changes in their acidity/basicity, with the result that proton transfer (PT) often accompanies electron transfer (ET). These types of reactions have come to be known as proton-coupled-electron-transfer or PCET reactions. While the prevalence of PCET reactions in reversible organic redox transformations is something that has been recognized for over 100 years, it is only much more recently that it has become apparent that H-bonding intermediates often play a significant role in the mechanism of these reactions. One way in which H-bonding can do so is through what we have termed a “wedge-scheme” mechanism in which ET and PT occur through the H-bond intermediate formed between the redox-active acid or base and the added base or acid.1 This can provide an easier path for ET-PT in either direction and thus increase the reversibility of the overall reaction. Previously we have demonstrated the utility of the wedge scheme in oxidation reactions involving phenylenediamines and bases.1,2 In new work, we are exploring the utility of the wedge scheme in explaining the chemically reversible PCET reaction occurring upon addition of acidic alcohols to the N-methyl-4,4'-bipyridinium redox couple (commonly called “monoquat” or MQ) in acetonitrile. This compound undergoes two reversible one electron reductions in acetonitrile, the first corresponding to reduction of the starting cation, MQ+, to an uncharged radical, MQ. This is followed by a second reduction at significantly more negative potentials of the radical to the anion, MQ-, which formally contains a negatively-charged nitrogen and is thus quite basic. Not surprisingly, cyclic voltammetry (CV) studies show that addition of trifluoroethanol (TFE), a fairly acidic alcohol, results in no change in E1/2 of the first reduction but a significant positive shift in the E1/2 of the second reduction, most likely due to protonation of the anion. The second reduction wave is slightly broad, but continues to shift positive and remain chemically reversible up to at least 400 equivalents of added alcohol, at which point it has shifted over 0.6 V and almost merged with the first reduction wave. What is perhaps surprising, in addition to the chemical reversibility, is that analysis of the observed E1/2 as a function of added alcohol concentration is consistent with a single 1e-, 2 H+ transformation (MQ + 2 HA + 1e- = MQH2+ + 2 A-) throughout the entire range. Qualitatively similar behavior observed with addition of alcohols to quinones in acetonitrile has commonly been attributed to strong H-bonding between the alcohol and the quinone dianion. However, a H-bonding only mechanism is not likely in this case both because of the expected basicity of MQ- and the uniformity of the reaction stoichiometry over the entire concentration range. (With H-bonding only, the stoichiometry would likely start 1:1, then go to 2:1 at higher concentrations.) On the other hand, a classic step-wise PCET mechanism, ET-PT-PT, can not explain the observed reversibility over the entire concentration range. What we believe can explain all of the above is initial 1 e-, 1H+ transfer occurring through the H-bond intermediate (wedge scheme), followed by the second 1H+ transfer. The results of current and future work on this system to further elucidate the mechanism, and either support or refute the wedge-scheme hypothesis, will be described in this presentation. These will include studies with other acids, CV simulations and analysis of deuterium isotope effects 1 L. A. Clare, A. T. Pham, F. Magdaleno, J. Acosta, J. E. Woods and D. K. Smith, Journal of the American Chemical Society 2013, 135, 18930-18941. 2B. T. Tamashiro, M. R. Cedano, A. T. Pham and D. K. Smith, Journal of Physical Chemistry C 2015, 119, 12865-12874.

An important goal in supramolecular chemistry is the development of stimuli-responsive systems in which the strength of the intermolecular interactions can be altered by external signals such as changes in light, temperature, pH or voltage of an electrode. Our group is exploring the latter possibility, primarily by trying to develop systems in which electron transfer perturbs the strength of H-bonding interactions between molecules. The underlying principle is straightforward: an oxidation that decreases the negative charge on a H-acceptor (A) or a reduction that decreases the positive charge on a H-donor (D) will weaken a H-bond. Alternatively, reduction that increases the negative charge on a H-acceptor or oxidation that increases the positive charge on a H-donor will increase the strength of a H-bond. However, in the latter case, it is possible that oxidation or reduction could also lead to full proton transfer. If this occurs across the H-bond, the primary H-bonds will remain, but the secondary H-bonds will change. This can lead to an increase in unfavorable secondary interactions, which would counteract the effect of the initial proton transfer. However, with proper design, proton transfer could lead to an increase in favorable secondary interactions, which would enhance the effect of initial transfer. The goal of this project is to do the latter. For this work, the 3 H-bond DAD array, 1, that contains a N-methyl-4,4'-bipyridinium or “monoquat” redox couple (see figure) has been synthesized. Compound 1 forms a three H-bond dimer with the non-electroactive ADA array, 2, in CH2Cl2. Typically, DAD-ADA arrays such as this have relatively weak association constants of ~102 M−1 in non-competitive solvents such as CH2Cl2 due to the three, favorable primary H-bonds (shown as solid, double-headed, green arrows in the figure) being counterbalanced by four unfavorable secondary interactions (shown as dashed, double-headed, red arrows). Preliminary NMR studies suggest stronger binding in this case (>103 M−1), probably due to the greater acidity of the amide NH’s in 1 because of the strongly electron-withdrawing nature of the pyridinium. In the absence of 2, initial cyclic voltammetry studies of 1 display the expected two, sequential, 1 e− reduction waves of the monoquat redox couple in CH2Cl2, corresponding to reduction of 1 to the radical and then the quinoidal anion. Addition of 2 results in no change in the E1/2 of the first reduction, but the second reduction shifts ~0.24 V positive. Further addition of 2 causes no additional change in the E1/2, consistent with a 1:1 complex. The 0.24 V positive shifts indicate a ~104 increase in binding strength upon overall 2 e− reduction of 1. Combined with the initial association constant of >103 M−1, this indicates an association constant of >107 M−1 in the fully reduced state. We believe the most likely explanation of such strong H-bonding in a 3 H-bond array is that the second reduction induces proton transfer across the central H-bond in the complex, thus converting the DAD-ADA array to a DDD-AAA array. The latter is expected to have significantly stronger H-bonding because all of the secondary interactions, in addition to the primary interactions, are favorable. If confirmed by further studies, the association constant in the reduced form would represent, to the best of our knowledge, the largest achieved to date in a redox-dependent H-bonding system, thereby demonstrating the utility of incorporating proton transfer to amplify the effect of electron transfer in these systems. Figure 1

An important goal in supramolecular chemistry is the design of stimuli-responsive supramolecular systems in which binding between components can be controlled by external signals. Supramolecular chemistry research in our group focuses in particular on using electron transfer to control the strength of H-Bonding interactions. In this presentation the redox-dependent H-bonding behavior of two simple phenylureas containing either a phenylenediamine redox couple, Figure 1a, or a ferrocene redox couple, FcUHH, Figure 1b, will be compared. The urea functional group provides two good H-donors (a DD array) which can be paired with two good H-acceptors (an AA array) in the PZD guest. Oxidation of either the ferrocene or phenylenediamine would be expected to increase the acidity of the attached urea NH, increasing the H-Bonding strength to PZD. Experimentally, this should result in a negative shift in the E1/2 of the cyclic voltammetric wave upon addition of PZD, with little change in wave shape or height. This behavior is indeed observed for both ureas in 0.1 M NBu4PF6/CH2Cl2. Plots of the ΔE1/2 of FcUHH vs [PZD] are fit well using a 1:1 binding isotherm, with initial results giving estimates of the binding constants to the oxidized and reduced forms as Kred = 80 M−1 and Kox = 500 M−1. The same type of experiments have been run with the phenylenediamine urea, UHH. The data also fits well to a simple 1:1 binding isotherm and gives similar values for the binding constants, Kred = 30 M−1 and Kox = 700 M−1. This is interesting because from our previous work1 we know that the electrochemistry is quite a bit more complex for UHH than FcUHH. In the former case the reversible, apparent 1 e− cyclic voltammetric (CV) wave observed in CH2Cl2 actually corresponds to 2 e− oxidation of half of the UHH’s to a quinoidal cation accompanied by proton transfer and electro-deactivation of the other half, Figure 1a. Given this, it is somewhat surprising that any potential shift is observed upon addition of PZD, since the oxidation product can only form a single H-bond with PZD. However, the key realization was that the other product, the protonated urea, can still form a 2 H-bond complex, and that should be stronger than the one formed with the starting urea due to the increased acidity of one of the urea NH’s. From this perspective, it is not surprising that we see such similar behavior between FcUHH and UHH. In the former, oxidation creates a +1 charge on one side of the urea (ferrocenium), Figure 1b, and, in the latter, proton transfer creates +1 charge, Figure 1a. This conclusion suggests an entirely new strategy to perturb H-bonding through electron transfer. Rather than doing this directly through oxidation/reduction introduced charges, it can also be done indirectly through oxidation-reduction induced proton transfer! 1 L. A. Clare, A. T. Pham, F. Magdaleno, J. Acosta, J. E. Woods and D. K. Smith, Journal of the American Chemical Society 2013, 135, 18930-18941. Figure 1

Strength and directionality are the hallmark characteristics of binding for molecules with multiple hydrogen bonds. This is important in the area of molecular recognition and supramolecular construction. When coupled to a redox center, oxidation or reduction can affect hydrogen bonds by either strengthening or weakening bonding, and thus allow for greater control over this type of binding. One class of model compounds which can be used to investigate the effects of manipulating the characteristics of hydrogen bonding are bi-substituted ureas. An example previously studied in our group is 1-phenyl-3-(4-dimethylamino)phenylurea (UHH), which has a H-donor-donor (DD) motif. This compound was coupled with two guests that have a H-acceptor-acceptor (AA) motif, 1,4-dimethylpiperizine-2,3-dione (PZD) and 1,8-naphthyridine (naph), to investigate the effects of oxidation on hydrogen binding. Cyclic voltammetry (CV) of UHH in the presence of PZD in methylene chloride shows a reversible, one electron oxidation wave whose E1/2 shifts negative with an increase in concentration of PZD. This indicates an increase in hydrogen binding strength upon oxidation. It was originally thought to be due to stronger binding to the UHH radical cation, but additional CV investigations of isolated UHH revealed that it actually undergoes a two electron per two urea process as shown in equation 1 to give the doubly oxidized UH+ plus the protonated, still reduced, HUHH+. Since only HUHH+ retains the DD motif, it is more likely that it is strong H-bonding between HUHH+ and PZD that cause the observed potential shift. Analyzing the potential shift data based on this assumption gives binding constants Kred = 30 M−1 and Kox =700 M−1. In contrast to PZD, CV studies of UHH in the presence of naph show the current for the oxidation increases with little change in E1/2. This is rationalized by proton transfer from the UHH radical cation to the naph rather than the dimethylamino group on a reduced UHH, equation 2. This allows full oxidation of all the UHH, but results in conversion of the more favorable DD-AA H-bond motif to the less favorable AD-DA plus creates electrostatic repulsion. These effects counteract the increased acidity caused by oxidation, apparently resulting in no significant change in H-bonding strength. In new studies, the phenylenediamine redox center in UHH has been replaced with a ferrocene redox center, producing the compound 1-ferrocenyl-3-phenylurea (FcUHH). This change limits the oxidation to a one electron process, equation 3, and removes the basic site that facilitates proton transfer. Initial CV experiments of FcUHH and PZD in methylene chloride show a reversible, one electron oxidation with a corresponding negative shift in E1/2 with increasing concentration of PZD. The magnitude of the shift is comparable to UHH with PZD, and analysis of this data gives binding constants Kred = 80 M−1 and Kox = 500 M−1. Since oxidation of the ferrocene to ferrocenium produces a +1 charge, as does protonation of the dimethyamino group of UHH, the similarity of this result to that observed with UHH and PZD supports the hypothesis that the potential shifts observed with UHH and PZD are due to H-bonding with the protonated, reduced urea. Similar studies will be performed with FcUHH and naph for comparison to the UHH and naph result. One complication in the studies with FcUHH is that, to date, we have been unable to fully remove a small amount of 1,3-diphenylurea side-product. Since the N-H’s in FcUHH and 1,3-diphenylurea likely have similar polarity and thus affinity for PZD and naph, our existing binding constant measurements are flawed due to competition between FcUHH and diphenylurea for binding to the guest. This issue is currently being addressed with NMR titrations of 1,3-Diphenylurea with PZD and naph to measure these binding constants. This will allow a more accurate mathematical model to be constructed to fit the potential shift data for FcUHH and the guests. Moving forward, we hope to optimize the reaction conditions for the synthesis of FcUHH to prevent formation of the diphenylurea. However, whether we are successful with this or not, it is clear that the studies with FcUHH will provide an interesting and informative comparison to the earlier work with UHH and add to our understanding of redox-dependent H-bonding in general. Figure 1

An important goal in supramolecular chemistry is the development of stimuli-responsive systems in which the strength of the intermolecular interactions can be altered by external signals such as changes in light, temperature, pH or voltage of an electrode. Our group is exploring the latter possibility, primarily by trying to develop systems in which electron transfer perturbs the strength of H-bonding interactions between molecules. The underlying principle is straightforward: an oxidation that decreases the negative charge on a H-acceptor (A) or a reduction that decreases the positive charge on a H-donor (D) will weaken a H-bond. Alternatively, reduction that increases the negative charge on a H-acceptor or oxidation that increases the positive charge on a H-donor will increase the strength of a H-bond. However, in the latter case, it is possible that oxidation or reduction could also lead to full proton transfer. If this occurs across the H-bond, the primary H-bonds will remain, but the secondary H-bonds will change. This can lead to an increase in unfavorable secondary interactions, which would counteract the effect of the initial proton transfer. However, with proper design, proton transfer could lead to an increase in favorable secondary interactions, which would enhance the effect of initial transfer. The goal of this project is to do the latter. In particular, 3-H-bond DAD-ADA arrays will be studied in which the DAD component contains a N-alkyl-4,4'-bipyridinium or “monoquat” redox couple (see figure). DAD-ADA arrays have relatively weak association constants (102 in non-competitive solvents such as CHCl3) due to the three favorable primary H-bonds (shown as solid double-headed arrows in the figure) being counterbalanced by four unfavorable secondary interactions (cross interactions shown as dashed double-headed arrows). As shown in the figure, reduction of the monoquat by one electron should lead to proton transfer, creating not only a strong ionic primary H-bond, but converting all the secondary H-bonds to favorable ones. Examples of similar ionic DDD-AAA complexes have been reported in the literature with association constants greater than 109, suggesting it may be possible see a binding enhancement on the order of 107 upon reduction in this system. Initial voltammetric studies with just monoquat itself (x = nothing) and the diimide ADA guest show a significant positive shift in the E1/2 of the second reduction of monoquat, but little to no shift in the first reduction in both acetonitrile and methylene chloride. The shift is larger in acetonitrile, consistent with it being due to proton transfer between the guest and the fully reduced monoquat. Currently, synthesis of the compete DAD array is underway, which will have X = NC(=O)R to make the H-donor a strongly donating amide NH. In contrast to the monoquat results, CV studies with this compound and the diimide ADA guest are expected to show a E1/2 shift in the first reduction that is significantly larger in methylene chloride than acetonitrile, indicating very strong H-bonding to the radical form. Figure 1

An important class of supramolecular systems are those capable of forming dimers held together by multiple linear H-bonds. Of these, the 4 H-bond arrays based on the ureidopyrimidone or UPy unit have found particularly use, having the right combination of ease-of-synthesis and appropriate binding strengths to be useful for the formation of supramolecular polymers and gels. The selling points for these materials are there self-healing properties. By application of heat or mechanical stress the polymer chains will break reversibly at the H-bonds making the material more fluid and allowing a defect to fill in. Upon cooling or relief of the mechanical stress, the H-bonds re-form locking the repair in place. For many applications the use of heat or mechanical stress as the external signal to alter structure is perfect, however, for other applications, a more selective stimulus such as electron transfer would be advantageous. To do this, a reversible redox couple needs to be attached to the UPy in such a way that oxidation or reduction perturbs the H-bond strength. Over the last several years, we have synthesized and investigated the electrochemistry of several such systems, shown below. With UPyH, there is a phenylenediamine redox couple attached to the urea allowing perturbation of the H-donating ability of the urea NH. In the other examples, the redox couple, either a pyridinium [UPy(MeP+)] or a ferrocene [UPy(Fc)], is attached to the pyrimidone ring, allowing perturbation of the H-accepting ability of the pyrimidine N and O. In addition to their potential use in redox-responsive supramolecular materials, these compounds also provide good platforms for fundamental investigations on the role of H-bonding in proton-coupled electron transfer. Indeed, our studies show that with both UPyH and UPy(MeP+) electron transfer leads to proton transfer, resulting in interesting, albeit rather complicated electrochemistry. In both these cases, electron transfer would be expected to increase H-bond strength. In contrast, with UPy(Fc), oxidation should decrease H-bond strength. The result is no proton transfer and greatly simplified electrochemistry, which is more amenable to future supramolecular applications. A thorough analysis of the voltammetry indicates a 107 decrease in Kdim upon oxidation. This result coupled with the high stability of the oxidized form suggests UPy(Fc) is highly amenable to supramolecular applications. Currently we are working on the synthesis of linked UPy(Fc)’s suitable for the formation of supramolecular polymers. Progress on this work will be reported. Figure 1

Proton transfer often accompanies electron transfer in organic redox couples due to the large changes in acidity or basicity that result from oxidation or reduction. Classically, the overall reaction is thought to occur in two separate steps, either proton transfer followed by electron transfer, PT-ET, or electron transfer followed by proton transfer, ET-PT. More recently it has become abundantly apparent that a third option is also possible in which the proton and electron move together in a single kinetic step. This is concerted proton-electron transfer, CPET, and is generally believed to occur within a H-bonded intermediate. It is straightforward to show that E° of this step will have a value in between the E° values corresponding to the oxidation of the fully protonated and fully deprotonated forms. Experimental evidence for CPET is typically the observance of a significant deuterium isotope effect for the apparent one electron transfer. However, it is important to note that the electron-proton transfer does not have to be concerted within the H-bond complex. It could also occur step-wise. If slow electron transfer occurred followed by rapid proton transfer within the H-bond complex, no significant deuterium isotope effect would be expected, yet, at the same time, the overall reversibility of the electron-proton transfer would still be aided by providing a mechanistic pathway through the H-bond complex with its intermediate E°. We believe that the second oxidation of a simple phenylenediamine, H2PD, in the presence of pyridines in acetonitrile provides a nice example of the significant role that step-wise electron-proton transfer within a H-bond intermediate can play. Without added base, the second oxidation corresponds to the reversible one electron oxidation of the radical cation, H2PD+, to the quinoidal dication, H2PD2+. CV data in the presence of one equivalent of pyridines of different basicity result in a significant negative shift in the E1/2 of the second CV wave with no loss in reversibility. Continued addition of the pyridine leads to smaller incremental shifts with still no change in reversibility. Analysis of the observed E1/2’s with one equivalent pyridine as a function of pyridine basicity indicates that, even with the weakest base, proton transfer to the pyridine occurs. This interpretation is supported by spectroelectrochemical data. Thus, the overall reaction occurring in the second CV wave is H2PD+ + B = HPD+ + HB+ + e-, where B is the pyridine base and HB+ is its conjugate acid. The shifts in E1/2 upon further additions of pyridine are nicely predicted by the Nernst equation with this as the overall reaction. However, attempts to simulate the CV data with just PT and ET steps fail at the higher concentrations of pyridine. This is because the rate of the ET step becomes exceedingly slow as the base concentration increases and the thermodynamic potential becomes increasingly removed from the E° of the actual ET step. This issue is simply solved by including in the mechanism the possibility of electron transfer occurring through the H-bond intermediate. This simple change leads to good fits to the experimental data, providing solid support for such a H-bonding intermediate being involved. However, a comparison of experiments with 10:1 cyanopyridine: H2PD run in 2% CH3OH to those run in 2% CD3OD show no significant difference in the ΔEp of the second oxidation, providing no evidence for the oxidation involving CPET. Thus, it appears that ET-PT within the H-bond complex still greatly aids the reversibility of the reaction even without CPET.

Oxidation or reduction of organic redox couples typically leads to large changes in acidity or basicity, with the result that proton transfer often accompanies electron transfer, particularly in aqueous solution. In less polar organic solvents, H-bonding also can play an important role. While it is generally appreciated that proton transfer will have a greater effect on the overall reaction than H-bonding, it is not always straight forward to distinguish between the two, and, despite a considerable amount of research, a complete, quantitative understanding of the relative roles that the two play in the voltammetry that is observed upon addition of acids or bases to organic redox couples in non-aqueous solution remains elusive. This study focuses on the electrochemistry of p-tetramethylphenylenediamine, H 2 PD, in acetonitrile in the presence of pyridine bases. In acetonitrile, this compound undergoes a reversible one electron oxidation to the radical cation, H 2 PD + , followed by a second reversible oxidation to the strongly acidic quinoidal dication, H 2 PD 2+ . With the addition of 1 equivalent of the pyridine base, no significant change in the first oxidation is observed, but there is a large negative shift in the E 1/2 of the second oxidation with no loss in reversibility. Subsequent additions show a smaller, incremental negative shift, still with no loss in reversibility. We believe that this behavior signals proton transfer is occurring between the pyridine, pyr, and the H 2 PD 2+ , so that the overall reaction occurring in the second oxidation corresponds to H 2 PD + + pyr = HPD + + Hpyr + + e-. In this case, the observed E 1/2 should depend linearly on the pK a of the Hpyr + with a slope of −60 mV/pH unit. To test this hypothesis, the voltammetry of H 2 PD was studied with different pyridines that cover a range of pK a values. A plot of E 1/2 of the wave observed with 1 eq pyridine vs. pKa is indeed linear with close to the predicted slope. Furthermore, it is found that the continued shift in E 1/2 of the second oxidation with increasing concentrations of pyridine is well-accounted for simply by applying the Nernst equation to the overall reaction. However, while proton transfer can explain the potentials of the CV waves in the presence of added pyridine, simulations of the voltammetry show that proton transfer by itself cannot explain the observed reversibility of the second oxidation wave in the presence of increasing amounts of added pyridine. This is where H-bonding can play a role. By including H-bonding steps, and allowing electron transfer and proton transfer to occur through the H-bond complex formed between H 2 PD + /pyr and HPD + /Hpyr + , the simulations nicely explain both the observed potential shifts and the reversibility of the waves. The next question is whether the electron and proton transfer within the H-bonding complex is concerted or step wise. To test this, CV’s were run with 10:1 cyanopyridine:H2PD in 2% D 4 -methanol/acetonitrile or 2% H 4 -methanol/acetonitrile. If the second oxidation involved concerted electron-proton transfer, a significant deuterium isotope effect would be expected, causing a larger ΔE p in the deuterated solvent resulting from slower electron transfer. However, no significant difference was observed. Therefore, there is no evidence that the electron-proton transfer is concerted. This result does not by any means rule out the hypothesis that the proton-electron transfer is occurring within the H-bond complex, it merely indicates that it is likely that the proton and electron transfer occur in a step-wise fashion within the H- bond complex, and that facilitation of proton-electron transfer by H-bonding can happen even if the process is not concerted.

The voltammetry of 2,3,5,6-tetramethyl-p-phenylenediamine, H2PD, has been studied and compared to that of its isomer N,N,N’N’-tetramethyl-p-phenylenediamine, Me2PD. Both undergo two reversible electron transfer processes in acetonitrile that nominally correspond to 1e- oxidation to the radical cations, Me2PD+ and H2PD+, and a second 1e- oxidation at more positive potentials to the quinonediimine dications, Me2PD2+ and H2PD2+. While the voltammetry of Me2PD agrees with this simple mechanism, that of H2PD does not. The second voltammetric wave is too small. UV/Vis spectroelectrochemical experiments indicate that the second wave does correspond to oxidation of H2PD+ to H2PD2+ in solution. The fact that the second wave is not present at all at the lowest concentrations (5 µM), and that it increases at longer times and higher concentrations, indicates that H2PD+ is not the initial solution product of the first oxidation. A number of lines of evidence suggest instead that the initial product is a mixed valent, H-bonded dimer between one H2PD in the the full reduced, fully protonated state, H4PD2+, and another in the fully oxidized, fully deprotonated state, PD. A mechanism is proposed in which this dimer is formed on the electrode surface through proton transfer and H-bonding. Once desorbed into solution, it breaks apart via reaction with other H2PD’s, to give 2 H2PD+, which is the thermodynamically favored species in solution.

The efficacy of current centrally acting obesity pharmacotherapies is limited by compensatory mechanisms that mitigate weight loss.Our objective was to determine whether opioid receptor antagonism (naltrexone) plus pro-opiomelanocortin activation (bupropion) causes greater weight loss than placebo or monotherapy.A randomized, placebo- and monotherapy-controlled, double-blind, dose-finding trial was conducted from August 2005 to December 2006 in seven U.S. outpatient clinics.A total of 419 patients with uncomplicated obesity participated.Interventions included 24 wk of sustained-release bupropion (400 mg/d), immediate-release naltrexone (48 mg/d), placebo, and three combination therapy [naltrexone/bupropion (NB)] groups consisting of immediate-release naltrexone, 16, 32, or 48 mg/d, plus sustained-release bupropion (400 mg/d) with a 24-wk extension. A minimal diet and exercise component was also included.Percent weight change from baseline at wk 24 in the intent-to-treat population for NB48 vs. placebo and monotherapy was assessed. Other measurements included body mass index, waist circumference, fasting lipids, glycemic variables, safety, and tolerability.At wk 24, placebo-subtracted weight loss was -4.62% [95% confidence interval (CI) -6.24 to -2.99; P0.001] for NB16, -4.65% (95% CI -6.20 to -3.09; P0.001) for NB32, and -3.53% (95% CI -5.15 to -1.90; P0.001) for NB48. Weight loss was statistically significant vs. monotherapy for all three NB combinations with the exception of NB48 vs. bupropion. Weight loss with NB continued after wk 24. The most common treatment-emergent adverse event was mild transient nausea.NB caused gradual sustained weight loss over 48 wk; NB32 and NB16 demonstrated greater weight loss in the intent-to-treat population due to lower attrition rates. Further study is needed to demonstrate long-term efficacy and safety of NB.

Existing obesity therapies are limited by safety concerns and modest efficacy reflecting a weight loss plateau. Here, we explore combination therapy with bupropion (BUP), a putative stimulator of melanocortin pathways, and an opioid antagonist, naltrexone (NAL), to antagonize an inhibitory feedback loop that limits sustained weight reduction. In vitro electrophysiologic experiments were conducted to determine the extent to which BUP+NAL stimulated hypothalamic pro-opiomelanocortin (POMC) neurons in mouse brain. A subsequent study further characterized the effect of combination BUP+NAL treatment on food intake in lean and obese mice. Finally, a randomized, blinded, placebo-controlled trial in obese adult subjects was conducted. Randomization included: BUP (300 mg) + NAL (50 mg), BUP (300 mg) + placebo (P), NAL (50 mg) + P or P+P for up to 24 weeks. BUP+NAL stimulated murine POMC neurons in vitro and caused a greater reduction in acute food intake than either monotherapy, an effect consistent with synergism. Combined BUP+NAL provided sustained weight loss without evidence of an efficacy plateau through 24 weeks of treatment. BUP+NAL completers diverged from NAL+P (P < 0.01) and P+P (P < 0.001) at week 16 and from BUP+P by week 24 (P < 0.05). The combination was also well tolerated. Translational studies indicated that BUP+NAL therapy produced synergistic weight loss which exceeded either BUP or NAL alone. These results supported the hypothesis that NAL, through blockade of β-endorphin mediated POMC autoinhibition, prevents the classic weight loss plateau observed with monotherapies such as BUP. This novel treatment approach (BUP+NAL) holds promise for the treatment of obesity.

Cyclic voltammetric studies of 1-phenyl-3-(4-N,N-dimethylamino)phenyl urea (UH) in CH2Cl2 in the presence of several pyridine derivatives are described. By itself, UH undergoes a reversible one electron oxidation to the radical cation, UH+, and a second less reversible oxidation at more positive potentials. Addition of ten equivalents of 1,8-naphthyridine or pyridine causes the second oxidation to shift 700 mV negative, resulting in a quasi-reversible two electron oxidation at the potential of the UH/UH+ wave. Possible explanations for the large shift are proton transfer between UH2+ and the guest or very strong hydrogen bonding between the two. Computer simulation studies and comparison to known pKa values suggest that neither by themselves are adequate to explain the observed voltammetry. A mechanism involving both hydrogen bonding and proton transfer is more likely.

To compare changes in weight in obese patients who received long-acting octreotide (octreotide LAR) at one of three dose levels (20, 40, or 60 mg) or placebo over 6 months and to identify the lowest dose of octreotide LAR that safely achieved optimal weight loss.Randomized, double-blind, placebo-controlled trial of octreotide LAR at three dose levels.A total of 172 adults (28 men and 144 women) with at least moderate obesity (body mass index (BMI) range 30-65 kg/m2) and evidence of insulin hypersecretion were enrolled. Patients were predominantly either Caucasian (50.0%) or African American (45.3%). The mean age (38 +/- 11 year), weight (110.7 +/- 23 kg), and BMI (39.8 +/- 6.5 kg/m2) were similar across the four treatment groups.Efficacy measures included weight, BMI, fasting serum glucose; triglycerides; percentage of total body fat and abdominal fat as measured by dual-energy X-ray absorptiometry; skin fold thickness; waist-to-hip circumference; leptin; percentage of carbohydrates, fat, and protein ingested; nutritional evaluation (including dietary analysis--3-day food record); quality of life (QoL; using the Impact of Weight on Quality of Life-Lite); Beck Depression Inventory; and Carbohydrate Craving Questionnaire. Safety measures included medical history, vital signs, physical examinations, hematology, blood chemistries, thyroid function tests, hemoglobin A1c, gallbladder ultrasound, electrocardiograms, and adverse events.After 6 months of treatment, patients receiving 40 or 60 mg of octreotide LAR experienced statistically significant weight loss compared to baseline, with mean differences from placebo in percent weight change of -1.98 and -1.87%, respectively. This finding was accompanied by statistically significant mean decreases in BMI compared to baseline, that is, a mean decrease of 0.73 and 0.79 kg/m2 for the 40 and 60 mg treatment arms, respectively. The observed weight loss was progressive during the 6-month treatment in the two higher dose groups. The lowest dose to reach statistical significance in weight loss after 6 months' treatment was 40 mg. Post hoc analysis revealed a 3.5-3.8% weight loss at month 6 in the two higher dose groups among Caucasian patients having insulin secretion greater than the median of the cohort, defined as CIR(gp) (corrected insulin response at the glucose peak)or = 1.43. There were no statistically significant changes in QoL scores, body fat, leptin concentration, Beck Depression Inventory, or macronutrient intake. Mean changes of blood glucose AUC(0-180 min) during an oral glucose tolerance test in patients taking octreotide LAR were 39-40 mg/dl h higher than those on placebo. A total of 7-21% of the patients taking octreotide LAR reached a 5% or greater decrease in body weight from Baseline, compared to 11% for the placebo group. This was not statistically significant. The most common adverse events included diarrhea, headache, cholelithiasis, nausea, and abdominal pain.Octreotide LAR given at 40 or 60 mg resulted in statistically significant weight loss. A post hoc analysis stratifying patients by race and CIR(gp) indicated that Caucasian patients with the greater degree of insulin hypersecretion appeared to derive the most benefit from treatment. The observed safety profile was consistent with the known effects of octreotide from previous studies.

Electrodes modified with Nafion films containing 2,7-dimethyldiazapyrenium (DAP2+) were prepared and characterized with voltammetry by themselves and in the presence of organic substrates. The large, planar, electron-poor aromatic surface in DAP2+ facilitates pi-stacking interactions with other planar aromatic molecules, particularly those that are negatively charged or electron-rich. Previous studies showed that the reduction of DAP2+ decreases the strength of these interactions, making the binding redox-dependent, and resulting in negative shifts in the E(1/2) of DAP2+/+. This study shows that the redox-dependent binding ability of DAP2+ is retained in Nafion, but the selectivity is considerably different. Most significantly, the electron-rich, neutral aromatic compounds that produced small shifts in the E(1/2) of DAP2+/+ in solution cause much larger shifts, up to -110 mV, with the modified electrodes. With indole as a substrate, Nernstian behavior is observed (-60 mV shift per log[indole]) between 10 and 0.5 mM.

Oxidation or reduction of organic redox couples typically leads to large changes in acidity or basicity, with the result that proton transfer often accompanies electron transfer, particularly in aqueous solution. In less polar organic solvents, H-bonding also can play an important role. While it is generally appreciated that proton transfer will have a greater effect on the overall reaction than H-bonding, it is not always straight forward to distinguish between the two, and, despite a considerable amount of research, a complete, quantitative understanding of the relative roles that the two play in the voltammetry that is observed upon addition of acids or bases to organic redox couples in non-aqueous solution remains elusive. Most of the research in this area has been done with quinones in the presence of H-donors. In contrast, this study focuses on p-tetramethylphenylenediamine, H2PD, in the presence of H-acceptors. While quinones undergo two successive reductions to form the increasingly basic radical anion, then dianion, phenylenediamines, which are weakly basic to begin with, undergo two successive oxidations to form an increasingly acidic radical cation, then quinoidal dication. Upon addition of DMF, a weak H-acceptor, to H2PD in acetonitrile, small negative shifts in potential of the second oxidation are observed in the cyclic voltammetry (CV), with little change in wave shape. This is consistent with H-bonding of the DMF to the quinoidal dication, H2PD2+. Somewhat similar behavior is observed when slightly more basic guests are added such as cyanopyridine or trifluoromethylpyridine, but, unlike with DMF, with just 1 equivalent pyridine guest, a significant shift in the potential of the second oxidation is observed, followed by smaller shifts with additional equivalents. We believe that this behavior signals proton transfer between the pyridine, pyr, and the H2PD2+, so that the overall reaction occurring in the second oxidation corresponds to H2PD+ + pyr = HPD+ + Hpyr+ + e- . In this case, the observed E1/2 should depend on the pKa of the Hpyr+. To test this hypothesis, the voltammetry of H2PD is currently being studied with different pyridines that cover a range of pKa values. If correct, the explanation for the continued shift in potential with increasing concentrations of pyridine is well-accounted for simply by applying the Nernst equation to the overall reaction. However, while proton transfer can explain the potentials of the CV waves in the presence of added pyridine, simulations of the voltammetry show that proton transfer by itself cannot explain the observed reversibility of the second oxidation wave in the presence of increasing amounts of added pyridine. This is where H-bonding can play a role. By including H-bonding steps, and allowing electron transfer to occur through the H-bond complex formed between H2PD2+ and pyr (the intermediate in the proton transfer), the simulations can nicely explain both the observed potential shifts and the reversibility of the waves.

In previous work in our lab, the redox-dependent binding behavior of a phenylenediamine urea, UHH, was investigated in the presence of different guests in CH2Cl2. The urea functionality contains two good H-donors in the two urea NH bonds (a DD motif) that are capable of H-bonding with two appropriately spaced H-acceptors (AA motif). Examples are the cyclic diamide PZD (see Figure) and 1,8-naphthyridine, naph. 1H NMR studies indicated a modest Kbinding of 100 M-1 with PZD and 30 M-1 with naph in CH2Cl2. The expectation was that oxidation of the phenylenediamine couple in UHH to the radical cation would increase the acidity of one of the NH’s leading to stronger H-bonding with the guests and a negative shift in the observed redox potential upon addition of these guests. A small negative shift was observed with PZD, but further research indicated that the actual overall reaction occurring upon oxidation of UHH was not the simple 1 electron oxidation to the radical cation expected, but rather 2 electron oxidation of one UHH accompanied by proton transfer to a reduced UHH to give UH+ and HUHH+ as shown in the Figure. Addition of PZD does not change this overall reaction. An alternative explanation of the potential shift is that PZD is actually H-bonding to the protonated HUHH+, which would also be a stronger H-donor. Interestingly, similar magnitude shifts are observed upon oxidation of the ferrocene-containing urea, FcUHH. Since Fc is only capable of being oxidized by 1 electron, oxidation of the Fc also makes a +1 charge, which appears to have a similar effect on the H-bonding ability of the urea as does protonation. In contrast to the PZD, addition of naph causes the current for the oxidation of UHH to increase with little change in peak potential. This can be explained by proton transfer from the UHH radical cation to the naphthyridine. The resulting neutral radical will be immediately oxidized by a second electron leading to the increase in current. The proton transfer will change the H-bonding motif from AA-DD to AD-DA, which is inherently weaker because of unfavorable secondary interactions. In addition, because of the two electron oxidation, both binding partners in the resulting complex, (naphH+)(UH+), will be positively charged. Therefore it is not surprising that the H-bonding does not strengthen upon oxidation. It will be of interest to compare this behavior to that of FcUHH and naph. Cyclic voltammetry experiments will be run with FcUHH and naphthyridine to see if eliminating the possibility of the second electron transfer prevents proton transfer and results in the expected stronger H-bonding to the oxidized form. Figure 1

An important goal in supramolecular chemistry is the development of stimuli-responsive systems in which the strength of the intermolecular interactions can be altered by external signals such as changes in light, temperature, pH or voltage of an electrode. Our group is exploring the latter possibility, primarily by trying to develop systems in which electron transfer perturbs the strength of H-bonding interactions between molecules. The underlying principle is straightforward: an oxidation that decreases the negative charge on a H-acceptor (A) or a reduction that decreases the positive charge on a H-donor (D) will weaken a H-bond. Alternatively, reduction that increases the negative charge on a H-acceptor or oxidation that increases the positive charge on a H-donor will increase the strength of a H-bond. However, in the latter case, it is possible that oxidation or reduction could also lead to full proton transfer. If this occurs across the H-bond, the primary H-bonds will remain, but the secondary H-bonds will change. This can lead to an increase in unfavorable secondary interactions, which would counteract the effect of the initial proton transfer. However, with proper design, proton transfer could lead to an increase in favorable secondary interactions, which would enhance the effect of initial transfer. The goal of this project is to do the latter. In particular, 3-H-bond DAD-ADA arrays will be studied in which the DAD component contains a N-alkyl-4,4'-bipyridinium or “monoquat” redox couple (see figure). DAD-ADA arrays have relatively weak association constants (102 in non-competitive solvents such as CHCl3) due to the three favorable primary H-bonds (shown as solid double-headed arrows in the figure) being counterbalanced by four unfavorable secondary interactions (cross interactions shown as dashed double-headed arrows). As shown in the figure, reduction of the monoquat by one electron should lead to proton transfer, creating not only a strong ionic primary H-bond, but converting all the secondary H-bonds to favorable ones. Examples of similar ionic DDD-AAA complexes have been reported in the literature with association constants greater than 109, suggesting it may be possible see a binding enhancement on the order of 107 upon reduction in this system. Initial voltammetric studies with just monoquat itself (x = nothing) and the diimide ADA guest will be reported in this presentation, along with our synthetic progress in making the complete DAD array, which will have X = NC(=O)R to make the H-donor a strongly donating amide NH. Figure 1

The design of stimuli-responsive systems is an important topic in supramolecular chemistry. These are systems in which an external signal, such as a change in light, temperature, pH or voltage on an electrode, results in structural changes in the supramolecular assembly. Applications include self-healing polymers and gels, controlled release of entrapped molecules and smart materials. Over the years our group has investigated a number of simple systems in which electron transfer is used to perturb the strength of H-bonding interactions. More recently we have expanded the research to include larger H-bond arrays. In particular we have been investigating several different electroactive ureidopyrimidones or UPy’s, which are capable of strong self-dimerization through the formation of 4 linear H-bonds. Non-electroactive UPy’s have already been widely used in the construction of supramolecular polymers and gels capable of self-repair; introduction of electroactive UPy’s in which dimerization can be controlled electrochemically will make even more types of applications possible. These systems are also of interest from a fundamental standpoint in that they provide a well-understood platform in which to investigate the role of intermolecular H-bonding (and proton transfer) on electron transfer. The principle behind redox-responsive H-bonding is straightforward: an oxidation that decreases the negative charge on a H-acceptor (A) or reduction that decreases the positive charge on a H-donor (D) will weaken a H-bond. Alternatively, reduction that increases the negative charge on a H-acceptor or oxidation that increases the positive charge on a H-donor will increase the strength of a H-bond. However, in the latter case, it is possible that oxidation or reduction could also lead to full proton transfer. If this occurs across the H-bond, the primary H-bonds will stay the same, but the secondary H-bonds will change, possibly in a way that has an undesired effect. It is also possible that proton transfer occurs between non-H-bonded sites. This can lead to a H-bonding pattern that is no longer self-complementary. An example of this is found with a UPy we have made that contains a pyridinium redox couple attached to the pyrimidine ring in the UPy, UPyH(MeP+). This UPy was expected and confirmed by NMR to exhibit relatively weak dimerization in CH2Cl2 in the oxidized state due to the electrostatic repulsion and a weaker DADA H-bond motif. Dimerization strength was expected to increase upon reduction to the uncharged radical due to the loss of electrostatic repulsion and increased basicity of the H-acceptor sites. However, the voltammetry observed in CH2Cl2 is not consistent with this result. Furthermore, very similar voltammetry is observed in CH3CN, even though NMR shows only the monomer is present at the same concentrations in this solvent. Studies with model compounds strongly suggest that the reason for the failure to dimerize upon reduction is that the initial reduction product, an uncharged radical, is basic enough to be protonated by the oxidized starting material, resulting in a H-bonding motif that is no longer self-complementary. While this result negates the original goal of strengthening dimerization through reduction, it also opens up the intriguing possibility of switching binding partners by including a guest molecule that is not complementary to the oxidized UPyH(MeP+) form, but is complementary to the reduced, protonated UPyH2(MeP)+ form.

Ureidopyrimidinone (UPy) derivatives, introduced by Meijer and co-workers, are capable of forming dimers linked by four strong H-bonds. They have been widely used as polymer crosslinkers in self-healing supramolecular polymers and gels. While their donor-donor-acceptor-acceptor (DDAA) H-bond motif contributes to polymer stretchiness, their relatively high association constant of 106 also contributes to their ability to autonomously self-repair. This presentation shows the result of substituting a UPy with a N,N- dimethyl-p-phenylenediamine and therefore adding a redox reactive dimension to self-repair. In particular, this work involves the electrochemical investigation of UPyH (shown below) with a series of different bases, specifically pyridine and substituted pyridines. The focus is on finding the appropriate electroinactive base that facilitates both the dimer break up upon oxidation and proton transfer and dimer reformation by reduction and back proton transfer. Cyclic voltammetry (CV) scans of UPyH only, conducted in CH2Cl2 at slower scan rates, show two reversible redox waves. The first wave is of one electron height, corresponding to a two-electron transfer per dimer. The second oxidation wave at more positive potential is one half the height of the first and less reversible than the first. At a faster scan rate, the second wave becomes irreversible and a new reduction wave appears at a more negative potential than the reduction wave for the first electron transfer process. This type of CV return wave behavior for UPyH at faster scan rates was previously seen in another dimethylphenylurea redox active compound with only two H-bonding sites, UHH. In the UHH investigation, CV and spectroelectrochemical results pointed to the dimethylamino moiety being basic enough to abstract an N-H proton in the radical cation oxidation state and therefore allowing for a two electron transfer wave to give the quinoidal cation. A similar reduction wave negative of the first redox wave was seen in acetonitrile at all scan rates, while in CH2Cl2 this wave was seen only at faster scan rates. At slower scans the first oxidation remained reversible. Based upon the UHH study, interpretation of the UPyH second oxidation wave is that it correspond with a second e- transfer per redox center. Each urea-phenylenediamine N-H site in the dimer then becomes acidic enough to transfer its proton to the carbonyl oxygen on the UPyH isocytosine. This would then change the favorable binding motif of DDAA to a weaker binding motif of ADAD. In addition, a 2+ charge forms on each monomer in the dimer contributing to strong repulsive forces. This results in dimer break-up, where upon the protonated isocytosine carbonyl on each monomer is exposed to the more basic dimethylamino site on a fully reduced UPyH dimer. Proton transfer from the isocytosine to the dimethyl amino renders the fully reduced dimer electroinactive, explaining the half wave height. Thus, the overall reaction, shown in the accompanying scheme, corresponds to break up of the half the starting dimers, with two electron oxidation and deprotonation of each monomer to give two UPy+. However, the other half of the UPy’s remain dimerized in the protonated form as (HUPyH+)2. Further investigation focuses on the addition of a series of substituted pyridines as a base and their effect on dimer break up and reformation as the added base competes with the dimethylamino moiety on the phenylenediamine. The goal is to find an external base that will allow full oxidation and break up of all the UPyH dimers, while retaining the chemical reversibility of the process. Figure 1

At one extreme, supramolecular polymers can have a high degree of randomness, existing as entangled coils with the mechanical properties of plastics and elastomers. At the other extreme, supramolecular polymers can be formed through directional intermolecular interactions among designed molecular subunits to achieve dynamic behavior as well as high degrees of internal order. The strength of these intermolecular interactions can be altered via external signals such as the absorption of light, temperature, pH, or the voltage of an electrode. Such is one of the important goals in supramolecular chemistry: the development of stimuli-responsive systems where the strength of intermolecular interactions can be perturbed through the use of external signals. Applications of supramolecular chemistry include self-healing polymers and gels, controlled release of entrapped molecules, and smart materials. In this presentation the results of electrochemical studies with a ferrocene-containing ureidopyrimidone (UPy) will be described. The UPy system – introduced by Meijer – can form two self-complementary 4 hydrogen bond arrays (AADD or ADAD) which dimerize in relatively non-polar organic solvents. By attaching an electroactive substituent such as ferrocene, the strength of electrostatic interactions can be perturbed to create a redox-responsive ureidopyrimidinone. Following earlier studies by Graham and co-workers, a ferrocene-functionalized ureidopyrimidinone UPy(Fc) was prepared. Concentration and scan-rate-dependent cyclic voltammetric studies indicate a square scheme type mechanism in dichloromethane. In the starting UPy(Fc) state, the AADD dimer is preferred. Oxidation to UPy(Fc+) results in conversion to another form with a less positive E1/2. Preference for this new form increases at lower concentrations which suggests that UPy(Fc+) exists as a monomeric species. This would indicate that oxidation breaks the dimer apart as desired. Alternatively it is possible that the observed concentration dependence results from rate-limiting dissociation of the dimer, but the monomer then tautomerizes to the ADAD enol form and re-dimerizes. To distinguish between these possibilities the relative diffusion coefficients (D’s) of UPy(Fc+) to UPy(Fc) were measured by steady-state voltammetry at a microdisk electrode. They were then compared to the relative D’s of ferrocenium to ferrocene. If UPy(Fc) is a dimer in both oxidation states, the DUPy(Fc+)/DUPy(Fc) ratio should be similar to that observed for ferrocene (DFc+/DFc), which will be a monomer in both oxidation states. However, if UPy(Fc+) is a monomer, the DUPy(Fc+)/DUPy(Fc) ratio will be larger than DFc+/DFc, since the D of the UPy(Fc+) monomer will be relatively smaller than the D of the UPy(Fc) dimer due to its smaller size. The latter is observed, supporting the conclusion that oxidation breaks apart the UPy(Fc) dimer at mM concentrations in dichloromethane. The results of these studies support the conclusion that UPy(Fc+) is monomeric at milli-molar concentrations in dichloromethane, meaning that oxidation of the UPy(Fc) breaks apart the dimer as hypothesized. Furthermore the process appears to be completely reversible. CV simulations were performed to determine if the square mechanism can quantitatively explain the observed voltammetry. The resulting simulated CV data supported the use of a square scheme mechanism to explain the cyclic voltammetry. However, the mechanism only accounts for the predominant keto AADD tautomer and excludes the enol ADAD tautomer. HNMR studies mimicking experimental conditions in cyclic voltammetry have been performed, revealing the presence of the enol ADAD tautomer in dichloromethane. Further HNMR studies in dichloromethane are underway, along with a comparative study in chloroform which does not appear to contain the enol ADAD tautomer. Cyclic voltammetry of the Upy(Fc) system will also be performed in chloroform. The results of these experiments will determine if the enol ADAD contributes to the redox mechanism of UPy(Fc). In addition, the synthesis of covalently linked UPy(Fc) oligomers is being initiated with the goal of creating a redox-responsive supramolecular polymer.