AmericanChemicalSociety.com

Progress in this year of the grant has focused on synthesis and characterization of radical cations 1^•+and 2^•+ and initial studies of their reactivity with hydrogen atom donors. Several sources in the chemical literature state that heterocyclic nitrogen radical cations rapidly decompose, as the nitrogen can act as a nucleophile.  For this grant, we have synthesized one radical cation 1^•+ that is stable in solution for at least an half an hour, perhaps longer.  Another radical cation 2^•+ appears to be stable on an electrochemical timescale, and we have made progress toward its synthesis.  Cyclic voltammetry, reactivity studies and computational studies preliminarily suggest that these agents can act as hydrogen atom abstraction agents with a concerted proton-electron transfer (CPET) mechanism.  These systems parallel important biological systems such as flavins and DNA and provide a better understanding of nitrogen radical chemistry.  Such compounds may have use in hydrocarbon functionalization and in artificial photosynthesis.

Student researchers Crina Sasaran, Victoria Polito and Jonathan Geruntho synthesized the precursors for radical cations (1 and 2) using procedures directly or modified from the literature.  Cyclic voltammetry for both compounds in 0.1 M n-Bu₄NPF₆ in acetonitrile affords oxidation and reduction waves, indicating clean electrochemical oxidations on the timescale of the scans.  The reduction potentials for 1^•+/0 and 2^•+/0 are 0.32 V and 0.87 V, respectively.  These potentials are similar to parallel compounds (3, 0.32 V¹ and 4, 0.67 V²).  The more positive potential of 2^•+/0 can be explained by the electron deficient nature of the pyridine ring.  On the electrochemical timescale, the oxidized species are stable.

Figure 1. Synthesis of 3-(2-pyridyl)-10-methylphenothiazium radical cation.

Figure 2. Synthesis of N,N-di-(4-bromophenyl)-N-(4-bromo-2'-pyridyl)aminium radical cation.

Brittney Tiley and Robert Richards performed one-electron oxidation reactions and initial reactivity studies.  Four different oxidizing agents were attempted to generate 1^•+ chemically: NOBF₄, thianthrenium, AgPF₆, tris(4-bromophenyl)aminium hexachloroantimonate in acetonitrile or methylene chloride.  Both the silver and aminium oxidants produced species with distinct absorptions in the visible spectrum (779 and 864 nm; further work is required to confirm the molar absorptivities) that parallel those in 10-methylphenothiazinium radical cation (780 and 870 nm).³   In addition, these reactions gave broad ¹H NMR spectra in CD₃CN. 

The reduction potential for 2^•+/0 indicates that only NOBF₄ is oxidizing enough out of the oxidants above to produce the radical cation from 2.    Reaction led to shifted but distinct NMR peaks and a product with an irreversible voltammogram.  Likely, the NO⁺ ion reacts with pyridyl rings for 1 and 2.  Two possible directions for compound 2 include reaction with SbCl₅, a common oxidant for aminium generation, or alteration of the substrate.  Substitution of bromine atoms with methoxy groups is synthetically possible.  This change will lower the reduction potential and hence allow for more possible oxidants. 

Radical cations can accept an electron, and pyridine can accept a proton.  Our hypothesis is that the combination of these two components will allow for the net transfer of a hydrogen atom by CPET.  Hence, we have explored the reactivity of 1^•+ and 2^•+ with two hydrogen atom donors, 9,10-dihydroanthracene (DHA) and 1,4-cyclohexadiene (CHD), in solution with UV-vis and NMR for 1^•+ and with cyclic voltammetry for both compounds.

In preliminary cyclic voltammetry experiments with 2 and DHA in solution, the current for reduction wave is reduced relative to that of the oxidation wave, suggesting an electrochemical-chemical (EC) mechanism in which the chemical mechanism is bimolecular.  Further studies in which DHA concentrations and scan rates are varied and the voltammograms are modeled may be able to afford a bimolecular rate constant for this reaction.  Changes in the voltammograms with 1 and DHA or CHD in solution are less apparent.

Addition of DHA or CHD to 2^•+ in acetonitrile leads to the decay of the absorptions in the visible region of the spectrum.  In reactions with DHA and CHD, peaks corresponding to anthracene and benzene, respectively, appear in ¹H NMR.  Planned control reactions with 3^•+ are required to establish that the aromatic products are not from one-electron oxidation.  Better control of quantities will establish the stoichiometry of reaction.

Figure 3. UV-vis scans for reaction of 2^•+ and CHD.

Computational studies by Jeffrey Wolbach and Samantha Cordisco suggest that these reactions occur with a CPET mechanism.⁴  The reactants and products for CPET mechanistic step for reaction between 1^•+ and CHD have been calculated.  The transition state has been identified and verified both with imaginary frequency and intrinsic reaction coordinate (IRC) calculations.   The barrier height is approximately 13 kcal/mol in the forward direction and 26.5 kcal/mol in the reverse direction.  Similar calculations with DHA are under way.  Some intermediates for stepwise proton transfer or electron transfer have also been calculated and tentatively indicate that the concerted mechanism has a much lower barrier for reaction.

Figure 4.  Calculated transition states for 1^•+ with DHA (left) and CHD (right).

In the next year of the grant, we plan to chemically generate 2^•+ or a similar aminium compound and continue reactivity studies with hydrogen atom donors for all compounds under study.  Kinetic studies will reveal the rate law and barriers for these reactions.  Kinetic isotope effect studies will determine if the hydrogen atom is involved in the mechanism, and computational studies will parallel the experiments. 

(1)   Rhile, I. J.; Markle,T. F.; Nagao, H.; DiPasquale, A. G.; Lam, O. P.;  Lockwood, M. A.; Rotter, K.; Mayer, J. M. J Am. Chem. Soc. 2006, 128, 6075-6088.

(2)   Connelly, N. G.; Geiger, W. E.  Chem Rev. 1996, 96, 877–910.

(3)   Rosokha , S. V.; Kochi, J. K. J. Am. Chem. Soc. 2007, 129, 3683-3697.

(4)   Geometries were calculated with B972/6-31+g(d,p) with  thermal corrections with B972/6-31+g(d,p) in the gas phase.  Energies for the stationary points were calculated with CAM-B3LYP/6-31++g(2d,p).  Transition states were calculated with ONIOM (CAM-B3LYP/6-31++g(2d,p):B972/3-21+g), and the frequencies and IRC were calculated with ONIOM (CAM-B3LYP/6-31++g(2d,p):B972/3-21+g).  The energies, transition states, frequencies and IRC were calculated with implicit solvation in CH₃CN.