DOI: 10.1002/cphc.201700560
Articles
Using Voltammetry to Measure the Relative HydrogenBonding Strengths of Pyridine and Its Derivatives in Acetonitrile Malcolm E. Tessensohn, Yu Rong Koh, Sihao Lim, Hajime Hirao, and Richard D. Webster*[a] The voltammetric behavior of 2,3,5,6-tetramethyl-1,4-phenylenediamine was found to be able to differentiate the hydrogen acceptor abilities of electroinactive pyridine compounds in acetonitrile. Weak and strong hydrogen acceptors were distinguished through the onset of a third oxidation process that came about at sub-stoichiometric amounts of strong hydrogen acceptors, but not in the presence of weak hydrogen acceptors. This additional oxidation reaction occurred at a potential between the two 1 e@-oxidation reactions that phenylenediamines typically undergo (i.e. EPox(1) < EPox(3) < EPox(2), with EPox(1) and EPox(2) representing the electrochemical conversion of the neutral phenylenediamine into the radical cation and thereafter to the quinonediimine dication) as well as at the expense
of the second electrochemical reaction. EPox(2) and EPox(3) were observed to shift towards less positive potentials with increasing concentrations of weak or strong hydrogen acceptors, respectively, whereas EPox(1) remained virtually unaffected. This allows the electrochemical parameters DEPox(1, 2) = j EPox(1)@EPox(2) j and DEPox(1, 3) = j EPox(1)@EPox(3) j to be employed as measures of the hydrogen-bonding strengths within each category, to which they were found to be highly reproducible and responsive to steric, electronic, inductive, and mesomeric effects. The electrochemical findings concur with available aqueous pKa data of the protonated pyridine compounds but were, however, in poor agreement with results obtained by density functional theory calculations.
1. Introduction Voltammetry has been found to be exceptionally sensitive for studying hydrogen-bonding interactions because the hydrogen bond contains some electrostatic character and so can be electrochemically perturbed. This electrochemical technique is particularly advantageous for it only requires one component of the hydrogen bond (e.g. hydrogen donor or hydrogen acceptor) to be electroactive to qualify for investigation and suffers from lesser interference as compared to infrared and nuclear magnetic resonance spectroscopic methods. In addition, it has been reported to be more sensitive than UV/Vis spectroscopy[1] and concur with the results obtained via theoretical calculations.[2] The experimental electrochemical results have even been thought to be more accurate than theoretical calculations because the matrix effects and other important interactions (e.g. hydrogen bonding with trace water in the solvent and ion-pairing with the supporting electrolyte) are taken into consideration. Phenylenediamines have been widely used as dyes,[3] antioxidants in rubber and fuels,[4] organic materials in spin electron[a] Dr. M. E. Tessensohn, Y. R. Koh, S. Lim, Prof. Dr. H. Hirao, Prof. Dr. R. D. Webster Division of Chemistry and Biological Chemistry School of Physical and Mathematical Sciences Nanyang Technological University 21 Nanyang Link, Singapore 637371 (Singapore) E-mail: [email protected] Supporting Information and the ORCID identification number(s) for the author(s) of this article can be found under: https://doi.org/10.1002/cphc.201700560.
ChemPhysChem 2017, 18, 2250 – 2257
ic[5] and electrochromic[6] devices and in the detection of ammonia,[7] chlorine,[8] paracetamol[9] and metal cations,[10] all of which are based on their electroactivity. In acetonitrile (CH3CN),[11] phenylenediamines generally undergo two chemically reversible 1 e@-oxidation reactions with the neutral compound (P) first oxidized to the radical cation (P· + ) at potential EPox(1) and thereafter into the quinonediimine dication (P2 + ) at a more positive potential EPox(2) (Scheme 1 and Figure 1).[6, 12]
Scheme 1. Oxidation mechanism of phenylenediamines in acetonitrile.
The neutral compound P largely behaves as a weak base and so preferentially functions as a hydrogen acceptor through the lone pairs of electrons on its nitrogen atoms, but can also serve as a hydrogen donor via the hydrogen atoms of the amino groups. The oxidized forms P· + and P2 + , on the other hand, tend to behave as hydrogen donors with the latter being more acidic and a stronger donor because of its greater positive charge. We have previously applied the electrochemical reduction of quinones for measuring the hydrogen donor abilities of alcohols and diols in aprotic organic solvents[13] and now report
2250
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles ceptors) that is known to be highly accurate and reproducible. DE is unaffected by small changes between experiments in the effects of uncompensated solution resistance, heterogeneous electron-transfer rates (ket), surface condition of the working electrode and independent of drifts in potential of the reference electrode and so only responds to hydrogen bonding between 2,3,5,6-tetramethyl-1,4-phenylenediamine and the pyridine compounds.[13] However, systems that experience slow heterogeneous electron transfer, undergo adsorption, or are studied under highly resistant conditions that result in distorted voltammograms are not amendable for correlating DE with hydrogen bonding effects. DE, either as the difference in potentials within a single molecule or between two molecules, has been used in the detection of oxygen[15] and measurements of moisture content,[1, 16] temperature[15, 17] and humidity.[17a]
Figure 1. Cyclic voltammograms of 1.0 mm 2,3,5,6-tetramethyl-1,4-phenylenediamine with 0.1 m nBu4NPF6 in CH3CN recorded at a scan rate of 0.1 V s@1 at 25 : 2 8C with a 1 mm diameter planar disk glassy carbon working electrode and trace moisture content of 17.2 mm.
the use of the electrochemical oxidation of 2,3,5,6-tetramethyl1,4-phenylenediamine[14] for determining the hydrogen acceptor abilities of pyridine compounds (Figure 2) in CH3CN. These qualitative measurements are based on the electrochemical parameter DE (i.e. DEPox(1, 2) = j EPox(1)@EPox(2) j for weak hydrogen acceptors and DEPox(1, 3) = j EPox(1)@EPox(3) j for strong hydrogen ac-
Figure 2. Pyridine compounds examined in this study.
2. Results and Discussion The effect of pyridine, 2,6-lutidine and 2,6-di-tert-butylpyridine on the voltammetric behavior of 2,3,5,6-tetramethyl-1,4-phenylenediamine were first investigated and the resulting cyclic voltammograms are shown in Figure 3. Although the first electron-transfer process (oxidation of P to P· + ) was not visibly affected by the presence of pyridine and 2,6-lutidine (Figures 3 a and 3b, respectively), the second electrochemical reaction (oxidation of P· + to P2 + ) was found to be significantly influenced and absent in the cyclic voltammograms of the phenylenediamine when their concentrations were 1.0 mm and greater. A third oxidation process, initially detected at sub-stoichiometric amounts (< 1.0 mm) of the hydrogen acceptors, was also noticed to occur at 250 mV negative of the second oxidation reaction and at its expense. This third process grew in height
Figure 3. Cyclic (solid lines) and square-wave (dashed lines) voltammograms of 1.0 mm 2,3,5,6-tetramethyl-1,4-phenylenediamine with 0.1 m nBu4NPF6 in CH3CN at different concentrations of a) pyridine, b) 2,6-lutidine and c) 2,6-di-tert-butylpyridine recorded at a scan rate of 0.1 V s@1 at 25 : 2 8C with a 1 mm diameter planar disk glassy carbon working electrode. The currents of the square-wave voltammograms were scaled by a factor of 0.5. The initial voltammograms obtained before the addition of the compounds are indicated in red together with their measured trace moisture content.
ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
2251
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles (i.e. peak current (iPox(3)) increased) and shifted to less positive potentials with increasing concentrations of the hydrogen acceptors. It subsequently merged with the first electron-transfer reaction at & 200 mm of pyridine and & 10 mm of 2,6-lutidine to give a single oxidation process that was chemically reversible, approximately two times the height of the first oxidation process (i.e. iPox(merged) & 2 V iPox(1)) and equal to the sum of the initial two 1 e@-oxidation reactions (i.e. iPox(merged) & iPox(1) + iPox(2)) thereby suggesting that it involves the 2 e@-oxidation of P directly to P2 + . This behavior was also observed on a platinum working electrode as well as at faster scan rates of 1 and 10 V s@1 (see the Supporting Information). The chemical reversibility of the oxidation process was, however, compromised at higher concentrations ( & 200 mm) of 2,6-lutidine, probably because of proton-transfer reactions with the quinonediimine dication, but could be regained by either an increase in scan rate or shortening of the potential window as shown in the Supporting Information. Pyridine and 2,6-lutidine, as well as all the other compounds studied, are apparently electroinactive within the applied potential range on the glassy carbon working electrode because no additional processes, other than those assigned to the phenylenediamine, are observed throughout the experiments, even when they are present in a two hundred-fold excess. Cyclic voltammograms of pyridine recorded in the presence and absence of acids are provided in the Supporting Information. Smith and co-workers have suggested that the third oxidation process describes the electrochemical conversion of P· + into P2 + coupled with either strong hydrogen bonding or a proton-transfer reaction between P2 + and pyridine/2,6-lutidine (aqueous pKa values: 1.5 for P2 + ,[18] 5.23 for pyridinium and 6.65 for protonated 2,6-lutidine[19]) and so explains it occurring more easily than in the absence of the additives (i.e. EPox(3) < EPox(2)).[12] The high chemical reversibility of the third oxidation reaction suggests that the rate of proton-transfer between P2 + and pyridine/2,6-lutidine, if it occurs, is sufficiently fast in either direction, which might be facilitated if the resulting quinonediimine cation and pyridinium/protonated 2,6-lutidine are also held together by hydrogen-bonding interactions. It is unlikely that a species more easily oxidizable than P· + is responsible for the third oxidation process because pyridine and 2,6-lutidine are unlikely to deprotonate P· + under the experimental conditions (aqueous pKa of P· + is 12[18]). As for 2,6-di-tert-butylpyridine (Figure 3 c), no additional oxidation processes of the phenylenediamine other than those initially detected and shown in Scheme 1 were observed across the studied concentration range. The first oxidation reaction did not seem to be affected by the presence of 2,6-di-tert-butylpyridine, while the second oxidation reaction was observed to shift towards less positive potentials with increasing concentrations, but it did not merge with the first process even at the highest concentration of 200 mm. Even though a proton-transfer between 2,6-di-tert-butylpyridine (aqueous pKa value of 4.95) and P2 + was thermodynamically feasible, the steric effect of the bulky tert-butyl groups appears to have dominated their own electronic contribution so preventing the reaction from occurring. Thus, the third oxidation process was not observed. ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
The shifts of EPox(2) towards less positive potentials indicate that 2,6-di-tert-butylpyridine is nevertheless hydrogen bonding with the phenylenediamine and its oxidized forms, albeit weaker than those formed by pyridine and 2,6-lutidine. The shifting of the oxidation potentials of 2,3,5,6-tetramethyl-1,4-phenylenediamine, and phenylenediamines in general, towards less positive potentials with extensive hydrogen bonding to the pyridine compounds was expected and indicates that the ionization energies of the individual electron-transfer reactions are lowered as a consequence of the interactions; a similar trend was previously observed for the electrochemical oxidation of phenols and hydroquinone and their corresponding adiabatic ionization energies (AIEs) with hydrogen bonding to water.[20] The ionization energy of P· + was also anticipated to be more susceptible to the effects of hydrogen bonding than the ionization energy of P because the interactions between the pyridine compounds and P2 + are the strongest while those with P are the weakest. Therefore, EPox(2) and EPox(3) are shifted by a greater magnitude than EPox(1) and the related DEP parameters concomitantly decrease. The AIEs of P and P· + at different levels of hydrogen bonding to pyridine were calculated through density functional theory (DFT) and the results show that the changes to the AIE of P· + were almost always two times larger than those for P (Table 1) and so concur with the electrochemical findings. The energies, 3D geometries and Cartesian coordinates of the optimized structures of the various hydrogen-bonded complexes (with n = 1 to 4) can be found in the Supporting Information.
Table 1. DFT-calculated adiabatic ionization energies of the phenylenediamine(pyridine)n hydrogen-bonded complexes in CH3CN.
No. of pyridine molecules (n)
Adiabatic ionization energy [eV] P(pyridine)n P· + (pyridine)n
0 1 2 3 4
4.460 4.345 4.226 4.135 4.046
5.739 5.511 5.317 5.168 5.033
More importantly, these initial results support the premise that the voltammetric behavior of phenylenediamines can be employed to measure the relative hydrogen acceptor abilities of pyridine and its derivatives in CH3CN. The occurrence of the third oxidation process in the voltammetric response of 2,3,5,6-tetramethyl-1,4-phenylenediamine at sub-stoichiometric concentrations of the additives could first be used to differentiate between weak and strong hydrogen acceptors, for which it would be absent in the former and present in the latter. Thereafter, the electrochemical parameters DEPox(1, 2) = j EPox(1)@EPox(2) j and DEPox(1, 3) = j EPox(1)@EPox(3) j would be applied to assess the hydrogen bonding abilities of weak and strong hydrogen acceptors, respectively. Based on these metrics, 2,6-di-tert-butylpyridine would be considered as a weak hydrogen acceptor whereas pyridine and 2,6-lutidine can be regarded as strong hydrogen acceptors. The overall order of their hydrogen ac-
2252
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles ceptor strengths is 2,6-di-tert-butylpyridine < pyridine < 2,6-lutidine, which is also in line with their basicity. The effect of the picolines and trifluoromethylpyridines (CF3Py) on the voltammetric behavior of 2,3,5,6-tetramethyl-1,4phenylenediamine was also investigated and the electrochemical results indicate that the picolines are strong hydrogen acceptors while the trifluoromethylpyridines are weak hydrogen acceptors; their corresponding cyclic voltammograms are provided in the Supporting Information. A plot of DEPox(1, 2) = j EPox(1)@EPox(2) j against concentrations (Figure 4) shows that the Figure 5. Plots of DEPox(1, 3) = j EPox(1)@EPox(3) j against concentrations of the strong hydrogen acceptors as measured by square-wave voltammetry during the oxidation of 1.0 mm 2,3,5,6-tetramethyl-1,4-phenylenediamine with 0.1 m nBu4NPF6 as the supporting electrolyte in CH3CN recorded with a 1 mm diameter planar glassy carbon working electrode at 25 : 2 8C. The scans were carried out with a pulse period (t) of 25 Hz, potential step of 5 mV and pulse amplitude of 20 mV. The plots for 2,6-lutidine and 2-picoline are not included. Legend for inset: (red) 4-picoline, (blue) isoquinoline, (green) pyridine and (pink) 3-picoline.
Figure 4. Plots of DEPox(1, 2) = j EPox(1)@EPox(2) j against concentrations of the weak hydrogen acceptors as measured by square-wave voltammetry during the oxidation of 1.0 mm 2,3,5,6-tetramethyl-1,4-phenylenediamine with 0.1 m nBu4NPF6 as the supporting electrolyte in CH3CN recorded with a 1 mm diameter planar glassy carbon working electrode at 25 : 2 8C. The scans were carried out with a pulse period (t) of 25 Hz, potential step of 5 mV and pulse amplitude of 20 mV.
trifluoromethylpyridines are weaker hydrogen acceptors than 2,6-di-tert-butylpyridine and the strength of their interactions follow the trend 4-CF3-Py > 3-CF3-Py > 2-CF3-Py, indicating that the electron-withdrawing inductive effect and, to some extent, the steric effect of the CF3 group have outweighed the mesomeric stability. 3-CF3-Py would be expected to show the strongest interactions if the mesomeric effect was dominant because its carbon atom that is directly attached to the CF3 group does not carry a partial positive charge in any of its resonance structures, and so its nitrogen atom would be more basic, whereas the same carbon atoms of interest in 2-CF3-Py and 4-CF3-Py do. In the case of the picolines, a comparison of their plots of DEPox(1, 3) = j EPox(1)@EPox(3) j against concentrations (Figure 5) as well as their amounts needed to merge the two oxidation processes revealed that 2-picoline is a stronger hydrogen acceptor than 4-picoline thereby indicating that the electronic effect of the methyl substituent has outweighed its own steric effect. As for 3-picoline, its methyl group is apparently not in a position to inductively stabilize the important mesomeric forms and so exhibits the weakest hydrogen acceptor properties. The electrochemical results have also indicated that 2,6-lutidine is a stronger hydrogen acceptor than 2-picoline, which is viable only if the electronic effects are more significant than the steric effects in governing the strength of the hydrogen bonds. This is supported by DFT calculations where the aromatic rings ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
of the pyridine compounds and the phenylenediamine were found to be orthogonal instead of co-planar to each other (see the Supporting Information) so reducing the steric effect of the methyl groups. This line of reasoning however, does not seem to apply for 2,6-di-tert-butylpyridine probably because the steric repulsion arising between its tertiary butyl groups and the methyl groups of the phenylenediamine are similar regardless of the manner of approach and so their steric effects outweigh their electronic contributions. It should be noted that the best fit line plots for DEPox(1, 2) (Figure 4) and DEPox(1, 3) (Figure 5) against concentrations for most of the pyridine compounds were obtained from three separate yet highly reproducible experiments and so attests that the electrochemical parameters are unaffected by the surface condition of the polished and unmodified glassy carbon working electrode, solution resistance as well as drifts in potential of the reference electrode, which the absolute potentials EPox(1), EPox(2) and EPox(3) are not immune to. The effect of the benzopyridines, quinoline and isoquinoline, on the electrochemical response of 2,3,5,6-tetramethyl-1,4-phenylenediamine were also studied (see the Supporting Information for the cyclic voltammograms), to which both isomers gave a third oxidation process at sub-stoichiometric concentrations and hence can be deemed as strong hydrogen acceptors; the related dibenzopyridines, acridine and phenanthrenedine, were unfortunately insoluble in CH3CN. It is shown in Figure 5, and also reported in the literature,[21] that the hydrogen acceptor ability of quinoline is weaker than isoquinoline since the first and third oxidation processes remained separated at 200 mm of quinoline whereas only a single oxidation process was obtained at the same concentration of isoquinoline. This difference in hydrogen bonding abilities is a consequence of the steric effects from the adjacent ring as well as the peri hydrogen atom that collectively prevents quinoline from coming into closer proximity to 2,3,5,6-tetramethyl-1,4-phenylenediamine.[22]
2253
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles The influence of two heteroatoms as in the case of the 2,2’bipyridine and the diazines (i.e. pyridazine, pyrimidine and pyrazine) on the voltammetric response of 2,3,5,6-tetramethyl-1,4phenylenediamine was also examined. The phenylenediamine underwent a third oxidation process in the presence of substoichiometric amounts of 2,2’-bipyridine (see the Supporting Information) and so it could be viewed as a strong hydrogen acceptor. The electrochemistry experiments however, indicate that 2,2’-bipyridine has weaker hydrogen-bonding abilities than pyridine, thereby suggesting that bidentate interactions with the phenylenediamine, where both nitrogen atoms of 2,2’-bipyridine are simultaneously interacting with a hydrogen atom of the phenylenediamine, are not occurring. This supposition is supported by DFT calculations where only one hydrogen-bonding interaction was detected in the hydrogenbonded complexes irrespective of the oxidation state of the phenylenediamine. It is conceivable that the electron-withdrawing character and action of one pyridine ring on the other may have reduced the overall hydrogen acceptor ability of 2,2’-bipyridine and so debilitate its propensity to participate in bidentate hydrogen bonding. The electrochemical results (see Figure 4 and Figure 5 as well as cyclic voltammograms that are included in the Supporting Information) indicate that pyridazine can be considered as a strong hydrogen acceptor while pyrimidine and pyrazine may be regarded as weak hydrogen acceptors. Pyridazine was however weaker than pyridine because the additional nitrogen atom exerts an electron-withdrawing inductive effect on the other nitrogen atom and so its overall binding ability is lessened. Among the diazines, pyridazine and pyrazine were the strongest and weakest hydrogen acceptors, respectively, which is in line with their reported basicity (see Table 2 and Table 3). Nevertheless, the observed trend for the diazines cannot be fully explained by mesomeric and inductive effects regardless whether they are independently or jointly considered. Pyrimidine should instead be the strongest if the mesomeric effects were overriding because no partial positive charge would reside on any of its nitrogen atoms. On the other hand, if the inductive effects were dominant then pyrazine would overshadow the others because its two nitrogen atoms are the fur-
Table 2. Relative binding strengths (in kJ mol@1) of the strong hydrogen acceptors with 2,3,5,6-tetramethyl-1,4-phenylenediamine in CH3CN.
Compound 2,6-Lutidine 2-Picoline 4-Picoline Isoquinoline 3-Picoline Pyridine 2,2’-Bipyridine Quinoline Pyridazine
P
Rel. hydrogen bond strength P· + P2 +
@1.66 @3.57 @3.69 @4.17 @3.83 @4.17 @12.59 @3.59 @3.87
@11.59 @14.28 @16.41 @15.49 @16.13 @15.19 @19.89 @13.14 @14.99
@30.41 @35.24 @41.12 @36.98 @39.06 @37.16 @34.06 @32.73 @35.14
Aqueous pKa values[19] 6.65 6.00 5.99 5.40 (20 8C) 5.70 5.23 – 4.90 (20 8C) 2.24 (20 8C)
The pKa values were measured at 25 8C unless otherwise indicated.
ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
Table 3. Relative binding strengths (in kJ mol@1) of the weak hydrogen acceptors with 2,3,5,6-tetramethyl-1,4-phenylenediamine in CH3CN.
Compound
Rel. hydrogen bond strength P P· + P2 +
Aqueous pKa values[19]
2,6-Di-tert-butylpyridine 4-CF3-Py Pyrimidine 3-CF3-Py Pyrazine 2-CF3-Py
3.04 @3.40 @3.44 @3.50 @3.34 @2.98
3.58 – 1.23 – 0.65 –
3.88 @13.08 @12.51 @13.11 @11.85 @10.12
7.77 @29.56 @28.27 @29.31 @25.99 @22.32
The pKa values were measured at 20 8C.
thest apart. Pyridazine was however determined to be the strongest hydrogen acceptor for the reason that hydrogen bonding with the phenylenediamine alleviates the electronic repulsion between the two lone electron pairs of the adjacent nitrogen atoms. Even so, the electronic repulsion between the lone electron pairs was insufficient to outweigh the electronwithdrawing inductive effect of the adjacent nitrogen atom and so pyridazine was the weakest among the compounds sorted as strong hydrogen acceptors. The trifluoromethylpyridines, pyrimidine, pyrazine and 2,6di-tert-butylpyridine, when present in sub-stoichiometric amounts, did not cause 2,3,5,6-tetramethyl-1,4-phenylenediamine to undergo a third oxidation process and so can be regarded as weak hydrogen acceptors. The order of their hydrogen acceptor ability, as determined by the electrochemical parameter DEPox(1, 2) = j EPox(1)@EPox(2) j (Figure 4), was found to follow 2-CF3-Py < pyrazine < 3-CF3-Py < pyrimidine < 4-CF3-Py < 2,6-di-tert-butylpyridine. The remaining compounds, on the other hand, may perhaps be considered as strong hydrogen acceptors because of the occurrence of the third oxidation reaction; their hydrogen acceptor ability was measured by the electrochemical parameter DEPox(1, 3) = j EPox(1)@EPox(3) j (Figure 5) and follows pyridazine < quinoline < 2,2’-bipyridine < pyridine < 3-picoline < isoquinoline < 4-picoline < 2-picoline < 2,6lutidine. The relative hydrogen-bonding strengths of pyridine and its derivatives with 2,3,5,6-tetramethyl-1,4-phenylenediamine were also calculated to which the data for the strong and weak hydrogen acceptors (together with their available pKa data) are reflected in Table 2 and Table 3, respectively; noting that the compounds are arranged in the order of hydrogen acceptor strengths obtained by the electrochemical procedure. The computational results were however, in poor agreement with the electrochemical findings and reported pKa values. For instance, the calculations signal that the binding strengths of the phenylenediamine-2,6-lutidine hydrogen-bonded complexes are the weakest and that formation of the phenylenediamine-2,6-di-tert-butylpyridine hydrogen-bonded complexes is unfavorable. The accuracy of DFT calculations in describing hydrogen bonds has been well scrutinized in the literature and reported to be dependent on, amongst other factors, the bond directionality,[23] exchange-correlation functionals and basis set.[24] In addition, although the solvation effects by
2254
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles CH3CN were taken into consideration in the calculations, the other important interactions such as ion pairing between the charged products and the supporting electrolyte as well as hydrogen bonding between the phenylenediamine, pyridine compounds and trace water in the solvent were disregarded. While effort was put into ensuring that most of the geometries were explored, it might be possible that some conformers were inadvertently left out because of the numerous conformational possibilities and so might explain the disagreement between the computational and experimental results. Nevertheless, the neutral and quinonediimine dicationic forms of the phenylenediamine correspondingly and expectedly formed the weakest and strongest hydrogen-bonding interactions. One advantage of this procedure for evaluating hydrogen acceptor strengths through the oxidation of phenylenediamines, is its high shared sensitivity towards electronic and steric effects as well as mesomeric and inductive stability that are not the same across the different classes of molecules (benzopyridines, diazines, picolines, trifluoromethylpyridines etc.). It is also experimentally simple to perform, fast and useful especially in cases when the theoretical considerations alone are insufficient to reach a decision. Take, for example, a comparison between 2,6-di-tert-butylpyridine and 4-CF3-Py where it is challenging to determine if the steric effect of the tert-butyl groups will dominate the electron-withdrawing inductive effect of the CF3 group, noting that the tert-butyl groups are also imparting an electron-donating inductive influence whereas the CF3 group in 4-CF3-Py does not impose any steric obstruction towards hydrogen bond formation. The electrochemical experiments indicate that the electron-withdrawing character of the CF3 group dominates and so 2,6-di-tert-butylpyridine is the stronger hydrogen acceptor. Lastly, and probably more applicably, the voltammetric method may also serve to measure the relative basicity of compounds in non-aqueous solvents. This is because the energy of a hydrogen bond is reported to correlate with the pKa of hydrogen donors and pKb of hydrogen acceptors in a given solvent and that the formation of a hydrogen-bonded complex is believed to precede proton-transfer reactions.[25] However, the voltammetric method only provides qualitative information about the hydrogen bond and so would require other treatments[26] to obtain quantitative data.
3. Conclusions In this study, the sensitivity of the electrochemical oxidation of 2,3,5,6-tetramethyl-1,4-phenylenediamine towards hydrogen bonding was used to qualitatively measure the hydrogen acceptor abilities of pyridine and its derivatives. The relative hydrogen bonding strengths were assessed by the electrochemical parameters DEPox(1, 2) = j EPox(1)@EPox(2) j for the weaker hydrogen acceptors and DEPox(1, 3) = j EPox(1)@EPox(3) j for stronger ones. These potential measurements are reproducible because the effects of solution resistance, migration, kinetics of the heterogeneous electron-transfer as well as variations in the surface condition of the polished bare unmodified glassy carbon working electrode and potential of the reference electrode are elimChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
inated because the potential measured is between two redox processes within the same compound. The electrochemical parameters respond to steric, electronic, inductive and mesomeric influences as well as number of heteroatoms present and are inversely related to the hydrogen acceptor ability of the compounds. The experimental trends agree with the reported pKa values of the protonated forms but are in poor accord with the DFT results. This voltammetric technique only requires small amounts of material for use, is fast and easy to perform, and the results obtained are not difficult to interpret. It is also applicable for electroinactive hydrogen acceptors and suffers from little interference. Nevertheless, the procedure only provides qualitative information about the hydrogen bond and so has to rely on other methods to acquire quantitative information. Nonetheless, the method may serve to indicate the relative basicity of compounds in non-aqueous solvents. The results presented in this study with a phenylenediamine and previously with quinones have shown that the hydrogen bonding abilities of electroinactive hydrogen donors and acceptors can be voltammetrically determined.
Experimental Section Chemicals The following chemicals were used as received: 2,3,5,6-tetramethyl-1,4-phenylenediamine (Sigma–Aldrich, 99 %), pyridine (Alfa Aesar, 99.5 + %), 2-picoline (Merck, + 98 %), 3-picoline (Alfa Aesar, 99 %), 4-picoline (Alfa Aesar, 98 %), 2-(trifluoromethyl)pyridine (2CF3-Py, Alfa Aesar, 99 %), 3-(trifluoromethyl)pyridine (3-CF3-Py, Alfa Aesar, 97 %), 4-(trifluoromethyl)pyridine (4-CF3-Py, Alfa Aesar, 97 %), 2,6-lutidine (Alfa Aesar, 98 + %), 2,6-di-tert-butylpyridine (Aldrich, + 97 %), quinoline (Alfa Aesar, 97 %), isoquinoline (Alfa Aesar, 98 %), pyridazine (Alfa Aesar, 98 + %), pyrimidine (Alfa Aesar, 99 %), pyrazine (Aldrich, 99 + %), 2,2’-bipyridine (Sigma–Aldrich, + 99 %) and CH3CN (RCI Labscan Limited, AR grade). The supporting electrolyte, tetra-n-butylammonium hexafluorophosphate (nBu4NPF6), was prepared by a literature procedure.[27] 1/16 in. rods with 3 a pore size molecular sieves (CAS: 308 080-99-1) were obtained from Fluka.
Measurement of the Initial Water Content Karl Fischer titrations were performed with a Mettler Toledo DL32 coulometer inside a humidity control box (Coy Laboratory Products) maintained at 30 % relative humidity, with (Riedel-deHa[n) HYDRANAL-Coulomat AG and CG solutions in the anolyte and catholyte compartments, respectively. Stabilization of the coulometer was achieved when the drift value of water was steady and close to 0 mg min@1. Each measurement was found to conclude within 1 min so indicating that the drift from atmospheric water was insignificant.
Voltammetry Cyclic voltammetry and square-wave voltammetry experiments were performed with a computer-controlled Eco Chemie Autolab PGSTAT302N potentiostat in a three-electrode cell with a 1 mm diameter planar glassy carbon disk (Cypress Systems) working electrode, a platinum wire (Metrohm) auxiliary electrode, and a silver
2255
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles wire (Cypress Systems) miniature reference electrode connected to the test solution via a salt bridge containing 0.5 m nBu4NPF6 in CH3CN. The internal filling solution of the liquid-junction reference electrode was replaced with a dried solution at the start of each experiment to reduce the interference of water.
Preparation of Dried Electrolyte Solutions Dried CH3CN was used to make a solution of 0.1 m nBu4NPF6 with the supporting electrolyte predried under vacuum at 160 8C for 6 h. The solution was then quantitatively transferred into a vacuum syringe (SGE Analytical Science) containing 3 a molecular sieves that were also predried under vacuum at 160 8C for 6 h. The syringe was subsequently stored under a nitrogen atmosphere for a minimum period of 48 h prior to further use.
Procedure At the start of each experiment, 15 mL of the dried electrolyte solution was used to make up a 1.0 mm solution of 2,3,5,6-tetramethyl-1,4-phenylenediamine in the humidity control box (Coy Laboratory Products) that was maintained at 30 % relative humidity. The phenylenediamine/electrolyte solution was then transferred into the electrochemical cell that had been previously heated at 100 8C for an hour and cooled under an argon atmosphere to room temperature, before deoxygenating the solution with argon gas. Consecutive cyclic voltammetry and square-wave voltammetry scans were subsequently performed. A 5 mL aliquot of the electrochemical solution was extracted with a vacuum syringe (SGE Analytical Science) and then subjected to Karl Fischer titration to obtain the initial moisture content of the solution. The working electrode was systematically polished on a Buehler polishing pad following which microliter volumes of the compounds from a concentrated stock solution were then carefully and accurately added with a micropipette and mixed with a steady stream of argon bubbles.
Computational Methods The molecular geometries were drawn by using GaussView 5 and optimized with the B3LYP hybrid density functional[28] with the 6– 311 + G(2df,p) Pople-style Gaussian basis set[29] (triple z quality with one set of diffuse s and p functions on heavy atoms, two additional d-type polarization functions and an additional f-type polarization function on heavy atoms, and one set of p-type polarization functions on hydrogen) with solvation effects taken into account within the Gaussian 09 suite of programs.[30] The final geometries were verified to stay at true energy minima through frequency calculations at the B3LYP/6–311 + G(2df,p) level that yielded no imaginary frequencies. The zero-point energies were scaled by a factor of 0.9889.[31] The adiabatic ionization energies were computed based on the differences in ground state energies of the respective oxidized states.[32] For example, the AIE of P was obtained by subtracting the ground state energy of P· + from that of P while the AIE of P· + can be acquired through the difference in ground state energies of P· + and P2 + . The relative hydrogen bond strengths of the pyridine compounds were calculated by subtracting the ground state energies (single point energy plus zero-point correction) of the components from the energy of the hydrogen-bonded complex: relative hydrogen bond strength = energy of hydrogenbonded complex—(energy of 2,3,5,6-tetramethyl-1,4-phenylenediamine + energy of pyridine compound). ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
Acknowledgements This work was supported by Singapore Government MOE Academic Research Fund Tier 1 Grant (RG109/15). M.E.T. thanks Nanyang Technological University for the award of the Nanyang President’s Graduate Scholarship (NPGS). Keywords: basicity · electrochemistry · hydrogen acceptor · phenylenediamine · voltammetry [1] Y. Hui, E. L. K. Chng, C. Y. L. Chng, H. L. Poh, R. D. Webster, J. Am. Chem. Soc. 2009, 131, 1523 – 1534. [2] a) M. Gjmez, F. J. Gonz#lez, I. Gonz#lez, J. Electrochem. Soc. 2003, 150, E527 – E534; b) J. Garza, R. Vargas, M. Gjmez, I. Gonz#lez, F. J. Gonz#lez, J. Phys. Chem. A 2003, 107, 11161 – 11168; c) C. Costentin, M. Robert, J.M. Sav8ant, C. Tard, Angew. Chem. Int. Ed. 2010, 49, 3803 – 3806; Angew. Chem. 2010, 122, 3891 – 3894; d) J.-M. Sav8ant, C. Tard, J. Am. Chem. Soc. 2014, 136, 8907 – 8910. [3] R. D. Theys, G. Sosnovsky, Chem. Rev. 1997, 97, 83 – 132. [4] a) J. Posp&sˇil, Aromatic and Heterocyclic Amines in Polymer Stabilization, Vol. 124, Springer Berlin Heidelberg, 1995; b) F. Cataldo, Polym. Degrad. Stab. 2001, 72, 287 – 296; c) F. Cataldo, Eur. Polym. J. 2002, 38, 885 – 893; d) N. M. Huntink, R. N. Datta, J. W. M. Noordermeer, Rubber Chem. Technol. 2004, 77, 476 – 511. [5] a) A. Ito, M. Urabe, K. Tanaka, Polyhedron 2003, 22, 1829 – 1836; b) A. Ito, Y. Nakano, M. Urabe, T. Kato, K. Tanaka, J. Am. Chem. Soc. 2006, 128, 2948 – 2953; c) A. Ito, R. Kurata, D. Sakamaki, S. Yano, Y. Kono, Y. Nakano, K. Furukawa, T. Kato, K. Tanaka, J. Phys. Chem. A 2013, 117, 12858 – 12867. [6] S. J. L. Lauw, X. Xu, R. D. Webster, ChemPlusChem 2015, 80, 1288 – 1297. [7] X. Ji, R. G. Compton, Anal. Sci. 2007, 23, 1317 – 1320. [8] E. H. Seymour, N. S. Lawrence, R. G. Compton, Electroanalysis 2003, 15, 689 – 694. [9] Y. Peng, Z. Wu, Z. Liu, Anal. Methods 2014, 6, 5673 – 5681. [10] a) A. J. Pearson, J.-J. Hwang, Tetrahedron Lett. 2001, 42, 3541 – 3543; b) J. W. Sibert, P. B. Forshee, Inorg. Chem. 2002, 41, 5928 – 5930; c) J. W. Sibert, P. B. Forshee, V. Lynch, Inorg. Chem. 2005, 44, 8602 – 8609; d) J. W. Sibert, G. R. Hundt, A. L. Sargent, V. Lynch, Tetrahedron 2005, 61, 12350 – 12357; e) J. W. Sibert, P. B. Forshee, G. R. Hundt, A. L. Sargent, S. G. Bott, V. Lynch, Inorg. Chem. 2007, 46, 10913 – 10925. [11] The voltammetric experiments were performed in acetonitrile only because the phenylenediamine exhibits complicated electrochemistry in dichloromethane and the potential separation between the two oxidation processes is insufficiently large in N,N-dimethylformamide to be of good use (see the Supporting Information). [12] L. A. Clare, L. E. Rojas-Sligh, S. M. Maciejewski, K. Kangas, J. E. Woods, L. J. Deiner, A. Cooksy, D. K. Smith, J. Phys. Chem. C 2010, 114, 8938 – 8949. [13] M. E. Tessensohn, M. Lee, H. Hirao, R. D. Webster, ChemPhysChem 2015, 16, 160 – 168. [14] The voltammetric behaviour of 2,3,5,6-tetramethyl-1,4-phenylenediamine, especially at low concentrations of 5 mm to 0.2 mm, has been reported by Smith et al. (reference 11) to be more complicated than the simple mechanism shown in Scheme 1, with the involvement of adsorption and desorption processes, hydrogen bonding and proton transfer steps to give a mixed valent hydrogen-bonded dimer as the kinetic product and P· + as the thermodynamic product of the first 1 e@oxidation reaction; whereas the second 1 e@-oxidation step was verified to be the conversion of P· + into P2 + . Although not investigated, we believe that the higher concentration of phenylenediamine used (1 mm), excess of pyridine compounds (200 mm), and hydrogen-bonding interactions between the compounds would simplify the phenylenediamine’s electrochemical behaviour and bypass some of these complex reactions. Therefore, the proposed method of using the DEPox of 2,3,5,6tetramethyl-1,4-phenylenediamine to measure the hydrogen acceptor ability of the pyridine compounds should still be applicable. [15] M. Zhang, L. Xiong, R. G. Compton, Anal. Methods 2013, 5, 3473 – 3481.
2256
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles [16] Y. Hui, E. L. K. Chng, L. P.-L. Chua, W. Z. Liu, R. D. Webster, Anal. Chem. 2010, 82, 1928 – 1934. [17] a) L. Xiong, A. M. Fletcher, S. G. Davies, S. E. Norman, C. Hardacre, R. G. Compton, Analyst 2012, 137, 4951 – 4957; b) L. Xiong, A. M. Fletcher, S. Ernst, S. G. Davies, R. G. Compton, Analyst 2012, 137, 2567 – 2573. [18] a) J. F. Corbett, J. Chem. Soc. B 1969, 213 – 216; b) M. Jonsson, J. Lind, T. E. Erikesen, G. Merenyi, J. Am. Chem. Soc. 1994, 116, 1423 – 1427; c) J. G. Speight, Langes Handbook of Chemistry, 16th ed., McGraw-Hill, New York, 2005. [19] CRC Handbook of Chemistry and Physics, 96th ed., CRC Press, 2016. [20] M. E. Tessensohn, H. Hirao, R. D. Webster, J. Phys. Chem. C 2013, 117, 1081 – 1090. [21] R. S. Hosmane, J. F. Liebman, Struct. Chem. 2009, 20, 693 – 697. [22] P. G. Seybold, K. B. Lipkowitz, Int. J. Quantum Chem. 1987, 31, 847 – 853. [23] J. Ireta, J. Neugebauer, M. Scheffler, J. Phys. Chem. A 2004, 108, 5692 – 5698. [24] a) K. Kim, K. D. Jordan, J. Phys. Chem. 1994, 98, 10089 – 10094; b) S. S. Xantheas, J. Chem. Phys. 1995, 102, 4505 – 4517; c) Y. Zhang, W. Pan, W. Yang, J. Chem. Phys. 1997, 107, 7921 – 7925; d) A. K. Rapp8, E. R. Bernstein, J. Phys. Chem. A 2000, 104, 6117 – 6128; e) S. Tsuzuki, H. P. Lethi, J. Chem. Phys. 2001, 114, 3949 – 3957. [25] P. Gilli, L. Pretto, V. Bertolasi, G. Gilli, Acc. Chem. Res. 2009, 42, 33 – 44. [26] a) M. E. Peover, J. D. Davies, J. Electroanal. Chem. 1963, 6, 46 – 53; b) Z. Galus, Fundamentals of Electrochemical Analysis, Halsted Press, Chichester, 1976; c) B. G. Chauhan, W. R. Fawcett, A. Lasia, J. Phys. Chem. 1977, 81, 1476 – 1481; d) W. R. Fawcett, M. Opallo, M. Fedurco, J. W. Lee, J. Am. Chem. Soc. 1993, 115, 196 – 200; e) N. Gupta, H. Linschitz, J. Am. Chem. Soc. 1997, 119, 6384 – 6391; f) M. Gjmez, F. J. Gonz#lez, I. Gonz#lez, Electroanalysis 2003, 15, 635 – 645. [27] A. J. Fry, W. E. Britton, (Eds.: P. T. Kissinger, W. R. Heineman), Marcel Dekker, New York, 1984.
ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
[28] A. D. Becke, J. Chem. Phys. 1993, 98, 5648 – 5652. [29] a) R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem. Phys. 1980, 72, 650 – 654; b) A. D. McLean, G. S. Chandler, J. Chem. Phys. 1980, 72, 5639 – 5648. [30] Gaussian 09 (Revision B.01), M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. J. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, N. J. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2010. [31] J. P. Merrick, D. Moran, L. Radom, J. Phys. Chem. A 2007, 111, 11683 – 11700. [32] a) L. T. Sein, Jr., A. L. Cederberg-Crossley, J. Mol. Struct. 2011, 1004, 319 – 328; b) N. Seeburrun, H. H. Abdallah, P. Ramasami, J. Phys. Chem. A 2012, 116, 3215 – 3223.
Manuscript received: May 22, 2017 Revised manuscript received: June 8, 2017 Accepted manuscript online: June 13, 2017 Version of record online: June 28, 2017
2257
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles
Using Voltammetry to Measure the Relative HydrogenBonding Strengths of Pyridine and Its Derivatives in Acetonitrile Malcolm E. Tessensohn, Yu Rong Koh, Sihao Lim, Hajime Hirao, and Richard D. Webster*[a] The voltammetric behavior of 2,3,5,6-tetramethyl-1,4-phenylenediamine was found to be able to differentiate the hydrogen acceptor abilities of electroinactive pyridine compounds in acetonitrile. Weak and strong hydrogen acceptors were distinguished through the onset of a third oxidation process that came about at sub-stoichiometric amounts of strong hydrogen acceptors, but not in the presence of weak hydrogen acceptors. This additional oxidation reaction occurred at a potential between the two 1 e@-oxidation reactions that phenylenediamines typically undergo (i.e. EPox(1) < EPox(3) < EPox(2), with EPox(1) and EPox(2) representing the electrochemical conversion of the neutral phenylenediamine into the radical cation and thereafter to the quinonediimine dication) as well as at the expense
of the second electrochemical reaction. EPox(2) and EPox(3) were observed to shift towards less positive potentials with increasing concentrations of weak or strong hydrogen acceptors, respectively, whereas EPox(1) remained virtually unaffected. This allows the electrochemical parameters DEPox(1, 2) = j EPox(1)@EPox(2) j and DEPox(1, 3) = j EPox(1)@EPox(3) j to be employed as measures of the hydrogen-bonding strengths within each category, to which they were found to be highly reproducible and responsive to steric, electronic, inductive, and mesomeric effects. The electrochemical findings concur with available aqueous pKa data of the protonated pyridine compounds but were, however, in poor agreement with results obtained by density functional theory calculations.
1. Introduction Voltammetry has been found to be exceptionally sensitive for studying hydrogen-bonding interactions because the hydrogen bond contains some electrostatic character and so can be electrochemically perturbed. This electrochemical technique is particularly advantageous for it only requires one component of the hydrogen bond (e.g. hydrogen donor or hydrogen acceptor) to be electroactive to qualify for investigation and suffers from lesser interference as compared to infrared and nuclear magnetic resonance spectroscopic methods. In addition, it has been reported to be more sensitive than UV/Vis spectroscopy[1] and concur with the results obtained via theoretical calculations.[2] The experimental electrochemical results have even been thought to be more accurate than theoretical calculations because the matrix effects and other important interactions (e.g. hydrogen bonding with trace water in the solvent and ion-pairing with the supporting electrolyte) are taken into consideration. Phenylenediamines have been widely used as dyes,[3] antioxidants in rubber and fuels,[4] organic materials in spin electron[a] Dr. M. E. Tessensohn, Y. R. Koh, S. Lim, Prof. Dr. H. Hirao, Prof. Dr. R. D. Webster Division of Chemistry and Biological Chemistry School of Physical and Mathematical Sciences Nanyang Technological University 21 Nanyang Link, Singapore 637371 (Singapore) E-mail: [email protected] Supporting Information and the ORCID identification number(s) for the author(s) of this article can be found under: https://doi.org/10.1002/cphc.201700560.
ChemPhysChem 2017, 18, 2250 – 2257
ic[5] and electrochromic[6] devices and in the detection of ammonia,[7] chlorine,[8] paracetamol[9] and metal cations,[10] all of which are based on their electroactivity. In acetonitrile (CH3CN),[11] phenylenediamines generally undergo two chemically reversible 1 e@-oxidation reactions with the neutral compound (P) first oxidized to the radical cation (P· + ) at potential EPox(1) and thereafter into the quinonediimine dication (P2 + ) at a more positive potential EPox(2) (Scheme 1 and Figure 1).[6, 12]
Scheme 1. Oxidation mechanism of phenylenediamines in acetonitrile.
The neutral compound P largely behaves as a weak base and so preferentially functions as a hydrogen acceptor through the lone pairs of electrons on its nitrogen atoms, but can also serve as a hydrogen donor via the hydrogen atoms of the amino groups. The oxidized forms P· + and P2 + , on the other hand, tend to behave as hydrogen donors with the latter being more acidic and a stronger donor because of its greater positive charge. We have previously applied the electrochemical reduction of quinones for measuring the hydrogen donor abilities of alcohols and diols in aprotic organic solvents[13] and now report
2250
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles ceptors) that is known to be highly accurate and reproducible. DE is unaffected by small changes between experiments in the effects of uncompensated solution resistance, heterogeneous electron-transfer rates (ket), surface condition of the working electrode and independent of drifts in potential of the reference electrode and so only responds to hydrogen bonding between 2,3,5,6-tetramethyl-1,4-phenylenediamine and the pyridine compounds.[13] However, systems that experience slow heterogeneous electron transfer, undergo adsorption, or are studied under highly resistant conditions that result in distorted voltammograms are not amendable for correlating DE with hydrogen bonding effects. DE, either as the difference in potentials within a single molecule or between two molecules, has been used in the detection of oxygen[15] and measurements of moisture content,[1, 16] temperature[15, 17] and humidity.[17a]
Figure 1. Cyclic voltammograms of 1.0 mm 2,3,5,6-tetramethyl-1,4-phenylenediamine with 0.1 m nBu4NPF6 in CH3CN recorded at a scan rate of 0.1 V s@1 at 25 : 2 8C with a 1 mm diameter planar disk glassy carbon working electrode and trace moisture content of 17.2 mm.
the use of the electrochemical oxidation of 2,3,5,6-tetramethyl1,4-phenylenediamine[14] for determining the hydrogen acceptor abilities of pyridine compounds (Figure 2) in CH3CN. These qualitative measurements are based on the electrochemical parameter DE (i.e. DEPox(1, 2) = j EPox(1)@EPox(2) j for weak hydrogen acceptors and DEPox(1, 3) = j EPox(1)@EPox(3) j for strong hydrogen ac-
Figure 2. Pyridine compounds examined in this study.
2. Results and Discussion The effect of pyridine, 2,6-lutidine and 2,6-di-tert-butylpyridine on the voltammetric behavior of 2,3,5,6-tetramethyl-1,4-phenylenediamine were first investigated and the resulting cyclic voltammograms are shown in Figure 3. Although the first electron-transfer process (oxidation of P to P· + ) was not visibly affected by the presence of pyridine and 2,6-lutidine (Figures 3 a and 3b, respectively), the second electrochemical reaction (oxidation of P· + to P2 + ) was found to be significantly influenced and absent in the cyclic voltammograms of the phenylenediamine when their concentrations were 1.0 mm and greater. A third oxidation process, initially detected at sub-stoichiometric amounts (< 1.0 mm) of the hydrogen acceptors, was also noticed to occur at 250 mV negative of the second oxidation reaction and at its expense. This third process grew in height
Figure 3. Cyclic (solid lines) and square-wave (dashed lines) voltammograms of 1.0 mm 2,3,5,6-tetramethyl-1,4-phenylenediamine with 0.1 m nBu4NPF6 in CH3CN at different concentrations of a) pyridine, b) 2,6-lutidine and c) 2,6-di-tert-butylpyridine recorded at a scan rate of 0.1 V s@1 at 25 : 2 8C with a 1 mm diameter planar disk glassy carbon working electrode. The currents of the square-wave voltammograms were scaled by a factor of 0.5. The initial voltammograms obtained before the addition of the compounds are indicated in red together with their measured trace moisture content.
ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
2251
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles (i.e. peak current (iPox(3)) increased) and shifted to less positive potentials with increasing concentrations of the hydrogen acceptors. It subsequently merged with the first electron-transfer reaction at & 200 mm of pyridine and & 10 mm of 2,6-lutidine to give a single oxidation process that was chemically reversible, approximately two times the height of the first oxidation process (i.e. iPox(merged) & 2 V iPox(1)) and equal to the sum of the initial two 1 e@-oxidation reactions (i.e. iPox(merged) & iPox(1) + iPox(2)) thereby suggesting that it involves the 2 e@-oxidation of P directly to P2 + . This behavior was also observed on a platinum working electrode as well as at faster scan rates of 1 and 10 V s@1 (see the Supporting Information). The chemical reversibility of the oxidation process was, however, compromised at higher concentrations ( & 200 mm) of 2,6-lutidine, probably because of proton-transfer reactions with the quinonediimine dication, but could be regained by either an increase in scan rate or shortening of the potential window as shown in the Supporting Information. Pyridine and 2,6-lutidine, as well as all the other compounds studied, are apparently electroinactive within the applied potential range on the glassy carbon working electrode because no additional processes, other than those assigned to the phenylenediamine, are observed throughout the experiments, even when they are present in a two hundred-fold excess. Cyclic voltammograms of pyridine recorded in the presence and absence of acids are provided in the Supporting Information. Smith and co-workers have suggested that the third oxidation process describes the electrochemical conversion of P· + into P2 + coupled with either strong hydrogen bonding or a proton-transfer reaction between P2 + and pyridine/2,6-lutidine (aqueous pKa values: 1.5 for P2 + ,[18] 5.23 for pyridinium and 6.65 for protonated 2,6-lutidine[19]) and so explains it occurring more easily than in the absence of the additives (i.e. EPox(3) < EPox(2)).[12] The high chemical reversibility of the third oxidation reaction suggests that the rate of proton-transfer between P2 + and pyridine/2,6-lutidine, if it occurs, is sufficiently fast in either direction, which might be facilitated if the resulting quinonediimine cation and pyridinium/protonated 2,6-lutidine are also held together by hydrogen-bonding interactions. It is unlikely that a species more easily oxidizable than P· + is responsible for the third oxidation process because pyridine and 2,6-lutidine are unlikely to deprotonate P· + under the experimental conditions (aqueous pKa of P· + is 12[18]). As for 2,6-di-tert-butylpyridine (Figure 3 c), no additional oxidation processes of the phenylenediamine other than those initially detected and shown in Scheme 1 were observed across the studied concentration range. The first oxidation reaction did not seem to be affected by the presence of 2,6-di-tert-butylpyridine, while the second oxidation reaction was observed to shift towards less positive potentials with increasing concentrations, but it did not merge with the first process even at the highest concentration of 200 mm. Even though a proton-transfer between 2,6-di-tert-butylpyridine (aqueous pKa value of 4.95) and P2 + was thermodynamically feasible, the steric effect of the bulky tert-butyl groups appears to have dominated their own electronic contribution so preventing the reaction from occurring. Thus, the third oxidation process was not observed. ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
The shifts of EPox(2) towards less positive potentials indicate that 2,6-di-tert-butylpyridine is nevertheless hydrogen bonding with the phenylenediamine and its oxidized forms, albeit weaker than those formed by pyridine and 2,6-lutidine. The shifting of the oxidation potentials of 2,3,5,6-tetramethyl-1,4-phenylenediamine, and phenylenediamines in general, towards less positive potentials with extensive hydrogen bonding to the pyridine compounds was expected and indicates that the ionization energies of the individual electron-transfer reactions are lowered as a consequence of the interactions; a similar trend was previously observed for the electrochemical oxidation of phenols and hydroquinone and their corresponding adiabatic ionization energies (AIEs) with hydrogen bonding to water.[20] The ionization energy of P· + was also anticipated to be more susceptible to the effects of hydrogen bonding than the ionization energy of P because the interactions between the pyridine compounds and P2 + are the strongest while those with P are the weakest. Therefore, EPox(2) and EPox(3) are shifted by a greater magnitude than EPox(1) and the related DEP parameters concomitantly decrease. The AIEs of P and P· + at different levels of hydrogen bonding to pyridine were calculated through density functional theory (DFT) and the results show that the changes to the AIE of P· + were almost always two times larger than those for P (Table 1) and so concur with the electrochemical findings. The energies, 3D geometries and Cartesian coordinates of the optimized structures of the various hydrogen-bonded complexes (with n = 1 to 4) can be found in the Supporting Information.
Table 1. DFT-calculated adiabatic ionization energies of the phenylenediamine(pyridine)n hydrogen-bonded complexes in CH3CN.
No. of pyridine molecules (n)
Adiabatic ionization energy [eV] P(pyridine)n P· + (pyridine)n
0 1 2 3 4
4.460 4.345 4.226 4.135 4.046
5.739 5.511 5.317 5.168 5.033
More importantly, these initial results support the premise that the voltammetric behavior of phenylenediamines can be employed to measure the relative hydrogen acceptor abilities of pyridine and its derivatives in CH3CN. The occurrence of the third oxidation process in the voltammetric response of 2,3,5,6-tetramethyl-1,4-phenylenediamine at sub-stoichiometric concentrations of the additives could first be used to differentiate between weak and strong hydrogen acceptors, for which it would be absent in the former and present in the latter. Thereafter, the electrochemical parameters DEPox(1, 2) = j EPox(1)@EPox(2) j and DEPox(1, 3) = j EPox(1)@EPox(3) j would be applied to assess the hydrogen bonding abilities of weak and strong hydrogen acceptors, respectively. Based on these metrics, 2,6-di-tert-butylpyridine would be considered as a weak hydrogen acceptor whereas pyridine and 2,6-lutidine can be regarded as strong hydrogen acceptors. The overall order of their hydrogen ac-
2252
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles ceptor strengths is 2,6-di-tert-butylpyridine < pyridine < 2,6-lutidine, which is also in line with their basicity. The effect of the picolines and trifluoromethylpyridines (CF3Py) on the voltammetric behavior of 2,3,5,6-tetramethyl-1,4phenylenediamine was also investigated and the electrochemical results indicate that the picolines are strong hydrogen acceptors while the trifluoromethylpyridines are weak hydrogen acceptors; their corresponding cyclic voltammograms are provided in the Supporting Information. A plot of DEPox(1, 2) = j EPox(1)@EPox(2) j against concentrations (Figure 4) shows that the Figure 5. Plots of DEPox(1, 3) = j EPox(1)@EPox(3) j against concentrations of the strong hydrogen acceptors as measured by square-wave voltammetry during the oxidation of 1.0 mm 2,3,5,6-tetramethyl-1,4-phenylenediamine with 0.1 m nBu4NPF6 as the supporting electrolyte in CH3CN recorded with a 1 mm diameter planar glassy carbon working electrode at 25 : 2 8C. The scans were carried out with a pulse period (t) of 25 Hz, potential step of 5 mV and pulse amplitude of 20 mV. The plots for 2,6-lutidine and 2-picoline are not included. Legend for inset: (red) 4-picoline, (blue) isoquinoline, (green) pyridine and (pink) 3-picoline.
Figure 4. Plots of DEPox(1, 2) = j EPox(1)@EPox(2) j against concentrations of the weak hydrogen acceptors as measured by square-wave voltammetry during the oxidation of 1.0 mm 2,3,5,6-tetramethyl-1,4-phenylenediamine with 0.1 m nBu4NPF6 as the supporting electrolyte in CH3CN recorded with a 1 mm diameter planar glassy carbon working electrode at 25 : 2 8C. The scans were carried out with a pulse period (t) of 25 Hz, potential step of 5 mV and pulse amplitude of 20 mV.
trifluoromethylpyridines are weaker hydrogen acceptors than 2,6-di-tert-butylpyridine and the strength of their interactions follow the trend 4-CF3-Py > 3-CF3-Py > 2-CF3-Py, indicating that the electron-withdrawing inductive effect and, to some extent, the steric effect of the CF3 group have outweighed the mesomeric stability. 3-CF3-Py would be expected to show the strongest interactions if the mesomeric effect was dominant because its carbon atom that is directly attached to the CF3 group does not carry a partial positive charge in any of its resonance structures, and so its nitrogen atom would be more basic, whereas the same carbon atoms of interest in 2-CF3-Py and 4-CF3-Py do. In the case of the picolines, a comparison of their plots of DEPox(1, 3) = j EPox(1)@EPox(3) j against concentrations (Figure 5) as well as their amounts needed to merge the two oxidation processes revealed that 2-picoline is a stronger hydrogen acceptor than 4-picoline thereby indicating that the electronic effect of the methyl substituent has outweighed its own steric effect. As for 3-picoline, its methyl group is apparently not in a position to inductively stabilize the important mesomeric forms and so exhibits the weakest hydrogen acceptor properties. The electrochemical results have also indicated that 2,6-lutidine is a stronger hydrogen acceptor than 2-picoline, which is viable only if the electronic effects are more significant than the steric effects in governing the strength of the hydrogen bonds. This is supported by DFT calculations where the aromatic rings ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
of the pyridine compounds and the phenylenediamine were found to be orthogonal instead of co-planar to each other (see the Supporting Information) so reducing the steric effect of the methyl groups. This line of reasoning however, does not seem to apply for 2,6-di-tert-butylpyridine probably because the steric repulsion arising between its tertiary butyl groups and the methyl groups of the phenylenediamine are similar regardless of the manner of approach and so their steric effects outweigh their electronic contributions. It should be noted that the best fit line plots for DEPox(1, 2) (Figure 4) and DEPox(1, 3) (Figure 5) against concentrations for most of the pyridine compounds were obtained from three separate yet highly reproducible experiments and so attests that the electrochemical parameters are unaffected by the surface condition of the polished and unmodified glassy carbon working electrode, solution resistance as well as drifts in potential of the reference electrode, which the absolute potentials EPox(1), EPox(2) and EPox(3) are not immune to. The effect of the benzopyridines, quinoline and isoquinoline, on the electrochemical response of 2,3,5,6-tetramethyl-1,4-phenylenediamine were also studied (see the Supporting Information for the cyclic voltammograms), to which both isomers gave a third oxidation process at sub-stoichiometric concentrations and hence can be deemed as strong hydrogen acceptors; the related dibenzopyridines, acridine and phenanthrenedine, were unfortunately insoluble in CH3CN. It is shown in Figure 5, and also reported in the literature,[21] that the hydrogen acceptor ability of quinoline is weaker than isoquinoline since the first and third oxidation processes remained separated at 200 mm of quinoline whereas only a single oxidation process was obtained at the same concentration of isoquinoline. This difference in hydrogen bonding abilities is a consequence of the steric effects from the adjacent ring as well as the peri hydrogen atom that collectively prevents quinoline from coming into closer proximity to 2,3,5,6-tetramethyl-1,4-phenylenediamine.[22]
2253
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles The influence of two heteroatoms as in the case of the 2,2’bipyridine and the diazines (i.e. pyridazine, pyrimidine and pyrazine) on the voltammetric response of 2,3,5,6-tetramethyl-1,4phenylenediamine was also examined. The phenylenediamine underwent a third oxidation process in the presence of substoichiometric amounts of 2,2’-bipyridine (see the Supporting Information) and so it could be viewed as a strong hydrogen acceptor. The electrochemistry experiments however, indicate that 2,2’-bipyridine has weaker hydrogen-bonding abilities than pyridine, thereby suggesting that bidentate interactions with the phenylenediamine, where both nitrogen atoms of 2,2’-bipyridine are simultaneously interacting with a hydrogen atom of the phenylenediamine, are not occurring. This supposition is supported by DFT calculations where only one hydrogen-bonding interaction was detected in the hydrogenbonded complexes irrespective of the oxidation state of the phenylenediamine. It is conceivable that the electron-withdrawing character and action of one pyridine ring on the other may have reduced the overall hydrogen acceptor ability of 2,2’-bipyridine and so debilitate its propensity to participate in bidentate hydrogen bonding. The electrochemical results (see Figure 4 and Figure 5 as well as cyclic voltammograms that are included in the Supporting Information) indicate that pyridazine can be considered as a strong hydrogen acceptor while pyrimidine and pyrazine may be regarded as weak hydrogen acceptors. Pyridazine was however weaker than pyridine because the additional nitrogen atom exerts an electron-withdrawing inductive effect on the other nitrogen atom and so its overall binding ability is lessened. Among the diazines, pyridazine and pyrazine were the strongest and weakest hydrogen acceptors, respectively, which is in line with their reported basicity (see Table 2 and Table 3). Nevertheless, the observed trend for the diazines cannot be fully explained by mesomeric and inductive effects regardless whether they are independently or jointly considered. Pyrimidine should instead be the strongest if the mesomeric effects were overriding because no partial positive charge would reside on any of its nitrogen atoms. On the other hand, if the inductive effects were dominant then pyrazine would overshadow the others because its two nitrogen atoms are the fur-
Table 2. Relative binding strengths (in kJ mol@1) of the strong hydrogen acceptors with 2,3,5,6-tetramethyl-1,4-phenylenediamine in CH3CN.
Compound 2,6-Lutidine 2-Picoline 4-Picoline Isoquinoline 3-Picoline Pyridine 2,2’-Bipyridine Quinoline Pyridazine
P
Rel. hydrogen bond strength P· + P2 +
@1.66 @3.57 @3.69 @4.17 @3.83 @4.17 @12.59 @3.59 @3.87
@11.59 @14.28 @16.41 @15.49 @16.13 @15.19 @19.89 @13.14 @14.99
@30.41 @35.24 @41.12 @36.98 @39.06 @37.16 @34.06 @32.73 @35.14
Aqueous pKa values[19] 6.65 6.00 5.99 5.40 (20 8C) 5.70 5.23 – 4.90 (20 8C) 2.24 (20 8C)
The pKa values were measured at 25 8C unless otherwise indicated.
ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
Table 3. Relative binding strengths (in kJ mol@1) of the weak hydrogen acceptors with 2,3,5,6-tetramethyl-1,4-phenylenediamine in CH3CN.
Compound
Rel. hydrogen bond strength P P· + P2 +
Aqueous pKa values[19]
2,6-Di-tert-butylpyridine 4-CF3-Py Pyrimidine 3-CF3-Py Pyrazine 2-CF3-Py
3.04 @3.40 @3.44 @3.50 @3.34 @2.98
3.58 – 1.23 – 0.65 –
3.88 @13.08 @12.51 @13.11 @11.85 @10.12
7.77 @29.56 @28.27 @29.31 @25.99 @22.32
The pKa values were measured at 20 8C.
thest apart. Pyridazine was however determined to be the strongest hydrogen acceptor for the reason that hydrogen bonding with the phenylenediamine alleviates the electronic repulsion between the two lone electron pairs of the adjacent nitrogen atoms. Even so, the electronic repulsion between the lone electron pairs was insufficient to outweigh the electronwithdrawing inductive effect of the adjacent nitrogen atom and so pyridazine was the weakest among the compounds sorted as strong hydrogen acceptors. The trifluoromethylpyridines, pyrimidine, pyrazine and 2,6di-tert-butylpyridine, when present in sub-stoichiometric amounts, did not cause 2,3,5,6-tetramethyl-1,4-phenylenediamine to undergo a third oxidation process and so can be regarded as weak hydrogen acceptors. The order of their hydrogen acceptor ability, as determined by the electrochemical parameter DEPox(1, 2) = j EPox(1)@EPox(2) j (Figure 4), was found to follow 2-CF3-Py < pyrazine < 3-CF3-Py < pyrimidine < 4-CF3-Py < 2,6-di-tert-butylpyridine. The remaining compounds, on the other hand, may perhaps be considered as strong hydrogen acceptors because of the occurrence of the third oxidation reaction; their hydrogen acceptor ability was measured by the electrochemical parameter DEPox(1, 3) = j EPox(1)@EPox(3) j (Figure 5) and follows pyridazine < quinoline < 2,2’-bipyridine < pyridine < 3-picoline < isoquinoline < 4-picoline < 2-picoline < 2,6lutidine. The relative hydrogen-bonding strengths of pyridine and its derivatives with 2,3,5,6-tetramethyl-1,4-phenylenediamine were also calculated to which the data for the strong and weak hydrogen acceptors (together with their available pKa data) are reflected in Table 2 and Table 3, respectively; noting that the compounds are arranged in the order of hydrogen acceptor strengths obtained by the electrochemical procedure. The computational results were however, in poor agreement with the electrochemical findings and reported pKa values. For instance, the calculations signal that the binding strengths of the phenylenediamine-2,6-lutidine hydrogen-bonded complexes are the weakest and that formation of the phenylenediamine-2,6-di-tert-butylpyridine hydrogen-bonded complexes is unfavorable. The accuracy of DFT calculations in describing hydrogen bonds has been well scrutinized in the literature and reported to be dependent on, amongst other factors, the bond directionality,[23] exchange-correlation functionals and basis set.[24] In addition, although the solvation effects by
2254
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles CH3CN were taken into consideration in the calculations, the other important interactions such as ion pairing between the charged products and the supporting electrolyte as well as hydrogen bonding between the phenylenediamine, pyridine compounds and trace water in the solvent were disregarded. While effort was put into ensuring that most of the geometries were explored, it might be possible that some conformers were inadvertently left out because of the numerous conformational possibilities and so might explain the disagreement between the computational and experimental results. Nevertheless, the neutral and quinonediimine dicationic forms of the phenylenediamine correspondingly and expectedly formed the weakest and strongest hydrogen-bonding interactions. One advantage of this procedure for evaluating hydrogen acceptor strengths through the oxidation of phenylenediamines, is its high shared sensitivity towards electronic and steric effects as well as mesomeric and inductive stability that are not the same across the different classes of molecules (benzopyridines, diazines, picolines, trifluoromethylpyridines etc.). It is also experimentally simple to perform, fast and useful especially in cases when the theoretical considerations alone are insufficient to reach a decision. Take, for example, a comparison between 2,6-di-tert-butylpyridine and 4-CF3-Py where it is challenging to determine if the steric effect of the tert-butyl groups will dominate the electron-withdrawing inductive effect of the CF3 group, noting that the tert-butyl groups are also imparting an electron-donating inductive influence whereas the CF3 group in 4-CF3-Py does not impose any steric obstruction towards hydrogen bond formation. The electrochemical experiments indicate that the electron-withdrawing character of the CF3 group dominates and so 2,6-di-tert-butylpyridine is the stronger hydrogen acceptor. Lastly, and probably more applicably, the voltammetric method may also serve to measure the relative basicity of compounds in non-aqueous solvents. This is because the energy of a hydrogen bond is reported to correlate with the pKa of hydrogen donors and pKb of hydrogen acceptors in a given solvent and that the formation of a hydrogen-bonded complex is believed to precede proton-transfer reactions.[25] However, the voltammetric method only provides qualitative information about the hydrogen bond and so would require other treatments[26] to obtain quantitative data.
3. Conclusions In this study, the sensitivity of the electrochemical oxidation of 2,3,5,6-tetramethyl-1,4-phenylenediamine towards hydrogen bonding was used to qualitatively measure the hydrogen acceptor abilities of pyridine and its derivatives. The relative hydrogen bonding strengths were assessed by the electrochemical parameters DEPox(1, 2) = j EPox(1)@EPox(2) j for the weaker hydrogen acceptors and DEPox(1, 3) = j EPox(1)@EPox(3) j for stronger ones. These potential measurements are reproducible because the effects of solution resistance, migration, kinetics of the heterogeneous electron-transfer as well as variations in the surface condition of the polished bare unmodified glassy carbon working electrode and potential of the reference electrode are elimChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
inated because the potential measured is between two redox processes within the same compound. The electrochemical parameters respond to steric, electronic, inductive and mesomeric influences as well as number of heteroatoms present and are inversely related to the hydrogen acceptor ability of the compounds. The experimental trends agree with the reported pKa values of the protonated forms but are in poor accord with the DFT results. This voltammetric technique only requires small amounts of material for use, is fast and easy to perform, and the results obtained are not difficult to interpret. It is also applicable for electroinactive hydrogen acceptors and suffers from little interference. Nevertheless, the procedure only provides qualitative information about the hydrogen bond and so has to rely on other methods to acquire quantitative information. Nonetheless, the method may serve to indicate the relative basicity of compounds in non-aqueous solvents. The results presented in this study with a phenylenediamine and previously with quinones have shown that the hydrogen bonding abilities of electroinactive hydrogen donors and acceptors can be voltammetrically determined.
Experimental Section Chemicals The following chemicals were used as received: 2,3,5,6-tetramethyl-1,4-phenylenediamine (Sigma–Aldrich, 99 %), pyridine (Alfa Aesar, 99.5 + %), 2-picoline (Merck, + 98 %), 3-picoline (Alfa Aesar, 99 %), 4-picoline (Alfa Aesar, 98 %), 2-(trifluoromethyl)pyridine (2CF3-Py, Alfa Aesar, 99 %), 3-(trifluoromethyl)pyridine (3-CF3-Py, Alfa Aesar, 97 %), 4-(trifluoromethyl)pyridine (4-CF3-Py, Alfa Aesar, 97 %), 2,6-lutidine (Alfa Aesar, 98 + %), 2,6-di-tert-butylpyridine (Aldrich, + 97 %), quinoline (Alfa Aesar, 97 %), isoquinoline (Alfa Aesar, 98 %), pyridazine (Alfa Aesar, 98 + %), pyrimidine (Alfa Aesar, 99 %), pyrazine (Aldrich, 99 + %), 2,2’-bipyridine (Sigma–Aldrich, + 99 %) and CH3CN (RCI Labscan Limited, AR grade). The supporting electrolyte, tetra-n-butylammonium hexafluorophosphate (nBu4NPF6), was prepared by a literature procedure.[27] 1/16 in. rods with 3 a pore size molecular sieves (CAS: 308 080-99-1) were obtained from Fluka.
Measurement of the Initial Water Content Karl Fischer titrations were performed with a Mettler Toledo DL32 coulometer inside a humidity control box (Coy Laboratory Products) maintained at 30 % relative humidity, with (Riedel-deHa[n) HYDRANAL-Coulomat AG and CG solutions in the anolyte and catholyte compartments, respectively. Stabilization of the coulometer was achieved when the drift value of water was steady and close to 0 mg min@1. Each measurement was found to conclude within 1 min so indicating that the drift from atmospheric water was insignificant.
Voltammetry Cyclic voltammetry and square-wave voltammetry experiments were performed with a computer-controlled Eco Chemie Autolab PGSTAT302N potentiostat in a three-electrode cell with a 1 mm diameter planar glassy carbon disk (Cypress Systems) working electrode, a platinum wire (Metrohm) auxiliary electrode, and a silver
2255
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles wire (Cypress Systems) miniature reference electrode connected to the test solution via a salt bridge containing 0.5 m nBu4NPF6 in CH3CN. The internal filling solution of the liquid-junction reference electrode was replaced with a dried solution at the start of each experiment to reduce the interference of water.
Preparation of Dried Electrolyte Solutions Dried CH3CN was used to make a solution of 0.1 m nBu4NPF6 with the supporting electrolyte predried under vacuum at 160 8C for 6 h. The solution was then quantitatively transferred into a vacuum syringe (SGE Analytical Science) containing 3 a molecular sieves that were also predried under vacuum at 160 8C for 6 h. The syringe was subsequently stored under a nitrogen atmosphere for a minimum period of 48 h prior to further use.
Procedure At the start of each experiment, 15 mL of the dried electrolyte solution was used to make up a 1.0 mm solution of 2,3,5,6-tetramethyl-1,4-phenylenediamine in the humidity control box (Coy Laboratory Products) that was maintained at 30 % relative humidity. The phenylenediamine/electrolyte solution was then transferred into the electrochemical cell that had been previously heated at 100 8C for an hour and cooled under an argon atmosphere to room temperature, before deoxygenating the solution with argon gas. Consecutive cyclic voltammetry and square-wave voltammetry scans were subsequently performed. A 5 mL aliquot of the electrochemical solution was extracted with a vacuum syringe (SGE Analytical Science) and then subjected to Karl Fischer titration to obtain the initial moisture content of the solution. The working electrode was systematically polished on a Buehler polishing pad following which microliter volumes of the compounds from a concentrated stock solution were then carefully and accurately added with a micropipette and mixed with a steady stream of argon bubbles.
Computational Methods The molecular geometries were drawn by using GaussView 5 and optimized with the B3LYP hybrid density functional[28] with the 6– 311 + G(2df,p) Pople-style Gaussian basis set[29] (triple z quality with one set of diffuse s and p functions on heavy atoms, two additional d-type polarization functions and an additional f-type polarization function on heavy atoms, and one set of p-type polarization functions on hydrogen) with solvation effects taken into account within the Gaussian 09 suite of programs.[30] The final geometries were verified to stay at true energy minima through frequency calculations at the B3LYP/6–311 + G(2df,p) level that yielded no imaginary frequencies. The zero-point energies were scaled by a factor of 0.9889.[31] The adiabatic ionization energies were computed based on the differences in ground state energies of the respective oxidized states.[32] For example, the AIE of P was obtained by subtracting the ground state energy of P· + from that of P while the AIE of P· + can be acquired through the difference in ground state energies of P· + and P2 + . The relative hydrogen bond strengths of the pyridine compounds were calculated by subtracting the ground state energies (single point energy plus zero-point correction) of the components from the energy of the hydrogen-bonded complex: relative hydrogen bond strength = energy of hydrogenbonded complex—(energy of 2,3,5,6-tetramethyl-1,4-phenylenediamine + energy of pyridine compound). ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
Acknowledgements This work was supported by Singapore Government MOE Academic Research Fund Tier 1 Grant (RG109/15). M.E.T. thanks Nanyang Technological University for the award of the Nanyang President’s Graduate Scholarship (NPGS). Keywords: basicity · electrochemistry · hydrogen acceptor · phenylenediamine · voltammetry [1] Y. Hui, E. L. K. Chng, C. Y. L. Chng, H. L. Poh, R. D. Webster, J. Am. Chem. Soc. 2009, 131, 1523 – 1534. [2] a) M. Gjmez, F. J. Gonz#lez, I. Gonz#lez, J. Electrochem. Soc. 2003, 150, E527 – E534; b) J. Garza, R. Vargas, M. Gjmez, I. Gonz#lez, F. J. Gonz#lez, J. Phys. Chem. A 2003, 107, 11161 – 11168; c) C. Costentin, M. Robert, J.M. Sav8ant, C. Tard, Angew. Chem. Int. Ed. 2010, 49, 3803 – 3806; Angew. Chem. 2010, 122, 3891 – 3894; d) J.-M. Sav8ant, C. Tard, J. Am. Chem. Soc. 2014, 136, 8907 – 8910. [3] R. D. Theys, G. Sosnovsky, Chem. Rev. 1997, 97, 83 – 132. [4] a) J. Posp&sˇil, Aromatic and Heterocyclic Amines in Polymer Stabilization, Vol. 124, Springer Berlin Heidelberg, 1995; b) F. Cataldo, Polym. Degrad. Stab. 2001, 72, 287 – 296; c) F. Cataldo, Eur. Polym. J. 2002, 38, 885 – 893; d) N. M. Huntink, R. N. Datta, J. W. M. Noordermeer, Rubber Chem. Technol. 2004, 77, 476 – 511. [5] a) A. Ito, M. Urabe, K. Tanaka, Polyhedron 2003, 22, 1829 – 1836; b) A. Ito, Y. Nakano, M. Urabe, T. Kato, K. Tanaka, J. Am. Chem. Soc. 2006, 128, 2948 – 2953; c) A. Ito, R. Kurata, D. Sakamaki, S. Yano, Y. Kono, Y. Nakano, K. Furukawa, T. Kato, K. Tanaka, J. Phys. Chem. A 2013, 117, 12858 – 12867. [6] S. J. L. Lauw, X. Xu, R. D. Webster, ChemPlusChem 2015, 80, 1288 – 1297. [7] X. Ji, R. G. Compton, Anal. Sci. 2007, 23, 1317 – 1320. [8] E. H. Seymour, N. S. Lawrence, R. G. Compton, Electroanalysis 2003, 15, 689 – 694. [9] Y. Peng, Z. Wu, Z. Liu, Anal. Methods 2014, 6, 5673 – 5681. [10] a) A. J. Pearson, J.-J. Hwang, Tetrahedron Lett. 2001, 42, 3541 – 3543; b) J. W. Sibert, P. B. Forshee, Inorg. Chem. 2002, 41, 5928 – 5930; c) J. W. Sibert, P. B. Forshee, V. Lynch, Inorg. Chem. 2005, 44, 8602 – 8609; d) J. W. Sibert, G. R. Hundt, A. L. Sargent, V. Lynch, Tetrahedron 2005, 61, 12350 – 12357; e) J. W. Sibert, P. B. Forshee, G. R. Hundt, A. L. Sargent, S. G. Bott, V. Lynch, Inorg. Chem. 2007, 46, 10913 – 10925. [11] The voltammetric experiments were performed in acetonitrile only because the phenylenediamine exhibits complicated electrochemistry in dichloromethane and the potential separation between the two oxidation processes is insufficiently large in N,N-dimethylformamide to be of good use (see the Supporting Information). [12] L. A. Clare, L. E. Rojas-Sligh, S. M. Maciejewski, K. Kangas, J. E. Woods, L. J. Deiner, A. Cooksy, D. K. Smith, J. Phys. Chem. C 2010, 114, 8938 – 8949. [13] M. E. Tessensohn, M. Lee, H. Hirao, R. D. Webster, ChemPhysChem 2015, 16, 160 – 168. [14] The voltammetric behaviour of 2,3,5,6-tetramethyl-1,4-phenylenediamine, especially at low concentrations of 5 mm to 0.2 mm, has been reported by Smith et al. (reference 11) to be more complicated than the simple mechanism shown in Scheme 1, with the involvement of adsorption and desorption processes, hydrogen bonding and proton transfer steps to give a mixed valent hydrogen-bonded dimer as the kinetic product and P· + as the thermodynamic product of the first 1 e@oxidation reaction; whereas the second 1 e@-oxidation step was verified to be the conversion of P· + into P2 + . Although not investigated, we believe that the higher concentration of phenylenediamine used (1 mm), excess of pyridine compounds (200 mm), and hydrogen-bonding interactions between the compounds would simplify the phenylenediamine’s electrochemical behaviour and bypass some of these complex reactions. Therefore, the proposed method of using the DEPox of 2,3,5,6tetramethyl-1,4-phenylenediamine to measure the hydrogen acceptor ability of the pyridine compounds should still be applicable. [15] M. Zhang, L. Xiong, R. G. Compton, Anal. Methods 2013, 5, 3473 – 3481.
2256
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Articles [16] Y. Hui, E. L. K. Chng, L. P.-L. Chua, W. Z. Liu, R. D. Webster, Anal. Chem. 2010, 82, 1928 – 1934. [17] a) L. Xiong, A. M. Fletcher, S. G. Davies, S. E. Norman, C. Hardacre, R. G. Compton, Analyst 2012, 137, 4951 – 4957; b) L. Xiong, A. M. Fletcher, S. Ernst, S. G. Davies, R. G. Compton, Analyst 2012, 137, 2567 – 2573. [18] a) J. F. Corbett, J. Chem. Soc. B 1969, 213 – 216; b) M. Jonsson, J. Lind, T. E. Erikesen, G. Merenyi, J. Am. Chem. Soc. 1994, 116, 1423 – 1427; c) J. G. Speight, Langes Handbook of Chemistry, 16th ed., McGraw-Hill, New York, 2005. [19] CRC Handbook of Chemistry and Physics, 96th ed., CRC Press, 2016. [20] M. E. Tessensohn, H. Hirao, R. D. Webster, J. Phys. Chem. C 2013, 117, 1081 – 1090. [21] R. S. Hosmane, J. F. Liebman, Struct. Chem. 2009, 20, 693 – 697. [22] P. G. Seybold, K. B. Lipkowitz, Int. J. Quantum Chem. 1987, 31, 847 – 853. [23] J. Ireta, J. Neugebauer, M. Scheffler, J. Phys. Chem. A 2004, 108, 5692 – 5698. [24] a) K. Kim, K. D. Jordan, J. Phys. Chem. 1994, 98, 10089 – 10094; b) S. S. Xantheas, J. Chem. Phys. 1995, 102, 4505 – 4517; c) Y. Zhang, W. Pan, W. Yang, J. Chem. Phys. 1997, 107, 7921 – 7925; d) A. K. Rapp8, E. R. Bernstein, J. Phys. Chem. A 2000, 104, 6117 – 6128; e) S. Tsuzuki, H. P. Lethi, J. Chem. Phys. 2001, 114, 3949 – 3957. [25] P. Gilli, L. Pretto, V. Bertolasi, G. Gilli, Acc. Chem. Res. 2009, 42, 33 – 44. [26] a) M. E. Peover, J. D. Davies, J. Electroanal. Chem. 1963, 6, 46 – 53; b) Z. Galus, Fundamentals of Electrochemical Analysis, Halsted Press, Chichester, 1976; c) B. G. Chauhan, W. R. Fawcett, A. Lasia, J. Phys. Chem. 1977, 81, 1476 – 1481; d) W. R. Fawcett, M. Opallo, M. Fedurco, J. W. Lee, J. Am. Chem. Soc. 1993, 115, 196 – 200; e) N. Gupta, H. Linschitz, J. Am. Chem. Soc. 1997, 119, 6384 – 6391; f) M. Gjmez, F. J. Gonz#lez, I. Gonz#lez, Electroanalysis 2003, 15, 635 – 645. [27] A. J. Fry, W. E. Britton, (Eds.: P. T. Kissinger, W. R. Heineman), Marcel Dekker, New York, 1984.
ChemPhysChem 2017, 18, 2250 – 2257
www.chemphyschem.org
[28] A. D. Becke, J. Chem. Phys. 1993, 98, 5648 – 5652. [29] a) R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem. Phys. 1980, 72, 650 – 654; b) A. D. McLean, G. S. Chandler, J. Chem. Phys. 1980, 72, 5639 – 5648. [30] Gaussian 09 (Revision B.01), M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. J. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, N. J. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2010. [31] J. P. Merrick, D. Moran, L. Radom, J. Phys. Chem. A 2007, 111, 11683 – 11700. [32] a) L. T. Sein, Jr., A. L. Cederberg-Crossley, J. Mol. Struct. 2011, 1004, 319 – 328; b) N. Seeburrun, H. H. Abdallah, P. Ramasami, J. Phys. Chem. A 2012, 116, 3215 – 3223.
Manuscript received: May 22, 2017 Revised manuscript received: June 8, 2017 Accepted manuscript online: June 13, 2017 Version of record online: June 28, 2017
2257
T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
