Article pubs.acs.org/JPCB
Raman Changes Induced by Electrochemical Oxidation of Poly(triarylamine)s: Toward a Relationship between Molecular Structure Modifications and Charge Generation Thiago Cervantes,†,‡ Guy Louarn,*,‡ Henrique de Santana,† Lukasz Skorka,§ and Irena Kulszewicz-Bajer*,§ †
Departamento de Química, Universidade Estadual de Londrina, 86057-970 Londrina, PR, Brazil Insitut des Materiaux Jean Rouxel, CNRS-University of Nantes, 2 rue de la Houssinière, 44322 Nantes, France § Faculty of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00-664 Warsaw, Poland ‡
S Supporting Information *
ABSTRACT: Linear, alternating polymers of aromatic amines (p-phenylenediamine, bis(p-aminophenyl)amine, and diaminocarbazole) and either m-phenylene or 3,5-pyridine have been synthesized and characterized by electrochemical and spectroscopic means. The presence of radical cations in their electrochemically oxidized forms is manifested by new bands in the UV−vis−NIR spectra whose appearance can be correlated with reversible redox couples registered in the corresponding cyclic voltammograms at approximately the same potentials as well as with pronounced evolutions of their resonance Raman spectra. Detailed analysis of the Raman data gives information about the locations and the distribution of radical cations and spinless dication in the macromolecule. This approach can provide a new insight into the formation of high-spin states in these polymers.
■
INTRODUCTION The search for purely organic magnetic materials, especially polymeric ones, which can be applied in molecular electronics or spintronics has attracted considerable attention in the past decade. The main design principles of high-spin polymeric compounds can be outlined as follows: their macromolecules should contain chemical units bearing unpaired spins, separated by conjugated units which can couple these spins in a ferromagnetic fashion.1 A large variety of spin bearing units has been developed and studied to date, as, for example, triarylmethyl,2−7 phenoxyl,8 nitronyl nitroxide,9,10 aminyl radicals11−13 and triarylamine radical cations.14−22 The most impressive result, concerning the fabrication of purely organic ferromagnetic materials, was obtained by Rajca et al. for a polymer containing triarylmethyl radicals which exhibited very high-spin states up to S = 5000 at low temperatures.5 However, considering potential applications of organic magnetic materials, the research should be directed toward the design and synthesis of compounds in which spin bearing units are more stable at room temperature, as, for example, triarylamine radical cations. Recently, several types of alternating oligo- and polyarylamines exhibiting a high-spin state have been synthesized.14−22 The highest spin state S = 9/2 was achieved for cross-linked polyarylamine.23 Many fundamental studies have also been devoted to linear polymers whose topology facilitates the determination of the principal factors influencing spin interactions. Linear polymers are also interesting due to their solubility and solution processability. The principal problem in this type of compounds is a limited spin © 2014 American Chemical Society
concentration causing a decrease in the extent of their magnetic interactions. Moreover, the results obtained to date show that only spins from two neighboring repeat units participate in the formation of the high-spin state and the magnetic interaction is not spread along the polymer chain.24,25 The main reason for this finding is related to an unfavorable local conformation which can change the nature of an interaction from ferromagnetic to antiferromagnetic. Additionally, the location of the radical cation within the conjugated amine segment can limit the magnetic interaction. Radical cation can be localized only in a specific part of the conjugated unit significantly decreasing spin density in the ferromagnetic coupler which, in turn, results in decreasing ferromagnetic interactions. In this context, the elucidation of the nature and location of the formed radical cations, possibly by hyphenated electrochemical−spectroscopic (UV−vis−NIR, Raman) techniques, is of crucial importance. In this paper, we describe the results of such detailed studies carried out for four linear polymers containing aromatic amine segments (p-phenylenediamine, bis(p-aminophenyl)amine, and diaminocarbazole) altered with either m-phenylene or 3,5-pyridine ferromagnetic couplers. Detailed theoretical and experimental analysis of the Raman spectra, obtained for these polymers at their different oxidation states, enabled us to gain a deeper insight into the spin and Received: November 5, 2014 Revised: December 23, 2014 Published: December 23, 2014 1756
DOI: 10.1021/jp511087j J. Phys. Chem. B 2015, 119, 1756−1767
Article
The Journal of Physical Chemistry B
Hz, 2H), 6.61 (d, J = 8.8 Hz, 4H), 3.67 (s, 2H), 3.18 (s, 6H), 1.48−1.40 (m, 2H), 1.12−1.04 (m, 6H), 0.81 (t, J = 6.9 Hz, 3H). 13C NMR (100 MHz, C6D6) δ, 155.8, 150.6, 142.0, 140.3, 138.3, 126.3, 125.8, 124.1, 119.0, 114.9, 113.3, 109.7, 54.8, 31.7, 29.0, 27.0, 22.8, 14.2. Anal. Calcd for C38H37N3O2: C, 80.42; H, 6.53; N, 7.41; O, 5.64. Found: C, 79.98; H, 6.52; N, 7.33. Mn = 6.9 kDa, Mw = 56.6 kDa. Spectroscopic Electrochemical and Spectroelectrochemical Techniques. XPS measurements were carried out at room temperature with an Axis Nova spectrometer from Kratos Analytical with an Al Kα line (1486.6 eV) as the excitation source. Thin layers of polymers were deposited on a glass substrate from chloroform solutions by drop-casting, then dried under a gentle nitrogen flux, and then directly introduced overnight in the sample exchange chamber of the spectrometer. Survey spectra were acquired at a pass energy of 80 eV and with an energy step of 1 eV. The core level spectra (C 1s, O 1s, and N 1s) were acquired using a constant pass energy mode of 10 eV, to obtain data in a reasonable experimental time (energy resolution of 0.28 eV). The pressure in the analysis chamber was maintained below 10−7 Pa. Data analysis was performed using CasaXPS software. The background spectra were considered as Shirley type, and curve fitting was carried out with a mixture of Gaussian−Lorentzian functions. Concerning the calibration, the binding energy for the C 1s hydrocarbons peak was set at 284.8 eV. No surface cleaning, using Ar sputtering, for example, was made. It is known from experience that carbonaceous atmospheric contamination usually occurs on the surface of a given material studied, but in the studied cases, ion sputtering can be suspected of changing the chemical composition of the surface and inducing structural damage. Raman spectra of neutral and oxidized polymers were recorded on a Renishaw InVia reflex spectrometer (λexc = 633.0 nm, laser power = 10 mW, 2 μm diameter spot, typical exposure times 60 s) or on a FT Raman Bruker RFS 100 spectrometer with the near-IR excitation line (1064 nm, laser power = 50 mW). All experiments were carried out on films of polymers freshly deposited on a platinum plate by drop-casting. The registered Raman signals frequently showed a pronounced baseline slope as a result of fluorescence, extending from 100 to 1700 cm−1. The baseline seriously drifted upon changes in the electrode polarization potential; in addition, it was affected by the energy of the excitation line. These two factors increased the difficulty in identifying characteristic peaks. Therefore, for the fluorescence correction, a “spline” linear interpolation method was rigorously executed to simulate the shape of the baseline, which significantly improved the quality of all spectra. Fourier transform infrared spectra (FTIR) were recorded on a Bruker Vertex 70 spectrometer (4 cm−1), using the KBr pellet method. Cyclic voltammograms were registered for thin polymer films drop-cast onto a platinum electrode. The experiments were carried out in a one-compartment electrochemical cell, in a solution of 0.1 M Bu4NBF4 in CH2Cl2 with Ag/0.1 M AgNO3 in acetonitrile as a reference electrode and a Pt counter electrode. The surface of the Pt disk electrode was 25 mm2. Raman spectroelectrochemical experiments were carried out in the oxidative mode. The potential of the working electrode was being increased in small increments. After each potential raise, a wait time was applied until the quasi-equilibrium state was reached and then the Raman spectrum was registered “in situ”. It was assumed that the equilibrium state was reached when the potential change induced current became negligible.
charge storage configurations, which is crucial for any prediction of their capability of high-spin-state formation.
■
EXPERIMENTAL SECTION Materials. 3,6-Bis(4′-methoxyphenyl)amino-9-hexyl-carbazole. 1.22 g (3 mmol) of 3,6-dibromo-9-hexyl-carbazole, 0.738 g (6 mmol) of p-anizidine, 274.7 mg (0.3 mmol) of Pd2(dba)3, 560.4 mg (0.6 mmol) of BINAP, and 0.864 g (9 mmol) of sodium tert-butoxide were dissolved in 20 mL of dry dioxane under an argon atmosphere. The mixture was stirred and heated at 100 °C for 20 h. Removal of the solvent followed by chromatography on silica gel (ethyl acetate/hexanes, 1:2, then 2:1) resulted in a pale-yellow powder. 1.1 g (2.23 mmol, 74.4% yield). 1H NMR (400 MHz, C6D6) δ, 7.65 (d, J = 2.0 Hz, 2H), 7.20 (dd, J = 8.6, 2.0 Hz, 2H), 7.11 (d, J = 8.6 Hz, 2H), 6.94 (d, J = 8.8 Hz, 4H), 6.82 (d, J = 8.8 Hz, 4H), 5.01 (s, 2H), 3.83 (t, J = 7.1 Hz, 2H), 3.38 (s, 6H), 1.58 (q, 2H), 1.22−0.99 (m, 6H), 0.81 (t, J = 6.9 Hz, 3H). 13C NMR (100 MHz, C6D6) δ, 154.5, 139.8, 137.5, 136.9, 123.9, 120.0, 119.0, 115.1, 111.6, 109.5, 55.2, 43.2, 31.8, 29.3, 27.2, 22.8, 14.1. IR (cm−1): 3350, 3045, 2952, 2927, 2858, 2830, 1509, 1475, 1439, 1309, 1278, 1256, 1230, 1031, 812. Anal. Calcd for C32H35N3O2: C, 77.86; H, 7.15; N, 8.51; O, 6.48. Found: C, 77.51; H, 7.03; N, 8.53. m/ z 493. Polymerization Procedure. All polymers studied were prepared using Hartwig−Buchwald catalytic amination of aryl halides. Polycondensation was carried out using the method described by Goodson et al.26 0.05 mmol of palladium(II) acetate and 0.15 mmol of tri-tert-butylphosphine were mixed with 3 mL of dry toluene and stirred for 15 min. Then, 2.5 mmol of 1,3-dibromoderivative, 2.5 mmol of diaminoderivative, 7.5 mmol of sodium tert-butoxide, and 12 mL of toluene were placed in a reaction flask containing the catalyst under an argon atmosphere. The mixture was stirred and heated at 90 °C for 4 days. The crude mixture was cooled to room temperature, and the product was precipitated with 300 mL of methanol. The solid was isolated via filtration and then washed with water and methanol. Low-molecular-weight products were removed by extraction with acetone and then with hexane. The remaining polymeric fraction was then dissolved in a small amount of toluene, precipitated in methanol, filtered, and finally dried in a vacuum. Characterization. PA1 and PA2. All spectroscopic and analytical data concerning these polymers can be found in ref 27. PA4. 2.5 mmol of 3,5-dibromopyridine and 2.5 mmol of bis[(4′-butylpheny)-4-aminophenyl]-4″-tert-butylphenylamine were used for the polycondensation reaction. 1H NMR (400 MHz, C6D6) δ, 8.40 (d, J = 2.4 Hz, 2H), 7.30 (t, J = 2.4 Hz, 1H), 7.21−7.19 (m, 4H), 7.08 (d, J = 8.8 Hz, 4H), 6.98 (s, 8H), 6.91 (d, J = 8.4 Hz, 4H), 2.41 (t, J = 7.8 Hz, 4H), 1.49− 1.41 (m, 4H), 1.27−1.21 (m, 13H), 0.85 (t, J = 7.4 Hz, 6H). 13 C NMR (100 MHz, C6D6) δ, 145.7, 145.5, 144.8, 144.0, 141.7, 138.2, 137.2, 129.6, 126.5, 126.0, 125.0, 124.7, 124.2, 120.7, 35.3, 34.2, 33.9, 31.4, 22.6, 14.0. Anal. Calcd for C47H50N4: C, 84.14; H, 7.51; N, 8.35. Found: C, 83.42; H, 7.38; N, 8.17. Mn = 8.9 kDa, Mw = 104.0 kDa. PK1. 2.5 mmol of dibromobenzene, 2.5 mmol of 3,6-bis(4′methoxyphenyl)amino-9-hexyl-carbazole, and 0.075 mmol of Pd2(dba)3 and 0.225 mmol of t-Bu3P were used for the polycondensation procedure. 1H NMR (400 MHz, C6D6) δ, 7.91 (s, 2H), 7.39 (d, J = 8.8 Hz, 2H), 7.23−7.17 (m, 5H), 7.08−7.00 (m, 2H), 6.97 (d, J = 8.8 Hz, 2H), 6.76 (d, J = 7.9 1757
DOI: 10.1021/jp511087j J. Phys. Chem. B 2015, 119, 1756−1767
Article
The Journal of Physical Chemistry B Computational Methods. Quantum chemical studies, and more precisely density functional theory calculations, were limited to PA1, PA4, and PK1 as representative examples of the two types of polymers studied. The first step within the computational approach was a conformational search of the elemental subunits of polymers with the molecular mechanics (Amber force field). An approach based on modification of selected torsion angles was used, as implemented in the HyperChem75 program. This first scrutiny enabled the most stable conformers of each sequence of polymers to be determined. Then, these optimized structures were used as input geometries for density functional theory (DFT) calculations, performed with the B3LYP functional, using the 6-31+G(d,p) basis set. Geometry optimizations and harmonic frequency calculations were carried out with the Gaussian 09 package.28 It should be stressed that the calculations here presented were performed in vacuo. However, from the reasonable agreement between the experimental and calculated frequencies, more expensive quantum chemical calculations were estimated unnecessary.
Figure 1. XPS (C 1s, N 1s, and O 1s) spectra of pristine polymers: (a) PA1, (b) PA4, and (c) PK1.
assigned to carbons of the methoxy group. In the N(1s) region, the peak assigned to nitrogen atoms of the triarylamine unit is located at 399.8 eV (fwhm = 1.05 eV). As expected, two additional components are observed in the spectra of PA4 and PK1, i.e., at 399.2 eV (numbered 2) and 400.5 eV (numbered 3), respectively. Unambiguously, they can be assigned to the nitrogen atoms in the pyridine ring and in the alkylcarbazole group. In the O(1s) spectra of PA1 and PA4, they can be attributed to surface contamination, since these molecules do not contain oxygen. The methoxy group in PK1 gives rise to a pronounced, narrow peak at 533.4 eV (fwhm = 1.5 eV). Another point of interest, the π−π* shakeup satellite peaks around 291.5 eV, a characteristic of aromatic or conjugated systems, constitute about 2.4−2.7% for PA1, PA4, and PK1, and significantly increase (4.0%) in the spectrum of PA2, which may indicate a higher conjugation in this polymer (see Figure S2, Supporting Information). Finally, Table 1 lists the surface
■
RESULTS The chemical structures of the studied polymers are depicted in Chart 1. In order to confirm their molecular structure but also Chart 1. Chemical Structures of the Investigated Polymers Containing Triarylamine Moieties
Table 1. XPS-Determined and Theoretically Calculated Contents in at% (±0.5%) of C, N, O, and Br in Thin Films of PA1, PA2, PA4, and PK1 C (%) PA1 PA2 PA4 PK1
N (%)
O (%)
expt
theo
expt
theo
expt
theo
Br (%)
94.4 94.3 92.6 88.4
94.1 94.1 92.2 88.4
5.6 5.7 7.4 6
5.9 5.9 7.8 6.9
