Theoretical study on the emeraldine salt – impact of the computational protocol
Introduction
Polyaniline (PANI) and its intriguing properties has been known to mankind for many years [1], [2]. It occupies a dominant place in the area of conducting polymers, the flourishing development of the latter being awarded with the Nobel Prize in Chemistry in 2000 [3]. The formulation of doped conjugated polymers (the so-called ‘fourth generation polymeric materials’) opened new opportunities for innovation in various modern fields, e.g., microelectronics, due to the unique combination of conductivity and plasticity. The main difference between polyaniline and the rest of the conducting polymers is that it exists in a variety of oxidation states – leukoemeraldine, pernigraniline and emeraldine, each of them possessing specific optical and conducting characteristics. A special feature of polyaniline is the unique method for obtaining of the highly conducting emeraldine salt – protonic acid doping, which is a reversible process depending on pH of the reaction medium. This tunable doping method opens new horizons for application of PANI.
Ever since its rediscovery, polyaniline with its unconventional and attractive properties challenges the theoretical community. Till the end of the last century, the theoretical approaches for simulation of polyaniline were based mostly on semi-empirical geometry optimization of oligomers or semi-empirical band structure calculations of the polymer [4], [5], [6], [7], [8], [9], [10], [11], [12], [13]. They reported theoretical estimates of the optical and geometric characteristics of different oxidation states and defect structures originating from protonic acid doping of polyaniline, comparing the theoretical predictions to experimental data. The effect of valence and torsion angles on PANI properties was discussed as well. In the same period, an adequate molecular model for description of conducting polyaniline, namely, the emeraldine salt, was formulated. Stafström еt al. [6] demonstrated two different possible configurations for protonation of emeraldine base with hydrochloric acid. One of them involves protonation of the imine nitrogens of the base. In this case the elementary unit of the polymeric emeraldine salt is a magnetically inactive tetramer with two positive charges – the bipolaron form. However, it does not reproduce correctly the EPR and UV–vis spectra of the experimentally obtained highly conducting hydrochloride (101 Ω−1 cm−1) [14]. Therefore, the authors exploit a second model of the salt where the elementary unit is reduced to an aniline dimer represented by a cation-radical and thus bearing a unit positive charge – the polaron form. The latter theoretical model reproduces the experimental characteristics of the polymer and has been acknowledged as accurate representation of the highly conducting form.
Since the beginning of this century, intensive investigation of polyaniline with Density Functional Theory methods has been initiated. This change with respect to the standard computational protocol at that time was dictated by the possibility for ‘time saving’ description of larger molecular models of the polymer and by the fact that non-dynamic correlation is taken into account by DFT. The latter is of utmost importance for simulation of conducting and optical properties of conjugated systems. In this period appeared theoretical studies encompassing different theoretical schemes applied to identical polyaniline models [15], [16]. Thereafter, the theoretical treatment of this attractive polymer was based exclusively on DFT calculations and even on their combination with CPMD. Moreover, some communications offered new model forms of the conducting emeraldine salt [17], [18], [19], [20], [21], [22], [23], [24], [25], [26]. All publications proved that DFT is sufficiently accurate for addressing geometry and other properties of polyaniline.
There exist also a number of theoretical reports, which consider the solvent effect and the influence of other factors from the medium, such as monomer excess, on PANI properties [13], [27], [28], [29], [30]. A third group of papers tackles the mechanism of aniline polymerization [31], [32].
In spite of the enormous interest towards studies on the conducting emeraldine salt, no detailed overview of the capacity of the various quantum chemical methods for its description is available. This is the main goal of the present study.
In our view, a suitable theoretical method and model of emeraldine salt should meet the following requirements:
- 1.
Yield precise geometric characteristics – bond lengths, valence and dihedral angles, comparable to the accessible experimental data.
- 2.
Take into account correlation effects.
- 3.
Describe accurately all possible multiplets of the emeraldine salt and their relative stability, i.e., produce wavefunctions without spin contamination, the latter contributing artifacts to the outcome.
- 4.
Consider explicitly the presence of counterions, allowing detailed description of the PANI-dopant interaction.
- 5.
Include medium effects such as solvent polarity, which are proven to be essential for the behaviour of the studied system.
- 6.
Guarantee transferability of the model to oligomers of different size.
The search for an appropriate computational protocol is usually validated on the smallest possible systems representative of the real target of study. In our case this is the tetramer. In order to accomplish our goal, we performed geometry optimization with HF and various DFT functionals of NH2-terminated tetramer model of emeraldine hydrochloride in singlet (bipolaron) and triplet (two polarons) form (Fig. 1). Protonation was interpreted as introduction of defect(s) in the oligomer chain and the distance to which it propagated along the chain – an important characteristic of such localized states in conducting polymers [33], [34], was estimated for models of octa-, dodeca-, and hexadecamer.
Section snippets
Model and method
In order to select the proper theoretical method, we carried out full geometry optimization in vacuum and in PCM (water) of tetramer model systems of emeraldine salt (Fig. 1) in singlet (bipolaron) and triplet (two polarons) state with the following methods: HF/6-31G∗, B3LYP/6-31G∗, BLYP/6-31G∗ and BH&H/6-31G∗. The singlet forms were modeled with the restricted versions of the methods and the triplet oligomers – with the unrestricted ones. Each of the optimized structures was subject to
Structural parameters
The bond lengths of the fully protonated singlet tetramer obtained after geometry optimization with HF and DFT and basis set 6-31G∗ in vacuum and in PCM are collected in Table 1. The bonds in Rings 2–4 (Fig. 1) were divided in two subgroups – along the oligomer axis (further on denoted as ‘shorter’) and in transversal direction (denoted as ‘longer’) and were averaged separately.
The HF calculations in vacuum and PCM show mild influence of ‘doping’ on the bond length distribution along the
Conclusion
Based on a detailed analysis, it can be concluded that the HF method is inappropriate for theoretical modeling of polyaniline: it does not describe adequately the effect of protonation in the emeraldine salt, localizes strictly the bipolaron and predicts heavily disturbed conjugation along the oligomer chain. These results contradict the available experimental data [38]. Moreover, the incapability to tackle various multiplicities and the inherent spin contamination disqualify the HF method as
Acknowledgements
Projects BYX-202/06, DO-02-136/08 and DO-02-82/08 of the Bulgarian National Science Fund and Grant No. SU-185/2009 of the Science Fund of the University of Sofia are acknowledged for the financial support; the participation grant (J.R.) from the Organizing Committee of DFT09 and the PhD ‘co-tutelle’ fellowship of the French Government are appreciated.
References (50)
- et al.
Synth. Metals
(1986) - et al.
Synth. Metals
(1986) Synth. Metals
(1991)- et al.
Synth. Metals
(1995) - et al.
Synth. Metals
(1998) - et al.
Chem. Phys.
(2000) - et al.
Synth. Metals
(1987)et al.Solid State Commun.
(1991)et al.Phys. Rev. B
(1994) - et al.
Phys. Rev. B
(2006) - et al.
J. Chem. Phys.
(2008) - et al.
Phys. Chem. Chem. Phys.
(2004)
ChemPhysChem
J. Phys. Chem. B
Phys. Chem. Chem. Phys.
Acc. Chem. Res.
J. Chem. Soc.
J. Am. Chem. Soc.
J. Phys. Chem. B
Phys. Rev. Lett.
Phys. Rev. B
Phys. Rev. B
J. Chem. Phys.
J. Phys. Chem. B
Phys. Rev. B
J. Phys. Chem. A
Cited by (26)
DFT study of iminodiacetic acid functionalised polyaniline copolymer interaction with heavy metal ions through binding energy, stability constant and charge transfer calculations
2021, Computational and Theoretical ChemistryCitation Excerpt :The bond lengths between carbon atoms (C-C), carbon-nitrogen (C-N), carbon-hydrogen (C-H) and nitrogen- hydrogen (N-H) and bond angles between the phenyl rings were noted. The polyaniline structure is well studied with DFT simulations by taking three, four, eight monomer units. [28–31] In the present study, we have considered the tetramer of aniline.
Polyaniline emeraldine salt as an ammonia gas sensor - Comparison of quantum-based simulation with experiment
2019, Physica E: Low-Dimensional Systems and NanostructuresCitation Excerpt :In comparison to other gases, polyaniline should be especially sensitive to ammonia gas. The molecular and electronic properties of different forms of polyaniline were analyzed using quantum mechanical calculations in Refs. [7,10]. Reis et al. [11] calculated band structure, charge density, transmittance and electrical current of polyaniline using the Density Functional Theory (DFT), with the generalized gradient approximation (CGA) of Perdew–Burke–Ernzerhof (PBE), for the exchange-correlation energy.
Time-dependent density functional theory study on the excited-state hydrogen-bonding characteristics of polyaniline in aqueous environment
2017, Spectrochimica Acta - Part A: Molecular and Biomolecular SpectroscopyCitation Excerpt :Matveeva and co-workers classified the PANI-water interaction into two certain types adsorbing on amine and imine nitrogen atoms, respectively. [25] Romanova et al. [27,28] proved that the protons bound tighter to the chains lead to stronger impact of the dopant on the structural parameters of PANI, and the charge density distribution was more realistic in aqueous environment compared with vacuum estimates. Canales et al. [29] studied the water absorption behavior of PANI by calculating the excess chemical potential, and predicted that the maximum content of water absorbed by PANI was 15% which was consistent with the experimentally measured value (14%).
DFT estimates of water environment impact on the reversible 2e<sup>-</sup> + 2H<sup>+</sup> oxidation of aniline tetramer
2016, Computational and Theoretical ChemistryCitation Excerpt :However, in the context of the observed electrochemical process of polyaniline, the quantities calculated in the gas phase were, in many aspects, far from the observed voltammetric characteristics. Since polyaniline is always obtained in solution and its influence on PANI properties is reported both experimentally [11–22] and theoretically [23–25], the importance of solvent for PANI calculation is clear. However, for the extended redox system ab initio dynamic calculations with explicit solvent molecules are not accessible.
Recent advances in polyaniline research: Polymerization mechanisms, structural aspects, properties and applications
2013, Synthetic MetalsCitation Excerpt :The results of a theoretical investigation with the polarisable continuum model showed that at the tetraaniline level the bipolaron form is always preferred, in accordance with available experimental results [380]. A theoretical study on the geometry characteristics and the electronic properties of bipolaron defects introduced in ES form of OANIs (tetraniline to hexadecaaniline) was reported by Romanova et al. [381]. Dane et al. recently reported a simple and effective ionic self-assembly approach to prepare thin tetraaniline and octaaniline films with well-defined structures [382].
N ⋯ H