Article pubs.acs.org/JPCC
Structural Insight into Electrogenerated Chemiluminescence of ParaSubstituted Aryl−Triazole−Thienyl Compounds Jacquelyn T. Price,†,∥ Michelle S. M. Li,†,∥ Allison L. Brazeau,† Danlei Tao,† Guiming Xiang,† Yanhua Chen,† Robert McDonald,‡ Nathan D. Jones,† and Zhifeng Ding*,† †
Department of Chemistry, The University of Western Ontario, London, Ontario N6A 5B7, Canada X-ray Crystallography Laboratory, Department of Chemistry, The University of Alberta, Edmonton, Alberta T6G 2G2, Canada
‡
S Supporting Information *
ABSTRACT: There has been a strong push to develop new thiophene-based monomers to tailor the electrical and optical properties of the resulting polymers. However, the synthesis of these elaborated thienyl compounds is difficult to realize. Here, we report the successful click coupling of thienyl azides and para-substituted aryl alkynes to synthesize eight thiophene-based luminophores intended for electrochemical and electrochemiluminescence (ECL) study. These thiophenes could be separated into two series: monothienyl and bithienyl analogues and further categorized based on the nature of the ligand attachment to their phenyl rings (electron-donating or -withdrawing characteristics: NMe2, OMe, H, or F) on the other side of the triazole bridge. The electrochemical experiments indicated these compounds lacked stability when they were oxidized or reduced, with the exception of those with a dimethylamine ligand attached (quasireversible oxidations). Cyclic and differential pulse voltammetries revealed the redox potentials of these compounds were affected by the extent of the conjugation and the nature of the ligands, while the electrochemical gaps correlated well with the energy differences between the excited and ground state species. ECL in the annihilation route confirmed the weak light-emitting nature of these thiophenes; however, great improvement was made with the use of a coreactant species (benzoyl peroxide or ammonium persulfate). ECL spectroscopy revealed that the excimer or polymeric excited states were more favorable in formation than their monomeric excited states, which was tunable based on the applied potentials. Structural insight into ECL will guide us to discover optimized thiophene-based luminophores.
■
INTRODUCTION
ization, but it has yet to be thoroughly investigated in this context. A current focus of organic materials chemistry is the straightforward design and synthesis of chromophores for optical components in optoelectronics applications.6 One motif is to conjugate a strongly electron-donating moiety (π-linker) to a strongly accepting one. In this sense, thiophenes have been used as the donor in conjugation;7 however, to the best of our knowledge, it has yet to be shown that thiophenes can be incorporated into a luminophore using triazole as the active πlinker that can be active in electrogenerated chemiluminescence or electrochemiluminescence (ECL) for chemosensors.5 Our research group has explored the ECL of four triazole-modified deoxycytidine (dC) nucleosides for a metal-free DNA sensor based on dC in combination with ECL.8 In another work, we synthesized and studied the electrochemical and spectroscopic properties of four potential blue-emitting thienyltriazoles, with the N in the thienyltriazole ring and O in a methoxy or phenol
Since the emergence of “plastic electronics” in the 1970s, thiophene-based conjugated polymers have attracted considerable attention for their potential use in a wide range of devices, such as field effect transistors and organic light emitting diodes.1,2 Considering the valuable applications of polythiophenes, there has been a strong push to develop new thiophene-based monomers to tailor the electrical and optical properties of the resulting polymers. However, the synthesis of these elaborated thienyl compounds usually relies on metalcatalyzed C−C bond forming reactions, which are less than ideal since they typically require expensive catalysts, extensive prefunctionalization (starting materials), and lengthy purifications (column chromatography). Huisgen 1,3-dipolar cycloadditions between an azide and an alkyne as the archetypal click reaction have significantly accelerated the development of libraries of small molecules and macromolecules with a variety of functional groups linked by a 1,2,3-triazole ring via its 1,4positions.3−5 This approach may prove to be a viable and advantageous alternative to the synthesis of desired thienyl monomers with varied functional roles for onward polymer© 2016 American Chemical Society
Received: July 11, 2016 Revised: August 13, 2016 Published: August 25, 2016 21778
DOI: 10.1021/acs.jpcc.6b06904 J. Phys. Chem. C 2016, 120, 21778−21789
Article
The Journal of Physical Chemistry C group attached to the trizole acting as the possible coordination atoms to form metal complexes.9 ECL has been proven as a powerful analytical technique in chemistry,10−19 which involves radicals generated in the vicinity of a working electrode (WE) undergoing highly energetic electron transfers to form excited states that emit light upon returning to the ground state. The annihilation and coreactant ECL systems are the two most common routes in this field of study. Annihilation ECL experiments typically consist of scanning the luminophore between its first oxidation and reduction potentials, which results in the formation of the radical precursors necessary for ECL generation. In the coreactant ECL systems, the applied potential is unidirectional such that the radical precursors necessary for ECL production are supplied from a coreactant (inducing the luminophore to form a radical) and the luminophore. Since they require a smaller potential window, there is an increased likelihood of electron transfer to occur and radical instability becomes less of an issue. This system also allows the ECL study of luminophores that have limited solubility, previously inhibited in annihilation ECL systems.10 There have been only a few reports on thiophene molecules which are capable of generating ECL from our group,8,9,20 and reports on donor−acceptor π-conjugated (D−π−A) systems consisting of thiophene, triazole, and electron acceptor are less common. In the above context, we report the synthesis of eight new compounds with this design via click coupling of 3azidothiophene and 4-azido-2−2′-bithiophene with a variety of aryl acetylenes containing either electron-withdrawing or -donating groups to give novel ECL active 1,4-disubstituted 1,2,3-triazoles (Figure 1). These compounds, in principle, were expected to permit charge transfer between the rings on either
side of the triazole bridge. The new hybrid luminophores were synthesized under ambient conditions and in high yields without resorting to typical C−C bond-forming reactions and did not require column chromatography for purification. Density function theory was carried out to gain insight into the structural−property relationship. Electrochemistry and ECL, in both annihilation and coreactant systems with benzoyl peroxide (BPO), ammonium persulfate (persulfate), and tri-npropylamine (TPrA) were also investigated.
■
EXPERIMENTAL SECTION Synthesis Preparations. The following materials were synthesized according to literature procedures: [1,1′-bis(diphenylphosphino)ferrocene]palladium(II) dichloride,21 tosylazide,22 3-azidothiophene,23 2,4-dibromothiophene, 2-thienylmagnesium bromide,24 4-bromo-2,2′-bithiophene,24 and 4azido-2,2′-bithiophene.23 All reagents were obtained from commercial sources and used as supplied, unless otherwise stated. General Synthetic Procedure. The azide 3-azidothiophene or 4-azido-2,2′-bithiophene (0.50 mmol) and the appropriate arylacetylene (0.50 mmol) were suspended in a 1:1 mixture of H2O and t-BuOH (3.0 mL total) in a 25 mL round-bottom flask. Sodium ascorbate (0.25 mmol) and CuSO4 (0.08 mmol) were added. The mixture was stirred for 24 h, during which time a brown solid precipitated from solution. The mixture was diluted with H2O (10 mL) and filtered. The solid residue was then washed several times with H2O (10 mL) and HCl (5 mL, 5% v/v) and finally with Et2O (10 mL) to yield the desired product (Scheme 1). Characterization of 1−8 is summarized in the Supporting Information (Figures S1− S16). Scheme 1. General Procedure for the Synthesis of Thiophene Compounds 1−8 via Cu(I)-Catalyzed Huisgen Cycloaddition Reactions
Crystal Structures. Crystals of 3, 5, and 6 suitable for X-ray diffraction analysis were grown at 22, 25, and 22 °C, respectively, from CH2Cl2/Et2O over a period of 1 week. Electrochemical Preparations. Cyclic voltammetry (CV), differential pulse voltammetry (DPV), and ECL experiments were conducted using a 2 mm Pt disc inlaid in a glass sheath as the WE, and two coiled Pt wires were used as the counter electrode (CE) and quasireference electrode (RE), respectively. Prior to each experiment, the WE was polished on a felt polishing pad (Buehler Ltd., Lake Bluff, IL) using 1.0, 0.3, and 0.05 μm alumina suspensions (Buehler Ltd.) in ultrapure water (18.2 MΩ.cm, Milli-Q, Millipore) consecutively for 5 min each to obtain a mirrorlike surface. This was followed by washing of
Figure 1. Molecular structures of thiophene-containing 1,2,3-triazole luminophores 1−8. These thiophenes can be separated into 2 series, (A) monothienyl compounds (1−4) and (B) bithienyl compounds (5−8), and further categorized based on its placement in the series; i.e., 1 and 5, 2 and 6, 3 and 7, and 4 and 8 have the same ligand attachment. 21779
DOI: 10.1021/acs.jpcc.6b06904 J. Phys. Chem. C 2016, 120, 21778−21789
Article
The Journal of Physical Chemistry C
ECL pulsing experiments were conducted using a potentiostat (model AFCBPI, Pine Instrument Co., Grove City, PA), an EG&G PAR 175 Universal Programmer (Princeton Applied Research, Trenton, NJ), and the PMT/picoammeter setup. This assembly allowed pulsing experiments to occur without a delay, at a relatively fast time pace (10 Hz). The data acquisition for the current, potential, and ECL signals was carried out using another homemade LabVIEW program (ECL_PAR.vi, National Instruments). For annihilation systems, the applied potential at the WE was pulsed between the first oxidation and reduction peak potentials, with a pulse width of 0.1 s or 10 Hz. For coreactant systems, the applied potential was pulsed in the cathodic region (BPO and persulfate) between the initial potential and the first reduction potential of 1−8 or in the anodic region, in the case of TPrA, between the initial potential and the first oxidation peak of 1−8. Accumulated ECL spectra were obtained by placing the electrochemical cell onto a spectrometer (Cornerstone 260, Newport, Canada) attached to a CCD camera (model DV420BV, Andor Technology, Belfast, UK). The camera was cooled to −55 °C prior to use and controlled by a computer for operation and data acquisition. Similar to the pulsing experiments, the samples were pulsed at 10 Hz within each compound’s redox potential window. The exposure time of the spectra was set to 300 s. Spooling ECL spectra were collected with an Acton 2300i spectrometer (Princeton Instruments, Trenton, NJ) with a CCD camera cooled to −65 °C (Andor Technology, model DU401-BR-DD-352) by gradually scanning between the initial potential and the first potential that resulted in ECL light emission. Collection parameters for the spooling spectra varied depending on the experimental conditions: exposure time and kinetic series lengths were optimized to produce the clearest ECL spectra. During all experiments, lights and computer monitors in the experimentation room were switched off to reduce the background interference. Blackout curtains were also positioned at the entryways to the lab and surrounding the electrochemical cell setup to prevent light interference. Wavelength calibration was accomplished using a mercury−argon source (Ocean Optics, HG-1). The data acquisition was carried out by means of an Andor Technology program. ECL Efficiency Calculations. ECL quantum efficiencies (Φx) were calculated relative to 9,10-diphenylanthracene (DPA) or tris(2,2′-bipyridine)ruthenium(II) hexafluorophosphate (Ru(bpy)32+) systems by taking their ECL efficiency as 100% or 1.0 in DMF.27,28 This was done by taking the sum of the integration of both the ECL intensity and current values (versus time) for each compound against the standards, as described in eq 129−31
the WE with copious amounts of ultrapure water. The WE was then electrochemically polished in 0.1 M H2SO4 for 400 segments between the approximate potentials of 1.400 V and −0.600 at 0.5 V s−1 to obtain a clean and more reproducible Pt surface.11 The WE was then washed repeatedly with ultrapure water and dried with Ar gas (ultrahigh purity, >99%, Praxair Canada Inc., London, ON). The CE and RE were rinsed with acetone, followed by deionized water. They were then sonicated in N,N-dimethylformamide (DMF, 15 min), ethanol (5 min), and deionized water (5 min) before thorough rinsing with ultrapure water. These electrodes were dried at 120 °C for 5 min and then left to cool to room temperature. Prior to experiments, the electrochemical cell was rinsed with acetone and deionized water, followed by immersion in a base bath of 5% KOH in isopropanol for 4 h. The cell was rinsed with copious deionized water and immersed in an acid bath of 1% HCl for 4 h. Following the acid bath, the electrochemical cell was rinsed thoroughly with ultrapure water, dried at 120 °C for 12 h, and then cooled to room temperature. This cleaning method was used since it provides a more thorough cleaning of our electrochemical cell. This electrochemical cell was specifically designed to have a flat Pyrex window at the bottom to allow the detection of generated light (ECL) from the WE. Solutions of 1−8 with a concentration range between 2.0 × 10−3 and 2.7 × 10−3 M in DMF containing 0.1 M tetranbutylammonium perchlorate (TBAP) as supporting electrolyte were prepared in a N2-filled glovebox. The assembly of the electrochemical cell was sealed using a custom-made Teflon cap with a rubber O-ring and was removed from the glovebox to perform electrochemistry and ECL experiments. For coreactant ECL studies, 5 mM BPO, 25 mM persulfate, or 25 mM TPrA was added to each solution. After completion of each experiment, the electrochemical potential window was calibrated using ferrocene (Fc); the redox potential of Fc+/Fc was taken as 0.470 V vs SHE.25 All potentials reported in this report are vs SHE, unless otherwise stated. Electrochemical Instrumentation. All electrochemical experiments were conducted on a CHI 610A electrochemical analyzer (CH Instruments, Austin, TX). The experimental parameters for CV were: initial potential of 0.000 V, positive or negative initial scan polarity, 0.1 V s−1 scan rate, four sweep segments, 0.001 V sample interval, 2 s quiet time, and (1−5) × 10−5 AV−1 sensitivity. The potential window was dependent on each individual compound’s redox potentials.11 Two DPV experiments26 were also conducted for each compound: one for the anodic scan and one for the cathodic scan, based on the redox potentials found in CV. The experimental parameters for DPVs are 0.004 V potential intervals, 0.05 V amplitude, 0.5 s pulse width, 0.0167 s sampling width, 0.2 s pulse period, 2 s quiet time, and (1−5) × 10−5 AV−1 sensitivity.11 ECL Instrumentation. For ECL measurements, the ECL− voltage curve and the CV were recorded simultaneously using a custom-made LabVIEW program (ECL_PMT610a.vi, National Instruments, Austin, TX). The ECL intensity was detected as a photocurrent by a photomultiplier tube (PMT, R928, Hamamatsu, Japan), held at −750 V with a high voltage power supply and transformed to a voltage signal using a picoammeter/voltage source (Keithley 6487, Cleveland, OH). The sensitivity on the picoammeter was set manually in order to avoid photosaturation. The ECL/CV signals were sent simultaneously through a DAQ board (DAQ 6052E, National Instruments, Austin, TX) in the computer workstation.
Φx = 100% ×
⎛ ∫ b ECL dt ⎞ ⎜ ba ⎟ ⎝ ∫a Current dt ⎠x
⎛ ∫ b ECL dt ⎞ ⎜ ba ⎟ ⎝ ∫a Current dt ⎠st
(1)
where x and st represent sample and standard, respectively. Equation 1 is based on the principle of generated photons per electron. ECL experiments for each compound were tested with a minimum of five different potential windows based on their redox potentials, tuning for the strongest ECL activity. The Φx 21780
DOI: 10.1021/acs.jpcc.6b06904 J. Phys. Chem. C 2016, 120, 21778−21789
Article
The Journal of Physical Chemistry C
Figure 2. Front (left) and side (right) prospectives of the ball-and-stick representations of 3, 5, and 6 are shown as (A), (B), and (C), respectively. Note that the monomeric units of 3 and 5 are more planar in orientation, while 6 is nonplanar containing disorder between the triazole and bithiophene rings. The building block in the crystal structure of 5 was found to be a trimeric unit.
relatively short N4−C14 bond length present, 1.387(3) Å, typical for the wide variety of reported donor−acceptor chargetransfer compounds containing the dimethylaniline group (e.g., 3-dimethylamino benzoic acid is 1.385(3) Å).34 This suggests that there is the possibility of a significant contribution of the charge-separated quinoid resonance form to the ground-state structure. However, the C−C single bond connecting the aniline and triazole rings (C(2)−C(11)) was found to be 1.468(4) Å, which is standard for a C−C single bond. The monomeric unit of compound 5 (CCDC# 1498983) was also essentially planar in the solid state, as revealed by the side view (Figure 2B). The solid state structure of 5 had three crystallographically independent, but similar molecules in its unit cell; we will get back to this point later in the ECL spectroscopy section. In contrast, the solid state structure of 6 (CCDC# 1498981, Figure 2C) revealed, while well-defined, it was not entirely planar: a torsional angle of 17.1(9)° exists between the two contiguous thiophene rings. Both the monomeric units of 5 and 6 contain one terminal thiophene that was disordered over two positions (Figure 2B and 2C, respectively). Electrochemistry and Its Correlation to Crystal and Electronic Structures. The electrochemical behaviors of 1−8 were first characterized by CV to determine their oxidation and reduction potentials required for ECL activity. Figure 3A shows a typical cyclic voltammogram of a thiophene luminophore (3) in DMF solution containing 0.1 M TBAP as the supporting electrolyte. When the potential is anodically scanned from the initial potential to 1.36 V, 3 underwent a reversible oxidation at 1.23 V (anodic peak) followed by its corresponding cathodic peak appearing at 1.13 V on the reverse scan. With further cathodic scanning, the reduction of 3 can be seen as an irreversible process with a peak potential of −1.74 V. Note that the voltammogram (Figure 3A) appears bumpy in the center of the scan; this is likely due to the electrochemical-active intermediates generated by the radicals undergoing chemical reactions in solution. In order to assess the redox potentials more clearly, DPV was conducted. For example, the DPV of 3 (Figure 3B) in the potential range of 1.51 and −2.01 V resolves the cathodic and anodic peaks more clearly than in the CV (Figure 3A). The formal redox potentials can be calculated from their peak
of 1−8 is listed as a range, identifying the minimum and maximum Φ. Theoretical Calculations. Optimized geometries and vibrational frequencies were calculated using density functional theory (DFT)32 with B3LYP/6-31+G** basis set using the Gaussian 03 suite of programs33 on SHARCNET. Representations of the HOMO and LUMO were determined using the cubegen utility. Spectroscopy. UV−visible absorption spectra were recorded over a range of 265−600 nm using a Varian Cary 50 spectrophotometer (Varian Inc., North Carolina), while photoluminescence spectra were recorded on a spectrometer (QM-7/2005, Photon Technology International, London, ON). The emission, excitation, and steady-state lamp slits were set to 0.4, 0.5, and 0.5 mm, respectively, and the integration time was set to 0.1 s. All solutions were prepared in DMF and analyzed in quartz cuvettes with a path length of 1 cm.
■
RESULTS AND DISCUSSION Synthesis. Thiophene azide precursors were synthesized using slight modifications from established procedures.23 These azides were successfully coupled to various aryl acetylenes using Cu(I)-catalyzed Huisgen cycloaddition reactions to synthesize 1−8 in high yields, ranging between 71 and 98% (Scheme 1, see SI for more details).3 The monothienyl thiophene species (1−4) were yellow, while their bithienyl analogues (5−8) were brown in color. The identities of 1−8 were confirmed by 1H and 13C NMR as well as HRMS (SI, Figures S1−S16). The 1H NMR spectra of 1−8 contained a diagnostic singlet at ca. 9 ppm, arising from the lone triazole proton. The successful synthesis of these triazoles was easily identified by the appearance of this peak. Crystal Structures. The ball-and-stick representations of 3, 5, and 6 are shown in Figure 2 (see Figure S17 for ORTEP representations). 3 (CCDC# 1498982) was the only thienylsubstituted triazole in this work that did not show any disorder of the thiophene ring (Figure 2A). Very small intermolecular torsion angles within 3 indicated that it was the most planar of the other thiophene compounds obtained (Supporting Information). We believe this is due to the NMe2 ligand, which allows resonance to occur. This is also supported by the 21781
DOI: 10.1021/acs.jpcc.6b06904 J. Phys. Chem. C 2016, 120, 21778−21789
Article
The Journal of Physical Chemistry C
Table 1. Determined Formal Potentials and Electrochemical Gaps (ΔEgap) of 1−8 formal redox potentials (V) compound
Eox0′
Ered0′
electrochemical energy gap (eV)
1 2 3 4 5 6 7
1.46 1.48 1.14 1.36 1.78 1.65 0.43a 0.96b 1.41
−1.44 −1.48 −1.77 −2.09 −1.63 −1.33 −1.69
2.90 2.96 2.91 3.45 3.41 2.98 2.12
−1.36
2.77
8 a
First oxidation potential. bSecond oxidation potential.
(LUMO). It is from the HOMO that the electrons are removed during the oxidation process, while the LUMO is where the electrons will be injected into upon electrochemical reduction. Therefore, it is deduced that both the oxidation and reduction of 1−8 will have some dependency on the nature of the substituent attached to the phenyl ring. Figure 3C, for example, illustrates that the electron density of the HOMO for 3 is distributed mainly across the triazole and benzene rings, as well as the NMe2 ligand. However, the electron density of the LUMO is distributed mainly over the thiophene and triazole rings. This image also suggests that upon injection of an electron during the reduction of 3 bond breaking can occur. This would most likely occur between the thiophene and triazole rings, rendering the instability of the radical anion. This is also confirmed through the comparison of the molecular conjugation between 3 and 1 (Figure S18). Effects of Ligand Attachment and Conjugation on Redox Potentials. Compounds 1−8 can be grouped into two series, the monothienyl analogues (Series A: 1−4) and bithienyl analogues (Series B: 5−8). Similarly, we have organized these compounds to have the same ligands depending on its placement in the series; i.e., 1 and 5, 2 and 6, 3 and 7, and 4 and 8 have the same ligands. These compounds mostly exhibited irreversible oxidation and reduction peaks (Table 1 and Figure S19), which suggested that both the radical cations and anions were unstable in solution, leading to the minimal ECL generation. The only exception to this was for the thiophenes containing the dimethylamine ligands attached to the phenyl group (3 and 7). These compounds underwent reversible oxidations, which resulted in more stable radical cations forming. It is likely that the introduction of these dimethylamine ligands to the thiophene luminophores allows for this stabilization upon oxidation (resonance). The bithienyl 7 underwent two consecutive quasi-reversible oxidation reactions with the lowest potential among the eight compounds (0.43 and 0.96 V, respectively), while the monothienyl 3 underwent its reversible oxidation reaction at a formal potential of 1.14 V (Figures 3 and S19). The additional thiophene ring in 7 is believed to have lowered the oxidation potential due to the extended πconjugation present, likewise lowering the reduction potential required to drive this reaction (−1.69 vs −1.77 V). The other analogues with an electron-donating ligand attached to the phenyl rings are 4 and 8, which have the methoxy ligand attached. These compounds are oxidized at slightly lower potentials (with respect to the other thiophenes):
Figure 3. (A) CV and (B) DPV of 3 in DMF solution with 0.1 M TBAP as the supporting electrolyte. In the CV, the circled cross indicates the starting potential. The arrows illustrate the scan direction, while for DPV, the arrows indicate the anodic (top trace) and cathodic (bottom trace) scan directions. (C) The molecular orbital isosurface plots for LUMO and HOMO of 3 based on the DFT/B3LYP/631+G** calculations.
potentials (Emax) and pulse heights (ΔE), using the following equation26 ΔE (2) 2 where ΔE is sign sensitive, i.e., positive in the anodic scan and negative in the cathodic scan. Using the DPV of 3 (Figure 3B) and eq 2, the first oxidation and reduction potentials can be calculated to be 1.14 and −1.77 V, respectively. The electrochemical energy gap (ΔEgap) for 3 was determined to be 2.91 eV from the product of electron and the difference between the formal redox potentials. In a similar fashion, the redox potentials and ΔEgap of the other compounds were assessed and summarized in Table 1 and visualized in Figure S19. As both the CV and DPV show the first oxidation peak of 3 as a reversible process (Figures 3A and B), the resulting radical cation required for ECL activity should be highly stable. This can be due to the extended conjugation or the introduction of a strong electron-donating group (NMe2) to the thiophene luminophore.35 The charge introduced from losing an electron is likely delocalized across the entire molecule as opposed to the breaking of a chemical bond. From its crystal structure (Figure 2A), 3 displays planar orientation which is further evidence of the molecular conjugation. We can further investigate this theory of molecular conjugation via DFT/B3LYP/6-31+G** calculations (Figures 3C and S18). These computations allow the visualization of the electron density across the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital E 0 ′ = Emax +
21782
DOI: 10.1021/acs.jpcc.6b06904 J. Phys. Chem. C 2016, 120, 21778−21789
Article
The Journal of Physical Chemistry C
Similarly, the other compounds displayed very low ECL intensity in their respective annihilation ECL systems (
