Highly crystalline and fluorescent BODIPY-labelled phenyl-triazole-coumarins as n-type semiconducting materials for OFET devices

Abstract

In this work, the synthesis of BODIPY-phenyl-triazole labelled coumarins (BPhTCs) using a two-step approach is described. The influence of the BODIPY appending on the photophysical, electrochemical and thermal properties of the phenyl-triazole-coumarin precursors (PhTCs) was investigated. Band gap energies were measured by absorption spectroscopy (2.20 ± 0.02 eV in the solid and 2.35 ± 0.01 eV in solution) and cyclic voltammetry (2.10 ± 0.05 eV). The results are supported by DFT calculations confirming the presence of lowest LUMO levels that facilitate the electron injection and stabilize the electron transport. Their charge-transport parameters were measured in Organic Field-Effect Transistor (OFET) devices. BPhTCs showed an ambipolar transistor behavior with good n-type charge mobilities (10⁻² cm²V⁻¹s⁻¹) allowing these derivatives to be employed as promising semiconducting crystalline and fluorescent materials with good thermal and air stability up to 250 °C.

1. Introduction

BODIPY derivatives, acronym of boron-dipyrromethene (4,4-difluoro-4-bora-3a,4a-diaza-s-indacene), have lately attracted wide research interest due to their remarkable photophysical properties: absorption-emission bands in the near UV–vis region, high fluorescence quantum yields (Φ_f) with an average lifetime between 1 and 10 ns, important molar absorption coefficients, and distinctive photochemical stability [[1], [2], [3]]. Consequently, the BODIPY unit has recently found potential application in a wide variety of fields such as laser technology, solar cell industry [[4], [5], [6], [7]], bioimaging acquisition [[8], [9], [10]], photodynamic therapy [11,12], viscosity measurements [13,14], chemosensors [15], and electroluminescent devices [16], among other applications. BODIPY derivatives constitute a new generation of outstanding functional materials with characteristics suitable to be used as organic semiconductor (OS) materials [[17], [18], [19]]. They show mechanical flexibility and good charge mobilities (10⁻⁵-10⁻² cm²V⁻¹s⁻¹) [5,20]. In the last decade, there has been a growing interest to develop BODIPY derivatives for optoelectronic applications [21], most of them focused on polymers. In this regard, a copolymer conjugated with thiophene-substituted BODIPY [22] is considered as the OS-BODIPY with the highest charge mobility (0.17 cm²V⁻¹s⁻¹). As far as BODIPY-based small molecules for OTFTs are concerned, their development has been more limited but with results equivalent to BODIPY-polymeric materials (0.01 cm²V^{- 1} s^{− 1}) [23].

The optical and electronic properties of BODIPYs are mostly relied on the type of substituent and the position at which they are attached (meso, α and β). Therefore, the BODIPY has been coupled to other derivatives with intrinsic biological or optical properties to increase scope for synergy. In this sense, coumarin compounds show high fluorescence quantum yields and stability under UV light [24]. Push-pull arrangement is developed in the coumarin system, through the combination of electron-withdrawing (EW) and electron-donor (ED) substituents on C-3 and C-7, respectively. The strong polarization of this combination of opposites allows tuning of their absorption-emission profiles [25]. Then, BODIPY-coumarin hybrids (BC) exhibit large bathochromic and Stokes shifts with improved quantum yields [[26], [27], [28]]. Besides, α,β−unsaturated compounds [29,30], amides [31], indole derivatives [32], phenyl [33], carbazole [34], and thiophene derivatives [35] have been used to link BODIPY and coumarin fluorophores to achieve specific properties.

On the other hand, BODIPY-triazole hybrids (BT) are recognized because of their rigidity, aromaticity, high chemical stability, and polarity as well as their capability of hydrogen bonding association [36]. The enhanced photo-physical properties of BT hybrids allowed their application as Hg²⁺, Ag⁺ and human leukocyte elastase fluorescent chemosensors [[37], [38], [39], [40]], for light harvesting [41], and in vitro labelling of cells and glycoproteins [42].

In addition, hybrids of triazole-coumarin moieties (TC) exhibit improved optical properties which can be modulated by the electronic nature and position on the coumarin substituents [[43], [44], [45]]. TC bifunctional hybrids have been applied as fluorescent sensors of physiological H₂O₂, H₂S, Cu²⁺, Pd²⁺, as well as a probe for HNO in living cells [[46], [47], [48], [49], [50]].

In addition, hybrids of triazole-coumarin moieties (TC) are known to exhibit improved optical properties tuned by the electronic nature and position on the coumarin substituents [[43], [44], [45]]. TC bifunctional hybrids have been applied as fluorescent sensors of physiological H₂O₂, H₂S, Cu²⁺, Pd²⁺, as well as a probe for HNO in living cells [[46], [47], [48], [49], [50]].

One approach to the novel materials is a combination of known luminophores, then and based on the improved above mentioned photophysical properties of BC, BT and TC hybrids, herein the 1,2,3-triazolyl-phenyl moiety (PhT), as donor (D), is used to connect the BODIPY (B) and coumarin (C) fragments, as the acceptors. Thus A–D dyads were prepared by modulating the acceptor nature of the coumarin unit with the appropriate substituent. Previously, we reported the synthesis, of the precursors 3-(4-formylphenyl)-triazole-coumarins and demonstrated their suitability for optoelectronic applications [51]. Taking advantage of the availability of the starting materials, BODIPY-1,2,3-triazolyl-phenyl-coumarin (BPhTC) hybrids 2a-g, Fig. 1, were herein prepared for the first time. The strong ED substituents (NEt₂, OR (R H, Me, Et)) and those EDs with negative inductive effect (Cl, Br), in positions 6–8 on the coumarin ring, offer the possibility to relate them with optoelectronic properties. It is worth mentioning that the combination of these three heterocycles, has not been reported before neither as luminophores nor as OS materials. The BPhTCs hybrid materials, 2a-g, are meso-substituted BODIPYs, this substitution is expected to lead to n-channel semiconducting characteristics [21], less frequently studied than the p-channel transistor behavior of small molecule OS materials.

Fig. 1.

BODIPY-1,2,3-triazolyl-phenyl-coumarin (BPhTC) hybrids 2a-g.

The synthesis of BPhTC-hybrids 2a-g was performed in two steps. They were spectroscopically, spectrometric, thermally, and electrochemically characterized, and their crystalline nature was established by X-ray diffraction. Furthermore, their photophysical properties, in solution and the solid state, were explored and supported by mechano-quantum calculations (DFT). Finally, films of BPhTCs were characterized by atomic force microscopy and their characteristics as OS materials were tested using organic field-effect transistor (OFET) devices as micro-labs. BPhTCs showed to be crystalline materials with good thermal and air stability suitable as luminophores and OS with good n-type charge mobilities.

2. Materials and methods

All reagents were purchased from Sigma-Aldrich. Aldehydes 1a-g were synthesized as reported [51]. Structural characterization of 1a-g is given in the supplementary material (S1). Standard procedures were followed to dry the organic solvents. Column chromatography was performed using standard silica gel 60. Melting points were determined with an Electrothermal AI 9100 apparatus or a Büchi Melting Point B540 and are uncorrected. FT-IR spectra were recorded on a PerkinElmer Spectrum GX Spectrophotometer and are reported as wavenumber of absorption (cm⁻¹). ¹H, ¹³C, ¹¹B and ¹⁹F NMR spectra were acquired on Varian NMR spectrometers, operating at 300 and 400 MHz, and on Bruker NMR spectrometer Avance III, operating a 750 MHz, using DMSO‑d₆ as solvent unless otherwise is specified. Chemical shifts (δ) are reported in parts per million (ppm) from the residual solvent peak used as reference and coupling constants (ⁿJ) are in Hz. The following abbreviations are used to indicate the multiplicity of the signals: s, singlet (s); d, doublet; t, triplet; q, quadruplet; m, multiplet. Agilent G1969A spectrometer was used to obtain high resolution mass spectra.

2.1. General procedure for the synthesis of BODIPY-1,2,3-triazolyl-phenyl-coumarin hybrids (BPhTCs 2a-g), and their characterization

Compounds 2a-g were prepared according to conventional two-step one pot procedure for BODIPYs synthesis [7,52,53] with modifications. A detailed protocol is described in the supplementary material (S2). NMR and HR-MS spectra can be found in the supplementary material (S3).

meso-(4-(1-(2-oxo-2H-chromen-3-yl)-1H-1,2,3-triazol-4-yl))phenyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (2a).

Orange crystalline solid (60 %). Mp: 281–283 °C. ¹H NMR (300 MHz, δ): 9.30 (s, 1H, H-9), 8.86 (s, 1H, H-4), 8.24 (d, 2H, ³J = 8.0, H-13), 8.18 (s, 2H, H-19), 7.99 (d, 1H, ³J = 7.5, H-5), 7.76 (m, 3H, H-7, H-12), 7.60 (d, 1H, ³J = 8.3, H-8), 7.51 (t, 1H, ³J = 7.5, H-6), 7.14 (m, 2H, H-18), 6.73 (m, 2H, H-17). ¹³C NMR (75.4 MHz, δ): 156.3 (C-2), 153.0 (C-8a), 146.8 (C-10), 146.1 (C-11), 145.2 (C-19), 135.8 (C-7), 134.4 (C-16), 133.5 (C-14), 133.1 (C-15), 133.0 (C-18), 132.2 (C-4), 131.1 (C-13), 130.0 (C-5), 126.1 (C-12), 125.8 (C-9), 123.6 (C-3), 123.5 (C-6), 119.8 (C-17), 118.6 (C-8), 116.8 (C-4a). ¹¹B NMR (128.3 MHz, δ): 0.25 (t, J_B-F = 28). ¹⁹F NMR (376.3 MHz, δ in ppm): 144.64 (q, J_B-F = 28. FT-IR (cm⁻¹): 3039, 2037, 1718, 1611, 1274, 1039, 932, 895, 760, 547. HR-MS (ESI-TOF) m/z calcd. for C₂₆H₁₆N₅O₂BF₂ [M+H]⁺: 480.1437, found 480.1437.

meso-(4-(1-((6-methoxy)2-oxo-2H-chromen-3-yl)-1H-1,2,3-triazol-4-yl))phenyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (2b).

Orange crystalline solid (54 %). Mp: 271–273 °C. ¹H NMR (400 MHz, CDCl₃, δ in ppm): 9.09 (s, 1H, H-9), 8.68 (s, 1H, H-4), 8.11 (d, 2H, ³J = 8.2, H-13), 7.96 (s, 2H, H-19), 7.69 (d, 2H, ³J = 8.2, H-12), 7.41 (d, 1H, ³J = 9.1, H-8), 7.24 (dd, 1H, ³J = 9.1, ⁴J = 2.9, H-7), 7.12 (d, 1H, ⁴J = 2.9, H-5), 7.01 (m, 2H, H-18), 6.57 (m, 2H, H-17), 3.91 (s, 3H, CH₃). ¹³C NMR (100.6 MHz, CDCl₃, δ): 157.0 (C-6), 155.9 (C-2), 147.2 (C-8a), 146.9 (C-10), 146.7 (C-11), 144.2 (C-19), 134.9 (C-16), 133.8 (C-15), 132.9 (C-4), 132.6 (C-14), 131.5 (C-18), 131.3 (C-13), 125.9 (C-12), 123.2 (C-3), 121.3 (C-9), 121.1 (C-7), 118.7 (C-17), 118.5 (C-4a), 117.9 (C-8), 110.4 (C-5), 55.9 (CH₃). ¹¹B NMR (128.3 MHz, CDCl₃, δ): 0.30 (t, J_B-F = 29). ¹⁹F NMR (376.3 MHz, CDCl₃, δ): 145.08 (q, J_B-F = 29). FT-IR (cm⁻¹): 3182, 1724, 1570, 1269, 1134, 795, 700, 657. HR-MS (ESI-TOF) m/z calcd. for C₂₇H₁₈N₅O₃BF₂ [M+H]⁺: 510.1544, found 510.1548.

meso-(4-(1-((7-diethylamino)2-oxo-2H-chromen-3-yl)-1H-1,2,3-triazol-4-yl))phenyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (2c).

Red solid (63 %). Mp: 236–237 °C. ¹H NMR (400 MHz, CDCl₃, δ): 8.97 (s,1H, H-9), 8.51 (s,1H, H-4), 8.09 (d, 2H, ³J = 8.5, H-13), 7.95 (s, 2H, H-19), 7.67 (d, 2H, ³J = 8.5, H-12), 7.45 (d, 1H, ³J = 9.0, H-5), 7.01 (d, 2H, ³J = 4.0, H-18), 6.70 (dd, 1H, ³J = 9.0, ⁴J = 2.5 Hz, H-6), 6.57 (m, 3H, H-8, H-17), 3.47 (q, 4H, ³J = 7.0, CH₂), 1.26 (t, 6H, ³J = 7.0, CH₃). ¹³C NMR (100.6 MHz, CDCl₃, δ): 156.9 (C-2), 155.9 (C-7), 151.6 (C-8a), 146.9 (C-10), 146.4 (C-11), 144.1 (C-19), 134.9 (C-16), 134.6 (C-4), 133.5 (C-15), 133.2 (C-14), 131.5 (C-18), 131.2 (C-13), 130.1 (C-5), 125.8 (C-12), 121.1 (C-9), 118.6 (C-17), 116.8 (C-3), 110.3 (C-6), 107.2 (C-4a), 97.2 (C-8), 45.1 (CH₂), 12.4 (CH₃). ¹¹B NMR (128.3 MHz, CDCl₃, δ): 0.30 (t, J_B-F = 29). ¹⁹F NMR (376.3 MHz, CDCl₃, δ): 145.51 (q, J_B-F = 29). FT-IR (cm⁻¹): 3413, 2347, 1635, 1319, 1261, 1154, 1042, 984, 758, 644, 565. HR-MS (ESI-TOF) m/z calcd. for C₃₀H₂₅N₆O₂BF₂ [M+H]⁺: 551.2173, found 551.2174.

meso-(4-(1-((6-bromo)2-oxo-2H-chromen-3-yl)-1H-1,2,3-triazol-4-yl))phenyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (2d).

Orange crystalline solid (54 %). Mp: 267–268 °C. ¹H NMR (750 MHz, δ): 9.28 (s, 1H, H-9), 8.77 (s, 1H, H-4), 8.22 (m, 3H, H-5, H-13), 8.15 (s, 2H, H-19), 7.89 (dd, 1H, ³J = 8.9, ⁴J = 2.3, H-7), 7.80 (d, 2H, ³J = 8.1, H-12), 7.55 (d, 1H, ³J = 8.9, H-8), 7.11 (d, 2H, ³J = 3.9, H-18), 6.71 (d, 2H, ³J = 2.7, H-17). ¹³C NMR (188.6 MHz, δ):155.9 (C-2), 152.0 (C-8a), 146.8 (C-10), 146.3 (C-11), 145.3 (C-19), 135.8 (C-7), 134.5 (C-16), 134.0 (C-4), 133.3 (C-15), 132.9 (C-14), 132.2 (C-18), 132.0 (C-13), 131.9 (C-5), 126.2 (C-12), 124.5 (C-3), 123.5 (C-9), 120.7 (C-6), 119.9 (C-17), 119.1 (C-8), 117.4 (C-4a). ¹¹B NMR (96.2 MHz, δ): 0.13 (t, J_B-F = 29). ¹⁹F NMR (282.2 MHz, δ): 141.39 (q, J_B-F = 29). FT-IR (cm⁻¹): 3031, 1735, 1537, 1391, 1385, 1256, 1110, 1070, 775, 745, 717. HR-MS (ESI-TOF) m/z calcd. for C₂₆H₁₅BBrF₂N₅O₂ [M+Na]⁺: 582.0350, found 582.0351.

meso-(4-(1-((6-chloro)2-oxo-2H-chromen-3-yl)-1H-1,2,3-triazol-4-yl))phenyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (2e).

Orange crystalline solid (63 %). Mp: >300 °C. ¹H NMR (750 MHz, δ): 9.29 (s, 1H, H-9), 8.79 (s, 1H, H-4), 8.22 (d, 2H, ³J = 8.2, H-13), 8.16 (s, 2H, H-19), 8.10 (d, 1H, ⁴J = 2.4, H-5), 7.80 (d, 2H, ³J = 8.2, H-12), 7.79 (dd, 1H, ³J = 8.8, ⁴J = 2.4, H-7), 7.62 (d, 1H, ³J = 8.8, H-8), 7.12 (d, 2H, ³J = 3.9, H-18), 6.72 (d, 2H, ³J = 2.6, H-17). ¹³C NMR (188.6 MHz, δ): 155.9 (C-2), 151.6 (C-8a), 146.9 (C-10), 146.3 (C-11), 145.3 (C-19), 134.6 (C-16), 134.1 (C-4), 133.3 (C-15), 132.9 (C-7), 132.3 (C-14), 132.2 (C-18), 132.0 (C-13), 129.6 (C-6), 128.9 (C-5), 126.2 (C-12), 124.6 (C-3), 123.5 (C-9), 120.2 (C-4a), 119.9 (C-17), 118.9 (C-8). ¹¹B NMR (96.2 MHz, δ): 0.15 (t, J_B-F = 29). ¹⁹F NMR (282.2 MHz, δ): 143.79 (q, J_B-F = 29). FT-IR (cm⁻¹): 3061, 2166, 1714, 1536, 1386, 1261, 1114, 1078, 980, 813, 744. HR-MS (ESI-TOF) m/z calcd. for C₂₆H₁₅BClF₂N₅O₂ [M+Na]⁺: 536.0873, found 536.0896.

meso-(4-(1-((8-ethoxy)2-oxo-2H-chromen-3-yl)-1H-1,2,3-triazol-4-yl))phenyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (2f).

Red crystalline solid (67 %). Mp: 247–249 °C. ¹H NMR (750 MHz, δ): 9.26 (s, 1H, H-9), 8.80 (s, 1H, H-4), 8.22 (d, 2H, ³J = 8.1, H-13), 8.15 (s, 2H, H-19), 7.80 (d, 2H, ³J = 8.1, H-12), 7.50 (d, 1H, ³J = 7.7, H-7), 7.42 (d, 1H, ³J = 7.7, H-5), 7.40 (t, 1H, ³J = 7.7, H-6), 7.12 (d, 2H, ³J = 3.9, H-18), 6.71 (d, 2H, ³J = 2.0, H-17), 4.23 (q, 2H, ³J = 6.9, CH₂), 1.43 (t, 3H, ³J = 6.9, CH₃). ¹³C NMR (188.6 MHz, δ):156.1 (C-2), 146.9 (C-11), 146.2(C-8, C-10), 145.3 (C-19), 142.4 (C-8a), 136.1 (C-4), 134.5 (C-16), 133.2 (C-15), 133.1 (C-14), 132.1 (C-18), 132.0 (C-13), 126.1 (C-12), 125.9 (C-6), 123.7 (C-3), 123.6 (C-9), 121.0 (C-7), 119.9 (C-17), 119.3 (C-4a), 116.6 (C-5), 65.1 (CH₂), 15.1 (CH₃). ¹¹B NMR (96.2 MHz, δ): 0.13 (t, J_B-F = 29). ¹⁹F NMR (282.2 MHz, δ): 143.79 (q, J_B-F = 29). FT-IR (cm⁻¹): 2983, 1717, 1541, 1385, 1260, 1150, 1075, 977, 775. HR-MS (ESI-TOF) m/z calcd. for C₂₈H₂₀BF₂N₅O₃ [M+H]⁺: 524.1700, found 524.1706.

meso-(4-(1-((7-hydroxy)2-oxo-2H-chromen-3-yl)-1H-1,2,3-triazol-4-yl))phenyl-4,4-difluoro-4-bora-3a,4a-diaza-s-indacene (2g).

Red crystalline solid (52 %). Mp: 239–241 °C. ¹H NMR (300 MHz, δ): 9.21 (s, 1H, H-9), 8.70 (s, 1H, H-4), 8.20 (d, 2H,³J = 8.3, H-12), 8.16 (s, 2H, H-19), 7.81 (d, 2H, ³J = 8.3, H-13), 7.78 (d, 1H, ³J = 8.6, H-5), 7.13 (m, 2H, H-18), 6.93 (dd, 1H, ³J = 8.6, ⁴J = 2.3, H-6), 6.89 (d, 1H, ⁴J = 2.3, H-8), 6.72 (m, 2H, H-17). ¹³C NMR (75.4 MHz, δ): 163.2 (C-7), 156.8 (C-2), 155.3 (C-8a), 146.9 (C-10), 145.9 (C-11), 145.2 (C-19), 137.4 (C-16), 134.5 (C-4), 133.2 (C-15), 133.1 (C-14), 132.2 (C-18), 131.9 (C-13), 131.5 (C-5), 125.9 (C-12), 123.7 (C-9), 119.8 (C-17), 119.5 (C-3), 114.9 (C-6), 110.7 (C-4a), 102.2 (C-8). ¹¹B NMR (96.2 MHz, δ): 0.28 (t, J_B-F = 29). ¹⁹F NMR (282.2 MHz, δ): 145.48 (q, J_B-F = 28). FT-IR (cm⁻¹): 3286, 2964, 2145, 1723, 1599, 1354, 1258, 1119, 1077, 774. HR-MS (ESI-TOF) m/z calcd. for C₂₆H₁₆BF₂N₅O₃ [M+H]⁺: 496.1387, found 496.1384.

2.2. Single crystal X-ray molecular structure

Single crystals of compound 2a, suitable for X-ray diffraction, were obtained from a saturated solution of 1,2-dichlorobenzene (DCB) as the monosolvate (2a∙DCB). Several efforts to obtain single crystals of 2b-g were unsuccessful. The monocrystal data of 2a∙DCB were recorded on a Bruker D8 VENTURE diffractometer using a graphite monochromator (Cu K_α, λ = 1.54184) at 185(2) K. The SAINT [54] and SORTAV software [55] were used to carry out the cell refinement and data reduction, respectively. The structures were solved by direct methods using the SHELXS-97 program [56] of the WINGX package [57]. The full-matrix least-squares methods were used to perform the final refinement with the SHELX97 program [56]. The H atoms on C, N, and O were geometrically positioned and treated as riding atoms with: C–H = 0.93−0.98 Å, Uiso(H) = 1.2 eq(C) for aromatic carbon atoms. Figures for publication were prepared with Platon [58]and Mercury [59]. General crystallographic data for compound 2a, has been deposited in the Cambridge Crystallographic Data Centre (CCDC) as supplementary publication number 1994818.

2.3. Hirshfeld surface (HS) analysis and energy-framework diagrams

These calculations were performed with the software CrystalExplorer17.5 [60]. The HS was mapped over d_norm [61]. The energy components E_elect (electrostatic potential), E_disp (dispersion), E_pol (polarization), and E_total (total energy) were calculated by CrystalExplorer17.5-Gaussian 09 software [62] tandem, using the B3LYP/6-31G (d,p) level of theory. Scale factors [60] and % contribution of individual components to the stabilization energy [63] were also calculated. The interaction energy calculations were performed with the CLP-PIXEL software [64].

2.4. UV–vis and emission spectra

The UV–vis and emission spectra were recorded with a PerkinElmer Lambda XLS spectrophotometer and PerkinElmer LS55 fluorescence spectrometer, respectively, using 1 cm path length cuvettes of quartz at room temperature. Thin films of BPhTCs 2a-2g were supported on a quartz surface by self assembly [65] through immersion in THF solutions (4 × 10⁻⁴ M, 517 nm) during 40–50 min, for absorption and emission measurements. Fluorescence quantum yields (φ) were measured in THF solutions, using fluorescein in 0.1 M NaOH (φ = 0.92) as a standard, as follows:

$${\phi }_{sample}={\phi }_{std}*\frac{A_{std}*F_{\left(area\right)sample}}{A_{sample}*F_{\left(area\right)std}}*{\left(\frac{{\eta }_{solventsample}}{{\eta }_{solventstd}}\right)}^2$$

Where A is absorbance, F is the area under emission spectrum, η is the refractive index and std the standard.

2.5. Thermal analysis

Q2000 equipment and a Thermobalance Q5000 IR, of TA instruments, were used to perform DSC and TG measurements, respectively. A quantity of approximately 3.0–5.0 mg of sample, a gradient of 5.0 °C/min, a temperature interval from room temperature to 350 °C, under air flux of 25 mL/min in an open (TG), or pin-holed panels (DSC) were used.

2.6. Powder X-ray diffraction

Room temperature X-ray powder diffraction data were collected on a PAN Analytical X'Pert PRO diffractometer with Cu Kα1 radiation (λ = 1.5405 Å, 45 kV, 40 mA) or on a D8 Focus Bruker AXS instrument using Cu Kα1 radiation (λ = 1.542 Å, 35 kV, 25 mA). The estimation of the crystallite size was carried out by the Scherrer equation [66]. The amorphousness and crystallinity percentages were calculated from the global and reduced area.

2.7. Computational methods

Gaussian 09w [62] package was used to perform density functional theory (DFT) calculations with the BVP86 method, the 6-31G (d,p) basis set theory level, and the DEF2TZV fitting set. The frontier molecular orbital (FMO) energy levels were calculated in the optimized geometry. Linear transitions of absorption were estimated from the optimized molecular structures using time-dependent DFT (TDDFT) methods [67] at the ωB97X-D/DGDZVP level of theory with CPCM model [68] in THF. The first ten electronic transitions were obtained from the optimized structures.

2.8. Electrochemical characterization

Cyclic voltammetry (CV) was performed with a PARSTAT® 2273 potentiostat-galvanostat. An electrochemical 3-electrodes cell was used: a saturated calomel electrode (SCE), a platinum wire and a graphite coated with 2a-g films as reference electrode, auxiliary electrode and working electrode, respectively. A solution of 0.1 M Bu₄NPF₆ in anhydrous acetonitrile was used as the electrolyte. The experiments were carried out at room temperature, with a prior injection of nitrogen to deoxidize the system, and at a scan rate of 50 mV/s. The films were deposited on the working electrode by self-assembly [65] on a graphite surface of 0.25 cm² using a saturated solution of the corresponding compound in THF as solvent. Ferrocene-ferrocenium (Fc/Fc⁺) couple was the internal standard. The energy levels of the FMOs (EH^EC and EL^EC) and electrochemical band gap (Eg^EC) values were calculated from the oxidation and reduction potentials in the corresponding onsets (E_onset(Ox) and E_onset(Red)), using the empirical equations: EH^EC = -e(E_onset(Ox) + 4.4), EL^EC = -e(E_onset(Red) + 4.4) and Eg^EC = -(EH^EC – EL^EC), as reported [69,70].

2.9. Film resistivity measurements

BPhTCs 2a-g were deposited on glass slides from THF solutions as thin films. Thickness measurements were performed with AMBIOS XP-100 Thickness Profilometer (A = 1 cm², thickness = 580–620 nm). Signatone Pro-4 of Lucas Labs, configured with inline four point probe [71], was used for sheet resistivity measurements with squared probes, the contact points were distanced such as: S₁₂ S₂₃ = S₃₄ = S = 1 mm.

2.10. Manufacture and electric characterization of OFET devices

OFET devices were manufactured following a top-contact/bottom-gate architecture with Indium Tin Oxide (ITO) films of 1.0 × 0.5 inch and 100 nm film thickness (surface resistivity 60 Ω/sq) as flexible substrate. The ITO films were previously sonicated in a mixture of tridistilled water and isopropyl alcohol (1:1) during 10 min and then ozonized under UV light during 15 min. Using a 12 μm pulsed CO₂ laser, one ITO substrate underwent patterning of source-drain contacts resulting in a channel of 1.0 cm wide (W) and 60 μm length (L). Next, solutions of OSs 2a-g in THF (2 × 10⁻⁴ M), were deposited by the Drop-Casting method [72] onto the channel and contacts, followed by the placement of a polydimethylsiloxane (PDMS) layer using the Doctor Blade technique [73]. The device was finally assembled with placing the ITO gate substrate. The electric characterization of compounds 2a-g, was performed in a system of two Keithley Source Meter (SMU) models 2400c y 2450 equipment, including the following: on/off ratio (I_ON/I_OFF) and field effect mobility at saturation regime (μ_FET). The reported values are the mean of at least three devices per each OS. All measurements were conducted under ambient temperature and atmosphere. The electric circuit and the architecture of OFETs are depicted in Fig. S1.

From the output characteristics, the lineal and saturation regimes are distinguished. In this work, the mobility of charges was calculated under the saturation regime when V_D > (V_G – V_T), in this case, the following equation is used to calculate the μ_FET from the transference plot:

$${\mu }_{FET}=\left(\frac{2L}{WC}\right){\left({\left(\frac{\delta \sqrt{I_D}}{\delta V_G}\right)}_{V_D}\right)}^2$$

Where L and W are the length and width of the channel, respectively, and C is the capacitance of the insulator per unit area (5.3 μFcm⁻²).

2.11. Characterization of BPhTCs films by AFM

AFM analysis was performed on films generated by spin coating, using 1 mg/mL solutions of each compound in chloroform. They were deposited at 3500 rpm for 1 min. The thicknesses were measured by AFM, ranging from 14 to 114 nm. Detailed values are listed in Table S7. The atomic force microscope (3D Bioscope Catalyst, Bruker, USA) coupled to an inverted optical microscope (Axio Observer Z1, Zeiss, Germany) was used to run the experiments. The samples were directly deposited on the microscope plate and scanned in ScanAsyst mode using RTESP-300 probes (k = 40 N/m, tip radius = 8 nm, f₀ = 300 kHz). The analyses were run under air at laboratory conditions (T = 20 °C). The average arithmetic roughness (R_a) was estimated in the height images with scan areas of 10 × 10 μm², using the NanoScope v. 1.40 software (Bruker).

3. Results and discussion

3.1. Synthesis and characterization

A conventional two-steps one pot protocol for BODIPYs synthesis was followed. Benzaldehydes 1a-g were reacted with pyrrole in the presence of TFA, as catalyst, to afford the corresponding dipyrromethanes, which were further oxidized in situ with DDQ, followed by a complexation step using BF₃–OEt₂ in the presence of Et₃N, leading to the formation of the corresponding BPhTCs 2a-g, Scheme 1. After column chromatography, compounds 2a-g were isolated in moderate yields, 52–67 %, which are in the reported range for other BODIPYs or even larger [74]. Evidence of the boron-fluorine coupling was obtained from all spectra in which a triple signal at δ 0.13 to 0.30 for ¹¹B-NMR (J_B-F = 28–29 Hz) and a quadruple signal (1:1:1:1) at −143.0 to −145.0 ppm for ¹⁹F (J_F-B = 29 Hz) were observed, in agreement with other meso-aryl-BODIPYs [75].

Scheme 1.

Two-steps, one pot synthesis of BPhTCs. (i) Pyrrole, TFA, DDQ (1.2 eq.), Et₃N (10 eq.), BF₃-(OEt)₂ (15 eq.) in CH₂Cl₂. Yields of 2a-g are in parenthesis.

3.2. Molecular and supramolecular structure of BPhTC 2a-DCB

The molecular structure of 2a was confirmed by single crystal X-ray. The ORTEP plot is depicted in Fig. 2(a); compound 2a crystallized in the triclinic system, space group P-1 and Z = 2 with one molecule of 1,2-dichlorobenzene as crystallization solvent (2a-DCB). The crystal data collection and refinement parameters of 2a-DCB are listed in Table S2 and selected geometric parameters in Table S3. Bond lengths and angles in the coumarin [76], triazole [45] and BODIPY [11,12] fragments are in the characteristic range for analogous molecules. Particularly the C19–C22 length (1.477(4) Å) is in the reported range for other meso-substituted BODIPYs [31]. The coumarin heterocycle is forming angles of 10.20(9)° with the triazole ring, 1.97(7)° with the phenyl ring and 55.50(5)° with the BODIPY ending, in agreement with other reported meso-aryl substituted BODIPYs (53–56°) [77]. Therefore, the molecule is almost planar in the 4-pheny-1,2,3-triazolyl-coumarin (PhTC) fragment, but the BODIPY ending is in synclinal disposition.

Fig. 2.

(a) ORTEP plot of BPhTC 2a at 50 % ellipsoid probability level with one molecule of 1,2-dichlorobenzene, used as crystallization solvent. (b) View in the ab plane, showing the supramolecular architecture developed by C–H⋅⋅⋅ A (A = N, O) interactions. (c) π-Stacking through carbonyl … carbonyl dipolar n→π* and parallel displaced π … π* interaction in the (1 1 1) family of planes, the BODIPY ending is drawn in solid lines.

The crystal lattice is structured through the participation of C5–H5⋯N13 and C8–H8⋯O1 weak hydrogen bonding interactions to develop 1D tapes along the [−1 1 5] direction, Fig. 2(b). Both interactions describe ring motifs whose graph set descriptors are R²₂(16) and R²₂(8), respectively. Coumarin carbonyls self-assemble the TC moieties through C2O2⋯C2O2 interactions of antiparallel type [78], directing the stacking in 2D along the (1 1 1) family of planes. The stacking is complemented by parallel displaced π-π* stacking interactions, one of them between the triazole and the benzofused coumarin ring, Cg1⋯Cg4, as has been observed in coumarins [76,79] and between two pyrrole rings, Cg3⋯Cg3. The third dimension is developed through several C–Y⋯A (Y H, O; A = π, F) close contacts, Fig. 2(c). Among non-covalent interactions, the π−π stacking is considered as the most efficient for transporting charge carriers [80]. The geometric parameters related to non-covalent interactions are listed in Tables S4 and S5.

3.3. Hirshfeld surface and quantitative interaction energy analysis of 2a∙DBC

The Hirshfeld surface (HS) of 2a·DCB, mapped over the d_norm (range = −0.2612 a 1.3011 atomic units), is shown in Fig. 3(a). The red bright spots highlight the dominant interactions, which are given by X⋯Y (X = H, C; Y N, O, F, Cl) interactions. Among all the close interactions, Y⋯H/H⋯Y (Y N, O, F, Cl) are the largest contributors to the surface (34.8 %), followed by H⋯H contacts (25.3 %), C⋯H/H⋯C (23.7 %), and at last the sum of C⋯Y/Y⋯C (14.8 %, Y C, N, O, F, Cl). The percent contributions to the HS are shown in Fig. 3(b).

Fig. 3.

(a) View of the Hirshfeld surface of BPhTC 2a, mapped over d_norm −0.2612 to 1.3011 range in a.u. Hydrogen bonding is represented by dashed lines. (b) Comparative percent contributions in the Hirshfeld surface, (Y N, O, F, Cl). (c) Energy-framework diagrams for E_ele (red), E_dis (green) and E_tot (blue) for a cluster of nearest-neighbor molecules in 2a, cylinder scale of 80 for energy and cut-off = 10 kJ mol⁻¹ in 2³ unit cells. Note the large contribution of dispersive energy, in green, to the stabilization of the crystal network. DCB was omitted for clarity. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Selected results of interaction energies are listed in Table 1. Those contacts with total interaction energy (E_tot) values larger than −20.0 kJ mol⁻¹ were omitted from discussion. The first directing energy contributor to the framework of 2a is the assembly of C21–H21···Cg2, C2–O2···Cg5 and Cg1···Cg4 dispersive interactions (%E_dis = 85.6 %) which sum an E_tot value of −64.5 kJ mol⁻¹. The second and third energy contributors are C5–H5⋯N13 (E_tot = −42.5 kJ mol⁻¹; %E_dis=43.8) and C2–O2⋯C2 (E_tot = 40.5 kJ mol⁻¹; %E_dis=74.4), respectively. Finally, the energies of C8–H8⋯O1, C30–H30⋯F34 and Cg3···Cg3 interactions contribute with E_tot in the −25.5 to −27.4 kJ mol⁻¹ each. Then, the lattice of 2a∙DCB is mainly stabilized by dispersive energy coming from the already non-covalent interactions early described. The larger contribution of E_dis is best appreciated in the Energy-framework diagrams for E_ele (red), E_dis (green) and E_tot (blue) for 2a, shown in Fig. 3(c).

Table 1.

Calculated interaction energies (kJ mol⁻¹), %E_comp contributions to stabilization energy for selected HB and close contacts in BTC 2a.

Interaction	-Eele	-Epol	-Edis	Erep	-Etot	%Eelea	%Edisa	Rb
C21–H21···Cg2iv,c	9.5	5.5	89.5	44.8	64.5	9.1	85.6	7.11
C2–O2···Cg5iv,c
Cg1···Cg4viii,c
C5–H5⋯N13i	35.8	9.3	35.1	52.9	42.5	44.6	43.8	11.9
C2–O2⋯C2vi	12.6	3.5	46.7	25.9	40.5	20.1	74.4	11.0
C8–H8⋯O1ii	8.9	2.7	30.4	20.0	25.5	21.2	72.4	16.5
C30–H30⋯F34iii	23.9	4.2	18.7	29.4	26.5	51.1	40.0	16.9
Cg3···Cg3vii,c	7.6	3.3	30.8	16.1	27.4	18.2	73.9	13.6
C25–H25···Cg4v,c	5.3	0.9	32.3	18.2	23.1	13.8	83.9	7.67

^aThe % contribution of E_comp (%E_comp) is calculated through (E_comp/E_stab) × 100 where E_stab = E_ele + E_pol + E_disp.

^bR = distance between centroids of the interacting molecules in Å.

^cCg is the centroid of the n ring formed by the atoms indicated in parenthesis: Cg1 (N11–C15), Cg2 (N27–C26), Cg3 (N29–C33), Cg4 (C5–C10) and Cg5 (C16–C21).

3.4. X-ray powder diffraction, thermal stability, and AFM morphological properties

The thermal stability was settled by thermogravimetry (TG) and differential scanning calorimetry (DSC). In both cases, the experiments were run under air. The thermograms of 2a-g are displayed in Fig. 4. Compounds 2b, 2c and 2g lose crystallization solvent before 150 °C, probably due to small amounts of surreptitious water and/or elution solvent remnants. However, BPhTCs 2a, 2c-g are thermally stable up to 250 °C, except for compound 2b (6-OMe) which decomposed at 180 °C. Decomposition is a multistep process, with a clear melting point for 2b and 2g, and the lack of solid phase transitions as shown by DSC analyses. The complete TG, ΔTG and DSC graphs are depicted in Figs. S2–8.

Fig. 4.

Thermograms of BPhTCs 2a-g recorded under air.

The microcrystalline-amorphousness nature of BPhTCs was estimated by X-ray, the powder diffraction patterns of 2a-g are depicted in Fig. 5. Broad lines characteristics of semicrystalline solids are observed in all diffractograms, allowing the estimation of 2a-g crystallinity in the 35–50 % range. The lines broadness is mostly due to the presence of small particles of BPhTCs whose crystallite sizes are in the 22–35 nm range, except for 2b of 61 nm. The estimated crystallinity and particle size values are included in Table S6. The XRPD pattern of 2a is quite different than that obtained by monocrystal X-ray diffraction of 2a∙DBC (1,2-dichlorobenzene) solvate, Fig. S9, in agreement the absence of DBC as crystallization solvent. Nevertheless, the molecular structures of 2a-g are expected to be similar between them since all compounds possess the same basic structural unit.

Fig. 5.

X-ray powder diffractograms of 2a-g.

Atomic force microscopy (AFM) measurements were performed to confirm the uniformity and morphology of the films formed by compounds 2a-g. The average arithmetic roughness (R_a) was estimated in the height images with squared scan areas of 10² μm². All compounds formed continuous films on glass surface after spin-coating, the films are characterized for roughness (R_a) values in the 2.8–18 nm range, being that formed by compound 2a the smoothest (R_a = 2.8 ± 0.5 nm), a complete list of R_a values is in Table S7 and AFM images in Fig. 6. However, some films show plateau-type (2c, 2e) or granulated (2f-g) aggregation domains within the continuous films, which could exert a fundamental role in determining the mobility of charges in thin film transistors [81].

Fig. 6.

AFM images of spin coated 2a-g films (10 × 10 μm).

3.5. Photophysical properties

The optical properties of 2a-g were investigated by absorption-emission spectroscopy. Measurements were performed both in solution and in solid thin films. The corresponding data are summarized in Table 2, Table 3. The UV–vis absorption profiles of BPhTCs in THF solutions are shown in Fig. 7(a). Compounds 2a, 2b, 2d, and 2f exhibit similar absorption profiles in solution, they show one absorption maximum (λ₁) centered at around 365–377 nm, attributed to the π–π* excitation from the PhTC moieties, and bathochromic shifted by 15–50 nm in relation to the parent PhTCs [51]. In contrast, the absorption profiles (λ₁) at high energy of BPhTCs 2c, 2e and 2g show two absorption maxima approximately at ∼361 and 418 nm each. In the region of longer wavelengths, BPhTCs show the absorption maxima (λ₂) at 500–502 nm, associated to the π–π* transition from the BODIPY unit [3,21].

Table 2.

Optical data of 2a-g in THF solutions at room temperature.

BPhTC	λabsa/nm		ε2/104 M−1 cm−1 (% error)	Egopt/eV	λemb /nm	Stokes shift/cm−1	Fluorescencec
	λ1	λ2
2a	365	500	4.78 (2.3)	2.35	520	769	5826
2b	377	501	4.87 (2.9)	2.36	521	766	7686
2c	363, 417	501	3.24 (0.9)	2.34	522	803	3269
2d	366	501	10.6 (0.8)	2.36	520	729	5514
2e	364, 417d	501	4.34 (1.0)	2.36	521	766	6392
2f	377	501	4.24 (0.6)	2.36	521	766	4581
2g	359, 419	502	1.52 (2.0)	2.36	520	690	4606

^aAbsorption in THF 2.5 X 10⁻⁵ M.

^bEmission maximum in THF at 10⁻⁶ M, excited at the absorption maximum (λ₂).

^cMeasured from experimental emission spectra as the area under the curve in area units.

^dShoulder.

Table 3.

Optical data of 2a-g in solid films at room temperature.

BPhTC	λabsa/nm		Egopt/eV	λemb/nm	Stokes shift/cm−1	Fluorescencec
	λ1	λ2
2a	367	526	2.18	607	2537	1456
2b	381	522	2.22	606	2655	3538
2c	365	523	2.19	592	2229	1111
2d	379	522	2.20	604	2601	3702
2e	364	521	2.19	575	1806	1742
2f	375	527	2.19	562	1182	870
2g	372	518	2.22	584	2182	1954

^aAbsorption of thin solid films.

^bEmission of thin solid film exited at the absorption maximum (λ₂).

^cMeasured from experimental emission spectra as area under the curve in area units.

Fig. 7.

Absorption and normalized emission spectra of BPhTCs 2a-g. (a) Absorption and (b) emission in THF solutions at 10⁻⁶ M. (c) Absorption and (d) emission in thin films formed after three submersions of 15 min in THF solutions at 2 × 10⁻⁴ M.

The ε₂ values are within the same order of magnitude (∼10⁴ M⁻¹ cm⁻¹), reaching the largest value in compound 2d (10.6 × 10⁴ M⁻¹ cm⁻¹), effect attributed to the contribution of additional electronic density from the bromine lone pairs of electrons. As far as fluorescence quantum yield is concerned, the values are in the 0.004–0.0117 range, very similar to the value reported for meso-aryl-BODIPY (φ = 0.065) [80]. The normalized emission spectra of 2a-g in THF, compared to fluorescein as a standard, as well as φ values are shown in Fig. S10. It is worth noting that the absorptions of the BODIPY unit in 2a-g are very similar to the reported value for the absorption of meso-aryl-BODIPY (λ_abs = 497 nm) [82]. These results are explained because of the absence of conjugation between the PhTC and the BODIPY unit, and the small effect exerted by weak electron acceptor or donor groups in the BODIPY meso-aryl substituent [82].

Furthermore, in thin films, the absorption profiles of 2a-g are quite different from solution, Fig. 7(c). They appear as two broad bands, the high energy band, associated to the coumarin transitions, exhibit hyperchromism whereas the BODIPY absorption hypochromism, compared to the solution spectra. This result indicates that the close packing of the molecular backbone within the thin films of BPhTCs improves the charge transfer into the coumarin. In general, the optical band-gap energies of 2a-g (E_g^opt), obtained from the absorption spectra in thin films (2.20 ± 0.02 eV), are smaller by 6–8% than the values in solution (2.35 ± 0.01 eV) and independent from the substitution pattern and nature of the substituent in the coumarin unit.

The emission profiles of BPhTCs in THF solutions are shown in Fig. 7(b), and in their solid films thereof in Fig. 7(d). The data indicate that all compounds exhibit fluorescence in solution at λ_em in the 520–522 nm range, Table 2. This value is independent from the substitution pattern in the coumarin unit and very similar to the reported value for meso-aryl-BODIPY (λ_em = 518 nm) [82]. The coumarin emission band was totally suppressed and only the emission from the BODIPY ending was observed. Furthermore, the emission maxima are red shifted in solid films as compared to THF-solutions by 84–87 nm (2a-b and 2d), 64–70 nm (2c and 2g) and 41–54 nm (2e-f), to reach the longest wavelength at 607 nm in compound 2a and the shortest at 562 nm in compound 2f, Table 3. The resulting fluorescence spectra are wide and almost symmetrical, the collection of which covers a larger part of the visible electromagnetic spectrum from green to red. In addition, the Stokes shift values in the solid, fold a minimum value of 1.8 (2f) to a maximum of 3.6 (2d) times the values in solution, indicating that decaying to the lowest vibrational level S₁ is highly favored in the solid, thus part of the energy is non-radiatively dissipated.

The emissive behavior of BPhTCs, in the solid, agrees with Twisted Intramolecular Charge Transfer (TICT) process characteristic of meso-aryl-BODIPYs [83]. The locally excited (LE) state of BPhTCs shapes a planar conformation in solution stabilized by electronic conjugation, which gives a sharp emission spectrum. In the solid, the intramolecular motion is restricted bringing the BPhTCs from the LE state to the TICT state, at which there is a total charge separation between BODIPY and the PhTC fragments. The TICT state has the effect of narrowing the band gap, red-shifting the emission spectrum but weakening the emission intensity through nonradiative quenching processes. The increased order in the solid, compared to solution, is given by the already described n→π* and pd π⋅⋅⋅ π* stacking interactions networking the crystal structure (vide supra).

3.6. Computational analysis

The calculated molecular structures, in the gas phase, of compounds 2a-g predict an almost planar arrangement relying in the PhTC fragment. The calculated torsion angles between coumarin and triazole are in the 9.7–11.5° range, and between coumarin and phenyl are in the 0.1–1.5° range, but the BODIPY endings are 47.5–50.2° out of the mean coumarin plane, in good agreement with the torsion angles measured in the single crystal X-ray structure of 2a-DCB (vide supra). Thus, the theoretical calculations are in good agreement with the geometry of the molecular structure observed by X-ray diffraction.

On the other hand, the challenge of acceptably predicting the electronic transitions of BODIPY derivatives has been reported in the literature [84,85]. The ωB97X-D model was chosen after comparing several commonly used models both in vacuum and with the CPCM model to estimate the electronic properties considering the contribution of the solvent [68,86]. The results of model comparisons can be found in the supplementary material S4. With all models, three main bands are predicted with a similar character of the transitions between the models, highlighting two more intense transitions. Acceptable approximations for BODIPYs derivatives have been reported with the ωB97X-D [84,[87], [88], [89]]. The most red-shifted electronic transition (around 430 nm) coincides with HOMO to LUMO (H→L) transition in six of the seven molecules studied, as expected. However, in the case of molecule 2c, this band corresponds to a transition from HOMO-1 to LUMO (H-1→L). This behavior has been described both theoretically and experimentally, where the lowest energy excited state level does not correspond to a HOMO/LUMO transition [[90], [91], [92]], including coumarin derivatives [93], such as 2c. In this case, the diethylamino group, as an electron-donating group, modifies the energy of the HOMO [94,95] approaching the energy levels between HOMO and HOMO-1 (−4.94 eV and −4.91 eV, respectively. See Table S8) and enabling the HOMO-1 to LUMO transition [96].

The lowest energy transitions, their energies, oscillator strengths, and probable MO nature are summarized in Table 4. A complete list of the main transitions are tabulated in Table S8 and a pictorial representation of the lowest excited electronic transitions of BPhTCs are visualized in Fig. S11. The oscillator strength is a physical parameter that shows the allowedness in the radiative electronic transition of molecules and is directly proportional to the absorptivity coefficient [33]. Given the non-planar disposition between the BODIPY ending and the PhTC fragments, the f values of the H→L transition, around 0.53, are explained because of the weak mixing of the electronic wave functions.

Table 4.

Computed lowest electronic transitions of BPhTCs 2a-2g obtained by TDDFT ωB797X-D/DGDZVP, CPCM model in THF. Wavelength of absorption (λ_abs) in nm, energy (E) in eV, oscillation strength (OS, f), and MO character (% probability).

Compound	λab / nm	E/eV	OS (ƒ)	MO Character (%)
2a	431.11	2.8759	0.5524	H→L (96 %)
2b	431.61	2.8726	0.5478	H→L (96 %)
2c	431.49	2.8734	0.5447	H-1→L (96 %)
2d	431.73	2.8718	0.5492	H→L (96 %)
2e	431.75	2.8717	0.5490	H→L (96 %)
2f	431.69	2.8721	0.5471	H→L (96 %)
2g	431.53	2.8731	0.5479	H→L (96 %)

The theoretical absorption spectra of the BODIPYs in THF solutions are shown in Fig. 8. The influence of the groups is reflected in the theoretical transitions, two main absorption regions, the first between 300 and 400 nm and the second between 400 and 500 nm are predicted. Molecules 2a, 2b (OMe), 2f (OEt) and 2g (OH) have similar theoretical spectra in terms of transitions and intensity, due to the influence of similar substituent groups. In addition, molecules 2d (Cl) and 2e (Br) differ only in the halide and show an almost identical spectrum. Finally, the spectrum of 2c exhibited the highest intensity and the largest bathochromic shift of the high energy band. The most red-shifted transition of BPhTCs is theoretically predicted at a wavelength around 431 nm in all compounds, regardless the coumarin substituent, 93-84 nm at smaller wavelength than the experimental spectra, as has been reported for calculated spectra of aryl meso-aryl-BODIPYs [97].

Fig. 8.

Theoretical absorption spectra of 2a-2g, obtained by TDDFT ωB797X-D/DGDZVP CPCM model in THF.

The pictorial representations of the HOMO, LUMO, HOMO-1 and LUMO+1 FMOs of compounds 2a, 2c, and 2d, as representative examples, are shown in Fig. 9. In general, the calculated absorption spectra of BPhTCs are consistent with what is observed in the experimental spectra. Then, the most red-shifted band, attributed to π→π* transition at 431 nm, defines the bandgap value of this family of compounds. The HOMO is localized on the BODIPY ending in all compounds, in agreement with reported calculations in other meso-aryl-BODIPYs [97,98]. On the other hand, the LUMO is centered in the whole molecule (2a,b and 2f,g), on the TC fragment (2d,e), or on the BODIPY ending (2c), depending on the coumarin substituent. As mentioned before, the H-1→L transition is predicted to define the main transition in 2c, H-1 is localized in the PhTC moiety. The effect of placing a BODIPY at one end of the PhTCs precursors (1a-g) [43] upshifts the energy of HOMO and downshifts the energy of the LUMO, to give bandgap values (E_H→L^DFT) around 1.9 eV.

Fig. 9.

Pictorial representation of HOMO, LUMO, HOMO-1, LUMO+1, energy distributions for 2a, 2c, and 2d, involved in the absorption processes.

According to theoretical calculations, only the 7-NEt₂ group in the coumarin is capable to transform the TC moiety into the donor, making isoenergetic the HOMO-1 and HOMO FMOs, thus reverting the D-A nature of the BODIPY and coumarin fragments in 2c.

In terms of resonance structures and donor-acceptor (D-A) properties of BPhTCs, that give rise to the experimental optical bandgaps (E_H→L^opt), it can be accounted that the PhT fragment behaves as the donor and both the coumarin and BODIPY fluorophores behave as the acceptors, depending upon the substituent in the coumarin moiety. The proposed resonance structures A and B are depicted in Fig. 10. In BPhTCs bearing EW or ED groups on the 6- or 8- positions, the coumarin fragment exerts the role of the acceptor to develop the resonance form A. This behavior is reversed by 7-NEt₂ and 7-OH ED groups, forming the resonance structure B.

Fig. 10.

Proposed resonance structures A and B for BPhTCs 2a-d. 2a (H), 2b (6-OMe), 2c (7-NEt₂), 2d (6-Br), 2e (6-Cl), 2f (8-OEt) and 2g (7-OH), the Coumarin and BODIPY endings act as the acceptors and the Ph-Triazole as the donor.

Besides, it is worth mentioning that the computed dipole moment (μ) is the largest for 2c (15.8 D) and the shortest are for the halogenated compounds 2d (8.2 D) and 2e (8.1 D) whereas the rest of compounds show intermediate values. These results agree with the above proposal. In general, the charge separation is more effective upon the incorporation of the BODIPY ending to the PhTC fragment. A complete list of μ values can be found in Table 5.

Table 5.

EC properties from CV data and theoretical dipole moment of 2a-2g.

BTC	Eonset(Red) (V)	Eonset(Ox) (V)	ELEC (eV)	EHEC (eV)	EH→LEC (eV)	Dipole moment (D)
2a	−0.68	1.40	−3.72	−5.80	2.13	10.7
2b	−0.58	1.48	−3.82	−5.88	2.06	12.2
2c	−0.66	1.59	−3.74	−5.99	2.25	15.8
2d	−0.46	1.62	−3.94	−6.02	2.08	8.2
2e	−0.53	1.49	−3.87	−5.89	2.02	8.1
2f	−0.42	1.67	−3.98	−6.07	2.09	12.2
2g	−0.74	1.30	−3.66	−5.70	2.10	11.8

3.7. Electrochemical characterization

Cyclic voltammetry (CV) allowed the evaluation of the electrochemical (EC) properties of 2a-g in solid films. EC data are summarized in Table 5 and cyclic voltammograms of BPhTCs are shown in Figs. S12–18. The redox behavior of BPhTCs is characterized by a quasi-reversible process. The formation of π-radical anion is observed in the reduction zone with E_onset(Red) values ranging between −0.42 and −0.74 V, attributed to the BODIPY unit, since it is easier to reduce than the coumarin fragment; these values are in the observed range for meso-aryl-BODIPYs [99]. In the oxidation region the formation of π-radical cation, attributed to the triazole donor, is observed. The E_onset(Ox) values are in the 1.30–1.67 V range, values in the range reported for the starting PhTCs [51]. The EC energy of the HOMO (E_H^EC) and LUMO (E_L^EC) MOs, were obtained from the oxidation and reduction potentials in the onset, respectively [100]. The EC H→L gap (E_H→L^EC) mean value is of 2.09 ± 0.08 eV for 2a-g. The BODIPY moiety largely increases the electron-accepting capability of BPhTCs in comparison to their PhTCs precursors, with more impact in the LUMO than in HOMO energy levels.

Thus, the BODIPY attachment to the PhTC backbone has the effect of lowering the LUMO energy level, allowing the mobility of charges in compounds 2a-g, as noted from optical and EC measurements, and theoretical calculations. In addition, the broad absorption spectrum in solid films, large extinction coefficients and Stokes shifts, could make BPhTCs potentially useful as electron transporting materials.

3.8. OFET electric characterization

To further study the electrical properties of compounds 2a-g, organic field-effect transistors (OFET) were manufactured. As part of a first electrical characterization, films of BPhTCs were deposited on glass slides to measure their electrical resistivity. All compounds showed resistivity values between 10⁴ and 10⁵ Ωcm, listed in Table S9, typical values for semiconducting organic materials [21]. The semiconducting properties of BPhTCs, were measured at room temperature and atmosphere, using a top contact/bottom gate TFT device architecture in which an indium tin oxide (ITO) film on a flexible PET was used as the substrate for the source, drain and gate contacts, and polydimethylsiloxane (PDMS) as the dielectric film.

I–V transfer characteristic curves were obtained by maintaining the source-drain voltage at 15 V and varying the gate voltage from −15 to 15 V forward and backward. The transfer curves are presented on both a linear scale in Fig. S19 and on a logarithmic scale (Log₁₀|I_DS|) in Fig. S20. Compounds 2a-g showed similar behavior between them although in different current scales, ranging from μA (2e-g) to a few hundred mA (2a-d). All compounds displayed ambipolar behavior with typical V-shaped (Log₁₀|I_DS|) transfer curves, where the p-type behavior, characteristic of coumarins and frequently observed in meso-substituted BODIPYs [21], is present in the I–V transfer curves of the 2a-g compounds. p-Type OS are more frequently described whereas the advances in n-type or ambipolar conjugated materials are scarce. Therefore, the n-channel behavior in the BPhTC-TFT devices at room temperature conditions was analyzed with the aim of contributing to this field of knowledge.

The coherent correspondence between the saturation currents, reached in the output curves, and the currents measured in the transfer curves of 2a-g devices, is observed in Fig. S21 for positive V_G values. The output curves are reliable between 4 and 10 V in the gate, other voltages were not tested due to the early degradation of the OS and insulating materials. The carrier mobilities (μ_FET) were calculated considering an n-type channel, from Fig. S22, where the I_DS and (I_DS)^1/2 currents are plotted versus V_G. The standard saturation regime quadratic model μ_FET = (2L/WC)((δ(I_D)^1/2/δV_G))V_D)² was used. From these plots, the I_ON/I_OFF current ratio of the 2a-g-TFT device was also estimated, ranging from 10² (2a-d) to 10⁵ (2e-g). A higher order I_ON/I_OFF ratio is not expected because the devices were operated at relatively low V_G voltages in contrast to the V_G values of up to 100 V used in polymer OFET devices [22]. The electrical characteristics of 2a-g-OFET devices are summarized in Table 6. Even when their complete electric characterization can be found in Figs. S19–22, the characteristic I–V transfer and output curves (drain current versus drain voltage) as well as the transfer I–V curves in logarithmic scale for |I_DS| and linear scale for the |I_DS|^1/2 are shown in Fig. 11(a) and (b) and 11(c),(d), respectively, for 2d- and 2f-OFET devices as representative examples.

Table 6.

Semiconductor characteristics of compounds 2a-g, measured under ambient temperature and atmosphere, in top-contact/bottom-gate TFT device architecture.

BPhTC	μFET/(cm2 V−1 s−1)	ION/IOF
2a	1.6 × 10−2	2.1 × 103
2b	5.0 × 10−2	4.6 × 102
2c	1.0 × 10−2	3.1 × 104
2d	2.2 × 10−2	2.4 × 105
2e	3.2 × 10−4	4.0 × 104
2f	0.77 × 10−4	4.3 × 105
2g	2.9 × 10−4	3.3 × 103

Fig. 11.

Characteristic I–V transfer plots measured in the range of 4–10 V in the gate voltage (V_G) at ambient temperature and atmosphere, and I–V transfer plots in logarithmic scale of |IDS| and linear scale for the |IDS|^1/2, of 2d (a), (b) and 2f (c), (d) OFET devices.

Regardless the chemical structure of the BPhTCs, the saturation regime is reached in the 6–9 V range. However, the mobility of the charge carriers seems to depend on the OS performance, compounds 2a-d exhibit μ_FET values of nearly ∼10⁻² cm² V⁻¹ s⁻¹, two orders of magnitude larger than the values exhibited by 2e-g (10⁻⁴ cm² V⁻¹ s⁻¹). Following a similar analysis for electron charge carriers, the hole mobilities were also estimated. The values are approximately one to two orders of magnitude smaller (7.4× 10⁻⁴ 2a, 6.3 × 10⁻³ 2b, 4.7 × 10⁻³ 2c, 1.0 × 10⁻³ 2d, 3.9 × 10⁻⁶ 2e, 2.5 × 10⁻⁶ 2f, and 6.5 × 10⁻⁶ 2g) than the corresponding electron mobilities in Tables 6 and in agreement with the better performance of compounds 2a-g as n-type OSs. These marked differences are attributed to the inherent properties of the 2a-g solutions to form a uniform and continuous deposit in the channel, whose integrity was tested by confocal microscopy taking advantage of their fluorescent properties. A cross-section view of a TFT device manufactured with OS 2c (in red) on PET-ITO substrate (in gray) is shown in Fig. 12(a–c). Compounds 2a-d were successfully layered in the channel from their THF solutions forming homogeneous deposits of ∼10 μm thick. In contrast, the deposits of OS 2e-g were stratified, generating defects that diminished the performance of the TFT devices. All our efforts to manufacture homogeneous deposits of compounds 2e-g using the Drop-Casting method were unsuccessful. It has been reported that the measured mobilities depend on the method used to place the OS [23]. In addition, the low mobility of 2e-g can be caused by structural defects, low charge interaction on the insulating-semiconductor surface [101] or the molecular packing in the solid state. A brief comparison among 2a-d OS, which were capable to form processable deposits, and the device performance, indicate that substituents in both 6- and 7-positions gave good μ_FET values. Nevertheless, soft ED substituents, 7-NEt₂ and 6-Br groups, showed the best values of I_ON/I_OF ratio. These results point to OSs 2c and 2d as the best materials among this group of compounds with the additional bonus that they can be monitored thanks to their fluorescent properties.

Fig. 12.

(a) Confocal Microscopy of a cross-section of the OFET device manufactured with OS 2c (in red) on PET-ITO substrate (in gray). Deposit thick (b) and channel wide (c) are shown. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Furthermore, OFETs made with BPhTCs 2a-d, showed negligible degradation after one month of storage in air under ambient conditions of light and humidity. Even though, the measured μ_FET values are in the range reported for BODIPYs-OFET devices [21], the charge carrier mobility values, herein measured for the 2a-d-OFET devices, are larger than those reported for other triazoles (3.8 × 10⁻⁵ cm² V⁻¹ s⁻¹) [102], thiophene-BODIPYs (5.3 × 10⁻⁴ cm² V⁻¹ s⁻¹) [17], triphenylamine-BODIPY-triphenylamine triads (10⁻⁵-10⁻⁷ cm² V⁻¹ s⁻¹) [103], carbazole-BODIPY-carbazole triads (10⁻⁵ cm² V⁻¹ s⁻¹) [104] but similar to BODIPY-quaterthiophene-BODIPY crystalline fibers (10⁻² cm² V⁻¹ s⁻¹) [23], dimeric isoindigo-aza-BODIPY (1.1 × 10⁻¹ cm² V⁻¹ s⁻¹) [105], naphthalene-imide-BODIPYs (0.10 cm² V⁻¹ s⁻¹) [106], among others.

Recalling the high degree of crystallinity of these compounds, estimated by XRPD measurements, the efficient charge transport and thus the performance of 2a-g crystals depend not only on the molecular structure but also on non-covalent intermolecular interactions. Therefore, the mobility properties observed for 2a-d can be attributed to their crystalline nature, whose crystal packing is ruled by C–H⋅⋅⋅A (A = N, O) interactions, π-stacking through carbonyl ⋅⋅⋅carbonyl dipolar n→π*, and parallel displaced π ⋅⋅⋅ π∗ interactions [77], as observed in the crystal network of the 2a (vide supra).

4. Conclusions

Herein we report a comprehensive research of a series of structurally related BODIPY-1,2,3-triazolyl-phenyl-coumarin hybrid compounds (BPhTCs). Seven BODIPY labelled phenyl-triazole-coumarin derivatives 2a-g have been synthesized, in moderate yields, using a simple methodology under mild conditions. The photophysical properties of BPhTCs depend on the nature of the coumarin substituent. Optical and electrochemical band-gap values in the solid are as small as ∼2.20 and ∼2.09 eV, respectively, ranking them as good prospects for semiconducting purposes which according with the substituent in the coumarin behave as A-D dyes being D the BODIPY or coumarin. Top gate/bottom-contact solution processed field-effect transistors of BPhTCs exhibit ambipolar behavior which exhibited n-channel charge carrier mobilities as high as 10⁻² cm² V⁻¹ s⁻¹ and current on/off ratios of 10²-10⁵ under ambient conditions. It was also demonstrated, through single crystal X-ray analysis of 2a∙DCB, that the crystal network is developed through highly dispersive interactions: n-π* and π-π* stacking interactions, as well as C–H ⋅⋅⋅ A (A = N, O, π, F) close contacts, which rule the stabilization energy of the crystal and are responsible for transporting the charge carriers. Further efforts to obtain single crystals of this class of OSs for X-ray diffraction should be done to a better understanding of those supramolecular features that rule the performance of the OFET devices. Thus, the potential of BPhTCs as crystalline building blocks for small molecules organic semiconductor purposes with large Stokes shift, good thermal and air stability up to 250 °C, and good shelf life was demonstrated. The OSs 2c and 2d are the best materials among this group of compounds with the additional bonus that they can be monitored thanks to their fluorescent properties. The medium yields as well as the lack of processability of the most polar derivatives 2e-g to form the OFET devices limit their high scale use. It is worth highlighting the use of THF both in optical studies and in the manufacturing of OFETs, which is considered an environmentally friendly solvent, out of the list of ‘substances of very high concern’ (SVHC). Finally, this work contributes on fundamental understanding between the relationship of the organic structures and charge transport properties. As a future proposal, the synthesis of a polymer (D-A)_n using the structures of 2a-g as a repeating unit and including intermolecular spacers is suggested. For this purpose, the synthesis of new precursors enabling this polymerization must be considered, using environmentally friendly synthetic methods such as microwave and/or mechano-synthesis. The polymer could have better processability and film formation, which could improve OFET performance.

Data availability statement

Authors agree to make our data publicly available. Most of them are available in the Supplementary Material and crystallographic data was deposited as CCDC-1994818.

CRediT authorship contribution statement

José Emilio de la Cerda-Pedro: Writing - original draft, Methodology, Investigation. Oscar Javier Hernández-Ortiz: Methodology, Investigation, Formal analysis. Rosa Angeles Vázquez-García: Supervision, Methodology, Formal analysis. Efrén V. García-Báez: Methodology, Formal analysis. Ramón Gómez-Aguilar: Validation, Methodology, Investigation, Formal analysis. Arián Espinosa-Roa: Software, Methodology. Norberto Farfán: Supervision, Funding acquisition, Conceptualization. Itzia I. Padilla-Martínez: Writing - review & editing, Writing - original draft, Supervision, Resources, Project administration, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Itzia Irene Padilla-Martinez reports financial support was provided by National Council on Science and Technology (CONACYT now CONAHCYT, National Council on Humanities, Science and Technology), Mexico. Norberto Farfan reports financial support was provided by National Council on Science and Technology, Mexico. Jose Emilio de la Cerda-Pedro reports financial support was provided by National Council on Science and Technology, Mexico. Oscar Javier Hernandez-Ortiz reports financial support was provided by National Council on Science and Technology, Mexico. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article.