Polyfluorenes with on-chain dibenzoborole units—Synthesis and anion-induced photoluminescence quenching


  • Vasco D. B. Bonifácio,

    Corresponding author
    1. Macromolecular Chemistry Group, University of Wuppertal, Gaußstraße 20, D-42119 Wuppertal, Germany
    2. Instituto Superior Técnico, Departamento de Eng. Química e Biológica, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
    • Macromolecular Chemistry Group, University of Wuppertal, Gaußstraße 20, D-42119 Wuppertal, Germany
    Search for more papers by this author
  • Jorge Morgado,

    1. Instituto Superior Técnico, Departamento de Eng. Química e Biológica, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
    2. Instituto de Telecomunicações, Instituto Superior Técnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
    Search for more papers by this author
  • Ullrich Scherf

    1. Macromolecular Chemistry Group, University of Wuppertal, Gaußstraße 20, D-42119 Wuppertal, Germany
    Search for more papers by this author


original image

The photoluminescence response of a novel, random 9,9-dialkylfluorene/dibenzoborole copolymer toward a series of nine inorganic anions was studied both in solution and in thin films. Photoluminescence quenching with sensitivity in the micromolar range was observed for fluoride, cyanide, and iodide. Hereby, the interaction of the copolymer with fluoride/cyanide or iodide follows different quenching mechanisms. Such copolymers are promising solid-state anion sensors. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]


Chemical sensing based on conjugated polymers using photoluminescent-based sensing schemes has emerged in the last few years.1 In this field, the search for highly selective and sensitive probes for the relevant analytes has been the primary target. A major challenge in combating terrorism that uses Chemical Warfare Agents (CWAs) is a reliable and affordable detection of toxic gases.2 Since nerve gases are odorless and colorless chemicals, a variety of detection methods has been developed, which included classic analytic methods, and also colorimetric or fluorimetric chemosensores.3 However, in both cases limitations such as low selectivity, low sensitivity, operational complexity, nonportability, difficulties in real-time monitoring, and false-positive readings need to be overcome. Fluorine and cyano substituents are present in military nerve gases, fluoride and cyanide are subsequently formed during their decomposition4 (Fig. 1). In addition, both anions are contaminants in drinking water.5

Figure 1.

Examples of military nerve gases which release fluoride (sarin and soman) and cyanide (tabun).

Many examples of low-molecular weight organoboron molecules, oligomers, and small molecules have been reported to selectively sense fluoride anions;6 however, only few examples of conjugated organoboron polymers have been reported.7, 8 The applicability of conjugated polymers in chemosensing processes results from their ability to generate an amplified response to analytes. Their increased sensitivity (signal amplification) is a consequence of a collective response in the system. Taking advantage of the signal amplification in conjugated polymers and the strong affinity of organoboron compounds toward inorganic anions as fluoride, the design of novel conjugated polymers with organoboron moieties is an attractive target.


Materials and Characterization

The reactions were carried out using dried solvents, under inert nitrogen atmosphere and using Schlenk techniques or a glove box. 2,7-Dibromo-9,9-di-n-octylfluorene9 and 4,4′-dibromo-6,6′-diiodobiphenyl101 (3-step synthesis from 2,5-dibromonitrobenzene) were obtained according to literature procedures. For the titration experiments, samples were prepared using a microbalance, a microliter syringe, and volumetric glassware. The photoluminescence quenching experiments in THF solution was performed by successive addition of solutions of the anions in THF. Films of submicrometer thickness were drop-cast from dilute (ca. 0.025 w/v %) solution of the polymer in toluene on glass slides. For the solid-state fluorescence quenching experiments, the films were dipped for 60 s into solutions of the corresponding salts in ethanol. 1H and 13C nuclear magnetic resonance (NMR) spectra were recorded on a Bruker ARX 400 using tetramethylsilane as an internal reference peak. 11B NMR spectra were recorded using BF3·Et2O as internal reference peak. Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) spectrum was recorded on a Brucker REFLEX mass spectrometer. The mass spectrum was obtained with an accelerating potential of 20 kV in positive ion and reflector mode. Dithranol was used as matrix and KCl as a cationating agent.

The UV–vis and fluorescence spectra were recorded on a Jasco V-550 spectrophotometer and a Varian-Cary Eclipse spectrometer, respectively. Molecular weight determinations via gel permeation chromatography (GPC) were performed using a Spectra 100 GPC column (5-μm particles) eluted with THF at 30 °C (flow rate 1 mL/min and concentration of the polymer ca. 1.5 g/L). The calibration was based on polystyrene standards with narrow molecular weight distribution. All the GPC analyses were performed on solutions of the polymers in THF at 30 °C. The thermogravimetric analyses (TGA) was carried out using a Mettler TG50 at 10 K/min under argon from 35 to 600 °C (Scheme 1).

Scheme 1.

Synthesis of the random fluorene/dibenzoborole copolymer BPFO-10.

Preparation of the Monomer

2,7-Dibromo-9-(4-cyanophenyl)borafluorene-2,7-diyl 2

To the solution of 4,4′-dibromo-2,2′-diiodo-biphenyl 1 (1 g, 1.77 mmol) in dry THF (15 mL), n-butyllithium (1.6 M in hexane, 2.3 mL, 3.64 mmol) was added at −100 °C. The resulting mixture was stirred at this temperature for 30 min, and then 4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)benzonitrile (1.77 mmol) was added. After the addition, the mixture was allowed to warm up to room temperature and stirred overnight. Then the reaction mixture was hydrolyzed with water and extracted with diethyl ether. The combined organic layers were dried over anhydrous MgSO4 and the solvent evaporated to dryness. The crude product was purified by column chromatography using silica gel, and dichloromethane/hexane (1:1) as an eluent to afford the target compound as a white solid (85 mg, 6% yield).

Rf: 0.40 (SiO2, hexane/CH2Cl2 1:1). mp: 97–99 °C. 1H NMR (400 MHz, CDCl3, δ, ppm): 7.61 (4H, d, J = 8.28 Hz, ArH), 7.32 (4H, d, J = 8.11 Hz, ArH), 7.14 (2H, s, ArH). 13C NMR (100.6 MHz, CDCl3, δ, ppm): 153.01; 138.38; 135.09; 132.41; 131.22; 130.62; 126.16; 124.29; 122.41; 122.39; 119.35. MS (MALDI-TOF) (m/ z): calcd for C19 H10 BBr2N: 422.93; found, 461.45 [C19 H10 BBr2N+K]+.

Preparation of the Polymer

Random (9,9-Di-n-octylfluorene-2,7-diyl)/[9-(4-cyanophenyl) borafluorene-2,7-diyl] Copolymer (BPFO-10)

A Schlenk-flask was charged under argon with 2,7-dibromo-9,9-di-n-octylfluorene (467.2 mg, 0.848 mmol), 2,7-dibromo-9-(4-cyanophenyl)borafluorene-2,7-diyl 2 (40 mg, 0.0944 mmol), Ni(COD)2 (624 mg, 2.270 mmol), and 2,2′-bipyridyl (354.4 mg, 2.270 mmol). COD (176 μL, 1.418 mmol) and THF (25 mL) were added and the mixture was heated up to 80 °C for 3 days. After this period, bromobenzene was added and the mixture heated for additional 3 h. The solution was poured into an aqueous 2 N hydrochloric acid and extracted with chloroform. The organic layer was washed with a saturated EDTA solution, dried over anhydrous MgSO4 and the solvent evaporated to dryness. The residue obtained was Soxhlet-extracted with methanol (1 day) and ethyl acetate (3 days), and then dissolved in chloroform and reprecipitated into cold methanol (1:100) to yield the random copolymer BPFO-10 as yellow pellets (295 mg, 83% yield). The copolymer showed thermal stability up to 350 °C in TGA.

GPC (THF): Mn = 22,888; Mw = 40,731, PD = 1.78 (after extraction).

1H NMR (400 MHz, CDCl3, δ, ppm): 0.81 (6H, br, CH3), 1.14 (20H, br, CH2), 1.52 (4H, br, CH2), 2.12 (4H, br, CH2), 7.67 (br, ArH), 7.83 (br, ArH). 13C NMR (100.6 MHz, CDCl3, δ, ppm: 151.84; 140.55; 140.06; 134.02; 130.98; 126.19; 121.54; 119.96; 55.37; 40.40; 31.80; 30.05; 29.22; 23.95; 22.60; 14.04. 11B NMR (128.4 MHz, CDCl3): δ (ppm) 43.18. Anal. calcd for (C280H370BN)n (3760.8)n: C 89.42, H 9.92. Found: C 88.62, H 10.57.


It has been postulated7b that ortho-substituents at one aryl group, usually 2,4,6-trimethylphenyl (mesityl) and 2,4,6-triisopropylphenyl (tripyl), are necessary for a good environmental stability of triarylboron compounds. The resulting “cage” around the boron center should protect it against nucleophilic attacks. In this case also, the Lewis acidity of the boron center is reduced due to the electron-donating nature of the alkyl protecting groups. This shortcoming has been overcome by an incorporation of the triarylboron building blocks into larger π-conjugated frameworks. The synthesis of “unprotected” 9-aryl-dibenzoboroles has been previously reported.11 In this study, we surprisingly found that also an “unprotected” dibenzoborole that is electronically stabilized by a cyano group in the para-position of the 9-phenyl substituent exhibits good environmental stability.12 Unfortunately, other “unprotected” dibenzoboroles synthesized during this investigation were not stable and decomposed during purification or storage (unpublished results).

The low yield obtained in the preparation of the dibenzoborole monomer should be mainly attributed to the steric hindrance of the pinacolboronate reagent.13 However, we did not systematically investigate the coupling reaction toward the desired boron monomer. The random 9,9-dialkylfluorene/dibenzoborole copolymer was synthesized in a Yamamoto-type aryl–aryl coupling using a 90:10 molar ratio of both comonomers 2,7-dibromo-9,9-di-n-octylfluorene and 2, respectively. Since the free access to the sensitive subunits and not their distribution is the most important issue in chemosensing applications, the simple approach to random copolymers by Yamamoto-type coupling was chosen in our investigations. The 1:9 molar ratio of both comonomers was selected to provide first evidence on the potential and versatility of this class of copolymers as anion-sensing materials without the aim to identify the detection limits. The copolymer is well soluble in chloroform and THF. The resolution of the 1H- and 13C NMR spectra of the copolymer BPFO-10 was insufficient to assign all proton and carbon signals of the minority component. However, the incorporation of 2 in the copolymer was confirmed by 11B NMR.

The quantitative response of the random copolymer BPFO-10 against nine different anions, I, Br, Cl, F, BF4, CN, HSO4, H2PO4, and NO3 (using the respective tetrabutylammonium salts in a THF solution), was recorded both in solution (28.7 nM in THF, calculated as the boron-unit concentration) and in drop-cast thin films. Hereby, the initial UV–vis absorption and fluorescence emission intensity were measured and the photophysical response, upon the addition of increasing concentrations of the anion, was monitored by absorption and fluorescence spectroscopy. The excitation wavelength used for solution and thin film experiments was 387 nm, within the long wavelength absorption band of BPFO-10. Some steric hindrance influence of the alkyl side chains of the dialkylfluorene unit on the nucleophilic attack of the anions toward the boron center can be expected, thereby reducing the photoluminescence response.14

Arylboron derivatives have been reported to bind to fluoride anions with high selectivity. However, little has been reported regarding their response to related anions (cyanide) and their interference with the fluoride sensing. We have found that BPFO-10 is also highly fluoride-sensitive but cyanide,15 iodide, and with a lower sensitivity also H2PO4 anions can act as photoluminescence quenchers and interfere with the fluoride detection. By contrast, BPFO-10 does not show any significant response toward Br, Cl, BF4, HSO4, and NO3−16.

When fluoride and cyanide anions are added to BPFO-10, defined anionic complexes are formed with the boron of the dibenzoborole unit showing characteristic isosbestic points at 490 (for F) and 510 nm (for CN). In both cases also the intensity of the long wavelength absorption maximum of the copolymer (388.5 nm) slightly decreases, and a new UV band at 273 nm is formed.

Figure 2(a,b) illustrates an only incomplete fluorescence quenching during fluoride and cyanide addition, which is in accordance with a static quenching mechanism. In the case of iodide, a nearly complete fluorescence quenching of BPFO-10 is observed [see Fig. 2(c)]. However, the absence of an isosbestic point and a different absorption behavior (increase of the long wavelength absorption band) point toward a collisional-quenching mechanism is this case.

Figure 2.

Photoluminescence quenching of BPFO-10 in THF solution, λexc = 387 nm. (a) TBAF, inset: absorption of BPFO-10 at: ▪ (λmax = 273 nm) and □ (λmax = 388.5 nm). (b) TBACN, inset: absorption of BPFO-10 at: ▪ (λmax = 273 nm) and □ (λmax = 388.5 nm). (c) TBAI, inset: absorption of BPFO-10 at: ▪ (λmax = 295.5 nm) and □ (λmax = 388.5 nm).

The Stern-Volmer quenching constants (KSV) were calculated from the initial slopes (Fig. 3 and Table 1). The highest KSV-value is observed for iodide quenching, while F and CN lead to significantly lower quenching constants.

Figure 3.

Stern-Volmer plots for the fluorescence quenching of BPFO-10 for different anions.

Table 1. Stern-Volmer Constants for the Fluorescence Quenching of BPFO-10 with F, CN, and I
QuencherKSV (×106) (M−1)
F0.49 ± 0.02
CN4.78 ± 0.72
I23.2 ± 0.90

The calculated KSV value for fluoride detection was found to be higher than the reported literature values for nonpolymeric organoboron compounds (ca. 6-fold,14 75-fold,6(f) and 6000-fold8), except for organocompounds with neighboring H-bounded donor groups.6(e,i)

To evaluate the potential applicability of BPFO-10 as a solid-state chemosensor, quenching experiments were also performed on drop-cast thin films. The obtained quenching plots follow a similar trend as observed for the THF solutions (see Fig. 4).

Figure 4.

Photoluminescence quenching of BPFO-10 spin-coated thin films (λexc = 387 nm).

In the solid-state quenching experiments, fluoride and cyanide could be detected down to a concentration of ∼ 0.1 mM. It should be mentioned that the boron content of the copolymer can be increased by a variation of the monomer feed ratio. For higher boron contents a higher sensitivity is expected. Based on the obtained results, a potential use in the construction of fluoride/cyanide or iodide-selective sensing devices could be envisaged, especially in combination with alternate sensing schemes. BPFO-10 may be useful in distinguishing between F/CN (fluorescence quenching) and Br/Cl (no response).


In conclusion, we have synthesized a novel, random conjugated dialkylfluorene/dibenzoborole copolymer containing “unprotected” B-aryldibenzoborole building blocks. The response of the copolymer toward a series of anions was evaluated in photoluminescence quenching measurements, both for THF solutions and drop-cast thin films. Following different quenching mechanisms, the copolymer BPFO-10 especially responses to fluoride/cyanide and iodide addition. The lack of selectivity between cyanide and fluoride may be of secondary importance concerning a potential application in the detection of toxic gases and their decomposition products (e.g., sarin vs. tabun). Iodide is not present in the analytical protocols developed for the hydrolysis of nerve gases, which uses a Cu(II)-bipy chelate made from a CuCl2 stock solution. As demonstrated, BPFO-10 does not show any response toward chloride. In future, studies toward anion-sensitive conjugated polymers related organoboron copolymers of varying composition will be studied.


V. D. B. Bonifácio thanks Fundação para a Ciência e Tecnologia (FC&T, Lisbon, Portugal) for the award of a research fellowship (SFRH/BPD/27148/2006).