Conductive Polymer Functionalization by Click Chemistry

Conductive Polymer Functionalization by Click Chemistry

(Parte 1 de 3)

Conductive Polymer Functionalization by Click Chemistry

Anders Egede Daugaard and Søren Hvilsted*

Danish Polymer Centre, Department of Chemical and Biochemical Engineering, Technical UniVersity of Denmark, Building 423 ProduktionstorVet, DK-2800 Kgs. Lyngby, Denmark

Thomas Steen Hansen and Niels B. Larsen

Polymer Department, POL-313, P.O. Box 49, FrederiksborgVej 399, DK-4000 Roskilde, Denmark ReceiVed December 7, 2007; ReVised Manuscript ReceiVed April 21, 2008

ABSTRACT: Click chemistry is used to obtain new conductive polymer films based on poly(3,4-ethylenedioxythiophene) (PEDOT) from a new azide functional monomer. Postpolymerization, 1,3-dipolar cycloadditions in DMF, using a catalyst system of CuSO4 and sodium ascorbate, and different alkynes are performed to functionalize films of PEDOT-N3 and copolymers prepared from EDOT-N3 and 3,4-ethylenedioxythiophene (EDOT). This approach enables new functionalities on PEDOT that could otherwise not withstand the polymerizationconditions.Reactionson the thin polymerfilms have been optimizedusing an alkynatedfluorophore, with reaction times of ∼20 h. The applicability of the method is illustrated by coupling of two other alkynes: a short chain fluorocarbon and a MPEG 5000 to the conductive polymer; this alters the advancing water contact angle of the surface by +20° and -20°/-25°, respectively. The targeted chemical surface modifications have been verified by X-ray photoelectron spectroscopy analysis.


Conductive polymers have been extensively studied during the past few decades for applications such as biosensors,1 strain gauges,2 organic solar cells,3 or organic light-emitting diodes.4 Many challenges have to be met in the development of new functional conducting polymers, including synthesis of monomers with the required functionality and establishment of polymerizationconditionscompatiblewith the targetedfunction. Specific problems may be instability of large biological molecules for biosensors under the typically harsh polymerization conditions or inhibition of the polymerization process by a side reaction with a functional group in the monomer. Functionalization of the conductive polymer may result either from incorporation of the functional entity into the monomer during monomer synthesis or by postpolymerization coupling of the targeted functionality to built-in generic coupling sites in the monomer unit. Recently,the term “click chemistry”was defined by Sharpless et al.5 as selection criteria for highly efficient coupling reactions. Ideally, the click reaction can be performed in water or organic solvents at room temperature, making the method suitable for most applications. Click chemistry has mainly been based on the use of 1,3-dipolar cycloadditions of azides and alkynes under copper catalysis6,7 and Diels-Alder reactions. In polymer chemistry it has been used in numerous ways e.g. for end-group functionalization,8 polymer to polymer couplings,9 functionalization of linear polymers with selected groups,10 dendrons,1 and poly(ethylene glycol) (PEG).12 Also, in reactions on surfaces the cycloaddition has been used with good results e.g. to bond metal surfaces together.13 It has been used with carbon nanotubes14 and for functionalization of selfassembled monolayers on gold.15,16 Recently click chemistry has also been used for surface reactions on cotton17 and glass.18,19 A bioactive surface has been prepared by introduction of biotin by click chemistryon a polymer substrate.20 For sensor applications especially biological systems are of great interest, and these often require mild reaction conditions, which can be obtained using click chemistry.21 A comprehensive review on macromolecular click chemistry has recently been published by Binder et al.2

The new developed conducting polymer is based on PEDOT, used in numerous applications due to its high electrical conductivityand high stability in ambient and aqueous environments.23 During polymerizationthe polymer becomes insoluble, and further functionalization becomes difficult. Combination of the conductive properties of PEDOT and the advantages of modularity, high selectivity, and high yields of click chemistry permits preparation of new PEDOTs for many different applications, e.g., surfaces with conductive properties or sensors. Both pre- and postfunctionalizationhave been studied by others. With regards to biological active molecules the most successful methods so far is physical absorption after polymerization or entrapment during polymerization.24 Covalent postfunctionalization has been achieved through peptide bonding.25,26 The click approach is a good alternative that gives a controlled functionalization without protective groups through high selectivity. In our group we have focused on the use of PEDOT for different applications27–30 and the use of click chemistry for polymer functionalization.31 Here we present a standardized method for postpolymerization functionalization of PEDOT to obtain conductive polymer surfaces with various functionalities.

Experimental Section

General Methods. Thin layer chromatography (TLC) was performed on Merck plates coated with silica gel F254. Kieselgel for column chromatography was Merck Kieselgel 60 (230-400 mesh). 1H NMR was run on a 250 MHz cryomagnet from Spectrospinand Bruckerat room temperature.Infraredspectroscopy (IR) was performed on a Perkin-Elmer Spectrum One model 2000 Fourier transform infrared system with a universal attenuated total reflection sampling accessory on a ZnSe/diamond composite. Differential scanning calorimetry (DSC) was performed on a DSCQ1000 from TA Instruments. The thermal analyses were performed at a heating and cooling rate of 10 °C/min. The melting temperatures (Tm) are reported as the peak temperatures of the endothermic melting peaks. The conductivity was measured with a four-point probe (Jandel Engineering Ltd., Linslade, UK) connected to a four-pointsource meter (Keithley2400, Cleveland,OH). Film thickneses were measured with an Ambios XP-2 (Ambios* Corresponding author. E-mail:

10.1021/ma702731k C: $40.75 2008 American Chemical Society Published on Web 05/29/2008

Technology,Inc., Santa Cruz, CA) profilometerusing a stylus force of 0.5 mg. Optical microscopy images was recorded with a AxioCam MRc 5 camera mounted on a Zeiss Axioskop 40 microscope (Oberkochen, Germany). Fluorescence analysis was conducted with a Zeiss Filter Set 09 (excitation 450-490 nm, emission >515 nm). XPS analysis was performed on a Thermo Fisher Scientific K Alpha (East Grinstead, UK) using monochromatized aluminum KR radiation in a 400 µm spot on the sample. Survey and high-resolution spectra were acquired and analyzed using the manufacturer’s Avantage software package. Spectra were generally acquired with electron charge compensation in operation to avoid sample charging, except for a series of measurements to determine the detrimental effects of electron flooding on azide functional gruops. Atomic force microscopy analysis proceeded on a PSIA XE-150 instrument operating in intermittent contact mode with BudgetSensor Tap-300 cantilevers.

Chemicals. Chemicals except for Baytron C were acquired from

Aldrich and were used as received unless otherwise specified. Baytron C was purchased from H.C. Starck. Fluorescein methyl ester was prepared in accordance with Moore et al.32

3,4-(1-Bromomethylethylene)dioxythiophene, 1 (EDOT-Br). 3,4-Dimethoxythiophene (0.41 g, 2.8 mmol), 3-bromo-1,2-propanediol (1.1 g, 7.2 mmol), and p-toluenesulfonic acid (0.08 g, 0.4 mmol) were dissolved in toluene (30 mL) and stirred at 100 °C for 48 h. Toluene was removed in vacuo, and the residue was dissolved in CH2Cl2 and extracted with Na2CO3 and H2O. The organic phase was dried with MgSO4, filtered, and concentrated in vacuo, and the crude product was purified by column chromatog- raphy with a gradient eluent of heptane/ethyl acetate (EtOAc). The product was isolated as a colorless oil (0.24 g, 37%). IR (cm-1):

(0.2 g, 0.9 mmol) and NaN3 (0.08 g, 1.2 mmol) were dissolved in DMF (10 mL) and stirred at room temperature (RT) for 17 h. The reaction mixture was diluted with H2O (15 mL), and the aqueous DMF was extracted with EtOAc (5 × 15 mL). The organics were

combined and extracted with H2O( 3 × 15 mL) and brine (1 × 15 mL), dried with MgSO4, filtered, and concentrated in vacuo to give the product as a colorless oil (0.18 g, 97%). IR (cm-1): 3114 (C-H

General Polymerization Method for 2, to PEDOT-N3,3 . The polymerization method was based on an earlier published method for the polymerization of EDOT.3 A number of microscope slides were thoroughly cleaned using acetone, isopropanol, ethanol, and water. The glass slides were surface modified by vapor phase hexamethyldisilazane (HMDS) in a dedicated oven (Yield Engineering Systems 6112). 2 (20 mg,0.15 mmol), Baytron C (0.48 mL, ∼40 wt % Fe(I)Tos in butanol), and butanol (0.48 mL) were mixed and spin-coated on the glass-slides (10 s at 1000 rpm). The samples were placed on a hot plate at 65 °C for 5 min and subsequently washed with water and blown dry in a nitrogen flow, yielding films with a thickness of 200-250 nm.

General Copolymerization Method for Poly(3,4-ethylenedioxythiophene-co-3,4-(1-azidomethylethylene)dioxythiophene). The copolymerization method was based on an earlier published method for the polymerization of EDOT.3 A solution of 3,4- ethylenedioxythiophene (EDOT, 0.2 mL), Baytron C (6.5 mL), butanol (6.5 mL), and pyridine (0.15 mL) was mixed with the

EDOT-N3 solution mentioned above to yield solutions containing

spin-coated onto the HDMS treated glass slides (10 s at 1000 rpm). The samples were placed on a hot plate at 65 °C for 5 min and subsequently washed with water and blown dry in a nitrogen flow.

General Ester Synthesis, 2,2,3,3,3-Pentafluoropropyl Pent- 4-ynoate, 10. A solution of 4-pentynoic acid (0.60 g, 6.1 mmol), dimethylaminopyridine (DMAP, 0.12 g, 0.9 mmol), and 2,2,3,3,3- pentafluoropropanol (0.97 g, 6.5 mmol) in CH2Cl2 (15 mL) was stirred at RT, and a solution of N,N′-dicyclohexylcarbodiimide

(DCC, 1.58 g, 7.6 mmol) in CH2Cl2 was added dropwise. The reaction mixture was stirred overnight at RT, filtered, and concen- trated in vacuo. The crude product was purified by column chromatography using a gradient eluent of pentane/ether and gave a colorless oil (1.24 g, 8%). IR (cm-1): 3314 (CtC-H stretch); 2119 (CtC stretch); 1761 (O-CdO stretch); 1197, 1143, 1107

Methyl 2-(3-Oxo-6-(prop-2-ynyloxy)xanthen-9-yl)benzoate, 1. Fluoresceinmethyl ester (2.0 g, 5.8 mmol), triphenylphosphine (TPP, 4.5 g, 17.3 mmol), and propargyl alcohol (0.98 g, 17.5 mmol) were stirred in acetonitrile/THF (50/50) at 0 °C. Diethylazodicarboxylate (DEAD, 3 mL, 17.3 mmol) was added slowly, and the mixture was stirred overnight at RT. The crude mixture was poured onto a Kieselgel column and chromatographed using an eluent of 80/20 CH2Cl2/ether, followed by a gradient of CH2Cl2/ EtOAC. The product was collected from the top of the column

3.59 (s, 3H, O-CH3); 5.0 (s, 2H, ≡C-CH2-O); 6.26 (s, 1H, Ar-H); 6.40 (m, 1H, Ar-H); 6.7-7.0 (m, 3H, Ar-H); 7.28 (s,

1H, Ar-H); 7.50 (m, 1H, Ar-H); 7.7-8.0 (m, 2H, Ar-H); 8.21 (s, 1H, Ar-H). r-Methoxypoly(ethylene glycol)-ω-pent-4-ynoate, 12. The product was prepared according to the general procedure for ester synthesis by DCC coupling, using a commercially available methoxypoly(ethylene glycol) (Mn ) 5010, PDI ) 1.1), 1.2 equiv of 4-pentynoic acid, DCC and 1 equiv of DMAP relative to the end group. The crude product was purified by precipitation in cold dry diethyl ether and dried in vacuo to give 12 as a solid polymer and 10 (62.8 mg, 0.27 mmol) were dissolved in H2O/THF (50/50,

25 mL), aqueous CuSO4 (0.21 mL, 1 M, 0.21 mmol) and sodium ascorbate (0.43 mL, 1 M, 0.43 mmol) were added, and the reaction mixture was stirred at RT overnight. THF was removed in vacuo, and the residue was dissolved in CH2Cl2, extracted with brine, H2O, dried with MgSO4, filtered, and concentrated in vacuo. The crude product contained a minor residue of starting material that could be removed by column chromatography (EtOAc/heptane); the product was a colorless oil (7.7 mg, 72%). IR (cm-1): 3112 (C-H stretch); 1756 (O-CdO stretch); 1190, 1140, 1106 (C-F stretch).

mL) and mixed with a solution of CuSO4 (10 µL, 0.1 M, 1 µmol) and sodium ascorbate (20 µL, 0.1 M, 2 µmol) in DMF (0.1 mL).

The reaction mixture was placed on the surface of 3 and left there for 20 h. The surface was rinsed with H2O and DMF and finally dried with pressurized air. The films were reoxidized by immersion min, followed by rinsing with H2O and drying with pressurized air.

Click Reaction of 3 and 10, 21. The product was prepared according to the general click procedure on 3 using 10(4.0 mg,

17.4 µmol), CuSO4 (10 µL, 0.1 M, 1 µmol), and sodium ascorbate (20 µL, 0.1 M, 2 µmol) in DMF for 20 h.

4322 Daugaard et al. Macromolecules, Vol. 41, No. 12, 2008

Click Reaction of 3 and 12, 2. The product was prepared according to the general click procedure on 3, using 12(12.3 mg,

Results and Discussion

In order to prepare a PEDOT for use in the click approach, it was necessary to synthesize either an alkyne or an azide functional monomer based on 3,4-ethylenedioxythiophene (EDOT). It was decided to prepare the azide monomer, since that was expected to be able to withstand the polymerization conditions. The new monomer was synthesized as shown in Scheme 1. A transetherification using 3-bromo-1,2-propanediol gives the bromine functional monomer, 1, and this is subse- quently substituted using NaN3 to give the product monomer

EDOT-N3, 2. The reactivity of the monomer was checked in a model reaction using standard click conditions and gave the expected product, 13.

The polymerization method published earlier33 was not efficient with this new monomer that appears to have a lower reactivity compared to ordinary EDOT. The polymerizationrate under standard conditions is limited by pyridine. However, by omitting pyridine from the mixture, the polymerization of

EDOT-N3 occurredunder otherwisesimilarconditionsas shown in Scheme 2.

The prepared polymer is insoluble and thus cannot be characterized using NMR or SEC. However, surface characterization using XPS gives a good correlation between the expected composition and the formed substrate.

To monitor the reaction on the surface, it was decided to produce a fluorescent alkyne and detect this by fluorescence microscopy/spectroscopy. Fluorescein was chosen as precursor since it is an inexpensive fluorophore and has a high quantum yield, even after modification of the phenol and the acid group. It was expected that only a surface layer would be functionalized, and this would be difficult to detect with other techniques. In addition to this, the UV/vis spectrum of fluorescein could also be used to estimate the degree of reaction at higher concentrations.

The alkyne functional fluorophore was synthesized as shown in Scheme 3. First the methyl ester of fluorescein was prepared through an ester synthesis in accordance with the method by Moore et al.32 In the second step the alkyne was introduced using a Mitsunobu ether synthesis in moderate yields, 48%.

At first the approach was to perform the click reactions on the polymerin H2O with a catalystsystemof CuSO4 and sodium ascorbate. To make the fluorophore water-soluble, it was attempted to deprotect 1 by the method of Balakirev et al.34 in order to perform the click reactions in water. This approach was ineffective and was abandoned. The solvent was replaced by DMF, and the catalyst system of CuSO4/sodium ascorbate was applied together with 1, which was much more efficient as shown in Scheme 4.

A reference reaction was performed without the CuSO4 catalyst but with otherwise equivalent conditions. Fluorescent microscopy images of the clicked surface and the reference surface are presented in Figure 1. The distinct fluorescence of the clickedsurfaceclearlyshows that the reactionhas proceeded, while the low fluorescence level of the nonclicked reference surface suggests that physical adsorption of the fluorescent reactant can be disregarded.

Fluorescent microscopy was performed on a pristine film of 1 spin-coated on a glass slide (not shown) and compared to the clicked sample. The detected fluorescence from 1 was significantly stronger than the clicked sample. This can be attributed to a combination of incomplete surface coupling and fluorescence quenching by the conducting polymer backbone. Ramanaviciene et al.35 report that the conducting polymer polypyrrole significantly quenches fluorescein and rhodamine B fluorescence. A similar behavior was observed when a thin layer of 1 was deposited on a PEDOT-N3 film. The thicknessof the clicked film and a referencefilm exposed to the same conditions except CuSO4 was measured with a mechanical profilometer before and several days after the reaction. The thickness of the clicked film increased from 240 to 420 nm ((10 nm), corresponding to an increase of 75%. The thickness of the reference film was approximately un- changedby the exposureto the reactionmixturewithoutCuSO4, and a thickness of 235 nm before and 220 nm after the exposure was found. The significantvolume expansion of the clicked film shows that the reaction not only is limited to the surface layer but also is occurring in the bulk phase. In addition to this the unchanged thickness of the reference film shows that the increase in thickness cannot be due to swelling or adhered reagents. The topography of the clicked sample (14) was investigated by atomic force microscopy (AFM) as shown in Figure 2. The surface is relatively uniform even after the large increase in thickness, and the roughness is only a few nanometers. This clearly shows that the reaction has been conducted evenly across the polymer film surface and that reactions in the bulk did not give rise to topographic differences on the surface.

X-ray photoelectron spectroscopy (XPS) was additionally used to investigatethe clickedsurfaces.The high-resolutionpeak of nitrogen shown in Figure 3 can be used to check for remaining azides on the surface due to the difference in binding energies of nitrogen in the triazole and the azide. The azide nitrogens exist in two different oxidation states resulting in two XPS peaks, one at 405 eV and one at 400 eV in a ratio of 1:2, whereas the triazole should give only one peak at 400 eV. The ratio of the two peaks in PEDOT-N3 shown in Figure 3 is between 1:3 and 1:4. This deviation is a result of degradation of the azides during the XPS analysis. The analyses were generallyundertakenwith electronchargecompensationto avoid charging of the less conductive film samples. XPS analysis without electron charge compensation on the highly conductive pure PEDOT-N3 showed a constant peak ratio of exactly 1:2 for extensive analysis times, followed by an immediate decrease

Scheme 1. Monomer Synthesis Scheme 2. Polymerization of the Azide Functional Monomer 2

Macromolecules, Vol. 41, No. 12, 2008 Conductive Polymer Functionalization 4323

in ratio to below 1:3 after switching on charge compensation. The relative nitrogen content also decreases with analysis time (see Table 1, second column), where especially the intensity of the peak at 405 eV is weakened. This deviation from the 1:2 ratio is in agreement with the results obtained by Shannon et al.19 The XPS results clearly show the difference in binding energies from PEDOT-N3 to the product triazole, where no residual azide was detected. The absence of azide nitrogens thus shows that the reaction has proceeded, though it cannot be concluded that all azides have reacted, since nonreacted azides could be degraded during the analysis. In combination with the data from the model reaction, fluorescence spectroscopy, thickness measurements, and UV/vis data (examples given in Figures 4, 5, and 8), this clearly shows that the click reaction has been performed on the polymer substrate.

The conductivity of the films is not surprisingly affected by the treatment. Before reaction the conductivity of the films is around 60 S/cm. During the reaction the films are reduced with sodium ascorbate, causing a dramatic reduction in conductivity to around 0.2-0.3 S/cm. Some of the conductivity is however regainedby reoxidationin an aqueoussolutionof Fe(I)tosylate ending at ∼15 S/cm. The actual loss in film conductance is

(Parte 1 de 3)