Poly(isothianaphthene) from 2,5-bis(trialkylsilyl )isothianaphthenes: preparation and spectroscopic characterization M. Lapkowski,a,b R. Kiebooms,c J. Gelan,c D. Vanderzande,c A. Pron,d,e T. P. Nguyen,f G. Louarnf and S. Lefrantf aDepartment of Chemistry, Silesian T echnical University, 44 100 Gliwice, Poland bDepartment of T extile Engineering and Environmental Sciences, T echnical University of £odz, Bielsko-Biala Campus, 43 300 Bielsko-Biala, Plac Fabryczny 1, Poland cL imburg University, Instituut voorMateriaalonderzoek (IMO), Department BSG, Universitaire Campus, B-3590 Diepenbeek, Belgium dDepartment ofMaterials Science and Ceramics, Academy of Mining and Metallurgy, 30 059 Krako�w, Mickiewicza 30, Poland eDepartment of Chemistry, T echnical University of Warsaw, 00 664 Warszawa, Noakowskiego 3, Poland fL aboratoire de Physique Cristalline, IMN, Universite� de Nantes, UMR 110, 2, rue de la Houssinie`re, 44072 Nantes Ce�dex 03, France A new method for the preparation of poly(isothianaphthene) is proposed, namely electropolymerisation of bis(tertbutyldimethylsilyl) isothianaphthene (BTBDMS)ITN.The advantage oered by this method is based on the fact that (BTBDMS)ITN is a stable monomer at ambient conditions whereas unsubstituted isothianaphthene is unstable and must be prepared prior to the polymerisation. Poly(isothianaphthene) (PITN) prepared from (BTBDMS)ITN shows an average polymerisation degree of 20 and spectral and spectroelectrochemical characteristics similar to classical PITN.FT Raman spectroelectrochemical studies show that during oxidative doping PITN undergoes similar changes to poly(alkylthiophenes) and poly(alkoxythiophenes).In modern molecular electronics significant research eort is which, in contrast to isothianaphthene, is a stable monomer directed towards the preparation of low band-gap conjugated and can be stored for extended times in a bottle in laboratory polymers due to several possible industrial applications of such conditions.In addition to the description of the electrochemical systems. Poly(isothianaphthene) (PITN) initially synthesized behaviour of PITN prepared from (BTBDMS)ITN we characby Wudl et al.1 exhibits the lowest band-gap of all polyconju- terize the obtained polymer by UV–VIS–NIR, XPS and Raman gated systems studied to date.Both PITN and its C6 ring spectroscopies. substituted derivatives can be prepared by electrochemical oxidative polymerisation of isothianaphthene or its deriva- Experimental tives.1–5 However the monomers used for these syntheses are rather unstable at ambient conditions and must be freshly For the synthesis of the monomer 2,5(tert-butyldimethylsilyl)- prepared prior to the polymerisation.In order to avoid the isothianaphthene the procedure described by Okuda et al. was inconveniences associated with electropolymerisation, chemical used.13 Electropolymerisation was carried out in a reaction oxidation procedures have been suggested involving the medium consisting of NBu4BF4 and (BTBDMS)ITN dissolved oxidation of dihydroisothianaphthene with FeCl3, O2,6 or in nitrobenzene.The concentrations of the electrolytic salt and N-chlorosuccinimide.7 the monomer were 0.2 and 0.1 M respectively. The reaction PITN can also be prepared by dehydrogenation of the PITN was performed in a three-electrode electrochemical cell with a precursor, namely poly(dihydroisothianapthene) (PDHITN) platinum counter electrode and Ag/AgCl wire as the reference with such dehydrogenation agents as SO2Cl2,8 or tert-butyl electrode.Three types of working electrodes were used, plati- hypochlorite.9 This last method is especially interesting because num, gold and ITO, dependingon the subsequent spectroscopic dehydrogenation of PDHITN in solution leads to a stable studies of the deposited polymer film. The polymerisation was PITN solution.Thus films of PITN can be prepared on an performed at a constant potential of 1.4 V vs. Ag/AgCl. appropriate substrate by casting. Here, we propose a new In addition to the deposition of the polymer on the working method for the preparation of PITN via electropolymerisation electrode, the formation of soluble oligomers took place which of disilyl derivatives of ITN. Silicon directed reactions have was manifested by a change of the colour of the electrolyte been widely used in organic chemistry with the goal of improv- solution in the vicinity of the working electrode.The deposition ing the selectivity of carbon–carbon bond formation.10 Recently of a homogeneous film required electrolysis times exceeding they have been applied to the preparation of polythiophene11 5 min.and poly(3-alkylthiophene).12 In order to prepare PITN we The electropolymerised film was then reduced electrochemi- have used 2,5-bis(tert-butyldimethylsilyl)isothianaphthene cally at E=-0.6 V vs. Ag/AgCl to give the neutral polymer (BTBDMS)ITN: of a distinct blue colour. The electrode with the film deposited on it was then carefully rinsed with pure nitrobenzene and then with acetonitrile in order to remove all oligomeric species.All operation were carried out in a dry nitrogen atmosphere. Cyclic voltammetry Cyclic voltammograms were recorded in a monomer free 0.2 M NBu4BF4 solution in acetonitrile using a PAR 273 potentiostat/ J. Mater. Chem., 1997, 7(6), 873–876 873galvanostat. The working electrode consisted of a layer of for the polymerisation reaction.The polymerisation carried out after 2 months using the same batch of ITN gave identical poly(isothianaphthene) deposited on a platinum foil. As in the electropolymerisation experiments a Pt counter electrode and polymer. In the neutral (i.e. reduced at E=-0.6 V) PITN film, XPS spectroscopy shows the presence of carbon, sulfur an Ag/AgCl reference electrode were used. and a small amount of oxygen in addition to the already mentioned silicon. The C 1s peak is located at 284.7 eV with UV–VIS–NIR spectroelectrochemistry a skewed arc at the high energy side and a full width at half UV–VIS–NIR spectroelectrochemical studies were performed maximum of 2 eV.The S 2p line shows a peak at 163.5 eV in the same electrolyte solutions as in the cyclic voltammetry with a widened base at low binding energy.These features are studies, using the same counter and reference electrodes. similar to those observed for polythiophene except that some Poly(isothianaphthene) was deposited on an indium tin oxide carbon atoms are aected by the silyl end groups.14 transparent electrode (ITO). Cyclic voltammograms of PITN prepared from (BTBDMS)ITN are shown in Fig. 1. We restricted ourselves FT Raman spectroelectrochemistry to oxidative (p-type) doping although there are literature reports of p- and n-type doping of PITN.15,16 The Raman spectra were recorded with a near-IR excitation Oxidative doping of PITN synthesized from the disilyl line (1064 nm) on a FT-Raman Bruker RFS spectrometer derivative of ITN gives rise to two broad strongly overlapping working in a back scattering geometry.For Raman studies the oxidation peaks (due to anion doping) and two reduction film of PITN was deposited on a Pt electrode. The same peaks associated with undoping. The shape of the current– electrolyte, counter and reference electrodes as in cyclic voltamvoltage curves is essentially the same as that reported by metry were used.Higgins et al.17 for poly(benzo[c]thiophene) cycled in NBu4X–acetonitrile electrolyte. XPS studies In the case of PITN the shape of the current–voltage curve The XPS measurements were performed on a Leybold LH 12 depends on the history of the sample and, more precisely, on analyser (CNRS Universite� de Nantes) using an Mg-Ka X-ray whether the scan range has been extended to n-doping prior source in a UHV system.The pressure of the chamber was to p-doping. In such case a pre-peak is observed.15 Onada kept in the 10-9 mbar range during experiments. The polymer et al.16 have ascribed the existence of these additional peak to film were deposited on ITO substrates and no charging eect the diculty in the removal of all negative charge upon due the beam radiation was observed.The binding energy was oxidapreviously n-doped polymer. The release of this referenced to the Au 4f7/2 line (84 eV) from a gold probe residual charge gives rise to this additional ‘pre-oxidation evaporation on the surface of the sample holder. The collected peak’. We do not observe this peak, in agreement with other data were treated by a computer program with satellite back- authors,16,17 in the experiments where PITN was not previously ground subtraction.Semi-quantitative determination of the n-doped. composition of the analysed surface was performed from It should be stressed here than our PITN has a very good the obtained spectra taking account of the sensitivity of the cycling stability. In addition the oxidative doping of our PITN elements present.seems to be more reversible than that reported in ref. 18. UV–VIS–NIR spectra registered for increasing electrode potentials are shown in Fig. 2. In the reduced polymer (E= Results and Discussion -400 mV) the dominant peak ascribed to the p–p* transition (BTBDMS)ITN readily polymerises electrochemically to give in the conjugated backbone shows a maximum at 690–700 nm, PITN.However, as has been stated before, the preparation of i.e. it is blue shifted by ca. 50 nm as compared to the PITN homogeneous films on the electrode requires higher concen- spectrum reported in ref. 15, and by ca. 90–100 nm in compari- trations of the silylated monomer as compared to unsubstituted son to the spectra recorded in refs. 9 and 16.Since the position isothianaphthene. In addition longer electrodeposition times of the p–p* transition band can be taken as a measure of are required. This may imply that in the case of (BTBTMS)ITN conjugation we conclude that PITN obtained from the electropolymerisation mechanism is significantly dierent (BTBDMS)ITN is less conjugated. This eect may be associ- than in the case of unsubstituted ITN.There is a low electro- ated with the influence of the silyl end groups taking into polymerisation yield and the presence of the oligomeric species in the vicinity of the electrode is only one of several electrooxidation products. Studies of the polymerisation mechanism are in progress. Due to the presence of the leaving silyl groups in the monomer the polymerisation degree in the polymer can be conveniently determined by end group analysis, and more precisely from the analytically determined S/Si molar ratio.XPS studies of the poly(isothianaphthene) film deposited on the ITO electrode show that this ratio is close to 10 which gives an average polymerisation degree equal to 20. Thus the trialkylsilyl groups are eciently eliminated during the electrochemical oxidation according to Scheme 1 It should be stressed that, in contrast to ITN, its silylated derivatives can be stored for extended times before their use Fig. 1 Cyclic voltammograms of PITN recorded in 0.1 mol dm-3 Scheme 1 (n#20) NBu4BF4 solution in acetonitrile 874 J. Mater. Chem., 1997, 7(6), 873–876Fig. 2 In situ UV–VIS absorbance curves of PITN recorded for selected electrode potentials (vs.Ag/AgCl): (a) -400; (b) 200; (c) 300; (d) 400; (e) 500; (f) 600; (g) 900; (h) 1000 mV account that the polymerisation degree in our polymer is rather low (average polymerisation degree of ca. 20). In conjugated polymers oxidative doping is usually manifested by bleaching of the p–p* transition peak with simultaneous growth of a peak (peaks) in the near-IR part of the spectrum.Qualitatively the same behaviour is observed for the PITN studied in this research. HoweverUV–VIS–NIR spectroelectrochemistry gives evidence of two distinctly dierent doping stages. Up to potentials E=0.6 V, which is in the vicinity of the maximum of the oxidative doping peak, two isosbestic points are observed at 365 and 430 nm.This means that only two optically dierent phases are present in the Fig. 3 Raman spectra with lex=1064 nm of PITN recorded during system. At E=0.6 V the polymer is only partially doped as the oxidative process (potentials vs. Ag/AgCl): (a) -300; (b) 300; (c) 600; (d) 1000 mV judged from a rather high absorbance at 690–700 nm. Complete doping manifested by total bleaching of the p–p* peak requires electrode polarization at potentials exceeding 0.9 V.In the second oxidation step no isosbestic points are even if minute amounts of dopant are present. A recently published PITN spectrum20 is very similar to that reported observed. It seems therefore plausible that during the first oxidation step polarons (radical cations) are formed which here; however, it contains one additional peak between the modes at 1301 and 1166 cm-1 which is absent in our spectrum.then recombine to bipolarons (dications) at higher potentials. Such a transformation clearly explains the existence of the Careful spectroscopic analysis of PITN and model compounds by Raman21 and NMR21,22 spectroscopies seems to indicate isosbestic points at the beginning of the oxidation and their absence at the end of the oxidative doping.that in the neutral state this polymer adopts the quinonoid sequence of bonds. Such a conclusion is also supported by This conclusion is also supported by the ellipsometric and FTIR results of Christiansen et al.19 who also claimed two- theoretical calculations.23 However, as has been pointed out by Kiebooms,24 the step oxidativedoping of PITN.However additional verification by joint cyclic voltammetry and EPR studies is required; such energy dierence between the quinonoid and aromatic states is very small (2.4 kcal mol-1, 1 cal=4.184 J) so the structure studies are in progress. We have also undertaken Raman spectroelectrochemical studies of PITN. In general Raman of the neutral polymer may be influenced by the nature of the end groups in the monomer used for the synthesis.Since studies of polyconjugated systems are complicated by resonance eects which result in a strong dependence of the Raman (BTBDMS)ITN monomer should favour the aromatic sequence of bonds in the resulting polymer it is highly probable line intensities (and sometimes their shape and positions) on the excitation wavelength energy.In addition the resonance that in PITN prepared from the disilyl derivative the aromatic structure is adopted in the neutral state; thus the doping conditions are altered significantly in the course of the doping reaction due to the doping induced changes in the electronic should lead to the quinonoid structure. This hypothesis is based on the close similarity of the Raman spectroelectro- spectra of conjugated polymers.We have selected the near-IR excitation line (lex=1064 nm) chemical behaviour of PITN and regioregular poly(3-alkylthiophenes) 25 and poly(dialkoxybithiophenes).26 In the last mainly because its position is very close to the maximum of the doping induced electronic transition. In addition at this two cases oxidative doping causes the transformation of the aromatic structure into the quinonoid one.wavelength we are also within the p–p* transition peak of the neutral form of PITN and asmall but not negligible absorbance As probed by Raman spectroscopy, doping induced spectral changes start at potentials E>0 V vs. Ag/AgCl. However is registered at 1064 nm for this compound. FT Raman spectra registered for increasing electrode poten- between 0.1 and 0.6 V the spectra are essentially the same showing the features of the doped sample.The main changes tials are shown in Fig. 3. The Raman spectrum of neutral (E= -0.3 V) PITN prepared from the disilyl derivative is dierent occurring upon doping can be characterized as follows. (i) The band at 1462 cm-1 in neutral PITN ascribed to CaMCb from that reported by Hoogmartens et al.18 In fact, the spectrum published in ref. 18 is essentially identical to our stretchings shifts to lower wavenumber (1417 cm-1) and decreases in intensity with respect to other bands. Identical spectrum of slightly doped PITN, i.e. that recorded at E= 0.3 V. Taking into account that lex=1064 nm almost matches behaviour has been observed for poly(3-alkylthiophenes)23 and poly(alkoxythiophenes).24 (ii) The band at 1440 cm-1 in the the maximum of the electronic absorption in the doped polymer, strong resonant enhancement of vibrations characteristic neutral PITN ascribed to the vibrations of the benzene ring condensed with the thiophene ring is essentially unaected by of the doped structure is expected.It is therefore highly probable that the features of the doped polymer may appear the doping. (iii) A band at 1195 cm-1 appears which can be J. Mater. Chem., 1997, 7(6), 873–876 8753 G. King, S. J. Higgins, S. E. Garner and A. R. Hillman, Synth.Met., interpreted as originating from the CaMCa¾ inter-ring 1994, 67, 241. vibrations of the quinonoid structure. 4 G. King and S.J. Higgins, J. Chem. Soc., Chem. Commun., 1994, At E>0.6 V further spectral changes occur which support 825. the hypothesis of the quinonoid structure formation upon 5 G. King and S. J. Higgins, J. Mater. Chem., 1995, 5, 447. doping. In particular the peak at 1301 cm-1 in the neutral 6 K. Jen and Elsenbaumer, Synth. Met., 1986, 16, 379. 7 I. Hoogmartens, D. Vanderzande, H.Martens and J. Gelan, Synth. polymer and ascribed to CbMCb¾ stretching broadens and Met., 1992, 47, 367. shifts to slightly higher wavenumber whereas the peak due to 8 T. L. Rose and M. C. Liberto, Synth.Met., 1989, 31, 395. CaMCb stretchings continues to shift towards lower wave- 9 S. A. Chen and C. C. Lee, Synth.Met., 1995, 75, 187. lengths (from 1462 cm-1 in the neutral polymer to 1400 cm-1 10 E.W. Colvin, in Silicon in Organic Synthesis, Butterworth, for the quinonoid CaMCa¾ sequence of bonds). The inter-ring Guildford, 1981. 11 J. L. Sauvajol, C. Chorro, J. P. Le`re-Porte, R. J. P. Corriu, J. J. E. stretching peak continues to grow. Moreau, P. The�pot and M. W. C. Man, Synth.Met., 1994, 62, 233. It should be stressed once more that these changes are 12 M.Bouachrine, J. P. Le`re-Porte, J. J. E. Moreau and M. W. C. qualitatively the same as those observed for poly(3-alkylthio- Man, J. Mater. Chem., 1995, 5, 797. phenes)25 and poly(alkoxythiophenes).26 The comparison with 13 Y. Okuda, M. V. Laksmikantham and M. P. Cava, J. Org. Chem., poly(3,3¾-dibutoxy-2,2¾-bithiophene) should be instructive. For 1991, 56, 6024. 14 G. Morea, C.Malitesta, L. Sabbatini and P. G. Zambonin, both polymers the near-IR excitation line (lex=1064 nm) is J. Chem. Soc., Faraday T rans., 1990, 86, 3769. located in the vicinity of the doping induced absorption 15 S. M. Dale, A. Glidle and A. R. Hillman, J. Mater. Chem., 1992, maximum and both polymers undergo very similar changes in 2, 99. the Raman spectra upon electrochemical doping. 16 M. Onoda, H. Nakayama, S. Morita and K. Yoshino, J. Electrochem. Soc., 1994, 141, 338. 17 S. J. Higgins, C. Jones, G. King, K. H. D. Slack and S. Petidy, Synth. Met., 1996, 78, 155. Conclusions 18 I. Hoogmartens, P. Adriaensen, R. Carleer, D. Vanderzande, M. Martens and J. Gelan, Synth. Met., 1992, 51, 219. We have demonstrated that electropolymerisation of bis(tert- 19 P. A.Christiansen, J. C. H. Kerr, S. J. Higgins and A. Hamnett, butyldimethyl)silylisothianaphthene (BTBDMS)ITN leads to Faraday Discuss. Chem. Soc., 1989, 88, 261. poly(isothianaphthene) with an average polymerisation degree 20 G. Zerbi, M. C. Magnoni, I. Hoogmartens, R. Kiebooms, R. Carleer, D. Vanderzande and J. Gelan, Adv. Mater., 1995, 7, of 20. The obtained polymer shows spectral features similar 1027. but not identical to poly(isothianaphthene) prepared by classi- 21 I. Hoogmartens, P. Adriaensen, D. Vanderzande, J. Gelan, cal methods. The main advantage of the procedure proposed C. Quattrocchi, R. Lazzaroni and J. L. Bredas, Macromolecules, here relies on the fact that (BTBDMS)ITN is a stable monomer 1992, 25, 7347. and can be stored for extended times whereas isothianaphthene 22 I. Hoogmartens, P. Adriaensen, D. Vanderzande and J. Gelan, Anal. Chim. Acta, 1993, 283, 1025. must be synthesized prior to the electropolymerisation. 23 L. Cu, M. Kertesz, J. Geisselbrecht, J. Ku�rti and H. Kuzmany, Synth. Met., 1993, 55, 564. 24 R. Kiebooms, PhD Thesis, University of Limburgs, 1995. 25 M. Trznadel, M. Zago�rska, M. Lapkowski, G. Louarn, S. Lefrant References and A. Pron, J. Chem. Soc., Faraday T rans., 1996, 92, 1387. 26 A. Pron, G. Louarn, M. Lapkowski, M. Zago�rska, I. Glo�wczyk- 1 F. Wudl, M. Kobayashi and A. J. Heeger, J. Org. Chem., 1984, Zubek and S. Lefrant, Macromolecules, 1995, 28, 4644. 49, 3382. 2 H. Yashima, M. Kobayashi, K. B. Lee, D. Chung, A. J. Heeger and F.Wudl, J. Electrochem. Soc., 1987, 134, 46. Paper 6/06868E; Received 7th October, 1996 876 J. Mater. Chem., 1997, 7(6), 873