Electrocatalytic Deposition of Polypyrrole** By Manfred Oberst and Fritz Beck*

Ever since it was demonstrated that durable films are produced on electropolymerization of pyrrole in aqueous"' and non-aqueous[*l electrolytes, polypyrrole has been the subject of intensive study. In the past few years particular attention has been paid to reversible discharging and re- charging of primarily co-deposited anions in view of a pos- sible application as battery electrodes or in electrochromic displays (cf. Refs. [3, 41). Other possible applications, e.g. in corrosion protection are being studied. The continuous preparation of polymer films by electrochemical deposi- tion on slowly rotating anode cylinders has been de- scribed.lS1 For the mechanism of the electrochemical depo- sition the primary formation of a radical cation is assumed as a direct electrode process [Reaction (a)].

pyrrole - pyrroleoG + eQ (a)

As second reaction step there follows a C-C coupling in the 2-position, either between two radical cations or be- tween a radical cation and a neutral m o l e ~ u l e . ~ ~ ~ ' ~ Since these dimers can (at less positive potentials) be readily fur- ther oxidized, the oxidative synthesis of the polymer pro- ceeds according to a scheme 2 + 2, 2 + 4, 4 + 4, etc. When the solubility limit of the oligomers is reached a solid phase forms.

Our voltammetric measurements for the electrodeposi- tion of polypyrrole at a rotating platinum disk electrode reveal a primary step (Fig. la), which is partially kineti- cally determined, even when using freshly distilled, highly pure pyrrole. Polypyrrole is deposited both in the region of the prewave as well as in the positive main branch. The prewave is also observed with Bu4NBF4 as supporting electrolyte. It is independent of the pyrrole concentration down to small values (0.01 M). From these findings we conclude that the pyrrole is reacting in the adsorbed state. A polypyrrole layer behaves thereby like platinum.

We have now surprisingly found that a bromine-cata- lyzed electrodeposition of polypyrrole is also possible. The

0.4 08 1.2 0.4 0.8 1.2 0 4 0.8 1.2

USSCE~"] - Fig. I . Current-voltage curves at the rotating R-disk electrode (A = 0.5 cmz at 200 rpm and a voltage scan rate of u , = 2 mV s - ' . Solvent: anhydrous aceto- nitrile; 20°C). a) Bromide-free system, 0.1 M pyrrole, 0.05 M NaBF,; b) 0.1 M

pyrrole, 0.1 M Bu4NBF4, 1 mM Bu,NBr; c) pyrrole-free system. Potentials "U\sct" versus the calomel electrode with NaCI,,$,, which is 236 mV more positive than the standard hydrogen electrode.

[*I Prof. Dr. F. Beck, Dip1.-Chem. M. Oberst Fachbereich 6- Elektrochemie der Universitat-Gesamthochschule Lotharstr. I , D-4100 Duisburg I (FRG)

[**I This work was supported by the Arbeitsgemeinschaft Industrieller For- schungsvereinigungen (AIF) and the Nordrhein-Westfalische Minister- ium fur Wissenschaft und Forschung.

current-voltage curve in the presence of 1 mM Br' is repro- duced in Figure lb. Once again a prewave can be recog- nized, but at still more negative potentials. In the course of this primary step no film formation is observed. The deep- colored polypyrrole is already clearly recognizable a t a layer thickness of 100 nm. During this prewave a polypyr- role film of about 0.5 pm thickness could have been formed. Polypyrrole first grows in the steeply rising part, which coincidentally occurs at the same potential as the prewave in Figure la. As experiments on large platinum sheets have shown, the current efficiency is almost 100% (referred to the theoretical value of 2.4 F/mol). The cur- rent-voltage curve in the absence of pyrrole, i.e. with l mM BrQ in acetonitrile (0.1 M Bu,NBF4 as supporting electro- lyte) shows two steps in the ratio 2 : 1 (cf. Fig. lc). Consis- tent with earlier measurements in a c e t ~ n i t r i l e [ * ~ ~ ~ we assign the first step, which appears at the same potential as the prewave in Figure Ib, to the formation of the tribromine anion [Reaction (b)].

3 BrQ - [Br3Ie + 2ee (b)

Only in the course of the second step is free bromine formed [Reaction (c)].

The free bromine, here formulated as tribromine com- plex can oxidize pyrrole radical cations according to [Reaction (d)].

pyrrole + Br3 -!A pyrroleO' + [Br?]' (4

The anion is reoxidized at the electrode. The homogene- ous reaction (d) takes place in a reaction boundary layer in front of the electrode. At higher pyrrole concentrations it is quasi first order. This means that the thickness of the reac- tion layer can be calculated using the simple relationship

p = @ E (e)

where D = diffusion coefficient and k = first order chemical rate constant; p concomitantly determines the rate of back diffusion of the [Br31Q to the electrode. That is, with the assumption of a rapid electrochemical partial step, there is a current amplification proportional to fi and to the con- centration of pyrrole.

When a dilute solution of bromine in acetonitrile is mixed with a solution of pyrrole in acetonitrile a sudden decoloration is observed, followed by a slow coloration to dark-green and subsequent separation of a black precipi- tate of pyrrole black; its formation with chemical oxidizing agents has been investigated by Gardini."']

Experiments with the rotating ring-disk electrode reveal the presence of reducible oligomeric radical cations, but no Brz. This means that reaction (d) is rapid. Our mecha- nism is also supported by the finding that, in the galvano- static deposition progressively less substance is deposited on the rotating disk with increasing speed of rotation. Since the height of the two steps in Figure Ic and that of the primary step in Figure l b increases proportionally to @ (o= angular velocity), and thus (according to Leoich) there is a limiting diffusion, it follows that an ever greater increasing percentage of the current is involved in the in- effective reaction (b) with increasing w.

A homogeneous redox catalysis by electrolytically gen- erated bromine is also observed in other electroorganic

Angew. Chem. In ( . Ed. Engl. 26 (1987) No. 10 0 VCH Verlagsgesellschnf! mbH, 0-6940 Weinheim. 1987 0570-0833/87/1010-1031 $ 02.50/0 1031

reactions, e.g. in the anodic methoxylation of furan"'] and the anodic oxidation of propylene to propylene oxide.1121 Even pyrrole has already been electropolymerized in aque- ous HBr1'31 and NaBr solution.['41 In this case, however, the polymer contains considerable amounts of bromine (up to 0.8 mol Br per pyrrole unit), which may in part be coval- ently bound. We use bromide in only small concentration, and the polymeric product contains practically no bro- mine; that is, bromide is a real catalyst. Bidan et al. have recently reported on the polymerization of 2-bromopyrrole to polypyrrole with cleavage of HBr.["' Under our condi- tions, the intermediary formation of a bromine-substitu- tion product can, however, be ruled out on kinetic grounds.

The practical significance of the electroanalytical depo- sition of polypyrrole lies in the realization of higher cur- rent densities. Whereas small current densities (below 1 mA cm-'), which obviously lie in the region of the pri- mary step in Figure la, are usually a de- position is possible at essentially higher current densities in the presence of small bromide concentrations. In this way the scan times can be substantially decreased.

Received: May 27, 1987 [Z 2263 1Ej German version: Angew. Chem. 99 (1987) 1061

CAS Registry numbers: [Br,]'. 14522-80.6; Bro, 24959-67-9; pyrrole, 109-97-7; pyrrole black, 62449- 68-7; polypyrrole, 30604-81-0; pyrrole radical cation, 34468-30-9.

[ I ] A. Dall'Olio, Y. Dascota, V. Varacca, V. Bocchi, C. R . Acad. Sci. Ser. C 267 (1968) 433.

[2] A. F. Diaz, K. K. Kanazawa, G. P. Gardini, J. Chem. SOC. Chem. Com- mun. 1979, 635; cf. A. Stanienda, Z. Nafutjorsch. B22 (1967) 1107.

[3] G. Wegner, Angew. Chem. 93 (1981) 352; Angew. Chem. Int. Ed. Engl. 20 (1981) 361.

[4] T. A. Skotheim (Ed.): Handbook ofConducting Polymers. Vol. 2, Marcel Dekker, New York 1986.

[5] H. Naarmann, Makromol. Chem. Macromol. Symp. 8 (1987) 1. 161 E. M. Genies, G. Bidan, A. F. Diaz, J . Electroanal. Chem. 149 (1983)

[7] J. Heinze, K. Hinkelmann, M. Dietrich, J. Mortensen, DECHEMA-Mon-

[8] M. Michlmayr, D. T. Sawyer, J. Electroanal. Chem. 23 (1969) 387. [9] T. Iwasita, M. C. Giordano, Electrochim. Acfn 14 (1969) 1045.

101.

ogr. 102 (1986) 209.

[lo] G. P. Gardini, Adu. Heferocycl. Chem. 15 (1973) 67. [ I 11 N. Clauson-Kaas, F. Limborg, K. Glens, Acfa Chem. Scand. 6 (1952)

[I21 F. Beck, Pure Appl. Chem. 5 (1974) 111. [I31 G. Mengoli, M. M. Musiani, M. Fleischrnann, D. Pletcher, J. Appl. Elec-

[I41 S. Tokito, T. Tsutsui, S. Saito, Chem. Lerf. 1985, 531. [ I S ] P. Audebert, G. Bidan, Synth. Met. 15 (1986) 9. 1161 A. F. Diaz, K. K. Kanazawa, J. 1. Castillo, J. A. Logan in R. B. Seymour

(Ed.): Conducfiue Polymers, Plenum Press, New York 1981, p. 149.

531.

rrochem. 14 (1984) 285.

cis- and trans-1,2-Dilithioethylene** By Adalbert Maercker,* Thornas Graule, and Wolfgang Demuth Dedicated to Professor Ulrich Schollkopf on the occasion of his 60th birthday

We recently found that vicina! dilithioalkenes are acces- sible by addition of lithium to alkynes."] When the reac- tion is carried out with open-chain alkynes in diethyl ether, insoluble trans-products are formed, whereas cyclooctyne

[*] Prof. Dr. A. Maercker, DipLChem. T. Graule, Dr. W. Demuth lnstitut fur Organische Chemie der Universitat Adolf-Reichwein-Strasse, D-5900 Siegen (FRG)

the Fonds der Chemischen 1ndustrie.-Part 5 : [I]. [**I Polylithium Organic Compounds, Part 6. This work was supported by

affords the soluble cis-l,2-dilithiocyclooctene. All attempts to prepare the 'parent compounds' cis- and frans- I ,2-dili- thioethylene have so far failed, even though at least the trans-compound has been estimated calculationally to have a higher stability than methyl- and vinyllithium.121 In- consistent with this proposal was the finding that only one stannyl moiety in the fruns-1,2-bis(tributylstannyl)ethylene 1 (R=n-C,H9) can be replaced by lithium upon reaction with n-butyl l i thi~m.~~] We are therefore of the opinion that even n-butyllithium is more stable than trans-1,2-dilithio- ethylene 6.l4]

Switching over to tert-butyllithium for the R,Sn/Li ex- change in 1 had little hope of success, however, since the ate-complex intermediate with four terf-butyl groups at- tached to the five-coordinate tin necessary for the second exchange would be sterically too demanding. Thus, e.g., even in the case of tetramethylstannane in this way maxi- mally two methyl groups can be replaced by fert-butyl resi- dues.['] We have therefore replaced the four-coordinate tin by two-coordinate mercury prior to the reaction with teri- butyllithium and thus indeed have been able to achieve a double metal-metal exchange. The compounds 6 and 7 al- ready formed at -75°C in diethyl ether or T H F were der- ivatized with dimethyl sulfate, which had the advantage that the gaseous reaction products trans- 9 and cis-2-bu- tene 10 could be driven out of the reaction mixture and quantitatively determined as meso- 8 and d,/-2,3-dibromo- butane 11, respectively, after passage into a solution of bromine in chloroform. The analysis was accomplished gas chromatographically as well as by combined G U M S .

One could argue that the derivatization might also have proceeded stepwise, i.e. that the second Hg/Li exchange has taken place only after the introduction of the first me- thyl group. This argument can easily be invalidated: It could be shown that the metal-metal exchange in the pres- ence of dimethyl sulfate functions very poorly, since ferf- butyllithium reacts much more rapidly with dimethyl sul- fate than with the mercury compounds.I6' The finding that the amount of di-tert-butylmercury in the reaction mixture before and after the derivatization is the same, also speaks against a mercury-lithium exchange during the addition of the dimethyl sulfate. Crystallization of the 1,2-dilithioethyl- enes and thus the separation from the di-tert-butylmercury has so far not been accomplished. No statements can be made therefore about the exact structure and degree of as- sociation of the compounds.

Interestingly, it was observed that reaction of cis- 1,2- bis(ch1oromercurio)ethylene 3 containing 7% 2 at - 75 "C does indeed afford the 2-butenes 10 and 9 in the ratio 93 :7, but, after 10 min at 0"C, the amount of trans product increases to 65%. This first suggested a rapid rearrange- ment of the cis-1,2-dilithioethylene 7 into the trans-isomer 6.171 In the meantime it has emerged that this cis-trans rear- rangement 7-6 is only simulated due to a very much higher kinetic stability of 6 in comparison to 7. Namely, cis- 1,2-dilithioethylene 7 decomposes particularly rapidly, splitting off lithium hydrideJ8I and, as expected, the li- thioacetylene 12 formed thereby is immediately metalated to dilithioacetylene 14. Not only is excess tert-butyllithium suitable as metalating reagent, but also the dilithioethyl- enes themselves. The resulting vinyllithium 13, like the di- lithioacetylene 14, was derivatized by dimethyl sulfate, so that finally the bromination products of propene 15 and 2-butyne 16 could be detected gas chromatographically in addition to 8 and 11. The methane formed by reaction of lithium hydride with dimethyl sulfate was detected mass spectrometricall y.

1032 0 VCH Verlagsgesellschafc mbH, 0.6940 Weinheim. 1987 0570-0833/87/1010-1032 $ 02.5010 Angew. Chem. I n f . Ed. Engl. 26 (1987) No. 10